In vitro metabolism and disposition of honokiol in rat and human livers

Article abstract of DOI:10.1002/jps.22536

The biotransformation of honokiol, a major constituent of the bark of Magnolia officinalis, was investigated in rat and human livers. When isolated, rat livers were perfused with 10μM honokiol and two metabolites, namely hydroxylated honokiol conjugated with glucuronic and sulfuric acid (M1) and honokiol monoglucuronide (M2), were quantified in bile and perfusate by high-performance liquid chromatography. The hepatic extraction ratio and clearance of honokiol was very high in rat liver (E: 0.99 ± 0.01 and 35.8 ± 0.04mL/min, respectively) leading to very low bioavailability (F = 0.007 ± 0.001). M2 formation was also highly efficient in human liver microsomes [V_max/K_m = 78.1 ± 6.73 μL/(min mg)], which appeared to be catalyzed mainly by UDP-glucuronosyltransferases 1A1, A3, 1A8, and 1A10, indicating hepatic and extrahepatic glucuronidation. Monosulfation of honokiol to the minor metabolite honokiol monosulfate [V_max/K_m = 27.9 ± 4.33 μL/(min mg)] by human liver cytosol was less pronounced and is mediated by sulfotransferases 1A1* 1, 1A1* 2, 1A2, 1A3, 1B1, and 1E1. P450-mediated oxidation of honokiol by liver microsomes, however, was below detection limit. In summary, this study established that glucuronidation and sulfation are the main metabolic pathways for honokiol in rat and human liver, suggesting their major contribution to clearance in vivo.

Full text of DOI:10.1002/jps.22536

In Vitro Metabolism and Disposition of Honokiol in Rat
and Human Livers
1
1
2
3
¨
MICHAELA BOHMDORFER, ALEXANDRA MAIER-SALAMON, BARBARA TAFERNER, GOTTFRIED REZNICEK,
4
2
5
6
1
¨
¨
¨
THERESIA THALHAMMER, STEFFEN HERING, ANTJE HUFNER, WOLFGANG SCHUHLY, WALTER JAGER
1Department of Clinical Pharmacy and Diagnostics, University of Vienna, A-1090 Vienna, Austria
2Department of Pharmacology and Toxicology, University of Vienna, A-1090 Vienna, Austria
3Department of Pharmacognosy, University of Vienna, A-1090 Vienna, Austria
4Institute of Pathophysiology, Medical University of Vienna, A-1090 Vienna, Austria
5Institute of Pharmaceutical Sciences, Department of Pharmaceutical Chemistry, Karl-Franzens-University Graz,
A-8010 Graz, Austria
6Institute of Pharmaceutical Sciences, Department of Pharmacognosy, Karl-Franzens-University Graz, A-8010 Graz, Austria
Received 6 October 2010; revised 31 January 2011; accepted 14 February 2011
Published online 14 March 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.22536
ABSTRACT: The biotransformation of honokiol, a major constituent of the bark of Magnolia
officinalis, was investigated in rat and human livers. When isolated, rat livers were perfused
with 10 :M honokiol and two metabolites, namely hydroxylated honokiol conjugated with glu-
curonic and sulfuric acid (M1) and honokiol monoglucuronide (M2), were quantified in bile and
perfusate by high-performance liquid chromatography. The hepatic extraction ratio and clear-
ance of honokiol was very high in rat liver (E: 0.99 ± 0.01 and 35.8 ± 0.04 mL/min, respectively)
leading to very low bioavailability (F = 0.007 ± 0.001). M2 formation was also highly efficient
in human liver microsomes [Vmax/K
alyzed mainly by UDP-glucuronosyltransferases 1A1, A3, 1A8, and 1A10, indicating hepatic
and extrahepatic glucuronidation. Monosulfation of honokiol to the minor metabolite honokiol
m
= 78.1 ± 6.73 :L/(min mg)], which appeared to be cat-
monosulfate [Vmax/K
m
= 27.9 ± 4.33 :L/(min mg)] by human liver cytosol was less pronounced
∗ ∗
and is mediated by sulfotransferases 1A1 1, 1A1 2, 1A2, 1A3, 1B1, and 1E1. P450-mediated
oxidation of honokiol by liver microsomes, however, was below detection limit. In summary,
this study established that glucuronidation and sulfation are the main metabolic pathways
for honokiol in rat and human liver, suggesting their major contribution to clearance in vivo.
©
2011 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 100:3505–3516,
2
011
Keywords: honokiol; hepatic metabolism; hepatic clearance; biliary excretion; glucuronida-
tion; sulfation
anxiolytic.7,8 In addition, honokiol is a promising
chemotherapeutic agent possessing antitumor activ-
ity without considerable toxicity. Recent studies show
that honokiol inhibits angiogenesis in vitro and tu-
INTRODUCTION
ꢀ
ꢀ
ꢀ
Honokiol (3 ,5-di-(2-propenyl)-1,1 -biphenyl-2,2 -diol)
is a bioactive neolignan constituent of cortex mag-
noliae, derived from the Chinese medicinal plant
Magnolia officinalis. The compound has several
pharmacological properties; it is anti-inflammatory
9
mor growth in vivo, induces apoptosis in human
1
0
colorectal carcinoma RKO cells, induces caspase-
dependent apoptosis in B-cell chronic lymphocytic
1
,2
3
via inhibiting cyclooxygenase-2,
antioxidative,
1
1
leukemia cells, and inhibits in vitro and in vivo
4
5
6
antiplatelet, antiarrhythmic, antidepressant, and
growth of breast cancer through induction of apop-
1
2
tosis and cell cycle arrest.
Correspondence to: Walter J a¨ ger (Telephone: +43-1-4277-55576;
Fax: +43-1-4277-9555; E-mail: walter.jaeger@univie.ac.at)
Although much is known about the biological
effects of honokiol, its metabolic pathway has not
been identified. A study of the biotransformation of
Journal of Pharmaceutical Sciences, Vol. 100, 3506–3516 (2011)
©
2011 Wiley-Liss, Inc. and the American Pharmacists Association
3
506
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

METABOLISM OF HONOKIOL IN RAT AND HUMAN LIVER
3507
ꢀ
ꢀ
its isomer magnolol (5,5 -diallyl-2,2 -biphenyldiol)
demonstrated phase I and phase II metabolism.
After magnolol was orally administered to rats,
chemicals and solvents were of analytical grade and
were used without further purification.
