Differential Effects of Naturally Occurring Isothiocyanates on the Activities of Cytochrome P450 2E1 and the Mutant P450 2E1 T303A

Archives of Biochemistry and Biophysics Vol. 391, No. 1, July 1, pp. 99 –110, 2001 doi:10.1006/abbi.2001.2390, available online at http://www.idealibrary.com on

Differential Effects of Naturally Occurring Isothiocyanates on the Activities of Cytochrome P450 2E1 and the Mutant P450 2E1 T303A Rosa L. Moreno,* Theunis Goosen,* ,1 Ute M. Kent,* Fung-Lung Chung,† and Paul F. Hollenberg* ,2 *Department of Pharmacology, The University of Michigan, Ann Arbor, Michigan 48109; and †Division of Carcinogenesis and Molecular Epidemiology, American Health Foundation, Valhalla, New York 10595

Received February 12, 2001, and in revised form March 23, 2001; published online June 4, 2001

benzoic acid when compared to the wild-type enzyme. The effects of benzyl (BITC) and phenethyl isothiocyanate (PEITC) on the activity of a P450 2E1 mutant where the conserved threonine at position 303 was replaced with an alanine residue (P450 2E1 T303A) were examined. PEITC inactivated the mutant enzyme with a K I of 1.6 ␮M. PEITC also inactivated the wildtype P450 2E1 as efficiently with a K I of 2.7 ␮M. The inactivation was entirely dependent on NADPH and followed pseudo-first-order kinetics. Previously we reported the mechanism-based inactivation of wild-type P450 2E1 by BITC with a K I of 13 ␮M. In contrast to the wild-type enzyme, the P450 2E1 T303A mutant was not inactivated by BITC but it was inhibited in a competitive manner with a K i of 3 ␮M. The binding constants determined by spectral binding studies were similar for both enzymes. The binding of BITC produced characteristic Type I spectral changes in the wild-type and mutant enzyme. A radiolabeled BITC metabolite bound to P450 2E1 and to P450 2E1 T303A when both enzymes were incubated with [ 14C]BITC and NADPH. Whole protein electrospray ion trap mass spectrometry indicated that a mass consistent with one molecule of benzylisocyanate and oxygen was adducted to the wild-type enzyme. The mass adducted to the T303A mutant was consistent with the addition of one hydroxylated BITC or of one benzylisocyanate moiety and one sulfur molecule. Analysis of the metabolites of BITC indicated that each enzyme produced similar metabolites but that the mutant enzyme generated significantly higher amounts of benzaldehyde and

1 Present address: Potchefstroom University for Christian Higher Education, Potchefstroom, South Africa. 2 To whom correspondence should be addressed at Department of Pharmacology, 2301 Medical Science Research Building III, 1150 West Medical Center Drive, Ann Arbor, MI 48109-0632. E-mail: [email protected].

0003-9861/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

© 2001 Academic Press

Key Words: P450 2E1; benzyl isothiocyanate; phenethyl isothiocyanate; cytochrome P450; mechanismbased inactivation; enzyme active site.

Benzyl- and phenethyl isothiocyanate belong to a group of naturally occurring compounds that are found in abundance in cruciferous vegetables such as cabbage, watercress, broccoli, and cauliflower. Isothiocyanates are derived from their parent glucosinolates through a hydrolytic process mediated by the enzyme myrosinase (1). These compounds have been frequently studied for their prominent anticarcinogenic effects that have been attributed to various properties such as the capacity to induce enzymes involved in detoxification reactions (glutathione S-transferase and quinone reductases) or in DNA repair and for their inhibitory and inactivating effects on Phase I enzymes such as the cytochrome P450s (1– 4). The anticarcinogenicity of isothiocyanates has been demonstrated in various animal models, mostly in the context of nitrosamine-induced tumorigenesis. BITC 3 was found to inhibit the forma3

Abbreviations used: BITC, benzyl isothiocyanate; PEITC, phenethyl isothiocyanate; P450, cytochrome P450; 7-EFC, 7-ethoxy-4(trifluoromethyl)coumarin; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NDEA, N-nitrosodiethylamine; 2-EN, 2-ethynylnaphthalene; 2-NA, 2-naphthylacetic acid; 2-AF, 2-aminofluorene; 7-HFC, 7-hydroxy-4-(trifluoromethyl)coumarin; BSA, bovine serum albumin; BIC, benzylisocyanate; DMSO, dimethyl sulfoxide; BAM, benzylamine; BZA, benzoic acid; BAH; benzaldehyde, DBU, N⬘, N-dibenzylurea; DBT, N⬘,N-dibenzylthiourea; BFU, N-benzyl-N⬘(2-fluorenyl)urea; BFT, N-benzyl-N⬘-(2-fluorenyl)thiourea; 7-EC, 7-ethoxycoumarin; p-NP, para-nitrophenol; 2-NITC, 2-naphthylisothiocyanate; 2-NIC, 2-naphthylisocyanate; 2-AN, 2-aminonaphthalene; DNU, N⬘,N-dinaphthylurea; PVDF, polyvinylidene difluoride; 99

100

MORENO ET AL.

