Aki T. Heikkinen

Analysis of Unstirred Water Layer in In Vitro Permeability Experiments

In vitro permeability experiments are used widely in drug discovery and other areas of pharmaceutical research. Much effort has been expended in developing novel epithelial models but generally much less attention has been paid to the hydrodynamic barrier in the actual experiments. The restricted liquid flow in the vicinity of solid surfaces leads to a zone where the diffusional movement of molecules exceeds the convection. This leads to formation of a concentration gradient between the bulk solution and the surface. The formed unstirred water layer (UWL) reduces the apparent permeability (P(app)) of compounds that rapidly pass through the actual epithelial layer. This lowers the resolution of P(app) versus fraction-absorbed assay, complicates the structure-permeability analysis and skews apparent kinetic parameters of transporter substrates. This review describes the techniques that can be used to determine the UWL thickness in permeability experiments and apparatuses described in the literature to control the in vitro hydrodynamics.

Application of PBPK modeling to predict human intestinal metabolism of CYP3A substrates – An evaluation and case study using GastroPlus™

First pass metabolism in the intestinal mucosa is a determinant of oral bioavailability of CYP3A substrates and so the prediction of intestinal availability (Fg) of potential drug candidates is important. Although intestinal metabolism can be modeled in commercial physiologically based pharmacokinetic (PBPK) software tools, a thorough evaluation of prediction performance is lacking. The current study evaluates the accuracy and precision of GastroPlus Fg predictions for 20 CYP3A substrates using in vitro and in silico input data for metabolic clearance and membrane permeation, and illustrates a potential impact of intestinal metabolism modeling on decision making in a drug Research and Development project. This analysis supports that CYP3A mediated metabolic clearance measured in human liver microsomes can be used to predict gut wall metabolism. Using values scaled from in vitro cell permeability as input for effective jejunal permeability resulted in good Fg prediction accuracy (no significant bias and ∼95% of predictions within 2 fold from in vivo estimated Fg), whereas simulations with in silico predicted permeability tended to overestimate gut metabolism (40% of Fg predictions under predicted more than 2 fold) ±2 fold range as an estimate of imprecision in metabolic clearance and permeability inputs propagated to >5 and <2 fold ranges of predicted Fg for compounds with <30% and >75% in vivo Fg, respectively, suggesting lower precision of predictions for high extraction compounds. Furthermore, parameter sensitivity analysis suggests that limitations in solubility or dissolution may either decrease Fg by preventing saturation of metabolism or increase Fg by shifting the site of absorption towards the colon where expression of CYP3A is low. The case example illustrates how, when accounting for the associated uncertainty in predicted pharmacokinetics and linking to predictive models for efficacy, PBPK modeling of intestinally metabolized compounds can support decision making in drug Research and Development.

In vitro-in vivo correlation in p-glycoprotein mediated transport in intestinal absorption

Oral administration is the most common route for drug administration. However, after oral administration, the absorption may be erratic and incomplete. P-glycoprotein, an efflux transporter localized in the enterocyte, limits the absorption of transported drugs extruding them back to the intestinal tract. The interaction between new drug candidates and P-glycoprotein is investigated in vitro during early stages of drug development. However, it is uncertain how well the in vitro studies actually predict the in vivo P-glycoprotein effect on the extent of oral absorption, since the in vitro and in vivo correlation has not been achieved. In the present review, the recent approaches to compare the in vitro and in vivo data are described and parameters are proposed that could be adequate for a reliable in vitro and in vivo correlation of P-glycoprotein contribution on intestinal absorption. The present article identifies an evident lack of suitable in vivo data. A significant in vitro and in vivo correlation would increase the value of in vitro studies and could reduce costs during the process of drug development.

