HRP-catalyzed bioactivation of carcinogenic hydroxamic acids. The greater reactivity of glycolyl-versus acetyl-derived hydroxamic acids

Chem -Bzol Interactzons, 63 (1987) 249-- 264 Elsevier Sclentdlc Pubhshers Ireland Ltd

249

HRP-CATALYZED BIOACTIVATION OF CARCINOGENIC HYDROXAMIC ACIDS. THE GREATER REACTIVITY OF GLYCOLYLVERSUS ACETYL-DERIVED HYDROXAMIC ACIDS

M I C H A E L D C O R B E T T and B E R N A D E T T E

R CORBETT

Pestw~de Research Laboratory, Department of Food Science and Human Nut~t~on, University of Flo~tda, GatneswUe, FL 3~611 (US A )

(Received February 16th, 1987) (Rewslon received June 25th, 1987) (Accepted June 25th, 1987)

SUMMARY

An analysis of the hydroxamlc acid oxidation reaction by H20 2 and horseradish peroxldase (HRP) was made with three pairs of hydroxamlc acids. Each pair consmted of the aceto- and glycolhydroxamic acid derivatives from one of three different arylhydroxylamines. The parent arylhydroxylammes were the known carcinogens, N-hydroxy-2-aminofluorene and N-hydroxy-4-ammoblphenyl and the noncarcinogen 4-chlorophenylhydroxylamine. All the hydroxamlc acids appeared to be converted to products that were expected on the basra of the previously-proposed mechanism of this peroxldatlve reaction. Each acetohydroxamlc acid gave the corresponding mtroso compound and O-acetyl ester of the starting material m approximately equal amounts. The glycolhydroxamlc acids gave the corresponding nitroso compound and a relatively unstable product that was proposed, by analogy, to be the O-glycolyl ester of the starting materml. A comparison of the initial rates of reaction of each hydroxamlc acid pair showed that the glycolhydroxamlc acid was much more susceptible to the peroxidation reaction than was the corresponding acetohydroxamlc acid. The imtial rate of the reaction was also highly dependent upon the nature of the aromatic ring in the order fluorene ~ blphenyl ~, 4-chlorophenyl. The relative degree of HRP-catalyzed covalent binding to DNA of the aceto- and glycolhydroxamic acids in the fluorene seines was studmd and found to Abbrewatlons HRP, horseradish peroxldase, N-AcO-2-AAF, N-acetoxy-2-acetylammofluorene, N OH-2-AAF, N-hydroxy-2-acetylammofluorene, N-OH-2-AF, N-hydroxy-2-ammofluorene, N-OH-2GAF, N-hydroxy-2-glycolylamlnofluorene,N-GlyO-2-GAF, N-glycoloxy-2-glycolylamlnofluorene,NAeO-4-AAB, N-acetoxy-4-acetylammoblphenyl. N-0H-4-AAB, N-hydroxy-4-acetylarnlnoblphenyl, N-OH-4-AB, N-hydroxy-4-amlnoblphenyl, N-OH-4-GAB, N-hydroxy-4-glycolylammoblphenyl, NGIyO-4-GAB, N-glycoloxy-4-glycolylam~noblphenyl, N-OH-4-C1Acet, N-hydroxy-N-(4chlorophenyl)acetamlde, N-OH-4-CIGIy, N-hydroxy-4-(chlorophenyl)glycolarmde, 2-NOF, 2mtrosofluorene, 4-NOB, 4-mtrosoblphenyl 0009-2797/87/$03 50

© 1987 Elsevier Sclentdlc Pubhshers Ireland Ltd Printed and Pubhshed m Ireland

250 parallel the relative rates of reaction of these substrates in the H202/HRP system It was proposed that glycolhydroxamic acids are likely to be more genotoxlc than are acetohydroxamic acids when subjected to peroxidatlve bloactlvation conditions.

Key words Peroxldase

--

Bmactivation -- Hydroxamic acid -- DNA binding

INTRODUCTION

The hydroxamtc acid functional group is known in many cases to be a proximate genotoxm metabohte of certain arylamldes and arylammes [1-4]. It appears that hydroxamIc acids require further bloactlvatlon before a chemical reaction with macromolecules can occur [1--4] The bioactlvatlon of hydroxam,c acids is currently thought to proceed by several possible pathways, including enzymatm hydrolysis to the hydroxylamme, O-sulfation to produce the N-O sulfate ester or N-O acyltransferase to give the putat,ve N-acetoxy arylamlne [2--4]. The latter two metabohtes are more reactive towards nucleic acids, and are probably ultimate genotoxlcants that can bind covalently to nucleic acids and protein An addlt,onal metabolic pathway that has been proposed as a mechanism for hydroxamlc acid bloactlvatlon involves the act,on of certain peroxidases [5]. In this reaction (Fig. I), the hydroxamlc acld (I) is oxldlzed by a 1 eprocess to give the mtroxyl free radical intermediate (II), which is then thought to undergo a blmolecular dismutatlon reaction to produce the corresponding aryl nItroso compound (Ill)and the O-acetate ester (IV) of the hydroxamm acid as products [5,6]. Both of these products are chemically more reactive than the hydroxamm acld precursor, and can react with certain nucleophlles. In many cases, the O-acetate ester (IVa) is a sufhciently reactive electrophlle, and covalent binding with D N A bases can occur [5,6] Although the nltroso functional group is known to react readily wlth protein sulfhydryls to form protein adducts [8], it is posslble that a further reaction

2

~

HO 0 I H

N- c-R - - - ' ~

2 [ ~ ~

[

o II

]1£

N-C-R - - ' ~

+

0 /C-R

1o b

R= CH 5 R=CH20H

IT

0

I 0

[ ~ N , ~ _ R

Z~'o

R= CH:5

b R= CH20H Fig 1 Proposed mechamsm for the peromdatlve or 1 e- omdatmn pathway for hydroxamm aclds Ia, N-OH 2-AAF, Ib, N-OH-2-GAF, III,2-NOF, IVa, N-AcO-2-AAF, IVb, N-GIyO-2-GAF

