- Email: [email protected]
Quercetin 3,7,4′-triglucoside formation from quercetin by Vitis hybrid cell cultures
Phytochemistry, Vol. 30, No. 3, pp. 829-831, 1991 Printed in Great Britain.
0031-9422/91 $3.00+0.00 © 1991 Pergamon Press plc
QUERCETIN 3,7,4'-TRIGLUCOSIDE FORMATION FROM QUERCETIN BY VITIS HYBRID CELL CULTURES TETSURO KOKUBO, MAYUMI NAKAMURA,TAKASHI YAMAKAWA,HIROSHI NOGUCHI* and TOHRU KODAMAt Department of Agricultural Chemistry, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan; *Faculty of Pharmaceutical Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan (Received in revisedform 21 August 1990) Key Word Index--Vitis hybrid; Vitaceae; grape; glucosylation; flavonol; quercetin; triglucoside.
Abstract--Suspension cultures of a Vitis hybrid converted quercetin to quercetin 3,7,4'-tri-O-glucoside which was characterized by spectroscopy and TLC analysis. Accumulation of quercetin 3,7,4'-triglucoside in the cells was stimulated when cultured in a high concentration of 2,4-dichlorophenoxyacetic acid.
INTRODUCTION In recent years glycosylation of phenolic compounds including flavonoids has been investigated in several laboratories [,14] in order to obtain further useful glucosides or reveal biochemical processes. From a biochemical point of view, there is insufficient information in spite of many enzymic studies [,7-10]. Thus, it is important to obtain a more detailed knowledge ofglycosylation at the molecular level for the efficient production of glycosides. Flavonoids are well known to possess many kinds of biological activity [11] as well as being natural plant pigments. In our laboratory we have attempted to develop a method to convert quercetin into a number of useful glucosides using Vitis hybrid cell cultures. As a first step to this goal, we reported on the structures of six glucosides formed from quercetin in a previous paper [,12]. During that investigation we observed a further unidentified product (1), which increased with time after a lag time of three days. The increase accompanied a gradual decrease of diglucoside and it was proposed that the compound is a quercetin triglucoside as judged from the R t on reverse phased HPLC. However 1 was not characterized because it was present in extremely small quantity. Thus, in order to establish the identity of this compound we have now devised a method to produce I in better yield. In the present paper we describe this method and the structural elucidation of 1. RESULTS AND DISCUSSION The structures of the six glucosides found in our previous study [,12] were: quercetin 3,7-diglucoside, 3,4'diglucoside and 3-glucoside, isorhamnetin 3,7-diglucoside, 3,4'-diglucoside, and 3-glucoside. As the amount of the unknown glycoside 1 was too small to elucidate its structure at that time, we have altered the concentration of phytohormones in the medium during the bioconversion in an attempt to increase production. These experiments indicated that the simultaneous addition of tAuthor to whom correspo~adenceshould be addressed.
2,4-D (10 ppm) with quercetin brought about a marked increase in the production of I (Fig. 1). Vitis sp. cell cultures at the late exponential growth phase were fed with quercetin and filter-sterilized 2,4-D (10 ppm), and subsequently incubated for a week before harvest. The methanol extract of the harvested cells was separated by Sephadex LH-20 column chromatography. The fraction containing 1 was collected and further separated by preparative HPLC to give a single peak. After acid hydrolysis 1 yielded quercetin as aglycone and only glucose moieties as linked sugars. The FAB mass spectrum of I gave a molecular ion at m/z 787 [ , M - H I - . The fragment peaks at m/z 625 [,(M - H ) - 1 6 2 ] - , 463 [ , ( M - H ) - 3 2 4 ] - and 301 [ ( M - H ) - 4 8 6 ] - were due to the consecutive loss of three glucosyl units. Absorption spectroscopy of 1 showed that the 3-, 7-, and 4'-hydroxyls are all substituted and the 5-hydroxyl is free [15, 16], thus defining it as quercetin 3,7,4'triglucoside. The 13C NMR spectrum (Table I) also supported the presence of one glucosyl unit at each of the C-3, C-7 and C-4' positions in comparison with the spectra of quercetin [,17], quercetin 3,7-diglucoside and 3,4'-diglucoside [12]. Both the assignments between C-3' and C-4' and between C-2' and C-5' were established on the basis of cross peaks in the C,H-COSY spectrum. The coupling constant (3=7.0-7.5 Hz) between H-1 glucose and H-2 glucose indicated that the linked sugars are all in the fl-configuration. The pyranoside form of the sugar was established on the basis of 13CNMR spectral data [,18]. All of the above results showed that 1 is quercetin 3,7,4'-tri-O-fl-D-glucopyranoside, a new natural product. As triglucoside formation by bioconversion is rarely observed [,19], it is important to understand the molecular mechanisms underlying increases in triglucoside production using 2,4-D and to know whether this is specific to cell cultures of Vitis sp. or universal in other plant species. These studies are in progress. EXPERIMENTAL
Tissue culture. The cell culture was initiated from anthers of a Vitis hybrid (Bailey Alicante A) and maintained for over 10 yr.
