Suppression of the DNA repair defects of BRCA2-deficient cells with heterologous protein fusions
The BRCA2 tumor suppressor plays an important role in the repair of DNA damage by homologous recombination, also termed homology-directed repair (HDR). Human BRCA2 is 3,418 aa and is composed of several domains. The central part of the protein contains multiple copies of a motif that binds the Rad51 recombinase (the BRC repeat), and the C terminus contains domains that have structural similarity to domains in the ssDNA-binding protein replication protein A (RPA). To gain insight into the role of BRCA2 in the repair of DNA damage, we fused a single (BRC3, BRC4) or multiple BRC motifs to the large RPA subunit. Expression of any of these protein fusions in Brca2 mutant cells substantially improved HDR while suppressing mutagenic repair. A fusion containing a Rad52 ssDNA-binding domain also was active in HDR. Mutations that reduced ssDNA or Rad51 binding impaired the ability of the fusion proteins to function in HDR. The high level of spontaneous chromosomal aberrations in Brca2 mutant cells was largely suppressed by the BRC-RPA fusion proteins, supporting the notion that the primary role of BRCA2 in maintaining genomic integrity is in HDR, specifically to deliver Rad51 to ssDNA. The fusion proteins also restored Rad51 focus formation and cellular survival in response to DNA damaging agents. Because as little as 2% of BRCA2 fused to RPA is sufficient to suppress cellular defects found in Brca2-mutant mammalian cells, these results provide insight into the recently discovered diversity of BRCA2 domain structures in different organisms. • double-strand break • mammalian cells • Rad51 • homologous recombination • BRC repeat
Suppression of the DNA repair defects of
BRCA2-deficient cells with heterologous
, Nicolas Siaud*
, Nicole Christ*, Wouter W. Wiegant
, Paul P. W. van Buul
, Mingguang Han*,
Małgorzata Z. Zdzienicka
, Jeremy M. Stark*
, and Maria Jasin*
*Molecular Biology Program, Memorial Sloan–Kettering Cancer Center, 1275 York Avenue, New York, NY 10021;
Department of Toxicogenetics, Leiden
University Medical Center, Einthovenweg 20, P.O. Box 9600, Postzone S4-P, 2300 RC, Leiden, The Netherlands;
Department of Molecular Cell Genetics,
The L. Rydygier Collegium Medicum, Nicolaus Copernicus University, ul. Sklodowskiej–Curie 9, 85-094, Bydgoszcz, Poland; and
Department of Radiation
Biology, City of Hope National Medical Center, Beckman Research Institute, Duarte, CA 91010
Edited by Charles M. Radding, Yale University School of Medicine, New Haven, CT, and approved April 24, 2006 (received for review January 11, 2006)
The BRCA2 tumor suppressor plays an important role in the repair
of DNA damage by homologous recombination, also termed ho-
mology-directed repair (HDR). Human BRCA2 is 3,418 aa and is
composed of several domains. The central part of the protein
contains multiple copies of a motif that binds the Rad51 recombi-
nase (the BRC repeat), and the C terminus contains domains that
have structural similarity to domains in the ssDNA-binding protein
replication protein A (RPA). To gain insight into the role of BRCA2
in the repair of DNA damage, we fused a single (BRC3, BRC4) or
multiple BRC motifs to the large RPA subunit. Expression of any of
these protein fusions in Brca2 mutant cells substantially improved
HDR while suppressing mutagenic repair. A fusion containing a
Rad52 ssDNA-binding domain also was active in HDR. Mutations
that reduced ssDNA or Rad51 binding impaired the ability of the
fusion proteins to function in HDR. The high level of spontaneous
chromosomal aberrations in Brca2 mutant cells was largely sup-
pressed by the BRC-RPA fusion proteins, supporting the notion that
the primary role of BRCA2 in maintaining genomic integrity is in
HDR, speciﬁcally to deliver Rad51 to ssDNA. The fusion proteins
also restored Rad51 focus formation and cellular survival in re-
sponse to DNA damaging agents. Because as little as 2% of BRCA2
fused to RPA is sufﬁcient to suppress cellular defects found in
Brca2-mutant mammalian cells, these results provide insight into
the recently discovered diversity of BRCA2 domain structures in
double-strand break 兩 mammalian cells 兩 Rad51 兩 homologous
recombination 兩 BRC repeat
oss of the BRCA2 tumor suppressor predispose s adult mutation
carriers to breast and ovarian cancer (1) and children with
biallelic mutations to multiple types of malignancies, including brain
and kidney tumors and leukemia (2–4). BRCA2 also plays a critical
role in the mouse during early embryonic development (5) and
during meiosis (6) and is critically important in cells for the
maintenance of genomic integrity (7). The BRCA2 protein pro-
vides an important function in promoting homology-directed repair
(HDR) of damaged DNA in cells (8, 9), presumably through its
interaction with the Rad51 recombinase (10, 11), which may
underlie its role in tumor suppression and development.
Mammalian BRCA2 proteins are ⬎3,300 aa in length and
contain several functional domains (Fig. 1a) (7). Although the role
of the N-terminal third of the protein is uncertain, the central
⬇1,000 aa of BRCA2 contain eight BRC repeats that bind Rad51
(10). In their normal context, the constellation of BRC repeats
presumably promote s HDR; however, cellular expression of indi-
vidual BRC repeats recapitulates phenotypes of Brca2 mutant cells,
including reduced HDR, implying that by themselves they interfere
with Rad51 function (12–14). C-terminal to the BRC repeats is a
region that binds ssDNA (15, 16). This region consists of four
globular domains, including two oligonucleotide兾oligosaccharide-
binding (OB) folds, which have close structural homology to two
OB folds in replication protein A (RPA) 70, the largest subunit of
the ssDNA-binding protein RPA (15), and an unusual tower
domain appended to one of the globular domains (16). The extreme
C terminus of BRCA2 contains an additional Rad51-binding motif
(11), which is distinct from the BRC repeats, and which has been
shown recently to undergo regulated Rad51 binding in response to
CDK phosphorylation (17).
