Human, rat, and mouse metabolism of resveratrol

Article abstract of DOI:10.1023/A:1021414129280

Purpose. Resveratrol, a phenolic phytoalexin occurring in grapes, wine, peanuts, and cranberries, has been reported to have anticarcinogenic, antioxidative, phytoestrogenic, and cardioprotective activities. Because little is known about the metabolism of this potentially important compound, the in vitro and in vivo metabolism of transresveratrol were investigated. Methods. The in vitro experiments included incubation with human liver microsomes, human hepatocytes, and rat hepatocytes and the in vivo studies included oral or intraperitoneal administration of resveratrol to rats and mice. Methanol extracts of rat urine, mouse serum, human hepatocytes, rat hepatocytes, and human liver microsomes were analyzed for resveratrol metabolites using reversed-phase high-performance liquid chromatography with on-line ultraviolet-photodiode array detection and mass spectrometric detection (LC-DAD-MS and LC-UV-MS-MS). UV-photodiode array analysis facilitated the identification of cis- and trans-isomers of resveratrol and its metabolites. Negative ion electrospray mass spectrometric analysis provided molecular weight confirmation of resveratrol metabolites and tandem mass spectrometry allowed structural information to be obtained. Results. No resveratrol metabolites were detected in the microsomal incubations, and no phase I metabolites, such as oxidations, reductions, or hydrolyzes, were observed in any samples. However, abundant trans-resveratrol-3-O-glucuronide and trans-resveratrol-3-sulfate were identified in rat urine, mouse serum, and incubations with rat and human hepatocytes. Incubation with β-glucuronidase and sulfatase to release free resveratrol was used to confirm the structures of these conjugates. Only trace amounts of cis-resveratrol were detected, indicating that isomerization was not an important factor in the metabolism and elimination of resveratrol. Conclusion. Our results indicate that trans-resveratrol-3-O-glucuronide and trans-resveratrol-3-sulfate are the most abundant metabolites of resveratrol. Virtually no unconjugated resveratrol was detected in urine or serum samples, which might have implications regarding the significance of in vitro studies that used only unconjugated resveratrol.

Full text of DOI:10.1023/A:1021414129280

Pharmaceutical Research, Vol. 19, No. 12, December 2002 (© 2002)
Research Paper
INTRODUCTION
Human, Rat, and Mouse Metabolism
of Resveratrol
Cancer chemoprevention may be regarded as the inges-
tion of nontoxic quantities of dietary or pharmaceutical
agents that are capable of preventing, inhibiting or reversing
the process of carcinogenesis (1,2). One dietary compound
under investigation as a chemoprevention agent is trans-
resveratrol (trans-3,5,4Ј-trihydroxystilbene) (Fig. 1) (3,4),
which is a naturally occurring phytoalexin produced by plants
in response to fungal infection or abiotic stresses, such as
heavy metal ions or ultraviolet light (UV) (5). Resveratrol has
attracted considerable attention because of its presence in
dietary sources, such as grapes, wine, peanuts, and cranberries
1
,2
1
3
Chongwoo Yu, Young Geun Shin, Anita Chow,
1
1
4
Yongmei Li, Jerome W. Kosmeder, Yong Sup Lee,
Wendy H. Hirschelman, John M. Pezzuto,
Rajendra G. Mehta, and Richard B. van Breemen
2
1,3
3
1,5
Received September 3, 2002; accepted September 9, 2002
(
6). In addition to cancer chemoprevention (3,4), other prop-
erties of resveratrol include antioxidative (7–9), antiplatelet
10–12), antifungal (13), phytoestrogenic (14,15), and cardio-
Purpose. Resveratrol, a phenolic phytoalexin occurring in grapes,
wine, peanuts, and cranberries, has been reported to have anticar-
cinogenic, antioxidative, phytoestrogenic, and cardioprotective activi-
ties. Because little is known about the metabolism of this potentially
important compound, the in vitro and in vivo metabolism of trans-
resveratrol were investigated.
(
protective activities (10,16,17).
The efficacy of orally administrated resveratrol will de-
pend on its absorption, metabolism, and tissue distribution.
Methods. The in vitro experiments included incubation with human Although many studies have implicated a role of resveratrol
liver microsomes, human hepatocytes, and rat hepatocytes and the in in disease prevention, only a few studies have addressed the
vivo studies included oral or intraperitoneal administration of resve-
bioavailability and metabolism of resveratrol (18–23). How-
ratrol to rats and mice. Methanol extracts of rat urine, mouse serum,
ever, none of these has provided a conclusive metabolic pro-
human hepatocytes, rat hepatocytes, and human liver microsomes
file for resveratrol including its metabolite structures.
were analyzed for resveratrol metabolites using reversed-phase high-
Human liver-derived experimental systems have been
performance liquid chromatography with on-line ultraviolet-
used extensively for the evaluation of drug metabolism, in-
cluding the use of intact cell systems, such as hepatocytes, and
cell-free systems, such as microsomes, or recombinant en-
photodiode array detection and mass spectrometric detection (LC-
DAD-MS and LC-UV-MS-MS). UV-photodiode array analysis facili-
tated the identification of cis- and trans-isomers of resveratrol and its
metabolites. Negative ion electrospray mass spectrometric analysis zymes, like specific cytochrome P-450 isozymes. Intact hepa-
provided molecular weight confirmation of resveratrol metabolites tocytes with full complements of enzymes and cofactors at
and tandem mass spectrometry allowed structural information to be physiologic levels and natural orientations should be more
obtained.
representative of the liver than cell-free systems with dis-
rupted membranes, incomplete cofactors, and enzymes (24).
