Integration of Physiologically-Based Pharmacokinetic Modeling into Early Clinical Development: An Investigation of the Pharmacokinetic Nonlinearity.

BMS-911543, a promising anticancer agent, exhibited time-dependent and dose-dependent nonlinear pharmacokinetics (PKs) in its first-in-human (FIH) study. Initial physiologically based pharmacokinetic (PBPK) modeling efforts using CYP1A2-mediated clearance kinetics were unsuccessful; however, further model analysis revealed that CYP1A2 time-dependent inhibition (TDI) and perhaps other factors could be keys to the nonlinearity. Subsequent experiments in human liver microsomes showed that the compound was a time-dependent inhibitor of CYP1A2 and were used to determine the enzyme inactivation parameter values. In addition, a rat tissue distribution study was conducted and human plasma samples were profiled to support the refinement of the PBPK model. It was concluded that the interplay between four BMS-911543 properties, namely, low solubility, saturation of the metabolizing enzyme CYP1A2, CYP1A2 TDI, and CYP1A2 induction likely resulted in the time-dependent and dose-dependent nonlinear PKs. The methodology of PBPK model-guided unmasking of compound properties can serve as a general practice for mechanistic understanding of a new compound's disposition.

In vitro characterization and pharmacokinetics of dapagliflozin (BMS-512148), a potent sodium-glucose cotransporter type II inhibitor, in animals and humans.

(2S,3R,4R,5S,6R)-2-(3-(4-Ethoxybenzyl)-4-chlorophenyl)-6-hydroxymethyl-tetrahydro-2H-pyran-3,4,5-triol (dapagliflozin; BMS-512148) is a potent sodium-glucose cotransporter type II inhibitor in animals and humans and is currently under development for the treatment of type 2 diabetes. The preclinical characterization of dapagliflozin, to allow compound selection and prediction of pharmacological and dispositional behavior in the clinic, involved Caco-2 cell permeability studies, cytochrome P450 (P450) inhibition and induction studies, P450 reaction phenotyping, metabolite identification in hepatocytes, and pharmacokinetics in rats, dogs, and monkeys. Dapagliflozin was found to have good permeability across Caco-2 cell membranes. It was found to be a substrate for P-glycoprotein (P-gp) but not a significant P-gp inhibitor. Dapagliflozin was not found to be an inhibitor or an inducer of human P450 enzymes. The in vitro metabolic profiles of dapagliflozin after incubation with hepatocytes from mice, rats, dogs, monkeys, and humans were qualitatively similar. Rat hepatocyte incubations showed the highest turnover, and dapagliflozin was most stable in human hepatocytes. Prominent in vitro metabolic pathways observed were glucuronidation, hydroxylation, and O-deethylation. Pharmacokinetic parameters for dapagliflozin in preclinical species revealed a compound with adequate oral exposure, clearance, and elimination half-life, consistent with the potential for single daily dosing in humans. The pharmacokinetics in humans after a single dose of 50 mg of [(14)C]dapagliflozin showed good exposure, low clearance, adequate half-life, and no metabolites with significant pharmacological activity or toxicological concern.

The use of radiolabeled compounds for ADME studies in discovery and exploratory development.

Radiolabeled compounds are excellent investigative tools and widely used to carry out ADME studies during drug discovery and development stages. The most commonly used radioisotopes are 14C and 3H. 3H materials are generally easier to synthesize than 14C materials. Therefore, a variety of probes and substrates used in in vitro assays are labeled with 3H. Since synthesis of 14C material requires intensive resources, it is usually not available until after a molecule is considered for potential development or after the molecule enters the development phase. Improvement in the technology in radiochemistry has enabled the use of radiolabeled compounds earlier in pre-clinical and clinical development to address mechanistic issues. For in vitro studies, radiolabeled probes are utilized to test affinity with various transporters, to perform metabolism comparison among species and to assess possible formation of reactive metabolites. For in vivo studies, radiolabeled compounds are employed to identify and elucidate metabolites formed, to investigate the extent of absorption, bioavailability, tissue distribution, mass balance, routes of excretion, and pre-systemic metabolism. Due to the significant impact of radiolabeled studies on drug development, these studies will be performed earlier than have been in the past and will continue to be an integral part of drug discovery and development.

