Multiple pathways for repair of oxidative DNA damages caused by X rays and hydrogen peroxide in Escherichia coli.

The responses of Escherichia coli to X rays and hydrogen peroxide were examined in mutants which are deficient in one or more DNA repair genes. Mutant cells deficient in either exonuclease III (xthA) or endonuclease IV (nfo) had normal resistance to X rays, but an xthA-nfo double mutant showed a sensitivity increased over that of either parental strain. A DNA polymerase I mutant (polA) was more sensitive than the xthA-nfo mutant. Cells bearing mutations in all of the polA, xthA, and nfo genes were more sensitive to X rays than polA and xthA-nfo mutants. Similar repair responses were obtained by exposing these mutant cells to hydrogen peroxide, with the exception of the xthA mutant, which was hypersensitive to this agent. The DNA polymerase III mutant (polC(Ts)) was slightly more sensitive to the agents than the wild-type strain at the restrictive temperature. The sensitivity of the polC-xthA-nfo mutant to X rays and hydrogen peroxide was greater than that of polC but almost the same as that of the xthA-nfo mutant. From these results it appears that there are at least four repair pathways, the DNA polymerase I-, exonuclease III/endonuclease IV and DNA polymerase I-, exonuclease III/endonuclease IV and DNA polymerase III-, and exonuclease III/endonuclease IV-dependent pathways, for the repair of oxidative DNA damages in E. coli.     

Endonuclease III and endonuclease IV protect Escherichia coli from the lethal and mutagenic effects of near-UV irradiation.

In contrast to the DNA damage caused by far-UV (lambda < 290 nm), near-UV (290 < lambda < 400 nm) induced DNA damage is partially oxygen dependent, suggesting the involvement of reactive oxygen species. To test the hypothesis that enzymes that protect cells from oxidative DNA damage are also involved in preventing near-UV mediated DNA damage, isogenic strains deficient in one or more of exonuclease III (xthA), endonuclease IV (nfo), and endonuclease III (nth) were exposed to increasing levels of far-UV and near-UV. All strains, with the exception of the nth single mutant, were found to be hypersensitive to the lethal effects of near-UV relative to a wild-type strain. A triple mutant strain (nth nfo xthA) exhibited the greatest sensitivity to near-UV-mediated lethality. The triple mutant was more sensitive than the nfo xthA double mutant to the lethal effects of near-UV, but not far-UV. A forward mutation assay also revealed a significantly increased sensitivity for the triple mutant compared to the nfo xthA deficient strain in the presence of near-UV. However, the triple mutant was no more sensitive to the mutagenic effects of far-UV than a nfo xthA double mutant. These data suggest that exonuclease III, endonuclease IV, and endonuclease III are important in protection against near-UV-induced DNA damage.     

Endonuclease IV (nfo) mutant of Escherichia coli.

A cloned gene, designated nfo, caused overproduction of an EDTA-resistant endonuclease specific for apurinic-apyrimidinic sites in DNA. The sedimentation coefficient of the enzyme was similar to that of endonuclease IV. An insertion mutation was constructed in vitro and transferred from a plasmid to the Escherichia coli chromosome. nfo mutants had an increased sensitivity to the alkylating agents methyl methanesulfonate and mitomycin C and to the oxidants tert-butyl hydroperoxide and bleomycin. The nfo mutation enhanced the killing of xth (exonuclease III) mutants by methyl methanesulfonate, H2O2, tert-butyl hydroperoxide, and gamma rays, and it enhanced their mutability by methyl methanesulfonate. It also increased the temperature sensitivity of an xth dut (dUTPase) mutant that is defective in the repair of uracil-containing DNA. These results are consistent with earlier findings that endonuclease IV and exonuclease III both cleave DNA 5' to an apurinic-apyrimidinic site and that exonuclease III is more active. However, nfo mutants were more sensitive to tert-butyl hydroperoxide and to bleomycin than were xth mutants, suggesting that endonuclease IV might recognize some lesions that exonuclease III does not. The mutants displayed no marked increase in sensitivity to 254-nm UV radiation, and the addition of an nth (endonuclease III) mutation to nfo or nfo xth mutants did not significantly increase their sensitivity to any of the agents tested.     

A mutant endonuclease IV of Escherichia coli loses the ability to repair lethal DNA damage induced by hydrogen peroxide but not that induced by methyl methanesulfonate.