Results. No resveratrol metabolites were detected in the microsomal
incubations, and no phase I metabolites, such as oxidations, reduc-
In vivo studies provide an overall indication of drug metabo-
tions, or hydrolyzes, were observed in any samples. However, abun-
lism resulting from all organ systems. In the present study,
dant trans-resveratrol-3-O-glucuronide and trans-resveratrol-3-
sulfate were identified in rat urine, mouse serum, and incubations
with rat and human hepatocytes. Incubation with ␤-glucuronidase
cryopreserved human and rat hepatocytes were compared
with human liver microsomes in the evaluation of resveratrol
metabolism (24,25). In addition, resveratrol metabolites were
and sulfatase to release free resveratrol was used to confirm the
structures of these conjugates. Only trace amounts of cis-resveratrol identified in rat urine and mouse serum. All metabolites were
were detected, indicating that isomerization was not an important characterized using high-performance liquid chromatography
factor in the metabolism and elimination of resveratrol.
Conclusion. Our results indicate that trans-resveratrol-3-O-
glucuronide and trans-resveratrol-3-sulfate are the most abundant
metabolites of resveratrol. Virtually no unconjugated resveratrol was
detected in urine or serum samples, which might have implications
regarding the significance of in vitro studies that used only unconju-
gated resveratrol.
(HPLC) with diode array detector (DAD) UV detection,
connected on-line with electrospray mass spectrometry or
tandem mass spectrometry (LC-DAD-MS and LC-UV-MS-
MS).
MATERIALS AND METHODS
KEY WORDS: trans-resveratrol; metabolism; LC-MS-MS; glucuro-
nides; sulfates.
Materials
trans-Resveratrol, NADPH, ␤-glucuronidase (type B-10
Department of Medicinal Chemistry and Pharmacognosy, Univer- from boving liver), and sulfatase (type IV: from Aerobacter
1
sity of Illinois at Chicago, 833 South Wood St., Chicago, Illinois
aerogenes, partially purified) were purchased from Sigma
60612.
Chemical (St. Louis, MO, USA). Formic acid (88%) was ob-
tained from J. T. Baker (Phillipsburg, NJ, USA) and was
diluted to the desired concentration using deionized distilled
water. HPLC-grade methanol, water, dimethyl sulfoxide
(DMSO), and acetonitrile were purchased from Fisher (Fair
Lawn, NJ, USA). Naringenin was purchased from Indofine
2
3
4
5
Department of Chemistry, University of Illinois at Chicago, Chi-
cago, Illinois 60607.
Department of Surgical Oncology, University of Illinois at Chicago,
Chicago, Illinois 60612.
Medicinal Chemistry, Research Center, Korea Institute of Science
and Technology, Seoul, 130-650, Korea.
To whom correspondence should be addressed. (e-mail address: Chemical (Somerville, NJ, USA). Human liver microsomes,
Breemen@uic.edu) cryopreserved human and rat hepatocytes, Krebs-Henseleit
1907
0724-8741/02/1200-1907/0 © 2002 Plenum Publishing Corporation

1
908
Yu et al.
0.1 mM trans-resveratrol in 10% DMSO was added. The 12-
well plate was incubated at 37°C for 4 h. During the incuba-
tion, 0.35-mL aliquots of the cell suspension (one aliquot per
well) containing trans-resveratrol metabolites were removed
at 1, 2, and 4 h and transferred to Eppendorf tubes. After the
addition of 0.35 mL of cold methanol for deactivation, each
tube was vortex-mixed for 30 s A control experiment was
performed using the same procedures except that the hepa-
tocytes were deactivated by the addition of cold methanol
before incubation with resveratrol. After centrifuging at
1
1
0,000 g for 3 min, the supernatant was transferred to another
.5-mL Eppendorf tube and evaporated to dryness in vacuo.
The residue was redissolved in 0.2 mL of methanol and cen-
trifuged at 10,000 g for 3 min. The supernatant was collected
and mixed with 0.4 mL of water. Rat hepatocyte samples were
prepared in the same manner as the human hepatocyte
samples.
Preparation of Rat Urine Samples
Resveratrol (20 mg/kg, dissolved in 6.5% ethanol in
neobee oil) was administered intraperitoneally (IP) to three
Sprague–Dawley female rats (75 days old, average weight ס
Fig. 1. Structures of resveratrol and resveratrol conjugates. (a) trans- 228 g), and urine was collected in metabolic cages for 2 h.
resveratrol; (b) trans-resveratrol-3-O-glucuronide; (c) trans- Also, control urine samples from two rats receiving no resve-
resveratrol-4Ј-O-glucuronide; (d) trans-resveratrol-3-sulfate; (e)
ratrol were obtained. A 100-␮L aliquot of each urine sample
trans-resveratrol-4Ј-sulfate; (f) cis-resveratrol; and (g) cis-resveratrol-
was mixed with 250 ␮L of acetonitrile and centrifuged at
1
3-O-glucuronide.
