Characterization of the cytochrome P450 isoenzymes involved in the in vitro N-dealkylation of haloperidol

AIMS: The present study was carried out to identify the cytochrome P450 isoenzyme(s) involved in the N-dealkylation of haloperidol (HAL). METHODS: In vitro studies were performed using human liver microsomes and c-DNA-expressed human P450 isoforms. N-dealkylation of HAL was assessed by measuring the formation of 4-(4-chlorophenyl)-4-hydroxypiperidine (CPHP). RESULTS: There was a tenfold variation in the extent of CPHP formation amongst the nine human liver microsomal preparations. The CPHP formation rates as a function of substrate concentration, measured in three livers, followed monophasic enzyme kinetics. Km and Vmax values ranged respectively from 50 to 78 microM and from 180 to 412 pmol mg-1 min-1 CPHP formation rates in the nine liver preparations were significantly correlated with dextromethorphan N-demethylase activity (a marker of CYP3A4 activity), but not with the activity of dextromethorphan O-demethylase (CYP2D6), phenacetin O-deethylase (CYP1A2) or tolbutamide hydroxylase (CYP2C9). Ketoconazole, an inhibitor of CYP3A4, inhibited competitively CPHP formation (Ki=0.1 microM), whereas sulphaphenazole (CYP2C9), furafylline (CYP1A2) and quinidine (CYP2D6) gave only little inhibition (IC50 > 100 microM). CPHP formation was, moreover, enhanced by apha-naphtoflavone, an effect common to CYP3A4 mediated reactions. Anti-CYP3A4 antibodies strongly inhibited CPHP formation, whereas no inhibition was observed in the presence of CYP2D6 antibodies. Among the recombinant human CYP isoforms tested, CYP3A4 exhibited the highest activity with respect to CPHP formation rate, with no detectable effect of other CYP isoforms (CYP1A2, CYP2D6 and CYP2C9). HAL inhibited dextromethorphan O-demethylase (CYP2D6) with IC50 values between 2.7 and 8.5 microM, but not (IC50 > 100 microM) dextromethorphan N-demethylase (CYP3A4), phenacetin O-deethylase (CYP1A2) or tolbutamide hydroxylase (CYP2C9). CONCLUSIONS: These results strongly suggest that the N-dealkylation of HAL in human liver microsomal preparations is mediated by CYP3A4.     

Effects of propofol on human hepatic microsomal cytochrome P450 activities

1. The potential of propofol to inhibit the activity of major human cytochrome P450 enzymes has been examined in vitro using human liver microsomes. Propofol produced inhibition of CYP1A2 (phenacetin O-deethylation), CYP2C9 (tolbutamide 4'-hydroxylation), CYP2D6 (dextromethorphan O-demethylation) and CYP3A4 (testosterone 6beta-hydroxylation) activities with IC50 = 40, 49, 213 and 32 microM respectively. Ki for propofol against all of these enzymes with the exception of CYP2D6, where propofol showed little inhibitory activity, was 30, 30 and 19 microM respectively for CYPs 1A2, 2C9 and 3A4. 2. Furafylline, sulphaphenazole, quinidine and ketoconazole, known selective inhibitors of CYPs 1A2, 2C9, 2D6 and 3A4 respectively, were much more potent than propofol having IC50 = 0.8, 0.5, 0.2 and 0.1 microM; furafylline and sulphaphenazole yielded Ki = 0.6 and 0.7 microM respectively. 3. The therapeutic blood concentration of propofol (20 microM; 3-4 microg/ml) together with the in vitro Ki estimates for each of the major human P450 enzymes have been used to estimate the extent of cytochrome P450 inhibition, which may be produced in vivo by propofol. This in vitro-in vivo extrapolation indicates that the degree of inhibition of CYP1A2, 2C9 and 3A4 activity which could theoretically be produced in vivo by propofol is relatively low (40-51%); this is considered unlikely to have any pronounced clinical significance. 4. Although propofol has now been used in > 190 million people since its launch in 1986, there are only single reports of possible drug interactions between propofol and either alfentanil or warfarin. Consequently, it is difficult to conclude from either the published literature or the ZENECA safety database whether there is any evidence to indicate that propofol produces clinically significant drug interactions through inhibition of cytochrome P450-related drug metabolism.     

