Glutathione s-transferase isoenzymes in relation to their role in detoxification of xenobiotics

The glutathione S-transferases (GST) are a family of isoenzymes serving a major part in the biotransformation of many reactive compounds. The isoenzymes from rat, man and mouse are divided into three classes, alpha, mu and pi, on the basis of similar structural and enzymatic properties.The main function of the glutathione S-transferases isthe catalysisof the conjugation of electrophilic, hydrophobic compounds with the tripeptide glutathione (GSH). In addition, some of the isoenzymes are capable of binding a number of non-substrate ligands non-covalently. The reaction with GSH generally results in the formation of less toxic compounds, however, several cases of activation are also known.Since the individual isoenzymes demonstrate differential though overlapping substrate selectivities, the actual isoenzyme pattern determines the extent to which detoxification and/or activation occurs. Consequently, an individual's susceptibility towards electrophilic compounds is not only dependent on substrate specificity, but also on factors determining the isoenzyme profile of the glutathione S-transferases, including generic factors as well as external factors causing changes in the levels or activities of individual isoenzymes.The studies in this thesis were aimed at the relationship: glutathione S-transferase isoenzyme patterns versus sensitivity towards alkylating agents, and focussed on several aspects of relevance in this context, i.e. substrate selectivity, induction, inhibition and the genetic deficiency of human class mu isoenzymes.In chapter I a review is presented on properties of the glutathione S-transferases. Functions of the enzyme system and the classification of isoenzymes are discussed in section 1.1. Section 1.2 deals with aspects of GST isoenzymes influencing an individual's susceptibility towards alkylating agents, i.e. tissue distribution, developmental patterns with age, hormonal influences, induction and inhibition. Sections 1.3, 1.4 and 1.5 describe the current knowledge on specific properties of class alpha, class mu and class pi isoenzymes respectively.Chapter 2 describes a study on the 9,10-mono-ozonide of methyl linoleate (MLO), a postulated intermediate in the toxicity of ozone, as a substrate for rat glutathione S-transferases. The reaction of MLO with GSH was found to result in the formation of oxidized glutathione and aldehydes as major products, two glutathione-conjugates being formed as unstable intermediates only. The reaction was catalyzed by rat liver cytosol, rat lung cytosol and rat liver microsomes. Relatively high activities were noticed for rat lung cytosol and rat liver microsomes when compared with rat liver cytosol, taking the respective activities towards 1-chloro-2,4-dinitrobenzene (CDNB) into account. Comparison of the specific activities of isoenzymes 1-1, 1-2, 2-2, 3-3, 3-4, 4-4 and 7-7 showed that isoenzyme 2-2 was most active with MLO, whereas 7-7 did not demonstrate detectable activity. The relatively high activity in rat lung cytosol may be partly explained by the fact that isoenzyme 2-2 constitutes a larger percentage of GST protein in rat lung than in rat liver.Although the ozonide was also found to be a substrate for glutathione peroxidase, the activity of this enzyme did not contribute significantly to the cytosolic conversion of MLO.From the results it was concluded that the activity of the glutathione S-transferases towards MLO is similar to their activity with lipid hydroperoxides.In chapter 3, effects of four inducing agents with different chemical structures on rat hepatic isoenzyme patterns were studied. Hexachlorobenzene, benzyl isothiocyanate, phenobarbital and 3-methylcholanthrene all caused an increase In the cytosolic GST activity towards CDNB, 1,2-dichloro-4-nitrobenzene and ethacrynic acid. Changes in the activity towards trans-4-phenyl-3-buten-2-one were relatively small and the activities with cumene hydroperoxide were essentially unchanged. The largest and smallest effects were noticed for hexachlorobenzene and 3- methylcholanthrene, respectively. Microsomal GST activity was not induced by any of the compounds.Isoenzyme patterns obtained by FPLC-chromatofocus ing showed that hexachlorobenzene and phenobarbital both cause an increase in the relative amounts of subunits 1 and 3 as compared with subunits 2 and 4, respectively. For 3-methylcholanthrene only an induction of subunit 1 was observed.Benzyl isothiocyanate differed from the other agents in that this compound did not enhance the relative amount of subunit 1, but caused an increase in subunit 2 instead. Subunit 3 was also induced by benzyl isothiocyanate.Alpha class subunits seemed to be enhanced preferentially: subunits I and 2 represented 53 to 60 % of the GST protein in treated animals, but only 38 to 45 % in controls.The exact mechanism of induction of glutathione S-transferases is unknown, but the differential behaviour of benzyl isothiocyanate suggests that there may be a relationship between chemical structure of the inducing agent and GST subunit induction.Chapter 4 deals with the inhibitory effects of quinones on rat glutathione S-transferases. In section 4.1, inhibition by a series of structurally related 1,4-benzoquinones (BQ) and 1,4- naphthoquinones (NQ) was studied towards a mixture of affinity- purified glutathione S-transferases. The nature of the inhibition was investigated for three quinones with different chemical structures, including 2-tert-butyl-BQ (a metabolite of the food additive 2(3)-tert-butyl-4-hydroxy-anisole), 5-hydroxy-NQ (a naturally occurring quinone) and 2,3-dichloro-NQ (a synthetic compound used as a fungicide). The sensitivities of individual rat hepatic isoenzymes towards these quinones were compared.The inhibitory activity of BQ and NQ was found to increase with an increasing number of electron-withdrawing substituents on the quinone ring, whereas the presence of electron-donating substituents resulted in a decrease in the extent of inhibition as compared with the parent quinone.The inhibition was of an irreversible nature and most likely due to covalent modification of a specific cysteine residue located in or near the active site. The three quinones tested were most inhibitory towards isoenzyme 