A conserved interaction between the replicative clamp loader and DNA

ligase in eukaryotes; implications for Okazaki fragment joining

David S. Levin1, 6, 8, Sangeetha Vijayakumar2, 8, Xiuping Liu1, 7, 8, Vladimir P.

Bermudez3, Jerard Hurwitz3 and Alan E. Tomkinson4, 5

1Department of Molecular Medicine, Institute of Biotechnology, The University of TexasHealth Science Center at San Antonio, 15355, Lambda Drive, San Antonio, TX 782452Molecular and Cellular Biology Graduate Program, University of Maryland School ofMedicine, 655 West Baltimore St, Baltimore, MD212013Program in Molecular Biology, Sloan Kettering Institute, Memorial Sloan-KetteringCancer Center, New York, NY100214Radiation Oncology Research Laboratory, Department of Radiation Oncology andGreenebaum Cancer Center, University of Maryland School of Medicine, 655 WestBaltimore St., Baltimore, MD212015Corresponding author: Radiation Oncology Research Laboratory, Department ofRadiation Oncology and Greenebaum Cancer Center, University of Maryland School ofMedicine, 655 West Baltimore St, Baltimore, Maryland

Tel: 410-706-2375; FAX: 410-706-3000; E mail, atomkinson@som.umaryland.edu6 Present address: GeneTex Inc., 14785 Omicron Drive, San Antonio, TX 782457 Present address: Department of Molecular Genetics, M.D. Anderson Cancer Center, The University of Texas, Houston, TX 77030, USA.

8These authors contributed equally.

Running Title: DNA Ligase I interacts with the clamp loader RFC.

Abbreviations footnote: BSA, bovine serum albumin; FEN-1, flap endonuclease 1; GST,glutathione S transferase; PCNA, proliferating cell nuclear antigen; RFC, replicationfactor C.

1

JBC Papers in Press. Published on October 23, 2004 as Manuscript M409250200

Copyright 2004 by The American Society for Biochemistry and Molecular Biology, Inc.

ABSTRACT

The recruitment of DNA ligase I to replication foci and the efficient joining of

Okazaki fragments is dependent on the interaction between DNA ligase I and

proliferating cell nuclear antigen (PCNA). Although the PCNA sliding clamp tethers

DNA ligase I to nicked duplex DNA circles, the interaction does not enhance DNA

joining. This suggests that other factors may be involved in the joining of Okazaki

fragments. In this study we describe an association between replication factor C (RFC),

the clamp loader, and DNA ligase I in human cell extracts. Subsequently we demonstrate

that there is a direct physical interaction between these proteins that involves both the N-

and C-terminal domains of DNA ligase I, the N-terminus of the large RFC subunit p140

and the p36 and p38 subunits of RFC. Although RFC inhibited DNA joining by DNA

ligase I, the addition of PCNA alleviated inhibition by RFC. Notably, the effect of PCNA

on ligation was dependent on the PCNA–binding site of DNA ligase I. Together these

results provide a molecular explanation for the key in vivo role of the DNA ligase

I/PCNA interaction and suggest that the joining of Okazaki fragments is co-ordinated by

pair wise interactions among RFC, PCNA and DNA ligase I.

2

INTRODUCTION

Mammalian cells contain multiple species of DNA ligase that are encoded by the

LIG1, LIG3 and LIG4 genes (1). There is compelling evidence linking the product of the

LIG1 gene, DNA ligase I, with the joining of Okazaki fragments generated during lagging

strand DNA synthesis. For example, DNA ligase I co-localizes with DNA replication

foci, co-purifies with a multisubunit DNA replication complex and efficiently joins

Okazaki fragments when DNA replication is reconstituted with purified replication

factors (2-4). Moreover, a DNA ligase I-deficient human cell line 46BR1G1 is defective

in the conversion of Okazaki fragments into larger DNA replication intermediates (5-8).

The identification and characterization of a specific interaction between DNA

ligase I and the homotrimeric sliding clamp, proliferating cell nuclear antigen (PCNA),

provided the first evidence physically linking DNA ligase I with a component of the

DNA replication machinery (9,10). Subsequent studies revealed that the 20 amino acid

sequence at the N-terminus of DNA ligase I is necessary and sufficient to bind to PCNA

and that this sequence is homologous to the canonical PCNA binding motif that has been

found in a growing number of PCNA-interacting proteins (7,11,12). Amino acid changes

that inactivate the PCNA binding site of DNA ligase I prevent the recruitment of DNA

ligase I to DNA replication foci and abolish the ability of this enzyme to efficiently join

Okazaki fragments (7,11). The interactions of PCNA with other DNA replication

3

enzymes such as replication factor C (RFC), the heteropentameric clamp loader, the

replicative DNA polymerases, Pol δ and Pol ε, and the 5 flap endonuclease (FEN-1)

suggest that this homotrimer plays a central role in co-ordinating and regulating the

actions of these enzymes during gap-filling DNA synthesis and ligation (12,13).

