In Vitro Formation of Quinoid Metabolites of the Dietary Supplement

Botanical dietary supplements containing Cimicifuga racemosa (Actaea racemosa; ... dietary supplement containing black cohosh, this study found no cau...
2 downloads 0 Views
Download PDF
838

Chem. Res. Toxicol. 2003, 16, 838-846

In Vitro Formation of Quinoid Metabolites of the Dietary Supplement Cimicifuga racemosa (Black Cohosh) Benjamin M. Johnson and Richard B. van Breemen* Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago, Illinois 60612-7231, and UIC/NIH Center for Botanical Dietary Supplements Research, College of Pharmacy, University of Illinois at Chicago, 833 South Woods Street, Chicago, Illinois 60612-7231 Received December 3, 2002

Botanical dietary supplements containing Cimicifuga racemosa (Actaea racemosa; black cohosh) are used commonly by women to assuage menopausal symptoms including hot flashes and sleep disorders. Despite the popularity of such supplements, little is known about the metabolism or possible toxicity of many compounds that could be concentrated therein. The aim of this study was to selectively identify phase I metabolites resulting from metabolic bioactivation of constituents of black cohosh in vitro and to determine whether evidence of such metabolites could be found in the urine of perimenopausal women taking black cohosh oral supplements. A variation of an ultrafiltration mass spectrometric assay devised previously was used to screen an extract of black cohosh for the formation of electrophilic phase I metabolites that had been trapped as GSH conjugates. Mercapturates (N-acetylcysteine conjugates) corresponding to the GSH conjugates identified during screening were synthesized and characterized using LC-MS/MS with product-ion scanning. During a phase I clinical trial of black cohosh in perimenopausal women, urine was collected from seven subjects, each of whom took a single oral dose of either 32, 64, or 128 mg of the black cohosh extract. These urine samples were analyzed for the presence of mercapturate conjugates using positive-ion electrospray LC-MS and LC-MS/MS. On the basis of their propensity to form GSH adducts following metabolic activation by hepatic microsomes and NADPH in vitro, a total of eight electrophilic metabolites of black cohosh were detected, including quinoid metabolites of fukinolic acid, fukiic acid, caffeic acid, and cimiracemate B. Additional quinoid metabolites were formed from hydroxytyrosol and dihydroxyphenyl lactic acid, neither of which had been isolated previously from black cohosh. However, mercapturate conjugates of these black cohosh constituents were not detected in urine samples from women who consumed single oral doses of up to 256 mg of a standardized black cohosh extract. Therefore, for moderate doses of a dietary supplement containing black cohosh, this study found no cause for safety concerns over the formation of quinoid metabolites in women.

Introduction In July 2002, a large clinical study of the risks and benefits of conjugated equine estrogens (0.625 mg/d) plus medroxyprogesterone acetate (2.5 mg/d) for the management of menopause was halted, 3.3 years ahead of schedule, due to an increased incidence of invasive breast cancer in the group that received hormone replacement therapy vs placebo. Pursuant to these results, the Women’s Health Initiative concluded that this course of therapy was not a safe option for women seeking relief of perimenopausal symptoms (1). In a separate clinical study, hormone therapy was found to have no clinically meaningful effect on health-related quality of life (2). Renewed public concern regarding the safety and efficacy of hormone replacement therapy is likely to create higher demand for alternatives to such therapy, including the use of botanical dietary supplements. Clinical evidence indicates that Cimicifuga racemosa (Actaea racemosa, * To whom correspondence should be addressed. Tel: (312)996-9353. Fax: (312)996-7107. E-mail: [email protected].

black cohosh) might provide relief for menopausal symptoms, including hot flashes, night sweats, insomnia, and anxiety (3-13). Despite its popularity and apparent efficacy, evidence regarding the safety of black cohosh in women remains inconclusive. A review of eight clinical trials concluded that extracts of the rhizome of black cohosh might be a safe alternative for women seeking alternatives to estrogen replacement therapy (14). Chronic toxicity was not observed in rats at a dose of 500 mg/kg/ day for 27 weeks or in dogs at 400 mg/kg/day for 26 weeks. Furthermore, a 40% 2-propanol extract of black cohosh was negative in the Ames test (15). However, adverse side effects, including nausea, vomiting, headaches, and dizziness, have been reported during clinical trials (6, 10-12). In mice, the LD50 of a black cohosh extract is 7.7 g/kg for intragastric administration and 1.1 g/kg for intravenous administration (15). Additionally, several compounds with untoward biological activities have been isolated from black cohosh. Actein and cimigenol are immunosuppressive in PHA-stimulated lymphocytes in vitro (16), and cimicifugoside has immunosuppressive, cytotoxic, and anticytotoxic properties (17, 18).

10.1021/tx020108n CCC: $25.00 © 2003 American Chemical Society Published on Web 05/21/2003

Formation of Quinoid Metabolites of Black Cohosh

Chem. Res. Toxicol., Vol. 16, No. 7, 2003 839

Figure 1. Scheme showing the ultrafiltration LC-MS/MS screening assay for the selective detection of electrophilic phase I metabolites formed from a botanical extract.

