P-Glycoprotein (P-gp) Efflux
P-glycoprotein efflux is the ATP-driven expulsion of drug molecules back out of cells by the ABCB1 transporter, and at the blood-brain barrier it is a major reason lipophilic drugs fail to reach the brain.
P-glycoprotein is an ATP-binding cassette transporter encoded by ABCB1, formerly MDR1, sitting in the luminal membrane of brain capillary endothelial cells with its business end facing the blood. It binds substrates within the lipid bilayer and uses ATP hydrolysis to expel them back into the circulation before they finish crossing. Its substrate range is unusually broad, favouring large, lipophilic, often cationic molecules, which means the barrier is an active pump as well as a set of tight junctions. The same transporter lines the intestinal epithelium, the hepatocyte canalicular membrane and the renal proximal tubule.
The defining experiments used mice lacking the gene. Knockout animals given ivermectin accumulated dramatically higher brain concentrations and died at doses their wild-type littermates tolerated, and the same collie-breed sensitivity to ivermectin traces to a natural ABCB1 deletion. The clinical corollary is loperamide, an opioid agonist that produces no central effect at normal exposure precisely because P-glycoprotein keeps it out of the brain. Digoxin serves as the standard clinical probe substrate; rifampicin induces the transporter, and inhibitors such as quinidine and ketoconazole raise substrate exposure.
This is why lipophilicity alone predicts central penetration poorly, and why efflux ratio in a transfected cell monolayer is measured alongside passive permeability during discovery. A high efflux ratio explains a compound that looks brain-penetrant on paper and is not, and it flags interaction risk with co-administered inducers or inhibitors.
The misapplication in the peptide literature is invoking P-glycoprotein to explain why a peptide does not reach the brain. Most peptides are not efflux substrates; they are excluded by size, charge and paracellular tightness. Blaming the pump implies an inhibitor would fix it, which it would not.