Polymorphic drug oxidation in humans

Genetic polymorphisms in the oxidative metabolism of debrisoquine, mephenytoin, phenformin, sparteine, and tolbutamide have been discovered during recent years. Among these pharmacogenetic conditions, polymorphic oxidation of debrisoquine and sparteine has been intensively studied. Two phenotypes, the extensive (EM) and the poor (PM) metabolizers, have been observed in all populations so far investigated. The PM phenotype exhibits a grossly impaired or nearly absent capacity to metabolize these drugs. The incidence of the PM phenotype in European populations ranges from 5 to 9%. Pronounced variations in the incidence of the PM phenotype have been demonstrated among different ethnic groups. The metabolism of debrisoquine and sparteine is determined by two alleles at a single gene locus; PMs are homozygous for an autosomal recessive gene. Because of markedly impaired metabolism, the PM phenotype develops side effects if normal doses of debrisoquine and sparteine are administered. Defective metabolism in the PM phenotype is not restricted to debrisoquine and sparteine. Impaired metabolism of guanoxan , phenformin, perhexiline, methoxyamphetamine, phenacetin, encainide, metoprolol, alprenolol, bufuralol, nortriptyline, and desipramine have been described. As a consequence of impaired metabolism of these drugs, toxicity and therapeutic failure are observed in the PMs. With regard to molecular mechanisms, studies with microsomes from human liver provide evidence that in the PM phenotype a cytochrome P-450 isozyme is either missing or functionally inadequate.     

Gas chromatography/mass spectrometry assays for the determination of debrisoquine and sparteine metabolites in microsomal fractions of rat liver.

Debrisoquine and sparteine are prototype substrates of a genetic deficiency in cytochrome P450-dependent drug metabolism. Sensitive assays of in vitro oxidation of sparteine and debrisoquine are required for evaluation of this polymorphism. The activities were measured by quantitative analysis of 2-dehydrosparteine and 4-hydroxydebrisoquine production, respectively, using capillary column gas chromatography coupled with mass selective ion detection. With a single extraction, separation of parent drug, metabolite, and a suitable internal standard was readily achievable. Time-dependent production of both metabolites could be detected from as little as 40 micrograms of microsomal protein. Both activities showed a maximal activity with a 240-min incubation period. The ability to simultaneously quantify the parent drug and its metabolite suggests it would also be useful for evaluation of in vivo metabolism.     

In vivo administration of H2 blockers,cimetidine and ranitidine,reduced the contents of the cytochrome P450IID (CYP2D) subfamily and their activities in rat liver microsomes

• 1.1. The effects of in vivo administration of H₂ blockers, cimetidine and ranitidine (0.6 mmol/kg body weight/day, for 5 days), on several P450 isozymes, the P450IID (CYP2D) subfamily, and their monooxygenase activities in rat liver microsomes were investigated.
• 2.2. In vivo administration of cimetidine and ranitidine decreased the contents of P450 isozymes and the activities of P450-linked monooxygenase systems; i.e., benzphetamine N-demethylase, aminopyrine N-demethylase, 7-ethoxycoumarine O-deethylase, debrisoquine 4-hydroxylase and bufuralol 1'-hydroxylase.
• 3.3. The inhibitory effect on the enzymatic activities of the P450IID (CYP2D)-linked monooxygenase systems was studied by Western blot analysis with serum containing antiCYP2D6 IgG, i.e., LKM1 autoantibody. The amount of P450IID (CYP2D) in liver microsomes decreased more remarkably in the group administered ranitidine or cimetidine in vivo than in controls.
• 4.4. The effects of cimetidine and ranitidine on the activities of the P450IID (CYP2D)-linked monooxygenase systems were investigated in vitro. The activities of debrisoquine 4-hydroxylase and bufuralol 1'-hydroxylase were inhibited in vitro by cimetidine or ranitidine at a higher concentration than that on in vivo administration of either H₂ blocker.
• 5.5. The kinetic parameters for cimetidine or ranitidine as to the activities of debrisoquine 4-hydroxylase and bufuralol 1'-hydroxylase in liver microsomes were determined by means of Lineweaver-Burk plots.
• 6.6. The suppressive effects of cimetidine and ranitidine on the activities of the P450IID (CYP2D)-linked monooxygenase systems in vivo were found to be due to a decrease of the content of the P450IID (CYP2D) protein.

