Portal de Periódicos da CAPES

DNA Polymerase I of Mycobacterium tuberculosis

2002; Elsevier BV; Volume: 277; Issue: 3 Linguagem: Inglês

10.1074/jbc.m108536200

1083-351X

Cindy Jo Arrigo, Kamalendra Singh, Mukund J. Modak,

HIV/AIDS drug development and treatment

The highly conserved GXD sequence present in the Mycobacterium tuberculosis DNA polymerase I corresponds to a hinge region in the finger subdomain connecting M and N helices of Escherichia coli pol I. An examination of the crystal structures of pol I family polymerases reveals that the invariant aspartate of the hinge forms a salt bridge with the conserved arginine of the O-helix and an H-bond with Gln-708. To clarify the role of this region, we generated and characterized conserved and nonconserved mutant derivatives of this aspartate, the preceding glutamate and the Gln in TB pol I. For comparison, D732A mutein of pol I was also included. The muteins representing conserved aspartate (Asp-707 of TB pol I or Asp-732 of pol I) showed a strongKm(dNTP) effect and minor alteration inKd(DNA), with about 10–20-fold decrease in overall catalytic efficiency. The TB muteins, E706A and Q683A, have less pronounced deviations from the wild-type enzyme. Further examination of D707A of TB pol I showed no alteration in the processivity or the dideoxynucleotide sensitivity patterns. However, both TB pol D707A and homologous E. coli D732A failed to form a stable E·DNA·dNTP ternary complex. These results suggest that the aspartate in the hinge region is catalytically important and is required for dNTP binding and in the formation of a prepolymerase ternary complex. The highly conserved GXD sequence present in the Mycobacterium tuberculosis DNA polymerase I corresponds to a hinge region in the finger subdomain connecting M and N helices of Escherichia coli pol I. An examination of the crystal structures of pol I family polymerases reveals that the invariant aspartate of the hinge forms a salt bridge with the conserved arginine of the O-helix and an H-bond with Gln-708. To clarify the role of this region, we generated and characterized conserved and nonconserved mutant derivatives of this aspartate, the preceding glutamate and the Gln in TB pol I. For comparison, D732A mutein of pol I was also included. The muteins representing conserved aspartate (Asp-707 of TB pol I or Asp-732 of pol I) showed a strongKm(dNTP) effect and minor alteration inKd(DNA), with about 10–20-fold decrease in overall catalytic efficiency. The TB muteins, E706A and Q683A, have less pronounced deviations from the wild-type enzyme. Further examination of D707A of TB pol I showed no alteration in the processivity or the dideoxynucleotide sensitivity patterns. However, both TB pol D707A and homologous E. coli D732A failed to form a stable E·DNA·dNTP ternary complex. These results suggest that the aspartate in the hinge region is catalytically important and is required for dNTP binding and in the formation of a prepolymerase ternary complex. Mycobacterium tuberculosis DNA polymerase I E. coli DNA polymerase I Klenow fragment of E. coliDNA polymerase I dithiothreitol DNA polymerases belong to a superfamily of enzymes referred to as nucleic acid polymerases (1Delarue M. Poch O. Tordo N. Moras D. Argos P. Protein Eng. 1990; 3: 461-467Crossref PubMed Scopus (574) Google Scholar). Nucleic acid polymerases from diverse families share a common structural architecture (2Braithwaite D.K. Ito J. Nucleic Acids Res. 1993; 21: 787-802Crossref PubMed Scopus (528) Google Scholar, 3Singh K. Modak M.J. Trends Biochem. Sci. 1998; 21: 186-190Google Scholar, 4Steitz T.A. J. Biol. Chem. 1999; 274: 17395-17398Abstract Full Text Full Text PDF PubMed Scopus (690) Google Scholar). Their structural elements or subdomains have been likened to the fingers, palm, and thumb of a half-opened right hand. All polymerases perform the nucleotidyl transfer reaction using a similar catalytic reaction pathway (4Steitz T.A. J. Biol. Chem. 1999; 274: 17395-17398Abstract Full Text Full Text PDF PubMed Scopus (690) Google Scholar) and contain at least two essential carboxylates at the catalytic site, which coordinate the divalent metal ions and induce nucleophilic attack by 3′ OH of the primer onto the α-phosphate of an incoming nucleotide substrate.DNA polymerase I of Mycobacterium tuberculosis (TB pol I)1 is a member of the family A or pol I family of polymerases whose canonical family memberEscherichia coli pol I, defines the group (1Delarue M. Poch O. Tordo N. Moras D. Argos P. Protein Eng. 1990; 3: 461-467Crossref PubMed Scopus (574) Google Scholar). This prototype pol I is a DNA-dependent DNA polymerase with a modular structure and contains both polymerase and nuclease activities in a single polypeptide chain. Our interest in TB pol I stems from the parent organism, M. tuberculosis, the etiologic agent of tuberculosis. This disease kills more people around the world each year than any other infectious disease. The DNA polymerase I of this pathogen, similar to other bacterial species (pol I family) is expected to be an important participant in genome stability and DNA repair functions, especially during stationary phase of the bacterial life cycle. Therefore, the TB polymerase provides an attractive target for possible chemotherapeutic intervention. Toward this ultimate goal, we have begun to investigate functional similarities of this polymerase with those of the prototype, E. coli pol I. We have purified both the 100-kDa holoenzyme (containing 5′-3′ exonuclease, 3′-5′ exonuclease, and 3′- 5′ polymerase domains) and its 63-kDa fragment equivalent to Klenow fragment of E. coli pol I and have compared their catalytic properties. A number of similarities in these properties have been noted as expected from ∼60% sequence homology (in polymerase domain) between the two species. 