abeta peptide and alzheimer's disease - c. barrow, d. small (springer, 2007) ww

8/15/2019 Abeta Peptide and Alzheimer's Disease - C. Barrow, D. Small (Springer, 2007) WW

http://slidepdf.com/reader/full/abeta-peptide-and-alzheimers-disease-c-barrow-d-small-springer 1/308



vivo and in AD brain, consistent with the notionthat A β(1-42) significantly contributes to theoxidative stress of AD brain.

5.3 Methionine-35 of A β(1-42):Role in A β(1-42)-InducedOxidative Stress and Neurotoxicity

Methionine 35 is a critical residue in A β(1-42)-mediated oxidative stress and neurotoxicity.Substitution of the sulfur atom of methionine 35 bya methylene group, –CH 2– (norleucine), signifi-cantly modulates the oxidative stress and neurotox-icity of A β(1-42), but the fibrilar morphology of

both peptides is similar [10]. Methionine 35 of Aβ(1-42) is also involved in the oxidative stress andneurotoxicity properties of this peptide in vivo . C.elegans expressing human A β(1-42) exhibited sig-nificantly increased protein oxidation, but replace-ment of the codon for Met by that for Cys in theDNA sequence for human A β(1-42) resulted in noincrease in protein oxidation in the worm comparedwith C. elegans expressing native human A β(1-42)[10]. Additionally, studies involving a temperatureinducible C. elegans model expressing humanAβ(1-42) revealed that protein oxidation preceedsthe deposition of fibrilar aggregates [8]. This find-ing is consistent with increasing evidence that smallsoluble aggregates of A β(1-42) are the toxic speciesof this peptide [66–68]. Moreover, that A β(1-42)containing the norleucine deriviative of A β(1-42),which through producing fibrils, was not oxidativeor neurotoxic supports our hypothesis that methion-ine is critically involved in the neurotoxic andoxidative properties of A β(1-42) [10, 69].

Lipid peroxidation is induced by A β(1-42) [15,27] and is found in AD brain [15, 19, 20]. Becauselipid peroxidation requires that the free radicalinvolved must be located in the immediate vicinityof the labile H-atoms of unsaturated acyl-chains onphospholipids, this requirement suggests that theMet residue of A β(1-42) is located in the bilayer[70], a suggestion confirmed by others [71]. It hasbeen proposed that, due to the hydrophobic car-boxy terminus of A β(1-42), the peptide inserts intothe lipid bilayer [70–72]. A β(1-42) adopts an α -helical conformation, similar to other proteins thatinsert into the lipid bilayer. A methionine sulfu-

ranyl radical (MetS .) on A β(1-42) is formed by aone-electron oxidation [12–14, 69, 72–75]. Thisradical, in turn, can abstract a hydrogen atom froma neighboring unsaturated lipid resulting in the for-mation of a carbon-centered lipid radical (L .). Viamechanisms described above (Fig. 5.1), the car-bon-centered radical on the lipid can readily reactwith molecular oxygen to form a peroxyl radical(LOO .). Hydrogen abstraction from a neighboringlipid results in the formation of a lipid hydroperox-ide (LOOH) and another carbon-centered lipid rad-ical (L .), thereby, propagating the free-radical chainreaction [69, 74, 75]. Both theoretical and experi-mental studies demonstrate that the α -helical sec-ondary structure of the peptide providesstabilization of the sulfuranyl radical formed by a

one-electron oxidation of methionine [72, 76].Mutation of isoleucine 31 in A β(1-42) to proline,an α-helix breaker, attenuated the oxidative stressand neurotoxic properties of the native peptide,suggesting that the amide oxygen of isoleucine 31in the α -helix conformation interacts with a lonepair of electrons on the sulfur atom of methionine35, priming this atom for a one-electron oxidation[72]. Subsequently, the sulfuranyl radical of methionine can react with other moieties of methionine to form an

α(alkylthio)alkyl radical of

methionine (–CH 2-CH 2-S-CH 2 or –CH 2-CH-S-CH 3) [69, 72, 74, 76]. Such carbon-centered radi-cals provide potential substrates for reaction withmolecular oxygen leading to the formation of per-oxyl radicals, and consequently, potentiation of free-radical generation and HNE formation [69, 75,77]. Recently, others have confirmed our hypothe-sis, directly demonstrating the existence of the sul-furanyl free radical in A β(1-40) [78]. Otherresearchers [79, 80] invoke Cu(II) reduction and

subsequent H 2O2 formation in the oxidative stressand neurotoxic properties of A β(1-42). Critical inthis scenario are the three His residues at positions6, 13, and 14 and the Tyr at position 10. The formerare the likely binding sites for Cu(II) on A β(1-42),while Tyr 10 is proposed to be the source of theelectron to reduce Cu(II) to Cu(I). However, sub-stitution of the three His residues by asparagine(which has at least a 100-fold less binding affinityof Cu(II) than does His) or substitution of Tyr 10

by aromatic Phe (which, though still aromatic, isincapable of providing an election to Cu(II)) leadsto peptides that are similarily toxic and oxidative as

88 D.A. Butterfield



native A β(1-42) [81, 82]. In contrast, substitutionof Met by norleucine, which still has the three Hisresidues and Tyr 10 present, is no longer toxic oroxidative [10]. Using the reverse peptide, A β(40-1), which is nontoxic, others showed that a Tyr freeradical could be formed [78]. That is, a central fea-ture required in mechanisms that involve Cu(II)reduction as a cardinal paradigm occur only in apeptide that is nontoxic [78].

Oxidative modification of methionine 35 tomethionine sulfoxide constitutes a major compo-nent of the various amyloid β-peptides isolatedfrom AD brain [83–85], consistent with the role of methionine in the oxidative properties of A β(1-42).In vitro oxidation of methionine to methionine sul-foxide has been shown to abolish the oxidative

stress and neurotoxic properties of A β(1-42) after a24-h incubation with neurons. Mitochondrial dys-function as measured by MTT reduction was alsoobserved [73]. This finding was confirmed in arecent study [80]. However, after a 96-h treatment,the methionine sulfoxide of A β(1-42) reportedlyresulted in neuronal death as observed by phasecontrast microscopy. A β(1-42) containing methion-ine sulfoxide does not associate itself with the lipidbilayer due to the hydrophilic oxidized sulfur atom[80]. It is conceivable that A β(1-42) containingmethionine sulfoxide may not form fibrils readilybut does so after a long enough period. Thus, toxic-ity of A β(1-42) containing methionine sulfoxidemay occur via a different mechanism than withnative A β(1-42), that is, fibril formation conceivablycould activate the receptor for advanced glycationend products (RAGE) leading to oxidative stressand neurotoxicity [86, 87].

5.4 ConclusionsAβ(1-42) plays a critical role in the oxidative stresspresent in AD brain and, consequently, may play acentral role in the pathogenesis of the disease.Aβ(1-42) induces protein oxidation and lipid per-oxidation both in vitro and in vivo. Methionine 35has been shown to play a vital role in the oxidativestress and neurotoxic properties of A β(1-42).Ongoing proteomic studies will lead to the identifi-cation of proteins that are specifically oxidativelymodified by A β(1-42), providing insight intomechanisms of A β(1-42)-induced neurodegenera-

tion and, consequently, a greater insight into therole that A β(1-42) plays in the pathogenesis of thisdementing disorder.

Acknowledgments This work was supported bygrants from NIH (AG-05119; AG-10836).

References

1. Katzman R and Saitoh T. Advances in Alzheimer’sdisease. FASEB J 1991; 5:278-86.

2. Selkoe DJ. Alzheimer’s disease results from the cere-bral accumulation and cytotoxicity of amyloid beta-protein. J Alzheimers Dis 2001; 3:75-80.

3. Butterfield DA and Lauderback CM. Lipid peroxida-tion and protein oxidation in Alzheimer’s disease

brain: potential causes and consequences involvingamyloid beta-peptide-associated free radical oxida-tive stress. Free Radic Biol Med 2002; 32:1050-60.

4. Butterfield DA, Drake J, Pocernich C, et al. Evidenceof oxidative damage in Alzheimer’s disease brain:central role for amyloid beta-peptide. Trends MolMed 2001; 7:548-54.

5. Hensley K, Hall N, Subramaniam R, et al. Brainregional correspondence between Alzheimer’s dis-ease histopathology and biomarkers of protein oxida-tion. J Neurochem 1995; 65:2146-56.

6. Markesbery MR. Oxidative stress hypothesis inAlzheimer’s disease. Free Radic Biol Med 1997;23:134-47.

7. Smith MA, Richey Harris PL, Sayre LM, et al.Widespread peroxynitrite-mediated damage inAlzheimer’s disease. J Neurosci 1997; 17:2653-57.

8. Drake J, Link CD, Butterfield DA. Oxidative Stressprecedes fibrillar deposition of Alzheimer’s diseaseamyloid β-peptide (1-42) in a transgenicCaenorhabditis elegans model. Neurobiol Aging2003; 24:415-20.

9. Subbarao KV, Richardson JS, Ang LC. Autopsy

samples of Alzheimer’s cortex show increased per-oxidation in vitro. J Neurochem 1990; 55:342-45.

10. Yatin SM, Varadarajan S, Link CD, et al. In vitro andin vivo oxidative stress associated with Alzheimer’samyloid β-peptide (1-42). Neurobiol Aging 1999;20:325-30.

11. Mohmmad-Abdul H, Wenk GL, Gramling M, et al.APP and PS-1 mutations induce brain oxidative stressindepend of dietary cholesterol: implications forAlzheimer’s disease. Neurosci Lett 2004; 368:148-50.

12. Butterfield DA. Amyloid beta-peptide (1-42)-

induced oxidative stress and neurotoxicity: implica-tions for neurodegeneration in Alzheimer’s diseasebrain. A review. Free Radic Res 2002; 36:1307-13.

5. Amyloid β-Peptide(1-42), Oxidative Stress, and Alzheimer’s Disease 89



13. Butterfield DA. Amyloid beta-peptide [1-42]-associated free radical-induced oxidative stress andneurodegeneration in Alzheimer’s disease brain:mechanisms and consequences. Curr Med Chem2003; 10:2651-59.

14. Varadarajan S, Yatin S, Aksenova M, et al. Review:

Alzheimer’s amyloid β-peptide-associated free radi-cal oxidative stress and neurotoxicity. J Struct Biol2000; 130:184-08.

15. Lauderback CM, Hackett JM, Huang F, et al. Theglial glutamate transporter, GLT-1, is oxidativelymodified by 4-hydroxy-2-nonenal in the Alzheimer’sdisease brain: the role of A β(1-42). J Neurochem2001; 78:413-16.

16. Yatin SM, Varadarajan S, Butterfield DA. Vitamin Eprevents Alzheimer’s amyloid β-peptide (1-42)-induced neuronal protein oxidation and reactive oxy-gen species production. J Alzheimers Dis 2000; 2:123-31.

17. Castegna A, Lauderback CM, Mohmmad-Abdul H,et al. Modulation of phospholipid asymmetry insynaptosomal membranes by the lipid peroxidationproducts, 4-hydroxynonenal and acrolein: implica-tions for Alzheimer’s disease. Brain Res 2004;1004:193-97.

18. Mohmmad Abdul H, Butterfield DA. Protectionagainst amyloid beta-peptide (1-42)-induced loss of phospholipid asymmetry in synaptosomal mem-branes by tricyclodecan-9-xanthogenate (D609) and

ferulic acid ethyl ester: implications for Alzheimer’sdisease. Biochim Biophys Acta 2005; 1741:140-148.

19. Lovell MA, Xie C, Markesbery WR. Acrolein isincreased in Alzheimer’s disease brain and is toxic toprimary hippocampal cultures. Neurobiol Aging2001; 22:187-94.

20. Markesbery WR and Lovell MA. Four-hydroxynone-nal, a product of lipid peroxidation, is increased inthe brain in Alzheimer’s disease. Neurobiol Aging1998; 19:33-36.

21. Subramaniam R, Roediger F, Jordan B, et al. The

lipid peroxidation product, 4-hydroxy-2-trans-none-nal, alters the conformation of cortical synaptosomalmembrane proteins. J Neurochem 1997; 69:1161-69.

22. LaFontaine MA, Mattson MP, Butterfield DA.Oxidative stress in synaptosomal proteins from mutantpresenilin-1 knock-in mice: implications for familialAlzheimer’s disease. Neurochem Res 2002; 27:417-21.

23. Mohmmad Abdul H, Sultana R, Keller JN, et al.Mutations in APP and PS1 genes increase the basaloxidative stress in murine neuronal cells and leadto increased sensitivity to oxidative stress mediatedby A β(1-42), H 2O2, and kainic acid: Implicationsfor Alzheimer’s disease. J Neurochem 2006; 96:1322-1335.

24. Pocernich CB and Butterfield DA. Acrolien inhibitsNADH-linked mitochondrial enzyme activity: impli-cations for Alzheimer’s disease. Neurotox Res 2003;5:515-20.

25. Gabbita SP, Lovell MA, Markesbery WR. Increasednuclear DNA oxidation in the brain in Alzheimer’s

disease. J Neurochem 1998; 71:2034-40.26. Drake J, Petroze R, Castegna A, et al. 4-

Hydroxynonenal oxidatively modifies histones:implications for Alzheimer’s disease. Neurosci Lett2004; 356:155-58.

27. Mark RJ, Lovell MA, Markesberry WR, et al. A rolefor 4-hydroxynonenal, an aldehydic product of lipidperoxidation, in disruption of ion homeostasis andneuronal death induced by amyloid β-peptide.J Neurochem 1997; 68:255-64.

28. Sultana R, Butterfield DA. Oxidatively modifiedGST and MRP1 in Alzheimer’s disease brain: impli-cations for accumulation of reactive lipid peroxida-tion products. Neurochem Res 2004; 29:2215-2220.

29. Masliah E, Alford M, De Teresa R, et al. Deficientglutamate transport is associated with neurodegener-ation in Alzheimer’s disease. Ann Neurol 1995;40:759-66.

30. Lovell MA, Xie C, Markesbery WR. Decreased glu-tathione transferase activity in brain and ventricularfluid in Alzheimer’s disease. Neurology 1998; 51:1562-1566.

31. Rosser MN, Svendsen C, Hunt SP, et al. The sub-

stantia innominata in Alzheimer’s disease: a histo-chemical and biochemical study of cholinergicmarker enzymers. Neurosci Lett 1982; 28:217-22.

32. Aksenov MY, Aksenova MV, Butterfield DA, et al.Oxidative modification of creatine kinase BB inAlzheimer’s disease brain. J Neurochem 2000;74:2520-27.

33. Butterfield DA and Stadtman ER. Protein oxidationprocesses in aging brain. Adv Cell Aging Gerontol1997; 2:161-91.

34. Castegna A, Thongboonkerd V, Klein JB, et al.

Proteomic identification of nitrated proteins inAlzheimer’s disease brain. J Neurochem 2003; 85:1394-01.

35. Good PF, Werner P, Hsu A, et al. Evidence for neu-ronal oxidative damage in Alzheimer’s disease. Am JPathol 1996; 149:21-27.

36. Sultana R, Poon HF, Cai J, et al. Identification of nitrated proteins in Alzheimer’s disease using aredox proteomics approach. Neurobiol Dis 2006;22:76-87.

37. Tohgi H, Abe T, Yamazaki K, et al. Alterations of 3-nityrosine concentration in the cerebrospinal fluidduring aging and in patients with Alzheimer’s dis-ease. Neurosci Lett 1999; 269:52-54.

90 D.A. Butterfield



38. Sultana R, Newman S, Mohammad-Abdul H, et al.Protective effect of the xanthate, D609, onAlzheimer’s amyloid β-peptide (1-42)-inducedoxidative stress in primary neuronal cells. Free RadicRes 2004; 38:449-58.

39. Sultana R, Ravagna A, Mohmmad-Abdul H, et al.

Ferulic acid ethyl ester protects neurons against amy-loid β-peptide (1-42)-induced oxidative stress andneurotoxicity: relationship to antioxidant activity.J Neurochem 2005; 92:749-758.

40. Drake J, Sultana R, Aksenova M, et al. Elevation of mitochondrial glutathione by gamma-glutamylcysteineethyl ester protects mitochondria against peroxyni-trite-induced oxidative stress. J Neurosci Res 2003;74:917-27.

41. Drake, J, Kanski, J, Varadarajan, S, et al. Elevation of brain glutathione by g-glutamylcysteine ethyl esterprotects against peroxynitrite-induced oxidativestress. J Neurosci Res 2002; 68:776-84.

42. Koppal, T, Drake, J, Yatin, S, et al. Peroxynitrite-induced alterations in synaptosomal membrane pro-teins: insight into oxidative stress in Alzheimer’sDisease. J Neurochem 1999; 72:310-17.

43. Castegna A, Aksenov M, Aksenova M, et al.Proteomic identification of oxidatively modified pro-teins in Alzheimer’s disease brain part I: creatinekinase BB, glutamine synthetase, and ubiquitin car-boxy-terminal hydrolase L-1. Free Radic Biol Med2002; 33:562-71.

44. Yatin SM, Aksenov M, Butterfield DA. The antioxi-dant vitamin E modulated amyloid β-peptide-inducedcreatine kinase activity inhibition and increased pro-tein oxidation: implications for the free radicalhypothesis of Alzheimer’s disease. Neurochem Res1999; 24:427-35.

45. Butterfield DA. Proteomics: a new approach to inves-tigate oxidative stress in Alzheimer’s disease brain.Brain Res 2004; 1000:1-7.

46. Castegna A, Aksnov M, Thongboonkerd V, et al.Proteomics identification of oxidatively modified

proteins in Alzheimer’s disease brain part II: dihy-dropyrimidinase related protein II, a-enolase, and heatschock cognate 71. J Neurochem 2002; 82:1524-32.

47. Sultana R, Boyd-Kimball D, Poon HF, et al.Oxidative modification and down-regulation of Pin1 in Alzheimer’s disease hippocampus: a redoxproteomics analysis. Neurobiol Aging 2006; 27:918-925.

48. Sultana R, Boyd-Kimball D, Poon HF, et al. Redoxproteomics identification of oxidized proteins inAlzheimer’s disease hippocampus and cerebellum:An approach to understand pathological and bio-chemical alterations in AD. Neurobiol Aging 2006;in press.

49. Sultana R, Perluigi M, Butterfield DA. Redox pro-teomics identification of oxidatively modified pro-teins in Alzheimer’s disease brain and in vivo andin vitro models of AD centered around A β(1-42).J Chromatogr B Analyt Technol Biomed Life Sci2006; 833:3-11.

50. Butterfield DA, Boyd-Kimball D, Castegna A.Proteomics in Alzheimer’s disease: insights intomechanisms of neurodegeneration. J Neurochem2003; 86:1313-1327.

51. Blass JP and Gibson GE. The role of oxidativeabnormalities in the pathophysiology of Alzheimer’sdisease. Rev Neurol 1991; 147:513-525.

52. Scheltens P and Korf ESC. Contribution of neu-roimaging in the diagnosis of Alzheimer’s disease andother dementias. Curr Opin Neurol, 2000; 13:391-96.

53. Shringarpure R, Grune T, Davies KJ. Protein oxidationand 20S proteasome-dependent proteolysis in mam-malian cells. Cell Mol Life Sci 2001; 58:1442-50.

54. Keller JN, Hanni KB, Markesbery WM. Impairedproteasome function in Alzheimer’s disease.J Neurochem 2000; 75:436-39.

55. Lubec G, Nonaka M, Krapfenbauer K, et al.Expression of the dihydropyrimidinase related pro-tein 2 (DRP-2) in Down syndrome and Alzheimer’sdisease brain is downregulated at the mRNA anddysregulated at the protein level. J Neural TransmSuppl 1999; 57:161-77.

56. Schonberger SJ, Edgar PF, Kydd R, et al. Proteomic

analysis of the brain in Alzheimer’s disease:Molecular phenotype of a complex disease process.Proteomics 2001; 1:1519-28.

57. Tsuji T, Shiozaki A, Kohno R, et al. Proteomic pro-filing and neurodegeneration in Alzheimer’s disease.Neurochem Res 2002; 27:1245-53.

58. Coleman PD and Flood DG. Neuron numbers anddendritic extent in normal aging and Alzheimer’s dis-ease. Neurobiol Aging 1987; 8:521-45.

59. Ojika K, Tsugu Y, Mitake S, et al. NMDA receptoractivation enhances the release of a cholinergic dif-

ferentiation peptide (HCNP) from hippocampal neu-rons in vitro. Neuroscience 1998; 101:341-352.60. Giacobini E. Cholinergic function and Alzheimer’s

disease. Int J Geriatr Psychiatry 2003; 8:S1-S5.61. Boyd-Kimball D, Poon HF, Lynn BC, et al.

Proteomic identification of proteins specifically oxi-dized by intracerebral injection of A β(1-42) into ratbrain: implications for Alzheimer’s disease.Neuroscience 2005; 132:313-324.

62. Boyd-Kimball D, Sultana R, Poon HF, et al. γ -Glutamylcysteine ethyl ester protection from A β(1-42)-mediated oxidative stress in neuronal cells:a proteomics approach. J Neurosci Res 2005; 79:707-713.

5. Amyloid β-Peptide(1-42), Oxidative Stress, and Alzheimer’s Disease 91



63. Boyd-Kimball D, Castegna A, Sultana R, et al.Proteomic identification of proteins oxidized by A β(1-42) in synaptosomes: implications for Alzheimer’sdisease. Brain Res 1044:206-215.

64. Boyd-Kimball D, Poon HF, Lynn BC, et al.Proteomic identification of proteins specifically oxi-

dized in Caenorhabditis elegans expressing humanAβ(1-42): implications for Alzheimer’s disease.Neurobiol Aging 2006; in press.

65. Poon HF, Farr SA, Banks WA, et al. Proteomic iden-tification of brain proteins in aged senescence accel-erated mice that have decreased oxidativemodification following administration of antisenseoligonucleotide directed at the A β region of amyloidprecursor protein. Mol Brain Res 2005; 138:8-16.

66. Walsh DM, Klyubin I, Fadeeva JV, et al. Naturallysecreted oligomers of amyloid beta protein potentlyinhibit hippocampal long-term potentiation in vivo.Nature 2002; 416:535-39.

67. Klein WL, Stine WB Jr, Teplow DB. Small assem-blies of unmodified amyloid beta-protein are theproximate neurotoxin in Alzheimer’s disease.Neurobiol Aging 2004; 25:569-80.

68. Bitan G, Tarus B, Vollers SS, et al. A molecularswitch in amyloid assembly: Met35 and amyloidbeta-protein oligomerization. J Am Chem Soc 2003;125:15359-65.

69. Butterfield DA, Boyd-Kimball D. The critical role of methionine 35 in Alzheimer’s amyloid β-peptide (1-

42)-induced oxidative stress and neurotoxicity.Biochim Biophys Acta 2005; 1703:149-156.

70. Kanski J, Aksenova M, Butterfield DA. The hydropho-bic environment of Met35 of Alzheimer’s A β(1-42) isimportant for the neurotoxic and oxidative propertiesof the peptide. Neurotox Res 2002; 4:219-223.

71. Curtin CC, Ali F, Volitakis I, et al. Alzheimer’s dis-ease amyloid-beta binds copper and zinc to generatean allosterically ordered membrane-penetratingstructure containing superoxide dismutase-like sub-units. J Biol Chem 2001; 276:20466-73.

72. Kanski J, Aksenova M, Butterfield DA. Substitutionof isoleucine-31 by helical-breaking proline abol-ishes oxidative and neurotoxic properties of Alzheimer’s amyloid beta-peptide. Free Radic BiolMed 2002; 32:1205-11.

73. Varadarajan S, Kanski J, Aksenova M, et al.Different mechanisms of oxidative stress and neuro-toxicity for Alzheimer’s A β(1-42) and A β(25-35).J Am Chem Soc 2001; 123:5625-31.

74. Schoneich C. Methionine oxidation by reactive oxy-gen species: reaction mechanisms and relevance toAlzheimer’s disease. Biochim Biophys Acta 2005;1703:111-119.

75. Butterfield DA and Bush AI. Alzheimer’s amyloid β-peptide (1-42): involvement of methionine residue 35in the oxidative stress and neurotoxicity properties of this peptide. Neurobiol Aging 2004; 25:563-68.

76. Pogocki D and Schöneich C. Redox properties of Met35 in neurotoxic b-amyloid peptide. A molecular

modeling study. Chem Res Toxicol 2002; 15:408-18.77. Schöneich C, Pogocki D, Hug GL, et al. Free radical

reactions of methionine in peptides: mechanisms rel-evant to beta-amyloid oxidation and Alzheimer’s dis-ease. J Am Chem Soc 2003; 125:13700-13.

78. Kadlcik V, Sicard-Roselli C, Mattioli T, et al. One-electron oxidation of β-amyloid peptide: sequencemodulation of reactivity. Free Radic Biol Med 2004;37:881-91.

79. Huang X, Cuajungco MP, Atwood CS, et al. Cu(II)potentiation of alzheimer abeta neurotoxicity.Correlation with cell-free hydrogen peroxide produc-tion and metal reduction. J Biol Chem. 1999;274:37111-16.

80. Barnham KJ, Ciccotosto GD, Tickler AK, et al.Neurotoxic, Redox-competent Alzheimer’s β-amy-loid is released from lipid membrane by methionineoxidation. J Biol Chem 2003; 278:42959-65.

81. Boyd-Kimball D, Abdul-Mohmmad H, Reed T, et al.Role of phenylalanine 20 in Alzheimer’s Amyloid b-peptide (1-42)-induced oxidative stress and neuro-toxicity. Chem Res Toxicol 2004; 17:1743-1749.

82. Boyd-Kimball D, Sultana R, Abdul-Mohammad H,

et al. Rodent A β(1-42) exhibits oxidative stressproperties similar to that of human A β(1-42):Implications for proposed mechanisms of toxicity.J Alzheimers Dis 2004; 6:515-525.

83. Dong J, Atwood CS, Anderson VE, et al. Metal bind-ing and oxidation of Amyloid- β within isolatedsenile plaque cores: Raman microscopic evidence.Biochemistry 2003; 42:2768-73.

84. Kou YM, Kokjohn TA, Beach TG, et al. Comparativeanalysis of amyloid- β chemical structure and amy-loid plaque morphology of transgenic mouse and

Alzheimer’s disease brains. J Biol Chem 2001;276:12991-98.85. Naslund J, Schierhorn A, Hellman U, et al. Relative

abundance of Alzheimer A β amyloid peptide vari-ants in Alzheimer’s disease and normal aging. ProcNatl Acad Sci U S A 1994; 91:8378-82.

86. Yan SD, Zhu H, Zhu A, et al. Receptor-dependentcell stress and amyloid accumulation in systemicamyloidosis. Nat Med 2000; 6:633-34.

87. Hou L, Kang I, Marchant RE, et al. Methionine 35oxidation reduces fibril assembly of the amyloidabeta-(1-42) peptide of Alzheimer’s disease. J BiolChem 2002; 277:40173-76.

92 D.A. Butterfield



6Amyloid Toxicity, Synaptic

Dysfunction, and the Biochemistry of Neurodegeneration in Alzheimer’sDiseaseJudy Ng, Marie-Isabel Aguilar, and David H. Small

93

6.1 Introduction

Despite considerable progress over the past fewyears in our understanding of β-amyloid protein(Aβ) production, aggregation, and degradation, lit-tle is known about the mechanism of A β-mediatedneurotoxicity. Although numerous targets of A β’saction have been reported [1], it has been difficult todetermine which, if any, of these targets is impor-tant for disease causation. In this article, we reviewwhat is known about the cellular and biochemical

mechanisms involved in A β neurotoxicity(Fig. 6.1).

6.2 Cellular Mechanisms of Neurotoxicity: Cell Loss versusSynaptic Dystrophy

Considerable attention has been paid to the mecha-nisms by which A β causes neuronal cell death.

Studies have implicated a variety of mechanisms(e.g., generation of reactive oxygen species, cas-pase activation, disturbanced in calcium homeosta-sis) in A β-induced cell death [1]. However,although the number of neurons is lower in the ADbrain compared with age-matched brains, there aregood reasons to believe that cell loss does not playan important role in cognitive decline in AD. First,cell loss is only a minor neuropathologic feature of AD, and it is poorly correlated with cognitivedecline [2]. Most of the brain atrophy can beaccounted for by synaptic loss, rather than adecrease in the number of cell bodies [2]. Second,

it may be argued on purely theoretical grounds thatthe pattern of retrograde amnesia that occurs in ADis unlikely to be caused by cell death.Computational studies involving attractor neuralnetwork models of memory suggest that synapticdysfunction is more likely to be the mechanismthat causes memory loss [1].

In contrast with cell death, neuritic dystrophy isan important diagnostic and pathologic feature of AD. Amyloid plaques are commonly surroundedby neurofibrillary tangle-bearing dystrophic neu-

rites. Aberrant neuronal sprouting can be seen inareas of synaptic loss in the hippocampal forma-tion and neocortex [3]. The dystrophic neurites area characteristic of AD brains and are typically, butnot exclusively, associated with A β deposition. A βhas been reported to induce neurite dystrophy inculture [4] as well as in mutant mouse models [5].For example, Tsai et al. [6] have recently demon-strated that microdeposits of A β amyloid cancause neuritic dystrophy and the breakage of neu-ronal branches in an APP transgenic mouse modelof AD.

6.3 A β Aggregation: The Searchfor Neurotoxic Species

Aggregation of A β is a key step in the generationof neurotoxic A β species. A β neurotoxicity isincreased when the peptide is incubated over manyhours to days, a process known as aging [7].Although there is a relationship between aggrega-tion and toxicity, the major toxic form of A β in AD



is not known. It has been demonstrated thataggregated A β in fibrillar form has neurotoxicproperties in cell culture as well as in vivo.However, more recent findings suggest a toxic roleof A β oligomeric species [8]. In vitro studies haveshown that oligomeric A β, particularly diffusiblelow-molecular-weight species, are neurotoxic [9,10]. This idea is reinforced by genetic studies,which demonstrate that familial AD mutationsfavor the production A β species that aggregatemore readily [11].

Aβ aggregation is a complex process that isinfluenced by incubation time, concentration, tem-perature, pH, and ionic strength. Initially,monomeric A β probably develops an abnormalconformation, after which a variety of differentaggregated structures, including oligomers,protofibrils, spheroids, and mature amyloid fibrils,can be produced. Protofibrils are thin 3- to 4-nm-diameter nonbranching linear aggregates [12],whereas fibrils are ~6 to 10 nm in diameter and arelong and semiflexible [13]. Fibril formation pro-

94 J. Ng et al.

Aβ (monomer)

Aβ (aggregated)

ROS

Lipid raft

ROS

Celldamage ?

Changes inmembrane

fluidity

RAGE p75 NTR Calcium channel

NFκ B

M-CSF Apoptosis

Ca 2+

nAChR

FIGURE 6.1. Possible mechanisms of A β-mediated neurotoxicity. A variety of different mechanisms have been pro-posed to explain the neurotoxic effects of A β. These mechanisms include the generation of ROS; binding to p75NTR,RAGE, or nAChRs. The interaction of A β with lipid rafts may disturb membrane fluidity and alter the function of membrane proteins such as calcium channels. It is still not clear which, if any, of these mechanisms may contributeto the synaptic dysfunction that is thought to underlie the cognitive decline in AD.



ceeds with a lag time, which has been interpretedas a nucleation-dependent process, where oligomerformation takes place through the initial formationof nuclei or seeds [14, 15]. This idea is supportedby studies where prepolymerized A β was added tomonomeric protein, which led to the immediateonset of fibril formation [7, 14].

In the past, it was thought that only fibrillar A βwas pathogenic. However, new evidence supportsthe hypothesis that prefibrillar structures may beeven more important in AD. Brain cell damage anddementia do not correlate well with plaque locationand quantity [16]. However, soluble A β oligomersare found in human AD cerebrospinal fluid, and thesoluble A β content of human brain is better corre-lated with the severity of the disease than plaque

density [17, 18]. Oxidative stress has been shownto precede fibrillar deposition of A β, suggestingthat oxidative stress observed in the AD brain maybe caused by nonfibrillar forms of A β [19]. It haseven been suggested that plaques may not be toxic,and that instead, they may have a protective role inAD by decreasing the amount of the more toxicprefibrillar A β species [20].

6.4 Biochemical Effects of A β

The exact sequence of events whereby A β causesneurodegeneration in AD is not known. In vitro,Aβ can cause oxidative stress, mitochondrial dys-function, disturbances in calcium homeostasis, andmicroglial activation [1]. However, the relativecontribution of these biochemical changes to neu-rodegeneration in vivo is unclear.

6.5 Oxidative Stress andMitochondrial Dysfunction

Aβ neurotoxicity is associated with oxidative stressand mitochondrial dysfunction [21]. Changes inmitochondrial enzymes have been described in theAD brain [22]. For example, cytochrome oxidaseactivity is decreased in AD [23], and defects inmitochondrial energy metabolism can lead toincreased production of reactive oxygen species(ROS). Increased A β is associated with increasednitric oxide (NO) and reduced ATP levels [24]. NOcan, in turn, interact with superoxide radicals to

form peroxynitrite, which can damage cells bypromoting membrane lipid peroxidation andapoptosis [25].

The interaction of metal ions with A β has beenproposed to accelerate peptide aggregation and ini-tiate hydrogen peroxide generation [26], althoughthere is not yet strong evidence for metal-A β inter-actions in vivo. During the process of aggregationin vitro, A β can generate hydrogen peroxide andfree radicals in the presence of Cu + or Fe 2+ [27].The binding of A β to Zn 2+ does not generate ROS,although Zn 2+ competes with Cu + or Fe 2+ for bind-ing to A β and therefore Zn 2+ could inhibit the oxi-dizing properties of metal-bound A β [28]. Theproduction of these ROS induces membrane lipidperoxidation, which can impair the function of

membrane enzymes [29, 30], which in turn cancause an elevation in intracellular calcium [29].The ability of antioxidants to prevent the loss of membrane enzyme function as well as to stabilizecalcium homeostasis in vitro supports the role of membrane lipid peroxidation by A β [31, 32]. Themajor antioxidant glutathione (GSH) is greatlyreduced in astrocytes and neurons exposed to A β[33, 34].

The role of oxidation in A β-induced neurode-generation in vivo still remains very unclear.Notwithstanding the success of the in vitro experi-ments and evidence from epidemiological studiesthat antioxidants may be of value for the treatmentof vascular dementia [35], antioxidants have yet toprove themselves in clinical trials for the treatmentof AD [36]. There are many possible reasons forthis failure. For example, the right drug may not yethave been found. However, it is also possible thatthe oxidative changes seen in vivo are the conse-quence of the neurodegeneration rather being than

the underlying cause .

6.6 The Role of the EndoplasmicReticulum

Some studies suggest that neuronal dysfunction inAD could arise from a defect in the endoplasmicreticulum (ER). As the ER is involved in proteinfolding and assembly, ER dysfunction could con-tribute to abnormal protein folding. It has been sug-gested that ER dysfunction could be due to a defectin the presenilins [37, 38]. Indeed, cells expressing

6. Amyloid Toxicity, Synaptic Dysfunction, and Biochemistry of Neurodegeneration in AD 95



mutant presenilins have an impaired ER responseto stress [39]. However, presenilin mutations mayalso cause an increase in A β production [13, 38],which is known to be linked to AD pathogenesis. Itis still unclear what role ER dysfunction plays infamilial AD caused by presenilin mutations.

6.7 A β-Membrane Interactions

The binding of A β to a component of the plasmamembrane may be the first event in A β-mediatedneurotoxicity [1]. A β has been shown to interacteither directly or indirectly with a number of dif-ferent membrane components including lipids, car-bohydrates, ion channels, and receptors. This

section describes some of the interactions and theirpotential roles in neuronal dysfunction.

6.7.1 Interaction of A β with MembraneLipids

Membrane lipids are localized in differentdomains: exofacial and cytofacial leaflets, choles-terol pools, annular lipids, and lipid rafts [40]. A βcan interact strongly with the lipid bilayer [41, 42].This binding causes an increase in A β fibrillogene-sis and modifications of bilayer properties [42]. A βbinds strongly to gangliosides and lipid rafts [43],which are also rich in cholesterol. Lipid rafts con-taining a ganglioside cluster serve as a conforma-tional catalyst or chaperone, helping to seed A βoligomerization after binding [44, 45]. In mice, A βdimers appear in lipid rafts at 6 months of age andthen continue to accumulate by 24–28 months of age [46].

Although it has been observed that A β binds

preferentially to acidic lipids, it has also been sug-gested that charge-charge interactions are notrequired for A β-membrane interactions [47].However, this idea is not supported by the results of Subasinghe et al. [42], which demonstrate that A βbinds exclusively to lipid membranes throughcharge-charge interactions. Liposomes composedof phosphatidylserine and phosphatidylcholineinduce rapid formation of A β aggregates [48].

The consequences of A β binding to membranesfor cell function are unclear. Biological membranesare fluid in nature, and membrane fluidity isimportant for the proper functioning of integral

membrane proteins and signal transduction path-ways. A β may disturb the acyl chain layer of themembrane [49]. A β reportedly decreases mem-brane fluidity so the membrane has a more rigidstructure, with the presence of gangliosidesincreasing this effect [50]. The addition of oligomeric A β to cultured neurons also causes therelease of lipid particles such as cholesterol, phos-pholipids, and monosialogangliosides [51],although the significance of this effect for thepathogenesis of AD is unclear.

6.7.2 Effects of A β on MembraneCalcium Permeability

Insertion of A β into the lipid membrane may set

off a series of independent events including dis-ruption of Ca 2+ homeostasis and free-radical for-mation, catalyzed by perturbation of theconformation of membrane proteins [52]. A β-mediated disruption of calcium homeostasis mayin turn produce downstream effects [53]. A β mayincrease membrane permeability by interactingwith membrane components to destabilize thestructure of the membrane [54, 55], or it may bedirectly inserted into the membrane to form a pore[56, 57]. A β aggregation is associated withenhanced ion permeability [58]. Sustainedincreases in intracellular calcium may alsoenhance the production and release of A β [59, 60].Aβ-induced destabilization of calcium can lead tocaspase activation and apoptosis [61], howeverthis effect may be caused by changes in the ERtransport of calcium rather than from calciumtransported across the plasma membrane.Reduction of calcium release from the ER mayprovide partial protection from A β toxicity by

reducing stress signals in the ER and decreasingthe increase in calcium triggered by A β [62].

6.7.3 Effect of A β on MembraneReceptors

Aβ may exert a toxic effect by binding to or alter-ing the normal function of cell-surface receptors.A number of receptors have been found to interactdirectly or indirectly with A β. There receptorsinclude the α 7-nicotine acetylcholine receptor, thereceptor for advanced glycation end products(RAGE), and the p75 neurotrophin receptor.

96 J. Ng et al.



6.7.3.1 α 7 Nicotinic Acetylcholine Receptor

The nicotinic acetylcholine receptor (nAChR) is amember of the pentameric ligand-gated ion chan-nel family of receptors [63]. In the central nervoussystem, most nicotinic receptors are of the α4β2 or

homomeric α 7 subtype. α7 nAChR receptors are of particular interest for AD because of their high cal-cium permeability, which suggests an importantrole in neuronal plasticity and cognition [64]. α 7nAChRs are mainly located at nerve terminals andare believed to be involved in regulating the neuro-transmitter release that mediates fast cholinergicneurotransmission [65, 66].

Several studies have shown that A β can bind toand influence the activity of α7 nAChRs [67–69]. α7nAChRs are present in senile plaques and A β42selectively and competitively binds α7 nAChRs withhigh affinity [67]. This binding may have functionalconsequences because A β40 and A β42 can impaircholinergic signaling and acetylcholine release [70].Although A β can block α7 nAChRs on neurons inculture [68], other studies suggest that, under certainconditions, A β may activate α7 nAChRs.

The interaction of A β with α 7 nAChRs mayexplain some of the biochemical changes thatoccur in the AD brain. For example, although

acetylcholinesterase (AChE) is decreased in thebrain of AD patients, AChE is increased around theamyloid plaques [71]. Fodero et al. [72] havedemonstrated that this increase may be due to inter-actions between A β and the α 7 nAChR. In primarycortical neurons, A β42 is more potent than A β40 inits ability to increase AChE [72]. Studies by Wanget al. [73] suggest that the binding of A β to α 7nAChRs may also influence phosphorylation path-ways leading to increased tau phosphorylation.

6.7.3.2 p75 Neurotrophin Receptor

The p75 neurotrophin receptor (p75 NTR ) is a mem-ber of the tumor necrosis factor receptor family thatbinds neurotrophins nonselectively and mediatesneuronal apoptosis and survival [74]. p75 NTR canbind A β and may thereby mediate some forms of Aβ toxicity [75–77]. However, notwithstandingthese findings, levels of p75 NTR have been found tocorrelate inversely with the degree of cognitiveimpairment in early AD, supporting the view thatp75 NTR may be protective for AD [78]. The idea

that p75 NTR is neuroprotective for AD is furthersupported by the observation that there areincreased levels of p75 NTR in the presence of extra-cellular A β deposits [79], that low concentrationsof A β increase the level of p75 NTR in primary cul-

tures of neurons, and that this increase protectsneurons from A β-induced toxicity [80].

6.7.3.3 RAGE

The receptor for advanced glycation and end prod-ucts (RAGE) is a member of the immunoglobulinfamily of cell-surface molecules that exhibits awide tissue distribution and interacts with a range of ligands. A β can bind to RAGE, and this binding

may influence neuronal and microglial function[81]. A β is not the only protein that binds to RAGE,as the receptor interacts broadly with β-sheet fibrils[82]. The interaction of A β with RAGE expressedon endothelial cells, neurons, and microglia report-edly causes oxidative stress and activation of thetranscription factor nuclear factor kappa B (NF- κ B)[81], which in turn enhances expression of macrophage-colony stimulating factor (M-CSF)[83]. A β-mediated M-CSF expression has also beendescribed in microglia, and anti-RAGE antibod-ies can block this effect. These findings suggestsa feedback loop may exist, whereby A β-RAGE–mediated microglial activation enhances theexpression of M-CSF and RAGE [84].

6.8 Conclusions

We still have a relatively poor understanding of themechanism(s) by which A β causes neurotoxicity.

There is increasing evidence to suggest that A βtoxicity is caused by synaptic dysfunction ratherthan cell death. It is clear that aggregation of A β isa key step in the generation of neurotoxic species.However, whether the toxic species are fibrils,protofibrils, amyloid β derived diffusible ligands(ADDLs), or some other aggregated form of A βremains to be established. It is also clear thatAβ can promote the formation of ROS as well asincrease oxidation. The central question is whetherthese changes in oxidation are the underlying causeof synaptic dysfunction or simply the effect of some neurodegenerative mechanism.




References

1. Small, D.H., Mok, S.S., and Bornstein, J.C.,Alzheimer’s disease and A β toxicity: from top to bot-tom. Nat. Rev. Neurosci., 2001; 2:595-598.

2. Terry, R.D., Cell death or synaptic loss in

Alzheimer’s disease. J. Neuropathol. Exp. Neurol.,2000; 59:1118-1119.3. Tolnay, M. and Probst, A., Review: tau protein pathol-

ogy in Alzheimer’s disease and related disorders.Neuropathol. Appl. Neurobiol., 1999; 25:171-187.

4. Postuma, R.B., He, W., Nunan, J., et al., Substrate-bound β -amyloid peptides inhibit cell adhesion andneurite outgrowth in primary neuronal cultures.J. Neurochem., 2000; 74:1122-1130.

5. Phinney, A.L., Deller, T., Stalder, M., et al., Cerebralamyloid induces aberrant axonal sprouting andectopic terminal formation in amyloid precursor pro-tein transgenic mice. J. Neurosci., 1999; 19:8552-8559.

6. Tsai, J., Grutzendler, J., Duff, K., et al., Fibrillaramyloid deposition leads to local synaptic abnormal-ities and breakage of neuronal branches. Nat.Neurosci., 2004; 7:1181-1183.

7. Jarret, J.T. and Lansbury, P.T.J., Seeding “one-dimensional crystallization” of amyloid: A patho-genic mechanism in Alzheimer’s disease and scrapie.Cell, 1993; 73:1055-1058.

8. Hartley, D.M., Walsh, D.M., Ye, C.P., et al.,

Protofibrillar intermediates of amyloid β-proteininduce acute electrophysiological changes and pro-gressive neurotoxicity in cortical neurons.J. Neurosci., 1999; 19(20):8876-8884.

9. Lambert, M.P., Barlow, A.K., Chromy, B.A., et al.,Diffusible, nonfibrillar ligands derived from A β1-42are potent central nervous system neurotoxins. Proc.Natl. Acad. Sci. U. S. A., 1998; 95(11):6448-6453.

10. Gong, Y., Chang, L., Viola, K.L., et al., Alzheimer’sdisease-affected brain: presence of oligomeric A βligands (ADDLs) suggests a molecular basis forreversible memory loss. Proc. Natl. Acad. Sci. U. S.A., 2003; 100(18):10417-10422.

11. Scheuner, D., Eckman, C., Jensen, M., et al.,Secreted amyloid β-protein similar to that in thesenile plaques of Alzheimer’s disease in increasedin vivo by the presenilin 1 and 2 and APP mutationslinked to familial Alzheimer’s disease. Nat. Med.,1996; 2:864-870.

12. Stine, W.B.J., Snyder, S.W., Ladror, U.S., et al., Thenanometer-scale structure of amyloid- β visualised byatomic force microscopy. J. Protein Chem., 1996; 15:193-203.

13. Fraser, P.E., Duffy, L.K., O’Malley, M.B., et al.,Morphology and anitibody recognition of synthetic

β-amyloid peptides. J. Neurosci. Res., 1991; 28:475-485.

14. Harper, J.D., Wong, S.S., Leiber, C.M., et al.,Observation of metastable A β amyloid protofibrilsby atomic force microscopy. Chem. Biol., 1997; 4:119-125.

15. Lansbury, P.T.J., The molecular mechanism of amy-loid formation in Alzheimer’s disease. Eur. J. Med.Chem., 1995; 30 (Suppl.): 621S-633S.

16. Terry, R.D., The neuropathology or Alzheimer’s dis-ease and the structural basis of its cognitive alter-ations. In: Terry, R.D., Katzman, R., Bick, K.L.,Sisodia, S.S. (Eds.), Alzheimer’s disease. LippincottWilliams & Wilkins, Philidelphia, 1999:187-206.

17. Kuo, Y.M., Emmerling, M.R., Vigo-Pelfrey, C.,et al., Water-soluble A β (N-40, N-42) oligomers innormal and Alzheimer’s disease brains. J. Biol.Chem., 1996; 271(8):4077-4081.

18. McLean, C.A., Cherny, R.A., Fraser, F.W., et al.,Soluble pool of A β amyloid as a determinant of severity of neurodegeneration in Alzheimer’s dis-ease. Ann. Neurol., 1999; 46(6):860-866.

19. Drake, J., Link, C.D., and Butterfield, D.A.,Oxidative stress precedes fibrillar deposition of Alzheimer’s disease amyloid β-peptide (1-42) in atrangenic Caenorhabditis elegans model. Neurobiol.Aging, 2003; 24:415-420.

20. Obrenovich, M.E., Joseph, J.A., Atwood, C.S., et al.,Amyloid- β: a (life) preserver for the brain.

Neurobiol. Aging, 2002; 23(6):1097-1099.21. Pike, C.J., Walencewicz, A.J., Glabe, C.G., et al., In

vitro aging of β-amyloid protein causes peptideaggregation and neurotoxicity. Brain Res., 1991;563:311-314.

22. Gibson, G.E., Sheu, K.F., and Blass, J.P., Abnormalitiesof mitochondrial enzymes in Alzheimer’s disease.J. Neural Transm., 1998; 105:855-870.

23. Swerdlow, R.H. and Kish, S.J., Mitochondria inAlzheimer’s disease. Int. Rev. Neurobiol., 2002; 53:341-385.

24. Keil, U., Bonert, A., Marques, C.A., et al., Amyloid-β induced changes in nitric oxide production andmitochondrial activity lead to apoptosis. J. Biol.Chem., 2004; in press.

25. Butterfield, D.A., Castegna, A., Lauderback, C.M.,et al., Evidence that amyloid β-peptide-induced lipidperoxidation and its sequelae in Alzheimer’s diseasebrain contribute to neuronal death. Neurobiol. Aging,2002; 23:655-664.

26. Barnham, K.J., Masters, C.L., and Bush, A.I.,Neurodegeneration diseases and oxidative stress.Nat. Rev. Drug Discov., 2004; 3:205-214.

27. Hensley, K., Carney, J.M., Mattson, M.P., et al.,A model for β-amyloid aggregation and neurotoxicity

98 J. Ng et al.



base on free radical generation by the peptide: rele-vance to Alzheimer’s disease. Proc. Natl. Acad. Sci.U. S. A., 1994; 91:3270-3274.

28. Cuajungco, M.P., Goldstein, L.E., Nunomura, A.,et al., Evidence that the β-amyloid plaques of Alzheimer’s disease represent the redox-silencing

and entombment of A β by zinc. J. Biol. Chem., 2000;275:19439-19442.

29. Mark, R.J., Hensley, K., Butterfield, D.A., et al.,Amyloid β-peptide impairs ion-motive ATPase activ-ities: evidence for a role in loss of neuronal Ca 2+

homeostasis and cell death. J. Neurosci., 1995; 15:6239-6249.

30. Mark, R.J., Pang, Z., Geddes, J.W., et al., Amyloid β-peptide impairs glucose uptake in hippocampal andcortical neurons: involvement of membrane lipid per-oxidation. J. Neurosci., 1997; 17:1046-1054.

31. Goodman, Y. and Mattson, M.P., Secreted forms of β-amyloid precursor protein protect hippocampalneurons against amyloid β-peptide-induced oxidativeinjury. Exp. Neurol., 1994; 128:1-12.

32. Goodman, Y., Bruce, A.J., Cheng, B., et al.,Estrogens attenuate and corticosterone exacerbatesexcitotoxicity, oxidative injury and amyloid β-pep-tide toxicity in hippocampal neurons. J. Neurochem.,1996; 66:1836-1844.

33. Abramov, A.Y., Canevari, L., and Duchen, M.R.,Changes in intracellular calcium and glutathione inastrocytes as the primary mechanism of amyloid neu-

rotoxicity. J. Neurosci., 2003; 23:5088-5095.34. Keelan, J., Allen, N.J., Antcliffe, D., et al.,

Quantitative imaging of glutathione in hippocampalneurons and glia in culture using monochlorobimane.J. Neurosci. Res., 2001; 66:873-884.

35. Masaki, K.H., Losonczy, K.G., Izmirlian, G., et al.,Association of vitamin E and C supplement use withcognitive function and dementia in elderly men.Neurology, 2000; 54:1265-1272.

36. Thal, L.J., Grundman, M., Berg, J., et al., Idebenonetreatment fails to slow cognitive decline in Alzheimer’s

disease. Neurology, 2003; 61:1498-1502.37. Katayama, T., Imaizumi, K., Sato, N., et al.,Presenilin-1 mutations downregulate the signallingpathway of the unfolded-protein response. Nat. CellBiol., 1999; 1(8):479-485.

38. Sato, N., Imaizumi, K., Manabe, T., et al., Increasedproduction of β-amyloid and vulnerability to endo-plasmic reticulum stress by an aberrant spliced form of presenilin 2. J. Biol. Chem., 2001; 276(3):2108-2114.

39. Guo, Q., Sopher, B.L., Furukawa, K., et al., Alzheimer’spresenilin mutation sensitizes neural cells to apoptosisinduced by trophic factor withdrawal and amyloid-peptide: involvement of calcium and oxyradicals.J. Neurosci., 1997; 17:4212-4222.

40. Tsui-Pierchala, B.A., Encinas, M., Milbrandt, J.,et al., Lipid rafts in neuronal signaling and function.Trends Neurosci., 2002; 25:412-417.

41. Terzi, E., Holzemann, G., and Seelig, J., Interactionof Alzheimer β-amyloid peptide(1-40) with lipidmembranes. Biochemistry (Moscow). 1997; 36:

14845-14852.42. Subasinghe, S., Unabia, S., Barrow, C.J., et al.,

Cholesterol is necessary both for the toxic effect of Aβ peptides on vascular smooth muscle cells and forAβ binding to vascular smooth muscle cell mem-branes. J. Neurochem., 2003; 84:471-479.

43. Kakio, A., Nishimoto, S., Kozutsumi, Y., et al.,Formation of a membrane-active form of amyloid β-protein in raft-like model membranes. Biochem.Biophys. Res. Commun., 2003; 303:514-518.

44. Kakio, A., Nishimoto, I., Kozutsumi, Y., et al.,Formation of a membrane-active form of amyloid β-protein in raft-like membranes. Biochem. Biophys.Res. Commun., 2003; 303:514-518.

45. Yip, C.M., Darabie, A.A., and McLaurin, J., A β42-peptide assembly on lipid bilayers. J. Mol. Biol.,2002; 318:97-107.

46. Kawarabayashi, T., Shoji, M., Younkin, L.H., et al.,Dimeric amyloid β protein rapidly accumulates inlipid rafts followed by apolipoprotein E and phos-phorylated tau accumulation in the Tg2576 mousemodel of Alzheimer’s disease. Neurobiol. Dis., 2004;24:3801-3809.

47. Kremer, J.J., Sklansky, D.J., and Murphy, R.M.,Profile of changes in lipid bilayer structure caused byβ-amloid peptide. Biochemistry (Moscow). 2001;40:8563-8571.

48. Zhao, H., Tuominen, E.K.J., and Kinnunen, P.K.J.,Formation of amyloid fibres triggered by phospho-tidylserine-containing membranes. Biochemistry(Moscow). 2004; 43:10302-10307.

49. Muller, W.E., Kirsch, C., and Eckert, G.P.,Membrane-disordering effects of β-amyloid pep-tides. Biochem. Soc. Trans., 2001; 29:617-623.

50. Kremer, J.J., Pallitto, M.M., Sklansky, D.J., et al.,Correlation of β-amyloid aggregate size andhydrophobicity with decreased bilayer fluidity of model membranes. Biochemistry (Moscow). 2000;39:10309-10318.

51. Michikawa, M., Gong, J.S., Fan, Q.W., et al., A novelaction of alzheimer’s amyloid β-protein (A β):oligomeric A β promotes lipid release. J. Neurosci.,2001; 21:7226-7235.

52. Kanfer, J.N., Sorrentino, G., and Sitar, D.S., Amyloid βpeptide membrane perturbation is the basis for its bio-logical effects. Neurochem. Res., 1999; 24:1621-1630.

53. Mattson, M.P., Cheng, B., Davis, D., et al., β-Amyloid peptides destabilize calcium homeostasis




and render human cortical neurons vulnerable toexcitotoxity. J. Neurosci., 1992; 12:376-389.

54. Muller, W.E., Koch, S., Eckert, A., et al., β-Amyloidpeptide decreases membrane fluidity. Brain Res.,1995; 674:133-136.

55. Mason, R.P., Estermyer, J.D., Kelly, J.F., et al.,

Alzheimer’s disease amyloid β peptide 25-35 inlocalized in the membrane hydrocarbon core: X-raydiffraction analysis. Biochem. Biophys. Res.Commun., 1996; 222:78-82.

56. Kawahara, M., Arispe, N., Kuroda, Y., et al.,Alzheimer’s disease amyloid β-protein forms Zn 2+-sensitive cation-selective channels across excitedmembrane patches from hypothalamic neurons.Biophys. J., 1997; 73:67-75.

57. Arispe, N., Rojas, E., and Pollard, H.B., Alzheimer’sdisease amyloid β-protein forms calcium channels inbilayer membranes: blockade by tromethamine andaluminium. Proc. Natl. Acad. Sci. U. S. A., 1993; 90:567-571.

58. Hirakura, Y., Lin, M.C., and Kagan, B.L., Alzheimeramyloid A β 1-42 channels: effects of solvent, pH,and Congo red. J. Neurosci. Res., 1999; 57:458-466.

59. Querfurth, H.W. and Selkoe, D.J., Calciumionophore increases amyloid β peptide production bycultured cells. Biochemistry (Moscow). 1994; 33:4550-4561.

60. Pierrot, N., Ghisdal, P., Caumont, A., et al.,Intraneuronal amyloid- β42 production triggered by

sustained increase of cytosolic calcium concentrationinduces neuronal death. J. Neurochem., 2004; 88:1140-1150.

61. Ferreiro, E., Oliveira, C.R., and Pereira, C.,Involvement of endoplasmic reticulum Ca 2+ releasethrough ryanodine and inositol 1,4,5-triphosphatereceptors in the neurotoxic effects induced by theamyloid- β peptide. J. Neurosci. Res., 2004; 76:872-880.

62. Suen, K.C., Lin, K.F., Elyaman, W., et al., Reductionof calcium release from the endoplasmic reticulum

could only provide partial neuroprotection again β-amyloid peptide toxicity. J. Neurochem., 2003; 87:1413-1426.

63. Broide, R.S. and Leslie, F.M., The alpha7 nicotinicacetylcholine receptor in neuronal plasticity. Mol.Neurobiol., 1999; 20:1-16.

64. Small, D.H. and Fodero, L.R., Cholinergic regulationof synaptic plasticity as a therapeutic target inAlzheimer’s disease. J. Alzheimers Dis., 2002; 4:349-355.

65. Gray, R., Rajan, A.S., Radcliffe, K.A., et al.,Hippocampal synaptic transmission enhanced bylow concentrations of nicotine. Nature, 1996; 383:713-716.

66. Chang, K.T. and Berg, D.K., Nicotinic acetylcholinereceptors containing α 7 subunits are required forreliable synaptic transmission in situ. J. Neurosci.,1999; 19:3701-3710.

67. Wang, H.Y., Lee, D.H., D’Andrea, M.R., et al., b-Amyloid (1-42) binds to alpha7 nicotinic acetyl-

choline receptor with high affinity. Implications forAlzheimer’s disease pathology. J. Biol. Chem., 2000;275:5626-5632.

68. Liu, Q., Kawai, H., and Berg, D.K., β-Amyloid pep-tide blocks the response of α7-containing nicotinicreceptors on hippocampal neurons. Proc. Natl. Acad.Sci. U. S. A., 2001; 98:4734-4739.

69. Dineley, K.T., Bell, K.A., Bui, D., et al., β-Amyloidpeptide activates α 7 nicotinic acetylcholine receptorsexpressed in Xenopus oocytes. J. Biol. Chem., 2002;277:25056-25061.

70. Kar, S., Issa, A.M., Seto, D., et al., Amyloid β-pep-tide inhibits high-affinity choline uptake and acetyl-choline release in rat hippocampal slices.J. Neurochem., 1998; 70:2179-2187.

71. Atack, J.R., Perry, E.K., Bonham, J.R., et al.,Molecular forms of acetylacholinesterase in seniledementia of Alzheimer type: selective loss of theintermediate (10S) form. Neurosci. Lett., 1983; 40:199-204.

72. Fodero, L.R., Mok, S.S., Losic, D., et al., α 7-Nicotinic acetylcholine receptors mediate an A β1-42-induced increase in the level of acetylcholinesterase

in primary cortical neurones. J. Neurochem., 2004;88:1186-1193.

73. Wang, H.Y., D’Andrea, M.R., and Nagele, R.G.,Cerebellar diffuse amyloid plaques are derived fromdendritic A β42 accumulations in Purkinje cells.Neurobiol. Aging, 2002; 23:213-223.

74. Barker, P.A., p75NTR is positively promiscuous:novel partners and new insights. Neuron, 2004; 42:529-533.

75. Rabizadeh, A., Bitler, C.M., Butcher, L.L., et al.,Expression of the low-affinity nerve growth factor

receptor enhances β-amyloid peptide toxicity.Proc. Natl. Acad. Sci. U. S. A., 1994; 91:10703-10706.

76. Yaar, M., Zhai, S., Pilch, P., et al., Binding of b-amy-loid to the p75 neurotrophin receptor induces apop-tosis. A possible mechanism for Alzheimer’s disease.J. Clin. Invest., 1997; 100:2333-2340.

77. Tsukamoto, E., Hashimoto, Y., Kanekura, K., et al.,Characterization of the toxic mechanism triggered byAlzheimer’s amyloid- β peptides bia p75 neu-rotrophin receptor in neuronal hybrid cells.J. Neurosci. Res., 2003; 73:627-636.

78. Mufson, E.J. and Kordower, J.H., Cortical neuronsexpress nerve growth factor receptors in advanced

100 J. Ng et al.



age and Alzheimer’s disease. Proc. Natl. Acad. Sci.U. S. A., 1992; 89:569-573.

79. Jaffar, S., Counts, S.E., Ma, S.Y., et al., Neuropathol-ogy of mice carrying mutant APP swe and/or PS 1M146Ltransgenes: alterations in the p75 NTR choinergic basalforebrain septohippocampal pathway. Exp. Neurol.,

2001; 170:277-243.80. Zhang, Y., Hong, Y., Bounhar, Y., et al., p75

Neurotrophin receptor protects primary cultures of human neurons against extracellular amyloid β pep-tide cytotoxicity. J. Neurosci., 2003; 23:7385-7394.

81. Yan, S.D., Chen, X., Fu, J., et al., RAGE and amy-loid- β peptide neurotoxicity in Alzheimer’s disease.Nature, 1996; 382:685-691.

82. Yan, S.D., Zhu, H., Zhu, A., et al., Receptor-depend-ent cell stress and amyloid accumulation in systemicamyloidosis. Nat. Med., 2000; 6:643-651.

83. Du Yan, S., Zhu, H., Fu, J., et al., Amyloid- β pep-tide-receptor for advanced glycation endproductinteraction elicits neuronal expression of

macrophage-colony stimulating factor: a proinflam-matory pathway in Alzheimer’s disease. Proc. Natl.Acad. Sci. U. S. A., 1997; 94:5296-5301.

84. Lue, L.F., Walker, D.G., Brachova, L., et al.,Involvement of microglial receptor for advanced gly-cation endproducts (RAGE) in Alzheimer’s disease:identification of a cellular activation mechanism.Exp. Neurol., 2001; 171:29-45.




7Aβ Variants and Their Impact on

Amyloid Formation and Alzheimer’sDisease ProgressionLaszlo Otvos, Jr.

102

7.1 Introduction

Alzheimer’s disease (AD) is characterized patho-logically by abnormal accumulation of amyloidplaques and neurofibrillary tangles in vulnerablebrain regions [1]. Although the main proteinaceouscomponent of the plaques is the amyloid β peptide(Aβ), the tangles are primarily made up fromhyperphosphorylated versions of the microtubule-associated protein tau [2]. Emerging evidence forthe overlap in the pathological and clinical features

of patients with brain amyloidosis suggests that theplaques and tangles may be linked mechanistically[3]. Increased levels of A β peptides in brain canpromote the formation of intracellular tau aggre-gates, although the mechanism for this process isstill unclear. These results indicate that one form of amyloid can directly or indirectly impact the forma-tion of another form of amyloid composed of differ-ent protein, likely contributing to the overlap inclinical and pathological features. A β is an approxi-mately 4-kDa peptide with a strong potential toaggregate during electrophoresis [4] and when iso-lated from amyloid deposits or control brain tissuerepresents a family of numerous peptide species [5].

It is increasingly believed that A β amyloidogen-esis and Alzheimer’s disease are causally related,and this notion derives from both genetic and cel-lular observations. On one hand, all four genesdefinitively linked to inherited forms of the diseaseto date have been shown to increase the productionand/or deposition of A β in the brain [6]. On the

other hand, drugs known to reduce the prevalenceof Alzheimer’s disease in epidemiological studiesalso reduce A β levels in cultured cells [7]. In gen-

eral, A β aggregates can directly and indirectlymediate neurotoxic effects, inflammatory respon-ses, and abnormal tau phosphorylation, the hall-marks of Alzheimer’s disease [8]. In spite of thiscorrelation, no major differences in A β concentra-tion between samples acquired from diseased ornormal tissues could initially be identified, at leastnot from the cerebrospinal fluid [9]. The explana-tion may rest in the insensitivity of early A β ana-lytical methodology [10] or more likely from theheterogeneity of the samples in Alzheimer’s dis-

ease–affected or normal brains.Aβ was originally isolated and sequenced as a 42(43) residue-long peptide with no sequence homol-ogy to proteins available at that time [11]:

H-Asp1-Ala2-Glu3-Phe4-Arg5-His6-Asp7-Ser8-Gly9-Tyr10-Glu11-Val12-His13-His14-Gln15-Lys16-Leu17-Val18-Phe19-Phe20-Ala21-Glu22-Asp23-Val24-Gly25-Ser26-Asn27-Lys28-Gly29-Ala30-Ile31-Ile32-Gly33-Leu34-Met35-Val36-Gly37-Gly38-Val39-Val40-Ile41-Ala42-(Thr43)-OH

Ensuing biochemical characterization and com-parison of soluble A β secreted by cells, solubleAβ in the cerebrospinal fluid, and insoluble A βisolated from the brains of affected individuals hasrevealed that there are numerous A β species withextensive amino and carboxyl-terminal hetero-geneity as well as featuring a series of mid-chainamino acid alterations [12]. As soon as the alter-ations were discovered, these genetic mutations orpost-translational modifications, including oxida-

tion by radicals, truncations, isomerization, andracemization, were speculated as modifiers of A βmetabolism and/or enhancers of aggregation and



hence as progression factors for familiar and spo-radic cases of Alzheimer’s disease. This articletries to unify the divergent views and provide acomprehensive account for the impact of A β vari-ations in the development of amyloid diseases.

Table 7.1 lists all known major A β sequence mod-ifications and their relevance in molecular or clin-ical pathogenesis.

After a short analysis into the origin of modi-fied A β forms in tissues and cultured cells, wewill concentrate on the major properties of theamyloid protein, as regulated by the amino acidalterations. The two dominant attributes of A β,the golden standards to which every derivative iscompared, are fibrillogenesis [13] and neurotoxi-city [14], this latter frequently related to oxidativestress [15]. Fibril formation can be vieweddirectly as true aggregation [16] or indirectly asthe ability of the peptide to assume β-pleatedsheet conformation, the prerequisite forfibrillogenesis [17]. More precisely, the character-istic α -helix/random coil → β-pleated sheet con-formational transition is considered an easilyobservable sign of increased ability to form aggre-gates [18]. Neurotoxicity can also be studied asdirect killing of cells [19] or as an outcome of

long-lived protein variants, unable to turn overwithin the life cycle of cells [20].

7.2 The Origin of Modified A βForms

Mid-chain modifications, concentrated aroundresidue Glu22, are clearly due to mutations in the

precursor gene. A β is a normally secreted prote-olytic product [21] of the amyloid precursor protein(APP), a 677–770 reside-long type 1 integral mem-brane protein [22]. A constitutive secretory meta-bolic pathway involves APP cleavage at A βposition 16 by the α -secretase enzyme producingtwo halves of A β. When the γ -secretase furthercleaves the product, a carboxy-terminal A β 17-40/42 fragment is formed, named p3 [23]. Duringan alternative proteolytic pathway, a third enzyme,the β-secretase, cleaves APP at the amino-terminusof A β [24] followed by γ -secretase action at the C-terminus producing the full-length amyloid peptide.C-terminal alterations are thought to originate frommutations in the APP gene. Processed from wild-type APP, the major 4-kDa A β species in both con-ditioned medium and human cerebrospinal fluid isAβ 1-40 (>60–70%), although some A β 1-42 is alsopresent ( ≈15%) along with minor amounts of otherAβ fragments [10]. However, when the APP geneincludes mutations immediately downstream of the

Aβ coding region, the production level of A β 1-42significantly increases [25].

7. Impact of A β Variants on Amyloid Formation and AD Progression 103

TABLE 7.1. A β variations known to affect Alzheimer’s disease development.

N-terminal truncations and isomerizationsFirst residue in truncated Ab Species abbreviation in text Presence in amyloid formsD-Asp1 rD-1 In plaques of controls with atherosclerosisisoAsp1 iD-1 Increased amyloid in parenchymapGlu3 pGlu3-Nterm Fifty percent in senile plaquesisoAsp7 iD-7 Increased amyloid in parenchymapGlu11 pGlu11-Nterm Thirty percent in serum

Leu17 p3 Early deposits in Down syndromeMid-chain genetic mutations

Mutated residue Species abbreviation in text Clinical phenotypeAla2 → Thr Thr2 Stroke and myocardial infarctionAla21 → Gly Flemish type Presenile dementia and cerebral hemorrhageGlu22 → Gln Dutch type Cerebral hemorrhageGlu22 → Gly Arctic type Early-onset Alzheimer’s diseaseGlu22 → Lys Italian type Presenile dementia and cerebral hemorrhageAsp23 → Asn Iowa type Early-onset Alzheimer’s diseaseAla42 → Val Val42 SchrizophreniaAla42 → Thr Thr42 Early-onset Alzheimer’s disease

C-terminal truncationLast residue in truncated A β Species abbreviation in text PresentVal40 1-40 When the precursor protein is not mutated

downstream



In wild-type APP, the fourth residue after A βAla42 is a valine; in familiar Alzheimer’s diseasein Anglo-Saxon, Italian, and Japanese kindreds,this Val is substituted with Ile, Phe, or Gly, respec-tively [26–28]. We compared A β production inhuman neuroblastoma (M17) cells transfected withconstructs expressing wild-type APP or theAPP717 mutants by either isolation of metaboli-cally labeled A β from conditioned medium, diges-tion with cyanogen bromide, and analysis of thecarboxyl-terminal peptides released, or by analysisof the amyloid peptide in conditioned medium withimmunosorbent assays that discriminate A β 1-40and 1-42. Both methods demonstrated that A βreleased from wild-type βAPP is primarily, but notexclusively, 40 residues long. The APP717 muta-

tions consistently caused a 1.5- to 1.9-fold increasein the percentage of 42-residue A β generated. Thepathological consequences of longer A β assemblywill be discussed later.

In general, peptides are subjected to endopepti-dase and exopeptidase cleavages with amino- andcarboxy-peptidases being the major culprits for pep-tide degradation [29]. Carboxy-terminal truncationsmay theoretically occur from the cleaved A β 1-42[43] peptides in tissues, but apparently genetic pro-cessing of APP is a more common explanation forexplaining heterogeneity at the C-terminus [30].Indeed, a novel expression system was developed,one that in the secretory pathway selectively gener-ates A β 1-40 or A β 1-42 fused to the transmembraneBRI protein. Significantly, expression of A β 1-42results in no increase in secreted A β 1-40, suggest-ing that the majority of A β 1-42 is not trimmed bycarboxypeptidase to A β 1-40. Yet, as the identityand role of secretases responsible for APP process-ing in the human brain have yet to be clarified [31],

the search for enzyme activities capable of cleavingnative brain APP in human hippocampus is under-way. A 40-kDa protein with proteolytic activity thatdegrades native brain APP in vitro was purified andcharacterized; molecular analysis identified it as anovel protease belonging to the carboxypeptidase Bfamily [32]. PC12 cells overexpressing this proteasegenerate a major 12-kDa A β-bearing peptide incytosol, a peptide that has also been detected in acell-free system using purified brain APP as sub-strate. Having said this, carboxypeptidase process-ing of longer A β variants enjoy much less attentionthan exopeptidase activity at the amino-terminus.

The amino acid sequence of wild-type A β startswith an N-terminal Asp residue, and a Glu residueis found two positions downstream; these aminoacids are the main substrates of aminopeptidase A[33]. When the activity of aminopeptidases as afunction of age or sex was studied, significant age-related increases were observed in glutamicaminopeptidase A activity in both human gendersand in aspartic aminopeptidase A activity infemales [34]. This may reflect the evolution of sus-ceptible circulating substrates during developmentand aging. In support, when specific soluble andmembrane-bound aspartyl-hydrolyzing activitieswere assayed in brain subcellular fractions from ratfetuses (19–20 days of gestation), and from 1- to260-week-old rats, significant age-related changes

were observed in all fractions for both enzymaticactivities [35]. Taken together, it is well conceiv-able that the amino terminal Asp1 and Glu3residues in A β undergo enzymatic degradation.

Alternatively, Asp is subject to a completelynonenzymatic processing pathway. It was hypothe-sized that Alzheimer’s disease is initiated by a pro-tein aging-related structural transformation insoluble A β [36]. According to this theory, sponta-neous chemical modification of aspartyl residues inAβ to transient succinimide induces a non-nativeconformation in a fraction of soluble A β, renderingit amyloidogenic and neurotoxic. As shown later,conformationally altered A β is characterized byincreased stability in solution and the presence of anon-native β-turn that determines folding.Formation of the succinimide from Asp is a resultof an intramolecular nucleophilic attack of the pep-tide amide-nitrogen on the side-chain carbonylgroup of Asp (Fig. 7.1). Hydrolysis of succinimideleads to accumulation of stable isoaspartyl sites

(isoAsp) in which a peptide bond is formed by theside-chain carboxyl group of Asp. A competinghydrolysis pathway leads to the production of pep-tides containing D-aspartic acid.

7.3 Different A β Variants inSpace and Time

In order to identify the proteolytic enzymes respon-sible for the formation of the distinct A β forms andthe organelles in which diverse forms of A β aregenerated and from which they are secreted, the A β

104 L. Otvos, Jr.



compositions of subcellular compartments wereinvestigated together with the compartments fromwhich the A β variants were secreted [37]. It wasfound that A β 1-40 (or A β x-40) is generatedexclusively within the trans-Golgi network andpackaged into post–trans-Golgi network secretoryvesicles; A β x-42 is made and retained within theendoplasmic reticulum in an insoluble state; all A β

42 forms are made in the trans-Golgi network andpackaged into secretory vesicles; and finally theamyloid peptides formed consist of two pools (asoluble population extractable with detergents anda detergent-insoluble form). It was concluded thatcell-free A β generation assays may distinguishbetween intracellular insoluble peptides andsecreted soluble analogues.


Methyltransferase

O

O

O

NH

O

O

CNH

O

NH CO −

L-Aspartate

D-Isoaspartate

D-Aspartate

O

O

CNH

NH

CO −

NH C

CO −

L-Isoaspartate

CNH

NH

O

NH

COMe

L-Isoaspartyl methyl ester

L-Succinimide

D-Succinimide

O

O

C

N

C

C

N

C

O

NH

O

O

NH

C

NH CO −

FIGURE 7.1. Formation of succinimide through spontaneous cyclization of aspartyl residues. Hydrolysis of the cyclicproducts leads to D-aspartate and L- and D-isoaspartates together with the unmodified L-asparate forms. Reprintedfrom Ref. 36, with permission of the Federation of American Societies for Experimental Biology.



To this extent, soluble A β and its variants, pro-duced by mouse neuroblastoma cells, were selec-tively isolated by immunoprecipitation withanti-A β monoclonal antibodies, and the identitiesof these isolated amyloid peptides were determinedby measuring their molecular masses using matrix-assisted laser desorption/ionization time-of-flightmass spectrometry. The relative signal intensitieswere used to estimate the concentrations of A β10.Although pharmacologically mass spectrometrywithout chromatographic quantitation steps is notfully defendable [38], this approach detected sev-eral novel A β variants and successfully quantifiedsoluble A β in conditioned media of cultured mam-malian cells. The identified 64 A β-related peptides(44 from human and 20 from murine amyloid

sequences) included a cascade of N- and C-termi-nal truncations with little preference of a givenstructural motif. The human APP samples featuredan increased abundance of peptides starting withAla2 and Phe4 (in agreement with the hypothe-sized aminopeptidase A activity on Asp1 andGlu3) but without major statistical significance. Atleast, analysis of degradation products of synthetichuman A β peptides revealed four primary cleavagesites (C-terminal to His13, Phe19, Lys28 andGly33) with three different endopeptidase substratespecificities. These A β variants may contribute tothe low levels of certain A β subpopulations nor-mally observed in cell culture media of transfectedcells.

Of course, these findings raise the question as towhich residues promote aggregation and whichendorse soluble A β derivatives. Because thisreview is concerned with natural A β variants, list-ing of all designer A β analogues falls outside thescope of this article. Yet, one study that claims to

represent an unbiased search for sequence determi-nants of A β amyloidogenesis may fit the bill. Thisscreen is based on the finding that fusions of thewild-type A β 1-42 sequence to green fluorescentprotein form insoluble aggregates in which thegreen fluorescent protein is inactive. Cells express-ing such fusions do not fluoresce as opposed to A βwith reduced tendencies to aggregate, which can beconstructed and screened from randomly mutatedAβ 1-42 green fluorescence protein libraries [39].Not surprisingly, most of the observed solubility-enhancer residues are replacements of hydrophobicamino acids in the Leu17-Phe19, Ile31-Ile32,

Leu34-Val36, and Val39-Ala42 fragments. Theonly notable finding is that some conservativeamino acid changes (Val18 → Ala, Phe19 → Leu,and Ile32 → Val) also increase solubility, and thesecuriously fall into or proximal to the detected pri-mary enzymatic cleavage sites of the previousparagraph.

7.4 Animal Models

A major obstacle to the pharmaceutical develop-ment of A β aggregation inhibitors is the lack of appropriate small animal models [40]. In most of the current mouse models of Alzheimer’s disease,the animals contain amyloid plaques in their brain,

but the amyloidosis is not accompanied by exten-sive tangle formation or massive neuronal loss.This is partially understandable if we compare theAβ sequences in different animal species and theirability to form aggregates. When the A β sequencesof human, dog, polar bear, rabbit, cow, sheep, pig,and guinea-pig are compared with the correspon-ding rodent sequences and a phylogenetic tree isgenerated, it is obvious that the A β amino acidsequence of human, dog, and polar bear and othermammals that may form amyloid plaques is con-served, and the mice and rats where amyloid hasnot been detected may be evolutionarily a distinctgroup [41, 42]. In addition, the predicted secondarystructure of mouse and rat A β lacks the propensityto form a β-pleated sheet secondary structure.

Compared with human A β, the amino acidsequence of mouse A β differs at three positions:Arg5 is replaced with Gly, Tyr10 is replaced withPhe, and His13 is replaced with Arg [43], with therat sequence being identical to that of mouse [44].

To study the preferred β-pleated sheet formingability of the human peptide compared with therodent analogue, we synthesized, purified, andcharacterized the two different A β sequences [45].Circular dichroism (CD) and Fourier-transformedinfrared spectroscopy were used with variousmembrane-mimicking solvents, different peptideconcentrations, and variable pH to identify thoseenvironmental conditions that promoted β-pleatedsheet formation of the human versus rodent amy-loid peptides. We found that higher β-pleated sheetcontent was observed for the rodent sequencein acetonitrile/water mixtures. In contrast, more

106 L. Otvos, Jr.



β-pleated sheets were detected for the human A β intrifluoroethanol/water mixtures at neutral pH.Remarkably, at relatively low peptide concentra-tions, only the human sequence assumed anextended secondary structure (Fig. 7.2). These datasuggest that subtle inter-species amino-acid differ-ences may account for the inability of the rodentpeptide to form amyloid fibrils in situ , when onlylow amounts of soluble peptides are available for

aggregation. However, if fibrils once formed, theseN-terminal amino acid differences have virtuallyno effect on the morphology or organization of thefibrils [46]. It needs to be added that in the currentarticle, altered peptide conformations are consid-ered as factors that promote disease pathogenesis.However, the opposite can be equally true: differ-ences in A β secondary structure may be aconsequence of disease progression.


[q ]MR

[q ]MR

20,000

−15,000

−20,000

20,000

180.0Wavelength [nm]

260.0

c

d

a

b

(a)

(b) 180.0 260.0Wavelength [nm]

d

c

b

a

FIGURE 7.2. Circular dichroism spectra of human (A) and rodent (B) A β peptides at different concentrations. The

rodent analogue forms β-pleated sheets at significantly higher concentration than the human version does: a, 0.5mg/mL; b, 0.25 mg/mL; c, 0.125 mg/mL; d, 0.0625 mg/mL. Reprinted from Ref. 45, with permission of theFederation of European Biochemical Society.



Earlier we briefly mentioned that in humanAlzheimer brain, the major C-terminal variant thatforms amyloid fibers is A β 1-42. In contrast, themajor fibrillar aggregates that present Congo redbirefrigence in rat brain consist of the A β 1-40 pep-tide, whereas A β 1-42 aggregates as a nonfibrillaramorphous material [47]. Thus, instead of the lack of deposition process per se , factors might exist inthe rat brain that inhibit the fibrillar assembly of themost pathogenic soluble A β 1-42 variant. In sup-port of differences in fibril assembly rather thanpostsecretory processing, freshly solubilizedhuman A β 1-40 or A β 1-42 were injected into ratbrains, and it was shown that both peptides wereequally processed at their amino-termini to yieldvariants starting at pGlu3 and at their C termini to

yield variants ending at Val40 and at Val39 [47].Contradictory to the previous argument, normal ratbrain can produce enzymes that mediate the con-version of A β 1-40/1-42 into processed variantssimilar to those in Alzheimer’s disease.

Obviously, the loss of the side-chain positivecharge at position 5 in the native rodent A β ana-logue can influence metal-binding, a well-studiedrisk factor in A β aggregation [48] and fibril forma-tion [49]. Indeed, Cu(II) (at concentrations lowerthan that associated with amyloid plaques) inducesthe generation of dityrosine cross-linked, sodiumdodecyl sulfate–resistant oligomers of human, butnot rat, A β peptides [50], and the alteration mustinvolve Tyr10 (also missing in rodent A β) becauseno detectable peroxidative modifications areobserved with A β 12-28 [51]. The coordination of metal ions for human and mouse N-terminal A βfragments starts from the N-terminal Asp residue,which stabilizes significantly the 1N complex as aresult of chelation through the side-chain carboxy-

late group [52]. In a wide pH range of 4–10, theimidazole nitrogen of His6 is coordinated to form amacrochelate. Results show that, in the pH range5–9, the human fragments form the complex withdifferent coordination mode compared with that of the mouse fragments. The low pK(1)(amide) val-ues (approximately 5) obtained for the mouse N-terminal A β fragments may suggest thecoordination of the amide nitrogen of His6 while incase of the human fragments the coordination of the amide nitrogen of Ala2 is a more likely sce-nario. The Gly → Arg residue replacement in posi-tion 5 of the A β peptide sequence changes the

coordination modes of a peptide to metal ion in thephysiological pH range. The mouse fragments of Aβ are much more effective in Cu(II) binding thanthe human fragments.

Human and rat variants of A β 1-42 were com-pared to determine whether they produce the sameamount of neuronal loss when combined with iron[53]. Coinjection of iron with either A β variantcaused significantly more neuronal loss than theAβ peptide alone, suggesting that iron may con-tribute to the toxicity associated with senileplaques. Rat A β 1-42 combined with iron was astoxic as iron alone, whereas iron combined withhuman A β 1-42 was significantly less toxic. Thislatter finding indicates that fibrillar human A β isable to reduce iron-induced neurotoxicity in vivo

and raises the interesting possibility that senileplaques in Alzheimer’s disease may represent aneuroprotective response to the presence of ele-vated metal ions.

When the human sequence is introduced intorodents, a thorough chemical and morphologicalcomparison of the A β molecules and the amyloidplaques present in the brains of APP transgenicmice and human Alzheimer’s disease patients showthat despite an apparent overall structural resem-blance to Alzheimer pathology, transgenic miceproduce amyloid cores that are completely solublein buffers containing sodium dodecyl sulfate,whereas human amyloid plaques are highly resist-ant to chemical and physical disruption [54]. It wassuggested that A β chemical alterations account forthe extreme stability of Alzheimer plaque coreamyloid. Curiously, the corresponding lack of post-translational modifications such as N-terminaldegradation, isomerization, racemization, pyroglu-tamyl formation, oxidation, and covalently linked

dimers, all the alterations we review in this article,in transgenic mouse A β may provide an explana-tion for the differences in solubility betweenhuman and APP transgenic mouse plaques. It washypothesized that either insufficient time is avail-able for A β structural modifications to take placeor the complex species-specific environment of thehuman disease is not precisely replicated in thetransgenic mice. The appraisal of therapeuticagents or protocols in these animal models must be

judged in the context of the lack of complete equiv-alence between the transgenic mouse plaques andhuman Alzheimer’s disease lesions.

108 L. Otvos, Jr.



However, perhaps there is light at the end of thetunnel. In transgenic mice overexpressing theLondon mutant of human APP, N- and C-termi-nally modified A β peptides were detected, similarto the modified A β versions in humans [55]. Theratios of deposited A β 1-42/1-40 were of the order2–3 for human and 8–9 for mouse peptides, indi-cating a preferential tendency for the deposition of the longer amyloid peptide. In protein extractsfrom soluble and insoluble brain fractions, the mostprominent peptides were truncated either at the car-boxyl- or the amino-termini yielding A β 1-38 andAβ 11-42, respectively, and the latter was stronglyenriched in the extracts of deposited peptides.These data indicate that plaques of APP-Londontransgenic mice consist of aggregates of multiple

human and mouse A β variants, possibly indeedcharacteristic for those in the brains of Alzheimer’sdisease patients.

Most recently, a similar transgenic mousemodel, named APP(SL)PS1KI, was presented [56].This transgenic mouse model carries knocked-inmutations in the presenilin-1 gene and overex-presses mutated human APP. Just like in the humancases, A β (x-42) is the major form of A β speciespresent in this model with progressive developmentof a complex pattern of N-truncated variants anddimers, similar to those observed in Alzheimer’sdisease brain. Significantly, an extensive neuronalloss (>50%) is present in the CA1/2 hippocampalpyramidal cell layer at 10 months of age togetherwith strong reactive astrogliosis. Due to the appear-ance of the critical A β variations, APP(SL)PS1KImice may provide a long-awaited tool to investi-gate therapeutic strategies designed to prevent neu-rodegeneration in Alzheimer’s disease.

7.5 N-Terminal Truncations andModifications

After so much about the modifications in general,let’s look at the variant human A β peptides indetail. We start with N-terminal modifications, fol-lowed by mid-chain alterations; finally, a brief dis-cussion of the differing fibrillogenesis by theC-terminal A β variants will be presented.

In a seminal report, A β peptides were isolatedfrom the compact amyloid cores of neuritic plaquesand separated from minor glycoprotein compo-

nents by size-exclusion high-performance liquidchromatography [57]. Parenchymal A β was shownto have a maximal length of 42 residues, but shorterforms with “ragged” amino-termini were also pres-ent. Most of the heterogeneity was found in A β 1-5 and A β 6-16 fragments, each of which eluted asfour peaks. Amino acid composition and sequenceanalyses, mass spectrometry, enzymatic methyla-tion, and stereoisomer determinations revealed thatthese multiple peptide forms resulted from struc-tural rearrangements of Asp1 and Asp7. The L-isoaspartyl form predominated at each of thesepositions, whereas the D-isoaspartyl, L-aspartyl,and D-aspartyl forms were present in lesseramounts. A β purified from the leptomeningealmicrovasculature contained the same structural

alterations as parenchymal A β, but at the C-termi-nus ended at Val40. It was suggested that the abun-dance of structurally altered aspartyl residuesaffect the conformation of the A β peptide withinplaque cores and thus significantly impact normalcatabolic processes designed to limit its deposition.

To this end, in a series of consecutive papers, wereported on the conformation-modifying effect of aspartic acid isomerization in general, and at theamino terminus of A β in particular. First we usedcircular dichroism and Fourier-transform infraredspectroscopy to characterize the conformationalchanges on human A β upon substitution of Asp1and Asp7 to isoaspartic residues [58]. We foundthat the intermolecular β-pleated sheet content ismarkedly increased for the post-translationallymodified peptide compared with that in the corre-sponding unmodified human or rodent A βsequences both in aqueous solutions in the pH7–12 range and in membrane-mimicking solvents(such as aqueous octyl- β-D-glucoside or aqueous

acetonitrile solutions). These findings underline theimportance of the originally α -helical N-terminalregions of the unmodified A β peptides in definingits secondary structure and may offer an explana-tion for the selective aggregation and retention of the isomerized A β variants in Alzheimer’s dis-ease–affected brains. For identifying the generaleffect of isoaspartic acid–bond formation on pep-tide conformation, we selected five sets of syn-thetic model peptides, each representing one of themajor secondary structures as the dominant spec-troscopically determined conformation: a type Iβ-turn, a type II β-turn, short segments of α - or




310-helices, or extended β-strands. We found thatboth types of turn structures are stabilized by theaspartic acid–bond isomerization. The isomeriza-tion at a terminal position did not affect the helixpropensity, but placing it in mid-chain broke thehelix structure [59]. Interestingly, when Asp wasalready part of a β-pleated sheet, this structure wasalso destabilized.

The physical-chemical explanation for the con-formational changes in A β upon isoAsp1 andisoAsp7 incorporation into the amino-terminaldecapeptide fragment was provided based onmolecular mechanics calculations [60]. The model-ing showed that insertion of the extra –CH 2– groupinto the decapeptide backbone results in the forma-tion of stable reverse-turns and destabilizes the hel-

ical conformer that competes with the extendedstructure at the full-sized peptide level (Fig. 7.3).The molecular modeling also revealed a limitedpropensity of the Asp1, Asp7 diisomerized peptide

to form extended structure directly. These basicfindings were later confirmed by reports from otherresearch groups. To test how changes in the aspar-tate forms influence peptide conformation, a seriesof designed peptides having the sequenceVTVKVXAVKVTV, where X represents asparticacid or its derivatives, were synthesized [61].Studies using circular dichroism showed that neu-tralization of the aspartate residue through the for-mation of a methyl ester or an amide, orreplacement of aspartate with glutamate led to anincreased β-sheet content at neutral and basic pH.A higher content of β-sheet structure correlatedwith increased propensity for fibril formation anddecreased solubility at neutral pH [61].

Anti-A β polyclonal antibody 2332 is more sen-

sitive for the non-isomerized status of the decapep-tide than that of the full-sized peptide [59].Monoclonal antibody 6E10, raised against unmod-ified A β recognizes only the unmodified decapep-tide or the peptide isomerized at the first asparticacid in a conformation-dependent manner but doesnot recognize the mid-chain isomerized or diiso-merized decapeptide in any circumstance. The di-isomerized decapeptide was used as immunogen togenerate polyclonal antibody 14943 that is notselective for the isomerized status of either the full-size peptide or the decapeptide but recognizes theisomerized peptides preferentially when the pep-tide antigen structures are conserved during theenzyme-linked immunoassay procedure [62].Owing to the poor peak shape of the full-sized A βpeptide during standard reversed-phase chromatog-raphy [63], serum stability studies that indicateextracellular stability can be more effectively per-formed on the decapeptide fragments. Remarkably,the diisomerized A β 1-10 peptide exhibits a signif-

icantly increased stability toward serum peptidasesthan the unmodified or monoisomerized peptides,suggesting a possible mechanism of the retentionof the isomerized A β peptide in the affected brains.

More contemporary techniques are able to iden-tify and quantitate the various A β forms withhigher accuracy. Although the protein is notdirectly Alzheimer’s disease related, serum amy-loid α -1 can be detected in serum as full-lengthprotein, as well as its well-characterized des-argi-nine and des-arginine/des-serine variants at the N-terminus by surface-enhanced laser desorptionionization mass spectroscopy [64]. The method is

110 L. Otvos, Jr.

FIGURE 7.3. Low-energy conformers of wild-type A β 1-10 and A β 1-10 containing isoaspartyl residues in posi-tions 1 and 7. The conformers for each subset aresuperimposed, and their peptide backbones are displayedas a line. For each conformer, the C α trace of helical orβ-turn regions are indicated by a ribbon and Asp andisoAsp residues in positions 1 and 7 by a ball and stick plot. Upper right: Type I β-turn with Glu in position i +1.Lower right: Type III β-turn with Phe at position i +1.Upper left: A β 1-10 with residues 3–9 and 5–9 posi-

tioned in a helix. Lower left: Type III β-turn with Arg inposition i +1. Reprinted from Ref. 60, with permissionfrom Blackwell Publishing.



sensitive enough to detect a low-abundant variantwith the first five N-terminal amino acids missing.Mass spectroscopy is reproducible, fast, and simplemode for the discovery and analysis of marker pro-teins of various diseases or for quality control of synthetic products.

This leads us to the quantification of the variousAβ forms in cells and tissues. We performed two-site enzyme-linked immunosorbent assay withantibodies specific for isomerized (i.e., A β with L-isoAsp at positions 1 and 7) and pGlu-modified(i.e., A β beginning with pyroglutamic acid at posi-tion 3) forms of A β to quantitate the levels of thesedifferent A β peptides in formic acid extracts of Alzheimer’s disease frontal cortex [65]. The majorspecies of A β in these samples were A β pGlu3-42

as well as A β x-42, whereas isomerized A β was aminor species. More specifically, across a panel of 14 samples, the µg/g wet tissue weight of the vari-ous A β species were as follows: A β 1-40 (1,7 di-isoAsp), 0.03; A β pGlu3-40, 0.14; A β 1-42 (1,7di-isoAsp), 0.61; A β x-40 (where x is 1 or 2), 1.66;Aβ x-42, 3.14; and A β pGlu3-42, 3.18. As seen, theforms ending with Ala42 greatly exceeded thoseending with Val40. This study was in line with anearlier report on cortical sections from 28 aged indi-viduals with a wide range in senile plaque density.According to these results, the major A β molecularspecies deposited in the brain contain PGlu3 as theN-terminal amino acid residue [66]. The abundanceof the pGlu N-terminal forms suggests that theseAβ variants can play important roles in the deposi-tion of amyloid in Alzheimer’s disease brains.

Of course, all quantitative data have to be viewedin light of the availability of the given A β analoguein the given sample. However, the hydrophobicityof the modified peptides is greatly different giving

rise to potential inaccuracy in concentration-deter-mination. After many years of trouble withreversed-phase chromatographic analysis of A βpeptides, a new protocol was developed that useshigh column temperature for optimal peak shapeand separation [67]. Coupled with mass spec-troscopy, the method is suitable for the quantifica-tion of A β isoforms in solution. Upon identicalseparation conditions, the recovery of the differentAβ species from the hydrophobic column were A β1-40, 36%; A β pGlu11-40, 34%; A β pGlu3-40,22%; and p3, 14%. It is obvious that the morehydrophobic the samples were, the lower recovery

yield was obtained. If this experiment can beextrapolated to tissue samples, there is a good pos-sibility that the total quantity of the lesshydrophilic variants is regularly underestimated.

How would the increase the pGlu3 amino-terminal forms influence the two major propertiesof Aβ, aggregation and neurotoxicity? Using circu-lar dichroism spectroscopy, it was determined thatthe pyroglutamic acid–containing peptides form β-sheet structure more readily than the correspondingfull-length A β peptides, both in aqueous solutionsand in 10% sodium dodecyl sulfate micelles [68].CD spectra taken in aqueous trifluoroethanol solu-tions indicated that the relative β-sheet to α -helicalstability is higher for the pGlu-containing peptides.The conformational differences were mirrored by

alterations in the level of precipitated A β speciesand the kinetics of the sedimentation (Fig. 7.4).According this, pGlu3 and pGlu11-N-terminal A β1-40 peptides have greater aggregation propensities


0

A β 1 -

2 8

A β 3 -

2 8

A β 1 1

- 2 8

A β 1 -

4 0

A β 3 -

4 0

A β 1 1

- 4 0

20

40

60

80

100

0

20

4060

80

100

(a)

% U

n a g g r e g a

t e d p e p

t i d e

(b)

FIGURE 7.4. Time-dependent aggregation of A β 1-28,pGlu3-28, pGlu11-28, 1-40, pGlu3-40, and pGlu11-40 ata concentration of 50 µM. Panel (A) corresponds with

studies at pH 7.2 and panel (B) with studies at pH 5.0.Reprinted from Ref. 68, with permission of theAmerican Chemical Society, Copyright 1999.



than the corresponding nonmodified peptides, withabout 4- to 5-fold reduction in the unaggregatedform at various pH and after three different incuba-tion periods. Comparison between peptides endingwith Val40 or Lys28 (the carboxy-terminal end of the extracellular domain) indicated that the greaterβ-sheet forming and aggregation propensities of the pyroglutamyl peptides are not simply due to anincrease in hydrophobicity [68]. As for the mecha-nistic explanation, it was suggested that the loss of N-terminal charges may facilitate β-sheet forma-tion by decreasing the level of unfavorable inter-strand charge repulsion, as long as the A β fibril isa hydrogen-bonded parallel β-sheet as previouslysuggested [69]. In addition, the loss of the Asp andGlu side-chain negative charges may destabilize

helix formation by eliminating favorable chargedipole interactions [70].

In another study, the toxic properties, fibrillo-genic capabilities, and in vitro degradation profileof Aβ 1-40, A β 1-42, A β pGlu3-40, and A β pGlu3-42 were compared [71]. The data show that thefiber morphology of the A β peptides is greatlyinfluenced by the C-terminus while toxicity, inter-action with cell membranes, and degradation areinfluenced by the N-terminus. A β pGlu3-40induces significantly more cell loss than the otherspecies both in neuronal and glial cell cultures. Thenumerical values are 23% decrease relative to con-trols at 0.1 µM, 31% loss at 1 µM, and 51% at 10µM, well within the range of modified A β level intissues (compare with the A β tissue concentrationsabove). Aggregated A β peptides starting withpyroglutamic acid in position 3 were heavily dis-tributed on plasma membrane and within the cyto-plasm of treated cells. The A β pGlu3-40/42peptides showed a significant resistance to degra-

dation by cultured astrocytes, while unmodifiedpeptides were partially degraded. These findingssuggest that formation of N-terminally modifiedpeptides enhance both β-amyloid aggregationand neurotoxicity, likely worsening the onset andprogression of Alzheimer’s disease.

The question arises whether the isomerized/ racemized forms are spatially and/or temporallyseparated from the unmodified A β isoform.Neuritic plaques in Alzheimer’s disease brain typi-cally immunostain with antibodies against noniso-merized A β and A β starting with pGlu3, but notAβ starting with Leu17 (p3) or Asp1 racemized

Aβ. Neuritic deposits in nondemented individualswith atherosclerotic and vascular hypertensivechanges could be identified with all three A β iso-forms [72]. The presence of A β with racemizedAsp1 in neuritic plaques in nondemented individu-als with atherosclerosis or hypertension, but not inAlzheimer’s disease, suggests a different evolutionof the plaques in the two conditions. In anotherantibody-based assay, the amino- and carboxyl-ter-minal properties of the various A β peptidesdeposited in diffuse plaques, one of the earliestforms of amyloid deposition, were examined [73].It was concluded that the amino termini of the A βspecies that initially deposit in diffuse plaquesbegin with Asp1 with or without structural modifi-cations (isomerization and racemization), as well

as with pGlu3, and terminate preferentially at A β1-42(43) rather than A β 40. This last paper wellrepresents a research trend that looks at modifica-tions in multiple positions along the A β sequence.In the end of this review, this approach will be scru-tinized in detail. Finally, here is an interestingobservation regarding the spatial relationshipbetween a 100-kDa unidentified “AMY” proteinand N-terminally modified A β peptides: AMYimmunoreactive plaques colocalized with amyloidplaques labeled by antibodies to A β starting atposition 3 with a pGlu, however AMY immunore-active deposits colocalized to a lesser degree withamyloid plaques labeled by antibodies to othervariants of the A β peptide [74] supporting the well-known finding that automatic water loss on naturaland synthetic peptides with glutamine amino ter-minus leads to massive pGlu production.

Isomerized A β variants are not restricted to theamino-terminus of the peptide. A specific antibodyrecognizing isoAsp23 of A β suggests the isomer-

ization of A β at Asp23 in vascular amyloid as wellas in the core of senile plaques [75]. The widespreadisomerization of aspartic acids in Alzheimer’s dis-ease is quite interesting, as biochemical analyses of neurofibrillary tangles also revealed L-isoaspartateat Asp193, Asn381, and Asp387 [76], indicating amodification, other than phosphorylation, that dif-ferentiates between normal tau and tau found in thepaired helical filaments of Alzheimer’s disease.Protein L-isoaspartyl methyltransferase is sug-gested to play a role in the repair of isomerized pro-teins containing L-isoAsp [77]. This enzyme isupregulated in neurodegenerative neurons and

112 L. Otvos, Jr.



colocalizes in neurofibrillary tangles [75]. Takentogether with the enhanced protein isomerization inAlzheimer’s disease brains, it is implicated thatupregulated isoaspartyl methyltransferase activitymay associate with increased protein isomerizationin Alzheimer’s disease. It needs to be added thataspartic acid isomerization occurs during syntheticglycosylation reactions of tau fragments a well,suggesting a chemical rather than enzymatic modi-fication in aged and post-translationally modifiedproteins [78]. Indeed, isomerization and racemiza-tion of aspartyl residues are often considered asproducts of spontaneous nonenzymatic reactionsthat give rise to many aspartyl forms, including L-and D-isoAsp and D-Asp [79].

7.6 Abundant Alterations atMid-Chain Positions

The appearance of isoaspartate at position 23 takesus to A β modifications in mid-chain positions.Assays with the isoAsp23-specific antibody docu-mented that A β isomerized at position 23 isdeposited on plaques and vascular amyloids [80].In vitro experiments showed that isomerization atposition 23, but not position 7, enhanced aggrega-tion. Furthermore, A β with the Dutch-type mid-chain mutation (Gln22), but not the Flemish-typemutation (Gly21), also showed greatly enhancedaggregation. These results suggest that mutationsor modifications at unmodified A β positions Glu22and Asp23 have a pathogenic role in amyloid dep-osition. The development and progression of spo-radic Alzheimer’s disease may be acceleratedby spontaneous isomerization at position 23.However, the pathological consequences of the

genetic mutation leading to the Flemish-type A βvariant need alternative explanation as the Flemishmutation fails to show potent aggregationproperties [80].

The previous study also showed that the aggre-gation rate of the Dutch-type mutation is moreextensive than that of unmodified A β in the pres-ence of Cu and Zn ions [80]. In support, in 8–28residue A β fragments, the Dutch-type mutationaccelerated fibril formation, this time around with-out metal ion addition [81]. The Gln22 Dutch,Asn23 Iowa, and Gln22, Asn23 Dutch/Iowa doublemutant A β 1-40 peptides rapidly assembled in

solution to form fibrils, whereas wild-type andGly21 Flemish A β 1-40 peptides exhibited littlefibril formation [82]. Similarly, the Dutch- andIowa-type peptides, especially the double mutantform, were found to induce robust pathologicresponses in cultured human cerebrovascularsmooth muscle cells, including elevated levels of cell-associated APP, proteolytic breakdown of smooth muscle cell α -actin, and cell death. Thesedata suggest that the different mid-chain mutationsin Aβ may exert their pathogenic effects throughdifferent mechanisms. Whereas the Gly21 Flemishmutation appears to enhance A β production, theGln22 Dutch and Asn23 Iowa mutations enhancefibrillogenesis and the pathogenicity of A β towardcultured cells. Very similar results with basically

identical conclusions were reported based on anexperiment in which the kinetics of aggregationwas followed by reversed-phase high-performanceliquid chromatography at 37˚C at pH 7.4 [83].

Using size-exclusion chromatography and cir-cular dichroism spectroscopy, kinetic and second-ary structural characteristics were compared withother A β 1-40 peptides and the extracellularAβ12-28 fragment, all having single amino acidsubstitutions in position 22 [84]. The A β 1-40Gly22 protofibrils are a group of comparativelystabile β-sheet–containing oligomers with a het-erogeneous size distribution, ranging from >100kDa to >3000 kDa. Salt promotes protofibril for-mation. When all the Glu22 substitutions werecompared, the rank order of protofibril formationof A β 1-40 and its variants was Val22 > Ala22 >Gly22 > Gln22 > Glu22 and correlated with thedegree of hydrophobicity of the substituent inposition 22. The conclusion was drawn that thephysical properties of A β 1-40 Gly22 suggest an

important role for the peptide in the neuropatho-genesis in the Arctic form of Alzheimer’s disease[84]. In support, a membrane-mimicking environ-ment generated in the presence of detergents or aganglioside is sufficient per se for amyloid fibrilformation from soluble A β and hereditary variantsof the A β peptide, including the Dutch, Flemish,and Arctic types. The peptides exhibit mutuallydifferent aggregation behavior in these environ-ments [85]. Notably, the Arctic-type A β peptide,in contrast with the wild-type and other variantforms, shows a markedly rapid and higher level of amyloid fibril formation in the presence of sodium




dodecyl sulfate or GM1 ganglioside. While in thepresence of a zwitterionic detergent, unmodifiedAβ forms 8- to 10-nm helical fibrils, and theDutch- and Flemish-type variants grow rather thin6- to 7-nm fibers. The Arctic-type A β peptideforms short and curved fibers with a diameter of 6–7 nm, and these can be defined as protofibrils(Fig. 7.5). These results underline the importanceof favorable local environments for fibrillogenesisof the amyloid peptide.

