DNA mediated disassembly of hRad51 and hRad52
proteins and recruitment of hRad51 to ssDNA by hRad52
Vasundhara M. Navadgi, Ashish Shukla, Rahul Kumar Vempati and Basuthkar J. Rao
Department of Biological Sciences, Tata Institute of Fundamental Research, Mumbai, India
Human Rad51 protein (hRad51), a homologue of
Escherichia coli RecA performs the fundamental role
of homologous pairing and strand exchange during
homologous recombination and double-strand break
repair [1,2]. Rad51 and Rad52 colocalize in distinct
nuclear foci in response to DNA damage [3]. Yeast
rad52 mutants show extensive degradation of the
DNA double-strand break ends suggesting that Rad52
is critically involved in stable maintenance of chromo-
somal integrity [4]. Cytological studies indicate that
Rad52 is required for Rad51 foci formation during
meiosis [5] and chromatin immunoprecipitation assays
demonstrated the requirement of Rad52 for association
of yeast Rad51 to HO induced double strand break
site at the MATa locus in vivo [6,7]. Biochemically,
Rad52 stimulates the strand exchange activity of
Rad51 [8–10] and is shown to displace Replication
Protein A (RPA) [11,12] and stabilize the Rad51-
single-stranded (ssDNA) filament [13].
hRad51 and hRad52 form ring-shaped structures
like other recombination proteins RecA, RecT, human
Dmc1 and b protein from bacteriophage k [14–18].
Rad51 and RecA bind DNA as helical filaments
whereas their meiosis specific homologue Dmc1 and
archaeal recombinase, RadA proteins, form stacked
octameric rings on DNA in the absence of ATP and as
helical filaments in the presence of ATP [18–20]. The

states, in vitro. Single-stranded DNA (ssDNA) renders high molecular
weight aggregates of both proteins into smaller and soluble forms that
include even the monomers. Consequently, these proteins that have a pro-
pensity to interact with each other’s higher order forms by themselves, start
interacting with monomeric forms in the presence of ssDNA, presumably
reflecting the steps of protein assembly on DNA. In the same conditions,
DNA binding assays reveal hRad52-mediated recruitment of hRad51 on
ssDNA. Put together, these studies hint at DNA-induced disassembly of
higher-order forms of Rad51 and Rad52 proteins as steps that precede
protein assembly during hRad51 presynapsis on DNA, in vitro.
Abbreviations
ATPcS, Adenosine 5¢-O-(3-thiotriphosphate); hRad51, human Rad51; hRad52, human Rad52; ssDNA, single-stranded DNA.
FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS 199
shorter and thicker nucleoprotein filaments when
ssDNA is added [28]. Tumour suppressors BRCA2
and p53 interact with the oligomerization domain of
Rad51 [24,29,30], and RPA is shown to inhibit the
higher order self association of hRad52 rings [31] sug-
gesting that oligomerization of Rad51 and Rad52 is
regulated by other molecules to control their activity.
In this work, we have studied the oligomeric states
of hRad51 and hRad52 in the presence and absence of
ssDNA. Our studies indicate that human Rad51, like
its bacterial homologue RecA exists in multiple aggre-
gation states [14,32]. DNA seems to dissociate the
higher order structures of both hRad51 and hRad52.
hRad52 interacts specifically with higher oligomeric
states of hRad51 in the absence of DNA, but with
hRad51 monomers when ssDNA is present. These
results are rationalized through a model where we pro-

