# A masking clamp for conditional activation of therapeutic antibodies

## TL;DR

Scientists developed a new way to control when antibodies attack tumors, reducing harm to healthy cells.

## Contribution

A human-derived, modular masking platform using calmodulin and a calmodulin-binding peptide for conditional antibody activation.

## Key findings

- The CaM-CBP peptide clamp reduced antibody binding potency by up to 410-fold in on-cell assays.
- Masking suppressed ADCC activation by up to 78-fold and restored function via MMP-9-mediated demasking.
- Structural optimization showed linker length and clamp positioning significantly affect masking efficiency.

## Abstract

Therapeutic monoclonal antibodies (mAbs) constitute cornerstone therapeutics in oncology, yet their clinical utility is often limited by on-target, off-tumor toxicity due to shared antigen expression in both tumor and healthy tissues. To counteract this issue, various approaches, including pH-dependent, as well as affinity-based and steric hindrance-based masked antibodies, have been developed. Several steric hindrance-based masking strategies have been proposed utilizing non-human proteins, potentially leading to an immunogenic response. To address this challenge, we engineered a modular protein-based masking platform leveraging the high-affinity interaction between human calmodulin (CaM) and a calmodulin-binding peptide (CBP). This strategy enables conditional activation of antibodies via tumor microenvironment (TME)-associated proteases (e.g., MMP-9), minimizing systemic off-tumor binding. The CaM-CBP peptide clamp, composed exclusively of human-derived protein domains, was fused to the amino termini of heavy and light chains of trastuzumab and cetuximab. On-cell binding assays demonstrated up to a 410-fold reduction in EC50 for masked constructs across multiple antigen-antibody systems. Functional validation using a reporter-cell-based antibody-dependent cellular cytotoxicity (ADCC) assay confirmed that masking abrogated effector cell activation, leading to up to 78-fold reduction of EC50 and no ADCC activation at concentrations corresponding to the onset of maximal ADCC activation by unmodified antibodies. Demasking via MMP-9-mediated linker hydrolysis restored antigen binding and ADCC potency. Structural optimization revealed that linker length and clamp positioning critically influenced masking efficiency. This human-derived, modular masking platform mitigates immunogenicity risks while enabling tumor-selective antibody activation. Its adaptability across antibody scaffolds underscores broad applicability for improving the therapeutic index of antibodies.