12-OH-nevirapine sulfate, formed in the skin, is responsible for nevirapine-induced skin rash

Nevirapine (NVP) treatment is associated with a significant incidence of skin rash in humans, and it also causes a similar immune-mediated skin rash in Brown Norway (BN) rats. We have shown that the sulfate of a major oxidative metabolite, 12-OH-NVP, covalently binds in the skin. The fact that the sulfate metabolite is responsible for covalent binding in the skin does not prove that it is responsible for the rash. We used various inhibitors of sulfation to test whether this reactive sulfate is responsible for the skin rash. Salicylamide (SA), which depletes 3'-phosphoadenosine-5'-phosphosulfate (PAPS) in the liver, significantly decreased 12-OH-NVP sulfate in the blood, but it did not prevent covalent binding in the skin or the rash. Topical application of 1-phenyl-1-hexanol, a sulfotransferase inhibitor, prevented covalent binding in the skin as well as the rash, but only where it was applied. In vitro incubations of 12-OH-NVP with PAPS and cytosolic fractions from the skin of rats or from human skin also led to covalent binding that was inhibited by 1-phenyl-1-hexanol. Incubation of 12-OH-NVP with PAPS and sulfotransferase 1A1*1, a human isoform that is present in the skin, also led to covalent binding, and this binding was also inhibited by 1-phenyl-1-hexanol. We conclude that salicylamide did not deplete PAPS in the skin and was unable to prevent covalent binding or the rash, while topical 1-phenyl-1-hexanol inhibited sulfation of 12-OH-NVP in the skin and did prevent covalent binding and the rash. These results provide definitive evidence that 12-OH-NVP sulfate formed in skin is responsible for NVP-induced skin rashes. Sulfotransferase is one of the few metabolic enzymes with significant activity in the skin, and it may be responsible for the bioactivation of other drugs that cause skin rashes.

Involvement of T helper 17 cells in D-penicillamine-induced autoimmune disease in Brown Norway rats

Idiosyncratic drug reactions (IDRs) are poorly understood, but their clinical characteristics suggest that they are immune mediated. Penicillamine-induced autoimmunity in Brown Norway rats has been utilized as an animal model for mechanistic studies of one type of IDR because it closely mimics the autoimmune syndromes that it causes in humans. Our previous work suggested that it is T-cell mediated. It has been shown that T helper 17 (Th17) cells play a central role in many types of autoimmune diseases. This study was designed to test whether Th17 cells are involved in the pathogenesis of penicillamine-induced autoimmunity and to establish an overall serum cytokine/chemokine profile for this IDR. In total, 24 serum cytokines/chemokines were determined and revealed a dynamic process. In sick animals, interleukin (IL) 6 and transforming growth factor-β1, known to be driving forces of Th17 differentiation, were consistently increased at both early and late stages of penicillamine treatment; however, no significant changes in these cytokines were observed in animals that did not develop autoimmunity. IL-17, a characteristic cytokine produced by Th17 cells, was increased in sick animals at both the messenger RNA and serum protein level. In addition, serum concentrations of IL-22, another characteristic cytokine produced by Th17 cells, were found to be elevated. Furthermore, the percentage of IL-17-producing CD4 T cells was significantly increased but only in sick animals. These data strongly suggest that Th17 cells are involved in penicillamine-induced autoimmunity. Such data provide important mechanistic clues that may help to predict which drug candidates will cause a relatively high incidence of such autoimmune IDRs.

Nevirapine hypersensitivity

Treatment of HIV-1 infections with nevirapine is associated with skin and liver toxicity. These two organ toxicities range from mild to severe, in rare cases resulting in life-threatening liver failure or toxic epidermal necrolysis. The study of the mechanistic steps leading to nevirapine-induced skin rash has been facilitated by the discovery of an animal model in which nevirapine causes a skin rash in rats that closely mimics the rash reported in patients. The similarity in characteristics of the rash between humans and rats strongly suggests that the basic mechanism is the same in both. The rash is clearly immune-mediated in rats, and partial depletion of CD4(+) T cells, but not CD8(+) T cells, is protective. We have demonstrated that the rash is related to the 12-hydroxylation of nevirapine rather than to the parent drug. This is presumably because the 12-hydroxy metabolite can be converted to a reactive quinone methide in skin, but that remains to be demonstrated. Although the rash is clearly related to the 12-hydroxy metabolite rather than the parent drug, cells from rechallenged animals respond ex vivo to the parent drug by producing cytokines such as interferon-gamma with little response to the 12-hydroxy metabolite, even when the rash was induced by treatment with the metabolite rather than the parent drug. This indicates that the response of T cells in vitro cannot be used to determine what caused an immune response. We are now studying the detailed steps by which the 12-hydroxy metabolite induces an immune response and skin rash. This animal model provides a unique tool to study the mechanistic details of an idiosyncratic drug reaction; however, it is likely that there are significant differences in the mechanisms of different idiosyncratic drug reactions, and therefore the results of these studies cannot safely be generalized to all idiosyncratic drug reactions.

D-penicillamine-induced autoimmunity: relationship to macrophage activation

Idiosyncratic drug reactions represent a serious health problem, and they remain unpredictable largely due to our limited understanding of the mechanisms involved. Penicillamine-induced autoimmunity in Brown Norway (BN) rats represents one model of an idiosyncratic reaction, and this drug can also cause autoimmune reactions in humans. We previously demonstrated that penicillamine binds to aldehydes on the surface of macrophages. There is evidence that an imine bond formed by aldehyde groups on macrophages and amine groups on T cells is one type of interaction between these two cells that is involved in the induction of an immune response. We proposed that the binding of penicillamine with aldehyde groups on macrophages could lead to their activation and in some patients could lead to autoimmunity. In this study, the transcriptome profile of spleen macrophages 6 h after penicillamine treatment was used to detect effects of penicillamine on macrophages with a focus on 20 genes known to be macrophage activation biomarkers. One biological consequence of macrophage activation was investigated by determining mRNA levels for IL-15 and IL-1 beta which are crucial for NK cell activation, as well as levels of mRNA for selected cytokines in spleen NK cells. Up-regulation of the macrophage activating cytokines, IFN-gamma and GM-CSF, and down-regulation of IL-13 indicated activation of NK cells, which suggests a positive feedback loop between macrophages and NK cells. Furthermore, treatment of a murine macrophage cell line, RAW264.7, with penicillamine increased the production of TNF-alpha, IL-6, and IL-23, providing additional evidence that penicillamine activates macrophages. Hydralazine and isoniazid cause a lupus-like syndrome in humans and also bind to aldehyde groups. These drugs were also found to activate RAW264.7 macrophages. Together, these data support the hypothesis that drugs that bind irreversibly with aldehydes lead to macrophage activation, which in some patients can lead to an autoimmune syndrome.

A study of the specificity of lymphocytes in nevirapine-induced skin rash

Nevirapine treatment can cause a skin rash. We developed an animal model of this rash and determined that the 12-hydroxylation metabolic pathway is responsible for the rash, and treatment of animals with 12-OH-nevirapine also leads to a rash. In the present study, we investigated the specificity of lymphocytes in nevirapine-induced skin rash. Brown Norway rats were treated with nevirapine or 12-OH-nevirapine to induce a rash. Lymph nodes were removed, and the response of lymphocytes to nevirapine and its metabolites/analogs was determined by cytokine production (enzyme-linked immunosorbent assay, enzyme-linked immunosorbent spot assay, and Luminex) and proliferation (alamar blue assay). Subsets of lymphocytes were depleted to determine which cells were responsible for cytokine production. Lymphocytes from animals rechallenged with nevirapine proliferated to nevirapine, but not to 12-OH-nevirapine or 4-chloro-nevirapine. They also produced interferon-gamma (IFN-gamma) when exposed to nevirapine, significantly less when exposed to 4-chloro-nevirapine, and very little when exposed to 12-OH-nevirapine, even though oxidation to 12-OH-nevirapine is required to induce the rash. Moreover, the specificity of lymphocytes from 12-OH-nevirapine-treated rats was the same, i.e., responding to nevirapine more than to 12-OH-nevirapine, even though these animals had never been exposed to nevirapine. A Luminex immunoassay showed that a variety of other cytokines/chemokines were also produced by nevirapine-stimulated lymphocytes. CD4(+) cells were the major source of IFN-gamma. The specificity of lymphocytes in activation assays cannot be used to determine what initiated an immune response. This has significant implications for understanding the evolution of an immune response and the basis of the pharmacological interaction hypothesis.