2C. J. Arrigo, K. Singh, N. Kaushik, and M. J. Modak, manuscript in preparation. 2C. J. Arrigo, K. Singh, N. Kaushik, and M. J. Modak, manuscript in preparation. We were particularly interested in the substrate-binding domain of TB pol I for it is perceived as an excellent target for the development of inhibitors. The primary amino acid sequence comparison (Fig. 1) shows the conservation of catalytically important amino acids in many members of the pol I family suggesting conserved catalytic mechanism. Thus, the commonality of reaction mechanism and possible structural similarity of the catalytic segment of TB pol I with the prototype, E. coli pol I is expected to permit the translation of new findings from one polymerase to other members of this family.The crystal structures of several pol I family members in apo and binary (dNTP or template/primer bound) complexes show a highly conserved aspartate (732 of E. coli pol I) present in the three-residue long hinge region between the M and N helices of pol I (5Korolev S. Nayal M. Di Barnes W.M. Cera E. Waksman G. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9264-9268Crossref PubMed Scopus (169) Google Scholar, 6Beese L.S. Derbyshire V. Steitz T.A. Science. 1993; 260: 352-355Crossref PubMed Scopus (450) Google Scholar, 7Beese L.S. Friedman J.M. Steitz T.A. Biochemistry. 1993; 32: 14095-14101Crossref PubMed Scopus (170) Google Scholar, 8Li Y. Kong Y. Korolev S. Waksman G. Protein Sci. 1998; 7: 1116-1123Crossref PubMed Scopus (95) Google Scholar). This aspartate appears to form a salt bridge with Arg-754 (E. coli pol I position) of the O-helix. These structural observations suggest that the conserved aspartate in the hinge between M and N helices stabilizes the position of arginine in the O-helix and should complement the function of Arg-754 in pol I and its equivalent in KlenTaq (Arg-659). To clarify the functional role of this conserved aspartate, we introduced conserved and nonconserved mutations within the M to N helical hinge of TB pol I to generate D707A, D707E, and E706A enzyme species.Since, in the ternary complex of KlenTaq, Asp-637 (equivalent to Asp-707 of TB pol I) also forms an H-bond with Gln-613 (TB Gln-683 orE. coli Gln-708 equivalent), we generated the Q683A mutant and characterized its biochemical properties. In addition, a double mutant, Q683N/D707E was constructed with the expectation that a putative defect in Q683N may be compensated for by increased side chain length in D707E. The results presented in this article suggest that the conserved Asp-707 is essential for catalytic function and that its properties are very similar to those reported for Arg-754 (10Astatke M. Grindley N.D.F. Joyce C.M. J. Biol. Chem. 1995; 270: 1945-1954Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 11Kaushik N. Pandey V.N. Modak M.J. Biochemistry. 1996; 35: 7256-7266Crossref PubMed Scopus (42) Google Scholar). Furthermore, this aspartate as well as the equivalent aspartate inE. coli pol I appears to be required for the substrate binding and ternary complex formation, the latter requiring the motion of the O-helix.DISCUSSIONThe sequence alignment of family A polymerase (including pol I) members and the examination of the available crystal structures of pol I from different organisms reveals the presence of a highly conserved aspartate within the GXD sequence located on the hinge region between M and N helices (Fig. 1). In addition, the adjoining region constituted by the “O-helix” in the prototype E. coli pol I also appears to be highly conserved among all the members of this family including TB pol I (Fig. 1). The available crystal structures of pol I family members, KF (6Beese L.S. Derbyshire V. Steitz T.A. Science. 1993; 260: 352-355Crossref PubMed Scopus (450) Google Scholar), KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar), and T7 DNA polymerase (22Doublie S. Tabor S. Long A.M. Richardson C. Ellenberger T. Nature. 1998; 391: 251-258Crossref PubMed Scopus (1099) Google Scholar), show that the connector function of this three-residue region is common and that the position of the hinge region is in the close physical vicinity of the O-helix of the “fingers” subdomain. In the GXD region, both glycine and aspartate represent conserved residues. Since glycine residues generally do not participate in the reaction but rather provide local structural integrity, we chose not to investigate the role of this glycine. To understand the function of the aspartate in the GXD sequence of TB pol I, we carried out site-directed mutagenesis at Asp-707. Since Asp-707 is preceded by another carboxylate (Glu-706) and in other members of this family, Asp-707 equivalent aspartate interacts with conserved Gln, we also included Glu-706 and Gln-683 of TB pol I in the present studies. The functional significance of these residues was