PAMPA study of the temperature effect on permeability

European Journal of Pharmaceutical Sciences 53 (2014) 45–49

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

PAMPA study of the temperature effect on permeability Gábor Vizserálek a, Tamás Balogh a, Krisztina Takács-Novák a, Bálint Sinkó a,b,⇑ a b

}gyes Endre Street, 1092 Budapest, Hungary Department of Pharmaceutical Chemistry, Semmelweis University, 9 Ho } lo }s Street, Budapest, Hungary SinkoLAB Scientific, 21 Nagyszo

a r t i c l e

i n f o

Article history: Received 14 May 2013 Received in revised form 18 October 2013 Accepted 5 December 2013 Available online 17 December 2013 Keywords: Temperature dependence Permeability Artificial membrane permeability assay Skin PAMPA™

a b s t r a c t The purpose of this work was to investigate the temperature dependence of permeability measured by PAMPA method. The effective permeability (log Pe) of seven drugs representing diverse structures and different acid–base properties was determined on three membrane models (GIT, BBB, Skin). The incubation temperature was varied in the range of 15–55 °C with ten degree steps. The intrinsic permeability (log P0) of the compounds is in linear relation with temperature (T). The slope of the log P0 = aT + b regression equation is a good measure of the temperature effect on permeability. Results show intensive and significant temperature dependence of permeability influenced by the properties of the compounds and also by the selected PAMPA model. The Skin PAMPA™ proved to be the most sensitive on temperature alteration, though GIT and BBB PAMPA results were also affected. The compound with acid function showed the lowest temperature dependence, while the permeability of bases increased considerably in response to the increasing temperature. The importance of human-relevant incubation conditions at in vitro assays is concluded for the better in vivo prediction. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Prediction of ADMET properties of drug candidates plays an indispensable role in drug discovery and development. Physicochemical profiling provides a useful approach for this purpose since pharmacokinetic behavior of drug molecules depends on inherent molecular properties. Permeability and solubility are two of the key parameters determining the pharmacokinetic behavior of orally administered drugs, therefore they also serve as the bases of the Biopharmaceutics Classification System (BCS), firstly introduced by Amidon and co-workers in 1995 (Amidon et al., 1995). The BCS sorts the active pharmaceutical ingredients (API) into four classes according to their permeability and solubility at highest dose strength and provides a basic estimation on ADME processes. Exact permeability value of a potentially active compound is essential for lead optimization, therefore fast and reliable permeability measurement strategies are needed to classify the molecules (Sinko et al., 2012). For this purpose, a 96-well microtiter plate based technology called PAMPA (Parallel Artificial Membrane Permeability Assay) was introduced by Kansy and coworkers (Kansy et al., 1998). PAMPA was aimed to serve as a rapid in vitro method for the evaluation of passive transcellular permeability. The original PAMPA system contained phosphatidylcholine ⇑ Corresponding author at: SinkoLAB Scientific, 21 Nagyszo} lo}s Street, 1113 Budapest, Hungary. Tel.: +36 1 7892896. E-mail address: [email protected] (B. Sinkó). 0928-0987/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejps.2013.12.008

(PC) dissolved in n-dodecane as a membrane barrier (Reis et al., 2010), while recent papers have described numerous tissue-specific methods that predict the gastrointestinal (GIT) absorption (Avdeef et al., 2007; Chen et al., 2008; Sugano et al., 2003; Zhu et al., 2002), the penetration through the blood–brain barrier (BBB) (Carrara et al., 2007; Di et al., 2003; Tsinman et al., 2011) or the barrier function of human skin (Sinko et al., 2012). Most of these models – except the skin penetration system – are based on phospholipids, cholesterol or porcine polar brain lipids dissolved in organic solvent like n-dodecane, octadiene or decadiene. The membrane of the skin penetration system (Skin PAMPA™) includes synthetic ceramide analogs, cholesterol and free fatty acids (Sinko et al., 2012). There are several factors influencing the PAMPA permeability values where the membrane’s lipid composition has been studied the most extensively (Carrara et al., 2007; Mensch et al., 2010; Seo et al., 2006). The unstirred water layer effect that can significantly influence the permeability values has also been investigated before (Avdeef, 2012; Avdeef et al., 2005; Ruell et al., 2003). Some other parameters like acceptor and donor solution composition, pH conditions and filter porosity have also been discussed (Avdeef, 2012). Although the incubation temperature plays an important role at the determination of physico-chemical (equilibrium or kinetic) constants, systematic study investigating the temperature dependence of permeability has not been published so far. The aim of our current work was to study the significance and extent of temperature dependence of permeability. We have investigated the permeability value of seven marketed drugs with