Reactive metabolites and agranulocytosis

Central to most hypotheses of the mechanism of idiosyncratic drug-induced blood dyscrasias is the involvement of reactive metabolites. In view of the reactive nature of the majority of such metabolites, it is likely that they are formed by, or in close proximity to the blood cells affected. The major oxidative system of neutrophils generates hypochlorous acid. We have demonstrated that the drugs associated with the highest incidence of agranulocytosis are oxidized to reactive metabolites by hypochlorous acid and/or activated neutrophils. There are many mechanisms by which such reactive metabolites could induce agranulocytosis. In the case of aminopyrine-induced agranulocytosis, most cases appear to involve drug-dependent anti-neutrophil antibodies, and these are likely to be induced by cell membrane antigens modified by the reactive metabolite of aminopyrine. The target of agranulocytosis associated with many other drugs is usually neutrophil precursors and may involve cytotoxicity or a cell-mediated immune reaction induced by a reactive metabolite. In the case of aplastic anaemia, there is evidence in some cases for involvement of cytotoxic T cells, which could either be induced by metabolites generated by neutrophils, or more likely, by reactive metabolites generated by stem cells.

Pathways of carbamazepine bioactivation in vitro. III. The role of human cytochrome P450 enzymes in the formation of 2,3-dihydroxycarbamazepine

Conversion of the carbamazepine metabolite 3-hydroxycarbamazepine (3-OHCBZ) to the catechol 2,3-dihydroxycarbamazepine (2,3-diOHCBZ) followed by subsequent oxidation to a reactive o-quinone species has been proposed as a possible bioactivation pathway in the pathogenesis of carbamazepine-induced hypersensitivity. Initial in vitro phenotyping studies implicated CYP3A4 as a primary catalyst of 2,3-diOHCBZ formation: 2-hydroxylation of 3-OHCBZ correlated significantly (r(2) > or = 0.929, P < 0.001) with CYP3A4/5 activities in a panel of human liver microsomes (n = 14) and was markedly impaired by CYP3A inhibitors (>80%) but not by inhibitors of other cytochrome P450 enzymes (< or = 20%). However, in the presence of troleandomycin, the rate of 2,3-diOHCBZ formation correlated significantly with CYP2C19 activity (r(2) = 0.893, P < 0.001) in the panel of human liver microsomes. Studies with a panel of cDNA-expressed enzymes revealed that CYP2C19 and CYP3A4 were high (S50 = 30 microM) and low (S50 = 203 microM) affinity enzymes, respectively, for 2,3-diOHCBZ formation and suggested that CYP3A4, but not CYP2C19, might be inactivated by a metabolite formed from 3-OHCBZ. Subsequent experiments demonstrated that preincubation of 3-OHCBZ with human liver microsomes or recombinant CYP3A4 led to decreased CYP3A4 activity, which was both preincubation time- and concentration-dependent, but not inhibited by inclusion of glutathione or N-acetylcysteine. CYP3A4, CYP3A5, CYP3A7, CYP2C19, and CYP1A2 converted [14C]3-OHCBZ into protein-reactive metabolites, but CYP3A4 was the most catalytically active enzyme. The results of this study suggest that CYP3A4-dependent secondary oxidation of 3-OHCBZ represents a potential carbamazepine bioactivation pathway via formation of reactive metabolites capable of inactivating CYP3A4, potentially generating a neoantigen that may play a role in the etiology of carbamazepine-induced idiosyncratic toxicity.

Testing the hypothesis that vitamin C deficiency is a risk factor for clozapine-induced agranulocytosis using guinea pigs and ODS rats

The use of clozapine is limited by a relatively high incidence of drug-induced agranulocytosis. Clozapine is oxidized by bone marrow cells to a reactive nitrenium ion. Although many idiosyncratic drug reactions are immune-mediated, the fact that patients with a history of clozapine-induced agranulocytosis do not immediately develop agranulocytosis on rechallenge suggests that some other factor may be responsible for the idiosyncratic nature of this reaction. The reactive nitrenium ion is very rapidly reduced back to clozapine by vitamin C, and many schizophrenic patients are vitamin C deficient. We set out to test the hypothesis that vitamin C deficiency is a major risk factor for clozapine-induced agranulocytosis using a vitamin C deficient guinea pig model. Although the vitamin C deficient guinea pigs did not develop agranulocytosis, the amount of clozapine covalent binding in these animals was less than we had previously observed in samples from rats and humans. Therefore, we studied ODS rats that also cannot synthesize vitamin C. Vitamin C deficient ODS rats also did not develop agranulocytosis, and furthermore, although covalent binding in the bone marrow was greater than that in the guinea pig, it was not increased in the vitamin C deficient ODS rats relative to ODS rats that had adequate vitamin C in their diet. Therefore, it is very unlikely that vitamin C deficiency is a major risk factor for clozapine-induced agranulocytosis.

Possible Bioactivation Pathways of Lamotrigine

The anticonvulsant lamotrigine is associated with idiosyncratic drug reactions, especially skin rashes. Most idiosyncratic reactions are believed to be caused by reactive metabolites. Previous studies have found evidence that an arene oxide is formed in rats; however, when we incubated radiolabeled lamotrigine with rat liver microsomes virtually no covalent binding was detected, and the expected downstream phenolic metabolites are not observed in humans. Rare cases of agranulocytosis have been associated with lamotrigine therapy, and we found that lamotrigine is oxidized to two different N-chloro products by HOCl. The more reactive N-chloro metabolite forms an adduct with N-acetylhistidine, and covalent binding was observed when radiolabeled lamotrigine was incubated with myeloperoxidase/H(2)O(2)/Cl(-). Another lamotrigine metabolite is an N-oxide. If this N-oxide were sulfated, it might be sufficiently reactive to bind to protein. The synthetic N-sulfate reacted with N-acetylserine; however, no covalent binding was detected when the radiolabeled N-oxide was incubated with sulfotransferase. We also investigated the possibility that lamotrigine might be oxidized to a free radical by other peroxidases or oxidized by other enzymes such as prostaglandin H synthase or tyrosinase, but no evidence of oxidation was found, and lamotrigine did not cause any detectable increase in lipid peroxidation in vivo. In view of the virtual lack of covalent binding to hepatic microsomes and the lack of any other likely pathway leading to metabolic activation in the skin, it is possible that the parent drug rather than a reactive metabolite causes lamotrigine-induced skin rashes.

Study of the Sequence of Events Involved in Nevirapine-Induced Skin Rash in Brown Norway Rats

Nevirapine, used for the treatment of HIV infection, is associated with development of skin rash and liver toxicity. The mechanism of these idiosyncratic reactions is unknown. We have previously reported the discovery of a new animal model of nevirapine-induced skin rash in rats. When treated with nevirapine, Brown Norway rats developed red ears on about day 7 and skin rash on about day 21. On rechallenge, ears turn red within 24 h, and skin lesions develop by day 9. In the current study, we analyzed the time course of the sequence of events involved in the development of skin rash. Rats were treated with nevirapine for 7, 14, or 21 days or rechallenged with it for 0, 1, or 9 days. This treatment led to an increase in the total number of auricular lymph node T, B, and macrophage cells. There was also an increase in the activation/infiltration marker ICAM-1 and activation/antigen presentation marker MHC II in these cells compared with those from control rats. Immunohistochemistry analysis showed macrophage infiltration and ICAM-1 expression in the ears of treated rats as early as day 7 of treatment. Macrophage infiltration preceded T cell infiltration, which was not apparent until the onset of rash. Both MHC I and MHC II expression increased in the skin of nevirapine-treated rats that developed rash. A major inducer of MHC is IFNgamma. Although rechallenge with nevirapine led to a large increase in serum levels of IFNgamma, this was not observed during the treatment of naïve rats with nevirapine. These observations provide further clues to the mechanism of nevirapine-induced skin rash.

