The role of the Escherichia coli mug protein in the removal of uracil and 3,N(4)-ethenocytosine from DNA.

The human thymine-DNA glycosylase has a sequence homolog in Escherichia coli that is described to excise uracils from U.G mismatches (Gallinari, P., and Jiricny, J. (1996) Nature 383, 735-738) and is named mismatched uracil glycosylase (Mug). It has also been described to remove 3,N(4)-ethenocytosine (epsilonC) from epsilonC.G mismatches (Saparbaev, M., and Laval, J. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 8508-8513). We used a mug mutant to clarify the role of this protein in DNA repair and mutation avoidance. We find that inactivation of mug has no effect on C to T or 5-methylcytosine to T mutations in E. coli and that this contrasts with the effect of ung defect on C to T mutations and of vsr defect on 5-methylcytosine to T mutations. Even under conditions where it is overproduced in cells, Mug has little effect on the frequency of C to T mutations. Because uracil-DNA glycosylase (Ung) and Vsr are known to repair U.G and T.G mismatches, respectively, we conclude that Mug does not repair U.G or T.G mismatches in vivo. A defect in mug also has little effect on forward mutations, suggesting that Mug does not play a role in avoiding mutations due to endogenous damage to DNA in growing E. coli. Cell-free extracts from mug(+) ung cells show very little ability to remove uracil from DNA, but can excise epsilonC. The latter activity is missing in extracts from mug cells, suggesting that Mug may be the only enzyme in E. coli that can remove this mutagenic adduct. Thus, the principal role of Mug in E. coli may be to help repair damage to DNA caused by exogenous chemical agents such as chloroacetaldehyde.     

Mutation by [5-3H]cytosine decay in DNA of Escherichia coli lacking uracil-DNA glycosylase activity

The mutagenic local effect of tritium decay at the 5 position of cytosine in DNA of Escherichia coli was determined in wild-type and in ung strains defective in uracil-DNA glycosylase. In the absence of this in vivo activity any genetic consequences of uracil residues formed in DNA should be enhanced. However, the mutation frequency response was no greater in the mutant strain than in the wild type. This finding is inconsistent with the earlier suggestion that efficient production of C to T transitions by the local effect of [5-3H]cytosine decay results from the formation of uracil in cellular DNA. Some other intermediate should be considered, one that is not a substrate for uracil-DNA glycosylase.     

Escherichia coli K-12 mutants deficient in uracil-DNA glycosylase.

A new assay specific for uracil-DNA glycosylase is described, Escherichia coli mutants partially and totally deficient in uracil-DNA glycosylase activity have been isolated by using this assay in mass-screening procedures. These have been designated ung mutants. The ung gene maps between tyrA and nadB on the E. coli chromosome. T4 phage containing uracil in their DNA grow on the most glycosylase-deficient hosts but are unable to grow on wild-type bacteria. This provides a simple spot test for the ung genotype. The ung mutants show slightly higher rates of spontaneous mutation to antibiotic resistance. Taken together, these results suggest a central role for uracil-DNA glycosylase in the initiation of an excision repair pathway for the exclusion of uracil from DNA.     

Role of the N-terminal proline residue in the catalytic activities of the Escherichia coli Fpg protein

