Enantioselective inhibition of the biotransformation and pharmacological actions of isoidide dinitrate by diphenyleneiodonium sulphate

Abstract

1. We have shown previously that the D- and L- enantiomers of isoidide dinitrate (D-IIDN and L-IIDN) exhibit a potency difference for relaxation and cyclic GMP accumulation in isolated rat aorta and that this is related to preferential biotransformation of the more potent enantiomer (D-IIDN). The objective of the current study was to examine the effect of the flavoprotein inhibitor, diphenyleneiodonium sulphate (DPI), on the enantioselectivity of IIDN action.

2. In isolated rat aortic strip preparations, exposure to 0.3 μM DPI resulted in a 3.6 fold increase in the EC₅₀ value for D-IIDN-induced relaxation, but had no effect on L-IIDN-induced relaxation.

3. Incubation of aortic strips with 2 μM D- or L-IIDN for 5 min resulted in significantly more D-isoidide mononitrate formed (5.0±1.5 pmol mg  protein⁻¹) than L-isoidide mononitrate (2.1±0.7 pmol mg protein⁻¹) and this difference was abolished by pretreatment of tissues with 0.3 μM DPI. DPI had no effect on glutathione S-transferase (GST) activity or GSH-dependent biotransformation of D- or L-IIDN in the 105,000×g supernatant fraction of rat aorta.

4. Consistent with both the relaxation and biotransformation data, treatment of tissues with 0.3 μM DPI significantly inhibited D-IIDN-induced cyclic GMP accumulation, but had no effect on L-IIDN-induced cyclic GMP accumulation.

5. In the intact animal, 2 mg kg⁻¹ DPI significantly inhibited the pharmacokinetic and haemodynamic properties of D-IIDN, but had no effect L-IIDN.

6. These data suggest that the basis for the potency difference for relaxation by the two enantiomers is preferential biotransformation of D-IIDN to NO, by an enzyme that is inhibited by DPI. Given that DPI binds to and inhibits NADPH-cytochrome P450 reductase, the data are consistent with a role for the cytochromes P450-NADPH-cytochrome P450 reductase system in this enantioselective biotransformation process.

Introduction

Since the introduction of glyceryl trinitrate (GTN) into clinical practice over a century ago, organic nitrates have become important clinical agents for the management of angina pectoris and congestive heart failure. Current hypotheses contend that organic nitrates are prodrugs, and that they require biotransformation to an active form (presumably nitric oxide (NO) or some derivative thereof) prior to initiating their pharmacological effects of increased cyclic GMP accumulation and vascular smooth muscle relaxation. Nitric oxide formation from GTN has been demonstrated in several vascular smooth muscle preparations (Chung & Fung, 1990; Feelisch & Kelm, 1991; Marks et al., 1992; Kurz et al., 1993). Vascular GTN biotransformation has been shown to be mediated by glutathione S-transferases (GSTs) (Tsuchida et al., 1990; Nigam et al., 1996) and cytochromes P450 (McDonald & Bennett, 1993). In addition, an unidentified microsomal protein mediates NO formation from GTN (Chung et al., 1992). However, the relative importance of these biotransformation systems in the formation of NO at pharmacologically relevant concentrations of organic nitrates is unclear.

It has been demonstrated that there is an enantioselective aspect to the organic nitrate response in isolated rat aorta, with the D-enantiomer of isoidide dinitrate (IIDN) being 10 fold more potent than L-IIDN for vascular relaxation and IIDN-induced cyclic GMP accumulation (Bennett et al., 1986). This was associated with a greater rate of vascular biotransformation of the more potent enantiomer (Stewart et al., 1989). When tissues were made tolerant to organic nitrates in vitro using high concentrations of GTN, enantioselectivity for relaxation by IIDN was lost (Bennett et al., 1988). In a subsequent study. NADPH-dependent hepatic microsome mediated biotransformation of IIDN was shown to be selective for the D-enantiomer, thereby raising the possibility that the cytochromes P450 system was involved in the enantioselective biotransformation of IIDN (McDonald & Bennett, 1990). Finally, in a study of the haemodynamic effects D- and L-IIDN in conscious rats, the magnitude of both the initial and sustained decrease in mean arterial pressure (MAP) was greater for D-IIDN (Bennett et al., 1991).