were migrating close to each other, there was enough
difference between the two to discern an increase in
the monomer level. Using 5¢
32
P-labelled ssDNA, we
mapped the positions of protein–DNA complexes in
this gel system (data not shown; compare Fig. 4).
Based on this comparison, the protein–DNA com-
plexes that entered into the gel mapped to the posi-
tion indicated by the asterisk in Fig. 1A. The
ssDNA mediated increase in monomer level remained
essentially unchanged in the presence of nucleotide
cofactors ADP, ATP or ATPcS (compare lanes 6–8,
10–12, 14–16 with of 2–4, respectively, Fig. 1A).
However, the signals associated with protein–DNA
complexes (position indicated by asterisk) appear to
diminish and that of higher oligomeric forms that
enter into the gel (as labelled in Fig. 1A) appear to
increase in sets containing nucleotide cofactors. This
trend is consistent with ATP induced effects reported
earlier, where much larger forms of hRad51 are dis-
aggregated into oligomeric complexes equivalent to
3–8 protein monomers [33].
In order to trap the oligomeric forms that do not
enter the gel, we used centrifugation assays. Following
the assay, we recovered a fraction of the protein in the
pellet (lane 1, Fig. 1B) and the remainder in the super-
natant (lane 5). In the presence of ssDNA, the protein
fraction that was pelletable became fully soluble, as no
signal was recovered in the pellet (lanes 2–4). This

from 30 to 80 nm sizes (grouped as b, c and d in
Fig. 1C). Upon ssDNA addition, the distribution shif-
ted towards smaller R
h
values (10–20 nm size, grouped
as a) with a concomitant drop in the levels of larger
ones (grouped as c and d). As a result, at the highest
concentration of DNA, the distribution revealed a high
preponderance of smaller protein particles and reduced
Disassembly and recruitment of hRad51 and hRad52 proteins V. M. Navadgi et al.
200 FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS
level of larger ones, thereby corroborating the effect of
ssDNA induced disaggregation of hRad51.
DNA/ATP induced changes in hRad51
We tested whether DNA induced disaggregation of
protein leads to changes in the pattern of limited pro-
teolysis of hRad51 by trypsin, where the extent of pro-
tease attacks is a function of both conformational as
well as overall organizational changes in the target
protein system. Comparison revealed that hRad51 is
relatively more protected in the presence of ssDNA
than in its absence (compare lanes 4 and 5 with lanes
2 and 3, respectively, Fig. 1D). This effect might
arise either due to steric hindrance imparted by ssDNA
binding or to changes in protein configuration ⁄
conformation following ssDNA binding or a combina-
tion of both. ATP has been shown to induce changes
in hRad51 such that accessibility to protease attacks is
altered [33]. We observed that ATP induced change
was somewhat different from that induced by ssDNA

nm) vs. intensity is plotted as histograms where groupings a–d depicts 10–80 nm range distribution. (D) Partial proteolysis experiment to
analyse DNA induced conformational changes on hRad51 protein. hRad51 (25 l
M) was incubated in binding buffer in the absence (lanes 2
and 3) or presence of 75 l
M ssDNA (lanes 4 and 5) and 1 mM ATP (lanes 3 and 5) for 1 h at 37 °C and then subjected to partial digestion
with trypsin (Sigma, Munich, Germany) (50 lgÆmL
)1
) for 1 min. The reaction was quenched by Laemmli buffer and analysed by SDS ⁄ PAGE
(20% acrylamide), followed by silver staining. Lane 6 consists of only ssDNA.
V. M. Navadgi et al. Disassembly and recruitment of hRad51 and hRad52 proteins
FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS 201
DNA-induced disassembly of higher oligomeric
forms of hRad52
A similar effect of disaggregation by ssDNA was also
observed with hRad52 protein. In the absence of
ssDNA, hRad52 protein appeared to be so highly
aggregated that it hardly entered into the gel (lane 5,
Fig. 2A) and was mostly in the pellet fraction following
a centrifugation assay (lane 1, Fig. 2B). Aggregation of
hRad52 appears to be highly salt sensitive and in the
present assay conditions with 20 mm KCl, the protein
remains highly aggregated. The addition of ssDNA ren-
dered the protein into forms that not only entered into
the gel, but also migrated as the monomeric form. This
effect was further evidenced by the complete recovery
of the protein in the supernatant fraction, as a function
of added ssDNA (lanes 7 and 8, Fig. 2B).
Protein aggregation/disaggregation changes vs.
ionic effects
In order to assess whether ssDNA induced diaggrega-

