Jun 25, 2016 - yeast enzymes (5 mg of protein) in a total volume of 1 ml for 20 min .... were pooled and evaporated to dryness at 30 "C under reduced pres- sure. ... and a hexose 1-phosphate with a chemical shift of 5.123 parts/million.
THEJOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 268,No. IS,Issue of June 25, pp. 13110-13117, 1993 Printed in lJ S.A.
0 1993 by The American Society for Biochemistry and Moleculat Biology, Inc
Mannosylphosphoryldolichol-mediatedReactions in OligosaccharideP-P-Dolichol Biosynthesis RECOGNITION OF THE SATURATED0-ISOPRENEUNIT MANNOSYLTRANSFERASES*
OF THE MANNOSYL DONOR BY PIG BRAIN
(Received for publication, January 21, 1993)
Jeffrey S. Rush, Judith G . Shelling, Nathan S. Zingg, Paul H.Ray$, and CharlesJ. Waechterj From the Department of Biochemistry, University of Kentucky College of Medicine, Lexington, Kentucky 40536 and the $Division of Molecular Genetics and Microbiology, Burroughs Wellcome Co., Research Triangle Park, North Carolina 27709
The specificity of Man-P-Dol:Mans_,GlcNAc2-P-P-P-The biosynthesis of N-linked oligosaccharides in eukarDol (Oligo-P-P-Dol) mannosyltransferase activity in yotes is initiated in the rough endoplasmic reticulum (RER)l pig brain was investigatedby comparing a variety of as the dolichyl pyrophosphate-bound precursor chain, mannosylphosphorylisprenols as mannosyldonors. Glc3MansGlcNAc2 (for reviews, see Refs. 1-3). In an early For thiscomparison the b-Man-P-isoprenols were syn- stage of the assembly process, the first five mannosyl units thesized usinga partially purified preparationof man- are derived directly from GDP-Man on the cytoplasmic face nosylphosphorylundecaprenol (Man-P-Undec) syn- of the RER. In the next stage of biosynthesis four more thase from Micrococcus luteus. The bacterial mannosyltransferaseefficientlycatalyzedthetransfer of mannose units areadded to ManSGlcNAc2-P-P-Dolwith manmannose from GDP-[’HIMan to a series of defined nosylphosphoryldolichol (Man-P-Dol) serving as the “activated” mannosyl donor in the lumenal compartment of the isoprenyl monophosphate substrates. Two a-Man-Pdolichols were synthesized chemically and also exam- RER (Fig. 1).The function of Man-P-Dol in this biosynthetic ined as substrates. When exogenous @-[3H]Man-P-Dole6 process was first established by demonstrating that the purias a mannosyl donor for was tested as a substrate forMan-P-Do1:Oligo-P-P-Dol fied mannolipid could act mannosyltransferase activity in pig brainmicrosomes, Man9GlcNAc2-P-P-Dolsynthesis when added exogenously to [‘Hlmannose was actively transferred to endogenous microsomal fractions from mammalian tissues and yeast (4, Oligo-P-P-Dol acceptors. The major enzymatically la- 5). The observation that mutants unable to synthesize Manlacking four spebeled product wasManeGlcNAcz-P-P-Dol. Under iden- P-Dol formed Glc3Man6GlcNAc2-P-P-Dol, tical conditions 8-[’H]mannosylphosphorylpolyprenol cific mannosyl residues, corroborated the role of the lipophilic (Man-P-Polysa) was an extremely poor substrate, in- mannosyl donor in the biosynthetic pathway (6, 7). dicating that the saturated a-isoprene unit of the doliMan-P-Dol is also the direct mannosyl donor for the manchyl moiety is critical for recognitionof the lipophilic nosylation of some serine residues in yeast glycoproteins (for mannosyl donor by the endoplasmic reticulum-associ- review, see Ref. 8) and in the biosynthesis of glycosylphosatedmannosyltransferase(s). When Man-P-dolichols phatidylinositol anchors (9). All of these mannosyl transfer containing 2, 11, or 19 isoprene units werecompared, reactions occur, presumably, after Man-P-Dol diffuses transthe initial rates for the mannosyl transfer reactions versely from the cytoplasmic leaflet of