Gel retarda-tion and DNA-dependent ATP hydrolysis measurements revealed that the substitution of the tyrosine residue at position 232 Tyr232 within the L1 loop with alanine, a short side
Trang 1Yusuke Matsuo1, Isao Sakane2, Yoshimasa Takizawa1, Masayuki Takahashi3
and Hitoshi Kurumizaka1,2
1 Graduate School of Science and Engineering, Waseda University, Tokyo, Japan
2 Institute for Biochemical Engineering, Waseda University, Tokyo, Japan
3 UMR 6204 Biocatalyse-Biotechnologie-Bioregulation, Centre National de la Recherche Scientifique, and University of Nantes, France
The Rad51 proteins are the eukaryotic orthologs of
the bacterial RecA protein [1], which promotes key
steps in homologous recombination [2–5] A RAD51
null mutation causes severe defects in meiotic
homol-ogous recombination and mitotic recombinational
repair of double strand breaks (DSBs) in
Saccharomy-ces cerevisiae [1] Rad51 is thus required for both the
meiotic and mitotic homologous recombination
pro-cesses, while another ortholog, Dmc1, is specific to
meiotic homologous recombination [6–8] In higher
eukaryotes, Rad51 is even essential for cell survival:
disruption of the RAD51 gene in mice results in early embryonic lethality [9,10] and the RAD51 gene knock-out in chicken DT40 cells causes cell death, with the accumulation of spontaneous chromosomal breaks [11]
Rad51 and RecA apparently use similar mechanisms
to promote homologous recombination [12–15] Dur-ing the homologous recombination process, Rad51 is thought to bind single-stranded tails produced at DSB sites, and to form a helical nucleoprotein filament The single-stranded DNA (ssDNA) and double-stranded
Keywords
DNA binding; DNA repair; Rad51; Rad51
mutant; recombination
Correspondence
H Kurumizaka, Graduate School of Science
and Engineering, Waseda University, 3-4-1
Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
Fax: +81 3 5292 9211
Tel: +81 3 5286 8189
E-mail: kurumizaka@waseda.jp
(Received 12 November 2005, revised
3 April 2006, accepted 16 May 2006)
doi:10.1111/j.1742-4658.2006.05323.x
The human Rad51 protein, a eukaryotic ortholog of the bacterial RecA protein, is a key enzyme that functions in homologous recombination and recombinational repair of double strand breaks The Rad51 protein con-tains two flexible loops, L1 and L2, which are proposed to be sites for DNA binding, based on a structural comparison with RecA In the present study, we performed mutational and fluorescent spectroscopic analyses on the L1 and L2 loops to examine their role in DNA binding Gel retarda-tion and DNA-dependent ATP hydrolysis measurements revealed that the substitution of the tyrosine residue at position 232 (Tyr232) within the L1 loop with alanine, a short side chain amino acid, significantly decreased the DNA-binding ability of human Rad51, without affecting the protein fold-ing or the salt-induced, DNA-independent ATP hydrolysis Even the conservative replacement with tryptophan affected the DNA binding, indicating that Tyr232 is involved in DNA binding The importance of the L1 loop was confirmed by the fluorescence change of a tryptophan residue, replacing the Asp231, Ser233, or Gly236 residue, upon DNA binding The alanine replacement of phenylalanine at position 279 (Phe279) within the L2 loop did not affect the DNA-binding ability of human Rad51, unlike the Phe203 mutation of the RecA L2 loop The Phe279 side chain may not
be directly involved in the interaction with DNA However, the fluores-cence intensity of the tryptophan replacing the Rad51-Phe279 residue was strongly reduced upon DNA binding, indicating that the L2 loop is also close to the DNA-binding site
Abbreviations
DSB, double strand break; dsDNA, double-stranded DNA; HsRad51, Homo sapiens Rad51; RPA, replication protein A; ScRad51,
Saccharomyces cerevisiae Rad51; ssDNA, single-stranded DNA; SSB, single stranded DNA-binding protein.
