Proteomic, Transcriptomic and Genomic Analysis of Esophageal Adenocarcinoma Uncovers Candidate Therapeutic Targets and Cancer-Selective Post-Transcriptional Regulation

Abstract Background Addressing the poor prognosis associated with esophageal adenocarcinoma (EAC) has been hindered by the absence of biomarkers for early disease detection and therapeutic targets. Despite extensive research on the somatic mutations associated with EAC, the impact of genomic aberrations on protein expression and cancer phenotype remains unclear. Methods We conducted a quantitative proteomic analysis using tumour tissue and patient-matched adjacent normal esophageal and gastric tissues from 23 patients undergoing resection for EAC. We further examined the relationship between transcript and protein levels using patient-matched whole transcriptome RNA sequencing and proteomic data from 7 patients and integrated these findings with data from a cohort of 264 EAC patients with RNA sequencing and 454 patients with whole-genome sequencing, as well as external published datasets. Results We quantified protein expression from 5897 genes in EAC and normal tissues. Several potential biomarkers showing selective expression in EAC, such as the transmembrane protein GPA33, were identified. We validated the EAC-specific expression of GPA33 in an external cohort of 115 patients, suggesting its utility as a diagnostic and therapeutic target. Additionally, integrated analysis of protein and RNA expression in EAC and normal tissues revealed genes with discordant protein and RNA levels, indicating post-transcriptional regulation of protein expression. These outlier genes, including SLC25A30, TAOK2, and AGMAT, exhibited rare somatic mutations, suggesting post-transcriptional mechanisms driving this EAC-specific phenotype. AGMAT, as an example, was found to be overexpressed at the protein level in EAC compared to adjacent normal tissues, with a proposed EAC-selective, post-transcriptional mechanism regulating protein abundance. Conclusion Through quantitative proteomic analysis, we identified GPA33 as a candidate EAC biomarker. Integrated analysis of the proteome, transcriptome, and genome in EAC uncovered genes with tumor-selective post-transcriptional regulation of protein expression, offering potential exploitable vulnerabilities for therapeutic intervention.