251 is necessary m order for mtroso compounds to react with nuclem acid bases Metabohc reduction of the mtroso group to the hydroxylamme ~s generally a rather facile reaction [8-10]. The hydroxylamme might react directly with nuclem acid bases, partmularly under slightly acldm cond,tlons [3], or an add,tlonal bloactlvatlon reaction might facilitate such covalent binding to macromolecules. Recently O-estemflcatmns of the hydroxylamme with acetate [11] or sulfate [12] have been shown to be possible contributors to hydroxylamine bmactlvatmn. The evidence that the peroxide-catalyzed bmactlvatmn reactmn actually occurs within certa,n tissues ,is fragmentary The observat,on that Nhydroxy-2-acetylammofluorene (N-OH-2-AAF) was more potent as a mammary gland tumorigen than was the corresponding am~de, hydroxylamme (N-OH-2-AF) or mtroso compound (2-NOF) suggested the posmbfllty of bmactlvatmn of this hydroxamm acid by a 1 e- oxidation reactmn [13,14]. The subsequent demonstratmn of the ability of a mammary gland perox~dase to convert N-OH-2-AAF to 2-NOF and N-AcO-2-AAF supports this posslbihty [15]. We have suggested the possibility that an unusual type of hydroxamm acid, whmh possesses the N-glycolyl group rather than the much more common N-acetyl group, might be a metabohte of arylamme xenobmtms m general [9,16] The extent to whmh such an unusual metabohte might be produced is probably small, and very much dependent on the chemical structure of the parent arylamme xenobmtm and also on the type of t~ssue under conslderatmn The conversmn of an arylamine xenobiotm ,n part to the mtroso metabohte must occur m order for a glycolhydroxamm acid to be produced. Although a glycolhydroxamlc acid may only be produced as a trace metabohte, it is possible that it contr,butes slgmficantly to the overall toxmlty of a xenoblotm. From a study of the metabohc chemistry of glycolhydroxamm acids, we found that this type of hydroxam,c acid is much more susceptible to perox~dase-catalyzed bloactlvat,on than is the corresponding acetohydroxamm acid We now report on the relative reactlwty of the glycol- and acetohydroxamlc acid der,vatlves of 2ammofluorene, 4-ammob,phenyl and 4-chloroamhne with H202 m the presence of the model enzyme HRP MATERIALS AND METHODS Chemwals Unlabelled substrates and metabohtes, whmh were prepared by previously descmbed methods were N-OH-2-AAF [17], N-OH-4-AAB [18], NOH-2-GAF [19], N-OH-4-GAB [9], N-AcO-2-AAF [20], 2-NOF [21], 4-NOB [9], NOH-4-C1Acet [22] and N-OH-4-C1GIy [9]. [9-14C]N-OH-2-AAF and 2-mtro[914C]fluorene were obta,ned from Chemsyn Scmnce Labs (Lenexa, Kansas} [9-14C]N-OH-2-GAF was prepared m the following manner: 2-mtro[9~4C]fluorene (2.1 rag, 10 ~mol) m 4 ml tetrahydrofuran (peroxide-free) was placed m a 10 ml round bottom flask eqmpped with a magnetm stirrer and

252 which was cooled to - 5 ° C in an ice-salt bath. A N2-atmosphere was employed and all solvents were thoroughly purged with N 2. To the solution was added 2.5 mg of 10°/o Pd/C and 5 mg of hydrazme hydrate, then the mixture was stirred for 90 mm while coohng and the N 2 atmosphere was maintained. (Actual reachon time was determined through periodic analysis of 5 pl aliquots by HPLC in order to follow consumphon of starting materml.) The reactmn mixture was filtered through an Acrodlsc CR, 0.45 ~m (Gelman Scmnces, Ann Arbor, MI) into a solution of FeCI~ • 6H20 (10.8 mg, 40/~mol) in 10 ml H20, whmh had been cooled at 0°C. After s h r r m g for 1 ram, the reactmn mixture was extracted with 10 ml CH2C12, then the orgamc layer was washed with 2 ml H20, dried over Na2S04 and filtered. To the filtrate was added 2 ml 95o/o ethanol, then the CH2C12 was removed m vacuo to leave the 2-NOF as a solution m ethanol, which was used immedmtely m the reactmn with transketolase [9]. The 2-NOF solution was added dropwlse m the course of 1 h to a s h r r e d solutmn of achvated [9] transketolase (100 units, type X, Slgrna Chemical Co.) and fructose 6-phosphate (561 rag, 1.5 mmol) m 150 ml 0.05 M Tris--HCl buffer (pH 7.4) and heated at 37°C. The reactmn was incubated for an additmnal 60 mm after the addltmn of 2-NOF was completed, then extracted twine with 200 ml of ethyl acetate. The organic extract was dried (Na2SO 4) and evaporated to dryness m vacuo. The residue was dissolved in 1 ml dlmethylformamlde and 100-pl ahquots were fractmnated by preparahve-scale HPLC, whmh uhhzed a ~Bondapak Cls column (30 cm × 3.9 mm) with a solvent of 60o/o aqueous methanol buffered to pH 3.5 with 0.01 M KH2PO 4 and which contained 0.01°/o desferal mesylate. Appropriate fractmns containing N-OH-2-GAF were collected and combined To remove traces of hpophfllC contaminants, the combined fractmns were washed twine with an equal volume of hexane, then the aqueous portmn was extracted twice with equal volumes of ethyl acetate. The ethyl acetate extract was drmd (Na2SO4) and evaporated to ymld 1.2 mg [9-14C]N-OH-2-GAF (47o/o yield), which had a radlochemical pumty in excess of 99o/o. The commercmlly-obtained [9-14C]N-OH-2-AAF was purified by HPLC m a similar manner.

Incubatwns of hydroxamw acwl substrates w~th H~O2 and HRP A typmal incubation was conducted at 20°C by addlhon of substrate as a solution m 100 ~1 ethanol to 10.0 ml 0.01 M KH2P04 buffer (pH 7.0}, contained m a 50-ml polypropylene centmfuge tube, followed by vigorous shaking to insure dissolution. HRP ( 1 - 2 0 0 pg) was added as a soluhon m 100 ~1 H20, then a 200-~ aliquot was taken for a T = 0 sample. The reachon was initiated by the addlhon of 100 ~l of a freshly prepared aqueous solution of H202. At demred times, ahquots of 1.0 ml were removed from the reachon and combined with 1.0 ml methanol, which had been pre-cooled to - 2 0 ° C . Analysis of the quenched ahquots for amounts of starting maternal and known metabohtes was achmved by HPLC employing a ~d3ondapak Cls column (30 cm × 3.9 mm) with aqueous methanolic solvents (50 - 60o/o) buffered to pH 3.5 with 0.01 M KH2PO 4 and whmh contained 0.01% desferal mesylate [23].

