Utility of 96 well Caco-2 cell system for increased throughput of P-gp screening in drug discovery

Abstract

The use of Caco-2 cells for screening of discovery compounds for their permeability characteristics and P-glycoprotein interactions is well established and used routinely in pharmaceutical industries world-wide. The screening model involves growing cells on 12 or 24 well transwell format. In this manuscript, we report the use of Caco-2 cells grown on 96 well transwell plates for screening compounds for their potential to interact with P-gp. Bi-directionality studies were performed with known P-gp substrates such as saquinavir, indinavir, vinblastine, vincristine, verapamil, digoxin and taxol. P-gp inhibition studies were also conducted using radiolabeled digoxin as the probe. The results demonstrated that P-gp substrates had efflux ratios (P_c (B to A)/P_c (A to B)) in the 96 well format that were comparable to the ratios seen in 12 and 24 well format. Inhibition of digoxin efflux transport in presence of the test compounds (P-gp substrates) demonstrated that 96 well cells express adequate amounts of efflux transporters and perform as well as the 12 and 24 well Caco-2 cells. Thus, the 96 well Caco-2 cell set-up presents a higher throughput permeability model capable of identifying compounds that interact with P-gp and has the potential to significantly increase the efficiency of P-gp screening in early drug discovery.

Introduction

For a new chemical entity (NCE) to become a successful drug there are a multitude of desirable characteristics it should possess: potency to a biological target, selectivity, good stability and physico-chemical properties, minimal toxicity and adequate ADME profile. There are diverse screens that are set-up in order to select the compounds with the most favorable ‘developability’ characteristics with regards to ADME. Good permeability through intestinal membranes that can lead to adequate systemic absorption is a property highly desirable in NCEs. Various assay systems are used in early discovery to screen the permeability properties of compounds [1]. Commonly used permeability assay systems include cell-based models such as Caco-2, MDCK, LLC-PK1 cells; tissue-based models such as Ussing chamber, single pass perfusion; whole animal models such as in vivo screens; and even non-empirical in silico models. Combinatorial chemistry and other advances in synthetic chemistry have led to a tremendous inflow of discovery compounds being fed into the screens for permeability assessment. Some key characteristics desirable in a permeability screen are: high efficiency, high accuracy, time, cost and space effectiveness, and capability for high throughput. Caco-2 cells grown on 12 or 24 well transwell plates have been the staple of the pharmaceutical industry for high-throughput permeability of discovery compounds. Since P-gp is known to play a significant role in the pharmacokinetics of compounds, bi-directional studies in Caco-2 cells are routinely done for identifying compounds that interact with P-gp.

Caco-2 cells are a sturdy cell line that have been used to predict permeability and absorption of compounds in human [2], [3], [4]. Traditionally, these cells are grown on 12 and 24 wells and have been extremely successful in screening discovery compounds. However, advances in the field of molecular biology and combinatorial chemistry have shifted the bottle-neck in drug discovery from the compound synthesis step to compound selection step. There is a tremendous challenge to improve the throughput of existing screening techniques to keep up with compound synthesis. Increasingly a lot of companies have incorporated sophisticated levels of automation into these assays to make it amenable to higher throughput. The throughput is normally a staggering 100s and even 1000s of compounds per week through this screen. However, there are a large number of reasons why even further miniaturization (i.e. 96 well Caco-2) would help in such permeability assays. Apart from the obvious increase in throughput it would lead to tremendous cost reduction (by virtue of decreased cost of media, plates and buffer). One other key advantage over 12 or 24 well set-up is also that much less discovery compound would be required for 96 well to perform the same assay. This can be a significant advantage keeping in mind the hectic pace at which diverse chemotypes are synthesized in early stages of discovery.

One typical problem that has been encountered during miniaturization of any assay has been the increased variability and progressive drop in the ‘quality’ of data generated because of increased ‘quantity’. In general, the levels of expression of transporters in cell models are known to be expressed more consistently in higher surface area wells (6 or 12 well) compared to smaller wells. Since the objective of this study was to optimize the use of 96 well Caco-2 cells, it was important to demonstrate the following week over week: (1) P-gp (and other efflux transporters such as MRP, BCRP and LRP) were consistently expressed to equivalent levels and (2) cell monolayer maintained their integrity. Studies done with reference compounds (metoprolol, cimetidine, mannitol, digoxin) demonstrated that the cells were functional and maintained their utility between passage numbers 40 and 80 (data not included in the manuscript).

The use of 96 well Caco-2 cell system that can perform as well as the 12 or 24 well Caco-2 can significantly increase the productivity needed in early drug discovery. Moreover, the fact that the 96 well Caco-2 cells retains adequate as well as consistent expression of efflux transporter (i.e. P-gp) make them an attractive model for P-gp screening.