% # , _ Chem. Res. Toxicol. 1991, 4, 330-334 _.,_ 0 i_jl)d Hydroxyl RadicalMediatedDemethylenationof , (Methylenedioxy)phenylCompounds *

\, Yoshito Kumagai,t Lena Y. Lin, Debra A. Schmitz, and Arthur K. Cho* M_/t _ _//A

! Department of Pharmacology, UCLA School o[ Medicine, Center [or the Health Sciences, Los Angeles, California 90024-1735 Received December 10, 1990

The oxidative demethylenation reactions of (methylendioxy)phenyl compounds (MDPs), (methylenedioxy)benzene (MDB), (methylenedioxy)amphetamine (MDA), and (methylenedi- oxy)methamphetamine (MDMA), were evaluated by using two hydroxyl radical generating systems, the autoxidation of ascorbate in the presence of iron-EDTA and the iron-catalyzed Haber-Welss reaction conducted by xanthine/xanthine oxidase with iron-EDTA. Reaction products generated when MDB, MDA, and MDMA were incubated with the ascorbate or xanthine oxidase system were catechol, dihydroxyamphetamine (DHA), and dihydroxymethnmphetemine (DHMA), respectively. The reaction required the presence of either ascorbic acid or xanthine oxidase. Levels of each catechol increased in proportion to ferric ion concentration and were suppressed by desferrio-nmine B methanesuLfonate (desferal). Catalase (CAT) inhibited the oxidation by the ascorbate system whereas superoxide dismutase (SOD) had little effect. The

,_; · _ wasadditinoton oinfitiahydterogd enby phyerodxirodgeen to ptheeroxidereactionalonmie, xtusurgge estimstingulatetdhat thheydrogoxidation,en peroxibutdetheacretsactaiosna _: _ '. precursor of hydroxyl radical. SOD and CAT suppressed the demethylenation reactions in the · xanthine oxidase system. Hydroxyl radical scavenging agents such as ethanol, benzoate, DMSO, _,', _ _ and thiourea effectively inhibited the oxidation by both systems. Urea, which has little effect on hydroxyl radical, was without any effect. Theose .results indicated that hydroxyl radical can

._ Introduction Experimental Procedures "'_ ' Although xenobiotics are generally oxidativeiy metabo- Chemicals. MDB and formaldehyde were obtained from lized by mixed-function oxidases (1), interaction of some Aldrich Chemical Co. Inc. (Milwaukee, WI). As the MDB drugs with hydroxyl radical has also been studied since it rained ,mn!! amounts of catechol, the catechol wasremoved with iSa common byproduct produced in biological systems and 1 N NaOH. MDA and MDMA were obtained from the Research i Technology Branch of The National Institute on Drug AboM ,._ hsextremely reactive (2). For example, it has been found that the dehydrogenation of ethanol to acetaldehyde (3), (Rockville, MD). DHA was obtained from Merck Sharp ,mci Dohme Laboratories (West Point, PA). DHMA wassynthesized demethylation of DMSO 1 to produce formaldehyde (4), accordingto published procedures (I1). Catechol, ferric chloric_

