Planta 136, 53-59 (1977)
Pl~)l~'~) 9 by Springer-Verlag 1977
Identification of an UDP-Glucose: Flavonol 3-O-Glucosyl-Transferase from Cell Suspension Cultures of Soybean (Glycine max L.) J.E. Poulton* and Margit Kauer Lehrstuhl f/ir Biochemie der Pflanzen am Biologischen Instltut II der Universit~it Freiburg, Sch~inzlestr. 1, D-7800 Freiburg, Federal Republic of Germany
Abstract. A glucosyltransferase, which catalyses the glucosylation of flavonols, using uridine diphosphateD-glucose as glucose donor, has been isolated and purified about 5-10 fold from cell suspension cultures of soybean ( G l y c i n e m a x L., var. Mandarin). The p H o p t i m u m for this reaction was ca. 8.5 in glycineN a O H buffer, and no additional cofactors were required. The enzyme glucosylated the following flavonols predominantly at the 3-position: quercetin (Kin 126 gM), kaempferol (Kin 172pM), isorhamnetin (Kin 200 pM) and fisetin (Kin 270 gM). With quercetin as substrate, the apparent K m value for uridine diphosphate-D-glucose was 0.3 M. Glucosylation of flavonols and flavones by this preparation occurred weakly also at the 7-position. N o activity was found with dihydroquercetin, naringenin, 4,2',4'-trihydroxychalcone, daidzein or texasin. The enzyme was specific for flavonoid compounds, since no activity was observed towards cinnamic acids or simple phenols. However, the preparation was contaminated by a vanillic acid glucosyltransferase, from which it could be partially separated by ionexchange chromatography. The specific activity of the flavonol 3-O-glucosyltransferase increased with age of the culture, reaching a m a x i m u m late in the growth cycle of the culture. Key words: Cell culture - Flavonoid O-glucosyltransferase - UDP-glucose.
Glycine
-
Introduction Recent experiments have confirmed the presence of two distinct SAM:3,4-dihydric phenol 3-O-methyl-
transferases* in extracts f r o m cell suspension cultures of soybean ( G l y c i n e m a x L.) (Poulton et al., 1976a, b). These enzymes were distinguishable by differences in their stability and substrate specificity. One enzyme (CMT) catalyses the 3-O-methylation of cinnamic acids and protocatechualdehyde, whereas the second methyltransferase (FMT) exhibited a pronounced preference for flavonoid substrates, such as quercetin and luteolin. Interestingly, it was found that the F M T could also methylate luteolin 7-O-glucoside (Poulton et al., 1977). It is at present unclear whether flavonoid glycosides are metabolically active in plants and it still has to be determined whether methylation of these flavonoids precedes glucosylation in vivo or vice versa. In an effort to clarify this position, a search was undertaken in extracts from soybean cell cultures for specific flavonoid glucosyltransferases, which could glucosylate the substrates of F M T and their respective products. We wish here to report the partial purification and characterization of an UDP-glucose:flavonol 3O-glucosyltransferase and to describe how the specific activity of this enzyme is dependent upon the age of the cell culture.
Materials and Methods Chemicab
UDP-D-[U-14C]glucose (296 mCi/mmol) was purchased from the Radiochemical Centre, Amersham, England, and diluted with nonradioactive UDP-glucose from Boehringer, Mannheim. Dihydroquercetin, aplgenin, luteolin, naringenin and fisetin were obtained from Roth (Karlsruhe) and quercetin from Merck (Darmstadt). Eriodictyol was a gift from Dr. Narstedt, Freiburg. Other flavonoid compounds were from our laboratory collection.
Present address: Department of Biochemistry and Biophysics, University of California, Davis, California 95616, USA
Cell Cultures
SAM, S-adenosyl-L-methionine;CMT, SAM :caffeate 3-O-methyltransferase; FMT, SAM:flavonoid O-methyltransferase; UDP-glucose, uridine diphosphate-D-glucose; PAL, phenylalanine ammonia-lyase
a: Cell cultures used for the purification of the enzyme. Soybean ceils (Glycine max L. var. Mandarin) were propagated in the dark in a fermenter containing 300 1 medium I (Hahlbrock et al., 1974; Hahlbrock, 1975). The cells used for the enzyme purification were
*
Abbreviations."
54
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
those which were harvested at a stage in the growth cycle of the culture (stage A) characterized by high phenylalanine ammonialyase and F M T levels (Poulton et al., 1976a; Hahlbrock etal., 1974).
b: Cell cultures used for the age-dependence experiments. The cell cultures used for the time-course experiment were propagated in the dark and sampled as described previously (Poulton et al., 1976a). At each sampling-time, determinations of cell fresh weight and conductivity in the medium were made (Hahlbrock et al., 1974; Ebel et al., 1974). The cells were immediately frozen by immersion in liquid nitrogen and then stored at - 2 0 ~ until all samples of the series had been cellected.
Buffer Solutions The following buffer solutions were used: (I) 0.1 M potassium phosphate buffer, pH 7.6, containing 1.45 m M fi-mercaptoethanot; (II) 20raM potassium phosphate buffer, pH7.6, containing 1.45 m M fi-mercaptoethanol.
Enzyme Assays a: Flavonol glucosyltransferase activity: The incubation mixture consisted of 50 nmol flavonoid acceptor (dissolved in 10 lal ethylene glycol monomethylether), 100 nmol UDP-D-[U-14C]glucose (containing 111,000 dpm), 0.25 mg bovine serum albumin and 25 lxmol potassium phosphate buffer, pH 7.0, in a total volume of 125 Ixl. Where indicated, 25 gmol potassium phosphate, pH6.0, or 25 gmol glycine-NaOH, pH 8.6, were used as the buffer in the assay mixture. The reaction was started by addition of enzyme (20 50 ~tl, depending on extent of purification) and was terminated after an incubation period of 30 rain or 60 min at 30~ by the addition of 10 lal glacial acetic acid. The reaction mixture was then applied in a 3 cm-band to Schleicher and Schtill (2043 b) chromatography paper and chromatographed with solvent system I. After viewing the paper under ultraviolet light, the product zones were cut out and counted in a toluene scintillation fluid (2,5diphenyloxazole/L toluene) in a liquid scintillation spectrometer (Beckmann LS 233). b: Flavonoid 7-O-Glucosyltransferase activity: This activity was assayed either as described above in paragraph (i) or in a modified assay mixture containing only 6.25 nmol flavonoid acceptor (dissolved in 5 gl ethylene glycol monomethylether), since this activity has been previously shown to be inhibited by excess substrate (Sutter et al., 1972). Incubation was carried out for 60 rain. After addition of 10 gl glacial acetic acid, chromatography was undertaken as described above.
graphy with an authentic sample on Schleicher and Schtill (2043b) paper using the following solvent systems: (I) 1% HC1 (RF=0.1), (II) 15% acetic acid (Rv=0.49), (III) n-butanol-acetic acid-H20 (4:1:5, by vol.) (Rv=0.62), (IV) water (RF=0.09). Furthermore, this product exhibited the same behaviour under ultraviolet light in the presence and absence of ammonia vapour as did the authentic 3-O-glucoside. The glucosylation of other flavonoid substrates by the enzyme preparation was routinely followed using solvent system I. The respective 3-0- and 7-O-glucosides of these compounds were identified by their Rv-valnes and by behaviour under ultraviolet light. The possible glucosylation of phenolic compounds and cinnamic acids was monitored by using solvent system III. UDPGlucose remains close to the origin, whereas the substrates had Rv-values in the range 0.75-0.95. The expected glucosylation products have intermediate Rv-values.