deduced by the examination of properties of mutant enzymes such as DNA polymerase activity, kinetics of nucleotide incorporation, dNTP and DNA binding affinity, ternary complex formation, processivity, and fidelity.The catalytic activity of the mutant TB polymerases showed that substitution of Asp-707 to Ala or Glu resulted in a significant decrease in DNA polymerase activity suggesting an important role for this residue in some catalytic function of pol I. However, no significant effect on the catalytic activity was apparent when its neighbor, Glu-706 was converted to Ala-706. Similarly, Gln-683 did not appear to be essential for the activity. Nevertheless, a double mutant D707E/Q683N was severely defective in catalytic function, probably representing the effect of D707E (see Table II). Very similar results were obtained for KF protein when its Asp-732 (equivalent of Asp-707 of TB pol I) was mutated to Ala-732. Therefore, Asp-707 of TB pol I is concluded to have a major role in the catalytic function in a manner similar to the equivalent aspartate present in the hinge region of M and N helices in other pol I family members. The kinetic analysis of D707A and other mutants clearly suggested that one of the defects with D707A, D707E, and D707E/Q673N is the decrease in the affinity for the dNTP substrate as judged by 10–30-fold increase in theKm(dNTP) (Table III) and loss of their ability to cross-link to substrate dNTP (Fig. 2). Most interestingly, the catalytic activity of these mutants was little affected when Mn2+ was substituted in place of Mg2+ as the divalent cation (Table II). These properties are reminiscent of those of O-helix muteins from E. coli pol I, namely Y766A, F762A, and R754A (10, 11). Further examination of Asp-707 muteins also showed the severe loss of pyrophosphorolysis activity (Fig. 3) and significant increase in the apparentKm(PPi), suggesting defect in the binding of PPi. The loss of binding affinity for substrate dNTP as well as for the product PPi by D707A is suggestive of the participation of Asp-707 in both of these functions. However, no change in the processivity of DNA synthesis or change in the sensitivity to dideoxynucleotides (Fig. 5) was observed with Asp-707 muteins. In addition, D707A seems to exhibit better fidelity properties compared with wild type (Fig. 4). Nevertheless, the persistent loss of catalytic activity for Asp-707 muteins, even in the presence of saturating concentrations of both template/primer and dNTP, was somewhat puzzling. We therefore, determined if the Asp-707 (Asp-732 of KF) muteins, were capable of forming a stable ternary complex. As described under “Results,” D707A of TB pol I and its E. coli counterpart D732A were unable to form stable ternary complexes (Fig. 6), even in the presence of high concentrations of substrate dNTP. Thus Asp-707 in the hinge region is concluded to be an essential component for the motion of finger subdomain (mainly O-helix) which is implicated in the ternary complex formation.From the above discussion, it is clear that most of the properties displayed by Asp-707 mutein of TB pol I are common to D732A of E. coli pol I (KF). Therefore, it seems likely that the process of dNTP binding and the ternary complex formation via the motion of the O-helix region is similar in both TB pol I and E. coli pol I. Unfortunately, the structure of TB pol I has not yet been resolved preventing us from drawing firm conclusions regarding a contribution of the hinge region aspartate in the catalytic process. However, structures of KF as well as KlenTaq have been firmly established and this has permitted us to draw valid conclusions for the role of Asp-732 in the catalytic process of pol I. Conserved D in the hinge region of M to N helices in pol I has Arg-754 of the O-helix as a major interacting partner. In the apo, DNA/dNTP-bound binary and DNA-dNTP-bound ternary complexes, a salt bridge between these two residues is consistently observed. Since the properties of muteins of either Asp-732 or Arg-754 of pol I are quite similar, we are tempted to suggest that the loss of the salt bridge between the Asp of the hinge and the Arg of the O-helix may be the major factor affecting catalytic function. Another interesting observation pertains to ∼3-fold decrease in the binding affinity for template/primer by Asp-732 muteins. This observation is suggestive of some role for Asp-732 in the binding/stabilization of template/primer. One possible scenario by which Asp-707 may indirectly participate in the process is as follows. Examination of the ternary complex structure of KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar) shows that its Asp-637 (equivalent to Asp-707 of TB pol I and Asp-732 of pol I) interacts with His-639 (equivalent to His-709 of TB pol I and His-734 of pol I). This histidine in pol I has been suggested to participate in the binding/stabilization of single-stranded template (10Astatke M. Grindley N.D.F. Joyce C.M. J. Biol. Chem. 1995; 270: 1945-1954Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). We have noted that a major factor in the stabilization of template/primer to DNA polymerases of this family is the single-stranded template overhang (to be published elsewhere). 