_ tdieonnitroosatif anionline oftoN-nitp-amroisnophenodimetholylam(ine6), and (5gen), heydroxyration ]a-of dEi,mDTAutas, ase c(SODorbic),accat,id,1, ,xan,e (OthiAne,T), exanthatnholin,ebeoxinzodaseic a,cid,supdeimethytroxide ., ethylene from methionai (7) can proceed by hydroxyl sulfoxide (DMSO), and urea were obtained from SigmaChemicg radical mediated pathways. Co. (St. Louis, MO). Desferal was obtained from CIBA Lib_ ' ' . (Methylenedioxy)amphetamine (MDA) and (methy- ratories (Horshnm, Sussex, U.K.). All other chemicals used were lenodioxy)methamphetamine (MDMA) are serotonergic of the highest grade available. neurotozins that cause a long-term depletion of 5- IaeulMtion. The ascorbatesystemconsisted of MDPs (MDB, hydroxytryptamine (5-HT) content in the brain (8, 9). It 2 mM; MDA and MDMA, 1 mM), 10_M ferric chloride, 20 _M EDTA, and 30 mM pota_ium ph_phate buffer, pH 7.4, in a final has been suggested that the neurotoxic effect is due to volume of 1.0 mL. The reaction was initiated by the addition metabolites (10). Recently, we have found in studies with of ascorbate (final concentration 1 mM). The zanthine oxidaM rat liver microsomes that MDMA is demethylenated to system consisted of the MDPs described above, 0.5mM xanthine, DHMA by cytochrome P-450 and that the catechol me- 10 _M ferric chloride, 20 ,M EDTA, and 30 mM potassium tebulite was further oxidized to a quinone or semiquinone phosphate buffer, pH 7.4, in a final volume of 1.0 mL. The which reacts with sulfhydryl functions (11). The reaction reactionwasinitiated by addition of 0.025unit of xanthine oxida_ has been reported for brain homogenates also (12), but as Both reactionswere carriedout at 37°C for 10rain and termlnAted the levels of cytochrome P-450 in brain are low, the pos- by the addition of 0.5 mi, of 7.5% perchloric acid containing 30 t sibility that demethylenation may occur by another mM thiourea (the ascorbic acid system mediated reaction could not be completely stopped by perchloric or trichloroacetic acid). pathway was considered. The purpose of the present study The reaction mixtures were centrifuged at 13500gfor 5 rain, and was to determine whether hydroxyl radical could promote the supernatants were analyzed by h/gh-performance liquid chemical cleavage of (methylenedioxy)phenyl compounds chromatography-electrochemical detection (HPLC-ECD). (MDPs) using (methylenedioxy)benzene (MDB), MDA, Determination of Metabolites. The catechol products were and MDMA as substrates (Figure 1). Two hydroxyl analytic1 by HPLC-ECD using a Biophase ODS (250 x 4.6 mm radical generating systems were used,the autoxidation of i.d., 5 mn particlesize;Bieanalytical Systems, Inc.,West Lafayette, ucorbate in the presence of iron and EDTA (4, 13) and the-iron-catalyzed Haber-Weiss reaction in which super- oxide is oxidized to oxygen and H202 reduced to hydroxyl (methylened_Abbreviationsioxy)ben: MzI)ene;Ps, (methMDAy,lencl(methylenediozy)ampioxy)phenycompoul n_ds;hetMDB,amiae; radical (12, I4). MDMA,(methylensd/ozy)meth-mphetamine;DHA,dihydrozyamphet_ ' rural.e; DHMA, dihydroxymethamphetamine; dederal, dederrio,_mine B methanmulfonate;SOD,superoxidedismutaoe;CAT,catalaoeDM; SO, t Merck postdoctoral fellow (1990-1991). dimethylsulfoxide;ECD,electrochemicaldetection.

0893-228x/91/2704'0330502.50/0 C 1991 American Chemical Society Oxidationsi MDPs by Hydroxyl Radical Chem. Res. Toxicol., Vol. 4, No. 3, 1991 331

] o_NH2 O.,_/%./-._,jNHCH3 MDB MDA MDMA _ 8o Figure 1. Structures of (methylenedioxy)benzene (MDB), (methylen.edioxy)amphetamine (MI)A), and (methylenedioxy)- mettlamphetamine (MDMA).

A B e C

.. ._ 0.45 0.55 0.65 0.75 0.85 *: = Electrode potential (V) Figure 4. Electrochemical response curves of eatechol and a reaction product generated during metabolism of MDB by the I ascorbate system°: (la) authentic catechol; (m)reaction product.

