To examine the essential nature of its activity, we analyzed wild-type and mutant RecD proteins with substitutions of important residues in each of the seven conserved helicase motifs..
Trang 1ATPase activity of RecD is essential for growth of the
Antarctic Pseudomonas syringae Lz4W at low temperature Ajit K Satapathy, Theetha L Pavankumar, Sumana Bhattacharjya, Rajan Sankaranarayanan and Malay K Ray
Centre for Cellular and Molecular Biology, Hyderabad, India
RecD is a 5¢ fi 3¢ helicase motor protein The primary
sequence contains the characteristic seven conserved
motifs (I, Ia, II, III, IV, V, and VI) of the
superfam-ily 1 (SF1) group of DNA helicases [1] (Fig 1) In
Esc-herichia coli, RecD displays ssDNA-dependent ATPase
and helicase activity in vitro [2,3] In vivo, it functions
as a component of the RecBCD complex (also known
as exonuclease V) that is involved in DNA repair and
recombination in many bacteria [4] RecBCD is a
highly processive helicase⁄ nuclease enzyme with dual
motor activity, in which RecB and RecD subunits,
with their respective (3¢ fi 5¢) and (5¢ fi 3¢) polar
movement, translocate the enzyme along the
anti-par-allel strands of dsDNA DNA unwinding by helicase
activity is accompanied by degradation of the strands
until the enzyme encounters the recombination hotspot
v (chi) sequence (5¢-GCTGGTGG-3¢) This changes the nuclease property of the enzyme, leading to the generation of 3¢-extended ssDNA and loading of RecA onto the DNA for homologous pairing and DNA strand exchange, producing recombination inter-mediates [5] Interestingly, however, RecBC alone is proficient for recombination and repair of DNA, and recD-inactivated mutants of E coli do not show any growth defects [6,7] Thus, the contribution of the RecD subunit is thought to be of less significance
in vivo Remarkably, RecD inactivation leads to the loss of exonuclease V activity in cells, despite the fact that the only nuclease catalytic center of RecBCD complex lies in the RecB subunit [8] Hence,
a role for RecD in regulating the nuclease activity of RecBCD has been advocated Recently, using ATP
Keywords
cold adaptation; Pseudomonas syringae;
RecBCD enzyme; RecD ATPase; RecD
helicase
Correspondence
M K Ray, Centre for Cellular and Molecular
Biology, Uppal Road, Hyderabad 500007,
India
Fax: +91 40 2716 0591
Tel: +91 40 2719 2512
E-mail: malay@ccmb.res.in
(Received 10 October 2007, revised 12
February 2008, accepted 18 February 2008)
doi:10.1111/j.1742-4658.2008.06342.x
RecD is essential for growth at low temperature in the Antarctic psychro-trophic bacterium Pseudomonas syringae Lz4W To examine the essential nature of its activity, we analyzed wild-type and mutant RecD proteins with substitutions of important residues in each of the seven conserved helicase motifs The wild-type RecD displayed DNA-dependent ATPase and helicase activity in vitro, with the ability to unwind short DNA duplexes containing only 5¢ overhangs or forked ends Five of the mutant proteins, K229Q (in motif I), D323N and E324Q (in motif II), Q354E (in motif III) and R660A (in motif VI) completely lost both ATPase and heli-case activities Three other mutants, T259A in motif Ia, R419A in motif IV and E633Q in motif V exhibited various degrees of reduction in ATPase activity, but had no helicase activity While all RecD proteins had DNA-binding activity, the mutants of motifs IV and V displayed reduced bind-ing, and the motif II mutant showed a higher degree of binding to ssDNA Significantly, only RecD variants with in vitro ATPase activity could complement the cold-sensitive growth of a recD-inactivated strain of
P syringae at 4C These results suggest that the requirement for RecD
at lower temperatures lies in its ATP-hydrolyzing activity
Abbreviations
ABM, Antarctic bacterial medium; ATPc-S, adenosine 5¢-O-(thiophosphate); EMSA, electrophoretic gel mobility shift assay;
SF1, superfamily 1.
