Inhibition of NADPH-Cytochrome P450 Reductaseand Glyceryl Trinitrate Biotransformation by
Diphenyleneiodonium SulfateJohn J. McGuire,* Diane J. Anderson,* Bernard J. McDonald,*
Ramani Narayanasami and Brian M. Bennett**DEPARTMENT OF PHARMACOLOGY AND TOXICOLOGY, QUEENS UNIVERSITY, KINGSTON, ONTARIO,
CANADA K7L 3N6; AND THE UNIVERSITY OF TEXAS HEALTH SCIENCE CENTER AT SAN ANTONIO,SAN ANTONIO, TX 78284-7760, U.S.A.
ABSTRACT. We reported previously that the flavoprotein inhibitor diphenyleneiodonium sulfate (DPI)irreversibly inhibited the metabolic activation of glyceryl trinitrate (GTN) in isolated aorta, possibly throughinhibition of vascular NADPH-cytochrome P450 reductase (CPR). We report that the content of CPRrepresents 0.03 to 0.1% of aortic microsomal protein and that DPI caused a concentration- and time-dependentinhibition of purified cDNA-expressed rat liver CPR and of aortic and hepatic microsomal NADPH-cytochromec reductase activity. Purified CPR incubated with NADPH and GTN under anaerobic, but not aerobicconditions formed the GTN metabolites glyceryl-1,3-dinitrate (1,3-GDN) and glyceryl-1,2-dinitrate (1,2-GDN). GTN biotransformation by purified CPR and by aortic and hepatic microsomes was inhibited . 90%after treatment with DPI and NADPH. DPI treatment also inhibited the production of activators of guanylylcyclase formed by hepatic microsomes. We also tested the effect of DPI on the hemodynamic-pharmacokineticproperties of GTN in conscious rats. Pretreatment with DPI (2 mg/kg) significantly inhibited the blood pressurelowering effect of GTN and inhibited the initial appearance of 1,2-GDN (15 min) and the clearance of1,3-GDN. These data suggest that the rapid initial formation of 1,2-GDN is related to mechanism-based GTNbiotransformation and to enzyme systems sensitive to DPI inhibition. We conclude that vascular CPR is a siteof action for the inhibition by DPI of the metabolic activation of GTN, and that vascular CPR is a novel siteof GTN biotransformation that should be considered when investigating the mechanism of GTN action invascular tissue. BIOCHEM PHARMACOL 56;7:881893, 1998. 1998 Elsevier Science Inc.
KEY WORDS. biotransformation; diphenyleneiodonium; flavoproteins; glyceryl trinitrate; NADPH-cyto-chrome P450 reductase; vascular smooth muscle
Biotransformation of GTN has been shown to be medi-ated by several proteins , including reduced unligandedhemoproteins , vascular glutathione S-transferases [3, 4],and the cytochromes P450-NADPH-cytochrome P450 re-ductase system . There is also an unidentified micro-somal protein isolated from bovine coronary artery thatmediates NO formation from GTN . In addition to theformation of the denitrated metabolites 1,2-GDN and1,3-GDN, both NO2
2 and NO may be formed during thebiotransformation reaction. However, the low vasodilatorpotency of NO2
2 indicates that it is of little importance inmediating the pharmacological actions of organic nitrates.
Hepatic and aortic microsomal biotransformation of
GTN to 1,3-GDN and 1,2-GDN has been shown to beNADPH dependent, inhibited by cytochrome P450 inhib-itors, such as carbon monoxide and SKF 525A, and in-creased in microsomes from phenobarbital-treated rats [6,7]. Furthermore, rat hepatic microsomal biotransformationof GTN results in the formation of an activator of guanylylcyclase (presumably NO) . Whether the vascular cyto-chromes P450-NADPH-cytochrome P450 reductase systemmediates bioactivation of GTN remains controversial,since classical inhibitors of cytochrome P450 such as SKF525A , cimetidine , and metyrapone  have noeffect on GTN-induced vasodilation. In contrast, the cyto-chrome P450 substrate 7-ethoxyresorufin inhibits GTN-induced relaxation, cyclic GMP accumulation and aorticbiotransformation of GTN .