Enzyme Purification The stored frozen cells (280 g stage A cells) were thawed out at 4~ and then homogenized with 30 g quarz sand and 360 ml buffer I in a mortar at this temperature. The resultant liquid (530 ml) was centrifuged at 30,000g for 25 min. Dowex 1 x 2 (0.15 g/ml), which had been equilibrated with buffer I, was added to the resulting supernatant. After stirring the mixture for 15 min at 4 ~ the ion-exchanger was removed by filtration through glass-wool. The supernatant liquid was brought to 40% saturation by addition of solid (NH4)2SO4 over a period of l0 min with continuous addition of K O H solution to maintain the pH at 7.0. The mixture was allowed to stand for a further 25 rain at 4~ and was then centrifuged at 30,000g for 15 rain. The supernatant was then brought to 80% saturation with solid (NH4)2SO 4 and the precipitate collected in the same way and dissolved in a minimum volume of buffer II (15 ml). Residual ammonium sulfate was removed from this fraction by applying it to a Sephadex G-25 column (40 c m x 3 cm), which was equilibrated and eluted with buffer II. Fractions containing protein were pooled and applied to a DEAEcellulose column (6 cm • 2 cm), which had been equilibrated with buffer II. After washing the column with 100 ml buffer II, eIution was continued with a linear 20 m M 250 m M potassium phosphate gradient (pH 7.5, containing 1.45 m M fl-mercaptoethanoi; 140 ml total volume). The most active glucosyltransferase fractions were combined and dialysed twice against buffer II for one hour. The enzyme was then concentrated by ultrafiltration and stored deepfrozen at - 2 0 ~ until required for kinetic studies.
Crude Extracts for Age-dependence Experiments e: Glucosyltransferase activity towards other phenolic substrates: 50 nmol of each phenolic substrate (dissolved in 10 ~tl ethylene glycol monomethylether) were tested at pH 7.0 (in phosphate buffer) and at pH 8.6 (in glycine-NaOH buffer) in the assay system as described in paragraph (i) above. After 45 min, the reaction was terminated by addition of 10 itl glacial acetic acid. Chromatography was carried out on Schleicher and Schiill (2043 b) chromatography paper, using solvent system III.
A crude macerate was obtained by stirring 1.0 g of cells with 2 ml of buffer I for 60 rain at 4~ The homogenate was centrifuged for 15 min at 20,000 g. Dowex 1 x 2 (0.3 g), previously equilibrated with buffer I, was added to the supernatant liquid and the mixture was stirred for 15 min at 4~ before centrifugation as described above. The resultant supernatant liquid was used for the estimation of protein content and of PAL and quercetin glucosyltransferase activities.
d.' Assay of PAL activity: Phenylalanine ammonia-lyase activity was determined by a spectrophotometric assay as described elsewhere (Zimmermann and Hahlbrock, 1975).
Protein Estimation Chromatographic Identifieation of Reaction Products Quercetin 3-O-glucoside was identified as the major product of the reaction between UDP-glucose and quercetin by co-chromato-
The protein content was estimated by the biuret method (Layne, 1957); crystalline bovine serum albumin, dessicated before use, was used as standard.
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
Results
Partial Purification of the Enzyme UDP : Glucose: flavonol 3-O-glucosyltransferase activity could be detected in crude homogenates of stage A soybean cells, using quercetin as substrate in the assay method described in the Methods section. Table 1 summarizes the partial purification of this enzyme, enabling a preparation to be obtained which had a specific activity towards quercetin about 5 times greater than that of the crude homogenate. The activTable 1. Partial purification of the soybean UDP-glucose: flavonol 3-O-glucosyltransferase Stage
Total Vol. pro- (ml) tein (mg)
Activity Specific Yield (pkat/ml) activity (% (~tkat/kg) initial)
774
530
2.26
1.55
100
Dowex Treatment
718
520
2.80
2.03
121
Sephadex G-25 eluate (after 40-80% ammonium sulphate step)
145
33
18.50
4.20
51
7.78
12
84
ity in the crude extract was initially concentrated and purified by a 40-80% ammonium sulphate fractionation step. Further purification was achieved by DEAE-cellulose ion-exchange chromatography (Fig. 1). The resultant enzyme preparation, after dialysis and concentration, was stored deep-frozen at -20~ until used for the kinetic experiments described below. No significant glucosyltransferase activity was observed at pH 7.0 with the crude homogenates towards luteolin. Likewise, glucosylation of two flavones (luteolin and apigenin) a flavanone (naringenin) and of two isoflavones (daidzein and texasin) was not detected at pH 7.0 within the 40-80% ammonium sulphate fraction (after removal of ammonium ions by gel-filtration) using the modified assay for 7-0glucosyltransferase activity.
Properties of the Partially-Purified Enzyme
Crude Homogenate
Concentrated DEAE-cellnlose eluate
55
3.3 43.6
Stability: Stability experiments indicated that the quercetin 3-O-glucosyltransferase activity was most stable when stored at - 2 0 ~ under these conditions, only 5% of the initial activity was lost after storage for 2 days. In contrast, 75% of the initial activity was lost after storage at 4~ for this period. Bovine serum albumin (2 mg/ml) was added to the incubation mixture to protect the enzyme from inactivation during incubation at 30~ pH Optimum: The pH optimum for the 3-O-glucosy-
-E i I-
lation of quercetin, determined in several different buffer systems, was found to be around pH 8.54.0 in glycine-NaOH buffer (Fig. 2). Since quercetin has
25(
i I
E
i I
200g -8O
:~ 100 80
150~i
60
10
8
loo i -90
o
~6
.o
40
2"
o
._>
5o2
2o
"/ /. o*o'~176
--
Q.
0
0 5
10 15 Froction
-100
20
Fig. l. Partial separation by DEAE-cellulose ion-exchange chromatography of flavonol 3-O-glucosyltransferase ( 9 and vanillic acid glucosyltransferase (zx zx) activities using a phosphate buffer gradient ( - - - ) at pH 7.5. The protein content of the collected fractions (7.5 ml) was monitored by absorbance at 280 nm ( ). For comparative purposes, the enzyme activities are expressed as the percentage of the maximum activities observed (kaempferol 3-O-glucosyltransferase activity, assayed at pH 7.0, 68 nmol/h/ml enzyme; vanillic acid glueosyltransferase activity, assayed at pH 8.6, 32 nmol/h/ml enzyme). The same elntion profile was obtained when the vanillic acid glucosyltransferase activity was assayed at pH 7.0
ffl
P
4
A
_
2 0
8
_z (.9
0
-
o/
I
I
I
4
6
8
10
pH
Fig. 2. Effect of pH on the quercetin 3-O-glucosyltransferase activity of the partially purified enzyme. Quercetin and UDPG-[14C] were incubated with the enzyme (0.3 rag) under conditions given in the Methods section using 0.2 M concentrations of the following buffers: glycine-NaOH (zx--z~), K H z P O 4 - K 2 H P O , ( e - - e ) and KzHPO4-citric acid ( o - - o )
56
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
Table 2. Substrate specificity of the flavonol glucosyltransferase preparation from soybean cell cultures Substrate
Amount substrate supplied (nmol)
Test pH
Product obtained
Amount product (cpm/30 rain/50 gl enzyme)
Amount product (gkat/kg)
Kaempferol
50
8.6
37Unidentified
5380 270 487
12.2 0.6 1.1
50
7.0
37Unidentified
3654 111 198
8.3 0.2 0.4
50
6.0
37Unidentified
3098 50 180
7.0 0.1 0.4
Quercetin
50
8.6
3 7-
6705 363
15.2 0.8
Cyanidin
50 50 50 25
8.6 7.0 6.0 6.0
333 3-
102 108 179 225
0.2 0.2 0.4 0.5
Apigenin
150 50 25 6.25
8.6 8.6 8.6 8.6
777 7-
195 285 200 8
0.4 0.6 0.4 0
Luteolin
50 2.5 6.25
8.6 8.6 8.6
77 7-
515 328 137
1.2 0.7 0.3
Vanillic acid
50 50
8.6 7.0
Unidentified Unidentified
2487 1537
5.6 3.5
The substrate specificity of the enzyme preparation was tested by incubating the above substrates with 50 gl of the purified enzyme (0.3 mg protein) at the pH value shown for 45 min in the assay system described in the Methods section. No detectable activity was found with the following substrates at either pH 8.6 or 7.0 : dihydroquercetin, naringenin, 4,2',4'-trihydroxychalcone, daidzein, texasin, catechol, hydroquinone, resorcinol, salicylic acid, phloroglucinol, caffeic acid, o-coumaric acid, p-coumaric acid, ferulic acid, coniferyl alcohol, vanillin, isovanillin and anisic acid.
been f o u n d to be u n s t a b l e at this p H ( P o u l t o n et al., 1977), reactions were n o r m a l l y carried out at p H 7.0 in phosphate buffer.
