Abstract
OBJECTIVE: The pathogenesis of inflammatory myopathies is poorly understood and there is a need to dissect the transcriptome in more granular ways beyond gene expression. METHODS: We used a set of muscle RNA-sequencing data from different myositis subtypes grouped by their specific autoantibodies (n = 152). We quantified annotated RNA transcripts for each myositis subtype and identified uniquely expressed RNA as well as transcriptional similarities among myositis types. In addition, we quantified event-based alternative splicing with predicted protein changes. And finally, we searched for cryptic exons. RESULTS: We saw considerable overlap in RNA expression among subtypes. In addition, MADCAM1 was previously shown to be uniquely expressed in Mi-2 myositis; we discovered it was two noncanonical transcripts that predominantly contributed to the observed increased expression. At the transcriptional level, dermatomyositis subtypes were least similar to inclusion body myositis (IBM) or Jo1, followed by HMGCR, then SRP and other dermatomyositis subtype. Additionally, we discovered many alternative splicing events that were unique by myositis subgroup, including events in muscle dystrophy genes and one event in SRP72, which was seen uniquely in SRP myositis. Finally, we looked for previously reported cryptic exons in IBM and did not find them. CONCLUSION: The large degree of transcriptional overlap among myositis subtypes reinforces the need to use disease (in addition to healthy) controls to find unique features of autoimmune disease. Unique alterations in the transcriptome that are seen in one myositis subtype and not others advance our understanding of distinct disease pathology.