Partial purification, properties, and kinetic studies of UDP-glucose:p-hydroxybenzoate glucosyltransferase from cell cultures of Lithospermum erythrorhizon

ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 288, No. 1, duly, pp. 39-47, 1991

Partial Purification, Properties, and Kinetic Studies of UDP-Glucose:p-Hydroxybenzoate Glucosyltransferase from Cell Cultures of Lithospermum erythrorhizon Andreas Bechthold,

Ursula Berger, and Lutz Heide’

Institut fiir Pharmazeutische Biologie der Rheinischen Friedrich- Wilhelms- Uniuersitiit Nussallee 6, 5300 Bonn 1, Federal Republic of Germany

Received October 23, 1990, and in revised form January

28, 1991

A glucosyltransferase, which catalyzed the transfer of glucose from UIJP-glucose (UDPG) top-hydroxyhenzoate (PHB) in cell cultures of Lithospermum erythrorhizon Sieb. et Zucc., Boraginaceae, was purified 21%fold by ammonium sulfate fractionation and chromatography on DEAE-Sephacel, Sephadex G-150, and phenyl-sepharose Cl-4B. p-Hydroxybenzoic acid G-@-D-glucoside (PHB-glc) was identified as a product of the enzymatic reaction. This glucosyltransferase has a molecular weight of 47,500 Da, an isoelectric point at pH 5.0, and a pH optimum of 7.8. The enzyme does not sediment at 100,OOOg. Enzyme activity did not require metal cofactors. The enzyme was highly specific for p-hydroxyben(K,,, 0.268 mM). zoate (K, 0.264 mM) and UDP-glucose Initial velocity studies suggest that the enzyme reaction mechanism is a sequential rather than a ping-pong mechanism. Product inhibition patterns are consistent with an ordered sequential bi-bi mechanism, where UDPG is the first substrate to bind to the enzyme and UDP the final product released. The data indicate the formation of a dead-end complex between PHB-glc and the enzyme. Uncompetitive inhibition by the substrate PHB can be put down to the formation of an abortive complex between E-UDP and PHB. o 1991 AcademicPress, IW.

Cell cultures of Lithospermum erythrorhizon are used in Japan for commercial production of the naphthoquinone pigment shikonin (1). The biosynthetic pathway of this compound is shown in Scheme 1. Under the influence of light (2), glutamine (3), and NH: (4, 5) the production of shikonin is repressed and p-hydroxybenzoic acid O-@-D-glucoside (PHB-glc) is accumulated instead (6). It has been demonstrated that this 1 To whom correspondence

should be addressed.

0003.9861/91$3.00

Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

Bonn,

metabolic change is regulated by the activities of p-hydroxybenzoate geranyltransferase and p-hydroxybenzoate glucosyltransferase (7). Some properties of the p-hydroxybenzoate geranyltransferase have been investigated (8). In the present study we have examined the substrate specificity, properties, and kinetic mechanism of PHB glucosyltransferase, which is involved in the regulatory branch of the secondary metabolic pathway leading to shikonin. MATERIALS

AND

METHODS

Chemicals UDP-D-[U-%]glucose was purchased from Amersham Buchler. DEAE-Sephacel, Sephadex G-150, phenyl-Sepharose Cl-4B, UDP-hexanolamine-agarose, UDP-glucuronic acid-agarose, and UDP-agarose were obtained from Pharmacia. RP-18 and cellulose TLC plates were from Merck. [U-‘%]PHB, UDPG, UDP, a-glucosidase (EC 3.2.1.90), and @-glucosidase (EC 3.2.1.21) were obtained from Sigma. PHB-glc was a gift of Dr. S. Tanaka (Kyoto University). All other chemicals and solvents were analytic grade reagents.

Plant Material

and Cell Cultures

The cell cultures of L. erythrorhizon Sieb. et Zucc. were derived from germinating seeds (2). By selection from the heterogeneous callus culture, strain Ml8 capable of producing large amounts of shikonin was obtained (9). Cell suspension cultures were initiated and maintained as described elsewhere (10). For the preparation of enzyme extracts, cells were cultured in LS medium supplemented with 10-a M IAA and 1O--5M kinetin.

