Medicinal Chemistry

2 downloads 0 Views 2MB Size Report
Medicinal. Chemistry. Research Article part of. Practical permeability-based hepatic ... Research Article Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh.
Future

Research Article

Medicinal Chemistry

For reprint orders, please contact [email protected]

Practical permeability-based hepatic clearance classification system (HepCCS) in drug discovery

Background: The use of liver microsomes and hepatocytes to predict total in vivo clearance is standard practice in the pharmaceutical industry; however, metabolic stability data alone cannot always predict in vivo clearance accurately. Results: Apparent permeability generated from Mardin–Darby canine kidney cells and rat hepatocyte uptake for 33 discovery compounds were obtained. Conclusion: When there is underprediction of in vivo clearance, compounds with low apparent permeability (less than 3 × 10 -6 cm/s) all exhibited hepatic uptake. A systematic approach in the form of a classification system (hepatic clearance classification system) and decision tree that will help drug discovery scientists understand in vitro–in vivo clearance prediction disconnect early is proposed.

The ability to accurately predict in vivo total clearance (CL) of small molecules in preclinical species and in humans is one of the major goals for DMPK scientists. The use of liver microsomes supplemented with cofactors as well as the use of hepatocytes has been widely applied to measure the metabolic stability of drugs in early drug discovery programs. The rate of substrate depletion due to intrinsic metabolism (CLint, met) in the abovementioned systems is typically incorporated into the well-stirred liver model (Equation 1), which also includes the concentration of free drug in plasma and incubations (fup and fuinc, respectively), the blood to plasma ratio correction (fub) and the hepatic blood flow (Qh). Many examples exist to show that the wellstirred model is effective at predicting in vivo CL from these in vitro parameters, especially for drugs that undergo primarily cytochrome P450-mediated metabolism [3–6] . CL =

Q h $ fu b $

CLint, met

fuinc CL + int, met Q h fub $ fuinc Equation 1

This method works well for compounds with high permeability since such compounds are not limited in their exposure to the liver. For perfusion rate-limited

10.4155/FMC.14.141  © 2014 Future Science Ltd

clearance, both microsomal and hepatocyte incubations should provide an accurate clearance prediction assuming that: the liver is the primary organ for metabolism; no enzyme activity is significantly lost during preparation or storage; the measurement of metabolic stability using the substrate depletion method is tested at a concentration below the test article’s Michaelis–Menten constant (K m) [7] ; and, finally, there is no inhibition of metabolic enzyme activity through timedependent formation of metabolites. When permeability is low, however, the rate limitation shifts away from one of perfusion to one of permeability. If nonhepatic elimination pathways are insignificant and if the rate of membrane penetration into hepatocytes is slow, a molecule that is stable in blood cannot have a liver blood flow clearance unless energy-dependent active uptake transporters are involved. Several groups have generated equations, such as Equation 2, to describe permeability rate-limited clearance [8–12] . CL int, app = CLint, met

Peter W Fan*,‡,1, Yang Song‡,2, Leonid M Berezhkovskiy1, Jonathan Cheong1, Emile G Plise1 & S Cyrus Khojasteh1 1 Department of Drug Metabolism & Pharmacokinetics, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA 2 Department of Analytical Chemistry & DMPK, ChemoCentryx, Inc., 850 W Maude Ave, Mountain View, CA 94043, USA *Author for correspondence: Tel.: +1 650 225 6808 [email protected] ‡ Authors contributed equally

CL int, trans uptake + CL int, pass

CL int, met + CLint, pass + CL int, trans eff

part of

Equation 2

Future Med. Chem. (2014) 6(18), 1995–2012

ISSN 1756-8919

1995

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh In this equation, CLint, app is the apparent hepatic intrinsic clearance, and CLint, met, CLint, trans uptake, CLint, and CLint, trans eff are the intrinsic clearance due to pass metabolism, active transporter-mediated uptake, passive permeability and active transporter-mediated efflux, respectively. When passive permeability does not limit a compound’s exposure to tissue (CLint, >> CLint, met, trans uptake and trans eff ), then CLint, app in pass Equation 2 equals the metabolic intrinsic clearance (CLint, met). However, if permeability (CLint, pass) is very low and active efflux (CLint, trans eff ) is also negligible, Equation 2 is solely dependent on the rate of active transporter-mediated uptake (CLint, trans uptake) [11] . In other words, a compound, regardless of its metabolic stability in the liver, is expected to have a clearance that is significantly lower than liver blood flow if it has low passive permeability and is not an active uptake substrate. Hepatocytes are not routinely utilized in drug discovery to measure the apparent permeability (Papp) of small molecule compounds. Instead, model cell systems such as Caco-2, Mardin-Darby canine kidney (MDCK) cells and pig kidney epithelial cells (LLC-PK1) are widely used to assess these data in small molecules. Despite their different cell origins, these cell lines share many common epithelial cell characteristics [13] . Transport across the cell membrane from the apical to basolateral side of MDCK cells is a good qualitative estimate of passive permeability because transporter activities are relatively low in MDCK cells [14,15] . Furthermore, compounds are typically dosed at a fairly high concentration (10 μM) in the permeability assay, and the assumption is that active uptake or efflux transporter involvement is saturated. A sigmoidal-like relationship exists between Papp from MDCK or Caco-2 cells and the percent of drug absorbed in human after an oral dose [14] . In general, there is a good correlation between Papp values from the two cell lines. Therefore, we selected 33 compounds from different early small molecule drug discovery programs to find a qualitative Papp cutoff point that could differentiate between compounds governed by permeability-rate-limited clearance (low Papp) and those by perfusion rate-limited clearance (high Papp). The 33 compounds were all weak bases and neutrals with diverse structures and Papp values ranging from 0.1 to 17 × 10 -6 cm/s in MDCK cell lines. Furthermore, recent advances have been made to measure the rate of active transporter-mediated uptake through either media loss or a modified oilspin method in suspended hepatocytes [16,17] . The latter method, though laborious, takes active transporter-mediated uptake, efflux and passive permeability into consideration and typically generates more reproducible data than the standard method; how-

