|Home | About | Journals | Submit | Contact Us | Français|
Conceived and designed the experiments: JPB ZTW. Performed the experiments: JPB SL XY GGX. Analyzed the data: JPB ZTW. Contributed reagents/materials/analysis tools: GGX ZTW. Wrote the paper: JPB ZTW.
The Rex repressor has been implicated in regulation of central carbon and energy metabolism in Gram-positive bacteria. We have previously shown that Streptococcus mutans, the primary causative agent of dental caries, alters its transcriptome upon Rex-deficiency and renders S. mutans to have increased susceptibility to oxidative stress, aberrations in glucan production, and poor biofilm formation. In this study, we showed that rex in S. mutans is co-transcribed as an operon with downstream guaA, encoding a putative glutamine amidotransferase. Electrophoretic mobility shift assays showed that recombinant Rex bound promoters of target genes avidly and specifically, including those down-regulated in response to Rex-deficiency, and that the ability of recombinant Rex to bind to selected promoters was modulated by NADH and NAD+. Results suggest that Rex in S. mutans can function as an activator in response to intracellular NADH/NAD+ level, although the exact binding site for activator Rex remains unclear. Consistent with a role in oxidative stress tolerance, hydrogen peroxide challenge assays showed that the Rex-deficient mutant, TW239, and the Rex/GuaA double mutant, JB314, were more susceptible to hydrogen peroxide killing than the wildtype, UA159. Relative to UA159, JB314 displayed major defects in biofilm formation, with a decrease of more than 50-fold in biomass after 48-hours. Collectively, these results further suggest that Rex in S. mutans regulates fermentation pathways, oxidative stress tolerance, and biofilm formation in response to intracellular NADH/NAD+ level. Current effort is being directed to further investigation of the role of GuaA in S. mutans cellular physiology.
Streptococcus mutans, the major causative agent of human dental caries, lives almost exclusively as tenacious biofilms on the tooth surface in an environment featured with fluctuating and often unpredictable conditions in nutrient availability, pH, temperature, oxygen tension, saliva, and shearing force , . It has evolved multiple mechanisms to survive various kinds of environmental insults, optimize carbohydrate catabolism, and accumulate significantly on the tooth surface under certain conditions, as seen in carious sites , . S. mutans is considered a homofermentative lactic acid bacterium, although the metabolic pathways of glucose by this bacterium vary, depending on the environmental conditions , , , . When glucose is available in excess or during aerobic growth, S. mutans catabolizes glucose through homofermentation pathways, yielding primarily lactate. During growth under glucose limiting or anaerobic conditions, S. mutans undergoes heterofermentation, generating also formate, acetate, acetoin, and ethanol.
Despite the abundance of oxygen in the oral cavity, plaque biofilms support growth of a wide range of bacteria, including facultative and obligate anaerobes. Aerobes on the top of the plaque can quickly utilize the oxygen, lowering the redox potential. In fact, deep within the plaque, the environment is thought to be mainly anaerobic , . Nevertheless, plaque bacteria, including anaerobes, have developed active defense mechanisms against oxygen and oxidative stress . S. mutans, a facultative anaerobe, lacks catalase, cytochrome oxidases, and a complete electron transport chain , . In S. mutans, oxygen and the deleterious reactive oxygen species (ROS) are scavenged through an array of cytoprotective enzymes, including alkylhydroperoxide reductases (AhpCF), glutathione reductase (GshR), superoxide dismutase (SodA), thioredoxin reductase (TrxA), and NADH oxidase (Nox) , , , .
S. mutans can utilize various carbohydrates at micro-concentrations , , , . One crucial process for continued carbohydrates catabolism is the regeneration of NAD+ , . Aerobically, with glucose in excess, the intracellular concentration of fructose 1,6-bisphosphate becomes elevated, and in turn, induces lactate dehydrogenase (Ldh) expression . Increased Ldh expression maintains the redox poise needed for continued glycolytic activity since pyruvate is converted into lactate with the concomitant oxidation of NADH to NAD+. On the other hand, when glucose concentration is low, intermediary fructose 1,6-bisphosphate concentration is low and Ldh is not induced . Under anaerobiosis, the bottleneck of catabolism is the re-oxidation of NADH to NAD+. Anaerobiosis induces the expression of pyruvate-formate lyase (Pfl), which can split pyruvate into formate and acetyl-CoA. The mechanism of Pfl does not directly impact the NADH:NAD+ ratio; however, the downstream action of alcohol-acetylaldehyde dehydrogenase (AdhE) can reduce acetyl-CoA into ethanol with the concomitant oxidation of NADH to NAD+. In Staphylococcus aureus, the shift to anaerobic conditions induced genes encoding exporters of the fermentation products lactate, formate, acetate, and 2,3-butanediol .
Rex is a member of the winged-helix family of DNA transcriptional repressors and plays a central role in regulation of energy metabolism by sensing the NADH:NAD+ ratio , . The Rex family of proteins has been shown to optimize energy production through activation of genes poised to act more efficiently anaerobically , , , , . In S. aureus, the shift from an aerobic to an anaerobic environment is associated with an increased glycolytic rate along with increased transcription of genes associated with lactate and formate export , . High NADH:NAD+ ratios can be avoided by increasing transcription of pyruvate dehydrogenase and AdhE , .
