Tyrosinase-Catalyzed Oxidation of 17â-Estradiol:Structure Elucidation of the Products Formed beyond

Catechol Estrogen Quinones

Alessandro Pezzella,* Liliana Lista, Alessandra Napolitano, and Marco d’Ischia

Department of Organic Chemistry and Biochemistry, University of Naples “Federico II” ComplessoUniversitario Monte S. Angelo, Via Cinthia 4, I-80126, Naples, Italy

Received February 28, 2005

This paper reports a systematic characterization of the products formed by oxidation of 17â-estradiol (1) with tyrosinase/O2 at low concentrations of physiological relevance. With thesubstrate at 1-10 nM concentration, the main reaction products included, beside the catecholestrogens 2-hydroxyestradiol (2) and 4-hydroxyestradiol (3), 6-oxo-2-hydroxyestradiol (4), 9,11-dehydro-2-hydroxyestradiol (6), 6,7-dehydro-2-hydroxyestradiol (7), and 9,11-dehydro-4-hydroxyestradiol (10). At higher estradiol concentrations, e.g., 1 µM, 6,7,8,9-dehydro-2-hydroxyestradiol (5) and the dimeric products 8 and 9 were also formed. The origin of theseproducts from oxidative routes of 2 and 3 was established. Overall, the results of this studydisclose novel aspects of the reactivity of 1 with the tyrosinase/O2 system and provide the firstinventory of the oxidation products of catechol estrogen quinones.

Introduction

Oxidative metabolism of the estrogen hormone 17â-estradiol (1) is believed to play a key role in the develop-ment of cancer in several tissues including breast andendometrium (1). Although the exact mechanisms forestrogen-induced carcinogenesis still await elucidation,current views implicate hydroxylation at the A ring bycytochrome P450, prostaglandin synthase, and otherenzymes, leading to 2-hydroxyestradiol (2) and 4-hydroxy-estradiol (3), commonly referred to as the catecholestrogens (CEs) (2-4).

These are oxidizable catechols, which may be convertedby peroxidase/P450 to the corresponding catechol estro-gen quinones (CEQs),1 commonly implicated as the

ultimate carcinogenic species (5-10). Little attention hasbeen paid to the nature of the products formed byoxidative conversion of CEQs under conditions mimickingthose within biological systems. Yet, several interestingconversion pathways may be predicted based, for ex-ample, on recent studies of 2-hydroxy-8,9-dehydroestroneoxidation, leading to 2- and 4-hydroxyequilenin (11, 12).

An oxidizing system that could model oxidative trans-formations of estradiol in vivo, thus enabling chemicalstudies of CEQs oxidation, is tyrosinase (13-15). Tyro-sinase, the key enzyme of melanogenesis (15), possessesboth monophenol monooxygenase activity (EC 1.14.18.1,tyrosine, 3,4-dihydroxyphenylalanine:oxygen oxido-reductase) and o-diphenoloxidase activity (EC 1.10.3.1,o-diphenol:oxygen oxidoreductase). Tyrosinase-like pro-teins have recently been identified in estrogen responsivetissues as well as in the nervous system (16-18). Suchenzymes were found to possess both cresolase and cat-echolase activity, as well as estrogen binding properties,and may thus play a role in the conversion of 1 into itshydroxylated metabolites.

Despite scattered observations in the literature (19-21), the products of the interaction of tyrosinase with 1after CE formation have never been investigated indetail.

In the frame of our studies on the mechanisms ofestrogen oxidation and toxicity (22-24), we have nowundertaken a systematic investigation of the tyrosinase-catalyzed oxidation of 1. The study was aimed to providean inventory of the products formed by conversion of theCEs 2 and 3 beyond the o-quinone stage.

Experimental Procedures

Caution: CEQs are hazardous chemicals and should behandled carefully in accordance with statutory procedures andhealth guidelines; IBX is explosive under impact or heating to>200° and should be handled carefully in accordance withstatutory procedures and health guidelines.

* To whom correspondence should be addressed. Tel: +39-081-674130. Fax: +39-081-674393. E-mail: alessandro.pezzella@unina.it.

1 Abbreviations: CE, catechol estrogen; CEQ, catechol estrogenquinone; HMQC, heteronuclear multiple quantum coherence; HMBC,heteronuclear multiple bond correlation; COSY, correlation spectros-copy; ROESY, rotating frame Overhauser effect spectroscopy; ESI,electrospray ionization.

1413Chem. Res. Toxicol. 2005, 18, 1413-1419

10.1021/tx050060o CCC: $30.25 © 2005 American Chemical SocietyPublished on Web 08/05/2005

Materials and Methods. 17â-Estradiol (1), 2-iodobenzoicacid, oxone (2 KHSO5-KHSO4-K2SO4), and sodium borohydridewere from Aldrich Chemie. Mushroom tyrosinase (EC 1.14.18.1)and tropolone were from Sigma. o-Iodoxybenzoic acid (IBX) wasfreshly prepared from 2-iodobenzoic acid according to a reportedprocedure (25).

