Curcuminoids form reactive glucuronides in vitro

Article abstract of DOI:10.1021/jf0623283

Curcumin is of current interest because of its putative anti-inflammatory, anticarcinogenic, and anti-Alzheimer's activity, but its pharmacokinetic and metabolic fate is poorly understood. The present in vitro study has therefore been conducted on the glucuronidation of curcumin and its major phase I metabolite, hexahydro-curcumin, as well as of various natural and artificial analogs. The predominant glucuronide generated by rat and human liver microsomes from curcumin, hexahydro-curcumin, and other analogs with a phenolic hydroxyl group was a phenolic glucuronide according to LC-MS/MS analysis. However, a second glucuronide carrying the glucuronic acid moiety at the alcoholic hydroxyl group was formed from the same curcuminoids, but not hexahydro-curcuminoids, by human microsomes. Curcuminoids without a phenolic hydroxyl group gave rise to the aliphatic glucuronide only. The phenolic glucuronides of curcuminoids, but not of hexahydro-curcuminoids, were rather lipophilic and, in part, unstable in aqueous solution, their stability depending strongly on the type of aromatic substitution. The phenolic glucuronide of curcumin and of its natural congeners, but not the parent compounds, clearly inhibited the assembly of microtubule proteins under cell-free conditions, implying chemical reactivity of the glucuronides. These novel properties of the major phase II metabolites of curcuminoids deserve further investigation.

Full text of DOI:10.1021/jf0623283

538 J. Agric. Food Chem. 2007, 55, 538−544
Curcuminoids Form Reactive Glucuronides In Vitro
ERIKA PFEIFFER,† SIMONE I. HOEHLE,† STEPHAN G. WALCH,† ALEXANDER RIESS,†
ANIKOÄ M. SOÄ LYOM,‡ AND MANFRED METZLER*,†
Institute of Applied Biosciences, University of Karlsruhe, P. O. Box 6980, D-76128 Karlsruhe,
Germany, and College of Pharmacy, University of Arizona, 1703 East Mabel Street, Tucson, Arizona
85721-0207
Curcumin is of current interest because of its putative anti-inflammatory, anticarcinogenic, and anti-
Alzheimer’s activity, but its pharmacokinetic and metabolic fate is poorly understood. The present in
vitro study has therefore been conducted on the glucuronidation of curcumin and its major phase I
metabolite, hexahydro-curcumin, as well as of various natural and artificial analogs. The predominant
glucuronide generated by rat and human liver microsomes from curcumin, hexahydro-curcumin, and
other analogs with a phenolic hydroxyl group was a phenolic glucuronide according to LC-MS/MS
analysis. However, a second glucuronide carrying the glucuronic acid moiety at the alcoholic hydroxyl
group was formed from the same curcuminoids, but not hexahydro-curcuminoids, by human
microsomes. Curcuminoids without a phenolic hydroxyl group gave rise to the aliphatic glucuronide
only. The phenolic glucuronides of curcuminoids, but not of hexahydro-curcuminoids, were rather
lipophilic and, in part, unstable in aqueous solution, their stability depending strongly on the type of
aromatic substitution. The phenolic glucuronide of curcumin and of its natural congeners, but not the
parent compounds, clearly inhibited the assembly of microtubule proteins under cell-free conditions,
implying chemical reactivity of the glucuronides. These novel properties of the major phase II
metabolites of curcuminoids deserve further investigation.
KEYWORDS: Curcumin; glucuronide; microsomes; microtubule proteins
INTRODUCTION
Curcuminoids are the yellow-orange pigments of turmeric,
which is obtained from the rhizomes of the east Indian plant
Curcuma longa and used extensively in traditional Indian
cooking and Ayurvedic herbal remedies (1-3). Turmeric
contains curcumin as the major component, together with
smaller amounts of demethoxy-curcumin and bisdemethoxy-
curcumin (4). The chemical structure of these curcuminoids is
that of a diarylheptanoid with differently substituted aryl rings
and a conjugated aliphatic structure (Figure 1). Turmeric has
long been recognized for its broad spectrum of biological
activities, in particular, its anti-inflammatory, antioxidant, and
anticarcinogenic effects (2, 5, 6). More recent evidence suggests
that curcumin may prevent or even cure Alzheimer’s disease
because it has been shown to decrease insoluble and soluble
â-amyloid and plaque burden in the brain of an Alzheimer
transgenic mouse model (7, 8). Accordingly, experimental and
clinical studies on the beneficial effects of curcuminoids are
increasing, and the number of nutritional supplements containing
Figure 1. Chemical structures of curcuminoids.
curcuminoids is rising. However, information on the metabolism
of curcuminoids and their pharmacokinetic fate is still scant.
Curcumin is known to be readily converted to several reductive
metabolites, mostly hexahydro-curcumin, and conjugates of both
the parent compound and its phase I metabolites with glucuronic
acid and sulfate have been identified in studies with rats and
mice in vivo (9-14).
In a recent study using rat liver slices and subcellular
fractions, we have shown that the three natural curcuminoids
differ in their chemical stability, whereas their hexahydro
metabolites are stable (15). In the present study, we have
investigated the glucuronidation of curcuminoids in more detail.
* To whom correspondence should be addressed. Tel.: +49-721-
6082132. Fax: +49-721-6087255. E-mail: manfred.metzler@chemie.uni-
karlsruhe.de.
† University of Karlsruhe.
