INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

DNA synthesis is catalyzed by a DNA polymerase (Pol). In humans, eight DNA polymerases have been identified thus far: Pol, -, -, -, -, -, -, and - (6, 10, 15, 17, 19, 32). All DNA polymerases studied so far from any biological source synthesize DNA by the Watson-Crick base-pairing rule, incorporating A, G, C, and T opposite the templates T, C, G, and A, respectively (37). Nuclear DNA replication involves Pol, -, and -, while mitochondrial DNA replication requires Pol (15). Pol is a major repair synthesis enzyme during base excision repair (16). Pol is believed to be involved in the damage-induced mutagenesis pathway for translesion DNA synthesis (6, 17, 27). Pol, encoded by the XPV gene, is a DNA lesion bypass polymerase capable of both error-free and error-prone translesion synthesis using several damaged DNA templates (11, 19, 39). Pol is predicted to be involved in repair of DNA interstrand cross-links (32).

Recently, four families of proteins have been identified as forming the UmuC superfamily (4, 23). The prototypic members of this superfamily are the Escherichia coli UmuC, E. coli DinB, yeast Saccharomyces cerevisiae Rev1, and S. cerevisiae Rad30. UmuC is a subunit of the E. coli DNA polymerase V, a polymerase required in the damage-induced mutagenesis pathway (30, 33). DinB is the E. coli DNA polymerase IV (34), which is involved in untargeted mutagenesis (1, 13, 34). Rev1 is a dCMP transferase, that is required in the damage-induced mutagenesis (26). Rad30 is the yeast DNA Pol (11). Human homologues of DinB (5, 28), Rev1 (7, 18), and Rad30 (10, 20, 23) have been isolated. In humans, two Rad30 homologues, XPV (10, 20) and RAD30B (23), have been reported based on protein sequence comparisons. Studies on the enzyme activity of XPV protein clearly indicate that it is the functional counterpart of the yeast Rad30 and thus has been named human Pol (19, 20).

The function of human Pol as a lesion bypass DNA polymerase has been unequivocally demonstrated by the clinical phenotypes of XPV patients (2), cellular characteristics of XPV cells (2, 24), and biochemical activities of XPV protein (19). The function of the human RAD30B protein, however, is not known. A very useful approach to understand RAD30B in human physiology is to elucidate its biochemical activities. In this report, we (i) show that the human RAD30B gene codes for the ninth DNA polymerase, Pol; (ii) demonstrate that human DNA Pol violates the Watson-Crick base-pairing rule opposite template T by preferentially incorporating a G rather than an A; and (iii) show efficient nucleotide incorporation by human Pol opposite a template abasic site.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. A mouse monoclonal antibody against the His₆ tag was purchased from Qiagen. Alkaline phosphatase-conjugated anti-mouse immunoglobulin G was from Sigma Chemical Co. Pfu DNA polymerase was from Stratagene. The yeast rad30 deletion mutant strain BY4741rad30 (MATa his3 leu2 met15 ura3 rad30) was purchased from Research Genetics. The 30-mer template containing a site-specific tetrahydrofuran (apurinic/apyrimidinic [AP] site analogue) was synthesized from a DNA synthesizer by Operon. Its sequences is 5'-GCGCGCTTCTGGCCAATXCTAGACGGTAGG-3', where X is the AP site analogue. Human Pol was expressed in the yeast rad30 deletion mutant cells and purified to apparent homogeneity as described elsewhere (Y. Zhang, F. Yuan, X. Wu, J.-S. Taylor, and Z. Wang, submitted for publication).

Overexpression plasmid of the human RAD30B gene. The human RAD30B cDNA was obtained by PCR amplification from human testis cDNAs using Pfu DNA polymerase and two primers, 5'-CGGGATCCATGGAACTGGCGGACG-3' and 5'-CCCAAGCTTACGCTTTGTGCCAGAATTTACTTC-3'. The resulting 2.3-kb PCR product was then cloned into the BamHI and HindIII sites of the vector pECUh6, yielding pECUh6-hRAD30B. The human RAD30B gene was verified by DNA sequencing. This expression construct contained the 2µm origin for multicopy plasmid replication, the URA3 gene for plasmid selection, the CUP1 promoter for inducible RAD30B gene expression, and six His codons preceding the ATG initiator codon of the human RAD30B gene.

