Development and application of a quantitative physicochemical model of P-gp substrate specificity

Area of Science:

• Pharmacology and Drug Discovery
• Computational Chemistry
• Biochemistry

Background:

• Active transport by ABC superfamily efflux pumps, like P-glycoprotein (P-gp), significantly impacts drug efficacy and distribution.
• Current computational methods for P-gp efflux prediction are often limited to substrate classification, lacking quantitative insights.
• Understanding P-gp efflux is critical for early-stage drug discovery to avoid drug-drug interactions and efficacy loss.

Purpose of the Study:

• To develop predictive models for quantitative assessment of P-gp efflux ratio (ER).
• To provide mechanistic insights into the interplay between passive diffusion and active efflux.
• To evaluate the impact of P-gp efflux on blood-brain barrier penetration and oral bioavailability.

Main Methods:

• Utilized a censored-regression statistical methodology for model development.
• Trained models on a dataset of approximately 3,500 compounds with exact or censored ER values.
• Employed fundamental physicochemical descriptors (LogP, pKa, McGowan Volume) for model interpretability.

Main Results:

• Developed predictive models yielding quantitative P-gp efflux ratio (ER) outputs.
• Models demonstrated interpretability, offering insights into passive diffusion and active efflux mechanisms.
• Successfully applied models to assess P-gp efflux effects on blood-brain barrier penetration and oral bioavailability.

Conclusions:

• The developed censored-regression models offer a quantitative approach to P-gp efflux prediction in drug discovery.
• These models provide valuable mechanistic insights and practical applications for evaluating drug transport properties.
• This methodology enhances early-stage drug candidate evaluation, improving predictions of in vivo drug behavior.