Ubc9- and mms21-mediated sumoylation counteracts recombinogenic events at damaged replication forks.
ABSTRACT The Ubc9 SUMO-conjugating enzyme and the Siz1 SUMO ligase sumoylate several repair and recombination proteins, including PCNA. Sumoylated PCNA binds Srs2, a helicase counteracting certain recombination events. Here we show that ubc9 mutants depend on checkpoint, recombination, and replication genes for growth. ubc9 cells maintain stalled-fork stability but exhibit a Rad51-dependent accumulation of cruciform structures during replication of damaged templates. Mutations in the Mms21 SUMO ligase resemble the ubc9 mutations. However, siz1, srs2, or pcna mutants altered in sumoylation do not exhibit the ubc9/mms21 phenotype. Like ubc9/mms21 mutants, sgs1 and top3 mutants also accumulate X molecules at damaged forks, and Sgs1/BLM is sumoylated. We propose that Ubc9 and Mms21 act in concert with Sgs1 to resolve the X structures formed during replication. Our results indicate that Ubc9- and Mms21-mediated sumoylation functions as a regulatory mechanism, different from that of replication checkpoints, to prevent pathological accumulation of cruciform structures at damaged forks.
- [show abstract] [hide abstract]
ABSTRACT: The RecQ helicases are highly conserved in evolution and are required for maintaining genome stability in all organisms. In humans, loss of RecQ helicase function is associated with predisposition to cancer and/or premature ageing. Recent data show that RecQ helicases have several roles during S phase of the cell cycle, ranging from facilitating the resumption of DNA synthesis at sites of replication fork breakdown to resolving structures during the process of homologous recombination.Trends in Cell Biology 10/2003; 13(9):493-501. · 11.72 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Normal eukaryotic cells do not enter mitosis unless DNA is fully replicated and repaired. Controls called 'checkpoints', mediate cell cycle arrest in response to unreplicated or damaged DNA. Two independent Schizosaccharomyces pombe mutant screens, both of which aimed to isolate new elements involved in checkpoint controls, have identified alleles of the hus5+ gene that are abnormally sensitive to both inhibitors of DNA synthesis and to ionizing radiation. We have cloned and sequenced the hus5+ gene. It is a novel member of the E2 family of ubiquitin conjugating enzymes (UBCs). To understand the role of hus5+ in cell cycle control we have characterized the phenotypes of the hus5 mutants and the hus5 gene disruption. We find that, whilst the mutants are sensitive to inhibitors of DNA synthesis and to irradiation, this is not due to an inability to undergo mitotic arrest. Thus, the hus5+ gene product is not directly involved in checkpoint control. However, in common with a large class of previously characterized checkpoint genes, it is required for efficient recovery from DNA damage or S-phase arrest and manifests a rapid death phenotype in combination with a temperature sensitive S phase and late S/G2 phase cdc mutants. In addition, hus5 deletion mutants are severely impaired in growth and exhibit high levels of abortive mitoses, suggesting a role for hus5+ in chromosome segregation. We conclude that this novel UBC enzyme plays multiple roles and is virtually essential for cell proliferation.Journal of Cell Science 03/1995; 108 ( Pt 2):475-86. · 5.88 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Ubiquitination of proteins was previously shown to modulate various processes of DNA metabolism. PCNA, a processivity factor with essential functions in replication and repair, is modified with ubiquitin at K164. In addition, PCNA is sumoylated at K127 and K164. We found that the rad18delta mutation suppresses the temperature sensitivity of the polymerase delta mutants hys2-1 and cdc2-1 as well as the synthetic lethality of cdc2-1 pol32delta mutants, suggesting a role for Rad18 in modulating DNA replication. As Rad18 mediates ubiquitination of PCNA, we examined whether PCNA modifications affected its function in replication. Multicopy PCNA alleviated the replication defects of rfc5-1 strains, but not those of poldelta mutants. In contrast, multicopy PCNA-K164R had reduced ability to suppress the replication defects of rfc5-1, but alleviated those of poldelta mutants. The roles of sumoylated and ubiquitinated PCNA in rfc5-1 and hys2-1 mutants were addressed by using mutant backgrounds that selectively affected sumoylation (siz1delta), ubiquitination (rad18delta), polyubiquitination (rad5delta, mms2delta), or the ability of cells to perform translesion synthesis (polzetadelta, poletadelta). Our results are consistent with the idea that the Rad18/Rad5/Mms2 polyubiquitination pathway is important for replication completion, perhaps by promoting a template switch type of DNA synthesis.Genes to Cells 12/2004; 9(11):1031-42. · 2.73 Impact Factor
Counteracts Recombinogenic Events
at Damaged Replication Forks
Dana Branzei,1,* Julie Sollier,1Giordano Liberi,1Xiaolan Zhao,2Daisuke Maeda,3Masayuki Seki,3
Takemi Enomoto,3,4Kunihiro Ohta,5and Marco Foiani1
1FIRC Institute of Molecular Oncology Foundation and Department of Biomedical Sciences and Biotechnology,
Universita ` degli Studi di Milano, Via Adamello 16, 20139 Milan, Italy
2Molecular Biology Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
3Molecular Cell Biology Laboratory, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba-ku,
Sendai 980-8578, Japan
4Tohoku University 21st Century COE Program ‘‘Comprehensive Research and Education Center for Planning of Drug
Development and Clinical Evaluation,’’ Sendai, Miyagi 980-8578, Japan
5Genetic System Regulation Laboratory, RIKEN Discovery Research Institute, 2-1, Hirosawa, Wako, Saitama 351-0198, Japan
The Ubc9 SUMO-conjugating enzyme and the
Siz1 SUMO ligase sumoylate several repair
and recombination proteins, including PCNA.
Sumoylated PCNA binds Srs2, a helicase coun-
teracting certain recombination events. Here
we show that ubc9 mutants depend on check-
point, recombination, and replication genes
for growth. ubc9 cells maintain stalled-fork sta-
bility but exhibit aRad51-dependent accumula-
damaged templates. Mutations in the Mms21
SUMO ligase resemble the ubc9 mutations.
However, siz1, srs2, or pcna mutants altered
in sumoylation do not exhibit the ubc9/mms21
phenotype. Like ubc9/mms21 mutants, sgs1
and top3 mutants also accumulate X molecules
at damaged forks, and Sgs1/BLM is sumoy-
lated. We propose that Ubc9 and Mms21 act
in concert with Sgs1 to resolve the X structures
formed during replication. Our results indicate
that Ubc9- and Mms21-mediated sumoylation
functions as a regulatory mechanism, different
from that of replication checkpoints, to prevent
pathological accumulation of cruciform struc-
tures at damaged forks.
Unscheduled recombination events during S phase may
lead to chromosomal rearrangements, which are associ-
ated with human diseases and cancer. The replication
process is often fraught with danger, as DNA lesions, de-
pools (dNTPs) can impede replication-fork progression.
Stalled forks are potentially unstable and, when unpro-
tected, can collapse, rearrange, or break (Branzei and
Foiani, 2005). Increasing evidence suggests that broken
forks can undergo aberrant recombination events to
form unstable chromosomal translocations. The replica-
tion checkpoint controls stalled fork stability (Boddy and
Russell, 2001; Lopes et al., 2001; Sogo et al., 2002;
Tercero and Diffley, 2001) and replisome-fork association
(Cobb et al., 2003; Lucca et al., 2004), thus preventing the
pathological formation of recombinogenic structures such
as reversed forks and hemireplicated DNA regions.
