Identification of Urinary Metabolites of Flavanone in the Rat Helge Busett and Ronald R. Schelinej Department of Pharmacology, School of Medicine, University of Bergen, 5016 Haukeland sykehus, Bergen, Norway
The urinary metabolites of flavanone in the rat were investigatedusing a gas chromatographicmass spectrometric method employing an OV-1 capillary column. Forty-three metabolites were detected and the most prominent of these were identified. The most common metabolic reactions were reductionof the keto group and hydroxylation at the 3- or 6-positions. Little hydroxylation occurred in ring B. The major metabolites were flavan-4a-01, trans -3-hydroxyflavan-40 -ol,6-hydroxyflavanone and 6-hydroxyflavan-40 -01. Gas chromatographic analysis of underivatized flavanone metabolites, especially with stainless steel columns, results in the formation of dehydrated products including flavone, flav-3-ene and flavanone itself.
INTRODUCTION Flavonoids furnish a large group of closely related plant compounds which have been extensively studied from a metabolic point of view. Much of this interest derives from the fact that fairly complex pathways of degradation are often involved. Additionally, many of the reactions encountered are carried out by the gastrointestinal microflora and this leads to an interplay between the gut and tissue reactions. A comprehensive summary of the mammalian metabolism of flavonoids has appeared recently.'
Flavone
Flavanone
Because nearly all plant flavonoids are hydroxylated derivatives, most of the metabolic investigations have dealt with these compounds. Very little information is available on the unsubstituted parent compounds. However, two reports on the metabolism of flavone2 and flavanone3 indicated that the small difference between their chemical structures (see above) was nonetheless sufficient to give rise to large differences in their metabolic fates. Briefly, flavone was metabolized in guinea pigs to derivatives hydroxylated in the B-ring, whereas flavanone underwent changes in rats involving the heterocyclic moiety adjacent to the A-ring. Significantly, the extensive degradative reactions which occur with the polyhydroxylated flavonoids were essentially zbsent with the parent compounds. Our interest in reinvestigating this subject was stimulated partly by the fact that only a relatively small portion of the dose had been accounted for in the previous studies. However, a much more interesting question concerned the apparent 7 Present address: Environmental Technology Section, Applied Chemistry Division, The Foundation of Scientific and Industrial Research at the University of Trondheim, 7034 Trondheim-NTH, Norway. $ Author to whom correspondence should be addressed.
large difference between the pathways of metabolism seen with flavone and flavanone. We were especially intrigued by the reported nature of the metabolic products of flavanone. Das et a ~employing , ~ a GCMS method, identified unchanged flavanone, flavone and flav-3-ene (differs from flavanone by the presence of a double bond at C-3-C-4 and the absence of oxygen at C-4) in the urine. All three of these compounds are of low polarity and it seemed somewhat surprising that they should be the major urinary metabolites of flavanone. Additionally, we found it unexpected that no acid hydrolysable conjugates of flavanone metabolites were detected in the u r i n e ~ .Based ~ on well known metabolic criteria, the flavanone structure contains several sites of potential metaboiic change. Thus, hydroxylation reactions might be expected in rings A and B as well as (Y to the carbonyl group (at C-3). Additionally, carbinols may be formed by reduction of the carbonyl group. In this communication we describe our findings on the nature of urinary flavanone metabolites in the rat.
EXPERIMENTAL Compounds When not specified otherwise, the compounds used in this study were obtained from commercial sources. These were checked for purity and purified if necessary. Samples of flavon-3-01 and 7-hydroxyflavanone were obtained from Dr L. Jurd (Western Regional Research Laboratory, US Department of Agriculture, Albany, California).
