| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 275, Issue 22, 16518-16529, June 2, 2000
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Braun Laboratory, California Institute of Technology, Pasadena, California 91125
Received for publication, November 24, 1999, and in revised form, March 3, 2000
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
Saccharomyces cerevisiae Dna2 protein
is required for DNA replication and repair and is associated with
multiple biochemical activities: DNA-dependent ATPase, DNA
helicase, and DNA nuclease. To investigate which of these activities is
important for the cellular functions of Dna2, we have identified
separation of function mutations that selectively inactivate the
helicase or nuclease. We describe the effect of six such mutations on
ATPase, helicase, and nuclease after purification of the mutant
proteins from yeast or baculovirus-infected insect cells. A mutation in
the Walker A box in the C-terminal third of the protein affects
helicase and ATPase but not nuclease; a mutation in the N-terminal
domain (amino acid 504) affects ATPase, helicase, and nuclease. Two
mutations in the N-terminal domain abolish nuclease but do not reduce
helicase activity (amino acids 657 and 675) and identify the putative
nuclease active site. Two mutations immediately adjacent to the
proposed nuclease active site (amino acids 640 and 693) impair nuclease activity in the absence of ATP but completely abolish nuclease activity
in the presence of ATP. These results suggest that, although the Dna2
helicase and nuclease activities can be independently affected by some
mutations, the two activities appear to interact, and the nuclease
activity is regulated in a complex manner by ATP. Physiological
analysis shows that both ATPase and nuclease are important for the
essential function of DNA2 in DNA replication and for its
role in double-strand break repair. Four of the nuclease mutants are
not only loss of function mutations but also exhibit a dominant
negative phenotype.
Yeast dna2-1 mutants were originally identified in a
screen for mutants defective in DNA replication in vitro (1)
and were then shown to be defective in DNA replication in
vivo (1, 2). Since that time, additional dna2 mutants
with similar phenotypes have been identified and characterized (3, 4).
Fluorescence-activated cell sorting analysis shows that
temperature-sensitive dna2 mutants can synthesize a full 2C
DNA content at 37 °C (3).1
The DNA synthesized is highly fragmented, however, indicating that,
although there is extensive DNA synthesis, DNA replication is
incomplete in some way (2). Strains with dna2 deletions are
inviable, showing Dna2 performs an essential function during DNA
replication (2-4). Recently, it has also been demonstrated that
dna2 mutants are defective in repair of x-ray-, bleomycin-, and methylmethane sulfonate-induced DNA damage (4, 5).
Dna2 is a 170-kDa protein with six motifs characteristic of DNA
helicases in the C-terminal third of the protein. A schematic diagram
of the protein is shown in Fig. 1. Genes homologous to DNA2
have been identified in Schizosaccharomyces pombe, Xenopus laevis, Caenorhabditis elegans, and humans (6-8).
Immunoaffinity-purified Dna2 has DNA-dependent ATPase
activity, DNA helicase activity that requires a 5' non-hybridized tail
adjacent to the duplex region unwound, and a potent endonuclease
activity (2, 9-11). The helicase and ATPase activities are required
for the essential function of Dna2 since mutation of the ATP binding
motif (K1080E) abolishes both ATPase and helicase and results in a gene
that does not complement either a dna2-1 or a
dna2
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
mutant (9). Other dna2 mutations that
reduce but do not abolish ATP binding and/or hydrolysis support growth
under some conditions, showing that the full helicase activity is not
essential for viability, and leading to the suggestion that it is the
nuclease activity that is essential (4). A recent analysis suggests
that Dna2 falls into the RecB class of helicase/nuclease proteins, with homology to the nuclease localized to a short motif in the N-terminal half of the protein, corresponding to the putative active site of RecB
nuclease (see Fig. 1).
View larger version (7K):
[in a new window]
Fig. 1.
Schematic diagram of the yeast
DNA2 gene.
Preliminary characterization of the nuclease activity associated with
immunoaffinity-purified Dna2 suggested that it was a so-called
structure-specific nuclease, in that it removed 5' single-stranded tails adjacent to a duplex region, but not 3' single-stranded tails and
did not digest duplex DNA (10). This suggested to us that Dna2 might
have an intrinsic nuclease and/or that it might copurify with yeast
flap endonuclease-1 (FEN-1), which has similar substrate specificity
(12, 13). Surprisingly, both interpretations appear to be correct.
FEN-1 is a 5' to 3' exonuclease that also functions as an endonuclease
on a 5' single-stranded flap structure adjacent to a duplex region,
cutting at (or near, depending on the context) the junction between the
single- and double-stranded region. The human FEN-1 homologue
participates in the maturation of Okazaki fragments synthesized during
an in vitro SV-40 DNA replication reaction (14-16). Yeast
FEN-1 is encoded by the RAD27 gene (17, 18), and the
phenotype of rad27 mutants suggests, albeit indirectly,
that the mutants are defective in Okazaki fragment maturation. For
instance, rad27 mutants are viable at 23 °C but not at
37 °C, cause a high level of mutagenesis, but are not defective in
repair (17, 18). rad27 mutants show an increased
frequency of duplications at repeated structures, which might occur due to faulty Okazaki fragment processing (19). rad27 mutants
accumulate expansions of di- and tri-nucleotide repeat DNA tracts,
which could be explained by fold-back of FLAP structures on Okazaki fragments at stalled replication forks (17, 20, 21). We have documented
strong genetic interactions between DNA2 and
RAD27. Overexpression of RAD27 complements
dna2-1 strains and overproduction of DNA2
suppresses the temperature-sensitive growth defect of rad27 strains. rad27 dna2-1
strains are inviable. Biochemical studies showed that the
affinity-purified Dna2 preparations contained FEN-1 and that FEN-1 and
Dna2 co-immunoprecipitated from yeast extracts. This genetic and
biochemical evidence led us to propose that Dna2, like FEN-1, functions
in Okazaki fragment maturation (10), but left the mechanism ill
defined. Dna2 could either substitute for FEN-1, or assist FEN-1 in its
function, or repair errors made by FEN-1. Overexpression of
RAD27 also suppresses this sensitivity of dna2
mutants to x-rays (5), suggesting again an interaction or functional
overlap between the two, although rad27 strains are
proficient in DSB2 repair
(17).