Although initially identified in mammalian cells, BRCA2 or-
thologs have been identified more recently in diverse organisms,
including Ustilago maydis (18), Caenorhabditis elegans (19), and
Arabidopsis thaliana (20). However, BRCA2 orthologs in different
species have widely diverse sizes. For example, in contrast to the
⬎3,300-aa vertebrate BRCA2 proteins, U. maydis Brh2 is 1,075 aa
(18) and C. elegans BRC-2 is 394 aa (19). The nearly 10-fold size
range of BRCA2 orthologs is due to highly variable and poorly
conserved N-terminal sequences, variation in the number of BRC
repeats, and domain differences in the DNA-binding region. De-
spite this variation, all BRCA2 orthologs have at least one BRC
repeat capable of binding Rad51 and apparently at least one
domain capable of binding ssDNA. In this report, we sought to
determine whether BRCA2 function in HDR in mammalian cells
could be contained essentially within these two identified activities,
that of Rad51 and ssDNA binding, by using a heterologous ssDNA-
binding protein fused to BRC motifs. We found that the se heter-
ologous fusion peptide s, which contain as little as 2% of BRCA2,
restored HDR to BRCA2 mutant cells, while concomitantly sup-
pressing genetic instability.
BRC-RPA Fusion Proteins Increase HDR in
Mutant Cells but Not
To te st whether a heterologous ssDNA-binding
protein could effectively substitute for the BRCA2 ssDNA-binding
domain in HDR, we fused human RPA70 to one or several BRC
repeats from BRCA2. For the single BRC repeat, we used the 70-aa
BRC3 (Fig. 1b), which interfere s with HDR when expressed on its
own (14, 21). As a control, we also fused RPA70 to BRC3⌬, which
is identical to BRC3 except for a 7-aa deletion that abrogates Rad51
binding (22). In addition, we fused larger segments of BRCA2
containing multiple repeats, in particular BRC1-2 and BRC1-4
(Fig. 1b). Expression vectors for the BRC-RPA fusion proteins and
Conﬂict of interest statement: No conﬂicts declared.
This paper was submitted directly (Track II) to the PNAS ofﬁce.
Abbreviations: DSB, double-strand break; HDR, homology-directed repair; IR, ionizing
radiation; MMC, mitomycin C; OB, oligonucleotide兾oligosaccharide-binding; RPA, replica-
tion protein A.
H.S. and N.S. contributed equally to this work.
To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
© 2006 by The National Academy of Sciences of the USA
June 6, 2006
no. 23 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0600298103
the BRC repeats were transiently transfected into the Brca2 mutant
V-C8 hamster cell line (23). By Western blot analysis, each of the
proteins was detected at the expected size, although the smaller
fusion proteins appeared to be better expressed than the larger one s
To assess HDR, we used the DR-GFP reporter (Fig. 1d).
With this reporter, HDR is detected when a double-strand
break (DSB) introduced into the chromosome by the I-SceI
endonuclease is repaired by HDR to give rise to GFP-positive
cells (24). HDR was reduced by ⬎20-fold in Brca2 mutant
V-C8 cells relative to wild-t ype V79 hamster cells (compare
vector control in Fig. 1 e and f ). Strik ingly, transient expression
of the BRC-RPA proteins substantially increased HDR in the
V-C8 cells (Fig. 1e). The increased HDR was estimated to be
4- to 6-fold, varying somewhat between the dif ferent BRC-
RPA proteins, although there was not a direct c orrelation
bet ween HDR level and the number of BRC repeats in the
fusion proteins. Improved HDR depended on the BRC-RPA
proteins having an intact Rad51-binding motif in the BRC
repeat, because BRC3⌬RPA ex pression did not increase
HDR. The presence of RPA70 also was critical, because the
BRC repeats themselves did not improve HDR. The increased
HDR observed in the Brca2 mut ant cells with BRC-RPA
ex pression was not due to the proteins conferring a general
hyperrec ombination phenotype to cells, because HDR was not
increased in wild-type V79 hamster cells (Fig. 1f ). Rather, the
BRC-RPA fusion proteins, including BRC3⌬RPA, reduced
HDR somewhat in w ild-type cells, although not as strongly as
the BRC repeats (Fig. 1f ). It is possible that the fusion proteins
interfere w ith endogenous RPA function.
The eight BRC repeats share only a c ore consensus sequence
and appear to bind Rad51 w ith different affinities (10, 22, 25).
Fig. 1. Transient expression of BRC-RPA fusion pro-
teins increases HDR in Brca2 mutant cells. (a) Human
BRCA2. BRCA2 has a central region containing eight
BRC repeats which bind Rad51. C-terminal to the BRC
repeats is a region of higher conservation that encom-
passes a DNA-binding domain (DBD). A distinct Rad51-
binding motif is found at the C terminus. (b) BRC-RPA
fusion proteins and BRC repeat peptides. RPA70 refers
to the entire human RPA70-coding sequence. BRC3⌬
contains a 7-aa deletion that abrogates Rad51 bind-
ing. A nuclear localization signal (nls) is present at the
N terminus and a FLAG epitope tag at the C terminus.
