The vasodilator isoamyl nitrite has been used for years to relieve angina pectoris . The structural isomer n-butyl nitrite (BN) is used as a drug of abuse to produce a euphoric- like state . Both of these distinctive features have been attributed to the vasodilatory action of nitric oxide (NO) formed during the metabolism of isoamyl nitrite , or when isoamyl nitrite is reduced by ascorbate or dithiothreitol . Reports have also appeared suggesting that the abuse of nitrite inhalants is a cofactor in AIDS [4, 5], or Kaposi’s sarcoma in AIDS patients .
BN has also been shown to be cytotoxic to lymphocytes . Furthermore, T-lymphocyte blastogenesis and antibody responsiveness is decreased 90% when mice are exposed to BN at levels similar to those used by chronic users [8, 9]. Mortality was found to occur in rats exposed to ≥ 600 ppm BN vapours for 14 days and hepatocellular cytoplasmic vac- uolization was noted . Previously, we showed that BN readily induces cytotoxicity in isolated hepatocytes. The molecular cytotoxic mechanisms involved the immediate formation of S- nitrosoglutathione (GSNO) as well as a con- comitant decrease in protein thiols, followed by a marked ATP depletion and finally lipid peroxidation . Furthermore, GSH- depleted hepatocytes were resistant to BN suggesting that GSNO formed in normal hepatocytes treated with BN contributes to ATP depletion, lipid peroxidation and cyto- toxicity. We also showed that the cytotoxici- ty of N-methyl-N’-nitro-N-nitrosoguanidine (MNNG) towards isolated hepatocytes could be attributed to GSNO formation and was prevented by cytochrome P-450 inhibitors . In the present study, we have also found that cytochrome P-450 inhibitors pre- vented BN cytotoxic effects including inhi- bition of mitochondrial respiration, ATP depletion, lipid peroxidation and plasma membrane disruption.
2. Materials and methods
Collagenase (from Clostridium his- tolyticum), HEPES and bovine serum albu- min were obtained from Boehringer- Mannheim (Montreal, P.Q., Canada). Trypan blue, metyrapone, imidazole and iso- niazid were obtained from the Sigma Chemical Co. (St. Louis, MS, USA). Desferoxamine was a gift from Ciba Geigy (Mississuga, ON, Canada). Isopropanol and toluene were purchased from BDH Chemicals (Toronto, ON, Canada) and ace- tone was obtained from Anachemia Canada Inc. (Toronto, ON, Canada). SKF 525A was obtained from Smith Kline Beecham Pharmaceuticals (King of Prussa, PA, USA) and HPLC grade solvents were purchased from Caledon (Georgetown, ON, Canada). n-Butyl nitrite was purchased from Aldrich Chemical Company (Milwaukee, WI, USA).
2-(4-Carboxyphenyl)-4,4,5,5-tetram - ethylimidazoline-3-oxide-1-oxyl (carboxy- PTIO) was purchased from Calbiochem- Novabiochem Corp. (La Jolla, CA, USA). Other chemicals were of the highest grade available.
2.2. Isolation and incubation of hepatocytes
Hepatocytes were isolated from male Sprague-Dawley rats (250-300g), main- tained on a standard chow diet, by collage- nase perfusion of the liver, as previously described . Cell viability was assessed by determining the percentage of the hepato- cytes which excluded trypan blue. Routinely, 85 to 90% of hepatocytes excluded trypan blue immediately after isolation. GSH- depleted hepatocytes were prepared by preincubating with n-bromoheptane as pre- viously described .
Hepatocytes (106 cells/ml) were suspended in Krebs-Henseleit buffer (pH 7.4) containing 12.5 mM HEPES. All incubations were performed in rotating, round bot- tomed flasks at 37 °C under a continuous flow of 95% O2 and 5% CO2 or 1% O2, 94% N2, and 5% CO2, where indicated. Reactions were started by the addition of BN. Aliquots of the incubation mixture were taken at various time points for biochemical analysis and cell viability determination.
2.3. Biochemical assays
Lipid peroxidation was measured by treating 1 ml aliquots of hepatocytes with 1 ml 20% trichloroacetic acid, 1 ml 0.8% thiobarbituric acid, and the mixture was heated for 20 min. Samples were centrifuged for 5 min at 2500 g and the supernatant was mon- itored at 535 nm . nATP levels in hepa tocytes were determined by alkaline extrac- tion and quantified by HPLC using a C18 µBondapak reverse phase column (Waters Associates, Milford, MA, USA) . Glycolysis was determined by measuring lactate formation from fructose. Lactate was measured by the formation of NADH from NAD+ by lactate dehydrogenase as previously described . Total GSH and GSSG content of hepatocytes were measured by an HPLC procedure using a µBondapak NH2 column, a Waters 6000 A solvent delivery system, a WISP 710A automatic injector and Data Module (Water Associates, Milford, MA, USA) .