We next purified Dna2 from a rad27 strain, and we found
that a potent nuclease activity was still present, suggesting that, in
addition to associating with FEN-1, DNA2 encodes an integral nuclease (10). That Dna2 is itself a nuclease was strongly supported by
more extensive characterization of a nuclease associated with recombinant yeast Dna2 protein produced in insect cells (11). These
workers found that highly purified Dna2 cleaves single-stranded DNA
endonucleolytically, that it prefers 5' single-stranded tails to 3'
single-stranded tails, and that it has very low endonuclease activity
on limited stretches of single-stranded DNA flanked by two regions of
duplex (11). Thus, it has the specificity to tailor the 5' ends of
Okazaki fragments during their maturation into continuous DNA strands,
providing a biochemical basis for a role in Okazaki fragment
maturation. Although several helicase/nucleases are required for DNA
repair, Dna2 is the first replication protein that contains both functions.
Herein, we further characterize the Dna2 nuclease/helicase
biochemically and genetically. Our results highlight the fact that Dna2
is a vigorous endonuclease, as opposed to a strictly structure-specific nuclease, and that it is regulated by ATP. More important, we show by
mutations that eliminate nuclease activity but that leave helicase
intact, that the nuclease is integral to Dna2 and that the nuclease
domain is essential both for viability of yeast and for repair of x-ray
induced damage.
| |
EXPERIMENTAL PROCEDURES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
Strains Used--
Yeast strains used were: BJ5459 (MATa
ura3-52 trp1 lys2-801 leu2200 his3200 pep4:1-1153
prb11.6r) and 4X154-2D (MATa ura3-52 trp1
his3 leu2). Strain dna2-1 for complementation analysis
was as described (10). The strain used for the x-ray sensitivity study
(Fig. 6E) was MB-2-2-5G-6A (MATa trp1 leu2 ura3 his3
dna2-2::LEU2 sgs1-3::TRP1).
Oligonucleotides-- Oligonucleotides used for construction of mutants and helicase and nuclease assays were: MB55, GGAATGCCAGGGACTGGGGAAACTACTGTTATCGCAGA; MB87, GTTCTTCTGTGGCGTTCCAGGACCACCCAAGCTAGCGTAGTCTGGGACGTCGTATGGGTACATATGGACGATCTCTTCAATTG; MB95, AAATAATACATCGGAATTTAGCACCAACAGGTT; MB94, GAAGCTCTTCTTATTCCCCGGATCCTCAATGGTGATGGTGATGGTGACTTTCATACTCTTGTAGAAT; MB140, TATTGGAGACCAAATGTTCGCTGCCGCTGCAATCACATTGGATATAGA; hpr3, AGCTCTTGATCGTAGACGTTGTAAAACGACGGCCAGTG; hpr7, AGCTAGCTCTTGATCGTAGACGTTGTAAAACGACGGCCAGTG; hpr8, AGCTCTTGATCGTAGACGTTGTAAAACGACGGCCAGTGCCAAGC (44-mer, 30 nt complementary to M13mp18).
Plasmids--
The plasmid pB/S:DNA2 has the 6110-bp
EcoRI fragment containing the DNA2 gene cloned
into the EcoRI site of pB/S SK
. The next B/S plasmids were
created by site-directed mutagenesis with the above oligonucleotides.
The plasmid B/S:87 has the hemagglutinin (HA) epitope YPYDVPDYASLGGP
fused to oligonucleotide AA1 and was created by hybridizing pB/S:DNA2
with MB87. In addition an EcoRI site was inserted 3 nucleotides upstream of the ATG. The plasmid B/S:87.94 was constructed
by hybridizing B/S:87 with MB94, creating a DNA2 gene with a
HA epitope tag at the N terminus and 6-histidine tag at the C terminus.
The plasmid B/S:87.94.55 was created by hybridizing MB55 to B/S:87.94,
resulting in a K1080E mutation in the ATP binding motif. B/S:87.94.95
was created by hybridizing MB95 to B/S:87.94, resulting in a P504S
dna2-1 mutation. The plasmid B/S:87.94.140 was made by
hybridizing MB140 to B/S:87.94, resulting in a DIEE640AAA mutation. The
plasmid pGAL:DNA2 was created by cloning the 5.4-kb EcoRI
fragment from B/S:87.94 into the EcoRI site of pGAL18. The
plasmid pGAL:Dna2:K1080E was created by cloning the 5.4-kb fragment
from B/S:87.94.55 into the EcoRI site of pGAL18. The plasmid
pGAL:Dna2:E640A was created by cloning the 5.4-kb EcoRI site
from B/S:87.94.140 into the EcoRI site of pGAL18. The pGAL
plasmids containing the tagged D657A, E675A, and Y293A mutant Dna2
proteins were created exactly as pGAL:Dna2:E640A. Oligonucleotides used
for mutagenesis are available on request. The plasmid pGAL:Dna2:P504S was created by cloning the 5.4-kb EcoRI fragment from
B/S:87.94.95 into the EcoRI site of pGAL18.
The baculovirus expression vector was prepared by inserting the
full-length DNA2 gene in pB/S SK using the 6-kb
EcoRI insert from Ycp154-2 containing the DNA2
gene (2). A BamHI site was engineered at the ATG translation
start site of the DNA2 gene by site-directed mutagenesis
with oligonucleotides GATCGTCAGGGGATCCATGCCCGG after
purification of U-rich single-stranded phagemid as described in the
Muta-Gene in vitro mutagenesis kit (Bio-Rad). Full-length DNA2 was transferred to the baculovirus expression vector
pFASTbac HTb (Life Technologies, Inc.), containing a 6xHis tag, using
BamHI and XhoI to give pbacDNA2. This yields
a 6xHis tag at the N terminus of DNA2. The plasmid
pbacDNA2N, containing the N-terminal 2980 bp of DNA2 was
prepared by deletion of the PvuII-ShoI fragment within the
DNA2 gene. pbacDNA2C was prepared by cloning of the C-terminal 1719 bp of DNA2, prepared by polymerase chain
reaction, into pFASTbacHTb.
Purification of Dna2 from Baculovirus-injected SF9
Cells--
Yeast Dna2 was expressed in insect SF9 cells using the Bac
to Bac baculovirus expression system (Life Technologies, Inc.). pFast
Dna2 was introduced into Escherichia coli DH10BAC and placed on LB agar containing 50 µg/ml kanamycin, 7 µg/ml gentamycin, 10 µg/ml tetracyclin, 100 µg/ml Bluo-gal, and 40 µg/ml
isopropyl-1-thio-
-D-galactopyranoside as described in
the supplier's manual. White colonies were picked and grown in LB
medium containing 50 µg/ml kanamycin, 7 µg/ml gentamycin, 10 µg/ml tetracyclin, and recombinant bacmid DNA was prepared. Bacmids
were prepared and introduced into SF9 (or HI5) insect cells using
Cell-fectin reagent; recombinant virus were collected and amplified.
