QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: jpet.103.059030v1 308/3/1012    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Zuideveld, K. P. Articles by Danhof, M. Search for Related Content PubMed PubMed Citation Articles by Zuideveld, K. P. Articles by Danhof, M.

NEUROPHARMACOLOGY

Mechanism-Based Pharmacokinetic-Pharmacodynamic Modeling of 5-HT_1A Receptor Agonists: Estimation of in Vivo Affinity and Intrinsic Efficacy on Body Temperature in Rats

Leiden/Amsterdam Center for Drug Research, Division of Pharmacology, Gorlaeus Laboratory, Leiden, The Netherlands (K.P.Z., M.D.); Pfizer Global Research and Development, Discovery Biology, Sandwich, Kent, United Kingdom (P.H.V., D.N., R.T., N.P.); F. Hoffmann-La Roche, Modeling and Simulation and Biometrics Groups, Basel, Switzerland (K.P.Z., P.J.); and Mathematical Institute, Leiden University, Leiden, The Netherlands (L.A.P.)

Received for publication August 25, 2003
Accepted November 25, 2003.

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
The pharmacokinetic-pharmacodynamic (PK-PD) correlations of seven prototypical 5-HT_1A agonists were analyzed on the basis of a recently proposed semi-mechanistic PK-PD model for the effect on body temperature. The resulting concentration-effect relationships were subsequently analyzed on the basis of the operational model of agonism to estimate the operational affinity (pK_A) and efficacy (log ) at the 5-HT_1A receptor in vivo. The values obtained in this manner were compared with estimates of the affinity (pK_i) and intrinsic efficacy (log[agonist ratio]) in a receptor-binding assay. Between 5-HT_1A agonists wide differences in in vivo affinity and efficacy were observed, with values of the pK_A ranging from 5.67 for flesinoxan to 8.63 for WAY-100,635 [N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-2-pyridinyl-cyclohexanecarboxamide] and of the log ranging from –1.27 for WAY-100,135 [N-(1,1-dimethylethyl)-4-(2-methoxyphenyl)--phenyl-1-piperazine-propanamide] to 0.62 for R-(+)-8-hydroxy-2-[di-n-propylamino)tetralin. Poor correlations were observed between the in vivo receptor affinity (pK_A) and the affinity estimates in the in vitro receptor binding assay (pK_i; r² = 0.55, P > 0.05), which could in part be explained by differences in blood-brain distribution. In contrast, a highly significant correlation was observed between the efficacy parameters in vivo (log ) and in vitro (log [agonist ratio]; r² = 0.76, P < 0.05). Thus by combining the previously proposed semi-mechanistic PK-PD model for the effect on body temperature with the operational model of agonism, a full mechanistic PK-PD model for 5-HT_1A receptor agonists has been obtained, which is highly predictive of the in vivo intrinsic efficacy.

The previously proposed model is semi-mechanistic in the sense that it uses a mechanistic model for the characterization of the transduction process. Specifically, 5-HT_1A agonists cause lowering of body temperature by modulation of the set-point for maintenance of the body temperature in a direct concentration-dependent manner (Zuideveld et al., 2001). Therefore a mechanism-based set-point model, which is based on dynamical systems analysis, was proposed and successfully implemented in the integrated PK-PD model to describe the complex time profile of the effects of 5-HT_1A agonists on body temperature (Zuideveld et al., 2001, 2002b). However, the existing model still contains the Hill equation to characterize the interaction of the 5-HT_1A receptor agonists at the receptor level. The Hill equation is an empirical equation describing concentration-effect relationships, since its parameters (E_max, EC₅₀, and the Hill factor) are dependent on both drug-specific properties (receptor affinity and receptor intrinsic efficacy) and system-related properties (e.g., receptor density). This complicates the use of this equation for extrapolation and prediction (i.e., from in vitro test systems to the in vivo situation, for interspecies extrapolation and for the prediction of intra- and inter-individual variability in drug response) (Van der Graaf and Danhof, 1997a). Therefore, in previous investigations it has been proposed to incorporate principles from receptor theory for characterization of the drug concentration-effect relationship in mechanism-based PK-PD modeling (Van der Graaf and Danhof 1997a; Cox et al., 1998; Tuk et al., 1999, 2002; Garrido et al., 2000; Visser et al., 2002, 2003). In these investigations it has been shown that in particular the operational model of agonism (Black and Leff, 1983) is a useful expression for the characterization of the drug-receptor interaction, allowing a separation between drug-related and system-related parameters. It has been demonstrated that this model is very useful for the prediction of in vivo concentration-effect relationships on the basis of results from in vitro bioassays (Black and Leff, 1983; Van der Graaf and Danhof, 1997a), the prediction of tissue selectivity of drug effects (Van der Graaf et al., 1997b, 1999) and the understanding of inter-individual variability in drug effects (Van der Graaf and Danhof, 1997a).

The objective of the present investigation was to link the previously developed semi-mechanistic PK-PD model for 5-HT_1A agonists with the operational model of agonism into a full mechanism-based PK-PD model. To this end the concentration-effect relationships of seven pro-typical 5-HT_1A receptor agonists were simultaneously analyzed. This allowed estimation of the unique system related parameters such as the tissue maximum (E_max) and slope factor of the transduction function (n). In addition for each compound the specific drug-related properties receptor affinity (K_A) and intrinsic efficacy () were estimated. These drug-related properties were compared with estimates of the receptor affinity and intrinsic efficacy as determined in an in vitro binding assay.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
In Vivo Pharmacological Experiments. The details of the pharmacokinetic-pharmacodynamic experiments have been described previously (Zuideveld et al., 2001, 2002a, 2002b, 2002c). Briefly, 8 days prior to the experiment, the rats were operated upon. Indwelling cannulae for drug administration and blood sampling were implanted into the right jugular vein and the left femoral artery, respectively. Furthermore, a telemetric transmitter (Physiotel implant TA10TA-F40 system; Data Sciences International (DSI), St. Paul, MN) was implanted into the abdominal cavity for the measurement of core body temperature. In the PK-PD experiments, conscious freely moving rats received an i.v. infusion of vehicle (saline) or active drug. R-8-OH-DPAT was administered in a wide range of different doses: 1 mg/kg in 5 min (n = 6), 3 mg/kg in 5 min (n = 7), in 15 min (n = 5) and in 30 min (n = 6) and by computer-controlled infusions, with the concentration targeted at 160 ng/ml in blood for 2 h (n = 6). S-8-OH-DPAT was administered in a 5 mg/kg in 15 min (n = 6) and a 15 mg/kg in 15 min (n = 6) infusion. Flesinoxan was administered in 3 mg/kg in 5 min (n = 6), 10 mg/kg in 5 and 15 min (both n = 6) infusions. Buspirone was administered in a 5 mg/kg in 15 min (n = 6) and a 15 mg/kg in 15 min (n = 7) infusion. 1-PP and WAY-100,135 were administered in a 10 mg/kg in 15 min infusion (both n = 6). WAY-100,635 (5-HT_1A antagonist) was administered in a 3 mg/kg in 15 min infusion and in computer-controlled infusions with the blood concentration targeted at 170, 85, and 20 ng/ml in blood, respectively (4 x n = 6) during which a 1 mg/kg in 15 min infusion of R-8-OH-DPAT was administered. In each experiment in each individual rat, approximately 15 to 18 serial blood samples of 50 µl were taken according to a fixed time schedule to determine the time course of the drug concentration. After the experiment samples were stored at –20°C pending HPLC analysis based on methods described previously (Zuideveld et al., 2000, 2002a, 2002b, 2002c) or by MS (e.g., WAY-100,135). Body temperature was measured continuously throughout the experiment using the telemetric system. Protein binding was determined ex vivo. Blood was collected and incubated with various compounds at 34° and 38°C. Concentrations of 50 and 1000 ng/ml for R-8-OH-DPAT and flesinoxan, 250 and 2500 ng/ml for buspirone and 1-PP, and 750 and 5000 ng/ml for S-8-OH-DPAT, buspirone, WAY-100,635, and WAY-100,135 were evaluated. Blood was centrifuged and from the remaining plasma, the free fraction was isolated using ultra filtration (Centrifree; Millipore Corporation, Bedford, MA).

