|Home | About | Journals | Submit | Contact Us | Français|
The causative agent of dental caries in humans, Streptococcus mutans, outcompetes other bacterial species in the oral cavity and causes disease by surviving acidic conditions in dental plaque. We have previously reported that the low-pH survival strategy of S. mutans includes the ability to induce a DNA repair system that appears to involve an enzyme with exonuclease functions (K. Hahn, R. C. Faustoferri, and R. G. Quivey, Jr., Mol. Microbiol 31:1489-1498, 1999). Here, we report overexpression of the S. mutans apurinic/apyrimidinic (AP) endonuclease, Smx, in Escherichia coli; initial characterization of its enzymatic activity; and analysis of an smx mutant strain of S. mutans. Insertional inactivation of the smx gene eliminates the low-pH-inducible exonuclease activity previously reported. In addition, loss of Smx activity renders the mutant strain sensitive to hydrogen peroxide treatment but relatively unaffected by acid-mediated damage or near-UV irradiation. The smx strain of S. mutans was highly sensitive to the combination of iron and hydrogen peroxide, indicating the likely production of hydroxyl radical by Fenton chemistry with concomitant formation of AP sites that are normally processed by the wild-type allele. Smx activity was sufficiently expressed in E. coli to protect an xth mutant strain from the effects of hydrogen peroxide treatment. The data indicate that S. mutans expresses an inducible, class II-like AP endonuclease, encoded by the smx gene, that exhibits exonucleolytic activity and is regulated as part of the acid-adaptive response of the organism. Smx is likely the primary, if not the sole, AP endonuclease induced during growth at low pH values.
Streptococcus mutans inhabits dental plaque, where, as a causative agent of dental caries, it creates and survives in an acidic milieu (20). Acidification of plaque is a direct result of the secretion of organic acids, by-products of carbohydrate metabolism by S. mutans and other bacterial inhabitants of the oral cavity. The survival of the microbe in the oral cavity is predicated on its ability to elaborate an acid-adaptive response (30). Since pH values in dental plaque have been reported to be as low as 4.0, the intracellular environment is potentially exposed to acidic conditions (11, 12, 15), which likely leads to an increased formation of abasic sites in DNA (19). This idea is indirectly supported by our previous observation that the process of acid adaptation in S. mutans also includes expression of a low-pH-inducible apurinic/apyrimidinic (AP) endonuclease (10, 27). These studies suggested that the AP endonuclease activity in S. mutans more closely resembles exonuclease III (Exo III) of Escherichia coli than endonuclease IV (10).
Based on this hypothesis, we expected that the gene encoding the AP endonuclease activity in S. mutans would show homology to exoA from Streptococcus pneumoniae and xth from E. coli. In the present study, we have identified and cloned a homologue of the E. coli class II AP endonuclease, exonuclease III, which we have named smx. The transcriptional start site was determined, and putative promoter elements in the 5′ untranslated region were identified. Insertional inactivation of the smx gene resulted in loss of the low-pH-inducible AP endonuclease activity previously reported (10).
Physiological studies of the smx mutant strain were undertaken to determine the magnitude of the contribution made by this inducible AP endonuclease to S. mutans and to determine whether additional AP site-cleaving enzymes were induced during growth at low pH values. Additionally, we expressed the S. mutans Smx protein in E. coli and found that activity of the purified, recombinant protein is similar to that of Exo III.
The bacterial strains used in this study are listed in Table Table1.1. E. coli strain BW9109, with the Exo III gene (xth) deleted, was the gift of Bernard Weiss (University of Michigan).