UV spectra were performed using a diode array Hewlett-Packard spectrophotometer model 8453E. 1H (13C) NMR spectrawere recorded at 400.1 (100.6) MHz using a Bruker DRX-400MHz instrument fitted with a 5 mm 1H/broadband gradientprobe with inverse geometry. For all products isolated, salientproton resonances are listed. Gradient selected versions ofinverse (1H detected) heteronuclear multiple quantum coherence(HMQC) and heteronuclear multiple bond correlation (HMBC)experiments were used. The HMBC experiments used a 100 mslong-range coupling delay. 1H, 1H correlation spectroscopy(COSY) and rotating Overhauser effect spectroscopy (ROESY)experiments were run using standard pulse programs from theBruker library. Electrospray ionization (ESI)-(-)-MS spectrawere recorded with a Waters ZQ quadrupole mass spectrometeron samples dissolved in methanol. HR ESI mass spectra wereobtained with a Finnegan MAT 90 instrument.

Analytical and preparative TLC analyses were performed onF254 0.25 and 0.5 mm silica gel plates using benzene-ethylacetate (50:50) containing 0.5% acetic acid (eluant A) or ethylacetate-cyclohexane (95:5) containing 0.5% acetic acid (eluantB). Silver nitrate-impregnated silica gel plates were preparedas described (26). Sephacryl S 200 HR from Amersham Bio-science was used for tyrosinase purification. The synthesis of 2and 3 was achieved by a modification of a reported procedure(27).

Analytical and preparative HPLC was performed with aShimadzu instrument equipped with a UV detector set at 280and 254 nm. Octadecylsilane-coated columns, 250 mm × 4.6mm, or 22 mm × 250 mm, 5 µm particle size (Sphereclone,Phenomenex), were used for analytical or preparative runs,respectively, at a flow rate of 1.0 or 15 mL/min, respectively,and H2O/acetonitrile/acetic acid 67:30:3 was used as the mobilephase.

Oxidation of 1 by Tyrosinase: General Procedure. Toa solution of 1 (5 mg, 1.9 × 10-5 mol) predissolved in methanol(1 mL), 0.1 M phosphate buffer (600 mL) (pH 7.4) was addedfollowed by tyrosinase in two aliquots at 3 h intervals up to a10 U/mL final concentration. After 18 h, the reaction mixturewas treated with NaBH4 (1 mg), carefully acidified to pH 5.0,and extracted with ethyl acetate (3 × 150 mL). The combinedorganic layers were dried over sodium sulfate and analyzed byHPLC. In other experiments, the reaction was carried out asabove with the substrate at concentrations varying in the rangeof 1 nM to 30 µM and tyrosinase at 1 U/mL final concentrationwhen the substrate was 1 nM to 1 µM. When required, tropolone(1 nM) was added to the incubation mixture with the substrateat 1 µM concentration.

Oxidation of 2 and 3 by Tyrosinase. To a solution of 2 or3 (5 mg, 1.7 × 10-5 mol) in methanol (1 mL), 0.1 M phosphatebuffer (300 mL) (pH 7.4) was added followed by tyrosinase intwo aliquots at 3 h intervals up to a 5 U/mL final concentration.The resulting pale yellow solution was taken under stirring for18 h and then treated with NaBH4, carefully acidified to pH5.0, and extracted three times with ethyl acetate (3 × 60 mL).The combined organic layers were dried over sodium sulfate andanalyzed by HPLC. In other experiments, the reaction wascarried out as above with the substrate at concentrationsvarying in the range of 1-60 µM.

Isolation of Compounds 4-10. For preparative purposes,the oxidation of 2 with tyrosinase was run as described aboveusing 100 mg (3.5 × 10-4 mol) of the starting material at 60µM concentration. After work up of the reaction mixture asabove, the residue obtained (90 mg) was fractionated by HPLCto give five fractions extracted with AcOEt. Fraction A (8 mg,tR 7 min) consisted of pure 4 (8% yield), while fraction B (2 mg,tR 26 min) contained compound 5 (2% yield). Fraction C (7 mg,

tR 31 min) was found to be a mixture of compounds 6 and 7based on NMR analysis and was further purified by fraction-ation on silver nitrate-impregnated TLC plates (eluant B) toafford compounds 7 (Rf 0.45, 3 mg, 3% yield) and 6 (Rf 0.50, 2mg, 2% yield). Fractions D (5 mg, tR 56 min) and E (5 mg, tR 82min) contained compounds 8 (5% yield) and 9 (5% yield).

The reaction of 3 was run under the same conditions using100 mg of the starting material at 60 µM concentration. Afterwork up as above, preparative HPLC fractionation followed byAcOEt extraction of the eluates afforded one main fraction (6mg, tR 35 min) consisting of compound 10 (6% yield).

2-Hydroxy-6-oxo-estra-1,3,5(10)-trien-3,17â-diol (4). UVλmax (CH3OH): 235, 280 and 321 nm. ESI(-)/MS: m/z 301 [M- H]. ESI(-)HRMS calculated mass for C18H21O4 [M - H],301.1440; found, m/z 301.1432. 1H NMR and 13C NMR (CD3OD)spectra were consistent with those reported (28).

2-Hydroxy-estra-1,3,5,6,8-pentaene-3,17â-diol (5). UV λmax

(CH3OH): 241 and 288 nm. 1H NMR (CD3OD) δ (ppm): 0.65(s, 3H), 3.01-3.11 (m, 2H), 3.89 (m, 1H), 6.93 (d, J ) 8.4 Hz,1H, H-7), 7.04 (s, 1H, H-4), 7.22 (s, 1H, H-1), 7.36 (d, J ) 8.4Hz, 1H, H-6). 13C NMR (CD3OD) δ (ppm): 11.5 (CH3), 24.3(CH2), 26.3 (CH2), 36.5 (CH2), 44.3 (C), 48.1 (CH), 81.0 (CH),113.3 (CH), 118.5 (CH), 127.4 (CH), 130.0 (C), 131.2 (C), 132.3(C), 137.2 (C), 148.3 (C), 149.0 (C). ESI(-)/MS: m/z 283 [M -H]. ESI(-)HRMS calculated mass for C18H19O3 [M - H],283.1334; found, m/z 283.1379.

