nitrobenzimidazoles as substrates for dt-diaphorase and redox cycling compounds: their enzymatic...

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 346, No. 2, October 15, pp. 219–229, 1997Article No. BB970285

Nitrobenzimidazoles as Substrates for DT-Diaphoraseand Redox Cycling Compounds: Their EnzymaticReactions and Cytotoxicity

Jonas Sarlauskas,* Egleg

Dickancaiteg

,* Ausra Nemeikaiteg

,† Zilvinas Anusevicius,*Henrikas Nivinskas,* Juan Segura-Aguilar,‡ and Narimantas Ce

g

nas*,1

*Institute of Biochemistry and †Institute of Immunology, Mokslininku 12, Vilnius 2600, Lithuania;and ‡Department of Pharmaceutical Bioscience, BMC, Uppsala S-751 23, Sweden

Received April 4, 1997, and in revised form July 7, 1997

diaphorase (260 U/mg protein) is partly prevented bydicumarol. That points out to partial determination ofWe have synthesized a number of nitrobenzimidaz-nitrobenzimidazole cytotoxicity by their reduction byoles containing nitro groups in the benzene ring andDT-diaphorase. Another important factor of nitro-found that they acted as relatively efficient substratesbenzimidazole toxicity to this cell line was oxidativefor rat liver DT-diaphorase (EC 1.6.99.2), their reactiv-stress, catalyzed by single-electron transfering en-ity exceeding reactivities of nitrofurans and nitroben-zymes. q 1997 Academic Presszenes. Nitrobenzimidazoles were competitive with

Key Words: nitrobenzimidazoles; DT-diaphorase;NADPH inhibitors of DT-diaphorase in menadione re-cytotoxicity; redox cycling; flavocytochrome b2;ductase reactions, their inhibition constant being un-ferredoxin:NADP/ reductase.changed in the presence of dicumarol and being in-

creased in the presence of 2 *,5*-ADP. These data indi-cate that the poor reactivity of nitrobenzimidazolesand other nitroaromatics in comparison to quinones

Nitroaromatic compounds such as nitrobenzenes, ni-could be determined by their binding in the adenosine-trofurans, and nitroimidazoles are widely used as phar-phosphate binding region of the NADPH-binding site,maceuticals, food additives, and explosives (1–4) andwhereas quinones bind at the nicotinamide-binding

pocket at the vicinity of FAD of DT-diaphorase. The comprise an important group of environment pollut-reduction of 4,5,6-trinitrobenzimidazol-2-one by DT-di- ants (5). For manifestation of their therapeutic and/oraphorase most probably involves reduction of 5-nitro cytotoxic properties, most nitroaromatics should un-group to 5-nitroso or 5-hydroxylamine derivative at dergo single- or two-electron enzymatic reduction inthe initial step. A certain parallelism existed between organism. Single-electron reduction of nitroaromaticsreactivities of nitrobenzimidazoles toward DT-diapho- is most frequently catalyzed by flavoenzyme dehydro-rase and their reactivities in single-electron reduc- genases–electrontransferases, e.g., NADPH:cytochrometion by Anabaena ferredoxin:NADP/ reductase (EC P-450 reductase (EC 1.6.2.4) (6–9), xanthine oxidase1.18.1.2) and Saccharomyces cerevisiae flavocyto- (EC 1.1.3.22) (10), ferredoxin:NADP/ reductase (ECchrome b2 (EC 1.1.2.3), the latter being determined by 1.18.1.2) (7), NADH:ubiquinone reductase (EC 1.6.99.3)electronic factors. However, we suppose that the rela- (11), bacterial oxygen-sensitive nitroreductases (12).tively high reactivity of polinitrobenzimidazoles to- Under aerobic conditions, single-electron reduction ofward DT-diaphorase was due not only to electronic nitroaromatics to their anion radicals results in theireffects, but also to a sterical crowding of nitrogroups

reoxidation by oxygen with formation of superoxide andby each other. The toxicity of nitrobenzimidazoles toother activated oxygen species that damage proteins,bovine leukemia virus-transformed lamb kidney fi-nucleic acids, and lipids. The cytotoxic hydroxylaminesbroblasts (line FLK) with a moderate amount of DT-might be formed as by-products of aerobic reduction,due to competition between reoxidation and dispropor-tionation of nitroradicals (6). Under hypoxic conditions,1 To whom correspondence should be addressed. Fax: 370 2 72 91

96. E-mail: [email protected]. single-electron transfering enzymes reduce nitroaro-

2190003-9861/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID ABB 0285 / 6b40$$$321 09-22-97 13:54:33 arcas

220 SARLAUSKAS ET AL.

matics to amines (12) or, less frequently, to hydroxyl- formed fibroblast culture. In addition, we have exam-ined single-electron enzymatic reduction reactions ofamines (13). Two-electron reduction of nitroaromatics to

nitroso compounds and, subsequently, to hydroxyl- these compounds.amines is catalyzed by bacterial oxygen-insensitive ni-troreductases (12, 14) and mammalian DT-diaphorase

MATERIALS AND METHODS(NAD(P)H:quinone reductase, EC 1.6.99.2) (15–17).Chemicals. The structural formulas of nitrobenzimidazoles usedDT-diaphorase is a cytosolic dimeric flavoprotein

in this work, are given in Fig. 1 together with the schemes of theircontaining one molecule of FAD per subunit (18). Ansynthesis. Compounds (1) and (10a) were obtained from 4-nitrophe-enzyme catalyzes the two-electron reduction of qui- nylene-1,2-diamine by published methods (34, 35). Compounds (2),

nones to hydroquinones using either NADH or NADPH (3), and (4) were synthesized by multistep nitration of compound (1),as electron donor and follows the ‘‘ping–pong’’ mecha- compound (9) (by multistep nitration of benzimidazole), and com-

pound (10) (by nitration of 2-trifluoromethyl-5(6)-nitrobenzimidazolenism at high turnover rates approaching 2000 s01 (19,(10a)), respectively, by published methods (36, 37), modified ac-20). Recently, a 0.21-nm-resolution crystal structure ofcording to the scheme in Fig. 1. 5-Amino, 5-aziridine, 5-hydroxyl-rat liver DT-diaphorase provided a rationale for the amine, and 5-nitroso derivatives of (3) (compounds (5), (6), (7), and