Trang 2DNA (dsDNA) molecules bind within the Rad51
nucleoprotein filament along the helical axis, thus
forming the ternary complex containing ssDNA,
dsDNA, and Rad51 In the ternary complex, the
homologous sequence between ssDNA and dsDNA is
aligned, and the ssDNA forms a heteroduplex with a
complementary strand of dsDNA (homologous
pair-ing) The heteroduplex region produced by
homolog-ous pairing is then extended by the Rad51-mediated
strand exchange Therefore, Rad51 should have at
least two DNA-binding sites, as in RecA [16], and the
identification of these sites is important for
under-standing the reaction mechanism and the regulation of
homologous recombination
So far, the crystal structures of bacterial RecA,
ar-chaeal Rad51 (RadA), yeast Rad51 (ScRad51), human
Rad51 (HsRad51), and human Dmc1 (HsDmc1) have
been solved [17–22] These structural analyses revealed
that these proteins have highly conserved
three-dimen-sional structures, especially in their ATPase domains
Two flexible loops, L1 and L2, which are involved in
DNA binding by Escherichia coli RecA [17], have also
been identified in these eukaryotic and archaeal
pro-teins (Fig 1A) Like the case of bacterial RecA, the L1
and L2 loops of the eukaryotic Rad51 proteins face
inside of their helical filaments, where the DNA should
be located, and are not found at the ATP-binding site
or the subunit–subunit interface of the Rad51 filament Therefore, the Rad51 L1 and L2 loops may also be involved in DNA binding
In the present study, we performed mutational and fluorescence spectroscopic analyses on HsRad51 to examine whether these loops are actually involved in DNA binding Because aromatic residues are involved
in the ssDNA binding by bacterial single stranded DNA-binding protein (SSB) and human replication protein A (RPA) [23,24], we performed mutational analyses on Tyr232 in the L1 loop and Phe279 in the L2 loop of HsRad51 We also performed tryptophan-scanning mutagenesis across the HsRad51-L1 loop and measured the fluorescence changes of the tryptophan residues upon DNA binding
Results Strategy of mutational analysis
In order to study the functions of the L1 and L2 loops of HsRad51, we examined the effect of replacing the aromatic residues in the L1 and L2 loops with
A
B
C
Fig 1 HsRad51 and the Rad51 mutants (A) Alignment of the HsRad51 domains to those of the Methanococcus voltae RadA (MvRadA), Saccharomyces cerevisiae Rad51 (ScRad51), and Escherichia coli RecA (EcRecA) domains The N-terminal domains, the conserved ATPase domains, and the C-terminal domain are indicated by shaded boxes The L1 and L2 loops are indicated by black boxes (B) Alignment of the HsRad51 sequence to those of MvRadA, Pyrococcus furiosus Rad51 (PfRad51), and ScRad51 around the L1 and L2 loops The L1 and L2 loops, which are invisible in the crystal structure of the ATPase domain of HsRad51 [21], are represented by boxes, and the Y232 and F279 residues are indicated by shaded boxes (C) Purified HsRad51 (lane 2), Y232A mutant (lane 3), F279A mutant (lane 4), Y232W mutant (lane 5), F279W mutant (lane 6), D231W mutant (lane 7), S233W mutant (lane 8), and G236W mutant (lane 9) were analyzed by 15% (w ⁄ v) SDS ⁄ PAGE with Coomassie Brilliant Blue staining Lane 1 indicates the molecular mass markers.
Trang 3alanine If the aromatic side chain is involved in the
interaction with DNA, then its replacement with
alanine, a short side chain amino acid residue,
should affect the DNA binding of the protein The
DNA-binding ability of these alanine-substituted
Rad51 mutants was evaluated by a gel retardation
tech-nique and by measurements of the DNA-dependent
ATPase activity To ensure that the defect in DNA
binding is not due to incorrect folding of the Rad51
mutants or a loss of binding cooperativity by changing
the subunit–subunit contacts, we measured their CD
spectra and carried out gel filtration chromatography
The DNA binding by Rad51 is highly cooperative, with
strong subunit–subunit contacts, and the protein can
form a polymer even in the absence of DNA [20,25]
We also prepared another type of Rad51 mutant, in
which one of these aromatic residues was replaced by
tryptophan, a fluorescent probe If the residue is within
or close to the DNA-binding site, then we would
expect a large change in its fluorescence upon DNA
binding Using such an approach, we previously
showed that Phe203 in the L2 loop of RecA is close to
the DNA-binding site [26] Furthermore, we performed
the tryptophan-scanning mutagenesis across the
HsRad51-L1 loop, and tested the interaction between
the L1 loop and DNA
Involvement of the L1 loop-Tyr232 residue in DNA binding
The L1 loop of HsRad51 contains an aromatic residue (Tyr232) that is highly conserved among the eukaryotic and archaeal Rad51 proteins (Fig 1B) We prepared the Rad51-Y232A and Rad51-Y232W mutants, in which the Tyr232 residues were replaced by alanine (Y232A) and tryptophan (Y232W), respectively, by site directed mutagenesis, and purified them to near homo-geneity by a four-step purification method based on HsRad51 purification, including nickel-nitrilotriacetic acid (Ni-NTA) agarose column chromatography, removal of the hexahistidine tag from HsRad51 with thrombin protease, spermidine precipitation, and MonoQ column chromatography (Fig 1C)
Rad51-Y232A yielded a CD spectrum similar to that
of HsRad51, indicating that the mutation did not affect either the folding or global structure of the protein (Fig 2A,B) Gel filtration chromatography revealed that Rad51-Y232A formed polymers by self association, like HsRad51: the protein eluted in the void volume from the Superdex 200 gel filtration column (data not shown) Therefore, the mutation did not appear to affect the subunit–subunit contact in the Rad51 polymer Rad51-Y232A also exhibited
F E
D
Fig 2 Circular dichroism analysis and ATPase activities of the Rad51 mutants (A–C) CD spectra of HsRad51 (6.7 l M ) and the Rad51 mutant (6.7 l M ) were recorded at 25 C HsRad51 (A); Y232A mutant (B); and F279A mutant (C) (D–F) The ATPase activities of the Rad51 mutants Time course experiments are shown d, m, and n indicate experiments in the presence of NaCl (1.6 M ), ssDNA (20 l M ), and dsDNA (20 l M ), respectively s indicate experiments in the absence of NaCl and DNA HsRad51 (D); Y232A mutant (E); F279A mutant (F).