2.4. Peroxynitrite/NO2 assay
The rate of peroxynitrite formation from BN was determined using the chemilumi- nescence method described by Radi et al. . BN was added to 1 ml of hepatocytes (106 cells/ml) in the presence of 1 mM luminol and the relative light unit (RLU) was recorded every 6 seconds, using a Luminometer LB 9501-Berthold Lumat.
2.5. Hepatocyte respiration measurement
Hepatocyte respiration was measured at determined time points after the addition of BN to the incubation of hepatocytes, using a Clark-type oxygen electrode (Model 5300; Yellow-Spring Instrument Co., Inc.) in a 2 ml chamber maintained at 37 °C. Prior to oxygen consumption measurement, hepato- cytes (106 cells/ml) were kept at 37 °C in Krebs-Henseleit buffer plus 12.5 mM HEPES, pH 7.4 under 95% O2 and 5% CO2.
Values are means ± SD of three separate experiments unless otherwise stated. Statistically significant differences between control and experimental groups were obtained using one way ANOVA.
BN cytotoxicity was dose-dependent and became marked above 100 µM of BN at 95% O2. Plasma membrane bleb formation was seen as early as 30 min after 200 µM BN was added, with cytotoxicity commenc- ing around 60 min and 60% cytotoxicity occurring at about 120 min. However, as shown in Table 1, BN cytotoxicity was prevented if the hepatocytes were preincubated for 15 min with the cytochrome P-4502E1 inhibitors acetone, and 2-hexanone. Similarly, the P-450 inhibitors phenylimida- zole, SKF 525A, metyrapone, isoniazid, piperonyl butoxide, and imidazole also pre- vented or markedly delayed cytotoxicity. On the other hand, carboxy-PTIO, which converts NO to NO2 radical , increased BN cytotoxicity. At low oxygen concentrations, BN was five times less toxic as the ED50 for 2 hr determined at 1% O2 (1 mM) was five times higher than that determined at 95% O2 (0.2 mM).
Table 1. Effects of cytochrome P-450 inhibitors on BN cytotoxicity.
Note: Hepatocytes (106 cells/ml) were incubated in Krebs-Henseleit buffer (pH 7.4) at 37 °C under an atmosphere of 95% O2 and 5% CO2 with P-450 inhibitors for 20 min before addition of butyl nitrite. Cytotoxicity was determined as the percentage of cells taken up trypan blue.Values are shown as means ± SD of at least three separate experiments.a Significantly different from untreated cells (p< 0.001) bSignificantly different from butyl nitrite treated cells (p< 0.005) c Carboxy-PTIO, a nitric oxide oxidizing agent, was not toxic at the concentration used.
Lipid peroxidation, as measured by mal- ondialdehyde formation, occurred only at BN concentrations which were toxic to the cell. As shown in Table 2, lipid peroxidation was also prevented by the P-450 inhibitors SKF 525A, phenylimidazole, piperonyl butoxide, and metyrapone, if added before BN but not if added 30 min after BN. Lipid peroxidation was increased by carboxy- PTIO and did not occur at a low 1% O2 concentration. The addition of BN to hepatocytes caused peroxynitrite formation as determined by luminol chemiluminescence. Furthermore, prior GSH depletion or low oxygen concentrations prevented peroxynitrite formation from BN. Antioxidants and P-450 inhibitors eg. Phenylimidazole decreased the rate of peroxynitrite formation(Figure 1).
Table 2. Effect of cytochrome P-450 inhibitors on BN-induced hepatocyte lipid peroxidation.
Note: Hepatocytes (106 cells/ml) were incubated in Krebs-Henseleit buffer (pH 7.4) at 37 °C under an atmosphere of 95% O2 and 5% CO2 with P-450 inhibitors for 20 min before addition of butyl nitrite. Values are shown as means (± SD) of at least three separate experiments.*MDA: Malondialdehyde.aSignificantly different from untreated cells (p < 0.01). bSignificantly different from butyl nitrite treated cells (p< 0.01).
Figure 1. Peroxynitrite formation from butyl nitrite. Butyl nitrite was added to 1 ml of hepatocytes (106 cells/ml) incubated with 1 mM luminol at 1 min and the relative light unit (RLU) was measured every 6 second. GSH-depleted hepatocytes were prepared by preincubation with bromoheptane; P-450 inhibition was achieved by preincubating hepatocytes with phenyl imidazole (0.3 mM) for 15 min. (_) No addition, (_) butyl nitrite 200 µM, (×) GSH depleted hepatocytes + butyl nitrite 200 µM, (_) P-450 inhibited hepatocytes + butyl nitrite 200 µM, (Æ) butyl nitrite 200 µM + morin 100 µM, (_) butyl nitrite 200 µM + hepatocytes incubated at 1% O2. Background chemiluminescence in GSH depleted or P-450 inhibited hepatocytes were not significantly different from control hepatocytes.