829
T. KOKU~O et al.
830
Table 1. 13CNMR data of compound 1 (100.5MHz, DMSO-dr, values in ppm from TMS)
-g
/
20
o
o
r..)
10
0
5
10
15
Incubation time ( day ) Fig. 1. Time course of conversion from quercetin to quercetin 3,7,4'-triglucoside with (@) and without (©) the addition of 2,4 D in the medium. At aero time, 2,4-D (10 #gml- 1) was added at the late exponential growth phase of Vitis sp. cell cultures.
C
1
2 3 4 5 6 7 8 9 t0 1' 2' 3' 4' 5' 6' 1"
155.9 134.1 177.6 160.7 99.3 162.8 94.4 155.9 105.7 124.4 116.6 146.2 147.5 115.6 120.8 99.8 100.7 101.6 73.0 73.1 74.0 75.8 76.3 76.4 69.6 69.8 69.9 77.0 77.1 77.4 60.6 60.7 60.9
2"
3''a The culture medium contained Murashige-Skoog's inorganic elements [13], 30 g sucrose, 100 mg myo-inositol, 1.0 mg thiamine HCI, 0.05 mg 2,4-dichlorophenoxyacetic acid (2,4-D) and 0.2 mg kinetin 1-1. Detailed subculture conditions have been described previously [14]. Biotransformation methods. A suspension culture (ca 2 g fr. wt) at the late exponential growth phase was transfered and further incubated (300 rpm, 2 cm stroke, 30°) in a test tube (10ml medium) containing 2 mg quercetin and 0.1-100 ppm 2,4-D, if necessary, for a week. Cells were periodically harvested on a filter paper and extracted with MeOH for 24hr. Aliquots of the extract were directly subjected to HPLC analysis. In the largescale cultivation using a 3 I jar fermentor (21 medium), cells at the late exponential growth phase (ca 400 g fr. wt) were incubated with quercetin (600 mg) and 10 ppm 2,4-D for a week at 30° and at a constant aeration rate of 1 vvm. Isolation ofglycosylated product. The cells incubated in a 3 1 jar fermentor were harvested and extracted with MeOH. The MeOH extract was concd in vacuo, suspended in H20 and then partitioned between hexane and H20. The aq. fraction was chromatographed on a column of Sephadex LH-20 using MeOH-H20 (2:3) as eluent. The fraction containing 1 was collected and further separated by prep. HPLC using a SensyuPak ODS-5251-S column (20 × 250 mm) and M e O H - H 2 0 HOAc (5:81:14) as eluent. Quantitative analysis by HPLC. Aliquots to be analysed were loaded onto the HPLC system (Shimadzu, LC-6A) fitted with a 4.6 x 250 mm SensyuPak ODS-1251-S column. The column was developed with a 7.5 ml linear gradient of solvent A (MeOH-H20-HOAc, 5:8l:14) to 100% solvent B (MeOHH20-HOAc, 3:6:1), followed by 22.5 ml of solvent B at a flow rate of 1.5 ml min -a. The eluting compounds were monitored with a UV/VIS detector at 370 nm to determine the conversion ratio of each product. Acid hydrolysis of compound 1. Compound 1 was refluxed in 2 M HCI for 1 hr. The aglycone was extracted with isoamyl alcohol and identified by HPLC to be quercetin. The aq, layer
4"
5''~
6"
aInterchangeable.
