Redox cycling of resorufin catalyzed by rat liver microsomal NADPH-cytochrome P450 reductase

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 268, No. 2. February 1, pp. 605-616,1989

Redox Cycling of Resorufin Catalyzed by Rat Liver Microsomal NADPH-Cytochrome P450 Reductase’

DAVID R. DUTTON. GREGORY A. REED, AND ANDREW PARKINSON’

Department of Pharmacology~ Toxicology. and Thera.peutics, lkiversity of Kansas hledical Center, Ka,nsas City, Kansas 66103

Received August 12,1988, and in revised form September 30,1988

The 0-dealkylation of 7-alkoxyresorufins to the highly fluorescent compound, resorufin (7-hydroxyphenoxazone), provides a rapid, sensitive, and convenient assay of certain forms of liver microsomal cytochrome P450. The results of this study indicate that NADPH-cytochrome P450 reductase catalyzes the reduction of resorufin (and the 7- alkoxyresorufins) to a colorless, nonfluorescent compound(s). The reduction of resorufin by NADPH-cytochrome P450 reductase was supported by NADPH but not NADH, and was not inhibited by dicumarol, which established that the reaction was not catalyzed by contaminating DT-diaphorase (NAD[P]H-quinone oxidoreductase). In addition to the rate of reduction, the exten.t of reduction of resorufin was dependent on the concentration of NADPH-cytochrome P450 reductase. The maintenance of steady-state levels of reduced resorufin required the continuous oxidation of NADPH, during which molecular O2 was consumed. When NADPH was completely consumed, the spectroscopic and fluorescent properties of resorufin were fully restored. These results indicate that the reduction of resorufin by NADPH-cytochrome P450 reductase initiates a redox cycling reaction. Stoichiometric measurements revealed a 1:l:l relationship between the amount of NADPH and O2 consumed and the amount of H202 formed (measured fluorometrically). The amount of O2 consumed during the redox cycling of resorufin decreased -50% in the presence of catalase, whereas the rate of O2 consumption decreased in the presence of superoxide dismutase. These results suggest that, during the reoxidation of reduced resorufin, O2 is converted to H202 via superoxide anion. Experiments with acetylated cytochrome c further implicated superoxide anion as an intermediate in the reduction of O2 to HzOz. However, the ability of reduced resorufin to reduce acetylated cytochrome c directly (i.e., without first reducing 0, to superoxide anion) precluded quantitative measurements of superoxide anion formation. Superoxide dismutase, but not catalase, increased the steady-state level of reduced resorufin and considerably delayed its reoxidation. This indicates that superoxide anion is not only capable of reoxidizing reduced resorufin, but is considerably more effective than molecular O2 in this regard. Overall, these results suggest that NADPH-cytochrome P450 reductase catalyzes the one-electron reduction of resorufin (probably to the corresponding semiquinoneimine radical)

’ This research was supported by Grants ES 03765 and GM 37044 (awarded to A.P.), by Grant ES 04092 (awarded to G.A.R.) from the National Institutes of reer Development Award (ES 00166) from the Na- Health, and by BRSG SO7 RR05373. D.R.D. is sup- tional Institutes of Health. A preliminary account of ported by Training Grant ES 07079 from the National this work was presented in abstract form (D. R. Dut- Institutes of Health, and by Stauffer Chemical Co., ton and A. Parkinson (1987) Fed. Proc. 46,1957). Farmington, CT. A.P. is a recipient of a Research Ca- *To whom correspondence should be addressed.

605 0003-9861189 $3.00 Copyright Q 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.

606 DUTTON. REED, AND PARKINSON

which can either undergo a second, one-electron reduction (presumably to the corresponding dihydroquinoneimine) or a one-electron oxidation by reducing molecular O2 to superoxide anion. The superoxide anion formed is then converted to hydrogen peroxide, either by a second, one-electron oxidation of reduced resorufin or by dismutation. In the accompanying paper, we show that the redox cycling of the 7-alkoxyresorufins by NADPH-cytochrome P450 reductase significantly affects their 0-dealkylation by purified isozymes of rat liver microsomal cytochrome P450. ii 1989 i\cademic Press. Inc.

The 0-dealkylation of various ‘7-alkoxyresorufins to the highly fluorescent compound, resorufin (7-hydroxyphenoxazone), provides a rapid, sensitive, and convenient method to study the induction of different forms of cytochrome P450 in liver microsomes (l-7). In rats, for example, the O-dealkylation of 7-ethoxyresorufin is catalyzed by cytochrome P45OcF whereas the 0-dealkylation of 7-pentoxy- and 7-benzyloxyresorufin is catalyzed by cytochrome P450b. The 30- to 60-fold induction of liver microsomal cytochrome P45Oc that results from treating rats with 3-methylcholanthrene (9, 16) is associated with a marked induction (>30 fold) of the 0-dealkylation of 7-ethoxyresorufin (l-5). Similarly, the 30- to 60-fold induction of cytochrome P450b that results from treating rats with phenobarbital (9, 16) is associated with a marked induction (>30 fold) of the O-dealkylation of 7-pentoxy- and 7-benzyloxyresorufin (4-7).

