Vol. 282, Issue 2, 715-726, 1997

Relationship between Chemical Structure and Physicochemical Properties of Series of Bulky Organic Cations and Their Hepatic Uptake and Biliary Excretion Rates¹

Groningen Institute for Drug Studies (GIDS),⁴ University Centre for Pharmacy, Department of Pharmacokinetics and Drug Delivery, University of Groningen, Groningen, The Netherlands (J.H.P., J.R., D.K.F.M.); and Research Group for Experimental Anesthesiology and Clinical Pharmacology, University Hospital, Department of Anesthesiology, Groningen, The Netherlands (J.H.P., J.R., J.M.K.H.W.)

	Abstract

Top Abstract Introduction Methods Results Discussion References

To obtain more insight in the relationship between physicochemical properties of cationic drugs and their hepatobiliary transport rate, a series of 12 aminosteroidal neuromuscular blocking agents (NMBAs), supplemented with data of four related NMBAs from the literature, were investigated in the isolated perfused rat liver. A significant correlation was found between plasma protein binding and the partition coefficient octanol/Krebs (log P), confirming results from the literature with other organic cations. Evidence was found for a saturable hepatic uptake of several NMBAs, indicating that carrier-mediated uptake processes are involved. Hepatic uptake rate was closely related to the lipophilicity of the compounds; the initial extraction ratio, the apparent clearance and the intrinsic clearance were significantly correlated to log P. We did not find a significant correlation between biliary clearance and lipophilicity in the current series of compounds. Pharmacokinetics analysis of perfusate disappearance and biliary excretion data revealed that a considerable fraction of the dose of these bulky organic cations is stored in the liver and seems to not be directly available for biliary excretion. This finding is in line with earlier observations showing a pronounced accumulation of this type of compounds in mitochondria and lysosomes.

	Introduction

Top Abstract Introduction Methods Results Discussion References

In the development of new short-acting NMBAs, the time course of action and their side effects are the principal parameters of interest (Baird and Viby-Mogensen, 1990; Savarese et al., 1995; Wierda et al., 1993; Wierda and Proost, 1992). Onset of action and duration of effect after intravenous administration are governed by pharmacokinetic and pharmacodynamic processes, which in turn are determined by the chemical structure and physicochemical properties (Neef and Meijer, 1984; Noy and Zakim, 1993; Rekker, 1977; Seydel and Schaper, 1982). Therefore, chemical structure, physicochemical properties, pharmacokinetics and pharmacodynamics should be studied to provide a rational basis for drug developmental research (Peck et al., 1992; Seydel and Schaper, 1982). Better knowledge of these relationships may also shed light on the basic mechanisms involved in membrane transport of such agents, (e.g., carrier-mediated uptake and biliary excretion in the liver).

Lipophilicity is the principal physicochemical property used to establish structure-activity relationship and structure-pharmacokinetics relationships (Neef and Meijer, 1984; Noy and Zakim, 1993; Rekker, 1977; Seydel and Schaper, 1982). Usually, lipophilicity is expressed as the partition coefficient between octanol and an aqueous Krebs' solution. Alternatively, the retention time over an HPLC column can be used as a measure of lipophilicity (Neef and Meijer, 1984; Tomlinson, 1975). This method can be applied easily to various quaternary and tertiary amines.

It has been shown previously that lipophilicity is a determinant factor for hepatobiliary transport of a series of monoquaternary ammonium (type 1) compounds, resulting in an increase of hepatic clearance with increasing lipophilicity (Neef and Meijer, 1984). The liver plays an important role in the distribution and elimination of more bulky cationic drugs, such as aminosteroidal NMBAs, and as a result, in the time course of action of NMBAs, as has been demonstrated in humans for vecuronium (Bencini et al., 1986). These bulky cations are supposed to be taken up via another carrier-mediated mechanism, as indicated by substrate specificity and inhibition studies; therefore, these bulky cations were categorized as type 2 organic cations (Meijer, 1976; Meijer et al., 1970; Mol et al., 1988). Experiments in cats with a portocaval shunt and liver exclusion and experiments with intraportal injection also demonstrated the predominant influence of the liver on the time course of action of NMBAs (Agoston et al., 1980; Bencini et al., 1985). Because the distribution, elimination and effect of drugs are related to the unbound concentration, protein binding may affect the potency and time course of action of drugs, as well as their hepatobiliary transport rates.

We investigated the relationship among chemical structure, lipophilicity, protein binding and pharmacokinetics in the isolated perfused rat liver of a series of 12 aminosteroidal NMBAs, supplemented with data of four related NMBAs from the literature. A better insight into this relationship may be helpful to reveal the factors governing the time course of action and potency of this class of drugs. Also, such studies may provide important model compounds for the functional characterization of cloned carrier proteins in the liver, such as OCT1 (Gründemann et al., 1994), OATP (Bossuyt et al., 1996a, 1996b), variants of P-glycoprotein (Oude Elferink et al., 1995) and the recently identified isoforms of MRP (Kartenbeck et al., 1996; Paulusma et al., 1996).

	Methods

Top Abstract Introduction Methods Results Discussion References

Chemicals. The NMBAs (fig. 1) and their putative metabolites, (i.e., their 3-OH-, 17-OH- and 3,17-di-OH-derivatives) were supplied by Organon Labs Ltd. (Newhouse, Scotland). The purity of the compounds, in the form of their bromides, was >98%, except for rocuronium bromide (purity, >95%). All other chemicals were of analytical grade and were obtained from commercial sources.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Chemical structures, including steroidal skeleton and list of compounds. R2 R3 R16 R17 Pancuronium N-Methyl-piperidino Acetyl N-Methyl-piperidino Acetyl Org 6368 N-Methyl-piperidino Acetyl N-Methyl-piperidino Hydrogen Pipecuronium 4-Dimethyl-1-piperazino Acetyl 4-Dimethyl-1-piperazino Acetyl Vecuronium Piperidino Acetyl N-Methyl-piperidino Acetyl Org 9489 Piperidino Acetyl N-Methyl-piperidino Propionyl Org 9487 Piperidino Acetyl N-Allyl-piperidino Propionyl Org 9453 Piperidino Acetyl N-Methyl-piperidino Butyryl Org 7617 Piperidino Acetyl N-Allyl-piperidino Butyryl Org 9616 Piperidino Acetyl N-Methyl-piperidino  -Butyryl Org 9991 Piperidino Acetyl N-Methyl-homopiperidino Butyryl Org 7268 Piperidino Hydroxyl N-Methyl-piperidino Acetyl Org 20297 Piperidino Hydroxyl N-Methyl-piperidino Butyryl Org 9955 Piperidino O-CO-NH-CH3 N-Methyl-piperidino Butyryl Org 9273 Morpholino Hydroxyl N-Methyl-piperidino Acetyl Rocuronium Morpholino Hydroxyl N-Allyl-pyrrolidino Acetyl Org 20059 Morpholino Hydrogen N-Methyl-piperidino Acetyl

Partition coefficient octanol/Krebs. The partition coefficient octanol/Krebs was determined as described by Neef and Meijer (Neef and Meijer, 1984) using n-octanol (saturated with water) and Krebs' solution without sodium bicarbonate (adjusted to pH 7.4 with sodium hydroxide and saturated with n-octanol). An aliquot of the octanol layer was evaporated to dryness and dissolved in the HPLC eluent. The concentrations in this solution and in the water layer were analyzed by HPLC as described below. The partition coefficient was calculated as the concentration in octanol divided by the concentration in the water layer.

