A Detailed Alternate Splicing Landscape in Multiple Myeloma with Significant Potential Biological and Clinical Implications

Alternative splicing (AS) is a critical post-transcriptional event, which affects number of cellular processes including cell growth and survival. Our group has previously shown that aberrant splicing of numerous genes in multiple myeloma (MM) affects the disease biology. Dysregulated expression or mutations in several splicing factors (SFs), including SF3B1, Fox2, SRSF1, NONO, in MM compared to normal plasma cells have been reported with impact on outcome, highlighting significant role of splicing factors in myeloma. We have previously reported global alternative splicing events in MM to understand biology. Here, we report, using a deep RNAseq data, a detailed analysis of the splicing events, including retained intron, cassette exon, Alternative 3' / 5' site, mutually exclusive exons, mutually alternative 3' / 5' exons and analyze significance of AS in predicting prognosis. We purified myeloma cells from 420 newly-diagnosed patients treated homogeneously with lenalidomide, bortezomib and dexamethsone (RVD) with or without high-dose melphalan followed by lenalidomide maintenance in the IFM/DFCI 2009 study. RNA isolated from these purified CD138+ MM cells and from 18 normal donor plasma cells were processed by RNAseq (80-100 million paired end reads on Illumina HiSeq) and analyzed using a custom computational and statistical pipeline.

Our results showed 1534 genes with one or more splicing events observed in at least 10% or more patients. Median alternative splicing event per patient was 595 (range 223 - 2735). Although we observed 1288 (~84%) of these 1534 genes with only one type of AS, remaining 16% of the genes had 2 or more AS events. Within the AS types, we observed retained intron (495 - ~32%) and cassette exons (727 - ~47%) as the most frequent events, whereas mutually exclusive exons were observed in less than 1% of 1534 genes. We found that 158 genes had same type of AS in more than 50% of the patients and 445 genes showed same AS in more than 30% of patients suggesting significant biological role for the spliced isoforms. Importantly, we have evaluated impact of each AS event with clinical outcome and observed that 43 AS events showed significant correlation with event free survival (EFS) (p value < 0.01), 27 AS events predicted overall survival (OS) and 5 AS events predicting both EFS and OS. We have also observed that although overall expression of number of these genes does not predict overall survival outcome, an isoform level expression predicts survival. We are now developing a composite model using gene expression and alternative splicing events that in our preliminary analysis has provided increased prediction power compared to gene expression model only which shows that alternative splicing events may carry independent significance for outcome prediction over overall gene expression alone.

In summary, this study for the first time reports a detailed splicing landscape in myeloma and highlights the biological and clinical importance of alternative splicing events. It suggests the need to further integrate study of spliced isoform in our biological studies, in understanding disease behaviors as well as develop targeted therapeutics.