Correlation of Kinetic Parameters of Nitroreductase ... - Semantic Scholar

THEJOURNAL OF BIOLOGICAL CHEMISTRY

Val. 264, No. 21, Issue of July 25, pp. 12379-12384, 1989 Printed in U.S.A.

Correlation of Kinetic Parametersof Nitroreductase Enzymes with Redox Properties of Nitroaromatic Compounds* (Received for publication, August 25, 1988)

Mary Virginia Ornaz and RonaldP. Mason From the Laboratory of Molecular Biophysics, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709

The well-established mechanism of regeneration of the parent nitro compound by the reaction of the nitro anion radical with oxygen in aerobic systems is the basis of the correlation of kinetic parameters of purified flavoenzymes with electron affinities of some selected nitroaryl and nitroheterocyclic compounds. We have found that there is a linear relationship between log V,,,/K, and the one-electron reduction potentials of these compounds and that the measured values of redox dependence for these compounds is similar to that determined by other methods. Our results support the proposal of a rate-determiningsingleelectrontransfer as the initial step in the reduction of nitro compounds by purified flavoenzymes and are discussed in terms of the Marcus electron transfer theory.

mation of reduction productsisnot a good index of free radical formation in uiuo because the stable productsof nitro anion radicals will always under-represent free radical formation. In addition, the reduction of oxygen to superoxide by the nitro anion radicals means that ithe n vitro rate of oxygen consumption canbe directly related to the kinetic parameters of the nitroreductaseenzymes. Wardman (14) has proposed that the reduction potential, E: for Reaction 2 in Fig. 1 in water at physiological pH is the most appropriate index of the redox properties of nitroaromatic compounds since it is the thermodynamic parameter which characterizes the relative ease of reduction of these compounds. Correlations between thesedatameasured by pulse radiolysis and other parameters, specifically, the logarithm of the rate of the reaction of nitro anion radicals with oxygen ( l l ) , yielded linear relationships. The purpose of the present investigationwas to establisha Nitroaryl and nitroheterocyclic compounds, including sub- correlation between the one-electron reduction potentials of stituted 2- and 5-nitroimidazoles,have enjoyed widespread nitro compounds with the kinetic parameters of nitroreducuse in medicine as antibiotics and radiosensitizers (1,2). The tase enzymes as measured by oxygen consumption. The nimost widely employed topical 5-nitrofuran, nitrofurazone, hastroaryl andnitroheterocyclic compounds (TableI) with a wide beenused as a food preservative, in adjunctive therapy of range of one-electron reduction potentials were selected on patients with second- and third-degree burns (3), and as an thebasis of their biomedicalsignificance. Another major antibacterial agent for the treatment or preventionof a wide consideration was that these compoundsbe soluble up to and variety of infections of the genito-urinary tract. Nitrofuran- preferably above their respective K,,, values. The two nitroreductase enzymes chosen were ferredoxin:NADP+ oxidoretoin is the substituted &nitrofuran administered most frequently for systemic infections, particularly those involving ductase and NADPH-cytochrome P-450 reductase, the former because of its potent nitroreductase activity and the latter theurinarytract(4). Benznidazole hasfound use as an antiprotozoal (5, 6), and metronidazole has been widely used because of its ubiquity in mammalian cells. for many years in the treatment of trichomoniasis, amoebiasis, EXPERIMENTALPROCEDURES and girardiasis, and a host of anaerobic bacterial infections (7). Pura-nitrobenzoate is a model substrate for nitroreducMaterials-Ferredoxin:NADP+ oxidoreductase (EC 1.18.1.2), nictase enzymes (8). otinamide adenine dinucleotide phosphate, reduced form (NADPH), Under anaerobic conditions, the major metabolic pathway nitrofurantoin, metronidazole, Trizma-7.7, buffer, diethylenetriafor biotransformationof nitroaryl andnitroheterocyclic com- minepentaacetic acid (DTPA),’ cytochrome c from horseheart (Type pounds appears tobe the reduction of the nitrogroup (9, 10). VI), pepstatin A, leupeptin hemisulfate, hypoxanthine, xanthine oxidase (EC 1.1.3.22), glucose 6-phosphate dehydrogenase (EC 1.1.1.49), The first intermediateof this reduction is generally the nitro bovine serum albumin,and D-glucose 6-phosphate were obtained from anion radical,whichrapidly air oxidizes (11). This set of Sigma. Nictotinamide adenine dinucleotide phosphate, catalase (from reactions leads to “futile metabolism” asproposed by Mason beef liver, EC 1.11.1.6) and superoxide dismutase (EC 1.15.1.1) were and Holtzman (8, 12), who first demonstrated by ESR spec- purchased from BoehringerMannheim. The 5-nitro-2-furaldehyde troscopy the formation of the nitro anion radical by a one- semicarbazone (nitrofurazone) was supplied by Aldrich, and the 4electron transfer from thereduced flavoenzyme according to nitrobenzoic acid was obtained from Fluka AG. The benznidazole was a gift from Hoffman-LaRoche and the deferoxamine (Desferal) mesthe sequence of Reactions 1 and 2 in Fig. 1 (13). Subsequently, ylate was a gift from Ciba Geigy. they proposed the productionof superoxide anion radical and Preparation of Microsomes-Hepatic microsomes were prepared the consequent regeneration of the parent nitro compound from 300-600 g, fed male C-D rats(Charles River). The animalswere (Reaction 3, Fig. 1).This futile metabolism implies the for- treated by daily intraperitoneal injection for 4 days with 80 mg/kg/ * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1. Permanent address: Dept. of Chemistry, College of New Rochelle, New Rochelle, NY 10801.

