Regulation of Recombination between gtfB/gtfC Genes in Streptococcus mutans by Recombinase A

Streptococcus mutans produces 3 types of glucosyltransferases (GTFs), whose cooperative action is essential for cellular adhesion. The recombinase A (RecA) protein is required for homologous recombination. In our previous study, we isolated several strains with a smooth colony morphology and low GTF activity, characteristics speculated to be derived from the GTF fusions. The purpose of the present study was to investigate the mechanism of those fusions. S. mutans strain MT8148 was grown in the presence of recombinant RecA (rRecA) protein, after which smooth colonies were isolated. The biological functions and sequences of the gtfB and gtfC genes of this as well as other clinical strains were determined. The sucrose-dependent adherence rates of those strains were reduced as compared to that of MT8148. Determination of the sequences of the gtfB and gtfC genes showed that an approximately 3500 bp region was deleted from the area between them. Furthermore, expression of the recA gene was elevated in those strains as compared to MT8148. These results suggest that RecA has an important role in fusions of gtfB and gtfC genes, leading to alteration of colony morphology and reduction in sucrose-dependent adhesion.


Introduction
Streptococcus mutans is known to be a primary causative agent of dental caries in humans [1]. One of the important virulence properties of the bacterium is its ability to form biofilm, known as dental plaque, on tooth surfaces [2]. S. mutans has been shown to produce 3 types of GTFs (GTFB, GTFC, and GTFD), whose cooperative action is essential for cellular adhesion [3]. Environmental conditions encountered by S. mutans in dental biofilms are highly variable, including frequent shifts in pH from above 7.0 to as low as 3.0 during the ingestion of dietary carbohydrates by the host [4]. Thus, pH exerts significant ecological pressure on S. mutans, and its ability to tolerate and grow in low pH environments is crucial for its survival and pathogenicity.
Recombinase A (RecA) is essential for transformation of both plasmid and chromosomal DNA in Streptococcus pneumoniae [5]. The recA gene is required for genetic transformation and is directly regulated by the cell signaling mechanism that induces competence in S. pneumoniae [6]. In addition, homologous recombination is strictly dependent upon the presence of the RecA protein [7]. However, RecA function remains uncharacterized due to the complexity of the recombination process. RecA-dependent recombination of the gtfB and gtfC genes has been reported to occur at a frequency of 10 −3 in S. mutans [8], while these activities of GTFB and GTFC were found to be significantly reduced as compared to the reference strains [9]. On the other hand, it was also reported that spontaneous gtfB-gtfC recombination in S. mutans is not dependent on RecA and that a variety of in vivo generated gtfB-gtfC recombinants have similar sites of recombination [10]. However, biofilm formation by a RecA-deficient mutant strain was reduced as compared to that of the parental strain [11], and RecA was shown to have a relationship with expressions of the genes or proteins involved in the response to pH level by S. mutans, as well as expressions of GTFB and GTFC activities [11,12]. In the present study, we demonstrated that the recA gene influences gtf expression, while the relationship between the recA gene and gtfB-gtfC recombination was shown using recombinant RecA protein and clinical isolates.

Bacteria Strains.
The strains used in this study are listed in Table 1. All of the procedures in the present study were approved by the Ethical Committee of the Osaka University Graduate School of Dentistry. Each was grown in brain heart infusion (BHI) broth, Todd Hewitt (TH) broth, or Mitis-Salivarius (MS) agar (Becton-Dickinson, Franklin Lakes, NJ, USA).

Assay for Recombination Frequency.
The test strains were grown in TH broth for 18 hours and inoculated into a 1/100 volume of fresh TH broth. rRecA protein (2 mg/mL) was added to the experimental group, then the strains were grown for 18 hours at 37 ∘ C in TH broth. Finally, they were inoculated onto MS agar plates and cultured anaerobically at 37 ∘ C for 48 hours to confirm culture purity and colony morphology. This experiment was independently repeated 3 times.