0,000 g for 3 min at 4°C. The supernatant was collected for
analysis. The research using rats and mouse adhered to the
buffer, and minimum essential media (MEM) were obtained “Principles of Laboratory Animal Care” (NIH publication
from In Vitro Technologies (Baltimore, MD, USA).
#85-23, revised in 1985).
To convert trans-resveratrol into cis-resveratrol, 0.7 mL
of a 0.01 mg/mL trans-resveratrol solution in methanol was Preparation of Mouse Serum Samples
exposed to UV light (␭ ס 254 nm) for 10 min.
Resveratrol was dissolved in a minimum volume of etha-
nol and then mixed with corn oil as a vehicle for drug delivery.
The final concentration of ethanol in the corn oil was 5.4%. In
one experiment, resveratrol at 20 mg/kg was administrated
via IP injection to 12 Balb/c female mice (4 weeks old, aver-
Methods
Incubation of Resveratrol in Human Liver Microsomes
Human liver microsomes (1 mg protein/mL) were incu- age body weight 19.5 g). In another experiment, resveratrol at
bated in 50 mM phosphate buffer (pH 7.4) with 5 mM resve- 20 mg/kg was administered via gavage to 12 Balb/c female
ratrol in the presence of 1 mM NADPH for 0, 15, 30, and 60 mice (4 weeks old, average weight 19.3 g). Control mice that
min. After incubation, two volumes of acetonitrile were did not receive resveratrol were used for both the IP injection
added to the reaction mixture, vortexed for 30 s, and then and gavage experiments.
centrifuged for 3 min at 10,000 g. The supernatant was col-
Blood samples (100–400 ␮L) were obtained by heart
lected and evaporated to dryness, and the residue was redis- puncture at 0.25, 0.5, 1, 2, 4, and 6 h, after treatment with
solved in 200 ␮L of methanol for analysis.
resveratrol. After clotting, the blood samples were centri-
fuged at 100 g for 20 min, and serum was collected from each
sample and stored at −70°C until analysis. A 25-␮L aliquot of
each mouse serum sample was mixed with 75 ␮L of acetoni-
trile and centrifuged at 10,000 g at 4°C for 3 min. Superna-
tants were collected for analysis.
The third experiment was conducted using gavage and
the same procedures except for the following modifications.
The dose of resveratrol was increased from 20 mg/kg to 60
mg/kg. Also, blood samples were obtained at 0.25, 0.5, 1, 2,
and 3 h after administration of resveratrol.
Incubation of Resveratrol with Hepatocytes
A cryopreserved sample of human hepatocytes was im-
mersed in a 37°C water bath and shaken gently for 80–90 s
The cell suspension was transferred to a 50-mL centrifuge
tube that had been precooled in ice. Minimum essential me-
dium (12 mL) was added to the cell suspension at the rate of
1
mL per 15 s. After each addition of 1 mL of medium, the
tube was shaken gently to prevent the cells from settling.
Next, the tube was centrifuged at 50 g at 4°C for 5 min. The
supernatant was removed by aspiration, and the cell pellet
was resuspended in 5 mL of Krebs-Henseleit buffer to pro-
Enzymatic Hydrolysis of Resveratrol Glucuronide and
Resveratrol Sulfate
6
vide a cell density of approximately 1 × 10 viable hepato-
cytes/mL. A 1-mL aliquot of the human hepatocytes suspen-
To confirm the presence of sulfate and glucuronide con-
sion was transferred to a well of a 12-well plate, and 0.1 mL of jugates of resveratrol, enzymatic cleavage was performed ac-

Resveratrol Metabolism
1909
cording to the method of Kuhnle et al. (20) and Sfakianos et was used from 10–80% acetonitrile. There was a 10-min re-
al. (26) with the following modifications. For resveratrol sul- equilibration period with the initial solvent mixture between
fate, aliquots of the extracts were evaporated to dryness and analyses. The flow rate was 0.2 mL/min.
reconstituted in 150 ␮L of 50 mM phosphate buffer (pH 7.4)
containing 0.5 U (30 ␮L) of aryl sulfatase. After incubating at using the Agilent mass spectrometer with the electrospray
7°C for 2.5 and 18.5 h, respectively, 100 ␮L of cold methanol capillary set at 4 kV. The flow rate of the nitrogen drying gas
Negative ion electrospray mass spectra were obtained
3
was added into each sample vial to stop the reaction. For was 6.0 L/min at a temperature of 300°C. Mass spectra were
resveratrol glucuronide, aliquots of extracts were incubated recorded over the range of m/z 70–800. Alternatively, se-
with ␤-glucuronidase (final concentration ס 8,000 U/mL) in lected ion monitoring was used for greater sensitivity by re-
5
0 mM phosphate buffer (pH 5.5, 150 ␮L) for 1 h. The reac- cording signals for ions of m/z 227, 307, and 403 with a dwell
tion was terminated by adding 150 ␮L of cold methanol. After time of 592 ms/ion. Negative ion electrospray tandem mass
centrifugation at 10,000 g for 3 min at 4°C, supernatants were spectra were recorded using the Micromass triple quadruple
collected, evaporated to dryness in vacuo, and reconstituted instrument with the electrospray capillary set at 2.5 kV and a
in 50 ␮L of 50% methanol. Control incubations were per- source block temperature of 100°C. Nitrogen was used as the
formed without addition of enzyme. All sample preparations drying and nebulizing gas at flow rates of approximately 8
were performed in dim light to minimize photochemical isom- L/min and 0.8 L/min, respectively. Argon at a pressure of 1.4
−
3
erization of trans-resveratrol to the cis-form.
