Polymerization activity of an alpha-like DNA polymerase requires a conserved 3''-5'' exonuclease active site.

Most DNA polymerases are multifunctional proteins that possess both polymerizing and exonucleolytic activities. For Escherichia coli DNA polymerase I and its relatives, polymerase and exonuclease activities reside on distinct, separable domains of the same polypeptide. The catalytic subunits of the alpha-like DNA polymerase family share regions of sequence homology with the 3'-5' exonuclease active site of DNA polymerase I; in certain alpha-like DNA polymerases, these regions of homology have been shown to be important for exonuclease activity. This finding has led to the hypothesis that alpha-like DNA polymerases also contain a distinct 3'-5' exonuclease domain. We have introduced conservative substitutions into a 3'-5' exonuclease active site homology in the gene encoding herpes simplex virus DNA polymerase, an alpha-like polymerase. Two mutants were severely impaired for viral DNA replication and polymerase activity. The mutants were not detectably affected in the ability of the polymerase to interact with its accessory protein, UL42, or to colocalize in infected cell nuclei with the major viral DNA-binding protein, ICP8, suggesting that the mutation did not exert global effects on protein folding. The results raise the possibility that there is a fundamental difference between alpha-like DNA polymerases and E. coli DNA polymerase I, with less distinction between 3'-5' exonuclease and polymerase functions in alpha-like DNA polymerases.     

Crystal structure of an archaebacterial DNA polymerase.

BACKGROUND: Members of the Pol II family of DNA polymerases are responsible for chromosomal replication in eukaryotes, and carry out highly processive DNA replication when attached to ring-shaped processivity clamps. The sequences of Pol II polymerases are distinct from those of members of the well-studied Pol I family of DNA polymerases. The DNA polymerase from the archaebacterium Desulfurococcus strain Tok (D. Tok Pol) is a member of the Pol II family that retains catalytic activity at elevated temperatures. RESULTS: The crystal structure of D. Tok Pol has been determined at 2.4 A resolution. The architecture of this Pol II type DNA polymerase resembles that of the DNA polymerase from the bacteriophage RB69, with which it shares less than approximately 20% sequence identity. As in RB69, the central catalytic region of the DNA polymerase is located within the 'palm' subdomain and is strikingly similar in structure to the corresponding regions of Pol I type DNA polymerases. The structural scaffold that surrounds the catalytic core in D. Tok Pol is unrelated in structure to that of Pol I type polymerases. The 3'-5' proofreading exonuclease domain of D. Tok Pol resembles the corresponding domains of RB69 Pol and Pol I type DNA polymerases. The exonuclease domain in D. Tok Pol is located in the same position relative to the polymerase domain as seen in RB69, and on the opposite side of the palm subdomain compared to its location in Pol I type polymerases. The N-terminal domain of D. Tok Pol has structural similarity to RNA-binding domains. Sequence alignments suggest that this domain is conserved in the eukaryotic DNA polymerases delta and epsilon. CONCLUSIONS: The structure of D. Tok Pol confirms that the modes of binding of the template and extrusion of newly synthesized duplex DNA are likely to be similar in both Pol II and Pol I type DNA polymerases. However, the mechanism by which the newly synthesized product transits in and out of the proofreading exonuclease domain has to be quite different. The discovery of a domain that seems to be an RNA-binding module raises the possibility that Pol II family members interact with RNA.     

Characterization of nuclease-dependent functions of Exo1p in Saccharomyces cerevisiae

Exo1p is a member of the Rad2p family of structure-specific nucleases that contain conserved N and I nuclease domains. Exo1p has been implicated in numerous DNA metabolic processes, such as recombination, double-strand break repair and DNA mismatch repair (MMR). In this report, we describe in vitro and in vivo characterization of full-length wild-type and mutant forms of Exo1p. Herein, we demonstrate that full-length yeast Exo1p possesses an intrinsic 5'-3' exonuclease activity as reported previously, but also possesses a flap-endonuclease activity. Our study indicates that Exo1p shares similar, but not identical structure-function relationships to other characterized members of the Rad2p family in the N and I nuclease domains. The two exo1p mutants we examined, showed deficiencies for both double-stranded DNA (dsDNA) 5'-3' exonuclease and flap-endonuclease activities. Examining the genetic interaction of these two exo1 mutations with rad27Delta suggest that the Exo1p flap-endonuclease activity and not the dsDNA 5'-3' exonuclease is redundant to Rad27p for viability. In addition, our in vivo results also indicate that many exo1Delta phenotypes are dependent on the complete catalytic activities of Exo1p. Finally, our findings plus those of other investigators suggest that Exo1p functions both in a catalytic and a structural capacity during DNA MMR.     

