| MultiScreen® Filter Plates for PAMPA and Permeability Assays | Application Note - AN1729EN00 > Product Catalogue | |
| Data review and optimization of PAMPA and Permeability Assays |
Abstract
The bioavailability of a drug is affected by a number of factors including its ability to be absorbed into the blood stream through the cells lining the intestines. There are a number of different in vitro assay options available to predict the gastrointestinal absorption property of drugs including a permeability assay1 that uses a hexadecane-filled membrane as a lipophilic barrier, and a method known as PAMPA2 (Parallel Artificial Membrane Permeability Assay), which uses a lipid filled membrane to simulate the lipid bilayer of various cell types, including intestinal epithelium. These non-cell based permeability assays are automation compatible, relatively fast (4–24 hours), inexpensive, and straightforward. They are being used with increasing frequency to determine the passive, transcellular permeability properties of potential drug compounds. The majority of drugs (>80%) enter the blood stream by passive diffusion through the intestinal epithelium3. Consequently, permeability assays that measure passive transport through lipophilic barriers correlate with human drug absorption values from published methods.
Introduction
Assays that predict passive absorption of orally administered drugs have become increasingly important in the drug discovery process. The ability of a molecule to be orally absorbed is one of the most important aspects in deciding whether the molecule is a potential lead candidate. Cell-based assays, like those using Caco-2 cells, are commonly used as a model for drug absorption; however, the technique is labor intensive and is often situated late in the drug discovery process. Assays described by Kansy2 and Faller1 have addressed these issues by providing rapid, low cost and automation friendly methods to measure a compound’s passive permeability. Both permeability and PAMPA assays use artificial membranes to model the passive transport properties of the cell membrane. Other researchers have presented variations on Kansy’s method, in some cases, improving on the correlation with a particular target (e.g., blood–brain barrier) or class of molecules. In general, the original assay has remained the same.
Methods and Data Review
The use of filter-immobilized artificial membranes (constructed of phospholipid bilayers supported on high-porosity microfilters), as model membranes to assess permeability can be traced back to the early 1960’s. In 1962, Mueller et al4 observed that when a small quantity of a phospholipid (2% wt/vol alkane solution) was carefully placed over a small hole in a thin sheet of Teflon® resin, a thin film formed at the center of the hole. The film eventually turned optically black as a single bilayer lipid membrane (BLM) formed over the hole. However, a serious drawback in using BLM’s as a model system is that they are extremely fragile and difficult to make. Efforts to overcome the limitations of the fragile membranes have evolved with the use of membrane supports such as porous microfilters.
Kansy et al from Hoffman-La Roche described a parallel artificial membrane permeation assay (PAMPA) method for determining permeability of drugs across phospholipid-coated filters. The assay is carried out in a Millipore 96-well MultiScreen-IP PAMPA plate (cat. ELIIP10SSP or MAIPN4510 with underdrain removed) and measures the ability of compounds to diffuse from a donor to an acceptor compartment separated by a 0.45 µm pore size PVDF membrane filter impregnated with a solution of egg lecithin in dodecane. Immediately following the addition of the artificial membrane to the PVDF membrane filter in the filter plate (known as the “Donor” plate), the 96-well Donor plate is filled with buffer solutions containing the compounds to be tested. The Donor plate is then placed upon a 96-well Acceptor plate filled with sufficient buffer so that there is liquid contact between the buffer in the Acceptor plate and the PVDF membrane filter. The Donor and Acceptor plates are incubated together for 12–16 hours after which, the plates are separated, and the concentration of compound in the Acceptor compartment is determined by UV/Vis measurements. The PAMPA method was used by the Roche investigators to predict permeability of their compound with good correlation to human absorption values for most compounds. The outliers in their assay were compounds known to be actively transported. Since the artificial membranes have no active transport system or metabolizing enzymes, only pure passive diffusion of the uncharged species is observed. There have been several modifications of the PAMPA assay designed to improve throughput, reproducibility and correlation with a particular target or class of molecule. The following sections outline some of this work and its impact.