The direct fluorometric assav of 7-alkoxyresorufin 0-dealkylase acti;ity devel- oped for microsomal preparations must be modified for isolated rat hepatocytes and postmitochondrial supernatant (S9) frac- tions to avoid interference from DT-diaphorase (NAD(P)H-quinone oxidoreductase, EC 1.6.99.2) (17, 18). This cytosolic flavoprotein interferes with the 7-alkoxyreso-

3 Recently, a nomenclature system for cytochrome P450 isozymes was proposed based on the genes of the cytochrome P450 superfamily whose cDNA and/or amino acid sequences have been determined (8). We have included the new committee on Standardized Nomenclature conventions in boldface in this foot- note. The nomenclature system of Ryan et al. (S-11) is used throughout the manuscript. Cytochrome P450b (P450IIBl) is also known as P450 PB-B (12,13), P450 PB-4 (14), or P450 I-C (15); and cytochrome P45Oc (P450IAI) as P450 BNF-B (12.13).

rufin 0-dealkylase assay by reducing the liberated resorufin to a nonfluorescent product. This problem can be eliminated by incubating samples in the presence of the DT-diaphorase inhibitor, dicumarol, or by measuring fluorescence intensity after reactions are terminated by addition of or- ganic solvents, such as methanol or ace- tone (17-20). The latter method is based on the observation that reduced resorufin re- oxidizes after inactivation of DT-diaphorase, with complete restoration of its fluorescent properties (17,18).

More recently, a second problem with the 7-alkoxyresorufin 0-dealkylation as- says has emerged from studies with purified isozymes of rat liver cytochrome P450 (7, 21, 22j. As expected, purified cytochrome P45Oc, when reconstituted with NADPH-cytochrome P450 reductase and lipid, catalyzes the 0-dealkylation of 7- ethoxyresorufin at a rate which exceeds that catalyzed by liver microsomes from 3- methylcholanthrene-induced rats (23, 24). In contrast, purified cytochrome P450b is an unexpectedly poor catalyst of the O-dealkylation of 7-pentoxy- and 7-benzyloxy- resort& (7, 21, 22). Purified cytochrome P450b is, however, an effective catalyst of many other biotransformation reactions, including the 0-dealkylation of 7-ethoxy- coumarin (23, 25). Furthermore, antibody against cytochrome P450b effectively in- hibits the high rate of 7-pentoxy- and 7- benzyloxyresorufin 0-dealkylation catalyzed by liver microsomes from phenobarbital-induced rats (6,7,22).

We have undertaken studies to determine why purified cytochrome P450b, when reconstituted with NADPH-cytochrome P450 reductase and lipid, is an unexpectedly poor catalyst of the O-dealkylation of 7-pentoxy- and 7-benzyloxyreso-

RAT LIVER MICROSOMAL NADPH-CYTOCHROME Pd50 REDUCTASE 607

rufin. The studies described in this paper show that NADPH-cytochrome P450 reductase catalyzes the reduction of resorufin and the 7-alkoxvresorufins which, in the presence of 02, initiates a redox cycling reaction. Evidence is presented which suggests that the reosidation of reduced resorufin involves a one-electron reduction of molecular 0, to superoxide anion, which in turn forms hydrogen peroxide either by a second one-electron oxidation of reduced resorufin or by dismutation. The studies described in the accompanying paper (22) show that reduction of the 7-alkoxyresorufins by NADPH-cytochrome P450 reductase impedes their 0-dealkylation by cytochrome P450b, but actually enhances their 0-dealkylation by cytochrome P45Oc.

EXPERIMENTAL PROCEDURES

CXenricn/s. Catalase (bovine liver), superoxide dismutase (bovine erythrocyte), horseradish peroxidase, cytochrome c, 3-(Ghydroxyphenyl)propionic acid, and diethylenetriaminepentaacetic acid (DETA- PAC? were purchased from Sigma Chemical Co. (St. Louis, MO). Resorufin (hgdroxyphenoxazone) and various 7-alkoxyresorufins were obtained from both Pierce Chemical Co. iRockford. IL) and Molecular Probes (Junction City, OR). Dicumarol [3.3’-methy- lenebis(f-hydroxycoumarin)] was purchased from Aldrich Chemical Co. (Milwaukee, WI). Acetylated cytochrome c was prepared according to the method of Wada and Okunuki (26).

Purijfcatio~ of ,~‘.~DPH-clltochrolrte P&s~J reductase. NADPH-cytochrome P450 reductase was purified from rat liver microsomes as described by Yasu- kochi and Masters (27). with modifications described previously (21). One nanomole of purified enzyme reduced -3 rmol cytochrome c per minute at 22°C in the presence of 330 mhl potassium phosphate buffer (pH 7.31, 1 mhf EDTA, 3 mM MgClz, 100 pM KCN, 50 pbr cytochrome c, and 100 pM NADPH.

Reduction qi resorufi n end j’-cllkoryresortili,cs. The reduction of resorufin (final concentration 2.5 pM in 100 rnbl potassium phosphate buffer, pH ‘7-1) was measured spectrophotometrically at room temperature (22°C) as a decrease in absorbance at 570 nm (17,28). Similarly, the reduction of ‘I-ethoxy-, 7-pentoxy-, and P-benzyloxgresorufin was monitored at 482, 433, and 138 nm, respectively. Spectra mere recorded on an SLM-Aminco DW-dC dual beam spectrophotometer.

’ Abbreviation used: DETAPAC, diethylenetriaminepentaacetic acid.

For chemical reduction, a few grains of sodium dithionite were added to the contents of the sample cuvette. For enzymatic reduction, NADPH-cytochrome P-150 reductase (0.1-1.0 nmol) was added to both the reference and sample cuvettes (final volume 1.0 ml) and reactions were initiated by addition of NADPH (final concentration 100 PM). The NADPH was added to the reference cuvette, so that a decrease in absorbance due to reduction of resorufin in the reference cuvette was displayed as an apparent increase in absorbance in the sample cuvette.

