The genetic control of sparteine and debrisoquine metabolism in man with new methods of analysing bimodal distributions

Sparteine oxidation polymorphism: a family study

Polymorphic oxidation of the pharmacogenetic probe drug sparteine was investigated in 35 parents and 29 siblings of 20 unrelated poor metabolizer (PM) probands. Phenotyping was carried out on the basis of metabolic ratio (MR) = sparteine/dehydrosparteines in the 12 h urine. The distribution of MR was bimodal: 47 relatives (20 siblings and 27 parents) had MR ranging from 0.22-12 and were defined as extensive metabolizers (EM) whereas MR ranged from 20-340 in nine siblings and eight parents thus defined as PM. The 20 pedigrees confirmed that poor metabolism of sparteine is inherited as an autosomal recessive character. The mean recovery of dehydrosparteines (% of dose) was 27% in 23 positively identified drug free heterozygotes compared with 37% in unrelated EM (both genotypes) (P less than 0.05) previously phenotyped. Degrees of dominance of 73% and 77% were calculated on basis of log excretion of dehydrosparteines (% of dose in 12 h urine) and log MR, respectively.

A sensitive capillary GC assay for the determination of sparteine oxidation products in microsomal fractions of human liver

A sensitive method for the assay of sparteine oxidase activity in vitro by microsomal fractions of human liver is described. The activity of sparteine oxidase was assessed by the formation of 2- and 5-dehydrosparteines, which were estimated by capillary gas chromatography with N2-FID detection. The limit of detection of the two metabolites, 2- and 5-dehydrosparteine, was 10 pmol (2.3 ng) per sample. Sparteine oxidase activity was linear with microsomal protein concentration ranging from 25 to 200 ug and with incubation times between 5 and 60 minutes. Omission of NADPH on incubation under an atmosphere of carbon monoxide inhibited formation of both metabolites, thus indicating that aforementioned metabolites arise in reaction catalyzed by cytochrome P-450. In three liver samples from humans classified as extensive (EM) metabolizers the formation of 2- and 5-dehydrosparteines was observed, 2-dehydrosparteine being the major metabolite. In these samples sparteine oxidase activity was characterised by Vmax = 136 +/- 53 pmol/min/mg and Km = 44 +/- 12 microM for 2-dehydrosparteine formation. For 5-dehydrosparteine formation the following values were obtained: Vmax = 57 +/- 18 pmol/min/mg and Km = 42 +/- 26 microM. In a liver sample from a poor metabolizer (PM) only the formation of 2-dehydrosparteine was detected with the method of analysis used. In this sample a great increase in Km (Km PM = 3033 microM) was noted, while Vmax was very similar to those obtained for 2-dehydrosparteine formation in EM subjects (Vmax PM = 147 pmol min/mg).

A simple borohydride/GC method for measuring sparteine metabolites in man

A simple borohydride/GC method was developed for phenotyping sparteine oxidation in man. The major metabolites of sparteine found in human urine, 2- and 5-dehydrosparteine, were converted quantitatively back to sparteine by sodium borohydride reduction. The amount of sparteine metabolites can be estimated from the difference of sparteine concentrations between the borohydride-treated and untreated urine samples. The coefficient of variation of this assay was estimated from repeated analyses to be +/- 3% within a day (intra-assay) and +/- 8% between days (inter-assay).

The genetic polymorphism of sparteine metabolism

The formation of the two major metabolites of the antiarrhythmic and oxytocic drug sparteine (2- and 5-dehydrosparteine) exhibits a genetic polymorphism. Two phenotypes, extensive (EM) and poor metabolizers (PM) are observed in the population. The frequency of the PM phenotype in various populations (Caucasian and Japanese) ranges from 2.3 to 9%. The metabolism of sparteine is determined by two allelic genes at a single gene locus. PM subjects are homozygous for an autosomal recessive gene. The metabolism of sparteine is predominantly under genetic control as treatment with drugs such as antipyrine and rifampicin known to induce oxidative drug metabolism elicited only marginal changes in sparteine metabolism. The formation of 2-dehydrosparteine in human liver microsomes from EM and PM subjects showed a more than 40-fold difference in Km between EM and PM subjects. However, Vmax-values were almost identical in both groups. These data indicate that the basis of the differences in oxidative capacity between EM and PM subjects is more likely to be due to a variant isozyme with defective catalytic properties than to a decreased amount of the isozyme.