titration rendered hRad52 highly sol-
uble, whereas hRad51 was highly insoluble (Fig. 3B). A
significant fraction of pelletable hRad52 was recovered
in the supernatant fraction following Mg
2+
treatment,
indicating that the protein was subject to solublization
not only by ssDNA (Fig. 2A and B), but also by Mg
2+
(Fig. 3B), a common effect facilitated by oppositely
charged ionic species. On the other hand, hRad51 exhib-
ited a behaviour opposite to that of hRad52 by under-
going high level of aggregation, which is akin to that
of E. coli RecA aggregation induced by Mg
2+
observed
earlier [32]. These studies indicate that hRad52 ⁄ hRad51
disaggregation ⁄ aggregation properties assayed here
reflect genuine modulations rendered by ssDNA ⁄
ATP ⁄ Mg
2+
, etc. rather than nonspecific ionic effects in
solution conditions.
hRad52 protein selectively interacts with higher
oligomeric forms of hRad51 in the absence of
DNA
As seen in earlier experiment, hRad51 exhibited
multiple oligomeric forms (lane 1, Fig. 4A) whereas
hRad52 was in an aggregated state and hardly entered
into the gel (lane 5, Fig. 4A). hRad51 (10 lm) was

also form large complexes that elute much earlier than
individual proteins in gel filtration chromatography
experiment [34].
Recruitment of hRad51 to ssDNA targets: the role
of hRad52
We addressed this issue by analysing the status of sol-
uble forms of hRad51 protein as a function of increas-
ing hRad52 protein in the presence of ssDNA. As
expected, native gel analyses revealed DNA (25 lm)
induced ‘monomerization’ of hRad51 protein (10 lm)
(compare lanes 2 and 7 with lanes 1 and 6, respect-
ively, Fig. 4B). In this native gel assay conditions, the
‘monomerized’ form of hRad52 essentially comigrates
with that of hRad51 monomer (compare lanes 2 and 7
with lane 11, Fig. 4B). Interestingly, addition of
hRad52 protein led to a measurable depletion rather
than a cumulative increase in the monomer signal of
both proteins (compare lanes 5 and 10 with 2–4 and
7–9, respectively, Fig. 4B). This was concomitantly
associated with the rise of a signal at high molecular
weight region in the gel (at asterisk position in lanes
5 and 10). This was observed both with and without
ATP.
In parallel, we studied protein binding to 5¢
32
P-
labelled ssDNA of 121-mer (used in the previous
experiments) and analysed the complex formation by
native gel electrophoresis. Increasing concentration of
hRad51 led to the generation of protein–DNA com-

incubated in buffer (30 m
M Tris ⁄ HCl pH 7.5, 20 mM KCl, 1 mM DTT) containing varying concentrations of MgCl
2
for 30 min at 37 °C, fol-
lowed by centrifugation and analyses of pellet ⁄ supernatant fractions by SDS PAGE.
V. M. Navadgi et al. Disassembly and recruitment of hRad51 and hRad52 proteins
FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS 203
presence of hRad52 renders much better binding of
hRad51 to ssDNA even at lower concentrations of the
latter, thereby implying that hRad52 plays a role in
the recruitment of hRad51 to ssDNA.
The results described in this study help us to under-
stand the transitions associated with the oligomeric
states of hRad51 and hRad52 proteins in the presence
of ssDNA and relate them to their DNA binding
activity. The observation that hRad51 protein in its
DNA-unbound form, exits in higher oligomeric forms,
poses a mechanistic challenge as to how such struc-
tures transform into right-handed helical filaments
during ⁄ following DNA binding. Whether higher oligo-
meric states of protein are directly recruited to DNA
or much smaller forms of the protein are generated
prior to active assembly, is an open question. Our
results suggest that transient contacts of DNA strands
with either protein create an effect of ‘protein disaggre-
gation’. It is important to note that all the effects
uncovered in the present study are from in vitro analy-
ses and it is not clear how these effects may relate to
the situations in vivo, the mechanistic description of
which is not very clear at present.