Previously, we showed that Rex in S. mutans is involved in biofilm formation and oxidative stress tolerance . This study was designed to elucidate members of the Rex regulon in S. mutans that contribute to the shift from homofermentation to heterofermentation and the oxidative stress tolerance. By combining protein-DNA interaction studies with RealTime-PCR and phenotype analysis, we have demonstrated that Rex can bind target promoters and regulate genes critical to the central carbon catabolism, oxidative stress tolerance, and biofilm formation. RT-PCR also revealed that rex in S. mutans is auto-regulated and co-transcribed with guaA (SMU.1054), for a putative glutamine amidotransferase. Deficiency of Rex and GuaA caused drastic defects in oxidative stress tolerance and biofilm formation.
Bacterial strains and plasmids used in this study are listed in Table 1. S. mutans strains were maintained in brain heart infusion (BHI) (Difco Laboratories) medium. Solid media were prepared similarly, but Bacto agar (Difco Laboratories) was added at a concentration of 1.5% (w/v). Kanamycin (1 mg mL−1) and/or spectinomycin (1 mg mL−1) were added to the growth medium, when needed. Unless stated otherwise, cultures were grown aerobically in a 37°C chamber containing 5% CO2 under static conditions. For growth studies, overnight cultures were diluted 1100 into fresh BHI and allowed to continue to grow. When reaching mid-exponential phase (optical density at 600 nm, OD0.4), cultures were diluted 1100 again and growth was then monitored using a Bioscreen C (Oy Growth Curves AB Ltd, Finland) at 37°C with and without sterile mineral oil overlay , . The OD was monitored every 30 minutes after gentle shaking for 10 seconds. All E. coli strains were grown in Luria Bertani medium at 37°C aerobically with or without inclusion of kanamycin (40 µg/ml) and/or ampicillin (100 µg/ml).
Strains deficient of GuaA were generated using a PCR-ligation-mutation strategy described elsewhere ,  (Table 1). The resulting mutant, TW263, was further analyzed by PCR and DNA sequencing to verify the deficiency of target gene and sequence accuracy. Similarly, a Rex- and GuaA-double mutant, JB314, was generated by replacing a large portion of both rex- and guaA-coding sequences, from nucleotide 136 (from translational initiation site ATG) of rex to nucleotide 598 of guaA, with a non-polar spectinomycin resistant determinant . TW239, a Rex-deficient mutant, and its complement strain, TW239C, were generated previously . For guaA mutant complementation, the coding sequence of guaA plus the cognate promoter region was amplified by PCR with the genomic DNA of strain TW239 as the template using high fidelity Phusion DNA polymerase (New England BioLabs, Ipswich, MA). The resulting amplicon with rex deleted and replaced with a non-polar kanamycin resistance element of similar size was cloned into the streptococcal shuttle vector pDL278 , , , , and the resulting construct was used to transform TW263 to generate the complement, TW263C. For complementation of JB314 with guaA in single copy, the guaA-coding sequence and its cognate promoter region were amplified by PCR using primers 1053-55′ and 1054-33′ with genomic DNA of TW239 as the template (Table 2). The resulting mutants isolated from BHI with spectinomycin and kanamycin would restore guaA in JB314 at its natural locus, and were designated as JB315. Similarly, JB314 was transformed with the amplicon generated similarly using TW263 as the template, and derivative, JB316, isolated from BHI with spectinomycin and kanamycin plates would have rex phenotypes restored. S. mutans UA159, TW239, and TW263 were also transformed with the empty vector pDL278 to calibrate growth rates in the presence of spectinomycin.
For Rex expression, the rex coding sequence was PCR amplified with the primers listed in Table 2, and then directly cloned in pQE30 (Qiagen, Inc.) following the manufacturer's instructions. E. coli DH10B (Invitrogen, CA) was used for plasmid amplification, and the insert was confirmed by DNA sequencing. The plasmid pQE-rex was transformed into expression host E. coli M15 (Qiagen, Inc.). The recombinant N-terminal His-tagged Rex protein (rRex) was expressed by IPTG (0.2 mM) induction and purified using a Ni-NTA column (Qiagen, Inc.) according to the manufacturer's instructions. Purified rRex was washed extensively with 20 mM Tris, pH 7.6, 250 mM NaCl, concentrated, and desalted. The concentration of Rex was determined using the extinction coefficient 10.32 mM−1 cm−1 while the concentration of NADH was determined using the extinction coefficient 6.3 mM−1 cm−1.
