 1998 Oxford University Press
Nucleic Acids Research, 1998, Vol. 26, No. 5
1173–1178
ATP-dependent interaction of human mismatch repair
proteins and dual role of PCNA in mismatch repair
Liya Gu1, Yu Hong1,+, Scott McCulloch2, Hiroyuki Watanabe3 and Guo-Min Li1,2,*
1Department
of Pathology and Laboratory Medicine, Lucille P.Markey Cancer Center, 2Graduate Center for Toxicology,
University of Kentucky, Lexington, KY 40536, USA and 3Department of Biochemistry and Biophysics,
University of North Carolina, Chapel Hill, NC 27599, USA
Received December 11, 1997; Accepted January 7, 1998
ABSTRACT
DNA mismatch repair ensures genomic stability by
correcting biosynthetic errors and by blocking homologous recombination. MutS-like and MutL-like proteins
play important roles in these processes. In Escherichia
coli and yeast these two types of proteins form a repair
initiation complex that binds to mismatched DNA.
However, whether human MutS and MutL homologs
interact to form a complex has not been elucidated.
Using immunoprecipitation and Western blot analysis
we show here that human MSH2, MLH1, PMS2 and
proliferating cell nuclear antigen (PCNA) can be
co-immunoprecipitated, suggesting formation of a
repair initiation complex among these proteins.
Formation of the initiation complex is dependent on
ATP hydrolysis and at least functional MSH2 and MLH1
proteins, because the complex could not be detected
in tumor cells that produce truncated MLH1 or MSH2
protein. We also demonstrate that PCNA is required in
human mismatch repair not only at the step of repair
initiation, but also at the step of repair DNA re-synthesis.
INTRODUCTION
DNA mismatch repair is an important mutation avoidance
pathway and defects in this pathway lead to genomic instability
and cancer predisposition (1–3). Escherichia coli methyl-directed
mismatch repair ensures genomic stability by correcting mispairs
generated from DNA biosynthetic errors and by blocking
homologous recombination (4). The E.coli repair pathway
involves mismatch recognition by MutS, followed by binding of
MutL to the DNA–MutS complex to form a large repair initiation
complex. Formation of this complex activates a latent endonuclease
activity associated with MutH, which incises the newly synthesized
DNA strand at a hemimethylated d(GATC) site to provide the
strand discrimination signal. After the ‘wrong’ base is removed
by exonucleases (Exo I, Exo VII or RecJ) repair DNA re-synthesis
is conducted by DNA polymerase III holoenzyme in concert with
single-stranded DNA binding protein and DNA ligase (for a
review see 4).
Human cells possess a homologous, but more complicated,
strand-specific mismatch repair pathway. Although the signal for
strand discrimination in vivo is unknown, strand-specific mismatch
repair in human cells can be directed by a strand break in the DNA
substrate (5,6). Six human mismatch repair genes (MSH2, MSH3,
MSH6, MLH1, PMS1 and PMS2) have been identified and their
products account for only two E.coli equivalents: MutS and MutL
(for reviews see 1–3). Biochemical studies have demonstrated
that functional human MutS and MutL homologs are heterodimers.
The MSH2 and MSH6 gene products compose a heterodimer
designated hMutSα that recognizes single base-base mismatches
and small insertion/deletion mispairs (7,8). The MSH2 and MSH3
gene products form a heterodimer recognizing insertion/deletion
mispairs (9–12). A human MutL complex, designated hMutLα,
consists of MLH1 and PMS2 and was isolated by virtue of its
ability to restore mismatch repair to nuclear extracts of a
MLH1-deficient colon tumor line (13). Proliferating cell nuclear
antigen (PCNA) has been shown to participate in mismatch repair
(14). When HeLa nuclear extracts were pre-incubated with p21
(also known as Cip1, WAF1 and Sdi1), a universal cyclin-dependent
kinase inhibitor and PCNA-interacting protein (15–18), strandspecific mismatch repair was inhibited at an early step prior to
excision (14). This inhibition can be reversed by addition of
exogenous PCNA (14). DNA polymerase δ has also been
implicated in human mismatch repair (19).
