Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

GTN is a common drug used in unstable angina, myocardial infarction and congestive heart failure for its potent and fast-onset action of vasorelaxation. It is believed that GTN requires a bioactivating process that occurs in intact blood vessels to biotransform the drug into an active species, which is presumably NO or a related compound. The active species stimulates soluble guanylyl cyclase and subsequently results in an increase in cGMP levels. The bioactivation pathway for GTN still remains unknown, although it is believed to be an enzymatic system in vascular smooth muscle. Thus far, the proposed GTN bioactivation pathways include GST (Tsuchida et al., 1990), CYP-NADPH CYP reductase system (McDonald and Bennett, 1993; Schroder, 1992) and a membrane-bound thiol containing protein that has not been identified (Chung and Fung, 1990). Evidence supporting the role of CYP in GTN vasodilation comes mostly from nonvascular tissues, such as hepatic microsomes, and nonvascular cell culture systems. In hepatic microsomes, the metabolism of GTN is found to be NADPH-dependent (Pistelli et al., 1994; Servent et al., 1989), is inhibited by CO, SKF 525 A, dioxygen (Bennett et al., 1992b; McDonald and Bennett, 1990) and diphyleneiodonium sulfate (McGuire et al., 1994) and is increased in various CYP inducer-treated rats (Bennett et al., 1994; Delaforge et al., 1993; McDonald and Bennett, 1993). In cell cultures of nonvascular sources (LLC-PK1 and RFL-6), inhibitors of CYP such as miconazole, cimetidine and SKF 525A were able to reduce cGMP stimulation by GTN, but not stimulation by another vasodilator, sodium nitroprusside (Schroder, 1992; Schroder and Schror, 1990). Recently, purified CYP from rabbit liver was shown to catalyze NO formation from GTN in a CYP-NADPH reductase- and NADPH-dependent fashion (Mülsch et al., 1995).

There are many enzymes in the CYP superfamily, and each of them could contribute to GTN metabolism. CYP3A, an enzyme that is responsible for more than 50% of the P450-dependent human xenobiotic metabolism, has been shown to have particularly high reactivity for GTN metabolism in rat liver (Delaforge et al., 1993). McDonald and co-workers (1994) incubated GTN with hepatic microsomes from DEX-treated rats (as the source of CYP) and found that GTN biotransformation was significantly greater than that seen in microsomes from control males. In the incubations of GTN, hepatic microsome and aortic guanylyl cyclase, they observed approximately 80% inhibition of GTN-induced guanylyl cyclase activity upon the addition of anti-CYP3A antibody. On the basis of these observations, they proposed that CYP3A was the major mediator of rat aortic guanylyl cyclase activation by GTN.

Although the bioactivation of GTN by P450 has been clearly demonstrated in nonvascular tissues and cells, most studies on vascular preparations fail to reach the same conclusion. Incubating aortic strips with P450 inhibitors, such as CO, SKF 525A and miconazole, has no effect on GTN-induced vasorelaxation (Braun et al., 1995; Liu et al., 1993; Salvemini et al., 1993). One study reported reduced relaxation of the GTN pharmacologic effect upon treatment with 7-ethoxyresorufin, an inhibitor of CYP1A1 and 1A2 (Bennett et al., 1992a). The characteristics of other vascular isoenzymes, especially their relationship to the bioactivation of GTN in vivo, remain to be examined. Here, we used microsomal preparations directly from rat aorta to test the hypothesis that CYP3A activity in these microsomes correlates with GTN bioactivation in vivo. The levels of cGMP in response to GTN were assessed as the measurement of the extent of GTN bioactivation. We employed KCZ, an imidazole antimycotic agent and a potent CYP3A inhibitor (Meredith et al., 1985), to reduce CYP3A activity in aorta. Because the P450 activity in control rat vasculature was below the detection limit, we were not able to measure directly the effect of KCZ on vascular P450 activity. To circumvent this problem, we treated rats with DEX to induce the activity of aortic P450 to detectable levels. When the rats were treated with the combination regimen of DEX and KCZ, the activity of P450 in aortic microsomes decreased to undetectable levels. Levels of cGMP paralleled the changes of CYP3A activity in the aorta by each treatment. The effects of KCZ and DEX on GST, another major GTN-metabolizing enzyme system in the vasculature, were also assessed.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