253

Preparation of D N A for use m binding studies Calf thymus DNA (Sigma type I) m batches of 100 mg m 20 ml H20 was heated m bolhng HeO for 10 min, followed by rapid cooling m an me bath. The denatured DNA was combined w~th 20 ml of 'cell lyres buffer' (10 mM EDTA, 0.1 M NaCl and 1% sodmm lauryl sulfate m 20 mM Tins base, then adjusted to pH 8), an additional 1 ml of 20% sodium lauryl sulfate m H20 and 5 mg of proteinase K, then heated at 37°C for 2 h The incubate was extracted twine for 30 and 15 mm with 40 ml of phenol/CHCIJlsoamyl alcohol/8-hydroxyquinohne (50 • 48 • 2 • 0.1) reagent that was saturated with 0.1 M Tms--HCI (pH 8), containing 0.2% dlthmthre~tol The aqueous layer was extracted with 40 ml of CHC1Jlso-amyl alcohol (24 : 1), then twine with 40 ml H20-saturated dmthyl ether. After addltmn of 2 ml 3 M NaCl solutmn, 100 ml of ethanol was added and the mixture was cooled overmght at 4°C. Following centrifugatmn, the purffmd DNA was drmd under high vacuum Solutmns of 2 mg/ml m H20 were prepared and stored at 4°C

HRP-med~ated binding of [9-1~C]N-OH-2-AAF and [9J~C]N-OH-GAF to DNA Purified and denatured DNA was adjusted to a concentration of 1 mg/ml m 0.01 M KH2PO 4 buffer (pH 7); then 9.0 ml was placed m each of two 50-ml polypropylene centrifuge tubes, followed by the addlhon of [9-14]N-OH-2-AAF to one tube and [9-14]N-OH-2-GAF to the other These substrates were added as solutions m dlmethyl formamlde (e g. 18 ~l of 10 mM soluhons gave 20 ~M reachon concentrahons with 0.70 ~Cl total radioactivity) Following vigorous mixing, 18 ~l of an HRP soluhon (1.0 mg/ml m H20) was added to the incubation tube, then an aliquot of 05 ml was removed for use m the determmahon of actual substrate concentration employed in each reachon To imhate the reactmns, 42 ~l 10 mM aqueous H202 solution was added to give a reactmn concentratmn of 50 ~M At desired hmes, ahquots of 2.0 ml each were taken and immediately extracted In 15 ml polypropylene centrifuge tubes with 4 ml cold H20-saturated ethyl acetate m order to stop the reactmn and to remove unreacted starting material and unbound metabohtes. The aqueous layer was extracted with 4 ml H20-saturated dmthyl ether, then once with 2 ml of the phenol reagent for 15 min, once with CHC1Jlsoamyl alcohol reagent and finally with 2 ml H20-saturated dmthyl ether. The residual dmthyl ether was removed with a N 2 stream, then the aqueous solutmn was treated with 0.1 ml 3 M NaC1 solution The DNA was precipitated by the addltmn of 5 ml of absolute ethanol and coohng at 4°C. After centrIfugatmn and drying of the resulting DNA pellet with a N 2 stream, the pellet was dissolved m 2.0 ml H20 Ahquots of 0 10 ml were analyzed by hqmd scmtfllatmn counting while DNA concentratmn was determined from A260 (by use of the relatmnshlp that 1 0 mg DNA = 20 A260 unats [24]. RESULTS

Ident~f~catwn of reaction products The mlhal products of the reactions of the hydroxamm acids with H202 m

254 the presence of HRP were ,dentlhed by chromatographm methods This was accomphshed by use of authentic standards for the mtroso compounds (2NOF and 4-NOB) and for N-AcO-2-AAF The peaks m HPLC chromatograms of the peroxidase products for each hydroxamm acid were matched w]th those of authentm standards on the bas,s of retenhon Umes and UV spectra generated with a rapid scanning UV detector attached to the HPLC system Since the O-acyl products expected from the N-glycolyl hydroxamm acids (Le N-GIyO-2-GAF and N-GIyO-4-GAB) could not be synthesized, we could only assume that these expected products were the source of the chromatographm peaks whmh eluted shortly after the hydroxam,c acids. For example, peak 2 in Fig. 2a was thought to be due to N-GlyO-2-GAF Th~s is considered to be a reasonable assignment on the basis that the nature of the 6

b

0

b x (n ¢.-

(D (J t-JQ

2

(n

<[

,

0 0

,

I0

~

0

,

,

I0

I

20

30

40

Mm F,g 2 HPLC chromatograms of ahquots from the reachon of fluorene-derwed hydroxamm acids with H2OJHRP The HPLC system was a ~Bondapak C,s column (30 cm × 3 9 ram) with 550/0 aqueous methanol buffered to pH 3 5 with 0 01 M KH2PO 4 and containing 001°/0 desferal mesylate Flow rate was 1 5 ml/mm and detector wavelength was 280 nm with attenuation set at 0 01 AUFS In]echon volumes of 60 ~1 methanol-quenched ahquots were used to generate each of the chromatograms Hydroxam,c ac,d substrates were 50 ~M, H202 was 50 ~M and HRP was present at 2 ~g/ml (a} Reactmn of N-OH-2-GAF at T = 2 mm peak 1, N-OH-2-GAF, peak 2, NGlyO-2-GAF (proposed) (b) reactmn of N-OH-2 AAF at T = 5 mm peak 1, N-OH-2-AAF, peak 2, N-AcO-2-AAF, peak 3, 2-NOF