Purification of human RAD30B protein. Yeast rad30 deletion strain BY4741rad30 harboring pECUh6-hRAD30B was grown in minimum medium containing 2% dextrose for 2 days. After 10-fold dilution in 16 liters of YPD medium (2% Bacto Peptone, 1% yeast extract, 2% dextrose) and growth for 6 h at 30°C, CuSO₄ was added to 0.3 mM to induce human RAD30B expression for 3 h. Collected cells (~100 g) were homogenized by zirconium beads in a Bead-Beater in an extraction buffer containing 50 mM Tris-HCl (pH 7.5), 600 mM KCl, 5 mM -mercaptoethanol, 10% sucrose, and protease inhibitors (38). The clarified extract (~120 ml) was loaded onto a HiTrap chelating column charged with NiSO₄ (10 ml; Amersham Pharmacia Biotech), the column was then washed sequentially with 100 ml of Ni buffer A (50 mM Tris-HCl [pH 7.5], 0.5 M NaCl, 10% glycerol, 5 mM -mercaptoethanol, protease inhibitors) containing 10 mM imidazole and 100 ml of Ni buffer A containing 35 mM imidazole. Bound proteins were eluted with a linear gradient of 35 to 108 mM imidazole. The His-tagged human RAD30B was identified by Western blotting using a mouse monoclonal antibody specific to the His₆ tag. The pooled sample was concentrated by polyethylene glycol (PEG) 10000 and desalted through five 5-ml Sephadex G-25 columns in FPLC (fast protein liquid chromatography) buffer A (50 mM Tris-HCl [pH 7.5], 1 mM EDTA, 10% glycerol, 5 mM -mercaptoethanol) containing 30 mM KCl. The resulting sample (~40 ml) was loaded onto an FPLC Mono S HR5/5 column and eluted with a 30-ml linear gradient of 30 to 500 mM KCl in FPLC buffer A. Human RAD30B was eluted at ~150 mM KCl. The Mono S fractions were concentrated by PEG 10000 and loaded onto a FPLC Superdex 200 gel filtration column that had been equilibrated with FPLC buffer A containing 300 mM KCl. Human RAD30B was eluted at a position of ~100 kDa.

Purification of human DNA Pol. Yeast SX46A cells (MATa ade2 his3-532 trp1-289 ura3-52) harboring pEGUh6-hPOLB were grown in minimal medium containing 2% sucrose for 2 days. Expression of Pol was induced by diluting the culture 10-fold in 16 liters of YPG (2% Bacto Peptone, 1% yeast extract, 2% galactose) medium supplemented with 0.5% sucrose and incubation for 15 h at 30°C with shaking. Preparation of cell extracts and purification by nickel column chromatography were performed similarly as described above for human RAD30B. To purify human Pol further, the nickel column fractions were concentrated by PEG 10000 and desalted through five 5-ml Sephadex G-25 columns in FPLC buffer A (50 mM Tris-HCl [pH 7.5], 1 mM EDTA, 10% glycerol, 5 mM -mercaptoethanol) containing 100 mM KCl. Then, the sample was loaded onto an FPLC Resource S column (1 ml) and eluted with a 30-ml linear gradient of 100 to 400 mM KCl in FPLC buffer A. Pol was eluted at ~230 mM KCl. The Resource S fractions were pooled and concentrated by PEG 10000. The sample was loaded onto an FPLC Superdex 200 gel filtration column equilibrated with FPLC buffer A containing 300 mM KCl and then eluted in the same buffer. Pol was eluted at a position of ~40 kDa.

DNA polymerase assays. A standard DNA polymerase reaction mixture (10 µl) contained 25 mM KH₂PO₄ (pH 7.0), 5 mM MgCl₂, 5 mM dithiothreitol, 100 µg of bovine serum albumin/ml, 10% glycerol, 50 µM each dATP, dCTP, dTTP, and dGTP, 50 fmol of a primed DNA template, and purified Pol. The DNA primer was labeled at its 5' end with ³²P, unless otherwise indicated. After incubation at 30°C for 10 min, reactions were terminated with 7 µl of a stop solution (20 mM EDTA, 95% formamide, 0.05% bromophenol blue, 0.05% xylene cyanol). Reaction products were separated on a 20% polyacrylamide gel containing 8 M urea and visualized by autoradiography. Primer extension was quantitated by scanning densitometry of the autoradiogram using the SigmaGel software (Sigma) for analysis.