the RER, where it is and the affinity of the enzyme(@ for the mannophos- formed, to the lumenal monolayer (10, 11). pholipid substrate increased with the length and hyRecently, the mannosyltransferase responsible for the syndrophobicity of the polyisoprenol chain. The anomeric thesis of Man-P-Dol in yeast has been cloned, and it appears configuration of the mannosyl moiety is apparently to contain an isoprenoid recognition domain (12). Thus, other essential because the brain mannosyltransferases ex- enzymes involved in lipid intermediateand dolichyl phoshibited a strong preference @-Man-P-dolichols for over may haverecognition sites for specific the corresponding chemically synthesized a-stereoiso- phate (Dol-P) synthesis mers. These results:1) describe a simple two-step pro- structural features of the polyisoprenol. Pertinent studies on cedure for obtaining a partially purified preparation a Chinese hamster ovary cell mutant, apparently defective in of Man-P-Undec synthase that efficiently synthesizes the reduction of the terminal isoprene unit of dolichol, have a variety of 8-Man-P-isoprenols; 2) indicate that pig provided indirect evidence that thepresence of a saturatedabrain Man-P-Do1:Oligo-P-P-Dol mannosyltransferase isoprene unit in Man-P-Dol may influence itsinteraction with Man-P-Dol:Man5_8GlcNAc2-P-P-Dol (Oligo-P-P-Dol) activity is relatively specific for lipophilic mannosyl donors containing 19 isoprene units with aD-Man 1-P mannosyltransferase (13, 14). Also consistent with the funcgroup attached to the saturated a-isoprene of unit dol- tional significance of the presence of a reduced terminal ichol; and 3)emphasize the importanceof the reduction isoprene unit in dolichol, fully unsaturated polyprenyl monof the a-isoprene unit in the biosynthesis and function of Dol-P in mammalian cells. The abbreviations used are: RER, rough endoplasmic reticulum;
* This work was supported by National Institutes of Health Grant GM 36065 (to C. J. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 3 To whom correspondence should be addressed.
Man-P-Dol., mannosylphosphoryldolichol ( n =the number of carbon atoms in the polyisoprenol chain); Oligo-P-P-Dol, Mans-sGlcNAczP-P-dolichol; Dol-P, dolichyl phosphate; Poly-P, fully unsaturated polyprenyl phosphate; CHAPS, (3-[(3-~holamididopropyl)dirnethy1ammonio)-1-propanesulfonate;PIPES,piperazine-N-N-bis-2ethanesulfonic acid; Endo H, endo-/3-glucosaminidaseH; Man-6-P, mannose 6-phosphate; Man-1-P, mannose 1-phosphate; Man-P-Undec synthase, GDP-MaxUndecaprenyl phosphate mannosyltransferase; ER, endoplasmic reticulum; cpm, counts/minute.
13110
Man-P-Dol-mediated Mannosyltransferases
i n Pig Brain
13111
(Danvers, MA). Undecaprenyl phosphate, dolichyl phosphate (c96)1 and polyprenyl phosphate (c96)were acquired from Dr. Tadeusz Chojnacki, Warsaw, Poland. Endo-N-acetylglucosaminidaseH (Endo H) was purchased from Boehringer Mannheim. Complete Counting Cocktail 3a70B is a productof Research Products International COW. (Mount Prospect, IL). All other chemicals and reagents were obtained from standard commercial sources. Enzymatic Synthesis and Purification of GDP-12-3HlManmsePP- G I c N A c ~ M A ~The labeled sugar nucleotide was prepared enzymatically by a modification of the procedure of Braell et al. (25). The conversion of 123H]Man-6-P, synthesized enzymatically (26), to the sugar nucleotide was catalyzed by a partially purified fraction containing phosphomannomutase and GDP-Man pyrophosphorylase obtained from baker's yeast. The yeast cytosolic fraction was applied to a TSK-GEL Toyopearl DEAE-650M column (10 ml) equilibrated with 10 mM Tris-HC1, pH 7.2. After washing the column with four column volumes of the equilibration buffer, the yeast enzymes