Application of structure-metabolism relationships in the identification of a selective endothelin A antagonist, BMS-193884, with favourable pharmacokinetic properties.

1. Based on binding affinity, 2'-amino-N-(3,4-dimethyl-5-isoxazolyl)-4'-(2-methylpropyl)[1,1'-biphenyl]-2-sulfonamide (2) was identified as an initial lead in a programme to identify selective endothelin (ET) receptor antagonists. However, the compound was extensively metabolized in preclinical animal species and human in vitro systems due to oxidative biotransformation. 2. To optimize this structural class, the site of metabolism of 2 was determined. This allowed for focussed structure-activity and structure-metabolism studies aimed at finding more metabolically stable analogues that maintained potency. New analogues were screened for their ET binding characteristics and their stability in rat and human liver microsomes. 3. The use of the microsomal stability screen was tested by the determination of the pharmacokinetic parameters of select analogues. A good correlation was found between reduced rates of rat microsomal metabolism and reduced clearance in the rat. 4. N-(3,4-dimethyl-5-isoxazolyl)-4'-(2-oxazolyl)[1,1'-biphenyl]-2-sulfonamide (3) was identified as an analogue with improved in vitro properties and further studies revealed that the compound had improved pharmacokinetic properties. 5. N-[[2'-[[(3,4-dimethyl-5-isoxazolyl)amino]sulfonyl]-4-(2-oxazolyl)[1,1'-biphenyl]-2-yl]methyl]acetamide (4) was subsequently identified as a compound with superior in vitro properties compared with compound 3, but when tested in vivo it had a substantially increased rate of clearance. Further studies demonstrated that the clearance of this closely related structural analogue was not dictated by metabolic processes, but was mediated by transport-mediated direct biliary excretion. 6. The utility of screening for in vitro liver microsomal stability as part of the lead optimization process for compounds with metabolic liabilities was shown. It was also shown that relatively small molecular changes can dramatically change the disposition of closely related analogues and care must be used when screening for a single property.

Oxidative activation of acylguanidine prodrugs: intestinal presystemic activation in rats limits absorption and can be inhibited by co-administration of ketoconazole.

1. The disposition of acyl prodrugs was studied to improve the delivery of a guanidine-containing parent compound with poor membrane permeability and poor absorption. 2. The prodrugs were evaluated in vitro and in vivo for conversion to drug. Prodrugs were evaluated for hydrolytic or oxidative bioactivation in intestinal homogenate and rat liver S9 or microsomes. The disposition of the prodrugs in vivo was monitored in bile duct-cannulated rats. 3. Compounds with n-alkylacyl groups were efficiently bioactivated, but were hydrolysed before absorption. 4. Hydrolytic bioactivation could be blocked in vitro by branching in the alkyl chain. These compounds showed modest improvements in absorption, despite favourable permeability. Experiments with liver microsomes demonstrated efficient NADPH-dependent oxidative bioactivation, which was proposed to occur through a CYP-mediated side chain oxidation followed by cyclization and release of parent compound. Ketoconazole co-administration yielded approximately a twofold increase in absorption. 5. The hydrolytically stable prodrugs were successful in increasing absorption of parent drug and were efficiently bioactivated, but they did not yield increased systemic levels of drug.

Thio- and oxoflavopiridols, cyclin-dependent kinase 1-selective inhibitors: synthesis and biological effects.

