Phytochemistry, Vol. 30, No.9, pp. 2865-2867, 1991 0031-9422/91 $3.00 +0.00 Printed in Great Britain. Pergamon Press pic CHARACTERIZATION OF MAIZE ENDOSPERM UDP-GLUCOSE: DOLICHOL-PHOSPHATE GLUCOSYLTRANSFERASE JAN A. MIERNYK and WALTER E. RIEDELL* Seed Biosynthesis Research Unit, USDA, Agricultural Research Service, National Center for Agricultural Utilization Research, Peoria, IL 61604, U.S.A. (Received ill revised form 12 February 1991) Key Word Index-Zea mays; Gramineae; maize; endosperm; tissue culture; dolichol cycle; glycolipid; glycosyl­ transferase. Abstract-Uridine-diphosphate-glucose:dolichol-phosphate glucosyltransferase [EC 2.4.1.1l7J was solubilized from maize endosperm microsomes by treatment with 0.25% Triton X-IOO. The solubilized enzyme was partially purified by chromatography on columns of hydroxylapatite, diethylaminoethyl-Sephacel, concanavalin A-Sepharose, and uridine-diphosphate-hexanolamine-agarose. The purified transferase had no activity unless a lipid-acceptor was included in the assays. K m values for uridine-diphosphate-glucose, dolichol-phosphate, and magnesium were 0.5 /lM, 1 1.2 /lg ml- , and 0.7 mM, respectively. The K j values for uridine-diphosphate, uridine-monophosphate and uridine­ diphosphate-glucuronic acid were 150, 650 and 7.6/lM, respectively. The in vitro activity of the transferase was inhibited 50% by 40 /lg ml- 1 of tunicamycin. INTRODUCTION were less efficient in terms of either solubilization or stabilization of the enzyme (data not presented). The We have previously presented the results of detailed results of a typical purification are presented in Table 1. studies on the localization of the dolichol-cycle enzymes The degree of purification ranged from 80- to 100-fold from maize endosperm cells [1]. During the preliminary with recoveries from 7 to 10%. The final preparations characterization of the glycosyltransferases it was ob­ were not homogeneous, and when examined by SDS­ served that uridine-5'-diphosphate-glucose (UDP-Glc): PAGE there were three major and three minor bands lipid Glc-transferase activity in vitro was not dependent (data not presented). We were not able to determine upon added dolichol-phosphate (dolichol-P) acceptor. At which band(s) corresponded to the Glc-transferase. Re­ that time we speculated that the activity assayed might be cently, it was reported that the subunit mass of the yeast a combination of two UDP-Glc:lipid Glc-transferases, UDP-Glc:dolichol-P Glc-transferase is 35000 [3J; how­ one using endogenous sterols [e.g. 2J as the lipid acceptor ever none of the bands in the purified maize transferase and the other using endogenous dolichol-phosphate [1]. corresponded to this M value. There has previously been The solubilization and separation of the Glc-transferases r difficulty in purifying the enzymes of the dolichol cycle was undertaken in order to characterize the dolichol­ because ofinstability after solubilization [e.g. 4]. We were cycle enzyme without complications in interpretation of able to achieve some degree of stabilization by the the results due to competing reactions. inclusion of protease inhibitors plus UDP-Glc in the isolation and solubilization buffers. RESULTS AND DISCUSSION Anion-exchange chromatography resolved solubilized Glc-transferase activity into two peaks (Fig. 1). The Maize UDP-Glc:dolichol-P Glc-transferase was solu­ activity in the first peak was relatively high even without a bilized from washed microsomes with 0.25% Triton lipid-acceptor included in the assay mixtures and was X-lOO, as previously described [lJ, and purified by increased less than two-fold by addition of dolichol-P. sequential hydroxylapatite-, anion-exchange-, immob­ The second peak, eluting at a conductivity of ca ilized lectin- and affinity-chromatography. The maize 7 mMho, had very low activity when assayed without a enzyme was not solubilized by treatment of the micro­ lipid-acceptor and was stimulated more than lOO-fold somes with 1.0 M KCI or alkaline sodium carbonate. when dolichol-P was added to an assay previously lack­ Other detergents, including Tween-20, Brij 35, Nonidet ing acceptor (data not presented). The later eluting peak P-40, and the tensides octyl-glucoside and dodecyl­ of Glc-transferase activity from the anion-exchange col­ maltoside, at concentrations ranging from 0.1 to 1%, umn is the dolichol-cycle enzyme, and further purification of this activity was undertaken. After two additional chromatography steps (Table 1) there was no detectable *Present address: USDA, Agricultural Research Service, Glc-transferase activity without addition ofa polyprenol­ Northern Grain Insects Research Laboratory, RR No.3, P acceptor to the enzyme assays and there was zero Brookings, SD 57006, U.S.A. activity when 10 /lg ml- 1 of soybean p-sitosterol was 2865 2866 1. A. MIERNYK and W. E. RIEDELL Table I. Partial purification of