Receptor Binding Assay. The interaction at the 5-HT_1A receptor was determined in vitro in recombinant HA 6 HeLa 5-HT_1A cells in which the human receptor is expressed (Pauwels et al., 1993). It has been demonstrated that the rat 5-HT_1A receptor is 89% similar to the human receptor (Albert et al., 1990). Recombinant HeLa 5-HT_1A cells were grown in adherent culture in roller bottle flasks in Dulbecco's modified Eagle's medium containing fetal bovine serum (50 ml/500 ml), L-glutamine (5 ml/500 ml), antibiotic-antimycotic (5 ml/500 ml), and the antibiotic Geneticin (5 ml/500 ml; BioVectra, Prince Edward Island, Canada).

Receptor binding was determined on the basis of a scintillation proximity assay (SPA). Briefly this assay was conducted as follows. On reaching confluence, the cells were harvested, cell pellets were prepared by centrifugation and stored at –80°C until required. For the actual analysis, cell pellets were retrieved from storage, thawed on ice, and resuspended in membrane preparation buffer (50 mM Tris-HCl, pH 7.5 (4°C), 4 mM CaCl₂, + 1 protease inhibitor tablet (per 50 ml; Roche Diagnostics, Indianapolis, IN) at approximately 10 ml of buffer per milliliter of pellet. The suspension was homogenized with a mechanical homogenizer using 15 full strokes on ice before centrifuging at 1000g for 20 min at 4°C. The supernatant was retained, and the pellet was homogenized and centrifuged at 1000g for 20 min at 4°C. The two supernatants were combined, incubated at 37°C for 15 min and centrifuged at 48,000g for 20 min at 4°C. The pellets were resuspended in a small volume of membrane preparation buffer using the homogenizer as before. The volumes were adjusted to allow storage at –80°C at a protein concentration of around 0.5 mg/ml, as determined on the basis of a protein assay using the Microprotein kit (Sigma Chemical, Dorset, UK). Membrane homogenates were thawed on ice, diluted in incubation buffer if required, and homogenized at low speed using a Powergen 125 homogenizer (Fisher Scientific, Loughborough, UK). Pll-Ysi SPA beads ([³H]8-OH-DPAT Assay) or WGA-Ysi SPA beads ([³H]WAY-100,635 assay), both from Amersham Biosciences UK, Ltd (Little Chalfont, Buckinghamshire, UK) were resuspended at 50 mg/ml in incubation buffer. Beads were precoupled with membranes by incubating 3 µg of protein/mg of bead on a tilting tube roller for 2 h at 4°C. The coupled beads/membranes were then centrifuged at 120g in a Heraeus Multifuge 3 benchtop centrifuge (Kendro, Bishop's Stortford, Hertfordshire, UK) for 2 min at 4°C. The supernatant was discarded, and the pellet was washed in incubation buffer and spun as before. The conjugated beads were then resuspended in incubation buffer at 15 mg of bead/ml. Radioligands [³H]8-OH-DPAT and [³H]WAY-100,635 were diluted in incubation buffer (50 mM Tris-HCl, 4 mM CaCl₂, 10 µM pargyline, 1 g/l ascorbic acid, 0.01% Tween 40, pH 7.5 at 25°C) to give a concentration of 3 nM and 1.5 nM, respectively (1 nM and 0.5 nM final assay concentration, respectively). For the determination of [³H]8-OH-DPAT binding, nonspecific binding (NSB) was defined using 10 µM WAY-100,635, while for [³H]WAY-100,635 binding NSB was defined using 10 µM R-8-OH-DPAT. Compounds were dissolved in 100% DMSO and diluted in assay buffer containing 3% DMSO using a Tecan Genesis liquid handling robot (Tecan, Maennedorf, Switzerland) to a top concentration of 30 µM in 3% DMSO (10 µM in 1% DMSO in well). Serial half-log dilutions of these stock solutions were made with assay buffer containing 3% DMSO. Samples (20 µl) were plated out in duplicate into 384 well optiplates to give 11-point concentration-effect curves. Twenty microliters of the total and NSB stocks were added to the plate, followed by the addition of 20 µl of the bead/membrane preparation to all wells using a multidrop. The bead/membrane preparation was kept in suspension using a stirring flask. Finally, 20 µl of the [³H]8-OH-DPAT or [³H]WAY-100,635 was added to each well of the Optiplate using the multidrop. The plates were incubated on a plate shaker at room temperature for a total incubation time of 2 h for [³H]8-OH-DPAT or 6 h for [³H]WAY-100,635, before being left to settle for 30 min before counting using the Topcount NXT for 45 s/well (PerkinElmer Life and Analytical Sciences, Boston, MA). All procedures were carried out at 4°C unless otherwise stated.