HPLC Assay for PHB-0-Glucosyltransferase

Activity

In a final volume of 100 gl, the incubation mixture contained PHB* (0.25 pmol), UDPG (2.0 pmol), Tris-HCl, pH 7.6 (4.25 pmol), DTT (85

’ Abbreviations used: Chaps, 3-((3.cholamidopropyl)dimethylammonio)-l-propane sulfonic acid, DTT, dithiothreitol; LS medium, Linsmaier-Skoog medium; PHB, p-hydroxybenzoic acid; PHB-glc, phydroxybenzoic acid 0-P-D-glucoside; PMSF, phenylmethylsulfonyl fluoride; PVPP, polyvinylpolypyrrolidone; TCA, trichloroacetic acid; UDPG, UDP-glucose; IAA, indole-3-acetic acid; enzyme used: UDPglucose:p-hydroxybenzoate glucosyltransferase (EC 2.4.1:). 39

40

BECHTHOLD,

OH0

(1) Solvent system 1: H,O:HCOOH 99:l by vol. (2) Solvent system 2: CH30H:HZO:HCOOH 100:99:1 by vol. (3) Solvent system 3: n-butanol:CH,COOH:HZO 4:1:5 by vol. upper phase.

OH 0

OH

OH

Shikonin

GHQ

GBA

1. Biosynthetic

pathway

of shikonin

and PHB-glc.

Radioisotope Assay for PHB-0-Glucosyltransferase Activity

nmol), PMSF (1.7 nmol), andpurified enzyme (7.5 pg). After incubation for 30 min at 34”C, reaction was terminated by addition of 10 ~1 TCA (240 mg/ml). 3,5Dihydroxybenzoic acid (20 nmol) was added as internal standard. After centrifugation, a sample (24 ~1) was subjected to HPLC

a

DEAE-SEPHACEL ---

HEIDE

TLC Solvent Systems

OH

@-L+&&-~

SCHEME

AND

(column: Multospher RP18 4 pm, 250 X 4 mm; solvent system: (A) HzO: HCOOH 99:l; (B) CH30H:H20:HCOOH 100:99:1, nonlinear gradient, O-SO% B; detection: 254 nm). At 34°C the reaction was linear with time for 90 min, and linear with the amount of purified protein between 6.5 and 120 pg per 100 ~1 incubation volume. No product formation was observed in the absence of one of the substrates or with heat-denaturated enzyme.

PHB-Glc

PHB

Cinnamic acid

Phe

BERGER,

Product inhibition studies with PHB-glc as inhibitors were carried out with radioactive substrates. Using low concentrations of PHB (<0.4 mM), the incubation mixtures contained, in a final volume of 100 ~1, [?Z]PHB (18,500 Bq) and other components as described above. After

b

Gradient:

SEPHADEX

G-150

1

E280 nm in TRIS0.05 H pll 7.6, 1 nUDIT, 20 pHPW’

240

ELUTION

t,

100

720 400 VOLUME (ml)

C

PHENYLSEPHAROSE

260 300 ELUTION VOL”iE

b yml)

CL-46

- - Eluent: 1. Tris 0.05 bi pH1.6,

nm

: 5 i= u a 40.-

-4125

I 75 ELUTION

VOLUME

150 (ml)

FIG. 1. Purification of PHB glucosyltransferase on (a) DEAE-Sephacel, (b) Sephadex G-150, and (c) phenyl-Sepharose Cl-4B. Activity was assayed as described under Materials and Methods, using 85 pl of each fraction. Absorbance at 280 nm is given as an indication of protein content.

Lithospermun TABLE

erythrorhizon

p-HYDROXYBENZOATE

TABLE

I Substrate

Partial

Purification of p-Hydroxybenzoate Glucosyltransferase from 144 g of Cultured Cells of L. erythrorhizon

Purification

step

Crude extract Protamine sulfate precipitation ammonium 33-48s sulfate DEAE-Sephacel Sephadex G-150 Phenyl-Sepharose Cl 4B

Total protein (mg)

Specific activity (pKat/mg)

Purification factor

(pKat)

(%)

466.2

680

100

1.46

1

356.7

850

125

2.38

1.6

80.4 13.2

221 294 97

19.4 151.5 169.2

13.3 103.8

3.9

1560 2000 660

1.5

480

71

320.0

219.2

Formation of O-glucoside (nmol/lOO pl enzyme solution)

p-Hydroxybenzoate o-Hydroxybenzoate m-Hydroxybenzoate 7-Hydroxy+methylcoumarin 6,7-Dihydroxycoumarin 3,3’,4’,5,7-Pentahydroxyflavone 4-Nitrophenol o-Hydroxybenzyl alcohol