Table 3. Substrate specificity of the soybean flavonol 3-O-glucosyltransferase Substrate
Km V (gM) (gkat/kg)
103 x V/Km (gkat kg ~M 1)
Quercetin Kaempferol Isorhamnetin Fisetin
126 172 200 270
103 86 51 25
Protein and Time Linearity." T h e rate of quercetin 3-O-glucosylation at p H 7.0, catalysed by 0.3 mg of the partially-purified enzyme, was linear for at least 80 min. The extent of 3-O-glucosylation at this pH was p r o p o r t i o n a l to the p r o t e i n a m o u n t added up to at least 0.3 m g of the partially-purified preparation.
Substrate specificity." The substrate specificity of the enzyme was investigated by i n c u b a t i n g various potential acceptors with the enzyme p r e p a r a t i o n at p H 7.0 (in p h o s p h a t e buffer) a n d at p H 8.6 (in g l y c i n e - N a O H buffer). Tables 2 a n d 3 indicate that, within the group of f l a v o n o i d c o m p o u n d s tested, strong glucosyltransferase activity was detected towards quercetin,
13.02 14.70 10.25 6.66
The above values were all obtained from kinetic experiments using the same enzyme preparation, which had been purified as described in the Methods section kaempferol, i s o r h a m n e t i n a n d fisetin. W e a k e r activity was observed with c y a n i d i n , a p i g e n i n a n d luteolin. I n contrast, n o glucosylation o f dihydroquercetin, n a r i n g e n i n or 4 , 2 ' , 4 ' - t r i h y d r o x y c h a l c o n e or of the isoflavones daidzein or texasin was observed.
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
As shown in Table 2, glucosylation of the supplied flavonols was not entirely restricted to the 3-position, since the production of the corresponding 7-O-glucosides from quercetin and kaempferol (amounting to approximately 2-5% of the 3-O-glucoside) was observed. Moreover, a third, as yet unidentified, glucosylation product was obtained with kaempferol. Kinetic parameters for the flavonol substrates of the enzyme were determined in phosphate buffer, pH 7.0. Under these conditions, all substrates are stable and the minor glucosylation reactions are reduced to a minimum. Quercetin was shown to be the best substrate, having the highest Vmax/Km value (Table 3). Glucosylation of the flavones luteolin and apigenin proceeded at a relatively slow rate at the 7-position (Table 2). Various simple phenolic compounds and cinnamic acids were also tested, using the assay system described, in order to determine whether the enzyme was specific for flavonoid compounds. No significant activity was detected with hydroquinone, catechol, resorcinol, salicylic acid, phloroglucinol, caffeic acid, o-coumaric acid (in presence or absence of 10 mM /~-mercaptoethanol), p-coumaric acid, ferulic acid, coniferyl alcohol, vanillin, isovanillin or anisic acid. In contrast, a radioactive product was detected after incubation with vanillic acid as acceptor substrate; the rate of its synthesis was greater at pH 8.6 than at pH 7.0. Positive identification of this product was not possible, since authentic glucosylation products were unavailable. This product had the indicated e F values when re-chromatographed on Schleicher and Schtill paper using the following solvents: (I) n-butanolacetic acid-water (4-1-5, v/v), RF=0.54, (II) 1% HC1, Rv=0.81, (III) 15% acetic acid, Rv=0.86, (IV) 30% acetic acid, Rv=0.85. An effort was made to determine whether the flavonol 3-O-glucosyltransferase was also responsible for the described 7-O-glucosylation reactions and for the glucosylation of vanillic acid, or whether this enzyme preparation was contaminated with other glucosyltransferases. These three activities were monitored in the fractions eluting from the DEAE-cellulose column. Figure 1 demonstrates that a partial separation of the flavonol 3-O-glucosyltransferase and of the vanillic acid glucosyltransferase activities was achieved, suggesting the existence of two distinct enzymes. A definite decision could not be made whether the 3-0glucosyltransferase could also glucosylate flavones at the 7-position, since, although both activities displayed a very similar profile upon elution from the column, only relatively low 7-O-glucosyltransferase activity (with luteolin as substrate, up to 250 cpm above background) was observed.
57 IO.O-
,o~ '~ - ~
zs- -~
_
,'o i
A
~
~
I
I
I
~
~ 9
8
?
20
>,
;'
--
5.0-
I
o
~
I \
~A ....
Z~-.
.,
"N
o
o
_
,
~
,
,
-1.0
,
:>-~'\'~.-.~ ~'--
_
'"
I 100
-1.5E
W.,i-
d
2.5- ~, 10
-213
-0.5
\
I
I ~ I 200 Time of incubation [h]
I 300
Fig. 3. Time course of the changes in the specific activities of PAL ( 9 1 6 9 and quercetin 3-O-glucosyltransferase ( o o) during the period of rapid increase in the cell fresh weight of a soybean culture. The conductivity of the medium ( A - - A ) was measured in addition to the fresh weight of the cells (z~--zs). The conductivity, expressed in the units mhos, is the reciprocal of the electrical resistance, expressed in ohms. The broken lines correspond to parts of the curves which were extrapolated according to previous results
Km for UDP-Glucose The apparent Km for UDP-glucose was 0.3 mM, when assayed at pH 7.0 in the presence of a saturating concentration of quercetin (0.4 mM).
Dependence of the 3-O-glucosyltransferase Activity upon the Age of the Cell Culture The specific activities of phenylalanine ammonia-lyase and of the quercetin 3-O-glucosyltransferase were analyzed during the growth in the dark of a soybean cell suspension culture. As has been reported in previous studies (Poulton et al., 1976a; Ebel et al., 1974), a sharp peak in the specific activity of phenylalanine ammonia-lyase was observed shortly before the conductivity of the culture medium reached its minimum value (Fig. 3). The specific activity of the quercetin 3-O-glucosyltransferase exhibited a distinct profile, in which, following an initial reduction it increased with age of the culture, reaching a maximum late in the growth-cycle.
Discussion
Many glucosyltransferases have been recognized within the Plant Kingdom which catalyse the transfer of a sugar residue from donor molecules to such acceptor molecules as cinnamic acids (Kleinhofs et al.,
58
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
1967), simple hydroxyphenols (Yamaha and Cardini, t 960), sterols (Fang and Baisted, 1976), coniferyl alcohol (Ibrahim and Grisebach, 1976), flavonoids (Sutter et al., 1972; Sutter and Grisebach, 1973), and anthocyanins (Saleh et al., 1976). Despite the accumulation of such knowledge and the fact that flavonoid glucosides are of widespread occurrence in plants (Harborne, 1975), the metabolic role of these compounds in plants is still unclear. Earlier work had suggested that they functioned as metabolic end-products (e.g. perhaps storage products or as pigments), but evidence is now forthcoming demonstrating that they may be further metabolized (H6sel and Barz, 1975; Ebel et al., 1972). The efficient methylation of luteolin, luteolin 7-O-glucoside and of quercetin by an SAM:flavonoid methyltransferase (FMT) purified from soybean cell suspension cultures has been observed in this laboratory (Poulton et al., 1977). Since the problem whether glucosylation of flavonoids may precede methylation is unsettled, a search was made in crude homogenates from these cells for enzymes comparable to the flavonoid 3-0- and 7-0glucosyltransferases from parsley cell cultures (Sutter et al., 1972; Sutter and Grisebach, 1973). A flavonol 3-O-glucosyltransferase activity was recognized in crude soya cell extracts, and partial purification was undertaken giving a preparation having a specific activity 5-10 times greater than that of the crude extract. The enzyme preparation glucosylated flavonols predominantly at the 3-position and was specific for flavonoid compounds, since, with the exception of vanillic acid (see below), it exhibited no activity towards cinnamic acids or simple phenols. A detailed study of the phenolic constituents of the cultured soybean cells has not been undertaken (Hahlbrock, 1972; Nimz et al., 1975). However, various kaempferol and quercetin glucosides are present in soybean plants (Buttery and Buzzell, 1975). It is therefore of great interest that these particular flavonols were the best substrates for the flavonol 3-O-glucosyltransferase isolated from soybean cell suspension cultures. Isorhamnetin too was a good substrate, possesing a Vmax/Km ratio around one half of that of kaempferol and quercetin. In contrast to the analogous enzymes from red cabbage (Saleh et al., 1976b) and Haplopappus gracilis cell suspension cultures (Saleh et al., 1976a), little activity towards cyanidin was detected here under all conditions tested. Furthermore, no activity was found with dihydroquercetin, naringenin or 4,2,4'-trihydroxychalcone, or with the isoflavones daidzein and texasin, The former result suggests not only the necessity for a C 2 - C 3 double bond in the heterocyclic ring for good enzyme activity, but also that in vivo 3-O-glucosylation occurs at the flavonol level of oxidation.