The Escherichia coli Fpg protein is a DNA glycosylase/AP lyase. It removes, in DNA, oxidized purine residues, including the highly mutagenic C8-oxo-guanine (8-oxoG). The catalytic mechanism is believed to involve the formation of a transient Schiff base intermediate formed between DNA containing an oxidized residue and the N-terminal proline of the Fpg protein. The importance and the role of this proline upon the various catalytic activities of the Fpg protein was examined by targeted mutagenesis, resulting in the construction of three mutant Fpg proteins: Pro-2 --> Gly (FpgP2G), Pro-2 --> Thr (FpgP2T), and Pro-2 --> Glu (FpgP2E). The formamidopyrimidine DNA glycosylase activities of FpgP2G and FpgP2T were comparable and accounted for 10% of the wild-type activity. FpgP2G and FpgP2T had barely detectable 8-oxoG-DNA glycosylase activity and produced minute Schiff base complex with 8-oxoG/C DNA. FpgP2G and FpgP2T mutants did not cleave a DNA containing preformed AP site but readily produced Schiff base complex with this substrate. FpgP2E was completely inactive in all the assays. The binding constants of the different mutants when challenged with a duplex DNA containing a tetrahydrofuran residue were comparable. The mutant Fpg proteins barely or did not complement in vivo the spontaneous transitions G/C --> T/A in E. coli BH990 (fpg mutY) cells. These results show the mandatory role of N-terminal proline in the 8-oxoG-DNA glycosylase activity of the Fpg protein in vitro and in vivo as well as in its AP lyase activity upon preformed AP site but less in the 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine-DNA glycosylase activity.     

Escherichia coli DNA glycosylase Mug: a growth-regulated enzyme required for mutation avoidance in stationary-phase cells

The Escherichia coli DNA glycosylase Mug excises 3,N(4)-ethenocytosines (epsilon C) and uracils from DNA, but its biological function is obscure. This is because epsilon C is not found in E. coli DNA, and uracil-DNA glycosylase (Ung), a distinct enzyme, is much more efficient at removing uracils from DNA than Mug. We find that Mug is overexpressed as cells enter stationary phase, and it is maintained at a fairly high level in resting cells. This is true of cells grown in rich or minimal media, and the principal regulation of mug is at the level of mRNA. Although the expression of mug is strongly dependent on the stationary-phase sigma factor, sigma(S), when cells are grown in minimal media, it shows only a modest dependence on sigma(S) when cells are grown in rich media. When mug cells are maintained in stationary phase for several days, they acquire many more mutations than their mug(+) counterparts. This is true in ung as well as ung(+) cells, and a majority of new mutations may not be C to T. Our results show that the biological role of Mug parallels its expression in cells. It is expressed poorly in exponentially growing cells and has no apparent role in mutation avoidance in these cells. In contrast, Mug is fairly abundant in stationary-phase cells and has an important anti-mutator role at this stage of cell growth. Thus, Mug joins a very small coterie of DNA repair enzymes whose principal function is to avoid mutations in stationary-phase cells.     

Uracil-DNA glycosylase inhibitor of bacteriophage PBS2: cloning and effects of expression of the inhibitor gene in Escherichia coli.

The uracil-DNA glycosylase inhibitor gene of bacteriophage PBS2 was cloned, and the effects of this inhibitor on Escherichia coli cells that contain uracil-DNA glycosylase activity were determined. A PBS2 genomic library was constructed by inserting EcoRI restriction fragments of PBS2 DNA into a plasmid pUC19 vector. The library was used to transform wild-type (ung+) E. coli, and the presence of the functional inhibitor gene was determined by screening for colonies that supported growth of M13mp19 phage containing uracil-DNA. A clone was identified that carried a 4.1-kilobase EcoRI DNA insert in the vector plasmid. Extracts of cells transformed with this recombinant plasmid lacked detectable uracil-DNA glycosylase activity and contained a protein that inhibited the activity of purified E. coli uracil-DNA glycosylase in vitro. The uracil-DNA glycosylase inhibitor expressed in these E. coli was partially purified and characterized as a heat-stable protein with a native molecular weight of about 18,000. Hence, we conclude that the PBS2 uracil-DNA glycosylase inhibitor gene was cloned and that the gene product has properties similar to those from PBS2-infected Bacillus subtilis cells. Inhibitor gene expression in E. coli resulted in (i) a weak mutator phenotype, (ii) a growth rate similar to that of E. coli containing pUC19 alone, (iii) a sensitivity to the antifolate drug aminopterin similar to that of cells lacking the inhibitor gene, and (iv) an increased resistance to the lethal effects of 5-fluoro-2'-deoxyuridine. These physiological properties are consistent with the phenotypes of other ung mutants.     

