Gidi`s Student presentations/8

REVIEWS
ATM AND RELATED PROTEIN
KINASES: SAFEGUARDING GENOME
INTEGRITY
Yosef Shiloh
Maintenance of genome stability is essential for avoiding the passage to neoplasia. The DNAdamage response — a cornerstone of genome stability — occurs by a swift transduction of the
DNA-damage signal to many cellular pathways. A prime example is the cellular response to DNA
double-strand breaks, which activate the ATM protein kinase that, in turn, modulates numerous
signalling pathways. ATM mutations lead to the cancer-predisposing genetic disorder ataxiatelangiectasia (A-T). Understanding ATM’s mode of action provides new insights into the
association between defective responses to DNA damage and cancer, and brings us closer to
resolving the issue of cancer predisposition in some A-T carriers.
The David and Inez Myers
Laboratory for Genetic
Research, Department of
Human Genetics and
Molecular Medicine,
Sackler School of Medicine,
Tel Aviv University,
Tel Aviv 69978, Israel.
e-mail: [email protected]
doi:10.1038/nrc1011
NATURE REVIEWS | C ANCER
‘Cancer is a genetic disease of the somatic cell’. This
cliché makes increasing sense with the discovery of
every new oncogene and tumour-suppressor gene. The
road to cancer is paved with alterations in the sequence
and organization of the cellular genome that range from
single-nucleotide substitutions to gross chromosomal
aberrations. These represent deviations from a prime
requisite for cellular homeostasis: maintenance of
genome stability. As the neoplastic cellular phenotype
progresses, genomic stability continues to deteriorate,
leading to a vicious cycle of genomic aberrations and
advancing malignancy1.
Although sequence variation of germline DNA is
essential for maintaining genetic variability, changes in
DNA sequence in somatic cells are usually unwanted,
and cells possess strict safeguards against such changes2.
The crucial event that signifies the onset of malignancy
is thought to be a single genomic alteration, the outcome of which might be as subtle as a slight change in
the amount of a protein, or the substitution of a single
amino acid. The occurrence and fixation of such an
alteration signifies the failure of a mechanism that
should have detected the DNA lesion or mismatch that
caused the mutation and evoked the response that is
required to restore the original sequence and leave the
cellular life cycle unperturbed.
Sequence alterations in DNA arise from spontaneous chemical changes in DNA constituents, replication errors and damage inflicted on the DNA3. The
greatest challenge to genome stability comes from
DNA-damaging agents that can be either endogenous
(they form during normal cell metabolism) or exogenous (they come from the environment). Damaging
agents such as radiation and reactive chemicals are capable of inducing a plethora of DNA lesions. Some are
extremely cytotoxic if not repaired, whereas others are
mutagenic and can affect the production, structure and
function of cellular proteins, with consequences ranging
from malfunction of the cell to malignant transformation (FIG. 1). It is not surprising, therefore, that many
mutagens are also carcinogens, and that there is a high
correlation between their carcinogenic and mutagenic
potencies4.
How does the cell defend itself against this serious
existential threat? The basic cellular response is to
repair the damage, but the type and amount of damage
might overwhelm the survival response machinery to
the extent that programmed cell death (apoptosis) is
initiated instead (FIG. 1). The mechanism of this important choice between attempts at survival and programmed death is not entirely clear, but here we will
focus on the road taken by the cell when chances for
VOLUME 3 | MARCH 2003 | 1 5 5
REVIEWS
The DSB model of the DNA-damage response
Summary
• The DNA-damage response is crucial for cellular life and for avoiding neoplasia. It
occurs by rapidly transducing the damage signal to many cellular systems, such as the
DNA-repair machinery and the cell-cycle checkpoints.
• Double-strand breaks (DSBs) in the DNA — deadly DNA lesions — mobilize an
intricate signalling network by activating the ATM protein kinase, which, in turn,
orchestrates this network by phosphorylating one or more key proteins in each of its
branches.
• Other protein kinases related to ATM carry out similar functions in response to other
genotoxic stresses, and some of them collaborate with ATM in the DSB response.
• ATM deficiency leads to ataxia-telangiectasia (A-T), a genomic instability syndrome,
the hallmarks of which — neurodegeneration, immunodeficiency, radiation sensitivity
and cancer predisposition — show the intimate connection between maintenance of
genome stability, cellular and tissue functioning, and cancer prevention.
• Certain types of ATM mutations seem to increase cancer predisposition in heterozygous
carriers. This adds ATM to the list of genes that have sequence variations with important
implications for public health, especially with regards to cancer epidemiology.
survival are better — when the cell activates an intricate
network of pathways in order to resume its normal life
cycle, unharmed.
The DNA-damage response
CELL-CYCLE CHECKPOINTS
Regulatory mechanisms that do
not allow the initiation of a new
phase of the cell cycle before the
previous one is completed, or
temporarily arrest cell-cycle
progression in response to stress.
DNA damage activates specific
checkpoints at the G1–S and
G2–M boundaries and in the
S phase, with each one based on
a different mechanism.
156
DNA damage initiates a response that switches on various repair mechanisms that recognize specific DNA
lesions. These are repaired in situ or removed and the
DNA is restored to its original sequence3. A wealth of
information has accumulated on DNA-repair mechanisms over the past 50 years. These mechanisms are different for different types of DNA lesions. Genetic defects
that perturb these mechanisms almost invariably cause
severe syndromes that are characterized by the degeneration of specific tissues (especially the nervous and
immune system), sensitivity to specific DNA-damaging
agents and predisposition to cancer5–8 (BOX 1).
Researchers are becoming aware, however, that the DNAdamage response is considerably broader than DNA
repair, and actually encompasses additional processes9.
This is especially true for the response to highly cytotoxic
DNA lesions, such as double-strand breaks (DSBs)10–12. A
hallmark of this response is the activation of CELL-CYCLE
13,14
CHECKPOINTS
. The sudden arrest of the cell cycle must
involve marked alterations in numerous physiological
processes. Indeed, DNA damage leads to changes in
gene-expression profiles15–17, and probably also in protein
synthesis, degradation and trafficking. This means that
the DNA-damage alarm must be conveyed swiftly and
precisely to numerous pathways across the cell.
What is the signal emanating from a DNA lesion and
how does it reach so many destinations at the same time?
Answers were found by studying the sophisticated damage response to DSBs, which has been conserved from
yeast to mammals. This is a fine-tuned, multi-branched
network of signalling pathways that act together to repair
the deadly DSB, while temporarily arresting the cell cycle
and balancing cellular metabolism accordingly. It turns
out that this intricate network is mobilized primarily
through the action of a single protein kinase — ATM.
| MARCH 2003 | VOLUME 3
DSBs are naturally formed and sealed during processes
such as meiotic recombination and the assembly of the
T-cell receptor and immunoglobulin genes via V(D)J
recombination, in T cells and B cells, respectively. It is
safe to assume that cellular DSB repair mechanisms
(BOX 2) maintain continuous, low-level activity, ensuring
that the occurrence and resealing of these breaks leave
the cell unharmed. But when DSBs are inflicted on the
genome by damaging agents, such as free radicals or
ionizing radiation, their threat to cell life is sufficiently
serious to set in motion, within minutes, a rapidly
mounting, decisive DNA-damage response10–12.
Recent models64 depict the DSB response as developing through a series of steps (FIG. 2). According to these
models, DSBs might first be detected by sensor proteins
that recognize the DNA lesion itself or possibly chromatin alterations that follow DNA breakage. The broken
ends are then processed — their chemical nature is random, so they cannot serve directly as substrates for repair
mechanisms. Then, the transducers are brought into
action; these convey the damage signal to downstream
effectors. It is this relay system from transducers to effectors that enables a single transducer to quickly affect the
operation of many pathways. The transducers might also
be involved in the assembly of DNA-repair complexes at
the site of the damage (BOX 2), so DSB repair and signalling are probably concomitant and functionally linked
(FIG. 2). In the case of DSBs, the initial and primary transducer is ATM — although related protein kinases are also
involved — which transmits the message via a standard
signalling mode: protein phosphorylation.
The PI3K-related protein kinase family
ATM belongs to a conserved family of proteins, most of
which possess a serine/threonine kinase activity18–22.
Interestingly, all of these proteins contain a domain with
motifs that are typical of the lipid kinase phosphatidylinositol 3-kinase (PI3K) (FIG. 3), so they are dubbed
‘PI3K-like protein kinases’ (PIKKs). The PI3K domain
harbours the catalytic site of the active protein kinases of
the PIKK family. The mammalian members of this family, at present, include five protein kinases — ATM, ATR,
ATX/SMG-1, mTOR/FRAP and DNA-PKcs (FIG. 3) —
and TRRAP, which is a component of histone acetyltransferase complexes23,24. The active protein kinases in
the family — which are conserved from yeast to mammals — respond to various stresses by phosphorylating
key proteins in the corresponding response pathways.
They could therefore simultaneously affect numerous
processes, depending on the spectrum of their substrates. Four mammalian PIKKs are known to be
involved in the DNA-damage response: the DNAdependent protein kinase (DNA-PK), ATM, ATR and
ATX (REFS 17–21; and R. T. Abraham, personal communication). Whereas ATM and DNA-PK respond primarily
to DSBs, ATR and ATX respond to both ultraviolet
(UV) light damage (possibly UV-light-induced replication arrest) and DSBs, and ATR also responds to stalled
replication forks. mTOR/FRAP is the only active kinase
in this family that is not involved in responding to DNA
www.nature.com/reviews/cancer
REVIEWS
damage. It responds to nutrient levels and mitogenic
stimuli by signalling to the protein translation machinery and probably to protein degradation systems, and it
is also involved in controlling cellular growth25,26. ATM
is the prototype transducer of the DNA-damage signal.
ATM: at the top of a web of DSB signalling
Ataxia telangiectasia. The ATM protein was identified as
the product of the gene that is mutated (lost or inactivated) in the human genetic disorder ataxia-telangiectasia
(A-T)27. A-T belongs to a group of human diseases that
are collectively known as ‘genomic instability syndromes’,
each of which results from a defective response to a
specific DNA lesion (BOX 1). A-T is characterized by cerebellar degeneration, which leads to severe, progressive
neuromotor dysfunction, immunodeficiency, genomic
instability, thymic and gonadal atrophy, a striking predisposition to lymphoreticular malignancies and extreme
sensitivity to ionizing radiation and DSB-inducing
agents28,29. This devastating human disorder typically
combines most of the hallmarks of a defective DNAdamage response, clearly pointing to the DSB as the
DNA lesion that elicits this defect. Indeed, cultured cells
from A-T patients show a broad defect in responding to
DSBs that span almost all of the known branches of this
response19. The striking clinical and cellular phenotype
that is caused by ATM loss clearly places this protein at a
top position in the DSB-response cascade.