Plasma protein binding. The binding to plasma proteins was determined in human plasma by means of an ultrafiltration system (Amicon micropartition system MPS-1 and membranes YMT30 with molecular weight cutoff of 30 kD, Amicon Corp., Danvers, MA), followed by HPLC analysis as described below. Undiluted human plasma was spiked with one of the compounds to a concentration of 2000 µg/liter (~ 3 µM). The loss of agent to the membrane was <10% in all experiments. The fraction unbound was calculated as the concentration in the ultrafiltrate divided by the concentration before filtration. Protein binding was expressed as the fraction bound and was calculated as 1 minus the fraction unbound.

Isolated perfused rat liver experiments. The isolated perfused rat liver experiments were carried out as described by Meijer et al. (1981) with some slight modifications. Male Wistar rats, weighing 220 to 320 g, were anaesthetized with pentobarbital sodium (60 mg/kg intraperitoneally) after a 16-hr fasting period. The bile duct was cannulated with PE tubing (i.d., 0.40 mm; o.d., 0.80 mm). After cannulation of the vena porta (PE tube, i.d., 1.57 mm; o.d., 2.08 mm), the liver was perfused with Krebs' solution to remove the blood. An outflow cannula (PE tube, i.d., 1.57 mm; o.d., 2.08 mm) was inserted in the vena cava superior, and the vena cava inferior and arteria hepatica were ligated. The liver was excised and placed in the perfusion apparatus. The recirculating perfusion medium (100 ml), containing 118 mM NaCl, 5 mM KCl, 1.1 mM MgSO₄, 2.5 mM CaCl₂, 1.2 mM KH₂PO₄, 25 mM NaHCO₃ and 2 g/liter glucose, 10 g/liter bovine serum albumin (Boseral, Organon Teknika), was constantly gassed with oxygen and carbon dioxide (95% O₂/5% CO₂). The perfusate flow was maintained at 3.5 ml/min/g liver at a hydrostatic pressure of 8 to 12 cm, the temperature was maintained at 37° to 38°C and the pH was maintained between 7.36 and 7.42. An infusion of sodium taurocholate (15 µmol/hr) was given to replace bile salts.

After a 30-min recovery period following the surgical procedure, 1 mg (~1.5 µmol) of the compound under study was added to the perfusion medium. During the experiment, the viability of the liver was checked by measuring bile flow and pH and flow of the recirculating perfusate. The perfusion medium was sampled during the experiment at 1- or 2-min intervals over 10 min and after 15, 20, 25, 30, 40, 50, 60, 80, 100 and 120 min. To prevent hydrolysis of the drug, the perfusate samples of 0.8 ml were acidified with 0.2 ml of 1 M sodium dihydrogenphosphate. Bile was collected at 5-min intervals for 30 min and at 10-min intervals for 120 min in an appropriate amount of 1 M sodium hydrogenphosphate and weighed. All samples were kept frozen at 20°C until analysis. It was ensured that the compounds were stable under the storage conditions.

Determination of the compounds and their hydroxy-derivatives in perfusate and bile homogenate. The determination of the compounds and their putative metabolites (3-OH, 17-OH and 3,17-di-OH analogs) in perfusate and bile was carried out by HPLC with postcolumn ion-pair extraction and fluorometric detection, as previously described (Kleef et al., 1993; Paanakker et al., 1987), with the following modifications. The pretreatment of the samples was carried out by a solvent extraction instead of the solid-phase extraction, using a phosphate buffer, pH 3.0, containing 0.44 mM DAS.

To 100 to 1000 µl of the sample (perfusate or bile) in a 20-ml glass-stoppered tube we added a solution of the internal standard (100-200 ng), 1.0 ml of extraction buffer and 6.0 or 7.0 ml of dichloromethane. After mixing for 15 sec on a Vortex mixer and centrifugation for 5 min, the upper layer was removed. The dichloromethane layer was poured into a smaller tube and evaporated to dryness under a nitrogen stream at 50°C. The residue was resolved in the eluent. Samples for the determination of the partition coefficient were injected without pretreatment.

The eluent consisted of 0.1 M sodium hydrogenphosphate and 0.11 mM DAS (in several experiments, 0.11 mM sodium heptane-sulfonate was included) in water/dioxane (16-19% dioxane) adjusted to pH 3.0 with phosphoric acid. After separation in the HPLC column, the eluent was extracted with dichloroethane, and the organic phase was led to a fluorimetric detector, operating at 385 nm (excitation) and 452 nm (emission). The retention times were within the range of 4 to 25 min. Calibration was carried out with samples spiked with the compounds and their hydroxy-derivatives in concentrations of 10 to 1000 µg/liter.

Concentrations and amounts of the drugs and metabolites refer to the parent compound as cationic entity, but administered doses refer to the weighed amounts. In the calculations, doses have been recalculated as the cationic equivalent.

Pharmacokinetic analysis. Perfusate concentrations and biliary excretion data were analyzed by nonlinear curve-fitting, using the program SimulFit (developed in our department and derived from programs used in earlier studies: Mol et al., 1992; Proost et al., 1993). The model is derived from the parallel-tube model (Bass and Keiding, 1988; Wilkinson, 1987; Winkler et al., 1973) and is depicted in figure 2. It consists of (1) a central compartment, composed of the perfusate reservoir with a volume V₁, (2) the sinusoidal space in the liver, perfused at flow rate Q, (3) a cytosolic compartment (i.e., the intracellular space in the liver) in which the drug is present after uptake by the hepatocytes and is available for excretion and metabolism, (4) a storage compartment (i.e., an intracellular space in the liver) in which the drug is temporarily stored, and (5) an excretory compartment (i.e., the biliary tree).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Pharmacokinetic model for the isolated perfused rat liver showing the pharmacokinetic model of the parent compound (top) and its metabolite (m) (bottom). V1 is the volume of the circulating perfusate; Q is the perfusate flow through the liver; Cin and Cout are the drug concentration entering and leaving the liver, respectively; is the average drug concentration in the sinusoids; 2 and 3 refer to the cytosolic compartment and the storage compartment in the liver, respectively; CLint is the intrinsic clearance from the sinusoidal space to the intracellular space; kab is the rate constant of transport from compartment a to compartment b; k20 is the rate constant of biliary excretion from compartment 2; kma is the rate constant of metabolite formation in compartment a and kab(m) is the rate constant of transport of the metabolite from compartment a to compartment b.

For compounds that are metabolized to their 3-desacetyl- or 17-desacetyl-derivatives, a similar model is assumed, as indicated in figure 2 (bottom). The formation of the metabolite is assumed to take place primarily in the liver compartments; however, some metabolite also may be formed in the perfusate due to chemical hydrolysis, which is known to take place at pH 7.4 and 37°C.⁵ Unless stated otherwise, all transport processes were assumed to be first order.

The kinetics of a drug in the model of figure 2 were described by a set of differential equations according to classic compartmental modeling. The rate of liver uptake was described as: Rate of uptake = CL_int (equation 1), where CL_int is the intrinsic uptake clearance (in ml/min), and is the average concentration of drug in the sinusoids according to the parallel-tube model:

(2)

where C_in is the perfusate concentration entering the sinusoids.

(3)

and C_out is the perfusate concentration leaving the sinusoids, which can be obtained from the mass balance of net drug uptake in the sinusoids:

(4)

where A₂ is the amount of drug in the liver (compartment 2).

From equations 1 to 4,  and C_out can be calculated by an iterative procedure if C_in and A₂, and the constants V₁, Q, CL_int, and k₂₁ are known.

The amount of drug excreted into bile during the interval t_i1 to t_i, Â_i, was obtained from:

(5)

where A_b(t) is the cumulative amount excreted into bile at time t, and t_lag is the time lag between the biliary excretion process at the canalicular membrane and the appearance at the sampling outlet.

