Autoregulation of the Streptococcus mutans SloR Metalloregulator Is Constitutive and Driven by an Independent Promoter

ABSTRACT

Streptococcus mutans, one of ∼600 bacterial species in the human oral cavity, is among the most acidogenic constituents of the plaque biofilm. Considered to be the primary causative agent of dental caries, S. mutans harbors a 25-kDa SloR metalloregulatory protein which controls metal ion transport across the bacterial cell membrane to maintain essential metal ion homeostasis. The expression of SloR derives in part from transcriptional readthrough of the sloABC operon, which encodes a Mn²⁺/Fe²⁺ ABC transport system. Here we describe the details of the sloABC promoter that drives this transcription as well as those for a novel independent promoter in an intergenic region (IGR) that contributes to downstream sloR expression. Reverse transcriptase PCR (RT-PCR) studies support the occurrence of sloR transcription that is independent of sloABC expression, and the results of 5′ rapid amplification of cDNA ends (5′ RACE) revealed a sloR transcription start site in the IGR, from which the −10 and −35 promoter regions were predicted. The results of gel mobility shift assays support direct SloR binding to the IGR, albeit with a lower affinity than that for SloR binding to the sloABCR promoter. The function of the sloR promoter was validated by semiquantitative real-time PCR (qRT-PCR) experiments. Interestingly, sloR expression was not significantly affected when bacteria were grown in the presence of a high manganese concentration, whereas expression of the sloABC operon was repressed under these conditions. The results of in vitro transcription studies support the occurrence of SloR-mediated transcriptional activation of sloR and repression of sloABC. Taken together, these findings implicate SloR as a bifunctional regulator that represses sloABC promoter activity and encourages sloR transcription from an independent promoter.

IMPORTANCE Tooth decay is a ubiquitous infectious disease that is especially pervasive in underserved communities worldwide. S. mutans-induced carious lesions cause functional, physical, and/or esthetic impairment in the vast majority of adults and in 60 to 90% of schoolchildren in industrialized countries. Billions of dollars are spent annually on caries treatment, and productivity losses due to absenteeism from the workplace are significant. Research aimed at alleviating S. mutans-induced tooth decay is important because it can address the socioeconomic disparity that is associated with dental cavities and improve overall general health, which is inextricably linked to oral health. Research focused on the S. mutans SloR metalloregulatory protein can guide the development of novel therapeutics and thus alleviate the burden of dental cavities.

INTRODUCTION

Dental caries are important indicators of oral health and overall general health in both children and adults. Despite significant public health efforts aimed at reducing caries incidence, approximately 60 to 90% of school-age children worldwide experience caries, with 91% of adult caries in the United States involving the permanent dentition (1, 2). Children of socioeconomically disadvantaged families are a particular concern, as they are twice as likely to experience rampant caries as their wealthier counterparts, and they often present with severe clinical outcomes later in life (3).

Among the early colonizers of the tooth surface is Streptococcus mutans, considered to be the primary causative agent of dental cavities in humans (4). Ongoing research aimed at alleviating or eliminating caries has given rise to valuable insights into the virulence properties of S. mutans, including genes that mediate its obligate biofilm lifestyle, its ability to tolerate acid and oxidative stress, and its ability to maintain metal ion homeostasis (5–12). The introduction of sucrose into the Western diet marked a turning point for S. mutans, which uses sucrose as a substrate for synthesizing extracellular polysaccharides that foster plaque formation and an environment that favors S. mutans-induced tooth decay (13, 14). Taken together, the arsenal of virulence attributes exhibited by S. mutans make it an especially resolute dental pathogen and, in dysbiotic plaque, a primary instigator of caries formation (15, 16).

Among the evolutionary responses that are paramount for S. mutans survival and pathogenesis in dental plaque is tight regulation of essential metal ion transport across the bacterial cell membrane. To this end, S. mutans is endowed with metal ion uptake machinery which enables the import of divalent cations, such as Mn²⁺ and Fe²⁺, that are essential for cellular and subcellular functions, and ultimately for bacterial cell viability. Aberrant metal ion uptake, however, can result in overaccumulation of intracellular metal ions and bacterial cell death owing to Fenton chemistry and the elaboration of toxic oxygen radicals (17–20). To counteract these deleterious effects and achieve intracellular homeostasis, S. mutans has evolved transport mechanisms that function to maintain appropriate metal ion uptake, which is especially paramount during periods of feast and famine in the oral cavity, where Mn²⁺ concentrations can vary greatly.

S. mutans metal ion uptake is mediated in large part by the sloABCR operon, which encodes a SloABC Mn²⁺/Fe²⁺ transporter and, via transcriptional readthrough, a 25-kDa SloR metalloregulatory protein. As a transcription factor, SloR modulates metal ion transport upon binding to DNA in response to manganese availability (7, 20, 21). For instance, between mealtimes, exogenous metal ions are not readily available because they are sequestered in host proteins, such as lactoferrin in saliva. Hence, under fasting conditions, S. mutans upregulates the sloABC gene products, which include ATP-binding and -hydrolyzing proteins as well as a transmembrane SloC lipoprotein that scavenges metal ions for uptake (20, 21). In contrast, during a meal, free metal ions are plentiful in the oral cavity, and in response, S. mutans downregulates its metal ion importers so as to avoid overaccumulation of intracellular metal ions and their associated cytotoxic effects. We believe that SloR mediates such metalloregulation by binding directly to Mn²⁺, which in turn promotes SloR dimerization and a subsequent conformational change at the N terminus of the protein to facilitate DNA binding. Specifically, in a previous report, we described SloR-DNA binding upstream of the sloABC locus to a promoter-proximal SloR recognition element (SRE) that represses sloABC transcription, presumably via a mechanism that involves promoter exclusion of RNA polymerase (RNAP) (22, 23). Hence, the sloABC-encoded metal ion uptake machinery in S. mutans is subject to transcriptional repression by SloR under conditions of Mn²⁺ availability and, conversely, to derepression when Mn²⁺ becomes limiting.