Protein-DNA interactions were analyzed using electrophoresis mobility shift assays (EMSA) similarly as described previously , . The promoter regions of selected genes identified previously in Rex-deficient mutant by DNA-microarray analysis  were amplified by PCR from the primers listed in Table 2. Purified PCR products were incubated with increasing amounts of rRex in the reaction buffer containing 20 mM Tris pH 8.0, 1 mM EDTA, 75 mM KCl, 2 mM DTT, and 10% glycerol for 10 minutes at room temperature , , , . When NAD+ or NADH was included for EMSA's, the samples were incubated for an additional 15 minutes upon the addition of NAD+ or NADH , , . The reaction mixtures were separated on a 5% polyacrylamide gel under native conditions for 40 minutes and DNA mobility shift was visualized by SYBR Gold staining (Invitrogen, CA) using a Bio-Rad ChemiDoc XRS Imager . For detailed dissection of the Rex-binding site (also Rex-box), oligonucleotides containing the putative Rex-binding site were synthesized, annealed, and then used as probes for EMSA (Table 2). Likewise, sets of oligos with mutations at particular nucleotides of the Rex-binding site were also used to determine Rex-binding specificity (Table 2). For cold competition assays, one oligonucleotide was 5′ biotinylated and annealed to its reverse complement. The cold probe was identical except it was not 5′ biotinylated. The EMSA reactions were carried out as described above. After electrophoresis, the gel was blotted to a Zeta™ Nylon membrane at 380 mA for 1 hour followed by UV cross-linking. Binding and the resulting mobility shift were visualized via LightShift Chemiluminescent EMSA Kit by following procedures recommended by the manufacturer (Pierce, Rockford, IL).
In an effort to identify the metabolic products of sugar catabolism, we measured and quantified the lactate generated by the wildtype and the mutant strains after 24 hour incubation in BHI by using EnzyChrom™ L-Lactate Assay Kit (Bioassay Systems, Inc.) . In comparison, we monitored the end-point pH using a standard pH electrode (Orion, Inc.). To assay the glycolytic pH drop , we grew the strains to OD0.5, washed twice with ice cold de-ionized water, and resuspended the cells with 50 mM KCl and 1 mM MgCl2. The pH was adjusted to 7.2 with 0.1 M KOH before the addition of 50 mM glucose. The pH drop was monitored every minute for a period of 30 minutes.
For biofilm formation, S. mutans strains were cultivated using a modified semi-defined biofilm medium with glucose (20 mM, BMG), sucrose (10 mM, BMS) or glucose (18 mM) and sucrose (2 mM) (BMGS) as the supplemental carbohydrate sources , , , . For quantitative and structure analysis, biofilms were grown on hydroxylapatite discs in 12-well plates in BMGS and BMS for 24 hours and 48 hours, and were then stained with Live/Dead BacLight fluorescent dye (Invitrogen, CA) before optical dissections using an Olympus Fluoview BX61 confocal laser scanning microscope at 600× magnification (Olympus) , . At least 5 random areas were scanned for each sample. For post-acquisition analysis, simulated xyz three-dimensional images were generated using SLIDEBOOK 5.0 (Olympus) , . Quantitative analysis, such as biovolume, biomass, surface area and mean height, was carried out using COMSTAT 2.0 with at least 5 independent scans for each sample , . The biovolume is defined as the volume (µm3) of the biomass per µm2 of substratum area.
The abilities of the cells to withstand acid and hydrogen peroxide challenge were carried out as described previously , , . Planktonic cultures were prepared from mid-exponential phase (OD600 nm0.5) cultures grown in BHI broth. For sessile populations, BMGS was used to support bacterial growth and glass slides were used as substratum , . After 48 hours, biofilms were harvested using a sterile spatula, and following sonication to gently disperse the biofilms and detach the streptococcal chains, the cells were washed and then subjected to either acid killing in 0.1 M glycine, pH 2.8 or 0.1 M glycine pH 7.0 plus 0.2% hydrogen peroxide for the hydrogen peroxide killing assays , , . Alternatively, methyl viologen (also paraquat, Sigma) at 10 mM or 25 mM was included in cultural medium, and the impact of superoxide generated on bacterial growth was monitored continuously using Bioscreen C.
For co-transcription and RealTime-PCR analysis, total RNA's were extracted from S. mutans strains prepared from either 50 mL of BHI broth or on glass slides grown in BMGS for 72 hours. Cells were collected by centrifugation at 4°C for 5 minutes, then immediately treated with RNAProtect (Qiagen, Inc.) as suggested by the manufacturer , , . Following brief sonication to de-chain and disperse biofilms, cells were treated by glass beadbeater, and total RNAs were isolated using hot phenol as described previously , , . To remove residual DNA, RNA preps were treated with DNaseI (Ambion, Inc.) and retrieved with the RNeasy purification kit (Qiagen, Inc.). For RealTime-PCR and co-transcription analysis, cDNA was synthesized with 0.25 µg of total RNA using the iScript cDNA synthesis kit (Bio-Rad) by following the procedures recommended by the manufacturer. RealTime-PCR was carried out with a Bio-Rad iCycler using procedures detailed elsewhere , , . Primers used for co-transcription analysis are listed in Table 2 and those for RealTime-PCR in Table S1.
All quantitative data were further analyzed using paired student's t-test.
