P-glycoprotein differentially affects escitalopram, levomilnacipran, vilazodone and vortioxetine transport at the mouse blood-brain barrier in vivo

P-glycoprotein (P-gp)-mediated brain efflux of xenobiotics is a well-known process, which may result in suboptimal target engagement and consequently reduced efficacy of drugs exerting their therapeutic effects in the central nervous system. In the present study the role of P-gp in transport across the blood-brain barrier (BBB) was investigated with a series of newer antidepressants (levomilnacipran, vilazodone and vortioxetine) and a control substrate (escitalopram) using P-gp knock-out (KO) and P-gp competent wild-type (WT) mice. Brain and plasma exposure time-courses were measured after an acute subcutaneous dose and at steady-state obtained after subcutaneous drug infusion by osmotic minipumps. Following acute dosing, the brain-to-plasma KO/WT exposure enhancement ratios ((AUCbrain ko/AUCplasma ko)/(AUCbrain WT/AUCplasma WT)) were 5.8 (levomilnacipran), 5.4 (vilazodone), 3.1 (escitalopram) and 0.9 (vortioxetine), respectively. At steady-state, assessment of Kp,uu (unbound brain concentrations/unbound plasma concentrations) revealed a restriction in the brain distribution in WT mice for all compounds except vortioxetine. Levomilnacipran exhibited the most pronounced efflux with a Kp,uu-value of 0.038 in WT mice which was increased to 0.37 in KO mice. Based on both the acute and steady-state distribution data, the results suggest that levomilnacipran, vilazodone and escitalopram are susceptible to P-gp mediated efflux at the BBB in vivo in mice, whereas vortioxetine was practically devoid of being affected by P-gp in vivo. The functional impact of the drug transport-controlling role of P-gp at the BBB was demonstrated by in vivo cortical serotonin transporter occupancy of vilazodone, which exhibited a 20-fold higher plasma EC50 in WT mice compared to KOs.