Streptococcal strains included S. mutans GS-5 (9), the S. mutans wild-type strain UA159 (23), and its recA derivative strain UR100 (26). S. mutans strains were maintained on brain heart infusion (BHI) medium plus agar (Difco Laboratories, Detroit, Mich.) supplemented, where appropriate, with antibiotics at the following concentrations: for erythromycin, 5 μg/ml, and for spectinomycin, 1,000 μg/ml. Batch cultures were incubated overnight in BHI medium at 37°C, in a 5% (vol/vol) CO2-enriched atmosphere. Steady-state cultures were grown in a BioFlo 2000 fermentor (New Brunswick, Edison, N.J.) on TY medium at a dilution rate of 0.24 h−1 as previously described (8, 10). Cells were limited by glucose (2.3 mM), and pH maintenance was achieved by the addition of 2 N KOH. Steady-state cells used in these studies were harvested after a minimum of 10 generations at each pH value.
E. coli DH10B (Invitrogen, Carlsbad, Calif.) and E. coli M15(pREP4) (QIAGEN, Valencia, Calif.) were used for cloning experiments as indicated below and transformed by electroporation. E. coli strains were grown on LB medium (33), and the following selective antibiotics were added, where needed: ampicillin (100 μg/ml), kanamycin (25 μg/ml), spectinomycin (100 μg/ml), and erythromycin (500 μg/ml). 5-Bromo-4-chloro-3-indolyl-β-d-galactosidase (X-Gal) was used at a final concentration of 40 μg/ml, and isopropyl β-d-thiogalactopyranoside (IPTG) was added to cultures at a final concentration of 1 mM to induce the expression of Smx.
Early work in cloning the smx gene was performed using the degenerate primers ExoFwd and DegExoRev (Table (Table2);2); sequences were based on the deduced amino acid sequences of the E. coli xth gene (32) and the Streptococcus pneumoniae exoA gene (25). The cloned smx amplicon, contained on plasmid pSMexo9, was used to probe a Southern blot of restriction enzyme-digested S. mutans UA159 genomic DNA. Subgenomic libraries of fragments hybridizing to the smx probe were created in pUC19 (41). E. coli DH10B transformants containing these constructs were selected on LB agar supplemented with ampicillin and X-Gal. Putative smx clones were screened by colony hybridization as previously described (34) and yielded a clone, pKHsmn4, containing the full-length smx gene flanked by 220 bp 5′ to the translational start codon and 280 bp 3′ to the stop codon. A partial open reading frame that contained a homologue to the E. coli nth gene, which encodes endonuclease III, was found in the 3′ flanking region.
RNA was extracted from overnight batch cultures of S. mutans UA159 with a QIAGEN RNeasy mini kit with modifications, as previously described (17). The Promega (Madison, Wis.) avian myeloblastosis virus reverse transcriptase primer extension system was used to label an oligonucleotide primer, SmnPE3 (Table (Table2),2), with [γ-32P]ATP (6,000 Ci/mmol) (Perkin Elmer, Boston, Mass.). The primer extension reaction mixture consisted of 58 μg of total RNA and 20 pmol of labeled primer. SmnPE3 was located 16 nucleotides upstream of the translational start codon of the smx gene.
The smx gene was PCR amplified from S. mutans UA159 chromosomal DNA with Pfu DNA polymerase (Stratagene, La Jolla, Calif.) and the oligonucleotide primer pair 5′ BamHI Smn and 3′ SalI Smn (Table (Table2).2). Amplified fragments were digested and cloned into plasmid pQE30 (QIAGEN). Transformants of E. coli DH10B were selected on LB agar containing ampicillin and screened for the presence of the smx gene by PCR. Nucleotide sequence determination was used to verify the appropriate construction of in-frame clones. One such clone was termed pKHsmnExpressI and was retained for use in producing recombinant Smx protein. E. coli M15, a strain carrying the pREP4 repressor plasmid for regulation of protein expression, was transformed with pKHsmnExpressI, and colonies were selected on LB medium containing ampicillin and kanamycin. Production of recombinant, His-tagged Smx from the E. coli M15 strain was performed according to the manufacturer's directions for native purification with Ni-nitrilotriacetic acid (NTA) agarose (QIAGEN). Fractions collected from each step in the protocol were analyzed on a sodium dodecyl sulfate (SDS)-12% polyacrylamide gel electrophoresis (PAGE) gel and stained with Coomassie brilliant blue (39) to monitor purification of the recombinant protein.