1996

Future Med. Chem. (2014) 6(18)

ever, this method is not suitable for early drug discovery because of its low throughput. We developed a high-throughput assay that can quickly screen compounds to determine whether they are subject to active uptake in cryopreserved hepatocytes. This method is ideal for the early drug discovery setting. Together with Papp values from MDCK cells and metabolic stability from either microsomes or hepatocytes, this high-throughput temperature-dependent uptake assay will aid early drug discovery scientists assess the major clearance pathway(s) and gain a clearer view of the complex permeability, transporter and metabolism interplay. Here we propose a practical clearance classification for discovery compounds dosed intravenously since data to determine whether a compound is directed by perfusion- or permeability-limited clearance is not readily available. By incorporating the Papp data with information from the high-throughput uptake assay, we intend to develop a new permeability-based hepatic clearance classification system (HepCCS) to categorize and simplify the complex interactions between permeability, transporters and metabolism and provide guidance to early drug discovery projects in addressing active uptake transporter liability. Methods Materials

All reagents used were of the highest grade commercially available. Atorvastatin, pravastatin, quinine, quinidine, bovine serum albumin (BSA), mineral and silicone oil and Krebs-Hensleit buffer modified were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Mardin-Darby canine kidney cells wild-type (MDCKI-WT) cells were purchased from the National Institutes of Health (NIH; Bethesda, MA, USA). Millicell 24-well cell culture plates (PET membrane, 1 μm pore size) were purchased from Millipore (Billerica, MA, USA). Earle’s balanced salt solution, Hank’s balanced salt solution (HBSS), 4- (2-hydroxyethyl) -1-piperazineethanesulfonic acid buffer and Eagle’s minimum essential medium (EMEM) were purchased from Invitrogen (Carlsbad, CA, USA). Lucifer yellow dilithium salt (LY), and (±)-propranolol were purchased from Sigma-Aldrich (St Louis, MO, USA). Cyclosporin A was purchased from Alexis Biochemicals (San Diego, CA, USA). Cryopreserved hepatocytes from male Sprague Dawley rat (Lot MTN) and hepatocyte thawing medium were all purchased from Celsis/In vitro Technologies (Baltimore, MD, USA). Genentech early discovery compounds and the internal standards were synthesized in-house by the medicinal chemistry group. Each internal standard was selected on the basis of structural

future science group

Practical permeability-based HepCCS in drug discovery 

similarity to the compounds of interest. Durapore, Solvenert and 96-well, glass fiber filter plates were from Millipore Co. (Billerica, MA, USA). In vitro metabolic stability assays & LC/MS-MS methods

The metabolic stability of test compounds 1–33 were measured via substrate disappearance in both rat liver microsomes and hepatocytes according to previously published methods [18,19] . Briefly, 0.5 mg/ml of protein was used in microsomal incubations and 0.5 × 106 cells/ml was used for hepatocytes. The final concentration of test articles was 0.5 μM. At various time points (0, 10, 20 and 30 min for microsomes and 0, 1, 2 and 3 h for hepatocytes), aliquots were taken from the incubation and quenched with acetonitrile containing internal standard (0.25 μM final). The amount of unchanged parent compound in the incubation was then analyzed. The generic sample analysis method from Halladay et al. [18] was applied to measure each test article’s concentration in the different matrices, except that a Triple Quad API 4000 mass spectrometer from AB SCIEX (Foster City, CA, USA) was used. In-house internal standards were chosen based on their structural similarity to the test article. During the LC–MS method development, chromatographic compatibility and mass spectrometric signal stability of the internal standard were optimized prior to and monitored during analysis. The physiological parameters used for rats were 44.8 mg of protein/g of liver weight for microsomes, 120 × 106 cells/ml for hepatocytes, 40 g of liver weight/kg body weight and 55.2 ml/min/kg liver blood flow rate. MDCKI cell culture

MDCKI cells were maintained at 37°C, 95% humidity and 5% CO2 in culture with EMEM and Earle’s balanced salt solution supplemented with 10% fetal bovine serum. Cells were subcultured once weekly in T75 flasks using 0.5% trypsin-ethylenediaminetetraacetic acid (EDTA) with two changes of media, one at mid-week and one at the end of the week. The cells were seeded at an initial concentration of 2.5 × 105 cells/ml in 24-well plates and allowed to grow for 4 days for all permeability experiments. Media was changed 2 days after seeding and the day prior to the experiment. Permeability assay

MDCK monolayers were equilibrated for 30 min in transport buffer (HBSS with 10 mM 4-(2-hydroxyethyl)1-piperazineethanesulfonic acid, pH 7.4) at 37°C with 5% CO2 and 95% humidity prior to the experiment. The permeability of each test compound was examined

future science group

Research Article

at 10 μM in the apical to basolateral (A-B) and basolateral to apical (B-A) directions. The dose solutions also contained the monolayer integrity marker LY (100 μM) and a final DMSO percentage of 1%. The apical and basolateral receiver chambers contained transport buffer. Samples were taken from the receiver chambers at 60, 90 and 120 min and analyzed directly by LCMS after addition of an internal standard. The generic sample analysis method from Halladay et al. [18] was applied to measure each test article’s concentration. The apparent permeability (Papp) was calculated using the following equation: Papp =

dQ dt

#

1 1 # C0 A Equation 3

where dQ/dt is the rate of compound appearance in the receiver compartment, C0 is the concentration in the donor compartment and A is the surface area of the insert. Fluorescence from the LY marker was quantitated on a SpectraMax M5 Microplate reader (Molecular Devices, Sunnyvale, CA, USA) with excitation and emission wavelengths set at 425 and 520 nm, respectively. All runs included the following high, moderate and low permeable controls: metoprolol (15.9 × 10-6 cm/s), labetalol (5.6 × 10-6 cm/s) and nadolol (0.1 × 10-6 cm/s), respectively. Their Papp were in good agreement with those reported in literature [14] suggesting 10 μM was sufficient to get reliable passive permeability value. Only compounds with good recovery of the parent drug from the assay, that is, within 75–125% of the Key terms Perfusion-rate limited clearance: Compounds with high passive permeability are virtually in equilibrium between the blood and tissue, and they can, therefore, diffuse freely in and out of cells such as hepatocytes, in which they can be metabolized by various enzymes including cytochrome P450s, glucuronosyltransferases, flavin-containing monooxygenases and aldehyde oxidases [2] . When hepatocyte membranes offer virtually no barrier to a drug, the rate-limiting step for drug clearance is perfusion. Permeability-rate limited clearance: In contrast with perfusion rate-limited clearance, when passive permeability of a compound is low, the rate-limiting step now lies in penetration of the cell membrane, rather than in the delivering and removing of compounds to and from tissues via the blood flow [2] . Hepatic clearance classification system: Permeability-based hepatic clearance classification system. It divides intravenously dosed compounds into four classes based on their apparent permeability. It was found that compounds with low apparent permeability but high in vivo clearance are all subject to active hepatic uptake (class T). Transporter involvement becomes less important as apparent permeability increases (class M or C).