N.lDPH and oqyen cowumpticm. NADPH and Oe consumption were measured to determine the stoichiometry of resorufin reduction catalyzed by NADPH- cytochrome Pd50 reductase. Nt\DPH consumption was measured spectrophotometrically as a decrease in absorbance at 3-10 nm, under the same conditions used to measure the reduction of resorufin (see above). Oxygen consumption was measured with a Clark oxygen electrode under slightly different conditions. Under identical conditions to those used to measure resoruhn reduction or NADPH oxidation. the rate of Oz consumption was too slow to measure reliably. To increase the rate of O2 consumption, the concentration of resorufin was increased to 25 pM, and the temperature raised to 37°C. The final concentration of NADPH-cytochrome Pd50 reductase ranged from 0.1 to 1.0 phi in a final volume of 1.8 ml. Reactions were initiated by the addition of 180 nmol NADPH (final concentration 100 picl). The concentration of O;, was estimated to be 197 pht assuming 100% satura- tion at 37°C in a 21”; O4 atmosphere.

Hydroyetr pemride J;wwo tiox Hydrogen peroxide formation was measured by the fluorometric method of Zaitsu and Ohkuma (29). Reactions were carried out in 0.5-ml incubation mixtures at room temperature, and contained 0.5 ~hl NADPH-cytochrome P150 reductase, 2.5 pM resorufln, and 50 PM NADPH. Reac- tions were terminated after 0. 5. 10, 15, or 20 min by the addition of 50 ~1 trichloroacetic acid (15% ). Each tube was treated with 200 ~1 of 7.5 mhl3-(l-hydroxy-

phengl jpropionic acid, 2 ml of 150 mM Tris-HCI buffer, pH 8.5, and 100 ~1 of 2 units/ml horseradish peroxidase. Aftera lo-min incubation at 22°C insolu- ble material was removed hy centrifugation (2000~ for 5 min). Fluorescence emission intensity was measured at 404 ntn, with escitation at 320 nm. Standards for recovery and quantitative analysis were prepared from commercially available H,Op (Sigma Chemical Co.), the concentration of which was verified by titra- tion against potassium permanganate according to American Chemical Society specifications (30).

Supero.ridc trniolr~h~,,lati0)2. Superoxide anion production was determined spectrophotometrically from the rate of reduction of acetylated cytochrome c at 550 nm in the presence and absence of superoxide dismutase (final concentration 100 units/ml). The superoxide dismutase-inhibitable rate of reduction of acet-

608 DUTTON, REED, AND PARKINSON

ylated cytochrome c was used to quantitate the amount of superoxide anion based on an extinction coefficient of ‘21 mM-’ em-’ for the reduced form of acetylated cytochrome c (31). It has been reported previously that acetylation of cytochrome c causes a much greater decrease in its rate of reduction by NADPH-cytochrome P450 reductase compared to its reactivity with superoxide anion, for which reason acetylated cytochrome c is preferentially reduced by superoxide anion (32). The acetylated eytochrome c prepared for these experiments was reduced by NADPH-cytochrome P450 reductase at 3-4s of the rate of reduction of native cytochrome c.

Effects of cat&use and superoxide dismutase. Where indicated, reaction mixtures also contained 20 rg/ml catalase or superoxide dismutase (480 and 60 units/ ml, respectively) to determine the effects of these en- zymes on the rate of NADPH oxidation, Oa consumption, and resorufin reduction catalyzed by NADPH- cytochrome P450 reductase.

RESULTS

It has been reported previously that the oxidized and reduced form of resorulin can be distinguished spectrophotometrically and fluorometrically: Only the oxidized form of resorufin absorbs visible light (X,,, - 570 nm) and is fluorescent (X em,s1** - 585 nm, Lcitation - 530 nm) (17,28). As expected, addition of a few grains of the chemical reductant, sodium dithionite, to a solution of 2.5 ~.LM resorufin caused a complete loss of absorbance at 570 nm, as shown in Fig. 1. However, this loss of absorbance was temporary, inasmuch as the absorbance at 570 nm was fully restored after 2-5 min (the time varied depending on how much solid sodium dithionite was added to the sample cuvette). The fluorescent properties of resorufin, which were lost after addition of sodium dithionite, were also restored after 2-5 min. Sodium dithionite also caused a temporary bleach- ing when added to solutions of oxidized ethoxyresorufin (A,,,,, - 482 nm), pentoxyresorufin (X,,, - 372 and 433 nm), and benzyloxyresorufin (X,,, - 375 and 438 nm).

As shown in Fig. 1, the absolute spectra of the alkoxyresorufins differed significantly from each other and from that of resorufin. In contrast to ethoxyresorufin and resorufin, a 10 FM solution of pentoxy- and benzyloxyresorufin displayed two distinct

RESORUFIN ETHOXVRESORUFIN

m

OXIDIZED OD5

0.05 I 5

D@

I 1

REDUCED

450 550 550 5s

PENTOXVRESORUFIN SENZVLOXYRESORUFIN

lD

5 I m

0.02 I 0.02

E!ilEg

5

1

1

350 450 550 650 550

WAVELENGTH (NM)

FIG. 1. Absolute spectra of resorufin and ?-alkoxyresorufins. Absolute spectra of resorufin (2.5 pM) and 7-ethoxy-, 7-pentoxy-, and T-benzyloxyresorufin (1,5, and 10 pM in 100 mM potassium phosphate buffer, pH 7.4) were recorded on an SLM-Aminco DW-2C spectrophotometer. A few grains of sodium dithionite were added to the sample prior to recording the absolute spectrum of reduced resorufin.

absorbance maxima, the ratio of which (h,,, - 375 nm to X,,, - 435 nm) increased when the concentration of pentoxy- and benzyloxyresorufin was increased from 1 to 10 PM (see Fig. l), or when the ionic strength of the potassium phosphate buffer was increased (results not shown). Conversely the ratio of the two absorbance peaks decreased when the polarity of the buffer was decreased by addition of etha- nol (results not shown). These results suggest that pentoxyresorufin and benzyloxyresorufin form aggregates or micelles at concentrations between 1 and 10 PM. Be- cause the absolute spectra of pentoxy- and benzyloxyresorufin were dependent on their concentration, and because the alkoxyresorufins absorbed light in the same region as NADPHcytochrome P450 reductase, most of the experiments were carried out with resorufin. It should be emphasized, however, that when experiments

