Streptococcus mutans UA159 was obtained from the Institute of Microbiology, Chinese Academy of Sciences and grown in brain–heart infusion broth (BHI) at 37°C under 5% CO₂ (v/v). Solid media were prepared by adding 1.5% (w/v) agar. For biofilm formation, S. mutans cells were grown in BHI supplemented with 1% sucrose (w/v) in 96-well plates under static conditions.

Sequences encoding full-length DexA (Signal peptide is not included) and DexA70 (residues 100–732) were amplified and cloned into pGL01, a vector modified from pET15b by introducing a PreScission Protease (PPase) cleavage site for the removal of His-tag. The resulting expression plasmids were transformed into Escherichia coli BL21 (DE3), which was cultured at 37°C until OD600 reached 0.8 and then induced overnight with 0.3 mM isopropyl β-D-thiogalactopyranoside at 16°C.

For protein purification, bacterial cells were harvested by centrifugation for 15 min at 5000 × g. The bacterial cells were resuspended in lysis buffer (25 mM Tris-HCl, pH 8.0, 200 mM NaCl) and lysed by sonication. The lysate was centrifuged at 25,200 × g for 45 min to remove the cell debris. The supernatant was loaded onto a Ni-NTA column (GE Healthcare) for affinity chromatography. Miscellaneous proteins were removed by washing several times with the lysis buffer. The resin binding the target protein was mixed with PPase (with a final concentration of 0.1 mg/ml) overnight at 4°C to remove the His-tag. The protein sample was then eluted with lysis buffer. For further purification, protein was loaded onto an ion-exchange column (Source 15Q HR 16/10, GE Healthcare) and eluted with a linear gradient of 0–1 M NaCl in 25 mM Tris-HCl buffer pH 8.0. Finally, the protein sample was purified by size-exclusion chromatography (Superdex 200 10/300 GL, GE Healthcare) in 10 mM Tris-HCl buffer, pH 8.0, containing 100 mM NaCl. Protein purity was examined by SDS-PAGE.

To determine the effects of chlorhexidine (CHX) and cetylpyridinium chloride (CPC) on planktonic bacteria, overnight cultures of S. mutans were diluted by 1:100, then inoculated into fresh BHI media containing different concentration of CHX or CPC, and grown at 37°C, 5% CO₂. The cultures were sampled at 1-h interval, and the absorbance at the wavelength of 600 nm was tested.

The effects of antimicrobial agents on biofilms were tested by calculating remaining viable cells after treatment with CHX and CPC. Twelve-hour biofilm was washed by saline solution and then different concentrations of CHX or CPC were added and incubated for 30 min. After reaction, biofilm was scraped off the tubes and resuspended in saline solution. Serially diluted suspensions were plated onto BHI agar plates and anaerobically incubated for 48 h at 37°C. The amount of viable bacterial cells from different conditions was expressed as CFU/ml of biofilm. All experiments were performed in triplicate.

Dextranase activity was measured by monitoring the release of reducing sugar during incubation of the mixture containing 200 nM dextranase, 1% dextran 40,000, and 50 mM sodium phosphate buffer (pH 5.0) at 37°C. The reaction was sampled at 2-min interval for 30 min. The sampled reaction was terminated by adding 2 ml of 3,5-dinitrosalicylic acid (DNS). Controls were made by preparing a solution containing the same ingredients except that the enzyme was boiled in advance. Mixtures of both the samples and controls were boiled for 5 min and then deionized water was added to a volume of 10 ml. Absorbance was measured at a wavelength of 540 nm. The gradients in the linear region were calculated to measure enzyme activity.

To determine the optimum temperature for the recombinant dextranase, enzymatic activity was measured at 0, 10, 20, 30, 35, 40, 45, 50, 55, 60, 70, and 80°C, respectively. To study the influence of pH on dextranase activity, enzymatic activity was measured in different buffers at different pH: 50 mM acetic acid sodium acetate buffer (pH 3.5–5.0), 50 mM phosphate buffer (pH 5.0–7.0), and 50 mM Tris-HCl buffer (pH 7.0–9.0).

