An Assay for the Activity of Base Excision Repair Enzymes in Cellular Extracts Using Fluorescent DNA Probes.

Damaged DNA bases are removed by the base excision repair (BER) mechanism. This enzymatic process begins with the action of one of DNA glycosylases, which recognize damaged DNA bases and remove them by hydrolyzing N-glycosidic bonds with the formation of apurinic/apyrimidinic (AP) sites. Apurinic/apyrimidinic endonuclease 1 (APE1) hydrolyzes the phosphodiester bond on the 5'-side of the AP site with generation of the single-strand DNA break. A decrease in the functional activity of BER enzymes is associated with the increased risk of cardiovascular, neurodegenerative, and oncological diseases. In this work, we developed a fluorescence method for measuring the activity of key human DNA glycosylases and AP endonuclease in cell extracts. The efficacy of fluorescent DNA probes was tested using purified enzymes; the most efficient probes were tested in the enzymatic activity assays in the extracts of A549, MCF7, HeLa, WT-7, HEK293T, and HKC8 cells. The activity of enzymes responsible for the repair of AP sites and removal of uracil and 5,6-dihydrouracil residues was higher in cancer cell lines as compared to the normal HKC8 human kidney cell line.

Characterization of Aspergillus niger endo-1,4-ß-glucanase ENG1 secreted from Saccharomyces cerevisiae using different expression vectors.

Heterologous expression of Aspergillus niger endo-1,4-ß-glucanase (ENG1) in Saccharomyces cerevisiae was tested both with an episomal plasmid vector (YEGAp/eng1) and a yeast vector capable of integration into the HO locus of the S. cerevisiae chromosome (pHO-GAPDH-eng1-KanMX4-HO). In both cases, eng1 gene expression in yeast, with its native signal sequence for secretion, was under the control of the strong glyceraldehyde 3-phosphate dehydrogenase (GAPDH) promoter. We aimed to verify how each expression system affects protein expression, posttranslational modification, and biochemical properties. Expression of eng1 from the episomal plasmid vector YEGAp/eng1 significantly slowed the growth of a yeast cell culture. However, expression of eng1 from the vector integrated into the HO locus of the chromosome did not cause growth suppression, and the enzyme activity in a culture supernatant was maintained throughout the incubation time. ENG1 has optimum catalytic activity at pH 6.0, and is stable in the pH range 5.0-9.0. The enzyme's optimum temperature for catalytic activity at pH 6.0 is 70°C; importantly, more than 95% of the enzyme's initial activity remained after a 2-h incubation at 60°C. The biochemical characterization of ENG1 confirmed the correct expression of the protein and showed that ENG1 expressed by the pHO-GAPDH-eng1-KanMX4-HO vector, in addition to its N-linked sites, is overglycosylated at its O-glycosylation sites compared with ENG1 expressed by the YEGAp/eng1 vector. It is likely that the O-glycosylated form of the A. niger ENG1 retains more stable activity during continuous cultivation of recombinant yeasts than the form that is only N-glycosylated.

Predictive biomarkers for cancer therapy with PARP inhibitors.

Poly(ADP-ribose) polymerase (PARP) inhibitors have raised high expectations for the treatment of multiple malignancies. PARP inhibitors, which can be used as monotherapies or in combination with DNA-damaging agents, are particularly efficient against tumors with defects in DNA repair mechanisms, in particular the homologous recombination pathway, for instance due to BRCA mutations. Thus, deficient DNA repair provides a framework for the success of PARP inhibitors in medical oncology. Here, we review encouraging results obtained in recent clinical trials investigating the safety and efficacy of PARP inhibitors as anticancer agents. We discuss emerging mechanisms of regulation of homologous recombination and how inhibition of DNA repair might be used in cancer therapy. We surmise that the identification of patients that are likely to benefit from PARP inhibition will improve the clinical use of PARP inhibitors in a defined target population. Thus, we will place special emphasis on biomarker discovery.