Avdeef et al5 from pION Inc. described an extension of the Roche approach, including a way to assess membrane retention. This method uses 1, 2-dioleoyl-sn-glycer-3-phosphocholine (DOPC) as the lipid component at 2% in dodecane. The experimental data obtained (Table 1, Figure 1) was compared with literature values for human jejunum permeability and the fraction absorbed in humans (BCS). This method was used to bin compounds into high or questionable permeability classes and identify those compounds which should advance to the in-depth cell culture screening. A major benefit of the work done by pION was the complete automation of the PAMPA assay from set-up through data collection.
Table 1: Comparison of PAMPA to Human Jejunal and BCS Permeabilities Report by pION5
Figure 1: PAMPA Log Pe7.4 vs. Human Absorption Values Reported by pION in Table 15.
Sugano et al6 endeavored to improve the composition of the lipid solution used in PAMPA for a more precise prediction of oral absorption. In their study, a mixture of several different phospholipids was used to more accurately resemble the mixture found in reconstituted brush-border lipids. The effects of a negative charge on the membrane and the chain length of the organic solvent were also investigated to determine their effect on predicting oral absorption. The addition of a negative charge was found to increase the permeability of basic compounds; however, their results suggested that another factor might play a role in permeability, such as, a compound’s hydrogen bond potential with the polar head group of the lipid. Recently, Sugano7 described the BAMPA assay, which is a bio-mimetic version of PAMPA from the viewpoint of lipid composition, inner barrier, and microclimate pH. To take the contribution of the paracellular pathway into account, BAMPA was combined with a prediction model based on a molecular sieving function (Renkin function). This resulted in better correlation to human absorption.
Di et al8 from Wyeth described a variation of PAMPA designed to predict blood–brain barrier permeation. The artificial membrane for the PAMPA-BBB(blood–brain barrier) assay is formed from a mixture of brain lipids selected to model the composition of the blood–brain barrier. The Wyeth investigators’ results show that the PAMPA-BBB assay predicts blood–brain barrier penetration by passive diffusion with greater than 90% certainty. The major benefits of this assay are its speed, simplicity, and cost as compared to tedious rat perfusion assays.
Faller et al1 from Novartis proposed an alternative to PAMPA called the High Throughput Permeability assay. This assay was designed to improve on PAMPA by increasing assay speed and reducing variability due to the quality of the phospholipids. The Permeability assay replaces the 120 µm thick PVDF structural support membrane filter used in PAMPA with a 10 µm thick track-etched polycarbonate support membrane filter (Millipore MultiScreen plate cat. MAPBMN310) and a 5% solution of hexadecane in hexane as the artificial membrane. By using this configuration, the incubation period can be reduced to only 5 hours and variability in artificial membrane formation is decreased. Faller’s results (Table 2, Figure 2) also show the benefit of assaying permeability at multiple pHs. By determining the maximum log Pe over a range of pH (4–8), better correlation with human absorption was observed.
Schmidt et al from Millipore investigated both the Permeability9 and PAMPA10 assays for data reproducibility and the affect of different assay conditions. Both assays were tested over the course of several days using different lots of filter plates. Several protocol variations (incubation times, volumes, temperature, etc.) were tested to simulate the analytical variability that might be encountered from lab to lab. In addition, the importance of pH profiling and time course experiments was also investigated. The results show that both assays are robust and generate reproducible results that correlate to human absorption values. In Table 3, reproducibility data for both High Throughput Permeability and PAMPA assays are shown. Higher variability is observed with low permeability drugs (i.e., methotrexate) due to the analytical method’s (UV) limit in detecting the small quantities of these drugs present in the acceptor compartment. The differences between Faller’s values and the permeability data reported by Schmidt are due to the expanded dynamic range of Schmidt’s analytical method, his higher starting (donor) concentration of drug, and differences in experimental pH (Schmidt at pH 7.4, Faller at pH 6.8).
