Planta

Planta (1982)155:162-165

9 Springer-Verlag 1982

Cell wall localization of dihydroflavonol-glucoside in flowers of Petunia hybrida

-glucosidase

A.W. Schram, E.J.M. A1, N. D o u m a , L.M.V. Jonsson, P. de Vlaming, A. Kooi, and G.J.H. Bennink* Section Biosynthesis of Flavonoids, Departments of Genetics and Plant Physiology, University of Amsterdam, Kruislaan 318, NL-1098 SM Amsterdam, The Netherlands

Abstract. Limbs o f flower buds f r o m Petunia hybrida were investigated for fi-glucosidase activity with dihydroflavonol-glucosides and 4-methyl-umbelliferylfl-D-glucoside as substrates. Dihydroflavonol-glucoside fl-glucosidase is localized in the cell wall. This activity has an acid p H o p t i m u m and is also active t o w a r d 4-methyl-umbelliferyl-fl-glucoside. Besides this activity a neutral fl-glucosidase is present. This activity is soluble and is n o t active t o w a r d dihydroflavonol-glucosides. Using starch gel electrophoresis it was shown that no difference in fl-glucosidase activity is present between m u t a n t s able to convert dihydroflavonols into anthocyanins and m u t a n t s accumulating dihydroflavonol-glucosides. It is concluded that flglucosidase activity is not involved in a n t h o c y a n i n synthesis. Key words: A n t h o c y a n i n D i h y d r o f l a v o n o l - fl-Glucosidase - M u t a n t (Petunia).

cyanin biosynthesis. To obtain information on this aspect we incubated flower limbs o f Petunia hybrida with glucosides o f dihydroquercetin and with dihydroquercetin aglucone. The glucosides appeared to be deglucosylated before being converted to a n t h o c y a n ins. This indicates the presence o f a fi-glucosidase activity, active t o w a r d dihydroquercetin-glucosides (Schram et al. 1981). We therefore investigated flowers o f Petunia hybrida on the presence o f such a fi-glucosidase activity. We used m u t a n t s h o m o z y g o u s recessive for the gene An3 which are, in feeding experiments, able to convert dihydroflavonols to anthocyanins ( K h o et al. 1975). Furthermore, we used mutants which are h o m o z y g o u s recessive for one o f the genes A n l , An2, An6, or An9. Flowers o f these mutants accumulate dihydroflavonol-glucosides because they are n o t able to convert dihydroflavonols to anthocyanins. The results of these experiments are described in this paper.

Materials and methods Introduction

Plant material. Clones of mutants of Petunia hybrida were grown in the greenhouse.

D i h y d r o f l a v o n o l s are the direct precursors for anthocyanins in flowers of Petunia hybrida ( K h o et al. 1975 ; T a b a k et al. 1978). Mutants, h o m o z y g o u s recessive for one of the genes An1, An2, An6, or An9 are not able to convert dihydroflavonols into a n t h o c y a n ins and they accumulate dihydroflavonols. These dihydroflavonols, however, accumulate as glucosides ( K h o et al. 1977; T a b a k et al. 1978). The m e c h a n i s m o f conversion o f dihydroflavonols to anthocyanins is n o t clear. One o f the problems is discerning whether the glucosides or the aglucones o f dihydroflavonols act as the substrate in antho-

Preparation of cell walls. Cell-flee extracts were isolated from about 50 limbs of flower buds from the white flowering mutant W80 or W37. The limbs were homogenized with an Ultraturrax in 40 ml potassium phosphate buffer (pH 6, 50 mM), containing 20 mM fl-mercaptoethanol and polyvinyl pyrrolidone (K15, Flnka) in an amount equal to the wet weight of the limbs. The homogenate was sonicated 4.15 s (Branson Sonic Power, max. power) and centrifuged for 20 min at 26,000 g. The pellet was washed two times in distilled water containing 20 mM fl-mercaptoethanol, sonicated, and washed again twice with distilled water. The pellet was finally suspended in 30 ml distilled water and designated as the cell wall fraction. Microscopic investigation confirmed that a fraction enriched in cell wall material was obtained.