Kinetic mechanism of the interaction of Saccharomyces cerevisiae AP-endonuclease 1 with DNA substrates.

The apurinic/apyrimidinic endonuclease from Saccharomyces cerevisiae Apn1 is one of the key enzymes involved in base excision repair of DNA lesions. A major function of the enzyme is to cleave the upstream phosphodiester bond of an apurinic/apyrimidinic site (AP-site), leading to the formation of a single-strand break with 3'-hydroxyl (OH) and 5'-deoxyribose phosphate (dRP) termini. In this study, the pre-steady-state kinetics and conformational dynamics of DNA substrates during their interaction with Apn1 were investigated. A stopped-flow method with detection of the fluorescence intensity of 2-aminopurine and pyrrolocytosine located adjacent or opposite to the damage was used. It was found that upon interaction with Apn1, both DNA strands undergo a number of rapid changes. The location of fluorescent analogs of heterocyclic bases in DNA does not influence the catalytic step of the reaction. Comparison of data obtained for yeast Apn1 and reported data (Kanazhevskaya, L. Yu., Koval, V. V., Vorobjev, Yu. N., and Fedorova, O. S. (2012) Biochemistry, 51, 1306-1321) for human Ape1 revealed some differences in their interaction with DNA substrates.

Initiation of 8-oxoguanine base excision repair within trinucleotide tandem repeats.

Trinucleotide repeat expansion provides a molecular basis for several devastating neurodegenerative diseases. In particular, expansion of a CAG run in the human HTT gene causes Huntington's disease. One of the main reasons for triplet repeat expansion in somatic cells is base excision repair (BER), involving damaged base excision and repair DNA synthesis that may be accompanied by expansion of the repaired strand due to formation of noncanonical DNA structures. We have analyzed the kinetics of excision of a ubiquitously found oxidized purine base, 8-oxoguanine (oxoG), by DNA glycosylase OGG1 from the substrates containing a CAG run flanked by AT-rich sequences. The values of k(2) rate constant for the removal of oxoG from triplets in the middle of the run were higher than for oxoG at the flanks of the run. The value of k(3) rate constant dropped starting from the third CAG-triplet in the run and remained stable until the 3'-terminal triplet, where it decreased even more. In nuclear extracts, the profile of oxoG removal rate along the run resembled the profile of k(2) constant, suggesting that the reaction rate in the extracts is limited by base excision. The fully reconstituted BER was efficient with all substrates unless oxoG was near the 3'-flank of the run, interfering with the initiation of the repair. DNA polymerase ß was able to perform a strand-displacement DNA synthesis, which may be important for CAG run expansion initiated by BER.

Kinetic mechanism of human apurinic/apyrimidinic endonuclease action in nucleotide incision repair.

Human major apurinic/apyrimidinic endonuclease (APE1) is a multifunctional enzyme that plays a central role in DNA repair through the base excision repair (BER) pathway. Besides BER, APE1 is involved in an alternative nucleotide incision repair (NIR) pathway that bypasses glycosylases. We have analyzed the conformational dynamics and the kinetic mechanism of APE1 action in the NIR pathway. For this purpose we recorded changes in the intensity of fluorescence of 2-aminopurine located in two different positions in a substrate containing dihydrouridine (DHU) during the interaction of the substrate with the enzyme. The enzyme was found to change its conformation within the complex with substrate and also within the complex with the reaction product, and the release of the enzyme from the complex with the product seemed to be the limiting stage of the enzymatic process. The rate constants of the catalytic cleavage of DHU-containing substrates by APE1 were comparable with the appropriate rate constants for substrates containing apurinic/apyrimidinic site or tetrahydrofuran residue, which suggests that NIR is a biologically important process.

Conformational dynamics of human AP endonuclease in base excision and nucleotide incision repair pathways.