Escherichia coli MutS and MutL proteins have been shown by
footprinting analysis to form an initiation complex in DNA
containing mismatches (20). A similar complex was demonstrated
with yeast MSH2, MLH1 and PMS1 by gel retardation analysis
(21). However, whether human MutS and MutL homologs
interact with each other has not been elucidated. Using immunoprecipitation and Western blot analysis we demonstrate here that
human MSH2, MLH1 and PMS2 can be co-immunoprecipitated
in the presence of DNA, suggesting that human MutS and MutL
homologs also form a repair initiation complex. Formation of the
initiation complex is ATP-dependent and requires at least
functional MSH2 and MLH1. In addition, we found that PCNA
was a member of this initiation complex, suggesting a role of
*To whom correspondence should be addressed at: Department of Pathology and Laboratory Medicine, Lucille P. Markey Cancer Center, University of Kentucky,
Suite MS-117, 800 Rose Street, Lexington, KY 40536-0084, USA. Tel: +1 606 257 7053; Fax: +1 606 323 2094; Email: [email protected]
+Present
address: Department of Animal Sciences, Michigan State University, East Lansing, MI 48823, USA
1174 Nucleic Acids Research, 1998, Vol. 26, No. 5
PCNA in repair initiation. We also found that when a limited
amount of PCNA was supplemented in p21-treated HeLa nuclear
extracts mismatch repair passed the excision step but was blocked
at the step of DNA re-synthesis, indicative of the requirement of
PCNA in the latter step as well.
MATERIALS AND METHODS
Cell lines and nuclear extract
HeLa S3 cells were purchased from the National Cell Culture
Center (Minneapolis, MN). Colorectal tumor line H6 was grown
in McCoy’s 5A medium (Gibco) with 10% fetal bovine serum
(HyClone) as described (22). LoVo cells were grown in
Dulbecco’s modified Eagle’s medium/Ham’s F-12 medium with
10% fetal bovine serum. Nuclear extracts from all cell lines were
prepared as described previously (5,22).
Antibodies and reagents
Antibodies against human MSH2 (polyclonal), MLH1 (polyclonal)
and PMS2 (monoclonal) were purchased from Oncogene Sciences
(Boston, MA). Antibody against human PCNA (monoclonal) was
obtained from Santa Cruz Biotechnology (Santa Cruz, CA).
Adenylyl imidodiphosphate (AMPPNP) was obtained from
Calbiochem (San Diego, CA). [γ-32P]ATP was from DuPont and
oligonucleotides were obtained from Bio-Synthesis (Houston, TX).
Immunoprecipitation and Western blot
All the immunoprecipitation steps were performed at 0–4C as
described (23), with some modifications. Nuclear extracts (250 µg)
were incubated with 10 µl protein A-positive Staphylococcus
aureus cells (Boehringer Mannheim) for 10 min and recovered by
centrifugation (16 000 g for 2 min). The resulting nuclear extracts
were added to a reaction containing 20 mM Tris–HCl, pH 7.6,
50 µg/ml bovine serum albumin (BSA), 5 mM MgCl2, 1.5 mM
ATP, 1 mM dithiothreitol (DTT), 0.1 M KCl and 1 µg
double-stranded (ds) f1MR3 DNA, followed by an overnight
incubation with 1 µg antibodies against either MSH2 or MLH1.
After incubation with 10 µl prewashed protein A–Sepharose
(Boehringer Mannheim) for 1 h the immunoprecipitates were
recovered by centrifugation (750 g, 2 min), washed three times
with fresh NP-40 lysis buffer [50 mM Tris–HCl, pH 8.0, 0.5%
Nonidet P-40, 0.15 M NaCl, 5 mM EDTA, 1 mM DTT, 1 mM
phenylmethylsulfonylfluoride (PMSF) and 1 mM NaVO4] and
resuspended in 10 µl SDS sample buffer (187.5 mM Tris–HCl,
pH 6.8, 6% SDS, 30% glycerol and 0.03% phenol red). Samples
were fractionated in an 8% SDS polyacrylamide gel and
transferred onto nitrocellulose membranes for Western blot
analysis. Membranes were blotted with antibody against MSH2,
MLH1, PMS2 or PCNA. Bound antibodies were detected by
chemiluminescence using a secondary antibody conjugated with
horseradish peroxidase (Amersham).
supernatant was mixed with 1 ml Ni–NTA–agarose (Qiagen) and
incubated for 1 h at room temperature. This mixture was packed into
a mini column and washed with decreasing concentrations of urea
lysis buffer. The p21 protein was eluted with 137 mM NaCl, 2.7 mM
KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4 and 200 mM imidazole.
