Microsoft word - crt_evans_metabolic activation_14oct03_final

Drug-Protein Adducts: An Industry Perspective on
Minimizing the Potential for Drug Bioactivation in Drug
Discovery and Development
David C. Evans†*, Alan P. Watt‡,
Deborah A. Nicoll-Griffith§ and Thomas A. Baillie††
†Drug Metabolism, Merck & Co., Inc., Rahway, NJ, USA.
‡Medicinal Chemistry, Drug Metabolism Section, Merck Sharp & Dohme, Terlings
Park, United Kingdom
§Medicinal Chemistry, Merck Frosst Center for Therapeutic Research, Kirkland,
Quebec, Canada
††Drug Metabolism, Merck & Co., Inc., West Point, PA, USA
David C. Evans, Ph.D.,Drug Metabolism,Merck & Co.,126 East Lincoln Avenue,Rahway, NJ 07065 Running title: Drug bioactivation and covalent binding to proteins It is generally accepted that there is neither a well defined nor consistent link between the formation of drug-protein adducts and organ toxicity. Since the potential does exist, however, for these processes to be causally related, the general strategy at Merck Research Laboratories has been to minimize reactive metabolite formation to the extent possible by appropriate structural modification during the lead optimization stage. This requires a flexible approach to defining bioactivation issues in a variety of metabolism vectors, and typically involves the initial use of small molecule trapping agents to define the potential for bioactivation. At some point, however, there is a requirement to synthesize a radiolabeled tracer and to undertake covalent binding studies in vitro, usually in liver microsomal (and sometimes hepatocyte) preparations from preclinical species and human, and also in vivo, typically in the rat. This Perspective article serves to provide one pragmatic approach to addressing the issue of bioactivation from an industry viewpoint based on protocols adopted by Merck Research Laboratories. The availability of a dedicated Labeled Compound Synthesis group, coupled to a close working relationship between Drug Metabolism and Medicinal Chemistry, represents a framework within which this perspective becomes viable; the overall aim being to bring I. Introduction
The concept that toxicities can stem from drug bioactivation in vivo continues to be problematic for pharmaceutical researchers, inasmuch as while the detection of a bioactivation process is relatively straightforward, the downstream consequences of this process remain indeterminable. This is due to the relatively little progress which has been made in understanding the molecular events which occur following drug haptenization of a protein. Since experimental tools are available to investigate drug bioactivation processes, and medicinal chemists are well versed in the art of optimizing structure-activity relationships, it is natural to assume that this latter skill set also can be exploited to prepare drugs which do not undergo metabolic activation. This goal has been the credo at Merck & Co., as one element of attempting to bring safe drugs to The concept that small organic molecules can undergo bioactivation in vivo, and that the resulting electrophiles can adduct to biological macromolecules and subsequently elicit organ toxicity, has its origins in studies performed during the early 1930’s to mid 1950’s. Such investigations included the work of Fieser (1) who investigated the hepatotoxicity of polycyclic hydrocarbons, and that of Miller and Miller (2, 3), who studied the hepatotoxic effects of p-dimethylaminoazobenzene in the rat and reported that aminoazo dyes become tightly bound to the protein constituents of liver tissues.
These early studies laid the foundation for the "Covalent Binding Theory" of xenobiotic induced liver and lung toxicity, which emerged during the 1970’s and 1980’s through a series of investigations performed at the National Institutes of Health by Brodie, Mitchell, Gillette and Boyd (4-8) who examined the bioactivation of a wide variety of small organic molecules and correlated this process with organ toxicity. The molecules investigated in these early studies, and in more recent work, include 4- ipomeanol (6-8), acetaminophen (9), halothane (10), vesnarinone (11), isoniazid (12, 13), furosemide (14), clozapine (15, 16), and bromobenzene (17-19). Many of these investigations shared the common experimental approach of using enzyme inducers and inhibitors to modulate the extent of bioactivation which, in turn, influenced the degree of covalent binding and tissue damage observed in vivo; the dogma postulated being that the processes of bioactivation, covalent binding and tissue damage were intrinsically linked.
The above dogma was challenged by several investigators during the 1980’s. For example, Tirmenstein and Nelson (20) reported that despite normalizing the dose level of acetaminophen (APAP) and its regioisomer, 3′-hydroxyacetanilide, to provide comparable levels of covalent binding in vivo in the mouse (~ 1.3 - 1.6 nmol drug equivalents/mg liver homogenate protein), APAP was hepatotoxic, but 3′- hydroxyacetanilide was not. From findings such as these, it was proposed that it was the adduction of "critical proteins", namely proteins required for specific cell function and viability, which was important for organ toxicity. While there are increasing numbers of reports of the identities of proteins which are modified by reactive drug metabolites (18, 21-23), few have attempted to distinguish "critical" from "non critical" proteins, or have investigated whether there are species differences in the types of proteins adducted. The absence of such information has essentially limited our ability to predict a priori whether bioactivation, and the accompanying covalent binding of drug-related material to proteins will, or will not, result in organ toxicity in preclinical species or human. Moreover, even if a drug candidate that is subject to metabolic activation were to fail to cause organ damage in a conventional preclinical safety assessment program, the concern remains that the drug-protein adducts formed in humans may have the potential to act as haptens and elicit an immune-mediated adverse event (24). This represents a significant concern for the pharmaceutical industry since it is generally acknowledged that no animal model exists for human immune-mediated toxicities. Thus, conventional preclinical safety studies may fail to identify those drug candidates which have the potential to elicit a hypersensitivity reaction in humans.
II. Covalent Binding Risk Assessment in Drug Discovery
In light of our inability to predict which reactive metabolites will be toxic, and which will not, the default position for Merck has been to minimize reactive intermediate formation to the extent possible by appropriate structure modification during the lead optimization stage. Initially, efforts are made to identify reactive intermediates by the use of chemical trapping agents, such as reduced glutathione (25) or cyanide (26, 27), to form stable adducts that are amenable to characterization by liquid chromatography- tandem mass spectrometry (LC-MS/MS) and/or NMR spectroscopy. These experiments, normally conducted in liver microsomal preparations, provide valuable indirect information on the identity of the original electrophile. Parallel efforts may involve the conduct of covalent binding studies, although these require the early availability of radiolabeled lead candidate compounds.
Regardless of the specific approach, there needs to be a readiness to identify the mechanism by which bioactivation occurs and a willingness to abrogate the potential liability through iterative structural modifications. Indeed, understanding the mechanism of metabolic activation is critically important if medicinal chemists are to minimize this process through rational changes in structure. It is these two phases, namely that of understanding the mechanism of bioactivation and applying chemical intervention to minimize bioactivation, which highlights the importance of a close working relationship between Drug Metabolism and Medicinal Chemistry from the point of project inception.
This proactive approach to addressing bioactivation is justified when considering the issue of patient safety and the economic risks in pursuing drugs which do covalently modify proteins. These risks are somewhat compounded by the typically low incidence of idiosyncratic reactions to drugs in the patient population (Table 1) which can result in safety issues being identified only after the patient population has become relatively large (Phase III clinical trials and beyond); a time when the economic commitment to drug development already has been considerable.