* Deceased Abbreviations: 4MU-fl-glc= 4-methylumbelliferyl-fi-D-glucopyranoside; dHQ-7-g = dihydroquercetin-7-glucoside; dHQ-4'-g = dihydroquercetin-4'-glucoside; dHM-4'-g = dihydromyricetin-4'-glucoside

Protein extraction. Fifty limbs were homogenized with an Ultraturrax in 40ml phosphate buffer (pH6.0, 50 mM) containing 20 mM fi-mercaptoethanol, Dowex l-X2 (Bio Rad), and polyvinyl pyrrolidone (K15, Fluka) in amounts equal to the wet weight of the limbs. The homogenate was centrifuged at 26,000 g for 20 rain.

0032-0935/82/0155/0162/$01.00

A.W. Schram et al. : fl-Glucosidases in Petunia The supernatant was purified on a Polyclar AT column (dimensions 30-150mm), equilibrated, and eluted with 10mM phosphate buffer, pH 6.0, containing 4 mM fl-mercaptoethanol. The fractions containing protein (and fi-glucosidase activity) were pooled and concentrated by ultrafiltration (Diaflo, PM30). All operations were carried out at 4~ C. In the experiments in which membrane-bound activities were investigated, Triton X-100 (final concentration 0.1%) was added to the extraction buffer as well as to the elution buffer. Protein was measured according to Lowry (1951) with Bovine serum albumin as the standard. Samples were treated with 5% trichloroacetic acid before protein measurements. The pellet was dissolved in 0.1 M NaOH and used in the protein assay.

Enzyme activity measurements. Activities were determined using either the artificial substrate 4-methyl umbelliferyl-fl-glucopyranoside (4-MU-fi-glc) or the flavonoid-glucosides dihydroquercetin-7glucoside, dihydroquercetin-4'-glucoside, and dihydromyricetin-4'glucoside. The concentration of substrates are given in Tables 1 and 2. The activity measurements with artificial substrate were carried out by incubating 4-MU-fl-glc and enzyme in 125 mM sodium acetate (pH 5.0) or potassium phosphate (pH 7.5). The incubation mixture had a final volume of 200 gl. Incubations were carried out at 37~ for 60 min and were stopped with 800 gl glycineNaOH (1 M, pH 10.6). Fluorescence of 4-methylumbelliferon was determined at )~ex 360 nm and )of 450 nm. Activity measurements with fiavonoid glucosides were carried out in a similar way. However, the incubation was terminated by adding 800 ixl chloroformmethanol (2:1, v/v), resulting in a Folch partition (Folch et al. 1957). Part of the upper phase was injected into the liquid chromatograph (Perkin Elmer, series 3). High Performance Liquid Chromatography was performed using a Lichrosorb 10 RP 18 column (24.4.6 ram) and eluted with 12.5% methanol and 5% acetic acid in water (v/v) at a flow rate of 3 ml/min at 30~ C. Detection of dihydroflavonol (glucosides) occurred at 290 nm. Concentrations were calculated from peak areas. Activities were lineair with time and protein concentration.

Starch gel electrophoresis. Four flower limbs from the mutants were homogenized in 200 gl potassium phosphate buffer (50 mM, pH 6.0), in a mortar, in the presence of quartz sand, polyclar AT and fl-mercaptoethanol. Part of these homogenates were subjected to gel electrophoresin on starch (12% in "gel buffer"). The electrophoresis buffer (pH 8.0) consisted of 61 g/1 trishydroxymethylaminomethane (Tris), 6 g/! ethylenediaminetctraacetic acid (EDTA), and 40 g/1 boric acid. A 10-fold diluted electrophoresis buffer was used as the gel buffer. After electrophoresis, the gel was cut lengthwise into two identical slices. One slice was preincubated in a 1 M acetate buffer, pH 5.0; and the other slice in a 1 M phosphate buffer, pH 7.5. After 5 min preincubation, the slices were transferred to the corresponding buffers of lower ionic strength (50 raM). To these buffers, 4-methylumbelliferyl-fl-glucopyranoside was added to a final concentration of 1 mg/ml. After 1-3 h incubation at 37~ C, the gels were transferred to a medium consisting of 1 M glycine-NaOH (pH 10.6) in water; the bands containing 4-methylumbelliferon were visualized by illuminating them with UV light (2ex 360 nm). These bands indicated the position of fi-glucosidase activities on the gel.