2-Hydroxy-estra-1,3,5(10),9-tetraen-3,17â-diol (6). UVλmax (CH3OH): 230, 270, 290, 316 nm. 1H NMR (CD3OD) δ(ppm): 0.77 (s, 3H), 2.55-2.75 (m, 2H), 3.75 (t, J ) 8.8 Hz, 1H),5.98 (m, 1H, H-11), 6.44 (s, 1H, H-4), 6.99 (s, 1H, H-1). 13C NMR(CD3OD) δ (ppm): 12.0 (CH3), 25.4 (CH2), 30.4 (CH2), 30.7 (CH2)31.1 (CH2), 40.8 (CH2), 43.2 (C), 51.2 (CH), 83.1 (CH), 111.8(CH), 116.7 (CH), 118.0 (CH), 128.5 (C), 129.8 (C), 137.5 (C),145.1 (C), 146.2 (C). ESI(-)/MS: m/z 285 [M - H]. ESI(-)HRMScalculated mass for C18H21O3 [M - H], 285.1491; found, m/z285.1517.

2-Hydroxy-estra-1,3,5,6-tetraen-3,17â-diol (7). UV λmax

(CH3OH): 222, 283, 316 nm. 1H NMR (CD3OD) δ (ppm): 0.76(s, 3H), 3.67 (t, J ) 8.4 Hz, 1H, H-17), 5.75 (dd, J ) 9.6, 2.0 Hz,1H, H-7), 6.29 (dd, J ) 9.6, 2.8 Hz, 1H, H-6), 6.50 (s, 1H, H-4),6.71 (s, 1H, H-1). 13C NMR (CD3OD) δ (ppm): 11.9 (CH3), 24.4(CH2), 26.0 (CH2), 31.3 (CH2), 40.9 (CH), 44.1 (CH), 45.3 (C),52.0 (CH), 82.8 (CH), 112.9 (CH), 115.1 (CH), 128.5 (C), 129.2(CH), 130.3 (CH), 133.2 (C), 144.4 (C), 145.7 (C). ESI(-)/MS:m/z 285 [M - H]. ESI(-)HRMS calculated mass for C18H21O3

[M - H], 285.1491; found, m/z 285.1479.2-Hydroxy-1-[[(17â)-2,17-dihydroxy-19-norpregna-

1,3,5(10)-trien-3-yl]oxy]estra-1,3,5(10)trien-3,17â-diol. (8).UV λmax (CH3OH): 290 nm. 1H NMR (CD3OD) δ (ppm): 0.64 (s,3H), 0.69 (s, 3H), 2.55 (m, 1H), 2.80 (m, 1H), 3.57 (m, 1H), 3.67(m, 1H), 6.08 (s, 1H, H-4′), 6.29 (s, 1H, H-4), 6.80 (s, 1H, H-1′).13C NMR (CD3OD) δ (ppm): 12.2 (CH3), 12.3 (CH3), 27.1 (CH2),28.7 (CH2), 20.3 (CH2), 31.5 (CH2), 38. 3 (CH2), 38.6 (CH2), 38.9(CH2), 39.0 (CH2), 41.8 (CH), 44.5 (C), 44.9 (C), 46. 2 (CH), 47.4(CH), 51.1 (CH), 51.2 (CH), 83.0 (CH), 118.2 (CH), 123.4 (C),127.1 (C), 132.3 (C), 135.5 (C), 142.6 (C), 145.2 (C), 145.1 (C).ESI(-)/MS: m/z 573 [M - H]. ESI(-)HRMS calculated mass forC36H45O6 [M - H], 573.3216; found, m/z 573.3203.

2-Hydroxy-1-[[(17â)-2,17-dihydroxy-19-norpregna-1,3,5(10)-trien-3-yl]oxy]estra-1,3,5,6,8 pentaen-3,17â-diol.(9). UV λmax (CH3OH): 238 and 289 nm. 1H NMR (CD3OD) δ(ppm): 0.65 (s, 3H), 0.77 (s, 3H), 3.08-3.11 (m, 2H), 3.66 (m,1H), 3.70 (m, 1H), 6.14 (s, 1H, H-4′), 6.62 (d, J ) 8.8 Hz, 1H,H-7), 6.81 (s, 1H, H-4), 7.17 (d, J ) 8.8 Hz, 1H, H-6), 7.19 (s,1H, H-1′). 13C NMR (CD3OD) δ (ppm): 10.5 (CH3) 11.0 (CH3),24.1 (CH2), 28.3 (CH2), 30.4 (CH2), 31.0 (CH2), 32.4 (CH2), 36.3(CH2), 38.3 (CH2), 44.2 (C), 45.1 (C), 48.4 (CH), 52.7 (CH), 105.0(CH), 113.1 (CH), 116.2 (CH), 122.5 (CH), 125.7 (CH), 127.2 (C),130.2 (C), 131.1 (C), 132.2 (C), 134.2 (C), 135.3 (C), 144.3 (C),145.0 (C), 146.2 (C), 147.5 (C). ESI/MS: m/z 569 [M - H].ESI(-)HRMS calculated mass for C36H41O6 [M - H], 569.2903;found, m/z 569.2935.