www.future-science.com

1997

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh

Table 1. Experimentally measured Papp values, calculated physicochemical properties, observed rat clearance and predicted (scaled) clearance (CL; r = 1 assumed) from rat liver microsomes or hepatocytes of 33 selected discovery compounds, all neutral or weak bases; compounds highlighted in bold were further tested for intrinsic uptake rate (CLint, influx). Compound

MDCK Papp LogP [× 10 -6 cm/s]

tPSA

HBDs/ HBAs

fu Rat plasma

Observed CL predicted in vivo rat CL from mic/hep† [ml/min/Kg] [ml/min/Kg]

Assigned Uptake substrate? ‡ class based on HepCCS

1

0.05. CL: Clearance; fu: Fraction unbound in plasma; HBA: Hydrogen bond acceptor; HBD: Hydrogen bond donor; HepCCS: Hepatic clearance classification system; hep: Hepatocytes; mic: Microsomes; ND: Not determined; Papp: Apparent permeability; tPSA: Total polar surface area. †

theoretical yield, were selected for the high-throughput temperature-dependent uptake assay. The recovery of compounds was calculated as the percent of analyte remaining in both the donor and recipient chambers at the final time point compared with the initial loading amount. Assessment of hepatocyte active uptake via the oil-spin method

The method from literature [17] was used to determine intrinsic active uptake clearance in cryopreserved rat hepatocytes (>80% viability), except that the test compounds were tested at a higher concentration (1 μM, 0.1% DMSO final) and incubated for slightly longer times (up to 150 s). Cryopreserved instead of fresh hepatocytes were used in the study to ensure the data consistency between different batches. Atorvastatin and pravastatin were used as positive controls. Compounds with ‘low’ (1, 5, 6 and 8) or ‘high’ passive permeability (25, 26 and 30) were initially selected to test the hepatocyte active uptake rate (Tables 1 & 2) . All samples were tested at both 37°C and 4°C in triplicates to assess temperature-dependent transport. Development of the high-throughput temperature-dependent uptake-filtration method

Prior to the development of an effective filtration method, 96-well filtering plates from different vendors were optimized in the presence and absence of pre-coated 4% BSA to test recovery of all test compounds (1 μM) with or without a cold buffer wash. Only compounds with greater than 75% recovery were further tested for

future science group

uptake, so no additional testing of nonspecific binding was performed. Cryopreserved rat hepatocytes (2 × 106 cells/ml final, >80% viability) were thawed, suspended in Krebs-Hensleit buffer and mixed with an equal volume of buffer containing only the test articles (1 μM, 0.1% DMSO final). For optimization and validation, an aliquot of 80 μl cell suspensions was transferred to a Durapore 96-well plate pre-coated with 4% BSA at 30, 60 and 90 s post-incubation at 37°C and filtered through the semi-permeable membrane to a 96-well receiver plate to collect the filtrate under vacuum. The vacuum was kept at or below 5 mmHg to minimize clogging of the membranes. The remaining cells on the membrane were first washed twice by cold buffer (100 μl × 2) and then lysed by methanol (100 μl). The substrate concentrations in the filtrate and lysate from the 37°C and 4°C experiments were measured by LC– MS. Standard curves were prepared in the corresponding matrix to determine the compounds’ concentration. Active uptake intrinsic clearances for atorvastatin, pravastatin and compounds 1 and 8 were measured and compared with that from the oil-spin method mentioned above using the same slope-intercept extraction method from Watanabe et al. [17] . After validation of the uptake-filtration method, only one incubation time point, 90 s, and the concentrations of each compound in the lysate at 37°C and 4°C were evaluated. OCT1-HEK cell culture & uptake measurement

Human OCT1 transfected HEK cells and empty-vector transfected HEK cells obtained from Professor KM Giacomini (UCSF, CA) were cultured in Dulbecco’s modified EMEM (Invitrogen, Carlsbad, CA, USA)

www.future-science.com

1999

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh

Table 2. Scaled intrinsic clearance from the sum of active uptake and passive permeability (CLint, influx) of selected compounds, including controls (atorvastatin and pravastatin) in cryopreserved rat hepatocytes.  Compound

Scaled CLint, influx at 37°C: Scaled CLint, influx at 37ºC: oil-spin method high-throughput method† (ml/min/g of liver) (ml/min/g of liver)

Scaled CLint, met Reported CLint, influx (ml/min/g of liver)‡  (ml/min/g of liver)

1

11.1

11.4

0.85

 

Pravastatin

1.04

2.54

 

2.59 (fresh)

5

1.20

 

1.12

 

6

4.80

 

0.33

 

8

4.57

3.28

< 0.1

 

Atorvastatin 12.6

10.5

 

23.1 (fresh)

25

Unable to determine

§

 

0.77

 

26

Unable to determine

§

 

0.87

 

30

Unable to determine §

 

0.37

 

The oil-spin method and intrinsic clearance calculations were taken from Watanabe et al. [16] . Time points 30, 90 and 150 s were taken. Assay optimized using Durapore 96-well plate pre-coated with 4% BSA and washed 2× with cold buffer. ‡ CLint, met from rat hepatocytes. A physiological scaling factor of 120 × 10 6 cells/g of liver was used for rat. § Slopes were approaching 0. CLint, influx: Intrinsic clearance from the sum of active uptake and passive permeability; CLint, met: Intrinsic clearance due to metabolism. †

supplemented with 10% fetal bovine serum and 200 μg/ml hygromycin B at 37°C with 5% CO2 and 95% humidity. The uptake studies were carried out between 48 and 72 h after seeding the cells at a density of 5.0 × 105 cells/well in 24-well tissue culture plates coated with poly-d-lysine and laminin. Cells were washed and incubated at 37°C with compound 1 (1 μM final) in 1 ml of HBSS for 1, 2, 5 and 10 min. The incubation was stopped by rinsing the cells with 1 ml of ice-cold HBSS buffer. The cells were sonicated with 200-μl HBSS for 5 min, followed by protein precipitation with an equal volume of acetonitrile. The supernatant of the samples was analyzed by LC–MS to determine the uptake of compound 1 into the cells. Inhibition of OCT1 uptake of compound 1 by quinine