tion of neoplasms induced by N-nitrosodiethylamine (NDEA) in the forestomach of A/J mice (5). In another study, BITC was found to inhibit mammary tumor formation in Sprague–Dawley rats induced by 7,12dimethylbenz[a]anthracene (6). PEITC prevented lung tumorigenesis associated with 4-(methylnitrosamino)1-(3-pyridyl)-1-butanone (NNK) treatment in rodents while BITC showed no effect in these studies. It was proposed that PEITC mediated this protection by inhibiting the P450 enzyme involved in the bioactivation of NNK to reactive intermediates (7–9). Previous studies have demonstrated a strong competitive inhibitory effect with a slow suicide inhibitory component of PEITC on P4501A2, and this enzyme was demonstrated to make a major contribution to the metabolism of NNK in human liver microsomes (10, 11). P4502E1 is of particular interest due to its possible involvement in chemically induced genotoxicity. This isozyme is abundant in human liver and has been implicated in the metabolic activation of carcinogens such as NDEA, N-nitrosodimethylamine, N-nitrosonornicotine, and N-methyl-N-benzylnitrosamine (12, 13). Previously, our laboratory reported that BITC inactivated P450 2E1 in a mechanism-based manner (14). The loss in enzymatic activity was primarily due to the binding of a reactive intermediate of BITC to the P450 2E1 apoprotein rather than the destruction of the heme moiety (14). Here we report a similar effect of PEITC on P450 2E1, as well as differential effects of BITC and PEITC on the activity of a mutated P450 2E1 where threonine 303 was replaced with alanine. This mutant provided us with the opportunity to explore the relevance of this residue in the process of inactivation by these two isothiocyanates. The threonine 303 residue corresponds to the conserved threonine identified by sequence alignments with the crystal structures of bacterial P450 enzymes such as P450 cam and BM-3. A role for this residue in substrate interactions and orientation and in catalysis by serving as a possible proton donor in an acid– base reaction are some of the functions that have been proposed (15–17). Previously, Roberts et al. found that 2-ethynylnaphthalene (2-EN) inactivated P4502B4 and P4502B1 in a mechanismbased manner. Inactivation occurred as a result of the formation of the reactive ketene intermediate that covalently modified the protein in the I helix region containing the conserved threonine 302 residue (18, 19) which corresponds to the threonine 303 in P450 2E1. Further studies with a P450 2B4 T302A mutant showed that while the ability to metabolize 2-EN to the ketene which was then hydrolyzed to 2-naphthylacetic

DLPC, dilauroyl-L-␣-phosphatidylcholine; ESI-LC-MS, electrospray ionization–liquid chromatography–mass spectrometry; TFA, trifluoroacetic acid; DTT, dithiothreitol.

acid was intact, the rate of inactivation and covalent binding were decreased in this enzyme, providing additional evidence for the involvement of the threonine residue in the inactivation event (20). In our system a similar phenomenon was observed, while BITC inactivated P4502E1 with a K I of 13 ␮M, the P4502E1 T303A mutant was only competitively inhibited by this isothiocyanate. Interestingly, both P4502E1 and the mutant were equally sensitive to the inactivating effects of PEITC. EXPERIMENTAL PROCEDURES Materials. BITC was from Aldrich Chemical Co. (Milwaukee, WI). 7-Ethoxy-4-(trifluoromethyl)coumarin (7-EFC) was from Molecular Probes (Eugene, OR). Catalase, purified from bovine liver, NADPH, p-nitrophenol (p-NP), and dilauroyl-L-␣-phosphatidylcholine (DLPC) were from Sigma Chemical Co. (St. Louis, MO). Topp3 Escherichia coli cells were obtained from Stratagene (La Jolla, CA). HPLC-grade acetonitrile was from Fischer Scientific (Pittsburgh, PA). [ 14C]BITC labeled at the ␣ carbon (56 mCi/mmol) was synthesized as previously described (21). Benzylamine, benzaldehyde, and benzoic acid were from Aldrich Chemical Co. N⬘,N-dibenzylurea (DBU), N⬘,N-dibenzylthiourea (DBT), N-benzyl-N⬘-(2-fluorenyl)urea (BFU), and N-benzyl-N⬘-(2-fluorenyl)thiourea (BFT) were provided by Dr. M.-S. Lee (Michigan Cancer Foundation). Enzymes. The cDNAs for rabbit P450 2E1 and P450 2E1 T303A (provided by Dr. M. J. Coon, University of Michigan) were expressed in E. coli cells. Expression and purification of the proteins were according to published methods (22). The cDNA for rat NADPH-P450 reductase within the expression plasmid pOR263 (23) was expressed in E. coli Topp3 cells. Expression and purification of the protein was as described by Hanna et al. (24). Inhibition of P450 2E1 T303A by BITC. P450 2E1 T303A was reconstituted with reductase at a 1:1 molar ratio (1.7 nmol of each enzyme) and lipid (300 ␮g/mL) for 25 min at 4°C. Following reconstitution, catalase (12,600 U/mL) and 0.5 M potassium phosphate buffer (pH 7.4) were added to a final concentration of 2500 U/mL and 50 mM, respectively. Primary reactions contained 1.9 ␮M P450 2E1 T303A and reductase together with increasing concentrations of BITC (5, 10, 20, 30, and 50 ␮M in a final concentration of 1% methanol). Control samples contained methanol instead of BITC. Reactions were initiated by adding NADPH to a final concentration of 1.2 mM and incubated at 30°C. The activity remaining at 0, 1, 2, 3, 4, and 8 min following the addition of NADPH was determined by measuring the O-deethylation of 7-EFC to 7-HFC spectrofluorometrically (25) on a SLM-Aminco Model SPF-500C spectrofluorometer (SLM-Aminco, Urbana, IL) with excitation at 410 nm and emission at 510 nm. For this assay, 20-pmol aliquots of P4502E1 T303A were transferred from the primary reaction mixture to secondary reaction mixtures containing 0.1 mM 7-EFC and 0.2 mM NADPH and 40 ␮g/mL BSA in 50 mM potassium phosphate buffer (pH 7.4) in a final volume of 1 mL. The reaction mixtures were incubated for 15 min and were terminated by adding 0.33 mL of ice-cold acetonitrile. Competitive inhibition of P450 2E1 T303A-catalyzed O-deethylation of 7-EFC by BITC. P450 2E1 T303A (1.7 nmol) and reductase (1.7 nmol) were reconstituted as above with 300 ␮g/mL lipid. Catalase and 0.5 M potassium phosphate buffer (pH 7.4) were added to a final concentration of 2500 U/mL and 50 mM, respectively. For this assay the P450 (40 pmol/sample) was incubated simultaneously with different concentrations of 7-EFC (15, 25, 50, 100, and 125 ␮M) and BITC (0, 5, 10, 15, and 20 ␮M). Reactions proceeded for 60 min at 30°C and were terminated with 500 ␮L of ice-cold acetonitrile. The amount of HFC produced was determined spectrofluorometrically as described above. The K I for the BITC inhibition and the K m for