‡ University of Arizona. Present address: Integrated Biomolecule
Corporation/IBC Labs, 2005 E. Innovation Park Dr., Tucson, AZ 85755-
1966. E-mail: amsolyom@integratedbiomolecule.com.
10.1021/jf0623283 CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/22/2006

Reactive Glucuronides of Curcuminoids
J. Agric. Food Chem., Vol. 55, No. 2, 2007 539
O,O-dimethyl-curcumin had a λmax at 280 nm (ꢀ ) 5100 in methanol)
and a molecular ion at m/z 402.2039 (calculated for C23H30O6:
402.2042). 1H NMR (500 MHz, CDCl3): δ 1.60-1.71 (m, 1H), 1.78-
1.85 (m, 1H), 2.57-2.61 (m, 2H), 2.61-2.68 (m, 1H), 2.75-2.77 (m,
2H), 2.73-2.80 (m, 1H), 2.86-2.89 (m, 2H), 3.10 (bs, 1H), 3.876 (s,
3H), 3.880 (s, 3H), 3.887 (s, 3H), 3.894 (s, 3H), 4.05-4.10 (m, 1H),
6.72-6.82 (m, 6H). 13C NMR (125 MHz, CDCl3): δ 29.2 (t), 31.4 (t),
38.3 (t), 45.3 (t), 49.4 (t), 55.86 (q, 2C), 55.94 (q, 2C), 66.9 (d), 111.3
(d), 111.3 (d), 111.7 (d), 111.8 (d), 120.1 (d), 120.2 (d), 133.3 (s),
134.4 (s), 147.2 (s), 147.5 (s), 148.88 (s), 148.94 (s), 211.3 (s).
â-Glucuronidase (type B-1 from bovine liver) and all other chemicals
and reagents were obtained from Sigma-Aldrich Chemical Co. (Stein-
heim, Germany). HPLC grade acetonitrile was from Carl Roth Co.
(Karlsruhe, Germany).
Male Sprague-Dawley rats were purchased from Harlan Winkel-
mann GmbH (Borchen, Germany). Animals were kept under a 12-h
light/dark cycle and received water and commercial lab chow ad libitum.
The livers of untreated male rats with 200-300 g weight and the liver
of a 63-year-old male white (kindly provided by Dr. J. Weymann,
former Knoll AG, Ludwigshafen, Germany) were used for the prepara-
tion of hepatic microsomes as described previously by Lake (17).
Human intestinal microsomes and supersomes, i.e., microsomes from
insect Sf-9 cells infected with a baculovirus strain containing the DNA
of human UGT1A1, 1A3, 1A6, 1A7, 1A8, 1A9, 1A10, or 2B7, were
from Gentest (Woburn, MA). Protein concentrations were measured
according to Bradford (18) with bovine serum albumin as standard.
The concentration of active cytochrome P450 was determined by the
method of Omura and Sato (19).
Microsomal Incubations. In vitro glucuronidation of curcuminoids
was carried out in a total volume of 0.2 mL of 0.1 M phosphate buffer
pH 7.4 containing 0.2 mg of hepatic microsomal protein. In a typical
incubation, the enzyme protein was first mixed with 5 µg of alamethicin
in about 50 µL of buffer and placed on ice for 15 min. Subsequently,
magnesium chloride (final concentration 10 mM), the â-glucuronidase
inhibitor saccharolactone (10 mM), UDPGA (20 mM), and finally the
curcuminoid (100 µM) dissolved in DMSO (final concentration 2%)
were added and the mixture was incubated at 37 °C for up to 60 min.
The workup procedure for HPLC analysis was either centrifugation of
the protein for 3 min at 1000g or addition of 0.2 mL of 0.7 M glycine/
chloride buffer pH 1.2 to yield pH 1.8 of the mixture, which was
subsequently extracted three times with 0.4 mL of ethyl acetate. The
extract was evaporated to dryness under reduced pressure and dissolved
in methanol. Control incubations were conducted without UDPGA. In
some experiments, saccharolactone was omitted, and an aliquot of the
incubation mixture was mixed with the same volume of 0.15 M acetate
buffer pH 5.0 containing 5000 Fishman units of â-glucuronidase per
milliliter and incubated for 2 h at 37 °C prior to extraction with ethyl
acetate and HPLC analysis.
For studies on the effects of curcuminoid glucuronides on micro-
tubule proteins, larger amounts of the glucuronides were required. As
attempts to scale up the 0.2 mL incubations failed, multiple parallel
0.2 mL incubations were conducted using protein concentrations of up
to 6 mg/mL and incubation times of 90 min. Ethyl acetate extracts
were combined and evaporated to dryness under reduced pressure, and
glucuronides were dissolved in 25 µL of DMSO. Their concentration
was in the low millimolar range as determined by HPLC analysis of
an aliquot diluted with methanol.
To facilitate the elucidation of the chemical structures and the
reactivities of the glucuronides, several non-natural analogs of
curcumin and their hexahydro derivatives were synthesized and
included in this study (Figure 1). Iso-curcumin is a positional
isomer of curcumin in which the hydroxyl and methoxyl group
at each aromatic ring have switched positions, whereas both
phenolic hydroxyl groups of curcumin are replaced by methoxyl
groups in O,O-dimethyl-curcumin. We have found that the
phenolic hydroxyl group is the preferred site of glucuronidation
for all curcuminoids. However, several of the phenolic glucu-
ronides, including that of curcumin, exhibited unexpected
reactivity and lipophilicity, which was strongly dependent on
the exact position of the phenolic groups and the presence of
olefinic double bonds in the aliphatic chain.
MATERIALS AND METHODS
Chemicals, Animals, and Cell Fractions. Demethoxy-curcumin was
isolated from turmeric by extraction, column chromatography, and
crystallization at the Arizona Center for Phytomedicine Research
(Tucson, AZ). It was >99% pure according to HPLC analysis.
Symmetrically substituted curcuminoids, i.e., curcumin, bisdemethoxy-
curcumin, iso-curcumin, and O,O-dimethyl-curcumin, were chemically
synthesized as described earlier (15), using the method of Pabon (16).