Kinetic analysis of human DNA Pol. Kinetic analysis of human DNA Pol was performed using a previously described method (3, 36). Briefly, DNA polymerase assays were performed using 50 fmol of a primed DNA template, 3.7 ng of purified human Pol, and increasing concentrations of each dNTP (dATP, dCTP, dTTP, or dTTP). The DNA primer, 5'-GGAAGAAGAAGTATGTT-3', was labeled at its 5' end with ³²P. The four DNA templates used were template A (5'-CCTTCTTCATTCAAACATACTTCTTCTTCC-3'), template G (5'-CCTTCTTCATTCGAACATACTTCTTCTTCC-3'), template C (5'-CCTTCTTCATTCCAACATACTTCTTCTTCC-3'), and template T (5'-CCTTCTTCATTCTAACATACTTCTTCTTCC-3') (the analyzed template base is underlined). After incubation for 10 min at 30°C under standard DNA polymerase assay conditions, reaction products were separated by electrophoresis on a 20% denaturing polyacrylamide gel. The percentage of primers extended by the polymerase was calculated following scanning densitometry of the extended DNA band(s) and the remaining primer band on the autoradiogram. Product formed (P) was derived from the calculation P = % primer extension × 50 fmol. Observed enzyme velocity (v) was obtained from the calculation v = P/10 min. Then, the observed enzyme velocity was plotted as a function of dNTP concentration. The plotted data was fitted by a nonlinear regression curve to the Michaelis-Menten equation, v = (V_max × [dNTP])/(K_m + [dNTP]), using the SigmaPlot software. V_max and K_m values for the incorporation of the correct and incorrect nucleotides were obtained from the fitted curves. Relative polymerase selectivity of incorrect versus correct nucleotides (f_inc) was finally calculated from the equation: f_inc = (V_max/K_m)_incorrect/(V_max/K_m)_correct (3).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Human RAD30B gene codes for the ninth DNA polymerase, Pol. To elucidate the biochemical activities of human RAD30B, we purified this protein to apparent homogeneity (Fig. 1A). To facilitate purification and detection, the human RAD30B protein was tagged by six histidine residues at its N terminus. The identity of the tagged human RAD30B protein was confirmed by Western blot analysis using a mouse monoclonal antibody specific to the His₆ tag (Fig. 1B). The purified human RAD30B migrated at 80 kDa on a sodium dodecyl sulfate (SDS)-10% polyacrylamide gel (Fig. 1A), consistent with its calculated molecular weight of 80,346. Using a 40-mer DNA template and a 17-mer primer containing a ³²P label at its 5' end, we found that human RAD30B possesses a DNA polymerase activity (Fig. 1C). Based on its sequence similarity to Pol, human RAD30B protein was suspected to be a DNA polymerase; the name DNA polymerase  (Pol) has been reserved for this protein in the human genome database. Our results show that human RAD30B protein is indeed a DNA polymerase, and thus it will be referred to as human Pol hereafter.

View larger version (42K): [in this window] [in a new window]   FIG. 1.   DNA polymerase activity of human RAD30B protein. (A) Purified human RAD30B protein (370 ng) was analyzed by electrophoresis on an SDS-10% polyacrylamide gel and visualized by silver staining. Protein size markers (lane M) are indicated on the left. (B) Purified human RAD30B protein (19 ng) was analyzed by Western blotting using a mouse monoclonal antibody against the His6 tag. Protein size markers (lane M) are indicated on the right. (C) DNA polymerase assays were performed without (lane 1) or with (lane 2) purified human RAD30B (19 ng), using the 40-mer template DNA, 5'-AAGGAAGGAAGGAAGGAACGAAGAACATACTTCTTCTTCC-3', annealed with the 5' 32P-labeled primer, 5'-GGAAGAAGAAGTATGTT-3'. Quantitation of extended primers is shown at the bottom of the gel. DNA size markers in nucleotides are indicated on the right.