ping–pong mechanism, indicating that quinones oc- (8)) were obtained according to the adapted procedure (36). All com-cupy the nicotinamide-binding pocket of the NAD(P)H- pounds were characterized by melting point, thin-layer chromatogra-

phy, 1H NMR and IR spectroscopy (to be published elsewhere).binding site in the vicinity of the isoalloxazine ring ofNADPH, 2 *,5*-ADP, cytochrome c, superoxide dismutase, dicu-FAD (21). The possible physiological functions of DT-

marol, glucose 6-phosphate, L-lactate, N,N*-diphenyl-p-phenylene di-diaphorase are participation in vitamin K cycle (22) amine (DPPD),2 desferrioxamine, menadione (Sigma), and glucose-and maintenance of the reduced antioxidant form of 6-phosphate dehydrogenase (Fermentas, Vilnius) were used as re-

ceived. Other aromatic nitrocompounds were obtained or synthesizedcoenzyme Q in membrane systems (23). DT-diaphoraseas described in our previous papers (9, 38).protects cells from the damaging action of certain qui-

Cytotoxicity studies. The culture of bovine leukemia virus-trans-nones, e.g., menadione, diverting them from single-formed lamb kidney fibroblasts (line FLK) was grown and main-electron reduction and reducing them to relatively sta-tained in Eagle’s medium supplemented with 10% fetal bovine serumble hydroquinones that can further undergo conjuga- at 377C as described previously (38, 39). In cytotoxicity experiments,

tion with glucuronic acid (24). The enzyme is inhibited cells (3 1 104/ml) were grown in the presence of various amounts ofby the anticoagulants dicumarol, warfarin (25), hydro- nitrocompounds for 24 h, trypsinized, and counted using a hematocy-

tometer with viability determined by exclusion of Trypan blue. Unat-xyanthraquinones (26, 27), and flavones (28), which aretached dead cells were removed prior to trypsinization. The activitycompetitive with NAD(P)H inhibitors in the quinoneof DT-diaphorase in cell homogenates was determined according toreductase reaction. a rate of dicumarol-sensitive reduction of dichlorophenolindophenol-

The activity of DT-diaphorase is strongly increased indophenol by NADPH as described previously (39) and coincidedwith a previously determined value, 260 { 20 U/mg protein, wherein certain tumor and virus-transformed cells (15, 29),1 unit corresponded to a number of nanomoles of dichlorophenolindo-which makes this enzyme an attractive target for biore-phenol reduced/min. The activities of NADPH:cytochrome c reduc-ductive activation of the anticancer quinones EO9 (3-tase and NADH:cytochrome c reductase were determined as de-

hydroxymethyl-5-aziridinyl-11-methyl-2(H-indole-4,7- scribed previously (39), and were equal to 43 { 1 and 141 { 8 U/indone)-propenol) or diaziquone (2,5-diaziridinyl-3,6- mg protein, respectively, where 1 unit corresponded to a number of

nanomoles of cytochrome c reduced/min. The lipid peroxidation dur-bis(carboethoxyamino)-1,4-benzoquinone) (15, 30, 31).ing 24 h of incubation of cells with nitrobenzimidazoles was moni-However, despite several examples of DT-diaphorase’stored according to a formation of malondialdehyde, using the thiobar-role in reductive activation of nitroaromatics (cytotoxic- bituric acid test (40).

ity of CB-1954 (5-(aziridin-1-yl)-2,4-dinitrobenzamide)Enzymatic assays. Rat liver DT-diaphorase (EC 1.6.99.2) wasand its analogs in Walker cells (15, 16), genetoxicity prepared as described previously (41). Enzyme concentration was

of dinitropyrenes (17) these reactions appear to be of determined spectrophotometrically using e460Å 11 mM01 cm01. Flavo-cytochrome b2 (L-lactate:cytochrome c reductase, EC 1.1.2.3) fromlimited significance due to low nitroreductase activitySaccharomyces cerevisiae expressed in Escherichia coli was purifiedof DT-diaphorase (15, 16). A possible approach for utili-as described (42). The concentration of reduced enzyme formed inzation of bioreductive potential of this enzyme is thethe presence of 10 mM L-lactate was determined using e423 Å 183

search for new classes of more easily reducible aromatic mM01 cm01. Ferredoxin:NADP/ reductase from Anabaena, preparednitrocompounds. as described previously (43), was a generous gift of Dr. M. Martinez-

Julvez and Professor C. Gomez-Moreno (Zaragoza University,Nitrobenzimidazoles represent a group of nitrocom-pounds that are characterized to an insignificant ex-tent, although some of them possess antitrychomonal

2 Abbreviations used: cL50, concentration of compound for 50% celland other types of antimicrobial activity (32, 33). Insurvival; E1

7 , single-electron reduction potential of compound at pHthis paper, we demonstrate that several polinitroben-7.0; DPPD, N,N*-diphenyl-p-phenylene diamine; kcat , steady-statezimidazoles are relatively efficient substrates for DT- catalytic constant of enzyme; kcat/Km, steady-state bimolecular rate

diaphorase and that this enzyme is partly responsible constant of enzymatic reaction; TLC, thin-layer chromatography; Cytb2, flavocytochrome b2, FNR, ferredoxin:NADP/ reductase.for their cytotoxicity to bovine leukemia virus-trans-

AID ABB 0285 / 6b40$$$321 09-22-97 13:54:33 arcas

221ENZYMATIC REACTIONS OF NITROBENZIMIDAZOLES

FIG. 1. The structure and synthesis scheme of the nitrobenzimidazoles investigated.

Spain). The enzyme concentration was determined using e459 Å 9.4 of flavocytochrome b2 was monitored according to the decrease ofO2 concentration using a Clark electrode in the presence of 10mM01 cm01.

