Checkpoint and Coordinated Cellular Responses to DNA Damage

Results Probl Cell Differ (42)
P. Kaldis: Cell Cycle Regulation
DOI 10.1007/b136684/Published online: 14 July 2005
© Springer-Verlag Berlin Heidelberg 2005
Checkpoint and Coordinated Cellular Responses
to DNA Damage
Xiaohong H. Yang
1
·LeeZou
1,2
(✉)
1
MGH Cancer Center, Harvard Medical School, Charlestown, MA 02129, USA

2
Department of Pathology, Harvard Medical School, Charlestown, MA 02129, USA

Abstract The DNA damage and replication checkpoints are signaling mechanisms that
regulate and coordinate cellular responses to genotoxic conditions. The activation of
checkpoints not only attenuates cell cycle progression, but also facilitates DNA repair
and recovery of faulty replication forks, thereby preventing DNA lesions from being con-
verted to inheritable mutations. It has become increasingly clear that the activation and
signaling of the checkpoint are intimately linked to the cellular processes directly in-
volved in chromosomal metabolism, such as DNA replication and DNA repair. Thus, the
checkpoint pathway is not just a surveillance system that monitors genomic integrity and
regulates cell proliferation, but also an integral part of the processes that work directly
on chromosomes to maintain genomic stability. In this article, we discuss the current
models of DNA damage and replication checkpoints, and highlight recent advances in the
field.
1
Introduction

The survival of organisms relies on faithful duplication and segregation of
their genomes. To ensure that the entire genome is accurately transmitted to
the next generation of cells, duplication and segregation of the genome must
be highly coordinated. The maintenance of genomic stability is a mounting
task for cells because the DNA content in every cell is constantly challenged
by both extrinsic insults and intrinsic stresses. For example, DNA lesions can
arise from failures of the DNA replication machine or from various types of
DNA-damaging molecules or radiations in the environment. If DNA damage
is not accurately and quickly repaired, or if the coordinated genome duplica-
tion and segregation cannot be maintained, detrimental events such as loss
of genetic information, deleterious chromosomal rearrangements, and mu-
tations disrupting the control of cell proliferation might result. To maintain
genomic stability in the presence of DNA damage or DNA replication in-
terference, cells use a complex signaling pathway called the checkpoint to
regulate and coordinate many cellular processes to remove DNA damage and
66 X.H. Yang · L. Zou
to alleviate stresses on their genomes (Hartwell and Weinert 1989). It was
initially discovered that activation of the checkpoint by DNA damage leads
to cell cycle arrest, probably providing more time for DNA repair (Hartwell
and Weinert 1989). Now it has become increasingly clear that the checkpoint
also plays vital roles in the regulation of DNA replication, DNA repair, chro-
matin structure, and many other cellular processes important for genomic
stability (Zhou and Elledge 2000; Osborn et al. 2002; Kastan and Bartek 2004).
Most interestingly, many of the processes regulated by the checkpoint might
also play roles in generating or transmitting DNA damage signals through
the checkpoint pathway (Osborn et al. 2002). Thus, the checkpoint pathway
is not only a surveillance system that monitors genomic integrity, but also an
integral part of the processes that directly work on chromosomes to main-
tain their stability. Recent studies have revealed important mechanisms by
which cells detect various types of DNA damage and activate the checkpoint

pathway, as well as the mechanisms by which the checkpoint regulates key
downstream processes such as DNA replication, DNA repair, and chromatin
modulation. In this review, we will discuss an updated model of checkpoint
signaling that begins to explain how different processes involved in the main-
tenance of genomic stability are integrated and coordinated by the checkpoint
pathway.
2
Sensing DNA Damage and DNA Replication Stress
ATM ( ataxia telangiectasia mutated)andATR(ATM- an d R a d3 - related)are
two large PI3K-like protein kinases that play central roles in the checkpoint-
signaling pathway (Abraham 2001). In response to DNA damage, ATM and
ATR phosphorylate Chk1 and Chk2, two downstream effector kinases, and
numerous substrates involved in various cellular processes (e.g. p53, Brca1,
Nbs1). The DNA damage specificities of ATM and ATR are distinct from
each other. While ATM primarily responds to double-strand DNA breaks
(DSBs), ATR is involved in the responses to DSBs as well as a broad spec-
trum of DNA damage caused by DNA replication interference. Although ATM
is important for genomic stability, patients, mice, and cells lacking ATM are
viable (Kastan and Bartek 2004), suggesting that ATM is not essential for nor-
mal cell proliferation in the absence of significant DSBs. On the other hand,
ATR is indispensable for the proliferation of human and mouse cells (Brown
and Baltimore 2000; Cortez et al. 2001). These findings indicate that ATR
has a critical role even in normal cell cycles, and that the function of ATR
might be regulated by certain DNA structures generated by intrinsic DNA
metabolism.
Checkpoint and Coordinated Cellular Responses to DNA Damage 67
2.1
Recruitment of ATR to DNA
What is the DNA structure sensed by the ATR kinase? Several important clues
came from the studies of Mec1, the budding yeast homologue of ATR (see