255 product which produced peak 2 must be less polar than N-OH-2-GAF (peak 1), yet significantly more polar than N-AcO-2-AAF. The peak produced by authentac N-AcO-2-AAF was found to be identical to peak 2 in Fig 2b. Furthermore, our observation that peak 2 m Fig 2a increased mlhally with time, then decreased m consistent with our proposal that N-GlyO-2-GAF is unstable Attempts to synthesize even the O-acetyl ester of N-OH-2-GAF were unsuccessful, which further illustrates the instability of such O-acyl derivatives including N-GIyO-2-GAF. The consumption of each hydroxamm acid and the formation of the O-acyl ester of each hydroxamic acid was determined on a ~Bondapak Cls column and a solvent consisting of 50% or 55% aqueous methanol for the blphenyl and fluorene systems, respectively The solvent contained desferal mesylate to suppress chemmorptlon effects on the hydroxamm acids [23] A second HPLC analysm was conducted to follow the production of the respective nltroso product. For the detectmn of 4-NOB, a ~Bondapak Cls column with 600/0 MeOH was employed, while the detectmn of 2-NOF required the use of a ~Bondapak phenyl column with 60% MeOH as the solvent These particular chromatographm systems were developed to allow for the chromatographm separatmn of mtroso product from the corresponding mtro compound This was done since we thought that the nitro compound would be a possible product of the perox~dahve reactmns, partmularly at later times during the course of the reactmn We have prevmusly observed that peroxldases can catalyze the oxidation of mtroso aromatics to the mtro ox~datmn state, although it is a relahvely sluggish reactmn [25] Neither 2-mtrofluorene nor 4-mtroblphenyl were detected as peroxldahve metabohtes of the fluorenyl or biphenyl hydroxamlc acids, respectively. The limit of detectmn for these nitro compounds was found to be equivalent to a 2% conversmn of the substrates at 50 ~M concentrations.

Analys~s of the k~net~cs of the perox~dat~ve reactwns The kmetms of the reachons of the hydroxamlc acids with H202 m the presence of HRP were determined on the basis of the disappearance of the starting matemals, which was measured by HPLC. No reaction was detected m the absence of either H202 or HRP with any of the hydroxamIc acids employed m this study The rates of the reachons of N-OH-4-AAB and N-OH4-GAB with H202 were found to increase m a hnear relationship with the concentrahon of HRP employed over the range of 1 - 2 0 ~g/ml All further studms were conducted within this linear range of enzyme concentration The effect of H202 concentration on the Initial rate of oxldahon of N-OH-4GAB was studied over the range of 0 0 1 2 5 - 1 0 mM The maximal initial reaction rate was attained when the H202 concentratmn was 0 05 mM or greater The effect of hydroxamic acid concentratmn on the initial rate of the reactmn was investigated for both N-OH-4-AAB and N-OH-4-GAB over the concentration range of 50--250 ~M In this study, the H~O2 concentration was 0.1 mM and the HRP concentration was 10 ~g/ml. The reactions with NOH-4-GAB and N-OH-4-AAB were analyzed at 30 and 120 s, respectively, at

256 which times approx. 15% of the substrates had been consumed. The initial rates of these reactions were found to display a linear relationship with hydroxamic acid concentration up to 100 /~M In excess of 100 ~M, only a slight tendency to enzyme saturation was observed. Subsequent kinetic studies were conducted with hydroxamic acid concentrations at or below 100 ~M, so that first order kinetic relationships were followed. The stolchiometry of the reaction of the hydroxamlc acids in the HRP system was expected to be 2 . 1 for hydroxamlc acid and He02, respectively, since such a relationship had been previously reported by Bartsch for N-OH2-AAF [5]. We confirmed that this same ratio also held for the glycolhydroxamlc acids. This study was conducted by varying the H202 concentration from 0.025 to 0.2 mM while maintaining a constant concentration of each hydroxamic acid at 0.2 raM. The total amount of hydroxamic acid which was consumed by the end of the reaction (10--20 rain) was determined by HPLC. Nearly complete disappearance of each hydroxamic acid was observed when the He02 concentration was 0 10 mM or greater, while at concentrations below 0.10 mM H202, the amount of hydroxamlc acid that had been consumed was approximately twice that of the original concentration of H2O2. The progress curves for the reactions of the aceto- and glycolhydroxamic acid demvatlves of N-OH-4-AB with H2O2 and HRP are shown in Fig. 3. Only the production of the nitroso compound is Illustrated for these two substrates, although an HPLC peak indicative of a second product unique to each bydroxamic acid was also observed. These products are probably the corresponding O-acyl esters of the particular hydroxamic acid [5]; however, synthetic standards were not available for absolute identification Figure 3 illustrates the greater rate of the peroxidatlve reaction of the glycolhydroxamic acid compared to the acetobydroxamic acid when conducted under identical conditions. Plots of log c vs time indicated the first order relationship of the reactions with respect to hydroxamic acid concentration, at least within the first few minutes of the reactions and prior to extensive depletion of H202. Pseudo first-order rate constants for the conditions described in Fig 3 were calculated to be 0 78 rain -1 and 0 18 rain -1 for the glycol- and acetohydroxamic acids, respectively. Thus, N-OH-4-GAB is about 4 times more reactive in this peroxldatlve system than is N-OH-4AAB. Figure 3 indicates that approximately twice as much N-OH-4-AAB was consumed relative to the amount of 4-NOB that was produced. This observation is consistent with that reported in a prior study of the reaction of N-OH-4-AAB with HRP [26]. This 2 1 ratio of bydroxamic acid consumed versus 4-NOB produced was not strictly followed for N-OH-4-GAB after about 5 rain. More 4-NOB was produced than predicted by the 2 1 ratio. The progress curves for the reactions of the aceto- and glycolbydroxamic acid derivatives of N-OH-2-AF with H2O2 and HRP are shown in Fig. 4 The glycolbydroxamic acid, N-OH-2-GAF, was nearly 6 times more reactive to the peroxidatlve reaction than was the acetohydroxamIc acid, N-OH-2-AAF. The pseudo first-order rate constants under the conditions described in Fig. 4

257

'2°I 100l 8O

:~ =L 6O 4O

2O

0

5

I0

15

20

J (mm) Fig 3 Progress curves for the reactions of blphenyi-derlved hydroxamlc acids with H202/HRP The incubations were conducted at 20°C in 0 01 M KH2PO4 buffer (pH 7 0) with 10 pg/ml HRP, 100 ~ I H202 and 110 ~M hydroxamlc acid e , N-OH-4-AAB, m, N-OH-4-GAB, O, 4-NOB from N-OH-4-AAB oxidation, D, 4-NOB from N-OH~4-GABoxldatson