were eluted with a linear gradient (40 ml) of 0-1 M NaC1. Each fraction was assayed for phosphomannomutase and GDP-Man pyrophosphorylase. Phosphomannomutase activity was assayed as described by Glaser (27) except that theformation of Man 1-Pwas measured using the Bartlett procedure (28) omitting digestion with perchloric acid. GDP-Man pyrophosphorylase was assayed as described by Preiss (29) except that GDP-[3H]Man was used as substrate. [3H]Man-l-Pformed in CYTOSOL LUMEN the presence of pyrophosphate was separated from the labeled sugar nucleotide by sedimentation following incubation with Norit A-activated charcoal. The fractions containingthe two yeast enzymes were pooled and concentrated (2-3 ml) with an Amicon Ultrafiltration cell using a YM-10 membrane filter. After dialysis, the labeled sugar RER nucleotide was synthesized by incubating [2-3H]mannose 6-P (5-10 FIG.1. Four mannosyl transfer reactionsinvolving Man-P- mCi) with 50 mM Tris-HC1, pH 7.8, 10 mMMgC12, 10 mM GTP, Dol on the lumenal surface of the RER following the trans- snake venom pyrophosphatase (Type II), and the partially purified membrane movement (dotted arrow) of the lipophilic man- yeast enzymes (5 mg of protein) in a total volume of 1 ml for 20 min nosy1 donor and the heptasaccharyl-lipid intermediatefrom at 37 "C. The reaction was terminated by thermal inactivation in a the site of synthesis on the cytoplasmicleaflet. boiling water bath, anddenatured protein was sedimented by centrifugation. The enzymatically synthesized sugar nucleotide was purified by ophosphates (Poly-P) were found to be poor substrates relasequential ion-exchange and concanavalin A-Sepharose chromatogtive to Dol-P for Man-P-Dol synthase in liver microsomes raphy. The labeled sugar nucleotide was applied to a DEAE-cellulose (15-17). Dol-P is also preferred over Poly-P as a substratefor column equilibrated and then washed with four column volumes of GlcNAc 1-P transferase in Chinese hamster ovary cells (18). 10 mM Tris-HC1, pH 7.2. GDP-[2-3H]Man was eluted with a linear More recently, the mannosyltransferase which converts gradient of 0-0.5 M NH4HC03. The fractions containing GDP-[2-3H] MansGlcNAc2-P-P-Dol to Man6GlcNAc2-P-P-Dol has been Man were pooled and concentrated and then applied to a concanahighly purified from pig aorta (19). Although the role of Man- valin A-Sepharose column equilibrated and eluted with 0.01 M ammonium phosphate, pH 4.0, containing 1 mM MgClz/CaClz/MnC12. P-Dol in the synthesis of Mam_9GlcNAc2-P-P-Do1interme- Under these conditions GDP-[2-3H]Man is retarded and clearly rediates is well documented, the specificity of the mannosyl- solved from Man-6-P, free mannose, and GTP which elute directly transferases has notbeen thoroughly characterized. through the column. The labeled sugar nucleotide was then desalted In this paper, we describe a simple two-step procedure for by gel filtration on a column (40 ml) of Sephadex G-10 eluted with the solubilization and partial purification of GDP- H20. The fractions containing GDP-[2-3H]Man were concentrated Man:undecaprenyl phosphate mannosyltransferase (Man-P- and stored at -20 "C. Solubilization and Partial Purification of Man-P-Undec Synthase Undec synthase) from Micrococcus luteus membranes (20,21). luteus was grown overnight in Luria from M. luteus Membranes-". The micrococcal enzyme efficiently transfers mannose from broth at 37 "C. The cells were harvested by tangential flow diafiltraGDP-Man to a variety of defined polyisoprenyl phosphate tion. Protoplasts were formed in 20 mM sodium phosphate, pH 7.2 substrates. Several P-Man-P-isoprenols, prepared enzymati- (buffer A), plus 1 M sucrose, by the addition of lysozyme (25 rg/ml) cally with the bacterial mannosyltransferase, and