Hepatocyte GSH levels were rapidly depleted upon addition of BN even if the hepatocytes were pretreated with P-450 inhibitors (data not shown).
ATP depletion preceded cytotoxicity and unlike lipid peroxidation, occurred with subtoxic concentrations of BN . As shown in Table 3, ATP depletion was not prevented by antioxidants or desferoxamine but was prevented if the hepatocytes were preincubated with the P-450 inhibitors SKF525A, phenylimidazole, piperonyl butox- ide, metyrapone or 2-hexanone. Hepatocyte ATP was not depleted at 1% O2 with 0.2 mM BN, even though initial ATP levels were lower than levels found at 95% O2.
Table 3. Effect of cytochrome P-450 inhibitors or reactive oxygen scavengers on butyl nitrite-induced ATP depletion.
Note: Hepatocytes (106 cells/ml) were incubated in Krebs-Henseleit buffer (pH 7.4) at 37 °C under an atmosphere of 95% O2 and 5% CO2 with P-450 inhibitors or antioxidants for 15 min before addition of butyl nitrite. Values are shown as means (± SD) of three separate experiments. a Significantly different from untreated cells (p < 0.01). bSignificantly different from butyl nitrite treated cells cells (p < 0.03).
As shown in Table 4, BN (300 µM) decreased hepatocyte respiration to about 50% and 31% of control value by 5 min and 30 min, respectively. Prior GSH depletion or P-450 inhibition prevented the inhibition of hepatocyte respiration by BN. On the other hand, the NO-oxidizing agent car- boxy-PTIO  slightly increased the inhibition of mitochondrial respiration by BN. The rate of glycolysis was also markedly inhibited by BN; however, the inhibition of glycolysis was little affected by most P-450 inhibitors or carboxy-PTIO (Table 5).
Table 4. Effect of butyl nitrite on hepatocyte respiration.
Hepatocytes (106 cells/ml) were incubated in Krebs-Henseleit buffer, pH 7.4 under 95% and 5% CO2. Oxygen uptake was measured with a Clark type electrode and a 2 ml chamber. GSH depleted hepatocytes were prepared by preincubating hepatocytes with bromoheptane. P-450 inhibition was achieved by preincubating hepatocytes with phenyl imidazole (300 µM) for 15 min.aSignificantly different from control. GSH depleted or P-450 inhibited hepatocytes were not significantly different from control.Carboxy-PTIO (100 µM) did not affect respiration of untreated hepatocytes.
Table 5. Effect of cytochrome P-450 inhibitors on butyl nitrite-induced inhibition of glycolysis.
Hepatocytes (106 cells/ml) were incubated in Krebs-Henseleit buffer, pH 7.4 at 37 °C under an atmosphere of 95% O 2 with P-450 inhibitors for 15 min before the addition of butyl nitrite. Fructose was added after butyl nitrite. Values are shown as means ± SD of at least three separate experiments. aSignificantly different from untreated cells (p < 0.001). bSignificantly different from frucctose treated cells ( p < 0.001)
It has been previously shown that BN cytotoxicity towards isolated hepatocytes involves ATP depletion, lipid peroxidation, and plasma membrane disruption . Membrane disruption and lipid peroxidation but not ATP depletion were prevented by antioxidants. However, BN-induced ATP depletion, cytotoxicity and lipid peroxidation could be prevented, if GSH was depleted before the addition of BN. This suggested that GSH was required for the cytotoxicity of BN. Similar results were also obtained for the NO donor MNNG .
S-Nitrosoglutathione formation from organic nitrites has been shown  to be catalyzed by human GSH transferases particularly M1a-1a and A 1-1 but not P1-1 (eq. 1).RCH2ONO + GSH _ GSNO + RCH2OH (eq. 1).
Furthermore, whilst stable in solution, GSNO slowly decomposes in the presence of GSH or ascorbate to form NO, apparently catalyzed by trace metals . GSNO may thus act as an intracellular NO store which contributes to the cytotoxicity of BN. 2 GSNO _ 2 NO. + GSSG (eq. 2)
The present study shows that cytochrome P-450 also contributes to the BN cytotoxic mechanisms as various P-450 inhibitors or substrates prevented cytotoxicity. Furthermore, P-450 inhibitors partially prevented the BN-induced inhibition ofmitochondrial respiration, ATP depletion, and lipid peroxidation. Other investigators have shown that sodium nitrite reacts with Fe2+-P-450 to form an unstable NO-Fe2+-P- 450 complex which is converted to a NOFe2+- P-420 complex [23, 24]. Our results suggest that a similar unstable complex is also formed when BN or GSNO reacts with reduced cytochrome P-450 which suggests that reduced P-450 can reduce BN or GSNO to form nitric oxide (eq. 3, 4).
