Chem. Res. Toxicol. 1994, 7, 792-799
792
Hepatic Biotransformation of Parathion: Role of Cytochrome P450 in NADPH- and NADH-Mediated Microsomal Oxidation in Vitro Michael Murray* and Alison M. Butler Liver Research Group, Department of Medicine, University of Sydney, Westmead Hospital, Westmead, NSW 2145, Australia Received May 12, 1994@
In vitro studies have established that cytochrome P450 (P450) is deactivated by the electrophilic sulfur atom released during the enzymic oxidation of parathion to paraoxon. However, in vivo studies in rats have been unable to demonstrate significant P450 loss. This study evaluated the possibility that there may be alternate pathways of parathion biotransformation in liver, other than those mediated by P450 and supported by NADPH. Initial experiments confirmed that parathion administration did not decrease microsomal P450 or testosterone hydroxylation activities. Subsequent in vitro experiments identified a n NADHdependent pathway of parathion biotransformation, and MS was used to confirm that paraoxon and 4-nitrophenol were the products of both the NADH- and NADPH-dependent reactions. The Michaelis constants of the NADH-dependent formation of paraoxon and 4-nitrophenol (26 f 6 pM and 53 f 10 pM,respectively) were approximately 3-fold greater than those for the NADPH-supported reactions (9 f 1pM and 18 & 3 pM, respectively). Induction of male rats with phenobarbital and dexamethasone, but not ,f?-naphthoflavone, produced similar increases in the rates of NADH- and NADPH-mediated parathion metabolism. Rates of NADHand NADPH-dependent metabolism were highly correlated in linear relationships. An antiNADPH-cytochrome P450 reductase (NADPH-P45O reductase) antibody partially inhibited microsomal parathion oxidation mediated by either cofactor, and the P450 inhibitor clotrimazole was similarly effective against the NADH- and NADPH-supported oxidation of parathion. Finally, a reconstituted system containing P450 2B1,NADPH-P450 reductase, and phospholipid supported parathion oxidation mediated by NADH. Michaelis constants for NADHsupported parathion metabolism to paraoxon and 4-nitrophenol were almost a n order of magnitude greater than those for NADPH-supported metabolism. Consistent with this finding, P450 inactivation by NADH was observed, but only a t higher cofactor concentrations ('2 mM), which suggests that the same mechanism may be operative in thionosulfur oxidation mediated by both cofactors. Considered together, these findings indicate that NADH and NADPH are both cofactors for the P450-mediated metabolism of parathion in rat hepatic microsomes. Rat liver reportedly contains higher concentrations of NAD(H) than NADP(H), but these are lower than the concentrations required for P450 loss. Thus, it is conceivable that the NADH-mediated reaction may function in vivo and serve to protect P450 from deactivation.
Introduction Phosphorothioate pesticides are of great commercial value in agriculture and are preferred over organochlorines, such as l,l'-(2,2,2-trichloroethylidene)bis[4-chlorobenzenel (DDT),' dieldrin, and chlordane, because they are less persistent in the environment. Phosphorothioates, such as parathion, chlorpyrifos, and fenitrothion, are themselves relatively inert and undergo bioactivation to the corresponding phosphate esters, or oxons, that are considerably more effective insecticides (1, 2 ) . It has been demonstrated in numerous studies that microsomal mixed-function oxidases participate in phosphorothioate bioactivation (3-6). In the process the thionosulfur atom in these compounds is transferred to cytochrome P450 (P450), the principal catalytic component in the mixed-function oxidase system (4, 6). Thus,
* Address carrespondence to this author. Phone: (61-2)-633-7704; Fax: (61-2)-635-7582. Abstract published in Advance ACS Abstracts, October 1, 1994. Abbreviations: CAD, collisionally activated dissociation;DM', 1,l'(2,2,2-trichloroethylidene)bis[4-chlorobenzene];IgG, immunoglobulin G, NADPH-P450 reductase; NADPH-cytochrome P450 reductase; P450, cytochrome P450. @
P450 enzymes are degraded and their associated activities are markedly decreased during oxidative desulfuration (6, 7). Although the inactivation process is readily detected in microsomes in vitro, correlations with in vivo findings have generally been poor. Thus, there have been reports that P450 activities in rat liver are essentially refractory to inactivation after in vivo administration (8, 9). One explanation that may account for the discrepancy between in vivo and in vitro findings is that alternate pathways of phosphorothioate oxidation may exist in vivo.