was analysed on silica gel TLC with n-BuOH-EtOH-H20 (120:33:57) as solvent and the sugar moiety detected with anisaldehyde-H2SO4 and identified using authentic sugars as standards. Quercetin 3,7,4'-tri-O-fl-D-glucopyranoside. Yellow powder. 1HNMR (500MHz) in DMSO-dr: 34.93 (1H, d, J=7.5 Hz, anomeric H), 5.15 (1H, d, J=7.5 Hz, anomeric H), 5.57 (1H, d, J = 7.0 Hz, anomeric H), 6.51 (1H, d, d = 1.5 Hz, 6-H), 6.87 (1H, d, J = 1.5 Hz, 8-H), 7.28 (1H, d, J = 8.0 Hz, 5'-H), 7.69 (1H, dd, J = 1.5 and 8.0 Hz, 6'-H), 7.74 (1H, d, J = l . 5 Hz, 2'-H). The 13CNMR spectrum is described in Table 2. FABMS data are also described in the text. UV 2 Me°" nm: 348, 266sh, 254; (NaOMe) 375, 265; (AICIa) 349, 268; (AIC13+ HC1) 395 sh, 347, 272; (NaOAc) 346, 263, 252 sh; (NaOAc + HaBO3) 348, 266 sh, 254. REFERENCES
1. Mizukami, H., Terao, T., Miura, H. and Ohashi, H. (1983) Phytochemistry 22, 679. 2. Braemer, R., Tsoutsias, Y., Hurabielle, M. and Paris, M. (1987) Planta Med. 53, 225.
Quercetin triglucoside formation by Vitis cultures 3. Furuya, T., Ushiyama, M., Asada, Y. and Yoshikawa, T. (1987) Phytochemistry 26, 2983. 4. Tabata, M., Umetani, Y., Ooya, M. and Tanaka, S. (1988) Phytochemistry 27, 809. 5. Berger, R. G. and Drawert, F. (1988) Z. Naturforsch. 43(2, 485. 6. Lcwinsohn, E., Berman, E., Mazur, Y. and Gressel, J. (1989) Plant Sci. 61, 23. 7. Petersen, M. and Seitz, H. U. (1986) J. Plant Physiol. 125, 383. 8. Kreis, W., May, U. and Reinhard, E. (1986) Plant Cell Rep. 5, 442. 9. Kerscher, F. and Franz, G. (1988) J. Plant Physiol. 132, 110. 10. Hrazdina, G. (1988) Biochim. Biophys. Acta 955, 301. 11. Racker, E. (1985) Progress in Clinical and Biological Research Vol. 213 Plant Flavonoids in Biology and Medicine (Cody, V., Middleton Jr., E. and Harborne, J. B. eds), pp. 257-272. Alan R. Liss, New York.
PE¥ 3 0 : 3 - a
831
12. Kodama, T., Ishida, H., Kokubo, T., Yamakawa, T. and Noguchi, H. (1990) Agric. Biol. Chem. 54, 3283. 13. Murashige, T. and Skoog, F. (1962) Physiol. Plant. 15, 473. 14. Yamakawa, T., Ishida, K., Kato, S., Kodama, T. and Minoda, Y. (1983) Agric. Biol. Chem. 47, 997. 15. Mabry, T. J., Markham, K. R. and Thomas, M. B. (1970) The Systematic Identification of Flavonoids. pp. 41-56, Springer, New York. 16. Markham, K. R. (1982) Techniques of Flavonoid Identification, pp. 36-51. Academic Press, New York. 17. Markham, K. R. and Chari, V. M. (1982) The Flavonoids:Advances in Research, (Harborne, J. B. and Mabry, T. J,, eds), pp. 37--45. Chapman & Hall, London. 18. Markham, K. R., Ternai, B., Stanley, R., Geiger, H. and Mabry, T. J. (1978) Tetrahedron 34, 1389. 19. Furuya, T. (1988) Yakugaku Zasshi 108, 675.