APE1 is a multifunctional enzyme that plays a central role in base excision repair (BER) of DNA. APE1 is also involved in the alternative nucleotide incision repair (NIR) pathway. We present an analysis of conformational dynamics and kinetic mechanisms of the full-length APE1 and truncated NDelta61-APE1 lacking the N-terminal 61 amino acids (REF1 domain) in BER and NIR pathways. The action of both enzyme forms were described by identical kinetic schemes, containing four stages corresponding to formation of the initial enzyme-substrate complex and isomerization of this complex; when a damaged substrate was present, these stages were followed by an irreversible catalytic stage resulting in the formation of the enzyme-product complex and the equilibrium stage of product release. For the first time we showed, that upon binding AP-containing DNA, the APE1 structure underwent conformational changes before the chemical cleavage step. Under BER conditions, the REF1 domain of APE1 influenced the stability of both the enzyme-substrate and enzyme-product complexes, as well as the isomerization rate, but did not affect the rates of initial complex formation or catalysis. Under NIR conditions, the REF1 domain affected both the rate of formation and the stability of the initial complex. In comparison with the full-length protein, NDelta61-APE1 did not display a decrease in NIR activity with a dihydrouracil-containing substrate. BER conditions decrease the rate of catalysis and strongly inhibit the rate of isomerization step for the NIR substrates. Under NIR conditions AP-endonuclease activity is still very efficient.

The HAP1 protein stimulates the turnover of human mismatch-specific thymine-DNA-glycosylase to process 3,N(4)-ethenocytosine residues.

When present in DNA, 3,N(4)-ethenocytosine (epsilon C) residues are potentially mutagenic and carcinogenic in vivo. The enzymatic activity responsible for the repair of the epsilon C residues in human cells is the hTDG protein, the human thymine-DNA-glycosylase that removes thymine in a T/G base pair [Proc. Natl. Acad. Sci., U.S.A., 95 (1998) 8508]. One of the distinctive properties of the hTDG protein is that it remains tightly bound to the AP-site resulting from its glycosylase activity. In this paper we report that the human AP endonuclease, the HAP1 (Ape1, APEX Ref-1) protein, stimulates the processing of epsilon C residues by the hTDG protein in vitro, in a dose-dependent manner. This property of HAP1 protein is specific since E.coli Fpg, Nfo and Nth proteins, all endowed with an AP nicking activity, do not show similar features. The results suggest that the HAP1 protein displaces the hTDG protein bound to the AP-site and therefore increases the turnover of the hTDG protein. However, using a variety of techniques including gel retardation assay, surface plasmon resonance and two-hybrid system, it was not possible to detect evidence for a complex including the substrate, the hTDG and HAP1 proteins.

Clustered DNA damages as dosemeters for ionising radiation exposure and biological responses.

Clustered DNA damages--two or more lesions (oxidised bases. abasic sites, or strand breaks) within a few DNA helical turns on opposing strands--are induced in DNA in solution and in vivo in human cells by ionising radiation. They have been postulated to be difficult to repair, and thus of potentially high biological significance. Since the total of clustered damages produced by ionising radiation is at about 3 to 4 times higher levels than double-strand breaks and are apparently absent in unirradiated cells, levels of clustered damages present immediately alter radiation exposure could serve as sensitive dosemeters of radiation exposure. Since some clusters may not be repairable and may accumulate in cells, they might also be useful as integrating dosemeters of biological effects of radiation damage.

AlkA protein is the third Escherichia coli DNA repair protein excising a ring fragmentation product of thymine.

