Schizosaccharomyces pombe proliferating cell nuclear antigen mutations affect DNA polymerase delta processivity.

We introduced nine site-directed mutations into seven conserved fission yeast proliferative cell nuclear antigen (PCNA) residues, Leu2, Asp63, Arg64, Gly69, Gln201, Glu259, and Glu260, either as single or as double mutants. Both the recombinant wild type and mutant PCNAs were able to form homotrimers in solution and to sustain growth of a null pcna strain (Δpcna). Wild type Schizosaccharomyces pombe PCNA and PCNA proteins with mutations in Asp63, Gln201, Glu259, or Glu260 to Ala were able to stimulate DNA synthetic activity and to enhance the processivity of calf thymus DNA polymerase δ holoenzyme similar to calf thymus PCNA. Mutations of Leu2 to Val or Arg64 to Ala, either singly or as a double mutant, yielded PCNA mutant proteins that had reduced capacity in enhancing the processivity of DNA polymerase δ but showed no deficiency in stimulation of the ATPase activity of replication factor C. S. pombe Δpcna strains sustained by these two mutant-pcna alleles had moderate defects in growth and displayed elongated phenotypes. These cells, however, were not sensitive to UV irradiation. Together, these in vitro and in vivo studies suggest that the side chains of Leu2 and Arg64 in one face of the PCNA trimer ring structure are two of the several sites involved in tethering DNA polymerase δ for processive DNA synthesis during DNA replication.

Proliferating cell nuclear antigen (PCNA) 1 is a multifunctional protein in the cell. It was originally identified as the auxiliary protein for DNA polymerase ␦ during DNA replication (1,2). In vitro, it is able to enhance the DNA synthetic processivity of both DNA polymerases ␦ and ⑀ (3,4) and is essential for synthesizing full-length SV40 DNA replication products (5)(6)(7)(8)(9)(10)(11). PCNA has also been found to be required for DNA excision repair and to be an essential component of the in vitro reconstituted DNA repair system (12)(13)(14). Furthermore, PCNA has been shown to associate with D-type cyclins and to interact with the cyclin kinase inhibitor p21. The interaction between PCNA and p21 leads to inhibition of DNA replication in vitro (15)(16)(17)(18), and this interaction has been shown to specif-ically inhibit the repair of DNA damage caused by either alkylating agents or by ultraviolet radiation (19). PCNA has also recently been found to physically bind flap endonuclease I, which is a 5Ј-flap DNA endonuclease and a nick-specific 5Јexonuclease (20), and to stimulate flap endonuclease I activity. This interaction is thought to be important for lagging strand synthesis and is implicated in a broad range of DNA metabolisms in which the flap endonuclease I nuclease family is involved (21).
The ␤-subunit of Escherichia coli DNA pol III holoenzyme is a functional homologue of PCNA (2,22,23). Budding yeast PCNA and the ␤-subunit of E. coli DNA pol III holoenzyme are structurally similar, one exception being that the ␤-subunit is composed of three globular domains forming a dimeric closed ring, whereas PCNA is composed of only two domains, thus forming a trimeric six-domain ring with a central cavity large enough to encircle DNA (24,25). Structural data have suggested that PCNA forms a homotrimer torus that encircles DNA and interacts with DNA polymerase ␦, tethering it for processive DNA synthesis. These structural data also suggest that the protein loops between the six domains are important for interacting with other proteins for various DNA metabolic processes (24,25).
There is substantial conservation in the primary sequences of PCNA proteins from human, mouse, rice, Drosophila, and budding and fission yeast (reviewed in Ref. 26). Double alanine scan mutagenesis of charged amino acid residues and random mutagenesis of budding yeast PCNA (POL30) have yielded varying degrees of growth defects and DNA damage sensitivity in vivo and defects in protein folding and protein-protein interactions in vitro. The pol30 mutant phenotypes have indicated distinct roles of budding yeast PCNA in DNA replication and repair (27). In addition, structure-function analysis of human PCNA has recently been described. Several amino acids were identified to be involved in stimulation of replication factor C (RF-C) ATPase activity or in enhancing the processivity of DNA synthesis by polymerase ␦ in vitro. Results of the human PCNA study suggest that some residues on the outer surface of PCNA function in protein-protein interactions (26).