Trang 2B
C
Fig 1 Schematic representation of P syringae RecD (A) Location of the seven conserved helicase motifs on linear RecD (shown as a hori-zontal bar) are indicated by shaded boxes, except motifs I and II, known as Walker motifs A and B, which are shown in black The amino acid substitutions (in single-letter code) that were introduced into the helicase motifs are shown above the bar, with the position number of the residues between the wild-type and mutated amino acids The location of the H386D mutation between motifs III and IV is also indi-cated (B) Alignment of the amino acids of the seven helicase motifs of RecD from P syringae (Ps) and E coli (Ec) and other well-studied members of DNA helicases belonging to SF1 (Rep of E coli, PcrA of Bacillus stearothermophilus (Bs) and UvrD of E coli), indicating the conserved nature of the residues The mutated residues of P syringae RecD are underlined Asterisks indicate amino acids that are identical
to the residue in P syringae RecD (C) Ribbon diagram of the structural model of P syringae RecD The model was built by homology mod-eling using the coordinates of the E coli RecD (D-chain of the RecBCD complex, Protein Data Bank code 1W36) The three domains of RecD, and the residues that were mutated in the seven conserved motifs, in addition to residue H386, are indicated The arrowheads mark the positions of the putative insertion sequences in P syringae RecD, which are absent from E coli RecD The 5¢-end of the DNA substrate has also been shown schematically to indicate the relative positions of domains 2 and 3 of RecD as seen in the structure of the DNA-bound RecBCD complex of E coli.
Trang 3hydrolysis-defective mutants of the helicase motif I in
RecD (RecDK177Q) and RecB (RecBK29Q) of E coli, it
has been concluded that there are subtle differences
between the properties of RecBC, RecBCDK177Q and
RecBK29QCD enzymes, and that the RecB motor is
absolutely required for v recognition and RecA
load-ing, while the RecD subunit is dispensable for motor
activity of the complex [9]
Psychrophilic and psychrotrophic bacteria from
Antarctica have evolved various novel adaptive features
that allow them to survive and grow at a very low
temperature [10–14] A molecular understanding of
these features would be important to our knowledge
regarding low-temperature-adapted biology We
previ-ously discovered that recD is essential for growth of the
Antarctic bacterium Pseudomonas syringae Lz4W at
low temperature [15] The peizophilic bacterium
Photo-bacterium profundum also required RecD function
during growth under high pressure [16] These two
studies suggested that the RecD protein might be
required for growth of bacteria under stress conditions,
as E coli does not show any growth defect due to recD
inactivation In addition, we observed that the
recD-inactivated cold-sensitive P syringae mutants
accumu-late DNA fragments in cells grown at 4C but not at
22C [15] Concurrently, the recD mutants were also
sensitive to DNA-damaging agents, such as UV and
mitomycin C, unlike in the case of mesophilic E coli
[6,7] This led us to believe that the Antarctic bacteria
are probably subjected to greater DNA damage at low
temperature, and RecD might play a direct role in
the RecBCD-dependent repair of such damage As
P syringae possesses genes for the RecB (recB) and
RecC (recC) subunits, we have initiated studies to
examine their role in cold adaptation A recent genetic
study (T L Pavankumar and M K Ray, unpublished
results) indicated that the recB and recC mutants of
P syringae also are cold-sensitive like the recD
mutants, suggesting that function of the entire RecBCD
machinery is important for growth In the case of
E coli, mutations in the recB or recC genes impair
homologous recombination, and the mutant cells have
reduced cell viability and reduced resistance to
DNA-damaging agents [4,6]
As a first step towards gaining an insight into why
RecD is essential in P syringae, we have analyzed the
in vitrobiochemical activities of this protein We report
here the comparative activities of the C-terminally
hexahistidine-tagged form of wild-type RecD (RecDHis)
and eight mutant proteins that were created by single
amino acid substitutions of important residues in each
of the seven conserved helicase motifs RecDHis
dis-played ATP-hydrolyzing as well as short DNA duplex
unwinding activity in vitro, but the mutations K229Q in motif I (Walker motif A), D323N and E324Q in
moti-f II (Walker motimoti-f B), Q354E in motimoti-f III and R660A
in motif VI caused complete loss of these activities in RecD However, the three mutant proteins of motifs Ia (T259A), IV (R419A) and V (E633Q) retained reduced ATPase activity to varying degrees, but showed no DNA-unwinding activity In the biological activity assay, only the wild-type and the three mutant proteins retaining ATPase activity were able to complement the growth defect of a recD-disrupted strain (CS1) of