All kinetic measurements were carried out in 0.1 M K-phosphate mM L-lactate, due to reoxidation of the nitroradical formed.NADPH:menadione reductase activity of DT-diaphorase was mon-buffer solution (pH 7.0) containing 1 mM EDTA at 257C. In experi-

ments with DT-diaphorase, Tween 20 (0.01%) and bovine serum itored according to the increase of absorbance of reduced cyto-chrome c (De550 Å 20 mM01 cm01, concentration of menadione, 10albumin (0.25 mg/ml) were used as activators (30). The rates of

nitroreductase activity of DT-diaphorase and ferredoxin:NADP/ mM, concentration of cytochrome c, 50 mM). At saturating NADPHconcentration (100 mM), activity of DT-diaphorase was equal toreductase were monitored following NADPH oxidation (De340 Å

6.2 mM01 cm01; concentration of NADPH, 15–100 mM) using a Hi- 2000 s01. The activities of flavocytochrome b2 (concentration of L-lactate, 10 mM) and ferredoxin:NADP/ reductase (concentrationtachi-557 spectrophotometer. The rate of nitroreductase reaction

AID ABB 0285 / 6b40$$$322 09-22-97 13:54:33 arcas


of NADPH, 200 mM), using 1 mM ferricyanide as acceptor (42, 43), TABLE Iwere 420 and 200 s01, respectively. Catalytic Constants (kcat), Bimolecular Rate ConstantsThe kinetic parameters of the nitroreductase reaction, the catalytic

(kcat/Km) of Reduction of Nitrobenzimidazoles byconstant (kcat) and the bimolecular rate constant (kcat/Km) correspondDT-diaphorase, and Competitive Inhibition Constants (Ki)to the reciprocal intercepts and slopes of the Lineweaver–Burk plots,

of Nitrobenzimidazoles vs NADPH in Menadionekcat expressing the number of NADPH oxidized or O2 reduced byReductase Reactions of Enzymeactive center of enzyme per 1 s. Typically, five or six concentrations

of nitrocompound were used for determination of kcat and kcat/Km,and experiments were performed in duplicate or triplicate. Compound

Thin-layer chromatography. During thin-layer chromatography No. kcat (s01) kcat/Km (M01 s01) Ki (mM)(TLC) analysis of compound (3) and its enzymatic reduction products,the reaction mixture in 0.1 M K-phosphate, pH 7.0, consisting of 200 2 0.7 { 0.1 5.0 { 0.8 1 103 30mM compound (3), 5 mM NADPH, 50 nM DT-diaphorase, 2 mM glucose 3 2.4 { 0.3 2.6 { 0.5 1 104 176-phosphate, and 30 mg/ml glucose-6-phosphate dehydrogenase (total 4 102 { 7a 5.0 { 1.0 1 106a 10volume, 2 ml), was extracted by three portions of a mixture of diethyl 5 1.1 { 0.2 8.0 { 1.5 1 102 n.d.ether and ethyl acetate (1:1, 0.2 ml each) at the end of the reaction 6 1.5 { 0.12 8.6 { 0.9 1 103 150(ca. 1.5 h). Organic fractions were combined, concentrated to 0.2 ml, 9 0.12 { 0.01 2.9 { 0.2 1 103 125and applied on silica gel plates (Silufol UV-254, Kavalier, Czech 10 1.5 { 0.2 6.0 { 1.0 1 103 n.d.Republic), and eluted by a mixture of methylene chloride, diethylether, and acetonitrile (2:2:1). Note. The numbers of compounds are from Fig. 1. The rate con-

stants are derived from NADPH oxidation rates, pH 7.0, 257C.a Determined according to absorbance changes of the compound atRESULTS

500 nm.We have found that nitrobenzimidazoles (2), (3), (5),

(6), (9), and (10) acted as substrates for DT-diaphorase,as revealed by oxidation of NADPH (10–50 mM) moni- of oxygen uptake accompanying reduction of nitrobenz-

imidazoles by NADPH and DT-diaphorase. In the pres-tored at 340 nm in the presence of DT-diaphorase andnitrobenzimidazole. After the completion of the reac- ence of 200 mM compounds (1, 5, and 9), the rates of O2

consumption did not differ from an intrinsic NADPHtion, the decrease in absorbance at 340 nm was slightlyhigher than expected for oxidation of a given amount oxidase activity of DT-diaphorase (0.05 s01). For other

compounds, the rates of O2 consumption were equal toof NADPH, e.g., by 22% (compound (3)) or by 8–12%(other compounds), thus pointing out parallell ab- 0.10–0.12 s01 (200 mM compounds (2, 3, 6, and 10)) and

0.3 s01 (200 mM compound (4)), being at least 10 timessorbance changes of nitrobenzimidazoles during aero-bic reduction by DT-diaphorase. Reaction rates did not lower than the rates of NADPH oxidation at the ex-

pense of nitrobenzimidazoles. This means that DT-di-depend on NADPH concentration (10–50 mM) and thuswere consistent with previously reported micromolar aphorase does not cause the redox cycling of nitrobenz-

imidazoles and that the observed rates of O2 consump-values of Km of NADPH in nitroreductase reactions ofDT-diaphorase (45). The kinetic parameters of reaction reflect much slower secondary processes.

Analogous to other nitroaromatic compounds (16,tions corrected for an intrinsic NADPH:oxidase activityof enzyme (0.05 s01) and for absorbance changes of ni- 38), nitrobenzimidazoles inhibited NADPH:menadione

reductase reaction of DT-diaphorase, acting as compet-trobenzimidazoles are presented in Table I. It is evi-dent that nitrobenzimidazoles possessing two or more itive inhibitors with respect to NADPH and uncompeti-

tive inhibitors to menadione (data not shown). The val-nitro groups in the benzene ring act as efficient sub-strates for DT-diaphorase, far exceeding the reactivity ues of competitive Ki of several compounds are pre-

sented in Table I. We have found that the Ki ofof nitrobenzenes and nitrofurans (38, 44). In contrast,5-nitrobenzimidazole (1) (300–400 mM) was found to be compound (3) was unchanged in the presence of 0.5 mM

dicumarol, a competitive to NADPH inhibitor; how-almost inactive. Compounds (7) and (8) rapidly decom-posed in aqueous media at pH 7.0; therefore, we did ever, it increased twice in the presence of 0.5 mM 2 *,5*-

ADP, another competitive to NADPH inhibitor (Ki Ånot investigate their enzymatic reactions. In general,aerobic reduction of nitrobenzimidazoles by DT-diapho- 0.4 mM) (Fig. 2).

