Steven D. Sheridan, PhD
Group Leader, Molecular and Cell Based Assay Development, Millipore Corp.
Sara Gutierrez
Research Scientist, Millipore Corp.
Andy Arena
Technical Service Representative, Millipore Corp.
Matthew Wilgo
Researcher ADG, Millipore Corp.
New high-throughput cell-based model systems for permeability screening of lead compounds.
Pressure has been mounting for techniques to screen the absorption, distribution, metabolism, elimination, and toxicity (ADME-tox) properties of drug candidates early in the discovery process in order to understand the pharmacokinetic properties of a drug candidate and ultimately, to predict how a compound might behave in clinical
 click to enlarge Apparent permeability (Papp) values for a panel of drug compounds were determined at 10 µM using radiometric methods. All Papp values listed have been multiplied by 1 × 106. All cell lines correctly rank-ordered all compounds based on permeability and historical classifications. The 21-day Caco-2 and 3-day MDCK-MDR1 cell monolayers both expressed P-gp to a similar degree based on efflux ratios ((B to A)/(A to B)) of digoxin and vinblastine (data not shown). |
trials. One primary screening parameter involves establishing the absorption properties of a compound. The absorption characteristics of an orally-administered drug typically are related to the drug’s permeability—its ability to cross intestinal cell barriers before being distributed throughout the body toward the site of action. This absorption can occur either passively through transcellular or paracellular diffusion or actively via carrier transport mechanisms. Absorption is not merely a function of compounds passing through cell membranes, since some cells come equipped with pumps that shunt the compounds back out of the cell again. To gain a thorough understanding of the absorption profile of a particular compound, one needs to evaluate both its influx and efflux mechanisms. Although absorption characteristics are typically assessed in animal models, there are now powerful in vitro methods available to model the absorption of chemical compounds.
Cell-based functional assays have become increasingly useful for screening the absorption properties of lead compounds. In particular, cell-based permeability models are widely used to predict the absorption rate of candidate drug compounds. The human colon adenocarcinoma line, Caco-2, is the most common cell-based model for assessing intestinal epithelial transport. This system has been demonstrated extensively to be an accurate predictor of compound permeability in vivo. Unlike artificial membrane systems that only evaluate passive diffusion of compounds across a lipid layer, Caco-2 assays provide more predictive information because the cells both selectively promote or resist the transport of compounds.
Although accurate and well-established, the Caco-2 model requires a considerable investment in both time and resources. Typically, Caco-2 cells can take up to three weeks to become fully differentiated at which time they express all relevant transporter molecules.
Another model uses Madin Darby Canine Kidney (MDCK) cells. MDCK cells are used primarily for identifying passive permeable compounds. Like Caco-2 cells, MDCK cells express the efflux pump P-glycoprotein (P-gp). Unlike Caco-2 cells that require complete differentiation in order to fully express the P-gp efflux pump, MDCK cells express significant P-gp levels in as little as three days. In order to avoid the potential for difference in the efflux pumps, the human P-gp was engineered into MDCK cells. Ira Pastan, MD, and colleagues at the National Cancer Institute (NCI) developed a genetically engineered MDCK model that over-expresses human P-gp.
The engineered MDCK system seems to provide the best of both models. Engineered MDCK cells that over-express the human P-glycoprotein pump, multi-drug resistance 1 (MDR1), provide a cell culture model system in which the polarized expression of human P-gp in integral monolayers can be obtained within one week. Combining the use of this MDCK-MDR1 cell line with 24-well or 96-well cell growth platforms provides a screening tool that generates permeability results that correlate well with the percent human absorption of many compounds without the need for long-term growth, costly specialty media, or extensive optimization.
Evaluating permeability assays
When growing cells on filter supports, it is essential to choose a good initial seeding density to ensure good cell attachment and eventual monolayer growth. Investigate a range of starting densities and evaluate each one based on monolayer formation. After cells attach to the filters, they start to grow towards confluency and form tight junctions as cell–cell interactions increase. These tight junctions initiate cell differentiation and polarization of surface proteins and transporters. Seeding densities for integral and polarized MDCK-MDR1 and Caco-2 monolayers were optimized for three and 21
 click to enlarge Western Blot analysis demonstrates MDCK cells stably transfected with the MDR1 gene product display consistently high levels of the ~170 kDa P-gp protein (A). P-gp is not detectable in wild type MDCK cells and only detectable in Caco-2 cells upon differentiation (21-day). The P-gp protein present in the 21-day Caco-2 cells appears as at least two distinct peptides (B and C). This is likely due to proteolytic cleavage of the P-gp protein at either Arg680 or Leu682 as described by S.L. Nuti and U.S. Rao [Journal of Biological Chemistry. vol. 33, pp. 29417-29423 (2002)]. |
days respectively on both MultiScreen Caco-2 96-well plates and 24-well plates from Millipore Corp., Billerica, Mass. Cell monolayer integrity and seeding density optimization were assessed by transepithelial electrical resistance (TEER). TEER measurements assess monolayer integrity by measuring the electrical resistance of the monolayer to a current set up by electrodes on either side of the monolayer on a filter membrane. TEER readings were measured and the averages were (in ohm × cm2) were: 400–500 for Caco-2 growth and 600–800 for MDCK-MDR1 growth. Combined with monolayer rejection of the paracellular dye lucifer yellow, these TEER values represented fully integral cell monolayers.
Once cell monolayer growth is optimized for integrity, apparent permeability (Papp) assays evaluate monolayer differentiation and polarization of transporters to the apical and basolateral surfaces of the cells. Efflux ratios derived from bi-directional transport data (the ratio of transport from basolateral to apical over the reverse direction) of known effluxed compounds assess the presence of efflux pump gene products such as P-gp and Human MDR1. We have used a standard set of drug compounds that includes paracellular, transcellular, and efflux. Caco-2 and MDCK-MDR1 correctly ranked all compounds based on permeability and historical classifications. More important, the Caco-2 and the MDCK-MDR1 cell monolayers both expressed P-gp to a similar degree based on high efflux ratios of digoxin and vinblastine.
Confirmation of P-gp expression in MDCK-MDR1 cells
Permeability assays rank order and flag efflux compounds. While digoxin and vinblastine are commonly used as standard markers for efflux activity with P-gp and MDR1, these compounds can be exported by other pumps if present. To correlate our efflux data to the presence of MDR1, we performed Western blot analysis using an anti-MDR1 antibody to confirm that our two cell models are expressing similar levels of MDR1. MDCK-MDR1 gene products display consistently high levels of ~170 kD P-gp protein after only three days with the same amount of protein expressed in Caco-2 cells after 21 days. Immunohistochemistry and Calcein AM uptake experiments were also performed, and both suggest the same similarity of MDR1 expression in cell monolayers. For these assays, MDCK-MDR1 and Caco-2 cells were grown on 1µm PET membrane, which allows for cell visualization through the membrane. In summary, the MDCK cell lines stably transfected with the human MDR1 gene offer a fast and effective alternative to the 21-day Caco-2 screening model by increasing throughput and decreasing material and resource costs.
This article was published in Drug Discovery & Development magazine: Vol. 10, No. 2, February, 2007, pp. 41-43.
Filed Under: Drug Discovery