Evidence of an immune-mediated mechanism for an idiosyncratic nevirapine-induced reaction in the female Brown Norway rat

Previously, we reported a new animal model of an idiosyncratic drug reaction in which nevirapine causes a skin rash in some rats that has characteristics similar to the reaction that occurs in humans. Strong evidence that the reaction is immune-mediated was found; specifically, low-dose pretreatment induced tolerance, while with rechallenge, the time to onset decreased and the severity increased. Furthermore, splenocytes from rechallenged rats transferred rash susceptibility to naïve recipients. We now report the results of studies to explore the immune aspects of this reaction. T cells were found to play an important role, as demonstrated by their ability to adoptively transfer susceptibility to the skin reaction. Of these T cells, CD4+ cells are the likely effectors because they were capable of transferring susceptibility and the reaction was delayed in rats partially depleted of CD4+ T cells. In contrast, it appears that CD8+ T cells are not essential, as CD8+ T cells were unable to transfer sensitivity to a naïve animal and rats depleted of CD8+ T cells still developed skin rash. Unlike the penicillamine model, where we have demonstrated that the tolerance induced by low-dose treatment is immune-mediated, tolerance induced by low-dose nevirapine appears to be largely due to induction of metabolism as it can be overcome by inhibition of cytochrome P450. Pretreatment with the immunosuppressants, cyclosporine and tacrolimus, prevented the rash and even led to resolution of the rash during nevirapine treatment. These studies reinforce the hypothesis that the reaction in this model is similar to that which occurs in humans. In particular, the finding that CD4+ T cells may play a central role in this model fits with the observation that the incidence of idiosyncratic reactions to nevirapine in humans appears to be lower in patients with low CD4+ counts.

In vitro and animal models of drug-induced blood dyscrasias

Drug-induced blood dyscrasias can be either acute and predictable or delayed and unpredictable (idiosyncratic). The predictable toxicity is relatively easy to reproduce with in vitro models, although they may not work for drugs that require bioactivation. It is very unlikely that idiosyncratic blood dyscrasias can be modeled in vitro, although some drugs (or their reactive metabolites) that cause idiosyncratic reaction are toxic to bone marrow cells in vitro. Although the mechanisms of idiosyncratic reactions are poorly understood, there is evidence that most are due to reactive metabolites and some are immune-mediated. Therefore screening drugs for their bioactivation by myeloperoxidase, the major oxidative enzyme in bone marrow, may provide some measure of the risk that a drug will cause blood dyscrasias. Several examples of drug-induced idiosyncratic agranulocytosis, aplastic anemia and thrombocytopenia are presented, but better in vivo models are clearly needed to gain a clearer understanding of these adverse reactions.

A comparison of the covalent binding of clozapine, procainamide, and vesnarinone to human neutrophils in vitro and rat tissues in vitro and in vivo

Covalent binding of drug reactive metabolites to neutrophils or their precursors is thought to play a role in the development of drug-induced agranulocytosis. In this study, we used immunochemical techniques to compare the covalent binding of clozapine, vesnarinone, and procainamide (three drugs associated with agranulocytosis) to phorbol-12,13-myristate acetate (PMA)-activated human neutrophils in vitro and rat tissues in vivo. In PMA-activated human neutrophils in vitro, clozapine and procainamide modified neutrophil proteins with molecular masses ranging from 30 to 200 kDa, while vesnarinone predominately formed adducts with molecular masses greater than 70 kDa. All three drugs formed adducts at 126, 98, and 58 kDa, and they all covalently bound to human myeloperoxidase when incubated with this enzyme and H2O2 in vitro. Covalent binding to PMA-activated neutrophils was inhibited by nucleophiles, such as glutathione and N-acetylcysteine, but not by N-acetyllysine. In the presence of the PMA, all three drugs covalently bound to activated rat bone marrow cells in vitro, while in its absence only clozapine did. Covalently modified liver proteins were observed in rats treated for 6 weeks with clozapine (25 or 50 mg/kg/day), vesnarinone (300 mg/kg/day), or procainamide (50 mg/kg/day). Clozapine extensively modified proteins in all subcellular fractions; procainamide formed a 99 kDa adduct in a membrane-containing fraction and 57, 47, and 36 kDa adducts in a cytosolic fraction, while vesnarinone formed liver-protein adducts with molecular masses of 82, 62, 49, and 40 kDa in membrane, cytosolic, and S9 fractions. In addition, clozapine and procainamide, but not vesnarinone, formed a 49 kDa drug-protein adduct in the bone marrow of treated rats. Furthermore, procainamide covalently bound to a 58 kDa protein in neutrophils of a patient treated with the drug. We suspect that covalent modification of common targets in the neutrophils by these three drugs plays a role in the development of drug-induced agranulocytosis.

Pathways of carbamazepine bioactivation in vitro: II. The role of human cytochrome P450 enzymes in the formation of 2-hydroxyiminostilbene

Conversion of the carbamazepine metabolite, 2-hydroxycarbamazepine, to the potentially reactive species, carbamazepine iminoquinone (CBZ-IQ), has been proposed as a possible bioactivation pathway in the pathogenesis of carbamazepine-induced hypersensitivity. Generation of CBZ-IQ has been proposed to proceed through the intermediate, 2-hydroxyiminostilbene (2-OHIS); however, data suggested that 2-hydroxycarbamazepine is oxidized by cytochromes P450 (P450s) directly to CBZ-IQ, followed by NADPH-mediated reduction to 2-OHIS. In vitro studies were conducted to identify the P450s responsible for converting 2-hydroxycarbamazepine to 2-OHIS and to determine functional consequences of this bioactivation pathway. Formation of 2-OHIS in human liver microsomes (HLMs) was consistent with monophasic, Michaelis-Menten kinetics. The sample-to-sample variation in the rate of 2-OHIS formation correlated significantly (r2 > or = 0.706) with CYP3A4/5 and CYP2B6 activities in a panel of HLMs (n = 10). Studies with a panel of cDNA-expressed enzymes revealed that CYP3A4 preferentially catalyzed 2-OHIS formation; CYP3A4 formed 2-OHIS at a rate >10 times that of other enzymes capable of forming 2-OHIS (CYP1A1, CYP2C19, and CYP3A7). Inhibitors of CYP3A enzymes markedly impaired 2-OHIS formation in HLMs, whereas inhibitors of other P450s resulted in < or = 20% inhibition. Although CYP3A4 was primarily responsible for converting 2-hydroxycarbamazepine to 2-OHIS, neither 2-hydroxycarbamazepine, 2-OHIS, nor CBZ-IQ caused time-dependent inactivation of CYP3A activity. No thiol adducts were formed directly from 2-hydroxycarbamazepine. However, glutathione- and N-acetylcysteine-conjugates were formed with 2-OHIS or CBZ-IQ as substrates. Thus, CYP3A4-dependent secondary oxidation of 2-hydroxycarbamazepine represents a potential carbamazepine bioactivation pathway leading to the formation of thiol-reactive metabolites, intermediates that may play a role in the etiology of idiosyncratic toxicity attributed to carbamazepine.

Gene Expression Profiling in a Model of d-Penicillamine-Induced Autoimmunity in the Brown Norway Rat: Predictive Value of Early Signs of Danger

Idiosyncratic drug reactions (IDRs) cause significant morbidity and mortality. In an animal model of IDRs, 50-80% of Brown Norway rats exposed to D-penicillamine develop an autoimmune syndrome after several weeks of treatment. The symptoms of the IDR are similar to that observed in humans who take D-penicillamine. The mechanism of this reaction is unknown, and no effective biomarkers have been identified to predict susceptibility. We postulate that cell stress caused by drugs is required to initiate the response. We used a high-throughput approach to identify factors that might represent danger signals by profiling hepatic gene expression 6 h after dosing with D-penicillamine (150 mg/kg). Our results show that the drug-treated animals cluster into two distinct groups. One group exhibits substantial expression changes relative to control animals. The most significantly altered transcripts have a role in stress, energy metabolism, acute phase response, and inflammation. We used quantitative reverse transcriptase polymerase chain reaction to measure transcript levels in liver biopsies of 33 rats and found that resistant animals cluster together. This "resistant" cluster of animals contains 87.5% (7/8) resistant animals but only 48% (12/25) "sensitive" animals. This separation is statistically significant at the p = 0.01 level.

