SB-242235, a selective inhibitor of p38 mitogen-activated protein kinase. I: preclinical pharmacokinetics

1. SB-242235 (1-(4-piperidinyl)-4-(4-fluorophenyl)-5-(2-methoxy-4-pyrimidinyl) imidazole) is a potent and selective p38 MAP kinase inhibitor that may be an effective therapy for cytokine-mediated diseases such as autoimmune or inflammatory diseases. The present studies were conducted to evaluate the pharmacokinetics of SB-242235 in several preclinical species, including rat, dog and monkey. 2. SB-242235 demonstrates generally favourable pharmacokinetic properties in all species examined. Systemic plasma clearance was high in rat, but in the non-rodent species SB-242235 demonstrated low to moderate clearance with plasma half-lives > 4h. Oral bioavailability in each preclinical species was high. In rat and monkey, SB-242235 demonstrated non-linear elimination kinetics that manifested as a decrease in clearance with increasing dose and apparent oral bioavailability > 100% at high oral doses. Furthermore, SB-242235 displayed concentration-dependent plasma protein binding over a concentration range of 1000-10,000 ng ml(-1). 3. In conclusion, SB-242235 demonstrates high oral bioavailability across the major preclinical species, and may thus be a useful tool compound for investigation of the role of p38 inhibition in various disease states. However, the observations of non-linear protein binding and disposition also suggest the need for caution in the design of and data interpretation from such studies.

SB-242235, a selective inhibitor of p38 mitogen-activated protein kinase. II: in vitro and in vivo metabolism studies and pharmacokinetic extrapolation to man

1. Inhibition of p38 MAP kinase has been investigated extensively as a potential therapy for cytokine-mediated diseases such as autoimmune and inflammatory diseases. SB-242235 (1-(4-piperidinyl)-4-(4-fluorophenyl)-5-(2-methoxy-4-pyrimidinyl) imidazole) is a potent and selective p38 MAP kinase inhibitor; the preclinical pharmacokinetics of SB-242235 have been described previously. The present studies were conducted to describe the in vitro metabolic rates and routes of SB-242235 metabolism, to characterize its in vivo preclinical metabolism, and to use these data to aid in the prediction of the pharmacokinetic behaviour of SB-242235 in man. 2. SB-242235 was metabolically stable in rat, dog, monkey and human hepatic microsomes, isolated hepatocytes and liver slices in vitro. The in vivo preclinical metabolism studies were consistent with the in vitro findings; SB-242235 was minimally metabolized, and was primarily excreted unchanged in the urine (45 and 67% of the administered dose in the rat and monkey, respectively). 3. Allometric scaling using various correction factors predicted that SB-242235 would have low clearance in man with a predicted half-life ranging from 11.5 to 18.7h. This prediction was consistent with the observed mean half-life of 16.4h in the first-in-man study for SB-242235. An allometric scaling method with a correction for interspecies differences in glomerular filtration rate provided the most accurate prediction of the pharmacokinetic behaviour of SB-242235 in humans, although the clinical data also highlight potential difficulties in conducting prospective allometry.

SB-239063, a potent and selective inhibitor of p38 map kinase: preclinical pharmacokinetics and species-specific reversible isomerization

PURPOSE

A series of studies was conducted to evaluate the preclinical pharmacokinetics of SB-239063 (trans-1-(4-hydroxycyclohexyl)-4-(4-fluorophenyl)-5-[(2-methoxy)pyrimidin-4-yl] imidazole), a potent and selective p38 MAP kinase inhibitor.

METHODS

SB-239063 was administered both i.v. and p.o. in the rat, dog, cynomolgus monkey, and rhesus monkey, with standard pharmacokinetic parameters generated from the concentration vs. time data.

