Chapter 26
Permeability Methods 26.1 INTRODUCTION Permeability is a key ADME (absorption, distribution, metabolism, and excretion) property that has an influence on in vitro drug discovery experiments (e.g., cell-based activity and ADME assays) and on in vivo pharmacokinetics, efficacy, and toxicity. Permeability tools can be applied strategically to the different drug discovery stages to efficiently provide the data needed to effectively impact each stage. Permeability methods should predict or measure the velocity of passage of a drug through a biological lipid membrane. Software predictions can be used early as a component in selecting among the large number of high-throughput screening (HTS) hits and in reviewing potential new compound synthesis. Since permeability has different mechanisms, it is useful to measure and understand the contribution of the different mechanisms in transport processes (e.g., passive diffusion, efflux, uptake) using in vitro methods and techniques. This knowledge informs medicinal chemistry design and structure modification to improve permeability. Uptake permeation can be helpful when attempting to increase exposure using influx transporters. Permeability values are used as input parameters to predict pharmacokinetics and model drug-drug interaction using physiologically based pharmacokinetics (PBPK) models.
26.2 COMPUTATIONAL PREDICTION OF PERMEABILITY There are a wide range of computational options with various levels of sophistication and cost to predict permeability. The purpose is to keep the drug discovery project focused on chemical space that has favorable probability of discovering a clinical candidate with good membrane permeability.
26.2.1
Permeability Prediction Using Structural Property Rules
As discussed in Chapter 4, several articles have described the structural properties of the drug-like property space [1–3]. These rules encompass properties related to good oral absorption, which is part of the drug-like concept. The properties include lipophilicity, hydrogen bonding, topological polar surface area (TPSA), molecular flexibility, and molecular size. See Section 4.4 for a compilation of the proposed ranges of these properties for well-absorbed drugs.
26.2.2
Permeability Prediction Using In Silico Methods
Computational models for intestinal drug absorption have been reported [4]. Several commercial software packages are available for the prediction of gastrointestinal (GI) absorption. A partial list of software is found in Table 26.1. As with all software, the user should evaluate the software using the compounds they are familiar with and for which internal data are available before purchasing and implementing. Permeability is a more complex process than lipophilicity and there are only a limited number of measurements of human intestinal absorption on which to develop algorithms. Thus, the predictions should be taken as a guide, but not over-interpreted. As with other software, predictions are likely best for comparing compounds in a series on a relative scale, as opposed to setting expectations for an exact value in vivo. One application of predictive permeability tools is for medicinal chemists to study the potential GI absorption effects of various substituents on a core scaffold during design phase. Another example of a permeability model used AlogP98 lipophilicity and TPSA descriptors (from Cerius2, Molecular Simulations). The model predicted well for the categories of >90% absorbed by passive intestinal absorption, 30-90% absorbed, and <30% absorbed [5]. Well-absorbed (>90%) compounds grouped in an ellipse in the region of (95% con˚ 2. Quantitative structure-activity relationship (QSAR) methods fidence limit) AlogP98 ¼ 1 to 5.9 and PSA ¼ 0 to 132 A Drug-Like Properties. http://dx.doi.org/10.1016/B978-0-12-801076-1.00026-5 Copyright © 2016 Elsevier Inc. All rights reserved.
325
326
Drug-Like Properties
TABLE 26.1 Partial List of Commercial Software for Permeability Name
Company
Website
Discovery Studio™
Accelrys
www.accelrys.com
Percepta™
Advanced Chemistry Development
www.acdlabs.com
COSMOquick™
COSMOlogic
www.cosmologic.de
Volsurf + ™
Molecular Discovery
www.moldiscovery.com
QikProp™
Schrodinger
www.schrodinger.com
ADMET predictor™
Simulations Plus
www.simulations-plus.com
for Caco-2, parallel artificial membrane permeability assay (PAMPA), PAMPA-BBB (PAMPA for blood-brain barrier), and human intestinal absorption have also been described [6,7]. The use of commercial computational tools for permeability model development has been discussed [8–10]. For example, the use of VolSurf descriptors to develop a permeation model has been described [11]. Commercial software models available for computationally predicting permeability in human intestine and in vitro monolayer methods, such as Caco-2 and MDCK, are listed in Table 26.1.
26.3
IN VITRO PERMEABILITY METHODS
Two major permeability methods, cell monolayer (e.g., Caco-2, MDCK) and PAMPA, are widely used in the pharmaceutical industry. Two other methods, liposomal and immobilized artificial membrane (IAM), are applied in some situations. Most of the methods model transcellular passive diffusion through lipid membranes, with the exception of cell monolayer assay, which also measures paracellular and transporter-mediated processes.
26.3.1
Liposomal Permeability Method
Liposomes are vesicles formed with a lipid bilayer membrane. They have been used to investigate passive trans-bilayer permeation rates, from outside to inside the liposome, under various conditions, such as different lipid compositions. The use of liposomes to determine passive trans-bilayer permeation kinetics (as opposed to passive transcellular double bilayer permeation of Caco-2) has been described [12]. pH-sensitive fluorophores are used. One is on the outside of the liposome and detects proton release from a basic cation as it reacts: BH+ ¼ B + H+, and B partitions into the lipid bilayer. The other pH-sensitive fluorophore is in inside and detects the reaction: B + H+ ¼ BH+, as the permeant depletes protons in the liposome lumen to establish equilibrium. Fluorosome-trans is a commercially available tool (see Table 26.2) that uses liposomes and fluorescence detection for permeability testing. The test compound permeates the liposome bilayer membrane by passive diffusion and then quenches the fluorescence of an internal proprietary fluorophore. The derived permeability values have good correlation to human fraction absorbed (Fa).