Table 2: Fraction Absorbed in the GI-Tract, HDM Permeability Data at pH 6.8, and Highest HDM Permeability Observed Between pH 4-8 for 32 Drug Compounds as Reported by Faller.1
 | Human Abs (%) | Log Pe @ pH 6.8 | Max log Pe |
| Acyclovir | 20 | –4.8 | –4.8 |
| Alprenolol | 93 | –4 | –4 |
| Amiloride | 50 | –4.6 | –4.4 |
| Atenolol | 50 | –4.5 | –4.5 |
| Carbamazepine | 100 | –3.9 | –3.9 |
| Chlorpromazine | 100 | –3.7 | –3.5 |
| Cimetidine | 95 | –4 | –4 |
| Clozapine | 100 | –3.6 | –3.6 |
| Desferrioxamine | 2 | –5.3 | –5.3 |
| Desipramine | 100 | –3.5 | –3.5 |
| Famotidine | 45 | –4.6 | –4.3 |
| Furosemide | 61 | –4.8 | –4.5 |
| Guanabenz | 75 | –5 | –4.4 |
| Hbed | 5 | –5.1 | –5.1 |
| Hydrochlorothiazide | 67 | –4.7 | –4.2 |
| Imipramine | 99 | –3.5 | –3.5 |
| Labetalol | 90 | –4.2 | –3.8 |
| Methotrexate | 20 | –4.7 | –4.7 |
| Metolazone | 64 | –4.6 | –4.6 |
| Metoprolol | 95 | –4.2 | –4.2 |
| Phenytoin | 90 | –4 | –4 |
| Piroxicam | 100 | –4 | –3.3 |
| Progesterone | 91 | –3.8 | –3.8 |
| Propranolol | 90 | –4.4 | –4 |
| Ranitidine | 50 | –4.6 | –4.5 |
| Sulfasalazine | 13 | –4.9 | –4.9 |
| Sulpiride | 35 | –4.6 | –4.6 |
| Terbutaline | 70 | –4.1 | –4.1 |
| Testosterone | 100 | –3.5 | –3.5 |
| Timolol | 72 | –4.4 | –4.2 |
| Valsartan | 55 | –5 | –4.5 |
| Warfarin | 98 | –3.8 | –3.8 |
Figure 2: Log Pe at pH 6.8 and Max Log Pe Between pH 4–8 vs. Human Absorption Values Reported by Faller1 in Table 2.
Reproducibility
The average and standard deviation for the permeability data was determined from 224 replicates of each drug (16 replicates each per plate for a total of 14 plates from 3 different lots). The average and standard deviation for the PAMPA data was determined from 288 replicates of each drug (16 replicates each per plate for a total of 18 plates from 3 different lots). Permeability and PAMPA data were collected at pH 7.4 and results are presented in Table 3. Permeability data in row 3 was collected at pH 6.8 and reported by Faller.
pH Profiling
The eight drugs listed in Table 4 were tested in a permeability assay at four pH points each. The observed change in log Pe for some drugs illustrates the benefit of pH profiling. The permeability of charged drugs is typically greatly reduced as compared to the neutral species. For bases like propranolol, permeability increases with increasing pH, and for acids like warfarin the opposite is true. In both cases, the relative abundance of the neutral species increases with the appropriate change in pH.
Table 3: Total Average Log Pe from Permeability9 and PAMPA10 Assay Reproducibility Experiments
| Average Log Pe | |||||
| Drug | Testosterone | Propranolol | Carbamazepine | Warfarin | Furosemide | Methotrexate |
| Permeability | –3.7 ± 0.12 | –4.0 ± 0.07 | –4.2 ± 0.07 | –4.9 ± 0.11 | –6.8 ± 0.59 | –7.0 ± 0.69 |
| PAMPA | –4.9 ± 0.07 | –5.0 ± 0.05 | –5.2 ± 0.05 | –6.0 ± 0.01 | –6.8 ± 0.09 | –7.9 ± 0.38 |
| Faller permeability pH 6.8 | –3.5 | –4.4 | –3.9 | –3.8 | –4.8 | –4.7 |
Table 4: Average Log Pe Values Where n=3 at Each pH
| Average Log Pe | |||||||
pH | Testosterone | Carbamazepine | Piroxicam | Propranolol | Warfarin | Furosemide | Sulpiride | Methotrexate |
3 | –3.5 | –3.9 | ³–3.3 | –4.5 | –3.5 | –5.2 | –5.2 | –5.4 |
5 | –3.7 | –4.0 | –3.7 | –5.3 | –3.6 | –5.7 | –5.8 | –5.9 |
7.4 | –3.7 | –4.2 | –4.9 | –3.9 | –5.2 | –5.7 | –5.6 | –6.6 |
9 | –3.5 | –4.0 | –5.8 | –3.9 | –6.0 | –5.6 | –5.8 | –6.6 |
Method Comparison
The average log Pe values for six drugs determined by several different methods are summarized in Table 5. Overall, the permeability ranking of the six drugs is similar, irrespective of the method.