The cultures were incubated statically in 96-well plates for 12 h at 37°C to allow biofilm formation. For the biofilm inhibition assay, enzymes were supplied into the growth medium at the beginning of inoculation. After incubation, non-adherent cells and matrix were removed by gently washing the plate with deionized water. The wells were stained with 100 μl of 0.1% (w/v) crystal violet for 10 min and then deionized water was added to rinse the wells three times to remove the unbound dye. The remaining dye was solubilized by addition of 200 μl of 30% (v/v) acetic acid and shaking for 15 min. The absorbance was measured at the wavelength of 595 nm. The amount of biofilm was proportional to the absorbance of the staining crystal violet.

For biofilm disassembly assays, 12-h-old biofilm was washed with distilled water to remove the non-adherent cells and medium. The wells were filled with 200 μl of 50 mM phosphate buffer (pH 5.0) with varying concentrations of DexA70 or lysozyme. Reactions were allowed to proceed for up to 30 min at 37°C, and then quenched by washing the plates with distilled water. The wells were stained with 100 μl of 0.1% (w/v) crystal violet for 10 min and then washed with distilled water. The crystal violet bound to biofilms was solubilized with 200 μl 30% acetic acid for quantification. Experiments were performed in triplicate separately.

To obtain the biomass from the biofilms, S. mutans UA159 biofilm was grown in sterile glass tubes in BHI broth at 37°C under 5% CO₂ (v/v) for 12 h. Biofilm was then washed twice with saline solution to remove unattached cells and then treated with or without DexA70 and lysozyme in 50 mM PBS buffer (pH 5.0) at 37°C for 30 min. The treated biofilm was washed twice with saline solution to remove dispersed cells. The remaining biofilm was scraped off the tubes and resuspended in 1 ml of saline solution. The suspension was transferred to pre-weighted tubes and incubated with 100% ethanol at −20°C for 10 min and then centrifuged at 25,200 × g for 30 min at 4°C. The procedure was repeated once, and the precipitate was washed with 75% ethanol and dried for 24 h in a desiccator. Biomass value was calculated by subtracting the initial weight of the empty tube from the final dry weight and expressed as milligrams per milliliter of biofilm suspension.

For confocal laser scanning microscopy observation, biofilm was grown in a petri dish with glass bottom and treated with enzymes. After treatment, the buffer was gently removed and the biofilm was stained with polysaccharide stain concanavalin A (ConA) lectin conjugated with tetramethylrhodamine for visualization of dextran and DNA stain SYTO9 for all bacteria and propidium iodide (PI) for dead cells (Molecular Probes, Invitrogen). Z-stack images were acquired on a Confocal Laser Scanning Microscope (CLSM) (LSM900, Zeiss, Germany) and further analyzed by imaris software package for the creation of three-dimensional visualizations. For systematic analysis of biofilm architecture and structure, a computer program COMSTAT with a wide variety of functions for handling, analyzing, and displaying images was used to quantify the elements of biofilm including bio-volume, thickness distribution, mean thickness, and surface-to-volume ratio (Heydorn et al., 2000).

Minimum biofilm inhibition concentration (MBIC) and minimum biofilm eradication concentration (MBEC) of biofilm were determined based on Clinical and Laboratory Standard Institute (CLSI) (2018) guidelines. Measurement was performed using a modified version of the Calgary biofilm device method (Ceri et al., 1999; Reiter et al., 2013). Biofilm was formed through overnight incubation at 37°C. Non-adherent cells were removed by gentle washing three times with sterile saline solution. Biofilm was exposed to media or media containing 500 nM DexA70 for 30 min before challenge of serial antibiotics and then washed again with sterile saline solution. The treated biofilm was left to air dry for 15 min. Serial twofold dilutions of each antimicrobial agent in BHI broth were then added and incubated at 37°C for 24 h. MBIC represents the minimal antimicrobial concentration at which there was no observable bacterial growth in wells containing treated biofilm (OD600 < 0.1).

For MBIC measurement, after being washed, biofilm treated with or without DexA70 was exposed to the serial diluted antimicrobial agents for 1 h. The media were then removed and wells were washed three times with sterile saline solution. Finally, antimicrobial-free BHI was added, followed by incubation for 24 h at 37°C. MBEC was defined as the minimal antimicrobial concentration at which bacteria fail to regrow after antimicrobial exposure (OD600 < 0.1).

