737-741 (1994); Received July 18/September 7, 1994

Home , Isoliquiritigenin

Formation of 6'-Deoxychalcone 4'-Glucosides by Enzyme Extracts from Petals of Dahlia variabilis Karl Stich3, Heidrun Halbwirtha, Friedrich Wurst3 and Gert Forkmannb a TU Wien, Institut für Angewandte Botanik, Technische Mikroskopie und Organische Rohstofflehre, Getreidemarkt 9, A-1060 Wien, Österreich b TU München-Weihenstephan, Lehrstuhl für Zierpflanzenbau, Blumenstraße 16, D-85350 Freising, Bundesrepublik Deutschland Z. Naturforsch. 49c, 737-741 (1994); received July 18/September 7, 1994

Dahlia variabilis, UDP-Glucose: 6 '-Deoxychalcone 4'-0-Glucosyltransferase, Isoliquiritigenin, Butein, Chalcone Yellow colouration of Dahlia variabilis is mainly provided by isoliquiritigenin 4'-glucoside and butein 4'-glucoside. Incubation of petal extracts with uridin 5'-diphosphoglucose and isoliquiritigenin or butein led to the formation of one product which was identified as the respective 6 '-deoxychalcone 4'-glucoside. Glucosylation of hydroxyl groups in other positions was not observed. Naringenin chalcone and eriodictyol chalcone were not accepted as sub strates. The 4'-glucosylation of isoliquiritigenin and butein showed a broad pH optimum ranging from pH 7 to 8 and was stimulated by Mg2+, Ca2+ and Mn2+. N-Ethylmaleimide, p-hydroxymercuribenzoate, Cu2+, Zn2+, and Fe2+ clearly reduced the activity of the enzyme. The apparent K m values for UDP-glucose, isoliquiritigenin and butein were 90, 26 and 276 [xm respectively.

Introduction lation of the B-ring and glycosylation of hydroxyl Yellow colouration of flowers is mainly due to groups at the ring A. Thus, the respective chal the presence of carotenoids (Harborne, 1988). cones were found to be accumulated as glycosides Other classes of yellow pigments are betaxanthins, in the flowers. This accumulation is particularly higher hydroxylated flavonols and the flavonoid- observed in genotypes lacking chalcone iso- related deep yellow chalcones and aurones. In the merase activity. petals of many plants, especially in some members While the biosynthesis of flavonoids has already of the Asteraceae, carotenoids co-occur with the been well-established (Heller and Forkmann, 1988, 1994), our knowledge on the formation of latter compounds. But there are some cases, where the differently hydroxylated and glucosylated chalcones and aurones are the only source of yellow colouration. Examples are Anthirrhinum, chalcones is still poor. Up to now, the respective Callistephus, Dahlia, Dianthus and Helichrysum hydroxylases and glucosyltransferases have not been characterized. Moreover, in vitro formation (Harborne, 1967). Two types of chalcones are formed in flowers, of 6'-deoxychalcones with enzyme preparations the 6'-hydroxychalcones (e.g. naringenin chal from flowers has not yet been demonstrated. cone), which are intermediates in the synthesis of In flowers of Dahlia variabilis the biosynthesis the common 5-hydroxyflavonoids including the of anthocyanins and other flavonoids has already anthocyanins, and the 6'-deoxychalcones (e.g. iso been studied in detail (Koch, 1992; Fischer et al., liquiritigenin), which are involved in the 5-deoxy 1988; Stich et al., 1988). As a rule, crude flower extracts exhibited high activity of all flavonoid en series of flavonoids. Consequently, the transforma zymes measured. Thus, D. variabilis seems to be tion of the chalcones to the respective flavanones well-suited for the elucidation of the biosynthesis is catalyzed by two different chalcone isomerases of different chalcone glucosides, too. Moreover, in (CHI) (Heller and Forkmann, 1988). Both types yellow-flowering cultivars, colouration is caused of chalcones can be modified by further hydroxy- by glucosides of the 6'-deoxychalcones isoliquiriti genin and butein (Nordstrom and Swain, 1956; Reprint requests to Dr. K. Stich. Giannasi, 1975; Harborne, 1990). These chalcones Telefax: (0043) 15874835. co-occur with glycosides of the flavones apigenin

