Metabolic fate of fraxin administered orally to rats

Article abstract of DOI:10.1021/np0580412

Naturally occurring fraxin (1) was administered orally to rats to investigate its metabolism. Urinary metabolites were analyzed by three-dimensional HPLC, and fraxetin-7-O-sulfate (2), fraxetin-7-O-β- glucuronide (3), fraxetin (4), 6,7,8-trihydroxycoumarin (5), and fraxidin (6) were isolated. Fraxin (1) was extensively metabolized to 4, which was partly metabolized to 5 in a rat fecal suspension after incubation for 24 h. Urinary excretion of 4 and 5 in rats administered orally with 1 was substantially reduced when the rats were treated with antibiotics to suppress their intestinal flora. Incubation of 1 with a rat liver S-9 mixture yielded 6. These results suggest that hydrolysis and demethylation of 1 are performed by intestinal microflora, while methylation occurs in the liver.

Full text of DOI:10.1021/np0580412

J. Nat. Prod. 2006, 69, 755-757
755
Metabolic Fate of Fraxin Administered Orally to Rats
Takaaki Yasuda, Mai Fukui, Takahiro Nakazawa, Ayumi Hoshikawa, and Keisuke Ohsawa*
Department of Phytochemistry, Tohoku Pharmaceutical UniVersity, 4-4-1, Komatsushima, Aoba-ku, Sendai,
Miyagi 981-8558, Japan
ReceiVed March 15, 2005
Naturally occurring fraxin (1) was administered orally to rats to investigate its metabolism. Urinary metabolites were
analyzed by three-dimensional HPLC, and fraxetin-7-O-sulfate (2), fraxetin-7-O-â-glucuronide (3), fraxetin (4), 6,7,8-
trihydroxycoumarin (5), and fraxidin (6) were isolated. Fraxin (1) was extensively metabolized to 4, which was partly
metabolized to 5 in a rat fecal suspension after incubation for 24 h. Urinary excretion of 4 and 5 in rats administered
orally with 1 was substantially reduced when the rats were treated with antibiotics to suppress their intestinal flora.
Incubation of 1 with a rat liver S-9 mixture yielded 6. These results suggest that hydrolysis and demethylation of 1 are
performed by intestinal microflora, while methylation occurs in the liver.
The dried bark of Fraxinus japonica Seringe Blume (Oleaceae)
is currently sold as the Kampo medicine “shinpi” in Japan. It has
traditionally been used as a diuretic, an antifebrile, an analgesic,
and an antirheumatic agent.1 The isolation of various coumarins,
such as fraxin (1), fraxetin (4), escletin, and esculin, from the bark
has been reported.2,3 Among these compounds, 1 is known to be
responsible for the diuretic and anti-inflammatory action of
“shinpi”.4-6 Although some biological effects of 1 have been
described, the metabolic fate of this compound has yet to be
reported. Here, we describe the identification of urinary metabolites
of fraxin (1) in rats. In addition, we report the biotransformation
of 1 using antibiotic-treated rats, rat fecal suspension, and a rat
liver S-9 mixture.
Scheme 1. Structure of 1 and Its Proposed Metabolic Pathway
in Rats
Results and Discussion
Two distinct HPLC peaks due to metabolites 2 and 3 were
detected in the untreated urine of rats orally administered with fraxin
(1), whereas three distinct HPLC peaks due to 4, 5, and 6 were
detected in â-glucuronidase/arylsulfatase-treated urine of rats orally
administered with 1. Unchanged 1 was not detected in either urine
sample. Compounds 2-6 were isolated from urine samples by
chromatography, as described in the Experimental Section.
Metabolite 2 was obtained as a white powder. Enzymatic
hydrolysis of 2 with arylsulfatase yielded 4, as confirmed by tR
agreement using HPLC. Intense absorption at 1049 cm-1 in the IR
spectrum and SO42- formation on carbonization suggested a sulfate-
conjugated structure for 2. Negative-ion FABMS of 2 showed a
base ion peak corresponding to (M - H)- at m/z 287 along with
fragment ion peaks at m/z 207 (M - H - SO3)-, thus indicating
one sulfate group in 2. A comparison of the 13C NMR spectrum of
2 with that of 4 showed that the C-7 signal of 2 has been shifted
6.6 ppm upfield, accompanied by downfield shifts of C-6 (4.1 ppm)
and C-8 (5.7 ppm). These shifts indicated a sulfate group at C-7.