Chemicals. Dihydroflavonol-glucosides were isolated from appropriate mutants, as described earlier (Tabak et al. t978).

Results and discussion The results of the extraction of fi-glucosidase from flower buds of the Petunia mutant W80 are summa-

163 Table 1. Extraction of fi-glucosidase from flowerbuds of W80 (1 g wet weight). Activities were measured using 4-methylumbelliferylfl-D-glucopyranoside (0.75 mM) as substrate Fraction

Protein (mg)

Enzyme activity (nkat) pH 5

Supernatant of homogenate Supernatant of homogenate with Triton X-100 Concentrate after polyclar AT Cell wall a

pH7.5

Specific enzyme activity (nkat/mg protein) pH 5

pH 7.5

14.5

1.7

1

0.12

0.07

15.9

1.6

1.1

0. t

0.07

3

0.48

0.27

0.16

0.09

1.5

3.8

0.49

3.1

11.7

a Determined in a separate experiment

100 ~/

'

'

,

IF 50

Oll~

Z~

I

6 8 pH Fig. 1. Effect of pH on fl-glucosidase activities of the mutant W80. o - o soluble activity; o - o cell-wall activity. Substrate: 0.75 m M 4 methylumbelliferyl-fl-glucoside

rized in Table 1. From 1 g (wet weight) material, 14.5 mg soluble protein were extracted. The cell wall fraction contained 3 mg protein. Identical protein concentrations were obtained after an extraction in the presence of Triton X-100. fl-Glucosidase activity was measured using different substrates (4-methylumbelliferyl-fl-glucoside, dihydroflavonol-glucosides). Total fl-glucosidase activity was determined with 4MU-fi-glc, whereas with dihydroflavonol-glucosides as the substrate only specific fi-glucosidases were detected. Over 85% of the total fi-glucosidase activity of W80 at pH 5.0, as measured with an artificial substrate, was recovered in the cell wall fraction, while the specific activity in this fraction was thirty times higher than in the supernatant. This activity is not extracted with Triton X-100, indicating that it is not membrane-bound but rather cell wall-bound. Different results were obtained when measuring the fi-glucosidase activity at pH 7.5. Only 60% of the total fi-glu-

164

A.W. Schram et al. :/~-Glucosidases in Petunia

Table 2. Activity of fl-glucosidase from flowerbuds of W80 (1 g wet weight) using flavonoid-glucosidesas substrate Fraction

Concentrate after polycIar AT Concentrate after polyclar AT with Triton X-100 Call walla

Enzyme activity (nkat) with dihydroquercetin7-g"

dihydroquercetin4'-g~

dihydromyricetin4'-g~

pH5 pH7.5

pH5

pH7.5

pH5

pH7.5

0.09

0.03

0.01

0

0.08 0.01

0.01

0

0.23

0

0.79

0.23

Final concentrations: a 0,23 mM ~ 0.93 mM b 0,45 mM a Determined in a separate experiment cosidase activity at this pH was recovered in the cell wall fraction. The effect of the pH on soluble and cell-wall, bound activities is shown in Fig. 1. The soluble en-