Black cohosh also contains several catechols, such as caffeic acid and piscidic acid, and fukiic acid esters that exhibit antioxidant properties, including fukinolic acid, cimicifugic acid A, and cimicifugic acid B (19). Such catechols are of significant concern in toxicology because of the possibility that they might be activated, either metabolically or chemically, to electrophilic quinones. The potential of such quinones to cause toxicity and carcinogenesis is well-documented, and might occur via arylation of cellular proteins and DNA or redox cycling leading to the formation of reactive oxygen species such as the hydroxyl radical (20). Caffeic acid is well-tolerated by humans but can induce forestomach squamous cell papillomas and carcinomas and renal tubular cell hyperplasias and adenomas in rats and mice (21). The antioxidant hydroxytyrosol, despite its benefits as an inhibitor of platelet aggregation and eicosanoid formation (22), was shown to promote DNA damage in the bleomycin-Fe(III) system (23). A pulsed ultrafiltration LC-MS/MS assay was recently developed to screen botanical extracts for the formation of GSH conjugates of electrophilic phase I metabolites, including quinones and reactive epoxides (24). In a process that mimics a common in vivo detoxification pathway, GSH traps electrophilic metabolites immediately following their formation by dexamethasoneinduced hepatic microsomes in the presence of NADPH. The stable GSH conjugates are then separated from the microsomal protein using ultrafiltration and subjected to positive-ion electrospray LC-MS/MS analysis using precursor-ion scanning. Precursor-ion scanning is used to detect GSH conjugates that fragment during collisioninduced dissociation to give product ions of m/z 130, corresponding to the protonated γ-glutamyl moiety of GSH. Hence, this method allows complex chemical mixtures, such as botanical extracts, combinatorial libraries, and their metabolites, to be screened for the appearance of GSH conjugates. Once such metabolites are detected, electrospray LC-MS/MS with product-ion scanning may be used to help elucidate their structures. When considered alongside output from the NAPRALERT database (25), which lists compounds that have been isolated from natural products and especially botanical sources, these tandem mass spectra frequently allow the identification of botanical compounds from which the electrophilic metabolites were formed. The aim of the present study is to determine whether constituents of a black cohosh extract are metabolically activated to electrophilic phase I metabolites. The in vitro screening assay, described above, is used to identify and characterize any such metabolites. Additionally, this

LC-MS/MS approach is extended to determine whether mercapturic acids (N-acetylcysteine (NAC) conjugates), the products formed during metabolism of GSH conjugates in the blood and kidney, can be recovered from the urine of women who consume black cohosh. If detected, mercapturic acids would provide evidence for the in vivo metabolic activation of compounds from black cohosh consumed at potentially therapeutic doses.

Experimental Procedures The 70% ethanol extract of black cohosh used in this study (the extract) was produced by PureWorld Botanicals (South Hackensack, NJ). All reagents and cofactors were purchased from Sigma Chemical Co. (St. Louis, MO), and all solvents were from Fisher Scientific (Pittsburgh, PA) and were HPLC grade or better. Stock solutions of 10 mM GSH, NADPH, and NAC in deoxygenated water were prepared in advance and stored in 100 µL aliquots at -20 °C until use. Dexamethasone-induced female Sprague-Dawley rat liver microsomes were prepared as reported previously (26) and stored at -80 °C until use. Screening Assay for GSH Conjugates of Electrophilic Phase I Metabolites. The screening assay for GSH conjugates of activated phase I metabolites was carried out as described previously (24) with minor modifications. A scheme of the ultrafiltration LC-MS/MS screening assay is shown in Figure 1. Briefly, 1 mg of the extract was dissolved in 300 µL of 50 mM phosphate buffer (pH 7.4) with vortexing and sonication. Next, ∼1 mg of microsomal protein was added followed by single 100 µL aliquots of NADPH and GSH, and the preparation was incubated at 37 °C for 1 h with intermittent shaking. Samples were then filtered at 12 000g using a Microcon centrifugal filter device (Millipore Corp., Bedford, MA) with a regenerated cellulose 30 000 MWCO membrane for 1 h at 4 °C to remove protein. The ultrafiltrate was stored at 4 °C until analysis within 24 h. Chromatographic separations of 15 µL aliquots of the ultrafiltrate were carried out using a Waters (Milford, MA) Alliance 2690 HPLC system with an Xterra MS C18 (2.1 mm × 150 mm) column. The solvent system consisted of a linear gradient (gradient 1) from 0.5% formic acid to acetonitrile at 120 µL/ min as follows: 5% acetonitrile for 5 min (eluent diverted to waste), 5-20% over 15 min, 20% isocratic for 10 min, 20-90% over 10 min, and 90% isocratic acetonitrile for 15 min. The HPLC was interfaced with a Micromass (Manchester, U.K.) Quattro II triple quadrupole mass spectrometer. During LC-MS/MS, positive-ion electrospray ionization was used, and precursor-ion tandem mass spectra were recorded over the ranges of m/z 400-750 and m/z 750-1100. Precursor-ion scanning was used to identify ions that fragmented to form products of m/z 130 corresponding to the b1 fragment ion of GSH. The collision energy was 24 eV, and the argon collision pressure was 1 µBar. Characterization of Quinoid Metabolites. The NAPRALERT database was searched to generate a list of