Various forms of oxidative stress lead to the formation of damaged bases including N-(2-deoxy-beta-D-erythro-pentofuranosyl)-N-3-(2R-hydroxyisobutyric acid)-urea or alphaRT, the fragmentation product of thymine formed from 5R-thymidine C5-hydrate upon hydrolysis. It was shown that alphaRT is excised by Escherichia coli Fpg and Nth proteins. Here we report that when present in DNA, alphaRT is, in addition, a substrate for the E. coli AlkA protein with an apparent K(m) value of congruent with170 nM. alphaRT positioned opposite T, dG, dC, and dA were efficiently excised by AlkA protein from duplex oligodeoxynucleotides in the following order: dA approximately T >> dC approximately dG. This is the first example of the excision of a ring opened form of a pyrimidine by AlkA protein and also the first example where the same DNA base lesion is excised by three different DNA glycosylases of the base excision repair pathway. The present results suggest possible structural similarity of the active site between E. coli AlkA, Fpg, and Nth proteins.

Interactions of the human, rat, Saccharomyces cerevisiae and Escherichia coli 3-methyladenine-DNA glycosylases with DNA containing dIMP residues.

In DNA, the deamination of dAMP generates 2'-deoxy-inosine 5'-monophosphate (dIMP). Hypoxanthine (HX) residues are mutagenic since they give rise to A.T-->G.C transition. They are excised, although with different efficiencies, by an activity of the 3-methyl-adenine (3-meAde)-DNA glycosylases from Escherichia coli (AlkA protein), human cells (ANPG protein), rat cells (APDG protein) and yeast (MAG protein). Comparison of the kinetic constants for the excision of HX residues by the four enzymes shows that the E.coli and yeast enzymes are quite inefficient, whereas for the ANPG and the APDG proteins they repair the HX residues with an efficiency comparable to that of alkylated bases, which are believed to be the primary substrates of these DNA glycosylases. Since the use of various substrates to monitor the activity of HX-DNA glycosylases has generated conflicting results, the efficacy of the four 3-meAde-DNA glycosylases of different origin was compared using three different substrates. Moreover, using oligo-nucleotides containing a single dIMP residue, we investigated a putative sequence specificity of the enzymes involving the bases next to the HX residue. We found up to 2-5-fold difference in the rates of HX excision between the various sequences of the oligonucleotides studied. When the dIMP residue was placed opposite to each of the four bases, a preferential recognition of dI:T over dI:dG, dI:dC and dI:dA mismatches was observed for both human (ANPG) and E.coli (AlkA) proteins. At variance, the yeast MAG protein removed more efficiently HX from a dI:dG over dI:dC, dI:T and dI:dA mismatches.

Enzymology of the repair of etheno adducts in mammalian cells and in Escherichia coli.

Exocyclic adducts are generated in cellular DNA by reaction with epoxides that are formed metabolically from various industrial pollutants and by reaction with activated aldehydes that arise during membrane lipid peroxidation. The etheno (epsilon) derivatives of purine and pyrimidine bases, e.g. 3,N4-ethenocytosine, 1,N6-ethenoadenine, N2,3-ethenoguanine and 1,N2-ethenoguanine, are probably involved in carcinogenesis because they are highly mutagenic and genotoxic. Therefore, the repair processes that eliminate exocyclic adducts from DNA should play a crucial role in maintaining the stability of the genetic information. The DNA glycosylases implicated in the repair of etheno adducts have been identified. Human and Escherichia coli 3-methyladenine-DNA-glycosylases excise 1,N6-ethenoadenine residues. We have identified two homologous proteins present in human cells and E. coli that remove 3,N4-ethenocytosine residues by DNA glycosylase activity. The human enzyme is an activity of the mismatch-specific thymine-DNA glycosylase, while the bacterial enzyme is an activity of the double-stranded uracil-DNA glycosylase, i.e., the homologue of the human enzyme. The fact that 1,N6-ethenoadenine and 3,N4-ethenocytosine are recognized and efficiently excised by DNA glycosylases in vitro suggests that these enzymes may be responsible for the repair of these mutagenic lesions in vivo and may contribute importantly to genetic stability.

3,N4-ethenocytosine, a highly mutagenic adduct, is a primary substrate for Escherichia coli double-stranded uracil-DNA glycosylase and human mismatch-specific thymine-DNA glycosylase.

