Bacterial biofilms are associated with numerous human infections. The predominant protein expressed in enteric biofilms is the amyloid curli, which forms highly immunogenic complexes with DNA. Infection with curli-expressing bacteria or systemic exposure to purified curli-DNA complexes triggers autoimmunity via the generation of type I interferons (IFNs) and anti-double-stranded DNA antibodies. Here, we show that DNA complexed with amyloid curli powerfully stimulates Toll-like receptor 9 (TLR9) through a two-step mechanism. First, the cross beta-sheet structure of curli is bound by cell-surface Toll-like receptor 2 (TLR2), enabling internalization of the complex into endosomes. After internalization, the curli-DNA immune complex binds strongly to endosomal TLR9, inducing production of type I IFNs. Analysis of wild-type and TLR2-deficient macrophages showed that TLR2 is the major receptor that drives the internalization of curli-DNA complexes. Suppression of TLR2 internalization via endocytosis inhibitors led to a significant decrease in Ifnβ expression. Confocal microscopy analysis confirmed that the TLR2-bound curli was required for shuttling of DNA to endosomal TLR9. Structural analysis using small-angle X-ray scattering revealed that incorporation of DNA into curli fibrils resulted in the formation of ordered curli-DNA immune complexes. Curli organizes parallel, double-stranded DNA rods at an inter-DNA spacing that matches up well with the steric size of TLR9. We also found that production of anti-double-stranded DNA autoantibodies in response to curli-DNA was attenuated in TLR2- and TLR9-deficient mice and in mice deficient in both TLR2 and TLR9 compared to wild-type mice, suggesting that both innate immune receptors are critical for shaping the autoimmune adaptive immune response. We also detected significantly lower levels of interferon-stimulated gene expression in response to purified curli-DNA in TLR2 and TLR9 deficient mice compared to wild-type mice, confirming that TLR2 and TLR9 are required for the induction of type I IFNs. Finally, we showed that curli-DNA complexes, but not cellulose, were responsible elicitation of the immune responses to bacterial biofilms. This study defines the series of events that lead to the severe pro-autoimmune effects of amyloid-expressing bacteria and suggest a mechanism by which amyloid curli acts as a carrier to break immune tolerance to DNA, leading to the activation of TLR9, production of type I IFNs, and subsequent production of autoantibodies.

No abstract

Bacterial biofilms, especially those associated with implanted medical devices, are difficult to eradicate. Curli amyloid fibers are important components of the biofilms formed by the Enterobacteriaceae family. Here, we show that a human monoclonal antibody with panamyloid-binding activity (mAb 3H3) can disrupt biofilms formed by Salmonella enterica serovar Typhimurium in vitro and in vivo. The antibody disrupts the biofilm structure, enhancing biofilm eradication by antibiotics and immune cells. In mice, 3H3 injections allow antibioticmediated clearance of catheter-associated S. Typhimurium biofilms. Thus, monoclonal antibodies that bind a pan-amyloid epitope have potential to prevent or eradicate bacterial biofilms.

No abstract

Reactive arthritis, an autoimmune disorder, occurs following gastrointestinal infection with invasive enteric pathogens, such as Salmonella enterica. Curli, an extracellular, bacterial amyloid with cross beta-sheet structure can trigger inflammatory responses by stimulating pattern recognition receptors. Here we show that S. Typhimurium produces curli amyloids in the cecum and colon of mice after natural oral infection, in both acute and chronic infection models. Production of curli was associated with an increase in anti-dsDNA autoantibodies and joint inflammation in infected mice. The negative impacts on the host appeared to be dependent on invasive systemic exposure of curli to immune cells. We hypothesize that in vivo synthesis of curli contributes to known complications of enteric infections and suggest that cross-seeding interactions can occur between pathogen-produced amyloids and amyloidogenic proteins of the host.