The role of cytochrome P450 enzymes in hepatic and extrahepatic human drug toxicity

The human cytochrome P450 enzyme system metabolises a wide array of xenobiotics to pharmacologically inactive metabolites, and occasionally, to toxicologically active metabolites. Impairment of cytochrome P450 activity, which may be either genetic or environmental, may lead to toxicity caused by the parent compound itself. In practise, this usually only applies to drugs that have a narrow therapeutic index and when their clearance is critically dependent upon the fraction normally metabolised by that pathway. P450 enzymes may also convert the drug to a chemically reactive metabolite, which, if not detoxified, may lead to various forms of hepatic and extrahepatic toxicity, including cellular necrosis, hypersensitivity, teratogenicity, and carcinogenicity, depending on the site of formation and the relative stability of the metabolite, and the cellular macromolecule with which it reacts. Variation in the regulation and expression of the drug metabolising enzymes may play a key role in both interindividual variation in sensitivity to drug toxicity and tissue-specific damage. Avoidance of toxicity may be possible in rare instances by prediction of individual susceptibility or by designing new chemical entities that are metabolised by a range of enzymes (both cytochromes P450 and others) and do not undergo bioactivation.

Species variation in the bioactivation of tacrine by hepatic microsomes

1. The metabolite profile of tacrine (1,2,3,4-tetrahydro-9-amino acridine) was similar in hepatic microsomes from man, rat, dog, rabbit, mouse and hamster. Major metabolites were 1-, 2-, 4- and 7-OH tacrine. Only quantitative differences in metabolite profile were evident between species. 2. Bioactivation to protein-reactive metabolite(s) was seen in microsomes from all species. 3. 7-Methyl tacrine was found to undergo significantly less bioactivation than either 7-OH tacrine or tacrine itself. 4. In the presence of hepatic microsomes and thiol-containing agents protein-reactive metabolite formation was significantly reduced. With mercaptoethanol present a stable thioether adduct was generated from both tacrine and 7-OH tacrine. 5. Analysis of the thioether adduct by mass spectrometry yielded a molecular ion of m/z 290 consistent with the presence of a covalent adduct of 7-OH tacrine complexed in a 1:1 molar ratio with mercaptoethanol. 6. We have therefore provided further evidence for a two-step mechanism in the bioactivation of tacrine involving an initial 7-hydroxylation followed by a postulated 2-electron oxidation to yield a reactive quinone methide. This mechanism of bioactivation appears to be identical in human and animal hepatic microsomes.

The role of active metabolites in drug toxicity

Adverse drug reactions can be caused by the parent drug or a metabolite of that drug. The metabolite may be stable or chemically reactive, the resultant toxicity being either a direct extension of the pharmacology of the drug, or unrelated to the known pharmacology of the drug and dependent on the chemical properties of the compound. Many different organ systems may be affected, and there are several mechanisms involved in determining organ-specific, and sometimes cell-selective, toxicity. An imbalance between bioactivation of a drug to a toxic metabolite and its detoxification is of prime importance in determining individual susceptibility. Such an imbalance may be genetically determined or acquired and, furthermore, may be systemic or tissue-specific. Prevention of metabolite-mediated toxicity is possible once the mechanism of toxicity has been elucidated.

The effect of enzyme inhibition on the metabolism and activation of tacrine by human liver microsomes