In the PCNA preparation, cell pellets (∼10 g) were resuspended
in 40 ml buffer A (20 mM Tris–HCl, pH 7.5, 0.2 M NaCl, 1 mM
EDTA and 1 mM 2-mercaptoethanol), followed by sonication and
centrifugation (25 000 g for 20 min). The supernatant was treated
with solid (NH4)2SO4 (0.2 g/ml) and the precipitate removed by
centrifugation (20 min at 25 000 g). The resulting supernatant was
then treated with 0.4 g/ml (NH4)2SO4 in a similar manner, the
precipitates collected by centrifugation (20 min at 25 000 g),
dissolved in 20 ml buffer A and dialyzed against this buffer for 1.5 h.
The dialysate was applied at 1 ml/min onto a DEAE–Sepharose
CL-6B column (2.5 cm2 × 25 cm) equilibrated with buffer A. The
column was washed with 600 ml buffer A and eluted with 20 mM
Tris–HCl, pH 7.5, 0.5 M NaCl and 1 mM EDTA. Fractions
containing PCNA were pooled and concentrated to 10 ml using
Amicon cell and ultrafiltration membrane YM10 (Amicon) and
loaded at a rate of 0.2 ml/min onto a Sephadex G100 column
(2.5 cm2 × 95 cm) equilibrated with a buffer containing 20 mM
Tris–HCl, pH 7.8, 0.2 M NaCl and 1 mM EDTA. The column was
developed with the same buffer and PCNA-containing fractions
were pooled, concentrated and frozen in liquid N2 in small
aliquots and stored at –80C.
Analysis of p21-inhibited mismatch repair intermediates
Mismatch repair intermediates produced in the presence of
exogenous p21 were analyzed by Southern hybridization as
described (25). Repair assays were performed as described (5)
using 50 µg HeLa nuclear extracts and 100 ng G-T heteroduplex
containing a strand break in the complementary strand at the
Sau96I site in the presence and absence of p21, PCNA and
aphidicolin. DNA samples were recovered and digested with
Bsp106, followed by phenol extraction and ethanol precipitation
prior to suspension in alkaline loading buffer (0.2 N NaOH,
40 mM EDTA, 30% glycerol and 0.1% bromocresol green). DNA
was electrophoresed at 1.5–2.5 V/cm through 1.5% agarose gel in
30 mM NaOH, 2 mM EDTA at 4C and then electrotransferred onto
nylon membranes (ICN Biotrans). Membranes were prehybridized
at 37C for at least 4 h in 50 mM Tris–HCl, pH 7.5, 2% SDS,
0.5% polyvinyl pyrrolidone, 0.2% heparin, 1 mM EDTA and 1 M
NaCl and then hybridized at 37C overnight in the same solution
containing 5′-[32P]d(ATGGTTTCATTGGTGACGTT) to probe
repair intermediates. After hybridization membranes were
washed twice (5 min each) with 1× SSC (0.15 M NaCl, 15 mM
sodium citrate, pH 7.0) containing 0.1% SDS and then twice with
0.1× SSC containing 0.1% SDS. Radioactivity was detected by
exposing membranes to Fujifilm.
RESULTS
Co-immunoprecipitation of MSH2, MLH1, PMS2 and PCNA
Preparation of human p21 and PCNA
The human p21 gene or PCNA gene was overexpressed under the
T7 promoter in E.coli strain BL21-DE3 as described (24). In the
preparation of p21 the cells were lysed in 50 mM Tris–HCl,
pH 8.0, 50 mM NaCl, 10% glycerol and 8 M urea for 30 min at
room temperature. After centrifugation (25 000 g for 20 min) the
To determine interactions between human MutS and MutL
homologs during the initiation step of mismatch repair we
performed immunoprecipitation combined with Western blot
analysis (IP-Western) on HeLa nuclear extracts, which are fully
active in strand-specific mismatch repair (5,6). As shown in
Figure 1, antibody against human MLH1 not only precipitated
1175
Nucleic Acids
Acids Research,
Research,1994,
1998,Vol.
