Introduction

Abstract Introduction Results Discussion References

In recent years, it has become evident that some apoenzymes of P450¹ are closely associated with the synthesis of NO. In the liver, microsomes catalyze the oxidative denitration of NOHA, thus producing NO and citrulline, oxidation that is not inhibited by NO synthase inhibitors, but rather by classical P450 inhibitors, such as miconazole and troleandomycin (1). NOHA is not the only substrate of rat liver microsomal P450 originating NO, because the reduction of pentamidine (2) or 18-nitro-oxyandrostenedione (3) also generate NO. In vascular tissues, the inhibition of the P450 of the carotid artery endothelium by inhaled anesthetics, clotrimazole, metyrapone, or SKF525A abolishes the NO-dependent relaxation induced by acetylcholine (4). Formation of NO seems to be rather specific to isoforms of the CYP3A family (5). Further supporting that P450 generates NO is the fact that cimetidine reduces in a dose-dependent manner the formation of NO induced by cytokines without affecting the mRNA of NO synthase (6).

The NO precursor, L-arginine, has been used clinically for a number of years as a vasodilator (7) in pathological conditions, such as angina pectoris, congestive heart failure, hypertensive emergencies, pulmonary hypertension, percutaneous angioplasty, and complications after cardiac catheterization (8). On the other hand, NO synthase inhibitors, such as L-NAME, are considered as therapeutic agents in the treatment of endotoxemia and in the prevention of the hypotension associated with the treatment with cytokines (9).

The fact that NO synthase presents structural similarites with specific isoforms of P450, and that both are able to synthesize NO from L-arginine analogs, raises the hypothesis that NO synthase substrates may inhibit the biotransformation of drugs catalyzed by selected isoforms of P450. The aims of this study were to assess whether L-NAME or L-arginine are able to reduce the rate of biotransformation of theophylline in vivo, and to confirm the results ex vivo and in vitro using cultured hepatocytes. Theophylline is used as substrate, because it is biotransformed by several apoenzymes of P450. As a consequence, it is a marker of the activity of a vast array of isozymes (10-12).

Materials and Methods

Animals. Male New Zealand rabbits (Ferme Cunicole, Les Lapins Léonard, Mirabel, Canada) weighing 2.0-2.2 kg were used throughout the study. Rabbits were kept in well-ventilated cages and were fed with dry food and water ad libitum. Rabbits were allowed 8 days for acclimatization before being included in the experiments.

In Vivo Studies. At time 0, rabbits from the three experimental groups (N = 6/group) received in a lateral vein of an ear 2.5 mg/kg of theophylline dissolved in 0.9% sodium chloride. Blood samples were withdrawn through a catheter (Butterfly-21, Abbot Ireland, Sligo, Ireland) inserted in the central artery of an ear before and at 5, 10, 15, 20, 30, 60, 120, 180, 240, 300, 360, 420, and 480 min after the injection of theophylline. Blood was immediately centrifuged for 5 min at 2500 rpm, and the plasma was stored at 20°C until analysis. A vesical catheter (Bardex Foley 8 ch/Fr, Mississauga, Ontario, Canada) was installed to collect urine for measurement of theophylline metabolites. Theophylline and metabolite concentrations in plasma and urine were quantified by HPLC as described elsewhere (13).

Ex Vivo Studies. Hepatocytes were isolated from the control (N = 4), and from rabbits pretreated with L-NAME (25 mg/kg) (N = 4) and L-arginine (150 mg/kg) (N = 4) using an in situ perfusion technique as described by Seglen (17), with minor modifications. Hepatocytes were separated from nonparenchymal cells by differential centrifugation at 50g and then passed over a 40% Percoll gradient to obtain a highly purified cell population. Viability was >90% as assessed by trypan blue exclusion. Hepatocytes (4 × 10⁶/ml) were placed into 12-well plastic culture plates (Falcon, Becton Dickinson Labware, Rutherford, NJ) coated with type I rat tail collagen (Sigma); cells were suspended in WME (Sigma), and supplemented with 10% calf serum (Sigma) and 1 µM insulin (Boehringer Mannheim Biochemica, Mannheim, Germany). Culture plates were incubated at 37°C in a humidifier with 95% O₂/5% CO₂.

In Vitro Studies. To investigate further the mechanism underlying the inhibition of P450 by L-NAME and L-arginine, in vitro studies were conducted with isolated hepatocytes from control rabbits (N = 4). The hepatocytes of each rabbit were segregated in three aliquots: two of the aliquots were incubated with 1 mM L-NAME or 1 mM L-arginine, and the third aliquot was used as a control. At time 0, 176 µM of theophylline dissolved in serum-free WME (50 µl) was added to the hepatocytes (4 × 10⁶). Four hours later, hepatocytes were harvested by scrapping with a spatula after two washes with phosphate-buffered saline to measure the amount of total P450 (15) and the content in proteins (16). Incubation media were collected to measure the concentration of theophylline, 3MX, 1MU, and 1,3DMU.

