A novel insulin-like growth factor II-based masking domain for conditional activation of therapeutic antibodies

Systemic target engagement constrains the therapeutic index of many antibody drugs, particularly in inflammatory disease and immuno-oncology. Protease-cleavable masking can localize activity to disease sites, but existing approaches often tradeoff masking strength, reversibility, and developability. Here, we evaluated the insulin-like growth factor II (IGF-II) as a compact, structurally defined steric masking domain for protease-dependent conditional activation of therapeutic antibodies. IGF-II masking domains, including a receptor-silent variant (IGF-II-s), were engineered as N-terminal fusions to therapeutic antibodies via protease-cleavable linkers and characterized using antigen-binding and cell-based functional assays. IGF-II masking attenuated anti-TNFα binding and neutralization potency in a protease-dependent manner and activity was largely restored following cleavage. In vivo, an MMP2/9-cleavable masked anti-TNFα antibody retained efficacy comparable to adalimumab and infliximab in a collagen antibody-induced arthritis model, yet, unlike the reference antibodies, did not measurably reduce survival in a Listeria monocytogenes infection challenge model at the dose tested. IGF-II-s showed no detectable IGF receptor binding and provided stronger masking than native IGF-II. This generalized approach was demonstrated across multiple antibodies, enabling efficient protease-dependent conditional activation of trastuzumab (anti-HER2), an anti-IL-1β antibody, and bevacizumab (anti-VEGF). Together, these results established an engineered IGF-II-based masking domain as a versatile, IGF-II receptor-silent platform enabling protease-dependent conditional activation of therapeutic antibodies, with potential to improve safety and expand therapeutic windows.