Animal models of idiosyncratic drug reactions

Idiosyncratic drug reactions represent a major problem. In most cases the mechanisms of these reactions are unknown, but circumstantial evidence points to the involvement of reactive metabolites and the characteristics of the reactions suggest involvement of the immune system. If progress is to be made in dealing with these adverse reactions it is essential that we have a better understanding of their mechanisms, and it is hard to imagine testing mechanistic hypotheses without good animal models. Unfortunately, idiosyncratic reactions are also idiosyncratic in animals so few good models exist. The best models, in which a rodent develops a clinical syndrome similar to that which occurs in humans, appear to be penicillamine-induced autoimmunity in Brown Norway rats and nevirapine-induced skin rash in rats. Sulfamethoxazole-induced hypersensitivity in dogs and propylthiouracil-induced autoimmunity in cats are also similar to adverse reactions that occur in people, but they have practical limitations. Halothane-induced liver toxicity in guinea pigs and amodiaquine-induced bone marrow and liver toxicity in rats represent models in which there is an immune response and mild, reversible toxicity. It is possible that the development of immune tolerance is what limits the toxicity in these models, and if this is true, interventions that prevent tolerance might lead to good models. Although the history of developing animal models of idiosyncratic drug reactions is mostly one of failure, such models are essential. A better understanding of immune tolerance may greatly facilitate the development of better models; transgenic technology may also provide an important tool.

Tolerance induced by low dose D-penicillamine in the brown Norway rat model of drug-induced autoimmunity is immune-mediated

Most patients taking drugs associated with idiosyncratic drug reactions tolerate the drug and do not develop adverse reactions. Understanding the mechanism of tolerance to drugs is important as it could provide insight into why some patients develop idiosyncratic reactions and others do not. The Brown Norway rat model of D-penicillamine-induced autoimmunity was used as a model of idiosyncratic drug-induced autoimmunity. Two weeks of low dose (5 mg/day) D-penicillamine pretreatment completely prevented all clinical signs of autoimmunity normally seen in 60-80% of rats treated with high dose (20 mg/day) D-penicillamine. Low dose pretreatment also prevented the increase in IgE and IL-4 mRNA characteristic of the response to high dose D-penicillamine. Experiments were conducted to determine whether low dose tolerance is metabolic or immunological. It was found that low dose tolerance possesses key characteristics of immune-mediated tolerance: memory, splenocytes that adoptively transfer tolerance, and regulatory cytokine production. To provide an understanding of the factors that can prevent or reverse established tolerance, the conditions for inducing and maintaining tolerance were investigated. Tolerance induction was investigated by manipulating the immune system during the period of low dose exposure. The induction of tolerance was partially prevented by depleting the macrophage subset of antigen presenting cells with clodronate-filled liposomes or by inhibiting T cells with tacrolimus during the period of low dose exposure. As well, the induction of tolerance was completely prevented by repeatedly stimulating the immune system throughout the period of low dose pretreatment with poly I:C. To investigate the permanence of tolerance, the immune system was stimulated after tolerance induction in an attempt to break tolerance. Both LPS and poly I:C reversed tolerance in a dose-dependent manner. These results demonstrate that immune tolerance to D-penicillamine autoimmunity can be induced by short-term low dose pretreatment.

Characterization of a potential animal model of an idiosyncratic drug reaction: nevirapine-induced skin rash in the rat

Idiosyncratic drug reactions are difficult to study in humans due to their unpredictability. Unfortunately, this characteristic also hinders the development of animal models needed for mechanistic studies. Nevirapine, used to treat human immunodeficiency virus (HIV) infections, results in a severe idiosyncratic skin rash in some patients. We found that nevirapine can also cause a significant rash in some strains of rats. At a dose of 150 mg/kg/day, the incidence in female Sprague-Dawley rats was 6/28 (21%), in female Brown Norway rats 32/32 (100%), and in female Lewis rats 0/6 (0%) while no male Sprague-Dawley or Brown Norway rats developed a rash. Female SJL mice 0/7 also did not develop nevirapine-induced skin lesions. The first sign of a reaction in Brown Norway rats was red ears at days 7-10 followed by a rash with scabbing mainly on the back; this was a shorter time to onset than in Sprague-Dawley rats. Light microscopy of the skin revealed a primarily mononuclear inflammatory infiltrate and lesions typical of self-trauma. Immunohistochemistry results suggest that the infiltrate was composed of CD4 and CD8 T cells as well as macrophages. A lower dose of either 40 or 75 mg/kg/day did not lead to a rash and, in fact, 2 weeks of the lower doses induced tolerance to the 150 mg/kg/day dose in female Brown Norway rats. A dose of 100 mg/kg/day resulted in rash in 2/4 (50%) of female Brown Norway rats. Rechallenge of Brown Norway rats that had been allowed to recuperate after a nevirapine-induced rash led to red ears in less than 24 h followed by hair loss and occasional skin lesions. Although the skin rash was less evident on rechallenge, microscopically, the cellular infiltrate was more prominent, especially surrounding the hair follicles. Moreover, there were lesions of interface dermatitis with apoptosis and satellitosis, indicative of a cell-mediated immune attack on the epidermis. While systemic signs of illness did not accompany the rash on primary exposure, on rechallenge, the animals appeared generally unwell and this forced sacrifice after 2 weeks or less of treatment. Importantly, splenocytes isolated from rechallenged animals were able to transfer susceptibility to nevirapine-induced skin rash to naïve female Brown Norway recipients, which was illustrated by a faster time to onset of rash in the recipients. The characteristics of this adverse reaction are similar to that seen in humans; that is, it is idiosyncratic in that it only occurs in some strains of animals, is delayed in onset, is more common in females, is dose-dependent, and appears to be immune-mediated. Therefore, it may represent a good animal model for the study of idiosyncratic drug reactions.

Modulation of D-penicillamine-induced autoimmunity in the Brown Norway rat using pharmacological agents that interfere with arachidonic acid metabolism or synthesis of inducible nitric oxide synthase

The D-penicillamine-induced autoimmune syndrome observed in Brown Norway (BN) rats is similar to an idiosyncratic reaction seen in some patients. We have previously shown that pretreatment of BN rats with aminoguanidine, an inducible nitric oxide synthase (iNOS) inhibitor, and misoprostol, a prostaglandin E (PGE) analog, completely prevented the development of D-penicillamine-induced autoimmunity. In an effort to further understand the role of arachidonic acid metabolism and iNOS in the pathogenesis of D-penicillamine-induced autoimmunity we had 3 objectives: (1) to test whether aminoguanidine and misoprostol could reverse D-penicillamine-induced autoimmunity; (2) whether BN rats that had previously developed D-penicillamine-induced autoimmunity could be protected on re-challenge with drug by pretreatment with aminoguanidine and misoprostol and (3) whether non-steroidal anti-inflammatory drugs, which inhibit PGE synthesis, would potentiate D-penicillamine-induced autoimmunity. We found that neither aminoguanidine nor misoprostol had any significant effect on the speed of recovery from D-penicillamine-induced autoimmunity. Prevention of disease on re-challenge after a 4 week recovery was less effective than on initial treatment with 7/8 animals pretreated with aminoguanidine getting sick again, while only 5/13 animals pretreated with misoprostol became ill. The effect of aminoguanidine was not significantly different from control (16/17) but that of misoprostol was (P=0.002). A single dose of the non-selective cyclooxygenase (COX) inhibitor, ketoprofen, decreased the time to onset of D-penicillamine-induced autoimmunity and continuous treatment significantly increased the incidence (P=0.024). Diclofenac, which is more selective, did not have a significant effect, and one dose of the selective inhibitor, rofecoxib, actually appeared to lower the incidence of D-penicillamine-induced autoimmunity (P=0.001). In this animal model of drug-induced autoimmunity, non-selective COX inhibitors appear to increase the incidence of disease. However, once the reaction occurs, prostaglandins are not effective for treatment and are only partially protective in an already sensitized animal.

