Impact of Bacillus subtilis Antibiotic Bacilysin and Campylobacter jejuni Efflux Pumps on Pathogen Survival in Mixed Biofilms

Abstract

The foodborne pathogen Campylobacter jejuni is typically found in an agricultural environment; in animals, such as birds, as an intestinal commensal; and also in food products, especially fresh poultry meat. Campylobacter interactions within mixed species biofilms are poorly understood, especially at the microscale. We have recently shown that the beneficial bacterium Bacillus subtilis reduces C. jejuni survival and biofilm formation in coculture by secreting the antibiotic bacillaene. We extend these studies here by providing evidence that besides bacillaene, the antagonistic effect of B. subtilis involves a nonribosomal peptide bacilysin and that the fully functional antagonism depends on the quorum-sensing transcriptional regulator ComA. Using confocal laser scanning microscopy, we also show that secreted antibiotics influence the distribution of C. jejuni and B. subtilis cells in the submerged biofilm and decrease the thickness of the pathogen's biofilm. Furthermore, we demonstrate that genes encoding structural or regulatory proteins of the efflux apparatus system (cmeF and cmeR), respectively, contribute to the survival of C. jejuni during interaction with B. subtilis PS-216. In conclusion, this study demonstrates a strong potential of B. subtilis PS-216 to reduce C. jejuni biofilm growth, which supports the application of the PS-216 strain to pathogen biofilm control. IMPORTANCE Campylobacter jejuni is a prevalent cause of foodborne infections worldwide, while Bacillus subtilis as a potential probiotic represents an alternative strategy to control this alimentary infection. However, only limited literature exists on the specific mechanisms that shape interactions between B. subtilis and C. jejuni in biofilms. This study shows that in the two species biofilms, B. subtilis produces two antibiotics, bacillaene and bacilysin, that inhibit C. jejuni growth. In addition, we provide the first evidence that specific pathogen efflux pumps contribute to the defense against B. subtilis attack. Specifically, the CmeDEF pump acts during the defense against bacilysin, while CmeR-dependent overexpression of CmeABC nullifies the bacillaene attack. The role of specific B. subtilis antibiotics and these polyspecific pumps, known for providing resistance against medically relevant antibiotics, has not been studied during bacterial competition in biofilms before. Hence, this work broadens our understanding of mechanisms that shape antagonisms and defense during probiotic-pathogen interactions.

Keywords: Bacillus subtilis; Campylobacter jejuni; antibiotics; bacillaene; bacilysin; biofilm formation; efflux pumps; secondary metabolites.

FIG 1

Inhibition of C. jejuni growth by B. subtilis PS-216 mutants in loci involved in the synthesis of the antibiotics bacillaene and bacilysin. PS-216 mutations are in loci involved in nonribosomal/polyketide synthesis (bacillaene [pks], bacilysin [bacA], and surfactin [srfAA]) and transcriptional regulatory protein ComA. (A) C. jejuni during mono- and coculture with B. subtilis mutant in the comA gene encoding transcriptional regulatory protein ComA. (B) C. jejuni during mono- and coculture with B. subtilis mutant in srfAA gene involved in nonribosomal peptide synthesis of surfactin. (C) C. jejuni during mono- and coculture with B. subtilis mutant in bacA gene in nonribosomal peptide synthesis of bacilysin. (D) C. jejuni during mono- and coculture with B. subtilis mutant in pks locus involved in polyketide synthesis of bacillaene. (E) C. jejuni during mono- and coculture with B. subtilis double mutants in loci involved in polyketide synthesis of bacillaene as nonribosomal synthesis of surfactin and bacilysin. All cocultures were grown in MHB medium under static microaerophilic conditions at 42°C for 24 h. Samples containing biofilm and broth were vortexed prior to plating. The results are presented as colony counts. Three biological and up to three technical repeats were used. The error bars represent the standard deviation of the mean. “a” and “b” represent statistically significant values, where “a” represents hypothesis testing between C. jejuni monoculture and C. jejuni in coculture with B. subtilis (mutant strains and WT), and “b” represents hypothesis testing between C. jejuni in coculture with B. subtilis mutant and C. jejuni in coculture with B. subtilis WT. Data were statistically evaluated using a two-sample t test (see Materials and Methods for details).

FIG 2

B. subtilis antibiotics bacillaene and bacilysin are antibiofilm mediators preventing C. jejuni from forming a submerged biofilm. (A) Schematic representations of an experimental model for investigating interactions between a pathogen (C. jejuni) and a potential antagonist (B. subtilis) under static conditions at 42°C in MHB medium using CLSM in the total volume of the well (left, height 1,800 μm of the well; right, schematic representing the ortho visualization of the submerged biofilm). The ortho view depicts the fluorescence in each cut section related to the x, y, and z dimensions of the submerged biofilm. The colored boxes (blue, red, green) each represent a different view through the biofilm. The larger panel labeled “z” is a two-dimensional distribution of the submerged biofilm in x-y dimension, where only the bottom z stack (3.5 μm) is presented. While the smaller side panels (x and y) represent combined z stacks through 100-μm depth of the submerged biofilm. (B) The CLSM images represent C. jejuni submerged biofilms incubated for 24 h in static microaerobic conditions at 42°C as a monoculture (control) compared to the phenomenon observed in coculture with PS-216 WT, PS-216 Δpks, PS-216 ΔbacA, and PS-216 Δpks ΔbacA strains. (C) Effect of 24 h of cultivation time on C. jejuni submerged biofilm formation expressed as biofilm thickness (μm). The results show the means and standard deviations for five independent experiments. Data were statistically evaluated using the Mann-Whitney test (see Materials and Methods for details). For CLSM analysis, we performed five biological experiments with five technical replicates (five wells). CLSM analysis was performed in three different position spots in each well where biofilms were grown.

FIG 3

C. jejuni loci for efflux apparatus system contribute to the defense against B. subtilis PS-216 in coculture. The growth of C. jejuni WT and C. jejuni ΔcmeB, ΔcmeF, ΔcmeG, ΔcmeR efflux pump mutants was measured as colony counts after 24 h of incubation under static conditions at 42°C in MHB medium in monoculture (results in yellow columns) and coculture (results in gray columns) with the B. subtilis PS-216 WT strain (A), the PS-216 mutant lacking the locus for polyketide antibiotic bacillaene (PS-216 Δpks) (B), the PS-216 mutant not producing the dipeptide antibiotic bacilysin (PS-216 ΔbacA) (C), and the PS-216 mutant lacking loci for both antibiotics: bacillaene and bacilysin (PS-216 ΔpksΔbacA) (D). In panel D, the PS-216 WT effect on C. jejuni growth (red column) was included. Experiments were performed in at least three (D), five (B and C), or eight (A) biological replicates. Each biological replicate was always performed in three technical replicates. Samples containing biofilm and broth were vortexed prior to plating. The results are presented as colony counts. The error bars represent the standard deviations of the mean. In panels B and C, “a” and “b” represents statistically significant values, where “a” represents hypothesis testing between C. jejuni monoculture and C. jejuni in coculture with B. subtilis, and “b” represents hypothesis testing between C. jejuni mutant in coculture with B. subtilis and C. jejuni WT in coculture with B. subtilis. In panel D, the asterisk (*) represents statistically significant values. Data were statistically evaluated using a two-sample t test (see Materials and Methods for details).