Dispatch R789

Cell-cycle signaling: Atm displays its many talentsChristoph Heiner Westphal

The discovery of multiple signaling cascadesdownstream of Atm may lead to a clearer understandingof the diverse defects seen in ataxia-telangiectasia.These pathways which include evolutionarily conservedChk1 and Atr, and non-conserved p21, p53 and Abl guard genomic integrity after DNA damage.

Address: Department of Genetics and Howard Hughes MedicalInstitute, Harvard Medical School, 200 Longwood Avenue, Boston,Massachusetts 02115, USA.E-mail: westphal@rascal.med.harvard.edu

Current Biology 1997, 7:R789R792http://biomednet.com/elecref/09609822007R0789

Current Biology Ltd ISSN 0960-9822

Ataxia-telangiectasia (AT) is a recessive childhood diseasecaused by mutations in the ATM (AT-mutated) gene [1].Symptoms include neurological abnormalities that causeunsteady posture (ataxia), dilated blood vessels (telangiec-tasia), infertility, radiation sensitivity, immune deficien-cies and lymphoid malignancies. It appears likely that thediverse defects seen in ATM null mice [25] and humanswith AT [6,7] are phenotypic manifestations of disparatesignal-transduction deficits. This interpretation issupported by recent evidence implicating p53 [8,9], Abl[10,11], Chk1 [1214] and Atr (AT- and Rad3-relatedkinase) [15] in a variety of pathways that include Atm.

Mammalian Atm turns out to be a member of a proteinkinase family that includes Atr, DNA-PK (DNA-depen-dent protein kinase) and FRAP [16], which are related toAtm by their carboxy-terminal kinase domains. Atmhomologs in other organisms, such as Mec1 and Tel1 ofSaccharomyces cerevisiae, Rad3 of Schizosaccharomyces pombe,and Mei-41 of Drosophila melanogaster, have shed lightupon diverse functions of this protein [17], the main roleof which appears to be the induction of DNA-damageresponses [6,7,16,1820]. These damage-control pathwaysare induced by both genotoxic insults such as ionizingradiation or anti-tumor medications and the pro-grammed DNA breaks of meiosis. This is consistent withthe phenotype of AT cells, which show chromosomalinstability and hypersensitivity to agents that cause DNA-strand breaks.

The description of an irradiation-induced pathwayleading from Atm to p53 [21] provided the first indicationthat the diverse disease manifestations seen in AT mayreflect impairment of signal-transduction cascades.Recent observations support this hypothesis [815] and

delineate a number of signal-transduction cascades ema-nating from the Atm protein kinase. Interestingly, theAtm signaling cascades involving Chk1 and Atr appear tobe conserved, at least in part, from yeast to mammals, andare likely to play a central role in the maintenance ofgenomic integrity.

Atm, p53 and p21: genetics links old friendsAtm and p53 have been postulated to interact in cellulargrowth control, the mediation of cell-cycle checkpoints,apoptosis and the organismal response to ionizingradiation [7]. The generation of knockout mice and thederivation of p53 and/or Atm deficient cells from thesemice [8,9] has made it possible to test these postulatedfunctions.

Atm-deficient embryonic fibroblasts, derived from ATMknockout mice, undergo premature growth arrest inculture, probably because of their increased genomicinstability [2,3,5]. In sharp contrast, fibroblasts deficient inboth Atm and p53 grow rapidly in culture, indicating thatp53 is a prominent modulator of the growth arrest seen inthe ATM null state, and consistent with Atm and p53being connnected by a linear pathway (Figure 1a). AsATM and p53 mutant cells are both defective in the DNA-damage-induced G1/S cell-cycle checkpoint, these twogenes may interact in this pathway as well [7]. Notably,the irradiation-induced checkpoint defect is no worse inATM/p53 double mutant fibroblasts than either singlemutant [9], again consistent with a linear cascade(Figure 1b), although it is likely that they interact in amore complex manner than portrayed here. These resultshighlight the paradox that Atm may either promote orinhibit growth, depending on the circumstances, and thatp53 may modulate both of these effects.

