Background Familial adenomatous polyposis (FAP) is characterized by hundreds of colorectal adenomas that inevitably progress into carcinomas. This study focused on the heterogeneity during the polyposis progression to identify new targets and signatures for therapeutic development. Methods An integrated analysis of single-cell sequencing ( GSE109308 ) and bulk transcriptomic data ( GSE79460 , GSE88945 , GSE94919 , GSE106500 , GSE109812 , GSE153385 , and GSE156172 ) of FAP patients from the Gene Expression Omnibus database was conducted. The heterogeneous features of epithelial cell clusters were described in terms of evolutionary trajectory, stemness, hypoxia, epithelial-mesenchymal transition (EMT), immune infiltration and metabolism. Three machine learning algorithms were applied to identify the key cell subset driving polyp heterogeneity, followed by functional validation with cell cycle and viability experiments. Cell-cell communication landscapes of this significant cell subset and its associations with prognosis and response to chemotherapy or immunotherapy were delineated. Results Thirteen epithelial cell clusters were determined and further classified into four heterogeneous phenotypes. A specific cell population, G protein-coupled receptor 182 (GPR182)-positive polyp stem cells (GPR182 + PSCs), was identified as a crucial contributor to heterogeneity. The GPR182 + PSCs showed tumor-priming capacity in evolutionary trajectory analysis and exhibited extensive cell-cell communication with immune cells, especially M2 macrophages and T cells. Importantly, high abundance of GPR182 + PSCs correlated with poor prognosis and elevated expression of immune checkpoints (PD-L1 and CTLA-4) in colorectal cancer. GPR182 + PSCs could also predict the responses of colorectal cancer patients to certain drug treatments, such as rapamycin and midostaurin. Conclusion Our findings map the epithelial heterogeneity in FAP and reveal that GPR182 + PSCs are crucial in driving heterogeneity and immune evasion. GPR182 + PSCs represent promising biomarkers for prognosis and drug responses, providing novel insights into FAP pathology and therapeutic development.
Wu et al. (Fri,) studied this question.