A continuous fluorometric assay for cytochrome P-450-dependent mixed function oxidases using 3-cyano-7-ethoxycoumarin

ANALYTICAL

BIOCHEMISTRY

172.304-3

10 ( 1988)

A Continuous Fluorometric Assay for Cytochrome P-450-Dependent Function Oxidases Using 3-Cyano-7-ethoxycoumarin IAN

N. H.

Mixed

WHITE

MRC Toxicology Unit, Woodmansterne Road, Curshalton, Surrey Shf5 4EF, United Kingdom Received

November

12. 1987

A direct fluorometric procedure for the continuous determination of cytochrome P-450-dependent mixed function oxidases. using 3-cyano-7-ethoxycoumarin substrate. is described. The reaction product, 3-cyano-7-hydroxycoumarin, is fluorescent at neutral pH values (excitation and emission wavelength maxima: 408 and 450 nm. respectively). Using hepatic microsomal preparations from control rats, the enzyme(s) had an apparent K”, of 16 PM. V,,,,, values (0.5 nmol/min/mg protein) were induced 6- and 2 1-fold by pretreatment of rats with phenobarbitone and 3-methylcholanthrene, respectively. Using microsomes from control rats, this procedure is about 50- to 1OO-fold more sensitive than the ethoxyresorufin deethylase assay. Reaction rates using 3-cyano-7-pentoxycoumarin as substrate were generally much lower than with the ethoxy analog. 3-Cyano-7-ethoxycoumarin can also be used as a substrate to measure mixed function oxidases in isolated hepatocytes. However, 3-cyano-7-hydroxycoumarin shows a timeand concentration-dependent loss of fluorescence when incubated with such cells. This causes an approximately 5% underestimate of the true reaction rates. (~8198x Acadcmac PWSS. IK KEY WORDS: fluorometry: cytochrome P-450; 3-cyano-7-alkyoxycoumarins: mixed function oxidases.

Two substratesare commonly used for the fluorometric determination of cytochrome P-450-dependent mixed function oxidase activities. These are 7-ethoxycoumarin (1,2) and 7-alkoxyphenoxazones (alkoxyresorufins (3.4)). 7-Ethoxycoumarin is O-deethylated to yield 7-hydroxycoumarin. However, at neutral pH the fluorescence intensity of this product and therefore the sensitivity of the assay is low. Maximum fluorescence is reached only at pH > 9.5, not conducive to studies in living cells. 7-Ethoxycoumarin is not normally used for the continuous analysisof mixed function oxidase activities except for specialized applications (5). In contrast, resorufin, the metabolite from the O-dealkylation of 7-alkoxyresorufins, is highly fluorescent at neutral pH values and hasbeen widely employed for the continuous kinetic determination of mixed function oxidase activities in microsomal and purified cytochrome P-450 preparations (3,4). However, in the presence of NAD(P)H, cytosolic DT’-diaphorases can 0003-2697/M

$3.00

Copyright 0 1988 hy Academic Press. Inc. All rights of reproductmn ,n any form resewed.

304

bring about the reduction of the 7-hydroxy group of resorufin to a nonfluorescent product (6). This property limits the use of these substratesin S9 systems. In this paper, the possibility of using 3-cyano-7-alkoxycoumatins such as 3-cyano7-ethoxycoumarin (Fig. 1, I) as substrates for the cytochrome P-450-dependent mixed function oxidases is investigated. The electron-withdrawing 3-cyano function of the expected metabolite, 3-cyano-7-hydroxycoumarin (Fig. 1, II), confers fluorescence at neutral pH values (7) and enhances fluorescence intensity by a 2.9-fold factor over that of 7hydroxycoumarin (8). MATERIALS

AND

METHODS

Materials 3-Cyano-7-hydroxycoumarin was synthesized by the method described by Wollbeis ’ Abbreviation cleotide specitic.

used: DT, di or triphosphopyridine

nu-

FLLJOROMETRIC

DETERMINATION

FIG. 1, Chemical structures of 3-cyano-7-ethoxycoumarin (I) and 3-cyano-7-hydroxycoumarin (II).

