|Home | About | Journals | Submit | Contact Us | Français|
NADH oxidase (Nox) is a flavin-containing enzyme used by Streptococcus mutans to reduce dissolved oxygen encountered during growth in the oral cavity. In this study, we characterized the role of the NADH oxidase in the oxidative and acid stress responses of S. mutans. A nox-defective mutant strain of S. mutans and its parental strain, the genomic type strain UA159, were exposed to various oxygen concentrations at pH values of 5 and 7 to better understand the adaptive mechanisms used by the organism to withstand environmental pressures. With the loss of nox, the activities of oxygen stress response enzymes such as superoxide dismutase and glutathione oxidoreductase were elevated compared to those in controls, resulting in a greater adaptation to oxygen stress. In contrast, the loss of nox led to a decreased ability to grow in a low-pH environment despite an increased resistance to severe acid challenge. Analysis of the membrane fatty acid composition revealed that for both the nox mutant and UA159 parent strain, growth in an oxygen-rich environment resulted in high proportions of unsaturated membrane fatty acids, independent of external pH. The data indicate that S. mutans membrane fatty acid composition is responsive to oxidative stress, as well as changes in environmental pH, as previously reported (E. M. Fozo and R. G. Quivey, Jr., Appl. Environ. Microbiol. 70:929–936, 2004). The heightened ability of the nox strain to survive acidic and oxidative environmental stress suggests a multifaceted response system that is partially dependent on oxygen metabolites.
The ability to metabolize oxygen is a nearly universal trait among bacteria. In many species, oxygen serves as an electron acceptor in the electron transport chain for production of ATP via oxidative phosphorylation, which prevents the formation of potentially dangerous metabolites (28). However, cellular respiration itself can lead to the production of reactive oxygen species (ROS), including superoxide radical (O2−), hydroxyl anion (HO−), and hydrogen peroxide (H2O2) (29). The accumulation of ROS in cells can lead to protein, DNA, and membrane lipid damage, along with enzyme inactivation, ultimately resulting in cell death. Bacteria have evolved various means of coping with the deleterious effects of respiration, including detoxification mechanisms such as catalase, superoxide dismutase, and various dehydrogenases and peroxidases (58).
The oral bacterium Streptococcus mutans is a facultative anaerobe found primarily on the human tooth surface in a multispecies biofilm known as dental plaque and to a lesser extent in saliva (2, 37, 41). Current models of dental plaque architecture are consistent with biofilm models of microbial environments, in that channels exist in biofilms that allow fluid movement, delivery of nutrients, and potential chemical challenges (31, 39, 61). Given the estimated numbers of bacterial species present in dental plaque (1, 2), it is perhaps contrary to expectation that oxygen tensions are not zero in much of plaque (37). Indeed, oxygen levels at approximately 10% of atmospheric values have been reported (37, 41). Oxygen, moving through plaque via saliva, is available for metabolism to reduced oxygen species, particularly O2 and H2O2. S. mutans lacks catalase, cytochrome oxidases, and an electron transport system (24); however, its genome encodes over 30 different dehydrogenases (7), suggesting an amplified role for oxygen-metabolizing enzymes in the effort to survive and adapt to the ever-changing environment of the human oral cavity.
In streptococci, a highly conserved mechanism of oxygen metabolism occurs via flavin-based enzymes, which act to reduce oxygen, one electron at a time, to either water (H2O) or H2O2, through the oxidation of NAD (NADH) to NAD+ (23, 24, 43, 50, 54, 59). The enzymatic functions of two flavin-containing enzymes, alkylhydroperoxidase (AhpF) and Nox, have been elucidated using mutants of S. mutans GS-5 (25, 50). Nox is clearly important for oxygen metabolism in S. mutans, as indicated by increased NADH-dependent respiration rates in aerobic cultures (43) and the inability of a nox mutant of S. mutans GS-5 to grow in aerated cultures (25, 62). Although both AhpF and Nox have been implicated in oxygen-mediated stress responses and the maintenance of cellular NAD+/NADH ratios, which contribute to the efficiency of sugar metabolism (25), the AhpF enzyme (a H2O2-forming NADH oxidase) seems to have little physiological effect on sugar metabolism. AhpF does play a role in peroxidation, in conjunction with the AhpC peroxidase, and has been studied extensively (25, 50).
The Nox enzyme (a H2O-forming NADH oxidase) contributes to efficient metabolism of sugar substrates to lactic acid via the regeneration of NAD+ and the maintenance of NAD+/NADH ratios (25). In S. mutans, sugar metabolism leads to the rapid acidification of dental plaque and initiation of the oral disease dental caries. However, studies have shown that acidic biofilm cultures of S. mutans exhibit reduced NADH oxidase activity compared to planktonic cultures (43). Furthermore, internal acidification of S. mutans cells, using membrane-permeative weak acids, also inhibits NADH oxidase activity (48), suggesting that in the acidic environment of dental plaque, NADH oxidase activity could be diminished. Thus, a question has been posed about the role of NADH oxidase in the pathophysiology of S. mutans during growth at low pH (37).
As a result of its acidogenic lifestyle, S. mutans utilizes a variety of adaptive strategies to survive the low-pH environments in dental plaque (34). These include upregulation of the F-ATPase activity (10, 32, 55), decreased phosphotransferase activity (4, 5), shifts in metabolic end products, and changes in the membrane fatty acid composition (19, 21). The loss of one or more of the acid-adaptive mechanisms can lead to a substantial reduction in pathogenic capability; for example, a mutant strain of S. mutans UA159 lacking the ability to produce unsaturated membrane fatty acids generated substantially less disease in a rat caries model (21). Loss of the NADH oxidase leads to changes in the relative proportions of metabolic end products in S. mutans GS-5 (25) and has been implicated in alterations of membrane fatty acid composition and pathogenic capability in other streptococcal species (47, 62).