To determine the half-maximal inhibitory concentration (IC50) of OCT1 uptake of compound 1, various concentrations of quinine, an OCT1 inhibitor, were added to compound 1 (1 μM final concentration) and incubated at 37°C for 5 min. The incubation was stopped by rinsing the cells twice with 1 ml of ice-cold HBSS buffer. After sonicating the cells with 200 μl HBSS for 5 min, an equal volume of acetonitrile was added for protein precipitation and the supernatant was analyzed by LC–MS. Determination of protein concentration

For each plate used in the OCT1 uptake study, three wells, seeded on the same day and with the same cell

2000

Future Med. Chem. (2014) 6(18)

density as all the wells used for the uptake study, were saved for protein analysis. Cells were washed with HBSS buffer and then lysed with 0.5 ml of ProteaPrep cell lysis solution (Protea Biosciences, Morgantown, WV). After 1 h, the solution was neutralized with 0.5 ml HBSS buffer, and 100 μl of the lysed cells were used to determine protein concentration by the standard Bradford protein assay (Thermo Scientific, Rockford, IL, USA) with BSA as a standard. In vitro data analysis

For active uptake rate measurements in hepatocytes, either areas under the curve from 0 to 150 s (AUC0-t) or, in the case of a single time point, lysate concentrations of each test article in triplicate, were compared at 37°C and 4°C. The student t-test was used to analyze the statistical difference between the two groups. p-values of less than 0.05 were considered statistically significant. Uptake values from the OCT1 transfected human embryonic kidney 293 cells (HEK-OCT1) cells are presented as mean ± standard deviation (SD) with a minimum of three wells used to generate each data point. IC50 values for quinine uptake into the cells was estimated by a sigmoidal inhibition model and was fitted to the following equation by nonlinear regression:

V =

V0

`l

+ log `

1 IC 50

n

jj

Equation 4

future science group

Practical permeability-based HepCCS in drug discovery 

where V is the uptake of compound 1 in the presence of the inhibitor (quinine), V0 is the uptake of compound 1 in the absence of the inhibitor, I is the inhibitor concentration and n is the hill slope. Determination of pharmacokinetic parameters of compounds 1–33 in rat & compound 1 in Oct1/2 constitutive knockout vs wild-type mice

All in vivo studies were carried out in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the USNIH. Briefly, for pharmacokinetic experiments conducted in rat, the animals were fasted overnight and given an intravenous bolus dose of test compound dissolved in either 10% DMSO/10% Cremophore EL in saline or 35% PEG400/65% water via jugular vein cannula. Blood samples were collected via the femoral artery, and plasma was isolated immediately in EDTA blood collection tubes. Plasma samples were frozen and stored at -80°C prior to analysis. Pharmacokinetic experiments performed in female Oct1/2 double targeted mutation knockout (KO) and wild-type (FVB/N) mice were 9–14 weeks of age (Taconic Farms, Hudson, NY; N = 3 for each group). Each animal was fasted overnight and given a 2 mg/kg intravenous bolus dose of compound 1 in the tail vein. A solution formulation of compound 1 containing 40% PEG400 in water was prepared, and the dosing volume was 5 ml/kg. Animals were kept in a temperature controlled environment with a 12 h light and 12 h dark cycle. Water was supplied ad libitum. Blood samples (15 μl) were collected at serial time points at pre-dose, 0.03, 0.25, 0.5, 1.0, 3.0 and 6.0 h post-dose via tail nick. EDTA solution (4 × the blood volume) was added to each of the blood samples, which were immediately frozen and stored at -80°C pending analysis. The blood concentration of compound 1 was analyzed after thawing and protein precipitation by acetonitrile treatment. Standard curve of compound 1 was constructed in diluted blood (same dilution factor) containing EDTA to correct for the any possible ion suppression due to EDTA. Pharmacokinetic analysis

Intravenous pharmacokinetic parameters were estimated using the WinNonLin software package (Pharsight, Mountain View, CA). Total blood clearance (CLiv), the apparent terminal half-life (t1/2) and steady state volume of distribution (Vss) of compound 1 was calculated based on Lau et al.’s method [20] . Simulation of plasma & intracellular water in liver concentration profiles based on permeability, transporter & metabolism interplay

To illustrate the combined influences of permeability, transporter and metabolism on drug plasma clearance

future science group

Research Article

and concentration–time profiles, a simplified threecompartment physiological model described in a recent publication was applied for simulations [21] . The central compartment includes blood and peripheral tissues except liver that are assumed to be in instantaneous equilibrium with systemic blood. The second and third compartments represent the liver as an elimination organ. The liver is considered a two-compartment system with instantaneous equilibration between organ blood, extracellular water, organ membranes and connective tissue, while the distribution between extracellular and hepatocyte intracellular water (where drug elimination occurs) is provided by passive diffusion and hepatic active uptake and efflux. A detailed description of the model is included in the Appendix. The described model was then applied to simulate the concentration profiles of a theoretical drug in a rat. Two opposite scenarios were considered: a metabolically stable compound and a labile compound. The compounds differed by 100-fold in stability in the liver, which was simulated by changing the rate of passive diffusion from low (100 ml/min, assumed) to high (5 × 106 ml/min, a value from the literature) [21,22] . The impact of active uptake and efflux transporters was also examined under a similar scenario. Protein binding and other physicochemical properties of the test article were assumed to be the same when comparing the two opposite situations. Sinusoidal efflux (from organ to blood) was considered to be at a minimum. Different classes of drugs were assigned according to the results of the simulation. The physiological parameters used in the simulations are shown in the Supplementary Information. Results Metabolic stability of 33 early discovery compounds & permeability in MDCK parental cells