× 10 mbar was used as the collision gas for collision-induced
dissociation (CID). UV spectra were monitored at 300 nm.
Product ion scans were carried out using the deprotonated
molecules of resveratrol and its metabolites as precursor ions.
Synthesis of Resveratrol Monosulfate Isomers
To determine the attachment site of sulfate in resveratrol When greater sensitivity was required, multiple reaction
sulfate, resveratrol monosulfate isomers were synthesized. monitoring was used as described in the Results and Discus-
Resveratrol sulfate was formed by standard reaction condi- sion section.
tions with one equivalent of sulfur trioxide-pyridine complex
For the quantitative analysis of trans-resveratrol, trans-
in dry pyridine (27). The yield of monosulfates was approxi- resveratrol-3-sulfate, and trans-resveratrol-3-glucuronide,
mately 30%, and the remainder was unreacted resveratrol. calibration curves were constructed by plotting the LC-MS-
The ratio of 3- and 4Ј-monosulfates was determined to be 7:3, MS peak area ratio of trans-resveratrol or trans-resveratrol-
and the mixture was separated by preparative HPLC. Struc- 3-sulfate to the internal standard naringenin (at 0.8 ␮M)
tures of the isomers were confirmed by nuclear magnetic against the analyte concentration. Note that no impurities and
resonance measurements. Spectra were recorded on a Bruker no cis-resveratrol were detected in the trans-resveratrol stan-
(Billerica, MA) 300 MHz instrument, and the following data dard, which was used for all studies. The linear regression
were obtained. analyses of these standard curves showed a correlation coef-
1
2
Resveratrol-3-sulfate: Amorphous white solid, H-NMR ficient of r ס 0.999. Because no resveratrol glucuronide stan-
(
CD OD) d 7.38 (d, J ס 8.6 Hz, H-2’,6’), 7.05 (d, J ס 16.5 Hz, dards were available, the concentration of trans-resveratrol-
3
trans-vinyl), 6.99 (dd, J ס 2.2 Hz, H-2), 6.87 (d, J ס 16.5 Hz, 3-glucuronide was estimated using the trans-resveratrol stan-
trans-vinyl), 6.78 (d, J ס 8.6 Hz, H-3Ј,5Ј), 6.75 (dd, J ס 2.1 Hz, dard curve.
H-6), 6.67 (dd, J ס 2.1 Hz, H-4), HPLC retention time 13.1
min.
RESULTS AND DISCUSSION
1
Resveratrol-4Ј-sulfate: Amorphous white solid, H-NMR
(
CD OD) d 7.51 (d, J ס 8.7 Hz, H-2’,6’), 7.30 (d, J ס 8.6 Hz,
3
H-3Ј,5Ј), 7.05 (d, J ס 16.3 Hz, trans-vinyl), 6.87 (d, J ס 16.5 Human Microsomes and Hepatocytes
Hz, trans-vinyl), 6.50 (d, J ס 2.1 Hz, H-2,6), 6.20 (dd, J ס 2.1
Hz, H-4), HPLC retention time 11.6 min.
After incubation of trans-resveratrol with human liver
microsomes, intact resveratrol but no metabolites were de-
tected using LC-DAD-MS. However, resveratrol incubated
with human hepatocytes for 4 h showed several new peaks
Liquid Chromatography-Mass Spectrometry (LC-MS)
LC-DAD-MS analyses were performed using an Agilent with abundant deprotonated molecules of m/z 403 and 307.
(Palo Alto, CA, USA) 1100 HPLC system equipped with a Therefore, LC-UV-MS-MS with product ion scanning (Fig. 2)
photodiode array detector and interfaced to a model G1946A and multiple reaction monitoring (Fig. 3) were performed for
single quadruple electrospray mass spectrometer. UV spectra greater sensitivity and selectivity to confirm these peaks as
were recorded from 200–360 nm. A Micromass (Manchester, metabolites of resveratrol. trans-Resveratrol was detected at
UK) Quattro II triple quadruple mass spectrometer equipped a retention time of 30.8 min, and the corresponding product
with a Waters (Milford, MA, USA) 2690 HPLC system and ion CID mass spectrum is shown in Fig. 2A.
2
487 UV detector was used for LC-UV-MS-MS. UV chro-
Resveratrol sulfate has been reported in studies with per-
matograms were recorded at 210 and 300 nm. HPLC separa- fused rat intestine and human liver (19,21,22). In our studies,
tions were obtained at 25°C using a Waters XTerra MS C18 resveratrol sulfate was detected in incubations with some hu-
column (2.1 mm × 100 mm, 3.5-␮m particle size).
man hepatocytes, as well as with rat hepatocytes and in mouse
Aliquots (10 ␮L) of urine or serum extracts were injected serum. For example, in the LC-UV-MS-MS analysis of a
directly onto the HPLC column and eluted with a solvent mouse serum extract (Fig. 2B), the ion of m/z 307 fragmented
system consisting of 0.1% (v/v) formic acid and acetonitrile. A to form a product ion of m/z 227 that corresponded to resve-
6
0-min linear gradient was used from 10–30% acetonitrile, ratrol itself after the loss of sulfate. Furthermore, only one
and for some mouse serum samples a 30-min linear gradient resveratrol monosulfate and no disulfate metabolites were

1
910
Yu et al.