Synthesis of reference compounds Chemical structures and mass spectral data for the following reference compounds (as TMS derivatives) are shown in Table 1. Flavan-4a -01 was prepared by the method of Freudenberg and Orthner4 using aluminium amalgam.' To the latter, obtained from 6 g fine aluminium powder, was
CCC-0306-042X/79/0006-0212$04.50 212 BIOMEDICAL MASS SPECTROMETRY, VOt. 6, NO. 5, 1979
@ Heyden & Son Ltd, 1979
URINARY METABOLITES OF FLAVANONE
added flavanone (2 g) dissolved in 80% ethanol (100 ml). The mixture was stirred for 5 h at 10-15 "C, filtered and diluted with an equal portion of water. This solution became turbid and produced a white precipitate upon the addition of a small quantity of 4 N HCl. After filtration, water (100 ml) was added to the filtrate which was set aside for 5 days at 4°C. The crystals which formed were recrystallized from 30% ethanol giving leaflets (45 mg, m.p. 115 "C (lit. 119 0C,4 127 "C6). GCMS analysis of this product on an OV-1 capillary column revealed that it consisted of a mixture of the 4 a and 4p-alcohols in approximately a 7 : 3 ratio. This ratio was not altered by recrystallization or TLC. Flavan-40 -01 was prepared from flavanone by reduction with H2 and a Pd (109'0) on charcoal catalyst using standard methods. This method gives the 4p-is0mer.~ trans -3-Hydroxyflavanone (hydroxyl group trans with respect to the phenyl ring) was prepared by the general method of Bokadia el aL8 2 N NaOH (0.35 ml) and 30% H 2 0 2(0.5 ml) were added to a cold solution of 2'-hydroxychalcone (400 mg) in methanol (5 ml) and allowed to stand overnight at 4 "C. The solution, which changed from reddish-yellow to pale yellow in colour, was acidified with dilute HCl and the precipitate filtered and recrystallized from ethanol to give needles (200 mg, m.p. 177-178 "C (lit. 178-180 "C)). trans-3-Hydroxyflavan-4a -01 and trans-3-hydroxyflavan-4p -01 were prepared from trans-3-hydroxyflavanone by H2reduction using a Pd (loo/,)on charcoal catalyst. This reaction was carried out on only 50 mg of the ketone and therefore the reaction mixture was separated on silica plates using benzene +ethyl acetate ( 5 : 1). Small portions of the developed plates were sprayed with KMn04 (0.5g) in concentrated H2S04 (15 ml) revealing bands at R f0.52 (unchanged ketone) and Rt 0.20. The material from the latter area was shown by OV-1 capillary GCMS to consist of two isomers of trans -3-hydroxyflavan-4-01 which differed by 1.1min in their retention times (as TMS derivatives, see Table 1). The isomer with the longest retention time comprised about 95% of the material and, in analogy with the results from the similar reduction of flavanone to flavan-4P-01 and the retention time differences between the latter compound and its 4a-isomer, it is evident that the major component is the 4p-hydroxy derivative of trans -3-hydroxyflavan and the minor component the corresponding 4 a -derivative. 6-Hydroxyflavanone was prepared by the method of Seikel et ~ 1 from . ~ 2,5-dihydroxyacetophenone(1.5 g) and benzaldehyde (1.05 8). These compounds were dissolved in ethanol (25 ml), the solution cooled to 2 "C and 40% aqueous KOH (15 ml) added under an atmosphere of N2 and with cooling. The deep-red solution was kept for 4 days at 4 "C under Nz. The precipitate obtained upon acidification of this solution was washed with small portions of saturated NaHC03 solution to remove benzoic acid and then taken up in ether. Removal of this solvent gave a yellow-brown oil which deposited yellowwhite crystals after standing for a few days at 4 "C. These were recrystallized from methanol and the pale yellow needles (125 mg) had m.p. 212-213 "C (lit. 213214 "C). 6-Hydroxyflavan-4a -01 and 6-hydroxyflavan-4p -01 were prepared from 6-hydroxyflavanone (60 mg) in the same way as were the 4 a - and 48-isomers of trans-3@ Heyden & Son Ltd, 1979
hydroxyflavan-4-01 described above. TLC under identical conditions allowed separation of the unchanged 0.44) from the flavanol derivative (Rf 0.14) ketone (Rf and GCMS of the latter again showed that the 4 a - and 4p- isomers were present. In this case the mixture consisted of about 80% 0-isomer. 4'-Hydroxyflavanone was prepared using the method of Geissman and Clinton." To a solution of 4-hydroxybenzaldehyde (1.8 g) and 2-hydroxyacetophenone (2.0 g) in ethanol (10 ml), 50% aqueous KOH (8 ml) was added dropwise and the solution, which became deepred, was kept at 50 "C for 18 h. The reaction mixture was poured over crushed ice and then acidified with 4 N HCl. The brown oil which separated was taken up in ether and then shaken with saturated sodium bisulfite solution (20 ml). After removal of the ether the oil was treated with cold ethyl acetate which gave a yellow precipitate. The filtrate was concentrated and placed at 4 "C overnight, giving more yellow material. Recrystallization of the yellow material from benzene + petroleum ether (60-80 "C) (1: 1) gave 2',4-dihydroxychalcone (1.1g, m.p. 143 "C (lit. 145 "C)). Ring closure of this chalcone (500 mg) was achieved by dissolving it in warm ethanol (10 ml) and adding sufficient 3% HCI with warming until the solution became turbid. The solution was then clarified with an addition of ethanol and refluxed for 24 h. The reaction mixture was cooled, filtered and the ethanol removed on a rotary evaporator. The residue was recrystallized from ethyl acetate +petroleum ether (60-80 "C) (1: 1);however, the yield was poor and only a few mg of product (m.p. 185°C (lit. 187°C)) were obtained. 4'-Hydroxyflavan-4a -01 and 4'-hydroxyflavan-4/3 -01 were prepared from 4'-hydroxyflavanone (10 mg) using HZand Pd/charcoal as described above. Removal of the ethanol from the filtered reaction mixture gave a white product which GCMS analysis showed to contain a mixture of the above 4 a - and 4p-isomers in a ratio of roughly 4 : 6.
