Localization of the deoxyribose phosphate lyase active site in human DNA polymerase iota by controlled proteolysis.

Human DNA polymerase iota (pol iota) is a member of the Y-family of low fidelity lesion bypass DNA polymerases. In addition to a probable role in DNA lesion bypass, this enzyme has recently been shown to be required for somatic hypermutation in human B-cells. We found earlier that human pol iota has deoxyribose phosphate (dRP) lyase activity and unusual specificity for activity during DNA synthesis, suggesting involvement in specialized forms of base excision repair (BER). Here, mapping of the domain structure of human pol iota by controlled proteolysis revealed that the enzyme has a 48-kDa NH2-terminal domain and a protease resistant 40-kDa "core domain" spanning residues Met79 to approximately Met445. A covalently cross-linked pol iota-DNA complex, representing a trapped intermediate in the dRP lyase reaction, was subjected to controlled proteolysis. Cross-linking was mapped to the 40-kDa core domain, indicating that the dRP lyase active site is in this region. To further evaluate the BER capacity of the enzyme, the dRP lyase and DNA polymerase activities were characterized on DNA substrates representing BER intermediates, and we found that pol iota was able to complement the in vitro single-nucleotide BER deficiency of a DNA polymerase beta null cell extract.

Human DNA polymerase (pol ) is a member of the Y-family of low fidelity lesion bypass DNA polymerases. In addition to a probable role in DNA lesion bypass, this enzyme has recently been shown to be required for somatic hypermutation in human B-cells. We found earlier that human pol has deoxyribose phosphate (dRP) lyase activity and unusual specificity for activity during DNA synthesis, suggesting involvement in specialized forms of base excision repair (BER). Here, mapping of the domain structure of human pol by controlled proteolysis revealed that the enzyme has a 48-kDa NH 2 -terminal domain and a protease resistant 40-kDa "core domain" spanning residues Met 79 to ϳMet 445 . A covalently crosslinked pol -DNA complex, representing a trapped intermediate in the dRP lyase reaction, was subjected to controlled proteolysis. Cross-linking was mapped to the 40-kDa core domain, indicating that the dRP lyase active site is in this region. To further evaluate the BER capacity of the enzyme, the dRP lyase and DNA polymerase activities were characterized on DNA substrates representing BER intermediates, and we found that pol was able to complement the in vitro single-nucleotide BER deficiency of a DNA polymerase ␤ null cell extract.
Human DNA polymerase (pol ) 1 is a member of the Yfamily of DNA polymerases (UmuC/DinB/Rev1/Rad30) (1)(2)(3)(4)(5). These polymerases are thought to facilitate DNA replication beyond certain types of damaged DNA blocking lesions by conducting translesion DNA synthesis (6 -12). The enzymes in the Y-family share little or no primary structural homology with DNA polymerases of the other families (13)(14)(15). Among the Y-family DNA polymerases, DNA polymerase (pol ), which is encoded by the POLH gene (also referred to as XPV/ RAD30A), is present in both Saccharomyces cerevisiae and humans, whereas two other Y-family polymerases, pol and DNA polymerase , are absent in S. cerevisiae (12). A physiological function of pol in protecting humans against UV light exposure was established by the finding that individuals with Xeroderma pigmentusum complementation group V carry mu-tations in the POLH gene, resulting in weak or inactive forms of pol (16,17).