The transcriptional regulatory properties of the SloR protein are not limited to the sloABC locus. In fact, work in our laboratory suggests that the SloR protein, a member of the DtxR family of metalloregulators, may be involved in regulating as many as 200 genes in the S. mutans genome, either directly or indirectly (24). The genes that are subject to SloR control belong to a variety of different functional categories beyond metal ion homeostasis and include genes whose products mediate S. mutans oxidative stress and acid tolerance, biofilm formation, and genetic competence, all of which contribute to S. mutans virulence (7, 8, 24, 25). In addition, accumulating evidence in our laboratory supports SloR as more than just a repressor of S. mutans gene expression. While the results of expression profiling studies support SloR-mediated repression of certain genes (called class I genes), the binding of SloR to other gene loci (called class II genes) can culminate in gene activation (24). The mechanism by which SloR encourages gene transcription is unknown and is under investigation in our laboratory.

Despite the central importance of SloR in promoting S. mutans survival and virulence gene expression, surprisingly little is known of the regulatory mechanism(s) that modulates SloR itself. To date, the regulation of SloR in S. mutans has been shown to be manganese dependent and driven, in part, by the sloABC promoter via transcriptional readthrough of a weak terminator that is located 3′ to the sloC gene (7, 20, 22). Hence, SloR levels that derive from sloABC promoter activity likely vary with Mn²⁺ availability between and during mealtimes. We propose, however, that some constitutive baseline level of SloR is likely necessary to facilitate scavenging of essential Mn²⁺ and/or Fe²⁺ by the SloC metal ion importer, regardless of exogenous metal ion availability, and that such fine-tuning may involve additional mechanisms of control.

In the present study, we set out to determine whether a 184-bp intergenic region (IGR) that is located immediately downstream of the sloC coding region and upstream of the sloR gene harbors a specific promoter that drives sloR expression in a manner that is independent of sloABC promoter activity. Taking this work together with an updated characterization of the sloABC promoter region, we propose an as-yet-unidentified SRE in this IGR and a bifunctional role for SloR as both a repressor and an activator of S. mutans gene expression (26).

RESULTS

SloR homodimers bind to a 72-bp SRE.

In a previous report, we described the thermodynamic properties of the binding of the S. mutans SloR metalloregulator to its cognate SRE within the sloABC promoter region. In that study, the results of fluorescence anisotropy studies conducted with 1 mM manganese and various SRE-containing DNA fragments revealed tight binding of the SloR protein to a 72-bp DNA fragment (K_d [dissociation constant] = 32 nM). Combining these results with the results of electrophoretic mobility shift assays (EMSAs) and DNA footprinting experiments which support the binding of at least two SloR dimers to this 72-bp sequence, we hypothesized that SloR may bind as a set of homodimers to each of three inverted hexameric repeats on this 72-bp target DNA, each with the consensus sequence AATTAA or some modification thereof (Fig. 1). To test this hypothesis, we used the SloR protein, 1 mM Mn²⁺, and the predicted 72-bp SloR recognition element in negative-staining and electron microscopy (EM) experiments. The resulting data set comprised 55 total images and was classified into two-dimensional class averages by use of the PARTICLE software package (www.image-analysis.net/EM/). The dominant pattern revealed by the class averages shares a shape with three distinct ellipsoidal regions, presumed to be SloR dimers (Fig. 1, inset, arrowheads), tilted off the DNA axis by ∼32°. The SloR binding pattern occupies a total length of 239 Å on the DNA, with each SloR dimer measuring ∼90 Å across, and the distance from the center of one dimer to the center of the next measures 75 Å (equal to 22 bp). Taken together, the binding pattern and low-resolution image of the SloR-SRE interaction are consistent with the binding of three SloR dimers to the 72-bp target sequence.

FIG 1.

SloR-SRE binding model for the region of the sloABC promoter. Shown are each of three inverted hexameric repeats (in blue, purple, and red) that span 72 bp of the sloABC promoter region and to which SloR presumably binds as three homodimers. Affinity binding studies support preferential binding of SloR homodimers to the hexameric repeat in region B (designated by the black bar), followed by cooperative binding of SloR dimers to regions A and C. (Inset) Negative staining of the SloR-SRE interaction in vitro supports homodimeric binding of SloR to a 72-bp SRE. Shown are three SloR homodimers bound to a DNA filament containing the 72-bp SloR binding element. The arrowheads denote the center of mass for each SloR homodimer.

SloR binding to the 72-bp SRE is cooperative.

Equilibrium binding of SloR to fluoresceinated duplex DNA-containing sequences from the sloABC promoter region (see Table S1 in the supplemental material) was measured using fluorescence anisotropy with SloR binding to fluorescently labeled duplex DNA containing either one, two, or three 22-bp sequences identified previously on the 72-bp SRE (labeled A, B, and C in Fig. 1). Region B, which includes a pair of inverted repeats, was the only site that demonstrated SloR-specific binding under NaCl concentration conditions as high as 250 mM (data not shown). Sequences A and C, which deviate from the consensus 22-bp sequence at two and three nucleotide positions, respectively, did not measurably bind SloR under the same assay conditions. Saturation binding to these sequences was observed, however, when the salt concentration was lowered to 50 mM, albeit along with some nonspecific interactions. With that said, when SloR was titrated into solutions containing 46-bp duplexes harboring two contiguous pairs of inverted repeats, either sloA_AB or sloA_BC, specific cooperative binding was observed under high-salt conditions, with Hill coefficients of 1.8 and 1.7, respectively. This result indicates that SloR is capable of binding to the A and C sites with high affinity if the B site is already occupied, consistent with SloR interactions at the B site that recruit the additional dimers to the flanking A and C sites. In addition, we measured 50% occupancy of the two sites on sloA_AB and sloA_BC at 26 nM and 10 nM SloR, respectively (data not shown). This result corroborates the observed binding of SloR dimers to all three sites on the 72-bp duplex described above for the negative-staining experiments.

Differential expression of the S. mutans sloABC and sloR genes suggests the presence of a sloR-specific promoter.

To determine whether expression of the sloABC and sloR genes is coordinated under conditions of low versus high manganese concentrations, we performed semiquantitative real-time PCR (qRT-PCR) experiments, the results of which reveal different transcription profiles despite the derivation of these genes from a single polycistronic mRNA (Table 1). Specifically, expression of the sloABC operon was repressed 3-fold under conditions of high manganese availability (P < 0.05 by Student's t test), whereas transcription of the sloR gene was not significantly altered under the same conditions (P > 0.05 by Student's t test). The observation that expression levels of the sloABC and sloR genes are different under high-Mn²⁺ conditions suggests an additional control mechanism for sloR transcription that may involve an independent promoter (Table 1). We predict that this promoter is located within the 184-bp IGR that separates the sloC gene of the sloABC operon from the downstream sloR gene on the S. mutans chromosome (Fig. 2).