The rex gene (SMU.1053) of S. mutans is flanked by downstream guaA (SMU.1054), encoding a putative glutamine amidotransferase, and by the upstream SMU.1052, encoding a conserved hypothetical protein (www.oralgen.lanl.gov) (Fig. 1A). While there is a large intergenic region (295 bp) between SMU.1052 and rex, the gap between rex and guaA is only 23 nucleotides, suggesting these two genes may be co-transcribed. Computer based analysis using BPROM, a bacterial sigma70 promoter recognition program, identified a putative −10 and −35 element of the rex promoter situated 26 nucleotides upstream of the translational start site of Rex. Interestingly, both rex and guaA, but not SMU.1052, were found to be down-regulated in response to the deficiency of BrpA, a predicted surface-associated protein with global role in regulation of stress responses and biofilm formation in S. mutans , , . To confirm that rex and guaA are transcriptionally linked, total RNAs were extracted from wildtype S. mutans, UA159, when the cultures were either grown planktonically in BHI (data not shown) and 3-day biofilms in BMGS. Reverse Transcription-PCR (RT-PCR) was carried out using reverse primer GuaA-Rv priming nucleotides 631–652 (relative to start codon ATG) of guaA (696 bp) and forward primer Rex-Fw priming nucleotides 286–308 within the 3′ end of rex (642 bp) (Fig. 1A, Table 2). As shown in Figure 1B, a single transcript of approximately 1.0 kb overlapping rex and guaA was amplified and identical in size to the PCR amplicon generated using genomic DNA as the template, further confirming that guaA is co-transcribed with rex under the conditions studied. In fact, our experimental evidence corroborates the bioinformatics data from Microbes Online (www. microbesonline.org/operons/gnc210007.html), which predicts rex and guaA to be the only genes within the rex operon. However, it cannot be ruled out that there are additional promoters driving the independent transcription of rex and guaA under certain conditions. The radC gene (SMU.1055) lies downstream of guaA, but is orientated in the opposite direction, signifying the end to the rex operon (Fig. 1A).
Computer-aided analysis of the rex promoter revealed two regions with high similarity to the consensus sequence of Rex-binding site that have been identified in Streptomyces coelicolor, Bacillus subtilis and S. aureus, TTGTGAAN4TTCACAA  (Fig. 1D). In fact, one of the putative Rex-binding sites, rex#1, actually overlaps the −35 element, providing an explanation as to how Rex feedback auto-regulation could occur (Fig. 1D). EMSA was then carried out using either the PCR-amplified promoter region of rex or oligonucleotides specific to the putative binding site (Table 2). As shown in Figure 1C, rRex was able to bind to the promoter sequence of rex, which further suggests that Rex is poised for auto-regulation. Consistently, our previous DNA microarray and RealTime-PCR data support the finding that the rex-guaA operon is auto-regulated. The guaA gene was up-regulated by approximately 5-fold in TW239, which had the rex-coding sequence deleted and replaced by a non-polar kanamycin resistance element , , .
Recent studies have shown that repressor Rex functions as a dimer and that binding to the operators of its target genes is modulated in response to the cellular NADH/NAD+ level , , , . To verify the direct role of Rex in the regulation of the altered genes, we used EMSA to assess the interactions between rRex and the promoters of selected genes identified from DNA microarray analysis of TW239 . We tested the promoters of two up-regulated genes in adhE (SMU.148) and frdC (SMU.1410 for fumarate reductase) and two down-regulated genes in gshR (SMU.140 for glutathione reductase) and tpn (SMU.1363 for a putative transposase). As seen in Figure 2, all of the promoters tested displayed a dose-dependent mobility shift to higher molecular mass of rRex-promoter complexes. For the adhE promoter, oligo of the Rex-binding sites were also chemically synthesized, and as shown by EMSA, they were able to bind to rRex (Fig. 2A). As expected in a cold competition assay, mobility shift was decreased with increasing amount of unlabeled DNA probe included in the EMSA reactions (Fig. S1, B). In addition, mutations of nucleotides that are considered critical within the Rex-box abolished rRex binding (Fig. S1, A). Consistently, no apparent mobility shift was observed when rRex was mixed with the promoter of levD, which is not regulated by Rex and whose promoter does not have an apparent Rex-box (Fig. 1S, C). These results further suggest the binding of rRex to the Rex-box of adhE promoter is nucleotides-specific.
Recent studies in B. subtilis and S. aureus have shown that Rex's ability to bind to the promoters is dependent upon the NADH:NAD+ ratio. When NADH>NAD+, Rex dissociates from the promoter, resulting in alleviation of Rex-mediated repression and transcription of Rex-regulated genes , , , , . To evaluate how nicotinamide moieties may affect Rex activity in S. mutans, we also examined the interactions of rRex with target promoters with inclusion of NAD+ and NADH in EMSA assays. In the EMSA pre-reaction, the target promoter of gshR and rRex were pre-incubated for 10 minutes, and then NADH and/or NAD+ (10 mM, final conc.) were added. As shown in Figure 3A, in the presence of NAD+, rRex bound to the promoter of gshR, as evidenced by the mobility shift. On the other hand, inclusion of NADH inhibited the formation of Rex-promoter complex. Alternatively, rRex bound the promoter of frdC under the same conditions even without any addition of exogenous NAD+ (Fig. 3B). The presence of NAD+ did not have major effect on mobility shift, but addition of NADH almost abolished such a binding activity (Fig. 3B). Effort was made to further evaluate the effect of NADH/NAD+ when mixed at different ratios on Rex-DNA binding efficiency, but the results were not evident as expected (data not shown), suggesting other factor(s) may be involved in Rex-mediated regulations in S. mutans.