     

The dopamine transporter and cytochrome P45OIID1 (debrisoquine 4-hydroxylase) in brain: resolution and identification of two distinct [3H]GBR-12935 binding proteins

Two [3H]GBR-12935 binding proteins, identified as the dopamine transporter and cytochrome P45OIID1, were solubilized in digitonin from canine striatal membranes, and were resolved following wheat germ agglutinin (WGA)-lectin column chromatography. Protein adsorbed to and specifically eluted from WGA-lectin with N-acetylglucosamine displayed saturable, high affinity (KD approximately 3 nM), and sodium-dependent binding of [3H]GBR-12935, which was inhibited in a concentration-dependent and stereoselective manner by dopamine uptake blockers and substrates with a pharmacological profile indicative of the dopamine uptake site. Protein not adsorbed to WGA-lectin also bound [3H]-GBR-12935 with high affinity (approximately 7 nM), in a sodium-independent manner, and was insensitive to classical dopamine uptake blockers and substrates such as mazindol or dopamine, corresponding to the so-called "piperazine acceptor" site seen in native membranes. [3H]GBR-12935 binding to this latter protein was, however, inhibited by various compounds with a pharmacological profile indicative of a form of cytochrome P450 designated P45OIID1 (debrisoquine/sparteine monooxygenase) with the following rank order of inhibitory potency: GBR-12909 greater than budipine greater than alpha-lobeline greater than quinidine greater than alpha flupenthixol greater than SKF-525A greater than sparteine greater than quinine. Ki values obtained for inhibition of [3H]-GBR-12935 binding to neuronal WGA passthrough fractions by these drugs correlate well with their respective Ki values for liver P45OIID1 activity. Western blotting and immunoprecipitation analysis with rabbit anti-rat P45OIID1 antibody also supported the identity of the mazindol-insensitive [3H]GBR-12935 binding site (or piperazine acceptor site) as P45OIID1. Furthermore, a [3H]GBR-12935 binding protein with pharmacological and immunological characteristics similar to those of P45OIID1 was solubilized from both bovine and human liver membranes, and GBR-12909 was found to be a potent competitive inhibitor (Ki approximately 100 nM) of sparteine monooxygenase activity in human liver microsomes. These data clearly indicate that [3H]GBR-12935 and its analogs display similar affinities for both the dopamine transporter and neuronal P45OIID1, and that this radioligand may be a useful probe of P45OIID1 activity in brain and liver. The exact molecular and functional association (if any) between these two distinct binding protein populations remains to be established; however, it is tempting to speculate that P45OIID1 is involved in the catabolism and processing of neurotransmitters subsequent to their reuptake into target cells.     

Deletion of the entire cytochrome P450 CYP2D6 gene as a cause of impaired drug metabolism in poor metabolizers of the debrisoquine/sparteine polymorphism.

The debrisoquine/sparteine polymorphism is associated with a clinically important genetic deficiency of oxidative drug metabolism. From 5% to 10% of Caucasians designated as poor metabolizers (PMs) of the debrisoquine/sparteine polymorphism have a severely impaired capacity to metabolize more than 25 therapeutically used drugs. The impaired drug metabolism in PMs is due to the absence of cytochrome P450IID6 protein. The gene controlling the P450IID6 protein, CYP2D6, is located on the long arm of chromosome 22. A pseudogene CYP2D8P and a related gene CYP2D7 are located upstream from CYP2D6. This gene locus is highly polymorphic. After digestion of genomic DNA with XbaI endonuclease, restriction fragments of 11.5 kb and 44 kb represent mutant alleles of the cytochrome CYP2D6 gene locus associated with the PM phenotype. In order to elucidate the molecular mechanism of the mutant allele reflected by the XbaI 11.5-kb fragment, a genomic library was constructed from leukocyte DNA of one individual homozygous for this fragment and screened with the human IID6 cDNA. The CYP2D genes were isolated and characterized by restriction mapping and partial sequencing. We demonstrate that the mutant 11.5-kb allele results from a deletion involving the entire functional CYP2D6 gene. This result provides an explanation for the total absence of P450IID6 protein in the liver of these PMs.     