Atm and p53 have both been implicated in apoptosis andin the organismal response to ionizing radiation [7]. Inter-estingly, ATM mutant thymocytes are only partially resis-tant to ionizing-radiation-induced apoptosis, whereas p53and ATM/p53 mutant thymocytes are profoundly resistant.Furthermore, the increase in p53 protein level that is nor-mally induced by ionizing radiation is significantlydelayed in ATM mutant thymocytes [8]. These results areconsistent with p53 and Atm being components of an ion-izing-radiation-induced apoptosis pathway, although inthis case there is substantial non-linearity in the Atmp53interactions (Figure 1b). Lastly, ATM mutant mice andhumans are exquisitely sensitive to ionizing radiation.Here again, p53 has been postulated to interact with Atm

[2,6,7]. Surprisingly, ATM/p53 double mutant mice are justas sensitive to ionizing radiation as the ATM singlemutants, arguing that the two genes do not interact inacute radiation toxicity.

A more detailed biochemical understanding of the interac-tions between Atm and p53 is the goal of intense currentresearch efforts. These interactions in mammalian cellsare not evolutionarily conserved in yeast, and so must benecessary only for the subtle control of the DNA-damageresponse in higher organisms, or, alternatively, not becrucial in the organismal DNA-damage response.

Radiation induces an Atm/Abl phosphorylation cascadeThe abl proto-oncogene, known to be involved in specifictranslocation-associated human tumors, has been impli-cated in the DNA-damage-induced G1/S cell cyclecheckpoint [22]. As Atm is also necessary for the damage-induced G1/S cell-cycle checkpoint [7], interactionsbetween these two genes and their protein products havebeen studied in some detail [10,11].

Atm binds to Abl constitutively in human cells via weakinteractions between a proline-rich region of Atm and theSrc-homology 3 (SH3) domain of Abl [11]. Gamma irradia-tion induces Abl tyrosine kinase activity, which is absentin ATM null cells of humans [11] and mice [10]. This tyro-sine kinase activity is dependent upon Atm-mediatedphosphorylation of Abl at Ser465 [10], but it is not clearwhether loss of this phosphorylation is associated withfailure of the G1/S cell cycle checkpoint. These data canbe summarized in a model of the interactions betweenAtm and Abl in mammalian cells (Figure 2).

The importance of these biochemical data will likely beclarified by further work analyzing their functional conse-quences. It will be especially important to confirm that ablmutant mouse embyonic fibroblasts are defective in aG1/S cell-cycle checkpoint, and to analyze the possibleinteractions of other proteins with the Atm/Abl signalingpathway outlined here. As with Atm and p53, AtmAblinteractions are not evolutionarily conserved in lowereukaryotes. This may mean either that these interactionsare not central to DNA-damage responses, or that moresophisticated regulation is afforded to mammalian cells byuse of this novel pathway.

Atm/Chk1 pathway conserved in yeast and mammals The chk1 gene is a central mediator of DNA-damage-induced cell-cycle checkpoints in S. pombe [23,24]. Chk1

R790 Current Biology, Vol 7 No 12

Figure 1

Pathways linking Atm and p53 in the mammalian DNA-damageresponse. (a) Loss of ATM leads to an abnormal accumulation of DNAdamage and so premature growth arrest in both mouse and humanfibroblasts, with concurrent elevation of p21 protein levels. This growtharrest is alleviated in ATM/p53 double mutant mouse fibroblasts, sothat Atm and p53 can be linked in a pathway. (b) Loss of both ATMand p53 has no greater effect on the G1/S cell-cycle checkpoint thanloss of either alone. This is again consistent with a linear pathway(pathway on left). In this case, however, Atm induces p53 activity andso cell-cycle arrest. In thymocyte apoptosis (pathway on right), ATMand p53 interact only partially, indicating substantial non-linearity ofthis pathway. Note that the actual pathways linking ATM and p53 arelikely to be much more complex, and less linear, than thosepresented here.

