Direct interactions with commensal streptococci modify intercellular communication behaviors of Streptococcus mutans

The formation of dental caries is a complex process that ultimately leads to damage of the tooth enamel from acids produced by microbes in attached biofilms. The bacterial interactions occurring within these biofilms between cariogenic bacteria, such as the mutans streptococci, and health-associated commensal streptococci, are thought to be critical determinants of health and disease. To better understand these interactions, a Streptococcus mutans reporter strain that actively monitors cell–cell communication via peptide signaling was cocultured with different commensal streptococci. Signaling by S. mutans, normally highly active in monoculture, was completely inhibited by several species of commensals, but only when the bacteria were in direct contact with S. mutans. We identified a novel gene expression pattern that occurred in S. mutans when cultured directly with these commensals. Finally, mutant derivatives of commensals lacking previously shown antagonistic gene products displayed wild-type levels of signal inhibition in cocultures. Collectively, these results reveal a novel pathway(s) in multiple health-associated commensal streptococci that blocks peptide signaling and induces a common contact-dependent pattern of differential gene expression in S. mutans. Understanding the molecular basis for this inhibition will assist in the rational design of new risk assessments, diagnostics, and treatments for the most pervasive oral infectious diseases.


SUPPLEMENTAL MATERIAL AND METHODS
Microtiter Plate Assays (Expanded) Figure 2C: 5 µM of synthetic XIP was added to CDM prior to strain inoculation to determine if exogenously added XIP could overcome ComRS signaling inhibition by competitor streptococci. Synthetic XIP (sXIP, aa sequence = GLDWWSL), corresponding to residues 11-17 of ComS, was synthesized and purified to 96% homogeneity by NeoBioSci (Cambridge, MA). The lyophilized sXIP was reconstituted with 99.7% dimethyl sulfoxide (DMSO) to a final concentration of 2 mM and stored in 100 μL aliquots at -20°C. Figure 5A: Selected strains were grown overnight in BHI with appropriate antibiotics (37C 5% CO2). Next morning, cells were harvested by centrifugation, washed and diluted 1:40 into fresh CDM. Cells were incubated for 5 h (till OD600 nm = 0.5). For thermal deactivation, cells (1 mL total volume in a microcentrifuge tube) were placed at 80C for 0.5 h in a heating block. After, treated cells were inoculated directly to begin the experiment. Figure 5B: Two sets of cells were cultured as in Figure 5A: (1) those to be inoculated without treatment for controls, and those to be treated with 4% paraformaldehyde. For paraformaldehyde treatment, cells were harvested by centrifugation and washed with 1X PBS. Cells were then resuspended in 4% paraformaldehyde PBS and kept on the benchtop to incubate for 1 h. After, cells were centrifuged, washed in CDM and resuspended in fresh CDM for inoculation to begin plate reader experiment. Figure 5C: For dilution series, 20 µL of S. sp. A12 was inoculated into 2 mL of fresh CDM after reaching mid-exponential growth phase (original 50% in Figure 5C). After vortexing to mix, 1 mL was removed and added to 1 mL of fresh CDM for a 1:2 serial dilution (25% S. sp. A12). This process was continued seven additional times to complete the S. sp. A12 dilution series. 0% represented no S. sp. A12 addition (same as fresh CDM). After all dilutions were completed, 10 µL of either S. mutans UA159 (fluorescent background subtraction) or S. mutans PcomX::gfp reporter strain was added, mixed and then plated in the 96 well plate to begin the experiment. Figure 5D-5F: These figures represent data from one experiment. Either S. mutans UA159 (fluorescent background subtraction) or S. mutans PcomX::gfp reporter strain were added to 1 mL of CDM and plated into a 96 well plate to begin the experiment. S. mutans was inoculated at different ratios so that tested cultures would enter mid-exponential growth at different time points and thus autoactivation of ComRS signaling would initiate separate from each other. 