High-performance liquid chromatographic assays for bufuralol 1'-hydroxylase, debrisoquine 4-hydroxylase, and dextromethorphan O-demethylase in microsomes and purified cytochrome P-450 isozymes of human liver

Bufuralol, debrisoquine, and dextromethorphan are three prototype substrates of the common genetic deficiency of oxidative drug metabolism in man known as debrisoquine/sparteine-type polymorphism. We describe assays for the in vitro metabolism of (+)- and (-)-bufuralol, debrisoquine, and dextromethorphan in human liver microsomes and reconstituted purified cytochrome P-450 isozymes. These assays combine nonextractive sample preparation by precipitation of protein with perchloric acid with reversed-phase inorganic ion-pair HPLC and fluorescence detection. The minimal detectable levels of the major metabolites formed are 1'-hydroxybufuralol, 0.1 ng/ml; 4-hydroxydebrisoquine, 0.8 ng/ml; and dextrorphan, 0.1 ng/ml. Formation of these metabolites is linear for at least 45 min and between 1 and 100 micrograms of microsomal protein. Comparative kinetic analysis of the three monooxygenase reactions in human liver microsomes revealed an apparent biphasicity of (+)- and (-)-bufuralol 1'-hydroxylation and dextromethorphan O-demethylation but monophasic formation of 4-hydroxydebrisoquine in the substrate concentration range (less than 1 mM) studied. These data, in combination with those obtained by purified human cytochrome P-450 isozymes indicate the involvement of the same enzyme in the metabolism of all three substrates investigated. However, additional and distinct activities contribute to the metabolism of (+)- and (-)-bufuralol and dextromethorphan.     

Purification and characterization of the human liver cytochromes P-450 involved in debrisoquine 4-hydroxylation and phenacetin O-deethylation, two prototypes for genetic polymorphism in oxidative drug metabolism

Two forms of cytochrome P-450 were purified to apparent homogeneity from several different preparations of human liver microsomes. One form, designated P-450DB, had relatively high catalytic activity towards the drugs debrisoquine, sparteine, bufuralol (both the (+)- and (-)-isomers), encainide, and propranolol and appears to be the enzyme involved in the polymorphic distribution of oxidative activities towards these substrates in humans. The other form, designated P-450PA, had relatively high phenacetin O-deethylase activity and appears to be involved in the variation of this activity among humans. Polyclonal antibodies raised to the two enzymes were specific for the antigens as judged by immunoelectrophoresis and immuno-inhibition studies. The two enzymes and their activities were distinguished by chromatographic separation, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, amino acid composition, immuno-inhibition studies, and steady-state kinetic assays. Immunochemical studies suggest that each form represents only a small fraction of the total cytochrome P-450 in human liver microsomes. These biochemical studies provide a basis for better understanding the mechanisms which underlie genetic polymorphisms involving P-450 cytochromes in humans.     

SKF 525-A and cytochrome P-450 ligands inhibit with high affinity the binding of [3H]dextromethorphan and sigma ligands to guinea pig brain.

The DM1/sigma 1 site binds dextromethorphan (DM) and sigma receptor ligands. The broad binding specificity of this site and its peculiar subcellular distribution prompted us to explore the possibility that this site is a member of the cytochrome P-450 superfamily of enzymes. We tested the effects of the liver microsomal monooxygenase inhibitor SKF 525-A (Proadifen), and other P-450 substrates on the binding of [3H]dextromethorphan, [3H]3-(-3-Hydroxyphenyl)-N-(1-propyl)piperidine and (+)-[3H]1,3-Di-o-tolyl-guanidine ([3H]DTG) to the guinea pig brain. SKF 525-A, l-lobeline and GBR-12909 inhibited the binding of the three labeled ligands with nM affinity. Each drug has identical nM Ki values for the high-affinity site labeled by the three ligands. This indicated that they displaced the labeled ligands from the common DM1/sigma 1 site. Debrisoquine and sparteine, prototypical substrates for liver debrisoquine 4-hydroxylase, displayed Ki values of 9-13 and 3-4 microM respectively against the three labeled ligands. These results, the broad specificity of the DM1/sigma 1 binding site, and its peculiar subcellular distribution, raises the possibility that this binding site is a member of the cytochrome P-450 superfamily of isozymes, rather than a neurotransmitter receptor. These findings may have important implications for the understanding of the therapeutic, side effects and toxicity of several neurotropic drugs.     