To test if conservation of several residues in the protein loop regions of PCNA are essential for biological function, we isolated the gene and cDNA of fission yeast PCNA and overexpressed functionally active recombinant PCNA protein in bacteria. In this study, we analyzed nine site-directed mutations in seven conserved residues of fission yeast PCNA for their function in enhancing DNA polymerase ␦ processivity in vitro and their biological effects in vivo.

EXPERIMENTAL PROCEDURES
Isolation of Schizosaccharomyces pombe PCNA Gene and cDNA-Degenerate oligonucleotides were generated based on conserved sequence regions of PCNA from human, mouse, and budding yeast and used as polymerase chain reaction primers to amplify fission yeast genomic DNA. The resulting polymerase chain reaction products were used to screen a genomic and cDNA library of S. pombe (28,29). An EcoRI genomic fragment containing the entire gene of PCNA was isolated, and a full-length cDNA clone was obtained from multiple overlapping cDNA fragments.
Plasmids-The full-length PCNA gene was cloned into pMT7H6p18 containing the M13 origin and named pMP1. This construct was used for generating site-directed mutants of S. pombe PCNA. A BamHI-SalI fragment containing either the wild type or mutant cDNAs as cassettes from site-directed mutagenesis reactions was constructed into pQE9, which contains a six-histidine tag (his 6 ) at the amino terminus, and plasmids named pQE9/x. An expression vector pREP181 (LEU2, ars1 ϩ ) was generated by modifying the multiple cloning site of pREP81 (30). Wild type and mutant PCNAs again as a BamHI-SalI cassette were constructed into the expression vector pREP181 for in vivo studies, and the resulting clones were named pREP181/x.
Yeast Strains and Methods-Fission yeast media, cell growth, transformation, and standard genetic manipulations were as described (31). Strains constructed and used in this study are listed in Table I. Diploid MAP112 containing a complete deletion of one copy of the PCNA coding sequence by replacement with the his3 ϩ gene was constructed, verified by genomic Southern analysis, and used for in vivo studies. MAP112 was transformed with the expression plasmid pREP181 containing either wild type (pREP181/pcna ϩ ) or mutant PCNA (pREP181/mutant pcna) and sporulated in medium lacking leucine and histidine to select for leucine and histidine prototrophic haploids. Haploids containing the chromosomal ⌬pcna sustained by either pREP181/pcna ϩ or pREP181/ mutant pcnas were analyzed. All cell growth was carried out at 30°C.
Construction of Site-directed Mutants of PCNA-Site-directed mutagenesis was performed as described (32). Mutations were verified by DNA sequencing. BamHI-SalI cassettes containing the mutations were cloned into the bacterial expression vector pQE9 and the yeast shuttle vector pREP181.
Protein Purification of Mutant PCNAs-Constructs containing wild type and mutant PCNAs were transformed into the E. coli M15[pREP4] (Nal s ,Str s ,rif s ,lac Ϫ ,ara Ϫ ,gal Ϫ ,mtl Ϫ ,F Ϫ ,recA ϩ ,uvr ϩ , [LacI]) for overexpression of protein. Cell cultures (500 ml) were grown in 2 ϫ YT medium at either 30 or 37°C to A 600 ϭ 0.7-0.9. Cells were then induced with 1 mM isopropyl-1-thio-␤-D-galactopyranoside and grown for 3 h at 30°C or 25°C. Cells were harvested, resuspended in 20 ml of lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 8.0, 3 mg/ml lysozyme, 1 g/ml aprotinin, and ␣2-macroglobulin), placed on ice for 10 min, and vortexed. All subsequent steps were carried out at 4°C. Cell lysates were spun at 15,000 ϫ g for 15 min. To the supernatants, Ni 2ϩ -nitrilotriacetic acid beads were added at 1 l/ml of cell lysates and rotated end-over-end for 15 min at 4°C. The protein-bound Ni 2ϩ -nitrilotriacetic acid beads were washed three times with high salt buffer (1 M NaCl, 50 mM Tris-HCl, pH 8.0) followed by equilibration in a low salt buffer (150 mM NaCl, 50 mM Tris-HCl, pH 8.0) and then packed into a 10-ml column. The Ni 2ϩ -nitrilotriacetic acid column was subsequently washed with 10 ml of low salt buffer. PCNA protein was eluted with 250 mM imidazole, 50 mM Tris-HCl, pH 9.0. Fractions containing the protein eluate were pooled and dialyzed in successive volumes of either processivity buffer ( Antibodies against S. pombe PCNA and Immunoblot Analysis.-S. pombe PCNA protein overexpressed and purified as described above was partially denatured by incubating half of the sample at 95°C in 0.1 mM dithiothreitol. Both native and denatured proteins of PCNA were used as antigen to immunize a chicken for IgY production (33). The polyclonal IgY specifically against S. pombe PCNA is named anti-FY-PCNA. Immunoblot analysis was performed as described in Ref. 34. After transferring either the wild type or mutant recombinant PCNA proteins (0.5 g) to Immobilon P membrane, they were blotted with anti-FY-PCNA in 1:100 dilution and detected by alkaline phosphataseconjugated anti-chicken IgG (Sigma).