Expression of GTF.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and western blot analyses were performed to determine the expression of the gtf gene in the tested strains, according to a method previously described by Aoki et al. [16]. The tested organisms were grown in BHI broth at 37 ∘ C to an OD550 value of 1.0. Cells were collected by centrifugation and the pellet was washed with PBS, then the bacterial cells were dissolved in SDS gel loading buffer. Next, an equal amount of each protein was separated by 7% SDS-PAGE and transferred NN2147 c smooth GTFBC fusion strain, isolated from 10-year-old girl [15] onto a polyvinylidene difluoride membrane (Immobilon, Millipore, Billerica, MA, USA). GTFB and GTFC were detected using a rabbit anti-GTF antibody [17], as well as swine anti-rabbit immunoglobulin conjugated with alkaline phosphatase (Dako, Glostrup, Denmark).

Determination of GTF Activity.
The enzyme activities of GTF protein were determined using polyacrylamide gels, as previously described [18]. Briefly, the strains were grown to the same OD550 value of 1.0 and cells were collected. Next, the cells were washed with PBS, then resuspended in phosphate-buffered saline buffer and adjusted to the same OD550 of 1.0. Fifteen microliters of each cell suspension was run on 7.5% SDS-PAGE gels. After electrophoresis, the gels were incubated overnight at 37 ∘ C in 3% sucrose, 0.5% Triton X-100, and 10 mg/mL dextran T10 in 10 mM sodium phosphate, pH 6.8, at 37 ∘ C, and the resulting glucan bands were treated with periodic acid and pararosaniline. The intensities of the stained bands were used to determine the activities of the GTF proteins.
2.6. Sucrose-Dependent Adherence Assay. Sucrose-dependent adherence of S. mutans growing cells was determined using a turbidimetric method, as previously reported [14].
The test strains were grown in BHI broth containing 1% sucrose at 37 ∘ C for 18 hours at a 30 ∘ angle. After incubation, the culture tubes were vigorously vibrated with a vortex mixer for 3 seconds, and nonadhesive cells were transferred to fresh tubes. Cells remaining on the glass surface (adhesive cells) were removed by a rubber scraper and suspended in 3 mL of distilled water. Both adhesive and non-adhesive cells were dispersed by ultrasonication; then the turbidity of the suspension was determined by reading optical density at 550 nm. The cells were defined as OD550 (adhesive cell plus non-adhesive cells) and the percentage of adherence was defined as 100 × OD550 (adhesive cells)/OD550 (total cells). All assays were performed 3 times, with the mean and standard deviation presented.  (Table 2). Data analysis was performed using Gene Works software (IntelliGENETICS). The sequences of each strain were compared using multiple alignment analysis with CLUSTAL W from the DNA Data Bank of Japan (DDBJ, Mishima, Japan) [19].

Real-Time Quantitative RT-PCR.
Primers for 16SrRNA were also designed as internal controls. Total RNA was isolated from mid-log-phase cell cultures (15 mL). After centrifugation, the cells were suspended in 0.3 mL of diethylpyrocarbonate-treated water. The samples were transferred to FastRNA tubes with blue caps (Qbiogene, Inc., Carlsbad, CA, USA); then 0.9 mL of TRIzol reagent (Invitrogen) was added. Cells were broken using a FastPrep FP120 homogenizer (Qbiogene) at a speed setting of 0.6 for 30 s. After the samples were placed on ice for 2 minutes, 0.2 mL of chloroform was added and the tubes were vortexed and centrifuged again, as described above. RNA was finally precipitated from the aqueous phase with isopropanol, and the resulting pellets were dried and resuspended in 20 L of diethylpyrocarbonate-treated water.

Statistical Analysis.
Intergroup differences of various factors were estimated by a statistical analysis of variance (ANOVA) for factorial models. Fisher's protected leastsignificant difference test was used to compare individual groups. Statistical computations were performed using STAT-VIEW II (Macintosh computer application).

Frequency of Smooth Colony
Morphology. The purified rRecA protein with a molecular weight of approximately 70 kDa is shown in Figure 1(a). Typical rough colonies were observed on MS-agar plates with no addition of rRecA in the test tubes after inoculation of the overnight grown culture of MT8148 was performed (Figure 1(b)). In contrast, the addition of rRecA to the test tubes after inoculation of MT8148 resulted in smooth colonies at a frequency of 0.2% (Figure 1(b)).  compared to that in MT8148 (approximately 9000 bp) (Figure 2(a)). In addition, western blot analysis of cell-associated GTFs of these strains with a smooth appearance revealed only a single band between the position of corresponding GTFB and GTFC (Figure 2(b)). Activity staining demonstrated that the levels of GTF activities of these strains with a smooth appearance were lower than that of MT8148 (Figure 2(c)). Furthermore, the rates of adherence of these strains to glass in the presence of sucrose were significantly lower than that of MT8148 (Figure 2(d)). Figure 3 shows the regions of recombination of gtfB and gtfC in RRA1, SP2, NN2051, NN2143, and NN2147. Sequence analyses revealed that the total length of the gtfB/gtfC region in these strains was 4368 bp. In addition, the nucleotides of gtfB at the 4235 bp position and those of gtfC at 788 bp were found to be deleted. The area of recombination of gtfB/gtfC is shown in Figure 3.