Confirming false adverse reactions to drugs by performing individualized, randomized trials

One-patient, randomized, double-blind, controlled trials (N-of-1 RCTs) have traditionally been used to assess the efficacy of treatment. At the Drug Safety Clinic, Toronto, this methodology is used to evaluate adverse effects related to medication use, specifically when the symptoms are vague and are in response to more than one medication. Two patients are described with histories of drug allergies to multiple medications; as well, guidelines for conducting N-of-1 trials are summarized. The first patient had a history of prolonged periorbital and generalized weakness lasting up to one week after exposure to a variety of drugs. Because of the ambiguous results of local anesthetic skin testing, an N-of-1 trial was performed using lidocaine without preservative. Two short-lived episodes of blepharospasm and lethargy were observed with placebo; no subjective or objective reaction occurred with active drug. The second patient had a history of prolonged weakness and drowsiness after exposure to many medications; she had been told that she was allergic to all drugs with a benzene ring. During the first N-of-1 trial, generalized weakness was observed with 10 mg of dimenhydrinate and all four placebo doses. During the second N-of-1 challenge using codeine, no unwarranted reactions occurred with either active or placebo drug. Traditional testing of these patients to disprove the clinical symptoms is often difficult because of the anxiety level associated with the patients' past experiences. N-of-1 trials provide a useful alternative for the management of patients with nonspecific symptomatology attributed to drug ingestion.

Predicting drug-induced agranulocytosis: characterizing neutrophil-generated metabolites of a model compound, DMP 406, and assessing the relevance of an in vitro apoptosis assay for identifying drugs that may cause agranulocytosis

DMP 406 is a clozapine analogue developed by Dupont-Pharma for the treatment of schizophrenia. Unfortunately it caused agranulocytosis in dogs during preclinical studies. Clozapine also causes agranulocytosis and this is believed to be due to a reactive nitrenium ion metabolite produced by neutrophils. We studied the oxidation of DMP 406 by activated neutrophils and found that the major reactive species that is produced is not a nitrenium ion but rather an imine. This metabolite is similar to the reactive metabolite that has been proposed to be responsible for mianserin-induced agranulocytosis. Therefore we also studied the oxidation of mianserin by activated neutrophils and found that, although the major species is an iminium ion, it also bears a lactam moiety in the piperazine ring resulting from further oxidation. We usually find that HOCl is a good model system for the production of reactive metabolites of drugs that are formed by activated neutrophils, but in the case of both DMP 406 and mianserin, the products produced were significantly different than those formed by activated neutrophils. In contrast, the combination of horseradish peroxidase and hydrogen peroxide (HRP/H(2)O(2)) formed a very similar pattern of products, and this system was used to produce sufficient quantities of metabolites to allow for identification. The reactive metabolites of both DMP 406 and mianserin reacted with a range of nucleophiles, but in many cases the reaction was reversible. The best nucleophile for trapping these reactive metabolites was cyanide. It has been demonstrated that the products of clozapine oxidation by HRP/H(2)O(2), presumably the nitrenium ion, induced apoptosis in neutrophils at therapeutic concentrations of clozapine. It has been suggested that this process is involved in the mechanism of clozapine-induced agranulocytosis. We tested DMP 406 and mianserin in this system to see if the ability of a reactive metabolite of a drug to cause apoptosis could predict the ability of that drug to cause agranulocytosis. We used clozapine as a positive control and we also tested olanzapine, a drug that forms a reactive metabolite similar to that of clozapine but is given at a lower dose and does not cause agranulocytosis. We found that DMP 406 did not increase apoptosis at concentrations below 50 microM, and although mianserin did increase apoptosis at 10 microM this is above the therapeutic concentration. Olanzapine caused an increase in apoptosis at the same concentration as clozapine (1 microM), but because its therapeutic concentration is lower, this concentration was above the pharmacological range. There was no increase in apoptosis with any drug in the absence of HRP/H(2)O(2). These results indicate that this assay is unable to reliably predict the ability of different types of drugs to cause agranulocytosis. This is not a surprising result given that different drugs may induce agranulocytosis by different mechanisms.

Metabolism of drugs by leukocytes

Neutrophils and monocytes can metabolize drugs to reactive metabolites, especially those drugs that have nitrogen or sulfur in a low oxidation state. The major system involved in this oxidation is the combination of NADPH oxidase and myeloperoxidase which generates HOCl. Although this system is unlikely to be quantitatively important, i.e. it is unlikely to have a significant effect on the pharmacokinetics of a drug, the reactive metabolites produced appear to have significant biological effects. Reactive metabolites, by their very nature, have short half-lives, and most of their effects will be exerted on the cells that formed them. Therefore, they are likely to be important for adverse reactions that involve leukocytes, such as agranulocytosis and immune-mediated reactions. However, the mechanism of these reactions is unknown and evidence for the association of leukocyte-derived reactive metabolites with such reactions is circumstantial at present. There is also circumstantial evidence to link the formation of such reactive metabolites to the antiinflammatory effects of some drugs. Possible mechanisms include the scavenging of other reactive species or inhibition of cells, especially neutrophils and macrophages, involved in inflammation. The oxidation of drugs by leukocytes requires activation of the cells; therefore, infection or other inflammatory conditions that activate leukocytes may represent one of the risk factors for idiosyncratic drug reactions.

Mechanism of idiosyncratic drug reactions: reactive metabolite formation, protein binding and the regulation of the immune system

Drug-induced adverse reactions, especially type B reactions, represent a major clinical problem. They also impart a significant degree of uncertainty into drug development because they are often not detected until the drug has been released onto the market. Type B reactions are also termed idiosyncratic drug reactions by many investigators due to their unpredictable nature and our lack of understanding of the mechanisms involved. It is currently believed that the majority of these reactions are immune-mediated and are caused by immunogenic conjugates formed from the reaction of a reactive metabolite of a drug with cellular proteins. It has been shown that most drugs associated with idiosyncratic reactions form reactive metabolites to some degree. Covalent binding of reactive metabolites to cellular proteins has also been shown in many cases. However, studies to reveal the role of reactive metabolites and their protein-adducts in the mechanism of drug-induced idiosyncratic reactions are lacking. This review will focus on our current understanding and speculative views on how a reactive metabolite of a drug might ultimately lead to immune-mediated toxicity.

Factors that modify penicillamine-induced autoimmunity in Brown Norway rats: failure of the Th1/Th2 paradigm

Idiosyncratic drug reactions appear to be immune-mediated. Immune responses are driven by helper T cells (Th); Th1 responses promote cell-mediated immunity, whereas Th2 responses drive antibody-mediated reactions. Th1 cytokines inhibit Th2 responses and Th2 cytokines inhibit Th1 responses; therefore, it may be possible to prevent idiosyncratic drug reactions by changing the Th1/Th2 cytokine balance. We tested this hypothesis in an animal model in which penicillamine causes an autoimmune syndrome in Brown Norway rats. This syndrome has the hallmarks of a Th2-mediated response and we tried to inhibit it with a polymer of inosine and cytosine (poly I:C), a Th1 cytokine-inducer. However, we found that a single dose of poly I:C, given at the onset of penicillamine treatment, significantly increased both the incidence (100 vs. 60%) and accelerated the onset (30+/-4 vs. 39+/-5 days) of penicillamine-induced autoimmunity when compared with controls. To rule out other effects of poly I:C that might overshadow the induction of Th1 cytokines, we directly tested the effects of the prototypic Th1 cytokine, interferon-gamma. Although not as dramatic, interferon-gamma-pretreatment also appeared to make the syndrome worse. Conversely, when we used misoprostol, a prostaglandin-E analog that inhibits Th1 cytokines, it completely protected the animals. Just one dose of misoprostol prior to initiation of penicillamine treatment was sufficient to provide this protection. The syndrome was also completely inhibited by aminoguanidine, an inhibitor of iNOS. These results, although dramatic, suggest that the effects of these agents were not mediated by their effects on Th1/Th2 balance, but rather by some other mechanism.