Metabolism experiments Male albino rats (230-270 g, Wistar strain derived) were switched from a commercial pellet diet to a purified diet" consisting of sucrose, casein, soya oil, salts and vitamins two or three days prior to and following dosing in order to reduce the number and amounts of normal urinary metabolites. An aqueous suspension of finely ground flavanone was prepared using a small amount of sodium deoxycholate and ultrasonic treatment. Following collection of a normal urine sample during a 24 h period, the dose (400 mg kg-') was given by stomach tube and urine samples were collected (at (0 "C) for two 24 h periods. These samples were adjusted to pH 5 , buffered with acetate buffer (pH 5.0) and incubated for 22-24 h at 37 "C with 1 vol.% of a mixture of pglucuronidase + sulfatase (Glusulase, Endo Laboratories, Garden City, New York). The incubates were acidified to p H 1-2, extracted with ether (5 x 25 ml) and the ether phase shaken with 5% NaHC03 solution (3 x 25 ml) to remove acidic components. The ether phase was dried over Na2S04, evaporated carefully to dryness and the samples then placed in a vacuum desiccator over silicagel for 1 h. These samples were then treated with a trimethylsilylating reagent.I2 BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 5, 1979 213
H. BUSET AND R. R. SCHELINE
Gas chromatography mass spectrometry GCMS was carried out using a Hewlett Packard model 5992A system in standard form except for a Hewlett Packard model 18740A capillary injector system (used in the splitless mode) and a special inlet into the mass spectrometer made from a platinum capillary and glasslined stainless steel (Scientific Glass Engineering, North Melbourne, Australia) tubes. A 20 m long x 0.3 mm OV-1 glass capillary column (H. & G. Jaeggi, Trogen, Switzerland) was used with a He flow rate of approx. 3 ml min-l. The oven temperature was programmed from 150 "C to 270 "C at 4 "C min-' and the injector temperature was 245 "C. The mass spectra were scanned from 100 to 600 amu. Experiments on the gas chromatographic stability of flavanone metabolites were carried out on a Varian MAT 111 GCMS system as described before.13 These experiments used glass (145 x 0.2 cm) or stainless steel (200 cm x 118 in) columns packed with 3% OV-17 on Chromosorb Q (80/100 mesh).
1 I
'i
RESULTS AND DISCUSSION As indicated above in the Experimental section, the flavanone metabolites studied in the present investigation are those found in the neutral fraction from urine. This fraction may contain alcohols, phenols and possibly less polar derivatives which are excreted as such or as glucuronide or sulfate conjugates. Enzymic hydrolysis of the samples prior to extraction prevents an assessment of the relative amounts and types of conjugates present or of the location of the conjugating moieties in those metabolites containing multi le sites of conjugation. I? However, other experiments employing [2-14C]flavanone indicated that these neutral compounds are excreted both as such and as their glucuronide and/or sulfate conjugates. Roughly half of the radioactivity excreted in the urine 24 h following administration of the radioactive compound (100 mg kg-', p.0.) was due to neutral metabolites and the free/conjugated ratio was approximately 1.7 : 1. Additionally, excretion of radioactivity in the urine (28% of the dose) was essentially complete within 48 h. Figure 1 shows the gas chromatograms of the total neutral fraction following derivatization with a TMS reagent of a control urine and urine samples collected 0-24 and 24-48 h after flavanone administration. Under the conditions used 43 metabolites were detected. These are numbered in order of their increasing retention times. Results obtained with a lower dose (100 mg kg-') were similar as regards the more abundant metabolites; however, some trace metabolites were poorly observed or absent. Additionally, metabolite excretion was more rapid and, in contrast to that seen in Fig. 1, most occurred during the 0-24 h period. The retention times and prominent mass spectral features of 26 of the flavanone metabolites are summarized in Tables 1 and 2. These tables separate the metabolites into two groups based on the degree of certainty of their structural identities. Thus, Table 1 includes the metabolites which are shown to be identical with authentic reference compounds, data for the latter 214 BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 5, 1979
Figure 1. Gas chromatograms of urine samples (neutral fraction, TMS derivatives) before and after administration of flavanone (400 m g kg-', p.0.) t o rats. (a) Control: (b) 0-24 h: (c) 24-48 h.