46

G. Vizserálek et al. / European Journal of Pharmaceutical Sciences 53 (2014) 45–49

diverse structures on three different PAMPA penetration models (GIT, BBB, Skin). The incubation temperature was varied in the range of 15–55 °C with ten degree steps. Besides the nature of temperature dependence on permeability of the compounds we also wanted to investigate how sensitive are the different membranes on various incubation temperatures. A correlation study using a larger set of compounds has also been performed comparing PAMPA permeability data and relevant in vitro human tissue models. 2. Materials and methods 2.1. Chemicals Carbamazepine, codeine, corticosterone, diclofenac, fentanyl, hydrocortisone, indomethacin, ketoprofen, lidocaine, morphine, piroxicam, prazosin, progesterone, propranolol, quinine, salicylic acid, verapamil, warfarin, zidovudine, cholesterol (03108HH-377) and phosphatidylcholine (PC) (057K5200) were purchased from Sigma–Aldrich, and were used without further purification. The Porcine Brain Lipid Extract (PBLE) (86088-88-2) was purchased from Avanti Polar Lipids Inc. DMSO (purity grade, >99.9%) was purchased from Reanal. N-dodecane was purchased from Merck (8.20543.0250). Standard Britton–Robinson buffers were made in-house, and were used for measurement. The pH values were measured using Meterlab PHM 220 pH meter (Radiometer AnalyticalÒ).

2.3. PAMPA equations The pION’s PAMPA Explorer™ software was used for all calculations. Considering membrane retention and gradient pH, the software uses the following published equation (Avdeef, 2012; Sinko et al., 2009):

Pe ¼ 

      2:303V D 1 1 þ ra C D ðtÞ  log10 r a þ   C D ð0Þ Aðt  sLAG Þ 1 þ r a 1R

where Pe is the effective permeability coefficient (cm/s), A is the filter area (0.3 cm2) multiplied by a nominal porosity of 70% according to the manufacturer, VD and VA are the volumes in the donor and acceptor phase (0.2 cm3), t is the incubation time, sLAG is the steady-state time (s), CD(t) is the concentration (mol cm3) of the compound in the donor phase at time t, CD(0) is the concentration (mol cm3) of the compound in the donor phase at time 0 and R is the membrane retention factor:

R¼1

C D ðtÞ V A C A ðtÞ  C D ð0Þ V D C D ð0Þ

ra ¼

  V D Pe ðA  DÞ V A Pe ðD  AÞ

ð3Þ

The PAMPA Evolution™ was used to calculate the intrinsic permeability data (P0) using the following equations (Avdeef, 2012):

P0 ¼ Pm ð1 þ 10pKapH Þ ™

ð2Þ

and ra is the sink asymmetric ratio (gradient-pH induced):