1. Tacrine (1,2,3,4-tetrahydro-9-aminoacridine-hydrochloride: THA) underwent metabolism in vitro by a panel (n = 12) of human liver microsomes genotyped for CYP2D6, in the presence of NADPH, to both protein-reactive and stable metabolites. 2. There was considerable variation in the extent of THA metabolism amongst human livers. Protein-reactive metabolite formation showed a 10-fold variation (0.6 +/- 0.1%-5.2 +/- 0.8% of incubated radioactivity mg-1 protein) whilst stable metabolites showed a 3-fold variation (24.3 +/- 1.7%-78.6 +/- 2.6% of incubated radioactivity). 3. Using cytochrome P450 isoform specific inhibitors CYP1A2 was identified as the major enzyme involved in all routes of THA metabolism. 4. There was a high correlation between aromatic and alicyclic hydroxylation (r = 0.92, P < 0.0001) consistent with these biotransformations being catalysed by the same enzymes. 5. Enoxacin (ENOX), cimetidine (CIM) and chloroquine (CQ) inhibited THA metabolism by a preferential decrease in the bioactivation to protein-reactive, and hence potentially toxic, species. The inhibitory potency of ENOX and CIM was increased significantly upon pre-incubation with microsomes and NADPH. 6. Covalent binding correlated with 7-OH-THA formation before (r = 0.792, P < 0.0001) and after (r = 0.73, P < 0.0001) inhibition by CIM, consistent with a two-step mechanism in the formation of protein-reactive metabolite(s) via a 7-OH intermediate. 7. The use of enzyme inhibitors may provide a useful tool for examining the relationship between the metabolism and toxicity of THA in vivo.

Lymphocyte microsomal epoxide hydrolase in patients on carbamazepine therapy

1. In order to determine whether carbamazepine is an inducer of lymphocyte microsomal epoxide hydrolase, the activity of the enzyme has been measured in the lymphocytes of 40 patients on continuous drug therapy using [3H]-cis stilbene oxide as a substrate. 2. Induction of the cytochrome P450 isoform, CYP3A, has been assessed in the same patients by measurement of the 24 h urinary excretion of 6 beta-hydroxycortisol by radioimmunoassay. The urinary concentrations of carbamazepine and its two metabolites, the 10,11-epoxide and trans-dihydrodiol, have also been measured by h.p.l.c. 3. The 24 h urinary 6 beta-hydroxycortisol excretion in the patients increased with the dose of carbamazepine (r = 0.57, P < 0.001) indicating induction of CYP3A. 4. The total amount of trans-dihydrodiol excreted in the urine increased with the dose of carbamazepine, and it was the most abundant urinary metabolite in all patients and at all dose-levels. There was no relationship between the dose of carbamazepine and the diol to epoxide ratio (r = 0.01, NS). 5. Lymphocyte microsomal epoxide hydrolase activity was marginally, but significantly (P = 0.02) higher in the patients (28.4 pmol diol min-1 mg-1 protein) than in drug-free controls (23.4 pmol diol min-1 mg-1 protein (95% CI for difference -9 to -0.8)). 6. The results indicate that at concentrations of carbamazepine which produce marked induction of hepatic CYP3A, an enzyme involved in the metabolism and bioactivation of carbamazepine, there is only a slight increase in lymphocyte microsomal epoxide hydrolase.

Structure-metabolism relationships of ring-A halogenated analogues of 17 alpha-ethynyloestradiol

The metabolic fates of 2-chloro-, 2-bromo-, 4-bromo- and 2-iodo-17 alpha-ethynyloestradiol (EE2) in rats were determined. 6,7-3H-labelled analogues (0.1-2.0 mumol/kg) were administered i.v. to anaesthetized animals. The metabolites of all four compounds were rapidly and extensively excreted in bile (79-93% of the dose over 6 h). Unlike EE2 and 2-fluoro-EE2 (2-FEE2), neither 2-chloro(Cl)-(2.0 mumol/kg),2-bromo(Br)-(0.1 mumol/kg), nor 2-iodo(I)-EE2-(0.1 mumol/kg) underwent C-2 hydroxylation in female rats; 2-BrEE2 was similarly refractory in male rats; females, was subject to approx. 2-fold greater C-2 hydroxylation than 2-FEE2 but this equalled only approx. 60% of that undergone by EE2. All three of the C-2 halogenated derivatives were substantially excreted unchanged except for conjugation. 2-ClEE2 alone was C-4 hydroxylated to an appreciable extent. The oxidative metabolism of 2- and 4-BrEE2 in rats was sexually differentiated: 2-BrEE2 yielded an alkyl hydroxylated metabolite and a two-component dihydroxylated fraction in the ratio 1:0.09 and 1:0.76 in males and females, respectively; 4-BrEE2 underwent C-2 and alicyclic (C-15) hydroxylation in the ratio 1:4.8 and 1:0.07 in males and females, respectively. 2-ClEE2 formed much less alkyl monohydroxylated metabolite (C-16 hydroxylated for 2-Cl- and 2-IEE2) than did either 2-BrEE2 or 2-IEE2. The observed structure-metabolism relationships are discussed.