(7) or was purchased, along with ethoxyresorufin, from Molecular Probes (4849 Pitchford Ave., Eugene, OR 97402). Ethyl iodide, pentyl iodide, silver(II), oxide 7-hydroxyphenoxazone, and quinoline came from Aldrich Chemical Co. (Gillingham, UK). All other chemicals were from BDH Ltd. (Poole, UK). Synthesis of 3-Cyano-7-ethoxycoumarin A procedure similar to that described by Miller (9) was used. To 2 mmol (374 mg) of 3-cyano-7-hydroxycoumarin in toluene (50 ml) was added 4 rnmol (0.93 g) of silver(I1) oxide, three drops of quinoline, and 4 mmol (0.62 g) of e’ihyl iodide. The mixture was stirred in the dark for 24 h at room temperature. Ethyl acetate (50 ml) was added and the mixture filtered into a separating funnel. After the mixture was washed with 0.5 M NaOH (2 X 20 ml) and water (2 X 20 ml), the organic phase was dried (anhydrous Na,SO,) and rotary evaporated to dryness. The crude product (348 mg, 8 1% yield) was recrystallized from hot ethyl acetate, mp 2 12°C. ‘H NMR (60 mHz, Perkin-Elmer RI2 B instrument, tetramethylsilane internal standard) 6 (ppm) ‘H chloroform: 8.7 (I H. m, H4), 8.1-6.9 (3H, aromatic protons). 4.2 (3H, q-CHJ, I .4 (3H. i. CHJ). Electron impact mass spectrometry (VG 70 SEQ instrument) gave a molecular ion of nz/: 2 15 (6 1.O% of base peak) and fragment ions of FYI/Z 187 (base peak, M+-C2H4) and 159 (R/I’-CZH,-CO). High resolution mass spectrometry gave a molecular ion of m/z 2 15.0696. Calcd for CZ7He,03N: 215.0582. 3-Cyano-7-pentoxycoumarin (mp 130°C) was prepared by analogous procedures using pentyl iodide. ‘H NMR 6(ppm) ‘H chloroform: 8.3 ( I H. m. H4). 7.7-6.3 (3H, m, aromatic protons). 4. I (2H, ml-CIZz). 1.6

OF OXIDASES

305

(6H, m, C3H6), 1.0 (3H, t, CH3-). Mass spectrometry (electron impact) gave a molecular ion of m/z 257 (19.3% of base peak) and fragment ions of m/z 257 (19.3% of base peak) and fragment ions of m/z 187 (M+-C5Hs, base peak) and 159 (M+-CSHs--CO). High resolution mass spectrometry gave a molecular ion of nz/~ 257.1088. Calcd for CIH,&N: 257.1051. These data and ‘H NMR were consistent with the proposed structure. TLC on silica gel developed with ethyl acetate toluene (3: 1) showed a single component with blue fluorescence under uv light. The 7-hydroxy, 7-ethoxy, and 7-pentoxy derivatives of 3-cyanocoumarin had R, values of0.36,0.52, and 0.58, respectively, in this solvent system. Preparation of Hepatocytes and Liver Microsomal Fractions Hepatocytes were prepared by collagenase perfusion of male Fischer F-344/N rats ( 150160 g) according to the method of Paine and Legg (10). Cells of viability >80% as assessed by trypan blue exclusion with the aid of a hemocytometer were used. Liver microsomal fractions were prepared as described previously ( 11). Microsomal protein concentrations were estimated using the Lowry procedure with bovine serum albumin as standard (12). In certain experiments rats were either pretreated with phenobarbitone (0.1%) in the drinking water for 7 days or were given 3methylcholanthrene (20 mg/kg, ip) once a day for 3 days.

7-ethc,s~wollnzarin 0-deethJlla.se (a) Liver microsomal assays. Incubation mixtures of 3-ml volume in 0.2 M phosphate buffer. pH 7.5. contained NADPH (0.5 mM), MgCl? (10 mM). and microsomal suspension (25 gg protein). Reactions were started wirh 3-cyano-7-ethoxycoumarin (0.03 mM) dissolved in dimethyl sulfoxide (10 ~1). Mixtures were stirred in a fluorometer cuvette thermostatted a.t 37°C. Increasing fluorescence with