(a) Atm

EndogenousDNA damage

p53

p21/others p21/others

Cell proliferation

Gamma irradiation

Atm Atm

p53 p53

G1/S checkpoint Thymocyte apoptosis

(b)

Current Biology

Other factors

Figure 2

The mammalian radiation-induced AtmAbl phosphorylation cascade.Atm binds to and phosphorylates Abl upon -irradiation. Thisphosphorylation makes Abl active as a tyrosine kinase and may, inconcert with other factors, play a role in the DNA damage-inducedG1/S cell-cycle checkpoint. Determining the functional importance ofthis pathway in the mediation of the DNA-damage response is the goalof current research efforts. This pathway, though presented as beinglinear here, may well be more convoluted.

Gamma irradiation

Atm

Abl Abl-phosphate(active tyrosine kinase)

? Other mediators

G1/S cell-cycle checkpoint

Current Biology

?

Dispatch R791

can be placed in a genetic pathway leading from rad3 theS. pombe homolog of both ATM and ATR to cdc2/cyclin B[18]. This signal-transduction cascade (Figure 3) is essentialfor cell-cycle arrest after DNA damage.

Recently, two groups have reported the cloning ofmammalian chk1 [12,13]. Chk1 expression patterns closelymirror those of ATM in mouse tissues [13], and Chk1protein expression is dependent upon ATM [12]. As rad3is the S. pombe homolog of ATM, these results indicate thatthe DNA-damage response upstream from chk1 appears tobe substantially conserved from fission yeast to mammals([12] and Figure 3).

Following DNA damage, Chk1 activity is necessary forthe phosphorylation of the dual-specificity phosphatasesCdc25A/B/C, thus preventing the activation of Cdc2/cyclin B which requires removal of an inhibitory phos-phate from Cdc2 and so mitotic entry [13,14], both infission yeast and in mammalian cells (Figure 3). Thesedata indicate that an entire yeast checkpoint signalingcascade is largely conserved in mammals. Thus, a largeamount of yeast work may help us to understand themammalian cell-cycle checkpoint machinery. Moreover,the extraordinary level of conservation of this pathwayindicates that it may be crucial to the maintenance ofgenomic stability.

ATM is necessary for ATR-associated protein kinase activityAtr is the closest mammalian homolog of Atm [15,25] andis most related to the S. pombe Rad3 protein kinase, animportant mediator of DNA-damage-induced cell-cyclecheckpoints [16,18]. Hence, a study of the AtmAtr inter-action in mammalian cells may shed light upon the mam-malian DNA-damage response.

Atr protein kinase activity is dependent upon ATM, whilethe expression of Atr protein is independent of both ATMand p53 in mouse testis (M.F. Hoekstra, personal commu-nication). This observation links the function of twoimportant checkpoint mediators in a regulatory cascade,similar to the results obtained with ATM and chk1(Figure 3). It appears likely that this cascade, involvingthe evolutionarily conserved genes ATM and ATR, willmediate some aspects of the DNA-damage response,though formal proof of the importance of Atr in thisresponse awaits studies of ATR null mice.

Loss of Atm affects multiple signaling cascadesThe pleiotropic nature of AT may reflect impairment ofdiverse signaling networks in the ATM null state. Recentbiochemical and genetic studies indicate that Atm inter-acts with a number of key regulators of the DNA-damageresponse, including p53, p21, Abl, Chk1 and Atr. Thereappear to be two classes of Atm-dependent signaling cas-cades: the first involving non-conserved proteins, such as

p21, p53 and Abl, and the second involving conserved pro-teins, such as Chk1 and Atr.

The pathways of the latter class are substantially con-served from yeast to mammals and are likely, therefore, tobe central to any DNA-damage response. By contrast, theAtm-dependent signal-transduction cascades involvingp21, p53 and Abl are not evolutionarily conserved, andmore detailed study will be needed to establish their func-tional importance. Elucidation of the precise interaction ofevolutionarily non-conserved cell-cycle checkpoint regula-tors with well-conserved cell-cycle proteins remains amajor goal of cell-cycle research.

