The use of CBZ and PHT is associated with the development of delayed onset hypersensitivity reactions in 5-10% of the patients. There is a high degree of cross reactivity between these drugs and the clinical symptoms include fever, malaise and skin rashes which may progress to lymphadenopathy, hepatitis and nephritis if the drug is not withdrawn. This clinical picture suggests an autoimmune phenomenon but the mechanism by which these ACDs may activate the immune system to provoke such immunological responses is poorly understood. For most of the characterised drug-induced autoimmune reactions, the drug may undergo metabolic bioactivation generating reactive metabolites which bind to cellular proteins (haptenation). This then triggers immunological reactions directed to the drug-bound proteins (either the hapten or the carrier or the combination), that become modified. To assess whether the pathogenesis of ACD-induced hypersensitivity conformed to
these general concepts, an investigation of some of the possible links between metabolic biotransformation of PHT and the initiation of immunological responses in hypersensitive patients are presented in this thesis. In particular, studies were conducted to characterise the enzymology behind the P450 mediated bioactivation of PHT. It was also of interest to determine pathways and chemical intermediates that may be important in the covalent modification of protein during the P450-mediated bioactivation of PHT. Since drug-induced adverse reactions may result from deficiencies in cellular detoxification of the reactive metabolites, the role of a GSTT1 and/or GSTM1 deletion polymorphism as a factor that may predispose patients to a detoxification deficiency and therefore the ACD-induced hypersensitivity syndrome were investigated. Lastly, studies were conducted to characterise antibody-mediated immune responses towards drug metabolising P450 enzymes in the patients with the ACD-induced
The hydroxylation of PHT is known to be the major metabolic route undergone by this drug in vivo and some drug metabolising P450s including P450s 2C9, 2C19, 3A4, 3A5 and 3A7 have been identified as catalysts for PHT bioactivation as well as targets of covalent adduct formation in vitro. At the commencement of the studies reported here, the role of many other hepatic as well as extrahepatic P450s in the metabolism and/or bioactivation of PHT was not known. Screening of many different human P450s including lAl, 1A2, IBl, 2A6, 2B6, 2E1, 2C8, 2C9, 2C18, 2C19, 3A4, 3A5, 3A7 and 4A11 for the metabolism of PHT confirmed results from previous studies which suggested that the primary hydroxylation of PHT is catalysed by P450s 2C9, 2C18 and P450 2C19 while the secondary hydroxylation is catalysed by P450s 2C8, 2C9, 2C19 and 3A4. For the first time however, studies were carried out to characterise the hydroxylation
of PHT by P450 2C18. Surprisingly, P450 2C18 catalysed the primary hydroxylation of PHT at a rate higher than that of P450s 2C9 and 2C19 by an order of magnitude. P450 2C18 was also found to be catalyst both the secondary hydroxylation of PHT as well as covalent drug-protein adduct formation from PHT and HPPH as substrates in vitro. Lack of expression in the liver and the extrahepatic distribution of P450 2C18 in tissues such as the skin suggest that this enzyme may not influence the pharmacokinetics of PHT but could be important for the extra hepatic tissue-specific bioactivation of PHT in vivo.
The P450-mediated bioactivation of PHT involves a series of oxidation steps that may involve arene oxides and quinone/semiquinones as the reactive metabolites. Many other studies have suggested that an epoxide is the reactive intermediate and the evidence for this was based on the fact that an increase in covalent protein binding
was observed when two mEH inhibitors TCPO and CyO were included in the incubations containing rat liver microsomes, NADPH and [14C]PHT. TCPO and CyO are reactive epoxides that are substrates for mEH and may inhibit other enzymes including P450s. An assessment of whether an epoxide intermediate was involved in covalent protein adduct formation during microsomal oxidation of PHT, required the use of nonepoxide, chemically stable and selective mEH inhibitors. This prompted the screening of novel non-epoxide EH inhibitors for their effect on the catalytic activity of P450s involved in the metabolism of PHT. Many of the mEH screened inhibited one or more of the P450s examined but N-dodecyl-N'-2-bromoethyl-urea, N-dodecyl-N'-2-fluoroethylurea and N-dodecyl-urea were identified as potent and relatively selective P450 2C9 and P450 2C19 competitive inhibitors. Hexadecylamine was identified as the most potent mEH inhibitor with no inhibitory effect on the activity of
P450s 2C9, 2C19, 2C18, 3A4 and 3A5 although it inhibited the hydroxylation of PHT by rat liver microsomes. Many of the sEH examined failed to show any potent effects on the catalytic activity of these P450s but N-cyclohexyl-N'-dodecyl-urea, 11-(3-cyclohexyl-vireido)-undecanoic add, 12- (3-cyclohexyl-ureido)-dodecanoic add, 12-(3-adamantyl-ureido)-dodecanoic acid and adamantyl-dodecyl-urea showed selective activation of the hydroxylation of PHT or the oxidation of indole by P450 2C19. This is the first study to document an effector-mediated potentiation in the catalytic activity of P450 2C19 but the mechanisms by which selective activation of P450 2C19 may occur have not been elucidated. Collectively, these data suggest that substituted ureas may be a useful group of compounds for studying structural features of P450 2C forms.