RESULTS

Initial rat studies suggested possible nonlinear disposition, however, assay refinement revealed an in vivo trans-cis isomerization of SB-239063 to a metabolite with nearly identical chromatographic and mass spectral properties. SB-239063 exhibited low to moderate clearance and good bioavailability in the rat and dog, but poor bioavailability in the cynomolgus monkey. Substantial in vivo trans-cis isomerization occurred in the rat and cynomolgus monkey, but occurred to a far lesser extent in the dog. The isomerization reaction was reversible, with a recycled fraction of 0.20 and 0.0003 in the rat and cynomolgus monkey, respectively. In the rhesus monkey, bioavailability was also poor. but no in vivo isomerization was observed. Conclusions. These studies demonstrate the necessity of exercising vigilance in conducting high-throughput analytical method development, and the importance of using a variety of preclinical species when evaluating the disposition of new drug candidates.

Development of an in vivo preclinical screen model to estimate absorption and bioavailability of xenobiotics

The purpose of this study was to develop an in vivo screening method for rapid preclinical characterization of absorption and bioavailability of large numbers of compounds. This effort involved several steps. First, a pharmacokinetic characterization of a reference compound was conducted in the monkey. These data were used to verify theoretical calculations of a maximal portal dose-normalized area under the concentration-time curve. Next, a monkey screen was implemented using mixtures of up to five compounds each (i.e., cassettes) to estimate the bioavailability of approximately 200 compounds. Cassettes were administered as a single intraduodenal dose to a single monkey followed by simultaneous portal and systemic blood sampling. Definitive studies were then conducted to determine absolute bioavailability of 14 of these compounds. The studies with the reference compound demonstrated that the theoretical methodology based on a single intraduodenal dose with portal and systemic sampling provided consistent estimates of bioavailability. In the screen studies, approximately 75% of the test compounds were excluded from further evaluation due to poor absorption. Of the 14 compounds selected for follow-up evaluation from both well and poorly absorbed compounds, the absolute bioavailability of 10 of them were correctly classified from the screening data. The remaining 4 compounds were false positives, which showed low bioavailability; no false negatives were encountered. This approach allows for a rapid and reliable screen to evaluate absorption and bioavailability using a single dose in a preclinical model.

The effect of rate of drug administration on the extent and time course of phencyclidine distribution in rat brain, testis, and serum

The goal of these studies was to examine the relationship between the rate of phencyclidine (PCP) administration and PCP tissue distribution. The time course of PCP distribution in serum, brain, and testis after rapid (i.v.) and slow (s.c.) administration was studied. Brain and serum PCP concentrations after an i.v. bolus dose (1 mg/kg at 900 microg/min) were highest at 30 s and decreased biphasically, with serum concentrations decreasing 30 times faster than brain concentrations during the early phase. Consequently, the brain-to-serum PCP concentration ratio increased from 8:1 at 30 s to 14:1 at 20 min before equilibrating at a ratio of 3:1 that remained constant from 1 to 8 h. In contrast, the testis-to-serum ratio increased slowly from 1:1 to 12:1 over 4 h, and then remained constant. In a separate group of animals, an s.c. infusion of PCP (18 mg/kg/day or 3.6 microg/min) produced a brain-to-serum ratio (6:1) that remained constant throughout the 96-h infusion. Testis-to-serum ratios increased from 4:1 at 1 h to 12:1 at 8 h and then remained constant for 96 h. Steady-state infusion of a pharmacologically inactive dose (2.5 mg/kg/day) produced a brain-to-serum ratio (3:1) that was significantly lower than the ratio (6:1) after infusion of the three pharmacologically active doses (10-25 mg/kg/day). The temporary high brain PCP concentrations and the dynamic disequilibrium between brain and serum concentrations after rapid i.v. administration could provide a better understanding of the preference of the human drug abuser for rapid rates (e.g., i.v. or smoking) of drug administration.