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschung This work has been digitalized and published in 2013 by Verlag Zeitschrift in Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung der für Naturforschung in cooperation with the Max Planck Society for the Wissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht: Advancement of Science under a Creative Commons Attribution Creative Commons Namensnennung 4.0 Lizenz. 4.0 International License. 738 K. Stich et al. • Formation of 6'-Deoxychalcone 4'-Glucosides from Petals of Dahlia variabilis and luteolin or even with anthocyanins, which pH = 8.00 for assays with isoliquiritigenin as leads to more or less yellow-orange flowers. Both substrates. the flavones and anthocyanins are synthesized via the 6'-hydroxychalcone (naringenin chalcone). Enzyme preparation The 6'-deoxychalcones isoliquiritigenin and butein All steps were performed at 4 °C. An aliquot of are not used as intermediates. 1.0 g petals was homogenized together with 0.5 g Yellow flowers accumulate appreciable amounts quartz sand, 0.5 g polyvinylpyrrolidone (Serva, of 6'-deoxychalcone glucosides but no naringenin Germany) and 6 ml buffer in a precooled mortar. chalcone glucoside. This is due to the fact that The homogenate was centrifuged for 10 min at naringenin chalcone is completely isomerized to lOOOOxg. To free the crude extract from phenolic naringenin by CHI providing the substrate for compounds and other low molecular weight sub flavone and anthocyanin formation. In contrast, stances the extract was passed through a Sephadex the 6'-deoxychalcones isoliquiritigenin and butein G-50 (fine) column (bed volume = 1 ml). are glucosylated but neither isomerized enzymati cally nor chemically to the respective flavanones. Enzyme assay This fact should allow the use of both 6'-deoxychalcones as stable substrates for the elucidation The reaction mixture contained in a total vol of the glucosylation reaction in flower extracts. ume of 100 (0,1: 50-82 ^1 buffer, 8-40 (il enzyme This paper reports for the first time on the extract (8-40 (.ig protein), 70 nmol of UDP-d - occurrence of a glucosyltransferase converting [U-l4C]glucose (1600 Bq/70 nmol) and 30 nmol of 6'-deoxychalcones with high specificity to their the respective chalcone or other flavonoid sub respective 4'-glucosides. strates (dissolved in 5 (il ethylene glycol mono methyl ether). Standard enzyme assays were Materials and Methods carried out with butein as substrate. The reaction was started by the addition of Plant material UDP-D-[U14C]glucose. After incubation for The investigations were performed on the yel 15 min at 25 °C the reaction was stopped by the low flowering commercial strain “Johann Nestroy” addition of 50 |il methanol. The mixture was chro (Dr. Wirth) of D. variabilis, containing mainly iso matographed on paper (Schleicher & Schiill liquiritigenin 4'-glucoside and butein 4'-glucoside 2043 b, Germany) using 30% acetic acid. The zone as pigments in the flowers (Nordstrom and Swain, containing the labelled product was localized with 1956; Giannasi, 1975). The plant material was cul a TLC linear analyzer (Berthold, Germany), cut tivated in the “Bundesgärten Schönbrunn” during out and transferred to a scintillation cocktail the summer period. (Ready Solve HP, Beckman, U.S.A.). Afterwards