Thirty-three compounds with very diverse chemical structures from different discovery programs at Genentech were selected to be tested for metabolic stability and passive permeability. Table 1 summarizes the Papp in MDCK cells (range between 0.1 and 17 × 10 –6 cm/s) as well as physicochemical properties including calculated LogP (0.55–3.7), total polar surface area (tPSA; 70–134), number of hydrogen bond donors (HBDs; 1–4) and hydrogen bond acceptor (HBAs; 5–12), measured fraction unbound (fu) in rat plasma and metabolic stability in rat microsomes (1–45 ml/min/kg) and rat hepatocytes (> CLint, pass and met; Table 4, class T, case II; Figure 2E). The rate limiting step in the disappearance of these compounds in plasma is mainly dependent on the rate of active plus passive influx into hepatocytes. This concept is demonstrated with our model compound 1. In the absence of Oct uptake transporters, total blood clearance was reduced by 50% versus the wild-type animals. Therefore, to correctly predict the clearance of these compounds in human, transportermediated uptake rate should be determined in human hepatocytes [12] . An accumulation of the parent drug in hepatocytes is also expected if the compound has a high metabolic stability (Figure 2C), and, in this

2006

Future Med. Chem. (2014) 6(18)

case, the hepatic intracellular water drug concentration is expected to be several fold higher than that in plasma immediately after dose. Enoxacin, a metabolically stable compound that is subject to active uptake did not cause any significant inhibition to theophylline oxidation in microsomes up to 1000 μM despite a competitive K i value of 120 μM in hepatocytes [28] . By contrast, an accumulation of metabolites is likely if the compound has low metabolic stability and if the metabolites are not rapidly removed from the cell by active efflux or passive diffusion. If a compound undergoes significant bioactivation or biotransformation after accumulation in the liver to generate either a reactive metabolite or an inhibitory metabolite against cytochrome P450 enzymes, then underprediction of bioactivation or of drug–drug interactions is likely because of an underestimation of inhibitor concentration as determined in plasma. In this case, the liver or intracellular drug concentration should be used if efflux of the parent drug or the perpetrating metabolite is negligible: Kp u =

CL int, trans uptake + CL int, pass CL int, pass

= Rate in = Rate out

6 cell @

6 medium @

Equation 5

Kpu is a measure of the relative ratio of drug concentration inside the cell versus that in the medium. Sensitivity analysis suggests that a more than 1000fold higher intracellular drug concentration compared with plasma concentration is expected when CLint, pass approaches 0.001 ml/min/million cells and CLint, trans uptake approaches 10 ml/min/million cells [11] . Class P: compounds with limited passive permeability & no active uptake – low in vivo clearance

The key differentiating characteristics of class P compounds are their low apparent permeability in MDCK cells and the fact that they are not subject to significant active uptake in hepatocytes, or the uptake rate is slow. The main driver for in vivo clearance for this class would be limited by the rate of permeation, and, therefore, the use of Equation 1 to predict the clearance of such a compound would be erroneous if the compound were incubated in microsomes, since microsomes have no membrane barrier. Protein binding is often speculated to be the cause of lower than expected in vivo clearance for metabolically unstable compounds; however, this cannot fully explain the disconnect in in vitro–in vivo clearance predictions. We observed two discovery compounds (G1 and G2) with virtually identical structures, metabolic stability (67 vs 66 ml/min/kg in mouse hepatocytes)

future science group

Practical permeability-based HepCCS in drug discovery 

and very high plasma protein binding (f u ≤ 0.01) and apparent permeability (4.4 vs 1.9 × 10 -6 cm/s), but the former is likely a class P molecule with a blood clearance of 7.2 ml/min/kg and the latter a class T molecule with a blood clearance of 51 ml/min/kg in mouse (Table 5) . G2 turned out to be an OCT1 substrate in human and an uptake substrate in mouse hepatocytes, but not G1. The only difference between the two molecules is a chloro-substituted aminopyrimidineon

Research Article

G1 and an unsubstituted aminopyrimidineon G2. The chloro-substitution in G1 possibly interfered with the hydrogen bond donating capacity of the neighboring amino group, thus eliminating OCT1 transporter recognition. The crystal structure of the OCT1 transporter is not readily available, so future work to examine the free energy of binding of these molecules to the transporter is needed to confirm this hypothesis. Nevertheless, these data taken together suggest that in

Table 4.  Permeability-based Hepatic Clearance Classification System (HepCCS). HepCCS

In vivo 

In vitro (suspended hepatocytes)

Class M CLint, pass >> CLint, uptake or efflux

Plasma

Medium Cell Parent drug

Parent drug CLint, pass

Unchanged drug

CLint, pass

Unchanged drug

Liver Efflux of parent drug into bile is insignificant; efflux of metabolites might be significant

Class T Case I: CLint, uptake >> CLint, pass or efflux

Plasma

Medium Cell Parent drug

Parent drug Unchanged drug

CLint, uptake CLint, pass

CLint, uptake

Unchanged drug

CLint, pass

Liver Efflux of parent drug into bile is insignificant; accumulation of parent drug or metabolite expected (if unstable and not excreted by efflux)

  Class T Case II: CLint, uptake or CLint efflux >> CLint, pass CLint, met and CLint, efflux may compete for parent drug in the liver, however, the ratedetermining step is the initial uptake.

Plasma

Medium Cell CLint, uptake

Parent drug Unchanged drug

CLint, pass

Parent drug

CLint, uptake

Unchanged drug

CLint, efflux

Liver CLint, efflux

 

Efflux of parent drug into bile is significant

The column on the left depicts the liver (triangle) in vivo and the column on the right depicts suspended hepatocytes (oval) in vitro. The size of the arrows represents the rate of movement across membranes or active secretion into bile. CLint, efflux: Intrinsic clearance due to active efflux; CLint, met: Intrinsic clearance due to metabolism; CLint, pass: Intrinsic clearance due to passive diffusion; CLint, uptake: Intrinsic clearance due to active uptake.

future science group

www.future-science.com

2007

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh

Table 4.  Permeability-based Hepatic Clearance Classification System (HepCCS) (cont.) HepCCS

In vivo 

Class P CLint, uptake, CLint efflux and CLint, pass are all low

In vitro (suspended hepatocytes) Plasma

Medium

Parent drug

Parent drug CLint, pass

Unchanged drug

Unchanged drug

CLint, pass Cell

Liver

 