with hRad51 monomers in the presence of DNA. hRad51 (10 l
M)
was incubated with 0, 2, 4 and 10 l
M (lanes 2–5 and 7–10) of
hRad52 in the presence of 25 l
M oligo PUC+ in binding buffer con-
taining 50 m
M KCl either in the absence (lane 1–5) or presence of
1m
M ATP (lane 6–10) and analysed by native PAGE (6% acryl-
amide) followed by silver staining. Lane 1 and 6 has hRad51
(10 l
M) without DNA and lane 11 had hRad52 (10 lM) with DNA.
The position of asterisk indicates the formation of large complex in
the presence of hRad51, hRad52 and ssDNA.
Fig. 5. hRad51 binding to ssDNA in the presence of hRad52.
32
P
labelled oligo PUC+ (1 l
M) was incubated with 0, 1, 2, 3, 6 lM
hRad51 (lanes 1–5) in the absence of hRad52 and 1 lM hRad51
with 0, 0.25 and 0.50 l
M hRad52 (lanes 6–9). Lanes 10 and 11 con-
tain DNA samples incubated with hRad52 alone. Samples were
resolved by native PAGE (6% acrylamide) and the gel was scanned
using a PhosphorImager.
Disassembly and recruitment of hRad51 and hRad52 proteins V. M. Navadgi et al.
204 FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS
provided some useful clues on important organiza-
tional changes that ensue in protein during their

cribed in the figure legends) were incubated in binding
buffer [30 mm Tris ⁄ HCl pH 7.5, 1 mm MgCl
2
,1mm di-
thiothreitol (DTT)] at 37 °C for 1 h. Samples were subjec-
ted to centrifugation at 14 000 r.p.m. for 10 min. The
supernatant and pellet were separated and heated in
Laemmli buffer at 90 °C for 10 min and analysed by
SDS ⁄ PAGE (10% acrylamide), followed by silver staining.
Native polyacrylamide gel electrophoresis
of proteins
Varying concentrations of Rad51, Rad52 and DNA were
incubated in specific conditions (as described in the figure
legends) at 37 °C for 1 h. Samples were subjected to native
PAGE (6% acrylamide) in TBE buffer at 200 V for 3 h at
room temperature (25 °C). Subsequently the proteins were
visualized by silver staining.
DNA binding by gel-shift assays
Labelled DNA substrate was incubated with various con-
centrations of hRad51 and hRad52 (as described in the fig-
ure legends) in a binding buffer (30 mm Tris ⁄ HCl pH 7.5,
1mm MgCl
2
,1mm DTT, 100 lgÆmL
)1
BSA) at 37 °C for
1 h. DNA–protein complexes were analysed by native
PAGE (6% acrylamide) in TBE buffer at 200 V for 3 h at
room temperature (25 °C). The radioactivity in the gels
was quantified by ImageQuant software on a Phosphor-

References
1 West SC (2003) Molecular views of recombination pro-
teins and their control. Nat Rev Mol Cell Biol 4, 435–445.
2 Sung P, Krejci L, Van Komen S & Sehorn MG (2003)
Rad51 recombinase and recombination mediators.
J Biol Chem 278, 42729–42732.
3 Liu Y & Maizels N (2000) Co-ordinated response of
mammalian Rad51 and Rad52 toDNA damage. EMBO
Rep 1, 85–90.
V. M. Navadgi et al. Disassembly and recruitment of hRad51 and hRad52 proteins
FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS 205
4 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.
5 Gasior SL, Olivares H, Ear U, Hari DM, Weichselbaum
R & Bishop DK (2001) Assembly of RecA-like recombi-
nases: distinct roles for mediator proteins in mitosis and
meiosis. Proc Natl Acad Sci USA 98, 8411–8418.
6 Sugawara N & Wang X, Haber JE (2003) In vivo roles
of Rad52, Rad54, and Rad55 proteins in Rad51-
mediated recombination. Mol Cell 12, 209–219.
7 Wolner B, Van Komen S, Sung P & Peterson CL (2003)
Recruitment of the recombinational repair machinery to
a DNA double-strand break in yeast. Mol Cell 12,
221–232.
8 Benson FE, Baumann P & West SC (1998) Synergistic
actions of Rad51 andRad52 in recombination and
DNA repair. Nature 391, 401–404.
9 Shinohara A, & Ogawa T (1998) Stimulation by Rad52