On the basis of these data, 2 was determined to be fraxetin-7-O-
sulfate.
Metabolite 3 was obtained as a white powder. Enzymatic
hydrolysis of 3 with â-glucuronidase yielded 4. The 1H NMR
spectrum of 3 showed one anomeric proton at δ 4.94 (d, J ) 9.6
Hz), and negative-ion FABMS showed a molecular ion at m/z 383
(M - H)- corresponding to a mono-glucuronide. A comparison of
the 13C NMR spectrum of 3 with that of 4 showed that C-7 of 3
was shifted 3.0 ppm upfield, accompanied by downfield shifts of
C-6 (3.6 ppm) and C-8 (2.9 ppm). These shifts indicated a
glucuronide group at C-7. HMBC revealed a correlation between
the anomeric proton and C-7. Thus, 3 was determined to be fraxetin-
7-O-â-glucuronide.
Metabolite 4 was identified as fraxetin, by direct comparison of
1
4 with an authentic sample by EIMS and H NMR data.
1
Metabolite 5 was obtained as a white powder. The H and 13C
NMR data of 5 were similar to those of 4, except for the dis-
appearance of an O-methyl resonance. These data indicate that 5
was a demethylated derivative of 4. Thus, 5 was 6,7,8-trihydroxy-
coumarin. Confirmation of the structure of 5 was made by direct
comparison using EIMS and 1H NMR data with an authentic
sample, which was prepared from 4 by BBr3 treatment.
Metabolite 6 was identified as fraxidin by direct comparison of
1
EIMS and H NMR spectral data with authentic samples.
To determine the contribution of the intestinal microflora to the
formation of 4-6, we examined the changes in urinary metabolite
excretion when rats were treated with antibiotics. Urinary excretion
of 4 and 5 was significantly decreased at 24 h after administration
of 1 in rats treated with antibiotics (4, p < 0.05; 5, p < 0.001),
while the excretion of 6 was not significantly different when
compared with nontreated rats (Figure 1). These results indicate
the importance of gut flora in generating urinary metabolites 4 and
* Corresponding author. Tel: +81-22-234-4181. Fax: +81-22-275-2013.
E-mail: ohsawa@tohoku-pharm.ac.jp.
10.1021/np0580412 CCC: $33.50
© 2006 American Chemical Society and American Society of Pharmacognosy
Published on Web 05/11/2006

756 Journal of Natural Products, 2006, Vol. 69, No. 5
Yasuda et al.
Figure 3. Time course of metabolite formation by rat S-9 mixture.
In general, orally administered phenolic compounds, particularly
those with low polarity, undergo hydroxylation and/or glucuronide
and sulfate conjugation primarily by intestinal microflora and
secondarily in the liver and other tissues. It is well known that
phenolic methoxyl groups of aromatic compounds, such as fla-
vonoids and coumarins, are partly demethylated in the liver.7-11
However, it is interesting to note that both methylated and
demethylated metabolites of 4 were produced after hydrolysis of 1
in the rat body.
Figure 1. Effects of antibiotics on urinary excretion from rats orally
administered 1 (100 mg/kg). The antibiotics phthalylsulfathiazole,
kanamycin sulfate, tetracycline hydrochloride, and bacitracin were
orally administered for 4 days before administration of 1. Statistical
analysis was carried out using the student’s t-test (n ) 6). **p <
0.001, *p < 0.05.
Experimental Section
General Experimental Procedures. Fraxin, arylsulfatase (Type
H-1), â-glucuronidase (Type H-2), glucose-6-phosphate, and proadifen
were purchased from Sigma (St. Louis, MO). Kanamycin sulfate,
tetracycline hydrochloride, and bacitracin were purchased from Wako
Pure Chemical Industries, Ltd. (Osaka, Japan). Phthalylsulfathiazole
was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan).
