Functional Analysis of the Amino-terminal 8-kDa Domain of DNA Polymerase β as Revealed by Site-directed Mutagenesis

The amino-terminal 8-kDa domain of DNA polymerase β functions in binding single-stranded DNA (ssDNA), recognition of a 5′-phosphate in gapped DNA structures, and as a 5′-deoxyribose phosphate (dRP) lyase. NMR and x-ray crystal structures of this domain have suggested several residues that may interact with ssDNA or play a role in the dRP lyase reaction. Nine of these residues were altered by site-directed mutagenesis. Each mutant was expressed inEscherichia coli, and the recombinant protein was purified to near homogeneity. CD spectra of these mutant proteins indicated that the alteration did not adversely affect the global protein structure. Single-stranded DNA binding was probed by photochemical cross-linking to oligo(dT)16. Several mutants (F25W, K35A, K60A, and K68A) were impaired in ssDNA binding activity, whereas other mutants (H34G, E71Q, K72A, E75A, and K84A) retained near wild-type binding activity. The 5′-phosphate recognition activity of these mutants was examined by UV cross-linking to a 5-nucleotide gap DNA where the 5′ terminus in the gap was either phosphorylated or unphosphorylated. The results indicate that Lys35 is involved in 5′-phosphate recognition of DNA polymerase β. Finally, the dRP lyase activity of these mutants was evaluated using a preincised apurinic/apyrimidinic DNA. Alanine mutants of Lys35 and Lys60 are significantly reduced in dRP lyase activity, consistent with the lower ssDNA binding activity. More importantly, alanine substitution for Lys72 resulted in a greater than 90% loss of dRP lyase activity, without affecting DNA binding. Alanine mutants of Lys68 and Lys84 had wild-type dRP lyase activity. The triple alanine mutant, K35A/K68A/K72A, was devoid of dRP lyase activity, suggesting that the effects of the alanine substitution at Lys72 and Lys35 were additive. The results suggest that Lys72 is directly involved in formation of a covalent imino intermediate and are consistent with Lys72as the predominant Schiff base nucleophile in the dRP lyase β-elimination catalytic reaction.

Genomic DNA is constantly exposed to various endogenous and external environmental agents leading to DNA base loss and/or damage. To remove such damage and retain genome stability, the base excision DNA repair pathway has been maintained in essentially all organisms. Base excision repair was initially described in Escherichia coli (1) and later in mammalian cells (2). This repair pathway is initiated by enzymatic removal of an inappropriate base or spontaneous hydrolysis of bases through cleavage of the N-glycosyl bond (3,4). The resulting apurinic/apyrimidinic (AP) 1 site is cleaved by a class II AP endonuclease (5), which incises the phosphodiester backbone 5Ј to the AP site resulting in a 3Ј-hydroxyl and 5Ј 2-deoxyribose 5-phosphate (dRP) containing termini. To complete repair, the dRP moiety is removed so that a single-nucleotide gap with a 3Ј-hydroxyl and 5Ј-phosphate is generated (6,7). DNA polymerase ␤ (␤-pol) replaces the missing nucleotide (7)(8)(9), and DNA ligase I seals the nicked product (9,10). These enzymatic activities should be coordinated for efficient base excision repair.
