Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Endothelin-1 is a potent vasoconstrictor and growth factor released from the vascular endothelium. It has been implicated in the pathophysiology of a number of diseases (Miyauchi and Masaki, 1999) and as such represents an attractive drug target, particularly in the cardiovascular area (Kentsch and Otter, 1999). Selective and mixed endothelin A (ET_A)¹ and endothelin B (ET_B) receptor antagonists have been investigated for the treatment of essential hypertension, chronic heart failure, pulmonary hypertension, and in the prevention of restenosis (Dupuis, 2000).

Based on chemical structure precedents (Williams et al., 1995), a rational drug discovery program was undertaken to provide a selective ET_A receptor antagonist (Rawson et al., 2001). Potency and selectivity were achieved in a series of indole-containing compounds; however, these were found to have poor oral bioavailability in rat pharmacokinetic studies. Based on additional studies examining the systemic pharmacokinetic properties of these agents and their physicochemical profiles, it became apparent that the poor oral pharmacokinetic profiles were primarily the result of poor permeation of the membranes of the gastrointestinal tract. In vitro and physicochemical evidence indicated that absorption was also likely to be limited in humans. The subsequent medicinal chemistry program and pharmacokinetic studies were therefore directed toward achieving improved oral absorption within this chemical series. Throughout the program, reference to physicochemical properties, in particular the H-bonding capacity of the molecules was used to improve the understanding of structural requirements for membrane permeation. Further pharmacokinetic studies were performed in both rat and dog to more fully profile the absorption properties of the eventual lead compound, compound 7a (single active enantiomer of compound 7). Pharmacological potency and selectivity were key reasons for focusing on the series of compounds discussed. However, for the purposes of this evaluation, potency was not a primary determinant of compounds that underwent pharmacokinetic testing and physicochemical profiling, hence potency values are not presented and only considered in broad terms.

The aims of the work described here were therefore 2-fold: first to provide physicochemical and pharmacokinetic guidance in the design of an oral ET_A antagonist; and second, to define the pharmacokinetic profile of the selected compound, compound 7a.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Compounds 1 to 16 (Fig. 1) were synthesized as part of a medicinal chemistry program at Pfizer Global Research and Development, Sandwich, UK. All compounds were initially synthesized as racemic mixtures. Compounds 6 and 7 were subsequently resolved into individual enantiomers by preparative chromatography and these single enantiomers subjected to further evaluation. Single enantiomers were then synthesized by inclusion of a chiral resolution step and these compounds were designated compound 6a [(S)-(+)-3-{1-(1,3-benzodioxol-5-yl)-2-[(2-methoxy-4-methylphenyl)sulfonamido]-2-oxoethyl}-6-(carboxy)-1-methyl-1H-indole] and compound 7a [(S)-(+)-3-{1-(1,3-benzodioxol-5-yl)-2-[(2-methoxy-4-methylphenyl)sulfonamido]-2-oxoethyl}-6-(hydroxymethyl)-1-methyl-1H-indole].

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Structures of ETA antagonists 1 to 16 that were evaluated for physicochemical and pharmacokinetic properties. Compound 6a and compound 7a are the single (+) enantiomers of compounds 6 and 7, respectively.

Rat Pharmacokinetic Studies. Male Sprague-Dawley rats (~250 g; Charles River, Manston, UK) were surgically prepared with jugular vein catheters at least 2 days before dose administration. For each compound (1-16) one animal received an intravenous dose (2 mg/kg) in a vehicle containing up to 20% dimethyl sulfoxide (DMSO) in saline and one animal received an oral dose (10 mg/kg) in a vehicle containing up to 20% DMSO in water. Dose volumes were approximately 1 ml/kg for intravenous doses and 2 ml/kg for oral doses. Blood samples (200 µl) were collected from the indwelling jugular vein catheter up to 24 h after dose administration and placed into heparinized tubes. These samples were centrifuged at 3000 rpm for 10 min and the plasma removed and stored frozen before analysis.