Structural mechanism for coordination of proofreading and polymerase activities in archaeal DNA polymerases

A novel mechanism for controlling the proofreading and polymerase activities of archaeal DNA polymerases was studied. The 3'-5'exonuclease (proofreading) activity and PCR performance of the family B DNA polymerase from Thermococcus kodakaraensis KOD1 (previously Pyrococcus kodakaraensis KOD1) were altered efficiently by mutation of a "unique loop" in the exonuclease domain. Interestingly, eight different H147 mutants showed considerable variations in respect to their 3'-5'exonuclease activity, from 9% to 276%, as against that of the wild-type (WT) enzyme. We determined the 2.75A crystal structure of the H147E mutant of KOD DNA polymerase that shows 30% of the 3'-5'exonuclease activity, excellent PCR performance and WT-like fidelity. The structural data indicate that the properties of the H147E mutant were altered by a conformational change of the Editing-cleft caused by an interaction between the unique loop and the Thumb domain. Our data suggest that electrostatic and hydrophobic interactions between the unique loop of the exonuclease domain and the tip of the Thumb domain are essential for determining the properties of these DNA polymerases.     

A conserved insertion in protein-primed DNA polymerases is involved in primer terminus stabilisation

Protein-primed DNA polymerases form a subgroup of the eukaryotic-type DNA polymerases family, also called family B or alpha-like. A multiple amino acid sequence alignment of this subgroup of DNA polymerases led to the identification of two insertions, TPR-1 and TPR-2, in the polymerisation domain. We showed previously that Asp332 of the TPR-1 insertion of phi29 DNA polymerase is involved in the correct orientation of the terminal protein (TP) for the initiation of replication. In this work, the functional role of two other conserved residues from TPR-1, Lys305 and Tyr315, has been analysed. The four mutant derivatives constructed, K305I, K305R, Y315A and Y315F, displayed a wild-type 3'-5' exonuclease activity on single-stranded DNA. However, when assayed on double-stranded DNA such activity was higher than that of the wild-type enzyme. This activity led to a reduced pol/exo ratio, suggesting a defect in stabilising the primer terminus at the polymerase active site. On the other hand, although mutant polymerases K305I and Y315A were able to couple processive DNA polymerisation to strand displacement, they were severely impaired in phi29 TP-DNA replication. The possible role of the TPR-1 insertion in the set of interactions with the nascent chain during the first steps of TP-DNA replication is discussed.     

Effects of mutations in the Exo III motif of the herpes simplex virus DNA polymerase gene on enzyme activities, viral replication, and replication fidelity.

The herpes simplex virus DNA polymerase catalytic subunit, which has intrinsic polymerase and 3'-5' exonuclease activities, contains sequence motifs that are homologous to those important for 3'-5' exonuclease activity in other polymerases. The role of one such motif, Exo III, was examined in this study. Mutated polymerases containing either a single tyrosine-to-histidine change at residue 577 or this change plus an aspartic acid-to-alanine at residue 581 in the Exo III motif exhibited defective or undetectable exonuclease activity, respectively, yet retained substantial polymerase activity. Despite the defects in exonuclease activity, the mutant polymerases were able to support viral replication in transient complementation assays, albeit inefficiently. Viruses replicated via the action of these mutant polymerases exhibited substantially increased frequencies of mutants resistant to ganciclovir. Furthermore, when the Exo III mutations were incorporated into the viral genome, the resulting mutant viruses displayed only modestly defect in replication in Vero cells and exhibited substantially increased mutation frequencies. The results suggest that herpes simplex virus can replicate despite severely impaired exonuclease activity and that the 3'-5' exonuclease contributes substantially to the fidelity of viral DNA replication.     