Glucose-6-phosphate dehydrogenase was obtained from Oriental Yeast
Co., Ltd. (Tokyo, Japan). For column chromatography, Sephadex
LH-20 (Pharmacia Biotech, Uppsala, Sweden) was used. All other
commercially available reagents were of the highest purity. Uncorrected
melting points were determined on a Yanagimoto micro melting point
apparatus. Infrared (IR) spectra were measured with a Perkin-Elmer
FT-IR1725X spectrometer. NMR spectra were recorded on a JEOL
JNM-EX 400 (1H, 400; 13C, 100 MHz) and JNM-LA 600 (1H, 600;
13C, 150 MHz) spectrometer. Chemical shifts are given in δ values
(ppm) downfield relative to tetramethylsilane. Electron impact (EI) and
fast atom bombardment (FAB)-MS were performed on a JEOL JMS-
DX 303 mass spectrometer. The HPLC system comprised a CCMP-II
pump, a CO-8020 column oven (Tosoh, Tokyo, Japan), and a model
MCPD-3600 photodiode array detector (Otsuka, Osaka, Japan). HPLC
conditions for analysis of metabolites were as follows: column,
CAPCELLPAK C18 (5 µm, 3.0 mm i.d. × 250 mm, Shiseido, Tokyo,
Japan); column temperature, 40 °C; flow rate, 0.5 mL/min; detection,
200-400 nm; mobile phase, linear gradient of 0.1% trifluoroacetic acid
in H2O (A) and CH3CN (B), A/B ) 95/5 (0 min) f 90/10 (80 min) f
60/40 (90 min).
Animal Experiments. Male Sprague-Dawley rats (6 weeks, 160-
180 g) were purchased from Japan SLC, Inc. Animals were housed at
22 ( 2 °C and 55 ( 10% humidity in a light-controlled room (light
from 9:00 to 21:00) with free access to water and commercial rodent
chow (CE-2, Clea Japan Inc., Tokyo, Japan). After 3 days of feeding,
food was withheld for 18 h and fraxin (100 mg/kg body weight)
uniformly dispersed in water was administered orally by direct stomach
intubation. Animals had free access to water and sugar during the
experiments.
Preparation of Urine Samples. Urine samples were collected over
a 24 h period using a metabolic cage. Urine was filtered through a
0.45 µm membrane filter, and 50 µL was subjected to HPLC.
Enzymatic Hydrolysis of Urine Samples. Urine (20 mL) was
transferred to a test tube to which 5.0 mL citrate buffer (pH 5.2) and
150 µL of â-glucuronidase solution were added followed by incubation
at 37 °C for 24 h. The incubated solution was extracted three times
with EtOAc (40 mL). The organic layer was dried over anhydrous
Na2SO4 and was evaporated to dryness at 40 °C. The residue was
dissolved in MeOH, and a 10 µL aliquot was subjected to HPLC.
Figure 2. Time course of metabolite formation by bacterial mixture
from rat feces.
5. On the other hand, the main site of formation of 6 was thought
to be the liver. These findings prompted us to investigate the
biotransformation of 1 by intestinal microflora and that of 4 by rat
liver S-9 fraction. During anaerobic incubation of 1 with rat fecal
microflora, 1 was almost completely consumed after 24 h, and
metabolites 4 and 5 were produced (Figure 2). Metabolite 4
increased progressively and peaked at 3 h, then decreased gradually
with prolonged incubation. Metabolite 5 was also produced, peaking
at 24 h and increasing with prolonged incubation. During incubation
of 4 with rat liver S-9 fraction in the presence of an NADPH-
generating system and S-adenosyl-L-methionine, 4 decomposed in
a time-dependent manner and after 8 h treatment was partly
converted to 6, which at its peak accounted for 29.1% of the initial
amount of 4 (Figure 3). Incubation of 4 without S-adenosyl-L-
methionine did not produce 6, suggesting that methylation of 4
occurred by catechol-O-methyltransferase (COMT) (data not shown).