For those compounds which possess the requisite characteristics of a development candidate (high degree of efficacy and potency in animal models, acceptable pharmacokinetics in preclinical species, appropriate physico-chemical properties, etc), a radiolabeled analog of the compound (preferably 14C-labeled) is prepared and in vitro covalent binding experiments are performed in liver preparations from animals and humans. Hepatic microsomes are employed to assess metabolic activation catalyzed primarily by oxidative enzymes, while the use of freshly isolated hepatocytes may serve to reveal metabolic activation processes that depend upon the presence of a full complement of cellular enzyme systems. Covalent binding studies also are carried out in intact rats, where the level of adducts to both liver and plasma proteins is determined at appropriate intervals after administration of a standard oral dose of the test compound Having determined the level of covalent binding of a drug candidate, the question then becomes, “How much apparent covalent binding is acceptable in deciding whether to advance a drug candidate into development?” Merck’s approach to this question has been to take the levels of covalent adducts typically found in the livers of animals given a dose of a prototypic hepatotoxin (e.g. APAP, bromobenzene, furosemide or 4-ipomeanol) (Table 2) associated with the expression of hepatic necrosis (approximately 1 nmol drug equivalents/mg protein), and to reduce this figure by 20-fold to give a conservative target ‘threshold’ value for in vivo covalent binding of 50 pmol drug equivalents/mg total liver protein. This value also corresponds to a level of radioactivity that is approximately 10- times background under normal conditions, and thus provides a suitable dynamic range for measurement of covalently bound drug-protein adducts. It should be emphasized that the figure of 50 pmol drug equivalents/mg protein is viewed as a target upper limit, and not as an absolute threshold above which a compound would not be advanced into development, since it is recognized that other factors must be taken into consideration in arriving at this decision. In this regard, it is important to point out that several drugs which are known to undergo bioactivation and have associated clinical adverse events (AEs) are still marketed. This fact provides some initial insight into how we should view covalent binding data in the context of risk assessment. The AEs for such drugs include agranulocytosis (amodiaquine & clozapine), hepatotoxicity (diclofenac), hepatic failure (bicalutamide), and Stevens-Johnson syndrome (lamotrigine). Clearly, in using these agents, the risks have to be weighed against the benefits, and a number of factors taken into consideration. In the context of new drug discovery and development, these factors, which can be regarded as "qualifying considerations" for the purposes of this text (Figure 1), include the intended clinical use of the drug candidate (is it an unmet medical need or are current therapies inadequate?), the severity of the indication (is the prognosis disabling or life-threatening?), the dosing regimen (will the drug be used acutely, or chronically/prophylactically?), and the intended clinical population (will the drug be used for a pediatric indication where the tolerance for toxicity will be less?).
When assessing covalent binding data in liver microsomes in vitro, the degree to which drug clearance is dependent on Phase 2 enzymes also should be borne in mind.
For instance, a compound providing levels of covalent binding in liver microsomes of > 50 pmol drug equivalents/mg, but which is metabolized almost exclusively in hepatocyte preparations or in vivo by Phase 2 enzymes, where covalent binding is < 50 pmol drug equivalents/mg protein, would still be considered viable. Also worthy of consideration is the projected therapeutic dose. Uetrecht has reported (28) that drugs given at a daily dose of 10 mg or less are rarely, if ever, associated with a significant degree of idiosyncratic drug reactions. From the foregoing considerations, it is apparent that the decision process regarding the weight to place on covalent binding data is not digital but multifactorial.
Thus, the propensity for metabolic activation should be viewed as one of several potential liabilities which need to be taken into account when making programatic decisions on a As an example of the application of the above criteria, it is instructive to consider ethinylestradiol (EE), the usual estrogenic component of the oral contraceptive pill.
While the daily oral dose is very low (35 µg), hence satisfying at least one of the risk- benefit considerations, EE produces high levels of covalent binding in vitro (1.2 nmol drug equivalents/mg protein following incubation of 250 µM EE for 20 min in rat liver microsomes) (Table 2). Understanding routes of metabolism (the compound is eliminated predominantly by sulfation and glucuronidation, with the gut making a significant contribution to this process) (29, 30), and the mechanism of bioactivation (in the absence of cytosolic catechol O-methyl transferase, a reactive o-quinone will be formed), an informed decision can be made not to view Phase 1-mediated metabolic Another practical issue that needs to be taken into account, namely that of appropriate location of the radiolabel, is perhaps obvious but is critical if false negative results are to be avoided. This is best exemplified in the case of isoniazid where, had the primary metabolite acetylisoniazid been labeled in the pyridine moiety (as opposed to the acetyl group) for covalent binding studies, no radioactivity would have become associated irreversibly with liver proteins (Table 2). This example highlights the potential need to undertake covalent binding studies with more than one radiolabeled form of the compound-of-interest, often prompted by a better understanding of the metabolic disposition of the drug candidate.
The process of covalent binding risk assessment in drug discovery can be summarized as attempting to minimize reactive metabolite formation to the extent possible by appropriate structure modification during the lead optimization stage. Studies performed in vitro (e.g. liver microsomes) allow species differences in covalent binding to be explored and an insight gained into the mechanism of metabolic activation. Studies performed in vivo (typically in the rat) provide information on the levels of covalent binding which will occur during evaluation in safety studies, and also provide an opportunity to assess the degree of metabolic activation in tissues other than the liver.
The decision tree employed at Merck (Figure 1), along with the risk-benefit (qualifying) considerations, provide a rationale for compound evaluation; target covalent binding values of < 50 pmol drug equivalents/mg protein both in vitro and in vivo are desirable, although values in excess of this figure in vitro and/or in vivo may be acceptable under III. Standardization of Covalent Binding Methods
Since covalent binding data have the potential to influence the fate of a program lead compound, it is essential that these data become available at an early stage when there is time to optimize the lead structure or to explore alternative chemical templates.
If covalent binding data from different research sites are to be viewed consistently, it is imperative that data be derived under standardized conditions and that assay conditions be validated accordingly. At Merck, several research sites participate in drug discovery efforts, and therefore an inter-site validation of a covalent binding method, based on the use of rat and human liver microsomes, was undertaken.
In brief, the method employed involves the incubation of 10 µM drug substrate with rat or human liver microsomes (RLM, HLM; 1mg protein/ml) (± NADPH) for a 1 hr incubation period (total final volume of 200 µl), at which point the reaction is stopped and protein precipitated by the addition of acetone (800 µl). Precipitated protein is collected using a Brandel cell/membrane harvester and the protein washed with 80% aqueous methanol (1.5 liter per 96 well plate; 30 ml per filter). Individual filter discs are punched from the Brandel harvester mat, the protein solubilized, and aliquots taken for protein assay and determination of radioactive content by liquid scintillation counting.
In addition to evaluating test compounds, two control compounds also are routinely investigated. The first of these is used by all sites and allows the performance of the assay in different laboratories, and on different sites, to be monitored. The second control is program specific, and allows the Project Team to monitor the impact of Medicinal Chemistry intervention on improving the stability of subsequent lead compounds towards metabolic activation.
In the context of standardizing methodology for assessing the covalent binding of drugs to liver microsomal protein in vitro, compounds were selected on the basis of a well characterized bioactivation process. Knowledge of the enzyme system(s) responsible for metabolic activation also was desirable. The compounds chosen covered a wide range of covalent binding values (~50 to >2500 pmol drug equivalents/mg) and included diclofenac, imipramine, L-746530 and MRL-A (Figure 2). Four Merck sites were involved in this validation process wherein four compounds were radiolabeled and rat and human liver microsomes prepared on one site were shipped to the remaining three sites for evaluation in a standard covalent binding protocol. The percentage coefficient of variations (CVs) obtained for this assay generally were < 20%, although there were some exceptions where CVs were as high as 40% (Tables 3 & 4).
Covalent binding studies in pooled hepatocyte preparations using the Brandel tissue harvester method also have been undertaken (10 µM drug, 2 hr incubation period with 1 x 106 cells/ml, N = 3 preparations, Brandel extraction). Due to the inherent variability of covalent binding values originating from the use of individual preparations of freshly isolated hepatocytes, or even from different batches of pooled cryopreserved hepatocytes, data from these studies are not necessarily considered to be on the critical path for decision making purposes. Rather, these data are viewed in the broader context of better defining the metabolic disposition of the compound, the mechanism of bioactivation, and the level of covalent binding observed in all experimental systems, both in vitro and in vivo. For information, however, the mean values (with CVs in parenthesis) of covalent binding obtained for the compounds under investigation in pooled (N=5) cryopreserved human hepatocytes were: 6 (17%) pmol drug equivalents/mg for imipramine, 81 (8%) pmol drug equivalents/mg for MRL-A, 31 (24%) pmol drug equivalents/mg for diclofenac, and ~359 (15%) pmol drug equivalents/mg for L-746530.