Liver Perfusion
1
3
Hattori et al. showed the formation of tetrahydro-
magnolol, 5-[(E)-1-propenyl]-5 -propyl-1,1 -biphenyl-
ꢀ
ꢀ
ꢀ
ꢀ
The study was approved by the Animal Experimenta-
tion Ethics Committee of the Austrian Ministry for
Education, Science and Culture. Male Wistar rats
(body weight = 275 ± 43.5 g; liver weight = 11.8
± 2.1 g) were purchased from the Department of
Animal Research and Genetics of the Medical Uni-
versity of Vienna, Austria. Animals were housed in
a temperature- and humidity-controlled room un-
der a 12-h light–dark cycle with free access to wa-
ter and food. Single-pass liver perfusion experiments
were carried out using the techniques previously
ꢀ
ꢀ
2
2
,2 -diol,
5-(2-propenyl)-5 -propyl-1,1 -biphenyl-
ꢀ
,2 -diol, isomagnolol, and 5-[(E)-1-propenyl]-5 -(2-
ꢀ
ꢀ
propenyl)-1,1 -biphenyl-2,2 -diol in the urine and
1
4
feces. An additional study by the same authors
exhibited rapid absorption of magnolol from the
gastrointestinal tract after oral administration of
1
4
[
ring- C]-magnolol to rats. Blood radioactivity levels
showed two peaks at 15 min and 8 h, suggesting
enterohepatic circulation of magnolol and its metabo-
lites. Radioactivity was distributed to a variety of
organs with [ring-14 C]-magnolol-2-O-glucuronide
being the major metabolite in the bile and blood.
In this study, we investigated the metabolism of
honokiol in rat and human livers by using liver mi-
crosomes, liver cytosol, and isolated perfused rat liver.
Furthermore, we identified the chemical structure of
honokiol metabolites by liquid chromatography–mass
spectrometry (LC/MS) and the isoenzymes responsi-
ble for their formation.
16
described. The total volume of the perfusion reser-
voir was 2500 mL. Perfusions were conducted using
a freshly prepared Krebs–Henseleit buffer (KHB) at
pH 7.4, equilibrated with 95% O2/5% CO2 during the
whole procedure at a constant flow rate of 36 mL/min.
The temperature of the perfusion cabinet and perfu-
◦
sion medium was thermostatically controlled at 37 C,
and the perfusion pressure was constantly monitored.
After 30 min of perfusion with KHB, stock solutions
of honokiol (10 mM in DMSO) were added to give
a final concentration of 10 :M. To test the stability
of honokiol under perfusion conditions, this solution
was applied in a control experiment to a perfusion
model without a liver. The viability of the liver was
assessed by its visual appearance and by checking
the perfusate flow. At the end of the experiment, the
MATERIALS AND METHODS
Materials
Honokiol (purity >98%) was obtained by demethyla-
tion (via Grignard reaction) of 4 -O-methylhonokiol.
Uridine 5 -diphosphoglucuronic acid (UDPGA;
ꢀ
15
◦
liver was dried, weighed, and frozen (−80 C) until
ꢀ
analysis.
9
3
8%–100% purity), dimethyl sulfoxide (DMSO),
-phosphoadenosine-5 -phosphosulfate (PAPS; 78%
Tissue Analysis
ꢀ
ꢀ
◦
purity), dithiothreitol (DTT), and the constituents
for the nicotinamide adenine dinucleotide phosphate
Liver tissue was thawed on ice at 4 C, weighed into
glass test-tubes (approx. 500 mg), mixed with a five-
fold amount of 50 mM potassium phosphate buffer
(pH 7.4), and then minced using an Ultra-Turrax
homogenizer (IKA, Staufen, Germany) four times for
10 s each on ice. Two milliliters of homogenates were
first centrifuged at 13,000 × g for 10 min, and pro-
teins were eliminated by a second centrifugation by
adding methanol (500 :L) to 250 :L of the super-
natant. Eighty microliters of the remaining super-
natant was injected onto the HPLC column.
(
NADPH)-generating system were obtained from
ꢀ
R
Sigma (Munich, Germany). Methanol and water were
of high-performance liquid chromatography (HPLC)
grade (Merck, Darmstadt, Germany). Human liver
microsomes (protein content 20 mg/mL) from one
male and two females (mean age 63 years, range
45–78 years), human liver cytosol (protein content
20 mg/mL) from one male and two females (mean age
65 years, range 53–78 years), baculovirus-infected
insect cells containing the cDNA for human UDP-
glucuronosyltransferases (UGTs) 1A1, 1A3, 1A4,
Determination of Honokiol and Its Metabolites
in Bile, Perfusate, and Liver Tissue
1A6, 1A7, 1A8, 1A9, 1A10, 2B4, 2B7, 2B15, and 2B17
were obtained from BD Biosciences (Woburn, Mas-
To study time-dependent first-pass metabolism, bile
(30–50 :L) and perfusate (approx. 1.5 mL) were col-
lected every 5 min for 90 min, immediately frozen
∗
sachusetts). Human sulfotransferases (SULTs) 1A1
∗
,
1
, 1A1 2 1A2, 1A3, 1B1, 1E1, and 2A1 expressed
in Escherichia coli were purchased from Cypex
Dundee, UK). The catalytic activities of human liver
◦
on dry ice, and stored at −80 C. Prior to analysis,
(
bile and perfusate samples were centrifuged (5000
× g for 5 min). Eighty microliters of diluted bile
samples (5 :L supernatant + 95 :L distilled water)
and 80 :L of perfusate supernatant were injected
microsomes, human liver cytosol, recombinant UGTs,
and recombinant SULTs were as described in the
data sheets provided by the manufacturers. All other
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

¨
BOHMDORFER ET AL.
3
508
onto the HPLC column. The determination of hon-
okiol and its metabolites was performed using an
UltiMate 3000 HPLC system (Dionex, Sunnyvale,
California) equipped with a Hypersil BDS-C18 col-
umn (5 :m, 250 × 4.6 mm ID; Thermo Fisher Scien-
tific, Inc., Waltham, Massachusetts), preceded by a
Hypersil BDS-C18 precolumn (5 :m, 10 × 4.6 mm
ID), at a flow rate of 1 mL/min and a column tem-
35, Brij 58, alamethicin, or digitonin. Then, magne-
sium chloride (10 mM), D-saccharic acid-1,4-lactone
(5 mM), and honokiol (final concentration 40 :M) in
DMSO were added to 50 mM potassium phosphate
buffer (pH 7.4) in a total volume of 100 :L. The glu-
curonidation assay was initiated by adding 3 mM
UDPGA, and the metabolism of honokiol was as-
sessed after 20-min incubation at 37 C as described
above. Control experiments in the absence of deter-
gent were run in parallel.
