Properties of the 3' to 5' exonuclease associated with calf DNA polymerase delta

The 3' to 5' exonuclease of calf thymus DNA polymerase delta has properties expected of a proofreading nuclease. It digests either single-stranded DNA or the single-stranded nucleotides of a mismatched primer on a DNA template by a nonprocessive mechanism. The distribution of oligonucleotide products suggests that a significant portion of the enzyme dissociates after the removal of one nucleotide. This mechanism is expected if the substrate in vivo is an incorrect nucleotide added by the polymerase. Digestion of single-stranded DNA does not proceed to completion, producing final products six to seven nucleotides long. Digestion of a long mismatched terminus accelerates when the mismatched region is reduced to less than six nucleotides. At the point of complementation, the digestion rate is greatly reduced. These results suggest that short mismatched regions are a preferred substrate. The use of a mismatched primer-template analogue, lacking the template single strand, greatly lowers digestion efficiency at the single-stranded 3'-terminus, suggesting that the template strand is important for substrate recognition. When oligonucleotides were examined for effectiveness as exonuclease inhibitors, (dG)8 was found to be the most potent inhibitor of single-stranded DNA digestion. (dG)8 was less effective at inhibiting digestion of mismatched primer termini, again suggesting that this DNA is a preferred substrate. Overall, these results indicate that the exonuclease of DNA polymerase delta efficiently removes short mismatched DNA, a structure formed from misincorporation during DNA synthesis.     

Interaction of Drosophila DNA polymerase alpha holoenzyme with synthetic template-primers containing mismatched primer bases or propanodeoxyguanosine adducts at various positions in template and primer regions.

We studied recognition and binding of synthetic template-primers by Drosophila DNA polymerase alpha (pol alpha) holoenzyme. The template-primers used contained either mismatched base pairs at various positions in the primer region or exocyclic propanodeoxyguanosine (PdG) adducts at various positions in both template and primer.pol alpha requires primer-terminal complementarity of greater than or equal to 4 base pairs for efficient binding and incorporation. When a mismatched base pair is at the -4 position relative to the 3'-primer terminus, minimal but detectable binding occurs. This is consistent with the ability of pol alpha to incorporate a single nucleotide on a template-primer containing a mismatch at this position, but at a rate of only 7% relative to incorporation on a perfectly matched template-primer. No binding or incorporation (less than 1% of incorporation on a perfectly matched template-primer) was evident when a mismatched base pair was at the -3 position or closer, relative to the 3'-primer terminus. Similar results were obtained when PdG was placed at various positions in the primer region. When a PdG residue was located in the template region (+ 3 position relative to the 3'-primer terminus), single-nucleotide incorporation was stimulated 3-4-fold. These observations suggest that there are intrinsic aspects to the mechanism of nucleotide incorporation by pol alpha which ensure the fidelity of DNA synthesis by this enzyme and may provide novel insights into the fundamental mechanism of polymerase translocation along templates.     

Structures of mismatch replication errors observed in a DNA polymerase

Accurate DNA replication is essential for genomic stability. One mechanism by which high-fidelity DNA polymerases maintain replication accuracy involves stalling of the polymerase in response to covalent incorporation of mismatched base pairs, thereby favoring subsequent mismatch excision. Some polymerases retain a "short-term memory" of replication errors, responding to mismatches up to four base pairs in from the primer terminus. Here we a present a structural characterization of all 12 possible mismatches captured at the growing primer terminus in the active site of a polymerase. Our observations suggest four mechanisms that lead to mismatch-induced stalling of the polymerase. Furthermore, we have observed the effects of extending a mismatch up to six base pairs from the primer terminus and find that long-range distortions in the DNA transmit the presence of the mismatch back to the enzyme active site, suggesting the structural basis for the short-term memory of replication errors.     

Deoxyribonucleic acid single-strand-specific endonucleases in human cells: partial purification of a salt-resistant endonuclease with an acidic isoelectric point

A second endonuclease with DNA single-strand specificity has been purified from KB cells, a continuous line of hunan epithelial cells. In contrast to other mammalian enzymes that cleave single-stranded DNA, this enzyme has an acidic isoelectric point (6.5 +/- 0.2). Its pH optimum is 9.5, it requires Mg2+ of Mn2+ for activity, and it has a sedimentation coefficient of 3.2 S, based on sucrose gradient centrifugation. The enzyme specifically catalyzes the endonucleolytic cleavage of synthetic DNA homopolymers and denatured viral DNA but does not attack linear duplex viral DNA. The rate of hydrolysis of poly(dT) is approximately 8-fold greater than that observed with denatured DNA. The relative rates of hydrolysis of homopolymers by the endonuclease are poly(dA) greater than poly(dT) greater than poly(dC) greater than poly(dG). Unlike other DNA single-strand-specific endonucleases isolated from human cells, this endonuclease is relatively insensitive to inhibition by KCl.     