a-Man-P- for 30 min. Protoplasts were lysed by a 10-fold dilution in buffer A dolichols, synthesized chemically (22, 231, have been tested in the presence of 10 mM MgSOl and DNase/RNase added at 10 mg/ as substrates toexamine the specificity of Man-P-Do1:Oligo- ml. A crude membrane fraction was sedimented by centrifugation at 100,000 X g for 30 min. The membranes were washed twice by P-P-Dol mannosyltransferase activity in heavy microsomes resuspension in buffer A and resedimentation. from pig brain. The implications of the results for the biosynCrude membrane fractions (2 mg/ml) were suspended in 0.1 M thesis and function of Dol-P are discussed. Some of these Tris-HC1, pH 7.4,lO mM 2-mercaptoethanol, 1 mM EDTA, leupeptin (5 wg/ml), 1 p M phenylmethylsulfonyl fluoride, 0.1 WMpepstatin, and results have been presented in preliminary form (24). 0.5% CHAPS and stirred on ice for 1 h. The CHAPS extract was separated from insoluble material by centrifugation (lo0,OOO X g, 10 EXPERIMENTAL PROCEDURES min) in aBeckman TL-100 ultracentrifuge. The CHAPS extract was Materi~k-[2-~H]Mannose(15 Ci/mmol) was obtained from Amer- then loaded on a TSK-GEL Toyopearl DEAE-650 M column (20 ml) ican Radiolabeled Chemicals (St. Louis, MO). (S)-Citronellol, equilibrated with Tris-HC1, pH 8.0, containing 0.1% CHAPS. After CHAPS, DEAE-cellulose, baker's yeast hexokinase (Type C-301), washing the column with four volumes of the equilibration buffer, bovine pancreatic DNase I, bovine pancreatic ribonuclease, snake Man-P-Undec synthase was eluted with a linear gradient (60 ml) of venom pyrophosphatase (Type II), a-mannosidase (jack bean), chem- 0-1 M NaC1. Fractions (3.5 ml)were collected, and Man-P-Undec ically hydrogenated dolichol (C&), concanavalin A-Sepharose and synthase activity was assayed in the presence of exogenous Undec-P synthetic GDP-Man (Type 111) were obtained from Sigma. AG 1-X8 essentially as described for Man-P-Dol synthase (30). The partially is a product of Bio-Rad. TSK-Gel Toyopearl DEAE-650M was ob- purified mannosyltransferase was stored at -20 "C until used for tained from Supelco Separation Technologies (Bellefonte, PA). Cel- Man-P-isoprenol synthesis. These preparations were stable for at lulose chromagrams are aproduct of Eastman Kodak Co. (Rochester, least 1year. NY). Phosphorus trichlorideoxide was obtained from Alfa Chemicals Chemicat Synthesis and Purification of Dol-P and Citronellyl Mon-
w m-
13112
Man-P-Dol-mediated Mannosyltransferases in Pig Brain
ophosphate-Citronellol and dolichol (CS5)were phosphorylated using phosphorus trichloride oxide and triethylamine inhexane as described by Danilov and Chojnacki (31). The final residue containing citronellyl phosphate was dissolved in CH30H and applied to a 10-ml column ofAG 1-X8 (acetate form). After washing with 5 column volumes of CH30H, the AG 1-X8 column was equilibrated with H 2 0 and the prenyl monophosphate was eluted with a linear gradient (30 ml) of NH4HC03 (0-1 M). Fractions containingcitronellyl phosphate were pooled and evaporated by rotary evaporation under reduced pressure at 35"C. Citronellyl monophosphate was separated from NHIHCOI by chromatography on a 100-ml column of Bio Gel P-2 eluted with H2O. Final residues containing Dol-P were dissolved in 5 mlof CHCI3/CH30H (2:1), and water-soluble contaminants were removed by the partition method of Folch et al. (32). The organic (lower) phase was concentrated and applied as a narrow band to a sheet of EDTA-treated SG-81 paper (33). Following development with CHC13/CH30H/Hz0 (65:35:6), the zone corresponding to Dol-P was located by exposure to iodine vapors and eluted with CHC13/ CHIOH/H~O (10103). Enzymatic Synthesis of P-Man-P-isoprenok-Man-P-polyisopren01s and