Forks encountering lesions on the template experience
continuous rounds of pausing and restart. The mecha-
nisms that regulate fork restart under these conditions
are less well understood. Several studies have implicated
the RecQ-family helicases, which include Sgs1 in budding
yeast,Rqh1 infission yeast,andBLMand WRNsyndrome
helicases in human, in facilitating the resumption of DNA
synthesis at sites of replication-fork breakdown or dealing
with homologous recombination events (Khakhar et al.,
2003). However, unlike replication-checkpoint mutants,
sgs1 cells do not accumulate pathological structures
when forks stall in response to hydroxyurea treatment
but rather accumulate structures resembling pseudo-
double Holliday junctions or hemicatenane-like mole-
cules, specifically when forks encounter a damaged
template (Liberi et al., 2005). The sgs1 defect is consistent
with the in vitro finding that the BLM/Top3 complex dis-
solves hemicatenane-like structures (Wu and Hickson,
2003). These results suggest that the replication check-
pointandRecQ helicasescooperate toallowforkresump-
tion. Sgs1 (Wohlschlegel et al., 2004) as well as WRN and
BLM (Eladad et al., 2005; Kawabe et al., 2000) is modified
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 509
ization to repair foci is modulated by this modification
(Eladad et al., 2005).
Sumoylation is a multistep process mediated by E1, E2,
and E3 enzymes (Johnson, 2004). Sumoylation relies on a
single E2 conjugating enzyme, Ubc9, encoded by a highly
conserved, essential gene. Three E3 SUMO ligases, Siz1,
Siz2, and Mms21, have been identified so far in budding
yeast and have been shown to promote the sumoylation
of different targets. However, different studies indicate
that Ubc9 may mediate sumoylation of target proteins
alone, without the need for a SUMO ligase (see Mo and
Moschos, 2005). Consistent with this, Ubc9 has been
two-hybrid system. Moreover, crystallographic studies
suggest that E3 ligases, in certain cases, interact more
with the Ubc9-SUMO complex than the target substrate
by optimally fixing the SUMO terminal glycine residue in
a favorable orientation to the substrate’s target lysine
(Reverter and Lima, 2005). In addition, cells express iso-
peptidases that are responsible for both SUMO matura-
tion and deconjugation.
Several lines of evidence suggest that Ubc9-mediated
sumoylation is involved in maintaining genome integrity.
First, the level of expression of UBC9 was found to be in-
creased in several tumor cell lines (Mo and Moschos,
2005), and inactivating mutations of UBC9 and isopepti-
dases are lethal in mice (Nacerddine etal., 2005; Yamagu-
chi et al., 2005). Second, proteins implicated in preserving
genome integrity, such as p53, the RecQ helicases BLM
and WRN, PCNA, DNA topoisomerases I and II, Smc1,
Smc3, Smc5, and Smc6, have been shown to be SUMO
targets (Andrews et al., 2005; Gill, 2004; Johnson, 2004;
Potts and Yu, 2005; Zhao and Blobel, 2005). Third, muta-
tions in budding yeast UBC9 and fission yeast hus5, as
well as mutations in the Mms21 SUMO ligase and the
Ulp1 and Ulp2 isopeptidases, cause hypersensitivity to
DNA-damaging agents and replication inhibitors (al-Kho-
dairy et al., 1995; Hoege et al., 2002; Li and Hochstrasser,
1999, 2000; Maeda et al., 2004; Soustelle et al., 2004;
Zhao and Blobel, 2005). Fourth, the recombination pro-
teins Rad51 and Rad52 interact with SUMO or Ubc9 (Ho
et al., 2001; Shen et al., 1996), and finally, mutations in
ULP1 result in a strong hyperrecombination phenotype
and synthetic lethality in combination with mutations in
genes involved in the homologous recombination path-
way (Soustelle et al., 2004).
Recently, sumoylation has been also implicated in
counteracting recombination events during replication
(Papouli et al., 2005; Pfander et al., 2005). In particular, it
has been shown that PCNA, once sumoylated through
a process dependent on Ubc9 and Siz1, recruits Srs2,
a helicase known to disrupt Rad51 filaments and prevent
certain recombination events (Krejci et al., 2003; Papouli
et al., 2005; Pfander et al., 2005; Veaute et al., 2003). In
a rad18 mutant background context, the recombination
events prevented by Srs2 and sumoylated PCNA are
thought to promote repair, thus allowing cell survival in
response to damage, as shown by the increased resis-
tance of rad18 srs2 or rad18 siz1 cells to DNA-damaging
agents (Aboussekhra et al., 1992; Papouli et al., 2005;
Pfander et al., 2005; Rong et al., 1991).
In this report, we examined whether Ubc9-mediated
sumoylation is important for replication-fork integrity. We
found that ubc9 mutants complete the bulk of DNA repli-
cation but accumulate DNA lesions and checkpoint
signals; for normal viability, they require a functional
replication-checkpoint response and homologous recom-
bination. Replication-fork stability was not dramatically
affected in ubc9-1 cells, but replication of damaged tem-
plates occurred with pathological accumulation of cruci-
form structures in a Rad51-dependent manner. This phe-
notype was also observed in the SUMO ligase mutant
mms21, but not in siz1, siz2, or siz1 siz2 mutants deficient
in different subpathways of sumoylation. Consistently, the
accumulation of X-shaped structures was not seen in mu-
tants defective in PCNA sumoylation. The ubc9/mms21
mutants show a phenotype similar to sgs1 and top3 mu-
tants (Liberi et al., 2005), and we found that Sgs1 interacts
with Ubc9 and is sumoylated. However, the bulk of Sgs1
sumoylation does not depend on Mms21. Our results
suggest that Ubc9- and Mms21-mediated sumoylation,
together with the Sgs1/Top3 complex, counteracts the
accumulation of cruciform intermediates at replication
forks during replication resumption processes.
ubc9 Mutants Are Slow Growing in Combination with
Mutations in Replication and Homologous
Since the highly conserved Ubc9 SUMO-conjugating en-
zyme has been implicated in maintaining genome integ-
rity, we used budding yeast as an experimental system
to investigate whether Ubc9-mediated sumoylation is im-
portant for maintaining the integrity of replication forks.
We first analyzed cell survival and cell-cycle progression
that are known to inhibit DNA replication.
ubc9-1 cells, which carry the Pro69Ser mutation (Seu-
fert et al., 1995), exhibit temperature sensitivity and hyper-
sensitivity to hydroxyurea (HU) and methyl methanesulfo-
nate (MMS) treatments (data not shown and Hoege et al.,
2002; Maeda et al., 2004). These phenotypes were com-
the mouse UBC9 (mUBC9) gene (data not shown). Previ-
ous work showed that, at nonpermissive temperatures,
ubc9 mutants arrest with large-budded cells, short prea-
naphase spindles, and a single nucleus (Seufert et al.,
1995). We found that temperature sensitivity was also as-
sociated with increased phosphorylation of the Rad53
thus suggesting that ubc9-1 cells activate checkpoint sig-
naling. We then tested whether ubc9-1 cells exhibit cell-
cycle progression defects when released at the nonper-
missive temperature (35?C) from a nocodazole-induced
510 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
G2/Mblock (see FigureS1intheSupplemental Dataavail-
able with this article online). ubc9-1 mutants completed
the first two rounds of replication with kinetics similar to
wild-type (WT) cells but began to accumulate cells with
tus of replication intermediates formed during the first
round of replication at 35?C at the ARS305 origin was
identical in WT and ubc9 cells (Figure S1B) as assayed
by 2-dimensional (2D) gel electrophoresis (Figure S2).
Certain mutants defective in DNA replication accumu-
late chromosomal abnormalities that cause checkpoint
activation and a checkpoint-dependent cell-cycle block.
Since ubc9 cells accumulate checkpoint signals and
exhibit a G2 cell-cycle arrest at high temperatures, we ad-
sive temperatures and the cell-cycle arrest at restrictive
nase.MEC1 isanessential gene,andits essential function
is suppressed by deleting SML1 (Zhao et al., 1998). Thus,
mec1 sml1 cells are viable but checkpoint deficient. We
found that the viability of ubc9-1 mutants was severely af-
fected in a sml1 mec1 background (Figure 1A). Further-
more, while ubc9-1 mutants progressively accumulated
checkpoint activation and cell-cycle arrest with a 2C DNA
content, ubc9-1 sml1 mec1 mutants no longer accumu-
lated in G2 (Figure 1A). These results indicate that the
problems accumulating in ubc9 cells lead to checkpoint
activation, which in turn contributes to cell-cycle delay
and is important for viability.
