A previous study reported cloning of a partial cDNA for the catalytic subunit (p260) of Xenopus Pol ε (xPol ε) 24, and in this report, the corresponding full-length cDNA was cloned by 5'

R

a

c

e

Deletion mutants of xPol ε p260 were constructed and used to identify protein regions involved in the interfaces between p260 and the other subunits of Pol ε holoenzyme. For these experiments, p260 deletion mutants were fused to the FLAG epitope tag (Fig. 1A) and the following mutants were generated: (a) FLAG-p260full, (b) FLAG-p260ΔC2157, (c) FLAG-p260ΔC1054, (d) FLAG-p260D860N, and (e) FLAG-p260ΔCat (equivalent to yeast Pol2-16p 151627). The p260 mutants were co-expressed with xPol ε p60, p12 and p17 using a baculovirus insect cell expression system and immunoprecipitated with anti-Flag antibodies. Figure 1B shows that anti-Flag antibodies co-immunoprecipitated p260, p60, p12, and p17 from crude extracts of insect cells expressing wild type p260, p60, p12, and p17. In contrast, in cells expressing p260ΔC2157, p12 and p17 co-immunoprecipitated with the catalytic subunit but p60 did not, and in cells expressing p260ΔC1054, no other subunits of Pol ε holoenzyme co-immunoprecipitated with the p260 mutant protein. These results suggest that p60, but not p12 or p17, interacts with p260 residues 2157 to 2285, which includes Zinc-finger motifs (Fig. 1C), and that p12 and p17 interact with a motif or region between residues 1054 and 2157 of p260 (Fig. 1C). Both p260D860N (a DNA polymerase-dead-point mutant) and p260ΔCat did not interfere with formation of the xPol ε holoenzyme, because all four Pol ε subunits co-immunoprecipitated from cells expressing these p260 mutants. In cells expressing histidine-tagged p60 (H-p60), p12, and p17, Ni²⁺ -chelating beads only precipitated H-p60, indicating that H-p60 does not interact directly with p12 or p17, or both subunits (Fig. 3), although it was known that both subunits interact each other and form a 1:1 complex in insect cells (data not shown) as S. cerevisiae Dpb3p-Dpb4p complex 28.

Previous studies suggested that the amino-terminal portion of yeast Pol ε is dispensable for DNA replication, DNA repair, and viability. Because this portion of the enzyme includes all known DNA polymerase and exonuclease motifs, these results suggest that the DNA polymerase activity of Pol ε is dispensable for chromosomal DNA replication in yeast 151627. To investigate this possibility further, the xPol ε mutant complexes described in Figure 2 as well as partial holoenzyme complexes lacking specific subunits were purified from insect cells and used in an in vitro Xenopus chromosomal DNA replication system with Pol ε-depleted Xenopus egg extracts. The following xPol ε holoenzyme complexes were used: (a) wild type r-xPol ε holoenzyme, (b) Pol ε-DN containing p260D860N, (c) Pol ε ΔCat containing p260ΔCat, (d) p260p60 sub-complex of xPol ε, (e) p260-p12-p17 sub-complex of xPol ε, and (f) p260 (Fig. 2 and 3). The specific activity of each enzyme complex is summarized in Table 1. Although we did not measure the specific activity of p260-p60 and p260-p12-p17 sub-complexes, they were similarly purified as r-xPol ε holoenzyme and p260 and might be the same specific activity as that of p260 alone. Pol ε-DN and Pol ε ΔCat holoenzymes do not have any significant DNA polymerase activity as predicted, although a slight contaminating DNA polymerase activity could be detected (Table 1). This may be due to contamination of insect DNA polymerase activity. It should be noted here that recombinant xPol ε holoenzyme complexes (r-xPol ε) purified from insect cells had minor contaminants, and all enzyme preparations except Flag-tagged p260 included p60, p12, and p17 in amounts that were detected by immunoblotting.

