dominant phenotypes produced by the hdmutation in ......hdh+ (+/+), sthdhq111/hdh+...

© 2000 Oxford University Press Human Molecular Genetics, 2000, Vol. 9, No. 19 2799–2809

ARTICLE

Dominant phenotypes produced by the HD mutation inSTHdhQ111 striatal cellsFlavia Trettel, Dorotea Rigamonti1, Paige Hilditch-Maguire, Vanessa C. Wheeler,Alan H. Sharp2, Francesca Persichetti, Elena Cattaneo1 and Marcy E. MacDonald+

Molecular Neurogenetics Unit, Massachusetts General Hospital, Building 149, 13th Street, Charlestown, MA 02129,USA, 1Institute of Pharmacological Sciences, University of Milano, 20133 Milano, Italy and 2Department of Psychiatry,Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA

Received 7 July 2000; Revised and Accepted 25 September 2000

Lengthening a glutamine tract in huntingtin confers a dominant attribute that initiates degeneration of striatalneurons in Huntington’s disease (HD). To identify pathways that are candidates for the mutant protein’sabnormal function, we compared striatal cell lines established from wild-type and HdhQ111 knock-in embryos.Alternate versions of full-length huntingtin, distinguished by epitope accessibility, were localized to differentsets of nuclear and perinuclear organelles involved in RNA biogenesis and membrane trafficking. However,mutant STHdhQ111 cells also exhibited additional forms of the full-length mutant protein and displayed domi-nant phenotypes that did not mirror phenotypes caused by either huntingtin deficiency or excess. Thesephenotypes indicate a disruption of striatal cell homeostasis by the mutant protein, via a mechanism that isseparate from its normal activity. They also support specific stress pathways, including elevated p53, endo-plasmic reticulum stress response and hypoxia, as potential players in HD.

INTRODUCTION

Huntington’s disease (HD) is a dominantly inherited disorderthat is characterized by the progressive loss of neurons in thestriatum (1). The HD mutation is an expanded CAG repeattract that lengthens a polyglutamine segment in huntingtin to∼37 or more residues (2). This novel protein is required fornormal nuclear organelles and perinuclear membrane compart-ments (3) and has been implicated in iron homeostasis (3),RNA biogenesis (3–7) and membrane trafficking (3,4,8–11).

The polyglutamine defect initiates the pathogenic processvia a mechanism that conforms to genetic criteria that havebeen determined by genotype–phenotype studies in HDpatients (12,13). These critical parameters include specificityfor striatal neurons, increased severity with polyglutaminelength and dominance over the wild-type protein.

Two mutant huntingtin phenotypes that fulfill these HDgenetic criteria have been identified in model systems, stronglyimplicating the underlying process in each case in the humandisorder. The promotion of mutant amino-terminal fragmentinto amyloid has implicated an abnormal property of theexpanded polyglutamine tract (14–16). This property may,however, act either from the full-length mutant protein or, aftercleavage, via an amino-terminal fragment (17,18). The otherphenotype is the nuclear ‘mis-localization’ of the full-lengthmutant protein in HdhQ92 and HdhQ111 striatal neurons, whichprecedes the later formation of truncated product (19). This

phenotype, produced by insertion of the HD mutation into themouse HD locus, has revealed alternate forms of full-lengthhuntingtin, distinguished by epitope accessibility, and has impli-cated a mechanism that involves the full-length mutant protein(19).

Moreover, the dominant phenotypes produced by the mutantprotein in HdhQ111 striatal neurons may provide clues to thedisrupted pathway that elicits eventual toxicity. One possibilityis a perturbation that ensues from increases or decreases in anessential huntingtin activity, predicting a phenotype thatentails abnormal organelles that require the protein’s function(3). This is because huntingtin excess and deficiency bothproduce a similar impact on the protein’s normal cellularpathway (3). Alternatively, the critical disruption may be elicitedindependently of any coincident change in the mutant protein’snormal activity, predicting phenotypes in homozygous andheterozygous mutant striatal cells that do not resemble alteredhuntingtin function.

Consequently, we have sought clues to the mutant protein’sabnormal pathway by assessing the dominant phenotypesexhibited by striatal cell lines established from mutant HdhQ111

knock-in compared with wild-type embryos. Our findingsdemonstrate alternate forms of huntingtin with properties thatsupport roles in the response to hypoxia, in RNA biogenesisand in membrane trafficking. However, mutant cells displaydominant phenotypes that imply toxicity due to a novel property

+To whom correspondence should be addressed: Tel: +1 617 726 5089; Fax: +1 617 726 5735; Email: [email protected]

2800 Human Molecular Genetics, 2000, Vol. 9, No. 19

of the mutant protein that elicits stress-response pathways to beexplored in HD pathogenesis.

RESULTS

A set of clonal striatal cell lines from HdhQ111 and wild-typelittermates

To identify candidates for the mutant protein’s abnormal path-ways in striatal cells, we established neuronal progenitor celllines from E14 striatal primordia of HdhQ111 knock-in (19) andwild-type littermate embryos using tsA58 SV40 large T antigen(20). Multiple clonal lines of nestin-positive cells from the wild-type (STHdh+/Hdh+) and the heterozygous (STHdhQ111/Hdh+)mutant embryos were established. In contrast, only two clonesfrom the mutant homozygote embryos were successfully‘passaged’. These STHdhQ111/HdhQ111 clones exhibited longerdoubling times at 33°C than did the wild-type clones (35 versus26 h), suggesting that the mutant protein may impair theestablishment pathway. However, heterozygous andhomozygous mutant cells, like their wild-type counterparts, wereconditionally ‘immortal’, as growth at 39°C resulted in the loss ofSV40 tsA58 protein and the cessation of proliferation (data notshown).

To test the differentiative capacity of the striatal cell lines, thegrowth medium was supplemented with a ‘dopamine cocktail’(20,21). As shown in Figure 1a for a wild-type clone, all of thecells in the culture were MAP-2-positive and elaborated longprocesses. These characteristics of post-mitotic neuronal-likecells confirmed that the cell lines in each case were establishedfrom neuronal striatal progenitor cells. Notably, genotype-specific differences were not found, indicating that the mutantprotein does not overtly affect the cellular pathways that mediateneuronal differentiation.