Briefly, acetylacetone was reacted with the appropriate substituted
benzaldehyde and boric acid anhydride as a catalyst. The final products
obtained after crystallization from methanol were >99% pure according
to HPLC analysis and exhibited the correct molecular ions in LC-MS
analysis. Iso-curcumin and O,O-dimethyl-curcumin are novel com-
pounds. Iso-curcumin had a melting point of 186 °C and a λmax at 419
nm with ꢀ ) 59200 in ethanol. The mass of its molecular ion as
determined by high-resolution electron impact mass spectrometry
(HREIMS) was 368.1257 (calculated for C21H20O6: 368.1260). The
1H and 13C NMR spectral data of iso-curcumin were as follows: 1H
NMR (500 MHz, acetone-d6): δ 3.92 (s, 6H), 6.05 (s, 1H), 6.71 (d,
2H, 3J ) 15.8 Hz), 7.03 (d, 2H, 3J ) 8.2 Hz), 7.17 (ddd, 2H, 4J ) 0.6
Hz, 4J ) 2.0 Hz, 3J ) 8.2 Hz), 7.23 (d, 2H, 4J ) 2.0 Hz), 7.60 (d, 2H,
3J ) 15.8 Hz), 7.85 (bs, 1H). 13C NMR (125 MHz, acetone-d6): δ
55.4 (q, 2C), 101.1 (d, 2C), 111.5 (d, 2C), 113.5 (d, 2C), 121.6 (d,
2C), 122.0 (d, 2C), 128.4 (s, 2C), 140.3 (d, 2C), 147.0 (s, 2C), 149.7
(s, 2C), 183.6 (s, 2C). A distinct difference between curcumin and iso-
curcumin was noted when hydroxide ions were added to a neutral
aqueous solution of the curcuminoid: whereas the curcumin solution
turned red, the solution of iso-curcumin remained yellow-orange due
to the meta positions of the hydroxyl groups which prevent their
participation in the conjugated π system. O,O-Dimethyl-curcumin had
a melting point of 128-130 °C and a λmax at 418 nm with ꢀ ) 56200
in ethanol. The mass of its molecular ion was 396.1571 (calculated for
1
C23H24O6: 396.1573). H NMR (400 MHz, CDCl3): δ 1.70 (bs, 1H),
3
3.92 (s, 6H), 3.94 (s, 6H), 5.82 (s, 1H), 6.50 (d, 2H, J ) 15.8 Hz),
3
4
6.88 (d, 2H, J ) 8.4 Hz), 7.08 (d, 2H, J ) 2.0 Hz), 7.14 (ddd, 2H,
4
3
3
4J ) 0.6 Hz, J ) 2.0 Hz, J ) 8.2 Hz), 7.61 (d, 2H, J ) 15.8 Hz).
13C NMR (100 MHz, CDCl3): δ 55.9 (q, 2C), 56.0 (q, 2C), 101.3 (d,
2C), 109.7 (d, 2C), 111.1 (d, 2C), 122.0 (d, 2C), 122.6 (d, 2C), 128.1
(s, 2C), 140.4 (d, 2C), 149.2 (s, 2C), 151.0 (s, 2C), 183.3 (s, 2C). Like
iso-curcumin, O,O-dimethyl-curcumin did not change its λmax when
the pH of the solution was raised.
For studies on the enzymatic activity of hepatic and intestinal
microsomes and of supersomes expressing human UGTs, incubations
were conducted in 50 mM Tris buffer pH 7.5, containing 0.05-1.0
mg of protein/mL and 20 µM curcumin or 4-(trifluoromethyl)-
umbelliferone (TFMU) for up to 2 h with linear product formation.
After being cooled to 0 °C, aliquots were directly analyzed by HPLC.
Glucuronides were quantified through their absorbance, assuming that
they have the same molar extinction coefficients as their aglycones.
NMR and HREIMS Analysis. 1H and 13C NMR spectra were
recorded in acetone-d6 or CDCl3 on a Bruker AM-400 and Bruker DRX-
500 spectrometer (Karlsruhe, Germany) at 400 or 500 MHz (1H) and
100 or 125 MHz (13C), using the residual solvent signal as internal
standard. Multiplicity determinations (DEPT 90, DEPT 135) were
acquired using standard Bruker pulse programs. HREIMS were obtained
From each curcuminoid, the respective hexahydro derivative was
obtained by hydrogenation in methanol with a Pd on charcoal catalyst
as previously described in detail (15). The chemical structures of the
synthetic hexahydro-curcuminoids, which had a purity of >95%
according to HPLC analysis, were confirmed by UV spectroscopy,
NMR spectroscopy, HREIMS analysis, LC-MS analysis, and GC-MS
analysis after trimethylsilylation (15). The novel compound hexahydro-
iso-curcumin had a λmax at 282 nm (ꢀ ) 7400 in methanol) and the
molecular ion had a mass of 374.1725 (calculated for C21H26O6:
374.1729). 13C NMR (100 MHz, CDCl3): δ 28.9 (t), 31.1 (t), 38.0 (t),
45.1 (t), 49.3 (t), 56.0 (q, 2C), 66.9 (d), 110.7 (d), 110.8 (d), 114.5 (d),
114.7 (d), 119.6 (d), 119.8 (d), 133.9 (s), 135.1 (s), 144.8 (s), 145.1
(s), 145.5 (s), 145.6 (s), 211.4 (s). The novel compound hexahydro-

540 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Pfeiffer et al.
by direct injection using a Finnigan MAT 95 instrument (Thermo
Finnigan, Austin, TX) with EI ionization at 70 eV.