Studies in B. subtilis suggest that Rex binds NADH and NAD+ tightly, with estimations of a KD lower than 100 nM for NADH . Based on this evidence, it could be envisioned that Rex is always bound to either NADH or NAD+. However, after purification of rRex, there was minimal NADH or NAD+ bound, which is likely an artifact of the over-expression and purification processes. Thus, in an effort to fully reconstitute rRex with either NAD+ or NADH to mimic the intracellular conditions, rRex was incubated with 20-fold excess of NAD+ or NADH overnight on ice, and then desalted using a Centricon (Millipore) to remove free NAD+ or NADH (Fig. 4A). As seen in Figure 4A, the UV-Vis spectrum of purified rRex (grey line) does not show the NADH associated shoulder peak at 340 nm. Indeed, full reconstitution of rRex with NAD+ did not change the UV-Vis spectrum. However, after reconstitution of rRex with NADH, the UV-Vis spectrum shows the typical 340 nm peak attributed to the absorbance of NADH (black line), confirming that NADH is directly bound to rRex. SDS-PAGE analysis was carried out to confirm that rRex itself was maintained and intact in the samples of the reconstitutions (Fig. 4B). The reconstituted samples were used to carry out the EMSA reactions with the promoter of adhE without the addition of exogenous nicotinamide. As seen in Figure 4C, NAD+-loaded rRex binds to the promoter of adhE, and NADH-loaded rRex is inhibited in binding to the adhE promoter.
It is apparent that Rex in S. mutans plays a major role in modulation of carbohydrates and energy metabolism. As shown previously by DNA microarray analysis and further confirmed by RealTime-PCR in this study  (Table 3), at least 17 genes with putative functions in carbohydrates metabolism were differentially regulated in response to Rex-deficiency. Of the genes up-regulated in TW239, several are known to be involved in heterofermentative catabolism , , , including pfl (SMU.402), adhE, pdh (SMU.1421/4 for pyruvate dehydrogenase complex), frdC and adhABCD (SMU.127-30 for the acetoin dehydrogenase complex) . The commonality among these enzymes is that they all function to maintain the NADH:NAD+ ratio needed to reposition the cells for re-entry into glycolysis. In response to carbohydrate limiting and/or high intracellular NADH/NAD+ level, the bacterial cells may divert the catabolic pathways away from homofermentation towards heterofermentation, resulting in a decreased production of acids, especially lactic acid by S. mutans. To this effect, we sought to determine the effects of Rex-deficiency on end-point pH and lactate production after 24 hours of growth and evaluate the glycolytic rate via pH drop experiments. As seen in Figure 5A, we noticed significant increases in the end-point pH of the Rex-deficient mutant, TW239, especially during growth under aerobic conditions. When grown in BMGS, the average end-point pH for UA159 was 6.04 (±0.03), as compared to 6.51 (±0.05) (P<0.001) for TW239, 6.13 (±0.04) (P>0.05) for GuaA-deficient mutant, TW263, and 6.33 (±0.05) (P<0.05) for the Rex-deficient and GuaA-deficient double mutant, JB314, respectively. EnzyChrom™ L-Lactate Assay also showed a reduction of approximately 1.5 mM in lactate production as a result of Rex-deficiency, with average of 24.21 (±0.55) mM) for TW239 vs 25.74 (±0.66) mM for UA159 (Fig. S2). In addition, pH drop experiments showed that TW239 had a slow pH drop upon initiation by addition of glucose and maintained a baseline threshold of pH 4.10, as compared to the baseline threshold of pH 3.80 for UA159 after 20 minutes (Fig. S3). As expected, complementation of the mutant with rex in pDL278 shuttle vector was able to partly restore these phenotypes to the wildtype levels (Fig. S2, S3). The differences between TW239 and UA159 in end-point pH, lactate production, and glycolytic rates could be attributed at least in part to the increases in ethanol and acetoin production as a result of the increased expression of AdhE and the acetoin dehydrogenase complex AdhABCD in response to Rex deficiency , . Similar trends were also observed when the strains were grown in BMS, although the differences were not as pronounced as in BMGS (data not shown). However, the exact metabolite pools await further investigation.
Previously, Rex-deficiency in S. mutans was shown to cause defects in oxidative stress tolerance, but not in acid tolerance response . When assessed by acid-killing at pH 2.8, TW239 was found slightly more susceptible to acid killing than UA159, although the differences were not statistically significant (P>0.05), consistent with our previous findings . Unlike TW239, the survival rate of the GuaA-deficient mutants, TW263, was slightly increased, but again the differences were not statistically significant from UA159 (Fig. 5B). However, deficiency of both Rex and GuaA in JB314 was shown to cause drastic increases in vulnerability to acid stress, with a reduction of survival rate in JB314 by more than 2-logs when compared to wildtype, UA159 (P<0.01). Complementation of this double mutant with rex plus its cognate promoter in JB316 and with guaA and its cognate promoter in JB315 were able to partly restore the phenotype to a level similar to TW263 and TW239, respectively (data not shown).