Normal and mutant huntingtin are expressed in thecultured striatal cells

Normal and mutant huntingtin were assessed by immunoblotanalyses of wild-type and mutant striatal cell lysates with thereagent monoclonal antibody (mAb) 2166. The results,presented in Figure 1b, revealed the appropriate genotype-specific bands of the endogenous normal and/or mutant proteinwith 111 glutamines, which migrates more slowly. Thesebands were detected in the supernatant, rather than the pelletfraction, revealing similar partitioning of the normal and themutant protein. However, the mutant band was always lessintense than the normal band in the heterozygote lysates,although this ratio varied between experiments, indicating anabnormal property of the mutant protein that has been notedpreviously in HD patient cells (22) and Hdh knock-in brain(19).

To track the long polyglutamine segment of the mutantprotein, the blot was probed with the polyglutamine mAb 1F8(Fig. 1b), which revealed only a full-length mutant band in thesoluble fraction. In this format, mAb 1F8 did not detect theshort glutamine tract of the normal protein as expected (15,23).In some experiments smears of 1F8 reactivity at approximatelythe mobility of the full-length mutant protein were detected,but reproducible low molecular weight bands were not evident(data not shown). Consistent with the unaltered half-life of the

mutant protein in HD patient lymphoblastoid cells (22), theseresults do not support greatly altered turnover of the mutantprotein. Rather they suggest that the long polyglutamine tractmay confer differential ‘extractability’ or ‘detectability’ in theimmunoblot format, although the basis of this abnormalbehavior of the mutant protein remains unexplained.

Alternate forms of huntingtin in the nucleus and in thecytoplasm

The distribution of the full-length normal protein in striatalcells was investigated by staining the STHdh+/Hdh+ progenitorcells with a panel of five specific antibody reagents that detectamino-terminal (AP194, AP229, EM48), downstream (HP1)and internal (HF1) epitopes. As shown in Figure 2, theconfocal images revealed overlapping yet reagent-specificpatterns of reactivity that varied from cell to cell. A comparisonof these patterns revealed alternate forms of the protein. Anamino-terminal-accessible form that was detected by AP194,AP229 and EM48, but with inaccessible HP1 and HF1 internalepitopes, was prominent in the nucleus and also in cytoplasmicdots. In contrast, HP1 and HF1 revealed an alternate form ofthe protein with internal-accessible epitopes, but with amino-

Figure 1. Striatal cell lines expressing endogenous normal and mutant protein.(a) Confocal images of wild-type STHdh+/Hdh+ cells stained for the neuronalprogenitor marker nestin (left). Differentiated neuron-like cells stained forMAP2, a marker of maturing post-mitotic neurons (right). Progenitors areMAP2 negative and differentiated cells are nestin negative (data not shown).(b) Immunoblot analysis of soluble (s) and pellet (p) fractions from STHdh+/Hdh+ (+/+), STHdhQ111/Hdh+ (111/+) and STHdhQ111/Hdh111 (111/111) cells.(Top) mAb 2166 detects normal (Q7) and less intensely mutant huntingtin(Q111). (Bottom) The blot probed with mAb 1F8, detecting the long glutaminetract of the mutant protein. Faint ‘smudges’ are gel artifacts. Equal protein isloaded in each lane.

Human Molecular Genetics, 2000, Vol. 9, No. 19 2801

terminal-inaccessible epitopes. This form was distributed tothe nucleolus and in the cytoplasm was found around thenucleus and in disperse dots. The distinct locations suggestedthat each version of the protein could associate with a differentset of dynamic organelles.

To determine whether post-mitotic neuron-like cells exhibitthe alternate versions of huntingtin, differentiated STHdh+/Hdh+ cells were stained with the antibody panel. The resultsindicated a similar subcellular distribution of reagent epitopesas was found for the striatal progenitor cells (Fig. 2). However,the rounded nuclei of the differentiated cells were denselystained and the amino-terminal-accessible version of theprotein, for example detected by AP194, was more evident inthe cytoplasm of the long projections.

Alternate versions of huntingtin colocalize with differentsets of organelles

To determine the locations of the alternate versions ofhuntingtin, the wild-type striatal progenitor cells werecostained for organelle-marker proteins. The results for theamino-terminal-accessible version, detected with AP194 orAP229 and markers of nuclear organelle, are shown in Figure3A. This form of the protein exhibited extensive overlap withboth ANA125-positive pre-mRNA splicing-speckles and theNuMA-reactive nuclear matrix. In contrast, the alternateinternal-accessible version, detected with HF1, colocalizedwith fibrillarin-positive nucleoli, although it also co-stainedwith some ANA125 splicing-speckles (data not shown). Asshown in Figure 3, co-staining revealed that this version of theprotein was also localized in the vicinity of Concanavalin A(ConA) stain in the rough endoplasmic reticulum (ER) andwith the GM130-reactive Golgi membrane. In some cells, theHF1 (or HP1) signals also displayed partial overlap with trans-ferrin receptor-positive perinuclear foci, although the HF1- (orHP1-) reactive dots in the cellular processes were distinct fromtransferrin receptor-reactive vesicles. Thus, amino-terminal-

accessible protein was located to splicing-speckles in thenucleus and the cytoplasm. In contrast, the alternate internal-accessible form was nucleolar and in the cytoplasm was foundprimarily in the proximity of membrane organelles near thenucleus.

Mislocalization of the mutant protein

To determine whether the polyglutamine expansion was asso-ciated with altered subcellular distribution of the mutantprotein, homozygous mutant STHdhQ111/HdhQ111 cells werestained with the antibody panel. The results for both theprogenitor cells and the differentiated neuronal-like cells areshown in Figure 4A. In general, the pattern of amino-terminaland internal reagent signals was similar to that observed in thecorresponding wild-type cells, indicating that the mutantprotein also exhibits alternate amino-terminal-accessible(internal-inaccessible) and internal-accessible (amino-terminal-inaccessible) forms. Furthermore, co-staining experi-ments revealed that each version of the mutant protein wasdetected in the same set of nuclear and cytoplasmic organellesas its normal counterpart in wild-type cells (data not shown).

However, the staining patterns also revealed notable differ-ences that are evident from the images in Figure 4A. Theamino-terminal-accessible form, detected by AP194 andAP229, was unevenly distributed in the nucleus and was notdetected in the processes of differentiated cells. Furthermore,the internal-accessible HF1- (and HP1-) reactive form of themutant protein exhibited abnormal nuclear and perinuclearpatches. These findings strongly suggested that a proportion ofeach version of the full-length mutant protein was ‘mislocal-ized’ due to the lengthened polyglutamine tract.