Table 1). In human cells, most ATR exists in a complex with its partner ATRIP
(Cortez et al. 2001). Yeast Mec1 also forms a similar complex with its part-
ner Ddc2 (Paciotti et al. 2000; Rouse and Jackson 2000; Wakayama et al. 2001).
Importantly, the checkpoint functions of ATR and Mec1 are dependent upon
ATRIP and Ddc2, respectively, indicating that ATR-ATRIP and Mec1-Ddc2
function as complexes in checkpoint signaling. In budding yeast, Mec1 can be
activated in the cdc13 mutant and by DSBs generated by the HO endonuclease
(Lydall and Weinert 1995; Pellicioli et al. 2001). Interestingly, single-stranded
DNA (ssDNA) is generated at telomeres in the cdc13 mutant and at DSBs after
they are recessed by exonucleases. Furthermore, Mec1 and Ddc2 have been
shown to localize to telomeres in the cdc13 mutant and to the HO-induced
DSBs (Kondo et al. 2001; Melo et al. 2001; Rouse and Jackson 2002; Zou and
Elledge 2003; Lisby et al. 2004). These findings indicated that ssDNA might
bepartoftheDNAstructurerecognizedbytheDNAdamagesensorsthat
Table 1 The proteins involved in checkpoint signaling in human and yeast cells
Human Budding Yeast
PI3K-like kinase ATR Mec1
ATM Tel1
ATR/Mec1 ATRIP Ddc2
regulatory partner
Replication protein A Rpa1-3 Rpa1-3
RFC-like complex Rad17 Rad24
Rfc2-5 Rfc2-5
PCNA-like complex Rad9 Ddc1
Rad1 Rad17
Hus1 Mec3
MRN complex Mre11 Mre11
Rad50 Rad50
Nbs1 Xrs2
Mediators Claspin Mrc1

and other Brca1/53BP1/Mdc1 Rad9
signaling molecules TopBP1 Dpb11
hTim1 Tof1
hTipin Csm3
Effector kinase Chk1 Chk1
Chk2 Rad53
68 X.H. Yang · L. Zou
recruit the ATR-ATRIP kinase complex. Consistent with this idea, increased
amountsofssDNAwasalsoobservedatDNAreplicationforkshaltedbyhy-
droxyurea (HU) treatment (Sogo et al. 2002), suggesting that ssDNA might
also be important for the activation of Mec1 by replication fork stalling.
Studies using Xenopus egg extracts have also revealed important clues
of how ATR is recruited to DNA. In Xenopus extracts, ATR associates with
chromatin during S-phase in a replication-dependent manner (Hekmat-Nejad
et al. 2000). Depletion of RPA, an ssDNA-binding protein complex essential
for DNA replication, abolished the chromatin association of ATR (You et al.
2002), suggesting that RPA is either directly or indirectly required for the re-
cruitment of ATR to chromatin. Similarly, in Xenopus extracts RPA is also
needed for the recruitment of ATR to DNA lesions generated by etoposide
(Costanzo et al. 2003), a DNA topoisomerase II inhibitor. This finding indi-
catesthatRPAitselfortheDNArepairprocessinvolvingRPAisrequiredfor
ATR recr uit ment.
In human cells, RPA is required for the localization of ATR to DNA
damage-induced nuclear foci and the efficient phosphorylation of Chk1 by
ATR (Zou and Elledge 2003). In yeast, depletion of RPA in the cells arrested
in G2 abolished the localization of Ddc2 to the HO-induced DSBs (Zou and
Elledge 2003), suggesting that RPA is required for the recruitment of Ddc2
to DNA damage in vivo and its function is independent of its role in DNA
replication. Indeed, rfa1-t11, a mutant of RPA that is proficient for DNA repli-
cation but partially defective for checkpoint responses (Umezu et al. 1998;