Crystal structure of a thwarted mismatch glycosylase DNA repair complex

The bacterial mismatch-specific uracil-DNA glycosylase (MUG) and eukaryotic thymine-DNA glycosylase (TDG) enzymes form a homologous family of DNA glycosylases that initiate base-excision repair of G:U/T mismatches. Despite low sequence homology, the MUG/TDG enzymes are structurally related to the uracil-DNA glycosylase enzymes, but have a very different mechanism for substrate recognition. We have now determined the crystal structure of the Escherichia coli MUG enzyme complexed with an oligonucleotide containing a non-hydrolysable deoxyuridine analogue mismatched with guanine, providing the first structure of an intact substrate-nucleotide productively bound to a hydrolytic DNA glycosylase. The structure of this complex explains the preference for G:U over G:T mispairs, and reveals an essentially non-specific pyrimidine-binding pocket that allows MUG/TDG enzymes to excise the alkylated base, 3, N(4)-ethenocytosine. Together with structures for the free enzyme and for an abasic-DNA product complex, the MUG-substrate analogue complex reveals the conformational changes accompanying the catalytic cycle of substrate binding, base excision and product release.     

Fidelity of uracil-initiated base excision DNA repair in Escherichia coli cell extracts

The error frequency and mutational specificity associated with Escherichia coli uracil-initiated base excision repair were measured using an M13mp2 lacZalpha DNA-based reversion assay. Repair was detected in cell-free extracts utilizing a form I DNA substrate containing a site-specific uracil residue. The rate and extent of complete uracil-DNA repair were measured using uracil-DNA glycosylase (Ung)- or double-strand uracil-DNA glycosylase (Dug)-proficient and -deficient isogenic E. coli cells. In reactions utilizing E. coli NR8051 (ung(+) dug(+)), approximately 80% of the uracil-DNA was repaired, whereas about 20% repair was observed using NR8052 (ung(-) dug(+)) cells. The Ung-deficient reaction was insensitive to inhibition by the PBS2 uracil-DNA glycosylase inhibitor protein, implying the involvement of Dug activity. Under both conditions, repaired form I DNA accumulated in conjunction with limited DNA synthesis associated with a repair patch size of 1-20 nucleotides. Reactions conducted with E. coli BH156 (ung(-) dug(+)), BH157 (ung(+) dug(-)), and BH158 (ung(-) dug(-)) cells provided direct evidence for the involvement of Dug in uracil-DNA repair. The rate of repair was 5-fold greater in the Ung-proficient than in the Ung-deficient reactions, while repair was not detected in reactions deficient in both Ung and Dug. The base substitution reversion frequency associated with uracil-DNA repair was determined to be approximately 5.5 x 10(-)(4) with transversion mutations dominating the mutational spectrum. In the presence of Dug, inactivation of Ung resulted in up to a 7.3-fold increase in mutation frequency without a dramatic change in mutational specificity.     

Human uracil-DNA glycosylase complements E. coli ung mutants.

We have previously isolated a cDNA encoding a human uracil-DNA glycosylase which is closely related to the bacterial and yeast enzymes. In vitro expression of this cDNA produced a protein with an apparent molecular weight of 34 K in agreement with the size predicted from the sequence data. The in vitro expressed protein exhibited uracil-DNA glycosylase activity. The close resemblance between the human and the bacterial enzyme raised the possibility that the human enzyme may be able to complement E. coli ung mutants. In order to test this hypothesis, the human uracil-DNA glycosylase cDNA was established in a bacterial expression vector. Expression of the human enzyme as a LacZ alpha-humUNG fusion protein was then studied in E. coli ung mutants. E. coli cells lacking uracil-DNA glycosylase activity exhibit a weak mutator phenotype and they are permissive for growth of phages with uracil-containing DNA. Here we show that the expression of human uracil-DNA glycosylase in E. coli can restore the wild type phenotype of ung mutants. These results demonstrate that the evolutionary conservation of the uracil-DNA glycosylase structure is also reflected in the conservation of the mechanism for removal of uracil from DNA.     