Steady-state concentrations of imipramine and its metabolites in relation to the sparteine/debrisoquine polymorphism

Thirty-five imipramine treated patients were phenotyped with regard to polymorphic drug oxidation using sparteine and/or debrisoquine. During treatment with 100 mg imipramine per day the mean steady-state concentrations and ratios in 28 extensive metabolizers were: imipramine 169 nmol/l; desipramine 212 nmol/l; 2-OH-imipramine/imipramine 0.25; 2-OH-desipramine/desipramine 0.57. The corresponding values in two poor metabolizers were: imipramine 455 and 302 nmol/l; desipramine 1148 and 1721 nmol/l; 2-OH-imipramine/imipramine 0.06 and 0.05; 2-OH-desipramine/desipramine: 0.09 and 0.04 respectively. The metabolic ratios (MR) sparteine/dehydrosparteine and debrisoquine/4-OH-debrisoquine (% of dose in 12-h urine samples) correlated poorly with the imipramine steady-state concentrations during administration of 100 mg per day, but quite well with the desipramine steady-state concentrations. Significant negative correlations were found between sparteine and debrisoquine MR and the 2-OH-imipramine/imipramine and 2-OH-desipramine/desipramine ratios. In most patients the initial dose was changed to obtain concentrations in the therapeutic range, and concentrations for imipramine + desipramine of (mean +/- SD) 713 +/- 132 nmol/l were achieved in 33 patients. The therapeutic dose was 50 mg per day in one poor metabolizer and ranged from 50-400 mg per day in 32 extensive metabolizers. There was a weak negative correlation between sparteine MR and daily dose. Treatment with imipramine inhibited metabolism of both sparteine and debrisoquine (MR values about doubled), but did not affect the interpatient correlations.

Sparteine oxidation polymorphism in Denmark

Sparteine oxidation was polymorphic among 301 healthy Danish volunteers. Hence 22 subjects or 7.3% were phenotyped as poor metabolizers (PM) whereas 279 subjects were classified as extensive metabolizers (EM). The metabolic ratio (MR) between sparteine and 2- and 5-dehydrosparteine (% of dose) in 12 hrs urine ranged from 0.11-12.6 in EM and from 30-394 in PM. Urinary excretion of 2- and 5-dehydrosparteine also discriminated between PM and EM. Age, sex, and smoking habits did not influence the MR. This study confirms that sparteine is a useful probe drug in pharmacogenetic investigations.

Lack of congruence of S-carboxymethyl-L-cysteine sulphoxidation and debrisoquine 4-hydroxylation in a Caucasian population

One-hundred-and-twenty volunteers and three families were investigated for possible association between the sulphoxidation of S-carboxymethyl-L-cysteine and the debrisoquine hydroxylation polymorphism. The observed individual variations in these two metabolic reactions were shown not to be concordant (rs = 0.068) and any heritable factors controlling the major aspects of these phenomena do not co-segregate.

Genetically determined variability in acetylation and oxidation. Therapeutic implications

The clinical significance of two separate genetic polymorphisms which alter drug metabolism, acetylation and oxidation is discussed, and methods of phenotyping for both acetylator and polymorphic oxidation status are reviewed. Particular reference is made to the dapsone method, which provides a simple means of distinguishing fast and slow - and possibly intermediate - acetylators, and to the sparteine method which allows a clear separation of oxidation phenotypes. Although acetylation polymorphism has been known for some time, definite indications for phenotyping are few. It is doubtful whether acetylator phenotype makes a significant difference to the outcome in most isoniazid treatment regimens, and peripheral neuropathy from isoniazid in slow acetylators is easily overcome by pyridoxine administration. However, in comparison with rapid acetylators, slow acetylators receiving isoniazid have an increased susceptibility to phenytoin toxicity, and perhaps also to carbamazepine toxicity. It is also possible that rapid acetylators receiving isoniazid attain higher serum fluoride concentrations from enflurane and similar anaesthetics than do similarly treated slow acetylators. Thus, when drug interactions of these types are suspected, phenotyping for acetylator status may be advisable. If routine monitoring of serum procainamide and N-acetylprocainamide concentrations is practised, phenotyping of subjects prior to therapy with these agents should not be necessary. Although acetylator phenotype influences serum concentrations of hydralazine, when this drug is given in combination with other drugs acetylator phenotype has not been shown to influence the therapeutic response. Slow acetylator phenotype along with female gender and the presence of HLA-DR antigens appear to be risk factors in the development of hydralazine-induced systemic lupus erythematosus (SLE). Determination of acetylator phenotype may therefore help determine susceptibility to this adverse reaction. In the case of sulphasalazine, adult slow acetylators require a lower daily dose of the drug than fast acetylators in order to maintain ulcerative colitis in remission without significant side effects. It is therefore advisable to determine acetylator phenotype prior to sulphasalazine therapy. Work on the association of acetylation polymorphism with various disease states is also reviewed. It is possible that a higher incidence of bladder cancer is associated with slow acetylation phenotype - especially in individuals exposed to high levels of arylamines. The question as to whether idiopathic SLE is more common in slow acetylators remains unresolved. There appears to be no difference between fa

Inter-ethnic difference in sparteine oxidation among Ghanaians and Germans

Sparteine oxidation, which exhibits genetic polymorphism in various Caucasian populations, was studied in 154 unrelated Ghanaian subjects. Mean total urinary recovery of sparteine and 2- and 5-dehydrosparteines (56.6 +/- 16.6% of the dose) was not different from Caucasians. In contrast to Caucasians, amongst whom 6.3 to 9% of the population are poor metabolizers (PM), no PM subject was observed in the Ghanaian population sample. The data appear to indicate allelic differences between Caucasians and Ghanaians in the gene encoding the synthesis of the P-450 isozyme involved in polymorphic sparteine oxidation.