Protein Determination Protein was determined according to the method of Bradford bovine serum albumin as standard (12).

using

UDPG(mM)

2.83



(20 mM)

115.9

incubation for 30 min at 34”C, reaction was terminated by addition of 10 ~1 HCl (2 M). The mixture was extracted with 1 ml of HzO-saturated Et20 to remove free [‘%]PHB. After centrifugation and removal of the Et,0 phase, 0.25 pmol PHB and 0.25 nmol PHB-glc were added to the water phase. This solution was applied to RP-18 TLC plates. After chromatography with solvent system 1 and 2 in succession, the spot of PHB-glc was detected under ultraviolet light and scraped out. Radioactivity was determined by scintillation counting. Using high concentrations of PHB (>0.4 mM), the incubation mixtures contained, in a final volume of 100 ~1, UDP-[“Clglucose (9250 Bq) and other compounds as described above. Chromatographic separation was carried out on cellulose TLC plates with solvent system 3. The spots of PHB-glc and UDPG were detected under ultraviolet light, scraped out, and assayed by scintillation counting.

II

Specificity of Purified p-Hydroxybenzoate Glycosyltransferase

Substrate (2.5 mM)

Total activity

41

GLUCOSYLTRANSFERASE

(nmol/lOO

~1 enzyme solution) 3.01 co.01
UDP-glucose TDP-glucose CDP-glucose ADP-glucose GDP-glucose Note. Incubations Methods.

Enzymatic

were carried out as described under Materials

and

Assay of a-Glucosidase and &Glucosidase

or-Glncosiduse. In a final volume of 400 ~1, the incubation mixture contained a-glucosidase (1.28 units), KH,PO,, pH 6.8 (13.4 pmol), and glutathione (0.33 wmol). After preincubation (10 min, 37°C) the respective glucoside (0.039 gmol) was added and incubated 120 min at 37°C.

/%Glucosidase. In a final volume of 400 gl, the incubation mixture contained P-glucosidase (1.63 units), NaCH,COO, pH 5.0 (40 pmol), and the respective glucoside (0.039 wmol). The mixture was incubated 120 min at 37°C. Aglycon formation was assayed by HPLC.

t

z :L m ;: 0.006

1

l/PHB

FIG. 2.

Double-reciprocal

(mM-‘)

plots of initial

l/UDPG

velocity

:

1

(mM-‘j

with UDPG as changing fixed substrate. Inset, slope and intercept

replot.

-

42

BECHTHOLD,

BERGER,

AND

HEIDE

UDP(mM) a

2.0t

P5

l/PHB b

UDP (mtvl)

(mM-l)

t

UDP (mM) ?

1

I

0

i/PHB

45

FIG. 3. Double-reciprocal plots of initial velocity with UDPG concentration constant, PHB as the variable (a) UDPG concentration, 4 mM. (b) UDPG concentration, 20 mM. Insets, slope and intercept replots.

CC-MS For GC-MS analysis, HPLC fractions of the glucoside of several incubations were pooled and evaporated to near dryness. After methylation with CHsJ (11). the methylated compounds were analyzed by CC (DB5 capillary column 30 m X 0.32 mm; 0.25 - ym film; temperature program, 120-320 Co (lO”C/min)); and by MS (Finnigan/Mat 1020 B, 70 eV).

Enzyme Preparation All operations were carried out at 4’C. Cells (144 g) were suspended in KP, buffer (288 ml, 0.1 M, pH 6.5) containing 10 mM DTT, 20 pM PMSF, and 57 g PVPP. The suspension was homogenized by a sonifier and then centrifuged at 25,000g for 30 min. A solution of protamine sulfate (2.33 ml; 20 mg/ml) was added. After centrifugation at 25,000g for 30 min the supernatant was subjected to ammonium sulfate precipitation. The pellet of the 33348% fraction was dissolved in Tris-HC1 buffer (0.05 M, pH 7.6) containing 1 mM DTT and 20 FM PMSF and passed through Sephadex G-25 (PD 10 column) equilibrated with the same buffer. The enzyme solution was chromatographed over DEAE-

5

UDP (mM)

(md)

substrate, and UDP as inhibitor.