Isolation of insertion, deletion, and nonsense mutations of the uracil-DNA glycosylase (ung) gene of Escherichia coli K-12.

Two uracil-DNA glycosylase (ung) mutation selection procedures based upon the ability of uracil glycosylase to degrade the chromosomes of organisms containing uracil-DNA were devised to obtain a collection of well-defined ung alleles. In an enrichment procedure, lysogens were selected from Escherichia coli cultures infected with lambda pKanr phage containing uracil in their DNA. (These uracil-DNA phage were prepared by growth on host cells deficient in both dUTPase and uracil-DNA glycosylase.) The lysogenic Kanr population was enriched for uracil glycosylase-deficient mutants by a factor of 10(4). In a phage suicide selection procedure, lambda pung+ phage were unable to form plaques on dut ung cells containing uracil-DNA in their chromosomes, and all of the progeny were lambda pung-. Deletion, insertion (ung::Mu and ung::Tn10), nonsense, and missense mutants were isolated by using these procedures. Extracts of three insertion mutants contained no detectable enzyme activity. All of the other mutant isolates had less than 1% of the normal uracil glycosylase specific activity. The previously studied ung-1 allele, which was derived by N-methyl-N'-nitro-N-nitrosoguanidine mutagenesis, produced about 0.02% of the normal amount of uracil glycosylase activity. No significant phenotypic differences between ung-1 and ung::Tn10 alleles were observed. Variations of the lysogen selection procedure may be helpful for isolating other DNA glycosylase mutations in E. coli and other organisms.     

Uracil-DNA glycosylase causes 5-bromodeoxyuridine photosensitization in Escherichia coli K-12.

An Escherichia coli uracil-DNA glycosylase-defective mutant (ung-1 thyA) was more resistant than its wild-type counterpart (ung+ thyA) to the killing effect of UV light when cultured in medium containing 5-bromouracil or 5-bromo-2'-deoxyuridine (BrdUrd). The phenotype of resistance to BrdUrd photosensitization and the uracil-DNA glycosylase deficiency appeared to be 100% cotransduced by P1 phage. During growth with BrdUrd, both strains exhibited similar growth rates and 5-bromouracil incorporation into DNA. The resistant phenotype of the ung-1 mutant was observed primarily during the stationary phase. In cells carrying 5-bromouracil-substituted DNA, mutations causing resistance to rifampin and valine were induced by UV irradiation at a higher frequency in the wild type than in the ung-1 mutant. This Ung-dependent UV mutagenesis required UmuC function. These results suggest that the action of the uracil-DNA glycosylase on UV-irradiated 5-bromouracil-substituted DNA produces lethal and mutagenic lesions. The BrdUrd photosensitization-resistant phenotype allowed us to develop a new, efficient method for enriching and screening ung mutants.     

The crystal structure of mismatch-specific uracil-DNA glycosylase (MUG) from Deinococcus radiodurans reveals a novel catalytic residue and broad substrate specificity

Deinococcus radiodurans is extremely resistant to the effects of ionizing radiation. The source of the radiation resistance is not known, but an expansion of specific protein families related to stress response and damage control has been observed. DNA repair enzymes are among the expanded protein families in D. radiodurans, and genes encoding five different uracil-DNA glycosylases are identified in the genome. Here we report the three-dimensional structure of the mismatch-specific uracil-DNA glycosylase (MUG) from D. radiodurans (drMUG) to a resolution of 1.75 angstroms. Structural analyses suggest that drMUG possesses a novel catalytic residue, Asp-93. Activity measurements show that drMUG has a modified and broadened substrate specificity compared with Escherichia coli MUG. The importance of Asp-93 for activity was confirmed by structural analysis and abolished activity for the mutant drMUGD93A. Two other microorganisms, Bradyrhizobium japonicum and Rhodopseudomonas palustris, possess genes that encode MUGs with the highest sequence identity to drMUG among all of the bacterial MUGs examined. A phylogenetic analysis indicates that these three MUGs form a new MUG/thymidine-DNA glycosylase subfamily, here called the MUG2 family. We suggest that the novel catalytic residue (Asp-93) has evolved to provide drMUG with broad substrate specificity to increase the DNA repair repertoire of D. radiodurans.     