The DNA-damage-induced signal-transduction pathwayslinking Atm, p21, p53, Abl, Chk1 and Atr, outlined in thisdispatch, are unlikely to be truly linear. Hence, more sys-tematic studies of these pathways are needed. These

Figure 3

Conservation of the Chk1 pathway from fission yeast (S. pombe) tomammals, and interactions between the Rad3 homologs Atm and Atr.S. pombe components of the DNA-damage-induced cell cycleresponse are shown on the left, and their mammalian homologs areshown on the right. Note that Atm is necessary for Chk1 proteinexpression, which, through direct or indirect mechanisms, inhibitsCdc25, thus preventing Cdc2 dephosphorylation and cell-cycleprogression. It appears that this pathway is substantially conservedfrom yeast to mammals, though the precise details of regulation aredifferent. For example, S. pombe Rad3 has two homologs in mammals,Atm and Atr. Atm is necessary for Atr-associated protein kinaseactivity, but not Atr protein expression. Mammalian cells may regulateChk1 activity and cell cycle progression via AtmAtr interactions(compare this to the apparent linearity of the Rad3Chk1 interaction inS. pombe). Atm/Atr interactions hence appear to represent a variationon the theme seen in yeast. Note that the actual DNA-damageresponse pathway is likely to be more intricate than portrayed here.

Gamma irradiation

Rad3

Chk1

Cdc25

Cdc2

Atm

Chk1

Cdc25

Cdc2

Mouse

Atr-associatedkinase activity

?

?

Fissionyeast

G2/M cell-cycle progression(?G1/S and ? S progression)

Current Biology

studies should uncover further clues to the pleiotropicnature of AT, and reveal some of the molecular defectsthat underlie carcinogenesis in general. Hopefully, theseinsights will lead to novel anti-tumor strategies, and toimproved therapeutic solutions for the many disordersthat plague children with AT.

AcknowledgementsI would like to express gratitude to Merl Hoekstra, Stephen Elledge, GailFlaggs, Kathy Keegan, Sylvia Pagn and Philip Leder for their comments,which greatly improved this manuscript. Special thanks to Philip Leder forhis kind scientific and personal mentorship. I apologize that not all pertinentcitations could be included due to space constraints.

References1. Savitsky K, Bar-Shira A, Gilad S, Rotman G, Ziv Y, Vanagaite L, et al.:

A single ataxia telangiectasia gene with a product similar to PI-3kinase. Science 1995, 268:1749-1753.

2. Elson A, Wang Y, Daugherty CJ, Morton CC, Zhou F, Campos-TorresJ, et al.: Pleiotropic defects in ataxia-telangiectasia protein-deficient mice. Proc Natl Acad Sci USA 1996, 93:13084-13089.

3. Barlow C, Hirotsune S, Paylor R, Liyanage M, Eckhaus M, Collins F, etal.: ATM-deficient mice: a paradigm of ataxia telangiectasia. Cell1996, 86:159-171.

4. Xu Y, Ashley T, Brainerd EE, Bronson RT, Meyn MS, Baltimore D:Targeted disruption of ATM leads to growth retardation,chromosomal fragmentation during meiosis, immune defects,and thymic lymphoma. Genes Dev 1996, 10:2411-2422.

5. Xu Y, Baltimore D: Dual roles of ATM in the cellular response toradiation and in cell growth control. Genes Dev 1996, 10:2401-2410.

6. Lavin MF, Shiloh Y: The genetic defect in ataxia-telangiectasia.Annu Rev Immunol 1997, 15:177-202.

7. Morgan SE, Kastan MB: p53 and ATM: cell cycle, cell death, andcancer. Adv Cancer Res 1997, 71:1-25.

8. Westphal CH, Rowan S, Schmaltz C, Elson A, Fisher DE, Leder P:ATM and p53 cooperate in apoptosis and suppression oftumorigenesis, but not in resistance to acute radiation toxicity.Nature Genet 1997, 16:397-401.

9. Westphal CH, Schmaltz C, Rowan S, Elson A, Fisher DE, Leder P:Genetic interactions between ATM and p53 influence cellularproliferation and irradiation-induced cell cycle checkpoints.Cancer Res 1997, 57:1664-1667.