also being given. Most of the major metabolites fall into this group. The second group (Table 2) includes metabolites whose structures have been deduced from GCMS data and, in some cases, from metabolic considerations. Although reference compounds are lacking for the metabolites in Table 2, the identities of these are considered to be reasonably certain as the interpretations are based on the data obtained from the closely related derivatives listed in Table 1. While the GCMS data for the remaining metabolites do allow for some structural features to be assigned, their identities nevertheless remain speculative. Therefore we have chosen not to include these minor metabolites in the following discussion. Mass spectra of flavanone derivatives and corresponding metabolites (Table 1) Flavan-4a-01 (metabolite 1). The mass spectrum shows a base peak at m / z 179 which corresponds to a fragment having structure a (Fig. 2). The prominent fragment pair at m / z 207 and m j z 208 is due to structures b and c, respectively. Formation of c corresponds to MHOTMS which is analogous to the well known loss of water from flavan-4-01s.'~ Ejection of H from c to give the fully conjugated species b is also characteristic of this class of compounds. Flavan-48-01 (metabolite 4). This compound, which has a retention time 1 min longer than the 4 a -isomer, gives the same fragmentation pattern as that of metabolite 1, but with lower abundances of the ion pair at m / z 207 and m / z 208. trans-3-Hydroxyflavanone (metabolite 6). The mass spectrum shows a prominent molecular ion at m / z 312 @ Heyden & Son Ltd, 1979
Table 1. Retention times and characteristic mass spectral features of TMS derivatives of flavanone metabolites and corresponding reference compounds Name lunderivatizedl or metabolite numbera
Compound structure (derivatized)
Flavan-4a-ol 1
OTMS
q
FIava n-4p-0I 4
OTMS
trans-3-Hydroxyflavanone 'OTMS
6
0
frans-3-Hydroxyflavan-4a-ol
Retention time (mid
207 179 161 (55) (100) (10) 207 179 161 (33) (100) (10)
10.7
298 (7)
208 (23)
10.8
298 (9)
208 (16)
207 179 161 (131) (19) (100) (9) (9) 207 179 161 131 (15) (100) (8) (7)
11.4
312 (60)
11.4
312 (74)
297 192 179 (29) (100) (44) 297 192 179 (30) (100) 144)
12.0
386 296 (4) (2) 193 192 (19) (100) 386 296 (5) (3) 193 192 (15) (100)
9.8
'OTMS
trans-3-Hydroxyflavan-4p-ol
I
12.0
13.0
'OTMS
OTMS
10
Flavon-3-01 OTMS
13
0
6-Hydroxyflavanone TMSO
14
0
6-H ydroxyf lavan-4a-01 TMSO
\
17
13.1
14.3 14.2
19 4'-Hydroxyflavanone 18
0
OTMS
20
4'-Hydroxyflavan-4p-oI W
a
OO T M S T
M
S
25
161 (15)
177 (24)
163 (19)
161 (18)
267 (46) 179 (63)
219 (10) 177 (9)
207 195 (14) (0.3) 161 147 (18) (21)
194 (7)
267 (59) 179 (53)
219 (10) 177 (11)
207 195 (15) (0.8) 161 147 (13) (14)
194 (7)
386 296 267 (1) (56) (2) 193 192 179 (17) (100) (55) 386 296 267 (2) (73) (4) 193 192 179 (18) (100) (56)
219 (3) 177 (9) 219 (1) 177 (10)
207 (5) 161 (12)
195 (1) 147 (10) 195 (1) 147 (8)
194 (6)
265 (12)
165 (4)
140 (6)
265 (15)
165 (8)
140 (9)
295 (100) 310 295 (3) (100)
310 (3)
207 (6) 161 (12)
311 (8)
297 (6)
235 208 (11) (100)
15.1
312 (49)
311 (9)
297 (6)
235 208 (10) (100)
15.5
386 (13)
295 (12)
282 267 (10) (100)
15.6
386 (19)
296 (23) 296 (18)
295 (14)
282 267 (10) (100)
16.6
386 (13)
296 (9)
295 (6)
282 267 (13) (100)
16.6
386 (13)
296 (9)
295 (6)
282 267 (10) (100)
15.8
312 (85)
311 (63)
15.8
312 (85)
311 (61)
297 192 179 114) (100) (92) 297 192 179 (25) (100) (91)
woTMs 4-Hydroxyflavan-4a-01
163 (14)
312 (44)
OTMS
(7)
177 (71)
151 (30)
177 (92)
151 (29) 177 (13)
267 (24) 267 (10) 267 (9)
192 179 (48) (100) 192 179 (45) (100)
177 (17) 177 (18)
295 (35)
267 (15)
295 (32)
17.8
386 (
The urinary metabolites of flavanone in the rat were investigatedusing a gas chromatographicmass spectrometric method employing an OV-1 capillary column. Forty-three metabolites were detected and the most prominent of these were identified. The most common metabolic reactions were reductionof the keto group and hydroxylation at the 3- or 6-positions. Little hydroxylation occurred in ring B. The major metabolites were flavan-4a-01, trans -3-hydroxyflavan-40 -ol,6-hydroxyflavanone and 6-hydroxyflavan-40 -01. Gas chromatographic analysis of underivatized flavanone metabolites, especially with stainless steel columns, results in the formation of dehydrated products including flavone, flav-3-ene and flavanone itself.