Fractions were tested for AP endonuclease activity in reaction mixtures containing 5 mM CaCl2; 66 mM Tris-HCl (pH 8.0); 8 fmol of an end-labeled, double-stranded tetrahydrofuran (THF)-containing substrate; and equivalent volumes of column fractions (1 μl). A THF-containing, double-stranded-DNA duplex, our model substrate for studying AP endonuclease activity, was prepared as previously described by annealing the 17-mer THF and 17-mer comp oligonucleotides (Table (Table2)2) (10). Reactions were carried out at 37°C for 15 min and terminated by the addition of stop solution (98% formamide, 10 mM EDTA, 0.025% bromophenol blue, and 0.025% xylene cyanol). Cleavage products were separated on 20% denaturing polyacrylamide gels. Lysates prepared from E. coli M15 cells containing the pQE30 vector alone served as a negative control to ensure that any activity observed was the result of recombinant Smx protein and not due to host cell factors.
THF-containing DNA substrates were prepared as described above. Reaction mixtures contained 66 mM Tris-HCl (pH 8.0), 5 mM MgCl2, 10 fmol of double-stranded THF-containing DNA substrate, and decreasing amounts of S. mutans Smx protein or E. coli Exo III (Invitrogen). Protein concentrations were determined using the Bio-Rad protein assay dye reagent (Bio-Rad, Hercules, Calif.), and 100 ng of each enzyme was 10-fold serially diluted to 20 fg with reaction buffer (66 mM Tris-HCl [pH 8.0], 5 mM MgCl2). Reactions were carried out at 37°C for 10 min. Termination of reactions and separation of cleavage products were performed as described above.
Complementation assays were based on a previously published procedure (4). E. coli strains DH10B (endA recA) and BW9109 (an xth-deficient strain) were transformed via electroporation with plasmids pKHsmn4 and pSMexoΔSpec17. The resulting strains were grown in LB medium supplemented with ampicillin and spectinomycin, respectively. E. coli strains carrying plasmids containing either an intact or an insertionally disrupted smx gene were grown as 5-ml overnight cultures and diluted 1:100 in 50 ml of fresh, prewarmed LB medium plus antibiotic. Following growth to mid-logarithmic phase (optical density at 600 nm, approximately 0.6), 3-ml culture samples were transferred to polypropylene tubes and 0.1 ml was removed for the zero time interval. The remaining sample was brought to 0.2% (vol/vol) hydrogen peroxide and held at 30°C. Aliquots were removed at 5, 10, and 15 min; serially diluted into LB medium; and plated on LB agar plates containing antibiotic. Viable cells were enumerated and used to calculate log(N/N0).
A SmaI site was introduced at position 231 of the 550-bp cloned fragment, contained on pSMexo9, by splice overlap extension PCR (13, 14). Briefly, the primer pairs SOEExoF-ExoSOERev and ExoSOEFwd-SOEExoR (Table (Table2)2) were used in independent PCRs with plasmid pSMexo9 DNA. The resultant amplicons were spliced together in a subsequent amplification reaction with the outside primer pair ExoSOEFwd-ExoSOERev. The reengineered smx gene fragment was cloned into pGEM-T and termed pSMexo9/SOE15. A spectinomycin resistance gene (Spr marker) from pGEM-Spc (3) was ligated into the SmaI site in pSMexo9/SOE15 and used to insertionally inactivate the smx gene. The resulting plasmid carrying the disrupted smx gene, termed pSMexoΔSpec17, was used to transform S. mutans UA159 by published procedures (24). Strains arising from the transformations were screened by Southern hybridization to verify the appropriate constructions (data not shown). One such strain, designated S. mutans UR101, contained the smx coding sequence interrupted by the spectinomycin antibiotic cassette.