This last report surveyed additional potentialchanges in the biochemical and biophysical prop-erties of A β, brought upon mid-chain modifica-tions [85]. In addition to the more extensivelystudied aggregation properties, the possible alter-ations included the formation of more toxic

oligomeric and fibrillar A β species correspondingwith the Dutch- and Arctic-type variants [86] oralteration in sensitivities to peptidase degradation[87]. The Dutch, Flemish, Italian, and Arcticmutations apparently make A β resistant to prote-olysis by neprilysin, the peptidase with the mostimportant role in catabolism of A β in the brain.Monomeric A β wild-type, Flemish, Italian(Lys22), and Iowa variants were readily degraded

by a rat insulin-degrading enzyme, an importantcomponent of the A β clearance process [88], withsimilar efficiency. However, the proteolysis of Dutch- and Arctic-type A β variants was signifi-cantly less extensive as compared with theunmodified or the rest of the mutant peptides [89].All of the A β variants were cleaved betweenGlu3-Phe4 and Phe4-Arg5 in addition to the pre-viously described major endopeptidase sitesaround positions 13–15 and 18–21. Detergent-stable A β dimers were highly resistant to proteol-ysis regardless of the variant, suggesting that theinsulin-degrading enzyme recognizes a conforma-tion that is available for interaction only inmonomeric A β.

What are the conformational differences

between unmodified and Dutch-type A β peptides?We used Fourier-transform infrared and circulardichroism spectroscopies on synthetic peptides todemonstrate that the Glu22 → Gln mutation resultsin altered secondary structure in membrane mim-icking solvents, characterized by a considerablyhigher β-structure content for the Dutch-type pep-tide [90]. Moreover, extreme high and low pH wereless effective in eliminating the β-conformation forthe Dutch-variant than for the normal humansequence (Fig. 7.6). The differences in the strengthand stability of the aggregates are attributed to the

114 L. Otvos, Jr.

FIGURE 7.5. Electron micrographs of A β 1-40 solutionsincluding wild-, Dutch-, Flemish-, and Arctic-type vari-

ants, incubated 24 h in the presence of 0.02%Zwittergent 3-14. Reprinted from Ref. 85, with permis-sion of the International Society for Neurochemistry.

17251700

1684

D(1-42)

H(1-42)1643

1625

16751650

Wavenumber, cm −1

162516001575

FIGURE 7.6. Infrared absorbance spectra of unmodified

(broken line) and Dutch-type (solid line) A β 1-42 pep-tides in D 2O at pH 11. Reprinted from Ref. 91, with per-mission of the Society for Applied Spectroscopy.



presence of varying (small) proportions of the clas-sical secondary structures [91]. Infrared spectra of material from autopsied human Alzheimer’s dis-ease brain show spectral features indicative of theformation of similar aggregates, which may berelated to plaque formation. These results werelater confirmed by additional spectroscopic, micro-scopic, and biochemical assays [92]. According tothese, in the Dutch-type peptide the propensity of the A β N-terminal domain to adopt an α -helicalstructure is decreased, with a concomitant increasein amyloid formation. It was proposed that A βexists in an equilibrium between two species: one“able” and another “unable” to form amyloid,depending on the secondary structure adopted bythe N-terminal domain. Thus, manipulation of the

Aβ secondary structure with therapeutic com-pounds that promote the α -helical conformationmay provide a tool to control the amyloid deposi-tion observed in Alzheimer’s disease patients.

In a more recent study, the cytotoxic propertiesof the Dutch- and Italian-type (Lys22) A β variantswere compared with the unmodified peptide oncultured human cerebral endothelial cells afterflow cytometry analysis [93]. Under the conditionstested, the Dutch-type Gln22-modified analogueexhibited the highest content of β-sheet conforma-tion and the fastest aggregation/fibrillization prop-erties. The Dutch variant also induced apoptosis of cerebral endothelial cells at a concentration of 25µM, whereas the wild-type A β and the Italianmutant had no effect. The data suggest that differ-ent amino acids at position 22 confer distinct struc-tural properties to the peptides that appear toinfluence the onset and aggressiveness of the dis-ease rather than the phenotype.

7.7 C-Terminal Forms: A β 1-40and A β 1-42

One of the studies concentrating on the amino-ter-minal modifications compared the fiber types asregulated by the length of the A β peptide [71].Peptides ending with Ala42 grew to a mature fibertype regardless of the N-terminal residue, forminga dense meshwork of long fibrils by the end of theaggregation process. In contrast, A β variants end-ing with Val40 assembled more slowly to generateshort, curly fibers.

To quantitate the various A β C-terminal formspresent in the brains of patients with Alzheimer’sdisease, cerebral cortex was homogenized in 70%formic acid, and the supernatant was analyzed bysandwich enzyme-linked immunoabsorbent assaysspecific for various forms of A β [94]. In 9 of 27brains examined, there was minimal congophilicangiopathy and virtually all A β (96%) ended atAla42 (Thr43). The other 18 Alzheimer’s diseasebrains contained increasing amounts of A β endingat Val40. From this set, 6 brains with substantialcongophilic angiopathy were separately analyzed.In these brains, the amount of A β 1-42(43) wasmuch the same as in brains with minimalcongophilic angiopathy, but a large amount of A β1-40 (76% of total A β) was also present.

Immunocytochemical analysis with monoclonalantibodies selective for the various A β C-terminalforms confirmed that, in brains with minimal con-gophilic angiopathy, virtually all A β species endedat Ala42 (Thr43) and this A β variant was depositedin senile plaques of all types. In the remainingbrains, A β 1-42(43) accumulated in a similar fash-ion in plaques, but, in addition, widely varyingamounts of A β 1-40 were also deposited, primarilyin blood vessel walls. The blood vessel also con-tained some A β 1-42(43) variants. These observa-tions indicate that A β ending at Val42 (Thr43),which are a minor component of the A β in humancerebrospinal fluid and plasma, are criticallyimportant in Alzheimer’s disease where theydeposit selectively in plaques of all kinds.

A postmortem cross-sectional study comparingthe deposition of A β variants in the prefrontal cor-tex of 79 nursing home residents having no, ques-tionable, mild, moderate, or severe dementiarevealed that all three A β forms, 1-40, 1-42, and

1-43 deposited in large quantities and the A β accu-mulation level could be correlated with the severityof the dementia [95]. The deposition of A β x-42and A β x-43 occurred very early in the diseaseprocess before Alzheimer’s disease could be actu-ally diagnosed. Levels of accumulated A β x-43appeared surprisingly high given the low amountsthat are constitutively synthesized. These data indi-cate that A β x-42/43 are important species associ-ated with early disease progression and suggestthat the physiochemical properties of the A βspecies may be a major determinant in amyloiddeposition. The results support an important role




for A β in mediating initial pathogenic events inAlzheimer’s disease dementia and reinforce thattreatment strategies targeting the formation, accu-mulation, or cytotoxic effects of A β should beequally pursued.

Incubation of A β solutions at 37˚C and pH 7.4produces soluble oligomers in a concentration-dependent manner [96]. On one hand, fresh A β1-42 solutions rapidly form soluble oligomers,whereas A β 1-40 solutions require prolonged incu-bation to produce oligomeric structures. On theother hand, fresh A β 1-42 solutions are more toxicto human neuroblastoma SH-SY5Y cells than A β1-40 solutions, possibly mediated by solubleoligomers. Thus, differences in solution-phase tox-icity between A β 1-42 and A β 1-40 could explain

the association of the longer form with familialearly-onset Alzheimer’s disease.

Because A β 1-42/43 appear early in the deposi-tion process, the question was asked whether theappearance of the other A β forms is dependentupon the longest form [97]. A β ending at residuesVal40, Ala42, and Thr43 have been identified inneuritic deposits, while the peptide in vascularamyloid appears to terminate at residue Val39 orVal40. Kinetic studies of aggregation by three nat-urally occurring A β variants (1-39, 1-40, 1-42) andfour model peptides (A β 26-39, A β 26-40, A β26-42, and A β 26-43) demonstrate that amyloidformation, like crystallization, is a nucleation-dependent phenomenon [98]. The length of theC-terminus is a critical determinant of the rate of amyloid formation (“kinetic solubility”) but hasonly a minor effect on the thermodynamic solubil-ity. Amyloid formation by the kinetically solublepeptides (e.g., A β 1-39, 1-40, 26-39, or 26-40) canbe nucleated, or “seeded,” by peptides that include

the critical C-terminal residues (A β 1-42, 26-42,26-43, and 34-42). These results suggest that nucle-ation may be the rate-determining step of in vivoamyloidogenesis and confirm that A β 1-42/43,rather than A β 1-40, is the pathogenic protein(s) inAlzheimer’s disease.

All we have left is a brief survey of the environ-ment in which the various C-terminal A β variantsform. We mentioned in the beginning of this reviewthat the carboxy-terminus of A β is generallyreleased from the precursor by γ -secretase.Whether the production of all A β peptide speciesrequires the action of γ -secretase was investigated

by a combination of surface-enhanced laser des-orption/ionization time-of-flight mass spectrome-try and a specific inhibitor of γ -secretase [99].Using this approach, it was demonstrated that theproduction of all truncated A β peptides exceptthose released by the action of the non-amyloido-genic α -secretase enzyme or potentially β-site APPcleaving enzyme 2 depends on γ -secretase activity.This indicates that none of these peptides are gen-erated by a separate enzyme entity, and a specificinhibitor of the γ -secretase should have the potentialto block the generation of all amyloidogenic vari-ants. The majority of the early onset Alzheimer’sdisease cases is inherited as autosomal dominantdisorders and cosegregate with mutations in thepresenilin genes 1 and 2 [100, 101]. Mutations in

presenilin (PS) 1 and 2 were found to be causativein ≈50% of pedigrees with early-onset familiarAlzheimer’s disease [102]. It was shown that theratio of A β 1-42(43) to A β 1-40 in conditionedmedia of N2a cell lines expressing three familiarAlzheimer’s disease–linked PS-1 variants is uni-formly elevated relative to cells expressing similarlevels of wild-type PS1 [103]. Similarly, the A β 1-42(43)/A β1-40 ratio is elevated in the brains of youngtransgenic animals coexpressing a chimeric amy-loid precursor protein and a PS-1 variant comparedwith brains of transgenic mice expressing APPalone or transgenic mice coexpressing wild-typehuman PS-1 and APP. These studies provide com-pelling support for the view that one mechanism bywhich these mutant PS-1 cause Alzheimer’s dis-ease is by increasing the extracellular concentra-tion of A β peptides terminating at 42(43), speciesthat foster A β deposition.

7.8 Multiple Mutations May Pointto a Unified Picture

As all the studies cited above indicate, single A βalterations affect various properties of the wild-type peptides without a clear view of the patholog-ical consequences of the modifications. Wesuggested that some A β species feature multipleamino acid residue changes, and the coexistence of these alterations may better define the role of cer-tain changes in the deposition or neurotoxicprocesses. The first, and quite obvious, doublemodification represents the appearance of cyclized

116 L. Otvos, Jr.



Asp residues (succinimidyl) both at the amino-terminus and in the middle of the A β chain, at posi-tions 7 and 23. A potential consequence of succinimide formation is a significant increase inthe water accessibility to the backbone and α -car-bon atoms of the succinimidyl-modified Asp7 andAsp23 residues [104]. If cell toxicity of A β ismediated by soluble forms [105], this wouldexplain the increased neurotoxicity of the multiplymodified peptide. It was also suggested that spon-taneous Asp → Suc transformation might lead toan increase of the racemization rates due to thehigher accessibility of water at these sites [104].Moreover, adjacent residues may influence theselectivity of the racemization to given Aspresidues, and these residues may indirectly control

the water accessibility at the modification sites.Increased solubility influences amyloidogenic

properties of the Flemish A β variant [106].Comparative biophysical and neurotoxicity studieson wild-type and Flemish (Gly21) A β 1-40, A β5-40, and A β 11-40 revealed that the Flemishamino acid substitution increases the solubility of each form of peptide, decreases the rate of forma-tion of thioflavin-T-positive assemblies, andincreases the sodium dodecyl sulfate stability of peptide oligomers. Although the kinetics of peptideassembly are altered by the Ala21 → Gly substitu-tion, all three Flemish variants form fibrils, as dothe wild-type peptides. The N-terminally truncatedpeptides were chosen on the basis of earlier cellculture studies, which detected increased amountsof N-terminally truncated peptides secreted bycells transfected with the Flemish APP [107].Importantly, toxicity studies using cultured pri-mary rat cortical cells showed that the Flemishassemblies were as potent a neurotoxin as were the

wild-type assemblies regardless of peptide length.These results are consistent with a pathogeneticprocess in which conformational changes in A βinduced by the Gly21 form would facilitate peptideadherence to the vascular endothelium, creatingnidi for amyloid growth. Increased peptide solubil-ity and assembly stability would favor formation of larger deposits and inhibit their elimination [108].In addition, increased concentrations of neurotoxicassemblies would accelerate neuronal injury anddeath.

The effects of amino-terminal truncations on theDutch-(Gln22) and Flemish-type A β peptides were

also compared with more conclusive data on thetoxicity induced by the various N-terminal forms[109]. At a concentration of 5 µM, the aggregationof the A β peptides followed the order A β 1-42unmodified > A β 12-42 normal mid-section >A β12-42 Flemish type> A β 12-42 Dutch type. Thelower level of aggregation of the shorter peptides,especially for the Dutch variant, could be due to theformation of smaller A β fibrils, and this is in accor-dance with previous studies that observed shorterand stubbier fibrils for the Dutch version [110].Apoptosis was induced in neuronal cells by thetruncated A β wild-type and Flemish peptides atconcentrations as low as 1–5 µM, as evidenced bypropidium iodide staining, DNA laddering, andcaspase-3 activity measurements. Even when

longer incubation times and higher peptide concen-trations were applied, the N-truncated Dutch-typepeptide did not induce apoptosis. Apoptosisinduced by the full-length A β 1-42 peptide wasweaker than that induced by its N-truncated vari-ant. These data suggest that N-truncation enhancedthe cytotoxic effects of unmodified A β andFlemish-type peptides, which may play a role inthe accelerated progression of dementia. When theeffects of the modifications at different parts of theAβ peptide are compared, it can be concluded thatwhile loss of charge at Glu22 (for either Gln orAla) enhances the pathogenic effects on cere-brovascular smooth muscle cells, the N-terminalresidues in the wild-type variant confer a neuropro-tective effect, partially in agreement with earlierfindings [111].

This latter study leads us to double modifica-tions at the two termini. A β variants starting withAsp1, Phe4, Ser8, Val12, and Leu17 and endingwith Val40 or Ala42 were synthesized and their

aggregation and neurotoxic properties were com-pared [111]. The N-terminally truncated peptidesexhibited enhanced peptide aggregation relative tofull-length species, as quantitatively assessed bysedimentation analyses. The sedimentation levelswere greater for peptides terminating at residue 42than for those terminating at residue 40. Theincreased aggregation properties of the N-terminalshort and C-terminal long peptides were accompa-nied by increased β-pleated sheet conformation,fibrillar morphology under transmission electronmicroscopy, and toxicity in cultures of rat hip-pocampal neurons. Indeed, decreased level of




in vitro solubility of N-terminally truncated A βpeptides were noted earlier [112], but the negativerelationship between peptide solubility and toxicityreported here is in contrast with the positive rela-tionship of these properties as discussed at thebeginning of this section. It has to be noted thatassessing the solubility and hydrophobic propertiesof different A β variants is not easy. In 8 M urea, theotherwise α -helical or β-pleated sheet A β peptidebecomes 100% random coil and remainsmonomeric [113]. However, during electrophoresisin this medium, the peptide and its truncated vari-ants do not obey the law of mass/mobility relation-ship that most proteins—including A β peptides—follow in conventional sodium dodecyl sulfategel electrophoresis. Rather, the smaller carboxy-

terminally truncated A β 1-38 or 1-40 peptidesmigrate slower than the larger A β 1-42 full-lengthpeptide, while the amino terminally truncated A β13-42 peptide does migrate faster than the full-length A β variant. Thus, despite their small size(2–4 kDa) and minor differences between theirlengths, the A β peptides display a wide separationin this low-porosity (12% acrylamide) gel. It wasfound that this unusual electrophoretic mobility in8 M urea is due to the fact that the quantity of labeled detergent bound to the A β peptides, insteadof being proportional to the total number of aminoacids, is rather proportional to the sum of thehydrophobicity consensus indices of the con-stituent amino acids. In turn, this underlines theimportance of the total number and each individualcharged residue in the sequence in defining thethree-dimensional shape and physical relationshipwith the immediate environment.

Photo-induced cross-linking was used to evalu-ate systematically the oligomerization of 34 physi-

ologically relevant A β variants, including thosecontaining familial Alzheimer’s disease–linkedamino acid substitutions, naturally occurring N-ter-minal truncations, and modifications altering thecharge, the hydrophobicity, or the conformation of the peptide [114]. The most important structuralfeature controlling early oligomerization was thelength of the C-terminus. Specifically, the side-chain of Ile41 in A β 1-42 was found to be impor-tant both for effective formation of paranuclei andfor self-association of paranuclei into largeroligomers. The side-chain of Ala42, and the C-terminal carboxyl group, affected paranucleus -

self-association. A β 1-40 oligomerization was par-ticularly sensitive to substitutions of Glu22 orAsp23 and to truncation of the N-terminus but notto substitutions of Phe19 or Ala21. A β 1-42oligomerization, in contrast, was largely unaffectedby substitutions at positions 22 or 23 or by N-ter-minal truncations but was affected significantly bysubstitutions of Phe19 or Ala21. These resultsreveal how specific regions and residues controlAβ oligomerization and show that these controllingelements differ between diverse A β C-terminalforms.

Both mid-chain and C-terminal A β modifica-tions were made in synthetic peptides to explainthe increase of cerebral amyloid angiopathy infamiliar Alzheimer’s disease [115]. All A β 1-40

mutants at positions 22 and 23, including those cor-responding with the Dutch (Gln22), Arctic(Gly22), Italian (Lys22), and Iowa (Asn23) types,showed stronger neurotoxicity than wild-type A β1-40. Similar tendency was observed for A β 1-42mutants at positions 22 and 23 whose toxic effectswere 50–200 times stronger than that of the corre-sponding A β 1-40 variants, suggesting that theseAβ 1-42 species are the ones that are mainlyinvolved in the pathogenesis of cerebral amyloidangiopathy. While the aggregation of Arctic- andIowa-type A β 1-42 was similar to that of wild-typeAβ 1-42, Gln22- and Lys22-containing A β pep-tides aggregated extensively, supporting the clini-cal evidence that Dutch and Italian patients arediagnosed as hereditary cerebral hemorrhage withamyloidosis. In contrast, the Flemish Gly21 muta-tion needs alternative explanation with theexception of altered physicochemical properties.Although attenuated total reflection–Fourier trans-form infrared spectroscopy spectra suggested that

the β-pleated sheet content correlated with A βaggregation, the enhanced β-turn around positions22 and 23 in the mutated versions also enhancedthe aggregative ability [115].

A noteworthy feature of the last report is theexceptional purity of the synthetic A β peptides,supported by mass spectroscopy data. It had previ-ously been reported that Gln22 A β 1-40 rather thanGln22 A β 1-42 plays a significant role in Dutch-type cerebral amyloid angiopathy because theDutch-type A β 1-42 did not show any cytotoxiceffects [116]. However, the newer report clearlydemonstrates the most potent cytotoxicity of Gln22

118 L. Otvos, Jr.



Aβ 1-42 among all the A β 1-42 variants. In addi-tion, in the newer paper, wild-type A β 1-42 aggre-gated far more rapidly than wild-type A β 1-40,differing from earlier data published by othergroups [47, 117]. Potentially novel and reliablesynthetic methods of pure A β 1-42 peptides [118,119] allowed more reliable measurements. If this isindeed true, the varying purity levels of syntheticAβ peptide preparations might be one of the majorreasons of the discrepancies in the biological data.

7.9 Conclusions

It is an undeniable fact that different A β variantspopulate the tissues in different amyloid diseases

and the N-terminal, mid-chain, or C-terminal mod-ifications are likely to contribute to the develop-ment of a given clinical phenotype. Due to the lack of naturally occurring material in quantities largeenough for detailed biochemical, biophysical, andbiological analysis, synthetic peptides correspon-ding with the isolated A β forms are prepared, andthe potential role of the modifications in the patho-genesis of the disease, mostly Alzheimer’s disease,is investigated on these synthetic products. In gen-eral, A β mutations enhance both typical propertiesof the amyloid peptide: fibrillogenesis and neuro-toxicity. The first is quite understandable becausedeletion of the amino-terminal hydrophilicresidues, addition of two carboxy-terminalhydrophobic residues, or elimination of chargedside-chains in mid-chain positions all likely con-tribute to the reduction of the α -helical conformerand to an increased β-pleated sheet formation aswell as aggregation. Less clear is the effect of thechanges on cell toxicity, especially as contrasting

views are present on the requirement for neurotoxicproperties. Peptide solubility is certainly one fac-tor, and while most modifications are expected todecrease aqueous solubility, N-terminal cyclizationof aspartyl residues actually increases it. Moreover,toxic properties associated with interactions withthe cell membrane or other hydrophobic cell-origi-nated components may play a role in the ability of the modified A β variants to disrupt cellular func-tions.

The modified A β forms are partly due to post-translational processing of the unmodified peptide;however, the mutations themselves may lend to

decreased sensitivity to further proteolytic degra-dation hence delayed turnover. One aspect is cer-tain: The A β peptide is a very difficult compoundto prepare and purify, and the purity of the syn-thetic products (and we are usually dealing withsingle amino acid mutations) can significantlyinfluence the results of comparative biologicalassays. A β peptides are notorious for irregularbehavior during chromatography or other separa-tion techniques, and single amino acid modifica-tions, often of charged residues as they are presentin the Dutch-, Italian-, Arctic-, or Iowa-type A βvariants, may dramatically change the physicalbehavior of the peptide and this reflects incontroversial biochemical data.

The development of reliable and reproducible

synthetic, separation, and analytical A β protocols aswell as the refinement of characteristic assays forfibrillogenesis and cell toxicity will allow the viewson the effects of the various A β forms to unify andprovide clues for molecular or cellular therapeuticinterventions to eliminate the pathogenic A βspecies.

References1. Klucken, J., McLean, P.J., Gomez-Tortosa, E., et al.

Neuritic alterations and neural system dysfunction inAlzheimer’s disease and dementia with Lewy bodies.Neurochem. Res. 2003; 28:1683-1691.

2. Ghiso, J., and Frangione, B. Amyloidosis andAlzheimer’s disease. Adv. Drug Deliv. Rev. 2002;54:1539-1551.

3. Giasson, B.I., Lee, V.M.Y., and Trojanowski, J.Q.Interactions of amyloidogenic proteins. Neuromol.Med. 2003; 4:49-58.

4. Knauer, M.F., Soreghan, B., Burdick, D., et al.Intracellular accumulation and resistance to degrada-

tion of the Alzheimer amyloid A4/ β protein. Proc.Natl. Acad. Sci. U. S. A. 1992; 89:7437-7441.

5. Tabaton, M., Nunzi, M.G., Xue, R., et al. Solubleamyloid β-protein is a marker of Alzheimer amyloidin brain but not in cerebrospinal fluid. Biochem.Biophys. Res. Commun. 1994; 200:1598-1603.

6. Selkoe, D.J. Alzheimer’s disease: genes, proteins,and therapy. Physiol. Rev. 2001; 81:741-766.

7. Weggen, S., Eriksen, J.L., Das, P., et al. A subset of NSAIDs lower amyloidogenic A β42 independentlyof cyclooxygenase activity. Nature 2001; 414:212-

216.8. Yankner, B.A. Mechanisms of neuronal degenerationin Alzheimer’s disease. Neuron 1996; 16:921-932.




9. Southwick, P.C., Yamagata, S.K., Echols, C.L., et al.Assessment of amyloid β-protein in cerebrospinalfluid as an aid in the diagnosis of Alzheimer’s dis-ease. J. Neurochem. 1996; 66:259-265.

10. Wang, R., Sweeney, D., Gandy, S.E., and Sisodia,S.S. The profile of soluble amyloid β protein in cul-

tured cell media. J. Biol. Chem. 1996; 271:31894-31902.

11. Glenner, G.G., and Wong, C.W. Alzheimer’s disease:initial report of the purification and characterizationof a novel cerebrovascular amyloid protein.Biochem. Biophys. Res. Commun. 1984; 120:885-890.

12. Golde, T.E., Eckman, C.B., and Younkin, S.G.Biochemical detection of A β isoforms: implicationsfor pathogenesis, diagnosis, and treatment of Alzheimer’s disease. Biochim. Biophys. Acta 2000;1502:172-187.

13. Gorman, P.M., and Chakrabartty, A. Alzheimer β-amyloid peptides: structures of amyloid fibrils andalternate aggregation products. Biopolymers 2001;60:381-394.

14. Paradisi, S., Sacchetti, B., Balduzzi, M., et al.Astrocyte modulation of in vitro β-amyloid neuro-toxicity. Glia 2004; 46:252-260.

15. Butterfield, D.A. Amyloid β-peptide (1-42)-inducedoxidative stress and neurotoxicity: implications forneurodegeneration in Alzheimer’s disease brain. Areview. Free Radic. Res. 2002; 36:1307-1313.

16. Harper, J.D., and Lansbury, P.T. Jr. Models of amy-loid seeding in Alzheimer’s disease and scrapie:mechanistic truths and physiological consequencesof the time-dependent solubility of amyloid proteins.Annu. Rev. Biochem. 1997; 66:385-407.

17. Barrow, C.J., and Zagorski, M.G. Solution structuresof β peptide and its constituent fragments: relation toamyloid deposition. Science 1991; 253:179-182.

18. Watson, A.A., Fairlie, D.P., and Craik, D.J. Solutionstructure of methionine-oxidized amyloid β-peptide(1-40). Does oxidation affect conformational switch-

ing? Biochemistry 1998; 37:12700-12706.19. Meda, L., Cassatella, M.A., Szendrei, G.I., et al.Activation of microglial cells by β-amyloid proteinand interferon- γ . Nature 1995; 374:647-650.

20. Fonseca, M.I., Head, E., Velazquez, P., et al. The pres-ence of isoaspartic acid in β-amyloid plaques indi-cates plaque age. Exp. Neurol. 1999; 157:277-288.

21. Shoji, M., Golde, T.E., Ghiso, J., et al. Production of the Alzheimer amyloid β-protein by normal prote-olytic processing. Science 1992; 258:126-129.

22. Goldgaber, D., Lerman, M.I., McBride, O.W., et al.Characterization and chromosomal localization of acDNA encoding brain amyloid of Alzheimer’s dis-ease. Science 1987; 235:877-880.

23. Busciglio, J., Gabuzda, D.H., Matsudaira, P., andYankner, B.A. Generation of β-amyloid in thesecretory pathway in neuronal and nonneuronalcells. Proc. Natl. Acad. Sci. U. S. A. 1993; 90:2092-2096.

24. Seubert, P., Oltersdorf, T., Lee, M.G., et al. Secretion

of β-amyloid precursor protein cleaved at the aminoterminus of the β-amyloid peptide. Nature 1993;361:260-263.

25. Suzuki, N., Cheung, T.T., Cai, X.D., et al. Anincreased percentage of long amyloid β-proteinsecreted by familial amyloid β-protein precursor(βAPP717) mutants. Science 1994; 264:1336-1340.

26. Goate, A., Chartier-Harlin, M.C., Mullan, M., et al.Segregation of a missense mutation in the amyloidprecursor protein gene with familial Alzheimer’s dis-ease. Nature 1991; 349:704-706.

27. Murrell, J. Farlow, M., Ghetti, B., and Benson, M.D.A mutation in the amyloid precursor protein associ-ated with hereditary Alzheimer’s disease. Science1991; 254:97-99.

28. Chartier-Harlin, M.C., Crawford, F., Houlden, H.,et al. Early-onset Alzheimer’s disease caused bymutations at codon 717 of the β-amyloid precursorprotein gene. Nature 1991; 353:844-846.

29. Medeiros, M.S., and Turner, A.J. Post-secretory pro-cessing of regulatory peptides: the pancreaticpolypeptide family as a model example. Biochimie1994; 76:283-287.

30. Lewis, P.A., Piper, S., Baker, M., et al. Expressionof BRI-amyloid β peptide fusion proteins: a novelmethod for specific high-level expression of amy-loid β peptides. Biochim. Biophys. Acta 2001;1537:58-62.

31. Hooper, N.M., and Turner, A.J. Protein processingmechanisms: from angiotensin-converting enzyme toAlzheimer’s disease. Biochem. Soc. Trans. 2000; 28:441-446.

32. Matsumoto, A., Itoh, K., and Matsumoto, R. A novelcarboxypeptidase B that processes native β-amyloid

precursor protein is present in human hippocampus.Eur. J. Neurosci. 1990; 12:227-238.33. Wu, Q., Li, L., Cooper, M.D., et al. Aminopeptidase

A activity of the murine B-lymphocyte differentia-tion antigen BP-1/6C3. Proc. Natl. Acad. Sci. U. S.A. 1991; 88:676-680.

34. Martinez, J.M., Prieto, I., Ramirez, M.J., et al. Sexdifferences and age-related changes in human serumaminopeptidase A activity. Clin. Chim. Acta 1998;274:53-61.

35. Arechaga, G., Sanchez, B., Alba, F., et al.Developmental changes of soluble and membrane-bound aspartate aminopeptidase activities in ratbrain. Rev. Esp. Fisiol. 1996; 52:149-154.

120 L. Otvos, Jr.



36. Orpiszewski, J., Schormann, N., Kluve-Beckerman,B., et al. Protein aging hypothesis of Alzheimer’sdisease. FASEB J. 2000; 14:1255-1263.

37. Greenfield, J.P., Tsai, J., Gouras, G.K., et al.Endoplasmic reticulum and trans-Golgi network generate distinct populations of Alzheimer β-amy-

loid peptides. Proc. Natl. Acad. Sci. U. S. A. 1999;96:742-747.

38. Prokai, L., Zharikova, A.D., Janaky, T., and Prokai-Tatrai, K. Exploratory pharmacokinetics and braindistribution study of a neuropeptide FF antagonist byliquid chromatography/atmospheric pressure ioniza-tion tandem mass spectrometry. Rapid Commun.Mass Spectrom. 2000; 14:2412-2418.

39. Wurth, C., Guimard, N.K., and Hecht, M.H.Mutations that reduce aggregation of the Alzheimer’sAβ42 peptide: an unbiased search for the sequencedeterminants of A β amyloidogenesis. J. Mol. Biol.2002; 319:1279-1290.

40. Janus, C., Chishti, M.A., and Westaway, D.Transgenic mouse models of Alzheimer’s disease.Biochim. Biophys. Acta 2000; 1502:63-75.

41. Selkoe, D.J. Biochemistry of altered brain proteins inAlzheimer’s disease. Annu. Rev. Neurosci. 1989; 12:463-490.

42. Johnstone, E.M., Chaney, M.O., Norris, F.H., et al.Conservation of the sequence of the Alzheimer’s dis-ease amyloid peptide in dog, polar bear and five othermammals by cross-species polymerase chain reac-

tion analysis. Brain Res. Mol. Brain Res. 1991; 10:299-305.

43. Yamada, T., Sasaki, H., Furuya, H., et al.Complementary DNA for the mouse homolog of thehuman amyloid β protein precursor. Biochem.Biophys. Res. Commun. 1987; 149:665-671.

44. Shivers, B.D., Hilbich, C., Multhaup, G., et al.Alzheimer’s disease amyloidogenic glycoprotein:expression pattern in rat brain suggests a role in cellcontact. EMBO J. 1988; 7:1365-1370.

45. Otvos, L. Jr., Szendrei, G.I., Lee, V.M., and Mantsch,

H.H. Human and rodent Alzheimer β-amyloid peptidesacquire distinct conformations in membrane-mimick-ing solvents. Eur. J. Biochem. 1993; 211:249-257.

46. Fraser, P.E., Nguyen, J.T., Inouye, H., et al. Fibrilformation by primate, rodent, and Dutch-hemor-rhagic analogues of Alzheimer amyloid β-protein.Biochemistry 1992; 31:10716-10723.

47. Shin, R.W., Ogino, K., Kondo, A., et al. Amyloid β-pro-tein (A β) 1-40 but not A β1-42 contributes to the exper-imental formation of Alzheimer’s disease amyloidfibrils in rat brain. J. Neurosci. 1997; 17:8187-8193.

48. Huang, X., Cuajungco, M.P., Atwood, C.S., et al.Alzheimer’s disease, β-amyloid protein and zinc.J. Nutr. 2000; 130:1488S-1492S.

49. Harris, J.R. In vitro fibrillogenesis of the amyloid β1-42 peptide: cholesterol potentiation and aspirininhibition. Micron 2002; 33:609-626.

50. Atwood, C.S., Perry, G., Zeng, H., et al. Coppermediates dityrosine cross-linking of Alzheimer’samyloid- β. Biochemistry 1998; 43:560-568.

51. Galeazzi, L., Ronchi, P., Franceschi, C., and Giunta,S. In vitro peroxidase oxidation induces stabledimers of β-amyloid (1-42) through dityrosine bridgeformation. Amyloid 1999; 6:7-13.

52. Kowalik-Jankowska, T., Ruta-Dolejsz, M.,Wisniewska, K., and Lankiewicz, L. Cu(II) interac-tion with N-terminal fragments of human and mouseβ-amyloid peptide. J. Inorg. Biochem. 2001; 86:535-545.

53. Bishop, G.M., and Robinson, S.R. Human A β1-42reduces iron-induced toxicity in rat cerebral cortex.J. Neurosci. Res. 2003; 73:316-323.

54. Kuo, Y.M., Kokjohn, T.A., Beach, T.G., et al.Comparative analysis of amyloid- β chemical struc-ture and amyloid plaque morphology of transgenicmouse and Alzheimer’s disease brains. J. Biol.Chem. 2001; 276:12991-12998.

55. Pype, S., Moechars, D., Dillen, L., and Mercken, M.Characterization of amyloid β peptides from brainextracts of transgenic mice overexpressing theLondon mutant of human amyloid precursor protein.J. Neurochem. 2003; 84:602-609.

56. Casas, C., Sergeant, N., Itier, J.M., et al. Massive

CA1/2 neuronal loss with intraneuronal and N-termi-nal truncated A β42 accumulation in a novelAlzheimer transgenic model. Am. J. Pathol. 2004;165:1289-1300.

57. Roher, A.E., Lowenson, J.D., Clarke, S., et al.Structural alterations in the peptide backbone of β-amyloid core protein may account for its depositionand stability in Alzheimer’s disease. J. Biol. Chem.1993; 268:3072-3083.

58. Fabian, H., Szendrei, G.I., Mantsch, H.H., et al.Synthetic post-translationally modified human A β

peptide exhibits a markedly increased tendency toform β-pleated sheets in vitro. Eur. J. Biochem.1994; 221:959-964.

59. Szendrei, G.I., Fabian, H., Mantsch, H.H., et al.Aspartate-bond isomerization affects the major con-formations of synthetic peptides. Eur. J. Biochem.1994; 226:917-924.

60. Szendrei, G.I., Prammer, K.V., Vasko, M., et al. Theeffects of aspartic acid-bond isomerization onin vitro properties of the amyloid β-peptide as mod-eled with N-terminal decapeptide fragments. Int.J. Pept. Protein Res. 1997; 47:289-296.

61. Orpiszewski, J., and Benson, M.D. Induction of -sheet structure in amyloidogenic peptides by




neutralization of aspartate: a model for β-amyloidnucleation. J. Mol. Biol.1999; 289:413-428.

62. Lang, E., Szendrei, G.I., Lee, V.M., and Otvos, L. Jr.Spectroscopic evidence that monoclonal antibodiesrecognize the dominant conformation of medium-sized synthetic peptides. J. Immunol. Methods 1994;

170:103-115.63. Barrow, C.J., Yasuda, A., Kenny, P.T., and Zagorski,

M.G. Solution conformations and aggregationalproperties of synthetic amyloid β-peptides of Alzheimer’s disease. Analysis of circular dichroismspectra. J. Mol. Biol. 1992; 225:1075-1093.

64. Tolson, J., Bogumil, R., Brunst, E., et al. Serum pro-tein profiling by SELDI mass spectrometry: detec-tion of multiple variants of serum amyloid α in renalcancer patients. Lab. Invest. 2004; 84:845-856.

65. Hosoda, R., Saido, T.C., Otvos, L. Jr., et al.Quantification of modified amyloid β peptides inAlzheimer’s disease and Down syndrome brains.J. Neuropathol. Exp. Neurol. 1998; 57:1089-1095.

66. Saido, T.C., Iwatsubo, T., Mann, D.M., et al..Dominant and differential deposition of distinct β-amyloid peptide species, A β N3(pE), in senileplaques. Neuron 1995; 14:457-466.

67. Thompson, A.J., Lim, T.K., and Barrow, C.J. On-linehigh-performance liquid chromatography/mass spec-trometric investigation of amyloid- β peptide variantsfound in Alzheimer’s disease. Rapid Commun. Mass.Spectrom. 1999; 13:2348-2351.

68. He, W., and Barrow, C.J. The A β 3-pyroglutamyl and11-pyroglutamyl peptides found in senile plaquehave greater β-sheet forming and aggregationpropensities in vitro than full-length A β .Biochemistry 1999; 38:10871-10877.

69. Benzinger, T.L., Gregory, D.M., Burkoth, T.S., et al.Propagating structure of Alzheimer’s β-amyloid (10-35) is parallel β-sheet with residues in exact register.Proc. Natl. Acad. Sci. U. S. A. 1998; 95:13407-13412.

70. Houston, M.E. Jr., Campbell, A.P., Lix, B., et al.

Lactam bridge stabilization of α -helices: the role of hydrophobicity in controlling dimeric versusmonomeric α -helices. Biochemistry 1996; 35:10041-10050.

71. Russo, C., Violani, E., Salis, S., et al. Pyroglutamate-modified amyloid β-peptides -A βN3(pE) -stronglyaffect cultured neuron and astrocyte survival.J. Neurochem. 2002; 82:1480-1489.

72. Tekirian, T.L., Saido, T.C., Markesbery, W.R., et al.N-terminal heterogeneity of parenchymal and cere-brovascular A β deposits. J. Neuropathol. Exp.Neurol. 1998; 57:76-94.

73. Iwatsubo, T., Saido, T.C., Mann, D.M., et al. Full-length amyloid- β (1-42(43)) and amino-terminally

modified and truncated amyloid- β 42(43) deposit indiffuse plaques. Am. J. Pathol. 1996; 149:1823-1830.

74. Schmidt, M.L., Saido, T.C., Lee, V.M., andTrojanowski, J.Q. Spatial relationship of AMY pro-tein deposits and different species of A β peptides inamyloid plaques of the Alzheimer’s disease brain.

J. Neuropathol. Exp. Neurol. 1999; 58:1227-1233.75. Shimizu, T., Watanabe, A., Ogawara, M., et al.

Isoaspartate formation and neurodegeneration inAlzheimer’s disease. Arch. Biochem. Biophys. 2000;381:225-234.

76. Watanabe, A., Takio, K., and Ihara, Y. Deamidationand isoaspartate formation in smeared τ in pairedhelical filaments. Unusual properties of the micro-tubule-binding domain of τ. J. Biol. Chem. 1999;274:7368-7378.

77. Clarke, S. Protein carboxyl methyltransferases: twodistinct classes of enzymes. Annu. Rev. Biochem.1985; 54:479-506.

78. Hoffmann, R., Craik, D.J., Bokonyi, K., et al. Highlevel of aspartic acid-bond isomerization during thesynthesis of an N-linked tau glycopeptide. J. Pept.Sci. 1999; 5:442-456.

79. Geiger, T., and Clarke, S. Deamidation, isomeriza-tion, and racemization at asparaginyl and aspartylresidues in peptides. Succinimide-linked reactionsthat contribute to protein degradation. J. Biol. Chem.1987; 262:785-794.

80. Shimizu, T., Fukuda, H., Murayama, S., et al.

Isoaspartate formation at position 23 of amyloid βpeptide enhanced fibril formation and deposited ontosenile plaques and vascular amyloids in Alzheimer’sdisease. J. Neurosci. Res. 2002; 70:451-461.

81. Wisniewski, T., Ghiso, J., and Frangione, B. Peptideshomologous to the amyloid protein of Alzheimer’sdisease containing a glutamine for glutamic acid sub-stitution have accelerated amyloid fibril formation.Biochem. Biophys. Res. Commun. 1991; 179:1247-1254.

82. Van Nostrand, W.E., Melchor, J.P., Cho, H.S., et al.

Pathogenic effects of D23N Iowa mutant amyloid β-protein. J. Biol. Chem. 2001; 276:32860-32866.83. Clements, A., Walsh, D.M., Williams, C.H., and

Allsop, D. Effects of the mutations Glu22 to Gln andAla21 to Gly on the aggregation of a synthetic frag-ment of the Alzheimer’s amyloid β /A4 peptide.Neurosci. Lett. 1993; 161:17-20.

84. Paivio, A., Jarvet, J., Graslund, A., et al. Uniquephysicochemical profile of β-amyloid peptide variantAβ1-40E22G protofibrils: conceivable neu-ropathogen in arctic mutant carriers. J. Mol. Biol.2004; 339:145-159.

85. Yamamoto, N., Hasegawa, K., Matsuzaki, K., et al.Environment- and mutation-dependent aggregation

122 L. Otvos, Jr.



behavior of Alzheimer amyloid β-protein.J. Neurochem. 2004; 90:62-69.

86. Dahlgren, K.N., Manelli, A.M., Stine, W.B., et al.Oligomeric and fibrillar species of amyloid- β pep-tides differentially affect neuronal viability. J. Biol.Chem. 2002; 277:32046-32053.

87. Tsubuki, S., Takaki, Y., and Saido, T.C. Dutch,Flemish, Italian, and Arctic mutations of APP andresistance of A β to physiologically relevant prote-olytic degradation. Lancet 2003; 361:1957-1958.

88. Eckman, E.A., Reed, D.K., and Eckman, C.B.Degradation of the Alzheimer’s amyloid β peptide byendothelin-converting enzyme. J. Biol. Chem. 2001;276:24540-24548.

89. Morelli, L., Llovera, R., Gonzalez, S.A., et al.Differential degradation of amyloid β genetic vari-ants associated with hereditary dementia or stroke byinsulin-degrading enzyme. J. Biol. Chem. 2003; 278:23221-23226.

90. Fabian, H., Szendrei, G.I., Mantsch, H.H., and Otvos,L., Jr. Comparative analysis of human and Dutch-typeAlzheimer β-amyloid peptides by infrared spec-troscopy and circular dichroism. Biochem. Biophys.Res. Commun. 1993; 191:232-239.

91. Fabian, H., Choo, L-P., Szendrei, G.I., et al. Infraredspectroscopic characterization of Alzheimer’splaques. Applied Spectrosc. 1993; 47:1513-1518.

92. Soto, C., Castano, E.M., Frangione, B., andInestrosa, N.C. The α -helical to β-strand transition in

the amino-terminal fragment of the amyloid β-pep-tide modulates amyloid formation. J. Biol. Chem.1995; 270:3063-3067.

93. Miravalle, L., Tokuda, T., Chiarle, R., et al.Substitutions at codon 22 of Alzheimer’s A β peptideinduce diverse conformational changes and apoptoticeffects in human cerebral endothelial cells. J. Biol.Chem. 2000; 275:27110-27116.

94. Gravina, S.A., Ho, L., Eckman, C.B., et al.Amyloid β protein (A β) in Alzheimer’s diseasebrain. Biochemical and immunocytochemical

analysis with antibodies specific for forms endingat A β 40 or A β 42(43). J. Biol. Chem. 1995; 270:7013-7016.

95. Parvathy, S., Davies, P., Haroutunian, V., et al.Correlation between A βx-40-, A βx-42-, and A βx-43-containing amyloid plaques and cognitivedecline. Arch. Neurol. 2001; 58:2025-2032.

96. El-Agnaf, O.M., Mahil, D.S., Patel, B.P., and Austen,B.M. Oligomerization and toxicity of β-amyloid-42implicated in Alzheimer’s disease. Biochem.Biophys. Res. Commun. 2000; 273:1003-1007.

97. Jarrett, J.T., Berger, E.P., and Lansbury, P.T., Jr. TheC-terminus of the β protein is critical in amyloidoge-nesis. Ann. N. Y. Acad. Sci. 1993; 695:144-148.

98. Jarrett, J.T., Berger, E.P, and Lansbury, P.T., Jr. Thecarboxy terminus of the β amyloid protein is criticalfor the seeding of amyloid formation: implicationsfor the pathogenesis of Alzheimer’s disease.Biochemistry 1993; 32:4693-4697.

99. Beher, D., Wrigley, J.D., Owens, A.P., and

Shearman, M.S. Generation of C-terminally trun-cated amyloid- β peptides is dependent on γ -secre-tase activity. J. Neurochem. 2002; 82:563-575.

100. St George-Hyslop, P., Haines, J., Rogaev, E., et al.Genetic evidence for a novel familial Alzheimer’sdisease locus on chromosome 14. Nat. Genet. 1992;2:330-334.

101. Levy-Lahad, E., Wasco, W., Poorkaj, P., et al.Candidate gene for the chromosome 1 familialAlzheimer’s disease locus. Science 1995; 269:973-977.

102. Schellenberg, G.D. Genetic dissection of Alzheimer’s disease, a heterogeneous disorder.Proc. Natl. Acad. Sci. U. S. A. 1995; 92:8552-8559.

103. Borchelt, D.R., Thinakaran, G., Eckman, C.B., et al.Familial Alzheimer’s disease-linked presenilin 1variants elevate A β1-42/1-40 ratio in vitro andin vivo. Neuron 1996; 17:1005-1013.

104. Lins, R.D., Soares, T.A., Ferreira, R., and Longo,R.L. Solvent accessibility to aspartyl and succin-imidyl residues at positions 7 and 23 in the amyloidβ 1-28 peptide. Z. Naturforsch. 1999; 54: 264-270.

105. Gentile, M.T., Vecchione, C., Maffei, A., et al.

Mechanisms of soluble β-amyloid impairment of endothelial function. J. Biol. Chem. 2004; 279:48135-48142.

106. Walsh, D.M., Hartley, D.M., Condron, M.M., et al.In vitro studies of amyloid β-protein fibril assemblyand toxicity provide clues to the aetiology of Flemish variant (Ala692 → Gly) Alzheimer’s dis-ease. Biochem. J. 2001; 355:869-877.

107. Haass, C., Hung, A.Y., Selkoe, D.J., and Teplow,D.B. Mutations associated with a locus for familialAlzheimer’s disease result in alternative processing

of amyloid β-protein precursor. J. Biol. Chem.1994; 269:17741-17748.108. Clements, A., Allsop, D., Walsh, D.M., and

Williams, C.H. Aggregation and metal-bindingproperties of mutant forms of the amyloid A β pep-tide of Alzheimer’s disease. J. Neurochem. 1996;66:740-747.

109. Demeester, N., Mertens, C., Caster, H., et al.Comparison of the aggregation properties, second-ary structure and apoptotic effects of wild-type,Flemish and Dutch N-terminally truncated amyloidβ peptides. Eur. J. Neurosci. 2001; 13:2015-2024.

110. Sian, A.K., Frears, E.R., El-Agnaf, O.M., et al.Oligomerization of β-amyloid of the Alzheimer’s




and the Dutch-cerebral-haemorrhage types.Biochem. J. 2000; 349:299-308.

111. Pike, C.J., Overman, M.J., and Cotman, C.W.Amino-terminal deletions enhance aggregation of β-amyloid peptides in vitro. J. Biol. Chem. 1995;270:23895-23898.

112. Hilbich, C., Kisters-Woike, B., Reed, J., et al.Aggregation and secondary structure of syntheticamyloid β A4 peptides of Alzheimer’s disease.J. Mol. Biol. 1991; 218:149-163.

113. Kawooya, J.K., Emmons, T.L., Gonzalez-DeWhitt,P.A., et al. Electrophoretic mobility of Alzheimer’samyloid- β peptides in urea-sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem.2003; 323:103-113.

114. Bitan, G., Vollers, S.S., and Teplow, D.B.Elucidation of primary structure elements control-ling early amyloid β-protein oligomerization.J. Biol. Chem. 2003; 278:34882-34889.

115. Murakami, K., Irie, K., Morimoto, A., et al.Neurotoxicity and physicochemical properties of

Aβ mutant peptides from cerebral amyloid angiopa-thy: implication for the pathogenesis of cerebralamyloid angiopathy and Alzheimer’s disease.J. Biol. Chem. 2003; 278:46179-46187.

116. Davis, J., and van Nostrand, W.E. Enhanced patho-logic properties of Dutch-type mutant amyloid β-

protein. Proc. Natl. Acad. Sci. U. S. A. 1996; 93:2996-3000.

117. Kirkitadze, M.D., Condron, M.M., and Teplow,D.B. Identification and characterization of keykinetic intermediates in amyloid β-protein fibrillo-genesis. J. Mol. Biol. 2001; 312:1103-1119.

118. Fukuda, H., Shimizu, T., Nakajima, M., et al.Synthesis, aggregation, and neurotoxicity of theAlzheimer’s A β1-42 amyloid peptide and its isoas-partyl isomers. Bioorg. Med. Chem. Lett. 1999; 9:953-956.

119. Tickler, A.K., Barrow, C.J., and Wade, J.D.Improved preparation of amyloid- β peptides usingDBU as N α -Fmoc deprotection reagent. J. Pept.Sci. 2001; 7: 488-494.

124 L. Otvos, Jr.



8Copper Coordination byβ-Amyloid

and the Neuropathology of Alzheimer’sDiseaseCyril C. Curtain and Kevin J. Barnham

125

8.1 IntroductionIt is nearly two decades since high concentrationsof the redox active transition metal ions Cu2+ andFe3+ found inβ-amyloid plaques were first pro-posed to play an important role in the pathology of Alzheimer’s disease (AD) (see review by Bush [1]).Over this time, a new field of metallo-neurobiologyrelating to AD and other neurodegenerative dis-eases has arisen with approximately 250 originalpapers and more than 1000 references in secondarypublications to date. At first, many neuroscientistsfailed to recognize the importance of this growingliterature. However, a recent pilot Phase II clinicaltrial of a blood-brain barrier permeable metal pro-tein attenuating compound (MPAC), clioquinol, inpatients with moderately severe AD has shownpromising results [2]. In a randomized sample of 36 subjects, the effect of treatment was significantin the more severely affected group, where thosetreated with clioquinol showed minimal deteriora-tion in their cognitive scores (Alzheimer’s disease

Assessment Scale≥25) compared with substantialworsening of the scores for the placebo group.Although subjected to the usual cautions applied tosmall-scale trials, this is an encouraging result thatrenders even more urgent the full elucidation of thepossible role of transition metals, particularly Cuand Zn, in AD. It must be stressed that, althoughthere is much experimental evidence on variousaspects of the interaction between Cu, Zn, and theconstituent of the amyloid plaques, theβ-amyloidpeptide (Aβ), the structural biology and elucidationof the neuropathological significance of metalbinding are very much works in progress.

The naturally occurringΑβ1–42, 1–41 and 1–39

peptides (sequence of Aβ1–42 given in Fig. 8.1)represent part of the putative trans-membranedomain of the amyloid precursor protein, liberatedfrom the membrane by proteolytic (secretase)action. Although its sequence is generally highlyconserved, the rat sequence has Arg5, Tyr10, andHis13 of human Aβ replaced by Gly5, Phe10, andArg13 (see highlighted residues in Fig. 8.1).Because the murine species do not develop amy-loid plaques in the brain with aging, it was recog-nized that these substitutions could be an importantpointer to mechanisms of plaque formation inhuman beings. The coordination of transition met-als by Aβ has been linked variously to their role inpromoting peptide aggregation to form amyloidplaques, in the production of cytotoxic reactiveoxygen species (ROS), and in promoting poten-tially cytotoxic interactions with cell membranes.

8.2 Cu2+ and Zn2+ Induced

Aggregation of AβTransition metal ion homeostasis is severely dys-regulated in the AD brain [3, 4] and the role of these metals has been the subject of continuingstudy [5–11]. The transition metal ions Cu2+, Fe3+,and Zn2+ have been reported to occur at high con-centrations in the neocortical parenchyma of healthy brain (total dry weight concentrations of 70, 340, and 350µM, respectively). These concen-trations may seem high but are not surprising whenone considers the intense bioenergetics of the brainand the fact that the transition metal ions are an



essential part of the redox systems involved. Theirlevels are far higher in the neuropil of the AD-affected brain, where they reach 0.4 and 1.0 mMfor Cu and Fe/Zn, respectively in the amyloidplaque deposits [12]. It is of interest that these havebeen termed “trace metals,” an evident misnomerbecause their concentrations in the gray matterare of the same order of magnitude as Mg(0.1–0.5 mM).

Miller et al [13] have imaged thein situ second-ary structure of the amyloid plaques in AD braintissue. Using synchrotron Fourier transforminfrared micro-spectroscopy and a synchrotronx-ray fluorescence microprobe on the same sam-ple, they showed a strong spatial correlationbetween elevatedβ-sheet content in Aβ plaquesand accumulated Cu2+ and Zn2+, emphasizing anassociation of metal ions with amyloid formationin AD. There was also a strong spatial correlationbetween the two ions. Higher Zn2+ concentrationshave also been seen histologically in plaquedeposits [14], and the importance of Zn2+ in plaqueformation has been emphasized by the finding thatage- and female sex-related plaque formation inAPP2576 transgenic mice was greatly reducedupon the genetic ablation of the zinc transporter 3protein, which is required for zinc transport intosynaptic vesicles [15].

Bush et al. [16] found that Aβ coordinated Cu2+,Zn2+, and Fe3+ with high affinity [17, 18], which

would explain the presence of these metals in amy-loid plaques. This study also showed stabilizationof an apparentΑβ1–40 dimer by Cu2+ on gel chro-matography suggesting an interaction betweenCu2+ and Aβ1–40. Clements et al. [19] observeddisplacement of65mZn2+ from Aβ when co-incubated with excess Cu2+, while Yang et al. [20]found that Cu2+ and Zn2+ shared a common bindingsite. Atwood et al. [21] found that Cu2+ was boundto soluble Aβ via histidine residues and that theprecipitation of soluble Aβ by Cu2+ was reversiblymodulated by pH with mildly acidic conditions(pH 6.6) greatly promoting Cu2+-mediated precipi-

tation, whereas raising the pH dissolved precipi-tated Αβ:Cu2+ complexes. Cherny et al. [22]observed that Zn2+ induced aggregation of solubleΑβ at pH 7.4 in vitro, which was totally reversiblewith chelation. They also found that marked Cu2+-induced aggregation ofΑβ1–40 occurred as thesolution pH was lowered from 7.4 to 6.8 and thatthe reaction was completely reversible with eitherchelation or raising the pH.Αβ1–40 was reported

to bind three to four Cu2+

ions when precipitated atpH 7.0. Rapid, pH-sensitive aggregation occurredat low nanomolar concentrations of bothΑ β1–40and Αβ1–42 with submicromolar concentrations of Cu2+. UnlikeΑβ1–40, Αβ1–42 was precipitated bysubmicromolar Cu2+ concentrations at pH 7.4. RatΑβ1–40 and histidine-modified humanΑβ1–40were not aggregated by Zn2+, Cu2+, or Fe3+, indi-cating that histidine residues are essential formetal-mediated Αβ assembly. Cherny et al. [23]also showed that Cu2+- and Zn2+-selective chelatorsenhanced the dissolution of amyloid deposits inpostmortem brain specimens from AD subjects andfrom amyloid precursor protein overexpressingtransgenic mice, confirming the part played bythese metal ions in cerebral amyloid assembly. Inparticular, Zn2+ efficiently induces aggregation of synthetic Αβ under conditions similar to the physi-ological ones in the normal brain, that is, atnanomolar and submicromolar concentrations of Αβ and free Zn2+, respectively [15–17].

Recently, it has been demonstrated that Aβ willnot precipitate when trace metal ions are rigorouslyexcluded [24]. On the other hand, the very strongprecipitating effect of Zn2+ implies that there aresome factors protectingΑβ from Zn2+-inducedaggregation in the normal brain. Certain metal ionssuch as Mg2+ and Ca2+, which do not exhibit a pre-cipitating effect, have been hypothesized to havethis protective effect [25]. However, the inhibitionof Zn2+-induced Αβ aggregation by these metalions has not yet been verified. The effect of Cu2+ onthe aggregation ofΑβ is ambiguous compared withZn2+. Cu2+ has been shown to be a strong inducer of

126 C.C. Curtain and K.J. Barnham

Human DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA

Rat DAEFGHDSGFEVRHQKLVFFAEDVGSNKGAIIGLMVGGVVIA

FIGURE8.1. Sequence of human Aβ compared with that of the rat.



Αβ aggregation under certain conditions [24]. Incontrast with the Zn2+-induced Αβ aggregation thatoccurs over a wide pH range (5.5–7.5), the Cu2+-induced aggregation occurs primarily at mildlyacidic pH [21].

Atwood et al. [21] determined a half-maximalbinding of Cu2+ for Αβ in the micromolar range(4.0 µM forΑβ1–40 and 0.3µM forΑβ1–42) byindirect spectrophotometric analysis. However, thisanalysis of binding affinities was limited by thesensitivity of the spectrophotometric technique andthe lack of competitive binding factors in the incu-bation that would emulate the physiological situa-tion more closely. Garzon-Rodriguez et al. [26]used a more sensitive fluorescence technique and asingle tryptophan (F4W) mutant of Aβ1-40 to

show that the relative affinities were Fe < Cu > Zn.Syme et al. [27] used the competitive effects of glycine andL-histidine to measure Cu2+ affinity forAβ by fluorescence spectroscopy. Adding Cu2+ toAβ1–28 caused marked quenching of the tyrosinefluorescence signal at 307 nm. Added glycine com-petes with Aβ for the Cu2+, and the tyrosine fluo-rescence signal reappears at a sufficiently highglycine levels. Cu2+ coordinates to glycine via theamino and carboxylate groups with an apparentpH-adjusted K

aof 1.8× 106 M−1, and two glycine

residues will bind to a single Cu2+ ion [27]. It tookmore than 100 mol equivalents of glycine to causethe tyrosine fluorescence signal to completelyreturn to its maximal strength. Half of the maximalquenching is achieved at approximately 18 ± 2 eq.of glycine. Finally, Huang et al. [24] had shownthat binding of Cu2+ to Αβ1– 42 promoted precipi-tation with so high an affinity that it was hard toavoid aggregation unless buffers were most rigor-ously treated with chelating agents. Even then, it isdifficult to remove the last traces of metal ion,which may account for many of the inconsistenciesreported in the Aβ metal binding literature.Extremely small changes in free or exchangeableCu2+ concentration are also likely to have a signif-icant effect onΑβ solubility in vivo.

8.3 Αβ StructuresThe structure of the metal binding site of Aβ mustbe considered in the context of the structure of thewhole molecule. Because it has been widely held

that Aβ exerts its neurotoxic action via interactionswith neuronal membranes additionally to or in con-cert with its redox activity, there have been manystudies on its structure in a variety of membranemimetic systems. A major obstacle to the determi-nation of definitive structures is the difficulty of obtaining reproducibly a random-structured start-ing material or, alternatively, of mimicking itstransmembrane conformation immediately aftersecretase cleavage. Furthermore, because aqueoussolutions of Aβ accumulate significant amounts of aggregates within a few hours, NMR studies can bedifficult. Nevertheless, early NMR studies of human Aβ1-40 showed a random coil structure inaqueous solution (pH 4) at micromolar concentra-tion [28]. The secondary structure of Aβ40 peptide

in 40% TFE buffered at pH 2.8 with 50 mM potas-sium phosphate was also studied by NMR. Underthese conditions, there was aggregation only after aweek and the NMR spectra were well resolved.Solution structures of Aβ1–40 in perdeuteratedsodium dodecyl sulfate (SDS-d 25 ) micellesobtained by Coles et al. [29] showed twoα -helicalsegments. The helical arrangement of residues15–25 and 29–37 was confirmed by intense NOEconnectivity (3–4 residues) while medium-rangeNOE for residues 25–29 were either weak or notobserved. The “break” between the two helices wassuggested by D2O exchange experiments, whereprotons on residues 25–29 were shown to exchangerapidly and, from quantitative structural and dihe-dral angle restraint calculation prediction, a kinkwas seen at residues 26–27 acting as a “hinge” forthe two helices.

Shao et al. [30] showed twoα -helical regionsbetween Tyr10-Val24 and Lys28-Val36 for bothAβ1–40 and Aβ1–42 in SDS-d 25 at pH 7.2. Thedata were supported by structural calculations indi-cating α-helices between residues 10–24 and28–42 with the region Gly25-Asn27 as a connect-ing loop. Similar downfield shifts of Aβ1–40 andAβ1–42 at Val39-Val40 and Val40-Ile41, respec-tively, suggested a structural preference for thepeptides at their C-terminus. This may be related toconformational averaging between a micelle boundα -helical structure andβ-sheet when the peptidesleave the micelle surface.

Most NMR studies in solution were done ineither trifluoroethanol (TFE) [31] or SDS-d 25 /D2Oto mimic a membrane environment, although an

8. Copper Coordination byβ-Amyloid and Neuropathology of AD 127



early study by Sorimachi and Craik [32] showedsome α-helical structure in dimethyl sulfoxide(DMSO). Theα -helical conformation found byNMR was further supported by far ultraviolet cir-cular dichrosim (CD) spectroscopy which showedthat Aβ1–28 in the presence of charged membrane-like surfaces, especially negatively charged SDS,preferred a helical structure. Other membrane-likespecies, zwitterionic dodecylphosphocholine(DPC) and dodecyltrimethylammonium chloride(DTAC), with heterogenous amphiphilic environ-ments similar to biological systems have beenused. Fletcher and Keire [33] used solution NMRand CD to study the conformation of Aβ12–28 indodecylphosphocholine (DPC) and SDS micellesas a function of pH and lipid type. Interaction with

micelles was weak but changed the conformationwhen compared with aqueous buffer alone.However, the peptide interacted strongly withanionic SDS micelles, where it was mostly bound,wasα -helical from Lys16 to Val24, and aggregatedslowly. The pH-dependent conformational changesof the peptide in solution occurred in the pH rangeat which the side-chain groups of Asp22, Glu23,His13, and HisI4 are deprotonated (pKs ~ 4 and6.5). The authors concluded that the interaction of Aβ12–28 with SDS micelles altered the pH-dependent conformational transitions of the pep-tide whereas the weak interaction with DPCmicelles caused little change.

These conformational changes indicate a rela-tionship between peptide structure and electro-static interactions involving protonation anddeprotonation of the micelle lipid head groups atdifferent pH. In experiments using Aβ1–40 withthe imidazole side chains of the histidine residues6, 13, 14 methylated, Tickler et al. [34] found that

the peptide-lipid interaction was modulated by thehistidine residues and, therefore, would be pHsensitive. Aβ1–28 appears to associate with thesurface of the membrane based on an irregularpattern in the amide chemical shift temperaturecoefficient dependence, suggesting that the amidebackbone is situated at the water and micelleinterface. Narrower NMR line widths indicatedconformational mobility at the micelle surfaceand the concentration of Aβ1–28 not affecting CDand NMR data suggested that theα -helical struc-ture is more likely to be stabilized by rapidexchange [33].

Jin et al. [35] used NMR spectroscopy to deter-mine the solution structure of rat Aβ1–28 (seeFig. 8.1) and its binding constant for Zn2+. Theyfound that the three-dimensional solution structureof rat Aβ1–28 was more stable than that of humanAβ1–28 in DMSO-d

6and that a helical region from

Gln15 to Val24 existed in the rat Aβ1–28. Theaffinity of Zn2+ for rat Aβ1–28 was lower than thatfor human Aβ1–28, and Arg13, His6, and His14residues provide the primary binding sites for Zn2+.They also found that Zn2+ binding to rat Aβ1–28caused the peptide to change to a more stableconformation.

Gröbner et al. [36] have outlined a method forstructure determination ofΑβ in membrane sys-tems. First, they used CD and31P magic angle spin-

ning (MAS) NMR spectroscopies to characterisethe peptide in a dimyristoyl phosphatidylcholine/dimyristoyl phosphatidyl glycerol vesiclesystem. Their most notable finding was that theycould get Aβ1–40 to give anα -helical structure if the peptide were dialyzed from TFE solution intothe vesicles. That is, it was given no opportunity toform β-structure inducing fibrils by contact withwater. Second, they used rotational resonance13CCP MAS NMR recoupling techniques to show thatthe membrane-penetrant part of the peptide wasα -helical before major aggregation had occurred.To gain further insights, these authors concluded,future MAS studies would have to be made on mul-tiple uniformly labeled peptides. Further advancesin spectral resolution and sensitivity are vital, as isdevelopment of labeling methodologies. The devel-opment of pulse sequences and appropriate algo-rithms to extract multiple distance and torsionangle constraints from these systems would also beneeded. Thus, the determination of the structure of

Aβ by NMR in a membrane environment is stillincomplete.

8.4 The Structure of Aβ in FibrilsConventionally, the supramolecular structure ofβ-sheet entities such as amyloid plaques can be con-sidered to be either parallel or antiparallel. Whichmode is likely to be important for determining theresidues involved in the metal-bridged cross-linksthat occur in amyloid plaques and for the subsequentredox chemistry.13mC multiple quantum SS-NMR




has been used to probe the structure of the full-length Aβ peptide [37]. Internuclear distances of approximately 4.8 Å would be observed for13mC-labeled residues if theβ-sheets form an in-registerparallel structure. An antiparallel structure, on theother hand, would have nearest neighbor residuesexhibiting far larger distances than 4.8 Å. Usingthese NMR techniques, Aβ1–40 was shown to forma parallelβ-structure [35]. This finding is similar tothat of Benziger et al. [38] for Aβ10–35.Comparison with their data shows evidence thatAβ10–35 fibrils have parallelβ-sheet organizationbeyond dimers. However, SS-NMR studies onAβ34–42 fibrils suggested an antiparallelβ-struc-ture, which was also observed for Aβ16–22 cappedat both ends [39]. Lansbury et al. [40] characterized

fibrils made from the C terminal fragment Aβ34–42.They found the alignment of Aβ34–42 fibrils to beantiparallel and two residues out of register usingrotational resonance experiments on doubly13C-labeled samples. Therefore, SS-NMR studies havepresented evidence for both parallel and antiparallelalignments of Aβ fragments, depending on the pep-tide sequence studied and the methodologyemployed.

In a different approach, Egnaczyk et al. [41]used photo cross-linking. They synthesized aphotoreactive Aβ1–40 ligand by substitutingL- p-benzoylphenylalanine (Bpa) for phenylalanineat position 4. This peptide was incorporated intosynthetic amyloid fibrils and exposed to near-UVradiation. Analysis of the fibrils showed a Bpa4-Met35 intermolecular cross-link, which was con-sistent with an antiparallel alignment of Aβpeptides within amyloid fibrils. Together, the aboveresults show that fibrils can adopt differentsupramolecular structures depending on the pep-

tide length and properties of the residues present.The differences are of considerable significance.For example, the photo cross-linking data showthat the Met35 could be very close to the metalbinding site, thus favoring redox reactions with theMet as an electron donor. On the other hand, it isquite conceivable that parallel alignment wouldgreatly favor metal-peptide cross-linking. It is pos-sible that physiologically both kinds of alignmentcould occur, the proportions being affected by dif-ferent environments, such as extracellular or mem-brane associated, the presence/absence of metalions or differing ratios of Zn to Cu.

8.5 The Metal-Binding Sites andthe Structure of AβThe randomness of the Aβ peptide in aqueous solu-tion makes it difficult to determine the nature of themetal-binding sites. The problem has beenapproached using various spectroscopic tech-niques, such as Raman, CD, and magnetic reso-nance. Miura et al. [42] used Raman spectroscopyto study the binding modes of Zn2+ and Cu2+ to Aβin solution and insoluble aggregates. They foundtwo different modes of metal-Aβ binding, onecharacterized by metal binding to the imidazole Nτatom of histidine, producing insoluble aggregates,the other involving metal binding to the Nπ, but not

the Nτ, atom of histidine as well as to main-chainamide nitrogens, giving soluble complexes. Zn2+

binds to Aβ only via the Nτ regardless of pH, whilethe Cu2+ binding mode is pH dependent. At mildlyacidic pH, Cu2+ binds to Aβ in the former mode,whereas the latter mode is predominant at neutralpH. Miura et al. [42] proposed that the transitionfrom one binding mode to the other explained thestrong pH dependence of Cu2+-induced Aβ aggre-gation. Dong et al. [43] also employed Ramanmicroscopy to study the metal-binding sites inamyloid plaque cores, using the spectra-structurecorrelations for Aβ–transition metal binding. Theyobserved that Zn2+ was coordinated to the histidineNτ and the Cu2+ to the Nπ, confirming that the metalbinding mode was the same in both the syntheticpeptide and its aggregates and the naturallyoccurring plaques.

Huang et al. [44] used multifrequency EPR(L-band, X- and Q-band) to show that copper coor-dinates tightly to Aβ1–40 and that an approxi-

mately equimolar mixture of peptide and CuCl2produced a single Cu2+-peptide complex.Computer simulation of the L-band spectrum withan axially symmetrical spin Hamiltonian and thegand A matrices (g ||, 2.295; g

⊥, 2.073; A||, 163.60; A

⊥,

10.0× 10−4 cm−1) suggested a tetragonally distortedgeometry, which is commonly found in type 2 cop-per proteins. Expansion of the M I = −1/2 resonancerevealed nitrogen ligand hyperfine coupling.Computer simulation of these resonances indicatedthe presence of at least three nitrogen atoms. Thisand the magnitude of theg || and A|| values, togetherwith Peisach and Blumberg [45] plots, are




consistent with a fourth equatorial ligand bindingto copper via an oxygen rather than a sulfur donoratom. Thus, the coordination sphere for the copper-peptide complex was considered to be 3N1O.These authors also used EPR spectroscopy tomeasure residual Cu2+ remaining after incubatingstoichiometric ratios of CuCl2 with Aβ1–40. Therewas a 76% loss of the Cu2+ signal, compatible withpeptide-mediated reduction of Cu2+ to diamagneticCu+, which is undetectable by EPR, agreeing withthe corresponding concentration of Cu+ measuredby bioassay. There was no evidence of free, unco-ordinated Cu2+ remaining after addition of the pep-tide, because unbound Cu2+ itself gives a differentmultiple resonance signal.

Using a combination of NMR and EPR spectro-

scopies, Curtain et al. [46] proposed a structure forthe high-affinity site and drew some conclusionsabout the interaction of the peptide with lipids andits modification by Cu2+, Zn2+, and pH. NMR stud-ies on Aβ1–28 and Aβ1–40/2 indicated that bothpeptides were undergoing significant conforma-tional exchange in aqueous solution. NMR andEPR spectra were also recorded for Aβ1–28 wherethe Nε2 nitrogens of the imidazole ring of the Hisresidues 6, 13, and 14 were methylated (Me-Aβ1–28). The NMR spectra of Me-Aβ1–28 werevirtually identical to Aβ1–28, the only significantdifferences being three strong singlets in the1Hspectrum at 3.80, 3.82, and 3.83 ppm from themethyl groups attached to the His imidazole rings.A precipitate formed when Zn2+ was added to thesolutions of Aβ1–28 or Aβ in PBS. NMR spectraof the supernatant of Aβ1–28 treated with Zn2+

showed that peaks assigned to C2H and C4H of His6, His13, and His14 of Aβ1–28 had broadenedsignificantly. However, there was little or no

change in the rest of the spectrum compared withAβ1–28 prior to the addition of Zn2+. This broad-ening of the NMR histidine residue peaks is theresult of the interaction of these residues with Zn2+.

The histidyl side chain is a well-established lig-and of zinc in proteins and peptides [47], so thisresult suggested that three of the ligands bound toZn2+ were most likely to be the imidazole rings of the histidine residues [48]. Indeed, His13 had beenestablished by Liu et al. [49] as a crucial residue inthe Zn2+-mediated aggregation of Aβ. The broad-ening of these peaks is the result of chemicalexchange between free and metal-bound states or

among different metal-bound states. The extent of broadening of the peaks indicated intermediateexchange, which on the NMR timescale suggeststhat the metal-binding affinity is in the micromolarrange, in agreement with the low-affinity sitedescribed by Bush et al. [16]. The absence of anychange in the rest of the spectrum suggested thatthe metal-bound form of the peptide wasmonomeric and that there was little or no signifi-cant amount of soluble oligomer in solution,because higher order aggregates would haveresulted in significantly broadened resonances.

When Cu2+ or Fe3+ was titrated into an aqueoussolution of Aβ1–28, similar changes were observedin the1H spectrum, with the peaks assigned to theC2H and C4H of His6, His13, and His14 disap-

pearing from the spectrum (Fig. 8.2). A slightbroadening of all peaks in the spectrum (associatedwith the paramagnetism of Cu2+ and Fe3+) was alsoobserved, but there were no other major changesafter the addition of the metal ions. Metal-inducedprecipitation blocked attempts to saturate themetal-binding site. The precipitate made the col-lection of NMR spectra difficult, and few conclu-sions could be drawn from spectra of peptideremaining in solution. When Cu2+ was added to anaqueous solution of Me-Aβ1–28, the changesobserved in the spectrum were identical to those


Aβ

9.0 8.0 7.0 ppm

Aβ + Cu 2+

FIGURE8.2. Amide and aromatic region of the 600 MHz1H NMR spectra of Aβ in aqueous PBS solution and fol-lowing the addition of Cu2+. Peaks caused by the C2Hand C4H of histidines 6, 13, and 14 have been broadenedbeyond detection because of coordination to the copper.

There is a generalized broadening of the rest of the spec-trum due to the paramagnetism of the added Cu2+. AfterCurtain et al. [46].



observed for Cu2+ added to Aβ1–28, but there wasno visible precipitate. In aqueous solution and lipidenvironments, coordination of metal ions toΑβ isthe same, with His6, His13, and His14 all involved.

The X-band EPR spectrum of Cu2+ bound to thepeptides had the unsplit intense g

⊥

resonance char-acteristic of an axially symmetric square planar3N1O or 4N coordination, g|| = 2.28 and g

⊥= 2.03,

Α|| = 173.8 gauss. Similar parameters were foundfor Cu2+ coordination by Aβ1–16, Aβ1–40, andAβ1–42, indicating that the site was not affected bythe size of the C-terminal regions of the peptides.A notable finding with peptides of all lengths wasthat increasing the Cu2+ above ~0.3 mol/mol pep-tide caused line broadening in the Cu2+ EPR spec-tra, over a pH range of 5.5 to 7.5, suggesting thepresence of dipolar or exchange effects (Fig. 8.3).These would be observed if two or more Cu ionswere within approximately 6 Å of each other.These effects could be explained if at Cu2+ /peptidemolar ratios >0.3,Αβ coordinated a second Cu2+

atom cooperatively. They were abolished if the his-tidine residues were methylated at either Nδ1 orΝε2, suggesting that bridging histidine residueswere being formed (Fig. 8.4) [32, 46].

One consequence of coordination by a metal ionto the Nδ1 of a histidine residue is a reduction inthe pKa of Nε2 NH, making this nitrogen moresuitable for metal binding [48], resulting in a histi-dine residue that can bridge metal ions; a goodexample being His63 at the active site of superox-ide dismutase [50]. Similar bridging histidineresidues have been proposed in the octarepeatregion of the prion protein [51], which has beenshown to possess significant SOD activity in thepresence of Cu2+ [52]. The line-broadening effectsobserved in the EPR spectra at Cu2+ /Aβ molar frac-tions up to 1.0 by Curtain et al. [46] were notobserved by Syme et al. [27], Huang et al. [44] orAntzutkin [52]. It is relevant that Huang et al. [54]

along with Narayanan and Reif [55] have shownthat NaCl has a marked effect on metal-inducedaggregation of Aβ. Huang et al. [44] and Curtainet al. [46] obtained their spectra from samples inphosphate-buffered saline at pH 7.4, Antzutkin[53] adjusted the pH of his sample to pH 7.4 anddialyzed against distilled water, while Syme et al.[27] used ethyl morpholine buffers.

Similar line-broadening phenomena to thatobserved by Curtain et al. [46] have been observedin the EPR spectra of imidazole-bridged coppercomplexes designed as SOD mimetics [56]. Thebridging histidine may be responsible for the


2500 2700 2900 3100Gauss

3300 3500

DC

BA

FIGURE8.3. EPR spectra (9.7 GHz) of Aβ1–28 to whichhad been added respectively: A, 0.2/1 M/M; B, 0.4/1M/M; C, 0.6/1 M/M; D, 0.8/1 M/M Cu2+ /peptide. Allspectra recorded at 130 K in pH 7.4 phosphate-buffered

saline. Spectra C and D show significant broadening of the g⊥

line. All lengths of Aβ studied give identical spec-tra (Curtain et al. [46, 79]).

A BBridging histidine

Tyr 10

~6A

Tyr 10

CuCu

FIGURE8.4. Model showing how two Aβ strands (A andB) could be linked by two copper atoms through a bridg-

ing histidine. The 6 Å distance between the copper atomsis within the range at which we would expect to see dipo-lar broadening of Cu2+ EPR spectra of the type seen inFigure 8.2.



reversible metal-induced aggregation that isobserved when Aβ is metallated with Cu2+ andZn2+. The bridging histidine residues may alsoexplain the multiple metal-binding sites observedfor each peptide and the high degree of cooperativ-ity evident for subsequent metal binding. Withthree histidines bound to the metal center, a largescope exists for metal-mediated cross-linking of the peptides leading to aggregation, which will bereversible when the metal is removed by chelation.It should be noted here that the bridging histidinehypothesis of peptide association would favor aparallel over an antiparallelβ-sheet structure forthe fibrils and plaques. It is quite possible thatmetal-induced precipitation of Aβ is quite differentfrom that induced by prolonged incubation of

monomeric peptide in the putative absence of metal. For example, Miura et al. [42] strongly sug-gested that the metal-induced aggregation of Aβwas promoted by cross-linking of the peptidesthrough metal-His[Nτ] bonds, most likely throughHis[Nτ]-metal-His[Nτ] bridges at three histidineresidues.

Observations that rat Aβ, which differs fromhuman Aβ by three substitutions (Fig. 8.1) [57],does not reduce Cu2+ and Fe3+, is not readily pre-cipitated by Zn2+ or Cu2+, does not produce ROS asstrongly as the human sequence, and does not pro-duce plaques highlight the importance of the threehistidines. Rat Aβ forms a metal complex via twohistidine residues and two oxygen ligands ratherthan three histidine residues and one oxygen lig-and, compared with human Aβ where the sidechain of His13 of human Aβ is ligated to the metalion. This was borne out by the EPR spectrum,which was typical of a square planar 2N2O Cu2+

coordination [44].

Syme et al. [27] and Antzutkin [53] both usedX-band EPR to study the interaction of Aβ withCu2+ in solution, confirming the axially symmetricbinding site. Syme et al. [27] obtained EPR spectraat pH 7.4 and higher that showed heterogeneityattributed to a second high-affinity binding site.This site became much more prominent when thepH was raised to 10.0. The heterogeneity at pH 7.4was not observed by Huang et al. [44], Curtainet al. [46], or by Antzutkin [53] and warrants fur-ther investigation. It is possible that the secondbinding site is a buffer ion effect. In order to definethe binding site, Syme et al. [27] also prepared

mutants of Aβ1–28 in which each of the histidineresidues had been replaced by alanine or in whichthe N-terminus was acetylated, and their data sug-gested that the N-terminus and His13 and His14are crucial for Cu2+ binding and that H6 also playeda part. On this basis, they proposed a square planarmodel with the Cu2+ coordinated to His13, His14,His6, and the amino N of the N-terminus. Althougha 4N model may be fitted to Syme et al.’s [27] X-band spectra, it is not compatible with the conclu-sions derived by Huang et al. [44] from L-bandspectra and their superhyperfine structure thatpoint to a 3N1O coordination.

Karr et al. [58] found that Aβ peptides lackingone to three N-terminal amino acids but containingHis6, His13, and His14 and Tyr10 did not coordi-

nate Cu2+ in the same environment as the nativepeptide, suggesting that these N-terminal residuesare significant for Cu2+ binding. They also con-firmed that the coordination is identical with anylength of peptide (Aβ1–16, Aβ1–28, Aβ1–40,Aβ1–42) that contained the first 16 amino acids.These authors also showed [59] that the coordina-tion of Cu2+ did not change during organization of monomeric Aβ into fibrils and that neither solublenor fibrillar forms of Aβ1–40 contained antiferro-magnetically exchange-coupled binuclear Cu2+

sites in which two ions were bridged by an inter-vening ligand. The latter conclusion was based ona temperature-dependence study of the EPR spec-tra for Cu2+ bound to soluble or fibrillar Aβshowing that the Cu2+ center displayed normalCurie behavior, indicating that the site wasmononuclear.

Further advances in understanding the N coordi-nation of Cu2+ will require more sophisticated EPRtechniques than have been used so far, supportedby input from other methods such as XAFS.Equally, there remains uncertainty as to the natureof the potential O ligand. Proton NMR dataobtained by Syme et al. [27] agreed with the find-ings of Huang et al. [44] and Curtain et al. [46] thathistidine residues are involved in Cu2+ coordina-tion, but they found that Tyr10 was not involved.Further, Karr et al. [58] found that the coordinationof Cu2+ in the Y10F mutant of Aβ remained 3N1Owith EPR spectra identical to the wild-type spectra.Isotopic labeling experiments showed that waterwas not the O-atom donor to Cu2+ in Aβ fibrils orin the Y10F mutant. However, the Raman data of




Miura et al. [42] suggest that the ligand was the Oof the tyrosine hydroxyl. They were able to assignthe 1504 cm−1 band in the Raman spectra of insol-uble Cu2+-Aβ1–16 aggregates to Cu2+-boundtyrosinate, and the high intensity of the 1604 cm−1

band was attributed to a contribution from the Y8aband of tyrosinate. Unlike Zn2+, Cu2+ binds to tyro-sine in the insoluble aggregates of Aβ1–16. Whenthe deprotonated phenolic oxygen of tyrosinate isbound to a transition metal ion such as Cu2+ andFe3+, the Y19a band shifts to about 1500 cm−1 andgains intensity through resonance with að (pheno-late) f d (metal) charge-transfer transition in thevisible. Such charge transfer does not occur forZn2+, thed orbitals of which are fully occupied. Itshould be noted that in these experiments, Miura

et al. [42] used phosphate-buffered saline, whichmight have had the effect of encouraging peptideassociation [54, 55].

In conclusion, although there is general agree-ment as to the nature of the monomeric binding siteinsofar as it is type two Cu2+ with a 3N10 coordi-nation, varying buffer conditions, peptide concen-tration, and conformation make it difficult tocompare one set of published data with another.There is a similarity here with the studies on thealignment of the peptide in fibrils. In consideringthe issue of monomeric versus dimeric Cu2+, it isimportant to remember that Aβ may formoligomers and multimers in a variety of ways,some more relevant to its neurotoxicity than others[60–64].

8.6 Aβ Redox Activity and theRole of Metal Coordination

Oxidative stress markers characterize the neu-ropathology both of Alzheimer’s disease and of amyloid-bearing transgenic mice. The neurotoxic-ity of Aβ has been linked to hydrogen peroxidegeneration in cell cultures by a mechanism that isstill being fully described but is likely to bedependent on Aβ coordinating redox active metalions. Huang et al. [65] showed that human Aβdirectly produces hydrogen peroxide (H2O2) by amechanism that involves the reduction of metalions, Fe3+ or Cu2+. They used spectrophotometry toshow that the Aβ peptide reduced Fe3+ and Cu2+ toFe2+ and Cu+ and that molecular oxygen is then

trapped by Aβ and reduced to H2O2 in a reactionthat is driven by sub-stoichiometric amounts of Fe2+ or Cu+. In the presence of Cu2+ or Fe3+, Aβproduced a positive thiobarbituric-reactive sub-stance, compatible with the generation of thehydroxyl radical [OH*]. Tabner et al. [66] used the5,5-dimethyl-1-pyrroline N -oxide (DMPO) spin-trap to identify the radical produced by Aβ in thepresence of Fe2+ and concluded that it was OH*.However, they also found OH* was produced in thepresence of Fe2+ by Aβ25–35, which does not con-tain a strong metal binding site. Because Fe2+ withtrace amounts of Cu2+ as low as 0.01 mol%, corre-sponding with the amount of adventitious Cu foundin the average peptide preparation, will produce anOH* adduct with the DMPO spin trap [Curtain

et al., unpublished], Tabner et al.’s [66] resultsshould be treated with caution even though theyappear to confirm the findings of Huang et al. [65].

In the course of metal-catalyzed redox activity,Aβ may undergo under a number of changes.Atwood et al. [67] found that Cu2+ induced the for-mation of SDS-resistant oligomers of Aβ that gavea fluorescence signal characteristic of the cross-linking of the peptide’s Tyr10. This finding wasconfirmed by directly identifying the dityrosine byelectrospray ionization mass spectrometry and bythe use of a specific dityrosine antibody. The addi-tion of H2O2 strongly promoted Cu2+-induced dity-rosine cross-linking of Aβ1–28, Aβ1–40, andAβ1–42, and it was suggested that the oxidativecoupling was initiated by interaction of H2O2 witha Cu2+ tyrosinate. The dityrosine modification issignificant because it is highly resistant to proteo-lysis and would be important in increasing thestructural strength of the plaques. Schoneich andWilliams [68], however, were unable to find anyevidence of tyrosine oxidation. They used ascor-bate/Cu2+-induced oxidation and electrosprayionization-time-of-flight MS/MS analysis to studythe oxidation products of Aβ1–16, Aβ1–28, andAβ1–40. Initial oxidation targets were His13 andHis14, which were converted to 2-oxo-His, whileHis6 and Tyr10 were unchanged, although His6was oxidized after longer oxidation times. The for-mation of 2-oxo-His suggests that a transient 2Ccentered His radical might have been formed. Suchradicals have been described in a number of bio-logical redox systems [69, 70], although not sofar in any of neuropathological significance.




Schoneich and Williams [68] explained the insen-sitivity of His6 to initial oxidation by suggestingthat histidine bridging of two Cu2+-Aβ moleculeslowered the electron density on His6, comparablewith similar results on a Cu2+- and Zn2+-bridgingHis61 residue of bovine Cu,Zn superoxidedismutase.

Barnham et al. [71] used density functional the-ory calculations to elucidate the chemical mecha-nisms underlying the catalytic production of H2O2by Aβ /Cu and the production of dityrosine. Here,Tyr10 was identified as the critical residue. Thisfinding accords with the growing awareness thatthe O2 activation ability of many cupro-enzymes isalso coupled to the redox properties of tyrosine andthe relative stability of tyrosyl radicals. The latter

play important catalytic roles in photosystem II,ribonucleotide reductase, COX-2, DNA pho-tolyase, galactose oxidase, and cytochrome-coxidase [72].

With ascorbate as the electron donor, the firststep in the catalytic production of H2O2 is thereduction of Cu2+ to Cu+. Barnham et al. [71] pro-posed that the transfer could take place via a pro-ton-coupled electron transfer (PCET) mechanism.

Reactions involving PCET are being increasinglyimplicated in a range of biological systems, includ-ing charge transport in DNA and enzymatic oxygenproduction [73]. In this system, the electron trans-fer involves both p- and d-orbitals on the ascorbate,Tyr10, and the copper ion, while proton transferinvolves p-orbitals on the O2-atom of ascorbate,and the side-chain oxygen of Tyr10 (Figs. 8.5A and8.5B). The significant change in electron spin onthe copper ion going from the ground state to thetransition state suggests that the proton and theelectron are transferred within different molecularorbitals, as is predicted to be necessary for PCET tooccur [73]. The activation energy for this oneelectron reduction step was computed to be only0.9 kcal/mol.

Barnham et al. [71] tested the Cu/tyrosinatehypothesis using an Aβ1–42 peptide with Tyr10substituted with alanine (Y10A). Both peptidesgave rise to similar65mCu EPR spectra with thestrong single g

⊥resonance characteristic of an axi-

ally symmetric square planar complex, althoughthere was a significant increase in the g|| value of Y10A. The increase was probably due to somedistortion of the coordination sphere because the


= O

= C

= N

= H

Ascorbicanion

Tyr 10

His 61.143

1.081

1.355

1.561

Water

1.2522.028

2.054

3.110 Cu

His 14His 13Ns(a) (b)

= O

= C

= N

= H

1.483

1.2952.364

1.403

FIGURE8.5. (A) The transition state that is formed when a hydrogen atom is transferred from ascorbate to the sidechain oxygen of Aβ Y10, which acts as a gate, and passes an electron to Cu2+ reducing it to Cu+ [71]. (B) An inter-mediate formed along the reaction path where Y10 has transformed into a tyrosyl radical giving up its side-chahydroxyl hydrogen atom to O2

● −⊥

via hydrogen atom transfer. Simultaneously, H3O+ has donated its proton to O2● −

via proton transfer, whereupon H2O2 has formed. Formed tyrosyl radical and water molecule are hydrogen bonded toH2O2. Ascorbyl radical anion coordinates via its O1-oxygen anion in an apical position to Cu2+. Figures based on dataof Barnham et al. [71].



fluid bilayer lipids (Fig. 8.7), which can be quanti-fied to give both the stoichiometry and selectivityof the first shell of lipids interacting directly withmembrane-penetrant peptides. The stoichiometricdata can give an estimate of the number of subunitsin a membrane-penetrant oligomeric structure.Using this approach, it was shown that Aβ1–40 and

Aβ1–42 bound to Cu2+ or Zn2+ penetrated bilayersof negatively charged, but not zwitterionic lipid,giving rise to such a partly immobilized componentin the spectrum (see Fig. 8.7 and its caption)[46, 79].

When the peptide:lipid ratio was increased, therelationship between the mole fraction of peptideand proportion of slow component was linear.Even at a fraction of 15%, all of the peptide wasassociated with the lipid, suggesting that the struc-ture penetrating the membrane lipid was welldefined, although at such a high peptide:lipid ratiofurther study would be needed to confirm whether

the lipid still retained a lamellar structure.Formation of non-lamellar structures in regions of the membrane associated with Aβ could well bethe cause of the peptide’s cytotoxicity. From thespin-label data, the first shell lipid:peptide wasapproximately 4:1. This stoichiometry can be sat-isfied by 6 helices arranged in a pore surroundedby 24 boundary lipids. This hypothetical structuregains credibility from atomic force microscopystudies of Aβ1–42 reconstituted in a planar lipidbilayer that showed multimeric channel-like struc-tures, many resembling hexamers, similar to thatmodeled in Figure 8.8 [81]. It was found [46] thatin the presence of Zn2+, Aβ1–40 and Aβ1–42 bothinserted into the bilayer over the pH range 5.5–7.5,as did Aβ1–42 in the presence of Cu2+. However,

Aβ40 only penetrated the lipid bilayer in the pres-ence of Cu2+ at pH 5.5–6.5; at higher pH, therewas a change in the Cu2+ coordination sphere thatinhibited membrane insertion. The addition of cholesterol up to 0.2 mole fraction of the total lipidinhibited insertion of both peptides under all con-


3250

A

B

C

3275 3300Gauss

3325 3350

FIGURE8.7. A: X-band EPR spectrum recorded at 305 Kof the negatively charged spin probe 1-palmitoyl-2-(16-doxyl stearoyl) phosphatidyl serine in negatively chargedLUV made from 50% palmitoyl oleoyl phophatidyl ser-ine and 50% palmitoyl oleoyl phophatidyl choline(probe/lipid 1/300). B: X-band spectra of system A at thesame temperature after the addition of Cu2+Aβ1–42(peptide/lipid 1/50), showing a shoulder (marked witharrow) to the left of the low field line. This is typical of peptide penetration into the bilayer core [73, 75]. C: Thedifference spectrum× 5 obtained when spectrum A issubtracted from spectrum B. This spectrum representsthe motionally restricted lipid in the boundary. Original

data given in Curtain et al. [46, 79].

FIGURE 8.8. Animation of hexameric pore formed byAβ1–40 helices calculated from annular lipid stoichiom-etry as determined from the EPR data shown in Figure8.5. Polar residues are shown as dark and nonpolar aslight. View from N-terminus. Peptide coordinates (inSDS) obtained from Barrow and Zagorski [31]. Model

prepared using Sculpt®

by aligning hydrophobic contactsbetween helices and orienting nonpolar residues insequence 21–40 to annular lipid.



ditions investigated. CD spectroscopy revealedthat the Aβ peptides had a highα -helix contentwhen membrane penetrant, but were predomi-nantly β-strand when not. Simulation of the spec-tra and calculation of the on-off rates suggestedthat the peptide was most likely penetrating as anα -helix [82].

In membrane-mimetic environments, coordina-tion of the metal ion is the same as in aqueous solu-tion, with the three-histidine residues, at sequencepositions 6, 13, and 14, all involved in the coordi-nation, along with an oxygen ligand. As had beenobserved at Cu2+ /peptide molar ratios >0.3 in aque-ous solution, line broadening was detectable in theEPR spectra, indicating that the peptide was coor-dinating a second Cu2+ atom in a highly coopera-

tive manner at a site 6 Å from the initial bindingsite. So, there appear to be two switches, metal ions(Zn2+ and Cu2+) and negatively charged lipids,needed to change the conformation of the peptidefrom β-strand nonpenetrant toα -helix penetrant.The closest parallel to this behavior is thatobserved with the B18 fusogenic sequence of thefertilization protein bindin [83] that, like Aβ, pos-sesses three histidine residues strategically placedto coordinate metals. In the absence of Zn2+, thispeptide forms nonfusingβ-sheet amyloid fibrils. Inthe presence of Zn2+, an α -helical conformation isimposed on its backbone and it forms fusogenicoligomers.

8.8 The Relevance of MembraneBinding to Aβ Cytoxicity:The Role of Methionine 35

In vitro, the methionine at position 35 can act as anelectron donor, and its conversion to the sulfoxideform has been the subject of several studies, giventhat the Met(O)Aβ. peptide has been isolated fromAD amyloid brain deposits [84, 85]. Furthermore,the Raman spectroscopic study by Dong et al. [43]of senile plaque cores isolated from diseased brainshas shown that much of the Aβ in these depositscontained methionine sulfoxide with copper andzinc coordinated to the histidine residues.

Although there are several potential electrondonors such as GSH and ascorbic acid, in vivo it islikely that Met35 occupies a privileged position

being part of the Aβ sequence. When it is missingas in Aβ1–28, the addition of exogenous methion-ine permits redox action to proceed, but withslower kinetics [46]. When Met35 is sequesteredwithin a lipid environment, there is also no metalreduction. Its oxidation also alters the physicalproperties of the peptide. Met(O)Aβ is more solu-ble in aqueous solution, and there is a disruption of the local helical structure when the peptide is dis-solved in SDS micelles [86].

The formation of trimers and tetramers byMet(O)Aβ is significantly attenuated and fibril for-mation is inhibited [87, 88]. Barnham et al. [89]showed by solid-state NMR that when Aβ coordi-nates and reduces Cu2+ to Cu+, the Met35 is oxi-dized. Although the Cu2+ coordination of the

oxidized peptide is identical to nonoxidized Aβand it will produce H2O2, it cannot penetrate lipidbilayers either in the presence or absence of Cu2+ orZn2+. On the other hand, Met(O)Aβ is toxic to neu-ronal cell cultures, a toxicity that is rescued bycatalase and the MPAC clioquinol. These resultssuggest that fibril formation and membrane pene-tration by Aβ could be epiphenomena, and that themain requirement for cytotoxicity is redox compe-tence. In this connection, it is important to note thatthe oxidized M35 has the potential for furtherreduction to the sulfone [90] and could thus still actas a Cu2+ reductant, acting in vivo in concert withagents such as ascorbic acid and GSH.

It might be legitimately asked whether Met35could act as a Cu2+ ligand. After all, there aremany instances of copper proteins where the ionis coordinated to a thioether, giving in most casesa type 1 binding site [91]. Such coordinationinvolving two nitrogens and an oxygen in additionto the sulfur is generally distorted tetrahedralrather than square planar and would favor Cu+over Cu2+. Because the former is EPR silent, thepossibility of this coordination might have beenoverlooked. However, in their Raman spectro-scopic studies, Miura et al. [42] were unable todetect any Cu-S bonds.

Ciccotosto et al. [92] further probed the role of Met35 by preparing Aβ1–42 in which it wasreplaced with valine (AβM35V). The neurotoxicactivity on primary mouse neuronal cortical cellsof this peptide was enhanced, and this diminishedcell viability occurred at a much faster rate com-pared with Aβ1–42. When cortical cells were




treated with the peptides for only a short 1-hduration so as to minimize the incidence of celldeath, and the amount of peptide bound to corticalcell extracts was quantitated by Western blotting,it was found that twice as much AβM35V com-pared with wild-type Aβ peptide bound to the cellsafter a 1-h cell exposure. It was suggested that theincreased toxicity was related to the increasedbinding.

AβM35V bound Cu2+ with the same coordina-tion sphere asw.t . Aβ and produced similaramounts of H2O2 as Aβ1–42 in vitro. The neuro-toxic activity was rescued by catalase. The redoxactivity of the mutated peptide was followed bymeasuring the decline in time of the strength of the Cu2+-AβM35V EPR signal, which showed

that the reduction of Cu2+ to the EPR silent Cu+was much slower compared with Aβ1-42, con-firming that the M35 residue in Aβ42 plays animportant part in the redox behavior of this pep-tide in solution. Like Cu2+-Aβ1–42, Cu2+-AβM35V inserted into a spin-labeled lipid bilayergave a partially immobilized component in theEPR spectrum. This component had a narrowerlinewidth than that found for the similar compo-nent obtained withw.t . Cu2+-Aβ1–42, suggestingthat the valine substitution made the mutant pep-tide less rigid in the bilayer region and possiblyeasier to insert, thus explaining the increased cellmembrane binding. The on- and off-rate constantsestimated from the simulation experimentsshowed that AβM35V had a higher affinity for thelipid bilayer as compared with Aβ42. CD analysisshowed that AβM35V had a higher proportion of β-sheet structure and random coil than Aβ1–42,which would also suggest a more flexible struc-ture in the bilayer [80, 82]. In summary, these and

the results described above tell us that the wild-type Aβ, its oxidized form, Met(O)Aβ, and themutant peptide, AβM35V, induce cell death viasimilar pathways that are metal-dependent andcan generate H2O2 in the absence of a methionineresidue. Fibril formation as a toxic species isnot responsible for cell death. Membrane associ-ation per se may play a part in localizing thepeptide, perhaps in domains particularly suscep-tible to oxidative damage. It follows, therefore,that elucidating the metal ion binding site of Aβmay provide a promising new therapeutic targetfor AD.

References1. Bush AI. Copper, zinc, and the metallobiology of

Alzheimer’s disease. Alzheimer Dis Assoc Disord2003;17:147-50.

2. Ritchie CW, Bush, AI, Mackinnon, A et al. Metal-

protein attenuation with iodochlorhydroxyquin (clio-quinol) targeting Abeta amyloid deposition andtoxicity in Alzheimer’s disease: a pilot phase 2 clini-cal trial. Arch Neurol 2003;60:1685-91.

3. Hershey CO, Hershey LA, Varnes A et al.Cerebrospinal fluid trace element content in demen-tia: clinical, radiologic and pathologic correlations.Neurology 1983;33:1350-53.

4. Ehmann WD, Markesbery WR, Alauddin M, et al.Brain trace elements in Alzheimer’s disease.Neurotoxicology 1986;7:195-206.

5. Thompson CM, Markesbery WR, Alaudin M et al.Regional brain trace-element studies in Alzheimer’sdisease. Neurotoxicology 1988;9:1-8.

6. Basun H, Forssell LG, Wetterberg L, et al. Metalsand trace elements in plasma and cerebrospinal fluidin normal aging and Alzheimer’s disease. J NeuralTransm Park Dis Dement Sect 1991;3:231-58.

7. Samudralwar DL, Diprete CC, Ni BF, et al.Elemental imbalances in the olfactory pathway inAlzheimer’s disease. J Neurol Sci 1995;130:139-45.

8. Deibel MA, Ehmann WD, Markesbery WR. Copper,iron, and zinc imbalances in severely degenerated

brain regions in Alzheimer’s disease: possiblerelation to oxidative stress. J Neurol Sci1996;143:137-42.

9. Cornett CR, Markesbery WR, Ehmann WD.Imbalances of trace elements related to oxida-tive damage in Alzheimer’s disease brain.Neurotoxicology 1998;19:339-45.

10. González C, Martin T, Cacho J, et al. Serum zinc,copper, insulin and lipids in Alzheimer’s diseaseepsilon 4 apolipoprotein E allele carriers. Eur J ClinInvest 1999;29:637-42.

11. Atwood, CS, Huang, X, Moir, RD, et al. Role of freeradicals and metal ions in the pathogenesis of Alzheimer’s disease. Met Ions Biol Syst1999;36:309-64.

12. Lovell MA, Robertson JD, Teesdale WJ, et al.Copper, iron and zinc in Alzheimer’s disease senileplaques. J Neurol Sci 1998;158:47-52.

13. Miller LM, Wang Q, Telivala TP, et al. Synchrotron-based infrared and X-ray imaging shows focalizedaccumulation of Cu and Zn co-localized with beta-amyloid deposits in Alzheimer’s disease. J StructBiol 2006; 155:30-37.

14. Suh SW, Jensen KB, Jensen MS, et al. Histologicalevidence implicating zinc in Alzheimer’s disease.Brain Res 2000;852:274-78.




15. Lee JY, Cole TB, Palmiter RD, et al. Contribution bysynaptic zinc to the gender disparate plaque forma-tion in human Swedish mutant APP transgenic mice.Proc Natl Acad Sci USA 2002;99:7705-10

16. Bush AI, Pettingell WH Jr, Paradis MD, et al.Modulation of Abeta adhesiveness and secretase site

cleavage by zinc. J Biol Chem 1994;269:12152-58.17. Bush AI, Pettingell WH, Multhaup G, et al. Rapidinduction of Alzheimer A beta amyloid formation byzinc. Science 1994;265:1464-67.

18. Bush AI, Moir RD, Rosenkranz KM, et al. Zincand Alzheimer’s disease -response. Science 1995;268:1921-23.

19. Clements A, Allsop D, Walsh DM, et al. Aggregationand metal-binding properties of mutant forms of theamyloid Aβ peptide of Alzheimer’s disease.J Neurochem 1996;66:740-47.

20. Yang DS, McLaurin J, Qin K, et al. Examining thezinc binding site of the amyloid-beta peptide. Eur JBiochem 2000;267:6692-8.

21. Atwood CS, Moir RD, Huang X, et al. Dramaticaggregation of Alzheimer A-beta by Cu(II) isinduced by conditions representing physiologicalacidosis J Biol Chem 1998;273:12817-26.

22. Cherny RA, Atwood CS, Xilinas ME, et al.Treatment with a copper-zinc chelator markedly andrapidly inhibits beta-amyloid accumulation inAlzheimer’s disease transgenic mice. Neuron2001;30:665-76..

23. Cherny RA, Legg JT, McLean CA, et al. Aqueousdissolution of Alzheimer’s disease A-beta amyloiddeposits by biometal depletion. J Biol Chem1999;274:23223-28

24. Huang X, Atwood CS, Moir RD, et al. Trace metalcontamination initiates the apparent auto-aggregation, amyloidosis, and oligomerization of Alzheimer’s A-beta peptides. J Biol Inorg Chem2004;9:954-60.

25. Basun H, Forssell LG, Wetterberg L, Winblad B.Metals and trace elements in plasma and cere-

brospinal fluid in normal aging and Alzheimer’s dis-ease. J Neural Transm Park Dis Dement Sect1991;3:231-58.

26. Garzon-Rodriguez W, Yatsimirsky AK, Glabe CG.Binding of Zn(II), Cu(II), and Fe(II) ions toAlzheimer’s A beta peptide studied by fluorescence.Bioorg Med Chem Lett 1999;9:2243-8.