Recent reports from this laboratory have described inhibition of the biotransformation and pharmacological actions of GTN by the flavoprotein inhibitor diphenyleneiodonium sulphate (DPI) (McGuire et al., 1994), and the in vitro inhibition of purified DNA expressed NADPH-cytochrome P450 reductase by DPI (McGuire et al., 1998). This suggests that the basis for the inhibitory effect of DPI on GTN action may be inhibition of the metabolic activation of GTN mediated by the cytochromes P450-NADPH-cytochrome P450 reductase system. DPI has also been shown to inhibit a number of other flavin-containing enzymes such as the NADPH-oxidase system responsible for superoxide production in macrophages and neutrophils (Cross & Jones, 1986; Doussière & Vignais, 1992), xanthine oxidase (O'Donnell et al., 1993) and macrophage NO synthase (Stuehr et al., 1991).

Since the enantioselectivity of IIDN action appears to be related to enantioselective biotransformation, and since the inhibitory action of DPI may be due to inhibition of the cytochromes P450-NADPH-cytochrome P450 reductase system, it was of interest to determine whether DPI affected the enantioselective pharmacological properties of IIDN.

Methods

Drugs and solutions

Krebs' solution was composed of the following (mM): NaCl, 118; KCl, 4.74; MgSO₄, 1.18; KH₂PO₄, 1.18; CaCl₂ 2.5; NaHCO₃, 24.9; and glucose, 10. The solution was aerated with 95% O₂-5% CO₂ and maintained at 37°C. Diphenyleneiodonium sulphate (DPI) was purchased from Colour Your Enzyme (Kingston, Ontario, Canada). Stock solutions of DPI were prepared in distilled water. D-isoidide dinitrate (1,4 : 3,6-dianhydro-D-iditol 2,5-dinitrate, D-IIDN), L-isoidide dinitrate (1,4 : 3,6-dianhydro-L-iditol 2,5-dinitrate, L-IIDN), D-IIMN and L-IIMN were obtained from D.H. Stereochemical Consulting Ltd. (Vancouver, British Columbia, Canada). Stock solutions of IIDN and IIMN were prepared by extraction of organic nitrate-lactose powder (50% w/w) with ethanol. Further dilutions were made with the appropriate buffer solution. The concentrations of IIDN and IIMN in stock solutions were determined by a modification of the spectrophotometric method of Dean & Baun (1975) as described previously (Bennett et al., 1988). Isosorbide dinitrate (ISDN) and isosorbide-2-mononitrate (ISMN) were gifts from Wyeth Ltd. (Toronto, Ontario, Canada). L-Phenylephrine hydrochloride, 1-chloro-2,4-dinitrobenzene (CDNB), reduced glutathione (GSH), β-NADPH, xanthine and xanthine oxidase were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). All other chemicals were of at least reagent grade and were obtained from a variety of sources.

Relaxation studies

Thoracic aortic strips were prepared from 250–300 g male Sprague-Dawley rats (Charles River, St. Constant, Quebec, Canada) as described (Stewart et al., 1989). Three aortic strip preparations were obtained from each animal. Tissues were contracted maximally with 10 μM phenylephrine to ensure the viability of the preparation. After a 30 min washout period, the tissues were contracted submaximally with 0.1 μM phenylephrine. Once the induced tone had stabilized, tissues were exposed to diluent (control) or 0.3 μM DPI (treated) and 10 min later cumulative concentration-response curves for D-IIDN (10 nM–0.1 mM) or L-IIDN (0.1 μM–0.1 mM) were obtained.

Tissue biotransformation studies

Aortae were divided in half and placed into individual tubes containing 1 ml of Krebs' solution at 37°C aerated with 95% O₂-5% CO₂. Tissues were exposed to 0.1 μM phenylephrine for 5 min and then to either diluent (control) or 0.3 μM DPI for an additional 10 min. Both tissue segments were incubated with either 2 μM D-IIDN or 2 μM L-IIDN for 5 min and then frozen between liquid nitrogen precooled clamps. The frozen tissues were placed in extraction tubes and 50 pmol of ISDN was added as the internal standard. Each tissue was extracted four times with 1 ml isomeric hexanes to remove IIDN and ISDN and the extracts combined. Each tissue was then spiked with 10 pmol of ISMN as the internal standard, extracted three times with 1 ml of anhydrous diethyl ether to remove the IIMN and ISMN, and the combined diethyl ether extracts dried with anhydrous magnesium sulphate. The tissues were digested with 2N NaOH (1 ml) for 48 h and aortic protein levels determined by the method of Lowry et al. (1951), using bovine serum albumin as the standard.