0031-9422/91 $3.00+0.00 © 1991 Pergamon Press plc
QUERCETIN 3,7,4'-TRIGLUCOSIDE FORMATION FROM QUERCETIN BY VITIS HYBRID CELL CULTURES TETSURO KOKUBO, MAYUMI NAKAMURA,TAKASHI YAMAKAWA,HIROSHI NOGUCHI* and TOHRU KODAMAt Department of Agricultural Chemistry, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan; *Faculty of Pharmaceutical Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan (Received in revisedform 21 August 1990) Key Word Index--Vitis hybrid; Vitaceae; grape; glucosylation; flavonol; quercetin; triglucoside.
Abstract--Suspension cultures of a Vitis hybrid converted quercetin to quercetin 3,7,4'-tri-O-glucoside which was characterized by spectroscopy and TLC analysis. Accumulation of quercetin 3,7,4'-triglucoside in the cells was stimulated when cultured in a high concentration of 2,4-dichlorophenoxyacetic acid.
INTRODUCTION In recent years glycosylation of phenolic compounds including flavonoids has been investigated in several laboratories [,14] in order to obtain further useful glucosides or reveal biochemical processes. From a biochemical point of view, there is insufficient information in spite of many enzymic studies [,7-10]. Thus, it is important to obtain a more detailed knowledge ofglycosylation at the molecular level for the efficient production of glycosides. Flavonoids are well known to possess many kinds of biological activity [11] as well as being natural plant pigments. In our laboratory we have attempted to develop a method to convert quercetin into a number of useful glucosides using Vitis hybrid cell cultures. As a first step to this goal, we reported on the structures of six glucosides formed from quercetin in a previous paper [,12]. During that investigation we observed a further unidentified product (1), which increased with time after a lag time of three days. The increase accompanied a gradual decrease of diglucoside and it was proposed that the compound is a quercetin triglucoside as judged from the R t on reverse phased HPLC. However 1 was not characterized because it was present in extremely small quantity. Thus, in order to establish the identity of this compound we have now devised a method to produce I in better yield. In the present paper we describe this method and the structural elucidation of 1. RESULTS AND DISCUSSION The structures of the six glucosides found in our previous study [,12] were: quercetin 3,7-diglucoside, 3,4'diglucoside and 3-glucoside, isorhamnetin 3,7-diglucoside, 3,4'-diglucoside, and 3-glucoside. As the amount of the unknown glycoside 1 was too small to elucidate its structure at that time, we have altered the concentration of phytohormones in the medium during the bioconversion in an attempt to increase production. These experiments indicated that the simultaneous addition of tAuthor to whom correspo~adenceshould be addressed.
2,4-D (10 ppm) with quercetin brought about a marked increase in the production of I (Fig. 1). Vitis sp. cell cultures at the late exponential growth phase were fed with quercetin and filter-sterilized 2,4-D (10 ppm), and subsequently incubated for a week before harvest. The methanol extract of the harvested cells was separated by Sephadex LH-20 column chromatography. The fraction containing 1 was collected and further separated by preparative HPLC to give a single peak. After acid hydrolysis 1 yielded quercetin as aglycone and only glucose moieties as linked sugars. The FAB mass spectrum of I gave a molecular ion at m/z 787 [ , M - H I - . The fragment peaks at m/z 625 [,(M - H ) - 1 6 2 ] - , 463 [ , ( M - H ) - 3 2 4 ] - and 301 [ ( M - H ) - 4 8 6 ] - were due to the consecutive loss of three glucosyl units. Absorption spectroscopy of 1 showed that the 3-, 7-, and 4'-hydroxyls are all substituted and the 5-hydroxyl is free [15, 16], thus defining it as quercetin 3,7,4'triglucoside. The 13C NMR spectrum (Table I) also supported the presence of one glucosyl unit at each of the C-3, C-7 and C-4' positions in comparison with the spectra of quercetin [,17], quercetin 3,7-diglucoside and 3,4'-diglucoside [12]. Both the assignments between C-3' and C-4' and between C-2' and C-5' were established on the basis of cross peaks in the C,H-COSY spectrum. The coupling constant (3=7.0-7.5 Hz) between H-1 glucose and H-2 glucose indicated that the linked sugars are all in the fl-configuration. The pyranoside form of the sugar was established on the basis of 13CNMR spectral data [,18]. All of the above results showed that 1 is quercetin 3,7,4'-tri-O-fl-D-glucopyranoside, a new natural product. As triglucoside formation by bioconversion is rarely observed [,19], it is important to understand the molecular mechanisms underlying increases in triglucoside production using 2,4-D and to know whether this is specific to cell cultures of Vitis sp. or universal in other plant species. These studies are in progress. EXPERIMENTAL
Tissue culture. The cell culture was initiated from anthers of a Vitis hybrid (Bailey Alicante A) and maintained for over 10 yr.