3R. Gangurde and M. J. Modak, manuscript in preparation. Thus Asp-732 by means of its van der Waals interaction with the β carbon of His-709, may indirectly contribute to the template/primer binding process in pol I family of polymerases.Most interestingly, the loss in the ability of mutein of pol I Asp-732 to form the closed ternary complex (Fig. 6) may be further examined by comparing the position of the O-helix in the binary (E·TP) and ternary (E·TP·dNTP) complexes (Fig.7). As Arg-754 of the O-helix is seen to form the salt bridge with Asp-732 of the hinge in all complexes, what follows is the implication that the “salt bridge” structure is a responsible factor for the motion of the O-helix. Indeed this will also explain slight repositioning of the M and N helices (due to a pull exerted in the hinge region) that can be seen in the ternary complexes.Figure 7Position of O-helix in the apo and ternary complex structures of KlenTaq. This figure shows the rotation of O-helix (rendered as a ribbon) upon the formation of the closed ternary complex (PDB numbers: 1KTQ and 3 KTQ). The Cα positions of Arg-754, Lys-758, Phe-762, and Tyr-766 in both forms are shown as spheres. The ddNTP in the ternary complex is represented in the ball and stick. As seen in the crystal structures, the position of O-helix is rotated ∼41 Å toward the cleft upon the ternary complex formation as shown by thearrow in the figure. This rotation of O-helix displaces the Cα positions of Arg-754 and Lys-758 by 9 and 5.4 Å, respectively. The side chains of four critical residues of O-helix (Arg-754, Lys-758, Phe-762, and Tyr-766) in the ternary complex are displayed insticks. The Arg-754 side chain in this figure is behind the ribbon represented O-helix. The displacement in the position of Asp-732, before and after ternary complex formation, is also shown by abroken arrow. The coordinate files were imported from Protein Data Bank (Research Collaboratory for Structural Bioinformatics, Rutgers University: www.rcsb.org). This figure was generated by MolMol (25Koradi R. Belleter M. Wuthrich K. J. Mol. Graph. 1996; 14: 51-55Crossref PubMed Scopus (6468) Google Scholar).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Examination of the O-helix motion seen in the crystal structures of KlenTaq, as depicted in Fig. 7, also indicates that the maximal motion (∼9.4 Å) occurs at Arg-754, while the catalytically important Lys-758 moves by about 5.8 Å. Little change is observed in the positions of Phe-762 and Tyr-766. It is therefore possible, that failure to form the ternary complex may also fail to bring Lys-758 close to the active site to interact with the α-phosphate of incoming dNTP, resulting in the catalytically inactive enzyme.In summary, the properties of conserved aspartate of TB pol I corresponding to the joining region of M and N helices, suggest that it is required for dNTP binding, pyrophosphorolysis as well as ternary complex formation. Many analogous properties observed with equivalent aspartate in E. coli pol I have permitted the elucidation of a tentative structural mechanism by which this aspartate may contribute to the process of catalysis in pol I family of enzymes. DNA polymerases belong to a superfamily of enzymes referred to as nucleic acid polymerases (1Delarue M. Poch O. Tordo N. Moras D. Argos P. Protein Eng. 1990; 3: 461-467Crossref PubMed Scopus (574) Google Scholar). Nucleic acid polymerases from diverse families share a common structural architecture (2Braithwaite D.K. Ito J. Nucleic Acids Res. 1993; 21: 787-802Crossref PubMed Scopus (528) Google Scholar, 3Singh K. Modak M.J. Trends Biochem. Sci. 1998; 21: 186-190Google Scholar, 4Steitz T.A. J. Biol. Chem. 1999; 274: 17395-17398Abstract Full Text Full Text PDF PubMed Scopus (690) Google Scholar). Their structural elements or subdomains have been likened to the fingers, palm, and thumb of a half-opened right hand. All polymerases perform the nucleotidyl transfer reaction using a similar catalytic reaction pathway (4Steitz T.A. J. Biol. Chem. 1999; 274: 17395-17398Abstract Full Text Full Text PDF PubMed Scopus (690) Google Scholar) and contain at least two essential carboxylates at the catalytic site, which coordinate the divalent metal ions and induce nucleophilic attack by 3′ OH of the primer onto the α-phosphate of an incoming nucleotide substrate. DNA polymerase I of Mycobacterium tuberculosis (TB pol I)1 is a member of the family A or pol I family of polymerases whose canonical family memberEscherichia coli pol I, defines the group (1Delarue M. Poch O. Tordo N. Moras D. Argos P. Protein Eng. 1990; 3: 461-467Crossref PubMed Scopus (574) Google Scholar). This prototype pol I is a DNA-dependent DNA polymerase with a modular structure and contains both polymerase and nuclease activities in a single polypeptide chain. Our interest in TB pol I stems from the parent organism, M. tuberculosis, the etiologic agent of tuberculosis. This disease kills more people around the world each year than any other infectious disease. The DNA polymerase I of this pathogen, similar to other bacterial species (pol I family) is expected to be an important participant in genome stability and DNA repair functions, especially during stationary phase of the bacterial life cycle. Therefore, the TB polymerase provides an attractive target for possible chemotherapeutic intervention. Toward this ultimate goal, we have begun to investigate functional similarities of this polymerase with those of the prototype, E. coli pol I. We have purified both the 100-kDa holoenzyme (containing 5′-3′ exonuclease, 3′-5′ exonuclease, and 3′- 5′ polymerase domains) and its 63-kDa fragment equivalent to Klenow fragment of E. coli pol I and have compared their catalytic properties. A number of similarities in these properties have been noted as expected from ∼60% sequence homology (in polymerase domain) between the two species. 