Phenol-Soluble Modulins From Staphylococcus aureus Biofilms Form Complexes With DNA to Drive Autoimmunity

Abstract

The bacterial amyloid curli, produced by Enterobacteriales including Salmonella species and Escherichia coli, is implicated in the pathogenesis of several complex autoimmune diseases. Curli binds to extracellular DNA, and these complexes drive autoimmunity via production of anti-double-stranded DNA autoantibodies. Here, we investigated immune activation by phenol-soluble modulins (PSMs), the amyloid proteins expressed by Staphylococcus species. We confirmed the amyloid nature of PSMs expressed by S. aureus using a novel specific amyloid stain, (E,E)-1-fluoro-2,5-bis(3-hydroxycarbonyl-4-hydroxy) styrylbenzene (FSB). Direct interaction of one of the S. aureus PSMs, PSMα3, with oligonucleotides promotes fibrillization of PSM amyloids and complex formation with bacterial DNA. Finally, utilizing a mouse model with an implanted mesh-associated S. aureus biofilm, we demonstrated that exposure to S. aureus biofilms for six weeks caused anti-double-stranded DNA autoantibody production in a PSM-dependent manner. Taken together, these results highlight how the presence of PSM-DNA complexes in S. aureus biofilms can induce autoimmune responses, and suggest an explanation for how bacterial infections trigger autoimmunity.

Keywords: PSM; Phenol Soluble Modulins; SLE; Staphycoccus aureus; autoimmune disease; biofilm; curli; mesh.

Figure 1

Staphylococcus aureus biofilms contain amyloid that can be detected by an amyloid specific stain. (A) Confocal Laser Scanning Microscopy (CLSM) images of in vitro biofilms of S. aureus SH1000 (lab strain) and SH1000 phenol soluble modulin mutant (Δpsm) were stained with syto9 (green) and FSB (red) to determine the expression of PSM amyloids. (B) Crystal Violet staining of pellicle biofilms grown in glass tubes of SH1000, SH1000 Δpsm, and UTI89 (E. coli clinical isolate). (C) CLSM surface images of in vitro biofilms of S. aureus SH1000 and E coli UTI89 were stained with syto9 (green) and FSB (red). (D) Overall biofilm thickness of SH1000 and UTI89 biofilms as determined by Leica TCS imaging software. Mean and SEM graphed; significance was calculated using Unpaired t test (*, P < 0.05).

Figure 2

DNA promotes PSM fibrillation via direct association. (A) Crystal structure of PSMα3 fiber (PDB: 5I55) with electric potentials of charge residues projected on surface model. The electric potentials were calculated using Chimera software. (B) Fluorescence staining of PSMα3-DNA complexes using nucleic acid dye BOBO-3 iodine (red) and Thioflavin T, which only binds to fibrillar amyloid structures. Scale bar: 3μm.(C) Fibrillization of synthetic PSMα3 with or without CpG DNA was monitored by Thioflavin T fluorescence and reported as relative fluorescent units (RFU). Mean and SEM graphed; significance was calculated using One-way ANOVA with Tukey’s multiple comparisons test (*, P < 0.05).

Figure 3

Intraperitoneal injection of synthetic PSM⍺3 fibrillized with CpG DNA elicits an autoantibody response. (A) NZBxW/F1 or (B) Balb/c mice were injected bi-weekly with PSMα3 fibrillzed alone (black bar) or fibrillzed in the presence of CpG DNA (blue bar) or control mice were injected with sterile PBS (gray bar). Mice were tail bled weekly and the production of anti-dsDNA autoantibodies were detected (Optical density 650-405nm). Black dotted line represents the maximum anti-dsDNA autoantibody production of positive control sera and the red dotted line represent autoantibody detection in naïve serum. Mean and SEM graphed; significance was calculated using a 2-way ANOVA and Tukey’s multiple comparisons tests (*, P < 0.05).

Figure 4

S. aureus-colonized mesh implantation induces autoantibody production. To determine the expression of PSM amyloids, Confocal Laser Scanning Microscopy (CLSM) images of in vitro biofilms of S. aureus (A) SH1000 (WT) and (C) Δpsm SH1000 stained with syto9 (green) and Congo red (red); or (B) SH1000 (WT) stained with syto9 (green) and amyloid-specific dye 1-Fluoro-2,5-bis[(E)-3-carboxy-4-hydroxystyryl]benzene (FSB). Biofilms visualized at 100x, solid scale bar represents 50µM and dashed scale bar represents 5µM. S. aureus WT (black bars) or Δpsm (blue bars) biofilm-colonized mesh or control mesh was implanted subcutaneously into the back flanks of (D) NZBxW/F1 or (E) Balb/c mice. Blood was collected via tail bleeding every week and sampled for production of dsDNA autoantibodies (optical density 650-405nm). Black dotted line represents the maximum autoantibody production of positive control sera and the red dotted line represents autoantibody detection in naïve serum. Mean and SEM graphed; significance was calculated using a 2-way ANOVA and Tukey’s multiple comparisons tests (*, P < 0.05). (F) Biofilms of S. aureus SH1000 wildtype (WT) (black bar) or Δpsm mutant (blue bars) were grown on mesh in either tryptic soy broth or peptone-based media (PNG) for 48 hours at 37°C. Biofilms were sonicated and the recovered bacteria were enumerated as colony forming units. Mean and SEM graphed; significance was calculated using a 1-way ANOVA and multiple comparisons tests. No statistical significance was determined.

Figure 5

S. aureus-colonized mesh implantation induces autoantibody production dependent on TLRs. (A) S. aureus WT colonized mesh or (B) control mesh was implanted subcutaneously into the back flanks of C57BL/6 wildtype (WT) (black bars), TLR2^-/- (blue bars), TLR9 mutant (TLR9M) (gray bars), and TLR2^-/- - TLR9M (TLR2^-/–9M) (red bars) mice. Blood was collected via tail bleeding every week and sampled for production of anti-dsDNA autoantibodies (optical density 650-405nm). Black dotted line represents the maximum autoantibody production of positive control sera and the red dotted line represent autoantibody detection in naïve serum. Mean and SEM graphed; significance was calculated using a 2-way ANOVA and Tukey’s multiple comparisons tests (***, P < 0.001).

Figure 6

Intraperitoneal infection with S. aureus wildtype SH1000 or Δpsm mutant does not induce autoantibody production. (A) NZBxW/F1 or (B) Balb/c mice were injected intraperitoneally with 10⁷ S. aureus WT (black bars) or psm mutant (Δpsm) (blue bars) or PBS-injected control mice (gray bars). Serum was collected weekly to measure the levels of anti-dsDNA autoantibody production (optical density 650-405nm). Black dotted line represents the maximum autoantibody production of positive control sera and the red dotted line represent autoantibody detection in naïve serum. Mean and SEM graphed; significance was calculated using a 2-way ANOVA and Tukey’s multiple comparisons tests. No statistical significance was determined.