829
T. KOKU~O et al.
830
Table 1. 13CNMR data of compound 1 (100.5MHz, DMSO-dr, values in ppm from TMS)
-g
/
20
o
o
r..)
10
0
5
10
15
Incubation time ( day ) Fig. 1. Time course of conversion from quercetin to quercetin 3,7,4'-triglucoside with (@) and without (©) the addition of 2,4 D in the medium. At aero time, 2,4-D (10 #gml- 1) was added at the late exponential growth phase of Vitis sp. cell cultures.
C
1
2 3 4 5 6 7 8 9 t0 1' 2' 3' 4' 5' 6' 1"
155.9 134.1 177.6 160.7 99.3 162.8 94.4 155.9 105.7 124.4 116.6 146.2 147.5 115.6 120.8 99.8 100.7 101.6 73.0 73.1 74.0 75.8 76.3 76.4 69.6 69.8 69.9 77.0 77.1 77.4 60.6 60.7 60.9
2"
3''a The culture medium contained Murashige-Skoog's inorganic elements [13], 30 g sucrose, 100 mg myo-inositol, 1.0 mg thiamine HCI, 0.05 mg 2,4-dichlorophenoxyacetic acid (2,4-D) and 0.2 mg kinetin 1-1. Detailed subculture conditions have been described previously [14]. Biotransformation methods. A suspension culture (ca 2 g fr. wt) at the late exponential growth phase was transfered and further incubated (300 rpm, 2 cm stroke, 30°) in a test tube (10ml medium) containing 2 mg quercetin and 0.1-100 ppm 2,4-D, if necessary, for a week. Cells were periodically harvested on a filter paper and extracted with MeOH for 24hr. Aliquots of the extract were directly subjected to HPLC analysis. In the largescale cultivation using a 3 I jar fermentor (21 medium), cells at the late exponential growth phase (ca 400 g fr. wt) were incubated with quercetin (600 mg) and 10 ppm 2,4-D for a week at 30° and at a constant aeration rate of 1 vvm. Isolation ofglycosylated product. The cells incubated in a 3 1 jar fermentor were harvested and extracted with MeOH. The MeOH extract was concd in vacuo, suspended in H20 and then partitioned between hexane and H20. The aq. fraction was chromatographed on a column of Sephadex LH-20 using MeOH-H20 (2:3) as eluent. The fraction containing 1 was collected and further separated by prep. HPLC using a SensyuPak ODS-5251-S column (20 × 250 mm) and M e O H - H 2 0 HOAc (5:81:14) as eluent. Quantitative analysis by HPLC. Aliquots to be analysed were loaded onto the HPLC system (Shimadzu, LC-6A) fitted with a 4.6 x 250 mm SensyuPak ODS-1251-S column. The column was developed with a 7.5 ml linear gradient of solvent A (MeOH-H20-HOAc, 5:8l:14) to 100% solvent B (MeOHH20-HOAc, 3:6:1), followed by 22.5 ml of solvent B at a flow rate of 1.5 ml min -a. The eluting compounds were monitored with a UV/VIS detector at 370 nm to determine the conversion ratio of each product. Acid hydrolysis of compound 1. Compound 1 was refluxed in 2 M HCI for 1 hr. The aglycone was extracted with isoamyl alcohol and identified by HPLC to be quercetin. The aq, layer
4"
5''~
6"
aInterchangeable.
