PDGFRβ Is a Novel Marker of Stromal Activation in Oral Squamous Cell Carcinomas

Carcinoma associated fibroblasts (CAFs) form the main constituents of tumor stroma and play an important role in tumor growth and invasion. The presence of CAFs is a strong predictor of poor prognosis of head and neck squamous cell carcinoma. Despite significant progress in determining the role of CAFs in tumor progression, the mechanisms contributing to their activation remain poorly characterized, in part due to fibroblast heterogeneity and the scarcity of reliable fibroblast surface markers. To search for such markers in oral squamous cell carcinoma (OSCC), we applied a novel approach that uses RNA-sequencing data derived from the cancer genome atlas (TCGA). Specifically, our strategy allowed for an unbiased identification of genes whose expression was closely associated with a set of bona fide stroma-specific transcripts, namely the interstitial collagens COL1A1, COL1A2, and COL3A1. Among the top hits were genes involved in cellular matrix remodeling and tumor invasion and migration, including platelet-derived growth factor receptor beta (PDGFR beta), which was found to be the highest-ranking receptor protein genome-wide. Similar analyses performed on ten additional TCGA cancer datasets revealed that other tumor types shared CAF markers with OSCC, including PDGFR beta, which was found to significantly correlate with the reference collagen expression in ten of the 11 cancer types tested. Subsequent immunostaining of OSCC specimens demonstrated that PDGFR beta was abundantly expressed in stromal fibroblasts of all tested cases (12/12), while it was absent in tumor cells, with greater specificity than other known markers such as alpha smooth muscle actin or podoplanin (3/11). Overall, this study identified PDGFR beta as a novel marker of stromal activation in OSCC, and further characterized a list of promising candidate CAF markers that may be relevant to other carcinomas. Our novel approach provides for a fast and accurate method to identify CAF markers without the need for large-scale immunostaining experiments.