GTN (in vehicle of ethanol/propylene glycol/water, 3:3:4 v/v) was purchased from American Regent Laboratories Inc. (Shirley, NY). SNAP was obtained from Research Biochemicals International (Natick, MA) and dissolved in 50% ethanol in sterile distilled water for injection. KCZ was purchased from U.S.P.C. (Rockville, MD). ³H-Labeled TSO was obtained from American Radiolabelled Chemicals Inc. (St. Louis, MO) at a specific activity of 15 Ci/mmol. Unlabelled TSO, glutathione, CDNB, DEX and EDTA were purchased from Sigma (St. Louis, MO). Immunoblotting agents were obtained from Bio-Rad Laboratory (Hercules, CA). Monoclonal antibody against CYP3A was kindly provided by Dr. S. Wrighton (Eli Lilly & Co., Indianapolis, IN), and polyclonal antibody against GST mu was a kind gift from Dr. C. Serabjit-Singh (Glaxo-Wellcome, Research Triangle Park, NC). The secondary antibody was obtained from Life Technologies Inc. (Gaithersburg, MD). Ketamine (100 mg/ml) as Ketaset (ketamine hydrochloride injection, USP) was obtained from Aveco (Fort Dodge, IA); xylazine sterile solution (20 mg/ml) as AnaSed (Lloyd Laboratories, Shenandoah, IA) was obtained from the Animal Care and Cell Culture Facility of the University of California, San Francisco.

Animals. Sprague-Dawley rats (300-350 g) were obtained from Charles River Animal Farm (Wilmington, MA) and kept under constant temperature and humidity conditions with a 12/12 hr light-dark cycle. The rats were maintained on a standard pellet diet with tap water given ad libitum and fasted for 8 hr before they were treated with vasodilators.

cGMP analysis. Portions of aortic tissues from each individual rat were weighed and homogenized in 2 ml of ice-cold modified Hanks' balanced salt solution consisting of (in g/l) NaCl, 8; KCl, 0.4; glucose, 1; KH₂PO₄, 0.06; Na₂HPO₄, 0.047; phenol red, 0.017, with the addition of 25 mM EDTA. The homogenate was centrifuged at 4000 × g for 10 min at 4°C, and the supernatant was transferred into a new test tube containing 1.5 ml of acetonitrile to precipitate the protein. The tubes were vortex-mixed briefly and then centrifuged at 4000 × g for 10 min at 4°C. The supernatant was transferred into a clean test tube and evaporated under nitrogen at 50°C to 60°C. The dried residue was reconstituted with Tris-EDTA buffer, pH 7.5 (0.05 M Tris, 4 mM EDTA, 10 µl buffer/mg tissue). Aliquots (100 µl) of the reconstituted solution, in duplicate, were used for cGMP measurements using a ³H-labeled cGMP radioimmunoassay kit (Amersham Corp., Arlington Heights, IL). The assay was performed as described by the manufacturer.

Preparation of microsomes or cytosol from rat aorta and liver. The frozen liver and aorta were cut into small pieces and immersed in Tris buffer containing EDTA (5 mM) and PMSF (100 µg/ml). Aortic samples from eight or nine rats in the same pretreatment group were pooled together for one microsomal preparation. Assessments of enzyme activities were made from three different sample pools of animals that received the same pretreatment. Aorta was homogenized in an Omni Micro Homogenizer for 20 to 30 sec at half power and then transferred to a Potter-Elvehjem glass-PTFE homogenizer. At this point, both liver and aorta underwent the same standard differential centrifugation procedure. After 105,000 × g centrifugation, the supernatant was saved as a cytosolic fraction and the pellet as microsomes. The microsomes were washed by another 105,000 × g spin at 4°C. All the samples were stored at 80°C.