The initial extraction ratio E, (i.e., the extraction at time 0 when the liver content is 0) was calculated from the following formula according to the parallel-tube model:

(6)

The apparent clearance from the central compartment CL_app was calculated as:

(7)

The biliary clearance CL₂₀ was calculated from:

(8)

where V₂ is the volume of compartment 2. Because this volume cannot be obtained by pharmacokinetic analysis, it was assumed that V₂ was equal to the total cytosolic volume of the hepatocytes, amounting 44.4% of the liver weight (Weibel et al., 1969).

In cases in which the uptake of drug in the liver does not obey first-order kinetics, the intrinsic clearance CL_int in equation 1 was replaced by the following expression according to Michaelis-Menten kinetics:

(9)

where V_max12 is the maximum liver uptake rate, and K_m12 is the perfusate concentration at which the rate reaches half of the maximum value.

In case of a saturable biliary excretion of the compound, the rate constant k₂₀ was replaced by the following expression according to Michaelis-Menten kinetics:

(10)

where V_max20 is the maximum biliary excretion rate, and K_m20 is the concentration in liver compartment 2 at which the biliary excretion rate reaches half of the maximum value.

For the presentation of data, the parameters CL_int, CL_app, CL₂₀, V_max12 and V_max20 were normalized for rat weight.

The perfusate concentration, the amount of drug in each compartment and the cumulative biliary excretion were obtained by numerical integration of the differential equations describing the model, with the initial condition that at time 0 A₁ equals the dose. Corrections were made for sampling of perfusate during the experiments for both the perfusate volume V₁ and the amount of drug in the perfusate (A₁) (Colburn et al., 1983). The time lag of the biliary excretion (equation 5) was estimated during the fitting procedure in a similar way as the model parameters. The numerical integration was performed by a fourth-order Runge-Kutta integration method (Press et al., 1986). It was confirmed that the numerical approximations gave results with an accuracy of at least four digits.

For each experiment the model parameters were estimated by minimizing the residual sum of squares (RSS), calculated from:

(11)

where n_p is the number of perfusate concentration measurements, C_i is the measured perfusate concentration at time t_i, _i is the calculated perfusate concentration at time t_i according to the mode, n_b is the number of intervals of the biliary excretion measurements, A_i is the measured biliary excretion over the interval t_i  1 to t_i and Â_i is the calculated biliary excretion over the interval t_i = 1 to t_i according to the model.

The correctness of the logarithmic transformation and weighting was tested by visual inspection of the graphs of the residuals plotted against time and against the concentration. Moreover, it is known that the relative error of the bioanalysis was almost independent of the concentration over the entire concentration range (Kleef et al., 1993; Paanakker et al., 1987).

The minimization procedure was performed by both the Simplex and Marquardt algorithms (Nelder and Mead, 1965; Press et al., 1986), using several sets of initial estimates, to avoid the occurrence of local minima. The fitting procedure stopped when the relative improvement of the residual sum of squares was <10¹⁰, and the relative change of each parameter was <10⁵. The standard errors of the estimated parameters was determined from the variance-covariance matrix (D'Argenio and Schumitzky, 1979; Draper and Smith, 1981; Sheiner and Beal, 1987; Veng Pedersen, 1977).

In each fitting procedure, the perfusate flow (Q) was set to the value measured during each individual experiment. First, the linear model was tested by fitting the following parameters: V₁, CL_int, k₂₁, k₂₀, t_lag, k₂₃ and k₃₂. If a metabolite was excreted into bile or was found in the perfusate, the formation of the metabolite in the liver (k_m2) and the transport of kinetics of the metabolite were included into a model by adding the parameters k_20(m), k_23(m), k_32(m), k_21(m) and CL_int(m); V₁ and t_lag were assumed to be the same for metabolite and parent compound. In this case, the perfusate concentration and biliary excretion of the metabolite were included in the calculation of the residual sum of squares (equation 11). If appropriate, formation of the metabolite in perfusate (k_m1), and in the storage compartment (k_m3) was tested. Finally, the fitting procedure was also carried out assuming that liver uptake and/or biliary excretion followed Michaelis-Menten kinetics according to equations 9 and 10.

The goodness of fit was expressed as the residual coefficient of variation calculated from:

(12)

when n is the total number of measurements used for calculation of RSS, and p is the number of estimated model parameters.

The choice between different models was based on the F test on the residual variance (Boxenbaum et al., 1974), accepting a more complex model as significantly better fitting if P < .05.

The outcome of the fitting procedure was accepted as a valid result only if (1) the residual coefficient of variation was <25%, (2) the relative standard error of each of the parameters was <40%, to avoid the problem of indistinguishability of models and parameters (Godfrey et al., 1994; Godfrey and Chapman, 1989; Jacquez, 1987), (3) there was no systematic deviation between the measured and calculated data and (4) the fitted value for V₁ did not deviate >15% from the measured perfusate volume.

The pharmacokinetic calculations were performed by the program SimulFit written in VisualBasic for MS-DOS (Microsoft Corp., Redmond, WA), using real numbers in double precision.

Correlation analysis. The correlation between various parameters was determined by standard linear regression analysis. Correlations were considered significant if P < .05. If indicated, x and/or y data were analyzed after logarithmic transformation.

	Results

Top Abstract Introduction Methods Results Discussion References

Chemical structure. The chemical structure of the investigated compounds is given in figure 1. The compounds are different at four substituents of the steroidal skeleton. Substitutions at R₂ and R₁₆ mimic the quaternary N-containing portion, whereas those at R₃ and R₁₇ correspond to the ester bond-containing domain of an acetylcholine molecule.

From the point of view of the chemical structure, Org 9453 can be considered the central compound. Seven compounds (Org 20297, Org 9955, vecuronium, Org 9489, Org 7617, Org 9616 and Org 9991) differ only at one position (3-, 16- or 17-) from this compound, and three compounds (pancuronium, Org 9487, and Org 7268) differ at two positions with Org 9453. Org 9616 is an isomeric form of Org 9453, differing only in the direction of the 16-substitutent.

Pancuronium, Org 6368 and pipecuronium are bisquaternary compounds. The remaining compounds have one quaternary and one tertiary amine group. The structure of pipecuronium differs from the other compounds in that the two quaternary nitrogen atoms are not attached to the steroidal skeleton but rather placed at the opposite side of the piperidine rings.

Partition coefficient octanol/Krebs. The partition coefficients octanol/Krebs are given in table 1. The bisquaternary compounds pancuronium and Org 6368 exhibit much lower partition coefficients than the tertiary compounds. The partition coefficient of pipecuronium could not be determined due to the extremely low concentration in the octanol phase; the estimated partition coeffcient was <0.02.

View this table: [in this window] [in a new window]   TABLE 1 Physicochemical data of the compounds

The 2-morpholino compounds are more lipophilic than their 2-piperidino analogs (Org 9273 vs. Org 7268). The 3-OH compounds Org 7268 and Org 20297 have much lower partition coefficients than the corresponding 3-acetyl compounds, vecuronium and Org 9453. Org 9955, containing a 3-carbamate group, takes an intermediate position. Org 20059, which does not contain a hydrophilic substituent at the 3-position, is more lipophilic than vecuronium.

The quaternary 16-N-substituent has also a mared effect on lipophilicity: the N-allyl compounds are more lipophilic than their N-methyl analogs (Org 7617 vs. Org 9453, Org 9487 vs. Org 9489 and rocuronium vs. Org 9273 (the latter also differ with respect to the ring at the 16-N position).

The partition coefficients of the 17-butyryl compounds Org 9453, Org 7617, Org 9616 and Org 9991 are markedly higher than that of the 17-propionyl compounds Org 9489 and Org 9487, which in turn are more lipophilic than the 17-acetyl compound vecuronium. The same shift in lipophilicity is found for the butyryl compound Org 20297 compared with the 17-acetyl compound Org 7268. Within the 17-butyryl compounds, Org 9991, containing the largest ring at the 16-N position, exhibits the highest partition coefficient.

