Monoclonal antibody inhibition of thrombin cleavage of osteopontin as a novel therapeutic to enhance host antitumor immunity and reduce cardiac fibrosis

Osteopontin (OPN), a circulating protein with pleiotropic functions, is highly expressed in many cancers and is known to have significant tumor-promoting activity. Thrombin cleaves OPN at arginine(Arg)153 and exposes a previously cryptic α4β1 and α9β1 integrin-binding site at the new C-terminus in the N-terminal thrombin-cleaved

OPN fragment, OPN-Arg (OPN-R). OPN-R can be further processed to OPN-Leu (OPN-L). We have shown elevated OPN-R and OPN-L levels in arthritis joint fluid, and cerebral spinal fluid in glioblastoma, demonstrating that these cleavages occur in vivo. To define the role of thrombin cleavage of OPN in cancer biology, we created

a thrombin cleavage-resistant OPN-knock in (KI) mouse in which Arg153 is substituted with alanine. In a murine B16 melanoma model, OPN-KI mice showed significant suppression of tumor growth and pulmonary metastasis. Tumor suppression was eliminated in severely immune deficient NOG-OPN-KI mice and was abolished by

macrophage depletion in OPN-KI mice. Thrombin inhibition by the direct oral thrombin inhibitor, dabigatran, replicated the B16 tumor suppression in wild type (WT) mice. Tumor-associated macrophages (TAMs) were increased in tumors from OPN-KO and OPN-KI mice, with a switch from M2 macrophages to TAMs with a new

activation profile. We also showed suppression of murine ovarian tumor in the OPN-KI mouse. Thus, thrombin cleavage of OPN initiates OPN’s tumor-promoting activity by suppressing the host-antitumor immune response mediated by modulation of the TAMs. We hypothesize that an anti-OPN monoclonal antibody (mAb) that blocks

OPN cleavage by thrombin would be a novel adjunctive immuno-oncologic therapeutic that enhances the host- antitumor immune response mediated by TAMs. We have generated rabbit mAbs targeting the thrombin cleavage site of OPN that inhibit thrombin cleavage of human and mouse OPN. We converted the rabbit mAb

A6 to a chimeric murine/rabbit mAb, and showed that chimeric mAb A6 suppressed B16 tumor growth and metastasis in WT mice, without any associated anticoagulant side effects, thus demonstrating proof-of-concept that A6 can be a lead therapeutic candidate. A6 has been humanized, and mAb hA6v3 binds human and mouse

OPN and prevents thrombin cleavage of OPN. In collaboration with CIEA (Japan), we have created NOG-OPN- KI and NOG-OPN-KO mice, and have reconstituted the NOGhIL3/GM-CSF mice with human cord blood CD34+ cells. We have also generated a series of mAbs against OPN-R, the OPN fragment that may play the key role

in suppressing the host-antitumor immune response. Anti-OPN-R mAb may be as effective as A6 with a major advantage over A6 in that the circulating pool of OPN-R is much smaller than OPN, and therefore the drug requirement as a therapeutic will be much less. We have identified a lead anti-OPN-R mAb C6r and it is a co-

lead for this translation project. Our goal is to convert the lead mAbs into a clinical candidate with the target of

filing an IND such that it will be an attractive drug candidate for out-licensing. Aim 1: Prove that inhibiting thrombin cleavage of OPN with the lead humanized anti-OPN mAb hA6v3 improves cancer outcomes. Subaim 1.1: Optimization of hA6v3 by determining its structure when bound to the immunogenic peptide. Subaim 1.2: Show

that hA6v3 reduces tumor growth in our humanized mouse cancer models. We will test hA6v3 on many human cancer cell xenografts including melanoma, ovarian, and breast carcinoma and non-small cell lung cancer. hA6v3 will be given in the presence or absence of anti-PAD-L1 mAb to determine if they have an additive effect.

Subaim 1.3: Testing hA6v3 in human melanoma in humanized mice in combination with DTIC. Aim 2: To develop a mAb against OPN-R as a co-lead to the anti-OPN mAb A6. Subaim 2.1: Test the efficacy of C6r in the mouse B16 model. Subaim 2.2: If Subaim 2.1 is accomplished, C6r will be humanized and its efficacy in suppressing

human tumor growth and metastasis will be tested in our humanized mouse model as described for hA6v3. We have extensively filed patent applications to protect the intellectual property in this project through the Stanford University Office of Technology Licensing (OTL) with explicit participation by the VA.