Abstract B25: Using species-specific RNAseq reveals stromal reprogramming in triple-negative breast cancer xenografts

The ability of cancer cells to survive and grow at primary and metastatic sites is influenced to a large degree by the surrounding cells in the tumor environment. In this process, tumor cells secrete signals that recruit specific cell types to their immediate vicinity and induce these cells to produce factors critical to tumor development, while the surrounding normal tissue, or stroma, detects and responds to developing lesions. While the clinical importance of this complex dialogue between the tumor and the surrounding tissue has long been appreciated, its molecular basis is not well understood. In particular, studies attempting to characterize interactions between tumors and the surrounding tissue have been confounded largely for technical reasons relating to the difficulty of precisely identifying and separating the developing tumor cells from normal cells nearby, and of confidently assigning the factors identified in later analyses to the proper tissue of origin. Recent developments in high-throughput sequencing technology offer an opportunity to overcome these difficulties and precisely identify factors associated with the stromal response to developing tumors on a genomic scale. Using a novel sequencing strategy applied to a xenograft model of triple-negative breast cancer, we are able to unambiguously assign 93.3% of mapped reads to tumor or stromal transcriptomes, with a misassignment rate approaching 0.01%. These methods have allowed us to directly compare transcript levels in stroma containing developing tumors to those levels observed in naive stroma, and to subsequently identify a set of candidate factors likely involved in the stromal response to developing lesions. We then extended this method to contrast stromal responses to near-isogenic tumors with differing metastatic potential and invasive phenotype conferred by expression of the RKIP metastasis suppressor, thereby identifying candidate factors likely to be specifically relevant to metastatic colonization. Surprisingly, we observed significant up-regulation of genes associated with macrophage chemotaxis and macrophage infiltration in the invasive stroma, which was not present in stroma surrounding tumors expressing RKIP. These findings were confirmed by immunohistochemistry using the original cell line as well as parallel experiments in a syngeneic model, suggesting that by utilizing RNA sequencing technology in an innovative way we can increase our understanding of the interplay between tumors and their local environment. Our model allows us to contrast simultaneous high-precision gene expression estimates in tumors and stroma to infer instances of crosstalk. In the invasive microenvironment, we observed coordinated suppression of genes involved in negatively regulating endothelial cell proliferation accompanied by elevated expression of genes whose products bind growth factors in both tumors and stroma. Finally, we extended these analyses to identify candidate paracrine interactions that are likely to be suppressed in noninvasive contexts involving stromal ALK and MDK which may directly contribute to observed differences in cell growth, migration and angiogenesis. Citation Format: Casey A. Frankenberger, Russell O. Bainer, Jyotsana Menon, Claudia Chavarria, Katelyn Michelini, Pall Melsted, Yoav Gilad, Marsha R. Rosner. Using species-specific RNAseq reveals stromal reprogramming in triple-negative breast cancer xenografts. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Invasion and Metastasis; Jan 20-23, 2013; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2013;73(3 Suppl):Abstract nr B25.