Catecholestrogens as mediators of carcinogenesis: correlation of aromatic hydroxylation of estradiol and its fluorinated analogs with tumor induction in Syrian hamsters

17 beta-Estradiol is known to induce kidney tumors in male Syrian hamsters when administered chronically, whereas 4-fluoroestradiol does so only after an extended induction period and 2-fluoroestradiol is not carcinogenic; both fluorinated analogs are hormonally active. Because C-4 and C-2 hydroxylations are, respectively, minor and major routes of estrogen metabolism in vivo, these observations suggest mediation of tumorigenesis by catecholestrogen metabolites. However, the analogs were reported to undergo oxidative defluorination in vitro. We have determined the metabolic fates of estradiol, 2-fluoroestradiol, and 4-fluoroestradiol in male hamsters. [6,7-3H]Estradiol was principally C-2 hydroxylated when given intravenously at either 0.1 mumol/kg or 50 mumol/kg; 2-hydroxyestradiol was eliminated in bile and urine, largely as a glucuronide, without undergoing extensive deactivation via O-methylation. Alicyclic alcohol metabolites were minor products. [6,7-3H]2-Fluoroestradiol underwent either glucuronylation or sequential dehydrogenation and alicyclic hydroxylation followed by glucuronylation but neither oxidative defluorination nor compensatory C-4 hydroxylation. [6,7-3H]4-Fluoroestradiol was also considerably dehydrogenated to the keto form and glucuronylated. Nevertheless, only 4-fluoroestradiol yielded appreciable quantities of C-2 hydroxylated metabolite at the lower dose; methylation was an insignificant pathway. No defluorinated products were observed. Dehydrogenation of both analogs and alicyclic hydroxylation of the 2-fluoroestrone metabolite were less extensive at the higher dose; all of the polar metabolites of 2-fluoroestradiol in bile, although not those in urine, declined to trace amounts. C-2 hydroxylation of 4-fluoroestradiol was greater at this dose. Thus, the rank order of catechol formation from estradiol and its fluoro analogs observed in vivo, unlike that found in microsomal incubations, was consistent with the hypothesis that catechols mediate estrogen-dependent renal carcinogenesis in hamsters.

Investigation of mechanisms in toxic epidermal necrolysis induced by carbamazepine

BACKGROUND

Erythema multiforme and toxic epidermal necrolysis can occur as serious and even life-threatening adverse drug reactions. The underlying mechanisms are unknown, but evidence suggests that affected individuals may have impaired capacity to detoxify reactive intermediate drug metabolites. Such intermediates may be directly toxic or may react with host tissues to form antigens, evoking an immune response. We describe our investigation of a patient with carbamazepine-induced erythema multiforme and toxic epidermal necrolysis. The inflammatory infiltrate was examined immunocytochemically in lesional skin specimens from the patient, in the patient's patch test response to carbamazepine, and in lesional skin specimens from five other patients with drug-induced erythema multiforme. The patient's lymphocytes were examined both for susceptibility to cytotoxic damage by liver microsome-induced carbamazepine metabolites and for proliferative responses to native carbamazepine, which might indicate cell-mediated immune sensitization.

OBSERVATIONS

Lesions of toxic epidermal necrolysis were more florid, but findings were essentially similar in all the skin samples examined. In the dermis there were CD14+ macrophages, CD1a+ Langerhans cells, and CD3+ CD45RO+ T cells. The CD4-CD8 T-cell ratio was 2:1, and 10% of the T cells were CD25+, suggesting activation by recent encounter with antigen. The epidermis contained CD14+ macrophages and T cells, but the CD8+ cells out-numbered the CD4+ cells. Up to 25% of the T cells were CD25+. Lymphocyte proliferation was not induced by native carbamazepine, but the patient's lymphocytes were significantly more susceptible to cytotoxic killing by liver microsome-induced carbamazepine intermediates.

