Antigen-driven T cell-macrophage interactions mediate the interface between innate and adaptive immunity in histidyl-tRNA synthetase-induced myositis

Figures

• 

  Figure 1. Role of T cells in HRS-induced myositis. (A, B) demonstrate H&E- and immunohistochemically-stained muscle tissue derived from n=4 C57BL/6 (B6) WT (A) and n=4 RAG1 KO (B) mice 17 days post immunization with recombinant HRS. Cell surface markers of infiltrating mononuclear cells (CD45), T cells (CD3, CD4, CD8), and macrophages (CD68) are indicated in the upper right hand corner of individual tissue sections. Scale bars (400 μm) are shown in the lower right hand corner of each photomicrograph. Box plots in panel (C) demonstrate the relative severity of muscle inflammation/cellular infiltration in B6 WT versus RAG1 KO mice following immunization with recombinant HRS. While horizontal bars designate median severity scores (median (IQR): 2.25 (2, 2.5) B6 WT vs. 0.75 (0.125, 1.375) Rag 1 KO), “x” symbols represent mean levels of inflammation (2.25 B6 WT vs. 0.75 RAG 1 KO). Minimum and maximum severity scores are indicated by whisker edges (min/max=2/2.5 B6 WT vs. 0/1.5 RAG1 KO); p-value determined by Mann-Whitney U-test.

• 

  Figure 2. Requirement for T cell-specific MyD88 signaling in HRS-induced myositis. While the box plots in (A) demonstrate the relative severity of HRS-induced myositis in B6 WT versus CD4-Cre.MyD88fl/fl conditional KO mice (n=5 mice/group) (median severity score (IQR)= 2.5 (1.75, 2.75) B6 WT vs. 1 (0.75, 1.25) CD4-Cre.MyD88fl/fl; min/max=1.5/3 B6 WT vs. 0.5/1.5 CD4-Cre.MyD88fl/fl; p-value determined by Mann-Whitney U-test), (B, C) depict immunohistochemical staining of muscle-infiltrating cell populations with antibodies recognizing CD45, CD3, CD4, and CD68. (D) demonstrates representative staining of muscle tissue derived from HRS-immunized RAG1 KO mice following adoptive transfer of unfractionated splenocyte populations from B6 WT versus CD4-Cre.MyD88fl/fl conditional KO mice. Scale bars (400 μm) are shown in the lower right hand corner of each photomicrograph.

• 

  Figure 3. Contribution of antigen-specific TCR signaling to HRS-induced myositis. (A) demonstrates the relative severity of muscle inflammation in n=5 B6 WT versus n=5 OT-II (Ovalbumin-specific TCR transgenic) mice 17 days post-immunization with recombinant HRS. Scale bars (400 μm) are shown in the lower right hand corner of each photomicrograph. Box plots represent the interquartile range of muscle severity scores and include median (horizontal bars) as well as mean (“x” symbol) values; median (IQR)=2 (2, 2.5) B6 WT vs. 0.5 (0.25, 1.25) OT-II; mean=2.2 B6 WT vs. 0.7 OT-II; min/max=2/2.5 B6 WT vs. 0/2 OT-II; p-value determined by Mann-Whitney U-test. Dot plots and summary tables in panels B-C illustrate flow cytometric profiles of pooled muscle-infiltrating (B) and spleen (C) cell populations derived from n=4 B6 WT and n=3 OT II TCR transgenic mice following HRS-mediated induction of myositis. While CD19+ B cells were gated as a percentage of CD45+CD3- populations, CD161+ NK cells and F4-80+ macrophages were quantified from CD3-CD19- and CD3-CD19-CD161- cells, respectively. Given the similar number of live cells isolated from muscle or spleen for each experimental condition, percentages of…

• 

  Figure 4. Dependence of HRS-induced myositis on myeloid/macrophage-specific MyD88 signaling. (A) depicts high (100x) power views of representative H&E-stained muscle tissue from Lyz2-Cre.MyD88fl/fl conditional KO mice versus B6.MyD88fl/fl (Cre-) littermate controls 17 days post immunization with recombinant HRS. Accompanying immunohistochemical stains demonstrate the distribution of CD4+ T cells, CD45+ leukocytes, and CD68+ macrophages in muscle-infiltrating cell populations. Scale bars (400 μm) are shown in the lower right hand corner of each photomicrograph. The box plots in (B) demonstrate the relative severity of muscle inflammation resulting from HRS immunization of n=5 B6.MyD88fl/fl littermate control and n=7 Lyz2-Cre.MyD88fl/fl conditional KO mice. While horizontal bars designate median severity scores, “x” symbols represent mean levels of inflammation; median (IQR)=2.5 (2.25, 3) B6.MyD88fl/fl vs. 1.5 (1, 2.5) Lyz2-Cre.MyD88fl/fl; mean=2.6 B6.MyD88fl/fl vs. 1.6 Lyz2-Cre.MyD88fl/fl; min/max=2/3 B6.MyD88fl/fl vs. 1/2.5 Lyz2-Cre.MyD88fl/fl; p-value determined by Mann-Whitney U-test.

• 

  Figure 5. T cell activation signals and their impact on macrophage phenotype in HRS-induced myositis. The UMAP plot in (A) represents composite single cell RNA sequencing (scRNAseq)-based clustering of both infiltrating and resident cell populations isolated from muscle tissue of B6 WT mice following immunization with recombinant HRS versus PBS. The UMAP plots and accompanying bar graphs in (B–D) show the relative distribution of macrophage and fibroblast subsets following HRS-mediated induction of myositis in B6 WT versus RAG1 KO mice (B), B6 WT versus OT-II TCR transgenic mice (C), and CD4-Cre.MyD88fl/fl conditional KO versus B6.MyD88fl/fl (Cre-) littermate control mice (D). Each of these comparative analyses represents scRNAseq profiling of muscle tissue from pools of n=3-5 mice; PBS controls have been omitted from the bar graphs of (B–D) due to limited populations of infiltrating macrophages/fibroblasts in these mice and overall ease of presentation. Additional clusters depicted in panel A include B cells/plasma cells (clusters 25/32), neutrophils (cluster 41), endothelial cells (clusters 0,3,9,12,24,26), and muscle satellite cells (cluster 4).

• 

  Figure 6. Association between T cell phenotype and differential activation of macrophage as well as fibroblast signaling pathways in HRS-induced myositis. Venn diagrams in panels (A, B) illustrate differential pathway induction in muscle-infiltrating macrophage and fibroblast populations (determined by Gene Ontology analysis of differentially expressed genes) following immunization of the indicated strains with recombinant HRS. Specific pathways demonstrating more than 5-fold enrichment in these inter-strain comparisons (WT vs. OT-II TCR transgenic, CD4-Cre.MyD88fl/fl conditional KO, and/or RAG1 KO) are enumerated in overlapping segments of the Venn diagrams and summarized in Supplementary Tables 3, 4.

• 

  Figure 7. NicheNet analysis of ligand-receptor interactions in HRS-induced myositis. (A–C) depict predicted ligands responsible for engagement between T cells, macrophages, and fibroblasts based on differential expression of target genes in macrophages and fibroblasts. While the accompanying heat maps correspond to fold change in ligand gene expression for “sender” populations in the indicated strains (condition of interest versus reference condition), dot plots represent the relative expression level (color) and percentage of specified cells (size) expressing genes encoding these ligands. Predicted ligands are ranked based on Pearson correlation coefficients.