Exocyclic DNA adducts are generated in cellular DNA by various industrial pollutants such as the carcinogen vinyl chloride and by endogenous products of lipid peroxidation. The etheno derivatives of purine and pyrimidine bases 3,N4-ethenocytosine (epsilonC), 1, N6-ethenoadenine (epsilonA), N2,3-ethenoguanine, and 1, N2-ethenoguanine cause mutations. The epsilonA residues are excised by the human and the Escherichia coli 3-methyladenine-DNA glycosylases (ANPG and AlkA proteins, respectively), but the enzymes repairing epsilonC residues have not yet been described. We have identified two homologous proteins present in human cells and E. coli that remove epsilonC residues by a DNA glycosylase activity. The human enzyme is an activity of the mismatch-specific thymine-DNA glycosylase (hTDG). The bacterial enzyme is the double-stranded uracil-DNA glycosylase (dsUDG) that is the homologue of the hTDG. In addition to uracil and epsilonC-DNA glycosylase activity, the dsUDG protein repairs thymine in a G/T mismatch. The fact that epsilonC is recognized and efficiently excised by the E. coli dsUDG and hTDG proteins in vitro suggests that these enzymes may be responsible for the repair of this mutagenic lesion in vivo and be important contributors to genetic stability.

Antimutagenic role of base-excision repair enzymes upon free radical-induced DNA damage.

As a consequence of oxidative stress, reactive oxygen species are generated in the cells. They interact with DNA and induce various modifications. Among them, oxidised purines (such as C8-oxoguanine and purines whose imidazole ring is opened), oxidised pyrimidines (such as thymine and cytosine glycols, ring saturated and fragmented pyrimidines), ethenobases and hypoxanthine. These various lesions have either miscoding properties or are blocks for DNA and RNA polymerases during replication and transcription, respectively. Most of these lesions are repaired by the base excision pathway in which the first step is mediated by specific DNA glycosylases. We review the various glycosylases involved in the repair of oxidised bases in Escherichia coli. The Fpg protein (formamidopyrimidine-DNA glycosylase) contains a zinc finger and excises oxidised purines whereas the Nth protein excises oxidised pyrimidines. The Nei protein excises a comparable spectra of pyrimidines and is believed to act as a back up enzyme to the Nth protein. The hypoxanthine-DNA glycosylase excises hypoxanthine residue and is one of the various activities of the AlkA protein (including formyluracil and ethenopurines residues). The Nfo protein was shown to have a novel activity that incises 5' to an alpha-deoxyadenosine residue (the anomer of deoxyadenosine formed by gamma-irradiation). The mechanism of action of the Fpg and Nth proteins are discussed. The properties of the human counterpart of the Fpg and Nth proteins the hNth and OGG1 proteins, respectively are also reviewed.

The ring fragmentation product of thymidine C5-hydrate when present in DNA is repaired by the Escherichia coli Fpg and Nth proteins.

Various forms of oxidative stress, including gamma-radiolysis and UV irradiation, result in the formation of damaged bases. (5R)-Thymidine C5-hydrate is one of several modified nucleosides produced from thymidine under these conditions. N-(2-Deoxy-beta-D-erythro-pentofuranosyl)-N-3-[(2R)-hydroxyisobutyric acid]urea or alphaRT is the respective fragmentation product formed from (5R)-thymidine C5-hydrate upon hydrolysis. This modified nucleoside has potential mutagenic or lethal properties. No enzymatic activity responsible for the removal of alphaRT has been identified. We report here that when present in DNA, alphaRT is a substrate for two purified enzymes from Escherichia coli involved in the repair of oxidized bases: the Nth and the Fpg proteins. The Fpg protein removes the alphaRT lesion more efficiently than the Nth protein. This is the first example of efficient excision of a ring-opened form of a pyrimidine by the Fpg protein. The high efficacy of the Fpg protein suggests that it is likely to be involved in vivo in the excision of alphaRT. The kinetics of the reaction of the Fpg protein with DNA containing alphaRT suggest substrate inhibition. Duplex oligodeoxynucleotides containing alphaRT positioned opposite T, dG, dC, and dA were cleaved efficiently by both enzymes, although the profiles of activity of the two enzymes were different. The Nth enzyme preferentially excises alphaRT when opposite a dG, followed by alphaRT.dA, alphaRT. T, and alphaRT.dC. For the Fpg protein, the order is alphaRT.dC >/= alphaRT.dG approximately alphaRT.T > alphaRT.dA. Moreover, we show that human cell extract exhibits an activity that excises alphaRT from an oligonucleotide, suggesting that human homologues of the Nth and/or Fpg proteins could be involved in repair of this lesion in human cells.