EFFECTS OF ISOTHIOCYANATES ON P450s 2E1 AND 2E1 T303A 7-EFC were calculated from a replot of the inhibitor concentration and the K m app. The values for K m app were obtained from a doublereciprocal plot of the substrate concentration versus the velocities of the reactions. Inactivation of P450 2E1 T303A by PEITC. Reconstitution and reaction conditions were as described above for BITC. Primary reactions contained increasing concentrations of PEITC (0.5, 1.25, 2.5, 5, and 10 ␮M in a final concentration of 1% methanol). The 7-EFC O-deethylation activity remaining was determined 0, 1, 2, 3, 4, and 5 min following the addition of NADPH as described above. Spectral binding constant determination for the interaction of P450 2E1 T303A and BITC. P450 2E1 T303A (1 nmol) was reconstituted with 100 ␮g of lipid for 45 min at 4°C. The reconstituted protein was diluted to 0.5 ␮M with 50 mM potassium phosphate buffer (pH 7.4). The sample was divided equally into the reference and sample cuvettes. Samples were scanned from 350 to 500 nm on a DW2-UV/Vis spectrophotometer (SLM-Aminco) equipped with an OLIS spectroscopy operating system (On-Line Instrument Systems, Inc., Bogart, GA). Following addition of 1-␮L aliquots of BITC stock dissolved in methanol to the sample cuvette and 1-␮L aliquots of methanol to the reference cuvette, scans were recorded for BITC concentrations ranging from 0.5 to 6.4 ␮M. The binding constant, K s, was calculated from a plot of the inverse of the change in absorption (380 – 420 nm) versus the inverse of the BITC concentration. The K s was obtained from the x-intercept of the linear regression line (26). HPLC analysis of BITC metabolites produced by P450 2E1 and P450 2E1 T303A. P450 2E1 and P450 2E1 T303A were reconstituted with reductase at a 1:1 molar ratio as described above with 300 ␮g/mL lipid. Following reconstitution, catalase (12,600 U/mL) and 0.5 M potassium phosphate buffer (pH 7.4) were added to a final concentration of 300 U/reaction and 50 mM, respectively. Primary reactions contained 2 ␮M P450 and 40 or 200 ␮M [ 14C]BITC. In order to trap the reactive isocyanate species (BIC), 2-AF was used in some experiments at a concentration of 200 ␮M (27). Reactions were carried out for 20 or 60 min at 30°C in the presence or absence of NADPH. Reactions were quenched with 1 mL ice-cold ethyl acetate, vortexed, and centrifuged for 10 min. The organic phases were transferred to a new tube and the extraction was repeated two more times. The organic phase was slowly evaporated under N 2 after adding DMSO (50 ␮L) to prevent the evaporation of the more volatile metabolites. The ethyl acetate extraction was repeated two more times under basic conditions (pH 11) and then two times under acidic conditions (pH 3). Concentrated samples were analyzed on a Dionex BioLC with variable wavelength detector (Dionex Corporation, Sunnyvale, CA) with a Vydac C 18 column (Hesperia, CA). Metabolites were separated using a solvent system of A, 5% acetonitrile, 1% acetic acid, pH 4.2; and B, 80% acetonitrile, 1% acetic acid, pH 4.2. The gradient was linear from 0 to 100% B over 32 min. Known standards were chromatographed under identical conditions to allow for the identification and quantification of products. In some experiments the specific activity of [ 14C]BITC was used to quantify the radiolabeled metabolites. Otherwise, metabolites were identified and quantified by their absorption at 254 nm. Labeling of P450 2E1 and P450 2E1 T303A by [ 14C]BITC. P450 2E1 or P450 2E1 T303A (2 nmol) were reconstituted with 2 nmol reductase and 300 ␮g/mL lipid as described above. Catalase (12,600 U/mL) and 0.5 M potassium phosphate (pH 7.4) were added to a final concentration of 1270 U/mL and 50 mM, respectively. Primary reactions contained 1.8 ␮M P450 and 80 ␮M [ 14C]BITC in a total volume of 490 ␮L. Reaction mixtures were incubated for 4 and 20 min at 30°C after adding 1.2 mM NADPH. The samples were cooled on ice and dialyzed in Slide-A-Lyzer 10K cassettes (Pierce, Rockford, IL) against three 250-mL changes of buffer containing 50 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA, and 0.4% cholic acid, for a total of 20 h. This was followed by dialysis against 50 mM potassium phosphate, pH 7.4, 10% glycerol, and 0.1 mM EDTA for 3 h. After dialysis, the samples were concentrated to 100 ␮L in a

