Excess single-stranded DNA inhibits meiotic double-strand break repair.
ABSTRACT During meiosis, self-inflicted DNA double-strand breaks (DSBs) are created by the protein Spo11 and repaired by homologous recombination leading to gene conversions and crossovers. Crossover formation is vital for the segregation of homologous chromosomes during the first meiotic division and requires the RecA orthologue, Dmc1. We analyzed repair during meiosis of site-specific DSBs created by another nuclease, VMA1-derived endonuclease (VDE), in cells lacking Dmc1 strand-exchange protein. Turnover and resection of the VDE-DSBs was assessed in two different reporter cassettes that can repair using flanking direct repeat sequences, thereby obviating the need for a Dmc1-dependent DNA strand invasion step. Access of the single-strand binding complex replication protein A, which is normally used in all modes of DSB repair, was checked in chromatin immunoprecipitation experiments, using antibody against Rfa1. Repair of the VDE-DSBs was severely inhibited in dmc1Delta cells, a defect that was associated with a reduction in the long tract resection required to initiate single-strand annealing between the flanking repeat sequences. Mutants that either reduce Spo11-DSB formation or abolish resection at Spo11-DSBs rescued the repair block. We also found that a replication protein A component, Rfa1, does not accumulate to expected levels at unrepaired single-stranded DNA (ssDNA) in dmc1Delta cells. The requirement of Dmc1 for VDE-DSB repair using flanking repeats appears to be caused by the accumulation of large quantities of ssDNA that accumulate at Spo11-DSBs when Dmc1 is absent. We propose that these resected DSBs sequester both resection machinery and ssDNA binding proteins, which in wild-type cells would normally be recycled as Spo11-DSBs repair. The implication is that repair proteins are in limited supply, and this could reflect an underlying mechanism for regulating DSB repair in wild-type cells, providing protection from potentially harmful effects of overabundant repair proteins.
- SourceAvailable from: Sarah Zanders[Show abstract] [Hide abstract]
ABSTRACT: In most organisms, the segregation of chromosomes during the first meiotic division is dependent upon at least one crossover (CO) between each pair of homologous chromosomes. COs can result from chromosome double-strand breaks (DSBs) that are induced and preferentially repaired using the homologous chromosome as a template. The PCH2 gene of budding yeast is required to establish proper meiotic chromosome axis structure and to regulate meiotic interhomolog DSB repair outcomes. These roles appear conserved in the mouse ortholog of PCH2, Trip13, which is also involved in meiotic chromosome axis organization and the regulation of DSB repair. Using a combination of genetic and physical assays to monitor meiotic DSB repair, we present data consistent with pch2Δ mutants showing defects in suppressing intersister DSB repair. These defects appear most pronounced in dmc1Δ mutants, which are defective for interhomolog repair, and explain the previously reported observation that pch2Δ dmc1Δ cells can complete meiosis. Results from genetic epistasis analyses involving spo13Δ, rad54Δ, and mek1/MEK1 alleles and an intersister recombination reporter assay are also consistent with Pch2 acting to limit intersister repair. We propose a model in which Pch2 is required to promote full Mek1 activity and thereby promotes interhomolog repair.Genetics 04/2011; 188(3):511-21. · 4.39 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Spo11 is the topoisomerase-like enzyme responsible for the induction of the meiosis-specific double strand breaks (DSBs), which initiates the recombination events responsible for proper chromosome segregation. Nineteen PCR-induced alleles of SPO11 were identified and characterized genetically and cytologically. Recombination, spore viability and synaptonemal complex (SC) formation were decreased to varying extents in these mutants. Arrest by ndt80 restored these events in two severe hypomorphic mutants, suggesting that ndt80-arrested nuclei are capable of extended DSB activity. While crossing-over, spore viability and synaptonemal complex (SC) formation defects correlated, the extent of such defects was not predictive of the level of heteroallelic gene conversions (prototrophs) exhibited by each mutant. High throughput sequencing of tetrads from spo11 hypomorphs revealed that gene conversion tracts associated with COs are significantly longer and gene conversion tracts unassociated with COs are significantly shorter than in wild type. By modeling the extent of these tract changes, we could account for the discrepancy in genetic measurements of prototrophy and crossover association. These findings provide an explanation for the unexpectedly low prototroph levels exhibited by spo11 hypomorphs and have important implications for genetic studies that assume an unbiased recovery of prototrophs, such as measurements of CO homeostasis. Our genetic and physical data support previous observations of DSB-limited meioses, in which COs are disproportionally maintained over NCOs (CO homeostasis).PLoS Genetics 10/2013; 9(10):e1003932. · 8.52 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: The Rad2/XPG family nuclease, Exo1, functions in a variety of DNA repair pathways. During meiosis, Exo1 promotes crossover recombination and thereby facilitates chromosome segregation at the first division. Meiotic recombination is initiated by programmed DNA double-strand breaks (DSBs). Nucleolytic resection of DSBs generates long 3' single-strand tails that undergo strand exchange with a homologous chromosome to form joint molecule (JM) intermediates. We show that meiotic DSB resection is dramatically reduced in exo1Δ mutants and test the idea that Exo1-catalyzed resection promotes crossing over by facilitating formation of crossover-specific JMs called double Holliday junctions (dHJs). Contrary to this idea, dHJs form at wild-type levels in exo1Δ mutants, implying that Exo1 has a second function that promotes resolution of dHJs into crossovers. Surprisingly, the dHJ resolution function of Exo1 is independent of its nuclease activities but requires interaction with the putative endonuclease complex, Mlh1-Mlh3. Thus, the DSB resection and procrossover functions of Exo1 during meiosis involve temporally and biochemically distinct activities.Molecular cell 12/2010; 40(6):1001-15. · 14.61 Impact Factor
Excess Single-Stranded DNA Inhibits Meiotic
Double-Strand Break Repair
Rebecca Johnson1[¤a, Vale ´rie Borde2[, Matthew J. Neale1[¤b, Anna Bishop-Bailey1, Matthew North1¤c,
Sheila Harris1¤d, Alain Nicolas2, Alastair S. H. Goldman1*
1 Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield, United Kingdom, 2 Institut Curie, Centre de Recherche, Recombinaison et Instabilite
Genetique UMR7147 CNRS Universite ´ P. et M. Curie, Paris, France
During meiosis, self-inflicted DNA double-strand breaks (DSBs) are created by the protein Spo11 and repaired by
homologous recombination leading to gene conversions and crossovers. Crossover formation is vital for the
segregation of homologous chromosomes during the first meiotic division and requires the RecA orthologue, Dmc1.We
analyzed repair during meiosis of site-specific DSBs created by another nuclease, VMA1-derived endonuclease (VDE), in
cells lacking Dmc1 strand-exchange protein. Turnover and resection of the VDE-DSBs was assessed in two different
reporter cassettes that can repair using flanking direct repeat sequences, thereby obviating the need for a Dmc1-
dependent DNA strand invasion step. Access of the single-strand binding complex replication protein A, which is
normally used in all modes of DSB repair, was checked in chromatin immunoprecipitation experiments, using antibody
against Rfa1. Repair of the VDE-DSBs was severely inhibited in dmc1D cells, a defect that was associated with a
reduction in the long tract resection required to initiate single-strand annealing between the flanking repeat
sequences. Mutants that either reduce Spo11-DSB formation or abolish resection at Spo11-DSBs rescued the repair
block. We also found that a replication protein A component, Rfa1, does not accumulate to expected levels at
unrepaired single-stranded DNA (ssDNA) in dmc1D cells. The requirement of Dmc1 for VDE-DSB repair using flanking
repeats appears to be caused by the accumulation of large quantities of ssDNA that accumulate at Spo11-DSBs when
Dmc1 is absent. We propose that these resected DSBs sequester both resection machinery and ssDNA binding proteins,
which in wild-type cells would normally be recycled as Spo11-DSBs repair. The implication is that repair proteins are in
limited supply, and this could reflect an underlying mechanism for regulating DSB repair in wild-type cells, providing
protection from potentially harmful effects of overabundant repair proteins.
Citation: Johnson R, Borde V, Neale MJ, Bishop-Bailey A, North M, et al. (2007) Excess single-stranded DNA inhibits meiotic double-strand break repair. PLoS Genet 4(11): e223.
In most organisms the success of meiosis is dependent on
the creation of molecular joints that serve to lock homolo-
gous chromosomes together until they mediate ordered
chromosome segregation at the first meiotic division. This
is achieved by the creation of crossovers, which creates a
covalent link between nonsister chromatids, and through the
forces of sister chromatid cohesion to maintain a link
between chromosome pairs until first anaphase (reviewed in
). Crossovers are formed during repair of programmed
DNA double-strand breaks (DSBs) created by the Spo11
protein [2,3]. DSBs can also be repaired by use of the sister
chromatid as template. But, because intersister repair does
not create links between homologous chromosomes, meiotic
cells have evolved a strong bias toward using the homologous
chromosome as donor template. Much attention is currently
focused on understanding how the meiotic cell enforces the
preference for interhomolog repair.