A mutant allele of the Escherichia coli nfo gene encoding endonuclease IV, nfo-186, was cloned into plasmid pUC18. When introduced into an E. coli xthA nfo mutant, the gene product of nfo-186 complemented the hypersensitivity of the mutant to methyl methanesulfonate (MMS) but not to hydrogen peroxide (H2O2) and bleomycin. These results suggest that the mutant endonuclease IV has normal activity for repairing DNA damages induced by MMS but not those induced by H2O2 and bleomycin. A missense mutation in the cloned nfo-186 gene, in which the wild-type glycine 149 was replaced by aspartic acid, was detected by DNA sequencing. The wild-type and mutant endonuclease IV were purified to near homogeneity, and their apurinic (AP) endonuclease and 3'-phosphatase activities were determined. No difference was observed in the AP endonuclease activities of the wild-type and mutant proteins. However, 3'-phosphatase activity was dramatically reduced in the mutant protein. From these results, it is concluded that the endonuclease IV186 protein is specifically deficient in the ability to remove 3'-terminus-blocking damage, which is required for DNA repair synthesis, and it is possible that the lethal DNA damage by H2O2 is 3'-blocking damage and not AP-site damage.     

Excision repair of thymine glycols, urea residues, and apurinic sites in Escherichia coli

The genetic requirements for the excision repair of thymine glycols, urea residues, and apurinic (AP) sites were examined by measuring the survival in Escherichia coli mutants of phi X174 replicative form (RF) I transfecting DNA containing selectively introduced lesions. phi X RF I DNA containing thymine glycols was inactivated at a greater rate in mutants deficient in endonuclease III (nth) than in wild-type hosts, suggesting that endonuclease III is involved in the repair of thymine glycols in vivo. phi X RF I DNA containing thymine glycols was also inactivated at a greater rate in mutants that were deficient in both exonuclease III and endonuclease IV (xth nfo) than in wild-type hosts, suggesting that a class II AP endonuclease is required for the in vivo processing of thymine glycols. phi X duplex-transfecting DNA containing urea residues or AP sites was inactivated at a greater rate in xth nfo double mutants than in wild-type, but not single-mutant, hosts, suggesting that exonuclease III or endonuclease IV is required for the repair of these damages and that either activity can substitute for the other. These data are in agreement with the known in vitro substrate specificities of endonuclease III, exonuclease III, and endonuclease IV.     

Repair of x-ray-induced deoxyribonucleic acid single-strand breaks in xth mutants of Escherichia coli.

An exonuclease III-deficient strain of Escherichia coli K-12, BW2001 (xthA11), was unable to perform rapid repair of X-ray-induced deoxyribonucleic acid single-strand breaks and appeared to have a defect in the priming of the 3'-termini necessary for initiation of repair synthesis at the breaks. This defect cannot be explained solely by the lack of exonuclease III activity, because other xth mutants tested, including a deletion mutant, repaired radiation-induced strand breaks at close to the normal rate.     

Drosophila Rrp1 complements E. coli xth nfo mutants: protection against both oxidative and alkylation-induced DNA damage.

Drosophila Rrp1 protein has four tightly associated enzymatic activities: DNA strand transfer, ssDNA renaturation, dsDNA 3'-exonuclease and apurinic/apyrimidinic (AP) endonuclease. The carboxy-terminal region of Rrp1 is homologous to Escherichia coli exonuclease III and several eukaryotic AP endonucleases. All members of this protein family cleave abasic sites. Rrp1 protein was expressed under the control of the E. coli RNA polymerase tac promoter (pRrp1-tac) in two repair deficient E. coli strains (BW528 and LG101) lacking both exonuclease III (xth) and endonuclease IV (nfo). Rrp1 confers resistance to killing by oxidative, antitumor and alkylating agents that damage DNA (hydrogen peroxide, t-butylhydroperoxide, bleomycin, methyl methanesulfonate, and mitomycin C). Complementation of the repair deficiency by Rrp1 provides up to a two log increase in survival and requires the C-terminal nuclease region of Rrp1, but not its N-terminal region. The AP endonuclease activity in extracts from the repair deficient strain LG101 is increased up to 12-fold when the strain contains pRrp1-tac. These results indicate that pRrp1-tac directs the synthesis of active enzyme, and that the nuclease activities of Rrp1 are likely to be the cause of the increased resistance to DNA damage of the mutant cells.     

Isolation of cDNA clones encoding a human apurinic/apyrimidinic endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III) mutants.