101

speed vacuum concentrator (Savant Instruments, Inc., Hicksville, NY) and the P450 concentrations were determined by CO difference spectra using a DW2-UV/Vis spectrophotometer (SLM-Aminco, Urbana, IL) equipped with an OLIS spectroscopy operating system (On-Line Instruments Systems, Inc.). Samples were electrophoresed using a 10 –20% polyacrylamide gradient gel and transferred to a PVDF membrane (Millipore, Bedford, MA) at 0.3 A for 2 h. The membrane was exposed to Biomax MR film supplemented with a Biomax intensifying screen (Eastman Kodak Co., Rochester, NY) for 2–3 weeks at ⫺80°C. Both the film and the Coomassie blue-stained membrane were analyzed by densitometry on a USB Sciscan 5000 (Cleveland, OH) densitometer equipped with Bioanalysis Software, Oberlin Scientific (Oberlin, OH). ESI-LC-MS Analysis of P450s 2E1 and 2E1 T303A. P450 2E1 or P450 2E1 T303A were reconstituted with reductase and lipid as previously described. The primary reaction mixtures contained 1 ␮M P450 2E1 or P450 2E1 T303A, 1 ␮M reductase, 30 ␮g DLPC, 43 units of catalase, and 30 –50 ␮M BITC in a total volume of 400 ␮L. The samples were incubated in the absence or presence of 1 mM NADPH for 10 –15 min at 30°C. The 7-EFC activity remaining was measured as described above. For ESI-LC-MS analysis 50 pmol of P450 in the reconstituted system was injected onto a 150 ⫻ 2.1-mm C 18 column (5 ␮m, Vydac) equilibrated with 40% CH 3CN and 0.1% TFA at a flow rate of 0.3 mL/min. The CH 3CN concentration was increased linearly to 50% during the first 10 min, to 60% by 22 min, to 70% by 23 min, and to 90% by 25 min. The column effluent was diverted for the first 10 min. The flow was then directed into the LCQ mass spectrometer (Thermoquest, Schaumburg, IL) and spectra were recorded. The retention times of the proteins in the reconstitution mixture were 12.7 min for short reductase, 16.2 min for the reductase, and 18.4 min for the P450. The ESI source conditions were optimized using myoglobin. The sheath gas was set at 90 (arbitrary units), the auxiliary gas was set at 30 (arbitrary units), the spray voltage was 4.2 kV, and the capillary temperature was 230°C. The protein envelopes were deconvoluted using the Thermoquest Excalibur 1.0 SR1 Qual Browser to obtain the mass associated with each protein envelope.

RESULTS

Inhibition of P450 2E1 T303A by BITC. The 7-EFC O-deethylation activity of purified P450 2E1 T303A was inhibited by BITC in the reconstituted system (Fig. 1). The P450 2E1 T303A activity decreased with increasing concentrations of BITC. However, this decrease was not time dependent. To determine if the observed inhibition was competitive, P450 2E1 T303A was incubated simultaneously with different concentrations of BITC together with 7-EFC as described under Experimental Procedures. The double-reciprocal plot of the 7-EFC concentrations versus the velocities was indicative of competitive inhibition (Fig. 2). The kinetic constants were determined from a replot of the K m app versus the inhibitor concentrations (Fig. 2, inset). The concentration of BITC that doubles the slope of the double reciprocal plot or the K i was approximately 3 ⫾ 0.9 ␮M (inset to Fig. 2). The K m for 7-EFC was 24 ⫾ 2.3 ␮M and was determined from the yintercept of the plot of the K m app versus the BITC concentration. BITC also competitively inhibited the P450 2E1 T303A p-NP hydroxylation with a K i of approximately 6 ␮M (data not shown).

102

MORENO ET AL.

FIG. 1. Concentration-dependent loss of P450 2E1 7-EFC O-deethylation activity following incubation with BITC and NADPH. Incubation conditions were as described under Experimental Procedures. The concentrations of BITC were (}) 0, (䊐) 5, (Œ) 10, (E) 20, (■) 30, and (‚) 50. The data shown represent the mean and SE from three to five different experiments.

Inactivation of P450 2E1 T303A by PEITC. PEITC inactivated P450 2E1 T303A in a time- and concentration-dependent manner (Fig. 3). The inactivation followed pseudo-first-order kinetics. The kinetic constants were calculated from a double-reciprocal plot of the inverse of the initial rates of inactivation as a

FIG. 3. Time- and concentration-dependent loss of P450 2E1 T303A 7-EFC O-deethylation activity due to incubation with PEITC. Incubation conditions were as described under Experimental Procedures. The concentrations of PEITC were (}) 0, (䊐) 0.5, (Œ) 1.25, (E) 2.5, (■) 5, and (‚) 10 ␮M. (Inset) Double-reciprocal plot of the rates of inactivation as a function of the PEITC concentrations. The data shown represent the mean of three to five separate determinations.

FIG. 2. Competitive inhibition of P450 2E1 7-EFC O-deethylation activity by BITC. Incubation conditions were as described under Experimental Procedures. (A) Double-reciprocal plots of the 7-EFC concentrations versus the velocity. The concentrations of 7-EFC were 15, 25, 50, 100, and 125 ␮M and the BITC concentrations were (F) 0, (䊐) 5, (}) 10, (‚) 15, and (■) 20 ␮M. (B) Secondary plot of K m app as a function of the BITC concentrations. Data shown represent the average of two determinations. The values for K m and K i reported are the average and SE of four separate experiments.

EFFECTS OF ISOTHIOCYANATES ON P450s 2E1 AND 2E1 T303A

FIG. 4. Spectral binding studies for the binding of BITC to P450 2E1 T303A. Spectral titrations were performed as described under Experimental Procedures. Lineweaver–Burk plot of the inverse of the BITC concentration (0.5– 6.4 ␮M) versus the inverse of the change in absorption (380 – 420 nm). A representative experiment is shown. The binding constant (K s) was calculated from the x-intercept, the final K s value reported is the mean of four separate determinations.

function of the reciprocal of the PEITC concentrations (inset to Fig. 3). The concentration of inactivator required for the half-maximal rate of inactivation (K I) was 1.6 ␮M. The maximal rate of inactivation at saturation (k inact) was 0.11 min ⫺1, and the time required to inactivate one-half of the enzyme molecules (t 1/ 2 ) was 6.3 min. PEITC inactivated the wild-type enzyme with similar kinetics (data not shown). The K I was 2.7 ␮M, the k inact was 0.14 min ⫺1, and the t 1/ 2 was 5 min. Binding constant. Binding of many substrates to P450 enzymes is accompanied by changes in the absorbance spectrum of the protein (26). Upon the addition

FIG. 5.

