Testing paradigm for prediction of development-limiting barriers and human drug toxicity.

The financial investment grows exponentially as a new chemical entity advances through each stage of discovery and development. The opportunity exists for the modern toxicologist to significantly impact expenditures by the early prediction of potential toxicity/side effect barriers to development by aggressive evaluation of development-limiting liabilities early in drug discovery. Improved efficiency in pharmaceutical research and development lies both in leveraging "best in class" technology and integration with pharmacologic activities during hit-to-lead and early lead optimization stages. To meet this challenge, a discovery assay by stage (DABS) paradigm should be adopted. The DABS clearly delineates to discovery project teams the timing and type of assay required for advancement of compounds to each subsequent level of discovery and development. An integrative core pathology function unifying Drug Safety Evaluation, Molecular Technologies and Clinical Research groups that effectively spans all phases of drug discovery and development is encouraged to drive the DABS. The ultimate goal of such improved efficiency being the accurate prediction of toxicity and side effects that would occur in development before commitment of the large prerequisite resource. Good justification of this approach is that every reduction of development attrition by 10% results in an estimated increase in net present value by $100 million.

The species-dependent metabolism of efavirenz produces a nephrotoxic glutathione conjugate in rats.

Efavirenz, a potent nonnucleoside reverse transcriptase inhibitor widely prescribed for the treatment of HIV infection, produces renal tubular epithelial cell necrosis in rats but not in cynomolgus monkeys or humans. This species selectivity in nephrotoxicity could result from differences in the production or processing of reactive metabolites, or both. A detailed comparison of the metabolites produced by rats, monkeys, and humans revealed that rats produce a unique glutathione adduct. The mechanism of formation and role of this glutathione adduct in the renal toxicity were investigated using both chemical and biochemical probes. Efavirenz was labeled at the methine position on the cyclopropyl ring with the stable isotope deuterium, effectively reducing the formation of the cyclopropanol metabolite, an obligate precursor to the glutathione adduct. This substitution markedly reduced both the incidence and severity of nephrotoxicity as measured histologically. Further processing of this glutathione adduct was also important in producing the lesion and was demonstrated by inhibiting gamma-glutamyltranspeptidase with acivicin pretreatment (10 mg/kg, IV) prior to dosing with efavirenz. Again, both the incidence and severity of the nephrotoxicity were reduced, such that four of nine rats given acivicin were without detectable lesions. These studies provide compelling evidence that a species-specific formation of glutathione conjugate(s) from efavirenz is involved in producing nephrotoxicity in rats. Mechanisms are proposed for the formation of reactive metabolites that could be responsible for the renal toxicity observed in rats.

Goals, design, timing, and future opportunities for nonclinical drug metabolism studies.

The nonclinical ADME (adsorption, distribution, metabolism, and excretion) studies employed during drug development are dependent on the regulatory expectations and the changing development focus from nonclinical to clinical issues. These evolving issues necessitate that the development goals for ADME studies also change during the development process. The rationale for these goal changes and their impact on the timing and design of the ADME studies are discussed in the context of drug development at Glaxo Inc. The progress in the technology and knowledge in drug metabolism and in biology provide new opportunities for pharmaceutical companies in predicting drug toxicities relevant to humans. Two case examples are discussed to illustrate this opportunity.

Drug residue formation from ronidazole, a 5-nitroimidazole. VIII. Identification of the 2-methylene position as a site of protein alkylation.

Ronidazole protein-bound adducts were generated by the in vitro anaerobic incubation of [2-methylene-14C]ronidazole with microsomes from the livers of male rats. Acid hydrolysis of the protein adducts yielded an imidazole ring fragment bearing the radiolabel and an amino acid residue derived from the proteins. This fragment has been identified as carboxymethylcysteine by co-chromatography of the amino acid and its dansyl derivative with known standards under a variety of conditions. The carboxymethylcysteine was estimated to represent at least 15% of the radioactivity derived from the protein-bound adducts and provides unequivocal evidence that nucleophilic attack by protein cysteine thiols occurred at the 2-methylene position of ronidazole.

(-)-6-Aminocarbovir pharmacokinetics and relative carbovir exposure in rats.