In this study, we further characterized the contribution of the NADH oxidase to oxygen metabolism in S. mutans and the relationship of the enzyme to stress responsiveness. We show that while the enzyme is a major mechanism for the reduction of oxygen in S. mutans, NADH oxidase is not solely responsible for oxygen metabolism. Utilizing controlled pH and oxygen concentrations in chemostat cultures of S. mutans UA159 and a nox mutant strain, we showed that NADH oxidase accounts for approximately 40 to 50% of respiration by the organism and that other, unidentified mechanisms contribute to the formation of an anaerobic state for the bacterium. The loss of Nox resulted in elevated levels of oxidative damage repair enzymes and substantial increases in the proportion of unsaturated membrane fatty acids, decoupled from external pHs, further expanding our view of the role of membrane fatty acid production as part of global stress responses (19). Finally, loss of nox affected adaptation and survival of S. mutans under highly acidic conditions, indicating a strong overlap between oxidative and acid stress responses, with Nox as a key contributor to the control of oxygen and its metabolites.
Strains used in this study are listed in Table 1. Streptococcus mutans strain UA159 (42) and subsequent derivatives were grown on brain heart infusion (BHI; Difco) agar at 37°C in a 5% (vol/vol) CO2–95% air environment. Antibiotics were added to a final concentration of 5 μg ml−1 for erythromycin and 1 mg ml−1 for kanamycin. Organisms were cultured in TY medium (3% tryptone, 0.1% yeast extract, 0.5% KOH, 1 mM H3PO4) and were grown in liquid culture or in continuous cultures in a BioFlo 2000 fermentor (New Brunswick Scientific, Edison, NJ) as described previously (19, 53). Continuous cultures were grown at a dilution rate of 0.24 h−1 under glucose-limiting conditions (2.3 mM), with a continuous impeller speed of 200 rpm, unless stated otherwise. Steady-state pH levels were maintained by the addition of 2 N KOH. Oxygen concentration was controlled by addition of air to the fermentor vessel at a rate of 1 vessel volume min−1, and impeller speed was automatically adjusted to regulate oxygen diffusion on demand. The culture pH and dissolved oxygen (DO) concentration were continuously monitored throughout the experiment by using an indwelling pH probe (Mettler Toledo, Columbus, OH) and an InPro 6000 oxygen sensor electrode (Mettler Toledo, Columbus, OH), respectively. After continuous cultures had been maintained for a minimum of 10 generations, aliquots of the culture were removed and cells were collected by centrifugation. Cell pellets were washed and stored frozen at −80°C prior to fatty acid and enzyme analysis.
Escherichia coli was grown on LB agar medium at 37°C. Liquid cultures were grown with shaking at 37°C. Antibiotics were added to a final concentration of 100 μg ml−1 for ampicillin, 50 μg ml−1 for kanamycin, 500 μg ml−1 for erythromycin, 20 μg ml−1 chloramphenicol, and 12.5 μg ml−1 for tetracycline.
Growth curves were recorded using an automated growth-monitoring device, a Bioscreen C (Growth Curves USA, Haverhill, MA), as described previously (30). The optical density at 600 nm (OD600) was measured every hour with a 10-s shaking period before each reading, to evenly suspend cultures. Microaerobic cultures contained a 50-μl mineral oil overlay.
Chromosomal DNA was isolated from S. mutans as previously described (52). Plasmid DNA was isolated from E. coli as previously described using a Qiagen miniprep kit (Qiagen, Chatsworth, CA). PCR was carried out with Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA). Amplicons were isolated and purified via gel electrophoresis as previously described (32). S. mutans UA159 was transformed by previously described methods (42, 46). Southern analysis was performed as previously described (56, 57) to validate mutant strains.
An insertional mutation was made in the S. mutans UA159 gene SMU.1117 (presumptively designated nox by NCBI [http://www.ncbi.nlm.nih.gov/nuccore/AE014133.2?report=genbank&from=1058760&to=1060133]) (7) using splice overlap extension (SOE) PCR as previously described (18, 20, 26, 27). To create a unique BglII site at position +450 of the predicted coding region, the primer pair NoxRevEco and NoxBglUp (Table 2) was designed to amplify the 5′ portion of the coding region. A PCR product containing the carboxyl terminus of the coding sequence was generated with the primer pair NoxFwdPst and NoxBglDwn (Table 2). Using the outside primer pair, a 1,170-bp amplicon was generated containing the unique BglII site, digested with PstI and EcoRI, and cloned into the compatible EcoRI and PstI sites of pBR322 (12). E. coli DH10B was transformed, and positive recombinants were selected on tetracycline-containing medium. Positive clones were confirmed by sequencing. The resulting plasmid, pBRNox (Table 3), was linearized with BglII, and an erythromycin resistance cassette, isolated from pTS19E (8), was inserted to create the construct pBRNoxErm (Table 3). Positive recombinants were selected on agar medium containing erythromycin and confirmed by restriction analysis. S. mutans UA159 was transformed with pBRNoxErm (Table 3), and correct constructions were selected on BHI medium containing erythromycin (46, 51). Putative mutant strains were confirmed by Southern hybridization (56, 57), and one strain was designated S. mutans UR110.