26.3.2
IAM High-Performance Liquid Chromatography Permeability Method
IAM is a convenient method because it uses the common HPLC format. Instead of octadecyl groups covalently bonded to the solid support, as in reversed phase HPLC, the IAM technique uses phospholipids bonded to the solid support. These contain the polar head groups and aliphatic side chains of the lipids. Pidgeon developed this new HPLC phase concept [13–22]. Test compounds partition between the aqueous mobile phase and the phospholipid phase. The chromatographic capacity factor (k) increases with increasing affinity for the phospholipid phase. Compounds are rank-ordered by k, which indicates a higher lipophilicity or phospholipid affinity. The parameters of this affinity correlate with permeation. Retention time is calibrated against permeability or absorption using standard compounds for which permeability or absorption has been measured using another technique. The HPLC format is commonly available in drug discovery laboratories, is convenient to use, and is readily automated using autosamplers to save scientist’s time. The correlation to permeability may be better than log D, because the lipids are
Permeability Methods Chapter 26
327
TABLE 26.2 Partial List of Commercial Instruments and Products for Permeability Method
Technology
Product Name
Company
Website
Caco-2, MDCK
Cells
Caco-2, MCDK cells
ATCC
www.atcc.org
®
Caco-2, MDCK
Plates
Transwell , Costar™
Corning
www.corning.com
Caco-2, MDCK
Plates
MultiScreen™
Millipore
www.millipore.com
Caco-2, MDCK
TEER
EVOM2, SYS-REMS
World Prec. Instr.
www.wpiinc.com
PAMPA
Pre-coated Plates
PAMPA plate system
Corning Gentest
www.corning.com
PAMPA
PAMPA system
Evolution™, Explorer™
Pion Inc.
www.pion-inc.com
PAMPA
Plates, lipid, buffer
PAMPA
Pion Inc.
www.pion-inc.com
IAM
IAM columns
Immobilized Art. Memb.
Regis Tech.
www.registech.com
TFC BioSciences
TFCbio.com
Beckman Coulter
www.beckman.com
Liposome
Liposomes
Caco-2, MDCK, PAMPA
Robotics
®
Fluorosomes ®
Biomek FX ®
Caco-2, MDCK, PAMPA
Robotics
Microlab Star
Hamilton
www.hamiltonrobotics.com
Caco-2, MDCK, PAMPA
Robotics
Freedom EVO
Tecan
www.tecan.com
close in structure to biological membrane composition. IAM columns are commercially available (Table 26.2). Because IAM only involves physicochemical interaction with the phospholipids, and not the polarity transitions and molecular volume space constraints of the lipid bilayer of a biological membrane, it is less predictive of permeability than some other methods. Traditionally, IAM uses an isocratic mobile phase and, therefore, has long retention times for highly lipophilic compounds. Valko et al. [22] developed a gradient IAM method that has higher throughput. Advantages of IAM are that it requires very little material and impurities do not interfere with the permeability prediction.
26.3.3
Caco-2 Monolayer Permeability Method
The first practical method for in vitro permeability assessment in discovery was the Caco-2 monolayer permeability method. This assay models the intestinal epithelial monolayer permeability barrier. Caco-2 is the most widely known and used cell line for permeability assessment [23–35]. Caco-2 has favorable characteristics as a model of human intestinal absorption. It is an immortal human colon adenocarcinoma cell line that is readily available from ATCC (Table 26.2). When cultured on a semipermeable membrane, the cells adhere to the membrane and grow to confluence, forming tight junctions. Caco-2 develops microvilli on the apical surface that resemble the morphology of GI epithelial cells that line the intestinal villi. Caco-2 cells differentiate to express certain cell membrane transporters, such as P-glycoprotein (P-gp), breast cancer resistance protein (BCRP), and multidrug resistance protein 2 (MRP2), in a polarized manner. This provides the opportunity to investigate passive transcellular and multiple efflux and uptake transporter permeability mechanisms. The strong correlation of Caco-2 monolayer permeability (Pc) and Fa after oral administration in humans for 21 compounds is shown in Figure 26.1 [8]. There are also less favorable characteristics of Caco-2. The cell line is not genetically homogeneous and contains multiple cell types. This can lead to divergence of the Caco-2 cultures with subsequent passages and between laboratories, even in the same institution. This will produce different Caco-2 values for different laboratories. Other factors can also cause differences between laboratories: culturing conditions (e.g., serum source, frequency of media changes) and practice (e.g., experimental apparatus, %dimethyl sulfoxide (DMSO), media components, type of transwell plate). Thus, while trends are likely to agree, the actual permeability values can vary among laboratories. Another limitation of Caco-2 is that 21-25 days are required for the Caco-2 cells to grow to confluence and differentiate to full transporter expression. This consumes considerable effort for cell culture work. It also increases the chances of bacterial contamination of the cultures. A shorter 5-day culture technique [29] is commercially available, but care must be taken to validate in-house full transporter functionality.
328
Drug-Like Properties
100 80
FA
60 40 20 0 –7
–8
–6 –5 log Pc (cm/s)
–4
–3
FIGURE 26.1 Relationship of Caco-2 cell monolayer permeability (Pc) and fraction absorbed (FA) after oral administration in humans for 21 compounds [8]. (Reprinted with permission from P. Stenberg, U. Norinder, K. Luthman, P. Artursson, Experimental and computational screening models for the prediction of intestinal drug absorption, J. Med. Chem. 44 (2001) 1927–1937. Copyright American Chemical Society.)
Caco-2 also has differences from small intestine enterocytes. The tight junction pores are smaller in Caco-2, thus, it is not a reliable indicator of paracellular permeability leading to under-prediction of human absorption for paracellular transport. For lower molecular weight (MW) polar compounds, paracellular investigations should use another approach (e.g., portal vein-cannulated (PVC) rat in vivo). In addition, Caco-2 has very low metabolic activity, so it is not a good model of combined permeability and metabolism effects in the intestine. Caco-2 also overexpresses P-gp compared to enterocytes, so efflux can be higher than with enterocytes.
26.3.3.1 General Protocol for Monolayer Permeability Cells are stored under liquid nitrogen. Careful attention is given to using cells of similar passage number to maintain comparability. After thawing they are grown in flasks before plating, then seeded into transwell inserts, shown in Figure 26.2. Plates of 12, 24, or 96 wells are used. The cells settle onto a porous filter support. They are cultured for a period of time appropriate to the cell type (i.e., 21-27 days for Caco-2, 5-7 days for MDCK), with media changes about every 3 days. They grow to confluence and cover the surface of the support. Monolayer integrity is checked before and after the assay with transepithelial electrical resistance (TEER), which checks for gaps that would allow the compounds an unimpeded path. If the TEER value is low, there will be much higher levels of compound transport through gaps in the monolayer. Alternatively, Lucifer yellow is added to each well to detect gaps (detection of Lucifer yellow uses a plate reader). Quality control (QC) compounds are included in the compound set to assure assay performance for low permeability (e.g., atenolol, mannitol) and high permeability (e.g., metoprolol). Test compounds are added to buffer (e.g., Hanks balanced salt solution (HBSS) with HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid)) containing glucose to make a specific concentration (e.g., 1 mM). The same test compound should be run in multiple wells (e.g., three replicates) to improve statistical reliability. Depending on the direction of permeation for the experiment, the test compound is added to a selected compartment at time 0 that is called the “donor” compartment and blank buffer is added to the other compartment that is called the “receiver.” Adding test compound to the apical compartment measures apical (A) to basolateral (B) permeability (A > B). Adding to the basolateral compartment measures B > A permeability. Typically, the pH is buffered in each compartment at pH 7.4, so as to avoid artificial efflux (of basic compounds to the apical compartment) or artificial uptake (of acidic compounds to the basolateral compartment). pH gradient can also be used across the membrane to mimic physiological conditions [donor of pH 6.5 (small intestine pH) and receiver of pH 7.4 (blood pH)]. The transwell plates are LC/MS/MS
Cell monolayer
Apical (A)
Basolateral (B)
A>B
Insert Semipermeable membrane B>A
FIGURE 26.2 Cell monolayer permeability method diagram.