radioactivity was measured in a scintillation coun Chemicals ter (LKB, Sweden). Naringenin, eriodictyol, apigenin, luteolin and Determination of the pH optimum their 7-glucosides as well as butein were purchased from Roth (Karlsruhe, Germany). Naringenin Enzyme assays for the determination of the pH chalcone, eriodictyol chalcone, isoliquiritigenin, optimum were carried out as described above, isoliquiritigenin 4'-glucoside and butein 4'-gluco- using 10 mM Tris/HCl buffer (pH 8.0, containing side were from our laboratory collection. UDP-d - 0.4% Na-ascorbate) for homogenization and 0.2 m [U-14C]glucose (12.0 GBq/mmol) was obtained Tris/HCl buffer (containing 0.4% Na-ascorbate) from Amersham International (Great Britain). with pH between 7.25 and 9.00 for the assay mixture. Buffer solutions Kinetics Unless otherwise stated the following buffers were used: 0.1 m Tris/HCl (containing 0.4% Na- Kinetic data were calculated from Lineweaver- ascorbate), pH = 7.5 for assays with butein and Burk plots. Determination of the apparent K. Stich et al. • Formation of 6'-Deoxychalcone 4'-Glucosides from Petals of Dahlia variabilis 739

Michaelis constant (K m) and maximal velocity firmed by co-chromatography with corresponding (Vmax) f ° r the chalcones was carried out using a reference substances (Table I). The 4'-glucosy- fixed UDP-glucose concentration of 700 [xm. lation of isoliquiritigenin and butein showed a Determination of K m and Vmax for UDP-glucose broad optimum between pH 7 and 8. were performed with a fixed concentration of Because of the higher reaction rate (see below), butein at 300 [xm. At these substrate concen the further characterization of the glucosyl- trations no substrate inhibition could be observed. transferase was carried out with butein as sub strate. Addition of bovine serum albumin (BSA) Analytical methods did not cause any effect. At 25 °C the formation The enzymatically formed chalcone glucosides, of butein 4'-glucoside was linear with protein con flavanone glucosides and flavone glucosides were centration up to 20 [ig protein per assay and with identified by TLC on precoated cellulose plates time for at least 60 min. The enzyme exhibited a (Merck, Germany), using the following solvent temperature optimum at 50 °C. At 0 °C and 60 °C systems: (1) 30% acetic acid; (2) chloroform/acetic the reaction rate reached only 10% of the acid/water (10:9:1); (3) «-butanol/acetic acid/ maximum. water (6:1:2); (4) f-butanol/acetic acid/water The effect of some bivalent cations and poten (3:1:1). Radioactivity was localized by scanning tial enzyme inhibitors on product formation is the plates with a TLC linear analyzer (Berthold, shown in Table II. Addition of Mg2+, Ca2+ and Wildbad, Germany). Mn2+ showed weak stimulatory effects, whereas Protein was determined by a modified Lowry heavy metal ions like Cu2+, Fe2+ and Zn2+ clearly procedure (Sandermann and Strominger, 1972), reduced enzyme activity. Moreover, /7-hydroxy- using crystalline BSA as standard. mercuribenzoate and N-ethylmaleimide did strongly affect the reaction. Diethyldithiocarba- Results mate (DDC) and diethylpyrocarbonate (DPC) mainly inhibited at higher concentrations. KCN Incubation of isoliquiritigenin and butein re and EDTA did not influence the reaction rate. spectively with UDP-D-[U-14C]glucose and en Using butein as substrate, a specific activity of zyme preparations from yellow petals of D. varia 216 [ikat/kg protein was measured, whereas with bilis led to the formation of a single product, which isoliquiritigenin as substrate only a specific activity was identified as the respective chalcone 4'-gluco- of 13 ^kat/kg protein was observed. The values for side by co-chromatography with authentic refer ence substances in four different solvent systems (Table I). Glucosylation of hydroxyl groups in Table II. Influence of divalent ions and inhibitors on 6 '-deoxychalcone 4'-0-glucosyltransferase. other positions was not observed. Moreover, naringenin chalcone and eriodictyol chalcone did Addition Concentra CHGT not serve as substrates. But the enzyme prep tion [ m M ] activity [%]

