The influence of passage number for Caco2 cell models when evaluating P-gp mediated drug transport

Caco2 cells are a human adenocarcinoma cell line that forms tight junctions and are widely used to examine bidirectional drug transport as well as P-glycoprotein mediated efflux. Unfortunately Caco2 cell lines can be very heterogeneous in nature. Our aim was to improve the Caco2 cell model for determination of P-glycoprotein mediated drug transport. Young passage Caco2 from ATCC had inadequate expression of P-glycoprotein, therefore three approaches were adopted to upregulate Caco2 P-glycoprotein expression to mimic that in vivo; a) incubation of mature Caco2 monolayer with rifampicin, b) prolonged exposure of Caco2 cells to vinblastine (generating the Caco2 VIN line), and c) splitting cells every 7 to 9 days until late passage numbers (over P80) were available. Upon development of the models, P-gp expression and activity was determined using western blotting and bidirectional transport studies of rhodamine123. All four models exhibited P-gp mediated efflux transport for rhodamine123. Incubation with rifampicin did not alter bidirectional transport compared to passage 44 cells. Increased passage number altered P-glycoprotein expression and the efflux ratio increased to 4.7 for passage 80 from 1.4 of passage 44. The highest basolateral to apical transport was observed for both passage 89 Caco2 and the Caco2 VIN model with an efflux ratio of 13 to 14. Western blot images confirmed the increased P-glycoprotein expression of late passage and Caco2 VIN. Caco2 cells are not ready for P-gp related research when first acquired from ATCC (Passage 18). Late passage Caco2 cell monolayers or Caco2 VIN models are needed to determine P-gp mediated efflux transport.