To address which DNA repair pathway (see below) deals
with the DNA damagespontaneously arising in ubc9-1 mu-
tants, we analyzed the growth rate and viability of ubc9-1
rad5, ubc9-1 rad18, ubc9-1 rad1, ubc9-1 rad51, ubc9-1
rad55, ubc9-1 rad59, ubc9-1 mms2, ubc9-1 rad30, and
ubc9-1 rev7 double mutants. Out of these combinations,
only the viability of ubc9-1 rad51 and ubc9-1 rad55 was af-
was also required for survival of HU-treated ubc9-1 cells at
25?C (see below and data not shown). Hence, Rad51-
dependent homologous recombination is the main repair
growth conditions or in response to HU-treatment.
In addition, we found that ubc9-1cells exhibited syn-
thetic growth defects at 30?C when combined with dele-
tions in the POL32 gene, which encodes the third subunit
of DNA polymerase d and is required for polymerase
d processivity (Burgers and Gerik, 1998). ubc9-1 cells
also showed synthetic growth defects with deletions of
the RRM3 but not the CHL1 helicase gene. Rrm3 facili-
tates fork progression in the presence of replication
blocks (Ivessa et al., 2002). Finally, we examined interac-
tions with two helicases implicated in preventing recombi-
nation events during S phase (Fabre et al., 2002; Liberi
et al., 2005; Rong et al., 1991). Srs2, which interacts phys-
ically and genetically with Pol32 (Huang et al., 2000) was
synthetic sick with ubc9-1, and ubc9-1 strains carrying
a partial deletion of SGS1 (Onoda et al., 2000), which dis-
rupts the helicase domain and shows recombination and
DNA-damage sensitivity phenotypes similar to sgs1D
strains, grew slowly (Figure 1C and data not shown).
Taken together, these results suggest that at least part
of the genomic problems arising in ubc9-1 cells result
from faulty replication.
ubc9 Mutants Accumulate Cruciform Structures
Specifically at Damaged Forks
The finding that no abnormal replication intermediates
occurred in ubc9 cells replicating at high temperatures
in ubc9 cells is generated gradually. Furthermore, the pre-
vious genetic interactions (Figure 1) suggest that ubc9
cells experience problems during replication. To under-
stand the nature of these problems, we examined the sta-
tus of replication intermediates in conditions that specifi-
cally induce replication-fork stalling (HU) or intra-S DNA
We first tested by 2D gel and cell-cycle cytometry
whether HU-treated ubc9-1 cells exhibit replication-fork
abnormalities and whether they recover properly from an
HU block. HU causesa reversible fork arrest,and WT cells
wereabletoresumereplication oncetheHUblock wasre-
moved. Origin firing generates intermediates that, on 2D
gels, migrate as a bubble arc containing forks proceeding
bidirectionally, large Y molecules resulting from forks mi-
grating outsideoftherestrictionfragment, andspecialized
X-shaped sister chromatid junctions (SCJs) resembling
hemicatenanes (Figure S2). The kinetics of accumulation
of replication intermediates at ARS305 in ubc9-1 and
WT cells treated with HU was identical (Figure 2A). In ad-
dition, FACS analysis of cells released in fresh medium af-
ter being blocked in HU showed that ubc9-1 cells com-
pleted replication, albeit with a small delay as compared
to WT cells, but then accumulated irreversibly in G2 with
phosphorylated Rad53 (Figure 2B and data not shown).
By pulsed-field gel electrophoresis (PFGE), we confirmed
that, in ubc9 cells recovering from HU, the appearance of
the typical chromosomal bands corresponding to fully
replicated chromosomes correlated with the completion
of S phase observed by FACS analysis (data not shown).
The status of replication intermediates and the cell-cycle
profiles of ubc9 cells replicating or recovering from HU
(Figure 2) suggest that stalled-fork stability is not signifi-
cantly impaired by the ablation of the Ubc9/SUMO path-
way butratherthat,inubc9cells,the problems mightarise
at late steps during replication, such as during replication
termination or postreplicative repair.
We then analyzed the ability of ubc9 cells to deal with
MMS-induced DNA damage; unlike forks encountering
HU-induced DNA lesions, forks encountering MMS-
induced DNA lesions need to undergo extensive restart to
resume replication. By PFGE, we found that, following
a transient MMS pulse, ubc9 mutants were delayed in re-
mulated in G2 (Figure S3), thus suggesting that ubc9 cells
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 511
Figure 1. ubc9-1 Mutants Accumulate Checkpoint Signals and Interact Functionally with Replication and Recombination Repair
analyzed by western blot for Rad53 phosphorylation. Strains Y0002 (WT), HY0512 (sml1), HY0514 (sml1 mec1), Y0174 (ubc9-1), HY0513 (ubc9-1
sml1), and HY0515 (ubc9-1 sml1 mec1) were grown in YPD plates for 2 days. Liquid cultures of the same strains were analyzed by FACS.
512 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
cificrepairpathway (Figure 3A).Mutations ingenes such as
RAD51 and RAD1 increased the MMS sensitivity of ubc9-1
sitivity of ubc9-1 only slightly, if at all (Figure 3B). We there-
fore conclude that the MMS sensitivity associated with the
ubc9-1 mutation is unlikely to be due primarily to defects
in homologous recombination, excision repair, and transle-
sion synthesis pathways. However, we found that the ubc9
(B) Strains Y0002 (WT), Y0174 (ubc9-1), HY0521 (ubc9-1 rad18), HY0517 (ubc9-1 rad5), HY0530 (ubc9-1 rad51), HY0532 (ubc9-1 rad55), HY0534
(ubc9-1 rad59), HY0506 (pol32), and HY0507 (ubc9-1 pol32) were streaked and grown in YPD plates for 2 days; strains Y0002 (WT), Y0174 (ubc9-
1), HY0503 (ubc9-1 sgs1), HY0505 (ubc9-1 srs2), HY0509 (ubc9-1 rrm3), and HY0511 (ubc9-1 chl1) were streaked and grown for 3 days. The growth
rate of strains HY0521 (ubc9-1 rad18), HY0517 (ubc9-1 rad5), HY0530 (ubc9-1 rad51), HY0532 (ubc9-1 rad55), HY0534 (ubc9-1 rad59), HY0503
(ubc9-1 sgs1), HY0505 (ubc9-1 srs2), HY0509 (ubc9-1 rrm3), and HY0511 (ubc9-1 chl1) was compared to the growth rate of single mutants in inde-
pendent experiments (data not shown).
Figure 2. ubc9-1 Mutants Are Proficient in Maintaining Fork Stability but, after a Transient HU Treatment, Accumulate in G2
(A) Wild-type (Y0002) and ubc9-1 (Y0174) strains were synchronized with nocodazole and released into medium containing HU. DNA samples were
cut with HindIII and EcoRV and analyzed by 2D gel with probes against ARS305.
3 hr and then released into YPD.