5O

4O

3C ~L 2O

2

4

6

8

10

(ram) Fig 4 Progress curves for the reactions of fluorene-denved hydroxamlc acids with H202/HRP The incubations were conducted at 20°C in 0 01 M K H 2 P O 4 buffer (pH 7 0) wlth i0 ~g/ml H I P , 100 p H H202 and 50 taM hydroxamtc acid e, N-OH-2-AAF, II, N-OH-2-GAF, O, 2 - N O F from N-OH-2A A F oxldatlon, I-I, 2 - N O F from N - O H - 2 - G A F oxldatlon, -- • --, N-AcO-2-AAF, - - - • - - -, proposed to be N-GIyO-2-GAF

258 w e r e c o m p u t e d f r o m plots of log c vs t i m e and w e r e 0 68 m m -1 for N-OH-2G A F and 0 12 m m -1 for N-OH-2-AAF In t h e case of N-OH-2-AAF, the t w o p r o d u c t s , 2-NOF and N - A c O - 2 - A A F , w e r e p r o d u c e d in e q m m o l a r a m o u n t s during the course of the reaction A q u a n t i t a t i v e maternal balance was not attained a f t e r a b o u t a 2-mm r e a c t i o n t~me and b y 10 m m , only a b o u t 7 0 % of the omgmal N-OH-2-AAF s u b s t r a t e could be a c c o u n t e d for as t h e k n o w n compounds In the case of the g l y c o l h y d r o x a m m acid, N-OH-2-GAF, the p r o d u c t m n of 2-NOF was found to be the a m o u n t e x p e c t e d d u r i n g the first m m of the reaction, but was less t h a n e x p e c t e d for longer r e a c t i o n times. Although a s y n t h e t m s t a n d a r d for the O-acyl e s t e r product, N-Gly0-2-GAF, was not available, we e s t i m a t e d its p r o d u c t i o n b y a s s u m i n g t h a t the p e a k height for this p u t a t i v e m e t a b o h t e at T = 30 s was e q u l v a l e n t to the a m o u n t of 2-NOF at this t~me I r r e s p e c t i v e of the a c c u r a c y of our assumption, it is obvious t h a t this p u t a t i v e m e t a b o h t e was not stable u n d e r the reaction conditions since its c o n c e n t r a t i o n dechned steadily a f t e r r e a c h i n g a m a x i m u m b e t w e e n 1 and 2 m m of r e a c t m n time. T h e m a t e r m l balance of k n o w n m e t a b o h t e s and s u b s t r a t e m the case of N-OH-2-GAF was e v e n less t h a n t h a t o b s e r v e d with N-OH-2-AAF By 2 m m , the m a t e r m l balance was 60% and only 40% by 10 m m Obviously, add~tmnal p r o d u c t s of u n k n o w n s t r u c t u r e s are f o r m e d d u r i n g t h e course of t h e r e a c t m n , and are p r o b a b l y the r e s u l t of s e c o n d a r y r e a c t m n s of t h e initial p r o d u c t s . The r e a c t m n k m e t m s of the aceto- and g l y c o l h y d r o x a m l c acids demved f r o m 4 - c h l o r o p h e n y l h y d r o x y l a m m e with H202 in the p r e s e n c e of H R P w e r e i n v e s t i g a t e d N-OH-4-AAB was i n v e s t i g a t e d s i m u l t a n e o u s l y u n d e r identical conditions so t h a t the r e l a t i v e r a t e c o n s t a n t s for t h e s e t w o m o n o a r o m a t i c s u b s t r a t e s (N-OH-4-C1Acet and N-OH-4-C1GIy) could be c o m p a r e d to the h y d r o x a m m acids m the blphenyl and fluorene series T h e s e c o m p o u n d s w e r e studied (data not shown) with an H R P c o n c e n t r a t m n of 10 ~g/ml, since t h e i r r a t e s of omdatlon w e r e c o n s i d e r a b l y less t h a n t h o s e o b s e r v e d in the fluorene series. T a b l e I i l l u s t r a t e s the r e l a t i v e r a t e s of reaction with the p e r o m d a t l v e s y s t e m of all the h y d r o x a m l c acids e m p l o y e d m this study.

TABLE I RELATIVE INITIAL RATES OF REACTION OF HYDROXAMIC ACIDS WITH H202/HRP Substrate

Relative rates a

N-OH-2-GAF N-OH-2-AAF N-OH-4-GAB N-OH-4-AAB N-OH-4-CIGIy N-OH-4-C1Acet

1 00 0 18 0 17 0 04 0 01 0 004

a Determined as the rate of consumption of starting material under hrst order enzymatic reaction conditions

259

HRP-catalyzed covalent binding of N-OH-2-AAF and N-OH-2-GAF to DNA We were particularly interested m determining whether the greater reactwlty of glycolhydroxamic acids in the HRP system resulted m greater covalent binding to DNA relatwe to aeetohydroxamm acids. The two hydroxamm acids m the fluorene series were selected to investigate the relatwe binding to DNA of the aceto- and glycolhydroxamic acids. We found an enzymatm reactmn to be useful m the radmsynthes~s of N-OH-2-GAF at the ~molar level. This labelled compound and commercially-obtained [9-14C]NOH-2-AAF were p u n h e d by preparative HPLC. The covalent binding of 14C-labelled N-OH-2-AAF and N-0H-2-GAF to DNA by the actmn of the HRP system was investigated by allowing the two substrates to incubate under ldentmal conditmns. Each incubate was sampled over a predetermined t~me period, which was selected on the bas~s of the expected half-hfe of each substrate in the HRP system. Concentratmns of 20 ~M were employed for each substrate and the entire experiment was repeated 3 times. A single experiment with each substrate at a 5 ~M concentratmn was also conducted. The results are illustrated m F~g. 5, which demonstrates a time-dependent binding of each substrate up to a maximum amount of covalent binding, whmh reqmred about 5 and 60 mm for the 30

2 X
20

15

E I0

5

0

5

I0

15

20

25

50

t (rain) Fig 5 Progress curves for the covalent binding of 14C-labelled hydroxamm acids to calf thymus DNA The incubations were conducted at 20°C m 0 01 M KH2PO 4 buffer (pH 7 0) with 2 #g/ml HRP, 50/~M H202, 1 mg/ml purlfmd calf thymus DNA and hydroxamm acids at both 5 #M and 20 #M B , N-OH-2-GAF at 20 #M, e , N-OH-2-AAF at 20 ~M, 13, N-OH-2-GAF at 5 pM, O, N-OH-2AAF at 5/~M Bars indicate - S D for 3 rephcates of the 20 ~ runs, and are shown only for selected time points where space allows