Damage
type
DNA
lesions
Radiations
Ultraviolet
radiation
Chemical agents
Endogenous Exogenous
(chemical
(oxygen
mutagens)
radicals)
Intrinsic
Spontaneous base
modifications,
replication errors
Base modifications,
abasic sites,
strand breaks
Base-pair
mismatches,
insertions, deletions,
strand breaks
Damage extent and severity
Amount and type of
damage can be handled
Cellular response
Activation of the
survival response network
Stress
responses
Consequence
Activation of ATM. ATM resides predominantly in the
nucleus in dividing cells, and responds swiftly and vigorously to DSBs by phosphorylating numerous substrates (see below). ATM-mediated phosphorylation
either enhances or represses the activity of its targets,
thereby affecting specific processes in which these proteins are involved. Similar to other active PIKKs (with
the exception of mTOR/FRAP), ATM targets serine or
threonine residues followed by glutamine (the ‘SQ/TQ’
motif)30,31. The hallmark of ATM’s response to DSBs is
a rapid increase in its kinase activity immediately following DSB formation32,33. Researchers have long been
impressed by the rapid phosphorylation of the many
ATM substrates, which converts them within minutes
to phosphorylated derivatives. A marked change in the
activity of ATM would account for this massive
process. Initial evidence indicated that ATM activation
might involve autophosphorylation34. A breakthrough
in our understanding of this process came in a landmark publication by Bakkenist and Kastan35. They
found that ATM molecules are inactive in undamaged
cells, being held as dimers or higher-order multimers.
In this configuration, the kinase domain of each molecule is blocked by the FAT domain (FIG. 3) of the other.
Following DNA damage, each ATM molecule phosphorylates the other on a serine residue at position
1981 within the FAT domain — a phosphorylation
that releases the two molecules from each other’s grip,
Cell-cycle
checkpoints
Cell survival
DNA adducts,
crosslinks,
strand breaks
Low fidelity
DNA repair
repair
Pyrimidine dimers,
base modifications
Ionizing
radiation
Abasic sites,
base modifications,
strand breaks
Damage is excessive
and/or irreparable
Activation of the
apoptotic pathway
Mutations,
chromosomal
aberrations
Malignant
transformation
Cell death
Figure 1 | Cellular responses to DNA damage. Different types of DNA damage cause different types of lesions, and these, in
turn, are handled by the cell in different ways. The outcome could be cell survival and resumption of the normal life cycle of the cell,
cell death or malignant transformation. The mechanism of choice between attempt at survival and programmed cell death is not
completely understood. The survival response is elaborate and encompasses many signalling pathways.
NATURE REVIEWS | C ANCER
VOLUME 3 | MARCH 2003 | 1 5 7
REVIEWS
turning them into fully active monomers. Within minutes after the infliction of as few as several DSBs per
genome, most ATM molecules become vigorously
active. Bakkenist and Kastan35 provide evidence that
the signal for ATM activation might be chromatin
alterations rather than direct contact of ATM with the
broken DNA. However, it has previously been shown
that, soon after damage infliction, a portion of the
Box 1 | Genomic instability syndromes
Genetic defects that affect specific DNA damage response pathways lead to syndromes that combine various degrees
of tissue degeneration, growth and developmental retardation, premature signs of ageing, chromosomal instability,
sensitivity to the corresponding DNA-damaging agents and cancer predisposition. The prominent genomic
instability syndromes3,5–8 listed below are all autosomal recessive and represent defects in the main damage-response
pathways, each of which is activated by a different class of damaging agent. Three diseases — xeroderma
pigmentosum, Cockayne’s syndrome and trichothiodystrophy — are associated with defects in the nucleotideexcision-repair (NER) pathway that deals with bulky, helix-distorting DNA lesions such as those inflicted by
ultraviolet-light (UV) radiation and certain chemicals.
Xeroderma pigmentosum (XP)
XP patients are extremely sensitive to the UV-light component of sunlight and show accelerated ageing of the skin and a
striking tendency to skin cancers. XP is genetically heterogeneous. Its classical form entails defects in various proteins
that act together in the nucleotide-excision-repair (NER) pathway. NER has a global repair arm and a transcriptioncoupled repair arm, and one or both might be defective in different XP patients. Some XP proteins are involved in DNA
repair and are also part of the transcription complex TFIIH, and one of them is a DNA polymerase that is responsible
for trans-lesion DNA synthesis.
Cockayne’s syndrome
This UV-light sensitivity syndrome is characterized by striking dwarfism and skeletal malformations, severe mental
retardation, deafness and photosensitivity, but no apparent cancer predisposition. The defective proteins are involved in
transcription-coupled repair of DNA damage that is induced by UV light and probably oxidative DNA damage.
Trichothiodystrophy (TTD)
This rare disease is characterized by brittle hair, dry skin, dysmorphic face, mental retardation and a certain degree of
photosensitivity. The mutations that are responsible for this occur in a specific region of the XPB and XPD genes that
encode DNA helicases involved in transcription — other mutations in these genes cause XP or Cockayne’s syndrome.
This is an interesting example of different mutations within the same gene leading to different phenotypic outcomes.
Bloom’s syndrome (BS)
BS patients have short stature, sun sensitivity, facial erythema, immunonodeficiency, decreased fertility and a
predisposition to various cancers. BS cells show striking elevation of spontaneous sister-chromatid exchanges,
chromosomal breakage and mild sensitivity to various DNA-damaging agents. The defective protein, BLM, is a DNA
helicase that resembles the bacterial recQ protein, which probably participates in homologous recombination repair and
repair of damage that occurs at stalled replication forks.
Werner’s syndrome (WS)
WS is characterized primarily by accelerated ageing, and also by a tendency for diabetes, delayed sexual
development, chromosomal instability and a predisposition to various cancers. WS cells show chromosomal
instability and are hypersensitive to camptothecin — an inhibitor of type I DNA topoisomerase — and certain
DNA alkylating and crosslinking agents. Of note, both BS and WS cells are hypersensitive to hydroxyurea-induced
apoptotic killing. The defective protein — WRN — is another DNA helicase of the recQ-like family with a unique
3′→5′ exonuclease activity.
Rothmund–Thompson syndrome (RTS)
RTS patients are recognized by their patchy skin coloration, stunted growth, skeletal abnormalities, early cataracts,
accelerated ageing, chromosomal instability and cancer predisposition. The culprit gene — RECQL4 — encodes yet
another recQ-like DNA helicase, the role of which in the DNA-damage response is unclear.
Fanconi’s anemia (FA)
The main characteristics of FA are bone-marrow depletion, which leads to aplastic anaemia (insufficient formation of all
blood cell types), skeletal malformations, deformities of various internal organs, reduced fertility, cutaneous
abnormalities and marked predisposition to myeloid leukaemias and squamous-cell carcinomas. FA cells show
characteristic chromosomal aberrations and sensitivity to DNA crosslinking and DNA-breaking agents. The disease is
genetically heterogeneous, with mutations affecting eight different ‘FA genes’, the products of which interact with each
other and with various components of the DNA-damage response systems, such as BRCA1 and ATM (see text and recent
review by D’Andrea and Grompe167). Recently, it was found that homozygosity for certain BRCA2 mutations is
responsible for one form of FA168.
Others
Additional genome instability syndromes are ataxia-telangiectasia (A-T), Nijmegen breakage syndrome (NBS) and
ataxia-telangiectasia-like disease (ATLD); these are discussed in detail in the main text.
158
| MARCH 2003 | VOLUME 3
www.nature.com/reviews/cancer
REVIEWS
nuclear ATM is recruited to DSB sites and strongly
adheres to them, possibly serving as a platform for further enzymatic reactions that take place at those sites36.
Both chromatin-bound and free ATM are autophosphorylated on serine 1981 (L. Moyal, unpublished
observations). So, following DSB induction, activated
ATM seems to divide between two fractions: one is
chromatin bound and the other is free to move
throughout the nucleus.
ATM substrates. The list of published ATM substrates
— more than a dozen at this time — is far from complete, and many ATM-dependent responses are likely to
involve ATM targets that are unknown at present.
However, the study of these pathways is gradually disclosing a remarkably broad cellular response to DSBs
that is meticulously orchestrated by ATM.
A close look at the network of ATM-mediated
pathways that is responsible for activation of the
Box 2 | Repair of DNA double-strand breaks
NATURE REVIEWS | C ANCER
Non-homologous
end-joining (NHEJ)
Homologous
recombination (HR)
Sister
chromatids
DSB
KU80
KU70
DNA-PKcs
?
RAD50
DSB
DNA-PKcs
The repair of DNA double-strand breaks10,11,49,50,64
(DSBs) is carried out by two different mechanisms: a
rapid, error-prone mechanism dubbed ‘nonhomologous end-joining’ (NHEJ) that quickly seals the
breaks at the expense of creating local microdeletions,
and a high-fidelity repair process based on homologous
recombination (HR) between sister chromatids. These
highly structured processes stem from the concerted
action of several multi-protein complexes.
The predominant repair mode in mammalian cells is
NHEJ. The exposed ends of the DNA strands are
detected by the KU70–KU80 heterodimer that recruits
the catalytic subunit of the DNA-dependent protein
kinase (DNA-PKcs). DNA-PKcs, in turn, might recruit
additional proteins to the damaged site and probably
phosphorylates some of them. The XRCC4–ligase IV
heterodimer finally seals the breaks. Despite the
disadvantage of its low fidelity, this pathway can act
quickly, as required of an emergency mechanism, and,
unlike HR, it does not depend on sister DNA molecules,
which exist in the cells only after DNA replication. NHEJ
is also important in sealing the breaks formed during
V(D)J recombination, in which the T-cell receptor and
immunoglobulin genes are rearranged169.
The high-fidelity, HR-based pathway is mediated by
the RAD51-associated proteins that include several
RAD51 paralogues, RAD52, the RAD54 helicase and the
BRCA2 tumour suppressor. Following initial end
resection and binding of RAD52 to the single-stranded
end, RAD51 forms a nucleoprotein filament on the
exposed strand. This process is probably initiated by a
RAD51–BRCA2 complex170,171 and is essential for the
main step in this pathway: strand invasion and strand
displacement. This is mediated by RAD51 and RAD54,
and allows use of the undamaged sister molecule as a
template for the resynthesis of the missing portions in
the broken molecule.
The MRE11–RAD50–NBS1 (MRN) complex
(reviewed by D’Amours and Jackson64) seems to be
essential for the HR pathway172, probably by carrying
out the initial processing of the DSB ends, and is also
involved in meiotic recombination, telomere
maintenance and checkpoint signalling. Following DSB
induction, the MRN complex rapidly forms prominent
foci at the damaged sites. These foci include additional
players in the DSB response, such as RAD51 and
BRCA1. MRE11 is a nuclease; RAD50 is an ATPase; and
the NBS1 protein, which is essential for the assembly of
the complex, interacts with histone H2AX173 and is
further involved specifically in the HR pathway172.
MRE11
NBS1
Nucleasemediated
resection
RAD52
KU80
KU70
XRCC4
BRCA1
RAD51–BRCA2
RAD54
Ligase IV
Strand
BRCA2
invasion
RAD51
Inaccurately
repaired DNA
DNA polymerase
DNA ligase
DNA synthesis;
branch migration
Ligation;
junction resolution
Accurately repaired DNA
VOLUME 3 | MARCH 2003 | 1 5 9
REVIEWS
Chromatin
alterations
Double-strand
break (DSB)
Sensors —
damage detection
Processing enzymes
(e.g. nucleases) —
initial processing
(e.g. end resection)
Processed
DSB
Transducers
Activation
Activated
transducers
Effectors
DNA
repair
Repair enzymes,
assembly of
repair complexes
Repaired
DNA
Cell-cycle
checkpoints
Stressresponse
pathways
Cell survival
Figure 2 | Basic steps in a current model of the cellular
survival response to DSBs. The damage sensors might
sense the DNA lesions themselves and/or alterations in
chromatin structure in their vicinity. The activation of the
transducers is probably mediated by post-translational
modifications (see text).
cell-cycle checkpoints (FIG. 4) explains how ATM precisely and decisively controls these pathways using several sophisticated strategies. The first strategy is to
approach the same effector from several different
directions. A main component in the G1–S cell-cycle
checkpoint is mediated by activation and stabilization
of p53, which, in turn, activates transcription of the
gene that encodes the CDK2–cyclin-E inhibitor WAF1
(also known as p21 and CIP1). The main target of
ATM in this pathway is p53, which is phosphorylated
by ATM on Ser15 (REFS 32,33,37). This contributes primarily to enhancing the activity of p53 as a transcription factor38–40. ATM also phosphorylates and activates
CHK2, a checkpoint kinase41,42 that phosphorylates
p53 on Ser20. This interferes with the p53–MDM2
interaction. The oncogenic protein MDM2 is both a
direct and indirect inhibitor of p53, as it serves as a
ubiquitin ligase in p53 ubiquitylation, which mediates
its proteasome-mediated degradation43,44. ATM also
directly phosphorylates MDM2 on Ser395, which
interferes with nuclear export of the p53–MDM2
complex, and hence the degradation of p53 (REFS 38,45).
