Abstract 5615: Anti-human CD47 antibody Hu5F9 inhibits MC38 tumor growth in hCD47/hSIRPα double knock-in mice

Introduction: CD47 is broadly expressed on all normal cells or tissues, such as hematopoietic cells, T cells, B cells, monocytes, thymocytes, and platelets as well as other cell lineages, but has reduced expression on Rh erythrocytes. Signal regulatory proteins (SIRP) are a multigene family of immune receptors encoded by a cluster of genes. The most conserved member, SIRPα, binds to CD47 to engage the “don't eat me” signal and avoid macrophage phagocytosis. Notably, CD47 is often highly expressed on various tumors, and is an important immune checkpoint inhibiting macrophage phagocytosis. Therefore, blockade of CD47/SIRPα signalling by an anti-CD47 antibody offers a promising therapeutic approach by unleashing the phagocytosis activity of macrophages against tumor cells. It is well known that the antitumor efficacy of anti-human CD47 antibodies is largely macrophage-dependent in immunocompromised mice (NOD-SCID or NOG®/NSG™) engrafted with xenograft tumors (PDX or cell line derived xenografts). However, the therapeutic effects of anti-CD47 antibodies in immunocompetent mice engrafted with syngeneic tumors have also shown the involvement of cytosolic DNA sensors, dendritic cells, cytotoxic T cells, and type I/II interferons (IFNs). A preclinical animal model with a competent immune system, and humanized CD47/SIRPα is needed to fully recapitulate the mode of action of anti-CD47 therapeutic antibodies. Methods: We developed the human CD47 and SIRPα double knock-in mouse model (CD47/SIRPα HuGEMM™) via CRISPR/Cas9 targeting. We have also engineered mouse syngeneic MC38 cells to express human CD47 (hCD47-MC38 HuCELL™). We engrafted the hCD47-MC38 HuCELL into CD47/SIRPα HuGEMM mice to evaluate the efficacy of the anti-human CD47 antibody Hu5F9. Results: FACS analysis verified that the expression patterns of chimeric h/m CD47 and h/m SIRPα in HuGEMM mice are consistent with the mouse proteins in wild-type C57BL/6 mice. We have also confirmed that an anti-human CD47 antibody (Hu5F9-hIgG4) can bind to human CD47 expressed on hCD47-MC38 HuCELL tumor cells. Finally, we demonstrated robust antitumor effects of an anti-human CD47 antibody by treating the HuGEMM mice bearing hCD47-MC38 tumors with Hu5F9 monotherapy or in combination with anti-PD-1 antibody. Conclusions: Collectively, our hCD47/hSIRPα double knock-in humanized model, along with hCD47-MC38 HuCELL, provides a powerful preclinical platform to evaluate the mode of action, and in vivo efficacy of human-specific