P syringae These results suggest that RecD with mod-est ATP-hydrolyzing activity, which does not support DNA unwinding in vitro, is sufficient for growth of the Antarctic P syringae at low temperature
Results
Selection of amino acid residues for mutational analysis of P syringae RecD
To dissect the biochemical activities of RecD with regard to its requirement during growth at low temper-ature, we used a mutational approach, assessing the roles of conserved amino acids in various helicase motifs of the RecD motor protein (Fig 1) Eight of the conserved residues (K229, T259, D323, E324, Q354, R419A, E633Q and R660) chosen in this study for mutational analysis are located on the seven heli-case motifs (I, Ia, II, III, IV, V, and VI) whose roles have been assessed in other helicases [17,18] One other residue, H386, which was mutated to D, is located out-side the conserved motifs of RecD (Fig 1), although
it was putatively identified to be on motif IV in a previous study [19] The incorrect identification was primarily due to blast and clustal w alignments of amino acid sequences of RecD proteins showing that the E coli RecD sequence 328QLSRLTGT335 and the P syringae RecD sequence 383WLEHVSGE390 align with the helicase motif IV sequence 284QNYRSTKR291 of PcrA and 281QNYRSTSN288
of UvrD, respectively [19–21].Using the recent crystal structure of RecD in the RecBCD complex obtained from E coli [22], we built a structural model of the
P syringae RecD protein (Fig 1C), which establishes that the RecD sequences 415RHSRRFGEG423 in
P syringae and 356QKSYRFGSD364 in E coli repre-sent motif IV The sequences are located in a structur-ally similar region of Rep and PcrA helicases [23,24] The H386D mutation located outside the conserved helicase motifs nevertheless gave us an opportunity to compare its biochemical and biological activities with those of the wild-type and other mutated proteins
Trang 4The structural model of the P syringae RecD
(Fig 1C) was built by homology modeling, using
E coliRecD [22] as the template The two proteins are
highly homologous, given that 519 Caatoms across the
length of the proteins could be superimposed with an
rmsd value of 1.21 A˚ (supplementary Fig S1) They are
also similar in their domain architectures, each
contain-ing three distinct domains (domains 1–3) Domain 2
(residues 159–417) and domain 3 (residues 418–682) of
P syringae RecD, corresponding to homologous
seg-ments (residues 110–358 and 359–593) of E coli RecD,
represent the motor domains 1A and 2A of other SF1
helicases [23,24] The N-terminal domain 1 that
consti-tutes the main interface between RecD and RecC in the
RecBCD complex is a little longer in P syringae (1–159
residues) compared to E coli RecD (1–110 residues)
Two more extra segments of amino acids within
domain 3 of P syringae RecD are also present (marked
by arrowheads in Fig 1C)
Expression and purification of RecD in soluble
form
To assess its biochemical activity, P syringae RecD
was initially expressed as a C-terminally
hexahistidine-tagged protein from the high-level expression vector
pET21D-His in E coli, in which it formed inclusion
bodies Therefore, the protein was subsequently
expressed in soluble form in Antarctic P syringae
(Fig S2) using the plasmid pRecDHis, a derivative of
the broad host range plasmid pGL10 (see
Experimen-tal procedures) The levels of expression of the soluble
form of RecD from pRecDHisin the recD-null mutant
of P syringae (CS1), although much lower than the
amount expressed from pET21D-His in E coli, were
satisfactory for purification under native conditions
Hence, the recombinant P syringae RecD was mainly
purified from the CS1 strain However, purification of
His-tagged RecD on Ni2+-agarose by a single-step
method led to the association of a few
co-contaminat-ing proteins Introduction of a heparin–Sepharose
chromatographic step prior to Ni2+-agarose
chroma-tography, as described in Experimental procedures,
eliminated such contamination Typically,
approxi-mately 1–2 mg of His-tagged RecD protein were
purified from 500 mL of overnight cultures of
CS1(pRecDHis) by this method The finally purified
protein was about 99% pure, as determined by SDS–
PAGE analysis with Coomassie blue staining
(supple-mentary Fig S1) Gel-filtration chromatography on a
Superose HR 10⁄ 30 column demonstrated that the
protein elutes as a discrete peak at about 76 kDa,
corresponding to the monomer form of the protein
All eight helicase motif mutants (K229Q, T259A, D323N, E324Q, Q354E, R419A, E633Q and R660A) and one (H386D) outside the conserved motifs of the
P syringae RecD protein (Fig 1) reported in this study were also expressed in CS1 cells Expression
of the proteins was confirmed by western analyses using anti-RecD and anti-His serum (supplementary Fig S1) Mutant proteins were purified by an identical method to that followed for RecDHis, with comparable yield and purity