Processivity Measurement-The processivity of calf thymus polymerase ␦ in the presence of calf thymus PCNA and wild type or mutant S. pombe PCNA was measured on poly(dA) 300 -oligo(dT) 16 using the DNA trap method (36,37). The reaction mixture (10 l) contained 100 nM primer-template, 0.5 units of purified calf thymus DNA polymerase ␦, and 0.2 g of either wild type or mutant PCNA. Reactions were commenced by incubation with dTTP for 1 min at room temperature, and DNA trap was subsequently added. Included as controls for each set of incubations were reactions with no dTTP substrate added, reactions with dTTP substrate but no DNA trap added, and reactions with dTTP substrate and DNA trap added simultaneously. The processivity is measured as the mean number of residues added by a DNA polymerase ␦ to a primer terminus per cycle of nucleotide addition as described in Refs. 36 and 37.
ATPase Assay-The assay for RF-C ATPase was as described (38)  Microscopy-Cells were ethanol fixed and stained with DAPI for microscopic analysis as described (39,40).
UV Radiation Measurements-UV sensitivity analysis was performed by plating a known density of mid log phase cells onto appropriate agar plates and exposing cells to UV irradiation at different

RESULTS
Mutation of S. pombe PCNA-Nine site-directed mutations were introduced into 7 conserved residues, Leu 2 , Asp 63 , Arg 64 , Gly 69 , Gln 201 , Glu 259 , and Glu 260 of S. pombe PCNA. The location of these seven residues in the PCNA structure are depicted in Fig. 1. Leu 2 is located at the amino terminus and is conserved among PCNA proteins of fission and budding yeast and rice. In mammalian cells and Drosophila, a hydrophobic residue, Phe, instead of Leu is located at this position. Asp 63 and Arg 64 are located in the protein loop between ␤-sheets ␤E1 and ␤F1 of budding yeast PCNA (25). Asp 63 and Arg 64 are conserved among PCNAs of mammalian cells, rice, Drosophila, and fission yeast, whereas budding yeast has a polar residue, His, in place of the charged residue, Arg. Gly 69 is not located in a protein loop region but is in ␤-sheet ␤F1. Gly 69 is highly conserved among PCNAs of all species. Mutation of this residue in budding yeast PCNA to Asp has been shown to suppress cold-sensitive alleles of CDC44, which encodes the large subunit of budding yeast replication factor C (41). Thus, Gly 69 of S. pombe PCNA was selected for similar mutation to Asp to test the mutational effect. Gln 201 is conserved only between budding and fission yeast. PCNA from most other species has the similar size residue Glu instead of Gln at this position. Gln 201 is located in the second domain of the protein loop between ␤E2 and ␤F2. Glu 259 and Glu 260 are located at the carboxyl terminus. Leu 2 was conservatively changed to Val, and Gly 69 was changed to Asp like that in budding yeast (41), whereas Asp 63 , Arg 64 , Gln 201 , Glu 259 , and Glu 260 were all mutated to Ala to abolish both the length and charge of the side chains in order to test their function. A double mutation of the two adjacent charged residues Glu 259 and Glu 260 was generated to test the combinational charge effect of these two carboxyl terminus residues. The three-dimensional structure of budding yeast PCNA indicates that residues Leu 2 , Asp 63 , and Arg 64 are located in one region of the PCNA homotrimer. The side chains of Leu 2 and Arg 64 are orientated toward the outer rim of one face of the PCNA molecule, whereas the side chain of Asp 63 is orientated in a different direction (Fig. 1). Thus, a double mutant L2V/R64A was also generated to test the combinational effect of mutation of both Leu 2 and Arg 64 .
