Determining the Prognostic Significance of Alternative Splicing Events in Soft Tissue Sarcoma Using Data From The Cancer Genome Atlas

Abstract

Background: Surgery, adjuvant chemotherapy, and radiotherapy remain the primary treatment options for soft tissue sarcomas (STSs). Identifying ways to improve the prognosis of patients with STS is a considerable challenge. Evidence shows that the dysregulation of alternative splicing events is involved in tumor pathogenesis and progression. The present study objectives were to identify survival-associated alternative splicing (AS) events that could serve as prognostic biomarkers and potentially serve as tumor-selective STS drug targets. Methods: STS-specific percent spliced in (PSI) values for splice events in 206 STS samples were downloaded from The Cancer Genome Atlas (TCGA) SpliceSeq®. Prognostic analysis was performed on seven types of AS events to determine their prognostic value in STS patients, for which prediction models were constructed with the formula Risk score = n ∑ PSIi*βi i To explore how the AS events function, mRNA level, protein level, copy number alteration, mutation, and methylation of several STS key genes were detected from TCGA RNA-seq data and relevant protein data, as well as Gene Expression Omnibus gene chip data. Additionally, Spearman’s rank correlation coefficients were calculated to evaluate any correlation between splicing factor expression and the PSI values of survival-associated AS events. Findings: Of the 40,184 AS events for 3,064 genes, 10,439 events were found to significantly correlate with patient survival rates. The area under the time-dependent receiver operating characteristic curve for the prognostic predictor of STS 2-year overall survival was 0·826. Notably, the splicing events of certain STS key genes were significantly associated with STS 2-year overall survival in the present study, including exon skip (ES) events in MDM2 and EWSR1 and alternate terminator events in CDKN2A and HMGA2 for dedifferentiated liposarcoma; ES in MDM2 and alternate promoter events in CDKN2A for leiomyosarcoma; and ES in EWSR1 for undifferentiated pleomorphic sarcoma. Moreover, splicing correlation networks between AS events and splicing factors revealed that almost all of the AS events associated with favorable prognosis were negatively correlated with the expression of splicing factors. Likewise, most of the AS events associated with poor prognosis were positively correlated with the expression of splicing factors. Interpretation: In-depth analysis of alternative RNA splicing could provide new insights into the mechanisms of STS oncogenesis, with the potential for novel avenues for this type of cancer therapy. Funding Statement: The Innovation Project of Guangxi Graduate Education (YCBZ2018038), Medical Excellence Award Funded by the Creative Research Development Grant from the First Affiliated Hospital of Guangxi Medical University, Guangxi Medical University Training Program for Distinguished Young Scholars, and Innovation Project of Guangxi Graduate Education (YCBZ20190xx). Declaration of Interests: The authors declare that they have no conflicts of interest.