The T stop property of human Pol. We consistently observed that human Pol extended several annealed primers by only a few nucleotides. It appeared that primer extension by human Pol often stopped at a template T. To further investigate this preliminary observation, we performed polymerase assays using three DNA templates that differ by one base 5 nt downstream from the end of the primer (Fig. 2). In the absence of a template T (Fig. 2, templates G22 and C22), longer DNA was synthesized by human Pol with increasing enzyme concentrations, nearly reaching the end of the template (Fig. 2, lanes 1 to 3 and 7 to 9). In contrast, DNA synthesis by human Pol was strongly inhibited by a template T (Fig. 2, template T22), as evidenced by aborted primer extension at 22 nt opposite the template T (Fig. 2, lanes 4 to 6). Even at a high Pol concentration, only a small fraction of the primer was extended beyond the template T (Fig. 2, lane 6). Further increase of human Pol concentration stimulated more DNA synthesis past the template T. Nevertheless, a strong DNA synthesis stop opposite the template T was evident (data not shown). These results demonstrate that DNA synthesis by human Pol is frequently aborted by a template T. We designate this phenomenon the T stop.

View larger version (52K): [in this window] [in a new window]   FIG. 2.   T stop during DNA synthesis by human Pol. Three DNA templates differ by one nucleotide at the underlined base were annealed to the 5' 32P-labeled 17-mer primer as indicated. Polymerase assays were then performed with increasing amounts of human Pol as indicated, using the primed G22, T22, and C22 templates. On template T22, DNA synthesis by human Pol was blocked at the template T (T stop), yielding a strong 22-nt band (lanes 5 and 6). Quantitation of the 22-nt bands is shown at the bottom of the gel. DNA size markers in nucleotides are indicated on the left.

Lack of a detectable 3'5' proofreading nuclease activity with human Pol. To examine whether human Pol contains a nuclease activity, we labeled several primers at their 5' ends with ³²P and annealed them to a 30-mer DNA template (Fig. 3A). These DNA substrates contained a T-C mismatch (Fig. 3A, primer P1), an A-C mismatch (Fig. 3A, primer P2), a C-T mismatch (Fig. 3A, primer P3), or a G-A mismatch (Fig. 3A, primer P4) at the primer end. Then, we incubated 50 fmol of these DNA substrates with 46 fmol of human Pol in polymerase reaction buffer without dNTPs. Even after prolonged incubation at 30°C for 30 min, rather than 10 min as in a standard DNA polymerase assay, there was no detectable exonucleolytic removal of the mismatched 3' nucleotides by human Pol (Fig. 3B). These results indicate that human Pol does not possess a detectable 3'5' proofreading nuclease activity under the conditions used.

View larger version (38K): [in this window] [in a new window]   FIG. 3.   Proofreading nuclease assays of purified human Pol. (A) The DNA template and four primers used for nuclease activity assays. The primers were labeled with 32P at their 5' ends as indicated by an asterisk. Each primer was annealed individually to the template. (B) DNA substrates (50 fmol) containing annealed primers as indicated were incubated without () or with (+) purified human Pol (46 fmol) for 30 min at 30°C in the DNA polymerase assay buffer without dNTPs. Reaction products were separated by electrophoresis on a 20% denaturing polyacrylamide gel. DNA size markers in nucleotides are indicated on the right.

Base-pairing specificity of human Pol. To examine the base-pairing property of Pol, we performed DNA polymerase assays in the presence of only one dNTP, using primed 30-mer DNA templates. A 17-mer primer was labeled with ³²P at its 5' end and annealed to templates C, G, T, and A (Fig. 4A). Opposite the template CC sequence, two G residues were incorporated by human Pol as the major event (Fig. 4B, lane 6). Less frequently, A, T, and C were also incorporated opposite the template C (Fig. 4B, lanes 3 to 5). Opposite the template G, a C residue was incorporated by human Pol (Fig. 4B, lane 10). Less frequently, a T residue was also incorporated (Fig. 4B, lane 11). Opposite the template A, a T residue was incorporated by human Pol (Fig. 4C, lane 11); G was also incorporated, although to a much lesser extent (Fig. 4C, lane 12). These results show that Pol preferentially incorporates the correct base opposite the template C, G, or A, according to the Watson-Crick base-pairing rule.