Biochemical assays measuring DNA synthesis and ligation in reactions either

reconstituted with purified replication factors or catalyzed by cell extracts have shown

that the ability of DNA ligase I to bind to PCNA is critical for efficient ligation and that

the extent of strand displacement DNA synthesis is limited by the functional interaction

between DNA ligase I and PCNA (7,14,15). Although DNA ligase I stably associates

with PCNA trimers that are topologically linked to duplex DNA circles (9), PCNA only

weakly stimulates DNA joining at very high concentrations (16). This suggests that

additional replication factors may be involved in the functional interaction between DNA

ligase I and PCNA. In this study we describe a physical interaction between DNA ligase I

and the large subunit of the clamp loader RFC that is conserved among eukaryotes.

Furthermore, we show that RFC modulates DNA joining by DNA ligase I in a reaction

that is dependent upon the interaction between DNA ligase I and PCNA.

EXPERIMENTAL PROCEDURES

Protein purification . Recombinant DNA ligase I was purified from Sf9 cells infected with

a baculovirus expressing human DNA ligase I (17). PCNA was overexpressed in and

4

purified from E. coli as described (9,18). Recombinant RFC complex containing full

length RFCp140 and a truncated version of RFCp140 lacking the N-terminal 555

residues (RFC∆Np140) were purified from baculovirus-infected insect cells (19). The

intact RFC complex was also obtained from Dr. Vladimir Podust (20).

DNA ligase I affinity chromatography. Nuclear and cytoplasmic extracts were prepared

from a frozen pellet of HeLa S3 cells (109 cells) as described (21). The nuclear extract

(20 mg) was fractionated by DNA ligase I affinity chromatography as described

previously (9). Fractions were analyzed for protein by immunoblotting after separation by

SDS-PAGE (22).

Purified RFC was incubated with either DNA ligase I or bovine serum albumin

(BSA) beads (9) in binding buffer (50 mM Tris-HCl pH 7.5, 100 mM KCl, 1 mM

dithiothreitol, 1% Nonidet P-40 and 5 µg BSA) for 30 min at 4°C with constant agitation.

After collection by centrifugation, the beads were washed with binding buffer and bound

proteins were eluted with SDS sample buffer (22). After separation by SDS-PAGE,

proteins were detected by immunoblotting with antibodies against RFC p37 (GeneTex

Inc) and RFC p140 (a gift from Dr. Bruce Stillman).

Immunoprecipitation. Hela cells (8 X 106 ) were lysed in IP buffer (50 mM Tris-HCl

pH 7.5, 150 mM KCl, 10 mM MgCl2, 1% Nonidet P-40 (Sigma), 1 mM DTT, 50 µg/ml

ethidium bromide, 1µg/ml leupeptin, 1 µg/ml pepstatin, 1 µg/ml chymostatin, 0.1 mM

5

phenylmethanesulfonyl fluoride, 50 mM NaF, 1 mM Na3VO4). Approximately 1 mg of

the clarified extract was used for each immunoprecipitation. The extracts were pre-

cleared by incubation for 1 hour at 4°C with 50 µl of Protein A Sepharose and Protein G

Sepharose beads (Amersham Biosciences, Piscataway, NJ) equilibrated with IP buffer,

prior to the addition of antibodies against DNA ligase I, RFC p37 or Cdc25 (all from

GeneTex Inc.). After incubation at 4°C for 2 hours, 50 µl of Protein A Sepharose and

Protein G Sepharose beads were added and the incubation continued for 1 hour. The

beads were collected by centrifugation, washed three times with IP buffer lacking

ethidium bromide and then resuspended in SDS- sample buffer. After separation by

SDS-PAGE, proteins were detected by immunoblotting.

Glutathione S transferase (GST)-fusion proteins. GST fusion proteins containing either

the N-terminal 118 residues of DNA ligase I (GST-N Lig 1-118) or residues 479-919

of DNA ligase I (GST-C Lig1) were expressed and purified as described previously (9).

Sequences encoding residues 1-584 and 479-1148 of RFC p140 were amplified from

pET16a-p140 (19) by the polymerase chain reaction and subcloned in-frame with the

GST open reading frame in the pGEX vector to generate pGST-Np140 and pGST-

Cp140, respectively. Similarly, the open reading frame encoding Cdc9 DNA ligase was

amplified from S. cerevisiae genomic DNA by the polymerase chain reaction and

subcloned in-frame with the GST open reading frame in the pGSTag vector (23). After

6

expression in E. coli, GST fusion proteins were purified from cell extract affinity

chromatography using glutathione Sepharose beads.

In Vitro Transcription and Translation. Coupled in vitro transcription and translation

reactions were performed using the TNT Quick Coupled Transcription/Translation system

(Promega, Madison, WI). The plasmids for the in vitro transcription and translation of

human RFC subunits have been described previously (19). The open reading frame

encoding Rfc1 was amplified from S. cerevisiae genomic DNA by the polymerase chain

reaction and subcloned into pET-28b for coupled in vitro transcription and translation.