The glutathione oxidoreductase-encoding gene, SMU.838 (7), referred to as gorB in this study, was deleted using a method similar to the one described above. A BglII restriction site was engineered into the gorB gene at position +672 by SOE PCR using the primer pair GorBFwdEcoRI and GorBSOEBglIIRev (Table 2), as well as GorBSOEBglIIFwd and GorBRevPstI (Table 2). The amplicon containing gorB with the BglII restriction site was cloned into pCR-Blunt using the Zero Blunt PCR cloning kit (Invitrogen, Carlsbad, CA) and named pCRgorB (Table 3). A gel-isolated DNA fragment containing an Ermr cassette (8) was ligated into the engineered BglII site, and the resultant plasmid was named pCRgorBErm (Table 3), which was then used to transform S. mutans UA159 (46, 51). Transformants were screened via Southern hybridization (56, 57) to verify appropriate construction, and one such strain, carrying the gorB mutation, was designated S. mutans UR142.
The alkyl hydroperoxidase gene complex, ahpCF, was PCR amplified from genomic DNA using the primer pair AhpCFFwdAatII and AhpCFRevPstI (Table 2). The ahpCF gene complex and flanking DNA (171 bp on the 5′ end and 230 bp on the 3′ end) were cloned into pCR-Blunt using the Zero Blunt PCR cloning kit (Invitrogen, Carlsbad, CA), and the resulting plasmid was named pCRAhpCF (Table 3). A BglII site was created in ahpC, using SOE PCR (26, 27) with the primer pairs AhpCFFwdAatII-AhpCFSOEBglIIRev and AhpCFSOEBglIIFwd-AhpCF1044Rev (Table 2). The SOE PCR amplicon was subcloned into pCRAhpCF, yielding a truncated ahpCF gene cluster containing a unique BglII site, resulting in a plasmid named pCRAhpMerge (Table 3). The truncated ahpCF gene cluster was digested from pCRAhpMerge with EcoRI and subcloned into pSU20 (9) to create pSUAhpMerge (Table 3). A 2.2-kb BamHI fragment containing a Kanr marker isolated from pKanSma (an omega kanamycin cassette , subcloned into pWhitescript [Stratagene, La Jolla, CA]), was ligated into BglII-linearized pSUAhpMerge, and the resulting construct was named pSUAhpKan (Table 3). Plasmid pSUAhpKan was used to transform S. mutans UA159 to Kanr using published procedures (46, 51). Transformants were screened via Southern hybridization (56, 57) to verify appropriate construction, and one correct strain was designated UR145.
The superoxide dismutase gene (SMU.629), sod, was PCR amplified from genomic DNA using the primer pair SodFwdPst and SodRevEco (Table 2). The amplicon was cloned into the PstI and EcoRI sites of pBR322, and the construct was then named pBRsod (Table 3). The Ermr cassette isolated from pTS19E (8) was cloned into the unique BglII site of pBRsod at position +18 of the coding region. The resulting plasmid, pBRsodErm (Table 3), was used to transform S. mutans UA159 to Ermr using published procedures (46, 51). Transformants were screened via Southern hybridization (56, 57) to verify appropriate construction, and one such strain, carrying the sod mutation, was designated S. mutans UR109.
A double-mutant strain containing insertional mutations in both the nox (SMU.1117) and sod genes (SMU.629) was created. A kanamycin resistance cassette was isolated from pKanSma (described above) by digestion with BamHI and the fragment was ligated into the plasmid pBRnox, linearized with BglII, to create pBRnoxKan. The plasmid pBRnoxKan was then transformed into S. mutans UR109 (described above). Transformants were screened via PCR (56) to verify appropriate construction and one such strain, carrying mutations in both nox and sod, was designated S. mutans UR197.
S. mutans strain UR110, the nox mutant strain, was complemented with a single-copy genomic insertion of the SMU.1117 locus, including the upstream 217 bases of the nox intergenic region, into the gtfA (SMU.881) locus. Under the growth conditions tested, no physiological impact from the disruption of the gtfA locus was detected, as reported previously (60). Complementation cloning was performed as previously described (20), with modifications. Primers containing BamHI sites were designed and named RnsDNoxA-RT-F1 and NoxAGtfA-R1 (Table 2). The primer pair was used to PCR amplify the nox coding region and cognate promoter fragment for subsequent cloning into pSUGK-Bgl (Table 3), an integration vector. The vector pBGK (Table 3), containing the gtfA gene of S. mutans interrupted by an ΩKan gene (60), was digested with EcoRI, and a 4.7-kb fragment was purified and ligated to a 2.3-kb EcoRI-digested fragment of pSU20 (9). Clones were selected on LB agar medium containing kanamycin and screened on medium containing chloramphenicol. A clone with the appropriate construction was named pSUGK (Table 3). The primer pair pSUGKQuickchangeF and pSUGKQuickchangeR (Table 2) was used to create a BglII site in pSUGK using the QuikChange XL site-directed mutagenesis kit (Stratagene, La Jolla, CA), and the resulting construct was named pSUGK-Bgl (Table 3). Proper integration of the nox promoter plus coding region into the BglII site of pSUGK-Bgl, in the opposite orientation to gtfA, was determined by colony PCR. The correct construct, pSUGKnox9, was transformed into S. mutans UR110 and selected on BHI agar medium containing kanamycin. The complemented strain was named S. mutans UR201.