Two Hydroxyl Radical Generating Systems"

conditions catechol DHA DHMA _, ___.g,[ l__ _' _ _ asTrecorblbatee I. syDstomemethylenation profoducMDBt . formMOatiA.on, andnmoMDMl/10 A minby complete 5.78 4- 0.57 9.11 4- 0.29 11.51 4- 0.48 -ascorbate 0 0 0 Figure 2. High-performance liquid chromatograms of MDB (A), -Fe s*- EDTA 0.53 4- 0.23 0.65 4. 0.14 0.49 4. 0.11 MDA(B), and MDMA (C) oxidation products obtained from the -Fe s - EDTA + 0 0 0 incubationmixture in the ascorbate system. The chromatographic desferal (100 _M) conditions are described under Experimental Procedures. +Fe8+- desferal6 0.34 4- 0.03 0 0 xanthine oxidase system A _ *' _ complete 4.33 4- 0.04 4.48 4-0.17 3.21 4-0.69 '_Go -- '* -xanthine ozidase 0 0 0

(1 f_ -Fe s+ - EDTA 0.06 4- 0.02 0.55 4- 0.06 0.65 4- 0.01 /_ d-Feesfa+era-l E(ID00TA,M+) 0 0 0 +Fes+- desferalb 0 0.27 _- 0.02 0.04 4. 0.01 from con- aEach MDP concentration was 1 mM, and ail incubations were ch Procedures. Each value is the mean 4- SD of two to six determi- nations, bIncubation was carried out in the presence of 100 _M 'p am desferal instead of EDTA. tesis_ loride coelution with authentic compounds. These reaction wiArozid,_use _ ; p_riedoductsout wundereernotthe condidetectedtions deswhecrribed_ theundinecrubaExperimtion ewasntal nethy '_ _ carried out in the absence of each substrate. The elec- emica trochemical responses due to authentic catechol and the Labs Figure 3. High-performance liquid chromatogr,ms of MDB (A), product formed from MDB by the ascorbate system were i weM MDA (B), and MDMA (C) oxidation products obtained from the compared as the oxidation potential of the detector was incubation mixture in the xanthine oxidase system. The chro- varied over the range 0.5-0.8 V. As shown in Figure 4, the MDB matographic conditions are described under Experimental Pro- K)#M cedures, product peak, (expressed as fraction of its height at 0.8 V) fi_ varied in exactly the same way,as the catechol standard. titio_ IN) column and a glassy carbon working electrode (LC-4, Bioa- When the comparison between authentic DHA or DHMA ddaM nalytical Systems, Inc.) set at 0.7 V (vs Ag/AgCI reference and the appropriate product was carried out, similar results thine, electrode). The mobile phase consisted of 0.1 M citrate, pH 3.5, were also obtained (data not shown). The identical Jsium contsining 1 mM octyl sodium sulfate/aeetonitrile/methanol (&Iff chromatographic and electrochemical behavior of the c_me_ v/v) at a flow rate of 0.7 mL/min. To confh'mcatechol production proposed reaction products with authentic standards ina"_ from MDB, 0.1 M citrate buffer, pH 4.0/acetonitrile (4:1v/v) was supports the conclusion that the MDPs are de- also employed as a solvent. The peak height of each metabolite methylenated to the corresponding catechols. insso was momtored by a Hewlett Packard 3390A recording integrator. The role of the components contained in the ascorbate _zould Formaldehyde formation from DMSO was determined by the and xanthine oxidase systems on the oxidation were ex- acid), method of Nash (15). stained, and the results are shown in Table I. No products were formed in the absence of ascorbate or xanthine ox- R®solt$ idase. The addition of desferal (100 _M), an iron chelator, Incubation of MDPs with the two hydroxyl radical markedly inhibited or blocked the oxidation. Although generating systems resulting in the formation of electro- small amounts of catechols were observed in the absence chemically active products. Figure 2 and 3 show HPLC- of added ferric ion and EDTA, their formation was blocked ECD product profdes for MDB, MDA, and MDMA oxi- by desferal. The formation of the products increased in dation by the ascorbate and xanthine oxidase reaction proportion to the iron added. For example, the addition system. The products (the ascorbate system, 9.9, 10.7, and of 5, 50, and 100 _M ferric chloride to the incubation 11.7 rain; the xanthine oxidase system, 10.1, 10.5, and 12.1 mixture, containing 100 _M EDTA, enhanced the pro- rain) generated from MDB, MDA, and MDMA were duction of catechol from MDB by 1.65-, 4.05-, and 5.82- identified as catechol, DHA, and DHMA, respectively, by fold, respectively, in 1 mM ascorbate. In the xanthine _32 Chem. Res. Toxicol., Vol. 4, No. 3, 1991 Kurnagai et at _/ Scheme 1. Proposed Hydroxyl Radical Mediated Demethylenation of MDP to Catechol Metabolite

lVD_P t oII

Ho'_O_] jR ' H/c- O'_ jR H20 _ HO"_R+ HCOOH H _,,,.._ _',,ff HO_ _ HO_Catechol_ _''_ Formate