The enzyme preparation also catalysed glucosylation at the 7-position of apigenin, luteolin, quercetin and kaempferol, albeit at a relatively slow rate. Although contamination by a minor 7-O-glucosyltransferase activity could not be definitely excluded, data obtained suggested that the flavonol 3-O-glucosyltransferase was responsible for these reactions. It should be noted in this connection that the 3-O-glucosyltransferase from parsley likewise failed to exhibit absolute position-specificity, since it glucosylated apigenin to a small extent at the 7-position (Sutter and Grisebach, 1973). However, no major 7-O-glucosyltransferase activity, analogous to that found in parsley cell suspension cultures, could be detected here at any stage of preparation of the enzyme extract. The inability to detect this enzyme activity perhaps correlates with the finding of apigenin itself and not of its 7-O-glucoside in these cell cultures (Hahlbrock, 1972). During investigation of the substrate specificity of the 3-O-glucosyltransferase preparation, a radioactive product was detected when vanillic acid was supplied as substrate. Since suitable reference compounds were not available, positive identification was not possible. However, the failure to glucosylate isovanillin and vanillin could indicate that the carboxyl group of vanillic acid is being glucosylated in this reaction. A vanillic acid glucosyltransferase has been reported in extracts from Geranium leaves (Comer and Swain, 1965). The soybean vanillic acid glucosyltransferase appeared to be quite distinct from the flavonol 3-O-glucosyltransferase activity, since a partial separation could be achieved (Fig. 1). It is clear that the flavonol glucosyltransferase preparation used for the experiments described here was contaminated by this second enzyme. Recent experiments demonstrated that the soybean FMT methylated luteolin, quercetin and luteolin 7-O-glucoside with comparable K m values (Poulton et al., 1977). In contrast, quercetin 3-O-glucoside and rutin were very poor substrates. These data point strongly to glucosylation as being the last stage in the biosynthesis of methylated flavonol 3-O-gtucosides. For example, isorhamnetin 3-O-glucoside is likely to be formed, not by the methylation of quercetin 3-O-glucoside, but by the glucosylation of isorhamnetin by the flavonol 3-O-glucosyltransferase. The two SAM:3,4-dihydric phenol O-methyltransferases in soybean cell suspension cultures were distinquishable by the variations in their specific activities during the growth cycle of the cell cultures (Poulton et al., 1976a). Whereas no change in the specific activity of the CMT occurred, a pronounced maximum in the specific activity of the FMT, concomitant with the peak in the phenylalanine ammo-
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
nia-lyase activity, was detected. It was suspected that other enzymes of the flavonoid pathway might be regulated in a similar fashion as for FMT. It was therefore surprising to demonstrate here that the specific activity of the soybean flavonol 3-O-glucosyltransferase exhibited a novel and distinct profile, in which it increased with age of the culture, reaching a maximum late in the growth-cycle of the culture. The significance of these results is at present unclear. The authors are very grateful to Prof. H. Grisebach for helpful suggestions during this work, which was undertaken in his laboratory. The work was supported by Deutsche Forschungsgemeinschaft (SFB 46) and by Fonds der Chemischen Industrie. One of us (J.E.P.) thanks the Alexander yon Humboldt-Stiftung for the award of a Fellowship.
References Buttery, B.R., Buzzell, R.I. : Soybean flavonol glycosides: identification and biochemical genetics. Canad. J. Bot. 53, 219-224 (1975) Corner, J.J., Swain, T.: Enzymatic synthesis of the sugar esters of hydroxy-aromatic acids. Nature 207, 634-635 (1965) Ebel, J., Hahlbrock, K., Grisebach, H. : Purification and properties of an o-dihydricphenol meta-O-methyltransferase from cell suspension cultures of parsley and its relation to flavonoid biosynthesis. Biochim. biophys. Acta 269, 313-326 (1972) Ebet, J., Schaller-Heketer, B., Knobloch, K.H., Wellmann, E., Grisebach, H., Hahlbrock, K. : Coordinated changes in enzyme activities of phenylpropanoid metabolism during the growth of soybean cell suspension cultures. Biochim. biophys. Acta 362, 417-424 (1974) Fang, T.-Y., Baisted, D.J.: UDPG:Sterol glucosyltransferase in etiolated pea seedlings. Phytochemistry 15, 273-278 (1976) Hahlbrock, K. : Isolation of apigenin from illuminated cell suspension cultures of soybean, Glycine max. Phytochemistry 11, 165-166 (1972) Hahlbrock, K.: Further studies on the relationship between the rates of nitrate uptake, growth and conductivity changes in the medium of plant suspension cultures. Planta 124, 311-319 (1975) Hahlbrock, K., Ebel, J., Oaks, A., Auden, J., Liersch, M. : Determination of specific growth stages of plant cell suspension cultures by monitoring conductivity changes in the medium. Planta 118, 75-84 (1974) Harborne, J.B., Williams, C.A.: Flavone and Flavonol Glycosides. In: The Flavonoids, p. 376 441. Harborne, J.B., Mabry, T.J., Mabry, H., eds. London: Chapman and Hall 1975
59
H6sel, W., Barz, W. : fl-Glucosidases from Cicer arietinum L. Purification and properties of isoflavone-7-O-glucoside-specific flglucosidases. Eur. J. Biochem. 57, 607 616 (1975) Ibrahim, R.K., Grisebach, H. : Purification and properties of UDPglucose: coniferyl alcohol glucosyltransferase from suspension cultures of Paul's Scarlet Rose. Arch. Biochem. Biophys. 176, 700-708 (1976) Kleinhofs, A., Haskins, F.A., Gorz, H.J.: Trans-o-hydroxycinnamic acid glucosylation in cell-free extracts of Melilotus alba. Phytochemistry 6, 1313-1318 (1967) Layne, E. : Spectrophotometric and turbidimetric methods for measuring proteins. In: Methods in Enzymology, vol. III, p. 447454. Colowick, S.P., Kaplan, N.O., eds. New York: Academic Press t957 Nimz, H., Ebel, J., Grisebach, H. : On the structure of lignin from soybean cell suspension cultures. Z. Naturforsch. 30c, 442-444 (1975) Poulton, J., Grisebach, H., Ebel, J., Schaller-Hekeler, B., Hahlbrock, K. : Two distinct S-adenosyl-L-methionine:3,4-dihydric phenol 3-O-methyltransferases of phenylpropanoid metabolism in soybean cell suspension cultures. Arch. Biochem. Biophys. 173, 301-305 (1976a) Poulton, J., Hahlbrock, K., Grisebach, H. : Enzymic Synthesis of lignin precursors: Purification and properties of the S-adenosylL-methionine:caffeic acid 3-O-methyltransferase from soybean cell suspension cultures. Arch. Biochem. Biophys. 176, 449-456 (I 976 b) Poulton, J., Hahlbrock, K., Grisebach, H. : O~Methylation of flavonoid substrates by a partially purified enzyme from soybean cell suspension cultures. Arch. Biochem. Biophys., in Press (1977) Saleh, N.A.M., Fritsch, H., Witkop, P., Grisebach, H.: UDPGlucose:cyanidin 3-O-glucosyltransferase from cell cultures of Haplopappus gracilis. Planta 133, 41-45 (1976a) Saleh, N.A.M., Poulton, J.E., Grisebach, H. : UDP-Glucose: cyanidin 3-O-glucosyltransferase from red cabbage seedlings. Phytochemistry, in Press (1976b) Sutter, A., Grisebach, H.: UDP-Glucose:flavonol 3-O-glucosyltransferase from cell suspension cultures of parsley. Biochim. biophys. Acta 309, 289-295 (1973) Sutter, A., Ortmann, R., Grisebach, H. : Purification and properties of an enzyme from cell suspension cultures of parsley catalysing the transfer of D-glucose from UDP-D-glucose to fiavonoids. Biochim. biophys. Acta 258, 71 87 (1972) Yamaha, T., Cardini, C.E : The biosynthesis of plant glycosides. I. Monoglucosides. Arch. Biochem. Biophys. 86, 127 132 (t960) Zimmermann, A., Hahlbrock, K.: Light-induced changes of enzyme activities in parsley cell suspension cultures. Arch. Biochem. Biophys. 166, 54-62 (1975)
Received 9 March; accepted 19 April 1977
Pl~)l~'~) 9 by Springer-Verlag 1977
Identification of an UDP-Glucose: Flavonol 3-O-Glucosyl-Transferase from Cell Suspension Cultures of Soybean (Glycine max L.) J.E. Poulton* and Margit Kauer Lehrstuhl f/ir Biochemie der Pflanzen am Biologischen Instltut II der Universit~it Freiburg, Sch~inzlestr. 1, D-7800 Freiburg, Federal Republic of Germany
Abstract. A glucosyltransferase, which catalyses the glucosylation of flavonols, using uridine diphosphateD-glucose as glucose donor, has been isolated and purified about 5-10 fold from cell suspension cultures of soybean ( G l y c i n e m a x L., var. Mandarin). The p H o p t i m u m for this reaction was ca. 8.5 in glycineN a O H buffer, and no additional cofactors were required. The enzyme glucosylated the following flavonols predominantly at the 3-position: quercetin (Kin 126 gM), kaempferol (Kin 172pM), isorhamnetin (Kin 200 pM) and fisetin (Kin 270 gM). With quercetin as substrate, the apparent K m value for uridine diphosphate-D-glucose was 0.3 M. Glucosylation of flavonols and flavones by this preparation occurred weakly also at the 7-position. N o activity was found with dihydroquercetin, naringenin, 4,2',4'-trihydroxychalcone, daidzein or texasin. The enzyme was specific for flavonoid compounds, since no activity was observed towards cinnamic acids or simple phenols. However, the preparation was contaminated by a vanillic acid glucosyltransferase, from which it could be partially separated by ionexchange chromatography. The specific activity of the flavonol 3-O-glucosyltransferase increased with age of the culture, reaching a m a x i m u m late in the growth cycle of the culture. Key words: Cell culture - Flavonoid O-glucosyltransferase - UDP-glucose.