TABLE 1.

Fold changes in expression for S. mutans sloABCR

Locus	Fold change in expression
	Expt 1		Expt 2
	UA159 with 125 μM Mn2+	UA159 with 5 μM Mn2+	UA159	GMS611	GMS611d
sloABC	0.33 ± 0.03a	1.02 ± 0.09	1.35 ± 0.10	0.14 ± 0.01b	0.05 ± 0.00b
sloR	0.75 ± 0.09	1.02 ± 0.19	1.35 ± 0.14	0.96 ± 0.03c	0.81 ± 0.10c

^aP < 0.05 by Student's t test.

^bP < 0.0001 by ANOVA with Tukey's test.

^cP < 0.05 by ANOVA with Tukey's test.

FIG 2.

The intergenic region between the sloR and sloC genes harbors a recognizable promoter. The nucleotide sequence of this region was aligned with the S. mutans UA159 genome sequence from the NCBI GenBank database (RefSeq accession number NC_004350.2). The +1 transcription start site (designated by an arrow) marks the transcription start site of the sloR gene as defined by 5′ RACE and defines a 19-bp 5′ untranslated region (UTR). Also shown are the predicted −35 and −10 promoter regions, the predicted ribosome binding site (RBS), and the start codon (SC) of the 654-bp sloR gene. A putative extended −10 element is denoted by the dashed line.

The results of 5′ RACE reveal the location of a transcription start site in the intergenic region between the S. mutans sloC and sloR genes.

To investigate whether a promoter might exist on the IGR that separates the sloABC and sloR genes, we performed 5′ rapid amplification of cDNA ends (5′ RACE) to identify a putative +1 transcription start site. The results revealed that transcription of the 654-bp sloR-specific transcript begins at an adenosine residue located 19 bp upstream of the ATG translation start codon. Mapping the cDNA sequence back to the S. mutans UA159 reference genome allowed us to predict and annotate the −10 and −35 sites of the predicted sloR promoter (Fig. 2). The putative −10 site aligns precisely with the conserved prokaryotic consensus sequence (TATAAT), whereas there is variation in the sequence between the predicted −35 site and the consensus sequence for other prokaryotes. A putative extended −10 element which is characterized by a TGN sequence and the presence of two poly(T) tracts in the spacer region may compensate for degeneracy in the −35 promoter.

The S. mutans sloR gene is transcribed even in the absence of a functional sloABC promoter.

To investigate the impact of promoter/SRE variants on transcription of the S. mutans sloABCR operon, we introduced transition mutations into the 72-bp SRE, at positions 11 and 11d, which overlap the −35 and −10 promoter sites upstream of sloABC, respectively. Notably, T-to-C mutations at these sites in the GMS611 and GMS611d strain variants culminated in significantly compromised sloABC transcription (P < 0.0001 by analysis of variance [ANOVA] with Tukey's test) in qRT-PCR experiments (Table 1), with quantification cycle (C_q) values approaching those for the no-template and reverse transcriptase-negative controls (data not shown). This is consistent with disruption of the sloABC promoter in the GMS611 and GMS611d strain variants. While sloABC transcription in these variants was greatly reduced, transcription of sloR was diminished to a much lesser extent (P < 0.05 by ANOVA with Tukey's test), likely owing to continued transcription from an independent, sloR-specific promoter (Table 1).

The S. mutans sloR gene is transcribed from the sloABC promoter as well as from an independent promoter on the 184-bp IGR.

To determine whether sloR transcription is indeed driven by an independent sloR-specific promoter, we performed reverse transcriptase PCR (RT-PCR) experiments with cDNAs that we generated from the S. mutans GMS611 and GMS611d sloABC promoter variants and their UA159 wild-type progenitor. As noted above, expression of the sloABC operon in GMS611 and GMS611d is virtually nil, indicating successful disruption of the sloABC promoter in these strains. In S. mutans, expression of the sloABC and sloR genes derives from a 3.4-kb polycistronic mRNA owing to transcription of the UA159 chromosome that is driven by the upstream sloABC promoter and transcriptional readthrough of a weak terminator at the 3′ end of the sloC gene (21, 27, 28). We expected to generate a 2.7-kb polycistronic mRNA by RT-PCR given the positioning of the P1 and P3 primers that span the sloABC operon and the downstream IGR (Fig. 3a). In fact, the results of RT-PCR indicate the presence of a 2.7-kb cDNA product for S. mutans UA159 and the absence of this product for GMS611 and GMS611d (Fig. 3b). In addition to this polycistronic mRNA, however, we noted the presence of a 250-bp cDNA product for all three S. mutans strains with the P2/P3 primer pair. The presence of this amplicon in GMS611 and GMS611d indicates the production of a sloR-specific transcript even in the absence of a functional sloABC promoter and supports sloR transcription from an independent promoter. Importantly, PCR products deriving from genomic DNA as the amplification template confirmed the specificity of the respective primer pairs (Fig. 3b).

FIG 3.

The S. mutans sloR gene is transcribed even in the absence of a functional sloABC promoter. (a) Map of the sloABCR operon and locations of primer annealing sites. Primer P1 (sloA.RT_PCR.F) anneals within the sloA coding sequence, and primers P2 and P3 (sloR.RT_PCR.F and sloR.RT_PCR.R, respectively) anneal within the sloR coding sequence. (b) Products of reverse transcriptase PCR resolved in a 0.8% agarose gel. Amplification of cDNA with the P1/P3 primer pair generated a 2,745-bp amplicon for UA159 but not for GMS611 or GMS611d, consistent with disruption of the sloABC promoter in the mutant strains. In contrast, cDNA amplification with the P2/P3 primer pair gave rise to a 250-bp amplicon even for the sloABC promoter mutants GMS611 and GMS611d, indicating the presence of a sloR-specific promoter in the 184-bp intergenic region that separates the sloABC operon from the downstream sloR gene. gDNA, genomic DNA; P/O, promoter/operator.

SloR binds directly to the intergenic region between the S. mutans sloC and sloR genes.

To determine whether the impact of SloR binding on sloR transcription is direct, we performed EMSAs, the results of which support direct SloR binding to the intergenic region between the S. mutans sloC and sloR genes. Specifically, we observed protein-IGR binding when SloR was provided at concentrations as low as 400 nM but not at concentrations of 200 nM or less (Fig. 4a). This contrasts with the SloR binding we observed at the sloABC promoter region, which occurred with as little as 60 nM SloR protein. These findings support SloR binding to the IGR with a lower affinity than that of SloR binding to the sloABC promoter region. In addition, SloR-IGR binding was abrogated upon the addition of 1.5 mM EDTA, consistent with the metal ion dependence of the interaction.