260 glycol- and acetohydroxamm acids, respectively. The maximum levels of DNA ,ncorporatlon of N-OH-2-AAF and N-OH-2-GAF were similar and were about 10--15% of the original starting materials. The progress curves m Fig. 5 are nearly identical reciprocals of those for the amount of substrate remaining vs. time as determ,ned from HPLC analysis of the ethyl acetate extracts, which were used to quench the timed ahquots. The initial rate of incorporation of substrate was about 10-fold faster for N-OH-2-GAF than for N-OH-2-AAF, which ,s consistent with the greater rate of the peroxldatlve reactmn of the glycolhydroxamm acid relative to the acetohydroxamlc acid. Also, the substrates at 20 ~M concentrat,ons gave about 4 times faster and total incorporation of label than when the substrates were employed at 5 ~M concentrations, which confirms an expected dependency on substrate concentration. DISCUSSION

In agreement w~th the results of Bartsch et al [26], we found that the acetohydroxamlc acid derivatives of N-OH-2-AF and N-OH-4-AB were good substrates for HRP The nature of the aromat,c r,ng system was found to have an important effect on the rate of oxldat,on of the acetohydroxamm ac,d functional group, since N-OH-2-AAF was oxidized approximately 4 times faster than was N-OH-4-AAB. In turn, the non-carcmogemc 4-chlorophenyl ring system resulted m a very slow react,on of the acetohydroxamm acid, NOH-4-C1Acet, with HRP since N-OH-2-AAF reacted 45 times faster than NOH-4-C1Acet An explanation for this pronounced effect of the r, ng system on the rate of the perox~datlon react,on is not obvious. Perhaps the major finding in th,s study was our observation that the glycolhydroxamm acid functional group ,s much more susceptible to HRP oxldat,on than is the acetohydroxamm acid group This observation was made with three different aromatic ring systems, and may be a fundamental difference between these two types of hydroxamm acids. The magnitude of this difference was greatest m the fluorene series m which N-OH-2-GAF was 5.6 times more reactive than N-OH-2-AAF, while the glycolhydroxamm acids m the blphenyl and 4-chlorophenyl systems were 4.2 and 2.5 times more reactive than the acetohydroxamlc acids, respectively. There are no hterature reports concerning the relative susceptibility of these two types of hydroxamic acids to one-electron chemical oxidants such as Ag20 and Fe(CN)~-. Only two reports have been made concerning the reactlwtms to such chemical oxidants of hydroxamm acids derived from acyl groups other than acetyl [6,27]. In those studies, no kinetic data were presented and a compamson of rates of perox~datlon reactions could not be made for various types of hydroxamm acids. The second major finding of ~hls study was our observat,on that the glycolhydroxamm acid, N-OH-2-GAF, was covalently bound to DNA at a rate that was nearly 10 times faster than the rate found for the acetohydroxamlc ac,d, N-OH-2oAAF Although the total amount of DNA binding of N-OH-2GAF and N-OH-2-AAF was similar when each reaction was allowed to

261

proceed to completion, the rate of such binding is more important for toxicological conslderahons. Obviously the glycolhydroxamlc acid would be predicted to be more genotoxlc than the acetohydroxamlc acid on the basis of relahve rates of nucleic acid binding by the 1 e- oxldahon pathway. However, this remains to be proven by more direct genotoxlclty testing methods, parltcularly since the role, if any, of the peroxldatlve bloachvatlon pathway for hydroxamlc acids has not been clearly estabhshed in any organ s~:stem. In fact, Kadlubar and Beland concluded that peroxldase-catalyzed bloachvahon of hydroxamlc acids does not occur even m tissues that contain high levels of peroxldases, since these hssues do not form acetylated carcmogerj-DNA adducts m vlvo [3]. Their conclusion, which Is based on very hmlted data, is vahd only insofar as the O-ester, N-AcO-2-AAF, is considered to be the olaly ulhmate electroph~le produced by the peroxldatlve pathway. A poslhve correlation between the rates of oxidation of the glycol- and acetohydroxam~c acids and the rates of binding of these two chemicals to DNA certainly suggests a close relahonshlp between these two processes. From what is currently known about the peroxldahve reachon of hydroxamlc acids, the mtroxlde species (II) or the O-acyl esters (IV) are the most suspect as serwng as ulhmate genotoxlc species. Bartsch et al [6] prewously suggested that the nltroxlde free radical (II) m:ght serve as the ulhmate electrophlhc species and attack hssue macromolecules directly On the other hand, N-AcO-2-AAF has long been considered to be a model for the ulhmate genotoxlcant of N-OH-2-AAF, although Kadlubar and Beland have raised semous doubts about the vahdlty of this possible model [3] The corresponding chemical species produced from glycolhydroxamlc acids by HRP should have major differences m chemical reactlwty compared to those from acetohydroxamlc acids. We cannot predict the actual effect upon r e a c h w t y of replacing the acetyl group m the mtroxlde w]th a glycolyl group Research on the relative reachwtles of such mtroxldes by use of electron spin resonance spectroscopy should provide an answer to this queshon. On the other hand, we predict that the O-acylated hydroxamlc acid product, N-GIyO-2-GAF (IVb) should be chemically more reachve than NAc0-2-AAF (IVa). As previously menhoned, the HPLC peak which we suspect to be due to N-GlyO-2-GAF disappears upon standing, which Is a probable reflechon of the high reachvlty of this putahve metabohte. In contrast, the HPLC peak due to N-AcO-2-AAF is qmte stable m comparison. Since glycohc amd is a stronger acid than acetic acid [pK a = 38 vs p K = 4.8] [28], it is reasonable to expect glycolate to be a better /H