3-3, isoenzyme 2-2 being least sensitive.Three strongly inhibitory quinones, tetrachloro-BQ, 5-hydroxy-NQ and 2,3-dichloro-NQ, were tested for their inhibitory capacity towards GST activity in a cellular system, using rat l135- hepatoma cells (section 4.2). The GST fraction of these cells mainly contained subunits 4 and 7, with subunits 2 and 3 present as minor constituents. Tetrachloro-BQ and 2,3-dichloro-NQ both inhibited GST activity in rat H35-hepatoma cells in an irreversible manner, 2,3-dichloro-NQ being most efficient. All isoenzymes present were presumably inhibited to some extent. All quinones caused a considerable depletion of cellular GSH-levels.The lack of inhibition noticed for 5-hydroxy-NQ may be explained by two mechanisms. First, this compound could possess a higher tendency for redox cycling. Alternatively, the inhibition may be at least partly mediated by glutathione-conjugat es . In contrast to the chlorinated quinones, conjugation of 5-hydroxy-NQ with GSH results in the formation of a hydroquinone-conjugate, which no longer possesses alkylating properties.Tetrachloro-BQ and 2,3-dichloro-NQ seem attractive starting points for the development of an in vivo inhibitor.In section 5.1 the production of monoclonal antibodies against rat GST isoenzymes 2-2 and 3-3 is described. One hybridoma against isoenzyme 2-2 and 2 hybridomas against isoenzyme 3-3 were capable of specifically differentiating their respective antigens from other rat isoenzymes as well as human isoenzymes in ELISA and on Western blot. Isoenzymes 1-1 and 4-4 did not elicit an immune response in Balb/c mice. However, high serum antibody titers were obtained for these isoenzymes in some other strains of mice, of different H-2 haplotype, notably CBA/BrARij mice and CBA/CaHRij-T6 mice for isoenzyme 1-1, and CBA/BrARij mice also for isoenzyme 4-4 after two injections with the antigen.In section 5.2 mononuclear lymphocytes from the blood of 12 human individuals were screened for the presence or absence of mu-class GST isoenzymes μand/or ψ, using a monoclonal antibody against human hepatic isoenzyme μ. These individuals had worked with a commercial preparation of the soil fumigant 1,3-dichloropropene (DCP), and had been exposed to the vapor of this agent in the field. Nine samples were found to be positive and 3 were negative for the presence of these muclass isoenzymes. In all samples a protein was noticed, staining positively with the anti-μantibody, with a somewhat lower molecular mass than the hepatic standard. This protein presumably constitutes an additional mu-class isoenzyme, since it was bound by the S-hexylglutathione affinity column.The data on the presence or absence of GST isoenzymes μand/or ψwere compared with excretion levels and excretion profiles of the mercapruric acids from cis-(Z-) and trans-(E-)DCP. No significant differences were observed between mu-class positive and mu-class-negative individuals, with respect to the levels and the half-lives of elimination of Z- and E-DCP mercapturic acids. The importance of establishing correlations between the mu-phenotype and mercapturic acid excretion and/or the occurrence of certain diseases is discussed in section 5.2.Most studies described in this thesis have focussed on glutathione S-transferases. from the rat. The recognition of the existence of three classes of isoenzymes common to several mammalian species, including man, and with a high degree of structural homology within the same class, could in principle simplify the extrapolation of data from rat to man. Main obstacles to overcome include the relative lack of knowledge on the substrate selectivity of human isoenzymes, the fact that isoenzymes α- εhave generally been studied as a group and not individually, and the genetic polymorphism of human class mu enzymes for which no equivalent is known in the rat.Studies on the substrate selectivities of individual human isoenzymes would therefore be quite useful. In addition, there is a need for studies on the mechanisms of transcriptional activation of genes coding for transferase subunits in the rat. In the context of induction it is interesting to note that in human liver, alpha class subunit B1 is generally present at much higher levels than subunit B2 1. B1B1 has a higher pI, displays a higher activity with Δ 5-androstenedione and a lower activity with cumene hydroperoxide, than isoenzyme B2B2 2. B1 and B2 may therefore be similar in function to subunits I and 2 respectively. It seems possible that the polymorphism noticed for alpha class subunits might partly reflect exposure to xenobiotics, with a preferential, increase in subunit B1, similar to the increase of subunit 1 in the rat.The development of resistance of tumour cells against chemotherapeutic agents presents a major problem in the treatment of various types of cancer. An increase in the levels of GST isoenzymes has been implicated as one of the possible causes. A selective in vivo inhibitor could possibly overcome this problem and would also be useful in the studies on in vivo metabolism of xenobiotics. The data presented in chapter 4 suggest that quinones may form a suitable starting point. Future studies will have to focus on the mechanism of inhibition, the aspect of selectivity and the structural requirements for the quinone to be capable of causing inhibition in a physiological environment.Evidence is increasing that the genetic deficiency of human class mu isoenzymes is more complicated than previously assumed. Besides the enzymes μand ψ, which are both subject to genetic polymorphism, an additional mu class isoenzyme exists with a slightly lower subunit molecular mass, present in all human individuals. The physiological and toxicological implications of this isoenzyme are unknown and certainly warrant further investigation.The determination of correlations of the genetic deficiency of isoenzymes μand ψwith certain diseases or with excretion profiles of mercapturic acids will be useful in increasing our understanding of the toxicological consequences of this polymorphism, in explaining interindividual differences in mercapturic acid excretion and in establishing safety regulations for the handling of xenobiotics in the work-environment. In this respect, the development of techniques capable of discriminating μfrom ψis of considerable interest.