A recA smx double mutant strain was created in the UR101 (smx) background with plasmid pRQ202, containing an insertionally disrupted recA gene fragment (26). Genomic DNAs were prepared from Spr Ermr transformants, and strains were confirmed by Southern hybridization (data not shown). One such strain, termed S. mutans UR102, was selected and used in further characterizations as part of this study.
Crude protein extracts were prepared as previously described from S. mutans cells grown at steady-state pH values of 5 and 7 (10). The lysate was dialyzed against 10 mM Tris-HCl (pH 7.0)-1 mM EDTA. Phenylmethylsulfonyl fluoride was added to a final concentration of 0.6 mM, and extracts were stored at −70°C. Total protein concentrations were determined with Bio-Rad (Hercules, Calif.) protein assay dye reagent. Crude cell extracts were assayed for AP endonuclease activity with the THF-containing, double-stranded-DNA duplex as the substrate. Reactions were performed in 5 mM CaCl2-66 mM Tris-HCl (pH 8.0)-25 mM NaCl-0.5 mM EDTA and contained 25 μg of total protein and 10 fmol of 5′-end-radiolabeled substrate. Reaction mixtures were incubated at 37°C for 30 min. Termination of the reactions and separation of cleavage products were performed as described above.
Samples were taken from steady-state cultures of S. mutans UA159 (wild type) and UR101 (smx mutant strain) for comparison of sensitivities to acid, hydrogen peroxide, and UV light irradiation, as described previously (27). Briefly, strains were tested for sensitivity to acid-mediated killing by harvesting samples from the chemostat, resuspending cell pellets in 0.1 M glycine (pH 2.5), and stirred at room temperature. Aliquots were removed at 0-, 15-, 30-, and 60-min intervals; serially diluted; and plated on solid BHI medium. For near-UV irradiation survival assays, cells were removed from the chemostat vessel and placed into plastic petri dishes. The uncovered dishes were placed under a UV light source (Stratalinker; Stratagene) for various lengths of time (0, 0.5, 1, 2, 5, and 10 min), serially diluted, and plated on BHI agar plates. Hydrogen peroxide sensitivity assays were also performed on the cells. Briefly, samples harvested from the chemostat were resuspended in BHI medium, and hydrogen peroxide was added to a final concentration of 0.2%. Aliquots (0.1 ml) were removed at 0, 15, 30, and 60 min; serially diluted; and plated on BHI agar. Viable cells from each condition were counted and used to calculate log(N/N0).
Overnight cultures in BHI medium were also used to determine the sensitivity of S. mutans wild-type, recA, smx, and recA smx strains to ferrous iron alone or in combination with hydrogen peroxide. Harvested cells were treated as described above for the peroxide assay conditions with the addition of 0.2% hydrogen peroxide, 10 mM FeCl2 (ferrous iron), or a combination of both agents, essentially as described previously (5).
The partial open reading frame that contained a homologue to the E. coli nth gene was deposited at GenBank and can be located using the accession number AF233280. The sequence of the smx gene has been deposited at GenBank under accession number AF233280.
A 550-bp gene fragment PCR amplified from S. mutans with degenerate primers served as a probe to identify a full-length clone of the AP endonuclease gene. The gene spanned 828 coding nucleotides and was designated smx for S. mutans Exo III homologue. Alignment of the deduced amino acid sequence of Smx revealed 80% identity and 90% similarity to S. pneumoniae exonuclease A, 25% identity and 39% similarity to E. coli exonuclease III, and 36% identity and 52% similarity to the human AP1 enzyme (Fig. (Fig.1).1). Downstream of smx, an open reading frame with a strong similarity (58%) to E. coli endonuclease III was identified (2). Subsequently, genomic information became available from the S. mutans UA159 genome sequencing project (1) which confirmed this finding and extended our knowledge to include the entire coding region for the endonuclease III homologue (Fig. (Fig.2),2), which appears to be transcribed convergently with the smx gene.