To accurately measure the effect of DexA70 on minimal concentrations of CHX and CPC to completely kill S. mutans in biofilm, biofilm was cultured in BHI broth with 1% sucrose for 12 h, and gently washed three times followed by antibiotics challenges with/without 500 nM DexA70 for 30 min. After treatment, biofilm was washed and scraped off the tubes and resuspended in 0.9% NaCl. Serially diluted suspensions were plated onto BHI agar plates in triplicate and anaerobically incubated for 48 h. Bacteria growth was measured by colony-forming units (CFU). Minimum biofilm viable cells eradication concentration (MBVEC) was defined as minimal antimicrobial concentration at which no colony grows on plate.

Chlorhexidine and cetylpyridinium chloride, agents commonly used in mouthwashes to treat dental caries, are known to be the most effective treatment currently applied in the dental field. It was reported that the concentrations of CHX and CPC contained in mouthwashes in the market (0.1 and 0.03% to 0.05%, respectively) are much higher than the sensitivity values of S. mutans (So Yeon and Si Young, 2019). To figure out if there is a possibility of reducing the antimicrobial agents in the mouthwashes, we tested the susceptibility of planktonic bacteria and biofilm to CHX and CPC respectively. 1 μg/ml (0.0001%) of CHX (Figure 1A) and 0.5 μg/ml (0.00005%) of CPC (Figure 1B) inhibit planktonic cells growth thoroughly, which are 1‰ of both agents used in mouthwashes. However, the effective concentrations to kill bacteria wrapped in biofilm is 0.5% for CHX, 5,000-fold higher than that against planktonic cells (Figure 1C) or 0.1% for CPC, 2,000-fold higher than that against planktonic cells (Figure 1D). This proves that biofilm formation dramatically promotes the bacterial drug resistance to antimicrobial agents.

Enzymatic targeting of biofilms by lysozyme has been considered as an alternative treatment for biofilm infection in certain bacteria, which generally have no side effect caused by antiseptics. The effect of lysozyme on S. mutans growth and biofilm formation was tested. The results showed that lysozyme had significant antimicrobial activity against the planktonic cells (Figure 2A) and hindered the biofilm formation at higher concentrations (Figure 2B), but it was not able to degrade the already formed biofilm (Figure 2C).

High resistance of S. mutans biofilm to antimicrobial agents and lysozyme imply that disrupting biofilm would be a better way to make an effective treatment. Since dextran is the major component of S. mutans biofilm EPS, dextranase and mutanase from various bacterial sources have been used to treat dental biofilms. However, most of them had little effect (Pleszczynska et al., 2010; Wiater et al., 2011; Wang et al., 2016).

The success of PlsG in P. aeruginosa biofilm degradation suggests that self-produced enzymes may be better candidates for combating the biofilm-related infections (Yu et al., 2015). Just like P. aeruginosa, S. mutans does encode a polysaccharide hydrolase, DexA, which plays an important role in biofilm formation. The functional similarity between PslG and DexA imply that DexA may also be a good treatment. DexA was then overexpressed and purified in E. coli BL21(DE3). The ability of DexA to hydrolyze S. mutans biofilm was tested. The results showed that although DexA degraded S. mutans biofilm to some extent, the effect was just not so good as expected (dispersed less than 50% of biofilm).

DexA belongs to glycoside hydrolase family 66, consisting of five regions from the N- to the C-termini: the N-terminal signal peptide sequence (N-terminal 24 amino acids), variable region (25–99), catalytic region (100–615), glucan-binding site (616–732), and C-terminal region (733–850). It has been reported that full-length DexA was easily digested into many peptide fragments by proteases, and the protein lacking of both N- and C-terminal regions showed higher catalytic activity (Kim et al., 2011). The crystal structure of this fragment (100–732, Dex70) has been determined, and the catalytic mechanism has also been elucidated (Suzuki et al., 2012). This reminds us that Dex70 may have better performance in S. mutans biofilm degradation.