Role of uracil-DNA glycosylase in the repair of deaminated cytosine residues of DNA in Escherichia coli.

Uracil-DNA glycosylase, which acts specifically on uracil-containing DNA, was purified 250-fold from an extract of Escherichia coli 1100. The enzyme releases free uracil from DNA, producing alkali-labile apyrimidinic sites in the DNA. The enzyme is active on both native and heat-denatured DNA of phage PBS1, which contains uracil in place of thymine. piX174 DNA which had been treated with bisulfite and then at alkaline pH was susceptible to the action of uracil-DNA glycosylase. Since DNA treated with bisulfite alone was less susceptible to the enzyme, it is likely that the enzyme recognizes deaminated cytosine, namely uracil, but not bisulfite adducts of uracil and cytosine in the treated DNA. DNA treated with nitrite or hydroxylamine was not attacked by the enzyme. Enzyme activity acting on bisulfite-treated DNA was absent from an extract of E. coli mutant BD10 (ung). The mutant exhibited higher sensitivity to bisulfite than did the wild-type strain and was unable to reactivate phage T1 pre-exposed to bisulfite and weak alkali.     

Identification of Escherichia coli Mismatch-specific Uracil DNA Glycosylase as a Robust Xanthine DNA Glycosylase

The gene for the mismatch-specific uracil DNA glycosylase (MUG) was identified in the Escherichia coli genome as a sequence homolog of the human thymine DNA glycosylase with activity against mismatched uracil base pairs. Examination of cell extracts led us to detect a previously unknown xanthine DNA glycosylase (XDG) activity in E. coli. DNA glycosylase assays with purified enzymes indicated the novel XDG activity is attributable to MUG. Here, we report a biochemical characterization of xanthine DNA glycosylase activity in MUG. The wild type MUG possesses more robust activity against xanthine than uracil and is active against all xanthine-containing DNA (C/X, T/X, G/X, A/X and single-stranded X). Analysis of potentials of mean force indicates that the double-stranded xanthine base pairs have a relatively narrow energetic difference in base flipping, whereas the tendency for uracil base flipping follows the order of C/U > G/U > T/U > A/U. Site-directed mutagenesis performed on conserved motifs revealed that Asn-140 and Ser-23 are important determinants for XDG activity in E. coli MUG. Molecular modeling and molecular dynamics simulations reveal distinct hydrogen-bonding patterns in the active site of E. coli MUG that account for the specificity differences between E. coli MUG and human thymine DNA glycosylase as well as that between the wild type MUG and the Asn-140 and Ser-23 mutants. This study underscores the role of the favorable binding interactions in modulating the specificity of DNA glycosylases.     

Quantitative determination of uracil residues in Escherichia coli DNA: Contribution of ung, dug, and dut genes to uracil avoidance

