Chem. Res. Toxicol. 1996,8, 82-91
82
Isozyme-Selective Metabolic Intermediate Complex Formation of Guinea Pig Hepatic Cytochrome P450 by N-Aralkylated Derivatives of 1-Aminobenzotriazole Christopher J. Sinal and John R. Bend* Department of Pharmacology and Toxicology, University of Western Ontario, London, Ontario, Canada N6A 5C1 Received March 21, 1994@ The capacity for metabolic intermediate (MI) complex formation as a mechanism of action for the isozyme-selective cytochrome P450 (P450) inhibitors N-benzyl-l-aminobenzotriazole (BBT), N-(a-methylbenzy1)-l-aminobenzotriazole(aMB), and N-(a-ethylbenzy1)-l-aminobenzotriazole (aEB) was investigated in hepatic microsomes from untreated, phenobarbital-induced, and /haphthoflavone-induced guinea pigs. Similar to other complex forming amines, MI complex formation was observed a s an absorbance maximum a t approximately 455 nm by optical-difference spectroscopy, was dependent upon incubation with NADP(H), and was dissociable by t h e addition of 50 pM potassium ferricyanide. MI complexes formed by BBT, aMB, and a E B were also dissociable by sedimentation and resuspension, as well as in the presence of limiting concentrations of NADP(H). Maximal complexation with the three compounds was observed in microsomes from phenobarbital-induced guinea pigs where the initial rate of complex formation was dependent upon inhibitor concentration and apparent K , values of 108 f 44,338 & 96, and 84 f 15pM for BBT, aMB, and aEB,respectively, were found. Inclusion of 1mM glutathione in the incubation mixtures had a significant attenuating effect upon complex formation, suggesting t h e involvement of a n electrophilic, reactive intermediate. Complex formation was not observed with the three inhibitors in pulmonary microsomes from either guinea pigs or rabbits. MI complexation is not likely to contribute t o the mechanism-based inactivation of guinea pig hepatic P450 2Bx, t h e homologue of rabbit P450 2B4, due to the irreversible inactivation of this isoform a t very low inhibitor concentrations, the lack of glutathione attenuation of this destruction, the instability of formed MI complexes, and t h e absence of MI complex formation with guinea pig or rabbit pulmonary P450.
Introduction The cytochrome P450 (P45O)l monooxygenase system catalyzes a wide variety of oxidative reactions important in the metabolism of a large number of lipophilic compounds of endogenous and exogenous origin. Examples of typical substrates include sterols, fatty vitamins, eicosanoids, fatty acids, drugs, pesticides, and carcinogens ( I 1. Oxidation of xenobiotics results in products of increased polarity that are more readily excreted directly or after conjugation with water-soluble agents such as glucuronic acid, sulfuric acid, or glutathione (2). Due to the wide variety of reactions catalyzed by P450 (31, inhibition or destruction of individual isozymes by exogenous compounds can not only alter the metabolism of other substances, but may also result in alterations in essential physiological processes. Many drugs and other chemicals have been identified as inhibitors of P450 (46). Of considerable interest are compounds that, as a result of oxidative metabolism, generate metabolic intermediates (MI) capable of forming a complex with the heme of the P450 isozyme which catalyzed their formation. The resultant stable ferrous-heme complex typi-
* Author to whom correspondence should be addressed.