103

of BITC to P450 2E1 a Type I spectral change was obtained and a binding constant of 4.5 ␮M was determined (14). The binding of BITC to P450 2E1 T303A also produced spectral changes characteristic of a Type I compound (data not shown). Changes in the absorbance spectrum at 380 and 420 nm were determined following the addition of increasing concentrations of BITC. The resulting changes in absorption were plotted as the inverse of the change in absorption versus the inverse of the BITC concentration to obtain the binding constant, K s, from the intercept of the linear regression line with the x-axis (Fig. 4). The concentration of BITC resulting in 50% of the theoretical maximal spectral change was 3.2 ␮M. HPLC analysis of BITC metabolites produced by P450 2E1 and P450 2E1 T303A. BITC was metabolized by P450 2E1 or P450 2E1 T303A in the reconstituted system to give benzylamine (BAM), benzoic acid (BZA), and benzaldehyde (BAH), as well as some DBU and DBT. The identity and quantity of the metabolites were determined using the elution times for known standards and standard curves (Fig. 5), or alternatively, the quantities were determined from the specific activities in experiments where [ 14C]BITC was used. Figure 6 shows two typical elution profiles for P450 2E1 and P450 2E1 T303A. Metabolites that were consistently observed were BAM, BZA, and BAH eluting at 8.2, 19.7, and 23.0 min, respectively (Fig. 6). BAM was produced at similar levels by both enzymes while BZA and BAH were produced in higher amounts by the T303A mutant enzyme as compared to the wild-type

HPLC elution pattern for BITC and authentic metabolite standards.

104

MORENO ET AL. TABLE II 14

Binding of [ C]BITC to P450 2E1 and P450 2E1 T303A a Enzyme

⫺NADPH

⫹NADPH

P450 2E1 P450 2E1 T303A

0.1 ⫾ 0.03 0.1 ⫾ 0.01

0.5 ⫾ 0.3 0.4 ⫾ 0.2

a Assay conditions were as described under Experimental Procedures. The data represents the average of densitometric values and standard deviations in arbitrary units from two separate experiments.

FIG. 6. Representative HPLC elution profile for metabolites produced from P450 2E1 (}) or P450 2E1 T303A (E) incubated with [ 14C]BITC in the presence of NADPH. Incubation conditions, extraction of metabolites, and HPLC conditions were as described under Experimental Procedures.

P450 2E1 (Table I). DBU and DBT, the urea and thiourea products, were not detected consistently. In experiments where they were observed, they were produced at similar levels by both enzymes (data not shown). In some experiments, 2-AF was included in the reaction mixtures to trap the isocyanate, BIC, resulting in the formation of BFU and BFT. Labeling of P450 2E1 and P450 2E1 T303A by [ 14C]BITC. Following incubation of both enzymes in the reconstituted system with [ 14C]BITC or [ 14C]BITC together with NADPH, samples were dialyzed to remove free and nonspecifically bound BITC. The majority of the counts were removed in the first two changes

of the cholate-containing dialysis buffer. The components in the reconstituted system were then separated by SDS–PAGE and the proteins were transferred to a PVDF membrane. Densitometric quantification of the Coomassie blue-stained PVDF membrane was used to normalize for the total amount of P450 per lane. In two separate experiments some background labeling was observed with both enzymes in the absence of NADPH. After incubating the samples for 20 min, the NADPHdependent 14C labeling increased fivefold for the wildtype enzyme and fourfold for the P450 T303A mutant (Table II). This observation demonstrates that both enzymes metabolize BITC to a reactive intermediate(s) that covalently binds to the apoprotein and that there are no significant differences in the labeling of the two proteins by BITC even though there are marked differences in the inactivation. ESI-LC-MS analysis of P450 2E1 and 2E1 T303A. LC-MS analysis of the BITC-inactivated P450 2E1 sample showed that the mass of the inactivated P450 (53,959 ⫾ 3 Da) had increased by an average of 155 mass units when compared to the control sample incubated with BITC in the absence of NADPH (53,804 ⫾ 1 Da) (Fig. 7). This mass difference is consistent within error of the LC-MS measurement with the mass of a BIC-derived protein adduct ⫹ 1 oxygen (134 ⫹ 16 Da).

TABLE I

Metabolism of BITC by P450 2E1 and P450 2E1 T303A a nmol product/nmol P450/min Metabolite

P450 2E1

P450 2E1 T303A

Product ratio for P450 2E1 T303A/2E1

Benzylamine (BAM) Benzoic acid (BZA) Benzaldehyde (BAH) Benzoic acid (BZA) b Benzaldehyde (BAH) b N-Benzyl-N⬘-(2-fluorenyl)urea (BFU) b

0.08 ⫾ 0.03 0.04 ⫾ 0.01 0.05 ⫾ 0.01 0.07 ⫾ 0.03 0.02 ⫾ 0.01 0.05 ⫾ 0.01

0.08 ⫾ 0.03 0.18 ⫾ 0.02 0.20 ⫾ 0.03 0.12 ⫾ 0.05 0.13 ⫾ 0.04 0.03 ⫾ 0.01

1 4.5 4 1.7 6.5 0.6

a Assay conditions were as described under Experimental Procedures. The data shown represent the mean and SE from two to four separate experiments. b Assayed in the presence of 200 ␮M BITC and 200 ␮M 2-AF. Under these conditions BAM eluted as a poorly defined peak early in the gradient and was not quantified. Results represent the mean and SE of two to four separate experiments.

FIG. 7. ESI-LC-MS analysis of BITC-inactivated P450 2E1. Incubations and separations were as described under Experimental Procedures. Deconvoluted spectrum of the BITC-inactivated sample showing the nonadducted apo-P450 2E1 (M r ⫽ 53,805 ⫾ 2 Da, n ⫽ 6), and the BITC-adducted apo-P450 2E1 (M r ⫽ 53,959 ⫾ 3 Da, n ⫽ 6). The inset shows the deconvoluted spectrum of the P450 2E1 protein envelope from the control sample incubated with BITC without NADPH (M r ⫽ 53,804 ⫾ 1 Da, n ⫽ 4).