Fig. 3. Negative ion electrospray multiple reaction monitoring LC-
UV-MS-MS analysis of resveratrol incubated for 4 h with human
hepatocytes or deactivated human hepatocytes (control). Peaks were
assigned to the following compounds. (a) trans-resveratrol-4Ј-O-
glucuronide; (b) trans-resveratrol-3-O-glucuronide; (c) cis-
resveratrol-3-O-glucuronide; (d) trans-resveratrol; (e) cis-resveratrol.
detected. This resveratrol sulfate isomer was synthesized and
confirmed to be trans-resveratrol-3-sulfate.
LC-UV-MS-MS with multiple reaction monitoring of the
transition of m/z 403 → 227 was performed and showed three
resveratrol glucuronide peaks eluting at 12.1, 18.1, and 28.6
min (Fig. 3). LC-UV-MS-MS with MRM of an identical con-
trol containing deactivated human hepatocytes showed no
signals for resveratrol glucuronide. Instead, only trans-
resveratrol and a trace of the cis-isomer were detected at 30.8
and 42.0 min (Fig. 3), respectively. The ion of m/z 403 (Fig.
Fig. 2. LC-UV-MS-MS product ion mass spectra obtained using
negative ion electrospray and CID during the analysis of resveratrol
metabolites. (A) Resveratrol eluting at 30.8 min (extract of human
hepatocyte incubation); (B) Resveratrol sulfate at 28.0 min retention
time (mouse serum extract); (C) Resveratrol glucuronide at 18.1 min
(human hepatocyte incubation)

Resveratrol Metabolism
1911
resveratrol (retention time 30.8 min) is shown in Fig. 2A.
Note the absence of glucuronic acid ions of m/z 175 or 113 in
the resveratrol tandem mass spectrum.
The use of a DAD in the LC-DAD-MS analysis allowed
the confirmation of the identity of each chromatographic
peak not only by its retention time but also by its spectrum.
The UV/DAD spectra pattern obtained for trans-resveratrol
and cis-resveratrol conjugates (Fig. 4) were consistent with
those obtained for trans-resveratrol and cis-resveratrol stan-
dards that have been reported in other studies (30,31). It
should be noted that trans-resveratrol produced the same
UV/DAD spectrum whether conjugated at the 3- or 4Ј-
position. Therefore, only one UV/DAD spectrum for these
two glucuronides of trans-resveratrol is shown in Fig. 4. These
spectra indicate that the conformations of the three resve-
ratrol glucuronides detected at 12.1, 18.1, and 28.6 min in Fig.
3
are trans-, trans-, and cis-, respectively.
Because synthetic resveratrol-3-sulfate was retained
longer than resveratrol-4Ј-sulfate during reversed-phase
HPLC (13.1 and 11.6 min, respectively), it is probable that
resveratrol-3-O-glucuronic acid will be retained longer than
resveratrol-4Ј-O-glucuronide. The longer retention times of
the conjugates at the C-3 position of the A ring are probably
the result hydrogen bonding of the sulfate or glucuronic acid
to the –OH group at C-5 of the A ring making the conjugate
less polar than the corresponding conjugates at the C-4Ј po-
sition. When the sulfate group or glucuronic acid is attached
Fig. 4. UV spectra obtained during LC-DAD-MS of human hepato-
cytes for (A) trans-resveratrol conjugates, and (B) cis-resveratrol at the C-4Ј of ring B, the two unconjugated phenolic OH
conjugates.
groups on the A ring are fully exposed rendering the molecule
more polar. As a result, resveratrol-4Ј-sulfate elutes earlier
than resveratrol-3-sulfate. By analogy, trans-resveratrol-4Ј-O-
2C), which was detected at 18.1 min fragmented to form prod- glucuronide is probably the earlier eluting glucuronide peak
uct ions of m/z 227 and m/z 175 that corresponded to resve- at 12.1 min, and the glucuronide at 18.1 min is assigned as
ratrol itself after the loss of glucuronic acid and dehydrated trans-resveratrol-3-O-glucuronide. The resveratrol glucuro-
glucuronic acid, respectively. Also, a product ion of m/z 113 nide eluting at 28.6 min is assigned as cis-resveratrol-3-O-
was observed, which can be considered characteristic of the glucuronide by comparison of the retention times of trans-
presence of a glucuronic acid moiety. The ion of m/z 113 was resveratrol, cis-resveratrol, and trans-resveratrol-3-O-glucu-
formed by further dissociation of the fragment ion of m/z 175 ronide, respectively, and also considering steric effects. The
and is common in the negative ion mass spectra of glucuro- structures of all of these compounds are shown in Fig. 1.
nide metabolites (28,29). Therefore, the peak at 18.1 min was
Because the amount of resveratrol sulfate was minor
identified as resveratrol monoglucuronide. No peaks corre- compared with the formation of the glucuronide in the incu-
sponding to a diglucuronide (mw ס 580) were detected. For bations with human hepatocytes, resveratrol sulfate seems to
comparison, the negative ion MS-MS product ion spectrum of be a minor human hepatic metabolite.