Abstract published in Advance ACS Abstracts, November 15,1994. B T ,l--nobenzotrimole; &B, N-(a-ethylbenzy1)-l-aminobenzotriazole;uMB,N-(a-methylbenzyl)-l-aminobenzotriazole; #?-NF, #?-naphthoflavone;BBT, N-benzyl-l-aminobenzotriazole; BD, 1,3-benzodioxole; ERF, 7-ethoxyresorufin O-dealkylation; MI, bbmdate; p450, *home p450; pB, so&um phenobarbital, PRF, 7-pentoxyresorufin O-dealkylation. @
1 Abbreviations:
cally converts the P450 into a catalytically nonfunctional state which can be stable both in vitro and in vivo (6). Many alkylamine compounds including orphenadrine (71, macrolide antibiotics such as troleandomycin (81, the tricyclic antidepressants imipramine, desipramine, amitriptyline, and nortriptyline (91,and the prototypical P450 inhibitor, proadiphen (SKF 525-A) (101, form MI complexes and effectively inhibit P450 in an isozymeselectivehpecific manner. When visualized by difference spectroscopy, these amine MI complexes display a characteristic absorbance maximum in the 450-460 nm range, the magnitude of which is dependent upon the proportion of P450 isozyme(s) present that form a complex with the MI (11). Three N-aralkyl derivatives of l-aminobenzotriazole (ABT),N-benzyl-l-aminobenzotriazole(BBT),N-(a-methylbenzy1)-l-aminobenzotriazole (aMB), and N-(a-ethylbenzyl)-l-aminobenzotriazole(&B), have previously been demonstrated to be potent and isozyme-selective inactivators of rabbit and guinea pig hepatic and pulmonary P450 in our laboratory (I2-16), both in vitro and in vivo. These compounds have also proved valuable as sensitive biochemical probes for the study of P45O-dependent physiological biotransformations (17). BBT, aMB, and prior to dB require NADP(H)-dependent inhibiting P450 and are consequently classified as mechanism-based inhibitors. The primary objectives of the present work were to determine the capacity and specificity of BBT, aMB, and aJ3B for the in vitro formation
0893-228x/95/2708-0082$09.00/00 1995 American Chemical Society
Chem. Res. Toxicol., Vol. 8, No. 1, 1995 83
Complexation of P45O by BBT, aMB, and aEB of MI complexes in hepatic microsomes of untreated, phenobarbital (PB)-induced, and p-naphthoflavone CpNF)-induced guinea pigs. The concentration dependence of complex formation was utilized as a n index to determine the apparent K, and hm(,,bs) of complex formation. In addition, the spectral characteristics of the interaction of BBT, aMB, and aEB with guinea pig hepatic microsomes were also compared with those obtained for the interaction of 1,3-benzodioxole(BD) and l-aminobenzotriazole (ABT), known mechanism-based inhibitors of P450 monooxygenase activities.
m y ; NN
ABT 1-aminobenzotriazole
I NH2
BBT N-benzyl-1-aminobenzotriazole
Materials and Methods OMB N- (a-methylbenzyl)- 1 -aminobenzotriazole
Materials. ABT, BBT, aMB, and aEB were synthesized and purified as previously described (12, 14). NADP(H) was purchased from Sigma Chemical Co. (St. Louis, MO);P-NF and BD were from Aldrich Chemical Co. (Milwaukee, WI); and GSH,
potassium ferricyanide, PB, MeZSO, sodium dithionite, and all other chemicals (reagent grade or better) were from BDH (Toronto, ON, Canada). Animal Treatment and Preparation of Microsomes. Male Hartley guinea pigs (250-300 g; Charles River Laboratories, St. Constant, PQ, Canada) were used. Some of the animals were treated intraperitoneally with 80 mgkg PB (2% in saline) or 80 mgkg p-NF (2%in corn oil) daily for 4 days and sacrificed 24 h following the last injection by asphyxiation with COz. All animals were allowed free access to food (Purina guinea pig chow) and water throughout the treatment period. Hepatic microsomeswere prepared by differential centrifugation as previously described (18). Microsomal protein concentrations were determined by the method of Lowry et al. (19)with bovine serum albumin as standard. Specific P450 content was determined from the dithionite difference spectrum of carbon monoxide-saturated microsomes using a Beckman DU-65 spectrophotometer with E = 100 mM-km-l (20). Time Course of Complex Formation. All spectral determinations were done at 25 "C. Experimental samples contained 1 mg/mL (unless otherwise noted in text) hepatic microsomal protein from untreated, p-NF-treated, or PB-treated guinea pigs. The test compounds (structures in Figure 1)were dissolved in either MezSO (BD, BBT, aMB, and aEB) or distilled water (ABT)and 5 pL was added to the microsomal suspensionto give a final inhibitor concentration of 200 pM. MezSO was shown to have no effect on the spectral data at the concentrations used. The final cuvette volume was made up t o 1 mL with 0.1 M potassium phosphate buffer (pH 7.4). A base-line scan of zero absorbance was recorded between 400 and 500 nm using a Beckman DU-8 single beam spectrophotometer (reference). In some experiments, GSH (1 mM final concentration) was also included in the incubation mixtures. Complex formation was started by the addition of 0.1 M NADP(H) (10pL) in distilled water (final concentration = 1mM). In some experiments, the concentration of NADP(H)was varied between 0.05 and 1.0 mM by the addition of 10 pL of an appropriate stock solution. Repetitive scans were recorded at intervals of approximately 90 s for a period of 15-30 min. The amount of complex formed was calculated by measuring the absorbance change at the wavelength of maximal absorption of each complex (455-460 nm) relative to the absorbance change at 490 nm (an isosbestic reference point). Estimates of the proportion of total P450 complexed were derived using an extinction coefficient of E = 64 mM-'.cm-l(11). Dissociation of the ferrous-P450 complexes was followed spectrophotometrically (Beckman DU-8) after the addition of potassium ferricyanide (50 pM) to microsomes from PB-treated guinea pigs (2 mg/mL) that had been preincubated with 200 pM inhibitor and 1 mM NADP(H) (1mL final volume) for 45 min at 37 "C. A n initial base-line scan (430-500 nm) of zero absorbance for the ferrous-complexed microsomes was recorded as a reference. After the addition of ferricyanide (10 pL of 5 mM) followed by a 5 min incubation period, a second scan was
aEB N-(a-ethylbenzyl ) - 1-aminobenzotriazole
cnZcn3
ao)
BD
0
1,3-benzodioxole
Figure 1. Chemical structures of the mechanism-based inhibi-
tors of P450 used in this study.