2C. J. Arrigo, K. Singh, N. Kaushik, and M. J. Modak, manuscript in preparation. 2C. J. Arrigo, K. Singh, N. Kaushik, and M. J. Modak, manuscript in preparation. We were particularly interested in the substrate-binding domain of TB pol I for it is perceived as an excellent target for the development of inhibitors. The primary amino acid sequence comparison (Fig. 1) shows the conservation of catalytically important amino acids in many members of the pol I family suggesting conserved catalytic mechanism. Thus, the commonality of reaction mechanism and possible structural similarity of the catalytic segment of TB pol I with the prototype, E. coli pol I is expected to permit the translation of new findings from one polymerase to other members of this family. The crystal structures of several pol I family members in apo and binary (dNTP or template/primer bound) complexes show a highly conserved aspartate (732 of E. coli pol I) present in the three-residue long hinge region between the M and N helices of pol I (5Korolev S. Nayal M. Di Barnes W.M. Cera E. Waksman G. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9264-9268Crossref PubMed Scopus (169) Google Scholar, 6Beese L.S. Derbyshire V. Steitz T.A. Science. 1993; 260: 352-355Crossref PubMed Scopus (450) Google Scholar, 7Beese L.S. Friedman J.M. Steitz T.A. Biochemistry. 1993; 32: 14095-14101Crossref PubMed Scopus (170) Google Scholar, 8Li Y. Kong Y. Korolev S. Waksman G. Protein Sci. 1998; 7: 1116-1123Crossref PubMed Scopus (95) Google Scholar). This aspartate appears to form a salt bridge with Arg-754 (E. coli pol I position) of the O-helix. These structural observations suggest that the conserved aspartate in the hinge between M and N helices stabilizes the position of arginine in the O-helix and should complement the function of Arg-754 in pol I and its equivalent in KlenTaq (Arg-659). To clarify the functional role of this conserved aspartate, we introduced conserved and nonconserved mutations within the M to N helical hinge of TB pol I to generate D707A, D707E, and E706A enzyme species. Since, in the ternary complex of KlenTaq, Asp-637 (equivalent to Asp-707 of TB pol I) also forms an H-bond with Gln-613 (TB Gln-683 orE. coli Gln-708 equivalent), we generated the Q683A mutant and characterized its biochemical properties. In addition, a double mutant, Q683N/D707E was constructed with the expectation that a putative defect in Q683N may be compensated for by increased side chain length in D707E. The results presented in this article suggest that the conserved Asp-707 is essential for catalytic function and that its properties are very similar to those reported for Arg-754 (10Astatke M. Grindley N.D.F. Joyce C.M. J. Biol. Chem. 1995; 270: 1945-1954Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 11Kaushik N. Pandey V.N. Modak M.J. Biochemistry. 1996; 35: 7256-7266Crossref PubMed Scopus (42) Google Scholar). Furthermore, this aspartate as well as the equivalent aspartate inE. coli pol I appears to be required for the substrate binding and ternary complex formation, the latter requiring the motion of the O-helix. DISCUSSIONThe sequence alignment of family A polymerase (including pol I) members and the examination of the available crystal structures of pol I from different organisms reveals the presence of a highly conserved aspartate within the GXD sequence located on the hinge region between M and N helices (Fig. 1). In addition, the adjoining region constituted by the “O-helix” in the prototype E. coli pol I also appears to be highly conserved among all the members of this family including TB pol I (Fig. 1). The available crystal structures of pol I family members, KF (6Beese L.S. Derbyshire V. Steitz T.A. Science. 1993; 260: 352-355Crossref PubMed Scopus (450) Google Scholar), KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar), and T7 DNA polymerase (22Doublie S. Tabor S. Long A.M. Richardson C. Ellenberger T. Nature. 1998; 391: 251-258Crossref PubMed Scopus (1099) Google Scholar), show that the connector function of this three-residue region is common and that the position of the hinge region is in the close physical vicinity of the O-helix of the “fingers” subdomain. In the GXD region, both glycine and aspartate represent conserved residues. Since glycine residues generally do not participate in the reaction but rather provide local structural integrity, we chose not to investigate the role of this glycine. To understand the function of the aspartate in the GXD sequence of TB pol I, we carried out site-directed mutagenesis at Asp-707. Since Asp-707 is preceded by another carboxylate (Glu-706) and in other members of this family, Asp-707 equivalent aspartate interacts with conserved Gln, we also included Glu-706 and Gln-683 of TB pol I in the present studies. The functional significance of these residues was deduced by the examination of properties