P0 ¼ Pm ð1 þ 10pKaþpH Þ 2.2. PAMPA method

ð1Þ

ðweak acidÞ

ð4Þ

ðweak baseÞ

ð5Þ

™

The PAMPA Explorer , PAMPA Evolution (pION INC) software packages were used for performing permeability measurements and data analysis. The 96-well STIRWELL™ PAMPA sandwiches (P/N: 110243) and the stirring bars (P/N: 110211) were from pION INC. The UV plates were from Greiner Bio-one (UV-star micro plate, clear, flat bottom, half area). The manufacturer-calibrated pipettes (EDP-3 and pipet lite series) were purchased from Mettler-Toledo. The Gut-Box™ from pION INC was used for stirring. Each well of the top (acceptor) compartment of STIRWELL™ PAMPA sandwich was coated with 5 lL of membrane components dissolved in n-dodecane. The GIT-PAMPA membrane consisted of 2% wt/vol phosphatidylcholine (PC) and 1% wt/vol cholesterol dissolved in n-dodecane. 5 lL of this freshly prepared solution was pipetted on each filter to form the PAMPA membrane. The BBBPAMPA method was adopted from Di and co-workers (Di et al., 2003), but additional cholesterol was added to the mixture (1% wt/vol) (Könczöl et al., 2013), and the same procedure was applied as indicated above. The Skin PAMPA™ sandwiches were purchased from pION INC, and were used after an overnight hydration. Before forming the sandwich, the bottom (donor) plate was filled with 180 lL of test compounds dissolved in the Britton Robinson buffer at different pH values, containing 1% v/v DMSO and the stirring bars. The acceptor plate was filled with 200 lL of fresh pH 7.4 Britton-Robinson buffer. The resultant sandwich was incubated at preadjusted temperature for 1–4 h. Lauda E100 termostat was used to adjust the desired temperature. Evaporation of the solutions was prevented. After the permeation time, the PAMPA sandwich was separated and 150 lL of both the donor and acceptor compartments were transferred to UV plates. UV absorption (230–500 nm) was measured with Tecan Infinite M200 driven by PAMPA Explorer software. Model compound concentrations were chosen according to their solubility and UV detection limits (50–300 lM). Data were measured with three replicates on every single plate.

  P0 ¼ Pm 1 þ 10pKa1pH þ 10pKa2þpH

ðampholyteÞ

ð6Þ

where Pm is the membrane permeability. 2.4. The effect of the temperature on permeability The membrane permeability (Pm) can be expressed by the following equation (Avdeef, 2012)

Pm ¼

Dm K d h

ð7Þ

where Dm is the diffusivity of the solute within the membrane, in units of cm2/s, Kd is the pH-dependent apparent partition coefficient, and h is the thickness of the membrane in centimeters. Based on Eq. (7), the temperature dependence of permeability can be expected hence both the diffusivity and the membrane partition coefficient are temperature dependent parameters. The relation between the diffusivity and the temperature can be written by the Stokes–Einstein equation (Bockris and Reddy, 1998):

Dm ¼

kT 6pgr

ð8Þ

where T is the temperature, k is the Boltzmann constant, g is the viscosity and r is the radius of the spheric particles.

dðlog K d Þ a ¼ þb dT T

ð9Þ

The van’t Hoff plot (9) (Sangster, 1997) describes the change of partition coefficient upon temperature variation, where a and b are fitting parameters. As temperature dependence of membrane partition coefficient depends on membrane-solute interactions and it can vary in a broad range, the estimation of significant temperature dependence of permability is also reasonable.

G. Vizserálek et al. / European Journal of Pharmaceutical Sciences 53 (2014) 45–49

47

Table 1 The measured intrinsic permeability in GIT-PAMPA and BBB-PAMPA at five various temperatures. log P0 t (°C)

15

25

37

45

55

Carbamazepine Diclofenac Ketoprofen Piroxicam Progesterone Propranolol Verapamil

GIT-PAMPA 4.93 2.11 3.65 3.27 4.21 2.75 2.52

4.94 1.98 3.61 3.16 4.34 2.63 2.60

4.75 1.93 3.55 3.13 4.17 2.18 2.28

4.64 1.94 3.44 3.16 4.04 2.06 2.10

4.49 1.88 3.38 3.07 3.64 1.84 1.78

Carbamazepine Diclofenac Ketoprofen Piroxicam Progesterone Propranolol Verapamil

BBB-PAMPA 4.91 2.08 3.49 3.37 4.55 2.74 3.13

4.78 1.96 3.39 3.26 4.35 2.46 2.85

4.52 1.80 3.19 3.10 4.21 2.33 2.76

4.39 1.76 3.04 3.04 4.08 2.17 2.62

4.21 1.68 2.96 2.99 3.94 2.00 2.33

Fig. 1. Log P0 vs. T in GIT-PAMPA.

