Home › Services › Affinity Attenuation / De-tuning

What if your antibody binds too well?

It is a real failure mode, not a paradox. In agonist antibodies, T-cell engagers and other formats where the therapeutic effect depends on a controlled interaction rather than maximal occupancy, excessive affinity narrows the therapeutic window: it drives on-target toxicity in healthy tissue, causes overstimulation, blocks serial engagement, and prevents tissue penetration. Abwiz Bio de-tunes affinity deliberately, selecting the combined variant library across all fast and slow on- and off-rate conditions at once so that stronger and weaker clones are captured simultaneously, then narrowing to the affinity window your application actually needs. You receive a ranked panel of affinity variants rather than a single clone.

Why more affinity stops helping

Antibody engineering has a strong default direction, and for a neutralizing antibody or a diagnostic reagent it is usually the right one. But several important formats have an optimum, not a maximum.

• T-cell engagers. Potency depends on serial engagement — one T cell killing many targets. An arm that binds too durably stays attached to a single target cell instead of moving on. Excessive affinity on the CD3 arm also drives cytokine release and redistributes the molecule into T-cell-rich tissue rather than tumour.
• Agonist antibodies. Agonism is a geometry problem. Signalling depends on how the receptor is clustered and for how long, and an antibody that binds too tightly can lock the receptor in a non-productive arrangement or overstimulate it. The functional optimum frequently sits well below maximal affinity.
• Therapeutic index in shared antigens. When the target is present on healthy tissue at lower density, moderate affinity plus avidity can create selectivity that a very high-affinity binder destroys — the tight binder saturates both compartments equally.
• Tissue penetration. Very slow release produces a binding-site barrier: the molecule saturates the first cells it meets and never distributes further into the tissue.

In each case the requirement is a number in a range, and the range is set by biology you can measure. That makes it an engineering target.

How we de-tune deliberately

Reducing affinity by guesswork — mutating contact residues and hoping to land in the window — usually damages specificity or stability on the way down. Selection is more controllable.

• Bidirectional selection. We run the combined library across all fast and slow on- and off-rate conditions, capturing stronger and weaker clones in the same campaign rather than pushing in one direction. This produces a continuum of binding strengths from a single library.
• Narrowing to the window. From that continuum we select the band your application requires, so the delivered clones are distributed around your target rather than clustered at one extreme.
• Specificity preserved. Every clone is still selected on your antigen, so weaker binding is achieved without loosening epitope specificity — the failure mode that makes hand-designed de-tuning mutations unreliable.
• Biophysical filters throughout. Thermostability and polyreactivity filters run at every stage, so an attenuated clone does not arrive with new developability problems.
• Ranked panel delivery. The panel is ranked by binding strength by ELISA, so you can test across the window and let your functional assay choose the clone rather than committing to a predicted optimum.

Designing a binding profile, not a number

These are the same platform run under different selection pressures, which is why they can be combined in one campaign: a T-cell engager arm can be attenuated, kinetically tuned and developability-filtered together rather than in three sequential projects.

You receive

• Sequences of the affinity-variant clone panel
• Ranking of the panel by binding strength (ELISA)
• Comparison against the parent antibody
• Recombinant IgG for selected clones; BLI kinetic data optional after purification
• No downstream royalties, stage-gated approval before each next step

Frequently asked questions

Why would anyone want a lower-affinity antibody?

Because several formats have an affinity optimum rather than a maximum. T-cell engagers depend on serial engagement, which an arm that binds too durably prevents; agonist antibodies depend on receptor geometry that excessive affinity can lock into a non-productive state; selectivity between diseased and healthy tissue can come from moderate affinity plus avidity, which a very tight binder destroys; and slow release produces a binding-site barrier that blocks tissue penetration.

How do you reduce affinity without losing specificity?

By selection rather than by design. Clones are still panned against your antigen, so epitope specificity remains a survival condition while binding strength is varied. Hand-designed de-tuning mutations often loosen specificity or destabilize the molecule, because the mutation is chosen for its effect on one contact rather than tested against the whole requirement.

Can you hit a specific KD window rather than just going weaker?

Yes, that is the point of the method. Selecting the combined library across fast and slow on- and off-rate conditions simultaneously produces a continuum of binding strengths, from which we narrow to the band your application needs and deliver a ranked panel spanning it.

What do you deliver?

Full sequences for the affinity-variant clone panel, ranked by binding strength by ELISA and compared against the parent, with recombinant IgG for selected clones and BLI kinetics available after purification.

Can attenuation be combined with other engineering?

Yes. Affinity attenuation, kinetic tuning, avidity engineering and developability filtering run on the same platform under different selection pressures, so a single campaign can address several axes at once instead of running sequential projects. Send your sequence and target window to info@abwizbio.com for a scoped quote.