Conformation-selective antibodies

In aggregation biology the interesting target is usually a species rather than a sequence: the oligomer, the fibril, the misfolded conformer. Since all of them share a sequence with the abundant, harmless monomer, a useful antibody has to discriminate on conformation. That is a real and achievable goal, and it is also one of the most over-claimed properties in the field.

Where selectivity comes from

Epitope exposure. A surface buried in the monomer and exposed in the aggregate, or the reverse, gives genuine conformational discrimination for a single binding site.

Repeat spacing. A fibril presents the same epitope every few angstroms along its length. A bivalent antibody can engage two copies at once, which a monomer cannot offer. Much of what is reported as conformational selectivity is this: avidity for a repeating array. It is real and useful, and it is a different mechanism with different consequences, since it depends on valency and on the geometry of the array.

Composite epitopes. Residues from two adjacent subunits form a surface that exists only in the assembled state.

Knowing which of the three you have changes how the molecule will behave in tissue, how it should be formatted, and whether a monovalent version will work at all.

How to raise them

Immunize or select against the aggregate, then deplete hard against monomer, in solution, at a concentration high enough to be a real counter-selection. Doing the depletion on a surface is weaker, because surface-bound monomer is itself partly ordered.

Selecting on the fibril and counter-selecting on monomer in alternating rounds works better than doing all the positive selection first. For display campaigns, sorting on the ratio of aggregate binding to monomer binding, rather than on aggregate binding alone, is the version that produces selective binders.

Proving it

The claim is a ratio, so the proof has to be a ratio measured in one format, on one day, with both species present at known concentrations:

• Titrate against monomer and against aggregate in the same assay, and report both curves.
• Do it in a monovalent format as well as the final one. If selectivity vanishes when valency is removed, say so, because that is avidity.
• Run a competition: does excess monomer displace binding to the aggregate?
• Confirm the aggregate preparation. Selectivity for heparin-induced recombinant fibrils may not extend to the polymorphs in tissue.
• Where possible, test on tissue, since that is the only material that contains the real thing.

A statement of the form "binds oligomers but not monomer" without both curves, both formats and the preparation described is not a measurement, and it will not survive contact with someone else's material.

The part people skip

Monomer is present at far higher concentration than the aggregate in vivo. A binder with tenfold selectivity is mostly bound to monomer at physiological ratios. The selectivity that matters is whatever exceeds the concentration ratio you expect to face, and that number should be estimated before a program commits, not after.