Sephacel(2.5 X 25 cm; gradient O-O.5 M KC1 in the same buffer). Active fractions were combined, concentrated, and chromatographed over Sephadex G-150 (2.5 X 55 cm) with 0.05 M Tris-HCl, pH 7.6, 1 mM DTT, 20 KM PMSF as eluent. Active fractions were combined and applied to a phenyl-Sepharose Cl-4B column (2.5 X 8 cm). The glucosyltransferase activity was eluted with 0.05 M Tris buffer, pH 7.6, containing 1 rnrvf DTT, 20 @M PMSF, and 1% Chaps. To remove Chaps, active fractions were passed through Sephadex G-25 (PD 10 column) equilibrated with 0.05 M Tris, pH 7.6, 1 mM DTT, 20 PM PMSF.

Analysis of Kinetic Data The pattern of the enzyme reaction was obtained from douhle-reciprocal plots for two-substrate reactions (13, 14). Reciprocals of velocities were plotted graphically against the reciprocals of substrate concentrations and the intercepts and slopes of the received straight lines were plotted against the reciprocal of the changing fixed substrate concentration. For product inhibition studies, one substrate was kept constant, the second substrate was varied, and the inhibitor (product) was present at

Lithospermum

erythrorhizon

p-HYDROXYBENZOATE

43

GLUCOSYLTRANSFERASE

UDP (mM)

T

1.6

,/

:0 v, 0.6

/’

/’

//’

? J’

I b

L 0 l(UDPG

2.5

5

J

UDP (mM)

(mM-‘)

t UDP (mM)

I

l/UDPG

(mM-‘)

UDP (mM)

FIG. 4. Double-reciprocal plots of initial velocity with PHB concentration constant, (a) PHB concentration, 0.4 mM. (b) PHB concentration, 2.5 mM. Insets, slope replots.

different fixed concentrations. In each case, primary plots of the obtained data and replots of the slopes or intercepts vs inhibitor concentration gave information about the existing type of inhibition. For substrate inhibition studies, data were obtained from double-reciprocal plots for two-substrate reactions using low, high, and very high concentrations of one substrate (13, 18).

Statistic Analysis of Data The data given in Fig. 2 were means of three independent measurements (SD < ?5%). The data were machine fit to the equation for the sequential mechanism using the computer program “Pennzyme” (23-25).

RESULTS

Product Identification The enzymatic product was first identified as PHB-Oglucoside by comparison with the authentic substance using HPLC with a diode array detector to compare both

UDPG as the variable

substrate, and UDP as inhibitor.

retention time and uv spectra. The enzymatic product is a ,&glucoside rather than an a-glucoside as shown by hydrolysis experiments with different glucosidases. While the enzymatic product, authentic PHB-P-0-D-glucoside, and salicin (a P-glucoside) were hydrolyzed by a /!I-glucosidase, but not by an ol-glucosidase, p-nitrophenyl-ocu-D-glucoside was hydrolyzed only by the ar-glucosidase. Using either UDP[14C]glucose or [14C]PHB in the incubation mixture, only one radioactive product was obtained by TLC separation. This product cochromatographed with PHB-glc in all TLC systems described under Materials and Methods. For GC-MS identification, the enzymatic product was isolated by HPLC and methylated with CH,J. Both retention time (13.59 min) and fragmentation pattern were the same as those obtained with the methylated authentic substance (m/z (relative intensity) = 219 (19%), 218 (25%), 187 (21%), 155 (8%), 127

BECHTHOLD,

a

BERGER,

T

PHB-Glc

I

I/UDPG

AND

HEIDE

(mM)

(rnkl:‘)

PHB-Glc

(mM)

b 0.3

I

l

1

0.5

0

l/UDPG

(mM-‘1

1

PHB-Glc

2

(mM)

FIG. 5. Double-reciprocal plots of initial velocity with PHB concentration constant, UDPG as the variable substrate, and PHB-glc (a) PHB concentration, 0.4 mM. (b) PHB concentration, 2.5 mM. Insets, slope and intercept replots.

(13%), 121 (7%), 111 (81%), 101 (loo%), 99 (15%), 95 (13%),89(33%),88(27%),85(8%),75(42%),74(12%), 73 (25%),71(52%),59

(13%),45

(52%)).