To determine the IIMN content of the Krebs' incubation medium, samples were first extracted five times with 2 ml isomeric hexanes to remove IIDN. Samples were spiked with 10 pmol ISMN as the internal standard, extracted three times with 2 ml anhydrous diethyl ether, and the combined diethyl ether extracts dried with anhydrous magnesium sulphate. Hexane or diethyl ether extracts were evaporated to dryness under a stream of nitrogen gas, the residue resuspended in 25 μl benzene and frozen at −20°C until analysis the following day. To control for any IIMN formed artifactually during the incubation or present in the stock IIDN solution, 2 μM D- or L-IIDN was incubated in Krebs' solution under the identical experimental conditions. The concentration of IIMN found in these samples was subtracted from the values obtained for the tissue samples, and ranged between 0.14 and 0.50% of the IIDN concentration. Organic nitrates were quantitated by megabore capillary column gas-liquid chromatography as described (McDonald & Bennett, 1990).

Cyclic GMP measurements

Each rat aorta was divided into four segments that were incubated individually in 5 ml of Krebs' solution at 37°C aerated with 95% O₂-5% CO₂. In some experiments, the endothelium was removed by gentle scraping with a scalpel blade. All tissues segments were exposed to 0.1 μM phenylephrine for 5 min followed by either diluent (control) or 0.3 μM DPI for an additional 10 min. Two tissue segments (one control and one DPI-treated) were incubated with 2 μM D-IIDN or 10 μM L-IIDN for 1 min while the other two tissues served as controls (basal and basal plus DPI). At the end of the incubation period, all tissue segments were frozen between liquid nitrogen precooled clamps. Tissues were homogenized in 1 ml of 6% trichloroacetic acid and centrifuged at low speed for 20 min. The supernatant fractions were extracted six times with 2 ml of water-saturated diethyl ether, acetylated (Harper & Brooker, 1975) and cyclic GMP was quantitated by radioimmunoassay (Steiner et al., 1972) The pellet was digested in 2N NaOH for 48 h and aortic protein was determined by the method of Lowry et al. (1951), using bovine serum albumin as the standard.

Glutathione S-transferase activity/biotransformation

The 105,000×g supernatant fraction of rat aorta was prepared as described (Nigam et al., 1996) and GST activity determined by the spectrophotometric method of Habig et al. (1974) using 1 mM CDNB and 1 mM GSH as substrates. To assess the inhibition of GST activity by DPI, 20 μg ml⁻¹ protein of rat aortic supernatant was incubated with 10 μM DPI or diluent (control) and 1 mM NADPH for 10 min before measuring GST activity. One unit of GST activity is defined as the amount of enzyme catalyzing the conjugation of 1 μmol CDNB per min at 25°C. To examine the inhibition of GST-dependent biotransformation of IIDN by DPI, samples of the 105,000×g supernatant fraction of rat aorta (200 μg ml⁻¹ protein) were pretreated for 15 min with 1 mM NADPH and either 10 μM DPI or diluent (control). Samples were then incubated with 1 mM GSH and 2.0 μM D- or L-IIDN for 60 min. Values for non-enzymatic denitration were determined by incubating D- or L-IIDN in the absence of protein and were subtracted from those obtained in the presence of aortic supernatant. The rates of non-enzymatic denitration of D- and L-IIDN were 0.15±0.05 pmol min⁻¹ and 0.18±0.07 pmol min⁻¹, respectively. IIMN was quantitated by megabore capillary column gas-liquid chromatography as described above. Protein content of aortic supernatant was determined by method of Bradford (1976) using bovine serum albumin as the standard.

Haemodynamic-pharmacokinetic studies

Two groups of male Sprague-Dawley rats (250–275 g) had catheters (PE-50 and 30-gauge Teflon fused to 0.02-inch inside diameter Tygon®) inserted into the abdominal vena cava and abdominal aorta, respectively. The animals were anaesthetized for surgery with a combination of ketamine (Rogarsetic®–70 mg kg⁻¹) and xylazine (Rompun®–5 mg kg⁻¹). The catheters were externalized between the scapulae, sutured into position, and filled with 50 IU ml⁻¹ heparin saline in order to help maintain patency during the ensuing 48 h post-surgery recovery period.

In one group of rats, continuous blood pressure measurements were made on conscious, unrestrained animals by connecting the aortic catheter to a Statham pressure transducer (P23D6, Grass Instruments, Quincy, MA, U.S.A.) coupled to a MacLab® data acquisition system for the recording of blood pressure and heart rate. An in vivo dose-response curve was done in order to compare the effects of a single dose of DPI (2 mg kg⁻¹) on varying doses of D- and L-IIDN. Animals were assigned at random to receive bolus i.v. doses of 0.2, 0.5, 1.0 and 2.0 mg kg⁻¹ D- or L-IIDN in saline through the venous catheter. The next dose of drug was administered when the blood pressure had stabilized at pre-injection control values. The control dose-response curve was followed by a 24 h washout period, after which a single 2 mg kg⁻¹ dose of DPI was administered. Fifteen minutes later, a second set of dose-response curve measurements were made using the protocol described above. Each animal received the same enantiomer of IIDN for both control and DPI-treated dose-response curves. Since DPI is an irreversible inhibitor, a second control dose response curve was performed in some animals following the 24 h washout period to control for any sequence effect of IIDN administration. There was no statistical difference between the control data obtained before or after the 24 h washout period (P>0.05).