27. Syme CD, Nadal RC, Rigby, SEJ, et al. Copper bind-ing to the amyloid-beta (Abeta) peptide associatedwith Alzheimer’s disease: folding, coordinationgeometry, pH dependence, stoichiometry and affinityof Abeta-(1-28) :insights from a range of comple-mentary spectroscopic techniques. J Biol Chem2004;279:18169-77.

28. Zagorski MG, Barrow CJ. NMR studies of amyloidbeta-peptides: proton assignments, secondary struc-ture and mechanism of an alpha-helix-beta-sheetconversion for a homologous, 28-residue, N-terminalfragment. Biochemistry 1992;31:5621-31.

29. Coles M, Bicknell W, Watson AA, et al. Solution

structure of amyloid beta-peptide (1-40) in a water-micelle environment. Is the membrane-spanningdomain where we think it is? Biochemistry1998;37:11064-77.

30. Shao H, Jao S.-C, Ma K, et al. Solution structures of micelle-bound amyloid beta-(1-40) and beta-(1-42)peptides of Alzheimer’s disease. J Mol Biol1999;285:755-73

31. Barrow CJ, Zagorski MG. Solution structures of betapeptide and its constituent fragments: relation toamyloid deposition. Science 1991;253: 179-82

32. Sorimachi K, Craik DJ. Structure determination of extracellular fragments of amyloid proteins involvedin Alzheimer’s disease and Dutch-type hereditarycerebral haemorrhage with amyloidosis. Eur JBiochem 1994;219:237-51

33. Fletcher TG, Keire DA. The interaction of beta-amy-loid protein fragment (12-28) with lipid environ-ments. Protein Sci 1997;6: 666-75

34. Tickler AK, Smith, DG, Ciccotosto, GD, et al.Methylation of imidazole side chains of theAlzheimer’s disease amyloid beta peptide results inabolition of SOD-like structures and inhibition of neurotoxicity. J Biol Chem 2005;280:13355-6335. Jin H, Yong Y, Jun L, et al. The solution structure of rat Abeta-(1-28) and its interaction with zinc ion:insights into the scarcity of amyloid deposition inaged rat brain. J Biol Inorg Chem 2004;9:627-35.

36. Gröbner G, Glaubitz C, Williamson PTF, et al.Structural insight into the interaction of amyloid-betapeptide with biological membranes by solid stateNMR. Focus Struct Biol 2001;1:203-14.

37. Antzutkin ON, Balbach JJ, Leapman RD, et al.Multiple quantum solid-state NMR indicates a paral-

lel, not antiparallel, organization ofβ-sheets inAlzheimer’s beta-amyloid fibrils. Proc Natl Acad SciUSA 2000;97:13045-50

38. Benzinger T, Gregory DM, Burkoth TS, et al.Propagating structure of Alzheimer’s beta-amyloid10-35 is parallel beta-sheet with residues in exactorder. Proc Natl Acad Sci USA 1998;95:13407-12.

39. Balbach JJ, Ishii Y, Antzutkin ON, et al. Amyloidfibril formation by Abeta16-22, a seven residue frag-ment of the Alzheimer’s beta-amyloid peptide, andstructural characterization by solid state NMR.Biochemistry 2000;39:13748-59.

40. Lansbury PT, Costa PR, Griffiths JM, et al. Structuralmodel for the beta-amyloid fibril based on inter-




strand alignment of an antiparallel-sheet comprisinga C-terminal peptide. Nat Struct Biol 1995;2:990-98.

41. Egnaczyk GF Greis KD, Stimson ER, et al.Photoaffinity cross-linking of Alzheimer’s diseaseamyloid fibrils reveals interstrand contact regionsbetween assembled beta-amyloid peptide subunits.

Biochemistry 2001;40:11706-1442. Miura T, Suzuki K, Kohata N, et al. Metal bindingmodes of Alzheimer’s amyloid beta-peptide ininsoluble aggregates and soluble complexes.Biochemistry 2000;39:7024-31.

43. Dong J, Atwood CS, Anderson VE, et al. Metal bind-ing and oxidation of amyloid-beta within isolatedsenile plaque cores: Raman microscopic evidence.Biochemistry 2003;42:2768-73.

44. Huang X, Cuajungco MP, Atwood CS, et al. Cu(II)potentiation of Alzheimer abeta neurotoxicity.Correlation with cell-free hydrogen peroxide pro-duction and metal reduction. J Biol Chem1999;274:37111-16.

45. Peisach J, Blumberg WE. Structural implicationsderived from the analysis of electron paramagneticresonance spectra of natural and artificial copper pro-teins. Arch Biochem Biophys 1974;165:691-708.

46. Curtain CC, Ali F, Volitakis I, Cherny RA, et al.Alzheimer’s disease amyloid-beta binds copper andzinc to generate an allosterically ordered membrane-penetrating structure containing superoxide dismu-tase-like subunits. J Biol Chem 2001;276:20466-73.

47. Alberts IL, Nadassy K, Wodak SJ. Analysis of zincbinding sites in protein crystal structures. Protein Sci1998;7:1700-16

48. Sundberg RJ, Martin RB. Interactions of histidineand other imidazole derivatives with transition metalions in chemical and biological systems. Chem Rev1974;74:471-517.

49. Liu S-T, Howlett G, Barrow CJ. Histidine-13 is a cru-cial residue in the zinc ion-induced aggregation of the Aβ peptide of Alzheimer’s disease. Biochemistry1999;38:9373-78.

50. Parge HE, Hallewell RA, Tainer JA. Atomic struc-tures of wild-type and thermostable mutant recombi-nant human Cu, Zn superoxide dismutase. Proc NatlAcad Sci USA 1992;89:6109-13.

51. Viles JH, Cohen, FE, Prusiner SB, et al. Copperbinding to the prion protein: Structural implicationsof four identical cooperative binding sites. Proc NatlAcad Sci USA 1999;96:2042-47.

52. Brown DR, Wong BS, Hafiz F, et al. Normal prionprotein has an activity like that of superoxide dismu-tase. Biochem J 1999;344:Pt 1:1-5.

53. Antzutkin ON. Amyloidosis of Alzheimer’s Apeptides: solid-state nuclear magnetic resonance,electron paramagnetic resonance, transmission elec-

tron microscopy, scanning transmission electronmicroscopy and atomic force microscopy studies.Magn Reson Chem 2004;42:231-46.

54. Huang X, Atwood CS, Moir RD, et al. Zinc-inducedAlzheimer’s Abeta1-40 aggregation is mediated byconformational factors. J Biol Chem 1997;272:

26464-70.55. Narayanan S, Reif B. Characterization of chemicalexchange between soluble and aggregated states of beta-amyloid by solution-state NMR upon variationof salt conditions. Biochemistry 2005;44:1444-52.

56. Ohtsu H, Shimazaki Y, Odani A, et al. Synthesis andcharacterization of imidazolate-bridged dinuclearcomplexes as active site models of Cu, Zn-SOD.J Am Chem Soc 2000;122:5733-41.

57. Shivers BD, Hilbich C, Multhaup G, et al.Alzheimers-disease amyloidogenic glycoproteinexpression pattern in rat-brain suggests a role in cellcontact. EMBO J 1988;7:1365-70.

58. Karr JW, Akintoye H, Kaupp LJ, Szalai VA.N-Terminal deletions modify the Cu2+ binding sitein amyloid-beta. Biochemistry 2005 12;44:5478-87

59. Karr JW, Kaupp LJ, Szalai VA. Amyloid-beta bindsCu2+ in a mononuclear metal ion binding site J AmChem Soc 2004 20;126:13534-8

60. Roher AE, Chaney MO, Kuo YM, et al. Morphologyand toxicity of Abeta-(1-42) dimer derived from neu-ritic and vascular amyloid deposits of Alzheimer’sdisease. J Biol Chem 1996;271:20631-5.

61. Walsh DM, Klyubin I, Fadeeva JV, et al. Naturallysecreted oligomers of amyloid beta protein potentlyinhibit hippocampal long-term potentiation in vivo.Nature 2002 ;416:535-9.

62. Roher AE, Chaney MO, Kuo YM, et al. Morphologyand toxicity of Abeta-(1-42) dimer derived from neu-ritic and vascular amyloid deposits of Alzheimer’sdisease. J Biol Chem 1996;271:20631-5.

63. Cleary JP, Walsh DM, Hofmeister JJ, et al. Naturaloligomers of the amyloid-beta protein specifi-cally disrupt cognitive function. Nat Neurosci

2005;8:79-8464. Karr JW, Akintoye H, Kaupp LJ, Szalai VA. Copperis implicated in the in vitro formation and toxicity of Alzheimer’s disease amyloid plaques containing thebeta-amyloid (A-beta) peptide. Proc Natl Acad SciUSA 2003;100:11934-40

65. Huang X, Atwood CS, Hartshorn MA, et al. The Abeta peptide of Alzheimer’s disease directly pro-duces hydrogen peroxide through metal ion reduc-tion. Biochemistry 1999;38:7609-16

66. Tabner BJ, Turnbull S, El-Agnaf OM, et al.Formation of hydrogen peroxide and hydroxyl radi-cals from A(beta) and alpha-synuclein as a possiblemechanism of cell death in Alzheimer’s disease and




Parkinson’s disease. Free Radic Biol Med2002;32:1076-83.

67. Atwood CS, Perry G, Zeng H, et al. Copper mediatesdityrosine cross-linking of Alzheimer’s amyloid-beta. Biochemistry 2004;43:560-68.

68. Schoneich C, Williams TD. Cu(II)-catalyzed oxida-

tion of beta-amyloid peptide targets His13 and His14over His6: Detection of 2-Oxo-histidine by HPLC-MS/MS. Chem Res Toxicol 2002;15:717-22.

69. Gunther MR, Peters, JA, Sivaneri MK. Histidinylradical formation in the self-peroxidation reaction of bovine copper-zinc superoxide dismutase. J BiolChem 2002;277:9160–66

70. Alvarez B, Demicheli V, Durán R, Trujillo M, et al.Inactivation of human Cu,Zn superoxide dismutaseby peroxynitrite and formation of histidinyl radicalFree Radic Biol Med 2004;37: 813–22.

71. Barnham KJ, Haeffner F, Ciccotosto GD, et al.Tyrosine gated electron transfer is key to the toxicmechanism of Alzheimer’s diseaseβ-amyloid.FASEB J 2004;18:1427-9.

72. Whittaker, JW. Free radical catalysis by galactoseoxidase. Chem Rev 2003;103:2347-63.

73. Cukier RI, Nocera DG. Proton-coupled electrontransfer. Annu Rev Phys Chem 1998;49:337-69.

74. Davies MJ, Hawkins CL. EPR spin trapping of pro-tein radicals. Free Radic Biol Med 2004;36:1072-86.

75. Choo-Smith LP, Surewicz WK. The interactionbetween Alzheimer amyloid beta(1-40) peptide andganglioside GM1-containing membranes. FEBS Lett1997;402:95-98.

76. McLaurin Jo-A, Franklin T, Fraser PE, et al.Structural transitions associated with the interactionof Alzheimerβ–amyloid peptides with gangliosides.J Biol Chem 1998;273:4506-15.

77. Ji S-R, Wu Y, Sui S-F. Cholesterol is an importantfactor affecting the membrane insertion of beta-amy-loid peptide (A beta 1-40), which may potentiallyinhibit the fibril formation. J Biol Chem2002;277:6273-79.

78. Del Angel VD, Dupuis F, Mornon J-P, et al. Viralfusion peptides and identification of membrane-interacting segments. Biochim Biophys ResCommun 2002;293:1153-60.

79. Curtain CC, Ali FE, Smith DG, et al. Metal ions, pH,and cholesterol regulate the interactions of Alzheimer’s disease amyloid-β peptide withmembrane lipid. J Biol Chem 2003;278:2977-82.

80. Marsh D, Horváth LI Structure, dynamics and com-position of the lipid-protein interface. Perspectives

from spin-labelling. Biochim Biophys Acta1998;1376:267-96.

81. Lin H, Bhatia R, Lal R. Amyloid beta protein formsion channels: implications for Alzheimer’s diseasepathophysiology. FASEB J 2001;15: 2433-44.

82. Horváth LI, Brophy PJ, Marsh D. Exchange rates at

the lipid-protein interface of myelin proteolipid pro-tein studied by spin-label electron spin resonance.Biochemistry 1988;27:46-52.

83. Ulrich AS, Tichelaar W, Förster G, et al.Ultrastructural characterization of peptide-inducedmembrane fusion and peptide self-assembly in thelipid bilayer. Biophys J 1999;77:829-41

84. Kuo YM, Kokjohn TA, Beach TG, et al. Comparativeanalysis of amyloid-beta chemical structure andamyloid plaque morphology of transgenic mouse andAlzheimer’s disease brains. J Biol Chem 2001;276:12991-98.

85. Naslund J, Schierhorn A, Hellman U, et al. Relativeabundance of Alzheimer A beta amyloid peptidevariants in Alzheimer’s disease and normal aging.Proc Natl Acad Sci USA 1994;91:8378-82.

86. Watson AA, Fairlie DP, Craik DJ. Solution structureof methionine oxidized amyloid beta-peptide (1-40).Does oxidation affect conformational switching?Biochemistry 1998;37:12700-06.

87. Palmblad M, Westlind-Danielsson A, BergquistJ. Oxidation of methionine 35 attenuates formationof amyloid beta-peptide 1-40 oligomers. J Biol Chem2002;277:19506-10.88. Hou L, Kang I, Marchant RE, et al. Methionine 35oxidation reduces fibril assembly of the amyloid A-beta-(1-42) peptide of Alzheimer’s disease. J BiolChem 2002;277:40173-76.

89. Barnham KJ, Ciccotosto GD, Tickler AK, et al.Neurotoxic, redox-competent Alzheimer’s beta-amy-loid is released from lipid membrane by methionineoxidation. J Biol Chem 2003;278:42959-65.

90. Ali FE, Separovic F, Barrow CJ, et al. Methionineregulates copper/hydrogen peroxide oxidation prod-

ucts of Abeta. J Pept Sci 2005;11:353-60.91. Boas JF. Electron paramagnetic resonance of copperproteins. In: Lontie R, editor. Copper Proteins andCopper Enzymes. Boca Raton, FL: CRC Press, 1984:5-62.

92. Ciccotosto GD, Tew D, Curtain CC, et al. Enhancedtoxicity and cellular binding of a modified amyloidbeta peptide with a methionine to valine substitution.J Biol Chem 2004;279:42528-34.




9Cholesterol and Alzheimer’s DiseaseJoanna M. Cordy and Benjamin Wolozin

142

9.1 Introduction

Recent studies indicate that cholesterol plays animportant part in the regulation of amyloid- β pep-tide (A β) production, with high cholesterol levelsbeing linked to increased A β generation and depo-sition. The mechanisms underlying the role(s) of cholesterol are not fully understood at present, butfrom the evidence currently available, it appearsthat there are many different ways in which abnor-malities in cholesterol metabolism can affect

the development of Alzheimer’s disease (AD).Polymorphisms in genes involved in cholesterolcatabolism and transport have been associated withan increased level of A β and are therefore potentialrisk factors for the disease. The best known of thesegenes is the apolipoprotein E gene (apoE), whichencodes a protein involved in cholesterol transport.The existence of a particular allele of apoE, ε4, isthe major genetic risk factor known for late-onsetAD. Other genes implicated include cholesterol24-hydroxylase (Cyp46), the LDL receptor relatedprotein (LRP), the cholesterol transporters ABCA1and ABCA2, acyl-CoA:cholesterol acetyl trans-ferase (ACAT), and the LDL receptor (LDLR).

In addition to this genetic evidence, epidemio-logical and biochemical findings also demonstraterelationships between cholesterol and AD and/orAβ. The prevalence of AD has been shown to bereduced among people taking 3-hydroxy-3-methyl-glutaryl (HMG)-CoA reductase inhibitors, such aslovastatin, which inhibit de novo cholesterol syn-

thesis, while levels of serum low-density lipopro-tein (LDL) and total cholesterol have been reportedto correlate with A β levels in the AD brain. These

studies are supported by work on transgenic miceoverexpressing the amyloid precursor protein(APP), demonstrating that increased dietary cho-lesterol results in higher levels of A β, and also byexperiments showing that cholesterol loading ordepletion of cells in culture leads to an increase ordecrease, respectively, in A β production.

In this chapter, all the evidence described abovewill be discussed in more detail to provide a pictureof our current understanding of the ways in whichcholesterol may affect the production of A β and the

development of AD.

9.2 Cholesterol Metabolism

9.2.1 Synthesis

Cholesterol performs many important functionswithin cells, particularly as a structural componentof cell membranes and as a precursor for the gen-eration of steroid hormones and bile salts. It isvital, however, that a balance is maintainedbetween cholesterol synthesis, uptake, and catabo-lism, as an excess of cholesterol is a major risk factor for the development of atherosclerosis.

Within the body, cholesterol is only synthesizedin the liver and brain and is the product of a com-plex multi-enzyme pathway. This pathway beginswith the condensation of acetyl-CoA with ace-toacetyl-CoA to form HMG-CoA. This is thenconverted to mevalonate by HMG-CoA reductase,

in the rate-limiting step of the process [1]. A cas-cade of other reactions then occurs to produce cho-lesterol (Fig. 9.1), and this pathway generates



many intermediate molecules that have importantbiological functions. For example, dolichol, whichis involved in synthesis of the oligosaccharidechains of glycoproteins, and ubiquinone, a compo-nent of the electron transport chain, are bothsynthesized from farnesyl pyrophosphate, a cho-lesterol intermediate.

After synthesis in the endoplasmic reticulum(ER), cholesterol builds up in membranes throughthe Golgi apparatus to the plasma membrane,which has the highest cholesterol content. Withinthese membranes, the distribution of cholesterol isnot uniform, but instead it clusters in regionsknown as lipid rafts, which are also enriched inglycosphingolipids and particular proteins [2–4].These domains will be discussed in more detailbelow.

In addition to the de novo synthesis of choles-terol by the brain and liver, dietary cholesterol canalso be absorbed from the gut. The identity of the

transporter(s) involved in this process is elusive,but one protein recently shown to have a criticalrole is the Niemann-Pick C1 like 1 (NPC1L1) pro-tein [5]. This protein shows ~50% homology toNPC1, the protein that is defective in the choles-terol storage disease Niemann-Pick type C [6].

9.2.2 Transport and Uptake

Cholesterol is insoluble in the blood and thereforemust be transported to and from cells by carriersknown as lipoproteins. Absorbed dietary choles-terol in the intestine is assembled into chylomi-crons, which then enter the bloodstream, whilecholesterol from the liver is released in very-low-density lipoproteins (VLDL). These particles con-tain triacylglycerols, phospholipids, and proteinsknown as apolipoproteins in addition to havingcholesterol. VLDL, LDL, and other lipoproteinscontain varying ratios of protein to lipid and also

9. Cholesterol and Alzheimer’s Disease 143

FIGURE 9.1. The biosynthesis of cholesterol. The synthesis of cholesterol begins with the condensation of acetyl-CoAwith acetoacetyl-CoA, to form HMG-CoA, which is then converted to mevalonate. A cascade of other reactionsoccurs to produce cholesterol and many biologically important intermediate molecules.

CoA+

S

Acetyl CoA Acetoacetyl CoA

HMG-CoA reductase

CoAS

CH3HO

OO

Geranylpyrophosphoric acid

Famesylpyrophosphoric acid Squalene

CholesterolHO

O O

OO

O OP PO 3−

PO3

−P

HOOC

Mevalonic acid

CH2OH



different species of apolipoproteins. ApoB, whichis present in VLDL and LDL, is the most importantlipoprotein in the periphery and is responsible forbinding to the LDL receptor. ApoD, E, and J arealso important, although animals with defects inthe apoD gene show normal cholesterol levels,while cholesterol uptake is impaired if apoB or Eare knocked out [7–9].

After their synthesis in the liver or intestine, bothVLDL and chylomicrons are converted, throughthe loss of triacylglycerol, to LDL, which is the pri-mary carrier of plasma cholesterol to extrahepatictissues. LDL is then taken up into cells via interac-tion with the LDL receptor, and cholesterol isreleased into the cells after degradation of the LDLparticle by lysosomal enzymes.

9.2.3 Storage and Catabolism

Cholesterol within the cell can either be stored asfree cholesterol (FC) in the membrane or it can beconverted to cholesteryl esters (CEs) and stored incytoplasmic droplets. An equilibrium exists

between these two pools of cholesterol controlledby acyl-CoA:cholesterol acyltransferase (ACAT),which catalyzes the formation of CEs from FC.ACAT is activated by a rise in FC levels, and con-versely, low FC levels promote the hydrolysis of CEs back to FC.

An alternative route of elimination of FC fromcells is oxidation. In the periphery, the majority of cholesterol is oxidized at the 7 α position (Fig. 9.2)and is then glycosylated and secreted as bile acids.Oxidation can also occur at the 24 or 27 positionsby the mitochondrial enzymes cholesterol 24 or 27hydroxylase (Cyp46 and Cyp27, respectively).This generates oxysterols, which diffuse from cellsinto the extracellular fluids and vasculature.Oxysterols play an important role in cholesterol

biology by acting as transcriptional regulators.They bind to and activate the liver X receptor(LXR), which then can dimerize with the retinoicacid receptor or retinoic X receptor to stimulatetranscription of genes important in cholesterolmetabolism. Genes regulated by LXR include apoE[10] and the ABCA1 transporter [11].

144 J.M. Cordy and B. Wolozin

Cholesterol

H

12

3

H

4

OH

24-hydroxycholesterol

27-hydroxycholesterol

H

OH7α -hydroxycholesterol

H

CholesterylesterROOC

OH

FIGURE 9.2. Cholesterol catabolism. Cholesterol can be converted into cholesteryl esters by the action of ACAT (1),or alternatively it can be converted into oxysterols by oxidation at the 7 α position by cholesterol 7 α hydroxylase (2),or the 24 or 27 position by Cyp 46 (4) or Cyp 27 (3), respectively.



9.2.4 Cholesterol Metabolism in theBrain

The brain contains approximately 20% of the totalcholesterol in the body, despite only accounting for2% of body mass. The majority of this cholesterolis found in myelin membranes, with some alsopresent in neurons and glial cells. Compared withthe periphery, the turnover of cholesterol in thehuman brain is very slow, with a half-life of almosta year, as opposed to a matter of hours in plasma,and this is largely due to the stability of the myelinsheaths. Most brain cholesterol is synthesizedin situ , and production of cholesterol in the brain islargely independent of plasma cholesterol levels.The extent of regulatory separation between brain

and periphery, though, might differ depending onthe species or conditions. Mice fed a high-lipid dietexhibit increased cholesterol levels in the CNS aswell as in plasma [12]. However, changes in dietarycholesterol do not appear to affect apoE levels [13].

Cholesterol metabolism in the brain differs fromthat in the periphery. Cholesterol is mainly gener-ated in glia and then transported to neurons. Aftersynthesis and secretion from glia via the ABCA1transporter, cholesterol is packaged into lipoproteinparticles resembling HDL. These HDL particlesdiffer from those in the periphery in that they con-tain apoE but no apoB, as occurs in the periphery.HDL is taken up into neurons through recognitionof ApoE by a variety of lipoprotein receptorsincluding the LDL receptor (LDLR), the LDLreceptor related protein (LRP), the apoE receptor,as well as other lipoprotein receptors. Eliminationof cholesterol from the brain occurs mainly via oxi-dation at the 24 and 27 positions to produce a classof compounds termed oxysterols, rather than being

oxidized at the 7 α position by Cyp7a to producebile acids, as occurs in the periphery. The twooxysterols 24(S) hydroxycholesterol and 27hydroxycholesterol are produced by enzymesCyp46 and Cyp27, respectively. As mentionedabove, 24(S) hydroxycholesterol is predominantlymade in the brain, and within the brain, predomi-nantly made by neurons. In contrast, 27 hydroxyc-holesterol is produced by many cells includingneurons and oligodendrocytes [14]. Oxysterols arefar more soluble than cholesterol and diffuse acrossthe blood-brain barrier (BBB) where they enter theperipheral circulation for excretion. Although the

enzymes that represent the first step in bile acidproduction, Cyp7a, is present in the brain, bileacids are not a major mechanism of cholesterolcatabolism in the CNS [15].

9.3 The Genetics of AD andCholesterol Metabolism

9.3.1 ApoE

Three genes associated with early-onset AD havebeen identified to date. These are the APP gene onchromosome 21 [16–18] and the genes encodingpresenilin 1 and 2 on chromosomes 14 and 1,respectively [19–21]. The only gene, however,

that has been unequivocally linked to late-onsetAD is the ApoE gene [22]. This gene, found onchromosome 19, has three common variants, ε2,ε3, and ε4, and it is the presence of the ε4 allele(apoE4) that is the most potent known risk factorfor late-onset AD, after age. The lifetime risk of AD for an individual without the ε4 allele isapprox. 9%, whereas the presence of at least oneε4 allele is believed to increase the risk to approx-imately 29% [23] and also to lower the averageage of onset of the disease [22, 24]. Conversely,the presence of the ε2 allele delays the onset of the disease and is thought to have a protectiveeffect [24].

The strongest hypothesis explaining how apoEimpacts on AD derives from the effects of apoE onAβ deposition and clearance. ApoE is believed toact as a chaperone protein and accelerate the for-mation of A β fibrils [25], with the apoE4 isoformbeing most efficient at promoting fibrillogenesisin vitro (Fig. 9.3) [26]. Results obtained from studies

with transgenic mice also support these data, show-ing that mice expressing apoE4 and APP haveaccelerated A β deposition compared with miceexpressing other apoE isoforms or no apoE [27,28]. More recently, experimental studies demon-strate that blocking the interaction of A β and apoEusing a synthetic peptide not only reduces A β fib-ril formation in vitro but also reduces A β load andplaque formation in a mouse model of AD [29].These studies provide experimental evidence thatthe ability of apoE4 to accelerate Aa aggregationand deposition represents an important mechanismby which apo E4 accelerates the progression of




AD. ApoE is also involved in A β clearance, in anisoform specific manner, with apo E2 and E3, butnot E4 being important for the removal of A β fromthe extracellular space (Fig. 9.3) [30].

The importance of apoE in cholesterol metabo-lism, though, remains a striking phenomenon thatraises the possibility that the presence of differentapoE isoforms may alter cholesterol homeostasis inthe brain and thereby influence the progression of AD. ApoE genotype is known to correlate withplasma cholesterol levels, with apoE4 being asso-ciated with the highest LDL cholesterol levels [31],a believed risk factor for AD [32, 33]. However,whether the association between apoE4 and ADderives from its effect on cholesterol metabolismremains a source of debate. Some studies suggestthat the effects of apolipoprotein E4 on AD are

independent of cholesterol while others show arelationship between cholesterol, apoE, and AD[32, 34–36]. In the periphery, apoE4 appears toassociate predominantly with VLDL particles,which contain a high percentage of cholesterol,whereas apoE2 prefers to associate with the lesscholesterol-rich high-density lipoprotein particles[37–39]. It is not known whether different apoEisoforms associate with different lipid particles inthe brain, but the occurrence of a similar effectcould alter cholesterol metabolism and help toexplain the increased risk of AD associated withapoE4.

9.3.2 Other Genes Linked to Late-OnsetAD and Cholesterol Metabolism

9.3.2.1 Cyp46

Cholesterol 24-hydroxylase, encoded by the Cyp46gene on chromosome 14, is expressed almost exclu-sively in the brain, with only very low levels of mRNA found in other tissues such as liver and testis[40]. The enzyme is a member of the cytochromeP450 family and is responsible for the catabolism of nearly all CNS cholesterol to 24S-hydroxycholes-terol. Knockout of the gene in mice results in adecrease of more than 98% in the level of 24S-hydroxycholesterol in the brain, however total braincholesterol remains unchanged, perhaps becausethere is a compensatory downregulation of de novocholesterol synthesis by approximately 40% [41].

Not surprisingly, knockout of Cyp46 produces noappreciable differences in the levels of peripheralcholesterol and lipoproteins in these mice.

In AD, and in mild cognitive impairment, thelevels of 24S-hydroxycholesterol in cerebral spinalfluid are elevated [42], however other studies sug-gest that plasma levels are decreased or unchanged[43–45]. The reason for the discrepancy might liein the dependence of plasma 24(S) hydroxycholes-terol levels on a variety of factors including diseasestate, cerebral injury, brain size, cerebro-vascularblood flow, and so forth. The integration of all of these factors might produce effects that counteract


Aβ

Apo E4Apo E2/3

LRP

Internalization

FIGURE 9.3. Possible roles of apoE isoforms in amyloid metabolism. The apoE4 isoform accelerates the aggregationand deposition of A β fibrils, whereas the apo E2 and E3 isoforms promote clearance of A β via LRP.



each other and limit the linkage between serum24(S) hydroxycholesterol and Alzheimer’s disease.Our own studies demonstrate that Cyp46 is selec-tively expressed around neuritic plaques, perhapsreflecting the need of neurons to remove excesscholesterol from degenerating neuritis [14].Recently, a number of studies have investigated thelink between polymorphisms in the Cyp46 geneand late-onset AD, with varied results. Two differ-ent intronic polymorphisms with potential associa-tion with AD, A β levels and/or phosphorylated tauhave been identified [46, 47], and these results havesince been corroborated in other populations [48,49]. The genotyping results are ambiguous,though, because other studies have failed to detectany associations between Cyp46 polymorphisms

and AD [50, 51]. These contradictory findingsleave the role of Cyp46 in AD developmentcontroversial.

9.3.2.2 ABCA1

The adenosine triphosphate–binding cassette trans-porter ABCA1 functions to secrete cholesterolfrom the cell and is an important regulator of cho-lesterol metabolism. The gene encoding this pro-tein, on chromosome 9, is another gene with apotential link to AD. In the periphery, ABCA1transports free cholesterol out of cells, and lack of this protein results in reduced plasma HDL levelsand an increased risk of cardiovascular disease[52–54]. Overexpression of the transporter in miceleads to opposite effects [55, 56]. In the brain,ABCA1 is also important in cholesterol trafficking,and it has been shown that its expression incerebral endothelial cells can be stimulated by 24S-hydroxycholesterol, suggesting a role in the

removal of excess brain cholesterol [57].An increasing number of studies suggest that

ABC proteins are important to the pathophysiologyof AD. A polymorphism in the ABCA1 gene,already known to be linked to a modified risk of coronary heart disease [58, 59], has recently beenshown to delay onset of AD by 1.7 years [60], anda larger study has provided further evidence thatvariants of ABCA1 alter the risk of developing AD[61]. ABCA1 has also been shown to directly alterproduction of A β. Transfectng ABCA1 or induct-ing ABCA1 via LXR reduces A β generation, pre-sumably by lowering cholesterol levels [62, 63].

Recently, a second ABC transporter that isexpressed in the brain has been cloned. ABCA2 isexpressed in the endolysosomal compartment, pri-marily in oligodendrocytes, but also in the cortex[64]. When expressed in cell culture ABCA2strongly regulates formation of cholesterol estersand expression of other proteins implicated in cho-lesterol metabolism, such as the LDLR. A poly-morphism in ABCA2 strongly increases the risk of AD, with a LOD score of 3.5 [65]. The associationof two different ABC transporters with AD, com-bined with the direct evidence that these proteinsmodulate A β metabolism, suggests that theseproteins could be particularly relevant to AD.

9.3.2.3 ACAT

Proteins like ABCA2 and Cyp46/LXR modulatemany other proteins important to cholesterol catab-olism or transport. One of these proteins is acyl-Coenzyme A:cholesterol acyl transferase (ACAT),which is a protein that converts cholesterol to cho-lesterol esters, which are highly insoluble and arethought to be used for storage. ACAT could be par-ticularly important for AD because pharmacologi-cal inhibitors of ACAT are available, and theseinhibitors have recently been shown to reduce A βproduction and decrease amyloid load in a trans-genic mouse model of AD [66]. Because relatedcompounds have also been investigated in humanclinical trials and found to be safe, thesecompounds hold great promise for therapy of AD.

9.3.3 LRP and LDLR

LRP is a member of the LDL receptor family and,in brain, is expressed predominantly on neurons

and reactive astrocytes [67, 68]. The main ligandfor LRP in the brain is apoE, although it can alsobind a number of different proteins, includingLDLR, urokinase-type plasminogen activator, andlactoferrin [69]. The fact that LRP is an importantneuronal receptor for apoE, which has long beenimplicated in AD, suggests that this protein mayalso be important in the disease. In addition, LRPand many of its ligands are found in senile plaques[70], suggesting that the function of LRP could beimpaired in AD, resulting in this buildup. Anotherinteresting link between LRP and AD is that it canbind APP and regulate its internalization and




processing [71, 72], thereby potentially affectingproduction of A β, as well as its clearance via apoE.

More evidence for a role for LRP in AD comesfrom genetic association studies. A polymorphismin exon 3 of the gene has been identified, which islinked to reduced AD susceptibility and decreasedamyloid burden [73]. This has since been corrobo-rated by other studies [74–77]. In addition, anotherpolymorphism in the LRP gene has also been iden-tified and linked to AD [78] providing furthergenetic evidence for a connection between LRPand AD. A meta-analysis of LRP polymorphismshas recently been done at the Alzgene website(http://www.alzforum.org/res/com/gen/alzgene/def ault.asp), which suggests a slight increased risk of AD associated with the C allele of the rs1799986

polymorphism. However, the main message pro-vided by the meta-analysis is that the effect of thispolymorphism, if real, is much, much smaller thanthe effect of apoE4.

9.3.4 α2M

One of the ligands for LRP is α-2-macroglobulin(α2M), a protein capable of binding A β with highspecificity [79, 80] and preventing its fibrillization.α2M is found in neuritic plaques in AD brain [81,82] and it may play a role in A β clearance via LRP,as it is known to be able to bind other ligands andtarget them for internalization and degradation[83]. The gene encoding α2M has also been iden-tified as a potential risk factor for AD in some stud-ies, but the overwhelming majority of studies havefailed to observe a linkage [84–86].

9.4 Cholesterol and APPProcessing

9.4.1 In Vitro Studies

A large number of experiments performed oncells in culture demonstrate that cellular process-ing of APP and production of A β can be modu-lated by cholesterol metabolism (Table 9.1). Kleinand colleagues were the first investigators toexamine this issue. They added cholesterol com-plexed with methyl- β-cyclodextrin to the cell lineHEK and demonstrated that the cholesteroldecreased APP secretion [87]. Next, Simons et al.

[88] used a combination of an HMG-CoA reduc-tase inhibitor and methyl- β cyclodextrin todeplete cholesterol levels in hippocampal neuronsby 70%. This caused a dramatic decrease in pro-duction of A β. Later studies using similar treat-ments confirmed these results [89, 90]. Thesystem appears to be reciprocal with respect tocholesterol levels because adding exogenous cho-lesterol to cells in culture upregulates A β produc-tion [89]. The mechanism underlying theregulation appears to depend in part on activity of β-secretase, because cholesterol depletionreduces CTF β [88, 90]. Regulation of APP pro-cessing by cholesterol is not limited to β-secretaseactivity; it appears to occur on multiple levels. Forinstance, α-secretase activity is also controlled by

cholesterol, with low cholesterol levels stimulat-ing production of sAPP α [91]. The third enzymeinvolved in APP processing, γ -secretase, couldalso be affected by cholesterol, as recent work hasshown that disruptions in cholesterol traffickingcause a redistribution of the presenilins and anassociated increase in A β generation [92, 93].However, γ -secretase activity appears to be theleast affected by cholesterol of all the enzymesregulating APP processing.

Cholesterol metabolism can also modulate APPprocessing through trafficking. There are many dif-ferent pools of cholesterol, cholesteryl esters(CEs), or free cholesterol (FC) present in cells. Inaddition, APP processing also occurs in many dif-ferent compartments. Modulation of particularenzymes in particular compartments or modulationof the distribution of APP among different vesiclescan alter generation of A β and APPs. For instance,the enzyme responsible for controlling the inter-conversion of these cholesterol pools is the ER-res-

ident enzyme ACAT, and it has been shown that theactivity of this enzyme can regulate A β generation,suggesting that it may be the distribution of intra-cellular cholesterol that is important rather than thetotal amount [94]. This investigation by Puglielliand co-workers [94] showed that the level of A βwas most closely correlated with cholesteryl esterlevels, although they could not rule out the possi-bility that it may be the ratio of FC to CEs that ismost important. It is likely that other types of cho-lesterol-related modulation also act by changing hevesicular distribution of components that affectAPP processing.




9.4.2 APP Processing and Lipid Rafts

A key to understanding how cholesterol mightmodulate APP processing lies in the concept of lipid rafts. Lipid rafts are small domains within cellmembranes consisting of sphingolipids in the outerleaflet of the bilayer and phospholipids with satu-

rated fatty acid chains in the inner leaflet, tightlypacked together with cholesterol (Fig. 9.4). Thesurrounding bilayer is less tightly packed due to theunsaturated nature of the phospholipid hydrocar-bon chains, with the result that the rafts formordered, although still fluid, platforms within thisliquid-disordered phase (for reviews, see Refs. 2–4,95). As well as containing particular classes of lipids, rafts can bind certain proteins. Different pro-teins are found to be associated with raft domainsto varying extents, for example proteins with a gly-cosylphosphatidyl inositol (GPI) membrane anchorand doubly acylated proteins such as Src family

tyrosine kinases tend to reside in rafts constitu-tively [96], whereas many proteins are able tomove in and out of rafts depending on ligand-bind-ing, oligomerization, or palmitoylation [97, 98].Because of this, the movement of proteins in andout of rafts, and their associations within thesedomains, can be tightly controlled.

Lipid rafts have been hypothesized to beinvolved in APP processing and could thereforehelp to explain how the connection between cho-lesterol and AD occurs [99]. Several proteins rele-vant to A β production have been shown to bepresent in raft domains including a small propor-tion of APP [100–103], the β-secretase BACE ( β-site APP cleaving enzyme) [104, 105], thepresenilins [101, 103, 106], and A β itself [101].These results, which were obtained from severaldifferent cell-lines and from samples of human,mouse, and rat brain, prompted the hypothesis thatamyloidogenic processing of APP may take place


TABLE 9.1. Summary of the effects of cholesterol modulation on amyloid precursor protein (APP) processing andamyloid- β peptide (A β) production.

In vitro / in vivo Modulation of cholesterol Effects Reference

In vitro ↑ Cholesterol Exogenous cholesterol added ↓ sAPP α production 87Exogenous cholesterol added ↑ Aβ production 89

↓ Cholesterol Cholesterol depleted using statin ↓ Aβ production 88

Cholesterol depleted using statin ↓ β-secretase cleavage products 89Cholesterol depleted using statin ↓ Aβ production 90Cholesterol depleted using statin or ↑ sAPP α production 91

methyl- β-cyclodextrin ↓ Aβ production

In vivo ↑ Cholesterol Primates fed high-fat diet ↑ Aβ deposition 108APP Tg mice fed high-fat diet ↑ Aβ deposition 109

Learning impairmentsAPP Tg mice fed high-fat diet ↑ Aβ and CTF β production 12

↓ sAPP α productionAPP Tg mice fed high-fat diet ↑ Aβ deposition 110APP Tg mice fed high-cholesterol diet ↓ Aβ and sAPP β production 112

↓ sAPP α production

APP Tg mice fed high-cholesterol diet ↓ Aβ deposition 113↓ sAPP α production↑ AICD

↓ Cholesterol Guinea pigs treated with simvastatin ↓ Aβ deposition 90APP Tg mice treated with cholesterol-

lowering drug ↓ Aβ and CTF β productionn 111↑ sAPP α production

APP Tg mice treated with lovastatin ↑ Aβ deposition in female mice 138No change in male mice

The in vitro studies suggest that increasing cholesterol levels results in an upregulation of amyloidogenic APP processing, whereaslowering cholesterol levels has the opposite effect. The majority of results from in vivo studies show the same pattern, however thereare some reports (highlighted) that contradict this trend.



within lipid rafts. The putative α-secretaseADAM10, however, is predominantly soluble afterdetergent extraction [91], leading to a model beingproposed in which amyloidogenic and non-amy-loidogenic processing of APP occur in separatemembrane compartments [99]. The existence of two pools of APP within the cell membrane, oneraft-localized and one present in phospholipiddomains [100, 101], fits in with this theory byallowing APP access to both α-secretase and β-and γ -secretases. According to this model of APPcleavage, a high concentration of membrane cho-lesterol would therefore favor A β production,whereas a reduced cholesterol level would favorthe non-amyloidogenic α-secretase pathway.

The studies described above, demonstrating thatdepletion of cellular cholesterol levels results ininhibition of A β production [88–90], support thishypothesis, as cholesterol removal disrupts lipid raft

domains. Further evidence that amyloidogenic APPprocessing, particularly by BACE, occurs in lipidrafts comes from recent work showing that antibodycross-linking of APP and BACE causes them to co-patch with known raft marker proteins, and that thisdramatically increases production of A β [107]. Inaddition, the direct dependence of BACE activity onlipid rafts has been demonstrated by targeting BACEexclusively to these domains using a GPI-anchor[104]. The production of A β and sAPP β wasincreased significantly by targeting BACE to lipidrafts, confirming that this environment is favorablefor the amyloidogenic processing of APP [104].

9.4.3 In Vivo Studies

A number of studies suggest that cholesterol alsomodulates APP processing in vivo (Table 9.1), butwhen interpreting the studies, one must considerthe added complexity of the in vivo situation. Whenanalyzing in vivo and human data, one must distin-guish between plasma cholesterol and cerebralcholesterol because the amount of cross-talk between the two pools of cholesterol and the mech-anism of cross-talk is unclear. One must also dis-tinguish between the type of animal beinginvestigated because lipid metabolism differsamong species such as mice, guinea-pigs, andhumans. For instance, mice generally have highlevels of LDL while humans tend to have higherlevels of HDL.

Despite these differences, several groups haveshown that changes in cholesterol metabolism

induced by pharmacological means (e.g., statins)or by feeding alter cholesterol metabolism. Thishas been shown in primates [108] and transgenicmouse models of AD [12, 109, 110]. For example,Refolo et al. [12] showed that both β-cleaved C-ter-minal APP fragments (CTF β) and A β wereincreased in the CNS of mice fed a high-choles-terol diet, whereas the production of α-cleaved sol-uble APP (sAPP α) was decreased, suggesting thatcholesterol was regulating APP processing. Otherin vivo studies have demonstrated that treatment of guinea-pigs or transgenic mice with cholesterol-lowering drugs resulted in lowered levels of A β


Glycosphingolipids

Transmembrane proteinAcylated protein

Cholesterol

Sphingomyelin

GPI-anchoredprotein

SaturatedPhospholipids

FIGURE 9.4. Schematic diagram of a lipid raft domain. The lipid raft is rich in cholesterol, sphingolipids, and sphin-gomyelin. Lipid-modified proteins such as acylated or GPI-anchored proteins tend to cluster in these regions, alongwith some transmembrane proteins.



[90, 111] and also increased sAPP α and decreasedCTF β production [111]. Each of these studies pres-ents cogent examples of the impact of cholesterolmetabolism on APP processing in vivo.

Although the results from these in vivo studiesindicate that hypercholesterolemia leads to anincrease in the amyloidogenic processing of APP,whereas reduced cholesterol level has the oppositeeffect, some studies have observed contradictoryevidence. Howland et al. [112] examined the effectof a high-cholesterol diet on a different transgenicmouse model of AD and found that levels of sAPP α, sAPP β, and A β were all reduced. Morerecently, another study has shown a similar effect[113]; the reasons for these apparent discrepanciesare not clear. Possible differences that could con-

tribute to these conflicting results could lie in thetransgenes present in the mouse models, thegenetic backgrounds of the mouse models, vari-ability in the ages, or differences in the sex of theanimals studied. Interestingly, the study by Georgeand colleagues [113] demonstrated that productionof the APP intracellular domain (AICD) isincreased in mice fed a high-cholesterol diet. Thisfragment appears to act as a transcriptional activa-tor [114, 115] and can induce apoptosis in neurons[116], leading to the possibility that cholesterolcould affect AD progression via the regulation of AICD production [113].

9.4.4 A β Aggregation and Toxicity

Cholesterol also appears to be important for theaggregation and toxicity of A β. Aggregated or fib-rillar A β is widely believed to be more toxic to neu-rons than the monomeric peptide [117], and there isevidence to suggest that polymerization of A β is

seeded by a species of the peptide that is tightlybound to GM1 ganglioside (GM1-A β) [118]. GM1-Aβ has been shown to accelerate amyloid fibril for-mation in vitro [119, 120], and the formation of thisspecies appears to be sensitive to the lipid environ-ment, with cholesterol being an important factor[121]. Kakio et al. [122] demonstrated that A βbound preferentially to clusters of GM1 moleculesand that these clusters formed in cholesterol-richenvironments such as lipid rafts, and this is sup-ported by a study reporting that depletion of cellu-lar cholesterol can protect cells from the toxiceffects of A β [123]. More recently, Subasinghe and

colleagues [124] have shown that binding of A β tomembrane lipids is important for toxicity of thepeptide and that both membrane-binding and toxic-ity were reduced by the removal of cholesterol.

9.5 Epidemiological and ClinicalEvidence

9.5.1 Cholesterol Levels and AD

Despite the strong genetic and biochemical evi-dence that points to a strong connection betweencholesterol and AD, epidemiological evidencelinking plasma levels of cholesterol and lipopro-teins with the development of AD is conflicting.

Some studies have demonstrated a link betweencholesterol level, particularly in mid-life, and AD.For example, Pappolla and colleagues [125] foundthat there was a strong correlation between totalcholesterol level and amyloid deposition in sub-

jects aged between 40 and 55 years, but this corre-lation became weaker as the age of the subjectsincreased. In another study, Finnish men who haddisplayed a high serum cholesterol level at age40–59 were found to be three times more likely tohave developed AD 30 years later [35]. Kivipeltoet al. [126, 127] also demonstrated a correlationbetween mid-life cholesterol level and the risk of developing AD later in life. These results, and thefact that in the study by Notkola et al. [35] the cho-lesterol level of men who developed AD decreasedbefore the disease manifested itself, suggest thathypercholesterolemia in mid-life could be a risk factor for AD, while cholesterol level in later lifeshows less correlation with the disease. Kuo et al.[33], however, examined serum levels of LDL and

HDL cholesterol at postmortem and found signifi-cantly higher LDL cholesterol and lower HDL cho-lesterol in AD patients than in control subjects.

In contrast with these studies, which have foundcorrelations between cholesterol levels and AD,other investigations have failed to find such a con-nection. Tan et al. [128] looked at total serum cho-lesterol levels from participants in the Framinghamstudy and found no association between average cho-lesterol level over a 30-year period and developmentof AD 10–20 years later. Another study investigatinga wide variety of serum markers in neurodegenera-tive diseases also found no correlation between




serum cholesterol and AD [129], although, interest-ingly, the levels of precursors to cholesterol synthesisappeared to be significantly different in AD patientscompared with controls.

9.5.2 Use of StatinsAn alternate approach to addressing the issue of cholesterol and AD is to shift the question fromwhether abnormal cholesterol metabolismincreases the risk of AD to the question of whethermodulating cholesterol metabolism can alter theincidence or progression of AD. HMG-CoA reduc-tase inhibitors, known collectively as statins, weredeveloped in the 1970s and have been widely usedsince the late 1980s to lower cholesterol levels in

patients at risk of coronary heart disease. Examplesof statins that are currently available include lovas-tatin (Mevacor, currently off patent), pravastatin(Pravacor), simvastatin (Zocor), rosuvastatin(Crestor), and atorvastatin (Lipitor). In 2000, tworetrospective studies suggested that the prevalenceof AD was reduced by approximately 70% amongpatients taking statins compared with control sub-

jects [130, 131]. Similar studies have since corrob-orated these findings in different groups of patients[132, 133].

More variable results have been obtained byprospective studies examining the use of statins aspotential therapeutic agents in AD. Simons et al.[134] observed a decrease in the CSF A β40 levels of patients suffering from mild AD after treatmentwith simvastatin for 26 weeks, but this was notseen in patients with a more severe form of the dis-ease. Cognitive decline appeared to be slowed inboth groups compared with subjects receiving aplacebo. Another small study of AD patients found

that CSF levels of sAPP α and sAPP β weredecreased after a 12-week treatment with simvas-tatin, but A β42 levels were unaltered [135]. Twolarger studies, looking primarily at the cardiovas-cular benefits of longer term (3- to 5-year) statintreatment, found that cognitive decline was notprevented by statins [136, 137], however, a recentpilot study of the effects of atorvastatin, reported atthe American Heart Association’s ScientificSessions 2004, has shown that it appears to slowmental decline and improve cognitive symptoms inAD patients (www.americanheart.org). These stud-ies have used a variety of statins with differing

lipophilicities, suggesting that the variable resultscannot be explained by the ability of the drug tocross the blood-brain barrier (BBB). The reason forthe mixed results obtained is unknown but have todo with the severity of AD or the cholesterol levelin the patients examined or the methods used to testfor cognitive function. Other clinical trials of statins in AD, such as the Cholesterol LoweringAgent to Slow Progression (CLASP) of AD Study,sponsored by the NIA, are currently in progress, sothese should provide more information about thepossible therapeutic benefits of these drugs.

9.6 Future Directions

Despite the current interest in determining theassociation between cholesterol and AD, there arestill many crucial questions that need to beaddressed before a complete picture of this com-plex relationship emerges. The effects of statins onAβ production appear to be clear in cell culture, butthe effects in vivo and the role of cholesterol in thepathogenesis of AD are by no means clear-cut, andif these drugs are to be used in the treatment of AD,many issues still need to be resolved. One impor-tant factor that has recently come to light is a pos-sible gender-related difference in response to statintreatment. When male and female APP transgenicmice were treated with lovastatin, both groupsshowed the expected reduction in cholesterol lev-els, but female mice showed an increase in both A βproduction and plaque load [138]. No changeswere seen in the male mice. These results suggestthat it will be important to reexamine the resultsfrom other studies and trials involving statins, totake into account gender differences. Another issue

that is currently being investigated is whether theneuroprotective effects of statins are due less totheir role as inhibitors of cholesterol synthesis andmore to other effects such as their anti-inflamma-tory properties [139, 140].

The fact that ageing leads to alterations in thelipid and cholesterol distribution within mem-branes could affect the number and stability of lipid rafts. Currently, however, no data existregarding changes in raft number, size, or compo-sition during aging or AD progression. If this issuecould be addressed, the results would be valuablein assessing exactly how lipid rafts are involved in




APP processing. Unfortunately, native rafts arevery difficult to study, as detergent isolation cancause individual rafts to coalesce [141] providingan inaccurate picture of the actual organization of rafts within the membrane. The development of new technologies to study lipid rafts may berequired before this question can be answeredsatisfactorily.

Despite all of these questions, there continues tobe a great deal of promise for cholesterol modula-tion in therapy of AD. Whether statins modulate A βin vivo remains a question, but increasing data sug-gest that statins have potent anti-inflammatory prop-erties, which could be valuable in treating AD [142].Other means of modulating cholesterol metabolismalso appear to be promising. For instance, ACAT

inhibitors appear to be very effective in reducing A βand plaque load in vivo. Other matters that requirefurther investigation include the relationshipbetween plasma and brain cholesterol. A betterunderstanding of brain cholesterol metabolism isrequired to clarify how modulating plasma choles-terol using diet or drugs could affect A β productionor deposition in the brain. In addition, the contribu-tion of different forms of cholesterol, free choles-terol, or cholesteryl esters, to the overall effect of cholesterol in AD needs to be examined further.

9.7 Conclusions

Over the past few years, an increasing amount of evidence has accumulated suggesting that choles-terol metabolism is strongly connected to thedevelopment of Alzheimer’s disease. This evidenceincludes studies showing linkages between genesinvolved in cholesterol metabolism, such as apoE

and cyp46, and AD and epidemiological evidencethat drugs aimed at lowering cholesterol levels maybe useful for treating AD. Additionally, there are alarge number of biochemical studies indicating thatcholesterol is involved in APP processing, possiblyby providing a favorable membrane environment inwhich the amyloidogenic secretase enzymes canact, and also in A β aggregation and toxicity. Thisevidence has led to the possibility that drugs affect-ing cholesterol metabolism, such as statins andACAT inhibitors, or the modulation of cholesterollevels by dietary control, may be beneficial in thetreatment of AD.

Despite this growing amount of evidence, we donot currently have a clear picture of the relation-ships between cholesterol and AD, and more work is needed to confirm the importance of cholesterolin the progression of the disease and to elucidatethe molecular basis of the relationship. Theadvances in our knowledge that will surely comeover the next few years may lead to the develop-ment of new strategies for both prevention andtreatment of Alzheimer’s disease.

References

1. Bloch K. The biological synthesis of cholesterol.Science 1965;150(692):19-28.

2. Hooper NM. Detergent-insoluble glycosphin-

golipid/cholesterol-rich membrane domains, lipidrafts and caveolae (review). Mol Membr Biol1999;16(2):145-56.

3. Simons K, Ikonen E. Functional rafts in cell mem-branes. Nature 1997;387(6633):569-72.

4. Simons K, Ikonen E. How cells handle cholesterol.Science 2000;290(5497):1721-6.

5. Altmann SW, Davis HR Jr, Zhu LJ, et al. Niemann-Pick C1 Like 1 protein is critical for intestinal cho-lesterol absorption. Science 2004;303(5661):1201-4.

6. Davies JP, Levy B, Ioannou YA. Evidence for aNiemann-pick C (NPC) gene family: identificationand characterization of NPC1L1. Genomics 2000;65(2):137-45.

7. Homanics GE, Smith TJ, Zhang SH, et al. Targetedmodification of the apolipoprotein B gene results inhypobetalipoproteinemia and developmental abnor-malities in mice. Proc Natl Acad Sci U S A 1993;90(6):2389-93.

8. Ishibashi S, Herz J, Maeda N, et al. The two-recep-tor model of lipoprotein clearance: tests of thehypothesis in “knockout” mice lacking the low den-sity lipoprotein receptor, apolipoprotein E, or both

proteins. Proc Natl Acad Sci U S A 1994;91(10):4431-5.

9. Srivastava RA, Toth L, Srivastava N, et al.Regulation of the apolipoprotein B in heterozygoushypobetalipoproteinemic knock-out mice expressingtruncated apoB, B81. Low production and enhancedclearance of apoB cause low levels of apoB. Mol CellBiochem 1999;202(1-2):37-46.

10. Laffitte BA, Repa JJ, Joseph SB, et al. LXRs controllipid-inducible expression of the apolipoprotein Egene in macrophages and adipocytes. Proc Natl Acad

Sci U S A 2001;98(2):507-12.11. Venkateswaran A, Laffitte BA, Joseph SB, et al.Control of cellular cholesterol efflux by the nuclear




oxysterol receptor LXR alpha. Proc Natl Acad SciU S A 2000;97(22):12097-102.

12. Refolo LM, Pappolla MA, Malester B, et al.Hypercholesterolemia accelerates the Alzheimer’samyloid pathology in a transgenic mouse model.Neurobiol Dis 2000;7(4):321-31.

13. Lafarga M, Crespo P, Berciano MT, et al.Apolipoprotein E expression in the cerebellum of normal and hypercholesterolemic rabbits. Brain ResMol Brain Res 1994;21(1-2):115-23.

14. Brown J, Theisler C, Silberman S, et al. Differentialexpression of cholesterol hydroxylases in Alzheimer’sdisease. J Biol Chem 2004;279:34674-81.

15. Lathe R. Steroid and sterol 7-hydroxylation: ancientpathways. Steroids 2002;67(12):967-77.

16. Chartier-Harlin M, Crawford F, Houlden H, et al.Early-onset Alzheimer’s disease caused by mutationsat codon 717 of the a-amyloid precursor proteingene. Nature 1991;353:844-6.

17. Goate A, Chartier-Harlin MC, Mullan M, et al.Segregation of a missence mutation in the amyloidprecursor protein gene with familial Alzheimer’s dis-ease. Nature 1991;349:704-6.

18. Murrell J, Farlow M, Ghetti B, Benson M. A mutationin the amyloid precursor protein associated with hered-itary Alzheimer’s disease. Science 1991;254:97-9.

19. Levy-Lahad E, Wasco W, Poorkaj P, et al. Candidategene for the chromosome 1 familial Alzheimer’s dis-ease locus. Science 1995;269(5226):973-7.

20. Rogaev EI, Sherrington R, Rogaeva EA, et al.Familial Alzheimer’s disease in kindreds with mis-sense mutations in a gene on chromosome 1 relatedto the Alzheimer’s disease type 3 gene. Nature1995;376(6543):775-8.

21. Sherrington R, Rogaev E, Liang Y, et al. Cloning of agene bearing missense mutations in early-onset famil-ial Alzheimer’s disease. Nature 1995;375:754-60.

22. Corder E, Saunders A, Strittmatter W, et al. Genedose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science

1993;261:921-3.23. Seshadri S, Drachman DA, Lippa CF. ApolipoproteinE epsilon 4 allele and the lifetime risk of Alzheimer’sdisease. What physicians know, and what they shouldknow. Arch Neurol 1995;52(11):1074-9.

24. Strittmatter WJ, Roses AD. Apolipoprotein E andAlzheimer’s disease. Proc Natl Acad Sci U S A1995;92(11):4725-7.

25. Tomiyama T, Corder EH, Mori H. Molecular patho-genesis of apolipoprotein E-mediated amyloidosis inlate-onset Alzheimer’s disease. Cell Mol Life Sci1999;56(3-4):268-79.

26. Wisniewski T, Castano EM, Golabek A, et al.Acceleration of Alzheimer’s fibril formation by

apolipoprotein E in vitro. Am J Pathol 1994;145(5):1030-5.

27. Carter DB, Dunn E, McKinley DD, et al. Humanapolipoprotein E4 accelerates beta-amyloid deposi-tion in APPsw transgenic mouse brain. Ann Neurol2001;50(4):468-75.

28. Holtzman DM, Bales KR, Tenkova T, et al.Apolipoprotein E isoform-dependent amyloid depo-sition and neuritic degeneration in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci U S A2000;97(6):2892-7.

29. Sadowski M, Pankiewicz J, Scholtzova H, et al. A syn-thetic peptide blocking the apolipoprotein E/beta-amy-loid binding mitigates beta-amyloid toxicity and fibrilformation in vitro and reduces beta-amyloid plaques intransgenic mice. Am J Pathol 2004;165(3):937-48.

30. Yang DS, Small DH, Seydel U, et al. ApolipoproteinE promotes the binding and uptake of beta-amyloidinto Chinese hamster ovary cells in an isoform-spe-cific manner. Neuroscience 1999;90(4):1217-26.

31. Menzel HJ, Kladetzky RG, Assmann G.Apolipoprotein E polymorphism and coronary arterydisease. Arteriosclerosis 1983;3(4):310-5.

32. Jarvik GP, Wijsman EM, Kukull WA, et al.Interactions of apolipoprotein E genotype, total cho-lesterol level, age, and sex in prediction of Alzheimer’s disease: a case-control study. Neurology1995;45(6):1092-6.

33. Kuo YM, Emmerling MR, Bisgaier CL, et al.

Elevated low-density lipoprotein in Alzheimer’s dis-ease correlates with brain abeta 1-42 levels. BiochemBiophys Res Commun 1998;252(3):711-5.

34. Evans RM, Hui S, Perkins A, et al. Cholesterol andAPOE genotype interact to influence Alzheimer’s dis-ease progression. Neurology 2004;62(10):1869-71.

35. Notkola IL, Sulkava R, Pekkanen J, et al. Serum totalcholesterol, apolipoprotein E epsilon 4 allele,and Alzheimer’s disease. Neuroepidemiology1998;17(1):14-20.

36. Prince M, Lovestone S, Cervilla J, et al. The associ-

ation between APOE and dementia does not seem tobe mediated by vascular factors. Neurology2000;54(2):397-402.

37. Gregg RE, Zech LA, Schaefer EJ, et al. Abnormalin vivo metabolism of apolipoprotein E4 in humans.J Clin Invest 1986;78(3):815-21.

38. Steinmetz A, Jakobs C, Motzny S, Kaffarnik H.Differential distribution of apolipoprotein E isoformsin human plasma lipoproteins. Arteriosclerosis1989;9(3):405-11.

39. Weisgraber KH. Apolipoprotein E distributionamong human plasma lipoproteins: role of the cys-teine-arginine interchange at residue 112. J Lipid Res1990;31(8):1503-11.




40. Lund EG, Guileyardo JM, Russell DW. cDNAcloning of cholesterol 24-hydroxylase, a mediator of cholesterol homeostasis in the brain. Proc Natl AcadSci U S A 1999;96(13):7238-43.

41. Lund EG, Xie C, Kotti T, et al. Knockout of the cho-lesterol 24-hydroxylase gene in mice reveals a brain-

specific mechanism of cholesterol turnover. J BiolChem 2003;278(25):22980-8.

42. Papassotiropoulos A, Lutjohann D, Bagli M, et al.24S-hydroxycholesterol in cerebrospinal fluid is ele-vated in early stages of dementia. J Psychiatr Res2002;36(1):27-32.

43. Bretillon L, Lutjohann D, Stahle L, et al. Plasma lev-els of 24S-hydroxycholesterol reflect the balancebetween cerebral production and hepatic metabolismand are inversely related to body surface. J Lipid Res2000;41(5):840-5.

44. Heverin M, Bogdanovic N, Lutjohann D, et al.Changes in the levels of cerebral and extracerebralsterols in the brain of patients with Alzheimer’s dis-ease. J Lipid Res 2004;45(1):186-93.

45. Kolsch H, Heun R, Kerksiek A, et al. Altered levelsof plasma 24S- and 27-hydroxycholesterol indemented patients. Neurosci Lett 2004;368(3):303-8.

46. Kolsch H, Lutjohann D, Ludwig M, et al.Polymorphism in the cholesterol 24S-hydroxylasegene is associated with Alzheimer’s disease. MolPsychiatry 2002;7(8):899-902.

47. Papassotiropoulos A, Streffer JR, Tsolaki M, et al.

Increased brain beta-amyloid load, phosphorylatedtau, and risk of Alzheimer’s disease associated withan intronic CYP46 polymorphism. Arch Neurol2003;60(1):29-35.

48. Borroni B, Archetti S, Agosti C, et al. Intronic CYP46polymorphism along with ApoE genotype in sporadicAlzheimer’s Disease: from risk factors to diseasemodulators. Neurobiol Aging 2004;25(6):747-51.

49. Johansson A, Katzov H, Zetterberg H, et al. Variantsof CYP46A1 may interact with age and APOE toinfluence CSF Abeta42 levels in Alzheimer’s dis-

ease. Hum Genet 2004;114(6):581-7.50. Desai P, DeKosky ST, Kamboh MI. Genetic variationin the cholesterol 24-hydroxylase (CYP46) gene andthe risk of Alzheimer’s disease. Neurosci Lett2002;328(1):9-12.

51. Kabbara A, Payet N, Cottel D, et al. Exclusion of CYP46 and APOM as candidate genes forAlzheimer’s disease in a French population.Neurosci Lett 2004;363(2):139-43.

52. Bodzioch M, Orso E, Klucken J, et al. The gene encod-ing ATP-binding cassette transporter 1 is mutated inTangier disease. Nat Genet 1999;22(4):347-51.

53. Brooks-Wilson A, Marcil M, Clee SM, et al.Mutations in ABC1 in Tangier disease and familial

high-density lipoprotein deficiency. Nat Genet1999;22(4):336-45.

54. Rust S, Rosier M, Funke H, et al. Tangier disease iscaused by mutations in the gene encoding ATP-bindingcassette transporter 1. Nat Genet 1999;22(4):352-5.

55. Singaraja RR, Bocher V, James ER, et al. Human

ABCA1 BAC transgenic mice show increased highdensity lipoprotein cholesterol and ApoAI-depend-ent efflux stimulated by an internal promoter con-taining liver X receptor response elements in intron1. J Biol Chem 2001;276(36):33969-79.

56. Singaraja RR, Fievet C, Castro G, et al. IncreasedABCA1 activity protects against atherosclerosis. JClin Invest 2002;110(1):35-42.

57. Panzenboeck U, Balazs Z, Sovic A, et al. ABCA1 andscavenger receptor class B, type I, are modulators of reverse sterol transport at an in vitro blood-brain bar-rier constituted of porcine brain capillary endothelialcells. J Biol Chem 2002;277(45):42781-9.

58. Clee SM, Kastelein JJ, van Dam M, et al. Age andresidual cholesterol efflux affect HDL cholesterollevels and coronary artery disease in ABCA1 het-erozygotes. J Clin Invest 2000;106(10):1263-70.

59. Clee SM, Zwinderman AH, Engert JC, et al. Commongenetic variation in ABCA1 is associated with alteredlipoprotein levels and a modified risk for coronaryartery disease. Circulation 2001;103(9):1198-205.

60. Wollmer MA, Streffer JR, Lutjohann D, et al.ABCA1 modulates CSF cholesterol levels and influ-

ences the age at onset of Alzheimer’s disease.Neurobiol Aging 2003;24(3):421-6.

61. Katzov H, Chalmers K, Palmgren J, et al. Geneticvariants of ABCA1 modify Alzheimer’s disease risk and quantitative traits related to beta-amyloid metab-olism. Hum Mutat 2004;23(4):358-67.

62. Koldamova RP, Lefterov IM, Ikonomovic MD, et al.22R-hydroxycholesterol and 9-cis-retinoic acid induceATP-binding cassette transporter A1 expression andcholesterol efflux in brain cells and decrease amyloidbeta secretion. J Biol Chem 2003;278(15):13244-56.

63. Sun Y, Yao J, Kim TW, Tall AR. Expression of liver X receptor target genes decreases cellular amy-loid beta peptide secretion. J Biol Chem2003;278(30):27688-94.

64. Chen ZJ, Vulevic B, Ile KE, et al. Association of ABCA2 expression with determinants of Alzheimer’s disease. FASEB J 2004;18(10):1129-31.

65. Mace S, Cousin E, Ricard S, et al. ABCA2 is a stronggenetic risk factor for early-onset Alzheimer’s dis-ease. Neurobiol Dis 2005;18(1):119-25.

66. Hutter-Paier B, Huttunen HJ, Puglielli L, et al. TheACAT inhibitor CP-113,818 markedly reduces amy-loid pathology in a mouse model of Alzheimer’sdisease. Neuron 2004;44(2):227-38.




67. Bu G, Maksymovitch EA, Nerbonne JM, SchwartzAL. Expression and function of the low densitylipoprotein receptor-related protein (LRP) inmammalian central neurons. J Biol Chem 1994;269(28):18521-8.

68. Ishiguro M, Imai Y, Kohsaka S. Expression and dis-

tribution of low density lipoprotein receptor-relatedprotein mRNA in the rat central nervous system.Brain Res Mol Brain Res 1995;33(1):37-46.

69. Strickland DK, Kounnas MZ, Argraves WS. LDLreceptor-related protein: a multiligand receptor forlipoprotein and proteinase catabolism. FASEB J1995;9(10):890-8.

70. Rebeck GW, Harr SD, Strickland DK, Hyman BT.Multiple, diverse senile plaque-associated proteinsare ligands of an apolipoprotein E receptor, thealpha 2-macroglobulin receptor/low-density-lipoprotein receptor-related protein. Ann Neurol1995;37(2):211-7.

71. Kounnas MZ, Moir RD, Rebeck GW, et al. LDLreceptor-related protein, a multifunctional ApoEreceptor, binds secreted beta-amyloid precursorprotein and mediates its degradation. Cell1995;82(2):331-40.

72. Ulery PG, Beers J, Mikhailenko I, et al. Modulationof beta-amyloid precursor protein processing by thelow density lipoprotein receptor-related protein(LRP). Evidence that LRP contributes to the patho-genesis of Alzheimer’s disease. J Biol Chem

2000;275(10):7410-5.73. Kang DE, Saitoh T, Chen X, et al. Genetic associa-

tion of the low-density lipoprotein receptor-relatedprotein gene (LRP), an apolipoprotein E receptor,with late-onset Alzheimer’s disease. Neurology1997;49(1):56-61.

74. Baum L, Chen L, Ng HK, et al. Low density lipopro-tein receptor related protein gene exon 3 polymor-phism association with Alzheimer’s disease inChinese. Neurosci Lett 1998;247(1):33-6.

75. Hollenbach E, Ackermann S, Hyman BT, Rebeck

GW. Confirmation of an association between a poly-morphism in exon 3 of the low-density lipoproteinreceptor-related protein gene and Alzheimer’s dis-ease. Neurology 1998;50(6):1905-7.

76. Kolsch H, Ptok U, Mohamed I, et al. Association of the C766T polymorphism of the low-density lipopro-tein receptor-related protein gene with Alzheimer’sdisease. Am J Med Genet 2003;121B(1):128-30.

77. Wavrant-DeVrieze F, Perez-Tur J, Lambert JC, et al.Association between the low density lipoproteinreceptor-related protein (LRP) and Alzheimer’s dis-ease. Neurosci Lett 1997;227(1):68-70.

78. Van Leuven F, Thiry E, Stas L, Nelissen B. Analysisof the human LRPAP1 gene coding for the lipopro-

tein receptor-associated protein: identification of 22polymorphisms and one mutation. Genomics1998;52(2):145-51.

79. Du Y, Ni B, Glinn M, et al. alpha2-Macroglobulin asa beta-amyloid peptide-binding plasma protein.J Neurochem 1997;69(1):299-305.

80. Hughes SR, Khorkova O, Goyal S, et al. Alpha2-macroglobulin associates with beta-amyloid peptideand prevents fibril formation. Proc Natl Acad SciU S A 1998;95(6):3275-80.

81. Bauer J, Strauss S, Schreiter-Gasser U, et al.Interleukin-6 and alpha-2-macroglobulin indicate anacute-phase state in Alzheimer’s disease cortices.FEBS Lett 1991;285(1):111-4.

82. Van Gool D, De Strooper B, Van Leuven F, et al.alpha 2-Macroglobulin expression in neuritic-typeplaques in patients with Alzheimer’s disease.Neurobiol Aging 1993;14(3):233-7.

83. Borth W. Alpha 2-macroglobulin, a multifunctionalbinding protein with targeting characteristics.FASEB J 1992;6(15):3345-53.

84. Blacker D, Wilcox M, Laird N, et al. Alpha-2macroglobulin is genetically associated withAlzheimer’s disease. Nat Gen 1998;19:357-60.

85. Myllykangas L, Polvikoski T, Sulkava R, et al.Genetic association of alpha2-macroglobulin withAlzheimer’s disease in a Finnish elderly population.Ann Neurol 1999;46(3):382-90.

86. Rogaeva EA, Premkumar S, Grubber J, et al. An

alpha-2-macroglobulin insertion-deletion polymor-phism in Alzheimer’s disease. Nat Genet1999;22(1):19-22.

87. Bodovitz S, Klein WL. Cholesterol modulates alpha-secretase cleavage of amyloid precursor protein.J Biol Chem 1996;271(8):4436-40.

88. Simons M, Keller P, De Strooper B, et al. Cholesteroldepletion inhibits the generation of β-amyloid in hip-pocampal neurons. Proc Natl Acad Sci U S A1998;95:6460-4.

89. Frears ER, Stephens DJ, Walters CE, et al. The role

of cholesterol in the biosynthesis of beta-amyloid.Neuroreport 1999;10(8):1699-705.90. Fassbender K, Simons M, Bergmann C, et al.

Simvastatin strongly reduces Alzheimer’s diseaseAβ42 and A β40 levels in vitro and in vivo. Proc NatlAcad Sci U S A 2001;98:5856-61.

91. Kojro E, Gimpl G, Lammich S, et al. Low cholesterolstimulates the nonamyloidogenic pathway by itseffect on the alpha-secretase ADAM 10. Proc NatlAcad Sci U S A 2001;98(10):5815-20.

92. Burns M, Gaynor K, Olm V, et al. Presenilin redistri-bution associated with aberrant cholesterol transportenhances beta-amyloid production in vivo. JNeurosci 2003;23(13):5645-9.




seed for Alzheimer’s amyloid. Biochemistry2002;41(23):7385-90.

121. Mizuno T, Nakata M, Naiki H, et al. Cholesterol-dependent generation of a seeding amyloid beta-protein in cell culture. J Biol Chem1999;274(21):15110-4.

122. Kakio A, Nishimoto SI, Yanagisawa K, et al.Cholesterol-dependent formation of GM1 ganglio-side-bound amyloid beta-protein, an endogenousseed for Alzheimer’s amyloid. J Biol Chem2001;276(27):24985-90.

123. Wang SS, Rymer DL, Good TA. Reduction in cho-lesterol and sialic acid content protects cells fromthe toxic effects of beta-amyloid peptides. J BiolChem 2001;276(45):42027-34.

124. Subasinghe S, Unabia S, Barrow CJ, et al.Cholesterol is necessary both for the toxic effect of Abeta peptides on vascular smooth muscle cells andfor Abeta binding to vascular smooth muscle cellmembranes. J Neurochem 2003;84(3):471-9.

125. Pappolla MA, Bryant-Thomas TK, Herbert D, et al.Mild hypercholesterolemia is an early risk factor forthe development of Alzheimer’s amyloid pathology.Neurology 2003;61(2):199-205.

126. Kivipelto M, Helkala EL, Laakso MP, et al. Midlifevascular risk factors and Alzheimer’s disease inlater life: longitudinal, population based study. BrMed J 2001;322(7300):1447-51.

127. Kivipelto M, Helkala EL, Laakso MP, et al.

Apolipoprotein E epsilon4 allele, elevated midlifetotal cholesterol level, and high midlife systolicblood pressure are independent risk factors for late-life Alzheimer’s disease. Ann Intern Med2002;137(3):149-55.

128. Tan ZS, Seshadri S, Beiser A, et al. Plasma totalcholesterol level as a risk factor for Alzheimer’s dis-ease: the Framingham Study. Arch Intern Med2003;163(9):1053-7.

129. Teunissen CE, De Vente J, von Bergmann K, et al.Serum cholesterol, precursors and metabolites and

cognitive performance in an aging population.Neurobiol Aging 2003;24(1):147-55.130. Jick H, Zornberg GL, Jick SS, et al. Statins and the

risk of dementia. Lancet 2000;356(9242):1627-31.131. Wolozin B, Kellman W, Ruosseau P, et al. Decreased

prevalence of Alzheimer’s disease associated with 3-hydroxy-3-methyglutaryl coenzyme A reductaseinhibitors. Arch Neurol 2000;57(10):1439-43.

132. Rockwood K, Kirkland S, Hogan DB, et al. Use of lipid-lowering agents, indication bias, and the risk of dementia in community-dwelling elderly people.Arch Neurol 2002;59(2):223-7.

133. Yaffe K, Barrett-Connor E, Lin F, Grady D. Serumlipoprotein levels, statin use, and cognitive function

in older women. Arch Neurol 2002;59(3):378-84.134. Simons M, Schwarzler F, Lutjohann D, et al.

Treatment with simvastatin in normocholes-terolemic patients with Alzheimer’s disease: a 26-week randomized, placebo-controlled, double-blindtrial. Ann Neurol 2002;52(3):346-50.

135. Sjogren M, Gustafsson K, Syversen S, et al.Treatment with simvastatin in patients withAlzheimer’s disease lowers both alpha- and beta-cleaved amyloid precursor protein. Dement GeriatrCogn Disord 2003;16(1):25-30.

136. Heart Protection Study Collaborative Group.MRC/BHF Heart Protection Study of cholesterollowering with simvastatin in 20,536 high-risk indi-viduals: a randomised placebo-controlled trial.Lancet 2002;360(9326):7-22.

137. Shepherd J, Blauw GJ, Murphy MB, et al.Pravastatin in elderly individuals at risk of vasculardisease (PROSPER): a randomised controlled trial.Lancet 2002;360(9346):1623-30.

138. Park IH, Hwang EM, Hong HS, et al. Lovastatinenhances Abeta production and senile plaque depo-sition in female Tg2576 mice. Neurobiol Aging

2003;24(5):637-43.139. Grip O, Janciauskiene S, Lindgren S. Pravastatin

down-regulates inflammatory mediators inhuman monocytes in vitro. Eur J Pharmacol2000;410(1):83-92.

140. Ortego M, Bustos C, Hernandez-Presa MA, et al.Atorvastatin reduces NF-kappaB activation andchemokine expression in vascular smooth musclecells and mononuclear cells. Atherosclerosis1999;147(2):253-61.

141. Mayor S, Maxfield FR. Insolubility and redistribu-

tion of GPI-anchored proteins at the cell surfaceafter detergent treatment. Mol Biol Cell1995;6(7):929-44.

142. Cordle A, Landreth G. 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors attenuate beta-amy-loid-induced microglial inflammatory responses. JNeurosci 2005;25(2):299-307.




10Amyloid β-Peptide and Central

Cholinergic Neurons: Involvementin Normal Brain Function andAlzheimer’s Disease PathologySatyabrata Kar, Z. Wei, David MacTavish, Doreen Kabogo, Mee-Sook Song,and Jack H. Jhamandas

159

10.1 Introduction

Alzheimer’s disease (AD), the most common formof dementia affecting individuals over 65 years of age, is a progressive neurodegenerative disorder. Itis characterized by a global deterioration of intel-lectual function that includes an amnesic type of memory impairment, deterioration of language, andvisuospatial deficits. Motor and sensory abnormali-ties are uncommon until the late phases of thedisease, and basic activities of daily living are grad-ually impaired as the disease enters advancedphases. Psychosis and agitation also develop duringmiddle or later phases of the disease. The averagecourse of AD from the onset of clinical symptomsto death is approximately a decade, but the rate of

progression is variable [1, 2]. Epidemiological datahave shown that AD afflicts about 8–10% of thepopulation over 65 years of age, and its prevalencedoubles every 5 years thereafter [3].

Although our understanding of the pathophysiol-ogy of AD still remains fragmentary, it is widelyaccepted that both genetic and environmental fac-tors can contribute to the development of the dis-ease. In the majority of cases, AD appears to occuras sporadic disease after the age of 65 years, but ina small proportion of cases the disease is inheritedas an autosomal dominant trait and appears as anearly-onset form prior to 65 years of age. To date,

mutations within three genes—the amyloid precur-sor protein (APP) gene on chromosome 21, the pre-senilin 1 (PS1) gene on chromosome 14, and thepresenilin 2 (PS2) gene on chromosome 1—havebeen identified as the cause of early-onset familialAD [4–6]. Although these findings are of impor-tance in elucidating the biological pathogenesis of AD, it is vital to recognize that mutations in thesethree genes may only account for 30–50% of allautosomal dominant early-onset cases. The inheri-tance of late-onset AD is more complex than that of the early-onset form. Various factors, includingconcomitant pathology and limited sample sizes,make it difficult to identify genetic causes oflate-onset disease by conventional linkage analysis.However, association studies have identified candi-

date genes that significantly increase the risk forlate-onset disease. The ε4 allele of the apolipopro-tein E (APOE) gene, on chromosome 19, is onesuch risk factor. Possessing a single copy of theallele may increase the chance of developing ADtwo- to fivefold, whereas having two ε4 allelesraises this probability to more than fivefold [5–8].Despite these advances in understanding the genet-ics of AD, the vast majority of cases has not yetbeen associated with any of the four genes impli-cated to date, thus suggesting that additionalcausative mutations and genetic risk factors remainto be identified [4–6, 9]. Other factors that may



play an important role in the pathogenesis of ADinclude age, head injury, and oxidative stress [10].

10.2 Neuropathological Features

of ADThe neuropathological changes of AD are character-ized by the presence of intracellular neurofibrillarytangles, extracellular parenchymal and cerebrovas-cular amyloid deposits, and loss of neurons andsynaptic integrity in specific brain areas. These fea-tures are also seen in Down syndrome (DS) brains(<40 years of age) and, to a limited extent, in thenormal aging brain [9–11].

10.2.1 Neurofibrillary Tangles andNeuritic Plaques

Neurofibrillary tangles in the AD brain areparticularly abundant in the entorhinal cortex, hip-pocampus, amygdala, association cortices of thefrontal, temporal, and parietal lobes, and certainsubcortical nuclei. This abnormal pathology, whichis evident in neuronal cell bodies, neuropil threads,and dystrophic neuritis, is composed of hyperphos-phorylated form of microtubule-associated proteintau. Accumulation of phospho-tau reduces the abil-ity of tau to stabilize microtubules, leading to dis-ruption of neuronal transport and eventually to thedeath of affected neurons [12–15]. The extent of neurofibrillary pathology, and particularly thenumber of cortical neurofibrillary tangles, corre-lates positively with the severity of dementia.However, tangles are also found in a variety of other neurodegenerative diseases without any evi-dence of amyloid deposits [9, 12, 13, 16]. Neuritic

plaques, on the other hand, are multicellular lesionscontaining a compact deposit of amyloid peptidesin a milieu of reactive astrocytes, activatedmicroglia, and dystrophic neurites. The major amy-loid peptides that are found in the plaques areβ-amyloid 1-42 (Aβ1-42 ) and A β1-40 , peptides that aregenerated by proteolytic cleavage of APP. The timerequired to develop a neuritic plaque is not known,but these lesions are believed to evolve graduallyover a period of time from “diffuse plaques” con-taining only A β

1-42[9, 17–19]. The diffuse plaques

are found in large numbers in areas that are not typ-ically affected in AD pathology (e.g., cerebellum,

striatum, and thalamus), whereas neuritic plaquesare usually seen in areas affected by neurodegener-ation such as entorhinal cortex, hippocampus, andassociation cortices [9, 17]. Neuritic plaque num-ber does not itself correlate with the severity of dementia, although a clinical correlation betweenelevated levels of the total A β peptide in the brainand cognitive decline has been reported [20].Recent investigations in animal models and humanbrain samples have placed a special emphasis onmeasurement of soluble A β species [9, 21, 22].

Diverse lines of evidence suggest that accumula-tion of A β peptide in the brain may, over time, ini-tiate and/or contribute to AD pathogenesis. Theseinclude the association of some AD cases withinherited APP mutations [4, 9, 11]; the elevation of

Aβ peptides and the appearance of amyloid plaquesin advance of other pathology in AD and DS brains[23]; the inheritance of APOE e4 allele(s) leads toenhanced A β deposition in the brain [5, 6, 9]; theincreased production of A β1-42 in vivo and in vitroby pathogenic mutations in PS1 and PS2 [9]; andthe in vitro neurotoxic potential of fibrillar A β pep-tides [9, 24, 25]. Recent studies of APP transgenicmice [26–29] and of intrathecally administered A βin nontransgenic adult animals [30–33] reinforcethe notion that overexpression of A β peptide, orinjection of aggregated A β, induces subcellularalterations or neuronal loss in selected brainregions. It has been suggested that overexpressionor injection of A β peptide may potentiate the for-mation of neurofibrillary tangles in tau transgenicmice [34, 35], a relationship first inferred from con-sideration of familial AD kindreds. Although theseresults implicate a role for A β peptides in the neu-rodegenerative process, both the role of A β in thenormal brain and the mechanisms by which it

causes neuronal loss and tau abnormalities in ADremain poorly understood.

10.2.2 Loss of Basal ForebrainCholinergic Neurons

Selective synapse loss along with neuronal dys-function and death are part of the elemental lesionsassociated with AD pathology. Evidence suggeststhat degenerating neurons and synapses are pre-dominantly located in neuroanatomic regions thateither project to or from the brain areas displayinghighest density of plaques and tangles. Regions

160 S. Kar et al.



that are severely affected in AD brains include thehippocampus, entorhinal cortex, amygdala, neo-cortex, some subcortical areas such as basal fore-brain cholinergic neurons, serotonergic neurons of the dorsal raphe, and noradrenergic neurons of thelocus coeruleus [36–38]. Biochemical investiga-tions of biopsy and autopsy tissues indicate thatvarious neurotransmitters/modulators, includingacetylcholine (ACh), serotonin, glutamate, nora-drenaline, and somatostatin, are differentiallyaltered in AD brains [11, 36, 39]. One of the mostconsistently reproduced finding is a profoundreduction in the activity of the ACh synthesizingenzyme choline acetyltransferase (ChAT) in theneocortex that correlates positively with the sever-ity of dementia [36, 38, 40]. Reduced choline

uptake, ACh release, and loss of cholinergic neu-rons from the basal forebrain region further indi-cate a selective presynaptic cholinergic deficit inthe hippocampus and neocortex of AD brains [39,41]. Some of the earlier studies have also reportedthat depletion of cholinergic markers in the corticalregions of the AD brain may occur early in thecourse of the disease, perhaps as initiating events.In contrast, the cholinergic markers of the striatum(originating from striatal interneurons) and of thethalamus (originating from the brain stem) areeither spared or affected only in late stages of thedisease [36, 38, 39]. Together with pharmacologi-cal evidence of cholinergic involvement in theaffected cognitive processes, these findings led tothe development of a “cholinergic hypothesis” of AD. This hypothesis posits the degeneration of thecholinergic neurons in the basal forebrain and theloss of cholinergic transmission in the cerebral cor-tex and other areas as the principal cause of cogni-tive dysfunction in AD patients [38, 39, 41–43].

The hypothesis is supported, in part, by evidencethat drugs that potentiate central cholinergic func-tion (such as donepezil, rivastigmine, and galanta-mine) have some value in symptomatic treatmentduring early stages of the disease [38, 44].However, some of the recent reports, all based onelderly subjects, have challenged the assumptionthat the cholinergic depletion is an early event inAD pathology [45]. Two of these studies report thatmild AD is not associated with a loss of corticalChAT activity [46, 47], whereas the third reportsuggests that the neurons containing ChAT andvesicular ACh transporter protein may not be

decreased in early AD [48]. Collectively, thesestudies have not only raised doubts over the valid-ity of the cholinergic hypothesis as it applies toearly AD but also raise the possibility that the mod-est efficacy of cholinesterase inhibitor drugs inmild-to-moderate AD may involve mechanismsother than simple upregulation of a central cholin-ergic deficit [49, 50]. While these studies have cre-ated a number of new questions related to the roleof the cholinergic system in the prodromal stage of AD, further investigations using in vivo imagingtechniques or biochemical analysis of autopsy tis-sue using complementary approaches are needed toevaluate other components of cholinergic function(e.g., high-affinity choline transporter and nicotinicreceptors) during aging and the progression of AD.

The loss of basal forebrain cholinergic neuronshas prompted extensive study of ACh receptors inAD brains [36, 38, 39, 41, 50, 51]. ACh exertseffects on the central nervous system by interactingwith G-protein–coupled muscarinic and ligand-gated cation channel nicotinic receptors. Five dis-tinct muscarinic receptor subtypes, m 1–m 5, havebeen cloned and shown to correspond with fivepharmacologically defined M1–M5 muscarinicreceptors. It is generally believed that M2 recep-tors, most of which are located on presynapticcholinergic terminals, are reduced in AD brains[38, 51]. The density of postsynaptic M1 receptorsremains unaltered, but there is some evidence fordisruption of the coupling between the receptors,their G-proteins, and second messengers [50–52].The profiles of M3 and M4 receptors in the ADbrain remain equivocal [53, 54]. For the nicotinicreceptor family, 11 genes encoding 8 α (α 2–α 9) andthree β receptor subunits ( β2–β4) have been identi-fied [38, 55]. High-affinity central nervous system

binding sites of the agonist nicotine are mostlycomposed of α 4β2 subunits, whereas homomers of the α7 receptor subunit contribute to the high-affin-ity binding of the antagonist α -bungarotoxin ( α -BgTx) [55, 56]. Epibatidine, a potent nicotineagonist, binds with high-affinity to a subtype of nicotinic receptor containing the α 3 subunit [55].Nicotinic receptors are predominantly located oncholinergic terminals. High-affinity nicotinic bind-ing sites are markedly reduced in the hippocampusand cortex of the postmortem AD brains, and theseobservations have been confirmed in vivo bypositron emission tomography [39, 57]. There is

10. Amyloid β-Peptide and Central Cholinergic Neurons 161



also evidence of a significant decrease in α 7 proteinexpression and α -BgTx binding sites in the hip-pocampus of AD brains [58]. However, a recentimmunocytochemical study demonstrated anincrease in the proportion of astrocytes expressingα

7immunoreactivity in the hippocampus and

entorhinal cortex of the AD brain relative to theage-matched controls [59]. Notwithstanding thesedata, no muscarinic or nicotinic receptor–basedtherapeutic approaches have provided convincingevidence of an adequate level of efficacy and relia-bility in AD balanced with an acceptable burden of side effects. Whether alterations in cholinergicreceptors play a pathogenic role in dysregulatingAPP processing or promoting tau phosphorylationassociated with AD pathology remains an area of

intense investigation.

10.3 Cholinergic System andAPP Processing

10.3.1 APP Processing

Aβ peptides, the principal component of amyloiddeposits, are a group of hydrophobic peptides of 39–43 amino acid residues. These peptides arederived by proteolytic cleavage of APP—a type 1integral membrane protein with a long N-terminalextracellular region, a single membrane-spanningdomain, and a short C-terminal cytoplasmic tail [9,11, 19, 60]. Multiple isoforms are produced from asingle APP gene by alternative mRNA splicing andencode proteins ranging from 365 to 770 aminoacids. In the nervous system, APP 695 isoform isexpressed predominantly in neurons, whereasAPP 770 and APP 751 isoforms are found in neuronal

as well as non-neuronal cells [9, 18, 19]. MatureAPP is proteolytically processed by mutually exclu-sive α -secretase or β-secretase pathways. The α -sec-retase activity cleaves the A β domain within Lys 16

and Leu 17 residues, thus precluding the formation of full-length A β peptide. This pathway yields a solu-ble N-terminal APP α and a 10-kDa C-terminal APPfragment that can be further processed by γ -secre-tase to generate A β17-40 or Aβ17-42 , also known as theP3 peptides. Three members of the disintegrin met-alloproteases family that can act as potential candi-dates for α -secretase are tumor necrosis factor alphaconverting enzyme (TACE or ADAM-17), ADAM-

10, and MDC-9 [9, 18]. The β-secretase pathway,which results in the formation of intact A β peptide,is carried out by the sequential actions of two dis-tinct proteases namely, β-secretase and γ -secretase.The β-secretase cleavage is mediated by a novelaspartyl protease referred to as the β-site APP cleav-ing enzyme (BACE), which generates a truncatedsoluble APP β and a membrane-bound A β-contain-ing C-terminal fragment. Further proteolysis of theC-terminal fragment by γ -secretase yields the full-length A β1-40 or A β1-42 peptide and a recentlydescribed C-terminal fragment termed γ -CTF [9, 18,19, 61]. γ -Secretase activity resides in a multimericprotein complex that contains PS, considered as aputative aspartyl protease [62] along with four com-ponents (nicastrin, PEN-2, APH-1, and CD147) that

are required for substrate recognition, complexassembly, and targeting the complex to its site of action [63, 64].

Assimilated evidence suggests that the majorityof A β1-40/1-42 is generated in the endosomal recy-cling pathway, whereas only a minority of A β1-40/1-

42 is produced in the secretory pathway, within theendoplasmic reticulum and Golgi apparatus [9, 18,19]. Once generated, A β peptide, depending on theconcentrations, can exist in multiple forms, includ-ing monomers, dimers, higher oligomers and poly-mers; the latter includes the fibrils that accumulatein amyloid deposits [9]. At present, the mecha-nisms by which APP processing is regulated undernormal or pathological conditions remain unclear.However, several lines of experimental data haveclearly shown that the discrete APP processingpathways can be influenced by a variety of factors,including the stimulation of receptors for ACh,serotonin, glutamate, estrogen, neuropeptides, andgrowth factors [65, 66]. The influence of choliner-

gic stimulation on amyloid formation is of particu-lar interest in view of the preferential vulnerabilityof the cholinergic basal forebrain in AD and thepossibility that maintenance of this cholinergictone might slow amyloid deposition in cholinergicterminal fields.

10.3.2 Cholinergic Regulations ofAPP Processing

Over the years, a clear connection has been estab-lished between the cholinergic system and APPmetabolism. Nitsch and colleagues first demon-

162 S. Kar et al.



strated cholinergic regulation of APP processing inhuman embryonic kidney (HEK) 293 cell lines thatwere stably transfected with human muscarinic m 1,m2, m 3, and m 4 receptors [67]. Carbachol, a nonse-lective muscarinic receptor agonist, significantlyincreased the release of soluble APP α in cellsexpressing m 1 and m 3, but not in cells expressingm2 or m 4 receptor subtypes. This response was bothatropine-sensitive and blocked by staurosporine,indicating the mediation of intracellular proteinkinases in receptor-controlled APP α secretion[67]. Activation of muscarinic m 1 receptor–trans-fected cells not only enhanced soluble APP α secre-tion but also reduced the secretion of A β peptide,thus suggesting that cholinergic agents may acti-vate the non-amyloidogenic α -secretase pathway

with the potential to prevent amyloid formation.Similarly, muscarinic m 1 and m 3 receptor agonistsstimulated soluble APP α release from rat corticalslices [68] as well as brain cultured neurons [69].Both m 1 and m 3 receptors activate signaling cas-cades involving phosphatidylinositol hydrolysis/ protein kinase C (PKC) as well as mitogen acti-vated protein (MAP) kinase pathways [70].Treating cells with phorbol esters mimicked theeffect of agonist administration on soluble APP αsecretion, and this effect was blocked by PKCinhibitors [65, 71]. There is also evidence from cul-tured SH-SY5Y cells that carbachol-mediated sol-uble APP α secretion could be mediated, at least inpart, by a MAP kinase–dependent pathway [69].The mechanism whereby PKC- or MAP kinase–dependent pathways increase soluble APP α secre-tion is still unknown but may involve additionalkinase steps and the eventual activation of the pro-teases that mediate APP cleavage [65, 66, 69, 71].Moreover, a variety of other neurotransmitter/hor-

mone receptors that activate PKC- or MAPkinase–dependent signaling pathways, includingthe vasopressin, bradykinin, estrogen, serotonin,and metabotropic glutamate receptors, share thiscapacity to stimulate soluble APP secretion andinhibit A β formation [65, 69, 71, 72].

In addition to the muscarinic receptor, somestudies have examined the influence of the nico-tinic receptor on APP processing. Treatment of PC12 cells with nicotine increases the release of soluble APP α without affecting A β secretion orexpression of APP mRNA [73]. The relativeincrease in soluble APP α was attenuated by the α 7

nicotinic receptor antagonist methyllycaconitineand also by EGTA, a Ca 2+ chelator. The nicotineantagonist chlorisondamine blocked in vivo eleva-tion of total soluble APP induced by exposure to ahigh dose (8 mg kg −1day −1) of nicotine [74]. Anicotine-induced increase in Ca 2+ influx was foundto correspond with the increase in soluble APPsecretion, suggesting that Ca 2+ influx through nico-tinic receptors may be involved in enhanced secre-tion. This result is in agreement with the findingsfrom several studies showing that increased cyto-plasmic Ca 2+ levels can stimulate soluble APPsecretion [66, 71, 75].

A number of studies have investigated whetheracetylcholinesterase (AChE) inhibitors, whichimprove central cholinergic neurotransmission, can

influence APP processing with the potential tomodulate the biochemical pathways involved in theAD pathogenesis. The effects of various AChEinhibitors on soluble APP α levels differ betweencell types and depend upon the specific drug, dura-tion of treatment and the dose tested. For example,metrifonate did not alter soluble APP or A β levelsin human SK-N-SH neuroblastoma cells [76],whereas acute treatment of the inhibitor couldincrease the secretion of soluble APP α in SH-SY5Y neuroblastoma cells, presumably by increas-ing the availability of ACh and thereby stimulatingmuscarinic receptors [69, 77]. Donepezil, a rever-sible AChE inhibitor, was found to increase thesecretion of soluble APP α in a neuroblastoma cellline and platelets from AD patients by altering theactivity/trafficking of α -secretase enzyme [78, 79].Physostigmine elevated soluble APP α secretion inrat cortical slices [80] but decreased soluble APPsecretion without altering A β levels in SK-N-SHneuroblastoma cells [76]. Tacrine, a potent

cholinesterase inhibitor, was found to attenuatesecretion of soluble APP α in glial, fibroblast, andPC12 cells. The addition of tacrine to neuroblas-toma cell lines resulted in reduction of the levels of total A β, A β1-40/1-42 along with soluble APP α [81].Other AChE inhibitors such as phenserine, cymser-ine, and tolserine decreased soluble APP α levels,whereas 3,4-diaminopyridine failed to affect solu-ble APP α levels in SK-N-SH neuroblastoma cells[76]. The differential effects of the AChE inhibitorson APP processing appear to be unrelated to theirselectivity for the cholinesterase enzymes but maydepend upon other mechanisms, such as their




influence on APP synthesis, expression, turnover,trafficking, or the regulation of APP processingenzymes [69, 71, 76, 82].

10.4 Regulation of CholinergicSystem by A β Peptides

10.4.1 Effects of A β on ACH Synthesisand Release

Several studies over the past decade have clearlyshown that nM concentrations of A β peptides,under acute as well as chronic conditions, can neg-atively regulate various steps of ACh synthesis andrelease, without apparent neurotoxicity. The high

potency and reversible nature of this effect, togetherwith the fact that pM to nM concentrations of A βpeptides are found constitutively in normal braincells, suggest that A β-related peptides may act as amodulator of cholinergic function under normalconditions (Table 10.1; Fig. 10.1) [41, 71, 83–86].A 1-h exposure to pM to nM concentrations of A βcan inhibit K +- or veratridine-evoked endogenousACh release from rat hippocampal and corticalslices. This effect is tetrodotoxin-insensitive, sug-gesting that A β peptide may act at the level or inclose proximity to the cholinergic terminals [87,88]. Structure activity studies reveal that inhibitoryeffects of A β-related peptides on ACh release fromrat hippocampal slices reside within the sequenceAβ25-28 (GSNK; the C-terminal domain of the non-toxic A β1-28 fragment). In contrast with the effectson hippocampal and cortical slices, striatal AChrelease is relatively insensitive to A β peptides [87].This regional selectivity indicates that factors otherthan transmitter phenotype, such as the distance

over which cholinergic axons project to their termi-nal fields and regional variation in the expression of Aβ binding sites, may contribute to the differencesin cellular responsiveness to A β-related peptides.However, the sensitivity to A β of cholinergic neu-rons in cortex, hippocampus, and striatum matchesthe pattern of regional vulnerability in AD.

The inhibitory effects of A β on ACh release havebeen confirmed in rat and guinea-pig corticalsynaptosomes [89], rat retinal neurons [90], and incholinergic synaptosomes from the electric organ of the electric ray Narke japonica [91]. These effects

may be affected by age-related cognitive deficits.Higher levels of A β1-40 were observed in the agedrat hippocampus than were found in young adultrats, and the cholinergic neurons of aged cogni-tively impaired rats may be more sensitive to A β-mediated inhibition of hippocampal ACh releasethan either cognitively unimpaired aged or youngadult rats [92]. This is supported in part by recentdata showing that administration of antibody to A βcan increase ACh levels in the hippocampus of 12-month SAMP8 mice that exhibit age-relatedincreases in A β levels and deficits in learning andmemory [93]. Lee et al. reported that inhibition of ACh release by A β25-35 could be reversed byginkgolide B and certain ginseng saponins at con-centrations that did not by themselves alter ACh

release [94, 95]. This effect was tetrodotoxin-insen-sitive, suggesting a direct interaction of ginseng atthe level of the cholinergic synapse.

At present, the cellular mechanisms by which A β-related peptides, under acute conditions, can attenu-ate ACh release from selected brain regions remainunclear. Given the nature and potency of the effects,several steps that are critical for ACh synthesis andrelease—ranging from precursor recruitment tovesicular fusion—could be impaired by A β peptides(Table 10.1; Fig. 10.1). Turnover of ACh in thecholinergic terminals is regulated so that increasedtransmitter release is associated with increased syn-thesis. When brain slices are exposed to submaximalconcentrations of depolarizing agents such as K + orveratridine, ongoing synthesis of ACh keeps pacewith release from the terminals [96]. ACh synthesisunder these conditions depends on the high-affinityuptake of choline from extracellular sources to intra-cellular acetyl CoA and ChAT. The availability of choline is a rate-limiting determinant of ACh

biosynthesis, whereas ChAT activity is not [96].Under acute treatment conditions, pM to nM con-centrations of A β1-40/1-42 do not affect ChAT activityin tissue homogenates or in slice preparations fromhippocampus, cortex, or striatum [88]. Additionally,it is also reported that soluble A β25-35 did not affectChAT activity, under acute conditions, in the adult oraged rat brain [97]. The phosphorylation of theChAT enzyme in IMR32 neuroblastoma cellsexpressing human ChAT is known to be regulated byAβ

1-42, but its significance to ACh synthesis and/or

release remains unclear [98].

164 S. Kar et al.




T A B L E 1 0 . 1 . E f f e c t s o f

A β - r e l a t e d p e p t i d e s o n c h o l i n e r g i c n e u r o n s .

P e p t i d e f r a g m e n t

E f f e c t o n

C o n c e n t r a t i o n

M o d e l

R e f s

A C h s y n

t h e s i s a n

d r e l e a s e

A β

1 - 4 2 , 1 - 4 0 , 1 - 2 8 , 2 5 - 3 5

D e c r e a s e

i n c h o l i n e u p t a k e

p M

t o µ M

C o r t i c a l a n d

h i p p o c a m p a l s y n a p t o s o m e s

8 8 , 9 9

A β

1 - 4 2

D e c r e a s e

i n P D H a c t i v i t y

n M

P r i m a r y s e p t a l c u l t u r e s

1 0 2

A β

1 - 4 2 , 1 - 4 0 , 1 - 2 8 , 2 5 - 3 5

D e c r e a s e

i n C h A T a c t i v i t y

n M

t o µ M

S N 5 6 c e l l

l i n e a n d p r i m

a r y s e p t a l c u l t u r e s

1 0 0 , 1 2 6

A β

1 - 4 2 , 1 - 2 8 , 2 5 - 3 5 , 2 5 - 2 8

D e c r e a s e

i n A C h c o n t e n t

p M

t o n M

S N 5 6 c e l l



1 0 0 – 1 0 2

A β

1 - 4 2 , 1 - 4 0 , 1 - 2 8 , 2 5 - 3 5

D e c r e a s e

i n A C h r e l e a s e

p M

t o µ M

C o r t i c a l a n d

h i p p o c a m p a l s l i c e s , c o r t i c a l a n d e l e c t r i c

8 7 – 9 5

o r g a n s y n a p t o s o m e s , r e t i n a l

n e u r o n s

N e u r o n a

l e x c i t a

b i l i t y

A β

1 - 4 2 , 2 5 - 3 5

D e c r e a s e

i n w

h o l e - c e l l c u r r e n t s a n d

n M

t o µ M

D i s s o c i a t e d c e l l s

f r o m d i a g o n a l b a n d o f

B r o c a

8 4

i n c r e a s e

i n e x c i t a b i l i t y

A C h r e c e p

t o r s

A β

1 - 4 0 , 2 5 - 3 5

D i s r u p t M 1 - l i k e r e c e p t o r s i g n a l i n g

n M

t o µ M

P r i m a r y c o r t i c a l c u l t u r e s

1 2 0

A β

1 - 4 2

I n t e r a c t s

w i t h n i c o t i n i c r e c e p t o r

p M

t o n M

A D h i p p o c a m p u s ,

t r a n s f e c t e d c e l l s , r a t a n d

1 0 8 , 1 0 9

g u i n e a - p i g h i p p o c a m p u s

A β

1 - 4 0 , 1 - 4 2 , 1 2 - 2 8

I n h i b i t s n i c o t i n i c r e c e p t o r c u r r e n t s

n M

t o µ M

R a t

h i p p o c a m p a l s l i c e s

a n d c u l t u r e d n e u r o n s ,

1 1 0 – 1 1 4

t r a n s f e c t e d c e l l s , a n d

X e n o p u s o o c y t e s

A β

1 - 4 0 , 1 - 4 2 , 2 5 - 3 5

S t i m u l a

t e s n i c o t i n i c r e c e p t o r c u r r e n t s

p M

t o µ M

D i s s o c i a t e d c e l l s

f r o m d i a g o n a l b a n d o f

B r o c a a n d

1 1 5 , 1 1 6

X e n o p u s o o c y t e s

N e u r o n a

l v u

l n e r a

b i l i t y

A β

1 - 4 2 , 1 - 4 0 , 2 5 - 3 5

I n d u c e t a u p h o s p h o r y l a t i o n

µ M

S N 5 6 c e l l



1 2 5 , 1 2 6

A β

1 - 4 2 , 1 - 4 0 , 2 5 - 3 5

I n d u c e t o x i c i t y

µ M

S N 5 6 c e l l

l i n e , R N 4 6 A

c e l l l i n e , a n d p r i m a r y r a t

1 2 4 – 1 2 8

s e p t a l c u l t u r e s

A β

, β - a m y l o i d p e p t i d e ;

A C h , a c e

t y l c h o l i n e ; A D

, A l z h e i m e r ’ s

d i s e a s e ; C h A T , c h o l i n e a c e t y l t r a n s f e r a s e ;

P D H , p y r u v a t e

d e h y d r o g e n a s e .



