Significant contribution of the 3′→5′ exonuclease activity to the high fidelity of nucleotide incorporation catalyzed by human DNA polymerase ?

Most eukaryotic DNA replication is performed by A- and B-family DNA polymerases which possess a faithful polymerase activity that preferentially incorporates correct over incorrect nucleotides. Additionally, many replicative polymerases have an efficient 3′→5′ exonuclease activity that excises misincorporated nucleotides. Together, these activities contribute to overall low polymerase error frequency (one error per 10⁶–10⁸ incorporations) and support faithful eukaryotic genome replication. Eukaryotic DNA polymerase ϵ (Polϵ) is one of three main replicative DNA polymerases for nuclear genomic replication and is responsible for leading strand synthesis. Here, we employed pre-steady-state kinetic methods and determined the overall fidelity of human Polϵ (hPolϵ) by measuring the individual contributions of its polymerase and 3′→5′ exonuclease activities. The polymerase activity of hPolϵ has a high base substitution fidelity (10⁻⁴–10⁻⁷) resulting from large decreases in both nucleotide incorporation rate constants and ground-state binding affinities for incorrect relative to correct nucleotides. The 3′→5′ exonuclease activity of hPolϵ further enhances polymerization fidelity by an unprecedented 3.5 × 10² to 1.2 × 10⁴-fold. The resulting overall fidelity of hPolϵ (10⁻⁶–10⁻¹¹) justifies hPolϵ to be a primary enzyme to replicate human nuclear genome (0.1–1.0 error per round). Consistently, somatic mutations in hPolϵ, which decrease its exonuclease activity, are connected with mutator phenotypes and cancer formation.     

Substrate specificity and proposed structure of the proofreading complex of T7 DNA polymerase

Faithful replication of genomic DNA by high-fidelity DNA polymerases is crucial for the survival of most living organisms. While high-fidelity DNA polymerases favor canonical base pairs over mismatches by a factor of ∼1 × 10⁵, fidelity is further enhanced several orders of magnitude by a 3′–5′ proofreading exonuclease that selectively removes mispaired bases in the primer strand. Despite the importance of proofreading to maintaining genome stability, it remains much less studied than the fidelity mechanisms employed at the polymerase active site. Here we characterize the substrate specificity for the proofreading exonuclease of a high-fidelity DNA polymerase by investigating the proofreading kinetics on various DNA substrates. The contribution of the exonuclease to net fidelity is a function of the kinetic partitioning between extension and excision. We show that while proofreading of a terminal mismatch is efficient, proofreading a mismatch buried by one or two correct bases is even more efficient. Because the polymerase stalls after incorporation of a mismatch and after incorporation of one or two correct bases on top of a mismatch, the net contribution of the exonuclease is a function of multiple opportunities to correct mistakes. We also characterize the exonuclease stereospecificity using phosphorothioate-modified DNA, provide a homology model for the DNA primer strand in the exonuclease active site, and propose a dynamic structural model for the transfer of DNA from the polymerase to the exonuclease active site based on MD simulations.     

Pre-steady-state Kinetic Analysis of a Family D DNA Polymerase from Thermococcus sp. 9°N Reveals Mechanisms for Archaeal Genomic Replication and Maintenance

Family D DNA polymerases (polDs) have been implicated as the major replicative polymerase in archaea, excluding the Crenarchaeota branch, and bear little sequence homology to other DNA polymerase families. Here we report a detailed kinetic analysis of nucleotide incorporation and exonuclease activity for a Family D DNA polymerase from Thermococcus sp. 9°N. Pre-steady-state single-turnover nucleotide incorporation assays were performed to obtain the kinetic parameters, k_pol and K_d, for correct nucleotide incorporation, incorrect nucleotide incorporation, and ribonucleotide incorporation by exonuclease-deficient polD. Correct nucleotide incorporation kinetics revealed a relatively slow maximal rate of polymerization (k_pol ∼2.5 s⁻¹) and especially tight nucleotide binding (K_d(dNTP) ∼1.7 μm), compared with DNA polymerases from Families A, B, C, X, and Y. Furthermore, pre-steady-state nucleotide incorporation assays revealed that polD prevents the incorporation of incorrect nucleotides and ribonucleotides primarily through reduced nucleotide binding affinity. Pre-steady-state single-turnover assays on wild-type 9°N polD were used to examine 3′-5′ exonuclease hydrolysis activity in the presence of Mg²⁺ and Mn²⁺. Interestingly, substituting Mn²⁺ for Mg²⁺ accelerated hydrolysis rates >40-fold (k_exo ≥110 s⁻¹ versus ≥2.5 s⁻¹). Preference for Mn²⁺ over Mg²⁺ in exonuclease hydrolysis activity is a property unique to the polD family. The kinetic assays performed in this work provide critical insight into the mechanisms that polD employs to accurately and efficiently replicate the archaeal genome. Furthermore, despite the unique properties of polD, this work suggests that a conserved polymerase kinetic pathway is present in all known DNA polymerase families.     

