Golgi-specific localization of transglycosylases engaged in glycoprotein biosynthesis in suspension-cultured cells of sycamore (Acer pseudoplatanus L.)

ARCHIVES

OF BIOCHEMISTRY

Vol. 251, No. 2, December,

AND BIOPHYSICS

pp. 421-431, 1986

Golgi-Specific Localization of Transglycosylases Engaged in Glycoprotein Biosynthesis in Suspension-Cultured Cells of Sycamore (Acer pseudoplatanus L.) MD. SHOWKAT ALI, TOSHIAKI

MITSUI,

AND

T. AKAZAWA’

Research Institute fw Biochemical Regulation, School of Agriculture, Nagwa University, Chikusa, Nagoya. 464, Japan Received May 15, 1986, and in revised form August 8,1986

Golgi complex and endoplasmic reticulum (ER) were isolated from suspension-cultured cells of sycamore (Acer pseudoplatanus L.) by stepwise sucrose density gradient centrifugation using protoplasts as starting material. The purity of the two organelle fractions isolated was assessed by measuring marker enzyme activities. Localization of glycolipid and glycoprotein glycosyltransferase activities in the isolated Golgi and ER fractions was examined; three glycosyltransferases, i.e., galactosyltransferase, fucosyltransferase, and xylosyltransferase, proved to be almost exclusively confined to the Golgi, whereas the ER fractions contained glycolipid glycosyltransferase. The Golgi complex was further subfractionated on a discontinuous sucrose density gradient into two components, migrating at densities of 1.118 and 1.127 g/cm3. The two fractions differed in their compositional polypeptide bands discernible from Na-dodecylsulfate gel electrophoresis. Galactosyltransferase distributed nearly equally between the two protein peaks and xylosyltransferase activities using the endogenous acceptor also appeared to be localized in the two subcompartments. By contrast, fucosyltransferase, engaged in the terminal stage of glycosylation, banded in the lower density fractions. Golgi-specific a-mannosidase, which is presumably engaged in the sugar trimming of Asn-N-linked glycoprotein carbohydrate core, was enriched fourfold in specific activity in the fractions of the higher density. The overall experimental results indicate that the cotranslational glycosylation of Asn-N-linked glyeoproteins, e.g., polyphenol oxidase (lactase), takes place in the ER, while subsequent post-translational processing of the oligosaccharide moiety proceeds successively in the two physically separable compartments of the Golgi complex. @ 1986 Academic PRESS.k.

The biosynthesis of glycoproteins in eukaryotic cells is carried out in successive stages. The polypeptide chain elongation directed by the polysomes attached to the ER’ membrane is coupled to its segregation

into the cisternae (l-5). The primary glycosylation reaction occurring cotranslationally in this step involves the en bloc transfer of the oligosaccharide moiety to the polypeptide, which is generally called the dolichol-P pathway. Before the subsequent transport to the Golgi complex through the transitional elements or vesicular buds, there occurs tne primary deglycosylation of outer residue of carbohydrate core, removal of glucose and mannose residues, in the ER. The Golgi complex is responsible for the secondary trimming of

’ To whom correspondence should be addressed. ’ Abbreviations used: cyt-c, cytocbrome c; EM, electronmicroscopy; GlcNAc, N-acetylglucosamine; Mes, 2-(N-morpholino)ethanesulfonic acid; PAGE, polyacrylamide gel electrophoresis; PNP-Man, pnitrophenol-cy-mannopyranoside; SDS, Na-dodecylsulfate; TCA, trichloroacetic acid; ER, endoplasmic reticulum. 421

0003-9861/86 Copyright All rights

$3.00

0 1986 by Academic Press, Inc. of reproduction in any form reserved.