In contrast with ChAT activity, high-affinity[3H]choline uptake is found to be decreased after20 minutes of preincubation with A β. This effect isparticularly marked in tissues from the hippocam-pus and cortex, mirroring the effect of A β on AChrelease in these regions [88]. Acute incubation of hippocampal synaptosomes with low nM A β1-40attenuates depolarization-induced high-affinitycholine uptake as well as [ 3H]hemicholinium-3([3H]HC-3) binding [99]. Further analysis of thesedata indicates that changes in the transport are dueto an alteration of V max , whereas the changes inspecific binding possibly involve alterations of both B max and K D. Micromolar concentrations of

Aβ1-40 decrease high-affinity choline uptake andthe [ 3H]HC-3 binding under basal conditions in atime-dependent manner [99]. These results indicatethat A β can affect acute ACh release, at least inpart, by regulating high-affinity choline uptake, butnot the activity of the ChAT enzyme. The possibleinvolvement of A β in the intracellular transport of newly synthesized ACh molecules and the fusionof ACh-containing vesicles with the presynapticmembrane remain to be investigated.

In addition to the acute effects, a 2-day exposureto pM to nM concentrations of A β1-42 , Aβ1-28 , Aβ25-

35, and to a lesser extent A β25-28 was found todecrease intracellular ACh concentrations in the

166 S. Kar et al.

FIGURE 10.1. Targets of β-amyloid (A β) peptide on central cholinergic neurons. 1, A β reduces high-affinity uptake of choline; 2, A β reduces activity of pyruvate dehydrogenase (PDH), an enzyme that generates acetyl-CoA from pyru-vate; 3, chronic exposure to A β reduces activity of the enzyme choline acetyltransferase (ChAT); 4, A β reducesacetylcholine (ACh) content; 5, A β reduces ACh release from presynaptic terminals; 6, A β interacts directly withnicotinic receptor; 7, A β impairs muscarinic M1-like signaling. AChE, acetylcholine-sterase; Ch U, site of cholineuptake; M2, presynaptic muscarinic M2 receptor; N, presynaptic nicotinic receptor. Modified from Kar et al. [94].

ACETYLCHOLINE

Membrane PtdCho

CHOLINE

Acetyl CoA

ChATChAT

ACETYLCHOLINEACETYLCHOLINE

CHOLINE

POSTSYNAPTIC NEURON

M 2

Ch U

Aβ peptides

Aβ targets on cholinergic neurons

PRESYNAPTIC NEURON

CHOLINE

N

M1

BRAINCAPILLARY

BRAINCAPILLARY

33

11

22

66

7

44

55 AChE



cholinergic hybrid SN56 cell line without causingtoxicity (Table 10.1; Fig. 10.1). The decrease in AChcould be attributed to reduced biosynthesis, as it wasaccompanied by a reduction in ChAT activity.Interestingly, the observed decrease could be pre-vented by a cotreatment with trans -retinoic acid, acompound that increases ChAT mRNA expressionin SN56 cells, or by coadministration of tyrosinekinase inhibitors [41, 100, 101]. However, inhibitionof DNA synthesis or treatment with antioxidants didnot alter ACh concentrations, thus suggesting thatneither gene transcription nor free-radical produc-tion is involved in mediating the long-term effect of Aβ on the cholinergic SN56 cell line [101]. In keep-ing with these results, treatment of rat primary sep-tal neurons with nM concentrations of A β1-42 was

found to decrease ACh production and reduce activ-ity of the acetyl-CoA biosynthesizing enzyme pyru-vate dehydrogenase (PDH) without affecting ChATactivity or neuronal survival. The decreased PDHactivity possibly results from A β activation of theglycogen synthase kinase-3 β (GSK-3 β), which canphosphorylate and inactivate PDH [102].Collectively these results suggest that chronic expo-sure to A β peptide may impair ACh synthesis/levelsby reducing the availability of acetyl CoA and/oractivity of the ChAT enzyme.

10.4.2 Effects of A β on Whole-CellCurrents in Cholineric Neurons

Apart from interacting with cholinergic terminalsin the hippocampal and cortical regions, A β pep-tide can also act at the level of cell body of cholin-ergic neurons within the basal forebrain toincrease neuronal excitability [84]. Application of 1 µM A β1-42/25-35 to acutely dissociated rat neurons

from the diagonal band of Broca decreased whole-cell voltage-sensitive currents in cholinergic neu-rons that were identified by single cell RT-PCR[84]. This reduction was observed for a suite of K +

currents, including the Ca 2+-activated K + currents(BK or Ic), the delayed rectifier current (I K), andtransient outward current (I A), but not for calciumor sodium currents. The responses were blockedby tyrosine kinase inhibitors, suggesting that A βinduces phosphorylation-dependent cascades toalter these currents [84]. These results indicate thatAβ peptides acutely modulating K + currents at thelevel of the cell body can increase excitability of the basal forebrain cholinergic neurons. More

recently, it has been demonstrated that the effectsof A β peptide on whole-cell currents are similar tothose evoked by human amylin, a 37-amino-acidpancreatic peptide that is deposited in the isletcells of patients with non-insulin-dependent dia-betes mellitus. A β evoked responses can beoccluded by human amylin and can be blocked byAC187—a specific amylin receptor antagonist.These data raise the intriguing possibility that theeffects of A β on basal forebrain cholinergic neu-rons may be expressed through the amylin recep-tor [103].

10.4.3 Effects of A β on CholinergicReceptors

Over the years, a variety of receptors (e.g., recep-tors for advanced glycation end products [RAGE],class A scavenger receptor [SR], the 75-kDa neu-rotrophin receptor [p75 NTR ], amylin receptor, andserpin-enzyme complex receptors) have beenshown to interact with A β in vitro [103–107].These interactions have attracted attention both forthe insights they may provide into the mechanismof A β action and also as potential targets for drugdesign. A number of recent studies suggest thatAβ

1-42can interact with the nicotinic ACh receptors

to mediate its acute as well as chronic effects. Thefirst reported observation of an interaction betweenAβ and α7 / α -BgTx nicotinic receptors showed thatthese proteins co-immunoprecipitated in samplesfrom postmortem AD hippocampus, and α 7 / α -BgTx nicotinic receptor antagonists compete forAβ1-42 binding to heterologously expressed α 7 / α -BgTx nicotinic receptors [108]. A subsequentstudy indicated that A β1-42 can bind with high affin-ity (Ki ~ 4–5 pM) to α 7 / α -BgTx nicotinic receptors

and with lower affinity (Ki ~ 20–30 nM) toα 4β2 /cytisine nicotinic (but not muscarinic) recep-tors in the rat and guinea-pig hippocampus andcerebral cortex [109]. This is supported by theobservation that nanomolar A β peptide was foundto inhibit nicotine-evoked currents via the α 7 / α -BgTx receptor and/or the non- α 7 nicotinic receptorin both rat hippocampal slices and cultured neu-rons, human SH-EP1 cells expressing α 4β2 nico-tinic receptor subunits, and in Xenopus oocytescontaining heterologously expressed rat or humanα 7 nicotinic receptor subunits [110–114]. However,there is also evidence that A β peptide can directlyactivate acutely dissociated rat basal forebrain




neurons via non- α 7 nicotinic receptors and in thecase of Xenopus oocytes expressing α 7 nicotinicreceptor subunit through the α 7 / α -BgTx receptors[115, 116]. In addition, it has been reported thatα 7 / α -BgTx receptors can facilitate internalizationof A β

1-42in transfected human SK-N-MC neurob-

lastoma cells [117] and can mediate A β-inducedtau phosphorylation in cultured SK-N-MC cellsand hippocampal synaptosomes [118]. The effectsof A β on the nicotinic receptor are consistent withreceptor involvement in A β-mediated inhibition of ACh release. In support of this notion, theinhibitory effects of A β1-40 on cortical ACh releasewere found to be restored by addition of α 7 agonist,such as nicotine and epibatidine, but not by α 4β2nicotinic receptor agonist cytosine [119]. However,

further studies are needed not only to define theprecise role of the α 7 nicotinic receptor in regulat-ing the inhibitory effects of A β peptides on AChrelease but also to establish its significance in rela-tion to AD pathology.

In addition to interacting with nicotinic AChreceptors, solubilized A β peptide has been shownto disrupt transduction of the muscarinic M1-likereceptor signal [120]. A 4-h exposure to nM- µMAβ1-40 reduced carbachol-induced GTPase activityin rat cortical cultured neurons without affectingmuscarinic receptor ligand binding parameters. Athigher concentrations, similar treatment with A βattenuated muscarinic M1 receptor signaling bydecreasing intracellular Ca 2+ and the accumulationof Ins(1)P, Ins(1,4)P 2, Ins(1,4,5)P 3, and Ins(1,3,4,5)P4 [120]. Exposure of rat cortical cultured neuronsto nM A β1-42 /Aβ25-35 inhibits carbachol-, but notglutamate-, induced increases in intracellular Ca 2+

and Ins(1,4,5)P 3 indicating that selective disruptionof the muscarinic M1-like signaling pathway is

another means by which A β can affect the functionof cholinoceptive neurons [121].

10.4.4 Effects of A β on CholinergicNeuron Survival

A number of in vitro studies have shown thatchronic exposure to A β peptides can induce toxic-ity in a variety of cell lines, as well as in primary ratand human cultured neurons. The toxicity of thepeptide is considered to be related to its ability toform insoluble aggregates [24, 25]. However, recentevidence suggests that the most detrimental forms

of A β peptides are the soluble oligomers and thatthe insoluble amorphous or fibrillar deposits repre-sent a less harmful form of the peptide [9, 122].Some neuronal phenotypes, such as GABAergicand serotonergic neurons, appear resistant to A βtoxicity, and various cell lines differ in their degreeof sensitivity [123, 124]. Differentiated SN56cholinergic cell lines are a susceptible line for toxi-city studies, and when exposed to A β1-40 , these cellsexhibit retraction of neurites, cell shrinkage, anddeath [125]. When treated with ciliary neurotrophicfactor, the RN46A cell line develops a cholinergicphenotype and is highly sensitive to A β peptides. Incontrast, stimulation of RN46A differentiation withbrain-derived neurotrophic factor yields an A β-insensitive cell population with a serotonergic trans-

mitter phenotype. 124 Prolonged exposure of ratprimary septal cultured neurons to µM A β peptidesinduces both cell death and a concomitant decreasein ChAT activity [126–128]. Collectively, theseresults suggest that cells expressing cholinergictransmitter phenotype are vulnerable to the toxiceffects of A β peptide.

The mechanisms by which A β induces choliner-gic cell death remains unclear but may involvealteration in intracellular calcium and/or the pro-duction of toxic and inflammatory mediators suchas nitric oxide, cytokines, and reactive oxygenintermediates [129–131]. Studies on a variety of cell lines and primary cultured neurons suggest thatAβ toxicity might be mediated either by interactionwith a hydroxysteroid dehydrogenase enzyme orby plasma membrane RAGE, SR, p75 NTR , amylin,or α7 nicotinic receptors [105–109, 127]. A role forthe death domain of p75 NTR in A β-induced celldeath was observed in neuroblastoma (SK-N-BE)cells expressing full-length or truncated forms of

p75 NTR , but recent evidence from primary humancultured neurons suggest that overexpression of p75 NTR can provide protection against A β-medi-ated toxicity by activating a phosphatidylinositide3-kinase–dependent but Akt-independent pathway[132, 133]. Studies of transfected neuroblastoma(SK-N-MC) cells indicate that expression of α 7nicotinic receptor may also have a critical role inthe degeneration by facilitating internalization andaccumulation of A β1-42 into neurons [117]. Giventhe marked expression of p75 NTR and of the α

7nicotinic receptor in the cholinergic basal fore-brain, their role in cholinergic cell death bears

168 S. Kar et al.



further investigation. More recently, it has beendemonstrated that the amylin receptor antagonistAC-187 can attenuate A β-induced toxicity in ratprimary septal cultured neurons by inhibiting a cas-pase-dependent pathway thus suggesting a possiblerole for this receptor in mediating the toxic effectsof A β [127].

Tau phosphorylation has long been considered tocontribute to neuronal vulnerability by destabilizingmicrotubules and impaired axonal transport [125,134–136]. Aggregated A β induces the phosphoryla-tion of tau protein in SN56 cholinergic cell lines[125]. Studies with rat septal cultured neurons haveindicated that aggregated A β increases levels of bothtotal tau as well as phosphorylated tau [126].Phosphorylated tau immunoreactivity could be

detected primarily in the distal axons of untreatedcells, whereas staining was evident in axons, soma,and dendrites of neurons exposed to A β [126].Hyperphosphorylated tau protein can lead to the neu-ronal death via disruption of the cytoskeletal network [13–15]; it is likely that the increase in tau phospho-rylation plays some role in A β-induced death of thecholinergic neurons. However, the mechanisms bywhich A β might induce the phosphorylation of thetau protein remain unclear. Reactive oxygen speciesand the lipid peroxidation product 4-hydroxynonenalmay be involved in A β-neurotoxicity and cross-link-ing of tau proteins [137]. Additionally, A β might alsoaffect tau phosphorylation by directly increasing rel-evant kinase activity or by decreasing phosphataseactivity [125, 134, 138–140]. Activation of GSK-3 β[136, 139, 141] and MAP kinase [138] induces tauprotein phosphorylation and cell death in a variety of cultured neuron paradigms, and prolonged exposureof rat septal cultured neurons to µM A β peptide hasbeen shown to induce tau phosphorylation by acti-

vating MAP kinase and GSK-3 β [126]. Variouskinases phosphorylate tau at discrete sites, and it islikely that the phosphorylation of tau protein incholinergic neurons is regulated by multiple kinases,including MAP kinase and GSK-3 β. Thus, it isimportant to explore both the biochemical potentialof additional tau kinases, such as cyclin-dependentkinase 5, PKC, and calcium-calmodulin kinase tophosphorylate tau [13–16], and the particular cellularexpression of these kinases by cholinergic neurons.

Tau phosphorylation can be regulated by cholin-ergic agonists, and control of tau hyperphosphory-lation by muscarinic receptor activation may

provide a side benefit of cholinomimetic therapeu-tics. Muscarinic agonists, carbachol and AF 102B,attenuate tau phosphorylation in cultured PC12cells stably transfected with muscarinic m 1 recep-tors [142]. On the other hand, activation of thenicotinic receptor by nicotine and epibatidineincreased the levels of phosphorylated as well asnon-phosphorylated tau in SH-SY5Y human neu-roblastoma cells [143]. The mechanisms by whichmuscarinic m 1 or nicotinic receptor activation mod-ify tau phosphorylation remain unclear, but recentdata suggest that stimulation of α 7 / α -BgTx nico-tinic receptors by A β1-42 can induce tau phosphory-lation in human neuroblastoma cells andhippocampal synaptosomes via extracellular recep-tor kinases (ERKs) and c-Jun N-terminal kinase

(JNK-1) [118]. These activities may likely involvealteration of other protein kinase/protein phos-phatase systems [71].

10.4.5 Effects of In Vivo Administrationof A β on Cholinergic Neurons

Attempts have been made to measure the impact of intracerebroventricular or local administration of A βon cholinergic system under in vivo conditions.Several studies have reported that A β peptides caninduce cholinergic hypofunction when administeredto the brain [31, 41, 83, 144, 145]. Injection of A β25-

35/1-40 into the rat medial septum causes a reductionin ACh release from the hippocampus in the absenceof toxicity [146]. Using a similar approach, Harkanyet al [31]. demonstrated that A β1-42 is toxic to cholin-ergic neurons, as indicated by reduction in ChAT-immunoreactive cell bodies in the basal forebrainand fibers in the cerebral cortex. This effect waspartly antagonized by the N -methyl- D-aspartate

(NMDA) receptor antagonist MK-801, thus suggest-ing a possible involvement of an excitotoxic path-way in mediating the effects of A β peptide [31].More recently, it has been shown that aging andhigh-cholesterol diet can enhance in vivo toxicity of Aβ peptide on cholinergic neurons [145]. Otherstudies have reported that infusion of A β into the lat-eral ventricles of adult rats impairs performance onlearning and memory tasks in a manner similar tothe effect of cholinergic inhibition [30, 32, 83, 144].Local injection of preaggregated A β

1-42into the

nucleus basalis magnocellularis (NBM) producescongophilic deposits and a strong inflammatory




response, characterized by activation of astrocytesand microglia and by induction of microglialp38MAP kinase activity [147]. These changes wereaccompanied by a decrease in the number of cholin-ergic neurons around the congophilic amyloiddeposit and hypofunction of the cortical cholinergicsystem [147]. Clearly, the influence of these astro-cytic and microglial responses must be considered inassessing in vivo effects of A β peptides on choliner-gic function.

10.4.6 Cholinergic System in TransgenicMice Overexpressing A β Peptide

Over the past few years, the central cholinergic sys-tem has been examined extensively in a variety of

mutant APP, PS1, or APP/PS1 transgenic mouselines, all of which exhibit elevated A β levels[148–163].

In mice expressing the hAPP V642I London mutanttransgene, a selective decrease was found in the sizeof medial septal cholinergic neurons, but not in NBMcholinergic neurons. At 17–22 months of age, thisline exhibits both reorganization of AChE-positivefibers in the hippocampus and dystrophic AChE-pos-itive fibers around amyloid plaques in the cortex[149]. Cerebral amyloidosis was found to cause asignificant cholinergic fiber loss and severe disrup-tion of neocortical cholinergic fiber networks in agedAPP23 mice expressing hAPP KM670/671NL Swedishmutant transgene [148]. Although the cholinergicneurons of the medial septum and vertical limb of thediagonal band of Broca were smaller in APP23 trans-genic mice than in non-transgenic controls, the num-ber and volume of ChAT-positive neurons in theNBM complex were not affected. Hippocampalcholinergic fiber density in APP23 mice has yet to be

reported [148]. Homozygous PDAPP mice express-ing the hAPP V717F mutant transgene showed an age-dependent decrease in hippocampal and corticalcholinergic fiber density without any evident loss of basal forebrain cholinergic neurons compared withthe non-transgenic controls. The degeneration of cholinergic nerve terminals in these transgenic micewas found to occur prior to the deposition of A β-con-taining neuritic plaques [159].

In another study, hAPP KM670/671NL mutant micedemonstrated an upregulation in the density of cholinergic synapses in the frontal cortex, parietal

cortex, and the hippocampus, whereas PS1 M146Ltransgenic mice showed no changes in either thesize or density of cholinergic synapses. Whencrossed to yield hAPP KM670/671NL /PS 1M146L doubletransgenic mice, extensive amyloid plaques werefound to be associated with decreased density andsize of cholinergic synapses in the frontal cortexand hippocampus [150]. A significant inverse rela-tionship was noted between the presynaptic cholin-ergic bouton density and size of A β-containingneuritic plaques located in the frontal cortex of thehAPP KM670/671NL /PS 1M146L double transgenic mice[160]. In one study, a selective increase inimmunostaining for p75 NTR (a marker of basal fore-brain cholinergic neurons) was evident in themedial septum of 12-month-old hAPP KM670/671NL or

PS1 M146L single transgenic mice but not inhAPP KM670/671NL /PS 1M146L double transgenic mice.Staining of p75 NTR -immunoreactive fibers in hip-pocampus was more robust in single transgenicmice, relative to non-transgenic controls, whiledouble transgenic mice displayed less intensep75 NTR fiber staining [151]. Whether the increasedimmunostaining in singly transgenic mice indicatesa trophic effect on the cholinergic neurons as a con-sequence of either hAPP KM670/671NL or PS1 M146Lgene overexpression remains to be investigated.However, a separate study revealed no differencesbetween hAPP KM670/671NL mice and non-transgeniccontrols in ChAT activity, AChE activity, vesicularACh transporter binding, or high-affinity cholineuptake sites in cortex, hippocampus, striatum, orcerebellum at multiple times up to 23 months of age[152]. Interestingly, a recent study showed thatextracellular hippocampal ACh levels, but not stim-ulated ACh release, were slightly but significantlyreduced (~26% decrease) in knock-in mice carrying

hAPP KM670/671NL /PS 1M146L transgenes compared withmice overexpressing hAPP KM670/671NL /PS wild-typetransgenes, thus suggesting expression of mutantAPP/PS1 genes may induce subtle alteration incholinergic transmission [164].

Densities of M1/[ 3H]pirenzepine, M2/[ 3H]AF-DX 384, or α 7 nicotinic/[ 125 I]α -BgTx receptorbinding sites in all brain regions of mutant PS1 L286Vtransgenic and wild-type PS1 transgenic mice arecomparable with those found in non-transgeniccontrols [153]. In hAPP

KM670/671NLmutant mice, a

decrease in M1/[ 3H]pirenzepine and α 4β2 nico-

170 S. Kar et al.



tinic/[ 3H]cytisine, but not M2/[ 3H]AF-DX 384,receptor binding was evident in the hippocampusand cortex compared with non-transgenic controls[157]. However, in other studies, elevated hip-pocampal α7 nicotinic receptor levels have beenreported in hAPP

K670N/M671Lsingle and two lines

(i.e., hAPP K670N/M671L /PS1 A246E and APP KM670/671NL +

V717F /PS1 M146L +L286V ) of double transgenic mice[154, 156]. In triple transgenic mice harboringhAPP KM670/671NL /PS1 M146V /Tau P301L transgenes, anage-dependent reduction of α 7 / α -BgTx nicotinicreceptor binding sites was observed in the hip-pocampus and cortical regions compared with non-transgenic mice. Additionally, chronic nicotineintake was found to exacerbate tau pathology inthese transgenic mice, suggesting an in vivo role

for the nicotinic receptor in the phosphorylation of tau protein [163]. Apart from receptor binding site,high-affinity [ 3H]HC binding (i.e., choline uptakesites) was found to be reduced in cortical regions of 5- and 17-month-old hAPP KM670/671NL mutant mice,whereas [ 3H]vesamicol binding (i.e., vesicular Achtransporter sites) was increased in 17-month-oldbut not in 5-month-old transgenic mice comparedwith littermate non-transgenic controls [162].However, the significance of the changes in thesepresynaptic cholinergic markers and their associa-tion with the amyloid pathology remains unclear.In sum, increased expression of A β peptides pro-duces a range of effects on cholinergic systems of mutant APP, PS1, or APP/PS1 transgenic mice.Establishing which of these effects are robustlyrelated to the type of pathogenic mutation, the levelof transgene expression, or to the intensity of amyloid deposits remains to be defined in futurestudies.

10.5 Significance of AmyloidInteractions with CholinergicNeurons

Earlier results have shown that A β-related pep-tides are produced constitutively by brain cells andare found in the pM to nM range in the cere-brospinal fluid of normal individuals [9, 165–167].These concentrations of A β can have a neuromod-ulatory role in the regulation of normal cholinergic

functions, possibly through their negative effectson ACh biosynthesis and release. Conversely,there is evidence that ACh can regulate APP syn-thesis and processing. For example, lesions of thebasal forebrain cholinergic neurons or transientinhibition of cortical ACh release could elevatelocal APP synthesis [65, 168–170], whereas ago-nist-induced activation of muscarinic m 1 and m 3receptor subtypes increases the secretion of solu-ble APP derivatives and reduces the production of amyloidogenic A β peptides [65–71, 171]. Theseresults suggest a reciprocal mechanism wherebynormal cholinergic innervation participates in thenonamyloidogenic maturation of APP via the α -secretase pathway, while the amyloidogenic A β-related peptides depress the activity of cholinergic

neurons. A shift in the balance between theseactivities may possibly be a key factor in the tar-geting of cholinergic neurons in AD. Insults thatreduce cholinergic transmission, increase A β gen-eration, or reduce A β clearance may enhance vul-nerability of neurons to direct toxicity of A βpeptide [9, 24, 25] or to choline limitation [83, 86,88, 99, 172, 173]. Because cholinergic neuronsutilize choline from membrane phosphatidyl-choline to synthesize ACh, it is likely that A β-induced alteration in intracellular choline levelsmight lead to an autocannibalistic process inwhich membrane turnover is disrupted to sustainneurotransmission [173]. Given the evidence thatAβ deposits precede any other lesions in ADbrains [23], it is possible that amyloid-induced tauphosphorylation may also play a critical role inneuronal loss. This is supported by some in vivostudies in which intrathecal administration, ortransgene-delivered expression of A β peptides wasshown to induce a loss of neurons, or a change in

presynaptic cholinergic markers, within selectedbrain regions [30–33, 148–150, 159]. The selec-tive interactions of A β with basal forebrain cholin-ergic neurons provide candidate mechanisms thatmay contribute, at least in part, to the vulnerabilityof these neurons and their projections in AD. Itremains to be determined whether changes incholinergic transmission alter APP processingpathways so as to further AD pathology. If so,appropriate cholinomimetic therapeutics might beexpected both to provide symptomatic benefit andto abrogate AD pathogenesis.




Acknowledgments The authors gratefully acknowl-edge support of the Canadian Institutes of HealthResearch and the many contributions of Drs. D.Westaway, H.T. Mount, and R. Quirion to thisresearch program. J.H.J. is a recipient of CanadaResearch Chair (CRC) in Alzheimer’s Research, andS.K. is a recipient of CRC in NeurodegenerativeDiseases and a Senior Scholar award from theAlberta Heritage Foundation for Medical Research.

References

1. Whitehouse PJ. Genesis of Alzheimer’s disease.Neurology 1997;48(Suppl 7):S2-7.

2. Katzman R. The prevalence and malignancy of Alzheimer’s disease. Arch Neurol 1976;33:217-8.

3. Cummings JL. Alzheimer’s disease. N Engl J Med2004;351:56-67.4. Holmes C. Genotype and phenotype in Alzheimer’s

disease. Br J Psychiatry 2002;180:131-4.5. Bertram L, Tanzi RE. The current status of

Alzheimer’s disease genetics: what do we tellpatients? Pharmacol Res 2004;50:385-396.

6. St George-Hyslop PH, Petit A. Molecular biologyand genetics of Alzheimer’s disease. C R Biologies2004;328:119-130.

7. Strittmatter WJ, Saunders AM, Schmeckel D, et al.Apolipoprotein E: High-avidity binding to β-amyloidand increased frequency of type 4 allele in late-onsetfamilial Alzheimer’s disease. Proc Natl Acad SciUSA 1993;90:1977-81.

8. Poirier J, Davingnon J, Bouthillier D, et al.Apolipoprotein E polymorphism and Alzheimer’sdisease. Lancet 1993;342:697-9.

9. Selkoe DJ. Alzheimer’s disease: genes, proteins andtherapy. Physiol Rev 2001;81:741-66.

10. Muller-Spahn F, Hock C. Risk factors and differentialdiagnosis of Alzheimer’s disease. Eur Arch PsychiatryClin Neurosci 1999;249(Suppl 3):III/37-III/42.

11. Price DL, Sisodia SS. Mutant genes in familialAlzheimer’s disease and transgenic models. AnnuRev Neurosci 1998;21:479-505.

12. Lee VM. Disruption of the cytoskeleton inAlzheimer’s disease. Curr Opin Neurobiol1995;5:663-8.

13. Iqbal K, Alonso Adel C, Chen S, et al. Tau pathologyin Alzheimer’s disease and other tauopathies.Biochim Biophys Acta 2005;1739(2-3):198-210.

14. Brion JP, Anderton BH, Authelet M, et al.Neurofibrillary tangles and tau phosphorylation.Biochem Soc Symp 2001;67:81-8.

15. Billingsley ML, Kincaid RL. Regulated phosphory-lation and dephosphorylation of tau protein: effects

on microtubule interaction, intracellular traffickingand neurodegeneration. Biochem J 1997;323:577-91.

16. Bierer LM, Hof PR, Purohit DP, et al. Neocorticalneurofibrillary tangles correlate with dementia sever-ity in Alzheimer’s disease. Arch Neurol 1995;52:81-8.

17. Dickson DW. The pathogenesis of senile plaques.

J Neuropathol Exp Neurol 1997;56:321-39.18. Clippingdale AB, Wade JD, Barrow CJ. The amy-

loid- β peptide and its role in Alzheimer’s disease.J Peptide Sci 2001;7:227-49.

19. Wisniewski T, Ghiso J, Frangione B. Biology of A βamyloid in Alzheimer’s disease. Neurobiol Dis1997;4:313-28.

20. Naslund J, Haroutunian V, Mohs R, et al. Correlationbetween elevated levels of amyloid beta-peptide in thebrain and cognitive decline. JAMA 2000;283: 1571-7.

21. McLean CA, Cherny RA, Fraser FW, et al. Solublepool of Abeta amyloid as a determinant of severity of neurodegeneration in Alzheimer’s disease. AnnNeurol 1999;46:860-6.

22. Lue LF, Kuo YM, Roher AE, et al. Soluble amyloidbeta peptide concentration as a predictor of synapticchange in Alzheimer’s disease. Am J Pathol 1999;155:853-62.

23. Tanzi RE. Neuropathology in the Down’s syndromebrain. Nat Med 1996;2:31-2.

24. Pike CJ, Burdick D, Walencewicz AJ, et al.Neurodegeneration induced by β-amyloid peptidesin vitro: the role of peptide assembly state. J Neurosci

1993;13:1676-87.25. Yankner BA. Mechanisms of neuronal degeneration

in Alzheimer’s disease. Neuron 1996;16:921-32.26. Games D, Adams D, Alessandrini R, et al.

Alzheimer’s type neuropathology in transgenic miceoverexpressing V717F β-amyloid precursor protein.Nature 1995;373:523-7.

27. Hsiao K, Chapman P, Nilsen S, et al. Correlativememory deficits, A β elevation, and amyloid plaquesin transgenic mice. Science 1996;274:99-102.

28. Calhoun M, Wiederhold K, Abramowski D, et al.

Neuron loss in APP transgenic mice. Nature1998;395:755-6.29. Bondolfi L, Calhoun M, Ermini F, et al. Amyloid-

associated neuron loss and gliogenesis in the neocor-tex of amyloid precursor protein transgenic mice. JNeurosci 2002;22:515-22.

30. Giovannelli L, Casamenti F, Scali C, et al.Differential effects of amyloid peptides β-(1-40) andβ-(25-35) injections into rat nucleus basalis.Neuroscience 1995;66:781-92.

31. Harkany T, Abraham I, Timmerman W, et al. β-amy-loid neurotoxicity is mediated by a glutamate-trig-gered excitotoxic cascade in rat nucleus basalis. EurJ Neurosci 2000;12:2735-45.

172 S. Kar et al.



32. Itoh A, Nitta A, Nadai M, et al. Dysfunction of cholinergic and dopaminergic neuronal systems in β-amyloid protein-infused rats. J Neurochem 1996;66:1113-7.

33. Geula C, Wu CK, Saroff D, et al. Aging renders thebrain vulnerable to amyloid β-protein neurotoxicity.

Nat Med 1998;4:827-31.34. Gotz J, Chen F, van Dorpe J, et al. Formation of neu-

rofibrillary tangles in P301l tau transgenic miceinduced by Abeta 42 fibrils. Science 2001;293:1491-5.

35. Lewis J, Dickson DW, Lin WL, et al. Enhanced neu-rofibrillary degeneration in transgenic mice express-ing mutant tau and APP. Science 2001;293:1487-91.

36. Geula C, Mesulam MM. Cholinergic system andrelated neuropathological predilection patterns inAlzheimer’s disease. In: Terry RD, Katzman R, Bick KL, editors. Alzheimer’s Disease. New York: RavenPress Ltd.; 1994:263-91.

37. DeKosky ST, Scheff SW, Styren SD. Structural cor-relates of cognition in dementia: quantification andassessment of synapse change. Neurodegeneration1996;5:417-21.

38. Lander CJ, Lee JM. Pharmacological drug treatmentof Alzheimer’s disease: the cholinergic hypothesisrevisited. J Neuropathol Exp Neurol 1998;57:719-31.

39. Francis PT, Palmer AM, Snape M, et al. The cholin-ergic hypothesis of Alzheimer’s disease: a review of progress. J Neurol Neurosurg Psychiatry 1999;66:137-47.

40. Davies P, Maloney AJF. Selective loss of centralcholinergic neurons in Alzheimer’s disease. Lancet1976;2:1403.

41. Blusztajn JK, Berse B. The cholinergic neuronal phe-notype in Alzheimer’s disease. Metab Brain Dis2000;15:45-64.

42. Perry EK, Tomlinson BE, Blessed G, et al.Correlation of cholinergic abnormalities with senileplaques and mental test scores in senile dementia. BrMed J 1978;2:1457-9.

43. Bartus RT, Dean RLIII, Beer B, et al. The choliner-

gic hypothesis of geriatric memory dysfunction.Science 1982;217:408-17.44. Trinh NH, Hoblyn J, Mohanty S, et al. Efficacy of

cholinesterase inhibitors in the treatment of neu-ropsychiatric symptoms and functional impairmentin Alzheimer’s disease: a meta-analysis. JAMA2003;289:210-6.

45. Morris JC. Challenging assumptions aboutAlzheimer’s disease; mild cognitive impairment andthe cholinergic hypothesis. Ann Neurol 2002;51:143-4.

46. Davis KL, Mohs RC, Marin D, et al. Cholinergicmarkers in elderly patients with early signs of Alzheimer’s disease. JAMA 1999;281:1401-6.

47. DeKosky ST, Ikonomovic MD, Styren SD, et al.Upregulation of choline acetyltransferase activity inhippocampus and frontal cortex of elderly subjectswith mild cognitive impairment. Ann Neurol2002;51:145-55.

48. Gilmor ML, Erickson JD, Varoqui H, et al.

Preservation of nucleus basalis neurons containingcholine acetyltransferase and the vesicular acetyl-choline transporter in the elderly with mild cognitiveimpairment and early Alzheimer’s disease. J CompNeurol 1999;411:693-704.

49. Terry AV, Buccafusco JJ. The cholinergic hypothesis of age and Alzheimer’s disease-related cognitive deficits:recent challenges and their implications for novel drugdevelopment. J Pharmacol Exp Ther 2003;306:821-7.

50. Mesulam M. The cholinergic lesions of Alzheimer’sdisease:pivotal factor or slide show? Learn Mem2004;11:43-9.

51. Nordberg A, Alafuzoff I, Winbald B. Nicotinic andmuscarinic receptor subtypes in the human brain:changes with aging and dementia. J Neurosci Res1992;31:103-11.

52. Warpman U, Alafuzoff I, Nordberg A. Coupling of muscarinic receptors to GTP proteins in postmortemhuman brain – alterations in Alzheimer’s disease.Neurosci Lett 1993;150:39-43.

53. Rodriguez-Puertas R, Pascual J, Vilaro T, et al.Autoradiographic distribution of M1, M2, M3, andM4 muscarinic receptor subtypes in Alzheimer’s dis-

ease. Synapse 1997;26:341-50.54. Mulugeta E, Karlsson E, Islam A, et al. Loss of mus-

carinic M4 receptors in hippocampus of Alzheimer’spatients. Brain Res 2003;960:259-62.

55. Colquhoun LM, Patrick JW. Pharmacology of neu-ronal nicotinic acetylcholine receptor subtypes. AdvPharmacol 1997;39:191-20.

56. Drisdel RC, Green WN. Neuronal α -bungarotoxinreceptors are α 7 subunit homomers. J Neurosci 2000;20:133-9.

57. Nordberg A, Lundqvist H, Hartvig P, et al. Kinetic

analysis of regional (S)(-)11

C-nicotine binding innormal and Alzheimer’s brains—in vivo assessmentusing positron emission tomography. Alzheimer’sDis Assoc Disord 1995;9:21-7.

58. Court J, Martin-Ruiz C, Piggott M, et al. Nicotinicreceptor abnormalities in Alzheimer’s disease. BiolPsychiatry 2001;49:175-84.

59. Teaktong T, Graham A, Court J, et al. Alzheimer’sdisease is associated with a selective increase in α 7nicotinic acetylcholine receptor immunoreactivity inastrocytes. Glia 2003;41:207-11.

60. Kang J, Lemaire GH, Unterbeck A, et al. The precur-sor of Alzheimer’s disease amyloid A4 protein resem-bles a cell surface receptor. Nature 1987;325:733-6.




61. Vassar R, Bennett BD, Babu-Khan S, et al. β-secretase cleavage of Alzheimer’s amyloid precursorprotein by the transmembrane aspartic proteaseBACE. Science 1999;286:735-41.

62. Kimberly WT, LaVoie MJ, Ostaszewski BL, et al.Gamma-secretase is a membrane protein complex

comprised of presenilin, nicastrin, Aph-1, and Pen-2.Proc Natl Acad Sci USA 2003;100:6382-7.

63. Haass C, Steiner H. Alzheimer’s disease γ -secretase:a complex story of GxGD-type presenilin proteases.Trends Cell Biol 2002;12:556-62.

64. Zhou S, Zhou H, Walian PJ, et al. CD147 is a regu-latory subunit of the gamma-secretase complex inAlzheimer’s disease amyloid beta-peptide produc-tion. Proc Natl Acad Sci USA 2005;102:7499-504.

65. Roberson MR, Harrell LE. Cholinergic and amyloidprecursor protein metabolism. Brain Res Rev1997;25:50-69.

66. Mills J, Reiner PB. Regulation of amyloid precursorprotein cleavage. J Neurochem 1999;72:443-60.

67. Nitsch RM, Slack BE, Wurtman RJ, et al. Release of Alzheimer’s amyloid precursor derivatives stimu-lated by activation of muscarinic cholinergic recep-tor. Science 1992;258:304-7.

68. Pittel Z, Heldman E, Barg J, et al. Muscarinic controlof amyloid precursor protein secretion in rat cerebralcortex and cerebellum. Brain Res 1996;742:299-304.

69. Racchi M, Mazzucchelli M, Porrello E, et al.Acetylcholinesterase inhibitors:novel activities of

old molecules. Pharmacol Res 2004;50:441-51.70. Guo FF, Kumahara E, Saffen D. A CalDAG-

GEFI/Rap1/B-Raf cassette couples M(1) muscarinicacetylcholine receptors to the activation of ERK1/2.J Biol Chem 2001;276:25568-81.

71. Hellstrom-Lindahl E. Modulation of β-amyloid pre-cursor protein processing and tau phosphorylation byacetylcholine receptors. Eur J Pharmacol 2000;393:255-63.

72. Manthey D, Heck S, Engert S, et al. Estrogen inducesa rapid secretion of amyloid beta precursor protein

via the mitogen-activated protein kinase pathway.Eur J Biochem 2001;268:4285-91.73. Kim SH, Kim YK, Jeong SJ, et al. Enhanced release

of secreted form of Alzheimer’s amyloid precursorprotein from PC12 cells by nicotine. Mol Pharmacol1997;52:430-6.

74. Lahiri DK, Utsuki T, Chen D, et al. Nicotine reducesthe secretion of Alzheimer’s β-amyloid precursorprotein containing β-amyloid peptide in the rat with-out altering synaptic proteins. Ann N Y Acad Sci2002;965:364-372.

75. Buxbaum JD, Ruefli AA, Parker CA, et al.Calcium regulates processing of the Alzheimer’s

amyloid protein precursor in a protein kinase C-independent manner. Proc Natl Acad Sci USA1994;91:4489-93.

76. Lahiri DK, Farlow MR, Hintz N, et al.Cholinesterase inhibitors, β-amyloid precursor pro-tein and amyloid β-peptides in Alzheimer’s disease.

Acta Neurol Scand Suppl 2000;176:60-7.77. Racchi M, Govoni S. The pharmacology of amyloid

precursor protein processing. Exp Gerontology2003;38:145-57.

78. Zimmermann M, Gardoni F, Marcello E, et al.,Acetylcholinesterase inhibitors increase ADAM10activity by promoting its trafficking in neuroblas-toma cell lines. J Neurochem 2004;90:1489-99.

79. Zimmermann M, Borroni B, Cattabeni F, et al.Cholinesterase inhibitors influence APP metabolismin Alzheimer’s disease patients. Neurobiol Dis2005;19:237-42.

80. Mori F, Lai CC, Fusi F, et al. Cholinesteraseinhibitors increase secretion of APPs in rat brain cor-tex. NeuroReport 1995;6:633-6.

81. Lahiri DK, Farlow MR, Sambamurti K. The secre-tion of amyloid beta-peptides is inhibited in thetacrine-treated human neuroblastoma cells. MolBrain Res 1998;62:131-40.

82. Shaw KT, Utsuki T, Rogers J, et al. Phenserine regu-lates translation of beta-amyloid precursor proteinmRNA by a putative interleukin-1 responsive ele-ment, a target for drug development. Proc Natl Acad

Sci USA 2001;98:7605-10.83. Auld DS, Kornecook TJ, Bastianetto S, et al.

Alzheimer’s disease and the basal forebrain choliner-gic system: relations to beta-amyloid peptides, cog-nition, and treatment strategies. Prog Neurobiol2002;68:209-45

84. Jhamandas JH, Cho C, Jassar B, et al. Cellularmechanisms for amyloid β-protein activation of ratbasal forebrain neurons. J Neurophysiol 2001;86:1312-20.

85. Dolezal V, Kasparova J. β-amyloid and cholinergic

neurons. Neurochem Res 2003;28:499-506.86. Kar S, Slowikowski SP, Westaway, Mount HT.Interactions between b-amyloid and central choliner-gic neurons: implications for Alzheimer’s disease.J Psychiatry Neurosci 2004;29:427-41.

87. Kar S, Seto D, Gaudreau P, et al. β-amyloid-relatedpeptides inhibit potassium-evoked acetylcholinerelease from rat hippocampal formation. J Neurosci1996;16:1034-40.

88. Kar S, Issa AM, Seto D, et al. Amyloid β-peptideinhibits high-affinity choline uptake and acetylcholinerelease in rat hippocampal slices. J Neurochem1998;70:2179-87.

174 S. Kar et al.



89. Wang HY, Wild KD, Shank RP, et al. Galanininhibits acetylcholine release from rat cerebral cor-tex via a pertussis toxin-sensitive G i protein.Neuropeptides 1999;33:197-205.

90. Melo JB, Agostinho P, Oliveira CR. Amyloid beta-peptide 25-35 reduces [ 3H]acetylcholine release in

retinal neurons. Involvement of metabolic dysfunc-tion. Amyloid 2002;9:221-8.

91. Satoh Y, Hirakura Y, Shibayama S, et al. Beta-amyloid peptides inhibit acetylcholine release fromcholinergic nerve endings isolated from an electricray. Neurosci Lett 2001;302:97-100.

92. Vaucher E, Amount N, Rowe W, et al. Amyloid βpeptide levels and its effects on hippocampal acetyl-choline release in aged, cognitively-impaired and -unimpaired rats. J Chem Neuroanat 2001;21:323-9.

93. Farr SA, Banks WA, Uezy K, et al. Antibody to β-amyloid protein increases acetylcholine in the hip-pocampus of 12 month SAMP8 male mice. Life Sci2003;73:555-62.

94. Lee TF, Shiao YJ, Chen CF, et al. Effect of ginsengsaponins on beta-amyloid-suppressed acetylcholinerelease from rat hippocampal slices. Planta Med2001;67:634-7.

95. Lee T, Chen C, Wang AC. Effect of Ginkolides onβ-amyloid-suppressed acetylcholine release fromrat hippocampal slices. Phytother Res 2004;18:556-60.

96. Wecker L. The synthesis and release of acetyl-

choline by depolarized hippocampal slices isincreased by increased choline available in vitroprior to stimulation. J Neurochem 1991;57:1119-27.

97. Zambrzycka A, Alberghina M, Strosznajder JB.Effects of aging and amyloid-beta peptides oncholine acetyltransferase activity in rat brain.Neurochem Res 2002;27:277-81.

98. Dobransky T, Brewer D, Lajoie G, et al.Phosphorylation of 69-kDa choline acetyltrans-ferase at threonine 456 in response to amyloid-betapeptide 1-42. J Biol Chem 2003;278:5883-93.

99. Kristofikova Z, Tekalova H, Klaschka J. Amyloidbeta peptide1-40 and the function of rat hippocam-pal hemicholinium-3 sensitive choline carriers:effects of a proteolytic degradation in vitro .Neurochem Res 2001;26:203-12.

100. Pedersen WA, Kloczewiak MA, Blusztajn JK.Amyloid β-protein reduces acetylcholine synthesisin a cell line derived from cholinergic neurones of the basal forebrain. Proc Natl Acad Sci USA1996;93:8068-71.

101. Pedersen WA, Blusztajn JK. Characterization of theacetylcholine reducing effect of the amyloid-beta peptidein mouse SN56 cells. Neurosci Lett 1997;239:77-80.

102. Hoshi M, Takashima A, Murayama M, et al.Nontoxic amyloid β peptide 1-42 supresses acetyl-choline synthesis. J Biol Chem 1997;272:2038-41.

103. Jhamandas JH, Harris KH, Cho C, et al. Humanamylin actions on rat cholinergic basal forebrainneurons: antagonism of beta-amyloid effects.

J Neurophysiol 2003;89:2923-30.104. Joslin G, Krause JE, Hershey AD, et al. Amyloid- β

peptide, substance, and bombesin bind to the ser-pin-enzyme complex receptor. J Biol Chem 1991;266:21897-902.

105. El Khoury J, Hickman SE, Thomas CA, et al.Scavenger receptor-mediated adhesion of microgliato β-amyloid fibrils. Nature 1996;382:716-9.

106. Yan SD, Chen X, Fu J, et al. RAGE and amyloid- βpeptide neurotoxicity in Alzheimer’s disease.Nature 1996;382:685-91.

107. Kuner P, Schubenel R, Hertel C. β-amyloid binds top75 NTR and activates NF κ B in human neuroblas-toma cells. J Neurosci Res 1998;54:798-804.

108. Wang HY, Lee DHS, D’Andrea MR, et al. β-amyloid1-42 binds α 7 nicotinic acetylcholinereceptor with high affinity: implications forAlzheimer’s disease pathology. J Biol Chem 2000;275: 5626-32.

109. Wang HY, Lee DHS, Davis CB, et al. Amyloid pep-tide A β1-42 binds selectively and with picomolaraffinity to α 7 nicotinic acetylcholine receptors.J Neurochem 2000;75:1155-61.

110. Liu Q, Kawai H, Berg DK. β-amyloid peptideblocks the response of α 7-containing nicotinicreceptors on hippocampal neurons. Proc Natl AcadSci USA 2001;98:4734-9.

111. Pettit DL, Shao Z, Yakel JL. β-amyloid 1-42 peptidedirectly modulates nicotinic receptors in the rat hip-pocampal slice. J Neurosci 2001;21:RC120-24.

112. Tozaki H, Matsumoto A, Kanno T, et al. Theinhibitory and facilitatory actions of amyloid-betapeptides on nicotinic ACh receptors and AMPAreceptors. Biochem Biophys Res Commun 2002;

294:42-5.113. Grassi F, Palma E, Tonini R, et al. Amyloid beta(1-42) peptide alters the gating of human and mousealpha-bungarotoxin-sensitive nicotinic receptors.J Physiol 2003;547:147-57.

114. Wu J, Kuo YP, George AA, et al. β-amyloid directlyinhibits α4β2-nicotinic acetylcholine receptorsheterologously expressed in human SH-EP1 cells.J Biol Chem 2004;279:37842-51.

115. Dineley KT, Bell K, Bui D, et al. beta-Amyloid pep-tide activates alpha 7 nicotinic acetylcholine recep-tors expressed in Xenopus oocytes. J Biol Chem2002;277:25056-61.




116. Fu W, Jhamandas JH. Beta-amyloid peptide acti-vates non-alpha7 nicotinic acetylcholine receptorsin rat basal forebrain neurons. J Neurophysiol2003;90:3130-6.

117. Nagele RG, D’Andrea MR, Anderson WJ, et al.Intracellular accumulation of beta-amyloid(1-42)

in neurons is facilitated by the alpha 7 nicotinicacetylcholine receptor in Alzheimer’s disease.Neuroscience 2002;110:199-211

118. Wang HY, Li W, Benedetti NJ, et al. α 7 nicotinicacetylcholine receptors mediate β-amyloid peptide-induced tau protein phosphorylation. J Biol Chem2003;278:31547-53.

119. Lee DHS, Wang HY. Differential physiologicresponses of α 7 nicotinic acetylcholine receptors toβ-amyloid 1-40 and β-amyloid 1-42 . J Neurobiol2003;55:25-30.

120. Kelly JF, Furukawa K, Barger SW, et al. Amyloid β-peptide disrupts carbachol-induced muscariniccholinergic signal transduction in cortical neurons.Proc Natl Acad Sci USA 1996;93:6753-8.

121. Huang HM, Ou HC, Hsieh SJ. Amyloid beta pep-tide impaired carbachol but not glutamate-mediatedphosphoinositide pathways in cultured rat corticalneurons. Neurochem Res 2000;25:303-12.

122. Selkoe DJ, Schenk D. Alzheimer’s disease: molecu-lar understanding predicts amyloid-based therapeu-tics. Annu Rev Pharmacol Toxicol 2003;43: 545-84.

123. Pike CJ, Cotman CW. Cultured GABA-immunore-

active neurons are resistant to toxicity induced by β-amyloid. Neuroscience 1993;56:269-74.

124. Olesen OF, Dago L, Mikkelsen JD. Amyloid β neu-rotoxicity in the cholinergic but not in the seroton-ergic phenotype of RN46A cells. Mol Brain Res1998;57:266-74.

125. Le W, Xie WJ, Kong R, et al. β-amyloid-induced neu-rotoxicity of a hybrid septal cell line associated withincreased tau phosphorylation and expression β-amy-loid precursor protein. J Neurochem 1997; 69:978-85.

126. Zheng WH, Bastianetto S, Mennicken F, et al.

Amyloid β peptide induces tau phosphorylation andneuronal degeneration in rat primary septal culturedneurons. Neuroscience 2002;115:201-11.

127. Jhamandas JH, MacTavish D. Antagonist of theamylin receptor blocks beta-amyloid toxicity in ratcholinergic basal forebrain neurons. J Neurosci2004;24:5579-84.

128. Jhamandas JH, Wie MB, Harris K, et al. Fucoidaninhibits cellular and neurotoxic effects of beta-amy-loid (A beta) in rat cholinergic basal forebrain neu-rons. Eur J Neurosci 2005;21:2649-59.

129. Behl C, Cole GM, Schubert D. Vitamin E protectsnerve cells from amyloid β protein toxicity.Biochem Biophys Res Commun 1992;186:944-50.

130. Mattson MP, Cheng B, Davis D, et al. β-amyloidpeptides destabilize calcium homeostasis and ren-der human cortical neurons vulnerable to excitotox-icity. J Neurosci 1992;12:376-89.

131. Hensley K, Carney JM, Mattson MP, et al. A modelfor β-amyloid aggregation and neurotoxicity based

on free radical generation by the peptide: relevanceto Alzheimer’s disease. Proc Natl Acad Sci USA1994;91:3270-74.

132. Perini G, Della-Bianca V, Politi V, et al. Role of p75neurotrophin receptor in the neurotoxicity by beta-amyloid peptides and synergistic effect of inflam-matory cytokines. J Exp Med 2002;195:907-18

133. ZhangY, Hong Y, Bounhar Y, et al., p75 neu-rotrophin receptor protects primary cultures of human neurons against extracellular amyloid betapeptide cytotoxicity. J Neurosci 2003;23:7385-94.

134. Busciglio J, Lorenzo A, Yeh J, et al. β-amyloid fib-rils induce tau phosphorylation and loss of micro-tubule binding. Neuron 1995;14:879-88.

135. Shea TB, Prabhakar S, Ekinci FJ. β-amyloid andionophore A23187 evoke tau hyper-phosphoryla-tion by distinct intracellular pathways: differentialinvolvement of the calpain/protein kinase C system.J Neurosci Res 1997;49:759-68.

136. Alvarez G, Munoz-Montano JR, et al. Lithium pro-tects cultured neurons against β-amyloid-inducedneurodegeneration. FEBS Lett 1999;453:260-64.

137. Mark RJ, Lovell MA, Markesbery WR, et al. A role

for 4-hydroxynonenal, an aldehydic product of lipidperoxidation, in disruption of ion homeostasis andneuronal death induced by amyloid β-peptide. JNeurochem 1997;68:255-64.

138. Greenberg SM, Kosik KS. Secreted β-APP stimu-lates MAP kinase and phosphorylation of tau inneurons. Neurobiol Aging 1995;16:403-8.

139. Takashima A, Honda T, Yasutake K, et al.Activation of tau protein kinase I/glycogen synthasekinase-3 beta by amyloid beta peptide (25-35)enhances phosphorylation of tau in hippocampal

neurons. Neurosci Res 1998;4:317-23.140. Alvarez A, Toro R, Caceres A, et al. Inhibition of tau phosphorylating protein kinase cdk5 prevents β-amyloid-induced neuronal death. FEBS Lett2001;459:421-6.

141. Hong M, Chen DCR, Klein PS, et al. Lithium reducestau phosphorylation by inhibition of glycogen syn-thase kinase-3. J Biol Chem 1997;40:25326-32.

142. Sadot E, Gurwitz D, Barg J, et al. Activation of m1muscarinic acetylcholine receptor regulates τ phos-phorylation in transfected PC12 cells. J Neurochem1996;66:877-80.

143. Hellstrom-Lindahl E, Moore H, Nordberg A.Increased levels of tau protein in SH-SY5Y cells

176 S. Kar et al.



after treatment with cholinesterase inhibitors andnicotinic agonists. J Neurochem 2000;74:777-84.

144. Yamaguchi Y, Kawashima S, Effects of amyloid-beta-(25-35) on passive avoidance, radial-arm mazelearning and choline acetyltransferase activity in therat. Eur J Pharmacol 2001;412:265-72.

145. Gonzalo-Ruiz A, Sang JM, Arevalo J, Amyloid betapeptide-induced cholinergic fibres loss in the cere-bral cortex of the rat is modified by diet high inlipids and by age. J Chem Neuroanat 2005;29:31-48.

146. Abe E, Casamenti F, Giovannelli L, et al.Administration of amyloid β-peptides into themedial septum of rats decreases acetylcholinerelease from hippocampus in vivo. Brain Res 1994;636:162-4.

147. Giovannini MG, Scali C, Prosperi C, et al. Beta-amyloid-induced inflammation and cholinergichypofunction in the rat brain in vivo: involvementof the p38MAPK pathway. Neurobiol Dis2002;11:257-74.

148. Boncristiano S, Calhoun ME, Kelly PH, et al.Cholinergic changes in the APP23 transgenicmouse model of cerebral amyloidosis. J Neurosci2002;22:3234-43.

149. Bronfman FC, Moechars D, Van Leuven F.Acetylcholinesterase-positive fiber deafferentationand cell shrinkage in the septohippocampal pathwayof aged amyloid precursor protein london mutanttransgenic mice. Neurobiol Dis 2000;7:152-68.

150. Wong TP, Debeir T, Duff K, et al. Reorganizationof cholinergic terminals in the cerebral cortex andhippocampus in transgenic mice carrying mutatedpresenilin-1 and amyloid precursor protein trans-genes. J Neurosci 1999;19:2706-16.

151. Jaffar S, Counts SE, Ma SY, et al. Neuropathologyof mice carrying mutant APPswe and/or PS1 M146Ltransgenes: alterations in the p 75NTR cholinergicbasal forebrain septohippocampal pathway. ExpNeurol 2001;170:227-43.

152. Gau JT, Steinhilb ML, Kao TC, et al. Stable β-

secretase activity and presynaptic cholinergic mark-ers during progressive central nervous systemamyloidogenesis in Tg2576 mice. Am J Pathol2002;160:731-8.

153. Vaucher E, Fluit P, Chishti MA, et al. Alteration inworking memory but not cholinergic receptor bind-ing sites in transgenic mice expressing human pre-senilin 1 transgenes. Exp Neurol 2002;21:323-9.

154. Dineley KT, Xia X, Bui D, et al. Accelerated plaqueaccumulation, associative learning deficits andupregulation of α 7 nicotinic receptor protein intransgenic mice co-expressing mutant human prese-nilin 1 and amyloid precursor proteins. J Biol Chem2002;227:22768-80.

155. Chishti MA, Yang DS, Janus C, et al. Early-onsetamyloid deposition and cognitive deficits in trans-genic mice expressing a double mutant form of amyloid precursor protein 695. J Biol Chem2001;276:21562-70.

156. Slowikowski SPM, Chishti MA, Zheng WH, et al.

Alterations in cholinergic parameters in the hip-pocampus of transgenic mice expressing mutatedamyloid precursor protein and/or presenilin-1 trans-genes. Soc Neurosci Abs 2002;295.18.

157. Apelt J, Kumar A, Schliebs R. Impairment of cholinergic neurotransmission in adult and agedtransgenic Tg2576 mouse brain expressing theSwedish mutation of human beta-amyloid precursorprotein. Brain Res 2002;953:17-30.

158. Hernandez D, Sugaya K, Qu T, et al. Survival andplasticity of basal forebrain cholinergic system inmice transgenic for presenilin-1 and amyloid pre-cursor protein mutant genes. NeuroReport 2001;12:1377-84.

159. German DC, Yazdani U, Speciale SG, et al.Cholinergic neuropathology in a mouse model of Alzheimer’s disease. J Comp Neurol 2003;462:371-81.

160. Hu L, Wong TP, Cote SL, et al. The impact of A β-plaques on cortical cholinergic and non-cholinergicpresynaptic boutons in Alzheimer’s disease-liketransgenic mice. Neuroscience 2003;121:421-32.

161. Feng Z, Chang Y, Cheng Y, et al. Melatonin allevi-ates behavioral deficits associated with apoptosis

and cholinergic system dysfunction in the APP 695transgenic mouse model of Alzheimer’s disease.J Pineal Res 2004;37:129-36.

162. Klingner M, Apelt J, Kumar A, Alterations incholinergic and non-cholinergic neurotransmitterreceptor densities in transgenic Tg2576 mousebrain with beta-amyloid plaque pathology. Int J DevNeurosci 2003;21:357-69.

163. Oddo S, Caccamo A, Green KN, et al. Chronicnicotine administration exacerbates tau pathologyin a transgenic model of Alzheimer’s disease. Proc

Natl Acad Sci USA 2005;102:3046-3051.164. Hartmann J, Erb C, Ebert U, et al., Central choliner-gic functions in human amyloid precursor proteinknock-in/presenilin-1 transgenic mice. Neuroscience2004;125:1009-17.

165. Haass C, Schlossmacher MG, Hung AY, et al.Amyloid β-peptide is produced by cultured cellsduring normal metabolism. Nature 1992;359:322-5.

166. Seubert P, Vigo-Pelfrey C, Esch F, et al. Isolationand quantification of soluble Alzheimer’s β-peptidefrom biological fluids. Nature 1992;359:325-7.

167. Shoji M, Golde TE, Ghiso J, et al. Production of Alzheimer’s β protein by normal proteolytic pro-cessing. Science 1992;258:126-9.




168. Iverfeldt K, Walaas SI, Greengard P. Altered pro-cessing of Alzheimer’s amyloid precursor protein inresponse to neuronal degeneration. Proc Natl AcadSci USA 1993;90:4146-50.

169. Wallace W, Ahlers ST, Gotlib J, et al. Amyloid pre-cursor protein in the cerebral cortex is rapidly and

persistently induced by loss of subcortical innerva-tion. Proc Natl Acad Sci USA 1993;90:8712-6.

170. Lin L, Georgievska B, Mattsson A, et al. Cognitivechanges and modified processing of amyloid pre-cursor protein in the cortical and hippocampal sys-tem after cholinergic synapse loss and muscarinic

receptor activation. Proc Natl Acad. Sci USA1999;96:12108-13.

171. Buxbaum JD, Oishi M, Chen HI, et al. Cholinergicagonists and interleukin 1 regulate processing andsecretion of the β /A 4 amyloid precursor protein.Proc Natl Acad Sci USA 1992;89:10075-8.

172. Allen DD, Galdzicki Z, Brining SK, et al. Beta-amy-loid induced increase in choline flux across PC12cell membranes. Neurosci Lett 1997;234:71-3.

173. Wurtman R. Choline metabolism as a basis for theselective vulnerability of cholinergic neurones.Trends Neurosci 1992;15:117-22.

178 S. Kar et al.



179

11Physiologic and Neurotoxic Properties

of A β PeptidesGillian C. Gregory, Claire E. Shepherd, and Glenda M. Halliday

11.1 Introduction

Alzheimer’s disease (AD) is characterized by agradual decline of numerous cognitive processes,culminating in dementia and neurodegeneration. Itis the most common form of dementia and a sig-nificant cause of death in the elderly. Definitivediagnosis of AD requires the presence of the extra-cellular accumulation of A β peptides in senileplaques in the cortex of the brain (Fig. 11.1) [1].β-Amyloid (A β) peptides are ~4-kDa polypeptides

with the main alloforms consisting of 40 and 42amino acids. Analysis of the insoluble protein frac-tion has identified the longer A β42 alloform as thepredominant peptide species in the neuropatho-logic accumulations (see [2]), although A β pep-tides of variable length accumulate within plaques[3–8]. The association between the abnormal accu-mulation of A β peptides in the brain and dementiais strong evidence that A β peptides are vital fornormal brain functioning.

Some of our understanding about A β and brainfunction has occurred after the identification of genetic mutations in the amyloid precursor protein(APP) that cause AD [9, 10] and the subsequentuse of molecular biology to study the cellularmechanisms involved in A β production and clear-ance. Initial reports using human APP695 mice andPDAPP mice with the APP717 mutation revealedthat these mutations caused A β levels to increasetwo to three times over control mice with A β dep-osition only occurring at these levels of production

[11–13]. Subsequent studies revealed that thesegenetic mutations increase the amount of the A β42alloform over other A β species [14–16]. The study

of these abnormalities in A β processing has led toa better understanding of the role A β peptides playwithin the brain.

11.2 Production of A β Peptides

The A β peptides are derived from the proteolyticprocessing of APP [17]. APP belongs to a het-erogenous group of ubiquitously expressedpolypeptides, with the heterogeneity arising from

alternative splicing and post-translational modifi-cations [18]. The pre-mRNA is spliced to producethree major isoforms APP 770 , APP 751 , and APP 695with the APP 695 isoform expressed at high levels inneurons (APP 770:751:695 mRNA ratio is 1:10:20in the cortex [19]). APP is a single membrane–spanning protein with a large extracellularN-terminal and small intracellular C-terminaldomain and is localized to numerous membranousstructures in the cell; the endoplasmic reticulum,Golgi compartments, and cell membrane [18]. Inthe axonal membrane, APP acts as a receptor forkinesin 1 during the fast axoplasmic transport of vesicles containing numerous proteins [20]. Inaddition to its possible role in membrane functions,APP undergoes considerable post-translationalmodifications including glycosylation and specificproteolytic cleavage to produce fragments that arebelieved to be extensively involved in adhesion,neurotrophic and neuroproliferative activity, inter-cellular communication, and membrane-to-nucleus

signaling [21].Proteolytic cleavage of APP occurs via at leasttwo pathways involving three secretases ( α , β,



and γ ), with only one pathway generating full-length A β peptide [18]. The α - and β-secretasecleavages are seen as mutually exclusive events,each releasing a large extracellular domain of theAPP protein, soluble APP (sAPP). α -Secretasecleavage precludes the formation of A β, insteadproducing a shortened fragment, together with γ -secretase cleavage, called p3 [22]. Production of these non-amyloidogenic sAPP and p3 fragmentsoccurs within the endoplasmic reticulum, the trans-Golgi apparatus, and at the cell membrane [23].

The A β peptides are generated early in the secre-tory trafficking of APP and at the cell surface. APP

not cleaved at the cell surface by α -secretase isreinternalized for processing in the endosome/lyso-some system by β-secretase [24, 25]. β-secretase,an aspartyl protease known as BACE ( β-site APPcleavage enzyme) [26], cleaves APP both withinthe endocytic and secretory pathways of the endo-plasmic reticulum and the Golgi [27]. The remain-ing APP fragment, the C-terminal fragment, issecured to the membrane. γ -Secretase cleavageoccurs in the hydrophobic transmembrane domain,after the α - or β-secretase cleavage events, and cre-ates the carboxyl terminus of the A β peptide.Studies suggest that A β peptides produced in theendoplasmic reticulum may not be secreted and areinstead retained and catabolized inside the cell[27]. Most A β, however, is believed to be secreted

into the extracellular space [18].The γ -secretase consists of a complex of proteins

made up of presenilin 1 and 2 (PS1 and PS2),nicastrin [28, 29], Aph-1 [30, 31], and pen-2 [31],though recent data suggest that different combina-tions of these proteins may exist [32]. This cleav-age event occurs at different sites in the C-terminalfragment producing the predominant A β1-40 andAβ1-42 fragments as well as A β1-39 and A β1-43 . It isnot clearly understood how the γ -secretase deter-mines its particular cleavage site in the C-terminalfragment and what regulates the production of onepeptide length over another. Such regulation islikely to have a substantial effect on overall A βfunction due to the different physicochemicalproperties of the peptides.

11.3 Detection and Tissue Locationof A β Peptides

The A β peptides can be detected in numerous bio-logical milieus, such as the CSF, plasma, and brain.Many studies have determined the concentrationsof the peptides in these different locations, pre-dominantly in the plasma and CSF because avail-ability and access to these areas is markedly easierthan brain tissue [33–46]. Comparisons and quan-tification of A β in plasma and CSF between controland AD samples have been performed for thedevelopment of biomarkers or objective predictorsof cognitive dysfunction [47]. However, conflictingresults have precluded any advances in this area

180 G.C. Gregory et al.

FIGURE 11.1. Tissue section from the temporal lobe of anearly-onset AD case immunohistochemically stained forAβ42. Initially, A β deposits in diffuse plaques that aretypically 10–200 µm in diameter with ill-defined bound-aries. Over time, the accumulating A β becomes fibrillaracquiring a

β-pleated sheet structure, and neuritic

plaques develop. These plaques are associated withaxonal and dendritic injury of pyramidal cells, known asdystrophic neurites, which occur both within this amy-loid deposit and immediately surrounding it. The accu-mulating A β in neuritic plaques develops further into theclassic senile plaques that have a distinct concentratedAβ core surrounded by a ring or “corona” of neuriticpathology.



because A β peptide concentrations in both CSFand plasma are highly variable [33, 35, 45, 48, 49].

The CSF bathes and drains from the brain, whichimplies that CSF A β mainly arises from brain tis-sue and in nondiseased states reflects brain tissueconcentrations of these peptides. In control CSF,Aβ40 is the dominant species, with concentrationsconsistently higher than A β42 [33–36]. This sug-gests that the dominant A β peptide secreted by thecells of the brain is A β40 and that γ -secretase cleav-

age preferentially produces this shorter A β peptide.It has been shown that CSF A β levels follow a nat-ural U-shaped course in normal aging (Fig. 11.2).Proportionately higher concentrations of both A β40and A β42 are detected in children compared withadults between 30 and 60 years of age [36, 37].Concentrations then increase proportionately withfurther aging [36]. Low levels of A β during adult-hood suggests that equilibrium has been reachedbetween the cellular synthesis and extracellular

11. Physiologic and Neurotoxic Properties of A β Peptides 181

Breakdown in one ofthese pathways

LDLRACEα 2M

Complexwith

Microglia

1. Proteolysis

α 2M

ApoE

LDLR

ACEApoE

3. BBB

CSF A βBrain A β

Brain A βCSF A β

Normal AD

2. Endocytosis

Pathways of clearance of A β

Accumulation of A β in the brain

AgeAge

A m o u n t

A m o u n t

Enzymes

FIGURE 11.2. Graphs depicting normal (left) and abnormal (right) A β brain levels and a diagram depicting the mech-anisms of A β clearance from the brain. The left-hand graph shows the natural U-shaped course of CSF A β duringnormal aging. Proportionately high concentrations of both A β40 and A β42 occur in childhood and are then downreg-ulated between the ages of 30 and 60 years. A β peptide levels then proportionately increase with subsequent aging.Low levels of A β during adulthood suggests that equilibrium has been reached between the cellular synthesis andextracellular clearance of these peptides, and that with older age this equilibrium is changed. The diagrams in thelower part of the figure depict A β clearance mechanisms. Normal removal of A β from the brain occurs via extracel-lular proteolysis, receptor-mediated endocytosis, and transport across the blood brain barrier (BBB) via angiotensin-converting enzyme (ACE) and α 2-macroglobulin ( α 2M) through interactions with LDL-receptor–related protein(LDLR) and apolipoproteins (ApoE). The right-hand graph shows that breakdown in one of the clearance pathways,and failure to clear the A β peptide, leads to increased brain A β and, hence, AD.



clearance of these peptides and that with older agethis equilibrium is changed (Fig. 11.2).

Numerous studies of CSF A β in AD show a con-sistent decrease in A β42 concentrations comparedwith controls [33–35, 37–44] and a negative corre-lation between A β

42levels and disease severity [40,

50]. A β40 levels in AD CSF remain the same[33–35, 37, 39] or decrease [40, 43] compared withcontrols. The lower A β42 CSF levels in AD arethought to be due to reduced A β42 clearance con-sistent with the preferential deposition of A β42 inAD brain [51]. However, there is an overlap in CSFAβ42 values between AD and control groups [35]with the clearance problem occurring primarily inearly disease [50]. More intriguing are studies thatshow low CSF A β42 levels in patients with a vari-

ety of other disorders, some of which do notdeposit A β in the brain. These include majordepression [40, 50] and Creutzfeldt-Jakob disease[52], suggesting a possible dissociation betweenAβ clearance and deposition. In addition, the samedeficit occurs in patients with dementia with Lewybodies [53] limiting the role of this measurement asa specific diagnostic marker for AD. Overall, thesefindings suggest that A β40 is preferentially clearedthrough the CSF at all ages and in all brain disor-ders compared with A β

42.

In the plasma of normal elderly, the A β40 peptideis the dominant species, with average concentra-tions of A β40 well above those of A β42 [35, 45, 46].Plasma A β originates from many sources, but par-ticularly blood-borne platelets, which preferen-tially produce A β40 [54]. Platelet activationreleases A β, and in patients with AD there is anincrease in the plasma concentrations of A β, par-ticularly A β42 [45, 55, 56]. The binding of platelet-activating factor to platelets in AD has been used to

measure platelet activation. This measure corre-lates with the degree of cognitive impairment inpatients with AD [57], with decreasing plateletAPP predicting conversion to dementia [58]. Thisraises the possibility that increased platelet activa-tion and plasma A β may play some role in thedementing process.

Aβ peptides complex with apolipoprotein E(ApoE) and apolipoprotein J (ApoJ) to cross theblood-brain barrier (BBB) [59]. In primates, infusedAβ

40readily crosses the BBB compared with other

peptides, with the rate of A β sequestration into thebrain parenchyma after a single exposure increasing

with age [60]. In rats, infusions of A β40 or A β42increase BBB permeability [61]. Enhancement of Aβ transport across the BBB along with reducedCSF clearance is thought to contribute to theincreased brain deposition of A β in a transgenicmodel of AD [62]. Alternatively, intravenous admin-istration of anti-A β antibody promotes a rapid effluxof Aβ from the CNS into plasma [63]. These studiesshow considerable flux of A β peptide across theBBB and suggest that a proportion of brain A βcould originate from the circulating pool found inplasma.

The “amyloid cascade” hypothesis proposes thatthe increased burden of A β in the brain is the pri-mary intrinsic pathogenic event in AD [64].Consequently, most studies analyzing brain A β

peptide levels have concentrated on AD tissue withfew studies focusing on A β levels in normal (dis-ease free) brain tissue [3, 15, 65–79]. In contrastwith the results obtained in CSF and plasma, alarge number of these studies show that A β40 levelsin elderly controls are low compared with the lev-els of A β42 (for review, see [2]). This suggests thatAβ40 is preferentially cleared from the brain, con-sistent with higher levels in the CSF. Despite theseconsistent findings, the literature commonly statesthat A β

40is the dominant peptide species in the

normal brain (for review, see [2]). This misconcep-tion is consistent with measurements from periph-eral tissues and supernatant from cell lines(equivalent of CSF) [80] but is not supported bydata from nondiseased human brain tissue.Unfortunately, this has also influenced researchinto AD pathogenesis to focus on changes in theproduction from the more “normal” A β40 peptideto the A β42 peptide that has been wrongly thoughtto only associate with AD.

11.4 Structure of A β Peptides

Aβ peptides exist as monomers, dimers, and higheroligomers, with aggregation producing protofibrilsand eventually fibrils, in a β-pleated sheet confor-mation. The A β oligomers are believed to play akey role in AD neurotoxicity [81–85]. The forma-tion of A β oligomers by the different alloformsoccurs through different pathways. A β

40aggregates

as monomers, dimers, trimers, and tetramers inrapid equilibrium, whereas A β42 preferentially




forms pentamer/hexamer units that are able toassemble further to form early protofibril structures[86, 87]. These differences suggest differentpeptide functions.

Recent experiments have established that themajor secondary structure adopted by A β dependson the environment [88]. The A β monomer con-tains an amphipathic sequence that favors an α -helix structure (Fig. 11.3) in a membrane ormembrane-mimicking environment [89, 90],whereas in an aqueous solution, a nontoxic randomcoil configuration with few components of α -helixand/or β-sheet conformations is preferred [91–94].The highly hydrophobic C-terminus of A β isembedded in the lipid membrane with itshydrophilic N-terminus protruding extracellularly

[95]. Two lipophilic regions (Lys16 to Ala21 andLys28 to Val40) are believed to be the main func-tional areas. The first region has an α -helical struc-ture and the second a β-pleated sheet structure,which is able to form hydrophobic forces withother β-sheets of A β peptides [91]. The twolipophilic helical regions are separated by a flexiblehinge or kink region (Fig. 11.3), which may beimportant for its membrane-inserting propertiesand conformational rearrangements [89, 95, 96].

The different lengths and structure of the A βpeptides contribute to their different oligomericstates. A β aggregation into oligomers occurs whenthe dominant structure of A β is converted from anα -helix or random coil to a β-sheet conformation[97, 98] through intermediates of mixed helices

and β-sheets [88, 92]. In contrast with A β42, A β40has a tendency to move out of the lipid environ-ment [88], possibly contributing to the smaller andmore soluble oligomers formed by this peptide.

In disease conditions, when A β fibrillogenesisoccurs, the structure of the A β peptides changessubstantially due to increased concentrations andconformational effects. Over time, the helical A βresidues 29–40 that are embedded into the stabiliz-ing cell membrane leave the lipid bilayer and enterthe extracellular environment where they have ahigh tendency to form short β-sheets in a concen-tration-dependent fashion thereby precipitatingpolymers [88, 92]. During the “lag phase” prior tothe development of A β fibrils, no A β precipitatesare detectable in brain tissue, suggesting that nucle-ation of a different structure is required, like seed-ing a crystallization process. The lag phase can beremoved by seeding A β monomers with preaggre-gated A β fibrils [99]. Using kinetic studies, A β42has been shown to form precipitated fibrils signifi-

cantly faster than A β40, leading to the frequentlycoined phrase that A β42 is more amyloidogenicthan A β40 [99]. This is probably due to its greaterpropensity for helical structures and lipid associa-tion. In fact, A β40 has been shown to be compara-tively neuroprotective against A β42 -inducedneurotoxicity in vitro and in vivo. The mechanismfor this neuroprotection may involve the A β40 pep-tide inhibiting the β-sheet transformation and fibrilformation of A β42 [100].

Comparison between the concentrations of solu-ble and insoluble A β peptides in control brain tis-sue [3, 66, 69, 72, 74, 78] suggests that A β40 is


Aβ1-40

Aβ1-42

FIGURE 11.3. Membrane-bound structure of the main A βpeptides, A β40 and A β42. Both peptides exhibit α -helicalconformations (shown as large arrows) in conditionsmimicking lipid membranes (in the presence of organic

modifiers such as SDS). A β42 has two α -helices, oneither side of the “kink” region, in contrast with A β40,which has only one α -helical domain.



greater in the soluble fraction, whereas A β42 is thepredominant species in the insoluble fraction [2],as may be expected based on the physiochemicalproperties of the two peptides. There is a signifi-cant change in the A β levels in the brain tissue of AD cases (both sporadic [3, 15, 65–67, 70, 72, 73,75, 78, 79, 101, 102] and familial [15, 65, 70, 77,101, 103, 104]), with significant increases in theamount and insolubility of A β42 in AD comparedwith controls (Fig. 11.2), in agreement with thedominant hypothesis that it is the pathogenicspecies in AD. In addition to the changes in A β42,Aβ40 levels are also increased in AD cases(Fig. 11.2), with greater increases in the amount of insoluble A β40 than insoluble A β42 in sporadic AD(for review, see [2]). These studies support the con-

cept that increases in A β peptide levels promotesignificant changes in their structure and thereforetheir solubility and that these structural changesproduce less soluble A β peptides and have signifi-cant pathogenic effects.

11.5 Other A β Binding Partners

Apart from concentration-dependent self-aggrega-tion, A β peptides readily bind to other molecules,including lipids, proteins, and metal ions. Threehistidine residues in the N-terminal hydrophilicregion provide primary metal binding sites on theAβ peptides. The binding of certain metal ions toAβ can promote aggregation. Zn 2+ induces A βaggregation at acidic to neutral pH and is the mostpowerful metal inducer of A β aggregation [105].Cu 2+ induces aggregation at mildly acidic pH com-parable with the pH-dependent effect of Cu 2+ oninsulin aggregation [105]. Under normal physio-

logic conditions, Cu 2+ protects A β against Zn 2+-induced aggregation by competing with Zn 2+ forthe histidine residues of A β [106]. A mildly acidicenvironment together with increased Zn 2+ and Cu 2+

are common features of inflammation, which sug-gests that A β aggregation by these factors may bea response to local injury [105].

Lipid membranes are important binding partnersfor A β as the peptide plays a role in the regulationof lipid membrane function, metabolism, andhomeostasis [107]. The binding efficacy of lipids toAβ increases when A β forms polymers [108] withthe lipids binding to the hydrophobic areas of

aggregated A β. Cholesterol is a key component of membranes and interacts with A β in a reciprocalmanner [107]. Aggregated A β40 in particular has ahigh affinity for cholesterol with oligomeric A βpeptides promoting the normal release of lipidfrom neurons [109]. These A β-lipid particles havea very low binding affinity for neurons, reducinglipid internalization and thereby affecting intracel-lular lipid metabolism. Gangliosides (sialylatedglycosphingolipids) are the predominant glycanson neuronal plasma membranes and are concen-trated into membrane rafts by cholesterol wherethey mediate important physiological functions.These lipid rafts (made of cholesterol, sphin-gomyelin, and glycosphingolipids such as GM1ganglioside) play an essential role in cell-cell com-

munications and signal transduction across mem-branes [110]. GM1 ganglioside associates withcholesterol and binds to A β peptides, with GM1ganglisoside–bound A β acting as a seed for A βfibrillogenesis [111].

In addition to the binding of A β to lipids, A βalso binds to lipid-trafficking lipoproteins. A βcomplexes with ApoJ, a universal lipoproteinexpressed in many cells throughout the body.Soluble A β also binds to normal human plasmahigh-density lipoprotein (HDL), includingapolipoprotein A (ApoA)-I, ApoA-II, ApoE, andApoJ [112]. A β. binding with ApoE, alleles E2 andE3, form stable membrane-bound complexes thatare more abundant than ApoE4-A β. complexes[113]. In contrast with neurons, A β-ApoE lipidparticles are internalized mainly by glia and vascu-lar cells presenting a clearance pathway throughwhich parenchymal A β is modulated [114].Exogenous ApoE3 but not ApoE4 prevents A β-induced neurotoxicity by a process requiring ApoE

receptors [113].A subset of plasma membrane proteins and recep-

tors also bind A β (for review, see [115]). Heparansulfate proteoglycans are cell-surface binding sitesfor A β. The serpin-enzyme complex receptor andthe insulin receptor can bind monomeric forms of Aβ peptides. The alpha7nicotinic acetylcholinereceptor, integrins, RAGE (receptor for advancedglycosylation end-products), and formyl peptidereceptor-like 1 are able to bind monomeric and fib-rillar forms of A β peptides. In addition, APP, colla-gen-like Alzheimer’s amyloid plaque componentprecursor/collagen XXV, the NMDA ( N -methyl- D-




aspartate) receptor, P75 neurotrophin receptor, scav-enger receptors A, BI, and CD36 and complexesbind fibrillar forms of A β peptides. It is thereforelikely that the function of A β differs depending onthe associated binding partners, which are modu-lated by its structure and solubility.

11.6 Function of the A β Peptides

The functional properties of the A β peptides havenot been completely elucidated to date, thoughnumerous studies suggest that the peptides possessa number of neurotrophic and neurotoxic proper-ties. As stated above, the divergent roles of A β seemdependent on their physicochemical properties,

aggregation state, and binding partners, with A β40function primarily studied (both neurotoxic andtrophic) due to its greater solubility. Recent studiessuggest that soluble A β plays important roles in thefacilitation of neuronal growth and survival, in themodulation of synaptic function, and in neurotoxicsurveillance and defense against oxidative stress[116, 117], whereas oligomeric and fibrillar A βhave less trophic and greater toxic properties.

11.6.1 Neurotrophic FunctionsRecent studies have shown that A β peptides may bevital for neuronal development, plasticity, and sur-vival due to its integral membrane interactions[118]. Neuronal viability appears to be dependent onAβ [117] with the peptide possessing neurogenicproperties [119]. Despite some controversy [120,121], there is increased differentiation of hippocam-pal neural stem cells treated with A β42, with nochange to the rate of cell death or proliferation.

Interestingly, this effect is only seen with solubleoligomeric A β42 peptide, as neither monomericAβ42, Aβ25-35 , nor A β40 (aggregated or not) increasedthe percentage of neurons [119]. This may suggestthat the formation of new neurons is induced by themore “soluble” forms of A β42 that form larger pen-tamer/hexamer subunits and membrane channels.

11.6.2 Physiologic Functions

Because A β binds to the plasma membranes inboth soluble and fibrillar forms, it changes thestructure and function of the membranes by

modifying the fluidity or forming ion channels[115]. Soluble A β40 increases voltage-gatedK(+) channel currents in cerebellar granular neu-rons without neurotoxic consequences [122].Neuronally released soluble A β selectivelydepresses excitatory synaptic transmission throughinteractions with NMDA receptors [116]. Themodification of membrane channels in vascularsmooth cells causes vasoconstriction, with A β40having significantly greater vasoconstrictive effectscompared with A β42 [123]. The negative feedback after synaptic excitation coupled with an ability toreduce local blood flow and oxygen and glucosedelivery would keep neuronal hyperactivity incheck [116]. This suggests that the nonpathologicsoluble forms of A β are important synaptic protec-

tors through their ability to change ionic channelfunctions within cell membranes [122].

Monomeric A β peptide is also thought to havean antioxidant function through its metal-bindingcapabilities, particularly capturing Zn, Cu, and Feions and preventing them from participating inredox cycling with other ligands [124]. A β produc-tion increases with oxidative stress [125–127], andthe peptides may be involved in altering ion fluxesby chelating metal ions in an attempt to preventoxidation [128]. This suggests that A β production,in conjunction with its neuroprotective and neu-rotrophic properties, may be a normal stressresponse to minimize oxidative damage [129]. Theformation of diffuse A β plaques in AD may be acompensatory event for the removal of reactiveoxygen species.

11.6.3 Neurotoxic Properties

The key to A β cellular toxicity appears to be its

aggregation state [130]. A β appears to promoteneuron degeneration only when the peptideassumes a particular β-pleated structure either inoligomeric and/or fibrillar forms. Yankner and col-leagues first showed that synthetic A βl-40 was neu-rotoxic in primary rat hippocampal cell cultures[131]. Roher et al. reported that A β isolated fromAD brains inhibited neurite sprouting and causedcell death in cultured sympathetic neurons [132].Further studies then demonstrated that the toxicityof the peptide was strongly correlated with itspropensity to form fibrillar aggregates [130,133–137]. However, more recent work has




indicated that oligomeric A β, the A β form requiredprior to fibrillization, may be the most toxic speciesinvolved in neuronal death [81–85]. Studies haveshown that oligomeric A β induces greater celldeath and apoptosis than soluble or fibrillar forms[138, 139], confirming that the structural confor-mation of the peptide is important in determiningits physiological action.

A change in the binding properties of A β pep-tides may induce significant toxicity. In particular,the interaction between oligomeric A β and lipidsmay be an important cause of neuronal degenera-tion and would certainly impact on lipid homeosta-sis and function [109]. Michikawa and colleaguespropose that the stimulation of lipid release fromneurons by the increase in oligomeric A β in AD

induces a disruption of cholesterol homeostasis andmembrane raft maintenance in the brain, with theconsequent neurotoxic changes such as an increasein tau phosphorylation [109, 140].

A change in the neurotrophic properties of A βpeptides may also induce considerable toxicity.Physiological levels of A β can interfere withfunctions critical for neuronal plasticity [141].Pretreatment of neurons with sublethal concentra-tions of the more amyloidogenic A β1-42 suppressesthe phosphorylation of cAMP-response elementbinding protein (CREB) and the downstream acti-vation of brain-derived neurotrophic factor(BDNF). As both CREB and BDNF play criticalroles in neuronal plasticity, an increase in the A β1-42suppression of this function may play a role in thecognitive deficits associated with AD [141].

Significant toxicity may also be induced by achange in the regulation of synaptic feedback andlocal blood flow by A β peptides. Increased releaseof A β from neurons significantly downregulates

synaptic activity [116], and increased A β bindingto vascular smooth muscle cells increases vasocon-striction and decreases local blood flow [123].These changes would reduce synaptic function andtherefore affect cognition. A β aggregation alsochanges synaptic properties due to downstreamincreases in intracellular free Ca 2+ and decreasedtransmitter manufacturing through lower enzymeactivities [142].

Changes in metal binding to A β peptides mayalso induce significant toxicity due to increasedoxidation [143–146] leading to mitochondrialdysfunction [147]. The methionine residue 35

(met-35) of A β is critical to its oxidative stress andneurotoxic properties, with its removal abolishingthe neurotoxic properties of A β1-42 [148]. AlthoughZn2+ binding induces the greatest A β aggregation,the oxidative toxicity of A β in cell culture is medi-ated through its interaction with Cu 2+ and Fe 3+

[149, 150]. A β catalyzes the reduction of Cu 2+ toCu+ and Fe 3+ to Fe 2+, generating H 2O2 from molec-ular oxygen and available biological reducingagents such as vitamin C, cholesterol, and cate-cholamines [150]. Any reduced activity of thedetoxifying enzymes, such as cytosolic Cu/Znsuperoxide dismutase (SOD1), catalase, and/orglutathione peroxidase, allows H 2O2 to furtherreact with reduced Fe 2+ and Cu + to generate toxichydroxyl radicals. A β42 has greater oxidative toxi-

city than A β40 [149] due to their relative Cu 2+ andFe3+ reducing potentials and the ability to catalyti-cally generate H 2O2 from biological reducingagents [150].

11.7 Clearance of A β Peptidesfrom the Brain

Aβ clearance occurs through at least three path-ways (Fig. 11.2): extracellular proteolysis bydegrading enzymes [151], transport across theBBB [152], and receptor-mediated endocytosis[152]. Several proteolytic enzymes have beenimplicated in the degradation of A β. Two metallo-proteinases; insulin-degrading enzyme (IDE) andendothelin-converting enzyme (ECE) 1 and 2[153], the plasmin system, and a neutral endopep-tidase known as neprilysin are involved in theextracellular degradation of A β [154–156]. IDEacts on soluble monomeric and particularly intra-

cellular A β [157, 158], whereas plasmin is capableof degrading aggregated A β [156]. The ECE zincmetallopeptidases are a class of type II integralmembrane protein named for their ability tohydrolyze a family of biologically inactive inter-mediate endothelins [159]. ECE-1 has been shownto cleave A β at multiple sites within the peptidesequence, with ECE inhibitors significantlyincreasing the accumulation of A β in culture, indi-cating a role for this protease in A β catabolism[153]. Neprilysin plays a major role in A β

42degra-

dation [160] with this enzyme concentrating in thebrain regions most vulnerable to AD [161]. A loss




of such clearance mechanisms may be responsiblefor the accumulation of A β with recent work show-ing that the degrading activity of neprilysin isinsufficient to clear brain A β accumulation ineither AD or pathologic aging [162].

Aβ transport across the BBB is less well under-stood. A β is thought to be able to move from theextracellular spaces into the perivascular pathways,along the small and large intracranial artery walls,possibly draining to the lymph nodes in the neck [163]. This mechanism of clearance occurs viathe endothelium, mediated by the enzymesangiotensin-converting enzyme and α 2-macroglobu-lin through interactions with LDL-receptor–relatedprotein and apolipoproteins [164, 165]. Microgliaand astrocytes also take up A β through receptor-

mediated mechanisms [166, 167]. A β-ApoJ com-plexes are transported over the BBB through theApoJ receptor megalin [59]. The high affinity of aggregated A β40 with cholesterol suggests that cho-lesterol bound peptide trafficking may also play arole in its removal from the extracellular space[108]. A β40 transport across the BBB is faster thanAβ42 [168] with A β40 the predominant constituentof abnormal A β peptide deposits in blood vesselwalls [169]. There is some evidence that age-asso-ciated changes in BBB transport stops the efflux of Aβ42 via this route [168].

Although still poorly understood, it appears thata number of regulatory mechanisms are importantfor modulating A β levels in the brain (Fig. 11.2).Under normal circumstances, local catabolism orclearance mechanisms efficiently prevent accumu-lation of these amyloidogenic peptides in the brain[170]. In AD, the considerable build-up of A β pep-tides suggests difficulties with A β clearance even if other production pathways are affected. In the

absence of knowing any common initiating eventor mechanism for AD, modification of clearancepathways provides the most obvious therapeutictargets for this disease.

11.8 Potential TherapeuticStrategies for A β Toxicity

Genetic and animal models of AD have provided animportant basis for the design and testing of thera-peutic strategies to alter A β production, aggregation,and/or accumulation. Strategies for lowering A β

production include secretase inhibitors [171].Strategies for reducing A β aggregation includemetal chelators [172], and strategies for ameliorat-ing A β accumulation include A β immunization,nonsteroidal anti-inflammatory drugs (NSAIDs),peroxisome proliferator-activated receptor- γ (PPAR)agonists, and statin medication [173].

11.8.1 Secretase Inhibitors

Since identifying the importance of β- and γ -secre-tase in the production of the A β alloforms, thera-peutics aimed at inhibiting these enzymes havebeen the focus of a great deal of research. Initialstudies of BACE1 therapy in mouse modelsappeared promising as, despite their role in normal

physiological functioning, BACE1/BACE2 doubleknockout animals do not show any phenotypicproblems (for review, see [174]). To date, no BACEinhibitors have been trialed in the literature,although significant numbers have been patented[175]. In contrast, models knocking out γ -secretasehave been more problematic behaviorally due tothe importance of PS1 in the γ -secretase proteincomplex and Notch signaling [176]. Fortunately,specific γ -secretase inhibitors have recently shownpromising results with a shift toward the produc-tion of the less toxic A β38 alloform and a reductionin A β40 and A β42 both in vitro and in transgenicmice [177, 178]. Importantly, these effects wereachieved without affecting other components of theγ -secretase complex, although clinical trials havenot yet been carried out. Unfortunately, clinical tri-als of 70 AD patients with the γ -secretase inhibitorLY450139, which showed promising results in ani-mal models, have failed to show a marked reduc-tion in CSF A β42 [179]. Although there is still great

promise for the development of specific and effica-cious γ -secretase inhibitors, many researchers arecalling on the development of BACE1 inhibitors asa safer alternative.

11.8.2 Metal Chelators

Given the interaction between A β and metal ions,and the suggestion that they may mediate A βaggregation and toxicity, therapeutic strategieshave focused on disrupting this interaction. Manyof these studies have generated promising data withthe demonstration that specific chelators of Zn and




Cu ions can solubilize A β plaques from Alzheimer’sdisease postmortem brain tissue [180]. The com-pound used, cloquinol, also substantially decreasedAβ deposition in the brains of transgenic mice after

just 9 weeks of treatment [181]. This drug alsoslowed the rate of cognitive decline in a clinicaltrial of AD and controls and appeared to be well-tolerated among patients [182]. Interestingly, thisimprovement was only reported as evident in indi-viduals who were more severely impaired andscored over 25 on the Alzheimer’s DementiaAssessment Scale–cognition subscale (ADAS-cog), although this could have been a type I errorand greater sample numbers need to be assessed. Incontrast, while no significant effect on cognitionwas seen in individuals who scored below 25 (the

authors suggest a lack of sensitivity in this measure[182]), their plasma A β42 levels were significantlydecreased. These discrepant results warrant furtherexperimental studies in this area, although giventhe heterogeneous roles of A β and the potentialantioxidant roles arising from an interactionbetween A β and metal ions, great caution isrequired when trialing such therapies.

11.8.3 A β Immunization

Recent evidence suggests that reducing A β deposi-tion in the brain by way of immunotherapy canreverse disease-associated functional deficits [183,184]. The immunization of transgenic APP micewith A β42 appears to prevent the formation of A β-containing plaques and subsequent AD-relatedneuropathologic changes in animals as young as6 weeks to 11 months [184]. This reduction in A βis associated with reductions in memory impair-ment [185]. Similar results occur with the adminis-

tration of other A β alloforms [186] and shorterpeptide fragments [187], as well as with peripheralimmunization with A β antibodies [188]. Clinicaltrials using active A β42 immunization, however,caused severe central nervous system inflammationin a small but significant number of subjects [189].Although no definitive data exists, it is generallyagreed that these side effects were attributable to acytotoxic T-cell–mediated response against A β,raising questions about immunizing against a self-protein and the effect of such a reaction on normalpeptide function [190]. An additional safety con-cern arises with the use of A β alloforms that are

capable of forming toxic fibrils and seeding plaqueformation [191]. Despite this data, neuropathologicstudies of patients treated with the AB vaccineshowed low levels of cortical A β [192]. In addition,those subjects who developed robust antibody titersdid show some clinical improvement [193]. Thesedata provide support for the continued develop-ment of immunization strategies in the treatment of AD.

Active immunization with nontoxic A β frag-ments may be more effective in clinical trials asthey have been shown to have reduced fibrillogenicproperties while maintaining immunogenicity intransgenic mice [187]. More recent studies havealso shown promising results from intracerbroven-tricular immunization of A β fragments in trans-

genic mice [194], thereby avoiding perivascularhemmorhage concerns associated with intravenousadministration. Despite promising results usingtransgenic murine models, these animals stillexpress endogenous APP and are therefore lesslikely to reflect the autoimmune problems that maybe associated with human A β vaccines. With this inmind, the serious adverse immune reactions seen inclinical trials highlights the need to test potentialtherapies in large primate cohorts [195] prior toclinical testing in patients.

11.8.4 NSAIDs and PPAR- γ Agonists

Epidemiological evidence indicates that NSAIDsmay lower the risk of developing AD [196, 197].Although a direct effect on reducing the damag-ing A β-stimulated inflammation has been postu-lated, recent studies have demonstrated thatNSAIDs are capable of directly affecting A β pro-duction via several mechanisms. Ibuprofen,

indomethacin, and sulindac sulfide are capable of reducing A β42 production, and increasing the lesstoxic A β38 alloform, in cultured cells [198].These effects have also been reported in trans-genic mice and are proposed to occur by shiftingγ -secretase activity [199]. Unfortunately, clinicaltrials of NSAIDs have been less fruitful [200],possibly due to the fact that most trials have beencarried out in AD patients where the disease istoo advanced for NSAID therapy to be effective.However, recent reports suggest that the dosesrequired to lower A β in patients may be toxic[201] and better results may be achieved through




8. Mann DM, Iwatsubo T, Pickering-Brown SM, et al.Preferential deposition of amyloid beta protein(Abeta) in the form Abeta40 in Alzheimer’s diseaseis associated with a gene dosage effect of theapolipoprotein E E4 allele. Neurosci Lett1997;221:81-4.

9. Goate A, Chartier-Harlin MC, Mullan M, et al.Segregation of a missense mutation in the amyloidprecursor protein gene with familial Alzheimer’s dis-ease. Nature 1991;349:704-6.

10. Van Broeckhoven C, Haan J, Bakker E, et al.Amyloid beta protein precursor gene and hereditarycerebral hemorrhage with amyloidosis (Dutch).Science 1990;248:1120-2.

11. Borchelt DR, Ratovitski T, van Lare J, et al.Accelerated amyloid deposition in the brains of trans-genic mice coexpressing mutant presenilin 1 andamyloid precursor proteins. Neuron 1997;19:939-45.

12. Games D, Adams D, Alessandrini R, et al.Alzheimer’s-type neuropathology in transgenic miceoverexpressing V717F beta-amyloid precursor pro-tein. Nature 1995;373:523-7.

13. Hsiao K, Chapman P, Nilsen S, et al. Correlativememory deficits, Abeta elevation, and amyloidplaques in transgenic mice. Science 1996;274:99-102.

14. Suzuki N, Cheung TT, Cai XD, et al. An increasedpercentage of long amyloid beta protein secreted byfamilial amyloid beta protein precursor (beta

APP717) mutants. Science 1994;264:1336-40.15. Tamaoka A, Odaka A, Ishibashi Y, et al. APP717

missense mutation affects the ratio of amyloid betaprotein species (A beta 1-42/43 and a beta 1-40) infamilial Alzheimer’s disease brain. J Biol Chem1994;269:32721-4.

16. Scheuner D, Eckman C, Jensen M, et al. Secretedamyloid beta-protein similar to that in the senileplaques of Alzheimer’s disease is increased in vivoby the presenilin 1 and 2 and APP mutations linkedto familial Alzheimer’s disease. Nat Med

1996;2:864-70.17. Haass C, Schlossmacher MG, Hung AY, et al.Amyloid beta-peptide is produced by cultured cellsduring normal metabolism. Nature 1992;359:322-5.

18. Selkoe DJ. Alzheimer’s disease: genes, proteins, andtherapy. Physiol Rev 2001;81:741-66.

19. Tanaka S, Shiojiri S, Takahashi Y, et al. Tissue-spe-cific expression of three types of beta-proteinprecursor mRNA: enhancement of proteaseinhibitor-harboring types in Alzheimer’s diseasebrain. Biochem Biophys Res Commun 1989;165:1406-14.

20. Kamal A, Almenar-Queralt A, LeBlanc JF, et al.Kinesin-mediated axonal transport of a membrane

compartment containing beta-secretase and prese-nilin-1 requires APP. Nature 2001;414:643-8.

21. Turner PR, O’Connor K, Tate WP, et al. Roles of amyloid precursor protein and its fragments in regu-lating neural activity, plasticity and memory. ProgNeurobiol 2003;70:1-32.

22. Haass C, Hung AY, Schlossmacher MG, et al. beta-Amyloid peptide and a 3-kDa fragment are derivedby distinct cellular mechanisms. J Biol Chem1993;268:3021-4.

23. Nunan J, Small DH. Regulation of APP cleavage byalpha-, beta- and gamma-secretases. FEBS Lett2000;483:6-10.

24. Brown MS, Ye J, Rawson RB, et al. Regulatedintramembrane proteolysis: a control mechanismconserved from bacteria to humans. Cell2000;100:391-8.

25. Ulery PG, Beers J, Mikhailenko I, et al. Modulationof beta-amyloid precursor protein processing by thelow density lipoprotein receptor-related protein(LRP). Evidence that LRP contributes to the patho-genesis of Alzheimer’s disease. J Biol Chem2000;275:7410-5.

26. Vassar R, Bennett BD, Babu-Khan S, et al. Beta-sec-retase cleavage of Alzheimer’s amyloid precursorprotein by the transmembrane aspartic proteaseBACE. Science 1999;286:735-41.

27. Cook DG, Forman MS, Sung JC, et al. Alzheimer’sA beta(1-42) is generated in the endoplasmic reticu-

lum/intermediate compartment of NT2N cells. NatMed 1997;3:1021-3.

28. Yu G, Nishimura M, Arawaka S, et al. Nicastrinmodulates presenilin-mediated notch/glp-1 signaltransduction and betaAPP processing. Nature2000;407:48-54.

29. Esler WP, Kimberly WT, Ostaszewski BL, et al.Activity-dependent isolation of the presenilin-gamma-secretase complex reveals nicastrin and agamma substrate. Proc Natl Acad Sci U S A2002;99:2720-5.

30. Goutte C, Tsunozaki M, Hale VA, et al. APH-1 is amultipass membrane protein essential for the Notchsignaling pathway in Caenorhabditis elegansembryos. Proc Natl Acad Sci U S A 2002;99:775-9.

31. Francis R, McGrath G, Zhang J, et al. aph-1 and pen-2 are required for Notch pathway signaling, gamma-secretase cleavage of betaAPP, and presenilin proteinaccumulation. Dev Cell 2002;3:85-97.

32. Gu Y, Sanjo N, Chen F, et al. The presenilin proteinsare components of multiple membrane-bound com-plexes that have different biological activities. J BiolChem 2004;279:31329-36.

33. Ida N, Hartmann T, Pantel J, et al. Analysis of het-erogeneous A4 peptides in human cerebrospinal




fluid and blood by a newly developed sensitiveWestern blot assay. J Biol Chem 1996;271:22908-14.

34. Tamaoka A, Sawamura N, Fukushima T, et al.Amyloid beta protein 42(43) in cerebrospinal fluid of patients with Alzheimer’s disease. J Neurol Sci1997;148:41-5.

35. Mehta PD, Pirttila T, Mehta SP, et al. Plasma andcerebrospinal fluid levels of amyloid beta proteins1-40 and 1-42 in Alzheimer’s disease. Arch Neurol2000;57:100-5.

36. Shoji M. Cerebrospinal fluid Abeta40 and Abeta42:natural course and clinical usefulness. Front Biosci2002;7:d997-1006.

37. Nakamura T, Shoji M, Harigaya Y, et al. Amyloidbeta protein levels in cerebrospinal fluid are elevatedin early-onset Alzheimer’s disease. Ann Neurol1994;36:903-11.

38. Motter R, Vigo-Pelfrey C, Kholodenko D, et al.Reduction of beta-amyloid peptide42 in the cere-brospinal fluid of patients with Alzheimer’s disease.Ann Neurol 1995;38:643-8.

39. Shoji M, Matsubara E, Kanai M, et al. Combinationassay of CSF tau, A beta 1-40 and A beta 1-42(43)as a biochemical marker of Alzheimer’s disease.J Neurol Sci 1998;158:134-40.

40. Samuels SC, Silverman JM, Marin DB, et al. CSFbeta-amyloid, cognition, and APOE genotype inAlzheimer’s disease. Neurology 1999;52:547-51.

41. Andreasen N, Hesse C, Davidsson P, et al.

Cerebrospinal fluid beta-amyloid(1-42) in Alzheimer’sdisease: differences between early- and late-onsetAlzheimer’s disease and stability during the courseof disease. Arch Neurol 1999;56:673-80.

42. Tapiola T, Pirttila T, Mehta PD, et al. Relationshipbetween apoE genotype and CSF beta-amyloid(1-42) and tau in patients with probable and defi-nite Alzheimer’s disease. Neurobiol Aging2000;21:735-40.

43. Tapiola T, Pirttila T, Mikkonen M, et al. Three-yearfollow-up of cerebrospinal fluid tau, beta-amyloid 42

and 40 concentrations in Alzheimer’s disease.Neurosci Lett 2000;280:119-22.44. Kanai M, Matsubara E, Isoe K, et al. Longitudinal

study of cerebrospinal fluid levels of tau, A beta1-40,and A beta1-42(43) in Alzheimer’s disease: a studyin Japan. Ann Neurol 1998;44:17-26.

45. Mayeux R, Tang MX, Jacobs DM, et al. Plasma amy-loid beta-peptide 1-42 and incipient Alzheimer’s dis-ease. Ann Neurol 1999;46:412-6.

46. Hoglund K, Wiklund O, Vanderstichele H, et al.Plasma levels of beta-amyloid(1-40), beta-amy-loid(1-42), and total beta-amyloid remain unaffectedin adult patients with hypercholesterolemia aftertreatment with statins. Arch Neurol 2004;61:333-7.