Kim and Brill 2001; Pellicioli et al. 2001), exhibits diminished ability to re-
cruit Ddc2 to DSBs and stalled replication forks in vivo (Zou and Elledge
2003; Lucca et al. 2004). All these findings suggest that RPA probably plays
a rather direct role in the recruitment of ATR-ATRIP and Mec1-Ddc2. The
direct role of RPA in the recruitment of ATR-ATRIP was eventually demon-
strated by a series of in vitro biochemical experiments (Zou and Elledge 2003)
(Fig.1a,b).First,thepurifiedhumanATRIP protein is efficiently recruited to
ssDNA only in the presence of RPA. Second, RPA renders ATRIP capable of
distinguishing ssDNA from double-stranded DNA (dsDNA). Finally, the ATR-
ATRIP complex, but not ATR alone, binds to ssDNA more efficiently in the
presence of RPA. Together, these findings strongly suggest that RPA-coated
ssDNA is sufficient for the recruitment of ATR-ATRIP to the sites of DNA
damage. Consistently, both purified yeast Ddc2 and Xenopus AT R I P are ef-
ficiently recruited to ssDNA in an RPA-dependent manner (Zou and Elledge
2003; Kumagai et al. 2004). Notably, the recruitment of Xenopus ATRIP to
the DNA structures formed by polyT-polyA oligomers, which can elicit ATR
signaling in Xenopus extracts, is also dependent upon RPA (Kumagai et al.
2004).
Although RPA is required for the recruitment of ATR-ATRIP to DNA dam-
age in vivo and RPA-coated ssDNA is sufficient for recruiting ATR-ATRIP in
vitro, the possibility that ATR-ATRIP can interact with RPA-ssDNA through
Checkpoint and Coordinated Cellular Responses to DNA Damage 69
Fig. 1 Models for checkpoint activation by DSBs and stalled replication forks. a Check-
point activation by DSBs. ATM is autophosphorylated in response to DSBs. The MRN
complex may recruit ATM to DSBs and activate ATM. ATM phosphorylates Nbs1, Brca1,
and Smc1 at DSBs. When DSBs are recessed, the resulting ssDNA is coated by RPA. ATR-
ATRIP is recruited to DSBs by RPA-ssDNA. RPA also stimulates the loading of 9-1-1
complexes by the Rad17 complex. b Checkpoint activation by stalled replication forks.
Long stretches of ssDNA are generated at stalled replication forks. ATR-ATRIP, Rad17, and
9-1-1 complexes are recruited to RPA-ssDNA and junctions of double/single-stranded

DNA at stalled forks. Although the checkpoint can be activated through different mech-
anisms, cell cycle arrest and inhibition of DNA replication are common results. The
activated checkpoint may play important roles at DSBs and stalled replication forks to
facilitate DNA repair and fork recovery
other proteins cannot be ruled out. Furthermore, it is also possible that
the interaction between ATR-ATRIP and RPA-ssDNA is regulated by other
proteins in vivo (see below). It was recently reported that ATR-ATRIP can
associate with proteins such as Claspin, Msh2 and Mcm7 (Chini and Chen
2003; Wang and Qin 2003; Cortez et al. 2004). It is possible that the interac-
tions of ATR-ATRIP with additional proteins on DNA also contribute to the
localization of ATR-ATRIP to specific types of DNA damage.
2.2
DNA Damage Recognition by the RFC- and PCNA-like Checkpoint Complexes
Although ssDNA plays a crucial role in the recruitment of ATR-ATRIP, ss-
DNA alone is not sufficient to elicit the checkpoint responses, suggesting that
additional DNA structures induced by DNA damage are also necessary for
the activation of ATR-ATRIP. In addition to ATR-ATRIP itself, several other
checkpoint proteins are also required for the initiation of checkpoint signal-
ing. In human cells, these proteins include Rad17, Rad9, Rad1, and Hus1.
Rad17 is a homologue of all five subunits of RFC, and its forms an RFC-like
protein complex with the four small subunits of RFC (Lindsey-Boltz et al.
2001; Shiomi et al. 2002; Ellison and Stillman 2003; Zou et al. 2003). Rad9,
Rad1, and Hus1, on the other hand, are all structurally related to PCNA
(Venclovas and Thelen 2000), and they assemble into a hetero-trimeric ring-
shaped complex (termed the 9-1-1 complex) resembling PCNA (Volkmer and
Karnitz 1999). Ablation of Rad17 or the 9-1-1 complex in human or mouse
cells resulted in severe chromosomal instability and failures in activating the
ATR-mediate checkpoint (Weiss et al. 2002; Zou et al. 2002; Roos-Mattjus et al.
2003; Wang et al. 2003; Bao et al. 2004; Loegering et al. 2004). During DNA
replication, RFC specifically recognizes the 3