Identification of a reactive metabolite of terbinafine: insights into terbinafine-induced hepatotoxicity

Oral terbinafine treatment for superficial fungal infections of toe and fingernails is associated with a low incidence (1:45000) of hepatobiliary dysfunction. Due to the rare and unpredictable nature of this adverse drug reaction, the mechanism of toxicity has been hypothesized to be either an uncommon immunological or metabolically mediated effect. However, there is little evidence to support either mechanism, and toxic metabolites of terbinafine have not been identified. We incubated terbinafine with both rat and human liver microsomal protein in the presence of GSH and were able to trap an allylic aldehyde, 7,7-dimethylhept-2-ene-4-ynal (TBF-A), which corresponds to the N-dealkylation product of terbinafine. TBF-A was also prepared synthetically and reacted with excess GSH to yield conjugates with HPLC retention times and mass spectra identical to those generated in the microsomal incubations. The major GSH conjugate, characterized by (1)H NMR, corresponds to addition of GSH in a 1,6-Michael fashion. There remains a second electrophilic site on this metabolite, which can bind either to a second molecule of GSH or to cellular proteins via a 1,4-Michael addition mechanism. Moreover, we demonstrated that the formation of the GSH conjugates was reversible. We speculate that this allylic aldehyde metabolite, formed by liver enzymes and conjugated with GSH, would be transported across the canalicular membrane of hepatocytes and concentrated in the bile. The mono-GSH conjugate, which is still reactive, could bind to hepatobiliary proteins and lead to direct toxicity. Alternatively, it could modify canalicular proteins and lead to an immune-mediated reaction causing cholestatic dysfunction.

Is it possible to more accurately predict which drug candidates will cause idiosyncratic drug reactions?

The unexpected occurrence of idiosyncratic drug reactions during late clinical trials or after a drug has been released can lead to a severe restriction in its use or failure to release/withdrawal. This leads to considerable uncertainty in drug development and has led to attempts to try to predict a drug's potential to cause such reactions. It appears that most idiosyncratic drug reactions are due to reactive metabolites; however, many drugs that form reactive metabolites are associated with a very low incidence of idiosyncratic drug reactions. Therefore. screening drug for their ability to generate reactive metabolites is likely to cause the rejection of many good drug candidates. There is evidence to suggest that an idiosyncratic drug reaction is more likely if there is some "danger signal'. Thus drugs that cause some degree of cell stress or damage may be more likely to lead to a high incidence of idiosyncratic drug reactions. The exact nature of the putative danger signals is unknown. However, a screen of the effects of drugs known to be associated with a high incidence of idiosyncatic reactions using expression genomics and proteomics may reveal a pattern or patterns of mRNA and protein expression that predict which drugs will cause a high incidence of idiosyncratic drug reactions. Although idiosyncratic drug reactions are not usually detected in animal tests because they are just as idiosyncratic in animals as they are in humans, it is likely that drug reactive metabolites would also cause similar cell stress in animals. It is more likely that in most cases it is differences in the immune response to the reactive metabolites that determine which individuals will develop an idiosyncratic reaction. If the expression of certain proteins in animals treated with a drug candidate could be used as a screening method to predict a drug's potential to cause a high incidence of idiosyncratic drug reactions, it would greatly facilitate the development of safer drugs.

Induction of metabolism-dependent and -independent neutrophil apoptosis by clozapine

Clozapine, an atypical antipsychotic used in the treatment of refractory schizophrenia, causes neutropenia and agranulocytosis in 3 and 0.8% of patients, respectively. Clozapine undergoes bioactivation to a chemically reactive nitrenium ion, which has been shown to cause neutrophil cytotoxicity. To define further the mechanism of cell death, we have investigated the toxicity of clozapine, its stable metabolites, and its chemically reactive nitrenium ion to neutrophils and lymphocytes. Clozapine was able to induce neutrophil apoptosis at therapeutic concentrations (1-3 microM) only when it was bioactivated to the nitrenium ion. The parent drug caused apoptosis at supratherapeutic concentrations (100-300 microM) only. Neutrophil apoptosis induced by the nitrenium ion, but not by the parent drug itself, was inhibited by antioxidants and genistein and was accompanied by cell surface haptenation (assessed by flow cytometry) and glutathione depletion. Dual-color flow cytometry showed that neutrophils that were haptenated were the same cells that underwent apoptosis. No apoptosis of lymphocytes was evident with the nitrenium ion or the parent drug, despite the fact that the former caused cell surface haptenation, glutathione depletion, and loss of membrane integrity. Demethylclozapine, the major stable metabolite in vivo, showed a profile that was similar to, although less marked than that observed with clozapine. N-oxidation of clozapine or replacement of the nitrogen (at position 5) by sulfur produced compounds that were entirely nontoxic to neutrophils. In conclusion, the findings of the study expand on potential mechanisms of clozapine-induced cytotoxicity, which may be of relevance to the major forms of toxicity encountered in patients taking this drug.

Metabolism of ticlopidine by activated neutrophils: implications for ticlopidine-induced agranulocytosis

Ticlopidine is associated with a relatively high incidence of agranulocytosis and aplastic anemia. We have shown that other drugs associated with agranulocytosis are metabolized to reactive metabolites by activated human neutrophils or by HOCl, which is the major oxidant produced by activated neutrophils. We set out to test the hypothesis that ticlopidine also fits this pattern and is oxidized to a reactive intermediate by activated neutrophils and HOCl. As much as 8% ticlopidine was metabolized by activated human neutrophils to a dehydro-ticlopidine; however, this product did not account for all of the decrease in ticlopidine concentration. The oxidation products of ticlopidine by the combination of myeloperoxidase and hydrogen peroxide were the same as those by HOCl: dehydrogenated ticlopidine and 2-chloroticlopidine. A neutrophil-derived reactive metabolite of ticlopidine was trapped with GSH and the same ticlopidine-GSH conjugate was found in both the myeloperoxidase and HOCl systems. Evidence for the identity of the reactive metabolite was obtained by reaction of ticlopidine with HOCl in a flow reaction system coupled to a mass spectrometer. The mass spectra suggested that the reactive metabolite was a thiophene-S-chloride. We conclude that ticlopidine follows the same pattern of reactive metabolite formation by activated neutrophils as other drugs associated with a high incidence of agranulocytosis, and the putative thiophene-S-chloride formed by activated neutrophils may be responsible for ticlopidine-induced agranulocytosis.

Bioactivation and covalent binding of hydroxyfluperlapine in human neutrophils: implications for fluperlapine-induced agranulocytosis

The use of fluperlapine and the structurally related clozapine has been associated with the induction of agranulocytosis in humans. Unlike clozapine, fluperlapine is relatively resistant to chemical and biochemical oxidations. In this study we demonstrated that 7-hydroxyfluperlapine, the major metabolite of fluperlapine in humans, is oxidized to a reactive intermediate by HOCl and by myeloperoxidase in the presence of H(2)O(2) and Cl(-). This reactive intermediate was identified as an iminoquinone species with a M + 1 ion at m/z 324 by mass spectrometry. The iminoquinone intermediate was trapped by N-acetyl-L-cysteine (NAC) as well as GSH. NMR spectra of the NAC adducts indicated that the NAC was bound to the 6 and 9 positions of the aromatic ring. This is the same orientation as the binding of nucleophiles to the reactive metabolite of clozapine. We were able to use an antibody against clozapine to demonstrate that 7-hydroxyfluperlapine, but not fluperlapine itself, covalently modifies human myeloperoxidase. Furthermore, we demonstrated that 7-hydroxyfluperlapine is metabolized by activated neutrophils to a reactive intermediate that covalently binds to neutrophils. In the presence of NAC or GSH, such covalent binding was inhibited and the NAC or GSH adducts were formed. Thus, the reactivity and even the orientation of the binding of the reactive metabolite of 7-hydroxyfluperlapine is very similar to that of clozapine. These results provide a mechanism for the formation of a reactive metabolite of fluperlapine similar to clozapine that may explain its ability to induce agranulocytosis.

