It is increasingly appreciated that alternative splicing plays a key role in generating functional specificity and diversity in cancer. However, the mechanisms by which cancer mutations perturb splicing remain unknown. Here, we developed a network-based strategy, DrAS-Net, to investigate over 2.
We identified over 40, driver variant candidates and their 80, putative splicing targets deregulated in 33 cancer types and inferred their functional impact. Strikingly, tumors with splicing perturbations show reduced expression of immune system-related genes, and increased expression of cell proliferation markers.
Tumors harboring different mutations in the same gene often exhibit distinct splicing perturbations. Further stratification of 10, patients based on their mutation-splicing relationships identifies subtypes with distinct clinical features, including survival rates. Our work reveals how single nucleotide changes can alter the repertoires of splicing isoforms, providing insights into oncogenic mechanisms for precision medicine.
Distinct target AS profiles help explain cancer heterogeneity and classify cancer patients into subtypes with distinct clinical features. Phenotypic variation and heterogeneity is far more complex in human compared to other species, even though there are similar numbers of genes in the genome.
This enigma could be at least partially addressed by studying the extent to which different protein isoforms can be encoded by each genome. It has been increasingly appreciated that alternative splicing is a key factor contributing to protein isoform diversity.
In human cancer, for instance, the problem of tumor heterogeneity across patient populations is known to involve alternative splicing. However, the fundamental question of how genomic mutations influence the splicing process leading to cancer is essentially unknown Figure 1A.
A Alternative splicing AS underlies the complexity of genotype-phenotype relationships. B Flowchart of the mutation-mediated alternative splicing AS analysis in cancer. Genome-wide mutational profiles of 10, samples and AS data from 10, samples across 33 types of cancer are integrated into functional networks.
Four types of analyses are shown: I Identification of genome-wide AS alternations in each type of cancer. Differential AS events are identified as cancer-specific splicing compared to controls; II Prioritization of driver somatic mutations based on the functional networks. The functional importance of mutations is evaluated; III Proposed mutation-AS model to explain principles of genetic heterogeneity; IV Clustering analysis based on AS to identify cancer subtypes with distinct clinical features.
C The average number of AS events per tumor detected in each cancer type from a total of 10, samples. Alternative splicing AS is a highly regulated process that adds complexity to human transcriptome, proteome and signal transduction networks in the cell Braunschweig et al. Tissue- and cell-type specific AS patterns have been shown to play critical roles Identifying and validating alternative splicing events in houston development and differentiation Buljan et al.
Aberrant AS events have been implicated in complex diseases, including various types of cancer David and Manley, ; Misquitta-Ali et al. AS alterations may confer a selective advantage to the tumor, such as cell proliferation, invasion and apoptosis evasion Dominguez et al. The determination of AS deregulation in cancer is therefore of utmost relevance reveal novel oncogenic mechanisms. Although considerable efforts have been made to study AS alterations in individual cancers, the extent to which aberrant AS perturbations contribute to cancer progression remains largely unknown.
Besides identification of aberrant AS events across cancer types, identifying molecular determinants and mechanisms that perturb AS in cancer is fundamental for the development of cancer-specific biomarkers for prognosis and therapy Barash et al. Lines of evidence have demonstrated that AS Identifying and validating alternative splicing events in houston in cancer may be caused by changes in expression, amplification and deletions in splicing factors and RNA-binding proteins Hollander et al.
Given the complexity of AS events, it is not surprising that they are particularly susceptible to genomic mutations implicated in human cancer Lu et al. Indeed, it has been increasingly appreciated that AS events are influenced by genomic mutations.
For instance, genetic variants that affect splicing have been inferred by deep learning algorithms Wan et al. Nevertheless, the general principles by which somatic mutations lead to AS alterations across diverse cancer types are unknown and have the potential to reveal oncogenic mechanisms in diverse cancers. Functional networks provide an informative platform to investigate properties of cellular systems Barabasi and Oltvai, ; Vidal et al.
Network-based approaches have been successfully applied to identifying cancer genes Barabasi et al. It is now clear that patients with the same cancer type have highly heterogeneous genotypes with diverse genomic alterations Vogelstein et al. Therefore, we urgently need methods to assess the impact of patient-specific mutations on AS events from individual tumors in order "Identifying and validating alternative splicing events in houston" discover personalized driver mutations.
Toward these goals, we developed an integrated, multi-scale framework hereafter referred to as DrAS-Net; Figure 1B and performed a large-scale systematic investigation of somatic mutation-mediated AS patterns across 33 types of cancer. Our integrated analysis revealed widespread AS changes across cancer types.
Cancer types with similar tissue origins form clusters based on differential AS patterns. By integrating genomic mutations and AS events into functional association networks, we describe a framework to identify patient-specific potential driver mutations that mediate AS alterations in cancer. The identified driver candidates were enriched in cancer hallmark genes, and that cancer subtypes with distinct clinical features could be identified by their AS profiles.