The Asymmetry of the Unstirred Water Layer in Permeability Experiments

PurposeTo elucidate the apical and basolateral components of the total unstirred water layer in regular permeability experiment.MethodsA novel stirring apparatus was constructed to remove the basolateral unstirred water layer. Caco-2 cells were used as the permeability barrier both in Transwell-type and side-by-side apparatuses. Permeability experiments were done with several ionisable compounds at various pH and stirring conditions. The permeabilities of the cell monolayer, the unstirred water layer and the polycarbonate filter were calculated either from experimental data or theoretically.ResultsThe unstirred water layer was thicker in the Transwell apparatus than in the side-by-side chamber even in the presence of vigorous basolateral magnetic stirring. Calculations indicated that the apical unstirred water layer is thicker than the basolateral layer. Different cellular permeability coefficients were obtained from the two permeability apparatuses.ConclusionsAn orbital shaker does not produce symmetric hydrodynamics in both chambers of Transwell apparatus. The asymmetric unstirred water layer may complicate the exact analysis of polarized transport.

Quantitative ADME Proteomics – CYP and UGT Enzymes in the Beagle Dog Liver and Intestine

PurposeBeagle dogs are used to study oral pharmacokinetics and guide development of drug formulations for human use. Since mechanistic insight into species differences is needed to translate findings in this species to human, abundances of cytochrome P450 (CYP) and uridine diphosphate glucuronosyltransferase (UGT) drug metabolizing enzymes have been quantified in dog liver and intestine.MethodsAbundances of enzymes were measured in Beagle dog intestine and liver using selected reaction monitoring mass spectrometry.ResultsSeven and two CYPs were present in the liver and intestine, respectively. CYP3A12 was the most abundant CYP in both tissues. Seven UGT enzymes were quantified in the liver and seven in the intestine although UGT1A11 and UGT1A9 were present only in the intestine and UGT1A7 and UGT2B31 were found only in the liver. UGT1A11 and UGT1A2 were the most abundant UGTs in the intestine and UGT2B31 was the most abundant UGT in the liver. Summed abundance of UGT enzymes was similar to the sum of CYP enzymes in the liver whereas intestinal UGTs were up to four times more abundant than CYPs. The estimated coefficients of variation of abundance estimates in the livers of 14 donors were separated into biological and technical components which ranged from 14 to 49% and 20 to 39%, respectively.ConclusionsAbundances of canine CYP enzymes in liver and intestine have been confirmed in a larger number of dogs and UGT abundances have been quantified for the first time. The biological variability in hepatic CYPs and UGTs has also been estimated.

Kinetics of Cellular Retention during Caco-2 Permeation Experiments: Role of Lysosomal Sequestration and Impact on Permeability Estimates

The permeability estimation from cell monolayer permeation data is usually based on 100% recovery assumption. However, poor recovery is often seen in such experiments in practice but often neglected in data interpretation. In the present study, the cellular retention kinetics during Caco-2 permeation experiments of three passively transported compounds, weakly basic propranolol [(±)-1-isopropylamino-3-(1-naphthyloxy)-2-propanol], weakly acidic ibuprofen [α-methyl-4-(isobutyl)phenylacetic acid], and neutral testosterone (17β-hydroxy-4-androsten-3-one), were determined. Furthermore, the effects of cellular retention kinetics on apparent permeability were evaluated, and the role of lysosomal sequestration in cellular retention of propranolol was explored. The cellular retention profiles were observed to be direction and concentration dependent, which may cause erroneous directionality and concentration dependence in permeability estimates. Furthermore, the lysosomal sequestration was demonstrated to contribute to the extent and kinetics of the cellular retention of propranolol.

Applications of a 7-day Caco-2 cell model in drug discovery and development.

Oral delivery is the preferred route of administration and therefore good absorption after oral dosing is a prerequisite for a compound to be successful in the clinic. The prediction of oral bioavailability from in vitro permeability assays is thus a valuable tool during drug discovery and development. Caco-2 cell monolayers mimic the human intestinal epithelium in many aspects. These monolayers form tight junctions between cells and have been widely used as a model of human intestinal absorption. Caco-2 cells also express a variety of transporter proteins although the transformed nature of the cells results in unpredictable differentiation markers, transport properties and enzyme expression. Thus various modifications of the Caco-2 assay are used in laboratories across the globe. The purpose of this paper is to provide an overview of a time and resource saving 7-day Caco-2 assay protocol. We also discuss the impact of various experimental conditions on permeability measurements and its applications during lead optimization in early discovery and for clinical candidate characterization, specifically for prediction of absorption in human, at a later stage in drug development.