(c) Western blot analysis of Brca2 mutant V-C8 hamster
cells transiently expressing the BRC-RPA fusion pro-
teins and BRC repeat peptides. Sizes of the expressed
peptides are as expected. Asterisks denote a back-
ground immunoreactive band. (d) Flow cytometric
analysis of V-C8 cells demonstrates increased HDR af-
ter transient expression of the BRC3RPA fusion protein
as compared with BRC3⌬RPA. V-C8 cells containing a
chromosomal DR-GFP reporter were cotransfected
with expression vectors for the I-SceI endonuclease
and the indicated BRC-RPA fusion protein. Cleavage of
the DR-GFP reporter at the I-SceI site in vivo at the
SceGFP gene and repair by HDR directed by the down-
stream iGFP repeat results in GFP-positive cells. (e) HDR
is increased in V-C8 cells with transient expression of
the BRC-RPA fusion proteins but not with the BRC
repeat peptides. Asterisks indicate a statistically signif-
icant difference from transfection with the empty ex-
pression vector by using an unpaired t test (BRC3RPA,
P ⫽ 0.013; BRC1–2RPA, P ⫽ 0.0004; BRC1–4RPA, P ⫽
0.0048; BRC4RPA, P ⬍ 0.0001) (see also Fig. 4a). (f) HDR
in wild-type V79 hamster cells with expression of the
BRC peptides or the BRC-RPA fusion proteins. Asterisks
indicate a statistically signiﬁcant difference from
transfection with the empty expression vector (BRC3,
P ⫽ 0.0002; BRC1-2, P ⫽ 0.0006; BRC1-4, P ⫽ 0.0007)
(see also Fig. 4b). (g) HDR in CAPAN-1 cells is increased
by transient expression of BRC1-2RPA. The human
pancreatic adenocarcinoma cell line CAPAN-1 carries a
6174delT mutation on one allele of BRCA2 with loss of
the other wild-type BRCA2 allele. Asterisk indicates a
statistically signiﬁcant difference in HDR between
transfection of the BRC1-2RPA expression vector and
the empty expression vector (P ⬍ 0.05). Error bars in
e–g indicate 1 SD from the mean. Results are derived
from three independent transfections for each sample
(n ⫽ 3).
Saeki et al. PNAS
June 6, 2006
In addition, BRC3 and BRC4, which have only 30% identity,
appear to interact with Rad51 nonequivalently, with BRC3
binding to the N-terminal domain and BRC4 binding to the
nucleotide-binding c ore domain (25, 26). To determine whether
BRC3 and BRC4 act differently when fused to RPA70, we
c onstr ucted a BRC4RPA fusion protein (Fig. 1b). Like
BRC3RPA, BRC4RPA expression increased HDR in the V-C8
cells (Fig. 1e; see also Fig. 4a, which is published as supporting
infor mation on the PNAS web site). As expected, the isolated
BRC4 repeat did not promote HDR in the V-C8 cells (Fig. 1e)
but did substantially reduce HDR in wild-type cells, as did
BRC4RPA to a lesser extent (Fig. 4b). These results indicate that
when BRC4 is tethered to a ssDNA-binding domain, it is
functional to promote HDR.
We also assessed the ability of the BRC-RPA proteins to correct
the HDR defect found in the human pancreatic adenocarcinoma
cell line CAPAN-1, which expre sse s a cytoplasmic BRCA2 peptide
truncated at BRC7 (27). As with the V-C8 hamster cells, HDR
increased in CAPAN-1 cells with expression of the BRC1–2RPA as
compared with the vector control (Fig. 1g), indicating that the
BRC-RPA enhancement of HDR is not specific to the Brca2
mutant hamster cells.
Stable Expression of BRC-RPA Fusion Proteins in
Increases HDR While Suppressing Mutagenic DSB Repair by Single-
Strand Annealing (SSA).
To further investigate the repair phenotype s
of Brca2 mutant cells expressing the fusion proteins, each of the
BRC-RPA proteins was stably expressed in the V-C8 cells (Fig. 2a).
Immunoprecipiation by using the FLAG epitope, followed by
immunoblotting for Rad51, confirmed that the fusion proteins with
intact BRC repeats bound Rad51, whereas the BRC3⌬RPA pro-
tein did not (Fig. 2a). As with transient expression, HDR was
increased in V-C8 cells stably expressing the BRC-RPA proteins
(Fig. 2b). The overall increase in HDR was 10- to 11-fold, which was
greater than that obtained by transient expression of the fusion
proteins, such that the absolute level of HDR approached that
found in wild-type cells (Fig. 2b). Cells that expressed BRC3RPA
and BRC1-2RPA had similar levels of the respective fusion protein
and also had similar levels of HDR, indicating that one or two BRC
repeats fused to RPA70 functioned equivalently. Interestingly,
BRC1-4RPA corrected the HDR defect as well as these other two
fusions, even though it was expressed at a lower level.
HDR appears to be in competition w ith a second DSB repair
pathway, ter med SSA (14, 28). When a DSB occurs bet ween
sequence repeats and is resected to ssDNA, annealing of the
Fig. 2. Correction of the HDR defect and phe-
notypes associated with impaired HDR in Brca2
mutant cells by stable expression of the BRC-
RPA proteins. (a) BRC-RPA expression vectors
were cotransfected into V-C8 cells with a neo-
mycin phosphotransferase gene (neo⫹). Immu-
noprecipitation with anti-FLAG antibody fol-
lowed by Western blot analysis shows BRC-RPA
-FLAG) and interaction with
-Rad51). (b) HDR is increased nearly to
wild-type levels in V-C8 cells stably expressing
BRC-RPA proteins. Asterisks indicate a statisti-
cally signiﬁcant difference from parental V-C8
cells (P ⬍ 0.0001; n ⫽ 6). (c) SSA is suppressed in
V-C8 cells stably expressing BRC-RPA proteins.
The 0.8-kb PCR fragment derived from primers
SA-F and SA-R2 speciﬁcally detects the SSA re-
pair product, whereas the 1.1-kb PCR fragment
from primers SA-F and SA-R1 detects a structur-
ally intact reporter, i.e., from HDR and NHEJ, as
well as the parental DR-GFP reporter. See Fig.
