Background

In addition to being a highly prevalent infectious bacterial disease, dental caries can have a serious adverse impact on patient quality of life and can lead to potentially severe complications including pain or tooth loss [1,2,3,4,5,6]. Streptococcus mutans (S. mutans) is a major cariogenic bacterial species that facilitates local acidification at tooth surfaces by metabolizing carbohydrates present in the oral microenvironment [7]. When the local pH falls below 5.5, this can lead to the demineralization of the tooth surface and the consequent development of dental caries [8].

Streptococcus gordonii (S. gordonii) is a Gram-positive bacterium that is thought to be a major initial colonizing species involved in the formation of dental plaque biofilms [9, 10]. S. gordonii metabolic activity can lead to the production of H2O2 [11], which inhibits S. mutans growth, as well as the production of ammonia which can counteract local tooth surface acidification to protect against the onset or progression of cariogenesis [12]. Previously reported research studies have highlighted a positive correlation between S. mutans detection rates in dental plaque and dental caries incidence [13, 14], whereas S. gordonii is negatively correlated with dental caries [15, 16].

Bacteriocins are a family of ribosomally synthesized peptide antibiotics that are produced by bacteria that resist adverse environments [17]. The bacteriocin produced by S. mutans is called mutacin [17, 18]. It has been reported that S. mutans produces several types of mutacins including mutacins I-IV, K8 and Smb. Mutacin IV positive S. mutans strains has been shown to readily suppress S. gordonii growth [19, 20]. S. mutans secretes mutacin IV, which is a bacteriocin encoded by nlmA and nlmB and regulated by the ComDE two-component system, enabling it to effectively inhibit S. gordonii microbial activity [21, 22]. In in vivo plaque biofilms, Tanzer et al. found that S. mutans was capable of outcompeting S. gordonii and was highly cariogenic [23]. CSP-21 is a 21-residue competence-stimulating peptide that can interact with the membrane-bound histidine kinase receptor ComD, leading to its autophosphorylation and the transfer of the phosphate group to its cognate cytoplasmic response regulator ComE, which is then capable of binding to the promoter regions upstream of genes encoding mutacin IV (nlmA and nlmB), promoting its upregulation [24]. SepM cleaves CSP-21 into CSP-18, which is more effective in activating the ComDE system than CSP-21 [25]. It is known that mutacin IV inhibits S. gordonii growth, and many studies on mutacin IV have used S. gordonii as an indicator bacterium [17, 21]. Therefore, SepM is a key factor in regulating the inhibitory effect of S. mutans on S. gordonii.

In S. mutans, the SepM protein is 346 amino acids in length and includes a minimum of one transmembrane domain (amino acids 10–26), a eukaryotic-type PDZ domain (amino acids 131–195), and a C-terminal Lon-like protease (S16) domain (amino acids 233–314) [26]. PDZ domains are most commonly present in multicellular organisms wherein they function as common modules that facilitate interactions between proteins through the recognition of short peptide sequences most often present in the C-terminal regions of proteins associated with the plasma membrane. These domains contribute to protein complex formation and spatial confinement, thereby exhibiting ligand specificity and regulating cellular signaling processes [27, 28]. The C-terminal domains of serine proteases also contain the residues that form the active site for substrate binding and the regulation of catalytic activity. Given that genetic mutations can impact virulence and thereby enhance or lower disease-related risk, the present study was developed to analyze sepM gene mutations of the PDZ and C-terminal domains across 286 serotype C S. mutans clinical isolates and explore the role that these mutations play in governing interactions between S. mutans and S. gordonii.