Flavopiridol analogues, thio- and oxoflavopiridols which contain a sulfur (16) or oxygen (18) atom linker between a chromone ring and the hydrophobic side chain, are selective cyclin-dependent kinase 1 (CDK1) inhibitors with an IC(50) of 110 and 130 nM. These analogues were prepared from key intermediate 7 by substituting the ethyl sulfoxide. Enantio pure intermediate piperidone 10 was obtained from the racemic piperidone 8 via a very efficient "dynamic kinetic resolution" in 76% yield. Hydrophobic side chains such as chlorophenyl or tert-butyl produced potent CDK1 inhibitory activity, while hydrophilic side chains such as pyrimidine or aniline caused a severe reduction in CDK inhibitory activity. These analogues are competitive inhibitors with respect to ATP, and therefore activity was dependent upon the CDK subunit without being affected by the cyclin subunit or protein substrate. Thio- and oxoflavopiridols 16 and 18 are not only selective within the CDK family but also discriminated between unrelated serine/threonine and tyrosine protein kinases. CDK1 selective thio- and oxoflavopiridol analogues inhibit the colony-forming ability of multiple human tumor cell lines and possess a unique antiproliferative profile in comparison to flavopiridol.

Continuous blood withdrawal as a rapid screening method for determining clearance of oral bioavailability in rats.

PURPOSE: To develop a methodology for continuous blood withdrawal in rats suitable for drug discovery screening purposes and perform limited validation studies with a series of test compounds. METHODS: A reliable methodology for continuous blood withdrawal in rats was developed. The method is dependent on continuous heparin infusion during withdrawal and the minimization of constrictive, thrombogenic sites. Plasma drug concentrations from either intermittent sampling or continuous withdrawal experiments were determined with HPLC analysis. RESULTS: The continuous withdrawal method was successfully adapted to rats such that blood samples could be reliably collected over a 6-hr experiment. The clearance and oral bioavailability values for theophylline, atenolol, propranolol, warfarin, BMS-182874 and BMS-A were determined from continuous withdrawal or intermittent sampling experiments. The results from the two methods were comparable, with each compound reliably placed in the same low, medium or high category based on clearance or oral bioavailability characteristics. CONCLUSIONS: The continuous withdrawal method proved to be a viable alternative to the classic intermittent sampling technique. The method should prove useful in drug discovery screening, where the evaluation of large numbers of compounds for systemic clearance or oral bioavailability is often necessary.

Cardioselective antiischemic ATP-sensitive potassium channel (KATP) openers. 5. Identification of 4-(N-aryl)-substituted benzopyran derivatives with high selectivity.

This paper describes our studies aimed at the discovery of structurally distinct analogs of the cardioprotective KATP opener BMS-180448 (2) with improved selectivity for the ischemic myocardium. The starting compound 6, derived from the indole analog 4. showed good cardioprotective potency and excellent selectivity compared to 2 and the first-generation KATP opener cromakalim (1). The structure-activity studies indicate that increasing the size of the alkyl ester leads to diminished potency as does its replacement with a variety of other groups (nitrile, methyl sulfone). Replacement of the ethyl ester of 6 with an imidazole gave the best compound 3 (BMS-191095) of this series which maintains the potency and selectivity of its predecessor 6. The results described in this publication further support that there is no correlation between vasorelaxant and cardioprotective potencies of KATP openers. Compound 3 is over 20- and 4000-fold more selective for the ischemic myocardium than 2 and cromakalim (1), respectively. The selectivity for the ischemic myocardium is achieved by reduction of vasorelaxant potency rather than enhancement in antiischemic potency. As for cromakalim (1) and 2, the cardioprotective effects of compound 3 are inhibited by cotreatment with the KATP blocker glyburide, indicating that the KATP opening is involved in its mechanism of cardioprotection. With its good oral bioavailability (47%) and plasma elimination half-life (3 h) in rats, compound 3 offers an excellent candidate to investigate the role of residual vasorelaxant potency of 2 toward its cardioprotective activity in vivo.

Aminodiol HIV protease inhibitors. Synthesis and structure-activity relationships of P1/P1' compounds: correlation between lipophilicity and cytotoxicity.

A series of novel aminodiol inhibitors of HIV protease based on the lead compound 1 with structural modifications at P1' were synthesized in order to reduce the cytotoxicity of 1. We have observed a high degree of correlation between the lipophilicity and cytotoxicity of this series of inhibitors. It was found that appropriate substitution at the para position of the P1' phenyl group of 1 resulted in the identification of equipotent (both against the enzyme and in cell culture) compounds (10l, 10m, 10n, and 15c) which possess significantly decreased cytotoxicity.