HPLC Analysis. A HP 1100 system equipped with a binary pump,
a photodiode array detector, and HP Chemstation software for data
collection and analysis (Agilent Technologies, Waldbronn, Germany)
was used and separation was carried out on a 250 × 4.6 mm i.d., 5
µm, reversed-phase Prodigy 5ODS(2) column (Phenomenex, Torrance,
CA). Samples were dissolved in methanol and the injection volume
was 10-20 µL. Four different linear gradients of acetonitrile (solvent
B) in deionized water (solvent A, adjusted to pH 3.0 with formic acid)
were used for different curcuminoids. Gradient I: 0-35 min 30-70%
B; gradient II: 0-5 min 35% B, 5-30 min 35-55% B; gradient III:
0-6 min 15% B, 7-17 min 30-35% B, 17-32 min 35-45% B;
gradient IV: 0-35 min 30-80% B. The flow rate was 1 mL/min. The
curcuminoids were recorded at 420 nm and the hexahydro-curcuminoids
at 280 nm.
LC-MS/MS Analysis. The HPLC system was the same as described
above for HPLC analysis but with a 250 × 4.6 mm i.d., 4 µm, reversed-
phase Synergy Hydro column (Phenomenex). The eluate of the HPLC
column was split, one part being analyzed by the photodiode array
detector and the other part by an Agilent MSD-Trap-SL ion trap mass
spectrometer operated in the negative electrospray ionization (ESI)
mode. The flow rate of the nitrogen drying gas was 10 L/min, the
nebulizer pressure was set to 60 psi, the drying temperature was
maintained at 200 °C, the capillary voltage was 4500 V, and the
capillary current was 18.3 nA. MS/MS experiments were performed
in AutoMS(n) mode. The first scan was a full MS scan, and then one
or two precursor ions were selected, isolated, and selectively fragmented
in the ion trap.
Hydrogenation of Glucuronides. The glucuronides of various
curcuminoids extracted from the microsomal incubations with ethyl
acetate were dissolved in methanol and stirred with a small amount of
PtO2 under a H2 atmosphere at 20 °C for 1 h to yield the glucuronides
of the respective hexahydro-curcuminoids. The reaction mixture was
centrifuged (5 min at 1000g) and the supernatant analyzed by LC-MS/
MS.
Figure 2. HPLC profiles of the incubations of curcumin with UDPGA and
hepatic microsomes from (A) rats and (B) humans.
of UDPGA and the incubation mixture analyzed by HPLC, a
product eluting earlier than curcumin from the reversed-phase
column was observed (Figure 2A). This product was absent
when the incubation mixture was treated with â-glucuronidase
prior to HPLC and it was therefore concluded to represent a
glucuronide of curcumin. The same glucuronide A was formed
upon incubation with human hepatic microsomes; however, a
smaller peak of an even more polar product eluted from the
HPLC column (Figure 2B), which could also be cleaved by
â-glucuronidase.
Chemical Hydrolysis of Glucuronides. The stability of the glucu-
ronides and aglycones of various curcuminoids and of ketoprofen was
determined in 0.1 M potassium phosphate buffer pH 7.4 at 37 °C, using
30 µM concentrations of the test compounds. Thirty microliters of a
0.5 mM stock solution of the aglycone or the corresponding volume
of the concentrated glucuronide solutions in DMSO were mixed with
the prewarmed buffer to obtain a total volume of 500 µL, and 20 µL
aliquots were analyzed by HPLC at 0, 1, 2, 3, and 24 h after mixing.
Interaction with Microtubule Proteins. Microtubule proteins
(MTP) were prepared from fresh bovine brain according to Shelanski
et al. (20) and the inhibition of microtubule polymerization was
determined in a total volume of 500 µL of reassembly buffer containing
10 µM MTP as previously described (21). The aglycones and
glucuronides of the test compounds were added as DMSO solutions.
The final concentrations were 2% (v/v) for DMSO, 40-100 µM for
the free curcuminoids and ketoprofen, 9-45 µM for the glucuronides,
and 2 µM for colchicine. The test compounds were first incubated with
the MTP solution for 20 min at 20 °C, followed by the addition of
guanosine triphosphate (GTP) dissolved in reassembly buffer (final
concentration 0.5 mM) and warming to 35 °C to start the assembly of
MTP, which was measured by the increase in absorbance (due to
turbidity) at 350 nm for 30 min in 5 min intervals. Subsequently, the
assembled MTP was depolymerized by cooling to 4 °C for 30 min,
followed by a second and third cycle of assembly and disassembly.
Control incubations contained DMSO but no test compound. The
absorbance of the curcuminoids with a fully conjugated double bond
system was compensated by measuring against an incubation mixture
containing the curcuminoid but no GTP. Colchicine and ketoprofen
glucuronide, which are known inhibitors of MTP assembly, were used
as positive controls.
LC-MS/MS analysis of curcumin glucuronides A and B
showed that both products were monoglucuronides (Table 1).
Because curcumin is a symmetrical molecule, it was assumed
that one glucuronide carries the glucuronic acid at a phenolic
and the other one at the alcoholic hydroxyl group. As the
daughter ion spectra obtained by MS/MS analysis from the
molecular ions of curcumin glucuronide A and B were virtually
identical and dominated by the loss of glucuronic acid (Table
1), it was not possible to assign the proposed structures to the
two glucuronides.
Demethoxy-curcumin, bisdemethoxy-curcumin, and iso-cur-
cumin behaved like curcumin upon microsomal glucuronidation,
yielding two glucuronides with human hepatic microsomes and
one with rat hepatic microsomes. The MS and MS/MS data of
the glucuronides A and B of bisdemethoxy-curcumin and iso-
curcumin are listed in Table 1. Again, the daughter ion spectra
of the glucuronides A and B were very similar. As expected,
only one glucuronide was obtained when O,O-dimethyl-cur-
cumin was glucuronidated. In this case, the glucuronic acid must
be attached to the alcoholic hydroxyl group because this is the
only site available for glucuronidation. The MS/MS data of this
glucuronide were consistent with those of the others, exhibiting
the preferential loss of glucuronic acid (Table 1).
When hexahydro-curcumin was glucuronidated with human
or rat hepatic microsomes, only one glucuronide was formed.
