Home › Services › Avidity Engineering

Your best clone by KD is not your best clone in the assay. Is avidity the reason?

Often, yes. Intrinsic affinity and functional potency come apart whenever binding is multivalent, because a weak monovalent interaction can become extremely tight once two arms engage a clustered target — and a strong monovalent binder can be a poor clustering agent. Abwiz Bio measures both properties separately: high-affinity candidates are identified by monovalent Fab ELISA, and weak-but-high-avidity candidates by bivalent ELISA, in which Fab is pre-complexed with a secondary antibody to form a pseudo-IgG. You get an avidity-binned clone panel, so valency becomes something you design rather than something that surprises you.

Why affinity ranking misleads

Screening campaigns rank clones by KD because KD is what the instrument reports. But a bivalent IgG binding a target that is dense, clustered, or arrayed on a membrane does not behave like the monovalent measurement predicts. Once the first arm is bound, the second arm searches a tiny local volume at enormous effective concentration; rebinding dominates, apparent off-rate collapses, and the functional affinity can exceed the monovalent value by orders of magnitude.

The consequence is that the two rankings diverge. The clone with the best KD may bind poorly in a cell assay because its geometry does not allow both arms to engage. A clone discarded as mediocre in BLI may be the strongest performer on cells because it cross-links efficiently. If the whole panel was ranked on monovalent affinity alone, the functionally best molecule may already have been thrown away.

This matters most in exactly the applications where antibodies are most valuable: receptor clustering and agonism, where function requires cross-linking; avidity-driven selectivity, where discrimination between high- and low-density tissue is the therapeutic window; and immunoassays, where valency sets both sensitivity and hook effect.

How we measure the two properties separately

• Monovalent Fab ELISA. Clones are screened as soluble Fab, giving intrinsic, single-site affinity uncontaminated by any rebinding contribution. Our phagemid vector is designed for exactly this — soluble Fab can be screened in 96-well plates before the costly IgG stage.
• Bivalent pseudo-IgG ELISA. The same Fabs are pre-complexed with a secondary antibody to create a bivalent pseudo-IgG, and re-measured. The difference between the two readouts is the avidity contribution, measured rather than modelled.
• Binning by the ratio, not the value. Clones are grouped by how much they gain from valency. Some gain little; some gain enormously. That ratio, not either number alone, is what predicts behaviour in a clustered-target assay.
• Panel delivery, not single-clone delivery. Because the right point on the affinity–avidity plane depends on your biology, we deliver the binned panel so you can test the hypothesis directly rather than accept our guess about it.

All of this happens at the Fab screening stage, before IgG conversion, so the information arrives while it can still change which clones you carry forward.

Where avidity engineering changes the answer

Avidity work pairs naturally with affinity attenuation, where monovalent binding is deliberately dialled down into a target window, and with kinetic tuning, where on- and off-rate are engineered independently. Together these three services let you specify a binding profile rather than merely maximize one number. For clustering-dependent function discovered from scratch, see functional and neutralizing antibody discovery.

You receive

• Monovalent and bivalent characterization for each clone
• An avidity-binned clone panel, grouped by valency gain
• Full VH/VL sequences
• Recombinant IgG for selected clones; BLI kinetic data optional after purification
• No downstream royalties, stage-gated approval before each next step

Frequently asked questions

What is the difference between affinity and avidity in antibodies?

Affinity is the strength of a single antigen-binding site interacting with one epitope. Avidity is the combined strength when multiple sites engage a multivalent or clustered target at once. Because a bound first arm makes engagement by the second arm far more likely, apparent binding can be orders of magnitude tighter than the monovalent affinity, which is why the two rankings of a clone panel often disagree.

How do you measure avidity separately from affinity?

Clones are screened twice: as soluble monovalent Fab, which reports intrinsic single-site affinity, and as bivalent pseudo-IgG created by pre-complexing the Fab with a secondary antibody. The difference between the two readouts is the avidity contribution, and clones are binned by how much they gain from valency.

Why would I want a weaker binder?

When function depends on cross-linking or when selectivity comes from target density, a clone with modest monovalent affinity but a large avidity gain can outperform a high-affinity clone that cannot engage two sites at once. Deliberately weaker monovalent binding is also used to widen the therapeutic window in agonists and T-cell engagers, which is the subject of our affinity attenuation service.

When in a project should avidity binning be done?

At the Fab screening stage, before IgG conversion. Our phagemid is designed to allow soluble Fab screening in 96-well plates ahead of the expensive IgG step, so the avidity data arrives while it can still change which clones you advance.

Can this be applied to an antibody I already have?

Yes. Send the VH/VL amino-acid sequence and we can characterize valency behaviour directly, or build and bin a variant panel around your lead if you need to move to a different point on the affinity–avidity plane. Contact info@abwizbio.com for a scoped quote.