Figure 4 shows that the extent and rate of DNA synthesis were greatly reduced in Pol ε-depleted Xenopus extracts, and this defect was fully complemented by addition of purified native (n-xPol ε)(data not shown, and 22) or wild type recombinant xPol ε (r-xPol ε). In contrast, addition of xPol ε ΔCat or xPol ε DN to Pol ε-depleted Xenopus extracts only partially (30–40%) complemented the replication defect and did not restore normal kinetics of the initiation of DNA replication (Fig. 4). Furthermore, neither p260, p260-p12-p17 nor p12-p17 complemented the defect in DNA synthesis in xPol ε-depleted Xenopus extracts. Of the partial holoenzyme complexes tested, only the p260-p60 sub-complex of xPol ε significantly complemented the defect (Fig. 4). These results clearly demonstrate that the DNA polymerase activity of xPol ε holoenzyme is required for normal, rapid and efficient chromosomal DNA replication in Xenopus egg extracts.

Previous studies suggested that Cdc45p, Cut5p, and GINS are critical for loading Pol α and Pol ε onto DNA replication origins and for initiating chromosomal DNA replication in yeast and Xenopus egg extracts 22224293031. However, those studies did not show any direct interaction between these initiation proteins and Pol ε holoenzyme. Therefore, in this study, we tested this possibility by measuring physical interactions between r-xPol ε holoenzyme and xCdc45p, xCut5p, and xGINS. The experiments were performed by incubating r-xPol ε holoenzyme with these proteins and immunoprecipitating protein complexes with anti-FLAG antibody. Figure 5A shows that xGINS was co-immunoprecipitated with Pol ε holoenzyme from reactions containing r-xPol ε holoenzyme and recombinant xGINS 29. Addition of xCdc45p, but not xCut5p 31, slightly disrupted the complex between r-xPol ε holoenzyme and xGINS (Fig 5C), and incubation of xCdc45, xCut5p, and xGINS in the absence of r-xPol ε holoenzyme resulted in formation of a heterotrimeric complex lacking xPol ε holoenzyme (Fig. 5B). xCdc45p also interacts directly with xPol ε holoenzyme in vitro (Fig. 5C), and the amount of this complex decreased slightly in the presence of xGINS, but not in the presence of xCut5p (Fig. 5C). These results suggest that the trimeric xCdc45/xCut5p/xGINS complex interacts with xPol ε holoenzyme, and this interaction may involve direct contacts with xCdc45p and xGINS but not with xCut5p (Fig. 5D).

Previous studies demonstrated a direct interaction between yeast GINS and yeast Pol ε holoenzyme and that this interaction stimulates the rate and processivity of DNA synthesis by Pol ε holoenzyme 32. A similar result was obtained using r-xPol ε and r-xGINS 29. As observed previously for yeast GINS, maximum stimulation by xGINS was obtained at a stoichimetry of 5–10 xGINS/xPol ε (Fig. 6). Under these conditions, however, xCdc45p or xCdc45p plus xCut5p did not stimulate the rate or extent of DNA synthesis by r-xPol ε holoenzyme (see Additional file 5).

The cDNA for the p260 subunit of Xenopus laevis Pol ε (GenBank accession no. AB259046) was isolated by 5'-RACE using Xenopus laevis ovary mRNA as templates and oligonucleotides primers synthesized based on the nucleotide sequence information of a partial cDNA for the p260 subunit of Xenopus Pol ε 24. Both strands of its cDNA insert were sequenced with the use of an Applied Biosystems Prism dye terminator cycle sequencing kit and a DNA sequencer (ABI377). The initiation methionine was postulated on the basis of a comparison with the amino acid sequence of HeLa Pol ε p260 (gi:62198237). For cloning of the p17 subunit of XPol ε, the IMAGE cDNA clone 4783571(5') (GenBank ACC# BI349483) was obtained and sequenced. This clone did not contain a complete cDNA insert, so the 3' portion of p17 cDNA was cloned by 3' RACE using Xenopus mRNA prepared from Xenopus ovary (GenBank ACC# AB259047). For cloning of the p12 subunit of XPol ε, the cDNA clone PBX0037B02(5') (GenBank ACC# AW635703) was obtained and sequenced (GenBank ACC# AB259048).