In the present study of in vivo metabolism of 1, five urinary
metabolites including sulfate and glucuronide conjugates of 4 were
detected in rat urine by HPLC. Furthermore, it was shown that 4
and 5 were produced by intestinal microflora, while 6 was produced
by methylation in the S-9 fraction. These data were obtained through
the use of antibiotic-treated rats and incubation of 1 with rat fecal
microflora and rat liver S-9 fraction, respectively. The following
metabolic pathways were deduced from the results of the present
study: 1 was hydrolyzed to produce 4, followed by demethylation
to produce 5 in the intestinal tract; 4 was partly methylated to 6 in
the liver.

Metabolic Fate of Fraxin in Rats
Journal of Natural Products, 2006, Vol. 69, No. 5 757
Isolation of Metabolites. For isolation of 2 and 3, urine (120 mL)
obtained from rats after oral administration of 1 (540 mg) was
lyophilized, dissolved in purified H2O, and centrifuged at 3000 rpm
for 20 min. The supernatant was subjected to Sephadex LH-20 column
chromatography (H2O f MeOH). The H2O-eluted fraction was again
subjected to column chromatography and was lyophilized to obtain the
metabolites 2 (4 mg) and 3 (2 mg). For isolation of 4-6, urine (130
mL) obtained from rats after oral administration of 1 (610 mg) was
incubated with â-glucuronidase/arylsulfatase and was extracted three
times with EtOAc. The organic layer was dried over anhydrous
Na2SO4 for 24 h and evaporated to dryness at 40 °C. The residue was
dissolved in a small amount of MeOH and subjected to chromatography
on Sephadex LH-20 with MeOH. Fractions containing metabolites 4-6
were subjected to preparative HPLC. The HPLC conditions were as
follows: column, Wakosil-II 5C18 (5 µm, 7.5 mm i.d. × 300 mm,
Wako Pure Chemicals Industries Ltd., Osaka, Japan,); mobile phase,
H2O (A) and MeOH (B), linear gradient, A/B ) 95/5 (0 min) f
80/20 (90 min) f 90/10 (100 min); flow rate, 1.0 mL/min; detection
wavelength, 220 nm. Each metabolite fraction was evaporated to
dryness at 40 °C in vacuo to afford 4 (58 mg), 5 (3 mg), and 6 (1 mg).
Compound 2: white powder (4 mg); IR (KBr) νmax 3200, 1714,
1617, 1505, 1461, 1049 cm-1; 1H NMR (DMSO-d6, 600 MHz) δ 7.95
(1H, d, J ) 9.0 Hz, H-4), 6.86 (1H, s, H-5), 6.41 (1H, d, J ) 9.6 Hz,
H-3), 3.78 (3H, s, OCH3); 13C NMR (DMSO-d6, 150 MHz) δ 159.5
(C, C-2), 149.3 (C, C-6), 143.9 (CH, C-4), 138.4 (C, C-8), 137.9 (C,
C-8a), 132.6 (C, C-7), 114.9 (CH or C, C-3 or 4a), 114.8 (CH or C,
C-3 or 4a), 100.1 (CH, C-5), 55.5 (CH3, OCH3); negative FABMS m/z
287 (M - H)-, 207.
Antibiotic Treatment of Animals. Gut sterilization was performed
according to the method of Goodwin et al.11 with minor modifications.
Rats were given a mixture of kanamycin sulfate (45 mg), tetracycline
hydrochloride (20 mg), bacitracin (1 mg), and phthalylsulfathiazole (0.5
mg) orally once daily for 4 days. One hour after the last dose on the
fourth day, 1 (100 mg/kg) was administered. Urine samples were
collected for 24 h, incubated with â-glucuronidase for 24 h, and then
extracted three times with EtOAc. The organic layer was dried over
anhydrous Na2SO4 and evaporated to dryness at 40 °C. The residue
was dissolved in a small amount of MeOH and subjected to HPLC.
Statistical analysis was carried out using the Student’s t-test (n ) 6).
P values of less than 0.05 were considered to indicate statistical
significance.
Incubation of 1 with Rat Fecal Suspension. Fresh feces (8.0 g)
obtained from male SD rats was homogenized in bicarbonate buffer
(pH 7.0, 25 mL) by bubbling with CO2 gas to eliminate air, and
sediment was removed by filtration through gauze. The filtrate was
used as a fecal suspension.