Results from primer extension analysis revealed that mRNA synthesis from the smx gene was initiated at an A residue, in the coding strand, 140 bp upstream from the translational start codon (Fig. (Fig.2).2). The presence of a single extension product and Northern analysis (data not shown) indicate that smx is transcribed as a single message. A putative ribosomal binding site (Shine-Dalgarno sequence) was located approximately 20 bp upstream of the translational start codon. A −10 promoter sequence was identified, and the corresponding −35 sequence shared 50% identity with the E. coli consensus sequence (TTGACA) (Fig. (Fig.22).
Purification of recombinant Smx was facilitated by subcloning the smx gene into the His tag expression vector pQE30. Nickel affinity column-based purification rendered Smx approximately 90 to 95% pure, as assessed by SDS-PAGE (Fig. (Fig.3A).3A). The apparent molecular mass of the fusion protein, which included a tag length of six histidine residues, was estimated to be 33 kDa. Fractions eluted from Ni-NTA affinity columns were assayed for endonuclease activity by measuring the conversion of a double-stranded-DNA substrate containing an abasic site (THF) into its corresponding cleavage products. Two experiments were performed: nickel column elution of the recombinant Smx protein produced in E. coli (Fig. (Fig.3B,3B, lanes 3 to 7) and elution of extracts from E. coli containing vector pQE30 alone (Fig. (Fig.3B,3B, lanes 8 to 12). The combined effects of recombinant Smx protein and endogenous nucleolytic activities were seen in the loading and wash fractions (Fig. (Fig.3B,3B, lanes 3 to 5), whereas the endogenous activities were seen in the corresponding lanes in the control experiment (Fig. (Fig.3B,3B, lanes 8 to 10). Removal of endogenous activities can be seen by comparing the cleavage products in extracts containing the Smx expression construct (Fig. (Fig.3B,3B, lanes 6 and 7) to the marked reduction of cleavage products in control extracts containing the pQE30 plasmid alone (Fig. (Fig.3B,3B, lanes 11 and 12). The lack of cleavage products in the final elution fraction from control extracts (Fig. (Fig.3B,3B, lane 12) indicated that the activity seen in the expression extracts likely represented Smx activity alone (Fig. (Fig.3B,3B, lane 7).
The AP endonuclease activity of the purified, recombinant protein was demonstrated using an enzyme dilution assay for comparison to commercially available E. coli Exo III (Fig. (Fig.4).4). The activities of Smx and Exo III were measured using a 17-bp DNA duplex substrate containing a model AP site (THF) in the presence of Mg2+ (Fig. (Fig.4).4). The results of the enzymatic assay indicated that the activities of the two enzymes are quite similar though not identical. Both enzymes are capable of cleaving 5′ of abasic sites, and both are also able to remove additional nucleotides at the cleavage site. However, we have observed that Exo III activity results in 5- and 6-bp products that are more abundant than the predominantly 7-bp product formed by Smx (Fig. (Fig.4).4). These findings suggest that there may be differences in the mechanisms of the two enzymes.
Based on the homology of the Smx and Exo III protein sequences and the similarity of the activities of the two enzymes, we asked whether Smx would be able to complement a xth deficiency in E. coli. Plasmid-borne copies of the smx gene were introduced into E. coli BW9109, deficient in xth, and DH10B, deficient in recA and endA. Hydrogen peroxide treatment of the strains showed that sensitivity to oxidative attack was substantially alleviated in the case of the xth-deficient strain bearing the smx gene from S. mutans (Fig. (Fig.5A).5A). Control experiments with the inactivated smx gene resulted in complete loss of complementing ability in either of the E. coli strains (Fig. (Fig.5),5), further underscoring the role of the smx gene in protection of the cells from oxidative damage. We observed that strain DH10B was not as inherently sensitive to peroxide as BW9109 (Fig. (Fig.5B),5B), probably owing to its wild-type xth gene. Nevertheless, the presence of an intact smx gene also provided protection to strain DH10B from hydrogen peroxide attack.