Man-P-Dollo were synthesized enzymatically using the partially purified preparation of Man-P-Undec synthase from M. luteus (see above). Reaction mixtures contained partially purified Man-PUndec synthase (25 pg of protein), 50 mM Tris-HCI, pH 7.4, 2 mM citronellyl monophosphate, 0.5 mM GDP-[2-3H]Man, prepared enzymatically as described above (500-2500 cpm/pmol), 10 mM MgCl,, 0.1% CHAPS, 5 @g/mlleupeptin, 1p M phenylmethylsulfonyl fluoride, 0.1 p M pepstatin, and 1 mM phosphatidylglycerol in a total volume of0.5ml. During incubation a t room temperature, aliquots of the synthetic reaction mixtures were removed and analyzed for the formation of Man-P-Dollo by thin layer chromatography on cellulose chromagrams developed in ethylacetate/butanol/acetic acid/H20 (4:3:2.5:4). Radioactlve products were detected with a Bioscan Imaging Scanner System 200-IBM, scraped into scintillation counting vials, and quantitated by liquid scintillation spectrometry in 1.0 ml of 1%SDS and10 ml of Complete Counting Cocktail 3a70B. For the synthesis of Man-P-dolichols and Man-P-polyprenols, the reaction mixtures were identical except 0.2 mM Dol-P or Poly-P and 70 p M GDP-Man were included. The progress of Man-P-Dol and Man-PPoly syntheses was monitored by the procedure used for Man-P-Dol synthase (30). When approximately 95% of the GDP-13H]Man (Rf =0.2) had been converted to mannolipid (R,= 0.81, the reactions were terminated by the addition of40 volumes ofCH3OH (for Man-PDollo synthesis) or CHC13/CH30H (2:l) (for Man-P-polyisoprenol synthesis). Purification of &Man-P-isoprenok-Reactions for the enzymatic synthesis of P-Man-P-Dol,o were terminated by the addition of CH30H (10 volumes) and centrifuged to sediment-precipitated protein. The supernatants were applied to an 8-ml column ofAG 1-X8 (acetate form) equilibrated with CH30H. The ion-exchange column was eluted with 5 column volumes of CH3OH and then equilibrated with H,O using a linear gradient (20 ml) of CH3OH/H20 (0-100%). 0Man-P-Dollo was then eluted with alinear gradient (40 ml) of NH4HC03 (0-1 M ) in H20. The fractions containing Man-P-Doll0 werepooled and reduced to a small volume(0.5-1 ml) by rotary evaporation at 30"C under reduced pressure. The samples were desalted by chromatography on a column (100 ml) of Bio-Gel P-2 eluted with H20.&Man-P-Doll0 was separated from unreacted citronellyl phosphate by ion-exchange chromatography on a10-ml column of DE52-cellulose equilibrated with H20. @-Man-P-Dollowas eluted with a linear gradient (40 ml) of 0-1 M NRHCOI and thendesalted by gel filtration on Bio-Gel P-2. Fractions containing Man-P-Dolt0 were pooled and evaporated to dryness at 30 "C under reduced pressure. Enzymatically synthesized Man-P-Dollo was finally dissolved in Hz0 and stored at -20 "C. 0-Man-P-polyisoprenols were purified as described previously for 0-Man-P-Dol (30). Chemical Synthesis and Purification of a-Man-P-Dolichols-The a-stereoisomers of Man-P-Dolg5and Man-P-Doll0 were synthesized basically as described by Warren and Jeanloz (22, 23).[1,2,3,4,6] Pentaacetyl-mannose was prepared by peracetylation of mannose using freshly redistilled pyridine (0.2 ml) and acetic anhydride (0.2 ml) at room temperature overnight. [2,3,4,6]Tetraacetyl-Man-l-P was prepared by stirring [1,2,3,4,6]pentaacetyl-mannose and crystalline phosphoric acid under reduced pressure at 65 "C. After 2 h, the reaction solution was diluted with 10 ml of Hz0 and extracted with CHCl3(3 ml) four times. The aqueous fraction was layered onto a 10ml column of AGl-X8 (acetate form) and the column eluted with approximately 5 column volumes of H20. [2,3,4,6]Tetraacetyl-Man-