Studies performed in vivo also have been standardized. Rats were dosed orally at 20 mg/kg and animals exsanguinated at 2, 6 and 24 hr (N=3 animals per time point), their livers removed, and the samples stored frozen until analysis. This in vivo assay may, in addition, be performed at a lower dose of 2 mg/kg to evaluate the impact of dose on the extent of metabolic activation. For guidance, approximately 20 µCi of 14C labeled compound or 40 µCi of 3H labeled compound typically is dosed to a 300 g rat. The dose levels of 2 and 20 mg/kg were chosen arbitrarily to reflect a typical “low” and “high” dose. It should be noted that the use of a tritium tracer carries with it the potential to underestimate levels of covalent binding, either through loss by exchange (generation of tritiated water) or through metabolism (a concern which also applies, albeit to a lesser extent, to a 14C tracer). Metabolism studies usually provide insights into the potential for such tracer loss, and investigators need to be prepared to undertake covalent binding and drug disposition studies with an alternative tracer, should the need arise.
It should be clear also from the foregoing discussion that our approach to assessing metabolic activation makes no effort to scale values of covalent binding determined in vitro (liver microsomes or hepatocytes) to those determined in vivo.
Rather, our experiments are designed to highlight the potential for metabolic activation in several metabolic systems so that a structure based resolution of this finding can be instigated. In this regard, our choice of relatively high drug concentrations, 10 µM for in vitro and up to 20 mg/kg for in vivo studies, reflect our desire to balance maximizing analytical sensitivity with standardizing protocols.
IV Characteristics of Compounds Investigated
Diclofenac typically is prescribed to patients at 50 mg tid and can undergo both CYP3A4- and CYP2C9-mediated bioactivation processes (31-33). Following incubation of 1 mM diclofenac with human liver microsomes, protein adduct formation was reported to be mediated by CYP3A4 (33) since the initial product of CYP3A4 catalysis, 5- hydroxydiclofenac, is subject to further oxidation to a reactive p-benzoquinoneimine intermediate. In support of this contention, an N-acetylcysteine adduct of 5- hydroxydiclofenac was identified in urine from human subjects dosed with diclofenac (34). On the other hand, Tang et al. (32) have reported that an isomeric p- benzoquinoneimine, derived from 4′-hydroxydiclofenac, is formed by CYP2C9 in vitro at more clinically relevant concentrations (< 50 µM), whereas the 5-hydroxylation of diclofenac, in agreement with Shen et al. (33), becomes a significant pathway in vitro at higher concentrations (> 100 µM). Indeed, CYP2C9 has been shown to mediate the formation of the putative quinoneimine which can be trapped with reduced glutathione to form 4′-hydroxy-3′-(glutathion-S-yl)diclofenac (32). Bougie et al. (35) have also shown that the product of CYP2C9-mediated oxidation and acyl glucuronidation, namely the 4′- hydroxydiclofenac acyl glucuronide metabolite, had an apparent role in diclofenac- Imipramine is used in the dosage range of 75 to 225 mg per day and has been associated with abnormalities in liver function which tend to occur during the second month of treatment (31, 36, 37). Imipramine 2-hydroxylation is a major route of metabolic clearance in humans (31), and is catalyzed almost exclusively by CYP2D6 (38). By analogy with rat (39), it is proposed that imipramine undergoes metabolic activation in humans to an arene oxide which covalently modifies proteins.
L-746530 undergoes CYP3A-mediated bioactivation (Section V). Following oxidation on the dioxobicyclo portion of the molecule at the methylene position alpha to the ring oxygen, an aldehyde is formed which can be trapped in vitro using semicarbazide. Alternatively, the aldehyde can form a Schiff base adduct to proteins in vitro resulting in levels of covalent binding of ~ 2 nmol drug equivalents/mg (Tables 3 & 4). The furan moiety of L-746530 also can undergo metabolic activation, as described in Section V for its close structural analog, L-739010.
The levels of covalent binding following incubation of 10 µM diclofenac with rat or human liver microsomes (~290 and ~57 pmol drug equivalents/mg) highlight a quantitative species differences, binding being higher in rat liver microsomes. Had diclofenac been a development candidate under the current Merck paradigm, efforts would have been made to explain this species difference in bioactivation, with the goal of more effectively assessing the potential risk to humans. The covalent binding values obtained for imipramine (~ 460 and ~130 pmol drug equivalents/mg in rat and human liver microsomes) would have warranted a similar discussion. The values for covalent binding of L-746530 and MRL-A (> 1500 pmol drug equivalents/mg) are sufficiently high that it is unlikely that these compounds would be advanced today as drug V. Biomarkers of Metabolic Activation
In some instances, reactive intermediates can form adducts with small molecule trapping agents. Characterization of these adducts using LC-MS/MS and NMR techniques can provide indirect information on the structure of the reactive species from which they are derived, thereby defining a potential bioactivation mechanism and hence a rationale on which to base a chemical intervention strategy. The use of trapping agents also provides a means by which a relatively large number of compounds can be evaluated rapidly, thereby allowing prioritization of compounds for radiolabeling prior to classical determination of covalent binding. A caveat to using a trapping agent is that there probably is no single small molecule which can serve as a universal surrogate for a complex protein macromolecule; nevertheless, several traps have found widespread utility and are discussed below. Also, certain reactive metabolites are so reactive that they are presumed to be unable to escape from their site of formation, or react rapidly with water of the medium, and therefore are not efficiently trapped. Products of hydrolysis can, however, yield information about their precursor reactive intermediates.
A good example is the trifluoroacetic acid metabolite of halothane which is excreted in the urine of patients receiving this agent and is formed, in part, by the hydrolysis of the reactive trifluoroacyl halide intermediate (31).
A. Glutathione
Reduced glutathione (GSH), an abundant physiological nucleophile by virtue of its cysteine sulfhydryl group, captures reactive xenobiotic metabolites to form S- substituted adducts. Once GSH stores are depleted, however, cellular proteins may be adducted, and it is believed that in certain cases cellular function is compromised and organ toxicity can ensue (40). Hence, GSH serves as a natural trapping agent for chemically reactive metabolites and has been used extensively in vitro to study a broad range of reactive intermediates including quinoneimines (acetaminophen), nitrenium ions (clozapine), arene oxides (carbamazepine), quinones (estrogens), imine methides (3- methylindole), and Michael acceptors (valproic acid metabolites), for which references Typically, in vitro experiments are conducted at 0.2 - 5 mM GSH concentration.
Glutathione adducts can be analyzed by LC-MS/MS using either the full scan mode to search for anticipated conjugates, or by constant neutral loss scanning for 129 Da (γ- glutamyl moiety) to detect all GSH-related species (Figure 3). However, it should be recognized that some GSH adducts are not stable and are subject to chemical degradation/rearrangement or enzymatic degradation (e.g., catalyzed by γ- glutamyltranspeptidase; SectionVI A), thereby escaping detection. In certain cases, N- acetyl cysteine (NAC) can be used in place of GSH as a nucleophilic trap, although this agent may be less efficient than GSH if the conjugation reaction in question is catalyzed (even in part) by glutathione-S-transferase enzymes.
B. Potassium Cyanide
The cyanide anion (CN-) is a “hard” nucleophile that can be used to trap certain electrophilic drug metabolites. An example is presented in Figure 4, in which 1 mM KCN (a mixture of CN and 13C15N at 1:1 ratio) was used as the trapping agent. The detection of cyano adducts by LC-MS was greatly facilitated by the presence of prominent isotopic “doublets” that differed in mass by 2 Da (mono-adducts) or 4 Da (bis- adducts); further, the MS/MS spectra of these adducts were characterized by a neutral loss of 27/29 Da (HCN/H13C15N). This technique has its origins in the work of Gorrod and co-workers (26, 27), and has been highlighted recently in a comprehensive review of bioactivation reactions of nitrogen-containing xenobiotics (41).
C. Methoxylamine and Semicarbazide
Both methoxylamine and semicarbazide can form a Schiff base with aldehydes, a process mimicking reactions between aldehyde metabolites with lysine residues on proteins. Typical conditions require the addition of 5 mM of either trapping agent to the incubation mixture followed by LC-MS/MS analysis. This approach was taken for both L-739010 and L-746530 (42, 43) and is presented in Figure 5.