FIG 4.

SloR binds directly to the intergenic region (IGR) between the S. mutans sloC and sloR genes. (a) Results of EMSA, which support direct SloR binding to 204-bp and 155-bp fragments of the sloC-sloR intergenic region at protein concentrations as low as 400 nM. SloR binding to a 95-bp IGR derivative was relatively compromised, however, and was completely absent when a 62-bp deletion derivative was used as the binding template. A 205-bp amplicon that included the sloABC promoter was used as a positive control for SloR binding. EDTA was added to select reaction mixtures in an attempt to abrogate metal ion-dependent binding. Nondenaturing polyacrylamide gels (12%) were subjected to electrophoresis at 300 V for 1.5 h. Film exposure in the presence of an intensifying screen proceeded for 48 h at −80°C before development. (b) SloR binding to serial deletion fragments of the S. mutans IGR. The arrowheads facing inward represent AATTAA hexameric repeats to which SloR putatively binds. The vertical bars denote the positioning of the −10 and −35 promoter sequences of the sloR-specific promoter. Whether or not SloR binds to the IGR fragment is shown with a “+” or “−” designation.

We also used EMSA to determine the region on the IGR to which SloR binds. To this end, we generated a series of IGR fragments with serial deletions at the 5′ or 3′ end (Fig. 4a). Interestingly, robust band shifts were generated with DNA fragments harboring promoter-distal hexameric repeat sequences, located at least 62 bp upstream of the +1 transcription start site, but not with DNA fragments less than 62 bp from the +1 start site, which lack these repeat sequences (Fig. 4b).

Expression of the S. mutans sloR gene is subject to positive autoregulation.

Direct binding of SloR to promoter-proximal sequences at the sloABC locus and to promoter-distal sequences at the sloR locus, coupled with their negative and positive effects on global gene expression, respectively, led us to hypothesize a role for SloR as a bifunctional regulator in S. mutans. To validate such a dual role for SloR, we performed in vitro transcription (IVT) experiments, the results of which demonstrate unequivocally that sloABC transcription is repressed by SloR, while transcription of the sloR gene is facilitated by SloR (Fig. 5). These findings, which were quantified by use of ImageJ software and in combination with the results of binding studies, indicate that SloR can either repress or encourage gene expression via direct binding to DNA. To our knowledge, this is the first experimental evidence to demonstrate bifunctional regulation of gene expression by the SloR metalloregulator in S. mutans.

FIG 5.

Results of sloABCR transcription experiments performed in vitro (IVT) support the hypothesis that SloR is a bifunctional regulator. Transcription of the sloABC and sloR genes was quantified using ImageJ software. SloR (75 nM) was added to select reaction mixtures containing an RNase inhibitor and either sloA or sloR template DNA. Pixel counting was performed with ImageJ software and does not include unincorporated [α-³²P]UTP. Higher pixel counts indicate greater band intensities. The data shown are the results of a single representative experiment (from a total of three independent experiments) and support repression of sloA and activation of sloR transcription by the SloR metalloregulator.

DISCUSSION

Until recently, work in our laboratory focused primarily on understanding the mechanism(s) of SloR binding at the sloABC locus, which is subject to transcriptional repression by SloR. Herein we present evidence that supports an updated model for SloR binding at this locus, consistent with the direct binding of SloR homodimers to binding sites that span a 72-bp region of DNA which includes the −10 and −35 sloABC promoter regions. In previous work (22), we described a pattern for SloR binding at the sloABC promoter site that involved two SloR dimers binding to two sets of inverted repeats, each 6 bp long and separated by 8 bp. Subsequent analysis of the 72-bp region upstream of the sloABC operon suggested that, in fact, SloR binds to three distinct but abutting 22-bp sites in that region, referred to herein as sites A, B, and C. Each binding site is comprised of two inverted hexameric repeats with an AATTAA consensus separated by 6 bp, thereby defining a putative 6-6-6 motif for SloR binding (Fig. 1). This sequence pattern (xxAATTAAxxxxxxTTAATTxx, where “x” is a nonconserved nucleotide) is similar to that of the SloR homolog in S. gordonii, called ScaR, which binds to two adjacent 22-bp sites upstream of the scaABC operon (29) and for which the binding pattern was confirmed by negative staining and electron microscopy (unpublished observations). We therefore expanded what we previously thought to be a 42-bp SRE in the sloABC promoter region to include at least 30 additional base pairs.

While SloR binds to the central 22-bp SRE (region B) with strong affinity when provided as a template in isolation, we measured cooperative interactions between SloR homodimers bound to adjacent SRE sequences (regions A and C). These cooperative interactions are strong enough to support high-affinity binding between SloR homodimers, presumably because of interactions involving the initial binding of SloR homodimers to the B site. Such cooperativity has likewise been observed for the S. gordonii ScaR protein (27), suggesting that this property may be a common feature among interactions involving other streptococcal manganese-dependent regulators and their promoter/operator sequences.

In the present study, we extend our observations on SloR binding to include the details of protein binding to the IGR that immediately precedes the S. mutans sloR gene. Based on accumulating evidence presented herein, we propose that the location of the SloR-DNA binding element relative to the promoter sequences that modulate downstream sloABC and sloR gene transcription contributes to SloR's ability to differentially downregulate sloABC promoter activity and upregulate sloR promoter activity. Interestingly, an in silico analysis of the 184-bp IGR that precedes the S. mutans sloR gene failed to reveal a recognizable SRE like the one we described above for the sloABC locus, consistent with differential regulation by SloR at these two loci.

The transcription start site for the S. mutans sloR gene occurs within the 184-bp IGR that separates sloR from the sloABC operon immediately upstream, as determined by 5′ RACE experiments. Based on the +1 transcription start site, −10 and −35 promoter regions were predicted, and a 19-bp 5′ untranslated region (UTR) was defined. Notably, the hexameric −10 region shares 100% sequence identity with the canonical prokaryotic consensus sequence for a −10 promoter (TATAAT) (30, 31). An in silico analysis of the sloR promoter revealed a putative extended −10 element in the IGR which is absent from the sloABC promoter region. Reports in the literature describe such a TGN motif immediately upstream of the −10 sequence as an element that may facilitate downstream transcription by stabilizing the open complex during initiation and by shortening the distance between the −10 and −35 sites (29, 32). The contact that RNA polymerase makes with the nonamer that defines the extended −10 site may compensate for the suboptimal contact that the polymerase makes with the degenerate −35 sequence (32). The TGN motif that we noted on the sloR-containing IGR is similarly located 14 to 16 nucleotides upstream of the transcription start site. Previous studies have also noted that short poly(T) tracts are characteristic of the spacer region of Escherichia coli TG promoters (33). We similarly noted the presence of two poly(T) tracts, centered at positions −18 and −29, in the spacer region of sloR's TG promoter.