Xenopus ovary mRNA was prepared as published 37, and was used for cloning of the 3'-untranslated region by using 3'-Full RACE Core Set (TAKARA, Japan). The PCR products were cloned into pBR322-based plasmid DNA and sequenced. The 5'-terminal sequence was obtained by 5'-Full RACE Core Set (TAKARA, Japan) using Xenopus ovary mRNA as templates.

Rabbit anti-Xenopus Pol ε p60 antibodies were affinity-purified as published 22 and used for making antigen-immobilized Affi-Gel 15 (Bio-Rad). The purified p60 antibodies or whole rabbit IgG (Pierce) as a control was crosslinked to Affi-Prep Protein A matrix (Bio-Rad) (1 mg of IgG per ml of matrix) and used for immunoprecipitation and immunodepletion experiments. Rabbit anti-Xenopus Pol ε p12 and p17 antibodies were raised by using p12 and p17, which were expressed in E. coli and purified, as antigens (K. Shikata, unpublished).

Xenopus laevis egg extracts (low-speed supernatant) were prepared as described previously 22. Immunodepletion was performed by mixing egg extracts three times with the antibody-crosslinked matrix at 4°C. DNA replication with membrane-removed sperm nuclei (2,000 sperm heads per ml of extract) was carried out at 23°C in the presence of [α-³²P]dATP as described 22. The reaction products were purified by RNase A digestion, proteinase K digestion, and phenol-chloroform extraction followed by ethanol precipitation and then separated by 0.8% agarose gel electrophoresis under neutral (Tris/borate/EDTA buffer) condition as described previously 22. After electrophoresis, the gel was fixed, dried, and subjected to autoradiography. The quantification of replication products was carried out with a Fuji image analyzer (BAS1500).

For expression of proteins, 150 × 10⁶ log phase sf21 insect cells were grown in suspension tissue culture flasks. The cells were infected with viral supernatant at a multiplicity of infection of 10 for p260 (a catalytic subunit of xPol ε complex), 5 for p60 (the second subunit of xPol ε complex), 5 for p12 (the third subunit of XPol ε complex), and 5 for p17 (the fourth subunit of xPol ε complex). After a 2-day incubation at 27 °C, cells were harvested, washed once with cold 1 × Tris-buffered saline, flash frozen in liquid N₂, and stored at -80 °C. The frozen cells were thawed once on ice and then resuspended in 5 ml of buffer A (50 mM Tris, 150 mM NaCl, 5 mM EDTA, 20% glycerol, pH 7.5) containing 20 μg/ml leupeptin. After lysis, cells were kept on ice for 15 min and centrifuged at 12,000 rpm in a microcentrifuge for 10 min, and supernatant was recovered. Pol ε complex was purified from insect cell extracts with DEAE sepharose column, followed by anti-FLAG antibody column. Native Xenopus Pol ε (n-xPol ε) was purified from Xenopus egg extract by 4 sequential column chromatographies as published 22. Throughout the purification, column fractions were assayed for DNA polymerase activity with the use of [α-³²P]dTTP and oligo(dT)₁₀/poly(dA)₄₀₀ (1:19; 0.04 mM nucleotides) as a primer/template and were analyzed by Western blotting. One unit of Pol ε supported the incorporation of 1 nmol of dTMP under the conditions described above.

Immunodepletion of xPol ε was performed by mixing Xenopus egg extracts three times with the xPol ε-p60-antibody-crosslinked matrix 22 at 4°C as published 2223, resulting in successful removal of more than 99% of xPol ε holoenzyme including the catalytic subunit p260 as prviously 22. Various amounts of the n-xPol ε obtained from Xenopus egg extracts or r-xPol ε holoenzyme purified from insect cells and the buffer were added back to Pol ε- or mock-depleted egg extracts as published before 22. Then DNA replication reaction was initiated by the addition of membrane-removed sperm nuclei (2,000 sperm heads per μl of extract. After incubation at 25°C for various times, aliquots were withdrawn and analyzed by neutral agarose gel electrophoresis followed by autoradiography as published before 2223.

6 × His-tagged Cut5p, 6 × Flag-tagged Cut5p, 6 × Flag-tagged Cdc45p, 6 × Flag-tagged Psf2p containing GINS were expressed and purified from insect cells as described previously 242731.