Labeled in vitro translated polypeptides were partially purified by ammonium sulfate

precipitation (24).

Pull down assays. To prepare GST and GST-Np140, GST-Cp140, GST-N Lig 1-118

and GST-C Lig1 beads, 5 µg of each purified protein was incubated with a 20 µl slurry of

Glutathione Sepharose beads (Amersham Biosciences) equilibrated in binding buffer for

30 min at 4°C with constant agitation. After washing with binding buffer, the beads were

resuspended in 500 µl binding buffer containing a labeled in vitro translated polypeptide

and then incubated at room temperature for 30 min with constant agitation. Next the

beads were collected by centrifugation and then washed extensively in binding buffer

prior to being resuspended in 20 µl of SDS-PAGE sample buffer. After separation by

SDS-PAGE, labeled polypeptides were visualized by phosphorImager analysis

7

(Molecular Dynamics).

Preparation of biotin-labeled linear DNA joining substrate. A 5’ biotinylated 90-mer

oligonucleotide, Bio-5-90, with a sequence corresponding to nucleotide positions 4881

and 4971 of M13mp19 single stranded DNA was purchased from Integrated DNA

Technologies, Inc. Bio15-1 5’-TGAGGCGGTCAG TAT-3’ and Bio15-2 5’-

AAGATAAAACAGAGG –3’ (Integrated Technologies, Inc) are complimentary to Bio-

5-90. Bio15-1 was 5 end labeled with 150 µCi of γ-32P ATP using T4 Polynucleotide

Kinase (New England Biolabs). After purification purified on Micro Bio-Spin 30 column

(Bio-Rad), labeled Biol5-1 and Bio15-2, were annealed to Bio-5-90 to generate a

partial duplex of 30 bp containing a single ligatable nick in the middle flanked by single-

stranded regions of 30 nucleotides.

DNA joining reaction with biotin-labeled linear substrate. Streptavidin-agarose beads

(10 µl, Pierce) were incubated with 1.6 pmol of the biotinylated linear DNA substrate in

PBS for 30 min at room temperature. After washing three times in ligation buffer (50 mM

Tris-HCl, pH 7.5, 10 mM MgCl2, 1 mM dithiothreitol, 0.25 mg/ml BSA, 100 µM ATP,

and 100 mM NaCl), the beads were incubated with 2 pmol of RPA/pmol of DNA in the

same buffer for 15 min at room-temparature. This substrate was then incubated with 2

pmol RFC or 2 pmol of ∆NRFC in the presence and absence of PCNA (2 pmol trimer) at

30°C for 2 minutes. DNA ligase I (2 pmol), either wild type or mutant, was added and the

8

reaction incubated at room-temperature for 5 minutes. The beads were then spun down

and the reaction terminated by addition of 10 µl of stop mix (50% glycerol, 1% SDS, 20

mM EDTA and 0.05% bromophenol blue). The beads were heated at 100° C for 3 min to

denature DNA. A 2 µl aliquot was mixed with 2 µl of denaturingPAGE dye (80 %

formamide, 0.05% bromophenol and 0.05% xylene cyanol). The samples were

electrophoresed through a 12% denaturing polyacrylamide gel. After drying, the gel was

exposed to a Storage Phosphor screen and subjected to phosphorImager analysis

(Molecular Dynamics).

Preparation of circular DNA joining substrate. The oligonucleotides,

5’CGTACGGGGAAGGACGTCAA3’ and 5’

CATGAAACCAACATAAACGTTATTGCCCGG 3’ (100 pmol of each), were end-

labeled with T4 polynucleotide kinase (New England Biolabs) in the presence of 3.2 mM

ATP and 70 µCi [γ32P] ATP. After purification from free nucleotides by passage through

a 1 ml G25 spin column, both of the labeled oligonucleotides were annealed to 25 µg of

circular ΦX174 single-stranded DNA by heating at 1000C for 15 minutes in 30 mM

Tris-HCl, pH 7.5, and 300 mM NaCl (200 µl total volume) followed by slow cooling to

room temperature. The labeled partial duplex circles were purified from free

oligonucleotides by passage through a 5 ml Bio-Gel A15m (Bio-Rad) column. Peak

fractions were ethanol precipitated, resuspended in 10 mM Tris-HCl, pH 7.5 and 1 mM

9

EDTA, and then run on a 1% agarose gel with ethidium bromide to verify the presence of

the substrate DNA. When annealed to their complementary sequences in ΦX174 single-

stranded DNA, the oligonucleotides form a partial duplex region containing a single

ligatable nick.