Apurinic/apyrimidinic (AP) sites in cellular DNA are considered to be both cytotoxic and mutagenic, and can arise spontaneously or following exposure to DNA damaging agents. We have isolated cDNA clones which encode an endonuclease, designated HAP1 (human AP endonuclease 1), that catalyses the initial step in AP site repair in human cells. The predicted HAP1 protein has an Mr of 35,500 and shows striking sequence similarity (93% identity) to BAP 1, a bovine AP endonuclease enzyme. Significant sequence homology to two bacterial DNA repair enzymes, E. coli exonuclease III and S. pneumoniae ExoA proteins, and to Drosophila Rrp1 protein is also apparent. We have expressed the HAP1 cDNA in E. coli mutants lacking exonuclease III (xth), endonuclease IV (nfo), or both AP endonucleases. The HAP1 protein can substitute for exonuclease III, but not for endonuclease IV, in respect of some, but not all, DNA repair and mutagenesis functions. Moreover, a dut xth (ts) double mutant, which is nonviable at 42 degrees C due to an accumulation of unrepaired AP sites following excision of uracil from DNA, was rescued by expression of the HAP1 cDNA. These results indicate that AP endonucleases show remarkable conservation of both primary sequence and function. We would predict that the HAP1 protein is important in human cells for protection against the toxic and mutagenic effects of DNA damaging agents.     

Development of T7 phage and T7 phage containing apurinic sites in an exonuclease III, endonuclease IV double mutant of Escherichia coli.

The development of bacteriophage T7 was examined in an Escherichia coli double mutant defective for the two major apurinic, apyrimidinic endonucleases (exonuclease III and endonuclease IV, xth nfo). In cells infected with phages containing apurinic sites, the defect in repair enzymes led to a decrease of phage survival and a total absence of bacterial DNA degradation and of phage DNA synthesis. These results directly demonstrate the toxic action of apurinic sites on bacteriophage T7 at the intracellular level and its alleviation by DNA repair. In addition, untreated T7 phage unexpectedly displayed reduced plating efficiency and decreased DNA synthesis in the xth nfo double mutant.     

Homogeneous Escherichia coli endonuclease IV. Characterization of an enzyme that recognizes oxidative damage in DNA

Agents that act via oxygen-derived free radicals form DNA strand breaks with fragmented sugar residues that block DNA repair synthesis. Using a synthetic DNA substrate with a single type of sugar fragment, 3'-phosphoglycolaldehyde esters, we show that in Escherichia coli extracts the only EDTA-resistant diesterase for these damages depends on the bacterial nfo (endonuclease IV) gene. Endonuclease IV was purified to physical homogeneity (Mr = 31,000) from an E. coli strain carrying the cloned nfo gene and in which the enzyme had been induced with paraquat. Although heat-stable and routinely assayed in the presence of EDTA, endonuclease IV was inactivated in the absence of substrate at 23-50 degrees C by either EDTA or 1,10-phenanthroline, suggesting the presence of an essential metal tightly bound to the protein. Purified endonuclease IV released phosphoglycolaldehyde, phosphate, and intact deoxyribose 5-phosphate from the 3'-end of DNA, all with apparent Km of 5-10 nM. The optimal KCl or NaCl concentration for 3'-phosphoglycolaldehyde release was 50-100 mM. The purified enzyme had endonuclease activity against partially depurinated DNA but lacked significant nonspecific nuclease activities. Endonuclease IV also activated H2O2-damaged DNA for repair synthesis by DNA polymerase I. Thus, endonuclease IV can act on a variety of oxidative damages in DNA, consistent with a role for the enzyme in combating free-radical toxicity.     

The relative importance of Escherichia coli exonuclease III and endonuclease IV for the hydrolysis of 3''-phosphoglycolate ends in polydeoxynucleotides.

In vitro, in the presence of Mg++, the 3'-phosphoglycolatase activity of endonuclease IV is about 4-times smaller than that of exonuclease III for the same AP endonuclease activity. It thus seems that endonuclease IV has only a minor role in the repair of strand breaks limited by 3'-phosphoglycolate ends in Escherichia coli even after the amount of enzyme has been increased by induction with O2 -generating agents.     