In a second group of rats, the pharmacokinetics after a bolus i.v. dose of D- or L-IIDN was compared in the presence or absence of DPI (2.0 mg kg⁻¹). Animals were assigned at random to receive a bolus i.v. dose of 2.0 mg kg⁻¹ D- or L-IIDN in 0.5 ml saline through the venous catheter. followed by 0.2 ml saline to flush the remaining drug from the catheter. Blood samples (0.25 ml) were obtained at 2, 5, 10, 15, 20, 30, 45, 60 and 90 min. All blood samples obtained were immediately transferred to tubes containing 5 U heparin and centrifuged at 13,000×g for 30 s. The plasma was separated from the erythrocytes and frozen at −20°C for later analysis. The red blood cells were collected and replaced at the end of the blood sampling protocol. After a 2-day washout period, a single i.v. dose of 2.0 mg kg⁻¹ DPI in 0.2 ml saline was administered. Fifteen minutes later, a bolus i.v. dose of 2.0 mg kg⁻¹ of the same enantiomer in 0.5 ml saline was administered through the venous catheter. The blood sampling protocol was repeated as described above. A second set of blood samples were taken in some animals following the 2-day washout period to control for any sequence effect of IIDN administration. There was no statistical difference between the control data obtained before or after the 2-day washout period (P>0.05).

The frozen plasma samples obtained were analysed for levels of IIDN and IIMN. Upon thawing, 100 μl aliquots of the plasma were placed in individual extraction tubes containing 5 μl of 1.0 μM ISDN as the internal standard. Each sample was extracted three times with 1 ml isomeric hexanes to remove IIDN and ISDN, the extracts combined and then evaporated to dryness under a stream of nitrogen gas. Each sample was then spiked with 5 μl of 1.0 μM ISMN as the internal standard. Each sample was extracted three times with 1 ml anhydrous diethyl ether to remove the IIMN and ISMN. The ether extract was dried with anhydrous magnesium sulphate and then evaporatated to dryness under a stream of nitrogen gas. The residue was resuspended in 50 μl of benzene and frozen at −20°C until analysis by megabore capillary column gas-liquid chromatography (Stewart et al., 1989) as described previously. A standard curve was prepared by spiking 100 μl aliquots of rat plasma with either IIDN or IIMN to produce final concentrations of 0.1–5.0 μM.

Data analysis

All data are presented as the means±standard deviation (s.d.). EC₅₀ values were determined from each concentration-response curve by interpolation. Pharmacokinetic data was analysed using the exponential curve-stripping program, ESTRIP (Brown & Manno, 1978), and pharmacokinetic parameters were obtained using the computer program, PKCALC (Shumaker, 1986). Data from all experiments was analysed by a one-way analysis of variance and Newman-Keuls post hoc test for multiple comparisons. The assumption of homogeneity of variance was tested in all cases using Bartlett's test. Due to inhomogeneity of variance, statistical analysis for the relaxation experiments was performed using logarithmically transformed data. A P value of 0.05 or less was considered statistically significant.

Results

Relaxation studies

The enantioselective relaxation of isolated rat aorta by D- and L-IIDN was reduced in the presence of 0.3 μM DPI (Figure 1a and b). The inhibition of D-IIDN-induced relaxation by DPI was characterized by a parallel, rightward shift of the concentration-response curve. The EC₅₀ value for D-IIDN-induced relaxation was increased 3.6 fold in the presence of 0.3 μM DPI. In contrast, DPI did not inhibit L-IIDN-induced relaxation. In control experiments, the EC₅₀ values for D-IIDN- and L-IIDN-induced relaxation were significantly different, however, the EC₅₀ values for D-IIDN- and L-IIDN-induced relaxation in the presence of DPI were not different.

Figure 1.