EFFECTS OF ISOTHIOCYANATES ON P450s 2E1 AND 2E1 T303A

105

106

MORENO ET AL.

SCHEME 1.

Products formed from BITC during metabolism by cytochrome P450 2E1.

The BITC metabolism outlined in Scheme 1 indicates two possible P450-dependent routes: one leading to the formation of benzaldehyde and benzoic acid and the other leading to the formation of benzylamine and a BIC reactive intermediate capable of binding to the apoprotein. With P450 2E1 an initial P450-dependent conversion of BITC (149 Da) to BIC (134 Da) with an additional hydroxylation of the ␣ carbon to ␣-hydroxy BIC could result in a reactive intermediate with a mass of 150 Da. Alternatively, an adduct resulting in the same mass increase would also be observed if BITC is first hydroxylated by P450 2E1 at the ␣ carbon to ␣-hydroxy BITC (165 Da) as shown in Scheme 1. The subsequent loss of sulfur to form the isocyanate could then again lead to a reactive ␣-hydroxy BIC intermediate with a mass of 150 Da. The LC-MS result also supports the 1:1 stoichiometry of binding of a radiolabeled BITC to P450 2E1 reported previously (14). LC-MS analysis of P450 2E1 T303A incubated with BITC in the presence or absence of NADPH showed a change in the mass of the mutant 2E1 of approximately 169 Da from 53,754 ⫾ 1 to 53,923 ⫾ 5 (Fig. 8). This mass difference could be accounted for by the addition of a BIC adduct (134 Da) together with a sulfur (32 Da)

adduct. Alternatively, an adduct consisting of an entire hydroxylated BITC molecule could result in the addition of 166 Da to the P450 2E1 apoprotein. DISCUSSION

P450 2E1 and P450 2E1 T303A exhibit similar catalytic activities with substrates such as 7-EFC, 7-EC, and p-NP and they also show similar kinetics for inactivation by tert-butyl isothiocyanate (28). However, Fukuda et al. have previously reported that a P450 2E1 mutant where the conserved threonine 303 was replaced by a serine residue exhibited a higher activity for aniline p-hydroxylation and nitrosodimethylamine N-demethylation when compared with the wild-type enzyme (29). Roberts et al. have reported a similar phenomenon where the hydroxylation of p-nitrophenol occurred at an enhanced rate in the P450 2E1 T303A mutant when compared to the wild-type P450 2E1 enzyme (30). We have also observed enhanced catalytic efficiency of the mutant P450 2E1 T303A with respect to the wild-type enzyme for the metabolism of 7-EFC to 7-HFC (unpublished results). These observations are in contrast to previous reports of bacterial P450 en-

FIG. 8. ESI-LC-MS analysis of P450 2E1 T303A incubated with BITC. Incubations and separations were as described under Experimental Procedures. Deconvoluted spectrum of a sample incubated in the presence of BITC and NADPH showing the BITC-adducted apo-P450 2E1 T303A (M r ⫽ 53,923 ⫾ 5 Da, n ⫽ 5) as well as the nonadducted protein. The inset shows the deconvoluted spectrum of the P450 2E1 T303A protein envelope from the control sample incubated with BITC without NADPH (M r ⫽ 53,754 ⫾ 1 Da, n ⫽ 5).

EFFECTS OF ISOTHIOCYANATES ON P450s 2E1 AND 2E1 T303A

107

108

MORENO ET AL.

zymes such as P450 cam where a mutation at the conserved threonine 252 resulted in the uncoupling of substrate hydroxylation from electron transport (16). In P450 2E1, either the threonine 303 residue may not play the same critical role in catalysis that has been previously postulated for P450 cam or, alternatively, a neighboring residue successfully compensates for the lack of a threonine at position 303. P450 EryF represents one example of a naturally occurring P450 where an alanine residue is found in place of this conserved threonine (31). In P450 EryF, a water molecule in place of the threonine residue is believed to allow this enzyme to maintain the hydrogen-bonding network which is common to P450 cam, P450 BM-3, and P450 terp. It has been proposed that this water molecule may act as the proton donor necessary for the scission of the oxygen molecule allowing the enzyme to be catalytically active (31). The findings reported here suggest that the threonine 303 residue in P450 2E1 is involved in the inactivation of this enzyme by BITC. While P450 2E1 is inactivated by BITC in a mechanism-based manner, the mutant 2E1 T303A is only competitively inhibited. This competitive inhibition of P450 2E1 T303A by BITC is in contrast to the mechanism-based inactivation observed when PEITC was used. PEITC is structurally similar to BITC except that it contains an additional CH 2 group. The K I values observed for the inactivation of P450 2E1 and P450 2E1 T303A by PEITC were similar suggesting that these two enzymes interact with PEITC in the same manner. This observation with PEITC prompted us to study in more detail the factors that may be involved in the differential interaction of BITC with the wild-type and the mutant P450 2E1. Several possibilities may account for the different effects of BITC on the activity of the mutant and wild-type enzyme. BITC binding to the mutant enzyme could either be different or impaired. Alternatively, the T303A mutation may have resulted in an altered orientation of BITC in the active site such that the metabolism of BITC to the reactive BIC intermediate no longer occurred. The possibility that BITC bound to P450 2E1 T303A differently than to P450 2E1 was addressed by analyzing the reversible binding of BITC to the two enzymes using spectral binding. Spectral binding studies suggested that the ability of the T303A mutant P450 2E1 to bind BITC remained virtually unchanged when compared to the wild-type P450 2E1 enzyme. BITC produced spectral changes characteristic of a Type I compound with both P450s exhibiting similar binding constants. However, P450 2E1 and 2E1 T303A behaved differently with respect to the metabolism of BITC. P450 2E1 T303A exhibited an enhanced ability to oxidize BITC at the ␣ carbon resulting in significantly higher