arations glucosylated the flavanones naringenin None _ 1 0 0 and eriodictyol as well as the flavones apigenin Mg2+ 1 1 2 1 and luteolin in the 7-position, which could be con- Ca2+ 1 108 Mn2+ 1 1 1 0 Cu2+ 1 0 Table I. R{ values (xlOO) of the enzymatically formed Zn2+ 1 0 products. Fe2+ 1 1 0 DDC 1 78 Solvent system 5 32 1 Compound 1 2 3 4 DPC 70 5 0 Isoliquiritigenin 4'-glucoside 32 42 69 59 KCN 1 97 Butein 4'-glucoside 24 58 45 36 EDTA 1 103 Naringenin 7-glucoside 75 74 8 6 89 N-Ethylmaleimide 1 0 Eriodictyol 7-glucoside 71 56 71 76 p-Hydroxymercuribenzoate 0 . 1 0 Apigenin 7-glucoside 38 60 67 71 Luteolin 7-glucoside 25 36 47 46 EDTA, ethylene diamine tetraacetate; DDC, diethyl- dithiocarbamate; DPC, diethylpyrocarbonate. 740 K. Stich et al. ■ Formation of 6'-Deoxychalcone 4'-Glucosides from Petals of Dahlia variabilis

K m and V^x calculated from the Lineweaver- (Larson and Lonergan, 1972; Kleinehollenhorst Burk plot were 267 |im and 400 [imol/skg for et al., 1982; Teusch et al., 1986; Cheng et al., 1994). butein and 26 ^im and 13 [imol/skg for isoliquiriti Surprisingly, however, the addition of BSA to the genin, respectively. The ratios VmaJ K m were 1.5 enzyme assay did not lead to a significant increase for butein and 0.5 for isoliquiritigenin, confirming of enzyme activity as reported for other glucosyl- butein as the preferred substrate. Using butein as transferases (Sutter et al., 1972; Saleh et al., 1976 a, substrate, the K m and Vm.dX for UDP-glucose were 1976 b). The inhibitory effect of the heavy metal

90 |a m and 87 [imol/skg, respectively. ions is most probably due to their interaction with A loss of 30% of enzyme activity was observed, SH groups of the enzyme. This assumption is con when the petals were frozen in liquid nitrogen. But firmed by the fact, that the addition of SH group- no further decrease was detected during storage binding reagents as /?-hydroxymercuribenzoate at -80 °C for several weeks. and N-ethylmaleimide did affect the reaction strongly, too. Discussion Maximal enzyme activity was observed at a sur The glycosylation of flavonoids has been well prisingly high temperature of 50 °C. Similar high studied in the case of flavones, flavonols and values were found for the chalcone synthase anthocyanidins (Heller and Forkmann, 1988, (CH S) of D. variabilis (Forkmann, unpublished), 1994). Up to now, however, there is little known which can be assumed to provide isoliquiritigenin about the enzymatic formation of chalcone gluco and butein and thereby the substrates for the sides. The only reports concern an enzyme prep glucosylation reaction. Moreover, a very high tem aration from Citrus paradisii seedlings, which was perature optimum was also found for the antho- found to catalyze 7-O-glucosylation of flavanones cyanidin 3-O-glucosyltransferase of Matthiola in- but also glucosylates the 4'-position of naringenin cana (Teusch et al., 1986). Actually, the high chalcone (McIntosh et al., 1990; McIntosh and optimum values measured in vitro are only of Mansell, 1990). theoretical interest, since they are clearly not Now in petal extracts of D. variabilis, we could within near-physiological ranges. More surprising, demonstrate for the first time an