Mechanisms of cytochrome P450 1A2-mediated formation of N-hydroxy arylamines and heterocyclic amines and their reaction with guanyl residues.

Amine oxidation reactions are catalyzed by cytochrome P450 (P450) and peroxidase enzymes; both types of enzymes appear to function via aminium radical intermediates. N-Dealkylation is favored over N-oxygenation for secondary and tertiary amines with both kinds of enzymes, but in the peroxidase-like enzymes N-oxygenation is even less favorable because of apparent restriction of the Fe-O complex in the active site. Among the rat liver P450s many of the carcinogenic primary arylamines and heterocyclic amines are N-oxygenated by P450 1A2 to form the N-hydroxy arylamine derivatives. Studies with human liver P450s also indicate that P450 1A2 plays a major role in such reactions, although some arylamines such as 4,4'-methylene-bis (3-chloroaniline) and dapsone are preferentially N-oxygenated by P450 3A4. Caffeine N3-demethylation has been developed as a useful marker of P450 1A2 levels in humans; the knowledge that P450 1A2 is the major phenacetin O-deethylase also allows insight into previous human interaction studies. 2-Ethynylnaphthalene is a useful mechanism-based inactivator of rat and rabbit P450 1A2 but not human P450 1A2 enzymes; the peptides labeled in the enzymes have been identified, along with the region in rat P450 1A2 that is modified with the photoaffinity label 4-azidobiphenyl. Microcrystals of rabbit P450 1A2 have been obtained as a first course to realizing the three-dimensional structures of these enzymes. Evidence is also presented that the major C8-guanyl DNA adducts resulting from these arylamines and heterocyclic amines in DNA may be formed via rearrangement of an initial N7-guanyl-2-arylamine adduct: reaction of N-acetoxy-2-aminofluorene with C8-methylguanine derivatives led to the formation of stable N7-substituted species, and reaction of N-acetoxy-2-aminofluorene with C8-bromoguanine yielded N-(C8-guanosinyl)-2-aminofluorene in a reaction best rationalized by such a mechanism.

Formation of 1,N2- and N2,3-ethenoguanine from 2-halooxiranes: isotopic labeling studies and isolation of a hemiaminal derivative of N2-(2-oxoethyl)guanine.

Vinyl halides are oxidized to 2-halooxiranes, which rapidly rearrange to 2-haloacetaldehydes. Both of these species can react with DNA to generate a variety of adducts, including the potentially mutagenic etheno (epsilon) products. Evidence was provided through kinetic studies that the epsilon-Gua adducts are formed primarily from 2-haloxiranes; consistent with this view, epoxide hydrolase inhibited the formation of N2,3-epsilon-Gua from vinyl chloride but alcohol dehydrogenase did not. Assignments of the NMR shifts of the etheno protons of 1,N2- and N2,3-epsilon-Gua were made with the use of 15N labeling and nuclear Overhauser effects, in revision of the literature. The H-5 proton of N2,3-epsilon-Gua showed facile exchange in acid or base; the H-7 proton of 1,N2-epsilon-Gua was exchanged at neutral or basic pH but not in acid. Reaction of Br2CHCH2OH (labeled at C1 with 2H or 13C) with Guo yielded 1,N2-epsilon-Gua and N2,3-epsilon-Gua, presumably through the intermediacy of 2-bromooxirane. 1H NMR analysis indicated that the labeled carbon was attached to the original Guo N2 atom in both cases. When N2-(2-oxoethyl)Gua was generated from a diethyl acetal or from a glycol, the major product was the cyclic derivative 5,6,7,9-tetrahydro-7-hydroxy-9-oxoimidazo[1,2-alpha]purine. This compound was also formed in considerable yield from the reaction of 2-chlorooxirane with Guo, dGuo 5'-phosphate, or DNA and is relatively stable in the presence of acid or mild base. It does not appear to be readily dehydrated to yield the etheno adducts but may be of significance as a DNA adducts in its own right.

Expression of mammalian glutathione S-transferase 5-5 in Salmonella typhimurium TA1535 leads to base-pair mutations upon exposure to dihalomethanes.