5a for quantitation. (d) Rad51 focus formation
in response to DNA-damaging agents is re-
stored in V-C8 cells stably expressing BRC-RPA
proteins. Representative wild-type (V79), Brca2
mutant (V-C8), or Brca2 mutant cells stably ex-
pressing the indicated BRC-RPA peptides are
shown after IR. Note that Rad51 is diffusely
present on the chromatin of the BRC3RPA and
BRC1–4RPA cell lines. Exponentially growing
cells were irradiated with 12 Gy of IR and ana-
lyzed 7 h later for Rad51 foci. See Fig. 5b for
quantitation. (e) Hypersensitivity of Brca2 mu-
tant cells to DNA-damaging agents is reduced
or eliminated with BRC-RPA expression. Sur-
vival was assessed by crystal violet staining after
treatment for 24 h (MMC) or by clonogenic
survival (IR). The percent survival of cells treated
with DNA-damaging agents was computed rel-
ative to that of untreated cells, which was set to
100% for each line. Each percentage shown is
the mean and error bars represent the SDs.
MMC treatments were triplicated except that
BRC3RPA and BRC1-2RPA stable cell lines were
treated six times. IR treatment for each dose
was performed once, except that 6-Gy treat-
ments were quadruplicated.
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0600298103 Saeki et al.
c omplement ary ssDNA for med at the repeats leads to SSA
(28), such that one repeat and the sequences between the
repeats are deleted (Fig. 2c). When HDR is impaired as a
result of BRCA 2 or Rad51 deficienc y, SSA levels are elevated
(9, 14), presumably because HDR and SSA compete for the
same ssDNA inter mediates.
To determine whether the BRC-RPA proteins, like BRCA2
itself, suppressed DSB repair by SSA, we used a PCR assay to
amplify the SSA deletion product after I-SceI expression (Fig. 2c;
ref. 29). The level of the SSA-specific product was compared with
the level of a control product from a structurally intact DR-GFP
reporter. Cell line s stably expressing intact BRC-RPA fusion pro-
teins suppressed SSA compared with parental V-C8 cells or cells
expre ssing BRC3⌬RPA (Fig. 2c; see also Fig. 5a, which is published
as supporting information on the PNAS web site). Thus, intact
BRC-RPA fusion proteins acted to restore the balance of SSA and
HDR in the Brca2 mutant cells.
Restored Rad51 Focus Formation and Decreased Sensitivity of
Mutant Cells to DNA Damaging Agents by BRC-RPA Expression. The
results of the DSB repair assays suggested that BRC-RPA proteins
restore other indicators of Rad51 function to Brca2 mutant cells. In
response to DNA damage, Rad51 exhibits a dynamic redistribution
within cells, localizing in nuclear foci to sites of DNA damage (30).
Brca2 mutant cells, including the V-C8 cell line (Fig. 2d), are
defective in DNA damage-induced Rad51 focus formation (13, 23).
To determine whether V-C8 cells stably expressing the intact
BRC-RPA proteins have recovered the ability to form Rad51 foci
after DNA damage induction, cells were treated with ionizing
radiation (IR) or the crosslinking agent mitomycin C (MMC). We
found that Rad51 focus formation was significantly restored by
BRC-RPA expression in re sponse to either agent (Figs. 2d and 5b).
The restoration of Rad51 focus formation depends on the ability of
the fusion proteins to bind Rad51, because cells expressing
BRC3⌬RPA do not form DNA damage-induced foci.
Interestingly, we noted that in addition to being present in bright
foci upon DNA damage, a portion of the Rad51 in V-C8 cells
expre ssing the intact BRC-RPA proteins also was diffusely present
in the nucleus, possibly on chromatin (e.g., compare BRC3RPA
with wild-type). This result suggested that full-length BRCA2
prevents Rad51 from interacting with undamaged chromatin, sup-
porting a previously proposed role for BRCA2 in regulating Rad51
Like other mammalian HDR mutants, Brca2 mutant cells display
an increased sensitivity to DNA damaging agents, especially
crosslinking agents (7, 23). We found that the hypersensitivity of
V-C8 cells to IR and MMC was greatly reduced or eliminated by
expre ssion of the intact BRC-RPA proteins (Fig. 2e), indicating that
the fusion proteins are proficient in the repair of DNA damage
produced by these agents.
Spontaneous Chromosomal Aberrations Are Reduced in
Cells Stably Expressing BRC-RPA Fusion Proteins.
Brca2 mutant cells,
including V-C8 cells, exhibit compromised genomic integrity, even
in the absence of DNA damaging agents (7, 23). We examined
spontaneous chromosomal aberrations in Brca2 mutant cells stably
expre ssing the BRC-RPA fusion proteins and found that they were
substantially reduced (Table 1). Whereas V-C8 cells had a 20-fold
increase in aberrations relative to V79 cells, cells expressing the
intact BRC-RPA proteins had only a 3- to 4-fold increased level of
aberrations. Thus, the BRC-RPA fusion proteins can partially or
fully correct many of the cellular phenotype s associated with
BRCA2 deficiency, including compromised genomic integrity
BRC-RPA Fusion Proteins Interact with Other RPA Subunits. RPA70
has roles in various cellular processe s as part of a heterotrimeric
complex (32). We investigated whether the BRC-RPA fusion
proteins interact with the other RPA subunits, RPA32 and RPA14.
Whole-cell extracts from V-C8 cells stably expressing either
BRC3RPA or BRC3⌬RPA were probed with antibodie s directed
against RPA70, RPA32, or RPA14 after immunoprecipitation with
an anti-FLAG antibody (Fig. 3a). The BRC-RPA fusion proteins
coimmunoprecipitated with RPA32 and with RPA14, suggesting
that the BRC-RPA proteins function in HDR as part of a hetero-
trimeric complex, like RPA70 itself.