day phenobarbital and killed on the fifth day by CO, asphyxiation. The livers were twice washed and subsequently homogenized with six volumes of 1.15% KC1 (J. T. Baker Chemical Co.), 0.25 mM HEPES The abbreviations used are: DTPA, diethylenetriaminepentaacetic acid; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

12379

12380

Correlationof Kinetic Parameters with Redox Properties

FIG. 1. Reactions 1and 2, the proposed mechanism of nitro radical anion formation demonstrating the one-electron transfer from the reduced flavine; Reaction 3 the proposed mechanism of nitro compound-mediatedproduction of superoxide anion radical, andthe associated regeneration of the parent nitrocompound. (From Mason and Holtzman, (13)).

TABLEI The nitro compounds chosen for this study Compound Structure E: (VP Versus NHE

Nitrofurazone

Nitrofurantoin

o,NQ;=N-~J~-NH.

-0.257

o z ~ $ ) , ~ =N-

yo

N kNH

-0.264

0 O H II

I

CH2 C"N"CH2

I

Benznidazoie y

y

0

Cg

Hs

-0.380

RESULTS

2

$00-

p-Nitrobenzoate

0

-0.415

CH2 CHZOH Metronidazole

I

-0.486

Ref. 14. (Sigma) buffer, pH 7.4. The homogenates were centrifuged at 15,000 g for 10 min, the pellet was discarded, and the supernatant was centrifuged at 100,000 X g for 1 h. The microsomal pellets were washed with 0.25 M sucrose (J. T. Baker Chemical Co.), -0.2 mM EDTA (Sigma) and resuspended in 2-5 ml of sucrose solution/pellet. The pooled pellets were rehomogenized in a Potter-Elvehjem type homogenizer with a Teflon pestle at 5 "C. The pooled homogenates were snap-frozen in liquid nitrogen and stored at -70 "C. Purification of Enzyme-NADPH-cytochrome P-450reductase was purified from hepatic microsomes by the method of Yasukochi and Masters (15) as modified by Serabjit-Singh and co-workers (16). Microsomes solubilized with sodium cholate (Sigma) and Emulgen 911 (Kao Atlas Co., Tokyo) were chromatographed on DEAE-cellulose (Whatman DE52) to yield cytochrome and NADPH-cytochrome P-450 reductase fractions. The NADPH-cytochrome P-450 reductase was purified by chromatography on 2',5'-ADP Sepharose (Pharmacia, Uppsala, Sweden),eluted with NADP', and concentrated by overnight dialysis against 100 mM phosphate buffer, pH = 7.7, containing 20% glycerol and 0.1 mM EDTA. Enzyme activity was determined by measuring the rate of cytochrome c reduction at 550 nm as described by Masters et al. (17) as modified by Vermilion and Coon (18). One unit of NADPH-cytochrome P-450 reductase activity at 550 nm at 30 "C in a 1-cm light path corresponds to reduction of 1.0 nmol of cytochrome c/min/ml of reaction mixture. The NADPHcytochrome P-450 reductase activities preparedhad anaverage activX