View larger version (83K): [in this window] [in a new window]   FIG. 4.   Base pairing specificity of human Pol. (A) DNA templates used for polymerase assays. Each DNA template (Temp) was annealed separately with the 17-mer primer that was labeled at its 5' end with 32P as indicated by an asterisk. (B and C) Polymerase assays were performed with 50 fmol of DNA and 3.7 ng of human Pol in the presence of a single dNTP or all four dNTPs (N4) as indicated. Lanes 1 and 7, controls without Pol. DNA size markers in nucleotides are indicated on the left. (D) The primed DNA template (from the human p53 gene sequence) as indicated on the right was incubated with purified human Pol (23 ng) or human Pol (1 ng). Primer extension assays were performed with either dATP (A), dCTP (C), dTTP (T), or dGTP (G) or all four dNTPs (N4) as indicated. Quantitation of extended primers is shown below the gels.

View larger version (46K): [in this window] [in a new window]   FIG. 5.   Nucleotide incorporation by human Pol opposite a template T. (A) DNA templates used for polymerase reactions. The annealed primers were not labeled with 32P at their 5' ends. (B and C) Using primed templates as indicated, DNA synthesis was performed with Pol (7.4 ng) in the presence of 5 µM dNTPs and 10 µCi of [-32P]dATP (A, lanes 1 and 5), [-32P]dCTP (C, lanes 2 and 6), [-32P]dTTP (T, lanes 3 and 7), or [-32P]dGTP (G, lanes 4 and 8) individually. Incorporation of 32P was quantitated relative to the lowest-density band detectable. In lane 1 of panel C, [-32P]dAMP incorporation was the sum of the 16-mer and 17-mer bands. DNA size markers in nucleotides are indicated on the left.

Kinetic measurements of base selection by human Pol. To obtain a quantitative measurement of base selections by human Pol, we measured its nucleotide incorporation fidelity using a steady-state kinetic analysis (3, 36). DNA polymerase assays were performed with increasing concentrations of a single dNTP using purified human Pol and 50 fmol of the primed template T (Fig. 4A). Opposite the template T, G incorporation was observed at low dGTP concentrations (Fig. 6A, lanes 7 to 12). However, for A incorporation, primer extension occurred at much higher dATP concentrations (Fig. 6A, lanes 1 to 6). These base incorporation results were quantitated, and the observed incorporation velocities (v) were calculated. Experimental v values were plotted against the dNTP substrate concentrations and fitted into the Michaelis-Menten equation, v = (V_max × [dNTP])/(K_m + [dNTP]), as shown in Fig. 6B. From the fitted curve, the kinetic parameters V_max and K_m were obtained. Using this method, we systematically measured the V_max and K_m values of all 16 possible base incorporations by human Pol opposite the four template bases in an identical sequence context (Fig. 4A). The results are shown in Table 1.

View larger version (21K): [in this window] [in a new window]   FIG. 6.   Kinetic analysis of A and G incorporations by human Pol opposite a template T. (A) Polymerase assays were performed at 30°C for 10 min with purified human Pol (3.7 ng) using 50 fmol of the template T (Fig. 4A) and increasing concentrations of dATP or dGTP as indicated. Primer extension products were separated from the 32P-labeled 17-mer primer by electrophoresis on a 20% denaturing polyacrylamide gel and visualized by autoradiography. DNA size markers in nucleotides are indicated on the sides. (B) Results in panel A were quantitated, and the rate of nucleotide incorporation (primer extension) was graphed as a function of dATP or dGTP concentration. The scattered plots were then fitted into the Michaelis-Menten equation as described in Materials and Methods. The Vmax and Km values obtained from the fitted curve are listed in Table 1.

View this table: [in this window] [in a new window]   TABLE 1.   Kinetic measurement of nucleotide incorporation by human Pol

2

2

View this table: [in this window] [in a new window]   TABLE 2.   Kinetic measurement of nucleotide incorporation opposite template T by human Pola

Mismatch extension by human Pol. To examine the ability of human Pol to extend mismatched bases, we performed primer extension assays using the 12 possible base pair mismatches. The T-G (template-primer) base pair was most effectively extended by Pol under standard DNA polymerase assay conditions containing 50 µM dNTPs and 46 fmol Pol (Fig. 7A). A-A and C-A (template-primer) mismatches were extended by Pol with lower frequencies (Fig. 7A). Compared among T-A, T-G, and A-T terminated template-primer substrates (50 fmol of each), 20, 6, and 24 fmol of primers, respectively, were extended by human Pol (Fig. 7B, lanes 1, 6, and 11). During the T-A or T-G extension by human Pol, the next base incorporated was a T as directed by the template A (Fig. 7B, lanes 4 and 9). These results show that T-G (template-primer) is a poor base pair for primer extension by human Pol compared to the Watson-Crick base pairs. Preferential G incorporation opposite the template T followed by inefficient extension of the resulting T-G base pair provided a molecular explanation for the T stop property of human Pol.