Vol.22,
26,No.
No.15
Nucleic
Figure 1. Co-immunoprecipitation of MSH2, MLH1 and PMS2. IP was
performed as described in Materials and Methods using 250 µg HeLa nuclear
extracts and 1 µg MLH1 polyclonal antibody in the presence or absence of 1 µg
DNA as indicated in a reaction containing 20 mM Tris–HCl, pH 7.6, 0.1 M KCl,
5 mM MgCl2, 1 mM DTT, 50 µg/ml BSA and 1.5 mM ATP. The heteroduplex
used was a single G-T mismatch-containing substrate constructed from the
f1MR phage series as described (5). The homoduplex was ds f1MR3 DNA.
Immunoprecipitates were electrophoresed in an 8% SDS–polyacrylamide gel
and transferred onto a nitrocellulose membrane for Western blot analysis by
chemiluminescence (see Materials and Methods). The same membrane was
sequentially blotted with antibodies against MLH1 (lanes 1–3), MSH2
(lanes 4–6) and PMS2 (lanes 7–9) without stripping the previous antibodies.
MLH1 protein (lanes 1–3), but also MSH2 (lanes 5 and 6) and
PMS2 (lanes 7–9) in the presence of DNA (homoduplex or
heteroduplex), Mg2+ and ATP. Although the precipitation of
MLH1 by its antibody was expected, there were two proteins
recognized by MLH1 antibody among the proteins precipitated.
It is not known whether this is due to a different modification
status of MLH1 (e.g. phosphorylation) or another human MutL
homolog closely related to MLH1. PMS2 was precipitated with
MLH1 in all reactions regardless of the presence of DNA
(lanes 7–9), further confirming that these two proteins form a
heterodimer (13). Unlike PMS2, MSH2 could not be co-precipitated
with MLH1 in the absence of DNA (lanes 4 and 7). Although the
presence of heteroduplex DNA in the reaction enhances complex
formation (lanes 5 and 8), homoduplex DNA is sufficient to induce
the interaction between MSH2 and MLH1–PMS2 (lanes 6 and 9).
These observations suggest that, as in E.coli and yeast, human MutS
and MutL homologs form an initiation complex in DNA.
Recently PCNA has been implicated in human mismatch repair
(14,26). Depletion of PCNA from HeLa nuclear extracts blocks
mismatch repair prior to excision (14), suggesting that PCNA
may be involved in repair initiation. To test this hypothesis we
examined the presence of PCNA in pellets precipitated by MLH1
antibodies. As shown in lane 3 of Figure 2, PCNA was indeed
co-precipitated with human MutS and MutL homologs, supporting
the idea that PCNA plays a role in mismatch repair initiation.
However, in nuclear extracts derived from either MSH2-defective
LoVo or MLH1-defective H6 cells, PCNA could not be precipitated
by antibodies against MLH1 or MSH2 (data not shown),
suggesting that formation of the initiation complex requires
functional MSH2 and MLH1 proteins.
1175
Figure 2. ATP is required for formation of the mismatch repair initiation
complex. IP-Western was performed as described in Materials and Methods
using 250 µg HeLa nuclear extracts, 1 µg MLH1 polyclonal antibody (except
lane 4), 1 µg ds f1MR3 DNA in the presence or absence of 1.5 mM ATP as
indicated. Proteins precipitated were fractionated in a SDS–polyacrylamide gel
and transferred onto a nitrocellulose membrane. The membrane was co-blotted
with antibodies against MSH2, PMS2 and PCNA in Western blot analysis.
AMPPNP is adenylyl imidodiphosphate, a non-hydrolyzable ATP analog.