Finally, it has been reported that phosphorylations of
160
| MARCH 2003 | VOLUME 3
p53 on Ser9 and Ser46 (REF. 46), and dephosphorylation
of Ser376 (REF. 47), are ATM dependent as well,
although the function of these changes is unknown.
This series of ATM-dependent modifications that
activate and stabilize p53 — although perhaps
not complete — illustrates the elaborate way in which
ATM handles a single effector, and indicates that ATM
might regulate several effectors within the same pathway. This principle is also seen in ATM-mediated
activation of the BRCA1 tumour-suppressor protein
following DNA damage. BRCA1 was found to be associated with large protein complexes that contain DSBrepair and mismatch-repair enzymes as well as ATM48,
but its mode of action in DSB repair is unclear at
present. It also activates the expression of certain damage-responsive genes and is involved in the S-phase and
G2–M checkpoints49–52. ATM phosphorylates BRCA1
on several sites53–55. Importantly, whereas ATMmediated phosphorylation of BRCA1 on Ser1387
activates BRCA1 as a regulator of the intra-S-phase
checkpoint56, its phosphorylation by ATM on Ser1423
spurs its involvement in the G2–M checkpoint57. So,
phosphorylation on different sites might direct an
effector to act in different pathways. The CHK2 kinase,
which is activated by ATM, adds yet another phosphorylation on the BRCA1 molecule58, while at the same
time ATM phosphorylates CtIP — an inhibitor of
BRCA1 — on two residues, which inhibits its function
and further stimulates BRCA1 (REF. 59).
ATM’s hold on downstream pathways is shown not
only by its multi-pronged grip of specific pathways, but
also by its ability to approach the same end point from
several different directions. A notable example is, again,
the intra-S checkpoint: several ATM-controlled pathways converge to control this process, as it is one of the
most crucial cellular responses to DSBs (FIG. 4). At least
five ATM-mediated pathways seem, at present, to be
involved in this checkpoint. In addition to BRCA1, ATM
phosphorylates NBS1 (REFS 60–63), a component of the
multi-functional MRE11–RAD50–NBS1 (MRN) complex (see BOX 2) that is involved in DSB repair64. Of several ATM phosphorylation sites on NBS1, Ser343 and
Ser278 seem to be particularly important for its as yet
unknown role in this checkpoint 60,61. Another ATM
substrate in the intra-S checkpoint pathway is the SMC1
(structural maintenance of chromosomes 1) protein.
This protein — known primarily for its involvement in
sister-chromatid cohesion — is phosphorylated by
ATM on two serine residues, and interference with this
phosphorylation abrogates the S-phase checkpoint 65,66.
A recently identified effector of ATM in the intra-S
checkpoint is the FANCD2 protein, which is phosphorylated by ATM on Ser222 following DSB induction and
undergoes BRCA1-mediated mono-ubiquitylation67.
FANCD2 is a member of a multi-protein complex,
defects of which lead to another genomic instability
syndrome — Fanconi’s anaemia (BOX 1). At the same
time, CHK2 and another ATM/ATR-activated kinase,
CHK1, take care of several checkpoints. Both of them
phosphorylate the checkpoint phosphatase CDC25A,
which marks it for degradation68–72. CDC25A’s duty is to
www.nature.com/reviews/cancer
REVIEWS
dephosphorylate — and hence maintain the activity of
— the cyclin-dependent kinases CDK2 and CDK1.
CDK2 drives both the G1–S transition and S phase, and
CDK1 mobilizes the G2 phase onto mitosis. So, the
destruction of CDC25A contributes to the G1–S, intra-S
and G2–M checkpoints (FIG. 4). Of interest is that the
CDC25A-mediated component of the G1–S checkpoint
is rapid and, unlike the p53-mediated component, is not
dependent on gene activation and protein synthesis70.
Notably, in addition to ATM’s versatility as a protein
kinase with numerous substrates, the ATM web contains protein kinases that are themselves capable of targeting several downstream effectors simultaneously, and
so concomitantly control subsets of pathways. CHK2,
for example, is known to phosphorylate p53, BRCA1,
CDC25A and CDC25C41,42 (FIG. 4). It has been proposed
that the CHK2–CDC25C pathway is similar to the previously mentioned CHK2–CDC25A pathway, but
might act instead at the G2–M transition. Here, phosphorylation of CDC25C by activated CHK2 is thought
to lead to cytoplasmic sequestration of CDC25C, which
prevents activation of CDK1. Noteworthy, the CHEK2
gene that encodes this protein was suggested to act as a
tumour suppressor in subsets of patients with the cancer-predisposing Li–Fraumeni syndrome73, but this
contention has not been substantiated by other
reports74. Germline CHEK2 mutations have also been
implicated in familial breast cancer75,76, and CHEK2
shows somatic mutations in various sporadic solid
tumours (reviewed in REFS 38,39).
Most of these pathways have not been completely
characterized, and the involvement of ATM substrates in them has been inferred simply from defective
activation of specific checkpoints following abrogation of
1
3056
ATM
1
2644
ATR
1
3031
ATX/SMG1
1
4128
DNA-PKcs
1
2549
mTOR/FRAP
1
3830
TRRAP
FAT
PI3K
FATC
Figure 3 | Size and common motifs in the human members
of the PIKK family. The number of residues is indicated for
each protein. This family comprises six proteins, all of which
(except for TRRAP) possess protein kinase activity. These
proteins share three motifs: the FAT and FATC domains are of
unknown functional significance, but the FAT domain of ATM
contains serine 1981 — the site that is autophosphorylated
during ATM activation (see text). The PI3K domain, which
contains the phosphatidylinositol 3-kinase motifs, harbours the
catalytic site in the active kinases of the family.
NATURE REVIEWS | C ANCER
ATM-mediated phosphorylation of these proteins. An
interesting pathway is mediated by phosphorylation of
RAD17 — the human orthologue of Rad17 from the fission yeast, Schizosaccharomyces pombe. RAD17 is phosphorylated in an ATM/ATR-mediated manner on two
serine residues, and this phosphorylation was reported by
different laboratories to be important for proper G1–S
and G2–M checkpoints77,78. This protein is involved in a
very early event in the DSB response: it loads a trimolecular complex onto the DNA that consists of the human
orthologues of the fission yeast proteins Rad9, Hus1 and
Rad1 (‘the 9-1-1 complex’). This complex acts as a sliding
clamp, probably a damage sensor79–82. Although this
process might be regarded as being upstream of the
transducer recruitment, two essential proteins in this
process — RAD17 and RAD9 — are also phosphorylated
in an ATM/ATR-dependent manner, and seem to be
involved in downstream checkpoint pathways77,78,83 (FIG. 4).
It is possible that such proteins have a dual role in
processes upstream and downstream of ATM. Such is the
case with NBS1 as well: on one hand, it is a component of
the MRN complex that is thought to be involved in the
initial processing of the DSB, and, on the other hand, it is
a downstream effector of ATM in a checkpoint pathway.
Upstream and downstream mingle. Indeed, there is a
growing notion that the DSB signal might be initially
amplified by means of a cyclic process rather than by a
series of steps with a linear hierarchy. One of the first
processes that is initiated by DSBs is the massive phosphorylation of the tail of a histone protein variant
called H2AX (reviewed by Redon et al.84). Foci of phosphorylated H2AX are rapidly formed at the DSB sites
and are thought to be essential for further recruitment
of repair factors, such as the MRN complex, RAD51
and BRCA1 (REF. 85). H2AX phosphorylation — a very
early event in the cascade induced by DSBs — was
reported to be ATM dependent following DSB induction86, and ATR dependent following replicative stress87.
This process could therefore serve as a rapid and powerful mechanism for amplifying the damage signal via
repeated cycles of H2AX phosphorylation, and recruitment of processing factors. These facilitate further
recruitment of damage transducers to the damaged
sites, along with repair proteins.
Another process of signal amplification might characterize the phosphorylation of certain ATM substrates.
The NBS1 protein, itself an ATM substrate, seems to
serve as an adaptor in the phosphorylation of other
ATM substrates, such as CHK2 (REFS 88,89) and SMC1
(REFS 65,66), particularly at low damage levels. So, in this
interesting mechanism, the phosphorylation of one
substrate is required for the phosphorylation of others.
This is in addition to the role of NBS1 in the MRN complex that acts as an early damage processor 64. Another
protein that seems to facilitate the phosphorylation of
certain ATM substrates is 53BP1. Like NBS1, 53BP1
contains a BRCT (breast cancer carboxyl terminus)
domain, is recruited to DSB-induced nuclear foci, is
required for proper activation of certain cell-cycle
checkpoints, and is itself an ATM substrate90–92. Yet
VOLUME 3 | MARCH 2003 | 1 6 1
REVIEWS
another newly discovered, BRCT-containing protein,
NFBD1, might have similar characteristics93–95. The
emerging complex relationships between ATM,
these three and other ATM substrates are drawing new
flow charts for the DNA-damage signal that deviate
from the traditional linear ones and assign to several
proteins more than one role ‘upstream’ or ‘downstream’
in this chart.
and functioning of the RAD51-associated protein
complexes in the HR arm of DSB repair96,97 (BOX 2).
Activation of this pathway has been suggested to be associated with phosphorylation of the RAD51 and RAD52
proteins by the ABL tyrosine kinase96,98,99, which
physically interacts with ATM and is activated in an
ATM-dependent manner following DSB induction100–102.
The essential rapidity of the damage response on
one hand, and the numerous functional interactions it
entails on the other, indicate that the relevant proteins
constantly remain in close physical proximity. Indeed,
a large complex containing BRCA1, ATM and numerous repair proteins was identified in human cells48.