Human DNA Polymerase ν Catalyzes Correct and Incorrect DNA Synthesis with High Catalytic Efficiency

DNA polymerase ν (pol ν) is a low fidelity A-family polymerase with a putative role in interstrand cross-link repair and homologous recombination. We carried out pre-steady-state kinetic analysis to elucidate the kinetic mechanism of this enzyme. We found that the mechanism consists of seven steps, similar that of other A-family polymerases. pol ν binds to DNA with a K_d for DNA of 9.2 nm, with an off-rate constant of 0.013 s⁻¹and an on-rate constant of 14 μm⁻¹ s⁻¹. dNTP binding is rapid with K_d values of 20 and 476 μm for the correct and incorrect dNTP, respectively. Pyrophosphorylation occurs with a K_d value for PP_i of 3.7 mm and a maximal rate constant of 11 s⁻¹. Pre-steady-state kinetics, examination of the elemental effect using dNTPαS, and pulse-chase experiments indicate that a rapid phosphodiester bond formation step is flanked by slow conformational changes for both correct and incorrect base pair formation. These experiments in combination with computer simulations indicate that the first conformational change occurs with rate constants of 75 and 20 s⁻¹; rapid phosphodiester bond formation occurs with a K_eq of 2.2 and 1.7, and the second conformational change occurs with rate constants of 2.1 and 0.5 s⁻¹, for correct and incorrect base pair formation, respectively. The presence of a mispair does not induce the polymerase to adopt a low catalytic conformation. pol ν catalyzes both correct and mispair formation with high catalytic efficiency.     

Characterization of Soluble Hepatitis C Virus RNA-Dependent RNA Polymerase Expressed in Escherichia coli

Production of soluble full-length nonstructural protein 5B (NS5B) of hepatitis C virus (HCV) has been shown to be problematic and requires the addition of salts, glycerol, and detergents. In an effort to improve the solubility of NS5B, the hydrophobic C terminus containing 21 amino acids was removed, yielding a truncated NS5B (NS5BΔCT) which is highly soluble and monodispersed in the absence of detergents. Fine deletional analysis of this region revealed that a four-leucine motif (LLLL) in the hydrophobic domain is responsible for the solubility profile of the full-length NS5B. Enzymatic characterization revealed that the RNA-dependent RNA polymerase (RdRp) activity of this truncated NS5B was comparable to those reported previously by others. For optimal enzyme activity, divalent manganese ions (Mn²⁺) are preferred rather than magnesium ions (Mg²⁺), whereas zinc ions (Zn²⁺) inhibit the RdRp activity. Gliotoxin, a known poliovirus 3D RdRp inhibitor, inhibited HCV NS5B RdRp in a dose-dependent manner. Kinetic analysis revealed that HCV NS5B has a rather low processivity compared to those of other known polymerases.     

T7 Protein Synthesis in F′ Episome-Containing Cells: Assignment of Specific Proteins to Three Translational Groups

Synthesis of many T7 proteins is prevented in F′ episome-containing cells. In order to quantitate the degree of inhibition, we measured the activity of several T7 proteins in extracts prepared from T7-infected F⁻ and F′ cells and cells containing F factors mutant in phage inhibition [F′(PIF⁻2A) and F′(PIF⁻2A,2B)]. In addition, we were able to assign specific T7 proteins to the three translational units previously defined by polyacrylamide gel analysis of T7 proteins made in F⁻ and episome-containing cells. After T7 infection, the presence of the wild-type F′ (PIF⁺) episome led to greater than 90% inhibition of T7 DNA polymerase (product of gene 5), T7 lysozyme (gene 3.5), and gene 10 capsid protein synthesis. Nearly normal amounts of T7 RNA polymerase (gene 1) were made in these cells. T7 infection of cells containing the mutant F′ (PIF⁻2A) episome led to normal synthesis of T7 RNA polymerase and T7 DNA polymerase; T7 lysozyme was synthesized at 30% of the maximal level in these cells; T7 gene 10 capsid protein synthesis was inhibited by 90%, and T7 DNA synthesis was arrested in these cells. T7 infection of cells containing the mutant F′ (PIF⁻2A,2B) episome led to synthesis of normal levels of the enzymes assayed.     