RAT LIVER MICROSOMAL NADPH-CYTOCHROME P450 REDUCTASE 609

0.00 ‘00 rM NADH I I I

0 1 2 3 Time (min)

FIG. 2. Effects of NADH and dicumarol on the reduction of resorufin by NADPH-cytochrome P450 reductase. A 950-PI sample containing 2.5 nmol resorufin and 1.0 nmol NADPH-cytochrome P450 reductase in potassium phosphate buffer (100 mM, pH 7.4) was placed in both the sample and reference cuvettes. A 50-p] aliquot of buffer containing 1 mM MgCl, was added to the sample cuvette, and 50 ~1 of 2 mM

NADPH or NADH in the same buffer was added to the reference cuvette. The change in absorbance at 570 nm was recorded as a function of time. The reaction initiated with NADPH was also monitored in the presence of 100 pM dicumarol (added to both euvettes in 10 ~1 of methanol).

were carried out with all four compounds, the results obtained with the alkoxyresorufins were qualitatively identical to those obtained with resorufin.

Enzymatic Reductiw of ResoruJn

Previous studies have shown that resorufin can be reduced by the cytosolic enzyme, DT-diaphorase (1’7, 18), which catalyzes a two-electron transfer from NADH or NADPH to a variety of quinones and azo-dyes (33,34). The results in Fig. 2 dem- onstrate that NADPH-cytochrome P450 reductase is also capable of catalyzing the reduction of resorufin. However, in contrast to the reaction catalyzed by DT- diaphorase, the reaction catalyzed by NADPH-cytochrome P450 reductase was not supported by NADH, nor was it inhibited by dicumarol (even at 10X the concentration known to inhibit DT-diaphorase completely (17). These results established that the ability of NADPH-cytochrome P450 reductase to catalyze the reduction of resorufin was not due to contamination of

the purified enzyme with DT-diaphorase. The reduction of resorufin by NADPH-cytochrome P450 reductase was unaffected by the presence of 10 /IM DETAPAC, which argues against a role for iron in this reaction (35,36).

As expected, the rate of reduction of resorufin (AA&min) was dependent on the concentration of NADPH-cytochrome P450 reductase. Unexpectedly, the extent of resorufin reduction (AAs& was also dependent on the concentration of NADPH- cytochrome P450 reductase, as shown in Fig. 3. At the lowest concentration of NADPH-cytochrome P450 reductase tested (0.1 nmol/ml), approximately 15% of the resorufin was in the reduced form during steady-state conditions, whereas -90% was in the reduced form at the highest concentration of NADPH-cytochrome P450 reductase tested (1.0 nmol/ ml). Similar results were obtained with ‘7- ethoxy-, 7-pentoxy-, and 7-benzyloxyresorufin. The results in Fig. 3 indicate that reduced resorufin does not continue to accu- mulate indefinitely, which suggests that the overall extent of reduction represents the balance between the rate of resorufin

[fteductase]

0 1 2 3 Time (min)

FIG. 3. Effects of NADPH-cytochrome P450 reductase concentration on the rate and extent of resorufin reduction. A 950-pl sample containing 2.5 nmol resorufin and 0.1 to 1.0 nmol of NADPH-cytochrome P450 reductase in potassium phosphate buffer (100 mM, pH 7.4) was placed in both the sample and reference cuvettes. A 50-p] aliquot of buffer containing 1 mM MgCl, was added to the sample cuvette, and 50 ~1 of 2 mM NADPH in the same buffer was added to the reference cuvette. The change in absorbance at 570 nm was recorded as a function of time.


A -ABSORBANCE 340 NM

0 -ABSORBANCE 570 NM

FIG. 4. Relationship between NADPH utilization and the reduction of resorufin by NADPH-cgto- chrome P450 reductase. A 950~rl sample containing 2.5 nmol resorufin and 0.1 to 1.0 nmol of NADPH-cytochrome P450 reductase in potassium phosphate buffer (100 mM, pH 7.4) was placed in both the sample and reference cuvettes. A 5Oyl aliquot of buffer containing 1 mM MgCla was added to the sample cuvette, and 50 ~1 of 2 mM NADPH in the same buffer was added to the reference cuvette. The reduction/reoxidation of resort& (0) was measured as the change in absorbance at 570 nm. The disappearance of NADPH (A) was measured as the change in absorbance at 340 nm.

reduction and the rate of reoxidation. In support of this conclusion, a Nz atmosphere increased the st,eady-state level of reduced resorufin, and NADPH-cytochrome P450 reductase (1 nmol/ml) completely reduced resorufin in the absence of O2 (results not shown).

NA DPH Oxidation

The relationship between the time course of resorufin reduction (and reoxidation) and the time course of NADPH consumption by NADPH-cytochrome P450 reductase is shown in Fig. 4. Addition of NADPH (100 FM) to a mixture of NADPH- cytochrome P450 reductase (1 PM) and resorufin (2.5 PM) caused a rapid reduction of resorufin, as indicated by an 80-85s decrease in the absorbance at 570 nm. Al- though resorufin was maximally reduced within 30 s, the oxidation of NADPH con- tinued for 8-9 min, at which time all of the added NADPH was consumed (based on the loss of absorbance at 340 nm). This re-

sult indicated that continual reduction by NADPH was required to maintain steady- state levels of reduced resorufin. The total amount of NADPH consumed (100 nmol) far exceeded the amount of resorufin present (2.5 nmol). Once all of the NADPH had been oxidized, the absorbance at 570 nm gradually increased to >99% of the origi- nal absorbance value, indicating the complete reoxidation of reduced resorufin. The results in Fig. 4 provide further evidence that resorufin undergoes a redox cycling reaction in the presence of NADPH and NADPH-cytochrome P450 reductase.