Table Il. Effect of SOD, CAT, and Hydrogen Peroxide on Table III. Effect of SOD, CAT, and Hydrogen Peroxide om MDB, MDA, and MDMA Demethylenation by the MDB, MDA, and MDMA Demethylenation by the Xanthine Aecorbate System a Oxidue System a demethylenation, demethylenation, % of control % of control addition MDB MDA MDMA addition MDB MDA MDMA none 100 100 100 none 100 100 150 SOD (10 units) 111 _- 7 107 4- 1 97 4-2 SOD (10 units) 43 4- 2 104- 1 114. 1 SOD (50 units) 114 4- 1 91 4- 3 98 8=1 SOD (50 units) 39 4. 1 0 0 CAT (20 units) 45 4. 1 46 4. 7 50 4. 1 CAT (20units) 38 4- 1 54 4- 2 40 4. 1 CAT (100 units) 21 '_ I 154- 5 18 4-1 CAT (100 units) 12 4- 1 13 4- 2 13 4. I _*' l · SOD (10 units) + CAT (20 unite) 46 4- 1 53 4- 1 49 4- 2 SOD (10 units) + CAT (20unite) 43 -_5 7 4- 4 11 -_I ,_w SOD (50 units) + CAT (100 units) 21 4- 1 154- 1 164- 2 SOD (50 units) + CAT (100units) 34 4- 2 0 0

'_.'_I_._:"' H_0_ (0.5 mM) 520 4-2 627 4- 18 592 4-5 H202(0.5 mM) 117 4-4 1134- 12 1074.8 aMDPs were incubated under the conditions described under aMDPs were incubated under the conditions described under Ig".',_I _ Experimental Procedures. Each valueis the mean 4-SD of two to Experimental Procedures. Each value is the mean _- SD of two to four determinations, four determinations.

,, ' oxidase system the same concentrations of ferric chloride Table IV. Effect of Hydroxyl Radical Scavenging Agents if' '_ increased catechol formation by 1.06, 2;18, and 3.23 times, on MDB, MDA, and MDMA Demethylenation by the [; respectively, in 0.5 mM xanthine and 0.025 unit of xan- Aecorbate System a ,,; ' thine oxidase over a control to which iron was not added, demethylenation, % of control Ferric ion could also be replaced with ferrous iron in both addition concn, mM MDB MDA MDMA "" ' systems. Removal of EDTA alone from the ascorbate or none 100 100 100 xanthine oxidase system caused a slight suppression of the ethanol 10 74 :l:1 60 4- 1 52 4- 1