Temperature effect on polymerase fidelity

The discovery of extremophiles helped enable the development of groundbreaking technology such as PCR. Temperature variation is often an essential step of these technology platforms, but the effect of temperature on the error rate of polymerases from different origins is underexplored. Here, we applied high-throughput sequencing to profile the error rates of DNA polymerases from psychrophilic, mesophilic, and thermophilic origins with single-molecule resolution. We found that the reaction temperature substantially increases substitution and deletion error rates of psychrophilic and mesophilic DNA polymerases. Our motif analysis shows that the substitution error profiles cluster according to phylogenetic similarity of polymerases, not the reaction temperature, thus suggesting that the reaction temperature increases the global error rate of polymerases independent of the sequence context. Intriguingly, we also found that the DNA polymerase I of psychrophilic bacteria exhibits higher polymerization activity than its mesophilic ortholog across all temperature ranges, including down to −19 ^°C, which is well below the freezing temperature of water. Our results provide a useful reference for how the reaction temperature, a crucial parameter of biochemistry, can affect DNA polymerase fidelity in organisms adapted to a wide range of thermal environments.     

The programmed death-1 and interleukin-10 pathways play a down-modulatory role in LP-BM5 retrovirus-induced murine immunodeficiency syndrome

Pathology due to the immune system's response to viral infections often represents a delicate balance between inhibition of viral pathogenesis and regulation of protective immunity. In susceptible C57BL/6 (B6) mice, the murine retroviral isolate LP-BM5 induces splenomegaly, hypergammaglobulinemia, profound B- and T-cell immunodeficiency, and increased susceptibility to opportunistic pathogens and terminal B-cell lymphomas. Here, we report that B6.PD-1 (programmed death-1) and B6.IL-10 knockout mice are substantially more susceptible to LP-BM5-induced disease than wild-type B6 mice. LP-BM5-infected B6.PD-1^−/− mice developed more severe splenomegaly, hypergammaglobulinemia, and immunodeficiency than infected B6 mice: PD-1^−/− mice are more susceptible to lower doses of LP-BM5 and show more exaggerated disease early postinfection. LP-BM5-infected B6.IL-10^−/− mice also develop exaggerated LP-BM5-induced disease, compared to B6 mice, without a significant change in the retroviral load. By reciprocal reconstitution experiments, comparing wild-type versus PD-1^−/− sources of the requisite cells for LP-BM5 pathogenesis—CD4 T and B cells, PD-1⁺ B cells appear to be crucial in the normal limitation of LP-BM5-induced disease in B6 mice. Also, infected B6 mice have increased CD11b⁺ spleen cells that express interleukin-10 (IL-10). However, PD-1^−/− mice, though showing an even greater expansion of CD11b⁺ cells after LP-BM5 inoculation, did not show an equivalent increase in IL-10-producing cells. Thus, it appears that PD-1/PD-L interactions and IL-10 are primarily important in moderating the effects of LP-BM5-induced disease in B6 mice.     

Differential and Site Specific Impact of B Cells in the Protective Immune Response to Mycobacterium tuberculosis in the Mouse