Compounds. R-8-OH-DPAT, S-8-OH-DPAT, and WAY-100,635 were purchased from Sigma/RBI (Natick, MA). Buspirone and 1-PP were generously donated by Bristol-Myers Squibb Co. (Princeton, NJ). Solvay Pharmaceuticals (Weesp, The Netherlands) generously donated flesinoxan, and WAY-100,135 was synthesized by Pfizer (Sandwich, Kent, UK). Ascorbic acid, cell dissociation solution, CaCl₂, Hepes, IBMX, KCl, L-glutamine, NaCl, pargyline, polyethylenimine, and TRIS (hydroxymethyl) methylamine were purchased from Sigma Chemical. Glucose, KH₂PO₄, and MgSO₄ were purchased from VWR International Ltd. (Dorset, UK). Forskolin was purchased from Calbiochem (San Diego, CA). Microscint "0" and GF/B unifilters were purchased from PerkinElmer Life and Analytical Sciences. Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS) antibiotic-antimycotic, and Geneticin were purchased from Life Technologies, Inc. (Paisley, UK). [³H]8-OH-DPAT and [³H]WAY-100,635 were obtained from Amersham Biosciences UK, Ltd.

Data Analysis. A population approach was used to quantify both the pharmacokinetics and pharmacodynamics of the 5-HT_1A receptor agonists. Modeling of the in vivo pharmacokinetic and pharmacodynamic data was performed using the nonlinear mixed effects modeling software NONMEM developed by Sheiner and Beal (version V 1.1, NONMEM Project Group, University of California, San Francisco, CA) (Boeckman et al., 1992). Individual predictions were obtained in a Bayesian post hoc step. The concentration-time profiles of the 5-HT_1A receptor ligand were described using 2- and 3-compartment pharmacokinetic models like those implemented in NONMEM's ADVAN3 and ADVAN11, respectively (Zuideveld et al., 2001, 2002a, 2002b, 2002c). The pharmacokinetic parameter estimates were used to calculate individual agonist blood concentrations at the times of the temperature measurements.

The pharmacokinetic data were used to quantify the relationship between the time profile of the agonist blood concentration and the time course of the hypothermic effect. For this purpose, the data on the time course of the hypothermic effect for each individual rat were fitted to the semi-mechanistic PK-PD model, which we have recently proposed (Zuideveld et al., 2001, 2002a). In this model, the hypothermic effect by the 5-HT_1A receptor agonists is considered the result of the attenuation of a set-point control by the drug receptor interaction. The model further uses the concept of an indirect physiological response model (Dayneka et al., 1993), and takes into account a 0th order rate constant associated with the warming of the body (k_in) and a first order rate constant associated with the cooling of the body (k_out). The thermostat-like regulation is implemented as a continuous process in which body temperature (T) is compared with a fixed reference or set point temperature (T_SP). 5-HT_1A agonists elicit hypothermia by decreasing the set point value, whereby the extent of the decrease is a function of the drug concentration C (see below). This yields the following system of equations.

(1)

in which X denotes the thermostat signal, which is driven by the difference between the body temperature T and the set-point temperature T₀ on a time scale governed by a. Where T_SP = T₀[1 – f(C)]. Hence, when the set-point value is lowered, the body temperature is perceived as too high and X is lowered. The thermostat signal affects the cooling of the body through an effector function X^–, which multiplies the first order rate constant k_out. Thus, a drop in set point temperature leads to an increase in thermostat signal and a lowering of the "effective rate constant," which governs the cooling of the body. With four system parameters to be estimated, the degree of parameterization in eq. 1 is high, and this may lead to parameter identifiability problems. It can be shown that that one parameter can be eliminated in a procedure involving the introduction of dimensionless variables (Zuideveld et al., 2001). The procedure results in the establishment of the parameters A and B defined by,

(2)

where T₀ and X₀ are the values for T and X when no drug is present. Thus, four physiological parameters are reduced to three, and the parameters become identifiable. Note that A and B represent the relative growth rate of, respectively, X and T when C = 0 and T = T₀. The maximal response, as defined by S_max, equals 1 for a full agonist, such as R-8-OH-DPAT and 0 for an antagonist. As a result of the introduction of dimensionless quantities, the dependent variable T is rescaled with respect to T₀, the average temperature during the hour prior to drug administration and the observed average minimal temperature of the individuals receiving the highest dose of R-8-OH-DPAT as described previously (Zuideveld et al., 2001).

In the original model, the effect of the 5-HT_1A receptor agonist at the receptor was described on the basis of the empirical Hill equation. Thereby the drug-receptor interaction is considered to generate a stimulus S, which in turn lowers the body's set point. The relationship between the concentration and the fractional lowering of the set point was formulated as a sigmoidal E_max model according to,

(3)

where S is the stimulus, i.e., the activity; is the maximum stimulus the drug can produce; C is the drug concentration or the potency, SC₅₀ is the concentration required to produce 50% of the maximum stimulus, and n_H is the slope factor, which determines the steepness of the curve (also known as the Hill factor). To obtain a full mechanistic model the sigmoidal E_max model was replaced by the operational model of agonism (Black and Leff, 1983);

(4)

where E_max is the maximum effect achievable in the system, K_A is the agonist dissociation equilibrium constant, n is the slope index for the occupancy-effect relation, and is the efficacy parameter, which is defined by the ratio of total receptor concentration and the concentration of agonist-receptor complex required to produce half-maximal effect. As f₁(C) and f₂(C) E_max · ⁿ/(ⁿ + 1) when C , the sigmoidal E_max equation parameters can be expressed in terms of the operational model of agonism as follows (Black and Leff, 1983);

(5)

and for f₂(C) = :

(6)

Inspection of eq. 6 shows that SC₅₀/K_A 1/(2^1/n – 1) when 0. Furthermore, with high-efficacy values, eq. 6 approximates to a simple linear relationship, SC₅₀/K_A = 1/ regardless of the value of n (Van der Graaf et al., 1999).