840

Chem. Res. Toxicol., Vol. 16, No. 7, 2003

all compounds that have been isolated from black cohosh, and the Beilstein CrossFire database was searched for a list of all previously isolated natural products with masses corresponding to those observed during the screening assay for GSH conjugates of electrophilic metabolites. Although the triple quadrupole mass spectrometer is ideal for precursor-ion scanning, subsequent LC-MSn experiments were carried out using a ThermoFinnigan (San Jose, CA) LCQ Deca ion trap mass spectrometer equipped with electrospray and a Surveyor HPLC system. This instrument was used to provide multiple stages of tandem mass spectrometry for structure confirmation. During positive-ion electrospray LC-MSn product-ion analysis of GSH conjugates, the gradient 1 solvent system was used. The heated capillary temperature was maintained at 350 °C, and the sheath and auxiliary nitrogen gas flow rates were 60 and 10 (arbitrary units), respectively. During LC-MSn product-ion scanning of GSH adducts (where n > 2), the y2 GSH fragment ions were selected for subsequent collision-induced dissociation. A 1 µg sample of the black cohosh extract was analyzed using negative-ion electrospray LC-MS/MS for the detection and identification of catechols that served as metabolic precursors of the GSH adducts detected during the ultrafiltration LC-MS/ MS screening assay. A standard of the catechol hydroxytyrosol was synthesized via the reduction of (3,4-dihydroxyphenyl)acetic acid as reported previously (27). For the catechol analysis, the mobile phase consisted of a linear gradient (gradient 2) from water to acetonitrile at 120 µL/min as follows: 5% acetonitrile for 5 min (eluent diverted to waste), 20% isocratic for 20 min, 20-90% over 15 min, and 90% isocratic acetonitrile for 15 min. The negative-ion electrospray parameters were as follows: spray voltage, -5 kV; capillary voltage, -4 V; tube lens offset, 15 V; capillary temperature, 350 °C; sheath gas flow rate, 60 (arbitrary units); and auxiliary gas flow rate, 10. If LC-MSn analyses of a GSH adduct and its corresponding catechol were insufficient for structure determination, then the cysteamine adduct of the same compound was prepared and analyzed. Cysteamine conjugates were synthesized using the same procedure that was used during the preparation of GSH conjugates, except that approximately 100 µg of solid cysteamine was added to the incubation mixture in place of GSH, and approximately 1000 u of tyrosinase was substituted for the hepatic microsomes. During LC-MSn analysis, the solvent system (gradient 3) was identical to the gradient 2 used for catechol analysis except that 0.5% formic acid was substituted for water. Synthesis of Mercapturic Acids. In preparation for LCMS/MS analysis of clinical urine samples, catechols from black cohosh were derivatized with NAC and characterized structurally. The procedure to form NAC adducts of the catechols was analogous to that used for GSH conjugation, except that NAC was substituted for GSH, the 50 mM phosphate buffer was used with a pH of 6.7, and approximately 1000 u of mushroom tyrosinase was added to the incubation mixture to promote quinone formation. Whenever possible, derivatization was also carried out using an authentic standard (caffeic acid) or a botanical fraction that was enriched in the compound of interest (fukinolic acid, cimiracemate B). Mass spectrometric tuning parameters were optimized using 1 µM NAC dissolved in 0.5% formic acid/20% acetonitrile at a flow rate of 120 µL/min. Selection and Treatment of Human Subjects. Six healthy, perimenopausal women, 45-59 years of age, were administered a single dose of either 32, 64, or 128 mg of the black cohosh extract. All subjects gave informed consent, and the human subjects protocol was approved by the UIC Institutional Review Board and the Scientific Advisory Committee of the UIC General Clinical Research Center. Women were excluded from the study using the following criteria: smoker, use of any prescription medicine within the last 2 months, body mass index > 30% above ideal, previous history of breast or reproductive cancer, consumption of more than five alcoholic beverages/week, chronic disease such as diabetes or hypertension, concurrent nondietary phytoestrogens or hormone use, vegan, concurrent participation

Johnson and van Breemen in other clinical trial(s), or unavailable for follow-up visit (28). A urine sample was obtained from each subject over the course of 24 h following oral administration of a single dose of the black cohosh extract. The urine samples were stored at -80 °C prior to analysis. As a positive control for the recovery and detection of mercapturic acids from urine, a healthy 27 year old male, who consumed approximately one alcoholic beverage/day during the month prior to treatment, was given a single oral dose of 500 mg of acetaminophen. Urine samples from this control subject were collected immediately before, and 8 h after, administration of the drug. Analysis of Urine Samples Using LC-MS/MS. Urine samples (10 mL) from each of the subjects were thawed and acidified with 100 µL of concentrated formic acid. RESPREP Drug Prep I solid phase extraction cartridges (Restek; Bellefonte, PA) were conditioned with 3 mL of 0.5% formic acid and then used to extract the urine samples. Following extraction, each cartridge was washed with 3 mL of 0.5% formic acid and eluted with 3 mL of formic acid/water/methanol (0.5:10:90, v/v/ v). The eluates were then evaporated to dryness under vacuum. After reconstitution in 200 µL of water, 10 µL aliquots were analyzed using positive-ion electrospray LC-MS/MS with solvent system gradient 1 and product-ion scanning for the selective detection of mercapturic acids. To maintain MS/MS detection efficiency, no more than two metabolites were monitored per analysis. To demonstrate the recovery of mercapturic acids using this extraction procedure, a blank urine sample (8 mL), which was collected from a human subject following a 12 h fast, was spiked with an ultrafiltrate (2 mL) from an incubation containing 1 mM caffeic acid and NAC and 1000 u of mushroom tyrosinase. The spiked urine sample was extracted as outlined above. An additional 2 mL aliquot of the ultrafiltrate was evaporated to dryness using a rotary evaporator and reconstituted in 200 µL of water. Positive-ion electrospray LC-MS analyses were carried out in triplicate for each of the two samples using the LCQ DECA ion trap mass spectrometer, and the peak in each chromatogram corresponding to the most abundant mercapturic acid conjugate of caffeic acid (m/z 342; retention time, 16.5 min) was integrated. For the analysis of the mercapturate conjugate of caffeic acid, the mobile phase consisted of a linear gradient (gradient 4) from 0.1% formic acid to acetonitrile at 120 µL/ min as follows: 5% acetonitrile for 5 min (eluent diverted to waste), 20-90% over 15 min, and 90% isocratic acetonitrile for 15 min. The percent recovery was calculated by dividing the average peak area of the mercapturic acid conjugate of caffeic acid in the extracted urine sample by the corresponding average peak area of the reconstituted ultrafiltrate.