POLI is phylogenetically related to the POLH gene and encodes the human enzyme pol (18). In vitro properties of DNA synthesis by pol are consistent with a role in low-fidelity translesion replication (18 -20). The polymerase is distributive and is error-prone even on normal DNA templates. Yet, pol has unique features that appear to differentiate it from other Y-family polymerases. First, it misincorporates G opposite template T with greater efficiency than that for insertion of the correct base, A. Second, it has a tendency to abort DNA synthesis at positions where the template residue is a T consistent with the frequent misinsertion of G at this templating base (18 -22). These features of pol suggest that it may have specialized DNA synthesis functions in cells. Another role of pol appears to be in somatic hypermutation of human immunoglobulin genes, since targeted gene disruption of pol in a Burkitt's lymphoma cell line strongly reduced somatic hypermutation (23). Finally, pol has been shown to have deoxyribose phosphate (dRP) lyase activity, a feature that is consistent with a role in base excision repair (BER) (24). This feature could, in fact, be related to the pol requirement for somatic hypermutation, since error-prone DNA synthesis during BER may be an integral event in the process (23,24). The dRP lyase activity of pol proceeds by a ␤-elimination mechanism involving formation of a Schiff base intermediate between an enzyme primary amine and the ClЈ carbon of the deoxyribose phosphate to be released (24). The location of the dRP lyase active site and the primary amine that serves as the nucleophile in Schiff base formation are unknown.
In view of the interesting biochemical features and biological importance of pol , we characterized its domain structure by controlled proteolysis. This approach has been applied with many DNA enzymes for preparation of domain peptides that can be used to gain mechanistic and structural information. Recombinant human pol was expressed as a fusion peptide with glutathione S-transferase (GST), purified to homogeneity, and then digested with a panel of proteases. The GST region was released rapidly, and the COOH-terminal region of the enzyme was found to be hypersensitive, as no protease resistant peptides were observed. However, a relatively protease resistant 48-kDa NH 2 -terminal domain of pol was identified. This domain could be further digested to a protease-resistant 40-kDa domain. We made use of this information to map the location of the dRP lyase active site on the domain structure of the enzyme. The implications of these results are discussed. ville, OR). [␣-32 P]ddATP (3000 Ci/mmol), RESOURCE Q column, glutathione-Sepharose, glutathione HCl, and dNTPs were from Amersham Biosciences; [␥-32 P]ATP (7000 Ci/mmol) was from ICN Biomedicals; T4 polynucleotide kinase and terminal deoxynucleotidyltransferase were from Promega. Human apurinic/apyrimidinic (AP) endonuclease and uracil-DNA glycosylase (UDG) with 84 amino acids deleted from the amino terminus were purified as described (25,26). Chymotrypsin was from Worthington Biochemical Corp. Cell extracts from DNA polymerase ␤ (pol ␤) null mouse fibroblasts were prepared as described previously (27).
Overproduction and Purification of GST-tagged Pol -SF9 insect cells were infected with a high titer stock of pJM299-derived baculovious. After 60-h incubation, the cells were collected by centrifugation. Pol was then purified as described previously (18), except that in the final step of purification a RESOURCE Q column (fast protein liquid chromatography) was used instead of hydroxylapatite ion-exchange chromatography.
3Ј-End Labeling-A 34-mer oligodeoxyribonucleotide containing uracil at position 16 was labeled on its 3Ј-end by terminal deoxynucleotyltransferase using [␣-32 P]ddATP. This strand was then annealed with a complementary 34-mer template strand with a G at a position that pairs with U, by heating the mixture at 90°C for 3 min, followed by slow cooling to 25°C. 32 P-Labeled duplex oligonucleotide was separated from unincorporated [␣-32 P]ddATP using a MicroSpin G-25 column (Amersham Biosciences). The radiolabeled oligonucleotide duplex was stored at Ϫ30°C.