Effect of DPI on D- and L-IIDN-induced relaxation of isolated rat aorta. (a) The EC₅₀ values for D-IIDN-induced relaxation were: Control, 0.23±0.18 μM; DPI, 0.83±0.47 μM. ^*Indicates P<0.05 vs control. (b) The EC₅₀ values for L-IIDN-induced relaxation were: Control, 2.39±1.35 μM; DPI, 1.40±0.52 μM P>0.05 vs control. Only the s.d. for the EC₅₀ value is shown (n=9).

Tissue biotransformation studies

Aortic biotransformation was assessed at an IIDN concentration (2 μM) that exhibited a differential relaxation response and that was high enough to permit quantitation of parent drug and metabolites. The tissue concentration of the parent drug was not different (32.7±12.14 and 24.1±9.0 pmol mg protein⁻¹ for D- and L-IIDN, respectively. P>0.05). Exposure of tissues to 0.3 μM DPI resulted in a significant decrease in D-IIMN formation, but had no effect on L-IIMN formation (Figure 2). IIMN formation from D- and L-IIDN in DPI-treated tissues was not different.

Figure 2.

Effect of DPI on rat aortic biotransformation of IIDN. The total IIMN formed in both the aortic tissue and Krebs' medium during a 5 min incubation with 2 μM D- or L-IIDN was: D-IIDN control, 5.0±1.5 pmol mg protein⁻¹; D-IIDN plus 0.3 μM DPI, 2.3±0.6 pmol mg protein⁻¹, L-IIDN control, 2.1±0.7 pmol mg protein⁻¹; L-IIDN plus 0.3 μM DPI, 2.3±1.1 pmol mg protein⁻¹. ^*Indicates P<0.05 vs control. Each value represents the means±s.d. (n=6).

Cyclic GMP measurements

IIDN-induced cyclic GMP accumulation was assessed at an IIDN concentration that resulted in 60–80% relaxation of isolated rat aorta (2 μM D-IIDN and 10 μM L-IIDN) and cyclic GMP levels were assessed at the time of maximal cyclic GMP accumulation (1 min). In the endothelium-intact aorta, cyclic GMP accumulation was increased 2.0±0.6 fold and 1.8±0.4 fold above basal levels for D- and L-IIDN, respectively. In endothelium-denuded aorta, cyclic GMP accumulation was increased 4.0±2.1 fold and 2.5±1.4 fold above basal levels for D- and L-IIDN, respectively. Treatment of tissues with 0.3 μM DPI significantly inhibited D-IIDN-induced cyclic GMP accumulation in both the endothelium-intact and endothelium-denuded tissues, whereas L-IIDN-induced cyclic GMP accumulation was not inhibited by 0.3 μM DPI (Figure 3). Decreased basal cyclic GMP accumulation in tissues treated with DPI was observed only in endothelium-intact tissues, suggesting an inhibitory effect of DPI on endothelial NO synthase resulting in decreased basal activity of smooth muscle guanylyl cyclase.

Figure 3.

Effect of DPI on IIDN-induced cyclic GMP accumulation in isolated rat aorta. Tissue levels of cyclic GMP were determined after a 1 min incubation with 2 μM D-IIDN or 10 μM L-IIDN. ^*Indicates P<0.05 versus Basal control and †indicates P<0.05 versus IIDN control). Each value represents the means±s.d., (n=7, endothelium-intact; n=6, endothelium-denuded).

Glutathione S-transferase activity/biotransformation

The GST activity in the 105,000×g supernatant fraction of rat aorta was 0.20±0.03 U mg protein⁻¹ and this was not significantly different after treatment of the supernatant fraction with 10 μM DPI (0.19±0.01 U mg protein⁻¹, P>0.05, Student's t-test for unpaired data). Similarly, GSH-dependent biotransformation of D- and L-IIDN occurred in the 105,000×g supernatant fraction of rat aorta, and this was not affected by treatment of the samples with 10 μM DPI. The control levels of GSH-dependent biotransformation of D- and L-IIDN were 1.94±0.23 pmol min⁻¹ mg protein⁻¹ (n=7) and 1.78±0.22 pmol min⁻¹ mg protein⁻¹ (n=5), respectively.