Human cytomegalovirus. III. Virus-induced DNA polymerase.

Infection of WI-38 human fibroblasts with human cytomegalovirus (CMV) led to the stimulation of host cell DNA polymerase synthesis and induction of a novel virus-specific DNA polymerase. This cytomegalovirus-induced DNA polymerase was purified and separated from host cell enzymes by DEAE-cellulose and phosphocellulose column chromatographies. It can be distinguished from host cell enzymes by chromatographic behavior, template primer specificity, sedimentation property, and the requirement of salt for maximal activity. This virus-induced enzyme has a sedimentation coefficient of 9.2S and is found in both the nuclei and cytoplasm of virus-infected cells, but not in uninfected cells. This enzyme could efficiently use activated calf-thymus DNA, oly(dA)-oligo(dT)12-18, and poly(dC)-oligo(dG)12-18 as template primers, especially poly(dA)-oligo(dT)12-18, but it could not use poly(rA)-oligo(dT)12-18, poly(rC)-oligo(dG)12-18, or oligo(dT)12-18. The enzyme requires Mg2+ for maximal activity, is sensitive to p-hydroxymercuribenzoate, and is not a zinc metalloenzyme. In addition, the cytomegalovirus-induced DNA polymerase activity can be enhanced by adding 0.06 to 0.12 M NaCl or 0.03 to 0.06 M (NH4)2SO4 to the reaction mixture.     

Characterization of a large form of DNA polymerase delta from HeLa cells that is insensitive to proliferating cell nuclear antigen

A large form of DNA polymerase delta from HeLa cells was recently purified in this laboratory as a factor required for conservative DNA synthesis in a reconstituted system utilizing UV-irradiated permeabilized human diploid fibroblasts (Nishida, C., Reinhard, P., and Linn, S. (1988) J. Biol. Chem. 263, 501-510). We have now purified this form of the enzyme utilizing its polymerase activity and further characterized it. The enzyme activity sediments at 11.1 S in low salt and 6.8 S in high salt. In both cases, activity cosediments with the major visible peptide displayed by sodium dodecyl sulfate-polyacrylamide gels which has an Mr of 215,000. This value is consistent with the molecular mass calculated from the sedimentation coefficient and gel filtration behavior in high salt. In low salt the apparent molecular mass was approximately double. The enzyme prefers poly(dA).oligo(dT) as template/primer in low salt, with which it has a processivity of several thousand nucleotides in 1 mM MgCl2. At isotonic KCl or potassium phosphate concentrations, the preferred template/primer is activated DNA. Proliferating cell nuclear antigen, also characterized as a DNA polymerase delta auxiliary protein, does not increase the activity of this preparation of the enzyme. An antibody to the proliferating cell nuclear antigen has no inhibitory effect, nor is it able to recognize any peptides in immunoblots of purified enzyme fractions. Under polymerizing conditions, the enzyme removes mismatched, but not matched nucleotides from the 3' terminus of oligo(dT) annealed to poly(dA) suggesting a proofreading function. The properties of this form of DNA polymerase delta distinguish it from other preparations reported in the literature.     

Arabinosylnucleoside 5'-triphosphate inhibits DNA primase of calf thymus

It has been shown that DNA primase activity is tightly associated with 10S DNA polymerase alpha from calf thymus and that the ribonucleotide-dependent DNA synthesis is more sensitive to araCTP than DNA-primed DNA synthesis (Yoshida, S., et al. (1983) Biochim. Biophys. Acta 741, 348-357). Here we measured DNA primase activity using poly(dT) template or M13 bacteriophage single-stranded DNA template and primer RNA synthesis was coupled to the reaction by Escherichia coli DNA polymerase I Klenow fragment. By this method, the primer RNA synthesis can be measured independently of the associating DNA polymerase alpha. Using poly(dT) template, it was found that arabinosyladenine 5'-triphosphate (araATP) strongly inhibited DNA primase in competition with rATP. The apparent Ki for araATP was 21 microM and the ratio of Ki/Km (for rATP) was as low as 0.015. With poly(dI, dT) or M13 DNA, it was shown that araCTP also inhibited DNA primase in the similar manner. Product analysis using [alpha-32P]rATP showed that araATP inhibited the elongation of primer RNA. However, it is not likely that arabinosylnucleotides act as chain-terminators, since incubation of primer RNA with araATP did not abolish its priming activity. From these results, it is suggested that arabinosylnucleotide inhibits the initiation as well as elongation of Okazaki fragments in mammalian cells.     