Different Behaviors In Vivo of Mutations in the β Hairpin Loop of the DNA Polymerases of the Closely Related Phages T4 and RB69

The T4 and RB69 DNA replicative polymerases are members of the B family and are highly similar. Both replicate DNA with high fidelity and employ the same mechanism that allows efficient switching of the primer terminus between the polymerase and exonuclease sites. Both polymerases have a β hairpin loop (hereafter called the β loop) in their exonuclease domains that plays an important role in active-site switching. The β loop is involved in strand separation and is needed to stabilize partially strand-separated exonuclease complexes. In T4 DNA polymerase, modification of the β-loop residue G255 to Ser confers a strong mutator phenotype in vivo due to a reduced ability to form editing complexes. Here, we describe the RB69 DNA polymerase mutant with the equivalent residue (G258) changed to Ser but showing only mild mutator activity in vivo. On the other hand, deletion of the tip of the RB69 β loop confers a strong mutator phenotype in vivo. Based on detailed mutational spectral analyses, DNA binding activities, and coupled polymerase/exonuclease assays, we define the differences between the T4 and RB69 polymerases. We propose that their β loops facilitate strand separation in both polymerases, while the residues that form the loop have low structural constraints.     

Crystal structure of a pol alpha family DNA polymerase from the hyperthermophilic archaeon Thermococcus sp. 9 degrees N-7

The 2.25 A resolution crystal structure of a pol alpha family (family B) DNA polymerase from the hyperthermophilic marine archaeon Thermococcus sp. 9 degrees N-7 (9 degrees N-7 pol) provides new insight into the mechanism of pol alpha family polymerases that include essentially all of the eukaryotic replicative and viral DNA polymerases. The structure is folded into NH(2)- terminal, editing 3'-5' exonuclease, and polymerase domains that are topologically similar to the two other known pol alpha family structures (bacteriophage RB69 and the recently determined Thermococcus gorgonarius), but differ in their relative orientation and conformation.The 9 degrees N-7 polymerase domain structure is reminiscent of the "closed" conformation characteristic of ternary complexes of the pol I polymerase family obtained in the presence of their dNTP and DNA substrates. In the apo-9 degrees N-7 structure, this conformation appears to be stabilized by an ion pair. Thus far, the other apo-pol alpha structures that have been determined adopt open conformations. These results therefore suggest that the pol alpha polymerases undergo a series of conformational transitions during the catalytic cycle similar to those proposed for the pol I family. Furthermore, comparison of the orientations of the fingers and exonuclease (sub)domains relative to the palm subdomain that contains the pol active site suggests that the exonuclease domain and the fingers subdomain of the polymerase can move as a unit and may do so as part of the catalytic cycle. This provides a possible structural explanation for the interdependence of polymerization and editing exonuclease activities unique to pol alpha family polymerases.We suggest that the NH(2)-terminal domain of 9 degrees N-7 pol may be structurally related to an RNA-binding motif, which appears to be conserved among archaeal polymerases. The presence of such a putative RNA- binding domain suggests a mechanism for the observed autoregulation of bacteriophage T4 DNA polymerase synthesis by binding to its own mRNA. Furthermore, conservation of this domain could indicate that such regulation of pol expression may be a characteristic of archaea. Comparion of the 9 degrees N-7 pol structure to its mesostable homolog from bacteriophage RB69 suggests that thermostability is achieved by shortening loops, forming two disulfide bridges, and increasing electrostatic interactions at subdomain interfaces.     

The multiple biological roles of the 3'-->5' exonuclease of Saccharomyces cerevisiae DNA polymerase delta require switching between the polymerase and exonuclease domains

Until recently, the only biological function attributed to the 3'-->5' exonuclease activity of DNA polymerases was proofreading of replication errors. Based on genetic and biochemical analysis of the 3'-->5' exonuclease of yeast DNA polymerase delta (Pol delta) we have discerned additional biological roles for this exonuclease in Okazaki fragment maturation and mismatch repair. We asked whether Pol delta exonuclease performs all these biological functions in association with the replicative complex or as an exonuclease separate from the replicating holoenzyme. We have identified yeast Pol delta mutants at Leu523 that are defective in processive DNA synthesis when the rate of misincorporation is high because of a deoxynucleoside triphosphate (dNTP) imbalance. Yet the mutants retain robust 3'-->5' exonuclease activity. Based on biochemical studies, the mutant enzymes appear to be impaired in switching of the nascent 3' end between the polymerase and the exonuclease sites, resulting in severely impaired biological functions. Mutation rates and spectra and synergistic interactions of the pol3-L523X mutations with msh2, exo1, and rad27/fen1 defects were indistinguishable from those observed with previously studied exonuclease-defective mutants of the Pol delta. We conclude that the three biological functions of the 3'-->5' exonuclease addressed in this study are performed intramolecularly within the replicating holoenzyme.     