Haemodynamic-pharmacokinetic studies

In the intact rat model, the haemodynamic study investigating the decrease in MAP after bolus i.v. doses of IIDN of 0.2, 0.5, 1.0 and 2.0 mg kg⁻¹ demonstrated that 2 mg kg⁻¹ DPI significantly inhibited the decrease in MAP by D-IIDN at doses of 0.5 mg kg⁻¹ or greater and eliminated the potency advantage of D-IIDN such that in the presence of DPI both D- and L-IIDN were equipotent vasodilators (Figure 4). DPI did not affect the decrease in MAP by any dose of L-IIDN. The absolute pressure values before DPI treatment were: D-IIDN, 89.9±12.2 mmHg and L-IIDN, 90.0±8.7 mmHg. The absolute pressure values after DPI treatment were: D-IIDN, 79.2±6.8 mmHg and L-IIDN, 75.6±8.2 mmHg. These values were not significantly different (P>0.05). Although the data in Figure 4 is expressed as the percentage decrease in MAP, the same results are obtained when absolute decreases in MAP are used (data not shown). The pharmacokinetic results following a bolus i.v. dose of 2 mg kg⁻¹ IIDN indicated that 2 mg kg⁻¹ DPI significantly increased the plasma levels of D-IIDN at 5, 10, 15, 20, 30, 45, 60 and 90 min while having no effect on plasma levels of L-IIDN (Figure 5). Evaluation of these data also indicated that the area under the curve (AUC) was significantly increased from control for D-IIDN (58.7±12.2 μM min vs treated, 151.8±16.6 μM min) while L-IIDN was unaffected (46.2±7.9 μM min vs treated, 57.5±4.6 μM min) (P<0.001). DPI also significantly extended the elimination half-life of D-IIDN from a control value of 15.3±2.4 min to 32.0±6.6 min, while the control and treated values for L-IIDN were not different at 17.2±3.3 min and 19.6±1.6 min, respectively (P<0.01).

Figure 4.

Effect of DPI on the per cent decrease in MAP caused by IIDN. The maximum decrease in MAP was measured in conscious rats following bolus i.v. doses of IIDN. A bolus i.v. dose of 2 mg kg⁻¹ DPI significantly inhibited the decrease in MAP caused by D-IIDN at all doses except 0.2 mg kg⁻¹. ^*Indicates P<0.001. D-IIDN control vs D-IIDN plus DPI. DPI had no effect on the decrease in MAP caused by L-IIDN (P>0.05, L-IIDN vs L-IIDN plus DPI). (D-IIDN, n=4; L-IIDN, n=5).

Figure 5.

Effect of DPI on the plasma levels of IIDN. In the intact animal model, plasma levels of IIDN were measured following a bolus i.v. dose of 2 mg kg⁻¹ IIDN. A bolus i.v. dose of 2 mg kg⁻¹ DPI significantly increased the plasma levels of D-IIDN. ^*Indicates P<0.001. D-IIDN control vs D-IIDN plus DPI). DPI had no effect on plasma levels of L-IIDN (P>0.05, L-IIDN vs L-IIDN plus DPI) (n=4).

Discussion

The major finding of this study was that DPI abolished the enantioselective differences between the D- and L-enantiomers of IIDN in both the intact tissue preparation and in the intact animal via an inhibitory effect on the biotransformation and pharmacological actions of D-IIDN. DPI caused a parallel, rightward shift in the concentration-response curve for D-IIDN and increased the EC₅₀ value for D-IIDN-induced relaxation, but had no effect on L-IIDN-induced relaxation. Vascular biotransformation of D-IIDN to its mononitrate metabolite was significantly inhibited by DPI, but DPI had no effect on L-IIMN formation. In both endothelium-intact and endothelium-denuded rat aorta. DPI markedly inhibited D-IIDN-induced cyclic GMP accumulation, but had no effect on L-IIDN-induced cyclic GMP accumulation. In vivo, DPI altered both the pharmacokinetic and haemodynamic properties of D-IIDN, but those of L-IIDN were not affected. Previous studies have suggested that DPI inhibited the biotransformation and pharmacological actions of GTN via an effect on the cytochromes P450-NADPH-cytochrome P450 reductase system (McGuire et al., 1994, 1998). The data reported herein suggest that the enantioselective biotransformation of D-IIDN to NO is mediated by an enzyme that is inhibited by DPI.

The biotransformation of organic nitrates can result in the formation of inorganic nitrite anion (NO₂⁻) or NO. The formation of NO₂⁻ predominates, but NO₂⁻ is not active at concentrations that could be formed from pharmacologically relevant concentrations of organic nitrates. This non-productive formation of NO₂⁻ from organic nitrates has been referred to as ‘clearance-based' biotransformation to distinguish it from ‘mechanism-based' biotransformation–the biotransformation of organic nitrates that results in the formation of activators of guanylyl cyclase. Since the greater potency of D-IIDN for vascular relaxation is associated with a greater rate of biotransformation, it would appear that mechanism-based biotransformation is the basis for enantioselectivity in vascular relaxation by these compounds. This component of the biotransformation of D-IIDN was abolished after treatment with DPI, suggesting that this compound inhibits an enzyme that mediates mechanism-based biotransformation of D-IIDN to NO.