␤-pol is a multifunctional enzyme consisting of an 8-kDa amino-terminal domain with dRP lyase activity (11,12) and a 31-kDa carboxyl-terminal domain with nucleotidyltransferase activity (13). The crystal and solution structures of the aminoterminal 8-kDa domain have been determined (14,15). The 8-kDa domain (residues 1-87) is formed by four ␣-helices, packed as two antiparallel pairs. The pairs of ␣-helices cross one another at 50°giving them a V-like shape (15,16). The 8-kDa domain of ␤-pol also contains a motif termed "Helixhairpin-Helix" (HhH) (17). This motif has been found in a number of DNA repair proteins, including several DNA glycosylases and AP lyases (18). Residues of the HhH motif have been proposed to contribute to recognition and excision of damaged nucleotides in DNA, as well as AP lyase chemistry (18,19). Alignment of the HhH motifs from ␤-pol and endonuclease III suggest that ␤-pol Lys 68 may be critically important in lyase chemistry since mutation of the analogous lysine residue in endonuclease III, Lys 120 , resulted in a dramatic reduction in k cat (20). Furthermore, ␤-pol protein residues involved in single-stranded DNA (ssDNA) binding have been identified by NMR using chemical shift changes (15). The Helix-3-hairpin-Helix-4 motif and residues in an adjacent ⍀-type loop form the ssDNA interaction surface (15,16). The x-ray crystal structure of ␤-pol bound to a template-primer substrate suggested that * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  four lysine residues (residues 35, 68, 72, and 84) in this region of the protein coordinate the DNA 5Ј-phosphate that may exist in a gapped DNA (17). However, in structures of ␤-pol bound to a one-nucleotide DNA gap only Lys 35 and Lys 68 coordinate the 5Ј-phosphate in the short gap (21). Based on information available from the crystal and NMR structures of the 8-kDa domain and from biochemical studies of the protein (14 -16, 22), we have now conducted site-directed mutagenesis to alter 9 residues in the 8-kDa domain of ␤-pol that appear to contribute key interactions to ssDNA binding, 5Ј-phosphate recognition in a DNA gap, and dRP lyase activity. The results allow us to identify two critical residues, Lys 35 and Lys 72 , for 5Ј-phosphate recognition and dRP lyase chemistry, respectively.
Mutagenesis, Expression, and Purification of the Recombinant Wildtype 8-kDa and Mutant Proteins-Oligonucleotide site-directed mutagenesis was performed using a procedure described previously (25). Recombinant amino-terminal 8-kDa domain and the mutant proteins were overexpressed and purified as described (22).
Circular Dichroism Spectroscopy-For CD analysis, the wild-type 8-kDa domain of ␤-pol and the mutants were further purified by gel filtration on HPLC using a Bio-Sel SEC-125 (300 ϫ 7.8 mm) size exclusion column (Bio-Rad). Buffer consisted of 5 mM Tris-HCl, pH 7.2, and 500 mM NaCl. The chromatogram for the mutants was compared with the chromatogram of the highly purified wild-type 8-kDa domain in selecting the pooled peak fraction for each mutant. The concentrations of the mutant proteins were determined by UV absorption at 280 nm (⑀ 280 ϭ 5440 M Ϫ1 cm Ϫ1 ). CD measurements were performed on a Jasco J715 spectropolarimeter in a 1-cm cell at 25°C. The CD spectra were collected from 260 to 200 nm at a resolution of 1 nm using up to 8 scans. The per residue molar ellipticity (deg⅐cm 2 dmol Ϫ1 ) was calculated from the concentration for the 87-residue polypeptide.
Template-Primer Annealing-Lyophilized oligonucleotides were resuspended in 10 mM Tris-HCl, pH 7.4, and 1 mM EDTA, and the concentrations were determined from their UV absorbance at 260 nm. Template-primers were annealed as described previously (27). The sequence of the 5-nucleotide gapped DNA used was as follows.
(P 1 , 17-mer) (P 2 ,17-mer) 5Ј-CGAGCCATGGCCGCTAG TTTTTTGCGGTGCCAGG-3Ј 3Ј-GCTCGGTACCGGCGATCCGTCTAAAAAACGCCACGGTCC-5Ј (T,39-mer) SEQUENCE 1 UV Cross-linking to Gapped DNA-Purified wild-type 8-kDa domain or the mutant protein (1.4 M) was mixed with the gapped DNA template-primer (0.7 M) in a reaction mixture containing 20 mM Tris-HCl, pH 8.0, 20 mM NaCl, 1 mM EDTA, and 5 mM MgCl 2 and incubated at room temperature for 15 min. The samples were irradiated, and the photochemical cross-linked 8-kDa protein-DNA complexes were separated and analyzed as described (28). The relative activities were determined by the ability of the mutant 8-kDa domain to discriminate between a 5Ј-phosphorylated and unphosphorylated terminus in a gap as compared with wild-type protein. The wild-type 8-kDa binds to a 5Ј-phosphorylated 5-nucleotide gap 2.5-fold more readily than when the gap has a 5Ј-hydroxyl (i.e. amount of cross-linked complex in a 5Јphosphorylated gap/amount of cross-linked complex in a unphosphorylated gap). The loss of 5Ј-phosphate recognition results in equal crosslinked mutant complexes in the presence or absence of a 5Ј-phosphate in the gap. To quantify the cross-linked complexes, the autoradiogram was scanned on an Imager Master VDS, and the data were analyzed using ImageMaster software.
