In a study published in Science Advances, the research group led by Dr. WANG Zefeng from Shanghai Institute of Nutrition and Health (SINH) of the Chinese Academy of Sciences and Dr. WANG Yang from Dalian Medical University revealed a widespread length dependent splicing regulation in cancer, and discovered a common feature of cancer-specific splicing dysregulation and provided important clinical implications in cancer diagnosis and therapies by analyzing various transcriptome data from The Cancer Genome Atlas (TCGA). 

Over 95% of human genes undergo alternative splicing (AS), generating multiple mRNA isoforms with distinct functions from a single gene. The widespread splicing dysregulation is one of the molecular hallmarks of cancer, and targeting the mis-spliced genes may be a powerful therapeutic strategy against cancers. Therefore, systematic study of AS in cancer is critical for cancer precision medicine.  

While several studies have provided useful information by analyzing various transcriptome data, there is still a lack of general trend for cancer-associated splicing dysregulation and the underlying mechanisms remain unclear. 

In this study, the researchers conducted a comprehensive analysis of AS changes using transcriptome data from thousands of patients in 18 types of cancers, and they revealed an unexpected length dependency in cancer-associated exons. Compared with typical exons, the short exons are more likely to be mis-spliced and preferentially excluded in almost all cancers. These cancer-associated short exons (CASEs) are more conserved, more likely to encode in-frame peptides and functional enrichment in GTPase regulators and cell adhesion.  

Then, the researchers developed machine learning algorithms with CASEs as diagnostic markers and defined a CASE-based risk factor to accurately predict the prognosis of cancer patients.  

Finally, they determined the possible mechanisms for this length-dependent regulation, which are regulated by elevated transcription rate and alteration of certain RNA binding proteins in cancers.   

This study reveals a general rule for splicing dysregulation in cancers, provides a simple and practical route for cancer stratification, and uncovers the potential mechanisms. Since the splicing dysregulation was recently found to associate with cancer immunotherapy and drug assistant, it provides valuable information for cancer prediction with clinical implications.