1414 Chem. Res. Toxicol., Vol. 18, No. 9, 2005 Pezzella et al.

4-Hydroxy-9,11-dehydroestra-1,3,5(10)-trien-3,17â-diol(10). UV λmax (CH3OH): 225 and 273 nm. 1H NMR (CD3OD) δ(ppm): 0.78 (s, 3H), 1.1-1.5 (m, 5H), 1.94-2.10 (m, 2H), 2.19(m, 1H), 2.50 (m, 2H), 2.97 (m, 2H), 3.74 (t, J ) 8.5 Hz, 1H),6.05 (m, 1H, H-10), 6.58 (d, J ) 9.0 Hz, 1H, H-2), 6.97 (d, J )9.0 Hz, 1H, H-1). 13C NMR (CD3OD) δ (ppm): 12.0 (CH3), 24.9(CH2), 25.4 (CH2), 31.2 (CH2), 31.3 (CH2), 40.4 (CH), 40.8 (CH2)43.2 (C), 50.0 (CH), 83.1 (CH), 114.4 (CH), 116.8 (CH), 118.1(CH), 125.7 (C), 129.0 (C), 137.8 (C), 144.2(C), 146.4 (C). ESI/MS: m/z 285 [M - H]. HREIMS calculated mass for C18H22O3

[M - H], 285.1491; found, m/z 285.1525.

Results and Discussion

Tyrosinase Promoted Oxidation of Estradiol.Mushroom tyrosinase was selected as a convenientoxidizing system for producing CEQs from 1 and formonitoring their fate under physiologically relevantconditions. At variance with the mammalian enzymefrom melanoma cells (20, 21), which is apparently unableto induce estradiol hydroxylation, mushroom tyrosinasedisplays efficient cresolase activity toward this estrogenand is highly effective in causing CE oxidation to CEQs.

The reaction of 1 with the tyrosinase/O2 system wasexamined with substrate concentration in the range of 1nM to 30 µM in phosphate buffer at pH 7.4. For productdetermination, a work up procedure was developed,involving mild reductive treatment of the reaction mix-ture with sodium dithionite to halt the oxidation andconvert quinonoid species to their reduced forms, followedby acidification to pH 5.0 and extraction with ethylacetate. Substrate consumption and product formationwere tracked by HPLC analysis of the ethyl acetateextractable fraction.

Typical elution profiles of mixtures obtained by oxida-tion of 1 at 1 nM and 30 µM concentrations are reportedin Figure 1a,b, respectively. A more complex reactionpathway is observed at a higher substrate concentration.A preliminary ESI(-)-MS analysis of the HPLC eluatesfollowing ethyl acetate extraction indicated the presenceof the CEs 2 and 3 along with a series of species withmass spectra suggestive of transformation products of 2and 3. To test this hypothesis, in another series ofexperiments, the tyrosinase-catalyzed oxidation of 2 and3 was investigated under similar conditions.

Tyrosinase Promoted Oxidation of 2 and 3. TheHPLC elution profiles of the oxidation mixtures obtainedby oxidation of 2 and 3 at 30 µM concentration are shownin Figure 1c,d. The data suggest that peaks I and III-VII in Figure 1b arise from 2 whereas peak II derivesfrom 3. Accordingly, the oxidation of these CEs wasrepeated on a preparative scale and the products elutedunder the main peaks were isolated by preparative HPLCand were characterized by extensive two-dimensional(2D) NMR and mass spectrometric analysis.

Preparative scale oxidation of 2 was carried out with60 µM substrate and 5 U/mL tyrosinase in vigorouslystirred phosphate buffer at pH 7.4. The structures of theproducts isolated from the oxidation mixture of 2 andspectrally characterized are provided in Figure 2.

The 1H and 13C NMR spectra, as well as 1H,13C HMBC,and HMQC experiments, suggested that the producteluted under peak I was 6-oxo-2-hydroxyestradiol (4).This conclusion was supported by a carbonyl carbonsignal at δ 200.3, the lack of the benzylic methyleneprotons resonating at δ 2.83, and the pseudomolecularion peak at m/z 301 {[M - H], ESI(-)}. Product 4 was

previously obtained by MnO2-promoted oxidation of 2 inan organic medium (28) but has never been describedamong the products formed by oxidation of 1 or 2 underphysiologically relevant conditions.

The mass spectrum of the product eluted under peakIII indicated a molecular mass 4 units lower than thatof 2. The aromatic region of the proton spectrum, featur-ing two singlets at δ 7.22 and δ 7.04 and two doublets atδ 6.93 and δ 7.36 (J ) 8.4 Hz), was suggestive of adihydroxynaphthalene moiety. This conclusion was cor-roborated by 1H, 13C HMBC experiments and comparisonof the UV absorbance with that of 2-hydroxyequilenin(29), allowing eventually formulation of the product as6,7,8,9-dehydro-2-hydroxyestradiol (5). Compound 5 bearsconsiderable resemblance to 2-hydroxyequilenin, whichis produced by oxidative metabolism of 8,9-dehydro-estrone via 2-hydroxy-8,9-dehydroestrone (11); however,to the best of our knowledge, it has never been obtainedby oxidation of 1 or 2.