Preparation of the dRP Lyase Substrate-32 P-Labeled uracil containing duplex DNA (62.5 nM) was pretreated with 10 nM UDG in 100 l of buffer containing 70 mM Hepes, pH 7.4, 0.5 mM EDTA, and 0.2 mM dithiothreitol. The reaction mixture was incubated at 37°C. After a 20-min incubation, the reaction mixture was supplemented with 5 mM MgCl 2 and 10 nM AP endonuclease, and the incubation was continued for another 20 min.
dRP Lyase Activity Assay-dRP lyase activity was performed in a reaction mixture (10 l) containing 50 mM Hepes, pH 7.4, 5 mM MgCl 2 , 2 mM dithiothreitol, and 20 nM preincised 32 P-labeled AP site containing DNA. The reaction was initiated by adding 10 nM wild-type 8-kDa domain or a mutant derivative and incubated at 37°C for 10 min. The reaction was terminated by transfer to 0 -1°C, and the DNA product was stabilized by addition of 2 M NaBH 4 to a final concentration of 340 mM and incubation 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 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 20% polyacrylamide gel containing 8 M urea in 89 mM Tris-HCl, 89 mM boric acid, and 2 mM EDTA, pH 8.8, and visualized by autoradiography. To quantify the product, gels were scanned on a PhosphorImager 450 (Molecular Dynamics), and the data were analyzed using ImageQuant software.

Structure-guided Site-directed Mutagenesis and Purification of Mutant
Proteins-To probe the functional importance of residues in the amino-terminal 8-kDa domain, 9 residues were selected from x-ray (14,17,21) and NMR (15,16) structural analyses and altered by site-directed mutagenesis as described previously (25). These residues were Phe 25 , His 34 , Lys 35 , Lys 60 , Lys 68 , Glu 71 , Lys 72 , Glu 75 , and Lys 84 . The residues and alterations selected for mutagenesis were based on the proposed role of each residue in DNA binding (15,22) or putative role in dRP lyase chemistry, as discussed in detail previously (11,12,28). Briefly, the primary structure of a portion of the HhH motif (residues 55-79) in the 8-kDa domain is similar to that of the HhH motif in E. coli endonuclease III glycosylase/AP lyase (17)(18)(19)(20). In addition, significant intermolecular nuclear Overhauser effects and chemical shift changes were observed in the 1 H-15 N heteronuclear single quantum correlation NMR spectra for surface-exposed residues upon addition of ssDNA (15), including residues Phe 25 , Lys 60 , Glu 71 , Lys 72 , and Glu 75 . Lys 68 is adjacent to residues showing chemical shift changes. Lys 84 is in the NMR unstructured linker region between the 8-kDa and 31-kDa domains but is coordinated with the 5Ј-phosphate in one of the crystal structures, along with Lys 35 , Lys 60 , and Lys 72 (16,17). In the NMR structure, the adjacent flexible ⍀-loop residue Lys 35 contributes to the basal surface charge potential and is adjacent to Lys 72 and Lys 68 (15). Also in the ⍀-loop, His 34 has been shown to cross-link with ssDNA (22) and to stack with a template strand base in the crystal structure of ␤-pol bound to DNA (21).