Dog Pharmacokinetic Studies. Compound 7a and compound 6a were intravenously administered to two male and one female dog (Pfizer colony, 12-18 kg) at dose levels of 0.2 and 0.5 mg/kg, respectively. Compound 7a was dissolved (0.2 mg/ml) in a vehicle of 1% DMSO (v/v) in saline containing 10% (w/v) hydroxypropyl -cyclodextrin and administered by intravenous infusion over 15 min into a saphenous vein. Compound 6a (0.5 mg/ml) was prepared in the same vehicle and similarly administered. Compound 7a was orally administered by gavage to the same three dogs 7 days after the intravenous dose. The dose level was 2 mg/kg and the vehicle (10-ml total volume) was the same as that used for intravenous dosing but with 10% (v/v) DMSO. Blood samples (5 ml) were collected from temporary indwelling saphenous vein catheters or by venepuncture of the cephalic vein. The blood samples were transferred to lithium heparin tubes, mixed, and centrifuged. Plasma samples were transferred to glass vials and stored frozen before analysis. Urine samples (0-7 h) were collected from male dogs after intravenous administration by catheterization of the bladder and stored frozen before analysis.

Analysis of ET_A Antagonists in Plasma Samples from Rats and Dogs. The concentrations of the ET_A antagonists were determined in rat and dog plasma by a solid phase extraction method followed by HPLC with ultraviolet detection. The method involved addition of internal standard (1 µg of analog with suitable HPLC retention time) to 0.1 ml of rat plasma or 1.0 ml of dog plasma. The plasma samples were diluted with 1.0 ml of 0.1 M sodium citrate buffer, pH 4.5, and applied to solid phase extraction cartridges (C₁₈ Isolute; Analytichem International, Harbor City, CA) and the cartridge washed with a further 1 ml of buffer. Analytes and internal standard were then eluted with 1 ml of methanol and evaporated to dryness under a stream of nitrogen at 37°C. The residues were reconstituted in 100 µl of methanol/water (1:1, v/v) and 80 µl injected onto the HPLC column (RPB 25 × 0.5 cm; Hichrom, Reading, UK). The column was eluted at 1 ml/min with a mobile phase of which the aqueous component comprised a solution of 10 mM sodium octane sulfonate, 20 mM sodium dihydrogen orthophosphate adjusted to pH 3 with concentrated hydrochloric acid. The aqueous phase was mixed with up to 60% acetonitrile (v/v) to elute the compounds within a retention time of 15 min. Detection was by ultraviolet absorption at a wavelength of 295 nm. The limits of quantitation were 50 to 100 ng/ml in rat plasma and 10 to 20 ng/ml in dog plasma. Formal validation of the analytical procedure was only carried out for compound 7a and compound 6a for which analysis of quality control samples provided values within 15% of the added value throughout calibration ranges of 50 to 10,000 ng/ml in rat plasma and 10 to 2,000 ng/ml in dog plasma. Dog and rat urine samples were analyzed by the same method with appropriate sample volumes.

Permeability across Caco-2 Cell Monolayers. The permeability of compounds 1 to 16 was assessed using Caco-2 cell monolayers prepared as previously described (Yee, 1997). Compound permeability was assessed at an initial concentration of 10 µM applied in the apical chamber of the apparatus. Determining the amount of compound present in the basolateral chamber after 3-h incubation at 37°C assessed the extent of permeation. All analyses were performed in duplicate. Quantitation of compound was by modification of the method described previously for plasma.

Plasma Protein Binding Determinations. Samples of rat and dog plasma (1 ml) containing ET_A antagonists at a concentration of 5 µg/ml were dialyzed (Spectrapor 1 dialysis membrane 6000-8000 molecular weight cut-off; Spectrum Medical Industries, Rancho Dominquez, CA) against isotonic Krebs-Ringer-bicarbonate buffer, pH 7.4 (1 ml), for 4 h at 37°C in a rotating dialyzer (Dianorm; NBS Biologicals, Huntingdon, UK). Triplicate determinations were performed on pooled samples of male rat plasma for each compound. After dialysis, the concentrations of drug in plasma and buffer were determined using solid phase extraction and HPLC analysis, as previously described. For compound 7a and compound 6a additional plasma protein binding determinations were made in control dog and human plasma.

Pharmacokinetic Analysis of Data. Terminal elimination rate constant (K_el) was determined by linear regression of the log plasma concentrations. The terminal elimination half-life (t_1/2) was calculated from 0.693/K_el. The area under the plasma concentration time curve (AUC) was calculated to the last time point at which drug could be measured using the linear trapezoidal rule and extrapolated to infinity by using K_el. Clearance (CL) was calculated by the relationship dose/AUC. The volume of distribution was calculated by the relationship CL/K_el. Oral bioavailability was calculated from the ratio of AUCs after oral and intravenous doses after normalizing for the dose. Unbound clearance was calculated by dividing the systemic clearance value after intravenous dosing (CL) by the free fraction measured in plasma. The extent of first-pass extraction (E) was estimated by reference to the well stirred model of hepatic clearance and assuming hepatic blood flow of 100 ml/min/kg in rat and 50 ml/min/kg in dog. Based on the bioavailability observed and the estimate of hepatic extraction the extent of absorption (A) was estimated (A = F/(1  E). For compound 7a, which was orally dosed to rats and plasma sampled from the hepatic portal vein, the extent of absorption was calculated from the relationship CL multiplied by portal vein AUC. This assumes that no significant extraction occurs in passage across the gut wall (Kwon and Inskeep, 1996) and that the clearance is solely hepatic. These assumptions are supported by observations of low renal clearance and stability to oxidative metabolism in liver microsomal preparations.