The Brucella abortus xthA-1 gene product participates in base excision repair and resistance to oxidative killing but is not required for wild-type virulence in the mouse model

Exonuclease III, encoded by the xthA gene, plays a central role in the base excision pathway of DNA repair in bacteria. Studies with Escherichia coli xthA mutants have also shown that exonuclease III participates in the repair of oxidative damage to DNA. An isogenic xthA-1 mutant (designated CAM220) derived from virulent Brucella abortus 2308 exhibited increased sensitivity to the alkylating agent methyl methanesulfonate (MMS) compared to the parent strain. In contrast, 2308 and the isogenic xthA-1 mutant displayed similar levels of resistance to the DNA cross-linker mitomycin C. These phenotypic properties are those that would be predicted for a strain defective in base excision repair. The B. abortus xthA-1 mutant also displayed reduced resistance to killing by H2O2 and the ONOO(-)-generating compound 3-morpholinosydnonimine (SIN-1) compared to strain 2308, indicating that the xthA-1 gene product participates in protecting B. abortus 2308 from oxidative damage. Introducing a plasmid-borne copy of the parental xthA-1 gene into CAM220 restored wild-type resistance of this mutant to MMS, H2O2, and SIN-1. Although the B. abortus xthA-1 mutant exhibited increased sensitivity to oxidative killing compared to the parental strain in laboratory assays, CAM220 and 2308 displayed equivalent spleen colonization profiles in C57BL/6 [corrected] mice through 8 weeks postinfection and equivalent intracellular survival and replication profiles in cultured murine macrophages. Thus, although the xthA-1 gene product participates in base excision repair and resistance to oxidative killing in B. abortus 2308, XthA-1 is not required for wild-type virulence of this strain in the mouse model.     

A Drosophila ribosomal protein contains 8-oxoguanine and abasic site DNA repair activities.

Ionizing radiation and normal cellular respiration form reactive oxygen species that damage DNA and contribute to a variety of human disorders including tumor promotion and carcinogenesis. A major product of free radical DNA damage is the formation of 8-oxoguanine, which is a highly mutagenic base modification produced by oxidative stress. Here, Drosophila ribosomal protein S3 is shown to cleave DNA containing 8-oxoguanine residues efficiently, The ribosomal protein also contains an associated apurinic/apyrimidinic (AP) lyase activity, cleaving phosphodiester bonds via a beta,delta elimination reaction. The significance of this DNA repair activity acting on 8-oxoguanine is shown by the ability of S3 to rescue the H2O2 sensitivity of an Escherichia coli mutM strain (defective for the repair of 8-oxoguanine) and to abolish completely the mutator phenotype of mutM caused by 8-oxoguanine-mediated G-->T transversions. The ribosomal protein is also able to rescue the alkylation sensitivity of an E.coli mutant deficient for the AP endonuclease activities associated with exonuclease III (xth) and endonuclease IV (nfo), indicating for the first time that an AP lyase can represent a significant source of DNA repair activity for the repair of AP sites. These results raise the possibility that DNA repair may be associated with protein translation.     

Effect of apurinic/apyrimidinic endonucleases and polyamines on DNA treated with bleomycin and neocarzinostatin: specific formation and cleavage of closely opposed lesions in complementary strands

Bleomycin and neocarzinostatin induce modified apurinic/apyrimidinic (AP) sites by oxidation of the sugar moiety in DNA. In order to quantitatively assess the susceptibility of these lesions to repair endonucleases, drug-treated 3H-labeled colE1 DNA was mixed with 14C-labeled heat-depurinated DNA, and endonuclease-susceptible sites in the mixture were titrated with various AP endonucleases or with polyamines. Single- and double-strand breaks were quantitated by determining the fractions of supercoiled, nicked circular, and linear molecules. Exonuclease III and endonucleases III and IV of Escherichia coli, as well as putrescine, produced a nearly 2-fold increase in single-strand breaks in bleomycin-treated DNA, indicating cleavage of drug-induced AP sites. The bleomycin-induced AP sites were comparable to heat-induced sites in their sensitivity to E. coli endonucleases III and IV but were cleaved by exonuclease III only at high concentrations. Bleomycin-induced AP sites were much more sensitive to cleavage by putrescine than heat-induced sites. Treatment with putrescine or very high concentrations of endonuclease III also increased the number of double-strand breaks in bleomycin-treated DNA, suggesting a minor class of lesion consisting of an AP site accompanied by a closely opposed break in the complementary strand. These complex lesions were resistant to cleavage by endonuclease IV. However, when colE1 DNA was treated with neocarzinostatin, subsequent treatment with putrescine, endonuclease IV, or very high concentrations of endonuclease III produced a dramatic increase in double-strand breaks but no detectable increase in single-strand breaks. These results suggest that virtually all neocarzinostatin-induced AP sites are accompanied by a closely opposed strand break.(ABSTRACT TRUNCATED AT 250 WORDS)     

Role of exonuclease III and endonuclease IV in repair of pyrimidine dimers initiated by bacteriophage T4 pyrimidine dimer-DNA glycosylase.