DNA joining reaction with circular substrate. The labeled circular DNA substrate DNA

(100 fmol) was incubated with RFC (0.6 pmol) in the presence or absence of PCNA (1.2

pmol trimer) at 30°C for 1 minute in ligation buffer. Human DNA ligase I (90 fmol),

either wild type or a mutant version with a disrupted PCNA binding site (7) was then

added. Aliquots were taken after 0, 30, 90 and 300 seconds and added directly to DNA

termination dye. After boiling, samples were electrophoresed through a 6% denaturing

acrylamide gel. Labeled oligonucleotides in the dried gel were detected and quantitated

by phosphorImager analysis (Molecular Dynamics).

RESULTS

Previously we have shown that DNA ligase I forms a stable complex with PCNA

molecules that are topologically linked to a nicked DNA circle (9). Since this interaction

did not significantly increase the efficiency of DNA joining (9), we suspected that

additional DNA replication factors may be involved in promoting DNA joining by DNA

ligase I. To identify such factors, we fractionated a HeLa nuclear extract by DNA ligase

I-affinity chromatography. As reported previously, PCNA was specifically retained by

10

the DNA ligase I resin (Fig. 1A, compare lanes 3 and 4). Analysis of the same fractions

by immunoblotting with antibodies specific for the p37 subunit and p140 subunit of the

clamp loader RFC revealed that these proteins were also present in the 150 mM NaCl

eluates from the DNA ligase I column but not in the equivalent eluates from the BSA

column (Fig. 1A, compare lanes 3 and 4). PCNA and the RFC subunits, p37 and p140,

were also present in the 300 mM NaCl eluate from the DNA ligase I but not from the

BSA column (data not shown). Since RFC also binds to PCNA (25), it is possible that

DNA ligase I-bound PCNA acts as a binding site for other PCNA binding proteins.

However, FEN-1, another PCNA-binding replication protein (26), was not retained by

the DNA ligase I resin (Fig. 1A), indicating that the binding of PCNA to the DNA ligase

I beads does not result in the non-specific retention of other PCNA binding proteins. To

provide further evidence for the specific association between RFC and DNA ligase I in

HeLa cell lysates, we performed immunoprecipitations in the presence of ethidium

bromide. As shown in Figure 1B, the DNA ligase I antibody co-immunoprecipitated the

p37 subunit of RFC (Fig. 1B, lane 3) and, in reciprocal experiments, the RFC p37

antibody co-immunoprecipitated DNA ligase I (Fig. 1B, lane 4).

Since RFC p37 and DNA ligase I can be co-immunprecipitated in the presence

of ethidium bromide, it appears likely that the association between these factors is

mediated by protein-protein interactions. To determine whether there is a direct

interaction between DNA ligase I and RFC, we performed pull down assays with DNA

11

ligase I affinity beads and purified recombinant human RFC (20) (Fig. 2A, lane 2). The

binding of RFC to DNA ligase I- but not BSA-beads (Fig. 2B, compare lanes 2 and 3),

demonstrates that there is a direct interaction between RFC and DNA ligase I.

RFC is composed of a large subunit, p140 and four smaller subunits p40, p38, p37

and p36 (27,28). Recently, several different variants of RFC, in which p140 has been

replaced by different polypeptides, have been described (29,30). Since p140 is the unique

component of the replicative clamp loader, we reasoned that the specific association with

RFC is likely to involve an interaction with p140. To test this idea we expressed the N-

terminal domain of p140, which can be removed without loss of catalytic activity (19),

and the C-terminal domain of p140, which is required for complex formation with the

four small subunits (19), as GST fusion proteins. In pull down assays, in vitro translated

DNA ligase I was specifically retained on glutathione beads liganded by the non-

catalytic N- terminal domain of p140 (Fig. 3A).

If the interaction between the clamp loader and the DNA ligase plays a critical

role in lagging strand DNA synthesis then it should be conserved other eukaryotes. This

prompted us to examine whether Cdc9, the S. cerevisiae DNA ligase I homolog, interacts

with the large subunit of S. cerevisiae RFC, Rfc1. As shown in Figure 3B, in vitro

translated Rfc1 was specifically retained on glutathione beads liganded by GST-Cdc9,

indicating that the interaction between the large subunit of the replicative clamp loader

and DNA ligase is conserved in eukaryotes.

12

Since previous studies have shown that the N-terminal 118 amino acid residues

of DNA ligase I contain the binding sites for both PCNA and DNA polymerase β (11,31),

we examined whether the same region was involved in the interaction with RFC p140. As

shown in Figure 4A, this fragment did bind specifically to the N-terminal domain of

RFC p140. This interaction was not disrupted by the substitution of the adjacent

phenylalanine residues by alanine residues in the PCNA binding site at the N-terminus of

DNA ligase I (data not shown). Unexpectedly, the N-terminal domain of RFC p140

interacted more efficiently with a fragment encompassing the catalytic domain of DNA

ligase I (Fig. 4A). This is the first example of a protein-protein interaction involving this

region of DNA ligase I.