Lipophilicity Determination. Distribution of the ET_A antagonists (compounds 1-16) between octanol and 0.1 M sodium phosphate buffer, pH 7.4, was determined by the method of Stopher and McClean (1990). Approximately 0.1 mg of compound was dissolved in 1 ml of octanol (octan-1-ol, specially pure; BDH, Poole, UK) and mixed with 1 ml 0.1 M sodium phosphate buffer, pH 7.4, on a rotary mixer at 30 rpm for 60 min. After centrifugation the two phases were separated and duplicate 5-µl aliquots of each phase directly injected onto the HPLC system described previously. The distribution coefficient (D_7.4) was calculated from the ratio of the concentration of compound in octanol to the concentration of compound in buffer. Calculated log P values (C log P) were calculated using the Medchem computer program (version 3.55; Biobyte Corp., Claremont, CA).

Assessment of H-Bond Free Energy. For each of the ET_A antagonists, the free energy-based hydrogen bond acceptor and donor factors, Ca and Cd, were computed with the HYBOT 5.0 program and database by using experimental data from 12,000 H-bonded complexes (Raevsky, 1997). The total H-bonding capability was determined from the sum of the Ca and Cd values to provide hydrogen-bond free energy (Cad) values, as previously described for the comparison of H-bonding capability and permeation (van de Waterbeemd et al., 1996). The number of H-bond acceptors (HBAs) was simply assessed from counting the total number of nitrogen and oxygen atoms in each molecule, whereas the total number of HBDs was obtained from the total number of ---NH and ---OH functions within each molecule (Lipinski et al., 1997).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pharmacokinetic Studies with ET_A Antagonists 1 to 16 in Rat. Pharmacokinetic parameters for racemic ET_A antagonists 1 to 6 (Fig. 1) obtained after single intravenous (2 mg/kg) and oral (10 mg/kg) doses (n = 1 per compound per route) to male jugular vein-cannulated rats are presented in Table 1. This table also includes the free fraction in plasma obtained from plasma protein binding experiments. The same parameters for racemic ET_A antagonists 7 to 16 are presented in Table 2.

Physicochemical Parameters. Physicochemical parameters of ET_A antagonists 1 to 16 obtained through experimental methods (Caco-2 permeability and log D_7.4) and by computational analysis (molecular weight, number of HBAs and HBDs, C log P, and Cad are presented in Table 3.

Pharmacokinetics of Compound 6a and Compound 7a in Rat. After single intravenous doses of compound 6a (single enantiomer of ET_A antagonist 6) to male jugular vein-cannulated rats (2 mg/kg, n = 3), mean values for plasma clearance, volume of distribution, and elimination half-life were 1.7 ml/min/kg, 0.16 l/kg, and 1.1 h, respectively. Compound 6a was observed as a circulating metabolite of compound 7a (single enantiomer of ET_A antagonist 7) after oral and intravenous administration of this compound. The plasma concentration time profiles of compound 7a and compound 6a after single intravenous (2 mg/kg, n = 3) and single oral (10 mg/kg, n = 2) administration of compound 7a are shown in Fig. 2. Mean pharmacokinetic parameters for the parent compound, 7a, after intravenous doses were plasma clearance of 20 ml/min/kg, volume of distribution of 1.9l/kg, and an elimination half-life of 1.1 h. The mean oral bioavailability of compound 7a (by comparison with dose-normalized AUC values) was 14%. Comparison with dose-normalized AUC values for compound 6a after oral administration of compound 7a and intravenous administration of compound 6a provided a value for the bioavailability of compound 6a from oral compound 7a of 4%. The pharmacokinetic parameters of compound 7a and compound 6a in male rats are summarized in Table 4. Compound 7a was not detectable in rat urine after intravenous dosing of either compound 7a or compound 6a. The extent of urinary excretion of compound 6a in male rats over 24 h amounted to 2.0 and 1.6% of the intravenous doses of compound 6a and compound 7a, respectively.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Plasma concentration time profiles of compound 7a and compound 6a after single intravenous doses (2 mg/kg, n = 3) of compound 7a (A) and single oral doses (10 mg/kg, n = 2) of compound 7a (B) to jugular vein-cannulated male rats.