ATP-hydrolyzing activity of RecDHisand its mutants
We first assessed the ATP-hydrolyzing activity of the recombinant wild-type RecD protein (RecDHis) of
P syringae, which is an essential activity of any heli-case motor The RecDHis displayed efficient ATPase activity in the presence of ssDNA Interestingly, RecD also displayed significant ATPase activity in the pres-ence of duplex DNAs with 5¢ overhangs and forked-end substrates, but much reduced ATPase activity with 3¢ overhangs and blunt-ended DNA (Fig 2A) No detectable intrinsic ATPase activity was associated with the protein
To compare the ATP-hydrolyzing activities of the mutant RecD proteins with those of RecDHis, a single-stranded 40-mer oligonucleotide was used as a stimula-tor under identical conditions (Fig 2B) The kinetic parameters of ATPase activity, obtained from analysis
of RecDHis and the various mutants, are shown in Table 1 RecDHis hydrolyzed ATP with a maximum velocity (Vmax) of 72 lmolÆs)1 and a Km of approxi-mately 147 lm for ATP Five of the mutant RecD pro-teins (K229Q, D323N, E324Q, Q354E, and R660A) had barely detectable ATPase activity However, the mutant proteins T259A, R419A, and E633Q exhibited reduced ATPase activity, about 72, 13 and 7% of the wild-type value, with Km values for ATP (Km(ATP))
of 217, 151 and 136 lm, respectively (Table 1) By varying the DNA concentration in the reaction, the
Km(DNA) values for ATPase stimulation were also determined, and were roughly similar (27–30 nm) to each other
It is generally believed that the ssDNA-dependent ATP hydrolysis activity of helicases is related to its translocation along the DNA strand Therefore, we tested the Vmax of ATPase activity of RecDHis in the presence of DNA oligomers of various lengths, e.g 15-mer, 25-mer and 40-mer (supplementary Table S1) As expected, the Vmax of the reactions increased as the length of the oligomeric DNA chain increased (Fig 3A) However, there was a reduction in the
Trang 5Km(DNA)values with the increase in DNA chain length,
which might be related to the increased residence time
of the proteins on longer DNA substrates Importantly,
the three mutated RecD proteins (T259A, R419A and E633Q) that had reduced ssDNA-dependent ATPase activities also showed a DNA chain-length-dependent increase in ATP hydrolysis (Fig 3B–D)
DNA-unwinding activity of RecD and its mutants Four types of duplex DNA substrate (supplementary Table S1) were used in the DNA strand unwinding assay RecDHis could unwind only the 5¢ overhang substrate (25 bp duplex DNA with a 15-base 5¢ exten-sion) and the forked-end substrate (17 bp duplex with
an 8 bp unpaired extension) (Fig 4A) Unwinding activity was barely detectable in assays with a 25 bp blunt-end DNA duplex or with the 25 bp duplex containing a 3¢ ssDNA tail (3¢ overhang substrate) Although the activity of P syringae RecD was mar-ginally better with the forked-end substrate, charac-terization of helicase activity was subsequently carried out using the 5¢ overhang DNA duplex substrate The helicase activity was found to be ATP- and
Mg2+-dependent, and maximum activity was observed with 2.5 mm ATP and 2.0 mm MgCl2, under our experimental conditions (data not shown) The RecD protein could catalyze the ATPase and DNA strand unwinding activities in the presence of Mg2+
or Mn2+ but not when Ca2+ or Zn2+ were used in the assays Addition of EDTA or removal of ATP from the reaction mixture abolished the DNA strand separation activity The non-hydrolysable ATP ana-logue (ATPc-S) also did not support the activity (data not shown)
To assess the importance of conserved residues in the helicase motifs on DNA-unwinding activity, mutant RecD proteins were tested for their ability to unwind the 5¢ overhang duplex DNA substrate at
25C None of the eight mutants of RecD helicase motifs (K229Q, T259A, D323N, E324Q, Q354E, R419A, E633Q and R660A) showed any measurable helicase activity (Fig 4B) However, the mutant (H386D) that had an alteration outside the conserved helicase motifs showed approximately 76% of the wild-type ATPase activity and approximately 80% of the helicase activity in vitro, under identical conditions (Tables 1 and 2)
DNA-binding activity of the mutant RecD proteins
To further examine whether the loss or reduction in activities of the mutant RecD proteins are due to their inability to bind DNA, the binding activity was assessed by an electrophoretic gel mobility shift assay
Table 1 Kinetic parameters of ssDNA-stimulated ATPase activity
of RecD The coupled NADH oxidation method with 6.4 n M protein
and 1 l M 40-mer ssDNA was used to determine ATPase activity at
25 C The activities of the RecD mutant proteins K229Q, D323N,
E324Q and Q354E, which were very low (0.41, 1.0, 0.43, and
0.86 lmol ATP hydrolysed per lmol RecD per second, respectively)
are not listed, and were not used for calculation of the Kmvalues.