Wild Type and Mutant PCNAs Can Form a Homotrimer Ring--Wild type and mutant PCNA proteins were expressed in E. coli and affinity purified to near homogeneity. Although the cDNA of S. pombe PCNA has a predicted molecular mass of 29 kDa, the recombinant PCNA resolves as a 36-kDa protein in SDS gel analysis. The purified proteins were further verified by immunoblotting with antibodies against S. pombe PCNA (Fig.  2). To test if the wild type or mutant PCNA proteins had any significant physical alterations, these were analyzed by gel filtration for trimer formation. The recombinant wild type PCNA protein and all of the PCNA mutant proteins eluted mainly as trimers with a molecular mass of 90 kDa with minor quantities of the protein as monomer with a molecular mass of 29 kDa (Fig. 3, upper panel). SDS gel analysis of gel filtration fractions of the wild type and mutant PCNA proteins L2V, D63A, R64A, L2V/R64A, Q201A, E259A, E260A, and E259A/ E260A showed the majority of protein in homotrimer form with the peak fraction of the protein eluting between fractions 27 and 29 (Fig. 3, lower panel, and data not shown). These results strongly suggest that the mutations introduced into these invariant residues either as single or as double mutants have no significant effect on homotrimer formation of the PCNA protein. Furthermore, all these eight PCNA proteins were expressed with a his-tag at the amino terminus. The ability of these PCNA proteins to form the homotrimer ring in solution strongly suggests that the amino-terminal his-tag does not affect the homotrimer ring formation (Fig. 3).
Mutants L2V, R64A, and L2V/R64A Have Reduced Ability to Stimulate DNA Polymerase ␦-The ability of heterospecies PCNA to stimulate calf thymus DNA polymerase ␦ in vitro has been previously demonstrated (42)(43)(44). Using poly(dA)-oligo(dT) as primer template, in vitro DNA synthetic activity of the highly purified two-subunit calf thymus DNA polymerase ␦ holoenzyme was analyzed either in the absence of PCNA, with calf thymus PCNA, or with wild type S. pombe PCNA. Wild type S. pombe PCNA was able to stimulate calf thymus DNA polymerase ␦ similar to calf thymus PCNA (Fig. 4A). Mutants D63A, Q201A, E259A, and E260A and the double mutant E259A/E260A were all able to stimulate DNA synthesis by polymerase ␦ at levels similar to that of wild type S. pombe PCNA (Table II). In contrast, mutants L2V and R64A showed approximately 50% of the wild type PCNA's ability to stimulate DNA polymerizing activity of polymerase ␦. The double mutant, L2V/R64A, had synergistic effects resulting in less than 10% of the wild type PCNA's ability in stimulating DNA polymerizing activity of polymerase ␦ (Fig. 4B and Table II). The deficiency of mutant PCNAs L2V, R64A, and L2V/R64A to stimulate DNA polymerase ␦ is specific to Leu 2 and Arg 64 in this particular region of PCNA because mutant D63A, despite being a highly conserved residue adjacent to Arg 64 and located in the same protein loop, had wild type PCNA capacity in stimulating DNA polymerase ␦ activity (Table II). These results suggest that the side chains of Leu 2 and Arg 64 but not the side chain of Asp 63 are specific in stimulating the DNA synthetic activity of DNA polymerase ␦. Lastly, mutant G69D was unable to stimulate DNA synthetic activity of polymerase ␦ beyond that of the holoenzyme alone. The abilities of each PCNA mutant to stimulate polymerase ␦ DNA synthetic activity are summarized in Table II.