The steady-state levels of uracil residues in DNA extracted from strains of Escherichia coli were measured and the influence of defects in the genes for uracil-DNA glycosylase (ung), double-strand uracil-DNA glycosylase (dug), and dUTP pyrophosphatase (dut) on uracil accumulation was determined. A sensitive method, called the Ung-ARP assay, was developed that utilized E. coli Ung, T4pdg, and the Aldehyde Reactive Probe reagent to label abasic sites resulting from uracil excision with biotin. The limit of detection was one uracil residue per million DNA nucleotides (U/10(6)nt). Uracil levels in the genomic DNA of E. coli JM105 (ung+ dug+) were at the limit of detection, as were those of an isogenic dug mutant, regardless of growth phase. Inactivation of ung in JM105 resulted in 31+/-2.6 U/10(6)nt during early log growth and 19+/-1.7 U/10(6)nt in saturated phase. An ung dug double mutant (CY11) accumulated 33+/-2.9 U/10(6)nt and 23+/-1.8U/10(6)nt during early log and saturated phase growth, respectively. When cultures of CY11 were supplemented with 20 ng/ml of 5-fluoro-2'-deoxyuridine, uracil levels in early log phase growth DNA rose to 125+/-1.7 U/10(6)nt. Deoxyuridine supplementation reduced the amount of uracil in CY11 DNA, but uridine did not. Levels of uracil in DNA extracted from CJ236 (dut-1 ung-1) were determined to be 3000-8000 U/10(6)nt as measured by the Ung-ARP assay, two-dimensional thin-layer chromatography of metabolically-labeled 32P DNA, and LC/MS of uracil and thymine deoxynucleosides. DNA sequencing revealed that the sole molecular defect in the CJ236 dUTP pyrophosphatase gene was a C-->T transition mutation that resulted in a Thr24Ile amino acid change.     

Repair of thymine.guanine and uracil.guanine mismatched base-pairs in bacteriophage M13mp18 DNA heteroduplexes

Repair of thymine.guanine (T.G) and uracil.guanine (U.G) mismatched base-pairs in bacteriophage M13mp18 replicative form (RF) DNA was compared upon transfection into repair-proficient or repair-deficient Escherichia coli strains. Oligonucleotide-directed mutagenesis was used to prepare covalently closed circular heteroduplexes that contained the mismatched base-pair at a restriction recognition site. The heteroduplexes were unmethylated at dam (5'-GATC-3') sites to avoid methylation-directed biasing of repair. In an E. coli host containing uracil-DNA glycosylase (ung+), about 97% of the transfecting U.G-containing heteroduplexes had the U residue excised by the uracil-excision repair system. With the analogous T.G mispair, mismatch repair operated on almost all of the transfecting heteroduplexes and removed the T residue in about 75% of them when the mismatched T was on the minus strand of the RF DNA. Similar preferential excision of the minus-strand's mismatched base was observed whether the heteroduplex RF DNA molecules had only one or both strands unmethylated at dcm (5'-CC(A/T)GG-3') sites and whether the RF DNA was prepared by primer extension in vitro or by reannealing mutant and non-mutant DNA strands. Also, the extent and directionality of repair was the same at a U.G mispair in ung- host cells as at the analogous T.G mispair in ung- or ung+ cells. Only in a mismatch repair-deficient (mutH-) host was the plus strand of the transfecting M13mp18 heteroduplex DNA preferentially repaired. It is suggested that the plus strand nick made by the M13-encoded gene II protein might be employed by a mutH- host to initiate repair on that strand.     

Specific mutator effects of ung (uracil-DNA glycosylase) mutations in Escherichia coli.

Studies of trpA reversions revealed that G:C leads to A:T transitions were stimulated about 30-fold in E. coli ung mutants, whereas other base substitutions were not affected. A dUTPase (dut) mutation, which increases the incorporation of uracil into DNA in place of thymine, had no significant effect on the rate of G:C leads to A:T transitions. The results support the proposal that the glycosylase functions to reduce the mutation rate in wild-type cells by acting in the repair of DNA cytosine residues that have undergone spontaneous deamination to uracil. Further support was provided by the finding that when lambda bacteriophages were treated with bisulfite, an agent known to produce cytosine deamination, the frequency of clear-plaque mutants was increased an additional 20-fold by growth on an ung host. Bisulfite-induced mutations of the cellular chromosome, however, were about equal in ung+ and ung strains; it was found that during the treatment of ung+ cells with bisulfite, the glycosylase was inactivated.     

