Regulation of autoimmune myocarditis by host responses to the microbiome

Abstract

The extensive, diverse communities that constitute the microbiome are increasingly appreciated as important regulators of human health and disease through inflammatory, immune, and metabolic pathways. We sought to elucidate pathways by which microbiota contribute to inflammatory, autoimmune cardiac disease. We employed an animal model of experimental autoimmune myocarditis (EAM), which results in inflammatory and autoimmune pathophysiology and subsequent maladaptive cardiac remodeling and heart failure. Antibiotic dysbiosis protected mice from EAM and fibrotic cardiac dysfunction. Additionally, mice derived from different sources with different microbiome colonization profiles demonstrated variable susceptibility to disease. Unexpectedly, it did not track with segmented filamentous bacteria (SFB)-driven Th17 programming of CD4⁺ T cells in the steady-state gut. Instead, we found disease susceptibility to track with presence of type 3 innate lymphoid cells (ILC3s). Ablating ILCs by antibody depletion or genetic tools in adoptive transfer variants of the EAM model demonstrated that ILCs and microbiome profiles contributed to the induction of CCL20/CCR6-mediated inflammatory chemotaxis to the diseased heart. From these data, we conclude that sensing of the microbiome by ILCs is an important checkpoint in the development of inflammatory cardiac disease processes through their ability to elicit cardiotropic chemotaxis.

Keywords: Autoimmune disease; Chemokines; Innate lymphoid cells; Microbiome; Myocarditis; Segmented filamentous bacteria; Th17 cells.

Fig. 1. Polyantibiotic treatment of BALB/cJ mice limits EAM, and progression to DCMI

A) Cardiac pathology of BALB/cJ mice fed polyantibiotic cocktail or control water at day 21 of EAM. Representative cardiac histopathology micrographs depict median individuals of each group, stained by H & E. B) Relative heart weight, normalized to body weight (left); total viable CD45⁺ cells infiltrating the hearts of BALB/cJ mice, determined cytometrically (right) at day 21 of EAM. C) Determination of cardiac function by M-mode echocardiographic imaging of mice at day 21 of EAM. Calculated parameters portray cardiac function: ejection fraction (left) and fractional shortening (right). D) Cardiac expression of fibrotic genes determined by realtime qPCR from total mRNA of hearts at day 21 of EAM. Data are representative of five independent experiments. Diamonds and circles depict individual animals, bars represent arithmetic means of each group.

Fig. 2. BALB/cJ mice develop more severe EAM than BALB/cAnNTac animals

A) Cardiac histopathology at day 21 of EAM in BALB/cJ (blue, JAX) and BALB/cAnNTac (red, Tac) mice. Representative cardiac histopathology micrographs depict median individuals of each group, stained by H & E. B) Relative heart weight, normalized to body weight (left); total viable CD45⁺ cells infiltrating the hearts of BALB/cJ and BALB/cAnNTac mice, determined cytometrically (middle); and cardiac hydroxyproline content (right) at day 21 of EAM. Data are representative of three independent experiments. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. A Th17 signature is associated with microbiome-sensitive cardiac inflammation

A) Cardiac expression of inflammatory cytokines and chemokines, determined by Linco multiplex interrogation of cardiac homogenates. Data are normalized to wet starting weight of sample. B) Absolute numbers of RORγt⁺CD4⁺ Th17 cells in the hearts of BALB/cJ and BALB/cAnNTac animals at day 21 of EAM. Representative bivariate plots depict viable TCRβ⁺CD45⁺-gated lymphocytes from concatenates of each group. C) Absolute numbers of Ly6G^hiCD11b⁺ neutrophils and Ly6C^hiCD11b⁺Ly6G⁻ inflammatory monocytes in the hearts of BALB/cJ and BALB/cAnNTac mice. D) Absolute numbers of RORγt⁺CD4⁺ Th17 cells in the hearts of polyantibiotic or control BALB/cJ animals at day 21 of EAM. Representative bivariate plots depict viable CD3ε⁺CD45⁺-gated lymphocytes from concatenates of each group.

