ABSTRACT
PURPOSE: Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy demonstrating aberrant and progressive expression of mucins. The contribution of individual mucins has been extensively investigated in PDAC; however, comprehensive mucin profiling including splice variants in PDAC tumors has not been reported. EXPERIMENTAL DESIGN: Using publicly available RNA sequencing (RNA-seq) datasets, we assess the expression of mucin family members and their splice variants (SV) in PDAC tumor samples for the first time. Mucin SVs that are correlated with PDAC patient survival are validated in a cohort of patient tumor samples. Further, we use computational methods to derive novel pancreatic tumor subtypes using mucin expression signatures and their associated activated pathways. RESULTS: Principal component analysis identified four novel mucin-based PDAC subtypes. Pathway analysis implicated specific biological signatures for each subtype, labeled (i) immune activated, (ii) progressive, (iii) pancreatitis-initiated, and (iv) anti-inflammatory/PanIN-initiated. Assessing mucin SVs, significantly longer survival is observed with higher expression of 4 MUC1 and 1 MUC13 SVs, whereas patients expressing 2 MUC4 and 1 MUC16 SVs had shorter survival. Using a whole-transcriptome correlation, a three-gene panel, including ESRP2, PTK6, and MAGEH1, is designated to assess PDAC tumor sample cellularity by PCR. One MUC4 SV and one MUC13 SV are quantified in a separate PDAC patient cohort, and their effects on survival are experimentally validated. CONCLUSIONS: Altogether, we demonstrate the unique expression pattern of mucins, four mucin-based PDAC subtypes, and the contribution of MUC1, MUC4, and MUC16 SVs in PDAC patient survival.