Methods

Biological phenotype

This study received ethical approval from the ethics committee of the First Affiliated Hospital of Bengbu Medical College (Approval No [2017] KY011) for the S. mutans clinical isolates collection. Informed consent was obtained from all subjects and/or their legal guardian(s). These 286 isolates were collected from samples of dental plaque obtained from 3-6-year-old children with or without dental caries that are preserved in our laboratory. The isolation and serotype c identification methods for these S. mutans clinical isolates, as well as the activity of these isolates against S. gordonii, were detailed in our prior study [29, 30]. All sequences of the 286 S. mutans clinical isolates were submitted to the National Center for Biotechnology Information (NCBI) under the BioProject accession number PRJNA1016128. Briefly, plaque samples were centrifuged for 30 s, plated 50 µL of saline on Trypticase Yeast-Extract Cysteine Sucrose Bacitracin (TYCSB) agar, and then cultured for 48 h at 37 °C under 5% CO2. Clinical S. mutans were confirmed according to their reaction to mannitol, sorbitol, raffinose, melibiose, aesculin, arginine hydrolase and arginine hydrolase control [31]. Then all isolates were cultured overnight in brain heart infusion (BHI) broth and then DNA was extracted from these isolates using a Bacterial Genome DNA Extraction Kit (Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. The confirmation of the C serotype was carried out by polymerase chain reaction (PCR) using serotype-specific primers. The PCR conditions for C serotype were as follows: denaturation at 96 °C for 2 min; 25 cycles consisting of 15 s of denaturation at 96 °C; 30 s of annealing at 61 °C; and 1 min of extension at 72 °C. The sequences of PCR primers for the identification of serotypes are listed in Table 1 [32]. The amplification products were then observed through electrophoresis in agarose gels. S. mutans UA159 was used as a positive reference. Bacteriocin assay was used to evaluate the activity of S. mutans against S. gordonii. Each isolate was incubated overnight in BHI broth, and then 10 µl of each clinical isolate with an OD600 of 0.3 were added to BHI agar. Equal amounts of S. gordonii (ATCC 10,558) were inoculated adjacent to the S. mutans sample after 12 h of incubation. The agar plate medium was then incubated for another 12 h. S gordonii clearing zones represent S. mutans’s activity against S. gordonii. The bacteriocin assay was conducted in three biological replicates.

Table 1 Primers used for the present study
Full size table
Table 2 The relationship between sepM mutations and the virulence of S. mutans against S. gordonii
Full size table

sepM polymorphism analyses

S. mutans DNA was extracted from the clinical isolates at stationary phase as in our prior study [33], after which it was used as a template for PCR reactions. PCR primers (sepM-F_mutation and sepM-R_mutation) were designed with Primer Premier 5.0 to amplify the DNA fragment corresponding to the sepM PDZ domain and C-terminal domain. Thermocycler settings were as follows: 5 min at 94 °C; 35 cycles of 94 °C for 30 s, 55 °C for 40 s and 72 °C for 30 s; and 72 °C for 5 min. The resultant DNA was analyzed by 1% agarose gel electrophoresis, and was subsequently sequenced by HuaXiao gene technology (Bengbu, China) with an ABI3730 instrument.

RNA extraction and qRT-PCR

Total RNA was extracted using 25 mg/mL lysozyme (TIANGEN Biotechnology, China) for 30 min and the RNeasy Mini Kit (Qiagen), and was then reverse transcribed into cDNA using the PrimeScript™ RT reagent Kit (Takara). The qRT- PCR of analyses of sepM, comD, comE, nlmA, and nlmB gene expression was performed using a 20 µL volume containing 10 µL of 2× SYBR Premix Ex TaqII, 1 µL each of the forward and reverse primers, 1 µL of cDNA, and 7 µL of RNase-free H2O. The amplification program settings were as follows: 95 °C for 30 s; 40 cycles of 95 °C for 5 s and 55 °C for 25 s. At the end of the reaction, a melting curve was generated. The sequences of PCR primers are listed in Table 1.