Yeast mitochondrial DNA mutators with deficient proofreading exonucleolytic activity.

The MIP1 gene which encodes yeast mitochondrial DNA polymerase possesses in its N-terminal region the three motifs (Exo1, Exo2 and Exo3) which characterize the 3'-5' exonucleolytic domain of many DNA polymerases. By site directed mutagenesis we have substituted alanine or glycine residues for conserved aspartate residues in each consensus sequence. Yeast mutants were therefore generated that are capable of replicating mitochondrial DNA (mtDNA) and exhibit a mutator phenotype, as estimated by the several hundred-fold increase in the frequency of spontaneous mitochondrial erythromycin resistant mutants. By overexpressing the mtDNA polymerase from the GAL1 promoter as a major 140 kDa polypeptide, we showed that the wild-type enzyme possesses a mismatch-specific 3'-5' exonuclease activity. This activity was decreased by approximately 500-fold in the mutant D347A; in contrast, the extent of DNA synthesis was only slightly decreased. The wild-type mtDNA polymerase efficiently catalyses elongation of singly-primed M13 DNA to the full-length product. However, the mutant preferentially accumulates low molecular weight products. These data were extended to the two other mutators D171G and D230A. Glycine substitution for the Cys344 residue which is present in the Exo3 site of several polymerases generates a mutant with a slightly higher mtDNA mutation rate and a slightly lower 3'-5' exonucleolytic activity. We conclude that proofreading is an important determinant of accuracy in the replication of yeast mtDNA.     

Sequence analysis of three family B DNA polymerases from the thermoacidophilic crenarchaeon Sulfurisphaera ohwakuensis.

Three family B DNA polymerase genes, designated B1, B2, and B3, were cloned from the thermoacidophilic crenarchaeon Sulfurisphaera ohwakuensis, and sequenced. Deduced amino acid sequences of B1 and B3 DNA polymerases have all exonuclease and polymerase motifs which include critical residues for catalytic activities. Furthermore, a YxGG/A motif, which is located between 3'-5' exonuclease and polymerization domains of family B DNA polymerases, was also found in each of the B1 and B3 sequences. These findings suggested that S. ohwakuensis B1 and B3 DNA polymerases have both exonuclease and polymerase activities. However, amino acid sequence of the B2 DNA polymerase of this organism contains several amino acid substitutions in Pol-motifs, and also lacks Exo-motif I and Exo-motif II. These substitutions and lack of certain motifs raise questions about polymerase and exonuclease activities of the corresponding gene product. The B3 sequence of S. ohwakuensis is more closely related to Pyrodictium, Aeropyrum, and Archaeoglobus DNA polymerase B3 sequences than to the Sulfolobus B3 sequences. Phylogenetic analysis showed that crenarchaeal B1 DNA polymerases are closely related to each other, and suggested that crenarchaeal B3, euryarchaeal family B, and eukaryal epsilon DNA polymerases may be orthologs.     

Cloning and characterization of a family B DNA polymerase from the hyperthermophilic crenarchaeon Pyrobaculum islandicum

In order to extend the limited knowledge about crenarchaeal DNA polymerases, we cloned a gene encoding a family B DNA polymerase from the hyperthermophilic crenarchaeon Pyrobaculum islandicum. The enzyme shared highest sequence identities with a group of phylogenetically related DNA polymerases, designated B3 DNA polymerases, from members of the kingdom Crenarchaeota, Pyrodictium occultum and Aeropyrum pernix, and several members of the kingdom Euryarchaeota. Six highly conserved regions as well as a DNA-binding motif, indicative of family B DNA polymerases, were identified within the sequence. Furthermore, three highly conserved 3'-5' exonuclease motifs were also found. The gene was expressed in Escherichia coli, and the DNA polymerase was purified to homogeneity by heat treatment and affinity chromatography. Activity staining after sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed an active polypeptide of approximately 90 kDa. For the recombinant DNA polymerase from P. islandicum, activated calf thymus DNA was used as a substrate rather than primed single-stranded DNA. The enzyme was strongly inhibited by monovalent cations and N-ethylmaleimide; it is moderately sensitive to aphidicolin and dideoxyribonucleoside triphosphates. The half-life of the enzyme at 100 and 90 degrees C was 35 min and >5 h, respectively. Interestingly, the pH of the assay buffer had a significant influence on the 3'-5' exonuclease activity of the recombinant enzyme. Under suitable assay conditions for PCR, the enzyme was able to amplify lambda DNA fragments of up to 1,500 bp.     