Permeability Methods Chapter 26
329
incubated at 37 °C in a cell culture incubator for transport studies using various time points (e.g., 1-2 h). For optimal measurements of highly permeable compounds (e.g., for use with in vivo predictions, for distinguishing between compounds) mixing is recommended using an orbital shaker to reduce the unstirred water layer (UWL, see Section 26.3.6.2). The maximum time of the experiment is set such that <10% of the test compound mass permeates into the receiver, so that “sink” conditions are in effect and the net forward permeation (donor to receiver) is not reduced by the backward permeation (receiver to donor). A small volume of sample of the apical and basolateral compartments is taken at the beginning, middle, and final time points. Alternatively, the inserts are moved to new basolateral compartments at each subsequent time point. Samples are measured using liquid chromatography/mass spectrometry (LC/MS/MS). The effective permeability is calculated as Papp ¼
dMr , dt A CD ð0Þ
where dMr is the mass of compound appearing in the receiver, dt is the time in seconds, A ¼ surface area of the cell monolayer, and CD(0) is the concentration in the donor at time ¼ 0 [36]. Flux across the monolayer (dMr/dt) can also be calculated from the slope of the plot of amount transported versus time. The A > B data provide a value for permeability in the absorptive direction.
26.3.3.2 Additional Considerations for Caco-2 Studies Here are some additional considerations and experimental variations for Caco-2. These can add information content and insight for data application. Caco-2 permeability values differ among laboratories; however, it is sometimes useful to have a benchmark for comparison. Here is one set of permeability ranges used for Caco-2: Papp < 2 106 cm/s
Low permeability
2 106 < Papp < 20 106 cm/s
Moderate permeability
Papp > 20 106 cm/s
High permeability
Efflux transport can be ascertained by performing the assay in both A > B and B > A directions and calculating the efflux ratio (ER): Papp, B>A ER ¼ : Papp, A>B If ER 1, then the compound primarily permeates by transcellular passive diffusion. If the rates are significantly different, then a transporter may be involved. An ER > 2-3 indicates that the compound is affected by an efflux transporter that is expressed by the Caco-2 cells. If a specific transporter is suspected, the experiment can be repeated with an inhibitor that is specific for that transporter. If the Papp and ER values change with the co-incubated inhibitor, then that is strong evidence for the identity of the transporter that is involved. These experiments are discussed in greater detail in Section 27.3.1. In general, it is understood that Caco-2 over expresses P-gp and under expresses uptake transporters relative to in vivo small intestine enterocytes and this should be considered in interpreting the data. Test compound concentration selection can affect the results. Many laboratories have used 1-10 mM concentration. Alternatively, it has been argued that the concentration of compound in the GI lumen following an oral dose is closer to 50-100 mM and some laboratories use this concentration. At this concentration, transporters are likely to be saturated. Thus, there might be differences in the measured permeability between low and high concentrations. If the test compound is an efflux transport substrate, the ratio of permeability by transporters relative to passive diffusion is lower as the concentration is increased and transporters are saturated. It is important, therefore, to specify the concentration of the experiment. Higher concentrations do not properly model other permeability barriers (e.g., BBB), where the free drug concentration is much lower (nM level) and the effects of transporters are much greater (e.g., P-gp efflux at the BBB). There are two types of pH conditions used in monolayer assays. The first uses different aqueous buffer pH in the apical and basolateral compartments to simulate pH differences across the epithelial layer. Acidic pH (e.g., pH 5-6.5) is used in the apical compartment and neutral pH 7.4 in the basolateral compartment. This models the upper small intestine, where the intestinal lumen is at an acidic pH. The pH-gradient condition is applied to predict oral absorption. For measuring ER, isopH in both compartments is used. The different conditions are diagrammed in Figure 26.3.
330
Drug-Like Properties
(a)
(b)
pH 7.4
pH 5.5
Solubilizer
Caco-2
Caco-2
Caco-2
pH 7.4
pH 7.4
Sequestering material
Iso-pH
pH Gradient
FIGURE 26.3 The monolayer assay has been run under various conditions that include: pH of the aqueous buffer, pH gradient, solubilizer, and sequestering materials.
Solubilizer, sequestering material
Another experimental variation is the use of solubilizers in the buffer. These help to solubilize low solubility compounds during the experiment to produce a measurable and more accurate value for permeability. Materials used for this purpose include bile salts in the 1-5% range. The use of fasted state simulated intestinal fluid (FaSSIF) in the apical compartment has been reported [30,31]. Sequestering materials can be used in the receiver compartment. These compounds sequester the compound once it has passed through the permeability barrier and are meant to model the conditions in the blood compartment. Materials used for this purpose include bovine serum albumin (BSA). Such conditions are commonly termed “sink conditions.” Another consideration is to determine the recovery. This is the total of test compound in both compartments at the end of the assay versus at the beginning. Low recovery can result from precipitation owing to low solubility, accumulation of compound in the Caco-2 cells, binding to the plate plastic, or metabolism by the cells. Recovery below 50% produces questionable data and permeability tends to be underpredicted. Caco-2 can be used to classify the permeability of a compound as high or low permeability for the Biopharmaceutics Classification System (BCS). If the Caco-2 permeability of the compound compares to a high permeability drug run in the same experiment, then the compound can be classified as high permeability. The common high permeability comparator drug is metoprolol (since it is known to be 95% absorbed from the GI). In previous years, Caco-2 was low throughput and consumed considerable resources. Long manual cell maintenance for 21-25 days, 24 transwell plates, manual pipetting, and LC/MS analysis that was as time-consuming as a full in vivo bioanalysis study were common. With experience and higher throughput technology advances, the resources needed have declined and throughput has improved. Such advances include the common use of 96-transwell plates, 96 pipette liquid handlers, robotics for automated cell maintenance and transport assays, automated MS/MS method development, and short run times of ultraperformance liquid chromatography (UPLC). Table 26.2 is a partial list of vendors of current supplies and instrumentation for monolayer permeability methods. Table 26.3 is a partial list of commercial laboratories that provide permeability services. TABLE 26.3 Partial List of Commercial Laboratories for Permeability Services Service
Company
Website
Caco-2, MDCK
Absorption Systems
www.absorption.com
Caco-2
Agilux
agiluxlabs.com
PAMPA
Analiza
www.analiza.com
Caco-2
BioReliance
www.bioreliance.com
Caco-2, MDCK
Charles River
www.criver.com
Caco-2, MDCK, PAMPA
Cyprotex
www.cyprotex.com
PAMPA
Nextar
www.aminolab-pharma.com
PAMPA
Pion Inc.