The observation that more than one region of DNA ligase I binds to the N

–terminal domain of RFC p140 prompted us to examine whether DNA ligase I also

interacts with the small RFC subunits. As expected, in vitro translated RFC p140 bound

specifically to both the N- (Fig. 4B, lane 19) and C-terminal (Fig. 4B, lane 20)

fragments of DNA ligase I expressed as GST fusion proteins. Similar, albeit weaker,

interactions were observed with in vitro translated p36 (Fig. 4B, lanes 3 and 4) and p38

(Fig. 4B, lanes 11 and 12).

The results of our protein-protein interaction experiments described above

together with published studies have identified a series of pairwise interactions among

PCNA, RFC and DNA ligase I (7,9,11). To examine the effect of RFC on DNA ligase I

13

catalytic activity, we constructed a partial duplex linear substrate containing a single

ligatable nick (Fig. 5A). Pre-incubation of this substrate with RFC inhibited the extent of

DNA joining by about 50% (Fig. 5B). In accord with previous studies (9), PCNA had

essentially no effect on the DNA joining activity of DNA ligase I (Fig. 5B). Interestingly,

when PCNA was pre-incubated with RFC, the inhibitory effect of RFC on DNA joining

was alleviated (Fig. 5B). Since we have shown earlier that the N –terminal region of p140

is involved in the interaction with DNA ligase I (Fig. 3), we asked whether the

catalytically active RFC complex containing a truncated version of p140 lacking the N-

terminal region had a similar effect on DNA ligase I activity. As shown in Fig. 5C, the

RFC complex containing a truncated version of p140 inhibited DNA joining by about

50% but, as with the intact complex, this inhibition was alleviated in the presence of

PCNA. Under these reaction conditions, the RFC complex containing a truncated version

of p140 was still specifically retained by DNA ligase I beads (Fig. 5D), presumably

because of the interactions between DNA ligase I and one or more of the small RFC

subunits (Fig. 4B).

The simplest explanation for the observations described above is that the ATP-

dependent loading of PCNA by RFC induces the dissociation of RFC from the nicked

DNA, thereby making the nick accessible for ligation. Pre-incubation of a circular

ligatable substrate (Fig. 6A) with RFC also caused an inhibition of DNA ligase I activity

(Fig. 6B). In this assay, the initial rate of DNA joining was reduced by about 4-fold but

14

this inhibition was alleviated by the inclusion of PCNA (Fig. 6B). However, in similar

experiments with a mutant variant of DNA ligase I in which the PCNA binding site had

been inactivated (7), preincubation of PCNA with RFC did not alleviate the inhibitory

effect of RFC (Fig. 6C). A similar result was obtained in assay with the linear ligatable

substrate (data not shown). Thus, the ability of DNA ligase I to efficiently join DNA in

the presence of RFC and PCNA is dependent upon its ability to bind to PCNA.

DISCUSSION

There is substantial evidence supporting the notion that multiprotein DNA

transactions such as DNA replication are co-ordinated by protein-protein interactions

among the participating factors. Notably, PCNA, which is a homotrimeric sliding clamp,

interacts with multiple DNA replication factors, indicating that it plays a central role in

directing the sequential actions of these proteins (12,13). In previous studies, we have

shown that the interaction between DNA ligase I and PCNA is critical for the efficient

joining of Okazaki fragments and the completion of the repair of DNA lesions by long

patch BER (7,9). Our failure to find conditions under which PCNA stimulated the

catalytic activity of DNA ligase I (9) prompted us to look for additional protein factors

that may contribute to DNA joining events involving DNA ligase I and PCNA. In this

study, we describe a direct interaction between DNA ligase I and RFC, the clamp loader.

Unlike PCNA binding, which is dependent on a 20-amino acid sequence at the N-

15

terminus of DNA ligase I (11), the interaction with RFC involves residues from both the

non-catalytic N-terminal and the catalytic C-terminal domains of DNA ligase I.

Moreover, DNA ligase I not only interacts with the non-catalytic N-terminal region of

the large subunit of p140 but also with two of the smaller RFC subunits. Together these

results indicate that there are multiple site of contact between DNA ligase I and RFC.

Previously we had shown that PCNA tethered DNA ligase I to a nicked circular

DNA duplex but did not stimulate ligation (9). This suggested that the free-sliding DNA

ligase I-PCNA complex required an additional factor(s) to position it at a nick. In this

regard, it is intriguing that RFC binds to the 3’ hydroxyl termini of primer-template

junctions (27) and so could potentially recruit DNA ligase I to nicks. However, the recent

structures of prokaryotic and eukaryotic RFC complexes determined by electron

microscopy and x-ray crystallography suggest that the binding of RFC would prevent

DNA ligase I from gaining access to the nick (32,33). Consistent with this prediction and

a previous biochemical study (2), RFC inhibited joining by DNA ligase I. This inhibitory

effect was alleviated when PCNA was included in the reaction but only when DNA ligase

I had a functional PCNA binding site. Thus, it appears that the pairwise interactions

among RFC, PCNA and DNA ligase I co-ordinate the joining step that links Okazaki

fragments and completes certain DNA excision repair pathways. The conservation of the

interaction between the functionally homologous Saccharomyces cerevisiae proteins

supports the notion that the interaction between the replicative DNA ligase and the

16

replicative clamp loader is functionally and biologically significant.