DexA70 was expressed and purified as previously mentioned (Suzuki et al., 2012). Hydrolysis activity was measured by using DNS colorimetric method under different pH values and different temperatures. The optimum pH of DexA70 is at pH 5.0 (Supplementary Figure 1A), and the optimum temperature of DexA70 is at 40°C (Supplementary Figure 1B).

To determine the biofilm-inhibiting activity of DexA70, DexA70 was added to the culture medium at the beginning of incubation. Biofilm biomass decreased significantly compared with the controls (Figure 3A). IC50 (the concentration that can inhibit 50% of biofilm biomass) of DexA70 was ∼1 nM. The biofilm biomass of UA159 treated with 50 nM DexA70 was similar to that of the negative controls, which had no sucrose as substrate, indicating that DexA70 at 50 nM completely prevents biofilm formation.

The effect of DexA70 was further quantitatively characterized by CSLM images. EPS was stained in red by concanavalin A conjugated with tetramethylrhodamine and viable cells were stained in green by SYTO9. Markedly, the pellicles with DexA70 treatment lost about 99% of dextran (Figure 3B). The topographic surfaces, biofilm organization, and architecture were then systematically analyzed by using the COMSTAT plugin of ImageJ (Figure 3C). The architectural analysis of DexA70-treated biofilm showed 90 ± 5.9% reduction of the biofilm biomass and 86.7 ± 6.3% reduction of the viable cells in the biofilm matrix as compared with the untreated biofilm (Figure 3Ci). Similar effects were also observed for the biofilm thickness (Figure 3Cii) and the average diffusion distance (Figure 3Ciii). The reduction of surface-to-volume ratio suggested that the presence of DexA70 inhibited biofilm formation and caused architecture disruption (Figure 3Civ). The biofilm inhibition effect of DexA70 is not due to the inhibition of the bacterial growth since the addition DexA70 did not change the growth rate of S. mutans (Figures 3D,E).

Dextran is the major component of S. mutans biofilm and confers adherence and drug resistance onto the bacteria. Since DexA70 is able to prevent biofilm formation by degrading dextran, we think that it may also be able to disperse the existing biofilm. The activity of DexA70 on S. mutans biofilm was then tested. The results showed that the biofilm disassembly activity of DexA70 increased with concentration. DexA70 dispersed 43 ± 2.2% of biofilm at 50 nM and 77 ± 2% of biofilm at 500 nM in 30 min. By comparison, full-length DexA had a much lower activity at the same concentration (Figure 4A). Deletion of the N- and C-terminal regions may make the catalytic site more exposed and get access to the substrate easier, thus resulting in more potent activity on S. mutans biofilm.

The measurement of biofilm formation ability and cell growth showed that lysozyme kills planktonic bacteria but not biofilm bacteria and has no degradation effect on preformed biofilm, while DexA70 disperses biofilm without affecting bacterial growth (Figures 2, 3). Thus, we speculate that co-administration of DexA70 and lysozyme should have a synergistic effect on biofilm. To test the synergistic effect of DexA70 and lysozyme,biofilm was cultured in BHI broth with 1% sucrose for 12 h, and treated with the following measures separately: 500 nM DexA70; 2 mg/ml lysozyme; 500 nM DexA70 plus 2 mg/ml lysozyme. After 30-min incubation at 37°C, the reaction buffers were removed and the treated biofilms were subjected to testing by three methods. The amount of remaining biofilm was quantitatively measured by staining with 0.1% crystal violet. The results showed that co-administration of DexA70 and lysozyme resulted in as low as 15.85 ± 2.4% remaining biofilm (Figure 4B). The degradation rate of the biofilm by co-administration of DexA70 and lysozyme was much higher than the sum effects of DexA70 and lysozyme.

