A Dual ICOS/CD28 Antagonist ICOSL Variant Ig Domain (vIgD) Potently Suppresses Human Mixed Lymphocyte Reactions and Human/NSG Mouse Xenograft Graft vs. Host Disease (GvHD)

Background/Purpose: Despite many recent advances in immunotherapy, graft versus host disease (GvHD), which reflects immune-mediated attack of recipient tissue by donor T cells, remains the major cause of morbidity and nonrelapse mortality after hematopoietic stem cell transplantation (HCT). The immunoglobulin superfamily (IgSF) is a large, diverse family of proteins that includes several key T cell-expressed members such as CD28, inducible T cell costimulator (ICOS), PD-1, and CTLA-4. CD28 and ICOS in particular play critical roles in T cell activation and adaptive immunity. Blockade of either pathway suppresses GvHD in syngeneic and xenograft mouse models; however, these two pathways are nonredundant and monotherapies directed against either pathway alone do not appear to be completely efficacious. Therefore, novel molecules that co-antagonize both CD28 and ICOS may provide more effective therapies for the prevention or treatment of GvHD. Methods: We used our proprietary variant Ig domain (vIgDTM) platform, which consists of directed evolution of select IgSF proteins, to create tailored ICOSL-Fc fusion proteins that can bind both ICOS and CD28 with high affinity, and co-inhibit both pathways. Mutant ICOSL-Fc molecules were evaluated in vitro with allogeneic mixed lymphocyte reactions (MLR) by co-culturing human pan T cells with activated human monocyte-derived dendritic cells. Lead candidate molecules were subsequently tested in multiple in vivo mouse models, including ovalbumin-induced delayed type hypersensitivity (DTH) and human xenograft PBMC-NSGTM GvHD. Results: ICOSL-Fc fusion proteins containing variant ICOSL domains significantly attenuate T cell activation in vitro when compared to wild-type (WT) ICOSL or belatacept, a clinically available CTLA4-Ig therapeutic, as assessed by suppressed proliferation and cytokine production in MLR, and ear swelling in DTH. In the human PBMC-NSG GvHD model, treatment with ICOSL-Fc significantly protected mice from the effects of xenogeneic T cell activation in vivo, with treated animals exhibiting greatly enhanced survival and reduced disease scores compared to WT ICOSL and belatacept. The level of protection in this model roughly correlated with the potency of the molecules in the in vitroMLR assay. Conclusion: The vIgD therapeutic platform has broad potential to generate potent biologics for the treatment of serious inflammatory conditions. Efficacy in vitro and in vivo of ICOSL-Fc vIgDs exceeds that of WT ICOSL-Fc or belatacept, correlating with their engineered increased affinities for cognate ligand (ICOS) and CD28. ICOSL-Fc vIgDs are therefore promising, novel therapeutic candidates for the prevention and/or treatment of GvHD. Preclinical development to enable clinical studies is underway.