The present study investigated the role of reduced pyridine nucleotides in microsomal parathion oxidation in rat liver (Scheme 1). The principal finding t o emerge was that NADH, as well as NADPH, is able to support the efficient biotransformation of parathion t o paraoxon and 4-nitrophenol in vitro. Evidence was obtained that P450 is involved in parathion oxidation mediated by both cofactors. The significance of the NADH-dependent reaction is that, at NADH concentrations likely to be found in vivo, parathion biotransformation occurs without the P450 deactivation noted during NADPH-mediated desulfuration.
0893-228x/94/2707-0792$04.50/00 1994 American Chemical Society
Chem. Res. Toxicol., Vol. 7, No. 6, 1994 793
Parathion Biotransformation in Hepatic Microsomes
,4\-0-0-1,
Scheme 1. Pathways of Parathion Biotransformation in Hepatic Microsomes ,0CpH5
1
0
DESULFURATION
6C2H5
PARAOXON
PARATHION
u 4-NITROPHENOL
Materials and Methods Chemicals. Parathion was provided by Rhone-Poulenc (Brisbane, Australia) and was a t least 99% pure by HPLC. The metabolites paraoxon and 4-nitrophenol and biochemicals were from Sigma Chemical Co. (St Louis, MO). [14ClTestosterone(sp act. 56 mCUmmo1) was purchased from Amersham Australia (Sydney, NSW), and steroid metabolites were obtained from Sigma or from the Steroid Reference Collection (Queen Mary's College, London, England). Silica gel TLC plates containing F254 indicator (E. Merck, Darmstadt, Germany) were used for the separation of testosterone metabolites. HPLC solvents were from Rhone-Poulenc and analytical reagent grade chemicals were from Selby-Anax (Sydney, Australia). Animals. Male Wistar rats (-250 g) and New Zealand rabbits were obtained from the Department of Animal Care, Westmead Hospital. Rats were held in cages and had free access to water and r a t chow. Some rats received either phenobarbital (100 m g k g in saline ip once daily for 3 days), dexamethasone (100 mgkg in corn oil ip once daily for 3 days), or p-naphthoflavone (45 m g k g in corn oil ip once daily for 3 days). The rats were anesthetized and killed 24 h after the last dose of inducer, and washed hepatic microsomes were prepared as before (10).In another experiment, rats received parathion according to two regimens: either 4 m g k g ip in corn oil once daily for 3 days (rats were killed 24 h after the third dose) or a single dose of 16 mg/kg ip prior to death. The final microsomal pellets were resuspended in potassium phosphate buffer (50 mM, pH 7.4) that contained 20% glycerol and 1 mM EDTA and were frozen rapidly in liquid nitrogen. Microsomal suspensions were stored a t -70 "C until required in experiments. Protein was determined according to Lowry et al. (11). Isolation of P450 2B1, NADPH-P450 Reductase, and Anti-NADPH-P450 Reductase IgG. P450 2B1 was purified from microsomal fractions obtained from phenobarbital-pretreated r a t liver, as described previously (12),based on procedures outlined by Guengerich and Martin (13). The purification of NADPH-P45O reductase from sodium cholate-solubilized r a t hepatic microsomes has also been described (14).The final preparations appeared homogeneous on 7.5% sodium dodecyl sulfate-polyacrylamide gels and had specific contents of 14 nmol of P450 2Bl/mg of protein and 49 ymol of cytochrome c reduced/(min.mg of protein). The NADPH-P450 reductase preparation contained no detectable cytochrome bg (assessed spectrophotometrically) or any low molecular weight proteins (by polyacrylamide gel electrophoresis). An antiserum to the rat NADPH-P450 reductase was raised in a female New Zealand rabbit by a standard protocol of three inoculations over a period of 4 weeks (15). The rabbit was bled via a n ear vein 2 weeks after the final inoculation and then fortnightly for 6 weeks. Immunoglobulin G (IgG) was obtained by ammonium sulfate precipitation and chromatography on DEAE-AfXgel Blue (Bio-Rad); preimmune IgG was isolated in analogous fashion from a rabbit that had not been immunized with any antigen. Assay of Parathion Biotransformation. The basic assay procedure has been described previously (7). Briefly, incubations in microsomal fractions (0.4 mL reaction volume) contained 250 yM parathion (5-250 pM in kinetic experiments) and 0.5