www.pion-inc.com
Caco-2
Solvo Biotech
www.solvobiotech.com
Caco-2
Wolfe Laboratories
www.wolfelabs.com
Permeability Methods Chapter 26
26.3.4
331
MDCK Monolayer Permeability Method
Another cell line that is widely used for the monolayer method, because of its advantageous characteristics, is MadinDarby Canine Kidney (MDCK). Before its use in ADME studies, MDCK was used in biological laboratories. MDCK is now actively used in drug discovery for passive diffusion permeability measurement [37–39]. One advantage is that MDCK cells grow to confluence, with tight junctions, and are ready for permeability studies in 3-5 days, thus minimizing maintenance resources and chances for contamination. It also has low expression levels of transporters and metabolic enzymes and is reproducible from day to day and lab to lab in both 24- and 96-transwell plates. Two MDCK cell types have been used, MDCKI and MDCKII. One difference is that the TEER values are higher for type I than for II [26]. For a test group of 55 compounds, MDCK and Caco-2 had a good correlation (r2 ¼ 0.79) to each other and they had a similar correlation to human Fa values (Spearman’s Rank Correlation Coefficient rs ¼ 0.58, 0.54 for MDCK and Caco-2, respectively) [39]. Another advantage is that MDCK cells can be transfected with human genes. These lines are stable and reproducible for side-by-side experiments with wild type cells, thus allowing comparison of permeation rates with and without the presence of the transporter. Many cell lines have been developed that express human transporters [40], including the multidrug resistance protein 1 (MDR1)-MDCKII cell line with P-gp. MDCKII-WT (wild type) or MDR1-MDCKII with a transporter inhibitor are used as control, depending on the purpose of the experiment. This makes the transporter-expressing cell lines very useful for diagnosis of a specific efflux transporter and study of the permeation effects of an efflux transporter on permeation of an individual compound. They can also be used to detect and study compounds that are efflux transporter inhibitors that might affect the disposition of coadministered compounds that are substrates for the efflux transporter. See Chapters 9 and 27 for in-depth discussion of transporters.
26.3.4.1 MDCKII-LE Monolayer Permeability Method Despite the low transporter expression levels in MDCK cells, a low to moderate ER has been observed for some compounds owing to non-human canine efflux transporter expression. A better measurement for passive transcellular diffusion would be performed if the canine transporters were attenuated. A low efflux (LE) MDCKII cell line, MDCKII-LE, was developed for this purpose [36]. It was produced by selecting MDCKII-WT cells with low P-gp function, using fluorescence-activated cell sorting (FACS) and calcein efflux labeling, through multiple passages, and selecting for the 1% of cells with the lowest P-gp function. The MDCKII-LE cell line was characterized using canine P-gp mRNA real-time polymerase chain reaction (PCR), which showed 200-fold reduction in canine P-gp compared to MDCKII-WT, and using bidirectional monolayer transport assay, which showed no ER for 60 compounds that had ERs from 1.5 to 8.5 in MDCKII-WT. Having very low P-gp level also enables use of these cells for a cassette assay approach, because there is no interference from saturation of transporters by cassette components. In other applications, this cell line should produce transporter-transfected MDCK lines with better performance due to the greatly reduced background P-gp expression. In addition, the MDCKII-LE cell line has been shown to provide comparable passive diffusion permeability data to passive diffusion into human hepatocytes [41], which is an important contributor to hepatic clearance, thus reducing the cost and increasing the speed of data access.
26.3.5
Monolayer Permeability Method with Other Cell Lines
The LLC-PK1 cell line was derived from porcine kidney proximal tubule epithelial cells. It also forms monolayers in transwell plates. It has also been transfected with human MDR1 and other transporter and metabolizing enzyme genes for studying drug disposition [42,43]. Another cell line better mimics paracellular permeation than Caco-2. 2/4/A1 cells are derived from fetal rat intestine epithelia. They form adherent monolayers in 5-7 days after plating. They effectively mimic human intestinal paracellular permeation [44,45]. In addition, they express low levels of some transporters and no levels of many transporters. They correlate well with human intestinal absorption for incompletely absorbed lower MW hydrophilic drugs. Therefore, 2/4/A1 can be used as a model of passive transcellular and paracellular permeation for more in-depth study of selected compounds. HT-29 is a mucus-producing goblet cell line that forms an adherent monolayer with higher paracellular than Caco-2. The line has been used alone and in co-culture with Caco-2 cells.
332
Drug-Like Properties
26.3.6
PAMPA—Parallel Artificial Membrane Permeability Assay
The parallel artificial membrane permeability assay, PAMPA [46], has lower cost and higher throughput for in vitro permeability screening. Instead of a layer of living cells, the PAMPA membrane is made of phospholipids (e.g., phosphatidyl choline, egg lecithin) solubilized in a long chain hydrocarbon (e.g., dodecane). A diagram of the PAMPA permeability experiment is shown in Figure 26.4. Variations of the PAMPA method have been reported [46–62].
96 Well filter plate
Acceptor
Buffer
FIGURE 26.4 Parallel artificial membrane permeability assay (PAMPA) diagram.