View larger version (33K): [in this window] [in a new window]   FIG. 7.   Mismatch extensions by human Pol. (A) Various primers labeled at their 5' ends with 32P were annealed to the indicated template, generating 12 possible mismatches at the primer 3' ends and at the underlined template positions. Mismatched substrates were incubated with (+) or without () human Pol (3.7 ng) under standard polymerase assay conditions. DNA size markers in nucleotides are indicated on the sides. (B) Three 5' 32P-labeled primers were annealed to the template as shown. Primer extensions by human Pol (1 ng) were then performed under polymerase assay conditions using either dATP (A), dCTP (C), dTTP (T), or dGTP (G) or all four dNTPs (N4) as indicated. DNA size markers in nucleotides are indicated on the right.

Response of human Pol to a template AP site. Since human Pol is a sequence homologue of the lesion bypass enzyme Pol (23), it is possible that human Pol may also play a role in DNA lesion bypass. To investigate this possibility, we examined the response of purified human Pol to an AP site in DNA. A 17-mer primer was labeled at its 5' end with ³²P and annealed to a DNA template right before a template T (template FY2) or an AP site (template AP) (Fig. 8). Primer extension was then performed with increasing amounts of purified human Pol. As shown in Fig. 8 (lanes 5 to 7), purified human Pol incorporated one nucleotide opposite the template AP site but was unable to extend DNA synthesis further. On the undamaged template, human Pol extended the primer by one nucleotide and stopped opposite the template T (Fig. 8, lanes 1 to 4), as a result of the T stop. Nucleotide incorporation opposite the template AP site was clearly more efficient than opposite the template T (Fig. 8). These results show that human Pol is able to incorporate one nucleotide opposite a template AP site very efficiently.

View larger version (41K): [in this window] [in a new window]   FIG. 8.   Efficient nucleotide insertion by human Pol opposite a template AP site. A 32P-labeled 17-mer primer was annealed to the undamaged template FY2 or annealed right before a template AP site as shown. The AP site is located at the X position. Polymerase assays were performed with increasing amounts of human Pol using 50 fmol of undamaged templates (FY2) (lanes 1 to 4) and the AP site-containing template (template AP) (lanes 5 to 7). Quantitation of extended primers is shown at the bottom of the gel. DNA size markers in nucleotides are indicated on the right.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have purified human RAD30B protein to apparent homogeneity and showed that it is a DNA polymerase, Pol. In a drastic contrast to any other DNA polymerases reported so far, human Pol violates the Watson-Crick base-pairing rule by incorporating G instead of A opposite the template T. In the yeast S. cerevisiae, there is a counterpart (RAD30) of the human XPV gene (10, 20, 22, 31). However, after searching the whole yeast genome, we did not find a RAD30B counterpart gene in this organism. Therefore, we chose to express the human RAD30B protein in yeast cells for its purification. Following expression of the human RAD30B protein, we detected the unique Pol activity of G incorporation opposite the template T in yeast cell extracts. In contrast, this activity was not detected in yeast cell extracts without the human RAD30B protein (data not shown). Hence, the unique polymerase activity is specific to the human RAD30B protein, and the activity of purified human Pol could not have derived from a contaminating yeast protein.

DNA polymerases are not 100% accurate when copying DNA templates. They are associated with certain levels of fidelity opposite template T. Thus, the high-fidelity E. coli polymerase III holoenzyme rarely incorporates G opposite the template T (29). The chance for the low-fidelity human Pol to incorporate G opposite the template T is much greater (Table 2) (12, 21). Regardless of the specific error rates of G misincorporation, all other DNA polymerases predominantly incorporate an A opposite the template T, thus following the Watson-Crick base-pairing rule. The difference among all other DNA polymerases with regard to G misincorporation opposite template T is thus one of degree. Human Pol is not in this category. Human Pol predominantly incorporates G opposite the template T, thus truly violating the Watson-Crick base-pairing rule. According to the kinetic analysis (Table 1), incorporation of A opposite the template T has become a misincorporation to human Pol, because it is kinetically unfavored compared to G incorporation. Indeed, for all other DNA polymerases, kinetically unfavored incorporations are defined as misincorporations (3).