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 513
rad18 mutants are as resistant to MMS treatment as ubc9
cells at both 25?C and 30?C (Figure 3C). This is somewhat
expected, as in ubc9-1 cells, PCNA is not sumoylated,
and pcna-sumo mutants also suppress the rad18 damage
sensitivity (Hoege et al., 2002; Papouli et al., 2005; Pfander
et al., 2005). We note that the ubc9-1 mutation suppressed
the MMS sensitivity, but not the HU sensitivity, of rad5 and
rad18 mutants (Figure 3C and data not shown). These find-
ings, together with the observation that ubc9 and ubc9
rad18 cells require Rad51 for growth (Figure 3B and data
not shown), suggest that the MMS sensitivity of rad5/
rad18 mutants is alleviated by the ubc9-1 mutation through
derepression of recombination. However, it is not clear
whether this relief reflects a postreplicative event or
whether the Ubc9/SUMO pathway actively functions at
damaged replication forks, during DNA synthesis, to
To address these issues, we analyzed replication forks
arising from ARS305 and invading adjacent genomic
regions by 2D gels in MMS-treated ubc9-1 and WT cells
(Figure S2 and Figure 4A). Cells were released from a G2
block in MMS-containing medium. The timing of the
appearance of bubbles, large Ys, and SCJs at ARS305
of X molecules. This abnormal accumulation of X mole-
cules also occurred at those forks invading ARS305-adja-
cent regions (Figure 4A and Figure S4).
The Abnormal Status of Replication Intermediates
of ubc9 Cells Resembles sgs1 Mutants
This ubc9-1 phenotype (Figure 4A) is reminiscent of the
one observed in MMS-treated sgs1 and top3 mutants
(Liberi et al., 2005). Like ubc9-1 cells, sgs1 cells show nor-
mal fork progression in response to HU blocks but accu-
mulate Rad51-dependent cruciform structures upon ex-
posure to MMS (Liberi et al., 2005) (see also Figure 5A).
We therefore tested whether the accumulation of X mole-
cules in MMS-treated ubc9-1 cells is dependent on
Rad51. In ubc9-1 rad51 cells, the X molecules appeared
Figure 3. The Main Repair Pathways and Their Epistatic Relationships with the ubc9-1 Mutation
(A) Schematic representation of the main repair pathways and the genes implicated.
(B) Y0002 (wt), Y0174 (ubc9-1), HY0528 (rad51), HY0530 (ubc9-1 rad51), HY0522 (rad1), HY0523 (ubc9-1 rad1), HY0524 (rad30), HY0525 (ubc9-1
rad5), HY0526 (rev7), and HY0527 (ubc9-1 rev7) strains were grown to log phase and analyzed for MMS and HU sensitivity by spot assay.
(C) ubc9-1suppressestheMMS sensitivity of rad18 strains. The MMSsensitivity of Y0002 (wt), Y0174 (ubc9-1),HY0520(rad18), and HY0521 (ubc9-1
rad18) strains grown to log phase was assessed by spot assay.
514 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
at the same time as in WT cells but, in contrast to ubc9-1
mutants, no longer accumulated at late time points
(Figure 4B). In addition, the accumulation of cruciform
structures in ubc9-1 cells in the regions flanking ARS305
and at ARS301 still required Rad51 (Figure S4). Hence,
we conclude that the X molecule accumulation in MMS-
treated ubc9-1 mutants depends on Rad51. The previous
findings (Figure 4), together with the observation that
the accumulation of X molecules is comparable in MMS-
treated sgs1, ubc9-1, and ubc9-1 sgs1strains (Figure
5A), suggest that both Ubc9 and Sgs1 act in counter-
acting the accumulation of cruciform structures at dam-
The Mms21 SUMO Ligase Counteracts X Molecule
Accumulation at Damaged Forks
Recently, Siz1-mediated PCNA sumoylation has been
implicated in the recruitment of Srs2 to prevent certain re-
combination events at replicating chromosomes (Papouli
et al., 2005; Pfander et al., 2005). In a rad18 context, the
recombination events prevented by Srs2, Siz1, and su-
moylated PCNA are likely required to promote repair and
important for viability (Aboussekhra et al., 1992; Papouli
et al., 2005; Pfander et al., 2005; Rong et al., 1991). We
found that srs2 cells (Figure S5 and Liberi et al., 2005);
pol30-RR mutants encoding nonsumoylatable PCNA pro-
Figure 4. X Molecule Accumulation in ubc9-1 Mutants Following Exposure to MMS Depends on Rad51
(A) Y0002 (wt) and Y0174 (ubc9-1) were synchronized with nocodazole and released in medium containing MMS. DNA samples were cut with HindIII
and EcoRV and analyzed with the ARS305 and dx304 probes.
(B) Y0002 (wt), HY0528 (rad51), Y0174 (ubc9-1), and HY0530 (ubc9-1 rad51) strains were treated as in (A) and analyzed with the ARS305 probe.
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 515
sumoylation of PCNA was previously shown to be ablated
(Hoege et al., 2002; Papouli et al., 2005; Pfander et al.,
2005), did not accumulate cruciform structures at the
forks when treated with MMS. Thus, we conclude that
for preventing the accumulation of cruciform intermedi-
ates at damaged forks. Moreover, as the relative accumu-
lation of X molecules in MMS-treated ubc9-1, ubc9-1 srs2
(Figure S5), and ubc9-1 siz1 (data not shown) strains was
an anomalous Srs2 or sumoylated PCNA activity is the
primary cause of the cruciform structures that become
substrates for Sgs1.
we tested whether the ablation of any of the known SUMO
ligases would cause an accumulation of recombinogenic
structures at the forks. We found that, similar to siz1 cells,
siz2 mutants did not accumulate cruciform structures at
damaged forks; neither did a siz1 siz2 double mutant (data
not shown). Thus, we do not observe an accumulation of
Figure 5. X Molecule Accumulation in ubc9-1 Mutants Resembles sgs1 Cells and Is Not Due to Lack of PCNA Sumoylation
Y0002 (wt), HY0501 (sgs1),Y0174 (ubc9-1),and HY0503 (ubc9-1 sgs1) strains(A) and Y1190 (POL30)and Y1194 (pol30-RR)strains (B)were synchro-
nized in G2 with nocodazole at 25?C, then released in fresh medium containing MMS 0.033% at 30?C, and samples were collected at the indicated
time points. DNA samples were cut with HindIII and EcoRV and analyzed using the ARS305 probe.
516 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
hemicatenane-like structures at damaged forks in mutants
affecting the functionofproteinssuchasPCNA (Figure5B),
Takahashi et al., 2006). However, like ubc9 cells, mms21
mutants, in which the C-terminal Siz/PIAS domain required
for the SUMO ligase activity is disrupted (Andrews et al.,
2005; Prakash and Prakash, 1977; Zhao and Blobel,
2005), accumulate X molecules at damaged replication
forks in a Rad51-dependent manner (Figure 6A and Fig-
ure S6). These results are consistent with findings that
to damaging agents, including MMS (Andrews et al., 2005;
Prakash and Prakash, 1977; Zhao and Blobel, 2005), and
that this hypersensitivity requires the SUMO ligase activity
which carries two point mutations (C200A and H202A) af-
fecting the catalytic SUMO ligase activity (Andrews et al.,
2005), also showed accumulation of X molecules at
damaged forks (Figure 6B). This result suggests that
Mms21-mediated sumoylation regulates the process that
counteracts cruciform-structure accumulation during DNA
Sgs1 Is Sumoylated, and the Bulk of Sgs1
Sumoylation Does Not Require Mms21
Taking into consideration that BLM/Top3 resolves hemi-
catenane-like structures (Wu and Hickson, 2003); that
SGS1 overexpression promotes the resolution of the cru-
ciform intermediates accumulating in MMS-treated sgs1
mutants (Liberi et al., 2005); and that X-shaped structures
accumulate in MMS-treated sgs1, ubc9, and mms21 cells
in a similar manner (Figure 5A; Figures 6A and 6B), it is
reasonable to think that Ubc9- and Mms21-mediated su-
moylation affects the activity of the Sgs1 complex—either
directly, through Sgs1-mediated sumoylation, or indi-
rectly, by means of sumoylation of other targets, which
could ultimately affect Sgs1 localization, recruitment, pro-
tein interaction, and/or activity. Interestingly, sumoylation
of BLM, the human ortholog of Sgs1, has been shown
to affect its function in DNA repair and maintenance of
genome integrity, likely by modulating BLM recruit-
ment to PML bodies (Eladad et al., 2005). Furthermore,
Sgs1 has been reported to be a SUMO target by a proteo-
mic study using a mass-spectrometry-based approach
(Wohlschlegel et al., 2004). We therefore addressed
whether Sgs1 is indeed a SUMO target and whether its
sumoylation depends upon Ubc9 and Mms21. We found
that Sgs1 interacts with Ubc9 and SUMO in the yeast
two-hybrid system and that the interaction with SUMO
is dependent on the terminal pair of glycines in yeast
SUMO, which are involved in forming thioester bonds
with the substrates (Figure 6C), suggesting that Sgs1
might be sumoylated. Furthermore, we could immunopre-
cipitate Sgs1 and its SUMO-modified forms from strains
having the endogenous Sgs1 tagged at its C terminus
with 3HA or 3FLAG (Figure 6D). The Sgs1 sumoylation is
dependent on Ubc9, as no SUMO-modified forms of
Sgs1 were visible in ubc9-1 mutants (Figure 6D). Sgs1 su-
diminished in the mms21 mutation context, under either
normal or DNA-damaging conditions (Figure 6D). Al-
though we cannot formally rule out, at least at this stage,
that certain site-specific sumoylation events depend on
Mms21, these results suggest that Mms21 is not respon-
sible for the bulk of Sgs1 sumoylation. However, taking
into account the previous results (Figures 6A and 6B),
the most likely scenario is that Mms21 has other targets,
whose sumoylation is important in either promoting or
cooperating with Sgs1 in this process.