RecD
Vmax (lmolÆs)1)
Km(ATP) (l M )
Km(ssDNA) (n M ) Wild-type (RecD His ) 72 ± 7 147 ± 26 29 ± 2
0
25
50
75
100
40-mer
5 ′-OV Fork-end
3 ′-OV Blunt-end
No DNA
1
10
100
T259A R419A E633Q
–1 )
A
B
Fig 2 ATPase activity of RecD and its mutants DNA-stimulated
ATPase activity was measured spectrophotometrically by the
NADH oxidation-coupled assay method (A) Activity of RecD His
pro-tein (6.6 n M ) in the presence of ssDNA (40-mer) and dsDNA with
various end structures (5¢ overhang, 3¢ overhang, blunt-end and
forked-end, as indicated in Supplementary Table S2) (B)
Compari-son of ATPase activity between RecD His and the three mutant
RecDs (T259A, R419A, E633Q) that displayed reduced activity.
Assays were performed in the presence of 1 l M 40-mer ssDNA
and 6.6 n M proteins Error bars indicate the standard deviation
based on a minimum of three experiments.
Trang 6(EMSA) in the presence of32P-labeled DNA substrates (Fig 5) Wild-type RecD protein displayed stronger binding to ssDNA than to dsDNA at both assay tem-peratures (4 and 25C) From quantification of the band intensities in EMSA (Fig 5A), it appears that the DNA duplexes with a 5¢ or 3¢ overhang or forked-end substrates were preferred (binding to approxi-mately 80–85% of the ssDNA) compared with the
Fig 3 ssDNA length-dependent ATP-hydrolyzing activity of P sy-ringae RecD and its mutants The activity of RecDHis and three mutant proteins (T259A, R419A and E633Q) was measured spec-trophotometrically using 6.6 n M protein in the presence of 1 l M
ssDNA of various lengths (15-, 25- and 40-mer) No activity was observed in the absence of ssDNA (not shown) The curves were obtained by nonlinear fitting of data using GRAPHPAD PRISM software The data are the mean of three independent experiments.
0
25
50
75
100
0
20
40
60
0.0
2.5
5.0
7.5
10.0
0 1 2 3 4 5 6
0.0
2.5
5.0
7.5
C 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
ds
ss
5 ′- Overhang
5′
3′- Overhang
3′
C1 C2 2 5 10 20 min
ds
ss C1 C2 2 5 10 20
Blunt - end 5′
Forked - end 5′
C1 C2 2 5 10 20 min
ds
ss C1 C2 2 5 10 20
A
B
Fig 4 DNA-unwinding activity of P syringae RecD and its mutants (A) RecD His protein (100 n M ) was incubated with 1 n M
32 P-labeled duplex DNA of various types (5¢ overhang, 3¢ overhang, blunt-end and forked-end, as shown in supplementary Table S2) Reactions were carried out at 25 C and analyzed by EMSA on native 15% polyacrylamide gel Shown here are the phosphor images of the gels The lanes marked as C1 and C2 contained con-trol samples with heat-denatured ssDNA and dsDNA substrates, respectively (B) Representative phosphor image of a gel showing the DNA-unwinding activity of RecDHis(WT) and mutant RecD pro-teins (K229Q, T259A, D323N, E324Q, Q354E, R419A, E633Q and R660A) Assays were carried out at 25 C using the 5¢ overhang DNA duplex substrate.
Trang 7blunt-end duplex DNA (approximately 60%)
Signifi-cantly, all the mutant RecD proteins retained the
abil-ity to bind the DNA substrates (Fig 5B) The
efficiency of binding to DNA was, however, variable
among the mutant proteins R419A and E633Q
dis-played weaker ssDNA binding activity (2- and 5-fold
less, respectively) compared to the wild-type RecD
protein (Fig 5C) As these two mutant proteins also showed reduced ATPase activity, the reduction might
be related to the weaker DNA binding On the other hand, the lack of or defective ATPase activity in the remaining mutants, such as K229Q, T259A, D323N, E324Q, Q354E and R660A, could not be related to any DNA-binding defect Surprisingly, two mutants of
Table 2 Summary of the properties of wild-type and mutant RecD proteins Biochemical activities of the wild-type protein (RecD His ) were taken as 100% for evaluation of the activities of the mutated proteins The V max values for ATPase activity of RecDHisat 25 and 4 C were
72 and 21 lmol ATPÆs)1,respectively, which were considered to be 100% for relative activity of the mutant proteins at the respective tem-peratures ND, not detectable under the experimental conditions.
Protein
Complements cold-sensitivity of CS1?
