Abstract A096: Glycosyltransferase B4GALNT3 drives LacdiNAc deposition as part of cellular reprogramming in pancreatic precancer

Abstract

The incidence of pancreatic cancer is steadily increasing and is projected to become the second leading cause of cancer-related deaths by 2030. Intraductal papillary mucinous neoplasms (IPMN) are macroscopic precancerous lesions in the pancreas that can be identified by routine imaging. This allows for an opportunity for surgical resection before cancer develops. Markers distinguishing benign from malignant IPMN, however, are lacking, leading to an abundance of high-risk, high-complication surgeries, with a non-insignificant risk of mortality. A better understanding of the mechanisms underlying IPMN progression is required to identify markers of progression to risk-stratify patients for curative-intent surgery. One defining mutation of IPMN is GNAS R201C . Using cell lines expressing Kras G12D +/- GNAS R201C , we previously showed that constitutive GNAS signaling drives an indolent phenotype, including a mesenchymal to epithelial (MET) transition and a global increase in glycosyltransferase (GT) expression, including beta-1,4-N-acetyl-galactosaminyltransferase 3 (B4galnt3). B4GALNT3 transfers GlcNAc to GalNAc to create the disaccharide LacdiNAc (LDN). Enhanced abundance of LDN in IPMN cell lines decreased invasive capacity, while cleavage by chitinase enhanced invasion. How LDN deposition and cleavage alone impact IPMN progression and PDAC development is currently unknown. To determine the functional role of LDN, we leveraged site-specific glycoproteomic, single-cell transcriptomic, and molecular approaches. Importantly, we identified critical cell-cell and cell-extracellular matrix receptors preferentially modified with LDN, such as integrins, a potential mechanism by which LDN inhibits invasion. Overexpression of B4GALNT3 in Kras G12D +/- GNAS R201C cells from murine IPMN-derived pancreatic cancer resulted in a decrease in expression of both mesenchymal (Fn1 and Zeb1) and epithelial (Epcam and Cldn2) markers, suggesting reprogramming of cell state driven by LDN deposition. Further, we saw a decrease in the glycan 3’ sulfated Lewis A/C, a known high-grade pancreatic cancer marker, suggesting that LDN competes with 3’ sulfated Lewis A/C. Finally, to investigate how LDN is lost during disease progression, we performed 10x Genomics Xenium-based spatial transcriptomics on 32 patient tissues spanning normal to IPMN, to IPMN-derived PDAC and identified an abundance of CHIT1 (chitinase)-secreting macrophages in high-grade and invasive IPMN as compared to low-grade, benign lesions. Overall, we’ve identified a putative mechanism by which LDN glycan modification of cell surface proteins inhibits IPMN progression while inflammation-induced chitinase activity from macrophages cleaves LDN and drives progression. These findings give insight into the drivers of IPMN pathogenesis and provide a foundation for glycans as both markers and mediators of pancreatic cancer development.