VI. Bioactivation - ‘Structure Alerts’ Past and Present
Over the last few decades, considerable literature has accumulated on the subject of chemical induced toxicity, a representative selection of which is summarized in Table 5. The chemicals listed share the common feature of being able to form a reactive intermediate capable of alkylating proteins. Table 5 also contains two recent examples from Merck Research Laboratories (L-754394 and L-739010) where metabolic activation issues were identified and simple modifications made to the parent structures to produce drug candidates with markedly reduced bioactivation potential.
A. Piperazine Bioactivation
MRL-A, a 3-acyl-N1-methylpiperazine derivative (Figure 6), was found to undergo CYP3A4-mediated piperazine bioactivation (44). A mechanism for this process was proposed whereby formation of the end-product metabolites M1 and M2 involved a six electron oxidation of the piperazine ring, attack by GSH, and hydrolysis of the glutamic acid residue to afford a cysteinylglycine conjugate of the piperazinone. Attack (aminolysis) by the cysteinyl amine moiety resulted in opening of the piperazinone ring leading to a thiazolidine thioaminal intermediate which, in turn, underwent ring closure to the imidazoline products observed. The structures of adducts M1 and M2, elucidated by 1H-NMR and high resolution mass spectrometry, were consistent with this sequence of events. In keeping with the proposed mechanism, it was found that while MRL-A alkylated proteins extensively, simple alkyl substitutions (methyl, isopropyl) alpha to the N1-methyl functionality afforded derivatives which did not undergo metabolic activation and therefore were more attractive as drug candidates.
B. Pyrazinone Bioactivation
When a substituted pyrazinone derivative (MRL-B, Figure 7) was incubated with liver microsomal preparations from rats and humans fortified with GSH, two isomeric GSH conjugates (GSH-1 and GSH-2) were detected using LC-MS/MS (45). The benzylic alcohol, resulting from hydroxylation of MRL-B at the 6-methyl position, also was observed. The proposed origins of GSH-1 and GSH-2 pointed to two distinct, cytochrome P-450-mediated metabolic activation processes. Thus, it was proposed that MRL-B had undergone a two-electron oxidation, either directly or via dehydration of the hydroxymethyl metabolite, to generate an electrophilic imine-methide. Capture of this intermediate by GSH afforded adduct GSH-1. In the competing activation process, it was proposed that oxidative attack of the pyrazinone ring generated an unstable epoxide intermediate, which was attacked regiospecifically by GSH to afford a cyclic carbinolamide. Tautomeric ring-opening of the latter species to the acyclic methyl ketone, followed by syn/anti isomerization of this substituted amidine, would lead to the formation of a new ring system which, following elimination of the elements of water, would generate the stable imidazole derivative, GSH-2. Subsequent studies performed in the rat in vivo using radiolabeled MRL-B indicated that this compound gave rise to appreciable levels of covalent adducts to plasma and liver proteins. However, replacement of the 6-methyl group by chlorine, which prevented formation of the imine methide intermediate and rendered this heterocycle less susceptible to epoxidation, led to an analog whose potential for reactive metabolite formation and protein alkylation was C. Bioactivation of a Series of Aryloxy Derivatives
In one of our drug discovery programs at Merck, sites of metabolic activation within a new structural series were identified rapidly using online LC-MSn on an ion trap mass spectrometer (46). Following incubation of MRL-C (Table 6) with rat or human liver microsomes, reactive intermediates were trapped as their corresponding glutathione conjugates. Mass spectral characterization of these conjugates allowed potential sites and mechanisms of bioactivation for MRL-C to be proposed. These studies, coupled to informed structural modification of the series to block sites susceptible to metabolic activation, resulted in a significant reduction in the propensity of this compound class to label liver microsomal proteins in vitro (Table 6).
One strategy to reduce metabolic activation of MRL-C was to explore aromatic fluorine substitution as a means of blocking sites of oxidative metabolism, the aim being to take advantage of the electron-withdrawing effects of fluorine to reduce the π-electron density of the aromatic ring and hence render this moiety less susceptible to cytochrome P-450 catalyzed metabolism (47, 48). In addition, the carbon-fluorine bond is stronger than the corresponding carbon-hydrogen bond, and thus is less prone to oxidative cleavage. Unfortunately, fluorinated derivatives such as MRL-D (Table 6) all underwent extensive metabolic activation that was accompanied by oxidative dehalogenation and formation of GSH adducts. Therefore, an alternative strategy was pursued whereby a pyridine ring was incorporated as an isosteric replacement for the phenyl ring; this modification resulted in reduced levels of metabolic activation of MRL-E. Despite this improvement, MRL-E remained sub-optimal and provided levels of covalent binding well in excess of the target threshold value of 50 pmol drug equivalents/mg. Further improvements were made via chlorination (MRL-F) or trifluoromethylation (MRL-G) of the pyridine, the latter substitution resulting in a compound which failed to produce GSH adducts in either rat or human liver microsomal incubations. Furthermore, the levels of covalent binding to liver microsomal protein determined for MRL-G were significantly lower than those obtained in any other members of the series (Table 6). Further structural refinements in other parts of the molecule (data not shown) provided levels of covalent binding which were <50 pmol drug equivalents/mg in both rat and human liver microsomal preparations, while retaining desirable pharmacokinetic and VII. Concluding remarks
Despite 70 years of research, it is still not possible to accurately predict the potential for toxicity of a compound which has been shown to undergo metabolic activation. However, considerable advances have been made in the areas of analytical methodology (mass spectrometry and NMR spectroscopy) and labeled compound synthesis. As a result, it has become easier to identify reactive intermediates, by using either small molecule trapping agents or by undertaking covalent binding studies using radiolabeled compounds, and these approaches have now become a routine part of candidate drug evaluation at Merck Research Laboratories. Importantly, these investigations are undertaken at an early stage, during the discovery phase, when a mechanistic understanding of a bioactivation problem enables Medicinal Chemistry to provide a structure-based solution. Clearly there is a need to standardize methods for evaluating the potential for a drug candidate to under-go bioactivation in order to ensure consistency of data and uniformity in the decision making process. Future advances in the areas of immunology, genetics and proteomics may well provide a more coherent relationship between the characteristics of a drug-protein adduct and a toxicity outcome.
Indeed, an important problem to address in the field of toxicology has been the characterization of protein targets of reactive electrophiles; in this regard, the work of Badghisi and Liebler offers new LC-MS/MS / bioinformatic based approaches to mapping protein modifications at the level of amino acid sequence and should prove of value in such studies (49). Our current practice of stabilizing drug candidates against metabolic activation hopefully will contribute to bringing equally effective, but safer Acknowledgments
The authors would like to thank Dr. Maria Beconi, Mr. Stephen Day, Ms. Ann Mao, Ms. Rebecca White, and Dr. Timothy Schulz-Utermohl for their contributions to the global covalent binding assay development and validation project at Merck Research Laboratories. Our thanks go also to Drs. Ronald Franklin, Ralph Stearns, Wei Tang, Sanjeev Kumar, and Mr. Randy Miller, for valuable discussions during the preparation of this manuscript. For the preparation of radiolabeled compounds our thanks go to Drs.
David Melillo, Dennis Dean, Matthew Braun, Frank Tang, Chad Elmore, Eric Soli, Allen Jones, Daniel Dubé and Mr. Wensheng Liu.
References
Fieser, L. F. (1938) Carcinogenic activity, structure, and chemical reactivity of polynuclear aromatic hydrocarbons. Am. J. Cancer 34, 37-124.
Miller, E. C., and Miller, J. A. (1947) The presence and significance of bound aminoazo dyes in the livers of rats fed p-dimethylaminoazobenzene. Cancer Research 7, 468-480.
Miller, E. C., and Miller, J. A. (1952) In vivo combinations between carcinogens and tissue constituents and their possible role in carcinogenesis. Cancer Research 12, 547-556.
Brodie, B. B., Reid, W. D., Cho, A. K., Sipes, G., Krishna, G., and Gillette, J. R.
(1971) Possible mechanism of liver necrosis caused by aromatic organic compounds Proc. Natl. Acad. Sci. (USA) 68, 160-164.