A tube containing 1 (1.0 mg) in DMSO (30 µL) and fecal suspension
(2 mL) was incubated at 37 °C in an anaerobic jar in which air was
replaced with oxygen-free CO2. The resulting mixture was adjusted to
pH ca. 3 with 0.05% trichloroacetic acid and was extracted three times
with EtOAc (10 mL). The EtOAc layer was concentrated to dryness in
vacuo, and the residue was dissolved in MeOH (1 mL). A 20 µL aliquot
of this solution was analyzed by HPLC.
Incubation of 1 with Rat Liver 9000g Supernatant. The S-9
fraction of male SD rats (control animal liver S9, In Vitro Technologies
Inc., MD) was used to investigate the metabolism of 1 according to
the method described by Cooper and Brodie,13 with the following
modifications. The incubation mixture contained S-9 fraction (8 mL),
1 mM S-(5′-adenosyl)-L-methionine chloride (3 mg), 20 mM MgCl2‚
6H2O, and 4 (5 mg) dissolved in 0.1 M PBS (pH 7.4, 7.5 mL).
Incubation was carried out at 37 °C for 0, 1, 2, 4, and 8 h. The mixture
was extracted twice with EtOAc, and the organic layer was evaporated
to dryness at 40 °C. The residue was dissolved in MeOH and filtered
through a 0.45 µm membrane filter, and 20 µL of the sample was
subjected to HPLC.
Quantitative Analysis of Metabolites. A calibration graph was
prepared from peak areas obtained by subjecting 10 µL of the sample
solution to HPLC over a concentration range 10-800 µg/mL for 1,
50-5000 µg/mL for 4, 30-500 µg/mL for 5, and 30-250 µg/mL for
6. The resulting calibration graphs were linear, and each quantitative
value represented the mean of three experiments. Recoveries of
standards added to each blank sample were 90.2-99.4%, and the
relative standard deviations were 2.12-3.34%.
Compound 3: white powder (2 mg); IR (KBr) νmax 3416, 1714,
1
1617, 1507, 1466 cm-1; H NMR (DMSO-d6, 400 MHz) δ 7.93 (1H,
d, J ) 9.6 Hz, H-4), 6.84 (1H, s, H-5), 6.39 (1H, d, J ) 9.6 Hz,
H-3), 4.94 (1H, d, J ) 7.6 Hz, H-1′), 3.78 (3H, s, OCH3); 13C NMR
(DMSO-d6, 100 MHz) δ 170.7 (C, C-6′), 159.5 (C, C-2), 148.8 (C,
C-6), 144.0 (CH, C-4), 137.7 (C, C-8a), 136.2 (C, C-7), 135.6 (C, C-8),
114.5 (CH and C, C-3 and 4a), 103.6 (CH, C-1′), 99.9 (CH, C-5), 75.7
(CH, C-3′), 75.1 (CH, C-5′), 72.8 (CH, C-2′), 71.1 (CH, C-4′), 55.8
(CH3, OCH3); negative FABMS m/z 383 (M - H)-, 207, 192.
Compound 4: pale yellow powder (58 mg); IR (KBr) νmax 3411,
1688, 1612, 1508, 1456 cm-1; 1H NMR (DMSO-d6, 600 MHz) δ 7.89
(1H, d, J ) 9.0 Hz, H-4), 6.80 (1H, s, H-5), 6.22 (1H, d, J ) 9.6 Hz,
H-3), 3.82 (3H, s, OCH3); 13C NMR (DMSO-d6, 150 MHz) δ 160.4
(C, C-2), 145.2 (C, C-6), 144.9 (CH, C-4), 139.2 (C, C-7, 8a), 132.7
(C, C-8), 111.7 (CH, C-3), 110.1 (C, C-4a), 100.2 (CH, C-5), 55.9
(CH3, OCH3); EIMS m/z 208 [M]+, 193, 180, 165, 137, 109; HREIMS
m/z 208.0355 (calcd for C10H8O5, 208.0372).