of mutant enzymes such as DNA polymerase activity, kinetics of nucleotide incorporation, dNTP and DNA binding affinity, ternary complex formation, processivity, and fidelity.The catalytic activity of the mutant TB polymerases showed that substitution of Asp-707 to Ala or Glu resulted in a significant decrease in DNA polymerase activity suggesting an important role for this residue in some catalytic function of pol I. However, no significant effect on the catalytic activity was apparent when its neighbor, Glu-706 was converted to Ala-706. Similarly, Gln-683 did not appear to be essential for the activity. Nevertheless, a double mutant D707E/Q683N was severely defective in catalytic function, probably representing the effect of D707E (see Table II). Very similar results were obtained for KF protein when its Asp-732 (equivalent of Asp-707 of TB pol I) was mutated to Ala-732. Therefore, Asp-707 of TB pol I is concluded to have a major role in the catalytic function in a manner similar to the equivalent aspartate present in the hinge region of M and N helices in other pol I family members. The kinetic analysis of D707A and other mutants clearly suggested that one of the defects with D707A, D707E, and D707E/Q673N is the decrease in the affinity for the dNTP substrate as judged by 10–30-fold increase in theKm(dNTP) (Table III) and loss of their ability to cross-link to substrate dNTP (Fig. 2). Most interestingly, the catalytic activity of these mutants was little affected when Mn2+ was substituted in place of Mg2+ as the divalent cation (Table II). These properties are reminiscent of those of O-helix muteins from E. coli pol I, namely Y766A, F762A, and R754A (10, 11). Further examination of Asp-707 muteins also showed the severe loss of pyrophosphorolysis activity (Fig. 3) and significant increase in the apparentKm(PPi), suggesting defect in the binding of PPi. The loss of binding affinity for substrate dNTP as well as for the product PPi by D707A is suggestive of the participation of Asp-707 in both of these functions. However, no change in the processivity of DNA synthesis or change in the sensitivity to dideoxynucleotides (Fig. 5) was observed with Asp-707 muteins. In addition, D707A seems to exhibit better fidelity properties compared with wild type (Fig. 4). Nevertheless, the persistent loss of catalytic activity for Asp-707 muteins, even in the presence of saturating concentrations of both template/primer and dNTP, was somewhat puzzling. We therefore, determined if the Asp-707 (Asp-732 of KF) muteins, were capable of forming a stable ternary complex. As described under “Results,” D707A of TB pol I and its E. coli counterpart D732A were unable to form stable ternary complexes (Fig. 6), even in the presence of high concentrations of substrate dNTP. Thus Asp-707 in the hinge region is concluded to be an essential component for the motion of finger subdomain (mainly O-helix) which is implicated in the ternary complex formation.From the above discussion, it is clear that most of the properties displayed by Asp-707 mutein of TB pol I are common to D732A of E. coli pol I (KF). Therefore, it seems likely that the process of dNTP binding and the ternary complex formation via the motion of the O-helix region is similar in both TB pol I and E. coli pol I. Unfortunately, the structure of TB pol I has not yet been resolved preventing us from drawing firm conclusions regarding a contribution of the hinge region aspartate in the catalytic process. However, structures of KF as well as KlenTaq have been firmly established and this has permitted us to draw valid conclusions for the role of Asp-732 in the catalytic process of pol I. Conserved D in the hinge region of M to N helices in pol I has Arg-754 of the O-helix as a major interacting partner. In the apo, DNA/dNTP-bound binary and DNA-dNTP-bound ternary complexes, a salt bridge between these two residues is consistently observed. Since the properties of muteins of either Asp-732 or Arg-754 of pol I are quite similar, we are tempted to suggest that the loss of the salt bridge between the Asp of the hinge and the Arg of the O-helix may be the major factor affecting catalytic function. Another interesting observation pertains to ∼3-fold decrease in the binding affinity for template/primer by Asp-732 muteins. This observation is suggestive of some role for Asp-732 in the binding/stabilization of template/primer. One possible scenario by which Asp-707 may indirectly participate in the process is as follows. Examination of the ternary complex structure of KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar) shows that its Asp-637 (equivalent to Asp-707 of TB pol I and Asp-732 of pol I) interacts with His-639 (equivalent to His-709 of TB pol I and His-734 of pol I). This histidine in pol I has been suggested to participate in the binding/stabilization of single-stranded template (10Astatke M. Grindley N.D.F. Joyce C.M. J. Biol. Chem. 1995; 270: 1945-1954Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). We have noted that a major factor in the stabilization of template/primer to DNA polymerases of this family is the single-stranded template overhang (to be published elsewhere). 