Lipid in filter pores (e.g., 2% phosphatidyl choline in dodecane) 96 Well plate Donor
Drug in buffer (25 µg/mL) Stirring bar
26.3.6.1 General Protocol for PAMPA Permeability The test compound is diluted in aqueous buffer (“donor solution”) and is placed in the donor compartment of a 96-well plate (e.g., 200 mL). A concentration of about 25 mg/mL (about 50 mM) is commonly used. A 96-well filter plate, which has a porous filter on the bottom, is placed on top of the donor plate so that it is in contact with the aqueous buffer. A few microliters (e.g., 4 mL) of a solution of artificial membrane fluid (e.g., 20% egg lecithin in dodecane) is placed onto the top of the porous filter and it soaks down into the pores of the filter to completely fill them and form the artificial membrane barrier that touches the buffer in the donor compartment. Blank buffer (e.g., 200 mL) is placed in the wells of the filter plate, on top of the artificial membrane and is called the “acceptor” compartment. This “sandwich” of 96-well plate and filter plate is maintained at a constant temperature and humidity for between 1 and 18 h, depending on the laboratory protocol and the permeability of the compounds. (As with the cell monolayer method, the initial rates, well before equilibrium is reached across the membrane, provide the best measurement of permeability.) Samples are then taken from the acceptor compartments, the filter plate is removed, and samples are taken from the donor compartments. The concentrations of compounds in the compartments are quantitated using an LC/MS, LC/UV, or an ultraviolet (UV) plate reader. The unused “donor solution,” that was not placed in the donor compartments, is used as a 100% standard for quantitating the concentration of compound in the donor and acceptor wells. The effective permeability (Peff) is calculated as discussed for the monolayer permeability method.
26.3.6.2 Additional Considerations for PAMPA Studies The PAMPA method measures only passive diffusion through a lipid membrane, but, as discussed in Chapter 8, this is the most important permeation mechanism for intestinal absorption of drugs. PAMPA also provides a way to evaluate this passive diffusion alone without other mechanisms occurring, whereas in Caco-2, specific permeability mechanism values require multiple experiments. PAMPA has been shown to correlate to human jejunal permeability with approximately the same correlation as Caco-2 [50]. An example of this relationship is shown in Figure 26.5. Thus, for projecting ahead to in vivo absorption, PAMPA provides a high-throughput approach that has adequate predictability at earlier stages of drug discovery. Advantages of PAMPA are lower cost and higher throughput. The artificial membrane is created at the time of the experiment, so there are no cell culture maintenance costs. A UV plate reader can be used for test compound quantitation, which reduces the price and increases the throughput compared to LC/MS/MS used in the monolayer method. PAMPA also has the advantage of greater flexibility of experimental conditions to customize the experiment for a particular research question. One variation of conditions is that buffers of various pHs can be used in the donor and acceptor. It is common to use the same pH on each side of the membrane, but to simulate the intestine, neutral pH might be used in the acceptor compartment and an acidic pH might be used in the donor compartment. This is helpful for pHs that are not well tolerated for Caco-2 cells, such as low pH. Solubilizing components, that might be toxic to cells, can be added to the PAMPA buffers for low solubility compounds. Sequestering materials can be used in the acceptor compartment, to simulate the “sink” conditions in the intestine. Also, the lipid components in the artificial membrane can be varied. Examples include 2% phosphatidyl choline in dodecane [50], 20% egg lecithin in dodecane [46], hexadecane alone
Permeability Methods Chapter 26
333
FIGURE 26.5 Prediction of human jejunal permeability using PAMPA. Used with kind permission of Dr. Alex Avdeef.
[56], and brain lipid in PAMPA-BBB (see Chapter 28). Pion Inc. offers a proprietary lipid mixture that has increased correlation of the PAMPA data to human intestinal absorption. Stirring of the donor is another variation that can be used [50]. The UWL (see Chapter 8), which acts as a barrier to permeation in static PAMPA, is greatly reduced when the donor compartment is stirred, so the PAMPA experiment can be completed in less time (e.g., 1 h rather than 10 h). Laboratories have also used various thicknesses of membrane by using filter plates with different filter membrane thicknesses [47,56]. Thinner membranes can reduce the time of the experiment, because the molecules have a shorter distance to diffuse. All of these variations affect the data, but they also demonstrate the flexibility of PAMPA to be modified to model a specific set of conditions of interest. It should be noted that the PAMPA membrane is not a bilayer and that it is much thicker (e.g., 125 mm) than a lipid bilayer (5 nm) or cell monolayer (20 mm). The microstructure of the PAMPA membrane is not certain, but likely to be multilamellar, disordered, and dependent on the composition.
26.3.7
Comparison of Caco-2 and PAMPA Methods
The widespread implementation of PAMPA in the pharmaceutical industry has prompted questions about how it compares to other methods and data. The previous section discussed the operational advantages of PAMPA. Figure 26.6 diagrams the data comparison of PAMPA and Caco-2 when Caco-2 is run at a low concentration that does not saturate the transporters. The Peff absolute values of all methods vary with the experimental conditions. However, for compounds that are primarily permeable by transcellular passive diffusion, the Caco-2 and PAMPA data fall along the central correlation line [62]. Compounds having strong P-gp efflux tend to plot in the lower right section of relatively higher
Uptake Caco-2 Peff
Passive diffusion
Efflux
PAMPA Peff FIGURE 26.6 Comparison of Caco-2 and PAMPA permeability values [62]. (Reprinted with permission from E.H. Kerns, L. Di, S. Petusky, M. Farris, R. Ley, P. Jupp, Combined application of parallel artificial membrane permeability assay and Caco-2 permeability assays in drug discovery, J. Pharm. Sci. 93 (2004) 1440–1453. Copyright 2004 Elsevier Science Ltd.)
334
Drug-Like Properties
PAMPA permeability than Caco-2 permeability. Compounds having active uptake transport tend to plot in the section (upper left) of relatively higher Caco-2 permeability compared to PAMPA.
26.4
IN-DEPTH PERMEABILITY METHODS
26.4.1
Ussing Chamber
When it is desirable to obtain more in-depth measurement of permeation by a compound or across an intact ex vivo animal or human tissue, the Ussing Chamber is used. Intact tissue membranes are dissected from the organ and species of interest. The membrane is clamped into a device in a polarized (apical/basolateral) direction between compartments containing aqueous buffer that can be controlled for temperature and other variables. As in the cell monolayer method, the test compound is added to either the apical or basolateral compartment and the concentration change over time in the other compartment is measured [63,64]. Despite the more elaborate experimental setup, differences are still observed between in vivo and Ussing Chamber data [63].
26.4.2
Cannulated In Vivo Hepatic Portal Vein
Investigators can cannulate the hepatic portal vein to measure the concentration of drug before first-pass liver metabolism. The hepatic portal vein drug concentration is determined by the intestinal absorption and intestinal metabolism. Bioavailability (F) is the product of fraction of dose absorbed (Fa) times fraction of dose not metabolized in the intestine (Fg) times fraction of dose not metabolized in liver (Fh). Thus, portal vein cannulation provides a method of separating intestinal absorption and metabolism from liver metabolism [65].