Retention time in HPLC analysis. In table 1, the HPLC retention times are given for two different eluents (i.e., using 16% and 19% dioxane). Increasing the amount of dioxane in the eluent shortens the retention times considerably, allowing analysis of the more lipophilic compounds within an acceptable time interval.

Linear regression analysis of the logarithms of the partition coefficient and the retention time revealed a significant correlation for both eluent systems (see table 4).

View this table: [in this window] [in a new window]   TABLE 4 Correlation between log P and physicochemical and pharmacokinetic parametersa

Plasma protein binding. The binding of the NMBAs to human plasma proteins was determined by ultrafiltration. The results, given in table 1, show a significant correlation (see data in table 4) between the logarithm of the partition coefficient and the plasma protein binding of the compounds used in this study (Proost et al., 1995; Wierda et al., 1993). The same applies to the ratio of the fraction unbound and the fraction bound (fu/fb), which may be considered a measure of the dissociation constant of the drug-protein complex (Neef and Meijer, 1984).

For several compounds, the protein binding in the perfusion medium containing 1% bovine serum albumin was measured by the same method. The fraction bound to albumin was 0.01 for pancuronium, 0.09 for vecuronium and 0.03 for rocuronium.

Kinetics in the isolated perfused rat liver. The results of the experiments in the isolated perfused rat liver are summarized in tables 2 and 3.

View this table: [in this window] [in a new window]   TABLE 2 Pharmacokinetic parameters in isolated perfused rat liver

View this table: [in this window] [in a new window]   TABLE 3 Cumulative biliary excretion

Disappearance from the perfusate. The profile of the perfusate concentration on a logarithmic scale exhibits a biphasic pattern only for Org 7268. For several compounds, including Org 9489, Org 9487, Org 9453, Org 9991, Org 20297, Org 9955 and Org 20059, the perfusate concentration profiles show a slightly convex pattern, as exemplified in figure 3 for Org 9487. The remaining compounds disappeared from the perfusate by an apparent monoexponential decay, without a distinct slower phase. The fastest decay of the perfusate concentration was observed for rocuronium. The rate of disappearance of Org 9489, Org 9487, Org 9453, Org 7617, Org 9616, Org 9991, Org 9273 and Org 20059 was roughly comparable to that of vecuronium (Bencini et al., 1988; Mol et al., 1992). whereas the concentration decay of Org 7268, Org 20297 and Org 9955 was slower and comparable to that of Org 6368 (Mol et al., 1992).

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Perfusate concentration (open symbols, left axis) and biliary excretion rate (closed symbols, right axis) of Org 9487 in the isolated perfused rat liver (data from one single experiment). The lines represent the calculated perfusate concentration and biliary excretion rate.

For each compound, the perfusate concentration and biliary excretion data were analyzed by pharmacokinetic modelling as described in Methods. The results are summarized in table 2. The hepatic uptake process has been characterized by the initial extraction ratio, apparent clearance and intrinsic clearance.

The slightly convex perfusate concentration profiles observed for Org 9489, Org 9487, Org 9453, Org 9991, Org 20297, Org 9955 and Org 20059 were also fitted according to a model with a liver uptake process obeying Michaelis-Menten kinetics (equation 9). In most cases, this model fitted significantly better to the data. However, the standard errors of the estimates of V_max12 and, in particular, of K_m12, were large in several experiments, and the standard deviations between experiments with one compounds were also significant; therefore, these parameters can be estimated only roughly. In general, the values for V_max12 were in the range of 0.4 to 0.9 µmol/min/kg, and K_m12 values ranged from 0.7 to 6 µM. Neither V_max12 nor K_m12 appeared to be related to lipophilicity.

Biliary excretion and metabolite formation. Table 3 summarizes the cumulative amounts of the parent compound and the desacetyl metabolites excreted into bile. In several chromatograms, relatively small peaks were observed, which did not coincide with one of the prospective hydroxy-derivatives. Assuming an extraction and detection efficiency comparable to that of the other compounds, these metabolites amount to <1% of the dose.

For rocuronium, Org 9273 and Org 20297, which are 3-hydroxy compounds, for Org 9955, a 3-carbamate, and for Org 20059, a 3-desoxy compound, no metabolites or only trace amounts of metabolites, probably the 17-OH-derivatives, were found. All compounds with a 3-acetyl group were partly hydrolyzed to the 3-OH derivatives. For several compounds, the 17-OH derivative also was found. Measurable amounts of the 3,17-di-OH derivatives were found only for Org 9991, Org 9487 and Org 9489.

The efficiency of the biliary excretion process is expressed in the biliary clearance CL₂₀ (table 2), as well as in the cumulative amounts excreted into bile (table 3). There was no apparent correlation between lipophilicity and the biliary clearance (table 4) or with the amounts excreted. Pharmacokinetic analysis using a model with biliary excretion obeying Michaelis-Menten kinetics (equation 10) did not provide evidence for a saturable excretory process at the canalicular membrane in the dose range chosen in the present experiments.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Chemical structure and lipophilicity. The results described in this study confirm that small changes in chemical structure have a significant influence on the measured partition coefficient, corresponding to the expected change in lipophilicity. Although quaternary ammonium compounds are relatively hydrophilic compounds, the presence of large hydrophobic structures may result in a surprisingly large lipophilic character, as was shown earlier for hexafluorenium (Meijer and Weitering, 1970), methyldeptropine (Lavy et al., 1972), thiazinamium and N-methyl-imipraminium (Neef and Meijer, 1984). The partition coefficient of these compounds is comparable to that of the 17-butyrate compounds used in the present study (table 1).

At first sight, our study does not support the existence of a correlation between the molecular weight and lipophilicity, as shown by Neef and Meijer (1984) for a series of monoquaternary organic cations. In our study, the molecular weight of the compounds was much higher than those in the study by Neef and Meijer and in contrast to this earlier study, the differences in molecular weight between the compounds were only small; the molecular weight of the presently studied bulky organic cations varied between 515 and 603. In the study by Neef and Meijer (1984), two compounds deviated from the observed relationship between lipophilicity and molecular weight (i.e., two compouds were much less lipophilic as predicted from the other compounds). These two compounds, thiazinamium sulfoxide and acetylprocainamide ethobromide, differ from the related agents thiazinamium and procainamid ethobromide by the presence of other hydrophilic parts in the molecule. The same applies for the compounds investigated in the present study; each of the compounds contains one or two ester groups. Therefore, the lack of an apparent relationship between partition coefficient and molecular weight in the present series of compounds can be explained by similar factors as indicated by Neef and Meijer (1984).

The significant correlation between the retention time in the HPLC column and partition coefficient indicates that the lipophilicity, in particular, the hydrophilicity-lipophilicity balance of the compounds, is at least partly reflected in the retention time. Therefore, the retention time can be used as a rough indicator of lipophilicity within a series of related compounds. However, other factors, such as stereochemistry, may also play a role in the chromatographic behavior. This is reflected in the difference between the retention times of Org 9453 and Org 9616, which are two isometric forms with comparable lipophilicity. Also, the retention times in our system may be affected by the use of ion-pair chromatography with an eluent containing the anionic compound DAS. Under these conditions, retention times may reflect the behavior of the ion-pairs rather than the cationic analytes on the column.