W-mutagenesis in the bisulfite-treated lambda phage

Survival of phage lambda cI857 inactivated by bisulfite (pH 5.6, 37 degrees C) is higher (the dose modification factor approx. 1.2) and frequency of bisulfite-induced c-mutations 2-4-fold lower on the lawn of the wild-type strain ung+, as compared to ung-1 mutant deficient in uracil-DNA glycosylase. Irradiation of host cells by a moderate UV dose inducing SOS repair system enhances the frequency of bisulfite-induced c-mutations 2-3-fold in the wild-type (ung+) host, but not in the ung-1 mutant. It is suggested that W-mutagenesis in bisulfite-treated lambda phage in the ung+ cells is due to SOS repair of apyrimidinic sites which are produced during excision of uracil residues, the products of cytosine deamination.     

The properties of a bacteriophage T5 mutant unable to induce deoxyuridine 5'-triphosphate nucleotidohydrolase. Synthesis of uracil-containing T5 deoxyribonucleic acid.

Bacteriophage T5 induces a deoxyuridine 5'-triphosphate nucleotidohydrolase (dUTPase) activity during infection of Escherichia coli. A T5 mutant (T5 dut) unable to induce this dUTPase activity has been isolated. Although this mutant is viable, the E. coli dUTPase activity is not sufficiently active to exclude uracil from the progeny DNA and about 3% of the thymine is replaced by uracil. When the mutant is grown in an E. coli dut host about 12% of the thymine in the progeny DNA is replaced by uracil. T5 phage containing 12% uracil can replicate in uracil-DNA glycosylase-deficient (ung) hosts with high efficiency, but fail to replicate in ung+ hosts. The amount of thymine replaced by uracil in the progeny produced in dut hosts is nearly independent of the ung genotype, indicating that the host uracil-DNA glycosylase-dependent repair pathway is not operating efficiently to remove uracil from T5 progeny DNA.     

Mutations in active-site residues of the uracil-DNA glycosylase encoded by vaccinia virus are incompatible with virus viability.

The D4R gene of vaccinia virus encodes a functional uracil-DNA glycosylase that is essential for viral viability (D. T. Stuart, C. Upton, M. A. Higman, E. G. Niles, and G. McFadden, J. Virol. 67:2503-2513, 1993), and a D4R mutant, ts4149, confers a conditional lethal defect in viral DNA replication (A. K. Millns, M. S. Carpenter, and A. M. DeLange, Virology 198:504-513, 1994). The mutant ts4149 protein was expressed in vitro and assayed for uracil-DNA glycosylase activity. Less than 6% of wild-type activity was observed at permissive temperatures, but the ts4149 protein was completely inactive at the nonpermissive temperature. Mutagenesis of the ts4149 gene back to wild type (Arg-179-->Gly) restored full activity. The ts4149 protein was considerably reduced in lysates of cells infected at the permissive temperature, and its activity was undetectable, even in the presence of the uracil glycosylase inhibitor protein, which inhibits the host uracil-DNA glycosylases but not that of vaccinia virus. Thus the ts4149 protein is thermolabile, correlating uracil removal with vaccinia virus DNA replication. Three active-site amino acids of the vaccinia virus uracil-DNA glycosylase were mutated (Asp-68-->Asn, Asn-120-->Val, and His-181-->Leu), producing proteins that were completely defective in uracil excision but still retained the ability to bind DNA. Each mutated D4R gene was transfected into vaccinia virus ts4149-infected cells in order to assess the recombination events that allowed virus survival at 40 degrees C. Genetic analysis and sequencing studies revealed that the only viruses to survive were those in which recombination eliminated the mutant locus. We conclude that the uracil cleavage activity of the D4R protein is essential for its function in vaccinia virus DNA replication, suggesting that the removal of uracil residues plays an obligatory role.