In contrast to the −10 promoter region, the predicted −35 site (TATCCA) shares only 50% sequence identity with the typical prokaryotic promoter sequence (TTGACA) (30). This is not surprising given frequent reports of sequence variation in and around the −35 promoter region within and across prokaryotic species (34). Since promoter strength is determined in part by conservation of the −10 and −35 promoter sequences (31), one might expect RNAP to have only moderate binding affinity for the relatively divergent sloR promoter (TATAAT and TATCCA) compared to that of RNAP binding at the more highly conserved sloABC promoter (TATATT and TTGACT) (22), and accordingly, weaker transcription from the former than from the latter. The results of DNA binding and expression profiling experiments reported herein support these predictions and suggest a role for divergent sloABC and sloR promoter sequences in fine-tuning metal ion transport and minimizing the toxic effects of metal ion hyperaccumulation. Additional layers of gene control involving SloR likely evolved at these loci given the importance of maintaining metal ion homeostasis under conditions as transient as those in the oral cavity.

The absence of a recognizable SloR binding motif in the IGR that precedes sloR is consistent with differential gene regulation and SloR binding at this locus versus those at the sloABC locus. That is, three adjacent palindromes on the 72-bp SRE appear to be absent from the IGR, although a pair of consensus palindromes with the sequence AATTAA appear to be uniquely located 44 to 50 bp and 94 to 100 bp distal to the sloR promoter. Interestingly, reports in the literature describe AT-rich sequences, including palindromes like those in the sloABC and sloR promoters, that can engender intrinsic curvatures in the DNA (24, 33). To assess inherent DNA curvature in the sloABC and sloR promoter regions, we applied a BEND algorithm (35) to the 72-bp sloABC SRE and the 184-bp IGR, the results of which revealed high-fidelity alignment of AT-rich palindromes with predicted sites for SloR binding (data not shown). Hence, DNA curvature that localizes to the paired palindromic repeats at the sloABC and sloR loci supports a SloR-DNA interaction that is not strictly defined by nucleotide sequence specificity but by the DNA conformation as well. Studies are under way to determine what impact, if any, these AT-rich palindromes may have on SloR-DNA binding and whether the DNA curvature they can instigate contributes to SloR's function as a repressor versus an enhancer of gene transcription.

Differential expression of the sloABC and sloR genes was especially pronounced in IVT assays in which we used the 5′ end of the sloA or sloR coding region and up to 200 bp of upstream DNA sequence as the DNA template. Pixel counting of the resulting mRNA transcripts on autoradiograms, performed with ImageJ, revealed considerably more robust sloR transcription in the presence of exogenous SloR than in its absence (Fig. 5). This contrasts with the in vitro transcription we observed for the upstream sloABC operon, which, as expected, was repressed by the presence of SloR. Taken together with the EMSA results that support direct SloR binding at these loci, these data demonstrably support a role for SloR as a bifunctional regulator of S. mutans gene transcription. While the mechanism for sloABC repression likely derives from SloR binding to an SRE that overlaps the sloABC promoter, thereby excluding RNAP from promoter access, sloR transcription is the likely result of derepression with SloR binding to promoter-distal sites. In future work, we will consider the Mn²⁺ status as a potential contributor to differential SloR binding and gene transcription outcomes. In fact, binding of the MntR metalloregulator to different sequences upstream of the mntABCD locus in Bacillus subtilis is known to be Mn²⁺ concentration dependent (36).

In summary, the results of the present study support SloR-mediated transcriptional events at the sloR locus that are different from those at the sloABC locus and lend further support to a role for SloR as a bifunctional regulator of gene transcription. It is tempting to suggest that the sloR-specific promoter on the 184-bp IGR evolved to ensure at least some level of constitutive SloR production. Accordingly, when free metal ions are introduced into the oral cavity during a meal, S. mutans is poised and ready to modulate the controlled uptake of the exogenous Mn²⁺ and Fe²⁺ it needs for survival. Hence, while the sudden introduction of metal ions into the mouth may prove damaging to some constituents of the oral microbiota, S. mutans can exploit these conditions by use of a metal ion-dependent SloR regulator that can repress the cytotoxic effects of excessive metal ion import while maintaining baseline levels of SloR from an ancillary promoter. Such fine-tuned gene regulation can have an impact on cell function beyond the scope of metal ion homeostasis and can influence processes, such as adherence, acid production, and the oxidative stress response, that more directly contribute to S. mutans-induced disease (7, 8, 24, 25). In conclusion, we propose that S. mutans coordinates the regulated expression of its metal ion transport machinery with that of its virulence attributes. An improved understanding of sloR autoregulation is significant because it can elucidate the mechanisms that fine-tune the regulated expression of metal ion homeostasis and virulence in an important oral pathogen. Moreover, from these investigations, we may better elucidate the details of SloR-mediated gene regulation that can benefit the design of an anti-SloR therapeutic aimed at alleviating and/or preventing caries.

MATERIALS AND METHODS

Bacterial strains, plasmids, and primers.

The bacterial strains used in this study are listed in Table 2. Working stocks of bacterial strains were prepared from overnight cultures and stored in 20% or 50% sterile glycerol at −20°C or −80°C, respectively.

TABLE 2.

Bacterial strains used in this study

Strain	Description or relevant characteristics	Source or reference
E. coli BL21(DE3)	hA2 [lon] ompT gal(λ) (DE3) [dcm] ΔhsdS	Thermo Fisher Scientific
S. mutans UA159	Wild type, serotype c; ATCC 700610	ATCC
S. mutans GMS602	UA159-derived strain; contains IFDC2 cassette in the sloA promoter region; Emr 4-Cl-Phes	22
S. mutans GMS611	Contains a markerless T → C mutation within the −35 region of the sloABC promoter	This study
S. mutans GMS611d	Contains a markerless T → C mutation within the −10 region of the sloABC promoter	This study

The primers used in this study are shown in Table S1 in the supplemental material. All primers were designed using the Primer BLAST tool from the NCBI website. A RefSeq record was used as the input, with forward and reverse primer locations specified by the user. Lack of secondary structure was confirmed using the Oligo Evaluator tool (Sigma), and primers were checked for specificity against the S. mutans UA159 genome by use of the NCBI Basic Local Alignment Search Tool (BLAST).