Leff et al. (1990) have shown that the operational model can be used to obtain estimates of affinity and efficacy of a partial agonist by comparison to full agonists. This comparative method (originally proposed by Barlow, 1967) is based on the idea that per definition, the intrinsic activity of a full agonist is identical to the maximum system response. Therefore E_max is constrained to the estimate of the sigmoidal E_max equation, , for a full agonists, and K_A and for partial agonists can be estimated by directly fitting the operational model of agonism to the concentration-effect data. Van der Graaf and Danhof (1997b) have shown that ignoring inter-individual variation in E_max may result in erroneous estimates of affinity and efficacy. Therefore, the pharmacodynamic models were fitted to the data using nonlinear mixed effects modeling with the NONMEM software package (Boeckman et al., 1992). The model was implemented in NONMEM using ADVAN6. Inter-individual variability on the parameters was modeled to an exponential equation,

(7)

where is the population value for parameter P, P_i is the individual value and _i is the random deviation of P_i from P. The values of _i are assumed to be independently normally distributed with mean zero and variance ². Interindividual variability of K_A was assumed to be insignificant because receptor affinity is generally considered to be constant across animals of the same strain. K_A and were estimated as pK_A (–log K_A) and log , respectively, because these parameters are assumed to be log normally distributed (Leff et al., 1990; Van der Graaf et al., 1997c). Residual error was characterized by a proportional error model

(8)

where yp_ij is the jth prediction for the ith individual predicted by the model, ym_ij is the measurement, and accounts for the residual deviance of the model predicted value from the observed value. The values for the population , ², and ², were estimated using the centering first-order conditional estimation method with the first-order model in NONMEM. A conditional estimation method was used due to the high degree of nonlinearity of the model and the high density of the data. The centering option gives the average estimate of each element of together with a P value which can be used to assess whether this value is sufficiently close to zero. The occurrence of an average that is significantly different from zero indicates an uncentered or a biased fit. This method was not chosen because the average estimates of each element of were expected to be different from zero, but rather to greatly decrease computing time as required with just the conditional estimation method (Lindstrom and Bates, 1990; Boeckman et al., 1992). To further decrease computing time, only 1/16th of the temperature data set was used for modeling, reducing the number of temperature measurements from over 900 measurements per individual to approximately 60. The implication of this reduction is that there is a data point every 8 min, as opposed to every 0.5 min. This reduction did not void the integrity of the data profiles, as judged by cross validation of fitting the split data files.

Goodness-of-fit was analyzed using the objective function and various diagnostic methods. Model selection was based on the Akaike Information Criterion (Akaike, 1974) and assessment of parameter estimates and correlations. Goodness-of-fit between the different models (Hill equation versus the operational model of agonism) was not formally compared since they are structurally different.

Data from the in vitro binding assays were analyzed using the Cheng-Prusoff equation (Cheng and Prusoff, 1973) to obtain estimates of the pK_i. Correlations between the apparent in vivo pK_A estimates with pK_i values from the in vitro 5-HT_1A receptor binding assays (in the presence of either of [³H]WAY-100,635 or [³H]8-OH-DPAT) and the in vitro parameter for efficacy in the receptor binding assay (log[agonist ratio]) to the in vivo measure for efficacy (log ) were calculated on the basis of mean values using an error-invariables (e.i.v.) approach (Casella and Berger, 1990). The e.i.v. approach differs from a "regular" linear regression since it allows error in both the x and y direction to be taken into account. The e.i.v. model assumes that the expected y values (E[y_i]) depend linearly on the expected or "true" x values, i.e., E[y] = a + b · E[x], which cannot be observed directly. The model used is the so-called functional relationship, where

(9)

where N(µ, ²) denotes the normal distribution with expectation µ and variance ². For indentifiability reasons, the usual assumption that , where is fixed and known, was made (Casella and Berger, 1990). A reasonable estimate for was obtained using;

(10)

where _x is the mean standard error of all the x values and _y of all the y values. The Pearson correlation coefficient between the expected and the observed y values was used as a measure of explained variability. Confidence limits for the estimated slope b and intercept a were obtained by bootstrap methods (3000 replicates). The value 1 for the slope b and 0 for the intercept a not contained in the corresponding confidence interval means that the values are significantly different from 1 and 0, respectively (at the given level of the confidence interval). The e.i.v. analysis was written in S-PLUS (v.6.1; Insightful Corporation, Seattle, WA), and a copy of the script can be obtained from the authors.

	Results

Top Abstract Materials and Methods Results Discussion References

 
In Vivo Concentration-Effect Relationships. Figure 1, A and B, shows the average hypothermic effects versus time profiles of R- and S-8-OH-DPAT, WAY-100,635 (without and with R-8-OH-DPAT), WAY-100,135, buspirone, its metabolite 1-PP, flesinoxan, and vehicle treatment. The average baseline temperature (±S.D., n = 109) was 37.95 ± 0.042°C. Upon drug administration, a significant decrease in the body temperature for all the 5-HT_1A receptor ligands was observed except for WAY-100,635, which did not show a significant decrease from baseline. Administration of R-8-OH-DPAT, S-8-OH-DPAT, flesinoxan, buspirone, 1-PP, and WAY-100,135 resulted in a maximum decrease in temperature (mean ± S.D.) of 4.0 ± 1.0°C at 40 to 60 min, 3.2 ± 0.7°C at 40 to 60 min, 3.8 ± 0.8°C at 20 to 60 min, 2.8 ± 1.0°C at 40 to 50 min, 1.6 ± 0.5°C at 30 to 40 min, and 0.8 ± 0.8°C at 20 to 40 min, respectively. After reaching a maximal decrease a rapid recovery was observed, followed by a plateau phase, before the body temperature returned to baseline for all compounds except for the higher doses of R-8-OH-DPAT (3 mg/kg administrations), in which the plateau phase was not observed, and the body temperature returned to baseline more gradually (Zuideveld et al., 2001, 2002a, 2002b, 2002c) (see Fig. 2 for examples of individual profiles).

View larger version (29K): [in this window] [in a new window]   Fig. 1. A, average temperature-time profiles (±S.E.M.) for R-8-OH-DPAT: , a 1 mg/kg in 5 min (n = 7); , , and : 3 mg/kg in 5 min (n = 7), in 15 min (n = 5), and in 30 min (n = 6), respectively; , computer-controlled infusions, where the concentration was clamped at 160 ng/ml in blood (n = 6); S-8-OH-DPAT: , 5 mg/kg in 15 min (n = 6) and , 15 mg/kg in 15 min (n = 6) administration; WAY-100,635; , 3 mg/kg in 15 min infusion (n = 6); , , and : 1 mg/kg administration of R-8-OH-DPAT during computer-controlled infusions of WAY-100,635 targeted at 170, 85, and 20 ng/ml in blood, respectively (3 times n = 6) and WAY-100,135; , 10 mg/kg in 15 min infusion (n = 6). All regular infusions started at t = 0. The vertical dotted line denotes time 0 and the horizontal lines denote the average temperature during a vehicle (B) treatment and the lowest temperature reached using R-8-OH-DPAT. B, average temperature-time profiles (±S.E.M.) for buspirone: , 5 mg/kg in 15 min (n = 6); , 15 mg/kg in 15 min (n = 7) infusion; 1-PP: , 10 mg/kg in 15 min infusion (n = 6); flesinoxan: , 3 mg/kg in 5 min (n = 6); , 10 mg/kg in 5 min (n = 6); and , 10 mg/kg in 15 min (n = 6) administration. Vehicle: , the average temperature profiles of 24, which received an equivalent amount of saline in 5, 15, 30, and 160 min. All regular infusions started at t = 0. The vertical dotted line denotes time 0, and the horizontal lines denote the average temperature during a vehicle treatment and the lowest temperature reached using R-8-OH-DPAT (A).