INTRODUCTION Flavonoids furnish a large group of closely related plant compounds which have been extensively studied from a metabolic point of view. Much of this interest derives from the fact that fairly complex pathways of degradation are often involved. Additionally, many of the reactions encountered are carried out by the gastrointestinal microflora and this leads to an interplay between the gut and tissue reactions. A comprehensive summary of the mammalian metabolism of flavonoids has appeared recently.'
Flavone
Flavanone
Because nearly all plant flavonoids are hydroxylated derivatives, most of the metabolic investigations have dealt with these compounds. Very little information is available on the unsubstituted parent compounds. However, two reports on the metabolism of flavone2 and flavanone3 indicated that the small difference between their chemical structures (see above) was nonetheless sufficient to give rise to large differences in their metabolic fates. Briefly, flavone was metabolized in guinea pigs to derivatives hydroxylated in the B-ring, whereas flavanone underwent changes in rats involving the heterocyclic moiety adjacent to the A-ring. Significantly, the extensive degradative reactions which occur with the polyhydroxylated flavonoids were essentially zbsent with the parent compounds. Our interest in reinvestigating this subject was stimulated partly by the fact that only a relatively small portion of the dose had been accounted for in the previous studies. However, a much more interesting question concerned the apparent 7 Present address: Environmental Technology Section, Applied Chemistry Division, The Foundation of Scientific and Industrial Research at the University of Trondheim, 7034 Trondheim-NTH, Norway. $ Author to whom correspondence should be addressed.
large difference between the pathways of metabolism seen with flavone and flavanone. We were especially intrigued by the reported nature of the metabolic products of flavanone. Das et a ~employing , ~ a GCMS method, identified unchanged flavanone, flavone and flav-3-ene (differs from flavanone by the presence of a double bond at C-3-C-4 and the absence of oxygen at C-4) in the urine. All three of these compounds are of low polarity and it seemed somewhat surprising that they should be the major urinary metabolites of flavanone. Additionally, we found it unexpected that no acid hydrolysable conjugates of flavanone metabolites were detected in the u r i n e ~ .Based ~ on well known metabolic criteria, the flavanone structure contains several sites of potential metaboiic change. Thus, hydroxylation reactions might be expected in rings A and B as well as (Y to the carbonyl group (at C-3). Additionally, carbinols may be formed by reduction of the carbonyl group. In this communication we describe our findings on the nature of urinary flavanone metabolites in the rat.
EXPERIMENTAL Compounds When not specified otherwise, the compounds used in this study were obtained from commercial sources. These were checked for purity and purified if necessary. Samples of flavon-3-01 and 7-hydroxyflavanone were obtained from Dr L. Jurd (Western Regional Research Laboratory, US Department of Agriculture, Albany, California).