Pharmacokinetics of Compound 7a and Compound 6a in Dog. After single intravenous doses of compound 6a to dogs (0.5 mg/kg, n = 3), the mean elimination half-life was 0.5 h, due to a plasma clearance of 5.8 ml/min/kg and a volume of distribution of 0.3 l/kg. Urinary excretion of 6a (0-7 h) accounted for 8.2% of the dose (n = 2 male dogs).

Plasma Protein Binding and Unbound Clearance Estimates of Compound 7a and Compound 6a. The plasma protein binding of compound 7a in rat and dog plasma was determined at 20 µg/ml (n = 2) by equilibrium dialysis. The compound was highly bound with protein binding values of 99.6 and 99.0% in rat and dog plasma, respectively. Unbound clearance values of compound 7a in rat and dog were therefore 5000 and 410 ml/min/kg, respectively. The protein binding of compound 6a was determined at a concentration of 10 µg/ml (n = 3) and was found to be 99.0% in rat plasma and 90.7% in dog plasma. Unbound clearance values of compound 6a in rat and dog were therefore 180 and 62 ml/min/kg, respectively. Protein binding determinations were also performed in control human plasma for both compounds at the same concentrations as those studied in animals. Human plasma protein binding of compound 7a and compound 6a was 99.3 and 99.7%, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Structure Pharmacokinetic Relationships for ET_A Antagonists. ET_A antagonists 1 to 6 all show extremely low oral bioavailability (less than 3%) in the rat (Table 1). On the basis that systemic clearance after intravenous dosing of all of these compounds was less than 12 ml/min/kg, which is less than 20% liver blood flow (estimated at 100 ml/min/kg), high first-pass extraction would not be expected to account for this. In addition compounds showed negligible turnover in standard in vitro metabolism screens (method as in Walker et al., 1999) in rat liver microsomes (data not shown), indicating low metabolic turnover. This strongly suggests that the compounds are poorly absorbed across the membranes of the gastrointestinal tract. In terms of the broad physicochemical properties of the molecules, all have molecular weights in excess of 500, a property previously related to a propensity for poor absorption (Lipinski et al., 1997).

6

1

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Graph showing the correlation between the calculated H-bonding capacity of ETA antagonists 1 to 16 and their permeability across Caco-2 monolayers. H-bonding capacity is expressed as Cad values, which represent the sum of absolute values of free energy H-bond factors, characterizing the total H-bond ability of each compound.

6

1

6

1

6

1

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Graph showing the correlation between the permeability of ETA antagonists 1 to 16 across Caco-2 monolayers and their estimated extent of absorption in rat. Absorption in the rat was estimated based on bioavailability after oral doses and accounting for expected hepatic first-pass extraction.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Scatter plot demonstrating the lack of a relationship between plasma clearance and total H-bonding capacity (Cad value) for the series of ETA antagonists 1 to 16.

Pharmacokinetics of Compound 7a and Compound 6a. The pharmacokinetic studies in rats with the single enantiomer of racemic ET_A antagonist 7 (compound 7a) provided a pharmacokinetic profile broadly similar to that observed for the racemate. However, the lower value obtained for plasma clearance (20 versus 72 ml/min/kg) and similar oral bioavailability (14 versus 16%) suggested the original estimate of absorption for the racemate (~55%) was too high. It is not known whether the different intravenous vehicles used for compounds 7 and 7a influenced the pharmacokinetics observed. Separate studies in which plasma concentrations of compound 7a were determined in the hepatic portal vein of nonjugular vein cannulated rats provided a measure of absorption of 22%, assuming no gut wall extraction (Kwon and Inskeep, 1996). Alternatively, the administration of compound into the hepatic portal vein allowed absorption to be estimated based on comparison of normalized systemic AUC values after oral and portal vein dosing and provided a value of 36%. Thus, in the rat the absorption of compound 7a would appear to be around 30%. In contrast, absorption of compound 7a in the dog was complete with 100% oral bioavailability observed. This is in keeping with comparisons of other compounds in rat and dog where dog has provided a higher extent of absorption than rat and where rat is the better predictor of human absorption (Beaumont et al., 2000; van de Waterbeemd et al., 2001).