Antibody Logic-Gated Design

The core mechanism requires dual antigen co-engagement for activation. A logic-gated antibody remains inert unless both target antigens (e.g., HER2 and EGFR) are present on the same cell surface. This molecular computation is achieved through split-receptor architectures or masked binding domains that are conditionally unmasked only upon binding of the first antigen.

• True Positive: Both antigens present → therapeutic activation
• True Negative: Single antigen on healthy tissue → no activation
• False Positive Rate: Dramatically reduced compared to monospecific antibodies

The primary clinical advantage is the reduction of systemic toxicity. Conventional monoclonal antibodies often target antigens that are overexpressed on tumors but also present at lower levels on healthy tissue (e.g., EGFR in skin, HER2 in cardiomyocytes). Logic-gated antibodies create a therapeutic window by requiring the unique combinatorial antigen signature of the tumor.

• Mechanism: Healthy cells expressing only Antigen A or Antigen B are ignored
• Result: Higher tolerated doses and reduced adverse events
• Example: A CD3-engaging bispecific masked to activate only when both PSMA and STEAP1 are present on prostate cancer cells

Two dominant engineering platforms enable logic-gated behavior:

Synthetic Notch (SynNotch) Receptors: A modular system where antigen binding to a synthetic receptor triggers proteolytic cleavage and release of a transcription factor or therapeutic payload. Multiple orthogonal SynNotch receptors can be layered to create AND, OR, and NOT gates.

Co-LOCKR (Colocalization-Dependent Latching Orthogonal Cage-Key pRoteins): A computationally designed protein switch that undergoes a conformational change only when two distinct antigen-binding domains colocalize on the same cell surface, exposing a functional domain.

Logic-gated bispecific T-cell engagers (BiTEs) represent the most clinically advanced application. These molecules contain:

• A masked CD3-binding domain that remains sterically occluded
• Two tumor antigen-binding arms that serve as logic inputs
• A protease-cleavable linker or conformational switch that activates only upon dual binding

This ensures T-cell redirection and cytotoxic granule release occur exclusively within the tumor microenvironment, preventing the cytokine release syndrome (CRS) associated with first-generation BiTEs.

Tumors evade monospecific therapies through antigen loss variants—subclones that downregulate the single targeted antigen. Logic-gated antibodies requiring two independent antigens drastically reduce the probability of simultaneous escape.

• Probability Calculation: If antigen loss rate is 10⁻⁶ per cell division, dual loss probability approaches 10⁻¹²
• Clinical Implication: Durable responses with reduced relapse from antigen-negative escape
• Strategy: Target antigen pairs with non-overlapping biological functions (e.g., a growth receptor and an adhesion molecule) to prevent coordinated downregulation

Identifying optimal antigen pairs for logic-gating requires multi-omics integration and machine learning:

• Single-cell RNA sequencing of tumor and normal tissue atlases to quantify co-expression patterns
• Graph neural networks to model antigen interaction networks and identify synthetic lethal-like pairings
• Generative models that simulate antigen density thresholds required for logic gate activation
• Negative selection: Antigen pairs must show zero co-expression on any vital normal tissue

The ideal pair exhibits high tumor co-expression and complete normal tissue orthogonality.

Physics-based and deep learning simulations that predict the three-dimensional binding pose of an antibody relative to its target antigen. For logic-gated designs, docking is critical to:

• Verify that both antigen-binding arms can simultaneously engage their targets without steric clash
• Model the spatial constraints of tumor-associated antigen pairs on the cell surface
• Predict whether the inter-antigen distance on the membrane is compatible with the antibody's reach
• Identify cooperative binding interfaces that enhance conditional activation

Engineering an antibody's binding affinity to be conditional on environmental pH. This mechanism is often combined with logic-gated designs to create multi-layered specificity:

• Antigen binding occurs at tumor microenvironment pH (~6.5) but not at physiological pH (7.4)
• Enables antigen release in acidic endosomes after internalization, allowing antibody recycling
• Extends circulatory half-life by preventing lysosomal degradation
• Adds a third conditional layer (pH AND Antigen A AND Antigen B) for ultra-specific targeting

Computational identification of the specific amino acid residues on an antigen recognized by an antibody's paratope. Essential for logic-gated design because:

• Epitope selection determines whether two antibodies can bind non-competitively to the same target cell
• Identifies epitopes that are co-expressed and spatially proximal on tumor cells but not healthy tissue
• Structural epitope data informs the linker length and geometry between binding arms
• Validates that the chosen antigen pair provides true Boolean AND-gate specificity rather than additive binding

Rational modification of the antibody constant region to modulate effector functions. In logic-gated therapeutics, Fc engineering determines what happens after conditional activation:

• Silenced Fc domains prevent ADCC until both antigens are engaged, reducing off-tumor killing
• Engineered Fc-gamma receptor selectivity directs immune cell recruitment only at the tumor site
• Heterodimerization mutations (e.g., knobs-into-holes) ensure correct chain pairing in bispecific formats
• pH-dependent FcRn binding modifications extend half-life of conditionally-activated antibodies

A comprehensive computational screening cascade that evaluates manufacturing suitability of engineered antibodies. Logic-gated designs face amplified developability risks:

• Bispecific formats are inherently more prone to aggregation and low expression yields
• Multiple CDR engineering steps increase the likelihood of hydrophobic patches and chemical liabilities
• Cross-pairing of heavy and light chains creates product-related impurities requiring removal
• Thermal stability assessment ensures the complex multi-domain architecture remains stable during formulation and storage