Analysis of Data. Theophylline kinetic parameters were estimated assuming noncompartmental kinetics. The AUC_T(0-t) was estimated by means of the trapezoidal method. The AUC_T(0-) was obtained by adding to the AUC_T(0-t) the value of C_t/z, wherein z is the theophylline rate constant of disposition estimated from the slope of the terminal phase of theophylline plasma concentrations. Systemic clearance, terminal half-life, and predicted apparent volume of distribution at steady-state of theophylline were calculated as described by Gibaldi and Perrier (18).

	Results

Abstract Introduction Results Discussion References

In vivo Studies.

Effect of L-NAME on Theophylline Metabolism. Pretreatment with L-NAME induced an increase in theophylline plasma concentrations (fig. 1). As a consequence, theophylline AUC_T(0-) was almost 42% greater (p < 0.05) than in control rabbits; the increase in AUC_T(0-) was secondary to the reduction in theophylline systemic clearance by 46% (table 1). The apparent volume of distribution of theophylline was not affected by the administration of L-NAME. In urine, only 1,3DMU was detected, and accounted for 80 ± 12% of the dose of theophylline injected in control rabbits. After pretreatment with L-NAME, the amount of 1,3DMU recovered in urine was decreased by 41% (p < 0.05).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Theophylline plasma concentrations after the intravenous dosage of 2.5 mg/kg to control rabbits (N = 6) (); rabbits pretreated with L-NAME, 25 mg/kg every 8 hr for 48 hr (N = 6) (); and rabbits pretreated with L-arginine, 150 mg/kg every 8 hr for 48 hr (N = 6) (). Vertical bars are means ± SE.

Effect of L-Arginine on Theophylline Metabolism. Forty-eight hours after the administration of 150 mg/kg of L-arginine, theophylline plasma concentrations were increased (fig. 1). As a consequence, theophylline AUC_T(0-) was almost 40% greater (p < 0.05) than in control rabbits; the increase in AUC_T(0-) was secondary to a reduction in theophylline systemic clearance by 42% (table 1). Volume of distribution at steady state of theophylline was not affected by the administration of L-arginine. Compared with control rabbits, L-arginine decreased the production of 1,3DMU by 37% (p < 0.05).

Ex Vivo Studies. After incubation of theophylline for 4 hr with cultured hepatocytes from control rabbits, 3MX, 1MU, and 1,3DMU were found in the media at concentrations of 0.25 ± 0.02, 0.28 ± 0.03, and 14.5 ± 0.4 µg/ml, respectively. Pretreatment of the rabbits with L-NAME reduced the formation of 3MX, 1MU, and 1,3DMU by 42, 45, and 32%, respectively (p < 0.05) (fig. 2). On the other hand, pretreatment of the rabbits with L-arginine reduced the formation of 3MX by 34% (p < 0.05). The production of 1MU was slightly reduced (15%); however, the difference did not reach statistical significance (fig. 2).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Concentration of 3MX, 1MU, and 1,3DMU after 4 hr of incubation with 176 µM of theophylline in hepatocytes of control rabbits (N = 4), and rabbits treated with 25 mg/kg of L-NAME (N = 4) or 150 mg/kg of L-arginine (N = 4) every 8 hr for 48 hr. Vertical bars are means ± SE. Asterisk is p < 0.05, compared with control.

In Vitro Studies. The presence of L-NAME or L-arginine during the 4-hr period of incubation with theophylline did not diminish the concentration of 3MX, 1MU, or 1,3DMU (i.e. it did not reduce the formation of the metabolites in vitro) (table 2). On the other hand, incubation of hepatocytes with L-NAME or L-arginine did not modify the concentration of total P450 (table 2).

	Discussion

Abstract Introduction Results Discussion References

The present results show that pretreatment of rabbits with L-NAME for 2 days reduces the clearance of theophylline by ~40%, mostly due to a decrease in its 8-hydroxylation. In ex vivo studies, in addition to the 8-hydroxylation of theophylline, L-NAME also inhibited the 1- and 3-demethylations. On the other hand, at the dosage of 75 mg/kg, L-arginine did not depress the clearance of theophylline in vivo (data not shown); dosages of 150 mg/kg of L-arginine were required to reduce the clearance of theophylline. At this dosage, the 8-hydroxylation of theophylline was reduced by L-arginine to the same extent as did L-NAME. However, ex vivo, L-arginine reduced significantly only the 1-demethylation of theophylline.