Protein extraction and Western blotting

Bacteria at the stationary phase were collected by centrifugation (15 min, 8,000 rpm) and washed using PBS. Lysozyme (TIANGEN Biotechnology, China) was diluted with lysozyme buffer (SL20734, COOLABER SCIENCE & TECHNOLOGY, China). This buffer was composed of 20 mM Tris (pH 8.0), 2 mM sodium Na2-EDTA and 1.2% Triton X-100. To lyse these cells, 100 µL of 25 mg/mL lysozyme was added per tube followed by incubation for 30 min at 37 °C, followed by the addition of 300 µL of RIPA buffer (P0013B, Beyotime Biotechnology, China), and incubation for 10 min on ice. The main composition of RIPA buffer was 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, sodium orthovanadate, sodium fluoride, EDTA and leupeptin. Samples were then centrifuged, and the supernatant was collected. A BCA Protein Assay Kit (Beyotime, China) was then used to measure protein concentrations in each sample and to dilute samples to 1 mg/mL [34]. Proteins were then combined with loading buffer, heated to 100 ℃ and boiled for 5 min, and subsequently stored at -20 ℃. Following SDS-PAGE separation these proteins were transferred to PVDF membranes (Sigma, MO, USA) that were blocked for 1 h using 5% skim milk in TBS containing 0.1% Tween-20 (TBST) at room temperature. After three washes with TBST, blots were probed overnight with antibodies specific for SepM, ComD, or ComE (diluted 1:500) at 4 ℃, rinsed with TBST, and incubated for 1 h with HRP-conjugated goat anti-Mouse IgG (diluted 1:2,000; S0002, Affinity Biosciences, Jiangsu, China) for SepM or HRP-conjugated Affinipure Goat Anti-Rabbit IgG (H + L) (diluted 1:2,000; SA00001-2, Proteintech, Wuhan, China) for ComD and ComE at room temperature. An ECL Western Blotting Substrate Kit (Cat: #KF005, Affinity Biosciences, Jiangsu, China) was used to detect protein bands. The antibodies specific for SepM, ComD, or ComE were made by HUABIO (Hangzhou, China). The brief steps are as follows: (1) Recombinant SepM, ComD, and ComE was expressed and purified; (2) Antibodies were prepared by immunizing New Zealand white rabbits three times at two-week intervals with 100 µg eluted SepM, ComD or ComE protein that was homogenized in complete Freund’s adjuvant; (3) A boost injection was given one week later; (4) Rabbit serum was collected seven days after the final immunization and stored at − 80 °C. The extraction steps of phosphorylated proteins were similar to those of conventional protein extraction, except that after adding RIPA buffer, 10 µL of Protease Inhibitor Cocktail (P1025, Beyotime Biotechnology, China) needed to be added for co incubation. For phosphorylated proteins expression analysis, a 10% phosphorylated gel was prepared using Beyotime SDS-PAGE Gel Preparation Reagent Kit (beyotime, Shanghai, China, CatNo.P0012A) and Phosbind Acrylamide (APExBIO Houston, USA, CatNo. F4002). 10 µL of the target protein sample and 2.5 µL of Prestained Protein Marker (APExBIO Houston, USA, Cat No. F4005) are loaded onto the gels. ComD antibody and Proteintech Affinipure Goat Anti-Rabbit IgG (H + L) were diluted in a 5% BSA (Solarbio, Beijing, China, Cat No. A8020) at a ratio of 1:250. The other steps were the same as routine Western blotting.

Molecular docking and molecular dynamics simulations

SepM protein crystal structures were not present in the RCSB Protein Data Bank (http://www.pdb.org/). The SepM protein structure was therefore predicted using Iterative Threading ASSEmbly Refinement (I-TASSER) (http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S689586/) [35, 36] followed by modification with the Schrodinger 2019.01 Protein Preparation Wizard model, including water molecule removal, hydrogen addition, amino acid optimization, and patching. Protein-substrate complex stability and dynamics were assessed through a slightly modified version of a previously reported molecular dynamics simulation approach [37]. Briefly, the simulation box was solvated with TIP3P water molecules together with appropriate counter ions to neutralize the system. System energy was minimized using optimized parameters for liquid simulation (OPLS3e) forcefield. In total, 100 ns of equilibration simulations were run in the NpT ensemble using the following parameters: temperature, 300 K; pressure,1.0 bar; integration time step, 1.2 fs. All bonds involving hydrogen atoms were constrained with the SHAKE method. Docking analyses on SepM and CSP-21 were performed with HDOCK (http://hdock.phys.hust.edu.cn/) [38], and the top-ranked docking model was visualized with PyMOL 2.1.