Although PCNA is a ring-shaped molecule, the two faces of the ring are not

equivalent. RFC loads PCNA onto DNA in a particular orientation and all of the

replication proteins appear to bind to the same face of the PCNA ring (12,13). Since

PCNA is a homotrimer, it is possible that up to three replication factors can bind to the

same PCNA trimer. Alternatively, the factors may bind sequentially with the DNA

structure presumably dictating their sequential action. Only 1 molecule of DNA ligase I

was bound per PCNA trimer topologically linked to DNA (9). Moreover, we have been

unable to detect formation of a ternary complex of PCNA, DNA ligase I and FEN-1

(Varkey, J. and A.E.T., unpublished results) suggesting that the binding of one molecule

DNA ligase I to a PCNA trimer occludes the other binding sites and that the binding of

DNA ligase I and FEN-1 to a PCNA trimer are mutually exclusive. In Okazaki fragment

maturation, the removal of flaps by FEN-1 linked to PCNA generates the nicked DNA

substrate for DNA ligase I. We propose that flap removal is the signal for the dissociation

of the FEN-1/PCNA complex. Since PCNA is then free to slide away from the nick, we

suggest that RFC acts to maintain PCNA at the nick site by binding to both the nick and

PCNA. It should be noted that, based on the structure of yeast RFC (33), it appears likely

that the interaction of RFC with a 3’ OH terminus will differ depending on whether there

is an adjacent strand or not. Since there are pair wise interactions among PCNA, RFC and

DNA ligase I, we suggest that DNA ligase I initially interacts with a ternary complex of

17

RFC and PCNA bound at a DNA nick via its interaction with the N-terminal region of

the large subunit of RFC. This protein-protein interaction may induce a conformational

change in RFC, enabling DNA ligase I to contact the PCNA ring and catalyze DNA

ligation. Finally, we speculate that nick ligation is the signal for RFC to unload PCNA

and for the dissociation of the ternary complex containing DNA ligase I, PCNA and RFC.

In summary, we have identified a conserved interaction between the replicative

clamp loader and the replicative DNA ligase. Moreover, we have shown that the

interaction between DNA ligase I and PCNA is required for PCNA to overcome the

inhibitory effect of RFC on DNA ligation, suggesting that pairwise physical and

functional interactions among RFC, PCNA and DNA ligase I co-ordinate the DNA

joining step that links adjacent Okazaki fragments. Further studies are needed to delineate

the functional significance of the protein-protein interaction between RFC and DNA

ligase I, and the molecular mechanisms involved in this key reaction in lagging strand

DNA synthesis.

ACKNOWLEDGEMENTS

We are grateful to Drs. Bruce Stillman and Vladimir Podust for reagents. We thank Dr.

Sean Post for help with the immunoprecipitation experiments. This study was supported

18

by grants (GM57479 to AET and GM38559 to JH) from the National Institutes of

Health.

19

REFERENCES

1. Tomkinson, A. E., & Mackey, Z. B. (1998) Mutat. Res. 407, 1-9.2. Waga, S., Bauer, G., & Stillman, B. (1994) J. Biol. Chem. 269, 10923-10934.3. Li, C., Goodchild, J., & Baril, E. F. (1994) Nucleic Acids Res. 22, 632-638.4. Lasko, D. D., Tomkinson, A. E., & Lindahl, T. (1990) J. Biol. Chem. 265, 12618-

12622.5. Prigent, C., Satoh, M. S., Daly, G., Barnes, D. E., & Lindahl, T. (1994) Molec.

Cell. Biol. 14, 310-317.6. Mackenney, V. J., Barnes, D. E., & Lindahl, T. (1997) J. Biol. Chem. 272,

11550-11556.7. Levin, D. S., McKenna, A., Motycka, T., Matsumoto, Y., & Tomkinson, A. E.

(2000) Current Biology 10, 919-922.8. Barnes, D. E., Tomkinson, A. E., Lehmann, A. R., Webster, A. D. B., & Lindahl,

T. (1992) Cell 69, 495-503.9. Levin, D. S., Bai, W., Yao, N, and O’Donnell, M & Tomkinson, A. E. (1997)

Proc. Natl. Acad. Sci. U.S.A. 94, 12863-12868.10. Jonsson, Z., Hindges, R., & Hubscher, U. (1998) EMBO J 17, 2412-2425.11. Montecucco, A., Rossi, R., Levin, D. S., Gary, R., Park, M. S., Motycka, T. A.,

Ciarrocchi, G., Villa, A., Biamonti, G., & Tomkinson, A. E. (1998) EMBO J. 17, 3786-3795.