Peak IV proved to be due to an intimate mixture oftwo related products sharing the molecular mass of adehydro-derivative of 2. Fractionation on silver nitrate-impregnated TLC plates allowed us eventually to isolatethe products in pure form. The sp2 region of the protonspectrum of the faster TLC eluting compound showed twosinglets (δ 6.44 and 6.99) and a multiplet at δ 5.98, apattern of resonance compatible with a 9,11-dehydro ora 8,9-dehydro 2-hydroxyestradiol. The position of thedouble bond at the 9-position was deduced by the NOEcontacts between the δ 5.98 multiplet due to the protonon the trisubstituted double bond and the aromatic H-1proton singlet at δ 6.99. These and other 2D homo- andheterocorrelation experiments allowed us to formulatethe compound as 9,11-dehydro-2-hydroxyestradiol (6).The carbon spectrum of the more polar compound showedin the sp2 region four quaternary carbons (δ 128.5, 133.2,144.4, and 145.7) and four CH carbons, at δ 112.9, 115.1,129.2, and 130.3 showing one-bond correlation withsinglets at δ 6.71 and 6.50 and double doublets at δ 6.29(J ) 9.0, 2.8 Hz) and 5.75 (J ) 9.0, 2.0 Hz) in that other.These features suggested that the compound was 6,7-dehydro-2-hydroxyestradiol (7). Products 6 and 7 werefairly stable in both acidic and alkaline media, withoutappreciable interconversion or degradation.

Mass spectra of products from peaks V and VI sug-gested dimeric structures, one of which (peak VI) was 4mass units lower than expected for a dimer. Extensivespectral analysis, including 2D NMR, eventually allowedcharacterization of the products as the dimers 8 and 9,respectively, in which the monomer units were linkedthrough ether bridges.

For product 8, substitution at the 1-position of onecatechol ring was deduced from distinct long-rangecouplings between the benzylic CH2 carbons on the 6- and6′-positions at δ 38.9 and 39.0 and the two aromaticprotons at δ 6.29 and 6.08, respectively, indicating thatthe adjacent 4- and 4′-positions were unsubstituted. Theinvolvement of the 3-OH group of the other catechol ringin the ether bridge was inferred from chemical shiftanalysis and, in particular, from the upfield shift of oneof the H-4 protons of ca. 0.5 ppm with respect to 2 (δ6.58), suggesting an aryl ring on the adjacent OH group.Brief inspection of the geometry optimized structure of8 (MM+) consistently indicated a spatial proximity of theH-4 proton to the aryl ring in a number of rotamers,whereby a shielding of at least 0.5 ppm can be expected

Tyrosinase-Promoted Oxidation of Estradiol Chem. Res. Toxicol., Vol. 18, No. 9, 2005 1415

(30). Furthermore, the less pronounced upfield shift ofthe other H-4 proton (δ 6.28) was in accord with theshielding effect of a para phenoxy group.

A survey of the literature indicated that a productrelated to 8, i.e., a 2-hydroxyestrone dimer featuring ano-quinone moiety, has been described (31). However, theclose similarity of the aromatic proton spin systems andthe facility of catechols to lose H2 in the EI source to give[M - 2]+ peaks (32) raises the possibility that the productdescribed by the previous authors was in fact theanalogue of 8 in the estrone series.

In compound 9, the three aromatic singlets at δ 6.14,δ 7.19, and δ 6.81 linked to the carbons at δ 116.2, δ105.0, and δ 133.1, in that order, were assigned to theH-1′, H-4′, and H-4 protons, respectively, based on 2D

heterocorrelation experiments. In addition, the doubletsat δ 6.62 and 7.17 showing one-bond correlation withcarbons at δ 122.5 and 125.7 were attributed to the C-6-and C-7-positions, respectively.

Oxidation of 3 under the reaction conditions adoptedfor 2 proceeded at a slightly lower rate with respect to 2(2.8 × 10-10 and 2.3 × 10-9 M s-1, respectively), support-ing previous observations on the higher oxidizability of2-hydroxyestrogens as compared to 4-hydroxyestrogens(4), to give a single major product (peak II Figure 1),which was isolated and subjected to NMR and massspectral analysis. Salient features of the proton spectrumwere two doublets (J ) 8.5 Hz) at δ 6.58 and δ 6.97 anda multiplet at δ 6.05 showing one bond correlation withcarbons at δ 114.4, δ 116.8, and δ 118.1 in that order.

Figure 1. HPLC profiles of the mixtures obtained by tyrosinase-catalyzed oxidation of 1 nM 1 (trace a), 30 µM 1 (trace b), 30 µM2 (trace c), and 30 µM 3 (trace d) at 3 h reaction time. Workup details are in the Experimental Procedures.

1416 Chem. Res. Toxicol., Vol. 18, No. 9, 2005 Pezzella et al.

Five additional quaternary sp2 carbons were clearlydiscernible in the 13C NMR spectrum indicating thepresence of a trisubstituted double bond in addition tothe aromatic A ring of 3. These data and the observedNOE contact between the doublet at δ 6.97 and themultiplet at δ 6.05 indicated a 9,11-dehydro-4-hydroxy-estradiol (10).