Various proteins have been implicated in directing DSB
repair toward the homologous chromosome and/or away
from the sister chromatid. These include the meiosis-specific
RecA homolog Dmc1. In the absence of Dmc1, DNA joint
molecules between homologous chromosomes fail to form,
causing unrepaired DSBs to accumulate [4,5]. Also implicated
in enforcing interhomolog DNA repair are members of a
meiosis-specific complex, Mek1-Hop1-Red1. Mek1 is a kinase
with similarities to Rad53. Loss of Mek1 function bypasses the
requirement for Dmc1, rendering meiotic DSB repair Rad54-
dependent and increasing the frequency of intersister
recombination events [6–9]. Hop1 and Red1 are phosphor-
proteins that localize to meiotic chromosome axes, and
mutants of both genes have been recovered in screens for
increased intersister chromatid repair [6,8,10–13].
Editor: Michael Lichten, National Cancer Institute, United States of America
Received May 11, 2007; Accepted October 22, 2007; Published November 30,
A previous version of this article appeared as an Early Online Release on October
24, 2007 (doi:10.1371/journal.pgen.0030223.eor).
Copyright: ? 2007 Johnson et al. This is an open-access article distributed under
the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author
and source are credited.
Abbreviations: ChIP, chromatin immunoprecipitation; DSB, double-strand break;
RNR, ribonucleotide reductase; RPA, replication protein A; SSA, single-strand
annealing; ssDNA, single-strand DNA; VDE, VMA1-derived endonuclease
* To whom correspondence should be addressed. E-mail: email@example.com
[ These authors contributed equally to this work.
¤a Current address: Helen L. and Martin S. Kimmel Center for Biology and Medicine,
Skirball Institute for Biomolecular Medicine and Department of Pathology, New
York University School of Medicine, New York, New York, United States of America
¤b Current address: Molecular Biology Program, Memorial Sloan-Kettering Cancer
Center, New York, New York, United States of America
¤c Current address: Department of Nutritional Sciences and Toxicology, University
of California Berkeley, Berkeley, California, United States of America
¤d Current address: Tumour Microcirculation Group, Royal Hallamshire Hospital,
University of Sheffield, Sheffield, United Kingdom
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e2232338
In addition to the classical homologous recombination
mechanisms, DSBs can be repaired by a second homology-
dependent mechanism, single-strand annealing (SSA) [14,15].
DSBs that form during mitotic or meiotic growth are resected
to create short tracts of 39 ending single-stranded DNA
(ssDNA). If unrepaired by strand invasion, resection can
extend for many kilobases. Extensive resection has the
potential to uncover repeated sequences that flank the initial
lesion, such that complementary strands anneal leaving a flap
of intervening DNA that is removed by Rad1/Rad10 flap
endonuclease activity . Little is known about proteins that
initiate resection or catalyze the formation of long tracts of
ssDNA. The Mre11 complex and Exo1 certainly contribute to
resection, and these proteins influence the likelihood of DSBs
repairing through SSA [17–22] (R. Johnson, M. J. Neale, A. S.
H. Goldman, unpublished data). Mitotic studies show SSA to
be independent of RAD51, RAD54, RAD55, and RAD57, but
dependent on RAD52 [17,18]. The single-stranded binding
protein complex replication protein A (RPA) is also required
for SSA, probably to help recruit Rad52 [19,23].
The activities of general and meiosis-specific recombina-
tion proteins are not restricted to repair of Spo11-induced
DSBs during meiosis. This has been determined from studies
of recombination induced by the HO-endonuclease and the
meiosis-specific homing VMA1-derived endonuclease, VDE.
Unlike Spo11, HO-endonuclease and VDE have strict
cleavage sequence-specificity and do not become covalently
bound to the DSB end, creating ‘‘clean’’ DSBs [2,24–27].
Despite these differences, the genetic requirements for repair
of HO- and VDE-induced DSBs are similar to those of Spo11-
DSBs [28–30]. For example, SAE2 is required for removal of
covalently bound Spo11 from DSB ends and for single-strand
resection, and VDE-DSB repair is slowed in sae2 mutants
because resection at VDE-DSBs is also retarded [28,31] (A.
Bishop-Bailey, A. S. H. Goldman, unpublished data). That
Sae2 would be important for repair of a clean DSB is
supported by more recent studies on HO-DSBs in mitotic
cells . DMC1 is required for gene conversion at a VDE-
DSB, indicating that the commitment of meiotic cells to
repair using a homologous chromosome template is not
restricted to Spo11-induced DSBs . Studying DSBs
created by an endonuclease other than Spo11 provides
insight into the regulatory significance of large numbers of
Spo11-DSBs and allows study of mutations at a stage beyond
the point where a phenotypic block would be observed at
Here we report that the requirement of Dmc1 for meiotic
DSB repair persists even when the VDE-DSB is flanked by
direct repeats, which allow repair without need for DNA
strand invasion. If repair by interhomolog gene conversion is
precluded, and SSA is the only pathway for VDE-DSB repair,
DSB repair is still strongly Dmc1-dependent. However, SSA-
mediated VDE-DSB repair becomes Dmc1-independent in the
absence of active Spo11, Hop1, or Sae2, all of which influence
levels of resected Spo11-DSBs. Titrating DSB-associated
ssDNA using hypomorphic spo11 alleles increases SSA repair
efficiency in the dmc1D cells. Analysis of Rfa1 binding supports
the view that extensive Spo11-DSB-associated ssDNA formed
in the absence of Dmc1 reduces the availability of ssDNA
binding proteins necessary for SSA. The implication is that
availability of proteins to perform DSB repair is limited, such
that mutations or physiological conditions that alter the
genomic distribution and/or timing of DSB formation may
have unforseen pleiotropic consequences.
Previous studies have shown that a DSB created by VDE
(described as VDE-DSB1 below) can be repaired either
through an interhomolog gene conversion event, or by using
direct repeats that flank the DSB site, causing deletion of
intervening sequences and creating ‘‘Dproduct’’ (; Figure
1A). Both sister chromatids containing the VDE-recognition
sequence are usually cleaved during meiosis, and tetrad
analysis revealed both chromatids are frequently repaired to
Dproduct, an outcome most compatible with repair by SSA
. The proportion of VDE-DSB1s repaired by interhomo-
log gene conversion versus repair using flanking repeated
sequences, creating Dproduct, varied significantly between
cells with mutations that either inhibited Spo11-DSB for-
mation or single-strand resection at Spo11-DSBs. The former
caused an increase in both the rate of VDE-DSB1 repair and
the proportion of Dproduct, whereas the latter slowed repair
and reduced the proportion of Dproduct. To further
determine how the presence of Spo11-DSBs can influence
VDE-DSB repair in trans, we have now investigated two VDE-
DSB sites in dmc1D meiotic cells. Mutating DMC1 allows
efficient Spo11-DSB formation and efficient single-strand
resection, but Spo11-DSBs fail to repair due to an inability to
form interhomolog joint molecules. Thus, in dmc1D cells the
VDE-DSBs are in the context of multiple hyper-resected
In addition to the published reporter construct , we
used a second VDE-DSB reporter construct, inserted on
Chromosome XV. In this cassette there are two pairs of
nested direct repeats flanking the VDE cleavage site (VDE-
DSB2). The proximal repeats contain URA3 sequence; the
distal repeats contain ADE2 sequence (Figure 1B). The second
Chromosome XV has a 645-bp deletion from the ADE2 locus.
PLoS Genetics | www.plosgenetics.orgNovember 2007 | Volume 3 | Issue 11 | e2232339
Limiting Factors in DSB Repair
During meiosis, DNA is deliberately damaged by formation of
double-strand breaks. Programmed breaks must be repaired for cell
division to be completed. Break repair enables reciprocal exchange
between parental chromosomes, and this exchange acts as a link
between chromosomes before anaphase separation. These links are
essential to ensure that maternal and paternal chromosomes
segregate into different daughter cells. Meiosis has special
mechanisms to ensure the repair creates sufficient reciprocal
exchanges between parental chromosomes; Dmc1 protein is
essential for these mechanisms to work. When Dmc1 is absent,
programmed breaks accumulate with excess single-stranded DNA
nearby. Using reporter constructs integrated into yeast, we
examined repair of an experimentally induced break expected not
to need Dmc1. When Dmc1 is absent, programmed breaks
accumulate in single-stranded form, and the experimental break is
not repaired. Either preventing formation of programmed breaks, or
stopping DNA near them from becoming single-stranded, relieves
this repair block. We conclude that repair proteins are likely to be in
limited supply during meiosis, and they run out in cells lacking
Dmc1 function. Limiting protein supply may be an important
regulatory mechanism, protecting DNA from potentially damaging
effects of oversupply.
No homology to repair VDE-DSB2 exists on the homologous
chromosome for approximately 4 kb to the left and 7 kb to
the right of VDE-DSB2, forcing DSB repair events to proceed
through an intrachromosomal route, which we show is SSA.
The VDE-DSB1 Accumulates in Cells Lacking Dmc1
In wild-type cells, VDE-DSB1 is repaired by both inter-
homolog gene conversion and SSA. We considered that
knocking out DMC1 could have very different influences on
VDE-DSB repair. On the one hand, RecA type strand invasion
function is not required for SSA, and so in dmc1D cells there
could have been efficient VDE-DSB repair by SSA, at the
expense of gene conversion. On the other hand, in dmc1D
cells the nucleus accumulates around 200 resecting Spo11-
DSBs and hundreds of kilobases of ssDNA . This might
create a competition for resources (such as resection proteins
or RPA) and therefore compromise VDE-DSB repair.