In this study, we analyzed the role of the Ubc9/SUMO
pathway in maintaining the integrity of replicating chromo-
somes. Based on our findings, we can conclude that, dur-
ing normal chromosome replication, ubc9-1 cells occa-
sionally accumulate lesions that trigger a checkpoint
response and require the homologous recombination ap-
paratus for repair. The nature and location of these lesions
are still elusive. These events are unlikely to occur at forks
in normal regions, as we could not detect any abnormali-
ties at the ARS305 origin in untreated ubc9-1 cells
(Figure S1). However, it might be impossible to detect
rare events by 2D gel analysis. We also analyzed whether
replication of ribosomal DNA (rDNA) regions occurs nor-
mally, but we could not detect any significant differences
in the status of replication intermediates between ubc9
and WT cells at these loci (data not shown).
We provide evidence suggesting that ubc9 mutants ex-
perience problems during replication. However, ubc9
cells are not impaired in maintaining the stability of stalled
forks in response to HU treatment (Figure 2A). We con-
firmed that forks also pause normally in untreated ubc9
cells at genomic locations containing replication-fork bar-
riers and pausing sites such as rDNA and tRNA (data not
shown). However, ubc9 cells fail to recover properly
from the extensive replication pausing caused by HU
treatment (Figure 2B) due to the generation of DNA dam-
age and checkpoint signals that cause a cell-cycle block
in G2. It is possible that a small proportion of replication
forks in ubc9-1 cells fail to efficiently restart during recov-
ery from HU. The synthetic interaction caused by the
RRM3 deletion in combination with the ubc9 mutation is
consistent with this idea. We envisage that ubc9 cells
recovering from HU blocks might occasionally generate
DNA breaks due to faulty replicative restart; alternatively,
replication of slow zones or certain steps of replication
termination might be affected in ubc9 cells (Figure 7A).
Our data indicate that Ubc9 controls some crucial
stepsrequiredforreplication ofdamagedtemplates. ubc9
mutants accumulate cruciform structures at MMS-
damaged forks through a Rad51-dependent process (Fig-
ure 4). This phenotype does not resemble the phenotypes
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 517
(A and B) Wild-type (W303 1-A), siz1 (HY0628), siz2 (HY0634), and mms21-SP (HY0638) strains (A) were synchronized in G1 with a factor at 25?C, and
wild-type (W303 1-A), mms21-11 (FY1012), and mms21-CH (FY1003) strains (B) were synchronized in G2 with nocodazole at 25?C and then released
518 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
of srs2 (Liberi et al., 2005; Figure S5), siz1 (Figure 6A), or
the pcna-sumo mutant pol30-K127R/K164R (Figure 5B).
These observations suggest that sumoylation of PCNA
and its interaction with Srs2 may not be crucial in hinder-
ing cruciform structures from accumulating at replication
forks in MMS-treated cells or that Srs2, together with
sumoylated PCNA, acts at damaged forks either infre-
quently or transiently, and consequently, its ablation
does not result in a large accumulation of X molecules.
Instead, our data support the idea that, in response to
DNA damage, the Ubc9- and Mms21-dependent sumoy-
lation processes ultimately influence the ability of Sgs1/
Top3 to prevent the accumulation of recombinogenic
structures on replicating chromosomes. First, MMS-
treated ubc9 cells behave like sgs1 and top3 mutants
(Figure 4; Liberi et al., 2005). Second, Sgs1/BLM is su-
moylated (Eladad et al., 2005; Wohlschlegel et al., 2004;
Figures 6C and 6D). Third, mms21 mutants, which affect
a subpathway of sumoylation, behave like ubc9 and
sgs1 cells (Figure 4; Figure 5A; Figure 6A and 6B; Fig-
ure S6). Finally, the phenotype of sgs1 appears to be
unique among DNA helicases, as mph1 (W. Carotenuto
and G.L., unpublished data), srs2, and rrm3 mutants do
not show the sgs1 phenotype (data not shown). Our find-
ings that the bulk of Sgs1 sumoylation depends on Ubc9
but not on the Mms21 ligase suggest either that sumoyla-
tion is directly mediated by the Ubc9 protein, with which
Sgs1 physically interacts (Figure 6C), or that another li-
gase is involved in this process. Furthermore, the results
showing that the ability of Mms21 to function as a
SUMO ligase is critical for the regulation of X molecule ac-
cumulation at damaged forks (Figures 6A and 6B) strongly
suggest that Mms21 has other targets, whose sumoyla-
tion is expected to promote crucial steps in this process
such as protein localization, protein-protein interaction,
or protein association with DNA. These sumoylation pro-
cesses may ultimately affect Sgs1 functionality or be
nane-like structures formed during replication (Figure 7A).
A crucial, although likely not exclusive, target may be the
Smc5-6 complex, which issumoylatedby Mms21 and ge-
netically interacts with Sgs1 (Miyabe et al., 2006; Mori-
kawa et al., 2004). It will be an important future task to
determine the Mms21-dependent targets and how they
interact or cooperate with Sgs1 and Top3 to allow the res-
olution of the hemicatenane-like structures, most likely
in fresh medium containing MMS 0.033% at 30?C. The DNA from the samples collected at the indicated time points was cut with HindIII and EcoRV
and analyzed using the ARS305 probe.
(C) BD-SGS1 in combination with AD-ySUMO, AD-ySUMO-G, AD-yUBC9, and pGAD424 (vector) was transformed in Y190 strains, and the interac-
tions between the proteins fused to the BD and AD domains was assayed by b-gal assay.
(D) Sgs1 is sumoylated. Wild-type (HY0608, HY0661), ubc9-1 (HY0607), mms21-11 (FY1014), and mms21-SP (HY0672) strains containing chromo-
somal 3HA- or 3FLAG-tagged Sgs1 were grown at 30?C to early log phase, and half of the cultures were treated with MMS 0.3% for 2 hr. The tagged
proteinswerepurifiedbyimmunoprecipitationasdescribed inZhao and Blobel(2005)usinganti-HAor anti-FLAG antibodies,respectively.Theeluate
was analyzed by western blot analysis using anti-SUMO and anti-HA or anti-FLAG antibodies, respectively.