In a series of preliminary experiments, we determined that O2 was consumed during the reoxidation of reduced resorufin. However, under the experimental conditions used to measure resorufin reduction and NADPH consumption, the rate of 0, consumption was too slow to measure accurately with a Clark oxygen electrode. To increase the rate of O., consumption, the concentration of resorufin was increased lo-fold to 25 PM, and the temperature was raised from 22 to 37°C. These conditions were used to determine the relationship between O2 consumption and the concentration of NADPH-cytochrome P450 reduct.ase. As shown in Fig. 5, the rate of Oe consumption, but not the amount of 0, consumed, was dependent on the concentration of NADPH-cytochrome P450 reductase. The total amount of O2 consumed was directly dependent on and propor- tional to the amount of NADPH added to initiate the reaction. The results in Fig. 5 indicate that addition of 180 nmol of NADPH resulted in the consumption of approximately 150 nmol Oz, which likely represents a 1:l stoichiometric relationship between the amount of NADPH oxidized and the amount of O2 consumed.

Oxygen consumption was also measured in the presence of catalase and superoxide dismutase, and the results are shown in Fig. 6. Catalase decreased the amount of O2 consumed by 50% which, together with the 1:l stoichiometry between NADPH oxidation and 0, consumption, indicated that O2


I I -1 0 1 2 3 4

Time (Min)

FIG. 5. Consumption of O2 during the reduction of resort& by varying concentrations of NADPH-cytochrome P-150 reductase. Oxygen consumption was measured with a Clark oxygen electrode. The reaction vessel contained resorufin (25 PM) and NADPH-cytochrome P450 reductase (0.1 to 1.0 pM) in 1.8 ml potassium phosphate buffer (100 mM, pH 7.4) at 37°C. Reac- tions were initiated by the addition of 180 nmol NADPH.

was being converted to H202 during the reduction of resorufin by NADPH-cytochrome P450 reductase. Superoxide dismutase had no effect on the amount of O;? consumed but significantly decreased the rate of O2 consumption, implicating superoxide anion as an intermediate in the formation of HY09.

Formation of Hydrogen Peroride

The fluorometric method of Zaitsu and Ohkuma (29) was used to verify that H202 was formed during the reoxidation of enzy- matically reduced resorufin. The amount of H202 formed was equivalent to the amount of NADPH added to initiate the reaction. For example, addition of 25 nmol NADPH resulted in the formation of 23-24 nmol HzOz. This 1:l stoichiometry between amount of NADPH added and the amount of Hz02 generated is consistent with the results of the O2 consumption experiments described above.

Formatiolt of Superoxide A,~im

As shown in Fig. 6, superoxide dismutase decreased the rate, but not the amount

of O2 consumed during the enzymatic reduction of resorufin. This effect of superoxide dismutase suggests that superoxide anion is an intermediate in the production of H202 during the reoxidation of reduced resorufin. Therefore, an attempt was made to measure the formation of superoxide anion spectrophotometrically by measuring the rate of reduction of acetylated cytochrome c at 550 nm in the presence and absence of superoxide dismutase. According to this method (31, 32, 39), the superoxide dismutase-inhibitable reduction of acetylated cytochrome c gives a measure of su- peroside anion levels. In preliminary experiments, we verified that acetylation of cyt.ochrome c decreased by more than 96% its rate of reduction by NADPH-cytochrome P450 reductase, and that the reduction of acetylated cytochrome c by a xanthineixanthine oxidase system (which generates superoxide anion) could be completely inhibited by 400 units/ml superoxide dismutase (results not shown).

In the presence of NADPH and NADPH-cytochrome P450 reductase, acetylated cytochrome c was slowly reduced, and, as expected, this slow rate of reduction was unaffected by the presence

Reductase alone

-1 0 1 2 3 4 Time (Min)

FIG. ti. Effects of catalase and superoxide dismutase on Oy consumption during the enzymatic reduction of resorufin. Oxygen consumption was measured with a Clark oxygen electrode. The reaction vessel contained resorufin (25 pchl), NADPH-cytochrome P450 reductase (1.0 PM), and either catalase or superoxide dismutase (20 pg/ml) in 1.8 ml potassium phosphate buffer (100 mM, pH 7.4) at 37°C. Reactions were initiated by the addition of 180 nmol NADPH.


30 60 so 120 TIME (SECONDS)

FIG. 7. Rate of reduction of acetylated cytochrome c by NADPHcytochrome P450 reductase in the presence and absence of resorufin and superoxide dismutase. A 990~pl sample containing 50 nmol of acetylated cytochrome c in 330 mM potassium phosphate buffer, pH 7.4, containing 3 mM MgCl?, 1 mM EDTA, and 100 pM KCN was placed in both the sample and reference cuvettes. In addition, each cuvette also contained 5 pmol of NADPH-cytochrome P450 reductase and 0, 2.5, or 25 nmol of resort&r, labeled 1, 2, and 3 respectively. A lo-p1 aliquot of 100 mM potassium phosphate was added to the reference cuvette and 10 ~1 of 10 mM NADPH in the same buffer was added to the sample cuvette. The rate of reduction of acetylated cytochrome c was measured in the presence (---) and absence (-) of superoxide dismutase by monitoring the change in absorbance at 550 nm.