In this case, LC-MS/MS analysis gave rise to a daughter ion
spectrum which was dominated by the loss of water from the
molecular ion (Table 1), strongly suggesting the presence of a
free alcoholic hydroxyl group and therefore a phenolic mono-
glucuronide. In contrast, the alcoholic glucuronide obtained from
hexahydro-O,O-dimethyl-curcumin did not release water upon
RESULTS AND DISCUSSION
Microsomal Glucuronides of Curcuminoids. When cur-
cumin was incubated with rat liver microsomes in the presence

Reactive Glucuronides of Curcuminoids
J. Agric. Food Chem., Vol. 55, No. 2, 2007 541
Table 1. Chromatographic and Mass Spectrometric Characteristics of Glucuronides of Various Curcuminoids Generated with Rat and Human Hepatic
Microsomes
negative ESI (m/z; % relative intensity)
retention time
(min)
in HPLCc
type of
yield of
glucuronidea
microsomesb
MS
MS/MS
glucuronide extractiond
-
-
-
-
-
-
-
-
-
-
-
curcumin A
curcumin B
bisdemethoxy-curcumin A
bisdemethoxy-curcumin B
iso-curcumin A
iso-curcumin B
O,O-dimethyl-curcumin
hexahydro-curcumin
RH, HH
HH
RH, HH
HH
RH, HH
HH
RH, HH
RH, HH
RH, HH
RH, HH
RH, HH
18.8 (gradient I)
8.8 (gradient I)
16.4 (gradient I)
8.3 (gradient I)
18.9 (gradient I)
13.0 (gradient I)
15.3 (gradient II) 571 [M
18.0 (gradient III) 549 [M
18.5 (gradient III) 489 [M
18.2 (gradient III) 549 [M
11.3 (gradient IV) 577 [M
543 [M
−H]
−H]
−H]
−H]
−H]
−H]
−H]
−H]
−H]
−H]
−H]
451 (12), 393 (25), 367 (100), 353 (13), 217 (39), 175 (80)
367 (100), 217 (86)
363 (68), 307 (100), 257 (7), 175 (64)
363 (33), 307 (77), 215 (21), 175 (100)
543 (100), 525 (12), 367 (83), 175 (11)
543 (100), 367 (43)
phenolic
alcoholic
phenolic
alcoholic
phenolic
alcoholic
alcoholic
phenolic
phenolic
phenolic
alcoholic
98
543 [M
483 [M
483 [M
543 [M
543 [M
n.d.e
98
n.d.
73
n.d.
99
36
n.d.
45
571 (7), 395 (100), 175 (6)
549 (19), 531 (100), 513 (12), 373 (13), 359 (23), 175 (29)
489 (39), 471 (100), 313 (64), 175 (95)
549 (26), 531 (100), 373 (18)
hexahydro-bisdemethoxy-curcumin
hexahydro-iso-curcumin
hexahydro-O,O-dimethyl-curcumin
392 (5), 193 (100), 175 (6)
n.d.
a A and B denote type of glucuronide. b RH, rat hepatic; HH, human hepatic. c Gradients I
−
IV, see Materials and Methods. d Percent of curcuminoid glucuronide extracted
with ethyl acetate from the microsomal glucuronidation mixture after incubation for 60 min and addition of glycine/chloride buffer to adjust to pH 1.8. e n.d., not determined.
LC-MS/MS analysis (Table 1). Thus, only the phenolic and
alcoholic glucuronides of the hexahydro-curcuminoids, but not
of the curcuminoids, differ in their daugther ion spectra.
To elucidate the exact structures of the two monoglucuronides
of curcumin, glucuronide A was prepared using rat liver
microsomes, extracted from the incubation mixture with ethyl
acetate and chemically hydrogenated in methanol using a
platinum catalyst. The product of this reaction had the same
HPLC retention time and daughter ion spectrum as the mi-
crosomal glucuronide of hexahydro-curcumin, clearly showing
that curcumin glucuronide A is a phenolic glucuronide. Cur-
cumin glucuronide B must then be the alcoholic glucuronide.
The small amount of this minor glucuronide prevented hydro-
genation and mass spectrometric confirmation of the structure.
The similar pattern of the microsomal glucuronides of
curcumin, bisdemethoxy-curcumin, and iso-curcumin, i.e.,
predominant and rather unpolar glucuronides A and minor but
more polar glucuronides B (Table 1), implies that the glucu-
ronides A of bisdemethoxy-curcumin and iso-curcumin are also
phenolic glucuronides. This was confirmed by hydrogenation
of these glucuronides A and comparison of the hydrogenation
products with the phenolic glucuronides of the respective
hexahydro-curcuminoids by HPLC and LC-MS/MS analysis
(Table 1).
During the experiments aiming to obtain larger amounts of
the glucuronides for chemical derivatization, several unusual
features of the glucuronides of the curcuminoids were noted.
First, the glucuronides of the curcuminoids appeared to be more
readily extractable from the aqueous incubation mixture than
the glucuronides of the hexahydro-curcuminoids. As shown in
Table 1, virtually complete extraction of the glucuronides A
of curcumin and some of its analogs, but not of hexahydro-
curcuminoids, was achieved with ethyl acetate when the pH of
the aqueous phase was adjusted to 1.8. Second, several of the
curcuminoid glucuronides but not the hexahydro-curcuminoid
glucuronides appeared to be unstable. The stability and chemical
reactivity of the glucuronides was therefore studied in more
detail.