primer-template junctions on
DNA and recruits PCNA onto DNA where it functions as the processivity
factor for DNA polymerases. Interestingly, in response to DNA damage, the
9-1-1 complex is recruited onto chromatin in an Rad17-dependent manner
(Zou et al. 2002), indicating that the RFC-like Rad17 complex might recog-
70 X.H. Yang · L. Zou
Checkpoint and Coordinated Cellular Responses to DNA Damage 71
nize certain damage-induced DNA structure and recruit the 9-1-1 complex in
a manner similar to the loading of PCNA by RFC.
What then is the DNA structure recognized by the Rad17 complex?
In Xenopus extracts, the 9-1-1 complex associates with chromatin during
S-phase like ATR does (You et al. 2002). However, unlike ATR, the chromatin
association of 9-1-1 complex requires DNA polymerase α (You et al. 2002),
indicating that the synthesis of RNA-DNA primer might be either directly or
indirectly required for the function of the Rad17 complex. In budding yeast,
both the HO-induced DSBs and the telomeres in cdc13 mutant cells are re-
cessed by 5

-to-3

exonucleases, resulting in 5

junctions of dsDNA and ssDNA
(Lydall and Weinert 1995; Lee et al. 1998). Furthermore, the yeast PCNA-like
checkpoint complex can be specifically recruited to the HO-induced DSBs
(Kondo et al. 2001), suggesting that the RFC-like checkpoint complex might
recognize the junctions of dsDNA/ssDNA. Interestingly, in yeast the efficient
recruitment of the PCNA-like complex to DSBs also requires the function of
RPA (Zou et al. 2003; Nakada et al. 2004). Consistent with these in vivo ob-

servations, purified Rad17 complexes recruit 9-1-1 complexes onto primed
ssDNA or gapped DNA structures in an RPA-dependent manner (Ellison and
Stillman 2003; Zou et al. 2003) (Fig. 1a, b). Unlike RFC, which only uses the
3

dsDNA/ssDNA junctions to load PCNA, the Rad17 complex can apparently
use the 5

dsDNA/ssDNA junctions to recruit 9-1-1 complexes, providing
a possible explanation of the DNA damage specificity of the Rad17 complex.
Together, the studies described above have revealed that the ATR-ATRIP,
Rad17, and 9-1-1 complexes can independently recognize damage-induced
DNA structures such as ssDNA and junctions of dsDNA/ssDNA. Moreover,
RPA appears to play important roles in the damage recognition by both the
ATR-ATRIP and the Rad17 complexes. Once recruited to the sites of DNA
damage, the Rad17 and 9-1-1 complexes might facilitate the recognition and
phosphorylation of ATR substrates by interacting with ATR-ATRIP and/or
other proteins at the damage sites. Alternatively, the Rad17 and 9-1-1 com-
plexes might stimulate the kinase activity of ATR on DNA. Although studies
using Xenopus extracts suggested that ATR can be activated on DNA (Ku-
magai et al. 2004), how ATR is stimulated by DNA or its regulators on DNA
remains to be elucidated.
2.3
Processing of DNA Lesions
Single-stranded DNA is a common DNA structure generated by DNA repli-
cation and many types of DNA repair processes. The discovery that ssDNA
plays a crucial role in the activation of ATR provides a plausible explanation
for why ATR can respond to different types of DNA damage. Furthermore, it
clearly suggests that the processing of DNA lesions plays an important role in
the activation of ATR. Many cellular processes involved in the maintenance