of superoxide dismutase (Fig. 7). The rate of reduction of acetylated cytochrome c increased when 2.5 PM resorufin was added to the incubation mixture, and this stimula- bory effect was partially reversed by superoxide dismutase (Fig. ‘7). These results are consistent with formation of superoxide anion during the reoxidation of reduced resorufin by Oz. As shown in Fig. 7, the ability of resorufin to stimulate the reduction of cytochrome c was not completely inhibited by superoxide dismutase. A likely explanation for this result is that the reduced form of resorufin can reduce acetylated cytochrome c directly, instead of reducing O2 to superoxide anion. In this case, acetylated cytochrome c and O2 would compete for electrons from reduced resorufin. In support of this explanation, the ability of superoxide dismutase to reverse the stimu-

latory effect of resorufin on the reduction of acetylated cytochrome c decreased when the concentration of resorufin was increased (Fig. 7) or when the concentration of acetylated cytochrome c was increased (results not shown). If the reduced form of resorufin can reduce acetylated cytochrome c directly, the results in Fig. 7 would underestimate the amount of superoxide formed during the reoxidation of reduced resorufin in the absence of acetylated cytochrome c. For this reason, no further attempt was made to determine the stoichiometry between NADPH consumption and superoxide anion formation during the enzymatic reduction of resorufin.

Eflects of Catakse a.nd Superoxide Dism.utase on the Enxymatic Reduction qf Resorujin

As shown in Fig. 8, NADPH-cytochrome P450 reduct.ase caused a more intense and prolonged reduction of resorufin when superoxide dismutase was added to the incubation mixture. In contrast, catalase had little effect on the steady-state level of reduced resorufin, and slightly increased, rather than decreased, its rate of reoxidation. These results suggest that the reoxidation of reduced resorufin involves a one- electron reduction of 0, to superoxide anion, and that the superoxide anion so formed can also oxidize the reduced form of resorufin (by a one-electron reduction of superoxide anion to hydrogen peroxide). The results further indicate that superoxide anion is considerably better than Oz at reoxidizing the reduced form of resorufin. Consequently, the reduced form of resorufin accumulated to a greater extent and persisted for a longer time in the absence of superoxide anion. Its utilization in the reoxidation of reduced resorulin is another reason why superoxide anion formation could not be quantitated accurately.

DISCUSSION

The results described here are from the first of a two-part study to determine why purified rat cytochrome P450b, when


Reductase +

Reductase + Catalase

Time (Min)

FIG. 8. Effects of catalase and superoxide dismutase on the reduction of resorufin by NADPH- cytochrome P450 reductase. A 950-J sample containing 2.5 pM resorufin, 0.5 pM NADPH-cytochrome P450 reductase, and either catalase or superoxide dismutase (20 pg/ml) was placed in both the reference and sample cuvettes. A 504 aliquot of buffer containing 1 tTIM MgCla was added to the sample cuvette, and the reaction was initiated by the addition of 50 ~1 of 2 mM NADPH in the same buffer to the reference cuvette.

reconstituted with NADPH-cytochrome P450 reductase and lipid, is an unexpectedly poor catalyst of 7-pentoxy- and 7-benzyloxyresorufin 0-dealkylation (7, 21, 22). This problem with purified cytochrome P450b is the second problem to arise with the fluorometric assay of 7-alkoxyresorufin 0-dealkylation; the first was an underesti- mation of enzyme activity in liver samples containing the cytosolic enzyme, DT-diaphorase, which reduces the reaction product, resorufin, to a nonfluorescent compound (17, 18). Interference from DT-diaphorase can be overcome by two methods (see the introduction), neither of which solves the problem of the unexpectedly low 7-alkoxyresorufin 0-dealkylase activity of purified, reconstituted cytochrome P450b (see accompanying paper).

The results of the present study indicate that purified NADPH-cytochrome P450 reductase catalyzes the reduction of resorufin (and 7-ethoxy-, 7-pentoxy-, and 7- benzyloxyresorufin) to a product or prod- ucts that readily reoxidize in the presence of molecular OZ. A possible scheme for this redox cycling reaction is shown in Fig. 9. According to this proposed scheme, NADPH-cytochrome P450 reductase cata-

lyzes the one-electron reduction of resorufin to a semiquinoneimine radical. Al- though the structures of the reduced form(s) of resorufin have not been determined, those shown in Fig. 9 are analogous to the structures of other quinone- and quinoneimine-containing compounds that undergo redox cycling reactions (38-41). Formation of a semiquinoneimine radical, which disrupts the conjugated double bond system of resorufin, is consistent with the observed loss of absorbance and fluorescence (see Fig. l), and is consistent with univalent electron transfer by NADPH- cytochrome P450 reductase (39, 42-44). It is possible that NADPH-cytochrome P450 reductase catalyzes two sequential one- electron reductions to convert resorufin to the corresponding dihydroquinoneimine (which also lacks a conjugated double bond system), as shown in Fig. 9. This dihydroquinoneimine is presumably the colorless, nonfluorescent product formed by the reduction of resorufin by DT-diaphorase, which is known to catalyze a two-electron transfer from NAD(P)H to various quinones and azo-dyes (33, 34). The ability of NADPH-cytochrome P450 reductase to reduce resorufin was not due to contamina-


DT- DIAPHORASE I /- Y

QUINONEIMINE SEMI-PUINONEIMINE RADICAL DlHYDFiOQUlNONElMlNE

FIG. 9. Proposed scheme for the redox cycling of resorufin catalyzed by NADPH-cytochrome P&O reductase. The quinoneimine is either resorufin (7-hydroxyphenoxazone, R = H), 7-ethoxgresorufin iR = C2H5-j, 7-pentoxyresorufin (R = C7Hi5-), or 7-benzyloxpresorufin (R = C6H5-CH2-). The structures of the semiquinoneimine radical and dihgdroquinone are inferred by analogy to other quinone and quinoneimine structures that undergo redox cycling reactions (38-41).

tion of the purified flavoprotein with DT- diaphorase because the reaction was unaffected by dicumarol, a known inhibiter of DT-diaphorase, and was not supported by NADH, which is a cofactor for DT-diaphorase but not NADPH-cytochrome P450 reductase (Fig. 2).