The role of exonuclease III and endonuclease IV in the repair of pyrimidine dimers in bacteriophage T4-infected Escherichia coli was examined. UV-irradiated T4 showed reduced survival when plated on an xth nfo double mutant but showed wild-type survival on either single mutant. T4 denV phage were equally sensitive when plated on wild-type E. coli or an xth nfo double mutant, suggesting that these endonucleases function in the same repair pathway as T4 pyrimidine dimer-DNA glycosylase. A uvrA mutant of E. coli in which the repair of pyrimidine dimers was dependent on the T4 denV gene carried on a plasmid was constructed. Neither an xth nor an nfo derivative of this strain was more sensitive than the parental strain to UV irradiation. We were unable to construct a uvrA xth nfo triple mutant. In addition, T4, which turns off the host UvrABC excision nuclease, showed reduced plating efficiency on an xth nfo double mutant.     

Phototriggered formation and repair of DNA containing a site-specific single strand break of the type produced by ionizing radiation or AP lyase activity

DNA strand breaks are produced by a variety of agents and processes such as ionizing radiation, xenobiotics, oxidative metabolism, and enzymatic processing of DNA base damage. One of the major types of strand breaks produced by these processes is a single nucleotide gap terminating in 5'- and 3'-phosphates. Previously, we had developed a method for sequence-specifically producing such phosphate-terminated strand breaks in an oligodeoxynucleotide by way of two photochemically activated (caged) building blocks placed in tandem. We now report the design and synthesis of a single caged building block consisting of 1,3-(2-nitrophenyl)-1,3-propanediol, for producing phosphate-terminated strand breaks, and its use producing such a break at a specific site in a double-stranded circular DNA vector. To produce the site-specific break in a duplex vector, a primer containing the caged single strand break was extended opposite the single strand form of a circular DNA vector followed by enzymatic ligation and purification. The single strand break could then be formed in quantitative yield by irradiation of the vector with 365 nm light. In contrast to a previous study, it was found that the strand break can be repaired by Escherichia coli DNA polymerase I and E. coli DNA ligase alone, though less efficiently than in the presence of the 3'-phosphate processing enzyme E. coli endonuclease IV. Repair in the absence of endonuclease IV could be attributed to hydrolysis of the 3'-phosphate in the presence of dNTP and to a lesser extent to exonucleolytic removal of the 3'-phosphate-bearing terminal nucleotide by way of the 3' --> 5' exonuclease activity of polymerase I. This work demonstrates that specialized 3'-end processing enzymes such as endonuclease IV or exonuclease III are not absolutely required for repair of phosphate-terminated gaps. In addition to preparing single strand breaks, the caged building block described should also be useful for preparing double strand breaks and multiply damaged sites that might otherwise be difficult to prepare by other methods due to their lability.     

Repair of DNA strand breaks by the overlapping functions of lesion-specific and non-lesion-specific DNA 3' phosphatases

In Saccharomyces cerevisiae, the apurinic/apyrimidinic (AP) endonucleases Apn1 and Apn2 act as alternative pathways for the removal of various 3'-terminal blocking lesions from DNA strand breaks and in the repair of abasic sites, which both result from oxidative DNA damage. Here we demonstrate that Tpp1, a homologue of the 3' phosphatase domain of polynucleotide kinase, is a third member of this group of redundant 3' processing enzymes. Unlike Apn1 and Apn2, Tpp1 is specific for the removal of 3' phosphates at strand breaks and does not possess more general 3' phosphodiesterase, exonuclease, or AP endonuclease activities. Deletion of TPP1 in an apn1 apn2 mutant background dramatically increased the sensitivity of the double mutant to DNA damage caused by H2O2 and bleomycin but not to damage caused by methyl methanesulfonate. The triple mutant was also deficient in the repair of 3' phosphate lesions left by Tdp1-mediated cleavage of camptothecin-stabilized Top1-DNA covalent complexes. Finally, the tpp1 apn1 apn2 triple mutation displayed synthetic lethality in combination with rad52, possibly implicating postreplication repair in the removal of unrepaired 3'-terminal lesions resulting from endogenous damage. Taken together, these results demonstrate a clear role for the lesion-specific enzyme, Tpp1, in the repair of a subset of DNA strand breaks.     