A chloramphenicol acetyltransferase gene (Promega, Madison, WI) was amplified using NewCATFwd and NewCATRev, followed by a BamHI digestion. The CAT gene was inserted into the BglII site of pSUGK-Bgl to create a single-copy chromosomal integration vector, pSUGKCAT. The proper integration of a promoter-CAT fusion in pSUGK-Bgl, in the opposite orientation to the gtfA gene, was determined by PCR and verified by nucleotide sequencing. The mutY (SMU.1865) promoter was PCR amplified using MutYCATFwd and MutYCATRev primers. The resulting amplicon, containing NdeI and BglII restriction sites, was ligated into pSUGKCAT, which was similarly digested. A positive construct, pSUGKmutY, was selected on LB medium containing kanamycin. The promoter-CAT fusion construct was integrated into the gtfA locus in the opposite orientation to gtfA by transforming Streptococcus mutans UA159 or the nox mutant strain, UR110, to kanamycin resistance, creating the CAT gene reporter strains UR230 and UR233, respectively.
The fpg promoter fusion was created using the Gateway system (Invitrogen, Carlsbad, CA) and the chloramphenicol acetyltransferase gene (Promega, Madison, WI) as detailed below. The CAT gene, described above, was amplified using the primer pair CATNotIFwd and CATEcoRVRev. The plasmid pENTR4 (Invitrogen, Carlsbad, CA) was digested with the restriction enzymes NotI and EcoRV and ligated with the digested CAT amplicon to create the plasmid pENTRCAT. The fpg (SMU.1614c) promoter was amplified using primers fpgBamHIFwd and fpgNotIRev. The fpg promoter amplicon was ligated to pGEM-T (Promega, Madison, WI) to create the plasmid pGEMfpg. The pGEMfpg plasmid was digested with NotI and BamHI and ligated to the similarly digested pENTRCAT to create the plasmid pEfpgCAT. The Gateway vector conversion system (Invitrogen, Carlsbad, CA) was used to change the integration vector pBGK (60) to a destination vector using the RfA cassette to create pBGKGW(−). When pEfpgCAT, containing the fpg promoter in an entry vector, and LR Clonase (Invitrogen, Carlsbad, CA) were used in conjunction with pBGKGW(−), the plasmid pBfpgCAT was created. This plasmid carried an fpg promoter-CAT fusion that would integrate into the gtfA locus in the opposite orientation of gtfA. Plasmid pBfpgCAT was transformed into UA159 and the nox mutant strain, UR110, to create strains UR116 and UR186, respectively.
Background activity of the CAT gene was determined by using a promoterless CAT construct created using the Gateway system described above. Plasmids pENTRCAT and pBGKGW(−) were used to perform a BP clonase reaction creating pBCAT. This plasmid was transformed into the parent strain, UA159, and the nox mutant strain, UR110, to create strains UR113 and UR187, respectively, containing the CAT gene without promoter control.
S. mutans strains to be tested were grown overnight in BHI broth. Cultures were diluted 1:10 into BHI medium and grown to mid-log phase. Assays were also performed on cells grown to steady state in a chemostat as described above. Cells were harvested by centrifugation at 3,452 × g for 15 min at 4°C and washed, and pellets were stored at −80°C. Cell pellets were resuspended in 1 ml lysis buffer (10 mM Tris [pH 7.5], 50 mM NaCl, 1 mM EDTA). An equal volume of 0.1-mm-diameter, acid-washed glass beads was added to the cell suspensions. The cells were homogenized by shaking twice for 2 min each time using a Mini-Beadbeater (Biospec Products, Inc., Bartlesville, OK). Disrupted cells were then centrifuged at 6,600 × g for 10 min at 4°C, and the cleared lysates were used for the assays. The protein concentration of the lysates was estimated by the method of Bradford (13).
NADH oxidase activity was measured as previously described (14, 49). The assays were performed at 25°C, and the assay mixture contained 100 mM pH 7 potassium phosphate buffer, 0.3 mM EDTA, and 50 μg protein. NADH was added to a final concentration of 0.16 mM to start the reaction. The decrease in absorbance at 340 nm was recorded over 4 min. One unit of NADH oxidase activity was defined as the amount of enzyme needed to catalyze the reduction of 1 mmol NAD+ minute−1 μg protein−1.
Gor activity assays were performed as described by Carlberg and Mannervik (15). Briefly, reactions were carried out at 25°C using 1-ml mixtures containing 100 mM potassium phosphate buffer (pH 7), 1 mM EDTA, 0.1 mM NADPH, and 50 μg protein. Oxidized glutathione (GSSG) (1 mM; Sigma, St. Louis, MO) was added to catalyze the reaction, and the decrease in absorbance was measured at 340 nm over 4 min. One unit of glutathione reductase activity was defined as the amount of enzyme needed to catalyze the reduction of 1 μmol NADP+ minute−1 μg protein−1.
Sod activity was measured by the cytochrome c-xanthine-xanthine oxidase method described by McCord and Fridovich (40) and Phan et al. (48). Briefly, reactions were carried out at 25°C using 1-ml mixtures that contained 50 mM potassium phosphate buffer (pH 7), 10 μM cytochrome c, 50 μM xanthine, and various amounts of cell extract. The concentration of xanthine oxidase (Sigma, St. Louis, MO) used in the assay is empirically determined by the amount necessary to produce a rate of reduction of cytochrome c of 0.025 U/min at an absorbance of 550 nm. Sod activity was recorded as units of Sod activity μg protein−1.
Chloramphenicol acetyltransferase assays for the measurement of transcriptional regulation were performed according to Kuhnert and Quivey (32). Results are given in nmol of chloramphenicol acetylated min−1 mg total protein−1.