Figure 7. The Role of Ubc9/Mms21-Me-
diated Sumoylation and the Checkpoint
Pathway in Preventing Recombinogenic
Structures from Accumulating during
(A) Schematic model for the role of the Ubc9/
Mms21/SUMO pathway in mediating the turn-
over of hemicatenane-like structures during
replication-bypass processes of damaged
templates and possibly when forks converge
during replication termination. The triangles
represent the Sgs1/Top3 complex.
(B) The two regulatory pathwayspreventing the
accumulation of recombinogenic structures
during replication. The replication-checkpoint
kinase Rad53 prevents reversed fork formation
at stalled forks, and the Ubc9/Mms21 sumoy-
lation pathway, in cooperation with Sgs1/
Top3, counteracts the accumulation of hemi-
catenane-like structures at damaged forks.
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 519
formed during damage bypass or gap-filling processes
Based on the characterization of the cruciform struc-
tures accumulating at damaged forks through a Rad51-
dependent process (Liberi et al., 2005), it has been
hypothesized that these structures represent pseudo-
double Holliday junctions in which nascent chains are
partially engaged into plectonemic pairings and the
corresponding parental tracts into paranemic junctions
(Figure 7A). This pathological accumulation of hemicate-
nane-like structures is likely to have tremendous conse-
quences for genomic stability, as their resolution might
stimulate recombination and genome rearrangements,
and failure to resolve them is expected to affect chromo-
some condensation, decatenation, and segregation pro-
cesses. Indeed, both sgs1 and ubc9/mms21/SUMO mu-
tants show defects in these processes (al-Khodairy
et al., 1995; Prakash and Prakash, 1977; Watt et al.,
1995), and recently, Ubc9 knockout mice were reported
to die at the early postimplantation stage with major chro-
mosome condensation and segregation aberrations (Na-
cerddine et al., 2005). We note that the role proposed for
Sgs1/Top3 in resolving the hemicatenanes formed by
quired for the resolution of hemicatenanes when two forks
converge at the end of replication (Figure 7A). Indeed,
Sgs1 and Top3 have been implicated in replication termi-
nation (Rothstein and Gangloff, 1995; Wang, 1991). With
this in mind, we speculate that, in addition to the prob-
lems encountered following exposure to DNA-damaging
agents, some of the genomic problems arising during nor-
mal replication in ubc9-1 cells could result from inefficient
resolution of hemicatenanes when the replicons fuse to-
gether (Figure 7A). It is expected that these structures
would occasionally give rise to DNA breaks and genomic
aberrations and that homologous recombination would
have to be involved to solve the termination problem in
ubc9 cells. Indeed, recombination is important for viability
in ubc9 mutants (Figure 1B) and has been already impli-
cated in replication termination (Horiuchi et al., 1994).
Considering thatubc9mutantsresemble sgs1cells with
regard to their accumulation of cruciform structures; that
both the ubc9 and sgs1 phenotypes are complemented
by the mammalian UBC9 and BLM genes, respectively;
and that BLM and Sgs1 are SUMO targets, it would not
bility disorders result from mutations in genes implicated
in the Ubc9/Mms21-mediated SUMO pathway.
In conclusion, this study identifies the Ubc9/Mms21/
SUMO posttranslational modification pathway as a novel
regulatory mechanism that prevents cruciform structures
from accumulating during replication (Figure 7B) and elu-
cidates some ofthe crucialeffectors ofthisprocess. Thus,
in addition to the roles of the Mec1/Rad53 replication
checkpoints (Branzei and Foiani, 2005), this is the second
pathway that opposes the accumulation of recombino-
genic structures (and consequently, genome rearrange-
ments) at the forks during DNA synthesis (Figure 7B). In-
terestingly, the mechanisms through which the Mec1/
Rad53 checkpoint and the Ubc9/SUMO pathway act to
prevent recombinogenic events during replication are dif-
ferent and can be seen from the outcomes: While the rep-
lication checkpoint counteracts fork reversal at stalled
forks (Branzei and Foiani, 2005; Lopes et al., 2001; Sogo
et al., 2002), ubc9, mms21, and sgs1/top3 mutants accu-
mulate pseudo-double Holliday junctions resembling
hemicatenanes at damaged forks (Figure 7B; Liberi
et al., 2005). Thus, cells have evolved at least these two
regulatory pathways to counteract the accumulation of re-
combinogenic structures during replication both when
forks stall and when they encounter damaged templates
(Figure 7B). It will be interesting in the future to determine
the effectors of these two pathways that prevent the path-
ological accumulation of these cruciform structures and
whether there is crosstalk between them.
Yeast Strains and Plasmids
The yeast strains used in this study were derivatives of DF5 (Seufert
et al., 1995) or W303 (Thomas and Rothstein, 1989); the relevant geno-
types are shown in Table S1. Combinations of deletion mutants were
constructed using standard yeast techniques (Branzei et al., 2004).
The mms21-SP allele was created by replacing the C-terminal Siz/
PIAS domain of Mms21 with the hphMX4 marker to give a mutation
genetically equivalent to that of mms21-11, containing a transposon
insertion in the Siz/PIAS domain (Zhao and Blobel, 2005). For con-
struction of the mms21-CH mutant, the wild-type MMS21 gene was
subcloned inpET15Band the C200A and H202A mutations were intro-
duced using the QuikChange Site-Directed Mutagenesis Kit (Strata-
gene), and the mutations were confirmed by sequencing. The
mms21-CH allele was integrated at the chromosomal locus using the
method previously described (Reid et al., 2002). For construction of
BD-SGS1, the SGS1 gene was amplified from the genome with oligos
containing BamHI and SalI sites and subcloned into pGBD-C3. pGAD-
contains the yeast SUMO gene, SMT3, subcloned into the SalI site
of pGAD424 (a gift from H. Shinagawa). The terminal GG residues
of Smt3 were mutated by using the QuikChange Site-Directed Muta-
genesis Kit to obtain pGAD-ySUMO-G. The mutant ySUMO protein
has G97D, and the G98 is mutated to a stop codon. The mutations
were confirmed by sequencing. YCp-RAD53-HA was described previ-
ously (Sugimoto et al., 1997).
Protein Techniques and Antibodies
For western blot analysis, yeast protein extracts were prepared by the
TCA method as described previously (Knop et al., 1999). The anti-
bodies used to detect Rad53 phosphorylation were rabbit anti-HA
antibody (MBL) when the YCp-RAD53-HA was used (Figure 1A) or
monoclonal mouse EL7 antibody (a gift from A. Pellicioli).
The immunoprecipitation and immunoblotting experiments con-
ducted to address the sumoylation of Sgs1, tagged at its C terminus
with 3HA or 3FLAG epitope, were performed largely as described in
Zhao and Blobel (2005), except that the antibodies used were 12CA5
(Covance) for HA and the M2 monoclonal antibody (Sigma) for FLAG.
Y190 cells were transformed with combinations of two plasmids, one
containing the SGS1 gene fused to the binding domain (BD) of GAL4
and the other containing one of the indicated genes fused to the acti-
vation domain (AD) of the GAL4 protein. The transformants were
520 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.
analyzed for b-galactosidase activity by a filter assay according to the
protocol described in the Clontech manual.
Growing Conditions, Cell-Cycle Arrests, Drug Treatments,
and Spot Assays
Unless otherwise indicated, strains were grown at 25?C in YPD me-
dium containing glucose at 2% w/v. Cell synchronization was per-
formed by adding a factor to a final concentration of 3 mg/ml or noco-
dazole to a final concentration of 10 mg/ml together with DMSO to
a final concentration of 1% v/v for about 2 hr and evaluating the per-
centage of unbudded (a factor) or large-budded (nocodazole) cells in
the culture. The release from a factor was performed as previously
described (Liberi et al., 2005). The release from nocodazole arrest
was performed by centrifugation, washing in YPD containing 1% v/v
DMSO, and resuspension of cells in fresh medium. The release was
performed at 30?C unless otherwise indicated, and MMS and HU
were used at final concentrations of 0.033% v/v and 0.2 M, respec-
tively. Analysis of drug sensitivity and growth ability by spot assay
was carried out as previously described (Branzei et al., 2004).