View larger version (32K): [in this window] [in a new window]   Fig. 2. Representative example of individual body temperature versus time profiles following administration of 6 protypical 5-HT1A receptor agonists. The open circles represent the observed values of the body temperature, the dotted line represents the fit with the semi-mechanistic PK-PD model (sigmoid Emax model in combination with the set point transduction model) and the solid line represents the fit obtained with the full mechanistic PK-PD model (operational model of agonism in combination with the set point transduction model) for R-8-OH-DPAT (1 mg/kg in a 5 min infusion), S-8-OH-DPAT (15 mg/kg in a 15 min infusion), flesinoxan (3 mg/kg in a 5 min infusion), buspirone (5 mg/kg in a 15 min infusion), 1-PP (10 mg/kg in a 15 min infusion) and WAY-100,135 (10 mg/kg in a 15 min infusion).

The pharmacokinetic behavior of all the 5-HT_1A receptor agonists could be adequately described using regular 2- and 3-compartmental models. The population pharmacokinetic parameters with the inter-individual variability expressed as a coefficient of variation are depicted in Table 1. The estimates of clearance (CL) and volume of distribution at steady state (Vd_ss) are CL = 22.8, 7.86, 2.96, 17.6, 8.22, 150, and 28.4 ml/min, and Vd_ss = 2820, 10,900, 583, 635, 932, 8750, and 1560 ml, which resulted in a terminal half-life (t) of 86, 900, 136, 25, 79, 20, and 33 min for R-8-OH-DPAT, S-8-OH-DPAT, flesinoxan, buspirone, 1-PP, WAY-100,135, and WAY-100,635, respectively (Zuideveld et al., 2001, 2002a, 2002b, 2002c).

The fraction unbound (mean ± S.D.) was 49 ± 4.2%, 52 ± 2.4%, 30 ± 4.0%, 34 ± 3.0%, 78 ± 2.6%, 20 ± 3.4%, and 19 ± 3.0% for R-8-OH-DPAT, S-8-OH-DPAT, flesinoxan, buspirone, 1-PP, WAY-100,135, and WAY-100,635, respectively. No differences were found between protein binding determined at 34°C or 38°C and across different concentrations (data not shown).

The pharmacokinetic parameter estimates were used to simultaneously fit the set point model to the individual (n = 6–7) body temperature versus time profiles for each agonist (eq. 1) in combination with the sigmoidal E_max model (eq. 3) to obtain estimates of the pharmacodynamic parameters (A, k_in, , , SC₅₀, and n_H, based on whole blood concentrations). See Zuideveld et al. (2001, 2002a, 2002b, 2002c) for the values of the physiological parameters. The drug-related parameters, , SC₅₀, and n_H are represented in Table 2. In the subsequent analysis with the operational model of agonism, values for A, k_in, and were constrained to the population estimates for each individual compound to avoid an increase in parameter precision due to over-parameterization of the model.

Estimation of Apparent Affinity and Efficacy in Vivo. Individual time-body temperature versus time profiles for all agonists were simultaneously analyzed on the basis of the full mechanistic model. Due to the relatively large between-experiment variability in the steepness of the concentration-effect curves, it was not possible to fit all data simultaneously with a single transducer slope parameter, n. Therefore the values of n and the associated variance describing the inter-individual variability were fitted and allowed to vary between compounds. The model converged and estimates of in vivo affinity (pK_A) and efficacy (log ) for each agonist were obtained (Table 3). Overall the population parameter estimates and the inter-individual variability were estimated with good precision. Figure 2 depicts representative fits for each of the compounds.

In Vivo-in Vitro Correlations. The in vivo-in vitro correlations for the different 5-HT_1A receptor agonists were analyzed both with regard to potency and efficacy. Figure 3 shows the correlation between the apparent in vivo pK_A estimates with pK_i values found in vitro in the 5-HT_1A receptor binding assay in the presence of either of [³H]WAY-100,635 and [³H]8-OH-DPAT, respectively. No statistically significant correlations were observed and the coefficients of determination were rather low ([³H]WAY-100,635: r² = 0.55; P > 0.05, and [³H]8-OH-DPAT: r² = 0.37; P > 0.1). The best fit line obtained with linear regression (pK_A = b · pK_i + a) was not equivalent to the line of identity. Thus the slope parameter, b ([³H]WAY-100,635: 0.64 ± 0.78 and [³H]-8-OH-DPAT: 0.60 ± 0.66) were neither significantly different (at P < 0.05) from unity nor zero. The y intercepts, a ([³H]WAY-100,635: 2.49 ± 7.09 and [³H]-8-OH-DPAT: 2.39 ± 6.35) were not significantly different (at P < 0.05) from zero.

View larger version (22K): [in this window] [in a new window]   Fig. 3. Relationship between the apparent in vivo pKA estimates for the effect on body temperature in the rat and pKi values for the 5-HT1A receptor in the presence of either [3H]WAY-100,635 (A) and [3H]8-OH-DPAT (B). Concentrations are not corrected for plasma protein binding. The dashed line represents the line of identity. The solid line was obtained by linear regression (A, r2 = 0.55; P < 0.10 and B, r2 = 0.37; P < 0.20), which yielded the following relation: pKA = (0.64 ± 0.78 · pKi) + 2.49 ± 7.09 (A) and pKA = (0.60 ± 0.66 · pKi) + 2.39 ± 6.35 (B). The symbols correspond to the agonist R-8-OH-DPAT (), S-8-OH-DPAT (), flesinoxan (), buspirone (), 1-PP (), WAY-100,135 (), and WAY-100,635 (). Expressed are population and mean values ± either the coefficient of variation or S.E.M. for the pKA and pKi, respectively. The S.E.M. values for the pKi determined in the presence of [3H]WAY-100,635 values are too small to depict. The coefficient of variation for WAY-100,135 () are large and imprecise and have been left out.