To visualize the effect of the treatments, the biofilms formed in each condition were stained by fluorescent dyes and characterized by a CLSM, and further analyzed by COMSTAT. The three-dimensional biofilm images also confirmed the synergistic effect of DexA70 and lysozyme in biofilm disruption. The biofilm untreated with enzymes were thick, compact, and uniform, whereas co-administration of DexA70 and lysozyme resulted in porous biofilms containing channels and voids (Figure 4C). Quantitative measurements of biofilm morphology are indicated in Figure 4D. The treatment of enzymes caused 81.8 ± 6.3% reduction of biofilm biomass and 76.6 ± 4.4% reduction of viable cells. The biofilm thickness and average diffusion distance were also decreased. Most strikingly, the surface-to-volume ratio of biofilm matrix treated with DexA70 and lysozyme increased greatly. The surface-to-volume ratio is the surface area divided by the bio-volume which reflects what fraction of the biofilm is in fact exposed to the environment. The significant erosion of the surfaces created a great deal of microcavities, which largely increased the contact of biofilm with the environment, causing higher sensitivity to environmental changes and challenge of antibiotics.

To make clear the synergistic mechanism of DexA70 and lysozyme, we measured the dry weights of biofilm and the number of viable cells of S. mutans in remaining biofilms after being treated in different ways. The biofilm treated with both DexA70 and lysozyme has the least remaining biomass (Figure 5A). To evaluate the viable cells, the remaining biofilm after treatment was washed with saline solution and then resuspended, serially diluted, and plated onto BHI agar plates in triplicate. The number of viable bacteria was counted by CFUs. As expected, both DexA70 and lysozyme treatments significantly reduced the total viable bacteria number in the remaining biofilms when added at the beginning of biofilm formation. Co-administration of DexA70 and lysozyme reduced the number of viable bacteria even further. In the case of existing biofilm, however, DexA70 was still effective while lysozyme completely lost killing activity. Strikingly, co-administration of DexA70 and lysozyme still showed a synergistic effect in reducing the number of viable bacteria in the remaining biofilm. This strongly suggests that co-administration of DexA70 and lysozyme makes a dispersing and killing strategy. Although S. mutans biofilm is fully resistant to lysozyme, hydrolysis of the glucan by DexA70 exposes the cells to lysozyme for killing (Figures 5B,C). This assay also eliminated the interference from planktonic cells, which provides a more accurate measurement for the effect of a specific treatment on biofilm.

To get a better understanding of the dispersing and killing mechanism, we carried out a LIVE/DEAD BacLight bacterial viability assay, in which all bacteria were stained in green by SYTO9 and dead bacteria were stained in red by PI. The untreated biofilm and biofilms treated by DexA70, lysozyme or both were examined on a CLSM (Figure 5D). A corresponding quantitative evaluation of live and dead cells from the morphological analysis of z-stacks is shown in Figure 5E. It is noteworthy that the synergistic effect of dispersing and killing is obvious since treatment with both enzymes resulted in a much more complete biofilm disruption; both the live and dead bacteria were mostly removed.

The effectiveness of the dispersing and killing strategy implies that DexA70 treatment may also sensitize S. mutans biofilm to commonly used antiplaque agents. To verify this hypothesis, we tested the MBIC (minimum biofilm inhibition concentration) of CHX and CPC against DexA70-treated 12-h-old biofilm. For both antimicrobial agents, DexA70-treated biofilm has a twofold lower MBIC in comparison of untreated biofilm (Figure 6A). To know whether antiplaque agents are able to eradicate the remaining surface-attached biofilm after 30 min of DexA70 treatment, we examined the MBEC (minimum biofilm eradication concentration). Similarly, for both antimicrobial agents, DexA70-treated biofilm has a twofold lower MBEC (Figure 6B).

The data of MBIC and MBEC also suggest a dispersing and killing mechanism. However, due to the complexity and heterogeneity of biofilm, MBIC and MBEC cannot quantitatively and accurately reflect the effectiveness of DexA70 treatment. We think that the number of viable S. mutans cells in the biofilm with different treatment would better reflect the effectiveness of dispersing and killing strategy. The result showed that, for both CHX and CPC, the minimal concentrations to completely kill S. mutans in biofilm [minimum biofilm viable cells eradication concentration (MBVEC)] treated with DexA70 were fivefold lower than those of untreated groups (Figure 6C). This indicated that DexA70 treatment sensitized S. mutans biofilm to antiplaque agents and formed a killing and dispersing strategy.