The physical interactions within such complexes are
ATM — a direct activator of repair? ATM’s command
over the damage response raises the obvious question as
to whether ATM directly controls the repair process itself
(BOX 2). A-T cells do show a subtle but distinct defect in
DSB repair, which was ascribed to suboptimal assembly
M
635
RAD17
P
P
645
272
(ATR)
G2
P RAD9
317
CDK1
DeP
CDC25C
CHK1
P
216
Ser
P
Additional
PTMs
(ATR)
P
ATM
(ATR)
345
15
P
1423
1423
P
20
P
BRCA1
P
P
1387
p53
G1
Thr 68
P
CHK2
988
395 P
MDM2
Ser
1387
745
WAF1
123
P
P
CDC25A
CtlP
P
966 957
664
P P
SMC1
DeP
222 P
CDK2
FANCD2
343 P
P 278
NBS1
S
Figure 4 | ATM-mediated activation of cell-cycle checkpoints in response to DSBs. Arrows indicate stimulation; T-shaped
lines mark inhibition. Inhibitory phosphorylations are marked by a line through the arrow. Dephosphorylation is marked by ‘DeP’. The
positions of the phosphorylated residues are indicated by their number. In most cases, the phosphorylated residues are serines,
except for ATM-mediated phosphorylation of CHK2, which targets threonine 68. Most of these pathways are regulated primarily by
ATM, particularly in the early stage of the DNA-damage response, with ATR probably becoming important at later stages to maintain
these pathways. However, the pathways mediated via RAD17 and CHK1 phosphorylation are primarily under ATR control. In some
of these pathways, the mechanism by which the cell-cycle machine is perturbed is not clear, but ATM-mediated phosphorylation of
the indicated proteins is important for these processes. Note the different BRCA1 phosphorylations directing this protein to act at
different checkpoints.
162
| MARCH 2003 | VOLUME 3
www.nature.com/reviews/cancer
REVIEWS
probably dynamic and respond rapidly to DNA damage. A recent example is the interaction between two
players in the damage response — BRCA1 and ABL
— which is quickly disrupted following DNA damage
in an ATM-dependent manner103.
ATM and gene transcription. Many damage responses
end up by modulating gene expression. Indeed, one of
the main ATM targets — p53 — is a transcription factor. Another transcription factor that has a central role
in cell-cycle control — E2F1 — has been reported to be
phosphorylated and stabilized in an ATM-dependent
manner104. In addition, it is becoming evident that
stress-response pathways that are best known for
responding to other triggers in an ATM-independent
manner, such as those mediated by the mitogen-activated protein kinases or the transcription factor NF-κB,
also contain a DNA-damage-responsive component14.
Interestingly, when these pathways are activated by
DSBs, their response becomes ATM dependent105–107.
The direct targets of ATM in these pathways have not
been elucidated, but these observations considerably
expand the ATM-dependent network.
ATM and damage-independent genomic stability. ATM
helps maintain genomic stability by mechanisms other
than responding to DSBs caused by damaging agents.
A-T cells have abnormally shortened telomeres
and show abnormal association of chromosome ends,
telomere clustering and altered interactions between the
telomeres and the nuclear matrix (reviewed by
Pandita108). Evidence is accumulating that ATM is functionally linked to the maintenance of telomere length
and integrity, a process that is crucial to ageing and cancer108–111. Here, ATM might be responding continuously
to an ongoing process rather than abruptly and vigorously to an acute insult112. The functional relationships
between ATM and the telomere-maintenance machinery
have recently been illuminated by a study that generated
mice that are doubly null for Atm and the telomerase
RNA component (Terc)113. These animals showed
increased genomic instability, enhanced ageing and premature death, with a general proliferation defect extending into stem- and progenitor-cell compartments.
Interestingly, the rate of T-cell malignancies in these mice
was reduced rather than enhanced.
A substrate of ATM-dependent phosphorylation
in the telomere maintenance system is TRF1. TRF1
negatively regulates telomere elongation and its phosphorylation seems to suppress TRF1-mediated apoptosis following DNA damage114. Interestingly, inhibition of TRF1
in A-T cells rescues telomere shortening and decreases
the radiosensitivity of these cells, providing further evidence of the link between telomere metabolism and the
DNA-damage response115.
Further evidence of ongoing activity of ATM at sites
of normally occurring breakage and reunion of DNA
was provided by its presence at sites of V(D)J recombination116. p53 phosphorylated at the ATM target
residue was also present at those sites, indicating that
ATM continuously surveys the V(D)J recombination
NATURE REVIEWS | C ANCER
process, and can induce a damage response if DSBs
are not sealed in a timely manner. Indeed, A-T
patients often show clonal translocations involving
the sites of the T-cell receptor and immunoglobulin
genes, showing the consequences of absence of this
surveillance. These translocations often herald the
onset of lymphoid malignancy117.
Cross-talk among genome instability syndromes
Defective responses to specific types of DNA lesions
underlie the genome instability syndromes (BOX 1). Not
surprisingly, the proteins that are defective in clinically
similar syndromes were found to maintain functional
interactions. Such is the case for A-T and disorders that
are caused by deficiencies of the different members of the
MRN complex. As mentioned above, the NBS1 protein is
an ATM substrate in the intra-S checkpoint pathway
(FIG. 4), and the gene that encodes it is mutated in
patients with Nijmegen breakage syndrome (NBS)118.
NBS is characterized by immunodeficiency, small head,
mental deficiency, genomic instability, radiation sensitivity and acute predisposition to lymphoid malignancies.
The NBS phenotype shows significant overlap with that
of A-T, with the important exception that NBS patients
do not have cerebellar degeneration. Another member of
the MRN complex — RAD50 — was recently found
to be mutated in a patient with an NBS phenotype
(T. Doerk, personal communication). Importantly, however, patients with defective MRE11 show yet another
phenotype that is even closer to that of A-T. This disorder
was therefore called ‘A-T-like disease’ (ATLD)119. ATLD
patients develop most of the hallmarks of A-T, albeit
onset at a later age and with slower progression. The different phenotypes that are associated with deficiencies in
NBS1 and MRE11 might indicate that these two proteins
have specific roles in the DNA-damage-response pathway, with different functional links to ATM. Furthermore,
the striking similarity between A-T and ATLD raises the
possibility that MRE11 deficiency might somehow affect
ATM activation — and hence cause defective phosphorylation of ATM substrates. If this contention proves to be
true, it will place the MRN complex upstream of ATM in
the DNA-damage response, in addition to NBS1’s role
downstream of ATM as an ATM target.
Less expected have been the functional links that were
recently reported between the ATM and the MRN complex, and the Fanconi’s anaemia (FA) proteins. The
ATM–FANCD2 connection (FIG. 4) links the DSBresponse network to the FA complex, known primarily
for its involvement in the response to DNA interstrand
crosslinks120,167. Conversely, the MRN complex — known
for its involvement in DSB repair — has recently been
linked to the DNA crosslinks response, as cells from NBS
patients are hypersensitive not only to DSB-inducing
agents, but also to chemicals that induce interstrand
crosslinks121. Furthermore, in response to this damage,
the MRN complex forms distinct nuclear foci that are
similar to those found in the radiation response, but that
also contain some of the FA proteins121,122. Finally, ATM
was recently reported to phosphorylate the BLM protein,
a DNA helicase that is defective in Bloom’s syndrome123
VOLUME 3 | MARCH 2003 | 1 6 3
REVIEWS
(BOX 1). Another link between BLM and the PIKK family
was indicated by findings that the phosphorylation of
BLM following replication arrest is dependent on
ATR124. Furthermore, the same study also provided evidence that the relocalization of the MRN complex at
sites of replication arrest requires functional BLM. These
examples highlight functional links between apparently
separate signalling pathways that might finally converge
to a unified, global DNA-damage response.
Other PIKK-family members
ATR. ATM-mediated pathways mainly depend on ATM
during the initial, rapid phase of the damage response,
which lasts for 1–2 hours. These pathways are therefore
not completely abolished in A-T cells, but rather are
attenuated and are turned on slowly later. Such a pattern indicates that other protein kinases act with ATM,
but at different kinetics and with different vigour, and
their activity is uncovered in the absence of ATM. Not
surprisingly, one of these protein kinases is thought to
be ATR, which shares several substrates with ATM20.
Suppression of endogenous ATR activity using overexpression of recombinant, inactive ATR indicated that
ATR is responsible for the phosphorylation of certain
ATM substrates at late stages following damage induction125,126. So, in the DSB response, ATM reacts first,
being responsible for the immediate, rapid phase of the
response, whereas ATR joins in later and maintains the
phosphorylated state of specific substrates. This important redundancy adds further finesse and complexity to
the fine-tuned DSB response.
Taking care of the late stage of the DSB response is
not, however, the main role of ATR. It also responds to
UV-light treatment, stalled replication forks and
hypoxia18,20,127,128 (which do not activate ATM) by phosphorylating at least a subset of ATM substrates, among
them p53 and BRCA1 (REFS 125,129). Indeed, ATR was
recently shown to preferentially bind to UV-light-damaged DNA130. ATR therefore links the ATM-mediated
web to signals other than DSBs. Not all downstream
effectors are shared equally by ATM and ATR. For example, whereas CHK2 is a preferred ATM substrate in the
cell-cycle-checkpoint pathways, CHK1 is a preferred ATR
target in the G2–M checkpoint pathway131–133 (FIG. 4). This
could have significant downstream consequences — for
example, if ATM, but not ATR, activates CHK2, p53
might not be stabilized by phosphorylation of Ser20 in
cells lacking ATM, despite the presence of ATR. Indeed,
recent work has shown that ATR does not have a significant influence on p53 stabilization and the activation of
the G1–S checkpoint. On the other hand, ATR is important for preventing premature chromatin condensation
following DNA damage134. This observation highlights
the functional distinction between ATM- and ATRdependent pathways and indicates that the relationship
between them might be less straightforward than merely
serving the same purpose at a different time.
Another important ATR substrate is the previously
mentioned RAD17 (REF. 126) (FIG. 4). ATR-mediated
phosphorylation of RAD17 and CHK1 is dependent
on the ‘9-1-1’ complex member HUS1 (REFS 126,135),
164
| MARCH 2003 | VOLUME 3
although recruitment of the 9-1-1 complex to chromatin does not depend on ATR or on the phosphorylation by ATR, of RAD17, which is the loader of the
9-1-1 clamp onto chromatin126. The binding of ATR
and the 9-1-1 complex therefore seem to be independent of each other, but once they are bound to the
damaged site, the 9-1-1 complex might act to enable
ATR to recognize its substrates on the chromatin.
Recent work based on analysis in Xenopus egg extracts
indicates that ATR might be involved in the important
replication checkpoint that continuously ensures that
mitosis is not initiated before replication is completed136.
Involvement in such a crucial process might explain why,
unlike ATM, loss of ATR is embryonic-lethal in mice137,138
and is not compatible with cellular viability in culture139.
Interestingly, in Xenopus extracts, the activation of the
ATR-dependent damage checkpoint requires initiation of
DNA replication140. Further support for the importance
of ATR in the response to replicative stress came recently
from a study showing that deficiency of ATR, but not of
ATM, markedly destabilizes common chromosomal fragile sites, the fragility of which are usually expressed under
such stress141.
Another important characteristic of ATR is its need
for an accessory protein, ATRIP (ATR-interacting
protein). ATR and ATRIP maintain physical interaction
and depend on each other for their stability, and ATRIP
is one of the immediate substrates of ATR following its
activation139. This system is conserved through evolution, as ATRIP is the functional homologue of the fission yeast Rad26 protein, which interacts with the ATR
orthologue Rad3 (REF. 142). In budding yeast, the ATR
orthologue, Mec1p, needs the ATRIP equivalent,
Ddc2/Pie1/Lcd1, for its recruitment to the damaged
sites on the DNA and its subsequent activation143–147.
ATX/SMG1. The ATX/SMG1 protein is the most recent
addition to the PIKK family. Interestingly, the first
physiological process that was found to involve this
protein was not the DNA-damage response, but mRNA
surveillance, also known as nonsense-mediated mRNA
decay (NMD). NMD is a conserved surveillance mechanism that eliminates mRNA species that contain premature termination codons, thereby preventing the
production of truncated proteins148. The human SMG1
protein is an important component of this system — it
phosphorylates another key player in NMD, the
hUpf1/SMG2 protein149,150. Unexpected of a protein
associated with NMD, but perhaps not surprising given
its PI3K motif, human ATX/SMG1 is also involved in
the DNA-damage response. Recent work from R.T.