was analysed on silica gel TLC with n-BuOH-EtOH-H20 (120:33:57) as solvent and the sugar moiety detected with anisaldehyde-H2SO4 and identified using authentic sugars as standards. Quercetin 3,7,4'-tri-O-fl-D-glucopyranoside. Yellow powder. 1HNMR (500MHz) in DMSO-dr: 34.93 (1H, d, J=7.5 Hz, anomeric H), 5.15 (1H, d, J=7.5 Hz, anomeric H), 5.57 (1H, d, J = 7.0 Hz, anomeric H), 6.51 (1H, d, d = 1.5 Hz, 6-H), 6.87 (1H, d, J = 1.5 Hz, 8-H), 7.28 (1H, d, J = 8.0 Hz, 5'-H), 7.69 (1H, dd, J = 1.5 and 8.0 Hz, 6'-H), 7.74 (1H, d, J = l . 5 Hz, 2'-H). The 13CNMR spectrum is described in Table 2. FABMS data are also described in the text. UV 2 Me°" nm: 348, 266sh, 254; (NaOMe) 375, 265; (AICIa) 349, 268; (AIC13+ HC1) 395 sh, 347, 272; (NaOAc) 346, 263, 252 sh; (NaOAc + HaBO3) 348, 266 sh, 254. REFERENCES
1. Mizukami, H., Terao, T., Miura, H. and Ohashi, H. (1983) Phytochemistry 22, 679. 2. Braemer, R., Tsoutsias, Y., Hurabielle, M. and Paris, M. (1987) Planta Med. 53, 225.
Quercetin triglucoside formation by Vitis cultures 3. Furuya, T., Ushiyama, M., Asada, Y. and Yoshikawa, T. (1987) Phytochemistry 26, 2983. 4. Tabata, M., Umetani, Y., Ooya, M. and Tanaka, S. (1988) Phytochemistry 27, 809. 5. Berger, R. G. and Drawert, F. (1988) Z. Naturforsch. 43(2, 485. 6. Lcwinsohn, E., Berman, E., Mazur, Y. and Gressel, J. (1989) Plant Sci. 61, 23. 7. Petersen, M. and Seitz, H. U. (1986) J. Plant Physiol. 125, 383. 8. Kreis, W., May, U. and Reinhard, E. (1986) Plant Cell Rep. 5, 442. 9. Kerscher, F. and Franz, G. (1988) J. Plant Physiol. 132, 110. 10. Hrazdina, G. (1988) Biochim. Biophys. Acta 955, 301. 11. Racker, E. (1985) Progress in Clinical and Biological Research Vol. 213 Plant Flavonoids in Biology and Medicine (Cody, V., Middleton Jr., E. and Harborne, J. B. eds), pp. 257-272. Alan R. Liss, New York.
PE¥ 3 0 : 3 - a
831
12. Kodama, T., Ishida, H., Kokubo, T., Yamakawa, T. and Noguchi, H. (1990) Agric. Biol. Chem. 54, 3283. 13. Murashige, T. and Skoog, F. (1962) Physiol. Plant. 15, 473. 14. Yamakawa, T., Ishida, K., Kato, S., Kodama, T. and Minoda, Y. (1983) Agric. Biol. Chem. 47, 997. 15. Mabry, T. J., Markham, K. R. and Thomas, M. B. (1970) The Systematic Identification of Flavonoids. pp. 41-56, Springer, New York. 16. Markham, K. R. (1982) Techniques of Flavonoid Identification, pp. 36-51. Academic Press, New York. 17. Markham, K. R. and Chari, V. M. (1982) The Flavonoids:Advances in Research, (Harborne, J. B. and Mabry, T. J,, eds), pp. 37--45. Chapman & Hall, London. 18. Markham, K. R., Ternai, B., Stanley, R., Geiger, H. and Mabry, T. J. (1978) Tetrahedron 34, 1389. 19. Furuya, T. (1988) Yakugaku Zasshi 108, 675.