DNA-binding activity (%)
ATPase activity at
25 C (%)
ATPase activity
at 4 C (%)
DNA unwinding
at 25 C (%)
C 1 2 C 1 2 C 1 2 C 1 2 C 1 2
DNA Complex
Free DNA
0
50
100
150
200
250
300
350
a b a b a b a b a b a b a b a b a b
WT K229Q T259A D323N E324Q Q354E R419A E633Q R660A
WT K229Q T259A DNA
complex
E324Q D323N Q354E R419A E633Q R660A
A
B C
Fig 5 DNA-binding activity of the wild-type and mutant RecD proteins (A) Binding activity of RecD His Single-stranded and double-stranded oligonucleotides with various end structures [5¢ overhang (5¢-OV), 3¢ overhang (3¢-OV), blunt-end and forked-end] were analyzed by EMSA on 8% polyacrylamide gel Lanes marked ‘C’ contained32P-end-labeled DNA substrates (2.5 n M ) alone, and lanes marked 1 and 2 contained labeled DNAs and RecD His protein (250 and 500 ng, respectively) (B) Relative ssDNA binding activity of RecD His and mutant RecD proteins Binding assays were performed with 32 P-labeled 25-mer single-stranded oligonucleotides (2.5 n M ) and 500 ng of RecD proteins, and analyzed
by EMSA as in (A) (C) Histogram showing the relative binding activity of various RecD proteins to ssDNA at 4 C (bar ‘a’) and 25 C (bar ‘b’) Binding values were obtained by quantifying the band intensities on gel phosphor images Error bars represent the standard deviation of the values obtained from three independent experiments The ssDNA binding activity of 500 ng RecD His protein was considered as 100% for the calculations.
Trang 8motif II, especially E324Q, displayed a consistently
higher degree of DNA-binding activity (> 2.5-fold)
compared to RecDHis under identical conditions
(Fig 5C) These residues in motif II are known to
interact with ATP and Mg2+ for hydrolysis of the
nucleotide substrate
Genetic complementation of the cold-sensitive
phenotype of CS1 by RecDHisand its mutants
We reported previously that the defect in the P
syrin-gae recD mutant CS1, which does not grow at low
temperature (4C), is complemented by the wild-type
recD gene in trans [15] We have tested the ability of
the recombinant RecDHis protein to support the
growth of CS1 at 4C, by expressing it from the
pRecDHis plasmid As expected, CS1, expressing
RecDHis, grew efficiently at 4C, both in ABM liquid
culture and on ABM agar plates In contrast, CS1 with
the empty plasmid pGL10 failed to grow in the
med-ium at 4C When the eight mutant RecD proteins
were tested in the trans complementation assay, five
(K229Q, D323N, E324Q, Q354E and R660A) failed to
support growth of CS1 at low temperature Only the
three mutant proteins (T259A, R419A and E633Q)
that displayed ATPase activity in vitro could
comple-ment the low-temperature-sensitive growth of the CS1
strain (Fig 6) The generation times (9–11 h) of the
complemented strains expressing the three mutant
proteins were roughly similar to that of the RecDHis
-complemented strain (approximately 8.5 h) It is
important to note that the levels of expression of all RecD proteins in CS1 were observed to be similar by western analysis (supplementary Fig S2) This rules out quantitative differences as an explanation for the observed difference in biological activities of the pro-teins
In vitro activities of RecD at various temperatures Mutant RecD proteins (T259A, R419A and E633Q) that displayed ATP hydrolysis activity but no DNA-unwinding ability in vitro were able to support growth
of the cold-sensitive, recD-disrupted strain (CS1) of
P syringae at 4C This raises some interesting ques-tions about the contribution of these two enzymatic activities to RecD function during growth at low temperature While it is impossible to directly assess these enzymatic activities of RecD in vivo, the relative
in vitro activities of the two enzyme reactions at low temperature (4 C) could be compared between the proteins to detect any correlation and⁄ or their relative importance during growth Towards this goal, we measured the ssDNA-dependent ATPase activity of wild-type RecDHis at various temperatures, using an enzyme-coupled NADH-oxidation assay The initial rate of ssDNA-induced ATP-hydrolyzing activity of RecDHiswas highest at 37C, as seen for many other enzymes from the bacterium [10,25,26] However, ATPase activity dropped sharply to about 2% of the activity at 25C, and could not be measured below
7C by this method (data not shown) To circumvent
Empty
RecD HIs
RecD HIs
K229Q
T259A
D323N
D323N E324Q E324Q
R419A R419A
E633Q E633Q
R660A R660A
Empty
RecD HIs
RecD HIs
T259A
T259A
D323N
D323N E324Q E324Q
R419A R419A
E633Q E633Q
R660A R660A
Fig 6 Complementation of cold-sensitive growth of CS1 by wild-type and mutant RecD proteins The recD-inactivated mutant of P syrin-gae (CS1) was transformed using empty plasmid pGL10, or pGL10-derived constructs expressing RecDHis(WT) or mutated RecD proteins (K229Q, T259A, D323N, E324Q, Q354E, R419A, E633Q and R660A) Growth of the resultant strains was determined at 22 and 4 C on ABM agar plates Only wild-type and mutant proteins T259A, R419A and E633Q could complement the growth defect of CS1 at 4 C.