Glycine
-
Introduction Recent experiments have confirmed the presence of two distinct SAM:3,4-dihydric phenol 3-O-methyl-
transferases* in extracts f r o m cell suspension cultures of soybean ( G l y c i n e m a x L.) (Poulton et al., 1976a, b). These enzymes were distinguishable by differences in their stability and substrate specificity. One enzyme (CMT) catalyses the 3-O-methylation of cinnamic acids and protocatechualdehyde, whereas the second methyltransferase (FMT) exhibited a pronounced preference for flavonoid substrates, such as quercetin and luteolin. Interestingly, it was found that the F M T could also methylate luteolin 7-O-glucoside (Poulton et al., 1977). It is at present unclear whether flavonoid glycosides are metabolically active in plants and it still has to be determined whether methylation of these flavonoids precedes glucosylation in vivo or vice versa. In an effort to clarify this position, a search was undertaken in extracts from soybean cell cultures for specific flavonoid glucosyltransferases, which could glucosylate the substrates of F M T and their respective products. We wish here to report the partial purification and characterization of an UDP-glucose:flavonol 3O-glucosyltransferase and to describe how the specific activity of this enzyme is dependent upon the age of the cell culture.
Materials and Methods Chemicab
UDP-D-[U-14C]glucose (296 mCi/mmol) was purchased from the Radiochemical Centre, Amersham, England, and diluted with nonradioactive UDP-glucose from Boehringer, Mannheim. Dihydroquercetin, aplgenin, luteolin, naringenin and fisetin were obtained from Roth (Karlsruhe) and quercetin from Merck (Darmstadt). Eriodictyol was a gift from Dr. Narstedt, Freiburg. Other flavonoid compounds were from our laboratory collection.
Present address: Department of Biochemistry and Biophysics, University of California, Davis, California 95616, USA
Cell Cultures
SAM, S-adenosyl-L-methionine;CMT, SAM :caffeate 3-O-methyltransferase; FMT, SAM:flavonoid O-methyltransferase; UDP-glucose, uridine diphosphate-D-glucose; PAL, phenylalanine ammonia-lyase
a: Cell cultures used for the purification of the enzyme. Soybean ceils (Glycine max L. var. Mandarin) were propagated in the dark in a fermenter containing 300 1 medium I (Hahlbrock et al., 1974; Hahlbrock, 1975). The cells used for the enzyme purification were
*
Abbreviations."
54
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
those which were harvested at a stage in the growth cycle of the culture (stage A) characterized by high phenylalanine ammonialyase and F M T levels (Poulton et al., 1976a; Hahlbrock etal., 1974).
b: Cell cultures used for the age-dependence experiments. The cell cultures used for the time-course experiment were propagated in the dark and sampled as described previously (Poulton et al., 1976a). At each sampling-time, determinations of cell fresh weight and conductivity in the medium were made (Hahlbrock et al., 1974; Ebel et al., 1974). The cells were immediately frozen by immersion in liquid nitrogen and then stored at - 2 0 ~ until all samples of the series had been cellected.
Buffer Solutions The following buffer solutions were used: (I) 0.1 M potassium phosphate buffer, pH 7.6, containing 1.45 m M fi-mercaptoethanot; (II) 20raM potassium phosphate buffer, pH7.6, containing 1.45 m M fi-mercaptoethanol.
Enzyme Assays a: Flavonol glucosyltransferase activity: The incubation mixture consisted of 50 nmol flavonoid acceptor (dissolved in 10 lal ethylene glycol monomethylether), 100 nmol UDP-D-[U-14C]glucose (containing 111,000 dpm), 0.25 mg bovine serum albumin and 25 lxmol potassium phosphate buffer, pH 7.0, in a total volume of 125 Ixl. Where indicated, 25 gmol potassium phosphate, pH6.0, or 25 gmol glycine-NaOH, pH 8.6, were used as the buffer in the assay mixture. The reaction was started by addition of enzyme (20 50 ~tl, depending on extent of purification) and was terminated after an incubation period of 30 rain or 60 min at 30~ by the addition of 10 lal glacial acetic acid. The reaction mixture was then applied in a 3 cm-band to Schleicher and Schtill (2043 b) chromatography paper and chromatographed with solvent system I. After viewing the paper under ultraviolet light, the product zones were cut out and counted in a toluene scintillation fluid (2,5diphenyloxazole/L toluene) in a liquid scintillation spectrometer (Beckmann LS 233). b: Flavonoid 7-O-Glucosyltransferase activity: This activity was assayed either as described above in paragraph (i) or in a modified assay mixture containing only 6.25 nmol flavonoid acceptor (dissolved in 5 gl ethylene glycol monomethylether), since this activity has been previously shown to be inhibited by excess substrate (Sutter et al., 1972). Incubation was carried out for 60 rain. After addition of 10 gl glacial acetic acid, chromatography was undertaken as described above.
graphy with an authentic sample on Schleicher and Schtill (2043b) paper using the following solvent systems: (I) 1% HC1 (RF=0.1), (II) 15% acetic acid (Rv=0.49), (III) n-butanol-acetic acid-H20 (4:1:5, by vol.) (Rv=0.62), (IV) water (RF=0.09). Furthermore, this product exhibited the same behaviour under ultraviolet light in the presence and absence of ammonia vapour as did the authentic 3-O-glucoside. The glucosylation of other flavonoid substrates by the enzyme preparation was routinely followed using solvent system I. The respective 3-0- and 7-O-glucosides of these compounds were identified by their Rv-valnes and by behaviour under ultraviolet light. The possible glucosylation of phenolic compounds and cinnamic acids was monitored by using solvent system III. UDPGlucose remains close to the origin, whereas the substrates had Rv-values in the range 0.75-0.95. The expected glucosylation products have intermediate Rv-values.
Enzyme Purification The stored frozen cells (280 g stage A cells) were thawed out at 4~ and then homogenized with 30 g quarz sand and 360 ml buffer I in a mortar at this temperature. The resultant liquid (530 ml) was centrifuged at 30,000g for 25 min. Dowex 1 x 2 (0.15 g/ml), which had been equilibrated with buffer I, was added to the resulting supernatant. After stirring the mixture for 15 min at 4 ~ the ion-exchanger was removed by filtration through glass-wool. The supernatant liquid was brought to 40% saturation by addition of solid (NH4)2SO4 over a period of l0 min with continuous addition of K O H solution to maintain the pH at 7.0. The mixture was allowed to stand for a further 25 rain at 4~ and was then centrifuged at 30,000g for 15 rain. The supernatant was then brought to 80% saturation with solid (NH4)2SO 4 and the precipitate collected in the same way and dissolved in a minimum volume of buffer II (15 ml). Residual ammonium sulfate was removed from this fraction by applying it to a Sephadex G-25 column (40 c m x 3 cm), which was equilibrated and eluted with buffer II. Fractions containing protein were pooled and applied to a DEAEcellulose column (6 cm • 2 cm), which had been equilibrated with buffer II. After washing the column with 100 ml buffer II, eIution was continued with a linear 20 m M 250 m M potassium phosphate gradient (pH 7.5, containing 1.45 m M fl-mercaptoethanoi; 140 ml total volume). The most active glucosyltransferase fractions were combined and dialysed twice against buffer II for one hour. The enzyme was then concentrated by ultrafiltration and stored deepfrozen at - 2 0 ~ until required for kinetic studies.
