Cross-kingdom inhibition of bacterial virulence and communication by probiotic yeast metabolites

Abstract

Background: Probiotic milk-fermented microorganism mixtures (e.g., yogurt, kefir) are perceived as contributing to human health, and possibly capable of protecting against bacterial infections. Co-existence of probiotic microorganisms are likely maintained via complex biomolecular mechanisms, secreted metabolites mediating cell-cell communication, and other yet-unknown biochemical pathways. In particular, deciphering molecular mechanisms by which probiotic microorganisms inhibit proliferation of pathogenic bacteria would be highly important for understanding both the potential benefits of probiotic foods as well as maintenance of healthy gut microbiome.

Results: The microbiome of a unique milk-fermented microorganism mixture was determined, revealing a predominance of the fungus Kluyveromyces marxianus. We further identified a new fungus-secreted metabolite-tryptophol acetate-which inhibits bacterial communication and virulence. We discovered that tryptophol acetate blocks quorum sensing (QS) of several Gram-negative bacteria, particularly Vibrio cholerae, a prominent gut pathogen. Notably, this is the first report of tryptophol acetate production by a yeast and role of the molecule as a signaling agent. Furthermore, mechanisms underscoring the anti-QS and anti-virulence activities of tryptophol acetate were elucidated, specifically down- or upregulation of distinct genes associated with V. cholerae QS and virulence pathways.

Conclusions: This study illuminates a yet-unrecognized mechanism for cross-kingdom inhibition of pathogenic bacteria cell-cell communication in a probiotic microorganism mixture. A newly identified fungus-secreted molecule-tryptophol acetate-was shown to disrupt quorum sensing pathways of the human gut pathogen V. cholerae. Cross-kingdom interference in quorum sensing may play important roles in enabling microorganism co-existence in multi-population environments, such as probiotic foods and the gut microbiome. This discovery may account for anti-virulence properties of the human microbiome and could aid elucidating health benefits of probiotic products against bacterially associated diseases. Video Abstract.

Keywords: Biofilms; Kluyveromyces marxianus; Microbiome; Probiotic microorganisms; Quorum sensing; Tryptophol acetate; Vibrio cholerae.

Fig. 1

Microorganism composition of the kefir. a Distribution pie chart of microorganisms in the kefir, obtained through total DNA shotgun sequencing based on a BLAST comparison in the One Codex data platform for applied microbial genomics. b Imagestream® flow cytometry analysis depicting microorganisms size/shape distribution. The dot plot of autofluorescence vs. bright field detail intensities of 19380 events analyzed. The brightfield cell images were used to draw an initial dot plot to identify cells of interest and exclude debris. The populations corresponding to initial hand-picked galleries of images are automatically gated on dot plot and each gate is indicated with a different color. Representative microscopic images of the cell subpopulations presented in the scatterplot: (i) bacterial cells; (ii) fungal cells; (iii–iv) fungal/bacterial cell aggregates. Scale bars correspond to 5 μm

Fig. 2

Effects of the kefir on bacterial quorum sensing and bacterial biofilms. a Schematic description of the generic bioluminescence induction mechanism in the reporter QS Gram-negative bacterial strains. b QS inhibition in the bioluminescent reporter strains Vibrio cholerae MM920, Agrobacterium tumefaciens A136; QS activation observed in the case of Vibrio harveyi MM30. Bioluminescence recorded upon addition of the kefir crude extract in different dilutions. The respective autoinducers (CAI-1, C8-HSL, DPD) correspond to the control experiments (no inhibition/activation of bioluminescence). c Bar diagrams showing biofilm volumes percentage generated in three wild-type bacterial strains upon addition of the kefir crude extract to the growth media (striped bars) and non-treated bacteria (solid bars). The graph displays the means (±SD; n = 3) of biofilm volume per area, generated from three independent sets of confocal fluorescence microscopy experiments calculated through the IMARIS software. p value: * p < 0.12, **p < 0.01, ***p < 0.002 calculated by ANOVA followed by Tukey’s post hoc analysis. d Viability analysis of bacterial cells in the biofilms. Staphylococcus aureus, Salmonella enterica, and Pseudomonas aeruginosa viable cells were stained in green while dead cells were stained red with the BacLight® Dead/Live Kit

Fig. 3

Identification of tryptophol acetate secreted by Kluyveromyces marxianus. a HPLC chromatogram of (i) tryptophol acetate extracted from the K. marxianus monoculture crude with a retention time of 13.84 min (peak indicated in red); tryptophol acetate molecular structure is indicated. (ii) Tryptophol acetate identified in the kefir crude extract (peak shown in red at the same retention time). b MS spectrum acquired in positive enhanced mass spectrometry for the identified tryptophol acetate peak in the kefir crude extract (between 13.72 and 14.05 min). c Calibration curve of tryptopol acetate in the kefir crude extract constructed with the synthetic compound. The broken lines indicate the concentration of tryptophol acetate in the kefir extract (210 μM)