P450 activity assay. CYP3A and 2C11 activities were assayed as the rate of hydroxylation of testosterone into 6- and 2- metabolites, as previously described (Bornheim et al., 1987). Briefly, ¹⁴C-testosterone (0.25 mM) was incubated with microsomes (0.1 mg of protein), NADPH, 1-isocitric acid (5 mM), 1-isocitrate dehydrogenase (1 unit) and 0.1 M phosphate buffer, pH 7.4, at 37°C for 10 min. Methylene chloride (3 ml) was added to stop the reaction, and a fixed amount of cold metabolites was added as carriers to the incubation solution. After three extractions with methylene chloride, the extracts were pooled and analyzed on a Rainin C18 reverse-phase HPLC column (Beckman system). The mobile phase was made of a gradient solution of methanol and acetonitrile The metabolites were individually collected and quantitated on the basis of scintillation counting (Beckman scintillation counter).

Total GST activity. Total GST activity was measured by spectrophotometric assay using CDNB as a general substrate (Habig et al., 1974). Briefly, the assay was carried out in a 1-ml incubation solution containing 1 mM CDNB, 5 mM glutathione, 0.1 mM potassium phosphate buffer, pH 6.65, and 10 µl of 4000 × g aortic supernatant. The increase in absorbance was measured at 340 nm at room temperature. The extinction coefficient is 9.6 mM¹ per cm.

GST mu isozyme activity. GST mu activity was determined by measuring the TSO-glutathione conjugation rate using a radiometric assay (Seidegard et al., 1984). The aorta was homogenized in phosphate-buffered saline. After 4000 × g centrifugation, the supernatant fraction (10 µl) was incubated in a final volume of 100 µl of phosphate buffer containing 4 mM glutathione and 250 µM ³H-TSO (specific activity = 15 Ci/mmol) at pH 7.4. The reaction mixture was incubated at 37°C for 10 min and the reaction was stopped by extraction with 2 volumes of hexanol. The glutathione conjugation rate in the aqueous phase was measured using liquid scintillation counting. Nonenzymatic rates of reaction were obtained from 10 µl of boiled tissue samples undergoing the same procedure. The enzyme activity was determined as the total conjugation rate corrected for nonenzymatic rates.

Immunoblot analysis. Microsomal and cytosolic samples (10 µg/lane of aorta samples and 2 µg/lane of liver samples) were separated by 12% SDS-polyacrylamide gel electrophoresis (Laemmli, 1970). Each sample is a pool of six or eight rats that received the same pretreatment. The separated proteins were transferred onto nitrocellulose membranes. After blotting and washing, the membranes were incubated with the appropriate primary antibody, either anti-GST mu (1:1000) or anti-CYP3A (1:5000), for 1 hr. Then the membranes were washed with 1% TTBS three times for 20 min each wash before they were incubated with secondary antibody for 1 hr. The ECL system (Amersham, CA) was used to develop the signal. The intensity of the bands was read by densitometry (Pharmacia LKB Technology, Alameda, CA).

Protein assay. Protein concentration was determined using the Biorad protein assay kit, at an absorbance of 595 nm at room temperature.

Data analysis. Data are expressed as mean ± S.D. Comparisons were made using one-way analysis of variance and the Student-Newman-Keuls test with statistical significance set at P < .05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

The effect of KCZ on cGMP elevation by GTN in vivo. Basal levels of cGMP in intracellular KCZ vehicle-treated rats were not significantly different from those in KCZ-treated rats (fig. 1). The 2-mg bolus dose of GTN induced a 3 to 5-fold increase in cGMP (146 ± 23 pmol/g of tissue). Treatment with KCZ significantly, but not completely, decreased the cGMP elevation produced by GTN (102 ± 12 pmol/g of tissue) (fig. 1). SNAP at 0.1 mg induced an increase in cGMP levels similar to that observed for GTN at 2 mg. Treatment with KCZ did not have any significant effect on SNAPinduced cGMP levels (fig. 2).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   The effect of KCZ on cGMP levels induced by GTN. Solid bar: rats receiving 1 ml of vehicle pretreatment 1 hr before the experiment. Hatched bar: rats receiving 1 ml of KCZ pretreatment before the experiment. Group 1: rats receiving 0.4 ml of GTN vehicle bolus injection. Group 2: rats receiving 2 mg of GTN bolus dose in 0.4 ml of vehicle. (Six rats in each treatment group, total of 24 rats.) * P < .05.