Gillette, J. R., Mitchell, J. R., and Brodie, B. B. (1974) Biochemical mechanisms of drug toxicity. Annu. Rev. Pharmacol., 14, 271-288.
Boyd, M.R., and Burka, L.T. (1978) In vivo studies on the relationship between target organ alkylation and the pulmonary toxicity of a chemically reactive metabolite of 4-ipomeanol. J. Pharmacol. Exp. Ther. 207, 687-697.
Boyd, M.R., Burka, L.T., and Wilson, B.J. (1975) Distribution, excretion, and binding of radioactivity in the rat after intraperitoneal administration of the lung- toxin furan, 4-ipomeanol. Toxicol. Appl. Pharmacol. 32, 147-157.
Boyd, M.R., Sasame, H.A., and Franklin, R B. (1980) Comparison of ratios of covalent binding to total metabolism of the pulmonary toxin, 4-ipomeanol, in vitro in pulmonary and hepatic microsomes, and the effects of pretreatments with phenobarbital or 3-methylcholanthrene. Biochem. Biophys. Res. Commun. 93,
Hinson, J. A., Pumford, N. R., and Roberts, D. W. (1995) Mechanisms of acetaminophen toxicity: immunochemical detection of drug-protein adducts. Drug Metab. Reviews. 27, 73-92.
(10) Satoh, H., Fukuda, Y., Anderson, D. K., Ferrans, V. J., Gillette. J. R., and Pohl, L.
R. (1985) Immunological studies on the mechanism of halothane-induced hepatotoxicity: Immunohistochemical evidence of trifluoroacetylated hepatocytes.
J. Pharmacol. Exp. Ther. 233, 857-862.
(11) Uetrecht, J. P., Zahid, N., and Whitefield, D. (1994) Metabolism of vesnarinone by activated neutrophils: implications for vesnarinone-induced agranulocytosis. J. Pharmacol. Exp. Ther. 270, 865-872.
(12) Snodgrass, W. R., Potter, W. Z., Timbrell, D. J., and Mitchell, J. R. (1975) Possible mechanisms of isoniazid-related hepatic injury. Clin. Res. 22, 323A.
(13) Mitchell, J. R., Nelson, S. D., Snodgrass, and W. R., Timbrell, J. A. (1977) Metabolic activation of hydrazines to highly reactive hepatotoxic intermediates. In Biological Reactive Intermediates, pp. 271-277, Ed. Jollow, D., Kocsis, J., and (14) Mitchell, J. R., Nelson, W. L., Potter, W. Z., Sasame, H. A., and Jollow, D. J.
(1976) Metabolic activation of furosemide to a chemically reactive, hepatotoxic metabolite. J. Pharmacol. Exp. Ther. 199, 41-52.
(15) Gardner, I., Leeder, J. S., Chin, T., Zahid, N., and Uetrecht, J. P. (1998) A comparison of the covalent binding of clozapine and olanzapine to human neutrophils in vitro and in vivo. Mol. Pharmacol. 53, 999-1008.
(16) Williams, D. P., Pirmohamed, M., Naisbitt, D. J., Maggs, J. L., and Park, B. K.
(1997) Neutrophil cytotoxicity of the chemically reactive metabolite(s) of clozapine: possible role in agranulocytosis. J. Pharmacol. Exp. Ther. 283, 1375-
(17) Fisher, R., Brendel, K., and Hanzlik, R. P. (1993) Correlation of metabolism, covalent binding and toxicity for a series of bromobenzene derivatives using rat liver slices in vitro. Chem.-Biol. Interact. 88, 191-208.
(18) Koen, Y. M., and Hanzlik, R. P. (2002) Identification of seven proteins in the endoplasmic reticulum as targets for reactive metabolites of bromobenzene. Chem. Res. Toxicol. 15, 699-706.
(19) Weller, P. E., Narasimhan, N., Buben, J. A., and Hanzlik, R. P. (1988) In vitro metabolism and covalent binding among ortho-substituted bromobenzenes of varying hepatotoxicity. Drug Metab. Dispos. 16, 232-237.
(20) Tirmenstein, M. A., and Nelson, S. D. (1989) Subcellular binding and effects on calcium homeostasis produced by acetaminophen and a nonhepatotoxic regioisomer, 3′-hydroxyacetanilide, in mouse liver. J. Biol. Chem. 264, 9814-9819.
(21) N. R. Pumford and N. C. Halmes (1997) Protein targets of xenobiotic reactive intermediates. Annu. Rev. Pharmacol. Toxicol., 37, 91-117.
(22) Y. Qiu, L. Z. Benet and A. L. Burlingame (1998) Identification of the hepatic protein targets of reactive metabolites of acetaminophen in vivo in mice using two- dimensional gel electrophoresis and mass spectrometry, J. Biol. Chem., 273, 17940-
(23) S. D. Cohen, N. R. Pumford, E. A. Khairallah, K. Boekelheide, L. R. Pohl, H. R.
Amouzadeh and J. A. Hinson (1997) Selective protein covalent binding and target organ toxicity, Toxicol. Appl. Pharmacol., 143, 1-12, 1997.
(24) Naisbitt, D. J., Gordon, S. F., Pirmohamed, M., and Park, B. K. (2000) Immunological principles of adverse drug reactions - The initiation and propagation of immune responses elicited by drug treatement. Drug Safety, 23, 483-507.
(25) Baillie, T. A., and Davis, M. R. (1993) Mass spectrometry in the analysis of glutathione conjugates. Biol. Mass. Spectrom. 22, 319-325.
(26) Gorrod J. W., and Aislaitner G. (1994) The metabolism of alicyclic amines to reactive iminium ion intermediates. Eur. J. Drug Metab. Pharmacokinet. 19, 209-
(27) Gorrod J. W., Whittlesea C. M., and Lam S. P. (1991) Trapping of reactive intermediates by incorporation of 14C-sodium cyanide during microsomal oxidation.
Adv. Exp. Med. Biol. 283, 657-664.
(28) Uetrecht, J. (1999) New concepts in immunology relevant to idiosyncratic drug reactions: the "danger hypothesis" and innate immune system. Chem. Res. Toxicol. 12, 387-395.
(29) Back, D. J., Breckenridge, A. M., MacIver, M., Orme, M., Purba, H. S., Rowe, P.H., and Taylor, I. (1982) The gut wall metabolism of ethinylestradiol and its contribution to the presystemic metabolism of ethyinylestradiol in humans. Br. J. Clin. Pharmacol. 13, 325-330.
(30) Rogers, S. M., Back, D. J., and Orme, M. L. (1987) Intestinal metabolism of ethinyloestradiol and paracetamol in vitro: studies using Ussing chambers. Br. J. Clin. Pharmacol. 23, 727-734.
(31) Boobis, A. R., Burley, D., Davies, D. M., Davies, D.S., Harrison, P. I., Orme, M.
L’E., Park, B. K., and Goldberg, L. I. (1991) In Therapeutic Drugs, Ed. Dollery, (32) Tang, W., Stearns, R. A., Bandiera, S. M., Zhang, Y., Raab, C., Braun, M. P., Dean, D.C., Pang, J., Leung, K. H., Doss, G. A., Strauss, J. R., Kwei, G. Y., Rushmore, T.
H., Chiu, S. L., and Baillie, T. A. (1999) Studies on cytochrome P-450-mediated bioactivation of diclofenac in rats and in human hepatocytes: identification of glutathione conjugated metabolites. Drug Metab. Dispos. 27, 365-372.
(33) Shen, S., Marchick, M. R., Davis, M. R., Doss, G. A., and Pohl, L. R. (1999) Metabolic activation of diclofenac by human cytochrome P450 3A4: role of 5- hydroxydiclofenac. Chem. Res. Toxicol. 12, 214-222.
(34) Poon, G. K., Chen, Q., Teffera, Y., Ngui, J. S., Griffin, P. R., Braun, M. P., Doss, G. A., Freeden, C., Stearns, R. A., Evans, D. C., Baillie, T. A., and Tang, W.