422

ALI,

MITSUI,

AND

mannose as well as the addition of terminal or capping sugars, i.e., GlcNAc, galactose, fucose, and sialic acid, to make up the complex-type glycoprotein molecules (3-6). The protein molecules are then transported to the cell surface by secretory vesicles and are finally discharged across the plasma membranes (5). The overall mechanism of the biosynthesis and transport of glycoproteins, frequently referred to as exocytosis, has been established mostly on the basis of experiments using mammalian systems. By contrast, experimental evidence with plant cells is fragmentary and rather ambiguous, because of technical difficulties in isolating functionally competent Golgi membranes (7). In recent years, efforts have been made to elucidate the role of lipid-linked saccharide in the biosynthesis of plant glycoproteins using the particulate enzyme preparations from different plant materials (8). Mellor et al. (9) reported that the ER membranes isolated from castor bean endosperm catalyzed the core glycosylation of denatured ribonuclease A molecules. The principal location of GlcNAc transferase and mannosyltransferase in the rER of pea cotyledons was prepared by Nagahashi and Beevers (10). Conder and Lord (11) found the heterogenous distribution of glycolipid glycosyltransferase activities in the two ER subfractions isolated from the castor bean endosperm; however, except for a few cases no detailed analyses have been carried out showing the pivotal role of the Golgi complex in the post-translational modification of glycoproteins in plants. Membranous fractions derived from the Golgi apparatus of carrot root (12) and onion stem (13) have been shown to contain glycoprotein glycosyltransferase activities. The enzyme activities transferring the fucosyl moiety from GDP-fucose to the membrane glycoprotein are almost exclusively confined to the Golgi vesicles in cucumber cotyledons (14). By employing [‘HIfucose pulse-chase experiments, Vitale and Chrispeels (15) reported the Golgi-mediated transport of phytohemagglutinin to the protein bodies in bean cotyledons. A similar pulse-chase in viva study on the glycoprotein processing was

AKAZAWA

carried out by Hori and Elbein (16) using soybean cells. Recently, Mitsui et al. (17) reported that the Golgi complex is most likely engaged in the Caz+-dependent secretion of a-amylase molecules in rice seedlings. There are well-developed Golgi complex and rER in sycamore cells as revealed by EM observation (Fig. 1). Since we have developed a rapid and simple method for isolating purified Golgi fractions (18, 19), it is deemed that the system provides us a good chance to reveal the mechanisms underlying the biosynthesis and intracellular transport of glycoprotein molecules, and furthermore compartmentation of glycosyltransferases in the Golgi complex. The extracellular secretion of a large amount of lactase-type polyphenol oxidase in the sycamore cells was reported previously by Bligny and Deuce (20), and recent structural investigations have disclosed the xylose-containing biantennary complextype structure of the enzyme molecules (21). In the present investigation, we first attempted to separate ER and Golgi fractions, and subsequently examined the association of various transglycosylases presumably engaged in the biosynthesis of glycoprotein with the Golgi membranes. Although little is known about the nature of the enzyme responsible for the conjugation of xylose to the N-linked oligosaccharide, based on our current knowledge of glycoprotein biosynthesis (3), it is most likely to be a post-translational event occurring in the Golgi. The basic procedure employed throughout this investigation is the stepwise sucrose density gradient centrifugation technique, which was used to demonstrate the compartmentation of the Golgi membrane system in animal cells (22-24). MATERIALS

AND

METHODS

Chaicals. NADH, cyt-c, UDP-glcNAc, and dolicholP were purchased from the Sigma Chemical Company (St. Louis, MO.). UDP$J-“Cjgalactose (309 mCi/ mmol), GDP-[“Clmannose (203 mCi/mmol), GDP[%]fucose (292 mCi/mmol), and UDP-[“Clxylose (9.9 Ci/mmol) were purchased from Amersham (England).

TRANSGLYCOSYLASE

LOCALIZATION

IN GOLGI

FIG. 1. EM structure of sycamore cell. A, amyloplast; CW, cell wall; rER, rough endoplasmic reticulum; G, Golgi complex; PM, plasma membrane; SV, secretion vesicle. In the picture, subcompartments of Golgi, cis, medial, and trans, are not defined. Photograph was taken by A. M. Tartakoff.