Cell-mediated immune responses are known to be critical for control of mycobacterial infections whereas the role of B cells and humoral immunity is unclear. B cells can modulate immune responses by secretion of immunoglobulin, production of cytokines and antigen-presentation. To define the impact of B cells in the absence of secreted immunoglobulin, we analyzed the progression of Mycobacterium tuberculosis (Mtb) infection in mice that have B cells but which lack secretory immunoglobulin (AID^−/−µS^−/−mice). AID^−/−µS^−/− mice accumulated a population of activated B cells in the lungs when infected and were more susceptible to aerosol Mtb when compared to wild type (C57BL/6) mice or indeed mice that totally lack B cells. The enhanced susceptibility of AID^−/−µS^−/− mice was not associated with defective T cell activation or expression of a type 1 immune response. While delivery of normal serum to AID^−/−µS^−/− mice did not reverse susceptibility, susceptibility in the spleen was dependent upon the presence of B cells and susceptibility in the lungs of AID^−/−µS^−/−mice was associated with elevated expression of the cytokines IL-6, GM-CSF, IL-10 and molecules made by alternatively activated macrophages. Blocking of IL-10 signaling resulted in reversal of susceptibility in the spleens and lungs of AID^−/−µS^−/− mice. These data support the hypothesis that B cells can modulate immunity to Mtb in an organ specific manner via the modulation of cytokine production and macrophage activation.     

Kinetic investigation of the polymerase and exonuclease activities of human DNA polymerase ε holoenzyme

In eukaryotic DNA replication, DNA polymerase ε (Polε) is responsible for leading strand synthesis, whereas DNA polymerases α and δ synthesize the lagging strand. The human Polε (hPolε) holoenzyme is comprised of the catalytic p261 subunit and the noncatalytic p59, p17, and p12 small subunits. So far, the contribution of the noncatalytic subunits to hPolε function is not well understood. Using pre-steady-state kinetic methods, we established a minimal kinetic mechanism for DNA polymerization and editing catalyzed by the hPolε holoenzyme. Compared with the 140-kDa N-terminal catalytic fragment of p261 (p261N), which we kinetically characterized in our earlier studies, the presence of the p261 C-terminal domain (p261C) and the three small subunits increased the DNA binding affinity and the base substitution fidelity. Although the small subunits enhanced correct nucleotide incorporation efficiency, there was a wide range of rate constants when incorporating a correct nucleotide over a single-base mismatch. Surprisingly, the 3′→5′ exonuclease activity of the hPolε holoenzyme was significantly slower than that of p261N when editing both matched and mismatched DNA substrates. This suggests that the presence of p261C and the three small subunits regulates the 3′→5′ exonuclease activity of the hPolε holoenzyme. Together, the 3′→5′ exonuclease activity and the variable mismatch extension activity modulate the overall fidelity of the hPolε holoenzyme by up to 3 orders of magnitude. Thus, the presence of p261C and the three noncatalytic subunits optimizes the dual enzymatic activities of the catalytic p261 subunit and makes the hPolε holoenzyme an efficient and faithful replicative DNA polymerase.     

Nucleostemin delays cellular senescence and negatively regulates TRF1 protein stability

Nucleostemin (NS) encodes a nucleolar GTP-binding protein highly enriched in the stem cells and cancer cells. To determine its biological activity in vivo, we generated NS loss- and gain-of-function mouse models. The embryogenesis of homozygous NS-null (NS^−/−) mice was aborted before the blastula stage. Although the growth and fertility of heterozygous NS-null (NS^+/−) mice appeared normal, NS^+/− mouse embryonic fibroblasts (MEFs) had fewer NS proteins, a lower population growth rate, and higher percentages of senescent cells from passage 5 (P5) to P7 than their wild-type littermates. Conversely, transgenic overexpression of NS could rescue the NS^−/− embryo in a dose-dependent manner, increase the population growth rate, and reduce the senescent percentage of MEFs. Cell cycle analyses revealed increased pre-G₁ percentages in the late-passage NS^+/− MEF cultures compared to the wild-type cultures. We demonstrated that NS could interact with telomeric repeat-binding factor 1 (TRF1) and enhance the degradation but not the ubiquitination of the TRF1 protein, which negatively regulates telomere length and is essential for early embryogenesis. This work demonstrates the roles of NS in establishing early embryogenesis and delaying cellular senescence of MEFs and reveals a mechanism of a NS-regulated degradation of TRF1.     