The positions of the 9 residues selected for mutagenesis in the crystal structure of ␤-pol bound to one-nucleotide gapped DNA are illustrated in Fig. 1. Single mutations were constructed at residues Phe 25 , His 34 , Glu 71 , and Glu 75 and at the lysines at positions 35, 60, 68, 72, and 84. His 34 was changed to glycine, and Glu 71 was changed to glutamine to remove the base-stacking potential and carboxylate moieties, respectively. Double mutants were constructed at Lys 68 and Lys 72 (K68A/ K72A) and Lys 68 and Glu 71 (K68A/E71Q), and a triple alanine mutant (K35A/K68A/K72A) was also examined. Each 8-kDa protein was expressed in E. coli, and the recombinant protein was found to be soluble in the crude cell extract. Following purification, SDS-PAGE analysis indicated that the mutant 8-kDa polypeptides were greater than 95% homogeneous (data not shown). During chromatography and electrophoresis, the mutant proteins behaved similarly to the wild-type 8-kDa domain, suggesting that the alteration did not cause gross changes in the structural properties of the proteins.
CD Spectra of Mutant Proteins-The CD spectra of single mutants K35A, K68A, K72A, K84A, E71Q, and E75A, the double mutant (K68A/K72A), and the triple mutant (K35A/ K68A/K72A) of the 8-kDa domain were similar to the CD spectrum of the wild-type 8-kDa domain. The comparable maximal negative ellipticities at 208 and 220 nm indicated that the overall helical structure in the mutants was similar to that seen in wild-type 8-kDa domain (Fig. 2). The differences in the absolute molar ellipticity at 220 nm for the mutants relative to wild type are due to error inherent in the protein concentration determinations. Positive ellipticity in the K35A and K72A single mutants at 240 -260 nm was due to the presence of a minor DNA contamination eluting from the ssDNA cellulose column during purification. This DNA contaminant, however, is not seen in the CD spectrum of the highly purified wild-type 8-kDa protein. There is an increase in the ratio of negative ellipticity at 222 nm to negative ellipticity at 208 nm in the K68A/K72A mutant in comparison with the relative ellipticities at these wavelengths observed with wild type (i.e. more negative ellipticity at 222 nm and less negative ellipticity at 208 nm in the double mutant). The increase in the ratio of the maximal negative ellipticity at 222 nm versus 208 nm in the double mutant suggests an increase in helical structure in comparison with the wild-type 8-kDa domain. Similar results for the K35A/ K68A/K72A triple mutant were also observed. On the basis of these results, we conclude that the substitutions chosen at these surface-exposed residues do not adversely affect the overall structure of the 8-kDa domain and that the effects of the mutations on ssDNA binding, 5Ј-phosphate recognition in gapped DNA, and dRP lyase activity are likely the result of the loss of specific side chain functionality.
Single-stranded DNA Binding Activity-The ssDNA binding activity of the 39-kDa enzyme has been demonstrated to reside in the 8-kDa domain (22). The ssDNA binding activities of the wild-type 8-kDa domain and mutant proteins were examined by an assay involving photochemical cross-linking to oligo(dT) 16 as described previously (22). To assay ssDNA binding activity, purified protein was mixed with [ 32 P](dT) 16 and irradiated with UV light to covalently cross-link the bound ligand. The cross-linked products were separated by SDS-PAGE and scored by autoradiography and PhosphorImager scanning ( Fig.  3 and Table I). Under the conditions of the assay, the level of cross-linking is proportional to the equilibrium association constant, K a (22). Quantitative analysis of the UV cross-linked products indicated that the single mutants K35A, K60A, K68A, and F25W, and the triple mutant K35A/K68A/K72A were reduced in ssDNA binding by approximately 60 -75% compared with wild type (Fig. 3 and Table I). The other mutants had similar ssDNA binding as wild type. Alanine substitution for Lys 35 , Lys 68 , and Lys 60 had the weakest binding. Interestingly, alanine substitution for Lys 72 did not affect ssDNA binding activity. In the intact enzyme, Lys 72 has been shown to be a target for pyridoxal 5Ј-phosphate modification and was protected by dNTP