Efflux of parent drug into bile might be significant

  Class C (for combined) CLint, pass is at a reasonable level but CLint, uptake is comparable to or higher than CLint, pass

Medium

Plasma

Parent drug

Unchanged drug

 

Parent drug

CLint, pass

CLint, pass CLint, uptake

CLint, uptake

Unchanged drug

CLint, pass or efflux Liver

 

CLint, pass or efflux Cell

Efflux of parent drug into bile might be significant; efflux of metabolites might be significant

The column on the left depicts the liver (triangle) in vivo and the column on the right depicts suspended hepatocytes (oval) in vitro. The size of the arrows represents the rate of movement across membranes or active secretion into bile. CLint, efflux: Intrinsic clearance due to active efflux; CLint, met: Intrinsic clearance due to metabolism; CLint, pass: Intrinsic clearance due to passive diffusion; CLint, uptake: Intrinsic clearance due to active uptake.

addition to plasma protein binding and metabolic stability, the rate of permeation (both passive and uptake) all play a role in the clearance of these two compounds. From a drug target point of view, class P molecules are not attractive for drug development, especially if the enzyme or receptor is located intracellularly, as the molecule might have difficulties reaching the drug receptor. However, class T molecules likely shift to become class P molecules in the absence of active transporters and, therefore, application of this concept to transporter KO animals can be an invaluable tool to study the impact of permeability in clearance. Finally, hepatocyte incubations of class P molecules should provide more accurate data for clearance prediction than microsomal incubations since hepatocytes contain a membrane barrier. Class C: compounds with high passive permeability & active uptake

This is a rare but interesting fourth class that was formulated after surveying the literature, and it is possible that the active uptake rate of these compounds in hepatocytes is comparable to or significantly higher than their passive permeation rate. For example, the overall influx rate for indomethacin is

2008

Future Med. Chem. (2014) 6(18)

higher (599 μl/min/10 6 cells) than that for atorvastatin (375 μl/min/10 6 cells); however, passive permeability for the former accounted for about half of the overall influx rate (237 μl/min/10 6 cells) [29] . In addition, while these compounds are reportedly extensively metabolized in human [1] , it is not known if their rate of active uptake is significantly higher than their rate of metabolism. However, because of their superior permeability (compared with class T drugs), these drugs are expected to diffuse out of the cell freely after active uptake. The compound’s liver and plasma concentrations should reach equilibrium rapidly after dosing and, therefore, perfusion, and ultimately metabolism, rather than permeability, will dictate its clearance in vivo. For example, if a metabolically unstable and highly permeable compound (class M) with a hepatic clearance (CL h) of 39.8 ml/min/kg (Figure 2B) is involved in uptake (class C), then the CL h is increased by only 22% to 48.8 ml/min/kg (Figure 2I) . Conclusion Unlike in drug development, the volume of compounds synthesized daily in drug discovery presents a challenge to collecting a comprehensive list

future science group

Practical permeability-based HepCCS in drug discovery 

Research Article

Table 5. Comparison of the properties of two structurally related analogs. In silico, in vitro and in vivo parameters

G1

G2

Calculated LogP

4.9

4.5

tPSA

98

98

HBDs/HBAs

3/8

3/8

CLblood predicted from mouse microsomes

67 ml/min/kg

66 ml/min/kg

fu in mouse

≤0.01

≤0.01

MDCK Papp (A→B)

4.4 × 10 -6 cm/s

1.9 × 10 -6 cm/s

Observed in vivo mouse CLblood

7.2 ml/min/kg

51 ml/min/kg

OCT1 substrate?

No

Yes

Uptake in mouse hepatocytes?

No

Yes

Permeability classification system

Class P (see discussion)

Class T

R = identical units. CLblood: Blood clearance; HBA: Hydrogen bond acceptor; HBD: Hydrogen bond donor; fu: Fraction unbound; MDCK: Madin-Darby canine kidney cells; tPSA: Total polar surface area.

of ADME-related data for every molecule, as many will not reach the clinic. Furthermore, many novel chemical entities have not been optimized to have ‘drug-like’ properties and are not well behaved like literature drugs. When in vivo clearance deviates from the in vitro-based prediction, quick decisions have to, therefore, be made on the basis of a limited amount of data. Discovery scientists use this limited set of information to intelligently cherry pick the assays to guide the design of the next set of molecules for medicinal chemists. Ultimately, good in vitro–in vivo correlations in preclinical species will increase the overall confidence in human clearance predictions. When a disconnect is seen in clearance predictions, the reasons for such behavior should be evaluated at an early stage in drug development to determine whether a unique clearance pathway is present in the preclinical species. The intent of the HepCCS is to provide a simplified picture of the complex permeability, transporter and metabolism mechanisms involved in the clearance of a drug [9] . Clearly, membrane composition and transporter expression level (e.g., blood–brain barrier vs hepatocytes), loss of transporter activity during hepatocyte isolation and storage, sinusoidal efflux and species differences will also need to be factored in. The objective in developing this system was to use Papp values from MDCK cells, or another in vitro cell system with low transporter expression, to determine a qualitative ‘cutoff ’ that can differentiate between perfusion-rate limited (high Papp) and permeability-rate limited (low Papp) clearance for early discovery compounds. When significant underprediction of clearance occurs for highly permeable compounds (e.g., compounds such as 18–25, 27–29, 31 and 33 in Table 1) , parameters such as blood sta-

future science group

bility, blood to plasma ratio, substrate concentration [S] tested in the depletion assay (i.e., [S] is not below Michaelis-Menten constant, K m), extrahepatic metabolism and non-cytochrome P450-mediated metabolism should be examined before hepatic uptake is considered. The prediction is also possibly limited by CLint, trans, uptake that can be experimentally measured because of the inherent loss of transporter activity during hepatocyte isolation and cryopreservation. Before any attempt on in vitro–in vivo correlation is undertaken, it is important to understand the limitations of any in vitro assay. For example, compounds might be misclassified if they have high nonspecific binding to the apparatus or cell membranes, or poor recovery in the MDCK permeability assay. In addition, the impact of sinusoidal efflux (from organ to blood), which would decrease plasma clearance in vivo, is not considered in our model since our objective was to understand the reasons behind underprediction of in vivo clearance from in vitro data. On the basis of our data, compounds with MDCK Papp values less than 3 × 10 -6 cm/s are predicted to be driven by permeability-rate limited clearance and undergo uptake in hepatocytes, while those with Papp values greater than 5 × 10 -6 cm/s are predicted to be driven by perfusion-rate limited clearance and not undergo uptake. Predicting whether a substrate is subject to active uptake is difficult, however, when the Papp value is in the range of 3 to 5 × 10 -6 cm/s. Key term In vitro–in vivo correlation: In vitro–in vivo correlation used in this manuscript is defined as a predictive mathematical model describing the relationship between in vitro metabolic and/or transporter-mediated clearance and in vivo clearance after protein binding correction.