Trang 4salt-induced ATPase activity very similar to that of
HsRad51 in the absence of DNA (Figs 2D,E) These
results suggest that the Tyr232 residue is within neither
the subunit–subunit interface nor the ATP-binding
site
In contrast to the salt-induced ATPase activity,
Rad51-Y232A did not exhibit DNA-dependent
ATPase activity (Fig 2E) Neither ssDNA nor dsDNA
induced the ATPase activity of Rad51-Y232A, while
the ATPase activity of HsRad51 was stimulated by
ssDNA and dsDNA (Fig 2D) These results indicate
that Rad51-Y232A was defective in DNA binding
Consistent with this finding, a gel retardation
experi-ment showed that Rad51-Y232A was defective in
ssDNA and dsDNA binding (Fig 3A,B, lanes 5–8), as
compared to HsRad51 (lanes 1–4) These results
dem-onstrate the importance of the Rad51-Tyr232 residue
in DNA binding The gel retardation experiments also
revealed that even the conservative replacement of
Tyr232 with tryptophan (Rad51-Y232W) caused
signi-ficant defects in dsDNA binding, although it possessed
the ssDNA-binding ability (Fig 3A,B, lanes 9–12) As
expected from the DNA binding defect, neither
Rad51-Y232A nor Rad51-Y232W promoted the
strand-exchange reaction (Fig 4B) These results suggest
that the Rad51-Tyr232 residue in the L1 loop is involved in the functional DNA binding during strand exchange
Tryptophan-scanning mutagenesis of the HsRad51-L1 loop
To gain further information about DNA binding by the L1 loop, we performed tryptophan-scanning mutagenesis across the L1 loop (from Thr230 to Gly236) Five mutant genes corresponding to the Rad51-D231W, Rad51-S233W, Rad51-G234W, Rad51-R235W, and Rad51-G236W mutants, in which Asp231, Ser233, Gly234, Arg235, and Gly236 were replaced by tryptophan, respectively, were constructed, and were expressed in E coli cells The Rad51-D231W, Rad51-S233W, and Rad51-G236W mutants were purified to near homogeneity by the same proto-col employed with the wildtype HsRad51 (Fig 1C, lanes 7–9), while the G234W and Rad51-R235W mutants could not be purified because they formed insoluble aggregates In contrast to the Rad51-Y232W mutant, the Rad51-D231W, Rad51-S233W, and Rad51-G236W mutants did not cause significant defects in ssDNA binding and dsDNA binding
A
B
Fig 3 The DNA binding activities of the Rad51 mutants (A) The ssDNA binding experiments The /X174 circular ssDNA (40 l M ) was incu-bated with HsRad51 or the Rad51 mutants at 37 C for 10 min (B) The dsDNA binding experiments Linearized /X174 DNA (20 l M ) was incubated with HsRad51 or the Rad51 mutants at 37 C for 10 min The samples were analyzed by 0.8% (w ⁄ v) agarose gel electrophoresis
in 1 · TAE buffer Lanes 1, 5, 9, 13, 17, 21, 25, and 29 indicate control experiments without HsRad51 The bands were visualized by ethi-dium bromide staining The protein concentrations used in the ssDNA binding experiments were 1 l M (lanes 2, 6, 10, 14, 18, 22, 26, and 30), 2 l M (lanes 3, 7, 11, 15, 19, 23, 27, and 31), and 4 l M (lanes 4, 8, 12, 16, 20, 24, 28, and 32).