Crude Extracts for Age-dependence Experiments e: Glucosyltransferase activity towards other phenolic substrates: 50 nmol of each phenolic substrate (dissolved in 10 ~tl ethylene glycol monomethylether) were tested at pH 7.0 (in phosphate buffer) and at pH 8.6 (in glycine-NaOH buffer) in the assay system as described in paragraph (i) above. After 45 min, the reaction was terminated by addition of 10 itl glacial acetic acid. Chromatography was carried out on Schleicher and Schiill (2043 b) chromatography paper, using solvent system III.
A crude macerate was obtained by stirring 1.0 g of cells with 2 ml of buffer I for 60 rain at 4~ The homogenate was centrifuged for 15 min at 20,000 g. Dowex 1 x 2 (0.3 g), previously equilibrated with buffer I, was added to the supernatant liquid and the mixture was stirred for 15 min at 4~ before centrifugation as described above. The resultant supernatant liquid was used for the estimation of protein content and of PAL and quercetin glucosyltransferase activities.
d.' Assay of PAL activity: Phenylalanine ammonia-lyase activity was determined by a spectrophotometric assay as described elsewhere (Zimmermann and Hahlbrock, 1975).
Protein Estimation Chromatographic Identifieation of Reaction Products Quercetin 3-O-glucoside was identified as the major product of the reaction between UDP-glucose and quercetin by co-chromato-
The protein content was estimated by the biuret method (Layne, 1957); crystalline bovine serum albumin, dessicated before use, was used as standard.
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
Results
Partial Purification of the Enzyme UDP : Glucose: flavonol 3-O-glucosyltransferase activity could be detected in crude homogenates of stage A soybean cells, using quercetin as substrate in the assay method described in the Methods section. Table 1 summarizes the partial purification of this enzyme, enabling a preparation to be obtained which had a specific activity towards quercetin about 5 times greater than that of the crude homogenate. The activTable 1. Partial purification of the soybean UDP-glucose: flavonol 3-O-glucosyltransferase Stage
Total Vol. pro- (ml) tein (mg)
Activity Specific Yield (pkat/ml) activity (% (~tkat/kg) initial)
774
530
2.26
1.55
100
Dowex Treatment
718
520
2.80
2.03
121
Sephadex G-25 eluate (after 40-80% ammonium sulphate step)
145
33
18.50
4.20
51
7.78
12
84
ity in the crude extract was initially concentrated and purified by a 40-80% ammonium sulphate fractionation step. Further purification was achieved by DEAE-cellulose ion-exchange chromatography (Fig. 1). The resultant enzyme preparation, after dialysis and concentration, was stored deep-frozen at -20~ until used for the kinetic experiments described below. No significant glucosyltransferase activity was observed at pH 7.0 with the crude homogenates towards luteolin. Likewise, glucosylation of two flavones (luteolin and apigenin) a flavanone (naringenin) and of two isoflavones (daidzein and texasin) was not detected at pH 7.0 within the 40-80% ammonium sulphate fraction (after removal of ammonium ions by gel-filtration) using the modified assay for 7-0glucosyltransferase activity.
Properties of the Partially-Purified Enzyme
Crude Homogenate
Concentrated DEAE-cellnlose eluate
55
3.3 43.6
Stability: Stability experiments indicated that the quercetin 3-O-glucosyltransferase activity was most stable when stored at - 2 0 ~ under these conditions, only 5% of the initial activity was lost after storage for 2 days. In contrast, 75% of the initial activity was lost after storage at 4~ for this period. Bovine serum albumin (2 mg/ml) was added to the incubation mixture to protect the enzyme from inactivation during incubation at 30~ pH Optimum: The pH optimum for the 3-O-glucosy-
-E i I-
lation of quercetin, determined in several different buffer systems, was found to be around pH 8.54.0 in glycine-NaOH buffer (Fig. 2). Since quercetin has
25(
i I
E
i I
200g -8O
:~ 100 80
150~i
60
10
8
loo i -90
o
~6
.o
40
2"
o
._>
5o2
2o
"/ /. o*o'~176
--
Q.
0
0 5
10 15 Froction
-100
20
Fig. l. Partial separation by DEAE-cellulose ion-exchange chromatography of flavonol 3-O-glucosyltransferase ( 9 and vanillic acid glucosyltransferase (zx zx) activities using a phosphate buffer gradient ( - - - ) at pH 7.5. The protein content of the collected fractions (7.5 ml) was monitored by absorbance at 280 nm ( ). For comparative purposes, the enzyme activities are expressed as the percentage of the maximum activities observed (kaempferol 3-O-glucosyltransferase activity, assayed at pH 7.0, 68 nmol/h/ml enzyme; vanillic acid glueosyltransferase activity, assayed at pH 8.6, 32 nmol/h/ml enzyme). The same elntion profile was obtained when the vanillic acid glucosyltransferase activity was assayed at pH 7.0
ffl
P
4
A
_
2 0
8
_z (.9
0
-
o/
I
I
I
4
6
8
10
pH
Fig. 2. Effect of pH on the quercetin 3-O-glucosyltransferase activity of the partially purified enzyme. Quercetin and UDPG-[14C] were incubated with the enzyme (0.3 rag) under conditions given in the Methods section using 0.2 M concentrations of the following buffers: glycine-NaOH (zx--z~), K H z P O 4 - K 2 H P O , ( e - - e ) and KzHPO4-citric acid ( o - - o )
56
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
Table 2. Substrate specificity of the flavonol glucosyltransferase preparation from soybean cell cultures Substrate
Amount substrate supplied (nmol)
Test pH
Product obtained
Amount product (cpm/30 rain/50 gl enzyme)
Amount product (gkat/kg)
Kaempferol
50
8.6
37Unidentified
5380 270 487
12.2 0.6 1.1
50
7.0
37Unidentified
3654 111 198
8.3 0.2 0.4
50
6.0
37Unidentified
3098 50 180
7.0 0.1 0.4
Quercetin
50
8.6
3 7-
6705 363
15.2 0.8
Cyanidin
50 50 50 25
8.6 7.0 6.0 6.0
333 3-
102 108 179 225
0.2 0.2 0.4 0.5
Apigenin
150 50 25 6.25
8.6 8.6 8.6 8.6
777 7-
195 285 200 8
0.4 0.6 0.4 0
Luteolin
50 2.5 6.25
8.6 8.6 8.6
77 7-
515 328 137
1.2 0.7 0.3
Vanillic acid
50 50
8.6 7.0
Unidentified Unidentified
2487 1537
5.6 3.5
The substrate specificity of the enzyme preparation was tested by incubating the above substrates with 50 gl of the purified enzyme (0.3 mg protein) at the pH value shown for 45 min in the assay system described in the Methods section. No detectable activity was found with the following substrates at either pH 8.6 or 7.0 : dihydroquercetin, naringenin, 4,2',4'-trihydroxychalcone, daidzein, texasin, catechol, hydroquinone, resorcinol, salicylic acid, phloroglucinol, caffeic acid, o-coumaric acid, p-coumaric acid, ferulic acid, coniferyl alcohol, vanillin, isovanillin and anisic acid.
been f o u n d to be u n s t a b l e at this p H ( P o u l t o n et al., 1977), reactions were n o r m a l l y carried out at p H 7.0 in phosphate buffer.
Table 3. Substrate specificity of the soybean flavonol 3-O-glucosyltransferase Substrate
Km V (gM) (gkat/kg)
103 x V/Km (gkat kg ~M 1)
Quercetin Kaempferol Isorhamnetin Fisetin
126 172 200 270
103 86 51 25
Protein and Time Linearity." T h e rate of quercetin 3-O-glucosylation at p H 7.0, catalysed by 0.3 mg of the partially-purified enzyme, was linear for at least 80 min. The extent of 3-O-glucosylation at this pH was p r o p o r t i o n a l to the p r o t e i n a m o u n t added up to at least 0.3 m g of the partially-purified preparation.