Figure 4A depicts the correlation between the in vitro parameter for efficacy in the receptor binding assay log[agonist ratio] to the in vivo measure for efficacy log . A rather strong correlation was observed with a coefficient of determination of r² = 0.76; P < 0.05. The best fit was obtained with linear regression on the basis of the equation log = b · log[agonist ratio] + a. In this analysis, the slope parameter b was 2.25 ± 2.44, and the y intercept a was –1.77 ± 3.75. Figure 4B depicts the relation between the intrinsic activity in vivo and in vitro efficacy log[agonist ratio] (WAY-100,635 was not included in the analysis since no value was obtained). The solid line shows the predicted relationship that was derived from the operational model of agonism fitting results.

View larger version (22K): [in this window] [in a new window]   Fig. 4. Relationship between the in vivo log estimates obtained for the effect on body temperature in the rat and the log[agonist-ratio] and intrinsic activity (). The solid line was obtained by linear regression (r2 = 0.81; P < 0.03, WAY-100,635 was excluded from the analysis since no value was obtained), which yielded the following relation: log = (2.25 ± 2.45 · log[agonist ratio]) – 1.77 ± 2.25 (A). Relationship between the intrinsic efficacy (log ) found for the 5-HT1A receptor and intrinsic activity () for the in vivo effect on body temperature in rats (B). The solid line shows the predicted relationship that was derived from the operational model of agonism-fitting results. The symbols correspond to the agonists R-8-OH-DPAT (), S-8-OH-DPAT (), flesinoxan (), buspirone (), 1-PP (), WAY-100,135 () and WAY-100,635 (). Expressed are population and mean values with error bars that represent a derived standard deviation for log (and ) and a 95% confidence interval for the log[agonist-ratio].

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
The recently developed semi-mechanistic PK-PD model for 5-HT_1A agonists using the effect on body temperature as a pharmacodynamic endpoint utilizes the empirical Hill equation to characterize the actions of the 5-HT_1A receptor agonists at the receptor in terms of potency (SC₅₀) and intrinsic activity () (Zuideveld et al., 2001, 2002a). Unfortunately the empirical Hill equation has only limited applicability as a model to predict the expression of agonism since both pSC₅₀ and contain mixed information on drug-specific properties and characteristics of the biological system. The prediction of the expression of agonism is of particular interest since it is believed that the pharmacological and therapeutical properties of 5-HT_1A agonists are closely related to the degree of intrinsic activity they display at the 5-HT_1A receptor (De Vry, 1995). It has been demonstrated that the operational model of agonism is a particularly useful tool to explain and predict differential expression of agonism in vivo (Black and Leff, 1983; Van der Graaf and Danhof, 1997a; Van der Graaf et al., 1997c, 1999; Cox et al., 1998; Garrido et al., 2000). The present study has therefore focused on the application of the operational model of agonism in the analysis of the 5-HT_1A receptor-mediated hypothermia. In this manner estimates of the in vivo affinity and efficacy of 5-HT_1A receptor agonists could be obtained and compared with the values found in the receptor binding assay and the cAMP assay.

Estimates of in vivo and in vitro affinity (pK_A and pK_i, respectively) are represented in Tables 3 and 4. The pK_i was estimated on the basis of displacement of both labeled antagonist and agonist. Since the antagonist binds to all the receptors in the inactive state, this pK_i is believed to be the most representative measure for affinity (Assie et al., 1999). The correlation found between the pK_A and pK_i based on [³H]WAY-100,635 was rather poor compared with similar in vivo-in vitro correlations observed for adenosine A₁ agonists, synthetic opiates, and GABA_A receptor agonists. In fact, the correlation was not statistically significant (P > 0.05), which was also the case when using the values obtained with [³H]8-OH-DPAT (P > 0.1) as the radioligand. Interestingly, no specific pattern or frame shift could be observed for the correlation. This indicates that the poor correlation cannot be explained by a systematic difference. Also, no improvement in the correlation was found after correction for the degree of protein binding in vivo ([³H]WAY-100,635: r² = 0.51; P > 0.1 and [³H]8-OH-DPAT: r² = 0.24; P > 0.2). However, close inspection of Fig. 3A shows that it is flesinoxan, which particularly deviates from the line of identity. In fact, the correlation between the pK_A and pK_i became statistically significant when flesinoxan was excluded from the analysis ([³H]WAY-100,635: r² = 0.84; P < 0.05 and [³H]8-OH-DPAT: r² = 0.66; P < 0.05). Recently Van der Sandt et al. (2001) have shown that active transport mechanisms (i.e., P-glycoprotein) at the blood-brain barrier are an important determinant of the brain distribution for flesinoxan. It appears therefore that the in vivo pK_A, which has been determined on the basis of blood concentrations, is not representative for the flesinoxan concentrations at the site of the 5-HT_1A receptor in the brain.

The estimates of in vivo and in vitro efficacy (log and agonist ratio) are shown in Tables 3 and 4. G protein-coupled receptors can exist in two states, with different affinities for agonists and inverse agonists but similar ones for neutral antagonists. The difference in affinity for a compound for the different states is believed to provide a measure of their intrinsic efficacy (Birdsall and Lazareno, 1997). It has been shown by Assie et al. (1999) and Watson et al. (2000) that this ratio is indeed representative for intrinsic activity displayed by agonists at the 5-HT_1A receptor. Between the in vivo and in vitro efficacy (log and log[agonist ratio]) a significant correlation was found (P < 0.05, r² = 0.76, Fig. 4A). The correlation between log and log[agonist ratio] shows further that the in vivo "test assay" is more sensitive in detecting 5-HT_1A activity than the agonist ratio. For example, the significant in vivo agonist activity demonstrated for WAY-100,135 was not detected in vitro. Interestingly, despite the strong correlation, the log[agonist ratio] for buspirone does not appear to be a good predictor of its log . The log[agonist ratio] of buspirone is higher than that of the full agonist R-8-OH-DPAT despite the fact that in vivo the intrinsic efficacy of buspirone is similar to that of the partial agonist S-8-OH-DPAT. This discrepancy might be caused by buspirone's activity at other receptors present in the in vivo test assay (New, 1990).

Figure 4B shows the relationship between the intrinsic activity in vivo for the effect on body temperature in rats found for the 5-HT_1A receptor and in vitro measure for efficacy log[agonist ratio]. The solid line shows the predicted relationship that was derived from the operational model of agonism fitting results (eq. 5). Interestingly Fig. 4B also shows that most log[agonist ratio] values are somewhat lower than predicted based on the model. This apparent rightward shift in the curve can be explained by weight of the log[agonist ratio] value for buspirone. In theory, the resulting "under-prediction" of the efficacy parameter log could be explained by antagonism of buspirone in these experiments by its own active metabolite 1-PP, as it is a more partial agonist than buspirone itself. However, we have previously shown that based on the in vivo potency and the blood concentrations of 1-PP during intravenous buspirone administration, this metabolite is not expected to contribute significantly (Zuideveld et al., 2002c).

