Abstract P1-05-16: Sushi Domain Containing 2 (SUSD2): a plasma membrane protein that increases immune evasion in breast tumorigenesis

Abstract

Abstract Routinely used therapies are not adequate to treat the heterogeneity of breast cancer, and consequently, more therapeutic targets are desperately needed. To identify novel targets, we generated a breast cancer cDNA library enriched for breast cancer genes that encode membrane and secreted proteins. From this library we identified SUSD2 (Sushi Domain Containing 2), which encodes an 822 amino acid protein containing a transmembrane domain and functional domains inherent to adhesion molecules. Previous studies describe the mouse homolog, mSVS-1, but there are no studies on the human gene associated with breast cancer. Reverse-Transcriptase PCR analysis using total RNA from frozen breast cancer samples showed that SUSD2 is expressed in 19 of 24 breast tumors. Immunohistochemistry (IHC) analysis was performed on human breast tissues using an anti-SUSD2 antibody. IHC showed weak or no expression of SUSD2 in normal epithelial cells, with the endothelial lining of vessels staining positive for SUSD2. However, staining was observed in pathological breast lesions and in lobular and ductal carcinomas. To explore the role of SUSD2 in breast tumorigenesis, we performed in vitro analyses of stable cell lines generated to over-express SUSD2. Our studies indicate that over-expression of SUSD2 increases the invasion of breast cancer cells through Matrigel and adhesion to fibronectin, an extracellular matrix protein. Additionally, contact between peripheral blood mononuclear cells (PBMCs) and breast cancer cells that express SUSD2 leads to diminished secretion of chemotactic cytokines and increased secretion of angiogenic cytokines compared to the empty-vector control. A similar co-culture experiment with Jurkat T cells showed increased induction of T cell apoptosis when cultured with cancer cells expressing SUSD2. Using an in vivo model with immunocompetent mice, we observed increased tumor growth and decreased survival in mice with mammary tumors expressing mSVS-1. Interestingly, increased tumor growth rate was not due to higher proliferative rates of the cells, as the Ki67 index was similar for all tumors. IHC staining with an anti-CD31 antibody revealed disordered and tortuous microvasculature in tumors expressing mSVS-1. We also characterized the population of lymphocytes in the blood, spleen and tumors of the mice as a measure of the immune response. While proportions of CD4 and CD8 cells were unchanged in the spleen and blood samples, we found significantly fewer CD4 tumor infiltrating lymphocytes in mice with tumors expressing mSVS-1. SUSD2 interacts with Galectin-1 (Gal-1), a 14-kDa secreted protein that is synthesized by carcinoma cells and promotes tumor immune evasion, angiogenesis and metastasis. Several previous studies suggest that cellular localization of Gal-1 is essential for its multiple roles in tumorigenesis. Interestingly, cell surface localization of Gal-1 was observed only in cell lines expressing SUSD2 but not the empty-vector control. Because the localization of Gal-1 on the surface of cells is dependent on the presence of SUSD2, using SUSD2 as a therapeutic target may disrupt this interaction and inhibit events required for tumorigenesis. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-05-16.