20 µL of S. mutans reporter strain was added to 1 mL of CDM in panel D, 15 µL in panel E and 10 µL in panel F. Simultaneously, cultures of either S. mutans UA159 or S. gordonii DL1 were inoculated 1:100 and incubated at 37˚C and 5% CO2 atmosphere at the beginning of the experiment. Data from the Synergy plate reader was then closely monitored. At a selected time point when competence activation could be fully detected over background in the higher inoculation culture and beginning to be detected in the middle inoculation cultures (= 4.5 h), the experimental protocol on the plate reader was paused, the 96-well plate carefully removed and 10 µL of either S. mutans UA159 or S. gordonii DL1 was added to their respective wells. The 96-well plate was then returned to the plate reader and protocol continued to run till completion. See supplemental figure 7. Figure 6A: Selected competitor strains were grown overnight in BHI with appropriate antibiotics. Next day, cells were harvested by centrifugation, washed and diluted 1:100 into fresh CDM. Cultures were incubated overnight (37C and 5% CO2). The next day, cultures were harvested by centrifugation and spent supernatant fluid moved into a conical tube. Supernates were treated by adjusting pH from ~6.3 to 7.0 using 6N sodium hydroxide and adding back 20 mM of glucose as a carbohydrate source. Supernatants were filter sterilized with a Millex ® GP 0.22 µm filter unit containing a Millipore Express ® polyethersulfone (PES) membrane prior to reporter inoculation. 10 µL of either S. mutans UA159 (fluorescent background subtraction) or S. mutans PcomX::gfp reporter strain were added to 1 mL of treated overnight supernatant fluid, mixed and then plated in the 96 well plate to begin the experiment. Figure 6B: Selected strains were grown overnight in BHI with appropriate antibiotics. Next morning, cells were harvested by centrifugation, washed and diluted 1:40 into fresh CDM. Cells were incubated for 5 h (till OD600 nm = 0.5). 10 µL of either S. mutans UA159 (fluorescent background subtraction) or S. mutans PcomX::gfp reporter strain were added to 1 mL of fresh CDM, mixed and then 0.1 mL was plated into the dark-sided, clear bottom 96-well plate. An HTS Transwell-96 Well insert (Corning) that contains a 0.4 µm polycarbonate membrane was then placed over the pipetted cultures. 0.1 mL of CDM that contained a 1/100 dilution of the competitor species was then added on top of the Transwell insert. After a mineral oil overlay was added, the 96-well plate containing the Transwell insert was then placed into the Synergy2 plate reader to begin the experiment. Figure 8B: After all inoculated cultures were added to two different 96-well plates, similar to previous experiments but with no mineral oil overlay, the plates were either placed (1) within a 37C and 5% CO2 incubator or (2) within a GasPak EZ Anaerobe Container System with Indicator (Becton, Dickinson and Company), and placed in a regular 37˚C incubator. After 12 h, plates were removed, and data recorded by the Synergy2 plate reader.
Construction of ∆oppA Strains DL1: The oppA deletion strain of S. gordonii DL1 was constructed using a PCR ligation mutagenesis approach as previously described (1). Briefly, 700 bp fragments both upstream and downstream of the gene of interest were amplified with designed primers and contained a BamHI restriction site in the region closest to the gene. After restriction digest, a non-polar kanamycin resistant cassette was ligated within the arms. This ligated fragment was then transformed into S. gordonii DL1 with selection on BHI agar containing 1 mg mL -1 kanamycin. Successful replacement was confirmed both by size selection colony PCR and Sanger sequencing (Eurofins Genomics, Louisville, KY, USA). Restriction and DNA-modifying enzymes were obtained from New England Biolabs. PCRs were carried out with 100 ng of chromosomal DNA by using Taq DNA polymerase, and PCR products were purified with the QIAquick kit (QIAGEN).
A12: The oppA deletion strain of A12 was constructed following methods previously described (2). Briefly, gene disruption was achieved via double-crossover recombination method using linear DNA assembled through Gibson Assembly (New England BioLabs, Beverly, MA). Two DNA fragments flanking the coding sequence of oppA were amplified using indicated primer sets. Selected primer sets also contained at least 25 bases of sequence that overlapped with 5' and 3' ends of a nonpolar kanamycin resistance cassette in pALH124. The two flanking DNA fragments and the kanamycin resistance cassette were mixed in equimolar concentrations in a single isothermal ligation reaction. For transformation of A12, overnight cultures of A12 were inoculated into fresh BHI medium and 0.5 ug of prepared DNA was used to transform A12 with 50 nM of A12 sCSP (2, 3) to induce competence. After 5 h of incubation, cells were plated onto BHI agar plates containing kanamycin (1 mg ml -1 ) and isolated colonies were picked for PCR verification. PCR products from positive colonies were sent for sequence verification. Supplemental Table 6: Plasmids used in this study.

