Plasma Protein Binding

Drug in plasma sits in equilibrium between a bound and an unbound form, and only the unbound fraction can leave the capillary, reach a receptor, be filtered at the glomerulus or be metabolised. Albumin, present at roughly 600 micromolar and itself surviving about three weeks in circulation, is the main partner for acidic and lipophilic drugs. The unbound fraction is a ratio rather than a rate, so a highly bound drug is not a slow-acting one; it is one whose active concentration is a small share of the measured total.

In peptide design the binding is the point, not a side effect. Liraglutide carries a C16 fatty acid and is more than 98 percent albumin-bound; semaglutide carries a C18 diacid on a spacer and exceeds 99 percent. The bound pool is too large to filter and is shielded from degradation, so albumin acts as a circulating reservoir releasing free peptide continuously. Much of the half-life gap between native GLP-1 and these analogues is exactly this.

The practical reading is that total concentrations are comparable only between drugs with similar binding, and that a change in albumin, whether from uraemia or inflammation, moves total concentration more than free concentration. For restrictively cleared drugs the free level tends to be restored by a compensating change in clearance, which is why total concentrations mislead in precisely the patients who are sickest.

The classic displacement interaction, in which one bound drug evicts another and causes toxicity, is largely a teaching artefact for such drugs. The peptide-specific error runs the other way: reading high protein binding on a datasheet as evidence that a compound is sequestered and inactive, when for acylated analogues it is the mechanism of their duration.