Abstract A161: Targeting gastrointestinal tumors with constant region engineered anti-glycan antibodies

Abstract

Abstract Post-translational modifications, for instance protein and lipid glycosylation, are attractive targets for therapeutic antibody (mAb) development. The altered tumor glyco-code drives oncogenic features such as the ability to proliferate, metastasize as well evade immune detection. Through immunizations with colorectal cancer cell membrane extracts we have generated a panel of anti-glycan mAbs with potential for application as cancer therapeutics. These mAbs selectively target Lewis or sialylated Lewis glycans on glycoproteins and/or glycolipids, binding to a large percentage of colorectal, pancreatic and gastric tumor tissues on tumor microarrays and induce a significant tumor volume reduction combined with a survival benefit in metastatic colorectal cancer xenograft models. Underlying these potent antitumor responses is the mAbs’ ability to induce direct tumor cell killing in the absence of complement or immune effector cells. In vitro, this direct cytotoxicity is characterized by mAb-induced cellular aggregation, pore formation, release of alarmins (ATP and high mobility group box 1 protein (HMGB1)), and maturation of immature dendritic cells, thereby constituting a form of inflammatory cell death (ICD). This mode of cell killing is linked to mAb cooperative binding on repeating antigen, a characteristic mostly associated with the murine mIgG3 isotype, thus human hIgG1 formats do not exhibit this form of direcT-cell killing. We have identified the residues required for the cooperative binding behaviour, through mAb constant region (CH2/CH3) screening, and have engineered improved (‘i’) hIgG1 variants, containing these selected residues. In silico immunogenicity prediction (IEDB) suggests limited immunogenicity, which requires further validation. Functional characterization of three improved anti-glycan hIgG1 mAbs in in vitro cell-based and Biacore assays demonstrates significant direct cancer cell killing, pore formation as well as increased functional glycan affinity. Classical immune effector functions (ADCC and CDC) were maintained or improved. Importantly, these improved hIgG1 variants now show significant tumor volume reduction in vivo in a mouse colorectal xenograft model.The multifaceted killing activity of our hIgG1 anti-glycan mAbs, has the potential to synergize with checkpoint blockade, thus holding tremendous therapeutic promise for the treatment of gastrointestinal tumors. Additionally, the activity of other therapeutic mAbs could be further enhanced with our Fc-engineering strategy for introducing cooperative binding. Citation Format: Mireille Vankemmelbeke, Thomas Kirk, Jia X. Chua, Richard McIntosh, Lindy G. Durrant. Targeting gastrointestinal tumors with constant region engineered anti-glycan antibodies [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A161.