306

IAN N. H. WHITE

time was recorded using a Perkin-Elmer LS5 fluorometer, excitation wavelength 408 nm, emission 455 nm. Excitation and emission bandpass slits of 10 nm were used. In some instances, the amount of protein or the substrate concentrations were changed. Estimation of K, and I’,,, values was computed using the procedures described by Wilkinson (13). Rates of ethoxyresorufin O-deethylase were determined under the reaction conditions described above. Fluorometer excitation and emission wavelengths were set to 530 and 586 nm, respectively, and an ethoxyresorufin concentration of 10 PM was used (3,4). Assays for DT-diaphorases were carried out in the presence of NAD(P)H (250 pm) and rat liver cytosol as described by Nims et a/. (6). In some instances resorufin (10 PM) was replaced by 3-cyano-7-hydroxycoumarin ( lo- 100 pm) using the appropriate excitation and emission wavelengths. (6) Hepatocvtes. Reactions were carried out in phosphate-buffered saline (3 ml) containing freshly isolated hepatocytes (5 X 1OS) and 3-cyano-7-ethoxycoumarin (0.03 mM). Conditions otherwise were the same as those described for the microsomal assay.

Kratos FS 970 fluorometer (excitation wavelength 408 nm, high pass emission filter 418 nm) and a Pye-Unicam DP88 integrator. Under these conditions 7-hydroxy and 7-ethoxy derivatives of 3-cyanocoumarin had retention times of 5.2 and 11.5 min, respectively: 7-hydroxycoumarin, with a retention time of 7.8 min, was used as the internal standard. RESULTS AND DISCUSSION

Fluorescence Characteristics of 3-cyano7-H.ydroxy- and 3-cyano7-Ethoxycownarins The fluorescence spectrum of 3-cyano-7hydroxycoumarin in phosphate buffer, pH 7.5, shows excitation and emission maxima at 408 and 450 nm, respectively (Fig. 2). 3Cyano-7-ethoxycoumarin has a lower fluorescence intensity on a molar basis and excitation and emission maxima are shifted to shorter wavelengths (355 and 4 10 nm. Fig. 2). The pH dependence of fluorescence intensity of 3-cyano-7-hydroxycoumarin is shown in Fig. 3. Fluorescence intensity is maintained over the pH range 8.0 to about 7.2 after which there is a pH-dependent decrease. Consider-

HPLC of 3-Cyano- 7-ethoxyconmarin and Its Metabolite Reactions, from microsomal incubation mixtures as described above, were stopped by adding an aliquot ( 100 ~1) to ice-cold CH3CN (200 ~1) in Eppendorf centrifuge tubes. Following mixing and centrifugation, the supernatant was used directly for HPLC. This was carried out using a reverse phase 5 PM Cl8 column (Lichrosorb Cl8 12.4 X 0.4 cm, E. Merck, Darmstadt, Germany). Starting conditions of solvent were methanol/O.01 M phosphate buffer, pH 7.4 (1:5), at a flow rate of 1.2 ml/min. A linear gradient (G6 Waters 660 programmer) was to 100% methanol over 15 min. Quantitation of metabolites based on fluorescence was achieved using a

FIG. 2. Excitation and emission spectra of 7-hydroxy and 7-ethoxy derivatives of 3-cyanocoumarin. Spectra were recorded using 100 nM solutions in 0.1 M phosphate buffer, pH 7.5. -. 3-Cyano-7-hydroxycoumarin; ---, 3cyano-7-ethoxycoumarin.

FLUOROMETRIC

DETERMINATION

FIG. 3. Dependence on pH of the fluorescence intensity of 3-cyano-7-hydroxycoumarin. Fluorescence was recorded using 0. I M KZHP04/KH1P04 buffers at the pH values indicated at 25°C. Excitation wavelength, 408 nm: emission. 450 nm

OF

307

OXIDASES

time-dependent increase in fluorescence intensity (Fig. 4a). Reaction time courses were Iinear for at least IO min. Similar, though much lower. rates are seen using 3-cyano7-pentoxycoumarin. Reaction rates were linearly related to the protein concentration between 10 and 150 pg/3 ml incubation mixture using 3-cyano-7-ethoxycoumarin as substrate. The lowest reaction rate which could be measured was on the order of 0.3 pmol/ min/3 ml incubation mixture. No metabolism was detected in the absence of NADPH. With ethoxyresorufin as substrate, protein concentrations approximately lOO-fold higher had to be used in order to achieve similar rates of increase of fluorescence with time. However, with ethoxyresorufin, baseline noise was lower due to more favorable longer excitation/emission wavelengths (3,4).