Plasma protein binding. We found a significant correlation between the logarithm of the partition coefficient and the plasma protein binding of bulky organic cations (Proost et al., 1995; Wierda et al., 1993) A similar correlation was found earlier for the binding of some quaternary ammonium compounds in rat plasma (Neef and Meijer, 1984) and for alpha-1 acid glycoprotein (Van der Sluijs and Meijer, 1985). From our results, it can be concluded that binding to albumin may play a role as well. These results indicate that the binding of quaternary ammonium compounds to plasma proteins is at least partly due to hydrophobic, rather than electrostatic, interactions (Van der Sluijs and Meijer, 1985). The data clearly indicate that the lipophilicity parameter, obtained from an in vitro octanol/Krebs system, has a biological correlate in the extent of protein binding. The relationship between plasma protein binding and lipophilicity has important consequences for the pharmacokinetic properties and, as a result, for the time course of action and potency of NMBAs (Poost et al., 1996). In the present study, we studied the protein binding in human plasma to facilitate the extrapolation to humans (Proost et al., 1995; Wierda et al., 1993; Wierda and Proost, 1992, 1995a, 1995b, 1995c).

Our measured protein binding of pancuronium and vecuronium agrees reasonably well with the values reported by Duvaldestin and Henzel (1982). On the contrary, Foldes and Deery (1983) reported a protein binding value of 87.5% for pancuronium and 90.6% for vecuronium by using a technique based on the inhibition of plasma butyrylcholinesterase. This indirect approach depends on many assumptions, and the traditional ultrafiltration and dialysis methods seem to be more reliable.

Hepatic uptake in isolated perfused rat liver. In earlier studies, it has been shown that in contrast to pancuronium (Bencini et al., 1988; Mol et al., 1992) and pipecuronium⁶ vecuronium is rapidly taken up in the isolated perfused rat liver; the uptakes of Org 7268 (3-hydroxy-vecuronium) and Org 6368 were intermediate between those of vecuronium and pancuronium (Mol et al., 1992). Several of the compounds investigated in the present study were taken up by the liver even more rapidly than vecuronium, such as rocuronium and the more lipophilic compounds with a propionyl- or butyryl-group at the 17-position and an allyl-group at the quaternary nitrogen atom at position 16.

The intrinsic clearance, apparent clearance and extraction ratio appear to be correlated to the partition coefficient, as can be seen in figure 4 (numerical correlation data given in table 4). The reason for such a correlation is probably the effect of hydrophobic interactions on the binding affinity for the carrier-mediated uptake system for these cationic drugs, resulting in a higher carrier occupancy and higher uptake rate of the more lipophilic compounds. Although the binding to albumin is low for these compounds, protein binding may play a role as well. Because both protein binding and clearance increase with lipophilicity, the increase of unbound clearance with increasing lipophilicity is even more pronounced than the increase in CL_int.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Correlation between lipophilicity (log P) and initial extraction ratio in the isolated perfused rat liver.

In accordance with the parallel-tube model (equations 6 and 7), the extraction ratio is low for compounds with an intrinsic clearance far below perfusate flow (e.g., pancuronium), and apparent clearance is only slighly lower than intrinsic clearance. For compounds with a high intrinsic clearance (e.g., rocuronium), the extraction ratio is high, and the apparent clearance is limited by the perfusate flow. A remarkable position was found for rocuronium, with an initial extraction ratio of almost 100%, indicating a very efficient uptake by the liver. A more efficient hepatic uptake of rocuronium compared with vecuronium was also reported in isolated human hepatocytes (Sandker et al., 1994). The high intrinsic clearance of rocuronium does not seem to be related to its lipophilicity, which has an intermediate value. This finding demonstrates that other factors also contribute to the intrinsic clearance process. Regarding the molecular structures (figure 1), the rapid clearance of rocuronium may be related to the five-member pyrrolidine-ring at the 16-N atom, which might cause less steric hindrance in binding to the carrier than the six-member piperidine-ring found in most compounds. This may explain also the relatively low hepatic clearance of the relatively lipophilic compound Org 9991, which compound has a seven-member homopiperidine-ring at the 16-N atom. Steric hindrance may contribute also to the difference in hepatic clearance between the isomers Org 9453 and Org 9616. Such an effect of steric factors has been demonstrated by the different hepatobiliary transport characteristics of the enantiomers of the organic cations oxyphenonium (Feitsma et al., 1989), N-methyl-dextrorphan and N-methyl-levorphanol.⁷

Many organic cations are taken up in the liver by carrier-mediated transport (Meijer, 1976; Meijer et al., 1989, 1990, 1991; Schanker, 1965). Different carrier systems have been identified for small, monovalent catios (type I compounds) and for relatively large, bivalent cations (type II compounds), like nondepolarizing NMBAs (Meijer et al., 1970). The uptake of type II compounds is strongly affected by cardiac glycosides and bile acids (Meijer et al., 1971; Mol et al., 1992; Steen et al., 1992) and occurs via both a saturable and a nonsaturable process (Mol et al., 1988; Steen et al., 1992). The slightly convex profile of the perfusate concentration for several compounds, as shown in figure 3 for Org 9487, also suggest a saturable uptake process in the liver. The design of our study was not well suited for an accurate estimation of the Michaelis-Menten parameters V_max and K_m. Studies with higher and lower doses should be performed before definite conclusions about this phenomenon can be drawn. The K_m values estimated from our results are of the same order of magnitude as the neuromuscular blocking concentrations for low potent NMBAs like rocuronium and Org 9487 in humans. Therefore, a saturable liver uptake may be important in clinical practice because of an unexpected increase of duration of action at an increase of the dose.

The profile of the perfusate concentration on a logarithmic scale exhibits the characteristic biphasic pattern only for Org 7268, as was found earlier for Org 6368 (Mol et al., 1992), indicating that after primary liver uptake, a portion of the drug is returned to the sinusoidal space. This was confirmed by relatively high values of k₂₁ for these compounds in the pharmacokinetics analysis. The highest value of k₂₁ was found for rocuronium, suggesting some correlation between the transport in both directions. Bidirectional transport of organic cations was recently demonstrated in reconstituted OCT1 expressed in Xenopus laevis oocytes (Busch et al., 1996).

Metabolism. All compounds containing an acetyl group at the 3-position were found to be metabolized to a considerable extent by ester hydrolysis. In general, the ester bond at the 17-position was more stable to hydrolysis; only for the 17-butyrate esters were detectable amounts of the 17-OH derivatives found (table 3).

Surprisingly, the variability of the amounts of the 3-OH, 17-OH and 3,17,di-OH derivatives between the experiments is very large for several compounds (e.g., for Org 9487, Org 9453 and Org 9991). Formation of the metabolites after excretion into bile is unlikely, considering the short residence time in the biliary tree and cannula (2 to 3 min). The bile was collected directly in a buffer at pH ~4, a condition at which these compounds are chemically stable, as was confirmed for each compound. This finding indicates that the ester hydrolysis of this type of compounds is highly variable, even within a single rat strain.

In principle, metabolic pathways other than hydrolysis of the esters at the 3- and 17-position cannot be excluded. For example, drugs may be metabolized to unknown derivatives that are not detected in the applied assay technique (i.e., more hydrophilic compounds that do not form lipophilic ion-pairs with the counter-ion DAS). Among others, the hydroxyl groups might be conjugated to glucuronides or sulfates. However, for Org 7617 it was found that pretreatment of the bile samples of the isolated perfused liver experiment with -glucuronidase and sulfatase did not result in higher levels of the hydroxy-derivatives of Org 7617.⁸ In several chromatograms, relatively small peaks were observed, which did not coincide with one of the three prospective hydroxy-derivatives. Assuming an extraction and detection efficiency similar to that of the other compounds, their amounts are small and do not influence the present conclusions.

Hepatic storage. The pharmacokinetic analysis revealed that a considerable fraction of the dose is stored within the liver, as can be concluded from the relatively large amounts present in the storage compartment at the end of the study period. As a result, the drug seems to not be directly available for biliary excretion, as may be inferred from the observation that the biliary excretion rate after 120 min of perfusion is very low despite considerable liver content. The rate of transport from the intracellular storage compartment back to the cytosol is rather slow. In fact, in most cases this rate constant was too low to be accurately determined within the duration of the experiment.

