Bacillus subtilis-derived peptides disrupt quorum sensing and biofilm assembly in multidrug-resistant Staphylococcus aureus

Abstract

Multidrug-resistant Staphylococcus aureus is one of the most clinically important pathogens in the world, with infections leading to high rates of morbidity and mortality in both humans and animals. The ability of S. aureus to form biofilms protects cells from antibiotics and promotes the transfer of antibiotic resistance genes; therefore, new strategies aimed at inhibiting biofilm growth are urgently needed. Probiotic species, including Bacillus subtilis, are gaining interest as potential therapies against S. aureus for their ability to reduce S. aureus colonization and virulence. Here, we search for strains and microbially derived compounds with strong antibiofilm activity against multidrug-resistant S. aureus by isolating and screening Bacillus strains from a variety of agricultural environments. From a total of 1,123 environmental isolates, we identify a single strain B. subtilis 6D1, with a potent ability to inhibit biofilm growth, disassemble mature biofilm, and improve antibiotic sensitivity of S. aureus biofilms through an Agr quorum sensing interference mechanism. Biochemical and molecular networking analysis of an active organic fraction revealed multiple surfactin isoforms, and an uncharacterized peptide was driving this antibiofilm activity. Compared with commercial high-performance liquid chromatography grade surfactin obtained from B. subtilis, we show these B. subtilis 6D1 peptides are significantly better at inhibiting biofilm formation in all four S. aureus Agr backgrounds and preventing S. aureus-induced cytotoxicity when applied to HT29 human intestinal cells. Our study illustrates the potential of exploring microbial strain diversity to discover novel antibiofilm agents that may help combat multidrug-resistant S. aureus infections and enhance antibiotic efficacy in clinical and veterinary settings.

Importance: The formation of biofilms by multidrug-resistant bacterial pathogens, such as Staphylococcus aureus, increases these microorganisms' virulence and decreases the efficacy of common antibiotic regimens. Probiotics possess a variety of strain-specific strategies to reduce biofilm formation in competing organisms; however, the mechanisms and compounds responsible for these phenomena often go uncharacterized. In this study, we identified a mixture of small probiotic-derived peptides capable of Agr quorum sensing interference as one of the mechanisms driving antibiofilm activity against S. aureus. This collection of peptides also improved antibiotic killing and protected human gut epithelial cells from S. aureus-induced toxicity by stimulating an adaptive cytokine response. We conclude that purposeful strain screening and selection efforts can be used to identify unique probiotic strains that possess specially desired mechanisms of action. This information can be used to further improve our understanding of the ways in which probiotic and probiotic-derived compounds can be applied to prevent bacterial infections or improve bacterial sensitivity to antibiotics in clinical and agricultural settings.

Keywords: Bacillus subtilis; Staphylococcus aureus; antibiotic resistance; biofilm; peptide; postbiotic; probiotic; quorum sensing interference.

Fig 1

B. subtilis 6D1 exhibits antibiofilm activity against S. aureus and harbors unique genetic traits not observed in closely related B. subtilis strains. (A) Antibiofilm activity of 16 Bacillus cell-free extracts against S. aureus 29213. Error bars represent mean ± SD of eight independent replicates. (B) The assembled B. subtilis 6D1 genome (C) B. subtilis 6D1 shares > 95% identity with a group of B. subtilis genomes (purple) and is more distantly related to common laboratory type strains (yellow dots). (D) Pangenome analysis identified 156 unique genes (green) in B. subtilis 6D1 that were not detected in any strain from this closely related group. Singleton gene clusters identified range from 0 to 233, and the number of gene clusters identified range from 0 to 5,000.

Fig 2

B. subtilis 6D1 inhibits S. aureus biofilm but not planktonic growth. (A and B) Compared with 0.5× MIC of Sulfamethoxazole/Trimethoprim, addition of 10% vol/vol B. subtilis 6D1 cell-free extract inhibits methicillin-susceptible (MSSA) and methicillin-resistant (MRSA) S. aureus biofilm growth, but not planktonic growth. Error bars represent mean ± SD of 12 independent replicates. Differences in biofilm inhibition were analyzed using a 2-tailed unpaired t-test where (***) P < 0.001, (****) P < 0.0001 compared with the untreated control. (C) Competition experiments performed in planktonic and biofilm conditions for 24 hours show B. subtilis 6D1 outcompetes S. aureus in a biofilm but not in a planktonic environment. Error bars represent mean ± SD of four independent replicates (**) P < 0.01. (D) Macroscopic observation of S. aureus ATCC 29213 biofilms grown for 24 hours and washed prior to SYTO 9 staining (E) SYTO 9 staining visualized by confocal microscopy confirmed antibiofilm activity of 10% vol/vol B. subtilis 6D1 CFE (F) compared with an untreated control of S. aureus ATCC 29213.