Results The ethanolic extract of black cohosh was screened for the formation of reactive metabolites that could be detected as GSH conjugates during ultrafiltration LC-MS/MS screening. The LC-MS/MS chromatogram obtained using precursor-ion scanning to detect distinctive GSH fragment ions of m/z 130 is shown in Figure 2. Multiple new peaks were detected in the incubation containing active rat liver microsomes as compared to the control containing no microsomes. The one significant peak in the control chromatogram, which is detected at 12.0 min, was probably the result of an autoxidation product that reacted with GSH, since the intensity of this peak was ∼5-fold greater in the microsomal incubation. To confirm these peaks as GSH conjugates, LC-MS/MS analyses of these samples were repeated using positiveion electrospray and CID with product-ion scanning. Except for the ions of m/z 523 and 530, each of these precursor ions formed product ions corresponding to a neutral loss of 129 u, which is characteristic of the y2

Formation of Quinoid Metabolites of Black Cohosh

Chem. Res. Toxicol., Vol. 16, No. 7, 2003 841

Figure 2. Positive-ion electrospray LC-MS/MS analysis for precursors of the GSH fragment ion of m/z 130 in the ultrafiltrate from a black cohosh incubation. The collision energy was 24 eV, and the argon collision gas pressure was 1 µbar. (A) Control in which microsomes were omitted from the incubation mixture so that no metabolic bioactivation could occur. (B) Analysis of an ultrafiltrate from an incubation of an extract of black cohosh, GSH, rat liver microsomes, and NADPH.

fragment ion of GSH adducts. Other product ions were observed corresponding to GSH conjugates including b2 and z2 ions. A search of the NAPRALERT database for compounds previously isolated from black cohosh indicated that the protonated GSH conjugates of m/z 486, 740, and 578 (see Figure 2) might correspond to the black cohosh constituents caffeic acid, fukinolic acid, and fukiic acid, respectively. These assignments were verified on the basis of LC-MS/MS and LC-MS3 experiments using product-ion scanning. In the first case, the retention times of the GSH adducts and the fragmentation pattern of the m/z 486 precursor ions matched those of GSH conjugates generated using an authentic caffeic acid standard (not shown). The presence of multiple chromatographic peaks for the ions of m/z 486 in both the black cohosh ultrafiltrate and the activated caffeic acid standard indicates that multiple isomers of the GSH adducts were formed. In the next example, the positive-ion electrospray MS/MS production spectra of the fukinolic acid-GSH conjugates of m/z 740 were sufficient by themselves to allow a positive identification. Although only one such conjugate was detected in the screening assay using precursor-ion scanning (Figure 2), a total of three fukinolic acid-GSH regioisomers were observed in the LC-MS/MS production chromatogram, and product-ion mass spectra of each of these three isomers are shown in Figure 3. The product ions of m/z 578 in the MS/MS spectrum of fukinolic acidGSH are consistent with a fragmentation pattern involving cleavage of the ester moiety of fukinolic acid. In the case of fukiic acid, an additional higher energy collisioninduced dissociation step, MS3, was necessary to produce unique fragment ions for identification. Figure 4 shows the positive-ion electrospray LC-MS3 product-ion analysis of the fukiic acid-GSH adduct, where the g type internal peptide fragment ion RSCH2CHNH2+ of m/z 346 was selected for additional CID and a third stage of MS. In the g type ion of a GSH adduct, most of the labile GSH bonds have already been eliminated leaving a charged imino group. Subsequent fragmentation of this ion gives product ions, most of which are formed from the moiety belonging to the botanical compound. Accordingly, the spectrum in Figure 4 is consistent with the g type fragment ion of fukiic acid-GSH, which contains a series of alcohol and acidic functional groups.

Although not yet in the NAPRALERT database, ultrafiltration LC-MS/MS indicated that a deesterified quinoid metabolite of cimiracemate B formed a GSH conjugate with a retention time of 10.1 min (see Figure 2) and a protonated molecule of m/z 488. The identification of cimiracemate B in black cohosh was reported recently (29). The GSH adduct of m/z 488 was probably formed by conjugation to the 2-keto-3-(3,4-dihydroxyphenyl)propan-1-ol moiety of cimiracemate B following ester hydrolysis and activation to an o-quinone. The fragmentation pattern and retention time of the GSH conjugate of m/z 488 at 10.2 min in Figure 2 are identical to those obtained during the ultrafiltration LC-MS/MS analysis of a botanical fraction that was enriched in cimiracemate B. This fraction was prepared during the isolation of cimiracemate B from black cohosh (29). It should be noted that ultrafiltration LC-MS/MS screening of reference cimiracemate B did not produce an ion of m/z 486 at 20.0 or 21.0 min (Figure 2), which indicates that the ion of m/z 486 was not formed by metabolism or oxidation of cimiracemate B. Identification of the GSH conjugate of m/z 460 eluting at 13.0 min in Figure 2 was accomplished via direct analysis of the black cohosh extract. Negative-ion electrospray LC-MS/MS with product-ion scanning was used to search the original extract for the presence of a deprotonated molecule corresponding to a catechol of m/z 153, and the results are shown in Figure 5. The fragmentation pattern and retention time of the ion detected at 13.3 min were identical to that of an authentic standard of hydroxytyrosol, which also formed the same GSH adduct of m/z 460 during ultrafiltration LC-MS/MS screening. The loss of 30 u during CID (Figure 5) is characteristic of primary alcohols during negative-ion electrospray tandem mass spectrometry. To our knowledge, this is the first time hydroxytyrosol has been identified in black cohosh. Although a similar analysis using negative-ion electrospray LC-MS/MS for m/z 197 indicated the presence of another catechol that might correspond to the GSH conjugate of m/z 504 eluting at 13.7 min in Figure 2, ionization of this compound was partially suppressed by other botanical compounds eluting simultaneously, resulting in poor sensitivity. To enhance ionization efficiency and overcome this problem, activated metabolites