12. Warbrick, E. (1998) Bioessays 20, 195-199.13. Jonsson, Z., and Hubscher, U. (1997) BioEssays 19, 967-97514. Matsumoto, Y., Kim, K., Hurwitz, J., Gary, R., Levin, D. S., Tomkinson, A. E., &

Park, M. (1999) J. Biol. Chem. 274, 33703-33708.15. Mossi, R., Ferrari, E., & Hubscher, U. (1998) J. Biol. Chem. 273, 14322-14330.16. Tom, S., Henricksen, L. A., Park, M. S., & Bambara, R. A. (2001) J. Biol. Chem.

276, 24817-24825.17. Wang, Y. C., Burkhart, W. A., Mackey, Z. B., Moyer, M. B., Ramos, W., Husain,

I., Chen, J., Besterman, J. M., & Tomkinson, A. E. (1994) J. Biol. Chem. 269, 31923-31928.

18. Fien, K., & Stillman, B. (1992) Molec. Cell. Biol. 12, 155-163.19. Uhlmann, F., Cai, J., Gibbs, E., O’Donnell, M., & Hurwitz, J. (1997) J. Biol.

Chem. 272, 10058-10064.20. Podust, V. N., & Fanning, F. (1997) J. Biol. Chem. 272, 6303-6310.21. Wu, Y., Hickey, R., Lawlor, K., Wills, P., Yu, F., Ozer, H., Starr, R., Quan, J. Y.,

Lee, M., & Malkas, M. (1994) J. Cell. Biochem. 54, 32-46.22. Laemmli, U. K. (1970) Nature 227, 680-685.23. Ron, D., & Dressler, H. (1992) Biotechniques 13, 866-868.24. Bardwell, L., Cooper, A. J., & Friedberg, E. C. (1992) Mol. Cell. Biol. 12, 3041-

20

3049.25. Zhang, G., Gibbs, E., Kelman, Z., O’Donnell, M., & Hurwitz, J. (1999) Proc.

Natl. Acad. Sci. U.S.A. 96, 1869-1874. 26. Li, X., Li, J., Harrington, J., Lieber, M. R., & Burgers, P. M. J. (1995) J. Biol.

Chem 270, 22109-22112.27. Tsurimoto, T., & Stillman, B. (1991) J. Biol. Chem. 266, 1950-1960.28. Lee, S. H., Kwong, A. D., Pan, Z. Q., & Hurwitz, J. (1991) J. Biol. Chem. 266,

594-602.29. Bermudez, V. P., Maniwa, Y., Tappin, I., Ozato, K., Yokomori, K., & Hurwitz, J.

(2003) Proc. Natl. Acad. Sci. U.S.A. 100, 10237-10242.30. Lindsey-Boltz, L. A., Bermudez, V. P., Hurwitz, J., & Sancar, A. (2001) Proc.

Natl. Acad. Sci. U.S.A. 98, 11236-11241.31. Dimitriadis, E. K., Prasad, R., Vaske, M. K., Chen, L., Tomkinson, A. E., Lewis,

M. S., & Wilson, S. H. (1998) J. Biol. Chem. 32, 20540-20550. 32. Miyata, T., Oyama, Y., Mayanagi, K., Ishino, S., Ishino, Y., & Morikawa, K.

(2004) Nat. Struct. Cell Biol. 11, 632-636.33. Bowman, G. D., O’Donnell, M., & Kuriyan, J. (2004) Nature 429, 724-730.

21

FIGURE LEGENDS

Figure 1. DNA ligase I associates with RFC in human cell extracts. A. DNA ligase I

affinity chromatography. A nuclear extract from HeLa cells was incubated with Affi-gel

beads covalently linked to either DNA ligase I (Affi-Lig I) or BSA (Affi-BSA). Bound

proteins were eluted with 150 mM NaCl. The eluate from Affi-Lig I (lane 3) and Affi-

BSA beads (lane 4) were subjected to immunoblotting using the antibodies indicated on

the right. Proteins in the nuclear extract (NE, 50 µg, lane 1) and cytoplasmic extract (CE,

50 µg, lane 2) were detected by direct immunoblotting. B. DNA ligase I (Lig I, lane 3)

and RFC p37 (p37, lane 4) were immunoprecipitated from HeLa cell extracts as

described in Experimental Procedures. Immunoprecipitation with anti-Cdc25C antibody

(mock, lane 2). Proteins in the whole cell extract (WCE, 50 µg, lane 1) and the

immunopecipitates were detected by immunoblotting with the antibodies indicated on the

22

right.

Figure 2. DNA ligase I interacts directly with RFC. A. Purified recombinant proteins.

Lane 1, DNA ligase I, 2 µg; lane 2, RF-C, 1µg; lane 3, PCNA, 2 µg of PCNA. Proteins

were separated by SDS-PAGE and then stained with Coomassie blue. The positions of

molecular mass standards are indicated on the right. B. Purified RFC was incubated either

with Affi-Lig I (lane 2) or Affi-BSA (lane 3) beads as described in Experimental

Procedures. Bound proteins were detected by immunoblotting with the indicated

antibodies. Lane 1, 10% of the input RFC detected by immunoblotting.