47. Hampel H, Mitchell A, Blennow K, et al. Core bio-logical marker candidates of Alzheimer’s disease—perspectives for diagnosis, prediction of outcome andreflection of biological activity. J Neural Transm2004;111:247-72.

48. Tamaoka A, Fukushima T, Sawamura N, et al.

Amyloid beta protein in plasma from patients withsporadic Alzheimer’s disease. J Neurol Sci1996;141:65-8.

49. Kuo YM, Emmerling MR, Lampert HC, et al. Highlevels of circulating Abeta42 are sequestered byplasma proteins in Alzheimer’s disease. BiochemBiophys Res Commun 1999;257:787-91.

50. Andreasen N, Minthon L, Vanmechelen E, et al.Cerebrospinal fluid tau and Abeta42 as predictors of development of Alzheimer’s disease in patients withmild cognitive impairment. Neurosci Lett1999;273:5-8.

51. Shoji M, Kanai M, Matsubara E, et al. Taps toAlzheimer’s patients: a continuous Japanese study of cerebrospinal fluid biomarkers. Ann Neurol2000;48:402.

52. Otto M, Esselmann H, Schulz-Shaeffer W, et al.Decreased beta-amyloid1-42 in cerebrospinal fluidof patients with Creutzfeldt-Jakob disease.Neurology 2000;54:1099-102.

53. Kanemaru K, Kameda N, Yamanouchi H. DecreasedCSF amyloid beta42 and normal tau levels indementia with Lewy bodies. Neurology

2000;54:1875-6.54. Li QX, Fuller SJ, Beyreuther K, et al. The amyloid

precursor protein of Alzheimer’s disease in humanbrain and blood. J Leukoc Biol 1999;66:567-74.

55. Wild-Bode C, Yamazaki T, Capell A, et al.Intracellular generation and accumulation of amyloidbeta-peptide terminating at amino acid 42. J BiolChem 1997;272:16085-8.

56. Hartmann T, Bieger SC, Bruhl B, et al. Distinctsites of intracellular production for Alzheimer’sdisease A beta40/42 amyloid peptides. Nat Med

1997;3:1016-20.57. Hershkowitz M, Adunsky A. Binding of platelet-activating factor to platelets of Alzheimer’s diseaseand multiinfarct dementia patients. Neurobiol Aging1996;17:865-8.

58. Borroni B, Colciaghi F, Caltagirone C, et al. Plateletamyloid precursor protein abnormalities in mild cog-nitive impairment predict conversion to dementia of Alzheimer’s type: a 2-year follow-up study. ArchNeurol 2003;60:1740-4.

59. Zlokovic BV. Cerebrovascular transport of Alzheimer’s amyloid beta and apolipoproteins J andE: possible anti-amyloidogenic role of the blood-brain barrier. Life Sci 1996;59:1483-97.




60. Mackic J, Ghiso J, Frangione B, et al. Differentialcerebrovascular sequestration and enhanced blood-brain barrier permeability to circulating Alzheimer’samyloid- β peptide in aged Rhesus vs. aged Squirrelmonkey. Vascular Pharmacol 2002;18:303-13.

61. Rhodin JA, Thomas TN, Clark L, et al. In vivo cere-

brovascular actions of amyloid beta-peptides and theprotective effect of conjugated estrogens. JAlzheimers Dis 2003;5:275-86.

62. Kawarabayashi T, Younkin LH, Saido TC, et al. Age-dependent changes in brain, CSF, and plasma amy-loid (beta) protein in the Tg2576 transgenicmouse model of Alzheimer’s disease. J Neurosci2001;21:372-81.

63. DeMattos RB, Bales KR, Cummins DJ, et al. Brainto plasma amyloid-beta efflux: a measure of brainamyloid burden in a mouse model of Alzheimer’sdisease. Science 2002;295:2264-7.

64. Hardy JA, Higgins GA. Alzheimer’s disease: the amy-loid cascade hypothesis. Science 1992;256:184-5.

65. Naslund J, Schierhorn A, Hellman U, et al. Relativeabundance of Alzheimer A beta amyloid peptidevariants in Alzheimer’s disease and normal aging.Proc Natl Acad Sci U S A 1994;91:8378-82.

66. Tamaoka A, Kondo T, Odaka A, et al. Biochemicalevidence for the long-tail form (A beta 1-42/43) of amyloid beta protein as a seed molecule in cerebraldeposits of Alzheimer’s disease. Biochem BiophysRes Commun 1994;205:834-42.

67. Gravina SA, Ho L, Eckman CB, et al. Amyloid betaprotein (A beta) in Alzheimer’s disease brain.Biochemical and immunocytochemical analysis withantibodies specific for forms ending at A beta 40 orA beta 42(43). J Biol Chem 1995;270:7013-6.

68. Kuo YM, Emmerling MR, Vigo-Pelfrey C, et al.Water-soluble Abeta (N-40, N-42) oligomers in nor-mal and Alzheimer’s disease brains. J Biol Chem1996;271:4077-81.

69. Shinkai Y, Yoshimura M, Morishima-Kawashima M,et al. Amyloid beta-protein deposition in the lep-

tomeninges and cerebral cortex. Ann Neurol1997;42:899-908.70. Tamaoka A, Fraser PE, Ishii K, et al. Amyloid-beta-

protein isoforms in brain of subjects with PS1-linked, beta APP-linked and sporadic Alzheimer’sdisease. Brain Res Mol Brain Res 1998;56:178-85.

71. Kuo YM, Emmerling MR, Bisgaier CL, et al.Elevated low-density lipoprotein in Alzheimer’s dis-ease correlates with brain abeta 1-42 levels. BiochemBiophys Res Commun 1998;252:711-5.

72. Lue LF, Kuo YM, Roher AE, et al. Soluble amyloidbeta peptide concentration as a predictor of synapticchange in Alzheimer’s disease. Am J Pathol1999;155:853-62.

73. Beffert U, Cohn JS, Petit-Turcotte C, et al.Apolipoprotein E and beta-amyloid levels in thehippocampus and frontal cortex of Alzheimer’sdisease subjects are disease-related andapolipoprotein E genotype dependent. Brain Res1999;843:87-94.

74. Wang J, Dickson DW, Trojanowski JQ, et al. The lev-els of soluble versus insoluble brain Abeta distin-guish Alzheimer’s disease from normal andpathologic aging. Exp Neurol 1999;158:328-37.

75. Naslund J, Haroutunian V, Mohs R, et al. Correlationbetween elevated levels of amyloid beta-peptidein the brain and cognitive decline. JAMA2000;283:1571-7.

76. Morishima-Kawashima M, Oshima N, Ogata H,et al. Effect of apolipoprotein E allele epsilon4 on theinitial phase of amyloid beta-protein accumulation inthe human brain. Am J Pathol 2000;157:2093-9.

77. Miklossy J, Taddei K, Suva D, et al. Two novel pre-senilin-1 mutations (Y256S and Q222H) are associ-ated with early-onset Alzheimer’s disease. NeurobiolAging 2003;24:655-62.

78. Ingelsson M, Fukumoto H, Newell KL, et al. EarlyAbeta accumulation and progressive synaptic loss,gliosis, and tangle formation in AD brain. Neurology2004;62:925-31.

79. Li R, Lindholm K, Yang LB, et al. Amyloid beta pep-tide load is correlated with increased beta-secretaseactivity in sporadic Alzheimer’s disease patients.

Proc Natl Acad Sci U S A 2004;101:3632-7.80. Turner RS, Suzuki N, Chyung AS, et al. Amyloids

beta40 and beta42 are generated intracellularly incultured human neurons and their secretion increaseswith maturation. J Biol Chem 1996;271:8966-70.

81. Klein WL, Krafft GA, Finch CE. Targeting smallAbeta oligomers: the solution to an Alzheimer’sdisease conundrum? Trends Neurosci 2001;24:219-24.

82. Walsh DM, Lomakin A, Benedek GB, et al. Amyloidbeta-protein fibrillogenesis. Detection of a protofib-

rillar intermediate. J Biol Chem 1997;272:22364-72.83. Hartley DM, Walsh DM, Ye CP, et al. Protofibrillarintermediates of amyloid beta-protein induce acuteelectrophysiological changes and progressiveneurotoxicity in cortical neurons. J Neurosci1999;19:8876-84.

84. Nilsberth C, Westlind-Danielsson A, Eckman CB,et al. The ‘Arctic’ APP mutation [E693G] causesAlzheimer’s disease by enhanced Abeta protofibrilformation. Nat Neurosci 2001;4:887-93.

85. Walsh DM, Klyubin I, Fadeeva JV, et al. Naturallysecreted oligomers of amyloid beta protein potentlyinhibit hippocampal long-term potentiation in vivo.Nature 2002;416:535-9.




conformation on structure and function. J MolNeurosci 2004;23:235-46.

114. Holtzman DM, Bales KR, Wu S, et al. Expressionof human apolipoprotein E reduces amyloid-betadeposition in a mouse model of Alzheimer’s dis-ease. J Clin Invest 1999;103:R15-R21.

115. Verdier Y, Zarandi M, Penke B. Amyloid beta-pep-tide interactions with neuronal and glial cell plasmamembrane: binding sites and implications forAlzheimer’s disease. J Pept Sci 2004;10:229-48.

116. Kamenetz F, Tomita T, Hsieh H, et al. APP pro-cessing and synaptic function. Neuron2003;37:925-37.

117. Plant LD, Boyle JP, Smith IF, et al. The productionof amyloid [beta] peptide is a critical requirementfor the viability of central neurons. J Neurosci2003;23:5531-5.

118. De Ferrari GV, Inestrosa NC. Wnt signaling func-tion in Alzheimer’s disease. Brain Res Brain ResRev 2000;33:1-12.

119. Lopez-Toledano MA, Shelanski ML. Neurogeniceffect of {beta}-amyloid peptide in the developmentof neural stem cells. J Neurosci 2004;24:5439-44.

120. Haughey NJ, Liu D, Nath A, et al. Disruption of neurogenesis in the subventricular zone of adultmice, and in human cortical neuronal precursorcells in culture, by amyloid beta-peptide: implica-tions for the pathogenesis of Alzheimer’s disease.Neuromolecular Med 2002;1:125-35.

121. Haughey NJ, Nath A, Chan SL, et al. Disruption of neurogenesis by amyloid beta-peptide, and per-turbed neural progenitor cell homeostasis, inmodels of Alzheimer’s disease. J Neurochem2002;83:1509-24.

122. Ramsden M, Plant LD, Webster NJ, et al.Differential effects of unaggregated and aggregatedamyloid beta protein (1-40) on K( +) channelcurrents in primary cultures of rat cerebellar gran-ule and cortical neurones. J Neurochem 2001;79:699-712.

123. Crawford F, Suo Z, Fang C, et al. Characteristics of the in vitro vasoactivity of beta-amyloid peptides.Exp Neurol 1998;150:159-68.

124. Zou K, Gong J-S, Yanagisawa K, et al. A novelfunction of monomeric amyloid beta -proteinserving as an antioxidant molecule against metal-induced oxidative damage. J Neurosci 2002;22:4833-41.

125. Misonou H, Morishima-Kawashima M, Ihara Y.Oxidative stress induces intracellular accumulationof amyloid beta-protein (Abeta) in human neurob-lastoma cells. Biochemistry 2000;39:6951-9.

126. Zhang L, Zhao B, Yew DT, et al. Processing of Alzheimer’s amyloid precursor protein during

H2O2-induced apoptosis in human neuronal cells.Biochem Biophys Res Commun 1997;235:845-8.

127. Paola D, Domenicotti C, Nitti M, et al. Oxidativestress induces increase in intracellular amyloidbeta-protein production and selective activation of betaI and betaII PKCs in NT2 cells. Biochem

Biophys Res Commun 2000;268:642-6.128. Cuajungco MP, Goldstein LE, Nunomura A, et al.

Evidence that the beta-amyloid plaques of Alzheimer’s disease represent the redox-silencingand entombment of abeta by zinc. J Biol Chem2000;275:19439-42.

129. Maynard CJ, Bush AI, Masters CL, et al. Metalsand amyloid-beta in Alzheimer’s disease. Int J ExpPathol 2005;86:147-59.

130. Pike CJ, Burdick D, Walencewicz AJ, et al.Neurodegeneration induced by beta-amyloid pep-tides in vitro: the role of peptide assembly state.J Neurosci 1993;13:1676-87.

131. Yankner BA, Duffy LK, Kirschner DA.Neurotrophic and neurotoxic effects of amyloidbeta protein: reversal by tachykinin neuropeptides.Science 1990;250:279-82.

132. Roher AE, Ball MJ, Bhave SV, et al. Beta-amyloidfrom Alzheimer’s disease brains inhibits sproutingand survival of sympathetic neurons. BiochemBiophys Res Commun 1991;174:572-9.

133. Pike CJ, Walencewicz AJ, Glabe CG, et al. Invitro aging of beta-amyloid protein causes peptide

aggregation and neurotoxicity. Brain Res1991;563:311-4.

134. Mattson MP, Tomaselli KJ, Rydel RE. Calcium-destabilizing and neurodegenerative effects of aggregated beta-amyloid peptide are attenuated bybasic FGF. Brain Res 1993;621:35-49.

135. Mattson MP, Rydel RE. beta-Amyloid precursorprotein and Alzheimer’s disease: the peptide plotthickens. Neurobiol Aging 1992;13:617-21.

136. Lorenzo A, Yankner BA. Beta-amyloid neurotoxic-ity requires fibril formation and is inhibited by

congo red. Proc Natl Acad Sci U S A 1994;91:12243-7.137. Busciglio J, Lorenzo A, Yankner BA.

Methodological variables in the assessment of betaamyloid neurotoxicity. Neurobiol Aging1992;13:609-12.

138. Gong Y, Chang L, Viola KL, et al. Alzheimer’s dis-ease-affected brain: presence of oligomeric A betaligands (ADDLs) suggests a molecular basis forreversible memory loss. Proc Natl Acad Sci U S A2003;100:10417-22.

139. Lacor PN, Buniel MC, Chang L, et al. Synaptic tar-geting by Alzheimer’s-related amyloid betaoligomers. J Neurosci 2004;24:10191-200.




uptake of amyloid-beta(1-42) peptide in microglia.Biochem Biophys Res Commun 2001;281:6-10.

168. Banks WA, Ronbinson SM, Verma S, et al. Effluxof human and mouse amyloid β proteins 1-40 and1-42 from brain: impairment in a mousemodel of Alzheimer’s disease. Neuroscience

2003;121:487-92.169. Iwatsubo T, Mann DM, Odaka A, et al. Amyloid

beta protein (A beta) deposition: a beta 42(43) pre-cedes A beta 40 in Down syndrome. Ann Neurol1995;37:294-9.

170. Calero M, Rostagno A, Matsubara E, et al.Apolipoprotein J (clusterin) and Alzheimer’s dis-ease. Microsc Res Tech 2000;50:305-15.

171. Pollack SJ, Lewis H. Secretase inhibitors forAlzheimer’s disease: challenges of a promiscuousprotease. Curr Opin Investig Drugs 2005;6:35-47.

172. Doraiswamy PM, Finefrock AE. Metals in ourminds: therapeutic implications for neurodegenera-tive disorders. Lancet Neurol 2004;3:431-4.

173. Lahiri DK, Farlow MR, Sambamurti K, et al. A crit-ical analysis of new molecular targets and strategiesfor drug developments in Alzheimer’s disease. CurrDrug Targets 2003;4:97-112.

174. Pietrzik C, Behl C. Concepts for the treatment of Alzheimer’s disease: molecular mechanismsand clinical application. Int J Exp Pathol2005;86:173-85.

175. Cumming JN, Iserloh U, Kennedy ME. Design and

development of BACE-1 inhibitors. Curr Opin DrugDiscov Devel 2004;7:536-56.

176. Kobayashi DT, Chen KS. Behavioral phenotypes of amyloid-based genetically modified mouse modelsof Alzheimer’s disease. Genes Brain Behav2005;4:173-96.

177. Anderson JJ, Holtz G, Baskin PP, et al. Reductions inbeta-amyloid concentrations in vivo by the gamma-secretase inhibitors BMS-289948 and BMS-299897.Biochem Pharmacol 2005;69:689-98.

178. Barten DM, Guss VL, Corsa JA, et al. Dynamics of

β-amyloid reductions in brain, cerebrospinal fluid,and plasma of β-amyloid precursor protein trans-genic mice treated with a γ -secretase inhibitor.J Pharmacol Exp Ther 2005;312:635-43.

179. Siemers E, Skinner M, Dean RA, et al. Safety, tol-erability, and changes in amyloid beta concentra-tions after administration of a gamma-secretaseinhibitor in volunteers. Clin Neuropharmacol2005;28:126-32.

180. Cherny RA, Legg JT, McLean CA, et al. Aqueousdissolution of Alzheimer’s disease Abeta amyloiddeposits by biometal depletion. J Biol Chem1999;274:23223-8.

181. Cherny RA, Atwood CS, Xilinas ME, et al.Treatment with a copper-zinc chelator markedlyand rapidly inhibits beta-amyloid accumulation inAlzheimer’s disease transgenic mice. Neuron2001;30:665-76.

182. Ritchie CW, Bush AI, Mackinnon A, et al. Metal-

protein attenuation with iodochlorhydroxyquin(clioquinol) targeting Abeta amyloid deposition andtoxicity in Alzheimer’s disease: a pilot phase 2 clin-ical trial. Arch Neurol 2003;60:1685-91.

183. DeMattos RB, Bales KR, Cummins DJ, et al.Peripheral anti-A beta antibody alters CNS andplasma A beta clearance and decreases brain A betaburden in a mouse model of Alzheimer’s disease.Proc Natl Acad Sci U S A 2001;98:8850-5.

184. Schenk D, Barbour R, Dunn W, et al. Immunizationwith amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature1999;400:173-7.

185. Gotz J, Streffer JR, David D, et al. Transgenic ani-mal models of Alzheimer’s disease and related dis-orders: histopathology, behavior and therapy. MolPsychiatry 2004;9:664-83.

186. Weiner HL, Lemere CA, Maron R, et al.Nasal administration of amyloid-beta peptidedecreases cerebral amyloid burden in a mousemodel of Alzheimer’s disease. Ann Neurol2000;48:567-79.

187. Sigurdsson EM, Scholtzova H, Mehta PD, et al.

Immunization with a nontoxic/nonfibrillar amyloid-beta homologous peptide reduces Alzheimer’s dis-ease-associated pathology in transgenic mice. AmJ Pathol 2001;159:439-47.

188. Bard F, Cannon C, Barbour R, et al. Peripherallyadministered antibodies against amyloid beta-pep-tide enter the central nervous system and reducepathology in a mouse model of Alzheimer’s dis-ease. Nat Med 2000;6:916-9.

189. Orgogozo JM, Gilman S, Dartigues JF, et al.Subacute meningoencephalitis in a subset of

patients with AD after Abeta42 immunization.Neurology 2003;61:46-54.190. Das P, Golde TE. Open peer commentary regarding

Abeta immunization and CNS inflammation byPasinetti et al. Neurobiol Aging 2002;23:671-4; dis-cussion 83-4.

191. Sigurdsson E, Wisniewski, T., Frangione, B.Infectivity of amyloid diseases. Trends Mol Med2002.

192. Nicoll JA, Wilkinson D, Holmes C, et al.Neuropathology of human Alzheimer’s diseaseafter immunization with amyloid-beta peptide: acase report. Nat Med 2003.




193. Hock C, Konietzko U, Streffer JR, et al. Antibodiesagainst beta-amyloid slow cognitive decline inAlzheimer’s disease. Neuron 2003;38:547-54.

194. Chauhan NB, Siegel GJ. Efficacy of anti-Abetaantibody isotypes used for intracerebroventricularimmunization in TgCRND8. Neurosci Lett

2005;375:143-7.195. Li SB, Wang HQ, Lin X, et al. Specific humoral

immune responses in rhesus monkeys vaccinatedwith the Alzheimer’s disease-associated beta-amy-loid 1-15 peptide vaccine. Chin Med J (Engl)2005;118:660-4.

196. Etminan M, Gill S, Samii A. Effect of non-steroidalanti-inflammatory drugs on risk of Alzheimer’s dis-ease: systematic review and meta-analysis of observational studies. BMJ 2003;327:128.

197. Szekely CA, Thorne JE, Zandi PP, et al.Nonsteroidal anti-inflammatory drugs for the pre-vention of Alzheimer’s disease: a systematic review.Neuroepidemiology 2004;23:159-69.

198. Weggen S, Eriksen JL, Das P, et al. A subset of NSAIDs lower amyloidogenic Abeta42 inde-pendently of cyclooxygenase activity. Nature2001;414:212-6.

199. Weggen S, Eriksen JL, Sagi SA, et al. Evidence thatnonsteroidal anti-inflammatory drugs decreaseamyloid beta 42 production by direct modulation of gamma-secretase activity. J Biol Chem2003;278:31831-7.

200. Aisen PS, Schafer KA, Grundman M, et al. Effectsof rofecoxib or naproxen vs placebo on Alzheimer’sdisease progression: a randomized controlled trial.JAMA 2003;289:2819-26.

201. Cole GM, Morihara T, Lim GP, et al. NSAID andantioxidant prevention of Alzheimer’s disease: les-sons from in vitro and animal models. Ann N YAcad Sci 2004;1035:68-84.

202. Lehmann JM, Kliewer SA, Moore LB, et al.Activation of the nuclear receptor LXR by oxys-terols defines a new hormone response pathway.

J Biol Chem 1997;272:3137-40.203. Jaradat MS, Wongsud B, Phornchirasilp S, et al.Activation of peroxisome proliferator-activated recep-tor isoforms and inhibition of prostaglandin H(2) syn-

thases by ibuprofen, naproxen, and indomethacin.Biochem Pharmacol 2001;62:1587-95.

204. Jiang C, Ting AT, Seed B. PPAR-gamma agonistsinhibit production of monocyte inflammatorycytokines. Nature 1998;391:82-6.

205. Ricote M, Li AC, Willson TM, et al. The peroxi-

some proliferator-activated receptor-gamma is anegative regulator of macrophage activation. Nature1998;391:79-82.

206. Heneka MT, Sastre M, Dumitrescu-Ozimek L, et al.Acute treatment with the PPARgamma agonistpioglitazone and ibuprofen reduces glial inflamma-tion and Abeta1-42 levels in APPV717I transgenicmice. Brain 2005;128:1442-53.

207. Wolozin B. Cholesterol and Alzheimer’s disease.Biochem Soc Trans 2002;30:525-9.

208. Refolo LM, Malester B, LaFrancois J, et al.Hypercholesterolemia accelerates the Alzheimer’samyloid pathology in a transgenic mouse model.Neurobiol Dis 2000;7:321-31.

209. Refolo LM, Pappolla MA, LaFrancois J, et al. Acholesterol-lowering drug reduces beta-amyloidpathology in a transgenic mouse model of Alzheimer’s disease. Neurobiol Dis 2001;8:890-9.

210. Sidera C, Parsons R, Austen B. The regulation of beta-secretase by cholesterol and statins inAlzheimer’s disease. J Neurol Sci 2005;229-230:269-73.

211. Stuve O, Youssef S, Steinman L, et al. Statins as

potential therapeutic agents in neuroinflammatorydisorders. Curr Opin Neurol 2003;16:393-401.

212. Sjogren M, Gustafsson K, Syversen S, et al.Treatment with simvastatin in patients withAlzheimer’s disease lowers both alpha- and beta-cleaved amyloid precursor protein. Dement GeriatrCogn Disord 2003;16:25-30.

213. Sparks DL, Sabbagh MN, Connor DJ, et al.Atorvastatin therapy lowers circulating cholesterolbut not free radical activity in advance of identifi-able clinical benefit in the treatment of mild-to-

moderate AD. Curr Alzheimer Res 2005;2:343-53.214. Rea TD, Breitner JC, Psaty BM, et al. Statin use andthe risk of incident dementia: the CardiovascularHealth Study. Arch Neurol 2005;62:1047-51.




12Impact of β-Amyloid on the Tau

Pathology in Tau Transgenic Mouse andTissue Culture ModelsJürgen Götz, Della C. David, and Lars M. Ittner

198

12.1 Introduction

Dementia is a generic term that describes chronicor progressive dysfunction of cortical and subcorti-cal functions that result in complex cognitivedecline. These cognitive changes are commonlyaccompanied by disturbances of mood, behavior,and personality. In developed countries with anincreasingly aging population, the prevalence of dementia is currently at around 1.5% at 65 years of age, which doubles every 4 years and reaches about

30% at the age of 80 [1].Of all age-related neurodegenerative disorders,Alzheimer’s disease (AD) is the most prevalent. Itis characterized histopathologically by β-amyloid(Aβ)-containing plaques, tau-containing neurofib-rillary tangles (NFTs), reduced synaptic densityand neuronal loss in selected brain areas [2]. Infamilial forms of AD (FAD), pathogenic mutationshave been identified in both the gene encoding theprecursor of the A β peptide, APP, itself and in thepresenilin genes, which encode part of the proteasecomplex involved in processing APP. This geneticevidence supports the amyloid cascade hypothesis,which claims that A β causes or enhances the NFTpathology.

Frontotemporal dementia (FTD) is the preferredterm for a spectrum of non-Alzheimer dementiascharacterized by focal atrophy of frontal and ante-rior temporal regions and NFTs in the absence of Aβ deposition. Recent epidemiological studiessuggest that FTD is the second most common

cause of dementia in persons younger than 65 years[3]. In familial forms of FTD (frontotemporaldementia with parkinsonism linked to chromosome

17; FTDP-17), pathogenic mutations have beenidentified in tau proving that tau dysfunction initself can lead to neurodegeneration and dementia.

AD and FTD have a distinct neuropathologicalprofile, but histopathological studies have shownthat mixed states (with people presenting with fea-tures of more than one type of dementia) are prob-ably more frequent than pure dementia syndromes[1, 4, 5]. Here, we discuss how aspects of thehuman pathology have been modeled in animals,with a special emphasis on tau transgenic mice.

Furthermore, we present experimental evidenceobtained in tau transgenic mouse and tissue-culturemodels that to some extent support the amyloidcascade hypothesis in mice.

12.2 Alzheimer’s Disease

The clinical presentation of AD is dominated byearly memory deficits, followed by gradual erosionof other cognitive functions such as judgment, ver-bal fluency, or orientation. Although this sequentialorder may vary, memory impairment is normallythe first and dominating feature.

In addition to a reduced synaptic density andneuronal loss in selected brain areas, AD is char-acterized by two forms of insoluble protein aggre-gates, the extracellular A β-containing plaques andthe intracellular NFTs. The major component of the plaques is a 40–42 amino acid aggregatedpolypeptide termed β-amyloid (A β; A β40 and

Aβ42), which is derived by proteolysis from thelarger amyloid precursor protein, APP (Fig. 12.1)[6, 7]. APP can be proteolytically cleaved by the



membrane-associated α-secretase, which cleaves

APP within the A β domain. This pathway is non-amyloidogenic, as this cleavage precludes the for-mation of A β. Alternatively, cleavage may occur

in the endosomal-lysosomal pathway, first byβ-secretase and then by γ -secretase, whichtogether generate the A β peptide. β-Secretaseactivity has been attributed to a single protein,BACE, whereas γ -secretase activity was shown todepend on the presence of a total of four compo-nents: presenilin, nicastrin, APH-1 and PEN-2 [8,9] (Fig. 12.1).

The second histopathological hallmark of ADare the neurofibrillary lesions that are found incell bodies and apical dendrites as NFTs, in dis-tal dendrites as neuropil threads, and in theabnormal neurites that are associated with someAβ plaques (neuritic plaques). NFTs develop inspecific sites and spread in a predictable, nonran-dom manner across the brain. This sequence of

the tau pathology is subjected to little inter-indi-vidual variation and provides a basis for distin-guishing six stages in the progression of thedisease [10, 11].

The major component of NFTs are abnormal fil-aments [12, 13]. The core protein of these fila-ments is tau, a microtubule-associated protein[14]. In the course of the disease, tau becomesabnormally phosphorylated, it adopts an alteredconformation and is relocalized from axonal tosomatodendritic compartments. Phosphorylationtends to dissociate tau from microtubules. Becausethis increases the soluble pool of tau, it might bean important first step in the assembly of tau fila-ments [5, 15–21]. Tau filaments have a clearβ-cross structure, which is the defining feature of amyloid fibers [22]. They share this structure withthe extracellular deposits present in the systemicand organ-specific amyloid diseases. It is thereforeappropriate to consider the diseases with filamen-tous tau aggregates, the so-called tauopathies, a

form of brain amyloidosis [23].Physiological functions of tau include the assem-

bly and stabilization of microtubules. Microtubulesare hollow, 25-nm-wide cylindrical polymers, assem-bled primarily from heterodimers of α - and β-tubulinand a collection of microtubule-associated proteins(MAPs). Microtubules have two general functions, asthe primary structural component of the mitotic spin-dle and in organizing the cytoplasm. Microtubulesisolated from cell extracts by multiple cycles of assembly/disassembly and differential centrifugationyield a final microtubule preparation of which about80% is tubulin, while the remaining 20% are MAPs.

12. Impact of β-Amyloid on Tau Pathology 199

APP mut

β-secretase

breeding approach

stereotaxic approach

5-7 × NFTinduction

S422-PT212-PS214-P

P301L

Tau mut

syntheticβ-amyloid

fibrilsβ-amyloidplaques

α -secretase

Aβ42Peptide

α -secretase

FIGURE 12.1. Cleavage of the amyloid precursor protein(APP) by the membrane-associated α -secretase is withinthe A β domain and thus precludes the formation of Aβ. Therefore, this pathway is non-amyloidogenic. Alter-natively, cleavage may occur in the endosomal-lysosomalpathway, first by β-secretase and then by γ -secretase gener-ating the A β peptide. A β is deposited around meningeal andcerebral vessels and in the gray matter as β-amyloidplaques. To determine the relationship between A β and theNFT/tau pathology in AD, two alternative approaches werepursued. One involved the intercrossing of APP and taumutant mice with a plaque and NFT pathology (“breedingapproach”), the other the stereotaxic injection of fibrillarpreparations of A β42 into mutant tau transgenic brains(“stereotaxic approach”). These approaches resulted in five-

to sevenfold increased NFT formation, which was associ-ated with phosphorylation of tau at the phospho-epitopesThr212/Ser214 and Ser422. Together, these studies provideevidence for the amyloid cascade hypothesis in mice. Thefinding that A β42 was not capable of inducing NFT forma-tion in non-NFT-forming wild-type tau transgenic micemay reflect species differences between mice and men.Alternatively, it may imply that, at least in mice, A β42 can-not induce NFT formation de novo .



Initially isolated from mammalian neurons, MAPswere named according to the three major size classesof polypeptides: MAP1 (>250 kDa), MAP2 (~200kDa), and tau protein (50–70 kDa). MAP2 and tauare expressed together in most neurons, where theylocalize to separate subcellular compartments.MAP2 is largely found in dendrites, whereas tau isconcentrated in axons. Tau has also been found inastrocytes and oligodendrocytes, although, underphysiological conditions, levels are relatively low[24]. Additional roles have been assigned to tau insignal transduction, the organization of the actincytoskeleton, intracellular vesicle transport, andanchoring of phosphatases and kinases [25–34]. Inthe adult human brain, six tau isoforms are producedby alternative mRNA splicing of exons 2, 3, and 10

(Fig. 12.2). They differ by the presence or absence of one or two short inserts in the amino-terminal half and have either three or four microtubule-bindingrepeat motifs in the carboxy-terminal half (3R and4R). All six brain tau isoforms are found in the neu-rofibrillary lesions of AD brains [35].

In early-onset familial forms of AD (FAD),mutations were identified in three genes: in theAPP gene itself and in the genes encoding prese-nilin 1 and 2 [36, 37]. Expression of FAD mutantforms of APP in transgenic mice by severalresearch groups caused A β-plaque formation andconcomitant memory deficits that progressed withage (reviewed in Ref. 5). These were more pro-nounced in transgenic mice coexpressing mutantforms of presenilin and APP, yet, NFT formationcould not be reproduced [5].

For late-onset sporadic AD (SAD), aroundtwo dozen risk-conferring genes have been iden-tified until today, but of these only theapolipoprotein E (APOE) gene has been con-

firmed unanimously and found to be associatedwith SAD [38]. When FAD is compared withSAD, the histopathological hallmarks are indis-tinguishable. This implies that lessons learnedfrom the familial forms of AD may be applicablealso to the sporadic forms.

12.3 Frontotemporal Dementias

Although AD is the most frequent form of demen-tia at high age, NFTs are, in the absence of β-amy-loid plaques, also abundant in additional

neurodegenerative diseases. The preferred term forthis spectrum of non-Alzheimer dementias is“frontotemporal dementia” (FTD) [39]. FTD ischaracterized by focal atrophy of frontal and ante-rior temporal regions. Three broad subdivisionshave been recognized, depending on the profile of immunohistochemical staining and the pattern of intracellular inclusions [39–42]: one with tau-posi-tive aggregates (Pick disease [PiD], progressive

200 J. Götz et al.

E2

1 R5H, R5L

K257TI260VL266VG272VN279K∆K280L284L∆N296, N296H, N296NP301L, P301SS305S, S305NL315RS320FQ336RV337ME342VS352LK369IG389RR406W

E3

E10

441

intronicmutations

FIGURE 12.2. By alternative mRNA splicing of exons E2,

E3, and E10, six tau isoforms are produced in the adulthuman brain. They differ by the presence or absence of one or two short inserts in the amino-terminal half (0N,1N, and 2N, respectively) and have either three or fourmicrotubule-binding repeat motifs in the carboxy-termi-nal half (3R and 4R). The microtubule-binding motifsare indicated in black. All six brain tau isoforms arefound in the neurofibrillary lesions of AD patients. InFTDP-17, the majority of the exonic mutations in tau areclustered around the microtubule binding domain,whereas the intronic mutations (indicated by the stemloop) result in a shift of 3R to 4R tau isoforms.



tant to know what triggers their formation and howthey are functionally related. Some insight may begained by the analysis of adult lifestyle risk factorscombined with the evidence of a genetic predispo-sition (as determined by the inheritance of risk alle-les of susceptibility genes), which together maycause SAD [96]. Although the etiology of FAD andSAD differ, the clinical picture and the morpholog-ical end stage in the brain appear to be the same.

12.5 Tau Transgenic Mice:Requirements for and Role ofNFT Formation

To better understand the role of β-amyloid plaquesand NFTs in AD and related disorders, experimen-tal animal models have been developed thatreproduce aspects of the neuropathological charac-teristics of these diseases (reviewed in Ref. 5).Their suitability largely depends on the purpose amodel has to suit. If one wants to modelhistopathological features, one has to discriminatebetween the precise anatomical “reproduction” of the pathology and modeling at the cellular level.This is important when the animals (in particulartransgenic mice) are employed in behavioral stud-ies intended to correlate the histopathology withdementia. These animal models may either offer ageneral proof of principle or reproduce more spe-cific aspects of the human disease. Animal modelsmay be used to identify disease modifiers, compo-nents of pathocascades, and susceptibility genes[97]. Furthermore, they may be employed in drugscreenings [5]. Finally, insight gained from thesemodels can be translated to human disease

and assist in the development of treatment thera-pies [99].

After the very first APP transgenic animals hadfailed to show an extensive AD-like neuropathol-ogy, in 1995 Games and co-workers successfullyexpressed high levels of the disease-linkedV717F mutant form of APP, under control of theplatelet-derived growth factor (PDGF) mini-pro-moter. These PDAPP mice showed many of thepathological features of AD, including extensivedeposition of extracellular amyloid plaques, astro-cytosis, and neuritic dystrophy [100]. Similar fea-tures were observed in a second transgenic model

by Hsiao and co-workers that expressed the APP sw

mutation inserted into a hamster prion protein(PrP) cosmid vector [101]. Then, by expressingthe Swedish double APP mutation under control of the mThy1.2 promoter, a research group atNovartis established the APP23 mouse model witha sevenfold overexpression of APP [102, 103].Subsequently, many more models have beendeveloped by both academic and industrialresearch groups (such as the TgCRND8 [104] orJ20 mice [105]). Using these mice, aspects of A βtoxicity have been addressed and therapies havebeen tested. The APP transgenic mice were alsocrossed with presenilin, BACE, ApoE, and TGF-β1 transgenic and/or knockout strains (reviewed inRef. 5).

The first tau transgenic models were establishedby us in 1995 (Table 12.1) and expressed thelongest human 4R brain tau isoform (2N4R), with-out a pathogenic mutation, in mice using the hThy1promoter for neuronal expression [106]. Despitethe lack of NFT pathology, these mice modeledaspects of human AD, such as the somatodendriticlocalization of hyperphosphorylated tau and, there-fore, represented an early pre-NFT phenotype. Thesubsequent use of stronger promoters caused amore pronounced phenotype in transgenic mice[107–109] (Table 12.1). In some strains, highexpression levels of the transgene in motor neuronscaused the formation of large numbers of patholog-ically enlarged axons with neurofilament- and tau-immunoreactive spheroids, a neuropathologicalcharacteristic of most cases of amyotrophic lateralsclerosis (ALS), where they are believed to impairslow axonal transport [110–112]. Tau proteinextracted from transgenic brain and spinal cord wasshown to be increasingly insoluble as the mice

became older. Despite the decreased solubility of tau, NFTs did not form with the exception of onestudy where they were reported to be present at lownumbers when the mice had reached a very old age[113]. Taken together, these findings demonstratethat overexpression of human tau can lead to anaxonopathy resulting in nerve cell dysfunction andamyotrophy [5, 20].

When the first pathogenic FTDP-17 mutationshad been identified in the tau gene in 1998, several

groups achieved NFT formation both in neurons[114–118] and in glial cells of transgenic mice[119–122] (Table 12.1).




motor neurons in spinal cord [119]. The clinicalphenotype of these mice was subtle. In contrast,when human wild-type tau was overexpressed inneurons and glial cells using the mouse T α 1 α-tubulin promoter, a glial pathology was foundresembling the astrocytic plaques in CBD and thecoiled bodies in CBD and PSP [120].

To reproduce the plaque and NFT pathology inone single animal model, triple-transgenic micewere developed harboring PS1 M146V, the APP Swe

and P301L tau transgenes. Instead of crossingindependent lines, the APP and tau transgenes weremicroinjected into transgenes embryos derivedfrom homozygous PS1 M146V knock-in mice,generating mice with the same genetic back-ground. In the triple transgenic mice, synaptic dys-

function, including LTP deficits, manifested in anage-related manner, but before plaque and NFTpathology [121].

To allow a better side-by-side comparison of wild-type and P301L mutant mice, a total of threestrains were generated by another research groupand analyzed in parallel [122]. First, they comparedtwo strains, both expressing the longest human tauisoform, one bearing the P301L mutation and onewithout mutations, at similar, moderate levels [122].The two strains developed very different pheno-types. Nonmutant mice became motor-impairedalready around at 6–8 weeks of age, accompanied byaxonopathy, but no tau aggregates, and survivednormally. In contrast, the mutant mice developedNFTs from 6 months of age, without axonal dilata-tions and, despite displaying only minor motor prob-lems, all succumbed before the age of 13 months.The authors concluded that excessive binding of wild-type human tau as opposed to reduced bindingof P301L mutant tau to microtubules may be respon-

sible for the development of axonopathy and tauopa-thy, respectively, in the two strains and that theconformational change of P301L tau is a majordeterminant in triggering the tauopathy. The thirdstrain (a tau knock-in of human wild-type tau-4R/2Naimed to inactivate the endogenous murine tau geneand to replace it with a single copy of the thy1 -tau-4R/2N expression construct) survived normally withminor motor problems late in life and without anyobvious pathology [122]. When these findings arecompared with those obtained by other researchgroups, it becomes obvious that the different strainsshow a range of phenotypes, possibly due to the use

of different promoters for transgene expression, theintegration site of the transgene, expression levels,and the mouse strain used for transgenesis [5].

In light of the neuropathological findings inhumans that only a subset of the neuronal loss canbe explained by NFTs, an important questionarises, namely whether NFTs are an incidentalmarker for the neurotoxic cascade in AD or ratherrepresent a protective neuronal response, allowingsequestration of neurotoxic species into a lessharmful stable form [125]. To address this ques-tion, P301L mice were generated where the trans-gene can be turned off (or at least reduced fromvery high to only high overexpression levels). Itwas found that mice expressing doxycycline-repressible human P301L mutant tau developed

progressive age-related NFTs, a remarkable neuronloss, and behavioral impairment. After the suppres-sion of transgenic tau from 13- to 2.5-fold overex-pression, memory function recovered, and neuronnumbers stabilized, but NFTs continued to accu-mulate. These data convincingly show that tau dys-function impairs memory, when massivelyoverexpressed. The data further imply that NFTsper se (as entities of fibrillar accumulation that arevisible by light microscopy) are not sufficient tocause cognitive decline or neuronal death in thismodel of tauopathy [125]. Not surprisingly, cogni-tive impairment in a second P301L tau transgenicmouse strain was shown to occur in the absence of NFT formation [126, 127]. As NFTs make up onlya small percentage of all neurons in any animalmodel published so far, and as they are by farexceeded by dysfunctional neurons with tau aggre-gates but lacking NFTs, it is not surprising that,considering the limited life-span of mice comparedwith humans, NFT numbers do not correlate with

functional impairment in these mice but rather thehigh number of cells that display tau aggregates.

12.6 Tau Transgenic Mice:Correlation of Histopathologyand Behavioral Impairment

Similar to the APP transgenic models, the tautransgenic mouse models have been assessed usinga wide range of behavioral tasks. Our mThy1.2promoter-driven P301L mice accumulate tau in




many brain areas but develop NFTs mainly in theamygdala. This brain area is involved in mediatingeffects of emotion and stress on learning and mem-ory [124, 128, 129]. Therefore, behavioral alter-ations and cognitive deficits of the P301L micewere investigated using an amygdala-specific testbattery for anxiety-related and cognitive behavior.These included an open-field, a light-dark box, fearconditioning, and a conditioned taste aversion(CTA) test [126]. The P301L mice showed anincreased exploratory behavior but normal anxietylevels and no impairment in fear conditioning. Inthe P301L mice, fear conditioning was unaffectedprobably due to the absence of tau aggregates in thecentral and lateral nucleus of the amygdala. In theCTA test, the mice learn to associate a novel taste

with nausea and, as a consequence, avoid con-sumption of this specific taste at the next presenta-tion. We found that acquisition and consolidationof CTA memory was not significantly affected bythe P301L transgene. However, transgenic miceextinguished the CTA memory more rapidly thandid wild-type mice [126]. This rapid extinctionmay be due to the presence of tau aggregates in thebasolateral nucleus of the amygdala, which hasbeen shown to be essential for the extinction of CTA memory, whereas acquisition is dependent onan intact central nucleus, where no tau aggregateswere found. When the P301L mice were assessedin hippocampus-dependent behavioral tests, theMorris water maze and Y-maze revealed intact spa-tial working memory but impairment in spatialreference memory at 6 and 11 months of age. Inaddition, a modest disinhibition of exploratorybehavior at 6 months of age was confirmed in theopen-field and the elevated O-maze and was morepronounced during aging [127].

The PrP promoter-driven P301L tau transgenicmice strongly overexpress mutant tau in severalneuronal cell-types, including motor neurons.Therefore, they develop a progressive motor phe-notype [114]. The V337M tau mutant mice show avery confined expression pattern as mutant tau wasdetected only in the hippocampus. These miceshow an increased locomotor activity and memorydeficits in the elevated plus maze, increased spon-taneous locomotion in the open-field, but no sig-nificant impairment in the Morris water maze[130]. R406W tau mutant mice express tau athighest levels in the hippocampus and, to a lesser

extent, in other cortical and subcortical brain areas.However, in the amygdala, only a few cellsstrongly express mutant tau, even in old animals[117]. These mice show a slight decrease in loco-motor activity during the first minutes of the open-field test and a significant impairment in thecontextual and cued fear-conditioning test.

When triple-transgenic mice (PS1 M146Vknock-in microinjected with APP sw and P301L tautransgenes) were analyzed, 2-month-old micewere cognitively unimpaired. The earliest cogni-tive impairment manifested at 4 months as a deficitin long-term retention and correlated with theaccumulation of intraneuronal A β in the hip-pocampus and amygdala. Plaque or NFT pathol-ogy was not apparent at this age, suggesting that

they contribute to cognitive dysfunction at latertime points [131].

In summary, these findings demonstrate that tauaggregation in distinct brain areas directly affectsthe performance in memory tests controlled bythese brain areas. They also show that tau aggrega-tion per se, in the absence of NFT formation, is suf-ficient to cause behavioral deficits.

12.7 Cross-Talk of β-Amyloid andTau in Experimental ModelSystems

Before NFT formation had been achieved in tautransgenic mice, the interaction of plaques andNFTs has been addressed in different non-trans-genic species such as rats and monkeys [132].Intracerebral injection of plaque-equivalent con-centrations of fibrillar, but not soluble, A β resulted

in profound neuronal loss, tau phosphorylation,and microglial proliferation in the aged rhesusmonkey cerebral cortex. In contrast, the samepreparations were not toxic in the young adult rhe-sus brain, indicating a role for age in A β toxicity.This toxicity was also highly species-specific as itwas neither observed in young nor in aged rats[132]. These results suggested that A β neurotoxic-ity in vivo is a pathological response of the agingbrain, which is most pronounced in higher orderprimates. Thus, longevity may contribute to theunique susceptibility of humans to AD by render-ing the brain vulnerable to A β neurotoxicity.

206 J. Götz et al.



In transgenic mice, the presence of the P301Lmutation appeared to accelerate tau filament for-mation as transgenic mice with high expressionlevels of human tau developed NFTs only at a highage [113–115]. P301L mutant mice are thereforesuitable models to determine whether A β affectsthe tau pathology in these mice. Synthetic prepara-tions of fibrillar A β42 were stereotaxically injectedinto the somatosensory cortex and the hippocampalCA1 region of P301L and wild-type human tautransgenic mice and non-transgenic littermate con-trols, causing a fivefold increase of NFTs in theamygdala of P301L transgenic, but not wild-typetau transgenic or control mice, 18 days after theinjections [124]. In contrast, when the non-fibrillo-genic reversed peptide A β42-1 was injected, levels

of NFTs were not affected (Fig. 12.1). NFT forma-tion in the A β42-injected P301L mice was tightlycorrelated with the pathological phosphorylation of tau at S422 and the epitope AT100 (T212/S214),but not AT8 (S202/T205). The finding that A β42was not capable of inducing NFT formation in non-NFT-forming wild-type tau transgenic mice mayreflect species differences between mice and men.Alternatively, it may imply that, at least in mice,Aβ42 cannot induce NFT formation de novo , whichwould be in disagreement with the amyloid cas-cade hypothesis. Interestingly, in cultured murinehippocampal neurons, toxicity of A β42 has beenshown to be dependent on the presence of tau[133].

An alternative approach was chosen by Lewisand co-workers who crossed A β-producing APP-mutant Tg2576 mice with their PrP promoter-driven P301L tau mutant mice [134]. Doubletransgenic mice showed a more than sevenfoldincrease in NFT numbers in the olfactory bulb, the

entorhinal cortex, and the amygdala compared withP301L single transgenic mice, whereas A β plaqueformation was unaffected by the presence of the taulesions (Fig. 12.1).

When both approaches are taken together, theyimply that not all brain areas are similarly suscep-tible to A β-mediated NFT induction. In both stud-ies, the amygdala is a hot spot of NFT induction.Unless tau levels are particularly high in the amyg-dala compared with other brain areas such asthe hippocampus or cortical areas, a differentmRNA/protein profile may account for theobserved differences. A recent study of amygdala-

specific gene expression provided a list of genes,some of which may confer an increased tau-relatedvulnerability of amygdaloid neurons to A β42 [135].Alternatively, it may be the nerve terminals, whichare susceptible to A β42, whereas direct exposure of the cell body or neurites may not pose a risk to thetau-expressing neuron. Whether A β is taken up byreceptor-mediated mechanisms or whether it formspores is still a matter of debate [136, 137](Fig. 12.3).

Antibody-directed approaches were pursued in arecent study to dissect the cross-talk of A β and tau.When triple transgenic mice (PS1 M146V knock-in microinjected with APP sw and P301L tau trans-genes) were intracerebrally injected with anti-A βantibodies or a γ -secretase inhibitor, this resulted in


(a)

(b)

(1a) Selectivevulnerability

(2) Uptake andtransport of A β

β-amyloid

β-amyloid

(4) Formationof pores

Ca 2+

(1b) Selectiveprotection

no NFTinduction

NFTinduction

(3) (Receptor-mediated) damageto nerve terminals

FIGURE 12.3. The mechanism of A β-mediated neurotox-icity is not understood at all. Whereas some neurons areparticularly vulnerable already early in disease (A, 1a),others are relatively spared (B, 1b). Possible mechanismsof A β neurotoxicity and downstream NFT formationinclude uptake and transport of A β (2), (receptor-medi-ated) damage to nerve terminals (3), and the formation of pores (4). The receptors may have a selective specificity

for A β or may, alternatively, bind peptides with a β-crossstructure as the defining feature of amyloid fibers suchas A β.



the disappearance of somatodendritic tau stainingin young, but not old, mice [138]. It thus appearsthat extracellular A β deposits can exacerbate theintraneuronal pathology caused by the expressionof mutant human tau protein [23].

An interaction between A β and tau was alsodemonstrated after the functional validation of pro-teomics findings in P301L tau transgenic mice[139]. Here, mainly mitochondrial proteins, antioxi-dant enzymes, and synaptic proteins were identifiedas modified in the proteome pattern of P301L taumice. Significantly, the reduction in mitochondrialcomplex V levels in the P301L tau mice found byusing proteomics was also confirmed as decreased inbrains derived from human carriers of the P301Lmutation of tau. Functional analysis demonstrated a

mitochondrial dysfunction in P301L tau micetogether with reduced NADH-ubiquinone oxidore-ductase activity and, with age, impaired mitochondr-ial respiration and ATP synthesis. Mitochondrialdysfunction was associated with higher levels of reactive oxygen species in aged transgenic mice.Increased tau pathology as in aged homozygousP301L tau mice revealed modified lipid peroxida-tion levels and the upregulation of antioxidantenzymes in response to oxidative stress. To investi-gate whether brain cells from P301L tau mice aremore susceptible to A β, we measured the mitochon-drial membrane potential of isolated cortical braincells with and without A β treatment [139]. Previousexperiments using PC12 cells had shown that extra-cellular A β treatment lead to a significant decreasein mitochondrial membrane potential [140]. Wefound that, interestingly, the basal mitochondrialmembrane potential was still conserved in cerebralcells from P301L tau mice. However a secondaryinsult with A β42 resulted in a higher reduction in

membrane potential in P301L tau mitochondria thanin wild-type controls. Importantly, this effect wasbrain region–specific and therefore probablydependent on the presence of P301L tau becausecells from the cerebellum with very low P301L tauexpression levels were not vulnerable to this damagewhereas cells from the cerebrum with high P301Ltau expression levels were. These data suggests asynergistic action of A β and tau pathology on mito-chondrial function. Moreover, it can be concludedthat the tau pathology involves a mitochondrial andoxidative stress disorder distinct from that caused byAβ [139].

The interaction between A β and tau has alsobeen addressed in cell lines. Several studies haveshown that tau-expressing cell lines are responsiveto different forms of pathogenic stimuli. For exam-ple, when human SH-SY5Y neuroblastoma cellswere incubated with okadaic acid (OA), a potentphosphatase inhibitor, together with HNE, a prod-uct of lipid oxidation found to be associated withNFTs in vivo [141–143], this resulted in theassembly of tau into aberrant polymers [144].Most of them had a diameter of 2–3 nm and werestraight, whereas PHFs have a diameter of 20 nmand are twisted. Fibrillar aggregates of tau werealso observed in Chinese hamster ovary (CHO)cells that have been transfected with mutant tauexpression constructs [145]. For example, ∆280K,

but not several other single tau mutants (such asV337M, P301L, and R406W), developed insolu-ble amorphous and fibrillar aggregates, whereas atriple tau mutant containing V337M, P301L, andR406W substitutions (VPR) also formed similaraggregates. Furthermore, the aggregates increasedin size over time. The formation of aggregated∆280K and VPR tau protein correlated with theirreduced affinity to bind microtubules. Reducedphosphorylation and altered proteolysis was alsoobserved in R406W and ∆280K tau mutants. Thus,distinct pathological phenotypes, including theformation of insoluble filamentous tau aggregates,result from the expression of different FTDP-17tau mutants in transfected CHO cells suggestingthat these missense mutations cause diverse neu-rodegenerative FTDP-17 syndromes by multiplemechanisms.

As mentioned above, in human tauopathies otherthan AD, tau-positive inclusions are not restrictedto neurons. They are found in oligodendrocytes and

are a consistent neuropathological feature of CBD,PSP, and some forms of FTDP-17. When an oligo-dendroglial cell line was engineered to stablyexpress high levels of the longest human tau iso-form, treatment with OA caused tau hyperphos-phorylation and a decreased binding of tau tomicrotubules. Transiently, tau-positive aggregatesformed that could be stained with the amyloid-binding dye thioflavin-S. However, when the pro-teasome was inhibited by MG-132 after OAtreatment, the aggregates were stabilized and werestill detectable after 18 h in the absence of OA.Incubation with MG-132 alone did not induce the

208 J. Götz et al.



formation of thioflavin-S-positive aggregates.Hence, although tau hyperphosphorylation inducedby protein phosphatase inhibition contributed topathological aggregate formation, only hyperphos-phorylation of tau followed by proteasome inhibi-tion led to stable fibrillar deposits of tau similar tothose observed in human tauopathies [146].Together, these studies demonstrate that tau iscapable of forming filamentous aggregates underspecific experimental conditions.

Previous stereotaxic injection experiments havedemonstrated principal differences between miceand men: Whereas A β induced NFT formation inhuman P301L mutant mice, it failed to do so inhuman wild-type tau transgenic mice. This is dif-ferent from the situation in human AD, where A β

aggregation and NFT formation occur in theabsence of pathogenic tau mutations. Therefore, toaddress the role of A β in tau fibrillogenesis in a tis-sue culture system, we chose the human SH-SY5Yneuroblastoma instead of a murine cell line. SH-SY5Y cells can be neuronally differentiated by thesequential treatment with retinoic acid and brain-derived neurotrophic factor (BDNF) [147](Fig. 12.4). They can be transplanted into mousebrain where they persist for a couple of days.

Moreover, they anatomically integrate into organ-otypic hippocampal slices where they expresssynaptic markers and fire action potentials after 20days in culture [O. Rainteau, A. Ferrari, andJ. Götz, unpublished observations]. We stablyexpressed human tau with and without pathogenicmutations in these cells and exposed them for 5days to aggregated synthetic A β42 (Fig. 12.4)[148]. This caused a decreased solubility of taualong with the generation of PHF-like tau contain-ing filaments, which were 20 nm wide and hadperiodicities of 130 to 140 nm in the presence of P301L mutant tau or 150 to 160 nm in the presenceof wild-type tau (Fig. 12.4). As the stereotaxicAβ42 injection experiments had linked the S422epitope of tau to NFT formation, we mutagenized

serine 422 into alanine (which was intended toabrogate phosphorylation) and glutamic acid(intended to mimic phosphorylation). To our sur-prise, both mutations prevented the A β42-mediateddecrease in solubility and the generation of PHF-like filaments suggesting a role of S422 or its phos-phorylation in tau filament formation. S422 islocated next to a putative caspase-3 cleavage site atposition 421, and altered caspase cleavage has beenshown to be involved in the rates of tau filamentformation [149–151]. Together, these data under-score a role of A β42 in the formation of PHF-likefilaments. These data are consistent with our previ-ous results of A β42-induced PHF-like tau filamentformation in P301L tau transgenic mice [124] butin contrast to the transgenic mice A β42-inducedPHF formation in tissue culture also occurred withwild-type mice. This may be related to the speciesdifference and points to the possibility that humancells in culture may be more susceptible to the for-mation of abnormal tau filaments than are murine

cells in vivo.The tissue culture system has since been used to

map additional phospho-epitopes of tau involved inPHF formation and revealed that mutagenesis of some sites is even inhibitory to tau filament forma-tion of endogenous, non-mutant tau [152]. Furtheradaptation of the system may allow the screeningand validation of compounds designed to preventPHF formation.

In summary, the above experiments demonstratepathological interactions between A β and tau thatled to increased NFT formation. Moreover, theregion-specific induction of A β-mediated NFT


FIGURE 12.4. The formation of PHFs in tissue culturewas reproduced by stably expressing human tau (bothwild-type and P301L mutant) in neuronally differenti-ated human SH-SY5Y cells and exposing them for5 days to aggregated synthetic A β42. An electron micro-graph of the fibrillar preparations of A β42 is included (onthe left). This incubation caused the generation of PHF-like tau containing filaments that were 20 nm wide and

had periodicities of 130 to 140 nm in the presence of P301L mutant tau or 150 to 160 nm in the presence of wild-type tau (on the right).



formation in P301L tau transgenic mice mirrors, tosome extent, the regional vulnerability observed inAD brains. Finally, besides their major advantagesfor an understanding of the pathophysiology of NFT formation, these models may assist in thedevelopment of therapies designed to reduce NFTformation and tau-related dysfunction, be they A β-mediated or not.

12.8 Outlook

The recent advent of transcriptomic and proteomictechnology and its application to transgenic mousemodels and tissue culture systems is likely to assistin the dissection of the pathocascade of AD and

FTD [153]. Transcriptomics and proteomics iden-tify individual, differentially regulated mRNAs andproteins and are in addition employed to dissectsignaling pathways and reveal networks by usingan integrated approach. This will undoubtedly leadto a redefinition and subdivision of disease entitiesbased on biochemical criteria rather than the clini-cal presentation. Moreover, it will determinewhether the pathogenesis of FAD and SAD areshared. Whether this can be reconciled with a uni-fying theory for AD remains to be determined. Inany case, the new knowledge will have importantimplications for treatment strategies [97, 98].

Acknowledgments J.G. is a Fellow of the MedicalFoundation. This work was supported by grantsfrom the University of Sydney and the MedicalFoundation (University of Sydney) to J.G.

References

1. Ritchie K, Lovestone S. The dementias. Lancet 2002;360:1759-66.

2. Arnold SE, Hyman BT, Flory J, et al. The topo-graphical and neuroanatomical distribution of neu-rofibrillary tangles and neuritic plaques in thecerebral cortex of patients with Alzheimer’s disease.Cereb Cortex 1991; 1:103-16.

3. Ratnavalli E, Brayne C, Dawson K, Hodges JR. Theprevalence of frontotemporal dementia. Neurology2002; 58:1615-21.

4. Kurosinski P, Guggisberg M, Gotz J. Alzheimer’sand Parkinson’s disease—Overlapping or synergisticpathologies? Trends Mol Med 2002; 8:3-5.

5. Gotz J, Streffer JR, David D, et al. Transgenic animalmodels of Alzheimer’s disease and related disor-ders: Histopathology, behavior and therapy. MolPsychiatry 2004; 9:664-683.

6. Glenner GG, Wong CW. Alzheimer’s disease: initialreport of the purification and characterization of a

novel cerebrovascular amyloid protein. BiochemBiophys Res Commun 1984; 120:885-90.

7. Masters CL, Simms G, Weinman NA, et al. Amyloidplaque core protein in Alzheimer’s disease andDown syndrome. Proc Natl Acad Sci U S A 1985;82:4245-9.

8. Vassar R, Bennett BD, Babu-Khan S, et al. Beta-sec-retase cleavage of Alzheimer’s amyloid precursorprotein by the transmembrane aspartic proteaseBACE. Science 1999; 286:735-41.

9. Edbauer D, Winkler E, Regula JT, et al. Reconstitutionof gamma-secretase activity. Nat Cell Biol 2003;5:486-8.

10. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol(Berlin) 1991; 82:239-59.

11. Braak H, Braak E. Staging of Alzheimer’s disease-related neurofibrillary changes. Neurobiol Aging1995; 16:271-8; discussion 278-84.

12. Crowther RA, Wischik CM. Image reconstruction of the Alzheimer paired helical filament. EMBO J1985; 4:3661-5.

13. Wischik CM, Crowther RA, Stewart M, Roth M.

Subunit structure of paired helical filaments inAlzheimer’s disease. J Cell Biol 1985; 100:1905-12.

14. Goedert M, Wischik CM, Crowther RA, et al.Cloning and sequencing of the cDNA encoding acore protein of the paired helical filament of Alzheimer’s disease: identification as the microtubule-associated protein tau. Proc Natl Acad Sci U S A1988; 85:4051-5.

15. Lichtenberg B, Mandelkow EM, Hagestedt T,Mandelkow E. Structure and elasticity of microtubule-associated protein tau. Nature 1988; 334:359-62.

16. Schweers O, Schonbrunn-Hanebeck E, Marx A,Mandelkow E. Structural studies of tau protein andAlzheimer paired helical filaments show no evidencefor beta-structure. J Biol Chem 1994; 269:24290-7.

17. Goedert M, Spillantini MG, Jakes R, et al. Moleculardissection of the paired helical filament. NeurobiolAging 1995; 16:325-34.

18. Buee L, Bussiere T, Buee-Scherrer V, et al. Tau pro-tein isoforms, phosphorylation and role in neurode-generative disorders. Brain Res Brain Res Rev 2000;33:95-130.

19. Lee VM, Goedert M, Trojanowski JQ.Neurodegenerative tauopathies. Annu Rev Neurosci2001; 24:1121-59.

210 J. Götz et al.



20. Gotz J. Tau and transgenic animal models. Brain ResBrain Res Rev 2001; 35:266-86.

21. Chen F, David D, Ferrari A, Gotz J. Posttranslationalmodifications of tau—Role in human tauopathiesand modeling in transgenic animals. Curr DrugTargets 2004; 5:503-15.

22. Berriman J, Serpell LC, Oberg KA, et al. Tau fila-ments from human brain and from in vitro assemblyof recombinant protein show cross-beta structure.Proc Natl Acad Sci U S A 2003; 100:9034-8.

23. Goedert M, Jakes R. Mutations causing neurodegen-erative tauopathies. Biochim Biophys Acta 2005;1739:240-50.

24. Tashiro K, Hasegawa M, Ihara Y, Iwatsubo T.Somatodendritic localization of phosphorylated tauin neonatal and adult rat cerebral cortex. Neuroreport1997; 8:2797-801.

25. Lee G, Rook SL. Expression of tau protein in non-neuronal cells: Microtubule binding and stabiliza-tion. J Cell Sci 1992; 102:227-37.

26. Reszka AA, Seger R, Diltz CD, et al. Association of mitogen-activated protein kinase with the micro-tubule cytoskeleton. Proc Natl Acad Sci U S A 1995;92:8881-5.

27. Morishima-Kawashima M, Kosik KS. The pool of map kinase associated with microtubules is small butconstitutively active. Mol Biol Cell 1996; 7:893-905.

28. Flanagan LA, Cunningham CC, Chen J, et al. Thestructure of divalent cation-induced aggregates of

PIP2 and their alteration by gelsolin and tau. BiophysJ 1997; 73:1440-7.

29. Ebneth A, Godemann R, Stamer K, et al.Overexpression of tau protein inhibits kinesin-dependent trafficking of vesicles, mitochondria, andendoplasmic reticulum: Implications for Alzheimer’sdisease. J Cell Biol 1998; 143:777-94.

30. Jenkins SM, Johnson GV. Tau complexes withphospholipase C-gamma in situ. Neuroreport 1998;9:67-71.

31. Sontag E, Nunbhakdi-Craig V, Lee G, et al.

Molecular interactions among protein phosphatase2A, tau, and microtubules. Implications for the regu-lation of tau phosphorylation and the development of tauopathies. J Biol Chem 1999; 274:25490-8.

32. Anderton BH, Dayanandan R, Killick R, LovestoneS. Does dysregulation of the Notch and wingless/ Wnt pathways underlie the pathogenesis of Alzheimer’s disease? Mol Med Today 2000; 6:54-9.

33. De F, Gv, Inestrosa NC. Wnt signaling function inAlzheimer’s disease [In Process Citation]. Brain ResBrain Res Rev 2000; 33:1-12.

34. Maas T, Eidenmuller J, Brandt R. Interaction of tauwith the neural membrane cortex is regulatedby phosphorylation at sites that are modified in

paired helical filaments. J Biol Chem 2000;275:15733-40.

35. Goedert M, Spillantini MG, Cairns NJ, CrowtherRA. Tau proteins of Alzheimer paired helical fila-ments: abnormal phosphorylation of all six brain iso-forms. Neuron 1992; 8:159-68.

36. Sherrington R, Rogaev EI, Liang Y, et al. Cloning of agene bearing missense mutations in early-onset famil-ial Alzheimer’s disease. Nature 1995; 375:754-60.

37. Van Broeckhoven C, Backhovens H, Cruts M, et al.Mapping of a gene predisposing to early-onsetAlzheimer’s disease to chromosome 14q24.3. NatGenet 1992; 2:335-9.

38. Rocchi A, Pellegrini S, Siciliano G, Murri L.Causative and susceptibility genes for Alzheimer’sdisease: A review. Brain Res Bull 2003; 61:1-24.

39. Hodges JR, Davies RR, Xuereb JH, et al.Clinicopathological correlates in frontotemporaldementia. Ann Neurol 2004; 56:399-406.

40. Dickson DW. Pick’s disease: A modern approach.Brain Pathol 1998; 8:339-54.

41. McKhann GM, Albert MS, Grossman M, et al.Clinical and pathological diagnosis of frontotempo-ral dementia: Report of the Work Group onFrontotemporal Dementia and Pick’s Disease. ArchNeurol 2001; 58:1803-9.

42. Hodges JR, Davies R, Xuereb J, et al. Survival in fron-totemporal dementia. Neurology 2003; 61:349-54.

43. Hutton M, Lendon CL, Rizzu P, et al. Association of

missense and 5 ′-splice-site mutations in tau with theinherited dementia FTDP-17. Nature 1998; 393:702-5.

44. Poorkaj P, Bird TD, Wijsman E, et al. Tau is a candi-date gene for chromosome 17 frontotemporaldementia. Ann Neurol 1998; 43:815-25.

45. Spillantini MG, Murrell JR, Goedert M, et al.Mutation in the tau gene in familial multiple systemtauopathy with presenile dementia. Proc Natl AcadSci U S A 1998; 95:7737-41.

46. Clark LN, Poorkaj P, Wszolek Z, et al. Pathogenicimplications of mutations in the tau gene in pallido-

ponto-nigral degeneration and related neurodegener-ative disorders linked to chromosome 17. Proc NatlAcad Sci U S A 1998; 95:13103-7.

47. Dumanchin C, Camuzat A, Campion D, et al.Segregation of a missense mutation in the micro-tubule-associated protein tau gene with familial fron-totemporal dementia and parkinsonism. Hum MolGenet 1998; 7:1825-9.

48. Spillantini MG, Crowther RA, Kamphorst W, et al.Tau pathology in two Dutch families with mutationsin the microtubule-binding region of tau. Am JPathol 1998; 153:1359-63.

49. Delisle MB, Murrell JR, Richardson R, et al. Amutation at codon 279 (N279K) in exon 10 of the




Tau gene causes a tauopathy with dementia andsupranuclear palsy. Acta Neuropathol (Berlin)1999; 98:62-77.

50. D’Souza I, Poorkaj P, Hong M, et al. Missense andsilent tau gene mutations cause frontotemporaldementia with parkinsonism-chromosome 17 type,

by affecting multiple alternative RNA splicing regu-latory elements. Proc Natl Acad Sci U S A 1999;96:5598-603.

51. Goedert M, Spillantini MG, Crowther RA, et al. Taugene mutation in familial progressive subcorticalgliosis. Nat Med 1999; 5:454-7.

52. Mirra SS, Murrell JR, Gearing M, et al. Tau pathol-ogy in a family with dementia and a P301L mutationin tau. J Neuropathol Exp Neurol 1999; 58:335-45.

53. Rizzu P, Van Swieten JC, Joosse M, et al. Highprevalence of mutations in the microtubule-associ-ated protein tau in a population study of frontotem-poral dementia in the Netherlands. Am J Hum Genet1999; 64:414-21.

54. Sperfeld AD, Collatz MB, Baier H, et al. FTDP-17:An early-onset phenotype with parkinsonism andepileptic seizures caused by a novel mutation [seecomments]. Ann Neurol 1999; 46:708-15.

55. Arima K, Kowalska A, Hasegawa M, et al. Twobrothers with frontotemporal dementia and parkin-sonism with an N279K mutation of the tau gene.Neurology 2000; 54:1787-95.

56. Delisle MB, Uro-Coste E, Murrell JR, et al.

Neurodegenerative disease associated with a muta-tion of codon 279(N279K) in exon 10 of Tau protein.Bull Acad Natl Med 2000; 184:799-809.

57. Tolnay M, Grazia Spillantini M, Rizzini C, et al. A newcase of frontotemporal dementia and parkinsonismresulting from an intron 10 +3-splice site mutation inthe tau gene: Clinical and pathological features.Neuropathol Appl Neurobiol 2000; 26:368-78.

58. Lippa CF, Zhukareva V, Kawarai T, et al.Frontotemporal dementia with novel tau pathology anda Glu342Val tau mutation. Ann Neurol 2000; 48:850-8.

59. Pastor P, Pastor E, Carnero C, et al. Familial atypicalprogressive supranuclear palsy associated withhomozigosity for the delN296 mutation in the taugene. Ann Neurol 2001; 49:263-7.

60. Hayashi S, Toyoshima Y, Hasegawa M, et al. Late-onsetfrontotemporal dementia with a novel exon 1 (Arg5His)tau gene mutation. Ann Neurol 2002; 51:525-30.

61. Goedert M, Crowther RA, Spillantini MG. Tau muta-tions cause frontotemporal dementias. Neuron 1998;21:955-8.

62. Delacourte A, Sergeant N, Wattez A, et al. Vulnerableneuronal subsets in Alzheimer’s and Pick’s disease aredistinguished by their tau isoform distribution andphosphorylation. Ann Neurol 1998; 43:193-204.

63. Mott RT, Dickson DW, Trojanowski JQ, et al.Neuropathologic, biochemical, and molecular char-acterization of the frontotemporal dementias.J Neuropathol Exp Neurol 2005; 64:420-8.

64. Kurosinski P, Gotz J. Glial cells under physiologicand pathological conditions. Arch Neurol 2002; 59:

1524-8.65. Cervos-Navarro J, Schumacher K. Neurofibrillary

pathology in progressive supranuclear palsy (PSP).J Neural Transm Suppl 1994; 42:153-64.

66. Delacourte A. Pathological Tau proteins of Alzheimer’s disease as a biochemical marker of neu-rofibrillary degeneration. Biomed Pharmacother1994; 48:287-95.

67. Ksiezak-Reding H, Morgan K, Mattiace LA, et al.Ultrastructure and biochemical composition of paired helical filaments in corticobasal degeneration.Am J Pathol 1994; 145:1496-508.

68. Wischik CM, Novak M, Thogersen HC, et al.Isolation of a fragment of tau derived from the coreof the paired helical filament of Alzheimer’s disease.Proc Natl Acad Sci U S A 1988; 85:4506-10.

69. Bondareff W, Wischik CM, Novak M, et al.Molecular analysis of neurofibrillary degeneration inAlzheimer’s disease. An immunohistochemicalstudy. Am J Pathol 1990; 137:711-23.

70. Murrell JR, Spillantini MG, Zolo P, et al. Tau genemutation G389R causes a tauopathy with abundantpick body-like inclusions and axonal deposits.

J Neuropathol Exp Neurol 1999; 58:1207-26.71. Rizzini C, Goedert M, Hodges JR, et al. Tau gene

mutation K257T causes a tauopathy similar to Pick’sdisease. J Neuropathol Exp Neurol 2000; 59:990-1001.

72. Neumann M, Schulz-Schaeffer W, Crowther RA,et al. Pick’s disease associated with the novelTau gene mutation K369I. Ann Neurol 2001; 50:503-13.

73. Rosso SM, van Herpen E, Deelen W, et al. A noveltau mutation, S320F, causes a tauopathy with inclu-

sions similar to those in Pick’s disease. Ann Neurol2002; 51:373-6.74. Kobayashi T, Ota S, Tanaka K, et al. A novel L266V

mutation of the tau gene causes frontotemporaldementia with a unique tau pathology. Ann Neurol2003; 53:133-7.

75. van Herpen E, Rosso SM, Serverijnen LA, et al.Variable phenotypic expression and extensive taupathology in two families with the novel tau mutationL315R. Ann Neurol 2003; 54:573-81.

76. Pickering-Brown SM, Baker M, Nonaka T, et al.Frontotemporal dementia with Pick-type histologyassociated with Q336R mutation in the tau gene.Brain 2004; 127:1415-26.

212 J. Götz et al.



104. Janus C, Pearson J, McLaurin J, et al. A beta pep-tide immunization reduces behavioural impairmentand plaques in a model of Alzheimer’s disease.Nature 2000; 408:979-82.

105. Mucke L, Masliah E, Yu GQ, et al. High-level neu-ronal expression of abeta 1-42 in wild-type human

amyloid protein precursor transgenic mice:Synaptotoxicity without plaque formation. JNeurosci 2000; 20:4050-8.

106. Gotz J, Probst A, Spillantini MG, et al.Somatodendritic localization and hyperphosphory-lation of tau protein in transgenic mice expressingthe longest human brain tau isoform. EMBO J1995; 14:1304-13.

107. Ishihara T, Hong M, Zhang B, et al. Age-dependentemergence and progression of a tauopathy in trans-genic mice overexpressing the shortest human tauisoform. Neuron 1999; 24:751-62.

108. Spittaels K, Van den Haute C, Van Dorpe J, et al.Prominent axonopathy in the brain and spinal cordof transgenic mice overexpressing four-repeathuman tau protein. Am J Pathol 1999; 155:2153-65.

109. Probst A, Gotz J, Wiederhold KH, et al.Axonopathy and amyotrophy in mice transgenic forhuman four-repeat tau protein. Acta Neuropathol(Berlin) 2000; 99:469-81.

110. Hirano A, Nakano I, Kurland LT, et al. Fine struc-tural study of neurofibrillary changes in a familywith amyotrophic lateral sclerosis. J Neuropathol

Exp Neurol 1984; 43:471-80.111. Munoz DG, Greene C, Perl DP, Selkoe DJ.

Accumulation of phosphorylated neurofilaments inanterior horn motoneurons of amyotrophic lateralsclerosis patients. J Neuropathol Exp Neurol 1988;47:9-18.

112. Rouleau GA, Clark AW, Rooke K, et al. SOD1mutation is associated with accumulation of neuro-filaments in amyotrophic lateral sclerosis. AnnNeurol 1996; 39:128-31.

113. Ishihara T, Zhang B, Higuchi M, et al. Age-depend-

ent induction of congophilic neurofibrillary tauinclusions in tau transgenic mice. Am J Pathol2001; 158:555-62.

114. Lewis J, McGowan E, Rockwood J, et al.Neurofibrillary tangles, amyotrophy and progres-sive motor disturbance in mice expressing mutant(P301L) tau protein. Nat Genet 2000; 25:402-5.

115. Gotz J, Chen F, Barmettler R, Nitsch RM. Tau fila-ment formation in transgenic mice expressingP301L tau. J Biol Chem 2001; 276:529-34.

116. Tanemura K, Akagi T, Murayama M, et al.Formation of filamentous tau aggregations in trans-genic mice expressing V337M human tau.Neurobiol Dis 2001; 8:1036-45.

117. Tatebayashi Y, Miyasaka T, Chui DH, et al. Taufilament formation and associative memorydeficit in aged mice expressing mutant (R406W)human tau. Proc Natl Acad Sci U S A 2002;99:13896-901.

118. Allen B, Ingram E, Takao M, et al. Abundant tau fil-

aments and nonapoptotic neurodegeneration intransgenic mice expressing human P301S tauprotein. J Neurosci 2002; 22:9340-51.

119. Gotz J, Tolnay M, Barmettler R, et al.Oligodendroglial tau filament formation in trans-genic mice expressing G272V tau. Eur J Neurosci2001; 13:2131-40.

120. Higuchi M, Ishihara T, Zhang B, et al. Transgenicmouse model of tauopathies with glial pathologyand nervous system degeneration. Neuron 2002;35:433-46.

121. Oddo S, Caccamo A, Shepherd JD, et al. Triple-transgenic model of Alzheimer’s disease withplaques and tangles. Intracellular abeta and synap-tic dysfunction. Neuron 2003; 39:409-21.

122. Terwel D, Lasrado R, Snauwaert J, et al. Changedconformation of mutant Tau-P301L underlies themoribund tauopathy, absent in progressive, non-lethal axonopathy of Tau-4R/2N transgenic mice.J Biol Chem 2005; 280:3963-73.

123. Sobrido MJ, Miller BL, Havlioglu N, et al. Noveltau polymorphisms, tau haplotypes, and splicing infamilial and sporadic frontotemporal dementia.

Arch Neurol 2003; 60:698-702.124. Gotz J, Chen F, van Dorpe J, Nitsch RM. Formation

of neurofibrillary tangles in P301L tau transgenicmice induced by Abeta 42 fibrils. Science 2001;293:1491-5.

125. Santacruz K, Lewis J, Spires T, et al. Tau suppres-sion in a neurodegenerative mouse model improvesmemory function. Science 2005; 309:476-81.

126. Pennanen L, Welzl H, D’Adamo P, et al.Accelerated extinction of conditioned taste aversionin P301L tau transgenic mice. Neurobiol Dis 2004;

15:500-9.127. Pennanen L, Wolfer D, Nitsch RM, GotzJ. Impaired spatial reference memory in P301L tautransgenic mice. Genes, Brain and Behavior 2006;5:369-379.

128. LeDoux JE. Emotion circuits in the brain. AnnuRev Neurosci 2000; 23:155-84.

129. Welzl H, D’Adamo P, Lipp HP. Conditioned tasteaversion as a learning and memory paradigm.Behav Brain Res 2001; 125:205-13.

130. Tanemura K, Murayama M, Akagi T, et al.Neurodegeneration with tau accumulation in atransgenic mouse expressing V337M human tau.J Neurosci 2002; 22:133-41.

214 J. Götz et al.



13Glial Cells and A β Peptides in

Alzheimer’s Disease PathogenesisGilbert Siu, Peter Clifford, Mary Kosciuk, Venkat Venkataraman,and Robert G. Nagele

216

13.1 Introduction

Alzheimer’s disease (AD) is a tragic neurodegenerativedisorder that targets the elderly and ultimately endsin dementia. Unfortunately, the ever increasinglength of the human life span in the United Statesand throughout the world is now being paralleledby corresponding increases in the incidence of ADas well as in the duration of this disease in individ-ual patients. AD is characterized symptomaticallyby progressive cognitive and memory loss, lan-guage deficits, impairment of judgment, deficientproblem solving, and reduced abstract thought. Atthe root of these symptoms is widespread loss of neurons and their synapses primarily in the cere-bral cortex, entorhinal area, hippocampus, ventralstriatum, and basal forebrain [1–5]. Other patho-logical features that make their appearance in thebrain tissue include a variety of different kinds of amyloid deposits collectively called amyloidplaques (Fig. 13.1), persistent accumulations of abnormal tau filaments referred to as neurofibril-

lary tangles, dense focal deposits of fibrillar amyloidin the walls of certain blood vessels (mostly smallarterioles), intraneuronal accumulation of amyloid,reactive gliosis, and inflammation [1, 2, 6–9].

The presence of numerous amyloid plaques inAD brains has attracted great interest because theyappear relatively early in the course of the diseaseand thus provide a potential early therapeutic tar-get. These plaques consist of amyloid deposits,microglial cells, dystrophic neurites, and bundlesof astrocytic processes. A principal component of plaques in human brain is amyloid β (Aβ) peptide,especially A β (1-42) (A β42), a 42-amino-acid pep-

tide fragment derived from the sequential prote-

olytic cleavage of the amyloid precursor protein bybeta- and gamma-secretases [1, 10]. An enormousnumber of studies have implicated A β42 as a keyplayer in the observed neurodegenerative cascade,and many investigators believe that it may bedirectly responsible for the rampant synaptic andneuronal loss observed during the course of thisdisease [11, 12]. Exactly how the accumulation of this “toxic” peptide is linked to the observed cog-nitive and memory decline remains to be eluci-dated, but this is an area of intense research interestwith the hope of changing the long-term outcomeof this disease or, better yet, eradicating the diseasealtogether.

It is now well-recognized that glial cells (espe-cially astrocytes and microglia) play a critical,dynamic role in inflammatory and neurodegenerativeevents that occur in the brain during the course of AD. Traditionally, astrocytes were assigned the roleof filling tissue voids caused by degenerative events,a process called glial scar formation, whereas

microglia were presumed to function primarily asbrain phagocytes, responsible for the removal of Aβ deposits and debris from degenerating neuronsand their processes. More recently, as will be dis-cussed here, it has become apparent that there maybe some sharing of phagocytic responsibilityamong these cell types and that their contributionto events occuring in the brain is considerably morecomplex than previously thought. In this chapter,we highlight the responses of astrocytes andmicroglia to intraneuronal A β accumulation, neu-ronal and synaptic degeneration, and amyloidplaque formation and focus on how their responses



are intimately and irrevocably integrated into thefate of A β peptides and evolving pathology in ADbrains.

13.2 Astrocytes and the Fate of A βin AD Brains

13.2.1 Astrocytes: Structure andFunction in Normal Healthy Brain

Astrocytes, the predominant glial cell type foundin the gray matter of the human CNS, extendnumerous cytoplasmic processes that contain

abundant bundles of intermediate filaments com-posed mainly of glial fibrillary acidic protein(GFAP) (Figs. 13.2A and 13.2B) [13]. In eachastrocyte, the fine, highly branched tips of thesecytoplasmic processes generally lack GFAP andcan come into contact with thousands of localsynapses [14]. In addition to structurally and func-tionally isolating synapses from events in the sur-rounding brain tissue, astrocytic processes are nowthought to play an active role in synaptogenesis,the construction of neuronal circuits during devel-opment, synaptic stability and plasticity in theadult brain, ensuring normal neuronal excitabilityby maintaining extracellular ion homeostasis, andin clearing potassium from the region of synapses[15–19]. In addition, astrocytes are able to take up

the excitatory amino acid glutamate from thesynaptic cleft to levels up to 10,000 times higherthan that in the extracellular space, a function thatis pivotal for optimal gluataminergic neurotransmis-sion and avoiding neuronal excitotoxicity [20–24].The interchange of metabolites between astrocytesand between astrocytes and neurons is complexand is not well-understood, but gap junctions arenow thought to be critical for this function [17].

13. Glial Cells and A β Peptides in AD Pathogenesis 217

FIGURE 13.1. Section through entorhinal cortex of ADbrain immunostained with anti-A β42 antibody (ChemiconInternational), which does not show appreciable reactivitywith A β40 in ELISA or APP, showing amyloid plaques(AP) confined to the pyramidal cell layer (PCL). Activatedastrocytes (AA) in both the molecular layer (ML) andPCL contain substantial quantities of A β42-positive mate-rial. The A β42-positive material in ML astrocytes is pre-sumed to be derived from their active role in clearingdebris associated with local synaptic and dendritic loss,which is rampant in this layer. Dendritic and synaptic lossin the ML appears to be temporally linked to the accumu-lation of A β42-positive material by the parent neurons inthe underlying PCLs.

FIGURE 13.2. (A) Section through entorhinal cortex of AD brain immunostained for glial fibrillary acidic pro-tein (GFAP) showing prominent GFAP-rich activatedastrocytes in the molecular layer (ML). PCL, pyramidalcell layer. (B) Higher magnification of similar region

showing the GFAP-rich processes of activated astrocytes(AA) with their associated end-feet in contact with thewall of a blood vessel (BV).



Lastly, the end feet of astrocytic processes encapsu-late brain capillaries that pass through the braintissue (Fig. 13.2B), most likely providing additionalstructural support for the blood-brain barrier andparticipating in regulation of the exchange betweenthe smaller blood vessels and the surrounding braintissue [25–27].

13.2.2 “Activation” of Astrocytes inResponse to Local AD PathologyCompromises Astrocytic Function

In addition to playing a critical role in the functionsdescribed above, all of which are ultimately devotedto the maintenance of normal neuronal activity,astrocytes are capable of responding to pathologi-

cal situations, where they engage in a series of structural and functional changes collectivelyreferred to as “activation,” “reactive astrogliosis,”or “astrocytosis” [28–32]. These “activated astro-cytes” exhibit a pronounced enlargement of theircell bodies and a dramatic thickening and lengthen-ing of their cytoplasmic processes. They are readilyidentifiable in regions of CNS trauma, hypoxia, andin many neurodegenerative conditions by virtue of the dramatically elevated expression of glial fibril-lary acidic protein (GFAP), vimentin, and nestin intheir cell bodies and in the main trunks andbranches of their cytoplasmic processes, comparedwith their more quiescent counterparts [13, 14, 28,33] (Figs. 13.2A and 13.2B). Unfortunately, thesechanges come with a price—“activation” forcesastrocytes to give up many of the activities men-tioned above that were essential for normal neu-ronal function. Physiological functions such as thebuffering of neuronally released potassium andglutamate from the extracellular space may be

impaired, favoring local nerve cell depolarization,excessive Ca 2+ influx, and excitotoxic damage toneurons [18, 34–35]. In addition, retraction of astrocytic end-feet and processes from synapsesand the walls of local blood vessels may jeopardizethe integrity of synapses and the local blood-brainbarrier. Thus, although astrocyte activation nodoubt is intended to be a protective response in thenormal day to day activities in the brain, the intenseand widespread astrocyte activation seen through-out AD brains may also exacerbate the extent of neuronal damage and even accelerate the rate of disease progression [36].

13.2.3 Astrocyte “Activation”Compromises the Blood-Brain Barrier,Leading to Leakage of Blood-BorneSubstances, Including Soluble A β42,into Brain Tissue

The blood-brain barrier (BBB) is a diffusion barrierthat blocks the movement of blood-borne sub-stances into the brain parenchyma [37]. The threemain components of the BBB are endothelial cells,the end-feet of astrocytes, and pericytes. Tight

junctions between the endothelial cells in cerebralvessels are thought to provide the structural basisfor the seal. Astrocyte end-feet tightly ensheath thevessel wall and most likely lend additional stabilityto the integrity of the barrier (Fig. 13.2B).

Activation of astrocytes causes them to pull awaymany of their processes from the walls of bloodvessels. The loss of astrocyte-endothelial cell con-tact can lead to breakdown of the BBB, resulting inan efflux of serum components into the brain tis-sue. Studies have shown that a significant pool of Aβ exists in the peripheral circulation [38–40].Because the breakdown of the BBB is unlikely tooccur uniformly throughout the brain, regionsshowing such leaks also exhibit increased levels of plasma components including serum immunoglob-

ulin, complement, and A β42 [R. Nagele, unpub-lished observations]. Leakage of these componentsinto AD brains can often be detected in AD brainas immunopositive perivascular “leak clouds” that,unexpectedly, are most often associated with smallarterioles rather than capillaries within the brainparenchyma (Fig. 13.3). Elevated levels of thesesubstances in the brain tissue may play an impor-tant role in the development of AD pathology asdescribed in more detail below and could conceiv-

ably explain the frequent observation of “hot spots”of AD pathology, especially in the brains of patients that are early in the course of the disease.

13.2.4 Activated Astrocytes AccumulateAβ42 in AD Brains

In early AD pathology, activated astrocytes areconspicuous in two regions: in the molecular layerof the cerebral cortex and in the immediate vicinityof amyloid plaques in the underlying pyramidalcell layers (Figs. 13.1 and 13.4A). What triggersthese cells to become activated in response to AD-

218 G. Siu et al.



related pathological changes is not clear, but invitro studies have shown that aggregated A β andthe cores of amyloid plaques isolated from humanAD brain tissue are effective in stimulating astro-cyte activation [41]. Once activated, these cells arecapable of internalizing and degrading A β42, sug-gesting that they may play a direct role in its clear-ance from the brain parenchyma. In support of thispossibility, activated astrocytes in AD brains posi-tioned in the cortical molecular layer as well asthose closely association with neuritic or dense-core plaques in the underlying pyramidal cell lay-

ers can accumulate substantial amounts of A β42(Figs. 13.1 and 13.4A) [42–46]. In the corticalpyramidal cell layers, astrocytes stationed outsideof amyloid plaques, just beyond the outer edge of the A β42-rich corona, extend thick, intenselyGFAP-immunopositive, cytoplasmic processes thatenvelop the amyloid plaque and thinner (mostlyGFAP negative) branches from these processes thatinfiltrate deep into the plaque interior. In additionto intense GFAP immunostaining, these cells oftenshow impressive intracellular accumulations of Aβ42-immunopositive material, suggesting thatthey are capable of internalizing A β42 via their

processes and transporting it back to the cell body,presumably for degradation within the lysosomalcompartment. In fact, most A β42-immunopositivematerial within astrocytes localizes to prominentgranules in the perinuclear cytoplasm, and thesegranules have the same distribution and size asthose that immunostain with antibodies specific forcathepsin D (Fig. 13.4B) [44].

13.2.5 The Amount of A β42 in ActivatedAstrocytes Is Linked to the LocalAbundance of Neurons ContainingSubstantial Intracellular A β42 Deposits

The amount of A β42-positive material contained

within activated astrocytes is not uniform through-out the cerebral cortex of AD brains but ratherappears to be both spatially and temporally corre-lated with the extent of local AD pathology [44]. Inthe pyramidal cell layers, the A β42 content withinindividual astrocytes is proportional to the amountof intracellular A β42-positive material withinnearby neurons as well as the presence and localdensity of plaques (Fig. 13.1). By contrast, corticalmolecular layer astrocytes contain abundant A β42-positive material despite the fact that this layer gen-erally lacks A β42-burdened neurons and plaques,especially in the early stages of AD pathogenesis


FIGURE 13.3. Section through the entorhinal cortex of anAD brain immunostained for A β42 showing and A β42-rich perivascular “leak cloud” surrounding a small bloodvessel (BV) (arteriole). These leak clouds are observedpreferentially around small arterioles and are only seenaround brain capillaries in regions showing advancedpathology and well-developed inflammation. AP, amy-loid plaque.

FIGURE 13.4. (A) Section through AD cortex immunos-tained with anti-A β42 antibodies showing large A β42-rich deposits in activated astrocytes (AA) in the

molecular layer (ML). These same cells also exhibitintense cathepsin D (Sigma) immunoreactivity, suggest-ing increased activity of their lysosomal compartment.PC, pyramidal cells; PCL, pyramidal cell layer.



(Fig. 13.1). Interestingly, the amount of A β42-positive material within these astrocytes correlatesclosely with the severity of pathology exhibited bythe pyramidal cell layers lying directly under thislayer. In brain regions where pyramidal cells lack significant intracellular A β42 deposits, most of theoverlying molecular layer astrocytes are quiescentand generally devoid of A β42-positive material[44]. Taken together, these observations emphasizethe temporal and spatial link between A β42 accu-mulation in pyramidal neurons and the appearanceof similar intracellular deposits in the overlyingmolecular layer astrocytes.

13.2.6 Activated Astrocytes AccumulateAβ42 While Clearing the Products of Neuronal and Synaptic Degenerationand Loss

The source of the A β42 and the mechanism bywhich it accumulates selectively in activated astro-cytes and not in their more quiescent counterpartsremains to be determined. Expression of the amy-loid precursor protein is either extremely low ornonexistent in astrocytes, thus internal productionis unlikely to be a major source of the A β42 thataccumulates in these cells. By contrast, exogenous(soluble) A β42 from the surrounding extracellularfluid is a much more likely source, and its accumu-lation in astrocytes could occur via receptor-medi-ated endocytosis and/or phagocytosis. In support of this possibility, the phagocytic capability of acti-vated astrocytes has already been demonstrated andincludes the removal of local synaptic material[47]. In addition, our previous study has providedstrong evidence that most (possibly all) of the accu-mulated A β42 within activated astrocytes posi-

tioned in the cortical molecular layer is of neuronalorigin and is derived from internalization of degen-erating synapses and dendrites belonging to neu-rons in the underlying pyramidal cell layers [44].Further evidence for this mode of astrocytic A β42accumulation comes from the fact that A β42 in acti-vated astrocytes colocalizes with other neuron-spe-cific proteins, including choline acetyltransferase(ChAT) and the alpha7 nicotinic acetylcholinereceptor ( α7nAChR) (Fig. 13.5A), neither of whichis synthesized by astrocytes [44]. The selectiveaccumulation of these neuronal proteins and A β42

in activated astrocytes is an expected consequenceof their debris-clearing activity in response to ele-vated levels of AD-related degeneration of localdendrites and synapses. The fact that accumulatedChAT- and α7nAChR-immunopositive material ismost prominent in astrocytes populating the corti-cal molecular layer is a reflection of the abundanceof synapses containing these proteins in this region[44]. Studies using electron microscopy have shown

that the corona of dense-core amyloid plaques inthe pyramidal cell layers and the amyloid aggre-gates associated with capillaries are extensivelyinfiltrated with astrocytic processes in both humanAD and APP tg mouse brains [48–50]. A β depositscan apparently be degraded by metalloproteases,including neprilysin and insulysin [51, 52], andneprilysin has been localized in astrocytes closelyassociated with amyloid plaques, suggesting thatthey possess the requisite elements for A β degra-dation [53]. In view of the above, the idea thatastrocytes may not become phagocytic until the

220 G. Siu et al.

FIGURE 13.5. (A) Section through entorhinal cortex of AD brain immunostained with rabbit polyclonal antibod-ies directed against the alpha7 nicotinic acetylcholinereceptor (alpha7) (Santa Cruz Biotechnology, sc-1447,raised against amino acids 367–502 mapping at the C-

terminus of human a7nAChR). Activated astrocytes(AA) in the cortical molecular layer are stronglyimmunopositive for alpha7. Alpha7 accumulation inthese cells is a by-product of their action in clearing localdendritic and synaptic debris. Confirmation of the speci-ficity of this antibody was obtained by Western blotanalysis and deletion of staining by preabsorption withthe immunogen peptide. (B) Death and lysis of A β42-overburdened, activated astrocytes leads to the formationof small astrocytic amyloid plaques (AP) in the corticalmolecular layer that are both A β42- and alpha7-immunopositive.



phagocytic capacity of brain microglia has becomesaturated [54] may have to be discarded. In fact,the reverse seems more likely—that microglia arenot activated until after the phagocytic activity of astrocytes is overwhelmed or, at least, sufficientlytaxed above some unknown threshold level.

13.2.7 Effects of Intracellular A β42Accumulation on the FunctionalActivity of Astrocytes

It is not known whether A β42-burdened, activatedastrocytes are capable of clearing internalized andaccumulated A β42. The fact that the total astro-cytic amyloid burden seems to increase in ADbrains with the degree of AD pathology suggests

that astrocytes are either not capable of clearinginternalized A β42 or that their clearance mecha-nism may be deficient. The effects of gradual intra-cellular A β42 accumulation on the functionalactivity of astrocytes is unknown, but it is likely tohave a progressively deleterious effect on thesecells throughout the accumulation process, eventu-ally ending in cell death and lysis. As mentionedabove, astrocytes are known to make contacts withmultiple neurons in their immediate vicinity. Thisposition between neurons allows astrocytes tofacilitate information transfer between neighboringneurons and other astrocytes, maintain neuronalexcitability by keeping close control over ionhomeostasis, and may contribute to synaptic plas-ticity [15, 16, 20, 55–57]. Recent work has led to anew appreciation of the active role of astrocytesand astrocyte-derived cytokines in the response toinjury and repair and their influence on theintegrity of the blood brain barrier [53, 58, 59].Degeneration of cortical dendrites and synapses in

AD brains may stimulate the conversion of “quies-cent” to “activated” astrocytes [31]. Such degener-ation would result in a severing of astrocyte-neuroncontacts, which may itself provide a signal for acti-vation of astrocytes, the clearing of local neuronaldebris, and, thus, drive the accumulation of neuron-derived materials, including A β42, in these cells.Another consequence is impairment of astrocyte-maintained extracellular ion homeostasis, whichfavors excitotoxic neuronal damage [32]. It ispossible that, as in many otherwise protectiveprocesses, this may get out of hand by favoring

oxidative neuronal damage and enhanced A β toxi-city, thus providing a therapeutic target to possiblyslow it down [31].

13.2.8 A β42-Overburdened Astrocytes

Can Undergo Lysis to Form Astrocyte-Derived Amyloid Plaques

The progressive and extensive synaptic loss in thecortical molecular layer appears to graduallyincrease the intracellular load of A β42-immunopositive material that has accumulated inlocal activated astrocytes (Fig. 13.1). We haveshown that this increased load is eventually accom-panied by the appearance of a new population of amyloid plaques within the cortical molecular layer

(Fig. 13.5B). This new population of plaquesappears to be derived from the death and lysis of Aβ42-overburdened astrocytes [44]. Upon lysis,cytoplasmic material from ruptured astrocytes isdispersed somewhat radially, including their con-tent of accumulated A β42. This dispersion may ini-tially be facilitated by the action of lysosomalenzymes that are also released at that time. Celllysis leaves in its wake a persistent, roughly spher-ical, A β42-rich residue that takes the form of a dis-tinctive population of amyloid plaques. That theseplaques are derived from the lysis of astrocytes isbolstered by the fact that they first appear in thesubpial portion of the cortical molecular layer andare observed only in regions where nearby astro-cytes contain large intracellular deposits of A β42-positive material (Fig. 13.1B) [44]. This proposedmode of “astrocytic” plaque formation is nearlyidentical to that which has been described previ-ously for the larger, spherical, neuron-derivedplaques that populate the underlying pyramidal cell

layers, many of which appear to be the lysis rem-nants of A β42-overburdened neurons [10, 60].Although both types of plaque are A β42-immunopositive, astrocytic plaques are readily dis-tinguished from neuron-derived plaques because of their location, much smaller size, and particularlyintense GFAP-immunoreactivity. The consistentspherical shape of most plaques (Fig. 13.1) and theclose relationship between the size of both neuron-and astrocyte-derived amyloid plaques and thecells from which they are presumably derivedargue strongly against proposed mechanisms for




amyloid plaque formation that describe the gradualgrowth of plaques from a seeding site or “nidus,” atleast for this morphological subset of plaques.

13.2.9 Astrocyte Activation May Be

Triggered by the IntraneuronalAccumulation of A β42 in AD Brains

The formation of large intracellular deposits of Aβ42 have been reported in several types of neu-rons in the cerebral cortex and cerebellum of ADand Down syndrome brains (Fig. 13.6) [8, 44,61–64]. Our recent studies suggest that their abilityto do so may be linked to neuronal expression of thealpha7 nicotinic acetylcholine receptor ( α7nAChR)[60]. Previous studies have shown that A β42 binds

with exceptionally high affinity to α7nAChRs onneuronal surfaces [63–64]. As described above, theleak of serum A β42 into the brain parenchymathrough local breaches in the BBB (cf. Fig. 13.3)would be expected to provide a constant source of exogenous A β42 to local neurons. Thus, neuronsthat are particularly well-endowed with α7nAChRs(e.g., cortical pyramidal cells) would form rela-tively high levels of A β42/ α7nAChR complex ontheir surfaces. It follows then that any membranerecycling or endocytic activity on the part of theneurons would tend to drive the internalization of Aβ42/ α7nAChR complex into neurons and target

this complex to the lysosomal compartment.Consistent with this mechanism, A β42 and theα7nAChR are invariably colocalized within intra-neuronal deposits in AD brains, and these depositsare also immunopositive for cathepsins, confirmingthat this accumulation occurs within the lysosomalcompartment [60]. We have suggested that the bind-ing of “exogenous” A β42 to the α7nAChR-bearingdendrite trees of neurons may not only facilitate inter-nalization and accumulation of A β42 in these cellsvia endocytosis but also provides a plausible expla-nation for the well-known selective vulnerability of cholinergic and cholinoceptive neurons to ADpathogenesis [60].

The accumulation of A β42/ α7nAChR complexin cortical pyramidal neurons is one of the earliest

signs of developing AD pathology, and work ontransgenic mice has temporally linked this eventwith early synaptic degeneration and loss [44,66–68]. It is likely that these events are alsodirectly linked to the observed early activation of astrocytes in the cortical molecular layer. Thislayer is densely packed with the fine, α 7nAChR-rich dendrite branches that extend from the maindendrite trunks of neurons positioned in the pyram-idal cell layers lying directly below. We have sug-gested that excessive accumulation of A β42 inneurons (Fig. 13.6) impairs the ability of these cellsto maintain their extensive dendritic arbors. If thisis the case, then the most distal dendrite branchesand their associated synapses, located in the corti-cal molecular layer, would be most vulnerable todegeneration and loss, which is consistent withwhat is observed. If A β42/ α 7nAChR complex ispresent on degenerating dendrites and synapses,clearing of this debris by local astrocytes viaphagocytosis/endocytosis and the targeting of this

material to the lysosomal compartment wouldexplain the source of A β42 seen in these cells(Fig. 13.5A). In addition, it would explain whyother neuron-specific proteins, such as α7nAChRand choline acetyl transferase (ChAT), are alsocolocalized within A β42-immunopositive depositsof astrocytes [44]. Thus, “activated” astrocytes arecapable of internalizing neuron-derived materials,including surface-bound A β42, presumablythrough their endocytic/phagocytic activity and, asin neurons, this activity is paralleled by a dramaticelevation of lysosomal cathepsin D levels [44]. Thegreat affinity of A β42, but not A β40, for the

222 G. Siu et al.

FIGURE 13.6. Section through the entorhinal cortex of an

AD brain immunostained with anti-A β42 antibodiesshowing A β42 localized to amyloid plaques (AP) andlarge intracellular deposits within pyramidal neurons (N).



α7nAChR also provides a straightforward explana-tion for A β42 as the dominant A β peptide speciesin astrocytic intracellular deposits and in amyloidplaques throughout AD brains [60]. The proposedmechanism described above pinpoints a few vari-ables that may dictate variations in both the natureof the pathology and rate at which it evolvesin individual AD patients. These variables couldinclude the serum levels of A β42, the location(s) of the breach in the BBB, whether the breach is focalor global, and whether the breach is sufficient toallow passage of materials from the blood into thebrain that could contribute to AD pathology (e.g.,Aβ42, immunoglobulin, and complement).

13.3 Microglia and the Fate ofAβ in AD Brains

Microglia are resident cells of monocyte-phagocytelineage in the brain that, when activated, are capa-ble of phagocytosis and participating in immuneresponses by presenting antigens to invadingimmune cells. In the normal healthy brain, they arereferred to as “resting microglia” and are widelyscattered, seeming to occupy their own individualdefined territory within the brain parenchyma. Thefunction of these cells in the resting state isunknown. However, in response to pathologicalchanges in the brain tissue, microglia can rapidlytransition to an activated state (Fig. 13.7A). In theactivated state, these cells take on a more amoeboidcharacter and migrate to the site of injury, wherethey can proliferate, launch a phagocytic attack onthe offending material including tissue debris, andrelease inflammatory mediators such as cytokinesinto the surrounding tissue [69–74]. Much of what

we know about the activity of microglia has beenderived from studies on the actions of these cells inthe culture environment. In cell cultures, microgliashow increased cell surface expression of MHCII[75], a classic marker for activated microglia, aswell as an increased secretion of inflammatorycytokines such as interleukin-1B (Il-1B), inter-leukin-6 (IL-6), and tumor necrosis factor- α (TNF-α ), and chemokines such as interleukin-8 (IL-8),macrophage inflammatory protein-1 α (MIP-1 α ),and monocyte chemoattractant peptide-1 (MCP-1)[76]. In addition, mRNAs encoding C1q, C3, C4,IL-1 receptor antagonist, and transforming growth

factor- β (TGF- β) have been detected in ADmicroglia [77–80]. Where the tissue devastation isparticularly great, brain microglia intermingle withadditional monocytic cells that appear to migrateinto the brain tissue from the blood (Fig. 13.7B).At this point, it is often difficult to distinguishmicroglia from these immigrant macrophages and,for this reason, it is probably best to refer to themas microglia/macrophages. The precise identityand nature of the signals that cause the initial acti-vation of microglia that are resident in the brain areunknown.