ATP hydrolysis is required for formation of the mismatch
repair initiation complex
The requirement for ATP in a mismatch repair initiation complex
is controversial in E.coli and in yeast. In E.coli ATP is essential
for interaction of MutS and MutL in mismatched DNA (20),
while complex formation of yeast MutS and MutL homologs in
DNA does not require exogenous ATP (21). To determine the role
of ATP in the human complex formation, IP-Western was
performed in the presence or absence of ATP. As demonstrated in
Figure 2, when ATP was eliminated from the immunoprecipitation
step (lane 1) both MSH2 and PCNA could not be co-precipitated
with MLH1 by the MLH1 antibody. The same was true when a
non-hydrolyzable ATP analog, adenylyl β,γ-imidodiphosphate
(AMPPNP) was used (lane 2). Complex formation of MSH2,
MLH1, PMS2 and PCNA could only be detected in the presence
of ATP (lane 3), suggesting a requirement for ATP hydrolysis in
this event. In addition, negative results in lanes 1, 2 and 4 of the
figure indicate that co-immunoprecipitation of MSH2 and PCNA
with MLH1–PMS2 is due to their specific interactions, not
because of their individual affinity for DNA molecules in the
reactions. Similar results were observed when MSH2 antibodies
were used during immunoprecipitation (data not shown).
PCNA is required for DNA re-synthesis in human
mismatch repair
The function of PCNA in mismatch repair can be abolished by
exogenous p21 (14). As an accessory factor for DNA polymerases
δ and ε PCNA is required for DNA replication. To test for a
possible role for PCNA in repair DNA re-synthesis in mismatch
1176 Nucleic Acids Research, 1998, Vol. 26, No. 5
A
Figure 4. PCNA is involved in repair DNA re-synthesis. Mismatch repair and
Southern analysis were performed as described in Figure 3 except that various
concentrations of PCNA were used, as indicated.
B
Figure 3. Analysis of p21-inhibited mismatch repair intermediates. (A) The
principle of detecting repair intermediates. (B) Southern blot analysis. Circular
G-T heteroduplex (100 ng) containing a complementary strand incision 125 bp
5′ of the mismatch was incubated with 50 µg HeLa nuclear extracts in the
presence or absence of aphidicolin, p21 and PCNA, as indicated. After cleavage
with Bsp106 products were subject to alkaline agarose gel (1.5%) electrophoresis
and electrotransferred onto a nylon membrane. The membrane was hybridized
with 5′-[32P]d(ATGGTTTCATTGGTGACGTT) to probe the excision tract
5′-termini. Corresponding indirectly end-labeled intermediates and electrophoretic
species are shown to the left of the gel, with the 3.2 kb marker indicating the
original location of the site-specific strand break, the 3.1 kb marker the location
of the mismatch and the solid bar the oligonucleotide probe. Normal repair
products and the ligated substrates map 6.4 kb, whereas the mismatch and the
strand break in the heteroduplex map 3.1 and 3.2 kb 5′ of the Bsp106 site
respectively. Products <3.1 kb are the excision intermediates.
repair Southern hybridization analysis was used to directly
visualize repair intermediates (25) produced during repair of the
heteroduplexes by nuclear extracts in the presence of p21. Under
conditions where repair DNA re-synthesis is inhibited, for
example in the presence of aphidicolin or in the absence of
exogenous dNTPs, mismatch-provoked excision intermediates
can be visualized as a single strand gap that spans the position of
the strand break and the original location of the mismatch (25).
These excision end points, which distribute near a point past the
mismatch, can be mapped relative to a restriction endonuclease
cleavage site by Southern blot analysis (25; see also Fig. 3A) after
electrophoresis through denaturing gels. Figure 3B demonstrates
an analysis of repair intermediates produced in aphidicolin- or
p21-treated HeLa nuclear extracts from an incised circular G-T
mismatch. In the normal reaction (lane 2 of Fig. 3B) the strand
that originally contained the incision was recovered as a
full-length 6.4 kb species (see normal repair pathway in Fig. 3A),
indicative of ligation or complete repair and ligation of the
substrate DNA. In reactions containing aphidicolin (lane 3) or
p21 (lane 4) 6.4 kb species were also observed, which was not due
to repair but to ligation of the substrate prior to excision (5,25).