glucuronides obtained with rat liver microsomes. The nonste-
roidal anti-inflammatory drug ketoprofen and its acyl glucu-
ronide, which is known to be unstable (22), were included for
comparison. The compounds were kept in 0.1 M phosphate
buffer pH 7.4 at 37 °C and analyzed by HPLC after various
time periods. The results are depicted in Figure 3. As expected,
ketoprofen was quite stable for up to 24 h, whereas its
glucuronide decomposed with a half-life of about 1 h. With
curcumin, a rapid disappearance from the aqueous solution was
observed for both the parent compound and its phenolic
glucuronide; after 1 h, more than 90% of the initial amount
had degraded, and the glucuronide appeared to be even more
unstable than the unconjugated curcumin. For bisdemethoxy-
curcumin, there was a clear difference between the unconjugated
and the glucuronidated form, the glucuronide decomposing
much faster than the aglycone. In contrast, no difference in
stability was observed between the glucuronides and parent
compounds of iso-curcumin and hexahydro-curcumin. However,
only hexahydro-curcumin and its glucuronide were completely
stable, whereas iso-curcumin and its glucuronide decomposed
slowly. Thus, pronounced differences in stability were obvious
for curcuminoids and their phenolic glucuronides as a result of
small alterations of the substituents at the aromatic rings. One
of the obvious questions is the nature of the degradation products
of the glucuronides. According to HPLC analysis, the aqueous
solutions of the glucuronides of ketoprofen, curcumin, and
bisdemethoxy-curcumin after 1 and 2 h did not contain the
parent compounds, and none of the known degradation products
of curcumin were detected in the solution of curcumin glucu-
ronide. Therefore, the decomposition of the curcuminoid glu-
curonides is probably not a simple hydrolysis of the glycosidic
bond, and further studies are needed to clarify the mechanism
of degradation.
To find out whether the unstable glucuronides are reactive
toward proteins, the glucuronides of various curcuminoids were
probed for their effects on microtubule proteins (MTP). MTP
consist predominantly of heterodimers of R- and â-tubulin as
well as smaller amounts of microtubule-associated proteins.
Stability and Reactivity of Curcumin Glucuronide. Cur-
cumin itself has long been known to degrade in aqueous solution
at pH values above 7, and some but not all of the degradation
products have been identified (15). Instability of curcumin
glucuronide has not been reported before. We have therefore
studied the stability of various curcuminoids and their major
Under cell-free conditions in the presence of GTP and Mg2+
,
MTP assemble at 35 °C to form microtubules, which disas-
semble again to MTP at 4 °C. It is known that chemical
modification of MTP, e.g., by reaction with the acyl glucu-
ronides of nonsteroidal anti-inflammatory drugs, causes MTP
to lose their ability to form microtubules (23). With native

542 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Pfeiffer et al.
Figure 4. Effect of curcumin and its phenolic glucuronide on the
polymerization of MTP under cell-free conditions. The second and third
cycle of assembly and disassembly are depicted.
Figure 5. Inhibitory effect of various curcuminoids and their phenolic
glucuronides on the polymerization of MTP under cell-free conditions. The
known MTP inhibitors colchicine and ketoprofen glucuronide were used
as positive controls. All experiments were independently repeated three
times. Error bars calculated from experiments with identical concentrations
represent standard deviations. Columns without error bars are from one
experiment, but were supported by two other experiments with similar
but not identical concentrations.
Figure 3. Stability of various curcuminoids and their phenolic glucuronides
in aqueous phosphate buffer at pH 7.4. Ketoprofen and its acyl glucuronide
were included for comparison. Data represent mean of two independent
experiments.
the unconjugated parent compounds did not affect MTP. This
suggests that the unstable curcuminoid glucuronides are reactive
toward proteins.
MTP, a second and third cycle of assembly and disassembly
can be conducted without adding new GTP by renewed warming
and cooling. The assembly and disassembly causes changes in
turbidity and can be measured by light scattering at 350 nm.
Because curcuminoids with a conjugated double bond system
and their glucuronides absorb light at 350 nm, the absorbance
of the complete polymerization mixture was measured against
a reference containing the same components but no GTP. A
typical assembly and disassembly of MTP in the presence of
curcumin and its phenolic glucuronide is depicted in Figure 4.
Whereas unconjugated curcumin had no effect on MTP polym-
erization, about 50% inhibition of MTP assembly was observed
with 40 µM curcumin glucuronide. About the same inhibitory
effect was obtained with the same concentration of ketoprofen
glucuronide, whereas the strong microtubule inhibitor colchicine
was about 10 times more effective (Figure 5). Among the tested
curcuminoids, the phenolic glucuronide of bisdemethoxy-
curcumin proved to be the strongest inhibitor and iso-curcumin
glucuronide the weakest one, whereas the glucuronide of
hexahydro-curcumin was devoid of any inhibitory effect (Figure
5). Thus, the inhibitory activities of the phenolic glucuronides
of various curcuminoids for MTP assembly paralleled their
instability in aqueous solution as shown in Figure 3, whereas
Although the formation of curcumin glucuronide in vivo and
in vitro has been reported before (9, 11-14), these publications
do not mention the chemical reactivity and unusual lipophilicity
of the glucuronide. Moreover, our study has, for the first time,
disclosed that two curcumin glucuronides are formed and
unambigously clarified their exact chemical structures, showing
that the phenolic hydroxyl group is the preferred site of
glucuronidation. Prior to the rigorous structure elucidation based
on daughter ion mass spectrometry, we were misled by the
chemical reactivity of the major glucuronide, which could be
more readily explained by assuming the structure of an alcoholic
glucuronide (24, 25). The alcoholic glucuronide of curcumin
represents a vinylogous acyl glucuronide. Acyl glucuronides are
formed from several nonsteroidal anti-inflammatory drugs, e.g.,
ketoprofen, ibuprofen, and diclofenac, and other carboxylic acid-
bearing compounds, and have been amply demonstrated to be
labile electrophiles (22).