3R. Gangurde and M. J. Modak, manuscript in preparation. Thus Asp-732 by means of its van der Waals interaction with the β carbon of His-709, may indirectly contribute to the template/primer binding process in pol I family of polymerases.Most interestingly, the loss in the ability of mutein of pol I Asp-732 to form the closed ternary complex (Fig. 6) may be further examined by comparing the position of the O-helix in the binary (E·TP) and ternary (E·TP·dNTP) complexes (Fig.7). As Arg-754 of the O-helix is seen to form the salt bridge with Asp-732 of the hinge in all complexes, what follows is the implication that the “salt bridge” structure is a responsible factor for the motion of the O-helix. Indeed this will also explain slight repositioning of the M and N helices (due to a pull exerted in the hinge region) that can be seen in the ternary complexes.Examination of the O-helix motion seen in the crystal structures of KlenTaq, as depicted in Fig. 7, also indicates that the maximal motion (∼9.4 Å) occurs at Arg-754, while the catalytically important Lys-758 moves by about 5.8 Å. Little change is observed in the positions of Phe-762 and Tyr-766. It is therefore possible, that failure to form the ternary complex may also fail to bring Lys-758 close to the active site to interact with the α-phosphate of incoming dNTP, resulting in the catalytically inactive enzyme.In summary, the properties of conserved aspartate of TB pol I corresponding to the joining region of M and N helices, suggest that it is required for dNTP binding, pyrophosphorolysis as well as ternary complex formation. Many analogous properties observed with equivalent aspartate in E. coli pol I have permitted the elucidation of a tentative structural mechanism by which this aspartate may contribute to the process of catalysis in pol I family of enzymes. The sequence alignment of family A polymerase (including pol I) members and the examination of the available crystal structures of pol I from different organisms reveals the presence of a highly conserved aspartate within the GXD sequence located on the hinge region between M and N helices (Fig. 1). In addition, the adjoining region constituted by the “O-helix” in the prototype E. coli pol I also appears to be highly conserved among all the members of this family including TB pol I (Fig. 1). The available crystal structures of pol I family members, KF (6Beese L.S. Derbyshire V. Steitz T.A. Science. 1993; 260: 352-355Crossref PubMed Scopus (450) Google Scholar), KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar), and T7 DNA polymerase (22Doublie S. Tabor S. Long A.M. Richardson C. Ellenberger T. Nature. 1998; 391: 251-258Crossref PubMed Scopus (1099) Google Scholar), show that the connector function of this three-residue region is common and that the position of the hinge region is in the close physical vicinity of the O-helix of the “fingers” subdomain. In the GXD region, both glycine and aspartate represent conserved residues. Since glycine residues generally do not participate in the reaction but rather provide local structural integrity, we chose not to investigate the role of this glycine. To understand the function of the aspartate in the GXD sequence of TB pol I, we carried out site-directed mutagenesis at Asp-707. Since Asp-707 is preceded by another carboxylate (Glu-706) and in other members of this family, Asp-707 equivalent aspartate interacts with conserved Gln, we also included Glu-706 and Gln-683 of TB pol I in the present studies. The functional significance of these residues was deduced by the examination of properties of mutant enzymes such as DNA polymerase activity, kinetics of nucleotide incorporation, dNTP and DNA binding affinity, ternary complex formation, processivity, and fidelity. The catalytic activity of the mutant TB polymerases showed that substitution of Asp-707 to Ala or Glu resulted in a significant decrease in DNA polymerase activity suggesting an important role for this residue in some catalytic function of pol I. However, no significant effect on the catalytic activity was apparent when its neighbor, Glu-706 was converted to Ala-706. Similarly, Gln-683 did not appear to be essential for the activity. Nevertheless, a double mutant D707E/Q683N was severely defective in catalytic function, probably representing the effect of D707E (see Table II). Very similar results were obtained for KF protein when its Asp-732 (equivalent of Asp-707 of TB pol I) was mutated to Ala-732. Therefore, Asp-707 of TB pol I is concluded to have a major role in the catalytic function in a manner similar to the equivalent aspartate present in the hinge region of M and N helices in other pol I family members. The kinetic analysis of D707A and other mutants clearly suggested that one of the defects with D707A, D707E, and D707E/Q673N is the decrease in the affinity for the dNTP substrate as judged by 10–30-fold increase in theKm(dNTP) (Table III) and loss of their ability to cross-link to substrate dNTP (Fig. 2). Most interestingly, the catalytic activity of these mutants was little affected when Mn2+ was substituted in place of Mg2+ as the divalent cation (Table II). These properties are reminiscent of those of O-helix muteins from E. coli pol I, namely Y766A, F762A, and R754A (10, 11). Further examination of Asp-707 muteins also showed the severe loss of pyrophosphorolysis activity (Fig. 3) and significant increase in the apparentKm(PPi), suggesting defect in the binding of PPi. The loss