Trang 5(Fig 3A,B) These results suggest that the Asp231,
Ser233, and Gly236 residues of HsRad51 are not in
direct contact with DNA As expected, the
Rad51-S233W and Rad51-G236W mutants were proficient
in strand exchange (Fig 4B) However, the
Rad51-D231W mutant was defective in strand exchange
(Fig 4B, lanes 17–19), suggesting that this acidic
resi-due (Asp231) may have some role in this process
Therefore, the Rad51-S233W and Rad51-G236W
mutants are suitable for the fluorescent spectroscopic
analysis, in contrast to the Rad51-Y232W mutant,
which is significantly defective in dsDNA binding and
strand exchange
Fluorescent spectroscopic analysis of the
HsRad51-L1 mutants
The fluorescence change of the D231W,
Rad51-S233W, and Rad51-G236W mutants upon DNA
bind-ing was examined, to confirm that the L1 loop is in
the DNA-binding site HsRad51 has no tryptophan
residue, and therefore, the fluorescence of these
mutants corresponded to that of the inserted
trypto-phan residue The fluorescence peaked at 341, 343 and
347 nm for the tryptophan residues inserted at
posi-tions 231, 233, and 236 of HsRad51, respectively The
peak positions indicate that residues 231 and 233 are
in a rather nonpolar environment (only partly exposed
to the solvent), while residue 236 is more exposed to
the solvent The fluorescence intensity decreased by
about 30, 50 and 60% for the tryptophan 231, 233 and
236 residues, respectively, in the presence of poly(dT),
a model ssDNA, with or without ATP (Table 1) These results confirm that the residues are close to the DNA-binding site
We then examined if these fluorescence changes occurred by the binding of the first or second DNA,
by titrating these modified Rad51 proteins with poly(dT) In the presence of ATP, Rad51 can bind at least two DNA strands, each with a stoichiometry of 3 bases per monomer However, the fluorescence changes
of these proteins were almost saturated at 3 bases per monomer of poly(dT), showing that the changes were mainly due to the binding of the first DNA, while the binding of the second DNA had less influence (Fig 5A) To ensure that the estimation of the pro-tein:poly(dT) ratio was correct, we performed the titra-tion in the absence of ATP, where Rad51 binds only one DNA strand, with a stoichiometry of about 4–5 bases per monomer [27] The titration of these proteins revealed that the fluorescence change was saturated at about 4–5 bases per monomer of poly(dT) for all of the Rad51 mutants (data not shown), as expected
By contrast, the change in fluorescence upon dissoci-ation of the Rad51 filament to monomers, by adding 2.5 m urea [28], was slight (Table 1) and shifted the fluorescence peak to a shorter wavelength We expected the peak position to move to a longer wavelength if the residue is involved in the subunit–subunit interaction, because of its exposure to solvent upon the
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
joint molecule (jm)
nicked circular DNA (nc)
ssDNA dsDNA
jm nc
A
B
Fig 4 The strand-exchange activities of the Rad51 mutants (A) A schematic diagram of the strand-exchange assay (B) The Rad51 concen-trations were 1 l M (lanes 2, 5, 8, 11, 14, 17, 20, and 23), 2 l M (lanes 3, 6, 9, 12, 15, 18, 21, and 24), and 4 l M (lanes 4, 7, 10, 13, 16, 19,
22, and 25) Lane 1 indicates a negative control experiment without the Rad51 protein Joint molecules and nicked circular DNA are indica-ted by jm and nc, respectively.
Trang 6ciation, like the case of Tyr188 of Xenopus laevis Rad51 [28] These results indicate that these residues in the modified Rad51 proteins are not strongly involved
in the subunit–subunit contacts However, the signifi-cant change in the peak position suggests that the L1 loop is close to the subunit–subunit interface
The binding of ATP only minimally affected the fluorescence intensity in the modified Rad51 proteins: only about a 5% increase was detected for Trp231 and Trp233, and a 10% decrease was found for Trp236 The peak position is also only slightly affected by ATP (less than 1 nm) The results confirm the conclusion obtained from the mutational analysis that the Rad51-L1 loop is not directly involved in ATP binding or ATP hydrolysis The slight change in the fluorescence intensity upon nucleotide binding indicates that some environmental change occurs around the L1 loop This nucleotide-induced allosteric effect on the L1 loop may explain the mechanism of DNA-binding regulation by nucleotide binding
The L2 loop is close to the DNA-binding site The L2 loop of HsRad51 contains only one aromatic residue, Phe279, which is highly conserved among the eukaryotic and archaeal Rad51 proteins (Fig 1B) For this residue, we performed analyses similar to those for Tyr232 of the L1 loop, to examine its role in DNA binding, by preparing two mutant proteins, Rad51-F279A and Rad51-F279W, in which the Phe279 residue was replaced by alanine and tryptophan, respectively (Fig 1C, lanes 3 and 5) The CD spectrum (Fig 2C), elution pattern from the gel filtration col-umn (data not shown), and salt-induced ATPase activ-ity (Fig 2F) of the purified Rad51-F279A were all similar to those of HsRad51, indicating that the muta-tion did not affect the global structure, the polymer formation, and the ATPase activity The mutation also did not affect the DNA-dependent ATPase, DNA binding, and strand-exchange activities of HsRad51 (Figs 2F, 3, and 4)