Synthesis of reference compounds Chemical structures and mass spectral data for the following reference compounds (as TMS derivatives) are shown in Table 1. Flavan-4a -01 was prepared by the method of Freudenberg and Orthner4 using aluminium amalgam.' To the latter, obtained from 6 g fine aluminium powder, was
CCC-0306-042X/79/0006-0212$04.50 212 BIOMEDICAL MASS SPECTROMETRY, VOt. 6, NO. 5, 1979
@ Heyden & Son Ltd, 1979
URINARY METABOLITES OF FLAVANONE
added flavanone (2 g) dissolved in 80% ethanol (100 ml). The mixture was stirred for 5 h at 10-15 "C, filtered and diluted with an equal portion of water. This solution became turbid and produced a white precipitate upon the addition of a small quantity of 4 N HCl. After filtration, water (100 ml) was added to the filtrate which was set aside for 5 days at 4°C. The crystals which formed were recrystallized from 30% ethanol giving leaflets (45 mg, m.p. 115 "C (lit. 119 0C,4 127 "C6). GCMS analysis of this product on an OV-1 capillary column revealed that it consisted of a mixture of the 4 a and 4p-alcohols in approximately a 7 : 3 ratio. This ratio was not altered by recrystallization or TLC. Flavan-40 -01 was prepared from flavanone by reduction with H2 and a Pd (109'0) on charcoal catalyst using standard methods. This method gives the 4p-is0mer.~ trans -3-Hydroxyflavanone (hydroxyl group trans with respect to the phenyl ring) was prepared by the general method of Bokadia el aL8 2 N NaOH (0.35 ml) and 30% H 2 0 2(0.5 ml) were added to a cold solution of 2'-hydroxychalcone (400 mg) in methanol (5 ml) and allowed to stand overnight at 4 "C. The solution, which changed from reddish-yellow to pale yellow in colour, was acidified with dilute HCl and the precipitate filtered and recrystallized from ethanol to give needles (200 mg, m.p. 177-178 "C (lit. 178-180 "C)). trans-3-Hydroxyflavan-4a -01 and trans-3-hydroxyflavan-4p -01 were prepared from trans-3-hydroxyflavanone by H2reduction using a Pd (loo/,)on charcoal catalyst. This reaction was carried out on only 50 mg of the ketone and therefore the reaction mixture was separated on silica plates using benzene +ethyl acetate ( 5 : 1). Small portions of the developed plates were sprayed with KMn04 (0.5g) in concentrated H2S04 (15 ml) revealing bands at R f0.52 (unchanged ketone) and Rt 0.20. The material from the latter area was shown by OV-1 capillary GCMS to consist of two isomers of trans -3-hydroxyflavan-4-01 which differed by 1.1min in their retention times (as TMS derivatives, see Table 1). The isomer with the longest retention time comprised about 95% of the material and, in analogy with the results from the similar reduction of flavanone to flavan-4P-01 and the retention time differences between the latter compound and its 4a-isomer, it is evident that the major component is the 4p-hydroxy derivative of trans -3-hydroxyflavan and the minor component the corresponding 4 a -derivative. 6-Hydroxyflavanone was prepared by the method of Seikel et ~ 1 from . ~ 2,5-dihydroxyacetophenone(1.5 g) and benzaldehyde (1.05 8). These compounds were dissolved in ethanol (25 ml), the solution cooled to 2 "C and 40% aqueous KOH (15 ml) added under an atmosphere of N2 and with cooling. The deep-red solution was kept for 4 days at 4 "C under Nz. The precipitate obtained upon acidification of this solution was washed with small portions of saturated NaHC03 solution to remove benzoic acid and then taken up in ether. Removal of this solvent gave a yellow-brown oil which deposited yellowwhite crystals after standing for a few days at 4 "C. These were recrystallized from methanol and the pale yellow needles (125 mg) had m.p. 212-213 "C (lit. 213214 "C). 6-Hydroxyflavan-4a -01 and 6-hydroxyflavan-4p -01 were prepared from 6-hydroxyflavanone (60 mg) in the same way as were the 4 a - and 48-isomers of trans-3@ Heyden & Son Ltd, 1979
hydroxyflavan-4-01 described above. TLC under identical conditions allowed separation of the unchanged 0.44) from the flavanol derivative (Rf 0.14) ketone (Rf and GCMS of the latter again showed that the 4 a - and 4p- isomers were present. In this case the mixture consisted of about 80% 0-isomer. 4'-Hydroxyflavanone was prepared using the method of Geissman and Clinton." To a solution of 4-hydroxybenzaldehyde (1.8 g) and 2-hydroxyacetophenone (2.0 g) in ethanol (10 ml), 50% aqueous KOH (8 ml) was added dropwise and the solution, which became deepred, was kept at 50 "C for 18 h. The reaction mixture was poured over crushed ice and then acidified with 4 N HCl. The brown oil which separated was taken up in ether and then shaken with saturated sodium bisulfite solution (20 ml). After removal of the ether the oil was treated with cold ethyl acetate which gave a yellow precipitate. The filtrate was concentrated and placed at 4 "C overnight, giving more yellow material. Recrystallization of the yellow material from benzene + petroleum ether (60-80 "C) (1: 1) gave 2',4-dihydroxychalcone (1.1g, m.p. 143 "C (lit. 145 "C)). Ring closure of this chalcone (500 mg) was achieved by dissolving it in warm ethanol (10 ml) and adding sufficient 3% HCI with warming until the solution became turbid. The solution was then clarified with an addition of ethanol and refluxed for 24 h. The reaction mixture was cooled, filtered and the ethanol removed on a rotary evaporator. The residue was recrystallized from ethyl acetate +petroleum ether (60-80 "C) (1: 1);however, the yield was poor and only a few mg of product (m.p. 185°C (lit. 187°C)) were obtained. 4'-Hydroxyflavan-4a -01 and 4'-hydroxyflavan-4/3 -01 were prepared from 4'-hydroxyflavanone (10 mg) using HZand Pd/charcoal as described above. Removal of the ethanol from the filtered reaction mixture gave a white product which GCMS analysis showed to contain a mixture of the above 4 a - and 4p-isomers in a ratio of roughly 4 : 6.
