Antibody specificity is actually a fluid concept
Polyreactivity refers to the ability of an antibody to bind, with varying but biologically relevant affinities, to multiple structurally unrelated antigens. Well-documented, it remains incompletely understood in both natural and therapeutic antibodies. In the natural repertoire, polyreactivity is associated predominantly with IgM natural antibodies produced by B-1 cells. Their polyreactivity is physiologically relevant and beneficial1.
The picture changes substantially when therapeutic mAbs are considered. Usually, they are derived after immune responses involving somatic hypermutation and affinity maturation. A substantial fraction of these somatically mutated antibodies retains off-target binding under physiological conditions.
Cross-interaction chromatography (CIC) – retention time on an affinity column prepared from serum bulk IgG. Initially designed as a correlate of antibody solubility. Later, the retention times were shown to also correlate with clearance times. Ultimately, it is used to screen for mAb developability, encompassing all possible interactions: rheumatoid factor-like, membrane-binding, and anti-idiotypic.
Baculovirus particles (BVP) Assay – an ELISA-based method measuring the binding to baculovirus particles - a test developed almost by chance in the course of using the baculovirus expression system to screen for antibodies against expressed target antigens. BVPs provide an inert membrane without key charged targets but with hydrophobic binding capacity and an envelope protein that is highly glycosylated with high-mannose glycans (a pathogen-associated molecular pattern).
Polyspecificity reagent (PSR) binding assay – A flow-cytometry-based assay measuring the capacity of yeast-expressed antibodies to bind biotinylated fragments of membranes from CHO cells. Unlike BVP, this membrane contains many of the antibody targets found on mammalian membranes.
MAb self-interaction by bio-layer interferometry (CSI-BLI) - a high-throughput method to detect antibody clone self-interaction (CSI) using BLI technology. Self-interaction causes high viscosity and aggregation, which, by themselves, are undesirable. Aggregates also tend to be sticky.
Affinity-Capture Self-Interaction Nanoparticle Spectroscopy (AC-SINS) and Salt-Gradient Affinity-Capture SINS (SGAC-SINS)– Self-association of antibody-covered gold nanoparticles red-shifts the adsorption spectra with a version in a salt gradient. These are other very sensitive self-association assays, with SGAC-SINS detecting aggregation-prone hydrophilic antibodies.
Melting temperature of the Fab by Differential Scanning Fluorimetry. A low melting temperature increases the exposure of novel binding sites, especially hydrophobic patches, in the Fab of mAbs, thereby promoting not only polyreactivity but also self-aggregation.
Standup monolayer chromatography (SMAC) and accelerated stability chromatography slope use size-exclusion chromatography to detect the tendency for aggregation.
Hydrophobic interaction chromatography (HIC) – another method for measuring propensity for hydrophobic interactions.
ELISA for a small number of structurally highly diverse antigens (e.g., cardiolipin, KLH, ssDNA, dsDNA, insulin, etc.). This is a traditional method that tests explicitly polyreactivity rather than “stickiness”, using a very limited set of target antigens.
If polyreactivity were just an indiscriminate stickiness, the traditional methods would have much greater correlation than observed2. The fact that they moderately correlate with each other and are all necessary as part of a diverse array of tests is evidence that polyreactivity is a phenomenon encompassing multiple forms of antibody interaction with multiple structures. There is no consensus on how the currently used assays should be combined, nor on the criteria for predicting clinical problems3,4. A systematic screening of approved and clinical-stage mAbs using a proteome-scale platform (binding assay with 6172 extracellular human proteins) found that 28% exhibited at least one confirmed off-target interaction3. Most of these interactions were related to epitope mimicry rather than “stickiness”.
Maybe it would be more useful to update the concept of antibody specificity5:
Specificity emerges from a continuum of affinities, and the distinction between monospecific and polyreactive behavior depends on arbitrary thresholds as well as the antigenic landscape.
The antibody repertoire is selected to avoid a range of self structures6 that are orders of magnitude fewer than the potential epitopes space. The GC reaction optimizes a single scalar quantity — affinity for the epitope of the immunizing antigen6.
Loss of polyreactivity may be a structural epiphenomenon of paratope rigidification, but is not an independent selection objective. It is not a necessary consequence of SHM7,8.
Autoreactivity checkpoints also impose just a boundary condition (self-tolerance) rather than optimizing an independent objective9,10.
Thus, specificity is never directly selected for. When it emerges, it is a consequence of the affinity driven geometric optimization on a particular paratope.