The potential for the metabolic conversion of (-)-6-aminocarbovir to (-)-carbovir, a potent reverse transcriptase inhibitor effective against human immunodeficiency virus, has been examined in male Sprague-Dawley rats. Plasma (-)-6-aminocarbovir concentrations declined rapidly in a biphasic manner following an iv bolus dose of 20 mg/kg. The total systemic clearance was 5.4 liter/hr/kg and the terminal t1/2 was 0.35 hr. Following iv dosing, approximately half of the dose was excreted into the urine and comprised equivalent quantities of (-)-carbovir and (-)-6-aminocarbovir. Orally administered (-)-6-aminocarbovir was rapidly absorbed (tmax of 0.39 hr and Cmax of 4.96 micrograms/ml) following a 60 mg/kg dose. Following oral administration, 32% of the dose was eliminated in the urine, and comprised (-)-carbovir (75%) and (-)-6-aminocarbovir (25%). The oral bioavailability of (-)-6-aminocarbovir was 46% by plasma AUC comparison and 33% based on urinary excretion data. Exposure to (-)-carbovir was lower following (-)-6-aminocarbovir dosing than observed following (-)-Carbovir dosing, by both the oral and iv routes.

The pharmacokinetics of (-)-carbovir in rats. Evidence for nonlinear elimination.

This study evaluated the pharmacokinetics and bioavailability of (-)-carbovir in rats following iv and po administration at doses of 10 to 120 mg/kg. The systemic clearance decreased 5-fold as the iv dose was increased from 10 to 120 mg/kg. At lower doses, (-)-carbovir was eliminated primarily by the renal route. The renal component of clearance became saturated as the dose was increased, leading to nonlinearity in the pharmacokinetics of (-)-carbovir. There were no changes in the metabolic clearance or the formation clearance of the major metabolite over the dose range of 10 to 60 mg/kg. The clearance value estimated following iv bolus administration was a "concentration averaged" value, and was not predictive of the steady-state concentrations. IV infusion studies indicated that at plasma concentrations less than 2500 ng/ml, the pharmacokinetics of (-)-carbovir were linear. The bioavailability was calculated for each treatment level and ranged from 43% at 10 mg/kg (iv and po) to 3% at 120 mg/kg (iv and po). The nonlinearity in the pharmacokinetics of (-)-carbovir must be taken into account when determining the bioavailability. At doses lower than 10 mg/kg, where the serum concentrations after iv administration would always remain in the linear range, the bioavailability may approach 50 to 60%. Poor bioavailability at high doses in the rat may not be reflective of the clinical situation, since the anticipated doses will be much lower than those administered in the rat.

The metabolism and excretion of carbovir, a carbocyclic nucleoside, in the rat.

The metabolism and disposition of carbovir [(1R-cis)-2-amino 1,9-(4-(hydroxymethyl)-2-cyclopenten-1-yl)-6H-purin-6-one], an antiviral agent, was examined in rats using an isolated perfused liver, and in vivo following iv and po administration at two dosing levels. HPLC analysis of perfusate and bile after perfusion of the isolated liver with racemic (+/-)[8-3H]carbovir showed conversion to two major metabolites. The major component in the bile was shown to be a glucuronide conjugate of carbovir. The perfusate contained a single major component that was isolated and identified as the 4'-carboxylic acid derivative. In vivo excretion balance studies were conducted with [8-14C](-)-carbovir using four animals in each dosing group. Following iv administration at 10 mg/kg doses, the majority of the dose (77%) was excreted in the urine. At 60 mg/kg iv dosing, this value dropped to 42% (the remainder appearing in the feces). With po administration at both doses, the bulk of the dose (41-61%) was excreted in the feces. HPLC profiling of the urine showed that in all cases, most of the radioactivity was accounted for as carbovir and the 4'-acid metabolite. This metabolite accounted for up to 25% of the administered dose following 60 mg/kg iv administration, and less than 3% following 10 mg/kg po dosing. A second urinary metabolite accounting for up to 5% of the dose also was seen in all samples. This was isolated and identified as the trans-diastereomer of the 4'-acid (resulting from epimerization at the 4' position).(ABSTRACT TRUNCATED AT 250 WORDS)

The role of metabolism and covalent binding in the cytotoxicity of a nitrogen-containing steroid toward rat hepatocytes.