The molecular basis of preferential G incorporation opposite template T by human Pol is not understood. Apparently, this unique base incorporation by human Pol is not simply determined by the hydrogen bonding property between T and G, since this polymerase predominantly incorporates C opposite the template G. Thus, the template T-primer G base pairing is likely determined specifically by the active site of human Pol. Furthermore, lack of preferential A incorporation opposite template T is not due to exonucleolytic action of human Pol. This conclusion is based on two observations. First, human Pol lacks a 3'5' exonuclease activity (Fig. 3). Second, on a DNA template containing a primer 3' A pairing with a template T, removal of this primer A by human Pol was not detected (Fig. 7B, lanes 2, 3, and 5). Instead, the primer was effectively extended by human Pol with a T addition according to the next template A (Fig. 7B, lanes 4).

After incorporating a G opposite the template T, the resulting T-G base pair is less efficiently extended by human Pol compared to Watson-Crick base pairs. Consequently, DNA synthesis frequently stops opposite a template T. We designate this novel property the T stop. T stop effectively restricts human Pol to a very short stretch of DNA synthesis. This unique feature additionally distinguishes human Pol from any other DNA polymerases studied so far.

Although human Pol predominantly incorporates the correct G opposite template C, it misincorporates T, A, and C with unprecedented high frequencies. According to the kinetic measurements, template bases were copied by human Pol with the fidelity order A > G > C > T (from most to least accurate). Recent studies on the DNA synthesis fidelity of human Pol have led to the conclusion that it is the least accurate DNA polymerase (21). Using purified human Pol, human Pol, and human Pol, we performed kinetic analyses with identical DNA templates (Tables 1 and 2 and our unpublished results). Comparison of the kinetic measurements clearly indicate that human Pol is the most inaccurate among the three DNA polymerases. Hence, human Pol may be the most inaccurate enzyme among all DNA polymerases studied.

Given the fact that Pol is a member of the UmuC superfamily and closely related to the lesion bypass enzyme Pol at the sequence level (23), it is possible that this polymerase may be involved in DNA lesion bypass. In this study, we found that nucleotide incorporation by human Pol opposite the template AP site is remarkably efficient, even more efficient than nucleotide incorporation opposite the template T. Further DNA synthesis, however, is blocked by the template AP site. This response of human Pol to template AP sites is similar to that of yeast Pol (39). It is conceivable that following human Pol action the blocked primer end may be extended by another DNA polymerase such as Pol in cells. Indeed, the yeast Pol is able to extend the blocked primer end following one nucleotide incorporation opposite the template AP site by yeast Pol in vitro (39). The human counterpart of Pol, consisting of the REV3 and REV7 gene products, has been identified (6, 17, 25). In a separate study, we found that human Pol predominantly incorporates a G opposite the template AP site (Zhang et al., submitted). Consistent with the notion that Pol plays a role in AP site bypass in humans, major G incorporation opposite AP sites was observed in a study using a shuttle vector that had been replicated in human lymphoblastoid cells (14). Additionally, our recent studies indicate that human Pol is able to incorporate a C opposite a bulky AAF-adducted guanine in DNA and an A opposite the 3' T of a template TT (6-4) photoproduct (Zhang et al., submitted). These observations support a role for human Pol in DNA lesion bypass. In contrast to Pol, which efficiently performs error-free lesion bypass opposite a template TT dimer (11, 19, 20, 35), human Pol is essentially unresponsive to this UV lesion (Zhang et al., submitted).

Preferential G incorporation opposite template T and the T stop feature set human Pol apart from Pol and other known DNA polymerases. These unique properties suggest a specialized function for Pol in human biology. Another possible function of human Pol might be in the somatic hypermutation during immunoglobulin development. The hallmarks of somatic hypermutation include high mutation rate involving both purines and pyrimidines, highly restricted occurrence at the immunoglobulin genes, and requirement for transcription (8, 9). Human Pol is highly error prone. DNA synthesis by human Pol is restricted to short stretches of template sequences as a result of the T stop. Therefore, the biochemical requirements for a hypermutation DNA polymerase could be satisfied by Pol. If enhancer-dependent transcription at the immunoglobulin gene is coupled to a factor that can occasionally incise DNA, mediate limited DNA degradation, and subsequently recruit Pol, somatic hypermutation would be expected.