Assays containing aphidicolin revealed a population of excision
intermediates 3.1 kb (lane 3 of Fig. 3B), indicative of a block
in DNA re-synthesis (see no re-synthesis pathway in Fig. 3A). In the
presence of p21 protein (lane 4 of Fig. 3) the yield of excision
intermediates (≤3.1 kb) was greatly reduced. Instead, products
were mapped near the strand break (compared with the original
substrate shown in lane 1), indicating that repair was blocked
prior to excision (see no initiation pathway in Fig. 3A). However,
in the reactions supplemented with PCNA protein (1 µM) both 3.2
and <3.1 kb species were observed (lane 5 of Fig. 3B). The
emergence of the latter species suggests that repair excision
occurred in these molecules but DNA re-synthesis was inhibited
(compared with the aphidicolin reaction). This result seems to
conflict with the previous report that p21-inhibited mismatch
repair in HeLa cell-free extracts can be reversed by addition of
exogenous PCNA (14). A possible explanation is that PCNA may
be required for both repair initiation and repair DNA re-synthesis
and that a limited amount of PCNA in the reaction could restore
repair initiation, but not repair re-synthesis. To address this issue
we performed experiments by increasing the PCNA concentration
1177
Nucleic Acids
Acids Research,
Research,1994,
1998,Vol.
Vol.22,
26,No.
No.15
Nucleic
in p21-treated reactions. As shown in Figure 4, as the concentration
of PCNA increased (≥2.0 µM, lanes 4–6) both 3.2 (uninitiated) and
<3.1 kb (re-synthesis blocked) species that were present in the
reaction containing 1 µM PCNA (lane 5 of Fig. 3 and lane 3 of
Fig. 4) disappeared, indicating conversion of both species into 6.4 kb
products through restoration of the whole repair process for the
uninitiated molecules and restoration of repair DNA re-synthesis
for the gapped molecules by excess PCNA.
DISCUSSION
Our co-immunoprecipitation data summarized above provide the
first evidence that human MutS and MutL homologs form an
initiation complex. In addition, we have found that PCNA is a
member of this initiation complex. Although yeast PCNA has
been shown to interact with yeast PMS1 in a two-hybrid system
(14) and interact with yeast MSH2–MSH3 using recombinant
proteins (26), such interactions in cell-free systems as well as in
human cells have never been reported. We have also demonstrated
that the interaction of human MSH2, MLH1, PMS2 and PCNA
requires ATP hydrolysis (Fig. 2). This is consistent with studies
in E.coli in which enhancement of the MutS–heteroduplex
footprint by MutL is ATP-dependent (20), but contrast with
observations in yeast where interactions among yeast mismatch
repair proteins do not require exogenous ATP (21). The
divergence for the ATP requirement between yeast and E.coli/
humans awaits further biochemical studies.
It has recently been demonstrated that in the presence of ATP
E.coli MutS promotes formation of α-shaped loop structures
from heteroduplex DNA (27). The formation of such loop
structures could bring a mismatch and a hemimethylated
d(GATC) site into close proximity, thereby activating the MutH
endonuclease activity (4). Therefore, the role of ATP in mismatch
repair initiation may be to provide energy for MutS translocation
and for recruiting MutL and MutH to the MutS–heteroduplex
complex. Although MutS possesses a weak ATPase activity
(28,29), it is not established whether it is responsible for ATP
hydrolysis in these reactions. Human MutS homologs, e.g. hMutSα,
are capable of interacting with oligonucleotides containing
mispairs or carcinogen–DNA adducts (7,30,31). Binding of
hMutSα to these heteroduplex-containing oligonucleotides is
abolished in the presence of ATP (7,31), implying an ATP-driven
translocation of the human MutS homologs, which causes MutS
homologs to slide off the small oligonucleotides. The present
study has demonstrated that formation of a human mismatch
repair initiation complex is dependent on ATP hydrolysis,
suggesting that mismatch recognition and initiation complex
formation in E.coli and humans are similar: both depend on an
ATP-driven interaction between mismatched DNA and MutS and
MutL homologs.
We were unable to determine whether MSH3 and MSH6 are
components of the repair initiation complex due to the unavailability
of quality antibodies against these human proteins. However,
both MSH6 and MSH3 have been shown to form heterodimers
with MSH2 in humans and yeast (7,8,10–12). Based on our
observation that the heterodimer of MLH1 and PMS2 exists as a
single unit regardless of reaction conditions (e.g. ATP and DNA),
MSH3 and MSH6 may always be associated with MSH2.
Therefore, they should be members of the eukaryotic mismatch
repair initiation complex.