13.3.1 Relationship Between thePhagocytic Activity of Microglia andAβ in AD Brains

In AD brains, activated microglia are widely distrib-uted throughout the brain parenchyma but are alsofocally concentrated within amyloid plaques wherethey are generally thought to be actively engaged inthe clearance of A β from the plaque interior viaphagocytosis [30, 55, 56, 70, 71, 81–89]. In culture,microglia derived from AD brains are not only ableto congregate around aggregated A β deposits, but


FIGURE 13.7. (A) Section through entorhinal cortex of AD brain double-immunostained with anti-A β42 anti-bodies and HLA-DR antibodies to immunolabel acti-vated microglia. Note the strong tendency for activatedmicroglia to congregate at the exact center of the amy-loid plaque (AP), a region known to contain a neuronalnuclear remnant and other debris associated with neu-ronal lysis. (B) Section through the pyramidal cell layerimmunostained with anti-HLA-DR antibodies showingmicroglia/macrophages (M), some of which appear to bein the process of entering into the brain tissue from localsmall blood vessels (BV).



they appear to be able to remove these depositsover a period of 2–4 weeks [90]. In addition, theintracellular accumulation of A β occurs more rap-idly and to a greater extent in these cells whenserum is added to the culture medium, suggestingthat serum contains some factor(s) that facilitatesAβ endocytosis [55]. Microglia applied to unfixedbrain sections in culture reportedly phagocytoseAβ deposits when anti-A β antibodies are includedin the culture medium, suggesting that opsoniza-tion of the A β facilitates this activity [55, 82]. TheAβ is subsequently found in phagosome-like intra-cellular vesicles [82].

Unfortunately, studies on the activities of microglia in the context of the AD brain have beenless revealing. Although ultrastructural studies

have reported that microglia in the AD cortex con-tain some intracytoplasmic A β fibrils, it is not dra-matic, and there have been reports to the contrary[85, 91, 92]. One possible explanation for thisapparent discrepancy is that microglia mightprocess internalized A β so rapidly that little of thismaterial can be demonstrated in a cell at any par-ticular time. Of course, another possibility is thatAβ internalization by microglia is a culture anom-aly and that they do not internalize A β at all in thebrain. If the latter proves to be true, we are still leftwithout assigning a definitive function to themicroglia that are stationed within amyloidplaques. In contrast to a role in the clearance of A βfrom plaques and the brain, it has also been sug-gested that microglia participate in the conversionof soluble or oligomeric A β into polymerized amy-loid fibrils in the parenchyma, within plaques andin the walls of blood vessels [85]. This idea isbased on the observation that plaque-associatedmicroglia display dilated intracellular channels of

endoplasmic reticulum that appear to contain amy-loid fibers [91, 92]. Also, largely because of theirlocation within plaques, the actions of plaque-asso-ciated microglia have been postulated to play a rolein the reported transformation of diffuse amyloidplaques into neuritic or dense-core A β plaques.However, this role seems to be unlikely in view of the fact that microglia are generally not found inassociation with diffuse plaques but rather clearlyprefer to congregate at the central portions of dense-core plaques in both AD brains and thebrains of APP-overexpressing transgenic mice [93](Fig. 13.7A). In addition, it has not yet been deter-

mined whether one morphological type of plaquecan evolve into another or whether they representdifferent plaque types with unique origins. Thegeneral lack of an obvious, well-defined functionfor plaque-associated microglia that is related toeither A β clearance or deposition inevitably leadsone to consider the possibility that their presencewithin plaques may have nothing at all to do withAβ clearance in the brain.

13.3.2 Microglial Chemotaxis: A β orDNA Fragments as Chemoattractants

What lures microglia to amyloid plaques isunknown, but their preferential association withdense-core plaques as well as the tendency for them

to be positioned at or near the dense core of plaquessuggests that there is something either at or emanat-ing from the plaque core that is strongly chemotac-tic to microglia. In elegant studies carried out byRogers and co-workers on cultured microglia origi-nally isolated from the brains of both AD and non-demented patients, these cells exhibited obviouschemotaxis to preaggregated A β42 deposits thatwere adherent to the culture substratum [94, 95] butwere not attracted to A β42 scrambled sequence[96]. It has been reported that A β can bind to sev-eral types of microglial cell surface receptors,including RAGE [97]. Although they have provideda wealth of information on the phagocytic actions of microglia, cultured microglia models also havesome limitations that raise questions about howaccurately and directly the actions of these cells inculture reflect those of their counterparts in the con-text of the brain. One obvious limitation is that theresponses of cultured microglia to various testagents or conditions are occurring in an artificial

environment that lacks their usual interactions withneurons, neuronal processes, astrocytes, and ele-ments of the local blood vasculature. Another limi-tation is that the culture environment alone issufficient to activate microglia, which makes it dif-ficult to determine the identity of factors that caneither induce or influence the activation state.Lastly, compared with what happens in a slowlyevolving disease state such as AD, studies on cul-tured microglia are of very short duration.

Direct extrapolation of the results on chemotaxisobtained from studies on cultured microglia to theactions of microglia in vivo does not seem to fit

224 G. Siu et al.



well with the apparent behavior of microglia andtheir response to A β peptides in the AD brain. Forexample, if A β42 is chemotactic to microglia inAD brains, one would expect to see abundantmicroglia near and within all types of A β42-containing plaques. Contrary to this expectation,microglia are generally not found either within orassociated with diffuse plaques, which containabundant A β42. In addition, these cells apparentlypass through the A β42-rich outer corona of dense-core plaques and take up residence preferentially atthe plaque core (Fig. 13.7A), which is also rich inAβ42. Together, these observations suggest that, inAD brains, something at or within the core of dense-core plaques is highly chemotactic to acti-vated microglia. One likely chemotactic factor is

DNA fragments. Microglia have been shown toaccumulate damaged DNA fragments in AD brain,and fragmented DNA has been suggested as apotent promoter of microglial activation [98]. Insupport of this possibility, our previous studieshave provided strong evidence that many (possiblyall) dense-core plaques in the pyramidal cell layersof the cerebral cortex are derived from the lysis of Aβ42-overburdened neurons. Neuronal lysisreleases the contents of the neuronal perikaryon,including A β42 and lysosomes. The local releaseof lysosomal enzymes probably facilitates theradial diffusion of neuron-derived A β peptide,which explains both the generally spherical shapeof all plaques as well as the fact that their individ-ual sizes seems to correlate with the size of localneurons (Fig. 13.6) [8, 44, 60]. Another conse-quence of neuronal lysis is the persistent presenceof a nuclear remnant at the core of the dense-coreplaque [8]. Here, we propose that the gradualdegradation of this nuclear material releases DNA

fragments that diffuse out from the plaque core intothe surrounding brain parenchyma. Becausemicroglia are capable of responding to DNA frag-ments, it is reasonable to suppose that the release of these fragments is chemotactic to microglia, draw-ing them ever closer to the source of the DNA posi-tioned at the plaque core (Fig. 13.7A). In addition,peripheral monocytes are often observed emigrat-ing from local small blood vessels into regionswhere dense-core plaques are nearby or adjacent,which is not observed in brain regions containingonly diffuse plaques [66, 99, 100] (Fig. 13.7B). Infact, it is entirely possible that most of the so-called

microglia/macrophages seen within dense-core amy-loid plaques in AD brains are immigrants from theblood and that the involvement of resting/residentmicroglia in the formation/evolution and eventualclearance of A β42 and plaques is minimal. Thepracticality of DNA fragments serving as the prin-cipal chemotactic signal attracting local microgliaand moncytic cells from local blood vessels isobvious because its release into the local milieu canonly occur via local cell death, thus making it anunambiguous marker indicating that local cellulardegeneration and death has actually occurred.

13.3.3 Mediators of MicroglialPhagocytosis

There are likely to be multiple mediators of microglial activation, chemotaxis, and phagocyticactivity in the brain, and some of these may dependon the nature of the pathology that develops inassociation with specific brain diseases. Theformyl peptide receptor (FPR), the macrophagescavenger receptors (MSR) [101], and the receptorfor advanced glycation end products (RAGE) areexpressed by microglia, have opsin-independentactivity, and appear to have A β as a ligand [102,103]. Microglia also express the complementopsonin receptors CR3 and CR4 and the anaphyla-toxins C3a and C5a [104–107]. Complement iswell-known to facilitate the phagocytosis of tissuedebris, and there is some evidence that complementcan opsonize A β fibrils, facilitating their removalby microglial phagocytosis. The well-known path-way for complement activation is initiated with theattachment of C1q to a target, its interaction with anumber of proteases (including C1r, C1s, C4, C2,C3) followed by the attachment of C4b and C3b,

which act as ligands for complement receptors onmicroglia and other phagocytic cells [108]. Whencompleted, complement terminal components(C5b–C9) are assembled into the membrane attack complex. Complement activation and opsonizationof fibrillar A β by C1q in amyloid plaques has beendemonstrated in AD brains [109–111]. The diffi-culties mentioned above in detecting significantamounts of phagocytosed A β within brainmicroglia raise a question as to the relevance of opsonization of A β fibrils within plaques. If thiswere, in fact, a driving influence for A β-mediatedmicroglial chemotaxis and the phagocytic activity




of these cells, it fails to explain why such microgliaare not found in association with A β42-rich diffuseplaques. Perhaps the idea of opsonization-enhancedphagocytosis is correct except for what is beingopsonized. The lack of microglia in diffuse plaquesand the preferential localization of microglia at thecore of dense-core plaques suggest that the opsonizedmaterial is located exclusively at the core of dense-core plaques.

13.3.4 Positive and Negative Aspectsof Microglial Activity in AD Brain

The intent of inflammation is to allow a series of specific cellular events to occur that will ultimatelyresult in the removal of the offending agent and its

associated cell and tissue debris from the affectedtissue, leaving the way open for either tissue repairor replacement (scar formation). The processseems to work well in instances where there areclear limits to the amount of offending agent andthe extent of tissue destruction caused by this agentand in situations where the vascularity of the tissuecan be restored. In this case, elimination of theoffending agent and tissue debris largely by phago-cytic activity can then be followed by a period of tissue repair or replacement without additionalinsults. On the other hand, this process does notseem to work well in cases of chronic diseases suchas AD, where the offending agent (presumably A β)is constantly supplied throughout the course of thedisease, leaving little opportunity for repair tooccur in an environment free of additional insultsand progressive tissue destruction. Unfortunately,AD seems to be one of those diseases where therate of tissue destruction exceeds the capacity of local cells (astrocytes and microglia) to resolve it.

Inevitably, such conditions lead to the recruitmentof additional cells (e.g., blood-borne monocyticcells) to the site of damage. When the brain tissuebecomes heavily populated with inflammatorycells (Figs. 13.8A and 13.8B), the additional pro-duction of unusually high levels of inflammatorymediators and the excessive phagocytic activity of these immigrant cells becomes more destructivethan beneficial. Thus, in AD, the chronic andprogressive nature of the disease eventually tipsthe balance of the resulting inflammation tothe destructive side, leading to the loss of irre-placeable neurons.

13.3.5 Microglia as Therapeutic Targets

As detailed above, microglia have been assigned arole in the inflammatory response associated withAD pathology and also possibly with the process-ing and/or clearance of A β from the brain. Theconcept that runaway inflammation in the brainmay actually precipitate some of the observed neu-rodegeneration in AD has raised the possibility thatat least some of this damage may be avoided oralleviated through the use of nonsteroidal anti-inflammatory drugs (NSAIDs). The results of anumber of clinical trails using NSAIDs, with someclaiming a reduced incidence of AD, have beensomewhat less than convincing [112–120]. Part of the problem may be that the levels of brain inflam-mation at the time the patient enters into the clini-cal trial may be too advanced. Another possibilityproposed by Streit and co-workers is that microgliain the AD brain show a loss or deterioration of function that may represent a type of cellular

senescence [121, 122]. If this is the case, then thecollective phagocytic capability of microglia/ macrophages in the brain both before and aftertreatment would be insufficient to keep up with therate of tissue destruction. This could explain themarginal benefit of NSAIDs for AD.

In the past few years, great attention has beengiven to the possibility that immune stimulation byvaccination with A β peptides (especially A β42)would lead to the production of anti-A β peptideantibodies. From the therapeutic standpoint, thehope is that this vaccination will ultimately resultin microglia/macrophages becoming more efficient

226 G. Siu et al.

FIGURE 13.8. Consecutive sections through the entorhinalcortex of an AD brain immunostained with anti-A β42(A) and anti-HLA-DR (B) antibodies. The brain tissueshows considerable inflammation with microglia/ macrophages occurring both individually (in the spacebetween plaques) and in clusters (within plaques).



at phagocytosing amyloid deposits, which are con-sidered by many to be the direct or indirect causeof the neurodegeneration that is associated withthis disease. Some success with this approach hasbeen reported in animal models of AD, where anti-bodies generated against A β42 caused a reductionin the amyloid load in the brain of transgenic mice[123–125]. In these experiments, the clearance of amyloid fibrils from the brain parenchyma wasdetermined to occur by the binding of A β42/ immunoglobulin complexes to immunoglobulinFc receptors on microglia/macrophages, whichenhanced the rate and extent of phagocytosis of these complexes. On the other hand, results of clin-ical trials with humans have not been encouraging,and the development of encephalitis has been prob-

lematic. Several potential problems with thisapproach are predictable and noteworthy. First, theability of anti-A β antibody to bind to anything inthe brain requires that the BBB not be intact, sothat the induced immunoglobulin can enter into thebrain from the blood. A question arises as towhether the long-term, global breach in the BBBcan ever be repaired in AD brains, even if the amy-loid load of the brain is successfully lowered.Second, as mentioned above, A β42 has great affin-ity for the α7nAChR, which is abundantly presenton the surfaces of many types of neurons through-out the brain. Thus, because A β42 is also able toenter into the brain from the blood, many neuronsin AD brains at the time of treatment will possessAβ42/ α 7nAChR complexes on their surfaces,which, of course, will be immunoreactive to theincoming anti-A β42 antibodies. In addition toinducing the formation of cell surface patches of aggregated anti-A β42-A β42/ α 7nAChR complex,this may prompt stripping of these complexes from

the cell surface via endocytosis. The net effect isthat binding of the anti-A β42 antibody to neuronalsurfaces could actually accelerate the rate of A β42internalization and accumulation within neurons.Another potential negative effect of the binding of anti-A β42 antibodies to neuronal surfaces is that itattracts complement (including the membraneattack complex), which can promote neuron degen-eration and death. Because accelerated neuronaldegeneration and death would be expected to elicitan enhanced inflammatory response, it is not sur-prising that the vaccination approach runs the risk of global brain inflammation.

13.4 Perspectives

The combined activities of astrocytes andmicroglia/macrophages eventually become delete-rious and make a major and direct contribution toevolution of AD pathology in the brain. Evaluationof recent data in the context of what is alreadyknown about these two important cell types and theformation of amyloid plaques has allowed us toconstruct a proposed pathological sequence thathighlights the entangled interactions of A β andthese cells and their involvement in the pathogene-sis of AD (Fig. 13.9). A key starting point for ADappears to be the focal or global compromise of theBBB. Of course, this can happen in association

with any head or brain trauma but can also evolveas a result of aging-associated changes in the wallsof blood vessels. The requirement for this step mayexplain why aging seems to be a prerequisite forone to express AD symptoms and pathology. Thechronic leak of serum-bound A β42 into the braintissue through the defective BBB provides a con-stant supply of exogenous A β42 that can bind withhigh affinity to neurons (especially cortical pyram-idal cells) abundantly endowed with α7nAChR.For unknown reasons, neurons begin to internalizeAβ42/ α7nAChR complex via endocytosis. Onceneurons have accumulated sufficient A β42-positivematerial to elicit distal synaptic and dendritic loss,first in the cortical molecular layer, local astrocytesare activated and begin to internalize the resultingneuronal debris, which includes neuron-specificproteins such as α7nAChR, ChAT and A β42 [44].Aβ42-overburdened neurons and astrocytes even-tually die and undergo lysis, releasing their contentof A β42-positive material [8, 44, 60]. The material

released by cell lysis is dispersed radially with theaid of the activity of released lysosomal enzymes,leading to the formation of a spherical depositionof cell residue in the form of a plaque. Both smallerastrocytic plaques and larger neuron-derivedplaques are rapidly infiltrated with macrophages/ microglia, many of which are derived from bloodmonocytes that immigrate into the brain parenchymafrom local capillaries. The lack of microglia/ macrophages in diffuse plaques and their directmigration through the A β42-rich corona and intothe cores of dense-core plaques suggest that DNAfragments gradually released from the nuclear




228 G. Siu et al.

Degeneratingdendrites

Role of Astrocytes and Microglia in Plaque Formation

Neuron (N1) Neuron (N2)

Astrocyte (A1) Soluble A β42 enters brain from blood

Aβ42 binds to neurons via α7nAChR

Aβ42 accumulates in neuron N1

Aβ42 displaces N1 cyoplasmic machinery

Reduced material support of N1 dendrites

Start of N1 distal dendrite degeneration

Astrocyte A1 begins to clear N1 debris

Start of N1-derived A β accumulation in A1

N1 perikaryon heavily loaded with A β42

Failed material support of N1 dendrites

Complete collapse of N1 dendrites

A1 "activated" to debris-clearing role

A1 retracts processes from N2 dendrites

A1 retracts processes from capillary

Accelerated A β42 accumulation in A1

Stability of N2 synapses compromised

Start of N2 distal dendrite degeneration

Start of A β42 accumulation in N2

Accumulation of N2 debris by A1

Death and lysis of N1 and A1

N1 and A1 release Iysosomes and A β42

Radial dispersal of A β42 from N1 and A1

Formation of N1- and A1- derived plaques

Immigration of macrophages from vessels

Chemotaxis of phagocytes to plaque core

Infiltration of plaques with astrocyte (A2) processes

More phagocytes recruited into plaque

CAP

(a)

(b)

(c)

A1N1 Debris

AccumulatedAβ42

N1-derived plaque

Macrophages

Plaquecore

Astrocyte (A2)

A1-derived plaque

N2

N2

N1

A ccumula tedA β42

FIGURE 13.9. Proposed scenario for the involvement of astrocytes and microglia/macrophages in AD pathogenesis inthe context of developing neuronal pathology.



remnant at the plaque core, and not A β peptides,may be chemotactic to microglia/macrophages.While at the plaque core, it is not clear if microglia/ macrophages ingest A β in AD brains. It is morelikely that their role is to clear remaining nucleardebris from the plaque core. Local activated astro-cytes that are positioned just outside the plaquemargin extend long, GFAP immunopositive cyto-plasmic processes toward the plaque and bothencapsulate it and infiltrate it with finer GFAP-neg-ative processes. In addition, plaque-associatedastrocytes clearly are able to internalize A β42-immunopositive material which accumulates intheir cell bodies.

The suggested ability of different cell types to inde-pendently give rise to amyloid plaques (especially

neurons and astrocytes) can account, at least in part,for the broad spectrum of plaque morphologiesobserved in AD brains. The proposed pathologicalsequence described in Figure 13.9 highlights the link between the loss of BBB integrity and the initiation of AD pathological changes. Equally important is thedramatic intraneuronal A β42 accumulation of A β42.The trigger for this phenomenon is unknown, but thepossibilities include one or more of the following;binding of serum-derived, exogenous A β42 toα7nAChR on neuronal surfaces, oxidative damage,reduced delivery of materials to distal dendrites,impaired neuronal A β42 clearance, or binding of neu-ron-specific immunoglobulins and complement thathave gained entry into the brain parenchyma via localor global breaches in the BBB. Regardless of thecause of neuronal A β42 accumulation, the fact that itleads to degeneration of distal dendrites and synapsesin the cortical molecular layer provides a plausibleexplanation for the early telltale signs of AD progres-sion (i.e., cognitive and memory decline), even prior

to the appearance of amyloid plaques within the braintissue. From a therapeutic perspective, maintaining orrestoring BBB integrity could be a first line of defenseagainst AD, and blocking the initial accumulation of Aβ42 in neurons is an obvious and early target.Success on either or both fronts would provide anopportunity to block or at least slow the progressionof AD pathology in the brains of the elderly.

Acknowledgments The authors wish to thank Drs.Michael D’Andrea and Hoau-Yan Wang for theirmany helpful discussions and Alison Rigby,

Jennifer Dubay, Seth Vatsky, Emily Sim, andJames Novak for their technical assistance. Thiswork is supported by grants from the NationalInstitute on Aging (AG00925), the Alzheimer’sAssociation., the New Jersey GerontologicalInstitute, and NJ Governor’s Council on Autism.

References

1. Selkoe D. Alzheimer’s disease: genes, proteins, andtherapy. Physiol Rev 2001; 81:741-66.

2. Wisniewski KE, Wisniewski HM, Wen GY.Occurrence of neuropathological changes anddementia of Alzheimer’s disease in Down’s syn-drome. Ann Neurol 1985; 17:278-82.

3. Selkoe DJ. Alzheimer’s disease: genotypes, pheno-

types, and treatments. Science 1997; 275:630-1.4. Terry RD, Masliah E, Slmon DP, et al. Physical basis

of cognitive alterations in Alzheimer’s disease:synapse loss is the major correlate of cognitiveimpairment. Ann Neurol 1991; 30:572-80.

5. Cummings JL, Vinters HV, Cole GM, et al.Alzheimer’s disease: etiologies, pathophysiology, cog-nitive reserve, and treatment opportunities. Neurology1998;51(1 Suppl 1):S2-17; discussion S65-7.

6. Dickson DW. The pathogenesis of senile plaques.J Neuropathol Exp Neurol 1997; 56:321-39.

7. Scheff SW and Price DA. Synaptic density in theinner molecular layer of the hippocampal dentategyrus in Alzheimer’s disease. J Neuropathol ExpNeurol 1998; 57:1146-53.

8. D’Andrea MR, Nagele RG, Wang HY, et al. Evidencethat neurones accumulating amyloid can undergolysis to form amyloid plaques in Alzheimer’s disease.Histopathology 2001; 38:120-34.

9. D’Andrea MR, Nagele RG, Wang HY, et al.Consistent immunohistochemical detection of intra-cellular beta-amyloid42 in pyramidal neurons of Alzheimer’s disease entorhinal cortex. Neurosci Lett

2002; 333:163-6.10. Citron M, Diehl TS, Gordon G, et al. Evidence that

the 42- and 40-amino acid forms of amyloid β pro-tein are generated from the β-amyloid precursor pro-tein by different protease activities. Proc Natl AcadSci U S A 1996; 93:13170-5.

11. Gendelman HE, Folks DG. Innate and acquiredimmunity in neurodegenerative disorders. J LeukocBiol 1999; 65:407-8.

12. Cotter RL, Burke WJ, Thomas VS, et al. Insights intothe neurodegenerative process of Alzheimer’s dis-

ease: a role for mononuclear phagocyte-associatedinflammation and neurotoxicity. J Leukoc Biol 1999;65:416-27.




13. Eng LF, Ghirnikar RS, Lee YL. Glial fibrillary acidicprotein: GFAP -thirty-one years (1969-2000).Neurochem Res 2000; 25:1439-51.

14. Bushong EA, Martone ME, Jones YZ, Ellisman MH.Protoplasmic astrocytes in CA1 stratum radiatumoccupy separate anatomical domains. J Neurosci

2002; 22(1):183-92.15. Ullian EM, Sapperstein SK, Christopherson KS,

et al. Control of synapse number by glia. Science2001; 291:657-61.

16. Martin ED, Araque A, Buno W. Synaptic regulationof the slow Ca2 +-activated K + current in hippocam-pal CA1 pyramidal neurons: implication in epilepto-genesis. J Neurophysiol 2001; 86:2878-86.

17. Fields RD, Stevens-Graham B. New insights intoneuron-glia communication. Science 2002; 298:556-62.

18. Hansson E, Ronnback L. Glial neuronal signalingin the central nervous system. FASEB J 2003;17:341-8.

19. Vernadakis A. Glia-neuron intercommunications andsynaptic plasticity. Prog Neurobiol 1996; 49:185-214.

20. Barres BA. New roles for glia. J Neurosci 1991;11:3685-94.

21. Danbolt NC. The high affinity uptake system forexcitatory amino acids in the brain. Prog Neurobiol1994; 44:377-96.

22. Rothstein JD, Martin L, Levey AI, et al. Localizationof neuronal and glial glutamate transporters. Neuron

1994; 13:713-25.23. Lehre KP, Levy LM, Ottersen OP, et al. Differential

expression of two glial glutamate transporters in therat brain: quantitative and immunocytochemicalobservations. J Neurosci 1995; 15:1835-53.

24. Hertz L, Zielke HR. Astrocytic control of gluta-matergic activity: astrocytes as stars of the show.Trends Neurosci 2004; 27:735-43.

25. Dong Y, Benveniste EN. Immune function of astro-cytes. Glia 2001; 36:180-90.

26. Hatten ME, Liem RK, Shelanski ML, et al. Astroglia

in CNS injury. Glia 1991; 4:233-43.27. Rubin LL, Staddon JM. The cell biology of the blood-brain barrier. Annu Rev Neurosci 1999; 22:11-28.

28. Pekny M, Nilsson M. Astrocyte activation and reac-tive gliosis. Glia 2005; 50:427-34.

29. Eng LF, Ghirnikar RS. GFAP and astrogliosis. BrainPathol 1994; 4:229-37.

30. Itagaki S, McGeer PL, Akiyama H, et al. Relation-ship of microglia and astrocytes to amyloid depositsof Alzheimer’s disease. J Neuroimmunol 1989; 24:173-82.

31. Schubert P, Ogata T, Marchini C, et al. Glia-relatedpathomechanisms in Alzheimer’s disease: a thera-peutic target? Mech. Ageing Dev 2001; 123:47-57.

32. Panickar KS, Norenberg MD. Astrocytes in cerebralischemic injury: morphological and general consid-erations. Glia 2005; 50:287-98.

33. Wilhelmsson U, Li L, Pekna M, et al. Absence of glialfibrillary acidic protein and vimentin prevents hyper-trophy of astrocytic processes and improves post-

traumatic regeneration. J Neurosci 2004;24:5016-21.34. Anderson CM, Swanson RA. Astrocyte glutamate

transport: review of properties, regulation, and phys-iological functions. Glia 2000; 32:1-14.

35. Parpura V, Scemes E, Spray DC. Mechanisms of glu-tamate release from astrocytes: gap junction“hemichannels,” purinergic receptors and exocytoticrelease. Neurochem Int 2004; 45:259-64.

36. Norenberg MD. The reactive astrocyte. In: AschnerM, editor. The role of glia in neurotoxicity. BocaRaton, FL: CRC Press, 2005:73-92.

37. Ge S, Song L, Pachter JS. Where is the blood-brainbarrier . . .really? J Neurosci Res 2005; 79:421-7.

38. Bush AI, Beyreuther K, Masters CL. The beta A4amyloid protein precursor in human circulation. AnnN Y Acad Sci 1993; 695:175-82.

39. Haas C, Hung AY, Citron M, et al. beta-Amyloid,protein processing and Alzheimer’s disease.Arzneimittelforschung 1995; 45:398-402.

40. Younkin SG. The role of A beta 42 in Alzheimer’sdisease. J Physiol (Paris) 1998; 92:289-92.

41. Dewitt DA, Perry G, Cohen M, et al. Astrocytes reg-ulate microglial phagocytosis of senile plaque cores

of Alzheimer’s disease. Exp Neurol 1998; 149:329-40.

42. Akiyama H, Mori H, Saido T, et al. Occurrence of thediffuse amyloid beta-protein (Abeta) deposits withnumerous Abeta-containing glial cells in the cerebralcortex of patients with Alzheimer’s disease. Glia1999; 25:324-31.

43. Kurt MA, Davies DC, Kidd MM. beta-Amyloidimmunoreactivity in astrocytes in Alzheimer’s dis-ease brain biopsies: an electron microscope study.Exp Neurol 1999; 158:221-8.

44. Nagele RG, D’Andrea MR, Lee H, Venkataraman V,et al. Astrocytes accumulate Amyloid beta 42 andgive rise to astrocytic amyloid plaques inAlzheimer’s disease brains. Brain Res 2003;971:197-209.

45. Thal DR, Schultz C, Dehghani F, et al. Amyloidbeta-protein (Abeta)-containing astrocytes arelocated preferentially near N-terminal-truncatedAbeta deposits in the human entorhinal cortex. ActaNeuropathol 2000; 100:608-17.

46. Thal DR, Hartig W, Schober R. Diffuse plaques inthe molecular layer show intracellular AB8-17-immunoreactive deposits in subpial astrocytes. ClinNeuropathol 1999; 18:226-31.

230 G. Siu et al.



47. Aldskogius H, Liu L, Svensson M. Glial responses tosynaptic damage and plasticity. J Neurosci Res 1999;58:33-41.

48. Wegiel J, Wang K-C, Tarnawski M, et al. Microglialcells are the driving force in fibrillar plaque forma-tion whereas astrocytes are a leading factor in plaque

degradation. Acta Neuropathol 2000; 100:356-64.49. Wegiel J, Wang K-C, Imaki H, et al. The role of

microglial cells and astrocytes in fibrillar plaque evo-lution in transgenic APPsw mice. Neurobiol Aging2001; 22:49-61.

50. Wegiel J, Imaki H, Wang K-C, et al. Origin andturnover of microglial cells in fibrillar plaques of APPsw transgenic mice. Acta Neuropathol 2003;105:393-402.

51. Kurochkin IV, Goto S. Alzheimer’s beta-amyloid pep-tide specifically interacts with and is degraded byinsulin degrading enzymes. FEBS Lett 1994; 345:33-7.

52. Qiu WQ, Walsh DM, Ye Z, et al. Insulin-degradingenzyme regulates extracellular levels of amyloid β-protein by degradation. J Biol Chem 1998; 273:32730-8.

53. Apelt J, Ach K, Schliebs R. Aging-related down-reg-ulation of neprilysin, a putative β-amyloid-degradingenzyme, in transgenic Tg2576 Alzheimer-like mousebrain is accompanied by an astroglial upregulation inthe vicinity of β-amyloid plaques. Neurosci Lett2003; 339:183-6.

54. Magnus T, Chan A, Linker RA, et al. Astrocytes are

less efficient in the removal of apoptotic lympho-cytes than microglia cells: implications for the role of glial cells in the inflamed central nervous system.J Neuropathol Exp Neurol 2002; 61:760-6.

55. Ard MD, Cole GM, Wei J, et al. Scavenging of Alzheimer’s amyloid β-protein by microglia in cul-ture. J Neurosci Res 1996; 43:190-202.

56. Wisniewski HM, Wegiel J, Wang K-C, et al.Ultrastructural studies of the cells forming amyloidfibers in classical plaques. Can J Neurol Sci 1989;16:535-42.

57. Wisniewski HM, Wegiel J. Spatial relationshipsbetween astrocytes and classical plaque components.Neurobiol Aging 1991; 12:593-600.

58. Perry VH, Gordon S. Macrophages in the nervoussystem. Int Rev Cytol 1991; 125:203-44.

59. Biernacki K, Prat A, Blain M, et al. Regulation of Th1 and Th2 lymphocyte migration by human adultbrain endothelial cells. J Neuropathol Exp Neurol2001; 60:1127-36.

60. Nagele RG, D’Andrea MR, Anderson WJ, et al.Intracellular accumulation of B-amyloid in neuronsis facilitated by the a7 nicotinic acetylcholine recep-tor in Alzheimer’s disease. Neuroscience 2002;110:199-211.

61. Gouras GK, Tsaiu J, Nalund J, et al. IntraneuronalAβ42 accumulation in human brain. Am J Pathol2000; 156:15-20.

62. Gyure KA, Durham R, Stewart WF, et al.Intraneuronal abeta-amyloid precedes developmentof amyloid plaques in Down syndrome. Arch Pathol

Lab Med 2001; 125:489-92.63. Wang H-Y, Lee DHS, D’Andrea MR, et al. β-

amyloid1-42 binds to α 7 nicotinic acetylcholinereceptor with high affinity: implications forAlzheimer’s disease pathology. J Biol Chem 2000;275:5626-32.

64. Wang H-Y, D’Andrea MR, Nagele RG. Cerebellardiffuse amyloid plaques are derived from dendriticAβ42 accumulations in Purkinje cells. NeurobiolAging 2002; 23:213-23.

65. Lee DH, Wang HY. Differential physiologic responsesof alpha7 nicotinic acetylcholine receptors to beta-amyloid1-40 and beta-amyloid1-42. J Neurobiol2003; 55:25-30.

66. Nagele RG, Wegiel J, Venkataraman V, et al.Contribution of glial cells to the development of amyloid plaques in Alzheimer’s disease. NeurobiolAging 2004; 25:663-74.

67. Oddo S, Caccamo A, Sheppard JD, et al. Triple-transgenic model of Alzheimer’s disease withplaques and tangles: intracellular Abeta and synapticdysfunction. Neuron. 2003; 39:409-21.

68. Mori C, Spooner ET, Wisniewsk KE, et al.

Intraneuronal Abeta42 accumulation in Down syn-drome brain. Amyloid 2002; 9:88-102.

69. Streit WJ, Walter SA, Pennell NA. Reactivemicrogliosis. Prog Neurobiol 1999; 57:563-81.

70. Griffin WS, Sheng JG, Roberts GW, et al.Interleukin-1 expression in different plaque types inAlzheimer’s disease: significance in plaque evolu-tion. J Neuropathol Exp Neurol 1995; 54:276-81.

71. Griffin WS, Stanley LC, Ling C, et al. Brain inter-leukin 1 and S-100 immunoreactivity are elevated inDown syndrome and Alzheimer’s disease. Proc Natl

Acad Sci U S A 1989; 86:7611-5.72. Summers WK. Alzheimer’s disease, oxidative injury,and cytokines. J Alzheimers Dis 2004; 6:651-7.

73. Perry VH, Andersson PB, Gordon S. Macrophagesand inflammation in the central nervous system.Trends Neurosci 1993; 16:268-73.

74. Streit WJ, Graeber MB, Kreutzberg GW. Functionalplasticity of microglia: a review. Glia 1988; 1:301-7.

75. Rogers J, Lue LF. Microglial chemotaxis, activation,and phagocytosis of amyloid beta-peptide as linkedphenomena in Alzheimer’s disease. Neurochem Int2001; 39:333-40.

76. Burudi EM, Regnier-Vigouroux A. Regional andcellular expression of the mannose receptor in the




post-natal developing mouse brain. Cell Tissue Res2001; 303:307-17.

77. Walker DG, Kim SU, McGeer PL. Complement andcytokine gene expression in cultured microglicalderived from postmortem human brains. J NeurosciRes 1995; 40:478-93.

78. Strohmeyer R, Shen Y, Rogers J. Detection of com-plement alternative pathway mRNA and proteins inthe Alzheimer’s disease brain. Brain Res Mol BrainRes 2000; 81:7-18.

79. Walker DG, Lue LF, Beach TG. Gene expressionprofiling of amyloid beta peptide-stimulated humanpost-mortem brain microglia. Neurobiol Aging 2001;22:957-66.

80. Shen Y, Li R, McGeer EG, et al. Neuronal expression of mRNAs for complement proteins of the classical path-way in Alzheimer brain. Brain Res 1997; 769:391-5.

81. Schenk D, Barbour R, Dunn W, et al. Immunizationwith amyloid-beta attenuates Alzheimer-disease-likepathology in the PDAPP mouse. Nature 1999;400:173-7.

82. Bard F, Cannon C, Barbour R, et al. Peripherallyadministered antibodies against amyloid beta-pep-tide enter the central nervous system and reducepathology in a mouse model of Alzheimer’s disease.Nat Med 2000; 6:916-9.

83. Haga S, Akai K, Ishii T. Demonstration of microglialcells in and around senile (neuritic) plaques in theAlzheimer brain. An immunohistochemical study

using a novel monoclonal antibody. Acta Neuropathol(Berlin) 1989; 77:569-75.

84. Perlmutter LS, Scott SA, Barron E, et al. MHC classII-positive microglia in human brain: association withAlzheimer lesions. J Neurosci Res 1992; 33:549-58.

85. Wisniewski HM, Wegiel J, Wang KC, et al.Ultrastructural studies of the cells forming amyloidin the cortical vessel wall in Alzheimer’s disease.Acta Neuropathol (Berlin) 1992; 84:117-27.

86. Frautschy SA, Cole GM, Baird A. Phagocytosis anddeposition of vascular beta-amyloid in rat brains

injected with Alzheimer beta-amyloid. Am J Pathol1992;140:1389-99.87. Kopec KK, Carroll RT. Alzheimer’s beta-amyloid

peptide 1-42 induces a phagocytic response inmurine microglia. J Neurochem 1998; 71:2123-31.

88. Weldon DT, Rogers SD, Ghilardi JR, et al. Fibrillarbeta-amyloid induces microglial phagocytosis,expression of inducible nitric oxide synthase, andloss of a select population of neurons in the rat CNSin vivo. J Neurosci 1998; 18:2161-73.

89. Paresce DM, Chung H, Maxfield FR. Slow degrada-tion of aggregates of the Alzheimer’s disease amy-loid beta-protein by microglial cells. J Biol Chem1997; 272:29390-7.

90. Rogers J, Lue LF, Walker DG, et al. Elucidatingmolecular mechanisms of Alzheimer’s disease inmicroglial cultures. Ernst Schering Res FoundWorkshop 2002; 39:25-44.

91. Frackowiak J, Wisniewski HM, Wegiel J, et al.Ultrastructure of the microglia that phagocytose

amyloid and the microglia that produce β-amyloidfibrils. Acta Neuropathol 1992; 84:225-33.

92. Wisniewski HM, Weigel J. Migration of perivascu-lar cells into the neuropil and their involvement inbeta-amyloid plaque formation. Acta Neuropathol(Berlin) 1993; 85:586-95.

93. Stalder M, Phinney A, Probst A, et al. Associationof microglia with amyloid plaques in brains of APP23 transgenic mice. Am J Pathol 1999;154:1673-84.

94. Lue LF, Walker DG, Rogers J. Modeling microglialactivation in Alzheimer’s disease with human post-mortem microglial cultures. Neurobiol Aging 2001;22:945-56.

95. Rogers J, Lue LF, Walker DG, et al. Elucidatingmolecular mechanisms of Alzheimer’s disease inmicroglial cultures. Ernst Schering Res FoundWorkshop 2002; 39:25-44.

96. Davis JB, McMurray HF, Schubert D. The amyloidbeta-protein of Alzheimer’s disease is chemotacticfor mononuclear phagocytes. Biochem Biophys ResCommun 1992; 189:1096-100.

97. Yan SD, Chen X, Fu J, et al. RAGE and amyloid-

beta peptide neurotoxicity in Alzheimer’s disease.Nature 1996; 382:685-91.

98. Kato K, Suzuki F, Morishita R, et al. Selectiveincrease in S-100 beta protein by aging in rat cere-bral cortex. J Neurochem 1990; 54:1269-74.

99. D’Andrea MR, Nagele RG. MAP-2 immunolabel-ing can distinguish diffuse from dense-core amyloidplaques in brains with Alzheimer’s disease. BiotechHistochem 2002; 77:95-103.

100. Fiala M, Zhang L, Gan X, et al. Amyloid-betainduces chemokine secretion and monocyte migra-

tion across a human blood–brain barrier model. MolMed 1998; 4:480-9.101. Loike JD, el Khoury J, Cao L, et al. Fibrin regulates

neutrophil migration in response to interleukin 8,leukotriene B4, tumor necrosis factor, and formyl-methionyl-leucyl-phenylalanine. J Exp Med 1995;181:1763-72.

102. Lorton D, Schaller J, Lala A, et al. Chemotactic-likereceptors and Abeta peptide induced responses inAlzheimer’s disease. Neurobiol Aging 2000;21:463-73.

103. Du Yan S, Zhu H, Fu J, et al. Amyloid-beta peptide-receptor for advanced glycation endproduct interac-tion elicits neuronal expression of macrophage-

232 G. Siu et al.



colony stimulating factor: a proinflammatory path-way in Alzheimer’s disease. Proc Natl Acad Sci U S A1997; 94:5296-301.

104. Gehrmann J, Schoen SW, Kreutzberg GW. Lesionof the rat entorhinal cortex leads to a rapidmicroglial reaction in the dentate gyrus. A light and

electron microscopical study. Acta Neuropathol(Berlin) 1991; 82:442-55.

105. Lacy M, Jones J, Whittemore SR, et al. Expressionof the receptors for the C5a anaphylatoxin, inter-leukin-8 and FMLP by human astrocytes andmicroglia. J Neuroimmunol 1995; 61:71-8.

106. Gasque P, Singhrao SK, Neal JW, et al. The recep-tor for complement anaphylatoxin C3a is expressedby myeloid cells and nonmyeloid cells in inflamedhuman central nervous system: analysis in multiplesclerosis and bacterial meningitis. J Immunol 1998;160:3543-54.

107. Kaur C, Chan YG, Ling EA. Ultrastructural andimmunocytochemical studies of macrophages in anexcitotoxin induced lesion in the rat brain. JHirnforsch 1992; 33:645-52.

108. Hauwel M, Furon E, Canova C, et al. Innate (inher-ent) control of brain infection, brain inflammationand brain repair: the role of microglia, astrocytes,“protective” glial stem cells and stromal ependymalcells. Brain Res Brain Res Rev 2005; 48:220-33.

109. Jiang H, Burdick D, Glabe CG, et al. beta-Amyloidactivates complement by binding to a specific

region of the collagen-like domain of the C1q Achain. J Immunol 1994; 152:5050-9.

110. McGeer PL, Akiyama H, Itagaki S, et al. Activationof the classical complement pathway in brain tissueof Alzheimer patients. Neurosci Lett 1989;107:341-6.

111. Tacnet-Delorme P, Chevallier S, Arlaud GJ. Beta-amyloid fibrils activate the C1 complex of comple-ment under physiological conditions: evidence for abinding site for A beta on the C1q globular regions.J Immunol 2001; 167:6374-81.

112. Akiyama H, Barger S, Barnum S, et al.Inflammation and Alzheimer’s disease. NeurobiolAging 2000; 21:383-421.

113. Anthony JC, Breitner JC, Zandi PP, et al. Reducedprevalence of AD in users of NSAIDs and H2

receptor antagonists: the Cache County study.Neurology 2000; 54:2066-71.

114. Colton CA, Chernyshev ON, Gilbert DL, et al.Microglial contribution to oxidative stress inAlzheimer’s disease. Ann N Y Acad Sci 2000;899:292-307.

115. Flynn BL, Theesen KA. Pharmacologic manage-ment of Alzheimer’s disease part III: nonsteroidalantiinflammatory drugs—emerging protective evi-dence? Ann Pharmacother 1999; 33:840-9.

116. in ‘t Veld BA, Launer LJ, Hoes AW, et al. NSAIDsand incident Alzheimer’s disease. The RotterdamStudy. Neurobiol Aging 1998; 19:607-11.

117. McGeer PL, McGeer EG. Inflammation, autotoxic-ity and Alzheimer’s disease. Neurobiol Aging 2001;22:799-809.

118. Mortimer JA. New findings consistent withAlzheimer’s-NSAIDs link. Neurobiol Aging 1998;19:615-6.

119. Pasinetti GM. Cyclooxygenase and inflammation inAlzheimer’s disease: experimental approaches andclinical interventions. J Neurosci Res 1998; 54:1-6.

120. Schubert P, Ogata T, Marchini C, et al. Glia-relatedpathomechanisms in Alzheimer’s disease: a thera-peutic target? Mech Ageing Dev 2001; 123:47-57.

121. Flanary BE, Streit WJ. Progressive telomere short-ening occurs in cultured rat microglia, but not astro-cytes. Glia 2004; 45:75-88.

122. Streit WJ, Sammons NW, Kuhns AJ, et al.

Dystrophic microglia in the aging human brain.Glia 2004; 45:208-12.

123. Wilcock DM, Gordon MN, Ugen KE, et al. Numberof Abeta inoculations in APP +PS1 transgenic miceinfluences antibody titers, microglial activation, andcongophilic plaque levels. DNA Cell Biol 2001;20:731-6.

124. Schenk D, Barbour R, Dunn W, et al. Immunizationwith amyloid-beta attenuates Alzheimer’s-disease-like pathology in the PDAPP mouse. Nature 1999;400:173-7.

125. Bard F, Cannon C, Barbour R, et al. Peripherallyadministered antibodies against amyloid beta-pep-tide enter the central nervous system and reducepathology in a mouse model of Alzheimer’s dis-ease. Nat Med 2000; 6:916-9.




14The Role of Presenilins in A β-Induced

Cell Death in Alzheimer’s DiseaseMaria Ankarcrona

234

14.1 Introduction

Neuronal death in specific brain regions is acommon feature of neurodegenerative disorders.Alzheimer’s disease (AD) is characterized bysynaptic loss and a substantial amount of neuronaldegeneration in regions involved in memory andlearning processes (e.g., temporal, entorhinal andfrontal cortex; hippocampus). The neuropathologichallmarks of AD include the accumulation of amy-loid plaques and hyperphosphorylated tau formingintracellular tangles. However, no correlation hasbeen established between the number of plaquesand the cognitive performance in AD patients [1,2]. Instead, synaptic failure and intracellular pro-duction of amyloid beta (A β) appears to correlatewell with the early cognitive dysfunction in ADpatients [3, 4]. This has also been tested in a tripletransgenic mouse model of AD where accumula-tion of intracellular A β1-42 corresponded with theearly cognitive impairment [5]. Interestingly, noextracellular deposits of A β1-42 were detected in

these mice at 4 months of age suggesting that A β1-42accumulate intracellulary early in the diseaseprocess. Moreover, intracellular accumulation of Aβ1-42 was cleared with administration of anti-A βantibodies and rescued the retention deficits seen inyoung 3 ×Tg AD mice. Together, results from thisand several other studies indicate that intracellularAβ1-42 -generation causes the primary toxicity toneurons in AD [6].

In this chapter, the functions of presenilin (PS)in A β-generation and toxicity will be described. PSappears to play several roles in cell death mecha-nisms associated with AD: (i) functional PS is

crucial for the generation of A β [7, 8], (ii) PS inter-

acts with proteins involved in cell signaling, regu-lation of calcium homeostasis, and apoptosis [9],and (iii) PS mutations sensitize cells to differentapoptotic stimuli in vitro [10] and increase the gen-eration of A β1-42 . Whether it is the overproductionof A β1-42 per se or other non-A β-related changesthat cause the increased sensitivity of cells carryingPS mutations is not clear, and the different possi-bilities will be discussed here.

14.2 Cell Death in AD Brain

The mechanisms of cell death in the AD brain arenot fully elucidated, however it is likely that severalforms of cell death are involved. Loss of synapsesis an early phenotypic manifestation in the pathol-ogy of AD, and synapse density is significantlydecreased in AD. Synapse loss and impaired long-term potentiation also precede accumulation of plaques and tangles in 3 ×Tg mice [11]. Cytosolic

extracts from synaptosomes exposed to A β inducedchromatin condensation and fragmentation of iso-lated nuclei showing that apoptotic signals can begenerated locally in synapses [12, 13]. Neuronsthat lose synapses and therefore also contact andcommunication with other cells are still alive butdo not function as before and will not survive in thelong run. Such cells could, however, stay in the tis-sue as “ghost cells” before they are cleared awayby, for example, apoptosis. There are several evi-dences for apoptosis in AD. Postmortem analysisof AD brain showed TUNEL positive neurons andglia in hippocampus and cortex indicating DNA



fragmentation [14–20]. Increased expression of Bcl-2 family members [21–25], as well as increa-sed caspase activities and cleavage of caspase sub-strates have been detected in AD brain [26–32].Cells that are triggered to die by apoptosis (e.g.,have active caspase 8 and 9, which are initiator cas-pases), but fail to complete the process becauseexecutor caspases such as caspase-3 and -7 are notactive, have also been detected in AD brain [33].This phenomenon is called “abortosis” and is as ananti-apoptotic mechanism that might try to protectneurons from death. However, this process is prob-ably finally overridden as many neurons still die inAD. There is also evidence for activation of cell-cycle proteins in AD brain [34, 35]. This may be adefense mechanism initiated to survive bad condi-

tions or toxic stimuli. However the neurons do notgo through mitosis, instead they are stuck in a cyclethey cannot complete and eventually die.Postmitotic neurons do not normally divide, but itis possible that reentry of the cell cycle is necessaryfor the completion of apoptotis. Normally prolifer-ating cells are regularly checked throughout thecell cycle and taken aside to die by apoptosis whendamaged. Maybe also postmitotic cells have to takethis way to death.

A cell dying by apoptosis leaves no traces in thetissue because it is silently disassembled andphagocytosed. Therefore, the main part of cells,which presumably have died by apoptosis duringthe course of AD, have already been cleared fromthe tissue at the time of autopsy. This is one of thedifficulties with proving the impact of apoptoticcell death in AD. It has also been argued that thegreat difference in time spans between the diseaseprocess (approximately 20 years) and the apoptoticprocess (approximately 24 hours), rules out apop-

tosis as a mechanism for cell death in AD.However, if cell death is triggered at different timesduring the course of the disease, it is very likelythat cells die by apoptosis in AD.

From a therapeutic point of view, it would of course be most attractive to target the early cogni-tive changes in AD presumably associated withintracellular accumulation of A β and synaptic fail-ure. When the neuron is dead, it is too late.Therefore, it is of great importance to understandthe mechanisms behind neuronal failure to be ableto design the best neuroprotection. The treatmentstrategies are also highly dependent on diagnostic

methods: the earlier a correct diagnosis can begiven, the earlier a potential treatment could start.

14.3 Presenilins, γ -SecretaseActivity, and APP Processing

Most AD cases are sporadic or have not so far beengenetically linked. Only a minor number of ADcases have been associated with mutations in spe-cific genes. These genes are presenilin 1 (PS1),presenilin 2 (PS2), and amyloid precursor protein(APP) [36]. All these mutations are autosomaldominant and fully penetrant. Generally, familialAlzheimer’s disease (FAD) cases have a lower ageof onset (PS1 mutation carriers 44 ± 7.8 years and

PS2 carriers 58.6 ± 7.0 years [37]) and show amore aggressive form of the disease compared withsporadic cases. PS1 and PS2 are encoded on chro-mosomes 14 and 1, respectively [38–40] and show63% homology. PS are membrane-bound proteinswith eight transmembrane domains and localizedto the endoplasmatic reticulum, Golgi apparatus[41– 44], plasma membrane [45], nuclear envelope[46], lysosomes [47], and mitochondria [48].Deficiency of PS1 inhibits A β generation from β-amyloid precursor protein (APP) suggesting thatPS1 is involved in γ -secretase cleavage [49]. The γ -secretase complex consists of at least PS1/PS2,nicastrin (Nct), presenilin enhancer-2 (Pen-2), andanterior pharynx defective-1 (Aph-1), and γ -secre-tase activity has been reconstituted by expressingthese four components in yeast [50] (Fig. 14.1).The γ -secretase complex is assembled in the ERand then trafficked to late secretory compartmentsincluding the plasma membrane where it exerts itsbiological function [51]. The four γ -secretase com-

ponents are assembled stepwise. Nct and Aph-1

14. The Role of Presenilins in A β-Induced Cell Death in AD 235

Nicastrin

Pen-2

CTFNTF

PresenilinAph-1Lumen

Cytosol

FIGURE 14.1. Illustration of Aph-1, nicastrin, PS, and Pen-2, which together form the γ -secretase complex. Courtesyof Dr. Jan Näslund, Karolinska Institutet, Sweden.



first form a stable subcomplex followed by theaddition of full-length PS. Then Pen-2 is added tothe complex and full-length PS is cleaved into C-terminal (CTF) and N-terminal (NTF) fragmentsforming the functional heterodimer of PS. Full-length PS, CTFs, and NTFs as well as Pen-2 aredegraded by the proteasome when not incorporatedinto the γ -secretase complex [52–54]. The impor-tance of PS for γ -secretase activity has beendemonstrated in several ways: (i) in PS-deficientcells [7, 8], (ii) by the use of γ -secretase inhibitorsthat bind to PS [55, 56], (iii) by the substitution of either of two aspartyl residues in transmembranedomains 6 (Asp257) and 7 (Asp385) of PS1 [57].All these studies showed inhibited γ -secretaseactivity and lower production of A β.

The γ -secretase complex cleaves APP and othertype I membrane proteins [58]. Before γ -secretasecleavage, the N-terminal part of APP either facingthe extracellular space or the lumen is cleaved byβ-site APP cleaving enzyme (BACE), a processreferred to as ectodomain shedding. BACE cleav-age releases secreted sAPP- β and leaves a 99-amino-acid C-terminal fragment (C99) in themembrane. Subsequently, the γ -secretase complexcleaves C99 generating APP intracellular domain(AICD) and A β. In the non-amyloidogenic path-way, α -secretase cleaves APP in the middle of theAβ region resulting in the formation of secretedsAPP α and an 83-amino acid C-terminal fragment(C83). γ -Secretase cleavage of C83 also results inAICD formation [36]. Other γ -secretase substratesthan APP include Notch, ErbB4, E-cadherin,Delta/Jagged, nectin-1 α , CD44, and LRP [59].Many of these substrates function in intercellularcommunication or adhesion. Notch signaling isimportant during development as γ -secretase cleav-

age of this receptor generates a Notch intracellulardomain (NICD). NICD translocates to the nucleusand activates transcription of the cell-fate deter-mining HES (Hairy/Enhancer of split) genes, thusinitiating a non-neuronal development of the cell[60, 61]. Similarly, AICD has been detected in the

nucleus where it interacts with the nuclear adaptorprotein Fe65 and the histone acetyltransferase andactivates transcription [62–64]. AICD has alsobeen implicated in the regulation of phosphoinosi-tide-mediated calcium signaling [65].

Mice knocked-out for both PS1/ PS2 die beforeembryonic day 13.5 [49, 66]. PS1 can compensatefor the loss of PS2 (PS1 + / + PS2 − / − and PS1 + / −PS2 − / − embryos survive), while PS2 cannot fullycompensate for the loss of PS1 (PS1 − / − PS2 + / + dieat birth; PS1 − / − PS2 + / − embryos die duringE9.5–E13.5). The results from these animal modelsemphasize the importance of PS1/ γ -secretase activ-ity during embryogenesis. In accordance, PS1 − / −PS2 + / + and PS K/O mouse embryonic fibroblasts(MEFs) accumulate C83/C99 showing that PS1 is

responsible for most of the γ -secretase activitycleaving APP (Fig. 14.2). The residual γ -secretaseactivity comes from PS2 that contributes to A βproduction to a lesser extent than PS1 [49].

14.4 PS Mutations, A β-Generation,and Apoptosis

To date, almost 150 mutations have been identified inPS1 and 11 mutations in PS2 (AD mutation database:http://www.molgen.ua.ac.be/ADmutations/default.cf m). All these are missense mutations that generatesingle amino acid substitutions in the protein primarystructure, with the exception of PS1 exon 9 deletionsplice mutation [67]. The different PS mutations leadto a similar phenotype: an increased ratio of A β1-42 toAβ1-40 , increased plaque deposition, and early age of onset [68]. Although the mutations are distributed allover the PS molecule, with a clustering of mutationsin the transmembrane regions, the effect on A β-gen-

eration is similar indicating a common mechanism.Fluorescent lifetime imaging microscopy (FLIM)[69] studies have suggested that PS1 mutations,spread in different regions, all cause a conformationalchange in PS1. The proximity between the N- andC-terminus of PS1 was increased in the mutant PS1

236 M. Ankarcrona

WT PS1 − / −PS2+/+

PS1+/+PS2 − / −

PS K/O

← APP C83/C99FIGURE 14.2. Western blot of cell lysates isolatedfrom mouse embryonic fibroblasts (obtained from

Prof. Bart de Strooper). Accumulation of APPC83/C99 fragments indicates lack of γ -secretaseactivity.



compared with PS1 wild-type. A consistent changewas also detected in the configuration of the PS1-APPcomplex in PS1 mutants and could explain the com-mon effect on A β generation [70]. In another study,using a random mutagenesis screen of PS1, fiveunique mutations that exclusively generated a highlevel of A β1-43 were identified [71]. Together, thesetwo studies show that PS1 mutations may change theactivity and specificity of γ -secretase through a com-mon mechanism.

PS1 is responsible for the major γ -secretaseactivity generating A β, and PS2 plays a minor role,still it has been shown that PS2 mutations alsoinfluence the ratio of A β42/40. Of the reported PS2mutations T122P, N141I, M239V, and M239I sig-nificantly increased the A β42/40 ratio similar to

very-early-onset PS1 FAD mutations [72]. Theshift toward the production of longer and moreamyloidogenic A β species induced both by PS1and PS2 mutations suggest that this common alter-ation in APP processing by γ -secretase contributesto the increased neuronal death in FAD.

FAD mutant proteins are expressed from birth,but it takes decades for AD to manifest itself, andFAD mutant carriers start to develop the disease asadults. This suggests that FAD mutants do notinduce neuronal cell death themselves but ratherincrease sensitivity to cell death stimuli [73].Indeed, several in vitro studies have shown that pre-senilin mutations contribute to neuronal death andsensitize cells to apoptotic stimuli [10]. It has beenreported that FAD-linked mutant PS1 enhances celldeath in T lymphocytes [74], PC12 cells [75, 76],SH-SY5Y neuroblastoma cells [77, 78], and pri-mary neurons [79]. However, another study failed todemonstrate that mutant PS1 increases sensitivity tocytotoxic insults in primary neurons [80].

Alterations in cells carrying PS1 mutationsinclude higher caspase-3 activity [81], increasedoxygen radical levels [82], induction of p53 andBax upregulation of calpain, mitochondrial mem-brane depolarization [83], enhanced phospholi-pase C activity [84], and altered intracellularcalcium regulation [75]. The PS2 mutant N141I-PS2 induces neuronal death in immortalized celllines and primary neurons [85, 86]. The inductionof apoptosis in PS2 mutant N141I-PS2 cells wasaccompanied by increased caspase-3 activity anddecreased Bcl-2 expression after serum-depriva-tion [87].

Whether it is the increased A β42/40 ratio thatcauses the cellular alterations detected in PSmutants or vice versa is not known. One possibilityis that PS mutations affect cellular functions inde-pendently of γ -secretase activity making such cellsmore vulnerable to A β and other cell death stimuli.Another possibility is that the high intracellularproduction of toxic A β species in PS mutant cellsdisturbs different cellular functions and therebyfinally renders the cells more susceptible to celldeath stimuli including A β.

14.5 PS Mutations, A β, andIntracellular Calcium Homeostasis

Many studies have shown dysregulation of intra-cellular calcium (Ca 2+) homeostasis in cells carry-ing PS mutations. Mutant forms of PS1 have beenshown to enhance Ca 2+ transients in several differ-ent cell systems including transfected PC12 cells[75, 88], fibroblasts from human FAD patients [89,90], mutant knock-in mouse fibroblasts [91], cul-tured hippocampal neurons [92], and oocytes over-expressing mutant PS1 [93]. The effect onintracellular Ca 2+ might be mediated by inositoltriphosphate (IP

3) as FAD-linked PS1 mutations

potentiate IP 3-mediated Ca 2+ release from the ER[93]. The number of IP 3 receptors are not increasedin cortical homogenates of PS1 knock-in mice,instead it has been suggested that the exaggregatedcytosolic Ca 2+ signals result from increased storefilling [94].

Increased intracellular calcium concentrations[Ca 2+]i result in enhanced A β generation [95] andat the same time cells treated with A β showincreased [Ca 2+]i [96]. One mechanism by which

Aβ could increase [Ca 2+]i is the formation of cal-cium-permeable pores in membranes [97, 98–100].More recently, Kayed and collegues suggested thatamyloid oligomers rather induce permeabilizationof membranes, without forming pores or channels,and thereby enhance the ability of ions to movethrough the lipid bilayer [101].

Other APP fragments have been shown to stabi-lize [Ca 2+]i and protect from A β toxicity. sAPP α isformed when α -secretase cleaves APP in the non-amyloidogenic pathway (Fig. 14.2). sAPP α hasbeen shown to stabilize calcium homeostasis andprotect neurons against excitotoxic, metabolic, and




mitochondria also lead to decreased ATP produc-tion and impaired calcium buffering capacity.Apoptosis can be triggered locally in synapses [12,13], and loss of synapses correlates well with theimpairment of cognitive functions early in AD.Local A β production in synapses may thereforedamage mitochondria and cause synapse loss.

Aβ accumulates in mitochondria in AD brainand in APP transgenic mice [124] and has beenshown to inhibit enzymes important for mitochon-drial functions in vitro, for example, cytochrome coxidase, β-ketoglutarate dehydrogenase, and pyru-vate dehydrogenase [125–127]. Another intracellu-lar target for A β is alcohol binding dehydrogenase(ABAD) [124] (for a review, see [128]). ABAD islocated to mitochondria where it binds to A β and

promotes A β-induced cell stress. ABAD is overex-pressed in AD brain and in brains from transgenicAPP mice.

Aβ-toxicity is dependent on a functional elec-tron transport chain [129], and A β has been shownto induce oxidative stress [130, 131] and inductionof permeability transition [132, 133] in differentcell models. A β also induces p53 and Bax activa-tion [134] associated with apoptosis signalingthrough the mitochondrial pathway. In addition,Aβ triggers the release of cytochrome c from mito-chondria [135]. Taken together, it seems that A βinduces cell death by affecting different mitochon-drial functions and triggering apoptotic mecha-nisms. As discussed above, cells carrying PSmutations have increased production of A β and aresensitized to apoptotic stimuli. Mitochondria seemto be an important target for A β-induced cell deathin agreement with the central role of mitochondriain apoptosis signaling.

At present, it is not clear whether A β is produced

in mitochondria or imported into mitochondria.Two studies have shown the localization of APP tomitochondria. First, APP immunoreactivity wasdetected by electron microscopy in the outer mem-brane of mitochondria [136]. Second, APP wasshown to be imported into the outer mitochondrialmembrane. However, the import is arrested by anacidic domain that spans sequence 220–290 of APP leaving a 73-kDa portion of the C-terminalside of the protein facing the cytoplasm. Accordingto this topology, the A β peptide region of APP isnot located to the membrane making it impossiblefor β- and γ -secretases to cleave out A β from APP

located to mitochondria [137]. We have shown thatPS, nicastrin, Pen-2, and Aph-1 form active γ -secretase complexes in mitochondria [138]. So far,no γ -secretase substrate has been identified in mito-chondria, and the function of the mitochondrialγ -secretase complex is not known. In conclusion, itis most likely that A β is taken up by mitochondriaand that the mitochondrial γ -secretase complexcleaves other substrates than APP. Exactly how A βgains access to mitochondria is not known, and thisissue has to be addressed in future studies. A β issecreted luminally and has been detected inER/Golgi, lysosomes/endosomes, and multivesicu-lar bodies. One possiblility is that, for example,ER-to-mitochondrial transfer might occur [139].

14.8 Conclusions

It has been established that PS is essential for γ -secretase activity, and PS is therefore mandatoryfor the generation of A β. Aβ is toxic and kills cellsby mechanisms involving perturbed intracellularcalcium homeostasis, oxidative stress, andimpaired mitochondrial functions. PS mutationssensitize cells to various toxic stimuli in vitro andincrease the production of A β. Whether it is theincreased A β load that causes the sensitization of PS mutant cells or if PS mutations cause cellularalterations independent of A β production have notbeen elucidated. Further studies have to be per-formed to shed more light on these complicatedmechanisms. Under all circumstances, it is becom-ing clear that it is the intracellular A β that is pri-marily toxic. Therefore, it is of great importance todecrease A β-generation and protect neurons fromAβ in order to block cell death in AD.

References

1. Arriagada PV, Growdon JH, Hedley-Whyte ET,Hyman BT. Neurofibrillary tangles but not senileplaques parallel duration and severity of Alzheimer’sdisease. Neurology 1992; 42:631-9.

2. Samuel W, Terry RD, DeTeresa R, et al. Clinical cor-relates of cortical and nucleus basalis pathology inAlzheimer dementia. Arch Neurol 1994; 51:772-8.

3. Terry RD, Masliah E, Salmon DP, et al. Physical

basis of cognitive alterations in Alzheimer’s disease:synapse loss is the major correlate of cognitiveimpairment. Ann Neurol 1991; 30:572-80.




4. Selkoe DJ. Alzheimer’s disease is a synaptic failure.Science 2002; 298:789-91.

5. Billings LM, Oddo S, Green KN, et al. IntraneuronalAbeta causes the onset of early Alzheimer’s disease-related cognitive deficits in transgenic mice. Neuron2005; 45:675-88.

6. Tseng BP, Kitazawa M, LaFerla FM Amyloid β-peptide: The inside story. Curr Alzheimer Res 2004;1:231-239.

7. Zhang Z, Nadeau P, Song W, et al. Presenilins arerequired for gamma-secretase cleavage of beta-APPand transmembrane cleavage of Notch-1. Nat CellBiol 2000; 2:463-5.

8. Herreman A, Serneels L, Annaert W, et al. Totalinactivation of gamma-secretase activity in prese-nilin-deficient embryonic stem cells. Nat Cell Biol2000; 2:461-2.

9. Chen Q, Schubert D. Presenilin-interacting proteins.Expert Rev Mol Med 2002; 2002:1-18.

10. Popescu BO, Ankarcrona, M. Neurons bearing pre-senilins: weapons for defense or suicide. J Cell MolMed 2000; 4:249-261.

11. Oddo S, Caccamo A, Shepherd JD, et al. Triple-transgenic model of Alzheimer’s disease withplaques and tangles. Intracellular abeta and synapticdysfunction. Neuron 2003; 39:409-21.

12. Mattson MP, Keller JN, Begley JG. Evidence forsynaptic apoptosis. Exp Neurol 1998; 153:35-48.

13. Mattson MP, Partin J, Begley JG. Amyloid beta-

peptide induces apoptosis-related events in synapsesand dendrites. Brain Res 1998; 807:167-76.

14. Su JH, Anderson AJ, Cummings BJ, Cotman CW.Immunohistochemical evidence for apoptosis inAlzheimer’s disease. Neuroreport 1994; 5:2529-33.

15. Lassmann H, Bancher C, Breitschopf H, et al. Celldeath in Alzheimer’s disease evaluated by DNAfragmentation in situ. Acta Neuropathol 1995;89:35-41.

16. Smale G, Nichols NR, Brady DR, et al. Evidence forapoptotic cell death in Alzheimer’s disease. Exp

Neurol 1995; 133:225-30.17. Dragunow M, Faull RL, Lawlor P, et al. In situ evi-dence for DNA fragmentation in Huntington’s dis-ease striatum and Alzheimer’s disease temporallobes. Neuroreport 1995; 6:1053-7.

18. Li WP, Chan WY, Lai HW, Yew DT. Terminal dUTPnick end labeling (TUNEL) positive cells in the dif-ferent regions of the brain in normal aging andAlzheimer patients. J Mol Neurosci 1997; 8:75-82.

19. Lucassen PJ, Chung WC, Kamphorst W, Swaab DF.DNA damage distribution in the human brain as shownby in situ end labeling; area-specific differences in agingand Alzheimer’s disease in the absence of apoptotic mor-phology. J Neuropathol Exp Neurol 1997; 56:887-900.

20. Sugaya K, Reeves M, McKinney M. Topographicassociations between DNA fragmentation andAlzheimer’s disease neuropathology in the hippo-campus. Neurochem Int 1997; 31:275-81.

21. Kitamura Y, Shimohama S, Kamoshima W, et al.Alteration of proteins regulating apoptosis, Bcl-2,

Bcl-x, Bax, Bak, Bad, ICH-1 and CPP32, inAlzheimer’s disease. Brain Res 1998; 780:260-9.

22. Giannakopoulos P, Kovari E, Savioz A, et al.Differential distribution of presenilin-1, Bax, andBcl-X(L) in Alzheimer’s disease and frontotempo-ral dementia. Acta Neuropathol (Berlin) 1999;98:141-9.

23. Su JH, Deng G, Cotman CW. Bax protein expressionis increased in Alzheimer’s brain: correlations withDNA damage, Bcl-2 expression, and brain pathol-ogy. J Neuropathol Exp Neurol 1997; 56:86-93.

24. Drache B, Diehl GE, Beyreuther K, et al. Bcl-xl-spe-cific antibody labels activated microglia associatedwith Alzheimer’s disease and other pathologicalstates. J Neurosci Res 1997; 47:98-108.

25. MacGibbon GA, Lawlor PA, Sirimanne ES, et al.Bax expression in mammalian neurons undergoingapoptosis, and in Alzheimer’s disease hippocampus.Brain Res 1997; 750:223-34.

26. Yang F, Sun X, Beech W, et al. Antibody to caspase-cleaved actin detects apoptosis in differentiated neu-roblastoma and plaque-associated neurons andmicroglia in Alzheimer’s disease. Am J Pathol 1998;

152:379-89.27. Uetsuki T, Takemoto K, Nishimura I, et al.

Activation of neuronal caspase-3 by intracellularaccumulation of wild-type Alzheimer amyloid pre-cursor protein. J Neurosci 1999; 19:6955-64.

28. LeBlanc A, Liu H, Goodyer C, et al. Caspase-6 rolein apoptosis of human neurons, amyloidogenesis,and Alzheimer’s disease. J Biol Chem 1999;274:23426-36.

29. Chan SL, Griffin WS, Mattson MP. Evidence for cas-pase-mediated cleavage of AMPA receptor subunits

in neuronal apoptosis and Alzheimer’s disease. JNeurosci Res 1999; 57:315-23.30. Stadelmann C, Deckwerth TL, Srinivasan A, et al.

Activation of caspase-3 in single neurons andautophagic granules of granulovacuolar degenerationin Alzheimer’s disease. Evidence for apoptotic celldeath. Am J Pathol 1999; 155:1459-66.

31. Rohn TT, Head E, Nesse WH, et al. Activation of caspase-8 in the Alzheimer’s disease brain.Neurobiol Dis 2001; 8:1006-16.

32. Pompl PN, Yemul S, Xiang Z, et al. Caspase geneexpression in the brain as a function of the clinicalprogression of Alzheimer’s disease. Arch Neurol2003; 60:369-76.

240 M. Ankarcrona



33. Raina AK, Hochman A, Zhu X, et al. Abortive apop-tosis in Alzheimer’s disease. Acta Neuropathol(Berlin) 2001; 101:305-10.

34. Yang Y, Mufson EJ, Herrup K. Neuronal cell death ispreceded by cell cycle events at all stages of Alzheimer’s disease. J Neurosci 2003; 23:2557-63.

35. Copani A, Sortino MA, Nicoletti F, Giuffrida SA.Alzheimer’s disease research enters a “new cycle”:how significant? Neurochem Res 2002; 27:173-6.

36. Mattson MP. Pathways towards and away fromAlzheimer’s disease. Nature 2004; 430:631-9.

37. Lippa CF, Swearer JM, Kane KJ, et al. FamilialAlzheimer’s disease: site of mutation influences clin-ical phenotype. Ann Neurol 2000; 48:376-9.

38. Levy-Lahad E, Wasco W, Poorkaj P, et al. Candidategene for the chromosome 1 familial Alzheimer’s dis-ease locus. Science 1995; 269:973-7.

39. Rogaev EI, Sherrington R, Rogaeva EA, et al. FamilialAlzheimer’s disease in kindreds with missense muta-tions in a gene on chromosome 1 related to the Alzhei-mer’s disease type 3 gene. Nature 1995; 376:775-8.

40. Sherrington R, Rogaev EI, Liang Y, et al. Cloning of agene bearing missense mutations in early-onset famil-ial Alzheimer’s disease. Nature 1995; 375:754-60.

41. Cook DG, Sung JC, Golde TE, et al. Expression andanalysis of presenilin 1 in a human neuronal system:localization in cell bodies and dendrites. Proc NatlAcad Sci U S A 1996; 93:9223-8.

42. Kovacs DM, Fausett HJ, Page KJ, et al. Alzheimer-

associated presenilins 1 and 2: neuronal expressionin brain and localization to intracellular membranesin mammalian cells. Nat Med 1996; 2:224-9.

43. Walter J, Capell A, Grunberg J, et al. The Alzheimer’sdisease-associated presenilins are differentially phos-phorylated proteins located predominantly within theendoplasmic reticulum. Mol Med 1996; 2:673-91.

44. De Strooper B, Beullens M, Contreras B, et al.Phosphorylation, subcellular localization, and mem-brane orientation of the Alzheimer’s disease-associ-ated presenilins. J Biol Chem 1997; 272:3590-8.

45. Dewji NN, Singer SJ. Cell surface expression of theAlzheimer’s disease-related presenilin proteins. ProcNatl Acad Sci U S A 1997; 94:9926-31.

46. Li J, Xu M, Zhou H, Ma J, Potter H. Alzheimer prese-nilins in the nuclear membrane, interphase kineto-chores, and centrosomes suggest a role in chromosomesegregation. Cell 1997; 90:917-27.

47. Pasternak SH, Bagshaw RD, Guiral M, et al.Presenilin-1, nicastrin, amyloid precursor protein, andgamma-secretase activity are co-localized in the lyso-somal membrane. J Biol Chem 2003; 278: 26687-94.

48. Ankarcrona M, Hultenby K. Presenilin-1 is locatedin rat mitochondria. Biochem Biophys Res Commun2002; 295:766-70.

49. De Strooper B, Saftig P, Craessaerts K, et al.Deficiency of presenilin-1 inhibits the normal cleav-age of amyloid precursor protein. Nature 1998;391:387-90.

50. Edbauer D, Winkler E, Regula JT, et al. Reconstitutionof gamma-secretase activity. Nat Cell Biol 2003;

5:486-8.51. Capell A, Beher D, Prokop S, et al. Gamma-secretase

complex assembly within the early secretory path-way. J Biol Chem 2005; 280:6471-8.

52. Steiner H, Capell A, Pesold B, et al. Expression of Alzheimer’s disease-associated presenilin-1 is con-trolled by proteolytic degradation and complex for-mation. J Biol Chem 1998; 273:32322-31.

53. Van Gassen G, De Jonghe C, Pype S, et al.Alzheimer’s disease associated presenilin 1 interactswith HC5 and ZETA, subunits of the catalytic 20Sproteasome. Neurobiol Dis 1999; 6:376-91.

54. Bergman A, Hansson EM, Pursglove SE, et al. Pen-2 is sequestered in the endoplasmic reticulum andsubjected to ubiquitylation and proteasome-mediateddegradation in the absence of presenilin. J Biol Chem2004; 279:16744-53. Epub 2004 Jan 14.

55. Li YM, Xu M, Lai MT, et al. Photoactivated gamma-secretase inhibitors directed to the active site cova-lently label presenilin 1. Nature 2000; 405:689-94.

56. Esler WP, Kimberly WT, Ostaszewski BL, et al.Transition-state analogue inhibitors of gamma-secre-tase bind directly to presenilin-1. Nat Cell Biol 2000;

2:428-34.57. Wolfe MS, Xia W, Ostaszewski BL, et al. Two trans-

membrane aspartates in presenilin-1 required for pre-senilin endoproteolysis and gamma-secretaseactivity. Nature 1999; 398:513-7.

58. Xia W, Wolfe MS. Intramembrane proteolysis bypresenilin and presenilin-like proteases. J Cell Sci2003; 116:2839-44.

59. Selkoe D, Kopan R. Notch and presenilin: regulatedintramembrane proteolysis links development anddegeneration. Annu Rev Neurosci 2003; 26: 565-97.

60. Jarriault S, Brou C, Logeat F, et al. Signalling down-stream of activated mammalian Notch. Nature 1995;377:355-8.

61. de la Pompa JL, Wakeham A, Correia KM, et al.Conservation of the Notch signalling pathway inmammalian neurogenesis. Development 1997;124:1139-48.

62. Kimberly WT, Zheng JB, Guenette SY, Selkoe DJ.The intracellular domain of the beta-amyloid precur-sor protein is stabilized by Fe65 and translocates tothe nucleus in a notch-like manner. J Biol Chem2001; 276:40288-92.

63. Kimberly WT, Zheng JB, Town T, et al.Physiological regulation of the beta-amyloid precur-




sor protein signaling domain by c-Jun N-terminalkinase JNK3 during neuronal differentiation. JNeurosci 2005; 25:5533-43.

64. Cao X, Sudhof TC. Dissection of amyloid-beta pre-cursor protein-dependent transcriptional transactiva-tion. J Biol Chem 2004; 279:24601-11.

65. Leissring MA, Murphy MP, Mead TR, et al. A phys-iologic signaling role for the gamma -secretase-derived intracellular fragment of APP. Proc NatlAcad Sci U S A 2002; 99:4697-702.

66. Donoviel DB, Hadjantonakis AK, Ikeda M, et al. Micelacking both presenilin genes exhibit early embryonicpatterning defects. Genes Dev 1999; 13:2801-10.

67. Perez-Tur J, Froelich S, Prihar G, et al. A mutation inAlzheimer’s disease destroying a splice acceptor site inthe presenilin-1 gene. Neuroreport 1995; 7:297-301.

68. Duering M, Grimm MO, Grimm HS, et al. Mean ageof onset in familial Alzheimer’s disease is determinedby amyloid beta 42. Neurobiol Aging 2005; 26:785-8.

69. Lleo A, Berezovska O, Herl L, et al. Nonsteroidalanti-inflammatory drugs lower Abeta42 and changepresenilin 1 conformation. Nat Med 2004; 10:1065-6.

70. Berezovska O, Lleo A, Herl LD, et al. FamilialAlzheimer’s disease presenilin 1 mutations causealterations in the conformation of presenilin andinteractions with amyloid precursor protein. JNeurosci 2005; 25:3009-17.

71. Nakaya Y, Yamane T, Shiraishi H, et al. Randommutagenesis of presenilin-1 identifies novel mutants

exclusively generating long amyloid beta-peptides. JBiol Chem 2005; 280:19070-7.

72. Walker ES, Martinez M, Brunkan AL, Goate A.Presenilin 2 familial Alzheimer’s disease mutationsresult in partial loss of function and dramaticchanges in Abeta 42/40 ratios. J Neurochem 2005;92:294-301.

73. Esposito L, Gan L, Yu GQ, Essrich C, Mucke L.Intracellularly generated amyloid-beta peptide coun-teracts the antiapoptotic function of its precursor pro-tein and primes proapoptotic pathways for activation

by other insults in neuroblastoma cells. J Neurochem2004; 91:1260-74.74. Wolozin B, Alexander P, Palacino J. Regulation of

apoptosis by presenilin 1. Neurobiol Aging 1998;19:S23-7.

75. Guo Q, Furukawa K, Sopher BL, et al. Alzheimer’sPS-1 mutation perturbs calcium homeostasis andsensitizes PC12 cells to death induced by amyloidbeta-peptide. Neuroreport 1996; 8:379-83.

76. Weihl CC, Ghadge GD, Kennedy SG, et al. Mutantpresenilin-1 induces apoptosis and downregulatesAkt/PKB. J Neurosci 1999; 19:5360-9.

77. Tanii H, Ankarcrona M, Flood F, et al. Alzheimer’sdisease presenilin-1 exon 9 deletion and L250Smutations sensitize SH-SY5Y neuroblastoma cells to

hyperosmotic stress-induced apoptosis. Neuro-science. 2000; 95:593-601.

78. Popescu BO, Cedazo-Minguez A, Popescu LM, et al.Caspase cleavage of exon 9 deleted presenilin-1 is anearly event in apoptosis induced by calciumionophore A 23187 in SH-SY5Y neuroblastoma

cells. J Neurosci Res 2001; 66:122-34.79. Czech C, Lesort M, Tremp G, et al. Characterization

of human presenilin 1 transgenic rats: increased sen-sitivity to apoptosis in primary neuronal cultures.Neuroscience 1998; 87:325-36.

80. Bursztajn S, DeSouza R, McPhie DL, et al.Overexpression in neurons of human presenilin-1 or apresenilin-1 familial Alzheimer’s disease mutant doesnot enhance apoptosis. J Neurosci 1998; 18:9790-9.

81. Kovacs DM, Mancini R, Henderson J, et al.Staurosporine-induced activation of caspase-3 ispotentiated by presenilin 1 familial Alzheimer’s dis-ease mutations in human neuroglioma cells. JNeurochem 1999; 73:2278-85.

82. Keller JN, Guo Q, Holtsberg FW, et al. Increased sen-sitivity to mitochondrial toxin-induced apoptosis inneural cells expressing mutant presenilin-1 is linkedto perturbed calcium homeostasis and enhancedoxyradical production. J Neurosci 1998; 18:4439-50.

83. Chan SL, Culmsee C, Haughey N, et al. Presenilin-1mutations sensitize neurons to DNA damage-induceddeath by a mechanism involving perturbed calciumhomeostasis and activation of calpains and caspase-

12. Neurobiol Dis 2002; 11:2-19.84. Cedazo-Minguez A, Popescu BO, Ankarcrona M,

et al. The presenilin 1 deltaE9 mutation givesenhanced basal phospholipase C activity and a result-ant increase in intracellular calcium concentrations. JBiol Chem 2002; 277:36646-55.

85. Wolozin B, Iwasaki K, Vito P, et al. Participation of presenilin 2 in apoptosis: enhanced basal activityconferred by an Alzheimer mutation. Science 1996;274:1710-3.

86. Araki W, Yuasa K, Takeda S, et al. Overexpression of

presenilin-2 enhances apoptotic death of cultured cor-tical neurons. Ann N Y Acad Sci 2000; 920:241-4.87. Mori M, Nakagami H, Morishita R, et al. N141I

mutant presenilin-2 gene enhances neuronal celldeath and decreases bcl-2 expression. Life Sci 2002;70:2567-80.

88. Guo Q, Sopher BL, Furukawa K, et al. Alzheimer’spresenilin mutation sensitizes neural cells to apopto-sis induced by trophic factor withdrawal and amyloidbeta-peptide: involvement of calcium and oxyradi-cals. J Neurosci 1997; 17:4212-22.

89. Ito E, Oka K, Etcheberrigaray R, et al. Internal Ca2 +mobilization is altered in fibroblasts from patientswith Alzheimer’s disease. Proc Natl Acad Sci U S A1994; 91:534-8.

242 M. Ankarcrona



90. Hirashima N, Etcheberrigaray R, Bergamaschi S,et al. Calcium responses in human fibroblasts: adiagnostic molecular profile for Alzheimer’s dis-ease. Neurobiol Aging 1996; 17:549-55.

91. Leissring MA, Akbari Y, Fanger CM, et al.Capacitative calcium entry deficits and elevated

luminal calcium content in mutant presenilin-1knockin mice. J Cell Biol 2000; 149:793-8.

92. Guo Q, Fu W, Sopher BL, et al. Increased vulnera-bility of hippocampal neurons to excitotoxic necro-sis in presenilin-1 mutant knock-in mice. Nat Med1999; 5:101-6.

93. Leissring MA, Paul BA, Parker I, et al. Alzheimer’spresenilin-1 mutation potentiates inositol 1,4,5-trisphosphate-mediated calcium signaling inXenopus oocytes. J Neurochem 1999; 72:1061-8.

94. Stutzmann GE, Caccamo A, LaFerla FM, Parker I.Dysregulated IP3 signaling in cortical neurons of knock-in mice expressing an Alzheimer’s-linkedmutation in presenilin1 results in exaggerated Ca2 +signals and altered membrane excitability. JNeurosci 2004; 24:508-13.

95. Querfurth HW, Selkoe DJ. Calcium ionophoreincreases amyloid beta peptide production by cul-tured cells. Biochemistry 1994; 33:4550-61.

96. Mattson MP, Cheng B, Davis D, et al. beta-Amyloidpeptides destabilize calcium homeostasis and ren-der human cortical neurons vulnerable to excitotox-icity. J Neurosci 1992; 12:376-89.

97. Arispe N, Rojas E, Pollard HB. Alzheimer’s diseaseamyloid beta protein forms calcium channels in bilayermembranes: blockade by tromethamine and aluminum.Proc Natl Acad Sci U S A 1993; 90:567-71.

98. Rhee SK, Quist AP, Lal R. Amyloid beta protein-(1-42) forms calcium-permeable, Zn2 +-sensitivechannel. J Biol Chem 1998; 273:13379-82.

99. Kagan BL, Hirakura Y, Azimov R, et al. The chan-nel hypothesis of Alzheimer’s disease: current sta-tus. Peptides 2002; 23:1311-5.

100. Bhatia R, Lin H, Lal R. Fresh and globular amyloid

beta protein (1-42) induces rapid cellular degenera-tion: evidence for AbetaP channel-mediated cellulartoxicity. FASEB J 2000; 14:1233-43.

101. Kayed R, Sokolov Y, Edmonds B, et al.Permeabilization of lipid bilayers is a common con-formation-dependent activity of soluble amyloidoligomers in protein misfolding diseases. J BiolChem 2004; 279:46363-6.

102. Mattson MP, Cheng B, Culwell AR, et al. Evidencefor excitoprotective and intraneuronal calcium-reg-ulating roles for secreted forms of the beta-amyloidprecursor protein. Neuron 1993; 10:243-54.

103. Furukawa K, Sopher BL, Rydel RE, et al. Increasedactivity-regulating and neuroprotective efficacy of alpha-secretase-derived secreted amyloid precursor

protein conferred by a C-terminal heparin-bindingdomain. J Neurochem 1996; 67:1882-96.

104. Guo Q, Robinson N, Mattson MP. Secreted beta-amyloid precursor protein counteracts the proapop-totic action of mutant presenilin-1 by activation of NF-kappaB and stabilization of calcium homeosta-

sis. J Biol Chem 1998; 273:12341-51.105. Pickel VM, Clarke CL, Meyers MB. Ultrastructural

localization of sorcin, a 22 kDa calcium bindingprotein, in the rat caudate-putamen nucleus: associ-ation with ryanodine receptors and intracellularcalcium release. J Comp Neurol 1997; 386:625-34.

106. Gracy KN, Clarke CL, Meyers MB, Pickel VM.N-methyl-D-aspartate receptor 1 in the caudate-putamen nucleus: ultrastructural localization andco-expression with sorcin, a 22,000 mol. wt cal-cium binding protein. Neuroscience 1999; 90:107-17.

107. Buxbaum JD, Choi EK, Luo Y, et al. Calsenilin: acalcium-binding protein that interacts with the pre-senilins and regulates the levels of a presenilin frag-ment. Nat Med 1998; 4:1177-81.

108. Jo DG, Kim MJ, Choi YH, et al. Pro-apoptoticfunction of calsenilin/DREAM/KChIP3. FASEB J2001; 15:589-91.

109. Jo DG, Chang JW, Hong HS, et al. Contribution of presenilin/gamma-secretase to calsenilin-mediatedapoptosis. Biochem Biophys Res Commun 2003;305:62-6.

110. Jo DG, Jang J, Kim BJ, Lundkvist J, Jung YK.Overexpression of calsenilin enhances gamma-sec-retase activity. Neurosci Lett 2005; 378:59-64.

111. Lilliehook C, Bozdagi O, Yao J, et al. Altered Abetaformation and long-term potentiation in a calsenilinknock-out. J Neurosci 2003; 23:9097-106.

112. Leissring MA, Yamasaki TR, Wasco W, et al.Calsenilin reverses presenilin-mediated enhancementof calcium signaling. Proc Natl Acad Sci U S A2000; 97:8590-3.

113. Maruyama K, Usami M, Kametani F, et al.

Molecular interactions between presenilin and cal-pain: inhibition of m-calpain protease activity bypresenilin-1, 2 and cleavage of presenilin-1 by m-,mu-calpain. Int J Mol Med 2000; 5:269-73.

114. van de Craen M, de Jonghe C, van den Brande I,et al. Identification of caspases that cleave prese-nilin-1 and presenilin-2. Five presenilin-1 (PS1)mutations do not alter the sensitivity of PS1 to cas-pases. FEBS Lett 1999; 445:149-54.

115. Vito P, Ghayur T, D’Adamio L. Generation of anti-apoptotic presenilin-2 polypeptides by alternativetranscription, proteolysis, and caspase-3 cleavage. JBiol Chem 1997; 272:28315-20.

116. Vezina J, Tschopp C, Andersen E, Muller K.Overexpression of a C-terminal fragment of




presenilin 1 delays anti-Fas induced apoptosis inJurkat cells. Neurosci Lett 1999; 263:65-8.

117. Alves da Costa C, Paitel E, Mattson MP, et al. Wild-type and mutated presenilins 2 trigger p53-dependentapoptosis and down-regulate presenilin 1 expressionin HEK293 human cells and in murine neurons. Proc

Natl Acad Sci U S A 2002; 99:4043-8.118. Araki W, Yuasa K, Takeda S, et al. Pro-apoptotic

effect of presenilin 2 (PS2) overexpression is asso-ciated with down-regulation of Bcl-2 in culturedneurons. J Neurochem 2001; 79:1161-8.

119. Roperch JP, Alvaro V, Prieur S, et al. Inhibition of presenilin 1 expression is promoted by p53 andp21WAF-1 and results in apoptosis and tumor sup-pression. Nat Med 1998; 4:835-8.

120. Hong CS, Caromile L, Nomata Y, et al. Contrastingrole of presenilin-1 and presenilin-2 in neuronal dif-ferentiation in vitro. J Neurosci 1999; 19:637-43.

121. Farmery MR, Tjernberg LO, Pursglove SE, et al.Partial purification and characterization of gamma-secretase from post-mortem human brain. J BiolChem 2003; 278:24277-84.

122. Brockhaus M, Grunberg J, Rohrig S, et al. Caspase-mediated cleavage is not required for the activity of presenilins in amyloidogenesis and NOTCH signal-ing. Neuroreport 1998; 9:1481-6.

123. LaFerla FM, Troncoso JC, Strickland DK, KawasCH, Jay G. Neuronal cell death in Alzheimer’s dis-ease correlates with apoE uptake and intracellular

Abeta stabilization. J Clin Invest 1997; 100:310-20.124. Lustbader JW, Cirilli M, Lin C, et al. ABAD

directly links Abeta to mitochondrial toxicity inAlzheimer’s disease. Science 2004; 304:448-52.

125. Crouch PJ, Blake R, Duce JA, et al. Copper-depend-ent inhibition of human cytochrome c oxidase by adimeric conformer of amyloid-beta1-42. J Neurosci2005; 25:672-9.

126. Casley CS, Canevari L, Land JM, et al. Beta-amy-loid inhibits integrated mitochondrial respirationand key enzyme activities. J Neurochem 2002; 80:

91-100.127. Canevari L, Clark JB, Bates TE. beta-Amyloid frag-ment 25-35 selectively decreases complex IV activ-ity in isolated mitochondria. FEBS Lett 1999;457:131-4.

128. Yan SD, Stern DM. Mitochondrial dysfunction andAlzheimer’s disease: role of amyloid-beta peptidealcohol dehydrogenase (ABAD). Int J Exp Pathol2005; 86:161-71.

129. Cardoso SM, Santos S, Swerdlow RH, Oliveira CR.Functional mitochondria are required for amyloidbeta-mediated neurotoxicity. FASEB J 2001; 15:1439-41.

130. Rodrigues CM, Sola S, Brito MA, et al. Amyloidbeta-peptide disrupts mitochondrial membrane lipid

and protein structure: protective role of taurour-sodeoxycholate. Biochem Biophys Res Commun2001; 281:468-74.

131. Abramov AY, Canevari L, Duchen MR. Beta-amy-loid peptides induce mitochondrial dysfunction andoxidative stress in astrocytes and death of neuronsthrough activation of NADPH oxidase. J Neurosci2004; 24:565-75.

132. Moreira PI, Santos MS, Moreno A, et al. Effect of amyloid beta-peptide on permeability transitionpore: a comparative study. J Neurosci Res 2002;69:257-67.

133. Parks JK, Smith TS, Trimmer PA, et al. NeurotoxicAbeta peptides increase oxidative stress in vivothrough NMDA-receptor and nitric-oxide-synthasemechanisms, and inhibit complex IV activity andinduce a mitochondrial permeability transitionin vitro. J Neurochem 2001; 76:1050-6.

134. Zhang Y, McLaughlin R, Goodyer C, LeBlanc A.Selective cytotoxicity of intracellular amyloid betapeptide1-42 through p53 and Bax in cultured pri-mary human neurons. J Cell Biol 2002; 156:519-29.

135. Kim HS, Lee JH, Lee JP, et al. Amyloid beta pep-

tide induces cytochrome C release from isolatedmitochondria. Neuroreport 2002; 13:1989-93.

136. Yamaguchi H, Yamazaki T, Ishiguro K, et al.Ultrastructural localization of Alzheimer amyloidbeta/A4 protein precursor in the cytoplasm of neu-rons and senile plaque-associated astrocytes. ActaNeuropathol (Berlin) 1992; 85:15-22.

137. Anandatheerthavarada HK, Biswas G, Robin MA,Avadhani NG. Mitochondrial targeting and a noveltransmembrane arrest of Alzheimer’s amyloid pre-cursor protein impairs mitochondrial function in

neuronal cells. J Cell Biol 2003; 161:41-54.138. Hansson CA, Frykman S, Farmery MR, et al.Nicastrin, presenilin, APH-1, and PEN-2 formactive gamma-secretase complexes in mitochon-dria. J Biol Chem 2004; 279:51654-60.

139. Caspersen C, Wang N, Yao J, et al. Mitochondrial A β:a potential focal point for neuronal metabolic dys-function in Alzheimer’s disease. FASEB J. 2005; 19:2040-2041.

244 M. Ankarcrona



15Immunotherapeutic Approaches

to Alzheimer’s DiseaseJosef Karkos

245

15.1 Concept of Immunotherapyfor Alzheimer’s Disease

The concept of immunotherapy for Alzheimer’sdisease (AD) is based on the molecular findingsthat place AD within the group of disorders called“protein-misfolding diseases.” These disorders arecaused by conformational changes coupled withthe aggregation of misfolded proteins outside of the cell [1–4]. The concept emerged after theresearch group of Salomon [5–8] demonstrated that

the immunologic approach in vitro was successfulin inducing conformational changes in both antigenand antibody. In particular, it was demonstratedthat the monoclonal antibodies were capable of stabilizing the conformation of an antigen againstincorrect folding and recognize an incompletelyfolded epitope, inducing native conformation in apartially unfolded protein.

Support for the in vivo relevance of the concepthas been provided by experiments published by theSchenk’s research group [9]. They found that vacci-nation of a transgenic mouse expressing the humanβ-amyloid protein with the β-amyloid peptide(Aβ42) significantly decreased the β-amyloid bur-den in areas of the brain important for cognition andmemory. Furthermore, the studies carried out bySchenk’s group indicated that the effect of the A β42peptide was mediated by antibodies it induced [10].

The functional relevance of the findings reportedby Schenk’s group was demonstrated in separated,independent follow-up studies carried out by Janus

and Morgan and their colleagues [11, 12]. Theyshowed that the β-amyloid peptide vaccine was

able to protect transgenic mice from the memory

deficits they normally develop and to amelioratethe preexisting behavioral and memory deficits.

After promising preclinical studies in severalspecies, clinical trials were initiated using A β42(vaccine’s name: AN-1792) in conjunction with theadjuvant QS-21 [13]. Despite numerous adverseeffects that occurred in some patients that led tosuspension of the study, preliminary data demon-strated that vaccination can reduce AD pathologyand mitigate progressive cognitive decline associ-ated with the disease.

The experimental and clinical data obtained todate indicate that the induction of the systemicadaptive response to A β42 is an effective way toinduce its clearance [14–17], supporting the amy-loid cascade hypothesis of AD and implying thatAβ42 deposition is driving the disease pathogenesis[18, 19]. Consistent with this hypothesis is therecent finding that the accumulation of A β is ableto induce the development of tau pathology [20].Aβ immunotherapy reduces first A β deposits and

subsequently clears aggregates of tau-protein [21].

15.2 Immune Responses to A β

15.2.1 Molecular Structure andImmunological Properties of A βIn the A β structure, two domains can be discrimi-nated: the N-terminal domain that encompassesamino acids 1 to 28 and C-terminal domain from

amino acids 29 to 42. In aqueous solution, theN-terminal region exhibits different conformations



and solubility properties depending on environ-mental conditions [22, 23]. The hydrophobicregion in the C-terminal domain forms a β-strandstructure in aqueous solutions, independently of pH and temperature. The amino acids sequences inthe N-terminal domain permit the existence of adynamic equilibrium between the α -helix and theβ-strand conformations. In addition, results of in vitro experiments indicate a steady-state equi-librium between A β in plaques and in solution[24]. The most important conclusion from experi-ments in vitro is that amyloid formation might besubjected to modulation in terms of changes inconformation.

The A β molecule exhibits antigenic andimmunogenic properties. Most of the A β42-anti-

body-producing epitopes were detected in theN-terminal region of the peptide A β42. The pre-dominance of T-cell epitopes lies in the central tocarboxy-terminal region of the peptide. Thereported differences in the location of epitopeswithin the A β peptide depend on the differentlength of the peptides used for the detection of epi-topes. The effects of antibody binding to variousepitopes may be different. As A β42 exists both insoluble and fibrillar forms, antibodies generatedagainst this antigen may recognize differentimmunogenic structures within it. It is important toidentify within A β42 antigenic determinants for Band T cells in order to design the most effectivevaccine.

Because the dominant B-cell and T-cell epitopeshave distinct location, the humoral and cellularimmune responses may be modulated. The modu-lation can be achieved for instance by using anantigen and various adjuvant combinations.Because the type of immune response generated

may be critical to the efficacy and safety of a poten-tial vaccine, a careful examination of the overallimmune response, especially of the T h1 and T h2responses, is of great importance [25].

15.2.2 Innate Immunoresponses to A βNaturally occurring anti-A β antibodies (autoanti-bodies) were found in plasma in the elderly popu-lation [26]. There were detectable but very lowlevels of anti-amyloid antibodies in just over 50%of all samples and modest levels in under 5% of allsamples. However, neither the presence nor the

level of anti-amyloid- β antibodies correlated withthe likelihood of developing dementia or withplasma levels of amyloid- β peptide. These findingssuggest that low levels of anti-amyloid- β autoanti-bodies are frequent in the elderly population but donot confer protection against developing dementia.

Another group detected anti-amyloid- β autoanti-bodies in the CSF of AD patients [27, 28]. Thetiters of the antibodies were significantly lower inAD patients than in age-matched controls. Thesedata indicate an impaired or reduced ability to gen-erate antibodies specific against AD. This hypoth-esis has been supported by the finding thattreatment of individuals with intravenous immuno-globulin preparation containing anti-A β antibodiesincrease both CSF and serum levels of anti-A β

antibodies and significantly lowered CSF levels,possibly by facilitating transport of A β from theCSF to the serum [29]. These findings suggest thathuman A β antibodies are able to lower the A β con-centration in the CSF, which may reduce A β depo-sition in brain. It seems that A β is recognized in theCNS as a molecule that needs to be cleared andprovokes activation of microglia and astrocytes.The innate immunoresponse is also supported bysuch findings in AD patients as activation of com-plement; secretion of proinflammatory cytokinessuch as interleukin (IL)-1 β and tumor necrosis fac-tor (TNF)- α ; expression of chemokines MIP-1 α ,MIP-1 β, and MCP-1; and the secretion of nitricoxide [30, 31].

Monsonego et al. [32] found that some healthy,elderly individuals, as well as individuals with AD,possess elevated baseline levels of A β-reactiveT-cells. While the general trend was toward adiminished immune response with aging, thisdemonstrates a selective increase in A β-reactive

T cells in older individuals with and withoutdementia. The reason for this selective expansionof A β-reactive cells in elderly individuals isunclear. T-cell reactivity may be considered as anendogenous reaction to A β deposition in the brainin the context of the local innate immune responsethat occurs in AD [32].

The epitopes for A β-reactive T cells in humansare primarily amino acids 16–42. As in studies of active immunization of humans and of mouse mod-els of AD, the primary epitope to which antibodiesare generated are residues 1–12 [33]. There existsthe possibility to influence both epitopes separately.

246 J. Karkos



The function of microglia in AD seems also tobe impaired. The role of microglia cells as a prin-ciple immune effector and phagocytic cells in theCNS is established. These cells are associated withplaques containing fibrillar β-amyloid found in thebrains of AD patients. The plaque-associated gliaundergo a phenotypic conversion into an activatedphenotype. It is believed that microglia are respon-sible for the development of a focal inflammatoryresponse that exacerbates and accelerates diseaseprocess. However, despite the presence of abundantactivated microglia in the brains of AD patients,these cells fail to mount a phagocytic response toAβ deposits but can efficiently phagocytose A βfibrils and plaques in vitro. It remains unclear whythe plaque-associated microglia in vivo are unable

to effectively phagocytose the amyloid depositsdespite their close physical vicinity to the plaques[34]. It could be assumed that other plaque con-stituents block the interaction of the microglia withthe plaque, as has been suggested for C1q [35].

15.2.3 Adaptive Immune Responseto A β15.2.3.1 Experience in Transgenic Animals

Although AD is associated with local innateimmune responses, they are not sufficient to protectagainst the development of the disease or to atten-uate the disease progression. The induction of sys-temic adaptive immune responses to A β in mousemodels of AD has been found to be beneficial forboth the neuropathologic and behavioral changesthat these mice develop.

Active immunization with synthetic A β peptideor passive transfer with A β antibodies has been

shown to prevent and reduce the cerebral amyloidload [9, 36, 37]. Using similar experimental settings,improvements in cognitive deficits in APP andAPP/PS1 transgenic mice were observed [11, 12, 38,39]. Schenk et al. [9] reported for the first time thatintraperitoneal injections of A β42 peptide, with com-plete or incomplete Freund`s adjuvant, almost com-pletely prevented plaque deposition when givenbefore initiation of plaque formation and signifi-cantly lowered cerebral levels if given after the initi-ation of plaque deposition in PDAPP transgenicmice. Evidence has been provided that the antibod-ies generated by active immunization with A β pep-

tide recognized an epitope within the amino-termi-nus of the A β protein [37, 40–44]. Active immu-nization was shown to be less effective in reducingcerebral A β levels in very old APP transgenic micewith abundant cerebral A β plaques [45].

Passive administration of selected A β antibod-ies achieved similar effects to active immuniza-tion [36]. Passive transfer with a monoclonalantibody directed at the midregion of A β (mAb266, recognizing A β13-28 ) has been shown to lowercerebral levels while increasing A β levels in theblood [46]. When a single dose of A β mAb 266was passively administered to aged transgenicmice, no reduction in A β levels in brain wasfound, nevertheless improvements in cognitivedeficits were observed [38].

Since the first report on the effect of immuno-therapy in animals, several formulations of A βhave been investigated, for example, geneticallyengineered filamentous phages displaying A β3-6(EFRH) [47], intranasal A β immunization [37, 41],a soluble non-amyloidogenic, nontoxic homologueof A β [48], microencapsulated A β [49], andrecombinant adeno-associated virus A β vaccineexpressing a fusion protein containing A β42 andcholera toxin B subunit [50]. Irrespective of theway of administration and the animal species used(mice, rabbits, guinea-pigs), the immunizationentailed reductions in cerebral amyloid load andimprovements in behavior.

Lemere et al. [51] immunized for the first time anon-human primate, the vervet monkey, with acocktail of human A β peptides (A β40, Aβ42). Thismonkey species develops cerebral amyloid plaqueswith aging, and the amyloid deposits are associatedwith gliosis and neuritic dystrophy. Immunizedanimals generated anti-A β antibodies that labeled

Aβ plaques in human, transgenic mouse, andvervet. Anti-A β antibodies bound to A β1-7 epitopeand recognized monomeric and oligomeric A β butnot full-length APP or C-terminal fragments of APP. The A β levels in the CNS were reduced,whereas they were increased in plasma. This find-ing confirms that A β can be moved from the cen-tral to peripheral compartment where the anti-A βantibodies bind them, enhancing clearance of A β[46]. In an experiment by Lemere et al. [51],immunization did not elicit any side effects. In par-ticular, no A β-reactive T-cell populations weredetected.

15. Immunotherapeutic Approaches to Alzheimer’s Disease 247



Plaque clearance can be invoked only by anti-bodies against epitopes located in the N-terminalregion of A β [52]. It has also been shown that theisotype of the antibody prominently influences thedegree of plaque clearance. For example, IgG2aantibodies against A β were more efficient thatIgG1 or IgG2b antibodies in reducing pathology.Moreover, it was shown that the high affinity of theantibody for Fc receptors on microglial cells seemsto be more important than high affinity for A βitself and that complement activation is notrequired for plaque clearance.

It was reported [53] that after intracranial anti-Aβ antibody injections into APP transgenic mice,there is a rapid removal of diffuse amyloid depositsapparently independent of microglial activation

and also a later removal of compact amyloiddeposits, which appears to require microglial acti-vation. After suppression of microglial activationwith dexamethasone, administration of anti-A βantibody inhibited the removal of compact,thioflavine-S-positive amyloid deposits [54].