Subspecies of DNA polymerase alpha from calf thymus with different fidelity in copying synthetic template-primers.

Three different subspecies of DNA polymerase alpha from calf thymus sedimenting at 9 S, 7 S and 5.7 S have been investigated with respect to their accuracy of in vitro DNA synthesis on poly (dA) (dT)16 and poly d(AT) as template-primers. Our results indicate that the structure of DNA polymerase alpha has a strong influence on the accuracy of DNA synthesis. The 9 S enzyme shows a misincorporation frequency of about 1:100 000. An error rate of 1:15 000 is measured for the 7 S species. The 5.7 S enzyme for which an error rate of 1:3 000 is determined, has to be considered as error prone. Lowering the rate of DNA synthesis leads to a decrease in fidelity. The single stranded DNA binding protein from E.coli increases the accuracy of the 5.7 S and the 7 S enzyme by a factor of two. Mn2+ decreases the fidelity of all three subspecies in a concentration dependent manner.     

Purification and characterization of the DNA polymerase and RNase H activities in Moloney murine sarcoma-leukemia virus.

Two RNase H (RNA-DNA hybrid ribonucleotidohydrolase, EC 3.1.4.34) activities separable by Sephadex G-100 gel filtration were identified in lysates of Moloney murine sarcoma-leukemia virus (MSV). The larger enzyme, which we have called RNase H-I, represented about 10% of the RNase H activity in the virion. RNase H-I (i) copurified with RNA-directed DNA polymerase from the virus, (ii) had a sedimentation coefficient of 4.4S (corresponds to an apparent mol wt of 70,000), (iii) required Mn-2+ (2 mM optimum) for activity with a [3-h]poly(A)-poly(dT) substrate, (iv) eluted from phosphocellulose at 0.2 M KC1, and (v) degraded [3-H]poly(A)-poly(dT) and [3-H]poly(C)-poly(dG) at approximately equal rates. The smaller enzyme, designated RNase H-II, which represented the majority of the RNase H activity in the virus preparation, was shown to be different since it (i) had no detectable, associated DNA polymerase activity, (ii) had a sedmimentation coefficient of 2.6S (corresponds to an apparent mol wt of 30,000), (iii) preferred Mg-2+ (10 to 15 mM optimum) over Mn-2+ (5 to 10 mM optimum) 2.5-fold for the degradation of [3-H]poly(A)-poly(dT), and (iv) degraded [3-H]poly(A)-poly(dT) 6 and 60 times faster than [3-H]poly(C)-poly(dG) in the presence of Mn-2+ and Mg-2+, respectively. Moloney MSV DNA polymerase (RNase H-I), purified by Sephadex G-100 gel filtration followed by phosphocellulose, poly(A)-oligo(dT)-cellulose, and DEAE-cellulose chromatography, transcribed heteropolymeric regions of avian myeloblastosis virus 70S RNA at a rate comparable to avian myeloblastosis virus DNA polymerase purified by the same procedure.     

Arg660Ser mutation in Thermus aquaticus DNA polymerase I suppresses T-->C transitions: implication of wobble base pair formation at the nucleotide incorporation step

We examined the replication fidelity of an Arg660Ser (R660S) mutant of Thermus aquaticus DNA polymerase I (Taq pol I). In a forward mutation assay, R660S showed a marked reduction in T→C transitions, one of the most frequent errors made by the wild-type enzyme. Steady-state kinetics showed that R660S discriminates against dGTP incorporation at a template T 13-fold better than the wild-type. R660S was also 3.2-fold less efficient than the wild-type at extending a T:dG mismatch. These results indicate that R660S has enhanced fidelity during incorporation and extension, which reduces its T→C transition frequency. Interestingly, R660S also discriminated correct from incorrect nucleotides at the incorporation step of C:dATP, A:dATP, G:dATP and C:8-OH-dGTP mispairs 28-, 6.0-, 4.1- and 6.8-fold better, respectively, than the wild-type, although it may not always be as accurate as the wild-type at the extension step. A structural model suggests that Arg660 may participate in two interactions that influence fidelity; the guanidinium group of Arg660 might interact with the incoming guanine base at the major groove and it might compete for forming another interaction with the primer terminus. Substituting Arg with Ser may eliminate or alter these interactions and destabilize the closed complex with incorrect substrates. Our data also suggest that T:dGTP and C:dATP base pairs form ‘wobble’ structures at the incorporation step of Taq pol I.     