Figure 3. The interaction between the large subunit of RFC and the replicative DNA

ligase is conserved between yeast and humans. A. In vitro translated full-length DNA

ligase I was incubated with glutathione sepharose beads liganded by; GST, lane 2; GST

N- terminal p140 (GST-N-p140) encoding amino acid residues 1-584, lane 3; GST C-

terminal p140 (GST-C-p140) encoding amino acid residues 496-1148, lane 4, as

described in Experimental Procedures. Lane 1, 10% of input of labeled DNA ligase I. B.

In vitro translated yeast Rfc1 (RF-C1) was incubated with GST (lane 2) and GST-Cdc9

(lane 3) beads as described in Experimental Procedures. Lane 1, 10% of input of labeled

Rfc1. Labeled bound proteins were separated by SDS-PAGE gel and detected by

PhosphorImager analysis (Molecular Dynamics).

23

Figure 4. Mapping the interacting regions of RFC and DNA ligase I. A. In vitro translated

N-terminal (residues 1-118, N-Lig I) and C-terminal (residues 479-919, C-Lig I) were

incubated with glutathione sepharose beads liganded by; GST, lanes 2 and 5;GST N-

terminal p140 (GST-N-p140), lanes 3 and 6. The eluate from GST beads GST-N-p140

beads (lanes 3 and 6) were analyzed on a 10% SDS-PAGE gel and detected by

PhosphorImager analysis (Molecular Dynamics). Lanes 1 and 4, 10% of the input of the

labeled polypeptide. B. In vitro translated subunits of RF-C consisting of p36, p37, p38,

p40 and p140 were incubated with GST, GST N-terminal Lig I (GST-N-Lig I (1-118))

and GST C-terminal Lig I (GST-C-Lig I (479-919)). The eluate from GST beads (lanes

2, 6, 10, 14, & 18), GST -N–Lig I beads (lanes 3, 7, 11, 15 & 19), GST-C–Lig I beads

(lanes 4, 8, 12, 16, & 20), were analyzed on a 10% SDS-PAGE gel and detected by

Phosphor Imager analysis. Lanes 1, 5, 9, 13 & 17 represent 10% of the translation

reaction mix.

Figure 5. Effect of RFC on DNA joining by DNA ligase I in the presence and absence of

PCNA. A. Schematic illustration of the labeled nicked linear DNA substrate. Two 15-

mer oligonucleotides were annealed to a biotin-labeled 90-mer as described in

Experimental Procedures, so as to generate a partial duplex with a single ligatable nick.

The asterisk indicates the labeled phosphate group. PCNA sliding off the DNA was

prevented by coating the single-stranded regions with RPA and also by blocking the

24

biotin-labeled end with streptavidin beads. B. The linear labeled substrate was incubated

as described in Experimental Procedures with equimolar ratios of RFC FL p140 (an RFC

complex carrying full-length p140, R), PCNA trimer (P) and DNA ligase I wild type

(wL) as indicated. C. The linear labeled substrate was incubated as described in

Experimental Procedures with equimolar ratios of RFC ∆Np140 (an RFC complex

carrying a truncated p140 with 555 amino acids deleted at the N-terminus, ∆R) PCNA

trimer (P) and DNA ligase I (wL) as indicated. Labeled oligonucleotides were separated

by denaturing polyacrylamide gel electrophoresis and quantitated by PhosphorImager

analyis. The data are representative of two independent experiments. D. Purified RFC

∆Np140 complex was incubated with either Affi-LigI (lane 2) or Affi-BSA (lane3) beads as

described in Experimental Procedures. Bound proteins were detected by immunoblotting

with the indicated antibodies. Lane I shows 10 % of the input RFC ∆Np140 complex.

Figure 6. Effect of the PCNA binding region of DNA ligase I on DNA joining in the

presence of RFC. A. Schematic illustration of the labeled nicked circular DNA substrate.

The asterisks indicate the position of the labeled phosphate groups. The labeled nicked

circular DNA substrate (100 fmol) was preincubated with PCNA trimer (P, 1.2 pmol) and

RFC (R, 0.6 pmol) as indicated. After the addition of 90 fmol of either wild type DNA

ligase I (wL) or mutant version of DNA ligase I that does not interact with PCNA (mL),

the incubation was continued. B. Wild type DNA ligase I. C. Mutant version of DNA

ligase I that does not bind to PCNA. Aliquots were taken after 0, 30, 90, and 300 seconds

25

and immediately added to a DNA dye stop solution. Labeled oligonucleotides were

separated by denaturing acrylamide gel electrophoresis and quantitated by

PhosphorImager analysis (Molecular Dynamics). The data are representative of three

independent experiments.

26