Abraham’s laboratory (personal communication) indicates that ATX/SMG1 also phosphorylates Ser15 of
p53, in response to UV light and ionizing radiation
treatment. Deficiency of this protein causes a rapid
decline in cell viability even in the absence of extrinsic
genotoxic stress. These exciting findings raise important questions about the nature of the functional link
between NMD and the DNA-damage response, and the
degree of overlap/collaboration between this newcomer
to the PIKK family and the veteran members.
www.nature.com/reviews/cancer
REVIEWS
DNA-PK. DNA-PK had been recognized as a protein
kinase that was activated by DSBs long before the other
PIKK-family members were identified. Despite its
apparently narrow spectrum of substrates, investigation
of this robust protein kinase has helped to shape the
concept of protein kinases as damage transducers.
DNA-PK is part of the non-homologous end-joining
arm of DSB repair151 (BOX 1). The DNA-PK holoenzyme
contains a large (450-kDa) catalytic subunit (DNAPKcs) bearing the PI3K motif, and two smaller accessory proteins, KU70 and KU80, which form a heterodimer. The KU heterodimer is thought to be a sensor
of DSBs that adheres to the damaged site and recruits
the catalytic subunit. DNA-PKcs, in turn, binds and
brings together the broken ends152, then undergoes
autophosphorylation on several serine residues that are
essential for its activation153,154.
Reduced activity of DNA-PKcs in mice leads to a
profound defect of the immune system called ‘severe
combined immunodeficiency’ (SCID), which is coupled
with extreme hypersensitivity to ionizing radiation151.
Complete loss of DNA-PKcs increases the risk of developing lymphomas, highlighting the strong link between
DSB repair, maintenance of genomic integrity, the
V(D)J recombination process and cancer formation.
The identification of DNA-PK’s downstream substrates is clearly important for understanding how it
transduces the DNA-damage signal, but, unfortunately,
the plethora of in vitro substrates of DNA-PK does not
represent its in vivo targets. Two proteins — the WRN
DNA helicase (BOX 1) and Artemis — were, however,
recently reported to be phosphorylated in cells in a
DNA-PK-dependent manner. Phosphorylation affects
WRN’s helicase activity, which might act in DSB
repair155,156. Artemis normally possesses a singlestranded 5′→3′ exonuclease activity, but following its
physical interaction with DNA-PK and its DNA-PKmediated phosphorylation, it becomes capable of opening the hairpin structures that are typically formed at
the ends of DSBs in the course of V(D)J recombination157. Loss of Artemis in humans causes SCID and
radiation sensitivity158. This pathway provides an elegant
example of a signalling process that is initiated by a
DNA lesion and ends up by activating an enzyme that is
involved in the direct processing of that lesion.
ATM mutations and cancer
Like the other genomic instability syndromes, A-T is a
cancer-predisposing disorder. Atm-deficient mice show
a striking predisposition to lymphoid malignancies,
particularly thymic lymphomas, to which they succumb
before the age of 1 year (reviewed in REFS 17–19).
However, much of the literature on ATM mutations
and cancer is not about A-T patients, but is, instead, on
heterozygous carriers of A-T mutations. For more than
two decades, ATM has been of interest to cancer epidemiologists and the public because of observations of cancer
predisposition among A-T carriers. These studies pointed
to a high incidence of malignancies, particularly breast
cancer, among unaffected members of A-T families
(reviewed by Khanna159). In view of the estimated 1–2%
NATURE REVIEWS | C ANCER
frequency of A-T carriers in the general population, this
observation has important implications for public health.
For example, should ATM mutations be regarded as a
genetic marker for cancer predisposition? Should people
that might potentially be exposed more than others to
radiation be screened for ATM mutations?
Intriguingly, epidemiological studies that were carried out in different communities to confirm this important observation have often produced conflicting results,
and the extent of cancer predisposition among A-T carriers seemed to be low at best. Important findings have
finally emerged that seem to help resolve this apparent
conflict. A-T mutations lead, in most cases, to truncated,
unstable protein products, and these ATM alleles therefore fail to produce any ATM at all. Carriers of such
mutations have a reduced amount of otherwise functional ATM. Other A-T mutations lead to amino-acid
substitutions (missense mutations) or in-frame deletions that produce a catalytically inactive protein. Cells
of carriers of these mutations are expected to contain
both functional and inactive ATM molecules in various
ratios. A meticulous study of the type of ATM mutations in A-T families with high incidence of cancers disclosed a high frequency of missense mutations160. The
importance of missense mutations in predisposing
carriers to cancer can be explained simply by the dominant-negative effect of the inactive version of the protein leading to a reduction in ATM function that is
greater than in carriers of truncation mutations161. This
model also explains why this important phenomenon
was not uniformly observed in many studies that presumably included families with the truncation mutations. An elegant experimental demonstration of this
principle was obtained by Scott et al.162, who expressed,
in cells, recombinant ATM that had been manipulated
to contain missense variations that were previously
found in a cohort of sporadic breast cancer patients163.
Some of these mutant ATM proteins indeed suppressed
the kinase activity of endogenous ATM and caused
radiosensitivity and genomic instability. Strong support
for this notion came from a recent study that used an
animal model of A-T164. Despite the tremendous cancer
predisposition of Atm-knockout mice, animals that
carry the null Atm mutations in a heterozygous state
show no cancer predisposition. However, Spring et al.165
established a knock-in mouse mutant in which an inframe deletion that was previously found to cause A-T
in humans was induced. Mice homozygous for this
mutation produce small amounts of inactive Atm and
usually show the hallmarks of the Atm-knockout phenotype. Significantly, mice heterozygous for this mutation are predisposed to various cancers, unlike the
animals that carry a single knockout allele that does not
produce any protein164.
Further evidence of the importance of ATM missense mutations as cancer-causing genomic alterations
came from the search for somatic ATM mutations in
sporadic human tumours. Indeed, ATM was found to
undergo somatic mutations in sporadic lymphoid
tumours — behaving like a tumour-suppressor gene in
these malignancies (reviewed by Stankovic et al.166).
VOLUME 3 | MARCH 2003 | 1 6 5
REVIEWS
Importantly, a significant proportion of somatic ATM
mutations in these tumours are of the missense
type, indicating that such mutations might indeed be
important in cancer formation.
These observations seem to reconcile the debate on
the role of ATM mutations in genetic predisposition to
cancer159,161, and place the gene encoding the ATM
protein well within the list of genes that are involved in
cancer morbidity in the general population.
Future directions
The ATM protein stands out as a prime model of a
damage signal transducer. As such, its investigation is
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
166
Hahn, W. C. & Weinberg, R. A. Modelling the molecular
circuitry of cancer. Nature Rev. Cancer 2, 331–341 (2002).
Levitt, N. C. & Hickson, I. D. Caretaker tumour suppressor
genes that defend genome integrity. Trends Mol. Med. 8,
179–186 (2002).
Hoeijmakers, J. H. Genome maintenance mechanisms for
preventing cancer. Nature 411, 366–374 (2001).
Davidson, J. F., Guo, H. H. & Loeb, L. A. Endogenous
mutagenesis and cancer. Mutat. Res. 509, 17–21 (2002).
Vessey, C. J., Norbury, C. J. & Hickson, I. D. Genetic disorders
associated with cancer predisposition and genomic instability.
Prog. Nucleic Acid Res. Mol. Biol. 63, 189–221 (1999).
Moses, R. E. DNA damage processing defects and disease.
Annu. Rev. Genom. Hum. Genet. 2, 41–68 (2001).
Thompson, L. H. & Schild, D. Recombinational DNA repair
and human disease. Mutat. Res. 509, 49–78 (2002).
O’Driscoll, M. & Jeggo, P. Immunological disorders and DNA
repair. Mutat. Res. 509, 109–126 (2002).
Rouse, J. & Jackson, S. P. Interfaces between the detection,
signaling, and repair of DNA damage. Science 297,
547–551 (2002).
Jackson, S. P. Sensing and repairing DNA double-strand
breaks. Carcinogenesis 23, 687–696 (2002).
Khanna, K. K. & Jackson, S. P. DNA double-strand breaks:
signaling, repair and the cancer connection. Nature Genet.
27, 247–254 (2001).
van Gent, D. C., Hoeijmakers, J. H. & Kanaar, R.
Chromosomal stability and the DNA double-stranded break
connection. Nature Rev. Genet. 2, 196–206 (2001).
Bartek, J. & Lukas, J. Mammalian G1- and S-phase
checkpoints in response to DNA damage. Curr. Opin. Cell.
Biol. 13, 738–747 (2001).
Pearce, A. K. & Humphrey, T. C. Integrating stress-response
and cell-cycle checkpoint pathways. Trends Cell Biol. 11,
426–433 (2001).
Amundson, S. A., Bittner, M., Meltzer, P., Trent, J. & Fornace,
A. J., Jr. Induction of gene expression as a monitor of exposure
to ionizing radiation. Radiat. Res. 156, 657–661 (2001).
Jelinsky, S. A., Estep, P., Church, G. M. & Samson, L. D.
Regulatory networks revealed by transcriptional profiling of
damaged Saccharomyces cerevisiae cells: Rpn4 links base
excision repair with proteasomes. Mol. Cell. Biol. 20,
8157–8167 (2000).
Sesto, A., Navarro, M., Burslem, F. & Jorcano, J. L. Analysis
of the ultraviolet B response in primary human keratinocytes
using oligonucleotide microarrays. Proc. Natl Acad. Sci.
USA 99, 2965–2970 (2002).
Shiloh, Y. ATM and ATR: networking cellular responses to
DNA damage. Curr. Opin. Genet. Dev. 11, 71–77 (2001).
Shiloh, Y. & Kastan, M. B. ATM: genome stability, neuronal
development, and cancer cross paths. Adv. Cancer Res.
83, 209–254 (2001).
Abraham, R. T. Cell cycle checkpoint signaling through the
ATM and ATR kinases. Genes Dev. 15, 2177–2196 (2001).
Durocher, D. & Jackson, S. P. DNA-PK, ATM and ATR as
sensors of DNA damage: variations on a theme? Curr. Opin.
Cell. Biol. 13, 225–231 (2001).
Khanna, K. K., Lavin, M. F., Jackson, S. P. & Mulhern, T. D.
ATM, a central controller of cellular responses to DNA
damage. Cell Death Differ. 8, 1052–1065 (2001).
McMahon, S. B., Wood, M. A. & Cole, M. D. The essential
cofactor TRRAP recruits the histone acetyltransferase
hGCN5 to c-Myc. Mol. Cell. Biol. 20, 556–562 (2000).
Nikiforov, M. A. et al. TRRAP-dependent and TRRAPindependent transcriptional activation by Myc family
oncoproteins. Mol. Cell. Biol. 22, 5054–5063 (2002).
| MARCH 2003 | VOLUME 3
expected to continue to yield valuable new information
about the DNA-damage response, one of the cardinal
processes for maintaining cellular homeostasis and
keeping the cell from taking the neoplastic path. The
DNA-damage response signifies a crossroads at which
tissue development, ageing and cancer converge.
Understanding the ATM-mediated network, from the
top — upstream of ATM — through the mechanism of
ATM activation, down to the last ATM effector will certainly have an impact on numerous fields of medicine.