DPI and other diaryliodonium compounds have been shown to inhibit a number of flavin-containing enzymes, including NADPH-oxidase (Cross & Jones, 1986; Doussière & Vignais, 1992), NO synthase (Stuehr et al., 1991) and xanthine oxidase (O'Donnell et al., 1993). The inhibition of bovine and rat hepatic NADPH-cytochrome P450 reductase by DPI has been reported (Doussière & Vignais, 1992; McGuire et al., 1998), and Tew (1993) has demonstrated the inhibition of rat hepatic NADPH-cytochrome P450 reductase by iodonium diphenyl (IDP), an analogue of DPI. DPI and IDP are thought to bind to the reduced redox center of the flavoprotein to form a reversible enzyme-inhibitor complex that is subsequently converted to an irreversible enzyme-inhibitor complex (O'Donnell et al., 1993; Tew, 1993). With respect to the cytochromes P450-NADPH-cytochrome P450 reductase system, sites of covalent binding of DPI may include the flavin or haeme moieties or adjacent amino acids near the active site. In this regard, incubation of rat aortic microsomes with ¹²⁵I-labelled DPI resulted in the labelling of two proteins (79 kDa and 50 kDa), which would be consistent with the binding of DPI to both NADPH-cytochrome P450 reductase and cytochromes P450 (McGuire & Bennett, 1996). The results of the current study suggest that it is an enzyme inhibited by DPI that is the basis for the potency difference between the two enantiomers of IIDN. Possible candidates for inhibition by DPI include both the NADPH-cytochrome P450 reductase and cytochromes P450 since both can mediate the biotransformation of organic nitrates (McDonald & Bennett, 1990, 1993; McGuire et al., 1998). Although biotransformation of organic nitrates by other flavoproteins has not been reported, evidence to exclude the involvement of several potential DPI targets exists. In a previous study by McGuire et al. (1994), the inhibitory effect of DPI on GTN relaxation, biotransformation and cyclic GMP accumulation was the same in endothelium-intact and endothelium-denuded aorta, suggesting that endothelial NO synthase was not involved in the bioactivation of GTN. Under the same experimental conditions, acetylcholine-induced relaxation in endothelium-intact tissues was significantly inhibited by DPI. In a study by McDonald & Bennett (1993), it was demonstrated that the aortic microsomal biotransformation of GTN could be inhibited greater than 90% by treatment of microsomes with carbon monoxide. Since microsomal flavin-containing monooxygenase (FMO) activity is unaffected by carbon monoxide (Stefek et al., 1989; Eriksson & Bostrom, 1988), this would exclude a role for FMO in organic nitrate biotransformation. NADPH-oxidase activity is assayed as cytochrome c reduction that is inhibitable by superoxide dismutase (the O₂⁻ generated by the enzyme is responsible for the reduction of cytochrome c). The NADPH-dependent cytochrome c reductase activity of aortic microsomes is decreased by about 20% by superoxide dismutase (unpublished observations) and therefore it is possible that NADPH-oxidase could contribute to microsomal organic nitrate biotransformation. We have performed experiments to assess the role of xanthine oxidase in organic nitrate biotransformation. Under anaerobic conditions only, incubation of purified xanthine oxidase and xanthine with GTN, D- or L-IIDN resulted in biotransformation that was inhibited by DPI, indicating that this flavoprotein can biotransform organic nitrates (data not shown). To assess whether xanthine oxidase contributed to vascular organic nitrate biotransformation, the 105,000×g supernatant fraction of rat aorta was incubated with GTN, D- or L-IIDN and added xanthine (1 mM). Organic nitrate biotransformation was not observed under anaerobic conditions and was not affected by the addition of xanthine, suggesting that vascular xanthine oxidase does not contribute to organic nitrate biotransformation.