Cells were washed with cold Tris-buffered saline buffer and resuspended in lysis buffer (25 mM Tris-HCl, pH 7.5, 1 mM EDTA, 100 mM NaCl, 10% glycerol, and protease inhibitor mix). Cells were lysed by sonication. After cell extracts were cleared by centrifugation at 10,000 rpm at 4 °C, the supernatant was loaded onto a 5-ml heparin-Sepharose column. Dna2-containing fractions were collected with lysis buffer containing 600 mM NaCl after several washes with lysis buffer. The fraction was incubated with Ni2+/NTA-agarose) equilibrated with binding buffer for 2 h at 4 °C. After thorough washing with lysis buffer (100 column volumes), Dna2 was eluted with a gradient from 0 mM to 150 mM imidazole in lysis buffer. The Dna2-containing fraction was dialyzed in 10 volumes of lysis buffer.
The Dna2-containing fraction was loaded onto a 1-ml FPLC Mono Q column. After washing, Dna2 was eluted with a gradient of 100-600 mM NaCl in lysis buffer. Dna2-containing fractions (eluting at 270 mM NaCl) were pooled and loaded onto a 5-ml 15-40% glycerol gradient and centrifuged at 45,000 rpm in a SW 50.1 rotor for 24 h. Dna2 was identified at all purification steps by DNA-dependent ATPase assay and Western blotting with anti-Dna2 polyclonal antibody. A summary of the purification is shown in Table I.
Purification of Dna2 from Yeast-- The pGAL18 plasmids described above were transformed into BJ5459, and cells were grown and extracts were prepared as described previously (2). To purify Dna2 on the Ni2+/NTA-agarose column, the ammonium sulfate precipitated extract was dialyzed in TBSG, pH 8.0, 0.1% Triton X-100 (0.025 M Tris, pH 8.0, 0.15 M NaCl, 10% glycerol, 0.1% Triton X-100). Protein (100 mg) was loaded onto a 1-ml Ni2+/NTA-agarose column. The column was washed with 50 volumes of TBSG, pH 8.0, 0.1% Triton X-100. The column was further washed with 20 mM imidazole in TBSG, pH 8.0, 0.1% Triton X-100. His-tagged Dna2 was eluted with 200 mM imidazole in TBSG, pH 7.6, 0.1% Triton X-100. The protein was concentrated to 1 mg/ml and frozen. Approximately 100-fold purification was obtained.
Dna2 was immunoprecipitated by incubating 5 µg of Ni2+/NTA-agarose-purified Dna2 with 5 µg of 12CA5 antibody for 1.5 h at 4 °C. 10 µl of 10% protein A beads was added followed by a 1-h incubation. Beads were washed five times with TBSG, pH 7.6, 0.1% Triton X-100, 1 mg/ml BSA, then two times with 2× assay buffer. 2× ATPase assay buffer is 20% glycerol, 0.08 M Tris, pH 7.6, 0.01 MgCl2, 0.05 M NaCl, 0.002 M DTT, 1 mg/ml BSA. 2× helicase assay buffer is 20% glycerol, 0.08 M Tris-HCl, pH 7.3, 0.01 M MgCl2, 0.05 M NaCl, 0.002 M DTT, 1 mg/ml BSA. 12CA5 was coupled to CL-4B-agarose beads at a concentration of 10 mg/ml. When 12CA5-coupled beads were used, 5 µg of Ni2+/NTA-purified protein was incubated with 5 µl of 12CA5-coupled beads for 2 h. The protein was then washed as above. When Dna2 was purified from ammonium sulfate precipitates of crude extracts, 0.5 mg of protein was incubated with 10 µg of 12CA5 for 1.5 h, followed by the addition of 10 µl of 10% protein A beads. The substrate was added in a volume of 10 µl and incubated at 37 °C.
Helicase and ATPase Assays-- The helicase assay contained 4 mM ATP, 0.01 pM oligonucleotide hpr3 hybridized to M13. 2× helicase assay buffer is 20% glycerol, 0.08 M Tris-HCl, pH 7.3, 0.015 M MgCl2, 0.05 M NaCl, 0.002 M DTT, 8 mM ATP. The standard ATPase assay contained 0.1 mg/ml poly(dA) and 0.2 mM ATP at 25 µM/1 µCi.
Preparation of Nuclease Substrates and Nuclease Assays--
For
5'-labeled substrate, oligonucleotides were labeled at the 5' end with
32P using polynucleotide kinase. Labeled oligonucleotides
were hybridized to M13mp18 by heating to 65 °C for 5 min and
annealing by cooling to room temperature. Free oligonucleotides were
removed by gel filtration using Sepharose CL-4B. For 3'-labeled
substrate, the oligonucleotide was annealed to M13mp18 DNA and labeled
using Klenow enzyme and [
-32P]dTTP. Unincorporated
dTTP was removed by gel filtration.
Purified Dna2 was incubated in a 20-µl reaction (same as ATPase
reaction) containing 15 fmol of substrate at 37 °C for 15 min. For
denaturing gel analysis, after mixing with sequencing gel loading
buffer, samples were boiled and loaded onto a 20% sequencing gel. The
gel was run for 90 min at 20 W. Gels were dried and the nuclease
products were analyzed with the PhosphorImager (Molecular Dynamics).
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
Expression of DNA2 in Baculovirus-infected Insect Cells-- The DNA2 gene used in this study contains the full-length DNA2, including the 105-amino acid, N-terminal segment absent from the gene used in our previous studies (2). This region does not appear to perform an essential function and shares no homology with DNA2 genes from other organisms. However, the 105-amino acid region may have a role in transcriptional silencing, since overexpressing this portion of the protein leads to derepression of genes normally silenced at telomeres (22). In evaluating the essential functions of DNA2, it seems more prudent to use the full-length protein.
The DNA2 gene was tagged at the N terminus with 6 histidines, cloned, expressed in SF9 insect cells, and extensively purified as described under "Experimental Procedures" and Table I. In addition to the wild type protein, a K1080E mutant protein, a protein consisting of the 963 N-terminal amino acids (120 kDa), and a protein consisting of the 573 C-terminal amino acids (65 kDa) were expressed and purified. Similar levels of expression and purity were obtained with each of the mutant proteins (Fig. 2). To determine if Dna2, like other helicases, is multimeric, gel filtration analysis was carried out. The Dna2 protein is found primarily as a dimer (Fig. 3). In addition, there is a significant peak of tetramers (Fig. 3) and even some hexamers (not visible in the tracing in Fig. 3). Only the dimeric form of the Dna2 protein shows DNA-dependent ATPase activity, suggesting that the form of the protein active as an ATPase is a dimer (Fig. 3).
|
|
|
DNA-dependent ATPase and Helicase Activity--
As
shown in Fig. 4A, the ATPase
activity of the wild type protein requires single-stranded DNA. Either
poly(dA) or single-stranded circular M13 DNA is an effective cofactor.