Moreover, the contribution of ATM gene mutations to
cancer predisposition in human populations will keep
this gene in the spotlight of cancer epidemiology.
25. Proud, C. G. Regulation of mammalian translation factors by
nutrients. Eur. J. Biochem. 269, 5338–5349 (2002).
26. Dennis, P. B. & Thomas, G. Quick guide: target of
rapamycin. Curr. Biol. 12, R269 (2002).
27. Savitsky, K. et al. A single ataxia telangiectasia gene with a
product similar to PI-3 kinase. Science 268, 1749–1753 (1995).
28. Crawford, T. O. Ataxia telangiectasia. Semin. Pediatr. Neurol.
5, 287–294 (1998).
29. Becker-Catania, S. G. & Gatti, R. A. Ataxia-telangiectasia.
Adv. Exp. Med. Biol. 495, 191–198 (2001).
30. Kim, S. T., Lim, D. S., Canman, C. E. & Kastan, M. B.
Substrate specificities and identification of putative
substrates of ATM kinase family members. J. Biol. Chem.
274, 37538–37543 (1999).
31. O’Neill, T. et al. Utilization of oriented peptide libraries to
identify substrate motifs selected by ATM. J. Biol. Chem.
275, 22719–22727 (2000).
32. Banin, S. et al. Enhanced phosphorylation of p53 by ATM in
response to DNA damage. Science 281, 1674–1677
(1998).
References 32, 33 and 37 provided the first complete
description of an ATM target — Ser15 of p53.
References 32 and 33 provided the first evidence for
ATM activation following DNA damage.
33. Canman, C. E. et al. Activation of the ATM kinase by ionizing
radiation and phosphorylation of p53. Science 281,
1677–1679 (1998).
34. Kozlov, S., Gueven, N., Keating, K., Ramsay, J. & Lavin, M. F.
ATP activates ATM in vitro: importance of
autophosphorylation. J. Biol. Chem. (in the press).
35. Bakkenist, C. J. & Kastan, M. B. DNA damage activates
ATM through intermolecular autophosphorylation and dimer
dissociation. Nature 421, 499–506 (2003).
This seminal work solved the basic mechanism
of ATM activation: ATM lies dormant as dimers
in undamaged cells, with its kinase domain
being blocked by this interaction. Rapid
intermolecular phosphorylation that follows DSB
induction immediately releases very active ATM
monomers.
36. Andegeko, Y. et al. Nuclear retention of ATM at sites of DNA
double strand breaks. J. Biol. Chem. 276, 38224–38230
(2001).
This work described the tight adherence of ATM to
the damaged sites in the DNA following DSB
induction: a fraction of ATM seems to be allocated to
duties to be carried out at the site of the damage.
37. Khanna, K. K. et al. ATM associates with and
phosphorylates p53: mapping the region of interaction.
Nature Genet. 20, 398–400 (1998).
38. Khosravi, R. et al. Rapid ATM-dependent phosphorylation of
MDM2 precedes p53 accumulation in response to DNA
damage. Proc. Natl Acad. Sci. USA 96, 14973–14977
(1999).
39. Ashcroft, M., Kubbutat, M. H. & Vousden, K. H. Regulation
of p53 function and stability by phosphorylation. Mol. Cell.
Biol. 19, 1751–1758 (1999).
40. Dumaz, N. & Meek, D. W. Serine15 phosphorylation
stimulates p53 transactivation but does not directly
influence interaction with HDM2. EMBO J. 18, 7002–7010
(1999).
41. McGowan, C. H. Checking in on Cds1 (Chk2): a checkpoint
kinase and tumor suppressor. Bioessays 24, 502–511
(2002).
42. Bartek, J., Falck, J. & Lukas, J. CHK2 kinase: a busy
messenger. Nature Rev. Mol. Cell. Biol. 2, 877–886 (2001).
43. Ryan, K. M., Phillips, A. C. & Vousden, K. H. Regulation and
function of the p53 tumor suppressor protein. Curr. Opin.
Cell. Biol. 13, 332–337 (2001).
44. Oren, M. et al. Regulation of p53: intricate loops and
delicate balances. Biochem. Pharmacol. 64, 865–871
(2002).
45. Maya, R. et al. ATM-dependent phosphorylation of Mdm2
on serine 395: role in p53 activation by DNA damage.
Genes Dev. 15, 1067–1077 (2001).
46. Saito, S. et al. ATM mediates phosphorylation at multiple
p53 sites, including Ser(46), in response to ionizing radiation.
J. Biol. Chem. 277, 12491–12494 (2002).
47. Waterman, M. J., Stavridi, E. S., Waterman, J. L. &
Halazonetis, T. D. ATM-dependent activation of p53 involves
dephosphorylation and association with 14-3-3 proteins.
Nature Genet. 19, 175–178 (1998).
48. Wang, Y. et al. BASC, a super complex of BRCA1associated proteins involved in the recognition and repair
of aberrant DNA structures. Genes Dev. 14, 927–939
(2000).
49. Venkitaraman, A. R. Cancer susceptibility and the functions
of BRCA1 and BRCA2. Cell 108, 171–182 (2002).
50. Jasin, M. Homologous repair of DNA damage and
tumorigenesis: the BRCA connection. Oncogene 21,
8981–8993 (2002).
51. Yarden, R. I., Pardo-Reoyo, S., Sgagias, M., Cowan, K. H. &
Brody, L. C. BRCA1 regulates the G2/M checkpoint by
activating Chk1 kinase upon DNA damage. Nature Genet.
30, 285–289 (2002).
52. El-Deiry, W. S. Transactivation of repair genes by BRCA1.
Cancer Biol. Ther. 1, 490–491 (2002).
53. Cortez, D., Wang, Y., Qin, J. & Elledge, S. J. Requirement of
ATM-dependent phosphorylation of BRCA1 in the DNA
damage response to double-strand breaks. Science 286,
1162–1166 (1999).
ATM-mediated phosphorylation of BRCA1 was
demonstrated in references 53–55, connecting two
cancer-predisposing genes in the web of the DNAdamage response.
54. Gatei, M. et al. Role for ATM in DNA damage-induced
phosphorylation of BRCA1. Cancer Res. 60, 3299–3304
(2000).
55. Gatei, M. et al. Ataxia telangiectasia mutated (ATM) kinase
and ATM and Rad3 related kinase mediate phosphorylation
of Brca1 at distinct and overlapping sites. In vivo
assessment using phospho-specific antibodies. J. Biol.
Chem. 276, 17276–17280 (2001).
56. Xu, B., O’Donnell, A. H., Kim, S. T. & Kastan, M. B.
Phosphorylation of serine 1387 in Brca1 is specifically
required for the Atm-mediated S-phase checkpoint after
ionizing irradiation. Cancer Res. 62, 4588–4591 (2002).
Work described in references 56 and 57 showed how
ATM-mediated phosphorylation of BRCA1 on different
sites directs it to act at different cell-cycle
checkpoints. These results show the finesse of ATM’s
control of downstream pathways.
57. Xu, B., Kim, S. & Kastan, M. B. Involvement of Brca1 in
S-phase and G(2)-phase checkpoints after ionizing
irradiation. Mol. Cell. Biol. 21, 3445–3450 (2001).
58. Lee, J. S., Collins, K. M., Brown, A. L., Lee, C. H. &
Chung, J. H. hCds1-mediated phosphorylation of BRCA1
regulates the DNA damage response. Nature 404,
201–204 (2000).
59. Li, S. et al. Functional link of BRCA1 and ataxia
telangiectasia gene product in DNA damage response.
Nature 406, 210–215 (2000).
www.nature.com/reviews/cancer
REVIEWS
60. Lim, D. S. et al. ATM phosphorylates p95/nbs1 in an
S-phase checkpoint pathway. Nature 404, 613–617 (2000).
Experiments in references 60–63 show ATM-mediated
phosphorylation of the NBS1 protein and provided the
first functional explanation of the partial overlap
between A-T and NBS1, two chromosomal instability
syndromes.
61. Zhao, S. et al. Functional link between ataxia-telangiectasia
and Nijmegen breakage syndrome gene products. Nature
405, 473–477 (2000).
62. Gatei, M. et al. ATM-dependent phosphorylation of nibrin in
response to radiation exposure. Nature Genet. 25, 115–119
(2000).
63. Wu, X. et al. ATM phosphorylation of Nijmegen breakage
syndrome protein is required in a DNA damage response.
Nature 405, 477–482 (2000).
64. D’Amours, D. & Jackson, S. P. The MRE11 complex: at the
crossroads of DNA repair and checkpoint signalling. Nature
Rev. Mol. Cell. Biol. 3, 317–327 (2002).
65. Kim, S. T., Xu, B. & Kastan, M. B. Involvement of the cohesin
protein, Smc1, in Atm-dependent and independent
responses to DNA damage. Genes Dev. 16, 560–570 (2002).
66. Yazdi, P. T. et al. SMC1 is a downstream effector in the
ATM/NBS1 branch of the human S-phase checkpoint.
Genes Dev. 16, 571–582 (2002).
67. Taniguchi, T. et al. Convergence of the fanconi anemia and
ataxia telangiectasia signaling pathways. Cell 109, 459–472
(2002).
Here, an important functional link between two
apparently remote chromosomal instability
syndromes was firmly established by showing ATMmediated phosphorylation of one of the Fanconi’s
anaemia proteins following DNA damage.
68. Mailand, N. et al. Rapid destruction of human Cdc25A in
response to DNA damage. Science 288, 1425–1429
(2000).
69. Falck, J., Mailand, N., Syljuasen, R. G., Bartek, J. &
Lukas, J. The ATM–Chk2–Cdc25A checkpoint pathway
guards against radioresistant DNA synthesis. Nature 410,
842–847 (2001).
70. Falck, J., Petrini, J. H., Williams, B. R., Lukas, J. & Bartek, J.
The DNA damage-dependent intra-S phase checkpoint is
regulated by parallel pathways. Nature Genet. 30, 290–294
(2002).
This work revealed two distinct mechanisms by which
the important intra-S cell-cycle checkpoint is fine-tuned.
71. Mailand, N. et al. Regulation of G(2)/M events by Cdc25A
through phosphorylation-dependent modulation of its
stability. EMBO J. 21, 5911–5920 (2002).
72. Zhao, H., Watkins, J. L. & Piwnica-Worms, H. Disruption of
the checkpoint kinase 1/cell division cycle 25A pathway
abrogates ionizing radiation-induced S and G2 checkpoints.
Proc. Natl Acad. Sci. USA 99, 14795–14800 (2002).
73. Bell, D. W. et al. Heterozygous germ line hCHK2 mutations
in Li–Fraumeni syndrome. Science 286, 2528–2531
(1999).
74. Sodha, N. et al. Increasing evidence that germline mutations
in CHEK2 do not cause Li–Fraumeni syndrome. Hum.
Mutat. 20, 460–462 (2002).
75. Meijers-Heijboer, H. et al. Low-penetrance susceptibility to
breast cancer due to CHEK2(*)1100delC in noncarriers of
BRCA1 or BRCA2 mutations. Nature Genet. 31, 55–59
(2002).
76. Vahteristo, P. et al. A CHEK2 genetic variant contributing to
a substantial fraction of familial breast cancer. Am. J. Hum.
Genet. 71, 432–438 (2002).
77. Bao, S. et al. ATR/ATM-mediated phosphorylation of human
Rad17 is required for genotoxic stress responses. Nature
411, 969–974 (2001).