binding (29). Lys 72 has also been implicated in forming a Schiff base intermediate with abasic site DNA (12). The mutant bearing a glycine substitution for His 34 was slightly reduced (Ϸ25%) in ssDNA binding activity. This histi-  (30). The molecular surface of only the 8-kDa domain is illustrated. The 31-kDa domain, which possesses nucleotidyltransferase activity, is omitted for clarity. The DNA is orange, and the 5Ј-deoxynucleoside monophosphate in the one nucleotide gap is yellow. The 5Ј-phosphate of this terminal nucleotide binds near a lysine-rich pocket. Residues altered by site-directed mutagenesis are indicated. Lysine residues 35, 60, 68, 72, and 84 (blue) were altered individually or in combination. Additionally, residues in red have been implicated in ssDNA binding (Phe 25 , His 34 , Glu 71 , and Glu 75 ) and were modified as outlined under "Experimental Procedures." The Na ϩ ion bound to the hairpin of the HhH motif is indicated (purple). The HhH structural motif has been identified by primary sequence alignment in several other DNA repair enzymes and suggested to play a role in DNA binding (17,18). dine residue was found to be covalently UV cross-linked to ssDNA (22), and in the crystal structure (Fig. 1), the imidazole ring was observed to stabilize the template base near the polymerase active site by stacking interactions (21). In contrast, alanine substitution for Lys 68 reduced ssDNA binding activity by 70%. Surprisingly, a double alanine mutation at Lys 68 and Lys 72 exhibited wild-type ssDNA binding activity, suggesting a compensating effect in the double mutant protein. The triple mutant, K35A/K68A/K72A, showed further reduction in ssDNA binding activity over that observed with the K68A/ K72A double mutant (Fig. 3 and Table I). These results confirm the importance of Lys 35 in ssDNA binding activity.
5Ј-Phosphate Recognition in Gapped DNA-Earlier, we had shown that cross-linking of ␤-pol to gapped DNA is dependent on a 5Ј-phosphate moiety in the gap. This DNA gap binding of ␤-pol was directed by the amino-terminal 8-kDa domain (27). Additionally, the crystal structure of ␤-pol in complex with gapped DNA substrates suggests that the 5Ј-phosphate in the gap is coordinated either by four lysines of the 8-kDa domain (residues 35, 68, 72, and 84) (17), or Lys 35 and Lys 68 (21), as illustrated in Fig. 1. These results were consistent with NMR studies showing protein-DNA interactions at or near these same lysine residues, among other residues (15). To identify the residue(s) involved in 5Ј-phosphate recognition in gapped DNA, a synthetic gapped DNA substrate was formed by annealing two 17-residue oligonucleotides (designated P 1 and P 2 ) to a 39-residue template (T) creating a 5-nucleotide gap between the 3Ј-hydroxyl of P 1 and the 5Ј-phosphate or 5Ј-hydroxyl of P 2 (see Fig. 4B and "Experimental Procedures"). This DNA substrate was incubated with wild-type or mutant proteins of the 8-kDa domain, and the complex was then photochemically cross-linked with UV light (27). To score the covalently crosslinked complexes, the 5Ј-end of the P 1 oligonucleotide was 32 P-labeled. After cross-linking, the mixture was separated by SDS-PAGE, and the gel was analyzed by autoradiography ( Fig.  4A and Table I). The results show that cross-linking between the 8-kDa domain, template (T), and primer (P 1 ) is strongly influenced by the phosphate group on the 5Ј-end of P 2 as previously demonstrated (27). Whereas the Lys 35 alanine mutant showed a strong decline in 5Ј-phosphate recognition activity, alanine substitutions of Lys 60 , Lys 68 , or Lys 72 retained the recognition activity ( Fig. 4A and Table I). The E75A mutant also displayed a diminished ability to discriminate between the phosphorylated and unphosphorylated gaps. The lower amount of cross-linking for the K60A and K68A mutants is consistent with the lower ssDNA binding activity of these mutants, but both mutants retained the 5Ј-phosphate recognition activity of wild type (Fig. 4A and Table I). Both NMR and crystallography data suggested that Lys 35 and Lys 68 are sites for 5Ј-phosphate contact (15,17,21). Our results on 5Ј-phosphate recognition with the K35A mutant support the conclusion that Lys 35 coordinates the 5Ј-phosphate group, whereas Lys 68 does not.