Effects of metal ion chelators on DNA strand breaks and inactivation produced by hydrogen peroxide in Escherichia coli: detection of iron-independent lesions.

In order to study the role of metallic ions in the H2O2 inactivation of Escherichia coli cells, H2O2-sensitive mutants were treated with metal ion chelators and then submitted to H2O2 treatment. o-Phenanthroline, dipyridyl, desferrioxamine, and neocuproine were used as metal chelators. Cell sensitivity to H2O2 treatment was not modified by neocuproine, suggesting that copper has a minor role in OH production in E. coli. On the other hand, prior treatment with iron chelators protected the cells against the H2O2 lethal effect, indicating that iron participates in the production of OH. However, analysis of DNA sedimentation profiles and DNA degradation studies indicated that these chelators did not completely block the formation of DNA single-strand breaks by H2O2 treatment. Thiourea, a scavenger of OH, caused a reduction in both H2O2 sensitivity and DNA single-strand break production. The breaks observed after treatment with metal chelators and H2O2 were repaired 60 min after H2O2 elimination in xthA but not polA mutant cells. Therefore, we propose that there are at least two pathways for H2O2-induced DNA lesions: one produced by H2O2 through iron oxidation and OH production, in which lesions are repaired by the products of the xthA and polA genes, and the other produced by an iron-independent pathway in which DNA repair requires polA gene products but not those of the xthA gene.     

Relationships between yeast Rad27 and Apn1 in response to apurinic/apyrimidinic (AP) sites in DNA.

Yeast Rad27 is a 5'-->3' exonuclease and a flap endo-nuclease. Apn1 is the major apurinic/apyrimidinic (AP) endonuclease in yeast. The rad27 deletion mutants are highly sensitive to methylmethane sulfonate (MMS). By examining the role of Rad27 in different modes of DNA excision repair, we wish to understand why the cytotoxic effect of MMS is dramatically enhanced in the absence of Rad27. Base excision repair (BER) of uracil-containing DNA was deficient in rad27 mutant extracts in that (i) the Apn1 activity was reduced, and (ii) after DNA incision by Apn1, hydrolysis of 1-5 nucleotides 3' to the baseless sugar phosphate was deficient. Thus, some AP sites may lead to unprocessed DNA strand breaks in rad27 mutant cells. The severe MMS sensitivity of rad27 mutants is not caused by a reduction of the Apn1 activity. Surprisingly, we found that Apn1 endonuclease sensitizes rad27 mutant cells to MMS. Deleting the APN1 gene largely restored the resistance of rad27 mutants to MMS. These results suggest that unprocessed DNA strand breaks at AP sites are mainly responsible for the MMS sensitivity of rad27 mutants. In contrast, nucleotide excision repair and BER of oxidative damage were not affected in rad27 mutant extracts, indicating that Rad27 is specifically required for BER of AP sites in DNA.     

Identification of two apurinic/apyrimidinic endonucleases from Caenorhabditis elegans by cross-species complementation

The Saccharomyces cerevisiae mutant strain YW778, which lacks apurinic/apyrimidinic (AP) endonuclease and 3'-diesterase DNA repair activities, displays high levels of spontaneous mutations and hypersensitivities to several DNA damaging agents. We searched a cDNA library derived from the nematode Caenorhabditis elegans for gene products that would rescue the DNA repair defects of this yeast mutant. We isolated two genes, apn-1 and exo-3, encoding proteins that have not been previously characterized. Both APN-1 and EXO-3 share significant identity with the functionally established Escherichia coli AP endonucleases, endonuclease IV and exonuclease III, respectively. Strain YW778 expressing either apn-1 or exo-3 shows parental levels of spontaneous mutations, as well as resistance to DNA damaging agents that produce AP sites and DNA single strand breaks with blocked 3'-ends. Using an in vitro assay, we show that the apn-1 and exo-3 genes independently express AP endonuclease activity in the yeast mutant. We further characterize the EXO-3 protein and three of its mutated variants E68A, D190A, and H279A. The E68A variant retains both AP endonuclease and 3'-diesterase repair activities in vitro, yet severely lacks the ability to protect strain YW778 from spontaneous and drug-induced DNA lesions, suggesting that this variant E68A may possess a defect that interferes with the repair process in vivo. In contrast, D190A and H279A are completely devoid of DNA repair activities and fail to rescue the genetic instability of strain YW778. Our data strongly suggest that EXO-3 and APN-1 are enzymes possessing intrinsic AP endonuclease and 3'-diesterase activities.