842

Chem. Res. Toxicol., Vol. 16, No. 7, 2003

Johnson and van Breemen

Figure 3. Positive-ion electrospray product-ion scans of the protonated molecules of m/z 740 corresponding to three fukinolic acidGSH regioisomers. Spectra A and C show GSH attachment to the fukiic acid moiety of fukinolic acid, while spectrum B shows attachment to the caffeic acid moiety. The normalized collision energy for all spectra is 28%. Part D shows the proposed structure of the most abundant fukinolic acid-GSH adduct (observed in spectrum C), in which the point of GSH attachment on the fukinolic acid ortho-quinone is the most electrophilic and least sterically hindered carbon in the aromatic ring.

Figure 4. Positive-ion electrospray LC-MS3 analysis of the fukiic acid-GSH conjugate eluting at 12.3 min in Figure 3. During MS3, the precursor-ion of m/z 578 was fragmented during CID to produce the g type peptide ion of m/z 346, which was selected as a precursor ion for CID in a third stage of MS. The normalized collision energies were 23 and 28%, respectively.

in the black cohosh extract were derivatized with cysteamine instead of GSH and analyzed using positive-ion electrospray. The fragmentation pattern and structure of a cysteamine adduct of m/z 274 formed during the incubation of a black cohosh extract with hepatic microsomes, NADPH and cysteamine, is shown in Figure 6. The retention time of this analyte (36.2 min) matches that of an adduct of (3,4-dihydroxyphenyl)lactate (DHPL) that is formed during the metabolism of rosmarinic acid in the presence of hepatic microsomes, NADPH and cysteamine. [Rosmarinic acid can undergo ester hydrolysis followed by metabolic oxidation of the remaining DHPL moiety in the presence of hepatic microsomes.

Figure 5. Negative-ion electrospray LC-MS/MS analysis of an extract of black cohosh for product ions of m/z 153. (A) LC chromatogram showing elution of deprotonated molecules of m/z 153. (B) MS/MS product-ion analysis of the ion of m/z 153 detected at 13.3 min (note that the MS/MS spectrum of the peak at 35.1 min was identical). During CID, loss of CH2O resulted in the formation of a product ion of m/z 123.0. The peak at 13.3 min was determined to correspond to hydroxytyrosol based on comparison to an authentic standard.

Accordingly, GSH conjugates have been observed during studies of botanical products that contain rosmarinic acid (24).] In addition to the MS/MS spectrum of the cysteamine adduct shown in Figure 6, positive-ion electrospray LC-MS3 analyses were carried out on the abundant fragment ions of m/z 256 and 257 (data not shown), which themselves were formed by elimination of water or

Formation of Quinoid Metabolites of Black Cohosh

Figure 6. (A) Positive-ion electrospray LC-MS/MS analysis of a preparation containing black cohosh, hepatic microsomes, NADPH, and cysteamine. Regioisomers of the DHPL-cysteamine conjugate eluted at 35.5, 36.2, and 37.0 min during HPLC. (B) Product ions are shown for the DHPL-cysteamine adduct, which eluted at 36.2 min. The ions of m/z 214, 256, and 257 were formed by the elimination of small neutral molecules of C2H4S, H2O, and NH3, respectively. The normalized collision energy was 34%.

ammonia, respectively. Although the ion of m/z 214 is the base peak in the MS/MS spectrum of m/z 274 (see Figure 6), it was not observed in the MS3 spectra. Therefore, the origin of the ion of m/z 214 was probably the neutral loss of C2H4S from the protonated molecule of m/z 274, accompanied by an intramolecular migration and retention of the amino group. Although DHPL is a substructure of fukiic acid and all fukiic acid esters, this is the first report of the presence of DHPL in black cohosh. The structures of DHPL and the other catechols in black cohosh that formed quinoid metabolites during screening are shown in Figure 7. During MS/MS and MS3 analysis of the GSH conjugates of m/z 442 and 682, unique product ions sufficient for structural determination of the botanical moieties were not detected. If the botanical compounds or metabolites giving rise to these conjugates were catechols, then their masses would be 136 and 376 u, respectively. The abundance of triterpenoids in black cohosh with masses in the range of 376 u, such as 27-deoxyactein, suggests that the GSH adduct m/z 682 might be formed from a quinoid or epoxide metabolite of a botanical compound with a triterpene-like structure. Mercapturic acids corresponding to the GSH conjugates observed in Figure 2, with the exception of the GSH adducts of m/z 442 and 682, were synthesized using NAC as a trapping agent. In general, the MS/MS product-ion spectra of these mercapturic acids were characterized by neutral losses of water and C2H2O (from the acetyl group) and a product ion of m/z 130 corresponding to the neutral loss of RSH. An example is shown in Figure 8 in which

Chem. Res. Toxicol., Vol. 16, No. 7, 2003 843

Figure 7. Catechols from black cohosh that could be activated to quinoid metabolites include caffeic acid (1), fukiic acid (2), fukinolic acid (3), cimiracemate B (4), DHPL (5), and hydroxytyrosol (6).

