Thrombin Cleavage of Osteopontin Suppresses Host-Anti-Tumor Immune Response in Cancer

点击翻译按钮获取中文摘要

Expression of osteopontin (OPN), a circulating matricellular protein with pleiotropic functions, is up-regulated in inflammation and cancer. OPN has multiple functional domains and a thrombin cleavage site (at Arg168 in human and Arg153 in mouse). Thrombin cleavage exposes a cryptic integrin-binding site for α4β1 and α9β1 integrins at the new C-terminus, SVVYGLR. We have shown that Arg168 is a bona fide thrombin cleavage site, and thrombin- cleaved OPN-Arg (OPN-R) has enhanced α4β1-dependent cell-binding activity which is abolished when the C- terminal arginine is cleaved by carboxypeptidase N (CPN) or thrombin-activatable CPB2, converting it to OPN- Leu (OPN-L). OPN-R and OPN-L levels are elevated in inflammatory joint fluid and cerebral spinal fluid in glioblastoma (GBM), based on specific ELISAs developed in our lab, demonstrating that these cleavages occur in vivo. OPN has been implicated in promoting invasive and metastatic progression of many cancers, including breast, lung, prostate and ovarian cancers, GBM and melanoma. However, the role of thrombin cleavage of OPN in cancer biology in vivo is undefined. We created a thrombin-resistant OPN knock-in (KI) mouse in which Arg153 is replaced by alanine (OPNR153A). OPNR153A KI mice are healthy and fertile. In preliminary studies, we showed: 1) decreased murine B16 melanoma growth and pulmonary metastasis in OPN KO and OPNR153A KI mice compared to WT mice. 2) OPN KO and OPNR153A KI B16 tumors had significantly increased infiltrating F4/80+ macrophages. 3) Tumor suppression in the OPNR153A KI mice was abolished by macrophage depletion in vivo and it was not observed in the immune deficient NOG-WT, NOG-OPN KO and NOG-OPNR153A KI mice. Thus, tumor suppression in OPNR153A KI mice is mediated by F4/80+ macrophages. 4) Oral administration of dabigatran, an orally active direct thrombin inhibitor, suppressed B16 tumor growth and metastasis in WT mice, replicating the OPNR153A KI phenotype. 5) Malignant ascites was suppressed in OPNR153A KI mice in a murine ovarian cancer model. Our overall hypothesis is that thrombin cleavage of OPN leads to suppression of the host-anti-tumor immune response, associated with a decrease in tumor-associated macrophages (TAMs), thus favoring tumor growth and metastasis. Blocking thrombin cleavage of OPN will lead to enhancement of the host-anti-tumor response, resulting in reduced tumor growth and metastasis. Specific Aim 1 will determine if B16 tumor suppression in the OPNR153A KI mouse is generalizable to other murine and human cancer models. We will test the OPNR153A KI mouse in the murine ovarian cancer and GBM models (Subaim 1.1). We will reconstitute the immune deficient NOG mice with human immune cells to create “humanized” NOG-OPN KO and NOG-OPNR153A KI mice. We will validate the tumor suppression observed in the murine cancer models with the corresponding human cancer models, and test additional human cancer models in these “humanized” mice. Specific Aim 2 will test whether dabigatran functions as an adjunctive therapy with standard melanoma therapy. We will test whether the OPNR153A KI mouse shows tumor suppression with murine YUMM3.1 melanoma cells carrying the BRAFV600E mutation (Subaim 2.1). We will then test whether dabigatran, in combination with vemurafenib (BRAF kinase inhibitor) or vemurafenib and cobimetinib (MEK kinase inhibitor), prolongs the survival of WT mice with YUMM3.1/BRAFV600E melanoma (Subaims 2.2 and 2.3), and whether dabigatran adds to dacarbazine in the survival of WT mice with B16 melanoma (Subaim 2.4). In Specific Aim 3, we will over-express OPN-R, OPN-L, and OPN-CTF (C-terminal fragment) in the OPN-KO and OPNR153A mice to determine which cleaved OPN form reverses the B16 tumor suppression phenotype (Subaim 3.1). We will characterize TAMs from B16 tumors in OPN KO and OPNR153A KI mice and compare them to that from WT B16 tumors (Subaim 3.2). We hypothesize that thrombin cleavage of OPN in WT mice reduces TAM infiltration and induces a “tumor promoting” M2-like phenotype, whereas TAMs in the OPN KO and OPNR153A mice will have a proinflammatory “tumor suppressing” M1-like phenotype. Our observations are novel, significant and relevant to cancer research and clinical practice.