Debrisoquine/sparteine hydroxylation genotype and phenotype: analysis of common mutations and alleles of CYP2D6 in a European population.

Four different mutations of the cytochrome P450 CYP2D6 gene associated with the poor metabolizer phenotype (PM) of the debrisoquine/sparteine polymorphism were analyzed by Xba I restriction fragment length polymorphism (RFLP) analysis and a polymerase chain reaction (PCR)-based DNA amplification method in DNA of 394 healthy European subjects; 341 of these were phenotyped by sparteine or debrisoquine administration and urinary metabolic ratios (MR). Our study demonstrates the efficiency of the PCR-test for phenotype prediction; 96.4% of individuals were correctly predicted, i.e., 100% of the extensive metabolizers (EMs) and 86.0% of the poor metabolizers (PMs). In contrast, Xba I RFLP analysis was far less informative, predicting the phenotype in only 26.8% of PMs. By combining both DNA tests, the prediction rate of the PM phenotype increased to 90.6%. A point mutation at a splice-site consensus sequence termed D6-B represented the most common mutant CYP2D6 gene and accounted for more than 75% of mutant alleles. In addition, other known mutations such as D6-D (14%), D6-A (5%), and the rare D6-C mutation bring the identified mutant alleles to greater than 95% of all mutant PM-alleles. Most of Xba I 44-kb alleles were confirmed as mutant alleles carrying the D6-B mutation. However, 9.7% did not have this mutation and may express a functional CYP2D6 gene. Moreover, all Xba I 16 + 9-kb alleles contained the D6-B mutation. Heterozygous EM individuals had a significantly higher MR when compared to homozygous EMs. Genotyping provides an important advantage for investigations of the influence of CYP2D6 activity on drug therapy and its association with certain diseases.     

Absence of hepatic cytochrome P450bufI causes genetically deficient debrisoquine oxidation in man

The common genetic deficiency of drug oxidation known as debrisoquine/sparteine-type polymorphism was investigated with bufuralol as prototype substrate. In human liver microsomes the 1'-hydroxylation of bufuralol is catalyzed by two functionally distinct P-450 isozymes, the high-affinity/highly stereoselective P450bufI and the low-affinity/nonstereoselective P450bufII. We demonstrate that P450bufI is unique in hydroxylating bufuralol in a cumene hydroperoxide (CuOOH) mediated reaction whereas P450bufII is active only in the classical NADPH- and O2-supported monooxygenation. In microsomes of liver biopsies of in vivo phenotyped poor metabolizers of debrisoquine or sparteine, the CuOOH-mediated activity was drastically reduced. Rabbit antibodies against a rat P-450 isozyme with high bufuralol 1'-hydroxylase activity (P450db1) precipitated exclusively P450bufI-type activity from solubilized microsomes. Western blotting of microsomes with these antibodies revealed a close correlation between the immunoreactive protein and CuOOH-mediated (+)-bufuralol 1'-hydroxylation. No immunoreactive protein was detected in liver microsomes of in vivo phenotyped poor metabolizers. These data provide evidence for a specific deficiency of P450bufI and are consistent with the complete or almost complete absence of this protein in the liver of poor metabolizers.     

Human debrisoquine 4-hydroxylase (P450IID1): cDNA and deduced amino acid sequence and assignment of the CYP2D locus to chromosome 22

The enzyme P450db1 (db1) is responsible for the common human defect in drug oxidation known as the "debrisoquine/sparteine polymorphism." Polyclonal antibody against the rat db1 protein was used to screen a human liver lambda gt11 library for the db1 cDNA clone. A cDNA containing the full protein coding sequence was isolated; the deduced NH2-terminal sequence of this cDNA was identical to that derived from direct sequencing of the purified human db1 protein. Comparison of the human db1 with rat db1 revealed 71 and 73% similarities of nucleotides and amino acids, respectively. By use of human-rodent somatic cell hybrids the db1 gene was localized to human chromosome 22 (CYP2D locus).     