Table I. Summary of Resveratrol Metabolites Detected Using LC-UV-MS-MS
Experiments
Detected metabolites
Retention time
Human microsome control
Human microsome 1 h
Human hepatocytes control
Human hepatocytes 4 h
None
None
None
—
—
—
trans-resveratrol-4Ј-O-glucuronide
trans-resveratrol-3-O-glucuronide
cis-resveratrol-3-O-glucuronide
None
trans-resveratrol-3-O-glucuronide
None
trans-resveratrol-3-O-glucuronide
trans-resveratrol-3-sulfate
None
12.1 min
18.1 min
28.6 min
—
18.0 min
—
18.4 min
32.9 min
—
Rat urine control
Rat urine 2 h
Rat hepatocytes control
Rat hepatocytes 4 h
Mouse serum control
Mouse serum gavage 15 min
trans-resveratrol-3-O-glucuronide
trans-resveratrol-3-sulfate
17.2 min
30.1 min

1
912
Yu et al.
Rat Urine and Hepatocytes
During LC-UV-MS-MS MRM analysis of rat urine
samples, the transition m/z 403 → 227 was monitored for
resveratrol glucuronide, m/z 307 → 227 for resveratrol sulfate,
and m/z 227 → 185 for resveratrol. However, only one peak
was detected at 18.0 min (data not shown) for the transition
m/z 403 → 227. This metabolite was identified as trans-
resveratrol-3-O-glucuronide and showed the same retention
time as in the analysis of resveratrol metabolites from human
hepatocytes. Analysis of rat hepatocytes was also performed
using LC-UV-MS-MS with MRM. In this case, two major
peaks were detected at 18.4 and 32.9 min (data not shown).
Based on the mass chromatogram and by comparing the re-
tention times, we have assigned these peaks as trans-
resveratrol-3-O-glucuronide, and trans-resveratrol-3-sulfate,
respectively. Resveratrol sulfate was the more abundant me-
tabolite in rat hepatocytes. All of the resveratrol metabolites
detected in these experiments are summarized in Table I.
Mouse Serum
Resveratrol was administered to mice at two different
doses using IP injection or gavage, and serial blood samples
were obtained various time points up to 4 h. Resveratrol gluc-
uronide and resveratrol sulfate were detected as the only res-
veratrol metabolites in the mouse serum samples, and both of
these metabolites were detected after IP or intragastric (IG)
administration. The time courses for the appearance of res-
veratrol sulfate and resveratrol glucuronide in mouse serum
following both routes of administration are shown in Fig. 5.
After the administration of 20 mg/kg, the maximum con-
centrations of both metabolites were observed in the serum
samples obtained at the first time point of 15 min (Fig. 5A). In
these samples, the concentration of resveratrol sulfate (13
␮
M) in the mouse serum was almost 3-fold greater than that
of resveratrol glucuronide (5 ␮M). Only traces of unconju-
gated resveratrol were observed. Furthermore, no resveratrol
or its metabolites were detected after 1 h (Fig. 5A).
To confirm the peak assigned to resveratrol sulfate and
resveratrol glucuronide, aliquots of mouse serum samples
were incubated with sulfatase and ␤-glucuronidase, respec-
tively and then re-analyzed using LC-UV-MS-MS. As a re-
sult, the resveratrol sulfate peak disappeared whereas a res-
veratrol peak appeared in the sulfatase incubation (Fig. 6),
and the resveratrol glucuronide peak disappeared whereas a
resveratrol peak appeared in the ␤-glucuronidase incubation.
The resveratrol peak intensity increased as the incubation
time increased in both cases. Finally, these observations were
confirmed using LC-DAD-MS with a UV photodiode array
detector and mass spectrometric scanning in the range m/z
Fig. 5. Resveratrol metabolites in mouse serum following doses of
(
A) 20 mg/kg via IP injection or gavage (IG), and (B) 60 mg/kg via
gavage (IG). Each data point represents the analysis of pooled blood
samples from five control or five treated mice. Because no resveratrol
or resveratrol conjugates were detected in the control serum, these
data are not plotted.
7
0–800. All of the UV/DAD spectra obtained were consistent
with the data obtained in the studies of rat urine, rat hepato-
cytes, and human hepatocytes samples (Fig. 4). Furthermore,
the mass spectra of these peaks were consistent with those uronide in resveratrol glucuronide is under investigation, it is
observed in the other experiments. Resveratrol sulfate was reasonable to speculate that the conformations of major res-
not affected by the ␤-glucuronidase incubation and resve- veratrol metabolites in mouse serum are trans-resveratrol-3-
ratrol glucuronide was not affected by sulfatase incubation sulfate and trans-resveratrol-3-O-glucuronide.
and they were still observed eluting at approximately 30 and
In the second set of experiments, 60 mg/kg of resveratrol
1
7 min, respectively. Therefore, our data confirm the pres- was given only via gavage because oral administration would
ence of resveratrol sulfate and resveratrol glucuronide in the be more relevant to future clinical investigations, and serum
mouse serum samples. Although the attachment site of gluc- was collected serially up to 3 h. LC-DAD-MS and LC-UV-

Resveratrol Metabolism
1913
to absorb the large volume that was administered. Note that
serum resveratrol disappeared after 30 min but that resve-
ratrol sulfate and resveratrol glucuronide were still detectable
3
h after the highest dosage (Fig. 5B). The prolonged detec-
tion of low levels of resveratrol sulfate and resveratrol gluc-
uronide in serum following the highest dose suggests that
resveratrol was distributed to various tissues and was being
cleared slowly. Perhaps only the highest dosage produced tis-
sue levels that were high enough for subsequent detection of
the cleared metabolites.