1177
Our results also showed that formation of the repair initiation
complex occurs in the presence of either homoduplex or
heteroduplex DNA, although the presence of heteroduplex DNA
strengthened complex formation (Fig. 1). This phenomenon was
previously shown in in vitro translated yeast proteins by gel
retardation analysis, where yeast MSH2 formed a ternary
complex with yeast MLH1 and PMS1 (21). These findings
suggest that the eukaryotic MutS homologs are regular DNA
binding proteins that constantly scan DNA for mispairs in an
energy-dependent fashion. Once a mismatch is identified by
MutS homologs other proteins, including MutL homologs and
PCNA, are recruited to the damage site to initiate mismatch
repair. This process is also energy-dependent, as formation of the
initiation complex requires ATP (Fig. 2).
Another novel discovery in this study is that we clearly show
a dual role for PCNA in human mismatch repair. Co-immunoprecipitation of PCNA with human MutS and MutL homologs
(Fig. 2) suggests involvement of PCNA in mismatch repair
initiation, supporting the previous observation that PCNA is
required at a step prior to excision (14). In addition, we
demonstrate here that PCNA plays a second role in human
mismatch repair, i.e. participating in repair DNA re-synthesis.
Although supplementation with exogenous PCNA can reverse
p21-inhibited mismatch repair (14), a significant amount of
unrepaired molecules (3.2 kb) and repair intermediates (<3.1 kb)
that were blocked at the step of DNA re-synthesis were observed
in a reaction with a limited amount of PCNA (Fig. 3). However,
both species were converted into full-length products in the
presence of excess PCNA (Fig. 4). Although we cannot
completely rule out other possibilities for these conversions, e.g. a
ligase activity stimulated by PCNA, it is most likely that
conversion of the species <3.1 kb is due to DNA polymerase
activity and that the 3.2 kb species are processed through the
whole repair cycle, i.e. initiation, excision and DNA re-synthesis.
Taken together, these findings strongly suggest that PCNA is
required in human mismatch repair not only in repair initiation,
but also in repair DNA re-synthesis.
The requirement for PCNA at the step of DNA re-synthesis in
the human system further supports the involvement of polymerase
δ or ε or both in eukaryotic mismatch repair. In fact, DNA
polymerase δ has recently been implicated in human mismatch
repair (19). The role of PCNA in the step prior to excision in
mismatch repair is not clear. Since it interacts with Mut homologs
(14,26; this study), PCNA may function as a ‘molecular
matchmaker’ (32) to stimulate interactions between MutS
homologs and MutL homologs. Alternatively, since PCNA
physically resides at DNA primer termini that can be used as a
starting point for mismatch repair (1), it may serve as the strand
discrimination signal and initiation target to induce formation of
the repair initiation complex. It is this specific interaction between
the mismatch repair enzymes and the replication apparatus that
leads to correction of biosynthetic errors (33).
ACKNOWLEDGEMENTS
We thank Yue Xiong and Henming Ke of the University of North
Carolina at Chapel Hill for reagents and discussion during the
course of the experiments and for supporting H.W. We thank Paul
Modrich and Isabel Mellon for critical evaluation of the
manuscript. This work was supported by grants IRG-163H from
the American Cancer Society, 4755 from the Council for Tobacco
1178 Nucleic Acids Research, 1998, Vol. 26, No. 5
Research Inc. and CA72856 from the National Cancer Institute
to G.-M.L. S.M. was supported in part by an NIEHS Training
Grant (ES07266) and H.W. was supported in part by a Showa
University School of Medicine Fellowship.
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Modrich,P. and Lahue,R. (1996) Annu. Rev. Biochem., 65, 101–133.
Kolodner,R. (1996) Genes Dev., 10, 1433–1442.
Umar,A. and Kunkel,T.A. (1996) Eur. J. Biochem., 238, 297–307.
Modrich,P. (1991) Annu. Rev. Genet., 25, 229–253.
Holmes,J., Clark,S. and Modrich,P. (1990) Proc. Natl. Acad. Sci. USA, 87,
5837–5841.
Thomas,D.C., Roberts,J.D. and Kunkel,T.A. (1991) J. Biol. Chem., 266,
3744–3751.
Drummond,J.T., Li,G.-M., Longley,M.J. and Modrich,P. (1995) Science,
268, 1909–1912.
Palombo,F., Gallinari,P., Iaccarino,I., Lettieri,T., Hughes,M., D’Arrigo,A.,
Truong,O., Hsuan,J.J. and Jiricny,J. (1995) Science, 268, 1912–1914.
