To evaluate the utility of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) as a cancer therapeutic, we created leucine zipper (LZ) forms of human (hu) and murine (mu) TRAIL to promote and stabilize the formation of trimers. Both were biologically active, inducing apoptosis of both human and murine target cells in vitro with similar specific activities. In contrast to the fulminant hepatotoxicity of LZ-huCD95L in vivo, administration of either LZ-huTRAIL or LZ-muTRAIL did not seem toxic to normal tissues of mice. Finally, repeated treatments with LZ-huTRAIL actively suppressed growth of the TRAIL-sensitive human mammary adenocarcinoma cell line MDA-231 in CB.17 (SCID) mice, and histologic examination of tumors from SCID mice treated with LZ-huTRAIL demonstrated clear areas of apoptotic necrosis within 9-12 hours of injection.

To define the normal physiological role for the TRAIL/Apo2L in vivo, we generated TRAIL/Apo2L gene‐targeted mice. These mice develop normally and show no defects in lymphoid or myeloid cell homeostasis or function. Although TRAIL/Apo2L kills transformed cells in vitro, TRAIL/Apo2L–/– mice do not spontaneously develop overt tumors at an early age. However, in the A20 B cell lymphoma‐transferred tumor model, TRAIL/Apo2L–/– mice are clearly more susceptible to death from overwhelming tumor burden, due to increased lymphoma load in the liver. A20 tumors are susceptible to TRAIL/Apo2L killing in vitro, indicating that TRAIL/Apo2L may act directly to control A20 cells in vivo. Despite the fact that TRAIL binds osteoprotegerin and osteoprotegerin‐transgenic mice are osteopetrotic, TRAIL/Apo2L–/– mice show no evidence of altered gross bone density, and no alterations in frequency or in vitro differentiationof bone marrow precursor osteoclasts. Moreover, leucine zipper TRAIL has no toxicity when repeatedly administered to osteoprotegerin–/– mice. Thus, TRAIL/Apo2L is important in controlling tumors in vivo, but is not an essential regulator of osteoprotegerin‐mediated biology, under normal physiological conditions.

4-1BB (CD137) is a member of the TNFR superfamily (TNFRSF9). T cell expression of 4-1BB is restricted to activated cells, and cross-linking has been shown to deliver a costimulatory signal. Here we have shown that treatment of tumor-bearing mice with agonistic 4-1BB-specific Abs can lead to T cell-mediated tumor rejection. In vivo mAb depletion experiments demonstrated that this rejection requires CD8+ cells but not CD4+ or NK cells. Both IFN-γ- and CD40-mediated signals were also required, because no benefit was observed on treatment with 4-1BB mAb in mice in which the genes for these molecules had been knocked out. Interestingly, 4-1BB-mediated stimulation of immune responses in CD40L−/− mice is effective (although at a reduced level), and may suggest the existence of an alternative ligand for CD40. Additional experiments in IL-15−/− mice indicate that IL-15 is not required for either the generation of the primary tumor-specific immune response or the maintenance of the memory immune response. In contrast, the presence of CD4 cells during the primary immune response appears to play a significant role in the maintenance of effective antitumor memory. Finally, in mice in which the number of dendritic cells had been expanded by Fms-like tyrosine kinase3 ligand treatment, the antitumor effects of 4-1BB ligation were enhanced.

Talimogene laherparepvec, a new oncolytic immunotherapy, has been recently approved for the treatment of melanoma. Using a murine version of the virus, we characterized local and systemic antitumor immune responses driving efficacy in murine syngeneic models. The activity of talimogene laherparepvec was characterized against melanoma cell lines using an viability assay. Efficacy of OncoVEX (talimogene laherparepvec with the mouse granulocyte-macrophage colony-stimulating factor transgene) alone or in combination with checkpoint blockade was characterized in A20 and CT-26 contralateral murine tumor models. CD8 depletion, adoptive T-cell transfers, and Enzyme-Linked ImmunoSpot assays were used to study the mechanism of action (MOA) of systemic immune responses. Treatment with OncoVEX cured all injected A20 tumors and half of contralateral tumors. Viral presence was limited to injected tumors and was not responsible for systemic efficacy. A significant increase in T cells (CD3/CD8) was observed in injected and contralateral tumors at 168 hours. analyses showed these cytotoxic T lymphocytes were tumor-specific. Increased neutrophils, monocytes, and chemokines were observed in injected tumors only. Importantly, depletion of CD8 T cells abolished all systemic efficacy and significantly decreased local efficacy. In addition, immune cell transfer from OncoVEX-cured mice significantly protected from tumor challenge. Finally, combination of OncoVEX and checkpoint blockade resulted in increased tumor-specific CD8 anti-AH1 T cells and systemic efficacy. The data support a dual MOA for OncoVEX that involves direct oncolysis of injected tumors and activation of a CD8-dependent systemic response that clears injected and contralateral tumors when combined with checkpoint inhibition. .

No abstract