SepMs expression and purification

Recombinant SepM_control, SepM_G178D, and SepM_D221N were purchased from Yangene Biological Technology (Wuhan, China). The brief steps are as follows: (1) Synthesized SepMs (SepM_control, SepM_G178D, and SepM_D221N) sequences were ligated into the pET-sumo vector, respectively. (2) The plasmid was validated by sequencing and subsequently used to transform E. coli BL21 (DE3) (Novagen, USA). (3) Following short-term culture and isopropyl-β-D-thiogalactopyranoside (IPTG) induction (0.8 mM) at 16 ℃ for 15 h, cells were collected by centrifugation (10 min, 4,000 rpm). (4) Pellets were resuspended in lysis buffer and disrupted by sonication on ice (30 min, 300 W work 5s-off 3s). (5) Lysates were then centrifuged at 18,000 rpm under 4 °C for 15 min. (6) The supernatant was then collected and incubated with Ni resin at 4 °C for 2 h. (7) The column was washed with a gradient of concentrations of imidazole solution. (8) 10 mL of elution buffer was added to the column to elute the bound proteins. (9) Recombinant proteins were then analyzed by 15% SDS-PAGE.

Microscale thermophoresis (MST) assay

SepM_control, SepM_G178D, and SepM_D221N binding to CSP-21 were monitored through an MST approach [39]. CSP-21 was synthesized by GenScript (Nanjing, China). These MST assays were conducted using a Monolith NT.115 instrument (NanoTemper Technologies GmbH, Munich, Germany). SepM_control, SepM_G178D, and SepM_D221N (final concentration, 50 nM) possessed a His-tag were mixed with His-Tag Labeling Kit-RED-tris-NTA (MO-L018, NanoTemper Technologies) at 25 °C for 30 min in the dark, and then individually mixed with CSP-21 (initial concentration, 100 µM) in a 16-point serial dilution series. The interactions between the proteins and CSP-21 were measured in Monolith NT.115 Standard Treated Capillaries (NanoTemper Technologies). Samples were prepared in PBS at a pH of 5.5 or 7.5. Measurements were made at 37 °C and 25 °C using 50% light-emitting diode (LED) power and medium MST power. Experiments were repeated in triplicate for all measurements. The KD values were analyzed using MO Affinity Analysis 2.3 software.

Statistical analysis

SPSS 22.0 (IBM, NY, USA) was used to examine the relationship between the inhibition of S. mutans on S. gordonii and sepM genetic polymorphisms, and the expression levels of genes between groups of mutation and mutation_free. Qualitative data were analyzed using the Pearson chi-square test for a theoretical frequency ≥ 5. For theoretical frequencies between 1 and 5, data were analyzed via continuity correction. All other results were analyzed with Fisher’s exact test. The Shapiro-Wilk test was used to assess whether the quantitative data were parametric. For parametric testing, a t test was used, for nonparametric testing, the Mann-Whitney U test was used. P < 0.05 was the threshold for statistical significance.

Results

We aimed to explore whether there is a sepM gene mutation in S. mutans clinical strains that regulates the ability of S. mutans to inhibit the growth of S. gordonii. The S. mutans clinical strains that can inhibit the growth of S. gordonii were defined as an inhibitory group, otherwise they are considered as the non-inhibitory group. When the sepM gene sequences from these 286 clinical isolates were compared to the reference S. mutans UA159 sequence, 16 single nucleotide polymorphisms were identified, including 3 silent mutations and 13 missense mutations. None of the sequenced sepM genes exhibited base insertions or deletions. Significant differences in mutation frequencies were observed at loci 482, 533, and 661 when comparing the inhibitory and non-inhibitory groups (P = 0.039, < 0.001, and 0.004, respectively), with all three of these sites being more frequently mutated in the antagonistic group (Table 2). Fig. S1 showed the results of S. mutans inhibiting the growth of S. gordonii and the serotype validation of S. mutans.