recorded over the same wavelengths. Complex dissociationwas followed as a loss of maximal absorbance in the 455 nm range. Determination of Spectral Binding to P460. All determinations were performed at 25 "C. Experimental samples contained 2 mg of guinea pig hepatic microsomal protein in a volume of 1.0 mL in 0.1 M potassium phosphate buffer (pH 7.4) in a spectrophotometercuvette. A base-line scan was recorded between 350 and 500 nm. Various dilutions of the inhibitors (BBT, aMB, and aEB) were prepared in MezSO. Sficient volumes of these dilutions were added sequentially to the cuvette to give the desired inhibitor concentrations (2.5-400 pM). The total final volume of inhibitorholvent added was 20 pL. Scans of 350-500 nm were recorded immediately after each addition. The magnitude of spectrally assayable inhibitorP450 binding was quantitated as the change in absorbance at 387 nm minus the change in absorbance at 425 nm relative to the initial reference scan. The apparent spectral K, and AA,, values were determined by double-reciprocal plots. Kinetics of Complex Formation. Experimental samples contained 1 mg/mL of hepatic microsomal protein from PBtreated guinea pigs in 0.1 M potassium phosphate buffer (pH 7.4). The test compounds(BBT, aMB, and aEB) were dissolved in MeZSO, and 5-10 pL was added to the samples (final concentration = 10-500 pM; 10-1000 pM for aMB). A baseline scan was recorded between 400 and 500 nm using a Beckman DU-8 single beam spectrophotometer (reference). Complex formation was started by the addition of 0.1 M NADP(H) (10 pL) in distilled water (final concentration = 1mM). The rate of complex formation was determined by least-squares linear regression of the linear portion of the absorbance change versus time curve. For BBT, aMB, and aEB, the amount of complex formed after 25 s was recorded. The rate of complex formation was determined by dividing the change in absorbance by 25 s. Apparent kinetic constants [K,,Dmax(obs)] were determined using double-reciprocal plots. Washing Protocol for Inhibited Microsomes. Experimental samples contained 2 mg/mL guinea pig hepatic microsomal protein. The test compounds (BD, BBT, aMB, and aEB)
Sinal and Bend
84 Chem. Res. Toxicol., Vol. 8,No. 1, 1995
aMB
BBT
I
0.01 AU
430
I
1 I
I
440
450
460
I
0.01 AU
I
I
1
470
450
490
I
500
(W aEB
BD
f
\
1
1
I
440
450
460
1
0.01 AU
I
I
1
470
480
490
0.01 AU
1
I
430
500
I
1
1
I
I
450
460
470
480
490
1
430
440
500
k 0" k Figure 2. Representative spectra obtained upon the incubation of 200 pM BBT, aMB, aEB, or BD with 1 mM NADP(H) in PBtreated guinea pig hepatic microsomes. Repetitive scans were recorded at approximately 90 s intervals. Experimental procedures are described in Materials and Methods.
