RNA-Sequencing Reveals a Distinct Transcriptomic Signature for Giant Cell Myocarditis and Identifies Novel Druggable Targets

Giant cell myocarditis (GCM) is a rare inflammatory heart muscle disease that is often rapidly progressive and carries a poor prognosis. The pathogenesis of GCM is incompletely understood, but is thought to be autoimmune in nature.¹ Histopathology of GCM is characterized by disproportionate myocyte necrosis and a myocardial infiltrate consisting of T-lymphocytes, macrophages, plasma cells, neutrophils, eosinophils, and multinucleated giant cells, the latter of which are often considered pathognomonic for this disease.² While limited data suggest a role for empiric immunosuppression to treat GCM, transplant-free survival remains poor using currently accepted therapies.¹ An improved mechanistic understanding of GCM is critical for identifying novel therapeutic targets and improving survival.

Ribonucleic acid sequencing (RNA-seq) allows for an unbiased analysis of the entire transcriptome. We sought to characterize the transcriptomic profile of histopathologically-proven GCM and compare it to that of healthy controls and two forms of lymphocytic myocarditis: 1) lymphocytic myocarditis due to viral or idiopathic causes and 2) acute cellular rejection (ACR) after heart transplantation. We hypothesized that GCM exhibits a distinct transcriptomic profile and, by querying a known database of drug-gene interactions, novel drug targets could be identified.

We performed RNA-seq utilizing formalin-fixed paraffin-embedded (FFPE) tissue from the Vanderbilt University Medical Center. This study was approved by the Institutional Review Board #201926. Out of 18 samples, four with GCM, four with grade 2R ACR, three with lymphocytic myocarditis, and three unused healthy donor heart samples were used in the final analysis; four samples failed quality control (QC) and were excluded. Specimens were obtained during endomyocardial biopsy or after native heart explant. Samples were obtained from the right ventricle and selected for regions with inflammation. RNA was isolated using the Maxwell 16 LEV RNA FFPE Kit (AS1260, Promega). Library preparation utilized a ribo-depletion total RNA library preparation kit and 150 bp paired-end sequencing was performed on the Illumina NovaSeq 6000 targeting an average of 50M reads per sample. QC was evaluated at different levels, including RNA quality, raw read data, alignment, and gene expression. Raw RNA-seq paired-ends were mapped to the human reference genome hg19 using STAR 2.7.3. Raw reads count matrix was calculated by featureCounts and used for downstream analysis. Differential gene expression and functional enrichment analyses were performed in R (version 3.6) with several Bioconductor packages, including DESeq2 and clusterProfiler, using a 5% false discovery rate threshold for significant results filtering

Demographics and clinical characteristics of the groups are highlighted in Panel A. 5,297 genes were differentially expressed in GCM in comparison to the other three groups. Principal component analysis (Panel B) based on complete expression profiles and unsupervised hierarchical cluster analysis of significant differentially expressed genes identified a distinct transcriptomic profile for GCM (Panel C). Consistent with shared histological features, overlap is noted between rejection and lymphocytic myocarditis (total differentially expressed genes between rejection and lymphocytic myocarditis: 213) Pathway enrichment analysis (Panel D) in GCM showed that upregulated pathways were enriched for neutrophil degranulation, multiple cytokine signaling pathways, and phagocytosis, while downregulated pathways included those involved in muscle contraction and cardiac conduction. The most significantly upregulated genes included RNASET2, ITGAX, TNFAIP2, TNFRSF14, PTPN7, and HLA-DRA, all of which play roles in innate and adaptive immune responses, while the most significantly downregulated genes included MYL2, CAMK2B, MLIP, and AKAP6, which play roles in cardiac contraction, conduction, and homeostasis (Panel E). Genes involved in apoptosis were upregulated while those involved in oxidative phosphorylation were downregulated in comparison to the other groups. Querying the Drug Gene Interaction Database (DGIdb) identified several novel biologically-plausible therapeutic targets (Panel F). For example, IL-1β antagonism with anakinra is currently being studied in patients with acute myocarditis (NCT03018834). Abatacept, a CTLA4-Ig fusion protein that targets B7, has been shown to be effective in a case of acute myocarditis due to immune-checkpoint inhibitor therapy³ and is currently being studied as a potential treatment for myocarditis (NCT03619876).

Two studies previously evaluated GCM with microarray and quantitative RT-PCR,^{4, 5} but to our knowledge, this is the first study using RNA-seq to interrogate the complete transcriptomic profile of GCM and highlights several important findings. First, our work shows the potential for using FFPE human heart samples to perform RNA-seq. Second, our results demonstrate that GCM has a gene expression profile that mirrors its clinical and histopathological features, and is distinct from the transcriptomic profiles seen in other healthy and inflammatory cardiac diseases. Downregulation of pathways involved in muscle contraction, ion homeostasis, and cardiac conduction help explain the typical patient presentation with acute heart failure and arrhythmias. Third, by querying a drug-gene interaction database, several novel drug targets were identified. While our results are limited by small sample size, they nonetheless lay the foundation for larger-scale transcriptomic evaluation of stored FFPE tissue and intersecting the results with the druggable genome to identify potential new therapies.

Figure 1.

A) Demographics and clinical characteristics. B) Principal component analysis (PCA) plot based on complete expression profiles and C) unsupervised hierarchical clustering of differentially expressed genes associated with giant cell myocarditis (GCM) reveal a distinct profile associated with GCM. D) Reactome pathway enrichment for up- and down-regulated genes of the three disease groups compared to control. E) Volcano plot of most upregulated and downregulated genes (by P.adj). F) Partial list of drug-gene interactions to identify potential novel drug targets. The IL1B drug-gene interaction is from the GCM vs lymphocytic myocarditis comparison, while all others are from the GCM vs control comparison. The full list can be found on https://prod.tbilab.org/gc_myocarditis. LVEF = left ventricular ejection fraction; HTN = hypertension; CKD = chronic kidney disease; EMBx = endomyocardial biopsy; P.adj = P-value adjusted for multiple comparisons by the Bonferroni-Hochberg method; GC = giant cell myocarditis; lym = lymphocytic myocarditis; rej = acute cellular rejection; con = control (unused donor heart tissue).

Data Availability.

Data and code are available at https://prod.tbilab.org/gc_myocarditis.