423

424

ALI,

MITSUI,

AND

IDP and other chemicals of reagent grade were purchased from the Wako Chemical Company (Tokyo). Plant materials. Sycamore (Acer pseudqhtanus L.) cells were grown under shaking conditions in 500-ml Erlenmeyer flasks containing 200 ml of culture medium as reported previously by Bligny and Deuce (20). Preparatkm of Go&i and?ER fractians The isolation method of the Golgi complex and ER involves the preparation of protoplasts from sycamore cells (ca. 50 g fr wt) followed by their mechanical disruption and subsequent differential centrifugation at 100,000~ for 1 h using a Beckman R30 rotor. The 100,000~ pellets, resuspended in a small volume of buffer containing 50 mrd glycylglycine (pH 7.5), 1 mM EDTA, and 0.5 M mannitol, were layered onto a discontinuous sucrose density gradient solution, containing 3 ml each of 50, 40,30,20, and 10% (w/w) sucrose dissolved in the same buffer but without mannitol. The gradients were centrifuged at 21,000 rpm for 3 h using a Beckman SW 25.3 rotor. In order to obtain homogenous fractions of ER and Golgi complex, the two fractions which banded at the interface between 10/20% and 20/30% sucrose solution, respectively, were collected separately, and after suitable dilution, were sedimented again by centrifugation at 40,000 rpm for 1 h using a Beckman R-50 rotor. After resuspending the resulting pellets in 1 ml of the same buffer, the fractions were applied again to the five-step gradient solution of sucrose as before and centrifugation was repeated under the same conditions. The ER and Golgi fractions obtained were fractionated separately using an ISCO density gradient fractionator (see Fig. 2). Preparation of Gdgi subf~actkms. The purified Golgi fractions were suspended in 3 ml of 40% sucrose, and gently homogenized using a Teflon homogenizer. The resulting homogenate was loaded onto a discontinuous sucrose gradient, consisting of 4 ml each of 30,20, and 10% sucrose dissolved in 50 mM glycylglycine (pH 7.5) and 1 mM EDTA. The gradient was centrifuged at 21,000 rpm for 16 h in a Beckman SW 25.3 rotor and fractionated for further assays (see Fig. 7). Enzyme assays. Three transglycosylases, i.e., galactosyltransferase, fucosyltransferase, and xylosyltransferase, were assayed using either endogenous membranes of GlcNAc as the acceptor molecules. Unless otherwise indicated in each case, the reaction system using GlcNAc as the glycosyl acceptor was exactly based on the method reported previously for galactosyltransferase (18), following the original report of Powell and Brew (13), whereas the assay system for the incorporation of “C-labeled sugar into the TCA-insoluble endogenous acceptor varied from one system to another but principally was also based on the method reported by Powell and Brew (13). In each assay, a zero time blank was taken, and the difference in radioactivities between the experimental and the control was expressed as the transferase activities.

AKAZAWA

(a) Ga4actosyltransjbrose. The standard reaction mixture for assaying the incorporation of [“Clgalactose from UDP-[14Clgalactose into the TCA-insoluble material using endogenous acceptor molecules contained in a total volume of 100 pl 30 mM sodium cacodylate buffer (pH 7.5), 30 mM MnCls, 0.001% Triton X-100, UDP-[i4Clgalactose (0.05 &i, 10 mCi/mmol), and 70 pl of Golgi or ER membrane fractions. The reaction was carried out at 30°C for selected time periods. For the radioactivity measurements of the TCAinsoluble fraction, an aliquot of the reaction mixture was placed on glass filter (Whatman GF/C) discs. After thorough washing in 5% TCA solution, ethanol, and ether, discs were subjected to liquid scintillation counting. (b) Fucosyltransferase. The incorporation of [i4C]fucose from GDP-[“Clfucose into the TCA insoluble endogenous acceptor was assayed in a reaction mixture (total volume of 100 ~1) containing 100 mM Tris-HCl (pH 7.4), 10 mM MgCla, 2 mM fl-mercaptoethanol, and 0.001% Triton X-100, GDP-[14Clfucose (0.05 pCi, 292 mCi/mmol), and 70 ~1 of Golgi or ER fractions. The reaction mixture was incubated at 30°C and radioactivities were measured following the same method as for galactosyltransferase. (c) Xylosyltransferase. The incubation mixture (100 ~1) contained 50 mM Tris-Mes (pH 5.0), 5 mM MnCla, 2 mM P-mercaptoethanol, 0.001% Triton-X 100, UDP[H31xylose (0.1 pCi, 9.9 CVmmol) and either Golgi or ER membranes. Unless otherwise indicated the reaction was carried out at 37°C and after 9 min incubation, an aliquot of the reaction mixture was placed on glass filter discs. After thorough washing in 5% TCA, ethanol, and ether, discs were subjected to the liquid scintillation counting. (d) Mannosyltransferase. Two hundred microliters of the ER fractions was added to the reaction mixture (total volume of 250 @I) containing 100 mM Tris-HCl (pH 7.4), 10 mM MgCl*, 2 mM j3-mercaptoethanol, 0.1% Triton X-100, and 20 pg of dolichol-P. The reaction was initiated by adding GDP-[“Clmannose (0.1 &i, 203 mCi/mmol) and the whole mixture was incubated at 30°C for 10 min. The reaction was stopped by adding 3 ml of chloroform:methanol (1:l) and 1 ml of HrO, and the incorporation of [“Clmannose into (i) chloroform:methanol (l:l)-soluble (monosaccharide lipid) and (ii) chloroform:methanol:HrO (1:1:0.3)-soluble (oligosaccharide lipid) fractions was determined as described previously (25). (e) Other enzymes. NADH-cyt-c reductase was assayed spectrophotometrically following the reduction of cyt-c in the presence of NADH and antimycin according to the method described by Lord et al (26). IDP-ase was assayed according to the method previously reported (18). a-Mannosidase was assayed basically following the method reported by Agrawal and Bahl(27); the standard reaction mixture (total volume 300 ~1) contained buffer, 25 mM PNP-Man, 0.001% 50 mM Na-citrate