Metabolism of trimethoprim to a reactive iminoquinone methide by activated human neutrophils and hepatic microsomes

The antibacterial agent, trimethoprim, is normally used synergistically with sulfonamides. Its use is associated with idiosyncratic reactions including liver toxicity and agranulocytosis. In this study, we demonstrated that trimethoprim was oxidized by activated human neutrophils, as well as a combination of myeloperoxidase/hydrogen peroxide/chloride or hypochlorous acid, to a reactive pyrimidine iminoquinone methide intermediate with a protonated molecular ion of m/z 289 as detected by mass spectrometry. In the presence of N-acetyl-L-cysteine (NAC), the pyrimidine iminoquinone methide could be trapped as three NAC adducts. The three NAC adducts were separable on HPLC, but showed the same protonated molecular ion of m/z 452. The proton NMR spectrum of the major adduct showed that the NAC group was at the 6 position of the pyrimidine ring. The mass spectra of the two minor NAC adducts indicated that they were the two diastereomers in which NAC was attached to the exo-cyclic prechiral carbon of the pyrimidine iminoquinone methide. Incubation of trimethoprim with isolated hepatic microsomes, both human and rat, in presence of NAC gave the same set of trimethoprim-NAC adducts. We propose that the formation of this pyrimidine iminoquinone methide by both hepatic microsomes and neutrophils may be responsible for trimethoprim-induced idiosyncratic hepatotoxicity and agranulocytosis.

Detection of 2-hydroxyiminostilbene in the urine of patients taking carbamazepine and its oxidation to a reactive iminoquinone intermediate

Carbamazepine is one of the most widely used anticonvulsants in North America; however, its use is associated with a range of serious idiosyncratic adverse reactions. These reactions are thought to result from the formation of chemically reactive metabolites. Carbamazepine is extensively metabolized in the liver and one of the major metabolites is 2-hydroxycarbamazepine, which has previously been detected as a urinary metabolite excreted by rats and humans along with its further metabolized product, 2-hydroxyiminostilbene. In this study, we found that the urine of patients taking carbamazepine appeared to contain more of the glucuronide of 2-hydroxyiminostilbene than that of 2-hydroxycarbamazepine. We have also demonstrated that 2-hydroxyiminostilbene can be oxidized readily to an iminoquinone species by HOCl, H2O2 or even on exposure to air. The reactivity of this iminoquinone as an electrophile was studied. It was shown to react with sulfhydryl-containing nucleophiles, such as glutathione and N-acetylcysteine. We also found a metabolite with the same molecular weight as 4-methylthio-2-hydroxyiminostilbene, but not the corresponding carbamazepine derivative, in the urine of patients taking carbamazepine and this presumably reflects the formation of a glutathione conjugate of the reactive iminoquinone. This iminoquinone intermediate may play a role in carbamazepine-induced idiosyncratic reactions.

Oxidation of a metabolite of indomethacin (Desmethyldeschlorobenzoylindomethacin) to reactive intermediates by activated neutrophils, hypochlorous acid, and the myeloperoxidase system

The use of indomethacin is associated with a relatively high incidence of adverse reactions such as agranulocytosis. Many other drugs associated with agranulocytosis are metabolized to reactive metabolites by activated neutrophils. Therefore, we studied the oxidation of indomethacin and its metabolites by activated neutrophils, myeloperoxidase (MPO) (the major oxidizing enzyme in neutrophils), and HOCl (the major oxidant produced by activated neutrophils). No oxidation of indomethacin by activated neutrophils was observed. However, desmethyldeschlorobenzoylindomethacin (DMBI), a major metabolite of indomethacin, was oxidized to a reactive iminoquinone that could be trapped with glutathione (GSH) or N-acetylcysteine (NAC) to form conjugates, with MH+ ions at m/z 511 and 367, respectively. No metabolism was detected in neutrophils that had not been activated, and the oxidation was inhibited by azide (which inhibits MPO) and by catalase (which catalyzes the breakdown of H2O2). In reactions with HOCl, the same reactive intermediate was formed; its mass spectrum, with a MH+ ion at m/z 204, was obtained by using a flow system in which the reactants were fed into a mixing chamber and the products flowed directly into the mass spectrometer. The same GSH and NAC conjugates were also observed when DMBI was oxidized by HOCl or by the MPO system, followed by addition of GSH or NAC. NMR data for the NAC conjugate indicated that the sulfur was substituted in the 4-position on the aromatic ring. The reactive intermediate generated from DMBI by activated neutrophils may be responsible for indomethacin-induced agranulocytosis.

A comparison of the covalent binding of clozapine and olanzapine to human neutrophils in vitro and in vivo

Covalent binding of a reactive metabolite of clozapine to neutrophils or their precursors is thought to play a role in the development of clozapine-induced agranulocytosis. Immunoblotting studies with an anti-clozapine antiserum detected covalent binding of clozapine to human neutrophils in vitro when HOCl was used to generate clozapine reactive metabolite (major clozapine adducts of 31, 49, 58, 78, 86, 126, 160, and 204 kDa). In addition, incubating neutrophils with clozapine and H2O2 (major clozapine adducts of 49 and 58 kDa) or clozapine, H2O2, and human myeloperoxidase (major clozapine adducts of 31, 49, 58, and 126 kDa) also resulted in covalent binding of clozapine to the neutrophils. The covalent binding of clozapine to neutrophils was inhibited by extracellular glutathione when HOCl, but not H2O2 was used to generate reactive metabolite. We found that the antiserum against clozapine also recognized olanzapine, an antipsychotic drug that forms a similar reactive metabolite to clozapine but has not been associated with induction of agranulocytosis. Repeating the in vitro experiments with olanzapine revealed that the major olanzapine-modified polypeptides had molecular masses of 96, 130-170, and 218 kDa. Only relatively low levels of 31, 49, and 58 kDa adducts were observed. Clozapine-modified polypeptides also were detected in neutrophils from patients being treated with clozapine. A major 58-kDa clozapine-modified polypeptide was detected in all patients tested. In contrast, no drug-modified polypeptides were detected in neutrophils from patients taking olanzapine. The differences in covalent binding exhibited by the two compounds and, in particular, the lack of olanzapine binding to human neutrophils in vivo may help to explain the difference in toxicity of these two drugs.

A comparison of the oxidation of clozapine and olanzapine to reactive metabolites and the toxicity of these metabolites to human leukocytes

Olanzapine was shown to be oxidized to a reactive intermediate by HOCl, which is the major oxidant produced by activated neutrophils. A mass spectrum obtained using a flow system in which the reactants were fed into a mixing chamber and the products flowed directly into a mass spectrometer revealed a reactive intermediate at m/z 311. This is 2 mass units less than the protonated molecular ion of parent olanzapine and suggests that the reactive intermediate is a nitrenium ion. The reactive intermediate could be trapped with glutathione or N-acetylcysteine to produce two conjugates. These data are analogous to results we reported previously with the structurally related atypical antipsychotic agent clozapine. However, the clozapine and olanzapine reactive metabolites showed differences in their ability to cause toxicity to human neutrophils. Toxicity to neutrophils was observed only at high concentrations of clozapine (>50 microM) when HOCl was used to generate reactive metabolite. In contrast, concentration-dependent toxicity (p < 0.05) was observed when neutrophils were incubated with clozapine (0-20 microM) and H2O2 to generate clozapine reactive metabolite. No toxicity was observed with clozapine alone (at concentrations of > 50 microM). Similar results were observed in monocytes and HL-60 cells. Olanzapine reactive metabolite only seemed to cause slight toxicity at the highest concentrations tested (20 microM), even when the reactive metabolite was generated using H2O2. Neutrophils from two patients with a history of clozapine-induced agranulocytosis seemed to be more sensitive to the toxic effects of the clozapine reactive metabolite; however, the numbers are too small to draw any definite conclusions.

Oxidative stress and thiol depletion in plasma and peripheral blood lymphocytes from HIV-infected patients: toxicological and pathological implications

OBJECTIVES

To determine, first, whether the plasma and lymphocytes of HIV-positive individuals and AIDS patients have alterations in the major thiols glutathione and cysteine, and/or their oxidative disulphide and mixed disulphide products; and, secondly, whether thiol/disulphide status differs in patients with sulphonamide drug hypersensitivity reactions.