Switching between polymerase and exonuclease sites in DNA polymerase ε

The balance between exonuclease and polymerase activities promotes DNA synthesis over degradation when nucleotides are correctly added to the new strand by replicative B-family polymerases. Misincorporations shift the balance toward the exonuclease site, and the balance tips back in favor of DNA synthesis when the incorrect nucleotides have been removed. Most B-family DNA polymerases have an extended β-hairpin loop that appears to be important for switching from the exonuclease site to the polymerase site, a process that affects fidelity of the DNA polymerase. Here, we show that DNA polymerase ε can switch between the polymerase site and exonuclease site in a processive manner despite the absence of an extended β-hairpin loop. K967 and R988 are two conserved amino acids in the palm and thumb domain that interact with bases on the primer strand in the minor groove at positions n−2 and n−4/n−5, respectively. DNA polymerase ε depends on both K967 and R988 to stabilize the 3′-terminus of the DNA within the polymerase site and on R988 to processively switch between the exonuclease and polymerase sites. Based on a structural alignment with DNA polymerase δ, we propose that arginines corresponding to R988 might have a similar function in other B-family polymerases.     

Kinetic mechanism of nick sealing by T4 RNA ligase 2 and effects of 3′-OH base mispairs and damaged base lesions

T4 RNA ligase 2 (Rnl2) repairs 3′-OH/5′-PO₄ nicks in duplex nucleic acids in which the broken 3′-OH strand is RNA. Ligation entails three chemical steps: reaction of Rnl2 with ATP to form a covalent Rnl2–(lysyl-Nζ)–AMP intermediate (step 1); transfer of AMP to the 5′-PO₄ of the nick to form an activated AppN– intermediate (step 2); and attack by the nick 3′-OH on the AppN– strand to form a 3′–5′ phosphodiester (step 3). Here we used rapid mix-quench methods to analyze the kinetic mechanism and fidelity of single-turnover nick sealing by Rnl2–AMP. For substrates with correctly base-paired 3′-OH nick termini, k_step2 was fast (9.5 to 17.9 sec⁻¹) and similar in magnitude to k_step3 (7.9 to 32 sec⁻¹). Rnl2 fidelity was enforced mainly at the level of step 2 catalysis, whereby 3′-OH base mispairs and oxoguanine, oxoadenine, or abasic lesions opposite the nick 3′-OH elicited severe decrements in the rate of 5′-adenylylation and relatively modest slowing of the rate of phosphodiester synthesis. The exception was the noncanonical A:oxoG base pair, which Rnl2 accepted as a correctly paired end for rapid sealing. These results underscore (1) how Rnl2 requires proper positioning of the 3′-terminal ribonucleoside at the nick for optimal 5′-adenylylation and (2) the potential for nick-sealing ligases to embed mutations during the repair of oxidative damage.     

Hepatitis C Virus NS5B and Host Cyclophilin A Share a Common Binding Site on NS5A

Nonstructural protein 5B (NS5B) is essential for hepatitis C virus (HCV) replication as it carries the viral RNA-dependent RNA polymerase enzymatic activity. HCV replication occurs in a membrane-associated multiprotein complex in which HCV NS5A and host cyclophilin A (CypA) have been shown to be present together with the viral polymerase. We used NMR spectroscopy to perform a per residue level characterization of the molecular interactions between the unfolded domains 2 and 3 of NS5A (NS5A-D2 and NS5A-D3), CypA, and NS5B_Δ21. We show that three regions of NS5A-D2 (residues 250–262 (region A), 274–287 (region B), and 306–333 (region C)) interact with NS5B_Δ21, whereas NS5A-D3 does not. We show that both NS5B_Δ21 and CypA share a common binding site on NS5A that contains residues Pro-306 to Glu-323. No direct molecular interaction has been detected by NMR spectroscopy between HCV NS5B_Δ21 and host CypA. We show that cyclosporine A added to a sample containing NS5B_Δ21, NS5A-D2, and CypA specifically inhibits the interaction between CypA and NS5A-D2 without altering the one between NS5A-D2 and NS5B_Δ21. A high quality heteronuclear NMR spectrum of HCV NS5B_Δ21 has been obtained and was used to characterize the binding site on the polymerase of NS5A-D2. Moreover these data highlight the potential of using NMR of NS5B_Δ21 as a powerful tool to characterize in solution the interactions of the HCV polymerase with all kinds of molecules (proteins, inhibitors, RNA). This work brings new insights into the comprehension of the molecular interplay between NS5B, NS5A, and CypA, three essentials proteins for HCV replication.     