zyme and the cell wall-bound activity exhibited maximal activity at pH 5.0-5.5. However, the shoulder at pH 7.5 in the curve for the soluble activity indicates the presence of a soluble enzyme with a neutral pH optimum. This explains the difference in recovery of enzyme activity in the cell wall fraction at pH 5.0 and pH 7.5 (Table 1). The activities with dihydroflavonol-glucosides as the substrate are summarized in Table 2. Similar to the results obtained with the artificial substrate most of the activity was recovered in the cell wall fraction. The activity at pH 5.0 was always higher than the activity at neutral pH. The recovery at pH 5.0 was identical to the recovery at pH 7.5, using dihydroquercetin-7-glucoside as the substrate. This indicates that the neutral soluble enzyme detected with an artificial substrate is not active with dihydroflavonol glucosides. No effect of Triton X100 was observed on the extraction of/?-glucosidase activity with dihydroquercetin-7-glucoside as the substrate. A slight inhibitory effect on the activity with dihydroflavonol-glucoside could be observed (results not shown). Besides the gene An6 (homozygous recessive in W80), the genes A n l , An2, An3, An5, and An9 regulate anthocyanin synthesis in Petunia hybrida. Mu-

Fig. 2. Starch gel electrophoresis of/%glucosidase extracted from buds of mutants homozygous recessive for An3 (W37, W39, W45), Anl (W20, W46), An2 (W13), An5 (W43), An6 (W58, W80) and An9 (A674-2, A698-1, W75, W84). /~-glucosidase is visualized using 4 methylumbelliferyl-/3-glucoside

A.W. Schram et al. :/~-Glucosidases in Petunia

rants homozygous recessive for one of these genes exhibited no or hardly any anthocyanin synthesis. To investigate whether one of these genes regulates the expression of/~-glucosidase, we compared the activities of homozygous-recessive mutants to one of these genes using starch gel electrophoresis. The results are shown in Fig. 2. Two anodal activities can be recognized: one slow-moving band, active at pH 5.0, and one fast-moving band, active at pH 7.5. Except for the mutant W37 (horn. rec. for An3), all mutants show the acid and the neutral/%glucosidase. The mutant W39 (also homozygous recessive for An3), however, contains the neutral /%glucosidase, which indicates that the gene An3 is not involved in the expression of the neutral /~-glucosidase. Obviously, no difference in fl-glucosidase is present between white flowering mutants able to convert dihydroflavonols into anthocyanins (W37, W39, W43) and mutants accumulating dihydroflavonol-glucosides (W13, W20, W46, W58, W75, W80, W84, A674-2,

A698-1). Conclusions

The cell wall localization of//-glucosidase with dihydroflavonol-glucosides as the substrate makes an intracellular deglucosylation of dihydroflavonol-glucosides unlikely. It explains, however, deglucosylations observed in feeding experiments with dihydroflavonol-glucosides and isolated petals of Petunia hybrida (Schram etal. 1981). Dihydroflavonol-glucosides are hydrolyzed during passage through the cell

165

wall and converted, as aglucone, into anthocyanins. We conclude that /~-glucosidase activity is not involved in anthocyanin synthesis. This confirms the conclusion that deglucosylation and glucosylation of dihydroflavonols is not involved in biosynthesis of anthocyanins. The authors are grateful to Mr. J. Bakker and T.L. Thio for growing the plants.

References Folch, J., Less, M., Sloane-Stanley, G H . (1957) A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 256, 497-509 Kho, K.F.F., Bennink, G.J.H., Wiering, H. (1975) Anthocyanin synthesis in a white flowering mutant of Petunia hybrida by a complementation technique. Planta 127, 1271 1279 Kho, K.F.F., Bolsman-Louwen, A.C., Vuik, J.C., Bennink, G.J.H. (1977) Anthocyanin synthesis in a white flowering mutant of Petunia hybrida. II. Accumulation of dihydroflavonol intermediates in white flowering mutants: uptake of intermediates in isolated corollas an conversion into anthocyanins. Planta 135, 109 118 Lowry, O.H., Roseborough, N.J., Farr, A.L., Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 256 275 Schram, A.W., Timmerman, A.W., Vlaming, P. de, Jonsson, L.M.V., Bennink, G.J.H. (1981) Glucosylation of flavonoids in petals of Petunia hybrida. Planta 153, 459-461 Tabak, A.J.H., Meyer, H., Bennink, G.J.H. (1978) Modifications in the B-ring during flavonoid synthesis in Petunia hybrida: introduction of the 3'-hydroxylgroup regulated by the gene Htl. Planta 139, 67-71 Received 27 December 1981; accepted 17 March 1982