Hydrogen peroxide killing assays were used to evaluate the impact of GuaA-deficiency on oxidative stress tolerance. Consistent with previous findings , TW239 was shown to be more susceptible to hydrogen peroxide killing than UA159, with a reduction in survival rate of more than 1-log (P<0.001), when compared to UA159 (Fig. 6A). When subjected to hydrogen peroxide killing, a reduction of more than 6-fold in survival rate was also measured with TW263 (P<0.001) when compared to UA159 (Fig. 6A). The survival rate of JB314, was further reduced by almost 2-logs, relative to UA159, which suggests an additive effect of Rex- and GuaA-deficiency on the susceptibility of JB314 to hydrogen peroxide. Consistently, when the strains were grown in the presence of methyl viologen (10 mM, final concentration), both TW239 and JB314 displayed greater challenge to methyl viologen than UA159 (Fig. 6B). Again, possession of a wildtype copy of the deleted gene plus the cognate promoter region in shuttle vector was able to partly complement the Rex-deficiency in TW239 (data not shown).
Previous studies showed that deficiency of Rex in S. mutans had no major effect on growth of the Rex-deficient mutant, but altered biofilm formation, especially when grown in sucrose-containing medium . Similarly, neither deficiency of GuaA alone nor both Rex and GuaA had any major effects on growth rates of the deficient mutants during growth in BHI and BMG (data not shown). To more fully evaluate the role of the rex operon in biofilm formation, biofilms of the Rex- and GuaA-deficient single mutants, TW239 and TW263, and the double mutant, JB314, were grown in BMS and BMGS on hydroxylapatite discs in 12-well plates simulating a static growth model under both aerobic and anaerobic conditions. Following Live/Dead BacLight fluorescent staining, biofilms were analyzed using a confocal microscope equipped with an Olympus LUMFLN 60× water immersion objective , , . When grown under anaerobic conditions on hydroxylapatite discs in BMS, wild-type UA159 formed robust biofilms after 48 hours (Fig. 7). Similar to UA159, TW263 also developed mostly viable biofilms after 48 hours, but unlike UA159, the biofilms of TW263 were flat and thin. When analyzed by COMSTAT, the bio-volume of TW263 biofilms was nearly half of that of UA159, with an average of 4.98 (±3.35) µm3/µm2 for TW263 vs. 9.10 (±3.54) µm3/µm2 for UA159, suggesting impaired growth of GuaA-deficient mutant in biofilms. Similar to our previous data, TW239 formed loose biofilms with porous structure, averaging a biovolume of 10.68 (±2.32) µm3/µm2 (P<0.001). However, relative to UA159 and the respective single mutants, biofilm formation by the Rex/GuaA-double mutant, JB314, were severely retarded during growth in BMS, accumulating an average biomass of just 0.21 (±0.18) µm3/µm2 (P<0.001) after 48 hours. Unlike the biofilms of UA159 and the respective single mutants, the biofilms of JB314 were mostly dead, showing red fluorescence after Live/Dead staining. Similar but less dramatic differences were observed between the different strains when grown under in the presence of glucose (Fig. S4).
A highly conserved transcriptional regulator, Rex has been implicated in regulation of key enzymes in carbon and energy metabolism , , , . Previously, we showed that deficiency of Rex in S. mutans causes substantial alterations in the transcriptional profile and the abilities of the deficient mutant to survive oxidative stress and form biofilms . Here, we provided further evidences that similar to B. subtilis and S. aureus, Rex in S. mutans also binds to target promoters and its binding efficiency is affected by the level of NADH:NAD+. Like S. coelicolor, but unlike B. subtilis , , Rex in S. mutans is subjected to Rex-mediated auto-regulation, a feedback mechanism commonly used to optimize the efficiency of cellular functions. These results further suggest that Rex in S. mutans plays a major role in modulation of the efficiency of carbon central metabolism and balance of the redox status.
Intensive studies mostly in B. subtilis and S. aureus have shown that repressor Rex binds to the target promoters as a dimer , , , . In B. subtilis, Rex binds to a region of 5′-WWTGTGAANTNNTNNNCAAW-3′, where “W” denotes either A or T , . In S. aureus, the consensus sequence is 5′-TTGTGAAWWWWTTCACAA-3′ . During the preparation of this manuscript, Ravcheev et al. also used comparative genomics approach to infer candidate Rex DNA-binding motifs, and found that the DNA-binding motifs of Rex orthologs from 11 taxonomic groups showed significant conservation with a generalized consensus of TTGTGAANNNNTTCACAA . Similarly, computer-aided analysis of the promoter regions of the de-repressed genes in S. mutans revealed region(s) with high similarity to the consensus sequence of the Rex-binding site deduced from other model microorganisms (Fig. 1D, 1E). EMSA analysis of selected promoters and/or synthetic oligos of the Rex-binding sites also demonstrated nucleotide-specific interactions with the recombinant Rex (Fig. 2–3, S1). However, further analysis of these promoter regions/Rex-binding sites also showed that differences exist between different promoters in nucleotide composition and number of Rex-binding sites (Fig. 1D, 1E). Consistently as seen in Figures 2 and and3,3, the promoters tested have also displayed differences in binding affinities for rRex, which signifies a hierarchal importance of Rex regulated genes. Indeed, our microarray data showed that Rex-deficiency effects transcription dissimilarly, suggesting either the involvement of other regulator(s) and/or tighter regulation depending on the nucleotide composition of the Rex-binding site. Two Rex-binding sites could be identified in the promoter of rex and adhE (Fig. 1D), although the exact significance of such possessions in Rex-mediated regulation remains unclear. Alignment of 41 proven and putative Rex-binding sites showed that the Rex-binding sites in S. mutans are similar to the established consensus and is more similar with those of B. subtilis (Fig. 1E) , , .