78. Post, S. et al. Phosphorylation of serines 635 and 645 of
human Rad17 is cell cycle regulated and is required for
G(1)/S checkpoint activation in response to DNA damage.
Proc. Natl Acad. Sci. USA 98, 13102–13107 (2001).
79. Thelen, M. P., Venclovas, C. & Fidelis, K. A sliding clamp
model for the Rad1 family of cell cycle checkpoint proteins.
Cell 96, 769–770 (1999).
80. Venclovas, C. & Thelen, M. P. Structure-based predictions of
Rad1, Rad9, Hus1 and Rad17 participation in sliding clamp
and clamp-loading complexes. Nucleic Acids Res. 28,
2481–2493 (2000).
81. Rauen, M., Burtelow, M. A., Dufault, V. M. & Karnitz, L. M.
The human checkpoint protein hRad17 interacts with the
PCNA-like proteins hRad1, hHus1, and hRad9. J. Biol.
Chem. 275, 29767–29771 (2000).
82. Roos-Mattjus, P. et al. Genotoxin-induced
Rad9–Hus1–Rad1 (9-1-1) chromatin association is an early
checkpoint signaling event. J. Biol. Chem. 277,
43809–43812 (2002).
83. Chen, M. J., Lin, Y. T., Lieberman, H. B., Chen, G. &
Lee, E. Y. ATM-dependent phosphorylation of human Rad9
NATURE REVIEWS | C ANCER
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
is required for ionizing radiation-induced checkpoint
activation. J. Biol. Chem. 276, 16580–16586 (2001).
Redon, C. et al. Histone H2A variants H2AX and H2AZ.
Curr. Opin. Genet. Dev. 12, 162–169 (2002).
Celeste, A. et al. Genomic instability in mice lacking histone
H2AX. Science 296, 922–927 (2002).
Burma, S., Chen, B. P., Murphy, M., Kurimasa, A. & Chen, D. J.
ATM phosphorylates histone H2AX in response to DNA
double-strand breaks. J. Biol. Chem. 276, 42462–42467
(2001).
Ward, I. M. & Chen, J. Histone H2AX is phosphorylated in an
ATR-dependent manner in response to replicational stress.
J. Biol. Chem. 276, 47759–47762 (2001).
Buscemi, G. et al. Chk2 activation dependence on Nbs1
after DNA damage. Mol. Cell. Biol. 21, 5214–5222
(2001).
Girard, P. M., Riballo, E., Begg, A. C., Waugh, A. & Jeggo, P. A.
Nbs1 promotes ATM dependent phosphorylation events
including those required for G1/S arrest. Oncogene 21,
4191–4199 (2002).
DiTullio, R. A. Jr et al. 53BP1 functions in an ATMdependent checkpoint pathway that is constitutively
activated in human cancer. Nature Cell Biol. 4, 998–1002
(2002).
Fernandez-Capetillo, O. et al. DNA damage-induced G2-M
checkpoint activation by histone H2AX and 53BP1. Nature
Cell Biol. 4, 993–997 (2002).
Wang, B., Matsuoka, S., Carpenter, P. B. & Elledge, S. J.
53BP1, a mediator of the DNA damage checkpoint.
Science 298, 1435–1438 (2002).
Peng, A. & Chen, P. L. NFBD1, like 53BP1, is an early and
redundant transducer mediating Chk2 phosphorylation in
response to DNA damage. J. Biol. Chem. 24 Jan 2003
[epub ahead of print].
Shang, Y. L., Bodero, A. J. & Chen, P. L. NFBD1, a novel
nuclear protein with signature motifs of FHA and BRCT, and
an internal 41 amino acid repeat sequence, is an early
participant in DNA damage response. J. Biol. Chem. 9 Dec
2002 [epub ahead of print].
Xu, X. & Stern, D. F. NFBD1/KIAA0170 is a chromatinassociated protein involved in DNA damage signaling
pathways. J. Biol. Chem. 23 Dec 2002 [epub ahead of
print].
Chen, G. et al. Radiation-induced assembly of Rad51 and
Rad52 recombination complex requires ATM and c-Abl.
J. Biol. Chem. 274, 12748–12752 (1999).
Morrison, C. et al. The controlling role of ATM in homologous
recombinational repair of DNA damage. EMBO J. 19,
463–471 (2000).
Yuan, Z. M. et al. Regulation of Rad51 function by c-Abl in
response to DNA damage. J. Biol. Chem. 273, 3799–3802
(1998).
Kitao, H. & Yuan, Z. M. Regulation of ionizing radiationinduced Rad52 nuclear foci formation by c-Abl-mediated
phosphorylation. J. Biol. Chem. 277, 48944–48948
(2002).
Shafman, T. et al. Interaction between ATM protein and
c-Abl in response to DNA damage. Nature 387, 520–523
(1997).
Baskaran, R. et al. Ataxia telangiectasia mutant protein
activates c-Abl tyrosine kinase in response to ionizing
radiation. Nature 387, 516–519 (1997).
Shangary, S. et al. Regulation of DNA-dependent protein
kinase activity by ionizing radiation-activated abl kinase is an
ATM-dependent process. J. Biol. Chem. 275,
30163–30168 (2000).
Foray, N. et al. Constitutive association of BRCA1 and c-Abl
and its ATM-dependent disruption after irradiation. Mol. Cell.
Biol. 22, 4020–4032 (2002).
Lin, W. C., Lin, F. T. & Nevins, J. R. Selective induction of
E2F1 in response to DNA damage, mediated by ATMdependent phosphorylation. Genes Dev. 15, 1833–1844
(2001).
Bar-Shira, A. et al. ATM-dependent activation of the gene
encoding MAP kinase phosphatase 5 by radiomimetic DNA
damage. Oncogene 21, 849–855 (2002).
Wang, X. et al. Involvement of the MKK6-p38gamma
cascade in gamma-radiation-induced cell cycle arrest. Mol.
Cell. Biol. 20, 4543–4552 (2000).
Li, N. et al. ATM is required for IkappaB kinase (IKKk)
activation in response to DNA double strand breaks. J. Biol.
Chem. 276, 8898–8903 (2001).
Pandita, T. K. The role of ATM in telomere structure and
function. Radiat. Res. 156, 642–647 (2001).
Maser, R. S. & DePinho, R. A. Connecting
chromosomes, crisis, and cancer. Science 297, 565–569
(2002).
Kim, S. H., Kaminker, P. & Campisi, J. Telomeres, aging and
cancer: in search of a happy ending. Oncogene 21,
503–511 (2002).
111. Chan, S. W. & Blackburn, E. H. New ways not to make ends
meet: telomerase, DNA damage proteins and
heterochromatin. Oncogene 21, 553–563 (2002).
112. Karlseder, J., Broccoli, D., Dai, Y., Hardy, S. &
de Lange, T. p53- and ATM-dependent apoptosis induced
by telomeres lacking TRF2. Science 283, 1321–1325
(1999).
113. Wong, K. K. et al. Telomere dysfunction and Atm deficiency
compromises organ homeostasis and accelerates ageing.
Nature 22 Jan 2003 [epub ahead of print].
Mice doubly deficient for Atm and the telomerase
RNA component showed a marked proliferative
defect, genome stability and premature ageing,
pointing at the importance of the defect in telomere
maintenance in some A-T symptoms.
114. Kishi, S. et al. Telomeric protein Pin2/TRF1 as an important
ATM target in response to double strand DNA breaks.
J. Biol. Chem. 276, 29282–29291 (2001).
115. Kishi, S. & Lu, K. P. A critical role for Pin2/TRF1 in ATMdependent regulation. Inhibition of Pin2/TRF1 function
complements telomere shortening, radiosensitivity, and the
G(2)/M checkpoint defect of ataxia-telangiectasia cells.
J. Biol. Chem. 277, 7420–7429 (2002).
116. Perkins, E. J. et al. Sensing of intermediates in V(D)J
recombination by ATM. Genes Dev. 16, 159–164 (2002).
117. Xu, Y. ATM in lymphoid development and tumorigenesis.
Adv. Immunol. 72, 179–189 (1999).
118. Tauchi, H., Matsuura, S., Kobayashi, J., Sakamoto, S. &
Komatsu, K. Nijmegen breakage syndrome gene, NBS1,
and molecular links to factors for genome stability.
Oncogene 21, 8967–8980 (2002).
119. Stewart, G. S. et al. The DNA double-strand break repair
gene hMRE11 is mutated in individuals with an ataxiatelangiectasia-like disorder. Cell 99, 577–587 (1999).
Here, the molecular basis for A-T-like disease (ATLD)
— MRE11 deficiency — was identified. This discovery
and the striking clinical similarity between A-T and
ATLD point to intimate functional relationships
between ATM and the MRN complex, which might be
related to the very early events in the DSB response.
120. Joenje, H. & Patel, K. J. The emerging genetic and
molecular basis of Fanconi anaemia. Nature Rev. Genet. 2,
446–457 (2001).
121. Nakanishi, K. et al. Interaction of FANCD2 and NBS1 in the
DNA damage response. Nature Cell Biol. 4, 913–920 (2002).
References 121 and 122 show the functional
interaction between FANCD2 and NBS1 in the
interstrand crosslink repair pathway, adding another
thread to the damage-response network and drawing
an additional line connecting different chromosomal
instability syndromes.
122. Pichierri, P., Averbeck, D. & Rosselli, F. DNA cross-linkdependent RAD50/MRE11/NBS1 subnuclear assembly
requires the Fanconi anemia C protein. Hum. Mol. Genet.
11, 2531–2546 (2002).
123. Beamish, H. et al. Functional link between BLM defective
in Bloom’s syndrome and the ataxia-telangiectasiamutated protein, ATM. J. Biol. Chem. 277, 30515–30523
(2002).
The finding that the Bloom’s syndrome gene product,
BLM, is an ATM target further intensifies the crosstalk between the chromosomal breakage syndromes.
124. Franchitto, A. & Pichierri, P. Bloom’s syndrome protein is
required for correct relocalization of RAD50/MRE11/NBS1
complex after replication fork arrest. J. Cell Biol. 157, 19–30
(2002).
125. Tibbetts, R. S. et al. A role for ATR in the DNA damageinduced phosphorylation of p53. Genes Dev. 13, 152–157
(1999).
126. Zou, L., Cortez, D. & Elledge, S. J. Regulation of ATR
substrate selection by Rad17-dependent loading of Rad9
complexes onto chromatin. Genes Dev. 16, 198–208
(2002).
127. Hammond, E. M., Denko, N. C., Dorie, M. J.,
Abraham, R. T. & Giaccia, A. J. Hypoxia links ATR and p53
through replication arrest. Mol. Cell. Biol. 22, 1834–1843
(2002).
128. Heffernan, T. P. et al. An ATR- and Chk1-dependent
S checkpoint inhibits replicon initiation following UVCinduced DNA damage. Mol. Cell. Biol. 22, 8552–8561
(2002).
129. Tibbetts, R. S. et al. Functional interactions between BRCA1
and the checkpoint kinase ATR during genotoxic stress.
Genes Dev. 14, 2989–3002 (2000).
130. Unsal-Kacmaz, K., Makhov, A. M., Griffith, J. D. & Sancar, A.
Preferential binding of ATR protein to UV-damaged DNA.
Proc. Natl Acad. Sci. USA 99, 6673–6678 (2002).
131. Zhao, H. & Piwnica-Worms, H. ATR-mediated checkpoint
pathways regulate phosphorylation and activation of human
Chk1. Mol. Cell. Biol. 21, 4129–4139 (2001).