Anti-phencyclidine monoclonal antibodies provide long-term reductions in brain phencyclidine concentrations during chronic phencyclidine administration in rats

These studies examined the hypothesis that a single large dose of monoclonal anti-phencyclidine (PCP) antibody could provide long-term reductions in brain PCP concentrations despite continuous PCP administration. PCP (18 mg/kg/day, s.c.) was infused to steady-state (24 h) and then a mole-equivalent dose of a short-acting anti-PCP antigen-binding fragment (Fab) or a long-acting anti-PCP IgG was administered i.v. The PCP infusion continued for up to 27 days, even though the binding capacity of the single dose of antibody used should have been saturated within the first day. At selected time points after antibody administration, brain, testis, and serum PCP concentrations were measured. Serum PCP concentrations rapidly increased approximately 100- and 300-fold after Fab or IgG administration, respectively. Based on the antibody-bound PCP concentrations in serum, the functional elimination half-life (t(1/2lambdaZ)) values for PCP-Fab and PCP-IgG complexes were 9.4 h and 15.4 days, respectively. Fab and IgG administration produced a complete removal of PCP from the brain within 15 min. Although brain PCP concentrations were significantly decreased for only 4 h in Fab-treated animals, IgG administration resulted in significant decreases in brain PCP concentrations lasting for at least 27 days. In contrast, testis PCP concentrations were not substantially affected by antibody administration, suggesting that redistribution of PCP from the testis is too slow to benefit from a limited dose of antibody. These results indicate that anti-PCP IgG can preferentially protect the brain for approximately 4 weeks after IgG administration, even when the antibody binding capacity should have been saturated with continuously administered PCP.

Pharmacokinetic mechanisms for obtaining high renal coelimination of phencyclidine and a monoclonal antiphencyclidine antigen-binding fragment of immunoglobulin G in the rat

Our purpose was to determine mechanisms and methods for significantly increasing the renal coelimination of phencyclidine (PCP) and an anti-PCP monoclonal antibody binding fragment (anti-PCP Fab). To accomplish this goal, we performed a series of experiments to examine the dose-dependence of Fab elimination, mechanisms for enhancing PCP and Fab urinary coelimination and the antigenicity of repeated Fab administration. The results showed that urinary elimination of PCP and anti-PCP Fab was linear over a 30-fold range of doses. Anti-PCP Fab serum pharmacokinetics were best described using bi- or tri-exponential curves with a terminal elimination half-life of approximately 8 hr. Nevertheless, under all experimental conditions the early, nonterminal phase(s) were responsible for the majority (60%) of intact Fab elimination, with only 40% of the Fab eliminated during the terminal phase. These data suggest that the early rapid decline in Fab serum concentrations was primarily due to passive filtration and excretion of intact Fab, and not due to extravascular distribution as previously described. In comparison of methods for enhancing renal coelimination of Fab and PCP, systemic alkalinization produced a significant increase in Fab urinary elimination, with 69% of the Fab dose and 41% of the PCP dose recovered intact in the urine. Finally, in studies of the antigenicity of Fab, repeated administration of Fab produced no significant immune response or renal impairment. Overall, these experiments suggest that careful attention to the physiological status of the kidney during early time periods is essential for maximum coelimination of Fab and bound chemicals.

Pharmacodynamics of a monoclonal antiphencyclidine Fab with broad selectivity for phencyclidine-like drugs

The development of treatment strategies for drug intoxication has been hindered in part by the lack of clinically useful antagonists. Consequently, the major goal of these studies was to determine whether a monoclonal antibody Fab fragment (of IgG) could be used as an effective drug class-selective antagonist and to understand better the dose-response relationships for reversing CNS drug toxicity. Changes in drug-induced locomotor effects in a rat model were used to assess the ability of the antiphencyclidine (anti-PCP) Fab to reverse the behavioral effects of PCP and other potent arylcyclohexylamines. In experiments to determine the pharmacodynamics of Fabinduced antagonism of behavioral effects, the Fab completely reversed all PCP-induced locomotor effects in a Fab dose-dependent manner with a minimal effective dose of 0.18 mole-equivalents of Fab and an ED50 value of about one-third mole-equivalent. The anti-PCP Fab also completely reversed the locomotor effects induced by two other structurally related potent analogs of PCP: 1-[1-(2-thienyl)cyclohexyl]piperidine and N-ethyl-1-phenylcyclohexylamine. In addition, pharmacological and immunological selectivity was further tested by treatment of the behavioral effects induced by the structurally unrelated locomotor stimulant (+)methamphetamine. The antibody did not effectively reverse the effects of methamphetamine-induced locomotor activity. These results indicate that antibody-based medications can be developed to treat toxicity caused by classes of drugs as well as by individual drugs.