Among the compounds investigated, 4,5,6-trinitrobenz-rase was accompanied by the increase in their ab-sorbance at 400–700 nm (see below). Since enzymatic imidazol-2-one (3) proved to be an efficient DT-diapho-

rase substrate (Table I). Since we were able to preparereduction of 3,4,5,6-tetranitrobenzimidazol-2-one (4)resulted in an almost unchanged absorbance level at compounds (5), (7), and (8) with different extents of

reduction of the 5-nitrogroup, we have tried to get some340 nm and absorbance increase at 440–700 nm (datanot shown), kinetic parameters of (4) (Table I) were insight into the mechanism of compound (3) reduction

by DT-diaphorase. In the presence of a NADPH regen-determined according to a product absorbance increase(De500 Å 2.6 mM01 cm01). eration system, we have observed an increase in ab-

sorbance of reaction mixture above 430 nm and a de-In parallel experiments, we have monitored the rates

AID ABB 0285 / 6b40$$$322 09-22-97 13:54:33 arcas


zol-2-one (3) enzymatic reduction (Fig. 3). Moreover,TLC analysis of the reaction mixture showed the samemajor product spot with RfÅ 0.25. This reaction did notproceed in the absence of DT-diaphorase, even when alarge excess NADPH (1 mM) was used. Thus, althoughthe product of enzymatic reduction of (3) was not yetidentified, one can conclude that reduction of (3) by DT-diaphorase proceeded with initial formation of unstable5-nitroso or 5-hydroxylamine intermediate. Evidently,the further reaction involves the reduction of anothernitro group.

For a more thorough insight into structure–activityrelationships in the enzymatic reduction of nitrobenz-imidazoles, and due to the importance of enzymaticsingle-electron reduction of nitroaromatics in their cy-totoxicity, we have investigated their reactions withtwo enzymes that reduce nitroaromatics in a single-electron way, ferredoxin:NADP/ reductase (7) and fla-vocytochrome b2 (9). Briefly, reactions of nitrobenzim-idazoles exhibited extensively described features of sin-

FIG. 2. Determination of the competitive to NADPH inhibition con- gle-electron reduction of other classes of nitroaromaticsstant of compound (3) in the absence (1) or in the presence of 0.5under aerobic conditions (7, 12), namely, a stoichiomet-mM 2 *,5*-ADP (2) or 0.5 mM dicumarol (3) during reduction of menadi-ric amount of O2 consumed per mole of NADPH oxi-one by DT-diaphorase. Concentration of menadione, 10 mM; concen-

tration of cytochrome c, 50 mM. The kcat/Km values for NADPH were dized by ferredoxin:NADP/ reductase, superoxide dis-calculated from slopes of Lineweaver–Burk plots, varying the con- mutase-sensitive reduction of added cytochrome c, andcentrations of NADPH (20–300 mM) in the absence or in the presence consumption of O2 amount exceeding the concentrationof inhibitors.

of nitrocompound. The second-order reaction rate con-stants of nitrobenzimidazoles are given in Table II, to-gether with the kcat/Km values of other nitroaromaticcrease of absorbance at lower wavelengths (Fig. 3). Thecompounds with available values of single-electron re-spectra obtained at the end of the reaction (ca. 1.5 h)duction potentials (E1

7).were stable overnight. After the completion of reaction,In order to determine the role of DT-diaphorase inTLC analysis indicated a major spot of product with Rf

cytotoxicity of nitrobenzimidazoles, we have used theÅ 0.25, which was sufficiently different from Rf Å 0.65bovine leukemia virus-transformed lamb fibroblast linefor compound (3), and minor spots with Rf Å 0.6 andFLK, which was exploited in our previous studies (38,Rf Å 0.1. The absorbance spectrum and TLC properties39). Analogous with other tumor- and virus-trans-of the main enzymatic reduction product of 4,5,6-trini-formed cell lines (15, 29) with elevated DT-diaphorasetrobenzimidazol-2-one (3) did not coincide with corre-activity, the FLK cell line possessed about 10-fold in-sponding parameters of its 5-amino derivative (5) (Rf

creased activity of DT-diaphorase compared to non-Å 0.6) and its 5-nitroso derivative (8) (Rf Å 0.4). Be-transformed cells (39). The concentrations of com-sides, the 5-nitroso derivative (8) rapidly decomposedpounds for 50% survival of the cells (cL50) after 24 h ofin aqueous media (pH 7.0, t1/2 Å 5–6 min), yielding aincubation were the following: 20 { 6 mM (4), 40 { 12single unidentified product (Rf Å 0.6) with an alteredmM (3), 50 { 17 mM (5), 90 { 35 mM (2), 120 { 35 mMabsorbance spectrum (data not shown). At pH 7.0, we(6), 190 { 60 mM (9), 220 { 60 mM (10), and 1250 { 440were unable to obtain a spectrum of the 5-hydroxyl-mM (1). The unstable compounds (7) and (8) were alsoamine derivative (7) (Rf Å 0.5), due to its rapid autoxi-toxic, with cL50 close to 100 mM. The data of Fig. 4dation to the 5-nitroso derivative (8) and parallel con-indicate that toxicity of 3,4,5,6-tetranitrobenzimidazol-version of the latter. However, we have found that the2-one (4) and its dinitro analogue (2) is decreased aboutproduct of conversion of the 5-nitroso derivative (8) intwice in the presence of dicumarol, a well-known inhibi-aqueous medium acted as a substrate for DT-diapho-tor of DT-diaphorase. The same effect was characteris-rase. After complete conversion of compound (8) (30–tic for toxicity of compound (3) (data not shown); how-40 min in 0.1 M phosphate, pH 7.0) and subsequentever, toxicity of 5-nitrobenzimidazol-2-one (1) that wasintroduction of DT-diaphorase, catalytic amounts ofinactive toward DT-diaphorase was not affected by di-NADPH, and the NADPH regeneration system, the ab-cumarol. Our preliminary experiments with nontrans-sorbance spectrum appearing after 1.5–2 h was analo-

gous to that of the product of 4,5,6-trinitrobenzimida- formed cells that possessed much lower amounts of DT-