TRANSGLYCOSYLASE

LOCALIZATION

Triton X-100, and 100 ~1 of Golgi or ER fractions. Incubation was carried out at 30°C for 15 min. The reaction was stopped by the addition of 700 ~1 of 0.2 M sodium carbonate and the released PNP was measured at 400 nm. The specific assay conditions are described in legend for each respective figure. SDS-PAGE. Protein compositions of the whole Golgi fraction and its subfractionated samples were examined by the SDS-PAGE according to the method of Laemmli (28). Protein an&&a Protein content was analyzed using Bio-Rad protein reagent, using bovine gamma globulin as a standard.

G&o&id

IN GOLGI

425

Glycosyltransferase

Employing mammalian tissues, it has been shown that an array of enzyme reactions engaged in the dolichol pathway is localized in the ER (3,6). ER fractions obtained in the present investigation were subjected to the assay of mannosyltransferase activity catalyzing the transfer of [‘*C]mannose from GDP-[‘4C]mannose to the lipid acceptor, dolichol-P, and the subsequent assembly of monosaccharide lipid soluble in chloroform:methanol to the chloroform:methanol:HzO-soluble oliRESULTS gosaccharide lipid (29). The distribution of these two enzyme reactions in the ER is Characterization of Golgi presented in Fig. 4A. Results show that acand ER Membranes tivities for the formation of both monosaccharide lipid and oligosaccharide lipid Employing the repeated discontinuous are localized in the fractions of d sucrose gradient centrifugation, as de- = 1.076 g/cm3. scribed in methods section, 100,000~ pelletable fractions obtained from the disrupted protoplasts of sycamore cells were Glycosyltransferase well separated into two fractions, Golgi and The distribution of enzymes catalyzing ER, as shown in Fig. 2A. Results of the marker enzyme assay clearly indicated the terminal processing of Asn-N-linked that each fraction was free from mutual oligosaccharides, i.e., galactosyltransfercontamination. As presented in Fig. 2B, ase, fucosyltransferase, and xylosyltransIDP-ase activity was localized in a single ferase, was determined by measuring the distinctive peak of Golgi at the 20130% su- incorporation of sugar into either TCA-incrose interface, whereas NADH-cyt-c re- soluble endogenous membranous acceptor ductase was localized in ER at the 10/20’S or GlcNAc. All the enzyme activities measucrose interface (A). Since the latter en- sured were found to reside in the d = 1.127 zyme activity was not sensitive to anti- g/cm3 fractions, but the ER membranes mycin A, it is evident that the mitochon- were depleted of glycosyltransferase acdrial enzyme was not contaminated in the tivities (Fig. 4B). In contrast, glycolipid ER fraction [cf. (19)]. glycosyltransferase activities were totally cu-Mannosidase, a marker enzyme of the absent in the Golgi membrane fractions. Asn-N-linked oligosaccharide processing, Results given in Fig. 4C show the effiwas assayed using PNP-Man as substrate cient incorporation of [14Clgalactose and (27). The enzyme activities were detectable [14Clfucose using GlcNAc as the acceptor in both the ER and Golgi fractions (Figs. (13,18), whereas xylose was not transferred 2A, C). However, the optimum pH was to GlcNAc. clearly distinguishable between ER and Results of Fig. 5 (A, B) show the proGolgi enzymes, 5.0 versus 6.0 respectively, portionalities between the enzyme activias shown in Fig. 3. Also it was found that ties (galactosyltransferase and fucosylswainsonine (50 ng) strongly inhibited the transferase using GlcNAc as acceptor) Golgi mannosidase activities (data not versus reaction time and the amount of enshown). From analogy to studies performed zyme proteins added. The results support using the animals, results appear to indi- the idea that membrane-bound enzymes cate that each enzyme is responsible for are likely involved in glycoprotein biosynthe primary and secondary mannose trim- thesis. The inertness of GlcNAc in the xyming of the core carbohydrate. losyltransferase reaction can be predicted

426

ALI,

MITSUI,

AND

AKAZAWA

310

- 0.