enzyme which however, is the strong reduction of enzyme ac glucosylates 6'-deoxychalcones. Thus, D. variabilis tivity at low temperatures, since yellow coloura flowers again proved to be a rich enzyme source. tion of D. variabilis flowers is not reduced in In agreement with the naturally present chalcone autumn. Thus, biosynthesis of the pigments has to glucosides, the glucosyltransferase transfers the proceed even at low temperatures. Apparently, glucosyl moiety of UDP-glucose specifically to the even a low glucosyltransferase activity is sufficient 4'-position of isoliquiritigenin and butein. The en for the glucosylation reaction since the 6'-deoxy- zyme preparation also catalyzed the glucosylation chalcones are stable substrates and they are not of flavanones and flavones in the 7-position. But converted by other enzymes. On the other hand, the 6'-hydroxy compounds naringenin chalcone in several plant species a number of flavonoid en and eriodictyol chalcone were not accepted as zymes were found to exhibit high activities also at substrates. low temperatures (Ruhnau and Forkmann, 1988; The kinetic data revealed that the glucosyl Teusch, 1986; Stich et al., 1992). Studies in Den- transferase shows high substrate specificity to dranthema grandiflora (Chrysanthemum morifo- wards the 6'-deoxychalcones isoliquiritigenin and lium) which are typical “autumn flowers” also butein indicating that the enzyme is specifically in revealed that the reaction rate of a number of en volved in the biosynthesis of chalcone glucosides. zymes involved in biosynthesis of anthocyanins Up to now, however, it cannot be excluded, that reached at 0 °C at least 50% of the optimum the glucosylation of 6'-deoxychalcones in the (Stich, unpublished). 4'-position and 5-hydroxyflavanones and -flavones The co-occurrence of isoliquiritigenin and in the related 7-position is catalyzed by one and butein glucosides in D. variabilis flowers raises the the same enzyme. question, how the additional hydroxy group in ring The stimulatory effect of divalent ions on en B of butein is introduced. There are two possible zyme activity corresponds well with reported data ways. On the one hand, isoliquiritigenin could be K. Stich et al. • Formation of 6'-Deoxychalcone 4'-Glucosides from Petals of Dahlia variabilis 741 hydroxylated to butein by an up to now unknown work will be concerned with the detailed eluci chalcone 3-hydroxylase. On the other hand butein dation of the biosynthesis of the 6'-deoxychal- could be formed directly by the use of caffeoyl- cones isoliquiritigenin and butein. CoA instead of p-coumaroyl-CoA during syn thesis of the 6'-deoxychalcone skeleton. 6'-Deoxy- Acknowledgements chalcone formation can be assumed to be cata lyzed by the common chalcone synthase in co The studies were supported by the “Hochschul- action with an NADPH-dependent polyketide jubiläumsstiftung der Stadt Wien”. We thank the reductase (Welle and Grisebach, 1988). However, “Bundesgärten Schönbrunn” under direction of this reaction has not yet been demonstrated with Dr. Fischer Colbrie and Dr. Wirth for the culti enzyme preparations from flowers. Thus, further vation of the plant material.