CONCLUSIONS

The inflammatory reaction in skin affected by erythema multiforme and toxic epidermal necrolysis was rich in CD8+ T cells, suggesting an immune cytotoxic reaction. The patient appeared to have a reduced capacity to detoxify reactive intermediates. This, together with the lack of lymphocyte response to native drug but a positive patch test response, suggests that the immune response may be directed at drug-modified epidermal cells.

The effect of fluorine substitution on the metabolism and antimalarial activity of amodiaquine

Amodiaquine (AQ) (2) is a 4-aminoquinoline antimalarial which causes adverse side effects such as agranulocytosis and liver damage. The observed drug toxicity is believed to be related to the formation of an electrophilic metabolite, amodiaquine quinone imine (AQQI), which can bind to cellular macro-molecules and initiate hypersensitivity reactions. 5'-Fluoroamodiaquine (5'-FAQ, 3), 5',6'-difluoroamodiaquine (5',6'-DIFAQ,4), 2',6'-difluoroamodiaquine (2',6'-DIFAQ,5), 2',5',6'-trifluoroamodiaquine (2',5',6'-TRIFAQ, 6) and 4'-dehydroxy-4'-fluoroamodiaquine (4'-deOH-4'-FAQ,7) have been synthesized to assess the effect of fluorine substitution on the oxidation potential, metabolism, and in vitro antimalarial activity of amodiaquine. The oxidation potentials were measured by cyclic voltammetry, and it was observed that substitution at the 2',6'- and the 4'-positions (2',6'-DIFAQ and 4'-deOH-4'-FAQ) produced analogues with significantly higher oxidation potentials than the parent drug. Fluorine substitution at the 2',6'-positions and the 4'-position also produced analogues that were more resistant to bioactivation. Thus 2',6'-DIFAQ and 4'-deOH-4'-FAQ produced thioether conjugates corresponding to 2.17% (SD: +/- 0.27%) and 0% of the dose compared with 11.87% (SD: +/- 1.31%) of the dose for amodiaquine. In general the fluorinated analogues had similar in vitro antimalarial activity to amodiaquine against the chloroquine resistant K1 strain of Plasmodium falciparum and the chloroquine sensitive T9-96 strain of P. falciparum with the notable exception of 2',5',6'-TRIFAQ (6). The data presented indicate that fluorine substitution at the 2',6'-positions and replacement of the 4'-hydroxyl of amodiaquine with fluorine produces analogues (5 and 7) that maintain antimalarial efficacy in vitro and are more resistant to oxidation and hence less likely to form toxic quinone imine metabolites in vivo.

Application of scanning electron microscopy in assessing the prevalence of some Setaria species in Korean cattle

The numbers of individual Setaria species in the peritoneal cavities of Korean cattle were estimated. The worms were tentatively identified under light microscopy, and then precisely classified by SEM on the basis of unique features at the anterior and posterior ends of the adult worms. The positive rate of Setaria species was 34.2% out of 1,074 Korean cattle surveyed; that of S. digitata was 25.1%, S. marshalli 2.9% and both species 6.2%, respectively. Out of a total of 1,254 worms collected, 66.8% were female and 19.1% were male S. digitata; 9.2% were female and 5.0% were male S. marshalli, respectively. The average length of S. marshalli was somewhat longer than that of S. digitata. S. marshalli was documented for the first time in Korea.

Cortisol metabolism by human liver in vitro--IV. Metabolism of 9 alpha-fluorocortisol by human liver microsomes and cytosol

The oxidative and reductive biotransformations of 9 alpha-fluorocortisol (fluorocortisol) by human liver microsomes and cytosol have been characterized. 9 alpha-Fluorination greatly simplified cortisol metabolism in microsomes: dehydrogenation of the 11 beta-hydroxyl group and A-ring (4-ene-5 beta and 3 alpha-keto) reduction, the principle pathways, were completely blocked. Fluorocortisol was essentially metabolized by the remaining pathways, 20 beta-reduction and 6 beta-hydroxylation. In cytosol, 20 beta-reduction replaced the A-ring reduction of cortisol as the sole biotransformation. The major structure-metabolism relationships of fluorocortisol in man, i.e. complete and extensive inhibition of 11 beta-dehydrogenation and 4-ene-5 beta-reduction, respectively, were attributed to hepatic enzyme systems. Their mechanistic basis is discussed with reference to the electronic and conformational changes induced by 9 alpha-fluorination.