Reoxidation of reduced resorufin ulti- mately resulted in the reduction of molecular O2 to H202. The 1:l:l stoichiometry among NADPH oxidation, O2 consumption, and H202 formation depicted in Fig. 9 was verified experimentally. The formation of H202, which was measured directly by a fluorometric assay, could also be inferred from the effects of catalase, which decreased O2 consumption by -50% (Fig. 6). Several lines of evidence indicated that the conversion of O2 to H202 involved the intermediacy of superoxide anion, as shown in Fig. 9. The rate but not the extent of O2 consumption was retarded in the presence of superoxide dismutase, and formation of superoxide anion during the en-

zymatic reduction of resorufin could be de- tected as the superoxide dismutase-inhibitable reduction of acetylated cytochrome c (Fig. 7). However, the ability of reduced resorufin to reduce acetylated cytochrome c directly (i.e., without first reducing O2 to superoxide anion) and the ability of superoxide anion to reoxidize the reduced form of resorufin precluded quantitative measurements of superoxide anion formation.

An important part of the reaction scheme proposed in Fig. 9 is the ability of superoxide anion to reoxidize the reduced form of resorufin, particularly because superoxide anion was considerably more effective than molecular O2 in this regard. In the absence of superoxide anion (i.e., in the presence of superoxide dismutase), en- zymatically reduced resorufin accumulated to a greater extent and persisted for a longer time, as shown in Fig. 8. These results indicate that the reoxidation of reduced resorufin involves a relatively slow one-electron reduction of O2 to superoxide

RAT LIS’ER MICROSOMAL NADPH-CYTOCHROME PJ50 REDUCTASE 615

anion, followed by a rapid one-electron reduction of superoxide anion to hydrogen peroside. It should be emphasized that, in terms of NADPH oxidation, O2 consumption, and H202 formation, the stoichiometry of the reactions shown in Fig. 9 is the same regardless of whether superoxide anion participates in the reoxidation of reduced resorulin.

In conclusion, we have established that NADPH-cytochrome P450 reductase catalyzes the redox cycling of resorufin (and various 7-alkoxyresorufins), which results in the conversion of O2 to HpOs via superoxide anion. The reduced form of resorufin (and the 7-alkoxyresorufins) is proposed to be a semiquinoneimine radical (one-electron reduction product) and/or the corresponding dihydroquinoneimine (two-electron reduction product), as shown in Fig. 9. We have also shown that the two sub- strates for cytochrome P450b, namely 7- pentoxy- and 7-benzyloxyresorufin, form aggregates or micelles at concentrations typically used in incubation mixtures (l-10 pbf), whereas the subst.rate for cytochrome P45Oc, ethoxyresorufin, does not (Fig. li. The results of this study provide several possible explanations for the unexpectedly low catalytic activity of purified cytochrome P450b toward 7-pentoxy- and 7- benzvlosyresorufin, which contrasts with the high catalytic activity of purified cytochrome P45Oc toward 7-ethoxyresorufin. These include destruction of cytochrome P450b (but not P45Ocj by the reactive oxygen species (superoxide anion and H,O,) generated during the redox cycling reactions, or decreased substrate availability due to the formation of aggregatesjmi- celles of 7-pentoxy- and 7-benzyloxyresorufin (but not 7-ethoxyresorufin). How- ever, in the accompanying paper (22) we show that the unexpectedly low catalytic activity of purified cytochrome P450b toward 7-pentoxy- and 7-benzylosyresorufin is related to the ability of NADPH-cytochrome P450 reductase to reduce these alkoxyresorufins. More specifically, we dem- onstrate that reduction of the 7-alkoxyreso&ins by NADPH-cytochrome P450 reductase impedes their 0-dealkylation by

cytochrome P450b, but actually enhances their 0-dealkylation by cytochrome P45Oc.

REFERENCES

1. BLIRKE, M. D.. ANn MATER, R. T. (1974) Drug Metab. Dispos. 2.583-588.

2. BURKE, M. D., ANn MAYER, R. T. (1975) Drug

Metub. Disp. 3,245-253. 3. BURKE, M. D., PROUGH, R. .4., ANn MAYER, R. T.

(1977) Drug M&b. Dispos. 5, l-8.

4. BURKE, M. D., ANn MAYER, R. T. (1983) Uwm.

Bid. Internet. 45.243-258. 5. BURKE, M. D., FALZON, M., AND MILTON, A. S.

(1983) Bioctwm Ph~~rmxd. 32,389-397. 6. BLIRRE, M. D., THOMPSON, S.. ELCOMBE, C. R.,

HALPERT, J., H.IAPARANTA, T., ANn MAYER,

R. T. (19851 Bid/em. Phamncd. 34, 3.737 3345.

7. LUBET, R. A., MAYER, R. T.. CA~IERON, J. IV., NIW,

R. IV’., BURKE, M. D., WOLFF, T., ANn GUENGE-

RICH, F. P. (1985) Arch. Biochetn. Biophys. 238, 13-18.

8. NEBERT. D. IV, ;ZnEsNIK, M., COON, M. J., ESTA-

BROOK, R. W., GONZALEZ, F. J., GUENGERICH,

F. P., G~JNSALU~, I. C., JOHNSON, E. F., KEMPER,

B., LEVIN, W., PHILLIPS, I. R., S~TO, R., .&Nn W.&TERMAN. hf. R. (1987) D-V-4 6, l-11.

