Preclinical development and evaluation of CD87-directed CART for colorectal cancer

Background: Chimeric antigen receptor T-cell (CAR-T) therapy has shown success in hematologic malignancies; however, its efficacy in solid tumors, including colorectal cancer (CRC), remains limited by poor tumor penetration, an immunosuppressive microenvironment, antigen heterogeneity, and a risk of on-target/off-tumor toxicities. CD87, the receptor for the urokinase plasminogen activator (uPAR), has emerged as a promising target, given its near-universal overexpression in CRC, and its near absence in normal tissues, except certain myeloid cells. Recent studies on the elimination of senescent myeloid cells in mice have supported the safety of CD87-directed CAR-T cells in vivo . Here, we pursued safety and efficacy studies of human CD87-directed CART (CART87) for use in CRC. We employed a Jurkat NFAT-Lucia reporter platform to functionally screen and optimize human CD87-specific CAR designs, then advanced the top candidates into CARTs for testing against a panel of CRC cell lines. Moreover, we employed bioinformatics to identify potential risks associated with targeting CD87 in humans. Methods: To assess the potential safety of targeting CD87, we analyzed public bulk and scRNAseq datasets from human and mouse tissues. Furthermore, we quantified CD87 surface expression in human blood leukocyte populations and assessed their susceptibility to CART87. To optimize the CAR87 design, we employed a Jurkat NFAT-Lucia reporter assay, which measured CAR activation upon stimulation with recombinant uPAR and CD87-expressing cells. Constructs with high tonic signaling and non-specific activation by CD87-negative cells were eliminated. We then engineered CD87-directed CAR-T cells (CART87) and assessed their activity against CD87-positive CRC cell lines compared with CD87-negative HEK293 cells and CD87 knockout (KO) cells. Results: scRNAseq analysis of bone marrow, whole blood leukocytes, and organ tissues revealed that CD87 mRNA expression is restricted to mature monocyte-lineage cells, with murine organs/tissues showing analogous results. Correspondingly, CD87 surface expression by flow cytometry was largely restricted to mature myeloid lineage cells. The Jurkat-NFAT reporter system enabled rapid comparison of single-chain variable fragments (scFv), structural and signaling domain configurations, and identified lead construct #3 (CAR87.3). CART87.3 demonstrated specific and potent cytolytic activity in vitro against CD87-positive CRC cell lines T84, SW480, LS174T, and DLD1. Conclusions: This work identifies CD87 as a safe, selectively-expressed target antigen for the development of CAR-T therapy in CRC. The in vitro results support the advancement of human CART87 towards in vivo validation. Furthermore, we describe the rational design of a CART87 construct using a high-throughput strategy that could be broadly applied in CAR design for other target antigens.