Abstract
To mitigate risks in central nervous system (CNS) drug development, we established a high-throughput in vitro blood-brain barrier (BBB) model using LLC-PK1-MOCK and LLC-PK1-MDR1 cells in a Transwell system, aiming to replicate in vivo brain distribution and elucidate permeability mechanisms. Model integrity was assessed via transepithelial electrical resistance (TEER) and efflux functionality using control drugs (atenolol, digoxin). Bidirectional transport studies of 41 compounds quantified permeability (Papp), efflux ratios (ER), and recoveries, while in vivo brain distribution parameters (Kp,uu,brain) were derived from literature and rat studies. The model demonstrated critical BBB features: tight junction integrity (TEER > 70 Ω·cm2), P-gp efflux activity (digoxin ER = 5.10 ~ 17.12), and discrimination of passive diffusion (63.41% of drugs) from transporter-mediated mechanisms (19.5% P-gp substrates). A training set of 20 randomly selected drugs revealed a robust correlation between MDR1-derived Papp(A-B) and Kp,uu,brain (R = 0.8886), with the remaining 21 compounds validating predictive accuracy (≤2-fold error). Four alkaloids exhibiting low recovery (<80%) due to lysosomal trapping were corrected using Bafilomycin A1, aligning their permeability with in vivo outcomes. These results position the LLC-PK1-MOCK/MDR1 model as a reliable surrogate tool for early CNS drug screening, enabling rapid prioritization of candidates based on BBB penetration potential. Its integration into preclinical workflows promises to accelerate the development of therapeutics for neurological disorders.