ity of 2000 units/ml. Total protein was determined by the method of Lowry et al. (19) using the sodium dodecyl sulfate modification of Dulley and Grieve (20). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was run using 7.5% gradient gels (Pharmacia, Uppsala, Sweden) and a Bio-Rad dual vertical slab apparatus. Electrophoresis was carried out for 7 h at room temperature (21). Analytical Procedures-K, and V,,, determinations for the nitroreductase enzyme-nitroaromatic systems were measured by the rate of oxygen consumption using a Clark electrode (YSI-5331, Yellow Springs Instrument Company) calibrated according to the method of Holtzman(22) by measuring the fractional decrease in electrode current for the hypoxanthine-xanthine oxidase-catalase system. A measured value of 214.1 f 1.6 (S.E.) pmol/liter of 0, -763 torr was taken as the saturation concentration. A 1.8-ml water-jacketed glass vessel (Gilson Medical Electronics No.01271002) fitted with the electrode was maintained at 37 "C and filled with varying concentrations of nitro substrates dissolved in 0.05 M Trizma-7.7 buffer, pH 7.4, containing 0.1 mM DTPA. Appropriate amounts of the nitroreductase enzyme were introduced through an injection port, and the reaction was initiated by addition of NADPH to yield a final concentration of 1.0 mM. The rate of oxygen consumption was taken as the initial slope. When catalase and superoxide dismutase were included, they were added before the nitroreductase enzyme. Measurements of K, and Vmaxwere also made on the ferredoxin:NADP+ oxidoreductase-nitrofurantoin-Desferal, in Trizma-7.7 buffer without a chelating agent and in Trizma-7.7 DTPA buffer. Experiments in each buffer system were made both in the absence and presence of bovine serum albumin (as a nonspecific protein), and in the presence of a NADPHgeneratingsystem as described by Mason and Holtzman (8). All experiments were run at least three times each in triplicate. The enzyme kinetic parameters K, and Vmaxwere calculated utilizing the Lineweaver-Burklinearization (double-reciprocal plot) of the Michaelis-Menten equation and the direct linear plot method of Eisenthal and Cornish-Bowden (23-26).

All oxygen consumption rates were corrected for basal rate in the absence of the nitroaromatic compound. V,, values for the ferredoxin:NADP+ oxidoreductase-nitro substrate systems are reported in micromol of O2 consumed/mg enzyme/ min, while those for NADPH-cytochrome P-450 reductasenitro substrate systems are reported in micromol of O2 consumed/unit of enzyme/min. K,,, values, in both cases, are reported in millimolar. A typical double-reciprocal plot for the nitrofurantoin-ferredoxin:NADP' oxidoreductase system for both catalase/superoxide dismutase-inhibited and uninhibited systems is shown in Fig. 2a, corresponding toan apparent case of mixed competitive-noncompetitive inhibition (27, 28). Fig. 2b illustrates that same data plotted in V,.,K,,, space of the direct linear plot method. Although the latter method is considered statistically more acceptable than the double-reciprocal plot method (29-31), the kinetic parameters calculated by both methods differed from each other by no more than 1-2%. The kinetic parameters calculated for ferredoxin:NADP+ oxidoreductase with the five chosen substrates are listed in Table 11; those for NADPH-cytochrome P-450 reductase with the five substrates are listed in Table111. Addition of a different chelating agent,Desferal, complete absence of any chelating agent, both with and without catalase and superoxide dismutase, and the presence of nonspecific protein in the form of bovine serum albumin, did not have a significant effect on the magnitude of K,,, (except in the case of the Tris/DTPA/superoxidedismutasesystem). On the other hand effects on the values of Vmaxcan be discerned (Table IV). The effects of superoxide dismutase or catalase on the V,, seem to correspond to an apparent noncompetitive or mixedmodel. Nitrofurantoin (0.50-2.00 mM) solutions containing 0.1 mM Desferal in place of DTPA, with and without 100 pg/ml superoxide dismutase and 1440 units cat-