Nuclear location entails 1F8-negative polyglutaminesegments

Therefore, we assessed the polyglutamine tract of the full-length protein by co-staining the striatal cells with HF1 and the

Figure 2. Alternate versions of full-length normal huntingtin. Confocal images of STHdh+/Hdh+ progenitors (a–d) and differentiated neuron-like cells (e–h)stained with a panel of specific huntingtin reagents demonstrating alternate patterns distinguished by reagents that detect amino-terminal (a and e) AP194 (aminoacids 1-17); (b and c) EM48 (amino acids 1–256) and (f) AP229 (amino acids 55–66); or internal (d and h) HF-1 (amino acids 1981–2580) and (g) HP-1 (aminoacids 80–113) epitopes. (Inset) MAP2 staining revealing long neuron-like processes.


polyglutamine mAb 1F8 (Fig. 4B). For wild-type cells, thebright 1F8 signal colocalized perfectly with the perinuclearHF1 stain, indicating that the short glutamine tract of thenormal protein exhibited detectable 1F8 epitope in this format.In contrast, in the nucleus of the same cells the HF1-reactivenucleoli and speckles were 1F8 negative, revealing adifference in the polyglutamine tract. Furthermore, mAb 1F8also failed to detect the speckle locations of the amino-terminal-accessible form of the protein prominent in thenucleus. Thus, 1F8 reveals a different property of the protein’spolyglutamine segment in the cytoplasm versus the nucleus,implying compartment-specific structural properties, modifi-cations or interactors.

Abnormal forms of the mutant protein

Homozygous mutant STHdhQ111/HdhQ111 cells were co-stainedwith HF1 and mAb 1F8 to assess the expanded polyglutaminesegment of the mutant protein (Fig. 4B). In general, these cellsexhibited a pattern of HF1–1F8 colocalization similar to thatfound in the wild-type cells. The cytoplasmic versions of themutant protein displayed 1F8-reactive expanded poly-glutamine tracts, whereas the nuclear forms exhibited 1F8-negative polyglutamine segments, implying the same nuclear/cytoplasmic difference displayed by the shorter tract in thenormal protein.

However, a proportion of the HF1-reactive protein in andaround the nucleus exhibited reciprocally ‘switched’ 1F8reactivities. For example, 1F8-reactive dots were embedded inthe generally 1F8-negative HF1-stained nucleoli and large1F8-negative sectors were found within the perinuclear HF1signal, which typically was 1F8 reactive. Moreover, 1F8-reactive‘worms’ in the nucleus suggested conglomerated splicing-speckles with abnormal expanded polyglutamine tracts. Thus,some proportion of the mutant protein in the nucleus and in thecytoplasm exhibited abnormal polyglutamine segments that mayreflect aberrant modifications, binding partners or conforma-tions.

Dominant phenotypes disclose a ‘stress’ response tomutant huntingtin

Although the growth properties were obviously different, wecompared the homozygous mutant cells and the wild-typestriatal cells to identify potential dominant phenotypes thatmay involve the protein’s normal function. This surveyincluded the perinuclear membrane organelles and the ironpathway, which require huntingtin (3). The abnormal pheno-types that were specific to the mutant striatal cells can begrouped into three classes: (i) an altered establishmentpathway due to elevated p53 levels; (ii) an enlarged ERcompartment; and (iii) heightened activity of the iron pathway.Assessment of these phenotypes in heterozygous STHdhQ111/Hdh+ striatal cell lines revealed that the mechanism was domi-nant over the wild-type protein, although the phenotype inmost cases was less severe than in the homozygous mutantcells. This finding may indicate a proposed protective effect ofthe wild-type protein (24) or may reflect the incomplete satura-tion of the mechanism by a single dose’s worth of the mutantprotein.

Figure 3. Distinct version-specific subsets of nuclear and cytoplasmicorganelles. (A) Confocal images of STHdh+/Hdh+ cells co-stained to reveallocalization of alternate amino-terminal-accessible and internal-accessible ver-sions of full-length huntingtin to discrete nuclear organelles: (a) AP229 (red)speckles partially colocalize with (b) splicing factor ANA125 (green) reactivespeckles in (c) the merged image. AP194 yields similar results (data notshown). (d) AP229 epitopes (red) partially colocalize with (e) nuclear matrixprotein NuMA (green) shown in (f) the merged image. (g) HF1-reactivenuclear clusters (red) overlap with (h) nucleolar fibrillarin (green), shown in(i) the merged image. (B) Confocal images of STHdh+/Hdh+ cells co-stainedto demonstrate colocalization of internal-accessible normal protein with peri-nuclear organelles. (a, d and g) HF1-reactive huntingtin (red) in the perinuclearregion that partially colocalizes with (b) ConA marking the ER (green), (e) theGolgi marker GM130 (green) and (h) perinuclear transferrin receptor (green)in (i) the perinuclear region but not (inset) with transferrin receptor puncta inthe cellular processes. (c, f and i) Merged images.


Elevated levels of p53 protein

Consistent with the longer generation times of the mutant cells,the results of propidium iodide flow cytometry (FACS)(Fig. 5a) revealed that STHdhQ111/HdhQ111 and STHdhQ111/Hdh+ cells exhibited twice the DNA content of STHdh+/Hdh+

cells, although the size of the cells was similar for allgenotypes. Increased ploidy is a hallmark of a blocked cellular

p53–SV40 large T antigen establishment pathway (25),prompting an assessment of levels of these proteins byimmunoblot analyses (Fig. 5b). Densitometry of the bandintensities detected in these experiments demonstrated that,when normalized to spectrin signal, the homozygous mutantextract exhibited ∼2-fold more SV40 tsA58 but >6-fold morep53 than the wild-type extract. As elevated p53 protein must be

Figure 4. Abnormal forms of the full-length mutant protein. (A) Confocal images of STHdhQ111/HdhQ111 (a–d) progenitor and (e–h) differentiated neuron-like cellsstained with (a and e) AP194 or (b and f) AP229 demonstrated nuclear speckles but (e and f) absence of the amino-terminal-accessible protein in the processes,indicating ‘mislocalization’. (c and g) HP1 detected amino-terminal-accessible protein in cytoplasmic puncta and nuclear speckles whereas (d and h) internal-accessible protein detected by HF1 was in the cytoplasm, perinucleus and in nucleoli. (B) Confocal images of (a–c) STHdh+/Hdh+ and (d–f) STHdhQ111/HdhQ111

cells co-stained with HF1 (red) and mAb 1F8 (green). In wild-type cells (a) HF1 reactivity and (b) bright 1F8 signal colocalize (c) in the cytoplasm. But HF1-reactive nucleoli were 1F8 negative, indicating ‘masked’ 1F8 epitope of this version of the protein in the nucleus. In mutant cells, the pattern was similar (d–f)except that (d) portions of the perinuclear HF1 reactivity were (e) 1F8 negative and in the nucleus HF1 reactive nucleoli were co-stained with 1F8 (f) revealingabnormal forms of the mutant protein.


overcome by high levels of SV40 large T antigen for ‘immor-talization’ (26), this finding is consistent with our inability toestablish more than two lines of STHdhQ111/HdhQ111 cells fromthe primordia of homozygous mutant embryos. Therefore,increased p53 levels in the mutant striatal cells may reflect anaugmentation of stress pathways that are normally activatedwhen cells are placed in tissue culture (26).