Because NADPH-cytochrome P450 reductase is essentialfor cytochrome P450 activity, inhibition of this flavopro-tein might be expected to inhibit the pharmacologicalactions of GTN. DPI is known to inhibit a number offlavoproteins including the NADPH oxidase system ofneutrophils and macrophages [13, 14], xanthine oxidase
Corresponding author. Tel. (613) 533-6473; FAX (613) 533-6412;E-mail: firstname.lastname@example.org
Abbreviations: DPI, diphenyleneiodonium sulfate; FMN, flavin mono-nucleotide; FMO, flavin-containing monooxygenase; 1,3-GDN, glyceryl-1,3-dinitrate; 1,2-GDN, glyceryl-1,2-dinitrate; GTN, glyceryl trinitrate;IDP, iodonium diphenyl; MAP, mean arterial pressure; and NO, nitricoxide.
Received 9 July 1997; accepted 26 May 1998.
Biochemical Pharmacology, Vol. 56, pp. 881893, 1998. ISSN 0006-2952/98/$19.00 1 0.00 1998 Elsevier Science Inc. All rights reserved. PII S0006-2952(98)00216-0
, and macrophage NO synthase . IDP, an analog ofDPI, is an irreversible inhibitor of bovine liver NADPH-cytochrome P450 reductase . In a previous study, wefound that DPI irreversibly inhibits GTN-induced relax-ation and cyclic GMP accumulation in isolated rat aortaand the selective biotransformation of GTN to 1,2-GDN. DPI also inhibits the enantioselectivity of isoididedinitrate action by selectively inhibiting biotransformationof, and cyclic GMP accumulation and relaxation inducedby, the more potent D-enantiomer . In addition, DPIhas no effect on GTN or isoidide dinitrate biotransforma-tion in aortic cytosol, suggesting that the site of action ofDPI for the inhibition of organic nitrate biotransformationis in the microsomal fraction of vascular tissue . Weproposed that DPI may be reduced by NADPH-cytochromeP450 reductase to form a radical that then forms an adductwith either the heme (cytochrome P450), the flavin cofac-tors of NADPH-cytochrome P450 reductase, or aminoacids of either cytochrome P450 or NADPH-cytochromeP450 reductase .
In the current study, we used kinetic analysis to assesswhether NADPH-cytochrome P450 reductase activity wasinhibited by DPI through an irreversible mechanism, andtested the effect of DPI on mitochondrial/microsomalNADH-cytochrome b5 reductase and the flavin-containingNADH-ubiquinone oxidoreductase (Complex 1), a compo-nent of the electron transport chain in mitochondria. Wealso examined the effect of DPI on GTN biotransformationby aortic and hepatic microsomes, and by purified NADPH-cytochrome P450 reductase. To assess the in vivo signifi-cance of DPI inhibition of GTN action, we tested the effectof DPI on hemodynamic and pharmacokinetic properties ofGTN in conscious rats.
MATERIALS AND METHODSMaterials
DPI was purchased from Colour Your Enzyme. Stocksolutions of DPI (1 mM) were prepared in distilled water forin vitro studies and in 5% dextrose solution for in vivostudies. GTN was obtained as a solution (TRIDILt,5 mg/mL) in ethanol, propylene glycol, and water (1:1:1.33) from DuPont Pharmaceuticals. 1,2-GDN and 1,3-GDN were prepared by acid hydrolysis and purified bythin-layer chromatography . The purified 1,2-GDN and1,3-GDN were quantified by the method of Dean and Baun as modified by Bennett et al. . Isosorbide-2-mono-nitrate was a gift from Wyeth Ltd. b-NADPH, b-NADH,rotenone, and bovine heart cytochrome c were purchasedfrom the Sigma Chemical Co. All other chemicals were ofat least reagent grade and were obtained from a variety ofsources.
Microsomal Protein Preparation
The microsomal fractions from rat aorta and liver wereprepared by standard differential centrifugation techniques.