Identification of cytochrome P450s in human glioma cell line

Five different isoforms of cytochrome P450 including 1A1, 1A2, 2E1, 2A and 2B6 have been identified in human glioma Hs 683 cell line using RT-PCR reaction. These isoforms belong to four distinct subfamilies. The effect of benzanthracene (Ba) as inducer was tested on the mRNA level of cytochrome P450 1A1. Northern blot analysis clearly showed an induction response from these cells to Ba in a proportion that is comparable to the induction seen in rat glioma cells.     

Hypothalamic aromatase cytochrome P450 and 5alpha-reductase enzyme activities in pregnant and female rats

The metabolism of steroid hormones in the medial basal hypothalamus (MBH) is known to play a critical role in neural development, the modulation of neuroendocrine function and regulating sexual behavior. While the important biological functions of the aromatase enzyme are well established, the importance of brain 5alpha-reductase has been revealed and elucidated only in the last few years. The distribution and regulation of brain aromatase and 5alpha-reductase enzyme activities have been investigated for the most part in male rats. Therefore, in the present study, MBH aromatase cytochrome P450 and 5alpha-reductase activities were characterized in pregnant and female rats during postnatal development under various hormonal conditions. MBH aromatase activity was determined in each tissue sample using the 'tritiated water release' assay, whereas, the 5alpha-reductase rates were determined by thin layer chromatography and scintillation counting of the isolated 5alpha-metabolites. Both activities were highest in infantile animals, then declined with increasing postnatal age; whereas, in aged non-cycling or ovariectomized/adrenalectomized (Ovx/Adx) rats high rates of androgen metabolism were seen in MBH tissue. No significant alterations in MBH aromatase were observed when the 5alpha-reductase pathway was blocked in pregnant animals during late gestation with a known 5alpha-reductase inhibitor (Proscar). However, plasma estradiol levels were significantly increased in the Proscar-treated animals. These results indicate that: 1) the decreasing MBH aromatase and 5alpha-reductase profile (in infantile to adult cycling animals) is developmentally regulated, 2) evidently, there is a divergent regulatory mechanism controlling MBH aromatase versus 5alpha-reductase in aged animals where the aromatase activity increased in aged non-cycling and Ovx/Adx rats while 5alpha-reductase rates remained at moderate levels and, 3) apparently, the 5alpha-reductase pathway is not involved in regulating MBH aromatase activity during late pregnancy.     

Inhibition of murine hepatic cytochrome P450 activities by natural and synthetic phenolic compounds

1. The effect of the phenolic compounds protocatechuic acid, chlorogenic acid, tannic acid, gallates and silybin on ethoxyresorufin O-dealkylase (CYP1A1), methoxyresorufin O-dealkylase (CYP1A2) and pentoxy-O-dealkylase (CYP2B) was examined in mouse liver microsomes from induced animals. 2. All compounds tested could inhibit cytochrome P450-mediated enzyme activities, but to different extents. Tannic acid was the most potent inhibitor, especially toward EROD activity with an IC50=2.6 microM. Synthetic dodecyl gallate was also relatively selective toward this enzyme activity with an IC50=120 microM. 3. Protocatechuic acid, chlorogenic and silybin were more selective towards PROD and MROD activities. Their relative inhibitory potency for PROD activity was as follows: chlorogenic acid > protocatechuic acid > silybin > dodecyl gallate > propyl gallate. Protocatechuic acid was a more effective inhibitor of MROD activity than chlorogenic acid, and propyl gallate more effective than dodecyl gallate. Thus, no clear structure-activity or selectivity relationship was observed. 4. Analysis of the kinetics of inhibition revealed that the inhibition in most cases was non-competitive in nature.     