DNA Extraction, 2D Gel Technique, Pulsed-Field Gel
Electrophoresis, and Cell-Cycle Analysis
Purification of DNA intermediates inthe presence of CTAB, 2D gel pro-
as described (Liberi et al., 2005; Lopes et al., 2003). FACS and chro-
mosome analysis by PFGE analysis were performed as described
previously (Branzei et al., 2004).
Supplemental Data include one table and six figures and can be found
with this article online at http://www.cell.com/cgi/content/full/127/3/
We thank S. Jentsch, K. Sugimoto, H. Shinagawa, A. Pellicioli, S. Rah-
man, V. Guacci, C. Newlon, A. Lehmann, and all members of the Foiani
laboratory for various reagents, procedures, and helpful discussions.
This work was mainly supported by grants from the Associazione Ital-
iana per la Ricerca sul Cancro and the Association for International
Cancer Research to M.F. and D.B. We also acknowledge grants
from the European Union, Telethon Italy, and MIUR to M.F.; Cancer
Center Support Grant NCI P30 CA-08478-41 to X.Z.; grants-in-aid
for scientific research from the Ministry of Education, Culture, Sports,
Science and Technology of Japan to T.E. and K.O.; and a basic re-
search grant from the Bio-oriented Technology Research Advance-
ment Institution to K.O. D.B. did part of this work in RIKEN, in K.O.’s
gram and is currently supported by the Buzzati-Traverso Foundation.
Received: April 11, 2006
Revised: July 19, 2006
Accepted: August 24, 2006
Published: November 2, 2006
Aboussekhra, A.,Chanet,R.,Adjiri, A.,and Fabre,F.(1992). Semidom-
inant suppressors of Srs2 helicase mutations of Saccharomyces cer-
evisiae map in the RAD51 gene, whose sequence predicts a protein
with similarities to procaryotic RecA proteins. Mol. Cell. Biol. 12,
al-Khodairy, F., Enoch, T., Hagan, I.M., and Carr, A.M. (1995). The
Schizosaccharomyces pombe hus5 gene encodes a ubiquitin conju-
gating enzyme required for normal mitosis. J. Cell Sci. 108, 475–486.
Andrews, E.A., Palecek, J., Sergeant, J., Taylor, E., Lehmann, A.R.,
and Watts, F.Z. (2005). Nse2, a component of the Smc5-6 complex,
is a SUMO ligase required for the response to DNA damage. Mol.
Cell. Biol. 25, 185–196.
Boddy, M.N., and Russell, P. (2001). DNA replication checkpoint. Curr.
Biol. 11, R953–R956.
Branzei, D., and Foiani, M. (2005). The DNA damage response during
DNA replication. Curr. Opin. Cell Biol. 17, 568–575.
diated polyubiquitination of PCNA is implicated in replication comple-
tion during replication stress. Genes Cells 9, 1031–1042.
Burgers, P.M., and Gerik, K.J. (1998). Structure and processivity of
two forms of Saccharomyces cerevisiae DNA polymerase delta.
J. Biol. Chem. 273, 19756–19762.
Cobb, J.A., Bjergbaek, L., Shimada, K., Frei, C., and Gasser, S.M.
(2003). DNA polymerase stabilization at stalled replication forks re-
quires Mec1 and the RecQ helicase Sgs1. EMBO J. 22, 4325–4336.
Eladad, S., Ye, T.Z., Hu, P., Leversha, M., Beresten, S., Matunis, M.J.,
and Ellis, N.A. (2005). Intra-nuclear trafficking of the BLM helicase to
DNA damage-induced foci is regulated by SUMO modification. Hum.
Mol. Genet. 14, 1351–1365.
Fabre, F., Chan, A., Heyer, W.D., and Gangloff, S. (2002). Alternate
pathways involving Sgs1/Top3, Mus81/ Mms4, and Srs2 prevent for-
mation of toxic recombination intermediates from single-stranded
gaps created by DNA replication. Proc. Natl. Acad. Sci. USA 99,
Gill, G. (2004). SUMO and ubiquitin in the nucleus: different functions,
similar mechanisms? Genes Dev. 18, 2046–2059.
Ho, J.C., Warr, N.J., Shimizu, H., and Watts, F.Z. (2001). SUMO mod-
ification of Rad22, the Schizosaccharomyces pombe homologue of
the recombination protein Rad52. Nucleic Acids Res. 29, 4179–4186.
Hoege, C., Pfander, B., Moldovan, G.L., Pyrowolakis, G., and Jentsch,
S. (2002). RAD6-dependent DNA repair is linked to modification of
PCNA by ubiquitin and SUMO. Nature 419, 135–141.
Horiuchi, T., Fujimura, Y., Nishitani, H., Kobayashi, T., and Hidaka, M.
(1994). The DNA replication fork blocked at the Ter site may be an en-
trance for the RecBCD enzyme into duplex DNA. J. Bacteriol. 176,
Huang, M.E., de Calignon, A., Nicolas, A., and Galibert, F. (2000).
POL32, a subunit of the Saccharomyces cerevisiae DNA polymerase
delta, defines a link between DNA replication and the mutagenic
bypass repair pathway. Curr. Genet. 38, 178–187.
Ivessa, A.S., Zhou, J.Q., Schulz, V.P., Monson, E.K., and Zakian, V.A.
(2002). Saccharomyces Rrm3p, a 50to 30DNA helicase that promotes
replication fork progression through telomeric and subtelomeric DNA.
Genes Dev. 16, 1383–1396.
Johnson, E.S. (2004). Protein modification by SUMO. Annu. Rev.
Biochem. 73, 355–382.
Kawabe, Y., Seki, M., Seki, T., Wang, W.S., Imamura, O., Furuichi, Y.,
Saitoh,H., and Enomoto, T. (2000).Covalent modification of the Wern-
er’s syndrome gene product with the ubiquitin-related protein,
SUMO-1. J. Biol. Chem. 275, 20963–20966.
Khakhar, R.R., Cobb, J.A., Bjergbaek, L., Hickson, I.D., and Gasser,
S.M. (2003). RecQ helicases: multiple roles in genome maintenance.
Trends Cell Biol. 13, 493–501.
Knop, M., Siegers, K., Pereira, G., Zachariae, W., Winsor, B.,Nasmyth,
K., and Schiebel, E. (1999). Epitope tagging of yeast genes using
a PCR-based strategy: more tags and improved practical routines.
Yeast 15, 963–972.
Krejci, L., Van Komen, S., Li, Y., Villemain, J., Reddy, M.S., Klein, H.,
Ellenberger, T., and Sung, P. (2003). DNA helicase Srs2 disrupts the
Rad51 presynaptic filament. Nature 423, 305–309.
Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc. 521
Li, S.J., and Hochstrasser, M. (1999). A new protease required for cell-
cycle progression in yeast. Nature 398, 246–251.
Li, S.J., and Hochstrasser, M. (2000). The yeast ULP2 (SMT4) gene en-
codes a novel protease specific for the ubiquitin-like Smt3 protein.
Mol. Cell. Biol. 20, 2367–2377.
Liberi, G., Maffioletti, G., Lucca, C., Chiolo, I., Baryshnikova, A., Cotta-
Ramusino, C., Lopes, M., Pellicioli, A., Haber, J.E., and Foiani, M.
(2005). Rad51-dependent DNA structures accumulate at damaged
replication forks in sgs1 mutants defective in the yeast ortholog of
BLM RecQ helicase. Genes Dev. 19, 339–350.