Fig. 4. Innate lymphoid cell (ILC) populations in the steady-state ileal lamina propria associate with susceptibility to EAM

A) Representative gating for innate lymphoid cell (ILC) populations from the lamina propria of the distal ilea in naïve, unimmunized BALB/cJ and BALB/cAnNTac mice; enumerated as T-bet⁺ ILC1, ST2/IL33R⁺ ILC2, and RORγt⁺ ILC3 subpopulations (right). Lineage cocktail includes CD3ε, CD11b, CD11c, CD19, CD45R/B220, CD49b/DX5, Gr1 (Ly6G/Ly6C), TCRγδ, and Ter119. Representative bivariate plots depict viable CD45⁺-gated lymphocytes from concatenates of each group. B) Realtime qPCR profiling of Th17/ILC3-associated transcripts from the distal lamina propria of naïve, unimmunized BALB/c mice. Data are represented as linearized fold difference values of BALB/cJ mice, relative to mean values from BALB/cAnNTac animals, normalized to Hprt. C) Expression of IL17A by viable CD45⁺-gated αβ T cells in the ileal lamina propria of naïve, unimmunized IL17A-GFP reporter mice or nontransgenic littermates treated with polyantibiotic or control water in representative bivariate plots of concatenates of (n = 3) groups. D) RORγt⁺ ILC3s in the ileal lamina propria of naïve, unimmunized polyantibiotic-treated or control BALB/cJ animals, expressed as a proportion of Lineage⁻CD90⁺IL7Rα⁺-gated ILCs. Representative bivariate plots depict viable ILC-gated lymphocytes from concatenates of each group.

Fig. 5. ILCs are critical for the recruitment of leukocytes to the inflamed heart

A) Schematic representation of depletion of ILCs from active AT-EAM recipients by allo-mAbs to CD90.2 (Thy1.2), clone 30H12. B, E) Representative bivariate plots depict donor and recipient-derived viable CD11b⁻CD45⁺-gated concatenates of each group from A) mAb depletion of ILCs or D) genetically defective development of ILCs at day 14 post-transfer. C, F) Enumerative cardiac cytometry of total CD45⁺ leukocytes, Ly6G^hiCD11b⁺ neutrophils, and CD90.1⁺CD4⁺TCRβ⁺ T cells from A) mAb depletion of ILCs or D) genetically defective development of ILCs at day 14 post transfer. D) Schematic representation of deletion of ILCs from RAG2^−/−γc^−/− recipients in active AT-EAM. Data are representative of two independent experiments.

Fig. 6. Microbiome profiles and ILCs regulate expression of CCL20/MIP3α in the ilea at steady-state, and in the inflamed heart

A) CCL20/MIP3α expression in the distal ilea of polyantibiotic-treated or control BALB/cJ mice at day 9 of EAM, as interrogated by realtime qPCR. Data are represented as fold change, relative to control animals, normalized to Hprt housekeeper. B) Proportional distribution of CCL20/MIP3α⁺ populations in the hearts of untreated BALB/cJ mice at day 21 of EAM, as determined by backcalculated values from cytometry in C–D. Slices indicate arithmetic mean of each population across the group. C) CCL20/MIP3α expression in viable CD31⁺CD45⁻ cardiac endothelial cells in dys-biotic or control BALB/cJ mice over the course of EAM. Representative histogram overlays depict concatenates of each group. D) CCL20/MIP3α expression, as a proportion of: CD64⁺CD11b⁺CD45⁺ macrophages, CD11c^hiCD45⁺ dendritic cells, and gp38/podoplanin-1⁺ and CD140a/PDGF-Rα⁺ fibroblasts at day 21 of EAM in dysbiotic or control BALB/cJ mice. E–F) CCL20/MIP3α expression, determined by quantitative sandwich ELISA interrogation of cardiac homogenates from E) α-CD90.2 mAb parking AT-EAM or F) RAG2^−/−γc^−/−ILC-deficient active AT-EAM recipients. Data are normalized to wet starting weight of sample.

Fig. 7. Microbiome profiles and ILCs are regulators of CCR6-driven recruitment of Th17-associated lymphocytes to the inflamed heart

A) Cytometric enumeration of RORγt-expressing γδ T cells and Lineage⁻CD90⁺IL7Rα⁺ ILC3s in the hearts of polyantibiotic-treated and control BALB/cJ mice at day 21 of EAM. Representative bivariate plots depict viable CD45⁺ lymphoid populations from concatenates of each group. B) CCR6 expression on cardiac RORγt⁺CD4⁺ Th17 cells in polyantibiotic-treated and control BALB/cJ mice at day 21 of EAM. Representative bivariate plots depict viable CD4⁺CD3ε⁺CD45⁺-gated lymphocyte populations from concatenates of each group. C) CCR6 expression on CD4⁺ and γδ T cells in the hearts of BALB/cJ (blue) and BALB/cAnNTac (red) mice at day 21 of EAM. Representative bivariate plots depict viable CD45⁺-gated lymphocyte subpopulations from concatenates of each group. D) CCR6 and RORγt expression on cardiac CD4⁺ T cells in ILC-deficient RAG2^−/−γc^−/− and control RAG2^−/− active AT-EAM recipients at day 14 post-transfer. Representative bivariate plots depict viable CD4⁺TCRβ⁺CD45⁺-gated T cells from concatenates each group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)