(2001) Bioactivation of diclofenac via benzoquinoneimine intermediates- identification of urinary mercapturic acid derivatives in rats and humans. Drug Metab. Dispos. 29, 1608-1613.
(35) Bougie, D., Johnson, S. T., Weitekamp, L. A., and Aster, R. H. (1997) Sensitivity to a metabolite of diclofenac as a cause of acute immune hemolytic anemia. Blood 90, 407-413.
(36) Weaver, G. A., Pavlinac, D., and Davis, J. (1977) Hepatic sensitivity to imipramine. Am. J. Digest. Diseases 22, 551-553.
(37) Short, M. H., Burns, J. M., and Harris, M. E. (1968) Cholestatic jaundice during imipramine therapy. J. Am. Med. Assoc. 206, 1791-1792.
(38) Koyama, E., Chiba, K., Tani, M., and Ishizaki, T. (1997) Reappraisal of human CYP isoforms involved in imipramine N-demethylation and 2-hydroxylation: a study using microsomes from putative extensive and poor metabolizers of S- mephenytoin and eleven recombinant human CYPs. J. Pharmacol. Exp. Ther. 281,
(39) Masubuchi, Y., Igarashi, S., Suzuki, T., Horie, T., and Narimatsu, S. (1996) Imipramine-induced inactivation of a cytochrome P450 2D enzyme in rat liver microsomes in relation to covalent binding of its reactive intermediates. J. Pharmacol. Exp. Ther. 279, 724-731.
(40) Rinaldi R., Eliasson E., Swedmark S., and Morgenstern R. (2002) Reactive intermediates and the dynamics of glutathione transferases. Drug Metab. Dispos. 30, 1053-1058.
(41) Kalgutkar, A., Dalvie, D. K., O’Donnell, J. P., Taylor, T. J., and Sahakian D. C.
(2002) On the diversity of oxidative bioactivation reactions on nitrogen-containing xenobiotics. Current Drug Metab. 3, 379-424.
(42) Chauret, N., Nicoll-Griffith, D., Friesen, R., Li, C., Trimble, L., Dube, D., Fortin, R., Girard, Y., and Yergey, J. (1995) Microsomal metabolism of the 5- lipoxygenase inhibitors L-746530 and L-739010 to reactive intermediates that covalently bind to protein: the role of the 6,8-dioxobicyclo[3.2.1]octanyl moiety.
Drug Metab. Dispos. 23, 1325-1334.
(43) Zhang, K. E., Naue, J. A., Arison, B., and Vyas, K. P. (1996) Microsomal metabolism of the 5-lipoxygenase inhibitor L-739010: Evidence for furan bioactivation. Chem. Res. Toxicol. 9, 547-554.
(44) Doss, G. A., R. Miller, R. R., Zhang, Z., Teffera, Y., Evans, D. C., Baillie, T. A., Stearns, R. S. et al. (2003) - unpublished observations.
(45) Singh, R., Silva-Elipe, M. V., Pearson, P. G., Arison, B. H., Wong, B. K., White, R., Yu, X., Burgey, C. S., Lin, J. H., and Baillie, T. A. (2003) Metabolic activation of a pyrazinone-containing thrombin inhibitor. Evidence for novel biotransformation involving pyrazinone ring oxidation, rearrangement, and covalent binding to proteins. Chem. Res. Toxicol. 16, 198-207.
(46) Samuel, K., Yin, W., Stearns, R. A., Tang, Y. S., Chaudhary, A. G., Jewell, J. P., Lanza, T., Lin, L. S., Hagmann, W. K., Evans, D. C., and Kumar, S. (2003) Rapidly addressing the metabolic activation potential of new leads in drug discovery: A case study using ion trap mass spectrometry and tritium labeling techniques. J. Mass. Spectrom. 38, 211-221.
(47) Park, B. K., Kitteringham, N. R., and O’Neill, P. M. (2001) Metabolism of fluorine-containing drugs. Annu. Rev. Pharmacol. Toxicol. 41, 443-470.
(48) Park, B. K., and Kitteringham, N. R. (1994) Effects of fluorine substitution on drug metabolism: pharmacological and toxicological implications. Drug Metab. Rev. 26, 605-643.
(49) Badghisi, H., and Liebler, D. C. (2002) Sequence mapping of epoxide adducts in human hemoglobin with LC-Tandem MS and the Salsa algorithm. Chem. Res. Toxicol. 15, 799-805.
(50) Physicians Desk Reference (PDR) (2003) Amiodarone p.3384; bicalutamide p.656; lamotrigine p. 1559. 57th Edition, published by Thomson PDR at Montvale.
(51) Jewell, H., Maggs, J. L., Harrison, A. C., O’Neill, P. M., Ruscoe, J. E., and Park, B.
K. (1995) Role of hepatic metabolism in the bioactivation and detoxication of amodiaquine. Xenobiotica 25, 199-217.
(52) Bigby, M., Jick, S., Jick, H., and Arndt, K. (1986) Drug-induced cutaneous reactions. A report from the Boston Collaborative Drug Surveillance Program on 15,438 consecutive inpatients, 1975 to 1982. J. Amer. Med. Assoc. 256, 3358-3363.
(53) Askmark, H., and Wiholm, B. E. (1990) Epidemiology of adverse reactions to carbamazepine as seen in a spontaneous reporting system. Acta Neurol. Scand. 81,
(54) Alvir, J. M., Lieberman, J. A., Safferman, A. Z., Schwimmer, J. L., and Schaaf, J.
A. (1993) Clozapine-induced agranulocytosis. Incidence and risk factors in the United States. New Eng. J. Med. 329, 162-167.
(55) Alarcon-Segovia, D. Drug-induced lupus syndromes. (1969) Mayo Clinic Proc. 44, 664-681.
(56) Park, B. K., Pirmohamed, M., and Kitteringham, N. R. (1998) Role of drug disposition in drug hypersensitivity: a chemical, molecular, and clinical perspective.
Chem. Res. Toxicol. 11, 969-988.
(57) Weiss, M. E., and Adkinson, N. F. (1988) Immediate hypersensitivity reactions to penicillin and related antibiotics. Clinical Allergy 18, 515-540.
(58) Shear, N. H., and Spielberg, S. P. (1988) Anticonvulsant hypersensitivity syndrome. In vitro assessment of risk. J. Clin. Invest. 82, 1826-1832.
(59) Lee, S. L., and Chase, P. H. (1975) Drug-induced systemic lupus erythematosus: a critical review. Seminars in Arthritis and Rheumatism 5, 83-103.
(60) Mandell, G. L., and Sande, M. A (1985) Anti-microbial agents: Sulfonamides, trimethoprim-sulphamethoxazole and agents for urinary tract infections. In The Pharmacological Basis of Therapeutics. (Gilman A.G., Goodman, L.S., Rall T.W.
and Marud R., Eds) pp. 1095-1114, McGraw-Hill.
(61) Ono, K., Kurohara, K., Yoshihara, M., Shimamoto, Y., and Yamaguchi, M. (1991) Agranulocytosis caused by ticlopidine and its mechanism. Am. J. Hematol. 37,
(62) Harrison, A. C., Kitteringham, N. R., Clarke, J. B., and Park, B. K. (1992) The mechanism of bioactivation and antigen formation of amodiaquine in the rat.
Biochem. Pharmacol. 43, 1421-1430.
(63) Maggs, J. L., Grabowski, P. S., and Park, B. K. (1983) Drug Protein conjugates- III. Inhibition of the irreversible binding of ethyinylestradiol to rat liver microsomal protein by mixed-function oxidase inhibitors, ascorbic acid and thiol.
J. Steroid Biochem. 19, 1273-1278.
(64) Maggs, J. L., Grabowski, P. S., Rose, M. E., and Park, B. K. (1982) The biotransformation of 17α-ethynyl[3H]estradiol in the rat: irreversible binding and biliary metabolites. Xenobiotica 12, 657-668.
(65) Maggs, J. L., Grabowski, P. S., and Park, B. K. (1983) Drug-protein conjugates-II.
An investigation of the irreversible binding and metabolism of 17α-ethinyl estradiol in vivo. Biochem. Pharmacol. 32, 301-308.