of binding affinity for substrate dNTP as well as for the product PPi by D707A is suggestive of the participation of Asp-707 in both of these functions. However, no change in the processivity of DNA synthesis or change in the sensitivity to dideoxynucleotides (Fig. 5) was observed with Asp-707 muteins. In addition, D707A seems to exhibit better fidelity properties compared with wild type (Fig. 4). Nevertheless, the persistent loss of catalytic activity for Asp-707 muteins, even in the presence of saturating concentrations of both template/primer and dNTP, was somewhat puzzling. We therefore, determined if the Asp-707 (Asp-732 of KF) muteins, were capable of forming a stable ternary complex. As described under “Results,” D707A of TB pol I and its E. coli counterpart D732A were unable to form stable ternary complexes (Fig. 6), even in the presence of high concentrations of substrate dNTP. Thus Asp-707 in the hinge region is concluded to be an essential component for the motion of finger subdomain (mainly O-helix) which is implicated in the ternary complex formation. From the above discussion, it is clear that most of the properties displayed by Asp-707 mutein of TB pol I are common to D732A of E. coli pol I (KF). Therefore, it seems likely that the process of dNTP binding and the ternary complex formation via the motion of the O-helix region is similar in both TB pol I and E. coli pol I. Unfortunately, the structure of TB pol I has not yet been resolved preventing us from drawing firm conclusions regarding a contribution of the hinge region aspartate in the catalytic process. However, structures of KF as well as KlenTaq have been firmly established and this has permitted us to draw valid conclusions for the role of Asp-732 in the catalytic process of pol I. Conserved D in the hinge region of M to N helices in pol I has Arg-754 of the O-helix as a major interacting partner. In the apo, DNA/dNTP-bound binary and DNA-dNTP-bound ternary complexes, a salt bridge between these two residues is consistently observed. Since the properties of muteins of either Asp-732 or Arg-754 of pol I are quite similar, we are tempted to suggest that the loss of the salt bridge between the Asp of the hinge and the Arg of the O-helix may be the major factor affecting catalytic function. Another interesting observation pertains to ∼3-fold decrease in the binding affinity for template/primer by Asp-732 muteins. This observation is suggestive of some role for Asp-732 in the binding/stabilization of template/primer. One possible scenario by which Asp-707 may indirectly participate in the process is as follows. Examination of the ternary complex structure of KlenTaq (9Li Y. Korolev S. Waksman G. EMBO J. 1998; 17: 7514-7525Crossref PubMed Scopus (653) Google Scholar) shows that its Asp-637 (equivalent to Asp-707 of TB pol I and Asp-732 of pol I) interacts with His-639 (equivalent to His-709 of TB pol I and His-734 of pol I). This histidine in pol I has been suggested to participate in the binding/stabilization of single-stranded template (10Astatke M. Grindley N.D.F. Joyce C.M. J. Biol. Chem. 1995; 270: 1945-1954Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). We have noted that a major factor in the stabilization of template/primer to DNA polymerases of this family is the single-stranded template overhang (to be published elsewhere). 3R. Gangurde and M. J. Modak, manuscript in preparation. Thus Asp-732 by means of its van der Waals interaction with the β carbon of His-709, may indirectly contribute to the template/primer binding process in pol I family of polymerases. Most interestingly, the loss in the ability of mutein of pol I Asp-732 to form the closed ternary complex (Fig. 6) may be further examined by comparing the position of the O-helix in the binary (E·TP) and ternary (E·TP·dNTP) complexes (Fig.7). As Arg-754 of the O-helix is seen to form the salt bridge with Asp-732 of the hinge in all complexes, what follows is the implication that the “salt bridge” structure is a responsible factor for the motion of the O-helix. Indeed this will also explain slight repositioning of the M and N helices (due to a pull exerted in the hinge region) that can be seen in the ternary complexes. Examination of the O-helix motion seen in the crystal structures of KlenTaq, as depicted in Fig. 7, also indicates that the maximal motion (∼9.4 Å) occurs at Arg-754, while the catalytically important Lys-758 moves by about 5.8 Å. Little change is observed in the positions of Phe-762 and Tyr-766. It is therefore possible, that failure to form the ternary complex may also fail to bring Lys-758 close to the active site to interact with the α-phosphate of incoming dNTP, resulting in the catalytically inactive enzyme. In summary, the properties of conserved aspartate of TB pol I corresponding to the joining region of M and N helices, suggest that it is required for dNTP binding, pyrophosphorolysis as well as ternary complex formation. Many analogous properties observed with equivalent aspartate in E. coli pol I have permitted the elucidation of a tentative structural mechanism by which this aspartate may contribute to the process of catalysis in pol I family of enzymes. We thank Dr. Valerie Mizrahi for the generous gift of the plasmid carrying TB pol I gene and general counsel during the early stages of the work.