Table 5: Average Log Pe Values for Various Drugs Where n=32
| Average Log Pe | |||||
| Method | Testosterone | Propranolol | Carbamazepine | Warfarin | Furosemide | Methotrexate |
| PAMPA 1% lecithin (100% dodecane) | –4.9 ± 0.06 | –5.0 ± 0.04 | –5.2 ± 0.01 | –6.0 ± 0.02 | –6.9 ± 0.08 | –7.6 ± 0.43 |
| PAMPA 1% lecithin (9:1 dodecane: decadiene) | –4.8 ± 0.18 | –5.0 ± 0.04 | –5.1 ± 0.01 | –5.9 ± 0.02 | –6.9 ± 0.1 | –7.6 ± 0.49 |
| Permeability (pH 7.4) | –3.7 ± 0.08 | –4.0 ± 0.06 | –4.2 ± 0.05 | –5.0 ± 0.05 | –6.9 ± 0.39 | –7.2 ± 0.05 |
Time Course Studies
Diffusion of the tested compounds from the donor to acceptor compartments of the Permeability and PAMPA assays was monitored at various times over a 48 hour period.9, 10 Drugs were added to successive rows of a plate for a set time (48, 32, 28, 24, 8, 6, 4, 2 hrs) prior to plate separation and analysis by UV/Vis. Each row of the plate had 2 replicates for each of the 6 drugs. Two plates were assayed resulting in a total of 4 replicates of each drug per time point for the Permeability assay. Three plates were assayed for a total of 6 replicates of each drug per time point for the PAMPA assay. The average of these replicates is plotted in Figures 3 and 4, respectively. The log Pe’s calculated for each time point are listed in Tables 6 and 7, respectively.
Tables 6 and 7 show the calculated log Pe for each time point during the 48-hour time study. This experiment is conducted to determine a drug's appropriate transport time, one that balances reproducibility and assay speed. The calculated log Pe for high and medium permeable drugs varied £0.4 log between time points; however, significant change was observed over time with low permeability compounds. The fact that the log Pe calculated for each point does not change significantly for high and medium permeable drugs over 4--24 hours is an indication of the robustness of the equation.
Figure 3: Ratio of OD Acceptor/OD Equilibrium (Permeability)
Figure 4: Ratio of OD Acceptor/OD Equilibrium (PAMPA)
Table 6: Average Log Pe for Different Incubation Times (Permeability)
| Average Log Pe | |||||
Time (hrs.) | Testosterone | Propranolol | Carbamazepine | Warfarin | Furosemide | Methotrexate |
2 | –3.7 | –3.9 | –4.2 | –4.9 | –5.5 | –5.3 |
4 | –3.7 | –4.0 | –4.2 | –5.0 | –6.0 | –6.3 |
6 | –3.7 | –4.0 | –4.2 | –5.0 | –6.1 | –6.2 |
8 | –3.8 | –4.1 | –4.3 | –5.0 | –6.2 | –6.3 |
24 | –3.8 | –4.3 | –4.3 | –5.0 | –6.7 | –7.0 |
28 | –4.0 | –4.3 | –4.3 | –5.0 | –6.9 | –7.2 |
32 | –4.0 | –4.4 | –4.3 | –5.1 | –6.8 | –6.9 |
48 | –4.2 | –4.3 | –4.4 | –4.9 | –7.1 | ------ |
Table 7: Average Log Pe at Different Incubation Times (PAMPA)
| Average Log Pe | |||||
Time (hrs.) | Testosterone | Propranolol | Carbamazepine | Warfarin | Furosemide | Methotrexate |
2 | –4.5 | –4.9 | –5.2 | –5.9 | –5.9 | –6.6 |
4 | –4.6 | –4.9 | –5.2 | –5.9 | –6.2 | –7.0 |
6 | –4.6 | –4.9 | –5.2 | –5.9 | –6.4 | –7.2 |
8 | –4.7 | –5.0 | –5.2 | –6.0 | –6.5 | –7.4 |
24 | –5.0 | –5.1 | –5.2 | –6.0 | –7.0 | –7.8 |
28 | –5.0 | –5.1 | –5.2 | –5.9 | –7.2 | –7.7 |