www.future-science.com

2009

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh

Key term Permeability–transporter–metabolism interplay: The concept of metabolism-transporter interplay was first described by Professor Leslie Benet to decipher the compensatory mechanisms in drug disposition as changes in elimination pathways [1] . Complementary to this concept, we propose that there is one more parameter important in the interplay, that is, passive permeability, which can either overcome or mask the importance of active transporters in some circumstances leading to elimination of xenobiotics by liver.

It is interesting that Huang et al. [30] , using mostly acids as test compounds, also proposed a Papp of 5 × 10 -6 cm/s from of MDCK passive permeability assay as a cutoff for active uptake transporter-mediated clearance. Hence, if used properly, this new classification system will offer users guidance and clarity of understanding of the complex permeability, transporter and metabolism interplay that are involved in drug clearance. Importantly, the HepCCS can aid in answering the question that if the clearance mechanism is understood in preclinical species and can be predicted using a suitable in vitro system, can one ultimately predict how a new molecular entity will be cleared in human? Future perspective Because of the fast moving environment in drug discovery, a practical and qualitative permeability-based classification system will help medicinal chemists and DMPK scientists in the future understand the potential mechanistic causes of the in vitro–in vivo clearance prediction disconnect early. Perhaps the most popular strategy used today to improve pharmacokinetics (PK) of new chemical entities is to fix metabolic liabilities alone. Often times preclinical species are used to screen

PK profiles because the in vitro–in vivo disconnects in clearance are not well understood. The reason for such bias is mainly because while there are plenty of literature examples of how to deal with metabolism-limited clearance but if metabolism is not the rate-determining step then there is very little guidance on what issues to fix for medicinal chemists. This is further complicated by our inability to recommend suitable in vitro systems early for them to probe mechanistically and qualitatively the outcome of the permeability–transporter–metabolism interplay. In other words, our current understanding of the rate-determining step in clearance, whether it might be permeability, metabolism or transporter, is still incomplete. Hence, our simple and qualitative classification system will help future discovery scientists minimize the risk of PK surprises later on in development. For instance, if a molecule falls into class M and if there is significant underprediction of clearance, we recommend users of this system to explore the causes elsewhere such as blood stability, in vitro saturation of metabolism, blood to plasma ratio and extrahepatic metabolism. Furthermore, our system will also allow quick assessment of the type of molecules (i.e., metabolically stable class T molecules) that will likely result in rapid plasma clearance but high tissue accumulation, which could have pharmacological and toxicity ramifications. Supplementary data To view the supplementary data that accompany this paper please visit the journal website at: www.future-science.com/ doi/full/10.4155/FMC.14.141

Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial

Executive summary • An easy to use decision tree to quickly determine if a new chemical entity is going through perfusion or permeability-rate-limited clearance is proposed based on the outcome of measured apparent permeability (Papp) from Mardin-Darby canine kidney passive permeability assay. • If compound has high passive permeability (in general, Papp >5 × 10 -6 cm/s), clearance is likely perfusion-rate limited (class M). For molecules in this category, classic well-stirred model (Equation 1) is suitable for clearance prediction. • If compound has low passive permeability (in general, Papp ≤5 × 10 -6 cm/s), clearance is likely permeability-rate limited (class T). For molecules in this category, hepatic uptake due to active transport should be evaluated. Underprediction of clearance is likely to occur if one relies on the classic well-stirred model (Equation 1) . Instead, Equation 2 should be used. • Plasma concentration could be misleading for class T molecules as liver concentration could be several fold higher due to active uptake transporter. This could have important ramifications in drug–drug interactions and toxicity. • Structure activity relationship for organic cation transporter 1 (Oct1) is subtle based on our data (Table 5). Clearance by hepatic uptake transporter (Oct1) is an important elimination pathway based on our knockout animal study.

2010

Future Med. Chem. (2014) 6(18)

future science group

Practical permeability-based HepCCS in drug discovery 

Research Article

interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research

References

12

Umehara K, Camenisch G. Novel in vitro–in vivo extrapolation (IVIVE) method to predict hepatic organ clearance in rat. Pharm. Res. 29(2), 603–617 (2012).

••

The authors took drugs from BDDCS classes 1, 2, 3 and 4, applied in vitro measured hepatic uptake, metabolism and efflux and they were able to accurately predict in vivo clearance.

13

Thwaites DT, Hirst BH, Simmons NL. Passive transepithelial absorption of thyrotropin-releasing hormone (TRH) via a paracellular route in cultured intestinal and renal epithelial cell lines. Pharm. Res. 10(5), 674–681 (1993).

14

Irvine JD, Takahashi L, Lockhart K et al. MDCK (MadinDarby canine kidney) cells: a tool for membrane permeability screening. J. Pharm. Sci. 88(1), 28–33 (1999).

15

Huang L, Berry L, Ganga S et al. Relationship between passive permeability, efflux, and predictability of clearance from in vitro metabolic intrinsic clearance. Drug Metab. Dispos. 38(2), 223–231 (2010).

16

Soars MG, Webborn PJ, Riley RJ. Impact of hepatic uptake transporters on pharmacokinetics and drug-drug interactions: use of assays and models for decision making in the pharmaceutical industry. Mol. Pharm. 6(6), 1662–1677 (2009).

17

Watanabe T, Kusuhara H, Maeda K et al. Investigation of the rate-determining process in the hepatic elimination of HMG-CoA reductase inhibitors in rats and humans. Drug Metab. Dispos. 38(2), 215–222 (2010).