◦
◦
perature of 15 C. The mobile phase consisted of a
continuous linear gradient mixed from 5 mM ammo-
nium acetate/acetic acid buffer, pH 7.4 (mobile phase
A), and methanol (mobile phase B); honokiol and its
metabolites eluted according to their lipophilicity. The
mobile phase was filtered through a 0.45-:M filter
Glucuronidation of Honokiol by Human Liver
Microsomes and Recombinant UGT Isoenzymes
Human liver microsomes ( 50 :g of protein) and mi-
crosomes prepared from baculovirus-infected insect
cells (BTI-TN-5B1–4; 25 :g of protein) containing
the cDNA for human UGTs 1A1, 1A3, 1A4, 1A6, 1A7,
1A8, 1A9, 1A10, 2B4, 2B7, 2B15, and 2B17, respec-
tively, were preincubated with alamethicin (25 :g/mg
protein) on ice for 30 min. Next, magnesium chloride
(10 mM), D-saccharic acid-1,4-lactone (5 mM), and
honokiol (final concentrations 10–400 :M) in DMSO
were added to 50 mM potassium phosphate buffer (pH
7.4) in a total volume of 100 :L. Reactions were initi-
ated by the addition of 3 mM UDPGA, and the sam-
(
HVLP04700; Millipore, Vienna, Austria). The gradi-
ent ranged from 20% B at 0 min to 40% B at 10 min,
and linearly increased to 90% B at 22 min, at which
it remained constant for 5 min. The percentage of
methanol was decreased within 1 min to 20%, in order
to equilibrate the column for 7 min before application
of the next sample. Quantification of honokiol and its
metabolites were monitored at 294 nm. Calibration of
the chromatogram was accomplished using the exter-
nal standard method. As standards of the metabolites
were not available, quantification of metabolites was
based on the assumption that the unknown metabo-
lites had molar extinction coefficients similar to that
of honokiol. Linear calibration curves were obtained
from the peak areas of honokiol and its metabolites
to the external standard honokiol by spiking drug-
free bile with standard solutions of honokiol to give a
concentration range of 0.1–100 :g/mL (average corre-
lation coefficients were >0.99).
◦
ples were incubated for 20 min at 37 C. After the re-
actions were terminated by the addition of 200 :L
methanol, the samples were centrifuged (13000 × g
for 5 min). A total of 80 :L of the supernatant was
injected onto the HPLC column and analyzed under
conditions identical to those mentioned above. Glu-
curonidation of honokiol was not observed when using
insect cells infected with wild-type baculovirus (Auto-
grapha californica) and in the absence of UDPGA.
Enzyme kinetics for glucuronidation of honokiol by
recombinant UGTs was further evaluated by incuba-
tion of UGTs 1A1, 1A3, 1A8, and 1A10 ( with honokiol
(10–400 :M) under conditions identical to those men-
tioned above.
Phase I Metabolism of Honokiol by Human Liver
Microsomes
Microsomes (1 mg/mL), NADPH (1 mM), isocitric acid
(
5 mM), and isocitric dehydrogenase (0.5 U/mL) were
◦
preincubated for 5 min at 37 C in 0.05 M phosphate
buffer, pH 7.4 (final volume 1.0 mL). The reaction was
started by the addition of honokiol in DMSO (40 and
Sulfation of Honokiol by Human Liver Cytosol
and Recombinant Sult Isoenzymes
8
0 :M) and its metabolism was assessed by HPLC
◦
after 30-min incubation at 37 C.
Human liver cytosol (50 :g of protein)and cytosolic
extracts prepared from E. coli (containing cDNA for
Effect of Detergents on Honokiol Glucuronidation
∗
∗
human SULTs 1A1 1, 1A1 2, 1A2, 1A3, 1B1, 1E1,
and 2A1; 25 :g of protein) were incubated in a reac-
tion mixture containing 1–400 :M of honokiol (final
concentration), 50 :M PAPS, 7.4 mg/mL DTT, 0.05 M
potassium phosphate (pH 7.4),150 mM KCl, 50 mM
Tris (pH 7.5), and 2 mM ethylenediaminetetraacetic
acid in a total volume of 100 :L. PAPS and DTT
stock solutions were prepared fresh each day. Reac-
tions were initiated by adding PAPS, and the samples
UDP-glucuronosyltransferases proteins are located in
the endoplasmic reticulum and are sometimes en-
zymatically latent. Detergents can be used in order
to obtain the maximal activity by disrupting the in-
tegrity of the membranes.17 We examined the ef-
fect of low solubilizing concentrations of some com-
mon detergents on the catalytic activity of UGTs
involved in honokiol glucuronidation. Human liver
microsomes (final concentration 0.5 mg proteins per
milliliter) were preincubated on ice for 30 min with
◦
were incubated for 20 min at 37 C. After the reac-
tions had been terminated by the addition of 200 :L
methanol, the samples were centrifuged (13000 × g
for 5 min) and 80 :L of the supernatant was injected
2
5, 50, or 100 :g detergent per milligram microsomal
protein; the detergent used was Triton X-100, Brij
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011
DOI 10.1002/jps

METABOLISM OF HONOKIOL IN RAT AND HUMAN LIVER
3509
onto the HPLC column. Sulfation of honokiol was not
observed in the absence of PAPS and when using con-
trol cytosol from E. coli containing the vector only.
Enzyme kinetics for honokiol sulfation by SULTs
where E is the extraction rate (1−F).
The amount of honokiol and its metabolites ex-
creted into bile during 90 min of perfusion, expressed
as a percentage of the administered dose (fbile), was
calculated as:
∗
∗
1A1 1, 1A1 2, 1A2, 1A3, 1B1, and 1E1 were evalu-
ated by incubation of the recombinant enzymes with
honokiol (1–400 :M) under identical conditions as
those mentioned above.
XC × bile flow × liver weight × 100
f bile
(3)
Xin
Identification of Honokiol Metabolites
where Xin is the amount of honokiol applied to the
liver during 90 min by single-pass perfusion and XC
is the cumulative amount of honokiol and the metabo-
lites M1 and M2 in bile.
The amount of honokiol and its metabolites ex-
creted into perfusate during 90 min of perfusion, ex-
pressed as a percentage of the administered dose
Liquid chromatography–mass spectrometry analy-
ses of diluted bile samples (5 + 95 :L water), hu-
man liver microsomes, and human liver cytosol (40
+
60 :L methanol) were performed on an Ulti-
Mate 3000 RSLC-series system (Dionex, Germering,
Germany) coupled to a mass spectrometer equipped
with an orthogonal electrospray ionization source.
Negative ion polarity was used to acquire mass spec-
tra from m/z equals to 150–900 amu in scanning
mode. Metabolites hydroxylated honokiol conjugated
with glucuronic and sulfuric acid (M1) and honokiol
monoglucuronide (M2) were analyzed on an AB Sciex
API 4000 triple quadrupole mass spectrometer us-
ing the following parameters: -4.1kV caplillary volt-
age, 20 psi nebulizer (N2), 10 L/min dry gas flow (N2),
(
fperfusate), was then calculated as:
XC × 100
f perfusate =
(4)
Xin
where Xin is the amount of honokiol applied to the
liver during 90 min by single-pass perfusion and XC
is the cumulative amount of honokiol and the metabo-
lites M1 and M2 in perfusate.
Metabolic clearance of honokiol via its metabolites
M1 and M2 (ClMet) was calculated as:
◦
and 340 C dry temperature, and scan m/z 100–1000/s.