HDR Correction and ssDNA-Binding Activity. It is formally possible
that the BRC-RPA heterotrimeric complex promotes HDR in
Table 1. Spontaneous chromosomal aberrations are reduced in
Brca2 mutant cells stably expressing BRC-RPA fusion proteins
⫾ SEM, %
V79 150 6 7 0 7 ⫾ 0.7
V-C8 150 73 137 2 141 ⫾ 17
⫹BRC3RPA 150 17 23 0 23 ⫾ 5
⫹BRC3⌬RPA 150 65 128 2 132 ⫾ 21
⫹BRC1-4RPA 150 27 33 0 33 ⫾ 5
Fig. 3. HDR and ssDNA binding. (a) Interaction of BRC-RPA fusion proteins
with the other RPA subunits. Whole-cell extracts from cells stably expressing
BRC3RPA or BRC3⌬RPA were probed after gel electrophoresis either directly
-RPA70, RPA32, and RPA14 antibodies or after immunoprecipitation
-FLAG antibody. Puriﬁed heterotrimeric RPA and whole-cell extracts
from MCF-7 cells also are included. The weak band below the BRC-RPA
proteins in the IP lanes is likely a degradation product that runs slightly above
the position of RPA70 itself. (b) RPA mutations that reduce (K263A) or abolish
(R234A兾K263A) ssDNA-binding interfere with the ability of the BRC-RPA
proteins to function in HDR. Asterisk indicates a statistically signiﬁcant differ-
ence for transfection of the wild-type BRC3RPA expression vector compared
with the other BRC3RPA expression vectors (K263A, P ⫽ 0.041 and R234A兾
K263A, P ⫽ 0.031; n ⫽ 3). (c) HDR is increased in V-C8 cells with transient
expression of a Brh2-Rad52 fusion protein. The Brh2-Rad52 fusion contains
the BRC repeat from U. maydis Brh2 with surrounding sequences (amino acids
89–551) fused to the conserved ssDNA binding domain of U. maydis Rad52
(amino acids 79 –314); Brh2 extends from amino acids 89 –955 and, hence, is
truncated in the DNA-binding domain (18). Asterisks indicate a statistically
signiﬁcant difference from transfection with the empty expression vector
(BRC3RPA, P ⫽ 0.0022 and Brh2-Rad52, P ⫽ 0.0027; n ⫽ 3).
Saeki et al. PNAS
June 6, 2006
Brca2 mutant cells independent of ssDNA binding. To rule out this
possibility, we incorporated RPA70 ssDNA binding mutations into
BRC3RPA (Fig. 3b). A strong DNA binding mutant (R234A兾
K263A; 0.9% ssDNA binding affinity; ref. 33) completely abolished
the ability of the BRC3RPA peptide to correct the HDR defect in
the V-C8 cells, whereas a weak mutant (K263A; 18% ssDNA
binding affinity; ref. 33) had residual activity (Fig. 3b). Thus, ssDNA
binding is required for the BRC-RPA fusions to promote HDR.
Although both RPA and BRCA2 contain OB folds, the number
and organization of ssDNA-binding domains in these proteins is
quite different (15, 16), raising the possibility that other ssDNA-
binding domains could promote HDR when fused to a BRC repeat.
We examined a Brh2-Rad52 fusion, which contains a portion of
Brh2 (BRC repeat and surrounding sequence) and the Rad52
ssDNA-binding domain, both derived from U. maydis (M. Kojic and
W. K. Holloman, unpublished results). The Brh2-Rad52 fusion
increased HDR ⬇2-fold in the V-C8 cells (Fig. 3c) while having no
effect in wild-type cells (data not shown). Increased HDR in the
V-C8 cells depended on the Rad52 ssDNA-binding domain, be-
cause a Brh2 peptide had no effect on its own (Fig. 3c). These results
demonstrate that a structurally intact ssDNA-binding domain can
promote HDR in the Brca2 mutant cells.
We have demonstrated that as little as 2% of BRCA2 fused to the
large subunit of the ssDNA-binding protein RPA is sufficient to
suppress cellular defects found in Brca2 mutant mammalian cells.
BRC-RPA expression restores precise repair of DSBs by HDR to
nearly wild-type levels, concomitantly reducing mutagenic DSB
repair by SSA. BRC-RPA expression also promotes DNA damage-
induced Rad51 focus formation and resistance to DNA damaging
agents and, importantly, suppresses spontaneous chromosome ab-
errations in the Brca2 mutant cells. Because the heterologous
BRC-RPA fusion proteins are unlikely to substitute for BRCA2 in
functions other than HDR (such as transcriptional transactivation)
(34), these results emphasize the role of BRCA2 in HDR for
maintaining genomic integrity.
An early step in HDR is the exonucleolytic processing of DSBs
to produce ssDNA tails, which rapidly become coated with RPA
(35). In vitro, RPA minimizes secondary structure in ssDNA to
promote Rad51 nucleoprotein filament assembly; however, RPA
must be displaced by Rad51 with the assistance of mediator proteins
for efficient filament assembly. The ability of the BRC-RPA
peptides to correct the HDR defect in Brca2 mutant cells, in a
manner dependent on Rad51 and ssDNA binding, implies that the
key role of BRCA2 in the cell is to deliver Rad51 to ssDNA.
Although the ability of the BRC-RPA proteins to complement
Brca2 mutants is quite surprising, our results provide insight into the
recently discovered plasticity of BRCA2 structures in different
organisms (18, 19, 36) and biochemistry on BRCA2 fragments (37).