Liver microsomal protein was prepared from individualmale SpragueDawley rats (275325 g) . Aortic micro-somes were prepared from the pooled aortae of 2040 rats. Aortic microsomes (105,000 g pellet) were resus-pended in 50 mM potassium phosphate buffer (pH 7.7)containing 0.1 mM EDTA. Protein concentrations weredetermined by the method of Lowry et al. .
The mitochondrial fraction from rat aorta and liver wasprepared using the isolation method of Kukielka et al. with the modification that livers from individual rats wereperfused with ice-cold 10 mM TrisHCl buffer (pH 7.4)containing 0.25 M sucrose and 1 mM EDTA prior tohomogenization (10%, w/v) in the same buffer. Mitochon-dria from rat aorta were obtained from a modified micro-somal preparation using 30 rats (see above). The modifica-tions were the use of the microsomal homogenization buffer and the inclusion of the 800 g centrifugation step.
Identification of NADPH-Cytochrome P450 Reductase
Purified cDNA-expressed rat liver NADPH-cytochromeP450 reductase (a gift from Dr. B. S. S. Masters, TheUniversity of Texas Health Science Center at San Anto-nio) [25, 26] and aortic microsomal and/or hepatic micro-somal proteins were separated by electrophoresis on 10%SDS polyacrylamide gels. Proteins were transferred electro-phoretically to PVDF membranes, and incubated overnightwith an anti-rat liver NADPH-cytochrome P450 reductaseantibody (also a gift from Dr. B. S. S. Masters); immuno-reactive bands were visualized using enhanced chemilumi-nescence.
NADPH-Cytochrome P450 Reductase Activity
NADPH-cytochrome P450 reductase activity of purifiedenzyme and microsomal fractions was measured by follow-ing the NADPH-dependent reduction of bovine heartcytochrome c . Samples (1 mL) contained 50 mMpotassium phosphate buffer (pH 7.7), 0.1 mM EDTA, 100mM NADPH, 36 mM cytochrome c and either 2 nMpurified NADPH-cytochrome P450 reductase, 110 mg/mLof hepatic microsomal protein, or 50 mg/mL of aorticmicrosomal protein. KCN (1 mM) was added to samplescontaining microsomal protein to inhibit possible contam-ination by mitochondrial cytochrome c oxidase. Reactionswere initiated by the addition of protein, and A550 wasmeasured using an HP 8451A diode array spectrophotom-eter. Reaction rates were calculated using the extinctioncoefficient for cytochrome c of 0.021 mM21 cm21. Inexperiments examining the effect of DPI on NADPH-cytochrome P450 reductase activity, protein was preincu-bated at 25 with 100 mM NADPH and DPI (0.03 to 10mM) for various times (15 sec to 5 min), prior to determin-ing cytochrome c reductase activity.
882 J. J. McGuire et al.
NADH-Cytochrome c Reductase Activity
The activity of NADH-cytochrome b5 reductase in intactmitochondria or microsomes was determined by measuringrotenone-insensitive NADH-cytochrome c reduction .The activity of NADH-ubiquinone oxidoreductase wasdetermined by measuring rotenone-sensitive NADH-cyto-chrome c reduction in mitochondria. Samples (1 mL)contained 0.2 M TrisHCl buffer (pH 7.4), 50 mM cyto-chrome c, 1 mM KCN, 100 mM NADH, 1030 mg ofmitochondrial protein (or 0.1 mg of microsomal protein)and either rotenone (5 mM) or DPI (10 mM, 1/10 dilutionof the preincubation sample concentration). Control sam-ples (minus rotenone) contained 2.5% (v/v) ethanol (ro-tenone vehicle). In experiments examining the effect ofDPI, protein was preincubated with 100 mM NADH and100 mM DPI (100 mL vol.) for 15 min prior to determiningcytochrome c reductase activity. Since NADH-ubiquinoneoxidoreductase activity is associated with the inner mem-brane of mitochondria, to measure exogenous cytochrome creduction by this enzyme component first required perme-abilizing the mitochondria to NADH and cytochrome c.Intact mitochondria were treated with 0.033% TritonX-100/mg of mitochondrial protein, which was found toincrease rotenone-sensitive activity without a loss of totalNADH-cytochrome c reductase activity.