Several mechanisms may be proposed to explain the inhibition of P450 produced in vivo by L-NAME and L-arginine. Pretreatment of rabbits with L-NAME did not result in the downregulation of P450 expression, because the concentration of total P450 was not modified by L-NAME. NO is able to suppress P450-dependent oxygenation reactions by reacting with both Fe²⁺ and Fe³⁺ hemes (20). It is conceivable that the administration of L-arginine may have induced the production of NO and depressed the activity of P450; however, we tend to believe that an unspecific reaction with the iron of the heme should have elicited a much broader inhibition of P450. Therefore, this mechanism cannot explain the effect of L-NAME.

The fact that L-NAME or L-arginine did not alter the in vitro biotransformation of theophylline discards a competitive inhibition as the mechanism underlying the in vivo inhibition of theophylline metabolism. L-NAME and L-arginine had to be administered in vivo to produce an inhibition, pointing out a catalysis-dependent inhibition. Several pieces of evidence support the view that inhibition is of the quasi-irreversible type (21). First, L-NAME and L-arginine contain several amine groups that could be oxidized to yield nitroso groups able to bind to the ferrous heme. Second, total P450 measured in rabbits pretreated with L-NAME was not decreased and that could be secondary to the complexation of the nitroso group to the heme resulting in the stabilization of the inactive enzyme that shows its maximal absorbance at 450 nm (22). Finally, L-NAME inhibits cytochrome c reductase by reacting with iron, and it has been proposed that L-NAME could affect other iron-containing systems in the cell (23). Therefore, a catalysis-dependent inhibition implies that both L-NAME and L-arginine will be oxidized in situ to a reactive intermediate that will complex with the heme. A similar mechanism explains the inhibition of selected isoenzymes of P450 elicited by erythromycin and troleandomycin (21).

L-NAME is metabolized to N--nitro-L-arginine and methanol (24, 25). L-Arginine is catalyzed to NO and citrulline by NO synthase (i.e. a P450 hemoprotein) (26). Besides NO synthase, other enzymatic systems are able to catalyze L-arginine analogs to NO and citrulline. NOHA is catalyzed by rat liver microsomes in the presence of NADPH and O₂ (1), a reaction not inhibited by specific inhibitors of NO synthase but rather by inhibitors of P450 (e.g. miconazole and troleandomycin); thus, suggesting that isozymes of the CYP3A family are able to catalyze NOHA. Further supporting the role of CYP3A, dexamethasone, a potent inducer of CYP3A, increases the formation of NO and citrulline from NOHA (27). Besides CYP3A, other isoforms are able to catalyze L-arginine (e.g. CYP2C3) (2, 26). These studies suggest that, in hepatic and extrahepatic (4) tissues, selected isoforms of the P450 catalyze L-arginine analogs, a catalysis that could be the source of nitroso-reactive intermediates.

In humans, the 1-demethylation of theophylline is to be performed by CYP1A2, the 3-demethylation by CYP1A1 and CYP1A2 (10), and the 8-hydroxylation by isoforms of the CYP1A, CYP2D, CYP2E, and CYP3A families (12). In rabbits, the metabolism of theophylline also involves several isoenzymes of P450 (28). Ex vivo, L-NAME reduced the formation of the three metabolites of theophylline, thus indicating that L-NAME is able to inhibit several isoforms of P450. Several mechanisms, not mutually exclusive, could explain why, in vivo, the dosages of L-arginine able to inhibit the metabolism of theophylline were 6-fold greater than those of L-NAME. First, in addition to primary amines, L-NAME contains a nitro group, all able to form reactive intermediates; L-arginine contains only two primary amine groups (21). Second, L-arginine may not be a substrate of P450, but only its hydroxylated metabolite (1), which is also rapidly converted into NO and citrulline by NO synthase. Finally, L-arginine could be catalyzed by parallel routes to nonreactive metabolites and reactive intermediates, and therefore high dosages of L-arginine would be required to generate reactive intermediates.

To explain the fact that ex vivo only the rate of formation of 3MX was significantly affected by L-arginine, and not that of 1,3DMU as observed in vivo, we may speculate that the complexation of L-arginine intermediates with the ferrous heme of the isozyme(s) responsible for the formation of 1,3DMU is weak and rapidly reversible. Thus, ex vivo, during the processes of cell isolation and culture, the isoform would recover its activity. Indeed, further studies are required to confirm the mechanisms underlying the differences between L-NAME and L-arginine.

In conclusion, L-NAME inhibits several isozymes of P450 and as such could interact with the biotransformation of a great number of drugs (29) with a narrow therapeutic range, such as theophylline, warfarin, cyclophosphamide, cyclosporin, lidocaine, quinidine, terfenadine, etc. An interaction between L-NAME and these drugs seems possible in humans, because the doses used clinically are higher than those administered in the present study (14). L-Arginine is a less potent inhibitor of P450 than L-NAME. However, in clinical practice, L-arginine has been used at dosages higher than those used in the present study [i.e. 200 mg/kg injected intravenously (30) or 30 g daily for 3-7 days (31, 32)] and, therefore, an interaction with other drugs may also be possible.