Cardiac autoantibodies promote a fibrotic transcriptome and reduced ventricular recovery in human myocarditis

Figures

• 

  Figure 1. Baseline autoantibodies to CM peptides are significantly elevated in non-recovered (NR) myocarditis patients compared with recovered (R) patients. Sera taken at initial clinical presentation from recovered and non-recovered myocarditis were reacted with whole CM molecule or 32 overlapping peptides of CM in ELISA. CM titers strongly trended towards significance when all patients and healthy subjects were compared (A) Non-recovered patients had significantly elevated titers to CM compared to healthy controls (B). In (C), non-recovered (n=5) myocarditis patients have significantly elevated autoantibodies to CM peptides when compared with recovered (n=28) patients and healthy subjects (n=27). In (C), each datapoint is the average serum IgG immunoreactivity (OD @ 1:100 serum dilution) with all 32 S2 CM fragment peptides for each individual (i.e. each black dot is the average from all patients for one peptide). Serum IgG titers to CM correlated with lower ejection fraction in non-recovered patients, but not in recovered patients [(D),s Supplementary Figure 2 ] NR, Non-recovered; R, Recovered.

• 

  Figure 2. Autoantibodies to CM peptides are elevated in non-recovered patients and AAbs recognize unique CM peptide epitopes in non-recovered patients. Overlapping peptides synthesized from the S2 region of CM were tested by ELISA for reactivity with serum IgG in healthy (n=30, panel B), recovered (n=28, panel C), and non-recovered patients (n=5, panel D). Recovered and non-recovered are shown in the top panel (A), dotted black line indicates mean healthy subject (n=30) serum IgG antibody reactivity at mean OD of 0.2809 ± 0.03. OD values normalized to healthy are plotted in (B-D). One-way ANOVA showed groupwise differences between healthy compared with recovered (panel C, p <0.0001) and with non-recovered (panel D, p <0.001). Post-hoc analysis showed that in recovered patients compared with healthy control serum IgG, peptides S2-3 (p <0.01), S2-9 (p <0.01), and S2-30 (p <0.0001) reached significance after multiple test correction. IgG AAbs to CM peptides specific to non-recovered patient sera were S2-1 (p <0.001), S2-3 (p <0.01), S2-17 (p <0.001), S2-25 and S2-29 (both p <0.01) reached significance above healthy subjects. ANOVA showed groupwise differences between recovered and…

• 

  Figure 3. Myocarditis/DCM sera signal PKA H9c2 cells treated with myocarditis/DCM sera (1:100 dilution). (A) Levels of PKA signaling in H9c2 cells at various timepoints are compared to healthy subject levels. (B) PKA signaling is significantly elevated in both recovery and non-recovery at the baseline time point. Recovered is shown in One-way ANOVA compared with healthy sera with Tukey’s multiple comparison test ****p<0.0001, non-recovered vs healthy sera: **p=0.0023. Although myocarditis sera strongly signal PKA, there is no difference in the recovered vs non-recovered.

• 

  Figure 4. Myocarditis mAb 2C.4 and sera promote a fibrotic transcriptome in H9c2 cells. (A) DE genes in mAb/sera vs. basal after treatments of H9c2 cells. Log2FC values DE genes show high concordance (same direction) in mAb/sera treated cells: DE genes in upper-right (177) vs lower left (61) quadrants. (B) KEGG Analysis of DE genes: mAb compared to basal treated H9C2 cells. (C) IPA of DE genes in myocarditis vs healthy (n=2) serum treated cells shows enriched fibrosis/hypoxia/apoptosis pathways. Red positive z score; blue negative z score indicates DE direction. (D) IPA of DE fibrosis related genes elevated in myocarditis serum treated cells (upper left) but not elevated in cells after treatment with healthy serum (lower right) p <0.05. (E) By qPCR, fibrosis/heart failure genes were upregulated in cells after treatment with myocarditis (PT, (n=6)) serum (CAMTA2, ITGB6, NFATC2, and XDH) vs healthy control (HC) sera (HC, n=4) treated H9C2 cells. Data analyzed with ΔΔCT method; β-actin reference gene. A gene was considered DE if it demonstrated Log2-fold change > or < 1 (red and blue, respectively).

• 

  Figure 5. Myocarditis-derived mAb 2C.4 induces fibrosis-related pathways in heart cells. Heart (H9c2) cells treated with human mAb 2C.4 or β-adrenergic agonist isoproterenol (ISO) underwent RNA sequencing, then pathway analysis (GSEA). Functional units (FU) visualized with emapplot() from enrichplot (top panel, left 2C.4, right ISO) identified fibrosis, respiration, and inflammation FUs from significantly upregulated pathways (Circled and labeled "A", "B", and "C". Redundant pathways were ruled out with leading edge analyses (Panels A and B, correspond to FU A: fibrosis and B: respiration). No shared pathways were observed in FU in C containing inflammatory pathways, therefore, (C) was not considered for further analysis. Thus, C is not shown below separately. (A, B) are shown separately to compare the shared fibrosis and respiration pathways in the H9c2 cells. Similarities between responses to mAb 2C4 vs isoproterenol treatment of the H9c2 cells are compared. (A) Unique and shared pathways between ISO and 2C.4 are shown in Venn-diagram (inset) and enrichment scores for shared fibrosis pathways compared between conditions (Lower Panel A: arrow). GSEA plots (Far Right Panels following…

• 

  Figure 6. Model of pathogenic mechanisms in human autoimmune myocarditis. Lytic viral infection of cardiomyocytes or other causes of cell damage releases cardiac myosin (CM) which acts as a danger signal via TLR2/8, causing cytokine secretion that can be suppressed with anti-human TLR2 blocking antibody in vitro. In human myocarditis, cytokine secretion responses from CD14+ monocytes are exaggerated in myocarditis in comparison to healthy controls and promote Th17 differentiation. Antibodies to CM are also produced, and agonize βARs on cardiomyocytes or cardiac fibroblasts directly inducing a pro-fibrotic response in the myocardium. Created with BioRender.com and modified from Myers 2016 (6).