Genomic and transcriptomic analysis of pituitary adenomas reveals the impacts of copy number variations on gene expression and clinical prognosis among prolactin-secreting subtype

Abstract

Pituitary adenomas (PAs) are slow growing and benign primary intracranial tumors that often cause occupying effects or endocrine symptoms. PAs can be classified into various subtypes according to hormone secretion. Although widespread transcriptional alterations that cause aberrant hormone secretion have been characterized, the impact of genomic variations on transcriptional alterations is unclear due to the rare occurrence of single- nucleotide variations in PA. In this study, we performed whole-genome sequencing (WGS) on 76 PA samples across three clinical subtypes (PRL-PAs; GH-PAs, and NFPAs); transcriptome sequencing (RNA-seq) of 54 samples across these subtypes was also conducted. Nine normal pituitary tissues were used as controls. Common and subtype-specific transcriptional alterations in PAs were identified. Strikingly, widespread genomic copy number amplifications were discovered for PRL-PAs, which are causally involved in transcriptomic changes in this subtype. Moreover, we found that the high copy number variations (CNVs) in PRL-PA cause increased prolactin production, drug resistance and proliferative capacity, potentially through key genes with copy number amplification and transcriptional activation, such as BCAT1. This study provides insight into how genomic CNVs affect the transcriptome and clinical outcomes of PRL-PA and sheds light on the development of potential therapeutics for aberrantly activated targets.