AID ABB 0285 / 6b40$$$323 09-22-97 13:54:33 arcas


FIG. 3. The spectral changes of 100 mM 4,5,6-trinitrobenzimidazol-2-one (3) during its aerobic reduction by 50 nM DT-diaphorase, 5 mM

NADPH, 2 mM glucose 6-phosphate, and 30 mg/ml glucose-6-phosphate dehydrogenase. The spectra were recorded every 10 min.

diaphorase (23 U/mg protein) indicated that compound electron transfering enzymes, e.g., NADH:cytochromec reductase (141 U/mg protein) and NADPH:cyto-(3) was much less cytotoxic in this case (cL50Å 200{ 35

mM); besides, its toxicity was unaffected by dicumarol. chrome c reductase (43 U/mg protein), that could initi-ate redox cycling of nitrobenzimidazoles, we shouldSince FLK cells contain a marked amount of single-

TABLE II

Bimolecular Rate Constants (kcat/Km) of Reduction of Nitroaromatics by Ferredoxin: NADP/ Reductase (FNR)and Flavocytochrome b2 (Cyt b2) Together with Their Single-Electron Reduction Potentials (E1

7)

kcat/Km (M01 s01)

Compound FNR Cyt b2 E17 (V)

Chinifur 6.0 { 0.5 1 104 1.54 { 0.11 1 104 00.20a

Nitrofurantoin 1.5 { 0.1 1 104 5.9 { 0.3 1 103 00.26b

Nifuroxime 7.5 { 1.7 1 103 3.7 { 0.2 1 103 00.26b

Nifurtimox 1.38 { 0.15 1 104 2.3 { 0.3 1 103 00.29a

p-Nitrobenzaldehyde 8.1 { 0.5 1 103 7.3 { 0.5 1 102 00.32b

p-Nitroacetophenone 1.0 { 0.15 1 103 2.5 { 0.18 1 102 00.35b

1 6.7 { 0.4 1 102 2.0 { 0.3 1 102 —2 1.9 { 0.2 1 104 3.2 { 0.4 1 103 —3 1.0 { 0.15 1 105 1.5 { 0.15 1 104 —4 3.1 { 0.4 1 105 3.2 { 0.3 1 104 —5 2.7 { 0.4 1 104 5.3 { 0.5 1 103 —6 2.8 { 0.4 1 103 2.1 { 0.25 1 103 —9 7.3 { 0.5 1 103 1.5 { 0.3 1 103 —10 8.0 { 1.0 1 103 2.8 { 0.5 1 103 —

Note. The numbers of nitrobenzimidazoles are from Fig. 1.a The values of reduction potentials were calculated in our previous work (9).b From Ref. (45).

AID ABB 0285 / 6b40$$$323 09-22-97 13:54:33 arcas


pound (4), nitrobenzimidazoles still remain ca. 103

times less reactive in comparison with quinone com-pounds (19, 20). It is of interest to note that the rateof reduction of nitrocompounds is so low that it doesnot depend on the NADPH concentration, even undernonsaturated conditions. The data presented providesome explanation for the modest DT-diaphorase nitro-reductase activity:

(i) The competitive inhibition of nitrobenzimidazolesto NADPH in menadione reductase reaction of DT-di-aphorase (Table I) indicates that these compounds bindat the NADPH-binding site of the enzyme. This typeof inhibition is characteristic of other classes of nitroar-omatics as well (16, 38, 44). The unchanged Ki of com-pound (3) in the presence of dicumarol (Fig. 2) indicatesthat (3) does not compete for nicotinamide-bindingpocket of NADPH-binding site of DT-diaphorase, a siteof binding of dicumarol (27). Contrary to this, an in-FIG. 4. The viability of FLK cells after 24 h of incubation with

various concentrations of compound (4) (1, 2) and compound (2) (3, crease in the Ki of compound (3) in the presence of 2 *,5*-4), in the absence (1, 3) and in the presence (2, 4) of 20 mM dicumarol ADP indicates that binding sites of 2 *,5*-ADP and (3)(n Å 3 or 4). overlap. According to X-ray analysis data (21), another

competitive to NADPH inhibitor, Cibacron blue, doesnot occupy nicotinamide-binding pocket. Besides, dicu-

take into account a possibility of oxidative stress as the marol and other competitive to NADPH inhibitors, fla-additional factor of nitrobenzimidazole toxicity. Evi- vones, bind at separate sites in the active center ofdently, this factor is also partly responsible for toxicity DT-diaphorase (28). Therefore, compound (3) and otherof nitrobenzimidazoles, since their action is partly pre- nitrobenzimidazoles may share the properties of thesevented by the antioxidant DPPD (46) and the iron-che- inhibitors and interact with the adenosine phosphate-lating agent desferrioxamine, the latter preventing the binding region of NADPH-binding site. It is interestingFenton reaction (Fig. 5). Other evidence of oxidativestress was an increase in the intracellular content ofthe lipid peroxidation product malondialdehyde. After24 h of incubation of cells with 70 mM compound (3) or200 mM compound (2) that resulted in 80–85% celldeath, the content of malondialdehyde was equal to 2.5{ 0.3 nmol/106 cells, whereas in untreated cells it wasequal to 0.7 { 0.2 nmol/106 cells.

DISCUSSION

The data presented indicate that nitrobenzimidaz-oles containing two or more nitro groups in the benzenering act as relatively efficient substrates for rat DT-diaphorase, their reactivity (Table I) markedly ex-ceeding the reactivity of CB-1954, other nitrobenzenes,and nitrofurans (16, 38, 44). In spite of the biomedicalsignificance of the metabolism of nitroheterocycles byhuman DT-diaphorase, nitrobenzimidazoles could de-serve attention as potential substrates for human en-zyme. Even after decrease by a factor of 7 (a ratio of

FIG. 5. The protective effect of desferrioxamine (300 mM) and DPPDnitroreductase activities of rat and human DT-diapho-(2 mM) against the toxicity of compound (4) (40 mM) toward FLK cellsrase (16, 47)), the reactivity of compounds (2–6, 10)during 24 h of incubation. Additions: desferrioxamine (1), DPPD (2),(Table I) will remain sufficiently high and will exceed desferrioxamine / DPPD (3), compound (4) (4), compound (4) / des-

the kcat of reduction of CB-1954 by human enzyme, ca. ferrioxamine (5), compound (4) / DPPD (6), compound (4) / desferri-oxamine / DPPD (7), n Å 3 to 4.0.01 s01 (16, 47). However, with the exception of com-