It is also worth noting that when using the consensus sequence to do a whole genome search allowing two or three mismatches, many more putative Rex-binding sites could be identified in intergenic regions and upstream of genes that were not identified in the TW239 microarray (data not shown), including lactate dehydrogenase gene ldh . Unlike S. aureus , but similar to B. subtilis , , ldh was not identified by DNA microarray analysis in TW239. Further analysis by RealTime-PCR also revealed no significant difference between UA159 and TW239 during growth under similar conditions (data not shown). As shown in Figure 1D, one putative Rex-binding site was positioned at nucleotide 64 relative start codon ATG, and when analyzed by EMSA, rRex was indeed able to bind to this region (Fig. 2E). BHI is a rich medium but contains only 11.05 mM dextrose. Thus, it is possible that the conditions used to grow these cultures were not optimal for Ldh expression and that other regulatory mechanisms are likely involved in the Ldh expression. Nevertheless, these results further suggest that a broader role exists for Rex in regulation of S. mutans physiology under different conditions, although details on the functions and regulation of these genes await further investigation.
Previously identified by DNA microarray analysis  and further confirmed by RealTime-PCR in this study (Table 3), Rex-deficiency in S. mutans also causes down-regulation of more than 32 genes , a preliminary indication that Rex may also act as a transcriptional activator. In a recent study of a Ldh-deficient E. faecalis , Mehmeti et al also reported that Rex-binding sites were not only identified in many up-regulated genes, but also some of the genes that were down-regulated in the Ldh-deficient mutant. It was then suggested that Rex in E. faecalis may function as an activator, although no other information was provided concerning the relationship between these altered genes and Rex. In an effort to pinpoint the putative binding site for activator Rex, a 95 bp promoter region of gshR and a 410 bp promoter region of tpn, were amplified by PCR and then subjected to EMSA analysis. As shown in Figures 2 and and3,3, both promoter regions bound well to rRex. Similarly, inclusion of NAD+ enhanced the protein-DNA interactions, while NADH decreased the bindings. These results further suggest that Rex in S. mutans actually binds to the promoter of gshR and tpn, modulating their expression as an activator in response to the NADH/NAD+ level and probably other factors, although the binding site for activator Rex appears to be different from that of the repressor Rex. Studies are underway to dissect the binding sites for Rex as an transcriptional activator.
S. mutans possesses the molecular weaponry to target bursts of oxidative and nitrosative stress . Reactive oxygen/nitrogen species (ROS/RNS) can be mitigated through small molecules, including free thiols, glutathione, ascorbate, and tocopherol or through enzymatic proteins and transcriptional regulators. Of the genes identified in TW239 by DNA-microarray analysis, several were known to play a role in oxidative stress tolerance, including the mleSP and gshR , . The malolactic fermentation system, including malolactic enzyme MleS and malate permease MleP, functions in malate decarboxylation, yielding lactate, CO2 and ATP . It is also recently shown to play a major role in S. mutans to protect from starvation and acid- and oxidative stress , . Glutathione reductase (GshR) is well documented for its role in ROS detoxification in both eukaryotes and Gram-negative bacteria, and similar results were recently reported in S. mutans . S. mutans possesses two homologues, gshR and gshR-2 (or gor) by SMU.838 (www.lan.oralgen.gov). Besides gshR, more than 2-fold reduction was also observed in gshR-2 by RealTime-PCR in response to Rex-deficiency (Table 3). Therefore, down-regulation of gshR's and mleSP could at least in part attribute to the observed defects in tolerance against MV and H2O2 by TW239 .
In S. mutans, NADH oxidase (Nox, SMU.1117c) and alkyl hydroperoxidase (AhpC and AhpF by SMU.764 and SMU.765, respectively) have been shown to play a major role in O2 metabolism and oxidative stress response , , , , . They function to reduce O2 to H2O (Nox) or H2O2 (AhpC and AhpF) with the concomitant oxidation of NADH to NAD+ . Nox's functionality is dually beneficial to S. mutans with alternative routes for NAD+ regeneration, contributing to the efficiency of carbohydrates metabolism and for detoxifying deleterious oxygen metabolites , , . When analyzed by RealTime-PCR, nox and ahpC were found to be down-regulated by 2.67- and 1.86-fold (P<0.01), respectively in TW239 (Table 3). However, no significant differences were measured in expression of ahpF. Decreased expression of Nox further supports the notion that Rex optimizes the catabolic pathways, ensuring efficient carbohydrate utilization without needing NADH oxidase to replenish the NAD+ pool for glycolysis re-entry. In addition, the down-regulation of these key anti-oxidant house-keepers also provides rationale as to why TW239 has an increased susceptibility to oxidative stress.