1 able advantage over 7-hydroxycoumarin is afforded since this compound shows maximum fluorescence intensity only at pH values ~9.5 (1) and on an equimolar basis is 2.9fold less fluorescent (8). The fluorescence intensity ratio at pH 7.4 in 0.1 M phosphate buffer between equimolar (10 nM) solutions of 3-cyano-7-hydroxycoumarin (408, 450 nm), 7-hydroxycoumarin (380, 452 nm), and resorufin (530, 586 nm) was 16.4: 1.4:1 (excitation and emission wavelengths used are given in parentheses). At pH 7.5 there was a linear relationship between fluorescence intensity and 3-cyano-7-hydroxycoumarin concentrations between 300 PM and 100 nM. Rates of 0-Dealkylation of 3-Cyano7-ethoxycoumarin bv Rat Liver Microsomes: Formation of 3Cvano- 7-hydrox~~coumarin When washed control rat liver microsomes are incubated with 3-cyano-7-ethoxycoumarin in a stirred fluorometer cuvette, there is a

FIG. 4. Time dependence of fluorescence intensity following incubation of 3-cyano-7-alkoxycoumarins with (al control rat liver microsomes and (b) hepatocytes. (a) Rat liver microsomal preparation (18 WC&protein) was incubated with NADPH (0.5 mM) in phosphate buffer. pH 7.5. in a stirred fluorometer cuvette at 37°C. Reactions were started with the addition of 3-cyano-7-ethoxycoumarin (0.03 mM, upper trace) or a similar concentration of 3-cyano-7-pentoxycoumarin (lower trace) dissoIved in dimethyl sulfoxide (10 ~1). (b) Freshly prepared rat hepatocytes (5 X 10’ cells) were incubated in phosphatebuffered saline (3 ml) and the appropriate 3-cyano-7-alkoxycoumarin substrate as described for (a).

308

IAN

N. H. WHITE

Reverse phase HPLC of protein free supernatants from microsome reaction mixtures incubated with 3-cyano-7-ethoxycoumarin for 5 min showed the formation of a single metabolite, as judged using fluorescence detection. This had a retention time of 5.2 min. the same as that of 3-cyano-7-hydroxycoumat-in. This metabolite, when purified by TLC and HPLC and subjected to electron impact MS, gave a molecular ion of m/z 187 with fragment ions of nz/z 159 (M’-CO) and 130 (M+-CO-CHO), the same as that of authentic 3-cyano-7-hydroxycoumarin. The reaction rate, as determined by HPLC using microsomal preparations from control rats, 3-cyano-7-ethoxycoumarin as substrate (0.03 IIIM), and a 5-min incubation time, was 0.46 ? 0.03 nmol/min/mg protein (mean + SE for four experiments), results which compare well with the direct fluorometric assay (cf. Table 1). In the present study. because radiolabeled 3-cyano-7-ethoxycoumarin was not available, it cannot be certain that all metabolites were extracted from the microsomal incubation mixtures. Nonfluorescent metabolites might also be formed which were not detected using the present HPLC procedures. Enzyme Kinetic Parameters Microsomal Preparations

Using Liver

Table 1 shows apparent K, and k’,,, values using ethoxy- or pentoxycyanocoumarins as

substrates for hepatic microsomal preparations from control, phenobarbitone-, or 3methylcholanthrene-pretreated rats. Using 3cyano-7-ethoxycoumarin as the substrate, phenobarbitone pretreatment causes a 6fold induction in the apparent F’,,,,, value, whereas there is a 2 1-fold increase after pretreatment with 3-methylcholanthrene. In contrast, v,,, values for the pentoxy analog are preferentially increased 2 1-fold by phenobarbitone pretreatment and only 5-fold by 3methylcholanthrene. The forms of cytochrome P-450 involved in the 0-dealkylation of these compounds appear similar to those involved in the dealkylation of the two corresponding alkoxyresorufins (3,4). No studies have been carried out using purified constitutive forms of cytochrome P450. Present results suggest, however, that the higher sensitivity of the current assay may be due in part to metabolism of 3-cyano-7-alkoxycoumarins by a broader spectrum of cytochrome P450 than is the case using alkoxyresorufins. This is reflected by the relatively higher levels of induction observed with the latter substrates. For example, using ethoxyresorufin, 3-methylcholanthrene pretreatment of rats causes a 145-fold increase in the rates of microsomal 0-deethylation while phenobarbitone pretreatment causes an 87-fold increase in the metabolism of pentoxyresorufin (15). The increases now observed with 3-cyano-7alkoxycoumarins are approximately 2 1-fold in the case of both pretreatments. Values for