AID ABB 0285 / 6b40$$$323 09-22-97 13:54:33 arcas


to note that carboxy-terminal portion of enzyme inter-acting with adenosine phosphate of NADPH (21) hasbeen proposed to be responsible for binding of flavonesand the regulation of nitroreductase activity of DT-di-aphorase (47). It seems that this mode of binding ischaracteristic of other types of nitroaromatics as well,since analogous effects of dicumarol and 2 *,5*-ADP(Fig. 2) have been observed for inhibition of DT-diapho-rase by vinylquinoline-substituted nitrofurans (48).Thus, the poor reactivity of nitroaromatics toward DT-diaphorase could be related to their less favorable bind-ing in comparison to quinones, the latter occupying nic-otinamide-binding pocket at 0.34 nm distance from iso-alloxazine ring of FAD (21). The reactivity of severalnitrobenzimidazoles studied toward DT-diaphorase in-creased upon the decrease of their competitive Ki toNADPH in the menadione reductase reaction (TableI). However, a comparison of inhibition constants andreduction rates of nitrobenzimidazoles and other typesof nitrocompounds (nitrobenzenes and nitrofurans) (38,44) indicates that the correlation between affinity andthe reduction rate is not a general feature of nitroaro-matics reduction.

FIG. 6. The relationship between reactivities of nitrobenzimidaz-(ii) An important factor determining reactivity of ni-oles and other nitroaromatics in their single-electron reduction by

trocompounds in enzymatic reduction is the value of ferredoxin:NADP/ reductase (kcat/Km (FNR)) and flavocytochrome b2

their single-electron reduction potential (E17). For reac- (kcat/Km (Cyt b2). The numbers of nitrobenzimidazoles are from Fig.

1: compound (4) (1), compound (3) (2), compound (5) (3), compoundtions of DT-diaphorase, relationship between reactivity(2) (4), compound (10) (5), compound (9) (6), compound (6) (7), andand redox potential of quinones or nitrocompounds hascompound (1) (8).not been observed (20, 38). However, reactivity of bacte-

rial oxygen-insensitive nitro-reductase increased uponincrease in E1

7 values of nitrocompounds (14). To ourknowledge, the directly determined E1

7 values of nitro- (Fig. 6). These values seem to be reasonable in view ofbenzimidazoles studied in this work are not available. available values of E1

7 of 1,2-dinitro- and 1,3-dinitroben-On the other hand, in single-electron reduction reac- zenes, 00.287 and 00.345 V, respectively (45). For ations by NADPH:cytochrome P-450 reductase (7–9), quantitative presentation, logs of kcat/Km and kcat of pol-ferredoxin:NADP/ reductase (7), flavocytochrome b2 initrobenzimidazoles in reactions with DT-diaphorase(9), and adrenodoxin (9), logs of kcat/Km of nitroaromat- were plotted against logs of the geometrical mean ofics exhibited linear dependences on their E1

7 values. their reactivities toward ferredoxin:NADP/ reductaseMoreover, an orthogonality existed between reactivi- and flavocytochrome b2, 0.5 log kcat/Km (FNR) / 0.5 logties of aromatic nitrocompounds in these systems (9), kcat/Km (Cyt b2) (Fig. 7). It is evident that reactivity ofindicating that reaction rates were governed by ener- DT-diaphorase tends to increase upon the increase ofgetics of single-electron transfer and not by substrate nitrobenzimidazole reactivities in single-electron enzy-structure specificity. The data in Fig. 6 indicate that matic reactions; however, these dependences are poorlyanalogous orthogonality exists between the reactivities expressed. Besides, this dependence is not characteris-of nitrobenzimidazoles and other nitrocompounds in tic of other types of nitrocompounds, e.g., nitrofurans,their single-electron reduction by ferredoxin:NADP/ that exhibited comparable activity towards ferredoxin:reductase and flavocytochrome b2. In our opinion, the NADP/ reductase and flavocytochrome b2 (Table II,comparison of reactivities of nitrobenzimidazoles and Fig. 6), and were almost inactive toward DT-diapho-reactivities of nitrocompounds with available values of rase (38). Thus, the increase in reactivity of nitrobenz-E1

7 enables us to define the orientational range of E17 of imidazoles with the increase in a number of ni-

nitrobenzimidazoles (Fig. 6). Thus, single-electron re- trogroups might be not entirely determined by theirduction potential of compounds (3) and (4) should be electron-accepting effects. One of the possible explana-close or above 00.2 V, the E1

7 of compound (1) should tions of the relatively high reactivity of compounds (2–be below 00.35 V, and the redox potentials of other 6, 9, and 10) (Fig. 1), in comparison to other nitroaro-

matics, could be the sterical crowding of nitro groupscompounds should be in the range of 00.26–00.32 V

AID ABB 0285 / 6b40$$$324 09-22-97 13:54:33 arcas


hand, one should take into account another importantfactor of cytotoxicity, oxidative stress (Fig. 5), sinceFLK cells contain a definite amount of single-electrontransferring enzymes (NADH:cytochrome c reductaseand NADPH:cytochrome c reductase) that could initi-ate redox cycling of nitrocompounds. An existence oflinear relationships between log cL50 and E1

7 of nitroaro-matics has been reported in several cases (38, 50–52),which pointed to redox cycling as the main factor oftheir cytotoxicity. Analogous to these data, the de-crease of log cL50 of nitrobenzimidazoles upon increaseof their reactivity in single-electron enzymatic reac-tions is observed (Fig. 8). We have failed to observe arelationship between the cytotoxicity of nitrobenzimid-azoles and the rates of O2 consumption accompanyingthe reduction of nitrobenzimidazoles by DT-diapho-rase. Since a certain parallelism exists between reac-tivity of nitrobenzimidazoles toward DT-diaphoraseand single-electron transferring enzymes (Fig. 7), thedata in Fig. 8 do not necessarily imply that redox cy-cling of nitrobenzimidazoles is the main factor of their