26.4.3
Perfusion In Vivo Methods
The human intestinal permeation values of a limited number of drugs have been measured using regional single-pass perfusion of the jejunum [63]. A multichannel tube with two distal latex balloons placed 10 cm apart is inserted via the esophagus through the stomach and into the intestine. Once in place, the balloons are inflated to isolate the intestinal section, and drug solution is delivered to the isolated section. The method used is called Loc-I-Gut. The results provide a valuable database of human absorption under controlled conditions for various research and method validation purposes. The in situ perfusion technique in living animals is most used during late drug discovery and development. A selected intestinal segment is exposed by dissection of an anesthetized animal and isolated by clamping. A drug solution is passed through this section of the living animal via syringe pump and the rate of absorption is measured. Such studies are performed rarely in discovery, but are more common in development, where they provide data for BCS studies that are submitted to the U.S. Food and Drug Administration (FDA) (see Section 7.2.3.1) [63].
26.4.4
In Vivo Pharmacokinetics Method
Of ultimate interest are Fa and bioavailability. These are derived from an in vivo dosing and bioanalysis. These methods are discussed in Chapter 37. An overview of the major permeability methods is in Table 26.4. TABLE 26.4 Methods for the Determination of Permeability Method
Type Assay
Speed (min/cpd)
Throughput (cpd/day/instrument)
IAM
High throughput
10
120
PAMPA
High throughput
0.5
200
Monolayer-MDCK
Moderate throughput
10
120
Monolayer-Caco-2
Moderate throughput
10
120
In situ intestinal perfusion
Low throughput
250
2
Hepatic portal vein cannulation
Low throughput
250
2
Permeability Methods Chapter 26
335
26.5 APPLICATIONS OF PERMEABILITY IN DRUG DISCOVERY (1) Review screening hits and compounds planned for synthesis using permeability software to be alerted early about predicted permeability limitations. (2) For hits of interest and all newly synthesized compounds, obtain high-throughput in vitro permeability assay data (e.g., PAMPA, MDCK) to assess the passive diffusion permeability. (3) If compounds have a much lower activity in cell-based biological assays than receptor and enzyme assays, test whether permeability may be causing the reduction. (4) Apply structure modification strategies to improve permeability for new compounds in the chemical series. (5) Obtain monolayer permeability data for selected chemical series examples using a cell line that contains efflux transporters to assess potential efflux limitations. (6) Compare in vivo PK screening data and in vitro permeability data on a selected chemical series example to assess whether unexpectedly poor pharmacokinetics might be due to a permeability issue (e.g., efflux). (7) Use Caco-2 for BCS categorization for regulatory filing.
PROBLEMS (1) For the following permeability methods, which permeation mechanisms contribute to the measured permeability value? Method
Passive Diffusion
Uptake
Efflux
IAM PAMPA Caco-2
(2) (3) (4) (5) (6)
What is convenient about IAM for permeability estimation? What factors add to the expense of Caco-2 compared to PAMPA? What additional information can be obtained from Caco-2 compared to PAMPA? How do IAM HPLC columns differ from other reversed phase HPLC columns? Compare Caco-2 and PAMPA permeability data for the following compounds:
Compound’s Permeability Mechanism(s)
PAMPA Relatively Higher than Caco-2
Caco-2 Relatively Higher than PAMPA
PAMPA and Caco-2 Relatively the Same
Passive diffusion only Passive diffusion and uptake Passive diffusion and efflux
REFERENCES [1] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings, Adv. Drug Deliv. Rev. 23 (1997) 3–25. [2] D.F. Veber, S.R. Johnson, H.Y. Cheng, B.R. Smith, K.W. Ward, K.D. Kopple, Molecular properties that influence the oral bioavailability of drug candidates, J. Med. Chem. 45 (2002) 2615–2623. [3] A.K. Ghose, V.N. Viswanadhan, J.J. Wendoloski, A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases, J. Comb. Chem. 1 (1999) 55–68. [4] T. Hou, J. Wang, W. Zhang, W. Wang, X. Xu, Recent advances in computational prediction of drug absorption and permeability in drug discovery, Curr. Med. Chem. 13 (2006) 2653–2667. [5] W.J. Egan, K.M. Merz, J.J. Baldwin, Prediction of drug absorption using multivariate statistics, J. Med. Chem. 43 (2000) 3867–3877. [6] C. Hansch, A. Leo, S.B. Mekapati, A. Kurup, QSAR and ADME, Bioorg. Med. Chem. 12 (2004) 3391–3400. [7] R.P. Verma, C. Hansch, C.D. Selassie, Comparative QSAR studies on PAMPA/modified PAMPA for high throughput profiling of drug absorption potential with respect to Caco-2 cells and human intestinal absorption, J. Comput.Aided Mol. Des. 21 (2007) 3–22. [8] P. Stenberg, U. Norinder, K. Luthman, P. Artursson, Experimental and computational screening models for the prediction of intestinal drug absorption, J. Med. Chem. 44 (2001) 1927–1937.