Dihalomethanes can produce liver tumors in mice but not in rats, and concern exists about the risk of these compounds to humans. Glutathione (GSH) conjugation of dihalomethanes has been considered to be a critical event in the bioactivation process, and risk assessment is based upon this premise; however, there is little experimental support for this view or information about the basis of genotoxicity. A plasmid vector containing rat GSH S-transferase 5-5 was transfected into the Salmonella typhimurium tester strain TA1535, which then produced active enzyme. The transfected bacteria produced base-pair revertants in the presence of ethylene dihalides or dihalomethanes, in the order CH2Br2 > CH2BrCl > CH2Cl2. However, revertants were not seen when cells were exposed to GSH, CH2Br2, and an amount of purified GSH S-transferase 5-5 (20-fold excess in amount of that expressed within the cells). HCHO, which is an end product of the reaction of GSH with dihalomethanes, also did not produce mutations. S-(1-Acetoxymethyl)GSH was prepared as an analog of the putative S-(1-halomethyl)GSH reactive intermediates. This analog did not produce revertants, consistent with the view that activation of dihalomethanes must occur within the bacteria to cause genetic damage, presenting a model to be considered in studies with mammalian cells. S-(1-Acetoxymethyl)GSH reacted with 2'-deoxyguanosine to yield a major adduct, identified as S-[1-(N2-deoxyguanosinyl)methyl]GSH. Demonstration of the activation of dihalomethanes by this mammalian GSH S-transferase theta class enzyme should be of use in evaluating the risk of these chemicals, particularly in light of reports of the polymorphic expression of a similar activity in humans.

Mechanism of C8 alkylation of guanine residues by activated arylamines: evidence for initial adduct formation at the N7 position.

Aromatic amines are bioactivated to electrophilic compounds that react with DNA, predominantly at the C8 position of guanine bases. This site is weakly nucleophilic and it has been proposed that the C8 adduct is the final product after initial N7-adduct formation. To consider this possibility, we reacted several C8-substituted guanine derivatives with N-acetoxy-2-aminofluorene, prepared in situ from 2-acetylsalicylic acid and N-hydroxy-2-aminofluorene. With C8,N9-dimethylguanine, an adduct was isolated in good yield that was consistent, by NMR and mass spectral characterization, with a structure involving carcinogen substitution at the N7 position of guanine and linked through the 2-aminofluorenyl nitrogen--N-(C8,N9- dimethylguanin-N7-yl)-2-aminofluorene. This adduct could be easily reduced with NaBH4, consistent with the proposed N7-adduct structure. The same reaction was also carried out with C8-methylguanosine and C8-methyldeoxyguanosine and similar adducts were isolated. In contrast, C8-bromoguanosine reacted with N-acetoxy-2-aminofluorene to yield the C8-substituted arylamine adduct N-(guanosin-C8-yl)-2-aminofluorene directly. These products are uniquely consistent with a scheme in which C8-adduct formation is preceded by an initial electrophilic substitution on the N7 atom, which is postulated to be a general reaction for activated arylamines and heterocyclic amines.

Mutation spectrum and sequence alkylation selectivity resulting from modification of bacteriophage M13mp18 DNA with S-(2-chloroethyl)glutathione. Evidence for a role of S-(2-N7-guanyl)ethyl)glutathione as a mutagenic lesion formed from ethylene dibromide.