CONCLUSION
Although many studies have implicated a role for resve-
ratrol in disease prevention, information on in vivo bioavail-
ability and metabolism is incomplete and the structures of the
resveratrol metabolites were uncertain. Our LC-DAD-MS
and LC-UV-MS-MS data for the human, rat, and mouse ex-
periments indicate that trans-resveratrol-3-O-glucuronide is
the primary metabolite of resveratrol in human liver, and that
trans-resveratrol-3-O-glucuronide and trans-resveratrol-3-
sulfate are both significant metabolites in rat urine, mouse
serum, and formed by rat hepatocytes. It is important to note
that no phase I metabolites of resveratrol such as oxidations,
reductions or hydrolyzes were detected in any of these sys-
tems. Furthermore, conjugated resveratrol and not its free
form were found to predominate in the circulation. These
data suggest that the potential biologic activity of resveratrol
conjugates should be considered in future investigations. Fur-
thermore, the form of resveratrol in cells and tissues after oral
or IP administration should be investigated to determine the
levels of conjugated and unconjugated resveratrol at these
potential points of action.
ACKNOWLEDGMENT
These studies were funded by grants R24 CA83124 and
P01 CA48112 awarded by the National Cancer Institute.
REFERENCES
1
. M. B. Sporn, N. M. Dunlop, D. L. Newton, and J. M. Smith.
Prevention of chemical carcinogenesis by vitamin A and its syn-
thetic analogs (retinoids). Fed. Proc. 35:1332–1338 (1976).
. W. K. Hong and M. B. Sporn. Recent advances in chemopreven-
tion of cancer. Science 278:1073–1074 (1997).
2
3
. M. Jang, L. Cai, G. O. Udeani, K. V. Slowing, C. F. Thomas, C.
W. W. Beecher, H. H. S. Fong, N. R. Farnsworth, A. D. King-
horn, R. G. Mehta, R. C. Moon, and J. M. Pezzuto. Cancer che-
mopreventive activity of resveratrol, a natural product derived
from grapes. Science 275:218–220 (1997).
4
5
. M. Jang and J. M. Pezzuto. Cancer chemopreventive activity of
resveratrol. Drugs Exp. Clin. Res 25:65–77 (1999).
. J. A. Bailey. Mechanism of phytoalexin accumulation. In Phytoa-
lexins, eds. J. A. Baily and J. W. Manfield, pp. 289–318, Wiley,
New York, 1982.
Fig. 6. LC-UV-MS-MS analysis of mouse serum collected after 15
min of administrating resveratrol via gavage and the same mouse
serum sample incubated with sulfatase for 18.5 h. Peaks were as-
signed to the following compounds. (a) trans-resveratrol-3-O-
glucuronide; (b) trans-resveratrol-3-sulfate; (c) trans-resveratrol.
6. Y. Wang, F. Catana, Y. Yang, R. Roderick, and R. B. van Bree-
men. Analysis of resveratrol in grape products, cranberry juice
and wine using liquid chromatography-mass spectrometry. J. Ag-
ric. Food Chem. 50:431–435 (2002).
MS-MS analysis were performed using the same conditions as
described above. The same resveratrol sulfate and resveratrol
glucuronide metabolites were detected in these analyses as
were observed in the analyses of serum samples following
administration of the lower dose of resveratrol. The resve-
ratrol sulfate concentration reached a maximum value in
mouse serum after 30 min instead of 15 min as observed for
the lower dosage, probably because more time was required
7
. E. N. Frankel, A. L. Waterhouse, and J. E. Kinsella. Inhibition of
human LDL oxidation by resveratrol. Lancet 341:1103–1104
(1993).
8. N. J. Miller and C. A. Rice-Evans. Antioxidant activity of resve-
ratrol in red wine—to the editor. Clin. Chem. 41:1789 (1995).
9
. B. Fuhrman, A. Lavy, and M. Aviram. Consumption of red wine
with meals reduces the susceptibility of human plasma and low-
density-lipoprotein to lipid-peroxidation. Am. J. Clin. Nutr. 42:
549–554 (1995).

1
914
Yu et al.
1
0. A. A. E. Bertelli, L. Giovannini, D. Giannesi, M. Migliori, W.
Bernini, M. Fregoni, and A. Bertelli. Antiplatelet activity of syn-
thetic and natural resveratrol in red wine. Int. J. Tissue React.
fici. Sulphation of resveratrol, a natural product present in grapes
and wine, in human liver and duodenum. Xenobiotica 30:609–617
(2000).
1
7:1–3 (1995).
22. C. De Santi, A. Pietrabissa, R. Spisni, F. Mosca, and G. M. Paci-
fici. Sulphation of resveratrol, a natural product present in wine,
and its inhibition by natural flavonoids. Xenobiotica 30:857–866
(2000).