Avian forms of cytochrome P450

Despite the importance of avian P450 forms in modulation of the toxicity of pesticides and other environmental chemicals, relatively little work has been done upon them, and very few forms have been fully characterised. An avian form that appears to belong to family 1A is readily inducible by planar molecules (e.g. coplanar PCB's, PCDD and certain PAHs) and has been the basis of a biomarker assay used in field studies. Although it is recognised by antibodies for mammalian P450 1A1, it evidently differs from the mammalian forms of the enzyme in catalytic properties. Phenobarbitone induces two forms of P450 in the domestic fowl (2H1 and 2H2) which have been purified, and these resemble P450 2B1 and P450 2B2 of the rat respectively. Two further phenobarbitone inducible forms, PB-A and PB-B have been partially purified. Also there is an acetone inducible form that resembles rodent P450 2E. In field studies evidence has been produced for the induction of P450s recognised by antibodies to mammalian forms of P450 1A1 and P450 2B in avian liver (adults and embryos), in response to environmental levels of PCBs. Fungicides which act as ergosterol biosynthesis inhibitors (EBI fungicides) such as prochloraz and propiconazole potentiate the toxicity of certain phosphorothionates to birds.     

Novobiocin inhibits both UDP-glucuronosyltransferase and cytochrome P450-mediated enzyme activities in pig liver microsomes

The effects of novobiocin (range 0.0125-2 mmol/L) on the hydroxylation of testosterone, the N-demethylation of erythromycin, and the glucuronidation of alpha-naphthol and paracetamol were studied using pig hepatic microsomes, pooled from five animals. The final concentrations of these substrates in the incubation mixtures were selected to meet Vmax conditions. Novobiocin caused a concentration-dependent inhibition of the glucuronidation of paracetamol; the formation of alpha-naphthol-glucuronide was reduced to a lesser degree. These results confirm and extend earlier findings in laboratory animal species that novobiocin inhibits UDP-glucuronosyltransferases (UDPGTs). Moreover, novobiocin strongly inhibited 6 beta-hydroxylation of testosterone. The microsomal N-demethylation of erythromycin and hydroxylation of testosterone at the 15 alpha position were less affected by novobiocin. These results suggest that novobiocin inhibits not only UDPGTs, but also cytochrome P450 (CYP) enzyme activities, probably those belonging to the CYP3A subfamily. More research is needed to reveal which CYPs and UDPGTs are affected by novobiocin in vivo, in order to improve the understanding the probably the predictability of potential drug interactions with this antibiotic.     

Physiological and pathophysiological regulation of cytochrome P450

This article is a report on a symposium sponsored by the American Society for Pharmacology and Experimental Therapeutics and held at the April 1998 Experimental Biology '98 meeting in San Francisco. The presentations focused on the mechanisms of regulation of cytochrome P450 gene expression by developmental factors and by hormones and cytokines, as well as on the interplay between physiological and chemical regulation. Approaches and systems used to address these questions included conditional gene knockouts in mice, primary hepatocyte cultures, immunofluorescence imaging of cells, and cell lines stably expressing reporter gene constructs.     

Semisynthetic flavocytochromes based on cytochrome P450 2B4: reductase and oxygenase activities

Endurance exercise training induces a rapid increase in the GLUT-4 isoform of the glucose transporter in muscle. In fasted rats, insulin-stimulated muscle glucose transport is increased in proportion to the increase in GLUT-4. There is evidence that high muscle glycogen may decrease insulin-stimulated glucose transport. This study was undertaken to determine whether glycogen supercompensation interferes with the increase in glucose transport associated with an exercise-induced increase in GLUT-4. Rats were trained by means of swimming for 6 h/day for 2 days. Rats fasted overnight after the last exercise bout had an approximately twofold increase in epitrochlearis muscle GLUT-4 and an associated approximately twofold increase in maximally insulin-stimulated glucose transport activity. Epitrochlearis muscles of rats fed rodent chow after exercise were glycogen supercompensated (86.4 +/- 4.8 micromol/g wet wt) and showed no significant increase in maximally insulin-stimulated glucose transport above the sedentary control value despite an approximately twofold increase in GLUT-4. Fasting resulted in higher basal muscle glucose transport rates in both sedentary and trained rats but did not significantly increase maximally insulin-stimulated transport in the sedentary group. We conclude that carbohydrate feeding that results in muscle glycogen supercompensation prevents the increase in maximally insulin-stimulated glucose transport associated with an exercise training-induced increase in muscle GLUT-4.     