SUPPLEMENTAL FIGURES AND FIGURE LEGENDS
Supplemental Figure 1 Autoactivation of ComRS signaling in CDM medium. Streptococcus mutans strain UA159 harboring plasmid pDL278 that contained a copy of the gfp gene under the control of either the PcomS (orange circles) or PcomX (blue squares) promoter was grown in CDM medium to midexponential log phase (OD600 nm = 0.5). Cultures were then back diluted 1/50 in CDM and monitored by both fluorescence signal and optical density for 18 h in a Synergy2 multimode plate reader (BioTek). The data shows that ComRS signaling spontaneously activates when either reporter strain is grown in CDM medium and is cell density dependent. sanguinis or (C) the A12-likes in CDM medium was monitored for 18 h in a Bioscreen C instrument. Cultures of all isolates were grown to mid-exponential log phase then back diluted 1/100 in CDM medium before being loaded in a Bioscreen C plate. For CFU determination, the 96 well plate was removed from the Synergy2 plate reader and individual replicates were serially diluted and plated on either BHI containing spectinomycin (pDL278 carries spectinomycin resistance) or BHI only (total CFU cell count). CFUs were enumerated to determine percentage of reporter strain remaining in the population after 18 h of monitoring.
Supplemental Figure 4 Screen of Smu Isolates cocultured with the S. mutans PcomX::gfp Reporter. (A & B) Selected Smu isolates were screened for their ability to inhibit ComRS signaling when cocultured with S. mutans UA159 PcomX::gfp reporter. The reporter cocultured with a reporterless UA159 served as a positive control (green dots) and coculturing with S. sp. A12 served as a negative control (blue squares). Coculturing of S. mutans with other S. mutans isolates did not inhibit ComRS signaling activity. Cocultures with two isolates, Smu107 and Smu109, displayed less reporter activity than with the other isolates. This is most likely due to the faster growth rates and higher carrying capacity of these two isolates as seen in (C), recorded growth rates of individual isolates in a Bioscreen C instrument. Additionally, (D) CFU enumeration showed that the UA159 reporter strain consisted of < 1% of all CFUs after 18 h of monitoring, consistent with the observed decreased reporter activity.
Supplemental Figure 5 Bacterial recovery after selected inactivation treatments. 1 mL cultures of S. mutans UA159 or S. sp. A12, grown to mid-exponential log phase, were treated either by heat inactivation (0.5 h at 80˚C in benchtop heat block) or suspension in 4% paraformaldehyde PBS for 1 h. To determine effectiveness of selected bacterial inactivation treatments, 10 µL out of 1 mL treated culture were spotted on square BHI agar plates using the drop-spot method, were the cultures are spotted and then the agar plate tilted up to let the spot run down the agar plate for resolution of individual bacterial colonies for colony forming unit enumeration. A lawn of bacteria was recovered for the heat inactivation treatment while individual colonies could be seen in the 4% paraformaldehyde treatment. C = control, non-treated S. mutans UA159 culture, followed by treated cultures of either S. mutans (used as controls in the fluorescent reporter experiment shown in Figure 5) or S. sp. A12.
Supplemental Figure 6 Recovery of different inoculation ratios of S. mutans PcomX::gfp reporter to S. sp. A12 after 18 h of growth. The S. mutans reporter strain PcomX::gfp and commensal competitor S. sp. A12 were inoculated in CDM growth medium at different inoculation ratios and then fluorescence of the coculture monitored in a Synergy2 plate reader for 18 h. For CFU determination, the 96 well plate was removed from the plate reader and individual replicates were serially diluted and plated on either BHI containing spectinomycin (reporter count, pDL278 carries spectinomycin resistance) or BHI only (total CFU cell count). CFUs were enumerated to determine percentage of reporter strain remaining in the population after 18 h of monitoring. Results showed that for cocultures of S. mutans and S. sp. A12, the population recovered after 18 h of growth remained close to the original inoculation ratio and was used in a competitor concentration experiment ( Figure 5C).