Mutants L2V, R64A, and L2V/R64A Have Reduced Capacity to Enhance DNA Polymerase ␦ Processivity but Are Not Deficient in Stimulating the ATPase Activity of RF-C-The processivity of DNA polymerase ␦ was measured to test whether the deficiency of PCNA mutants L2V, R64A, and L2V/R64A to stimulate polymerase ␦ DNA synthesis activity was due to their inability to enhance DNA polymerase ␦ processivity. Using poly(dA) 300 -oligo(dT) 16 as primer template, polymerase ␦ processivity was analyzed by a DNA trap method either in the absence of PCNA or in the presence of calf thymus PCNA, wild type S. pombe PCNA, or S. pombe mutant PCNAs L2V, R64A, L2V/R64A, and D63A (Fig. 5). In the absence of added PCNA, DNA polymerase ␦ incorporated merely 3 nucleotides per binding of the primer terminus. In the presence of calf thymus PCNA, polymerase ␦ was able to incorporate Ͼ30 nucleotides per binding event of the primer terminus. With wild type S. pombe PCNA, DNA polymerase ␦ showed a processivity of Ͼ20 nucleotides. In contrast, with mutant L2V, polymerase ␦ showed a greater than 50% reduction of its processivity as compared with the wild type PCNA with only 9 nucleotides incorporated per primer binding event. With mutant R64A, DNA polymerase ␦ had a 25% reduction in its processivity incorporating, 15 nucleotides per primer binding event. With the double mutant, L2V/R64A, DNA polymerase ␦ showed a processivity of 5 nucleotides incorporated per primer binding event, which was nearly that of polymerase ␦ holoenzyme alone with no PCNA added. Again, the reduced capacity of mutants L2V and R64A and the double mutant L2V/R64A to enhance polymerase ␦ processivity is specific, because mutant D63A enhanced polymerase ␦ processivity similar to wild type S. pombe PCNA with Ͼ20 nucleotides incorporated per primer binding event. These results, quantitatively summarized in Table III, strongly suggest that the side chains of Leu 2 and Arg 64 play a role in tethering DNA polymerase ␦ for processive DNA synthesis in vitro.
Similar to the E. coli pol III ␤-subunit, PCNA requires loading onto chromosomal DNA by the accessory protein RF-C in order to interact with polymerase ␦ for processive DNA synthesis in vivo. In vitro studies have shown that ATP hydrolysis promotes the interaction between PCNA and RF-C for this loading (10,45). To test if mutations in Leu 2 and Arg 64 affect the ability of PCNA to be loaded onto DNA by RF-C, we assayed L2V, R64A, and the double mutant L2V/R64A for their ability to stimulate RF-C ATPase activity. Wild type S. pombe PCNA was found to stimulate the ATPase activity of budding yeast RF-C similar to budding yeast PCNA. Mutants L2V, R64A, and L2V/R64A stimulated the ATPase activity of RF-C similar to wild type fission and budding yeast PCNA (Fig. 4C). These results suggest that mutations in Leu 2 and Arg 64 of fission yeast PCNA do not affect the interaction of PCNA with RF-C but do specifically affect PCNA's interaction with DNA polymerase ␦.
Recombinant pcna ϩ Is Able to Complement a Null pcna Strain-A diploid strain MAP112 containing one null allele of  PCNA by replacement of pcna ϩ with the his3 ϩ marker was generated (see "Experimental Procedures"). After sporulation in medium lacking histidine, only spores prototrophic for histidine and containing ⌬pcna germinated. Germinating spores were stained with DAPI, examined microscopically and compared with the parental wild type strain (Fig. 6A). Germinating spores with ⌬pcna displayed an elongated cell division cycle (cdc) phenotype similar to that previously described (46) (Fig. 6B).
To test whether the wild type recombinant pcna ϩ isolated in this study was functional in vivo, recombinant pcna ϩ was constructed into an expression vector containing the LEU2 selectable marker (pREP181) in which expression is controlled from a weakened nmt1 promoter (29,47,48). pREP181/pcna ϩ was transformed into the diploid MAP112. Upon sporulation and selection for histidine and leucine prototrophs, a haploid strain MAP300 was generated with ⌬pcna sustained by pREP181/ pcna ϩ . MAP300 had a normal growth rate (see results described below) and phenotype compared with the parental wild type strain (Fig. 6C). This indicates that the recombinant pcna ϩ isolated in this study is functional in vivo.
Cells   Tables I and IV. Strains MAP300, MAP301, MAP304, MAP305, and MAP326 were grown in medium lacking thiamine to allow full expression of each of the mutant-pcnas. Microscopic examination of DAPI-stained cells showed that MAP300 sustained by wild type pcna ϩ and MAP304 cells sustained by the mutant pcna-4 allele (with D63A mutation) (see Table IV) had wild type phenotypes (Fig. 7, A and B). Mutant strains MAP301, MAP305, and MAP326, sustained by pcna-1 (with L2V mutation), pcna-5 (with R64A mutation), and pcna-26 (with L2V/R64A mutations) alleles, respectively, displayed elongated phenotypes with cells on average approximately 5-10 m longer than cells sustained by either wild type pcna ϩ or pcna-4 (Fig. 7, compare  C, D, and E with A and B).