0 0 II \CH2''" C ~ 0

I

Ar_N_C_CH2 0 H II 0

0 It HOCH2_C~.o I

Ar-N~

H~O I

/CH 2 C II 0

Fig 6 Intramolecular H-bonding m O-glycoloxy esters of glycolhydroxamlc acids

262 leaving group than is acetate Thus, the breaking of the hydroxamate N-O bond should occur more readdy for IVb than for IVa, and production of the nitrenmm 1on specms should be facilitated The - C H 2 0 H groups m NGIyO-2-GAF might also faclhtate N-O bond breaking as the result of mtramolecular H-bonding forms such as those illustrated m Fig 6. The reactivltms of the putative m t r e n m m ions from IVa and IVb may also be quite different since the - C H 2 0 H group will probably destabdlze the nitrenmm ion relative to the - C H 3 group, as the result of Inductive electron w~thdrawal. It was recently shown t h a t O-acetyl esters of hydroxamic acid can react either through heterolyt~c cleavage of the N-O bond to produce a mtrenmm Ion or through hydrolysis of the ester bond to release the hydroxamm acid [29]. Perhaps the major factor that determines which of these two pathways predominates is the nature of the aromatic ring system on the hydroxamlc acid mtrogen. N-AcO-2-AAF r~ edly only undergoes N-O bond cleavage at or below pH 7, while N-AcO-4-AAB undergoes ester hydrolyms [29] Our observation that the total amount of 4-NOB produced from N-OH-4-GAB actually exceeds 50% of theory during the HRP-catalyzed oxidation is readily explained on the basis of the ester hydrolysis pathway The putative metabohte, N-GIyO-4-GAB, probably undergoes ester hydrolysis preferentially (typical of the blphenyl ring system), which results m the regeneratmn of N-OH-4-GAB which, upon further oxidation, generates additmnal 4-NOB and N-GlyO-4-GAB This cycling of the reactive ester product eventually results m an excess production of the mtroso metabohte No such excess production of 2-NOF was seen m the fluorene series N-O bond cleavage ~s probably the major secondary reactmn of the esters m the fluorene series and this results m an as yet unknown product and a corresponding lack of a material balance HRP has been found to oxidlze certam substrates in a manner similar to that of prostaglandm synthase Thus HRP has been considered as a model for the peroxldatm function of prostaglandm synthase, particularly for certain arylamlnes [30--33]. The contmbutlon of prostaglandin synthase coomdatmns of xenobiotlcs is thought to play a major role m the overall metabolic chemistry of many xenoblotlcs [34] It has been reported that the peroxidase action of prostaglandm synthase catalyzes the 1 e- omdatlon of NOH-2-AAF to give the same products that are produced by HRP [35] On the basis of that report, we predict that glycolhydroxamm acids such as N-OH-2GAF should react m a similar manner w~th this peromdase. Other bmchemmal ~ystems which display peromdatlve activity, and which are known to metabolize acetohydroxam~c acids in a manner similar to the HRP system include methemoglobm, cytochrome P-420 [36] and cytochrome c [37] Recently tissue peroxldases from rat mammary gland and uterus, along with lactoperomdase, were found to effect the oxidation of N-OH-2-AAF via the 1 e- oxidation pathway and also wa an unusual Br--dependent peroxIdatlon [38] Our results from the HRP system suggest that appropriate glycolhydroxamlc acids should be particularly reactive m the presence of such peromdase activity, and that one or more highly reactive metabohtes will be produced.

263 ACKNOWLEDGEMENT

This research was supported by Grant No. ES 03631 from the National Institute of Environmental Health Scmnces, DHHS. REFERENCES 1

2 3

4

5

6

7

8

9

10 11

12

13 14

15 16

17

J H Weisburger and E K Welsburger, Biochemical formation and pharmacological, toxicologlcal and pathologlcal propertms of hydroxylamines and hydroxamlc acids, Pharmacol Revs, 25 (1973) 1-66 S S Thorgeirsson, I B Glowinskl and M E McManus, Metabolism, mutagemclty and carclnogeniclty of aromatm amines, Rev Blochem Toxmol, 5 (1983) 349-386 F F Kadlubar and F A Beland, Chemmal properties of ultlmate carcinogenic metaboIites of arylammes and arylamldes, m R G Harvey (Ed), Polycychc Hydrocarbons and Carcinogens, A C S Symposium Series 283, American Chemlcal Society, Washington, D C, 1985, pp 341-370 P E Hanna and R B Banks, Arylhydroxylamines and arylhydroxamic acids conjugation reactions, in M W A n d e r s (Ed), Bloactlvatlon of Foreign Compounds, Academic Press, Orlando, 1985, pp 375-402 H Bartsch and E Hecker, On the metabohc actlvatlon of the carcinogen N-hydroxy-N-2acetylamlnofluorene Ill Oxidation wlth horseradish peroxidase to yield 2-nltrosofluorene and N-acetoxy-N-2-acetylamlnofluorene, Biochlm Biophys Acta, 237 (1971) 567-578 H Bartsch, M Traut and E Hecker, On the metabolic activation of N-hydroxy-N-2acetylammofluorene II Simultaneous formatlon of 2-nItrosofluorene and N-acetoxy-N-2acetylammofluorene from N-hydroxy-N-2-acetylammofluorene via a free radical intermediate, Biochlm BIophys Acta, 237 (1971) 5 5 6 - 5 6 6 E Krlek and J G Westra, Metabolic activation of aromatic amines and amides and interactions with nucleic acids, in P L Grover (Ed), Chemical Carcinogens and DNA, Vol 2, CRC Press, Boca Raton, 1979, pp 1 - 2 8 P Eyer, Reactions of mtrosoarenes with sulfhydryl groups reaction mechamsm and biological significance, in J W Gorrod and L A Damani (Eds), Biological 0xldation of Nitrogen in Organic Molecules, Ellis Horwood, Chlcbester, 1985, pp 3 8 6 - 3 9 9 M D Corbett and B R Corbett, Effect of ring substituents on the transketolase-catalyzed conversion of nitroso aromatics to hydroxamic acids, Blochem Pharmacol, 35 (1986} 3613-3621 AR Becker and L A Sternson, Nonenzymatlc reduction of nitrosobenzene to phenylhydroxylamme by NAD{P)H, Bmorg C h e m , 9 (1980) 3 0 5 - 3 1 2 T J Flammang and F F Kadlubar, Acetyl coenzyme A-dependent metabolic activation of Nhydroxy-3,2'-dlmethyl-4-aminoblphenyl and several carcinogenic N-hydroxy arylamlnes in relatmn to tissue and specms differences, other acyl donors and arylhydroxamic aciddependent acyltransferases, Carcinogenesm, 7 (1986) 9 1 9 - 9 2 6 C C Lai, J A Miller, E C Miller and A Llem, N-Sulfooxy-2-ammofluorene is the major ultimate electrophlhc and carcinogenic metabohte of N-hydroxy-2-acetylammofluorene in the livers of infant male mice, Carcinogenesis, 6 (1985) 1037-1045 D Malejka-Gigantl, H R Gutmann and R E Rydell, Mammary carcinogenesis in the rat by topical apphcatIon of fluorenylhydroxamlc acids, Cancer R e s , 33 (1973) 2489--2497 D Malejka-Glganti and H R Gutmann, N-Hydroxy-2-fluorenylacetamide, an active intermediate of the mammary carcinogen N-hydroxy-2-fluorenylbenzene sulfonamide, Proc Soc Exp Biol Med, 150 (1975) 92--97 D L Relgh, M Stuart and R A Floyd, Activation of the carcinogen N-hydroxy-2 acetylaminofluorene by r a t mammary peroxidase, Expermntm, 34 (1978) 107--108 M D Corbett and B R Corbett, The reactions of C-nltroso aromatics with a-oxo acids, m J W Gorrod and L A DamanI (Eds), Biological Oxidation of Nitrogen in Orgamc Molecules, Ellis Horwood, Chlchester, 1985, pp 400--408 J G Westra, A rapid and simple synthesis of reactive metabohtes of carcinogenic aromatic amines in high yield, Carcinogenesis, 2 (1981) 3 5 5 - 3 5 7