Enzyme Purification The enzyme was partially purified by ammonium sulfate fractionation and chromatography on DEAE-Sephaccl, Sephadex G-150, and phenyl-Sepharose Cl-4B (Fig. 1). This procedure led to a 219-fold purification with a yield of 71% (Table I). No purification could be achieved using UDP-hexanolamine-agarose, UDP-glucuronic acid-agarose, or UDP-agarose. Enzyme Stability At 4-8”C, the dialyzed crude extract 90% of its activity within 24 h.

lost more than

as inhibitor.

A solution of the purified enzyme containing 0.01 M Tris-HCl pH 7.2, lost 76% of its activity within 15 days at 4-8°C. Stability could be increased by adding DTT (l10 mM) or PMSF (0.02-l mM) to the enzyme solution. In dilute solutions, addition of bovine serum albumin or glycerol increased stability. The purified enzyme preparations could be stored for 3 months at -20°C in TrisHCl buffer (0.05 M, pH 7.6) containing 1 mM DTT and 20 PM PMSF. During this time the enzyme lost 10% of its activity. pH and Temperature Dependency In Tris-HCl, the pH optimum was at 7.8, with halfmaximal reaction velocity at pH 6.6 and 8.6. Under the conditions described, maximal product formation was obtained at 45’C.

Lithospermum

erythrorhizon

p-HYDROXYBENZOATE

PHB-Glc 3

I

Influence

l/PHB

Double-reciprocal plots of initial velocity Inset, slope and intercept replots.

with UDPG concentration

The enzyme did not require a metal cofactor. Enzyme activity was slightly inhibited by Mg2+, Mn2+, and Ca2’ and was strongly inhibited by Fe2+, Co2+, Ni2+, Cu2+, and Zn2’ at concentrations of 1 and 10 mM. High concentrations of EDTA (10 mM) had a negative influence on the enzyme activity, which was reversible,

TABLE Inhibition

III of

ns

PHB-glc

NC S and I linear

NC S and I linear

UDP

C S linear

C S linear

S

Variable substrate: PHB Constant substrate: UDPG

UDP

when EDTA column). Molecular

as

was removed by Sephadex G-25 (PD 10

Weight

The apparent molecular weight of the transferase was determined as 47,500 + 1500 Da by a chromatography on a Sephadex G-150 column, calibrated with albumin, ovalbumin, ribonuclease, chymotrypsinogene, and Blue Dextran 2000 as reference substances.

Substrate Specificity

Product

PHB-glc

constant (4 mM), PHB as the variable substrate, and PHB-Glc

The enzyme showed an isoelectric point at 5.0, determined by the method of Harzer (19).

Variable substrate: UDPG Constant substrate: PHB

ns

(mM)

Isoelectric Point

p-Hydroxybenzoate Glucosyltransferase Patterns

PHB-Glc

(mM-‘1

of Metal Cofactors

Product

(mM)

2.5

;

FIG. 6. inhibitor.

45

GLUCOSYLTRANSFERASE

s

NC S parabolic I linear

Not detectable

NC S and I linear

No inhibition

In crude extracts, not only PHB but also other phenolic substances (4-nitrophenol, 7-hydroxy-4-methylcoumarin) were glucosylated, indicating the presence of unspecific glucosyltransferases. The purified enzyme fraction, in contrast, was specific for PHB as glucose acceptor and UDPG as glucose donor (Table II). Some glucosylation was found with p-nitrophenol, which is structurally very similar to PHB, but not with any of the other tested compounds. The KIM for p-nitrophenol was determined as 0.6 mM from a Lineweaver-Burk plot, whereas the KM for PHB was 0.264 mM. Kinetic Studies

Note. Abbreviations used: ns, nonsaturating; s, saturating; I, intercept; NC, noncompetitive; C, competitive.

S, slope;

Initial velocity studies. The patterns obtained from double-reciprocal plots for two-substrate reactions gave a family of intersecting straight lines (Fig. 2). Intercept and slope replots versus reciprocal fixed substrate concentrations generated straight lines (inset of Fig. 2).

BECHTHOLD,

BERGER,

AND

HEIDE

PHB

(mM)

f

l/UDPG

(mM-‘)

T

PHB

FIG. 7.