VOLUME 3 | MARCH 2003 | 1 6 7
REVIEWS
132. Guo, Z., Kumagai, A., Wang, S. X. & Dunphy, W. G.
Requirement for Atr in phosphorylation of Chk1 and cell
cycle regulation in response to DNA replication blocks and
UV-damaged DNA in Xenopus egg extracts. Genes Dev. 14,
2745–2756 (2000).
133. Lopez-Girona, A. et al. Serine-345 is required for Rad3dependent phosphorylation and function of checkpoint
kinase Chk1 in fission yeast. Proc. Natl Acad. Sci. USA 98,
11289–11294 (2001).
134. Nghiem, P., Park, P. K., Kim, Y. S., Desai, B. N. & Schreiber, S. L.
ATR is not required for p53 activation but synergizes with
p53 in the replication checkpoint. J. Biol. Chem. 277,
4428–4434 (2002).
135. Weiss, R. S., Matsuoka, S., Elledge, S. J. & Leder, P.
Hus1 acts upstream of chk1 in a mammalian DNA
damage response pathway. Curr. Biol. 12, 73–77
(2002).
136. Hekmat-Nejad, M., You, Z., Yee, M. C., Newport, J. W. &
Cimprich, K. A. Xenopus ATR is a replication-dependent
chromatin-binding protein required for the DNA replication
checkpoint. Curr. Biol. 10, 1565–1573 (2000).
137. de Klein, A. et al. Targeted disruption of the cell-cycle
checkpoint gene ATR leads to early embryonic lethality in
mice. Curr. Biol. 10, 479–482 (2000).
138. Brown, E. J. & Baltimore, D. ATR disruption leads to
chromosomal fragmentation and early embryonic lethality.
Genes Dev. 14, 397–402 (2000).
139. Cortez, D., Guntuku, S., Qin, J. & Elledge, S. J. ATR and
ATRIP: partners in checkpoint signaling. Science 294,
1713–1716 (2001).
The discovery of ATRIP — an ATR-associated protein
that is essential for its stability and activity — was
important for understanding how ATR is set in motion
following DNA damage.
140. Lupardus, P. J., Byun, T., Yee, M. C., Hekmat-Nejad, M. &
Cimprich, K. A. A requirement for replication in activation of
the ATR-dependent DNA damage checkpoint. Genes Dev.
16, 2327–2332 (2002).
141. Casper, A. M., Nghiem, P., Arlt, M. F. & Glover, T. W. ATR
regulates fragile site stability. Cell 111, 779–789 (2002).
ATR — but not ATM — was found to be important for
preventing the expression of chromosomal fragile sites
in the face of replicative stress. This draws another
important distinction between ATM and ATR and points
out that ATR is involved in the checkpoint associated
with DNA replication rather than with DNA damage.
142. Edwards, R. J., Bentley, N. J. & Carr, A. M.
A Rad3–Rad26 complex responds to DNA damage
independently of other checkpoint proteins. Nature Cell
Biol. 1, 393–398 (1999).
143. Rouse, J. & Jackson, S. P. LCD1: an essential gene involved
in checkpoint control and regulation of the MEC1 signalling
pathway in Saccharomyces cerevisiae. EMBO J. 19,
5801–5812 (2000).
144. Paciotti, V., Clerici, M., Lucchini, G. & Longhese, M. P. The
checkpoint protein Ddc2, functionally related to S. pombe
Rad26, interacts with Mec1 and is regulated by Mec1dependent phosphorylation in budding yeast. Genes Dev.
14, 2046–2059 (2000).
145. Kondo, T., Wakayama, T., Naiki, T., Matsumoto, K. &
Sugimoto, K. Recruitment of Mec1 and Ddc1 checkpoint
proteins to double-strand breaks through distinct
mechanisms. Science 294, 867–870 (2001).
146. Wakayama, T., Kondo, T., Ando, S., Matsumoto, K. &
Sugimoto, K. Pie1, a protein interacting with Mec1,
controls cell growth and checkpoint responses in
Saccharomyces cerevisiae. Mol. Cell. Biol. 21, 755–764
(2001).
168
| MARCH 2003 | VOLUME 3
147. Rouse, J. & Jackson, S. P. Lcd1p recruits Mec1p to DNA
lesions in vitro and in vivo. Mol. Cell 9, 857–869 (2002).
148. Maquat, L. E. & Carmichael, G. G. Quality control of mRNA
function. Cell 104, 173–176 (2001).
149. Yamashita, A., Ohnishi, T., Kashima, I., Taya, Y. & Ohno, S.
Human SMG-1, a novel phosphatidylinositol 3-kinaserelated protein kinase, associates with components of the
mRNA surveillance complex and is involved in the regulation
of nonsense-mediated mRNA decay. Genes Dev. 15,
2215–2228 (2001).
150. Denning, G., Jamieson, L., Maquat, L. E., Thompson, E. A.
& Fields, A. P. Cloning of a novel phosphatidylinositol kinaserelated kinase: characterization of the human SMG-1 RNA
surveillance protein. J. Biol. Chem. 276, 22709–22714
(2001).
151. Smith, G. C. & Jackson, S. P. The DNA-dependent protein
kinase. Genes Dev. 13, 916–934 (1999).
152. DeFazio, L. G., Stansel, R. M., Griffith, J. D. & Chu, G.
Synapsis of DNA ends by DNA-dependent protein kinase.
EMBO J. 21, 3192–3200 (2002).
153. Chan, D. W. et al. Autophosphorylation of the DNA-dependent
protein kinase catalytic subunit is required for rejoining of DNA
double-strand breaks. Genes Dev. 16, 2333–2338 (2002).
Work described in references 153 and 154 shows, for
the first time, that DNA-PK activation by DNA damage
involves autophosphorylation.
154. Douglas, P. et al. Identification of in vitro and in vivo
phosphorylation sites in the catalytic subunit of the DNAdependent protein kinase. Biochem. J. 368, 243–251
(2002).
155. Yannone, S. M. et al. Werner syndrome protein is regulated
and phosphorylated by DNA-dependent protein kinase.
J. Biol. Chem. 276, 38242–38248 (2001).
156. Karmakar, P. et al. Werner protein is a target of DNAdependent protein kinase in vivo and in vitro, and its catalytic
activities are regulated by phosphorylation. J. Biol. Chem.
277, 18291–18302 (2002).
157. Ma, Y., Pannicke, U., Schwarz, K. & Lieber, M. R. Hairpin
opening and overhang processing by an Artemis/DNAdependent protein kinase complex in nonhomologous end
joining and V(D)J recombination. Cell 108, 781–794
(2002).
Here, an important DNA-PK substrate was identified
after a long search for physiological substrates of this
central player in DSB repair. It is the Artemis protein,
which is involved in processing the hairpin structures
at the site of DSBs formed during V(D)J
recombination.
158. Moshous, D. et al. Artemis, a novel DNA double-strand
break repair/V(D)J recombination protein, is mutated in
human severe combined immune deficiency. Cell 105,
177–186 (2001).
159. Khanna, K. K. Cancer risk and the ATM gene: a continuing
debate. J. Natl Cancer Inst. 92, 795–802 (2000).
160. Stankovic, T. et al. ATM mutations and phenotypes in ataxiatelangiectasia families in the British Isles: expression of
mutant ATM and the risk of leukemia, lymphoma, and breast
cancer. Am. J. Hum. Genet. 62, 334–345 (1998).
161. Gatti, R. A., Tward, A. & Concannon, P. Cancer risk in ATM
heterozygotes: a model of phenotypic and mechanistic
differences between missense and truncating mutations.
Mol. Genet. Metab. 68, 419–423 (1999).
A model was suggested to reconcile the apparently
conflicting data on cancer predisposition of A-T
carriers. The model attributes different degrees of
cancer predisposition to different hererozygous ATM
mutations. Recent results seem to substantiate this
model.
162. Scott, S. P. et al. Missense mutations but not allelic variants
alter the function of ATM by dominant interference in
patients with breast cancer. Proc. Natl Acad. Sci. USA 99,
925–930 (2002).
163. Dork, T. et al. Spectrum of ATM gene mutations in a
hospital-based series of unselected breast cancer patients.
Cancer Res. 61, 7608–7615 (2001).
164. Spring, K. et al. Mice heterozygous for mutation in Atm, the
gene involved in ataxia-telangiectasia, have heightened
susceptibility to cancer. Nature Genet. 32, 185–190 (2002).
This work provided strong support for the model
associating different ATM mutations with different
degrees of cancer predisposition in heterozygotes.
165. Spring, K. et al. Atm knock-in mice harboring an in-frame
deletion corresponding to the human ATM 7636del9
common mutation exhibit a variant phenotype. Cancer Res.
61, 4561–4568 (2001).
166. Stankovic, T. et al. ATM mutations in sporadic lymphoid
tumours. Leuk. Lymph. 43, 1563–1571 (2002).
167. D’Andrea, A. D. & Grompe, M. The Fanconi anaemia/BRCA
pathway. Nature Rev. Cancer 3, 23–34 (2003).
168. Howlett, N. G. et al. Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 297, 606–609 (2002).
169. Gellert, M. V(D)J recombination: rag proteins, repair factors,
and regulation. Annu. Rev. Biochem. 71, 101–132 (2002).
170. Pellegrini, L. et al. Insights into DNA recombination from the
structure of a RAD51–BRCA2 complex. Nature 420,
287–293 (2002).
171. Yang, H. et al. BRCA2 function in DNA binding and
recombination from a BRCA2–DSS1–ssDNA structure.
Science 297, 1837–1848 (2002).
172. Tauchi, H. et al. Nbs1 is essential for DNA repair by
homologous recombination in higher vertebrate cells. Nature
420, 93–98 (2002).
173. Kobayashi, J. et al. NBS1 localizes to gamma-H2AX foci
through interaction with the FHA/BRCT domain. Curr. Biol.
12, 1846–1851 (2002).
Acknowledgements
The author wishes to thank the members of the David and Inez Myers
Laboratory for Genetic Research for their dedication to A-T research.
Work in the author’s laboratory is supported by research grants from
the A-T Medical Research Foundation, The A-T Children’s Project,
The A-T Medical Research Trust, The National Institutes of Health and
the Israel Ministry of Science, Culture and Sport.
Online links
DATABASES
The following terms in this article are linked online to:
LocusLink: http://www.ncbi.nih.gov/LocusLink/
53BP1 | ABL | ATM | ATRIP | ATR | ATX | BLM | BRCA1 | BRCA2 |
CDC25A | CDC25C | CDK2 | CHK1 | cyclin E | E2F1 | FANCD2 |
H2AX | KU70 | KU80 | MDM2 | MRE11 | mTOR | NBS1 | NF-κB |
p53 | PI3K | RAD9 | RAD17 | RAD50 | RAD51 | RAD52 | RAD54 |
SMC1 | TRF1 | TRRAP | WAF1
OMIM: http://www.ncbi.nlm.nih.gov/Omim/
ataxia telangiectasia | Bloom’s syndrome | Fanconi’s anaemia |
Nijmegen breakage syndrome
FURTHER INFORMATION
The ATM entry in the NCBI web site:
http://www.ncbi.nlm.nih.gov/disease/ATM.html
Information on A-T at the NIH site:
http://www.ninds.nih.gov/health_and_medical/disorders/a-t.htm;
http://www.medhelp.org/lib/ataxia.htm
Access to this interactive links box is free online.
www.nature.com/reviews/cancer