Effects of nitrous acid treatment on the survival and mutagenesis of Escherichia coli cells lacking base excision repair (hypoxanthine-DNA glycosylase-ALK A protein) and/or nucleotide excision repair.

Deoxyinosine occurs in DNA by spontaneous deamination of adenine or by incorporation of dITP during replication. Hypoxanthine residues (HX) are mutagenic and give rise to A-T-->G-C transition. They are substrates for the Escherichia coli product of the alkA gene, the 3-methyl-adenine-DNA glycosylase II (ALK A protein). In mammalian cells and in yeast, HX is excised by the counterpart of ALK A protein, the ANPG or the MAG proteins respectively. We have investigated in vivo the contribution of the alkA gene to counteract the lethal and/or mutagenic effects of HX residues induced by nitrous acid treatment. Using an E.coli strain allowing the detection of A-T-->G-C transition, we show that the alkA mutant has a slightly increased spontaneous rate of mutation and about the same sensitivity when treated with HNO2 as compared with the wild-type strain. Using the E.coli alkA mutant carrying a multicopy plasmid expressing the ALK A protein or the ANPG protein, we barely observe any effect of HNO2 treatment on sensitivity and mutation rate of the bacteria. In contrast, the same experiment performed with a uvrA- strain, deficient in nucleotide excision repair (NER), shows that this mutant is extremely sensitive to HNO2 treatment. Furthermore, the sensitivity and the spontaneous mutation rate observed in the double mutant alkA- uvrA- are almost identical to those of the uvrA- mutant. Hence, NER has the major role in vivo for the repair of lethal and mutagenic lesions induced by HNO2.

Dimerization of plasmid DNA accelerates selection for antibiotic resistance.

Dimerization of multicopy plasmids is widely assumed to be disadvantageous both for plasmid maintenance and for the host cell. It is known that dimerization causes plasmid instability; dimer-containing cells grow slower than their monomer-containing counterparts. However, as we demonstrate here, under conditions of selective stress, dimers provide an advantage for bacteria. Dimers facilitate segregation of mutants from numerous copies of the parental plasmid. Accelerated segregation greatly increases the rate of accumulation of plasmids carrying mutations that are adaptive for bacteria. In contrast, resolution of dimers by site-specific recombination decreases, 10(3)-10(5)-fold, the efficiency of selection of spontaneous reversions in the tet gene of pBR327.

Requirement of mismatch repair genes MSH2 and MSH3 in the RAD1-RAD10 pathway of mitotic recombination in Saccharomyces cerevisiae.