_-}_ _ frerreactionsic chlor(dideata (10not_Msh)ownand). EDTAAt fix(2ed0 _cMon),centrationsthe amountsof benzoate 1005 6221 4-:_ 31 3514 :l:_- 1 4123 4-1 _ of ascorbate, xanthine, and xanthine oxidase were varied 50 21 4- 1 6 4- 1 9 4-1 , in order to determine optimal conditions for de- DMSO 505 41294-2_- i 364-7 4-11 387:_4-12 ( _ methylenation reaction. Catechol production depended thiourea 1 _3 4- 1 60 4- 1 62 4-3 on ascorbate concentrations. In the xanthine oxidase 10 25 4- 1 11 4- 1 14 4-1 "· (0.025 unit) system, the reaction reached a plateau level urea 1 108 _- 2 103 4. 3 100 4. 4 · · of (1,5mM xanthine. When the demethylenation of MDB 10 108 :_ 1 106 4. 2 101 · 5 Was compared to demethylation of DMSO, a typical sub- aMDPs were incubated under the conditions described under strate for hydroxyl radical mediated oxidation, at the same Experimental Procedures. Each value is the mean 4-SD of twoto concentration (1 mM), catechol production from MDB and four determinations. formaldehyde production from DMSO were 5.78 and 4.67 nmol/10 rain, respectively, in the ascorbic acid system. Table V. Effect of Hydroxyl Scavenging Agents on MDB, Since superoxide and hydrogen peroxide, in addition to MDA, and MDMA Demethylenation by the X_athine hydroxyl radical, are generated from the ascorbate and Oxidase Syatem · xanthine oxidase systems (2, 16), the effects of scavengers demethylenation, % of control for these active oxygen species on the cleavage of MDPs addition conch, mM MDB MDA MDMA were examined. Table II shows the oxidation of MDB, none I00 100 100 MDA, and MDMA by the ascorbic acid system in the ethanol 10 71 4- 4 47 4-3 314-1 presence of SOD and CAT. CAT inhibited cateehol, DHA, 100 21 4- 2 14 4-! 104- 1 and DHMA formation whereas SOD had no effect on the benzoate 5 58 4. 1 32 _-2 29 4. 1 demethylenation reactions. The effect of simultaneous 50 22 4. 1 11 4. 1 124. 1 addition of SOD and CAT was similar to that of CAT only. DMSO 5 37 4. 3 22 4. 2 21 4. 1 50 74-1 4_-1 74-1 The addition of hydrogen peroxide (0.5 mM) resulted in thiourea 1 68 4- 4 45 4-I 38 4- 4 the stimulation of each oxidation but did not promote 10 15 4- 1 3 4- 2 0 demethylenation by itself. The -nnthine oxidase mediated urea 1 97 4- 4 90 4. 4 98 4. 1 oxidation was significantly inhibited by both SOD and 10 98 4- 1 90 4- 2 99 4-5 CAT (Table III). Exogenous hydrogen peroxide did "MDPs were incubated under the conditions described under stimulate the xanthine oxidase driven reaction to some Experimental Procedures. Each value is the mean 4. SD of twoto extent (Table III). four determinations. Table IV shows the effects of hydroxyl radical sca- venging agents on the MDB, MI)A, and MDMA oxidation thiourea inhibited all reactions in a concentration-de- by ascorbate system. Ethanol, benzoic acid, DMSO, and pendent manner. Thiourea dramatically suppressed de- :i et al. Oxidation o[ MDPs by Hydroxyl Radical Chem. Res. Toxicol., Vol. 4, No. 3, 1991 333