Substrate specificity." The substrate specificity of the enzyme was investigated by i n c u b a t i n g various potential acceptors with the enzyme p r e p a r a t i o n at p H 7.0 (in p h o s p h a t e buffer) a n d at p H 8.6 (in g l y c i n e - N a O H buffer). Tables 2 a n d 3 indicate that, within the group of f l a v o n o i d c o m p o u n d s tested, strong glucosyltransferase activity was detected towards quercetin,
13.02 14.70 10.25 6.66
The above values were all obtained from kinetic experiments using the same enzyme preparation, which had been purified as described in the Methods section kaempferol, i s o r h a m n e t i n a n d fisetin. W e a k e r activity was observed with c y a n i d i n , a p i g e n i n a n d luteolin. I n contrast, n o glucosylation o f dihydroquercetin, n a r i n g e n i n or 4 , 2 ' , 4 ' - t r i h y d r o x y c h a l c o n e or of the isoflavones daidzein or texasin was observed.
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
As shown in Table 2, glucosylation of the supplied flavonols was not entirely restricted to the 3-position, since the production of the corresponding 7-O-glucosides from quercetin and kaempferol (amounting to approximately 2-5% of the 3-O-glucoside) was observed. Moreover, a third, as yet unidentified, glucosylation product was obtained with kaempferol. Kinetic parameters for the flavonol substrates of the enzyme were determined in phosphate buffer, pH 7.0. Under these conditions, all substrates are stable and the minor glucosylation reactions are reduced to a minimum. Quercetin was shown to be the best substrate, having the highest Vmax/Km value (Table 3). Glucosylation of the flavones luteolin and apigenin proceeded at a relatively slow rate at the 7-position (Table 2). Various simple phenolic compounds and cinnamic acids were also tested, using the assay system described, in order to determine whether the enzyme was specific for flavonoid compounds. No significant activity was detected with hydroquinone, catechol, resorcinol, salicylic acid, phloroglucinol, caffeic acid, o-coumaric acid (in presence or absence of 10 mM /~-mercaptoethanol), p-coumaric acid, ferulic acid, coniferyl alcohol, vanillin, isovanillin or anisic acid. In contrast, a radioactive product was detected after incubation with vanillic acid as acceptor substrate; the rate of its synthesis was greater at pH 8.6 than at pH 7.0. Positive identification of this product was not possible, since authentic glucosylation products were unavailable. This product had the indicated e F values when re-chromatographed on Schleicher and Schtill paper using the following solvents: (I) n-butanolacetic acid-water (4-1-5, v/v), RF=0.54, (II) 1% HC1, Rv=0.81, (III) 15% acetic acid, Rv=0.86, (IV) 30% acetic acid, Rv=0.85. An effort was made to determine whether the flavonol 3-O-glucosyltransferase was also responsible for the described 7-O-glucosylation reactions and for the glucosylation of vanillic acid, or whether this enzyme preparation was contaminated with other glucosyltransferases. These three activities were monitored in the fractions eluting from the DEAE-cellulose column. Figure 1 demonstrates that a partial separation of the flavonol 3-O-glucosyltransferase and of the vanillic acid glucosyltransferase activities was achieved, suggesting the existence of two distinct enzymes. A definite decision could not be made whether the 3-0glucosyltransferase could also glucosylate flavones at the 7-position, since, although both activities displayed a very similar profile upon elution from the column, only relatively low 7-O-glucosyltransferase activity (with luteolin as substrate, up to 250 cpm above background) was observed.
57 IO.O-
,o~ '~ - ~
zs- -~
_
,'o i
A
~
~
I
I
I
~
~ 9
8
?
20
>,
;'
--
5.0-
I
o
~
I \
~A ....
Z~-.
.,
"N
o
o
_
,
~
,
,
-1.0
,
:>-~'\'~.-.~ ~'--
_
'"
I 100
-1.5E
W.,i-
d
2.5- ~, 10
-213
-0.5
\
I
I ~ I 200 Time of incubation [h]
I 300
Fig. 3. Time course of the changes in the specific activities of PAL ( 9 1 6 9 and quercetin 3-O-glucosyltransferase ( o o) during the period of rapid increase in the cell fresh weight of a soybean culture. The conductivity of the medium ( A - - A ) was measured in addition to the fresh weight of the cells (z~--zs). The conductivity, expressed in the units mhos, is the reciprocal of the electrical resistance, expressed in ohms. The broken lines correspond to parts of the curves which were extrapolated according to previous results
Km for UDP-Glucose The apparent Km for UDP-glucose was 0.3 mM, when assayed at pH 7.0 in the presence of a saturating concentration of quercetin (0.4 mM).
Dependence of the 3-O-glucosyltransferase Activity upon the Age of the Cell Culture The specific activities of phenylalanine ammonia-lyase and of the quercetin 3-O-glucosyltransferase were analyzed during the growth in the dark of a soybean cell suspension culture. As has been reported in previous studies (Poulton et al., 1976a; Ebel et al., 1974), a sharp peak in the specific activity of phenylalanine ammonia-lyase was observed shortly before the conductivity of the culture medium reached its minimum value (Fig. 3). The specific activity of the quercetin 3-O-glucosyltransferase exhibited a distinct profile, in which, following an initial reduction it increased with age of the culture, reaching a maximum late in the growth-cycle.
Discussion
Many glucosyltransferases have been recognized within the Plant Kingdom which catalyse the transfer of a sugar residue from donor molecules to such acceptor molecules as cinnamic acids (Kleinhofs et al.,
58
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
1967), simple hydroxyphenols (Yamaha and Cardini, t 960), sterols (Fang and Baisted, 1976), coniferyl alcohol (Ibrahim and Grisebach, 1976), flavonoids (Sutter et al., 1972; Sutter and Grisebach, 1973), and anthocyanins (Saleh et al., 1976). Despite the accumulation of such knowledge and the fact that flavonoid glucosides are of widespread occurrence in plants (Harborne, 1975), the metabolic role of these compounds in plants is still unclear. Earlier work had suggested that they functioned as metabolic end-products (e.g. perhaps storage products or as pigments), but evidence is now forthcoming demonstrating that they may be further metabolized (H6sel and Barz, 1975; Ebel et al., 1972). The efficient methylation of luteolin, luteolin 7-O-glucoside and of quercetin by an SAM:flavonoid methyltransferase (FMT) purified from soybean cell suspension cultures has been observed in this laboratory (Poulton et al., 1977). Since the problem whether glucosylation of flavonoids may precede methylation is unsettled, a search was made in crude homogenates from these cells for enzymes comparable to the flavonoid 3-0- and 7-0glucosyltransferases from parsley cell cultures (Sutter et al., 1972; Sutter and Grisebach, 1973). A flavonol 3-O-glucosyltransferase activity was recognized in crude soya cell extracts, and partial purification was undertaken giving a preparation having a specific activity 5-10 times greater than that of the crude extract. The enzyme preparation glucosylated flavonols predominantly at the 3-position and was specific for flavonoid compounds, since, with the exception of vanillic acid (see below), it exhibited no activity towards cinnamic acids or simple phenols. A detailed study of the phenolic constituents of the cultured soybean cells has not been undertaken (Hahlbrock, 1972; Nimz et al., 1975). However, various kaempferol and quercetin glucosides are present in soybean plants (Buttery and Buzzell, 1975). It is therefore of great interest that these particular flavonols were the best substrates for the flavonol 3-O-glucosyltransferase isolated from soybean cell suspension cultures. Isorhamnetin too was a good substrate, possesing a Vmax/Km ratio around one half of that of kaempferol and quercetin. In contrast to the analogous enzymes from red cabbage (Saleh et al., 1976b) and Haplopappus gracilis cell suspension cultures (Saleh et al., 1976a), little activity towards cyanidin was detected here under all conditions tested. Furthermore, no activity was found with dihydroquercetin, naringenin or 4,2,4'-trihydroxychalcone, or with the isoflavones daidzein and texasin, The former result suggests not only the necessity for a C 2 - C 3 double bond in the heterocyclic ring for good enzyme activity, but also that in vivo 3-O-glucosylation occurs at the flavonol level of oxidation.