5Ј-End Labeling-A uracil-containing 19-mer oligodeoxyribonucleotide was 5Ј-phosphorylated with T4 polynucleotide kinase and [␥-32 P]ATP. The 34-mer template strand (described above) was then annealed with complementary 15-mer and 19-mer 32 P-labeled oligonucleotides. The duplex was purified by a MicroSpin G-25 column and stored as described above.
dRP Lyase Assay for Column Fractions-dRP lyase activity of pol was measured using a 34-base pair substrate DNA as described previously (28). Briefly, the 32 P-labeled uracil-containing duplex DNA was pretreated with 10 nM UDG and AP endonuclease. The reaction mixture (10 l) contained 50 mM Hepes, pH 7.5, 20 mM KCl, 5 mM MgCl 2 , 0.5 mM EDTA, 2 mM dithiothreitol (DTT), and 50 nM 32 P-labeled AP sitecontaining DNA. The reaction was initiated by adding 1 l of the indicated fraction from a RESOURCE Q column ( Fig. 1) and incubated at 37°C for 30 min. The reaction was terminated by transfer to 0 -1°C, and the DNA product was stabilized by addition of 2 M sodium borohydride (NaBH 4 ) to a final concentration of 340 mM. Incubation was continued for 30 min on ice. The stabilized DNA products were recovered by ethanol precipitation in the presence of 0.1 g/ml tRNA and resuspended in 10 l of gel-loading dye buffer (95% formamide, 20 mM EDTA, 0.02% bromphenol blue, and 0.02% xylene cyanol). After incubation at 75°C for 2 min, the reaction products were separated by electrophoresis in a 15% polyacrylamide gel containing 8 M urea in 89 mM Tris-HCl, pH 8.8, 89 mM boric acid, and 2 mM EDTA. Gels were scanned on a PhosphorImager 450 (Amersham Biosciences), and the data were analyzed using ImageQuant software.
Kinetic Measurement of dRP Lyase Activity-Kinetic analysis of dRP lyase activity of pol was performed using a 34-base pair substrate DNA that contained a 32 P-labeled uracil at position 16 and a nick between positions 15 and 16. The duplex DNA (100 nM) was pretreated (20 min at 30°C) with 20 nM UDG in a buffer containing 50 mM Hepes, pH 7.5, 20 mM KCl, 0.5 mM EDTA, and 2 mM DTT. The dRP lyase reaction mixture (50 l) was assembled at 0 -1°C in the above buffer. Reactions were initiated by adding appropriate dilutions of pol and incubated at 37°C. Aliquots were withdrawn at different time intervals and transferred to 0 -1°C to stop the reaction. The DNA product was stabilized by addition of 20 mM NaBH4 and incubated 30 min on ice. Then, an equal volume of gel-loading buffer was added, and the reaction mixture was incubated at 75°C for 2 min. The reaction products were separated by electrophoresis in a 15% polyacrylamide TBE-Urea gel (Invitrogen, precast gel) for 30 min at constant voltage (200 V). To quantify the reaction products, gels were scanned on a PhosphorImager and the data were analyzed using ImageQuant software. Due to dRP instability and the slow turnover number for the pol -dependent dRP lyase reaction, high enzyme concentrations were utilized. Under these conditions, the first-order rate constants of the exponential time courses are dependent on enzyme concentration. A secondary plot of the concentration dependence of k obs was fitted to the Michaelis equation.
Covalent Cross-linking of DNA to Pol -To prepare the covalently cross-linked pol -DNA complex, the NaBH 4 trapping technique was used (29). Briefly, the reaction mixture (10 l) contained 50 mM Hepes, pH 7.5, 20 mM KCl, 0.5 mM EDTA, 2 mM DTT, 200 nM 32 P-labeled UDG-treated duplex DNA, and 1 M pol . NaBH 4 was added immediately to the reaction mixture (1 mM final concentration) and incubated for 30 min on ice. SDS-PAGE sample buffer (10 l) was then added. The sample was boiled for 2 min, and the covalent complexes were resolved by NuPAGE, 4 -12% BisTris gel (Invitrogen), and the gel was scanned on a PhosphorImager.