Double-reciprocal

plots of initial

velocity

I

12.5

25 (mM)

with PHB as changing fixed substrate. Insets, slope and intercept

The experimentally obtained data fit the rate equation for a sequential bireactant mechanism (13):

V(A)03) ’ = K&MB + Km(B) + Km(A) + (A)(B) The kinetic data obtained for the PHB glucosyltransferase (A = UDPG, B = PHB) were V = 129 pkat/m.g (SD -t 2%); KiA = 0.191 mM (SD + 14%); KM* = 0.268 mM (SD 2 11%); and Km = 0.264 mM (SD f 3%). Product inhibition studies. Product inhibition of the enzyme was observed with both UDP and PHB-glc. In order to gain more information about the mechanism, product and substrate inhibition patterns of the enzyme were determined (13, 15, 16). UDP is a noncompetitive inhibitor with respect to PHB at nonsaturating levels of UDPG (Fig. 3a). Saturation with UDPG could not be achieved, but inhibition by UDP apparently disappeared at high concentrations of UDPG. As the level of UDPG (constant substrate) was raised from 4 mM (Fig. 3a) over 20 mM (Fig. 3b) to 92 mM (data not shown), the apparent inhibition constant for UDP was elevated from 0.2 mM over 1.1 mM to 6.0 mM, as calculated from slope replots. With respect to UDPG, UDP is a competitive inhibitor as shown at different concentrations of PHB (Figs. 4a, 4b). Therefore, UDP and UDPG combine with the same enzyme form.

replots.

PHB-glc is a noncompetitive inhibitor with respect to both UDPG (Figs. 5a, 5b) and PHB (Fig. 6). The facts that PHB-glc shows noncompetitive inhibition at high concentrations of PHB as constant substrate (Fig. 5b) and that the slope effect is parabolic when UDPG concentration is constant and low, PHB is the variable substrate, and PHB-glc is the inhibitor (Fig. 6) indicate the formation of a dead-end complex between PHB-glc and the enzyme. A summary of the kinetic patterns obtained is given in Table III. Substrate inhibition studies. With UDPG no substrate inhibition could be observed. PHB showed uncompetitive substrate inhibition. While the slopes of the plots shown in Fig. 7 decrease with increasing PHB, the intercepts go

PHB-Glc

t

UP

f

FIG. 8. Proposed kinetic mechanism for the p-hydroxybenzoate cosyltransferase reaction.

glu-

Lithospermum

erythrorhizon

p-HYDROXYBENZOATE

through a minimum and increase again at high concentration of PHB. Reuersibitity of the reaction. In order to test the reversibility of the enzymatic reaction different concentrations of PHB-Glc (up to 0.5 mM) and UDP (up to 10 XnM) were incubated with the purified enzyme. NO UDP-dependent formation of PHB could be observed after various times up to 48 h. DISCUSSION

Investigations of the secondary metabolism in L. erythrorhizon cell cultures had indicated that the glucosylation of PHB may be involved in the regulation of shikonin biosynthesis (7). In crude extracts, an enzymatic glucosylation of PHB could be demonstrated, but clearly unspecific glucosyltransferases contributed to this effect. A purification over several steps (Table I) yielded a glucosyltransferase which was specific for both UDPG and PHB as substrates (Table II). Therefore, this enzyme is to be regarded as a specific secondary metabolic enzyme. The kinetic data of this transferase have been studied in detail. The initial velocity studies (Fig. 2) suggest that the enzyme reaction mechanism is a sequential rather than a ping-pong mechanism (13,17). The data obtained from product inhibition studies (Fig. 3-6) are consistent with an ordered bi-bi mechanism, where UDPG is the first substrate to bind to the enzyme and UDP the final product to be released. Indications for the formation of a dead-end complex between PHB-glc and the enzyme have been provided. The fact that inhibition by PHB-glc with respect to PHB is noncompetitive clearly shows that one central enzyme complex is present and rules out a mechanism of the Theorell-Chance type (13). Substrate inhibition studies finally prove that the enzymatic mechanism is an ordered sequential one. Neither a ping-pong nor a rapid equilibrium random mechanism would show uncompetitive substrate inhibition (13, 18). The kinetic mechanism for this transferase reaction is proposed in Fig. 8. This mechanism is identical to that described for flavonol-ring-B 0-glucosyltransferase from Chrysoplenium americanun with respect to the order of binding of substrates and release of products (20). In that study, product inhibition of substrate B (flavonol) was also observed, but the nature of this inhibition could not be determined. In contrast, the coniferyl-alcohol glucosyltransferase from Picea abies showed a mono-iso-ordered bi-bi mechanism in which a conformational change of the enzyme during the course of the reaction is assumed (21). A PHB-glucosylating activity has been reported from germinating Pinus pollen (22). Other phenolic substrates, however, had a higher affinity to that enzyme than PHB itself, and it therefore appears likely that the Pinus enzyme is not a specific PHB glucosyltransferase.