Abnormal ER response

Two of the dominant phenotypes were first identified byconfocal images of the stained striatal cells, which indicatedabnormal expansions of select perinuclear secretoryorganelles. As shown in Figure 6a, the mutant cells exhibitedexcessive ConA-reactive ER membrane. This was especiallyevident in cells that had been treated with the membrane fusioninhibitor, Brefeldin A. In contrast, the GM130-reactive Golgimembrane in these untreated or Brefeldin A-treated mutantcells was similar to wild-type, implying that the traffickingbetween the ER and Golgi apparatus may not be greatlyaffected. This relatively selective impact on the amount of ERmembrane implies an ER-stress response, which may involveabnormal forms of the mutant protein that were detected in thenucleus and in the cytoplasm of the mutant striatal cells.

Heightened basal activity of the iron pathway

The images in Figure 6b illustrated the intense perinuclear trans-ferrin receptor staining exhibited by the mutant cells, whetheruntreated or treated with Brefeldin A, which indicated that therecycling endosome compartment was enlarged and alsosuggested that transferrin receptor levels may be elevated. Thiswas confirmed by immunoblot analyses (Fig. 6c), whichrevealed a more intense transferrin receptor band in the mutantcell extracts. When normalized to spectrin, the band was ∼5-fold

increased in heterozygous cell lysates compared with the wild-type level. However, whereas the confocal images indicated themost dramatic increase in the homozygous cells, these extractstypically exhibited only an ∼2-fold elevation above the wild-type levels by immunoblot. This reproducible finding remains tobe explained, although it may suggest an interaction of trans-ferrin receptor with the mutant protein, which is also difficult todetect in this format.

Furthermore, the activity of the iron pathway in the mutantcells was dramatically elevated. The results of the uptake ofFITC-tagged transferrin from the growth medium are shown inFigure 7a. Increased uptake compared with wild-type cells wasevidenced by intense cytoplasmic and perinuclear FITC signal.Despite the already elevated activity of this pathway in mutantcells, treatment with deferoxamine mesylate led to increasedtransferrin uptake and transport for these cells and for the wild-type cells, indicating that the mutant protein does not alter theproper modulation of the iron pathway. These results implythat the heightened basal level of this pathway that wasobserved in the mutant striatal cells could reflect a normalresponse to hypoxia that has been elicited by the mutantprotein.

Increased levels of the normal and the mutant protein inresponse to hypoxia

Huntingtin is an iron-response protein that has been implicatedin cellular homeostasis (3), suggesting that the levels ofhuntingtin may be increased in the hypoxic mutant striatal cells.This possibility was assessed by immunoblot analyses of lysatesfrom cells of all genotypes, without and with deferoxaminemesylate treatment. As shown in Figure 7b, increased levels ofthe normal protein were disclosed by the more intense mAb2166-reactive bands in the treated compared with the untreated

Figure 5. Dominant mitigation of the SV40 oncogene signal reveals elevated p53. (a) Propidium iodide FACS analysis reveals that compared with wild-type cells(+/+) the majority of homozygous (111/111) and heterozygous (111/+) mutant cells exhibit doubled DNA content. (b) Immunoblot analysis of protein extracts(bottom) from homozygous mutant (111/111) and wild-type (+/+) cells reveals increased intensity of SV40 large T antigen (top) and greatly elevated p53 (middle)in mutant extract. Equal protein loading is confirmed by probing for spectrin. Densitometry of bands from two experiments reveals an SV40 large T antigen:spec-trin ratio of 5.7 U for wild-type and 13.0 U for mutant extracts and a p53:spectrin ratio of 0.2 U for wild-type and 1.4 U for mutant extracts.


Figure 6. Dominant high transferrin receptor reveals upregulated iron pathway. (a) Confocal images of ConA (white) ‘stain’ in wild-type (+/+), homozygous (111/111) and heterozygous (111/+) mutant striatal cells, reveal abnormal ER blobs in mutant cells, more evident after Brefeldin A (BFA) treatment. (Insets) GM130-reactive Golgi in untreated and BFA-treated cells (asterisk) are relatively similar in all genotypes. (b) Confocal images of transferrin receptor stained homozygous(111/111) and heterozygous (111/+) mutant cells reveal exaggerated perinuclear recycling compartment foci compared with wild-type (+/+) cells, also apparentafter BFA treatment (higher magnification). (c) Immunoblot analysis of tranferrin receptor levels in proteins extracted from STHdh+/Hdh+ (+/+), STHdhQ111/Hdh+

(111/+), STHdhQ111/Hdh111 (111/111) cells. Transferrin receptor TnfR bands, with differing mobility due to glycosylation, in the soluble (s) and pellet (p) fraction,are intensely reactive in mutant striatal cell extracts, indicating upregulation. This effect is most apparent in heterozygous cell extracts, as recovery fromhomozygous mutant cells is variable. Fodrin (spectrin) bands reveal comparable protein in each lane.


extracts for both the wild-type and the heterozygous mutantcells. Densitometry of the band intensities, normalized to thespectrin signal, indicated an ∼2-fold increase in the wild-typeand ∼5-fold increase in the mutant heterozygous cell extract,indicating that the levels of the normal protein are modulated instriatal cells in response to hypoxia. Furthermore, despite thedifficulty in detecting the mutant protein in this format, the bandintensities of mutant huntingtin in the treated versus theuntreated heterozygous and homozygous mutant cell lysates was

reproducibly increased by ∼3-fold, demonstrating iron regula-tion of the mutant allele.