Metabolite honokiol monosulfate (M3) was identified
by a 3D quadrupole ion trap mass spectrometer (HCT;
Bruker Daltonics, Bremen, Germany) with the follow-
Xbile + Xperfusate
ClMet
=
× perfusate flow rate (5)
Xin
ing settings: 3.7 kV capillary voltage, 22 psi nebulizer
◦
(
N2), 8 L/min dry gas flow (N2), and 340 C dry tem-
where Xin is the amount of honokiol applied to the
liver during 90 min by single-pass perfusion, Xbile is
the cumulative concentration of the metabolites M1
and M2 in bile, and Xperfusate is the cumulative con-
centration of the metabolites M1 and M2 in perfusate.
Unless otherwise indicated, values are expressed as
mean value ± SD averaged from three individual ex-
periments.
perature. The HPLC column, injection volume, mobile
phase, gradient, and flow rate were identical to those
used for the analytical HPLC assay (see above).
Data Analysis
Pharmacokinetic parameters of the liver perfusion
were calculated using data recorded after steady-
state conditions were reached. Inflow and outflow
molar concentrations and biliary excretion were av-
eraged before inclusion into the pharmacokinetic
equations given below.
In the various glucuronidation and sulfation as-
says, each incubation step was performed at least in
triplicate, and the reSULTs were expressed as mean
±
SD. Formation rates calculated for the recombinant
The availability (F) of honokiol was calculated us-
ing the following equation:
enzymes were mathematically adjusted to the indi-
vidual catalyzing activities of liver microsomal and
liver cytosolic samples as given by the company. The
data were fitted to Michaelis–Menten (hyperbolic)
and substrate inhibition models. The coefficient of de-
Xout
F =
(1)
Xin
2
termination (R ) was used to determine the quality of
where Xin is the administered dose (amount of hon-
okiol applied to the liver during 90 min) and Xout is the
amount of honokiol excreted in the effluent perfusate
over 90 min.
The hepatic clearance (Cl) of honokiol was calcu-
lated as follows:
a fit to a specific model. Kinetic parameters were esti-
mated using Prism (version 5.0; GraphPad Software
Inc., San Diego, California) with Michaelis–Menten
(Eq. 6) and substrate inhibition kinetics (Eq. 7).
S
V = Vmax ×
(6)
(
Km + S)
Cl = E × perfusate flow rate
(2)
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

¨
BOHMDORFER ET AL.
3
510
Vmax
V = ꢀ
ꢁ
(7)
Km
S
Ki
1
+
+
S
where V is the rate of reaction, Vmax is the maximum
velocity, Km is the Michaelis constant, S is the sub-
strate concentration, and Ki is the inhibition constant.
The enzymatic efficacy, which is defined as the Vmax/
Km ratio, quantifies the glucuronidation capacity and
corresponds to the intrinsic clearance.
RESULTS
Metabolism and Disposition of Honokiol in Perfused
Rat Livers
After perfusion of isolated livers of male Wistar rats
with 10 :M honokiol for 90 min, native honokiol (re-
tention time TR = 26.9 min) as well as the two metabo-
lites M1 (TR = 16.8 min) and M2 (TR = 21.9 min) could
be detected in bile and perfusate (Fig. 1a). The ki-
netics of the excretion of honokiol and its metabo-
lites after application of honokiol are demonstrated
in Figures 2a–2c. Biliary secretion of unconjugated
honokiol reached an early steady state after 10 min,
with an average excretion rate of 8.23 ± 1.03 pmol/g
liver per minute. However, excretion of the metabo-
lites M1 and M2 showed a maximum after 45 min
(
mean values = 369 ± 107 and 798 ± 179 pmol/g liver
per minute, respectively) followed by a distinct de-
crease until the end of perfusion (mean values after
9
0 min = 180 ± 40.5 and 493 ± 178 pmol/g liver per
minute, respectively). In contrast to the rapid steady
state seen in the bile, efflux of the parent drug into
perfusate reached a plateau only after 40 min (max-
imum value = 217 ± 116 pmol/g liver per minute).
Efflux of M1 and M2 continuously increased during
9
0 min of perfusion to 314 ± 87 and 1496 ± 690 pmol/
g liver per minute, respectively.
The cumulative biliary excretion of honokiol, M1,
and M2 was moderate, representing only 0.03 ± 0.01,
0
.87 ± 0.38, and 2.03 ± 0.36% (0.66 ± 0.13, 21.1
±
5.86, and 52.8 ± 15.7 nmol/g in liver per 90 min,
respectively), respectively, of the administered dose
of 10 :M (see Table 1; Fig. 3). The total amounts
of M1 and M2 excreted into effluent perfusate over
Figure 1. Representative HPLC chromatograms of hon-
okiol (HON) and its metabolites M1, M2, and M3
9
0 min were 4.5- and 1.5-fold lower compared with
(
1
a) in bile from Wistar rats after 60 min of perfusion with
0 :M honokiol, (b) in human liver microsomes after in-
cubation with 40 :M honokiol in the presence of uridine
their biliary excretion (4.53 ± 3.47, 34.4 ± 21.1 nmol/g
in liver per 90 min, respectively) indicating preferen-
tial excretion of metabolites into bile. The total efflux
of honokiol, however, was 25-fold higher than in bile
ꢀ
5
-diphosphoglucuronic acid, and (c) in human liver cytosol
after incubation with 40 :M honokiol in the presence of
ꢀ
ꢀ
(
16.5 ± 0.91 nmol/g in liver per 90 min). Remarkably,
3 -phosphoadenosine-5 -phosphosulfate.
the total amount of native honokiol but not of M1 and
M2 in the liver tissue of Wistar rats over 90 min of
perfusion was high, amounting up to 80.0 ± 5.5% of
the applied dose. Only 1.8 ± 0.29% and 10.4 ± 4.5%
of M1 and M2 were found in liver tissue, respectively
Pharmacokinetic parameters describing the dispo-
sition of honokiol in the liver of Wistar rats were
determined by standard techniques (Eqs. 1 and 2)
using data from effluent perfusate samples collected
(
Fig. 3).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011
DOI 10.1002/jps

METABOLISM OF HONOKIOL IN RAT AND HUMAN LIVER
3511
Table 1. Disposition of Honokiol and Its Metabolites M1 and M2 in the Isolated Perfused Wistar Rat Liver
Cl (mL/min)
F
E
fbile (%)
fperfusate (%)
fliver (%)
ClMet (mL/min)
M1
M2
Honokiol
–
–
–
–
–
–
0.87 ± 0.38
2.03 ± 0.36
0.03 ± 0.01
0.17 ± 0.11
1.26 ± 0.48
0.66 ± 0.11
1.80 ± 0.29
10.4 ± 4.51
80.0 ± 5.53
0.31 ± 0.12
0.92 ± 0.26
–
35.8 ± 0.04
0.007 ± 0.001
0.99 ± 0.01
Perfusions were performed with 10 :M honokiol for 90 min using a single pass model.