View larger version (36K): [in this window] [in a new window]   Fig. 2.   The effect of KCZ on cGMP levels induced by SNAP. Solid bar: rats receiving 1 ml of vehicle pretreatment 1 hr before the experiment. Hatched bar: rats receiving 1 ml of KCZ pretreatment 1 hr before the experiment. Group 1: rats receiving 0.4 ml of SNAP vehicle bolus injection. Group 2: rats receiving 0.1 mg of SNAP in 0.4 ml of vehicle as a bolus dose. (Six rats in each treatment group, total of 24 rats.)

The effects of DEX and DEX/KCZ combination treatment on cGMP elevations by GTN in vivo. Pretreatment with DEX resulted in an increased cGMP response to GTN (249 ± 26 vs. 156 ± 22 pmol/g of tissue, DEX-treated rats vs. control rats, respectively) (fig. 3). This elevated response to GTN was suppressed to the control level (135 ± 22 pmol/g of tissue) when the rats received an i.p. injection of KCZ after DEX treatment. The basal and inducible levels of cGMP were not affected by pretreatment of the vehicle for DEX or by the combination treatment of vehicles for DEX and KCZ (fig. 3). No significant effect of DEX was observed on SNAP-induced cGMP levels (data not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 3.   The effects of DEX or the combination treatment of DEX and KCZ on GTN-induced cGMP levels. Solid bars in groups 1 and 2: DEX vehicle-pretreated rats. Solid bar in group 3: DEX vehicle- and KCZ vehicle-pretreated rats. Hatched bars in groups 1 and 2: DEX-pretreated rats. Hatched bar in group 3: DEX/KCZ-pretreated rats. Group 1: rats received 0.4 ml of GTN vehicle bolus injection. Groups 2 and 3: rats received 2 mg of GTN in 0.4 ml of vehicle as a bolus dose. (Six rats in each treatment group, total of 36 rats.) * P < .05.

The effects of KCZ and DEX on CYP3A activity and expression in rat aorta. Microsomal P450 activity in control rat aorta was below measurable levels, so the effect of KCZ on aortic P450 was undetectable. However, when the rats were treated with DEX for 4 days, both CYP3A and 2C11 activities were measurable, though still at low levels (table 1). KCZ treatment after DEX induction suppressed CYP3A activity to undetectable levels in the combination-treatment group, whereas 2C11 activity was not significantly affected (table 1). Using monoclonal antibody against rat CYP3A, Western blot analysis showed that DEX pretreatment increased the expression of CYP3A in aortic tissue by ~10-fold (fig. 4A, lane 2 vs. lane 1). The combination treatment did not alter the 3A expression induced by DEX (fig. 4A, lane 3 vs. lane 2).

View this table: [in this window] [in a new window]   TABLE 1 The effects of DEX and KCZ on aortic CYP activities Each data point is the mean ± S.D. of three different sample pools, subject to the same pretreatment, with some samples taken from the same tissues that underwent cGMP analysis. 6-OHT (6 testosterone hydrolase activity) and 2-OHT (2 testosterone hydrolase activity) are the markers for CYP3A and CYP2C activities, respectively.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   The effects of DEX or DEX/KCZ treatment on CYP3A and GST mu expression in rat aorta and rat liver. A) CYP3A expression in rat aorta or liver. Lane 1: control rat aorta (0.14); lane 2: DEX-pretreated rat aorta (1.52); lane 3: DEX/KCZ-pretreated rat aorta (1.7); lane 4: control rat liver (0.43); lane 5: DEX-pretreated rat liver (2.89); lane 6: DEX/KCZ-pretreated rat liver (2.84). B) GST mu expression in rat aorta and liver. Lane 1: control rat aorta (0.08); lane 2: DEX-pretreated rat aorta (0.18); lane 3: DEX/KCZ-pretreated rat aorta (0.15); lane 4: control rat liver (0.47); lane 5: DEX-pretreated rat liver (2.73); lane 6: DEX/KCZ-pretreated rat liver (3.10). The numbers in parentheses are densitometry readings in arbitrary units. The blot presented is representative of three separate experiments.