RCH2ONO + P-450-Fe2+ _ P-450-Fe3+ + NO. + RCH2O- (eq. 3)
GSNO + P-450-Fe2+ _ NO. + GS- + P-450-Fe3+ (eq. 4)
BN also markedly increased peroxynitrite/NO2. formation in normal hepatocytes which was prevented if GSH depleted hepatocytes were used. This suggests that peroxynitrite formation from BN catalyzed by hepatocytes is dependent on GSNO formation and could be responsible for the cytotoxicity. Peroxynitrite has been shown to cause lipid peroxidation  and could therefore be responsible for BN-induced lipid peroxidation and cytotoxicity at high O2 concentrations. Lipid peroxidation could also be initiated by nitroxyl radicals or nitrogen dioxide formed from the peroxynitrite anion  as the NO oxidizing agent carboxy- PTIO, which converts NO to NO2., further increased cytotoxicity and lipid peroxidation induced by BN. The high oxygen requirement for BN-induced lipid perox idation and ATP depletion further suggests that peroxynitrite could be the toxic species responsible for the inhibition of mitochondrial respiration and lipid peroxidation. Thus GSH was instantly depleted to the same extent upon addition of BN at 1% or 95% O2 (data not shown), whereas peroxynitrite formation, ATP depletion or lipid peroxidation did not occur at 1% but at 95% O2. Furthermore, the NO-oxidant carboxy-PTIO also increased the inhibition of mitochondrial respiration and ATP depletion induced by BN at 95% O2, further suggesting that an oxidation product of NO was responsible for the inhibition of mitochondrial respiration.
Cytochrome P-450 inhibitors, reactive oxygen scavengers, or desferoxamine partially inhibited peroxynitrite/NO2formation from BN. It is, therefore, more likely that cytochrome P-450 plays a role in BN cyto- toxicity b y catalyzing the formation of per- oxynitrite (eq. 5, 6) or by acting as a source of superoxide radicals which are involved in the formation of cytotoxic peroxynitrite as shown in equation 7.
NO. + P-450-Fe2+_ P-450-Fe2+NO (eq. 5)
P450-Fe2+NO + O2 _ P-450-Fe3+ + ONOO-(eq. 6)
NO. + O .-_ ONOO-(eq. 7)
It is possible that cytochrome P-450 activates peroxynitrite by catalyzing its conversion to NO2 which also readily causes lipid peroxidation . Such a reaction (eq. 9) has also been suggested to explain the activation of peroxynitrite by peroxidases in which per- oxidase compound II is formed .
ONOOH + Fe3+ _ NO2 + Fe4+O (complex II) (eq. 8) BN markedly depletes hepatocytes ATP prior to lipid peroxidation even at noncytotoxic doses . As hepatocyte respiration was partially inhibited when ATP depletion occurred, it is possible that ATP depletion
arose because: (a) NO complexed the ironsulfur proteins of the respiratory chain ;
(b) peroxynitrite inhibited mitochondrial respiration by inactivating cytochrome C oxidase or succinate dehydrogenase [29, 30] or by inactivating aconitase of the citric acid cycle [31, 32]; and/or (c) GSNO (probably via NO+) inhibits glycolysis by inactivating glyceraldehyde-3-phosphate dehydrogenase by S-nitrosylating the active site which causes nonenzymatic auto-ADP-ribosylation by NAD+ [33-35]. Cytochrome P-450 inhibitors prevented mitochondrial respiration inhibition and ATP depletion induced by BN, without affecting the inhibition of glycolysis by BN, which suggests that ATP depletion primarily results from mitochondrial toxicity caused by peroxynitrite, whereas, glycolysis inhibition can be attributed to NO. Previously, it has been shown that prior hepatocyte GSH depletion prevented the inhibition of glycolysis by BN .
In conclusion, a metabolite of BN, most likely peroxynitrite (or its breakdown products)formed from GSNO, is responsible for the inhibition of mitochondrial respiration,ATP depletion, lipid peroxidation and cytotoxicity induced by BN. Furthermore, cytochrome P-450 also seems to be involved in the reductive activation of BN or GSNO to form reactive metabolites eg. Peroxynitrite which causes lipid peroxidation that mediates BN cytotoxicity. Cytochrome P-450 inhibitors or substrates prevented the inhibition of hepatocyte respiration but did not prevent the inactivation of glycolysis by BN. It is interesting to speculate whether NO or GSNO inhibits glycolysis whereas peroxynitrite inhibits mitochondrial respiration.