Endocrine Abstracts

Neuroendocrine neoplasms (NENs) are a highly heterogeneous group of tumors that arise from the diffuse neuroendocrine system and whose incidence has increased over the last years. Among them, pancreatic NENs (PanNENs) are relatively common and one of the most studied NENs. PanNENs are characterized by a low number of mutations, with some genes frequently mutated, such as MEN1, ATRX/DAXX, and mTOR signaling pathway genes. Despite genomics, transcriptomics and epigenomics studies that have helped improved understanding of the molecular features of PanNENs, there is still vast unexplored ground for better comprehension of this disease. In this context, the process of RNA splicing, which is the removal of introns from pre-RNA molecules, is known to be dysregulated in many diseases and it has even been shown to contribute to every hallmark of cancer. Our group has previously demonstrated that RNA splicing is also dysregulated in PanNENs, where it opens up new therapeutic avenues. However, clinical and molecular implications of this dysregulation are still very poorly understood. In this work, we aimed to describe the spliceosomic landscape, that sums alternative splice variants, events and splicing machinery components expression, as well as their potential relationship, how it is related to the pathology of PanNENs and its contribution to their high heterogeneity, tumorigenesis and behavior. With this aim, we explored two different cohorts of 175 and 66 PanNENs samples, including RNA-seq, mutational and clinicopathological data, integrated to better understand the disease. The expression levels of splicing variants/events and splicing machinery components in each sample were quantified using SUPPA-2 and DESeq-2, respectively. We used dimensionality reduction techniques to classify samples according to their spliceosomic landscape and later explored their relationship with pathological and mutational data. Results revealed important associations between clinical data, such as hormone secretion (tumor functionality), tumor grade or metastasis, and specific splicing profiles, implying a relevant role of alternative splicing in PanNENs pathophysiology. Additionally, the most common mutations in PanNENs showed specific splicing landscapes. Further analysis and experiments will be required to ascertain whether splicing alterations are merely a subsidiary effector of hierarchically higher molecular changes or, contrarily, function as tumor drivers. In conclusion, our study represents the first spliceosomic study in PanNENs and provides original information to better understand the heterogeneity and pathophysiology of these tumors.