Using Southern analysis that isolates a parental arg4-vde
DNA fragment from all other arg4-containing species, we
determined that mutating DMC1 had no significant effect on
the efficiency of forming VDE-DSB1 (Figure 2A and 2B). By
contrast, mutating DMC1 causes a major delay in turnover of
the VDE-DSB1 DNA (Figure 2C and 2D). In wild-type cells,
the VDE-DSB1 signal accumulates up to 5 h after induction of
meiosis and gradually disappears as the rate of DSB repair
outpaces residual DSB formation. The VDE-DSB1 signal
accumulated in dmc1D cells similarly to wild type up to 4 h
after induction of meiosis, but did not diminish. Consistent
with the observed accumulation of VDE-DSB1 molecules in
dmc1D cells, very little Dproduct was created compared to
wild-type cells (Figure 2E). As there is no significant differ-
ence in the kinetics of VDE-DSB1 formation in the different
strains, the observed variation in VDE-DSB1 levels must be
due to a difference in repair.
In dmc1D cells, the amount of VDE-DSB1 DNA detected on
Southern blots was maintained from 4 h to 8 h at a steady
level representing approximately 35% of chromatids (Figure
2F). However, by 8 h of meiosis, less than 10% of chromatids
Figure 1. The arg4-vde-Containing Reporter Constructs
(A) The ura3::arg4-vde reporter cassette containing the VDE-DSB1 site has been described previously . This cassette is in a heterozygous state with a
nearly identical insertion containing the arg4-bgl allele  on the opposite Chromosome V. Repair of the VDE-DSB1 is possible by gene conversion
after short resection. Long resection of approximately 2 kb and 6.5 kb on the left and right of the VDE-DSB uncover flanking homology (URA3
sequences) that can be used for repair by SSA yielding Dproduct (the grey area within URA3::Ty is a naturally disrupting Ty element).
(B) The ade2::arg4-vde reporter cassette is hemizygous, inserted into one ADE2 locus. The opposite Chromosome XV has an internal deletion in the ADE2
locus. Resection of approximately 3.0 kb to both the left and right will uncover the proximal URA3 repeated DNA sequences. An SSA event between
these yields Dproximal. Further resection to 4.5 kb and approximately 7.0 kb to the left and right will uncover the ADE2 repeated sequences; SSA
between them yields Ddistal. Repair of VDE-DSB2 by gene conversion is unlikely as homology with the ade2D chromosome is over 3 kb and 7 kb away,
on the left and right sides, respectively, leading to long nonhomologous 39 ends.
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e223 2340
Limiting Factors in DSB Repair
were detectable in the Dproduct band, even though approx-
imately 90% of chromatids had been broken (Figure 2B). The
discrepancy between these values suggested that standard
Southern analysis fails to detect some VDE-DSB1 DNA. This
could be due to extensive single-strand resection, which is
known to cause Spo11-DSB bands to become diffuse when
DNA is extracted from dmc1D cells .
Accumulated VDE-DSBs Are Resected
To test whether or not the VDE-DSB1 band was under-
represented due to diffuse smearing, we compared the
amount of resected DNA that accumulates in wild-type and
dmc1D cells using a loss-of-restriction site assay. This assay
resolves DNA under denaturing conditions and quantitatively
detects resected molecules as discrete bands by using a strand-
specific probe (Figure 3A). In the wild-type culture after 8 h,
approximately 11% of arg4-vde chromatids were detected in
bands representing VDE-DSB1 resected DNA (Figure 3B). By
contrast, 8 h after induction of meiosis in the dmc1D culture,
84% of VDE-DSB1 DNA was detected in the resection bands
(Figure 3B). Since 90% of arg4-vde chromatids are cut by VDE
by 8 h meiosis (Figure 2B), virtually all VDE-DSB1s in dmc1D
cells persist in a resected, unrepaired state.
Persistence of resected DNA around VDE-DSB1 is highly
reminiscent of the behavior of Spo11-DSBs in dmc1D cells. We
found most of the resected DNA appears in two apparent
pause sites for resection (Figure 3A and 3C). One pause site,
with resection up to 1.8 kb, is also present in earlier time
points for wild-type cells, and in dmc1D cells contains
approximately 49% of VDE-DSB1 chromatids at 8 h (Figure
3C). The second pause site, with up to 2.7 kb of resection,
appears to be a major block to further processing in dmc1D
cells with 23% of VDE-DSB1 chromatids collecting there by 8
h (Figure 3C). Repair of the VDE-DSB1 by SSA requires that
resection extends as far as both flanking copies of the
repeated sequence. To uncover the repeat sequence on the
right side of VDE-DSB1 requires over 6 kb of resection. Thus,
the difficulty in repairing the VDE-DSB1 by SSA may be
caused by the inability to create sufficiently extensive
resection tracts in the dmc1D cells.
Another important feature of this analysis is a 5-fold
increase in the band representing resection between 3.3 kb
and 8.5 kb in dmc1D cells, compared to wild-type (Figure 3A
and 3C). This band is barely visible in wild-type cells,
presumably due to rapid repair by SSA of DSBs with long
resection tracts. Thus, even when there is sufficient resection
for SSA, repair is inhibited in dmc1D cells.
Repair of VDE-DSB1 Is Not Dependent on Meiotic
dmc1D cells arrest meiosis prior to exit from prophase, due
to checkpoint-induced inactivation of the Ndt80 transcrip-
tion factor [7,33–35]. To test the possibility that this arrest
causes inhibition of resection and SSA, we repeated the assay
in cells lacking NDT80 (Figure 4).
The kinetics of VDE-DSB1 induction and repair were
similar in wild type and ndt80 meioses (Figure 4A and 4B).
While ndt80 dmc1D mutants displayed the same defects in
VDE-DSB1 repair as dmc1D cells, ndt80 cells exhibited
amounts of Dproduct comparable to wild type (Figure 4F).
Thus, the meiotic prophase arrest imposed by dmc1D is
unlikely to be the cause of the inefficient VDE-DSB1 repair.
Mutants That Reduce the Cellular Load of Resected
Spo11-DSBs Restore SSA of VDE-DSBs in dmc1D Cells
We next considered the possibility that VDE-DSB repair is
reduced in dmc1D cells because of the large quantity of ssDNA
produced by the resection of unrepaired Spo11-DSBs. To test
this, we measured the efficiency of repairing VDE-DSB1 in
dmc1D cells with few Spo11-DSBs (hop1D), completely lacking
Spo11-DSBs (spo11-Y135F-HA3His6 , referred to as spo11,
forthwith), and in sae2D cells with wild-type levels of Spo11-
DSBs but no single-stranded resection (Figure 4A and 4C–4F).
We previously reported that mutation of HOP1, SPO11, or
SAE2 alters the relative proportion of VDE-DSB1 molecules
Figure 2. Repair of VDE-DSB Is Inhibited in dmc1 Cells
(A) DNA from 0 h to 8 h of meiotic culture was digested with EcoRV and
BglII and probed close to the VDE-DSB1 site to create a DNA fragment of
unique size containing uncut parental arg4-vde DNA (bgl ¼ parental
arg4-bgl chromatids plus gene conversion products creating further
arg4-bgl; vde ¼ parental arg4-vde chromatids; VIII ¼ arg4-nsp;bgl
chromatids in both natural Chromosome VIII loci; DSB&GC¼ VDE-DSB1
molecules plus gene conversion products creating ARG4).
(B) Quantification of the parental arg4-vde band normalized to 50% of
signal in band VIII. Diminution of the arg4-vde band is a consequence of
VDE-DSB1 formation, which occurs at similar rates in wild-type and dmc1
(C) DNA was digested with SpeI and probed distal to the URA3 locus to
isolate DNA fragments of unique size containing VDE-DSB1 molecules
and Dproduct (P&GC ¼ parental arg4-vde and arg4-bgl chromatids plus
gene conversion products; DSB ¼ chromatids with VDE-DSB1; D ¼
Dproduct; Spo11 ¼ natural Spo11-DSB site close to the arg4 insert).
(D) Quantification of VDE-DSB1 signal expressed as a proportion of arg4-
vde-containing chromatids, symbols as in (B).
(E) Quantification of Dproduct expressed as a proportion of parental
(F) Quantification of VDE-DSB1 signal expressed as a proportion of
cumulative parental arg4-vde chromatids that have received a VDE-DSB1,
symbols as in (B).
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e223 2341
Limiting Factors in DSB Repair
that repair using the flanking repeats versus repair by gene
conversion. In that study, and here, we found that mutating
HOP1 or SPO11 increased both the speed of repair and the
proportion of VDE-DSB1 that repaired using flanking
homology  (Figure 4A, 4C, 4D, and 4F). In sae2D cells,
VDE-DSB1 repair is slower and the proportion of molecules
that repair using flanking repeats is reduced, probably due to
a direct role of Sae2 on resection, even at DSBs lacking
covalently bound protein  (Figure 4A, 4E, and 4F) .