Double-stranded DNA fails to stimulate the ATPase. The K1080E mutation
abolishes the ATPase activity (Fig. 4A), identifying the
Walker A box as part of the active site of the
DNA-dependent ATPase. The N-terminal 120-kDa protein and
the C-terminal 65-kDa protein both lack ATPase (Fig. 4A),
suggesting that the two domains may interact to activate the ATPase or
that the C-terminal fragment does not fold properly, even though it is
expressed at levels similar to the wild type (data not shown). The
Km for the ATPase activity was determined from the
data in Fig. 4B and is 165 µM. Since we have previously shown that 4 mM ATP is required for the DNA
helicase activity (2), the low Km was surprising.
Below, we reconcile this discrepancy by showing that much lower
concentrations of ATP (100 µM) are required for helicase
activity when the nuclease is inactivated by mutation. It is noteworthy
that the specific activity of the ATPase associated with yeast Dna2 is
almost 100 fold higher than that found in X. laevis Dna2
(8), suggesting that there may be some differences between the enzymes
from different species.
|
As reported previously for yeast Dna2 expressed in insect cells and for
Xenopus Dna2 expressed in insect cells (8, 11), at the
levels of protein used here, there was no detectable helicase activity
in our preparations on any of the substrates used to assay the activity
of the Dna2 protein purified from yeast (for example Fig.
5A). The difficulty of
detecting helicase in recombinant Dna2 is further addressed under
"Discussion" but is not the subject of this paper, which focuses on
the nuclease activity.
|
Nuclease Activity of Dna2-- The Dna2 preparations are active as nucleases. To further characterize the substrates and products of the Dna2 nuclease activity, we chose substrates, labeling configurations, and reaction conditions not investigated previously with the recombinant yeast Dna2 protein from insect cells (11). In addition, we compared the wild type Dna2 protein with the K1080E mutant protein. First, we used the helicase substrate, an M13 single-stranded DNA circle hybridized to a 38-nt oligonucleotide with a 14-base noncomplementary 5' tail (2). The 38-mer was labeled at the 5' terminus with 32P. The digestion products were analyzed on a native gel. As shown in Fig. 5A (lanes 2, 4, 7, 9, 11, and 12), the major labeled product of the nuclease was 10 nt, with some smaller products, suggesting a preference for cutting this substrate 4 nt away from the junction of duplex and single-stranded DNA (but see below). Lanes 1 and 6 are controls with protein from mock-infected insect cells carried through the same purification procedure. The K1080E mutant was as active as the wild type protein (compare lanes 2 and 4, 3 and 5, 7 and 9, 11 and 12). When a substrate with an 18-nt 5' tail was used, the major product was 14 nt (data not shown). These results could suggest that the enzyme monitors distance from the fork. A similar spectrum of products is seen with the enzyme purified from yeast (see Fig. 10).
Since we were concerned that the nuclease might mask the helicase activity by destroying the substrate, we searched for conditions that might inhibit the nuclease preferentially. As shown in Fig. 5A, a combination of 2 mM ATP and 2 mM Ca2+ completely inhibited the nuclease (lanes 3 and 5), yet helicase activity still was not seen. The boiled substrate is shown in lane 13 to indicate where the product of the helicase would appear. Comparison of lanes 3 and 5 in Fig. 5A, which contain both ATP and Ca2+, to lanes 8 and 10, which contain only Ca2+, shows that Ca2+ alone was less inhibitory than Ca2+ plus ATP. (Lanes 11 and 12 of Fig. 5A are similar to lanes 3 and 5, but contain only 1 mM Ca2+.) Comparison of lane 2 (with ATP) with lane 7 (no ATP) shows that ATP is inhibitory by itself. Note the increase in smaller products in lane 7. We have consistently observed that ATP is inhibitory to the nuclease activity of wild type Dna2 (2, 9), and see below. ATP inhibition is not a result of titration of Mg2+, since 1 mM excess Mg2+ was added in the reactions containing ATP and no Ca2+.
In Fig. 5B, we show the labeled products of the reaction when Dna2 is incubated with the same configuration of substrate but with the radioactive label at the 3' terminus of the oligonucleotide, which is in a duplex structure. In this experiment, the duplex region is 30 bp long and the tail is 14 nt. As opposed to the native gel shown in Fig. 5A, a denaturing polyacrylamide gel was used in order to detect potential cutting in the duplex portion of the substrate. The major product, as revealed after denaturation, is clearly 30 nt in length, with minor products of 29 and 31 nt. The size of the products labeled at the 3' end suggests that there is significant cutting at the junction of the single-stranded 5'tail and duplex DNA, that is at the fork. This activity is due to Dna2 protein since it is inhibited by Dna2 antibody, shown in Fig. 5C. Taking the results of Fig. 5, panels A and B, together, we infer that the endonuclease cleaves at the junction, releasing the single-stranded 5' product, which may then be further degraded. The smaller products are not detectable when label is only at the 3' end. The 30-nt duplex, by contrast, is stable to further digestion by Dna2, but detectable only when label is at the 3' end. Comparison of the amount of degradation in Fig. 5 (A and B) suggests that the enzyme prefers the single-stranded tail to the junction, although it is difficult to quantify the difference in preference. It is clear that Dna2, like FEN-1, can cleave a flap structure. However, Dna2, unlike FEN-1 (13), has a potent activity as a single-stranded endonuclease, yielding products of 10 nt in length and smaller. Thus, the specificities of Dna2 and FEN-1 are overlapping but not identical.
To verify that Dna2 cleaves single-stranded DNA, the 42-nt oligonucleotide, in the absence of M13, was labeled at the 5' end. After incubation with enzyme, the major products were 10 nucleotides or less (data not shown). Completely single-stranded circular DNA was then used to determine whether the enzyme had any preference for ends, as might be found on immature Okazaki fragments. When M13 circular DNA was used as substrate, it was cleaved to linear, full-length DNA (Fig. 5D). Over 60% of the substrate was converted to linear DNA before any fragments smaller than unit length were observed. This is the same pattern of cleavage observed with the RecB nuclease, and suggests that on long DNA molecules Dna2 has no preference for DNA ends over internal sites and that the enzyme is distributive under these assay conditions. This does not rule out that the endonuclease may prefer short tails adjacent to short single-stranded regions within duplex DNA, where the single-stranded DNA may be too short for efficient binding of Dna2 protein. Other studies suggest that single-stranded tails are preferred to single-stranded regions flanked by duplex DNA when oligonucleotides are used as substrates (11).