Structural and biochemical investigation of the role in proofreading of a beta hairpin loop found in the exonuclease domain of a replicative DNA polymerase of the B family

Replicative DNA polymerases, as exemplified by the B family polymerases from bacteriophages T4 and RB69, not only replicate DNA but also have the ability to proofread misincorporated nucleotides. Because the two activities reside in separate protein domains, polymerases must employ a mechanism that allows for efficient switching of the primer strand between the two active sites to achieve fast and accurate replication. Prior mutational and structural studies suggested that a beta hairpin structure located in the exonuclease domain of family B polymerases might play an important role in active site switching in the event of a nucleotide misincorporation. We show that deleting the beta hairpin loop in RB69 gp43 affects neither polymerase nor exonuclease activities. Single binding event studies with mismatched primer termini, however, show that the beta hairpin plays a role in maintaining the stability of the polymerase/DNA interactions during the binding of the primer DNA in the exonuclease active site but not on the return of the corrected primer to the polymerase active site. In addition, the deletion variant showed a more stable incorporation of a nucleotide opposite an abasic site. Moreover, in the 2.4 A crystal structure of the beta hairpin deletion variant incorporating an A opposite a templating furan, all four molecules in the crystal asymmetric unit have DNA in the polymerase active site, despite the presence of DNA distortions because of the misincorporation, confirming that the primer strand is not stably bound within the exonuclease active site in the absence of the beta hairpin loop.     

Kinetics of error generation in homologous B-family DNA polymerases

The kinetics of forming a proper Watson–Crick base pair as well incorporating bases opposite furan, an abasic site analog, have been well characterized for the B Family replicative DNA polymerase from bacteriophage T4. Structural studies of these reactions, however, have only been performed with the homologous enzyme from bacteriophage RB69. In this work, the homologous enzymes from RB69 and T4 were compared in parallel reactions to determine the relative abilities of the two polymerases to incorporate correct nucleotides as well as to form improper pairings. The kinetic rates for three different exonuclease mutants for each enzyme were measured for incorporation of an A opposite T and an A opposite furan as well as for the formation of A:C and T:T mismatches. The T4 exonuclease mutants were all ~2- to 7-fold more efficient than the corresponding RB69 exonuclease mutants depending on whether a T or furan was in the templating position and which exonuclease mutant was used. The rates for mismatch formation by T4 were significantly reduced compared with incorporation opposite furan, much more so than the corresponding RB69 mutant. These results show that there are kinetic differences between the two enzymes but they are not large enough to preclude structural assumptions for T4 DNA polymerase based on the known structure of the RB69 DNA polymerase.     

Family A and family B DNA polymerases are structurally related: evolutionary implications.

In order to establish the evolutionary relationship between the family A and B DNA polymerases, we have closely compared the 3'-->5' exonuclease domains between the Klenow fragment of E.coli DNA polymerase I (a family A DNA polymerase) and the bacteriophage PRD1 DNA polymerase, the smallest member of the DNA polymerase family B. Although the PRD1 DNA polymerase has a smaller 3'-->5' exonuclease domain, its active sites appear to be very similar to those of the Klenow fragment. Site-directed mutagenesis studies revealed that the residues important for the 3'-->5' exonuclease activity, particularly metal binding ligands for the Klenow fragment, are all conserved in the PRD1 DNA polymerase as well. The metal binding ligands are also essential for the strand-displacement activity of the PRD1 DNA polymerase. Based on these results and the studies by others in various systems, we conclude that family A and B DNA polymerases, at least in the 3'-->5' exonuclease domain, are structurally as well as evolutionarily related.     