These findings strongly suggested that the levels of hunt-ingtin in the untreated mutant striatal cells might be elevated,as they exhibited a heightened basal activity of the ironpathway compared with the naïve wild-type cells. Althoughthe differential ‘recovery’ of the mutant protein precludes anaccurate comparison between the naïve wild-type and themutant cell extracts, densitometry revealed that the ratio of the

Figure 7. Dominant high activity of the iron pathway reveals hypoxic response. (a) Confocal images of functional uptake of FITC-tagged transferrin by transferrinreceptor reveals increased signal in attached homozygous (111/111) and heterozygous (111/+) mutant cells (untreated), compared with wild-type cells (+/+).Deferoxamine mesylate (DM) appropriately stimulates increased FITC–transferrin uptake in wild-type (+/+), homozygous (111/111) and heterozygous (111/+)mutant cells. (b) Immunoblot of proteins extracted from deferoxamine mesylate (DM)-treated wild-type, homozygous and heterozygous mutant striatal cells (+)revealing strong mAb 2166-reactive huntingtin (httn) bands, compared with untreated extract (–). The mutant band, migrating more slowly, is less intense than thenormal protein due to differential ‘recovery’ but is upregulated by DM, indicating appropriate modulation. (c) Deferoxamine mesylate (DM) treatment is toxic toSTHdhQ111/Hdh111 (111/111) and STHdhQ111/Hdh+ (111/+) cells, evident from the numerous pyknotic FITC–transferrin-positive cells observed in uptake experi-ments. Wild-type cells (+/+) are not sensitive and remain firmly attached to the culture dish.


band of normal protein to the spectrin band in wild-type cellextracts is typically ∼1:1 (n = 2). In contrast, for the same expo-sures the ratio of the normal protein band to the spectrin bandin the heterozygous mutant cell lysates was ∼3:1 (n = 2). Theseratios disclose upregulation of huntingtin in the untreatedmutant striatal cells, implying that the basal levels of themutant protein may also be elevated perhaps setting up aharmful cycle that might further stimulate stress-responsepathways.

Sensitivity of the mutant striatal cells to hypoxic stress

During the course of the FITC–transferrin uptake experimentswe discovered that exposure to deferoxamine mesylatefollowed by repeated changes of growth medium to removeexcess FITC–ligand was associated with toxicity. The confocalimages in Figure 7c demonstrated that wild-type striatal cellswere refractory to the procedure. In contrast, the majority ofthe homozygous mutant cells and a proportion of theSTHdhQ111/Hdh+ cells had retracted their processes and werepyknotic or had already detached from the culture dish. Thus,additional hypoxia was harmful to the mutant striatal cells,implicating the already stimulated p53, ER and hypoxia stress-response pathways in this ultimately deleterious consequenceof the mutant protein.

DISCUSSION

Mutant huntingtin provokes a biochemical disruption thatinitiates the eventual demise of striatal neurons in HD. Thisintriguing dominant mechanism may, but need not, involvesome aspect of the protein’s normal activity. Consequently wehave assessed dominant phenotypes in mutant STHdhQ111

striatal cells to determine whether they resemble phenotypescaused by altered huntingtin function. Our results indicate that,although it shares properties of the normal protein that may beinvolved in neuronal selectivity, the full-length mutant proteinmay elicit eventual striatal cell toxicity via a mechanism that isseparate from its normal activity.

The alternate forms of the full-length protein that weredistinguished by their differing epitope accessibility supportprevious reports of distinct versions of the full-length proteinin medium spiny neurons in the mouse striatum (19). Theseforms were localized to a different set of organelles in thenucleus and in the cytoplasm. Our data also demonstrated thatnuclear and cytoplasmic forms of the protein weredistinguished by different ‘availability’ of the polyglutaminesegment. These alternate versions may reflect distinctmodifications, binding partners and/or conformationalproperties. They are also likely to involve huntingtin’s HEATprotein interaction domains (27), found in nuclear–cytoplasmic proteins such as the importins (28,29), which mayprovide a scaffold for the organization of components offunctional complexes.

An amino-terminal-accessible form of the full-length proteinlocalized to nuclear sites that are addresses of huntingtin’sspliceosome, transcription factor and polyadenylation complexpartners (3). This version may therefore fulfill the huntingtinrequirement for normal compartmentalization of transcriptionfactor partners (3), supporting a proposed role in RNA biogenesis(3–7). In contrast, the internal-accessible form of the protein was

found in the vicinity of perinuclear membranes, consistent withhuntingtin’s vesicle interactors (4,8–11), but was also associatedwith nucleoli. These organelles all require huntingtin, implicatingthis version of the protein in normal perinuclear membranefunction, in a pathway that is linked with the nucleolus (3).

Although the mutant protein shared this spectrum ofalternate forms, in support of the fulfillment of its intrinsicactivities (30), the mutant striatal cells also displayed abnormalversions that were ‘mislocalized’. Detected in the nucleus andin the cytoplasm, these abnormal forms of the full-lengthmutant protein exhibited aberrant polyglutamine segment‘accessibility’. This strongly suggests aberrant modifications,complexes and/or conformations of the full-length mutantprotein, which are promoted by the expanded polyglutamineproperty. This dominant phenotype mirrors the early‘mislocalization’ of the full-length mutant protein in theHdhQ111 knock-in striatum, which conforms to HD geneticcriteria (19). The formation of abnormal nuclear and cyto-plasmic forms of the full-length mutant protein may, therefore,be an early event in the pathogenic process in man.

The dominant mutant striatal cell phenotypes did not mirrorthe effects of altered huntingtin function, which yieldsdiminished nucleoli and perinuclear secretory organelles.Instead, the abnormal phenotypes pointed to potentiallyharmful metabolic stress. This finding argues for an aberrantproperty of the full-length mutant protein that has the capacityto disrupt striatal cell homeostasis via a mechanism that isseparate from its normal activity. This may involve, forexample, an abnormal interaction with a protein whosefunction is essential to striatal neurons.

Several candidates for the disrupted cellular pathway aresupported by these stress responses. An ‘ER unfolded proteinresponse’ (31,32), implying elevated chaperones andproteasome subunits that mitigate the effects of mutant poly-glutamine protein (33–35), implicate abnormal protein foldingand degradation (36,37). p53-mediated homeostasis pathways(38), which may include a response to DNA damage, areconsistent with aberrant gene transcription (39–41). Thehypoxic response suggests disruption of pathways involved inenergy metabolism that may also be awry in HD (42).