At variance with previous observations on the oxida-tion of 2-hydroxyestrone in an organic solvent (31), wewere unable to identify C-C coupling products of 2 and

3 in the oxidation mixtures. With the above productsavailable, separate analyses were run to determinechanges in product distribution with estrogen concentra-tions. The results showed that with 1 at 1-10 nMconcentration, main reaction products included 6-oxo-2-hydroxyestradiol (4), 9,11-dehydro-2-hydroxyestradiol (6),6,7-dehydro-2-hydroxyestradiol (7), and 9,11-dehydro-4-hydroxyestradiol (10), as apparent from Figure 1a, while6,7,8,9-dehydro-2-hydroxyestradiol (5) and the dimericproducts 8 and 9 were formed only at higher concentra-tions of 1, e.g., 1-30 µM.

Mechanistic Remarks. Products 4-10 conceivablyarise by conversion of CEs to the corresponding CEQs,which may subsequently partition among different com-peting pathways, as schematically outlined in Scheme1. Most of these pathways are proposed to involvetransient quinone methide intermediates. These mayundergo addition of water, as in the oxidative route to 4,or may isomerize to give the unsaturated derivatives 5-7

Figure 2. Main products isolated by tyrosinase-catalyzed oxidation of 2.

Scheme 1. Schematic Outline of the Proposed Mechanism of Formation of Products 4-10

Tyrosinase-Promoted Oxidation of Estradiol Chem. Res. Toxicol., Vol. 18, No. 9, 2005 1417

and 10 after oxidation/aromatization steps. Isomerizationto p-quinonemethides (33) is an established conversionroute of o-quinones, which usually does not requireenzymatic assistance (7, 34). However, the possibilitythat this reaction was promoted by laccase, which ispresent as a contaminant in the commercial preparationsof mushroom tyrosinase (35), was considered. To this aim,the oxidation reactions were also carried out with care-fully purified enzyme preparations according to reportedprocedures (35) and were found to give product patternsthat were superimposable to those obtained with thecrude commercial enzyme. Moreover, addition of a spe-cific tyrosinase inhibitor (tropolone) (36) resulted innearly complete suppression of substrate consumptionand product formation.

Whereas the 2,3-CEQ may give rise to two isomericquinone methide intermediates, which appear to contrib-ute equally to product formation when generated in thenanomolar range, the 3,4-quinone can give rise to onlyone quinone methide, and it is possible that this differ-ence accounts for the simpler oxidation mixture of 3. InScheme 1, dimeric derivatives 8 and 9 are proposed toarise from different coupling reactions. This conclusionwas supported by separate experiments showing that 8is by no means converted to 9 when exposed to tyrosinaseat pH 7.4, under the usual reaction conditions. However,redox cycling with intermediate oxidation products can-not be ruled out. Indeed, dimer formation from catecholcompounds may occur either by a one-electron transfervia a charge transfer complex (31, 37) or by addition of 2to the o-quinone via the 3-OH group. Structures 8 and 9would reveal a similar pattern of reactivity of theo-quinones of 2 and 5, favoring coupling at the hindered1-position. In the case of 2, this behavior was largelyanticipated on the basis of the mode of dimerization of2-hydroxyestrone (31). Yet, it was less obvious for 5,which would behave on oxidation like a highly reactive2,3-naphthoquinone (38, 39).

Concluding Remarks

The results of this study yield further insights into theconversion pathways of the CEs 2 and 3 beyond theo-quinone stage. A battery of products have been isolatedand spectrally characterized and may be used as stand-ards in future studies on the identification of oxidativemetabolites of estrogens in vivo.

Although some of the products described here havealready been reported, either as such or as their estroneanalogues, to the best of our knowledge, none of themhas been obtained by tyrosinase-catalyzed oxidation of1, nor by reaction of 1 with other enzymes or chemicaloxidants under biomimetic conditions. The generation ofproduct 5 by oxidation of 2 and the formation of the noveldimer 9 by attack of 2 to the o-quinone of 5 appear to beof particular interest from the chemical viewpoint.

The use of commercially available mushroom tyrosi-nase, although of questionable relevance to mammaliansystems, was necessary for preparative scale reactions.This enzyme may yet be taken as a convenient model ofthe tyrosinase-like activity described in uterine andestrogen responsive tissues, since it displays an efficientmonophenol monooxygenase (cresolase) activity. Fur-thermore, most of the previous studies on the tyrosinase-estradiol interaction were performed with the mushroomenzyme (12, 20, 40).

While products 4, 6, 7, and 10 were formed by oxidationof the estrogens at nanomolar concentrations, products5, 8, and 9 became detectable at substrate concentrationsin the micromolar range. Thus, their formation may beimplicated only under circumstances in which abnormallyhigh concentrations of 1 are present in settings ofoxidative stress, e.g., during estrogen replacement therapy.Clearly, more studies are required before the new prod-ucts can be ranked among estradiol metabolites, andtheir toxicities or DNA binding properties have to betested in comparison with CEs and the CEQs.

Acknowledgment. This study was carried out in theframe of the MIUR [“Sostanze naturali ed analoghisintetici ad attivita antitumorale” (PRIN 2003)] andRegione Campania (L.R. 5, a. 2002) projects. We thankSilvana Corsani for technical assistance and the “CentroInterdipartimentale di Metodologie Chimico-Fisiche”(CIMCF, University of Naples Federico II) for NMR andmass facilities.