A Single Mutation in Human Mitochondrial DNA Polymerase Pol γA Affects Both Polymerization and Proofreading Activities of Only the Holoenzyme

Common causes of human mitochondrial diseases are mutations affecting DNA polymerase (Pol) γ, the sole polymerase responsible for DNA synthesis in mitochondria. Although the polymerase and exonuclease active sites are located on the catalytic subunit Pol γA, in holoenzyme both activities are regulated by the accessory subunit Pol γB. Several patients with severe neurological and muscular disorders were reported to carry the Pol γA substitutions R232G or R232H, which lie outside of either active site. We report that Arg²³² substitutions have no effect on independent Pol γA activities but show major defects in the Pol γA-Pol γB holoenzyme, including decreased polymerase and increased exonuclease activities, the latter with decreased selectivity for mismatches. We show that Pol γB facilitates distinguishing mismatched from base-paired primer termini and that Pol γA Arg²³² is essential for mediating this regulatory function of the accessory subunit. This study provides a molecular basis for the disease symptoms exhibited by patients carrying those substitutions.     

Variation in RNA Virus Mutation Rates across Host Cells

It is well established that RNA viruses exhibit higher rates of spontaneous mutation than DNA viruses and microorganisms. However, their mutation rates vary amply, from 10⁻⁶ to 10⁻⁴ substitutions per nucleotide per round of copying (s/n/r) and the causes of this variability remain poorly understood. In addition to differences in intrinsic fidelity or error correction capability, viral mutation rates may be dependent on host factors. Here, we assessed the effect of the cellular environment on the rate of spontaneous mutation of the vesicular stomatitis virus (VSV), which has a broad host range and cell tropism. Luria-Delbrück fluctuation tests and sequencing showed that VSV mutated similarly in baby hamster kidney, murine embryonic fibroblasts, colon cancer, and neuroblastoma cells (approx. 10⁻⁵ s/n/r). Cell immortalization through p53 inactivation and oxygen levels (1–21%) did not have a significant impact on viral replication fidelity. This shows that previously published mutation rates can be considered reliable despite being based on a narrow and artificial set of laboratory conditions. Interestingly, we also found that VSV mutated approximately four times more slowly in various insect cells compared with mammalian cells. This may contribute to explaining the relatively slow evolution of VSV and other arthropod-borne viruses in nature.     

Spin-label Studies of Membranes in Rye Protoplasts during Extracellular Freezing

Protoplasts isolated from epicotyls of nonhardened winter rye seedlings were spin-labeled with the N-oxyl-4-4-dimethyloxazolidine derivatives of 5-ketostearic (5NS) and 16-ketostearic (16NS) acids. Spectra of the membrane-bound labels showed motional broadening with a rotational correlation time of 1.5 × 10⁻⁸ second for 5NS and 1.5 × 10⁻¹⁰ second for 16NS at 0 C. A procedure was developed to follow membrane changes in these protoplasts during extracellular freezing. With freezing, molecular motion of 5NS, but not of 16NS, spin probes was restricted. The increase in molecular order near the hydrated end of the membrane did not result from lowered temperatures inasmuch as no such change was observed in supercooled samples. These changes are probably due to dehydration of protoplast membranes during extracellular freezing. Similar results were obtained with multilayered egg yolk lecithin and are consistent with previous observations of changes in lecithin multibilayers during dehydration. Such alterations in membrane order might lead to irreversible membrane damage during extracellular freezing of plant cells.     

The fidelity of DNA synthesis by yeast DNA polymerase zeta alone and with accessory proteins

DNA polymerase zeta (pol ζ) participates in several DNA transactions in eukaryotic cells that increase spontaneous and damage-induced mutagenesis. To better understand this central role in mutagenesis in vivo, here we report the fidelity of DNA synthesis in vitro by yeast pol ζ alone and with RFC, PCNA and RPA. Overall, the accessory proteins have little effect on the fidelity of pol ζ. Pol ζ is relatively accurate for single base insertion/deletion errors. However, the average base substitution fidelity of pol ζ is substantially lower than that of homologous B family pols α, δ and . Pol ζ is particularly error prone for substitutions in specific sequence contexts and generates multiple single base errors clustered in short patches at a rate that is unprecedented in comparison with other polymerases. The unique error specificity of pol ζ in vitro is consistent with Pol ζ-dependent mutagenic specificity reported in vivo. This fact, combined with the high rate of single base substitution errors and complex mutations observed here, indicates that pol ζ contributes to mutagenesis in vivo not only by extending mismatches made by other polymerases, but also by directly generating its own mismatches and then extending them.