Abstract B28: Evidence for the antagonistic form of CXCL10 in high grade serous epithelial ovarian tumours

Abstract

Abstract The chemokine CXCL10 is a potent angiostatic and chemoattractive agent, whose primary role is to recruit activated T-cells into sites of inflammation. We have investigated the expression of CXCL10 in serous epithelial ovarian cancers (SOC), with a focus on specific post-translational modifications that control its function. The study group was comprised of a small cohort of patients diagnosed with high grade (HGSOC) or low grade (LGSOC) ovarian tumors, benign SOC, non-ovarian benign gynecological disease or no pathology. Patients were post-menopausal and had not undergone any prior treatment. CXCL10 expression in tumor tissue was assessed by quantitative PCR and western blotting. Immunoassay was used to analyse plasma CA125, and plasma and urine CXCL10 levels in matched patient samples. Localization and correlation with leukocyte recruitment was determined using immunohiostochemical staining of tumor tissue for CXCL10 and the leukocyte cell surface antigen CD45. Proteolytic processing of CXCL10 was analyzed using MALDI imaging mass spectrometry (IMS), and de novo sequence information obtained by immunoprecipitation and MS/MS analysis. We identified significantly elevated expression of both CXCL10 mRNA and protein in HGSOC, with the highest levels evident in early stage (FIGO stage Ib-c) disease. CXCL10 expression was localized to the cytoplasm of tumor epithelial cells. Despite high levels of CXCL10, however, significant leukocyte infiltration was only observed in low grade tumors or within benign cysts; HGSOC contained significantly fewer CD45+ infiltrating cells in tumor tissue. The proteolytic removal of 2 N-terminal amino acids converts CXCL10 into a potent antagonist of chemotaxis and T-cell recruitment. We therefore investigated the presence of proteolytic modifications to CXCL10 using mass spectrometry. MALDI IMS revealed masses suggesting a 2-amino acid and a 9-amino acid truncation at the N- or C-termini of CXCL10, respectively. These masses also displayed strikingly different localization; N-terminally truncated CXCL10 was present only in tumor epithelium, whilst C-terminally truncated CXCL10 was located only in the intervening regions of stromal tissue. Immunoprecipitation and MS/MS analysis confirmed the presence of N-terminally truncated CXCL10 in HGSOC. Our data identify, for the first time, the presence of the “antagonistic” form of CXCL10 in HGSOC. Antagonistic CXCL10 is present at a clinically early stage of tumor progression and correlates with poor leukocyte infiltration into tumors, suggestive of a direct influence on patient outcomes. We hypothesize that tumor-specific conversion of CXCL10 from agonist to antagonist represents a novel mechanism by which HGSOC may partially evade the early immune response. Our data also suggest inhibition of CXCL10 cleavage may represent a new avenue for the treatment of patients diagnosed with HGSOCs. Citation Format: Andrew N. Stephens, Jyothsna R. Rao, Santanu Deb-Choudhury, Adam Rainczuk. Evidence for the antagonistic form of CXCL10 in high-grade serous epithelial ovarian tumors. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr B28.