To examine the role that the sepM G533A mutation in the ability of S. mutans to inhibit S. gordonii, the levels of sepM and downstream genes were analyzed in 30 selected clinical isolates including 15 harboring the G533A mutation (G533A group) and 15 control isolates lacking this mutation (G533A_free group). Isolates in the G533A group were selected from the inhibitory group which can inhibit the growth of S. gordonii, while the bacteria in the G533A_free group were selected from non-inhibitory group. Overall, 98% of the analyzed isolates (281/286) shared the G419A mutation, so the G419A point mutation was present in all 30 selected isolates, but with no interference of other sepM silent or missense mutation sites in these isolates. We found that nlmA was expressed at significantly (P < 0.001) higher levels in the G533A group relative to the control group, while no differences between these groups were observed with respect to sepM, comD, comE, or nlmB expression levels (P = 0.330, 0.141, 0.237, and 0.395, respectively) (Fig. 1A). We subsequently assessed the expression levels of SepM, ComD, ComE, phosphorylated ComD, and phosphorylated ComE in 16 clinical isolates with G533A mutation (G533A group, n = 8) and lacking G533A mutation (control group, n = 8). These isolates were selected randomly from the isolates in mRNA analysis. We found SepM and ComE protein levels were higher in the G533A group relative to the control group; although ComD protein levels did not differ substantially between these groups but phosphorylated ComD protein levels were higher in the G533A group relative to the control group (Fig. 1B). Full-length gels of Fig. 1B, Fig S1 A and Fig S2 were included in Supplementary material. Phosphorylated ComE failed detection after multiple experiments. Fig S3 showed the reaction of clinical strains with mannitol, sorbitol, raffinose, melibiose, aesculin, arginine hydrolase and arginine hydrolase control.

Fig. 1
figure 1

Gene and protein expression levels in S. mutans clinical isolates. (A) Relative sepM, comD, and comE, nlmA and nlmB gene expression levels were compared between groups that did and did not harbor the G533A mutation. (B) The expression levels of SepM, ComD, ComE, and phosphorylated ComD compared across 16 isolates. To better display the differences between groups, we displayed samples from the same group on a piece of glue, we also presented a pair of inter group samples on a piece of glue to reduce the impact of exposure factors on the comparison of inter group samples. ComD-P represents phosphorylated ComD. The red arrow represents phosphorylated ComD, and the blue arrow represents unphosphorylated ComD

Full size image

Mutation of G533A at the gene level corresponds to G178D, the latter is a change at amino acid level. Mutation of G661A at the gene level corresponds to D221N at the amino acid level. Potential binding sites and stability were also compared among SepMs (SepM_control, SepM_G178D (G533A), SepM_D221N (G661A)) and CSP-21 through molecular docking and molecular dynamics simulation approaches. A root-mean-square displacement (RMSD) cutoff of 2Å is frequently used as a criterion when predicting correct bond structures. The RMSD value of < 0.6 in this study indicated that the SepM-CSP-21 complex was stable (Fig. 2A). Root-mean-square fluctuation (RMSF) values further demonstrated that the structure of this SepM_G533A-CSP-21 complex exhibited more fluctuations for 175–200 residues relative to other binding sites, the structure of the SepM_G661A-CSP-21 complex had more fluctuations for 220–225 residues relative to other binding sites. This indicates that mutations can affect the binding mode through which SepM and CSP-21 interact. All binding energy values were below − 200, indicating a stable binding interaction between all three analyzed SepM isoforms and CSP-21. Identified binding sites between SepM_control and CSP-21 included ALA-76, TYR-116, LYS-280, and LYS-301, while binding sites between SepM_D221N and CSP-21 included TYR-116, PHE-162, LYS-160, ASP-278, and LYS-280, and binding sites between SepM_G178D and CSP-21 included LYS-50, GLU-51, LYS-55, ILE-231, ASP-278, and LYS-301 (Fig. 2B). Of these complexes, SepM_G178D exhibited three residues (ILE-231, ASP-278, and LYS-301) that were very close to the active center (S235 and K280), while SepM_control and SepM_D221N each harbored two such residues (LYS-280, and LYS-301 for SepM_control; ASP-278 and LYS-280 for SepM_G661A) close to the active center.