An investigation of the formation of cytotoxic, genotoxic, protein-reactive and stable metabolites from naphthalene by human liver microsomes

Chemically reactive epoxide metabolites have been implicated in various forms of drug and chemical toxicity. Naphthalene, which is metabolized to a 1,2-epoxide, has been used as a model compound in this study in order to investigate the effects of perturbation of detoxication mechanisms on the in vitro toxicity of epoxides in the presence of human liver microsomes. Naphthalene (100 microM) was metabolized to cytotoxic, protein-reactive and stable, but not genotoxic, metabolites by human liver microsomes. The metabolism-dependent cytotoxicity and covalent binding to protein of naphthalene were significantly higher in the presence of phenobarbitone-induced mouse liver microsomes than with human liver microsomes. The ratio of trans-1,2-dihydrodiol to 1-naphthol was 8.6 and 0.4 with the human and the induced mouse microsomes, respectively. The metabolism-dependent toxicity of naphthalene toward human peripheral mononuclear leucocytes was not affected by the glutathione transferase mu status of the co-incubated cells. Trichloropropene oxide (TCPO; 30 microM), an epoxide hydrolase inhibitor, increased the human liver microsomal-dependent cytotoxicity (19.6 +/- 0.9% vs 28.7 +/- 1.0%; P = 0.02) and covalent binding to protein (1.4 +/- 0.3% vs 2.8 +/- 0.2%; P = 0.03) of naphthalene (100 microM), and reversed the 1,2-dihydrodiol to 1-naphthol ratio from 6.6 (without TCPO) to 2.6, 0.6 and 0.1 at TCPO concentrations of 30, 100 and 500 microM, respectively. Increasing the human liver microsomal protein concentration reduced the cytotoxicity of naphthalene, while increasing its covalent binding to protein and the formation of the 1,2-dihydrodiol metabolite. Co-incubation with glutathione (5 mM) reduced the cytotoxicity and covalent binding to protein of naphthalene by 68 and 64%, respectively. Covalent binding to protein was also inhibited by gestodene, while stable metabolite formation was reduced by gestodene (250 microM) and enoxacin (250 microM). The study demonstrates that human liver cytochrome P450 enzymes metabolize naphthalene to a cytotoxic and protein-reactive, but not genotoxic, metabolite which is probably an epoxide. This is rapidly detoxified by microsomal epoxide hydrolase, the efficiency of which can be readily determined by measurement of the ratio of the stable metabolites, naphthalene 1,2-dihydrodiol and 1-naphthol.

The effect of fluorine substitution on the hepatotoxicity and metabolism of paracetamol in the mouse

The widely used analgesic paracetamol (P) produces fulminant hepatocellular necrosis in humans when taken in overdose. The toxicity is mediated by drug oxidation and depletion of hepatic glutathione. We have, therefore, explored the effects of fluorine substitution on the hepatotoxicity of P in female CD1 mice. 3-Fluoro-4-hydroxyacetanilide (1FPO), 3,5-difluoro-4-hydroxyacetanilide (2FPO), 2,6-difluoro-4-hydroxyacetanilide (2FPN) and 2,3,5,6-tetrafluoro-4-hydroxyacetanilide (4FP) were synthesized, characterized and investigated for their potential to cause hepatotoxicity in the mouse. Introduction of fluorine into P increases the oxidation potential of the drug. The oxidation potentials of paracetamol and its fluorinated analogues were measured by cyclic voltametry and found to increase in the order P < 1FPO < 2FPO < 2FPN < 4FP. Serum transaminase (ALT) and hepatic glutathione were measured 24 and 6 hr, respectively, after administration of a single dose (2.65 mmol/kg) of each compound to female CD1 mice. There was significant elevation of ALT in mice given P, 1FPO and 2FPO, but not in those which received either 2FPN or 4FP. Hepatic glutathione was reduced significantly by administration of P and IFP, but not after administration of 2FPO, 2FPN or 4FP. Accordingly, glucuronide and sulphate conjugates, but not thioether metabolites, were detected in urine after administration of 14C-labelled 2FPO, 2FPN and 4FP. These data indicate that introduction of fluorine into the 2 and 6 positions increases the oxidation potential of paracetamol which in turn reduces the propensity of the molecule to undergo oxidative bioactivation, and thereby reduces the in vivo toxicity of the molecule.