The enzyme preparation also catalysed glucosylation at the 7-position of apigenin, luteolin, quercetin and kaempferol, albeit at a relatively slow rate. Although contamination by a minor 7-O-glucosyltransferase activity could not be definitely excluded, data obtained suggested that the flavonol 3-O-glucosyltransferase was responsible for these reactions. It should be noted in this connection that the 3-O-glucosyltransferase from parsley likewise failed to exhibit absolute position-specificity, since it glucosylated apigenin to a small extent at the 7-position (Sutter and Grisebach, 1973). However, no major 7-O-glucosyltransferase activity, analogous to that found in parsley cell suspension cultures, could be detected here at any stage of preparation of the enzyme extract. The inability to detect this enzyme activity perhaps correlates with the finding of apigenin itself and not of its 7-O-glucoside in these cell cultures (Hahlbrock, 1972). During investigation of the substrate specificity of the 3-O-glucosyltransferase preparation, a radioactive product was detected when vanillic acid was supplied as substrate. Since suitable reference compounds were not available, positive identification was not possible. However, the failure to glucosylate isovanillin and vanillin could indicate that the carboxyl group of vanillic acid is being glucosylated in this reaction. A vanillic acid glucosyltransferase has been reported in extracts from Geranium leaves (Comer and Swain, 1965). The soybean vanillic acid glucosyltransferase appeared to be quite distinct from the flavonol 3-O-glucosyltransferase activity, since a partial separation could be achieved (Fig. 1). It is clear that the flavonol glucosyltransferase preparation used for the experiments described here was contaminated by this second enzyme. Recent experiments demonstrated that the soybean FMT methylated luteolin, quercetin and luteolin 7-O-glucoside with comparable K m values (Poulton et al., 1977). In contrast, quercetin 3-O-glucoside and rutin were very poor substrates. These data point strongly to glucosylation as being the last stage in the biosynthesis of methylated flavonol 3-O-gtucosides. For example, isorhamnetin 3-O-glucoside is likely to be formed, not by the methylation of quercetin 3-O-glucoside, but by the glucosylation of isorhamnetin by the flavonol 3-O-glucosyltransferase. The two SAM:3,4-dihydric phenol O-methyltransferases in soybean cell suspension cultures were distinquishable by the variations in their specific activities during the growth cycle of the cell cultures (Poulton et al., 1976a). Whereas no change in the specific activity of the CMT occurred, a pronounced maximum in the specific activity of the FMT, concomitant with the peak in the phenylalanine ammo-
J.E. Poulton and M. Kauer: Glucosyltransferase from Suspension Cultures
nia-lyase activity, was detected. It was suspected that other enzymes of the flavonoid pathway might be regulated in a similar fashion as for FMT. It was therefore surprising to demonstrate here that the specific activity of the soybean flavonol 3-O-glucosyltransferase exhibited a novel and distinct profile, in which it increased with age of the culture, reaching a maximum late in the growth-cycle of the culture. The significance of these results is at present unclear. The authors are very grateful to Prof. H. Grisebach for helpful suggestions during this work, which was undertaken in his laboratory. The work was supported by Deutsche Forschungsgemeinschaft (SFB 46) and by Fonds der Chemischen Industrie. One of us (J.E.P.) thanks the Alexander yon Humboldt-Stiftung for the award of a Fellowship.
References Buttery, B.R., Buzzell, R.I. : Soybean flavonol glycosides: identification and biochemical genetics. Canad. J. Bot. 53, 219-224 (1975) Corner, J.J., Swain, T.: Enzymatic synthesis of the sugar esters of hydroxy-aromatic acids. Nature 207, 634-635 (1965) Ebel, J., Hahlbrock, K., Grisebach, H. : Purification and properties of an o-dihydricphenol meta-O-methyltransferase from cell suspension cultures of parsley and its relation to flavonoid biosynthesis. Biochim. biophys. Acta 269, 313-326 (1972) Ebet, J., Schaller-Heketer, B., Knobloch, K.H., Wellmann, E., Grisebach, H., Hahlbrock, K. : Coordinated changes in enzyme activities of phenylpropanoid metabolism during the growth of soybean cell suspension cultures. Biochim. biophys. Acta 362, 417-424 (1974) Fang, T.-Y., Baisted, D.J.: UDPG:Sterol glucosyltransferase in etiolated pea seedlings. Phytochemistry 15, 273-278 (1976) Hahlbrock, K. : Isolation of apigenin from illuminated cell suspension cultures of soybean, Glycine max. Phytochemistry 11, 165-166 (1972) Hahlbrock, K.: Further studies on the relationship between the rates of nitrate uptake, growth and conductivity changes in the medium of plant suspension cultures. Planta 124, 311-319 (1975) Hahlbrock, K., Ebel, J., Oaks, A., Auden, J., Liersch, M. : Determination of specific growth stages of plant cell suspension cultures by monitoring conductivity changes in the medium. Planta 118, 75-84 (1974) Harborne, J.B., Williams, C.A.: Flavone and Flavonol Glycosides. In: The Flavonoids, p. 376 441. Harborne, J.B., Mabry, T.J., Mabry, H., eds. London: Chapman and Hall 1975
59
H6sel, W., Barz, W. : fl-Glucosidases from Cicer arietinum L. Purification and properties of isoflavone-7-O-glucoside-specific flglucosidases. Eur. J. Biochem. 57, 607 616 (1975) Ibrahim, R.K., Grisebach, H. : Purification and properties of UDPglucose: coniferyl alcohol glucosyltransferase from suspension cultures of Paul's Scarlet Rose. Arch. Biochem. Biophys. 176, 700-708 (1976) Kleinhofs, A., Haskins, F.A., Gorz, H.J.: Trans-o-hydroxycinnamic acid glucosylation in cell-free extracts of Melilotus alba. Phytochemistry 6, 1313-1318 (1967) Layne, E. : Spectrophotometric and turbidimetric methods for measuring proteins. In: Methods in Enzymology, vol. III, p. 447454. Colowick, S.P., Kaplan, N.O., eds. New York: Academic Press t957 Nimz, H., Ebel, J., Grisebach, H. : On the structure of lignin from soybean cell suspension cultures. Z. Naturforsch. 30c, 442-444 (1975) Poulton, J., Grisebach, H., Ebel, J., Schaller-Hekeler, B., Hahlbrock, K. : Two distinct S-adenosyl-L-methionine:3,4-dihydric phenol 3-O-methyltransferases of phenylpropanoid metabolism in soybean cell suspension cultures. Arch. Biochem. Biophys. 173, 301-305 (1976a) Poulton, J., Hahlbrock, K., Grisebach, H. : Enzymic Synthesis of lignin precursors: Purification and properties of the S-adenosylL-methionine:caffeic acid 3-O-methyltransferase from soybean cell suspension cultures. Arch. Biochem. Biophys. 176, 449-456 (I 976 b) Poulton, J., Hahlbrock, K., Grisebach, H. : O~Methylation of flavonoid substrates by a partially purified enzyme from soybean cell suspension cultures. Arch. Biochem. Biophys., in Press (1977) Saleh, N.A.M., Fritsch, H., Witkop, P., Grisebach, H.: UDPGlucose:cyanidin 3-O-glucosyltransferase from cell cultures of Haplopappus gracilis. Planta 133, 41-45 (1976a) Saleh, N.A.M., Poulton, J.E., Grisebach, H. : UDP-Glucose: cyanidin 3-O-glucosyltransferase from red cabbage seedlings. Phytochemistry, in Press (1976b) Sutter, A., Grisebach, H.: UDP-Glucose:flavonol 3-O-glucosyltransferase from cell suspension cultures of parsley. Biochim. biophys. Acta 309, 289-295 (1973) Sutter, A., Ortmann, R., Grisebach, H. : Purification and properties of an enzyme from cell suspension cultures of parsley catalysing the transfer of D-glucose from UDP-D-glucose to fiavonoids. Biochim. biophys. Acta 258, 71 87 (1972) Yamaha, T., Cardini, C.E : The biosynthesis of plant glycosides. I. Monoglucosides. Arch. Biochem. Biophys. 86, 127 132 (t960) Zimmermann, A., Hahlbrock, K.: Light-induced changes of enzyme activities in parsley cell suspension cultures. Arch. Biochem. Biophys. 166, 54-62 (1975)
Received 9 March; accepted 19 April 1977