mg of microsomal protein and were initiated by the addition of 1 mM NADH (0.5-10 mM in kinetic experiments) or NADPH (0.05-0.2 mM in kinetic experiments). f i r a 2 min incubation in a shaking water bath (37 "C), reactions were terminated by removal to a dry ice/acetone bath. In some experiments, paraoxon (12.5 and 250pM) was employed as substrate in place of parathion. The capacity of NADPH and NADH to support the metabolism of paraoxon to 4-nitrophenol was determined. In some experiments, clotrimazole (in 5 pL of dimethylformamide) was included in incubations (400 pL; final concentration 1.25%); solvent alone was added to control incubations. This concentration of dimethylformamide did not affect parathion metabolism significantly. In incubations that included anti-NADPH-P450 reductase, the IgG was preincubated with the microsomal fractions (10 mg of IgG/mg of microsomal protein) for 10 min prior to inclusion of the other reaction components. The reaction was initiated with NADH or NADPH and then terminated as described above. Studies of parathion metabolism were also undertaken in reconstituted enzyme systems that included homogeneous P450 2B1 (200 nmol), NADPH-P45O reductase (400 nmol), and sonicated dilauroylphosphatidylcholine (20 yg) in potassium phosphate buffer (0.1 M, pH 7.41, that also contained 20% glycerol and 1mM EDTA. Preincubation of the components at 37 "C was undertaken for 5 min, after which parathion (250 pM) and either NADPH or NADH (1 mM) were added. Parathion and its metabolites were extracted by the method of Sultatos et al. (16);this method enables the quantitative extraction of each compound (7). Separation of Parathion Metabolites by HPLC. Parathion, paraoxon, and 4-nitrophenol, as well a s the internal standard 4,4'-dihydroxybiphenyl, were separated on an Ultrasphere-Si HPLC column (5 pm, 4.6 mm i.d. x 25 cm, Beckman, San Ramon, CA) using a Waters Associates system. The mobile phase was dichloromethane/acetonitrile/glacialacetic acid, 93/ 7/0.02, the flow rate was 1 m u m i n , and the detection wavelength was 254 nm. Peak area ratios were used t o calculate metabolite formation ratios on a Waters 730 data module. Standard curves were prepared using known quantities of authentic metabolites. Assay of Microsomal Testosterone Hydroxylation. The assay of microsomal testosterone metabolism is based on methods described by Waxman et al. (17).Incubations (37 "C, 2.5 min) contained microsomal protein (0.15 mg/mL) and [14C]testosterone (50 pM, 0.18 yCi) and were initiated with NADPH (final concentration 1mM). In immunoinhibition experiments, anti-NADPH-P450 reductase IgG was added to incubations in the ratio 10 mg of IgG/mg of microsomal protein and preincubated a t 37 "C for 10 min, and then reactions were conducted. Reactions were stopped with chloroform (5 mL) and removed to ice. Metabolites were extracted and applied to TLC plates and then resolved by sequential development in &chloromethane/ acetone (4/1) and chlorofordethyl acetate/ethanol (4/1/0.7). Radioactive zones on TLC plates were located by autoradiography (Hyperfilm-MP, Amersham), and metabolite formation was estimated by liquid scintillation spectrometry (ACS 11, Amersham).
794 Chem. Res. Toxicol., Vol. 7, No. 6, 1994
Murray and Butler
1
loo
A
1123
93
I
0
11
0
11
50
100
RETENTION TlMEIminl
mlr
Figure 1. HPLC profile of parathion and its metabolites obtained from microsomal incubations with (A) NADPH and (B) NADH as cofactors. The numbers 1-4 indicate the respective elution positions of parathion, 4-nitrophenol, the internal standard (4,4'-dihydroxybiphenyl), and paraoxon. Mass Spectrometry. Metabolites of parathion produced in NADPH- and NADH-fortified microsomal incubations were separated by HPLC as descrited above, collected into vials, and analyzed by MS (Finnegan TSQ-46 triple stage quadrupole instrument, San Jose, CA) in the Department of Pharmacy, University of Sydney. Material that eluted from the Ultrasphere-Si column with the same retention time as paraoxon was subjected to direct MS (source temperature 140 "C; methane chemical ionization-MS). Material that eluted with the same retention time as 4-nitrophenol was subjected to collisionally activated dissociation (CAD)-MS using argon a s the reactant gas in the second quadrupole (collision pressure 2 mTorr, collision energy -20 eV). Other Assays. P450 content in microsomes was determined by the ferrous P450-carbonyl procedure of Omura and Sato (18). Data Analysis and Statistics. Data obtained for NADPHand NADH-supported parathion oxidation in rat hepatic microsomes were treated according to Lineweaver-Burk (1/ V versus 1/S, where S is the parathion concentration and V is the rate of metabolite formation). Data are expressed as mean f SEM throughout. Differences between two groups were detected using Student's t-test whereas differences between more than two groups were detected using Dunnett's test after single factor ANOVA.