Metabolism experiments Male albino rats (230-270 g, Wistar strain derived) were switched from a commercial pellet diet to a purified diet" consisting of sucrose, casein, soya oil, salts and vitamins two or three days prior to and following dosing in order to reduce the number and amounts of normal urinary metabolites. An aqueous suspension of finely ground flavanone was prepared using a small amount of sodium deoxycholate and ultrasonic treatment. Following collection of a normal urine sample during a 24 h period, the dose (400 mg kg-') was given by stomach tube and urine samples were collected (at (0 "C) for two 24 h periods. These samples were adjusted to pH 5 , buffered with acetate buffer (pH 5.0) and incubated for 22-24 h at 37 "C with 1 vol.% of a mixture of pglucuronidase + sulfatase (Glusulase, Endo Laboratories, Garden City, New York). The incubates were acidified to p H 1-2, extracted with ether (5 x 25 ml) and the ether phase shaken with 5% NaHC03 solution (3 x 25 ml) to remove acidic components. The ether phase was dried over Na2S04, evaporated carefully to dryness and the samples then placed in a vacuum desiccator over silicagel for 1 h. These samples were then treated with a trimethylsilylating reagent.I2 BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 5, 1979 213
H. BUSET AND R. R. SCHELINE
Gas chromatography mass spectrometry GCMS was carried out using a Hewlett Packard model 5992A system in standard form except for a Hewlett Packard model 18740A capillary injector system (used in the splitless mode) and a special inlet into the mass spectrometer made from a platinum capillary and glasslined stainless steel (Scientific Glass Engineering, North Melbourne, Australia) tubes. A 20 m long x 0.3 mm OV-1 glass capillary column (H. & G. Jaeggi, Trogen, Switzerland) was used with a He flow rate of approx. 3 ml min-l. The oven temperature was programmed from 150 "C to 270 "C at 4 "C min-' and the injector temperature was 245 "C. The mass spectra were scanned from 100 to 600 amu. Experiments on the gas chromatographic stability of flavanone metabolites were carried out on a Varian MAT 111 GCMS system as described before.13 These experiments used glass (145 x 0.2 cm) or stainless steel (200 cm x 118 in) columns packed with 3% OV-17 on Chromosorb Q (80/100 mesh).
1 I
'i
RESULTS AND DISCUSSION As indicated above in the Experimental section, the flavanone metabolites studied in the present investigation are those found in the neutral fraction from urine. This fraction may contain alcohols, phenols and possibly less polar derivatives which are excreted as such or as glucuronide or sulfate conjugates. Enzymic hydrolysis of the samples prior to extraction prevents an assessment of the relative amounts and types of conjugates present or of the location of the conjugating moieties in those metabolites containing multi le sites of conjugation. I? However, other experiments employing [2-14C]flavanone indicated that these neutral compounds are excreted both as such and as their glucuronide and/or sulfate conjugates. Roughly half of the radioactivity excreted in the urine 24 h following administration of the radioactive compound (100 mg kg-', p.0.) was due to neutral metabolites and the free/conjugated ratio was approximately 1.7 : 1. Additionally, excretion of radioactivity in the urine (28% of the dose) was essentially complete within 48 h. Figure 1 shows the gas chromatograms of the total neutral fraction following derivatization with a TMS reagent of a control urine and urine samples collected 0-24 and 24-48 h after flavanone administration. Under the conditions used 43 metabolites were detected. These are numbered in order of their increasing retention times. Results obtained with a lower dose (100 mg kg-') were similar as regards the more abundant metabolites; however, some trace metabolites were poorly observed or absent. Additionally, metabolite excretion was more rapid and, in contrast to that seen in Fig. 1, most occurred during the 0-24 h period. The retention times and prominent mass spectral features of 26 of the flavanone metabolites are summarized in Tables 1 and 2. These tables separate the metabolites into two groups based on the degree of certainty of their structural identities. Thus, Table 1 includes the metabolites which are shown to be identical with authentic reference compounds, data for the latter 214 BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 5, 1979
Figure 1. Gas chromatograms of urine samples (neutral fraction, TMS derivatives) before and after administration of flavanone (400 m g kg-', p.0.) t o rats. (a) Control: (b) 0-24 h: (c) 24-48 h.