[Comparative susceptibility of different cell lines for culture of Toxoplasma gondii in vitro]

In order to establish a useful cell culture system for T. gondii, we compared the degree of proliferation of T. gondii tachyzoites among 8 different cell lines; 2 kinds of normal animal cells (MDCK-canine kidney cells; Vero-monkey kidney cells) and 6 kinds of human tumor cells (A 549, PC 14-lung cancer cells; SNU 1, SNU 16, MKN 45-stomach cancer cells; HL-60-promyelocytic leukemia cells), through morphological observation and 3H-uracil uptake assay. The degree of susceptibility to infection with T. gondii tachyzoites was highest in A 549 and PC 14 cells, medium in Vero, HL-60, MDCK and SNU 1, and lowest in SNU 16 and MKN 45 cells. The kinetics of T. gondii multiplication during the post-infection 60 hours were highly dependent upon the dose of tachyzoites administered and the duration of cultivation. These results show that A 549 and PC 14 are the most suitable cell lines among the 8 tested for the growth and multiplication of T. gondii in vitro.

The effect of fluorine substitution on the physicochemical properties and the analgesic activity of paracetamol

The physicochemical properties and analgesic action of six fluorinated analogues of 4-hydroxyacetanilide (paracetamol) have been investigated. Fluorine substitution adjacent to the hydroxyl group increased lipophilicity and oxidation potential whilst substitution adjacent to the amide had little effect on lipophilicity but led to a greater increase in oxidation potential. Lack of coplanarity and conjugation of the amide group and aromatic ring was also apparent with the analogues that had fluorine in the 2 and 6 positions. Introduction of fluorine into the amide group of paracetamol increased the lipophilicity 4-fold and also increased the oxidation potential of paracetamol. ED50 values for analgesic activity in the phenylquinone-induced abdominal constriction test on male Swiss White mice showed that ring substitution by fluorine reduced activity, especially at the 2,6-positions. Introduction of fluorine into the amide group enhanced activity significantly. Correlation of the analgesic activity with the physicochemical properties indicated that conjugation (and planarity) of the amide group with the aromatic ring is essential for activity and that ease of oxidation may also be an important factor.

An investigation into the formation of stable, protein-reactive and cytotoxic metabolites from tacrine in vitro. Studies with human and rat liver microsomes