were dissolved in MezSO, and 15 pL was added to the samples to give a final inhibitor concentration of 200 pM. The h a l volume was made up to 2.97 mL with 0.1 M potassium phosphate buffer (pH 7.4). Complex formation was started by the addition of 0.1 M NADP(H)(30 pL) in distilled water (final concentration = 1mM) and was allowed to proceed for 10 min at 37 "C. At the end of the incubation time, aliquots of the samples were cooled on ice, followed by either spectrophotometric determination of spectral complex (unwashed) or washing by ultracentrifugation (washed). Briefly, washed samples were sedimented at 100000 RPM (412160g)for 10 min at 4 "C (BeckmanTL-100 ultracentrifuge, TLA 100.3rotor) followed by resuspension and resedimentation in 0.1 M potassium phosphate buffer (pH 7.4). In some experiments, the microsomal samples were reduced by the addition of 10 mg of sodium dithionite prior to washing. Control samples for each group were treated identically to the samples but did not contain NADRH). Washed and unwashed inhibited microsomal samples were added to a spectrophotometric cuvette to give a final concentration of 2 mg/mL. A base-line scan of zero absorbance was recorded between 400 and 500 nm. Control samples were used to obtain reference scans. Repetitive scans were recorded, and the amount of complex present was calculated by measuring the absorbance change at the wavelength maximal absorption of each complex (455-460 nm) relative to the absorbancechange at 490 nm. Statistics. Statistical analyses were performed on raw data using ANOVA. This was followed by Dunnett's test for multiple comparisons where appropriate (Tables 1, 2, and 4). The criterion for significance was chosen as p < 0.05.
Results Metabolic Intermediate Complex Formation. Formation of MI complexes was observed upon incubation of BD, BBT, aMB, or aEB, but not ABT, with guinea pig hepatic microsomes containing P450 and NADP(H). The maximal absorbance of the spectral complexes of the N-aralkylated derivatives of 1-aminobenzotriazole occurred in the 455-457 nm region (Figure 2). Representative scans were recorded from incubations with hepatic microsomes from PB-treated guinea pigs, as these pro-
vided the greatest rate and extent of complex formation for these inhibitors. With BD, the peak absorbance occurred at approximately 460 nm. Induction with PB or P-NF had an affect on both the rate and maximal extent of complex formation observed for BBT, aMB, aEB, or BD. Hepatic microsomes from PB-treated animals had the fastest rate and greatest extent of complex formation in all cases (Figure 3). At equimolar concentrations (200 pM), the ABT analogues formed complexes a t a faster rate than was observed for BD. Complex formation was minimal with BBT and aEB and was not detected with aMB, using microsomes from P-NF-treated guinea pigs (Figure 3). MI complex formation with BD was fastest and greatest in magnitude with microsomes from PB- and P-NF-induced guinea pigs. The extent of complex formation with BBT, aMB, and &B apparently reflected a selectivity for complexation with a PB-inducible P450 isozyme(s) in guinea pig liver. In contrast, the extent of complex formation with BD reflected the total P450 content (PB > P-NF > untreated) of the microsomal samples. The order of magnitude of NADP(H)-dependent MI complex formation with BBT, aMB, and &B in hepatic microsomes was PB-treated > untreated > ,8-NF-treated (Table 1).In hepatic microsomes from PB-treated guinea pigs, 2.4- (BBT), 3.0- (aMB), and 3.9-fold (&B) increases in the total (% complexed x [P4501) amount of MI complex formed were noted. On the other hand, P-NF induction markedly decreased MI complex formation with all three compounds. Spectral Binding with Cytochrome P450. The inhibitors BBT, aMB, and aEB produced a type I spectral change, characterized by the appearance of an absorption peak a t approximately 387 nm and a trough near 425 nm, when added to hepatic microsomes from PB-treated guinea pigs. The magnitude of the spectral change was dependent upon the concentration of the added compounds (Figure 4). Apparent spectral dissociation constants (K,)were calculated from the data by double-
Chem. Res. Toxicol., Vol.8, No.1, 1995 86
Complexation of P450 by BBT, aMB, and aEB BBT
0
200
151 aMB
400
600
800
1000
0
200
400
600
800
1000
800
1000
Time (s)
Time (s)
207 BD
0
200
400
600
800
1000
0
200
400
600
Time (s) Time (s) Figure 3. Time course of MI complex formation by 200 pM BBT, aMB, aEB, or BD with PB-treated (O), ,8-NF-treated (O), and control (A) guinea pig hepatic microsomes.The change in absorbance was measured as AA465-490 for BBT, aMB, and aEB and +60-490 for BD. Each point is representative of the mean of at least three experiments with microsomes pooled from 4-5 animals. Expenmental procedures are described in Materials and Methods. Table 1. Percentage of Total Cytochrome P450 Involved in NADP(H)-DependentMI Complex Formation with BBT, aMB, and aEB in Hepatic Microsomes from Untreated, b-NF-Treated,and PB-Treated Guinea Pigsa % of total P450 complexedb