S. mutans cultures were grown to steady state in a chemostat and as an overnight culture. Cell pellets were washed with H2O and stored at −80°C until processed for fatty acid methyl ester analysis (performed by Microbial ID, Newark, DE).
Acid and hydrogen peroxide survival assays were performed as previously described (53). Cells harvested from steady-state cultures of S. mutans UA159 and UR110 were exposed to an acid challenge of 0.1 M glycine, pH 2.5, or an oxidative challenge of 16.3 mM H2O2. Aliquots were sampled at 0, 15, 30, 60, and 90 min, diluted, and plated for enumeration. Surviving colonies were counted to estimate the percent survival at each time compared to that at time zero.
The Streptococcus mutans gene SMU.1117 (nox [nox-2]; http://www.ncbi.nlm.nih.gov/nuccore/AE014133.2?report=genbank&from=1058760&to=1060133) is annotated as the sole gene for H2O-forming NADH oxidase (7). To determine if the presumptive nox gene was uniquely responsible for the NADH oxidase activity in S. mutans UA159, coding regions of genes annotated as oxygen-metabolizing or oxygen-mediated stress response genes were disrupted as described above (see Materials and Methods) and used in the experiments described below.
Cell extracts of strains UA159, and mutant strains carrying insertions in the genes encoding superoxide dismutase (sod), glutathione oxidoreductase (gorB), alkylhydroperoxidase (ahpCF), and NADH oxidase (nox) were assayed for NADH oxidase activity. The results showed an approximately 10-fold decrease in NADH oxidase activity in the nox strain, the only mutant strain in the test group that showed a marked decrease (Fig. 1). The genetically complemented nox strain (nox+) restored NADH oxidase activity, indicating that SMU.1117 encodes the Nox enzyme and that the protein represents the major NADH oxidase in S. mutans (Fig. 1). Moreover, strains containing mutations in ahpCF, sod, and gorB maintained the ability to metabolize NADH using oxygen as a sole substrate and, in fact, demonstrated increases in NADH oxidase activity over the parent strain of 1.7-fold, 1.3-fold, and 1.5-fold, respectively. A double mutant strain containing mutations in nox and ahpCF exhibited NADH oxidase activity similar to that of the nox mutant (data not shown).
Previous work indicated that levels of NADH oxidase activity differed with variations in environmental oxygen conditions in S. mutans (43). The nox mutant strain of S. mutans UA159 provided an opportunity to determine whether NADH oxidase activity levels, attributable to the nox gene product, were also affected by culture pHs. Cells were grown in chemostat cultures, using glucose limitation to hold the cultures at steady state, at either a neutral pH of 7 or an acidic pH of 5. Extracts prepared from the UA159 and the nox strains showed elevated NADH oxidase activity levels in the extracts prepared from the pH 5 cultures (Fig. 2), with 1.2-fold and 1.4-fold increases, respectively, compared to the values from cells grown at pH 7. The data suggested that the ability of S. mutans to oxidize NADH, in an oxygen-dependent reaction, was primarily, but not solely, due to the nox gene product. The data also suggested that the acid-adaptive response of S. mutans could include regulation of nox.
The data from the cell extract measurements indicated that levels of NADH oxidase activity were influenced by culture pHs. We hypothesized that S. mutans strains lacking NADH oxidase may have difficulty growing under low-pH conditions. The parent strain, nox mutant, and nox-complemented strain were grown in the presence of an acid stress (medium buffered to pH 5.5) or an oxygen stress (medium containing 0.5 mM H2O2), either aerobically or microaerobically (maintained by a mineral oil overlay), to analyze the contribution of nox to growth of the organism in physiologically relevant environments. The mutant strain grown at low pH under aerobic conditions demonstrated a statistically significant increase in doubling time, indicating slower growth, and a decrease in final yield (Fig. 3A). Both phenotypes were restored in the complemented strain grown under the same condition (Fig. 3B). In all microaerobic cultures, there was also an increase in doubling time of the nox mutant strain, which was restored in the complemented strain (Fig. 3D and F). These data indicate minor metabolic changes that are compensated for during growth in aerobic cultures, indicated by essentially no differences in the growth rates between strains (Fig. 3C and E). The growth phenotypes were not due to changes in the glycolytic capabilities of the nox mutant, as the glycolytic output was unaffected in the nox mutant strain (data not shown).
We determined the ability of S. mutans to metabolize oxygen, using chemostat cultures to control oxygen levels and culture pHs. Since oxygen is the major substrate of NADH oxidase-dependent respiration, the rate of dissolved oxygen (DO) consumption in the nox mutant strain was monitored from the point of inoculation throughout growth in a chemostat vessel. To calibrate the DO probe, ambient air (containing oxygen at a concentration of 21% or 0.21 mmol liter−1) was added to saturate the medium in an uninoculated chemostat vessel. Thus, the oxygen saturation level of the medium was set at 0.21 mmol liter−1. The parent strain, UA159, was able to metabolize nearly all DO in the medium (Fig. 4A). Following an initial drop in DO via respiration, the nox mutant strain maintained an elevated DO concentration of approximately 0.08 mmol liter−1 (Fig. 4A). The ability to respire oxygen was completely restored in the nox-complemented strain (Fig. 4A). When an excess of glucose was added to the vessel to allow acidification of the medium to a pH of 5, via glycolysis, a decrease in respiration rate was observed for all strains (Fig. 4B). As cultures approached steady-state growth at pH 5, the respiration rates and, in turn, the dissolved oxygen concentrations returned to the levels observed during growth at pH 7 for each strain (Fig. 4B). To verify that the bacteria were indeed consuming oxygen present in the medium, the dissolved oxygen in an uninoculated vessel was measured. In uninoculated medium, the dissolved oxygen was present at near-saturated levels of approximately 0.20 mmol liter−1 at a pH of 7 and approximately 0.18 mmol liter−1 at a pH of 5 (Fig. 4C), indicating that pH effects on DO were negligible during the course of these experiments.