Expression of recA Gene.
Quantitative RT-PCR analyses revealed that the recA expression was significantly higher in RRA1, SP2, NN2051, NN2143, and NN2147 as compared to MT8148 (Figure 4).

Discussion
In our previous study, the recA gene was shown to have a relationship with the activities of GTF and expression of genes encoding GTF [11]. It has been reported that homologous recombination is strictly dependent on the presence of the RecA protein [7]. In the present study, the addition of rRecA protein increased the frequency of spontaneous recombination of gtfB and gtfC. Furthermore, the morphological appearance of those strains was altered to smooth and their GTF activities were reduced. We also confirmed the location of the recombination of gtfB and gtfC by sequence analysis. It is of great interest that the recombination region of RRA1 and those of the clinical strains with smooth colony morphology, such as SP2 and NN2051, were the same. These results suggest that recombination of gtfB and gtfC genes is associated with RecA, and that excess gene expression may cause fusions between gtfB and gtfC. GTFs of S. mutans are cell surface proteins that facilitate adherence and colonization. Although the roles of GTFB and GTFC in S. mutans for virulence related to dental caries are well established, the mechanisms that control expression of these proteins are poorly understood. The gtfB and gtfC genes appear in tandem sequence, whereas the promoter regions are upstream of the GTFB and GTFC open reading frames. In our previous studies, we found several clinical strains that developed smooth colonies with low GTF activities [9]. Morphological changes of mutant colonies observed after culturing on sucrose-containing agar plates suggest alterations in the production of extracellular polysaccharides or cell surface proteins. In addition, we examined the expressions of genes that encode various surface proteins such as glucan-binding proteins of S. mutans and observed changes in gtfB and gtfC expressions in the clinical strains as compared to the reference strain MT8148.
In the present study, we revealed expressions of the recA gene and other stress response genes in clinical strains using real-time RT-PCR. Interestingly, recA gene expression was elevated in all of the clinical strains tested, while the levels of the other stress response genes such as the dnaK [20] and groEL [21] genes were unchanged in those strains (data not shown). Hazlett et al. [10] reported that inactivation of the gbpA gene encoding the glucan-binding protein of S. mutans promoted in vivo recombination of gtfB and gtfC, and that this recombination was independent of the recA gene, while gtfBC recombinants had similar sites of recombination. Although they did not determine the mechanism for accumulation of gbpA mutants with gtfB-gtfC fusion, the cell surface structure seems to be important for maintaining a steady state of intact gtfB and gtfC genes. However, the recA gene is one of the stress response genes involved in quorum sensing in S. pneumoniae [6]. Thus, inactivation of gbpA may influence the expression of genes associated with quorum sensing relative, including recA. Furthermore, gtfB and gtfC fusion was also found in rml mutant strains, in which inactivation of the rml gene caused the fusion [22]. However, the present clinical strains with smooth colonies also demonstrated rml gene expression (data not shown). Taken together, gtfB and gtfC fusion may occur by some alterations of gene expression in S. mutans.
We speculated that fusion between gtfB and gtfC may occur under an acidic condition, since recA gene expression was shown to be elevated at acidic pH levels [11]. It was also reported that excess recA gene expression related to low pH is caused by drastically changed pH in the human oral cavity. Such a condition leads to increased gtfB-gtfC fusions by binding RecA to the junction of the gtfB and gtfC genes. S. mutans utilizes quorum-sensing systems to  modulate environmental stress responses, such as low pH, antibiotics, and oxidative stress [23]. Therefore, we speculated that DNA recombination and phenotypic change through the uptake of extracellular RecA may have a relationship with the signal transduction system. Additional studies are required to clarify this point.