Proteolysis of the Pol -DNA Complex with Chymotrypsin-After NaBH 4 reduction to achieve cross-linking of pol to 32 P-labeled DNA, the reaction mixture was supplemented with 10 mM CaCl 2 , and the complex was treated with micrococcal nuclease (0.5 g/ml). After a 20-min incubation at 30°C, the pH of the reaction mixture was adjusted to 8.0 by adding 1 M Tris-HCl, pH 8.0, to a final concentration of 100 mM. Cross-linked complex was then treated with chymotrypsin by mixing chymotrypsin to complex (w/w; 1:6) and incubating at 30°C. Aliquots were withdrawn at 1-, 2-, 5-, 10-, and 20-min intervals, mixed with B and C, the dRP lyase and DNA synthesis activities of the indicated fractions from the RESOURCE Q column were examined, as described under "Experimental Procedures." To examine dRP lyase activity of pol , the uracil-containing DNA strand was 3Ј-endlabeled with [ 32 P]ddAMP and annealed to its complementary DNA strand. To prepare the DNA substrate for the dRP lyase reaction mixture, 32 P-labeled DNA was pretreated with UDG and AP endonuclease. Thus, the resulting DNA substrate with a dRP group in a singlenucleotide gap, and with the dRP-containing strand 32 P-labeled at the 3Ј-terminus, was reacted with an equal volume (1 l) of the indicated fractions eluted from the RESOURCE Q column. The release of 5Ј-dRP from the 32 P-labeled substrate strand was detected by gel electrophoresis and quantified using a PhosphorImager and ImageQuant software. DNA synthesis activity of these fractions was assayed using a template/ primer in the presence of all four dNTPs. The reaction products were analyzed as above. The relative values of dRP lyase and DNA synthesis activities were plotted (C). SDS-PAGE sample buffer, and boiled for 3 min. The digested products were analyzed by NuPAGE 4 -12% BisTris gel. When pol alone was digested with chymotrypsin, a ratio of 1:50 (chymotrypsin to pol ) was used.
DNA Polymerase Assay-Radiolabeled template-primer DNA was prepared by annealing the 5Ј-32 P-labeled 21-mer primer to the 40-mer oligodeoxyribonucleotide template at a molar ratio of 1:1. The standard DNA polymerase reaction mixture (10 l) contained 40 mM Tris-HCl, pH 8.0, 5 mM MgCl 2 , 5 mM DTT, 100 g/ml bovine serum albumin, 100 M dNTPs, 2.5% glycerol, and 200 nM 32 P-labeled DNA. The reactions were initiated by adding the appropriate dilution of pol . After incubation at 37°C for 10 min, reactions were terminated by the addition of gel-loading buffer and the reaction products were analyzed as above.
Steady-state and Single Turnover Measurements of Pol Utilizing Gapped DNA-Steady-state (enzyme Ͻ Ͻ substrate DNA) and single turnover (enzyme Ͼ Ͼ substrate DNA) measurements were performed with a templating G residue in a single-nucleotide gapped DNA substrate described previously (30). The sequence of the single-nucleotide gapped DNA substrate was: primer, 5Ј-CTGCAGCTGATGCGC-3Ј; downstream oligonucleotide, 5Ј-GTACGGATCCCCGGGTAC-3Ј; and template, 3Ј-GACGTCGACTACGCGGCATGCCTAGGGGCCCATG-5Ј, where the underlined guanine serves as the templating base. Briefly, GST-tagged pol and DNA substrate were preincubated for 5 min at 37°C, and the reaction was initiated with Mg 2ϩ ⅐dCTP. The final concentrations in the reaction mixture were 50 mM Tris-HCl, pH 7.4, 100 mM KCl, 5 mM MgCl 2 , 5 mM DTT, 100 g/ml bovine serum albumin, 10% glycerol, and 100 M dCTP. Aliquots were withdrawn at various time intervals and quenched by the addition of 0.5 M EDTA. For the steadystate time course, the concentrations of GST-pol and DNA substrate were 5 and 200 nM, respectively, in contrast to the single turnover time course where the ratio of GST-pol /DNA was 5 (i.e. 250 nM/50 nM). Products were separated by 12% denaturing polyacrylamide gel electrophoresis and the reaction products were analyzed as above. Data were fitted to the appropriate equations by nonlinear least squares methods.