Because the Rad51-F279A mutant did not show a deficiency in DNA binding, we next tested the fluores-cence changes of Rad51-F279W upon DNA binding Rad51-F279W was confirmed to bind DNA like HsRad51, according to the gel retardation experiments (Fig 3A,B, lanes 17–20) The fluorescence of Rad51-F279W peaked at 340 nm, with a rather large intensity (Table 1) This fluorescence feature indicates that the residue is only partly exposed to the solvent or exists
in a rather nonpolar environment, like Trp231 and Trp233, but does not strongly contact other residues The fluorescence of Rad51-F279W was strongly
(kmax
(Imax
kmax
I max
kmax
I max
kmax
I max
kmax
I max
Trang 7decreased (more than 30% decrease) upon poly(dT)
binding, without a change in peak position, in both
the presence and absence of ATP (Table 1) Although
a smaller change was observed upon dsDNA binding
(15%), it was accompanied by a change in the peak
position ()1 nm) These results suggest that the residue
is close to the DNA-binding site The titration of
Rad51-F279W with poly(dT) revealed that the
fluores-cence change became saturated at about 4.5 bases per
monomer of poly(dT) in the absence of ATP (data not
shown), as observed for Xenopus Rad51 [27] In the
presence of ATP, the fluorescence change was almost
saturated with 3 bases per monomer of poly(dT), the
amount needed to saturate the first DNA-binding site
(Fig 5B) The binding of the second poly(dT) did not
change the fluorescence of Rad51-F279W, in contrast
to the results obtained with RecA-F203W, with a
tryp-tophan inserted in the L2 loop of RecA [26], which
displayed changes in fluorescence with the second poly(dT) The addition of poly(dA):poly(dT) duplex DNA to the preformed Rad51-Y279W–ATP–poly(dT) complex also did not affect the fluorescence (Table 1), suggesting that residue 279 is not close to the second DNA
The fluorescence was not significantly changed by the addition of nucleotides (ATP and ADP) or by the addition of 2.5 m urea, which dissociates the protein to monomers, confirming the conclusion from the muta-tional analysis that the residue is involved in neither the subunit–subunit contacts nor ATP hydrolysis (Table 1)
Discussion
We have investigated the DNA-binding sites of HsRad51 to understand the mechanism of homologous pairing and strand exchange catalyzed by this protein for homologous recombination Because several rela-tionships exist between homologous recombination and cancer [29–32], and Rad51 is thus a potential tar-get for anticancer treatment [33], this study would also contribute to its development Our mutational and fluorescent spectroscopic analyses indicated the involvement of the HsRad51 L1 and L2 loops in DNA binding, like the case of RecA However, there could
be some mechanistic differences between the proteins
The Rad51-L1 loop
In the present study, we found that the replacement of Tyr232 with alanine strongly reduced the ssDNA- and dsDNA-binding abilities of HsRad51 The fact that even the conservative replacement with tryptophan affects the dsDNA binding clearly indicated that Tyr232 is involved in DNA binding by HsRad51 The tryptophan-scanning mutagenesis suggested that other residues, such as Asp231, Ser233 and Gly236, within the L1 loop are less important for DNA binding, although the D231W mutation affects the strand-exchange reaction The proximity of the L1 loop to the DNA-binding site was also verified by fluorescent spectroscopic analyses of tryptophan residues inserted
in this loop The importance of the L1 loop for DNA binding by RecA⁄ Rad51 family proteins was observed
by mutational and photocrosslinking analyses of RecA [34–37] and a mutational analysis of HsDmc1 [22] The Dmc1-F233A mutation, which corresponds to the Y232A mutation of HsRad51, also affected DNA binding However, interestingly, the Dmc1-F233A mutation affected only ssDNA binding, but not dsDNA binding [22], in contrast to the case of the
A
B
Fig 5 Fluorescence changes of tryptophan probes inserted in the
L1 and L2 loops of Rad51 upon poly(dT) binding The fluorescence
intensities of 1 l M modified Rad51, in which a tryptophan probe
was inserted within the L1 (A) or L2 loop (B), were measured at
350 nm, after each stepwise addition of poly(dT) in the presence of
1 m M ATP The intensity was normalized to that of the
correspond-ing protein without poly(dT), and is presented as a function of the
poly(dT):protein ratio (A) Graphic representation of fluorescence
intensities with Rad51: Rad51-D231W (d), Rad51-S233W (m) and
Rad51-G236W (n) (B) Graphic representation of the fluorescence
intensities with Rad51-F279W (s).