inhibitor BBT aMB crEB
untreated
B-NF-induced
PB-induced
13 f 1 9*4 9*1
5 f 2c
20 3 c 16 f 2c 21 1c
NDd
*
3
* 1c
* *
Values represent the means SD (n = 3-4). Determined from the absorbance change at 455 nm relative to 490 nm using E = 64 mM-l-cm-l; microsomal P450 content (nmol/mg) was determined as 0.69 f 0.06,untreated; 1.01f 0.09,#?-NF;1.12& 0.09, PB. Significantly different from untreated. dND = not detected (A4< 0.001). a
reciprocal plots yielding values of 22 f 5, 16 f 2, and 5 f 1pM for BBT, aMB, and aEB, respectively (Table 2). The corresponding iL4- values were determined to be 0.0047 f 0.002, 0.0049 f 0.004, and 0.0058 f 0.005, respectively. Kinetics of Complex Formation. To compare the kinetics of NADP(H)-dependent MI complex formation for BBT, aMB, and aEB, various concentrations (101000 pM) were incubated with NADP(H) and hepatic microsomes from PB-treated guinea pigs. Initial rates of complex formation were observed to be concentration dependent and appeared to approach near-maximal values over the concentration range studied. Values for the apparent kinetic constants of complex formation were derived by double-reciprocal plots (Figure 5). All of the test compounds displayed linear relationships over the moderate to high concentration values studied. At very
low inhibitor concentrations, the relationships became curvilinear, with large decreases in the rate of complexation associated with relatively small changes in concentration (data not shown). As a consequence of this behavior, the apparent kinetic constants (Table 2) were derived using those concentrations at which linearity was observed. The apparent K, values decreased in the order aMB > BBT > aEB. The smaller apparent K, values for BBT and aEB when compared with aMB suggest a higher substrate aEnity or efficiency of complex formation. Dissociation of MI Complexes. The addition of 50 pM potassium ferricyanide to microsomes from PBtreated guinea pigs that had been incubated for 45 min with 200 p M inhibitor and 1mM NADP(H) resulted in a loss in absorbance which was maximal at approximately 455 nm for BBT, aMB, and aEB (not shown). This indicates that the MI complexes formed with these compounds are more stable with the ferrous form of the heme than with its ferric form. In contrast, the addition of ferricyanide to samples that had been incubated with BD and NADP(H) did not result in a loss of absorbance in the region in which the complex displays maximum absorption (approximately 460 nm). Effect of Washing and NADP(H) Concentration. To further investigate the stability of MI complexes of the inhibitors with P450, samples of guinea pig hepatic microsomal protein were incubated with NADP(H) and the inhibitors a t a concentration of 200 pM for 10 min. At the end of the incubation period, half of the sample volume was washed to remove unreacted inhibitor and
86 Chem. Res. Toxicol., Vol. 8, No. 1, 1995
Sinal and Bend
BBT
aEB
aMB
II
0.06
"d
o,02ii 0.01
;
0
0 0
100
200
300
400
500
,
,
,
I
,
,
,
,
,
0 0
100
200
300
400
500
0
100
200
300
400
500
Concentration (pM) Figure 4. Spectral binding curves of BBT, aMB, and aEB with P450 of hepatic microsomes from PB-treated guinea pigs in the absence of NADP(H). Each point is representative of the mean of at least three experiments with microsomes pooled from 4-5 animals. Experimental procedures are described in Materials and Methods. Table 2. Apparent Spectral Constants for Binding to Microsomal P450 in the Absence of NADP(H) (IC. and AA& and for MI Complex Formation in the Presence of NADP(H) lg,and U - ( ~ ~ ) I for BBT, aMB, and aEB Using Hepatic Microsomes from PB-Treated Guinea Pig@
Table 3. Effect of Resuspension and Resedimentation of Microsomes on the Maintenance of MI Complexes Formed by 200 pM BBT, aMB,aEB, and BD with Guinea Pig Hepatic Microsomes (2 mg) Incubated with 1 mM NADP(H)
inhibitor BBT
25 f 3 BBT 22 f 5 0.049 f 0.004 108 f 44 21f4 aMB 16 f 2 0.047 f 0.002 338 f 966 21 f 2 (rEB 5 & l c 0.058f0.005 9 7 f 4 8 Determined as described in the Materials and Methods section. Reported values are the means f SD (n = 3-4). Significantly different from BBT and aEB. CSignificantlydifferent from aMB and BBT.
condition M m g of protein % lossa unwashed 0.012 i 0.001 0.002 f 0.001b washed 81 0.010 f 0.002 aMB unwashed ND > 95 washed 0.009 f 0.002 aEB unwashed 0.001 i 0.001b 89 washed BD unwashed 0.010 f 0.003 0.009 f 0.003 10 washed % loss of mean. Significantly different from unwashed. Not detected (