FIG. 7. The dependence of kcat/Km (A) and kcat (B) of nitrobenzimid- cytotoxicity. Most probably, we observe the superposi-azoles in their reactions with DT-diaphorase upon their reactivity

tion of two sufficiently similar structure vs cytotoxicityin single-electron enzymatic reactions: a geometrical mean of theirrelationships. These data point to the existence of cer-kcat/Km in reactions with ferredoxin:NADP/ reductase (FNR) and

flavocytochrome b2 (Cyt b2). The numbers of nitrobenzimidazoles are tain limits in biomedical exploitation of specificity offrom Fig. 1: compound (4) (1), compound (3) (2), compound (5) (3), nitrobenzimidazoles toward DT-diaphorase, i.e., thatcompound (2) (4), compound (10) (5), compound (9) (6), and compound one should take into account that the best nitrobenzim-(6) (7).

by each other, causing their deconjugation with thebenzene ring (4, 36). Although deconjugation of the ni-trogroup with the aromatic system should decrease thesingle-electron reduction potential of the nitrocom-pound (49), other factors, e.g., formation of less bulkynitroso or hydroxylamine groups after nitro group re-duction, might drive the reaction forward. In a mole-cule of 4,5,6-trinitrobenzimidazol-2-one (3), the 5-nitrogroup most easily participates in nucleophilic substitu-tion reactions (4, 36). Our studies on this compound(Fig. 3) point to the possibility that 5-nitro group isalso the most readily reducible by DT-diaphorase.Thus, various classes of nitroaromatics bearing two orthree nitrogroups in o-position to each other could prob-ably act as relatively efficient substrates for DT-diaph-orase.

The data of this work indicate that toxicity of nitro-benzimidazoles toward the FLK cell line, which con-tains a moderate amount of DT-diaphorase, 260 U/mgprotein (39), is partly determined by their bioactivation FIG. 8. The dependence of cL50 of nitrobenzimidazoles toward FLK

cells upon their reactivity in single-electron enzymatic reactions, aby DT-diaphorase (Fig. 4). In general, the toxicity ofgeometrical mean of their kcat/Km in reactions with ferredoxin:NADP/nitrobenzimidazoles tends to increase upon increase inreductase (FNR) and flavocytochrome b2 (Cyt b2). The numbers oftheir reactivity toward DT-diaphorase, although the nitrobenzimidazoles are from Fig. 1: compound (4) (1), compound (3)

correlations between logs cL50 and logs kcat or kcat/Km (2), compound (5) (3), compound (2) (4), compound (10) (5), compound(9) (6), compound (6) (7), and compound (1) (8).are poorly expressed (data not shown). On the other

AID ABB 0285 / 6b40$$$324 09-22-97 13:54:33 arcas


Archakov, A. I., and Ollinger, K. (1994) Arch. Biochem. Biophys.idazole substrates for DT-diaphorase could also act as315, 400–406.best redox cycling agents in reactions with single-elec-

10. O’Connor, C. J., McLennan, D. J., Sutton, E. M., Denny, W. A.,tron transferring enzymes. A more thorough investiga-and Wilson, W. R. (1991) J. Chem. Soc. Perkin Trans. 2, 951–tion of the role of DT-diaphorase in the toxicity of nitro- 954.

benzimidazoles to other cell lines enriched in DT-di- 11. Bironaiteg

, D. A., Ceg

nas, N. K., and Kulys, J. J. (1991) Biochim.aphorase to different extents is currently in progress. Biophys. Acta 1060, 203–209.

12. Peterson, F. J., Mason, R. P., Hovsepian, J., and Holtzman, J. L.(1979) J. Biol. Chem. 254, 4009–4014.CONCLUSIONS

13. Leung, K. H., Yao, M., Stearns, R., and Chiu, S.-K. L. (1995)In this paper, we have shown that certain polinitro- Chem.–Biol. Interact. 97, 37–51.

benzimidazoles acted as relatively efficient substrates 14. Bryant, C., and DeLuca, M. (1991) J. Biol. Chem. 266, 4119–4125.for rat DT-diaphorase, far exceeding the reactivity of

15. Riley, R. J., and Workman, P. (1992) Biochem. Pharmacol. 43,nitrofurans and nitrobenzenes. These compounds ap-1657–1669.pear to bind at the adenosine phosphate-binding region

16. Knox, R. J., Friedlos, F., and Boland, M. P. (1993) Cancer Metas-of the NADPH-binding site of DT-diaphorase. The re-tasis Rev. 12, 195–212.duction rates of nitrobenzimidazoles by DT-diaphorase

17. Hajos, A. K. D., and Winston, G. W. (1991) J. Biochem. Toxicol.can not be entirely predicted by the single-electron re- 6, 277–281.duction potential, but may instead be governed by ste- 18. Ernster, L. (1987) Chem. Scripta 27A, 1–13.ric factors. In bovine leukemia virus-transformed lamb 19. Hosoda, S., Nakamura, W., and Hayashi, K. (1974) J. Biol.kidney fibroblast cell line, DT-diaphorase may be in- Chem. 249, 6416–6423.volved in the metabolism of nitrobenzimidazoles, re- 20. Tedeschi, G., Chen, S., and Massey, V. (1995) J. Biol. Chem. 270,sulting in their cytotoxicity, as indicated by protective 1198–1204.effects of an inhibitor of DT-diaphorase, dicumarol. An- 21. Li, R., Bianchet, M. A., Talalay, P., and Amzel, L. M. (1995) Proc.

Natl. Acad. Sci. USA 92, 8846–8850.other important factor of nitrobenzimidazole cytotoxic-22. Wallin, R., Gebhardt, O., and Prydz, H. (1978) Biochem. J. 169,ity is their redox cycling catalyzed by single-electron

95–101.transferring enzymes, e.g., NAD(P)H:cytochrome c re-23. Beyer, R. E., Segura-Aguilar, J., DiBernardo, S., Cavazzoni, M.,ductases. This factor may limit the biomedical exploita-

Fato, R., Fiorentini, D., Galli, M. C., Setti, M., Landi, L., andtion of specificity of nitrobenzimidazoles toward DT- Lenaz, G. (1996) Proc. Natl. Acad. Sci. USA 93, 2528–2532.diaphorase. 24. Lind, C., Hochstein, P., and Ernster, L. (1982) Arch. Biochem.