12381

Correlation of Kinetic Parameters withRedox Properties

TABLE I11 Kinetic parametersfor NADPH-cytochrome P-450 reductase with the five nitro compounds chosenfor this study No catalase or superoxide dismutase was added in this study. Compound

v,.:

K,

x 10-4

mM

log V,.lK,

-1.27 0.10 f 0.01 53.4 k 0.40 Nitrofurazone -2.00 0.24 2 0.05 25.3 f 2.40 Nitrofurantoin -3.14 1.31 f 0.25 9.42 2 1.92 Benznidazole -4.33 3.87 rf: 1.91 1.82 f 0.57 p-Nitrobenzoate -4.91 78.5 f 44.8 9.59 f 3.88 Metronidazole a pmol 02/unit of enzyme/min f S.E. based on cytochrome c assay units. 0

1.5

0.5

TABLEIV Effect of varying buffers and additionof bovine serum albumin, catalase, and superoxide dismutase onthe kinetic parameters of ferredoxin:NADP+ oxidoreductase with nitrofurantoin (0.05-2.00 m M ) solutions The abbreviations used are: Tris, Trizma -7.7 buffer; DES, deferoxamine (Desferal) mesylate; BSA, bovine serumalbumin;CAT, catalase; SOD, superoxide dismutase.

2.5

l/[NFT] (rnM)-1

Vma:

Tris DTPA DES CAT BSA

K,

mM 57.13 f 2.30 1.36 2 0.14 50.18 f 1.72 1.25 f 0.04 53.90 f 4.68 1.29 2 0.06

31.11 f 2.68 31.69 2 0.66 36.36 f 5.78 36.98 k 2.25 2.27 24.04 k 1.35 48.25 k 0.68 -2 5

-1 5

-0 5

05

1.5

25

K, (mM)

FIG. 2. Lineweaver-Burk double-reciprocal plot (a)of l / v versus l/[NFT] and direct linear plot ( b ) (V,,,.,K, space) for nitrofurantoin (NF!l')-ferredoxin: NADP+ oxidoreductase in Trizma, pH = 7.7,DTPA buffer for both inhibited (*) and uninhibited (0)systems. Inhibition was brought about by inclusion of catalase and superoxide dismutase in the incubates.

TABLE I1 Kinetic parameters for ferredoxin:NADP+ oxidoreductase with the five nitro compounds chosen forthis study V,.X"

Compound

K, +b

_c

log VroaxlKm +b

-e

mM

Nitrofura31.2 k 3.7 0.671.67 f 0.09 zone Nitrofu29.0k 1.1 49.0k 2.5 1.3k0.2 1.35 1.58 rantoin Benzni4.27f1.05 4.1420.95 7.43f2.17 -0.24 0.25 dazole p-Nitroben1.7f0.2 2.4k0.2 9.0f0.1 -0.57 -0.72 zoate Metroni3.30f 0.04 3.9 k0.8 139 f 30 -1.70 -1.52 dazole a pmol Oz/mg enzyme/min f S.E. * +, with catalase and superoxide dismutase. -, without catalase and superoxide dismutase.

alase/ml incubate, yielded the results summarized in Table IV. Fig. 3 illustrates the effects of addition of catalase and/or superoxide dismutase both before and after initiation of oxygen consumption by nitro anion radical. According to the scheme in Fig. 4, for every mole of superoxide anion radical