Decreased placental monooxygenase activities associated with birth defects

We have measured monooxygenase activities in placentas from 82 women who smoked throughout their pregnancies and correlated these with the presence or absence of major somatic anomalies. Monooxygenase activities toward benzo(a)pyrene and ethoxyresorufin in placentas from 18 abnormal infants were compared with activities in placentas from 64 concurrently studied normal infants. Placentas from normal infants were found to have high levels of monooxygenase activities and low apparent Kms toward ethoxyresorufin (10(-7) M), reflecting induction of cytochrome P-450 enzymes usually associated with maternal cigarette smoking. Placentas from the abnormal infants, however, had significantly lower monooxygenase activities and higher apparent Kms toward ethoxyresorufin (10(-5) M), indicating that induction of specific cytochrome P-450 systems occurred less frequently among placentas from abnormal infants. The reasons for this association are unclear. Apparent lack of induction of monooxygenase activity occurred most frequently in placentas from anencephalic infants but was neither exclusively nor consistently found with this defect. No specific maternal condition or environmental exposure associated with lack of monooxygenase induction was identified.     

DNA haplotype dependency of debrisoquine 4-hydroxylase (CYP2D6) expression among extensive metabolisers

Deficient debrisoquine/sparteine type oxidation is inherited as an autosomal recessive trait. Of all Caucasians, 5–10% are poor metabolisers, due to the absence of cytochrome P4502D6. Extensive metabolisers (EMs) exhibit highly variable metabolic activity. We investigated the relationship between CYP2D6 activity and genotypes of the CYP2D locus in a large set of French Caucasian families. Genotypes concern both common mutations affecting the enzyme activity and linkedBamHI polymorphisms of the locus. We found, like other authors, that in EMs part of the heterogeneity is explained by a subgroup of individuals heterozygous for a mutant allele. However, a second level of heterogeneity was detected among individuals not carrying mutations, and this was related to a polymorphicBamHI-defined DNA haplotype. Different combinations of haplotypes are associated with differences in CYP2D6 metabolic activity. This finding might help to clarify the conflicting data on the relation between CYP2D6 activity and susceptibility to lung cancer.     

Effects of Coumarin and Sparteine on Attachment to Cellulose and Cellulolysis by Neocallimastix frontalis RE1

The plant secondary metabolites coumarin and sparteine reduced attachment to cellulose, cellulose solubilization, and the proportion of lactate in the fermentation products of the anaerobic fungus Neocallimastix frontalis RE1. Neither compound directly inhibited the endoglucanase or lactate dehydrogenase activities of cell extracts of the fungus.     

Spectrophotometric assay of the flavin-containing monooxygenase and changes in its activity in female mouse liver with nutritional and diurnal conditions

A highly sensitive spectrophotometric assay was developed for measuring flavin-containing monooxygenase activity using methimazole (N-methyl-2-mercaptoimidazole) as the substrate. With the procedure described, flavin-containing monooxygenase activity can be accurately measured in whole cell homogenates without interference due to NADPH oxidase activities. The effects of detergents and octylamine on female mouse liver flavin-containing monooxygenase activity were characterized for whole homogenates and microsomes prepared under conditions which tend to cause or minimize microsomal aggregation. A small activation was observed with 0.2% (v/v) Emulgen 913 with nonaggregated microsomes; higher levels of detergents gave maximal activity with aggregated microsomes. Variations in the activity of the female mouse liver enzyme with nutritional state and time of day were evaluated. Higher specific activities were observed in homogenates and microsomes of livers from fed animals than from livers of 24-h starved animals, and higher specific activities were present in samples from livers of animals sacrificed in late afternoon than in the early morning. In the period where activity increased in fed animals (i.e., the AM to PM transition), a portion of flavin-containing monooxygenase was more resistant to thermal inactivation. Other properties are described which suggest structural differences for at least a portion of the flavin-containing monooxygenase. The possibility that these differences may be related to turnover of the flavin-containing monooxygenase is discussed.     

Hepatic biotransformation in the buzzard (Buteo buteo) and the Japanese quail (Coturnix coturnix): effect of PCBs

Hepatic biotransformation was studied in microsomal (cytochrome P-450-dependent monooxygenase activities) and cytosolic (glutathione S-transferase activities) fractions from Japanese quail (Coturnix coturnix) and buzzard (Buteo buteo). Monooxygenase activities were not very different apart from a high 7-ethoxycoumarin de-ethylase activity in quail as compared to buzzard. Glutathione S-transferase activities were higher in quail than in buzzard. DP5 (a commercial mixture of PCBs containing 50% chlorine) produced a marked increase in monooxygenase activity from quail liver. In contrast, no induction was found in buzzard under the same conditions. Glutathione S-transferase activities were not modified in both species.     