Enzymatic activation of chemicals to toxic metabolites

A variety of enzymes function in the oxygenation, oxidation-reduction, conjugation, and hydrolysis of drugs and other foreign chemicals. Often these enzymes detoxicate chemicals to prevent detrimental effects. In this review we will, however, concentrate on cases in which metabolism activates chemicals to reactive species which cause cellular damage. Particular attention will be given to mixed-function oxidases, which carry out a variety of oxygenations, as well as other reactions. (We will focus on cellular toxicity as opposed to initiation of tumorigenesis in this review.) In many cases, considerable circumstantial evidence exists linking these enzymes to enhanced toxicity of chemicals, although causal relationships have seldom been demonstrated. Further, in very few cases is the explicit cause of toxicity known. Modification of critical protein residues is suspected, although oxidative stress may also be involved in some cases. We discuss general aspects of mechanisms of toxic action, briefly list all cases in which metabolism is suspected to play a role in enhancing toxicity, and review a few examples in detail where substantial chemical and enzymatic information is available. The latter instances would involve knowledge of the enzymes involved, chemical evidence on the structures of the reactive metabolites, identification of adducts, and some inference into the biological processes which are effected to elicit toxicity. We consider, in this regard, vinyl halides (which have been a focus in our own laboratory), acetaminophen, pyrrolizidine alkaloids, and fluoroxene.

Metabolic oxidation phenotypes as markers for susceptibility to lung cancer

That bronchial carcinoma is not an inevitable consequence of cigarette smoking has stimulated the search for host factors that might influence the susceptibility of the individual smoker. One plausible host factor would be a polymorphic gene controlling the metabolic oxidative activation of chemical carcinogens, giving rise to wide inter-subject variation in the generation of cancer-inducing and/or promoting species. Recently, three genetic polymorphisms of human metabolic oxidation have been demonstrated (as characterized by debrisoquine, mephenytoin and carbocysteine), with the metabolism of several substrates exhibiting the phenomenon. Debrisoquine 4-hydroxylation segregates into two human phenotypes, each comprising characteristic metabolic capability. We report here the frequency of debrisoquine 4-hydroxylation phenotypes in age-, sex- and smoking history-matched bronchial carcinoma and control patients. Cancer patients showed a preponderance of probable homozygous dominant extensive metabolizers (78.8%) with few recessive poor metabolizers (1.6%) compared with smoking controls (27.8% and 9.0% respectively). We conclude that the gene controlling debrisoquine 4-hydroxylation may be a host genetic determinant of susceptibility to lung cancer in smokers and that it represents a marker to assist in assessing individual risk.

Genetic aspects of the polymodally distributed sulphoxidation of S-carboxymethyl-L-cysteine in man

Interindividual variation in the sulphoxidation of S-carboxymethyl-L-cysteine (750 mg p.o.) was investigated in 200 healthy volunteers. Nearly a 100-fold difference was observed between individuals with respect to the amount of sulphoxide metabolites detected in their 0-8 h urine (0.6 to 59.1% recovery). Such a difference was shown to be reproducible over several months in 40 subjects who spanned the entire range of capacities. Cumulative plots and maximum likelihood analysis of the distribution indicated that a bimodal model was most probable. Analysis of pedigree data obtained from 12 families suggested a genetic effect with overlying environmental influences.

Polymorphic hydroxylation of perhexiline maleate in man

Long term perhexiline maleate therapy causes peripheral neuropathy and hepatic damage in certain subjects. An association between these adverse reactions and a genetically determined relative inability to hydroxylate debrisoquine has been described. This association could indicate either that the effects of perhexiline impair debrisoquine oxidation thus producing a phenocopy, or that perhexiline is polymorphically hydroxylated and that the polymorphism is controlled by the same alleles as control the debrisoquine polymorphism. To test the second possibility, a study investigating the hydroxylation status of a population of healthy volunteer subjects has been performed using perhexiline maleate. Hydroxylation phenotyping was performed on 50 normal volunteers. A standard oral dose was given and plasma and urinary perhexiline, 4-monohydroxyperhexiline (MI metabolite), and 4'monohydroxyperhexiline (MIII metabolite) was measured. The 24-hour plasma perhexiline concentration, the 24-hour plasma MI metabolite concentration, and 12 to 24-hour urinary MI metabolite excretion were clearly bimodal, suggesting the existence of a polymorphism for perhexiline hydroxylation. Poor metabolisers represent 6% of the population studied. Known poor metabolisers of debrisoquine are also poor metabolisers of perhexiline, while known extensive metabolisers of debrisoquine are also extensive metabolisers of perhexiline, indicating that in white British subjects the hydroxylation polymorphism is under identical genetic control for both compounds. The poor metaboliser sub-group exhibited the highest plasma perhexiline levels. Perhexiline phenotyping separates the poor and extensive metaboliser phenotypes much more clearly than other tests and defines a sub-group at risk from perhexiline toxicity. Pretreatment phenotyping using this test, followed by exclusion of poor metabolisers from perhexiline therapy, should substantially reduce the incidence of major adverse effects.