In conclusion, by combining the previously proposed semimechanistic PK-PD model for the hypothermic effect of 5-HT_1A agonists with the operational model of agonism, a full mechanistic PK-PD model has been obtained, which is highly predictive of the in vivo intrinsic activity of ligands at this receptor. This is important, since the pharmacological and therapeutical properties of 5-HT_1A agonists are closely related to the degree of intrinsic activity at the 5-HT_1A receptor. The ability of the in vivo assay to detect weak partial agonism, given the poor signal from the in vitro assay (WAY-100,135) underscores the importance of the use of in vivo models in the development of 5-HT_1A antagonists as clinical agents.

	Acknowledgements

 
The generous donation of buspirone and 1-PP by Bristol-Myers Squibb and flesinoxan by Solvay Pharmaceuticals is highly appreciated.

	Footnotes

 
DOI: 10.1124/jpet.103.059030.

ABBREVIATIONS: 5-HT, 5-hydroxytryptamine, serotonin; R-8-OH-DPAT, R-(+)-8-hydroxy-2-(di-n-propylamino)tetralin; S-8-OH-DPAT, S-(+)-8-hydroxy-2-(di-n-propylamino)tetralin; PK-PD, pharmacokinetic-pharmacodynamic; 1-PP, 1-(2-pyrimidinyl)-piperazine; HPLC, high performance liquid chromatography; MS, mass spectrometry; SPA, scintillation proximity assay; NSB, nonspecific binding; DMSO, dimethyl sulfoxide; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; e.i.v., error-in-variables; , upper asymptote of the concentration effect relationship; AIC, Akaike Information Criterion; , residual error; , inter-individual variation; , amplification of the set point signal; SC₅₀, concentration at 50% maximum stimulus; S_max, maximum stimulus the drug can produce; WAY-100,135, N-(1,1-dimethylethyl)-4-(2-methoxyphenyl)--phenyl-1-piperazinepropanamide; WAY-100,635, N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-2-pyridinyl-cyclohexanecarboxamide.

Address correspondence to: Prof. Meindert Danhof, Leiden/Amsterdam Center for Drug Research, Division of Pharmacology, Gorlaeus Laboratory, P.O. Box 9502, 2300 RA, Leiden, The Netherlands. E-mail: M.Danhof{at}LACDR.LeidenUniv.nl

	References

Top Abstract Materials and Methods Results Discussion References

 