levels of benzaldehyde and benzoic acid compared to the wild-type enzyme. Roberts et al. reported similar differential effects for the metabolism of 5-phenyl-1pentyne by P450 2E1 and 2E1 T303A (30). The oxidation of the internal carbon of 5-phenyl-1-pentyne resulted in the inactivation of P450 2E1 predominantly through a modification of the heme moiety. In contrast, P450 2E1 T303A was inactivated at a slower rate and this was a result of an enhanced oxidation of the external carbon of 5-phenyl-1-pentyne thereby resulting in a higher production of 5-phenylvaleric acid (30). Therefore, the slower rate of inactivation of the 2E1 mutant lacking the threonine residue was explained by a shift in metabolism from the internal to the terminal acetylenic carbon. While a shift in the identity of the carbon atom targeted for hydroxylation by the T303A mutant was also observed in our system, that shift did not lead to an inactivation of the enzyme but rather a competitive inhibition was observed. The extraction procedure and HPLC analysis of BITC metabolites used in this study was a modification of Goosen et al. (31) based on a previously published method for the separation and quantitation of the metabolites of 2-naphthyl isothiocyanate (2-NITC) developed by Lee (27). The metabolism of 2-NITC by rat liver microsomes was P450-dependent and resulted in the formation of the more reactive, isocyanate 2-naphthyl isocyanate (2-NIC), with the subsequent hydrolysis to 2-aminonaphthalene (2-AN). The formation of 2-NIC was confirmed by using 2-AF as a trapping agent to form the conjugated urea N⬘,N-dinaphthylurea (DNU). The identities of both 2-NIC and 2-AN were also confirmed by mass spectral analysis (27). The metabolism of BITC in our system followed a similar pathway as the one described for 2-NITC. Scheme 1 shows the two main routes for the metabolism of BITC by P450. Oxidation at the ␣ carbon leads to the formation of BAH and BZA without resulting in inactivation of P450 2E1. Similar studies by Goosen et al. (31) have also indicated that BITC is metabolized by P450 2B1 to the metabolites shown in Scheme 1. This was demonstrated by HPLC using authentic standards and by GC-MS analysis. As with P450 2E1, the inactivation of P450 2B1 by BITC was due primarily to modification of the apoprotein with little heme loss. Oxidation at the isothiocyanate carbon could lead to the formation of the proposed reactive isothiocyanate species resulting in protein modification. Interestingly, both enzymes produced the proposed reactive BIC intermediate as monitored by the generation of BAM and the urea derivative BFU. This urea derivative was also detected in earlier studies where BITC and 2-AF were incubated together with liver microsomes of rats treated with Aroclor 1254 (33, 34). In addition, a radiolabeled metabolite of BITC was bound to both P450 2E1 and P450 2E1 T303A.

EFFECTS OF ISOTHIOCYANATES ON P450s 2E1 AND 2E1 T303A

ESI-LC-MS analysis of the BITC-inactivated P450 2E1 showed an increase in molecular mass of P450 2E1 of approximately 155 Da. This increase would be consistent with the size of a BIC adduct containing 1 oxygen (Scheme 1). Alternatively, this size adduct could have also been generated by first hydroxylating the ␣-carbon of BITC as is expected to generate benzaldehyde. The hydroxylated intermediate could then be converted to the isocyanate reactive intermediate, leading to protein adduction. Incubation of the P450 2E1 T303 mutant with BITC and NADPH also resulted in the formation of an adducted protein. These observations confirmed the labeling studies described above. The increase in mass of approximately 169 Da would be consistent either with a modification by the postulated BIC reactive intermediate plus a sulfur molecule or by a hydroxylated BITC intermediate. The latter possibility seems to be the more likely. Preliminary experiments indicate that P450 T303A incubated with BITC and NADPH followed by exposure to 5 mM DTT for 2 h on ice brought about the complete release of the adduct. As mentioned previously, even though the ability of P450s to catalyze certain reactions is conserved, specific differences in the rate of the reactions catalyzed by the wild-type and mutant enzymes have been reported. Evidence for metabolic shifts induced by single amino acid mutations have been reported previously by Vaz et al., where mutation of the conserved threonine 302 to alanine in P450 2B4 suppressed both epoxidation and allylic hydroxylation reactions and resulted in an enhancement in oxidative deformylation reactions supporting the role of a nucleophilic species, the peroxoiron, as the oxidant (35). Similarly, P450 2E1 T303A showed an enhanced ability to catalyze epoxidation and oxidative deformylation reactions. Vaz et al. (35) proposed that the P450 2E1 T303A mutant forms predominantly the hydroperoxo-iron species over the oxenoid species. These studies provide supporting evidence for the hypothesis that more than one oxidant species may be involved in P450 catalysis (35). The differential interaction of BITC by these two enzymes may occur as a result of a modified active site conformation due to the threonine to alanine mutation. While the mutant enzyme remained catalytically intact and able to metabolize the same substrates (p-NP, 7-EC, 7-EFC) as the wild-type enzyme, it is evident that specific changes have taken place that alter the way the mutant enzyme interacts with at least some substrates (5-phenyl-1-pentyne and BITC). Although both enzymes appear to generate the reactive species responsible for inactivation, only the wild-type P450 2E1 containing the T303 was inactivated. This could indicate that the reactive BITC intermediate binds covalently to T303 resulting in a loss in activity. In the T303A mutant, the catalytic role of T303 may be performed by a neighboring residue that

109

does not interact with the reactive BITC intermediate or by a water molecule in the active site. Studies are under way to determine if the BITC-labeled amino acid residue(s) in P450 2E1 and P450 2E1 T303A are the same or different and if T303 is labeled in the wild-type enzyme. ACKNOWLEDGMENTS We thank Dr. Mei-Sie Lee for providing us with the standards for the BITC metabolic studies and Dr. Minor J. Coon for the plasmids for P450 2E1 and P450 2E1 T303A. We are grateful to Chitra Shridar for her help with the purification of P450 2E1, P450 2E1 T303A, and reductase. This publication was supported in part by NIH Grants CA 16954 (P.F.H.) and CA 46535 (F-L.C.) from the National Cancer Institute.