GLUCOSYLTRANSFERASE

47

Shikonin biosynthesis in L. erythrorhizon cell cultures is blocked under the influence of light by a suppression of PHB geranyltransferase activity and an increase of PHB glucosyltransferase activity (7). Therefore it may be speculated that the physiological importance of the PHB glucosyltransferase lies in the detoxification of accumulated PHB (see Scheme 1). ACKNOWLEDGMENTS We are grateful to Professor M. Tabata (Faculty of Pharmaceutical Sciences, Kyoto University) for the generous gift of L. erythrorhizon cell cultures; to Dr. S. Tanaka (Faculty of Pharmaceutical Sciences, Kyoto University) for providing us with PHB-glc; and to C. Lehmann for excellent technical assistance with cell cultures. This work has been supported by the Deutsche Forschungsgemeinschaft.

REFERENCES 1. Tabata, M., and Fujita, Y. (1985) in Biotechnology in Plant Science (Day, P., Zaitlin, M., and Hollaender, A., Eds.), pp. 207-218, Academic Press, San Diego. 2. Tabata, M., Mizukami, H., Hiraoka, N., and Konoshima, M. (1974) Phytochemistry 13,927-932. 3. Yazaki, K., Fukui, H., Kikuma, M., and Tabata, M. (1987) Plant CellRep. 6,131L134. 4. Fujita, Y., Hara, Y., Ogino, T., and Suga, C. (1981) Plant Cell Rep.

1,59-60. 5. Mizukami, H., Konoshima, M., and Tabata, M. (1977) Phytochemistry 16, 1183-1186. 6. Yazaki, K., Fukui, H., and Tabata, M. (1986) Phytochemistry 25, 1629-1632. 7. Heide, L., Nishioka, N., Fukui, H., and Tabata, M. (1989) Phytochemistry 28, 1873-1877. 8. Heide, L., and Tabata, M. (1987) Phytochemistry 26, 1651-1655. 9. Mizukami, H., Konoshima, M., and Tabata, M. (1978) Phytochem-

istry 17,95-97. 10. Fukui, H., Yoshikawa, N., and Tabata, M. (1984) Phytochemistry 23,301&305. 11. Ciucanu, I., and Kerek, F. (1984) Carbohydr. Res. 131, 209-217. 12. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. 13. Cleland, W. W. (1970) in The Enzymes (Bayer, P. D., Ed.), Vol. 2, pp. l-65, Academic Press, New York. 14. Cleland, W. W. (1963) Biochim. Biophys. Acta 67, 104-137. 15. Cleland, W. W. (1963) Biochim. Biophys. Acta 67, 173-187. 16. Cleland, W. W. (1963) Biochim. Biophys. Acta 67, 188-196. 17. Lasch, J. (1987) Enzymkinetik, pp. 84-92, Springer, Berlin/Heidelberg/New York. 18. Segel, I. H. (1975) Enzyme Kinetics, pp. 819-826, Wiley, New York. 19 Harzer, K. (1970) 2. Anal. Chem. 252, 170-174. 20. Khouri, H., and Ibrahim, R. K. (1984) Eur. J. Biochem. 142, 559564. 21. Schmid, G., and Griesebach, H. (1982) Eur. J. Biochem. 123, 363370. 22 Katsumata, T., Shige, H., and Ejiri, S. (1989) Phytochemistry 28, 359-362. 23 Eisenthal, R., and Cornish-Bowden, A. (1974) Biochem. J. 139, 715-720. 24 Cornish-Bowden, A., and Eisenthal, R. (1974) Biochem. J. 139, 721-730. 25 Cornish-Bowden, A. (1977) Biochem. J. 165, 55-59.