We now report that all the double mutants, spo11 dmc1D,
hop1D dmc1D, and sae2D dmc1D behave almost the same as
their DMC1 single mutant counterparts with respect to the
kinetics and efficiency of VDE-DSB1 repair (Figure 4A and
4C–4F). In other words, the requirement for Dmc1 to repair a
VDE-DSB using flanking homology is relieved by mutations
that reduce the large quantity of resected Spo11-DSBs and
ssDNA usually seen in dmc1D meiosis.
Lowering the Quantity of Resected Spo11-DSBs Reduces
the Requirement for Dmc1 to Repair VDE-DSB1 by SSA
We next checked if intermediate levels of VDE-DSB repair
could be achieved in dmc1D cells when intermediate levels of
ssDNA accumulate in the nucleus. The number of Spo11-
DSBs was reduced by using a heterozygous diploid expressing
the hypomorphic SPO11-HA3His6 allele and the null allele,
spo11-Y135F-HA3His6 . This combination of SPO11 alleles
is reported to reduce recombination and Spo11-DSB levels to
approximately 50% of wild type . We confirmed this to be
the case in dmc1D cells by measuring DSB formation at the
ARE1 hotspot (unpublished data). This reduction in Spo11-
DSB formation increased the VDE-DSB1 repair efficiency in
dmc1D cells approximately 6-fold (Figure 4F), supporting the
contention that inefficient SSA repair of VDE-DSBs in dmc1D
cells is related to accumulation of resecting Spo11-DSBs.
Dependency on Dmc1 for VDE-DSB1 Repair Does Not
Reflect the Requirement for an Unexpected DNA Strand
Theoretically, VDE-DSB1 could create the Dproduct
following an unequal strand exchange event with the
homologous chromosome, which could be Dmc1-dependent
in meiosis. It was important therefore to test repair of a VDE-
DSB in a context that ruled out the possibility of Dmc1-
dependent interhomolog repair. To test this, we used a
second reporter cassette that contains the VDE cleavage site
(VDE-DSB2) inserted on Chromosome XV between two pairs
of nested direct repeats (Figure 1B). Repair of VDE-DSB2 by
strand invasion with the homolog is precluded by large tracts
of heterology present in the hemizygous reporter insert.
Repair of VDE-DSB2 was also reduced by loss of DMC1,
although not to the same extent as VDE-DSB1 (Figure 5A and
5B). By 8 h in the dmc1D culture, approximately 33% of VDE-
DSB2 broken chromatids had repaired compared to 86% in
wild type (Figure 5B). The repair level did not increase in
dmc1D cells by 12 h (unpublished data). As for VDE-DSB1,
repair of VDE-DSB2 was increased in dmc1D cells when the
DSB forming function of Spo11 was removed (Figure 5B).
By mutating the RAD54 gene, which is important for strand
exchange between sister chromatids, we confirmed our
expectation that repair of VDE-DSB2 using flanking repeated
sequences was not by unequal intersister repair [6,9,38,39].
For rad54 cells, there was a slight delay in meiosis indicated by
delayed induction of VDE-DSB2 and late onset of the first
division (unpublished data). Despite this delay, 87% of
broken chromatids molecules were repaired by 8 h in rad54
meiosis (Figure 5B). Because VDE-DSB2 repair was as
efficient in rad54 cells as in wild type, we conclude that
repair using the flanking repeats does not require either a
homolog or Rad54. In other words, VDE-DSB2 repair occurs
by SSA, rather than via a mechanism that requires DNA
strand invasion of a donor duplex.
Figure 3. Most Broken arg4-vde-Containing Chromatids in dmc1 Cells
Remain Unrepaired in a Resected State
(A) DNA from 0 h to 8 h of meiotic culture DNA was digested with HaeII
and fractionated by alkaline denaturing gel electrophoresis. Bands have
increasing molecular weight as the restriction sites are progressively
destroyed by resection. The numerals to the left indicate distance from
the VDE-DSB1 site to an HaeII cut-site, and represent the maximum
extent of resection for molecules in the respective bands (*¼nonspecific
band; other labels are as described in Figure 2; a maximum of 1/4 of the
signal in the P&GC&VIII band represents the parental arg4-vde fragment;
full details in Materials and Methods).
(B) Quantification of the resected bands totaled after 8 h of meiosis
expressed as a proportion of parental arg4-vde chromatids. In dmc1D
cells practically all VDE-DSB1s created remain in a resected state.
(C) Quantification of individual resected bands after 8 h of meiosis
expressed as a proportion of parental arg4-vde chromatids. Accumu-
lation of resected molecules is punctuated in dmc1D cells indicating
likely pause sites through which resection is nonprocessive. (P&GC
represents remaining parental chromatids plus gene conversion product;
the numbers refer to band sizes).
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e223 2342
Limiting Factors in DSB Repair
Evidence that VDE-DSBs cannot repair using the sister
chromatid in meiosis is supported by analysis of a his4-vde
allele in a hybrid strain containing a Saccharomyces cerevisiae
Chromosome III opposite a Saccharomyces carlsbergensis Chro-
mosome III. Due to sequence divergence between the two
chromosomes, interhomolog gene conversion is prohibited
[40–42]. Since no repeats flank the his4 VDE-DSB, the only
possible route to homologous repair is through invasion of
the sister chromatid. However, all tetrads are two-spore
viable due to loss of the his4-vde Chromosome III, indicating
that intersister strand invasion and repair does not happen
(K. Tittcomb and A. S. H. Goldman, unpublished data).
Resection at VDE-DSB2 Can Be Extensive in dmc1D Cells
Our analysis of resection intermediates at VDE-DSB1
suggested that DSB repair might be inhibited in dmc1D cells
because of a failure to perform sufficient extensive ssDNA
resection (Figure 4). Because VDE-DSB2 is flanked by nested
direct repeats (Figure 1B), this site can be used to estimate
how much resection had occurred in the repaired molecules.
Repair by SSA using the proximal repeats (Dprox) creates a 5
kb deletion, detected on Southern blots as a 10 kb band.
Repair using distal repeats (Ddistal) creates a 10.5 kb deletion
and a 4.5 kb band (Figure 5A). By comparing the relative
proportion of these repaired products, it is possible to
estimate the proportion of molecules that repaired with
longer resection tracts (Figure 5C). In wild-type cells,
approximately 22% of repaired VDE-DSB2 molecules used
the distal repeats. In dmc1D cells, the distal repeats were used
more frequently, with 33% of repaired molecules being of the
Ddistal type. Thus, the reduction in resection in dmc1D cells is
not uniform across the population of cells, possibly reflecting
a stochastic inhibition created by competition for resources.
Rfa1, a Component of the ssDNA Binding Complex RPA,
Has Limited Access to the Repeated Sequences Flanking
VDE-DSB1 in dmc1D Cells
The negative correlation between ability to repair the
VDE-DSB by SSA and the number of DSBs undergoing
extensive resection could reflect inability of ssDNA binding
proteins to access homologous sequences flanking the VDE-
DSBs. Failure of ssDNA binding proteins to bind DNA could
result from a combination of reduced resection and ssDNA
binding proteins being in limited supply at the VDE-DSBs,
because they are sequestered to the long ssDNA tracts that
accumulate at the many Spo11-DSBs present in the cell.
The ssDNA binding protein complex, RPA (also known as
RF-A in yeast), is required to remove secondary structure in
ssDNA and to recruit recombination proteins such as Rad52
during homologous recombination . Using chromatin
immunoprecipitation (ChIP), we compared (in wild-type and
dmc1D cells) the association of an RPA component (Rfa1) to
DNA coupled with a Spo11-DSB hotspot (YCR047C/BUD23
ORF, ) or with VDE-DSB1 (Figure 6A–6C). Close to the
BUD23 Spo11-DSB hotspot, a small amount of DNA is
enriched from wild-type cells by ChIP, the changing levels
through time reflect the kinetics of appearance and dis-
appearance of Spo11-DSBs at this site (Figure 6A). In dmc1D
cells, the Rfa1 signal is not reduced in later time points
because Spo11-DSBs are not repaired. Normally Spo11-DSBs
are formed up to 6 h after induction of meiosis. Interestingly,
Rfa1 does not accumulate to higher levels after 4 h as more
Spo11-DSBs are created, perhaps because as Rfa1 becomes
limiting it is competed away from a proportion of the sites
close to the Spo11-DSB.
We have already shown that almost all VDE-DSBs created (in
close to 100% of cells; Figure 2B) in dmc1D remain broken and
in a resected state throughout the time course (Figure 3). We
reasoned that if Rfa1 coats all ssDNA in dmc1D cells, the Rfa1
ChIP signal should be higher in dmc1D cells versus wild type.
Figure 4. Meiotic Prophase Arrest Does Not Inhibit VDE-DSB1 Repair, but
the Quantity and State of Spo11-DSBs Affects the Need for Dmc1
In all cases spo11 refers to spo11-Y135F-HA3His6.
(A) DNA from 0 h to 8 h of meiotic culture was digested and displayed as
described in Figure 2C.