(66) Nelson, S. D., Mitchell, J. R., Sodgrass, W. R., and Timbrell, J. A. (1978) Hepatotoxicity and metabolism of iproniazid and isopropylhydrazine. J. Pharmacol. Exp. Ther. 206, 574-585.
(67) Nelson, S. D., Mitchell, J. R., Timbrell, J. A., Sodgrass, W. R., and Corcoran, G. B.
(1976) Isoniazid and iproniazid: activation of metabolites to toxic intermediates in man and rat. Science 193, 901-903.
(68) Hinson, J. A., and Mays, J. B. (1986) Covalent binding of the phenacetin metabolite p-nitrosophenetole to protein. J. Pharmacol. Exp. Ther. 238, 106-112.
(69) Buhler, D. R., Unlu, F., Thakker, D. R., Slaga, T. J., Conney, A. H., Wood, A. W., Chang, R. L., Levin, W., Jerina, D. M. (1983) Effect of a 6-fluoro substituent on the metabolism and biological activity of benzo[a]pyrene. Cancer Research 43,
(70) Monks, T. J., Lau, S. S., Highet, R. J., and Gillette, J. R. (1985) Glutathione conjugates of 2-bromohydroquinone are nephrotoxic. Drug Metab. Dispos. 13,
(71) Lillibridge, J. H., Amore, B. M., Slattery, J. T., Kalhorn, T. F., Nelson, S. D., Finnell, R. H., and Bennett, G. D. (1996) Protein-reactive metabolites of carbamazepine in mouse liver microsomes. Drug Metab. Dispos. 24, 509-514.
(72) Munns, A. J., De Voss, J. J., Hooper, W. D., Dickinson, R. G., Gillam, E. M. J.
(1997) Bioactivation of phenytoin by human cytochrome P450: Characterization of the mechanism and targets of covalent adduct formation. Chem. Res. Toxicol. 10,
(73) Cuttle, L., Munns, A. J., Hogg, N. A., Scott, J. R., Hooper, W. D., Dickinson, R. G., and Gillam, E. M. J. (2000) Phenytoin metabolism by human cytochrome P450: Involvement of P4503A and 2C forms in secondary metabolism and drug-protein adduct formation. Drug Metab. Dispos. 28, 945-950.
(74) Buckpitt, A. R., Castagnoli, N., Nelson, S. D., Jones, A. D., Bahnson, L. S. (1987) Stereoselectivity of naphthalene epoxidation by mouse, rat, and hamster pulmonary, hepatic, and renal microsomal enzymes. Drug Metab. Dispos. 15, 491-498.
(75) Wilson, A. S., Davis, C. D., Williams, D. P., Buckpitt, A. R., Pirmohamed, M., and Park, B. K. (1997) Characterization of the toxic metabolite(s) of naphthalene.
Toxicology, 114, 233-242.
(76) Iverson, S. L., Shen, L., Anlar, N., and Bolton, J. L. (1996) Bioactivation of estrone and its catechol metabolites to quinoid-glutathione conjugates in rat liver microsomes. Chem. Res. Toxicol. 9, 492-499.
(77) Shen, L., Qiu, S., Chen, Y., Zhang, F., van Breemen, R. B., Nikolic, D., and Bolton, J. L. (1998) Alkylation of 2’-deoxynucleosides and DNA by the premarin metabolite 4-hydroxyequilenin semiquinone radical. Chem. Res. Toxicol. 11, 94-
(78) Fan, P. W., Zhang, F., and Bolton, J. L. (2000) 4-Hydroxylated metabolites of the antiestrogens tamoxifen and toremifene are metabolized to unusually stable quinone methides. Chem. Res. Toxicol.13, 45-52.
(79) Zhang, F., Fan, P.W., Liu, X., Shen, L., van Breeman, R. B., and Bolton, J. L.
(2000) Synthesis and reactivity of a potential carcinogenic metabolite of tamoxifen: 3,4-dihydroxytamoxifen-o-quinone. Chem. Res. Toxicol. 13, 53-62.
(80) Dehal, S. S., and Kupfer, D. (1999) Cytochrome P-450 3A and 2D6 catalyze ortho hydroxylation of 4-hydroxytamoxifen and 3-hydroxytamoxifen (Droloxifene) yielding tamoxifen catechol: Involvement of catechols in covalent binding to hepatic proteins. Drug Metab. Dispos. 27, 681-688.
(81) Chen, Q., Ngui, J. S., Doss, G. A., Wang, R. W., Cai, X., DiNinno, F. P., Blizzard, T. A., Hammond, M. L., Stearns, R. A., Evans, D. C., Baillie, T. A., and Tang, W.
(2002) Cytochrome P450 3A4-mediated bioactivation of Raloxifene: Irreversible enzyme inhibition and thiol adduct formation. Chem. Res. Toxicol., 15, 907-914.
(82) Orton, T. C., and Lowery, C. (1981) Practolol metabolism. IV. Irreversible binding of [14C]practolol metabolite(s) to mammalian liver microsomes. J. Pharmacol. Exp. Ther. 219, 207-212.
(83) Khojasteh-Bakht, S. C., Koenigs, L. L., Peter, R. M., Trager, W. F., and Nelson, S.
D. (1998) (R)-(+)-Menthofuran is a potent, mechanism-based inactivator of human liver cytochrome P450 2A6. Drug Metab. Dispos. 26, 701-704.
(84) Sahali-Sahly, Y., Balani, S. K., Lin, J. H., and Baillie, T. A. (1996) In vitro studies on the metabolic activation of the furanopyridine L-754394, a highly potent and selective mechanism-based inhibitor of cytochrome P450 3A4. Chem. Res. Toxicol. 9, 1007-1012.
(85) Bartolone, J. B., Beierschmitt, W. P., Birge, R. B., Hart, S. G. E., Wyand, S., Cohen, S. D., Khairallah, E. A. (1989) Selective acetaminophen metabolite binding to hepatic and extrahepatic proteins: An in vivo and in vitro analysis.
Toxicol. Appl. Pharmacol. 99, 240-249.
(86) Meyers, L. L., Beierschmitt, W. P., Khairallah, E. A., and Cohen, S. D. (1988) Acetaminophen-induced inhibition of hepatic mitochondrial respiration in mice.
Toxicol. Appl. Pharmacol. 93, 378-387.
(87) Kretz-Rommel, A., and Boelsterli, U.A. (1994) Mechanism of covalent adduct formation of diclofenac to rat hepatic microsomal proteins. Retention of the glucuronic acid moiety in the adduct. Drug Metab. Dispos. 22, 956-61.
(88) Kretz-Rommel, A., and Boelsterli, U.A. (1994) Selective protein adducts to membrane proteins in cultured rat hepatocytes exposed to diclofenac: radiochemical and immunochemical analysis. Mol. Pharmacol. 45, 237-244.
(89) Pumford, N. R., Myers, T. G., Davila, J. C., Highet, R. J., and Pohl, L. R. (1993) Immunochemical detection of liver protein adducts of the nonsteroidal antiinflammatory drug diclofenac. Chem. Res. Toxicol. 6, 147-150.
(90) Kretz-Rommel, A., and Boelsterli, U. A. (1993) Diclofenac covalent protein binding is dependent on acyl glucuronide formation and is inversely related to P450-mediated acute cell injury in cultured rat hepatocytes. Toxicol. Appl. Pharmacol. 120, 155-161.
(91) Kassahun, K., Pearson, P. G., Tang, W., McIntosh, I., Leung, K., Elmore, C., Dean, D., Wang, R., Doss, G., and Baillie, T. A. (2001) Studies on the metabolism of troglitazone to reactive intermediates in vitro and in vivo. Evidence for novel biotransformation pathways involving quinone methide formation and thiazolidinedione ring scission. Chem. Res. Toxicol. 14, 62-70.
(92) Tettey, J. N., Maggs, J. L., Rapeport, W. G., Pirmohamed, M., and Park, B. K.
(2001) Enzyme-induction dependent bioactivation of troglitazone and troglitazone quinone in vivo. Chem. Res. Toxicol. 14, 965-974.