In Vitro Methods to Study the Interplay of Drug Metabolism and Efflux in the Intestine

This review provides an overview of the in vitro methods currently used in studies of intestinal drug metabolism and active efflux with a special emphasis on the efflux- metabolism interplay. These methods include e.g. expressed enzymes or efflux transporters, fractionated intestinal cells, cell lines, primary cells, intestinal segments and other tissue preparations. Pharmacokinetics of effluxmetabolism interplay is often very complicated, possibly involving saturation, stimulation and/or inhibition of one or both of these mechanisms. Parent drug and/or metabolite(s) can be substrates for several enzymes and/or efflux proteins. These detoxifying proteins may alter the exposure of drugs to each other and, consequently, their contributions to the overall drug elimination. Depending on the complexity of the in vitro system used, different kinds of information can be extracted from the results. Simple methods concentrating on single mechanisms provide easily interpretable information, but neglect the interplay between various mechanisms influencing the kinetics in a whole organism. More complex experimental systems mimic the mechanistic complexity of in vivo setting better, but at the same time the interpretation and utilization of the results becomes more challenging. Advantages and limitations of various in vitro systems are addressed and consideration is given to the physiological relevance of the results obtained and there is discussion of approaches for in vitro - in vivo translation of the data.

Modelling of Drug Disposition Kinetics in In Vitro Intestinal Absorption Cell Models

One major prerequisite for an orally administered drug is the ability to cross the intestinal epithelia from intestinal lumen into the blood circulation. Therefore, the absorption potential of molecules is studied early on during the drug development process. Permeation experiments using cultured cell monolayers are one of the most often applied methods to screen and also to predict in more detail the intestinal absorption potential of molecules in preclinical phase. Furthermore, these studies are also used to screen the molecules for transporter interactions as well as for more detailed mechanistic studies of the transfer routes involved. Several mathematical and computational models with complexity varying from simple non-mechanistic single barrier models to mechanistically more detailed compartmental models have been developed to describe the drug disposition during these in vitro permeation experiments. This MiniReview gives an overview of these models and their applications. Also the implications of these models to the prediction of intestinal absorption in vivo are discussed.

Effects of experimental setup on the apparent concentration dependency of active efflux transport in in vitro cell permeation experiments.

P-Glycoprotein mediated efflux is one of the barriers limiting drug absorption from the intestine. Predictions of the intestinal P-glycoprotein function need to take into account the concentration dependency because high intestinal drug concentrations may saturate P-glycoprotein. However, the substrate binding site of P-glycoprotein lies inside the cells and the drug concentration at the binding site cannot be measured directly. Therefore, rigorous determination of concentration dependent P-glycoprotein kinetics is challenging. In this study, the effects of the aqueous boundary layers, extracellular pH and cellular retention on the apparent saturation kinetics of P-glycoprotein mediated transport of quinidine in an in vitro cell permeation setting were explored. The changes in the experimental conditions caused 1 order of magnitude variation in the apparent affinity to P-glycoprotein (K(m,app)) and a 5-fold difference in the maximum effective P-glycoprotein mediated transport rate of quinidine (V(max,app)). However, fitting the concentration data into a compartmental model which accounted for the aqueous boundary layers, cell membranes and cellular retention suggested that the P-glycoprotein function per se was not altered, it was the differences in the passive transfer of quinidine which changed the apparent transport kinetics. These results provide further insight into the dynamics of the P-glycoprotein mediated transport and on the roles of several confounding factors involved in in vitro experimental setting. Further, the results confirm the applicability of compartmental model based data analysis approach in the determination of active transporter kinetics.