10. Baskaran R, Wood LD, Whitaker LL, Canman CE, Morgan SE, Xu Y,et al.: Ataxia telangiectasia mutant protein activates c-Abl tyrosinekinase in response to ionizing radiation. Nature 1997, 387:516-519.

11. Shafman T, Khanna KK, Kedar P, Spring K, Kozlov S, Yen T, et al.:Interaction between ATM protein and c-Abl in response to DNAdamage. Nature 1997, 387:520-523.

12. Flaggs G, Plug A, Dunks KM, Mundt KE, Ford JC, Quiggle MRE,Taylor EM, Westphal CH, Ashley T, Hoekstra MF, Carr AM: Atm-dependent interactions of a mammalian Chk1 homolog withmeiotic chromosomes. Curr Biol 1997, 7:this issue.

13. Sanchez Y, Wong C, Thoma RS, Richman R, Wu Z, Piwnica-WormsH, et al.: Conservation of the Chk1 checkpoint pathway inmammals: linkage of DNA damage to Cdk regulation throughCdc25. Science 1997, 277:1497-1501.

14. Furnari B, Rhind N, Russell P: Cdc25 mitotic inducer targeted byChk1 DNA damage checkpoint kinase. Science 1997, 277:1495-1497

15. Keegan KS, Holtzman DA, Plug AW, Christenson ER, Brainerd EE,Flaggs G, et al.: The ATR and ATM protein kinases associate withdifferent sites along meiotically pairing chromosomes. Genes Dev1996, 10:2423-2437.

16. Hoekstra MF: Responses to DNA damage and regulation of cellcycle checkpoints by the ATM protein kinase family. Curr OpinGenet Dev 1997, 7:170-175.

17. Jackson SP: Cancer predisposition. Ataxia-telangiectasia at thecrossroads. Curr Biol 1995, 5:1210-1212.

18. Elledge SJ: Cell cycle checkpoints: preventing an identity crisis.Science 1996, 274:1664-1672.

19. Carr AM: Control of cell cycle arrest by the Mec1sc/Rad3sp DNAstructure checkpoint pathway. Curr Opin Genet Dev 1997, 7:93-98.

20. Hawley RS, Friend SH: Strange bedfellows in even strangerplaces: the role of ATM in meiotic cells, lymphocytes, tumors, andits functional links to p53. Genes Dev 1996, 10:2383-2388.

21. Kastan MB, Zhan Q, el-Deiry WS, Carrier F, Jacks T, Walsh WV, etal.: A mammalian cell cycle checkpoint pathway utilizing p53 andGADD45 is defective in ataxia-telangiectasia. Cell 1992, 71:587-597.

22. Yuan ZM, Huang Y, Whang Y, Sawyers C, Weichselbaum R,Kharbanda S, et al.: Role for c-Abl tyrosine kinase in growth arrestresponse to DNA damage. Nature 1996, 382:272-274.

23. Walworth N, Davey S, Beach D: Fission yeast chk1 protein kinaselinks the rad checkpoint pathway to cdc2. Nature 1993, 363:368-371.

24. al-Khodairy F, Fotou E, Sheldrick KS, Griffiths DJ, Lehmann AR, CarrAM: Identification and characterization of new elements involvedin checkpoint and feedback controls in fission yeast. Mol Biol Cell1994, 5:147-160.

25. Bentley NJ, Holtzman DA, Flaggs G, Keegan KS, DeMaggio A, FordJC, et al.: The Schizosaccharomyces pombe rad3 checkpoint gene.EMBO J 1996, 15:6641-6651.

R792 Current Biology, Vol 7 No 12

Atm displays its many talentsAtm, p53 and p21: genetics links old friendsRadiation induces an Atm/Abl phosphorylation cascadeAtm/Chk1 pathway conserved in yeast and mammals ATM is necessary for ATR-associated protein kinase activityLoss of Atm affects multiple signaling cascadesAcknowledgementsReferences

FiguresFigure 1 and Figure 2Figure 3