The major DNA adduct (greater than 95% total) resulting from the bioactivation of ethylene dibromide by conjugation with GSH is S-(2-(N7-guanyl)ethyl)GSH. The mutagenic potential of this adduct has been uncertain, however, because the observed mutagenicity might be caused by other adducts present at much lower levels, e.g. S-(2-N1-adenyl)ethyl)GSH. To assess the formation of other potential adducts, S-(2-(N3-deoxycytidyl)ethyl)GSH, S-(2-(O6-deoxyguanosyl)ethyl)GSH, and S-(2-(N2-deoxyguanosyl)ethyl)GSH were prepared and used as standards in the analysis of calf thymus DNA modified by treatment with [1,2-14C]ethylene dibromide and GSH in the presence of rat liver cytosol; only minor amounts (less than 0.2%) were found. A forward mutation assay in (repair-deficient) Salmonella typhimurium TA100 and sequence analysis were utilized to determine the type, site, and frequency of mutations in a portion of the lacZ gene resulting from in vitro modification of bacteriophage M13mp18 DNA with S-(2-chloroethyl)GSH, an analog of the ethylene dibromide-GSH conjugate. An adduct level of approximately 8 nmol (mg DNA)-1 resulted in a 10-fold increase in mutation frequency relative to the spontaneous level. The spectrum of spontaneous mutations was quite varied, but the spectrum of S-(2-chloroethyl)GSH-induced mutations consisted primarily of base substitutions of which G:C to A:T transitions accounted for 75% (70% of the total mutations). All available evidence implicates S-(2-(N7-guanyl)ethyl)GSH as the cause of these mutations inasmuch as the levels of the minor adducts are not consistent with the mutation frequency observed in this system. The sequence selectivity of alkylation was determined by treatment of end-labeled lac DNA fragments with S-(2-chloroethyl)GSH, cleavage of the DNA at adduct sites, and electrophoretic analysis. Comparison of the sequence selectivity with the mutation spectrum revealed no obligate relationship between the extent of adduct formation and the number of mutations which resulted at different sites. We suggest that the mechanism of mutagenesis involves DNA sequence-dependent alterations in the interaction of the polymerase with the (modified) template and incoming nucleotide.

Preparation and characterization of oligonucleotides containing S-[2-(N7-guanyl)ethyl]glutathione.

S-[2-(N7-Guanyl)ethyl]glutathione is the major adduct derived from modification of DNA with 1,2-dibromoethane in biological systems and is postulated to be a mutagenic lesion [Humphreys, W. G., Kim, D.-H., Cmarik, J. L., Shimada, T., & Guengerich, F. P. (1990) Biochemistry 29, 10342-10350]. Oligonucleotides containing this modified base were prepared by treatment of oligonucleotides with S-(2-chloroethyl)glutathione and purified by chromatography. The self-complementary oligonucleotide d(ATGCAT), when thus modified at the single guanine, appeared to associate with itself as judged by UV measurements, but CD and NMR measurements indicated a lack of hybridization, with a decrease in the melting temperature of greater than 10 degrees C. The same lack of self-association was noted when d(ATGCAT) was modified to contain an N-acetyl-S-[2-(N7-guanyl)ethyl]cysteine methyl ester moiety. The oligomer d-(C1A2T3G4C5C6T7) was modified to contain a single S-[2-(N7-guanyl)ethyl]glutathione moiety at the central position, and UV, CD, and 1H NMR studies indicated that this oligomer hybridized to its normal complement d(A8G9G10C11A12T13G14), although the binding was considerably weakened by adduction (imino proton NMR spectroscopy in the presence of H2O indicated that the hydrogen bond signals seen in the oligomer were all broadened upon modification). All proton resonances were identified using two-dimensional 1H NMR spectroscopy. Adduct formation affected the chemical shifts of the base and 1', 2', and 2" protons of T3 and C5, the 2" proton of C6, and the 8 and 1' protons of C11, while little effect was observed on other protons. No cross-peaks were detected between the glutathione and oligomer moieties in two-dimensional nuclear Overhauser enhanced NMR studies. These results suggest that a rather local structural perturbation occurs in the DNA oligomer upon modification and that the glutathione moiety appears to be relatively unperturbed by its placement in the duplex. When the cytosine in the normal d(AGGCATG) complement to d-(CATGCCT) was changed to each of the other three potential bases at the central position, no hybridization with the oligomer d(CATGCCT) containing S-[2-(N7-guanyl)ethyl]glutathione was detected. We conclude that these N7-guanyl derivatives destabilize hybridization and that bases other than cytosine do not appear to show preferential thermodynamic bonding to these adducts, at least in the sequences examined to date.(ABSTRACT TRUNCATED AT 400 WORDS)