The effects of KCZ and DEX on CYP3A activity and expression in rat liver. From the same groups of rats that underwent assessment of aortic P450 activity, hepatic microsomes were prepared and their P450 activities were measured. KCZ markedly reduced the CYP3A activity, as compared with that in the control group, but had no significant effect on CYP2C11 activity (table 2). DEX pretreatment alone increased hepatic CYP3A activity by ~6 fold. When the combination pretreatment of both DEX and KCZ was examined, hepatic CYP3A activity was reduced to the same level as observed in the control group. Western blot analysis showed that the constitutive expression of CYP3A in liver is much higher than that in the aorta (fig. 4A, lane 4 vs. lane 1). In the liver of rats that received either DEX alone or the DEX/KCZ combination treatment, the expression of CYP3A was increased (fig. 4A, lane 5 vs. lane 4). DEX appeared to inhibit CYP2C11 activity, but the effect was not statistically significant. The combination treatment yielded 2C11 activity similar to that for DEX alone (table 2).

View this table: [in this window] [in a new window]   TABLE 2 The effects of DEX and KCZ on hepatic CYP activities The data are means ± S.D. (n = 3 in each treatment). Data obtained for the livers are from the same rats for which aortic results are given in Table 1. 6-OHT (6 testosterone hydrolase activity) and 2-OHT (2 testosterone hydrolase activity) are the markers for CYP3A and CYP2C activities, respectively.

The effects of KCZ and DEX on aortic GST activities and expression. In rat aorta, neither KCZ nor DEX had any significant effect on total GST activity toward CDNB (table 3). DEX significantly induced GST mu activity toward TSO in vascular tissue, whereas KCZ had no significant effect. The GST mu activity in the combination DEX/KCZ treatment group was the same as found for the DEX-treated group alone, a result consistent with the lack of effect noted for KCZ pretreatment. The induction of GST mu expression by DEX was assessed by Western blot analysis using a monoclonal antibody against GST mu (fig. 4B). GST mu expression in DEX-pretreated rat aortic cytosol was 2 times higher than that in control rats. KCZ treatment did not affect GST protein expression (data not shown). In the liver, the constitutive levels of GST mu were much higher than those found in the aorta. DEX pretreatment increased the expression of GST mu in liver by ~6-fold (fig. 4B).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study, to the best of our knowledge, is the first to demonstrate the correlation of CYP3A activity in rat aortic microsomes with GTN bioactivation in vivo. Previously, the ability of CYP3A to bioactivate GTN has been suggested. Using immunoinhibition studies with CYP3A antibody, McDonald et al. (1994) have shown that 80% of hepatic microsome-mediated biotransformation of GTN into an activator of guanylyl cyclase is carried out by CYP3A. They also demonstrated that biotransformation of GTN and the resulting ability to stimulate guanylyl cyclase was greater in hepatic microsomes prepared from DEX-treated rats than in hepatic microsomes from control rats. To examine further the role of this enzyme in vasculature to bioactivate GTN, we directly measured the CYP3A activity in rat aorta.

We were able to detect CYP3A in rat aorta by using a monoclonal antibody against this enzyme via a Western blot analysis (fig. 4), but the activity was too low to be measured (table 1). However, in the DEX-treated rats, both the activity and the expression of CYP3A were readily detected in the aorta. Treatment with KCZ drastically reduced the CYP3A activity in DEX-treated rat aorta (table 1). It also almost completely inhibited CYP3A activity in rat hepatic microsomes. Thus treatment with KCZ alone presumably decreased the CYP3A activity in aorta, though this effect is not directly detectable. We have been unable to measure the CYP levels in aorta using UV-Vis difference spectra, which made the comparison of aortic and hepatic activity normalized for CYP content impossible. However, the trend in CYP3A changes for aortic microsomes was similar to that for hepatic microsomes. DEX increased hepatic CYP3A activity by 6.3-fold (table 2). In both control and DEX-pretreated groups, KCZ treatment caused marked inhibition of CYP3A in hepatic microsomes (6.5-fold and 8-fold, respectively) (table 2), as previously reported for both in vitro and in vivo functional studies of metabolism (Meredith et al., 1985). cGMP levels induced by GTN were found to change in different treatment groups, in parallel to the CYP3A activity in aorta. In DEX-pretreated aorta, GTN-induced cGMP levels were elevated, whereas in the DEX/KCZ combination-treatment group, GTN-induced cGMP levels were reduced to those of control (fig. 2). We expected that the reduction in GTN-induced cGMP levels in the KCZ-pretreated group would also correlate with CYP3A activity in aorta, but these levels were below our detection limit. By contrast, the cGMP accumulation induced by the nitrovasodilator SNAP was not affected by either DEX or KCZ treatment. This indicates that a direct effect of DEX or KCZ on guanylyl cyclase activity is unlikely.