Effect of sparteine sulfate on insulin secretion in normal men

This study aimed at evaluating the influence of sparteine sulfate either upon basal plasma glucose and insulin or glucose-induced insulin secretion in normal man. Thirteen overnight fasted volunteers took part in this study; five of them were submitted to sparteine sulfate bolus (15 mg in 10 ml of saline solution) followed by a slow infusion (90 mg/100 ml X 60 min) and eight subjects underwent two different glucose pulses (20 gr. i.v.) in absence or in presence of sparteine, infused as described above. In basal conditions, along with sparteine infusion, plasma glucose showed a progressive and significant decrease (P less than 0.0001) and plasma insulin was significantly higher from min 10 to 120' (P less than 0.0005-0.001). Even during the glucose-induced insulin secretion, in the presence of sparteine infusion, plasma glucose levels were significantly lower while plasma insulin levels were significantly higher when compared to those observed after glucose alone. The acute insulin response (AIR) was 42 +/- 10 microU/ml after glucose alone vs 67 +/- 9 microU/ml after glucose plus sparteine (P less than 0.05). Total insulinemic areas were significantly different being 1410 +/- 190 vs 2250 +/- 310 microU/ml/min (P less than 0.001) during glucose and glucose plus sparteine infusion, respectively. This study thereby, demonstrates that in normal man sparteine sulfate, administrated by intravenous infusion, is able to increase either basal or glucose-induced insulin secretion.     

Interaction of alkaloids with plant transfer ribonucleic acids. Effect of sparteine on lupin arginyl-tRNA formation

The effect of the alkaloid sparteine on arginyl-tRNA formation was studied. It was demonstrated that sparteine sulfate in the concentration range 10–60 mM inhibits the charging reaction when amino acid, ATP and tRNA are used as variable substrates. The mode of action is different for all substrates studied. It was concluded that at high sparteine concentration the pattern of inhibition for all varied substrates is generally uncompetitive. A non-competitive mechanism for amino acid and tRNA was observed at low sparteine concentration, but in the case of ATP it is also uncompetitive.     

Inhibition of sparteine oxidation in human liver by tricyclic antidepressants and other drugs

Testing for competitive inhibition of sparteine oxidation in the 9000 × g supernatant fraction from human liver provides an in vitro means to identify drugs which can bind to the same form of cytochrome P450 which oxidizes sparteine. There has so far been only two outcomes of this test: either the drug examined competed with sparteine for a common binding site, or it did not inhibit the reaction. The results of such in vitro testing implicated the involvement of guanoxan, nortriptyline, desipramine, imipramine, amitriptyline and chlorpromazine with this enzyme. Amobarbital, tolbutamide and guanethidine in therapeutic concentrations did not interfere with sparteine oxidation by this preparation.     

Fluorescence assay for per-cell estimation of cytochrome P-450-dependent monooxygenase activities in keratinocyte suspensions and cultures

An assay was characterized that facilitated per-cell estimation of cytochrome P-450-dependent monooxygenase activities in whole-cell suspensions and cultures of murine epidermal keratinocytes (MEKs). 7-Ethoxycoumarin O-deethylase (7-ECD), 7-ethoxyresorufin O-deethylase (7-ERD), and 7-pentoxyresorufin O-deethylase (7-PRD) activities were monitored by fluorescent detection of their products. MEKs were made permeable by a freeze-thaw cycle, and xenobiotic metabolism occurred in situ. Analyses of cultured MEKs were made with the cells attached to the culture dishes. Product formation was proportional with MEK cell number and assay time and was dependent upon a NADPH-generating system. The three monooxygenase activities were inhibited to various degrees, in a dose-dependent manner, by the P-450 inhibitors alpha-naphthoflavone and metyrapone. The number of MEKs obtained from a single mouse was sufficient for multiple analyses. The assay was also used to determine monooxygenase activities in whole-cell suspensions of rat hepatocytes. Constitutive per hepatocyte 7-ECD, 7-PRD, and 7-ERD activities were 357-, 96-, and 1926-fold greater, respectively, than the activities measured in suspensions of dorsal MEKs.