also being given. Most of the major metabolites fall into this group. The second group (Table 2) includes metabolites whose structures have been deduced from GCMS data and, in some cases, from metabolic considerations. Although reference compounds are lacking for the metabolites in Table 2, the identities of these are considered to be reasonably certain as the interpretations are based on the data obtained from the closely related derivatives listed in Table 1. While the GCMS data for the remaining metabolites do allow for some structural features to be assigned, their identities nevertheless remain speculative. Therefore we have chosen not to include these minor metabolites in the following discussion. Mass spectra of flavanone derivatives and corresponding metabolites (Table 1) Flavan-4a-01 (metabolite 1). The mass spectrum shows a base peak at m / z 179 which corresponds to a fragment having structure a (Fig. 2). The prominent fragment pair at m / z 207 and m j z 208 is due to structures b and c, respectively. Formation of c corresponds to MHOTMS which is analogous to the well known loss of water from flavan-4-01s.'~ Ejection of H from c to give the fully conjugated species b is also characteristic of this class of compounds. Flavan-48-01 (metabolite 4). This compound, which has a retention time 1 min longer than the 4 a -isomer, gives the same fragmentation pattern as that of metabolite 1, but with lower abundances of the ion pair at m / z 207 and m / z 208. trans-3-Hydroxyflavanone (metabolite 6). The mass spectrum shows a prominent molecular ion at m / z 312 @ Heyden & Son Ltd, 1979
Table 1. Retention times and characteristic mass spectral features of TMS derivatives of flavanone metabolites and corresponding reference compounds Name lunderivatizedl or metabolite numbera
Compound structure (derivatized)
Flavan-4a-ol 1
OTMS
q
FIava n-4p-0I 4
OTMS
trans-3-Hydroxyflavanone 'OTMS
6
0
frans-3-Hydroxyflavan-4a-ol
Retention time (mid
207 179 161 (55) (100) (10) 207 179 161 (33) (100) (10)
10.7
298 (7)
208 (23)
10.8
298 (9)
208 (16)
207 179 161 (131) (19) (100) (9) (9) 207 179 161 131 (15) (100) (8) (7)
11.4
312 (60)
11.4
312 (74)
297 192 179 (29) (100) (44) 297 192 179 (30) (100) 144)
12.0
386 296 (4) (2) 193 192 (19) (100) 386 296 (5) (3) 193 192 (15) (100)
9.8
'OTMS
trans-3-Hydroxyflavan-4p-ol
I
12.0
13.0
'OTMS
OTMS
10
Flavon-3-01 OTMS
13
0
6-Hydroxyflavanone TMSO
14
0
6-H ydroxyf lavan-4a-01 TMSO
\
17
13.1
14.3 14.2
19 4'-Hydroxyflavanone 18
0
OTMS
20
4'-Hydroxyflavan-4p-oI W
a
OO T M S T
M
S
25
161 (15)
177 (24)
163 (19)
161 (18)
267 (46) 179 (63)
219 (10) 177 (9)
207 195 (14) (0.3) 161 147 (18) (21)
194 (7)
267 (59) 179 (53)
219 (10) 177 (11)
207 195 (15) (0.8) 161 147 (13) (14)
194 (7)
386 296 267 (1) (56) (2) 193 192 179 (17) (100) (55) 386 296 267 (2) (73) (4) 193 192 179 (18) (100) (56)
219 (3) 177 (9) 219 (1) 177 (10)
207 (5) 161 (12)
195 (1) 147 (10) 195 (1) 147 (8)
194 (6)
265 (12)
165 (4)
140 (6)
265 (15)
165 (8)
140 (9)
295 (100) 310 295 (3) (100)
310 (3)
207 (6) 161 (12)
311 (8)
297 (6)
235 208 (11) (100)
15.1
312 (49)
311 (9)
297 (6)
235 208 (10) (100)
15.5
386 (13)
295 (12)
282 267 (10) (100)
15.6
386 (19)
296 (23) 296 (18)
295 (14)
282 267 (10) (100)
16.6
386 (13)
296 (9)
295 (6)
282 267 (13) (100)
16.6
386 (13)
296 (9)
295 (6)
282 267 (10) (100)
15.8
312 (85)
311 (63)
15.8
312 (85)
311 (61)
297 192 179 114) (100) (92) 297 192 179 (25) (100) (91)
woTMs 4-Hydroxyflavan-4a-01
163 (14)
312 (44)
OTMS
(7)
177 (71)
151 (30)
177 (92)
151 (29) 177 (13)
267 (24) 267 (10) 267 (9)
192 179 (48) (100) 192 179 (45) (100)
177 (17) 177 (18)
295 (35)
267 (15)
295 (32)
17.8
386 (