Strain MAP306, sustained by mutant allele pcna-6 (with G69D mutation), failed to grow beyond one or two cell divisions and displayed a phenotype similar to ⌬pcna (Fig. 6B), generating highly elongated cells (depicted by arrow in Fig. 7F). Thus, mutation of S. pombe PCNA Gly 69 to Asp not only yields mutant protein unable to stimulate polymerase ␦ DNA synthetic activity in vitro (Table II) but also fails to sustain growth of cells containing ⌬pcna in vivo.

Cells Sustained by Mutant-pcna That Are Defective in Enhancing DNA Polymerase ␦ Processivity in Vitro Exhibit Moderate Growth Defects but No Altered Sensitivity to UV-We
analyzed the growth rate of the mutant strains that are defective in enhancing DNA polymerase ␦ processivity and display elongated phenotypes. MAP300 (with ⌬pcna sustained by pREP81pcna ϩ ) and the parental strain KG3 had generation times of ϳ2.5 h at 30°C, whereas MAP301 (with L2V mutation), MAP305 (with R64A mutation), and MAP326 (with L2V/ R64A mutations) had reproducible generation times 40 min longer than MAP300 or KG3 (Fig. 8). In contrast MAP304, sustained by the pcna-4 allele (with D63A mutation), had a growth rate similar to MAP300 (data not shown). These results indicate that strains sustained by mutant pcna alleles encoding PCNA proteins defective in their ability to enhance polymerase ␦ processivity in vitro not only display elongated phenotype but also have moderate growth defects in vivo.
Because PCNA has been shown to play a role in DNA repair in vitro, strains MAP301, MAP304, MAP305, and MAP326 were also tested for their sensitivity to UV irradiation, a DNA damaging agent. MAP301, MAP304, MAP305, and MAP326 showed no significant difference in UV sensitivity compared with MAP300 or the parental strain (Fig. 9). Under the same conditions, rad26 showed high UV sensitivity. DISCUSSION To begin to understand how PCNA interacts with replication factors for S phase progression, we used the S. cerevisiae PCNA structure (25) as a guiding model and performed structurefunction analyses of S. pombe PCNA both in vitro and in vivo.
Rationale for the Mutations Introduced-Structural data of the E. coli pol III ␤-subunit and budding yeast PCNA suggest that residues in the protein loops are most likely involved in protein-protein interactions, whereas residues in either the ␤-sheets or ␣-helices are important for maintenance of the ring structure of these two molecules (24,25). We introduced nine site-directed mutations into seven residues located in the protein loop regions of S. pombe PCNA. The rationale for the   mutations is to target those residues in the protein loop regions that are conserved among PCNA proteins of different phylogenetic species. Furthermore, the mutations introduced must not alter the global physical homotrimer ring structure of PCNA. Leu 2 , Asp 63 , and Arg 64 have been chosen for mutational analysis because these three residues are conserved and clustered in one region of PCNA (Fig. 1). Because Leu 2 is located at the amino terminus, which might be important for the homotrimer ring formation of PCNA, Leu 2 was conservatively changed to another hydrophobic residue, Val, to test length functionality of the leucine side chain. Based on the structure data of budding yeast PCNA (25), Asp 63 and Arg 64 side chains protrude in different directions, so these two residues were mutagenized to Ala to test not only the functionality of the side chains but also the effect of side chain directionality on protein interactions.
Gly 69 is located within ␤F1, and mutation of Gly 69 to Asp yielded a mutant that was unable to stimulate the DNA synthetic activity of polymerase ␦ (Table II). In addition, ectopically expressed G69A mutant PCNA was unable to sustain growth of null pcna cells, displaying a severely elongated phenotype similar to null pcna cells, suggesting a loss of function mutation (Fig. 7F). Mutant G69D protein expressed in bacteria was mostly insoluble, and the quantities of soluble G69D protein were insufficient to be analyzed for structural integrity by gel filtration. However, the in vivo phenotype of this mutant suggests that mutation of Gly 69 to Asp in ␤F1 causes some structural perturbation resulting in nonfunctional PCNA protein. In budding yeast, the pol30 -36 allele (G69D mutation) is able to suppress a cold-sensitive mutation of CDC44 (cdc44-cs) but does not affect growth of budding yeast cells independent of cdc44-cs (41). Our in vivo and in vitro results suggest that the pcna-6 allele (G69D) in S. pombe is lethal. Sequence analysis has indicated that there is no other mutations in the strain besides G69D mutation. Thus, this is contrary to that seen in budding yeast.