Characterization of a replicative DNA polymerase mutant with reduced fidelity and increased translesion synthesis capacity

Changing a highly conserved amino acid in motif A of any of the four yeast family B DNA polymerases, DNA polymerase alpha, delta, epsilon or zeta, results in yeast strains with elevated mutation rates. In order to better understand this phenotype, we have performed structure-function studies of homologous mutants of RB69 DNA polymerase (RB69 pol), a structural model for family B members. When Leu415 in RB69 pol is replaced with phenylalanine or glycine, the mutant polymerases retain high-catalytic efficiency for correct nucleotide incorporation, yet have increased error rates due to increased misinsertion, increased mismatch extension and inefficient proofreading. The Leu415Phe mutant also has increased dNTP insertion efficiency opposite a template 8-oxoG and opposite an abasic site. The 2.5 A crystal structure of a ternary complex of RB69 L415F pol with a correctly base-paired incoming dTTP reveals that the phenylalanine ring is accommodated within a cavity seen in the wild-type enzyme, without steric clash or major change in active site geometry, consistent with retention of high-catalytic efficiency for correct incorporation. In addition, slight structural differences were observed that could be relevant to the reduced fidelity of L415F RB69 pol.     

Invariant Asp-1122 and Asp-1124 are essential residues for polymerization catalysis of family D DNA polymerase from Pyrococcus horikoshii

Family D DNA polymerase has recently been found in the Euryarchaeota subdomain of Archaea. Its genes are adjacent to several other genes related to DNA replication, repair, and recombination in the genome, suggesting that this enzyme may be the major DNA replicase in Euryarchaeota. Although it possesses strong polymerization and proofreading activities, the motifs common to other DNA polymerase families are absent in its sequences. Here we report the mapping of the catalytic residues in a family D DNA polymerase from Pyrococcus horikoshii. Site-directed alanine mutants for 28 conserved aspartic acid or glutamic acid residues were screened for polymerization and 3'-5' exonuclease activities. We identified the invariant aspartates Asp-1122 and Asp-1124 within the most conserved motif as the catalytic residues involved in DNA polymerization. Alanine mutation at either site caused a loss of polymerization activity, whereas the conserved mutants, D1122E, D1124N, and D1124E, had slightly reduced polymerization activity. We also found that the 3'-5' exonuclease activity remains in D1122A and D1124A, indicating that the catalytic residues of DNA polymerization are different from those of the 3'-5' exonuclease activity. Furthermore we determined the molecular mass of the recombinant enzyme by gel filtration and proposed a heterotetrameric structure for this enzyme.     

Accuracy, lesion bypass, strand displacement and translocation by DNA polymerases

The structures of DNA polymerases from different families show common features and significant differences that shed light on the ability of these enzymes to accurately copy DNA and translocate. The structure of a B family DNA polymerase from phage RB69 exhibits an active-site closing conformational change in the fingers domain upon forming a ternary complex with primer template in deoxynucleoside triphosphate. The rotation of the fingers domain alpha-helices by 60 degrees upon dNTP binding is analogous to the changes seen in other families of polymerases. When the 3' terminus is bound to the editing 3' exonuclease active site, the orientation of the DNA helix axis changes by 40 degrees and the thumb domain re-orients with the DNA. Structures of substrate and product complexes of T7 RNA polymerase, a structural homologue of T7 DNA polymerase, show that family polymerases use the rotation conformational change of the fingers domain to translocate down the DNA. The fingers opening rotation that results in translocation is powered by the release of the product pyrophosphate and also enables the Pol I family polymerases to function as a helicase in displacing the downstream non-template strand from the template strand.     

Mutagenic DNA repair: insertion of nucleotides opposite non-coding template structures by a reversed 3'-5' exonuclease reaction? A hypothesis.

1. The enzymatic mechanism of mutagenic DNA repair is unknown. None of the characterized DNA polymerases is capable of polymerization past non-coding template structures. 2. A hypothesis is proposed according to which polymerization opposite non-coding template structures is catalyzed by the DNA-polymerase-associated 3'-5' exonuclease under conditions which shift the equilibrium of the 3'-5' exonuclease reaction DNAn + H2O in equilibrium DNAn-1 + dNMP to the left, i.e. to the incorporation of deoxynucleoside monophosphates. 3. Conditions which favor the incorporation of dNMP by the reversed 3'-5' exonuclease reaction include a high dNMP concentration, a coupled H2O-consuming reaction and a hydrophobic enzyme environment. 4. The statements of the hypothesis are supported by published work on the biochemistry of DNA polymerases and their associated 3'-5' exonucleases, the genetics of mutagenic DNA repair and the involvement of Escherichia coli DNA polymerase III in this process. 5. The hypothesis offers an explanation of the mutator and antimutator properties of certain genes, in particular of DNA polymerase genes, and also explains how some drugs act mutagenically during DNA replication and antimutagenically against mutagenic DNA repair.