Specificity of self-assembly of cytochrome P450

Under certain conditions, hexamers of microsomal cytochrome P450 can self-assemble from the subunits of different isoforms. However, the possibility for free choice results in recognition between identical subunits of each form of cytochrome P450 which provides preferential association of identical monomers into corresponding hexamers. The specificity of self-assembly suggests hexameric arrangement of cytochrome P450 in native membranes as we proposed earlier. In the present study, highly purified cytochrome P450 2B4 and cytochrome P450 1A2 (CYP 2B4 and CYP 1A2), including those immobilized by covalent attachment to an insoluble carrier of one protomer of each hexamer, were employed.     

Flavodoxin and NADPH-flavodoxin reductase from Escherichia coli support bovine cytochrome P450c17 hydroxylase activities

Two soluble flavoproteins, purified from Escherichia coli cytosol and identified as flavodoxin and NADPH-flavodoxin (ferredoxin) reductase (flavodoxin reductase), have been found in combination to support the 17 alpha-hydroxylase activities of heterologously expressed bovine 17 alpha-hydroxylase cytochrome P450 (P450c17). Physical characteristics of the two flavoproteins including absorbance spectra, molecular weights, and amino-terminal sequences are identical with those reported previously for E. coli flavodoxin and flavodoxin reductase. Flavodoxin reductase, possessing FAD as a cofactor, is able to reconstitute P450c17 activities only in the presence of flavodoxin, an FMN-containing protein, and NAD(P)H. Reducing equivalents are utilized more effectively from NADPH than NADH by flavodoxin reductase. E. coli flavodoxin binds P450c17 directly and with relatively high affinity (apparent Ks approximately 0.2 microM) at low ionic strength, as evidenced by a change in spin state of the P450c17 heme iron upon titration with flavodoxin. This apparent spin shift is attenuated at moderate ionic strengths (100-200 mM KCl). In addition, bovine P450c17 binds reversibly to flavodoxin Sepharose in an ionic strength-dependent manner. These data implicate charge pairing as being important for the interaction between flavodoxin and P450c17. We propose that the amino acid sequence similarity between E. coli flavodoxin-flavodoxin reductase and the putative FMN, FAD, and NAD(P)H binding regions of cytochrome P450 reductase provides the basis for the reconstitution of P450c17 activities by this bacterial system.     

Design of a novel P450: a functional bacterial-human cytochrome P450 chimera

We report the construction of a functional chimera from approximately 50% bacterial (cytosolic) cytochrome P450cam and 50% mammalian (membrane-bound) cytochrome P450 2C9. The chimeric protein shows a reduced CO-difference spectrum absorption at 446 nm, and circular dichroism spectra indicate that the protein is globular. The protein is soluble and catalyzes the oxidation of 4-chlorotoluene using molecular oxygen and reducing equivalents from bacterial putidaredoxin and putidaredoxin reductase. This chimera provides a novel method for addressing structure-function issues and may prove useful in the design of oxidants for benign and stereospecific synthesis, as well as catalysts for bioremediation of polluted areas. Furthermore, these results provide the first evidence that bacterial P450 enzymes and mammalian P450 enzymes are likely to share a common tertiary structure.     

Enhanced expression of cytochrome P450 in stomach cancer

The cytochromes P450 have a central role in the oxidative activation and detoxification of a wide range of xenobiotics, including many carcinogens and several anti-cancer drugs. Thus the cytochrome P450 enzyme system has important roles in both tumour development and influencing the response of tumours to chemotherapy. Stomach cancer is one of the commonest tumours of the alimentary tract and environmental factors, including dietary factors, have been implicated in the development of this tumour. This type of tumour has a poor prognosis and responds poorly to current therapies. In this study, the presence and cellular localization of several major forms of P450, CYP1A, CYP2E1 and CYP3A have been investigated in stomach cancer and compared with their expression in normal stomach. There was enhanced expression of CYP1A and CYP3A in stomach cancer with CYP1A present in 51% and CYP3A present in 28% of cases. In contrast, no P450 was identified in normal stomach. The presence of CYP1A and CYP3A in stomach cancer provides further evidence for the enhanced expression of specific forms of cytochrome P450 in tumours and may be important therapeutically for the development of anti-cancer drugs that are activated by these forms of P450.     