Biophys. 216, 178–185.25. Hollander, P. M., and Ernster, L. (1975) Arch. Biochem. Biophys.ACKNOWLEDGMENTS

169, 560–567.We thank Professor C. Gomez-Moreno and Dr. M. Martinez-Julvez

26. Ma, Q., Cui, K., Xiao, F., Lu, A. Y. H., and Yang, P. S. (1992) J.(Zaragoza University, Spain) for their generous gift of ferredoxin:Biol. Chem. 267, 22298–22304.NADP/ reductase. The generous assistance of the laboratory of Dr.

27. Prestera, T., Prochaska, H. J., and Talalay, P. (1992) Biochemis-F. Lederer (CNRS Gif-sur-Yvette, France) in preparation of flavocy-try 31, 824–833.tochrome b2 is gratefully acknowledged. This work was supported in

28. Chen, S., Hwang, J., and Deng, P. S. K. (1993) Arch. Biochem.part by EC Grant IC15CT961004.Biophys. 302, 72–77.

29. Rushmore, T. H., Morton, M. R., and Pickett, C. B. (1991) J. Biol.REFERENCES Chem. 266, 11632–11639.

1. Grundberg, E., and Titsworth, E. H. (1973) Annu. Rev. Microbiol. 30. Fisher, G. R., Donis, J., and Gutierrez, P. L. (1992) Biochem.27, 317–346. Pharmacol. 44, 1625–1635.

2. Tazima, Y., Kata, T., and Murakami, M. (1975) Mutat. Res. 32, 31. Gustafson, D. L., Beall, H. D., Bolton, E. M., Ross, D., and Wal-55–80. dren, C. A. (1996) Molec. Pharmacol. 50, 728–735.

32. Grundberg, E., and Titsworth, E. (1965) Antimicrob. Agents Che-3. Kedderis, G. L., and Miwa, G. (1988) Drug. Metab. Rev. 19, 31–62. mother. 5, 478–480.

33. Alcalde, E. (1994) Adv. Heterocycl. Chem. 60, 197–259.4. Orlova, E. J. (1973) Chemistry and Technology of High Explo-sives. Khimiya, Leningrad. [In Russian] 34. Clark, R. L., and Pessolano, A. A. (1958) J. Am. Chem. Soc. 80,

1657–1664.5. Fukuhara, K., Takei, M., Kageyama, H., and Miyata, N. (1995)Chem. Res. Toxicol. 8, 47–54. 35. Samuel, E. L. (1972) Aust. J. Chem. 25, 2725–2729.

6. Holtzman, J. L., Crankshaw, D. L., Peterson, F. J., and Polnas- 36. Efros, L. S., and Jeltsov, A. V. (1957) J. Gen. Chem. USSR 27,zek, C. F. (1981) Mol. Pharmacol. 20, 669–673. 127–135. [In Russian]

7. Orna, M. V., and Mason, R. P. (1989) J. Biol. Chem. 264, 12379– 37. Buchel, K. H. (1970) Z. Naturforsch. 25b, 934–944.12384. 38. Ce

g

nas, N., Nemeikaiteg

, A., Dickancaiteg

, E., Anusevicius, Z., Nivi-8. Butler, J., and Hoey, B. M. (1993) Biochim. Biophys. Acta 1161, nskas, H., and Bironaite

g

, D. (1995) Biochim. Biophys. Acta 1268,73–78. 159–164.

39. Nemeikaiteg

, A., and Ceg

nas, N. (1993) FEBS Lett. 326, 65–68.9. Ceg

nas, N., Anusevicius, Z., Bironaiteg

, D., Bachmanova, G. I.,

AID ABB 0285 / 6b40$$$324 09-22-97 13:54:33 arcas


40. Ramanathan, R., Das, N. P., and Tan, C. H. (1994) Free Rad. P. S. K., Fung, M., Ebenstein, D., Wu, K., and Tsai, T.-M. (1995)Mol. Pharmacol. 47, 934–939.Biol. & Med. 16, 43–48.

41. Linderson, Y., Baez, S., and Segura-Aguilar, J. (1994) Biochim. 48. Anusevicius, Z. (1996) Interactions of Quinones and AromaticBiophys. Acta 1200, 197–204. Nitrocompounds with NAD(P)H-oxidizing Flavoenzymes. Ab-

stract of Ph.D. Thesis, Institute of Biochemistry, Vilnius.42. Pueyo, J. J., and Gomez-Moreno, C. (1991) Prep. Biochem. 21,191–204. 49. Boyd, M., Lee, H. H., Anderson, R. F., and Denny, W. A. (1994)

J. Chem. Soc. Perkin Trans. 2, 291–295.43. Black, M. T., White, S. A., Reid, G. A., and Chapman, S. K. (1989)Biochem. J. 258, 255–259. 50. Adams, G. E., Clarke, E. D., Jakobs, R. S., Stratford, I. J., Wal-

44. Knox, R. J., Boland, M. P., Friedlos, F., Coles, B., Southan, C., lace, R. G., Wardman, P., and Watts, M. E. (1976) Biochem. Bio-and Roberts, J. J. (1988) Biochem. Pharmacol. 37, 4671–4677. phys. Res. Commun. 72, 824–829.

45. Wardman, P. (1989) J. Phys. Chem. Ref. Data 18, 1637–1755. 51. Guissany, A., Henry, A., Lougmani, N., and Hickel, B. (1990)Free Rad. Biol. Med. 8, 173–189.46. Miccadie, S., Kyle, M. E., Gilford, D., and Farber, J. L. (1988)

Arch. Biochem. Biophys. 265, 311–320. 52. Siim, B. G., Atwell, G. J., and Wilson, W. R. (1994) Biochem.Pharmacol. 48, 1593–1604.47. Chen, S., Knox, R., Lewis, A. D., Friedlos, F., Workman, P., Deng,

AID ABB 0285 / 6b40$$$325 09-22-97 13:54:33 arcas

nitrobenzimidazoles as substrates for dt-diaphorase and redox cycling compounds: their enzymatic...

Documents