Akaike H (1974) A new look at the statistical model identification. IEEE (Inst Electr Electron Eng) Trans Automatic Control 19: 716–723.
Albert PR, Zhou QY, Van Tol HH, Bunzow JR, and Civelli O (1990) Cloning, functional expression, and mRNA tissue distribution of the rat 5-HT_1A receptor gene. J Biol Chem 265: 5825–5832.[Abstract/Free Full Text]
Assie M, Cosi C, and Koek W (1999) Correlation between low/high affinity ratios for 5-HT_1A receptors and intrinsic activity. Eur J Pharmacol 386: 97–103.[CrossRef][Medline]
Barlow RB (1967) Use of an antagonist for estimating the degree of agonist stimulation during physiological release. Trends Pharmacol Sci 16: 262–264.[CrossRef]
Birdsall NJ and Lazareno S (1997) To what extent can binding studies allow the quantification of affinity and efficacy? Ann N Y Acad Sci 812: 41–47.[CrossRef][Medline]
Black JW and Leff P (1983) Operational models of pharmacological agonism. Proc R Soc Lond B Biol Sci 220: 141–162.[Medline]
Boeckman A, Sheiner LB, and Beal SL (1992) NONMEM Users guide, NONMEM project group, University of California, San Francisco.
Casella G and Berger RL (1990) Statistical Inference, Wadsworth & Brooks, Pacific Grove.
Cheng Y and Prusoff WH (1973) Relationship between the inhibition constant (K_i) and the concentration of inhibitor which causes 50 per cent inhibition (I₅₀) of an enzymatic reaction. Biochem Pharmacol 22: 3099–3108.[CrossRef][Medline]
Cox EH, Kerbusch T, Van der Graaf PH, and Danhof M (1998) Pharmacokinetic-pharmacodynamic modeling of the electroencephalogram effect of synthetic opioids in the rat: correlation with the interaction at the mu-opioid receptor. J Pharmacol Exp Ther 284: 1095–1103.[Abstract/Free Full Text]
Dayneka NL, Garg V, and Jusko WJ (1993) Comparison of four basic models of indirect pharmacodynamic responses. J Pharmacokinet Biopharm 21: 457–478.[CrossRef][Medline]
De Vry J (1995) 5-HT_1A receptor agonists: recent developments and controversial issues. Psychopharmacology Ser (Berl) 121: 1–26.
Garrido M, Gubbens-Stibbe J, Tukker E, Cox E, von Frijtag J, Kunzel D, IJzerman A, Danhof M, and Van der Graaf PH (2000) Pharmacokinetic-pharmacodynamic analysis of the EEG effect of alfentanil in rats following beta-funaltrexamine-induced mu-opioid receptor "knockdown" in vivo. Pharm Res 17: 653–659.[CrossRef][Medline]
Leff P, Prentice DJ, Giles H, Martin GR, and Wood J (1990) Estimation of agonist affinity and efficacy by direct, operational model-fitting. J Pharmacol Methods 23: 225–237.[CrossRef][Medline]
Lindstrom ML and Bates DM (1990) Nonlinear mixed effects models for repeated measures data. Biometrics 46: 673–687.[CrossRef][Medline]
New JS (1990) The discovery and development of buspirone: a new approach to the treatment of anxiety. Med Res Rev 10: 283–326.[Medline]
Pauwels PJ, Van Gompel P, and Leysen JE (1993) Activity of serotonin (5-HT) receptor agonists, partial agonists and antagonists at cloned human 5-HT_1A receptors that are negatively coupled to adenylate cyclase in permanently transfected HeLa cells. Biochem Pharmacol 45: 375–383.[CrossRef][Medline]
Tuk B, van Gool T, and Danhof M (2002) Mechanism-based pharmacodynamic modeling of the interaction of midazolam, bretazenil, and zolpidem with ethanol. J Pharmacokinet Pharmacodyn 29: 235–250.[CrossRef][Medline]
Tuk B, Van Oostenbruggen MF, Herben VM, Mandema JW, and Danhof M (1999) Characterization of the pharmacodynamic interaction between parent drug and active metabolite in vivo: midazolam and -OH-midazolam. J Pharmacol Exp Ther 289: 1067–1074.[Abstract/Free Full Text]
Van der Graaf PH and Danhof M (1997a) Analysis of drug-receptor interactions in vivo: a new approach in pharmacokinetic-pharmacodynamic modelling. Int J Clin Pharmacol Ther 35: 442–446.[Medline]
Van der Graaf PH and Danhof M (1997b) On the reliability of affinity and efficacy estimates obtained by direct operational model fitting of agonist concentration-effect curves following irreversible receptor inactivation. J Pharmacol Toxicol Methods 38: 81–85.[CrossRef][Medline]
Van der Graaf PH, Van Schaick EA, Mathot RAA, IJzerman AP, and Danhof M (1997c) Mechanism-based pharmacokinetic-pharmacodynamic modelling of the effects of N6-cyclopentyladenosine analogues on heart rate in rat: estimation of in vivo operational affinity and efficacy at adenosine A₁ receptors. J Pharmacol Exp Ther 283: 809–816.[Abstract/Free Full Text]
Van der Graaf PH, Van Schaick EA, Visser SA, de Greef HJ, IJzerman AP, and Danhof M (1999) Mechanism-based pharmacokinetic-pharmacodynamic modeling of antilipolytic effects of adenosine A₁ receptor agonists in rats: prediction of tissue-dependent efficacy in vivo. J Pharmacol Exp Ther 290: 702–709.[Abstract/Free Full Text]
Van der Sandt ICJ, Smolders R, Nabulsi L, Zuideveld KP, De Boer AG, and Breimer DD (2001) Active efflux of the 5-HT_1A receptor agonist flesinoxan via P-glycoprotein at the blood-brain barrier. Eur J Pharm Sci 14: 81–86.[CrossRef][Medline]
Visser SA, Gladdines WW, van der Graaf PH, Peletier LA, and Danhof M (2002) Neuroactive steroids differ in potency but not in intrinsic efficacy at the GABA_A receptor in vivo. J Pharmacol Exp Ther 303: 616–626.[Abstract/Free Full Text]
Visser SA, Wolters FL, Gubbens-Stibbe JM, Tukker E, Van Der Graaf PH, Peletier LA, and Danhof M (2003) Mechanism-based pharmacokinetic/pharmacodynamic modeling of the electroencephalogram effects of GABA_A receptor modulators: in vitro-in vivo correlations. J Pharmacol Exp Ther 304: 88–101.[Abstract/Free Full Text]
Watson J, Collin L, Ho M, Riley G, Scott C, Selkirk JV, and Price GW (2000) 5-HT_1A receptor agonist-antagonist binding affinity difference as a measure of intrinsic activity in recombinant and native tissue systems. Br J Pharmacol 130: 1108–1114.[CrossRef][Medline]
Zuideveld KP, Maas HJ, Treijtel N, Hulshof J, Van der Graaf PH, Peletier LA, and Danhof M (2001) A set-point model with oscillatory behavior predicts the time course of 8-OH-DPAT-induced hypothermia. Am J Physiol Regul Integr Comp Physiol 281: R2059–R2071.[Abstract/Free Full Text]
Zuideveld KP, Rusiç-Pavletiç J, Maas HJ, Peletier LA, Van der Graaf PH and Danhof M (2002a) Pharmacokinetic-pharmacodynamic modeling of the hypothermic effect induced by buspirone and its metabolite 1(2-pyrimidinyl)-piperazine. J Pharmacol Exp Ther 303: 1130–1137.[Abstract/Free Full Text]
Zuideveld KP, Treijtel N, Maas HJ, Gubbens-Stibbe JM, Peletier LA, Van der Graaf PH, and Danhof M (2002b) A competitive interaction model predicts the effect of WAY-100,635 on the time course of R-(+)-8-hydroxy-2-(di-n-propylamino)tetralin-induced hypothermia. J Pharmacol Exp Ther 300: 330–338.[Abstract/Free Full Text]
Zuideveld KP, Treijtel N, Van der Graaf PH, and Danhof M (2000) Enantioselective high-performance liquid chromatographic analysis of the 5-HT_1A receptor agonist 8-hydroxy-2-(di-n-propylamino)tetralin. Application to a pharmacokinetic-pharmacodynamic study in rats. J Chromatogr B Biomed Sci Appl 738: 67–73.[CrossRef][Medline]
Zuideveld KP, Van Gestel A, Peletier LA, Van der Graaf PH, and Danhof M (2002c) Pharmacokinetic-pharmacodynamic modelling of the hypothermic and corticosterone effects of the 5-HT_1A-receptor agonists flesinoxan. Eur J Pharmacol 445: 43–54.[CrossRef][Medline]

This article has been cited by other articles:

				A. Yassen, J. Kan, E. Olofsen, E. Suidgeest, A. Dahan, and M. Danhof Mechanism-Based Pharmacokinetic-Pharmacodynamic Modeling of the Respiratory-Depressant Effect of Buprenorphine and Fentanyl in Rats J. Pharmacol. Exp. Ther., November 1, 2006; 319(2): 682 - 692. [Abstract] [Full Text] [PDF]

				J. H. De Maeyer, N. H. Prins, J. A. J. Schuurkes, and R. A. Lefebvre Differential Effects of 5-Hydroxytryptamine4 Receptor Agonists at Gastric versus Cardiac Receptors: An Operational Framework to Explain and Quantify Organ-Specific Behavior J. Pharmacol. Exp. Ther., June 1, 2006; 317(3): 955 - 964. [Abstract] [Full Text] [PDF]

				S. A. G. Visser, B. Sallstrom, T. Forsberg, L. A. Peletier, and J. Gabrielsson Modeling Drug- and System-Related Changes in Body Temperature: Application to Clomethiazole-Induced Hypothermia, Long-Lasting Tolerance Development, and Circadian Rhythm in Rats J. Pharmacol. Exp. Ther., April 1, 2006; 317(1): 209 - 219. [Abstract] [Full Text] [PDF]

				A. Yassen, E. Olofsen, A. Dahan, and M. Danhof Pharmacokinetic-Pharmacodynamic Modeling of the Antinociceptive Effect of Buprenorphine and Fentanyl in Rats: Role of Receptor Equilibration Kinetics J. Pharmacol. Exp. Ther., June 1, 2005; 313(3): 1136 - 1149. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: jpet.103.059030v1 308/3/1012    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Zuideveld, K. P. Articles by Danhof, M. Search for Related Content PubMed PubMed Citation Articles by Zuideveld, K. P. Articles by Danhof, M.