Tacrine (1,2,3,4-tetrahydro-9-aminoacridine hydrochloride; THA) is known to undergo extensive oxidative metabolism to a variety of mono- and dihydroxylated metabolites in animals and humans. The potential for tacrine to undergo metabolism to stable, protein-reactive and cytotoxic metabolites has been investigated in incubations with human and rat liver microsomes. Using lymphocytes as sensitive markers to quantify cytotoxicity, THA (50 microM) underwent NADPH-dependent bioactivation to a cytotoxic metabolite(s). NADPH-dependent cytotoxicity in the presence of rat and human microsomes was 9.8 +/- 3.1% (P < 0.05 cf. -NADPH control) and 6.2 +/- 2.0% (P < 0.05 cf. -NADPH control), respectively. Stable and protein-reactive metabolites were also formed in microsomes from both species. These accounted for 28.2 +/- 12.7% and 1.22 +/- 0.79% of incubated radioactivity in human microsomes and 6.4 +/- 2.2% and 0.4 +/- 0.1% of incubated radioactivity in rat microsomes. In microsomes pooled from six human livers the NADPH-dependent cytotoxicity was 9.4 +/- 1.1%. Formation of stable and protein-reactive metabolites accounted for 29.2 +/- 2.3% and 1.2 +/- 1.0% of incubated radioactivity. Reduced glutathione (500 microM) completely blocked NADPH-dependent cytotoxicity and inhibited protein-reactive metabolite formation by 60% (P < 0.05). Ascorbic acid (500 microM) inhibited the generation of cytotoxic and protein-reactive metabolites by 75% (P < 0.05) and 35% (P < 0.05), respectively. Cyclohexene oxide was without effect. Human serum albumin was found to protect the lymphocytes against toxicity. In microsomes prepared from the livers of four donors known to have been smokers there were no significant differences in the generation of metabolites from THA compared with microsomes prepared from livers of non-smokers. Enoxacin, a specific inhibitor of cytochrome P450 1A2 significantly inhibited all routes of THA metabolism. We have therefore demonstrated that THA may be oxidatively metabolized to stable, protein-reactive and cytotoxic metabolites in human and rat liver microsomes. A number of inhibitors may affect these process, whilst inhibition by enoxacin indicates a role for cytochrome P450 1A2 in THA metabolism.

Elucidation of the structural requirements for the bioactivation of mianserin in-vitro

The aim of this study was to investigate structure-activity relationships among a series of compounds related to the antidepressant drug, mianserin, with respect to their ability to produce cytotoxic metabolites. Human peripheral lymphocytes were used as target cells and these were exposed to the individual compounds, in the presence or absence of a drug metabolizing system derived from human liver. The individual enantiomers of mianserin showed differences in their cytotoxicity profiles; the R-(-) isomer giving NADPH-dependent cytotoxicity while the S-(+) isomer showed direct cytotoxicity at high concentrations. Cytotoxicity was reduced by removal from mianserin of the nitrogen atom at the 5 position and by substitution of a methyl group for a hydrogen atom at position 14b. In contrast, insertion of an oxygen atom at position 10 of the drug molecule, precluding the formation of a carbonium ion, had little effect on cytotoxic metabolite formation. The data are consistent with the proposal that one or more iminium ions derived from mianserin are responsible for the cytotoxicity observed in this in-vitro system and that appropriate chemical modification may preclude bioactivation of mianserin by P450 enzymes.

The use of a three compartment in vitro model to investigate the role of hepatic drug metabolism in drug-induced blood dyscrasias

1. N-hydroxylation is thought to be an essential step in the haemotoxicity of dapsone (DDS). To investigate both metabolism-dependent and cell-selective drug toxicity in vitro we have developed a three-compartment system in which an hepatic drug metabolizing system is contained within a central compartment separated by semipermeable membranes from compartments containing mononuclear leucocytes (MNL) and red blood cells (RBC). 2. Metabolism of dapsone (100 microM) by rat liver microsomes resulted in toxicity to RBC cells (47.3 +/- 2.1% methaemoglobin), but there was no significant toxicity toward MNL (3.7 +/- 1.3% cell death) compared with control values (1.6 +/- 0.9%). However, when RBC were replaced with buffer in the third compartment there was significantly greater (P < 0.001) white cell toxicity (17.6 +/- 0.6% cell death), demonstrating the protection of MNL by RBC. Metabolism of dapsone by human liver microsomes again resulted in RBC toxicity (12.5 +/- 3.3% methaemoglobin) but no significant MNL toxicity (2.9 +/- 0.8% cell death). Replacement of RBC resulted in a significant (P < 0.001) increase in MNL toxicity (6.5 +/- 0.7% cell death). Addition of synthetic dapsone hydroxylamine (30 microM) in the absence of a metabolizing system and with no RBC in the third compartment resulted in significant (P < 0.001) toxicity toward MNL (43.36 +/- 5.82% cell death) compared with control (1.8 +/- 1.1%). The presence of RBC in the third compartment resulted in a significant (P < 0.001) decrease in MNL toxicity (17.6 +/- 2.2% cell death), with 40.1 +/- 3.7% methaemoglobin in the RBC.(ABSTRACT TRUNCATED AT 250 WORDS)