Importantly, our findings indicate that the mutant protein,like its normal counterpart, is upmodulated in response tohypoxia. Striatal neurons are exquisitely sensitive to oxidativestress, suggesting a cycle of mutant protein upregulation thatmay hasten the harmful effect of the polyglutamine property.Thus, two aspects of the mutant protein are suggested: anabnormal ‘toxic’ activity that is separate from its intrinsicactivity and its role as a hypoxia response protein, which maycontribute the selective vulnerability of striatal neurons to thisultimately deleterious insult.

MATERIALS AND METHODS

Establishment of E14 striatal precursor cell lines, cellculture and differentiation

Striatal primordia were dissected from genotyped E14 litter-mates of a timed pregnant HdhQ111/Hdh+ intercross, in ice-coldcalcium- and magnesium-free Hanks’ balanced salt solution(HBBS; Gibco BRL, Gaithersburg, MD), supplemented with3.9 g/l HEPES. Dissociated cells at ∼ 2–5 × 104 cells/cm2 were


plated in Dulbecco’s modified Eagle’s medium (DMEM) (LifeTechnologies, Gaithersburg, MD) plus 10% fetal calf serum(FCS; Imperial, UK). Infection with a defective retrovirustransducing the tsA58/U19 large T antigen, selection of G418resistant colonies at the permissive temperature 33°C andascertainment of nestin-positive colonies by immunostainingwas as described (20). The genotype of resultant cell lines wasconfirmed by PCR analysis to determine HD CAG repeat size(18). All experiments were carried out with two independentcell lines for each genotype.

Striatal cells were grown at 33°C in 10% FCS in DMEM.Occasional cells, of all genotypes, exhibit micronuclei and cyto-plasmic vacuoles that are not correlated with mutant huntingtin.Neuronal differentiation was induced by incubating for 12 h inserum-free DMEM with α-FGF (10 ng/ml) (Promega, Madison,WI), IBMX (240 µM), TPA (20 µM), forskolin (48.6 µM) anddopamine (5 µM) (Sigma, St Louis, MO) (20,21).

Protein extracts and immunoblot analysis

Protein extracts were prepared from phosphate-buffered saline(PBS) washed cell pellets by lysis at 4°C in RIPA buffer (50 mMTris–HCl pH 8, 1% NP40, 150 mM NaCl, 12 mM deoxycholicacid sodium salt, 0.1% SDS) with protease inhibitors. The super-natant was collected after centrifugation (25 min at 10 000 g),whereas the pellet in PBS was sonicated. For immunoblots 60 µgof total protein, boiled in SDS sample buffer, was loaded on a 6%SDS–polyacrylamide gel and blotted to nylon membrane.Protein concentration was by modified Bradford assay (BioRad,Hercules, CA). Blots were probed with primary antibody andspecific signal was detected with appropriate secondary reagentsby enhanced chemiluminescence.

Immunostaining, nuclear matrix preparation and confocalmicroscopy

For immunostaining PBS-washed cells were fixed in 4% para-formaldehyde, permeabilized in 0.1% Triton X-100 in PBS,blocked in 1% bovine serum albumin in PBS and were thenincubated with primary antibody (90 min), washed in PBS andexposed to secondary antibody (60 min). Similar results wereobtained with methanol fixation. Fluorescein-labeled trans-ferrin (Molecular Probes, Eugene, OR) uptake was performedover 30 min (43), in some cases 4 µM deferoxamine mesylatewas added to growth medium 24 h earlier (Sigma). Brefeldin A(Calbiochem, La Jolla, CA) treatment (5 µM) was for 30 min.Nuclear matrices were prepared by serial extraction as describedin Kern-matrix buffer (44). Secondary antibodies were Cy3-conjugated goat anti-rabbit (Jackson ImmunoResearch, WestGrove, PA) and Oregon Green-conjugated goat anti-mouse.Confocal analyses was on a BioRad MRC-1024 laser confocalmicroscope using a 40× objective lens. Digitized images weresaved as separate files for each channel and merged usingAdobe Photoshop.

Primary antibodies

The specific huntingtin reagents were as follows: EM48(amino acids 1–256) (45), AP194 (amino acids 1–17), AP229(amino acids 55–66), HP1 (amino acids 80–113), HF1 (aminoacids 1981–2580) (14,23), mAb 2166 (amino acids 181–810)(Chemicon International, Temecula, CA) and mAb 1F8 (3,35).

Other reagents were as follows: MAP-2 (BoehringerMannheim, Indianapolis, IN), nestin (Developmental StudiesHybridoma Bank, Univeristy of Iowa, IA), NuMA (Transduc-tion Laboratories, San Diego, CA), ANA125 (a gift of Dr M.Bedford, Harvard Medical School, Boston, MA), fibrillarin72B9 (a gift of Dr R. Terns, University of Georgia, Athens,GA), FITC–ConA (Molecular Probes), GM130 (TransductionLaboratories), transferrin receptor (Biosource International,Camarillo, CA) and SV40 large T antigen (Pab 416), p53(Pab 421) (46) (gifts from Dr E. Harlow, MassachusettsGeneral Hospital, Charlestown, MA).

FACS analysis for DNA content

DNA content was measured by flow cytometry of cells stainedwith propidium iodide (Calbiochem) and the data analyzed withDNA analysis CellQuest software. Briefly, after trypsinizationcells were filtered through a 37 µm mesh, centrifuged at 90 g for10 min and resuspended in HEPES–HBSS. Counted cells werefixed for 15 min in 4% paraformaldehyde, permeabilized in 0.1%Triton X-100 for 3 min at 4°C and resuspended in RNAse A(180 U/ml) for 40 min at 37°C. DNA staining was on 1–2 × 106

cells in 1 ml of propidium iodide (50 µg/ml propidium iodide/PBS/3% PEG-8000).

ACKNOWLEDGEMENTS

We thank Drs M. Bedford, E. Harlow, R. Terns and X.J. Li forgifts of antibody reagents, Drs K. McGinnis and J.F. Gusellafor discussion and Mr V. Vrbanac and Ms L. Lebel for experttechnical assistance. We are grateful to the MGH FACS CoreFacility for FACS analyses. This work was supported byNINDS grants NS32765 and NS16367 (Huntington’s DiseaseCenter Without Walls), grants to E.C. from the Foundation forthe Care and Cure of Huntington’s Disease and Telethon, Italy.V.C.W. is the recipient of a postdoctoral fellowship from theHereditary Disease Foundation.