(B) Cells do not require Ndt80 function to turnover the VDE-DSB in DMC1
cells, and dmc1D is epistatic to ndt80 for VDE-DSB1 repair. The repair
defect imposed by loss of Dmc1 function is fully rescued by (C and D)
removing all or most Spo11-DSB formation in either spo11 or hop1
mutant cells and by (E) inhibiting resection at Spo11-DSBs by mutating
(F) The epistatic relationships between the various mutations and dmc1
are also displayed by quantification of Dproduct after 8 h of meiotic
culture, expressed as a proportion of parental arg4-vde chromatids. Wild
type and dmc1 are shown for comparison. Reducing the quantity of
Spo11-DSBs to 50% of wild-type levels using a strain heterozygous for
different SPO11 alleles leads to partial relief of Dmc1 dependence for
PLoS Genetics | www.plosgenetics.orgNovember 2007 | Volume 3 | Issue 11 | e223 2343
Limiting Factors in DSB Repair
Using two different primer sets, one close to the VDE-DSB1
(Figure 6B) and one close to a flanking repeated sequence
(Figure 6C), we found that the ChIP signal was not
significantly enhanced in dmc1D cells close to the VDE-DSB
and is slightly reduced far from the VDE-DSB1. Taken
together, these results support the view that when Spo11-
DSBs remain in a resected state, Rfa1 does not accumulate at
specific ssDNA sites, even though the quantity of ssDNA is
demonstrably increasing (Figure 3). Thus, for those cells, in
which resection does uncover the repeated sequences
flanking the VDE-DSB1, there may not be sufficient RPA to
mediate repair by SSA.
We noted that the ChIP signal is not significantly reduced
by 7 h in wild-type cells, when a large proportion of VDE-
DSB1s are repaired. In part this is probably because 40% of
the peak numbers of VDE-DSB1s are still present at this time
(Figure 2D). It is also possible that RPA is not removed as
rapidly from the DNA following SSA as it would be following
classical homologous recombination.
In budding yeast, repair of DSBs induced by VDE is a
natural process, which propagates VDE-containing genetic
elements from one chromosome to another during meiosis
. Previous studies have shown the timing of DSB induction
by VDE and the mechanisms of repair parallel closely those of
Spo11-induced recombination [28,31]. Deleting DMC1 is
expected to prevent gene conversion events at VDE-DSB
sites, but in addition we have found it also inhibits repair by
SSA, a process that in mitosis neither requires strand invasion
nor the RecA ortholog Rad51 .
The failure in VDE-DSB repair in dmc1D mutant cells can
be relieved by mutations that eliminate Spo11-DSBs, reduce
their frequency, or prevent their resection. Furthermore, in
dmc1D cells resection at VDE-DSBs is reduced, and Rfa1 (a
component of the ssDNA binding complex, RPA) is unable to
access the repeated sequences flanking the VDE-DSB sites.
Repair Proteins Are a Limiting Factor
Losing Dmc1 function has an enormous impact on the
normal balance of DNA transactions taking place in the
meiotic nucleus. Spo11-DSBs are formed with wild-type
kinetics, but breaks remain unrepaired and accumulate
genome-wide, so that at later time points each cell would
contain about 200 lesions . Under these conditions,
resection continues for many hundreds or thousands of
nucleotides further than normal, and therefore the demand
for both resection complex and RPA is likely to be extremely
high and critical (Figure 7).
The data from two VDE-DSB-containing cassettes support
the idea that either resection complex or ssDNA binding
proteins can become limiting. Insufficient access to resection
complex could reduce resection below the lengths needed to
render flanking repeats single-stranded. Insufficient RPA to
coat long resection tracts could prevent SSA even if resection
has uncovered the repeated sequences. Whether a protein
complex becomes limiting at a particular DSB may be
stochastic, though genome location and immediate environ-
ment may also be important. This would explain why at VDE-
DSB1 resection appears extremely limited, but less so at VDE-
DSB2, which also repairs poorly but more efficiently than
The coordinate induction and repair of DNA damage is a
vital part of the meiotic developmental pathway. Ensuring
that the many proteins required for DSB repair are in
appropriate supply and in active form is a major task for the
cell. Spo11-DSB formation is temporally linked to replica-
tion, with DSBs appearing about 1.5 h after replication has
passed through . Thus, like replication, the formation of
Spo11-DSBs across the genome is asynchronous. In between
Spo11-DSB formation and repair, Spo11-DSBs are processed
by 59 to 39 resection. Resulting ssDNA subsequently forms
joint molecules with the homologous chromosome, and DSBs
Figure 5. Southern Analysis of the VDE-DSB2 Cassette in a Hemizygous
In all cases spo11 refers to spo11-Y135F-HA3His6.
(A) DNA from wild-type and mutant cells, as indicated, was digested with
SpeI, which isolates unique fragments representing (P) the uncut parent
arg4-vde chromatid, (Dprox) the product of SSA between URA3 proximal
repeated sequences, (DSB) VDE-DSB2 molecules, and (Ddistal) the
product of SSA between ADE2 distal repeated sequences. Cutting by
VDE is less efficient in this assay, as indicated by the remaining parental
DNA. In all but the dmc1D mutant cells, little VDE-DSB DNA is visible by 8
h. The amount of product accumulating is significantly reduced in
dmc1D cells in which the VDE-DSB2 becomes smeared due to prolonged
(B) The total amount of repair product (sum of Ddistal and Dproximal)
visible has been measured as a proportion of arg4-vde-containing
chromatids that have received a VDE-DSB by 8 h in meiosis (details in
Materials and Methods).
(C) The quantity of Ddistal expressed as a proportion of sum of total
repair product gives an indication of how often repaired molecules used
the longer resection tract.
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e223 2344
Limiting Factors in DSB Repair
are repaired 1–1.5 h after Spo11-DSB formation [47,48]. It is
reasonable to assume that by this time resection would have
ceased and ssDNA-binding complexes such as RPA and Dmc1/
Rad51 may be liberated for reuse. Since a large number of
such molecules are required at every DSB, the asynchronous
induction of recombination may have evolved to spread the
workload of proteins, which otherwise would have to be
produced at levels that might result in pleiotropic negative
There is precedence for proteins involved in nucleic acid
metabolism being in limiting supply to avoid pleiotropic
effects from over abundance. Activity of ribonucleotide
reductases (RNRs) is tightly controlled to ensure nucleotide
pools are sustained at appropriate levels. Control is main-
tained at the transcriptional level and by an inhibitor protein,
Sml1 ( and within). The Mec1 DNA damage response
pathway, acting through Rad53 and Dun1, regulates both
controls. DNA damage increases transcription of RNR genes
and phosphorylates Sml1 causing its inactivation [50,51].
Regulation of active RNR protein supply is so tight that
mutations in the Mec1 kinase cascade cause SML1 expression
to become toxic, due to insufficiency of nucleotides needed
for routine DNA repair [50,52,53]. Stringent control of RNRs
probably reflects the fact that overabundance of nucleotides
can be mutagenic due to increased risk of misincorporation.
Our data suggest that ssDNA binding proteins must also be
kept in limiting supply. This suggestion fits well with the fact
that the transcription regulation of RPA components is
linked to the cell cycle, peaking at the G1/S boundary .
Furthermore, reports on in vitro DNA binding and activity of
Rad52 and Rad51 indicate that RPA at low concentrations
promotes strand exchange, yet at high concentrations it
limits access of repair proteins to the DNA [55–57].
Limited Supply of Repair Proteins Could Have Wide-
Ranging Impact on Regulating Meiotic Processes
That Rfa1 can become a limiting factor in an experimental
situation raises the possibility that in wild-type cells pertur-
bations of repair efficiency or replication could have indirect
effects on other repair/replication processes. Such effects can
be at sites other than the original lesion. For example, a level
of DNA damage similar to that experienced by meiotic cells
might perturb DNA replication, because damage repair
sequesters protein factors such as RPA that are also critical
One reason for limiting the number of Spo11-DSBs in yeast
meiosis could be to ensure that a safe and sustainable balance
Figure 6. ChIP Reveals That RPA Is a Limiting Factor in dmc1D Cells
(A–C) DNA associated with immunoprecipitated Rfa1, a component of RPA, was amplified and quantified by qPCR. The values are averages from
duplicate experiments. Each diagram shows the distance of PCR primers from the relevant DSBs; grey arrow represents YCR047C, the cassette
containing the VDE-DSB1 site is as in Figure 1A, the small black bars represent the position of the PCR products at the indicated distance from the DSB
site. The left axis of the graphs show ChIP enrichment (i.e., ChIP signal relative to input signal; %); on the right axis the values have been corrected to
account for overrepresentation of substrate in the input compared to the proportion of probed chromatids that can receive a DSB (see Materials and
Methods). The bar charts show the ratio of corrected ChIP enrichment values to proportion of chromatids that are expected to contain ssDNA in the
PCR-amplified region. In each bar graph the ratio of corrected ChIP enrichment to estimated levels of ssDNA are always higher in wild type compared to
dmc1D, showing that the immunoprecipitation of Rfa1 is more efficient in wild-type cells. The difference in ChIP efficiencies for wild-type and dmc1D
cells is greatest for the site far from VDE-DSB1, implying that as resection proceeds it becomes more difficult to compete for Rfa1, which may be stably
bound to ssDNA created earlier, close to DSB sites.
(A) Close to a Spo11-DSB hotspot, the ChIP enrichment decreases with repair in wild-type cells but not in dmc1D cells.