Fig. 2
figure 2

The binding mode for interactions between SepMs and CSP-21. (A) A root-mean-square displacement (RMSD) plot, a root-mean-square fluctuation (RMSF) plot for SepM and a RMSF plot for CSP-21. RMSF diagrams demonstrated that the structure of the SepM_G533A-CSP-21 complex exhibited more fluctuations at 175–200 residues relative to other binding sites, the structure of the SepM_G661A-CSP-21 complex had more fluctuations at 220–225 residues relative to other binding sites. This suggests mutations in SepM exhibit effective fluctuations conducive to its binding with CSP-21; (B) The binding mode for interactions between SepM_control and CSP-21

Full size image

The expression and purification of SepMs (SepM_control, SepM_G178D, SepM_D221N) were illustrated in Fig. S2. Interactions between the three analyzed SepMs (SepM_control, SepM_G178D, SepM_D221N) and CSP-21 were assessed through an MST assay approach (Fig. 3). This analysis revealed that: (1) at 37 ℃ and pH 7.5, the respective affinities of these three proteins for CSP-21 were 29.3 ± 34.5 µM, 85.0 ± 240 µM, and 28.7 ± 24.1 µM; (2) at 37 ℃ and pH 5.5, the respective affinities of these three proteins for CSP-21 were 12.8 ± 10.3 µM, 21.4 ± 26.8 µM, and 7.52 ± 4.62 µM; (3) at 25 ℃ and pH 7.5, the respective affinities of these three proteins for CSP-21 were 15.9 ± 16.5 µM, 3.02 ± 2.27 µM, and 20.8 ± 47.1 µM; and (4) at 25 ℃ and pH 5.5, the respective affinities of these three proteins for CSP-21 were 33.1 ± 25.5 µM, 44.2 ± 86.8 µM, and 8.25 ± 8.65 µM.

Fig. 3
figure 3

The affinity (KD) between SepMs (SepM_control, SepM_G178D, and SepM_D221N) and CSP-21

Full size image

Discussion

Genetic variations in bacteria can alter their virulence phenotypes in a manner that may lower or raise the risk of disease. S. mutans is an important cariogenic bacterium. While S. gordonii is a dominant bacterium for early colonization of dental plaque, and is negatively correlated with the onset of dental caries. SepM plays an important role in the interaction between S. mutans and S. gordonii. In this study, we focused on SepM mutation in S. mutans clinical isolates and related function. Of these 286 S. mutans isolates, 149 and 137 isolates originated from the caries site and caries_free site, respectively. No correlation was found between the source of the isolates (caries or caries_free) and the inhibition of S. mutans against S. gordonii (P = 0.697). This may be because the inhibition of S. gordonii growth by S. mutans is not a direct factor in its cariogenic effect. The direct cariogenicity of S. mutans is to lower the local pH and induce demineralization [40]. Considering that the ability of S. mutans to inhibit S. gordonii is not a necessary factor for the occurrence of dental caries, we did not include the analysis of the source (caries or caries_free) of the strains in mutation analysis.

The G533A (G178D) mutation distribution differed most significantly when comparing the inhibitory and non-inhibitory groups, and the levels of SepM and associated regulatory protein and genes were thus analyzed in clinical isolates that did or did not harbor this G533A mutation. Gene expression level results suggested that none of the analyzed genes exhibited any differential expression patterns. As information is limited regarding appropriate internal reference proteins for use when studying S. mutans, protein levels were instead measured based on general protein quantification [34, 41]. Protein analyses revealed that the expression levels of the SepM and ComE proteins in the G533A group were elevated relative to the control group, although ComD protein levels did not vary between these groups, whereas the expression levels of the phosphorylated ComD in the G533A group were elevated relative to the control group. These indicate the possibility that the sepM G533A mutation may increase SepM protein expression without having any effect on sepM gene expression. The increased SepM will bind to more ComD proteins, inducing the latter to produce more phosphorylation levels, higher expression of ComE (phosphorylated ComE), and a stronger phenotype that inhibits the growth of S. gordonii. Unfortunately, we were unable to obtain clear phosphorylated ComE bands under various conditions. The expression levels of genes and their corresponding proteins in the same strains are not completely consistent. This does not rule out the possibility of another mechanism promote the translation of these genes or the discrepancy may be due to different protein stability.