Mutations Affecting the Nuclease Activity of Dna2 and Their Effects
on Growth of Yeast and Repair of Double-strand Breaks--
Two lines
of evidence suggest but do not prove that Dna2 encodes the nuclease
activity. First, the nuclease activity copurifies with Dna2 from a
rad27 mutant and with Dna2 from insect cells expressing
only yeast Dna2 and not yeast FEN-1. Second, the substrate specificity
of FEN-1 differs from that of Dna2. To obtain further evidence that the
nuclease is intrinsic to Dna2, to examine the location of the nuclease
domain in the protein, and to determine if the activity is part of the
essential function of Dna2, we have made mutations that target the
nuclease. Since at the time this work was begun, we could detect no
homology within Dna2 with any known nuclease motifs, our rationale was
as follows. The nuclease is probably associated with the N terminus,
since helicase occupies the C terminus. Therefore, the first mutation
we introduced into the full-length Dna2 protein is the N-terminal
dna2-1 mutation, P504S, originally identified by us (9).
This amino acid falls in a region strictly conserved among all Dna2
orthologs. The second mutation was designed to target putative nuclease
catalytic residues. Dna2 is a Mg2+-dependent
nuclease. Metal catalyzed nuclease reactions often require aspartic and
glutamic acids as ligands for Mg2+. In the Exo1 domain of
E. coli DNA polymerase I, there are two aspartic acid
residues in a row and these are conserved in Exo domains from other
bacterial and eukaryotic DNA polymerases with 3' to 5' exonuclease
activity (23). Altering the nuclease activity of Dna2 might involve
mutating two or more conserved aspartic or glutamic amino acid
residues. The residues DIEE beginning at amino acid 640 are conserved
among human, C. elegans, and X. laevis Dna2 and
were therefore deemed candidates for essential catalytic residues.
These amino acids were changed to AAAA by site-directed mutagenesis as
described under "Experimental Procedures."
Since the time that we constructed the latter mutant, a study has appeared proposing that a region 11 amino acids downstream of amino acid 640 is homologous to the putative nuclease active site of the RecB nuclease family (24). Three additional mutations were then introduced on the basis of the conservation with RecB nuclease. Asp657, Glu675, and Tyr693 were changed to alanine. Mutation of the amino acid in RecB equivalent to Dna2 Glu675 eliminates the nuclease activity of RecB protein, D1080A (24, 25).
A sixth mutation, K1080E, was introduced into the Walker ATP binding loop. We have already described both the catalytic and functional properties affected by this mutation in the N-terminally 105 amino acid truncated form of Dna2. However, since we had not characterized the full-length Dna2 protein containing this change, the new K1080E mutant serves as an important control (2).
The wild type and six mutant genes, P504S, E640A, D657A, E675A, Y690A,
and K1080E, tagged with 6xHis at the C terminus and with the HA epitope
at the N terminus, were cloned into the multicopy vector pGAL18
downstream of the galactose-inducible, glucose-repressible yeast
GAL10 promoter. As illustrated in Fig.
6A, the DNA2 gene complemented the dna2-1 strain even without induction
(glucose plate). We conclude that when Dna2 is cloned on a multicopy
plasmid there is sufficient expression for complementation of
dna2-1 even under glucose repression. None of the remaining
genes, E640A, P504S, D657A, E675A, Y640A, or K1080E, complemented the
dna2-1 mutant at 37 °C on glucose plates (Fig. 6,
A and E).
|
When the various DNA2 genes were induced with galactose (Fig. 6B), both the wild type and the P504S protein were able to restore growth to the dna2-1 mutant strain at 37 °C, whereas the K1080E, E640A, D657A, E675A, and Y693A mutants were unable to complement. (Only K1080E and E640A are shown here, but see Fig. 6F.) Complementation by high levels of mutant dna2-1 protein indicates that there is some residual active dna2-1 protein at 37 °C, and simply increasing the amount allows cell growth. Since none of the other proteins complemented under any conditions, these are likely loss of function mutations.
In addition to being loss of function mutations, the E640A, D657A,
E675A, and Y693A mutations are dominant negative when the proteins are
overproduced (Fig. 6, C, D, and F).
Overexpression of the E640A protein at 37 °C in a wild type strain
(Fig. 6C) and at the permissive temperature in
dna2-1 strains (Fig. 6D) is lethal. The relative
plating efficiency on galactose versus glucose for the
dna2-1 strain expressing the E640A mutant from the
GAL promoter was quantitated (Table
II) and found to be 3 × 10
4, whereas the relative plating efficiency of plasmid
free dna2-1 cells or dna2-1 cells carrying
plasmids expressing the P504S and K1080E mutants was unity (Table II).
The D657A, E675A, and Y693A mutations are also lethal in
dna2-1 plated on galactose at the permissive temperature
(Fig. 6F). The ability to function as dominant negatives can
be explained if the mutant proteins can assemble in replication
complexes but are inactive.
|
Sensitivity of Mutants to X-rays--
dna2-1 and
dna2-2 mutants are sensitive to x-rays and to methylmethane
sulfonate (4, 5). Since this x-ray sensitivity is suppressed by
overproduction of FEN-1 (5), and since the nuclease activity of Dna2 is
similar to that of FEN-1, we tested the ability of the new mutants to
restore x-ray resistance to dna2-2 mutants. The wild type
but not K1080E or E640A rescued the x-ray sensitivity of
dna2-2. Thus, as shown in Fig.
7, the activities affected by mutations
in both the C-terminal ATP domain and the N-terminal domain are
required for efficient repair of x-ray damage. As shown under
"Experimental Procedures," the dna2-2 strain used here
also contained a mutation in an additional helicase, sgs1. However, sgs1 mutants are not
sensitive to x-rays (26).
|
Expression of Mutant Proteins in Yeast and Definition of the
Enzymatic Activities of the Mutant Proteins--
The DNA2
genes were expressed in yeast, rather than in insect cells, so that the
helicase activity could also be monitored. To allow for two affinity
purification steps, the Dna2 proteins were fused to a 6xHis tag at the
C terminus and an HA tag at the N terminus as described under
"Experimental Procedures." As illustrated in Western blots shown in
Fig. 8D, the wild type, E640A,
P504S, and K1080E proteins were expressed and recovered after
purification at approximately equal levels. The other proteins were
expressed at similar levels (data not shown).
|
DNA-dependent ATPase-- Ni2+/NTA affinity-purified proteins (see Experimental Procedures) were assayed for DNA-stimulated ATPase at 23 °Cand 37 °C. As shown in Fig. 8A, the Dna2 and E640A proteins have nearly equivalent levels of ATPase. The K1080E protein, as expected, is defective (9). The K1080E curve serves as control to demonstrate that Dna2 is the only ATPase in the purified fractions. What was surprising is that the activity of the P504S (the dna2-1 mutation) protein is only slightly greater than the K1080E protein, since the P504S mutation does not map to the predicted ATPase domain. The P504S protein does not appear to be temperature-sensitive, but rather is defective at all temperatures. The reduced ATPase activity of the N-terminally located mutation suggests that the N terminus may be a positive effector of the ATPase activity of Dna2p, perhaps a DNA binding domain. Although global unfolding cannot be ruled out, it is worth noting that the P504S protein is active in vivo when overproduced, suggesting that at least in vivo the residual activity is sufficient for function.