REFERENCES

1. Vonsattel, J.P. and DiFiglia, M. (1998) Huntington disease. J. Neuropathol.Exp. Neurol., 57, 369–384.

2. Huntington’s Disease Collaborative Research Group (1993) A novel genecontaining a trinucleotide repeat that is expanded and unstable on Hunt-ington’s disease chromosomes. Cell, 72, 971–983.

3. Hilditch-Maguire, P., Trettel, F., Passani, L.A., Auerbach, A., Persichetti,F. and MacDonald, M.E. (2000) Huntingtin: an iron-regulated proteinessential for normal nuclear and perinuclear organelles. Hum. Mol.Genet., 9, 2789–2797.

4. Faber, P.W., Barnes, G.T., Srinidhi, J., Chen, J., Gusella, J.F. andMacDonald, M.E. (1998) Huntingtin interacts with a family of WW domainproteins. Hum. Mol. Genet., 7, 1463–1474.

5. Boutell, J.M., Thomas, P., Neal, J.W., Weston, V.J., Duce, J., Harper, P.S.and Jones, A.L. (1999) Aberrant interactions of transcriptional repressorproteins with the Huntington’s disease gene product, huntingtin. Hum.Mol. Genet., 8, 1647–1655.

6. Passani, L.A., Bedford, M.T., Faber, P.W., McGinnis, K.M., Sharp, A.,Gusella, J.F., Vonsattel, J.-P. and MacDonald, M.E. (2000) Huntingtin’sWW domain partners in HD post-mortem brain fulfill genetic criteria fordirect involvement in HD pathogenesis. Hum. Mol. Genet., 9, 2175–2182.

7. Takagaki, Y. and Manley, J.L. (2000) Complex protein interactions withinthe human polyadenylation machinery identify a novel component. Mol.Cell Biol., 20, 1515–1525.

8. Li, X.J., Li, S.H., Sharp, A.H., Nucifora Jr, F.C., Schilling, G., Lanahan, A.,Worley, P., Snyder, S.H. and Ross, C.A. (1995) A huntingtin-associated pro-tein enriched in brain with implications for pathology. Nature, 378, 398–402.


9. Kalchman, M.A., Koide, H.B., McCutcheon, K., Graham, R.K., Nichol,K., Nishiyama, K., Kazemi-Esfarjani, P., Lynn, F.C., Wellington, C.,Metzler, M. et al. (1997) HIP1, a human homologue of S.cerevisiae Sla2p,interacts with membrane-associated huntingtin in the brain. NatureGenet., 16, 44–53.

10. Wanker, E.E., Rovira, C., Scherzinger, E., Hasenbank, R., Walter, S., Tait,D., Colicelli, J. and Lehrach, H. (1997) HIP-I: a huntingtin interacting pro-tein isolated by the yeast two-hybrid system. Hum. Mol. Genet., 6, 487–495.

11. Velier, J., Kim, M., Schwarz, C., Kim, T.W., Sapp, E., Chase, K., Aronin, N.,DiFiglia, M. (1998) Wild-type and mutant huntingtins function in vesicle traf-ficking in the secretory and endocytic pathways. Exp. Neurol., 152, 34–40.

12. Gusella, J.F., Persichetti, F. and MacDonald, M.E. (1997) The geneticdefect causing Huntington’s disease: repeated in other contexts? Mol.Med., 3, 238–2468.

13. Gusella, J.F. (2000) Huntington’s disease: two decades from mystery tomodels. NeuroScience News, 3, 15–22.

14. Scherzinger, E., Lurz, R., Turmaine, M., Mangiarini, L., Hollenbach, B.,Hasenbank, R., Bates, G.P., Davies, S.W., Lehrach, H. and Wanker, E.E.(1997) Huntingtin-encoded polyglutamine expansions form amyloid-likeprotein aggregates in vitro and in vivo. Cell, 90, 549–558.

15. Huang, C.C., Faber, P.W., Persichetti, F., Mittal, V., Vonsattel, J.-P.,MacDonald, M.E. and Gusella, J.F. (1998) Amyloid formation by mutanthuntingtin: threshold, progressivity and recruitment of normal poly-glutamine proteins. Somatic Cell Mol. Genet., 24, 217–233.

16. Scherzinger, E., Sittler, A., Schweiger, K., Heiser, V., Lurz, R., Hasenbank,R., Bates, G.P., Lehrach, H. and Wanker, E.E. (1999) Self-assembly ofpolyglutamine-containing huntingtin fragments into amyloid-like fibrils:implications for Huntington’s disease pathology. Proc. Natl Acad. Sci.USA, 96, 4604–4609.

17. Goldberg, Y.P., Nicholson, D.W., Rasper, D.M., Kalchman, M.A., Koide,H.B., Graham, R.K., Bromm, M., Kazemi-Esfarjani, P., Thornberry, N.A.,Vaillancourt, J.P. et al. (1996). Cleavage of huntingtin by apopain, a pro-apoptotic cysteine protease, is modulated by the polyglutamine tract. NatureGenet., 13, 442–449.

18. Leavitt, B.R., Wellington, C.L. and Hayden, M.R. (1999). Recent insightsinto the molecular pathogenesis of Huntington disease. Semin. Neurol.,19, 385–395.

19. Wheeler, V.C., White, J.K., Gutekunst, C.A., Vrbanac, V., Weaver, M.,Li, X.J., Li, S.H., Yi, H., Vonsattel, J.P., Gusella, J.F. et al. (2000) Longglutamine tracts cause nuclear localization of a novel form of huntingtin inmedium spiny striatal neurons in HdhQ92 and HdhQ111 knock-in mice.Hum. Mol. Genet., 9, 503–513.

20. Cattaneo, E. and Conti, L. (1998). Generation and characterization ofembryonic striatal conditionally immortalized ST14A cells. J. Neuro.Res., 53, 223–234.

21. Du, X., Stull, N.D. and Iacovitti, L. (1994) Novel expression of the tyro-sine hydroxylase gene requires both acidic fibroblast growth factor and anactivator. J. Neurosci., 14, 7688–7694.

22. Persichetti, F., Carlee, L., Faber, P.W., McNeil, S.M., Ambrose, C.M.,Srinidhi, J., Anderson, M., Barnes, G.T., Gusella, J.F. and MacDonald,M.E. (1996) Differential expression of normal and mutant Huntington’sdisease gene alleles. Neurobiol. Dis., 3, 183–190.