Saswati Biswas et al. reported that when the mutation is located at the 18th and 19th amino acids of CSP-21, it can affect the activity of CSP-21 binding to SepM, while mutations at other amino acid sites do not significantly affect its binding to SepM [26]. The comC gene has a total length of 141 bp and encodes 46 amino acids. CSP-21 is located at the 76–138 bp of the comC gene in S. mutans. We also analyzed gene mutations in the CSP-21 fragment in these 286 S. mutans clinical isolates and found that the CSP-21 peptide segment only shared 6 mutations including 5 missense mutations and 1 synonymous mutation. However, there was no significant difference in the distribution of these missense mutations between the two groups. Considering that the site of these mutations is located at the first four amino acids of the CSP-21, which is far from the 18th and 19th amino acids of CSP-21, we think there is no difference in CSP-21 production by the various mutant strains in this study.

Molecular docking analyses were used to predict ligand interactions with the target binding site and to gauge the stability of this interaction [42]. However, protein flexibility was lacking in most instances [43], and the reliability of the resultant protein-peptide complexes is unclear. Molecular dynamics simulations can complement these docking analyses [44]. Here, both strategies were employed in parallel, revealing that the G533A mutation did not impact the stability of SepM binding to CSP-21, whereas it did impact the binding site of SepM where it interacts with this ligand. Specifically, the CSP-21 binding sites in the SepM_G533A protein included the LYS-50, ASP-278, LYS-301, and ILE-231 residues. SepM is a serine protease that localizes to the cell surface and harbors a Ser-Lys dyad (S235 and K280) active site [25]. The ASP-278 position is close to this active site K280 residue, and it may thus influence SepM activity.

In caries-free individuals, the oral pH is typically neutral [45, 46], while a pH of 5.5 is the critical threshold value for the demineralization of the enamel and caries development [47]. Accordingly, we assessed SepMs activity under these two pH levels. The equilibrium dissociation constant KD is used to describe the binding affinity between a given protein and its target ligand [48], with greater KD values indicating weaker affinity. Generally, the oral temperature is 37 ℃, and our results showed that there was no significant difference between the mutant protein and the control group under 37 ℃. But at 25 ℃ and a pH of 5.5, SepM_D221N exhibited high affinity for its target substrate, whereas the control SepM protein exhibited lower affinity. Similarly, at 25 ℃ and a pH of 7.5, the affinity of SepM_G178D for its cognate substrate was relatively high as compared to that of the control SepM. This suggests that S. mutans carrying mutant SepM proteins may play an important role in assisting directly cariogenic S. mutans strains in antagonizing other threatening species such as S. gordonii to gain competitive growth advantages on the tooth surface.

While this study offers new insight into the role that genetic modification in bacterial interactions, it is subject to limitations. In this study, we also tried to construct mutants in UA159, but the resultant strains were unstable and had poor vitality. We were puzzled whether the phenotypic changes were caused by mutations or because of the repeated freeze-thaw recovery and clonal screening selection. Relevant techniques need to be improved in the future to observe the specific effect of mutation on phenotype. Therefore, this study used protein prokaryotic expression and affinity experiments to compensate for the impact of gene mutations on the SepM protein. More experiments are needed to confirm the impact of mutations on the stability of the SepM protein.

Conclusions

In summary, this study is the first to have systematically analyzed mutations present in sepM across S. mutans clinical isolates, enabling an assessment of the interplay between these sepM mutations and the inhibitory activity of S. mutans against S. gordonii. This research facilitated the verification of the effects of differentially distributed mutations on gene expression, protein expression, and SepM binding affinity for its substrates. Notably, these findings demonstrated that the sepM G533A (G178D) and G661A (D221N) mutations may impact protein translation levels or the stability of the protein, and the binding affinity of this protein for its substrate CSP-21, underscoring their relevance when exploring interactions between S. mutans and S. gordonii and the impact of these interactions on caries development from a genetic perspective.