Figure 8. Positive-ion electrospray MS/MS product-ion mass spectrum of m/z 344, corresponding to the protonated molecule of a mercapturate that was biosynthesized using a botanical fraction enriched in cimiracemate B. The peaks of m/z 284, 266, and 248 were formed by the successive loss of water molecules from the fragment ion of m/z 302 during CID. The normalized collision energy was 30%.

cimiracemate B has undergone ester hydrolysis and metabolic oxidation of the 2-keto-3-(3,4-dihydroxyphenyl)propan-1-ol moiety, followed by conjugation to NAC. Although in vitro screening assays had demonstrated the potential for the formation of several GSH conjugates of activated metabolites of black cohosh, none of the corresponding mercapturic acids were detected in the urine of women during the first 24 h after consuming a black cohosh extract. To verify the method for the detection of urinary mercapturic acids, a positive control assay was carried out on the urine of a healthy man following the oral administration of a single 500 mg dose of acetami-

844

Chem. Res. Toxicol., Vol. 16, No. 7, 2003

Johnson and van Breemen

Figure 9. Scheme showing the bioactivation and detoxication of acetaminophen in vivo. Acetaminophen is oxidized by cytochrome P4502E1 to give a quinone imine, which undergoes nuclephilic attack in the presence of GSH. The hepatic enzymes cysteinylglycine S-conjugate dipeptidase and γ-glytamyl transpeptidase convert the GSH adduct to a Cys adduct, and subsequent acetylation of the Cys adduct in the kidney produces the observed urinary NAC conjugate.

nophen. As shown in Figure 9, acetaminophen can be oxidized to a reactive quinone imine (30, 31), which can undergo nucleophilic attack by GSH and be metabolized subsequently to a mercapturic acid. An abundant LC-MS/MS signal for an acetaminophen-mercapturic acid was detected at 20.5 min (Figure 10) in urine obtained after acetaminophen administration but not in control urine collected before administration. The corresponding tandem mass spectrum shown in Figure 10 is consistent with this structure. Mass spectrometric characterization of mercapturic acid conjugates of acetaminophen has been carried out previously using multiple ionization methods including electron impact, chemical ionization, and field desorption (32). To further verify the extraction procedure, a blank urine sample was spiked with an ultrafiltrate of a reaction mixture containing a mercapturate conjugate of caffeic acid, and both the ultrafiltrate and the extracted urine sample were analyzed using LC-MS. Comparison of the peaks corresponding to the mercapturate conjugate of caffeic acid in the two samples indicated that the recovery of this metabolite from the urine sample using solid phase extraction was approximately 81%.

Discussion The methods used in this study represent a general approach to the screening of botanical natural products for possible toxicity due to the formation of electrophilic phase I metabolites, such as quinones or epoxides. Reactive metabolites that are generated by hepatic microsomes in vitro are trapped using a biological nucleophile (such as GSH) that is detected selectively during mass spectrometric screening analysis using either precursor-ion scanning or constant neutral-loss scanning. Mercapturic acids that correspond to these metabolites are synthesized and characterized using LC-MS/MS with product-ion scanning as reference standards for subsequent LC-MS/MS analysis of urine samples. The detection of such mercapturic acids in urine would provide evidence for the formation of corresponding reactive

Figure 10. As a positive control for the detection of urinary mercapturic acids, urine samples from a healthy man were analyzed using positive-ion electrospray LC-MS/MS with production scanning. Urine samples were collected (A) immediately before and (B) 8 h after the patient took an oral dose of 500 mg of acetaminophen. (C) Product-ion MS/MS spectrum of the ion of m/z 313 confirming the identification of the peak eluting at 20.5 min in part B as an acetaminophen-mercapturic acid. The product ion of m/z 271 was formed by the elimination of C2H2O from the acetyl group, and the product ion of m/z 184 corresponds to acetaminophen with a thiol group attached to the aromatic ring. The normalized collision energy was 24%.

metabolites in vivo. Similar strategies for the selective detection of GSH conjugates and mercapturic acids have been employed previously (33, 34). Although our in vitro assay indicated the possibility of formation of reactive quinoid metabolites from various constituents of black cohosh, no corresponding mercapturic acids were found in the urine of women who took black cohosh during our clinical trial. There are several possible explanations for the absence of mercapturic acids in the urine of women who took black cohosh. For example, the botanical compounds that were identified during the screening assay for activated phase I metabolites might not have been absorbed from the gut following administration, or alternative metabolic pathways including sulfate and glucuronic acid conjugation might have predominated during hepatic first pass metabolism. Alternatively, the amounts of these compounds that were administered during treatment might have been too small to produce quantities of GSH conjugates sufficient for detection. Finally, the botanical compounds, which formed GSH conjugates following in vitro oxidation by dexamethasone-induced rat liver microsomes, might not be substrates for metabolic oxidation by P450 isozymes in humans. It is interesting to note that several of the botanical catechols discussed here have been identified previously as antioxidantssa term that carries a positive connota-