Cl, hepatic clearance of honokiol; F, bioavailability of honokiol; E, hepatic extraction rate of honokiol; ClMet, metabolic clearance.
Figure 3. Disposition of honokiol and its metabolites M1
and M2 in Wistar rats after 90 min perfusion with 10 :M
honokiol. Data are expressed as nmol/g liver per 90 min
±
SD (n = 3).
over 90 min after addition of honokiol. On the basis
of the extensive accumulation in liver tissue, hon-
okiol has an extremely low bioavailability (F: 0.007
±
0.001); more than 99% of the parent drug is ex-
tracted from the liver within the first liver passage
Table 1). Thus, the almost quantitative extraction re-
(
SULTs in a high clearance rate of 35.8 ± 0.04 mL/min,
which approached the average perfusate flow rates. A
proposed metabolic pathway of honokiol in the liver
of Wistar rats is shown in Figure 6.
Metabolism of Honokiol in Human Livers
Metabolism of honokiol was further assessed in
human livers. In contrast to rat liver, NADPH-
dependent formation of phase I biotransformation
products was below detection limit in human liver
microsomes. However, when human liver microsomes
and cytosol were incubated with honokiol in the pres-
ence of UDPGA and PAPS, respectively, two metabo-
lites, namely M2 and M3 could be identified (Figs.
1
b and 1c). ReSULTs from preliminary studies to
optimize the incubation conditions for honokiol glu-
curonidation and sulfation showed that formation of
M2 and M3 was linear with time up to 60 min at mi-
crosomal protein concentrations of 0.25–2 mg/mL. As
shown in Figure 4, both glucuronidation and sulfation
of honokiol by human livers exhibited substrate inhi-
bition with comparable Ki values of 163 ± 22.6 and
Figure 2. Time course of biliary excretion and efflux into
perfusate of (a) honokiol and its metabolites (b) M1 and
(
c) M2 in Wistar rats when 10 :M honokiol was applied to
the liver. Data are expressed as pmol/g liver per minute ±
SD (n = 3).
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

¨
BOHMDORFER ET AL.
3
512
Figure 4. Kinetics of formation of (a) M2 in human
liver microsomes and (b) M3 in human liver cytosol nor-
malized to protein content as a function of honokiol con-
centration. Liver microsomes and cytosol were incubated
Figure 5. Rate of (a) M2 formation in recombinant hu-
man UDP-glucuronosyltransferases (UGTs) and (b) M3 for-
mation in recombinant human sulfotransferases (SULTs)
normalized to protein content as a function of honokiol
concentration. Recombinant human UGTs and SULTs
◦
with honokiol for 20 min at 37 C in the presence of uri-
ꢀ
ꢀ
dine 5 -diphosphoglucuronic acid and 3 -phosphoadenosine-
5
◦
were incubated with 40 :M honokiol for 20 min at 37 C
ꢀ
-phosphosulfate, respectively. Data are expressed as mean
ꢀ
in the presence of uridine 5 -diphosphoglucuronic and
±
SD (n = 3).
ꢀ
ꢀ
3
-phosphoadenosine-5 -phosphosulfate, respectively. Data
are shown as mean ± SD (n = 3).
2
07 ± 88.2 :M, respectively. The Km value for hon-
okiol glucuronidation, however, was 14-fold higher
than that for sulfation (85.5 ± 10.2 vs. 6.14 ± 2.32 :M)
indicating a preferable sulfate conjugation at lower
substrate concentrations (see Tables 2 and 3). Fig-
ure 6 shows the proposed biotransformation pathway
of honokiol in human liver.
1
.01 pmol/(mg min)] with UGTs 1A1 and 1A8, respec-
tively. Only a minor amount (∼5%) of the metabolite
was formed with UGTs 2B17 and 2B7. UGTs 1A6 and
1
A9 exhibited low catalytic activity for the formation
of M2, each contributed to less than and equal to 1%
glucuronidation of the substrate. No detectable for-
mation of M2 was seen with UGTs 1A4 and 2B4.
A comparative kinetic analysis was performed us-
ing honokiol concentrations ranging from 1 to 400 :M
to investigate the catalytic activities of UGT1A1,
UGT1A3, UGT1A8, and UGT1A10 for honokiol glu-
curonidation. The apparent enzyme kinetic parame-
ters for M2 of the recombinant isoenzymes UGT1A1,
UGT1A8, and UGT1A10 were estimated by fitting to
the substrate inhibition model (Eq. 7), whereas the
formation of the honokiol glucuronide in UGT1A3
followed classical Michaelis–Menten kinetics (Eq. 6).
Data on kinetics are reported in Table 2. Interestingly,
the formation of honokiol glucuronide resulted in
Glucuronidation of Honokiol by Recombinant UGTs
Twelve recombinant human UGTs of family 1 and 2
were incubated with honokiol (40 :M) for 20 min after
pretreatment with the solubilizing agent alamethicin.
This initial screening showed that UGTs 1A1, 1A3,
1
A6, 1A7, 1A8, 1A9, 1A10, 2B7, 2B15, and 2B17 glu-
curonidated honokiol to a single metabolite to varying
extents (Fig. 5a). Among these isoforms, the forma-
tion of the metabolite was preferentially catalyzed by
UGTs 1A10 [35.8%; 3293 ± 2.55 pmol/(mg min)] and
1
9
A3 [28.2%; 2601 ± 16.3 pmol/(mg min)], but reached
.8% [901 ± 7.50 pmol/(mg min)] and 9.3% [858 ±
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011
DOI 10.1002/jps

METABOLISM OF HONOKIOL IN RAT AND HUMAN LIVER
3513
Table 2. Kinetic Parameters of M2 Formation by Human Liver Microsomes and Recombinant Human UGT1A1, UGT1A3,
UGT1A8, and UGT1A10
Model
Km (:M)
Vmax (pmol/(mg min))
Vmax/Km (:L/(min mg))
Ki (:M)
Liver microsomes
Substrate inhibition
Substrate inhibition
Michaelis–Menten
Substrate inhibition
Substrate inhibition
85.5 ± 10.2
1.82 ± 0.43
3.67 ± 0.62
0.84 ± 0.46
1.32 ± 0.17
6677 ± 525
932 ± 19.9
3046 ± 52.5
747 ± 18.3
3592 ± 31.4
78.1 ± 6.73
512 ± 43.1
830 ± 16.7
893 ± 27.2
2729 ± 103
163 ± 22.6
1246 ± 202
NA
1
1
1
1
A1
A3
A8
A10
1709 ± 411
1392 ± 101
Data are shown as mean ± SD of three determinations.