(B) Close to VDE-DSB1, the ChIP enrichment in dmc1D cells is no higher than in wild-type, even though by 7 h around 80% of arg4-vde-containing
chromatids are in a resected state (see Figures 2 and 3) and (C). Distant from VDE-DSB1, but close to a flanking repeated sequence used for SSA, again
the ChIP enrichment in dmc1D cells is no higher than in wild type, even though significant amounts of DNA accumulate with resection beyond this
point (see Figure 3).
PLoS Genetics | www.plosgenetics.orgNovember 2007 | Volume 3 | Issue 11 | e223 2345
Limiting Factors in DSB Repair
can be reached between repair protein supply and demand
within a suitable time frame. Studies on mammalian and yeast
vegetative cells support the view that homologous recombi-
nation is extremely sensitive to protein supply, as over-
expression of Rad52 epistasis group proteins can inhibit
repair of an experimentally induced DSB [58,59]. Consistent
with the notion that protein supply and demand is finely
balanced in meiosis, moderately hypomorphic mutants of
DSB repair genes such as MRE11, NBS1, or RAD51C have
significantly reduced fertility in mice due to inefficient DSB
repair [60,61]. Similarly in yeast meiosis, deleting one RecA
ortholog, DMC1, causes a complete block in DSB repair that
can be largely rescued by either overexpressing or releasing
inhibition of RAD51 [62,63].
It was recently shown that synchrony of the first meiotic
division could be influenced by temporary chemical inhib-
ition of a specific Cdc7 activity required for induction of
Spo11-DSBs . Nearly half of the population underwent
the first meiotic division sometime between 3 h and 4 h after
inhibitor wash out, with no loss of viability. This might seem
to counter our argument that repair proteins are limiting.
But across the population at the single Spo11-DSB analyzed,
break induction was still spread over some hours. Thus, while
the Wan et al. data  demonstrate some improved
synchrony of the first division across a population of cells,
it does not directly address the synchrony of DSB formation
within each cell. Further analysis of highly synchronized
populations, in which the time of inducing Spo11-DSBs and
their life spans is well defined, will help to clarify the time
limits within which Spo11-DSB formation must be limited to
avoid problems of repair protein supply.
Limiting protein supply could serve a useful function other
than protecting against ill effects of overproduction, such as
to direct the proportion of events that take one biochemical
pathway rather than another . One possible example of
this comes from maize, which produces more than 20-fold
excess of DSBs during meiosis compared to the known
crossover frequencies . These DSBs are identified cyto-
logically as Rad51 foci that often appear as opposite pairs on
homologous chromosomes and may be used as pairing sites
[65,66]. What prevents a much higher proportion of DSBs in
maize meiosis from becoming crossovers is unknown. Perhaps
a protein required for crossover-associated repair is in
limited supply, so the majority of DSBs can serve a function
other than forming crossovers.
The impact single mutations have on biochemistry,
nucleus-wide, is rarely known. This study highlights the fact
that indirect effects can easily arise from changing the
Figure 7. Limited Availability of DNA Repair Proteins Explains the Requirement of Dmc1 for SSA at VDE-DSBs
In wild-type meiosis there is sufficient resection complex and Rfa1 to create and bind to long tracts of ssDNA at the VDE-DSB so that SSA is possible. In
part, the ready supply of such proteins is likely created by the asynchronous nature of Spo11-DSB formation and repair in the nucleus, thus when some
Spo11-DSBs are using these proteins others may have moved to a biochemical step that allows their release. In dmc1D cells, so much resection complex
and Rfa1 is sequestered to multiple unprepared Spo11-DSBs that insufficient resection complex is available to create long resection tracts at the VDE-
DSB; or in cases where long resection tracts appear there is not enough free Rfa1 to bind the repeated sequences. Mutating either SPO11 or SAE2
relaxes the demand on both resection proteins and ssDNA binding proteins such that resection and repair of the VDE-DSB is no longer limiting.
PLoS Genetics | www.plosgenetics.orgNovember 2007 | Volume 3 | Issue 11 | e223 2346
Limiting Factors in DSB Repair
balance of protein supply and substrate in the nucleus. In
nature, tight control of protein supply may be essential for
avoiding pleiotropic effects from oversupply, or for limiting
the number of events passing down a specific pathway. In the
laboratory, phenotypes ascribed to mutations may often not
be caused by a direct biochemical impact of mutating a gene,
but may be due to broader biochemical imbalances created
across the nucleus. Genes involved in DNA metabolism may
be particularly likely to cause pleiotropic phenotypes.
Mutations altering the amount of damage present in a
nucleus are certainly prone to pleiotropic effects that are
worth serious consideration when defining protein function.
Materials and Methods
Media, genetic methods, and strains. Diploid yeast strains of the
SK1 background relevant genotypes are listed in Table S1.
The ura3::arg4-vde cassette (Figure 1A) was created as described
previously in . The ade2::arg4-vde reporter cassette (Figure 1B) was
created by transforming parent strains with pAG408, a derivative of
pBR322 based pAG137 . The URA3 of pAG137 was modified by a
59 (?17 to þ129) deletion between restriction sites SdaI to XcmI. A
functional URA3 (HindIII fragment) was inserted into an NruI site in
the PBR322 backbone. A section of the ADE2 ORF (þ250 to þ1695)
generated by PCR from yeast genomic DNA was inserted into the
pBR322 EcoRI site and used for integration into the yeast genome
following linearization at the AflII site.
Relevant mutant strains containing either the ura3::arg4-vde or
ade2::arg4-vde reporter cassettes were made by mating and dissection
with appropriate SK1 haploids. The source of the spo11-Y135F-
HA3His6; hop1; sae2D; and ndt80 haploids is reported in .
For the ura3::arg4-vde reporter construct, dmc1 disruption was
made in pAG64 using primers 59-gccattctatgtctgatcccgg-39 and 59-
tcgcttagttcacctctaccgc-39 to amplify a 1,466-bp region of the dmc1
locus from haploid SK1 genomic DNA. MfeI-cleaved PCR product
was ligated into EcoRI-linearized pUC19. A 2.2 kb ADE2-containing
BglII fragment of pAG52 ( pMJ412 from M. Lichten) was ligated at the
BglII site located 90 bp inside the DMC1 ORF. For the ade::arg4-vde
reporter cassett, dmc1::ARG4 was obtained from D. Bishop in SK1 and
was crossed into our experimental strains. The rad54D mutation was
obtained from D. Bishop and transformed into relevant strains.
DNA isolation and Southern blot analysis. 40 ml samples of culture
were removed at hourly intervals and processed for storage and DNA
isolation according to Allers and Lichten (2001); hexamine cobalt (III)
chloride was excluded from solutions.
Restriction endonuclease–digested DNA was separated under
native conditions or denaturing conditions. Separated DNA was
blotted to Zetaprobe membrane (Bio-Rad) under denaturing con-
ditions with a Vacugene-XL system. Analysis of VDE-DSB1 Dproduct
and resection intermediates was undertaken as described . For
VDE-DSB2 the probe used to display SpeI- digested DNA after native
separation is specific to Chromosome XV coordinates 566120–
Quantification and calculations for Southern analyses. Quantifica-
tion was as described ; briefly, for VDE-DSB1 the amount of DNA
in the VDE-DSB1 band and Dproduct were determined as a
proportion of arg4-vde-containing chromatids by dividing the signal
in each band by half of total signal (which represents both arg4-vde
and arg4-bgl chromatids). For the denaturing gels, the amount of
signal attributable to chromatids that contained parental arg4-vde
insert was calculated taking into account the presence of signal from
six other chromatids (two chromatids with arg4-bgl on the homolo-
gous Chromosome V and four chromatids with arg4-nsp,bgl at the
natural ARG4 locus on both Chromosome VIII homologues) and the
fact that chromatids repairing to Dproduct do not contribute signal.
Dr¼Recorded proportion of arg4-vde chromatids repaired by SSA
Sr¼ Signal recorded in lane on denaturing gel
Dm¼ Dproduct missing from lane on denaturing gel
Rr¼ Resection band recorded signal on denaturing gel
1/4 of total possible signal comes from arg4-vde chromatids
Therefore signal calculated in lane attributable to arg4-vde
Then proportion of arg4-vde chromatids in each resection band;
For VDE-DSB2 gels, the DNA in each band was quantified.
T ¼ Total signal in lane
P ¼ Signal in parent band
Dprox ¼ signal in Dprox band
Ddistal ¼ signal in Ddistal band
Then DNA repaired, as a proportion of breaks made equals (Dprox
þ Ddistal)/(T ? P) and the proportion of DNA repaired to Ddistal
equals Ddistal/(Dprox þ Ddistal).