The effective storage of these bulky organic cations in the liver confirms the results of several previous studies. Earlier experiments in intact rats and isolated perfused rat liver indicated that d-tubocurarine, pancuronium, Org 6368 and vecuronium are stored in subcellular particles, in particular, in mitochondria, lysosomes and cell nuclei (Mol and Meijer, 1990; Weitering et al., 1975, 1977). Several organic cations, including rocuronium, were found to be taken up in mitochondria by a potential-driven active transport mechanism (Steen, 1992; Steen et al., 1993). Organic cations like vecuronium can be sequestrated in lysosomes and endosomes; this accumulation was found to be related to proton gradients (Moseley and Van Dyke, 1995; Van Dyke et al., 1992). Taking into account the relatively large contribution of mitochondria in total cell volume, it is likely that the observed hepatic storage is primarily due to accumulation into mitochondria. This intracellular sequestration keeps the cytoplasmatic concentration of these agents low. Consequently, the chemical driving force for excretion back into plasma and into bile will be low, and the biliary excretion rate may be largely dependent on these storage processes. Therefore, a saturation of this transport process will not be easily demonstrated.

Biliary excretion. No significant correlation was found between the biliary clearance CL₂₀ and lipophilicity; the data even suggest that the biliary excretory function decreases with increasing lipophilicity (table 4). A possible explanation for the apparent low value for the biliary clearance, in particular, for the relatively lipophilic 3-acetyl,17-butyrates and 3-acetyl,17-propionates, may be an interaction with the metabolites competing for the same carrier system.

Recently, two potential candidates for the carrier-mediated hepatic uptake of bulky organic cations in the rat were cloned: OCT1 (Gründemann et al., 1994) and OATP (Oude Elferink et al., 1995). The latter carrier protein, surprisingly, also seems to accommodate amphipathic organic cations. Preliminary studies indicate that OCT1 transports the relatively hydrophilic organic cations, whereas OATP may take more hydrophobic substrates.⁹ The present series of structurally related organic cations with widely varying physicochemical properties may provide a unique series of model compounds to relate transport characteristics in the intact organ with that in reconstructed systems of cloned carrier proteins.

In conclusion, we demonstrated that lipophilicity of aminosteroidal NMBAs, which are relatively large, bivalent cations, is a predominant factor for their degree of protein binding. Lipophilicity is also a determinant of their hepatic uptake rate, but other factors may play a significant role as well. The biliary excretory process does not seem to be related to lipophilicity. The observed relationships among lipophilicity, protein binding and hepatic uptake may be helpful for a better understanding of the relationship between chemical structure and the time course of action of NMBAs. Also, this study may provide important model compounds for the functional characterization of cloned carrier proteins in the liver.

	Acknowledgments

The authors wish to thank Mr. R. Oosting for his skillful technical assistance with the isolated perfused rat liver experiments, Mr. J. Visser for his advice on the bioassays and Dr. J. E. Paanakker and Mrs. S. Tjepkema (Organon International, Oss, The Netherlands) for performing the bioassays of the perfusion experiments with rocuronium and Org 9616. Organon Teknika (Turnhout, Belgium) is gratefully acknowledged for financial support and for providing the NMBAs.

	Footnotes

Accepted for publication March 17, 1997.

Received for publication November 15, 1996.

¹   This work was supported by Organon Teknika (Turnhout, Belgium).

²   Present address: Groningen Institute for Drug Studies (GIDS), University Centre for Pharmacy, Department of Pharmacokinetics and Drug Delivery, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands.

³   Present address: Research Group for Experimental Anesthesiology and Clinical Pharmacology, University Hospital, Department of Anesthesiology, Hanzeplein 1, 9713 GZ Groningen, The Netherlands.

⁴   GIDS is part of the research school Groningen Utrecht Institute for Drug Exploration (GUIDE).

⁵   J. H. Proost, Groningen Institute for Drug Studies, unpublished observations.

⁷   Lanting, A. B. L., Ensing, K., Drenth, B. F. H., Meijer, D. K. F. and De Zeeuw, R. A.: Stereochemical factors in the hepatic transport and metabolism of the organic cations N-methyl dextrorphan and N-methyl levorphanol in the isolated perfused rat liver, submitted manuscript.

⁸   J. H. Proost, Groningen Institute for Drug Studies, unpublished observations.

⁹   H. Koepsell and P. J. Meijer, personal communication.

⁶   D. K. F. Meijer, Groningen Institute for Drug Studies, unpublished observations.

Send reprint requests to: Dr. J. H. Proost, University Centre for Pharmacy, Department of Pharmacokinetics and Drug Delivery, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands.

	Abbreviations

NMBA, neuromuscular blocking agent; P, partition coefficient; OCT1, organic cation transporter 1; OATP, organic anion transporting polypeptide; MRP, multidrug resistance-related protein; HPLC, high-performance liquid chromatography; DAS, 9,10-dimethoxyanthracene-2-sulfonate.

