Intervening in Symbiotic Cross-Kingdom Biofilm Interactions: a Binding Mechanism-Based Nonmicrobicidal Approach

Abstract

Early childhood caries is a severe oral disease that results in aggressive tooth decay. Particularly, a synergistic association between a fungus, Candida albicans, and a cariogenic bacterium, Streptococcus mutans, promotes the development of hard-to-remove and highly acidic biofilms, exacerbating the virulent damage. These interactions are largely mediated via glucosyltransferases (GtfB) binding to mannans on the cell wall of C. albicans Here, we present an enzymatic approach to target GtfB-mannan interactions in this cross-kingdom consortium using mannan-degrading exo- and endo-enzymes. These exo- and endo-enzymes are highly effective in reducing biofilm biomass without killing microorganisms, as well as alleviating the production of an acidic pH environment conducive to tooth decay. To corroborate these results, we present biophysical evidence using single-molecule atomic force microscopy, biofilm shearing, and enamel surface topography analyses. Data show a drastic decrease in binding forces of GtfB to C. albicans (∼15-fold reduction) following enzyme treatment. Furthermore, enzymatic activity disrupted biofilm mechanical stability and significantly reduced human tooth enamel demineralization without cytotoxic effects on gingival keratinocytes. Our results represent significant progress toward a novel nonbiocidal therapeutic intervention against pathogenic bacterial-fungal biofilms by targeting the interkingdom receptor-ligand binding interactions.IMPORTANCE Biofilm formation is a key virulence factor responsible for various infectious diseases. Particularly, interactions between a fungus, Candida albicans, and a bacterium, Streptococcus mutans, have been known to play important roles in the pathogenesis of dental caries. Although some antimicrobials have been applied to treat fungal-involved biofilm-associated diseases, these often lack targeting polymicrobial interactions. Furthermore, these may not be appropriate for preventive measures because these antimicrobials may disrupt ecological microbiota and/or induce the prevalence of drug resistance over time. By specifically targeting the interaction mechanism whereby mannoproteins on the C. albicans surface mediate the cross-kingdom interaction, we demonstrated that mannoprotein-degrading enzymes can effectively disrupt biofilm interactions without microbiocidal effects or causing cytotoxicity to human cells. This suggests a potential application as a targeted approach for intervening a pathogenic cross-kingdom biofilm associated with a costly and unresolved oral disease.

Keywords: Candida albicans; Streptococcus mutans; mannan-degrading enzymes; nonmicrobicidal approach; polymicrobial interaction.

FIG 1

Activity profiles for MDEs in MES buffer. Enzyme activities were measured at different time points for α-mannosidase (A), β-mannosidase (B), and β-mannanase (C). All MDEs had similar activity profiles for all time points. Activities of enzymes in acidic to neutral pH ranges were also determined for α-mannosidase (D), β-mannosidase (E), and β-mannanase (F). (n ≥ 3).

FIG 2

Effect of MDEs on the cell wall of C. albicans and its binding potential with GtfB. Dose-dependent degradation of the cell wall mannans in the supernatant (A) and a corresponding decrease in the pellet of C. albicans (B), and the amount of glucans formed on each C. albicans with or without MDE treatment (C). The amount of mannans on MDE-treated C. albicans in supernatant increased while it decreased from the microbial pellet. In the presence of sucrose, smaller amounts of bound GtfB in MDE-treated C. albicans led to reduced glucan formation. Panel C statistics employed one-way ANOVA with P < 0.0001 post hoc; ***, P < 0.001; **** P < 0.0001 against vehicle control using Dunnett’s method (n ≥ 3).

FIG 3

Efficacy of MDEs against S. mutans-C. albicans biofilms. (A) The pH of biofilm supernatant. (B) Dry weight per biofilm. (C) CFU of S. mutans per biofilm. (D) CFU of C. albicans per biofilm. At optimal units, all MDEs had a significant antibiofilm effect on S. mutans-C. albicans biofilms as measured at 18, 28, and 42 h. (E) Representative confocal images of untreated and MDE-treated biofilms at 18 h. Scale bar indicates 20 μm. Statistics: *** represents P < 0.001 for unpaired t tests against the vehicle control (n ≥ 3).

FIG 4

Effect of MDE treatment on the mechanical stability of S. mutans-C. albicans biofilms. (A) Schematic diagram of shear-induced biofilm mechanical strength tester. (B) Remaining biofilm biomass before and after applying shear stress of 0.18 N/m² for 10 min. (C) Representative confocal images of biofilms after shearing. Scale bar indicates 20 μm. Statistics: *** represents P < 0.001 for unpaired t tests against the intact biofilms (n ≥ 3).

FIG 5

Demineralization of human enamel surface by S. mutans-C. albicans biofilms with or without β-mannanase treatment. (A) Schematic diagram for human enamel slab preparation. Shown are representative confocal surface-topography of intact enamel (B), enamel after forming biofilms without β-mannanase treatment (C), and enamel after forming biofilms with β-mannanase treatment (D). Panels (C) and (D) were scanned after removing biofilms from the enamel. (E) Average surface roughness values of intact enamel slabs and slabs that had biofilms with or without β-mannanase treatment; ***, P < 0.001.

FIG 6

Binding forces of GtfB to MDE treated C. albicans. Shown are adhesion force histograms of GtfB to untreated C. albicans (A), α-mannosidase treated C. albicans (B), β-mannosidase treated C. albicans (C), and β-mannanase treated C. albicans (D) at optimal units and 5 min of treatment. All MDEs drastically reduced the binding forces between GtfB and the cell wall of C. albicans. Kolmogorov-Smirnov tests were used to compare frequency distributions of untreated versus each MDE (P < 0.0001 in each case) (n ≥ 3).

FIG 7

Toxicity assay of MDEs on microbes and human gingival keratinocytes. (A) Normalized cell viability for HGKs after exposure to optimal units of MDEs for 1 h and 24 h, and CFU values for S. mutans (B) and C. albicans (C) after treatment with optimal units of MDEs for 5 min. No loss in HGK cell viability was observed for MDE treatments. Negative control and positive control represent vehicle control and 3% H₂O₂ control, respectively. There were no significant differences in microbial cell viability with or without MDE treatment. Statistics: one-way ANOVA with P < 0.01; post hoc: ** represents P < 0.0001 against vehicle control using Dunnett’s method (n ≥ 3); ns, not significant.