Formation of Quinoid Metabolites of Black Cohosh

tion in the context of public health and nutrition. However, identification of these compounds during the screening assay for GSH conjugates of electrophilic phase I metabolites suggests that some phenolic antioxidants might also be oxidized to electrophilic and potentially toxic compounds. Thus, the formation of these electrophilic oxidation products might be either catalyzed (e.g., activation by cytochrome P450 enzymes) or uncatalyzed (e.g., oxidation during the quenching of a reactive oxygen species). The former pathway might create compounds capable of redox cycling, leading to the production of additional reactive oxygen species including hydrogen peroxide and the hydroxyl radical. Therefore, phenolic antioxidants that are substrates for metabolic oxidation by cytochromes P450, or other oxidizing enzymes, might actually potentiate oxidative stress in tissues where these enzymes are abundant. Observations to this effect have been made previously (35). The structure of cimiracemate B is worthy of an additional note. Although the regioselectivity of the reaction between the oxidized 2-keto-3-(3,4-dihydroxyphenyl)propan-1-ol moiety and the GSH was not investigated in this study, the carbonyl group β to the aromatic ring would help to conjugate the corresponding quinone methide. If this metabolite exists and exhibits intermediate stability, it might be capable of alkylating cellular material that is located at considerable distance from its site of formation (36), thereby increasing the likelihood that acute toxicity might be associated with cimiracemate B at high doses. Therefore, the metabolism of this compound will be the subject of additional study. Despite the results of our in vitro screening analysis, our clinical data did not indicate the in vivo formation of any reactive metabolites of black cohosh constituents at the doses used in this study. On the basis of this evidence, dietary supplements prepared from black cohosh should not form electrophilic phase I metabolites when used in moderation. However, the in vitro data demonstrate the possibility of metabolic activation of black cohosh constituents that might be a safety concern at high doses.

Acknowledgment. We thank Aleksej Krunic for assistance with organic synthesis; Dejan Nikolic for assistance with the interpretation of mass spectra; Suzanne Banuvar, Stacie E. Geller, and Lee Schulman for acquiring clinical samples; and ThermoFinnigan for providing the ion trap LC-MS instrumentation used in this study. The clinical studies were supported in part by the General Clinical Research Center at the University of Illinois at Chicago, which is funded by NIH Grant M01 RR13987. This project was also funded by Grant P50 AT00155 provided to the UIC/NIH Center for Botanical Dietary Supplements Research by the Office of Dietary Supplements, the National Institute of General Medical Sciences, the Office for Research on Women’s Health, and the National Center for Complementary and Alternative Medicine. Its contents are the responsibility of the authors and do not necessarily represent the official views of the sponsors.

References (1) Rossouw, J. E., Anderson, G. L., Prentice, R. L., LaCroix A. Z., Kooperberg, C., Stefanick, M. L., Jackson, R. D., Beresford, S. A., Howard, B. V., Johnson, K. C., Kotchen, J. M., Ockene, J., and Writing Group for the Women’s Health Initiative Investigators (2002) Risks and benefits of estrogen plus progestin in healthy postmenopausal women, principal results from the

Chem. Res. Toxicol., Vol. 16, No. 7, 2003 845

(2)

(3) (4) (5)

(6) (7)

(8) (9) (10) (11) (12) (13) (14) (15) (16)

(17)

(18) (19) (20) (21)

(22)

(23)

(24) (25) (26)

(27)

women’s health initiative randomized controlled trial. J. Am. Med. Assoc. 288, 321-332. Hays, J., Ockene, J. K., Brunner, R. L., Kotchen, J. M., Manson, J. E., Patterson, R. E., Aragaki, A. K., Shumaker, S. A., Brzyski, R. G., LaCroix, A. Z., Granek, I. A., and Valanis, B. G. (2003) Effects of estrogen plus progestin on health-related quality of life. N. Engl. J. Med. In press. Mahady, G. B., Fong, H. H. S., and Farnsworth, N. R. (2001) Botanical Dietary Supplements: Quality, Safety and Efficacy, Swets & Zeitlinger B. V., Lisse, The Netherlands. Daiber, W. (1983) Klimakterische Beschwerden: Ohne Hormone zum Erfolg. A ¨ rztliche Praxis 35, 1946-1947. Du¨ker, E. M., Kopanski, L., Jarry, H., and Wuttke, W. (1991) Effects of extracts from Cimicifuga racemosa on gonadotripin release in menopausal women and ovariectomized rats. Planta Med. 57, 424-427. Fo¨ldes, J. (1959) Die Wirkungen eines Extraktes aus Cimicifuga racemosa. A ¨ rztliche Forsch. 13, 623-624. Lehmann-Willenbrock, E., and Riedel, H. H. (1998) Klinische und endokrinologische Untersuchungen zur Therapie ovarieller Ausfallerscheinungen nach Hysterektomie unter Belassung der Adnexe. Zentralbl. Gyna¨ kol. 110, 611-618. Liske, E. (1998) Therapeutic efficacy and safety of Cimicifuga racemosa for gynecological disorders. Adv. Ther. 15, 45-53. Petho¨, A. (1987) Klimakterische Beschwerden. Umstellung einer Hormonbehandlung auf ein pflanzliches Gynakologikum mo¨glich? A ¨ rztliche Praxis 38, 1551-1553. Stoll, W. (1987) Phytopharmakon beeinflusst atrophisches Vaginalepithel Doppelblindversuch Cimicifuga vs O ¨ strogenpraparat. Therapeutikon 1, 7-15. Stolze, H. (1982) Der andere Weg, klimakterische Beschwerden zu behandeln. Gyne 1, 14-16. Vorberg, G. (1984) Therapie klimakterischer Beschwerden. Z. Allgemeinmedizin 60, 626-629. Warnecke, G. (1985) Influencing menopausal symptoms with a phytotherapeutic agent. Die Med. Welt 36, 871-874. Liberman, S. (1998) A review of the effectiveness of Cimicifuga racemosa (black cohosh) for the symptoms of menopause. J. Women’s Health 7, 525-529. Beuscher, N. (1995) Cimicifuga racemosa L. Black cohosh. Z. Phytother. 16, 301-310. Hemmi, H., Kitame, F., Ishida, N., Kusano, G., Kondo, Y., and Nozoe, S. (1979) Inhibition of thymidine transport into phytohemagglutinin-stimulated lymphocytes by triterpenoids from Cimicifuga species. J. Pharmacobio-Dyn. 2, 339-349. Hemmi, H. (1981) Differential cytotoxicity of cytosine arabinoside toward murine leukemia L1210 cells and murine bone marrow progenitor cells inhibited in nucleoside transport by cimicifugoside. Res. Commun. Chem. Pathol. Pharmacol. 32, 565-568. Hemmi, H., and Ishida, N. (1980) The immune response of splenic lymphocytes after cimicifugoside treatment in vitro and pretreatment in vivo. J. Pharmacobio-Dyn. 3, 643-648. Takahira, M., Kusano, A., Shibano, M., Kusano, G., and Miyase, T. (1998) Piscidic acid and fukiic acid esters from Cimicifuga simplex. Phytochemistry 49, 2115-2119. Bolton, J. L., Trush, M. A., Penning, T. M., Dryhurst, G., and Monks, T. J. (2000) Role of quinones in toxicology. Chem. Res. Toxicol. 13, 135-160. Hagiwara, A., Hirose, M., Takahashi, S., Ogawa, K., Shirai, T., and Ito, N. (1991) Forestomach and kidney carcinogenicity of caffeic acid in F344 rats and C57BL/6N x C3H/HeN F1 mice. Cancer Res. 51, 5655-5660. Petroni, A., Blasevich, M., Salami, M., Papini, N., Montedoro, G. F., and Galli, C. (1995) Inhibition of platelet aggregation and eicosanoid production by phenolic components of olive oil. Thromb. Res. 78, 151-160. Aeschbach, R., Loliger, J., Scott, B. C., Murcia, A., Butler, J., Halliwell, B., and Aruoma, O. I. (1994) Antioxidant actions of thymol, carvacrol, 6-gingerol, zingerone and hydroxytyrosol. Food Chem. Toxicol. 32, 31-36. Johnson, B. M., Bolton, J. L., and van Breemen, R. B. (2001) Screening botanical extracts for quinoid metabolites. Chem. Res. Toxicol. 14, 1546-1551. NAPRALERT, A database on natural products. Available on-line through the Scientific and Technical Services of Chemical Abstracts Services, Columbus, Ohio. Nikolic, D., Fan, P. W., Bolton, J. L., and van Breemen, R. B. (1999) Screening for xenobiotic electrophilic metabolites using pulsed ultrafiltration-mass spectrometry. Comb. Chem. High Throughput Screening 2, 165-175. Tuck, K. L., Tan, H.-W., and Hayball, P. J. (2000) Synthesis of tritium-labeled hydroxytyrosol, a phenolic compound found in olive oil. J. Agric. Food Chem. 48, 4087-4090.