23. C. De Santi, A. Pietrabissa, F. Mosca, and G. M. Pacifici. Gluc-
uronidation of resveratrol, a natural product present in grape and
wine, in the human liver. Xenobiotica 30:1047–1054 (2000).
1
1. C. R. Pace-Asciak, S. E. Hahn, E. P. Diamandis, G. Soleas, and
D. M. Goldberg. The red wine phenolics trans-resveratrol and
quercetin block human platelet-aggregation and eicosanoid syn-
thesis—Implications for protection against coronary heart-
disease. Clin. Chim. Acta 235:207–219 (1995).
1
2. Y. Kimura, H. Okuda, and S. Arichi. Effects of stilbenes on ara-
chidonate metabolism in leukocytes. Biochim. Biophys. Acta 834: 24. A. P. Li, C. Lu, J. A. Brent, C. Pham, A. Fackett, C. E. Ruegg,
275–278 (1985).
and P. M. Silber. Cryopreserved human hepatocytes: character-
ization of drug-metabolizing enzyme activities and applications in
higher throughput screening assays for hepatotoxicity, metabolic
stability, and drug-drug interaction potential. Chem. Biol. Inter-
act. 121:17–35 (1999).
1
3. S. Sotheeswaran and V. Pasupathy. Distribution of resveratrol
oligomers in plants. Phtyochemistry 32:1083–1092 (1993).
4. P. Kopp. Resveratrol, a phytoestrogen found in red wine. A pos-
sible explanation for the conundrum of the “French paradox”?
Eur. J. Endocrinol. 138:619–620 (1998).
1
25. A. P. Li. Overview: Hepatocytes and cryopreservation—a per-
sonal historical perspective. Chem. Biol. Interact. 121:1–5 (1999).
26. J. Sfakianos, L. Coward, M. Kirk, and S. Barnes. Intestinal uptake
and biliary excretion of the isoflavones genistein in rats. J. Nutr.
127:1260–1268 (1997).
1
1
1
1
5. R. Lu and G. Serrero. Resveratrol, a natural product derived
from grape exhibits antiestrogenic activity and inhibits the
growth of the human breast cancer cells. J. Cell Physiol. 179:297–
304 (1999).
6. D. K. Das, M. Sato, P. S. Ray, G. Maulik, R. M. Engelman. 27. N. Kawai, Y. Fujibayashi, S. Kuwabara, K.-I. Takao, Y. Ijuin, and
A. A. E. Bertelli, A. Bertelli. Cardioprotection of red wine: Role
S. Kobayashi. Synthesis of a potential key intermediate of akat-
erpin, specific inhibitor of PI-PLC. Tetrahedron 56:6467–6478
(2000).
of polyphenolic antioxidants. Drugs Exptl. Clin. Res. 25:115–120
(1999).
7. C. R. Pace-Asciak, O. Rounova, S. E. Hahn, E. P. Diamandis, 28. L. Debrauwer, E. Rathahao, G. Boudry, M. Baradat, and J. P.
and D. M. Goldberg. Wines and grape juices as modulation of
platelet aggregation in healthy human subjects. Clin. Chim. Acta.
Cravedi. Identification of major metabolites of prochloraz in
rainbow trout by liquid chromatography and tandem mass spec-
trometry. J. Agric. Food Chem. 49:3821–3826 (2001).
246:163–182 (1996).
8. A. A. E. Bertelli, L. Giovannini, R. Stradi, S. Urien, J.-P. Tille- 29. P. Manini, R. Andreoli, A. Mutti, E. Bergamaschi, I. Franchini,
ment, and A. Bertelli. Kinetics of trans- and cis-resveratrol
3,4Ј,5-trihydroxystilbene) after red wine oral administration in
rats. Int. J. Clin. Pharm. Res. 16:77–81 (1996).
9. W. Andlauer, J. Kolb, K. Siebert, and P. F u¨ rst. Assessment of
and W. M. A. Niessen. Determination of glucuronide molecules
of toxicological interest by liquid chromatography negative-ion
mass spectrometry with atmospheric pressure chemical ioniza-
tion. Chromatographia 47:659–666 (1998).
(
1
resveratrol bioavailability in the perfused small intestine of the 30. D. M. Goldberg, E. Ng, A. Karumanchiri, J. Yan, E. P. Diaman-
rat. Drugs Exptl. Clin. Res. 26:47–55 (2000).
0. G. Kuhnle, J. P. Spencer, G. Chowrimootoo, H. Schroeter, E. S.
Debnam, S. K. S. Srai, C. Rice-Evans, and U. Hahn. Resveratrol
dis, and G. J. Soleas. Assay of resveratrol glycosides and isomers
in wine by direct-injection high-performance liquid chromatog-
raphy. J. Chromatogr. A 708:89–98 (1995).
2
is absorbed in the small intestine as resveratrol glucuronide. Bio- 31. B. C. Trela and A. L. Waterhouse. Resveratrol: Isomeric molar
chem. Biophys. Res. Commun. 272:212–217 (2000).
1. C. De Santi, A. Pietrabissa, R. Spisni, F. Mosca, and G. M. Paci-
absorptivities and stability. J. Agric. Food Chem. 44:1253–1257
(1996).
2