How to Validate Dual Binding for a Bispecific Antibody

"Dual binding" is a decision ladder, not one experiment: (1) does each arm bind its target with the intended affinity, (2) can both targets be engaged simultaneously where the mechanism requires it, (3) are the numbers free of avidity and surface artifacts, and (4) is binding enough to move the program forward? This guide walks those four decisions as delivered by the S02 Dual Binding service (SPR/BLI).

Decision 1 — Confirm each arm binds its target (single-arm affinity)

The first BsAb failure mode is mispairing: a "dual binding" signal can come from a mixture of two monospecific molecules rather than one bispecific. So the first experiment measures the monovalent KD of each arm, not bivalent apparent affinity.

• Capture at low density (Decision 3), fit a 1:1 Langmuir model, report KD, kon, and koff per target.
• Compare each arm against the parental mAb (or its Fab): an arm >~2–3× weaker than its parent points to a chain-pairing, linker, or format defect — fix that before trusting anything downstream.
• Typical expectations: tumor-antigen arms are commonly low-nM to pM; CD3 arms of T-cell engagers are often engineered in the ~10–100 nM range (reported/typical). Absolute values matter less than agreement with design intent and parental controls.
• Sanity-check the sample: SEC monomer % and absence of aggregates, or every KD is suspect.

Choose the per-arm affinity run as Decision 1 when you need to prove each arm is functional and rule out mispairing before any dual-binding claim.

Decision 2 — Simultaneous vs sequential engagement (and when it matters)

Once each arm binds, ask whether the molecule engages both targets at once.

• SPR double-injection: capture the BsAb, saturate with Target A, then inject Target B; additional response above the Target-A plateau is direct evidence of simultaneous binding.
• BLI sandwich: load the BsAb on a biosensor, dip into Target A, then Target B, watching for the second step-increase.
• A silent second injection is usually steric — epitope occlusion, a bulky ECD blocking the second paratope, or arm geometry (Fab vs scFv linker) — not a dead arm.

Whether this matters depends on the mechanism of action. Cross-linking-dependent mechanisms — T-cell engagers building the immune synapse, agonistic antibodies cross-linking receptors — require simultaneous engagement by definition. Bispecifics acting as two independent blockers (dual cytokine or dual ligand neutralization) do not; sequential data suffice.

Choose the simultaneous-engagement assay when the mechanism depends on forming a ternary complex (retargeting, receptor cross-linking); choose single-arm/sequential data when the arms act independently.

Decision 3 — Controls that prevent false conclusions (avidity, regeneration, low-density)

Three artifacts routinely produce fake "dual binding" results:

• Avidity inflation. On a dense capture surface a bivalent molecule binds two ligands at once and appears far higher affinity than any single arm has. Keep the surface low-density — a typical Rmax budget of ~30–80 RU for the analyte — and confirm the fit stays 1:1. Kinetic KD claims are only valid at low density.
• Weak controls. Always run monovalent controls (Fab, or single-arm mutants with one paratope knocked out by CDR mutation or a dead arm) to prove each signal comes from its intended arm; negative controls (isotype, single-target-only surfaces, an irrelevant ECD) to rule out non-specific binding; and parental mAbs as affinity references.
• Regeneration damage. Strip analyte with a mild condition (e.g., 10 mM glycine pH 1.5–2.5, typical for antibody–antigen pairs) and verify the capture ligand survives repeated cycles — if it does not, every later cycle quietly lies to you.

Choose low-density capture plus monovalent/single-arm controls when the decision is a kinetic KD; choose regeneration QC when running multi-sample panels where cycle-to-cycle drift would corrupt comparisons.

Decision 4 — When dual binding is not enough (escalate to functional)

Binding in range plus simultaneous engagement does not mean the molecule works. Escalate to functional assays when:

• The mechanism is engagement-dependent: T-cell engagers → T-cell activation and killing; agonistic receptors → signaling; blocking pairs → ligand-blockade or receptor-activation readouts; internalizing BsAbs (BsAb-ADC) → internalization.
• The next decision is candidate selection or an in vivo study — functional data are the gate before those commitments.

A molecule with clean dual binding but no function usually fails on geometry, epitope, or format, not affinity — redirect to engineering, not more binding runs. This matches the PRD §20 assay workflow (Binding → Dual Binding → Functional), with S03 functional screening opened only once mature delivery capability exists.

Choose escalation to functional assays when the mechanism is engagement-dependent and candidate selection or in vivo work is next; do not re-run binding assays to explain a functional failure.

Tool set this implies

A dual-binding program needs, in priority order:

This maps one-to-one onto S02 and the reagent bundles recommended on the GP pair pages (e.g., EGFR×MET: both ECDs + biotin + reference antibodies). Add functional reagents (cells, effector cells, reporter lines) only when Decision 4 triggers.

Choose this reagent set when starting dual-binding validation; add functional reagents only when Decision 4 triggers escalation.

References

• Labrijn, A.F. et al. "Bispecific antibodies: a mechanistic view of the promise and the challenge." Nat Rev Drug Discov 18 (2019). https://www.nature.com/articles/s41573-019-0028-1
• Brinkmann, U. & Kontermann, R.E. "The making of bispecific antibodies." MAbs 9 (2017). https://www.tandfonline.com/doi/full/10.1080/19420862.2017.1333797
• "SPR-based assays enable the full functional analysis of bispecific molecules." J Pharm Biomed Anal (2017). https://www.sciencedirect.com/science/article/abs/pii/S0731708516307002
• Sartorius (Octet BLI) — mAb discovery and characterization resources, including bispecific simultaneous-binding workflows. https://www.sartorius.com/en/applications/life-science-research/biologics-discovery/mab-discovery-and-development-resources/acclerating-drug-discovery-with-biomolecular-interaction-analysis-webinar
• BLI-based bispecific binding example (Octet RED96). https://pmc.ncbi.nlm.nih.gov/articles/PMC10361135/