methylenation of MDPs at 10 mM. But comparable preparations in vitro. The enzymatic nature of the reaction concentrations of urea, which reacts poorly with hydroxyl in the brain has not been characterized, and although radical (4, 17), had little effect on the demethylenation, mediation by cytochrome P450 is likely, the levels of this Similar results were obtained from experiments with enzyme are very low. An alternative is an oxygen-mediated zanthine oxidase system (Table V). pathway such as the one reported here in which a reactive oxygen source is connected to hydroxyl radical which could Discussion then cleave the methylenedioxy group. The results presented here indicate that attack by hy- Acknowledgment. We thank Mrs. Emma W. Di Ste- droxyl radical, generated by the two typical generating fano for her excellent contributions to this work. This systems, upon MDB, MI)A, and MI)MA produces the research was supported by USPHS Grant DA 04206. corresponding demethylenated products, catechol, DHA, Registry No. MDB, 274-09-9; MDA, 4764-17-4; MDMA, _de om and DHMA, respectively. 54946-52-0; DHA, 3583-05-9;DHMA, 20521-18-0;catechol, 120- maiu Previously, Marshall and Wilkinson (I8) had speculated 80-9. that inhibition of Fenton's reagent mediated epoxidation n' ofaldrin by insecticide synergists such as MDPs appeared References to be due to the interaction of the MDPs with hydroxyl (1) Jakoby, W. B. (1980)Enzymatic basiso/detoxification, p 414, AcademicPress, New York. MI)MA radical, and the present data confirm the hypothesis. (2) Trager,W. F. (1982)The postenzymatic chemistry of activated L00 Product formation was dependent on substrate con- oxygen. Drug.Metab. Rev. 13,51--69. 11· I centration and required addition of ascorbate or xanthine (3) Cederbaum, A. I., Dicker, E., and Cohen, G. (1978)Effect of 400_-1 0xidase. When ferric ion-EDTA was not added to the hydroxyl radical scavengerson microsomal oxidation of alcohol 13 _ 1 incubation mixture of ascorbate or xanthine and xanthine and on associated microsomal reactions. Biochemistry 17, 11 _ I 0xidase, limited demethylenation did occur, probably due 3o58-3064. 0 to the presence of iron in the pho6phats buffer used (19). (4)ofKleiformaln, Sd. ehM.,ydeCohduenri,ngG.me,antadbCeolisdemrbofaudm,imA.ethI.yl(1980)sulfoziProdde byuctihyo-n L07i 8 The addition of desferal, which inactivates ferric ion (20, droxyl radical generating system. Biochemistry 20, 6006--6012. i under 21),to the reaction mixtures caused complete or marked (5) Heur, Y.-H., Streeter, A. J., Nims, R. W., and Keefer, L. K. [ two to inhibition of product formation (Table I). All reactions (1989)The Fenton asa nonensymatic model formicrosomal de- wereaccelerated in proportion to the amount of ferric ion nitroeation of Nonitro_odiethylamine. Chem. Res. Toxicol. 2, added. These results demonstrated the requirement of 247-253. ,gents ironsalts for the active oxygen mediated demethylenation. (6)drIongzylaelmtean-d Sbyunthdbeerg,cytoMch.,roandmeEkP-4st50r6-mde, pGe.nde(1982)nt hAydnrilineoxylisrahdi-y- the We found that CAT suppressed both the ascorbic acid cai-mediated oxygenation mechanism. Biochem. Biophys. Res. and xanthine oxidase mediated product formation. Al- Commun.106,625-631. _ol though hydrogen peroxide stimulated demethylenation, (7) Beauchsmp, C., and Fridovich,I. (1970)A mechanism for the it did not initiate the reaction. These data suggest that production of ethylene from methional: the generation of the hydrogen peroxide is a precursor to hydroxyl radical, 4hy6d41rox-46yl46rad. ical by xanth/ne oxidase. J. Biol. Chem. 245, ! Consistent with this notion, the oxidation of MDB, MDA, (8) Schmidt, C.J. (1987)Neurotoxicity of the psychedelic amphe. 1 and MDMA was inhibited by typical hydroxyl radical tamine, methylenedioxymethamphetamine. J. Pharmacol.Exp. I scavenging agents such as ethanol, benzoic acid, DMSO, Ther.240, 1-7. 