Structural Integrity of Mutant PCNAs-Mutations introduced into S. pombe PCNA in this study were targeted to residues that would not be expected to affect the overall structure of PCNA with the exception of G69D. The structural integrity of the mutants was proven by two criteria. First, gel filtration analysis of six single mutant PCNA proteins L2V, D63A, R64A, Q201A, E259A, and E260A as well as the two double mutants E259A/E260A and L2V/R64A showed that all of these mutant PCNA proteins maintained the homotrimer ring structure (Fig. 3 and data not shown). Second, each of these six single mutants and the two double mutants were able to sustain growth of ⌬pcna cells. These results indicate that the introduced mutations do not perturb the homotrimer torus structure of S. pombe PCNA.
Hetero-species Interactions-Several lines of evidence have shown that PCNA is able to interact with protein factors of hetero-species. The primary sequence of S. pombe PCNA is 52% identical to human and rat PCNAs. The disrupted S. pombe PCNA gene can be functionally complemented by ectopically expressed human PCNA (46). Furthermore, Drosophila PCNA with approximately 70% identity to the human protein (49) is able to substitute for human PCNA in SV40 replication in vitro (43). Budding yeast PCNA, with 35% identity to the human protein, is able to enhance the processivity of mammalian DNA polymerase ␦ in vitro but cannot substitute for human PCNA in SV40 in vitro replication (42). Nonetheless, it has been shown that PCNA proteins and polymerase ␦ from budding yeast and calf thymus are mutually interactive for processive DNA synthesis (42,50). In this study, we showed that S. pombe PCNA is able to stimulate calf thymus DNA polymerase ␦ enzymatic activity and enhance calf thymus DNA polymerase ␦ processivity similar to calf thymus PCNA (Figs. 4A and 5). Furthermore, S. pombe PCNA is able to stimulate the DNA-dependent ATPase activity of budding yeast RF-C similar if not identical to budding yeast PCNA (Fig. 4C). These results strongly sug-gest that particular sites or domains of the PCNA protein are functionally conserved and able to interact with DNA replication and/or repair proteins from hetero-species.
We have overexpressed an enzymatically active catalytic subunit of S. pombe DNA polymerase ␦ in recombinant baculovirus-infected insect cells. The recombinant S. pombe PCNA described in this study, however, was unable to stimulate the single catalytic subunit of S. pombe polymerase ␦. 2 This is similar to the observation in Drosophila, where PCNA is unable to stimulate the DNA polymerizing activity of the single catalytic subunit of polymerase ␦ (51), and to the findings in mouse (52). In contrast, budding yeast PCNA is able to stimulate the budding yeast polymerase ␦ single catalytic subunit overexpressed in bacteria (53). Because S. pombe PCNA is unable to stimulate the single catalytic subunit of homo-species polymerase ␦, we used calf thymus polymerase ␦ for our study.
Leu 2 and Arg 64 Are Involved in Enhancing DNA Polymerase ␦ Activity and Processivity in Vitro-We identified two residues, Leu 2 and Arg 64 , that are specifically involved in interacting with polymerase ␦ for processive DNA synthesis but not with RF-C (Figs. 4, B and C, and 5 and Tables II and III). According to the structural data of S. cerevisiae PCNA (25), the side chain directions of Leu 2 and Arg 64 are also specific for interaction with polymerase ␦, because mutation of Asp 63 , adjacent to Arg 64 but with its side chain oriented in a different direction than Arg 64 , did not affect the processivity of polymerase ␦ (Fig. 5 and Tables II and III). This suggests that the face of PCNA with the protruding Leu 2 and Arg 64 side chains is involved in the functional interaction between polymerase ␦ and PCNA. These results are supported by the in vitro structure-function analysis of human PCNA by Fukuda et al., which identified residues on the outer surface and carboxyl-and amino-terminal regions of PCNA that are responsible for interaction with multiple DNA replication factors (26). In the study of human PCNA, mutation of Arg 64 but not Asp 63 to Ala reduced the ability of mutant PCNA to stimulate human DNA polymerase ␦ activity by 50%. However, neither mutation had demonstrable effects on the ATPase activity of RF-C. Our findings that null pcna cells sustained by mutant PCNA R64A display abnormal elongated phenotypes and have moderate growth defects, further supports the notion that Arg 64 indeed has a role in interacting with polymerase ␦ for processive DNA synthesis.