Trang 8Rad51-Y232A mutant Therefore, the DNA-binding
mode of HsRad51 should be somewhat different from
that of HsDmc1 The octameric ring form [38,39] and
the helical filament form [40] of HsDmc1 are capable
of binding DNA, but Rad51 does not form such an
octameric ring with DNA These differences in the
higher ordered structures of Rad51 and Dmc1 with the
DNA may reflect the different DNA binding modes of
these proteins
Interestingly, even the conservative replacement of
Rad51-Tyr232 by tryptophan affects the dsDNA
bind-ing, indicating direct contact between Tyr232 and
DNA Aromatic residues, such as Tyr, Phe, and Trp,
can stack with the base moieties of ssDNA for the
inter-action Such contacts have been observed in the
interac-tions of the SSB and RPA proteins with ssDNA [23,24]
The Rad51-Tyr232 side chain could thus stack with
DNA bases for its interaction with DNA A reduction
in the tyrosine fluorescence intensity of human Rad51
upon DNA binding was reported [41] The fluorescence
of Tyr232 could be the source of this fluorescence
change Consistent with its importance, the Tyr232
resi-due is highly conserved as an aromatic resiresi-due among
the eukaryotic and archaeal Rad51 and Dmc1 proteins
In contrast to Rad51, Tyr232 is not conserved in the
L1 loop of E coli RecA This fact suggests that the L1
loop of Rad51 interacts with DNA in a different
man-ner from that of RecA His163 is the only residue with
a ring structure similar to that of tyrosine in the
L1 loop of RecA A chemical interference analysis
revealed the protection of one of the two histidine
resi-dues, His97 and His163, of RecA by DNA binding
[42] His163 could be the protected histidine residue
However, its chemical modification did not affect the
DNA binding by RecA [42], and the residue
appar-ently could be replaced with another amino acid [35],
unlike the Rad51-Tyr232 residue Therefore, the
His163 residue of RecA is not functionally equivalent
to the Tyr232 residue of HsRad51
The Rad51-L2 loop
In contrast to the Tyr232 residue, the direct
involve-ment of the Phe279 residue within the HsRad51-L2
loop in DNA binding is less evident, because the
F279A mutation in the L2 loop did not reduce the
DNA-binding ability of HsRad51 It has been reported
that some other mutations of residues in the ScRad51
and HsRad51 L2 loops did not affect the
DNA-bind-ing abilities [43,44] In addition, most of the Rad51
mutants with a mutation in the L2 loop displayed
enhanced DNA-binding abilities [43,44] This
enhance-ment may be caused by an allosteric effect induced by
mutations on the DNA-binding site of Rad51, suggest-ing that the L2 loop of Rad51 is not far from the DNA-binding site Consistent with this idea, the fluor-escence of the tryptophan inserted in the place of Rad51-Phe279 strongly decreased upon poly(dT) binding, suggesting that this residue is close to the DNA-binding site The fluorescence change upon DNA binding is not strong enough for a stacking interaction of the residue with a DNA base, but is large enough to indicate DNA binding in its proximity Several experimental methods, including photocros-slinking, mutational analysis, and fluorescence meas-urements, have been used to show that the RecA-L2 loop is involved in DNA binding [26,36,37,45] Satura-tion mutagenesis of the RecA-L2 loop revealed that mutations in the L2 amino acid residues result in recombination defects in vivo [46] In addition, 20 resi-due peptides that comprise the L2 loop region can bind DNA by forming filamentous beta structures [47–49]
In the L2 peptide, an aromatic residue, which corres-ponds to Phe203, was found to be absolutely required for the DNA binding [47] Therefore, the L2 loop may be a functional DNA-binding site among the RecA⁄ Rad51 class of proteins, although its DNA-bind-ing mode differs somewhat between RecA and Rad51
Experimental procedures Preparation of the human Rad51 mutants
The Rad51 mutant genes, inserted at the NdeI site of