	References

Top Abstract Introduction Methods Results Discussion References

• Agoston, S., Houwertjes, M. C. and Salt, P. J.: A new method for studying the relationship between hepatic uptake of drugs and their pharmacodynamic effects in anaesthetized cats. Br. J. Pharmacol. 68: 637-643, 1980[Medline].
• Baird, W. L. M. and Viby-Mogensen, J.: Clinical considerations in the rational use of neuromuscular blocking drugs. In Muscle Relaxants, ed. by S. Agoston and W. C. Bowman, pp. 181-197, Elsevier Science Publishers, Amsterdam, 1990.
• Bass, L. and Keiding, S.: Physiologically based models and strategic experiments in hepatic pharmacology. Biochem. Pharmacol. 37: 1425-1431, 1988[Medline].
• Bencini, A. F., Houwertjes, M. C. and Agoston, S.: Effects of hepatic uptake of vecuronium bromide and its putative metabolites on their neuromuscular blocking actions in the cat. Br. J. Anaesth. 57: 789-795, 1985[Abstract/Free Full Text].
• Bencini, A. F., Mol, W. E. M., Scaf, A. H. J., ersten, U. W., Wolters, K. T. P., Agoston, S. and Meijer, D. K. F.: Uptake and excretion of vecuronium bromide and pancuronium bromide in the isolated perfused rat liver. Anesthesiology 69: 487-492, 1988[Medline].
• Bencini, A. F., Scaf, A. H. J., Sohn, Y. J., Kersten-Kleef, U. W. and Agoston, S.: Hepatobiliary disposition of vecuronium bromide in man. Br. J. Anaesth. 58: 988-995, 1986[Abstract/Free Full Text].
• Bossuyt, X., Müller, M., Hagenbuch, B. and Meier, P. J.: Polyspecific steroid and drug clearance by an organic anion transporter of mammalian liver. J. Pharmacol. Exp. Ther. 276: 891-896, 1996a[Abstract/Free Full Text].
• Boussuyt, K., Müller, M. and Meier, P. J.: Multispecific amphipathic substrate transport by an organic anion transporter of human liver. J. Hepatol. 25: 733-738, 1996b[Medline].
• Boxenbaum, H. G., Riegelman, S. and Elashoff, R. M.: Statistical estimation in pharmacokinetics. J. Pharmacokinet. Biopharm. 2: 123-148, 1974[Medline].
• Busch, A. E., Quester, S., Ulzheimer, J. C., Waldegger, S., Gorboulev, V., Arndt, P., Lang, F. and Koepsell, H.: Electrogenic properties and substrate specificity of the polyspecific rat cation transporter rOCT1. J. Biol. Chem. 271: 32599-32604, 1996[Abstract/Free Full Text].
• Colburn, W. A., Brazzell, R. K. and Bekersky, I.: Organ perfusion studies. J. Pharm. Sci. 72: 970, 1983[Medline].
• D'Argenio, D. Z. and Schumitzky, A.: A program package for simulation and parameter estimation in pharmacokinetic systems. Comp. Prog. Biomed. 9: 115-134, 1979.
• Draper, N. R. and Smith, H.: Applied Regression Analysis, pp. 44-85, John Wiley & Sons, New York, 1981.
• Duvaldestin, P. and Henzel, D.: Binding of tubocurarine, fazadinium, pancuronium and Org NC 45 to serum proteins in normal man and in patients with cirrhosis. Br. J. Anaesth. 54: 513-516, 1982[Abstract/Free Full Text].
• Feitsma, K. G., Drenth, B. F. H., De Zeeuw, R. A., Oosting, R. and Meijer, D. K. F.: Unequal disposition of enantiomers of the organic cation oxyphenonium in the rat isolated perfused liver. J. Pharm. Pharmacol. 41: 27-31, 1989[Medline].
• Foldes, F. F. and Deery, A.: Protein binding of atracurium and other short-acting neuromuscular blocking agents and their interaction with human cholinesterases. Br. J. Anaesth. 55: 31S-34S, 1983.
• Godfrey, K. R. and Chapman, M. J.: The problem of model indistinguishability in pharmacokinetics. J. Pharmacokinet. Biopharm. 17: 229-267, 1989[Medline].
• Godfrey, K. R., Chapman, M. J. and Vajda, S.: Identifiability and indistinguishability of nonlinear pharmacokinetic models. J. Pharmacokinet. Biopharm. 22: 229-251, 1994[Medline].
• Gründemann, D., Gorboulev, V., Gambaryan, S., Veyhl, M. and Koepsell, H.: Drug excretion mediated by a new prototype of polyspecific transporter. Nature 372: 549-552, 1994[Medline].
• Jacquez, J. A.: Identifiability: The first step in parameter estimation. FASEB Fed. Proc. 46: 2477-2480, 1987.
• Kartenbeck, J., Keuschner, U., Mayer, R. and Keppler, D.: Absense of the canalicular isoform of the MRP gene-encoded conjugate export pump from the hepatocytes in Dubin-Johnson syndrome. Hepatology 23: 1061-1066, 1996[Medline].
• Kleef, U. W., Proost, J. H., Roggeveld, J. and Wierda, J. M. K. H.: Determination of rocuronium and its putative metabolites in body fluids and tissue homogenates. J. Chromatogr. 621: 65-76, 1993[Medline].
• Lavy, U. I., Hespe, W. and Meijer, D. K. F.: Uptake and excretion of the quaternary ammonium compound deptropine methoiodide in the isolated perfused rat liver. Naunyn-Schmiedebergs Arch. Pharmacol. 275: 183-192, 1972[Medline].
• Meijer, D. K. F., Bos, E. S. and Van Der Laan, K. J.: Hepatic transport of mono- and bisquaternary ammonium compounds. Eur. J. Pharmacol. 11: 371-377, 1970[Medline].
• Meijer, D. K. F., Arends, J. W. and Weitering, J. G.: The cardiac glycoside sensitive step in the hepatic transport of the bisquaternary ammonium compound, hexafluorenium. Eur. J. Pharmacol. 15: 245-251, 1971[Medline].
• Meijer, D. K. F.: The mechanisms for hepatic uptake and biliary excretion of organic cations. In Intestinal Permeation, ed. by M. Kramer and F. Lauterbach, pp. 196-207, Excerptia Medica, Amsterdam, 1976.
• Meijer, D. K. F., Mol, W. E. M., Müller, M. and Kurz, G.: Mechanisms for hepato-biliary transport of cationic drugs studied with the intact organ and on the membrane level. In Hepatic Transport in Organic Substances, ed. by E. Petzinger, R. K. H. Kinne and H. Sies, pp. 344-367, Springer-Verlag, Berlin/Heidelberg, 1989.
• Meijer, D. K. F., Mol, W. E. M., Müller, M. and Kurz, G.: Carrier-mediated transport in the hepatic distribution and elimination of drugs, with special reference to the category of organic cations. J. Pharmacokinet. Biopharm. 18: 35-70, 1990[Medline].
• Meijer, D. K. F., Mol, W. E. M., Müller, M., Steen, H. and Kurz, G.: Carrier-mediated transport in the hepatic distribution and elimination of organic cations. In Hepatic metabolism and disposition of endo- and xenobiotics, ed. by K. W. Bock, S. Matern, W. Gerok and R. Schmid, pp. 259-270, Kluwer Academic Publishers, Dordrecht/Boston/London, 1991.
• Meijer, D. K. F. and Weitering, J. G.: Curare-like agents: relation between lipid solubility and transport into bile in perfused liver. Eur. J. Pharmacol. 10: 283-289, 1970[Medline].
• Mol, W. E. M., Fokkema, G. N., Weert, B. and Meijer, D. K. F.: Mechanisms for the hepatic uptake of organic cations: Studies with the muscle relaxant vecuronium in isolated rat hepatocytes. J. Pharmacol. Exp. Ther. 244: 268-275, 1988[Abstract/Free Full Text].
• Mol, W. E. M. and Meijer, D. K. F.: Hepatic transport mechanisms for bivalent organic cations: Subcellular distribution and hepato-biliary concentration gradients of some steroidal muscle relaxants. Biochem. Pharmacol. 39: 383-390, 1990[Medline].
• Mol, W. E. M., Rombout, F., Paanakker, J. E., Oosting, R., Scaf, A. H. J. and Meijer, D. K. F.: Pharmacokinetics of steroidal muscle relaxants in isolated perfused rat liver. Biochem. Pharmacol. 44: 1453-1459, 1992[Medline].
• Moseley, R. H. and Van Dyke, R. W.: Organic cation transport by rat liver lysosomes. Am. J. Physiol. 268(Gastrointest. Liver Physiol. 31): G480-G486, 1995[Abstract/Free Full Text].
• Neef, C. and Meijer, D. K. F.: Structure-pharmacokinetics relationship of quaternary ammonium compounds: Correlation of physicochemical and pharmacokinetic parameters. Naunyn-Schmiedebergs Arch. Pharmacol. 328: 111-118, 1984[Medline].
• Nelder, J. A. and Mead, R.: A simplex method for function minimization. Comput. J. 7: 308-313, 1965.