Mutations of Gln 201 , Glu 259 , and Glu 260 and double mutations of Glu 259 and Glu 260 to Ala did not appear to affect the DNA polymerizing activity of calf thymus DNA polymerase ␦ (Table II). Because a hetero-species polymerase ␦ was used in our study, we cannot rule out the possibility that these residues may have species specificity for interacting with polymerase ␦. Deletion analysis of a short carboxyl-terminal stretch of human PCNA from Lys 254 to Glu 256 was found to be necessary for stimulation of RF-C ATPase activity but not for stimulation of DNA polymerase ␦ activity (26). A double mutation, D256A/ E257A, of acidic residues at the carboxyl-terminal region of budding yeast PCNA equivalent to those that were introduced in our fission yeast study have been analyzed in vivo. A strain containing the mutant allele pol30 -22 showed slightly higher sensitivity to UV and methylmethane sulfonate than the wild type strain but had a growth rate similar to wild type. However, this budding yeast PCNA mutant has not been studied in vitro for its effect on polymerase ␦ activity (27). Preliminary in vivo studies of S. pombe PCNA with single or double mutations of Glu 259 and Glu 260 to Ala show that mutations in these two residues do not significantly effect growth similar to that ob- served with the budding yeast pol30 -22. mutant (27). Together, these budding and fission yeast results suggest that these charged residues at the carboxyl-terminal region might not play a role in S phase progression.
In Vivo Effect of Ectopically Expressed pcna ϩ and pcna Mutants in Null pcna Cells-Waseem et al. (46) previously reported that neither pcna ϩ nor POL30 expressed from pREP1 were able to complement ⌬pcna cells. Moreover, overexpression of pcna ϩ rendered cells with increased generation times and abnormal phenotypes (46). We found that pcna ϩ isolated in this study and expressed from a weakened nmt1 promoter (pREP81 or pREP181) is able to complement ⌬pcna cells under fully induced conditions (Fig. 6). This indicates that although overexpression of PCNA from pREP1 at levels approximately 400fold greater than endogenous leads to abnormal germination and inability to rescue ⌬pcna cells (46), moderate expression of PCNA at levels 7-10-fold over endogenous (47) does not appear to compromise germination of ⌬pcna spores nor growth of ⌬pcna cells.
Strains MAP301, MAP305, and MAP326, sustained by ectopically expressed PCNA mutants that are defective in vitro for enhancing polymerase ␦ processivity, have abnormal elongated phenotypes and slower growth rates (Figs. 7 and 8). These findings suggest that cells sustained by mutant PCNA proteins that have a reduced capacity to enhance polymerase ␦ for processive DNA synthesis in vitro might cause a delay in S phase progression, thus resulting in longer doubling times and elongated phenotypes in vivo. The moderate in vivo effects of Leu 2 , Arg 64 , and Leu 2 /Arg 64 mutants suggest that Leu 2 and Arg 64 are only two of several residues that interact with polymerase ␦ for processive DNA synthesis. These in vivo results reflect the in vitro results that mutations in Leu 2 or Arg 64 reduce the capacity of mutant PCNA to stimulate DNA polymerase ␦ activity but do not abolish stimulation completely ( Fig.  4B and Table II).
The lack of UV sensitivity of MAP301 (with L2V mutation), MAP305 (with R64A mutation), and MAP326 (with L2V/R64A mutations) suggests that the observed moderate growth defects and elongated phenotypes of these mutant strains are primarily due to defects in replication and not to UV irradiation repair defects.
Conclusion-In this study, we defined two residues in a region on one face of the S. pombe PCNA trimer ring structure that are involved in interacting with polymerase ␦ for processive DNA synthesis. Correlating with the in vitro findings, S. pombe cells containing these mutant alleles have elongated phenotypes and moderate growth defects. Cells with these mutant alleles, however, have no apparent altered sensitivity to UV. Together, these in vitro and in vivo results strongly suggest that residues Leu 2 and Arg 64 , with their side chains protruding toward one face of the S. pombe PCNA trimer ring, functionally interact with DNA polymerase ␦ for DNA replication in cells.