We determined the contribution of smx to AP endonuclease activity in S. mutans by using an insertionally inactivated smx gene in strain S. mutans UR101. This strain was characterized for its ability to cleave duplex DNA containing a model abasic site, in this case, THF. Protein extracts, prepared from steady-state cultures of wild-type and smx mutant strains, were used to degrade a radiolabeled 17-bp duplex containing THF at position 9. Reaction mixtures containing extracts from the wild-type strain displayed the expected cleavage products. However, a dramatic reduction in product formation was observed in samples containing extracts from the smx mutant strain (Fig. (Fig.6,6, lanes 4 and 5). Our previous studies indicated that exonuclease activity increased during growth at low pH, specifically pH 5 (10). The results of the digestions showed, as predicted, that extracts from the smx mutant strain grown at pH 5 were no longer able to cleave the model substrate. Enzymatic activity in extracts from pH 7-grown smx-deficient cultures was greatly diminished compared to that in the wild type (compare Fig. Fig.6,6, lanes 4 and 2, respectively). Interestingly, minute levels of activity were still observed, indicating that there may be other endonucleases present in the organism (see Discussion). Nevertheless, the absence of cleavage products in reaction mixtures with the extracts from pH 5-grown cultures of the smx-deficient strain suggested that smx is the main component of abasic-site DNA repair in the acid-adaptive response of S. mutans.
It has been well established that mutations in xth, the gene encoding Exo III of E. coli (40), result in hypersensitivity to hydrogen peroxide treatment and near-UV irradiation (4). We have shown that the expression of Smx is related to acid adaptation in a RecA-independent manner. Adapted cells are protected from insults such as near-UV irradiation, acid-mediated damage, and hydrogen peroxide treatment, unlike their unadapted counterparts (10, 27). Here, we were interested in examining whether a mutation in smx affected the ability of S. mutans to resist DNA-damaging agents following adaptation. The results showed that the smx-deficient strain was more sensitive to hydrogen peroxide treatment than wild-type cultures (Fig. (Fig.7B).7B). The loss of smx had relatively modest, if any, effect on the ability of adapted (pH 5 growth) or unadapted (pH 7 growth) cells to resist acid-mediated killing (Fig. (Fig.7A).7A). The small difference in acid sensitivity between the wild-type and smx mutant strains suggests that the organism may respond to acid-mediated damage and attack by oxidative agents through different mechanisms. In fact, this possibility has been previously suggested from work involving differential expression patterns in two-dimensional gels (35). Finally, loss of smx had modest, if any, effect on sensitivity to near-UV irradiation, considerably less than that seen for acid- or hydrogen peroxide-mediated damage (Fig. (Fig.7C).7C). The lack of sensitivity to near-UV irradiation was unexpected, since this characteristic had been reported to occur in an xth-deficient strain of E. coli (4); however, in general, we have found that S. mutans is somewhat less sensitive to near-UV irradiation than E. coli (29).
Previous work has shown that S. mutans is extremely sensitive to killing by ferrous iron and that this killing is enhanced in the presence of peroxides (5). The sensitivity of the organism to iron had been observed in the absence of oxygen, suggesting that reactive oxygen species were not necessarily involved in toxicity. Here, we have asked whether Fe2+ plays a role in DNA damage and whether the smx mutation sensitizes cells to the presence of iron. For these experiments, we used batch-grown cells, both for convenience and because it has been shown that the organism becomes acid adapted under the conditions described (22). Batch-grown cultures were assayed for their ability to survive the presence of hydrogen peroxide alone to establish baseline sensitivity. Exposure of wild-type, smx, recA, and recA smx strains to hydrogen peroxide again showed very clearly that strains containing the smx mutation were several orders of magnitude more sensitive to H2O2 than the wild type (Fig. (Fig.8A).8A). Loss of RecA apparently played no role in the sensitivity of S. mutans to hydrogen peroxide, supporting our earlier observations that RecA-dependent processes are evidently not involved in protection of S. mutans from oxidative damage (27). Further, the difference in H2O2 sensitivity between the smx and wild-type strains was much more pronounced in batch-grown cultures than in steady-state cultures (compare Fig. Fig.8A8A and and7B,7B, respectively).