REFERENCES 1. Zhang, Y., and Talalay, P. (1994) Cancer Res. 54(7 Suppl.), 1976 –1981. 2. Yang, C. S., Smith, T. J., and Hong, J. Y. (1994) Cancer Res. 54(7 Suppl.), 1982–1986. 3. Zhang, Y., Talalay, P., Cho, C. G., and Posner, G. H. (1992) Proc. Natl. Acad. Sci. USA 89(6), 2399 –2403. 4. Conaway, C. C., Jiao, D., and Chung, F. L. (1996) Carcinogenesis 17(11), 2423–2427. 5. Wattenberg, L. W. (1987) Carcinogenesis 8(12), 1971–1973. 6. Wattenberg, L. W. (1977) J. Natl. Cancer Inst. 58(2), 395–398. 7. Morse, M. A., Amin, S. G., Hecht, S. S., and Chung, F. L. (1989) Cancer Res. 49(11), 2894 –2897. 8. Morse, M. A., Wang, C. X., Stoner, G. D., Mandal, S., Conran, P. B., Amin, S. G., Hecht, S. S., and Chung, F. L. (1989) Cancer Res. 49(3), 549 –553. 9. Morse, M. A., Eklind, K. I., Hecht, S. S., Jordan, K. G., Choi, C. I., Desai, D. H., Amin, S. G., and Chung, F. L. (1991) Cancer Res. 51(7), 1846 –1850. 10. Smith, T. J., Guo, Z., Guengerich, F. P., and Yang, C. S. (1996) Carcinogenesis 17(4), 809 – 813. 11. Smith, T. J., Guo, Z., Gonzalez, F. J., Guengerich, F. P., Stoner, G. D., and Yang, C. S. (1992) Cancer Res. 52(7), 1757–1763. 12. Yamazaki, H., Inui, Y., Yun, C. H., Guengerich, F. P., and Shimada, T. (1992) Carcinogenesis 13(10), 1789 –1794. 13. Guengerich, F. P., Kim, D. H., and Iwasaki, M. (1991) Chem. Res. Toxicol. 4(2), 168 –179. 14. Moreno, R. L., Kent, U. M., Hodge, K., and Hollenberg, P. F. (1999) Chem. Res. Toxicol. 12(7), 582–587. 15. Raag, R., Martinis, S. A., Sligar, S. G., and Poulos, T. L. (1991) Biochemistry 30(48), 11420 –11429. 16. Imai, M., Shimada, H., Watanabe, Y., Matsushima-Hibiya, Y., Makino, R., Koga, H., Horiuchi, T., and Ishimura, Y. (1989) Proc. Natl. Acad. Sci. USA 86(20), 7823–7827. 17. Tan, Y., White, S. P., Paranawithana, S. R., and Yang, C. S. (1997) Xenobiotica 27(3), 287–299. 18. Roberts, E. S., Hopkins, N. E., Alworth, W. L., and Hollenberg, P. F. (1993) Chem. Res. Toxicol. 6(4), 470 – 479. 19. Roberts, E. S., Pernecky, S. J., Alworth, W. L., and Hollenberg, P. F. (1996) Arch. Biochem. Biophys. 331(2), 170 –176. 20. Roberts, E. S., Hopkins, N. E., Zaluzec, E. J., Gage, D. A., Alworth, W. L., and Hollenberg, P. F. (1994) Biochemistry 33(12), 3766 –3771. 21. Eklind, K. I., Morse, M. A., and Chung, F. L. (1990) Carcinogenesis 11(11), 2033–2036.

110

MORENO ET AL.

22. Larson, J. R., Coon, M. J., and Porter, T. D. (1991) J. Biol. Chem. 266(12), 7321–7324. 23. Shen, A. L., Porter, T. D., Wilson, T. E., and Kasper, C. B. (1989) J. Biol. Chem. 264(13), 7584 –7589. 24. Hanna, I. H., Teiber, J. F., Kokones, K. L., and Hollenberg, P. F. (1998) Arch. Biochem. Biophys. 350(2), 324 –332. 25. Buters, J. T., Schiller, C. D., and Chou, R. C. (1993) Biochem. Pharmacol. 46(9), 1577–1584. 26. Estabrook, R. W., and Werringloer, J. (1978) Methods Enzymol. 52, 212–220. 27. Lee, M. S. (1992) Chem. Res. Toxicol. 5(6), 791–796. 28. Kent, U. M., Roberts, E. S., Chun, J., Hodge, K., Juncaj, J., and Hollenberg, P. F. (1998) Chem. Res. Toxicol. 11(10), 1154 –1161.

29. Fukuda, T., Imai, Y., Komori, M., Nakamura, M., Kusunose, E., Satouchi, K., and Kusunose, M. (1993) J. Biochem. (Tokyo) 113(1), 7–12. 30. Roberts, E. S., Alworth, W. L., and Hollenberg, P. F. (1998) Arch. Biochem. Biophys. 354(2), 295–302. 31. Goosen, T. C., Mills, D. E., and Hollenberg, P. F. (2000) J. Pharmacol. Exp. Ther. 296(1), 198 –206. 32. Cupp-Vickery, J. R., and Poulos, T. L. (1995) Nature Struct. Biol. 2(2), 144 –153. 33. Lee, M. S. (1996) Chem. Res. Toxicol. 9(7), 1072–1078. 34. Lee, M. S. (1994) Environ. Health Perspect. 102(Suppl. 6), 115–118. 35. Vaz, A. D., McGinnity, D. F., and Coon, M. J. (1998) Proc. Natl. Acad. Sci. USA 95(7), 3555–3560.