RESULTS AND DISCUSSION
Enzymatic Activities-To further examine the relationship between the polymerase and lyase activities of the recombinant pol protein, we purified the protein to near-homogeneity and then subjected the preparation to RESOURCE Q column chromatography (Fig. 1A). Fractions from the RESOURCE Q column containing most of the protein (fractions 9 -13) were pooled and represented the final purified enzyme. There is a single peak for both the lyase and polymerase activities observed in these fractions (Fig. 1, B and C). Since most of the 108 kDa pol protein elutes in these fractions, the results are consistent with the interpretation that these enzymatic activities are intrinsic to pol .
Domain Mapping of Pol -To examine the domain organization of pol , we subjected the purified enzyme (i.e. GST-Pol fusion protein) to controlled proteolysis with endoproteinase Glu-C, endoproteinase Lys-C, chymotrypsin, or trypsin. Results after chymotrypsin digestion are described in the legend to Fig. 2A. After 5 or 10 min, this digestion produced four major polypeptides (designated I-IV; Fig. 2A) with molecular masses of about 8-, 28-, 40-, and 48-kDa. Amino-terminal sequence analysis of these four polypeptides was undertaken (Fig. 2B). The results revealed that the 8-and 48-kDa polypeptides corresponded to the NH 2 terminus of the pol sequence, whereas the 40-kDa polypeptide started at residue Met 79 of pol . Se-quencing of the 28-kDa proteolytic fragment revealed it to be the GST tag. With longer incubation, the 48-kDa polypeptide was no longer observed, whereas the 40-kDa pol polypeptide and the 28-kDa GST tag polypeptide persisted. These results indicated that the 40-kDa polypeptide is a protease-resistant domain of pol derived from the NH 2 -terminal and central regions of the 80-kDa enzyme. The 40-kDa region is designated as the core domain (Fig. 2C). This 40-kDa core domain was also observed after digestion with other proteases, e.g. trypsin and endoproteinase GluC (not shown). Finally, we note that no stable proteolytic fragments were obtained from the COOHterminal region, ϳ35% of the pol sequence, indicating that Then, an equal volume of gel-loading buffer was added, and the reaction products were separated by polyacrylamide gel electrophoresis. A, schematic representation of the DNA substrate (18-merϩ 32 P-dRP) that was generated by UDG pretreatment from the internally 32 P-labeled 34-bp oligonucleotide duplex. B, photograph of the polyacrylamide gel, illustrating product analysis for a reaction mixture (100 nM pol is shown as an example). C, gels were scanned on a PhosphorImager, and the data were analyzed using ImageQuant software. The apparent activities were fitted to a hyberbolic equation by nonlinear least squares methods. this region is hypersensitive to proteolytic digestion. The COOH-terminal region has been shown to interact with pol and is necessary for "targeting" to stalled replication forks following DNA damaging treatment (31).
dRP Lyase Active Site-In one phase of the ␤-elimination mechanism proposed for the dRP lyase reaction, an active site primary amine nucleophile attacks the electrophilic ClЈ atom of the substrate sugar yielding a Schiff base. Routinely, this intermediate can be covalently trapped by reduction with NaBH 4 . The covalent cross-linking approach has been widely exploited for the identification of the domain or amino acid residue bearing the primary amine nucleophile in many DNA repair lyases (32). Similarly, pol can be covalently cross-linked to a dRP-containing DNA substrate by NaBH 4 reduction (24). To localize the dRP lyase active site in pol , we subjected the cross-linked enzyme-DNA complex first to micrococcal nuclease digestion and then to chymotrypsin digestion (Fig. 3A). Micrococcal nuclease digestion was used to remove extraneous substrate DNA, and hence this preserves the gel migration properties of the labeled enzyme (32). With shorter periods of proteolytic digestion, label was observed mainly in the 48-kDa polypeptide (peptide I (I), Fig. 3A). But, this species was strongly diminished with prolonged digestion, whereupon most of the labeled material observed corresponded to the 40-kDa core domain (peptide II (II), Fig. 3A). Controls for this experiment confirmed that the DNA substrate was nuclease sensitive, as expected (Fig. 3A, lanes 1-3), and that migration of the cross-linked enzyme was slightly faster after removal of the DNA substrate by nuclease digestion. These results indicate that the dRP lyase active site is in the 40-kDa core domain.