72 X.H. Yang · L. Zou
of genomic stability, such as DNA replication, DNA repair, and chromatin
modulation, have been implicated in the processing of DNA lesions. Thus, the
activation of ATR appears to be an effective way to integrate the DNA dam-
age signals generated by a variety of processes monitoring the integrity of the
genome.
During S-phase of the cell cycle, the entire genome is duplicated by DNA
replication forks. When replication forks encounter certain types of DNA
damage or interference, they will activate the ATR-mediated checkpoint.
Using yeast and Xenopus extracts, it has been shown that the activation of
checkpoint by DNA damage generated by UV (ultraviolet light) or MMS
(methyl methanesulfonate), a DNA-alkylating agent, is primarily through
mechanisms dependent upon DNA replication (Lupardus et al. 2002; Stokes
et al. 2002; Tercero et al. 2003). Furthermore, direct inhibition of DNA synthe-
sis by aphidicolin, a DNA polymerase inhibitor, or HU, an inhibitor of dNTP
synthesis, also elicits the ATR-mediated checkpoint. Thus, in addition to du-
plicating the genome, replication forksalsofunctiontoscanthegenomefor
DNA damage or other types of interference, and to translate these replica-
tion stresses into DNA structures that can be recognized by the checkpoint
sensors. Several studies using human, Xenopus,andyeastsystemshavesug-
gested that increased amounts of RPA-ssDNA are commonly induced at the
replication forks encountering various types of DNA damage and interfer-
ence (Tanaka and Nasmyth 1998; Michael et al. 2000; Lupardus et al. 2002;
Zou and Elledge 2003). The accumulation of ssDNA at the stressed replica-
tion forks might be a result of the uncoupling of different fork components,
such as helicases and DNA polymerases. Interestingly, the recruitment of
ATRIP to RPA-ssDNA in vitro is dependent on the length of ssDNA (Zou and
Elledge 2003), suggesting a quantitative mechanism for monitoring the stress
that a replication fork encounters. Furthermore, elevated levels of DNA poly-
merase α were also observed on chromatin after DNA damage (Michael et al.

2000; Lupardus et al. 2002), suggesting that junctions of dsDNA/ssDNA might
also accumulate at the stressed forks and they might facilitate the functions of
Rad17 and 9-1-1 complexes.
DNA repair also plays important roles in the processing of DNA lesions
and the activation of checkpoint. In yeast, the DSBs generated by the HO en-
donuclease are recessed by exonucleases in the 5

-to-3

direction. The reces-
sion of DSBs requires Xrs2, the yeast homologue of the human Nbs1 protein,
and Exo1 (exonuclease I) (Nakada et al. 2004). The generation of ssDNA at
the ends of DSBs not only presents a signal for Mec1-Ddc2 activation, but
also creates an important DNA intermediate for homologous recombination
(Wang and Haber 2004). Furthermore, the recession of DSBs is controlled
by the cyclin-dependent kinase Cdk1 (Ira et al. 2004). Hence, the processing
of DSBs in yeast cells has presented a clear example how the activation of
checkpoint by a particular type of DNA damage is coupled to a specific DNA
repair pathway as well as the cell cycle regulatory apparatus. Certain types of
Checkpoint and Coordinated Cellular Responses to DNA Damage 73
DNA damage might be sensed by both replication-dependent and replication-
independent mechanisms. For example, in the cells arrested in G0, G1 or
G2, UV-induced DNA damage can activate the checkpoint in a NER (nu-
cleotide excision repair)-dependent manner (Neecke et al. 1999; O’Driscoll
et al. 2003). Moreover, the endonuclease and helicase involved in NER are
needed to process the DNA lesions into structures that can elicit checkpoint
responses (Giannattasio et al. 2004). In addition to processing DNA lesions,
many DNA repair proteins also physically interact with the checkpoint sen-
sors. For example, the NER protein Rad14 associates with the 9-1-1 complex
in yeast (Giannattasio et al. 2004), and the mismatch repair protein Msh2