Methodology to the oil-spin method explained in detail and validated in vivo. It was shown that the rate-determining step for the clearance of statins tested is active hepatic uptake in rat and human.

18

Halladay JS, Wong S, Jaffer SM, Sinhababu AK, KhojastehBakht SC. Metabolic stability screen for drug discovery using cassette analysis and column switching. Drug Metab. Lett. 1(1), 67–72 (2007).

19

Jacobson L, Middleton B, Holmgren J et al. An optimized automated assay for determination of metabolic stability using hepatocytes: assay validation, variance component analysis, and in vivo relevance. Assay Drug Dev. Technol. 5(3), 403–415 (2007).

20

Lau YY, Okochi H, Huang Y, Benet LZ. Pharmacokinetics of atorvastatin and its hydroxy metabolites in rats and the effects of concomitant rifampicin single doses: relevance of first-pass effect from hepatic uptake transporters, and intestinal and hepatic metabolism. Drug Metab. Dispos. 34(7), 1175–1181 (2006).

21

Berezhkovskiy LM. The influence of hepatic transport on the distribution volumes and mean residence time of drug in the body and the accuracy of estimating these parameters by

The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.

Papers of special note have been highlighted as: • of interest; •• of considerable interest 1

Wu CY, Benet LZ. Predicting drug disposition via application of BCS. Transport/absorption/elimination interplay and development of a biopharmaceutics drug disposition classification system. Pharm. Res. 22(1), 11–23 (2005).

2

Rowland M, Tozer T. Clinical Pharmacokinetics Concepts and Applications (3rd Edition). Lippincott Williams & Wilkins, PA, USA, 109–117 (1995).

3

Obach RS. Prediction of human clearance of twenty-nine drugs from hepatic microsomal intrinsic clearance data: an examination of in vitro half-life approach and nonspecific binding to microsomes. Drug Metab. Dispos. 27(11), 1350–1359 (1999).

4

Jones HM, Houston JB. Substrate depletion approach for determining in vitro metabolic clearance: time dependencies in hepatocyte and microsomal incubations. Drug Metab. Dispos. 32(9), 973–982 (2004).

5

Mcginnity DF, Soars MG, Urbanowicz RA, Riley RJ. Evaluation of fresh and cryopreserved hepatocytes as in vitro drug metabolism tools for the prediction of metabolic clearance. Drug Metab. Dispos. 32(11), 1247–1253 (2004).

6

Austin RP, Barton P, Cockroft SL, Wenlock MC, Riley RJ. The influence of nonspecific microsomal binding on apparent intrinsic clearance, and its prediction from physicochemical properties. Drug Metab. Dispos. 30(12), 1497–1503 (2002).

7

Obach RS, Reed-Hagen AE. Measurement of Michaelis constants for cytochrome P450-mediated biotransformation reactions using a substrate depletion approach. Drug Metab. Dispos. 30(7), 831–837 (2002).

8

Pang KS, Weiss M, Macheras P. Advanced pharmacokinetic models based on organ clearance, circulatory, and fractal concepts. AAPS J. 9(2), E268–E283 (2007).

9

Pang KS, Maeng HJ, Fan J. Interplay of transporters and enzymes in drug and metabolite processing. Mol. Pharm. 6(6), 1734–1755 (2009).

10

Kusuhara H, Sugiyama Y. Pharmacokinetic modeling of the hepatobiliary transport mediated by cooperation of uptake and efflux transporters. Drug Metab. Rev. 42(3), 539–550 (2010).

11

Houston JB. Transporters and cytochrome P450 interplay in defining hepatic drug clearance. In: 4th European Cyprotex Drug Discovery Workshop. London, UK, 9 June 2010.

••

Concept of permeability, transporter and metabolism interplay is illustrated in simple to understand Power Point slides. Highly relevant to this paper.

future science group

www.future-science.com

2011

Research Article  Fan, Song, Berezhkovskiy, Cheong, Plise & Khojasteh 27

Amidon GL, Lennernas H, Shah VP, Crison JR. A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res. 12(3), 413–420 (1995).

28

Brown HS, Wilby AJ, Alder J, Houston JB. Comparative use of isolated hepatocytes and hepatic microsomes for cytochrome P450 inhibition studies: transporter-enzyme interplay. Drug Metab. Dispos. 38(12), 2139–2146 (2010).

29

Soars MG, Grime K, Sproston JL, Webborn PJ, Riley RJ. Use of hepatocytes to assess the contribution of hepatic uptake to clearance in vivo. Drug Metab. Dispos. 35(6), 859–865 (2007).

Paine SW, Parker AJ, Gardiner P, Webborn PJ, Riley RJ. Prediction of the pharmacokinetics of atorvastatin, cerivastatin, and indomethacin using kinetic models applied to isolated rat hepatocytes. Drug Metab. Dispos. 36(7), 1365–1374 (2008).

30

25

Feng B, Mills JB, Davidson RE et al. In vitro P-glycoprotein assays to predict the in vivo interactions of P-glycoprotein with drugs in the central nervous system. Drug Metab. Dispos. 36(2), 268–275 (2008).

Huang L, Chen A, Roberts J et al. Use of uptake intrinsic clearance from attached rat hepatocytes to predict hepatic clearance for poorly permeable compounds. Xenobiotica 42(9), 830–840 (2012).

••

26

Hallifax D, Houston JB. Saturable uptake of lipophilic amine drugs into isolated hepatocytes: mechanisms and consequences for quantitative clearance prediction. Drug Metab. Dispos. 35(8), 1325–1332 (2007).

Examined CLint, trans uptake in plated rat hepatocytes of 15 proprietary compounds and two controls (mostly acids) with low passive permeability in MDCK and found a good correlation between permeability and active hepatic uptake.

the traditional pharmacokinetic calculations. J. Pharm. Sci. 100(11), 5031–5047 (2011). 22

23

24

2012

Fagerholm U. Presentation of a modified dispersion model (MDM) for hepatic drug extraction and a new methodology for the prediction of the rate-limiting step in hepatic metabolic clearance. Xenobiotica 39(1), 57–71 (2009). Koepsell H, Lips K, Volk C. Polyspecific organic cation transporters: structure, function, physiological roles, and biopharmaceutical implications. Pharm. Res. 24(7), 1227–1251 (2007).

Future Med. Chem. (2014) 6(18)

future science group