BIOM-35. COMPREHENSIVE IDENTIFICATION OF MRNA ISOFORMS AND THEIR DIAGNOSTIC POTENTIAL IN GLIOMAS

Abstract Most genes generate multiple mRNA isoforms which may differ in their protein-coding capacity. Also, cancer and its prognosis are associated with the altered expression of specific isoforms of a variety of genes as cancer cells can manipulate regulatory mechanisms to express isoforms that provide a survival advantage and/or resistance to therapy. However, the underlying mechanisms behind their pathology are poorly understood. Fifty seven glioma samples that could be categorized unambiguously into Astrocytoma (n=17), Oligodendroglioma (n=16) and Glioblastoma (n=24) based on pathology reports and according to WHO standards were selected for this analysis. Total RNA-Sequencing (RNA-Seq) was performed on RNA isolated from these 57 samples. RNA-Seq data analysis was performed using the DRAGEN RNA application in Illumina BaseSpace and transcript counts were generated after alignment. From these counts, isoforms that were significantly differentially expressed in Astrocytoma (Astro) vs Oligodendroglioma (Oligo), and Astrocytoma vs Glioblastoma (GBM) were identified using the R package, IsoformSwitchAnalyzeR. Approximately 93,000 transcripts were screened in both comparisons of which 4388 and 44 transcript switches were identified as significant at FDR<0.05 in the Astro vs GBM and Astro vs Oligo comparisons, respectively. Applying an additional cut-off based on the difference in isoform fraction or dIF of a transcript between the two groups being compared (dIF>1.0 and dIF<-0.1), the most significant top switches were identified.. Similarly, in the Astro vs GBM comparison, the isoform switch in genes such as MDF1 and RER1 whose overexpression is known to be involved in tumorigenesis and metastasis were among the 155 top switches. Among the six top switches in the Astro vs. Oligo comparison were transcripts of the RENBP and GCNT2 genes which play an important role in the progression of melanoma, gastric and lung cancers. These results indicate that there may be isoform-specific dysregulation within gliomas, and further suggest that their biological significance and diagnostic potential can be explored.