Negative staining and EM analysis.

A SloR–Mn²⁺–72-bp SRE complex was prepared in vitro, assuming a 3:1 stoichiometry of dimers to DNA. The reaction mixtures were then systematically diluted to 25, 50, and 100 nM for negative-stain EM grid preparation using 1% uranyl acetate. After screening for the optimal staining quality and particle concentration, single-particle data were collected on the 50 nM specimen grid by use of an FEI T12 electron microscope at 120 keV and a nominal magnification of ×68,000, producing 108 micrographs at 3.15 Å/pixel in the image, with various defocus values between 0.9 and 1.8 μm. Next, 8,500 particles of the SloR-DNA complex were selected for two-dimensional (2D) classification using the PARTICLE (www.sbgrid.org/software/titles/particle) program.

Construction of S. mutans promoter variants.

To generate specific mutations in the promoters that drive sloABC transcription, we adopted a markerless mutagenesis strategy similar to that described by Xie et al. (37). This involved constructing promoter variants with an IFDC2 cassette inserted within the sloABC promoter region for subsequent competence-stimulating peptide (CSP) transformation into S. mutans UA159 to generate the erythromycin-resistant and p-4-chlorophenylalanine-sensitive GMS602 strain. The double-crossover event was confirmed by PCR and Sanger sequencing. Derivatives of GMS602 were generated by overlap extension PCR (OE-PCR) with reverse primers harboring a single point mutation in the predicted 72-bp SRE that precedes the sloABC genes (Table S1). S. mutans GMS611 was generated by use of degenerate primers that introduced a point mutation into the SRE at nucleotide position 11, which we predict to overlap the sloABC-specific −35 promoter region (22). Likewise, GMS611d was generated by use of a different degenerate primer set that introduced a point mutation into the SRE at position 11d, which we predict overlap the sloABC-specific −10 promoter region. Thymine-to-cytosine transition mutations were generated in both S. mutans strains and validated by Sanger sequencing (22).

Chromosomal DNA isolation.

S. mutans grown overnight at 37°C and 5% CO₂ in 14 ml Todd-Hewitt yeast extract (THYE) broth was pelleted by centrifugation at 7,000 rpm for 5 min in a Sorvall RCB centrifuge, after which the cells were resuspended in 1 ml Tris-EDTA buffer (10 mM Tris, 1 mM EDTA) and chemically disrupted according to established protocols (8, 22). Genomic DNA was purified by subsequent rounds of phenol-chloroform extraction, ethanol (EtOH) precipitated, and resuspended overnight in nuclease-free water (Ambion) at 4°C with gentle agitation (38). The nucleic acid yield and purity were assessed by use of a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific), and the samples were stored at −20°C.

RNA isolation.

RNA was isolated from S. mutans strains according to established protocols (8). Cells were grown to mid-logarithmic phase (optical density at 600 nm [OD₆₀₀] = 0.4 to 0.6) before pelleting by centrifugation as described above and resuspension in RNAProtect (Qiagen). Total intact RNA was purified following cell lysis and DNase I treatment by use of a Qiagen RNeasy kit, after which the nucleic acid yield and purity were assessed by use of a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific). RNA samples were analyzed for integrity by agarose gel electrophoresis and were stored at −80°C.

RT-PCR.

Reverse transcriptase PCR was performed with cDNAs that were reverse transcribed from RNAs by use of a First Strand cDNA synthesis kit according to the recommendations of the supplier (Thermo Fisher Scientific) or else with chromosomal DNA isolated from S. mutans UA159, GMS611, or GMS611d according to a modification of the method of Idone et al. (39). Q5 Hi-Fi polymerase was used for PCR in accordance with the recommendations of the supplier (New England BioLabs). Each 50-μl PCR mixture consisted of 10 μl 5× Q5 reaction buffer, 1 μl deoxynucleoside triphosphates (dNTPs; 10 mM [each]), 2.5 μl sloA.RT_PCR.F or sloR.RT_PCR.F, 2.5 μl sloA.RT_PCR.R or sloR.RT_PCR.R (Table S1), 200 ng of genomic DNA or 1 μl of cDNA product, nuclease-free water up to 49.5 μl, and 0.5 μl of Q5 Hi-Fi polymerase (added to the reaction mixture just prior to amplification). PCR conditions were as follows: initial denaturation at 98°C for 30 s followed by 25 cycles of denaturation at 98°C for 10 s, annealing at 67°C for 30 s, and extension at 72°C, concluding with a final extension at 72°C for 80 s. An aliquot of each sample was visualized in 0.8% agarose gel as described above.

5′ RACE.

To reveal the sloR transcription start site and predict the −10 and −35 regions of the sloR promoter, we performed 5′ RACE with an Invitrogen 5′ RACE kit in accordance with the manufacturer's protocol, unless otherwise specified. RNAs isolated from S. mutans GMS611 and GMS611d were reverse transcribed into cDNAs by use of the sloR.[R].GSP1.A reverse primer and 200 U of Superscript II reverse transcriptase. After S.N.A.P. column purification, the cDNA was poly(dC) tailed using terminal deoxynucleotidyltransferase as described in the manufacturer's protocol. The cDNA products were PCR amplified using the kit-supplied abridged anchor primer (AAP), the sloR.[R].GSP2.B reverse primer, and Platinum Hi-Fi Taq polymerase in a Bio-Rad PCR machine programmed to perform 94°C for 2 min followed by 35 cycles of 94°C for 15 s, 58°C for 30 s, and 68°C for 1 min, with a final hold at 4°C. An aliquot of the resulting PCR product was analyzed by agarose gel electrophoresis, and the remainder was purified using a MinElute PCR purification kit in accordance with the supplier's recommendations (Qiagen). The purified amplicons were quantified using a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific) and sequenced (Eurofins) using the reverse primer sloR.GSP2.A or sloR.nested.GSP. The 5′ sequence of the mRNA transcript was aligned with the S. mutans UA159 reference genome (RefSeq accession number NC_004350.2) from the NCBI database to identify the nucleotide that immediately follows the poly(dC) tail as the transcription start site (+1 site), predict the −10 and −35 promoter sequences in the 184-bp intergenic region, and reveal the 5′ untranslated region of the sloR transcript (40).

