Understanding the spatial heterogeneity of the tumor microenvironment (TME) in pancreatic cancer (PC) remains challenging.
In this study, we performed spatial transcriptomics (ST) to investigate the gene expression features across one normal pancreatic tissue, PC tissue, adjacent tumor tissue, and tumor stroma. We divided 18,075 spatial spots into 22 clusters with t-distributed stochastic neighbor embedding based on gene expression profiles. The biological functions and signaling pathways involved in each cluster were analyzed with gene set enrichment analysis.
The results revealed that KRT13+FABP5+ malignant cell subpopulation had keratinization characteristics in the tumor tissue. Fibroblasts from adjacent tumor tissue exhibited a tumor-inhibiting role such as “B-cell activation” and “positive regulation of leukocyte activation.” The FGG+CRP+ inflammatory cancer-associated fibroblasts replaced the islets in tumor stroma. During PC progression, the damage to pancreatic structure and function was heavier in the pancreatic exocrine (AMYA2+PRSS1+) than in the endocrine (INS+GCG+).
Our results revealed the spatial heterogeneity of dynamic changes and highlighted the significance of impaired exocrine function in PC.