The RAD1 and RAD10 genes of Saccharomyces cerevisiae are required for nucleotide excision repair and they also act in mitotic recombination. The Rad1-Rad10 complex has a single-stranded DNA endonuclease activity. Here, we show that the mismatch repair genes MSH2 and MSH3 function in mitotic recombination. For both his3 and his4 duplications, and for homologous integration of a linear DNA fragment into the genome, the msh3 delta mutation has an effect on recombination similar to that of the rad1 delta and rad10 delta mutations. The msh2 delta mutation also reduces the rate of recombination of the his3 duplication and lowers the incidence of homologous integration of a linear DNA fragment. Epistasis analyses indicate that MSH2 and MSH3 function in the RAD1-RAD10 recombination pathway, and studies presented here suggest an involvement of the RAD1-RAD10 pathway in reciprocal recombination. The possible roles of Msh2, Msh3, Rad1, and Rad10 proteins in genetic recombination are discussed. Coupling of mismatch binding proteins with the recombinational machinery could be important for ensuring genetic fidelity in the recombination process.

Escherichia coli, Saccharomyces cerevisiae, rat and human 3-methyladenine DNA glycosylases repair 1,N6-ethenoadenine when present in DNA.

The human carcinogen vinyl chloride is metabolized in the liver to reactive intermediates which generate various ethenobases in DNA. It has been reported that 1,N6-ethenoadenine (epsilon A) is excised by a DNA glycosylase present in human cell extracts, whereas protein extracts from Escherichia coli and yeast were devoid of such an activity. We confirm that the human 3-methyladenine-DNA glycosylase (ANPG protein) excises epsilon A residues. This finding was extended to the rat (ADPG protein). We show, at variance with the previous report, that pure E.coli 3-methyladenine-DNA glycosylase II (AlkA protein) as well as its yeast counterpart, the MAG protein, excise epsilon A from double stranded oligodeoxynucleotides that contain a single epsilon A. Both enzymes act as DNA glycosylases. The full length and the truncated human (ANPG 70 and 40 proteins, respectively) and the rat (ADPG protein) 3-methyladenine-DNA glycosylases activities towards epsilon A are 2-3 orders of magnitude more efficient than the E.coli or yeast enzyme for the removal of epsilon A. The Km of the various proteins were measured. They are 24, 200 and 800 nM for the ANPG, MAG and AlkA proteins respectively. These three proteins efficiently cleave duplex oligonucleotides containing epsilon A positioned opposite T, G, C or epsilon A. However the MAG protein excises A opposite cytosine much faster than opposite thymine, guanine or adenine.

Excision of hypoxanthine from DNA containing dIMP residues by the Escherichia coli, yeast, rat, and human alkylpurine DNA glycosylases.

The deamination of adenine residues in DNA generates hypoxanthine, which is mutagenic since it gives rise to an A.T to G.C transition. Hypoxanthine is removed by hypoxanthine DNA glycosylase activity present in Escherichia coli and mammalian cells. Using polydeoxyribonucleotides or double-stranded synthetic oligonucleotides that contain dIMP residues, we show that this activity in E. coli is associated with the 3-methyladenine DNA glycosylase II coded for by the alkA gene. This conclusion is based on the following facts: (i) the two enzymatic activities have the same chromatographic behavior on various supports and they have the same molecular weight, (ii) both are induced during the adaptive response, (iii) a multicopy plasmid bearing the alkA gene overproduces both activities, (iv) homogeneous preparation of AlkA has both enzymatic activities, (v) the E. coli alkA- mutant does not show any detectable hypoxanthine DNA glycosylase activity. Under the same experimental conditions, but using different substrates, the same amount of AlkA protein liberates 1 pmol of 3-methyladenine from alkylated DNA and 1.2 fmol of hypoxanthine from dIMP-containing DNA. The Km for the latter substrate is 420 x 10(-9) M as compared to 5 x 10(-9) M for alkylated DNA. Hypoxanthine is released as a free base during the reaction. Duplex oligodeoxynucleotides containing hypoxanthine positioned opposite T, G, C, and A were cleaved efficiently. ANPG protein, APDG protein, and MAG protein--the 3-methyladenine DNA glycosylases of human, rat, and yeast origin, respectively--were also able to release hypoxanthine from various DNA substrates containing dIMP residues. The mammalian enzyme is by far the most efficient hypoxanthine DNA glycosylase of all the enzymes tested.