3. Results and discussion Seven commercially available drugs with diverse structure and different acid–base properties were used as test compounds. Diclofenac and ketoprofen represented the acids, propranolol and verapamil represented the bases, carbamazepine and progesterone were used as neutral molecules, and piroxicam has been selected as an ampholyte. The effective permeability data (Pe) were measured in three different PAMPA models, i.e. GIT-PAMPA, BBB-PAMPA and Skin PAMPA™. Phosphatidylcholine was used as main membrane component in GIT-PAMPA, while Porcine Brain Lipid Extract served as membrane in BBB-PAMPA. The Skin PAMPA™ model is a phospholipid-free system. The three examined PAMPA systems differ considerably, therefore the effect of the membrane components on the permeability can be observed according to the various incubation temperatures. The permeability measurements were performed at five different incubation temperatures: 15, 25, 37, 45, 55 °C. In case of Skin PAMPA™ 32 °C has been included to simulate the temperature conditions of Franz cell measurement. The logarithm of the obtained intrinsic permeability values (log P0) in the three systems are shown in Tables 1 and 2. The standard deviation (SD) of log P0 values falls into range 0.01– 0.04 in all three PAMPA models, the average value is 0.02. The results show that the intrinsic permeability of the compounds is linearly proportional to temperature. However, the slope of the permeability-temperature curve is strongly influenced by the compounds and also the PAMPA membrane used for the measurements. Figs. 1–3 show the plot of log P0 versus T while Table 3 demonstrates the results of the linear regression analysis (log P0 = aT + b). The correlation coefficients (R) vary in the range of 0.84–0.99 in GIT-PAMPA, 0.98–0.99 in BBB-PAMPA and 0.91–0.99 in Skin PAMPA™, respectively. The slopes of the linear regression

Fig. 2. Log P0 vs. T in BBB-PAMPA.

Table 2 The measured intrinsic permeability in Skin PAMPA™ at six various temperatures. log P0 t (°C)

15

Carbamazepine Diclofenac Ketoprofen Piroxicam Progesterone Propranolol Verapamil

Skin PAMPA™ 6.00 5.95 3.46 3.46 4.60 4.29 4.69 4.28 4.78 4.58 4.74 3.96 4.13 3.59

25

32

37

45

55

5.58 3.16 3.98 4.27 4.75 3.30 3.44

5.26 2.94 3.73 4.22 4.53 3.12 3.40

4.88 2.77 3.52 3.93 4.40 2.92 3.20

4.59 2.60 3.24 3.47 4.23 2.57 2.99

Fig. 3. Log P0 vs. T in Skin PAMPA™.

equations in Table 3 (‘a’ values) are equal to the logP0 change caused by 1 °C of temperature increase. Therefore, an increase of incubation temperature from 25 °C to 37 °C will increase the log P0 value by 12  a units. The a values represent the behavior of the

48

G. Vizserálek et al. / European Journal of Pharmaceutical Sciences 53 (2014) 45–49

Table 3 The linear regression analysis between the intrinsic permeability and the temperature. a is the slope, b is the intercept, and R is correlation coefficient. GIT-PAMPA