1.15 f 0.19 1.09 f 0.21 0.92 k 0.24 k 0.17 0.83 2 0.06 1.28 f 0.03

+ + + + +

-

+

-

+ + +

+ + + +

-

-

SOD

"

+

"

-

"

-

-

-

+

+

-

"

-

-

+ + + + -

+ + + + -

pmol 02/mg enzyme/min f S.E.

that forms water by the action of superoxide dismutase and catalase (Reactions 7 and 8), 3h mol of molecular 0, is regenerated:

0,+

H+ + V z 0 2

'hH202

-+

+ '/zH202

(7)

+ 'A02

(8)

%Hz0

Since superoxide is not stable even in the absence of superoxide dismutase, one cannot expect the recovery of oxygen to be entirely stoichiometric (Fig. 3b). In any case, catalase itself should decrease the net rate of oxygen consumption by half as is found (Table IV). From Table IV, it can be seen that the use of Desferal as a chelatorhas no effect on the values of the V,,, and K,,, in the absence of catalase and superoxide dismutase. However, there is a significant difference in the Vmax and a slight reduction in the value of the K,,, when catalase and superoxide dismutase are present. Since Desferal is able to bind transition metal ions in such a way that theircatalytic activity is inhibited, it has been used extensively as a metal ion scavenger (32). Morehouse and Mason (33) have also shown that the formation of hydroxyl-free radical during the reduction of molecular oxygen in the ferredoxin:NADP' oxidoreductase system is transition metal-mediated (Fig. 4, Reaction 9). The decreased v,,, for the system under study in this work suggests that the catalase/superoxide dismutase inhibition of oxygen consumption may in turn be inhibited by competitive metalion-mediated mechanisms notyet clearly understood. If the hydrogen peroxide formed by the two-electron reduction of oxygen (Reaction 5) or by disproportionation of superoxide radical (Reaction 7 ) can undergo a one-electron reduction to the .OH radical via a transition metal-mediated mechanism (33, 34), then suppression of this

Correlation of Kinetic Parameters

12382

with Redox Properties

a.

CAT+SOD

SOD CAT

-05 "

9

2

0

4

6

Time (min.)

-0.3

-0.4

-0.2

E: (V) FIG. 5. Log (V,JK,,,) versus E7(mV) for ferredoxin:NADP+ oxidoreductase interaction with five nitrocompounds as substrates in theabsence (0) and presence (0)of catalase/super-

oxide dismutase. FNR

CAT SOD

NADPH

-1

b.

-2 n

E

sx E

>

CAT+SOD

v

CAT SOD

-w

-3

0

-4

k

y = 2.29+15.0~ 2

1

0

3

4

-5

Time (min.)

FIG. 3. The oxygen consumption curves for the NADPHnitrofurantoin-ferredoxin:NADP* oxidoreductase system (FNR)showing the effects of superoxide dismutase (SOD) and/or catalase (CAT).1.0 mM nitrofurantoin in Trizma, pH 7.7, 0.1 mM DTPA buffer, pH 7.4, 37"C, was placed in the incubation chamber, and the other components were added at the points indicated. Final concentrations were 0.0033 pg/ml ferredoxin:NADP+ oxidoreductase, 1.0 mM NADPH, 5000 units/ml catalase, and 100 pg/ ml superoxide dismutase (500 units/ml). a, catalase and superoxide dismutase added prior to NADPH, b, catalase and superoxide dismutase added 1 min after NADPH.