• Noy, N. and Zakim, D.: Physical-chemical basis for the uptake of organic compounds by cells. In Hepatic Transport and Bile Secretion: Physiology and Pathophysiology, ed. by N. Tavoloni and P. D. Berk, pp. 313-335, Raven Press, New York, 1993.
• Oude Elferink, R. P. J., Meijer, D. K. F., Kuipers, F., Jansen, P. L. M., Groen, A. K. and Groothuis, G. M. M.: Hepatobiliary secretion of organic compounds; molecular mechanisms of membrane transport. Biochim. Biophys. Acta 1241: 215-268, 1995[Medline].
• Paanakker, J. E., Thio, J. S. L. M., Van de Wildenberg, H. M. and Kaspersen, F. M.: Assay of vecuronium in plasma using solid-phase extraction, HPLC and post-column ion-pair extraction with fluorimetric detection. J. Chromatogr. 421: 327-335, 1987[Medline].
• Paulusma, C. C., Bosma, P. J., Zaman, G. J. R., Bakker, C. T. M., Otter, M., Scheffer, G. L., Scheper, R. J., Borst, P. and Oude Elferink, R. P. J.: Congenital jaundice in rats with a mutation in a multidrug resistance-associated protein gene. Science 271: 1126-1128, 1996[Abstract].
• Peck, C. C., Barr, W. H., Benet, L. Z., Collins, J., Desjardins, R. E., Furst, D. E., Harter, J. G., Levy, G., Ludden, T., Rodman, J. H., Santhanan, L., Schentag, J. J., Shah, V. P., Sheiner, L. B., Skelly, J. P., Stansky, D. R., Temple, R. J., Viswanathan, C. T., Weissinger, J. and Yacobi, A.: Opportunities for integration of pharmacokinetics, pharmacodynamics, and toxicokinetics in rational drug development. Clin. Pharmacol. Ther. 51: 465-473, 1992[Medline].
• Press, W. H., Flannery, B. P., Teukolsky, S. A. and Vetterling, W. T.: Numerical Recipes, Cambridge University Press, Cambridge, 1986.
• Proost, J. H., Nijssen, H. M. J., Strating, C. B., Meijer, D. K. F. and Groothuis, G. M. M.: Pharmacokinetic modeling of the sinusoidal efflux of anionic ligands from the isolated perfused rat liver: The influence of albumin. J. Pharmacokinet. Biopharm. 21: 375-394, 1993[Medline].
• Proost, J. H., Wierda, J. M. K. H. and Meijer, D. K. F.: Structure-pharmacokinetic and pharmacodynamic relationships of steroidal neuromuscular blocking agents. In Muscle Relaxants: Physiologic and Pharmacologic Aspects, ed. by K. Fukushima and R. Ochiai, pp. 205-212, Springer-Verlag, Tokyo, 1995.
• Proost, J. H., Wierda, J. M. K. H. and Meijer, D. K. F.: An extended pharmacokinetic/pharmacodynamic (PK/PD) model describing quantitatively the influence of plasma protein binding, tissue binding, and receptor binding on the potency and time course of action of drugs. J. Pharmacokinet. Biopharm. 24: 45-77, 1996[Medline].
• Rekker, R. F.: Hydrophobic aspects of binding. In Biological Activity and Chemical Structure, ed. by J. A. Keverling Buisman, pp. 107-130, Elsevier Scientific Publishing Company, Amsterdam, 1977.
• Sandker, G. W., Weert, B., Olinga, P., Wolters, H., Slooff, M. J. H., Meijer, D. K. F. and Groothuis, G. M. M.: Characterization of transport in isolated human hepatocytes: A study with the bile acid tauocholic acid, the uncharged ouabain and the organic cations vecuronium and rocuronium. Biochem. Pharmacol. 47: 2193-2200, 1994[Medline].
• Savarese, J. J., Lien, C. A., Belmont, M. R., Maehr, R. B. and Wastila, W. B.: Future goals of the benzylisoquinolinium ester program. In Muscle Relaxants: Physiological and Pharmacological Aspects, ed. by K. Fukushima and R. Ochiai, pp. 335-341, Springer-Verlag, Tokyo, 1995.
• Schanker, L. S.: Hepatic transport of organic cations. In The Biliary System, ed. by W. Taylor, pp. 469-480, Blackwell, Oxford, 1965.
• Seydel, J. K. and Schaper, K. J.: Quantitative structure-pharmacokinetic relationships and drug design. Pharmacol. Ther. 15: 131-182, 1982.
• Sheiner, L. B. and Beal, S. L.: A note on confidence intervals with extended least squares parameter estimates. J. Pharmacokinet. Biopharm. 15: 93-98, 1987[Medline].
• Steen, H.: Mechanisms for the hepatic disposition of cationic drugs: Molecular aspects of membrane transport in the hepatocyte, pp. 125-132. Thesis, University of Groningen, 1992.
• Steen, H., Maring, J. G. and Meijer, D. K. F.: Differential effects of metabolic inhibitors on cellular and mitochondrial uptake of organic cations in rat liver. Biochem. Pharmacol. 45: 809-818, 1993[Medline].
• Steen, H., Merema, M. and Meijer, D. K. F.: A multispecific uptake system for taurocholate, cardiac glycosides and cationic drugs in the liver. Biochem. Pharmacol. 44: 2323-2331, 1992[Medline].
• Tomlinson, E.: Chromatographic hydrophobic parameters in correlation analysis of structure-activity relationships. J. Chromatogr. 113: 1-45, 1975[Medline].
• Van der Sluijs, P. and Meijer, D. K. F.: Binding of drugs with a quaternary ammonium group to alpha-1 acid glycoprotein and asialo alpha-1 acid glycoprotein. J. Pharmacol. Exp. Ther. 234: 703-707, 1985[Abstract/Free Full Text].
• Van Dyke, R. W., Faber, E. D. and Meijer, D. K. F.: Sequestration of organic cations by acidified hepatic endocytic vesicles and implications for biliary excretion. J. Pharmacol. Exp. Ther. 261: 1-11, 1992[Abstract/Free Full Text].
• Veng Pedersen, P.: Curve fitting and modeling in pharmacokinetics and some practical experiences with NONLIN and a new program FUNFIT. J. Pharmacokinet. Biopharm. 5: 513-531, 1977[Medline].
• Weibel, E. R., Stäubli, W., Gnägi, H. R. and Hess, F. A.: Correlated morphometric and biochemical studies on the liver cell. I. Morphometric model, stereologic methods and normal morphometric data for rat liver. J. Cell Biol. 42: 68-91, 1969[Abstract/Free Full Text].
• Weitering, J. G., Mulder, G. J., Meijer, D. K. F., Lammers, W., Veenhuis, M. and Wendelaar Bonga, S. E.: On the localization of d-tubocurarine in rat liver lysosomes in vivo by electron microscopy and subcellular fractionation. Naunyn-Schmiedebergs Arch. Pharmacol. 289: 251-256, 1975[Medline].
• Weitering, J. G., Lammers, W., Meijer, D. K. F. and Mulder, G. J.: Localization of d-tubocurarine in rat liver lysosomes: Lysosomal uptake, biliary excretion and displacement by quinacrine in vivo. Naunyn-Schmiedebergs Arch. Pharmacol. 299: 277-281, 1977[Medline].
• Wierda, J. M. K. H., Proost, J. H., Muir, A. W. and Marshall, R. J.: Design of drugs for rapid onset. Anaesth, Pharmacol. Rev. 1: 57-68, 1993.
• Wierda, J. M. K. H. and Proost, J. H.: (Pre-)clinical development of new aminosteroidal nondepolarizing neuromuscular blocking agents. In The Development of Aminosteroidal Neuromuscular Blocking Agents, ed. by P. Denissen, pp. 43-82, Interface, Turnhout, Belgium, 1992.
• Wierda, J. M. K. H. and Proost, J. H.: Structure-onset relations of steroidal neuromuscular blocking agents. In Muscle Relaxants: Physiologic and Pharmacologic Aspects, ed. K. Fukushima and R. Ochiai, pp. 163-166, Springer-Verlag, Tokyo, 1995a.
• Wierda, J. M. K. H. and Proost, J. H.: Structure-pharmacodynamic-pharmacokinetic relationships of steroidal neuromuscular blocking agens. Eur. J. Anaesthesiol. 12 (Suppl. 11): 45-54, 1995b.
• Wierda, J. M. K. H. and Proost, J. H.: Molecular design of steroidal muscle relaxants. In Proceedings of the 24th Central European Congress on Anesthesiology, Volume of the Refresher, Main Lectures, Symposia, ed. by W. K. Ilias, A. Hammerle, W. List, W. Mauritz, N. Mutz, V. Schalk and M. Zimpfer, pp. 543-548, Monduzzi Editore, Bologna, 1995c.
• Wilkinson, G. R.: Clearance approaches in pharmacology. Pharmacol. Rev. 39: 1-47, 1987[Medline].
• Winkler, K., Keiding, S. and Tygstrup, N.: Clearance as a quantitative measure of liver function. In The liver: Quantitative Aspects of Structure and Function, ed. by G. Paumgartner and R. Preisig, pp. 144-155, Karger, Basel, 1973.