Carbamazepine Diclofenac Ketoprofen Piroxicam Progesterone Propranolol Verapamil

Skin PAMPA™

BBB-PAMPA log P0 = aT + b

a

b

R

a

b

R

a

b

R

0.012 0.005 0.007 0.004 0.014 0.024 0.020

5.17 2.15 3.77 3.30 4.58 3.14 2.95

0.97 0.92 0.98 0.88 0.84 0.99 0.94

0.018 0.010 0.014 0.010 0.015 0.018 0.018

5.20 2.21 3.71 3.50 4.75 2.97 3.38

0.99 0.99 0.99 0.98 0.99 0.99 0.98

0.039 0.024 0.035 0.028 0.013 0.053 0.026

6.74 3.91 5.11 5.11 5.01 5.29 4.38

0.97 0.97 0.99 0.96 0.91 0.96 0.96

compounds in different PAMPA systems at various incubation temperatures. The lowest slopes in GIT-PAMPA (0.004-0.007) belong to the two acids, diclofenac and ketoprofen, and to the ampholyte, piroxicam. In case of these compounds the temperature dependence of permeability is minimal, only exceeding the standard deviation by 5 °C of temperature raise. Contrarily, the permeability of bases (verapamil and propranolol) show significant temperature dependence, where the slopes of the plots are 0.020 and 0.024, respectively. This means if the temperature increases by 5 °C, the change in log P0 exceeds the SD by five times. The BBB-PAMPA model showed the lowest effect on temperature changes, the slopes were varying between 0.010 (diclofenac and piroxicam) and 0.018 (carbamazepine, propranolol and verapamil). The above established trend can be also observed in this model as well, i.e. the permeability of the bases is more sensitive on temperature raise. The largest differences in the slopes can be observed in case of Skin PAMPA™. Progesterone (a = 0.013) has the lowest slope that is four times lower than the slope determined at propranolol (a = 0.053). The average slope in the different PAMPA models is 0.012 (from 0.004 to 0.024; D:0.020) in GIT-PAMPA, 0.015 (from 0.010 to 0.018, D:0.008) in BBB- PAMPA and 0.031 (from 0.013 to 0.053, D:0.040) in Skin PAMPA™. From this we can conclude that the Skin PAMPA™ is the most sensitive on the temperature alteration comparing to the other models, since it shows five times higher variability than BBB-PAMPA and double variability than GIT-PAMPA. This can be caused by the completely different (phospholipid-free) membrane composition of Skin PAMPA™ as discussed above. The difference in variability of GIT- and BBB-PAMPA models is probably caused by phase transition temperature that is usually higher for lipid mixtures (like BBB-PAMPA) then for mono-component systems (like GIT-PAMPA). The largest variation of temperature dependence among the tested systems can be observed at piroxicam, i.e. the slope of the temperature vs. permeability curve was found to be a = 0.004 in GIT-PAMPA model, while a slope of almost three times higher in BBB-PAMPA (a = 0.010) and seven times higher in Skin PAMPA™ (a = 0.028) has been measured. Contrarily, no difference was found at progesterone (MW = 314.46), the slopes of the plot is 0.014 in GIT-PAMPA, 0.015 in BBB-PAMPA and 0.013 in Skin PAMPA™. The different behavior can be explained rather by the molecular size of progesterone that is the largest in the group, than the neutral character of the molecule. Since, the permeability of the other neutral molecule of our test compounds, carbamazepine with lower molecular weight (MW = 236.27) showed three times larger temperature dependence in Skin PAMPA™ (a = 0.039) than in GIT-PAMPA (a = 0.012). Due to the above described fact incubation temperature appears to be an important parameter of in vitro permeability assays and can have a significant effect on the quality of in vitro/in vivo correlations. To demonstrate this consideration, our test set has been extended to 16 compounds with known log BB (brain to plasma

Table 4 The measured effective permeability at pH 7.4 at 25 and 37 °C, and the log BB data. t (°C)

Propranolol Quinine Fentanyl Codeine* Progesterone Lidocaine Carbamazepine Warfarin Morphine* Corticosterone Verapamil* Zidovudine* Hydrocortisone Prazosin Salicylic acid Indomethacin a

log BBa

BBB-PAMPA pH 7.4 log Pe 25

37

4.39 4.51 4.32 5.13 4.35 4.86 4.78 5.74 6.06 5.01 4.17

4.17 4.34 3.97 4.94 4.21 4.31 4.52 5.64 5.31 4.68 3.92 6.52 5.32 4.8 6.66 5.52

5.46 4.95 7.25 5.44

0.64 0.60 0.60 0.55 0.20 0.10 0.00 0.00 0.16 0.50 0.70 0.89 0.90 0.90 1.10 1.26

From Platts et al. (2001) and Tsinman et al. (2011).