Rfa1 ChIP. 20 ml cells (4 3 109cells) were treated with 1% fresh
formaldehyde for 15 min at room temperature and 125 mM glycine
for 5 min. ChIP was performed as described  using magnetic
protein G beads (Dynal) and a polyclonal rabbit anti-Rfa1 antibody
(supplied by S. Brill). Enrichment of DNA bound by RPA was
estimated by quantitative PCR using an Applied Biosystems 7500
Real-Time PCR system with 0.4 lM primers, SYBR Green PCR master
mix (Applied Biosystems), and the PCR program: 95 8C for 15 s; 60 8C
for 1 min; 40 cycles. The following primers were used: YCR047c; 59-
CAGCGGTTGATGAGG-39 10 bp from VDE-DSB1; 59-GCGAAT-
GAAAGACGTCTTGG-39 and 59-CGGCCCTCTTAATTAGAACTTC-
39 5.2 kb from VDE-DSB1; 59 -CGCACATTTCCCCGAAAA-39 and 59-
Primers close to the YCR047C promoter regions were used to
measure recovery of Spo11 DSB-associated sequences. The approx-
imate distance from the PCR product to the nearest DSB is 0.6 kb
(Buhler et al., unpublished data). A dilution series of genomic DNA
from dAG206 strain was used to establish a standard curve. The ChIP
enrichment was calculated as percentage of the target locus present
in the immunoprecipitated sample relative to the amount in the
starting input material. Corrected ChIP enrichment (Figure 6) was
calculated based on estimates of the maximum proportion of probed
chromatids likely to receive a DSB (Pmax). At YCR047C this value is
0.12 . For VDE-DSB1 the correction value accounts for knowledge
that up to 95% (Figure 2B) of all arg4-vde chromatids can receive a
VDE-DSB1, and the proportion of chromatids with homology to the
PCR primers and containing the arg4-vde allele varies with primer
site. For primers close to and far from VDE-DSB1, the respective
proportions of input chromatids containing the arg4-vde allele are
0.25 and 0.50 creating Pmaxvalues of 0.24 and 0.48. The corrected
ChIP enrichment is the original ChIP enrichment value/Pmaxand thus
reports on the ChIP enrichment as a proportion of chromatids that
will receive a DSB during the time course.
To determine the ratio between corrected ChIP enrichment and
proportion of chromatids expected to be single-stranded at the
primer site, ssDNA values close to the Spo11-DSB at YCR047C were
taken to be equal to the proportion of DSBs visible by Southern
analyses in wild-type and dmc1D strains  (unpublished data). For
the 10 bp from VDE-DSB1, the proportion of chromatids with ssDNA
close to the break site was taken as the total proportion of arg4-vde
chromatids present in resection bands at the relevant time points
(Figure 3A; unpublished data). For the 5.2 kb from VDE-DSB1, the
proportion of chromatids with ssDNA 8.5 kb from the break site was
calculated (Figure 3A; unpublished data) and used as a conservative
estimate of the proportion of chromatids that would be single-
stranded where the primers lie. The DNA used to determine the
proportion of chromatids with ssDNA was derived from different
time courses used in ChIP experiments.
Table S1. List of the Diploid Yeast Strains Used in This Work
Found at doi:10.1371/journal.pgen.0030223.st001 (21 KB XLS).
The accession numbers from the NCBI (http://www.ncbi.nlm.nih.gov)
gene database for genes discussed in this paper are DMC1 (856926),
HOP1 (854738), MEK1 (854533), NDT80 (?856524), RAD54 (852713),
RFA1 (851266), RED1 (850968), SAE2 (852700), and SPO11 (856364).
We thank members of the Goldman lab, Enrique (Fadri) Martinez-
Perez, anonymous referees, and the editor for comments that
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e2232347
Limiting Factors in DSB Repair
improved the manuscript; Douglas Bishop for strains and plasmids;
and Steven Brill for anti-Rfa1 antibody.
Author contributions. ASHG conceived and designed the experi-
ments. RJ, VB, MJN, AB-B, and MN performed the experiments. SH
performed the preliminary ChIP experiments. RJ, VB, MJN, AB-B,
MN, and ASHG analyzed the data. AN contributed reagents/materials/
analysis tools. ASHG wrote the paper with contributions from RJ, VB,
MJN, and AN.
Funding. This work was funded by Yorkshire Cancer Research and
BBSRC grants to ASHG and the Agence Nationale pour la Recherche
(06-BLAN-0120–01) to AN.
Competing interests. The authors have declared that no competing
1.Petronczki M, Siomos MF, Nasmyth K (2003) Un menage a quatre: the
molecular biology of chromosome segregation in meiosis. Cell 112: 423–
2. Keeney S, Giroux CN, Kleckner N (1997) Meiosis-specific DNA double-
strand breaks are catalyzed by Spo11, a member of a widely conserved
protein family. Cell 88: 375–384.
3.Sun H, Treco D, Schultes NP, Szostak JW (1989) Double-strand breaks at an
initiation site for meiotic gene conversion. Nature 338: 87–90.
4.Schwacha A, Kleckner N (1997) Interhomolog bias during meiotic
recombination: meiotic functions promote a highly differentiated inter-
homolog-only pathway. Cell 90: 1123–1135.
5.Bishop DK, Park D, Xu L, Kleckner N (1992) DMC1: a meiosis-specific yeast
homolog of E. coli recA required for recombination, synaptonemal
complex formation, and cell cycle progression. Cell 69: 439–456.
6.Niu H, Wan L, Baumgartner B, Schaefer D, Loidl J, et al. (2005) Partner
choice during meiosis is regulated by Hop1-promoted dimerization of
Mek1. Mol Biol Cell 16: 5804–5818.
7. Xu L, Weiner BM, Kleckner N (1997) Meiotic cells monitor the status of the
interhomolog recombination complex. Genes Dev 11: 106–118.
8.Thompson DA, Stahl FW (1999) Genetic control of recombination partner
preference in yeast meiosis. Isolation and characterization of mutants
elevated for meiotic unequal sister-chromatid recombination. Genetics
9.Bishop DK, Nikolski Y, Oshiro J, Chon J, Shinohara M, et al. (1999) High
copy number suppression of the meiotic arrest caused by a dmc1 mutation:
REC114 imposes an early recombination block and RAD54 promotes a
DMC1-independent DSB repair pathway. Genes Cells 4: 425–444.
10. Bailis JM, Roeder GS (1998) Synaptonemal complex morphogenesis and
sister-chromatid cohesion require Mek1-dependent phosphorylation of a
meiotic chromosomal protein. Genes Dev 12: 3551–3563.
11. Smith AV, Roeder GS (1997) The yeast Red1 protein localizes to the cores
of meiotic chromosomes. J Cell Biol 136: 957–967.
12. Hollingsworth NM, Byers B (1989) HOP1: a yeast meiotic pairing gene.
Genetics 121: 445–462.
13. Rockmill B, Roeder GS (1988) RED1: a yeast gene required for the
segregation of chromosomes during the reductional division of meiosis.
Proc Natl Acad Sci U S A 85: 6057–6061.
14. Lin FL, Sperle K, Sternberg N (1984) Model for homologous recombination
during transfer of DNA into mouse L cells: role for DNA ends in the
recombination process. Mol Cell Biol 4: 1020–1034.
15. Ozenberger BA, Roeder GS (1991) A unique pathway of double-strand
break repair operates in tandemly repeated genes. Mol Cell Biol 11: 1222–
16. Ivanov EL, Haber JE (1995) RAD1 and RAD10, but not other excision repair
genes, are required for double-strand break-induced recombination in
Saccharomyces cerevisiae. Mol Cell Biol 15: 2245–2251.
17. Ivanov EL, Sugawara N, Fishman-Lobell J, Haber JE (1996) Genetic
requirements for the single-strand annealing pathway of double-strand
break repair in Saccharomyces cerevisiae. Genetics 142: 693–704.
18. Sugawara N, Haber JE (1992) Characterization of double-strand break-
induced recombination: homology requirements and single-stranded DNA
formation. Mol Cell Biol 12: 563–575.
19. Shinohara A, Shinohara M, Ohta T, Matsuda S, Ogawa T (1998) Rad52
forms ring structures and co-operates with RPA in single-strand DNA
annealing. Genes Cells 3: 145–156.
20. Moreau S, Morgan EA, Symington LS (2001) Overlapping functions of the
Saccharomyces cerevisiae Mre11, Exo1, and Rad27 nucleases in DNA
metabolism. Genetics 159: 1423–1433.
21. Llorente B, Symington LS (2004) The Mre11 nuclease is not required for 59
to 39 resection at multiple HO-induced double-strand breaks. Mol Cell Biol
22. Khazanehdari KA, Borts RH (2000) EXO1 and MSH4 differentially affect
crossing over and segregation. Chromosoma 109: 94–102.
23. Umezu K, Sugawara N, Chen C, Haber JE, Kolodner RD (1998) Genetic
analysis of yeast RPA1 reveals its multiple functions in DNA metabolism.
Genetics 148: 989–1005.
24. Liu J, Wu TC, Lichten M (1995) The location and structure of double-strand
DNA breaks induced during yeast meiosis: evidence for a covalently linked
DNA-protein intermediate. EMBO J 14: 4599–4608.
25. Ohta K, Shibata T, Nicolas A (1994) Changes in chromatin structure at
recombination initiation sites during yeast meiosis. EMBO J 13: 5754–5763.
26. Miyamoto S, Mizutani R, Satow Y, Kawasaki M, Ohya Y, et al. (1999)
Recognition and cleavage of double-stranded DNA by yeast VMA1-derived
endonuclease. Nucleic Acids Symp Ser 197–198.
27. Nickoloff JA, Chen EY, Heffron F (1986) A 24-bp DNA sequence from the
MAT locus stimulates intergenic recombination in yeast. Proc Natl Acad
Sci U S A 83: 7831–7835.