The finding that Lys 35 is a key residue in 5Ј-phosphate recognition is further supported from the results obtained with the triple mutant (K35A/K68A/K72A). The triple mutant failed to discriminate between phosphorylated and unphosphorylated forms of the gapped DNA ligand (lanes 14 -15). Taken together, these results indicate that lysine residues 60, 68, and 35 are  16 ) and free probe are indicated. The gel was quantified as outlined under "Experimental Procedures," and the results are summarized in Table I.
ϩϩϩϩ ϩϩϩϩ ϩϩϩϩ a Activity relative to wild-type: ϩϩϩϩ, 75-100%; ϩϩϩ, 50 -75%; ϩϩ, 25-50%; ϩ, 1-25%; Ϫ, Ͻ1%; ND, not determined. Data were analyzed as described under "Experimental Procedures" and tabulated from the data in Figs. 3-5 for ssDNA binding, 5Ј-phosphate recognition, and dRP lyase activities, respectively. b The relative activity is determined by the ability of the mutant 8-kDa domain to discriminate between a 5Ј-phosphorylated and unphosphorylated terminus in a gap as compared with wild type as described under "Experimental Procedures." important for ssDNA binding, whereas Lys 35 is critical for 5Ј-phosphate recognition. Lys 84 and the exposed residues displaying significant NMR chemical shift changes upon binding (dT) 8 , such as Glu 71 , had little or no influence on the DNA binding activities tested when mutated to alanine or glutamine (Figs. 3 and 4 and Table I).
dRP Lyase Activity-To examine the dRP lyase activity of wild-type and mutant enzymes, we utilized a 49-residue oligonucleotide duplex DNA, which contained a uracil residue at position 21. The uracil-containing strand was 3Ј-end labeled with [␣-32 P]ddAMP and annealed to its complementary DNA strand. To prepare DNA substrate for the dRP lyase reaction, the 32 P-labeled duplex DNA was pretreated with UDG and AP endonuclease. Thus, the resulting DNA substrate contains a 5Ј-dRP group and a 32 P-labeled ddAMP residue at the 3Ј-end of the downstream DNA strand (Fig. 5B). Wild-type and mutant enzymes were incubated with this pretreated 32 P-labeled duplex DNA, and at the end of each reaction period the DNA product was stabilized by NaBH 4 . The release of 5Ј-dRP from the 32 P-labeled substrate was determined by the appearance of a new radioactive electrophoretic band migrating approximately one-half nucleotide faster than the substrate (Fig. 5A). Results of dRP lyase activity of alanine substitutions are shown (Fig. 5A), but the results of all the mutants have been summarized ( Fig. 5C and Table I). Our results indicate that Lys 68 retained wild-type dRP lyase activity when mutated to alanine. We had previously considered that this residue was a candidate nucleophile involved in Schiff base formation during the lyase reaction, based on sequence alignment with endonuclease III (28). Mutagenesis of the corresponding lysine (residue 120) in the HhH motif of endonuclease III strongly reduced AP site lyase activity (20). In contrast, our results indicate that Lys 68 is not a candidate for Schiff base formation in the dRP lyase reaction catalyzed by the 8-kDa domain of ␤-pol. While mutants E75A, K35A, H34G, and K60A retained approximately 40 -75% dRP lyase activity, the K72A had less than 10% activity of wild type (Fig. 5C). There was no further decline in dRP lyase activity with the K68A/K72A double mutant over that of the K72A single mutant, suggesting that a Schiff base nucleophile role for Lys 68 in the K72A mutant is unlikely.