For example, our smallest fusion, BRC3RPA, may be considered
analogous to BRCA2 orthologs, which contain a single BRC repeat
[U. maydis Brh2 (18) and C. elegans BRC-2 (19)] or have truncated
DNA-binding regions (e.g., C. elegans BRC-2, which does not have
a tower domain). Because our results indicate that much of mam-
malian BRCA2 may be dispensable for HDR when substituted with
a heterologous ssDNA binding protein, a challenge will be to
determine the physiological roles for the remaining portions of this
Phenotype s of U. maydis expressing a Brh2-RPA70 fusion pro-
tein were reported recently (38). Unlike the BRC-RPA proteins in
mammalian cells, the Brh2-RPA70 fusion caused a hyperrecombi-
nation phenotype in U. maydis. The Brh2-RPA70 fusion differs
from the BRC-RPA proteins in that it includes the entire N-
terminal half of the Brh2 protein (551 aa) rather than an isolated
BRC repeat or sets of BRC repeats. Therefore, the hyperrecom-
bination phenotype caused by Brh2-RPA70 may point to additional
roles for the N terminus in regulating Rad51 function, although it
should be noted that this region of the protein is highly diverged
among BRCA2 orthologs.
A crystal structure has been solved for one BRC repeat, BRC4,
fused to the Rad51 nucleotide-binding core domain (26). In this
structure, the fused BRC4 interacts with Rad51 by assuming the
same structure as the Rad51 oligomerization motif, thereby pre-
venting the incorporation of Rad51 into nucleoprotein filaments.
Because BRC4RPA is proficient at HDR correction, as are other
BRC repeats fused to RPA, the inhibitory role that BRC4 has been
proposed to have is likely to be a function of its isolation from other
parts of BRCA2, namely the DNA-binding domain.
HDR is restored to nearly wild-type levels in the Brca2 mutant
cells by stable expre ssion of the BRC-RPA proteins. The efficient
correction of the HDR defect in the se cells raises the possibility that
these fusion proteins mimic BRCA2, by delivering Rad51 to ssDNA
coated with endogenous RPA. Alternatively, the BRC-RPA pro-
teins may simply bypass the cellular requirement for BRCA2,
loading Rad51 directly onto ssDNA. A bypass mechanism would
not be surprising given the distinctions between RPA and BRCA2.
For example, RPA contains six OB folds within its three subunits
to bind ssDNA with high affinity (15), whereas BRCA2 contains
only three OB folds in its single subunit (16). Moreover, an unusual
nonglobular tower domain interrupts one of the BRCA2 OB folds
(16) and may act to promote BRCA2 binding to dsDNA–ssDNA
junctions (39). Consistent with a bypass mechanism, we found that
the structurally distinct ssDNA-binding domain from Rad52 (40,
41) can function in HDR correction in the Brca2 mutant cells.
Beside s the DNA-binding domain, other parts of the BRCA2
protein may provide more complex levels of control of BRCA2
function, including the recently identified cell cycle-regulated phos-
phorylation site that modulates Rad51 binding to the C terminus of
BRCA2 (17). Our protein fusions are likely to provide an important
approach for delineating roles for these other BRCA2 domains
Materials and Methods
Cell Lines and Plasmids. The DR-GFP reporter (24, 42) was inte-
grated into V-C8 hamster cells (23). Several clones with a single
copy of DR-GFP were identified by Southern blotting and found to
give similar re sults in HDR assays. One clone was used in the
experiments presented here. To create stable cell lines, 5
pMC1neo was cotransfected with 30
g of the BRC-RPA expres-
sion vector, and cells were selected in 1.5 mg兾ml G418 (GIBCO) for
2–3 weeks. Clone s were screened by We stern blotting for expression
of the BRC-RPA proteins, and one was used for subsequent
manipulations. CAPAN-1 cells with DR-GFP reporter were de-
scribed in ref. 8.
Expre ssion vectors for human BRC3 and BRC3⌬ (21) were
modified to contain a nuclear localization signal and FLAG tag.
BRC4, BRC1-2, and BRC1-4 vectors were generated similarly by
using cDNAs from human (BRC4, amino acids 1511–1578) and
mouse (BRC1-2, amino acids 923-1252; BRC1-4, amino acids
923-1563). Human RPA70 sequences were PCR amplified from
p11d-tRPA (43) and cloned C-terminal of the BRC peptide in the
respective expression vectors. The integrity of the RPA70-coding
sequence was verified by DNA sequencing. RPA70 mutations (33)
were introduced into the BRC-RPA vectors by swapping an MfeI兾
MscI restriction fragment containing the mutations. The RPA70
R234A兾R263A mutant (33) also carries a third mutation (N239K),
which is inferred to be a silent mutation for DNA binding (M. Wold,
personal communication). The Brh2 expression plasmid was con-
structed by cloning the Brh2 BamHI fragment, encoding amino
acids 89–955, into pCAGGS. To create Brh2-Rad52, Brh2 amino
acids 89–551 were fused to U. maydis Rad52 amino acids 79–314,
which is inferred by sequence conservation to contain the ssDNA-
binding domain (M. Kojic and W. K. Holloman, personal
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0600298103 Saeki et al.
DSB Repair Assays. To measure the repair of an I-SceI-generated
DSB, 40 or 50
g of the I-SceI expression vector pCBASce (44) or
the empty pCAGGS vector was mixed with 4–5 ⫻ 10
suspended in 650
l of opti-MEM medium (Invitrogen) in a 0.4-cm
cuvette, followed by pulsing the cells at 280 V, 1,000
F for hamster
cell lines or 250 V, 950
F for the CAPAN-1 cell line. For transient
expre ssion, 40 or 50
g of the BRC peptide s or BRC-RPA fusion
proteins expression vectors were additionally added. To specifically
determine the amount of HDR, the percentage of GFP-positive
cells was quantified by flow cytometric analysis 2 days after elec-
troporation on a Becton Dickinson FACScan (24). To assay SSA,
we used a combined PCR-Southern blotting method (ref. 29; see
Fig. 5 legend).