Treatment of the strains with 10 mM ferrous chloride showed that sensitivity to iron alone was not traceable to the loss of Smx or RecA activity, in that all strains, regardless of DNA repair genotype, were adversely affected by the presence of iron (Fig. (Fig.8B).8B). Indeed, the kinetics of killing were quite severe, but cells were still viable at significant numbers at the 60-min time point. However, addition of Fe2+ in the presence of hydrogen peroxide showed very different survival kinetics. Viable cells were not recovered from beyond the 30-min time point in strains containing DNA repair mutations, indicating that the combination of agents was quite lethal for S. mutans, particularly in the absence of the Smx protein (Fig. (Fig.8C8C).
The role of S. mutans in the development of human dental caries has been well defined (20). Many studies, using a variety of biochemical approaches, have identified several properties of S. mutans that contribute to the cariogenic potential of this organism, including aciduricity and acidogenicity. The organism is able to persist in the oral cavity by adapting to environmental stresses, particularly the acidification of its milieu (18, 30). It has been established that the F-ATPase is transcriptionally up-regulated during growth at pH values below 7 (16, 17). Recently, we have shown that the membrane fatty acid composition of S. mutans changes dramatically during growth at low pH values (7, 8, 28). In an effort to contribute to the characterization of additional factors involved with the acid adaptation process, we have looked to identify proteins that allow S. mutans to survive and grow in an acidified environment. Previously, we identified an AP endonuclease activity that is induced when S. mutans is grown at acidic pH values (10). The work presented here describes the initial characterization of an S. mutans AP endonuclease activity, as well as the role of that enzyme, Smx, in the physiology of the organism.
Based on our earlier work, we presumed that the endonuclease activity observed in cell extracts was similar to Exo III. We identified a gene, termed smx, that possessed a high degree of similarity to the E. coli xth and S. pneumoniae exoA genes. On the basis of this amino acid identity, the AP endonuclease of S. mutans identified here is likely another member of the family of class II AP endonucleases. Promoter analysis of the 5′ untranslated region of smx, determination of the transcriptional start site, and Northern analysis (data not shown) reveal that the smx gene is transcribed as a monocistronic mRNA.
To determine whether the similarities between Smx and the other class II endonucleases extends to biochemical functionality, recombinant Smx protein was expressed and utilized in a cleavage assay with a model abasic substrate. Indeed, the gene cloned and expressed in this study encoded the enzymatic activity that we reported previously. Purified Smx protein was able to catalyze the cleavage of a THF-containing duplex in a manner similar to that of the other class II enzymes (specifically, Exo III), that is, 5′ to the abasic residue. However, Smx differs from Exo III in its efficiency of removal of additional bases. These results indicate that Smx may catalyze auxiliary functions and/or that the kinetic profile of this enzyme may be different from that of Exo III. Studies are under way in the laboratory to extend our knowledge of the biochemical characteristics of Smx.
A hallmark phenotype of E. coli xth mutant strains is peroxide sensitivity (4), which proved to be an effective means of determining whether the S. mutans smx gene would be able to complement an E. coli xth-deficient strain. In fact, smx expressed from a plasmid was able to complement the E. coli xth-deficient strain by alleviating sensitivity to peroxide attack. These findings further demonstrate similar functionalities of the two enzymes. In addition, the smx gene was able to complement an endA recA mutant strain of E. coli. Taken together, these data suggest the possibility that Smx is able to affect other types of DNA repair in S. mutans.