UDP-glucose:dolichol-phosphate glucosyltransferase from maize endosperm cultures Total Specific Units protein activity Fold- Yield I Fraction (nkat) (mg) (units mg- ) purification (%J Solubilized microsomes 116 750 0.15 1 100 Hydroxyapatite 102 283 0.36 2.4 88 DEAE-Sephacel 27 6.8 3.97 26.4 23 Con A-Sepharose 21 4.8 4.37 29.1 18 UDP-hexanolamine-agarose 8 0.6 13.11 87.3 7 Starting material was 1.5 kg of seven-day-old cells. Table 2. Some characteristics of partially purified maize endosperm UDP-glucose: dolichol-phosphate glucosyltransfer­ ase Optimum [Triton X-I00] (%) 0.Q15 pH optimum 7.5 K m UDP-glucose (tIM) 0.5 1 K m dolichol-P (pgml- )* 1.2 K m MgCl z (mM) 0.7 K i UDP (pM) 150.0 K i UMP(pM) 650.0 K i UDP-glucuronic acid (pM) 7.6 Fraction Number 10 .5 tunicamycin (pg ml- 1) 40.0 Fig. 1. Separation of the maize endosperm UDP-glucose: lipid *The kinetic constants for dolichol-P and tunicamycin are glucosyltransferase by anion-exchange chromatography. both presented as mass volume - 1 values rather than as molar Proteins solubilized from microsomal membranes by treatment concentrations because both compounds are mixtures of mo­ with 0.25% Triton X-IOO were chromatographed on a 1.6 x 22 lecular species. column of DEAE-Sephacel previously equilibrated with 10 mM Kinetic constants were derived from initial-rate studies ana­ TES pH 7.5 containing 0.25% Triton X-lOO. After washing, lysed by iterative curve-fitting using nonlinear regression [11]. bound proteins were eluted with a linear gradient of 0 to 0.5 M KCl in equilibration buffer. Maximum activity was 2.4 nkat fraction -I. Recovery of activity in fractions 12-34 was 69% of that loaded onto the column. Tunicamycin is a potent inhibitor of UDP­ GlcNAc:dolichol-P GlcNAc-transferase activity both in vivo and in vitro [5]. It is thought that tunicamycin resembles a transition-state reaction intermediate. The K i substituted for dolichol-P. At equal substrate concentra­ value for the chick embryo GlcNAc-transferase is 5 nM tions, activity with decaprenol-P was 43% of that with [5J, and we previously reported that the K i for the maize dolichol-P (data not presented). Preparations purified endosperm GlcNAc-transferase is 14 ng ml-! [1]. While through the affinity-chromatography step were used for our observation that tunicamycin also inhibited maize in vitro catalytic and kinetic analyses. endosperm UDP-Glc: dolichol-P Glc-transferase was un­ In vitro catalytic activity of the purified dolichol-cycle expected, the previous results were verified with the transferase was maximal at a Triton X-100 concentration purified enzyme (Table 2). Because the kinetic mechanism of 0.015%, and pH value of 7.5 (Table 2). This pH of inhibition of the maize enzyme is unknown, we are optimum is significantly more alkaline than the pre­ presenting an 10 .5 value instead of a K i value. It was viously reported value [lJ which was the combination of previously reported that relatively high concentrations of dolichol and sterol transferase activities. Additionally, the tunicamycin nonspecifically inhibited protein synthesis, K m value for UDP-Glc was 10-times lower than the mRNA synthesis, and even the transport of UDP-galac­ previously reported value [lJ, which also was the combi­ tose across Golgi membrane vesicles [5J, and it could be nation of the two enzyme activities. Other kinetic con­ that the effect upon the maize endosperm Glc-transferase stants (Table 2) for the purified dolichol transferase are is this type of nonspecific inhibition. similar to those previously reported. UDP-glucuronic During purification, maize endosperm Glc-transferase acid was a potent inhibitor of the maize dolichol trans­ was adsorbed to immobilized concanavalin A and then ferase (Table 2). Inhibition was competitive with respect specifically eluted with C(-methyl-D-mannoside. This to UDP-Glc (data not presented) and the K i value was lectin has a binding selectivity for high-mannose type 7.6 tiM. UDP-glucuronic acid is also a potent inhibitor of glycans [6]. The enzymes in plant cells that convert high­ the mammalian GIc-transferase, with Shailubhai et al. [4J mannose type glycans to the complex-type are localized reporting 10 .5 of 15 tlM. within the Golgi apparatus [7]. The occurrence of Maize endosperm glucosyltransferase 2867 high-mannose glycans on maize endosperm UDP­ directly onto a 1.6 x 20 cm column of con A-Sepharose pre­ Glc:dolichol-P Glc-transferase is consistent with the viously equilibrated with 10 mM TES, pH 7.5, containing 0.25% localization of this enzyme as a resident protein of the Triton X-IOO, 100 mM KCl, 1 mM MgCI 2 , 1 mM CaCl2 and rough-ER [1]. It has been reported that the plant UDP­ 1 mM MnCI2 . After washing with equilibration buffer the bound Glc:sterol Glc-transferase is localized in the Golgi appar­ proteins were specifically eluted with 250 mM x-methyl-D­ atus and plasma membrane [8].