TABLE

1

APPARENTK~NETIC~F~RTHEO-DEALKYLATIONOF~-C~ANO-~-ALKOXYCOUMARIN~BYRATLIVER MICROSOMES:EFFECTOFPRETREATMENTWITHMIXEDFUNCTIONOXIDASE~NDUCERS Substrate 3-Cyano-7-ethoxycoumarin

Pretreatment None (controls) Phenobarbitone 3-Methylcholanthrene

L (PM) 16.1 t 2.6 16.4 k 5.7 5.5 jI 0.8

3-Cyano-7-pentoxycoumarin I’

(nmol/min/kg

protein)

0.53 + 0.03 3.34kO.14 11.4 kO.51

Kn (PM) 20.9 It 3.0 2.67 ? 0.40 4.03 f 0.49

1 Ln (nmol/min/mg

protein)

0.037 f 0.0 1 0.80 k 0.03 0.19 kO.08

FLUOROMETRIC

DETERMINATION

toward ihe present substrates after in~rnax duction (Table 1) are generally lower than those reported for the alkoxyresorufins (3). In contrast, in microsomes from control rats, the apparant K,,, values toward the 3-cyano7-alkoxycoumarins are nearly an order of magnitude higher than those toward the alkoxyresorufins. This may be due to the relatively greater polarity of the former compounds associ,ated with the presence of the 3cyan0 function. Metabolism o,f3-Cyano- 7-alkoxycoumarins by Isolated Hepatocytes When freshly prepared hepatocytes are incubated with 3-cyano-7-ethoxycoumarin, there is a time-dependent increase in fluorescence similar to that seen in microsomal systems. However, at the start of the reaction, there was a nonlinear lag period of about 0.5 min using ethoxycyanocoumarin or 3-5 min using pentoxycyanocoumarin as substrate (Fig. 4b). A similar lag period has been reported using 7-ethoxycoumarin as substrate, attributed to the membrane permeability of the hepatocytes (2). After the initial lag period, reaction rates were linear for 10 min. Reaction rates were also linearly related to the number 0.f cells between 1 X lo4 and 1 X 106/3 ml incubation mixture. Table 2 shows relative rates of metabolism using saturating concentrations of the ethoxy- and pentoxy-3-cyanocioumarins and the effects of the mixed function oxidase inducers, phenobar-

OF

309

OXIDASES

bitone, or 3-methylcholanthrene. Results using hepatocytes were similar to those seen with microsomal systems although the extent of induction by phenobarbitone or 3-methylcholanthrene is somewhat greater in the case of the isolated cells. Metabolism of 3-Cyano-7-hydroxycoumarin by Isolated Hepatocytes In hepatocyte incubation mixtures, 3-cyano-7-hydroxycoumarin is itself slowly metabolized in a time-dependent manner to nonfluorescent products (not shown). Initial rates of loss are linearly related to substrate concentration over the range 8-100 nM (Fig. 5). Over this range, no saturation of enzymatic activities is observed. Loss of 3cyano-7-hydroxycoumarin fluorescence is not enhanced using hepatocytes from rats pretreated with phenobarbitone or 3-methylcholanthrene. Although it is difficult to extrapolate from substrates added to the incubation mixture and those formed intracellularly, the present results suggest that destruction of 3-cyano-7-hydroxycoumarin in this way would lead to an approximately 5% underestimate of mixed function oxidase activities using the 3-cyano-7-alkoxycoumarin substrates. Although the nature of the enzyme involved has not been established, it is not thought to be DT-diaphorase. Losses are neither increased using hepatocytes from 3-methylcholanthrene pretreated rats nor inhibited by the presence of dicoumarol (l-30