binds to ATR in human cells (Wang and Qin 2003). The interactions among
the repair and checkpoint proteins might contribute to the recruitment of
checkpoint sensors to specific types of DNA damage.
Both DNA damage and the activation of checkpoint can lead to changes in
the chromatin structure at the sites of DNA damage (see discussion below). It
was recently reported that yeast histone H2A is phosphorylated by Mec1 and
Tel1 (the yeast homologue of ATM) in the vicinity of the HO-induced DSBs
(Shroff et al. 2004), and that the phosphorylated histone H2A functions to re-
cruit the INO80 chromatin remodeling complex to the DSBs (Morrison et al.
2004; van Attikum et al. 2004). Interestingly, the INO80 complex is required
for the extensive recession of DSBs (van Attikum et al. 2004), revealing a reg-
ulatory role of chromatin modulation in the processing of DNA lesions and
the control of checkpoint signaling.
2.4
MRN Complex and Activation of ATM and ATR
Unlike ATR that responds to a variety of types of DNA damage, ATM is
primarily involved in the responses to DSBs. The phosphorylation of ATM
substrates can be observed in a very short time after DNA damage, suggest-
ing that compared with ATR, the activation of ATM is less dependent on the
processing of DNA lesions. In human cells, the Mre11-Rad50-Nbs1 protein
complex (the MRN complex), a complex that exhibits both exo- and endonu-
clease activities in vitro (Paull and Gellert 1999), plays a crucial role in the
activation of ATM (Carson et al. 2003; Uziel et al. 2003; Kitagawa et al. 2004).
ATM fails to associate with chromatin and to phosphorylate many of its sub-
strates in the cells lacking functional Nbs1 or Mre11 (Stewart et al. 1999;
Carson et al. 2003; Uziel et al. 2003).
Recent studies by two laboratories have revealed important clues of how
ATM is activated by DNA damage. A study by Kastan’s laboratory reported
that ATM is autophosphorylated on serine 1981 after DNA damage (Bakkenist
and Kastan 2003) (Fig. 1a). Following this autophosphorylation, ATM disso-

ciates from its multimeric form to become monomers (Bakkenist and Kastan
2003). The monomerization of ATM appears to be an important step for its
74 X.H. Yang · L. Zou
activation. Intriguingly, the autophosphorylation of ATM can be induced by
very low doses of DNA damage or treatments disrupting chromatin structures
in the absence of detectable DSBs, leading to the hypothesis that ATM might
be activated by changes of chromatin structures (Bakkenist and Kastan 2003).
Although monomeric ATM is capable of phosphorylating non-DNA-bound
substrates such as p53, the phosphorylation of other substrates at the sites of
DSBs requires the MRN complex and Brca1 (Kitagawa et al. 2004). The mech-
anisms by which the autophosphorylation of ATM is regulated are yet to be
elucidated. Proteins including Nbs1, 53BP1, Mdc1, PP5, PP2A, and p18 might
be involved in the regulation of ATM autophosphorylation at various stages
of signaling (Mochan et al. 2003; Uziel et al. 2003; Ali et al. 2004; Goodarzi
et al. 2004; Park et al. 2005). Elevated kinase activity of ATM can be detected
in vitro after DNA damage (Canman et al. 1998). A study by Paull’s laboratory
demonstrated that the MRN complex stimulates the phosphorylation of ATM
substrates in vitro even in the absence of DNA (Lee and Paull 2004). Although
this study did not reveal the contribution of DSBs in the activation of ATM,
it clearly shows a direct role of the MRN complex in the stimulation of ATM.
Together these recent findings suggest that the activation of ATM is a multi-
step process that involves the autophosphorylation of ATM, the interactions
of ATM with other regulatory factors, and the localization of ATM to DSBs.
The MRN complex is not only important for the localization of ATM to DSBs,
but also critical for the activation of the kinase activity of ATM.
The MRN complex might also be involved in the activation of ATR (Car-
son et al. 2003; Pichierri and Rosselli 2004; Stiff et al. 2005). The budding and
fission yeast mutants lacking the MRN complex display defective checkpoint
responses after HU or MMS treatments (D’Amours and Jackson 2001; Chah-
wan et al. 2003). In human cells, like ATR, the MRN complex is implicated in