Trang 9this problem, we employed a TLC method to measure
ATP hydrolysis activity using [c-32P]ATP as the
sub-strate (Fig 7A) This method was also suitable for
measuring ssDNA-dependent ATPase activity under
identical buffer and salt conditions to those employed
in the DNA-unwinding assays in vitro Results from
TLC assays also demonstrated that wild-type RecDHis
has the highest activity at 37C (data not shown), but,
more importantly, that RecD displayed about 30% of
the 25C activity even at a lower temperature (4 C)
in vitro Like the wild-type, the mutant RecD proteins
(T259A, R419A and E633Q) also hydrolyzed ATP
efficiently at 4C, at about 25–35% of their 25 C
activity (Fig 7B)
We then examined the helicase activity of P
syrin-gae RecDHis in vitro at various temperatures, using
identical buffer conditions to the TLC-based ATPase
assay method Again, maximum DNA unwinding was
observed at 37C, and was about 10-fold higher than
that at 25C (Fig 7C) However, at 4 C, RecDHis
(100 nm) failed to show any detectable
DNA-unwind-ing activity When the amount of RecDHis was
increased up to 800 nm and the reaction time to
30 min, the protein could unwind DNA duplex at only
0.8–1% of the 25C activity (data not shown) This is
surprising, considering the fact that the protein
dis-played approximately 30% ATPase activity at 4C
This suggests that the DNA strand separation assay
in vitroprobably underestimates RecD helicase activity
at lower temperatures to a considerable extent
None-theless, the method is robust enough to measure
heli-case activity at higher temperatures (25–37C), and is sufficient to establish that the T259A, R419A and E633Q mutants lack helicase activity, at least under these in vitro conditions
With regard to the DNA-binding activity of the RecD proteins at various temperatures, it appears that,
by and large, the in vitro assay temperatures (4 and
25C) do not affect the binding (Fig 5C) Table 2 summarizes the key biochemical and biological activi-ties of the wild-type and mutated RecD proteins obtained in this study
Discussion Our results establish that recombinantly produced
P syringae RecD has ssDNA-dependent ATPase and 5¢ fi 3¢ helicase activity, like that of the mesophilic
E coli RecD [2,3] However, the Vmax (72 lmolÆs)1) for the ATP-hydrolyzing activity of P syringae RecD
is much higher than the reported value (5 lmolÆs)1) for mesophilic E coli RecD at 25C [3] The Km(DNA) (29 nm) for the P syringae RecD towards ss-DNA, for stimulation of ATPase activity, is lower than the reported value (9 lm) for E coli RecD The higher activity of RecD from P syringae could be due either
to its inherent efficient activity or due to its isolation
in native soluble form, unlike the insoluble form of
E coli RecD that required unfolding and refolding in order to recover the active protein [3] It is perhaps important to point out here that the Km(DNA) values for RecD obtained from the ATPase stimulation
0 10 20 30 0
25 50 75 100
37 ºC
25 ºC
4 ºC
Incubation time (min)
C
32 Pi
WT T259A R419A E633Q 1
10
100
25 °C
4 °C
A
Fig 7 Activity of P syringae RecD protein
at various temperatures ATPase assays
were performed by a TLC method at various
temperatures in the presence of 16.6 n M
RecD His , 1 l M 40-mer ssDNA and 100 l M
[c- 32 P]ATP (A) Representative phosphor
images of the TLC plates from the 25 and
4 C assays (B) Histogram showing the
rel-ative ATPase activities of RecD His (WT) and
three mutant proteins (T259A, R419A and
E633Q) at 4 and 25 C Error bars represent
the standard deviation of the values based
on three experiments (C) Relative
DNA-unwinding activity of RecD His at various
temperatures Reactions were carried out
with 10 n M RecDHisprotein on 1 n M 5¢
over-hang duplex DNA at three temperatures (37,
25 and 4 C) and analyzed as in Fig 4.