Near wild-type dRP lyase activity was observed for F25W, K84A, E71Q, and the K68A/E71Q double mutant. The triple mutant, K35A/K68A/K72A, was essentially devoid of dRP lyase activity (Ͻ1%), and use of a 10-fold higher protein concentration also failed to reveal detectable activity (data not shown).
Our data suggest that Lys 72 is the best candidate as the Schiff base-forming nucleophile. This interpretation must be considered in light of the results obtained by Piersen et al. (12), who demonstrated that a Schiff base intermediate could be formed with the K72A mutant of intact ␤-pol, albeit at a strongly reduced efficiency. Reduction in the dRP lyase activity of the alanine mutant of Lys 60 may be attributed to a significant loss in ssDNA binding. A role of His 34 in ␤-elimination chemistry has also been proposed, and the loss in dRP lyase activity of the H34G mutant is consistent with a role for this residue in lyase chemistry, as suggested previously (28).
The activities characterized for each 8-kDa domain mutant are summarized in Table I. Although Lys 72 is the candidate for Schiff base formation during the dRP lyase reaction, the K72A mutant retained significant residual activity. Thus, Lys 72 may be the preferred, but not obligatory, Schiff base nucleophile. The residual activity with K72A may be contributed from an alternate Schiff base nucleophile in the lysine-rich pocket formed in the wild type by Lys 35 , Lys 68 , Lys 72 , and Lys 84 (21). In the absence of Lys 72 , partial deprotonation at a nearby amine, such as Lys 35 , could activate this amine and allow it to function as a backup Schiff base nucleophile. This explanation is strengthened by the observation of complete loss of dRP lyase activity in the triple mutant where lysines 35, 68, and 72 are changed to alanine. Taken together, our results clearly demonstrate that Lys 72 and Lys 35 are involved in the dRP lyase reaction catalyzed by the 8-kDa domain and that Lys 35 is involved in 5Ј-phosphate recognition. Lys 35 may stabilize the leaving group in the dRP lyase reaction through its interaction with the 5Ј-phosphate of the DNA terminus, and Lys 35 could be activated in this role by His 34 .
In conclusion, it is interesting to compare the results of our site-directed mutagenesis with NMR and x-ray crystallography studies of the 8-kDa domain. Alanine substitution for lysine at residues 35, 60, and 72 resulted in loss of function for 5Јphosphate recognition, DNA binding, and dRP lyase, respectively. In each case, this loss of function was perfectly consistent with the structural data, e.g. Fig. 1. Similarly, Lys 84 and Glu 71 are distal to the dRP lyase active site and also to the DNA FIG. 5. Relative dRP lyase activity of mutants of the ␤-pol 8-kDa domain. The reaction conditions and product analysis are described under "Experimental Procedures." A, a photograph of the autoradiogram illustrating the dRP lyase activity for wild-type and mutant forms (identified at the top of the photograph) of the 8-kDa domain; B, a representation of the AP site DNA (49-mer), 3Ј-labeled substrate after AP endonuclease cleavage (29-mer with the 5Ј-sugar phosphate), and the dRP lyase product (29-mer without sugar-phosphate). The 3Ј-end of the oligodeoxyribonucleotide was labeled with [␣-32 P]ddAMP, and as a result of addition of ddAMP at the 3Ј-end the AP site-cleaved DNA product will be one nucleotide longer (29-mer), as shown. C, a histogram of the dRP lyase activity of all mutants of the ␤-pol 8-kDa domain relative to wild type is shown. The identity of the substrate and the product is indicated. binding interface, and the K84A and E71Q mutants are not altered in any of the activities tested. The H34G mutant shows loss of a portion of the dRP lyase activity, and this is consistent with predictions from NMR studies (27). In contrast, the E75A mutant has lower 5Ј-phosphate recognition, and this is not suggested by the structural data. The results with Lys 68 are surprising. The K68A mutant, as with K35A and K60A, is lower in DNA binding activity. However, K68A is not lower in either 5Ј-phosphate recognition or dRP lyase activity, as expected.