The data from the oxygen uptake measurements in chemostat cultures clearly indicated that S. mutans has an additional mechanism(s) for reducing oxygen, beyond the activity of Nox. However, the role of Nox-independent oxygen reduction in the nox strain and its relationship to stress survival were unclear. Previously, we and others reported that growth at acidic pHs leads to an acid-adaptive response, which includes elevated resistance to hydrogen peroxide treatment (34). Here, we determined the ability of the nox mutant strain to respond to hydrogen peroxide exposure following steady-state growth at pHs of 5 and 7 (see Materials and Methods). Although Nox is the major mechanism for the reduction of environmental oxygen, we observed that the loss of nox did not restrict the ability of the organism to deal with oxygen stress in the form of H2O2 (Fig. 5A). In fact, the loss of nox correlated with increased resistance to H2O2 in cells grown under either neutral- or acidic-pH conditions (Fig. 5A). Control experiments, without hydrogen peroxide, did not result in cell death during these assays (data not shown), indicating that cell death in the treatment groups correlated strongly with the presence of hydrogen peroxide and the strains' relative abilities to cope with its effects.
We showed that the nox mutant strain was unable to metabolize oxygen at the same rate as the parent strain (Fig. 4A). In effect, the nox strain was exposed to elevated levels of oxygen, compared to the parent strain. On the basis of the elevated hydrogen peroxide resistance exhibited by the nox strain, we hypothesized that the increase in H2O2 resistance was, potentially, a response to the relatively high levels of oxygen experienced by the Nox-defective strain. To test this hypothesis, exogenous air was added to UA159 cultures at levels sufficient to maintain oxygen saturation of the culture medium. The level of dissolved oxygen was chosen to mimic the oxygen-mediated stress observed in the nox cultures. Following growth under these conditions, we determined the ability of the strains to survive H2O2 treatment. Here, we observed an elevated resistance to H2O2 when the parent strain was grown in the presence of sustained high levels of exogenously supplied oxygen (Fig. 5B), suggesting that exposure to levels of oxygen identical to those observed in the nox mutant stimulated an elevated response to oxidative stress, in particular, resistance to hydrogen peroxide.
The mechanism of elevated H2O2 resistance in the nox mutant strain was initially unclear. The possibility existed that the nox mutant strain was in a general state of heightened responsiveness to oxidative stress. Previously, the enzymatic activities of superoxide dismutase (Sod), glutathione oxidoreductase (Gor), and Nox demonstrated similar responses to oxygen when tested in spx mutant strains carrying deletions in genes encoding global regulators of the oxygen stress response (30). To test the hypothesis that the general stress response systems in the nox strain were elevated, we determined the enzymatic activity of the specific oxygen stress response proteins Gor and Sod. These enzymes were assayed from extracts of the nox mutant and the parent strain, UA159, grown to steady state at both neutral and acidic pHs. Sod activity in the nox mutant strain grown at pH 7 was approximately 3-fold higher than that of the parent strain; however, no significant difference in Sod activity was noted in cultures of the nox mutant strain grown at low pHs (Fig. 6A). Gor activity was elevated in the nox mutant strain, compared to the parent strain, at both neutral and acidic pHs, approximately 2-fold and 1.2-fold, respectively (Fig. 6B). Additionally, the parent strain exhibited increases in Sod activity of nearly 3-fold when cultures were grown to a steady-state pH of 5 compared to 7, indicating that the acid-adaptive response includes enzyme activity pertinent to oxygen metabolism (Fig. 6A). Further, the elevated Sod levels suggested a potential need to reduce superoxide radicals in the absence of Nox. In contrast, no acid-adaptive change in Gor activity was observed in either the nox mutant strain or the parent strain (Fig. 6B).
To further elucidate the importance of nox-encoded NADH oxidase activity in cellular homeostasis, we created a double mutant strain defective for both nox and sod. The goal was to establish the impact on the organism of loss of NADH oxidase, in the absence of superoxide-reducing capability (via Sod). If the nox sod double mutant was incapable of growth, the data would suggest that nox is the primary enzyme involved in efficient oxygen reduction. We observed that the double mutant was unable to grow aerobically and grew at a lower rate under microaerobic growth conditions than the UA159 parent strain (see Fig. S1A and B in the supplemental material). The decreased ability of the nox mutant strain to metabolize environmental oxygen, the increase in Gor and Sod enzymatic activity in the nox mutant strain, and the inability of the nox sod double mutant to grow under aerobic conditions all indicate that loss of nox results in a significant deficiency in oxygen metabolism that may lead to damage in other metabolic pathways.