Cheng G. W., Malencik D. A. and Breen P. J. (1994), Nordstrom C. G. and Swain T. (1956), The flavonoid UDP-glucose: flavonoid O-glucosyltransferase from glycosides of Dahlia variabilis. II. Glycosides of yel strawberry fruit. Phytochemistry 35, 1435-1439. low varieties “Pius IX” and “Cotton”. Arch. Biochem. Fischer D., Stich K., Britsch L. and Grisebach H. (1988), Biophys. 60, 329-344. Purification and characterization of (+)-dihydrofla- Ruhnau B. and Forkmann G. (1988), Flavan-3,4-diols in vonol (3-hydroxyflavanone) 4-reductase from flowers anthocyanin biosynthesis, enzymatic formation with of Dahlia variabilis. Arch. Biochem. Biophys. 264, flower extracts from Callistephus chinensis. Phyto 40-47. chemistry 27, 1035-1039. Giannasi D. E. (1975), The flavonoid systematics of the Saleh N. A. M., Fritsch H., Witkop P. and Grisebach H. genus Dahlia (Compositae). Memoirs of the New (1976a), UDP-glucose: cyanidin 3-O-glucosyltrans- York Botanical Garden 26, 1-125. ferase from cell cultures of Haplopappus gracilis. Harborne J. B. (1967), in: Comparative Biochemistry of Planta 133, 41-45. Flavonoids. Academic Press, London. Saleh N. A. M., Poulton J. E. and Grisebach H. (1976b), Harborne J. B. (1988), in: Plant Pigments (T. W. Good UDP-glucose: cyanidin 3-O-glucosyltransferase from win, ed.). Academic Press, London, p. 337. red cabbage seedlings. Phytochemistry 15, 1865-1868. Harborne J. B., Greenham J. and Eagles J. (1990), Sandermann H. and Strominger L. (1972), Purification Malonylated chalcone glycosides in Dahlia. Phyto and properties of C55-isoprenoid alcohol phospho- chemistry 29, 2899-2900. kinase from Staphylococcus aureus. J. Biol. Chem. 247, Heller W. and Forkmann G. (1988), in: The Flavonoids 5123-5131. (J. B. Harborne, ed.). Chapman and Hall, London, Stich K., Ebermann R. and Forkmann G. (1988), Einfluß pp. 399-425. Cytochrom P-450-spezifischer Inhibitoren auf die Heller W. and Forkmann G. (1994), in: The Flavonoids Aktivität von Flavonoid 3'-Hydroxylase und Flavon- (J. B. Harborne, ed.). Chapman and Hall, London, synthase II bei verschiedenen Pflanzen. Phyton (Aus pp. 499-535. tria) 28, 237-247. Kleinehollenhorst G., Behrens H., Pegels G., Srunk N. Stich K., Eidenberger T., Wurst F. and Forkmann G. and Wiermann R. (1982), Formation of flavonol 3-0- (1992), Enzymatic conversion of dihydroflavonols to diglycosides and flavonol 3-O-triglycosides by enzyme flavan-3,4-diols using flower extracts of Dianthus extracts from anthers of Tulipa cv. Apeldoorn. Z. caryophyllus L. (carnation). Planta 187, 103-108. Naturforsch. 37c, 587-599. Sutter A., Ortmann R. and Grisebach H. (1972), Puri Koch I. (1992), Enzyme der Flavonoidbiosynthese fication and properties of an enzyme from cell sus höherer Pflanzen: Enzymologische und immunologi pension cultures of parsley catalyzing the transfer of sche Charakterisierung der Chalkonsynthase, Flava- D-glucose from UDP-D-glucose to flavonoids. Bio non 3-Hydroxylase und Dihydroflavonol 4-Reduk- chim. Biophys. Acta 258, 71-87. tase. Ph.D. Thesis, University of Tübingen, Germany. Teusch M. (1986), Uridine 5'-diphosphate-xylose: antho- Larson R. L. and Lonergan C. M. (1972), Glucosyl- cyanidin 3-O-glucose-xylosyltransferase from petals of transferase activity in a water extract of maize pollen. Matthiola incana R. Br. Planta 169, 559-563. Planta 103, 361-364. Teusch M., Forkmann G. and Seyffert W. (1986), UDP- McIntosh C. A., Latchinian L. and Mansell R. L. (1990), glucose: anthocyanidin/flavonol 3-O-glucosyltrans- Flavanone-specific 7-O-glucosyltransferase activity in ferase in enzyme preparation from flower extracts Citrus paradisi seedlings: purification and characteri of genetically defined lines of Matthiola incana R. Br. zation. Arch. Biochem. Biophys. 282, 50-57. Z. Naturforsch. 41c, 699-706. McIntosh C. A. and Mansell R. L. (1990), Biosynthesis Welle R. and Grisebach H. (1988), Isolation of a novel of naringenin in Citrus paradisi: UDP-glucosyltrans- NADPH-dependent reductase which co-acts with ferase activity in grapefruit seedlings. Phytochemistry chalcone synthase in the biosynthesis of 6 '-deoxychal- 29, 1533-1538. cones. FEBS Lett. 236, 221-225.