Results Kinetics of Parathion Oxidation in Microsomes. The oxidative biotransformation of parathion to paraoxon (by desulfuration) and 4-nitrophenol (by dearylation) occurred in hepatic microsomes from adult male rats. The HPLC traces in Figure 1 suggested that the formation of both metabolites was supported by both NADPH and NADH (cofactor concentrations 1 mM). Kinetic studies were undertaken to derive the K , and V,, values for product formation, and as summarized in Table 1, NADPH-mediated metabolism appeared more efficient than that supported by NADH. Thus, the Michaelis constants for NADPH-supported paraoxon and 4-nitrophenol formation were 9 f 1 pM and 18 f 3 pM, respectively. By comparison, the constants from NADH-
140
m/z Figure 2. Mass spectra obtained for (A) 4-nitrophenol and (B) paraoxon formed in microsomal incubations with NADH as cofactor. CAD-MS was used to obtain spectrum A.
mediated metabolism were 26 f 6 pM and 53 f 10 pM, respectively. In contrast, Vm, values were quite similar regardless of whether NADPH or NADH was employed (Table 1). V,,lK, ratios, which reflect enzymic efficiency, were calculated for each pathway and suggested that, under these conditions, the NADPH-mediated reactions are about 3-fold more efficient than those mediated by NADH and that this is due primarily to differences in enzyme &ty (reflectedby the K, constant& Paraoxon formation was about 2.5-fold more efficient than 4-nitrophenol formation, regardless of the cofactor employed. MS of Parathion Metabolites. MS was used to confirm that the products of NADH-mediated microsomal parathion metabolism were 4-nitrophenol and paraoxon (Figure 2, panels A and B, respectively); CAD-MS was used in the case of the former compound. Thus, MS analysis indicated that the materials that were eluted in the HPLC separation procedure were indistinguishable from 4-nitrophenol and paraoxon (spectra not presented). The CAD-mass spectrum of authentic 4-nitrophenol exhibited fragments at m l z 140 (M 1, 17% of base peak), 123 (M 1- OH, base peak), 93 (M 1- OH -
+
+
+
Table 1. Kinetic Parameters of NADPH- and NADPH-Mediated Parathion Metabolism in Rat Liver Microsomes NADPH NADH Vmax VmaxlK, Km Vmax VmJKm KmbM) [nmoV(min*mgof protein)] [Ll(minng of protein)] MM) [nmoV(minmg of protein)] [L/(min*mgof protein)] paraoxon 9 f lQ 3.21 f 0.25 3.6 10-4 26 f 6 2.86 f 0.43 1.1x 10-4 2.58 f 0.37 1.4 10-4 53 f 10 2.26 f 0.48 4.3 x 10-6 4-nitrophenol 18 f 3 Data are mean f SE of estimates from three to seven individual microsomal fractions.