Wilcock et al. [55] using antibody 2286 (mousemonoclonal anti-human A β28-40 IgG1) for passiveimmunization in a transgenic mouse model showedthat the antibody is able to enter the brain and bindto the amyloid deposits, likely opsonizing the A βand resulting in Fc γ receptor-mediated phagocyto-sis. This group also showed that passive immuniza-tion improved behavioral performance. Suchimprovement might reflect rapid reduction of theAβ pool, closely linked to memory impairmentsyet not easily detected by immunochemistry. Asimilar phenomenon was previously reported byDodart et al. [38] and Kotilinek et al. [39]. Theyobserved rapid reversal of memory deficits intransgenic mice after passive immunization with-

out significant reduction in brain A β.The clearance of various types of amyloid

plaque depends on the isotype of the administeredantibody [56]. It was shown that IgG2a antibodiesare efficacious in clearing fibrillar, thio-S-positiveplaque. The high efficacy of IgG2a antibodies isconsistent with their ability to best stimulatemicroglial and peripheral macrophage phagocyto-sis. This finding also supports a crucial role formicroglial Fc receptor-mediated phagocytosis inthe clearance of at least fibrillar plaques. However,because Fc knockout mice show a reduction of plaque burden after A β immunotherapy [57], alter-

native clearing mechanisms should be taken inconsideration.

Mechanisms by which antibodies act are notentirely understood. Suggested mechanismsinclude (i) microglial-mediated phagocytosis (Fc-dependent, Fc-independent, or combination of Fc-dependent and Fc-independent mechanisms[53–55, 58]), and β1 integrin-dependent [59]; (ii)direct interaction of antibodies with A β with sub-sequent disaggregation of amyloid deposits [8, 53,55]; and (iii) removal of A β from the brain bybinding circulating A β in plasma with the anti-A βantibodies (so-called peripheral sink hypothesis)[38, 46, 60].

All three proposed mechanisms of anti-A β anti-body-mediated amyloid removal are not mutually

exclusive. They are likely to be synergistic if mul-tiple mechanisms are elicited by a single antibodyor serum. Other possible mechanisms of amyloidremoval would include activation of scavengerreceptors [61, 62] or receptors for advanced glyca-tion end products [63].

The effect of immunization on vascular A βdeposits has recently been addressed [64]. Thisissue seems to be important in light of a studyshowing that passive immunization of APP23transgenic mice, characterized by prominent vascu-lar A β deposition, with anti-A β IgG1 antibody,resulted in a twofold increase in the rate of hemor-rhages [65]. To better understand this potential sideeffect, Racke et al. [64] characterized the bindingproperties of several monoclonal anti-A β antibod-ies to deposited A β in brain parenchyma and cere-bral vessels (CAA; cerebral amyloid angiopathy).They observed an increase in both the incidenceand severity of CAA-associated microhemorrhageswhen PDAPP transgenic mice were treated with N-

terminally directed 3D6 antibody, whereas micetreated with central domain antibody 266 wereunaffected. In this context, the question ariseswhether the amyloid angiitis that has been recentlyreported [66] would augment the risk of such hem-orrhages. Taken together, circulating antibodieselicited by active immunization or administeredpassively cross the blood-brain barrier [67, 68].Moreover, administration to transgenic animals of monoclonal A β antibodies against defined A β epi-topes reduces plaque burden and improves cogni-tive deficits to the same degree as activeimmunization [8].

248 J. Karkos



Assessment of morphological and behavioralchanges in animals is a very important issue forcomparative purposes and for effectivity and safetymeasurements of investigated agents. Assessmentof behavioral deficits observed in transgenic micemay be particularly difficult, because these deficitsare only in part related to amyloid deposition. Ashistological analyses by Dodart et al. [69] indicate,the behavioral deficits are also related to neu-roanatomical alterations secondary to overexpres-sion of the APP transgene and are independent of amyloid deposition.

Gandy and Walker [70] suggest the use of non-human primates as adjunctive models for assessingthe efficacy and safety of immunotherapeutics forAD. Use of this animal model could contribute to

further clarification of potential damage caused byimmunization to the cerebral vessels.

15.2.3.2 Clinical Experience: HumanTrials of A β Vaccination

The finding that active and passive vaccinationwith A β exerts remarkable A β-reducing effects inanimal models of AD led to clinical trials in whichan A β42 synthetic peptide was administered par-enterally with a previously tested adjuvant (QS-21)to patients with mild to moderate AD.

In a long-term phase I clinical trial [71], thesafety, tolerability, and immunogenicity of AN1792(human aggregated A β42) and exploratory evi-dence of efficacy in patients with mild to moderateAD were evaluated. Twenty patients were enrolledinto each of four dose groups and randomlyassigned to receive intramuscularly AN1792 (50 or225 µg with QS-21 adjuvant 50 or 100 µg) or QS-21 only (control) in a 4:1 active-control ratio on day

0 and at weeks 4, 12, and 24. Patients were allowedto receive up to four additional injections of polysorbate 80 modified formulation at weeks 36,48, 60, and 72.

During the period of the first four injections,23.4% of AN1792-treated patients had a positiveanti-AN1792 antibody titer (an anti-AN1792 anti-body titre of ≥1:1000). This increased to 58.8%after additional injections with the modified for-mulation. With regard to efficacy, DisabilityAssessment for Dementia scores showed lessdecline among active compared with controlpatients at week 84 (p = 0.002).

No treatment differences were observed in threeother efficacy measures. Treatment-related sideeffects were reported in 19 (23.8%) patients, but norelationship was observed between AN1792 doseand their incidence. One patient developed menin-goencephalitis 219 days after discontinuing fromthe study. Diagnostics of meningoencephalitis wasmade postmortem, and the cause of death was con-sidered non-treatment related. Another five deathsoccurred during the study follow-up, but none wasdeemed directly related to study treatment.

Although no severe side effects occurred duringthe course of the phase I trials, phase IIa trialswere halted when 18 of 298 patients immunizedwith AN-1792 presented with symptoms consis-tent with meningoencephalitis [72]. The symp-

toms and signs of encephalitis included headache,confusion, and changes on magnetic resonanceimaging scans. Of the 18 patients in the phase IIstudy, 12 have returned to their baseline status andsix have experienced some type of prolonged neu-rological deficit. The majority of patients had IgGresponses to A β, and all patients mounted at leasta small IgM response. There was no correlation of the severity of encephalitis with either the level orepitope specificity of the antibody response.Moreover, the vast majority of individuals whomounted the antibody response to A β did notdevelop encephalitis.

A cohort of 30 patients who participated in thephase IIa multicenter trial was followed up aftersuspension of treatment [73]. The group of patientswho generated antibodies against β-amyloidshowed a marked and long-lasting increase inserum antibodies against aggregated A β42 in bothIgG and IgM classes.

AD patients who generated antibodies against

Aβ performed markedly better on the Mini MentalState Examination (MMSE) 8 months and 1 yearafter the immunization, as compared with controlpatients, and they remained unchanged after 1 year,as compared with baseline. Within this period,patients in the control group worsened signifi-cantly. Taken together, the patients who generatedantibodies exhibited slower rates of cognitivedecline 1 year after the last immunization.

The neuropathologic findings in 3 patients whoreceived AN1792/QS21 were reported to date[74–76]. Nicoll et al. [76] found infiltrates of lym-phocytes in the leptomeninges that were identified




as being composed of T lymphocytes (CD3 + andCD45RO +); the majority were CD4 + and very fewwere CD8 +. B lymphocytes were not present. Thelarge areas of neocortex contained very few A βplaques or they were devoid of plaques. In someregions devoid of plaques, A β-immunoreactivitywas associated with microglia immunoreactive forCD68 and human leukocyte antigen DR. Moreover,in the neocortical areas devoid of plaques, densitiesof tangles, neuropil threads, and cerebral amyloidangiopathy similar to unimmunized AD patientswere found. The plaque-associated dystrophic neu-rites and astrocyte clusters were not seen. Atimmunohistochemistry, the plaques were sur-rounded by IgG and C3 complement. Interestingly,cerebral white matter showed marked reduction in

the density of myelinated fibers and extensive infil-tration with macrophages that were not immunos-tained for A β.

Neuropathological data reported by Ferrer et al.[74] showed some differences in comparison withthe above described case. A focal depletion of dif-fuse and neuritic plaques was observed, but not of amyloid angiopathy. In the cerebral white matter,there was loss of myelin that was accompanied bymoderate microgliosis and astrogliosis. Moreover,multinucleated giant cells filled with dense A β

42and A4 β40 were seen.Interestingly, severe small cerebral blood vessel

lesion (lipohyalinosis) and multiple cortical hemor-rhages, including acute lesions and lesions withmacrophages filled with hemosyderin, were found.Focal inflammatory infiltrates were seen in themeninges as well as in the cerebrum and they werecomposed mostly of CD8 +, less often of CD4 +,CD3 +, CD5 +, and, rarely, CD7 + lymphocytes. Blymphocytes and the detected T cytotoxic markers

were negative.Masliah et al. [75] reported the results of neu-

ropathologic examination of the patient withoutclinical symptoms and signs of meningoencephali-tis. They found that vaccination with A β42 resultedin a considerable reduction of plaque burden andpromoted amyloid phagocytosis in the frontal cor-tex and to a lesser extent in the temporal lobe.Plaque associated neuritic dystrophy in the frontalcortex was undetectable. Neurofibrillary pathologyand CAA were unchanged. Only minimal lympho-cytic reaction was observed in the leptomeningesand the white matter was unaffected.

In summary, it can be said that the clinical andpathologic data of these two trials support the con-cept of using immunization in the treatment of AD.However, many questions remain unanswered.First, the responder population needs to be charac-terized. Indeed, assuming that the anti-A β antibod-ies mediate the reduction in the observed amyloidpathology, only about half of the patients benefitfrom the treatment. Second, the risk to benefit ratiocannot be determinated until an analysis of thephase IIa trial data is completed and the pathogen-esis of the side effects is definitively determined.Inflammatory response, demyelination, and intrac-erebral bleeding would be severe and intolerableside effects of the immunization. Current data indi-cate that the meningoencephalitis may be due to a

T-cell response rather than the anti-A β antibodies.Immunization with the full-length A β42 peptide,

containing both B- and T-cell epitiopes, appearsnot to be optimual, because it brings about anextensive T-cell activation. The cerebral bleeding ispossibly due to cerebral amyloid angiopathy(CAA). The cerebral hemorrhages were reportedafter passive anti-A β immunotherapy in mice [65].Investigation into the pathogenesis of meningoen-cephalitis induced by vaccination with amyloid- βpeptide should now be possible using a recentlyconstructed appropriate animal model [77].

It cannot be excluded that the differences insafety results obtained in transgenic animals and inclinical trials depend, at least to some extent, on thedifferent adjuvants used in protocols. In the studiesin mice, the adjuvants CFA (complete Freund’sadjuvant) and IFA (incomplete Freund’s adjuvant)were used, whereas in clinical trials the immuno-gen was formulated in adjuvant QS21, a saponinederivative. Moreover, in clinical trial a detergent

(polysorbate-80) was added to aid the manufactur-ing and stability of the A β peptide [13].

15.3 Current Directions inExperimental and ClinicalResearch

The experimental evidence indicates that the clear-ance of A β from the brain is dependent on anti-A βantibody and not on T cell–mediated mechanisms.These mechanisms were probably responsible for

250 J. Karkos



side-effects observed in the first clinical trials. It isclear that alternative approaches must be developedthat bias the immune response toward a T h2-phenotype and/or replace the A β T-cell epitopewith a foreign T-cell epitope.

These goals may be attained through modifica-tions of the A β molecule, synthesis of newimmunogens, and by choice of suitable adjuvants.The use of humanized monoclonal anti-A β anti-bodies will entirely eliminate a cellular response toAβ, with comparable effectiveness to active immu-nization. The development of new delivery systemscan also contribute to the improvement of efficacyand safety aspects of immunization. Some of thecurrent approaches are discussed below.

15.3.1 Active Immunization

An immunization procedure was developed forthe production of effective anti-aggregating A βmonoclonal antibodies based on filamentousphages displaying only one epitope, the EFRHepitope, as a specific and nontoxic antigen.Effective autoimmune responses were obtainedafter phage administration as an antigen inguinea-pigs, in which the amino acids sequence inthe A β molecule is identical to that in humans.Because of the high antigenicity of the phage, noadjuvant was required to obtain high affinity anti-aggregating IgG antibodies [7].

The development of immunoconjugates seemsto be a very promising strategy. The immunoconju-gates are typically composed of a fragment of theAβ peptide derived from either the amino-terminalor central region linked to a carrier protein that pro-vides T-cell help. An epitope vaccine has beenengineered composed of the B-cell epitope from

the immunodominant region of A β42, Aβ1-15 in tan-dem with a universal synthetic T-cell epitope, panHLA DR-binding peptide (PADRE). Immunizationof BALB/c mice with the PADRE-A β1-15 epitopevaccine produced high titers of anti-A β antibodies[78].

Seabrook et al. [79] have designed two multi-antigen peptides (MAP) composed of either 8copies of A β1-7 or 16 copies of A β1-15 and investi-gated the immune response in B6D2F1 mice. TheMAP were formulated with the adjuvant LT(R192G). As the mice receiving A β1-15 MAP gen-erated very high anti-A β antibody titers of the

mainly IgG isotype, it was suggested that this MAPmay have potential as an AD vaccine.

Immunization with A β40 fibrils generated twoconformation-specific monoclonal antibodies inBALB/c mice [80]. The monoclonal antibodiesWO1 and WO2 bound to the amyloid fibril state of the A β40 peptide but not to its soluble, monomericstate. This new class of antibodies appears to rec-ognize a common conformational epitope with lit-tle apparent dependence on amino acid side-chainconformation. Reduction in brain levels of solubleAβ42 by 57% was detected after immunization witha soluble non-amyloidogenic, nontoxic A β homol-ogous peptide in Tg2576 mice. The cortical andhippocampal brain amyloid burden was reduced by89% and 81%, respectively [48].

Although compelling evidence has been pro-vided that the reduction of plaque burden afterimmunization is mediated through anti-A β antibod-ies, Frenkel et al. [81] reported that nasal vaccina-tion with a proteasome-based adjuvant (IVX-908)and glatiramer acetate, a synthetic copolymer usedin the treatment of multiple sclerosis, clears β-amy-loid in a mouse model of AD in an antibody-inde-pendent fashion. Vaccinated animals developedactivated microglia (CD11b + cells), and the extentof microglial activation correlated strongly with thedecrease in A β fibrils. They also found a strong cor-relation between CD11b + cells and IFN- γ secretingcells and increased numbers of T cells, which mayplay a role in promoting microglial activation.

15.3.2 Passive Immunization

Passive immunotherapy has advantages over activeimmunization from both efficacy and safety per-spectives. Particularly, passive immunotherapy

using a humanized monoclonal anti-A β antibodywill entirely eliminate a cellular response to A β.The use of polyclonal anti-A β antibodies can beconsidered as a promising alternative. Polyclonalanti-A β antibodies can be delivered by healthyindividuals because they have circulating autoanti-bodies against A β-peptide.

Bard et al. [52] determinated prerequisites formonoclonal antibodies to prevent neuropathologiclesions in transgenic mice. For this purpose,immune sera with reactivity against different A βepitopes and monoclonal antibodies with differentisotypes were examined for efficacy ex vivo and




in vivo. They found that only antibodies against theN-terminal regions of A β were able to invokeplaque clearance. Plaque binding correlated with aclearance response, whereas the ability of antibod-ies to capture soluble A β was not necessarily cor-related with efficacy. The isotype of the antibodyinfluenced the degree of plaque clearance. Highaffinity of the antibody for Fc receptors seemedmore important that high affinity for A β itself.

High-affinity anti-aggregating monoclonal anti-Aβ antibodies were obtained in human APP trans-genic mice after a short immunization time withphage-EFRH. A dose-response relationship wasobserved between antibody-titer and reduced amy-loid load. High immunogenicity of the phageenables intranasal administration without use of

adjuvant [40].Rangan et al. [82] have identified recombinant

antibody light-chain fragments with proteolyticactivity, capable of hydrolyzing A β in vitro.Although these fragments currently demonstratebroad substrate specificity, they may prove thera-peutically useful if the antibody could be engi-neered to specifically target pathogenic forms of Aβ, such as oligomers or protofibrils.

By screening a human single-chain antibody(scFv) library for A β immunoreactivity, Fukuchiet al. [83] have isolated a battery of scFvs thatspecifically react with amyloid plaques in thebrain. The efficacy of human scFv was tested in amouse model of AD. It was observed that relativeto control mice, injections of the scFv into the brainof transgenic mice reduced A β deposits andimproved spatial learning in Morris water maze.They concluded that human scFvs against A β maybe useful to treat AD patients without elicitingbrain inflammation because scFvs lack the Fc-por-

tion of the immunoglobulin molecule.Frenkel et al. [6] suggested a novel approach,

where intracellular expression of a site-directedsingle-chain antibody, which has been shown toinhibit fibrillogenesis and cytotoxicity in vitro,could target A β before it is released from the cell.

Reducing the ability of an amyloidogenic pro-tein to form partly unfolded species has been sug-gested as an effective method of preventing itsaggregation [84]. It was shown that a single-domain fragment of a camelid antibody raisedagainst wild-type human lysosyme inhibits the invitro aggregation of its amyloidogenic variant,

D67H. The binding of the antibody achieves itseffect by restoring the structural cooperativity char-acteristic of the wild-type protein. This appeared tooccur at least in part through the transmission of long-range conformational effects to the interfacebetween the two structural domains of the protein.

Ultrastructural investigation into structure of human classical plaques in different stages of development showed that in the early plaque, theleading pathology is fibrillar A β deposition bymicroglial cells. In the late plaques, microglialcells retract and activation of astrocytes predomi-nate [85]. In line with these findings, Wyss-Corayet al. [86] found that adult mouse astrocytesdegrade amyloid- β in vitro and in situ . Further-more, it was demonstrated [87] that a modest

increase in astroglial production of transforminggrowth factor β1 (TGF- β1) in aged transgenic miceexpressing the human APP (hAPP) results in athreefold reduction in the number of parenchymalamyloid plaques, a 50% reduction in the overall A βload in the hippocampus and neocortex, and adecrease in the number of dystrophic neurites. Inmice expressing hAPP and TGF- β1, the reductionof parenchymal plaques was associated with astrong activation of microglia and an increase ininflammatory mediators. Taken together, the stim-ulation of astrocytes and/or microglia could beconsidered an alternative approach for the treat-ment of AD. However, it was found [88] that over-activation of microglia induces apoptosis.Interestingly, in the experiment reported by Weineret al. [37], the lowering of A β burden was associ-ated with decreased local microglial and astrocyticactivation after nasal administration of A β40 toPDAPP mice. In serum, anti-A β antibodies of theIgG1 and Ig2b classes were detected, both of which

are characteristic of the T h2-type immune response.It is possible to generate anti-A β antibodies that

are capable of exerting their selective effect on A βfibrils. In the study by McLaurin et al. [43], theTgCRND8 mice were vaccinated with protofibril-lar/oligomeric assemblies of A β42 that reducedcerebral A β deposits and cognitive impairmentsand induced immunoglobulins of IgG2b isotypeagainst residues 4–10 of A β. The generated anti-Aβ antibodies were able to inhibit A β fibrilsassembly and toxicity without activating microglialor other cellular inflammatory responses. In thelight of the above-mentioned results, both stimula-

252 J. Karkos



tion and inhibition of either microglia or astrocytesmight be of therapeutic relevance in dependence,among others, of the stage in classical plaquedevelopment. Schmechel et al. [89] suggest thatmonoclonal antibody recognizing A β42 homod-imers, which are potentially the earliest form of synaptotoxic A β oligomers, might be useful for A βamyloid related therapeutic approaches by imped-ing its precipitation into existing plaques. A multi-antibody based approach, with one antibodytargeted against A β and one against tau, was sug-gested by Oddo et al. [21].

Specific polyclonal anti-A β-IgG in both the serumand the CNS from non-immunized humans wereidentified [27, 29]. The distribution of the differentIgG subclasses in the A β antibody sample were as

follows: IgG1, 63.8%; IgG2, 19.9%; IgG3, 9%; andIgG4, 7.3%. These antibodies were able to block fib-ril formation, disrupt formation of fibrillar structures,and prevent neurotoxicity of A β in vitro [90]. Inanother experiment [52], purified anti-A β antibodiescould disaggregate both preformed A β40 as well asactive truncated A β 25-35 and also block neurotoxicityinduced by both peptides. These results indicate thatthe investigated antibody fractions include antibodiesnot only against the N-terminal of A β but alsoagainst the middle portion of A β.

In a pilot study [91], IgG were administeredintravenously (IVIgG) in patients with AD. Fivepatients with AD were enrolled and receivedmonthly IVIgG (0.4 g intravenous IG per kg bodyweight) over a 6-month period. After IVIgG, totalAβ levels in the CSF decreased by 30.1% com-pared with baseline. Total A β increased in theserum by 233%. No effect on A β42 levels wasobserved. In addition, stabilization or a mildimprovement in cognitive function was observed in

the patients as detected using ADAS-cog.(improvement of 3.7 ± 2.9 points). It was postu-lated that the effects of IVIg in the AD patientswere due to altered cytokine production bymicroglial cells. However, the patient populationincluded in this study was too small to make defi-nite conclusions regarding the efficacy of IVIg inAD. From the safety point of view, it is importantthat polyclonal antibodies do not bind complement.Taken together, the available data indicate thatadministration of polyclonal human anti-A β anti-bodies isolated from plasma might be a potentialtherapeutic agent in AD.

15.3.3 Gentechnologic Approaches

It could be expected that efficacy and safety issuesassociated with immunotherapy for AD could beimproved using DNA vaccines or viral vectors [92,93]. Among the most important goals of the work

being done in the field are (i) the limitation of extension of amyloid accumulation through gener-ation of high titers of epitope-specific anti-A β anti-bodies with favorable isotype-profiles; (ii)reduction of side effects related to T h1-responses;(iii) induction of T h2-based immune response; and(iv) breaking of self-tolerance to A β. Some of thesegoals have already been achieved in animals. Forexample, Qu et al. [94] have demonstrated thatgene-gun–mediated genetic immunization with

Aβ42 gene can efficiently elicit humoral immuneresponses against mouse A β42 peptide in wild-typeBALB/c mice as well as against human A β42 intransgenic mice. It was shown that induction of thehumoral immune response did not induce a signif-icant cellular immune response. A study is under-way to detect whether this novel immunizationapproach leads to reduction of A β burden in thebrains of mice.

Dodart et al. [95] investigated whether genedelivery of the three common human apoE iso-

forms can directly alter the brain A β pathology inPADPP transgenic mice. They demonstrated thatintracerebral gene delivery of the lentivirus encod-ing apoE-constructs resulted in efficient and sus-tained expression of human apoE in thehippocampus as well as in a significant isoform-dependent effect of human apoE on hippocampalAβ burden and amyloid formation. This experi-mental data suggests that gene delivery of humanapoE2 may prevent and/or reduce brain A β burden

and the subsequent formation of neuritic plaques. Itis possible that the use of gene technology couldenable the construction of new transgenic animalsmodels suitable for further investigating the effi-cacy and safety of immunotherapy [96].

15.3.4 Role of Adjuvant

The choice of appropriate adjuvant canstrengthen the antibody response to A β42 and shiftthe type of the immune response generated (T

h1

vs. T h2). To investigate the role of adjuvant in thehumoral and cell-mediated immune response to




Aβ42, immunization with A β42 formulated in fourdifferent adjuvants, complete Freund’s adjuvant(CFA), incomplete Freund’s adjuvant (IFA), sapo-nine QS21, alum, and TitreMax Gold (TMG), wasperformed in BALB/c mice [25]. All adjuvantsinduced a strong anti-A β

42antibody response after

the first boost, and the antibody titers increasedconsiderably after the second and third boosts withfibrillar A β42. A significant difference in the mag-nitude of the antibody response to A β42 immu-nization with the different adjuvants was observed.The highest titers of antibody were generated inmice immunized and boosted with A β42 formu-lated in QS21 followed by CFA/IFA > alum >TMG.

To provide a relative measure of the contribu-

tion of T h2- and T h1-type humoral responses, theratios of IgG1 to IgG2a antibody generated inresponse to A β immunization were examined. Allmice immunized with A β42 formulated in alumhad IgG1:IgG2a ratios >1, indicating that thisadjuvant induced primarily T h2-type antibodyresponse against A β42. On the other hand, CFA,TMG, and QS21 shifted the humoral immuneresponses toward a T h1 phenotype. Promisingresults in terms of antibody generation and theirisotypes were obtained in B6D2F1 mice afterimmunization with A β formulated in adjuvantsmonophosphoryl lipid A (MPL)/trehalose dico-rynomycolate (TDM), cholera toxin B subunit(CTB), and LT (R192G) [97].

15.4 Other Suggested TreatmentApproaches Targeting A β

Amyloid binding ligands (ABL) has been sug-

gested as an alternative, non-immunological thera-peutic strategy to delay the onset or slow theprogression of AD [98]. The ABL represent deriv-atives of known amyloid-binding molecules suchas Congo red, chrysamine G (CG), and thioflavin S(TS). The generated derivatives of CG and TSspecifically recognize fibrillar A β in vitro, arrestthe formation of A β fibrils, and contrary to the par-ent substances, they cross the blood-brain barrier of transgenic mice after intravenous administration. Itwas demonstrated that CG derivative IMSB bindsto amyloid plaques composed of A β40 with muchhigher affinity than A β42, whereas TS derivative

TDZM shows the opposite affinity. Furthermore,IMSB but not TDZM bound selectively to neu-rofibrillary tangles.

As the microglia activated by A β exert theirtoxic effects through NMDA receptors in vitro, theblocking of these receptors may be an effectivetherapeutic approach [99]. It is possible that small,bifunctional molecules that reveal antifibrillogenicproperties may be of relevance in vivo [100]. Zinc-copper chelation resulting in the solubilization of Aβ offers promise as a new therapeutic approachfor AD [101, 102]. Curcumin, the unconventionalNSAID/antioxidant, has multiple anti-amyloidactions. Curcumin, targeting directly A β, may actas a “peripheral sink” [103].

15.5 Conclusions

Although transgenic animals are not the mostfavorable models of AD in terms of morphologicand immunologic aspects, compelling evidenceexists that immunotherapy can prevent or reduceneuropathology and improve cognitive perform-ance. The preventative effects of immunization aremediated by anti-A β antibodies, with titer, isotype,and epitope specificity playing crucial roles in theireffects. Experimentally, the anti-A β antibodiesreduced or prevented plaque formation, actedagainst aggregation and neurotoxicity, favored dis-aggregation, and promoted recovery of neuronaldamage. Compelling experimental evidence alsoindicates that A β immunization may be useful forclearing aggregates of tau protein, another hall-mark lesion of AD neuropathology, on conditionthat the treatment occurs early in the disease pro-gression. Clinically, the primary concern is the

safety of immunotherapy, especially the cause of side effects, including subacute meningoencephali-tis, microhemorrhages, and demyelination. Withregard to efficacy, slowing down of cognitivedeficits after suspension of vaccine administrationin a cohort study was observed. Modifications of Aβ-antigen, synthesis of new immunogens, gener-ation of epitope-specific monoclonal antibodies,development of new adjuvants and delivery sys-tems may contribute to future favorable efficacyand safety profiles of immunotherapy. In thisrespect, gentechnology seems to be a particularlypromising approach.

254 J. Karkos



References

1. Dobson CM. Protein folding and misfolding. Nature2003, 426, 884-890.

2. Holtzman DM. A β conformational change is centralto Alzheimer’s disease. Neurobiol Aging 2002, 23,

1085-1088.3. Selkoe DJ. Folding proteins in fatal ways. Nature2003, 426, 900-904.

4. Soto C. Unfolding the role of protein misfolding inneurodegenerative diseases. Nat Rev Neurosci 2003,4, 49-60.

5. Frenkel D, Balass M, Solomon B. N-terminal EFRHsequence of Alzheimer’s β-amyloid peptide repre-sents the epitope of its anti-aggregating antibodies. JNeuroimmunol 1998, 88, 85-90.

6. Frenkel D, Solomon B, Benhar I. Modulation of Alzheimer’s beta-amyloid neurotoxicity by site-directed single-chain antibody. J Neuroimmunol2000, 106, 21-31.

7. Solomon B. Generation of anti- β-amyloid antibodiesvia phage display technology towards Alzheimer’sdisease vaccination. Vaccine 2005, 23, 2327-2330.

8. Solomon B. Immunotherapeutic strategies for pre-vention and treatment of Alzheimer’s disease. DNACell Biol 2001, 20, 697-703.

9. Schenk D, Barbour W, Dunn W, et al. Immunizationwith amyloid- β attenuates Alzheimer’s-disease-likepathology in the PDAPP mouse. Nature 1999, 400,

173-177.10. Schenk D. Amyloid- β immunotherapy for Alzhei-

mer’s disease: The end of the beginning. Nat RevNeurosci 2002, 3, 824-828.

11. Janus C, Pearson J, McLaurin J, et al. A β peptideimmunization reduces behavioural impairment andplaques in a model of Alzheimer’s disease. Nature2000, 408, 979-982.

12. Morgan D, Diamond DM, Gottschall PE, et al. A β -peptide vaccination prevents memory loss in an ani-mal model of Alzheimer’s disease. Nature 2000, 408,982-985.

13. Schenk D, Hagen M, Seubert P. Current progress inbeta-amyloid immunotherapy. Curr Opin Immunol2004, 16, 599-606.

14. Cirrito JR, Holtzman DM. Amyloid- β and Alzheimer’sdisease therapeutics: the devil may be in the details.J Clin Invest 2003, 112, 321-323.

15. Gelinas DS, DaSilva K, Fenili D, et al. Immuno-therapy for Alzheimer’s disease. Proc Natl Acad SciU S A 2004, 101 (suppl 2), 14657-14662.

16. Karkos J. Immuntherapie der Alzheimer-Krankheit.Experimentelle Untersuchungsergebnisse und

Behandlungsperspektiven. Fortschr Neurol Psychiat2004, 72, 204-219.

17. Selkoe DJ, Schenk D. Alzheimer’s disease: molecu-lar understanding predicts amyloid-based therapeu-tics. Annu Rev Pharmacol Toxicol 2002, 43,545-584.

18. Gandy S. Molecular basis for anti-amyloid therapy inthe prevention and treatment of Alzheimer’s disease.

Neurobiol Aging 2002, 23, 1009-1016.19. Selkoe DJ. Alzheimer’s disease results from the cere-

bral accumulation and cytotoxicity of amyloid β-pro-tein. J Alzheimer’s Dis 2001, 3, 75-81.

20. Hutton M, McGowan E. Clearing tau pathology withAβ immunotherapy-reversible and irreversible stagesrevealed. Neuron 2004, 43, 293-294.

21. Oddo S, Billings L, Kesslak JP, et al. A βimmunotherapy leads to clearance of early, but notlate, hyperphosphorylated tau aggregates via protea-some. Neuron 2004, 43, 321-332.

22. Barrow CJ, Zagorski MG. Solution structures ofβ-peptide and its constituent fragments: relation toamyloid deposition. Science 1991, 253, 179-182.

23. Hollosi M, Otvos L, Kaijtar J, et al. Is amyloid dep-osition in Alzheimer’s disease preceded by an envi-ronment induced double conformational transition?Peptide Res 1989, 2, 109-113.

24. Maggio JE, Mantyh PW. Brain amyloid – a physico-chemical perspective. Brain Pathol 1996, 6, 147-162.

25. Cribbs DH, Ghochikyan A, Vasilevko V, et al.Adjuvant-dependent modulation of T h1 and T h2responses to immunization with β-amyloid. Int

Immunol 2003, 15, 505-514.26. Hyman BT, Smith C, Buldyrev I, et al.

Autoantibodies to amyloid- β and Alzheimer’s dis-ease. Ann Neurol 2001, 49, 808-810.

27. Du Y, Dodel R, Hampel H, et al. Eastwood B.Reduced levels of amyloid beta-peptide antibody inAlzheimer’s disease. Neurology 2001, 57, 801-805.

28. Weksler ME, Relkin N, Turkenich R, et al. Patientswith Alzheimer’s disease have lower levels of serumanti-amyloid peptide antibodies than healthy elderlyindividuals. Exp Gerontol 2002, 37, 943-948.

29. Dodel R, Hampel H, Depboylu C, et al. Human anti-bodies against amyloid β peptide: a potential treat-ment for Alzheimer’s disease. Ann Neurol 2002, 52,253-256.

30. McGeer EG, McGeer PL. Inflammatory process inAlzheimer’s disease. Progr NeuropsychopharmacolBiol Psychiatry 2003, 27, 741-749.

31. Monsonego A, Weiner HL. Immunotherapeuticapproaches to Alzheimer’s disease. Science 2003,302, 834-838.

32. Monsonego A, Zota V, Karni A, et al. Increased Tcell reactivity to amyloid β protein in older humansand patients with Alzheimer’s disease. J Clin Invest2003, 112, 415-422.




33. Town T, Tan J, Sansone N, et al. Characterization of murine immunoglobulin G antibodies against humanamyloid- β1-42. Neurosci Lett 2001, 307, 101-104.

34. Stalder M, Deller T, Staufenbiel M, Jucker M. 3D-Reconstruction of microglia and amyloid in APP23transgenic mice: no evidence of intracellular amy-

loid. Neurobiol Aging 2001, 22, 427-434.35. Webster SD, Yabg AJ, Margol L, et al. Complement

component C1q modulates the phagocytosis of Abetaby microglia. Exp Neurol 2000, 161, 127-138.

36. Bard F, Cannon C, Barbour R, et al. Peripherallyadministered antibodies against amyloid β-peptideenter the central nervous system and reduce pathol-ogy in a mouse model of Alzheimer’s disease. NatMed 2000, 6, 916-919.

37. Weiner HL, Lemere CA, Maron R, et al. Nasaladministration of amyloid- β peptide decreases cere-bral amyloid burden in a mouse model of Alzheimer’sdisease. Ann Neurol 2000, 48, 567-579.

38. Dodart J-C, Bales KR, Gannon KS, et al.Immunization reverses memory deficits withoutreducing brain Abeta burden in Alzheimer’s diseasemodel. Nat Neurosci 2002, 5, 452-457.

39. Kotilinek LA, Bacskai B, Westerman M, et al.Reversible memory loss in a mouse transgenic modelof Alzheimer’s disease. J Neurosci 2002, 22, 6331-6335.

40. Frenkel D, Dewachter F, Van Leuven F, Solomon B.Reduction of beta-amyloid plaques in brain of trans-

genic mouse model of Alzheimer’s disease byEFRH-phage immunization. Vaccine 2003, 21,1060-1065.

41. Lemere CA, Maron R, Spooner ET, et al. Nasal A βtreatment induces anti-A β antibody production anddecreases cerebral amyloid burden in PD-APP mice.Ann N Y Acad Sci 2000, 920, 328-333.

42. Lemere CA, Spooner E, Leverone J, Mori C,Clements J. Intranasal immunotherapy for the treat-ment of Alzheimer’s disease: Escherichia coli LT andLT (R192G) as mucosal adjuvants. Neurobiol Aging

2002, 23, 991-1000.43. McLaurin J, Cecal R, Kierstead ME, et al.Therapeutically effective antibodies against amyloid-β peptide target amyloid- β residues 4-10 and inhibitcytotoxicity and fibrillogenesis. Nat Med 2002, 8,1263-1269.

44. Spooner E, Desai R, Mori C, et al. The generation andcharacterization of potentially therapeutic A β anti-bodies in mice: differences according to strain andimmunization protocol. Vaccine 2000, 21, 290-297.

45. Das P, Murphy M, Younkin L, et al. Reduced effec-tiveness of A β1-42 immunization in APP transgenicmice with significant amyloid deposition. NeurobiolAging 2001, 22, 721-727.

46. DeMattos RB, Bales KR, Cummins DJ, et al.Peripheral anti-A β antibody alters CNS and plasmaclearance and decreases brain A β burden in a mousemodel of Alzheimer’s disease. Proc Natl Acad SciU S A 2001, 98, 8850-8855.

47. Frenkel D, Katz O, Solomon B. Immunization

against Alzheimer’s β-amyloid plaques via EFRHphage administration. Proc Natl Acad Sci U S A2000, 97, 11455-11459.

48. Sigurdsson EM, Scholtzova H, Mehta PD, et al.Immunization with non-toxic/nonfibrillar amyloid- βhomologous peptide reduces Alzheimer’s disease-associated pathology in transgenic mice. Am J Pathol2001, 159, 439-447.

49. Brayden D, Templeton L, McClean S, et al.Encapsulation in biodegradable microparticlesenhances serum antibody response to parenterallydelivered β-amyloid in mice. Vaccine 2001, 19,4185-4193.

50. Zhang J, Wu S, Qin C, et al. A novel recombinantadeno-associated virus vaccine reduces behaviouralimpairment and β-amyloid plaques in a mousemodel of Alzheimer’s disease. Neurobiol Dis 2003,14, 365-379.

51. Lemere CA, Beierschmitt A, Iglesias M, et al.Alzheimer’s disease A β vaccine reduces central nerv-ous system A β levels in a non-human primate, theCaribbean vervet. Am J Pathol 2004, 165, 283-297.

52. Bard F, Barbour R, Cannon C, et al. Epitope and iso-

type specificities of antibodies to beta-amyloid pep-tide for protection against Alzheimer’s disease-likeneuropathology. Proc Natl Acad Sci U S A 2003,100, 2023-2028.

53. Wilcock DM, DiCarlo G, Henderson D, et al.Intracranially administered anti-A β antibodiesreduce β-amyloid deposition by mechanisms bothindependent of and associated with microglial activa-tion. J Neurosci 2003, 213, 3745-3751.

54. Wilcock DM, Munireddy SK, Rosenthal A, et al.Microglial activation facilitates A β plaque removal

following intracranial anti-A β administration.Neurobiol Dis 2004, 15, 11-20.55. Wilcock DM, Rojiani A, Rosenthal A, et al. Passive

amyloid immunotherapy clears amyloid and tran-siently activates microglia in a transgenic mousemodel of amyloid deposition. J Neurosci 2004, 24,6144-6151.

56. Bussière T, Bard F, Barbour R, et al. Morphologicalcharacterization of thioflavin-S-positive amyloidplaques in transgenic Alzheimer mice and effect of passive A β immunotherapy on their clearance. Am JPathol 2004, 165, 987-995.

57. Das P, Howard V, Loosbrock N, et al. Amyloid- βimmunization effectively reduces amyloid deposition

256 J. Karkos



in FcR γ − / − knock-out mice. J Neurosci 2003, 23,8532-8538.

58. Bacskai BJ, Kajdasz S, McLellan ME, et al. Non-Fc-mediated mechanisms are involved in clearance of amyloid- β in vivo by immunotherapy. J Neurosci2002, 22, 7873-7878.

59. Koenigsknecht J, Landreth G. Microglial phagocyto-sis of fibrillar β-amyloid through a β

1 integrin-depend-ent mechanism. J Neurosci 2004, 24, 9838-9846.

60. Lemere CA, Spooner E, LaFrancois J, et al. Evidencefor peripheral clearance of cerebral A β protein fol-lowing chronic, active A β immunization in PSAPPmice. Neurobiol Dis 2003, 14, 10-18.

61. Huang F, Buttini M, Wyss-Coray T, et al. Eliminationof the class A scavenger receptor does not affectamyloid plaque formation or neurodegeneration intransgenic mice expressing human amyloid proteinprecursors. Am J Pathol 1999, 155, 1741-1747.

62. Husemann J, Loike JD, Anankov R, et al. Scavengerreceptors in neurobiology and neuropathology: theirrole on microglia and other cells of the nervous sys-tem. Glia 2002, 40, 195-205.

63. Münch G, Thome J, Foley P, et al. Advanced glyca-tion end products in ageing and Alzheimer’s disease.Brain Res Rev 1997, 23, 134-143.

64. Racke MM, Boone LI, Hepburn DL, et al.Exacerbation of cerebral amyloid angiopathy-associ-ated microhemorrhage in amyloid precursor proteintransgenic mice by immunotherapy is dependent on

antibody recognition of deposited forms of amyloidβ. J Neurosci 2005, 25, 629-636.

65. Pfeifer M, Boncristiano S, Bondolfi L, et al. Cerebralhemorrhage after passive anti-A β immunotherapy.Science 2002, 298, 1379.

66. Scolding NJ, Joseph F, Kirby PA, et al. A β-relatedangiitis: primary angiitis of the central nervous sys-tem associated with cerebral amyloid angiopathy.Brain 2005, 128, 500-515.

67. Banks W.A., Terell B., Farr S.A., et al. Passage of amyloid β protein antibody across the blood-brain

barrier in a mouse model of Alzheimer’s disease.Peptides 2002, 23, 2223-2226.68. Hock C, Konietzko U, Papassotiropoulos A, et al.

Generation of antibodies specific for β-amyloid byvaccination of patients with Alzheimer’s disease. NatMed 2002, 8, 1270-1275.

69. Dodart J-C, Mathis C, Saura J, et al.Neuroanatomical abnormalities in behaviourallycharacterized APP V717F transgenic mice. NeurobiolDis 2000, 7, 71-85.

70. Gandy S, Walker L. Toward modelling hemorrhagicand encephalitic complications of Alzheimer amy-loid- β vaccination in nonhuman primates. Curr OpinImmunol 2004, 16, 607-615.

71. Bayer AJ, Bullock R, Jones RW, et al. Evaluation of the safety and immunogenicity of syntheticA[beta]42 (AN1792) in patients with AD. Neurology2005, 64, 94-101.

72. Orgogozo J-M, Gilman S, Dartigues J-F, et al.Subacute meningoencephalitis in a subset of patients

with AD after A [beta] 42 immunization. Neurology2003, 61, 46-54.

73. Hock C, Konietzko U, Streffer JR, et al. Antibodiesagainst β-amyloid slow cognitive decline inAlzheimer’s disease. Neuron 2003, 38, 547-554.

74. Ferrer I, Boada Rovira M, et al. Neuropathology andpathogenesis of encephalitis following amyloid- βimmunization in Alzheimer’s disease. Brain Pathol2004, 14, 11-20.

75. Masliah E, Hansen L, Adame A, et al. A [beta] vac-cination effects on plaque pathology in the absenceof encephalitis in Alzheimer’s disease. Neurology2005, 64, 129-131.

76. Nicoll JAR, Wilkinson D, Holmes C, et al.Neuropathology of human Alzheimer’s disease afterimmunization with amyloid- β peptide: a case report.Nat Med 2003, 9, 448-452.

77. Furlan R, Brambilla E, Sanvito F, et al. Vaccinationwith amyloid- β peptide induces autoimmuneencephalomyelitis in C57/BL6 mice. Brain 2003,126, 285-291.

78. Agadjanyan MG, Ghochikyan A, Petrushina I, et al.Prototype Alzheimer’s disease vaccine using the

immunodominant B cell epitope from β-amyloid andpromiscuous T cell epitope Pan HLA DR-bindingpeptide. J Immunol 2005, 174, 1580-1586.

79. Seabroook TJ, Bloom JK, Spooner ET, Lemere CA.The use of A β1-15 multi-antigen peptide as a poten-tial vaccine for Alzheimer’s disease. Program No.830.1. 2004 Abstract Viewer/Itinerary Planner .Washington, DC: Society for Neuroscience.

80. O’Nuallain B, Wetzel R. Conformational Abs recog-nizing a generic amyloid fibril epitope. Proc NatlAcad Sci U S A 2002, 99, 1485-1490.

81. Frenkel D, Maron R, Burt DS, Weiner HL. Nasal vac-cination with a proteasome-based adjuvant and glati-ramer acetate clears β-amyloid in a mouse model of Alzheimer’s disease in an antibody-independent fash-ion. Program No. 674.4. 2004 Abstract Viewer/itineraryPlanner . Washington, DC: Society for Neuroscience.

82. Rangan SK, Ruitian L, Brune D, et al. Degradationof β-amyloid by proteolytic antibody light chains.Biochemistry 2003, 42, 14328-14334.

83. Fukuchi K, Accaviti–Loper M., Kim H.D, et al.Human single chain antibodies for treatment of Alzheimer’s disease. Program No. 675.13. 2004

Abstract Viewer/Itinerary Planner . Washington, DC:Society for Neuroscience.




84. Dumoulin M., Last AM., Desmyther A., et al. Acamelid antibody fragment inhibits the formation of amyloid fibrils by human lysozyme. Nature 2003,424, 783-788.

85. Wegiel J, Wang K-C, Tarnawski M, Lach B.Microglial cells are the driving force in fibrillar

plaque formation, whereas astrocytes are a leadingfactor in plaque degradation. Acta Neuropathol2000, 100, 356-364.

86. Wyss-Coray T, Loike JD, Brionne TC, et al. Adultmouse astrocytes degrade amyloid- β in vitro andin situ. Nat Med 2003, 9, 453-458.

87. Wyss-Coray T, Lin C, Yan F, et al. TGF- β1 pro-motes microglial amyloid- β clearance and reducesplaque burden in transgenic mice. Nat Med 2001, 7,612-618.

88. Liu B, Wang K, Gao HM, et al. Molecular conse-quences of activated microglia in the brain: overac-tivation induces apoptosis. J Neurochem 2001, 77,182-189.

89. Schmechel A, Zentgraf H, Scheuermann S, et al.Alzheimer β-amyloid homodimers facilitate A β fib-rillization and the generation of conformationalantibodies. J Biol Chem 2003, 278, 35317-35324.

90. Du Y, Wei X, Dodel R, et al. Human anti- β-amyloidantibodies block β-amyloid fibril formation andprevent β-amyloid-induced neurotoxicity. Brain2003, 126, 1935-1939.

91. Dodel RC, Du Y, Depboylu C, et al. Intravenous

immunoglobulins containing antibodies against β-amyloid for the treatment of Alzheimer’s disease. JNeurol Neurosurg Psychiatry 2004, 75, 1472-1474.

92. Ghochikyan A, Vasilevko V, Petrushina I, et al.Generation and characterization of the humoralimmune response to DNA immunization with achimeric beta-amyloid-interleukin-4 minigene. EurJ Immunol 2003, 33, 3232-3241.

93. Tsuji S. DNA vaccination may open up a newavenue for treatment of Alzheimer’s disease. ArchNeurol 2004, 61, 1832-1832.

94. Qu B, Rosenberg RN, Li L, et al. Gene vaccinationto bias the immune response to amyloid- β peptideas therapy for Alzheimer’s disease. Arch Neurol2004, 61, 1859-1864.

95. Dodart J-C, Marr RA, Koistinaho M, et al. Genedelivery of human apolipoprotein E alters brain A β

burden in a mouse model of Alzheimer’s disease.Proc Natl Acad Sci U S A 2005, 102, 1211-1216.

96. Chan AW, Chong KY, Martinovich C, et al.Transgenic monkeys produced by retroviral genetransfer into mature oocytes. Science 2001, 291,309-312.

97. Maier M, Seabrook TJ, Bloom JK, Lemere CA.Immune responses to alternative adjuvants and A βimmunogens for A β immunotherapy. Program No.830.2. 2004 Abstract Viewer/Itinerary Planner .Washington, DC: Society for Neuroscience.

98. Lee V M-Y. Amyloid binding ligands asAlzheimer’s disease therapies. Neurobiol Aging2002, 23, 1039-1042.

99. Takeuchi H, Mizuno T, Zhang G., et al. Neuriticbeading induced by activated microglia is an earlyfeature on neuronal dysfunction toward neuronaldeath by inhibition of mitochondrial respiration andaxonal transport. J Biol Chem 2005, 280, 10444-10454.

100. Gestwicki JE, Crabtree GR, Graef IA. Harnessingchaperones to generate small-molecule inhibitors of amyloid β aggregation. Science 2004, 306, 865-

869.101. Bush AI. The metallobiology of Alzheimer’s dis-

ease. Trends Neurosci 2003, 26, 207-214.102. Rosenberg RN. Metal chelation therapy for

Alzheimer’s Disease. Arch Neurol 2003, 60, 1678-1679.

103. Cole GM, Morihara T, Lim GP, et al. NSAID andantioxidant prevention of Alzheimer’s disease.Lessons from in vitro and animals models. Ann N YAcad Sci 2004, 1035, 68-84.

258 J. Karkos



16Mouse Models of Alzheimer’s DiseaseDwight C. German

259

16.1 Introduction

Transgenic mouse models have been created thatmimic many of the neuropathologic and behav-ioral phenotypes of Alzheimer’s disease (AD).Using mutations found in familial AD, the mousemodels exhibit some of the cardinal features of thehuman disease. Wong et al. [1] and Higgins andJacobsen [2] have written reviews of this topic.The current review extends a previous one [3] andwill describe the similarities in the neuropathology

of AD and the mouse models of the disease,specifically regarding neurodegeneration, and alsodescribe treatments being developed using themouse models.

16.2 Neuropathology of Alzheimer’s Disease

AD is characterized by extensive cortical and hip-pocampal neuropathology [4], including extracel-lular neuritic plaques composed of β-amyloid (A β)protein. There are also neurofibrillary tangles(NFTs), which accumulate within neurons in corti-cal and subcortical regions. In addition, severalsubcortical nuclei degenerate in AD, and many of the affected nuclei have been shown to project tothe cerebral cortex. The first subcortical nucleusfound to degenerate in AD was the nucleus basalisof Meynert [5–8], which contains cholinergic neu-rons that project to cortical and hippocampal

regions [9]. Further study indicated that there isalso degeneration of other cortical-projecting sub-

cortical nuclei: for example, the serotonergic dorsalraphe nucleus [10], the dopaminergic ventraltegmental area [11], and the noradrenergic locuscoeruleus [12–14].

AD is also characterized by inflammation;microglia are located near neuritic plaques andundergo a phenotypic activation [15]. Microglialactivation results in the expression of a wide rangeof proinflammatory molecules that may activelydamage/destroy neurons. Astrocytes are also acti-vated in AD.

Neurogenesis is abnormal in AD. Adult neuro-genesis occurs in brain structures that have a highdegree of neuronal plasticity, such as the hip-pocampus and olfactory bulb [16–20]. In the adultrat hippocampus, it is estimated that more than9000 new neurons are born each day [21].Although the number of newly born neurons isthought to be much lower in human and non-human primates [22–24], the presence of adult neu-rogenesis in a wide range of species suggests a rolefor new neurons in shaping the form and functionof the adult brain [25]. Adult neurogenesis is regu-lated by myriad environmental and physiologicalstimuli [26, 27]. In vivo, chronic stress, aging,inflammation, and repeated exposure to drugs of abuse decrease adult hippocampal neurogenesis[28–31].

Neurogenesis takes place in the hippocampus of the adult primate brain [17, 22, 24]. The first reportof neurogenesis in AD postmortem brain indicatesthat it is abnormal [32] and is also abnormal in

other neurodegenerative diseases like Parkinsondisease [33] and Huntington disease [34].



16.3 Mouse Models of Alzheimer’sDisease

Several transgenic mouse models of AD have beendeveloped. Although the various models exhibit

some of the neuropathologic features of the humandisease, so far none exhibits all of the features.Table 16.1 summarizes the gene mutations used tocreate nine AD mouse models and how the neu-ropathology in the mouse models compares withthat in AD. The early mouse models of AD con-tained mutant genes such as APP 717 (PDAPPmouse, [35]) and APP 695 (Tg2576 mouse, [36]).Additional bigenic models have been developedthat contain mutant APP and PS1 [37, 38], mutantAPP and tau [39], and a triple transgenic mousethat carries mutant amyloid precursor protein(APP), presenilin-1 (PS1), and tau [40]. Mice lack-ing PS1 and PS2 function also exhibit some ADneuropathology [41].

16.3.1 Amyloid- β Plaques

All APP and PS1 mouse models exhibit diffuseand/or neuritic A β-plaques in the cortex and hip-pocampus (Table 16.1), as illustrated in Figure 16.1.Two APP mouse models have been shown to exhibitan age-related development of neuritic plaques inthe cerebral cortex and hippocampus [35, 36].

One of the earliest AD mouse models was devel-oped in 1995 by Games et al. [35]: the PDAPPmouse. This mouse was generated using theplatelet-derived growth factor- β promoter driving ahuman APP minigene encoding the APP 717V-Fmutation associated with familial AD [42].Between 6 and 9 months of age, hemizygous

PDAPP mice exhibit thioflavin-S-positive A βdeposits and neuritic plaques. The A β-containingplaques are directly associated with reactive gliosisand dystrophic neurites, suggesting that theplaques may induce neurodegenerative changes.Some of the A β plaque pathology in the dentategyrus appears to originate from nerve terminalswhose axons traverse the perforant pathway, aslesions of this pathway in mouse models of ADresult in a reduction in hippocampal plaque pathol-ogy [43, 44].

Protofibrils are precursors to the formation of fibrilar neuritic plaques, and evidence suggests

that they play a role in the neurodegenerativeprocess. Protofibrils are short assemblies, 5–200nm in length, that assemble into A β plaques. Theprotofibrils have been shown to be neurotoxic [45,46]. The A β oligomers, but not monomers, inhibithippocampal long-term potentiation in the rat [47,48]. The homozygous PDAPP mouse containsvery high levels of soluble A β in both CSF andplasma [49]. That there are region-specificamounts of the oligomers in APP mouse models issuggested by the regional differences in splicevariants of β-secretase enzyme, which mayexplain why A β-extracellular plaques are formedonly in certain brain regions in AD and in ADmouse models [50].

16.3.2 Neurofibrillary Tangles

In some of the AD mouse models that express APPand/or PS mutations, there is an age-related hyper-phosphorylation of tau protein, which comes afterthe formation of A β-plaques [51–55]. However,none of these models exhibit NFTs as defined bythe presence of paired helical filaments (PHF)(Table 16.1). Kurt et al. [54] found evidence of PHF-like structures in the 24-month-old APP/PS1mouse but not in younger animals, however,whether they represent PHF or Hirano bodies is notclear. In hemizygous PDAPP animals up to 20months of age, no PHFs were observed [35, 51].Even in transgenic mice that express mutant APP,PS1, and tau [40], and in those expressing APP andtau [39], the NFTs that occur within neurons in theneocortex and hippocampus are defined solely byimmunostaining with phospho-specific tau anti-bodies and not by the presence of PHF. In a studyusing conditional knock-out of PS1 in PS2 KO

mice (PS cDKO mice), there is hyperphosphoryla-tion of tau in the cortex of 9-month-old mice andmarked cortical shrinkage [41]. These studies indi-cate that mouse models containing mutant APP,PS, and/or tau accumulate abnormally phosphory-lated tau in an age-related manner, but whetherthere is progression to PHF formation in older ani-mals must await further study.

16.3.3 Glial Activation

In APP transgenic mouse models of AD thatexhibit neuritic plaques in the cortex and hip-

260 D.C. German



16. Mouse Models of Alzheimer’s Disease 261

T A B L E

1 6 . 1 . N e u r o p a t h o l o g y

i n m o u s e m o d e l s o f

A l z h e i m e r ’ s

d i s e a s e .

N a m e

( A l t e r n a

t e n a m e )

N S E A P P

P D A P P

T g 2 5 7 6

T A P P

( A P P l o n )

( A P P s w e

)

A P P 2 3

T g C

R N D 8

( A P P / t a u

)

P S A P P

P S 1

A β - A r c

T r a n s g e n e o r m u t a t i o n

A P P

7 5 1 2 4

A P P

V 7 1 7 F

3 5

A P P

6 9 5 3 6

A P P

7 5 1 5 3

A P P

6 9 5 + V 7 1 7 F

1 3 0

A P P

6 9 5

A P P

6 9 5 × P S 1 7 0

P S 1 1 3 3 o r

A β

E 2 2 G +

× J N P L 3 3 9

c P S 1 9 4

A P P

6 9 5 / V 7 1 7 F

1 3 4

A m y l o

i d - β p l a q u e s

Y 1 2 4

Y 3 5

Y 3 6

Y 5 3

Y 1 3 0

Y 3 9

Y 7 0

Y 1 3 3

Y 1 3 4

N e u r o f

i b r i l l a r y t a n g l e s

N 5 1

Y 3 9

N 5 4

( P a i r e d

h e l i c a l f i l a m e n t s )

G l i a l a c t i v a t i o n

Y 1 2 5 , 1 2 6

Y 3 5

Y 1 2 8

Y 5 3

Y 1 3 0

Y 3 9

Y 7 0

Y 6 7

H i p p o c a m p a l a n d / o r

N 3 5 , 5 8 , 1 2 7

Y 5 9 , 1 2 9

N 1 3 1 , 1 3 2

N 9 4

c o r t i c a l c e l l

l o s s

C h o l i n e r g i c c e l l

l o s s

N 6 9

N 1 2 8

N 6 8

N 6 3

N o r a d r e n e r g i c c e l l

l o s s

N 8 8

A b n o r m

a l a d u l t

Y 9 5

Y 9 2

Y 9 4

h i p p o c a m p a l n e u r o g e n e s i s

A b b r e v i a t i o n s : Y

, y e s ;

N ,

n o ; b l a n k , n o t

d e t e r m

i n e d .

R e f e r e n c e n u m

b e r s a p p e a r a s s u p e r s c r i p t s .

D e t a i l s o n m u t a t i o n s a r e

f o u n d i n o r i g i n a l p u b l i c a t i o n s

( s e e r e f e r e n c e s ) .



pocampus, there is an activation of microglia inregions containing neuritic plaques (Fig. 16.2) [35,56, 57]. Also, there is activation of astrocytes in theregion of A β-containing plaques. Even in the mod-els that lack mutant APP, astrocytes are still acti-vated [41]. These data suggest that glial activationand inflammation are not solely related to the pres-ence of neuritic plaques.

16.3.4 Hippocampal and Cortical CellLoss

Modest neuron loss in hippocampus and cortex hasbeen reported in some AD mouse models.Hemizygous 18-month-old PDAPP mice have beenexamined for cortical cell loss, but there was noneeven in regions that contained a high density of plaques [58]. However, Calhoun et al. [59] reporteda moderate loss of cortical neurons in old APP23mice. The cell loss was correlated with amyloidplaque density in this study. In the PS cDKOmouse, there is an age-related cortical atrophy andthinning of the cortical mantle, although nodetailed quantitative cell counts have yet beenreported [41]. In addition, a hallmark of AD, amarked shrinkage of the hippocampus, has beenobserved in the PDAPP mouse [60, 61]. There isalso a loss of CA1 neurons in the hippocampus inAPP +PS1 mutant animals [38].

Neurodegeneration becomes prominent in APPmouse models with impaired PS function. Using amouse model that expresses mutations in both APP(KM670/671NL and V717I) and PS1 (M146L),significant neurodegeneration has been reported inthe hippocampal CA1 region. The neurodegenera-tion appears to be age-related [38], and the neuronsthat are destined to degenerate accumulate A β pro-tein within the somata [62]. These data suggest thatneurodegeneration can occur from intracellularaccumulation of A β protein. In mice with mutanttau and APP, there are NFTs in entorhinal cortexand hippocampus CA1 that increase in number

262 D.C. German

FIGURE 16.1. A β-containing plaques accumulate withage in the PDAPP mouse brain. Brain sections werestained with an antibody against human A β. At 2 yearsof age, there are many mature and diffuse plaques in thecerebral cortex and hippocampus (A). A lower numberof compacted plaques are also found in subcorticalregions (B), such as the caudate-putamen, and in whitematter regions. Compared with the 2-year-old mouse, thenumber of compacted plaques is less in the 1-year-oldPDAPP mouse cortex and hippocampus (C), and thereare no plaques in subcortical regions like the thalamus(D). There are very few A β-containing plaques in the 4-month-old PDAPP brain (E) and none in the 2-year-oldnon-transgenic control brain (not illustrated).Abbreviations: CPu, caudate-putamen; df, dorsal fornix;DG, dentate gyrus; fmj; forceps major corpus callosum;RS, retrosplenial cortex; Th, thalamus. Marker, 150 µmin (A), (C), (E), and 300 µm in (B) and (D). Reproducedfrom German et al. [69].

FIGURE 16.2. Microglial cells surround neuritic plaquesin the PDAPP mouse cerebral cortex. Notice the numer-ous microglial cells (arrow points to one of several) sur-rounding the neuritic plaque (P). This section is stainedwith an antibody against ChAT (black fibers), and thesection is counterstained with cresyl violet. Marker,6 µm.



with age, especially in female transgenic animals[39]. Detailed cell counting was not performed inthis study, however. It will be interesting to makequantitative measurements of neurodegeneration incortical, hippocampal, and subcortical regions inanimal models that exhibit NFTs to determinewhether the NFTs play a role in the degeneration of these neurons.

16.3.5 Cholinergic Cell Loss

Cholinergic nerve terminal abnormalities are com-mon in the hippocampus and cortex of APP mousemodels [63–69]. Cholinergic degenerative changesoccur specifically in regions that eventually exhibitneuritic plaque deposition (Fig. 16.3). In 2-year-old

homozygous PDAPP mice, for example, there is avery high density of A β-containing neuriticplaques in the cingulate cortex but only a low den-sity in the striatum. At this same time point, thereis a significant reduction in cholinergic enzymeactivity in the cingulate cortex, but no significantreduction in enzyme activity or cholinergic celldensity in the striatum [69].

Neocortical cholinergic nerve terminals degen-erate prior to A β plaque deposition. There is a sig-nificant reduction in the number of cholinergicnerve terminal varicosities in young homozygousPDAPP mice versus age-matched controls, at atime when only a very few A β plaques are present[69]. Other types of studies support this conclu-sion. For example, behavioral impairments [70,71], synaptic transmission deficits [72], and loss of cortical nerve terminal markers in the PDAPPmouse [73] precede the formation of neuriticplaques in APP mouse models of AD. These find-ings are consistent with the hypothesis that nerve

terminal toxicity comes from extracellular solubleforms of A β.

There are markedly swollen ChAT-containingcholinergic nerve terminal varicosities in proximityto mature A β-containing plaques. The morpholog-ical similarity to the APP-positive neuritic plaquesfound in the PDAPP mouse [35] and human ADtissue [74] indicates that neuritic dystrophy associ-ated with A β deposition affects cortical cholinergicnerve terminals. The swollen cholinergic nerve ter-minals are more than twice the normal size, andtheir density is extensive within the cortex and hip-pocampus of 2-year-old homozygous PDAPP

mice. Similar morphological abnormalities havebeen observed in cholinergic synapses in mice car-rying a mutation in APP [64, 68] and double muta-tions in APP and PS1 [63, 66, 67]. Likewise, theswollen cholinergic nerve terminals have beenidentified using antibodies against ChAT [66, 68,69], the p75 nerve growth factor (NGF) receptor[67], the vesicular acetylcholine transporter [63],and immunostaining for acetylcholinesterase [64].The swelling may be related to the induction of brain-derived neurotrophic factor in plaque-associ-ated glial cells in the APP mouse models [75].

Because cholinergic synaptic transmission isimportant for learning and memory [76, 77] reduc-tions in cholinergic nerve terminals may play a partin the learning deficits observed in APP-transgenic

mice [78, 79] and in the PDAPP mouse [80]. Thesevere cholinergic pathology in the PDAPP mouseis similar to that in end-stage AD postmortem brainwhere there are marked decreases in the density of cholinergic nerve terminals and ChAT enzymeactivity [81, 82].

Neurodegeneration of the basal forebraincholinergic neurons is one of the cardinal featuresof AD; however, in AD mouse models these neu-rons do not degenerate. In the PDAPP mouse,there is no reduction in the number of basal fore-brain cholinergic somata in the aged homozygousPDAPP mouse (Fig. 16.4) [69]. At 2 years of age,there are a similar number of basal forebraincholinergic somata in homozygous PDAPP miceversus 2-month-old homozygous PDAPP mice.The basal forebrain cholinergic somata collec-tively within the medial septal nucleus and in thevertical and horizontal limbs of the diagonal bandof Broca project to the cingulate cortex and hip-pocampus in the rodent [83, 84], both of which are

regions that contain dense accumulations of A β-containing neuritic plaques in the 2-year-old ani-mals. In hemizygous APP transgenic mice, there isalso no loss of basal forebrain cholinergic neurons[64, 68], nor in APP SWE /PS1 M146L transgenic mice[67]. The lack of reduction in the number of basalforebrain cholinergic somata in the APP mousemodels differ from that observed in AD patients,perhaps because the pathologic process in the ani-mals lasts for a much shorter time period than istypical in man. It is also possible that expressionof genes or activation of proteins that play a role inneuroprotection occur in the APP mouse models




264 D.C. German

FIGURE 16.3. There is a marked decrease in cortical cholinergic markers in the PDAPP mouse. (A) The density of nerve fibers, immunostained for ChAT, is decreased in the cingulate cortex and hippocampus of the 2-year-oldPDAPP mouse. ChAT fiber density is illustrated in the control 2-year-old mouse and in a 2-year-old PDAPP animalin both the cingulate cortex and hippocampus. Arrows in the hippocampus of the PDAPP mouse illustrate CA1 and

CA3 regions, which contain clear losses of ChAT immunostained fibers. Abbreviations: CA1, CA1 field of the hip-pocampus; CA3, CA3 field of the hippocampus; DG, dentate gyrus. Marker, 70 µm. (B) There is an age-relateddecrease in the density of cholinergic varicosities in the PDAPP mouse. Homozygous PDAPP mice and age-matchedcontrol mice were examined at 2 months, 4 months, 1 year, and 2 years of age. Data represent ChAT varicosity den-sity (varicosities × 106 /mm 3) for individual animals in the cingulate cortex as a percent of the age-matched controlmice. (C) ChAT enzyme activity is significantly decreased in the cingulate cortex, but not in the striatum, of 2-year-old PDAPP compared with age-matched control mice. Data represent values for individual mice (nmol mg protein −1

h−1). There was a significant 18% average reduction (asterisk) in enzyme activity in the cingulate cortex (Student’s t =3.27, p < 0.04), but no change in enzyme activity in the striatum (Student’s t = 1.42). From German et al. [69].

that counter the neurotoxic effects of A β, as reportedfor the APP sw mouse model of AD [85, 86]. It is alsopossible that NFTs are important for neurodegener-ation to occur, and thus it will be interesting to deter-

mine whether the cholinergic neurons degenerate inmouse models that have NFTs [39, 40].

The loss of cholinergic nerve terminals in ADmouse models, without a loss of basal forebrain



cholinergic somata, is consistent with the hypothesisthat the neuropathology begins in the cerebral cortexand hippocampus prior to spreading in a retrogradefashion to subcortical regions [87]. The density of cholinergic nerve terminals in the cortex is reduced byapproximately 65% in the 2-year-old PDAPP mouseversus age-matched non-transgenic controls, yet there

is no reduction in the number of basal forebraincholinergic somata that innervate this cortical region[69]. Likewise, in 2-year-old APP23 mutant mice,which carry a lower A β burden than in the homozy-gous PDAPP mice, there is a 29% reduction in totalcholinergic fiber length in the cerebral cortex and noloss of basal forebrain cholinergic somata [68].


FIGURE 16.4. There is no age-related change in the number of basal forebrain cholinergic neurons in the PDAPPmouse. Basal forebrain cholinergic neurons were examined in the medial septal (MS) nucleus, and in the vertical(VDB) and horizontal limb (HDB) of the diagonal band of Broca. Representative sections, immunostained with anantibody against ChAT, are illustrated at rostral (A), middle (B), and caudal (C) locations where the basal forebraincells were counted. Abbreviations: aca, anterior commissure, anterior; acp, anterior commissure, posterior. Marker,

300 µm. (D) There is no difference in the number of basal forebrain somata per tissue section throughout the rostral-caudal 1.0 mm of the basal forebrain in 2-month-old versus 2-year-old PDAPP mice. Illustrated are the mean totalnumber of somata per tissue section ± SEM (n = 6/group) for sections from rostral (0 µm distance) to caudal (1000µm) within the basal forebrain complex of the two mouse groups. From German et al. [69].



16.3.6 Noradrenergic Cell Loss

There is significant loss of LC neurons in AD[12–14], however, it is not found in the one ADmouse model reported to date, the PDAPP mouse[88]. Comparing 2-year-old homozygous animals

with 2-month-old homozygous animals, the rostral-caudal distribution of LC neurons is similar. It isinteresting that there is a cell shrinkage selectivelywithin the region of the LC where cells reside thatproject to the cortex and hippocampus [88], sug-gesting that these neurons are in the early stage of degeneration. It will be interesting to determinewhether AD mouse models that exhibit NFTs willexhibit loss of LC neurons that project selectively tothe forebrain regions where A β-pathology exists, as

in AD [13]. The NFTs, however, do not appear to beresponsible for all of the neurodegeneration thatoccurs in mouse models as some loss of hippocam-pal neurons occurs in APP mouse models that donot express NFTs [38, 59]. In addition, in Neimann-Pick type C (NPC) disease, there is neurodegenera-tion and NFT formation in man [89]; however, inthe NPC mouse there is marked neurodegenerationwithout tangle formation [90, 91].

16.3.7 NeurogenesisWith AD mouse models, changes in adult hip-pocampal neurogenesis can actually be quantified,in contrast with the qualitative approach requiredin human postmortem studies. Using quantitativeanalysis, adult neurogenesis has been observed tobe decreased in several AD mouse models.Neurogenesis is decreased in an APP mousemodel of AD (Tg2576 mouse) in the subependy-mal zone, a region of the brain that gives rise to

olfactory neurons [92]. Notably, adult neurogene-sis is also decreased in the hippocampal subgranu-lar zone (SGZ), which gives rise to dentate gyrusneurons, in three different AD mouse models[93–95]. The Tg2576 mouse [93] and the PDAPPmouse [95] show an age-related decrease in SGZneurogenesis. In the homozygous PDAPP mouse,neurogenesis is markedly decreased in the hip-pocampus of 1-year-old animals, and there is a38% decrease in the number of granule cells in thedentate gyrus [95]. Given that the PDAPP mousemodel of AD shows decreased hippocampal vol-ume, an age-related loss of cholinergic input to the

cortex and hippocampus (e.g., Ref. 69), anddeficits in hippocampal function [78, 96], it will beinteresting to determine whether treatments thatrestore learning and memory and reduce A β-plaque neuropathology can ameliorate the deficitin hippocampal neurogenesis.

16.4 Future TreatmentPossibilities

At least six strategies have been proposed for thetreatment of AD, which have been tested in ADmouse models. The first potential therapeutic treat-ment for AD used the PDAPP mouse model anddemonstrated that immunization with the human

Aβ42peptide caused a marked reduction in plaquepathology when given to older animals. In addition,when immunization was given to young animals, itblocked the development of plaque pathology as theanimals aged [97]. A β-immunization also reducesamyloid deposition in the Tg2576 mouse model of AD [98]. Similar findings were reported afterimmunization with antibodies against A β42. Forexample, Janus et al. [99] found that A β antibodyimmunization reduced memory impairment andplaque pathology in an AD mouse model, andDodart et al. [49] found that immunization with A β-antibody m266 reversed the memory impairment inthe PDAPP mouse even before there were reduc-tions in A β-plaque neuropathology. Kotilinek et al.[100] demonstrated that immunization with A β-antibody BAM10 reversed the memory impairmentin the Tg2576 mouse model of AD. Because thecognitive impairments are improved after such ashort antibody treatment, it is unlikely that theimprovement was due to structural changes in the

brain and perhaps reflects removal of extracellularAβ42 oligomers from the synaptic environment [47].

When the A β peptide immunization approachwas used on AD patients, aseptic meningoen-cephalitis occurred in 6% of the patients, and thetrial was stopped [101, 102]. However, recent datafrom a group of the immunized patients indicatethat after 1 year, the patients still had high levelsof A β42 antibody in blood, and the “dementiascore” was no different from a year previouslyversus a decline in dementia score in controlpatients that were not immunized [103]. Thesedata suggest that some form of immunization

266 D.C. German



therapy may be of benefit to AD patients; how-ever, the success may depend upon the degree of cerebral amyloid angiopathy (CAA) in specificpatients. Recent data suggest that the antibodytarget (N-terminal vs. central domain directed)has an effect on the induction of CAA in thePDAPP mouse [104], which may provide insightinto the optimal design for future A β-antibodiesfor immunization therapy.

Epidemiological data indicate that long-termnonsteroidal anti-inflammatory drug (NSAID)treatment has dramatic effects on the incidence of AD [105] resulting in a reduction of risk by asmuch as 60–80% [106, 107]. The NSAID ibupro-fen has been used in the Tg2576 mouse model of AD and found to significantly decrease A β-neuritic

plaques, and decrease brain levels of A β42 peptide,by a mechanism independent of its anti-inflamma-tory effects [56, 108]. Similar beneficial effects of reducing AD neuropathology have been found withdifferent NSAID drugs in an APP mouse model(e.g., Ref. 109). However, additional work isneeded to identify which NSAIDs will provideanti-AD effects because some compounds (e.g.,celecoxib) increase brain A β42 in the brains of Tg2576 mice via effects of γ -secretase [110].

Treatments have been proposed that would slowthe production of the A β42 peptide. Inhibitors of thetwo proteases, β- and γ -secretase, which cleave A βfrom APP have been developed. However, the cur-rent β-secretase inhibitors do not easily cross theblood-brain barrier, and γ -secretase inhibition canpotentially inhibit Notch signaling [111] and pro-duce adverse effects. In mice that have signifi-cantly reduced levels of PS function, there isseborrheic keratosis and autoimmune disease[112]. This treatment strategy will require careful

testing in AD mouse models.Another approach for the treatment of AD

involves modulation of cholesterol homeostasis.Chronic use of cholesterol-lowering drugs, thestatins, is associated with a lowered incidence of AD [113, 114]. High-cholesterol diets have beenfound to increase A β neuropathology in APPmouse models [115, 116], and cholesterol-lower-ing drugs reduce neuropathology in APP mice[117]. However, a recent study questions the use of statins in females because although lovastatin low-ered cholesterol in both male and female Tg2576mice, it increased the number of plaques in the hip-

pocampus and cortex of females but not males[118]. In addition, the beneficial effects of statinsfor AD may also derive from their ability to reducethe microglial inflammatory response [119].

Another strategy for lowering A β concentrationsin brain is based on the observation that A β aggre-gation is partly dependent upon the metal ions Cu 2+

and Zn 2+. Aβ deposition was reduced in APP trans-genic mice treated with the antibiotic clioquinol,which is a chelator of Cu 2+ and Zn 2+ [120]. Humanclinical trials with clioquinol are in progress.

Recent studies have also examined the effects of environmental enrichment and dietary supplementson AD neuropathology in mouse models of the dis-ease. Two studies have examined whether volun-tary exercise has an effect on A β plaque load and

brain peptide levels and also cognitive function[121, 122]. One of the studies used the TgCRND8mouse, which expresses two mutations in APP, andfound that 5 months of voluntary exercisedecreased amyloid plaque load and improved cog-nitive function, and the effect was related to alteredAPP processing [121]. The other study used theTg2576 mouse model of AD and found that 6months of voluntary exercise improved cognitivefunction, but amyloid plaque pathology wasenhanced [122]. The latter study demonstrates thatcognitive function is not positively correlated withplaque pathology, and both studies support clinicaldata showing that people leading a physicallyactive life have a lower incidence of AD. Finally,using the aged Tg2576 mouse model of AD, it hasbeen demonstrated that increased intake of theomega-3 polyunsaturated fatty acid docosa-hexaenoic acid reduces brain levels of A β [123].

The current AD mouse models are being usedfor testing putative AD therapies and their effects

on specific aspects of AD neuropathology. SeveralAD mouse models exhibit an age-related reductionin the density of cholinergic nerve terminal vari-cosities without a reduction in the numbers of basalforebrain cholinergic somata (e.g., Ref. 69). Willearly administration of therapies that reduce plaquepathology and restore learning/memory in ADmouse models, like NSAIDs and immunizationwith A β42 peptides, block cholinergic nerve termi-nal degeneration? In the bigenic AD mouse modelof Schmitz et al. [38], which exhibits degenerationof CA1 hippocampal neurons, will some of theabove AD therapies block and/or reduce the mag-




nitude of NFTs and neurodegeneration? Becauseadult hippocampal neurogenesis is abnormal in AD[32] and abnormal in APP mouse models [93, 95],will therapies that reduce brain concentrations of Aβ42 normalize neurogenesis? Once a mousemodel is developed that mimics all of the majorneuropathologic features of the human disease(Aβ-plaques, NFTs, and neurodegeneration), theseand numerous other questions can be more fullyaddressed in the process of finding novel therapiesfor the treatment of the human condition.

Acknowledgments This work was supported byThe Carl J. and Hortense M. Thomsen Chair inAlzheimer’s Disease Research and the NIH/NIACenter.

References

1. Wong PC, Cai H, Borchelt DR, et al. Geneticallyengineered mouse models of neurodegenerative dis-eases. Nat Neurosci 2002; 5:633-39.

2. Higgins GA, Jacobsen H. Transgenic mouse modelsof Alzheimer’s disease; phenotype and application.Behav Pharmacol 2003; 14:419-38.

3. German DC, Eisch AJ. Mouse models of Alzheimer’s disease: insight into treatment. Rev

Neurosci 2004; 15:353-69.4. Braak H, Braak E. Neuropathological stageing of

Alzheimer-related changes. Acta Neuropathol 1991;82:239-59.

5. Bowen DM, Smith CB, White P, et al. Neuro-transmitter-related enzymes and indices of hypoxiain senile dementia and other abiotrophies. Brain1976; 99:459-96.

6. Davies P, Maloney AJ. Selective loss of centralcholinergic neurons in Alzheimer’s disease. Lancet1976; 2:1403.

7. Perry EK, Perry RH, Blessed G, et al. Necropsy evi-dence of central cholinergic deficits in senile demen-tia. Lancet 1977; 1:189.

8. Whitehouse PJ, Price DL, Struble RG, et al. Alzheimer’sdisease and senile dementia: loss of neurons in the basalforebrain. Science 1982; 215:1237-39.

9. Mesulam M-M, Geula C. Nucleus basalis (Ch4) andcortical cholinergic innervation in the human brain:observations based on the distribution of acetyl-cholinesterase and choline acetyltransferase. J CompNeurol 1988; 275:216-40.

10. Yamamoto T, Hirano A. Nucleus raphe dorsalis inAlzheimer’s disease: neurofibrillary tangles and lossof large neurons. Ann Neurol 1985; 17:573-77.

11. Mann DMA, Yates PO, Marcyniuk B. Dopaminergicneurotransmitter systems in Alzheimer’s disease andin Down’s syndrome at middle age. J NeurolNeurosurg Psychiatry 1987; 50:341-44.

12. Iversen LL, Rossor MN, Reynolds GP, et al. Loss of pigmented dopamine- β-hydroxylase positive cells

from locus coeruleus in senile dementia of Alzheimer’s type. Neurosci Lett 1983; 39:95-100.

13. German DC, Manaye KF, Smith WK, et al. Disease-specific patterns of locus coeruleus cell loss. AnnNeurol 1992; 32:667-76.

14. Chan-Palay V, Asan E. Alterations in catecholamineneurons of the locus coeruleus in senile dementia of the Alzheimer’s type and in Parkinson’s disease withand without dementia and depression. J CompNeurol 1989; 287:373-92.

15. Kalaria RN. Microglia and Alzheimer’s disease. CurrOpin Hematol 1999; 6:15-24.

16. Altman J, Das GD. Autoradiographic and histologi-cal evidence of postnatal hippocampal neurogenesisin rats. J Comp Neurol 1965; 124:319-35.

17. Ericksson PS, Perfilieva E, Bjork-Eriksson T, et al.Neurogenesis in the adult human hippocampus. NatMed 1998; 4:1313-317.

18. Cameron HA, Woolley CS, McEwen BS, et al.Differentiation of newly born neurons and glia in thedentate gyrus of the adult rat. Neuroscience 1993;56:337-44.

19. Hastings NB, Gould E. Rapid extension of axons into

the CA3 region by adult-generated granule cells. JComp Neurol 1999; 413:146-54.

20. Markakis EA, Gage FH. Adult-generated neurons inthe dentate gyrus send axonal projections to fieldCA3 and are surrounded by synaptic vesicles. JComp Neurol 1999; 406:449-60.

21. Cameron H A, McKay RD. Adult neurogenesis pro-duces a large pool of new granule cells in the dentategyrus. J Comp Neurol 2001; 435:406-17.

22. Gould E, Reeves AJ, Fallah M, et al. Hippocampalneurogenesis in adult old world primates. Proc Natl

Acad Sci U S A 1999; 96:5263-67.23. Gould E, Vail N, Wagers M, et al. Adult-generatedhippocampal and neocortical neurons in macaqueshave a transient existence. Proc Natl Acad Sci USA2001; 98:10910-17.

24. Kornack DR, and Rakic P. Continuation of neuroge-nesis in the hippocampus of the adult macaque mon-key. Proc Natl Acad Sci U S A 1999; 96:5768-73.

25. Boonstra R, Galea L, Matthews S, et al. Adult neu-rogenesis in natural populations. Can J PhysiolPharmacol 2001; 79:297-302.

26. Eisch AJ, Nestler EJ. To be or not to be: adult neuro-genesis and psychiatry. Clin Neurosci Res 2001; 2:93-108.

268 D.C. German



27. Eisch AJ. Adult neurogenesis: implications for psy-chiatry. Prog Brain Res 2002; 138:317-44.

28. Abrous DN, Adriani W, Montaron MF, et al.Nicotine self-administration impairs hippocampalplasticity. J Neurosci 2002; 22:3656-62.

29. Kempermann G, Kuhn HG, Gage FH. Experience-

induced neurogenesis in the senescent dentate gyrus.J Neurosci 1998; 18:3206-12.

30. Kuhn HG, Dickinson-Anson H, Gage FH.Neurogenesis in the dentate gyrus of the adult rat:age-related decrease of neuronal progenitor prolifer-ation. J Neurosci 1996; 16:2027-33.

31. Vallieres L, Campbell IL, Gage FH et al. Reducedhippocampal neurogenesis in adult transgenic micewith chronic astrocytic production of interleukin-6. JNeurosci 2002; 22:486-92.

32. Jin K, Peel AL, Mao XO, et al. Increased hippocam-pal neurogenesis in Alzheimer’s disease. Proc NatlAcad Sci U S A 2004; 101:343-47.

33. Jordan-Sciutto KL, Dorsey R, Chalovich EM, et al.Expression patterns of retinoblastoma protein inParkinson’s disease. J Neuropathol Exp Neurol 2002;62:68-74.

34. Curtis MA, Penney EB, Pearson AG, et al. Increasedcell proliferation and neurogenesis in the adulthuman Huntington’s disease brain. Proc Natl AcadSci U S A 2003; 100:9023-27.

35. Games D, Adams D, Alessandrini R, et al.Alzheimer-type neuropathology in transgenic mice

overexpressing V717F β-amyloid precursor protein.Nature 1995; 373:523-27.

36. Hsiao K, Chapman P, Nilsen S, et al. Correlative mem-ory deficits, Abeta elevation, and amyloid plaques intransgenic mice. Science 1996; 274:99-102.

37. Borchelt DR, Ratovitski T, van Lare J, et al.Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1and amyloid precursor proteins. Neuron 1997; 19:939-45.

38. Schmitz C, Rutten BPF, Pielen A, et al. Hippocampal

neuron loss exceeds amyloid plaque load in a trans-genic mouse model of Alzheimer’s disease. Am JPathol 2004; 164:1495-502.

39. Lewis J, Dickson DW, Lin W-L, et al. Enhanced neu-rofibrillary degeneration in transgenic mice expressionmutant tau and APP. Science 2001; 293:1487-91.

40. Oddo S, Caccamo A, Shepherd JD, et al. Triple-transgenic model of Alzheimer’s disease withplaques and tangles: intracellular A β and synapticdysfunction. Neuron 2003; 39:409-21.

41. Saura CA, Choi S-Y, Beglopoulos V, et al. Loss of presenilin function causes impairments of memoryand synaptic plasticity followed by age-dependentneurodegeneration. Neuron 2004; 42:23-36.

42. Murrell J, Farlow M, Ghetti B, et al. A mutation inthe amyloid precursor protein associated with hered-itary Alzheimer’s disease. Science 1991; 254:97-99.

43. Lazarov O, Lee M, Peterson DA, et al. Evidence thatsynaptically released [beta]amyloid accumulates asextracellular deposits in the hippocampus of trans-

genic mice. J Neurosci 2002; 22:9785-93.44. Sheng JG, Price DL, Kotiatsos VE. Disruption of

cortico-cortical connections ameliorates amyloidburden in terminal fields in a transgenic model of A[beta] amyloidosis. J Neurosci 2002; 22:9794-99.

45. Walsh DM, Hartley DM, Kusumoto Y, et al. Amyloidβ-protein fibrillogenesis: structure and biologicalactivity of protofibrillar intermediates. J Biol Chem1999; 274:25945-52.

46. Hartley DM, Walsh DM, Ye CP, et al. Protofibrillarintermediates of amyloid β-protein induce acute elec-trophysiological changes and progressive neurotoxic-ity in cortical neurons. J Neurosci 1999; 19:8876-84.

47. Walsh DM, Klyubin I, Fadeeva JV, et al. Naturallysecreted oligomers of amyloid beta protein potentlyinhibit hippocampal long-term potentiation in vivo.Nature 2002; 416:535-39.

48. Klyubin I, Walsh DM, Lemere CA et al. Amyloid βprotein immunotherapy neutralizes A β oligomersthat disrupt synaptic plasticity in vivo. Nat Med2005; 11:556-61.

49. Dodart J-C, Bales KR, Gannon KS, et al.Immunization reverses memory deficits without

reducing brain A β burden in Alzheimer’s diseasemodel. Nat Neurosci 2002; 5:452-57.

50. Zohar O, Cavallaro S, Agata VD, et al.Quantification and distribution of β-secretase alter-native splice variants in the rat and human brain. MolBrain Res 2003; 115:63-68.

51. Masliah E, Sisk A, Mallory M, et al. Neurofibrillarypathology in transgenic mice overexpressing V717Fbeta-amyloid precursor protein. J Neuropathol ExpNeurol 2001; 60:357-68.

52. Tomidokoro Y, Ishiguro K, Harigaya Y, et al. Abeta

amyloidosis induces the initial stage of tau accumu-lation in APP(Sw) mice. Neurosci Lett 2001; 299:169-72.

53. Sturchler-Pierrat C, Abramowski D, Duke M, et al.Two amyloid precursor protein transgenic mousemodels with Alzheimer’s disease-like pathology.Proc Natl Acad Sci U S A, 1997; 94:13287-92.

54. Kurt MA, Davies DC, Kidd M, et al.Hyperphosphorylated tau and paired helical fila-ment-like structures in the brains of mice carryingmutant amyloid precursor protein and mutant prese-nilin-1 transgenes. Neurobiol Dis 2003; 14:89-97.

55. Schwab C, Hosokawa M, McGeer PL. Transgenicmice overexpressing amyloid beta protein are an




incomplete model of Alzheimer’s disease. ExpNeurol 2004; 188:52-64.

56. Yan Q, Zhang J, Liu H, et al. Anti-inflammatory drugtherapy alters β-amyloid processing and depositionin an animal model of Alzheimer’s disease. JNeurosci 2003; 23:7504-09.

57. Richards JG, Higgins GA, Ouagazzal A-M, et al.PS2APP transgenic mice, coexpressing hPS2mut andhAPPswe, show age-related cognitive deficits associ-ated with discrete brain amyloid deposition andinflammation. J Neurosci 2003; 23:8989-03.

58. Irizarry MC, Soriano F, McNamara M, et al. A β dep-osition is associated with neuropil changes, but notwith overt neuronal loss in the human amyloid pre-cursor protein V717F (PDAPP) transgenic mouse. JNeurosci 1997; 17:7053-59.

59. Calhoun ME, Wiederhold K-H, Abramowski D, et al.Neuron loss in APP transgenic mice. Nature 1998;395:755-56.

60. Gonzales-Lima F, Berndt JD, Valla JE, et al.Reduced corpus callosum, fornix and hippocampusin PDAPP transgenic mouse model of Alzheimer’sdisease. NeuroReport 2001; 12:2375-79.

61. Redwine JM, Kosofsky B, Jacobs RE, et al. Dentategyrus volume is reduced before the onset of plaqueformation in PDAPP mice: a magnetic resonancemicroscopy and stereologic analysis. Proc Natl AcadSci U S A 2003; 100:1381-86.

62. Casas C, Sergeant N, Itier J-M, et al. Massive CA1/2

neuronal loss with intraneuronal and N-terminal trun-cated Abeta42 accumulation in a novel Alzheimertransgenic model. Am J Pathol 2004; 165: 1289-300.

63. Wong TP, Debeir T, Duff K, et al. Reorganization of cholinergic terminals in the cerebral cortex and hip-pocampus in transgenic mice carrying mutated pre-senilin-1 and amyloid precursor protein transgenes.J Neurosci 1999; 19:2706-16.

64. Bronfman FC, Moechars D, Van Leuven F.Acetylcholinesterase-positive fiber deafferentationand cell shrinkage in the septohippocampal pathway

of aged amyloid precursor protein London mutanttransgenic mice. Neurobiol Dis 2000; 7:152-68.65. Sturchler-Pierrat C, Staufenbiel M. Pathogenic

mechanisms of Alzheimer’s disease analyzed in theAPP23 transgenic mouse model. Ann N Y Acad Sci2000; 920:134-39.

66. Hernandez D, Sugaya K, Qu T, et al. Survival andplasticity of basal forebrain cholinergic systems inmice transgenic for presenilin-1 and amyloid precur-sor protein mutant genes. NeuroReport 2001; 12:1377-84.

67. Jaffar S, Counts SE, Ma SY, et al. Neuropathology of mice carrying mutant APP swe and/or PS1 M146L trans-genes: alterations in the p75 NTR cholinergic basal

forebrain septohippocampal pathway. Exp Neurol2001; 170:227-43.

68. Boncristiano S, Calhoun ME, Kelly PH, et al.Cholinergic changes in the APP23 transgenic mousemodel of cerebral amyloidosis. J Neurosci 2002; 22:3234-43.

69. German DC, Yazdani U, Speciale SG, et al.Cholinergic neuropathology in a mouse model of Alzheimer’s disease. J Comp Neurol 2003; 462:371-81.

70. Holcomb L, Gordon MN, McGowan E, et al.Accelerated Alzheimer-type phenotype in trans-genic mice carrying both mutant amyloid precursorprotein and presenilin 1 transgenes. Nat Med 1998;4:97-100.

71. Moechars D, Dewachter I, Lorent K, et al. Early phe-notypic changes in transgenic mice that overexpressdifferent mutants of amyloid precursor protein inbrain. J Biol Chem 1999; 274:6483-92.

72. Hsia A, Masliah E, McConlogue L, et al. Plaque-independent disruption of neural circuits inAlzheimer’s disease mouse models. Proc Natl AcadSci U S A 1999; 96:3228-33.

73. Mucke L, Masliah E, Yu GQ, et al. High-level neu-ronal expression of A β1-42 in wild-type human amy-loid protein precursor transgenic mice:synaptotoxicity without plaque formation. J Neurosci2000; 20:4050-58.

74. Joachim C, Games D, Morris J, et al. Antibodies to

non-beta regions of the beta-amyloid precursor pro-tein detect a subset of senile plaques. Am J Pathol1991; 138:373-84.

75. Burbach GJ, Hellweg R, Haas CA, et al. Induction of brain-derived neurothophic factor in plaque-associ-ated glial cells in aged APP23 transgenic mice.J Neurosci 2004; 24:2421-30.

76. Coyle JT, Price DL, DeLong MR. Alzheimer’s dis-ease: a disorder of cortical cholinergic innervation.Science 1983; 219:1184-90.

77. Fibiger HC. Cholinergic mechanisms in learning,

memory and dementia: a review of recent evidence.Trends Neurosci 1991; 14:220-23.78. Chapman PF, White GL, Jones MW, et al. Impaired

synaptic plasticity and learning in aged amyloid pre-cursor protein transgenic mice. Nat Neurosci 1999;2:271-76.

79. Morgan D, Diamond DM, Gottschall PE, et al. A βpeptide vaccination prevents memory loss in an ani-mal model of Alzheimer’s disease. Nature 2000; 408:982-85.

80. Chen G, Chen KS, Knox J, et al. A learning deficitrelated to age and β-amyloid plaques in a mousemodel of Alzheimer’s disease. Nature 2000; 408:975-79.

270 D.C. German



81. DeKosky ST, Ikonomovic MD, Styren SD, et al.Upregulation of choline acetyltransferase activity inhippocampus and frontal cortex of elderly subjectswith mild cognitive impairment. Ann Neurol 2002;51:145-55.

82. Geula C, Mesulam M-M. Systematic regional varia-

tions in the loss of cortical cholinergic fibers inAlzheimer’s disease. Cerebral Cortex 1996; 6: 165-77.

83. McKinney M, Coyle JT, Hedreen JC. Topographicanalysis of the innervation of the rat neocortex andhippocampus by the basal forebrain cholinergic sys-tem. J Comp Neurol 1983; 217:103-21.

84. Rye DB, Wainer BH, Mesulam MM, et al. Corticalprojections arising from the basal forebrain: a studyof cholinergic and noncholinergic componentsemploying combined retrograde tracing andimmunohistochemical localization of choline acetyl-transferase. Neuroscience 1984; 13:627-43.

85. Stein TD, Johnson JA. Lack of neurodegeneration intransgenic mice overexpressing mutant amyloid pre-cursor protein is associated with increased levels of transthyretin and the activation of cell survival path-ways. J Neurosci 2002; 22:7380-88.

86. Stein TD, Anders NJ, DeCarli C et al. Neutralizationof transthyretin reverses the neuroprotective effectsof secreted amyloid precursor protein (APP) inAPP SW mice resulting in tau phosphorylation andloss of hippocampal neurons: support for the amyloidhypothesis. J Neurosci 2004; 24:7707-717.

87. Saper CB, Wainer B, German DC. Axonal andtransneuronal transport in the transmission of neuro-logical disease: potential role in system degenera-tions, including Alzheimer’s disease. Neuroscience1987; 23:389-98.

88. German DC, Nelson O, Liang F, et al. The PDAPPmouse model of Alzheimer’s disease: locuscoeruleus neuronal shrinkage. J Comp Neurol 2005;492:469-76.

89. Love S, Bridges LR, Case CP. Neurofibrillary tanglesin Niemann-Pick disease type C. Brain 1995; 118:

119-29.90. Tanaka J, Nakamura H, Miyawaki S. Cerebellarinvolvement in murine sphingomyelinosis: a newmodel of Niemann-Pick disease. J Neuropathol ExpNeurol 1988; 47:291-00.

91. German DC, Quintero EM, Liang C-L, et al Selectiveneurodegeneration, without neurofibrillary tangles,in a mouse model of Niemann-Pick C disease. JComp Neurol 2001; 433:415-25.

92. Haughey NJ, Nath A, Chan SL, et al. Disruption of neurogenesis by amyloid β-peptide, and perturbedneural progenitor cell homeostasis, in models of Alzheimer’s disease. J Neurochem 2002; 83:1509-24.

93. Haughey NJ, Liu D, Nath A, et al. Disruption of neurogenesis in the subventricular zone of adultmice, and in human cortical neuronal precursorcells in culture, by amyloid beta-peptide: implica-tions for the pathogenesis of Alzheimer’s disease.Neuromol Med 2002; 1:125-35.

94. Feng R, Rampon C, Tang YP, et al. Deficient neuro-genesis in forebrain-specific presenilin-1 knockoutmice is associated with reduced clearance of hip-pocampal memory traces. Neuron 2001; 32:911-26.

95. Donovan MH, Yazdani U, Norris RD, et al.Decreased adult hippocampal neurogenesis in thePDAPP mouse model of Alzheimer’s disease. JComp Neurol 2006; 495:70-83.

96. Lanz TA, Carter DB, Merchant KM. Dendriticspine loss in the hippocampus of young PDAPP andTg2576 mice and its prevention by the ApoE2genotype. Neurobiol Dis 2003; 13:246-53.

97. Schenk D, Barbour R, Dunn W, et al. Immunizationwith amyloid-beta attenuates Alzheimer’s disease-like pathology in the PDAPP mouse. Nature 1999;400:173-77.

98. Das P, Howard V, Loosbrock N, et al. Amyloid- βimmunization effectively reduces amyloid deposi-tion in FcR γ − / − knock out mice. J Neurosci 2003;23:8532-38.

99. Janus C, Pearson J, McLaurin J, et al. A β peptideimmunization reduces behavioural impairment andplaques in a model of Alzheimer’s disease. Nature

2000; 408:979-82.100. Kotilinek LA, Bacskai B, Westerman M, et al.

Reversible memory loss in a mouse transgenicmodel of Alzheimer’s disease. J Neurosci 2002; 22:6331-335.

101. Schenk D. Amyloid- β immunotherapy forAlzheimer’s disease: the end of the beginning. NatRev Neurosci 2002; 3:824-28.

102. Orgogozo J-M, Gilman S, Dartigues J-F, et al.Subacute meningoencephalitis in a subset of patients with AD after A β42 immunization.

Neurology 2003; 61:46-54.103. Hock C, Konietzko U, Streffer JR, et al. Antibodiesagainst β-amyloid slow cognitive decline inAlzheimer’s disease. Neuron 2003; 38:547-54.

104. Racke MM, Boone LI, Hepburn DL, et al.Exacerbation of cerebral amyloid antiopathy-asso-ciated microhemorrhage in amyloid precursor pro-tein transgenic mice by immunotherapy isdependent on antibody recognition of depositedforms of amyloid β. J Neurosci 2005; 25:629-36.

105. McGeer PL, Schulzer M and McGeer EG. Arthritisand anti-inflammatory agents as possible protectivefactors for Alzheimer’s disease: a review of 17 epi-demiologic studies. Neurology 1996; 47:425-32.




106. Stewart WF, Kawas C, Corrada M, et al. Risk of Alzheimer’s disease and duration of NSAID use.Neurology 1997; 48:626-32.

107. in t’ Veld BA, Ruitenberg A, Hofman A, et al.Nonsteroidal anti-inflammatory drugs and the risk of Alzheimer’s disease. N Engl J Med 2001; 345:

1515-21.108. Lim GP, Yang F, Chu T, et al. Ibuprofen suppresses

plaque pathology and inflammation in a mousemodel of Alzheimer’s disease. J Neurosci 2000; 20:5709-14.

109. Jantzen PT, Connor KE, DiCarlo G, et al.Microglial activation and β-amyloid deposit reduc-tion caused by a nitric oxide-releasing nonsteroidalanti-inflammatory drug in amyloid precursor pro-tein plus presenilin-transgenic mice. J Neurosci2002; 22:2246-54.

110. Kukar T, Murphy MP, Eriksen JL, et al. Diversecompounds mimic Alzheimer’s disease-causingmutations by augmenting A β42 production. NatMed 2005; 11:545-50.

111. Haass C, DeStrooper B. The presenilins inAlzheimer’s disease -proteolysis holds the key.Science 1999; 286:916-19.

112. Tournoy J, Bossuyt X, Snellinx A, et al. Partial lossof presenilins causes seborrheic keratosis andautoimmune disease in mice. Hum Mol Genet2004; 13:1321-31.

113. Wolozin B, Kellman W, Ruosseau P, et al.

Decreased prevalence of Alzheimer’s disease asso-ciated with 3-hydroxy-3-methyglutaryl coenzymeA reductase inhibitors. Arch Neurol. 2000; 57:1439-443.

114. Jick H, Zornberg GL, Jick SS, et al. Statins and therisk of dementia. Lancet 2000; 356:1627-31.

115. Sparks DL, Kuo YM, Roher A, et al. Alterations of Alzheimer’s disease in the cholesterol-fed rabbit,including vascular inflammation. Preliminary obser-vations. Ann N Y Acad Sci 2000; 903:335-44.

116. Refolo LM, Malester B, LaFrancois J, et al.

Hypercholesterolemia accelerates the Alzheimer’samyloid pathology in a transgenic mouse model.Neurobiol Dis 2000; 7:321-31.

117. Refolo LM, Pappolla MA, LaFrancois J, et al. Acholesterol-lowering drug reduces beta-amyloidpathology in a transgenic mouse model of Alzheimer’s disease. Neurobiol Dis 2001; 8:890-99.

118. Park IH, Hwang EM, Hong HS, et al. Lovasatinenhances Abeta production and senile plaque depo-sition in female Tg2576 mice. Neurobiol Aging2003; 24:637-43.

119. Cordle A, Landreth G. 3-Hydroxy-3-methylglu-taryl-coenzyme A reductase inhibitors attenuate β-

amyloid-induced microglial inflammatory respon-ses. J Neurosci 2005; 25:299-07.

120. Cherny RA, Atwood CS, Xilinas ME, et al.Treatment with a copper-zinc chelator markedlyand rapidly inhibits beta-amyloid accumulation inAlzheimer’s disease transgenic mice. Neuron 2001;

30:665-76.121. Adlard PA, Perreau VM, Pop V, et al. Voluntary

exercise decreases amyloid load in a transgenicmodel of Alzheimer’s disease. J Neurosci 2005; 25:4217-4221.

122. Jankowshy JL, Melnikova T, Fadale DJ, et al.Environmental enrichment mitigates cognitivedeficits in a mouse model of Alzheimer’s disease. JNeurosci 2005; 25:5217-224.

123. Lim GP, Calon F, Morihara T, et al. A diet enrichedwith the Omega-3 fatty acid docosahexaenoic acidreduces amyloid burden in an aged Alzheimer’smouse model. J Neurosci 2005; 25:3032-40.

124. Quon D, Wang Y, Catalano R, et al. Formation of beta-amyloid protein deposits in brains of trans-genic mice. Nature 1991; 352:239-41.

125. Higgins LS, Holtzman DM, Rabin J, et al.Transgenic mouse brain histopathology resemblesearly Alzheimer’s disease. Ann Neurol 1994; 35:598-607.

126. Higgins LS, Rodems JM, Catalano R, et al. EarlyAlzheimer’s disease-like histopathology increasesin frequency with age in mice transgenic for beta-

APP751. Proc Natl Acad Sci U S A 1995; 92:4402-406.

127. Dodart JC, Mathis C, Saura J, et al.Neuroanatomical abnormalities in behaviorallycharacterized APP(V717F) transgenic mice.Neurobiol Dis 2000; 7:71-85.

128. Luth HJ, Apelt J, Ihunwo AO, et al. Degeneration of beta-amyloid-associated cholinergic structures in trans-genic APP SW mice. Brain Res 2003; 977:16-22.

129. Bondolfi L, Calhoun M, Ermini F, et al. Amyloid-associated neuron loss and gliogenesis in the neo-

cortex of amyloid precursor protein transgenicmice. J Neurosci 2002; 22:515-22.130. Chishti MA, Yang DS, Janus C, et al. Early-onset

amyloid deposition and cognitive deficits in trans-genic mice expressing a double mutant form of amyloid precursor protein 695. J Biol Chem 2001;276:21562-70.

131. Takeuchi A, Irizarry MC, Duff D, et al. Age-relatedamyloid beta deposition in transgenic mice overex-pressing both Alzheimer mutant presenilin 1 andamyloid beta precursor protein Swedish mutant isnot associated with global neuronal loss. Am JPathol 2000; 157:331-39.

272 D.C. German



132. Urbanc B, Cruz L, Le R, et al. Neurotoxic effects of thioflavin S-positive amyloid deposits in transgenicmice and Alzheimer’s disease. Proc Natl Acad SciU S A 2002; 99:13990-95.

133. Duff K, Eckman C, Zehr C, et al. Increased amy-loid-beta42(43) in brains of mice expressing mutant

presenilin-1. Nature 1996; 383:710-13.

134. Cheng IH, Palop JJ, Esposito LA, et al. Aggressiveamyloidosis in mice expressing human amyloidpeptides with the Arctic mutation. Nat Med 2004;10:1190-192.




