Immune Recognition of S. Typhimurium Biofilms via Amyloids and Extracellular DNA

Abstract

Salmonella enterica serovar Typhimurium is an important cause of gastroenteritis in the United States and the developing world. Biofilm growth is an significant mechanism, which S. Typhimurium utilizes to contaminate food products and survive in the environment. Biofilms are also an important part of the infectious process for many pathogenic bacteria. As part of the biofilm, S. Typhimurium produces an extracellular matrix consisting of cellulose, extracellular DNA, and most importantly, the amyloid protein curli. Similar to amyloids associated with human diseases, curli is recognized by the innate immune system through Toll-Like Receptors (TLRs). Here, we studied the immune receptors recognizing curli as well as interactions between eDNA and curli during biofilm development in order to glean a better understanding of these complex bacterial communities and the immune response to them. Recently, our lab demonstrated that curli fibers are recognized by the TLR2/TLR1 complex. CD14 has been shown to be a common adaptor protein for TLR2/TLR1 complex in response to one of its ligands, tri-acylated lipopeptide, Pam3CSK4. In order to study the role of CD14 in the immune receptor complex recognizing curli, we utilized HeLa 57A cells, a human cervical cancer cell line that has a stably transfected luciferase reporter for Nf-κB activation. When these cells were transiently transfected with TLR2 and TLR1 together or with the addition of membrane-bound CD14, NfκB activation was enhanced by the presence of CD14 in response to purified curli, GST-tagged curli subunit (GST-CsgA), and the control lipopeptide Pam3CSK4. Soluble CD14 also increased NfκB activation in response to purified curli. Bone marrow derived macrophages (BMDM) from wild type (C57BL/6) mice produced more IL-6 and nitric oxide in response to stimulation with purified curli, GST-CsgA, and Pam3CSK4, than BMDMs deficient in CD14. Binding assays demonstrated direct binding of curli to all members of this hypothesized trimolecular complex, TLR2, TLR1, and CD14. Utilizing synthetic peptides corresponding to the fourth and fifth repeat of the CsgA monomer, CsgA R4-5, and its modified version, CsgA R4-5N122A deficient in forming amyloid fibers, we also showed that binding to CD14, and CD14 enhancement of IL-6 production required the fibrillar amyloid structure of curli. To study interactions between curli and eDNA in biofilms and the resulting immune response generated to composites formed by these ECM components, we analyzed biofilms of GFP expressing S. Typhimurium using confocal laser scanning microscopy (CLSM). Staining for amyloids with Congo Red revealed the presence of curli in the biofilms and staining with propidium iodide demonstrated the presence of extracellular DNA in the biofilms. Co-staining with TOTO-1, a nucleic acid stain, and Congo Red showed co-localization of the fluorescent signal for these molecules within the biofilms. DNase I treatment of the biofilms produced no significant change in biofilm thickness by confocal microscopy signifying that the biofilm, possibly eDNA, was resistant to DNase treatment. This was further confirmed by the presence of DNA in purified curli fibers, which were treated twice with DNase and RNase. Polymerization assays showed acceleration of amyloid polymerization in the presence of DNA from both bacteria and salmon sperm. CLSM of bone marrow derived dendritic cells demonstrated that DCs are able to sample antigens from biofilms. BMDCs also produced robust quantities of proinflammatory cytokines in response to wild type, msbB, and ΔfliCfljB S. Typhimurium biofilms and purified amyloid/DNA composites as measured by ELISA. Using BMDCs deficient in TLR2 and TLR9, we found that this cytokine production was partially dependent on TLR2, but did not require TLR9. Together, these findings significantly broaden our understanding of S. Typhimurium biofilms and the immune response to ECM components present in its biofilms. We now understand that a trimolecular complex of TLR2/TLR1/CD14 is required for full response to curli by innate immune cells. We also discerned that interactions between biofilm components aid biofilm development and create composites that are highly immunogenic. This new information enhances the need to explore the interaction between composite ligands and the immune system rather than only studying ligands individually.