Mechanisms of host-commensal interactions with Roseomonas mucosa that modulate innate immune responses to infection with Staphylococcus aureus and support homeostasis of human skin

National Institute Of Allergy And Infectious Diseases

Investigators

Abstract

We are interested in how commensal bacteria of the skin support its homeostasis and innate immune responses to elucidate the mechanisms of host-microbe interactions in the context of commensal, rather than pathogenic, bacteria. Such interactions are foundational to the physiology of barrier sites, but there is only limited understanding of their mechanisms and how they influence health and disease. Our investigation thus far has focused on using in vitro models with primary human keratinocytes and fibroblasts (epithelial cells of the skin) to evaluate cytokines, chemokines and antimicrobial peptides (AMPs) produced in response to Roseomonas mucosa and how R. mucosa influences responses to the skin pathobiont Staphylococcus aureus, a common cause of skin infections that is also present in high numbers in atopic dermatitis lesions. The first step of any host-microbe interaction is adhesion, wherein the microbe attaches to the host cell thereby initiating responses in both host and microbe that are specific to that interaction. Our research has confirmed the interaction between R. mucosa and TLR5, first identified by Dr. Ian Myles, and shown that R. mucosa requires TLR5 for adhesion to skin both in vitro with human primary keratinocytes (KCs) and in vivo using TLR5 KO mice that showed significantly reduced adhesion of R. mucosa compared to wild-type mice. We have also found that R. mucosa is a potent activator of TLR2. It is noteworthy that despite binding to and/or activating these innate immune receptors, R. mucosa does not induce the strong pro-inflammatory response typically associated with pathogen-induced activation of TLRs. Subsequent to adhesion of a given microbe, human cells will initiate intracellular signaling in response to detection of said microbe. This is almost exclusively studied in the context of pathogenic microbes, however, the body has daily interactions with commensal microbes that also presumably elicit intracellular signaling that likely modulates human physiology. Again, we have discovered that after adhering to KCs, R. mucosa does not induce the pro-inflammatory signaling typically induced by pathogens. Specifically, we found that it does not activate p38 MAPK or ERK1/2 during incubation with KCs, but R. mucosa does induce canonical NFkB activation. However, the nature of this activation is distinct from its activation by pathogens like S. aureus. When NFkB is activated by R. mucosa, associated inhibitors of inflammation, namely TNFAIP3/A20 and IkBa, are also strongly activated. This appears to occur through a variation in the assembly of the Myddosome in response to TLR activation. Instead of the typical Myddosome assembly, which activates ERK1/2 and p38 MAPK along with NFkB, R. mucosa causes the Myddosome to incorporate IRAK3. Assembly of this version of the Myddosome is known to lead to anti-inflammatory signaling and shows how R. mucosa modulates anti-inflammatory signaling via TLR activation. Furthermore, as a result of interaction with R. mucosa, KCs in vitro and mouse skin in vivo were found to have significant changes in their expression of select pro-inflammatory cytokines and chemokines like IL-1b, IL-8, TNFa and CXCL10. These changes were not to the same degree as that observed in response to pathogens and may represent a low-level inflammation that can be protective in response to challenge with pathogenic bacteria; CXCL10 in particular is known to have direct antimicrobial activity on S. aureus. To test this hypothesis, we developed an in vitro model of infection with S. aureus of human KCs colonized with R. mucosa. Results from this model showed that R. mucosa protects human KCs from cell death induced by S. aureus, and blocks pro-inflammatory signaling (p38 MAPK and ERK1/2) induced by this pathogen while also enhancing the production of the AMP hBD-3. In this model, we also showed that R. mucosa alone does not cause cell death of KCs, and it does not out-grow S. aureus in this model. Evaluation of cytokine and chemokine responses also showed that R. mucosa modulates levels of certain pro-inflammatory cytokines in response to S. aureus; namely, R. mucosa increases CXCL10 and TNFa, while reducing IL-1b and IL-18. Finally, to more broadly understand the effects of R. mucosa on skin homeostasis, we obtained quantitative proteomics data from human KCs co-cultured overnight with R. mucosa. Drs. Aleksandra Nita-Lazar and Nathan Manes assisted with analysis of this data which showed that R. mucosa caused significant changes in expression to 30% of proteins in human KCs. This data helped highlight the proteins that mediate the anti-inflammatory effects of R. mucosa, such as IRAK3 and TNFAIP3/A20. Further elucidation of the mechanism by which R. mucosa activates these anti-inflammatory and protective pathways through TLR signaling requires identification of the ligand(s) expressed by this commensal bacteria. We have shown that R. mucosa activates both TLR2 and TLR5 using reporter cells. We are currently collaborating with Dr. Stephen Leppla and Rasem Fattah to perform bioassay guided fractionation of the outer membrane components of R. mucosa to identify these ligands. Fractionated bacterial components are added to reporter cells to identify which ones activate each receptor, these fractions are subjected to finer separation and active components are identified via further analysis using mass spectrometry. Understanding how R. mucosa provides therapeutic benefits through its activation of TLRs will provide valuable guidance for identifying other therapeutic microbes for safe effective treatment of prevalent allergic and inflammatory diseases.