Abstract 291: Unveiling heterogeneity of cancer-associated fibroblasts (CAFs) across multiple solid cancer types

Abstract Background: Stromal cells, including CAFs, are a heterogeneous group of cells in the tumor microenvironment. Many models describe their heterogeneity like the classification into 3 main subtypes in breast cancer: myCAF (myofibroblasts), iCAF (inflammatory CAF), and PVL (perivascular-like) cells. However, a pan-cancer model has not been established to exhaustively characterize stromal cell heterogeneity, undermining efforts in targeting CAFs with anti-tumor therapies. We investigated the heterogeneity of stromal cells in diverse diagnoses with single-cell RNA sequencing (scRNA-seq) data and bulk RNA-seq of sorted stromal cells to define consistent subtypes. Design: First, a meta-cohort of 11 scRNA-seq open-source datasets (>500,000 cells, >100 patients, 6 diagnoses) was analyzed. We also generated a scRNA-seq dataset of CD90+ cells derived from tumor tissue of an ovarian cancer patient. Additionally, bulk RNA-seq of 6 PVL, 2 iCAF, and 4 myCAF sorted cell samples across four solid cancers was performed. scRNA-seq data was analyzed with the SCANPY Python package. Results: Stromal cell clusters were defined as positive for PDGFRB, THY1 (CD90), and COL1A1 expression, and negative for the expression of known non-stromal and endothelial cell markers, by separately analyzing the open-source scRNA-seq datasets. Analysis of stromal cell heterogeneity revealed that the CAFs were consistently classified into 3 subtypes across all datasets: myCAF (high expression: FAP and ACTA2), PVL (high expression: MCAM (CD146) and ACTA2), and iCAF (high expression: APOD and cytokines IL6, CXCL12, CXCL2). Differential gene expression analysis of the scRNA-seq meta-cohort allowed us to develop functional gene expression signatures to define each subtype. All scRNA-sequenced CD90+ cells from ovarian cancer tissue were classified into the CAF subtypes. The myCAF and PVL signatures were specifically upregulated in the corresponding cell types (p<0.01, Mann-Whitney U test). While the iCAF signature was enriched in both iCAFs and myCAFs, myCAF and PVL signatures were downregulated in iCAFs. Finally, we developed a fluorescence-activated cell sorting (FACS) panel to separate CAF subtypes, using negative markers (CD45, CD31, EPCAM), a positive marker for stromal cells (CD90) and cell-type surface markers (CD146, FAP, CD91, CD44) for myCAF, iCAF and PVL cells. The high expression of CAF-specific genes in the sorted CAF populations denoted high sample purity. Importantly, the 3 CAF subpopulations separated by gating based on different levels of cell-type surface markers showed expression patterns congruent with the developed signatures across diagnoses. Summary: Altogether, the consistent classification of stromal cells with our cell-specific expression profiles demonstrated that the 3 CAF subtypes are conserved in many cancer types, offering groundwork for a pan-cancer model of stromal cell heterogeneity. Citation Format: Andrey Kravets, Boris Shpak, Anastasia Zotova, Maria Savchenko, Sofya Kust, Olga Golubeva, Mary Abdou, Fedor Grigoryev, Irina Bulusheva, Beniamin Sargsyan, Polina Pavlovich, Vlad Maximov, Vladimir Kushnarev, Michael Goldberg, Nikita Kotlov, Alexander Bagaev. Unveiling heterogeneity of cancer-associated fibroblasts (CAFs) across multiple solid cancer types [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 291.