336
Drug-Like Properties
[9] H. Lennernaes, Human intestinal permeability, J. Pharm. Sci. 87 (1998) 403–410. [10] S. Winiwarter, N.M. Bonham, F. Ax, A. Hallberg, H. Lennernaes, A. Karlen, Correlation of human jejunal permeability (in vivo) of drugs with experimentally and theoretically derived parameters. A multivariate data analysis approach, J. Med. Chem. 41 (1998) 4939–4949. [11] G. Cruciani, M. Pastor, W. Guba, VolSurf: a new tool for the pharmacokinetic optimization of lead compounds, Eur. J. Pharm. Sci. 11 (Suppl. 2) (2000) S29–S39. [12] K. Eyer, F. Paech, F. Schuler, P. Kuhn, R. Kissner, S. Belli, P.S. Dittrich, S.D. Kra¨mer, A liposomal fluorescence assay to study permeation kinetics of drug-like weak bases across the lipid bilayer, J. Controlled Release 173 (2014) 102–109. [13] C. Pidgeon, S. Ong, H. Liu, X. Qiu, M. Pidgeon, A.H. Dantzig, J. Munroe, W.J. Hornback, J.S. Kasher, IAM chromatography: an in vitro screen for predicting drug membrane permeability, J. Med. Chem. 38 (1995) 590–594. [14] S. Ong, H. Liu, C. Pidgeon, Immobilized-artificial-membrane chromatography: measurements of membrane partition coefficient and predicting drug membrane permeability, J. Chromatogr. A 728 (1996) 113–128. [15] C.Y. Yang, S.J. Cai, H. Liu, C. Pidgeon, Immobilized artificial membranes—screens for drug-membrane interactions, Adv. Drug Deliv. Rev. 23 (1997) 229–256. [16] S. Ong, H. Liu, X. Qiu, G. Bhat, C. Pidgeon, Membrane partition coefficients chromatographically measured using immobilized artificial membrane surfaces, Anal. Chem. 67 (1995) 755–762. [17] H. Liu, S. Ong, L. Glunz, C. Pidgeon, Predicting drug-membrane interactions by HPLC: structural requirements of chromatographic surfaces, Anal. Chem. 67 (1995) 3550–3557. [18] B.H. Stewart, O.H. Chan, Use of immobilized artificial membrane chromatography for drug transport applications, J. Pharm. Sci. 87 (1998) 1471–1478. [19] J.A. Masucci, G.W. Caldwell, J.P. Foley, Comparison of the retention behavior of b-blockers using immobilized artificial membrane chromatography and lysophospholipid micellar electrokinetic chromatography, J. Chromatogr. A 810 (1998) 95–103. [20] G.W. Caldwell, J.A. Masucci, M. Evangelisto, R. White, Evaluation of the immobilized artificial membrane phosphatidylcholine. Drug discovery column for high-performance liquid chromatographic screening of drug-membrane interactions, J. Chromatogr. A 800 (1998) 161–169. [21] F. Beigi, I. Gottschalk, C. Lagerquist Hagglund, L. Haneskog, E. Brekkan, Y. Zhang, T. Osterberg, P. Lundahl, Immobilized liposome and biomembrane partitioning chromatography of drugs for prediction of drug transport, Int. J. Pharm. 164 (1998) 129–137. [22] K. Valko, C.M. Du, C.D. Bevan, D.P. Reynolds, M.H. Abraham, Rapid-gradient HPLC method for measuring drug interactions with immobilized artificial membrane: comparison with other lipophilicity measures, J. Pharm. Sci. 89 (2000) 1085–1096. [23] I.J. Hidalgo, T.J. Raub, R.T. Borchardt, Characterization of the human colon carcinoma cell line (Caco-2) as a model system for intestinal epithelial permeability, Gastroenterology 96 (1989) 736–749. [24] P. Artursson, J. Karlsson, Correlation between oral drug absorption in humans and apparent drug permeability coefficients in human intestinal epithelial (Caco-2) cells, Biochem. Biophys. Res. Commun. 175 (1991) 880–885. [25] P. Artursson, R.T. Borchardt, Intestinal drug absorption and metabolism in cell cultures: Caco-2 and beyond, Pharm. Res. 14 (1997) 1655–1658. [26] A. Braun, S. Hammerle, K. Suda, B. Rothen-Rutishauser, M. Gunthert, S.D. Kramer, H. Wunderli-Allenspach, Cell cultures as tools in biopharmacy, Eur. J. Pharm. Sci. 11 (Suppl. 2) (2000) S51–S60. [27] P. Artursson, K. Palma, K. Luthman, Caco-2 monolayers in experimental and theoretical predictions of drug transport, Adv. Drug Deliv. Rev. 46 (2001) 27–43. [28] I.J. Hidalgo, Assessing the absorption of new pharmaceuticals, Curr. Top. Med. Chem. 1 (2001) 385–401. [29] P.V. Balimane, K. Patel, A. Marino, S. Chong, Utility of 96 well Caco-2 cell system for increased throughput of P-gp screening in drug discovery, Eur. J. Pharm. Biopharm. 58 (2004) 99–105. [30] F. Ingels, S. Defermec, E. Destexhe, M. Oth, G. Van den Mooter, P. Augustijns, Simulated intestinal fluid as transport medium in the Caco-2 cell culture model, Int. J. Pharm. 232 (2002) 183–192. [31] L. Fossati, R. Dechaume, E. Hardillier, E. Chevillon, C. Prevost, S. Bolze, N. Maubon, Use of simulated intestinal fluid for Caco-2 permeability assay of lipophilic drugs, Int. J. Pharm. 360 (2008) 148–155. [32] S. Chong, S.A. Dando, R.A. Morrison, Evaluation of Biocoat intestinal epithelium differentiation environment (3-day cultured Caco-2 cells) as an absorption screening model with improved productivity, Pharm. Res. 14 (1997) 1835–1837. [33] L.-S.L. Gan, D.R. Thakker, Applications of the Caco-2 model in the design and development of orally active drugs: elucidation of biochemical and physical barriers posed by the intestinal epithelium, Adv. Drug Deliv. Rev. 23 (1997) 77–98. [34] B. Press, D. Di Grandi, Permeability for intestinal absorption: Caco-2 assay and related issues, Curr. Drug Metab. 9 (2008) 893–900. [35] I. Hubatsch, E.G.E. Ragnarsson, P. Artursson, Determination of drug permeability and prediction of drug absorption in Caco-2 monolayers, Nat. Protoc. 2 (2007) 2111–2119. [36] L. Di, C. Whitney-Pickett, J.P. Umland, H. Zhang, X. Zhang, D.F. Gebhard, Y. Lai, J.J. Federico 3rd, R.E. Davidson, R. Smith, E.L. Reyner, C. Lee, B. Feng, C. Rotter, M.V. Varma, S. Kempshall, K. Fenner, A.F. El-Kattan, T.E. Liston, M.D. Troutman, Development of a new permeability assay using low-efflux MDCKII cells, J. Pharm. Sci. 100 (2011) 4974–4985. [37] M.J. Cho, D.P. Thompson, C.T. Cramer, T.J. Vidmar, J.F. Scieszka, The Madin-Darby canine kidney (MDCK) epithelial cell monolayer as a model cellular transport barrier, Pharm. Res. 6 (1989) 71–77. [38] M.J. Cho, A. Adson, F.J. Kezdy, Transepithelial transport of aliphatic carboxylic acids studied in Madin-Darby canine kidney (MDCK) cell monolayers, Pharm. Res. 7 (1990) 325–331. [39] J.D. Irvine, L. Takahashi, K. Lockhart, J. Cheong, J.W. Tolan, H.E. Selick, J.R. Grove, MDCK (Madin-Darby canine kidney) cells: a tool for membrane permeability screening, J. Pharm. Sci. 88 (1999) 28–33.