To determine if the findings in the intact tissue preparation translated into significant alterations in the enantioselective haemodynamic-pharmacokinetic properties of IIDN that have been observed in the intact animal (Bennett et al., 1991), rats were treated with 2 mg kg⁻¹ DPI in vivo. DPI clearly abolished the enantioselective differences between D- and L-IIDN and results showed that inhibition of the in vivo biotransformation of D-IIDN related to a decrease in its haemodynamic effects (Figures 4 and 5). Biotransformation of D-IIDN, which is thought to be necessary for its vasodilator activity, was inhibited in the presence of DPI as indicated by an increase in both plasma levels of the parent compound as well as the AUC and elimination half-life. None of the haemodynamic or pharmacokinetic parameters measured for L-IIDN were inhibited by DPI. In effect, there were no differences between the haemodynamic properties of D- and L-IIDN in the presence of DPI, thereby confirming the results obtained in the intact tissue preparation. In the whole animal, IIDN undergoes both mechanism-based and clearance-based biotransformation. Based on the elevated plasma levels of D-IIDN following DPI administration. DPI must also inhibit the clearance-based biotransformation of D-IIDN. Considering the haemodynamic data, where DPI inhibited the D-IIDN-induced decrease in MAP, one might argue that there should have been an increased MAP lowering with DPI because there were significantly higher plasma levels of the parent compound available for mechanism-based biotransformation. However, the haemodynamic data confirms that DPI also inhibited the mechanism-based biotransformation of D-IIDN since the D-IIDN-induced decrease in MAP was less than in control animals. Due to the fact that the pharmacokinetic data are based on whole body arterial blood samples, the increase in plasma levels of D-IIDN reflects the combined effect of DPI on both mechanism- and clearance-based biotransformation. Separating the relative contribution of the mechanism- and clearance-based biotransformation from the overall IIDN biotransformation was not feasible in this system. It was also evident that DPI did not have an effect on the clearance-based biotransformation of L-IIDN meaning that clearance must also be an enantioselective process. Taken together, there is strong evidence that DPI inhibits an enzyme or enzyme system that gives D-IIDN the enantioselective potency advantage over L-IIDN that has been observed both in vitro and in vivo.

Although DPI inhibited the enantioselectivity of IIDN action, D-IIDN-induced relaxation still occurred in the presence of DPI, and the responses to L-IIDN were not altered by DPI. This suggests that other biotransformation systems can mediate NO formation from organic nitrates, assuming that the cytochromes P450 system is the target for the inhibitory actions of DPI and that it is effectively inhibited at the concentration of DPI employed. An alternative interpretation would be that NO- and cyclic GMP-independent mechanisms for relaxation by organic nitrates exist (although none have been reported). Other vascular enzymes reported to mediate organic nitrate biotransformation are the GSTs (Tsuchida et al., 1990; Nigam et al., 1996) and an unidentified microsomal protein that mediates NO formation from GTN (Chung et al., 1992). The GSTs from several different tissues and species exhibit enantioselectivity for the conjugation of glutathione with a variety of substrates (Mannervik & Danielson, 1988). However, the rate of biotransformation of D- and L-IIDN in the 105,000×g supernatant fraction of rat aorta was not different. Furthermore, DPI did not inhibit GST activity in the 105,000×g supernatant fraction of rat aorta nor did it inhibit GSH-dependent biotransformation of D- and L-IIDN, suggesting that the enantioselective component of biotransformation is not mediated by the GSTs. Whether the GSTs are responsible for the non-enantioselective biotransformation of IIDN to NO in intact cells remains to be determined.

In conclusion, we found that the flavoprotein inhibitor, DPI, inhibited D-IIDN-induced relaxation and cyclic GMP accumulation, and vascular D-IIDN biotransformation, but had no effect on the biotransformation and pharmacological actions of L-IIDN. These results were confirmed in the intact rat where DPI inhibited the blood pressure response to D-IIDN and altered the pharmacokinetic properties of D-IIDN, but had no effect on L-IIDN. DPI had no effect on the GST activity or the GSH-dependent biotransformation of D- or L-IIDN in the 105,000×g supernatant fraction of rat aorta. A specific enzyme system responsible for the enantioselective differences between D- and L-IIDN and that is inhibited by DPI was not identified in the current study. However, the accumulated evidence for the cytochromes P450 system as a target for the inhibitory effects of DPI on GTN action (McGuire et al., 1994, 1998; McGuire & Bennett, 1996) would make this system a likely candidate. Since the enantioselective relaxation response of IIDN is abolished in organic nitrate-tolerant tissues (Bennett et al., 1988), the enzymes targeted by DPI may also be targets for the tolerance-inducing effects of organic nitrates.

Abbreviations

1,2-GDN

glyceryl-1,2-dinitrate

1,3-GDN

glyceryl-1,3-dinitrate

AUC

area under the curve

CDNB

1-chloro-2,4-dinitrobenzene

DPI

diphenyleneiodonium sulphate

GSH

reduced glutathione

GST

glutathione S-transferase

GTN

glyceryl trinitrate

IDP

iodonium diphenyl

IIDN

isoidide dinitrate

IIMN

isoidide mononitrate

ISDN

isosorbide dinitrate

ISMN

isosorbide-2-mononitrate

MAP

mean arterial pressure

NO

nitric oxide.