Subject Index

274

AAβ; see also β–amyloid protein—adaptive immune response, 247–250—aggregation, 93–95, 94f, 151—biochemical effects, 95—Cu 2+ and Zn 2+ induced aggregation,

125–127—cytoxicity, 137–138—effect on membrane receptors, 96–97—immune responses, 245–250—immunological properties, 245–246—induced neurodegeneration, 95—innate immunoresponses, 246–247

—interaction with membrane lipids, 96—metal binding effect on interactionwith membranes, 135–137

—metal-binding sites and structure,129–133

—metal coordination role, 133–135,134f

—molecular structure, 245–246—neurotoxicity, 133—redox activity, 133–135—relevance of membrane binding

cytoxicity, 137–138—sequence of, 125, 126—structures, 127–133

—toxicity, 151—treatment approaches targeting, 254Aβ amyloidogenesis and Alzheimer’s

disease, 102ACH synthesis and release—A β effect on, 164–167, 165t, 166f AD brains—A β42 accumulation, 218–219,

219f —astrocytes and A β, 217–223—astrocytes structure and function,

217–218—cell death mechanisms, 234–235—intracellular A β42-deposits in

neurons, 219–220—intraneuronal A β42-accumulation,222–223, 222f

—microglia and A β, 223–227—microglial activity, 226Adhesion molecules, 58–59Aβ(1-42)-induced oxidative stress—Methionine-35 of A β(1-42) role in,

88–89Alzheimer’s disease (AD)—age as risk factor, 6–7—A β amyloidogenesis and, 102—amyloid β–peptide(1-42) and, 83–89—amyloid β–peptide in pathology of,

159–171—amyloid toxicity, 93–97

—anti-A β therapies, 3—anti-inflammatory drug therapy, 70—astrocytes and A β in, 217–223—basal forebrain cholinergic neurons

loss, 160–162—biochemistry of neurodegeneration,

93–97—brain changes, 56–61—cell death, 234–239—central cholinergic neurons in

pathology, 159–171—A β centric pathway, 5–19—cholesterol and, 142–153—cholesterol levels and, 151–152

—clinical diagnosis, 5—clinical presentation of, 198–200,199f–200f

—clinical symptoms and inflammation,67–68

—diagnosis, 5–6—epidemiological and clinical evidence,

151–152—epidemiological and inflammation

findings, 64–65, 72—experimental and clinical research,

250–254—failure of anti-inflammatory drug

treatment, 70

—genetics and cholesterol metabolism,145–148—glial cells and A β peptides in

pathogenesis of, 216–229—histopathology, 1, 198–199—history of β–Amyloid protein, 1–3—A β hypothesis/theory, 3, 6, 8–9—illo tempore, 6—immunotherapeutic approaches to,

245–254—inflammation and etiology, 65–67, 67f —inflammation-related systemic

changes, 63–64—inflammatory response, 53–55, 61–72—mutations, 2–3—neuritic plaques, 160

—neurodegeneration, 93–97—neurofibrillary tangles, 160—neuroimaging, 15, 17–18, 18f —neuroinflammatory response, 43–44, 70—neuropathology, 6, 72, 125–138,

160–162, 259—novel therapeutic approaches, 15–17,

16f —oxidative stress, 12, 88–89—pathogenesis, 2, 6–8, 59–61, 216–229—role of β–amyloid protein, 1–3—statins clinical trials, 152—synaptic dysfunction, 93–97—therapeutic aspects and inflammation,

68–71, 69f —therapeutic strategies, 15–17, 16f —traditional therapeutic approaches, 15,

16f —transgenic mouse models, 11–12,

61–62, 259–268—treatment, 15–18—treatment approaches targeting A β,

254—A β variations impact, 102–119—vascular variant, 68Aβ(1-42)-mediated lipid peroxidation—protein oxidation and, 83–88Aβ–mediated neurotoxicity, 93–95, 94f

—mechanism of, 207f Aβ–membrane interactions, 96–97—with membrane lipids, 96

Page numbers followed by f and t indicate figures and tables, respectively.



Aβ membrane receptors, 96–97—α 7 nicotinic acetylcholine receptor,

97—p75 neurotrophin receptor, 97—RAGE, 96–97β–amyloid—copper coordination, 125–138—impact on Tau pathology, 198–210Amyloid β–peptide—central cholinergic neurons, 159–171Amyloid β–peptide(1-42)—Alzheimer’s disease and, 83–89Amyloid- β peptide ( Αβ) production—cholesterol and, 148–151, 149f Amyloid- β plaques—transgenic mouse models, 260Amyloid precursor protein; see also

β–Amyloid precursor protein—APLP1 gene activities, 41—APLP2 gene activities, 40–41—cellular processing, 39

—cholesterol and, 148–151—cholinergic regulations, 162–164—cholinergic system and, 162–164—cuproprotein, 43–44—function, 39–44—isoforms, 38–39—knockout mice, 41–43—metabolism, 39—modulator of synaptogenesis, 39–40—multidomain molecule, 37–38, 38f —neuroprotective activity, 39—physiological function, 43–44—processing and cholesterol, 148–151,

149f

β–amyloid precursor protein (APP), 7,9–10; see also Amyloid precursorprotein

—functions, 9, 37–44—mismetabolism of, 62, 63f —proteolytic processing of, 1–2, 2f β–amyloid protein—Alzheimer’s disease, 1–3—as cuproprotein, 43–44—genetics of, 9–12—history, 1–3—hypothesis of AD, 3, 6–9—neuroinflammatory response, 43–44,

52–72

—neurotoxicity, 14–15, 19—physiological function of, 43–44—potential probes, 18—role in AD, 1–3—theory behind the hypothesis, 7–9—toxicity, 12–15, 13f, 19—transition metals and, 14–15, 19Aβ neurotoxicity, 93–95, 94f —α 1-antichymotrypsin, 54–55, 57Aβ peptides—ACH synthesis and release, 164–167,

165t, 166f —astrocytes in AD brains, 217–223—binding to other molecules, 184–185

—chemoattractants, 224–225—cholinergic neuron survival, 168–169—cholinergic receptors, 167–168

—cholinergic system regulation by,164–171, 165t, 166f

—clearance from brain, 186–187—detection and tissue location of,

180–182, 181f —functional properties, 185–186—glial cells and, 216–229—microglia in AD Brains, 223–227—neurotoxic properties, 185–186—neurotrophic functions, 185–186—perspectives, 227–229—physiologic and neurotoxic properties,

179–189—physiologic functions, 185—production of, 179–180—structure of, 182–184, 183f —therapeutic strategies for toxicity,

187–189—toxicity, 238–239—whole-cell currents in cholinergic

neurons, 167APLP1 gene activities—knockout mice, 42–43APLP2 gene activities—knockout mice, 42–43ApoE, 11, 54—biological functions, 54—genetics of A β, 11—genetics of AD, 145–146, 146f Apolipoprotein E, see ApoEApoptosis, 236–237—PS processing A βgeneration, 238APP, see β–amyloid precursor proteinAPP processing, 235–236—cholesterol, 148–151

Astrocytes—A β42 accumulation in AD brains,218–219, 219f

—activation of, 218—AD pathology and, 218—blood–brain barrier, 218—A β in AD brains, 217–223—intracellular A β42–accumulation

effects, 221—intraneuronal A β42–accumulation,

222–223, 222f —structure and function, 218–219Atherosclerosis, 67Atorvastatin (Lipitor), 152

Aβ toxicity—cholesterol and statins, 189—immunization, 188—NSAIDs, 188–189—PPAR– γ agonists, 188–189—secretase inhibitors, 187—therapeutic strategies, 187–189Auguste D.’s case clinical history, 6, 52Aβ vaccination—clinical experience, 249–250—human trials, 249–250Aβ variants—A β(1-40), 115–116—A β(1-42), 115–116

—alterations at mid–chain positions,113–115—Alzheimer’s disease progression,

102–119—amyloid formation, 102–119—animal models, 106–109—C-terminal forms, 115–116—modified forms, 102–104—multiple mutations, 116–119—n-terminal truncations and

modifications, 109–113—relevance in molecular or clinical

pathogenesis, 103t—in space and time, 104–106

BBasal forebrain cholinergic neurons loss

in—Alzheimer’s disease, 160–162Blood-brain barrier (BBB), 218Brain; see also AD brains—adhesion molecules, 58–59—A β–associated proteins, 57–58, 57f,

62—changes during Alzheimer’s disease,

56–61—cholesterol metabolism in, 145—leakage of blood–borne substances

into, 218—microglia, 56–57—neuronal changes, 59–61—A β peptides clearance from, 186–187—tumor, 5

CCelecoxib, 69Cell death in AD brains—A β–induced, 238–239

—mechanisms, 234–235Cell loss versus synaptic dystrophy, 93Central cholinergic neurons—Alzheimer’s disease pathology and,

159–171—amyloid β–peptide, 159–171Cerebral hemorrhage with amyloidosis-

Dutch disorder, 68Chemoattractants—DNA fragments as, 224–225Chemotaxis, 224–225Cholesterol—A β aggregation and toxicity, 151—Alzheimer’s disease and, 142–153

—APP processing, 148–151, 149t, 150f —biosynthesis, 143f —levels and AD, 151–152—metabolism, 142–148—statins and, 189—storage and catabolism, 144, 144f —synthesis, 142–143—transport and uptake, 143–144Cholesterol metabolism, 142–148—ApoE, 145–146, 146f —in brain, 145—genes linked to late–onset AD,

146–147—genetics of AD, 145–148

Cholinergic cell loss—transgenic mouse models, 263–265,264f–265f

Subject Index 275



Cholinergic neuron—amyloid interactions, 171—A β effects, 168–169—in vivo administration of A β effects

169–170Cholinergic receptorsAβ effects, 167–168Cholinergic system—ACH synthesis and release, 164–167,

165t, 167f —amyloid precursor protein (APP)

processing, 162–164—cholinergic neuron survival, 168–170—cholinergic receptors, 167–168—A β peptides regulation of, 164–171,

165t, 166f —whole–cell currents in cholinergic

neurons, 167Clinical experience—adaptive immune response to A β,

247–249Clinical symptoms and—inflammation, 67–68Clioquinol (CQ), 17, 125Clusterin, 54, 57Conditioned taste aversion (CTA) test,

206Congophilic angiopathy, 7Copper coordination by β–amyloid—neuropathology of Alzheimer’s

disease and, 125–138Cu 2+ and Zn 2+ induced aggregation A β,

125–127Cuproprotein, 43–44Cyclooxygenase–2 (COX–2)

—role in AD pathogenesis, 59–60,70–71

DDementia, 198Dementia pugilistica, 5Depression, 5DNA fragments as chemoattractants,

224–225Down syndrome, 7–8, 66

EEndoplasmic reticulum (ER), 95–96Epidemiological findings and

inflammation, 64–65FFamilial Alzheimer’s disease (FAD), 2,

198, 200Frontotemporal dementia (FTD), 5, 198,

200–202Functional magnetic resonance imaging

(fMRI), 17

GGenes linked to late-onset AD—ABCA1, 147—ACAT, 147

—cholesterol metabolism and, 146–147—Cyp46, 146–147—LRP and LDLR, 147–148

Genetics of A β, 9–12—ApoE, 11—APP, 9–10, 10f —presenilins, 10–11—transgenic mouse models of AD,

11–12, 61–62, 259–268Genetics of AD—ABCA1, 147—ACAT, 147—ApoE, 145–146, 146f —cholesterol metabolism, 145–148—Cyp46, 146–147—genes linked to late–onset AD,

147–148—LRP and LDRP, 147–148—α2M, 148Glial activation—transgenic mouse models, 260, 262,

262f Glial cells and A β peptides, 216–229Glial fibrillary acidic protein (GFAP),

217–218, 217f HHCHWA-D, see Cerebral hemorrhage

with amyloidosis–Dutch disorderHeparan sulfate proteoglycans, 54–55Hippocampal and cortical cell loss—transgenic mouse models, 262–263HMG-CoA reductase inhibitor, see StatinsHydroxychloroquine, 69

IICAM-1, 54Immunization

—active, 251—adjuvant role, 253–254—aspects of, 251–254—gentechnologic approaches, 253—passive, 251–253Immunotherapy—concept for Alzheimer’s disease, 245Inflammation—clinical symptoms and, 67–68—epidemiological findings and, 64–65—etiology of AD subtypes, 65–67, 67f —related systemic changes in AD

patients, 63–64—therapeutic aspects and, 68–71, 69f

—in transgenic mice models of AD,61–62Intracellular calcium homeostasis,

237–238

LLipid peroxidation—mechanism of, 83–88, 84f Lipid rafts—APP processing and, 149–150, 150f London mutation, 2Lovastatin (Mevacor), 152

M

α2-macroglobulin, 54Magnetic resonance spectroscopy(MRS), 17

Magnetoencephalography (MEG), 17Methionine 35 (A βM35V), 137–138Methionine-35 of A β(1–42)—A β cytoxicity, 137–138—A β(1-42)-induced oxidative stress,

88–89—neurotoxicity, 88–89Microglia, 56–57—and A β in AD brains, 223–227—phagocytic activity of, 223–224—therapeutic targets, 226–227Microglial chemotaxis, 224–225Microglial phagocytosis, 225–226Mild cognitive impairment (MCI), 6Mitochondria—A β–induced cell death, 238–239Mitochondrial dysfunction—oxidative stress and, 95

NNaproxen, 69Neprilysin, 186–187Neuritic plaques and—Alzheimer’s disease, 160Neurodegeneration nature, 5Neurofibrillary tangles—transgenic mouse models, 260, 262f Neurofibrillary tangles (NFTs)—Alzheimer’s disease, 160—plaques and, 198, 202–203—structural component, 5—tau transgenic mouse model, 203–205Neurogenesis—transgenic mouse models, 266Neuroimaging for

—Alzheimer’s disease, 15, 17–18, 18f Neuroinflammation concept, 55–56Neuroinflammatory response—β–amyloid protein involvement in,

52–72Neurons—intracellular A β42-deposits in,

219–220Neuropathology—Alzheimer’s disease, 6, 72, 125–138,

259Neurotoxicity—cellular mechanisms, 93—methionine-35 of A β(1–42) role,

88–89Neurotoxic species—search for, 93–95α7 Nicotinic acetylcholine receptor, 97Nonsteroidal anti-inflammatory drugs

(NSAIDs), 15, 16f, 70–71—A β toxicity, 188–189Noradrenergic cell loss—transgenic mouse models, 266Normal brain function—amyloid β–peptide involvement in,

159–171—central cholinergic neurons

involvement in, 159–171

Normal pressure hydrocephalus, 5NSAIDs, see Nonsteroidalanti–inflammatory drugs

276 Subject Index



OOxidative stress—Alzheimer’s disease and, 83–89—mitochondrial dysfunction, 95Oxidative stress and neurotoxicity—Methionine-35 of A β(1–42) role,

88–89

PParkinson dementia complex, 5Parkinson disease, 7Pathological cascade and—inflammation, 62–63, 63f Plaques—neurofibrillary tangles (NFTs) and,

202–203—senile dementia patients, 52–53, 53f P75 neurotrophin receptor, 97Positron emission tomography (PET),

17–18, 18f PPAR- γ agonists—A

βtoxicity, 188–189

Pravastatin (Pravacor), 152Prednisone, 69Presenilins (PS), 10–11, 235–236—apoptosis, 236–237—A β–generation, 236–237—intracellular calcium homeostasis,

237–238—mutations, 236–238—processing A βgeneration, 238Protein oxidation and—a β(1-42)-mediated lipid peroxidation,

83–88Proteoglycans, 57

RReceptor for advanced glycation end

products (RAGE), 96–97Rofecoxib, 69Rosuvastatin (Crestor), 152

Sγ –secretase complex, 235–236Senile dementia, 6, 52—plaque formation, 52–53, 53f Senile plaques, 52–53, 53f —chronic inflammatory response,

53–55, 54f —A β peptides in, 179, 180f Senium Praecox, 7Serum amyloid P component (SAP),

54–55, 57Simvastatin (Zocor), 152Single photon emission tomography

(SPECT), 17–18Sporadic AD (SAD), 200Statins—cholesterol and, 189—clinical trials, 152Subacute sclerosing panencephalitis, 5Succinimide formation, 104, 105f Synaptic dystrophy versus cell loss, 93

TTau pathology—β–amyloid impact on, 198–210Tau phosphorylation, 169Tau protein, 5Tau transgenic mouse model—β–Amyloid and Tau interaction in,

206–210—conditioned taste aversion test, 206—histopathology and behavioral

impairment correlation, 205–206—neurofibrillary tangles role, 203–205,

204t

—requirements for, 203–205, 204tTherapeutic aspects and—inflammation, 68–71, 69f Tissue culture model—β–amyloid impact on Tau pathology

in, 198–210

Toxicity of β–amyloid protein—energy metabolism, 13–14—inflammatory processes, 13—mechanism, 12–15, 13f —metal homeostasis, 14–15—RNS generation, 13—ROS generation, 12–13Transgenic animals—adaptive immune response to A β,

247–249Transgenic mice overexpressing A β

peptide—cholinergic system in, 170–171Transgenic mouse models, 11–12—Alzheimer’s disease, 259–268—Amyloid- β plaques, 260—cholinergic cell loss, 263–265,

264f–265f —future treatment possibilities,

266–268—gene mutations used, 260, 261t

—glial activation, 260, 262, 262f —hippocampal and cortical cell loss,262–263

—inflammation in, 61–62—neurofibrillary tangles, 260, 262f —neurogenesis, 266—noradrenergic cell loss, 266

WWhole-cell currents in cholinergic

neurons—A β effects on, 167

Subject Index 277



Author Index

278

A

Abdul-Mohmmad, H., 92Abe, E., 177Abe, K., 46Abe, T., 90Abraha, A., 215Abraham, C. R., 74Abraham, I., 20, 172Abraham, W. C., 24Abramov, A. Y., 99, 244Abramowski, D., 172, 213, 269, 270Abrous, D. N., 269Accaviti–Loper, M., 257Ach, K., 231Ackermann, S., 156

Adame, A., 257Adams, D., 26, 77, 172, 190, 213, 269Adams, W. J., 32Adlard, P. A., 272Adlerz, L., 50Adriani, W., 269Adunsky, A., 191Agadjanyan, M. G., 257Agata, V. D., 269Agdeppa, E. D., 35Agosti, C., 155Agostinho, P., 175Ahlers, S. T., 178Aisen, P. S., 76, 82, 197

Akagi, T., 214Akai, K., 232Akbari, Y., 243Akintoye, H., 140Akiyama, H., 73, 230, 233Aksenov, M. Y., 90–91Aksenova, M. V., 90–92Aksnov, M., 91Alafuzoff, I., 173Alauddin, M., 138Alba, F., 120Alberghina, M., 175Albert, M. S., 20, 211Alberts, I. L., 140

Aldskogius, H., 231Alessandrini, R., 26, 77, 172, 190, 213,269

Alexander, P., 25, 242

Alford, M., 90Ali, F. E., 31, 141Ali, F., 27, 92, 140Ali, S. M., 24Allen, B., 214Allen, D. D., 178Allen, N. J., 99Allen, R. G., 28Allinquant, B., 157Allinson, T. M., 47Allsop, D., 72, 81, 123, 139Almenar-Queralt, A., 190Alonso Adel, C., 3, 172Altman, J., 268

Altman, R. A., 24, 75Altmann, S. W., 153Alvarez, A., 176Alvarez, B., 141Alvarez, G., 176Alvaro, V., 244Alves da Costa, C., 244Alzheimer, A., 20Amari, M., 78Amarnath, V., 28Ames, D., 19Amount, N., 175Anandatheerthavarada, H. K., 244Anankov, R., 257

Anders, N. J., 271Andersen, E., 243Anderson, A. J., 75, 240Anderson, C. M., 230Anderson, D. J., 215Anderson, J. J., 196Anderson, J. P., 4, 47Anderson, V. E., 31, 92, 140Anderson, W. J., 176, 231Andersson, P. B., 231Anderton, B. H., 172, 211Ando, K., 50Andreasen, N., 191Andrews, N. C., 24

Ang, L. C., 89Ankarcrona, M., 240–242Ann, N. Y., 197

Annaert, W., 46–47, 240

Antcliffe, D., 99Anthony, J. C., 233Antzutkin, O. N., 139–140Aono, M., 26Apelt, J., 77, 177, 231, 272Arai, H., 80Araki, S., 22Araki, W., 242, 244Araque, A., 230Arawaka, S., 190Archetti, S., 155Ard, M. D., 231Arechaga, G., 120Arends, Y. M., 75

Arendt, T., 76Arevalo, J., 177Argentiero, V., 78Argraves, W. S., 156Aria, H., 80Arima, K., 212Arispe, N., 30, 100, 243Arlaud, G. J., 233Arlt, S., 28Arnold, S. E., 210Arriagada, P. V., 213, 239Arrigada, V., 80Asai, M., 47Asami-Odaka, A., 189

Asan, E., 268Aschner, M., 230Assmann, G.Atack, J. R., 100Atwood, C. S., 26, 30–31, 33, 92, 98,

121, 138–141, 195–196, 272Aubert, A., 81Auld, D. S., 31, 174Austen, B. M., 123Austen, B., 197Authelet, M., 172Avadhani, N. G., 244Avdulov, N. A., 193Averill, D., 26

Avila, J., 215Axelman, K., 24Azimov, R., 243



Botting, R. M., 76Boudreau, M., 76Bouillot, C., 157Bouman, L., 75Bounhar, Y., 101, 176Bouthillier, D., 24, 172Bowen, D. M., 31, 268Bowser, R., 76Boyd-Kimball, D., 91–92Boyle, J. P., 194Bozdagi, O., 243Braak, E., 23, 210, 268Braak, H., 23, 210, 213, 268Braam, A. W., 81Brachova, L., 29, 81, 101Brady, D. R., 240Brambilla, E., 257Brandt, R., 211Brayden, D., 256Brayne, C., 210Brazier, M. W., 50Breen, K. C., 75Breitner, J. C., 82, 197, 233Breitschopf, H., 240Brendza, R. P., 82Bresjanac, M., 35Breteler, M. M. B., 80Bretillon, L., 155Brewer, D., 175Brewer, J. W., 77Bridges, L. R., 271Brining, S. K., 178Brion, J. P., 172Brion, J. P., 215Brionne, T. C., 258

Brito, M. A., 244Brockhaus, M., 244Broide, R. S., 100Bronfman, F. C., 177, 270Brook, J. D., 50Brooks, D. J., 75Brooks, W. S., 213Brooks-Wilson, A., 155Brophy, P. J., 141Brou, C., 241Brown, D. A., 157Brown, D. R., 140Brown, G. C., 29Brown, J., 154

Brown, M. D., 29Brown, M. S., 190Broytman, O., 33Broze, G. J. Jr., 47Bruce, A. J., 99Bruce, J., 215Bruckner, M. K., 76Bruhl, B., 191Brun, A., 75Brune, D., 257Brunkan, A. L., 242Brunst, E., 122Bryant-Thomas, T. K., 158Bu, G., 156

Buccafusco, J. J., 173Buciak, J., 36Budavari, A., 81

Buee, L., 78, 210Buee-Scherrer, V., 210Bugiani, O., 213Bui, D., 100, 175, 177Buldyrev, I., 255Bullock, R., 257Bunch, C., 80Buniel, M. C., 194Bunke, D., 4, 45Buno, W., 230Bupp, K., 50Burbach, G. J., 270Burdick, D., 27, 119, 172, 194, 233Burgermeister, P., 77Burke, W. J., 229Burkoth, T. S., 122, 139Burns, A., 19Burns, M., 156Bursztajn, S., 242Burt, D. S., 257Burudi, E. M., 231Busciglio, J., 47, 76, 120, 176, 194Bush, A. I., 21–22, 24, 27, 30, 33,

45–46, 48, 92, 98, 138–139, 194–196,230, 258

Bushong, E. A., 230Busser, J., 76Bussière, T., 210, 213, 256Bustos, C., 158Butcher, L. L., 100Butler, S. M., 28Butterfield, D. A., 28, 30, 89–92, 98–99,

120, 195Buttini, M., 257Buxbaum, J. D., 4, 25, 47, 174, 178, 243

Buxbaum, J. N., 22Byrne, E., 29

CCaccamo, A., 177, 214, 240, 243, 269Caceres, A., 176Cacho, J., 138Cadenas, E., 29Cagnin, A., 75Cai, H., 268Cai, J., 90Cai, X. D., 23, 120, 190Cairns, N. J., 28, 211, 213Caldwell, J. N., 195

Calero, M., 81, 196Calhoun, M. E., 177, 270Calhoun, M., 172, 272Callahan, L. M., 74Calon, F., 272Caltagirone, C., 191Camacho, I. E., 82Camakaris, J., 25, 45Camarkis, J., 45Cameron, H. A., 268Campbell, A. P., 122Campbell, I. L., 269Campion, D., 211Camuzat, A., 211

Canevari, L., 244Canevari, L., 99Cannady, S. B., 76

Cannella, B., 73Cannon, C., 33, 78, 196, 232–233, 256Canova, C., 233Cao, D., 157Cao, L., 232Cao, X., 46, 157, 242Capell, A., 191, 241Capizzano, A. A., 34Cappai, R., 22, 25, 45, 48, 50–51Card, J. P., 48Cardoso, S. M., 244Carlson, G., 23Carman, M. D., 21Carnero, C., 212Carney, J. M., 28, 98, 176Caromile, L., 244Carroll, R. T., 232Carson, J. A., 78Carter, D. B., 154, 271Caruso, A., 76Casadei, V. M., 79Casamenti, F., 172, 177Casas, C., 121, 270Case, C. P., 271Casley, C. S., 244Caspersen, C., 244Cassarino, D. S., 29Cassatella, M. A., 29, 120Castano, E. M., 22, 123, 154, 193Castegna, A., 28, 30, 90–92, 98Castellani, R. J., 3Caster, H., 123Castro, G., 155Catalano, R., 272Cato, A. M., 76

Cattabeni, F., 174Catton, M., 36Caughey, B., 213Caumont, A., 100Cavallaro, S., 269Cecal, R., 32, 256Cedazo-Minguez, A., 242Cervilla, J., 154Cervos-Navarro, J., 212Chai, H., 46Chakrabartty, A., 120Chalmers, K., 155Chalovich, E. M., 76, 269Chambaz, J., 25

Champain, D., 213Chan, A. W., 258Chan, A., 231Chan, P. H., 28Chan, S. L., 29, 194, 240, 242, 271Chan, W. Y., 240Chan, Y. G., 233Chandra, S., 157Chaney, M. O., 27, 121, 140Chang, C. Y., 34, 36Chang, J. W., 243Chang, J. Y., 28Chang, K. T., 100Chang, L., 98, 194

Chang, Y., 177Chan-Palay, V., 268Chapman, P. F., 270

280 Author Index



Chapman, P., 26, 77, 172, 190, 213, 269Charlton, B. G., 78Chartier-Harlin, M. C., 21, 120, 154,

190Chauhan, N. B., 197Chavis, J. A., 28Checler, F., 23Chen, A., 29Chen, C. F., 175Chen, D. C. R., 176Chen, D., 174Chen, F., 173, 190, 211, 214–215Chen, G., 270Chen, H. I., 178Chen, H. Y., 75Chen, K. C., 77Chen, K. S., 196, 270Chen, L., 156Chen, Q., 240Chen, S., 3, 172Chen, X., 73, 101, 156, 175, 232Chen, Y., 47Chen, Z. J., 155Cheng, B., 47, 75, 99, 176, 243Cheng, F., 51Cheng, I. H., 273Cheng, Y., 177Cherian, K., 32Cherny, R. A., 20, 30–31, 33, 98,

139–140, 172, 195–196, 272Chernyshev, O. N., 29, 233Cheung, T. T., 23, 120, 190Chevallier, S., 233Chi, C. W., 80Chiarle, R., 123

Chishti, M. A., 121Chishti, M. A., 177, 272Chleboun, J. O., 22Cho, C., 174–175Cho, H. S., 122Chochina, S. V., 193Choi, E. K., 25, 243Choi, S. -Y., 269Choi, Y. H., 243Chong, K. Y., 258Choo, L. -P., 123Choo-Smith, L. P., 141, 157Christen, Y., 81Christie, G., 157

Christie, R. H., 78Christopherson, K. S., 230Chromy, B. A., 27, 98Chu, T., 22, 33, 82, 272Chui, D. H., 214Chung, H., 232Chung, W. C., 240Chyung, A. S., 192Ciallella, J. R., 48Ciccotosto, G. D., 31, 51, 92, 139, 141Cirilli, M., 195, 244Cirrito, J. R., 82, 255Citron, M., 21, 23–24, 229–230Civin, W. H., 81

Clark, A. W., 214Clark, J. B., 244Clark, L. N., 211, 192

Clarke, C. L., 243Clarke, E. E., 33, 82Clarke, S., 121–122, 189Clarris, H. J., 48Cleary, J. P., 140Clee, S. M., 155Clements, A., 122–123, 139Clements, J., 256Clippingdale, A. B., 172Close, D. R., 28Cohen, A. S., 22Cohen, F. E., 140Cohen, M., 230Cohn, J. S., 192Cohn, R., 20Colangelo, V., 77Colciaghi, F., 191Cole, G. M., 19, 35, 47, 176, 197, 229,

231–232, 258Cole, T. B., 30, 139Cole, W. C., 28Coleman, P. D., 91Coleman, R. E., 34Coles, M., 139, 193Collatz, M. B., 212Collin, R. W. J., 50Collins, M. T., 34Colquhoun, L. M., 173Colton, C. A., 29, 233Combarros, O., 79Combrinck, M., 76Combs, C., 76Condorelli, F., 76Condron, M. M., 123–124, 193Confaloni, A. M., 50

Connor, D. J., 197Connor, K. E., 33, 78, 272Contreras, B., 241Conway, K. A., 32Cook, D. G., 157, 190, 241Cooper, M. D., 120Cooper, N. R., 73Copani, A., 241Copani, A., 76Corder, E. H., 21, 74, 154Cordle, A., 158, 272Cordy, J. M., 157Cornett, C. R., 138Corrada, M., 272

Corral-Debrinski, M., 29Correia, K. M., 241Corsa, J. A., 196Costa, P. R., 139Cote, S. L., 177Cotman, C. W., 21, 26, 29, 75, 124, 176,

193, 240Cottel, D., 155Cotter, R. L., 229Coulson, E. J., 46Counts, S. E., 101, 177, 270Court, J., 173Cousin, E., 155Cowburn, R. F., 48

Coyle, J. T., 270–271Crabtree, G. R., 258Craddock, S. D., 48

Craessaerts, K., 25, 241Craik, D. J., 120, 122, 139, 141Cras, P., 24, 75Crawford, A. W., 32Crawford, F., 24, 120, 154, 194Crescenzi, O., 193Crespo, P., 154Cribbs, D. H., 75, 255Crook, R., 25Crouch, P. J., 244Crowther, R. A., 20. 210–212Crutcher, K. A., 26Cruts, M., 21, 211Cruz, L., 273Crystal, H., 213Csajbok, L., 46Cuadros, R., 215Cuajungco, M. P., 27, 30, 92, 99, 121,

140, 194Cukier, R. I., 141Culmsee, C., 242Culotta, V. C., 24Culpan, D., 79Culwell, A. R., 47, 75, 243Cumming, J. N., 196Cummings, B. J., 240Cummings, J. L., 19, 33–34, 36, 172,

192, 229, 256Cunningham, C. C., 211Curb, J. D., 79Curran, M. D., 79Curtain, C. C., 27, 31, 140–141Curtin, C. C., 92Curtis, M. A., 269Cutler, N. R., 32

Cuzin, F., 50Czech, C., 46, 242

DD’Adamio, L., 243D’Adamo, P., 214D’Andrea, M. R., 100, 175–176, 229,

231–232D’Souza, I., 212Daffner, K. R., 19, 31Dago, L., 176Dahlgren, K. N., 123Daigle, I., 46Dal Forno, G., 80

Dalfo, E., 46Dalton, A. J., 27Danbolt, N. C., 230Danton, G. H., 81Dantzer, R., 81Darabie, A. A., 99Dartigues, J. -F., 196, 257, 271Das, C., 32Das, G. D., 268Das, P., 33, 82, 119, 157, 196–197, 256,

271DaSilva, K., 255David, D. C., 215David, D., 196, 210–211, 213

David, J. P., 213Davidsson, P., 191Davies, D. C., 230, 269

Author Index 281



Davies, J. P., 153Davies, K. J., 28–29, 91Davies, M. J., 141Davies, P., 123, 173, 268Davies, R. R., 211Davignon, J., 24Davingnon, J., 172Davis, C. B., 175Davis, D., 99, 176, 243Davis, H. R, Jr., 153Davis, J. B., 30, 232Davis, J., 77, 124Davis, K. L., 82, 173Davis, L. G., 48Davis, P. B., 74Dawes, L. R, 4, 21Dawson, H. N., 26, 215Dawson, K., 210Dayanandan, R., 211De Berardinis, M. A., 76de Figueiredo, R. J., 26De Groot, C. J. A., 74, 78De Jonghe, C., 241de Jonghe, C., 243de la Pompa, J. L., 241De Strooper, B., 4, 25, 32, 46–47, 156,

241De Teresa, R., 90De Vente, J., 158de Waal, R. M. W., 81De, F., 211De, Ferrari, G. V., 194Dean, R. A., 196Dean, R. L. III., 173Dean, R. L., 22

Dean, Y. D., 75Deane, R., 77Debeir, T., 177, 270Debnath, M. L., 34DeCarli, C., 271Deckwerth, T. L., 240Deelen, W., 212Dehghani, F., 230Deibel, M. A., 138DeKosky, S. T., 155DeKosky, S. T., 36, 78–79, 173, 271Del Angel, V. D., 141Delacourte, A., 212–213Delisle, M. B., 211–212

Della-Bianca, V., 176Deller, T., 98, 256DeLong, M. R., 192, 270DeMattos, R. B., 33, 77, 196, 256Demeester, N., 123Demicheli, F., 46Demicheli, V., 141Deng, G., 29, 240Depboylu, C., 255, 258Desai, P., 155Desai, R., 256Desikan, R., 28Desmyther, A., 258DeSouza, R., 242

DeStrooper, B., 272DeTeresa, R., 239Devanand, D. P., 34, 81DeVos, N., 78

Dewachter, F., 256Dewachter, I., 270Dewitt, D. A., 230Dewji, N. N., 241Diamond, D. M., 82, 255, 270DiCarlo, G., 33, 77–78, 256, 272Dickinson-Anson, H., 269Dickson, D. W., 4, 23, 27, 73, 172–173,

192, 211–213, 215, 229, 269Diehl, G. E., 240Diehl, J. A., 77Diehl, T. S., 229Dietrich, W. D., 81Dik, M. G., 79Dillen, L., 121Diltz, C. D., 211Dineley, K. T., 100, 175, 177Ding, W. H., 78Dingwall, C., 157Diprete, C. C., 138Dizdaroglu, M., 28Dobransky, T., 175Dobrowsky, R. T., 19Dobson, C. M., 255Dodart, J. -C., 256–258, 269, 272Dodel, R. C., 4, 27, 77, 79, 255, 258Doh-ura, K., 45Dolezal, V., 174Domenicotti, C., 194Doms, R. W., 22Dong, J., 31, 92, 140Dong, Y., 230Donnelly, R. J., 45Donovan, M. H., 271Donoviel, D. B., 242

Doody, R. S., 36Doody, R., 31Doraiswamy, P. M., 32, 34, 196Dority, M. D., 74, 195Dorsey, R., 76, 269Dowling, M. M., 20Drache, B., 240Drachman, D. A., 154Dragunow, M., 240Drake, J., 28, 30, 89–91, 98Dressner-Pollak, R., 79Drew, P. D., 28Drisaldi, B., 25, 45Drisdel, R. C., 173

Drouet, B., 25Du Yan, S., 101, 232Du, A. T., 34Du, Y., 79, 156, 255, 258Dubas, F., 73Duce, J. A., 244Duchen, M. R., 99, 244Dudal, S., 77Dudek, D. M., 25Duering, M., 242Duff, D., 272Duff, K., 177Duff, K., 23, 26, 77, 98, 270, 273Duffy, L. K., 19–20, 98, 194

Duke, M., 213, 269Dulubova, I., 45Dumanchin, C., 211Dumitrescu-Ozimek, L., 197

Dumoulin, M., 258Dunn, E., 4, 32, 78, 154, 196, 232–233,

255, 271Dupont, J. L., 48Dupuis, F., 141Durán, R., 141Durham, R., 231Duyckaerts, C., 74–75Dynan, K. B., 79Dyrks, T., 24, 46

EEanes, E. D., 21Eastwood, B. J., 79Eastwood, B., 255Ebert, U., 177Ebneth, A., 211Echols, C. L., 120Eckert, A., 100Eckert, G. P., 99, 193Eckman, C., 4Eckman, C. B., 24, 120, 123, 192, 195Eckman, C., 22–24, 77, 98, 273, 190Eckman, E. A., 123, 195Edbauer, D., 210, 241Edgar, P. F., 91Edmonds, B., 243Edwards, D. R., 29Edwards, J. K., 19Efthimiopoulos, S., 46Eggert, S., 46Egnaczyk, G. F., 140Ehehalt, R., 157Ehl, C., 79Ehmann, W. D., 28, 138

Eidenmuller, J., 211Eikelenboom, P., 72–75, 78–82Einstein, G., 34Eisch, A. J., 268–269Ekinci, F. J., 176El Khoury, J., 29, 73, 175, 232El-Agnaf, O. M., 123, 140Ellisman, M. H., 230Elyaman, W., 100Emerich, D. F., 22Emile, J., 73Emilien, G., 31Emmerling, M. R., 98, 154, 191–192Emmons, T. L., 124

Emsley, J., 74Encinas, M., 99, 193, 215Eng, L. F., 230Engelhart, M. J., 79Engert, J. C., 155Engert, S., 174Engler, H., 35Englund, E., 75Erb, C., 177Erickson, J. D., 173Ericksson, P. S., 268Eriksen, J. L., 33, 82, 119, 197, 272Ermak, G., 28Ermini, F., 172, 272

Ernesto, C., 31Ertekin-Taner, N., 25Esch, F. S., 4Esch, F., 22, 47, 177

282 Author Index



Esiri, M. M., 76, 213Esler, W. P., 190, 241Esposito, L. A., 273Esposito, L., 242Esselmann, H., 191Essrich, C., 242Estermyer, J. D., 100Estus, S., 46–47Etcheberrigaray, R., 242–243Etminan, M., 197Eur, J., 193Evans, D. A., 20Evans, R. M., 154Evin, G., 48

FFabian, H., 121, 123Fadale, D. J., 272Fadeeva, J. V., 27, 92, 140, 192, 269Fairlie, D. P., 120, 141Fallah, M., 268Familian, A., 75Fan, Q. W., 99, 193, 195Fang, C., 194Fanger, C. M., 243Farlow, M. R., 32, 174, 196Farlow, M., 120, 154, 269Farmery, M. R., 244Farr, S. A., 92, 175, 257Farrow, J. S., 46Fassbender, K., 33, 73, 156Fasulo, L., 215Faull, R. L., 240Fausett, H. J., 25, 241Fedorchak, K., 26

Feenstra, M. G. P., 79Feil, S. C., 45Felsenstein, K. M., 32Feng, R., 271Feng, Z., 177Fenili, D., 255Fernandez-Madrid, I. J., 21Ferrari, A., 211, 215Ferreira, A., 48Ferreira, R., 123Ferreiro, E., 100Ferrell, R. E., 80Ferrer, I., 257Ferrero, M., 79

Ferri, C., 79–80Ferrier, I. N., 78Fiala, M., 232Fibiger, H. C., 270Fidani, L., 79Fields, R. D., 230Fievet, C., 155Filley, C. M., 29Fillit, H., 78Finch, C. E., 192Finefrock, A. E., 196Fischbeck, K. H., 77Fischer, O., 72Fisher, S., 4, 21

Flaherty, D. B., 76Flanagan, L. A., 211Flanary, B. E., 233Fletcher, T. G., 139

Flood, D. G., 91Flood, F., 242Florey, E. E., 33Florey, K., 33Flory, J., 210Fluit, P., 177Flynn, B. L., 233Fodero, L. R., 100Foley, P., 257Folks, D. G., 229Fonseca, M. I., 120Fonte, V., 34Ford, J. M., 79Forloni, G., 46Forman, H. J., 28Forman, M. S., 77, 190Forrest, G. L., 80Forssell, L. G., 138–139Förster, G., 141Forsyth, P., 29Frackowiak, J., 232Frackowiak, R. S., 34Frances, P. T., 31Franceschi, C., 121Francis, P. T., 173Francis, R., 190Frangione, B., 22, 24, 122–123, 172,

192–193, 196Frank, R. A., 19Franklin, T., 141Fraser, F. W., 20, 98, 172Fraser, H., 81Fraser, P. E., 121, 98, 141, 192Frautschy, S. A., 232Frears, E. R., 123, 156

Frenkel, D., 255–257Friedland, R. P., 3, 36Fritz, L. C., 45Froelich, S., 242Frosch, M. P., 195Frykman, S., 244Fu, J., 73, 101, 175, 232Fu, W., 25, 176, 243Fukuchi, K., 46, 257Fukuda, H., 122, 124Fukumoto, H., 189, 192Fukushima, T., 191Fukutani, Y., 213Fuld, P., 213

Fuller, S. J., 48, 191Funato, H., 189Funke, H., 155Funkenstein, H. H., 20Furlan, R., 257Furon, E., 233Furukawa, K., 48, 99, 176, 242–243Furuya, H., 121Fusi, F., 174Fuson, K. S., 28

GGabbita, S. P., 29, 90Gabuzda, D. H., 47, 120

Gage, F. H., 268–269Galasko, D., 19Galatis, D., 45, 47, 50Galdzicki, Z., 178

Galea, L., 268Galeazzi, L., 121Galindo, M. F., 26Gambhir, S. S., 36Gamblin, T. C., 215Games, D., 26, 77, 172, 190, 213,

269–270Gan, L., 242Gan, X., 232Gandy, S. E., 120Gandy, S., 255, 257Ganguli, M., 36, 79Gannon, K. S., 256, 269Gao, H. M., 258Gao, Y., 157Garber, D. W., 157Garcia-Jimenez, A., 48Gardoni, F., 174Gartner, U., 29, 76Garzon-Rodriguez, W., 139, 157Gasparini, L., 73Gasque, P., 75, 233Gau, J. T., 177Gaudreau, P., 174Gaunt, M. J., 32Gauthier, S., 29Gaynor, K., 156Ge, N., 48Ge, S., 230Gearing, M., 212Geddes, J. W., 28, 75, 77, 99Geerlings, M. I., 79, 81Gehrmann, J., 233Geiger, T., 122Geldmacher, D. S., 76

Gelinas, D. S., 255Gendelman, H. E., 229Genthe, A. M., 21Gentile, M. T., 123Gentleman, S. M., 48, 80George, A. A., 175George, A. J., 157Georgievska, B., 178German, D. C., 177, 268, 270–271Gerner-Beuerle, E., 80Gestwicki, J. E., 258Geula, C., 19, 173, 215, 268, 271Ghadge, G. D., 242Ghayur, T., 243

Ghetti, B., 26, 120, 154, 269Ghilardi, J. R., 35–36, 232Ghirnikar, R. S., 230Ghisdal, P., 100Ghiso, J., 22, 24, 74, 119–120, 122, 172,

177, 192Ghochikyan, A., 255, 257–258Ghosh, R. N., 73Ghoshal, N., 215Giaccone, G., 213Giacobini, E., 91Giambarella, U., 48Giannakopoulos, P., 213, 240Giasson, B. I., 119

Gibson, G. E., 28, 91, 98Gibson, Wood, W., 193Gilbert, D. L., 29, 233Gill, S., 197

Author Index 283



Hayashi, I., 32Hayashi, S., 212Hayes, A., 76, 79He, W., 98, 122Head, E., 120, 140Hecht, M. H., 121Heck, S., 174Hedley-Whyte, E. T., 213, 239Hedreen, J. C., 271Heffernan, D., 48Heldman, E., 174Helkala, E. L., 158Hellman, U., 31, 92, 141, 192Hellstrom-Lindahl, E., 174, 176Hellweg, R., 270Helmuth, L., 35Hemmer, W., 32Hemnani, T., 27, 31–32Hempelman, S. R., 82Hendershot, L. M., 77Henderson, D., 33, 78, 256Henderson, J., 242Hendriks, L., 24Heneka, M. T., 82, 197Henry, A., 47, 51Henryk M., 22Hensley, K., 28, 89, 98–99, 176Hepburn, D. L., 257, 271Herbert, D., 158Herl, L. D., 242Herl, L., 82, 242Hernandez, D., 177, 270Hernandez, F., 215Hernandez-Presa, M. A., 158Herreman, A., 240

Herrmann, F. R., 213Herrup, K., 76, 241Hershey, A. D., 175Hershey, C. O., 138Hershey, L. A., 138Hershkowitz, M., 191Hertel, C., 175Hertz, L., 230Herz, J., 153Herzig, M. C., 77Hesse, C., 191Hesse, L., 30, 45Heun, R., 155Heverin, M., 155

Hickey, G. A., 35Hickman, S. E., 29, 73, 175Higgins, G. A., 4, 23, 192, 268, 270Higgins, L. S., 272Higuchi, D. A., 47Higuchi, M., 214Hilbich, C., 24, 80, 121, 124, 140Hill, J. L., 78Himes, C. S., 32Hinds, T. R., 78Hintz, N., 174Hirai, S., 22, 78Hirakura, Y., 100, 175, 243Hirano, A., 214, 268

Hirashima, N., 243Hla, T., 78Ho, A., 45Ho, L., 46, 76, 123, 192

Hoblyn, J., 173Hochman, A., 241Hock, B. J. Jr., 4Hock, C., 4, 32, 172, 197, 257, 271Hodges, J. R., 210–212Hoerndli, D., 213Hoerndli, F., 213, 215Hoes, A. W., 233Hof, P. R., 172, 213Hoffmann, R., 122Hofman, A., 73, 80, 272Hofmeister, J. J., 140Hogan, D. B., 158Hoglund, K., 191Holback, S., 50Holcomb, L., 270Hollenbach, E., 156Hollister, R., 213Hollosi, M., 255Holmans, P., 25Holmes, C., 4, 78, 172, 196, 257Holness, M. J., 213Holsboer, F., 32Holsinger, R. M., 48, 157Holt, D. P., 35Holton, J., 213Holtsberg, F. W., 29, 242Holtz, G., 196Holtzman, D. M., 74, 154, 194, 255, 272Holzemann, G., 99Holzer, M., 29, 76Homanics, G. E., 153Honda, T., 176Hong, C. S., 244Hong, H. S., 158, 243, 272

Hong, M., 176, 212, 214Hong, Y., 101, 176Hoogendijk, W. J. G., 79, 81–82Hooper, N. M., 47, 120, 153, 157Hoozemans, J. J. M., 74–78, 82Horsburg, H., 75Hortobagyi, T., 20Horton, T., 29Horváth, L. I., 141Hoshi, M., 175Hosley, J. D., 32Hosoda, R., 122, 193Hosokawa, M., 269Hotamisligil, G. S., 81

Hou, C., 34–35Hou, L., 31, 92, 141Houlden, H., 21, 25, 120, 154, 193Houston, M. E. Jr., 122Hovland, A. R., 28Howard, V., 256, 271Howell N., 30Howland, D. S., 157Howlett, D. R., 47Howlett, G., 140Hoyer, S., 81Hsia, A. Y., 27Hsia, A., 270Hsiao, K., 26, 77, 172, 190, 213, 269

Hsieh, H., 194Hsieh, S. J., 176Hsu, A., 28, 90Hu, J., 26

Hu, L., 177Huang, F., 90, 257Huang, G. F., 35Huang, H. M., 176Huang, H. W., 36Huang, S. C., 35Huang, X., 30–31, 92, 121, 138–140,

193, 195Huber, G., 48Huell, M., 73Hug, G. L., 92Hughes, S. R., 156Hui, S., 154Hulshof, S., 78Hultenby, K., 241Hung, A. Y., 22–23, 47, 123, 177, 190,

230Hunt, S. P., 90Huryeva, I., 45Husemann, J., 257Hussain, I., 47, 157Husseman, J. W., 77Hutter-Paier, B., 155Hutton, M., 26, 211, 255Huttunen, H. J., 155Hwang, E. M., 158, 272Hyman, B. T., 24, 34, 80, 156, 195, 210,

213, 239, 255

IIda, N., 190Igbavboa, U., 193Iglesias, M., 215, 256Ihara, Y., 20, 122, 157, 194, 211Ihunwo, A. O., 272

Iijima, K., 46Ikeda, M., 242Ikezu, T., 47Ikonen, E., 153Ikonomovic, M. D., 48, 78, 155, 173,

271Ile, K. E., 155Imai, Y., 156Imaizumi, K., 99Imaki, H., 77, 231In’t Veld, B. A., 73, 233, 272Inestrosa, N. C., 22, 123, 194Infante, J., 79Ingelsson, M., 192

Ingram, D. M., 82Ingram, E., 214Inoue, S., 22Inouye, H., 121Ioannou, Y. A., 153Iqbal, K., 3, 172Irie, K., 124Irizarry, M. C., 26, 270, 272Isacson, O., 19Iserloh, U., 196Ishibashi, S., 153Ishibashi, Y., 190Ishiguro, K., 27, 244, 269Ishiguro, M., 156

Ishihara, T., 214Ishii, K., 192Ishii, T., 73, 81, 232Ishii, Y., 139

Author Index 285



Ishikawa, T., 29Islam, A., 173Isoe, K., 191Isohara, T., 50Issa, A. M., 100, 174Itabashi, S., 80Itagaki, S., 73, 230, 233Itier, J. M., 121Itier, J. -M., 270Ito, E., 242Itoh, A., 173Itoh, K., 120Itoh, Y., 80–81Itokawa, Y., 30Iverfeldt, K., 50, 178Iversen, L. L., 268Ivnik, R. J., 19Iwasaki, K., 242Iwata, N., 195Iwatsubo, T., 122, 189–190, 196Izmirlian, G., 99

JJabbour, W., 73Jacobs, D. M., 34, 191Jacobs, R. E., 270Jacobsen, G., 20Jacobsen, H., 268Jacobsen, T. M., 32Jaeger, S., 50Jaffar, S., 101, 177, 270Jakes, R., 210–211Jakobs, C., 154James, E. R., 155Janaky, T., 121

Janciauskiene, S., 158Jang, J., 243Jankowshy, J. L., 272Jansen, W., 80Janssen, I., 74, 82Jantzen, P. T., 33, 78, 272Janus, C., 33, 121, 177, 214, 255,

271–272Jao, S. -C., 139Jaradat, M. S., 197Jarret, J. T., 98Jarrett, J. T., 4, 123, 193Jarriault, S., 241Jarvet, J., 122

Jarvik, G. P., 154Jassar, B., 174Jay, G., 244Jeffrey, M., 75Jelic, V., 31Jellinger, K. A., 19Jellinger, K., 19Jenkins, E. C., 21Jenkins, S. M., 211Jenner, A., 28Jensen, K. B., 138Jensen, M. S., 138Jensen, M., 4, 22, 98, 190Jeong, S. J., 174

Jhamandas, J. H., 174–176Jhee, S., 32Jhoo, J. H., 80

Ji, S. -R., 141Jiang, C., 197Jiang, H., 233Jicha, G., 76Jick, H., 33, 185, 272Jick, S. S., 33, 158, 272Jin, H., 139Jin, K., 269Jin, L. W., 47, 157Jo, D. G., 243Joachim, C. L., 20, 46Joachim, C., 270Jobst, K. A., 213Johansson, A., 155Johnson, C. J., 34Johnson, G. V., 211Johnson, J. A., 271Johnson, S. A., 74Johnstone, E. C., 81Johnstone, E. M., 121Jones, J., 233Jones, M. W., 270Jones, R. W., 257Jones, Y. Z., 230Jonker, C., 79, 82Joosse, M., 212Jordan, B., 90Jordan, J., 26Jordan-Sciutto, K. L., 76, 269Joseph, F., 257Joseph, J. A., 98Joseph, S. B., 153Joslin, G., 175Jucker, M., 256Jun, L., 139

Jung, Y. K., 243Junn, E., 76

KKabbara, A., 155Kadlcik, V., 92Kaffarnik, H.Kagan, B. L., 100, 243Kaijtar, J., 255Kajdasz, S. T., 35, 78Kajdasz, S., 257Kakimura, J., 195Kakio, A., 99, 157–158, 193Kalaria, R. N., 26, 74, 268

Kalimo, H., 25Kamal, A., 190Kamboh, M. I., 80, 155Kameda, N., 191Kamenetz, F., 194Kametani, F., 243Kamoshima, W., 240Kamphorst, W., 74–75, 211, 240Kanai, M., 191Kanaya, A., 79Kane, K. J., 241Kanekura, K., 100Kanemaru, K., 191Kanfer, J. N., 99

Kang, D. E., 156Kang, I., 31, 92, 141Kang, J., 3, 21, 44, 72, 173

Kanno, T., 175Kanski, J., 91–92Kao, T. C., 177Kapell, D., 26Kar, S., 100, 174Karkos, J., 255Karlsson, E., 173Karni, A., 255Karr, J. W., 140Karran, E. H., 157Kasparova, J., 174Kastelein, J. J., 155Katayama, T., 99Kato, K., 232Katz, M. J., 195Katz, O., 256Katzman, R., 27, 89, 98, 172–173Katzov, H., 155Kaufman, R. J., 77Kaul, M., 29Kaupp, L. J., 140Kaur, C., 233Kawahara, M., 100Kawai, H., 100, 175Kawai, M., 75Kawarabayashi, T., 99, 192Kawarai, T., 212Kawas, C. H., 75, 244Kawas, C., 272Kawashima, S., 177Kawooya, J. K., 124Kay, D. W. K., 20Kayed, R., 243Keelan, J., 99Keil, U., 98, 215

Keim, P. S., 4Keire, D. A., 139Keller, J. N., 28–29, 90–91, 240, 242Keller, P., 156–157Kellman, W., 33, 158, 272Kelly, J. F., 100, 176Kelly, P. H., 177, 270Kempermann, G., 269Kennedy, A. M., 34, 75Kennedy, M. E., 196Kennedy, S. G., 242Kenney, M., 80Kenny, P. T., 122Kepe, V., 35

Kerksiek, A., 155Kerr, M. L., 50Kesslak, J. P., 215, 255Key, B., 48Khachaturian, Z. S., 19Khatri, A., 30Kholodenko, D., 191Khorkova, O., 156Ki, C. S., 80Kidd, M. M., 230Kidd, M., 20, 269Kierstead, M. E., 32, 256Kihiko, M., 195Killick, R., 211

Kim, B. J., 243Kim, D., 193Kim, H. D., 257

286 Author Index



Kim, H. S., 244Kim, J. W., 80Kim, K. R., 34Kim, K. S., 29, 47Kim, K. W., 80Kim, M. J., 243Kim, S. H., 174Kim, S. U., 232Kim, T. W., 155Kim, Y. K., 174Kim, Y. M., 76Kimberly, W. T., 25, 50, 174, 190, 241Kimura, M., 30Kincaid, R. L., 172King, G. D., 32Kinnunen, P. K. J., 99Kins, S., 24Kinsella, G. J., 3Kirby, L. C., 82Kirby, P. A., 257Kirkitadze, M. D., 124, 193Kirkland, S., 158Kirsch, C., 99Kirschner, D. A., 19Kish, S. J., 98Kisilevsky, R., 22Kisters-Woike, B., 124Kitaguchi, N., 45Kitamura, Y., 195, 240Kitazawa, M., 240Kitt, C. A., 50Kivipelto, M., 158Kladetzky, R. G., 154Klaschka, J., 175Klauss, E., 33

Klein, J. B., 90Klein, P. S., 176Klein, W. L., 92, 156, 192Kleinschmidt-DeMasters, B. K., 29Kliewer, S. A., 197Klingner, M., 177Kloczewiak, M. A., 175Klomp, L. W., 24Klucken, J., 119, 155Klunk, W. E., 34–35Kluve-Beckerman, B., 121Klyubin, I., 27, 92, 140, 192, 269Knapp, M., 79Knauer, M. F., 119

Knopman, D. S., 36Knox, J., 270Kobayashi, D. T., 196Kobayashi, T., 212Koch, S., 100Koelsch, G., 4Koenigsknecht, J., 257Kogure, K., 46Kohata, N., 140, 193Kohli, B. M., 46Kohno, R., 91Kohsaka, S., 156Koistinaho, J., 81Koistinaho, M., 81, 258

Kojro, E., 4, 47, 156Kokjohn, T. A., 31, 92, 121, 141Koldamova, R. P., 155

Kolsch, H., 79, 155–156Kondo, A., 121Kondo, T., 192Kong, R., 176Konietzko, U., 4, 32, 197, 257, 271Konig, G., 4, 46König, G., 45Konopka, G., 157Kontush, A., 28Konya, C., 20Koo, E. H., 22, 48, 75Kopan, R., 241Kopec, K. K., 232Koppal, T., 91Kordon, C., 81Kordower, J. H., 100Korf, E. S. C., 91Kornack, D. R., 268Kornecook, T. J., 31, 174Kornilova, A. Y., 32Kosik, K. S., 19–20, 176, 211Kosmoski, J., 4Kosofsky, B., 270Kotiatsos, V. E., 269Kotilinek, L. A., 256, 271Kotti, T., 155Kotula, L., 3Kou, Y. M., 92Koudinov, A. R., 193Koudinov, A., 74Koudinova, N. V.Kounnas, M. Z., 156Kovacs, D. M., 25, 75, 241–242Kovari, E., 213, 240Kovelowski, C. J., 78

Kowalik-Jankowska, T., 121Kowalska, A., 212Kowalska, M. A., 76Kozutsumi, Y., 99Kraal, G., 73Kraepelin, E., 20Krafft, G. A., 192Kraftsik, R., 213Kranenburg, O., 76Krapfenbauer, K., 91Krause, J. E., 175Kreiman, G., 215Kremer, J. J., 99Kreng, V. M., 75

Kreutzberg, G. W., 74, 231, 233Kril, J. J., 213Kristjansson, G. I., 50Kristofikova, Z., 175Kriz, J., 76Kruman, II., 29Krzywkowski, P., 26, 77Ksiezak-Reding, H., 212Kuentzel, S. L., 24Kuhl, S., 73Kuhn, H. G., 269Kuhns, A. J., 233Kukar, T., 272Kukull, W. A., 78, 154

Kumahara, E., 174Kumar, A., 177Kume, H., 25

Kummer, C., 50Kuner, P., 175Kung, H. F., 34–35Kung, M. P., 34–35Kunishita, T., 47Kuo, Y. M., 23, 27, 31, 33, 79, 98, 121,

140–141, 154, 172, 191–192, 272Kuo, Y. P., 175Kurihara, A., 36Kurland, L. T., 214Kurochkin, I. V., 231Kuroda, Y., 100Kuroiwa, M., 22Kurosinski, P., 210, 212Kurt, M. A., 230, 269Kusui, K., 189Kusumoto, Y., 269Kwok, J. B., 25Kydd, R., 91

LLaakso, M. P., 158Lach, B., 258Lacor, P. N., 194Lacy, M., 233Ladror, U. S., 98LaDu, M. J., 26Lafarga, M., 154LaFerla, F. M., 28, 240, 243–244Laffitte, B. A., 153LaFontaine, M. A., 90LaFrancois, J., 33, 157, 197, 257, 272Lagalwar, S., 215Lahiri, D. K., 32, 47, 174, 196Lai, C. C., 174

Lai, H. W., 240Lai, M. T., 241Laird, N. M., 80Laird, N., 156Lajoie, G., 175Lal, R., 141, 243Lala, A., 232Lam, L. C. W., 79Lamanna, J. C., 36Lamb, B. T., 4Lambert, J. C., 25, 156Lambert, M. P., 27, 98Lambris, J. D., 77Lammich, S., 4, 47, 156

Lampert, H. C., 191Lampert, M. P., 80Lampert-Etchells, M., 74Lamperti, E. D., 45Land, J. M., 244Lander, C. J., 173Landreth, G. E., 82Landreth, G., 158, 257, 272Lang, E., 122Langdown, M. L., 213Langui, D., 3Lanke, J., 20Lankiewicz, L., 121Lansbury, P. T. J., 98

Lansbury, P. T. Jr., 4, 22, 27, 120, 123,193Lansbury, P. T., 139, 213

Author Index 287



Lanz, T. A., 32, 271Larson, E. B., 19Larsson, T., 20Lashuel, H. A., 215Laskowitz, D., 26Lasrado, R., 214Lassmann, H., 240Last, A. M., 258Lathe, R., 154Lau, K. F., 50Lau, T. L., 31Lauderback, C. M., 28, 89–90, 98Lauer, D., 195Launer, L. J., 233LaVoie, M. J., 174Law, A., 29Lawlor, B. A., 79Lawlor, P. A., 240Lawlor, P., 240Lazarov, O., 269Le, R., 273Le, W., 176Leapman, R. D., 139LeBlanc, A. C., 47, 75LeBlanc, A., 240, 244LeBlanc, J. F., 190LeDoux, J. E., 214Lee, C. W., 34–35Lee, D. H. S., 175–176, 231Lee, D. S., 46, 100Lee, G., 211Lee, H. G., 33Lee, H., 230Lee, J. H., 26, 244Lee, J. M., 173

Lee, J. P., 244Lee, J. Y., 30, 139Lee, K. U., 80Lee, M. G., 24, 120Lee, M., 269Lee, S. C., 73, 76Lee, S. J., 157Lee, T. F., 175Lee, T., 175Lee, V. M. Y., 119Lee, V. M., 22–23, 35, 48, 77, 121–122,

172, 210Lee, V. M. -Y., 258Lee, Y. L., 230

Lee, Y., 26Lefterov, I. M., 155Legg, J. T., 31, 139, 196Lehmann, J. M., 197Lehre, K. P., 230Leiber, C. M., 98Leissring, M. A., 242–243Leiter, L. M., 78Lemaire, G. H., 173Lemaire, H. G., 3, 21Lemaire, H., 44, 72Lemere, C. A., 23, 33, 72, 196,

256–258, 269Lemieux, N., 80

Lendon, C. L., 26, 211Lerman, M. I., 24, 44, 72, 120Lesley, R., 30, 195Leslie, F. M., 100

Lesort, M., 242Leuba, G., 213Leunissen, J. A. M., 50Leverone, J., 256Levey, A. I., 230LeVine, H. III, 32Levy, B., 153Levy, E., 21Levy, L. M., 230Levy-Lahad, E., 4, 21, 123, 154,

241Lewen, A., 28Lewis, H., 196Lewis, J, 4, 27, 173, 214–215, 269Lewis, P. A., 120Li, A. C., 197Li, C., 46Li, G., 19Li, J., 241Li, L., 120, 157, 230, 258Li, Q. X., 46–47, 191Li, R., 74, 192, 232Li, S. B., 197Li, W. P., 240Li, W., 176Li, X. Q., 79Li, Y. M., 241Liang, C. -L., 271Liang, F., 271Liang, X., 82Liang, Y., 4, 21, 154, 211, 241Liao, P. C., 79Libby, P., 81Licastro, F., 78–80Lichtenberg, B., 210

Lieber, C. M., 27Liem, R. K., 230Lilliehook, C., 243Lim, G. P., 33, 197, 258, 272Lim, T. K., 122Lin, A. H., 77Lin, C., 79, 195, 244, 258Lin, F., 158Lin, H., 141, 243Lin, K. F., 100Lin, L., 19, 178Lin, M. C., 100Lin, W. L., 4, 27, 173Lin, W. -L., 215, 269

Lin, X., 4, 197Linder, R., 28Lindgren, G. H., 20Lindgren, S., 158Lindholm, K., 192Lindquist, K., 79Ling, C., 29, 73Ling, E. A., 233Link, C. D., 34, 89, 98, 195Linker, R. A., 231Lins, R. D., 123Lipp, H. P., 214Lippa, A. S., 22Lippa, C. F., 27, 154, 212, 241

Lipton, R. B., 195Liu, B., 258Liu, D. X., 76Liu, D., 194, 271

Liu, H. C., 79Liu, H., 240, 270Liu, J., 35Liu, K. N., 4, 47Liu, L. Z., 28Liu, L., 231Liu, Q., 100, 175Liu, S. -T., 140Liu, T. Y., 79–80Liu, T., 30Liu, Y., 215Lix, B., 122Liyanage, U., 157Lleo, A., 82, 242Llovera, R., 123, 195Loeloff, R., 47Loffler, C., 50Löffler, J., 48Logeat, F., 241Loike, J. D., 232, 257–258Lomakin, A., 192–193London, E., 157Longo, R. L., 123Lontie, R., 141Loosbrock, N., 256, 271Lopez-Toledano, M. A., 194Lorent, K., 270Lorenzo, A., 19, 27, 176, 194Lorton, D., 232Losic, D., 100Losonczy, K. G., 99Lott, M. T., 29Love, S., 28, 271Lovell, M. A., 28–30, 90, 138, 176Lovestone, S., 19, 23, 154, 210–211

Lowenson, J. D., 121, 189Lowery, D. E., 75Lowery, D., 75Lu, D. C., 157Lu, M., 19Lubec, G., 91Luber-Narod, J., 73Lucassen, P. J., 78, 240Ludwig, M., 155Lue, L. F., 23, 81, 101, 172, 192,

231–232Lund, E. G., 155Lundkvist, J., 243Lundqvist, H., 173

Luo, L. Q., 47Luo, X., 193Luo, Y. X., 47Luo, Y., 4, 25, 243Lustbader, J. W., 195, 244Luth, H. J., 29, 272Lutjohann, D., 155, 158Lyckman, A. W., 50Lynch, T., 46Lynn, B. C., 91–92Lyras, L., 28

MMa, J., 241

Ma, K., 139Ma, S. L., 79Ma, S. Y., 101, 177, 270Maas, T., 211

288 Author Index



Maat Schieman, M. L. C., 81Mace, B., 157Mace, S., 155Macfarlane, S., 26MacGarvey, U., 28MacGibbon, G. A., 240MacGowan, S. H., 79Mackic, J., 192Mackinnon, A., 33, 138, 196MacTavish, D., 176Maeda, N., 153Maes, M., 78Maffei, A., 123Magendantz, M., 50Mager, P. P., 193Maggio, J. E., 36, 255Magnus, T., 231Maher, F., 45, 50Mahil, D. S., 123Mahlapuu, R., 48Maier, M., 258Maiorini, A. F., 32Majocha, R. E., 36Maksymovitch, E. A., 156Malester, B., 33, 77, 154, 197, 272Malfoy, B., 80Mallory, M., 26, 48, 269Maloney, A. J. F., 20, 173Maloney, A. J., 268Maloteaux, J. M., 31Malter, J. S., 33Mammen, A. L., 47Manabe, T., 99Manaye, K. F., 268Mancini, R., 242

Mandelkow, E. M., 210Mandelkow, E., 210Mander, A., 82Manelli, A. M., 123, 193Manfra, D., 32Mann, D. M. A., 268Mann, D. M., 23, 122, 189–190, 196Manthey, D., 174Mantsch, H. H., 121, 123Mantyh, P. W., 255Mao, X. O., 269Maquet, P., 34Marcello, E., 174Marchant, R. E., 31, 92, 141

Marchini, C., 230, 233Marcil, M., 155Marcon, G., 80Marcyniuk, B., 23, 268Margol, L., 75, 256Marin, D. B., 191Marin, D., 173Mark, R. J., 28, 90, 99, 176, 195Markakis, E. A., 268Markesbery, W. R., 28–29, 89–91, 122,

138, 176Maron, R., 33, 196, 256–257Marques, C. A., 98, 215Marques, M. A., 26

Marr, R. A., 258Marsh, D., 141Marshall, H., 25Marshall, J. R., 35

Marten, G. J. M., 50Martin, E. D., 230Martin, L., 230Martin, R. B., 140Martin, T., 138Martinez, J. M., 120Martinez, M., 242Martin-Morris, L., 47Martinovich, C., 258Martin-Ruiz, C., 173Martins, R. N., 22, 30, 48Martone, M. E., 230Maruyama, K., 25, 243Marx, A., 210Marx, F., 78Marzloff, K., 80Masaki, K. H., 99Masino, L., 193Masliah, E., 26–27, 48, 75, 90, 214,

229, 239, 257, 269–270Mason, R. P., 100Masters, C. L., 3, 21–25, 30–31, 44–45,

47–48, 50, 98, 194, 210, 230Masure, S., 75Matera, M. G., 80Mathe, A. A., 78Mathis, C. A., 34–35Mathis, C., 257, 272Matsubara, E., 74, 78, 191, 196Matsudaira, P., 47, 120Matsui, T., 48, 80Matsumoto, A., 120, 175Matsuoka, Y., 77Matsushita, S., 80Matsuura, Y., 48

Matsuzaki, K., 122Matthews, S., 268Mattiace, L. A., 73, 212Mattioli, T., 92Mattson, M. P., 28–29, 31, 47, 75, 90,

98–99, 176, 194–195, 240–241,243–244

Mattson, M., 29Mattsson, A., 178Matz, P., 28Maurer, K., 19Maxfield, F. R., 73, 158, 232May, P. C., 28Mayeux, R., 26, 80, 191

Maynard, C. J., 25, 45, 194Mayor, S., 158Mazziotta, J. C., 22, 34Mazzucchelli, M., 174McBride, O. W., 24, 44, 120McBride, O., 72McClatchey, A. I., 45McClean, S., 256McConlogue, L., 27, 270McCusker, S. M., 79McEwen, B. S., 268McGann, K., 82McGeer, E. G., 73–74, 195, 232–233,

255, 271

McGeer, P. L., 73–74, 195, 230,232–233, 255, 269, 271McGowan, E., 25, 193, 214, 255, 270McGrath, G., 190

McGreal, E. P., 75McKay, R. D., 268McKeel, D. W Jr., 75McKenzie, J. E., 48McKhann, G. M., 211McKinley, D. D., 154McKinney, M., 240, 271McKinstry, W. J., 25, 45McLachlan, C., 27McLachlan, D. C., 26McLaughlin, R., 244McLaurin, J., 32–33, 99, 139, 214,

255–256, 271McLaurin, Jo-A., 141McLean, C. A., 3, 20, 23, 25, 31, 98,

139, 157, 172, 196McLean, P. J., 119McLellan, M. E., 257McMurray, H. F., 232McNamara, M., 26, 270McNeill, T. H., 74McPhee, J., 78McPhie, D. L., 242Mead, T. R., 242Meade, R. P., 48Mecocci, P., 28Meda, L., 29, 120Medeiros, M. S., 120Medina, M., 46Mega, M. S., 22Mehne, P., 80Mehta, N. D., 24Mehta, P. D., 191, 196, 256Mehta, S. P., 191Meijer, J., 79

Meiner, Z., 79Melchor, J. P., 122Mellon, A., 22Mellow, A. M., 78Melnikova, T., 272Melo, J. B., 175Mendoza-Ramirez, J. L., 32Mennicken, F., 27, 176, 195Menzel, H. J., 154Mercer, J. F., 25, 45Merchant, K. M., 32, 271Mercken, M., 121Mertens, C., 123Mesulam, M. M., 173, 271

Mesulam, M., 173Mesulam, M-M., 268, 271Metchnikoff, E., 74Meyers, M. B., 243Miao, J., 77Michaelis, M. L., 19Michikawa, M., 99, 193Mikhailenko, I., 156, 190Mikkelsen, J. D., 176Mikkonen, M., 191Miklossy, J., 23, 192–193Milbrandt, J., 99, 193Miller, B. L., 214Miller, C., 48

Miller, J., 31Miller, L. M., 138Miller, R. J., 26Mills, J., 174

Author Index 289



Milward, E., 48Minster, R. L., 80Minthon, L., 78, 191Miravalle, L., 123Mirjany, M., 76Mirra, S. S., 212Misonou, H., 194Mitake, S., 91Mitchell, A., 191Mittsou, V., 79Miura, T., 140, 193Miyasaka, T., 214Miyawaki, S., 271Mizuno, T., 158, 258Mobius, H. J., 31Moechars, D., 121, 177, 270Mohamed, I., 156Mohanty, S., 173Mohmmad-Abdul, H., 89–91Mohs, R. C., 78, 173Mohs, R., 23, 172, 192Moir, R. D., 21, 26, 30, 45–46,

138–140, 156, 193Mok, S. S., 4, 48, 50–51, 98, 100Molchan, S. E., 78Monaghan, D. T., 75Monning, U., 46Monsonego, A., 255Montaron, M. F., 269Montine, K. S., 28Moore, H., 176Moore, L. B., 197Moosmann, B., 31Morag, M., 81Moreira, P. I., 244

Morelli, L., 123, 195Moreno, A., 244Morgan, D., 82, 255, 270Morgan, K., 212Morgan, T. E., 74Mori, C., 256Mori, F., 174Mori, H., 27, 154, 230Mori, M., 242Morihara, T., 82, 197, 258, 272Morimatsu, M., 22Morimoto, A., 124Morin, P. J., 46Morini, M. C., 78

Morishima-Kawashima, M., 192, 194,211Morishita, R., 232, 242Morita, A., 30Mornon, J-P., 141Morris, H. R., 213Morris, J. C., 23, 74–75, 173Morris, J., 270Mortimer, J. A., 233Mott, R. T., 212Motter, R., 191Motzny, S., 154Mrak, R. E., 79Mucke, L., 26–27, 78, 214, 242, 270

Mudher, A., 23Mufson, E. J., 76, 100, 241Mullan, M., 21–24, 120, 154, 190Muller, K., 243

Muller, U., 24Muller, W. E., 99–100Muller-Spahn, F., 172Mullin, K., 25, 80Multhaup, G., 21, 24–25, 27, 30, 45–46,

121, 139–140Mulugeta, E., 173Muma, N. A., 213Münch, G., 257Munch, G., 33Munireddy, S. K., 256Munoz, D. G., 214Munoz-Montano, J. R., 176Murakami, K., 124Muramatsu, T., 80Murayama, M., 25, 175, 214Murayama, O., 25, 30Murayama, S., 122Murayama, Y., 48Murphy, L. J., 195Murphy, M. B., 158Murphy, M. P., 242, 272Murphy, M., 256Murphy, R. M., 99, 193Murray, D., 81Murrell, J. R., 211–213Murrell, J., 120, 154, 269Murri, L., 26, 211Myers, A., 25Myllykangas, L., 156

NNa, D. L., 80Nacmias, B., 80Nadai, M., 173

Nadal, R. C., 139Nadassy, K., 140Nadeau, P., 240Nagele, R. G., 100, 176, 229–232Nagy, Z., 76, 213Naiki, H., 158Nakagami, H., 242Nakajima, M., 124Nakamura, H., 271Nakamura, T., 191Nakamura, Y., 48Nakano, I., 214Nakata, M., 158Nakaya, Y., 242

Nalund, J., 231Narayanan, S., 140Narindrasorasak, S., 75Naslund, J., 23, 92, 141, 172, 192Nath, A., 194, 271Neal, J. W., 233Nee, L. E., 27Needham, B. E., 51Neill, S., 28Nelissen, B., 156Nelson, O., 271Nerbonne, J. M., 156Nerl, C., 80Nesse, W. H., 240

Nestler, E. J., 268Neumann, M. A., 20, 212Newell, K. L., 192Newman, S. K., 34

Newman, S., 91Newton, A. C., 157Ng, H. K., 156Nguyen, J. T., 121Nguyen, M. D., 76Ni, B. F., 138Ni, B., 156Nichols, N. R., 240Nicoletti, F., 241Nicoll, J. A. R., 78–79, 257Nicoll, J. A., 4, 196Niigawa, H., 48Nilsberth, C., 192Nilsen, S., 26, 77, 172, 190, 213, 269Nilsson, L. N. G., 77Nilsson, M., 230Nishimoto, I., 48, 99Nishimoto, S. I., 158Nishimoto, S., 99, 157, 193Nishimura, I., 240Nishimura, M., 190Nissl, F., 20Nitsch, R. M., 174, 214Nitta, A., 173Nitti, M., 194Nobrega, J., 34Nocera, D. G., 141Nochlin, D., 74, 77Nomata, Y., 244Nonaka, M., 91Nonaka, T., 212Nordberg, A., 173, 176Nordberg, A., 35Norenberg, M. D., 230Norris, F. H., 121

Norris, R. D., 271Notkola, I. L., 154Nottet, H. S., 73Novak, M., 20, 212Nukina, N., 20Nunan, J., 4, 98Nunbhakdi-Craig, V., 211Nunomura, A., 27, 99, 194Nunzi, M. G., 119Nurcombe, V., 45Nyborg, A. C., 33, 157

OO’Banion, M. K., 76

O’Barr, S., 74O’Brien, J., 19O’Connor, K., 24, 190O’dell, M. A., 77O’Hara, B. P., 74O’Malley, M. B., 98O’Meara, E., 19O’Neill, G. J., 80O’Nuallain, B., 257Obbili, A., 28Oberg, K. A., 211Obrenovich, M. E., 30, 98Octave, J. N., 215Odaka, A., 157, 189–190, 192–193, 196

Odani, A., 140Oddo, S., 177, 214–215, 240, 255, 269Odo, S., 77Ogata, H., 192

290 Author Index



Ogata, T., 230, 233Ogawara, M., 122Ogino, K., 121Ohm, T. G., 48, 74Ohtsu, H., 140Oishi, M., 178Ojika, K., 91Oka, A., 74Oka, K., 242Okado, H., 46Okamoto, H., 46Okamoto, T., 48Okutsu, J., 46Olesen, O. F., 176Oliveira, C. R., 100, 175, 244Olm, V., 156Olson, J. M., 25Olson, M. I., 24Olson, S. J., 28Olsson, A., 46Oltersdorf, T., 21, 24, 45, 120Omar, R. A., 29Ongini, E., 73Ono, M., 34Onyewuchi, O., 47Opazo, C., 30, 195Opdenakker, G., 75Oppermann, M., 215Orantes, M., 213Orgogozo, J. -M., 196, 257, 271Orpiszewski, J., 121Orso, E., 155Ortego, M., 158Osaki, Y., 213Oshima, N., 192

Ost, M., 46Ostaszewski, B. L., 25, 174, 190, 241Ota, S., 212Oten, R., 26Ott, A., 80Otte-Holler, I., 81Ottersen, O. P., 230Otto, M., 191Otvos, L. Jr., 121–123, 193Otvos, L., 255Ou, H. C., 176Ouagazzal, A. -M., 270Overman, M. J., 124, 193Owens, A. P., 123

PPachter, J. S., 230Pack-Chung, E., 157Padmanabhan, G., 33Paetau, A., 25Page, K. J., 25, 241Paitel, E., 244Paivio, A., 122Palacino, J., 25, 242Paliga, K., 46Paljug, W. R., 48Pallante, G., 32Pallitto, M. M., 99

Palma, E., 175Palmblad, M., 31, 141Palmer, A. M., 31, 173Palmert, M. R., 46

Palmgren, J., 155Palmiter, R. D., 30, 139Palop, J. J., 273Pals, S. T., 73Panegyres, P. K., 45Pang, Z., 28, 99Pangalos, M. N., 45–46Panickar, K. S., 230Pankiewicz, J., 154Pantel, J., 190Pantelakis, S., 20Panzenboeck, U., 155Paola, D., 194Papadopoulos, M., 47Papadopoulos, R., 48Papassotiropoulos, A., 32, 79–80, 155,

257Pappolla, M. A., 29, 33, 154, 157, 197,

272Paquette, J., 77Paradis, M. D., 50, 139Paradis, M., 21Paradisi, S., 120Pardridge, W. M., 36Paresce, D. M., 73, 232Parge, H. E., 140Parihar, M. S., 27, 31–32Park, I. H., 158, 272Park, L., 48, 75Parker, C. A., 174Parker, I., 243Parker, W. D. Jr., 29Parkin, E. T., 47, 157Parks, J. K., 29, 244Parks, J., 29

Parnetti, L., 78Parpura, V., 230Parsadanian, M., 77Parsons, R., 197Partin, J., 240Parvathy, S., 123Pascual, J., 78, 173Pasinetti, G. M., 74, 76, 82, 233Pasternak, S. H., 241Pastor, E., 212Pastor, P., 212Pastore, A., 193Patel, B. P., 123Patel, S., 213

Patrick, J. W., 173Patrico, D., 82Patterson, P., 81Paul, B. A., 243Payet, N., 155Pearson, A. G., 269Pearson, J., 33, 214, 255, 271Pedersen, W. A., 175Pedrini, S., 80Peel, A. L., 269Pehar, M., 27Peisach, J., 140Pekkanen, J., 154Pekna, M., 230

Pellegrini, S., 26, 211Penke, B., 194, 215Pennanen, L., 214–215Penney, E. B., 269

Penninx, B. W., 79Pereira, C., 100Perez, M., 215Perez-Tur, J., 25, 156, 242Perfilieva, E., 268Perini, G., 176Perkins, A., 154Perkinton, M. S., 50Perl, D. P., 19, 214Perlmutter, L. S., 232Perluigi, M., 91Perreau, V. M., 272Perry, E. K., 22, 268, 100, 173Perry, G., 3, 27–28, 30, 36, 121, 141,

230Perry, R. H., 268, 231Persidsky, Y., 73Perusini, G., 20Peskind, E., 78Pesold, B., 241Peters, J. A., 141Petersen, R. C., 19–20, 36Peterson, D. A., 269Petit, A., 172Petit-Turcotte, C., 192Petre, B. M., 215Petrella, J. R., 34Petri, A., 35Petroze, R., 90Petrushina, I., 257–258Pettegrew, J. W., 34Pettigrew, L. C., 48Pettingell, W. H. Jr., 139Pettingell, W. H., 21, 139Pettit, D. L., 175

Pfeifer, M., 257Pfister, K. K., 75Phelps, M. E., 34Phinney, A. L., 25, 45, 98Phinney, A., 77, 213, 232Phornchirasilp, S., 197Piccardo, P., 81Picciano, M., 77Pickel, V. M., 243Pickering-Brown, S. M., 190, 212Pielen, A., 269Pierrot, N., 100Pietrzik, C. U., 50Pietrzik, C., 196

Piggott, M., 173Pike, C. J., 4, 21, 27, 98, 124, 172, 176,193–194

Pilch, P., 100Pillot, T., 25Pimplikar, S. W., 157Pincon-Raymond, M., 25Piper, S., 120Pirttila, T., 191Pittel, Z., 174Plant, L. D., 194Pocernich, C. B., 90Pocernich, C., 30, 89Podlisny, M. B., 46

Pogocki, D., 92Poirier, J., 24, 26, 172Polinsky, R. J., 21Politi, V., 176

Author Index 291



Pollack, S. J., 196Pollak, Y., 81Pollard, H. B., 30, 100, 243,Pollwein, P., 45, 47Polvikoski, T., 24, 156Pompl, P. N., 240Ponte, P., 45Poon, H. F., 90–92Poorkaj, P., 21, 26, 123, 154, 211–213,

241Pop, V., 272Popescu, B. O., 240, 242Popescu, L. M., 242Porrello, E., 174Postina, R., 4, 47Postuma, R. B., 98Potter, H., 74, 241Poulard-Barthelaix, A., 73Poulet, F. M., 32Pouplard, A., 73Powell, D., 47Power, C., 29Prabhakar, S., 176Prammer, K. V., 121Prasad, K. N., 28Prat, A., 231Pratico, D., 48Premkumar, S., 156Price, D. A., 229Price, D. L., 30, 36, 172, 268–270Price, J. L., 23, 74Prieto, I., 120Prieur, S., 244Prihar, G., 242Prince, M., 154

Pritchard, A., 76, 79Probst, A., 3, 77, 98, 213–214, 232Prochiantz, A., 157Prokai, L., 121Prokop, S., 241Pronk, J. C., 80Prosperi, C., 177Prusiner, S. B., 140Psaty, B. M., 197Ptok, U., 79, 156Pufahl, R. A., 51Puglielli, L., 155, 157Puig, B., 46Purohit, D. P., 172, 213

Pursglove, S. E., 241, 244Pype, S., 121, 241

QQin, C., 256Qin, K., 139Qiu, W. Q., 48, 195, 231Qiu, Z., 195Qu, B., 258Qu, T., 177, 270Quaranta, V., 21Quast, T., 50Querfurth, H. W., 24, 243, 100Quintero, E. M., 271

Quirion, R., 29, 31Quist, A. P., 243Quon, D., 272

RRaadsheer, F. C., 78Rabin, J., 272Rabizadeh, A., 100Rabizadeh, S., 157Racchi, M., 174Racke, M. M., 257, 271Radcliffe, K. A., 100Rae, T. D., 51Rahmati, T., 25Raina, A. K., 241Raine, C. S., 73Rainero, I., 79Rajan, A. S., 100Rakic, P., 268Ramakrishna, N., 24Ramassamy, C., 26Ramirez, M. J., 120Rampon, C., 271Ramsden, M., 194Rangan, S. K., 257Ranganathan, S., 76Rapoport, M., 26, 215Rapoport, S. I., 34Raskind, M. A., 78Rasool, C. G., 45Rassoulzadegan, M., 50Ratnavalli, E., 210Ratovitski, T., 190, 269Ravagna, A., 91Ravindra, C. R., 23Rawson, R. B., 190Rea, T. D., 197Rebeck, G. W., 23–24, 79, 156Redwine, J. M., 270

Reed, D. K., 123, 195Reed, G., 45Reed, J., 124Reed, T., 92Reeves, A. J., 268Reeves, M., 240Refolo, L. M., 24, 33, 154, 157, 197,

272Regnier-Vigouroux, A., 231Regula, J. T., 210, 241Reichenbach, A., 82, 195Reif, B., 140Reiner, P. B., 174Reines, S. A., 82

Reinhard, F. B., 46Reisberg, B., 31Reiter, J. S., 24Reiter, R. J., 27Reitz, A. B., 32Relkin, N., 255Reno, J. M., 36Repa, J. J., 153Reszka, A. A., 211Retallack, R., 80Reyes, R., 25, 45Reynolds, C. H., 19Reynolds, G. P., 268Rhee, S. K., 243

Rhodin, J. A., 192Ribaut-Barassin, C., 48Ricard, S., 155

Richards, J. G., 270Richardson, J. S., 89Richardson, R., 211Richey Harris, P. L., 89Richey, P. L., 28Richter-Landsberg, C., 215Ricote, M., 197Riddell, D. R., 157Riederer, P., 19Rietdorf, J., 157Rigby, S. E. J., 139Ripellino, J. A., 45Ritchie, C. W., 33, 138, 196Ritchie, D., 75Ritchie, K., 210Rivard, M. F., 78Rizzini, C., 212, 213Rizzu, P., 26, 211–212Robakis, N. K., 21, 24, 29, 47Roberson, M. R., 174Roberts, G. W., 48Robertson, J. D., 30, 138Robin, M. A., 244Robino, G., 28Robinson, N., 243Robinson, P. J., 22Robinson, S. R., 23, 33, 121Robinson, W. G., 28Rocchi, A., 26, 211Rockwood, J., 214Rockwood, K., 158Rodel, L., 76Rodems, J. M., 272Rodrigues, C. M., 244Rodriguez-Puertas, R., 173

Roediger, F., 90Rogaev, E. I., 4, 21, 23, 123, 154, 211,241, 154

Rogaeva, E. A., 23, 154, 156, 241Rogawski, M. A., 31Rogers, J. T., 78Rogers, J., 29, 33, 73–74, 78, 82, 174,

231–232Rogers, S. D., 232Roher, A. E., 23, 27, 33, 121, 140, 172,

189, 192, 194Rohn, T. T., 240Rohrig, S., 244Rojas, E., 30, 100, 243

Rojiani, A., 256Rollins, J., 74Romano, D. M., 26Ronbinson, S. M., 196Ronchi, P., 121Ronnback, L., 230Rook, S. L., 211Rooke, K., 214Roos, R. A. C., 81Roperch, J. P., 244Rorsman, B., 20Rosado, M., 76Rosen, D. R., 47Rosenberg, R. N., 258

Rosenkranz, K. M., 139Rosenmann, H., 79Rosenthal, A., 256

292 Author Index



Seydel, U., 154Shaffer, L. M., 74, 195Shank, R. P., 175Shao, H., 139Shao, Z., 175Shaw, C. M., 24Shaw, K. T., 174Shea, S., 26Shea, T. B., 176Shearman, M. S., 123, 194Shelanski, M. L., 230Shen, J., 193Shen, Y., 74, 232Sheng, J. G., 269Shepherd, J. D., 214, 240, 269Shepherd, J., 158Sheridan, S., 35Sherr, C. J., 77Sherrington, R., 4, 21, 23, 154, 211, 241Sheu, K. F., 98Shi, J., 36Shiao, Y. J., 175Shibayama, S., 175Shie, F. S., 157Shimadzu, H., 35Shimazaki, Y., 140Shimizu, T., 122, 124Shimohama, S., 240Shin, R. W., 121Shinkai, Y., 192Shinozaki, K., 25Shioi, J., 45–47Shiojiri, S., 190Shiovitz, T., 32Shiozaki, A., 91

Shiozawa, M., 213Shirahama, T., 22, 74Shiraishi, H., 242Shivers, B. D., 121, 140Shoghi-Jadid, K., 35Shoji, M., 22, 99, 120, 177, 191Shringarpure, R., 91Shuck, M. E., 4, 47Sian, A. K., 123Sicard-Roselli, C., 92Siciliano, G., 26, 211Sidera, C., 197Siegel, G. J., 197Siemers, E., 196

Sigurdsson, E. M., 196, 256Silberman, S., 154Silverman, D. H., 34, 36Silverman, J. M., 191Siman, R., 23Simms, G., 3, 44, 21, 210Simons, A., 25, 45, 51Simons, K., 153, 157Simons, M., 156, 158Simons, M., 33Singaraja, R. R., 155Singer, S. J., 241Singh, V. K., 78Singhrao, S. K., 233

Sinha, S., 4, 47Sirimanne, E. S., 240Sisk, A., 26, 269Sisodia, S. S., 30, 98, 120, 172

Sitar, D. S., 99Sivaneri, M. K., 141Sjogren, H., 20Sjogren, M., 158, 197Sjogren, T., 20Skinner, M., 196Skinner, M., 22Sklansky, D. J., 99Skoch, J., 35Skovronsky, D. M., 22, 34–35, 48Slack, B. E., 174Slavin, M. J., 3Slmon, D. P., 229Slowikowski, S. P. M., 174, 177Slunt, H. H., 50Smale, G., 240Small, D. H., 3–4, 45, 50, 98, 100, 154,

190Small, G. W., 34–35Small, G., 34Smid, L. M., 35Smith, C. B., 268Smith, C. D., 28Smith, C., 255Smith, D. G., 31, 139, 141Smith, G. E., 19–20Smith, I. F., 194Smith, M. A., 3, 28, 30, 89, 215Smith, M. J., 25Smith, R. P., 47Smith, T. E., 82Smith, T. J., 153Smith, T. S., 29, 244Smith, W. K., 268Smith-Swintosky, V. L., 48

Snape, M., 173Snauwaert, J., 214Snell, J., 29Snellinx, A., 272Snow, A. D., 74Snyder, S. W., 73, 98Soares, T. A., 123Sobrido, M. J., 214Sodeyama, N., 80Sokolov, Y., 243Sola, S., 244Solomon, B., 255Song, K. S., 47Song, L., 230

Song, W., 240Sontag, E., 211Sopher, B. L., 48, 99, 242–243Soreghan, B., 119Soria, J. P., 76Soriano, F., 26, 270Sorimachi, K., 139Sorrentino, G., 99Sortino, M. A., 241Soto, C., 22, 123, 193, 255Southwick, P. C., 120Sovic, A., 155Sparks, D. L., 33, 197, 272Speciale, S. G., 177, 270

Sperfeld, A. D., 212Spillantini, M. G., 26, 210–214Spina, M. B., 31Spires, T., 214

Spittaels, K., 214Spittaels, K., 82Spooner, E. T., 256–257Spray, D. C., 230Squinto, S. P., 31Sramek, J. J., 32Srinivasan, A., 240Srivastava, N., 153Srivastava, R. A., 153St George-Hyslop, P. H., 21, 23–24,

26–27, 123, 172Staddon, J. M., 230Stadelmann, C., 240Stadtman, E. R., 90Stahle, L., 155Stalder, M., 77, 98, 213, 232, 256Stam, F. C., 72–73Stamer, K., 211Standen, C. L., 50Stanford, P. M., 213Stanley, L. C., 29, 73Starke-Reed, P. E., 28Stas, L., 25, 156Staufenbiel, M., 256, 270Stein, T. D., 271Steiner, E., 78Steiner, H., 174, 241Steinhilb, M. L., 177Steinman, L., 197Steinmetz, A., 154Stephens, D. J., 156Stern, D. M., 244Stern, S. E., 28Stevens, F. J., 22Stevens, J. C., 36

Stevens-Graham, B., 230Stewart, M., 210Stewart, W. F., 75, 231, 272Stimson, E. R., 35–36, 140Stine, W. B. Jr., 92, 98Stine, W. B., 123, 193Stoffler, A., 31Stokes, G. B., 30Storandt, M., 75Storey, E., 3, 48Strain, J., 24Straus, S., 73Strauss, M., 51Strauss, S., 156

Streffer, J. R., 155, 4, 196–197, 210,257, 271Streit, W. J., 74, 231, 233Strickland, D. K., 24, 156, 195, 244Strittmatter, W. J., 4, 21, 26, 74, 154,

172Strohmeyer, R., 78, 232Strosznajder, J. B., 175Struble, R. G., 268Sturchler-Pierrat, C., 213, 269–270Stutzmann, G. E., 243Stuve, O., 197Styren, S. D., 173, 271Styren, S. D., 73

Su, J. H., 29, 240Su, J., 26Subasinghe, S., 99, 158Subbarao, K. V., 89

294 Author Index



Subramaniam, R., 89–90Sudhof, T. C., 46, 157, 242Sudoh, S., 195Suematsu, N., 81Suemoto, T., 35Suen, K. C., 100Sugaya, K., 177, 240, 270Sugden, M. C., 213Suh, S. W., 138Sui, S. -F., 141Sulkava, R., 24, 154, 156–157Sultana, R., 90–92Summers, W. K., 231Sun, X., 240Sun, Y. X., 78Sun, Y., 155Sundberg, R. J., 140Sung, J. C., 190, 241Suo, Z., 194Surewicz, W. K., 141, 157Suva, D., 23, 192Suzuki, F., 232Suzuki, K., 140, 193Suzuki, M., 35Suzuki, N., 23, 120, 157, 189–190, 192Suzuki, T., 50Suzuki, Y. J., 28Svendsen, C., 90Svensson, M., 231Swaab, D. F., 240Swanson, R. A., 230Swearer, J. M., 241Sweeney, D., 120Swerdlow, R. H., 29, 98, 244Syme, C. D., 139

Syversen, S., 158, 197Szabo, B., 47Szalai, V. A., 140Szekely, C. A., 73, 197Szendrei, G. I., 120Szendrei, G. I., 121Szendrei, G. I., 29, 121–123

TTabaton, M., 119Tabner, B. J., 140Tacnet-Delorme, P., 233Taddei, K., 23, 25, 192Taddeo, M. A., 27

Tago, H., 73Tainer, J. A., 140Takahashi, I., 30Takahashi, S., 30Takahashi, Y., 32, 45, 190Takaki, Y., 123, 195Takao, M., 214Takashima, A., 25, 175–176Takashima, S., 74Takeda, A., 215Takeda, M., 48Takeda, S., 48, 242, 244Takemoto, K., 240Takeuchi, A., 272

Takeuchi, H., 258Takio, K., 122Tall, A. R., 155Talwalker, S., 82

Tamagno, E., 28Tamaoka, A., 190–193Tan, J., 256Tan, R., 22Tan, Z. S., 158Tanaka, J., 271Tanaka, K., 212Tanaka, S., 190Tanemura, K., 30, 214Tang, M. X., 34, 81, 191Tang, N. L. S., 79Tang, Y. P., 271Taniguchi, T., 195Tanii, H., 242Tanzi, R. E., 4, 21, 30, 44–46, 50, 72

172Tanzi, R., 30Tapiola, T., 191Tarnawski, M., 231, 258Tarus, B., 92Tashiro, K., 211Tate, W. P., 24, 190Tatebayashi, Y., 214Taylor, G. M., 80Taylor, J. P., 77Teaktong, T., 173Tedde, A., 80Teesdale, W. J., 30, 138Tekalova, H., 175Tekirian, T. L., 122Telivala, T. P., 138Templeton, L., 256Temussi, P. A., 193Tenkova, T., 154Teplow, D. B., 23, 92, 123–124, 193

Terell, B., 257Terry, A. V., 173Terry, R. D., 20, 27, 30–31, 98, 173,

229, 239Terwel, D., 214Terzi, E., 99Teunissen, C. E., 158Tew, D., 141Thal, D. R., 213, 230Thal, L. J., 99Theesen, K. A., 233Theisler, C., 154Thiele, C., 157Thienhaus, O. J., 29

Thinakaran, G., 23–24, 50, 123Thiry, E., 156Thogersen, H. C., 20, 212Thomas, C. A., 29, 73, 175Thomas, R. G., 31Thomas, T. N., 192Thomas, V. S., 229Thome, J., 257Thompson, A. J., 122Thompson, C. M., 138Thongboonkerd, V., 90–91Thorne, J. E., 73, 197Thornton, P. L., 195Tichelaar, W., 141

Tickler, A. K., 21, 92, 124, 139, 141Tilders, F. J. H., 81Timmerman, W., 172Ting, A. T., 197

Tjernberg, L. O., 244Tohgi, H., 90Tokuda, T., 123Tokushima, Y., 45Tolnay, M., 98, 212, 214Tolson, J., 122Tomaselli, K. J., 31, 194Tomaselli, S., 193Tomasiewicz, H. G., 76Tomic, I., 157Tomidokoro, Y., 27, 269Tominari, Y., 32Tomita, T., 194Tomiyama, T., 154Tomlinson, B. E., 20, 22, 173Tong, L., 195Tonini, R., 175Toomre, D., 157Toro, R., 176Toth, L., 153Tournoy, J., 272Town, T., 241, 256Townsend, K. P., 82Toyoshima, Y., 212Tozaki, H., 175Trapp, B. D., 47Treiber, C., 51Tremp, G., 215, 242Tresini, M., 28Trimmer, P. A., 29, 244Trinh, N. H., 173Trojanowski, J. Q., 23, 77, 119, 122,

192, 210, 212Troncoso, J. C., 244Trujillo, M., 141

Trusko, S. P., 157Tsai, J., 98, 121Tsai, S. J., 79Tsaiu, J., 231Tschopp, C., 243Tseng, B. P., 240Tsien, R. Y., 157Tsolaki, M., 155Tsubuki, S., 123, 195Tsuda, T., 26Tsugu, Y., 91Tsui-Pierchala, B. A., 99, 193Tsuji, S., 258Tsuji, T., 91

Tsukamoto, E., 100Tsunozaki, M., 190Tucker, H. M., 195Tuominen, E. K. J., 99Turkenich, R., 255Turnbull, S., 140Turner, A. J., 47, 78, 120, 157, 195Turner, P. R., 24, 190Turner, R. S., 32, 192

UUbeda, O., 82Uetsuki, T., 240Ugen, K. E., 233

Ugolini, G., 215Ugoni, A., 82Ulery, P. G., 156, 190Ullian, E. M., 230

Author Index 295



White, P., 268Whitehouse, P. J., 172, 268Whittaker, J. W., 141Whittemore, S. R., 233Wie, M. B., 176Wiederhold, K. H., 214Wiederhold, K., 172Wiederhold, K. -H., 270Wijsman, E. M., 21, 24, 154Wijsman, E., 26, 211Wiklund, O., 191Wilcock, D. M., 33, 233, 256Wilcock, D., 78Wilcock, G. K., 28Wilcox, M. A., 80Wilcox, M., 156Wild, K. D., 175Wild-Bode, C., 191Wilhelmsson, U., 230Wilkinson, D., 4, 78, 196, 257Williams, A. E., 73, 75Williams, C. H., 122–123Williams, J., 76Williams, T. D., 141Williamson, P. T. F., 139Williamson, T. G., 51Willoughby, D. A., 74Willson, T. M., 197Wilquet, V., 4Wilson, A., 34Winbald, B., 31, 173Windelspecht, M., 21Winkler, D. T., 77Winkler, E., 210, 241Wischik, C. M., 20, 210, 212

Wisniewska, K., 121Wisniewski, H. M., 3, 21, 24, 229,231–232

Wisniewski, K. E., 27, 229Wisniewski, T., 24, 122, 154, 172, 196Witker, D. S., 46Wodak, S. J., 140Wolf, B. A., 195Wolfe, G., 24Wolfe, M S., 25, 32, 241Wolfer, D., 214Wollmer, M. A., 155, 215Wolozin, B., 25, 33, 157–158, 197, 242,

272

Wong, B. S., 140Wong, C. W., 3, 20–21, 44, 72, 120,189, 210

Wong, G. T., 32Wong, P. C., 268Wong, S. S., 27, 98Wong, T. P., 177, 270Wongsud, B., 197Woolley, C. S., 268Wrigley, J. D., 33, 82 123Wszolek, Z., 211Wu, C. K., 19, 173, 215Wu, J., 175Wu, Q., 76, 120

Wu, S., 4, 194, 256Wu, Y., 141Wujek, J. R., 77

Wurth, C., 121Wurtman, R. J., 174, 178Wyss-Coray, T., 77–78, 257–258

XXia, W., 23, 25, 32, 241Xia, X., 177Xiang, Z., 76, 240Xie, C., 90, 155Xie, J., 25Xie, W. J., 176Xie, Y., 46Xilinas, M. E., 33, 139, 196, 272Xu, F., 77Xu, H., 32Xu, M., 241Xu, Y., 193Xue, R., 119Xuereb, J. H., 211Xuereb, J., 211

YYaar, M., 100Yabg, A. J., 256Yaffe, K., 79, 158Yakel, J. L., 175Yamada, M., 80–81Yamada, T., 121Yamagata, S. K., 120Yamaguchi, H., 22–23, 244Yamaguchi, Y., 177Yamakawa-Kobayashi, K., 80Yamamoto, K., 29Yamamoto, N., 122Yamamoto, T., 268

Yamane, T., 242Yamanouchi, H., 191Yamasaki, T. R., 243Yamatsuji, T., 48Yamazaki, K., 90Yamazaki, T., 191, 244Yan, B., 29Yan, F., 77, 258Yan, Q., 270Yan, R., 4, 47Yan, S. D., 101Yan, S. D., 73, 92, 175, 232, 244Yanagisawa, K., 157–158, 193–194Yang, A. J., 75

Yang, D. S., 139, 154, 177, 272Yang, F., 240, 272Yang, F., 33Yang, J., 36Yang, L. B., 192Yang, S., 47Yang, Y. H., 50Yang, Y., 76, 241Yankner, B. A., 4, 47, 19, 21, 27,

119–120, 172, 194Yao, J., 155, 243–244Yao, Z., 46Yarowsky, P., 34Yasojima, K., 195

Yasuda, A., 122Yasutake, K., 176Yates, P. O., 23, 268

Yatin, S. M., 89–91, 195Yatin, S., 90–91Yatsimirsky, A. K., 139Yazdani, U., 177, 270–271Ye, C. P., 27, 98, 192, 269Ye, J., 190Ye, Z., 195, 231Yeh, J., 176Yemul, S., 240Yermakova, A. V., 74, 76Yew, D. T., 194, 240Yip, C. M., 99Yirmiya, R., 81–82Yong, V. W., 29Yong, Y., 139Yoon, I. S., 50Yoshida, H., 213Yoshiike, Y., 30Yoshikai, S., 45Yoshimura, M., 189, 192Yoshizawa, T., 80Younkin, L. H., 25, 47, 99, 192Younkin, L., 256Younkin, S. G., 23, 120, 230, 46Youssef, S., 197Yu, G. Q., 27, 214, 242, 270Yu, G., 190Yu, W., 195Yuan, M., 19Yuasa, K., 242, 244

ZZacharias, D. A., 157Zagorski, M. G., 31, 120, 122, 139, 255Zambrzycka, A., 175

Zandi, P. P., 73, 82, 197, 233Zantema, A., 76Zarandi, M., 194Zarski, R., 74Zech, L. A., 154Zehr, C., 23, 77, 273Zeidler, M., 81Zelasko, D. A., 215Zemlan, F. P., 29Zeng, H., 121, 141Zentgraf, H., 258Zetterberg, H., 155Zhai, S., 100Zhan, S. S., 75, 78

Zhang, B., 214Zhang, B., 35Zhang, C., 80Zhang, G., 258Zhang, J. W., 79Zhang, J., 34, 190, 256, 270Zhang, L., 194, 232Zhang, S. H., 153Zhang, Y., 76, 101, 176, 244Zhang, Z. X., 79Zhang, Z., 240Zhao, B., 194Zhao, H., 99Zharikova, A. D., 121

Zheng, H., 50Zheng, J. B., 50, 241Zheng, W. H., 27, 176–177, 195

Author Index 297

abeta peptide and alzheimer's disease - c. barrow, d. small (springer, 2007) ww

Documents