NaF and mononucleotides as inhibitors of 3'-5'-exonuclease activity and stimulators of polymerase activity of E. coli DNA polymerase I Klenow fragment

It has been shown that, in the absence of dATP in the poly(dT).oligo(dA) template-primer complex, the rate of primer cleavage by the E. coli DNA polymerase I Klenow fragment equals 4% of polymerization rate, while in the presence of dATP it equals as much as 50-60%. NaF and NMP taken separately inhibit exonuclease cleavage of oligo(dA) both with and without dATP. The addition of NaF (5-10 mM) or NMP (5-20 mM) increases the absolute increment of polymerization rate 5-9-fold relative to the absolute decrement of the rate of nuclease hydrolysis of primer. This proves the assumption that not more than 10-20% of primer molecules, interacting with the exonuclease center of polymerase, are cleaved by the enzyme. Presumably, NaF and nucleotides disturb the coupling of the 3'-end of oligonucleotide primer to the exonuclease center of the enzyme. As the primers mostly form complexes with the polymerizing center, the reaction of polymerization is activated.     

Interactions of azidothymidine triphosphate with the cellular DNA polymerases alpha, delta, and epsilon and with DNA primase.

The interactions of azidothymidine triphosphate, the metabolically active form of the anti-AIDS drug azidothymidine (zidovudine), with the cellular DNA polymerases alpha, delta, and epsilon, as well as with the RNA primer-forming enzyme DNA primase were studied in vitro. DNA polymerase alpha was shown to incorporate azidothymidine monophosphate into a growing polynucleotide chain. This occurred 2000-fold slower than the incorporation of natural dTTP. Despite the ability of polymerase alpha to use azidothymidine triphosphate as an alternate substrate, this compound was only marginally inhibitory to the enzyme (Ki greater than 1 mM). Furthermore, the DNA primase activity associated with DNA polymerase alpha was barely inhibited by azidothymidine triphosphate (Ki greater than 1 mM). Inhibition was more pronounced for DNA polymerases delta and epsilon. The type of inhibition was competitive with respect to dTTP, with Ki values of 250 and 320 microM, respectively. No incorporation of azidothymidine monophosphate was detectable with these two DNA polymerases because their associated 3'- to 5'-exonuclease activities degraded primer molecules prior to any measurable elongation. Template-primer systems with a preformed 3'-azidothymidine-containing primer terminus inhibited the three replicative polymerases rather potently. DNA polymerase alpha was inhibited with a Ki of 150 nM and polymerases delta and epsilon with Ki values of 25 and 20 nM, respectively. The type of inhibition was competitive with respect to the unmodified substrate poly(dA).oligo(dT) for all DNA polymerases tested. Performed 3'-azidothymidine-containing primers hybridized to poly(dA) were rather resistant to degradation by the 3'- to 5'-exonuclease of DNA polymerases epsilon and more susceptible to the analogous activity that copurified with DNA polymerase delta. It is proposed that the repair of 3'-azidothymidine-containing primers might become rate-limiting for the process of DNA replication in cells that have been treated with azidothymidine triphosphate.     

Mismatch recognition and uracil excision provide complementary paths to both Ig switching and the A/T-focused phase of somatic mutation

AID-mediated deamination of dC residues within the immunoglobulin locus generates dU:dG lesions whose resolution leads to class-switch recombination and somatic hypermutation. The dU:dG pair is a mismatch and comprises a base foreign to DNA and is, thus, recognized by proteins from both base excision (uracil-DNA glycosylase, UNG) and mismatch recognition (MSH2/MSH6) pathways. Strikingly, while antibody diversification is perturbed by single deficiency in either UNG or MSH2, combined UNG/MSH2 deficiency leads to a total ablation both of switch recombination and of IgV hypermutation at dA:dT pairs. The initiating dU:dG lesions appear not to be recognized and are simply replicated over. The results indicate that the major pathway for switch recombination occurs through uracil excision with mismatch recognition of dU:dG providing a backup; the second phase of hypermutation (essentially introducing mutations solely at dA:dT pairs) is triggered by mismatch recognition of the dU:dG lesion with uracil excision providing a backup.