Protein Manipulations. Whole-cell extracts were prepared 24 h after
transfection, and total protein was determined by Bio-Rad assays.
For Western blot analysis of the BRC peptides, 30–50
g of lysate
was separated by 4–12% Bis-Tris PAGE (Invitrogen). For BRC-
RPA fusion proteins, 100
g of lysate was separated by Tris-glycine
10% PAGE. Immunoprecipitations were performed with 1.2 mg
whole-cell extracts and 35
l of anti-FLAG-M2 agarose affinity gel
(A-2220; Sigma) for4hat4°C.Immune complexes were washed
with lysis buffer and separated by 4–12% Bis-Tris PAGE for Rad51,
RPA32, and RPA14 and by 8% Tris-glycine 10% PAGE for
RPA70. Purified RPA was a gift from Jerry Hurwitz (Memorial
Sloan–Kettering Cancer Center). Membranes were probed with
anti-FLAG-M2 antibody (A-8592; Sigma), anti-Rad51 antibody
(ab-1; Calbiochem), anti-RPA70 antibody (ab12320; Abcam, Inc.,
Cambridge, MA), anti-RPA32 antibody (ab2626; Abcam, Inc.), and
anti-RPA14 antibody (Tom Kelly, Memorial Sloan–Kettering Can-
cer Center). Anti-
-actin antibody (ab8227; Abcam, Inc.) was used
to asse ss protein loading.
MMC and IR Sensitivity Assays. MMC sensitivity was assayed by
seeding hamster cells into 12-well plates at 6 ⫻ 10
cells per well.
After attaching overnight, cells were incubated with media con-
taining MMC at various concentrations for 24 h. After incubation
for another 4 days, monolayers were washed once with PBS and
fixed for 5 min at room temperature in 10% methanol and 10%
acetic acid. Adherent colonies were stained for 5 min at room
temperature with 1% crystal violet (Sigma) in methanol. Plate s
were rinsed in water, and the adsorbed dye was resolubilized with
methanol containing 0.1% SDS by gentle agitation for 30 min at
room temperature. Dye solution was transferred to 96-well plates
and diluted (1:2) in methanol. OD at 595 nm was measured
photometrically in a model 3550 microplate reader (Bio-Rad). For
quantification, the ODs of each well were normalized to those
obtained from untreated cells (100% cell survival) and a well
without any cells (0%).
For IR sensitiv ity assays, hamster cells were seeded in 10-cm
plates and ex posed to various doses of IR by varying the duration
of exposure to a
Cs source. After 10 days, the clonogenic
survival was determined for a given concentration of cells that
were plated by dividing the number of colonies on each treated
plate by the number of colonies on each untreated plate.
Analysis of Chromosomal Aberrations. Frequencies of spontaneous
chromosomal aberrations were determined in exponentially grow-
ing cell cultures. Cells were harvested by trypsinization after2hof
incubation with 1
g兾ml Colcemid and fixed after treatment with
hypotonic solution (0.6% sodium citrate) in ethanol-glacial acetic
acid (3:1). Air-dried preparations were made and stained with
We thank Milorad Kojic and Bill Holloman (Weill Medical College of
Cornell University, New York), Marc Wold (University of Iowa, Iowa
City), Jerry Hurwitz, and Tom Kelly for reagents and helpful discussions.
This work was supported by the Human Frontier Science Program (to
N.C.), an Emerald Foundation grant and National Institutes of Health
Grant R01 GM54668 (to M.J.), and National Cancer Institute Grant P01
CA94060 (to L. Norton, Memorial Sloan–Kettering Cancer Center).
1. Wooster, R. & Weber, B. L. (2003) N. Engl. J. Med. 348, 2339–2347.
2. Reid, S., Renwick, A., Seal, S., Baskcomb, L., Barfoot, R., Jayatilake, H.,
Pritchard-Jones, K., Stratton, M. R., Ridolfi-Luthy, A. & Rahman, N. (2005)
J. Med . Genet. 42, 147–151.
3. Offit, K., Levran, O., Mullaney, B., Mah, K., Nafa, K., Batish, S. D., Diotti, R.,
Schneider, H., Deffenbaugh, A., Scholl, T., et al. (2003) J. Natl. Cancer Inst . 95,
4. Wagner, J. E., Tolar, J., Levran, O., Scholl, T., Deffenbaugh, A., Satagopan, J.,
Ben-Porat, L., Mah, K., Batish, S. D., Kutler, D. I., et al. (2004) Blood 103, 3226–3229.
5. Moynahan, M. E. (2002) Oncogene 21, 8994–9007.
6. Sharan, S. K., Pyle, A., Coppola, V., Babus, J., Swaminathan, S., Benedict, J., Swing,
D., Martin, B. K., Tessarollo, L., Evans, J. P., et al. (2004) Development (Cambridge,
U.K.) 131, 131–142.
7. Jasin, M. (2002) Oncogene 21, 8981–8993.
8. Moynahan, M. E., Pierce, A. J. & Jasin, M. (2001) Mol. Cell 7, 263–272.
9. Tutt, A., Bertwistle, D., Valentine, J., Gabriel, A., Swift, S., Ross, G., Griffin, C.,
Thacker, J. & Ashworth, A. (2001) EMBO J. 20, 4704–4716.
10. Wong, A. K. C., Pero, R., Or monde, P. A., Tavtigian, S. V. & Bartel, P. L. (1997)
J. Biol. Chem. 272, 31941–31944.
11. Sharan, S. K., Morimatsu, M., Albrecht, U., Lim, D.-S., Regel, E., Dinh, C., Sands,
A., Eichele, G., Hasty, P. & Bradley, A. (1997) Nature 386, 804–810.