Nitric oxide generation by renal proximal tubules: role of nitric oxide in the cytotoxicity of lipid A

Lipid A, the biologically active component of lipopolysaccharide, stimulated nitric oxide (NO) production by isolated rat proximal tubules (as measured by NO2- release) in a time-dependent manner. At a concentration of 50 micrograms/ml, lipid A stimulated NO2- generation and guanosine 3',5'-cyclic phosphate (cGMP) production within 5 min. Both of these effects were blocked by NG-methyl-L-arginine (L-NMMA), an inhibitor of NO synthase or by 8-(N,N'-diethylamino)-octyl-3,4,5-trimethoxybenzoate (TMB-8), an inhibitor of intracellular Ca++ release. Because an increase in NO production may be cytotoxic, we examined the cytotoxic potential of lipid A. At 90 min, lipid A (50 micrograms/ml) produced significant lactate dehydrogenase release (42 +/- 5%) compared to control (25 +/- 5%; P < .05). Both L-NMMA (1 mM) and TMB-8 (100 microM) completely protected against lipid A-induced cytotoxicity. TMB-8 but not L-NMMA inhibited the rise intracellular Ca++ concentration ([Ca++]i) in isolated proximal tubules elicited by lipid A. L-NMMA but not TMB-8 inhibited proximal tubule soluble NO synthase activity. Thus, in the proximal tubule, lipid A stimulates a rise in [Ca++]i that in turn activates constitutive NO synthase. Furthermore, these events lead ultimately to NO-dependent cytotoxicity. Therefore, these findings suggest the potential for lipopolysaccharide to have a direct impact on proximal tubule physiology and renal function in vivo and support the potential therapeutic benefits of NO synthase inhibitors in the treatment of endotoxemia.

Effects of lipid A on calcium homeostasis in renal proximal tubules

It is clear that lipopolysaccharides (LPS) are responsible for the multiorgan failure often associated with endotoxemia. However, little is known of the direct effects of LPS on kidney cells. We examined the effects of lipid A, the biologically active component of LPS, on rat proximal tubule Ca++ homeostasis. Lipid A produced a rapid, transient, concentration-dependent rise in intracellular Ca++ concentration, [Ca++]i, as monitored by fura-2. At 50 micrograms/ml [Ca++]i rose to 138 +/- 12 nM (n = 4) above basal [Ca++]i levels. The response to lipid A was not significantly inhibited by chelating extracellular Ca++ with EGTA (5 mM). However, the rise in [Ca++]i was significantly inhibited by 8-(N,N-dimethylamino)octyl-3,4,5-trimethoxybenzoate hydrochloride) and thapsigargin (17 +/- 7 nM and 13 +/- 9 nM rise, respectively; P < .05). These data indicate that the rise in [Ca++]i induced by lipid A is due to release of intracellular stores, and not extracellular influx. We also examined the role of inositol 1,4,5-trisphosphate in the lipid A response. Lipid A caused a time-dependent increase in inositol 1,4,5-trisphosphate that paralleled the rise in [Ca++]i, suggesting the release in [Ca++]i is through an inositol 1,4,5-trisphosphate-mediated release of intracellular stores. The ability of lipid A to alter Ca++ homeostasis suggests a potential for LPS to directly alter proximal tubule physiology and renal function in vivo.