Interactions of substrate and product with cytochrome P450: P4502B4 versus P450cam

OBJECTIVE: To determine the relative abilities of somatostatin receptor scintigraphy (SRS) and conventional imaging studies (computed tomography, magnetic resonance imaging, ultrasound, angiography) to localize gastrinomas before surgery in patients with Zollinger-Ellison syndrome (ZES) subsequently found at surgery, and to determine the effect of SRS on the disease-free rate. SUMMARY BACKGROUND DATA: Recent studies demonstrate that SRS is the most sensitive imaging modality for localizing neuroendocrine tumors such as gastrinomas. Because of conflicting results in small series, it is unclear in ZES whether SRS will alter the disease-free rate, which gastrinomas are not detected, what factors contribute to failure to detect a gastrinoma, or whether the SRS result should be used to determine operability in patients without hepatic metastases, as recently recommended by some investigators. METHODS: Thirty-five consecutive patients with ZES undergoing 37 exploratory laparotomies for possible cure were prospectively studied. All had SRS and conventional imaging studies before surgery. Imaging results were determined by an independent investigator depending on surgical findings. All patients underwent an identical surgical protocol (palpation after an extensive Kocher maneuver, ultrasound during surgery, duodenal transillumination, and 3 cm duodenotomy) and postoperative assessment of disease status (fasting gastrin, secretin test imaging within 2 weeks, at 3 to 6 months, and yearly), as used in pre-SRS studies previously. RESULTS: Gastrinomas were detected in all patients at each surgery. Seventy-four gastrinomas were found: 22 duodenal, 8 pancreatic, 3 primaries in other sites, and 41 lymph node metastases. The relative detection order on a per-patient or per-lesion basis was SRS > angiography, magnetic resonance imaging, computed tomography > ultrasound. On a per-lesion basis, SRS had greater sensitivity than all conventional studies combined. SRS missed one third of all lesions found at surgery. SRS detected 30% of gastrinomas < or =1.1 cm, 64% of those 1.1 to 2 cm, and 96% of those >2 cm and missed primarily small duodenal tumors. Tumor size correlated closely with SRS rate of detection. SRS did not increase the disease-free rate immediately after surgery or at 2 years mean follow-up. CONCLUSIONS: SRS is the most sensitive preoperative imaging study for extrahepatic gastrinomas in patients with ZES and should replace conventional imaging studies as the preoperative study of choice. Negative results of SRS for localizing extrahepatic gastrinomas should not be used to decide operability, because a surgical procedure will detect 33% more gastrinomas than SRS. SRS does not increase the disease-free rate. In the future, more sensitive methods to detect small gastrinomas, especially in the duodenum and in periduodenal lymph nodes, or more extensive surgery will be needed to improve the postoperative disease-free rate in ZES.     

Escherichia coli expression and substrate specificities of canine cytochrome P450 3A12 and rabbit cytochrome P450 3A6

High level Escherichia coli expression of cytochromes P450 3A12 and 3A6 has facilitated the characterization of proteins which exhibit limited activity as purified hepatic enzymes in reconstituted systems. Three 3A12 and two 3A6 constructs modified at the 5'-end to encode the bovine 17 alpha-sequence (Barnes et al., Proc. Natl. Acad. Sci. U.S.A. 88: 5597-5601, 1991), or related sequences, exhibited expression levels ranging from 2 to 89 nmol of cytochrome P450 liter-1. Recombinant canine 3A12 catalyzed steroid 6 beta-hydroxylation and erythromycin demethylation at rates comparable to those obtained in phenobarbital-induced canine liver microsomes. In contrast, 3A12 troleandomycin demethylase activity (2.5 nmol/min/nmol) was significantly lower than that of canine phenobarbital-induced liver microsomes (6.6 nmol/min/nmol). This difference in activity suggests that at least two 3A forms, which may differ functionally, are present within the canine liver. Purification of recombinant rabbit 3A6 revealed that homogeneous and E. coli-solubilized membrane preparations of 3A6 exhibit similar metabolic rates and identical substrate specificities; 3A activity was modulated by 25 microM alpha-naphthoflavone, which stimulated an unidentified progesterone metabolite 9-fold in 3A6 reconstituted systems in contrast to the 4-fold stimulation of 3A12. Furthermore, 25 microM alpha-naphthoflavone inhibited erythromycin demethylation 64 and 33% by purified recombinant 3A6- or 3A6-solubilized membrane fractions, respectively; 3A12-mediated erythromycin demethylation in solubilized membrane fractions was resistant to flavonoid inhibition. These results indicate that, although 3A substrate specificities are highly conserved between species, functional differences exist between canine 3A12 and rabbit 3A6, which may be utilized to better understand 3A structure-function relationships.     