References

(1) Lippert, T. H., Seeger, H., and Mueck, A. O. (2000) The impactof endogenous estradiol metabolites on carcinogenesis. Steroids65, 357-369.

(2) Cavalieri, E., Rogan, E., and Chakravarti, D. (2004) The role ofendogenous catechol quinones in the initiation of cancer andneurodegenerative diseases. Methods Enzymol. 382, 293-319.

(3) Klaunig, J E., Xu, Y., Isenberg, J. S., Bachowski, S., Kolaja, K.L., Jiang, J., Stevenson, D. E., and Walborg, E. F., Jr. (1998) Therole of oxidative stress in chemical carcinogenesis. Environ.Health Perspect. 106 (S1), 289-295.

(4) Dwivedy, I., Devanesan, P., Cremonesi, P., Rogan, E., andCavalieri, E. (1992) Synthesis and characterization of estrogen2,3- and 3,4-quinones. Comparison of DNA adducts formed by thequinones versus horseradish peroxidase-activated catechol estro-gens. Chem. Res. Toxicol. 5, 828-833.

(5) Cavalieri, E., Frenkel, K., Liehr, J. G., Rogan, E., and Roy, D.(2000) Chapter 4: Estrogens as endogenous genotoxic agentssDNA adducts and mutations. J. Natl. Cancer Inst. Monogr. 27,75-93.

(6) Bolton, J. L., Pisha, E., Zhang, F., and Qiu, S. (1998) Role ofquinoids in estrogen carcinogenesis. Chem. Res. Toxicol. 11, 1113-1127.

(7) Bolton, J. L., Yu, L., and Thatcher, G. R. J. (2004) Quinoids formedfrom estrogens and antiestrogens. Methods Enzymol. 378, 110-123.

(8) Cavalieri, E. L., Stack, D. E., Devanesan, P. D., Todorovic, R.,Dwivedy, I., Higginbotham, S., Johansson, S. L., Patil, K. D., andGross, M. L., et al. (1997) Molecular origin of cancer: catecholestrogen-3,4-quinones as endogenous tumor initiators. Proc. Natl.Acad. Sci. U.S.A. 94, 10937-10942.

(9) Nutter, L. M., Wu, Y.-Y., Ngo, E. O., Sierra, E. E., Gutierrez, P.L., and Abul-Hajj, Y. J. (1994) An o-quinone form of estrogenproduces free radicals in human breast cancer cells: Correlationwith DNA damage. Chem. Res. Toxicol. 7, 23-28.

(10) Bolton, J. L. (2002) Quinoids, quinoid radicals, and phenoxylradicals formed from estrogens and antiestrogens. Toxicology 177,55-65.

(11) Zhang, F., Chen, Y., Pisha, E., Shen, L., Xiong, Y., van Breemen,R. B., and Bolton, J. L. (1999) The major metabolite of equilin,4-hydroxyequilin, autoxidizes to an o-quinone which isomerizesto the potent cytotoxin 4-hydroxyequilenin-o-quinone. Chem. Res.Toxicol. 12, 204-213.

(12) Zhang, F., Yao, D., Hua, Y., van Breemen, R. B., and Bolton, J.L. (2001) Synthesis and reactivity of catechol metabolites fromequine estrogen 8,9-dehydroestrone. Chem. Res. Toxicol. 14, 754-763.

(13) Seo, S. Y., Sharma, V. K, and Sharma, N. (2003) Mushroomtyrosinase: Recent prospects. J. Agric. Food. Chem. 51, 2837-2853.

(14) Sanchez-Ferrer, A., Rodriguez-Lopez, J. N., Garcia-Canovas, F.,and Garcia-Carmona, F. (1995) Tyrosinase: A comprehensivereview of its mechanism. Biochim. Biophys. Acta 1247, 1-11.

(15) del Marmol, V., and Beermann, F. (1996) Tyrosinase and relatedproteins in mammalian pigmentation. FEBS Lett. 381, 165-168.

1418 Chem. Res. Toxicol., Vol. 18, No. 9, 2005 Pezzella et al.

(16) Garai, J., Tiller, A. A., and Clark, J. H. (1992) Tyrosinase-likepolypeptides in the uterus and in the central nervous system ofrats. Steroids 57, 183-188.

(17) Garai, J., and Clark, J. (1992) Tyrosinase-like activity andestradiol binding in rat uterine nuclear extracts. Steroids 57, 248-256.

(18) Densmore, C. L., Schauweker, T. H., Gregory, R. R., Webb, B.,Garcia, E., and Markaverich, B. M. (1994) Chromatographicresolution of the type II estrogen binding site and a tyrosinase-like enzymic activity from rat uterine nuclei. Steroids 59, 282-287.

(19) Li, K.-M., Todorovic, R., Devanesan, P., Higginbotham, S.,Koefeler, H., Ramanathan, R., Gross, M. L., Rogan, E. G., andCavalieri, E. L. (2004) Metabolism and DNA binding studies of4-hydroxyestradiol and estradiol-3,4-quinone in vitro and infemale ACI rat mammary gland in vivo. Carcinogenesis 25, 289-297.

(20) Jacobsohn, M. K., Byler, D. M., and Jacobsohn, G. M. (1991)Isolation of estradiol-2,3-quinone and its intermediary role inmelanin formation. Biochim. Biophys. Acta 1073, 1-10.