The extent of the oxidative-stress response in the nox mutant strain was further assessed by characterizing the transcription of two base excision DNA repair enzymes, MutY and Fpg, which are highly conserved in bacteria and known to be involved in repair of oxygen metabolite-damaged DNA (11, 16, 17). In these experiments, promoter fusions of mutY (SMU.1865) and fpg (SMU.1614c) were created using a chloramphenicol acetyltransferase gene as the reporter to assay for potential oxygen stress related to the loss of NADH oxidase. Promoters for both enzymes were transcriptionally upregulated in the nox mutant background, compared to the UA159 background (Table 4). The results indicated an increase in DNA damage repair capacity in the nox strain correlating with the loss of NADH oxidase. Together with the increase in activity of Gor and Sod, the data suggest a stress response to environmental oxygen that contains oxygen-metabolizing enzymes (Gor and Sod) and DNA repair enzymes (MutY and Fpg), at least to cope with the decrease in oxygen-reducing capacity observed in the nox mutant.
Previous research suggested that an increase in oxygen-mediated stress resistance occurs in conjunction with at least some elements of the acid stress response of S. mutans (53). The nox mutant strain provided a means to further define the overlap between the two stress responses. The test strains UA159 and UR110 were assayed for survival after severe acid challenge following growth in steady-state chemostat cultures at fixed pHs of 5 and 7. Phenotypically, the loss of nox resulted in an increase in acid resistance following growth at both neutral and low pHs, compared to the parent strain, UA159 (Fig. 7A). It was important to determine whether the loss of nox was directly responsible for the resistance to acid stress or whether the elevated environmental oxygen concentration in the nox cultures was the cause of the increased oxygen-mediated stress response (and potentially the acid stress response). Thus, the acid challenge was also performed with samples of the parent strain grown in the presence of exogenous air, at levels mimicking those of the nox mutant strain (i.e., 0.08 mmol liter−1) (Fig. 4A). The results showed that cells from these cultures were also acid resistant, signifying the importance of oxygen metabolites as potent regulating components of the global stress response (Fig. 7B).
Previously, we reported substantial changes in membrane fatty acid composition in S. mutans dependent on external pH, involving a shift to long-chain unsaturated fatty acids (UFAs) during growth at pH 5 (19). Loss of UFAs in the membrane resulted in greatly reduced virulence in S. mutans (21). Here, we wished to determine whether the increased resistance to acid stress and decreased ability to metabolize oxygen exhibited by the nox mutant strain were correlated with changes in the membrane fatty acid composition of the organism. Using cells grown in chemostat cultures and fatty acid methyl ester analysis, we determined the membrane fatty acid composition as a function of external pHs. The nox mutant strain maintained an elevated proportion of unsaturated fatty acids (approximately 52% at pH 7 and approximately 72% at pH 5) compared to the UA159 parent strain (37% at pH 7 and 63% at pH 5) (Fig. 8A to D). We followed these measurements with additional experiments to determine the effect of environmental oxygen concentration on membrane fatty acid composition. As mentioned above, the parent strain, UA159, was grown in a culture with a dissolved oxygen concentration maintained at 0.08 mmol liter−1. Membrane fatty acid analysis of the strain grown with exogenous oxygen clearly showed that the increase in oxygen correlated with a substantial increase in the proportion of unsaturated fatty acids, approximately 65% at pH 7 and approximately 71% at pH 5, and a corresponding decrease in saturated fatty acids, compared to cultures grown without exogenous oxygen (Fig. 8E and F).
The data from experiments conducted with the chemostat-grown cells strongly indicated that oxygen plays a substantial role in affecting membrane fatty acid composition. Here, cultures of the parent strain, UA159, were grown in batch cultures in an anaerobic chamber or in a 5% CO2-enriched incubator, and membrane fatty acid compositions of culture samples were determined. No significant differences in membrane composition, in either the type or length of the fatty acids, were seen between samples from cells grown in a carbon dioxide-enriched atmosphere and those grown anaerobically (Fig. 9), stressing the importance of chemostat-grown cultures.
S. mutans has evolved a variety of overlapping, and redundant, adaptive mechanisms to respond to stresses in the dynamic environment of the oral cavity (34). The fluctuating carbon availability (3), oxygen tension changes (37), and the necessity of competing with hundreds of other microbial species in dental plaque, some of which are net producers of ROS (37), suggest that it is not surprising that adaptation to changes in acidic conditions and oxygen levels would be important to the survival of the organism. In previous studies, we have shown that the acid stress response includes upregulation of the membrane-bound F-ATPase (32, 33) and a recA-independent DNA repair system (18, 22) and, more recently, that S. mutans acts to increase the proportion of unsaturated membrane fatty acids in response to external acidification (19, 20) and, further, that loss of unsaturated fatty acids results in a substantial loss of virulence in the rat model of dental caries (21).
In this study, we focused on the role of NADH oxidase in the potential overlap of the acid stress response with the oxidative-stress response. The basis for this investigation has been published in previous studies showing that, in streptococci, the NADH oxidase is at the intersection of key metabolic attributes of this facultative organism (25, 62). These processes include contributions to the formation of lactic acid by way of redox cycling of NADH/NAD+ and the reduction of oxygen in single-electron reactions (25). The results of the enzyme's activity are thought to be a key mechanism in the creation of an acidic, highly reduced environment in dental plaque, establishing a pathogenic niche on tooth surfaces (37). However, studies have also shown that the NADH oxidase is inhibited at low pH in model biofilms (43) and that weak acids are capable of reducing respiration by S. mutans (48). Thus, there remained a question about the extent of the role of NADH oxidase in respiration by S. mutans.
Here, using tightly controlled environmental conditions in chemostat cultures, we determined the ability of S. mutans to respire oxygen in the presence and absence of the NADH oxidase. The use of a nox-defective mutant strain also facilitated the characterization of an induced oxygen stress response. Maintenance of culture pH and oxygen concentration in continuous-flow growth allowed further examination of the oxidative-stress response as it relates to acid resistance and changes in membrane fatty acid composition.