Fig 3

B. subtilis 6D1 cell-free extracts (CFEs) reduce S. aureus ATCC 29213 biofilm growth, disassemble mature biofilm, and improve biofilm inhibition when applied in conjunction with low doses of antibiotics. (A) S. aureus (5 × 10⁵ CFU/mL) was grown on polystyrene plates at 37°C for 0, 1, 6, 12, or 18 hours, followed by treatment with B. subtilis 6D1 CFE (10% vol/vol) at 37°C for up to 24 hours (B) S. aureus (5 × 10⁵ CFU/mL) was grown on polystyrene plates at 37°C for 1, 6, 12, 18, or 24 hours in the presence or absence of B. subtilis 6D1 CFE (10% vol/vol) (C) B. subtilis 6D1 CFE (20%, 10%, 5%, or 2.5% vol/vol) was applied concurrently with S. aureus at T0 prior to staining biofilm at 24 hours (D) B. subtilis 6D1 CFE (20%, 10%, 5%, or 2.5% vol/vol) was applied to S. aureus 24 hour biofilms and incubated at 37°C shaking at 100 rpm for 2 hours prior to staining residual biofilm. (E–G) B. subtilis 6D1 CFE (10% vol/vol) was applied in conjunction with 0.5 µg/mL antibiotic at T0 prior to measuring S. aureus biofilm growth at 24 hours; error bars represent mean ± SD in 12 independent replicates. Differences in biofilm inhibition and disruption were analyzed using a 2-tailed unpaired t-test (A–D) or one-way ANOVA, followed by Tukey’s multiple comparisons (E–G) where (*) P < 0.05; (**) P < 0.01; (***) P < 0.001.

Fig 4

B. subtilis 6D1 derived peptides inhibit biofilm growth in all Agr backgrounds. (A) B. subtilis 6D1 ethyl acetate extracts exhibited dose-dependent antibiofilm activity against S. aureus strains with different AgrC receptors (agr I-IV) (*) P<0.0001 (B) Antibiofilm activity of B. subtilis 6D1 ethyl acetate (EtOAc) extracts (500 µg/mL) further separated by combiflash fractionation. Error bars represent mean ± SD in 12 independent replicates. Differences in biofilm inhibition were analyzed using one-way ANOVA, followed by Tukey’s multiple comparisons where (***) P<0.001, (****) P<0.0001 (C) Liquid Chromatography with tandem mass spectrometry spectra of B. subtilis 6D1 fraction 100%P1 m/z values that matched known molecules in the Global Natural Product Social Molecular Networking (GNPS) database. Black (observed) spectra represent peaks identified in fraction 100% P1, and green (reference) spectra represent GNPS database spectra. The x-axis represents m/z and the y-axis represents the relative abundance of the various ions.

Fig 5

B. subtilis 6D1 100%P1 fraction exhibits stronger antibiofilm activity compared with commercial HPLC grade surfactin obtained from B. subtilis. (A) Antibiofilm activity of titrated concentrations of commercial surfactin and B. subtilis 6D1 fraction 100% P1 against S. aureus ATCC 29213. (B) Live (SYTO 9)/dead(PI) staining and confocal image analysis comparing the arithmetic mean fluorescence intensity of S. aureus ATCC 29213 biofilms treated with 500 µg/mL commercial surfactin and B. subtilis 6D1 fraction 100% P1, scale bar = 70 µm. Error bars represent mean ± SEM of 15 independent fields of view (C) Antibiofilm activity of commercial surfactin and B. subtilis 6D1 fraction 100% P1 against S. aureus strains with different AgrC receptors (agr I-IV) (D) Antibiofilm activity of commercial surfactin and B. subtilis 6D1 fraction 100% P1 against clinical S. aureus (E) and S. epidermidis strains. Error bars represent mean ± SD of 12 independent replicates. Differences in biofilm inhibition were analyzed using a 2-tailed unpaired t-test where (*) P < 0.05; (**) P < 0.01; (***) P < 0.001; (****) P < 0.0001.

Fig 6

B. subtilis 6D1 modulates gene expression associated with S. aureus Agr quorum sensing and biofilm development. (A) B. subtilis 6D1 (blue) and 10% vol/vol B. subtilis 6D1 cell-free extract (teal) and (B) B. subtilis 6D1 fraction 100%P1 (light blue) and commercial surfactin (pink) were applied to S. aureus ATCC 29213 for 24 hours before RNA extraction and (left) Agr associated and (right) biofilm and stress-associated gene expression analysis via RT-qPCR. Error bars represent mean ± SD of nine independent replicates. Differential gene expression data were analyzed using one-way ANOVA followed by Tukey’s multiple comparisons using log2-transformed data where (*) P < 0.05; (**) P < 0.01; (***) P < 0.001 (****) P < 0.0001.

Fig 7

B. subtilis 6D1 reduces S. aureus ATCC 29213 virulence in a human intestinal cell line. (A) Vero cells pretreated for 2 hours with increasing concentrations of B. subtilis 6D1 CFE reduced cytotoxicity of S. aureus ATCC 29213 (Sa29213) induced cell death; data normalized to Sa29213 positive infection control. (B) Cytotoxicity of Vero cells pretreated for 2 hours with increasing concentrations of fraction 100%P1 and surfactin delivered in DMSO in the presence and (C) absence of a Sa29213 challenge; data normalized to LDH positive control. (D) HT29 cells pretreated for 2 hours with increasing concentrations of B. subtilis 6D1 CFE inhibit Sa29213 induced cytotoxicity; data normalized to Sa29213 positive infection control (E) HT29 cells pretreated for 2 hours with B. subtilis 6D1 CFE increase production of the anti-inflammatory cytokine IL-10 (F) and inhibits the production of the pro-inflammatory cytokine TNFα brought on by Sa29213 challenge. Error bars represent mean ± SD in 6 independent replicates. Vero and HT29 cell cytotoxicity results were analyzed using one-way ANOVA, followed by Tukey’s multiple comparisons relative to the positive infection control. Cytokine levels were compared using Welch’s non-parametric tests relative to the positive infection control where (*) P<0.05; (**) P<0.01; (***) P<0.001 (****) P<0.0001, ND = not detected.