With the chemical structure of curcumin glucuronide A
established, the chemical instability of this phenolic glucuronide
in aqueous solution and its reactivity with proteins remain to
be explained. Comparison with the glucuronides of other
curcuminoids used in this study clearly show that two prereq-
uisites must be fulfilled for reactivity, i.e., a para position of

Reactive Glucuronides of Curcuminoids
J. Agric. Food Chem., Vol. 55, No. 2, 2007 543
after very high doses of several grams per day (28). In
conclusion, our study has demonstrated that the glucuronides
of curcuminoids, which are major phase II metabolites, exhibit
novel properties that should be further investigated.
Table 2. Glucuronidation of Curcumin by Microsomes Expressing
Human Recombinant UGT Isoforms and by Microsomes from Various
Tissues
enzymatic activity (pmol/min/mg protein)
microsomes
for curcumin
for TFMUa
ABBREVIATIONS
UGT1A1
UGT1A3
UGT1A6
UGT1A7
UGT1A8
UGT1A9
UGT1A10
UGT2B7
human hepatic
human intestinal
rat hepatic
rat intestinal
1858
540
20
187
1537
97
1535
306
4641
12687
4589
3933
±
±
±
±
±
±
±
±
±
±
±
±
77
14
2
4
146
15
92
426
422
3886
8735
341
±
±
±
±
±
±
±
±
±
±
±
±
33
49
DMSO, dimethyl sulfoxide; ESI, electrospray ionization;
GTP, guanosine triphosphate; HREIMS, high-resolution electron
impact mass spectrometry; MTP, microtubule proteins; TFMU,
4-(trifluoromethyl)umbelliferone; UDPGA, uridine-5′-diphos-
phoglucuronic acid; UGT, uridine-5′-diphosphoglucuronosyl-
transferase.
227
306
24
278
35
108
1361
499
859
640
5547
421
38
1537
24079
10326
29654
13817
126
1138
170
104
LITERATURE CITED
(1) Scartezzini, P.; Speroni, E. Review on some plants of Indian
traditional medicine with antioxidant activity. J. Ethnopharmacol.
2000, 71, 2.
(2) Ammon, H. P.; Wahl, M. A. Pharmacology of Curcuma longa.
Planta Med. 1991, 57, 1.
a TFMU represents a nonspecific substrate for various UGT isoforms and was
used as reference.
(3) Sharma, R. A.; Gescher, A. J.; Steward, W. P. Curcumin: the
story so far. Eur. J. Cancer 2005, 41, 1955.
(4) Govindarajan, V. S. Turmeric - chemistry, technology, and
quality. Crit. ReV. Food Sci. Nutr. 1980, 12, 199.
(5) Aggarwal, B. B.; Kumar, A.; Bharti, A. C. Anticancer potential
of curcumin: preclinical and clinical studies. Anticancer Res.
2003, 23, 636.
(6) Sharma, O. P. Antioxidant activity of curcumin and related
compounds. Biochem. Pharmacol. 1976, 25, 1811.
the phenolic hydroxyl group to the aliphatic chain, and the
presence of a conjugated double bond system in the aliphatic
chain. The glucuronides of curcuminoids lacking one of these
structural requirements, e.g., iso-curcumin with meta-positioned
phenolic groups or hexahydro-curcumin with a saturated
aliphatic chain, are markedly more stable and less reactive. In
contrast, the phenolic glucuronides A of the natural congeners
demethoxy- and bisdemethoxy-curcumin, which meet the
requirements, exhibit the same or even higher reactivity as
compared to that of curcumin glucuronide A. Recent high-level
ab initio molecular orbital calculations have been conducted on
curcumin and have shown that curcumin has numerous elec-
trophilic regions over the entire molecule and a low molecular
hardness, implying electrophilic reactivity (26). Possibly similar
calculations for the glucuronide A of curcumin as well as for
other curcuminoids and their glucuronides might help to better
understand why some curcuminoids and/or their phenolic
glucuronides are stable whereas others are not. Furthermore,
the synthesis and study of other curcuminoids, e.g., a “mixed
curcumin” carrying one curcumin-type and one iso-curcumin-
type aromatic ring, and their glucuronides might prove useful.
UGT Isoforms Involved in the Glucuronidation of Cur-
cumin. No information is available to date about the activity
of different isoforms of UGTs for curcuminoids. We have
therefore determined the enzymatic activity of eight human
UGTs for the in vitro glucuronidation of curcumin (Table 2).
All UGTs gave rise to the formation of the phenolic glucuronide
A of curcumin, but with some isoforms, small amounts of
glucuronide B were also obtained. For the calculation of the
enzymatic activity, the amounts of glucuronide A and B were
added. The nonspecific UGT substrate TFMU was included in
this experiment for comparison. In addition to human UGT
isoforms, hepatic and intestinal microsomes from humans and
rats were assayed for their glucuronidation activity.
(7) Yang, F.; Lim, G. P.; Begum, A. N.; Ubeda, O. J.; Simmons,
M. R.; Ambegaokar, S. S.; Chen, P. P.; Kayed, R.; Glabe, C.
G.; Frautschy, S. A.; Cole, G. M. Curcumin inhibits formation
of amyloid beta oligomers and fibrils, binds plaques, and reduces
amyloid in vivo. J. Biol. Chem. 2005, 280, 5892.
(8) Lim, G. P.; Chu, T.; Yang, F.; Beech, W.; Frautschy, S. A.;
Cole, G. M. The curry spice curcumin reduces oxidative damage
and amyloid pathology in an Alzheimer transgenic mouse. J.
Neurosci. 2001, 21, 8370.
(9) Holder, G. M.; Plummer, J. L.; Ryan, A. J. The metabolism and
excretion of curcumin (1,7-bis-(4-hydroxy-3-methoxyphenyl)-
1,6-heptadiene-3,5-dione) in the rat. Xenobiotica 1978, 8, 761.
(10) Ravindranath, V.; Chandrasekhara, N. Metabolism of curcumin
- studies with [3H]curcumin. Toxicology 1982, 22, 337.