- 0.

-0

- 0.

- 0.

- 0.

-0

4

6

12 Fraction

16

20

24

Number

FIG. 2. Separation profile of ER and Golgi complex. Two fractions separated from the 100,OOOg pellets of disrupted protoplasts, employing the repetitive discontinuous sucrose density gradient centrifugation as explained in the text, are shown in the top frame. Distribution pattern of marker enzyme activities of each of ER and Golgi fractions are given in (A)-(C) (see test). (A) NADH-cytc reductase (@), a-mannosidase (0), and protein (0). (B) IDP-ase (A) and protein (0). (C) a-mannosidase. In (A), IDP-ase activities in ER fractions are shown (A), whereas NADH-cyt-c reductase activities in Golgi fractions are shown in (B) (a).

from the configuration of the xylose conjugation to the oligosaccharide molecule [cf. (21)], but as presented the proportionalities of the [‘4C]xylose incorporation using the endogenous acceptor were observed (Fig. 5). The enzyme reaction exhibiting a broad pH optimum 4.0 to 6.0 was dependent on divalent cations in the order of Mn’+

> Me > CO’+ > Ca2+ (data not shown). The optimum temperature of the enzyme reaction was 37°C. Subcompartmentaticm of Golgi and Enzyme Dist?-ibuticms

We next attempted to determine whether Golgi-associated glycosyltransferase ac-

TRANSGLYC

Na-citrate

SYLASE

LOCALIZATION

Ma-phosphate

0.25 0.20 z 0.15 0.10

zL f 5 E a

~0.05 ‘0 4

5

8

7

427

IN GOLGI

gradients as presented in Fig. 6. Both galactosyltransferase and xylosyltransferase activities, using the endogenous acceptors are distributing between the two peaks (1.118 and 1.127 g/cm3), whereas fucosyltransferase band mostly in the lighter fractions (1.118 g/cm3) with only a marginal leading edge in the denser fractions (Fig. 6C). Essentially the identical distribution profiles were obtained with both galactosyl- and fucosyltransferases using GlcNAc as the sugar acceptor (Fig. 6D). cY-Mannosidasecomigrated as a distinctive peak at a density of 1.127 g/cm3, while the

8

PH

FIG. 3. Activities

vs pH relationship of ER and Golgi mannosidase. Mannosidase activities of ER and Golgi fractions as shown in Fig. 2 (A and C) were measured at various pHs using different buffers as given. For basic assay conditions, see the text.

tivities could be further separated by subfractionation of the Golgi membranes. Employing homogenization of purified Golgi followed by discontinuous sucrose density gradient centrifugation, as described in the text, two protein peaks which banded at densities of 1.118 and 1.127 g/ cm3 were obtained (Fig. 6A). In order to ascertain that the two subfractions did not result from aggregation and/or disaggregation, each fraction was separately collected and subjected to recentrifugation. It was confirmed that the sedimentation of each subfraction was not altered, ruling out the possible occurrence of such phenomena during the centrifugation step (data not shown). As a means to clarify that the two separated Golgi subcompartments are really distinguishable, we employed SDS-PAGE separation of two fractions together with the original whole Golgi fractions, and the Coomassie-stained gels are shown in Fig. 7. Results show that there are common bands between two but several bands were found to be clearly characteristic to each fraction as marked. Three glycosyltransferases activities were shown to distribute differently in the

A

ER

.E 0 5 $ s 1

’

c

Golgibndogenous)

a 2 g 2t 0

4

a

12 Fraction

10

20

24

Number

FIG. 4. (A, B) Distribution patterns of glycolipid and glycoprotein glycosyl transferases in ER and Golgi. Basic experimental conditions for ER and Golgi separations were the same as those described in Fig. 2, and enzyme activities were measured. (A) mannosyltransferase; monosaccharide lipid (0) and oligosaccharide lipid (A). (B) fucosyltransferase (m), galactosyltransferase (0), and xylosyltransferase (V) using the endogenous acceptor. (C) Galactosyltransferase (0) and fucosyltransferase (0) using GlcNAc as the acceptor. For experimental details see the text.

