Epithelial cells in culture as a model for the intestinal transport of antimicrobial agents.

The bioavailabilities of orally administered drugs depend to a great extent on their capability of being transported across the intestinal mucosa. In an attempt to develop an in vitro model for studying the intestinal transport of drugs, we used an intestinal epithelial cell line (Caco 2) derived from a human colon adenocarcinoma. A renal epithelial cell line (MDCK) was also used to determine the tissue specificity of drug transport. These cell lines, which were grown on filters, form a monolayer of well-polarized cells coupled by tight junctions and can be used for transcellular transport experiments. We studied the transport of nine antimicrobial agents with different physicochemical and pharmacokinetic characteristics using these epithelial cell monolayers to determine whether this model could be predictive of oral bioavailability. The transepithelial passage was assayed from the apical (AP) to the basolateral (BL) side and in the opposite direction (BL to AP) in both cell lines. Radioactively labeled mannitol was used to monitor the intactness of the cell monolayer during drug passage. The results indicated that all antimicrobial agents tested tended to behave in vitro generally according to their known in vivo absorptive characteristics. In addition, the use of epithelia from different tissues enabled us to divide the drugs into four groups according to their behaviors and suggested the existence of different transport mechanisms. In particular, two antibiotics, gentamicin and teicoplanin, showed no passage in either direction or cell line, in accordance with their very poor in vivo absorbances after oral administration. In contrast, rifapentine, rifampin, and nalidixic acid passed very efficiently at similar rates in both directions and cell lines in a concentration-dependent, nonsaturable manner, which is suggestive of passive diffusion down a concentration gradient. Of the remaining drugs, isoniazid and novobiocin sodium showed some differences in passage between the two cell lines and, given their ionized state at the pH that was used, may use the paracellular route. Finally, trimethoprim and D-cycloserine exhibited differences in passage both with respect to polarity and cell line; in particular, trimethoprim had a faster rate of passage only in Caco 2 cells and in the BL to AP direction, while D-cycloserine was exclusively transported by Caco 2 cells in the AP to BL direction. In both cases it is possible that active transport mechanisms are involved.