Since hexadecylamine was identified as a potent, non-epoxide and chemically unreactive mEH inhibitor, experiments were designed to
quantify covalent adduct formation in liver microsomes using [14C]PHT as a substrate in the presence or absence of hexadecylamine. Under the experimental conditions outlined in the current study, no covalent protein adduct formation was detected in human liver microsomes from [14C]PHT in the presence or absence of mEH inhibitors CyO and hexadecylamine. Failure to demonstrate even low levels of adduct formation precluded a proper reassessment of the role of an epoxide as a reactive intermediate responsible for covalent adduct formation. When recombinant P450s 2C19 or 2C18 were used in assessing covalent protein adduct formation from [14C]PHT or [14C]HPPH as substrates, both enzymes were able to catalyse covalent drug-protein adduct formation from [14C]HPPH but only P450 2C18 could catalyse adduct formation from [14C]PHT. The P450 2C18 catalysed
covalent protein adduct formation from [14C]HPPH was significantly higher than that from [14C]PHT. These results are consistent with the fact that P450 2C18 catalysed the primary hydroxylation of PHT more efficiently than P450s 2C9 and 2C19. These results also support the contention that initial hydroxylation of PHT to HPPH is required for bioactivation of PHT and fail to support the hyothesis that an epoxide is the reactive intermediate. Reactive metabolites of PHT and CBZ which include epoxides and quinones/semquinones are likely to be substrates for a wide range of detoxification enzyme systems such as mEH, GSTs, NQO1 and COMT. It is conceivable therefore, that polymorphisms or genetic variations that alter the expression or function of these enzymes may contribute to interindividual differences in the elimination of reactive metabolites thus rendering some individuals susceptible to ACD-induced adverse reactions. Since the role of
mEH, NQO1, COMT and GSTM1 have been examined in previous studies, it was hypothesised that a genetic deletion of polymorphic GSTT1 or both GSTT1 and GSTM1 may be a risk factor in the development of ACD-induced hypersensitivity. Using a PCR-based approach, eight hypersensitive patients and twenty-seven control subjects were genotyped for GSTT1 and GSTM1. Using a one-tailed Fisher's exact test, the proportion of patients with the GSTT1 null genotype was found to be significantly higher (P<0.05) than that of control subjects. There were no significant differences in the frequency of the null GSTM1 or a double (GSTT1 and GSTM1) null genotype in patients and control subjects. These results are consistent with the hypothesis that patients were predisposed to ACD-induced hypersensitivity due to a deficiency in GSTT1 with or without a further deficiency in GSTM1.
Having identified individual P450 enzymes that are involved in the bioactivation of PHT, it was
also of interest to determine whether some of these enzymes are targets of autoimmune reactions in their native forms or following modification by reactive metabolites of PHT in vitro. ELISA was used for the detection of immunoreactivity of sera from hypersensitive patients or control subjects towards P450s 2C9, 2C19,3A4,3A5 and TXAS. For all the native or drug modified proteins tested, there was a low but detectable immimoreactivity with no significant differences in the patients and controls group. This may suggest that conformational epitopes are not an important feature of the ACD-induced hypersensitivity syndrome or presence of other determinants of immunoreactivity that were not investigated in the current study. It is also likely that ACD-induced hypersensitivity in the group of patients examined here may involve mechanisms other than the humoral antibody mediated responses.