428

ALI, MITSUI, AND AKAZAWA

Takahashi et al. (21) have disclosed that the enzyme molecule is a typical N-linked glycoprotein consisting of oligosaccharides which are xylose-containing biantennary “complex-type” having the common structural core, xylose 61 + 2 (mannose arl --* 6) mannose fil--* 4 GlcNAc /31+ 4 (fucose al --* 3) GlcNAc, and with additional molecules of GlcNAc, galactose, and fucose in the outer chain. 0 5 10 15 20 In the present investigation, we first atTime (mid tempted to clarify the crucial role of the Golgi complex in the glycoprotein formation in the sycamore cells, because in a v o -1.5 6study using castor bean endosperm it was n reported that the ER entails the princi- 1.0 4pal role in the biosynthetic process (11). vw Glycosyltransferases, i.e., galactosyl-, fu2n cosyl-, and xylosyltransferases, were found (mg)(v) o’5 /& / to be present in the Golgi using endogenous 0.1 02 0.3 0.4 ’ 0 0 membranes or exogenous GlcNAc as ac0 80 100 20 40 60 ceptors, whereas the ER fractions conProtein (rg) tained glycolipid glycosyltransferase exclusively. The overall results are consistent FIG. 5. Time dependency (A) and enzyme concenwith the currently accepted mechanisms tration dependency (B) of Golgi glycosyltransglycosylases. (A) Time-dependent transfer of r’CJgalact.ose operating in the glycoprotein biosynthesis and [‘%lfucose to the exogenous GlcNAc and that of and processing based on studies using the [3H]xylose to the endogenous membrane acceptor were animal system (3, 5, 6). In light of the remeasured employing the assay system described in cent observation that sycamore polyphenol the text. (B) The basic experimental conditions were oxidase contains xylose-containing carbothe same as those shown in (A) except for varying hydrate unit (21), the presence of xylothe amount of protein (mug)added to the assay mixture syltransferase activities in the Golgi as shown. membranes is of particular interest. In attempting to further amplify our knowledge of the Asn-N-linked oligosacremaining activity migrated at a density charide processing enzymes localized in the of 1.118 g/cma (Fig. 6B). The enzyme was Golgi membrane system, we examined enriched about fourfold in the denser fracsubcompartmentation of these enzyme actions. IDP-ase sedimented in nearly equal tivities. Upon a gentle mechanical disrupquantities in the lighter and denser regions tion of the Golgi followed by the stepwise of the gradients, sucrose gradient centrifugation, two distinguishable fractions were separated, DISCUSSION among which enzyme activities showed Experimental exploration of the mech- disproportionate distributions. Although CYanisms underlying glycoprotein biosyn- galactosyl- and xylosyltransferase, thesis using the suspension-cultured cells mannosidase, and IDP-ase were shown to of sycamore offers a number of advantages. be distributed between both lighter and denser fractions, fucosyltransferase apOf particular interest is the biosynthesis and secretion of lactase-type polyphenol peared to be predominantly localized in the lighter fractions. It was previously reoxidase, which comprises approximately 2% of the total cellular protein synthesized ported that there are ois and tram com(20, 30). It is a blue copper protein of M, partments in the Golgi of animal origins 97,000 containing 45% carbohydrate, and using the discontinuous sucrose density

TRANSGLYCOSYLASE

5-

LOCALIZATION

429

IN GOLGI

A

J5)5)-

,

I

I I

1

3- B

I I

y mannosidase

)j-

C

s

1 -I

ump

Fuc-transferase

4.5

Xyl-transferase Gal-transferase

2L T

i- 22 )-

5 E

,

D ?y

Fuc-transferase

-f -1

4

12 Fraction

16

24

Number

FIG. 6. Distribution patterns of enzyme activities in the Golgi subfractions. Mechanically disrupted Golgi fractions (see Fig. 2) were subfractionated employing the discontinuous sucrose density gradient according to the procedures described in the text (A), and enzyme activities were measured as shown. (A) Protein (0) and sucrose density (-). (B) a-Mannosidase (0) and IDP-ase (A). (C) Fucosyltransferase (B), galactosyltransferase (0), and xylosyl transferase (V) using the endogenous membranous acceptors. (D) Galactosyltransferase (0) and fucosyltransferase (m) using the exogenous GlcNAc as the acceptor. For the basic experimental details, see the text.

