Off-Target Toxicity

Off-target toxicity is harm arising from binding something the molecule was not designed to engage. It is the counterpart of on-target toxicity, where the adverse effect is the intended pharmacology taken too far. The separation rests on selectivity arithmetic: if a peptide binds its intended receptor at low nanomolar affinity and a related receptor a hundredfold more weakly, that margin holds only while exposure stays inside the range where the weaker site remains largely unoccupied.

The melanocortin family shows how narrow those margins can be. Receptor subtypes share a small ligand pocket, so agonists aimed at MC4R also touch MC1R, which is why pigmentary changes accompany compounds developed for other melanocortin indications. Peptides are generally cleaner than small molecules, which are routinely screened against panels of kinases, ion channels and transporters, but a closely related receptor subtype is exactly the kind of neighbour a peptide cannot easily avoid.

Deciding whether a toxicity is on-target or off-target determines what can be done about it. On-target harm follows the dose everywhere and can only be traded against benefit; off-target harm can in principle be engineered out with a more selective analogue or avoided by keeping peak concentrations lower. The same finding therefore points either at the molecule or at the exposure.

Two errors recur. The first is quoting a clean in-vitro selectivity panel as proof of safety, when the panel tested the parent molecule at one concentration and says nothing about metabolites, accumulation or tissue distribution. The second is labelling any unwanted effect off-target because it sounds like a manufacturing problem: incretin nausea is on-target, and calling it otherwise implies a purer product would remove it.