NA, not applicable.
similar Vmax values for UGT1A3 and UGT1A10 (3046
and 3592 pmol/mg proteins per minute, respectively),
but showed a 2.8-fold higher Km value for UGT1A3
compared with UGT1A10. In addition, the intrinsic
clearance estimates (Vmax/Km) revealed 3.1- to 5.3-
fold higher values for UGT1A10 compared with those
for UGT isoenzymes 1A1, 1A3, and 1A8. The Ki es-
timates for UGTs 1A1, 1A10, and 1A8 were all high,
with values of 1246, 1392, and 1709 :M, respectively,
indicating only minor substrate inhibition.
Sulfation of Honokiol by Recombinant SULTs
The screening of seven recombinant human SULTs
of family 1 and 2 for honokiol sulfation activity re-
∗
1
∗2
vealed that SULTs 1A1 , 1A1 , 1A2, 1A3, 1B1, and
E1 are able to catalyze the sulfation of honokiol to
the monosulfate M3, however, to varying extents (Fig.
1
5
b). At 40 :M, honokiol sulfation was almost equally
∗
catalyzed by SULT1A1 1 [27.8%; 450 ± 7.88 pmol/
∗
(
(
(
mg min)], SULT1A1 2 [27.9%; 452 ± 6.13 pmol/
mg min)], and SULT1A2 [31.2%; 504 ± 11.8 pmol/
mg min)] but reached 7.65% [123 ± 0.89 pmol/(mg
min)] with SULT1B1. The isoforms SULT1A3 and
SULT1E1 exhibited only low catalytic activities for
M3 formation (≤4%). No formation of M3 was seen
with SULT2A1. A similar sulfation pattern by all
SULT isoforms was also seen at 5 :M honokiol (data
not shown).
A comparative kinetic analysis was performed us-
ing honokiol concentrations ranging from 1 to 400 :M
∗
to investigate the catalytic activities of SULTs 1A1
1
∗
, 1A1 2, 1A2, 1A3, 1B1, and 1E1 for honokiol sul-
fation (kinetic parameters are reported in Table 3).
∗
∗
M3 formation in SULTs 1A1 1, 1A1 2, 1A2, and
E1 exhibited distinct substrate inhibition (Eq. 7)
with Ki values ranging from 24.8 ± 11.3 to 262 ±
9.8 :M. In contrast, formation of M3 by SULTs 1A3
1
5
and 1B1 displayed Michaelis–Menten kinetics. Km
and Vmax values for SULTs 1A3, 1B1, and 1E1 were
∗
∗
much lower compared with SULTs 1A1 1, 1A1 2,
and 1A2. Intrinsic clearance values (Vmax/Km) quan-
tifying the sulfation capacity of the enzyme source,
Figure 6. Proposed metabolic pathway of honokiol in
rat and human liver.
∗
Table 3. Kinetic Parameters of M3 Formation by Human Liver Cytosol and Recombinant Human SULT1A1 1,
SULT1A1 2, SULT1A2, SULT1A3, SULT1B1, and SULT1E1
∗
Model
Km (:M)
Vmax (pmol/(mg min))
Vmax/Km (:L/(min mg))
Ki (:M)
Liver Cytosol
Substrate inhibition
Substrate inhibition
Substrate inhibition
Substrate inhibition
Michaelis–Menten
Michaelis–Menten
Substrate inhibition
6.14 ± 2.32
48.0 ± 17.4
42.1 ± 25.2
92.3 ± 39.4
1.07 ± 0.36
0.41 ± 0.19
0.85 ± 0.32
171 ± 24.4
1270 ± 295
1509 ± 520
1668 ± 504
23.5 ± 5.72
117 ± 32.1
59.5 ± 17.4
27.9 ± 4.33
26.5 ± 9.71
35.9 ± 13.8
18.1 ± 5.78
22.0 ± 6.91
285 ± 86.7
69.9 ± 21.9
207 ± 88.2
32.6 ± 10.7
24.8 ± 11.3
79.8 ± 20.4
NA
NA
262 ± 59.8
∗
1
1
1
1
1
1
A1 1
A1 2
A2
A3
B1
E1
∗
Data are shown as mean ± SD of three determinations.
NA, not applicable.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

¨
BOHMDORFER ET AL.
3
514
∗
∗
however, were as follows: 1B1 > 1E1 > 1A1 2 > 1A1
> 1A3 > 1A2.
monohydroxyhonokiol (M1) and a monoglucuronide
(M2). Low recovery of these compounds did not allow
further structural identification by NMR. Interest-
ingly, glucuronidation of honokiol leading to M2 for-
mation was clearly favored in the rat liver, amounting
up to 13.5% of the applied dose, whereas M1 forma-
tion seemed to be a minor metabolic pathway (3% of
the administered drug).
1
Identification of Honokiol Metabolites
Structural identification of M1–M3 was conducted by
LC/MS. Negative ion mass spectra of bile samples,
human liver microsomes, and human liver cytosol re-
vealed stable molecular ions at m/z 265, 537, 441, and
3
45 amu., in agreement with the molecular weights of
native honokiol, M1 (honokiol + 16 amu + 80 amu +
76 amu), M2 (honokiol + 176 amu), and M3 (honokiol
80 amu). Because of insufficient amounts of these
Extensive glucuronidation is in accordance with
1
4
studies of Hattori et al., who showed pronounced
glucuronidation and high accumulation of the struc-
tural isomer magnolol in rat liver. Pharmacokinetic
studies also revealed very similar half-lives of hon-
okiol and magnolol in rat plasma (49.22 ± 6.68
and 56.24 ± 7.30 min vs. 49.05 ± 5.96 and 49.58 ±
6.81 min) after intravenous administration of 5 and
10 mg/kg, respectively, indicating similar biotransfor-
1
+
metabolites in bile and human liver microsomes and
cytosol and a lack of commercially available stan-
dards, it was not possible to determine the exact
structure of honokiol metabolites by nuclear magnetic
resonance (NMR).
2
0,21
mation and disposition.
Analogous to rat liver, phase II metabolism was
also predominant in human liver with about threefold
higher intrinsic clearance values for monoglucuronide
(M2) formation than for monosulfation (M3). Metabo-
lite formation, however, was dose dependent as
sulfation clearly prevailed in the lower honokiol con-
centrations (1–40 :M). When the initial honokiol con-
centration further increased, total sulfate formation
in human liver cytosol dropped dramatically (200 :M
honokiol) based on noncompetitive substrate inhibi-
tion. This is in accordance with recent data from our
laboratory, in which we demonstrated that in human
liver cytosol and in human intestinal Caco-2 cells, sul-
fation of the natural polyphenolic compound resvera-
trol was also best characterized by the substrate in-
DISCUSSION
Honokiol is one of the main components of M. of-
1
8
ficinalis bark, which has been extensively used as
a traditional medicine in China and Japan. Recent
studies demonstrated that honokiol exhibits vari-
ous pharmacological properties without considerable
toxicity.