430

ALI,

MITSUI,

AND

AKAZAWA

phenol oxidase, the localization of the transfucosylation reaction in the trans compartment is a likely mechanism. Clearly, the elucidation of the topological orientation of both substrates and enzyme molecules bound to the Golgi membranes is crucial to our studies, and the final answer as to the intrinsic compartmental localization of transglycosylases in the organelle having assymmetric organization 67must wait until both the enzyme molecule and the natural sugar acceptor are characterized. 43The detailed enzymic mechanisms of the surgar trimming during the step of plant glycoprotein biosynthesis is another area 30needing further investigation. In the present investigation, we have not studied the 2014enzymic properties of mannosidases of two different organelles using either the purifled preparations or the high-mannose lighter denser whole type oligosaccharides [(Mann)gGlcNAc fraction fraction Golgi or GlcNAc(Man)s(GlcNAc)z] as subI 1 strate. However, based on the currently FIG. 7. SDS-PAGE (Coomassie-stained bands) of available information on the specific subsubfractionated Golgi membranes. Equal amounts cellular localization of the enzyme, we can (100 pg) of lighter (1.118 g/cm*) and denser (1.127 g/ put forward hypothetical consideration. cm’) Golgi subfractions (see Fig. 6A) and the whole Using rat liver enzyme, both ER and Golgi Golgi fractions (see Fig. 2) were subjected to the SDSPAGE as described in the text. II mannosidases are reported to hydrolyze PNP-Man, whereas Golgi I mannosidase does not (38). On the other hand, ER mangradient centrifugation techniques (5, 22, nosidase is distinguishable from mannos23, 31-33). However, it now appears that idase II, as the substrate of the latter enthe Golgi complex consists of at least three zyme, GlcNAc(Man&( GlcNAc)z is inert to distinct subcompartments, cis, medial, and the former, and also it is insensitive to harts, differing in their structural, histo- swainsonine (38). Furthermore, both rat chemical, enzymatic, and functional prop- liver lysosomal mannosidase (39) and waerties (6,24,34,35). There is no report re- ter-soluble plant enzyme (27) are known to garding the spatial separation of glycopro- have the acidic optimum pH. Above all, it is likely that the mannosidase studied here tein glycosyltransferase activities within contains at least Golgi-specific mannosithe plant Golgi complex, although multisite localization of nucleoside diphosphatase or dase II activities; but a more thorough IDP-ase has been reported in both plant characterization of the enzyme molecule is and animal cells (36, 37). It is difficult to needed. Recently the purification of plant compare the separation profiles of the sy- a-mannosidase has been reported by Forcamore Golgi subcompartments with those see (40) and Szumilo et aZ.(41), both as mireported in the animal system, but the crosomal enzymes isolated from mung lighter compartment (d = 1.118 g/cm3) in bean hypocotyls. Finally, another most important queswhich fucosyltransferase was found to be predominantly localized may well corre- tion remaining to be answered by addispond to the tram Golgi. From the struc- tional work is whether the sugar moiety ture of the carbohydrate moiety of poly- (xylose, galactose, or fucose) incorporation Mr(kDa:

TRANSGLYCOSYLASE

LOCALIZATION

examined in the present study truly reflects the formation of the carbohydrate linkage as characterized in the lactase molecule [cf. (21)]. Extracellular secretion of the cell wall polysaccharides in sycamore cells has been studied (42), and the structural characterization of xyloglucan has been established. In order to obtain further insight into the crucial role of transglycosylase(s) in glycoprotein biosynthesis, an immunochemical approach employing specific antisera raised against the purified lactase molecule is in progress.

19. 20. 21.

22.

23. 24.