28. Fukuda T, Nogami S, Ohya Y (2003) VDE-initiated intein homing in
Saccharomyces cerevisiae proceeds in a meiotic recombination-like manner.
Genes Cells 8: 587–602.
29. Malkova A, Klein F, Leung WY, Haber JE (2000) HO endonuclease-induced
recombination in yeast meiosis resembles Spo11-induced events. Proc Natl
Acad Sci U S A 97: 14500–14505.
30. Fukuda T, Ohya Y (2006) Recruitment of RecA homologs Dmc1p and
Rad51p to the double-strand break repair site initiated by meiosis-specific
endonuclease VDE (PI-SceI). Mol Genet Genomics 275: 204–214.
31. Neale MJ, Ramachandran M, Trelles-Sticken E, Scherthan H, Goldman ASH
(2002) Wild-type levels of Spo11-induced DSBs are required for normal
single-strand resection during meiosis. Mol Cell 9: 835–846.
32. Clerici M, Mantiero D, Lucchini G, Longhese MP (2005) The Saccharomyces
cerevisiae Sae2 protein promotes resection and bridging of double-strand
break ends. J Biol Chem 280: 38631–38638.
33. Xu L, Ajimura M, Padmore R, Klein C, Kleckner N (1995) NDT80, a meiosis-
specific gene required for exit from pachytene in Saccharomyces cerevisiae.
Mol Cell Biol 15: 6572–6581.
34. Chu S, DeRisi J, Eisen M, Mulholland J, Botstein D, et al. (1998) The
transcriptional program of sporulation in budding yeast. Science 282: 699–
35. Lydall D, Nikolsky Y, Bishop DK, Weinert T (1996) A meiotic recombina-
tion checkpoint controlled by mitotic checkpoint genes. Nature 383: 840–
36. Diaz RL, Alcid AD, Berger JM, Keeney S (2002) Identification of residues in
yeast Spo11p critical for meiotic DNA double-strand break formation. Mol
Cell Biol 22: 1106–1115.
37. Henderson KA, Keeney S (2004) Tying synaptonemal complex initiation to
the formation and programmed repair of DNA double-strand breaks. Proc
Natl Acad Sci U S A 101: 4519–4524.
38. Arbel A, Zenvirth D, Simchen G (1999) Sister chromatid-based DNA repair
is mediated by RAD54, not by DMC1 or TID1. EMBO J 18: 2648–2658.
39. Shinohara M, Shita-Yamaguchi E, Buerstedde JM, Shinagawa H, Ogawa H,
et al. (1997) Characterization of the roles of the Saccharomyces cerevisiae
RAD54 gene and a homologue of RAD54, RDH54/TID1, in mitosis and
meiosis. Genetics 147: 1545–1556.
40. Resnick MA, Zgaga Z, Hieter P, Westmoreland J, Fogel S, et al. (1992)
Recombinant repair of diverged DNAs: a study of homoeologous
chromosomes and mammalian YACs in yeast. Mol Gen Genet 234: 65–73.
41. Nilsson-Tillgren T, Gjermansen C, Holmberg S, Peterson JGL, Kielland-
Brandt MC (1986) Analysis of Chromosome V and the ILV1 gene from
Saccharomyces carlsebergensis. Carlsberg Res Commun 51: 309–326.
42. Kielland-Brandt MC, Gjermansen C, Nilsson-Tillgren T, Peterson JGL,
Holmberg S (1981) Genetic differences between Saccharomyces carlsebergensis
and S. cerevisiae. Analysis of Chromosome III by single chromosome
transfer. Carlsberg Res Commun 46: 65–76.
43. Krogh BO, Symington LS (2004) Recombination proteins in yeast. Annu
Rev Genet 38: 233–271.
44. Goldway M, Sherman A, Zenvirth D, Arbel T, Simchen G (1993) A short
chromosomal region with major roles in yeast Chromosome III meiotic
disjunction, recombination, and double- strand breaks. Genetics 133: 159–
45. Fukuda T, Nagai Y, Ohya Y (2004) Molecular mechanism of VDE-initiated
intein homing in yeast nuclear genome. Adv Biophys 38: 215–232.
46. Borde V, Goldman ASH, Lichten M (2000) Direct coupling between meiotic
DNA replication and recombination initiation. Science 290: 806–809.
47. Allers T, Lichten M (2001) Differential timing and control of noncrossover
and crossover recombination during meiosis. Cell 106: 47–57.
48. Schwacha A, Kleckner N (1994) Identification of joint molecules that form
frequently between homologs but rarely between sister chromatids during
yeast meiosis. Cell 76: 51–63.
49. Chabes A, Domkin V, Thelander L (1999) Yeast Sml1, a protein inhibitor of
ribonucleotide reductase. J Biol Chem 274: 36679–36683.
50. Zhao X, Rothstein R (2002) The Dun1 checkpoint kinase phosphorylates
and regulates the ribonucleotide reductase inhibitor Sml1. Proc Natl Acad
Sci U S A 99: 3746–3751.
51. Zhao X, Chabes A, Domkin V, Thelander L, Rothstein R (2001) The
ribonucleotide reductase inhibitor Sml1 is a new target of the Mec1/Rad53
kinase cascade during growth and in response to DNA damage. EMBO J 20:
52. Koc A, Merrill GF (2007) Checkpoint deficient rad53–11 yeast cannot
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e2232348
Limiting Factors in DSB Repair
accumulate dNTPs in response to DNA damage. Biochem Biophys Res
Commun 353: 527–530.
53. Zhao X, Muller EG, Rothstein R (1998) A suppressor of two essential
checkpoint genes identifies a novel protein that negatively affects dNTP
pools. Mol Cell 2: 329–340.
54. Brill SJ, Stillman B (1991) Replication factor-A from Saccharomyces cerevisiae
is encoded by three essential genes coordinately expressed at S phase.
Genes Dev 5: 1589–1600.
55. Sugiyama T, Zaitseva EM, Kowalczykowski SC (1997) A single-stranded
DNA-binding protein is needed for efficient presynaptic complex
formation by the Saccharomyces cerevisiae Rad51 protein. J Biol Chem 272:
56. Sung P (1997) Yeast Rad55 and Rad57 proteins form a heterodimer that
functions with replication protein A to promote DNA strand exchange by
Rad51 recombinase. Genes Dev 11: 1111–1121.
57. Sung P (1997) Function of yeast Rad52 protein as a mediator between
replication protein A and the Rad51 recombinase. J Biol Chem 272: 28194–
58. Kim PM, Paffett KS, Solinger JA, Heyer WD, Nickoloff JA (2002)
Spontaneous and double-strand break-induced recombination, and gene
conversion tract lengths, are differentially affected by overexpression of
wild-type or ATPase-defective yeast Rad54. Nucleic Acids Res 30: 2727–
59. Kim PM, Allen C, Wagener BM, Shen Z, Nickoloff JA (2001) Overexpression
of human RAD51 and RAD52 reduces double-strand break-induced
homologous recombination in mammalian cells. Nucleic Acids Res 29:
60. Kuznetsov S, Pellegrini M, Shuda K, Fernandez-Capetillo O, Liu Y, et al.
(2007) RAD51C deficiency in mice results in early prophase I arrest in
males and sister chromatid separation at metaphase II in females. J Cell
Biol 176: 581–592.
61. Cherry SM, Adelman CA, Theunissen JW, Hassold TJ, Hunt PA, et al. (2007)
The Mre11 complex influences DNA repair, synapsis, and crossing over in
murine meiosis. Curr Biol 17: 373–378.
62. Tsubouchi H, Roeder GS (2006) Budding yeast Hed1 downregulates the
mitotic recombination machinery when meiotic recombination is im-
paired. Genes Dev 20: 1766–1775.
63. Tsubouchi H, Roeder GS (2003) The importance of genetic recombination
for fidelity of chromosome pairing in meiosis. Dev Cell 5: 915–925.
64. Wan L, Zhang C, Shokat KM, Hollingsworth NM (2006) Chemical
inactivation of Cdc7 kinase in budding yeast results in a reversible arrest
that allows efficient cell synchronization prior to meiotic recombination.
Genetics 174: 1767–1774.
65. Franklin AE, McElver J, Sunjevaric I, Rothstein R, Bowen B, et al. (1999)
Three-dimensional microscopy of the Rad51 recombination protein during
meiotic prophase. Plant Cell 11: 809–824.
66. Pawlowski WP, Golubovskaya IN, Cande WZ (2003) Altered nuclear
distribution of recombination protein RAD51 in maize mutants suggests
the involvement of RAD51 in meiotic homology recognition. Plant Cell 15:
67. Prieler S, Penkner A, Borde V, Klein F (2005) The control of Spo119s
interaction with meiotic recombination hotspots. Genes Dev 19: 255–269.
68. Borde V, Lin W, Novikov E, Petrini JH, Lichten M, et al. (2004) Association
of Mre11p with double-strand break sites during yeast meiosis. Mol Cell 13:
69. Wu TC, Lichten M (1995) Factors that affect the location and frequency of
meiosis-induced double-strand breaks in Saccharomyces cerevisiae. Genetics
PLoS Genetics | www.plosgenetics.org November 2007 | Volume 3 | Issue 11 | e2232349
Limiting Factors in DSB Repair