ratio) data (Platts et al., 2001; Tsinman et al., 2011), and this extended set was studied in BBB-PAMPA at two different incubation temperature (25 and 37 °C) to perform a correlation analysis between the measured effective permeability (log Pe) at pH 7.4 and the log BB values. The log BB data have been determined at 37 °C. The log Pe and the log BB data are summarized in Table 4. Four outliers have appeared in the correlation analysis italicized in Table 4 (codeine, morphine, verapamil, and zidovudive). These compounds are known P-gp substrates (Cunningham et al., 2008; Orlowski et al., 1996; Quevedo et al., 2011), so they were excluded from further study. The resulted linear regression equations and the correlation coefficients are shown as following:

log P e ¼ 0:831  log BB  4:91 R ¼ 0:70

ð25  CÞ

log P e ¼ 0:864  log BB  4:66 R ¼ 0:75 ð37  CÞ

ð10Þ ð11Þ

The better correlation (R = 0.75) was found between the log BB data and the BBB-PAMPA permeability results measured at 37 °C, so identical incubation temperature resulted in a higher correlation. A similar trend can be observed in case of Skin PAMPA™ and Franz cell measurement. The measured log Pe at 32 °C and the Franz cell data (32 °C) showed the best correlation (data not shown). Our study suggests that higher correlation between in vitro artificial and in vivo models is possible if bio-relevant incubation temperature is selected. Most of the published PAMPA assays are performed at room temperature while they are correlated to gastro-intestinal or blood brain barrier data that has been determined at 37 °C or to skin data that has been determined at 32 °C in most cases. A better selection of incubation temperature could provide a higher correlation to in vivo data.

G. Vizserálek et al. / European Journal of Pharmaceutical Sciences 53 (2014) 45–49

4. Conclusion Results demonstrated a significant effect of temperature on permeability measured by PAMPA method. This effect is influenced by the properties of the compounds and also by the selected PAMPA model. The Skin PAMPA™ appeared to be the most sensitive on temperature alteration, though GIT and BBB PAMPA results were also affected. The permeability of the compounds with acid function showed the lowest temperature dependence, however, the permeability of the bases increased considerably in response to the increasing temperature. As an extension of this study it can be predicted that permeability as a physico-chemical parameter has a temperature dependence that must be considered at every in vitro permeability assay. In general, the better prediction of in vivo permeability processes can be reached by using human-relevant incubation conditions at in vitro assays. References Amidon, G.L., Lennernas, H., Shah, V.P., Crison, J.R., 1995. A theoritical basis for a biopharmaceutic drug classification – the correlation of in-vitro drug product dissolution and in-vivo bioavailability. Pharm. Res. 12, 413–420. Avdeef, A., 2012. Absorption and Drug Development, Second ed. Avdeef, A., Artursson, P., Neuhoff, S., Lazorova, L., Grasjo, J., Tavelin, S., 2005. Caco-2 permeability of weakly basic drugs predicted with the double-sink PAMPA pKa(flux) method. Eur. J. Pharm. Sci. 24, 333–349. Avdeef, A., Bendels, S., Di, L., Faller, B., Kansy, M., Sugano, K., Yamauchi, Y., 2007. PAMPA – critical factors for better predictions of absorption. J. Pharm. Sci. 96, 2893–2909. Bockris, J.O.M., Reddy, A.K.N., 1998. Modern elechtrochemistry. Kluwer Academic/ Plenum Publishers, New York. Carrara, S., Reali, V., Misiano, P., Dondio, G., Bigogno, C., 2007. Evaluation of in vitro brain penetration: optimized PAMPA and MDCKII-MDR1 assay comparison. Int. J. Pharm. 345, 125–133. Chen, X., Murawski, A., Patel, K., Crespi, C.L., Balimane, P.V., 2008. A novel design of artificial membrane for improving the PAMPA model. Pharm. Res. 25, 1511– 1520. Cunningham, C.W., Mercer, S.L., Hassan, H.E., Traynor, J.R., Eddington, N.D., Coop, A., 2008. Opioids and efflux transporters. Part 2: P-glycoprotein substrate activity of 3- and 6-substituted morphine analogs. J. Med. Chem. 51, 2316–2320. Di, L., Kerns, E.H., Fan, K., McConnell, O.J., Carter, G.T., 2003. High throughput artificial membrane permeability assay for blood – brain barrier. Eur. J. Med. Chem. 38, 223–232.