The nitrogen-containing steroid N,N-diethyl-4-methyl-3-oxo-4-aza-5 alpha-androst-1-ene-17 beta-carboxamide (I) causes concentration- and time-dependent cytotoxicity toward freshly isolated F-344 rat hepatocytes. Because hepatocytes extensively metabolize I to both stable and reactive products, the role of metabolism and covalent binding in the cytotoxicity of I was investigated. Concentration-dependent covalent binding of I-related material to hepatocyte macromolecules was detected. Treatment of rats with phenobarbital increased the rate and extent of hepatocyte metabolism of I, increased the covalent binding of I-related material to hepatocyte macromolecules, but decreased the cytotoxicity of I. Addition of testosterone to incubations of hepatocytes and I inhibited the metabolism of I, decreased the covalent binding of I-related material to hepatocyte macromolecules, but potentiated the cytotoxicity of I. These results indicate that the covalent binding of reactive metabolites is not involved in the cytotoxicity of I. Moreover, the two major metabolites of I, the 4-carbinolamide and monoethyl analog, were much less cytotoxic toward hepatocytes than I. These data suggest that the metabolism of I represents detoxication and that the parent compound is the cytotoxicant.

Metabolism of a nitrogen-containing steroid by rat hepatocytes and hepatic subcellular fractions.

Freshly isolated rat hepatocytes metabolized the nitrogen-containing steroid N,N-diethyl-4-methyl-3-oxo-4-aza-5 alpha-androst-1-ene-17 beta- carboxamide (I) to six products analyzed by reverse phase HPLC. The major metabolite, which could account for greater than 50% of the total I-related material, exhibited chromatographic, NMR, and mass spectral properties identical to those of the authentic 4-carbinolamide of I. Thus, the major biotransformation pathway in hepatocytes was hydroxylation of the N-methyl group of I to form a stable carbinolamide intermediate of N-demethylation. Desmethyl-I was observed as a minor metabolite. Another metabolite which accounted for approximately 10% of the total I-related material had chromatographic and spectral properties identical to those of the authentic monoethyl analog of I. The other three metabolites were formed in variable quantities and were unstable when isolated. Mass spectral data suggested that one metabolite was the carbinolamide intermediate of N-deethylation. Treatment of rats with phenobarbital or dexamethasone increased the formation of the monoethyl metabolite of I in hepatocytes but had no effect upon the formation of the 4-carbinolamide metabolite. Rat hepatic microsomes catalyzed the NADPH-dependent metabolism of I to the same metabolites in the same relative amounts as observed with intact hepatocytes. Studies with alternative substrates and inhibitors demonstrated that microsomal cytochrome P-450 was responsible for the metabolism of I. Dog and human hepatic microsomes metabolized I to the same products as rat hepatic microsomes, but monoethyl I was the major metabolite.

Covalent interaction of 5-nitroimidazoles with DNA and protein in vitro: mechanism of reductive activation.

Human hepatic microsomal enzymes catalyzed the NADPH-dependent anaerobic reductive activation of [1-14C]metronidazole [1-(2-hydroxyethyl)-2-methyl-5-nitroimidazole] and [4,5-14C]ronidazole [(1-methyl-5-nitroimidazole-2-yl)methyl carbamate] to species that became covalently bound to proteins. Due to the low efficiency of the enzyme-catalyzed covalent binding of metronidazole, the stoichiometry of anaerobic reductive activation was studied with dithionite as the reductant. Two moles of dithionite was consumed per mole of [1-14C]metronidazole for maximal covalent binding to either DNA or immobilized sulfhydryl groups, demonstrating that four electrons are required for the reductive activation of metronidazole. These data implicate the involvement of a hydroxylamine in covalent binding. Maximal covalent binding of [4,5-14C]ronidazole to DNA also required four-electron reduction, consistent with previous studies of the covalent binding of this agent to immobilized sulfhydryl groups [Kedderis et al. (1988) Arch. Biochem. Biophys. 262, 40-48]. Studies of the covalent binding of variously radiolabeled ronidazole molecules to DNA suggested that the imidazole ring was intact while greater than 80% of the 2-carbamoyl group and the C4 proton were not present in the DNA adduct. Studies of both the chemical and human hepatic microsomal reduction of [4-3H]metronidazole demonstrated that covalent binding occurred with the stoichiometric loss of this label, implicating binding at the C4 position. These results suggest that the reductive activation of 5-nitroimidazoles generally proceeds via four-electron reduction to form hydroxylamines followed by nucleophilic attack at C4.

Studies with nitrogen-containing steroids and freshly isolated rat hepatocytes: role of cytochrome P-450 in detoxication.