In the present study, hepatic biotransformation and
disposition of honokiol and its metabolites were in-
vestigated for the first time by perfusing isolated
male Wistar rat livers with 10 :M of honokiol. This
concentration was chosen based on daily intakes of
2
50–750 mg of an extract standardized for the pri-
2
2,23
mary active ingredients (typically 1%–2% honokiol
and magnolol). During 90 min the perfused liver ex-
tracted more than 99% of applied honokiol; steady-
state concentrations were achieved within 30 min af-
ter commencement of infusion. However, only 0.03%
of the administered dose was found in bile and
hibition model.
In contrast to rat liver, cytochrome P450-dependent
hydroxylation was below detection limit in human
liver indicating species-related differences in hon-
okiol metabolism.
Using recombinant UGT isoenzymes, we further
demonstrated that honokiol is mainly glucuronidated
by UGTs 1A1, 1A3, 1A8, and 1A10. The kinetic pro-
files of M2 formation by UGTs 1A1, 1A8, and 1A10
were similar, but that of UGT1A3 was markedly dif-
ferent. M2 formation by UGT1A3 was consistent with
classical Michaelis–Menten kinetics, with a 2.0- to
4.4-fold higher Km value compared with the isoforms
UGTs 1A1, 1A8, and 1A10, which displayed substrate
inhibition kinetics.
0.66% in the effluent perfusate. This is in accor-
dance with a previous study in rats, also demon-
strating very low plasma levels of unconjugated hon-
okiol (Cmax < 90 ng/mL) after rectal administration
of 13.5 mg/kg honokiol,19 indicating low bioavailabil-
ity based on extensive biotransformation. The high
hepatic clearance of honokiol (35.8 ± 0.04 mL/min)
may also be explained by its marked tendency to
accumulate in liver tissue (80.0% of the applied
dose).
On the contrary, hepatic biotransformation of hon-
okiol was high, amounting to 16.5% of the applied
dose; 22.4% of total metabolites were retained in the
liver tissue, 50.6% were excreted into bile, and 27%
were excreted into effluent perfusate. LC/MS analy-
sis of bile and perfusate revealed the formation of two
metabolites, namely a glucuronidated and sulfated
Our data also demonstrated hepatic as well as ex-
trahepatic glucuronidation of honokiol. UGTs 1A1
and 1A3 have been observed in the liver and small
intestine. UGTs 1A8 and 1A10 are not expressed in
the liver but are expressed in the small intestine and
2
4,25
colon.
All of these UGTs will therefore contribute
to the honokiol glucuronidation in human and rat
intestines.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011
DOI 10.1002/jps

METABOLISM OF HONOKIOL IN RAT AND HUMAN LIVER
3515
Analogously, human recombinant SULT isoforms
were screened for honokiol sulfation. We found that
several of these isoenzymes are responsible for the
conjugation of honokiol with sulfuronic acid. Although
SULTs 1A1 1, 1A1 2, and 1A2 demonstrated far
higher sulfation rates in vitro at 5 and 40 :M, in vivo
only SULTs 1A3, 1B1, and 1E1 may contribute sig-
nificantly to honokiol sulfation based on very low Km
values (0.41–1.07 :M).
CONCLUSION
Our data demonstrated that honokiol is extensively
metabolized in rat and human liver. Glucuronidation
and sulfation will therefore play a key role in the
elimination of dietary honokiol in humans after oral
uptake. Measurements of both parent compound and
metabolites will be necessary to correlate pharma-
cokinetic with pharmacological activity.
∗
∗
As demonstrated for glucuronidation, we could also
show hepatic as well as extrahepatic sulfation of hon-
okiol. SULT1A1 levels are very high in the liver and
ACKNOWLEDGMENTS
This study was supported by grants of the
2
6
is also present at lower levels in many other tissues,
¨
Jubil a¨ umsfonds der Osterreichischen Nationalbank
whereas lower mRNA levels have been found for
SULT1A2 than for other SULT1A members in the
liver, kidney, brain, ovary, and some other sections
of the gastrointestinal tract. As recently published
data also show the expression of SULT1A2 protein in
Caco-2 cells as well as in individual samples of the
liver and caecum,27,28 our study is indicative of the
physiological role of SULT1A2 in honokiol biotrans-
formation. SULT1A3 expression is extremely high in
the gut, detectable in many other extrahepatic tis-
(
12600) and Fonds zur wissenschaftlichen Forschung
(
FWF) (P21083-B11) to W. J a¨ ger.
REFERENCES
1
. Liou KT, Shen YC, Chen CF, Tsao CM, Tsai SK. 2003. The
anti-inflammatory effect of honokiol on neutrophils: Mecha-
nisms in the inhibition of reactive oxygen species production.
Eur J Pharmacol 475:19–27.
2
3
4
5
. Sch u¨ hly W, Khan SI, Fischer NH. 2009. Neolignans from
North American Magnolia species with cyclooxygenase-2 in-
hibitory activity. Inflammopharmacology 17:106–110.
. Shen CC, Ni CL, Shen YC, Huang YL, Kuo CH, Wu TS,
Chen CC. 2009. Phenolic constituents from the stem bark of
Magnolia officinalis. J Nat Prod 72:168–171.
. Teng CM, Chen CC, Ko FN, Lee LG, Huang TF, Chen YP, Hsu
HY. 1988. Two antiplatelet agents from Magnolia officinalis.
Thromb Res 50:757–765.
. Tsai SK, Huang CH, Huang SS, Hung LM, Hong CY. 1999.
Antiarrhythmic effect of magnolol and honokiol during acute
phase of coronary occlusion in anesthetized rats: Influence of
L-NAME and aspirin. Pharmacology 59:227–233.
2
6
sues, and essentially not expressed in the liver,
whereas SULT1E1 mRNA and protein have been
detected in liver, small intestine, and in hormone-
dependent tissues (e.g., endometrium and breast).
SULT1B1 mRNA has been primarily identified in the
2
9
liver, small intestine, colon, and blood leucocytes.
2
7
Teubner et al. reported a 17-fold higher expression
of SULT1B1 in the ileum than in the liver. During
absorption, honokiol may be sulfated by several in-
testinal SULT enzymes, which may contribute to the
biotransformation of honokiol.
6
. Xu Q, Yi LT, Pan Y, Wang X, Li YC, Li JM, Wang CP,
Kong LD. 2008. Antidepressant-like effects of the mixture of
honokiol and magnolol from the barks of Magnolia officinalis
in stressed rodents. Prog Neuropsychopharmacol Biol Psychi-
atry 32:715–725.
As observed for glucuronidation, kinetic profiles of
M3 formation by four out of six investigated SULT
isoforms also exhibited substrate inhibition, but with
distinct lower IC50-values of 25–262 :M. The ob-
served decrease in metabolite formation, especially
that of the sulfate conjugate at higher honokiol con-
centrations, may have a major impact on the oral
bioavailability of honokiol in vivo as saturation and
inhibition of M3 formation will not only occur in
the intestine but also in the liver. Concomitant oral
administration of SULT inhibitors like quercetin or
resveratrol with honokiol may substantially increase
its bioavailability.