Fig. 4

Effects of tryptophol acetate on Vibrio cholerae quorum sensing and biofilm assembly. a Tryptophol acetate concentration-dependent inhibition of the CAI-1 QS system in V. cholerae MM920. Experiments were performed in triplicate and error bars represent standard deviation of the mean. b Concentration-dependent effect of tryptophol on the CAI-1 QS system in V. cholerae MM920. Experiments were performed in triplicate and error bars represent standard deviation of the mean. c Confocal fluorescence microscopy z-stacks showing V. cholerae biofilms. Excitations were at 488 nm and 561 nm; emission 490–588 nm and 604–735 nm, respectively. V. cholerae viable cells were stained in green while dead cells were stained red with the BacLight® Dead/Live Kit. Sizes of the biofilm images are 500 μm × 500 μm. (i) V. cholerae VC1 (wild-type); (ii) V. cholerae VC1 grown in the presence of 100 μM tryptophol acetate; (iii) V. cholerae MM920 mutant; (iv) V. cholerae MM920 incubated with 900 nM CAI-1; (v) V. cholerae MM920 incubated with both 900 nM CAI-1 and 100 μM tryptophol acetate. The graph (top right) displays the means (±SD) of biofilm volume per area, generated from three independent sets of confocal fluorescence microscopy experiments calculated through the IMARIS software. d V. cholerae biofilm mass analysis at different concentrations of tryptophol acetate (μM) obtained through crystal violet staining (The concentrations in μM are indicated by the different bar colors). Biofilms were stained after 24-hr growth. Error bars indicate the standard deviations of 4 measurements. *p < 0.001, **p < 0.0001, ***p < 0.000001 versus the control calculated by ANOVA followed by Tukey’s post hoc analysis

Fig. 5

Effect of tryptophol acetate on the expression of genes associated with quorum sensing in V. cholerae. a Relative gene expression in V. cholerae VC1 WT assessed by RT-qPCR. The blue bars correspond to bacteria untreated with tryptophol acetate, while the grey bars indicate gene expression levels recorded following addition of the compound. The relative magnitude of gene levels was defined as the copy number of cDNA of genes in the QS pathway normalized in relation to the expression of a reference housekeeping gene not affected by the treatment. Error bars indicate standard deviations of four independent cultures. *p < 0.002, **p < 0.0001, ***p < 0.05 versus the untreated bacteria calculated by ANOVA followed by Tukey’s post hoc analysis. b Scheme depicting the effects of tryptophol acetate upon QS gene regulation of biofilm formation and virulence of V. cholerae in high cell density conditions (as recorded in the RT-qPCR experiments). The solid black arrows indicate transcription activation, the grey arrows indicate inactivation, while the dashed lines account for transcription repression induced by tryptophol acetate. Genes shown in grey background with red borders—hapR and ctxA—were downregulated by tryptophol acetate, while blue backgrounds account for genes that were upregulated due to repression of hapR (mimicking low cell density conditions), see text

Fig. 6

Effect of tryptophol acetate on V. cholerae toxin secretion. a Cholerae Toxin B (CTB) level expression analyzed by anti-CTB western blotting. V. cholerae bacterial cells (VC1 WT strain) were cultured for 16 h at 30 ^oC without tryptophol acetate, and in the presence of 100 μM tryptophol acetate. The treated and untreated bacterial cultures were used to isolate the CTB (see “Methods” section). The samples were immunoblotted for CTB protein levels. Beta-actin used as loading control. b V. cholerae toxin production upon adding different concentrations of tryptophol acetate to the V. cholerae growing medium quantified with ELISA assay for CTX using GM-1 (the CTX receptor). The Y axis corresponds to percentage of CTX production in the growing medium. Error bars indicate standard deviations of four independent cultures. *p < 0.0001, versus the untreated bacteria calculated by ANOVA followed by Tukey’s post hoc analysis. c Confocal fluorescence microscopy images of HeLa cells grown for 16 h and exposed to extraction from VC1 WT cell lysate in the absence (i) and presence (ii) of 100 μM tryptophol acetate for 2 h. Cells were fixed and labeled with monoclonal anti-cholerae toxin, subunit B antibodies (red-excitations were at 633; emission 681 nm) and imaged. Nuclei were visualized upon co-staining of the cells with Hoechst 33342 (blue-excitations were at 405 nm; emission 445 nm). Individual channels and merged confocal images are shown. Scale bar: 50 μm. d Confocal fluorescence microscopy images showing the effects of CT extracted from VC1 WT cell lysate; cells were grown in the absence and presence of 100 μM tryptophol acetate. The HeLa cells were treated with CT for 16 h at 37 ^oC and CO₂ conditions. Propidium iodide staining (red) indicates dead cells and Syto 9 staining (green) indicates viable cells. Excitations were at 488 nm and 561 nm; emission 490–588 nm and 604–735 nm, respectively. (i) Viability staining of HeLa cells exposed to CT. (ii) Viability staining of HeLa cells exposed to CT under treatment of 100 μM tryptophol acetate. Scale bar: 100 μm