1 and thiourea. The production of catechol metabolite (9) Stone, D. M., Merchant, K. M., Hanson, G. R., and Gibb, J. W. 2 generated during the interaction of MDP with hydroxyl (1988)Immediate and long-termeffects of 3,4-methylenedioxy- 1 radical is thought to occur as shown in Scheme I. The methamphetamine on serownin pathways in brain of rat. Neu- _' 3 ropharrnacology247, 79-87. . i I ascorbate and xanthine oxidase systems which are based (10) Schmidt, C. J., and Taylor, V. L. (1988)Direct central effects 4 on Fenton' (22) and iron-catalyzed Haber-Weiss (23) re- of acute methylenedioxymethamphetamine on serotoninergic 5 actions require a continuous generation of ferrous from neurons. Eur. J. Pharmacol. 156, 121-131. ! under ferric ion to produce hydroxyl radical. In the ascorbate (11) Hiramateu, M., Kumagai, Y., Unger, S. E., and Cho, A. K. two to system, ascorbic acid reduces ferric ion. In the xanthine (1990) Metabolism of methylenedioxymethamphetamine 0xiclasesystem, superoxide anion functions as the reducing (MDMA): formation of dihydroxymethamphetamine and a qui- agent. SOD markedly suppressed the catechol formation nonTheer.id2e5nt4,ffied521-52as7its. glutathione adduct. J. Pharmacol.Exp. MDB, in the xanthine oxidase system but had no effect in the (12) Lira, H. K., and Foltz, R. L. (1988) In vivoand in vitro me- lee secorbic acid system. The reason for the difference in SOD tabolism of 3,4-(methylenedioxy)methamphetamine in the rat: sensitivity is that superoxide anion is not an essential identification of metabolites using an ion trap detector. Chem. · ol reducing agent for ferric ion in the ascorbate system. Res. Toxicol. 1,370--378. Klein et al. (4) have proposed measuring formR!dehyde (13) Cederbanm, A. I., and Berl, L. (1982)Pyrasole and 4-methyl- es a marker for hydroxyl radical reaction with DMSO. The pyrazole inhibit oxidation of ethanol and dimethyl sulfoxide by hydroxyl radicalsgeneratedfrom ascorbate,xanthine oxidase,and 1 present data show that the rate of the ascorbate system rat liver micrcaomes. Arch. Biochern.Biophys. 216, 530-543. k 1 mediated demethylenation of MDB was greater than that (14) Puntrarulo, S., and Cederbaum, A. I. (1987) Production of 1 of demethylation of DMSO, so that the oxidation of MDPs 4-hydroxypyrazole from the interaction of the alcohol de- I_1 may be a useful tool in studies of the generation of hy- hydrogenase inhibitor pyrazole with hydroxyl radical. Arch. L 1 droxyl radical in biological systems. Biochem.Biophys. 255, 217-225. k I Lim and Foltz (12) reported that demethylenation of (15) Nssh, T. (1953)The calorimetric estimation of formaldehyde by means of the Hantzsch reaction. Biochem. J. 55, 416-421. 4 MDMA occurs in brain homogenate. The present results (16) Kellogg, E. W., HI, and Fridovich,I. (1975)Superoxide, hy- 1 raise the possibility that oxidative demethylenation of drogen peroxide, and singlet oxygen in lipid peroxidation by a 5 MI)MA in brain could involve participation of not only xanthine oxidase system. J. Biol. Chem.250, 8812-8817. under cytochrome P-450 but locally generated hydroxyl radical (17) Cederbaum, A. I., Dicker, E., Rubin, _, and Cohen, G. (1977) two to aSwell. The effect of dimethylsulfoxide and other hydroxyl radicalsca. vengers on the oxidation of ethanol by rat liver micrcaomas. The demethylenated products of MDA and MDMA are Biochern.Biophys. Res. Commun.78, 1254-1262. highly polar compounds whose access to the CNS from the (18) Marshall,R. S., and Wilkinson, C. F. (1973)The interactionof m-de. periphery should be very lirnited. However, lam and Foltz insectidesynergists with nonenzymatic model oxidation systems. _l de- (12) have shown that the reaction can occur in brain Pestic. Biochem. Biophys.2, 425-436. · 334 Chem. Res. Toxicol. 1991, 4, 334-340