4

+

le

141

0:-

,

FIG. 6. Log (VmJKm) uersus C, (mV) for NADPH-cytochrome P-460reductaseinteractionwithfive nitrocompounds as substrates.

returning one-half mol of 02 to the reaction system. In fact in the presence of Desferal, catalase/superoxide dismutase decreases V,, by one-half (Table IV). Figs. 5 and 6, which show plots of log V,,,/K, versus E : (V), were constructed from the data in Tables 1-111. Log V,,J K,,, gave excellentcorrelationswith E: (Figs. 5 and 6). In contrast, no correlation of log V,,. VmaX,log K,, or K, with E: could be found. DISCUSSION

FIG. 4. Proposed pathways for the reduction of oxygen and disposition of the reduction products.

pathway by binding of the transition metal by a powerful chelator such as Desferal should promote the disproportionation reaction of H202 to H 2 0 and regenerated 02,thus

The steady-state kineticbehavior of enzymes as a function of the concentration of one substrate generally obeys the Michaelis-Menten equation in which the two kinetically independent constants are V, and Vmax/Km.Both of these constants are complex functions consisting of several rate constants. Northrop (35) has shown that if one assumes the following very simplified reaction mechanism consisting of a minimal number of components,

Correlation of Kinetic Parameters with

Redox Properties

12383

concentration of lo-' M in atypicalaerobic incubate, we obtain a Rate 3/Rate 11 ratio of approximately 2000/1. Similar calculations for the range of compounds of this study using estimated second-order rate constants for the disproportionation (Reaction 11)of the orderof 104-105M" s" (37) yields Rate 3/Rate 11 ratios of 5 x 105/1 and 1.4 X 106/1 for nitrofurantoinand metronidazole,respectively. Thus,the oxygen present in aerobic systems competes successfully for the steady-state concentration of RNO; generated by reduced flavoenzymes. Cytotoxicity enhancement for the 2-nitroimidazole derivative,benznidazole, duetoheatstimulation(38)has been reportedto follow Michaelis-Menten kinetics. Numerous studies have suggested that cytotoxicity and rate of reduction of nitroaryl compounds and nitroheterocycles have a similar redox dependence ( i e . rate of change of rate constant with respect to change in the one-electron reduction potential) (39-43). Wardman and Clarke (44, 45) have shown on the basis of Marcuselectrontransfertheory (46, 47) that an electron transferreaction was compatiblewith the magnitude of the observed redox dependencefor nitroaromaticcompounds, thus implying that the formation of the nitro anion radical is the obligate key event for such biologically critical processes asradiosensitization,anti-microbialactivityand host toxicity. We have found not only that the nitroaromatics and nitroheterocyclesselectedovera wide range of E:(RNO,/ RNO;) followed Michaelis-Menten kinetics but that the log correlated linearly with the measured E: values as 2 RNO; + 2H+ -+ RNOZ + RNO + H 2 0 (11) ( Vmax/Km) shown in Figs. 5 and 6. The slopes of these lines (Table V) One might reasonably expect that thetwo major decay path- are analogous to the d(1og K ) / d ( A E )parameter derived by ways for RNO; are Reactions3 and 11: Wardman and Clarke (39) from the theoretical treatment of Marcus and canbe taken as a measure of the redox dependRate 3 = k3 [RNO;][02] (12) ence correlating biological response with E:. Our values for (13) redox dependence (log V,,,/K, uersus E:) fall well within the Rate 11 = kll [RNO;]' range reported (39,48,49)for other reduction-mediatedprocTaking p-nitroacetophenone as a mid-range example with a esses (Table V). measured k3 of 1.4 X lo6 M" s" and a kll of 1.1X lo7 M" s-' Clarke et al. (41) studied the anaerobic reduction of 0.1 mM ( l l ) , and using the measured [02] from this study, 2 X nitroimidazoles by xanthine-xanthine oxidase and demonM, we have strated a linear dependence of the logarithm of the rate of Rate 3 = (1.4 X lo6) (2 X [RNO;] M" s? (14) anaerobic nitro reduction upon the one-electron potential, E:, for 11 neutral nitroimidazoles. Initially, this result seems and to be contradictory to those reported in this paper.How can Rate 11 = (1.1x 10') [RN0;I2 M" s" (15) the logs of the rate of reaction at a single, arbitrary substrate concentration and V,,,/K, both correlate with E:? The anThusRate3/Rate 11 = 2.5 X [RNO;]". Under aerobic swer lies in the natureof Michaelis-Menten kineticswhere if conditions, RNO; is not detectable by ESR. Since the detec[RN02]