7
8
9
. Maruyama Y, Kuribara H, Morita M, Yuzurihara M,
Weintraub ST. 1998. Identification of magnolol and honokiol as
anxiolytic agents in extracts of saiboku-to, an oriental herbal
medicine. J Nat Prod 61:135–138.
. Kuribara H, Kishi E, Hattori N, Okada M, Maruyama Y. 2000.
The anxiolytic effect of two oriental herbal drugs in Japan at-
tributed to honokiol from Magnolia bark. J Pharm Pharmacol
52:1425–1429.
. Bai X, Cerimele F, Ushio-Fukai M, Waqas M, Campbell PM,
Govindarajan B, Der CJ, Battle T, Frank DA, Ye K, Murad E,
Dubiel W, Soff G, Arbiser JL. 2003. Honokiol, a small molecu-
lar weight natural product, inhibits angiogenesis in vitro and
tumor growth in vivo. J Biol Chem 278:35501–35507.
It is not known yet whether honokiol conjugates
also exhibit pharmacological activity. However, like
the biologically inactive estrogen sulfate and glu-
curonide that is transformed by cellular sulfatases
or by $-glucuronidase into the biological bioactive
10. Wang T, Chen F, Chen Z, Wu YF, Xu XL, Zheng S, Hu X.
2004. Honokiol induces apoptosis through p53-independent
pathway in human colorectal cell line RKO. World J Gastroen-
terol 10:2205–2208.
11. Battle TE, Arbiser J, Frank DA. 2005. The natural
product honokiol induces caspase-dependent apoptosis in
B-cell chronic lymphocytic leukemia (B-CLL) cells. Blood
3
0
estradiol, honokiol conjugates may also serve as in-
active pools for honokiol. Furthermore, conjugation
might also prevent honokiol from enzymatic oxida-
tion, extending its half-life in the cell and maintaining
its biological properties.31
106:690–697.
1
2. Wolf I, O’Kelly J, Wakimoto N, Nguyen A, Amblard F,
Karlan BY, Arbiser JL, Koeffler HP. 2007. Honokiol, a nat-
ural biphenyl, inhibits in vitro and in vivo growth of breast
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011

¨
BOHMDORFER ET AL.
3
516
cancer through induction of apoptosis and cell cycle arrest. Int
J Oncol 30:1529–1537.
monolayers of the human intestinal Caco-2 cells is medi-
ated by inhibition and saturation of metabolites. Pharm Res
23:2107–2115.
1
3. Hattori M, Sakamoto T, Endo Y, Kakiuchi N, Kobashi K,
Mizuno T, Namba T. 1984. Metabolism of magnolol from mag-
noliae cortex. I. Application of liquid chromatography-mass
spectrometry to the analysis of metabolites of magnolol in rats.
Chem Pharm Bull 32:5010–5017.
4. Hattori M, Endo Y, Takebe S, Kobashi K, Fukasaku N, Namba
T. 1986. Metabolism of magnolol from Magnoliae cortex. II.
Absorption, metabolism and excretion of [ring-14C]magnolol
in rats. Chem Pharm Bull 34:158–167.
23. Miksits M, Maier-Salamon A, Aust S, Thalhammer T,
Reznicek G, Kunert O, Haslinger E, Szekeres T, Jaeger W.
2005. Sulfation of resveratrol in human liver: Evidence of a
major role for the sulfotransferases SULT1A1 and SULT1E1.
Xenobiotica 35:1101–1119.
24. King CD, Rios GR, Green MD, Tephly TR. 2000. UDP-
glucuronosyltransferases. Curr Drug Metab 1:143–161.
25. Nagar S, Remmel RP. 2006. Uridine diphosphoglucurono-
syltransferase pharmacogenetics and cancer. Oncogene
25:1659–1672.
1
1
5. El-Feraly FS, Li WS. 1978. Phenolic constituents of Magnolia
grandiflora L. seeds. Lloydia 41:442–449.
1
6. Thalhammer T, Stapf V, Gajdzik L, Graf J. 1994. Bile canalic-
ular cationic dye secretion as a model for P-glycoprotein medi-
ated transport. Eur J Pharmacol 270:213–220.
26. Glatt, H. 2000. Sulfotransferases in the bioactivation of xeno-
biotics. Chem Biol Interact 129:141–170.
27. Teubner W, Meinl W, Florian S, Kretzschmar M, Glatt H. 2007.
Identification and localization of soluble sulfotransferases
in the human gastrointestinal tract. Biochem J 404:207–
215.
28. Meinl W, Ebert B, Glatt H, Lampen A. 2008. Sulfotransferase
forms expressed in human intestinal Caco-2 and TC7 cells at
varying stages of differentiation and role in benzo[a]pyrene
metabolism. Drug Metab Dispos 36:276–283.
1
7. Soars MG, Ring BJ, Wrighton SA. 2003. The effect of
incubation conditions on the enzyme kinetics of udp-
glucuronosyltransferases. Drug Metab Dispos 31:762–767.
8. Fujita M, Itokawa H, Sashida Y. 1973. Studies on the compo-
nents of Magnolia obovata Thunb. 3. Occurrence of magnolol
and honokiol in M. obovata and other allied plants. Yakugaku
Zasshi 93:429–434.
1
1
9. Wu X, Chen X, Hu Z. 2003: High-performance liquid chromato-
graphic method for simultaneous determination of honokiol
and magnolol in rat plasma. Talanta 59:115–121.
29. Wang J, Falany JL, Falany CN. 1998. Expression and char-
acterization of a novel thyroid hormone-sulfating form of cy-
tosolic sulfotransferase from human liver. Mol Pharmacol
53:274–282.
30. Otake Y, Nolan AL, Walle UK, Walle T. 2000. Quercetin and
resveratrol potently reduce estrogen sulfotransferase activity
in normal human mammary epithelial cells. J Steroid Biochem
Mol Biol 73:265–270.
31. Pasqualini JR, Chetrite GS. 2005. Recent insight on the con-
trol of enzymes involved in estrogen formation and transfor-
mation in human breast cancer. J Steroid Biochem Mol Biol
93:221–236.
2
0. Tsai TH, Chou CJ, Cheng FC, Chen CF. 1994. Pharmacoki-
netics of honokiol after intravenous administration in rats
assessed using high-performance liquid chromatography. J
Chromatogr Biomed Appl 655:41–45.
2
1. Tsai TH, Chou CJ, Chen CF. 1996. Pharmacokinetics and
brain distribution of magnolol in the rat after intravenous bo-
lus injection. J Pharm Pharmacol 48:57–59.
22. Maier-Salamon A, Hagenauer B, Wirth M, Gabor F, Szekeres
T, J a¨ ger W. 2006. Increased transport of resveratrol across
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 8, AUGUST 2011
DOI 10.1002/jps