Creation of an smx mutant strain resulted in the inability of cell extracts from pH 5-grown cultures to cleave a THF-containing, model abasic substrate. Some residual AP endonuclease-like activity is still, however, detected in cell extracts from pH 7-grown cultures. We had demonstrated the low likelihood that an endonuclease IV-like activity exists in S. mutans (10), and a search of the UA159 genome database (1) was unsuccessful in locating deduced amino acid sequences with sufficient similarity to endonuclease IV from E. coli. A candidate to explain the low levels of AP endonuclease activity in the pH 7 extracts is the endonuclease III homologue, which we have located immediately downstream of the smx gene (Fig. (Fig.2).2). Experiments are under way to investigate the role, if any, that this putative nth-like gene might play in the acid base physiology of S. mutans.
The smx-deficient strain UR101 was characterized with respect to sensitivity to environmental and DNA-damaging agents: acid, hydrogen peroxide, and near-UV irradiation. Loss of smx rendered the mutant strain more sensitive to oxidative damage than the wild-type strain when cultures were grown at pH values of 5. Given that peroxide sensitivity is a hallmark phenotype of E. coli strains deficient in xth, this finding further corroborates the conclusion that smx encodes a major, if not the sole, low-pH-inducible AP endonuclease activity expressed in S. mutans. Acid challenge experiments showed that while the mutant strain still acid adapts, the response to acid-mediated damage is not affected by the loss of smx. The difference in sensitivities of the smx strain to acid and hydrogen peroxide suggest the possibility of different mechanisms for the formation of DNA damage by the two agents that we used in this study, at least in S. mutans.
All the strains tested in this study were adversely affected by the presence of iron, confirming earlier results (5) which have suggested that the mechanism of metal ion toxicity in the absence of oxygen is complex and likely includes mechanisms beyond the involvement of Fenton chemistry. Iron in the form of Fe2+, plus the addition of hydrogen peroxide to the reaction mixture, significantly altered the survival characteristics and resulted in complete killing of cells. The data strongly suggest that the bulk of iron had probably been converted to Fe3+ via Fenton chemistry (for a review, see reference 36). Along with the conversion of Fe2+ to Fe3+, the concomitant formation of hydroxyl radical (OH·) likely leads to substantial DNA damage, with the formation of AP sites (37, 38). Clearly, strains containing a mutation in smx were far less able than the wild-type or the recA-deficient strain to cope with oxidative damage. The presence of iron in the reduced state served only to intensify the effects of the damage seen with peroxide alone, confirming our earlier hypothesis that RecA-dependent processes are not involved in the protection of the organism from oxidative damage (27). The mechanism of iron-mediated killing is still unclear, but the action of iron as a pro-oxidant (iron facilitating hydrogen peroxide killing) was well supported by our observations.
Our results demonstrate that Smx is the major AP endonuclease in S. mutans and that it is capable of removing AP sites in DNA. Moreover, the data show that acid adaptation involves at least some aspects of the oxidative-stress response, in the sense that the smx mutant strain was sensitive to hydrogen peroxide. What remains to be established is how extensively oxidative-stress gene regulation coincides with other attributes of acid adaptation in S. mutans or whether a specific subset of genes, such as smx, overlaps with those products participating in the acid response repertoire of oral streptococci. Experiments designed to provide insights into the regulation of Smx production are being conducted presently. The construction of the expression vector described in this study allows the convenient preparation of purified Smx protein, potentially facilitating the characterization of mutant forms of the enzyme. This will, in turn, aid in a more detailed understanding of the requirements of the Smx enzyme for substrate recognition and cleavage as well as further biochemical characterization. These data will enable us to gain a more complete picture of the overall role of Smx in the survival strategies of S. mutans.
We thank Wendi Kuhnert and Kelly Monahan for their assistance with nucleotide sequencing of the clones containing the smx and endonuclease III genes. We also thank Elizabeth Fozo for critical reading of the manuscript.
This work was supported by NIH/NIDCR grants DE10174, DE13683, and DE06127.