Several nitrogen-containing steroids produced concentration- and time-dependent decreases in the viability of freshly isolated F-344 rat hepatocytes. N,N-Diethyl-4-methyl-3-oxo-4-aza-5 alpha-androst-1-ene-17 beta-carboxamide (I) was not cytotoxic at or below 0.3 mM but produced decreases in cell viability at higher concentrations. In contrast, the desmethyl analog of I was essentially nontoxic, demonstrating that relatively small structural changes result in substantial differences in cytotoxicity. Testosterone and other steroids specifically potentiated the cytotoxicity of I in a concentration-dependent manner, while having no effect upon the toxicity of other chemical agents. Pargyline and methimazole had no effect upon the cytotoxicity of I, suggesting that monoamine oxidase and flavin-containing monooxygenase are not involved. The cytochrome P-450 inhibitors octylamine and metyrapone potentiated the cytotoxicity of I. Induction of cytochrome P-450 isozymes by phenobarbital and beta-naphthoflavone treatment protected the cells against the cytotoxicity of I, while acetone or dexamethasone treatment had no effect. The initial rates of hepatocyte metabolism of the six nitrogen-containing steroids investigated did not correlate with cytotoxicity. Dithiothreitol and other thiol compounds had no effect upon the cytotoxicity of I, suggesting that sulfhydryl oxidation is not involved. Galactosamine and sulfate-free media had no effect upon the cytotoxicity of I. These results suggest that cytochrome P-450 is involved in the detoxication of I by rat hepatocytes while conjugative metabolism does not play a significant role.

Mechanism of reductive activation of a 5-nitroimidazole by flavoproteins: model studies with dithionite.

The flavoprotein nitroreductases NADPH:cytochrome P-450 reductase and xanthine oxidase catalyzed the cofactor-dependent anaerobic nitro group reduction and covalent binding to protein sulfhydryl groups of the 5-nitroimidazole substrate ronidazole [1-methyl-5-nitroimidazole-2-yl)-methyl carbamate). Studies with variously radiolabeled ronidazole molecules demonstrated that the imidazole ring was intact while greater than 80% of the C-4 3H and 2-carbamoyl group were lost from the covalently bound product. The stoichiometry of cofactor consumption during the enzyme-catalyzed reduction of the substrate could not be determined, so a model nitroreductase system which utilized dithionite as the reductant and agarose-immobilized cysteine as the target for alkylation was developed. Two moles of dithionite was consumed per mole of substrate for maximal reduction of uv absorbance due to the nitro group, for maximal release of C-4 3H, and for maximal covalent binding to agarose-immobilized cysteine. These results indicate that four electrons are required for the reductive activation of the substrate, consistent with formation of a hydroxylamine reactive intermediate. Covalent binding of variously radiolabeled substrate molecules after dithionite reduction exhibited the same labeling pattern as flavoprotein-catalyzed covalent binding, suggesting that covalent binding is mediated by the same species in both chemical and biological systems. The data are consistent with a mechanism where the substrate undergoes four-electron reduction to form a hydroxylamine, which is susceptible to nucleophilic attack at C-4. When water attacks C-4, the 2-carbamoyl group can eliminate to form a Michael-like acceptor which adds thiols at the 2-methylene position.

Structural alterations that differentially affect the mutagenic and antitrichomonal activities of 5-nitroimidazoles.

Two approaches have been used to develop nonmutagenic 5-nitroimidazoles. Both approaches are based on knowledge of the likely mechanisms by which this class of compounds cause mutagenicity. The first approach involved incorporating readily oxidizable gallate derivatives into the molecule. In one case, a very weakly mutagenic active antitrichomonal agent was obtained. The second approach involved incorporating a substituent at the C4 position of the ring. This generally resulted in a large reduction in mutagenicity and a lowering of antitrichomonal activity in vitro. In certain cases, however, mutagenicity was dramatically reduced while moderate antitrichomonal activity was retained. For example, 1,2-dimethyl-4-(2-hydroxyethyl)-5-nitroimidazole (5) showed good antitrichomonal activity in vitro (ED50 = 2 micrograms/kg) while possessing only 4% of the mutagenicity of metronidazole.

Studies on the mechanism of activation and the mutagenicity of ronidazole, a 5-nitroimidazole.

Substantial evidence implicates the obligatory nucleophilic attack by water at C4 for the elimination of the carbamate and subsequent immobilization by electrophilic attack on protein thiols. Consequently, the strong correlation between the structural requirements for protein alkylation and for mutagenicity in TA100 suggests a possible role of nucleophilic addition at C4 or at the 2-methylene carbon for the expression of mutagenicity. Further studies directed at evaluating this possibility are currently in progress.