(11) Pan, M. H.; Huang, T. M.; Lin, J. K. Biotransformation of
curcumin through reduction and glucuronidation in mice. Drug
Metab. Dispos. 1999, 27, 486.
(12) Ireson, C.; Orr, S.; Jones, D. J.; Verschoyle, R.; Lim, C. K.;
Luo, J. L.; Howells, L.; Plummer, S.; Jukes, R.; Williams, M.;
Steward, W. P.; Gescher, A. Characterization of metabolites of
the chemopreventive agent curcumin in human and rat hepato-
cytes and in the rat in vivo, and evaluation of their ability to
inhibit phorbol ester-induced prostaglandin E2 production.
Cancer Res. 2001, 61, 1058.
(13) Ireson, C. R.; Jones, D. J.; Orr, S.; Coughtrie, M. W.; Boocock,
D. J.; Williams, M. L.; Farmer, P. B.; Steward, W. P.; Gescher,
A. J. Metabolism of the cancer chemopreventive agent curcumin
in human and rat intestine. Cancer Epidemiol. Biomarkers PreV.
2002, 11, 105.
The human UGT isoforms with the highest activity for
curcumin were UGT1A1, 1A8, and 1A10 (Table 2). High levels
of all three UGTs are expressed in the human gastrointestinal
tract (27). In fact, the activity for the glucuronidation of
curcumin was 3-fold higher in human intestinal microsomes than
in human hepatic microsomes, whereas no such difference was
noted for the rat intestinal and hepatic microsomes (Table 2).
Thus, curcumin may be efficiently glucuronidated during
absorption, and the lipophilicity of the glucuronide may facilitate
its distribution into various organs, possibly explaining, in part,
the extremely low plasma levels of curcumin observed even
(14) Asai, A.; Miyazawa, T. Occurrence of orally administered
curcuminoid as glucuronide and glucuronide/sulfate conjugates
in rat plasma. Life Sci. 2000, 67, 2785.
(15) Hoehle, S. I.; Pfeiffer, E.; Solyom, A. M.; Metzler, M.
Metabolism of curcuminoids in tissue slices and subcellular
fractions from rat liver. J. Agric. Food Chem. 2006, 54, 756.
(16) Pabon, H. J. J. A synthesis of curcumin and related compounds.
Recl. TraV. Chim. Pays-Bas. 1964, 83, 379.
(17) Lake, B. G. Preparation and characterization of microsomal
fractions for studies on xenobiotic metabolism. In Biochemical
Toxicology, A Practical Approach; Snell, K., Mulloch, B., Eds.;
IRL Press: Oxford, U.K., 1987; pp 183.

544 J. Agric. Food Chem., Vol. 55, No. 2, 2007
Pfeiffer et al.
(18) Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utulizing the principle of
protein-dye binding. Anal. Biochem. 1976, 72, 248.
(19) Omura, T.; Sato, R. The carbon monoxide-binding pigment of
liver microsomes. I. Evidence for its hemoprotein nature. J. Biol.
Chem. 1964, 239, 2370.
(20) Shelanski, M. L.; Gaskin, F.; Cantor, C. R. Microtubule assembly
in the absence of added nucleotides. Proc. Natl. Acad. Sci. U.S.A.
1973, 70, 765.
(21) Pfeiffer, E.; Metzler, M. Interaction of p-benzoquinone and
p-biphenoquinone with microtubule proteins in vitro. Chem. Biol.
Interact. 1996, 102, 37.
(22) Wang, J.; Davis, M.; Li, F.; Azam, F.; Scatina, J.; Talaat, R. A
novel approach for predicting acyl glucuronide reactivity via
Schiff base formation: development of rapidly formed peptide
adducts for LC/MS/MS measurements. Chem. Res. Toxicol. 2004,
17, 1206.
(25) Pfeiffer, E.; Walch, S. G.; Riess, A.; Hoehle, S. I.; Solyom, A.
M.; Metzler, M. Curcumin glucuronide inhibits cell-free micro-
tubule assembly. Toxicologist 2006, 90, 139.
(26) Balasubramanian, K. Molecular orbital basis for yellow curry
spice curcumin’s prevention of Alzheimer’s disease. J. Agric.
Food Chem. 2006, 54, 3512.
(27) Tukey, R. H.; Strassburg, C. P. Human UDP-glucuronosyltrans-
ferases: metabolism, expression, and disease. Annu. ReV. Phar-
macol. Toxicol. 2000, 40, 581.
(28) Cheng, A. L.; Hsu, C. H.; Lin, J. K.; Hsu, M. M.; Ho, Y. F.;
Shen, T. S.; Ko, J. Y.; Lin, J. T.; Lin, B. R.; Ming-Shiang, W.;
Yu, H. S.; Jee, S. H.; Chen, G. S.; Chen, T. M.; Chen, C. A.;
Lai, M. K.; Pu, Y. S.; Pan, M. H.; Wang, Y. J.; Tsai, C. C.;
Hsieh, C. Y. Phase I clinical trial of curcumin, a chemopreventive
agent, in patients with high-risk or pre-malignant lesions.
Anticancer Res. 2001, 21, 2895.
(23) Bailey, M. J.; Worrall, S.; de Jersey, J.; Dickinson, R. G.
Zomepirac acyl glucuronide covalently modifies tubulin in vitro
and in vivo and inhibits its assembly in an in vitro system. Chem.
Biol. Interact. 1998, 115, 153.
(24) Pfeiffer, E.; Hoehle, S. I.; Solyom, A. M.; Timmermann, B. N.;
Metzler, M. Stable and unstable glucuronides in curcumin
metabolism. Toxicol. Sci. 2003, 72, 313.
Received for review August 12, 2006. Revised manuscript received
October 6, 2006. Accepted October 6, 2006. This work was supported,
in part, by NIH Grant P50 AT00474 to the Arizona Center for
Phytomedicine Research.
JF0623283