Induction of Ad4BP/SF-1, steroidogenic acute regulatory protein, and cytochrome P450scc enzyme system expression in newly established human granulosa cell lines

We have established immortalized human granulosa cells by triple transfection of primary cells obtained from in vitro fertilization patients with SV40 DNA, Ha-ras oncogene, and a temperature sensitive (ts) mutant of the tumor suppressor gene p53 (p53val135). Forty-one clones were isolated, and their steroidogenic responses were analyzed. While all the cell lines proliferate rapidly and show only traces of progesterone production, upon stimulation with 50 microM of forskolin (FK), which elevates intracellular cAMP, they become steroidogenic as evidenced by progesterone production. The steroidogenic response of the cell lines was stable even after 20 generations and several cycles of freezing and thawing. A highly responsive cell line (HO-23) was further examined for characteristics of the steroidogenic response. Cells stimulated with FK and 8-Br-cAMP produced high levels of pregnenolone, progesterone, and 20alpha-hydroxy-4-pregnen-3-one (20alpha-OH-progesterone) comparable with amounts produced by highly differentiated primary human granulosa-luteal cells. Hydrocortisone and dexamethasone highly augment the cAMP-stimulated progesterone production, whereas testosterone and PRL enhanced cAMP-induced progesterone synthesis only moderately. Estradiol, insulin-like growth factor I, and insulin showed no significant effect on cAMP-induced steroidogenesis. The phorbol ester TPA, and basic fibroblast growth factor, dramatically suppress cAMP-induced production of progesterone, whereas bovine corneal endothelial cell ECM (BCE/ECM) enhanced cAMP-induced progesterone and antagonized basic fibroblast growth factor suppression of cAMP-induced steroidogenesis. Steroidogenic factor 1 (Ad4BP/SF-1) was expressed in control cells, and its expression was augmented by FK, whereas the steroidogenic acute regulatory protein showed low expression in the nonstimulated cells but was clearly elevated upon cAMP stimulation and was slightly decreased by TPA in cAMP-stimulated cells. Expression of the electron carrier adrenodoxin (ADX), which is a part of the cytochrome P450scc enzyme system, was very low in nonstimulated cells but was dramatically elevated in FK- and 8-Br-cAMP-stimulated cells, whereas no reduction of ADX was evident in cells costimulated with FK and TPA. Immunocytochemical studies revealed a weak staining of ADX in mitochondria of nonstimulated cells and intensive staining in highly clustered mitochondria of FK- or 8-Br-cAMP-stimulated cells. Only moderate reduction in ADX staining was evident in cells costimulated with FK and TPA. These unique cell lines can provide a useful model for the investigation of induced steroidogenesis in human granulosa cells.     

The substrate specificity of cytochrome P450cam

The catalytic turnover of xenobiotics by cytochrome P450cam results in both the formation of organic metabolites and the uncoupled production of H2O2, and H2O. Previous studies have shown that a receptor-constrained three-dimensional screening program (DOCK) can be used to identify potential ligands (ergo substrates) for the enzyme (De Voss, J. J.; Sibbesen, O.; Zhang, Z.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1997, 119, 5489). A new set of 10 compounds has now been examined to further test the substrate specificity of P450cam and the ability of DOCK to identify substrates for this enzyme. The results expand the known specificity of P450cam and define limitations in the use of DOCK to predict its substrate specificity.