Germany), were constructed using a Quik-Change kit (Stratagene, La Jolla, CA, USA) The hexahistidine-tagged HsRad51 and Rad51 mutants were expressed in the E coli JM109(DE3) strain, which also carries an expression vector for the minor tRNAs (Codon(+)RIL, Novagen) The proteins were purified on nickel-nitrilotriacetic acid (Ni-NTA) agarose (Qiagen, Hilden, Germany) The hexahisti-dine tag was then removed from the Rad51 portion with thrombin protease (Amersham Biosciences, Piscataway, NJ, USA) Then, the HsRad51 and the Rad51 mutants without
HsRad51 and the Rad51 mutants were precipitated (sper-midine precipitation) [50], and the proteins were dissolved
in 100 mm potassium phosphate buffer (pH 7.0) containing
150 mm NaCl, 1 mm EDTA, 2 mm 2-mercaptoethanol, and
were further purified by chromatography on a MonoQ col-umn (Amersham Biosciences) The purified HsRad51 and
Trang 9buffer (pH 7.5), containing 150 mm NaCl, 0.1 mm EDTA,
(Hercules, CA, USA) protein assay kit with bovine serum
albumin as the standard protein
DNAs
The /X174 phage ssDNA and dsDNA used in the DNA
binding and ATPase assays were purchased from New
England Biolabs (Ipswich, MA, USA) Poly(dT) and
poly(dA):poly(dT) were obtained from Amersham
Bio-sciences All of the DNA concentrations are expressed in
moles of nucleotides
Assays for DNA binding
The /X174 circular ssDNA (40 lm) or the PstI-digested
linear /X174 dsDNA (20 lm) was mixed with the Rad51
protein or the Rad51 mutants in 10 lL of 25 mm Hepes
0.1 mm EDTA, 1 mm 2-mercaptoethanol, 1 mm
1· TAE buffer (40 mm Tris ⁄ acetate and 1 mm EDTA) at
bromide staining
Assays for strand exchange
The /X174 circular ssDNA (40 lm) was incubated with the
in 10 lL of 20 mm potassium phosphate buffer (pH 7.4),
kinase After this incubation, 2 lm RPA and 0.2 m KCl
were added to the reaction mixture, which was incubated at
addition of 20 lm /X174 linear dsDNA, and were
contin-ued for 1 h The reactions were stopped by the addition of
Applied Science, Basel, Switzerland), and were further
dye, the deproteinized reaction products were separated by
gold (Invitrogen, Carlsbad, CA, USA) staining
Gel filtration
Rad51 (150 lg) and Rad51 mutants (150 lg) were analyzed
filtration chromatography The elution buffer contained
CD measurements
or the Rad51 mutants was measured on a JASCO J-820 spectropolarimeter (Japan Spectroscopic Co., Ltd, Tokyo, Japan) using a 1 cm pathlength quartz cell All of the CD
(pH 7.5), containing 150 mm NaCl, 0.1 mm EDTA, 2 mm
Fluorescence measurements
Fluorescence was measured with an FP-6500 spectrofluo-rometer (Japan Spectroscopic Co., Ltd), in 20 mm potas-sium phosphate buffer (pH 7.4), containing 50 mm NaCl,
in the presence or absence of 1 mm ATP The emission spectra were measured (bandwidth: 3 nm; response time:
with continuous stirring (300 r.p.m per min), or in a
excitation wavelength was 295 nm (bandwidth: 3 nm) for selective excitation of the tryptophan residue The spectra were measured at least twice to verify the absence of signifi-cant photobleaching, and were averaged to increase the sig-nal to noise ratio All of the spectra were corrected for the Raman signal and background by subtracting the spectrum
of the buffer
ATPase activity
Rad51 (5 lm) or a Rad51 mutant (5 lm) was incubated with 1 mm ATP (Roche, ATP sodium salt) in 25 mm Hepes
0.1 mm EDTA, 1 mm 2-mercaptoethanol, 1 mm
ssDNA or dsDNA In the ssDNA-dependent reaction, the /X174 circular ssDNA (20 lm) was used as a substrate In the dsDNA-dependent reaction, the /X174 RF I DNA (20 lm) (supercoiled dsDNA) was used as a substrate In the high salt conditions, the reaction mixture contained
10 min preincubation in the absence of ATP, the reaction was initiated by adding 1 mm ATP Then, a 20 lL aliquot
of the reaction mixture was mixed with 30 lL of 100 mm EDTA, to quench the reaction at the indicated time The amount of inorganic phosphate released was determined by
a colorimetric assay [51,52] Briefly, 500 lL of a malachite
Trang 101.05% (w⁄ v) hexaammonium heptamolybdate tetrahydrate,
with 50 lL of sample solution (i.e., the reaction mixture
sodium citrate dihydrate was added to stop further color
development The absorbance at 655 nm was measured
phosphate ion standard solution (Wako Pure Chemicals,
Osaka, Japan) was used to prepare phosphate standards
Acknowledgements
We thank Dr Chantal Prevost (CNRS-UPR) and Mr