REFERENCES 25. 1. PALADE, G. E. (1975) S&n.% 189,347-358. 2. BLOBEL, G., AND DOBBERSTEIN, B. (1975) J. CeU Bill 67,835851. 3. LENNARZ, W. J. (1980) in The Biochemistry of Glycoproteins and Proteoglycans (Lennam, W. J., ed.), pp. 35-83, Plenum, New York. 4. HUBBARD, S. C., AND IVAN, R. J. (1981) Anna Rev. Biochem 50,555-583. 5. FARQUHAR, M. G., ANDPALADE, G. E. (1981) J. Cell Biol 61, ‘7%1035. 6. KORNFELD, R., AND KORNFELD, S. (1985) Annu Rev. B&&em. 54,631-664. 7. ROBINSON, D. G. (1985) in Plant Membranes, pp. 51-79, Wiley, New York. 8. ELBEIN, A. D. (1979) Annu. Rev. Plant PhysioL 30, 239-272. 9. MELLOR, R. B., ROBERTS, L. M., AND LORD, J. M. (1979) Biochem J 182,629-631. 10. NAGAHASHI, J., AND BEEVERS, L. (1968) Plant PhysioL 61,451-459. 11. CONDER, M. J., AND LORD, J. M. (1983) Plant PhysioL 72,547-552. 12. GARDINER, M., AND CHRISPEELS, M. J. (1979) Plant PhysioL 55,536-541. 13. POWELL, J. T., AND BREW, K. (1974) Biochem J. 142,203-209. 14. STURM, A., AND KINDL, H. (1983) FEBS L&t. 160, 165-169. 15. VITALE, A., AND CHRISPEELS, M. J. (1983) J. CeU Bid 99.133-140. 16. HORI, H., AND ELBEIN, A. D. (1983) Arch, Bio&m Biophys. 220,133-140. 17. MITSUI, T., AKAZAWA, A., CHRISTELLER, J. T., AND TARTAKOFF, A. (1985) Arch. Biochmn. Biophys. 241,315-328. 18. ALI, M. S., NISHIMURA, M., MITSUI, T., AKAZAWA,

26. 27.

28. 29. 30. 31. 32. 33.

34. 35. 36. 37. 38. 39. 40. 41. 42.

IN GOLGI

431

T., AND KOJIMA, K. (1985) Plant Cell PhysioL 26, 1119-1133. ALI, M. S., AND AKAZAWA, T. (1986) Plant PhysioL 81.222-227. BLIGNY, R., ANDDOUCE, R. (1983) Biochem J. 209, 489-496. TAKAHASHI, N., HOTTA, T., ISHIHARA, H., MORI, M., TEJIMA, S., BLIGNY, R., AKAZAWA, T., ENDO, S., AND ARATA, Y. (1986) Biochemistry 25,388395. DLJNPHY, W. G., FRIES, E., ARBANI, L. F., AND R~THDUN, J. E. (1981) Proc Nat1 Ad sci USA 78.7453-7457. DUNPHY, W. G., AND ROTHMAN, J. E. (1983) J. CeU BioL 97,270-275. DUNPHY, W. G., AND ROTHMAN, J. E. (1985) Cell 42,13-21. MELLOR, R. B., ANDLORD, J. M. (1979) Planta 146, 91-99. LORD, J. M., KAGAWA, T., MORRB, T. S., AND BEEVERS, H. (1973) J. CeU Biol 57,659-667. AGRAWAL, K. M. L., AND BAHL, 0. P. (1972) in Methods in Enzymology (Ginsburg, V. ed.), Vol. 28, pp.720-728, Academic Press, New York. LAEMMLI, U. K. (1979) Nature (London) 227,680685. FORSEE, V. T., VALKOVICH, G., AND ELBEIN, A. D. (1976) Arch Biochem. Biophys. 174,469-479. AKAZAWA, T., AND HARA-NISHIMURA, I. (1985) Anna Rex Plant Physid 36,441-472. ROTHMAN, J. E. (1981) Science 213,1212-1219. GOLDBERG, D. E., AND KORNFELD, S. (1983) J. BioL Chem 258,3159-3165. DEUTSCHER, S. L., CREEK, K. E., MERION, M., AND HIRSCHBERG, C. B. (1983) Proc. NatL Acad Sci USA 80,3938-3942. DUNPHY, W. G., BRANDS, R., AND ROTHMAN, J. E. (1985) CeU 40.463-472. ROTHMAN, J. E. (1985) Sci Amer. 253,84-95. TULKENS, P., BEAUFAY, H., ANDTROUET, A. (1974) J. Cell BioL 63,383-401. BOWELS, D. J., AND KAUSS, H. (1976) Biochim Bie phys. Acta 443.360-374. BISHOFF, J., AND KORNFELD, S. (1983) J. Bid Chem 258,7907-7910. OPHEIM, D. J., AND TOUSTER, 0. (1978) J. BioL Chem 253,1017-1023. FORSEE, W. T. (1986) Arch. Biochem, Biophys. 242, 48-57. SZUMILO, T., KANSHAL, G. P., HORI, H., AND ELBEIN, A. D. (1986) Plant PhysioL 81,383-389. STEVENSON,T. T., MCNEIL, M., DARVILL, A. G., AND ALBERSHEIM, P. (1986) Plant PhysioL 80,10121019.