846

Chem. Res. Toxicol., Vol. 16, No. 7, 2003

(28) Single-dose study of black cohosh and red clover. January 2002. Internet: http://clinicaltrials.gov/ct/gui/c/w2r/show/ NCT00010959?order)1&JServSessionIdzone_ct)g43rlopiv1 (accessed 20 May 2002). (29) Chen, S.-N., Fabricant, D. S., Lu, Z.-Z., Zhang, H., Fong, H. H. S., and Farnsworth, N. R. (2002) Cimiracemates A-D, novel phenylpropanoid esters from the rhizomes of Cimicifuga racemosa. Phytochemistry 61, 409. (30) Calder, I. C., Hart, S. J., Healey, K., and Ham, K. N. (1981) N-hydroxyacetaminophen: a postulated toxic metabolite of acetaminophen. J. Med. Chem. 24, 988-993. (31) Koymans, L., van Lenthe, J. H., van de Straat, R., Donne-Op den Keldner, G. M., and Vermeulen, N. P. (1989) A theoretical study on the metabolic activation of paracetamol by cytochrome P-450: indications for a uniform oxidation mechanism. Chem. Res. Toxicol. 2, 60-66. (32) Nelson, S. D., Vaishnav, Y., Kambara, H., and Baillie, T. A. (1981) Comparative electron impact, chemical ionization and field desorption mass spectra of some thioether metabolites of acetaminophen. Biomed. Mass Spectrom. 8, 244-251. (33) Kassahun, K., Davis, M., Hu, P., Martin, B., and Baillie, T. A.

Johnson and van Breemen (1997) Biotransformation of the naturally occurring isothiocyanate sulforophane in the rat: identification of phase I metabolites and glutathione conjugates. Chem. Res. Toxicol. 10, 1228-1233. (34) Baillie, T. A., Pearson, P. G., Rashed, M. S., and Howald, W. N. (1989) The use of mass spectrometry in the study of chemically reactive drug metabolites. Application of MS/MS and LC/MS to the analysis of glutathione- and related S-linked conjugates of N-methylformamide. J. Pharm. Biomed. Anal. 7, 1351-1360. (35) Nakamura, Y., Torikai, K., Ohto, Y., Murakami, A., Tanaka, T., and Ohigashi, H. (2000) A simple phenolic antioxidant protocatechuic acid enhances tumor promotion and oxidative stress in female ICR mouse skin: dose-and-timing-dependent enhancement and involvement of bioactivation by tyrosinase. Carcinogenesis 21, 1899-1907. (36) Thompson, D. C., Perera, K., Krol, E. S., and Bolton, J. L. (1995) o-Methoxy-4-alkylphenols that form quinone methides of intermediate reactivity are the most toxic in rat liver slices. Chem. Res. Toxicol. 8, 323-327.

TX020108N