Knowing what a monoclonal antibody (mAb) is made of is only half the picture. What ultimately matters is how it behaves. A subtle structural change can alter how the antibody binds its antigen, and that difference can separate an effective therapeutic from one that underperforms in the clinic.

Established binding assays such as enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance (SPR) remain indispensable, but neither reveals which structural feature is driving a change in binding. As a result, investigating a drop in potency typically means a lengthy, multi-step workflow: chromatographic fractionation, mass spectrometry and an offline bioassay, repeated until the cause emerges.

Recent work we have done addresses this gap by turning the binding event itself into the basis for separation. Miniaturized high-performance antigen affinity chromatography (HPAC) resolves mAbs according to how they bind. In addition, it integrates directly into UHPLC and MS-based workflows.

Turning affinity into a separation

Affinity chromatography has long been a powerful way to study antibody-receptor interactions. The principle is well established, but its limitations were practical: columns were bulky, ligand consumption was high, and chromatographic efficiency fell short.

So our team miniaturized it. We immobilized HER2 on narrow-bore HPAC columns (2.1 × 20 mm) using under 100 µg of ligand while remaining fully UHPLC-compatible. The smaller format proved advantageous on every front: lower cost, markedly reduced ligand demand, higher-efficiency separations, and clean coupling to mass spectrometry.

Using trastuzumab as a well-characterized model antibody, HPAC resolved its proteoforms by HER2 binding strength rather than by charge or hydrophobicity.

Catching small differences that count

Temperature-stressed trastuzumab (which is known to lose potency) resolved into two partially separated affinity populations on HER2-HPAC. A conventional linear gradient barely distinguished them, but optimizing the binding pH and gradient brought them clearly into view.

HPAC-MS then identified the cause. It was not a high-mass modification such as oxidation or glycation, but a low-mass chemical change within the CDRs, consistent with known degradation pathways of IgG1 antibodies.

Structure-function insight, straight from multidimensional LC

The real power of HPAC emerges when it is embedded into multidimensional LC workflows.

2D CEX-HER2-HPAC assessed each charge variant functionally as it eluted, eliminating the usual offline handling. The result was unambiguous: heavy-chain Asp102 isomerization markedly reduces HER2 binding, whereas deamidation at heavy-chain Asn55 and light-chain Asn30 has little to no functional impact.

4D-LC-MS extended this further. By coupling HPAC with online desalting, reduction, and tryptic digestion, it linked each affinity difference directly to a specific sequence-level PTM, even labile modifications such as deamidation and isomerization.

Together, these workflows achieve what averaged functional readouts cannot: they identify precisely which variant is responsible while preserving the antibody’s native modification state.

Why miniaturized HPAC matters in biopharma development

Taken together, our work positions miniaturized HPAC as far more than a niche separation technique. By embedding antigen affinity directly within the LC-MS workflow, it becomes a functional separation dimension, alongside charge, size, and hydrophobicity, hereby:

• Dramatically reduced ligand consumption
• Full UHPLC and MS compatibility
• Automated workflows that minimize artifacts
• Direct identification of functionally relevant CQAs

This makes HPAC-enabled multidimensional LC-MS a scalable, sensitive approach for early-stage developability assessment, precisely the stage where material is scarce and the functional differences that determine a candidate’s progression are subtle but decisive.

Interested in a pdf of our corresponding poster? You can request your poster download here. Want to learn more about our advanced analytics? Let’s talk or subscribe to our newsletter below!