TABLE

2

Substrate 3-Cyano-7-ethoxycoumarin Pretreatment None (controls) Phenobarbilone 3-Methylcholanthrene

3-Cyano-7-pentoxycoumarin (pmol/min/lOh

9 54* 362 k 37 1579 k 179

cells) 1.5 k 0.6 48 ?9 IO *I

IAN N. H. WHITE

310

line of the krypton ion laser where low detections limits are required. ACKNOWLEDGMENTS I thank Susan Baillie for technical assistance. Rebecca Jukes for NMR spectra, and Dr. Peter Farmer and John Lamb for mass spectrometry.

REFERENCES I, Greenlee. W. F., and Poland, A. ( 1978) J. Phurmacol Esp. Thu. Substrate

cone”.

I nM I

FIG. 5. Loss of fluorescence of 3-cyano-7-hydroxycoumarin on incubation with rat hepatocytes: dependence on substrate concentration. Rat hepatocytes (5 x lo5 cells) were incubated at 37°C in phosphatebuffered saline (3 ml). Initial rates of fluorescence intensity loss (excitation wavelength. 408 nm; emission. 450 nm) with time were plotted using different concentrations of 3-cyano-7-hydroxycoumarin. Results represent the mean of two experiments using 0, control hepatocytes; a, those from 3-methylcholanthrene-pretreated rats; or 0, those from phenobarbitone-pretreated rats.

PM) in the incubation

mixtures ( 16) as would be expected if DT-diaphorases were involved. CONCLUSIONS

The use of 3-cyano-7-alkoxycoumarins provides a sensitive continuous fluorometric assay for mixed function oxidase assays for use both in microsomal preparations and in intact cells. The isoenzymatic forms of cytochrome P-450 involved appear broadly similar to those associated with the O-dealkylation of the alkoxyresorufins. The excitation maximum wavelength of the reaction product 3-cyano-7-hydroxycoumarin is potentially suited for use with the 407-nm emission

205.596-605.

2. Rogiers. V., Adrianenssens. L.. Vandenborghe. Y.. Gepts, E., Callaerts, A., and Vercruysse, A. ( 1986) Xmohioticu

16,8 17-826.

3. Burke, M. D., and Mayer. R. T. ( 1974) Drug Metuh. Dispos.

2, 583-588.

4. Burke, M. D., Thompson, S., Elcome. C. R.. Halpet-t. J.. Haaparanta. T.. and Mayer. R. T. (1985) Biochem.

Phurmacol.

34,3337-3345.

5. Ji. S.. LeMasters. J. L.. and Thurman, R. G. ( 198 I) Mol.

Pharmacol.

19. 5 13-5 16.

6. Nims. R. W., Prough, R. A., and Lubet, R. A. ( 1984) At&.

Biochem.

BiuphJss. 229, 459-465.

7. Woltbeis. 0. S. (1977) Z. Natztr$xxh. 32a, lO651067. 8. Sherman, W. R.. and Robins, E. (1968),lnaL Chrm 40,803-805. 9. Miller, A. G. (1983)iinal. Biochem. 133,46-57. 10. Paine. A. J.. and Legg. R. F. (1978) Bioehcwz. Biophys. Res. Commur~. 1 I,

12. 13. 14. 15. 16.

81, 672-697.

White, 1.N. H. ( 1980) C/rem. Bid. Interact. 29. 103115. Lowry. 0. H.. Rosebrough, N. J.. Farr. A. L.. and Randall, R. J. (1951) J. Bid. Chem. 193. 265275. Wilkinson, G. N. ( 196 I ) Biochem. J 80, 324-332. Lubet. R. A.. Mayer, R. T.. Cameron. W.. Nims, R. W.. Burke, M., Wolff. T.. and Guengerich. F. P. (1985) Arch. Biochem. Biophyx 238,43-48. Godden. P. M. M., Kass. G., Mayer, R. T., and Burke, M. D. ( 1987) Biochem. Pharmacol. 36, 3393-3398. Hojeberg, B., Blomberg, K., Stenberg. S.. and Lind. C. (198 1) A4rch. Biochem. BiophJx 207.205-Z 16.