Data continues to accumulate, suggesting that development of mature biofilms in S. mutans requires coordination of various cellular functions in response to environmental conditions, including acid and oxidative stress responses , , , . The weakened tolerance to oxidative stress can certainly be attributed in part to the observed defects in biofilm formation by TW239, and especially the Rex/GuaA-double mutant, JB314. It is known that acid and oxidative stress induce DNA damages, including formation of abasic sites in DNA , . The guaA gene is annotated to encode a putative glutamine amidotransferase of Class I family enzyme with a potential role in purine ribonucleotide biosynthesis (www.oralgen.lanl.gov), although genes encode enzymes that catalyze the transfer of the amino group from glutamine in purine ribonucleotide biosynthesis, such as amidophosphoribosyltransferase (SMU.32) and phosphoribosyl-formylglycinamidine synthetase (SMU.30), are clearly identified in an apparent operon. While the exact role of GuaA as well as the rationale for transcriptionally linking rex and guaA awaits further investigation, elevated expression of enzymes for ribonucleotide biosynthesis in TW239 could partly compensate the need for the repair of DNA damaged by oxidative stress as a result of Rex-deficiency. The additive effects of GuaA- and Rex-deficiency in acid and oxidative stress tolerance and biofilm formation as seen in JB314 further support this notion.
In summary, we further demonstrated that rex and guaA in S. mutans are co-transcribed and auto-regulated. As a redox-sensor, Rex can function as a transcriptional repressor as well as an activator in response to intracellular NADH/NAD+ level and metabolic state and plays a major role in regulation of central carbon metabolism, oxidative stress tolerance and biofilm formation. Further investigation is underway to pinpoint the binding site of Rex activator and elucidate the role of GuaA in S. mutans cellular physiology.
Rex-DNA interactions are specific. (A) EMSA stained with SYBR gold showing that Rex binding to the promoter of adhE is dependent on highly conserved nucleotides within the Rex-box. When incubated with annealed synthetic DNA oligo 148Fw and 148Rv, which contain the conserved Rex-box of the adhE promoter, the recombinant Rex caused mobility shift. As expected, no apparent binding and shift were observed when mixed with oligos with the conserved nucleotides mutated (m148Fw and m148Rv) (Table 2). (B) In a “cold” competition assay, rRex (25 ng) was incubated with a 5′ biotinylated oligo with intact Rex-Box of adhE promoter (1 ng) (Table 2) and increasing amounts of the non-biotiny|ated DNA probe. As expected, inclusion of cold probe decreased the mobility shift of the rRex-biotiny|ated promoter complex, as visualized by the LightShift Chemiluminescene EMSA Kit (Pierce, Rockford, IL), and such a effect is dose-dependent. (C) EMSA stained with SYBR gold showing results when Rex was mixed with the promoter of levD, which does not contain a Rex-box.
End-point L-lactate determination. S. mutans strains were grown aerobically in BHI. After 24 hours, L-lactate concentrations in the culture supernates were quantified by using EnzyChrom™ L-Lactate Assay Kit (BioAssay Systems, Hayward, CA). Preliminary results showed that the Rex-deficient mutant, TW239, produced 1.5 mM less lactate than the wildtype, UA159. Similar results were also seen with TW263, a GuaA-deficient mutant, and JB314, a Rex/GuaA double mutant. Complementation of TW239 and TW263 with the respective wild-type copy plus the cognate promoter in shuttle vector pDL278 in strain TW239C and TW263C, respectively, was able to partially restore the wildtype lactate levels.
Glycolytic pH drop. S. mutans strains were grown to an O.D. of 0.5, washed twice in ice cold water, and resuspended in 50 mM KCI and 1 mM MgCl2 before glucose (50 mM) was added to initiate glycolysis as detailed in Materials and Methods. Results showed that relative to the wild-type, UA159 (solid square), the Rex-deficient mutant, TW239 (open triangle), displayed a slower pH drop. Additionally, the resting pH of TW239 after 20 minutes was higher than that of UA159, as can be seen more clearly in the inset. Complementation of the mutant with the wild-type rex plus its cognate promoter in shuttle vector pDL278 (TW239C, solid triangle) was able to partly restore the phenotype to the wild-type. Similar trends but in a less degree was also seen with TW263, a GuaA-deficient mutant. TW263C, TW263 carrying shuttle vector pDL278 with the wild-type guaA; JB314, a Rex and GuaA-deficient double mutant.
Hydrogen Peroxide Killing of Anaerobic Biofilms. S. mutans biofilms were grown on glass slides anaerobically in BMGS. After 48 hours, biofilms were harvested, briefly sonicated, and washed with 0.1 M glycine, pH 7.0. The cells were then resuspended with 0.1 M glycine, pH 7.0 containing 0.2% (w/v) and hydrogen peroxide, and incubated for periods of 0, 90, and 110 minutes. At each time point, samples were taken, serially diluted, and plated in triplicate on agar plates. Results showed that mutants deficient of Rex (TW239) and GuaA (TW263) were more than 1-log more susceptible to hydrogen peroxide after 110 minutes, as compared to the wild-type, UA159. Relative to UA159, the Rex/GuaA double mutant, JB314, is 3-logs more susceptible to hydrogen peroxide.
Primers used for RealTime-PCR in this study.
This work was supported in part by grant R01 DE19452 from the National Institute for Dental and Craniofacial Research, United States of America (http://www.nidcr.nih.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.