(19) Cohen, G. (1977)An acetaldehyde artifact in studies of the lron-catelyzed hydroxyl radical formation. J. Biol. Chem.2fm, interact/on of ethanolwith biegenicsmh_,eu_tem_, the oxidl_ 3620-3624. / of ethanol by smcorbicacid. J. Neurochem. 29, 761-762. (22) Sheldon, 1_ A., and Koch/, J. K. (1981)Metal-CatalyzedOzi. (20) Gutteride, J, C., Richmond, R., and Halliwell, B. (1979)In- dations o[ OrganicCompounds,pp 35-38, Academic Press,New h/bition of the iron-catalyzedformation of hydroxyl radicalsfrom York. superoxide and of lipid peroxidation by desferrioumine. Bio- (23) Haber, F.,and Weiss, J. (1934)The catalytic decomposit/ooof chem. J. 184, 469-472. hydrogenperoxideby iron salt_ Proc.It. Soc. LondonSer.A147, (21) Graf,F_.,Mahoney,J. 1_,Bryant, R G.,and Eaton, J. W. (1984) 332-351.

1,3-Dlalkyltrlazenes: Reactive Intermediates and DNA Alkylatlon

M. B. Kroeger-Koepke, t R. H. Smith, Jr., _ E. A. Ooodnow,_ J. Brashears,_ L. Kratz, _ A. W. Andrews,! W. G. Alvord, wand C. J. Michejda*,t ABL-Basic Research Program, Program Resources, Inc., and Data Management Services, NCI-Frederick Cancer Research and Development Center, Frederick, Maryland 21702, and ,u,_ .. Department o[ Chemistry, Western Maryland College, Westminster, Maryland 21157 ,_,,Jw · f _a , mlBi Received December 20, 1990

g,_ nitTherosourreactionsea (MNU)of, ca1,3lf-dithymethylutris azen(cet) DN(DAETw),ithN-e1t,h3-diyl-Nm-enitthyrosltriazourenae ((DENMTU)),, anNd-metl-ethhyl-yl-N3-- "* _ methyltriazene (MET) were studied as a function of concentration of the alkylating agents, of , - · various buffers, and of ionic strength. The amount of alkylation at the 7- and OS-positions of r ' _ guanine increased linearly with dose over a 10-fold concentration range. The slopes of the DMT ·_· and MNU curves were identical as were those of DET and ENU. These data suggest that both "_, types of compounds alkylate DNA via a similar intermediate, presumably the corresponding alkanediazonium ion. MET methylates and ethylates DNA, the amount of each product being a function of the competitive formation of the two diazonium ions possible from MET. The MET product ratios could be reproduced by an appropriate mixture of DET and DMT. The !_ alkylation of DNA by DMT and by MET is very sensitive to ionic strength, to the nature of ' the buffer, and to the identity of the salt used to balance ionic strength. In general, the reaction _ is favored by low ionic strength, by amine rather than oxy acid buffers, and by doubly charged inert anions. The alkylation of DNA is inversely proportional to the logarithm of the ionic _ strength over a wid e range. The mutagenic activity of triazenes in Salmonella typhimurium · · is correlated very well with the ability of the triazenes to form adducts, partiod_rly OS-guanine adducts. Thus, symmetrical 1,3-dialkyltriazenes are mutagens in the order of methyl >> ethyl , > butyl = isopropyl, and unsymmetrical l-alkyl-3-methyltriazenes are mutagens in the order ethyl > butyl > isopropyl. The latter order follows the rate of production of the methanedi- azonium ion, the most mutagenic of the diazonium ions.

I ittroductlon final products resulting from the reaction of the diazonium Although the first preparation of 1,3-dlslk-yltriazenes was ion with water are alkyl alcohols and, in some cases, al- reported by Dimroth in 1906 (1), this class of compounds kenes. In addition, we (8) described kinetics and product remained essentially unstudied until a more convenient distributions of the acid-catalyzed decompositio n of un- means of preparation was developed (2). The preparation symmetrical 1,3-dialkyltriazenes. These triazanes exist as (3), hydrolytic chemistry (4, 5), and biological activity (6) a mixture of two rapidly equilibrating tautomeric forms. of a related class of compounds, the 1,3,3-trlalkyltriazenes, They decompose in aqueous buffers by a specific acid have also been investigated recently, catalyzed pathway to give rise competitively to a mlrture The proteolyfic decomposition of 1,3-dialkyltriazenss in of two _!i:A-edlazonium ions. The product distribution aqueous buffers has been reported by our laboratory (7). is determined largely by the stability of the diasonium ion The hydrolytic decomposition is initiated by rapid and formed. Comparative rate data for a series of l.alkyl-_ reversible protonation followed by rate-determining het- methyltriazenes show that the species produced in the erolyzis to an alkanediszonium ion and an alkylsmine. The rate-determining step are the corresponding alkanedi- azonium ions, rather than the carbocations, even in cases such as benzyl where the carbocation is stabilized by tABL_B_ic _ Program,NCI-Freder/ck CancerResearch resonance (8). and Development Center. Alkanedlazonium ions are the putative electrophilic in. !:ProgramWestern MarylandResources,CollegInc., NCI-e. Frederick Cancer Research and termediates responsible for the carcinogenic and cytoWxic Development Center. properties of several important classes of drugs, including IData Management Services, NCI-Frederick Cancer Research alkyltriazenes (9, 10). Triazenes have been shown to al- smd Development Center. _ kylate DNA in vitro and in vivo (11-14), and, consistent

0893-228x/91/2704-0334502.50/0 C 1991 American Chemical Saeietv