To demonstrate that the ATPase of the P504S mutant was partially active, Ni2+/NTA-purified Dna2 and the P504S protein were immunoprecipitated with CL-4B beads coupled with 12CA5 in order to further purify and concentrate the protein. The additional purification step removed the small fraction of non-DNA-dependent ATPases from the Dna2 so that longer incubation times and higher specific activity ATP could be used in the assay. When this preparation was assayed, and the ATP concentration was reduced to 50 µM, the P504S protein now showed approximately 50% the extent and rate of ATPase activity as wild type, whereas the K1080E mutant remained inactive (Fig. 8C).
As shown in Fig. 8B, the Glu675 and
Tyr693 mutants show similar, though slightly reduced ATPase
activity compared with this preparation of wild type. The
Asp657 mutant shows a lower rate and extent, about 40% of
wild type. The significance of these small differences is not clear
since the helicase of these mutants does not appear different from wild type (see Figs. 9 and
10). There could be variations in
recovery of the proteins that prevent accurate quantitation within a
2-fold range.
|
|
DNA Helicase and Nuclease-- In order to test the mutant proteins for the helicase and nuclease activities, Dna2 proteins were purified from the Ni2+/NTA column and immunoprecipitated with 12CA5 antibody. Note that these purification steps are different from those used previously and that two affinity purification steps are employed rather than one (2, 9). As shown in Fig. 9 (lane 5), the reaction containing Dna2 and ATP yields three bands: a major one of 38 nt, a minor product of approximately 34 nt, and a major product consisting of oligonucleotides of products of approximately 10 nt. We propose that the 38-mer is due to helicase activity acting at the fork between the 5' tail and the duplex region. Controls show that in the absence of Dna2 protein, no helicase is observed (Fig. 9, lane 12). In the presence of Dna2 and absence of ATP, there is also no helicase activity (Fig. 9, lane 4, no 38-nt band).
We propose that the 34-mer is due to helicase at the 3' end of the duplex followed by nucleolytic degradation of the exposed 3' single-stranded region, since it is not seen in the minus ATP reaction (Fig. 9, lane 4). This activity was suppressed in our former work by the introduction of a thiophosphate ester in the 3' terminal phosphodiester bond of the oligonucleotide (2), but was reported by Bae et al. (11).
The much more abundant labeled product migrating with oligonucleotides of 14 nt or less in length indicates that Dna2 nuclease activity cuts the 5' flap region in the presence of ATP. This nuclease activity is not only ATP-independent (lane 4), it is actually inhibited by ATP (compare lanes 4 and 5). In the mock immunoprecipitate, that is, in the absence of 12CA5, a small amount of helicase and nuclease activity is observed, but much less than in the presence of antibody. It is likely a small amount of protein precipitates in the presence of high concentrations of BSA that are needed to stabilize the ATPase activity and helicase activity in the purified fraction. BSA is not needed when crude Dna2 is immunoprecipitated, probably because other proteins in the extract serve to stabilize Dna2 (data not shown).
The E640A mutant retains significant helicase activity in the presence of ATP (Fig. 9, lane 7, 38-nt band). The helicase is approximately 60% as active as wild type, estimated from three experiments, consistent with the estimated reduction in ATPase of the E640A protein compared with wild type (see Fig. 8). The E640A change has a much greater effect on the nuclease activity than on the helicase and ATPase. The E640A mutant shows less than 15% of the amount of endonuclease as wild type in the absence of ATP (lane 6) and has no detectable nuclease activity in the presence of ATP (lane 7). This is even more clear evidence than that presented in Fig. 5 with the insect cell preparation of Dna2, that the nuclease is regulated by ATP.
Nuclease Activity of Mutations Affecting the RecB Homology Domain-- Mutations in the residues conserved in RecB nuclease have a greater effect on the nuclease than the E640A change. D657A and E675A show no detectable nuclease activity either in the absence or in the presence of ATP (Fig. 10A). In these assays we have used a more sensitive denaturing gel nuclease assay than was used in the experiment in Fig. 9 to refine the analysis of the products. Products range from 14 nt down to a few nt. The Y693A mutant protein shows some, although reduced, nuclease activity in the absence of ATP, but no activity in the presence of ATP (Fig. 10A), similar to E640A (Fig. 9). Overexposure of the Dna2 protein lane (panel on the right) shows that wild type nuclease is inhibited by ATP but also shows that there is some ATP-dependent cutting (faint bands between 32 and 14 nucleotides in length). We propose that the ATP binding domain can modulate the activity of the nuclease, since nuclease is always reduced or altered in the presence of ATP, and is completely abolished by ATP in the E640A and Y693A mutants.
Helicase Activity Is Intact in All Four Nuclease Mutants-- An important point is that the E640A, D657A, E675A, and Y693A mutant proteins, although deficient in nuclease, are all active helicases (Fig. 10B). Thus, the inability of these mutants to support growth (Fig. 6, A-F, Table II) is clearly due to a nuclease defect. In addition, the inability of the E640A mutant to complement the x-ray repair defect in a dna2-2 mutant is also likely due to the nuclease deficiency (Fig. 7). This ability to selectively inactivate nuclease but not helicase also shows that the helicase activity of Dna2 on forked molecules is not dependent on the nuclease. We are aware in drawing these conclusions that there is a small reduction in DNA-dependent ATPase in some of these mutants (Fig. 8). However, others have provided convincing evidence that full ATPase activity is not required for viability (4).
The availability of the E675A mutant allowed us to reinvestigate the Km for ATP of the Dna2 helicase activity in the absence of nuclease. Using the assay conditions described in Fig. 10B, we determined a Km of between 70 and 100 µM, similar to that for ATPase. The 4 mM ATP Km described previously for the wild type enzyme might be required to inhibit the nuclease so that helicase can be observed (2).