Acharaya,S., Wilson,T., Gradia,S., Kane,M.F., Guerrette,S., Marsischky,G.T.,
Kolodner,R. and Fishel,R. (1996) Proc. Natl. Acad. Sci. USA, 93,
13629–13634.
Marsischky,G.T., Filosi,N., Kane,M.F. and Kolodner,R. (1996) Genes Dev.,
10, 407–420.
Johnson,R.E., Kovvali,G.K., Prakash,L. and Prakash,S. (1996) J. Biol. Chem.,
271, 7285–7288.
Habraken,Y., Sung,P., Prakash,L. and Prakash,S. (1996) Curr. Biol., 6,
1185–1187.
Li,G.-M. and Modrich,P. (1995) Proc. Natl. Acad. Sci. USA, 92, 1950–1954.
Umar,A., Buermeyer,A.B., Simon,J., Thomas,D.C., Clark,A.B.,
Liskay,R.M. and Kunkel,T.A. (1996) Cell, 87, 65–73.
Xiong,Y., Zhang,H. and Beach,D. (1992) Cell, 71, 505–514.
Xiong,Y., Hannon,G., Zhang,H., Casso,D., Kobayashi,R. and Beach,D.
(1993) Nature, 366, 701–704.
17 Waga,S., Hannon,G.J., Beach,D. and Stillman,B. (1994) Nature, 369,
574–578.
18 Flores,R.H., Kelman,Z., Dean,F.B., Pan,Z.Q., Harper,J.W., Elledge,S.J.,
O’Donnell,M. and Hurwitz,J. (1994) Proc. Natl. Acad. Sci. USA, 91,
8655–8659.
19 Longley,M.J., Pierce,A.J. and Modrich,P. (1997) J. Biol. Chem., 272,
10917–10921.
20 Grilley,M., Welsh,K.M., Su,S.-S. and Modrich,P. (1989) J. Biol. Chem.,
264, 1000–1004.
21 Prolla,T.A., Pang,Q., Alani,E., Kolodner,R.D. and Liskay,R.M. (1994)
Science, 265, 1091–1093.
22 Parsons,R., Li,G.M., Longley,M.J., Fang,W.H., Papadopoulos,N., Jen,J.,
de la Chapelle,A., Kinzler,K.W., Vogelstein,B. and Modrich,P. (1993) Cell,
75, 1227–1236.
23 Jenkins,C.W. and Xiong,Y. (1995) In Pagano,M. (ed.), Cell Cycle:
Materials and Methods. Springer-Verlag, New York, NY, pp. 250–263.
24 Pan,Z.Q., Reardon,J.T., Li,L., Flores,R.H., Legerski,R., Sancar,A. and
Hurwitz,J. (1995) J. Biol. Chem., 270, 22008–22016.
25 Fang,W.H. and Modrich,P. (1993) J. Biol. Chem., 268, 11838–11844.
26 Johnson,R.E., Kovvali,G.K., Guzder,S.N., Amin,N.S., Holm,C.,
Habraken,Y., Sung,P., Prakash,L. and Prakash,S. (1996) J. Biol. Chem.,
271, 27987–27990.
27 Allen,D.J., Makhov,A., Grilley,M., Taylor,J., Thresher,R., Modrich,P. and
Griffith,J.D. (1997) EMBO J., 16, 4467–4476.
28 Haber,L.T., Pang,P.P., Sobell,D.I., Mankovich,J.A. and Walker,G.C. (1988)
J. Bacteriol., 170, 197–202.
29 Haber,L.T. and Walker,G.C. (1991) EMBO J., 10, 2707–2715.
30 Duckett,D.R., Drummond,J.T., Murchie,A.I.H., Reardon,J.T., Sancar,A.,
Lilley,D.M. and Modrich,P. (1996) Proc. Natl. Acad. Sci. USA, 93,
6443–6447.
31 Li,G.-M., Wang,H. and Romano,L.J. (1996) J. Biol. Chem., 271,
24084–24088.
32 Sancar,A. and Hearst,J.E. (1993) Science, 259, 1415–1420.
33 Wagner,R. and Meselson,M. (1976) Proc. Natl. Acad. Sci. USA, 73,
4135–4139.