Compound 5: yellow powder (3 mg); IR (KBr) νmax 3233, 1684,
1620, 1531, 1478 cm-1; 1H NMR (acetone-d6, 600 MHz) δ 7.75 (1H,
d, J ) 9.6 Hz, H-4), 6.63 (1H, s, H-5), 6.13 (1H, d, J ) 9.6 Hz, H-3);
13C NMR (acetone-d6, 150 MHz) δ 161.0 (C, C-2), 145.2 (CH, C-4),
143.7 (C, C-6), 139.1 (C, C-7 or 8a), 139.0 (C, C-7 or 8a), 133.6 (C,
C-8), 113.2 (CH, C-3), 111.9 (C, C-4a), 104.3 (CH, C-5); EIMS m/z
194 [M]+, 166, 138, 110; HREIMS m/z 194.0216 (calcd for C9H6O5,
194.0215).
Supporting Information Available: Three-dimensional HPLC
chromatograms of rat urine excreted for 24 h after oral administration
of 1 (100 mg/kg). This information is available free of charge via the
Internet at http://pubs.acs.org.
References and Notes
Compound 6: colorless needles (CH3OH) (1 mg); mp 170-180 °C
(1) Kariyone, T.; Kimura, Y. Saishin Wakanyakuyoshokubutsu; Hirokawa
Shoten: Tokyo, 1976; p 105.
(ref 172-174 °C); IR (KBr) νmax 3288, 1690, 1613, 1501, 1464 cm-1
;
1H NMR (DMSO-d6, 600 MHz) δ 7.93 (1H, d, J ) 9.6 Hz, H-4), 6.83
(1H, s, H-5), 6.37 (1H, d, J ) 9.6 Hz, H-3), 3.82 (3H, s, 6-OCH3),
3.77 (3H, s, 7-OCH3); 13C NMR (DMSO-d6, 150 MHz) δ 160.0 (C,
C-2), 149.6 (C, C-6), 144.6 (CH, C-4), 140.0 (C, C-7), 138.4 (C, C-8
or 8a), 138.3 (C, C-8 or 8a), 114.4 (C, C-4a), 114.2 (CH, C-3), 100.0
(CH, C-5), 60.4 (CH3, 7-OCH3), 55.8 (CH3, 6-OCH3); EIMS m/z 222
[M]+, 207, 179; HREIMS m/z 222.0538 (calcd for C11H10O5, 222.0528).
Demethylation of 4. BBr3 (3.85 mL) was added dropwise to a
solution of 4 (200 mg) in CH2Cl2 (20 mL) at 0 °C and was then stirred
at room temperature under N2 for 24 h. The reaction mixture was
evaporated in vacuo, and the residue was dissolved in H2O (100
mL). The solution was extracted with EtOAc, dried over anhydrous
Na2SO4, and evaporated in vacuo. The residue was dissolved in MeOH
and subjected to chromatography on Sephadex LH-20 using MeOH as
the eluant. The fraction containing 5 was again subjected to column
chromatography to obtain 5 (160 mg, yield 86%).
(2) Shimada, G. Yakugaku Zasshi 1940, 60, 508-510.
(3) Morita, N.; Hori M. Yakugaku Zasshi 1953, 73, 771-772.
(4) Yamagami, I.; Suzuki, Y.; Itoh, K. Folia Pharmacol. Jpn. 1968, 64,
714-729.
(5) Iida, G. Tohoku J. Exp. Med. 1935, 25, 454-470.
(6) Okui, T. Tohoku J. Exp. Med. 1938, 32, 225-232.
(7) Canivenc-Lavier, M. C.; Brunold, C.; Siess, M. H.; Suschetet M.
Xenobiotica 1993, 23, 259-266.
(8) Mennes, W. C.; Van Holsteijn, C. W. M.; Timmerman, A.;
Noordhoek, J.; Blaauboer, B. J. Biochem. Pharmacol. 1991, 41,
1203-1208.
(9) Mennes, W. C.; Luijcky, N. B. L.; Wortelboer, H. M.; Noordhoek,
J.; Blaauboer, B. J. Arch. Toxicol. 1993, 67, 92-97.
(10) Horsmans, Y.; Desager, J. P.; Harvengt, C. Life Sci. 1993, 53, 421-
426.
(11) Yue, Q. Y.; Sa¨we, J. Eur. J. Clin. Pharmacol. 1997, 52, 41-47.
NP0580412