Comprehensive Analysis Of Alternative Splicing Events Involved InProneural-Mesenchymal Transition Of Glioblastoma

Objective:Alternative splicing helps to create complex phenotypes in higher eukaryotes by increasing the diversity of transcriptomes and proteomes,and participates in various biological evolution processes.Once deregulated,it can lead to disease occurrence and transformation.The main features of glioblastoma(GBM)are high malignancy,treatment resistance,easy postoperative recurrence,and high mortality.Previous studies have shown that intratumoral heterogeneity of GBM plays a major role,and genome-wide transcriptome analysis determined Four subtypes have been identified:proneural(PN),classical(CL),mesenchymal(MES),neuronal(NE),and GBM can shift from one subtype to another during disease progression.A subtype in which the proneural-mesenchymal transition(PMT)is most prominent.This study aimed to systematically analyze the functional role and regulatory mechanism of alternative splicing in the proneural-mesenchymal transition based on high-throughput sequencing and its application in transcriptomic research,as well as potential impact on the prognosis of GBM patients.Methods:49 cases of mesenchymal GBM and 38 cases of proneural GBM with complete clinical information and transcriptome data in the TCGA database were selected,and the "limma" R software package was used to determine differentially expressed splicing events.The differentially expressed splicing events related to prognosis were obtained by univariate Cox.LASSO regression and multivariate Cox regression analysis were used to construct prognostic risk model and performance evaluation.Based on Pearson correlation analysis and GSVA analysis,the regulatory network of splicing factor-alternative splicing event-KEGG signaling pathway was established.Unsupervised clustering analysis was used to verify PMT-related splicing events,the ssGSEA algorithm was used to obtain PMT splicing scores to further explore the relationship between PMT splicing scores and splicing-related tumor subtypes with immune infiltration and treatment responsiveness.Results:By integrated analysis of alternative splicing events in PN and MES GBM,we obtained 1342 differentially expressed alternative splicing events,of which the most significant changes were in the alternative promoter events,and the corresponding functional analysis explained neuronal differentiation,cell morphological changes and adhesion that were related to subtype transformation processes.By evaluating the prognostic association of alternative splicing events with GBM patients,we defined 75 prognosis-relevant differentially alternative splicing events as PMT-related splicing events,and constructed an independent prognostic risk model consisting of 6 splicing events(LRRC29,HDAC9,RPE,ERC1,PLCD1,FAM13A).Among them,high-risk GBM was significantly concentrated in MES subtype,while low-risk GBM was concentrated in PN subtype,and the survival analysis of different risk groups had significant differences.We further explored the abnormal changes of splicing factors,among which the expression of some PN splicing factors markers,including CELF and ELAVL family,decreased in MES GBM,while the expression of MES splicing factors markers RBM46/47 was increased.By mapping the PMT-related splicing regulatory network,it was found that the splicing factor RBM47/NOVA1 regulates the exon 20 skipping event of PPFIBP1 and participates in the signaling pathway of ECM receptor interaction.Since PMT-related splicing events can accurately and efficiently assess GBM isoform status,we calculated the PMT splicing scores of GBM and found that high PMT splicing scores and MES splicing type have significant in immune infiltration,containing a variety of immunosuppressive cells,expressing immune checkpoint-related proteins,and showing efficacy against immunosuppressive therapy.Conclusion:Based on the above analysis results,the prognostic model we constructed can quantitatively assess the prognostic risk of patients at the individual level.By drawing the PMT-related splicing network map,we found that the exon 20 skipping event of PPFIBP1 is a dangerous cancer-related alternative splicing event.The MES splicing type identified based on PMT-related splicing events are significantly associated with immune infiltration.In-depth study of the splicing regulatory mechanism will help to reveal the mechanism of GBM mesenchymal transition.And provide a new theoretical basis for tumor prognosis diagnosis and treatment.