12. Chen, P.-L., Chen, C.-F., Chen, Y., Xiao, J., Sharp, Z. D. & Lee, W.-H. (1998) P roc.
Natl. Acad. Sci. USA 95, 5287–5292.
13. Yuan, S. S., Lee, S. Y., Chen, G., Song, M., Tomlinson, G. E. & Lee, E. Y. (1999)
Cancer Res. 59, 3547–3551.
14. Stark, J. M., Pierce, A. J., Oh, J., Pastink, A. & Jasin, M. (2004) Mol. Cell. Biol. 24,
15. Bochkarev, A. & Bochk areva, E. (2004) Curr. Opin. Struct. Biol. 14, 36–42.
16. Yang, H., Jeffrey, P. D., Miller, J., Kinnucan, E., Sun, Y., Thoma, N. H., Zheng, N.,
Chen, P. L., Lee, W. H. & Pavletich, N. P. (2002) Science 297, 1837–1848.
17. Esashi, F., Christ, N., Gannon, J., Liu, Y., Hunt, T., Jasin, M. & West, S. C. (2005)
Nature 434, 598 –604.
18. Kojic, M., Kostr ub, C. F., Buchman, A. R. & Holloman, W. K. (2002) Mol. Cell 10,
19. Martin, J. S., Winkelmann, N., Petalcorin, M. I., McIlwraith, M. J. & Boulton, S. J.
(2005) Mol. Cell. Biol. 25, 3127–3139.
20. Siaud, N., Dray, E., Gy, I., Gerard, E., Takvorian, N. & Doutriaux, M. P. (2004)
EMBO J. 23, 1392–1401.
21. Stark, J. M., Hu, P., Pierce, A. J., Moynahan, M. E., Ellis, N. & Jasin, M. (2002) J. Biol .
Chem. 277, 20185–20194.
22. Davies, A. A., Masson, J. Y., McIlwraith, M. J., Stasiak, A. Z., Stasiak, A.,
Venkitaraman, A. R. & West, S. C. (2001) Mol. Cell 7, 273–282.
23. Kraakman-van der Zwet, M., Overkamp, W. J., van Lange, R. E., Essers, J., van
Duijn-Goedhart, A., Wiggers, I., Swaminathan, S., van Buul, P. P., Errami, A., Tan,
R. T., et al. (2002) Mol. Cell. Biol. 22, 669–679.
24. Pierce, A. J., Johnson, R. D., Thompson, L. H. & Jasin, M. (1999) Genes Dev. 13,
25. Galkin, V. E., Esashi, F., Yu, X., Yang, S., West, S. C. & Egelman, E. H. (2005) Proc.
Natl. Acad. Sci. USA 102, 8537–8542.
26. Pellegrini, L., Yu, D. S., Lo, T., A nand, S., Lee, M., Blundell, T. L. & Venkit araman,
A. R. (2002) Nature 420, 287–293.
27. Spain, B. H., Larson, C. J., Shihabuddin, L. S., Gage, F. H. & Verma, I. M. (1999)
Proc. Natl. Acad. Sci. USA 96, 13920–13925.
28. Paˆques, F. & Haber, J. E. (1999) Microbiol. Mol. Biol. Rev. 63, 349– 404.
29. Nakanishi, K., Yang, Y.-G., Pierce, A. J., Taniguchi, T., Digweed, M., D’Andrea,
A. D., Wang, Z.-Q. & Jasin, M. (2005) Proc. Natl. Acad. Sci. USA 102, 1110–1115.
30. Haaf, T., Golub, E. I., Reddy, G., Radding, C. M. & Ward, D. C. (1995) Proc. Natl.
Acad. Sci. USA 92, 2298–2302.
31. Yu, D. S., Sonoda, E., Takeda, S., Huang, C. L., Pellegrini, L., Blundell, T. L. &
Venkitaraman, A. R. (2003) Mol. Cell 12, 1029–1041.
32. Wold, M. S. (1997) Annu. Rev. Biochem. 66, 61–92.
33. Wyka, I. M., Dhar, K., Binz, S. K. & Wold, M. S. (2003) Biochemistry 42, 12909–12918.
34. Venkitaraman, A. R. (2002) Cell 108, 171–182.
35. Sung, P., Krejci, L., Van Komen, S. & Sehorn, M. G. (2003) J. Biol. Chem. 278,
36. Warren, M., Smith, A., Partridge, N., Masabanda, J., Griffin, D. & Ashworth, A.
(2002) Hum. Mol. Genet. 11, 841–851.
37. Filippo, J. S., Chi, P., Sehorn, M. G., Etchin, J., Krejci, L. & Sung, P. (2006) J. Biol.
Chem. 281, 11649 –11657.
38. Kojic, M., Zhou, Q., Lisby, M. & Holloman, W. K. (2005) Mol . Cell. Biol. 25,
39. Yang, H., Li, Q., Fan, J., Holloman, W. K. & Pavletich, N. P. (2005) Nature 433,
40. Kagawa, W., Kurumizaka, H., Ishitani, R., Fuk ai, S., Nureki, O., Shibat a, T. &
Yokoyama, S. (2002) Mol. Cell 10, 359–371.
41. Singleton, M. R., Wentzell, L. M., Liu, Y., West, S. C. & Wigley, D. B. (2002) Proc.
Natl. Acad. Sci. USA 99, 13492–13497.
42. Pierce, A. J., Hu, P., Han, M., Ellis, N. & Jasin, M. (2001) Genes Dev. 15, 3237–3242.
43. Henricksen, L. A., Umbricht, C. B. & Wold, M. S. (1994) J. Biol. Chem. 269,
44. Richardson, C., Moynahan, M. E. & Jasin, M. (1998) Genes Dev. 12, 3831–3842.
Saeki et al. PNAS
June 6, 2006