Permeability Methods Chapter 26
337
[40] A.H. Schinkel, E. Wagenaar, C.A.A.M. Mol, L. van Deemter, P-glycoprotein in the blood-brain barrier of mice influences the brain penetration and pharmacological activity of many drug, J. Clin. Invest. 97 (1996) 2517–2524. [41] R. Li, Y.-A. Bi, Y. Lai, K. Sugano, S.J. Steyn, P.E. Trapa, L. Di, Permeability comparison between hepatocyte and low efflux MDCKII cell monolayer, AAPS J. 16 (2014) 802–809. [42] R. Ohashi, Y. Kamikozawa, M. Sugiura, H. Fukuda, H. Yabuuchi, I. Tamai, Effect of P-glycoprotein on intestinal absorption and brain penetration of antiallergic agent bepostastine besilate, Drug Metab. Dispos. 34 (2006) 793–799. [43] M. Iwai, T. Minematsu, Q. Li, T. Iwatsubo, T. Usui, Utility of P-glycoprotein and organic cation transporter 1 double-transfected LLC-PK1 cells for studying the interaction of YM155 monobromide, novel small-molecule survivin suppressant, with P-glycoprotein, Drug Metab. Dispos. 39 (2011) 2314–2320. [44] S. Tavelin, J. Taipalensuu, L. S€oderberg, R. Morrison, S. Chong, P. Artursson, Prediction of the oral absorption of low-permeability drugs using small intestine-like 2/4/A1 cell monolayers, Pharm. Res. 20 (2003) 397–405. [45] S. Tavelin, J. Taipalensuu, F. Hallbook, K.S. Vellonen, V. Moore, P. Artursson, An improved cell culture model based on 2/4/A1 cell monolayers for studies of intestinal drug transport: characterization of transport routes, Pharm. Res. 20 (2003) 373–381. [46] M. Kansy, F. Senner, K. Gubernator, Physicochemical high throughput screening: parallel artificial membrane permeation assay in the description of passive absorption processes, J. Med. Chem. 41 (1998) 1007–1010. [47] C. Zhu, L. Jiang, T.-M. Chen, K.-K. Hwang, A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential, Eur. J. Med. Chem. 37 (2002) 399–407. [48] F. Wohnsland, B. Faller, High-throughput permeability pH profile and high-throughput alkane/water log P with artificial membranes, J. Med. Chem. 44 (2001) 923–930. [49] A. Avdeef, High-throughput measurement of permeability profiles, Methods Principles Med. Chem. 18 (2003) 46–71. [50] A. Avdeef, Absorption and Drug Development: Solubility, Permeability, and Charge State, second ed., John Wiley & Sons Inc, Hoboken, USA, 2012. [51] K. Sugano, Y. Nabuchi, M. Machida, Y. Aso, Prediction of human intestinal permeability using artificial membrane permeability, Int. J. Pharm. 257 (2003) 245–251. [52] M. Bermejo, A. Avdeef, A. Ruiz, R. Nalda, J.A. Ruell, O. Tsinman, I. Gonzalez, C. Fernandez, G. Sanchez, T.M. Garrigues, V. Merino, PAMPA-a drug absorption in vitro model 7. Comparing rat in situ, Caco-2, and PAMPA permeability of fluoroquinolones, Eur. J. Pharm. Sci. 21 (2004) 429–441. [53] M. Kansy, A. Avdeef, H. Fischer, Advances in screening for membrane permeability: high-resolution PAMPA for medicinal chemists, Drug Discov. Today Technol. 1 (2004) 349–355. [54] M. Kansy, H. Fischer, S. Bendels, B. Wagner, F. Senner, I. Parrilla, V. Micallef, Physicochemical methods for estimating permeability and related properties, in: R.T. Borchardt, E.H. Kerns, C.A. Lipinski, D.R. Thakker, B. Wang (Eds.), Pharmaceutical Profiling in Drug Discovery for Lead Selection, AAPS Press, Arlington, VA, 2004, p. 197. [55] A. Avdeef, The rise of PAMPA, Expert Opin. Drug Metab. Toxicol. 1 (2005) 325–342. [56] B. Faller, H.P. Grimm, F. Loeuillet-Ritzler, S. Arnold, X. Briand, High-throughput lipophilicity measurement with immobilized artificial membranes, J. Med. Chem. 48 (2005) 2571–2576. [57] K. Obata, K. Sugano, R. Saitoh, A. Higashida, Y. Nabuchi, M. Machida, Y. Aso, Prediction of oral drug absorption in humans by theoretical passive absorption model, Int. J. Pharm. 293 (2005) 183–192. [58] Sugano, K., Obata, K., Saitoh, R., Higashida, A., Hamada, H. (2006). Processing of biopharmaceutical profiling datbba in drug discovery. Pharmacokinetic Profiling in Drug Research: Biological, Physicochemical, and Computational Strategies, [LogP2004, Lipophilicity Symposium], 3rd, Zurich, Switzerland, Feb. 29-Mar. 4, 2004, 441-458. [59] P.V. Balimane, E. Pace, S. Chong, M. Zhu, M. Jemal, C.K. Van Pelt, A novel high-throughput automated chip-based nanoelectrospray tandem mass spectrometric method for PAMPA sample analysis, J. Pharm. Biomed. Anal. 39 (2005) 8–16. [60] T. Loftsson, F. Konradsdottir, M. Masson, Development and evaluation of an artificial membrane for determination of drug availability, Int. J. Pharm. 326 (2006) 60–68. [61] A. Avdeef, S. Bendels, L. Di, B. Faller, M. Kansy, K. Sugano, Y. Yamauchi, PAMPA—critical factors for better predictions of absorption, J. Pharm. Sci. 96 (2007) 2893–2909. [62] E.H. Kerns, L. Di, S. Petusky, M. Farris, R. Ley, P. Jupp, Combined application of parallel artificial membrane permeability assay and Caco-2 permeability assays in drug discovery, J. Pharm. Sci. 93 (2004) 1440–1453. [63] H. Lennerna¨s, Animal data: the contributions of the Ussing Chamber and perfusion systems to predicting human oral drug delivery in vivo, Adv. Drug Deliv. Rev. 59 (2007) 1103–1120. [64] A. Sj€ oberg, M. Lutz, C. Tannergren, C. Wingolf, A. Borde, A.-L. Ungell, Comprehensive study on regional human intestinal permeability and prediction of fraction absorbed of drugs using the Ussing chamber technique, Eur. J. Pharm. Sci. 48 (2013) 166–180. [65] Y. Matsuda, Y. Konno, M. Satsukawa, T. Kobayashi, Y. Takimoto, Y. Morisaki, S. Yamashita, Assessment of intestinal availability of various drugs in the oral absorption process using portal vein-cannulated rats, Drug Metab. Dispos. 40 (2012) 2231–2238.