Helicase Is Deficient in the P504S and K1080E Mutants-- The Dna2 P504S protein, which gives a temperature-sensitive in vivo phenotype, retains helicase activity, although like its ATPase activity, the helicase activity is significantly reduced compared with wild type and E640A (Fig. 9). The P504S mutant also retains nuclease activity, although it is reduced by about 95% compared with wild type (Fig. 9, compare lanes 8 and 4). The effect of this mutation is different from that of the E640A and Y693 mutations, since the nuclease activity is not completely abolished in the presence of ATP, although it may be slightly further reduced (lane 9). Since the P504S mutation is temperature-sensitive, the mutation may destroy the conformation of the protein, rather than identifying amino acids important for one or the other of the activities of Dna2 protein.
For the K1080E protein, helicase is not observed but nuclease is
retained at high levels in the presence of ATP, consistent with our
previous results (9). The Dna2 K1080E mutant retains full nuclease
activity (compared with wild type) in the absence of ATP, but unlike
the other Dna2 proteins, the K1080E nuclease activity appears to be
stimulated by ATP. We do not know if there is another ATP binding site
in this protein, but this stimulation is reproducible.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
Since Dna2 is a multifunctional enzyme, it has been important to determine which of the enzymatic activities is required for the functions of the protein in the cell. We have previously shown that the ATPase, and presumably the helicase, is essential. We have now shown by generating separation of function mutations that the associated single-strand specific endonuclease activity is integral to the Dna2 protein and that it is also essential for the viability of yeast. Mutations in a conserved region of the N terminus abolished nuclease activity and only helicase activity remained. Changing either of two amino acids in the RecB homology region abolished nuclease in both the presence and absence of ATP, suggesting that the homology has functional significance. The nuclease mutants failed to complement the growth defect of a dna2-1 mutant and showed a dominant negative affect on growth. This suggests that the four mutant proteins can associate with the essential replication complexes in vivo. Thus, we propose that it is the impaired enzymatic activity that is giving rise to repair and replication phenotypes rather than a defect in protein/protein interactions. Although it has previously been proposed that only the helicase is essential for Dna2-mediated repair of damage due to double-strand breaks (4), the nuclease defective E640A mutant is unable to rescue the x-ray repair defect of dna2-2 mutants.
The nuclease active site of Dna2 appears to map to the N-terminal part of the protein in the RecB homology domain and can be separated from the ATPase domain, in the sense that nuclease-inactivating mutations do not appear to diminish ATPase activity. It has been shown previously that mutations affecting the ATPase domain lie in the C terminus of the protein that is also essential for viability. Nevertheless, the ATPase and helicase activities of the N-terminal mutant, P504S, are reduced, suggesting that the N terminus and the C-terminal ATPase domain may somehow interact. This conclusion is also supported by the inhibitory effect of ATP on the nuclease activity in wild type Dna2 and two of the N-terminal mutants, as well as the ability of ATP to stimulate the nuclease of the P504S mutant. A summary of the phenotypes of the mutants and the activities of the mutant proteins studied here is provided in Table III.
|
FEN-1 prefers to cleave a 5' tail adjacent to a duplex region at bases distinct from the single-strand double-strand junction when the region 5' to the flap structure is single-stranded, although on other substrates it cleaves at the junction (12, 13). Our results show that Dna2 shares this specificity, and this might explain how the two proteins could compensate for each other in DNA replication and double-strand break repair. FEN-1, however, does not cut single-stranded oligonucleotides, whereas Dna2 does cut single-stranded DNA (12, 13, 27). Thus, Dna2 has both structure-specific activities and simple endonuclease activity, whereas FEN-1 is structure-specific. The in vivo substrates of the two proteins may overlap, since all single-stranded DNA in the cell is linked to double-stranded DNA.
Although DNA helicase activity is observed with Dna2 expressed in S. cerevisiae, helicase is weak or unobservable in baculovirus expressed yeast Dna2 or baculovirus expressed X. laevis Dna2 (8, 11). Nevertheless, the observation of helicase at high levels of recombinant protein (11), the absence of DNA helicase in the K1080E protein (expressed in yeast), as well as the association of helicase with six other mutant forms of Dna2 (see Figs. 9 and 10B) and under two different purification regimes continue to suggest that a helicase activity is intrinsic to Dna2 and not the result of a co-migrating protein that is removed by the K1080E mutation. Proof of this point will require finding conditions that restore more efficient helicase activity to the Dna2 produced in insect cells. Dna2 may require additional proteins, present only in yeast, for optimal helicase activity. Given the inhibition of nuclease activity by ATP and by analogy to RecBC helicase/nuclease, we favor a model in which there may be an inhibitor of the nuclease associated with Dna2 purified from yeast that allows detection of the helicase. For RecBC helicase/nuclease, an inhibitor of the nuclease converts the enzyme from a nuclease into an active helicase. However, there is no evidence for this with Dna2 as yet.
Another possibility is that Dna2 requires a yeast-specific modification for helicase activity. G1-specific cyclin-dependent kinase (28) and Tor1 DNA-dependent protein kinase (3) both show genetic interactions with Dna2 and therefore are candidates for Dna2 regulators. Yet a third protein kinase, casein kinase 1, Hrr25, which regulates DNA double-strand break repair (29, 30), may also modify Dna2, which we have shown participates in double-strand break repair (5). In S. pombe, overexpression of Dna2 and Hhp1, a homolog of Hrr25, suppresses the temperature-sensitive phenotype of cdc24 (discussed below) (31).
Role of the Dna2 Helicase/Nuclease in Okazaki Fragment Metabolism: Direct or Indirect?-- DNA synthesized at the restrictive temperature in dna2-1 strains is severely fragmented (2). This and the genetic and physical interactions with FEN-1 originally led us to propose that Dna2 functions in Okazaki fragment processing (10). To join nascent fragments, the RNA-containing primer must be removed to prepare for ligation of DNA into continuous chains. The specificity of the Dna2 enzymatic activities can be at least partially reconciled with a role in primer removal. Dna2 can remove a 5'-flap structure, as we have shown here. Dna2 prefers a single-stranded tail adjacent to a duplex region and is less active on a substrate with short stretches of single-stranded DNA flanked by duplex DNA (11). An unwinding event, however, is required to produce a substrate for the flap endonuclease, be it Dna2 or FEN-1 or one of the many other endonucleases found in the yeast genome. Dna2 possesses (or copurifies with) a helicase activity that can unwind a short segment of duplex DNA (24 nucleotides) at a forked junction, but Dna2 helicase cannot unwind fully duplex substrates (2). It is therefore not clear how Dna2 could process the 5' end of an Okazaki fragment, unless one invokes displacement of the RNA by the oncoming polymerase, as has been proposed by Murante et al. (27, 32, 33).