Trang 10experiments are much lower than the apparent values
calculated from the EMSA data, for which 2.5 nm
DNA was used to bind to 66.6–333 nm of RecD We
believe that the EMSA method depends largely on
stability of the DNA–protein complex under the
elec-trophoretic conditions used, and hence is likely to be
less sensitive than the ATPase stimulation method for
determination of the Km(DNA)value
Our study also shows that wild-type RecD of
P syringae is very active in unwinding 5¢ overhangs
and forked-end short DNA duplexes (15–25 bp)
in vitro However, RecD fails to unwind duplexes of
> 100 bp (A K Satapathy & M K Ray, unpublished
results), suggesting that RecD, on its own, is not a
‘strong’ helicase in vitro The helicase activities of RecD
from P syringae and E coli could not be compared
due to the different DNA substrates used in the studies
with E coli [3] Additionally, detailed analysis on the
helicase activity of E coli RecD protein alone has not
been reported However, the ability of P syringae
RecD to unwind both 5¢ overhangs and forked-end
duplex DNAs is similar to the behavior of RecD
protein from the radio-resistant bacterium Deinococcus
radiodurans[27] However, the RecD of this bacterium
belongs to RecD2 subgroup, which is present in
bacteria lacking RecBC protein homologues [28]
Mutational effects of conserved residues in the
helicase motifs of RecD
One conserved residue from each of the seven
heli-case motifs (except motif II in which two residues
were changed) has been altered in the present study
to dissect the biochemical activities of RecD The
roles of these conserved residues have been assessed
previously in other helicases by structural and
func-tional analyses, including ATP binding and hydrolysis
(motifs I and II), ssDNA binding (motifs Ia, III, IV,
and V), and coupling of ATPase and helicase
(DNA-unwinding) activities to translocation on ssDNA
(motifs III, IV, V, and VI) [17,18] In the context of
RecD, only the role of the conserved lysine residue
in motif I (Walker motif A) has been investigated
previously in E coli [2,3,9] The lysine residue in
other helicases, including PcrA and UvrD, makes
contact with the b-phosphate of ATP-Mg2+ and
thereby plays a role in the catalytic reaction [17,18]
Consistent with these results, our data show that the
K229 residue is essential for ATP hydrolysis and
DNA-unwinding activities, and, as expected, the
K229Q mutant protein is biologically inactive with
respect to support of the growth of P syringae at
low temperature
Similar to K229, the residues (D323 and E324) in motif II (Walker motif B) are also conserved in RecD
In other helicases, these residues co-ordinate the ATP-associated Mg2+ion and active water molecule for the hydrolytic reaction, and their alteration causes reduc-tion in the ATPase and DNA-unwinding activities [17,18] Consistent with this, the present study demon-strates that D323N and E324Q mutants of P syringae RecD do not display ATPase and helicase activities
in vitro However, a surprising finding here is that the D323N and E324Q mutant proteins bind ssDNA 2.5–3.0-fold more than the wild-type RecD under iden-tical conditions (Fig 5) The implication is that these residues might normally interfere with binding of DNA
In the crystal structure of the DNA-bound RecBCD complex of E coli [22], the 5¢-end of the bound DNA molecule was not in the vicinity of the DNA-binding pocket of RecD Therefore, it is not clear how the invariant D and E residues of the ATP-binding pocket would affect nucleic acid binding Interestingly, Walker motif A and B mutants of RuvB, a 5¢ fi 3¢ hexameric helicase, were also reported to be defective in DNA binding in addition to the ATP-binding defect [29,30] The T259A mutation was created in RecD based on
an analysis showing that the RecD sequence (PTGKAAAR) from both P syringae and E coli is found in similar locations in PcrA and Rep helicase, and they include a conserved threonine residue in their
Ia motifs (64FTNKAAR70 and 55FTNKAAR61, respectively) The conserved threonine in PcrA and Rep proteins was shown to interact with the phosphate backbone of ssDNA [23,24] Although the DNA bind-ing role of residues in motif Ia has been corroborated
by mutational analysis of the UL9 protein (SF2 group
of helicases) of HSV-1 virus [31], our study shows that the RecD T259A mutant protein retains ssDNA bind-ing However, the protein displays reduced ATPase activity (72%) and lacks DNA-unwinding activity
in vitro This might result from the uncoupling of ATP hydrolysis and DNA unwinding, which has been shown previously in the case of the RecBCD enzyme when inter-strand cross-linked DNA duplexes or DNA:RNA hybrids were used as substrates [32,33] Importantly, however, T259A is active in supporting growth of CS1 at 4C Retention of the biological activity suggests that the uncoupled ATPase activity in the mutant protein might have other significance, as discussed below Only a limited number of mutational studies have been carried out on the residues in motif
Ia from various helicases [34,35]
Two highly conserved residues, a glutamine in motif III and an arginine in motif VI of PcrA helicase (cor-responding to Q354 and R660 of P syringae RecD),