32 | –5.1 | –5.1 | –5.2 | –6.0 | –7.2 | –7.8 |
48 | –5.0 | –5.3 | –5.2 | –6.0 | –7.5 | –7.9 |
Conclusions
Both the Permeability and PAMPA assays are simple and robust, and data generated by these assays correlate with human absorption data. The Permeability assay provides the benefits of increased speed due to shorter incubation times and improved data reproducibility. The improved reproducibility is due to the elimination of a phospholipid membrane, which may contribute to assay variations depending on the quality of the phospholipids and how successfully the lipid layers form. The PAMPA assay provides the benefits of a more biologically relevant membrane and the ability to tailor the membrane to a specific target such as the blood–brain barrier. However, due to the use of a complex phospholipid membrane, PAMPA may require extensive optimization before reproducible correlative data is obtained. Unlike Caco-2 experiments, both Permeability and PAMPA assays are compatible with pH profiling. Due to the high variability of pH in the lumen of the G.I. tract, pH profiling can help better predict drug behavior.
References
1. Wohnsland, F.; Faller, B. High-throughput Permeability pH Profile and High-throughput Alkane/Water Log P With Artificial Membranes, J. Med. Chem., 2001; 44, p. 923–930.
2. Kansy, M.; Senner, F.; Gubernator, K. Physicochemical High Throughput Screening: Parallel Artificial Membrane Permeation Assay in the Description of Passive Absorption Processes, J. Med. Chem., 1998; 41, p. 1007–1010.
3. Brennan, M.B., Drug Discovery (Filtering out Failures Early in the Game), Chem. & Eng. News, 2000; 78, 63.
4. Mueller, P.; Rudin, D.O.; Tien, H.T.; Westcott, W.C. Nature, 1962; 194, 979.
5. Ruell, J.A., Avdeef, A., Du, C. and Tsinman, K., A Simple PAMPA Filter for Passively Absorbed Compounds, Poster, ACS National Meeting, Boston, August 2002.
6. Sugano, K.; Hamada, H.; Machida, M.; Ushio, H. High Throughput Prediction of Oral Absorption: Improvement of the Composition of the Lipid Solution Used in Parallel Artificial Membrane Permeation Assay, J. Biomolecular Screening, 2001; 6, p. 189–196.
7. Sugano, K. BAMPA: Beyond the Phyisicochemical Profiling, PAMPA; 2002.
8. Di, L.; Kerns, E.; McConnell, O.; Carter, G. High Throughput Artificial Membrane Permeability Assay for Blood–Brain Barrier, PAMPA; 2002.
9. Schmidt, D.; Lynch, J. Evaluation of the Reproducibility of Passive, Transcellular Drug Permeability Assays Using MultiScreen™ Permeability Assay Plates, Millipore Corporation Application Note, 2002; Lit. No. AN1725EN00.
10. Schmidt, D.; Lynch, J. Evaluation of the Reproducibility of Parallel Artificial Membrane Permeation Assays, Millipore Corporation Application Note, 2002; Lit. No. AN1728EN00.
Application Notes
Additional Application Notes for PAMPA and Permeability assays using MultiScreen filter plates:
AN1725EN00: Evaluation of the reproducibility of passive, transcellular drug permeability assays using MultiScreen® Permeability assay plates
AN1728EN00: Evaluation of the reproducibility of Parallel Artificial Membrane Permeation Assays (PAMPA)