CYPs have previously been shown to be capable of reducing GTN. The formation of P450 Fe(II)-NO and P450 Fe(III)-GTN complexes was demonstrated in incubations of liver microsomes with GTN (Servent et al., 1989). Bennett et al. (1992b) found that a product of this incubation was able to stimulate aortic guanylyl cyclase to elevate cGMP levels in cell-free systems, and recently, the identity of this product as NO was confirmed by ESP spectroscopy (Mülsch et al., 1995). Using aortic microsomes pooled from 200 aortas, McDonald and Bennett (1993) reported a NADPH-dependent biotransformation of GTN that was inhibited by CO, SKF 525A and oxygen and that increased in microsomes from phenobarbital-treated rats. However, phenobarbital is known to induce multiple isoforms of P450, including 2B, 3A and 2C, as well as GST, another major xenobiotic metabolic enzyme that has been implicated in GTN biotransformation (Guengerich et al., 1982; Hales and Neims, 1977). Oxygen and CO are general inhibitors for P450-dependent reductive reactions. SKF-525A forms a metabolic intermediate complex with 2B1, 2C11 and 3A1/2 (Murray and Reidy, 1990). Further attempts to incubate more selective inhibitors of P450namely cimetidine, miconazole, proadifen, metyrapone and 7-ethoxyresorufinwith aortic strips yielded no effects on cGMP levels or on measurements of relaxation induced by GTN (Braun et al., 1995; Liu et al., 1992; Liu et al., 1993). Thus, even though previous biotransformation studies of GTN were in agreement with inhibitor and inducer studies in hepatic microsomes, two questions remained to be addressed: 1) whether this biotransformation of GTN contributes to GTN bioactivation in the vasculature, and 2) which isoenzyme of vascular P450 is predominant in the bioactivation process. Transformation of GTN (where the covalence value of nitrogen is +5) to NO (where the covalence value of nitrogen is +2) is a reductive process that is sensitive to dioxygen tension. In vitro experiments on GTN denitration and GTN-induced vasorelaxation reveal an increased sensitivity of these processes under low-oxygen conditions (Bennett et al., 1994; Bennett et al., 1992b), which appears to explain the preferential vasodilation action of GTN on the venous side of the circulation. However, the relevance of CYP in GTN pharmacological action in vivo has not been tested. The high sensitivity of the reductive process of P450 toward O₂ leads us to question whether the previous in vitro experiments represent the in vivo condition, especially in testing vasorelaxation action (most of these in vitro experiments were conducted under oxygenated buffer conditions). The present in vivo study was designed to measure the bioactivation of GTN directly and thus to avoid the controversies related to previous experiments.

Although the immunoinhibitory evidence of CYP3A from the study of McDonald et al. (1994) is in agreement with our result of CYP3A being able to bioactivate GTN, the authors of the former study reached a different conclusion concerning the effects of DEX on GTN action. McDonald et al. (1994) observed no effect of DEX treatment on the vasorelaxation produced by GTN. We conjecture that this discrepancy results from different experimental methodologies. McDonald and co-workers (1994) reported that under aerobic conditions, GTN biotransformation does not differ between the control and DEX-pretreated group. However, DEX treatment shows enhancement of GTN biotransformation only under anaerobic conditions (McDonald et al., 1994). Thus it should not be surprising that in vitro vasorelaxation, which is performed under aerobic conditions in their study, is the same for control rats and DEX-treated rats. In another study, using difference spectra, Delaforge et al. (1993) observed less NO production from GTN incubation with hepatic microsomes from DEX-treated rats, compared with control rats. This result is contradictory to that of the present study. The reason for this difference is not clear, but it may be that the in vivo activator of GC is a product from GTN other than NO. Our speculation is supported by recent evidence that detectable NO production from GTN does not correlate with GTN-induced vasorelaxation (Marks et al., 1995). Thus we believe that the levels of cGMP at an early time-point, as measured in our study, are probably a better indicator of the extent of GTN bioactivation.

