Fc engineering of CD9 antibodies overcomes platelet toxicity and preserves superior potency against refractory acute lymphoblastic leukemia

Abstract Introduction Advances in CD19/CD22-directed immunotherapies have revolutionized the treatment landscape of relapsed/refractory acute lymphoblastic leukemia (ALL). However, relapses after immunotherapy are common, and a significant proportion of patients will display antigen loss with limited salvage options. Our earlier studies showed that the tetraspanin family protein CD9 was broadly expressed in ALL cases and associated with a high relapse risk, where antibody blockade markedly suppressed leukemia progression in preclinical models (Leung et al, Leukemia, 2020/2024). Yet, lethal platelet toxicities incurred by existing CD9 antibodies have largely hampered their clinical translation despite its clear prognostic and therapeutic relevance. This study aims to develop a safe and effective CD9 biologic for refractory ALL through antibody engineering. Methods Murine CD9 antibodies were generated by standard immunization/hybridoma technologies and humanized via complementarity-determining region (CDR) grafting. Fc engineering was performed via site-directed mutagenesis. Antibody affinity and specificity were assessed against soluble CD9 antigen by biolayer interferometry and on CD9-expressing/depleted ALL cell lines by flow cytometry, while efficacy was confirmed with cell line-derived (CDX) and patient-derived xenograft (PDX) models. Immunogenicity was evaluated by interferon-gamma (IFN-γ) release assay and CD4⁺ T cell proliferation by CFSE labelling following antibody stimulation. Platelet toxicity was evaluated by CD62P expression, optical aggregometry and in FcγRIIA transgenic mice. Mechanisms were explored by cell proliferation, apoptosis, imaging and antibody-dependent cellular cytotoxicity (ADCC) assays. Results Two isolated murine monoclonal antibodies (mAbs) displayed distinct binding affinity and specificity towards the CD9 antigen, with biophysical properties preserved following humanization (hAbs). Limited CD4⁺ T cell activation was observed upon in vitro hAbs stimulation, suggesting minimal immunogenicity. Administration of both hAbs in 697-grafted NOD/SCID mice markedly suppressed medullary (98.6±0.7% reduction, P=0.021) and extramedullary leukemia load (brain: 94.5±1.7%, P=0.040; testes: 99.4±0.2%, P=0.036), alongside with a significant extension of animal survival (1.6-1.8-fold, P<0.01). Consistent with previous observations, platelets were fully activated within 30 minutes after exposure to hAb IgGs but not their Fab or scFv fragments, indicating that CD9-mediated platelet activation is a Fc-dependent event. By introducing a N297A mutation to the antibody Fc domain, platelet activation and aggregation was largely reduced when compared with the non-engineered version as reflected by the absence of phenotypic CD62P expression (5.1±0.9% vs. 89.6±4.8%, P<0.001) and physical aggregation (AUC: 18.9-29.5 vs. 382-395, P=0.002). Additionally, no thrombocytopenia was observed in FcγRIIA transgenic mice with platelet counts in the hAb arm similar to those in the human IgG control arm (929±84 vs. 968±118 x 103/uL, P=0.794). Notably, single-agent N297A-engineered hAbs treatment profoundly reduced multi-organ leukemia burden by >98% (P<0.01) and introduced clear survival benefits (53-62 days vs. 27 days, P=0.003) in PDXs developed from ALL patients who relapsed from CAR T or blinatumomab therapies displaying CD19 antigen loss. Fc-engineered hAbs triggered homotypic aggregation in CD9+ ALL cells, a typical feature associated with type II antibodies, leading to direct cell death and proliferation inhibition. Intriguingly, hAbs exerted its anti-leukemic effects independently of classical caspase-dependent apoptosis and antibody-dependent cellular cytotoxicity (ADCC), revealing a non-canonical mechanism of action. Conclusion In sum, Fc engineering effectively mitigates platelet toxicity of CD9 antibodies while retaining remarkable anti-leukemia activity against refractory ALL. This study potentially breaks the bottleneck of CD9-targeted therapy development and puts forward a new agent for refractory ALL patients failing upfront immunotherapies. Further validations in large animal models are ongoing.