Sebastien Conilleau for discussions, and Dr Takashi
Kinebuchi (RIKEN) for the CD measurements The
fluorometer was kindly provided by Jasco
Interna-tional This work was partly supported by a grant
from the Association pour la Recherche contre le
Can-cer (No 4813) to MT, and Grants-in-Aid from the
Jap-anese Society for the Promotion of Science (JSPS), and
the Ministry of Education, Sports, Culture, Science,
and Technology, Japan to HK HK and IS were
supported by the Consolidated Research Institute for
Advanced Science and Medical Care, Waseda
Univer-sity
References
1 Shinohara A, Ogawa H & Ogawa T (1992) Rad51
pro-tein involved in repair and recombination in S
2 Shibata T, DasGupta C, Cunningham RP & Radding
CM (1979) Purified Escherichia coli recA protein
cata-lyzes homologous pairing of superhelical DNA and
sin-gle-stranded fragments Proc Natl Acad Sci USA 76,
1638–1642
3 McEntee K, Weinstock GM & Lehman IR (1979)
Initiation of general recombination catalyzed in vitro by
the recA protein of Escherichia coli Proc Natl Acad Sci
USA 76, 2615–2619
4 Cox MM & Lehman IR (1981) recA protein of
dis-tinct phase of DNA strand exchange Proc Natl Acad
Sci USA 78, 3433–3437
5 Cox MM & Lehman IR (1981) Directionality and
polarity in recA protein-promoted branch migration
Proc Natl Acad Sci USA 78, 6018–6022
6 Bishop DK, Park D, Xu L & Kleckner N (1992)
DMC1: a meiosis-specific yeast homolog of E coli recA
required for recombination, synaptonemal complex
for-mation, and cell cycle progression Cell 69, 439–456
7 Pittman DL, Cobb J, Schimenti KJ, Wilson LA, Cooper
DM, Brignull E, Handel MA & Schimenti JC (1998)
Meiotic prophase arrest with failure of chromosome
synapsis in mice deficient for Dmc1, a germline-specific RecA homolog Mol Cell 1, 697–705
8 Yoshida K, Kondoh G, Matsuda Y, Habu T, Nishi-mune Y & Morita T (1998) The mouse RecA-like gene Dmc1 is required for homologous chromosome synapsis during meiosis Mol Cell 1, 707–718
9 Lim DS & Hasty P (1996) A mutation in mouse rad51 results in an early embryonic lethal that is suppressed
by a mutation in p53 Mol Cell Biol 16, 7133–7143
10 Tsuzuki T, Fujii Y, Sakumi K, Tominaga Y, Nakao K, Sekiguchi M, Matsushiro A, Yoshimura Y & Morita T (1996) Targeted disruption of the Rad51 gene leads to lethality in embryonic mice Proc Natl Acad Sci USA
93, 6236–6240
11 Sonoda E, Sasaki MS, Buerstedde JM, Bezzubova O, Shinohara A, Ogawa H, Takata M, Yamaguchi-Iwai Y
& Takeda S (1998) Rad51-deficient vertebrate cells accu-mulate chromosomal breaks prior to cell death EMBO
J 17, 598–608
12 Sung P (1994) Catalysis of ATP-dependent homologous DNA pairing and strand exchange by yeast RAD51 protein Science 265, 1241–1243
13 Baumann P, Benson FE & West SC (1996) Human Rad51 protein promotes ATP-dependent homologous pairing and strand transfer reactions in vitro Cell 87, 757–766
14 Maeshima K, Morimatsu K & Horii T (1996) Purifica-tion and characterizaPurifica-tion of XRad51.1 protein, Xenopus RAD51 homologue: recombinant XRad51.1 promotes strand exchange reaction Genes Cells 1, 1057–1068
15 Gupta RC, Bazemore LR, Golub EI & Radding CM (1997) Activities of human recombination protein Rad51 Proc Natl Acad Sci USA 94, 463–468
16 Takahashi M & Norden B (1994) Structure of RecA-DNA complex and mechanism of RecA-DNA strand exchange reaction in homologous recombination Adv Biophys 30, 1–35
17 Story RM, Weber IT & Steitz TA (1992) The structure
of the E coli recA protein monomer and polymer Nature 355, 318–325
18 Shin DS, Pellegrini L, Daniels DS, Yelent B, Craig L, Bates DYuDS, Shivji MK, Hitomi C, Arvai AS, Volk-mann N, Tsuruta H, Blundell TL, Venkitaraman AR & Tainer JA (2003) Full-length archaeal Rad51 structure and mutants: mechanisms for RAD51 assembly and control by BRCA2 EMBO J 22, 4566–4576
19 Wu Y, He Y, Moya IA, Qian X & Luo Y (2004) Crys-tal structure of archaeal recombinase RADA: a snap-shot of its extended conformation Mol Cell 15, 423– 435
20 Conway AB, Lynch TW, Zhang Y, Fortin GS, Fung
CW, Symington LS & Rice PA (2004) Crystal structure
of a Rad51 filament Nat Struct Mol Biol 11, 791–796
21 Pellegrini L, Yu DS, Lo T, Anand S, Lee M, Blundell
TL & Venkitaraman AR (2002) Insights into DNA