P-glycoprotein: the pump that keeps drugs out of the brain

You can design a molecule that hits its target at picomolar potency, crosses membranes beautifully, and still watch it fail completely in brain metastases. The usual culprit is not the target and not permeability — it is a molecular bouncer called P-glycoprotein. P-gp sits in the blood-brain barrier and pumps drugs back out into the blood faster than they can diffuse in. For oncology programs chasing CNS disease, it is one of the most underappreciated reasons a good compound never reaches the tumor.

What P-glycoprotein is

P-glycoprotein (gene ABCB1, also called MDR1) is an ATP-driven efflux transporter in the ABC family. It threads its substrate out of the cell membrane and hydrolyzes ATP to power the stroke, working like a one-way pump. It is expressed on the luminal (blood-facing) side of brain capillary endothelial cells, in the gut epithelium, and in the kidney and liver. In the brain, P-gp and its partner transporter BCRP (ABCG2) form the active half of the blood-brain barrier: even a lipophilic molecule that should diffuse across will be grabbed and ejected before it accumulates.

P-gp is also gloriously promiscuous. Its binding cavity is large and flexible, and it recognizes a broad swath of mid-sized, moderately lipophilic, often hydrogen-bond-rich molecules — which describes a large fraction of kinase inhibitors.

The efflux ratio, and what counts as a problem

The standard readout is the efflux ratio from a bidirectional permeability assay across a polarized monolayer (Caco-2 or MDCK-MDR1 cells). You measure apparent permeability in both directions and divide:

• Efflux ratio = Papp(B→A) / Papp(A→B). A ratio near 1 means the compound moves equally both ways — not a substrate. A ratio above ~2 flags active efflux; above ~3 is a real liability for CNS targets.
• Confirm with an inhibitor. Repeat the assay with a P-gp inhibitor (elacridar, zosuquidar) added. If the efflux ratio collapses toward 1, P-gp was responsible. If it does not, suspect BCRP or another transporter.
• Read it against the target. For a peripheral tumor a high efflux ratio may not matter, and gut P-gp can even be saturated at clinical dose. For a brain target it is often decisive, because the barrier transporters are hard to saturate.

Oncology drugs that P-gp shuts out of the brain

• Ibrutinib — ABCB1 markedly restricts its brain penetration; transporter-deficient mice show roughly 5-fold higher brain-to-plasma ratios, while CYP3A separately caps its oral exposure.
• Vemurafenib — the BRAF V600E inhibitor is a substrate of both P-gp and BCRP, with brain-to-plasma ratios about 20-fold higher in mice lacking both pumps, and the efflux can be reversed with elacridar.
• Palbociclib — the CDK4/6 inhibitor is a dual P-gp/BCRP substrate; brain exposure rises roughly 100-fold in transporter-knockout models, which is why it underperforms against intracranial tumors despite strong systemic activity.

The contrast that makes the point is osimertinib, which was deliberately optimized for low efflux and consequently delivers real intracranial activity. Low P-gp liability is not a footnote — for a CNS-directed program it can be the whole ballgame.

How you fix it (and why it is hard)

Designing out P-gp efflux usually means reducing the count of hydrogen-bond donors and acceptors and trimming topological polar surface area, since P-gp recognition tracks with H-bonding capacity. The tension is that those same features often drive solubility and potency, so efflux optimization is a balancing act rather than a clean win. Co-dosing a P-gp inhibitor works in mice but has repeatedly disappointed in the clinic, because you cannot selectively unlock the brain barrier without also raising systemic and gut exposure.

Try the prediction yourself

Liganx’s ADMET panel scores P-glycoprotein substrate likelihood alongside permeability, BBB penetration, and the other transport properties, so you can spot an efflux liability before committing synthesis. Read it together with the blood-brain barrier penetration call: a compound that looks permeable but flags as a P-gp substrate is the classic CNS trap.

Open Studio and dock a candidate, then open the ADMET pill on the result row to see the transport profile. Liganx brings molecular docking online into the browser and runs the ADMET panel on every pose, so you can pair the molecular docking score with the efflux readout and catch a brain-barrier problem before it costs you a campaign.

Primary sources

• P-Glycoprotein (MDR1/ABCB1) restricts brain penetration of the Bruton’s tyrosine kinase inhibitor ibrutinib, while CYP3A limits its oral bioavailability. Mol Pharm 15, 5103–5113 (2018). doi:10.1021/acs.molpharmaceut.8b00702
• Durmus S, Sparidans RW, Wagenaar E, et al. Oral availability and brain penetration of the BRAF-V600E inhibitor vemurafenib can be enhanced by the P-glycoprotein (ABCB1) and breast cancer resistance protein (ABCG2) inhibitor elacridar. Mol Pharm 9, 3236–3245 (2012). PMID:23020847
• Parrish KE, Pokorny J, Mittapalli RK, et al. Efflux transporters at the blood-brain barrier limit delivery and efficacy of the CDK4/6 inhibitor palbociclib (PD-0332991) in an orthotopic brain tumor model. J Pharmacol Exp Ther 355, 264–271 (2015). PMC4613960