264 18

19

20

21

22 23

24 25 26

27 28 29

30 31 32 33 34

35

36 37

38

T Shlral, J M Fysh, M Lee, J B Vaught and C M King, Relatlonshlp of metabolic activation of N-hydroxy-N-acylarylamlnes to biologicalresponse in the liver and m a m m a r y gland of the female C D rat, Cancer Res, 41 (1981)4346-4353 M D Corbett, B R Corbett, C I Wel and J D Owusu-yaw, The relative mutagenlcltles of glycohc and acetic acld amlde and hydroxamlc acld derivatives of the carcinogens 2amlnofluorene and 4-amlnoblphenyl, (submitted for pubhcatlon) P D Lothkar, J D Scrlbner, J A Miller and E C Miller,Reaction of esters of aromatic Nhydroxylamlnes and amldes with methionme in vitro A model for in VlVO binding of amine carcinogens to protein, Life Scl, 5 (1966) 1263-1269 P D Lotllkar,E C Miller, J A Miller and A Margreth, The enzymatic reduction of the Nhydroxy derivatives of 2-acetylamlnofluorene and related carcinogens by tissue preparations, Cancer Res, 25 (1965) 1743--1752 MD Corbett and B R Corbett, Metabohsm of 4-chloronltrosobenzene by the yeast Rhodosportdtum sp, Appl Environ Mlcroblol, 41 (1981) 942--949 M D Corbett and B R Chlpko, Quantitative determination of N-arylaceto- and N arylglycolhydroxamIc acids in biochemical reaction mixtures, Anal Biochem, 98 (1979) 169-- 177 J M Clark and R L Swltzer, Expemmental Biochemmtry, 2nd edn, W H Freeman, San Francisco, 1977, p 215 M D Corbett and B R Corbett, Arylamine N-oxidation by the microsomal fraction of germinating pea seedlings (Ptsum sat~vum), J Agrlc Food C h e m , 31 (1983) 1276--1282 H Bartsch, J A Miller and E C Miller, N-Acetoxy-N-acetylaminoarenes and mtrosoarenes One-electron non-enzymatic and e n z y m a t i c oxidation p r o d u c t s of various c a r c i n o g e m c aromatic acetohydroxamic acids, Blochlm Biophys Acta, 273 (1972) 40--51 A R Forrester, M M Ogilvy and R H Thomson, Mode of action of carcinogenic amines P a r t 1 Oxidation of N-arylhydroxamlc acids, J Chem Soc C, (1970) 1081-1083 J March, Advanced Organic Chemistry, 3rd e d n , J Wiley, New York, 1985, p 230 G R Underwood and R B Kirsch, The solvolysis of N-acetoxy-2-acetylamlnofluorene and Nacetoxy-4-acetylamlnoblphenyl delicate balance between mtrenlum ion formation and hydrolysis, J Chem Soc Cbem Commun, (1985)136-138 R W Wise, T V Zenser and B B Davis, Peroxldase metabolism of the urinary bladder carcinogen 2-amino-4-(5-nitro-2-furyl)thlazole, Cancer R e s , 43 (1983) 1518--1522 R S Krauss and T E Eling, Formation of unique arylamlne DNA adducts from 2amlnofluorene activated by prostaglandln H synthase, Cancer Res, 45 (1985) 1680-1686 J A Boyd and T E Ehng, Metabolism of aromatic amines by prostaglandin H synthase, Environ Health P e r s p e c t , 64 (1985) 4 5 - 5 1 Y Tsuruta, P D Josephy, A D Rahimtula and P J O'Brlen, Peroxldasecatalyzed benzidine binding to DNA and other macromolecules, Chem -Biol I n t e r a c t , 54 (1985) 143--158 L J Marnett, M J Blenkowski, M Lelthauser, W R Pagels, A Panthanamckal and G A Reed, Prostaglandin synthetase-dependent cooxygenatlon, in T J Powles, R S Bockman, K V Honn and P Ramwell (Eds), ProstaglandIns and Cancer First International Conference, A R Liss, New York, 1982, pp 97--111 P K Wong, M J Hampton and R A Floyd, Evidence for hpoxygenase-peroxldase activation of N-hydroxy-2-acetylaminofluorene by r a t mammary gland parenchymal cells, in T J Powles, R S Bockman, K V Honn and P Ramwell (Eds), Prostaglandlns and Cancer First International Conference, A R Llss, New York, 1982, pp 1 6 7 - 1 7 9 R A Floyd, Free radical events in chemical and biochemical reactlons revolving carclnogemc arylamlnes, Radlat R e s , 86 (19811 243--263 C L Ritter, D Malejka-Glganti and C F Polnaszek, Cytochrome c/H202-medlated one electron oxidation of carcinogenic N-fluorenylacetohydroxamic acids to mtroxyl free radicals, Chem -Biol I n t e r a c t , 46 (1983) 3 1 7 - 3 3 4 D Malejka-Glganti, C L Ritter, R W Decker and J M Sullman, Peroxidatlve metabohsm of a carcinogen, N-hydroxy-N-2-fluorenylacetamide, by rat uterus and mammary gland in vitro, Cancer R e s , 46 (1986) 6 2 0 0 - 6 2 0 6