(21) Jacobsohn, G. M., Chiartas, P. L., Hearing, V. J., and Jacobsohn,M. K. (1998) Role of estradiol and 2-hydroxyestradiol in melaninformation in vitro. Biochim. Biophys. Acta 966, 222-230.

(22) Pezzella, A., Lista, L., Napolitano, A., and d’Ischia, M. (2004)Oxidative coupling of 17â-estradiol: Inventory of oligomer prod-ucts and configuration assignment of atropoisomeric C4-linkedbiphenyl-type dimers and trimers. J. Org. Chem. 69, 5652-5659.

(23) Pezzella, A., Manini, P., Di Donato, P., Boni, R., Napolitano, A.,Palumbo, A., and d'Ischia, M. (2004) 17â-Estradiol nitration byperoxidase/H2O2/NO2

-: A chemical assessment. Bioorg. Med.Chem. 12, 2927-2936.

(24) Pezzella, A., Manini, P., Napolitano, A., Crescenzi, O., Barone,V., and d’Ischia, M. (2005) Oxidative chemistry of 2-nitro and4-nitroestradiol: Dichotomous behavior of radical intermediatesand novel potential routes for oxyfunctionalization and B-ringfission of steroidal scaffolds. Steroids 70, 543-550.

(25) Frigerio, M., Santagostino, M., and Sputore, S. (1999) A user-friendly entry to 2-iodoxybenzoic acid (IBX). J. Org. Chem. 64,4537-4538.

(26) Wilson, R., and Sargent, J. R. (2001) Chain separation ofmonounsaturated fatty acid methyl esters by argentation thin-layer chromatography J. Chromatogr. A 905, 251-257.

(27) Pezzella, A., Lista, L., Napolitano, A., and d’Ischia, M. (2005) Anexpedient one-pot entry to catecholestrogens and other catecholcompounds via IBX-mediated phenolic oxygenation. TetrahedronLett. 46, 3541-3544.

(28) Convert, O., Van Aerden, C., Debrauwer, L., Rathahao, E.,Molines, H., Fournier, F., Tabet, J. C., and Paris, A. (2002)

Reactions of estradiol-2,3-quinone with deoxyribonucleosides:Possible insights in the reactivity of estrogen quinones with DNA.Chem. Res. Toxicol. 15, 754-764.

(29) Zhang, F., and Bolton, J. L. (1999) Synthesis of the equineestrogen metabolites 2-hydroxyequilin and 2-hydroxyequilenin.Chem. Res. Toxicol. 12, 200-203.

(30) Jackman, L. M., and Sternhell, S. (1969) Theory of chemical effectsin NMR spectroscopy. In Applications of Nuclear MagneticResonance Spectroscopy in Organic Chemistry, II ed., pp 94-95,Pergamon Press, Oxford.

(31) Tabakovic, K., Gleason, W., Ojala, W. H., and Abul-Hajj, Y. J.(1996) Oxidative transformation of 2-hydroxyestrone. Stabilityand reactivity of 2,3-estrone quinone and its relationship toestrogen carcinogenicity. Chem. Res. Toxicol. 9, 860-865.

(32) Aplin, R. T., and Pike, W. T. (1966) Mass spectra of 1,4-dihydroxybenzenes and p-benzoquinones. Quinol-quinone inter-conversion in the heated inlet system of a mass spectrometer.Chem. Ind. (London) 48, 2009-2012.

(33) Sugumaran, M., and Bolton, J. (1995) Direct evidence for quinone-quinone methide tautomerism during tyrosinase catalyzed oxida-tion of 4-alkylcatechol. Biochem. Biophys. Res. Commu. 213, 469-474.

(34) Bolton, J. L., and Shen, L. (1996) p-Quinone methides are themajor decomposition products of catechol estrogen o-quinones.Carcinogenesis 17, 925-929.

(35) Sugumaran, M., and Bolton, J. L. (1998) Laccasesand nottyrosinasesis the enzyme responsible for quinone methide pro-duction from 2,6-dimethoxy-4-allyl phenol. Arch. Biochem. Bio-phys. 353, 207-212.

(36) Espin, J. C., and Wichers, H. J. (1999) Slow-binding inhibition ofmushroom agaricus bisporus tyrosinase isoforms by tropolone. J.Agric. Food. Chem. 7, 2638-2644.

(37) Becker, H. D., and Turner, A. B. (1988) Quinones as oxidants anddehydrogenating agents. Chem. Quinonoid Compd. 2, 1351-1384.

(38) Horak, V., Foster, F. V., De Levie, R., Jones, J. W., and Svoronos,P. (1981) Generation and trapping of 2,3-naphthoquinone.Tetrahedron Lett. 22, 3577-3578.

(39) Menting, K. H., Eichel, W., Schmand, H. L. K., and Boldt, P.(1983) Quinones. Part 9. Stability of quinones toward water.Synthesis of a 1,7-naphthoquinone. J. Org. Chem. 48, 2814-2820.

(40) Liu, X., Zhang, F., Liu, H., Burdette, J. E., Li, Y., Overk, C. R.,Pisha, E., Yao, J., Van Breemen, R. B., Swanson, S. M., andBolton, J. L. (2003) Effect of halogenated substituents on themetabolism and estrogenic effects of the equine estrogen,equilenin. Chem. Res. Toxicol. 16, 741-749.

TX050060O

Tyrosinase-Promoted Oxidation of Estradiol Chem. Res. Toxicol., Vol. 18, No. 9, 2005 1419