Lopes, M., Cotta-Ramusino, C., Pellicioli, A., Liberi, G., Plevani, P.,
Muzi-Falconi, M., Newlon, C.S., and Foiani, M. (2001). The DNA repli-
cation checkpoint response stabilizes stalled replication forks. Nature
Lopes, M., Cotta-Ramusino, C., Liberi, G., and Foiani, M. (2003).
Branch migrating sister chromatid junctions form at replication origins
through Rad51/Rad52-independent mechanisms. Mol. Cell 12,
Lucca, C., Vanoli, F., Cotta-Ramusino, C., Pellicioli, A., Liberi, G.,
Haber, J., and Foiani, M. (2004). Checkpoint-mediated control of
replisome-fork association and signalling in response to replication
pausing. Oncogene 23, 1206–1213.
T. (2004). Ubc9 is required for damage-tolerance and damage-in-
duced interchromosomal homologous recombination in S. cerevisiae.
DNA Repair (Amst.) 3, 335–341.
Miyabe, I., Morishita, T., Hishida, T., Yonei, S., and Shinagawa, H.
(2006). Rhp51-dependent recombination intermediates that do not
generate checkpoint signal are accumulated in Schizosaccharomyces
pombe rad60 and smc5/6 mutants after release from replication
arrest. Mol. Cell. Biol. 26, 343–353.
Mo, Y.Y., and Moschos, S.J. (2005). Targeting Ubc9 for cancer ther-
apy. Expert Opin. Ther. Targets 9, 1203–1216.
Morikawa, H., Morishita, T., Kawane, S., Iwasaki, H., Carr, A.M., and
Shinagawa, H. (2004). Rad62 protein functionally and physically asso-
ciates with the Smc5/Smc6 protein complex and is required for chro-
mosome integrity and recombination repair in fission yeast. Mol. Cell.
Biol. 24, 9401–9413.
Nacerddine, K., Lehembre, F., Bhaumik, M., Artus, J., Cohen-Tan-
noudji, M., Babinet, C., Pandolfi, P.P., and Dejean, A. (2005). The
SUMO pathway is essential for nuclear integrity and chromosome
segregation in mice. Dev. Cell 9, 769–779.
Onoda, F., Seki, M., Miyajima, A., and Enomoto, T. (2000). Elevation of
sister chromatid exchange in Saccharomyces cerevisiae sgs1 disrup-
tants and the relevance of the disruptants as a system to evaluate mu-
tations in Bloom’s syndrome gene. Mutat. Res. 459, 203–209.
Papouli, E., Chen, S., Davies, A.A., Huttner, D., Krejci, L., Sung, P., and
mediated by recruitment of the helicase Srs2p. Mol. Cell 19, 123–133.
Pfander, B., Moldovan, G.L., Sacher, M., Hoege, C., and Jentsch, S.
(2005). SUMO-modified PCNA recruits Srs2 to prevent recombination
during S phase. Nature 436, 428–433.
Potts, P.R., and Yu, H. (2005). Human MMS21/NSE2 is a SUMO ligase
required for DNA repair. Mol. Cell. Biol. 25, 7021–7032.
Prakash, S., and Prakash, L. (1977). Increased spontaneous mitotic
segregation in MMS-sensitive mutants of Saccharomyces cerevisiae.
Genetics 87, 229–236.
terations in Saccharomyces cerevisiae using adaptamer-mediated
PCR. Methods Enzymol. 350, 258–277.
Reverter, D., and Lima, C.D. (2005). Insights into E3 ligase activity re-
vealed by a SUMO-RanGAP1-Ubc9-Nup358 complex. Nature 435,
Rong, L., Palladino, F., Aguilera, A., and Klein, H.L. (1991). The hyper-
gene conversion hpr5-1 mutation of Saccharomyces cerevisiae is an
allele of the SRS2/RADH gene. Genetics 127, 75–85.
Rothstein, R., and Gangloff, S. (1995). Hyper-recombination and
Bloom’s syndrome: microbes again provide clues about cancer.
Genome Res. 5, 421–426.
Seufert, W., Futcher, B., and Jentsch, S. (1995). Role of a ubiquitin-
conjugating enzyme in degradation of S- and M-phase cyclins. Nature
Shen, Z., Pardington-Purtymun, P.E., Comeaux, J.C., Moyzis, R.K.,
and Chen, D.J. (1996). Associations of UBE2I with RAD52, UBL1,
p53, and RAD51 proteins in a yeast two-hybrid system. Genomics
Sogo, J.M., Lopes, M., and Foiani, M. (2002). Fork reversal and ssDNA
accumulation at stalled replication forks owing to checkpoint defects.
Science 297, 599–602.
Soustelle, C., Vernis, L., Freon, K., Reynaud-Angelin, A., Chanet, R.,
Fabre, F., and Heude, M. (2004). A new Saccharomyces cerevisiae
strain with a mutant Smt3-deconjugating Ulp1 protein is affected in
DNA replication and requires Srs2 and homologous recombination
for its viability. Mol. Cell. Biol. 24, 5130–5143.
Stead, K., Aguilar, C., Hartman, T., Drexel, M., Meluh, P., and Guacci,
V. (2003). Pds5p regulates the maintenance of sister chromatid cohe-
sion and is sumoylated to promote the dissolution of cohesion.
J. Cell Biol. 163, 729–741.
Sugimoto, K., Ando, S., Shimomura, T., and Matsumoto, K. (1997).
Rfc5, a replication factor C component, is required for regulation of
Rad53 protein kinase in the yeast checkpoint pathway. Mol. Cell.
Biol. 17, 5905–5914.
Takahashi, Y., Yong-Gonzalez, V., Kikuchi, Y., and Strunnikov, A.
(2006). SIZ1/SIZ2controlof chromosometransmission fidelity is medi-
ated by the sumoylation of topoisomerase II. Genetics 172, 783–794.
Tercero, J.A., and Diffley, J.F. (2001). Regulation of DNA replication
fork progression through damaged DNA by the Mec1/Rad53 check-
point. Nature 412, 553–557.
Thomas, B.J., and Rothstein, R. (1989). Elevated recombination rates
in transcriptionally active DNA. Cell 56, 619–630.
Veaute, X., Jeusset, J., Soustelle, C., Kowalczykowski, S.C., Le Cam,
E., and Fabre, F. (2003). The Srs2 helicase prevents recombination by
disrupting Rad51 nucleoprotein filaments. Nature 423, 309–312.
Wang, J.C.(1991).DNA topoisomerases: why somany? J.Biol. Chem.
Watt, P.M., Louis, E.J., Borts, R.H., and Hickson, I.D. (1995). Sgs1:
II in vivo and is required for faithful chromosome segregation. Cell 81,
Wohlschlegel, J.A., Johnson, E.S., Reed, S.I., and Yates, J.R., III.
(2004). Global analysis of protein sumoylation in Saccharomyces cer-
evisiae. J. Biol. Chem. 279, 45662–45668.
Wu, L., and Hickson, I.D. (2003). The Bloom’s syndrome helicase sup-
presses crossing over during homologous recombination. Nature 426,
Yamaguchi, T., Sharma, P., Athanasiou, M., Kumar, A., Yamada, S.,
and Kuehn, M.R. (2005). Mutation of SENP1/SuPr-2 reveals an essen-
tial role for desumoylation in mouse development. Mol. Cell. Biol. 25,
Zhao, X., and Blobel, G. (2005). A SUMO ligase is part of anuclearmul-
tiprotein complex that affects DNA repair and chromosomal organiza-
tion. Proc. Natl. Acad. Sci. USA 102, 4777–4782.
Zhao, X., Muller, E.G., and Rothstein, R. (1998). A suppressor of two
essential checkpoint genes identifies a novel protein that negatively
affects dNTP pools. Mol. Cell 2, 329–340.
522 Cell 127, 509–522, November 3, 2006 ª2006 Elsevier Inc.