9. Ry.4~. D. E., THOMAS. P. E., REIK, L. M., ~NL)

LEVIN,!~. il988),~~llt,ltirtticrr 12,727-'i14.

10. CONNEY, A. H. (1986) Life Sci. 39,2-193-2518.

11. RYAN, D. E., Irrw S., WOOD, A. W’., THOMAS,

P. E., LIEBER, C. S., ANn LEVIN, IV. (1984) J.

Bid. Chenc. 259, 1239-1250.

12. GUENGERICH, F. P., DANNAN. G. A., WRIGHT,

S. T., MARTIN, M. W., AND KAMINSK~, L. S.

(1982) X~rmbioticcr 12, 701-716.

13. GUENGERICH, F. P., DANN.%N, G. A., WRIGHT.

S. T.. MARTIN, M. W., ANn KAMINSKT, L. S.

(19821 Birdzemistry 21,6019-6030. 1-i. ‘IlrAXfUN. D. J.. E;O, A., ;\Nn WALSH, C. (1983) J.

Bid. C/ie)n. 258, 11937-119-17.

15. KAMATXI, T., MAEn.& Ii., Y~M~ZOE, Y., NAGAI,

T., .wn KATO, R. (1982) Life Sri. 31,2603-2610.

16. THOMAS, P. E., REIK, L. hl., RYAN. D. E., ANn

LEVIN, W. (1981) J. Bid. Chem. 256,104+1052. 17. NIMS, R. W., PR~UGH, R. A., .wn LUBET, R. A.

(1984) Arch. Biochem. Bi~php. 299.459-469. 18. LUBET. R. A., NIIVIS, R. IV., MAYER, R. T., CAM-

ERUN, J. W.. ANL) SCHECHTM.IN, L. M. (1985)

Mutut. Res. 142 127-131.

19. POHL, R. J., AND ‘FOUTS, J. R. 11980) A&. Biw diem. 107,150-155.

20. GORSKI, J. R., ARLOTTO, M. P., KLAASSEN, C. D..

.wn PARK[NSON, A. (19851 Ccfrcingqenesis 6,

617-624.

21. DLITTON, D. R., AND PARKINSON. A. (1987) Fed. Proc. 46.1957.


22. Du~~oN, D. R., AND PARKINSON, A. (1989) Arch. Biochem. Bioph ys. 617-629.

23. GUENGERICH, F. P., DANNAN, G. A., WRIGHT, S. T., MARTIN, M. W., AND KAMINSKY, L. S. (1982) Biochemistry 21,6019-6030.

24. DUTTON, D. R., MCMILLEN, S. K., SONDERFAN, A. J.. THOMAS, P. E., AND PARKINSON, A. (1987) Arch. B&hem Biophys. 255,316-328.

25. RYAN, D. E., THOMAS, P. E., KORZENIOWSKI, D., AND LEVIN, W. (1979) J. BioL Chem. 254,1365- 1374.

26. WADA, K.. AND OKUNUKI, K. (1968) J. Biochem. 64, 667-681.

27. YASUKOCHI, Y., AND MASTERS, B. S. S. (1976) J. BioL Chem. 251,5337-5344.

28. KLOTZ, A. V., STEGMAN, J. J., AND WALSH, C. (1984) Anal. Biochem. 140,138-145.

29. ZAITSU, K., AND OHKUMA, Y. (1980) Anal. Bio- them. 109,109-113.

30. SCHMIDT, W. E., Ed. (1968) ill Reagent Chemicals: American Chemical Society Specifications, 4th ed., pp. 290-292, American Chemical Society Publications, Washington, DC.

31. MOREHOUSE, L. A., THOMAS, C. E., AND AUST, S. D. (1984) Arch. Biochem. Biophys. 232,366- 377.

32. AZZI, A., MONTECUCCO, C., AND RICHTER, C. (1975) Biochem. Biophys. Res. Common. 65,597-603.

33. ERNSTER, L., DANIELSON, L., AND LJUNGGREN, M. (1962) B&him. Biophys. Acta 58,171-188.

34. HUANG, M. T., MIWA, G. T., AND Lu, A. Y. H. (1979) J. BioL Chem. 254.3930-3934.

35. BUETTNER, G. R., OBERLEY, L. W., AND LEUT- HAUSER, S. W. H. C. (1978) Photochem. Photrr biol. 28,693-695.

36. HALLIWELL, B. (1978) FEBS L&t. 92,321-326. 37. FINKELSTEIN, E., ROSEN, G. M., PATTON, S. E., Co-

HEN, M. S., AND RAUCKMAN, E. J. (1981) Bie them Biophys. Res. Commun 102,1008-1015.

38. SMITH, M. T., EVANS, C. G., THOR, H., AND ORREN- IUS, S. (1985) in Oxidative Stress (Sies, H., Ed.), pp. 91-113, Academic Press, San Diego.

39. KAPPUS, H. (1986) B&hem. Pharmacol. 35,1-6. 40. Musso, H., VONGIZYC, V., KRAMER, H., AND DOPP,

H. (1965) Chem. Ber. 98,3952-3963. 41. Musso, H., AND ZAHORSKY, U. I. (1965) Chem. Ber.

98,39643980. 42. VERMILION, J. L., AND COON, M. J. (1978) J. Biol.

Chem. 253,8812-8819. 43. LAI, C. S., GROVER, T. A., AND PIETTE, L. H. (1979)

Arch. B&hem. Bioph ys. 193,373-378.

44. B~STERLING, B., AND TRUDELL, J. R. (1981) Bie them. Biophys. Res. Com mun. g&569-575.

Documents

Redox cycling of resorufin catalyzed by rat liver microsomal NADPH-cytochrome P450 reductase