49

Kansy, M., Senner, F., Gubernator, K., 1998. Physicochemical high throughput screening: parallel artificial membrane permeation assay in the description of passive absorption processes. J. Med. Chem. 41, 1007–1010. Könczöl, Á., Müller, J., Földes, E., Béni, Z., Végh, K., Kéry, Á., Balogh, T.Gy., 2013. Applicability of a blood-brain barrier specific artificial membrane permeability assay at the early stage of natural product-based CNSDrug discovery. J. Nat. Prod. 76, 655–663. Mensch, J., Melis, A., Mackie, C., Verreck, G., Brewster, M.E., Augustijns, P., 2010. Evaluation of various PAMPA models to identify the most discriminating method for the prediction of BBB permeability. Eur. J. Pharm. Biopharm. 74, 495–502. Orlowski, S., Mir, L.M., Belehradek, J., Garrigos, M., 1996. Effects of steroids and verapamil on P-glycoprotein ATPase activity: progesterone, desoxycorticosterone, corticosterone and verapamil are mutually nonexclusive modulators. J. Biochem. 317, 515–522. Platts, J.A., Abraham, M.H., Zhao, Y.H., Hersey, A., Ijaz, L., Butina, D., 2001. Correlation and prediction of a large blood-brain distribution data set – an LFER study. Eur. J. Med. Chem. 36, 719–730. Quevedo, M.A., Nieto, L.E., Brinon, M.C., 2011. P-glycoprotein limits the absorption of the anti-HIV drug zidovudine through rat intestinal segments. Eur. J. Pharm. Sci. 43, 151–159. Reis, J.M., Sinko, B., Serra, C.H.R., 2010. Parallel Artificial Membrane Permeability Assay (PAMPA) – is it better than Caco-2 for human passive permeability prediction? Mini-Rev. Med. Chem. 10, 1071–1076. Ruell, J.A., Tsinman, K.L., Avdeef, A., 2003. PAMPA – a drug absorption in vitro model: 5. Unstirred water layer in iso-pH mapping assays and pKaflux – Optimized design (pOD-PAMPA). Eur. J. Pharm. Sci. 20, 393–402. Sangster, J., 1997. Octanol-Water Partition Coefficients: Fundamentals and Physical Chemistry. Seo, P.R., Teksin, Z.S., Kao, J.P.Y., Polli, J.E., 2006. Lipid composition effect on permeability across PAMPA. Eur. J. Pharm. Sci. 29, 259–268. Sinko, B., Kokosi, J., Avdeef, A., Takacs-Novak, K., 2009. A PAMPA study of the permeability-enhancing effect of new ceramide analogues. Chem. Biodivers. 6, 1867–1874. Sinko, B., Garrigues, T.M., Balogh, G.T., Nagy, Z.K., Tsinman, O., Avdeef, A., TakacsNovak, K., 2012. Skin-PAMPA: a new method for fast prediction of skin penetration. Eur. J. Pharm. Sci. 45, 698–707. Sugano, K., Nabuchi, Y., Machida, M., Aso, Y., 2003. Prediction of human intestinal permeability using artificial membrane permeability. Int. J. Pharm. 257, 245– 251. Tsinman, O., Tsinman, K., Sun, N., Avdeef, A., 2011. Physicochemical selectivity of the BBB microenvironment governing passive diffusion–matching with a porcine brain lipid extract artificial membrane permeability model. Pharm. Res. 28, 337–363. Zhu, C., Jiang, L., Chen, T.M., Hwang, K.K., 2002. A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential. Eur. J. Med. Chem. 37, 399–407.