(93) Yost, G. S. (1989) Mechanisms of 3-methylindole pneumotoxicity. Chem. Res. Toxicol. 2, 273-279.
(94) Ruangyuttikarn, W. A., Martin, L., and Yost, G.S. (1991) Metabolism of 3- methylindole in human tissues. Drug Metab. Dispos. 19, 977-984.
(95) Ruangyuttikarn, W., Skiles, G. L., Yost, G.S. (1992) Identification of a cysteinyl adduct of oxidized 3-methylindole from goat lung and human liver microsomal proteins. Chem. Res. Toxicol. 5, 713-719.
(96) Tang, W., and Abbott, F. S. (1996) Bioactivation of a toxic metabolite of valproic acid, (E)-2-propyl-2,4-pentadienoic acid, via glucuronidation. LC/MS/MS characterization of the GSH-glucuronide diconjugates. Chem. Res. Toxicol. 9, 517-
(97) Sadeque, A. J. M., Fisher, M. B., Korzekwa, K. R., Gonzalez, F. J., and Rettie, A.
E. (1997) Human CYP2C9 and CYP2A6 mediate formation of the hepatotoxin 4- ene-valproic acid. J. Pharmacol. Exp. Ther. 283, 698-703.
(98) Porubek, D. J., Grillo, M. P., and Baillie, T. A. (1989) The covalent binding to protein of valproic acid and its hepatotoxic metabolite, 2-n-propyl-4-pentenoic acid, in rats and in isolated rat hepatocytes. Drug Metab. Dispos. 17, 123-130.
(99) Baillie, T. A. (1988) Metabolic activation of valproic acid and drug-mediated hepatotoxicity. Role of the terminal olefin, 2-n-propyl-4-pentenoic acid. Chem. Res. Toxicol. 1, 195-199.
(100) Lasser, K. E., Allen, P. D., Woolhandler, S. J., Himmelstein, D. U., Wolfe, S. M., and Bor, D. H. (2002) Timing of new black box warnings and withdrawals for prescription medications. J. Amer. Med. Assoc. 287, 2215-2220.
(101) Mouelhi, M., Ruelius, H. W., Fenselau, C., and Dulik, D. M. (1987) Species- dependent enantioselective glucuronidation of three 2- arylpropionic acids.
Naproxen, ibuprofen, and benoxaprofen. Drug Metab. Dispos. 15, 767-772.
(102) Castillo, M., and Smith, P. C. (1995) Disposition and reactivity of ibuprofen and ibufenac acyl glucuronides in vivo in the rhesus monkey and in vitro with human serum albumin. Drug Metab. Dispos. 23, 566-572.
(103) Smith, P. C., Benet, L. Z., and McDonagh, A. F. (1990) Covalent binding of zomepirac glucuronide to proteins: evidence for a Schiff base mechanism. Drug Metab. Dispos. 18, 639-644.
(104) Smith, P. C., McDonagh, A. F., and Benet, L. Z. (1986) Irreversible binding of zomepirac to plasma protein in vitro and in vivo. J. Clin. Invest. 77, 934-939.
(105) Smith, P. C., and Benet, L. Z. (1986) Characterization of the isomeric esters of zomepirac glucuronide by proton NMR. Drug Metab. Dispos. 14, 503-505.
(106) Lecoeur, S., Bonierbale, E., Challine, D., Gautier, J.-C., Valadon, P., Dansette, P.
M., Catinot, R., Ballet, F., Mansuy, D., and Beaune, P. H. (1994) Specificity of in vitro covalent binding of tienilic acid metabolites to human liver microsomes in relationship to the type of hepatotoxicity: comparison with two directly hepatotoxic drugs. Chem. Res. Toxicol. 7, 434-442.
(107) Koenigs, L. L., Peter, R. M., Hunter, A. P., Haining, R. L., Rettie, A. E., Friedberg, T., Pritchard, M. P., Shou, M., Rushmore, T. H., and Trager, W. F. (1999) Electrospray ionization mass spectrometric analysis of intact cytochrome P450: Identification of tienilic acid adducts to P450 2C9. Biochemistry 38, 2312-2319.
Table 1 Incidence of clinical toxicity for drugs withdrawn from use or are still marketed Compound
Marketed / Withdrawn
Toxicity
Incidence
e.g.CefotaximeCefradineCefaclorCeftriaxoneCefuroximeCimetidine (trimethoprim-sulfamethoxazolecombination)Diazepam Table 3 Covalent binding of 10 µM imipramine, MRL-A, diclofenac, and L-746530following incubation with rat liver microsomes Covalent binding, pmol drug equiv/mg proteina
Intersite
Compound &
Study Day
[3H]Imipramine
[3H]MRL-A
[14C]Diclofenac
[14C]L-746530
a Values are means (N=3) with %CV in parenthesis.
Table 4 Covalent binding of 10 µM imipramine, MRL-A, diclofenac, and L-746530following incubation with human liver microsomes Covalent binding, pmol drug equiv/mg proteina
Intersite
Compound &
Study Day
[3H]Imipramine
[3H]MRL-A
[14C]Diclofenac
[14C]L-746530
a Values are means (N=3) with %CV in parenthesis.
Table 5 Examples of compounds with toxicities which are known to undergo metabolicactivation Compound
Toxicity
Reference
Aryl oxidation to either an epoxide or a quinone Drug induced hypersensitivity, teratogenic.
Lung, but covalent binding higher in liver and kidney.
Carcinogenicity (Breast, liver, endometrial, kidney) Jaundice accompanied by elevated liver enzymes Furan epoxidation, ring opening to yield an aldehyde Formation of an acyl glucuronide conjugate Hepatotoxicity. Withdrawn in 1998 – 6 deaths Hepatotoxic in humans following N-acetylation and Agranulocytosis - nitrenium ion implicated Irreversible binding of radioactivity to human and rat liver microsomal protein upon incubation with a series of tritiated analogs (MRL-C though MRL-G) at a10 µM concentration for 1 hr in the presence of an NADPH-regenerating system. Theirreversible binding was assessed both in the absence and presence of 5 mM GSH.
Covalent binding to liver microsomal protein a Parallel experiments demonstrated that covalent binding of radioactivity to microsomal proteinat time zero and in the absence of an NADPH-regenerating system was <5 pmol drugequivalents/mg protein incubation in all cases.
b For all compounds, the tritium label was incorporated at the same position within the functional Figure Legend
Decision tree for assessing the suitability of lead compounds for development based on metabolic activation considerations. (As indicated in the text, a liability with regard to metabolic activation is viewed as only one of a number of criteria in the selection process).
Compounds used for the inter-site evaluation of covalent binding in vitro Proposed mechanism of bioactivation of acetaminophen and tapping of the electrophilic intermediate by glutathione. The origin of the characteristic neutral loss of 129 Da upon collisional activation of the MH+ species of the glutathione conjugate in a tandem mass spectrometer is as indicated.
Use of cyanide to trap an iminium ion intermediate.
Bioactivation of L-739010 or L-746530 (Fig. 2) and trapping of an aldehyde intermediate using (A) semicarbazide, and (B) methoxylamine.
Novel bioactivation of a 3-acyl-N1-methyl piperazine derivative. The structures shown in brackets represent proposed intermediates which were Novel bioactivation of a substituted pyrazinone derivative (45). Structures shown in brackets represent proposed intermediates which were not In vitro: Human and Rat LM (10 µM, 1 hr) and/or In vivo: Rat (20 mg/kg P.O.; 2, 6, 24 hr, liver & plasma) Qualifying considerations
½ Chemical tractability of structural series? - What potential exists to modify structure?- Evidence of informed chemical intervention?- Has metabolic activation been minimized relative to the preceding compound(s)? ½ Is the prognosis disabling or life-threatening? ½ Is the anticipated clinical daily dose < 10 mg? ½ Are the metabolic clearance routes predominantly non-Phase 1? - Will the drug will be used chronically/prophylactically? ½ What is the intended target population? - Is the drug intended for a pediatric indication?

Source: http://chemdepo.com/reactive_intermediate_protein_binding_drug_design.pdf