DESIGN

Thiols provide critical cellular defence against toxic drug reactive intermediates and endogenous oxidative stress, and may modulate HIV replication. Glutathione is reported to be low in HIV-positive individuals and AIDS patients, but this is controversial and the mechanism responsible is unknown. Also unknown is whether altered thiol/disulphide status determines the predisposition of HIV-positive and AIDS patients to drug reactions.

METHODS

Thiols and disulphides were measured by high-performance liquid chromatography.

RESULTS

Both plasma thiols were decreased by approximately 58% in HIV-positive individuals and AIDS patients compared with uninfected controls (P < 0.05), with increases of up to threefold in oxidized products (P < 0.05). Similarly, in lymphocytes, thiols were decreased by 30-35% (P < 0.05), with apparent increases in oxidized products. For both glutathione and cysteine, the thiol/disulphide ratios also were decreased (P < 0.05). The plasma and lymphocyte glutathione thiol/disulphide ratios were highly correlated (r = 0.7661; P = 0.0001) among all subjects. No parameters differed in patients with drug reactions, or with antiretroviral therapy.

CONCLUSIONS

The enhanced thiol oxidation in HIV-positive individuals and AIDS patients indicates oxidative stress, which also contributes to thiol depletion, and may enhance damage to macromolecular targets. These mechanisms may contribute to enhanced viral replication and other pathological outcomes. HIV-positive individuals' and AIDS patients' predisposition to drug hypersensitivity reactions appears to be unrelated to thiol/disulphide status.

Current trends in drug-induced autoimmunity

Idiosyncratic adverse drug reactions have characteristics that suggest involvement of the immune system. In particular, drug-induced lupus which is an autoimmune syndrome, must be immune-mediated. A major working hypothesis for the first step in the mechanism of drug-induced autoimmunity is that the drug, or more commonly a reactive metabolite of the drug, must irreversibly bind to some structure. In view of the reactive nature of these metabolites, in most cases it is likely that the metabolite must be formed in the organ where toxicity occurs. The liver is the major site of drug metabolism and it is a common target for idiosyncratic drug reactions. In the case of immune reactions directly involving leukocytes, the enzyme system most likely responsible for the formation of reactive metabolites is the NADPH oxidase/myeloperoxidase system found in neutrophils and monocytes. In some cases, the reactive metabolite results in the production of antibodies or T-cells directed against the altered structure. However, in many other cases, the mechanism appears to be more complex than this. In some cases, true auto-antibodies are produced that do not require the presence of the drug, and furthermore, the antibodies produced often are the same as those induced by other stimuli, such as viruses. This suggests either molecular mimicry or a common alteration in the processing and presentation of antigens such that cryptic antigens are presented. Another possibility is that the reactive metabolite directly alters the class II MHC molecule leading to a graft-vs-host reaction.

Oxidation of diclofenac to reactive intermediates by neutrophils, myeloperoxidase, and hypochlorous acid

Diclofenac is associated with a low, but significant, incidence of hepatotoxicity and bone marrow toxicity. It has been suggested that this could be due to a reactive acyl glucuronide. An alternative hypothesis is that an oxidative reactive metabolite could be responsible for such reactions and such metabolites formed by the enzymes present in neutrophils could be responsible for bone marrow toxicity. Others had reported the formation of 2,2'-dihydroxyazobenzene during the oxidation of diclofenac by myeloperoxidase/hydrogen peroxide. In contrast, in similar experiments we did not find evidence for the formation of 2,2'-dihydroxyazobenzene, but we did find several products, including a reactive iminoquinone. The same iminoquinone was formed by the oxidation of 5-hydroxydiclofenac. This iminoquinone was also formed by oxidation of diclofenac by HOCl or by activated neutrophils. It reacted with glutathione to form a conjugate. 5-Hydroxydiclofenac is also a major hepatic metabolite of diclofenac, and we found that rat hepatic microsomes oxidized 5-hydroxydiclofenac to the iminoquinone which was trapped with glutathione. This reactive metabolite represents another possible cause of the idiosyncratic reactions associated with the use of diclofenac.

Clozapine is oxidized by activated human neutrophils to a reactive nitrenium ion that irreversibly binds to the cells

Clozapine was oxidized to a reactive intermediate by HOCI, which is the major oxidant produced by activated neutrophils. A mass spectrum was obtained of this reactive intermediate by using a flow system in which the reactants were fed into a mixing chamber and the products flowed directly into a Sciex API III mass spectrometer. The intermediate was observed at m/z 325, which is 2 mass units less than the protonated molecular ion of the parent drug. This intermediate reacted with water to form several products with a m/z at 343. The same products were produced by the oxidation of clozapine by the combination of myeloperoxidase, hydrogen peroxide and chloride ion. The reactive intermediate was trapped by glutathione (GSH) and several conjugates were formed. Nuclear magnetic resonance spectra of the two major conjugates indicated GSH bound to the 6 and 9 positions of the aromatic ring. These data provide further evidence for the formation of a formal nitrenium ion in which the positive charge is highly delocalized. Clozapine was also oxidized by activated neutrophils, and in the presence of GSH, the same GSH conjugates were formed. When therapeutic concentrations of radiolabeled clozapine were used, up to 7% of the drug became irreversibly bound to the neutrophils. Covalent binding was inhibited by about 30% in the presence of 1 mM GSH but was almost abolished at 5 mM GSH. The putative nitrenium ion formed by activated leukocytes could be responsible for clozapine-induced agranulocytosis.

Myeloperoxidase as a generator of drug free radicals

Reactive metabolites are believed to be responsible for many types of toxicity, including idiosyncratic drug reactions. Bone marrow is a frequent target of idiosyncratic reactions, and, since these reactions have characteristics that suggest involvement of the immune system, the formation of reactive metabolites by leucocytes could also play a role in the aetiology of idiosyncratic drug reactions. The major oxidation system in neutrophils and monocytes is a combination of NADPH oxidase and myeloperoxidase. This system oxidizes primary arylamines, such as sulphonamides, to reactive metabolites and these drugs are also associated with a high incidence of agranulocytosis, generalized idiosyncratic reactions and/ or drug-induced lupus. Clozapine is oxidized by this system to a relatively stable nitrenium ion; clozapine is also associated with a high incidence of agranulocytosis. Arylamines that have an oxygen or nitrogen in the para position, such as amodiaquine, vesnarinone and 5-aminosalicylic acid, are oxidized to quinone-like reactive intermediates. Aminopyrine is oxidized to a very reactive dication. Such reactive metabolites could also inhibit neutrophil function and mediate some of the therapeutic effects of these drugs: for example, the use of dapsone for dermatitis herpetiformis and the use of 5-aminosalicylic acid for inflammatory bowel disease.

Covalent binding of carbamazepine oxidative metabolites to neutrophils

Carbamazepine therapy is associated with several types of idiosyncratic drug reactions, including hematological disorders. In previous studies, we found that carbamazepine was metabolized by the myeloperoxidase/H2O2 system of activated neutrophils, and covalent binding of the drug to neutrophils was observed. Several metabolites were identified, including 9-acridine carboxaldehyde. Iminostilbene, a minor hepatic metabolite of carbamazepine, was metabolized to a much greater extent than carbamazepine to similar metabolites, including 9-acridine carboxaldehyde. In the present study, the covalent binding of iminostilbene to activated neutrophils was also found to be 10-fold greater than that of carbamazepine. In addition, the binding of 9-acridine carboxaldehyde to neutrophils was 100-fold that of carbamazepine and did not require activation of the neutrophils. This suggests that this aldehyde is the reactive intermediate responsible for much of the binding. To understand possible mechanisms of covalent binding, we investigated the reaction of 9-acridine carboxaldehyde with nucleophiles and found that a reaction occurs with primary amines, such as n-butylamine and N-alpha-acetyllysine, with the formation of an imine. Sodium cyanoborohydride was used to reduce the imine to a stable secondary amine. This suggests a possible mechanism for 9-acridine carboxaldehyde binding to neutrophils that could involve physiological reducing systems in place of the borohydride. 9-Acridine carboxaldehyde may be responsible for some of the adverse reactions associated with carbamazepine, especially those that involve bone marrow.