17 Passy SI., YuX, Li Z, Radding CM & Egelman EH
(1999) Rings and filaments of B protein from bacterio-
phage 1 suggest a superfamily of recombination pro-
teins. Proc Natl Acad Sci USA 96, 4279–4284.
18 Kinebuchi T, Kagawa W, Enomoto R, Tanaka K,
Miyagawa K, Shibata T, Kurumizaka H & Yokoyama
S (2004) Structural basis for octameric ring formation
and DNA interaction of the human homologous-pairing
protein Dmc1. Mol Cell 14, 363–374.
19 Sehorn MG, Sigurdsson S, Bussen W, Unger VM &
Sung P (2004) Human meiotic recombinase Dmc1 pro-
motes ATP-dependent homologous DNA strand
exchange. Nature 429, 433–437.
20 Yang S, Yu X, Seitz EM, Kowalczykowski SC & Egel-
man EH (2001) Archaeal RadA protein binds DNA as
both helical filaments and octameric rings. J Mol Biol
314, 1077–1085.
21 Shin DS, Pellegrini L, Daniels DS, Yelent B, Craig L
& Bates D., YuDS, Shivji MK, Hitomi C, Arvai AS,
Volkmann N, Tsuruta H, Blundell TL, Venkitaraman
AR & Tainer JA (2003) Full length archaeal Rad51
structure and mutants: mechanisms for Rad51 assem-
bly and control by BRCA2. EMBO J 22,.4566–
4576.
22 Kagawa W, Kurumizaka H, Ishitani R, Fukai S, Nureki
O & Shibata T & Yokoyama S (2002) Crystal structure
of the homologous-pairing domain from the human
Rad52 recombinase in the undecameric form. Mol Cell
10, 359–371.
23 Singleton MR, Wentzell LM, Liu Y, West SC & Wigley

3868–3874.
Disassembly and recruitment of hRad51 and hRad52 proteins V. M. Navadgi et al.
206 FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS
31 Jackson D, Dhar K, Wahl JK, Wold MS & Borgstahl
GE (2002) Analysis of thehuman replication protein A:
Rad52 complex: evidence for crosstalk between
RPA32,RPA70, Rad52 and DNA. J Mol Biol 321,
133–148.
32 Brenner SL, Zlotnick A & Griffith JD (1988) RecA pro-
tein self-assembly. Multiple discrete aggregation states.
J Mol Biol 204, 959–972.
33 Tombline G, Heinen CD, Shim KS & Fishel R (2002)
Biochemical characterization of the human Rad51 pro-
tein. Modulation of DNA binding by adenosine nucleo-
tides. J Biol Chem 277, 14434–14442.
34 Song B & Sung P (2000) Functional Interactions among
yeast Rad51 recombinase, Rad52 mediator, and replica-
tion protein A in DNA strand exchange. J Biol Chem
275, 15895–15904.
35 Navadgi VM, Dutta A & Rao BJ (2003) Human Rad52
facilitates three-stranded pairing that follows no strand
exchange: A novel pairing function of the protein.
Biochemistry 42, 15237–15251.
36 Kurumizaka H, Aihara H, Kagawa W, Shibata T &
Yokoyama S (1999) Human Rad51 amino acid residues
required for Rad52 binding. J Mol Biol 291, 537–548.
V. M. Navadgi et al. Disassembly and recruitment of hRad51 and hRad52 proteins
FEBS Journal 273 (2006) 199–207 ª 2005 The Authors Journal compilation ª 2005 FEBS 207