Supplemental Figure 7
Relative Fluorescent Unit Measurements During Competitor Addition Experiments From Figure 5D-F: S. mutans reporter strain PcomX::gfp was added at different dilutions to begin the experiment, either at a high (1:50, represented in Figure 5D), medium (1:66, represented in Figure 5E) or low (1:100, represented in Figure 5F) dilution, and activation of ComRS signaling by the reporter was monitored using a Synergy 2 multimode plate reader. (A) At 4.5 h, different levels of activation were seen, with the high dilution being activated, the medium dilution beginning to be activated and the low dilution not activated yet. At 4.5 h, either S. mutans UA159 (control) or S. gordonii DL1 (competitor) was added to the reporter strain and readings in the Synergy 2 plate reader were resumed. (B) Differences in reporter activities between dilutions when S. gordonii DL1 was added at 12 h into the experiment. The lower dilution, not activated for ComRS signaling at 4.5 h, showed significantly less RFUs compared to both the high and medium dilutions (Student's T-Test, * = P-value of 0.03 for both comparisons).
The experiment was performed with four biological replicates recorded in technical quadruplets.
Supplemental Figure 8 Score Plot of Principal Component Analysis on RNA-Seq Data. Principal component analysis of transcriptome profiles between three different culturing conditions of S. mutans UA159: in its own spent supernatant fluid (UA159sup, green), in S. sp. A12 spent supernatant fluid (A12sup, red), and directly cocultured with S. sp. A12 (Dual, blue).
Supplemental Figure 9 qRT-PCR of S. mutans differentially expressed genes during coculture with select S. sp. A12 mutants. qRT-PCR of selected upregulated (top row) or downregulated (bottom row) genes from coculture with selected S. sp. A12 strains. Cocultures of S. mutans UA159 and selected mutants were grown in CDM medium to OD600 nm = 0.5 before harvest and RNA extraction. Data represents fold change in gene expression compared to S. mutans UA159 cocultured with the wild-type S. sp. A12 strain. Data was normalized to internal SMU.996 and SMU.1616c controls, similar to Figure 7. Three independent cocultures were analyzed and plotted. Red star denotes statistical significance by Student's T-Test within the GraphPad Prism software, P < 0.05.
Supplemental Figure 10 ComRS Signaling Inhibition in Commensal Supernatants. Comparison of relative fluorescent units (RFUs) after 12 h of incubation for the S. mutans reporter strain PcomX::gfp grown within different supernatants. Fresh CDM (green bars) is inoculation of reporter strain into fresh (non-supernatant) CDM to begin the experiment as serves as a positive control. A12 Direct (red bars) is inoculation of S. sp. A12 with the reporter strain into fresh CDM and serves as the negative control. In this experiment, two different supernatants were tested. Supernatant from S. sp. A12 monoculture grown overnight (blue bars, similar to Figure 3A) or supernatant from a S. sp. A12 and S. mutans UA159 coculture grown overnight (orange bars). Overnight cultures of selected strains where centrifuged, spent supernates removed, filter sterilized, the pH was adjusted to 7.0 and 20 mM additional glucose was added. The PcomX::gfp reporter strain was then inoculated and monitored for 18 h in a Synergy 2 multimode plate reader. The experiment was performed with four biological replicates recorded in technical quadruplets.

Movie S1. S. mutans PcomX::gfp Active Cell Within a Cocultured Biofilm. Movie of the 3D
rendering of a 40 µm x 40 µm area within a SYTO 42-stained (total cells, blue) S. mutans (PcomX::gfp, green) and S. gordonii (P23::dsRed2, red) cocultured biofilm. Two PcomX-active cells are located at/near the substratum of the biofilm, within a S. mutans microcolony. The image in the movie is the same as that displayed in the top panel of Figure 2B.