Kinetic Analysis of Enzyme Inactivation
We have based the kinetic analysis of enzyme inactivationon the following enzyme reaction equation :
Ki k3E 1 I ^ EzI 3 E z I* (1)
where E represents active enzyme, I represents inhibitor, E zI represents the reversible enzymeinhibitor complex, E z I*represents the irreversible enzymeinhibitor complex, Ki isthe apparent equilibrium constant for formation of E z I, andk3 is the rate of conversion of E z I to E z I*. To determinethe values for apparent Ki and k3, we used the method ofKitz and Wilson  to solve the equation:
1/kinactivation 5 1/k3 1 ~Ki/k3! z ~1/@I#!. (2)
GTN Biotransformation Studies
Purified expressed rat liver NADPH-cytochrome P450reductase (5 mg/mL of protein; 60 nM NADPH-cyto-chrome P450 reductase) was incubated with 0.2 mM GTNand 1 mM NADPH in 50 mM potassium phosphate buffer(pH 7.7) containing 0.1 mM EDTA. Incubations wereperformed in an Instrumentation Laboratory 237 tonometerat 37 under either aerobic (normal atmospheric oxygen) oranaerobic (humidified N2 at a flow rate of 250 mL/min)conditions. An aliquot (1 mL) was taken immediately afterthe addition of GTN and then after 1 hr of incubation.Extraction of GTN and the GTN metabolites 1,2-GDN
and 1,3-GDN and quantitation by megabore capillarycolumn gas liquid chromatography were performed as de-scribed . In experiments examining inhibition of GTNbiotransformation by DPI, aortic microsomes (0.5 mg/mL ofprotein), hepatic microsomes (0.9 mg/mL of protein), orpurified rat liver NADPH-cytochrome P450 reductase (60nM) were treated with 100 mM NADPH and either 10 mMDPI or diluent for 10 min (hepatic microsomes and purifiedreductase) or 30 min (aortic microsomes) prior to theaddition of 0.2 mM GTN and 1 mM NADPH. Hepaticmicrosomal biotransformation of GTN was determinedover a 10-min time course. GTN biotransformation byaortic microsomes and by purified rat liver NADPH-cytochrome P450 reductase was assessed after a 1-hr incu-bation with GTN. Nonenzymatic denitration of GTN wasdetermined by incubating samples in the absence of pro-tein. NADH-dependent GTN biotransformation was as-sessed using the same procedures as NADPH-dependentGTN biotransformation under both aerobic and anaerobicconditions. Samples (1 mL) containing liver mitochondria(0.6 mg/mL of protein), aortic mitochondria (0.1 mg/mL ofprotein), or liver microsomes (0.6 mg/mL of protein) weretreated with NADH (1 mM) and GTN (0.2 mM) for 10min (liver proteins) or 60 min (aortic protein). In experi-ments to determine the effect of DPI on GTN metabolism,samples were preincubated with NADH (100 mM) andeither DPI (10 mM) or diluent prior to initiating thereactions with GTN and NADH.
Guanylyl Cyclase Activity
The preparations of crude aortic guanylyl cyclase andenzyme assay conditions have been described . In addi-tion to the usual assay components, samples containedGTN (1 mM to 0.3 mM), DPI (10 mM), and either 1 mMNADPH or 1 mM NADH. Reactions were initiated by theaddition of 1823 mg of aortic supernatant protein and2130 mg of hepatic microsomal protein (20 pmol cyto-chrome P450) or 2030 mg of mitochondrial protein.Reactions were terminated by the addition of 0.8 mL of 50mM sodium acetate (pH 4.0) followed by heating at 90 for3 min. Aliquots were assayed for cyclic GMP content byradioimmunoassay . Aortic supernatant protein wasdetermined by the method of Bradford . CytochromeP450 content of hepatic microsomes was measured by themethod of Omura and Sato .
Surgical Preparation and Instrumentation of Rats
Male SpragueDawley rats (275325 g) were anaesthetizedwith ketamine (Rogarsetict, 70...