the checkpoint signaling elicited by cross-linked DNA (Pichierri and Rosselli
2004). Very recently, it was shown that in the cells lacking functional Nbs1,
ATR is unable to stably associate with RPA-ssDNA and to phosphorylate Chk1
after HU treatment (Stiff et al. 2005), raising the possibility that the MNR
complex might regulate the interaction between ATR-ATRIP and RPA-ssDNA,
and might contribute to the activation of ATR.
3
Transduction of DNA Damage Signals
After ATM and ATR are activated by DNA damage, they phosphorylate nu-
merous substrates including two downstream checkpoint kinases, Chk1 and
Chk2 (Liu et al. 2000; Matsuoka et al. 1998). Existing evidence suggests that
Chk2 is primarily phosphorylated by ATM in response to DSBs (Matsuoka
et al. 2000), whereas Chk1 is mainly phosphorylated by ATR after DNA dam-
Checkpoint and Coordinated Cellular Responses to DNA Damage 75
age interfering with DNA replication (Brown and Baltimore 2003; Cortez
2003). The phosphorylation of Chk1 and Chk2 by ATR and ATM are re-
quired for the activation of these kinases (Matsuoka et al. 2000; Zhao and
Piwnica-Worms 2001). Once activated, Chk1 and Chk2 can then phosphory-
late substrates such as p53, Brca1, and Cdc25A (Matsuoka et al. 1998; Hirao
et al. 2000; Zhang et al. 2004), providing another layer of regulation on the
cellular processes involved in the maintenance of genomic stability. Chk1 has
been shown to play crucial roles in the regulation of cell cycle progression and
the stability of DNA replication forks (Zhao et al. 2002; Zachos et al. 2003).
Chk2, one the other hand, appears to play a less significant role in arresting
the cell cycle but is important for the regulation of cell death (Hirao et al.
2002; Takai et al. 2002).
The phosphorylation of Chk1 and Chk2 by ATR and ATM is regulated by
several other proteins in vivo. For example, in addition to the MRN complex
(Buscemi et al. 2001), the phosphorylation of Chk2 requires 53BP1, Mdc1, and
Brca1 (Foray et al. 2003; Peng and Chen 2003; Wang et al. 2002). Similarly, in

addition to the Rad17 and 9-1-1 complexes, the phosphorylation of Chk1 re-
quires Claspin, Mcm7, Brca1, CtIP, and TopBP1 (Kumagai and Dunphy 2000;
Yarden et al. 2002; Yamane et al. 2003; Cortez et al. 2004; Lin et al. 2004; Tsao
et al. 2004; Yu and Chen 2004). Because these groups of proteins function to
mediate the DNA damage signals between ATM/ATR and C h k 1 /Chk2, they
were appropriately termed “mediators”. However, the exact biochemical ac-
tivities of these proteins in checkpoint signaling are not known. The mediator
proteins do share several features that might be important for their func-
tions in signaling. First, most if not all mediators are present at the sites of
DNA damage or stalled replication forks. Some mediators, such as Claspin
and Mcm7, are probably components of DNA replication forks (Chini and
Chen 2003; Lee et al. 2003), whereas others, such as Mdc1, 53BP1, and Brca1,
are recruited to sites of DNA damage (Scully et al. 1997; Schultz et al. 2000;
Goldberg et al. 2003; Lou et al. 2003; Peng and Chen 2003; Stewart et al. 2003;
Xu and Stern 2003). Second, most mediator are themselves substrates of ATM
or ATR after DNA damage (Tibbetts et al. 2000; Chini and Chen 2003; Stew-
art et al. 2003). Third, many mediators possess phospho-serine/threonine-
binding motifs, such as the BRCT and the FHA motifs (Durocher et al. 1999;
Manke et al. 2003; Rodriguez et al. 2003; Yu et al. 2003). It has been shown that
certain mediators can associate with each other or with downstream kinase
in a phosphorylation-dependent manner (Kumagai and Dunphy 2003; Yu and
Chen 2004), suggesting that the damage-induced phosphorylation of these
mediators and their phospho-Ser/Thr-binding motifs might be involved in
organizing the protein complex at the sites of DNA damage and/or recruiting
downstream kinases. Consistent with the idea that mediators are important
for the recruitment of downstream kinases to ATM/AT R, cer tain mediators
in budding and fission yeast can be partially bypassed by fusing the down-