Only limited studies have been conducted to demonstrate metabolic reactions via the P450 system in the vasculature, although the existence of this enzyme system in aorta has been reported (Irizar and Ioannides, 1995). The ability of aortic CYP to metabolize exogenous and endogenous substrates has been demonstrated (Bond et al., 1979; Finnen et al., 1986; Juchau et al., 1976). We observed much less CYP3A expression in rat aorta than in rat liver (fig. 4A). In the DEX-treated animals, we found that aortic microsomes possessed 90-fold less 6-testosterone hydroxylase activity than that in hepatic microsomes (tables 1 and 2). By comparison, McDonald and Bennett (1993) reported 7-ethoxycoumarin demethylase activity in aortic microsomes to be 100-fold less than that in hepatic microsomes. Similarly, Juchau et al. (1976) observed 97-fold less activity of aryl 4-monoxygenase in rabbit aorta than in rabbit liver.

Besides CYP3A, CYP2C was reported to account for 20% of GTN bioactivation (McDonald et al., 1994). Although the use of cimetidine, a selective 2C11 inhibitor, also suggested a role for 2C11 in cultured cells (Schroder and Schror, 1990), in vivo treatment of rats with cimetidine showed no effect on GTN-induced vasorelaxation (Bennett et al., 1992a). In our study, DEX did induce 2C11 in rat aorta. In animals pretreated with DEX and KCZ, GTN-induced cGMP was similar to that of the control group, although 2C11 activity was elevated. This finding suggests that a role for 2C11 in bioactivating GTN in vivo is unlikely. Attempts to immunoblot this isoform in rat aorta were unsuccessful.

GST has also been reported to produce NO from GTN, though these enzymes are much less efficient than P450 (Kurz et al., 1993). GSTs are present in rat aortic smooth muscle, and the mu isoform has been shown to be the predominant form to metabolize GTN (Lau et al., 1992). Contradictory observations have been reported for GST-directed vasorelaxation in vitro (Chung et al., 1992; Kenkare et al., 1994; Lau and Benet, 1992; Yeates et al., 1989). Furthermore, Haefeli et al. (1993) observed no correlation between GST mu polymorphism (as measured in lymphocytes) and nitroglycerin responsiveness in humans. Here, we demonstrate that GTN-induced cGMP levels in vivo were not consistent with aortic GST activity. KCZ did not affect GST activity, yet it reduced GTN-induced cGMP levels (table 3; fig. 1). In rats pretreated with DEX and KCZ, GST activity in aorta was elevated, but the level of GTN-produced cGMP was reduced. Our observation that DEX treatment increased GST mu expression in rat liver (fig. 4) is consistent with other reported studies (Flomerfelt et al., 1993; Waxman et al., 1992). Our study is the first to demonstrate the inducing ability of DEX on GST mu expression in rat aorta. A similar induction of GST mu in rat aorta and liver by phenobarbital was recently reported by Kashfi et al. (1994).

In conclusion, we have evaluated the effects of DEX and KCZ treatment on vascular P450 and GST activities, as well as their effects on GTN-induced cGMP levels. We observed a correlation of GTN bioactivation with CYP3A, but not with GST mu in vivo. However, the contribution of CYP3A is relatively small (20%-30%). Whether the major portion of GTN bioactivation is carried out by another P450 isozyme that has not been systematically examined or by another, unidentified biotransformation pathway remains to be determined. Future studies are also needed to examine the role of P450 in development of tolerance to GTN, a process previously suggested to be due to a reduced bioactivation of GTN in vasculature (Forster et al., 1991).