EMSA.

EMSAs were performed according to established protocols to determine whether SloR binding to the intergenic region upstream of the sloR gene is direct and to narrow down the region of SloR binding at this locus (8, 22, 23). Primer design for the DNA binding template spanned the 184-bp intergenic region (IGR) between sloC and sloR and included serial deletions thereof. PCR amplification was performed with Q5 polymerase according to the manufacturer's protocol (New England BioLabs), using the following thermal cycling conditions: initial denaturation at 98°C for 30 s, 35 cycles of 98°C for 10 s, annealing at 60°C for 30 s, and extension at 72°C for 30 s, and a final extension at 72°C for 2 min. The resulting amplicons were PCR purified as described above, confirmed by agarose gel electrophoresis, and quantified by NanoDrop spectrophotometry.

The resulting amplicons were end labeled with [γ-³²P]dATP (Perkin-Elmer) in the presence of T4 polynucleotide kinase (New England BioLabs), after which they were centrifuged through a TE Select-D G-25 spin column (Roche Applied Science) to remove the unincorporated [³²P]dATP. Binding reaction mixtures were prepared as described previously (23) in a 16-μl total volume containing 1 μl of end-labeled amplicon, purified native SloR protein at concentrations ranging from 60 nM to 400 nM, and 3.2 μl of 5× binding buffer (42 mM NaH₂PO₄, 58 mM Na₂HPO₄, 250 mM NaCl, 25 mM MgCl₂, 50 μg/ml bovine serum albumin, 1 mg sonicated salmon sperm DNA, 50% [vol/vol] glycerol, and 37.5 μM MnCl₂). EDTA was added to a separate reaction mixture at a final concentration of 1.5 mM to validate that SloR binding is metal ion dependent. An end-labeled sloA promoter-containing amplicon was used as a positive control for SloR binding (7, 22, 23). Samples were loaded onto 12% nondenaturing polyacrylamide gels (3 ml 20× bis-Tris-borate [pH 7.4], 74 μl 100 nM MnCl₂, 1.5 ml 100% glycerol, 24 ml 30% acrylamide [37.5:1 acrylamide-bis], 31 ml Millipore H₂O, 300 μl 15% ammonium persulfate [APS], 90 μl TEMED [N,N,N′,N′-tetramethylethylenediamine]) and resolved at 300 V for 1.5 h. Gels were exposed to Kodak BioMax film for 24 to 72 h at −80°C in the presence of an intensifying screen prior to autoradiography.

Fluorescence anisotropy.

Equilibrium binding of SloR to regions within the sloABC promoter/operator region was probed by titrating SloR onto fluoresceinated DNA and monitoring binding by fluorescence anisotropy. Duplex DNA fragments containing relevant sequences were prepared from oligonucleotides obtained from Integrated DNA Technologies (Coralville, IA). A 5′-fluoresceinated oligonucleotide was annealed with a 10% excess of its unlabeled complementary strand in 25 mM HEPES, pH 7.9, 50 mM NaCl by heating to 90°C and cooling slowly to room temperature (Table S1). Titrations were performed in either high- or low-salt buffer; the high-salt buffer contained 25 mM HEPES, pH 8.0, 250 mM NaCl, 10% glycerol, and 1 mM MnCl₂, whereas the low-salt buffer contained 25 mM HEPES, pH 7.9, 50 mM NaCl, and 1 mM MnCl₂. SloR was titrated into 1 ml of 1 nM fluoresceinated duplex DNA in buffer, and anisotropy measurements were made at 25°C by use of a Beacon 2000 fluorescence polarization instrument. Data were fit to one of several equations, related to simple 1:1 binding stoichiometry if the K_d was >10 nM (equation 1) or <10 nM (equation 2) or to the Hill equation (equation 3), as follows:

$$r\mathrm{=\Delta }r\times \frac{P}{K_d+P}+r_{\min }$$ (1)

$$r\mathrm{=\Delta }r\times \frac{K_d+D+P-\sqrt{{(K_d+D+P)}^2-4\mathrm{DP}}}{2D}+r_{\min }$$ (2)

$$r\mathrm{=\Delta }r\times \frac{P^n}{K_h^n+P^n}+r_{\min }$$ (3)

where r is anisotropy, Δr is the total change in anisotropy at saturation, K_d is the dissociation constant, K_h is the concentration of SloR dimers giving 50% maximal binding under cooperative conditions, n is the Hill coefficient, r_min is the anisotropy obtained before addition of SloR, P is the concentration of SloR dimers, and D is the concentration of duplex DNA. In addition, where nonspecific binding prevented signal saturation, a term to model the slow linear increase in anisotropy was added to equation 1 or 2, with the form K_nsP, where K_ns is the nonspecific binding constant. All curve fitting was performed in R software.

Preparation of S. mutans RpoD.

An E. coli DH5α strain harboring plasmid pIB611 (a kind gift from Indranil Biswas, University of Kansas) was streaked for isolation on L-agar plates supplemented with ampicillin (100 μg/ml) and incubated overnight at 37°C. Resident on pIB611 is the S. mutans rpoD gene, cloned directly downstream of an inducible operon on vector pET-23d(+) and upstream of a C-terminal His tag (41). Importantly, the 6×His tag was shown in in vitro transcription experiments to have no physiological impact on RpoD functionality (41). Plasmid pIB611 was purified by use of a Qiagen miniprep kit according to the recommendations of the supplier and mapped by restriction digestion (New England BioLabs).

Next, pIB611 was used to transform E. coli BL21(DE3) cells in accordance with the manufacturer's protocol (Invitrogen). Successful transformants were selected after overnight growth on L-agar plates supplemented with 100 μg/ml ampicillin and used to inoculate a 25-ml starter culture for yet another overnight incubation. This culture was then used to inoculate 500 ml of prewarmed l-ampicillin (100 μg/ml) broth in a 2.8-liter Fernbach flask, which was incubated at 37°C with continuous shaking at 200 rpm. When the cells reached mid-log phase (OD₆₀₀ = 0.4), isopropyl-β-d-thiogalactopyranoside (IPTG) was added to a final concentration of 0.5 mM to induce expression of T7 RNA polymerase in the BL21(DE3) cells. Protein induction proceeded with continuous shaking for an additional 3.5 h, after which the cells were centrifuged as previously described and stored as dry pellets at −80°C.

