Bispecific Antibody Formats — Structural Designs, Challenges, and Emerging Approaches

What is an epitope and why does mapping matter?

An epitope (or antigenic determinant) is the specific molecular surface on an antigen that an antibody paratope contacts. Epitopes fall into two categories: linear epitopes, formed by a contiguous stretch of amino acids, and conformational epitopes, formed by residues that are distant in primary sequence but brought together by protein folding. Most therapeutic antibodies recognize conformational epitopes because they bind the native target on cell surfaces or in circulation.

Epitope mapping is valuable at multiple stages of antibody development. During lead selection, it identifies which candidates bind distinct, non-overlapping epitopes — critical for selecting antibody pairs for sandwich ELISA or developing matched-pair immunoassays. For therapeutic programs, the epitope determines mechanism of action (receptor blocking vs. allosteric modulation) and defines freedom-to-operate relative to competitor antibodies. Regulatory agencies increasingly expect epitope data in biosimilar and bispecific antibody filings.

What are the main epitope mapping methods?

How does peptide array mapping work?

The antigen sequence is divided into overlapping synthetic peptides (typically 15-mers offset by 1–3 residues), which are spotted onto a membrane or glass slide. The antibody of interest is incubated with the array, and binding is detected by a labeled secondary antibody. Positive spots identify the minimal linear sequence the antibody recognizes. This method is fast, cost-effective, and well-suited for screening large antibody panels. However, it cannot detect conformational epitopes because the peptides are presented in a denatured, surface-bound context. An antibody that shows no binding on a peptide array likely recognizes a conformational or discontinuous epitope.

How does HDX-MS reveal conformational epitopes?

Hydrogen–deuterium exchange mass spectrometry (HDX-MS) compares the deuterium uptake of an antigen alone versus the antigen–antibody complex. Regions that are shielded by antibody binding exchange deuterium more slowly, producing a mass shift detectable by LC-MS. Because the antigen remains in solution in its native fold, HDX-MS captures conformational epitopes that peptide arrays miss. Resolution is limited to the peptic peptide level (typically 5–15 residues), but this is sufficient to define the epitope region and compare binding sites between antibody clones. HDX-MS has become the most widely used method for conformational epitope mapping in the biopharmaceutical industry because it requires modest amounts of protein, no crystallization, and can be completed in weeks rather than months.

When is structural determination (X-ray or cryo-EM) necessary?

X-ray co-crystallography of the Fab–antigen complex provides the definitive epitope map: every contact residue, hydrogen bond, and van der Waals interaction is resolved at atomic level. This data is the gold standard for patent filings and regulatory submissions but requires milligram quantities of purified complex, successful crystallization (which can take months), and significant structural biology infrastructure. Cryo-EM has emerged as an alternative for large or flexible complexes and membrane targets that resist crystallization, achieving near-atomic resolution without crystals. For most discovery-stage programs, HDX-MS or competition binning provides sufficient resolution; structural methods are reserved for advanced candidates entering IND-enabling studies.

What is epitope binning and how does it complement fine mapping?

Epitope binning groups antibodies by competitive binding: two antibodies that block each other’s binding belong to the same bin and likely share overlapping epitopes. SPR-based tandem injection experiments or BLI are the standard platforms for binning. In a classic sandwich format, antibody A is captured on the sensor, antigen is flowed to saturate binding, and then antibody B is injected. If B produces an additional signal, A and B bind non-overlapping epitopes. If no additional signal is observed, they compete for the same site.

Binning is fast enough to classify panels of 50–100+ clones and is typically the first epitope characterization step after initial screening. However, binning reveals only whether epitopes overlap; it does not identify which residues are involved. Projects that require residue-level detail proceed from binning to HDX-MS or alanine scanning for representative clones from each bin. This tiered approach — bin first, fine-map selected leads — is the most efficient strategy for large hybridoma campaigns or nanobody discovery projects.

How should I interpret epitope mapping data?