parathion metabolite
Parathion Biotransformation in Hepatic Microsomes
w
I5-O1
aa
Chem. Res. Toxicol., Vol. 7, No. 6,1994 796
crosomes from control rat liver exhibited very good correlations (r = 0.949 and 0.889, respectively) that were highly significant. Similar strong linear relationships ( r = 0.96) were obtained when data on NADH- and NADPHmediated parathion metabolism were correlated in hepatic microsomes from rats treated with inducing agents or when all the data were combined (control plus induced). Thus, 79-95% of the data variance ($) could be explained in this series of linear correlations between NADH- and NADPH-dependent activities. Inhibition of Microsomal Parathion Oxidations by Anti-NADPH-P46O-F&ductaseand Clotrimazole. The effects of an anti-NADPH-P450 reductase IgG preparation on NADPH- and NADH-dependent parathion oxidation in rat hepatic microsomes were determined. As shown in Figure 5, NADPH-supported metabolism of parathion to 4-nitrophenol and paraoxon was inhibited by 10 mg of IgGImg of microsomal protein by 39% (p < 0.02) and 35% (p < 0.02) relative to enzyme activities in the presence of preimmune IgG. Microsomal 4-nitrophenol and paraoxon formation from parathion mediated by NADH was decreased by 22% ( p < 0.05) and 28% (p < 0.05) from preimmune control activities. By comparison, the anti-NADPH-P450 reductase IgG produced decreases in the four principal testosterone hydroxylations to 3 5 4 6 % of activities in the presence of preimmune IgG (Figure 6). The antimycotic imidazole clotrimazole, a potent inhibitor of most P450 activities, was found to inhibit NADH- and NADPH-supported parathion oxidation in microsomes in a concentration-dependent manner (Figure 7). With NADH as cofactor, I C 5 0of ~ 10 pM and 4.5 pM against paraoxon and 4-nitrophenol formation were estimated; the corresponding IC5,,s were 24 pM and 2.9 pM, when NADPH was employed as the cofactor. Thus, 4-nitrophenol formation from parathion appeared slightly more susceptible than paraoxon formation to inhibition by clotrimazole. This appeared to be the case regardless of the cofactor used in the incubations. Finally, the noninvolvement of either cytochrome b5 or NADH-cytochrome bg reductase in the NADH-dependent reaction was established by the addition of sodium cyanide (1 mM) or potassium ferricyanide (1 mM) to microsomal incubations. Neither compound had any effect on parathion metabolism supported by NADH (not shown). Parathion Oxidation by P450 2B1 in a Reconstituted Enzyme System in Vitro. Studies were undertaken in a reconstituted enzyme system incorporating P450 2B1 and NADPH-PI50 reductase and indicated that NADH (1mM) supported parathion oxidation at a
aa aa
paraoxon Cnitrophenol paraoxon Cnitrophenol
NADPH NADH Figure 3. Microsomal parathion metabolism in liver isolated from untreated male rats (a),and rats pretreated for 3 days with phenobarbital (hatched 0,slanting down to the right), dexamethasone (O),or p-naphthoflavone (hatched 0,slanting
up to the right), as described under Materials and Methods. aSignificant difference from activity in untreated liver p < 0.01.
+
NO, 78%), and 65 (M 1- OH - NO - CO, 23%). The mass spectrum of paraoxon exhibited fragments at m l z 276 (M 1, base peak) and 246 (M 1 - NO, 66%). Correlation between NADPH- and NADH-Mediated Parathion Metabolism in Rat Hepatic Microsomes. In view of the finding that NADH appeared to support P450-like oxidation of parathion in rat liver, further studies were undertaken. Rates of NADH- and NADPH-dependent metabolism were measured in hepatic microsomes from control and differently-pretreated rats. Compared with control, pretreatment of rats with phenobarbital produced approximate 2.4-fold increases in the rates of NADPH-supported conversion of parathion to paraoxon (Figure 3). Corresponding increases in the NADH-mediated reactions were 3.1- and 3.7-fold, respectively, over control. Pretreatment of rats with dexamethasone also increased the activities of the NADPHand NADH-dependent pathways of parathion metabolism in a similar fashion to phenobarbital. However, 6-naphthoflavone, an inducer of P450s 1Al and 1A2 (191, was without significant effect on parathion oxidation (Figure 3). Similar ratios of paraoxon:4-nitrophenol formation were estimated in all of the fractions when NADPH was employed as the cofactor; paraoxon formation was more extensive than 4-nitrophenol formation in each case. In contrast, however, the NADH-mediated formation of 4-nitrophenol appeared to be favored slightly over paraoxon formation in microsomes from phenobarbital- and dexamethasone-induced rat liver. As shown in Figure 4, the linear relationships between NADH- and NADPHmediated paraoxon and 4-nitrophenol formation in mi-
+
+
=
d
PARAOXON FORMATION(NADPH)
E
4-NITROPHENOLFORMATION(NADPH)
5
Figure 4. Linear correlations of NADH- and NADPH-mediatedparathion metabolism to (left) paraoxon and (right) 4-nitrophenol. Units of metabolite formation [nmoUminmg of protein)] were omitted from axis labels for clarity.
796 Chem. Res. Toxicol., Vol. 7, No. 6,1994 w
Murray and Butler Table 2. NADPH- and NADH-Mediated Parathion Metabolism by a Reconstituted Enzyme System Containing Purified P450 2B1 4-nitro phenol paraoxon [nmol/(min-nmol [nmoV(min-nmol incubation of P450)lU of P450)la NADPH as Cofactor complete 5.50 f 0.27 9.97 f 0.19
6.0
1
NADPH NADH
NADH as Cofactor complete 4.82 f 0.16 minus P450 2B1