Cell pellets were resuspended in His binding buffer (0.5 M NaCl, 20 mM Tris-HCl, 5 mM imidazole) with Halt EDTA-free protease inhibitor at a 1× concentration (Thermo Fisher Scientific). The cell suspension was sonicated (model 2000U; Ultrasonic Power Corporation) at 60% power for six 30-s cycles, using 0.5-s pulses, with samples maintained on ice between runs. Cells were pelleted by centrifugation at 10,000 × g for 30 min at 4°C, after which aliquots of the supernatant and 2× Laemmli buffer (4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue, and 0.125 M Tris-HCl, pH 6.8) were mixed in equal proportions and resolved in 10% Bis-Tris polyacrylamide gels in morpholinepropanesulfonic acid (MOPS) buffer. Proteins were visualized by use of Sypro Ruby gel stain (Thermo Fisher Scientific) according to the manufacturer's instructions. Polyacrylamide gels were fixed for 45 min in fixative solution (50% methanol, 7% acetic acid) with gentle shaking on an orbital shaker. The fixative was subsequently decanted and replaced with 80 ml of Sypro Ruby gel stain. Gels were covered and left to stain on an orbital shaker overnight at room temperature. The gel stain was decanted and wash solution (10% methanol, 7% acetic acid) added before UV visualization.

The remaining cell lysate was further purified by Ni²⁺-nitrilotriacetic acid (Ni-NTA) column chromatography (Thermo Fisher Scientific) at 4°C according to the manufacturer's instructions. The columns were placed on a rotating platform for 30 min at 4°C to encourage RpoD binding to the Ni-NTA resin before centrifugation to remove unbound protein from the column. After elution, the protein yield was determined using both NanoDrop Lite spectrophotometry and a bicinchoninic acid (BCA) protein determination assay (Thermo Fisher Scientific). RpoD purity was assessed in SDS-PAGE gels following Sypro Ruby staining. Select fractions containing RpoD were dialyzed using G2 Slide-A-Lyzer cassettes with a 10-kDa cutoff (Thermo Fisher Scientific) in dialysis buffer (25% glycerol in 1× phosphate-buffered saline [PBS]). The RpoD concentration was assayed as described above, and RpoD was stored at −20°C.

In vitro transcription.

In vitro transcription was performed according to an adaptation of the method of Kajfasz et al. (9). First, genomic DNA spanning approximately 100 to 200 bp of the sloA or sloR coding region and about 150 to 200 bp of upstream sequence was amplified by PCR with Q5 polymerase (New England BioLabs) and either PM.IVT.sloA.F and PM.IVT.sloA.R or PM.IVT.sloR.F and PM.IVT.sloR.R (Table S1) according to the following thermal cycling conditions: initial denaturation at 98°C for 30 s, 35 cycles of 98°C for 10 s, annealing (60°C for sloA and 67°C for sloR) for 30 s, and extension at 72°C for 30 s, and a final extension at 72°C for 2 min. The resulting amplicons were PCR purified as described earlier and confirmed by agarose gel electrophoresis. The DNA concentration was determined spectrophotometrically on a NanoDrop Lite spectrophotometer. Next, in 1.5-ml microcentrifuge tubes, 10 nM DNA template (sloA or sloR), 1 U E. coli RNA polymerase core enzyme (New England BioLabs), 20 U of SUPERase RNase inhibitor (Thermo Fisher Scientific), 25 nM S. mutans RpoD extract (based on BCA assay), and 75 nM purified SloR or no SloR were mixed in reaction buffer (10 mM Tris-HCl [pH 8.0], 50 mM NaCl, 5 mM MnCl₂, 50 μg/ml bovine serum albumin) to yield a final volume of 17.8 μl and then incubated at 37°C for 10 min. To generate an mRNA transcript, 2.2 μl of nucleotide mixture (200 μM ATP, 200 μM GTP, 200 μM CTP, and 10 μM UTP) and 5 μCi [α-³²P]UTP (PerkinElmer) were then added, and the reaction mixture was incubated at 37°C for 10 min. Ten microliters of stop solution (1 M ammonium acetate, 0.1 mg/ml yeast tRNA [Ambion], 0.03 M EDTA) was added to terminate transcription, after which 90 μl of ice-cold 99% EtOH was added for ethanol precipitation overnight at 4°C. The following day, samples were pelleted at 16,200 × g for 30 min, followed by three rounds of washing with 70% EtOH, with additional centrifugation between washes. A final wash with 99% EtOH was performed, after which the samples were lyophilized in a vacuum centrifuge (Eppendorf) for 10 min. The radiolabeled cell pellet was resuspended in 5 μl of formamide dye (0.3% xylene cyanol, 0.3% bromophenol blue, and 12 mM EDTA, dissolved in formamide), and the samples were heated to 70°C in a water bath for 2 min before being placed on ice for gel loading. Samples were resolved in a Novex 10% TBE-7 M urea gel in 1× TBE (10.8 g Tris, 5.5 g boric acid, 0.01 M EDTA [pH 8.0]), and mRNA transcripts were visualized via autoradiography using Biomax XAR film (Thermo Fisher Scientific) exposed for up to 4 h at −80°C in the presence of an intensifying screen. Film was developed according to standard protocols, and ImageJ software was used to quantify the band intensities for the samples.

Semiquantitative real-time PCR (qRT-PCR).

Total intact RNA was isolated from mid-logarithmic-phase S. mutans cultures of the wild-type UA159 strain grown in a semidefined medium (SDM) (22) supplemented with either 5 μM (low) or 125 μM (high) MnSO₄. The resulting RNAs were analyzed for integrity in 0.8% agarose gels before reverse transcribing 100 ng of each RNA sample into cDNA as described above. The cDNAs were used as templates for qRT-PCR, which was performed in accordance with established protocols (22) in a CXR thermal cycler (Bio-Rad). Specifically, sloABC and sloR transcription was assessed in three independent qRT-PCR experiments, each performed in triplicate and normalized against the expression of gyrA (8, 22).

To assess the impacts of transition mutations in the SRE that precedes the sloABC operon on downstream sloR transcription, S. mutans GMS611 and GMS611d were grown as described above, but without supplemental Mn²⁺. Total intact RNA was isolated and reverse transcribed as described above, and the results of qRT-PCR were normalized against those for hk11, whose expression does not change under the experimental test conditions.