The initial data from this study, obtained with the nox mutant strain, showed that Nox catalyzes the majority of NADH-requiring oxygen consumption in S. mutans, though an additional ability to reduce oxygen, via NADH, was also evident. It has been shown that FabK, a member of the fatty acid biosynthesis pathway, exhibits NADH oxidase activity in Streptococcus pneumoniae (38). However, it is unknown if FabK is the source of the remaining NADH oxidase activity in the nox mutant strain of S. mutans, and mutation of fabK appears to be lethal (our unpublished observation). Furthermore, it is important to note that the nox sod double mutant strain was incapable of aerobic growth, indicating that Nox is the predominant NADH oxidase in S. mutans. Mutation of established oxidative-stress response genes, including those encoding glutathione oxidoreductase (gorB), superoxide dismutase (sod), and alkylhydroperoxidase (ahpCF), resulted in elevated levels of Nox activity. We interpret the results as indicating that loss of these stress response genes leads to elevated cellular damage and the need to reduce oxygen by increasing levels of Nox.
Previous reports indicated that the loss of Nox had little noticeable physiological effect on the growth of the organism on glucose-containing medium (25), which has been reproduced in the present study. However, when grown microaerobically, the nox mutant demonstrated slower growth, which led us to believe that there is a modest change in sugar metabolism in the S. mutans UA159-derived nox mutant strain. The loss of nox resulted in elevated resistance to oxygen stress, in the form of H2O2 sensitivity, and to acid stress. The observations suggested that overlapping stress-regulatory mechanisms act to counter the effects of hydrogen peroxide and acidification. Investigation of potential participants in the overall stress response revealed elevated levels of enzymes known to be protective against oxidative stress, superoxide dismutase and glutathione oxidoreductase, along with elevated levels of transcription for the DNA repair genes fpg and mutY. We interpret our findings with Gor, Sod, fpg, and mutY as indicating that they constitute a part of the plausible mechanism(s) for the increased ability of S. mutans to survive severe challenges following growth in acidic and oxidative stress.
It has been shown that Nox activity in S. mutans decreases in model biofilms, which are highly acidic, and in the presence of proton donors, such as organic weak acids and sodium fluoride (43, 48). However, we observed that Nox activity was abundant in extracts of cells grown in the chemostat at pH 5 or 7, indicating that while the enzyme is inhibited at low pH, assay of the enzyme in buffer at pH 7 restores enzymatic activity. Moreover, there was a consistent observation of elevated Nox activity in extracts derived from pH 5 cultures, suggesting a difference in nox regulation at the genetic level at low pH.
In this study, we report elevated levels of unsaturated membrane fatty acids in the nox mutant strain and in the parent strain of S. mutans following growth in high levels of oxygen. Previously, we had shown the importance of unsaturated fatty acids in protecting S. mutans from acidic pHs (19–21). The data here show that loss of nox results in increased proportions of unsaturated membrane fatty acids, similar to proportions previously observed as a result of pH-dependent changes. The physiological data, discussed above, demonstrate that loss of nox substantially increases the stress response mechanisms of the organism. When grown in oxygen levels mimicking the concentrations observed in the nox mutant strain, the parent strain exhibited increased proportions of unsaturated fatty acids, regardless of culture pH. We conclude from these measurements that oxygen-mediated stress, or the loss of Nox, can act to control unsaturated membrane fatty acid concentrations in a manner similar to pH-mediated shifts. Alterations in membrane fatty acid composition due to changes in oxygen concentrations have been observed in other streptococcal species (47, 62). The ratio of saturated to unsaturated fatty acids in streptococci is thought to be controlled by the competition for fatty acid substrates between FabK and FabM (36). If FabK, a flavin-containing enzyme (38), is inactivated or damaged by high levels of oxygen, the balance of fatty acids would shift to the production of unsaturated fatty acids.
Collectively, the results from this study suggest that the NADH oxidase enzyme in S. mutans likely functions to create an anaerobic environment, rather than contributing to the alteration of metabolic products. The observations of oxygen-mediated stress responsiveness by the nox strain and its parent were more substantial than effects on growth rate and yield. The genome of the organism contains a substantial number of enzymes with predicted flavin-binding clefts, suggesting the possibility of a substantial capacity for single electron transfer to oxygen, potentially creating oxygen radicals. The role, then, of the NADH oxidase would be primarily to protect the cell from oxygen radical formation by reducing oxygen to water and secondarily to contribute to sugar metabolism, through the oxidization of NADH.
In conclusion, the data indicate that multiple systems are affected by the loss of nox and/or an increase in environmental oxygen. Although there are reports indicating the involvement of a variety of regulators and sensing systems (6, 30, 35, 44), direct evidence for a specific regulatory mechanism for NADH oxidase is still a matter for investigation. Current work is directed to expanding previously published observations that indicated a role for Spx in the transcriptional control of NADH oxidase production (30). Finally, the effects of the nox mutation on the global transcriptome are under investigation to account for the ability of the organism to mount an oxidative and acid stress response in the absence of NADH oxidase.
This study was supported by NIH/NIDCR DE-13683 (R.E.M.), DE-17157 (R.G.Q.) and the Training Program in Oral Sciences T32 DE-07165 (A.M.D. and K.G.).
We thank Alan Smith, Jeremiah Baldeck, Ross Heffer, and Tracey Householder for technical assistance.
Published ahead of print 16 December 2011
Supplemental material for this article may be found at http://aem.asm.org/.