In view of the results described to this point, we examined the question of whether any of the chymotryptic fragments retained enough enzymatic activity to be cross-linked in situ. After chymotryptic digestion (see Fig. 2), the peptide fragments were added to a dRP lyase reaction mixture and subjected to cross-linking with radiolabeled substrate DNA. Interestingly, the 48-kDa domain was labeled, but the 40-kDa core domain and other peptides were not (Fig. 3B). These results suggested that the 48-kDa domain was capable of dRP lyase activity and that its 8-kDa NH 2 -terminal segment, which was removed by prolonged proteolytic digestion, was required for this domain peptide to be active.
Base Excision Repair Activities of Pol -We examined several features of purified human pol related to BER. First, kinetic features of single-nucleotide gap-filling DNA synthesis and dRP lyase were measured and compared with those of human pol ␤. To study kinetic parameters of dRP removal by pol , a duplex DNA (34-bp) that contained uracil at position 16 and a nick between positions 15 and 16 was prepared by annealing a 15-mer and 5Ј-end 32 P-lableled 19-mer to the 34-mer complementary DNA strand; the resulting duplex DNA was treated with UDG to create a dRP-containing DNA substrate. Thus, this DNA substrate contained a 32 P-lableled dRP flap in a single-nucleotide gap ( Fig. 4A; 18-merϩ 32 P-dRP). The time course of dRP removal from this substrate by pol was assayed as a function of enzyme concentration. The apparent K m and k cat for the dRP lyase of pol were 0.4 M and 0.002 s Ϫ1 , respectively (Fig. 4), resulting in a catalytic efficiency (k cat /K m ) of 5 nM Ϫ1 s Ϫ1 . Utilizing an optimized protocol to measure dRP lyase activity, this is ϳ30-fold lower than that found earlier for pol ␤ (28).
To determine the steady-state rate of single-nucleotide gapfilling by pol , a time course was performed using a saturating concentration of DNA and dCTP (Fig. 5A). The observed rate (i.e. k cat ) of insertion was 0.02 s Ϫ1 . A time course performed under single turnover conditions (i.e. enzyme/DNA ϭ 5) at saturating dCTP concentration yielded a rate of nucleotide insertion (k pol ) of 0.05 s Ϫ1 (Fig. 5B). The similarity between FIG. 5. Kinetic measurements of DNA synthesis and BER activity of pol . Steady-state rate (A) and single turnover rate (B) of pol were performed as described under "Experimental Procedures." Pol and DNA substrate were preincubated for 5 min at 37°C, and the reaction was initiated with Mg⅐dCTP. Aliquots were withdrawn at the indicated time intervals and quenched by EDTA. The reaction products were separated by 12% denaturing polyacrylamide gel electrophoresis and analyzed by using ImageQuant software. C, the reaction conditions and product analysis of BER assay were as described under "Experimental Procedures." The positions of ligated (35-mer) and unligated single-nucleotide (15-mer) products are indicated. A photograph of a PhosphorImager scan, illustrating the incorporation of [ 32 P]dCMP into DNA, is shown. A 35-bp duplex DNA containing a uracil residue at position 16 was incubated without (lane 1) or with (lane 2) purified pol , [␣-32 P]dCTP, ␤-pol null extract for 30 min at 37°C. The DNA products were separated by 15% polyacrylamide gel and scanned using a PhosphorImager.