Indirect Role for Dna2 in Lagging Strand Events--
Several
observations make a direct role in Okazaki fragment processing more
difficult to support. The potent single-stranded endonuclease
associated with Dna2 would appear to be antagonistic to lagging strand
integrity. Second, rad27exo1 mutants and
rad27mre11 mutants are, like rad27dna2
mutants, synthetically lethal. Exo1 is a 5' to 3' double-stranded
exonuclease that interacts with Msh2 in two hybrid assays (34, 35).
Mrell is also a nuclease. Thus, either of these might be a backup for
rad27. Overexpression of Exo1 did not suppress the
temperature-sensitive phenotype of dna2-1 mutants and
dna2-1exo1 double mutants are not synthetically lethal.3 If Exo1 processes
Okazaki fragments in the absence of FEN-1, then what is the role of
Dna2? We must more seriously examine the idea that Dna2 may play a
role, which is essential in every cell cycle, in correcting errors made
by FEN-1. Failure of FEN-1 to properly process Okazaki fragments may
lead to broken replication forks, and the function of Dna2 may involve
repairing such collapsed forks. rad27 rad52 strains
are inviable as a likely consequence of the inability of cells to
repair rad27-dependent errors by recombination
(19). Broken replication forks (and double-strand breaks) are, at least
in some cases, repaired by a mechanism that involves the assembly of
new replisomes in an origin-independent reaction. In bacteria, an
enzyme with activities similar to Dna2, the PriA 3' to 5' helicase, is
involved in assembling the new primosome (36, 37). This may occur at
collapsed replication forks, where the replisome has disassembled or
where double-strand breaks occur as forks encounter DNA damage. In
yeast, DNA polymerase , primase, and DNA polymerase 
are also
involved in recombinational repair of double-strand breaks (38). This
may also be related to the break-induced replication described by the
latter workers (39). It is interesting to note that dna2
mutants are sensitive to agents that cause double-strand breaks, such
as x-rays and bleomycin, and that this sensitivity is suppressed by
overproduction of FEN-1 (5). Overproduction of FEN-1 may simply allow
it to function more efficiently, reducing the need for Dna2 on the
lagging strand and freeing it for repair of the exogenously induced
damage. dna2 mutants also have a hyperrecombination
phenotype, suggesting double-strand breaks are not being efficiently
repaired leading to increased recombination (3). Support for a role for
Dna2 in repair of broken replication forks also is found in other
organisms. S. pombe cdc24 has been implicated in repair of
collapsed replication forks. Arrest of cdc24-G1
strains at the restrictive temperature is followed by loss in viability
and chromosome fragmentation (7). The fragmented chromosomes likely
arise from unrepaired broken replication forks. cdc24
mutants arrest at the restrictive temperature with 2C DNA content (31).
The temperature-sensitive growth defect of
cdc24-G1 mutant is suppressed by overproduction of DNA2 (7), which could be allowing repair of the
fragmented chromosomes present in cdc24-G1 cells
at 37 °C. Another allele of cdc24, cdc24-M38,
is suppressed by overproduction of proliferating cell nuclear antigen
(PCNA) and replication factor C, establishing a link to the replication
apparatus (31).
Other Helicase/Nucleases Involved in DNA Replication and Repair-- E. coli RecB contains nuclease and helicase encoded in the same polypeptide (40 for review) and is required for recombination-dependent replication in E. coli and possibly for repair of DSBs (41). The RecBCD helicase/nuclease is a powerful helicase that unwinds double-stranded DNA at the rate of 1000 bp/s, creating a single-stranded DNA as a substrate for the nuclease. In E. coli, broken replication forks can be repaired in a reaction requiring recA, recBCD, priA, dnaB, dnaG, and Ssb and the PolIII holoenzyme (reviewed in Ref. 42). The initial step of repair is catalyzed by RecBCD, which produces a 3' single-stranded tail for invasion of another duplex. In E. coli inhibition of DNA replication results in significant DNA degradation, which is recB-dependent (43). Strand invasion is initiated by the helicase activity of RecB when the nuclease is inactivated after encountering a Chi site in the DNA or induction of an inhibitor by the SOS response (44). Dna2 may be involved in such strand invasion. Alternatively, it may be more like priA and be creating the entry site for the replicative helicase and the rest of the replisome (41).
The human Werner syndrome protein is also a helicase with intrinsic
nuclease activity (45, 46). The purified protein has a 3' to 5'
exonuclease activity and 3' to 5' DNA helicase activity. A direct
involvement in DNA replication is suggested by the observation that S
phase is lengthened in Werner cells (47). Werner helicase is homologous
to S. cerevisiae sgs1 and E. coli
recQ. Sgs1p is a 3' to 5' helicase and probably creates a
substrate for Top3, perhaps in termination of DNA replication (48).
Recently, Sgs1 has been shown to be essential for DNA replication when
the srs2 gene, which also encodes a helicase, is also
deleted (49). Both sgs1 and dna2 mutants have
hyperrecombination and chromosomal instability phenotypes. Unlike Dna2,
Sgs1 apparently lacks nuclease activity, and sgs1-ts srs2
mutants fail to synthesize any DNA at the restrictive temperature,
suggesting that Sgs1 and Srs2 helicases are directly involved in moving
the replication fork. In summary, three proteins that function as a
helicase and nuclease, Dna2, RecBCD, and Werner protein, appear to be
required for processing intermediates in replication, which can lead to
deletions, amplifications, translocations, and chromosome loss if not resolved.
| |
FOOTNOTES |
|---|
* This work was supported by United States Public Health Service Grants GM25508 and NSF 9507352.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
To whom correspondence should be addressed: Braun Laboratory
147-75, California Institute of Technology, Pasadena, CA 91125. Tel.:
626-395-6053; Fax: 626-405-9452; E-mail:
jcampbel@cco.caltech.edu.
Published, JBC Papers in Press, March 23, 2000, DOI 10.1074/jbc.M909511199
1 M. E. Budd, L. M. Hoopes, and J. L. Campbell, unpublished results.
3 M. E. Budd, W.-C. Choe, and J. L. Campbell, unpublished results.
| |
ABBREVIATIONS |
|---|
The abbreviations used are: DSB, double-strand break; HA, hemagglutinin; 12CA5, anti-HA monoclonal antibody; nt, nucleotide(s); kb, kilobase pair(s); bp, base pair(s); BSA, bovine serum albumin; DTT, dithiothreitol; NTA, nitrilotriacetic acid; TBSG, Tris-buffered saline plus glycerol.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
< |
|---|
, Dna2; 
, ATPase activity.