23. Persichetti, F., Trettel, F., Huang, C.C., Fraefel, C., Timmers, H.T.M.,Gusella, J.F. and MacDonald, M.E. (1999) Mutant huntingtin forms invivo complexes with distinct context-dependent conformations of thepolyglutamine segment. Neurobiol. Dis., 6, 364–375.

24. Rigamonti, D., Bauer, J.H., De-Fraja, C., Conti, L., Sipione, S., Sciorati,C., Clementi, E., Hackam, A., Hayden, M.R., Li, Y. et al. (2000) Wild-type huntingtin protects from apoptosis upstream of caspase-3. J.Neurosci., 20, 3705–3713.

25. Lehman, J.M., Perry, M.B. and Laffin, J. (1996). Induction of tetraploidDNA content by mutant simian virus 40 T antigen that has reduced com-plex formation with p53. Cytometry, 23, 126–130.

26. Artandi, S.E. and DePinho, R.A. (2000) Mice without telomerase: whatcan they teach us about human cancer? Nature Med., 6, 852–855.

27. De Rooij, K.E., Dorsman, J.C., Smoor, M.A., Den Dunnen, J.T. and VanOmmen, G.J. (1996) Subcellular localization of the Huntington’s diseasegene product in cell lines by immunofluorescence and biochemicalsubcellular fractionation. Hum. Mol. Genet., 5, 1093–1099.

28. Vetter, I.R., Arndt, A., Kutay, U., Gorlich, D. and Wittinghofer, A. (1999)Structural view of the Ran-Importin beta interaction at 2.3 Å resolution.Cell, 97, 635–646.

29. Seedorf, M. and Silver, P.A. (1997) Importin/karyopherin protein familymembers required for mRNA export from the nucleus. Proc. Natl Acad.Sci. USA, 94, 8590–8595.

30. White, J.K., Auerbach, W., Duyao, M.P., Vonsattel, J.P., Gusella, J.F.,Joyner, A.L. and MacDonald, M.E. (1997) Huntingtin is required for neuro-genesis and is not impaired by the Huntington’s disease CAG expansion.Nature Genet., 17, 404–410.

31. Hampton, R.Y. (2000) Getting the UPR hand on misfolded proteins. Curr.Biol., 10, 518–521.

32. Welihinda, A.A., Tirasophon, W. and Kaufman, R.J. (1999) The cellularresponse to protein misfolding in the endoplasmic reticulum. Gene Exp.,7, 293–300.

33. Cummings, C.J., Mancini, M.A., Antalffy, B., DeFranco, D.B., Orr, H.T.and Zoghbi, H.Y. (1998) Chaperone suppression of aggregation andaltered subcellular proteasome localization imply protein misfolding inSCA1. Nature Genet., 19, 148–154.

34. Chai, Y., Koppenhafer, S.L., Bonini, N.M. and Paulson, H.L. (1999)Analysis of the role of heat shock protein (Hsp) molecular chaperones inpolyglutamine disease. J. Neurosci., 19, 10338–10347.

35. Cummings, C.J., Reinstein, E., Sun, Y., Antalffy, B., Jiang, Y.,Ciechanover, A., Orr, H.T., Beaudet, A.L. and Zoghbi, H.Y. (1999)Mutation of the E6-AP ubiquitin ligase reduces nuclear inclusion fre-quency while accelerating polyglutamine-induced pathology in SCA1mice. Neuron, 24, 879–892.

36. Paulson, H.L. (1999) Protein fate in neurodegenerative proteinopathies:polyglutamine diseases join the (mis)fold. Am. J. Hum. Genet., 64, 339–345.

37. Orr, H.T. and Zoghbi, H.Y. (2000) Reversing neurodegeneration: a promiseunfolds. Cell, 101, 1–4.

38. Ashcroft, M., Taya, Y. and Vousden, K.H. (2000) Stress signals utilizemultiple pathways to stabilize p53. Mol. Cell. Biol., 20, 3224–3233.

39. Steffan, J.S., Kazantsev, A., Spasic-Boskovic, O., Greenwald, M., Zhu,Y.Z., Gohler, H., Wanker, E.E., Bates, G.P., Housman, D.E. andThompson, L.M. (2000) The Huntington’s disease protein interacts withp53 and CREB-binding protein and represses transcription. Proc. NatlAcad. Sci. USA, 97, 6763–6768.

40. Luthi-Carter, R., Strand, A., Peters, N.L., Solano, S.M., Hollingsworth,Z.R., Menon, A.S., Frey, A.S., Spektor, B.S., Penney, E.B., Schilling, G.et al. (2000) Decreased expression of striatal signaling genes in a mousemodel of Huntington’s disease. Hum. Mol. Genet., 22, 1259–1271.

41. McCampbell, A., Taylor, J.P., Taye, A.A., Robitschek, J., Li, M., Walcott,J., Merry, D., Chai, Y., Paulson, H., Sobue, B. and Fischbeck, K.H. (2000)CREB-binding protein sequestration by expanded polyglutamine. Hum.Mol. Genet., 9, 2197–2202.

42. Browne, S.E., Ferrante, R.J. and Beal, M.F. (1999) Oxidative stress inHuntington’s disease. Brain Pathol., 9, 147–163.

43. Ghosh, R.N., Gelman, D.L. and Maxfield, T.R. (1994) Quantification oflow density lipoprotein and transferrin endocytic sorting in Hep2 cellsusing confocal microscopy. J. Cell Sci., 107, 2177–2189.

44. Staufenbiel, M. and Deppert, W. (1984) Preparation of nuclear matricesfrom cultured cells: subfractionation of nuclei in situ. J. Cell Biol., 98,1886–1894.

45. Gutekuns, C.-A, Li, S.H., Yi, H., Mulroy, J.S., Kuemmerle, S., Jones, R.,Rye, D., Ferrante, R.J., Hersch, S.M. and Li, X.J. (1999) Nuclear andneuropil aggregates in Huntington’s disease: relationship to neuropathol-ogy. J. Neurosci., 19, 2522–2534.

46. Shmieg, F.I. and Simmons, D.T. (1988) Characterization of the in vitrointeraction between SV40 T antigen and p53: mapping the p53 bindingsite. Virology, 164, 132–140.

dominant phenotypes produced by the hdmutation in ......hdh+ (+/+), sthdhq111/hdh+...

Documents