1-Pwas then eluted with a linear gradient of NH,HCO, (0-1 M). The fractions containing acetylated Man-1-P were pooled, concentrated to 1 ml by rotary evaporation under reduced pressure at 30 "C, and desalted by gel filtration on a 100-ml column of Bio-Gel P-2 eluted was condensed with dolichol with HzO. [2,3,4,6]Tetraacetyl-Man-l-P or citronellol by stirring at room temperature for 72 h in anhydrous pyridine using dicyclohexylcarbodiimide (5 mg) as catalyst. The reaction mixture was diluted with 20 volumes of CHC13 and applied to a 5-ml column of silica gel. The silica gel column was eluted sequentially with CHC13/CHaOH in the proportions (50:1),(20:1),(10:1), were eluted quanand (5:l). [2,3,4,6]Tetraacetyl-a-Man-P-isoprenols titativelyin the CHC13/CH30H (5:l) fraction. The product was deacetylated by mild alkaline methanolysis (34). Man-P-DolI0 was purified by ion-exchange chromatography on DEAE-cellulose (acetate form) and desalted by gel filtration on Bio-Gel P-2 eluted with Hz0 as described previously for p-Man-P-Doll0. a-Man-P-Dolpewas purified by DEAE-cellulose chromatography as described for @-ManP-Dol (30). Characterization of a/P-Man-P-Dollo-The enzymatic product was identified as P-Man-P-Dollo based on its chemical and chromatographic properties. First, the enzymatic product migrated as a single compound on cellulose chromagrams developed with ethyl acetate/ butanol/acetic acid/HZO(4:3:2.5:4) and on Silica Gel G thin layer plates or EDTA-treated SG-81 paper developed with CHCb/CH,OH/ H20 (65:35:6). When sprayed with an anisaldehyde reagent (35), the product stained a blue color identical to a citronellol standard. The product was also detectable with a phospholipid spray reagent (36). The compound yielded mannose and citronellyl phosphate following mild acid hydrolysis (0.1 M HCl, 50 "C, 60 min), consistent with the proposed structure. Free citronellol was liberated by sequential treatment with mild acid and E. coli alkaline phosphatase. The chemically and enzymatically synthesized compounds contained equimolar amounts of phosphate (28) and mannose. Mannose was identified and quantitated by high performance anion-exchange chromatography at alkaline pH using a Dionex Bio LC Chromatography System ~ at a equipped with an HIPC-ASG column eluted with 150 r n NaOH flow rate of 0.6 ml/min. 'H NMR analysis, performed as described below, yielded a spectrum consistent with the presence of citronellol and a hexose 1-phosphate with a chemical shift of 5.123 parts/million for the anomeric proton, relative to the acetone internal standard, and with coupling constants J H I . H ~ = 1.085 and J~~.p=8.225.These values are in agreement with those reported by O'Connor et al. (37) for a P-Man-1-P. a-Man-P-Dollo, synthesized chemically, was characterized based on the criteria described above for P-Man-P-Dollo.'H NMR analysis yielded a chemical shift of 5.4 parts/million for the anomeric proton, relative to the acetone internal standard, with coupling constants J H ~ flZ = 1.804 and J H ~ =. 7.93. ~ These values are consistent with the presence of an a-Man-1-P(37). These preparations contained a small (5-7%) amount of the p-isomer. Characterization of Enzymatically Synthesized P-Man-P-Polyisoprenok-The products synthesized enzymatically by the Man-P-Und synthase preparations from M. luteus contained equimolar amounts of mannose and phosphate and were chromatographically identical to [3H]Man-P-Dol,prepared by metabolic labeling of B lymphocytes (38), when analyzed on Silica Gel G thin layer plates in CHCl,/ CH30H/H20 (65:35:4) and CHC~/CH~OH/HZO/NH,OH (65:354:4). Mild acid hydrolysis with 0.1 N HCl in 50% n-propyl alcohol at 50 "C for 1 h liberated mannose and Dol-P from Man-P-Dol, but mannose and free phosphate were released from Man-P-Poly, consistent with the properties of Man-P-Dol and Man-P-Poly, respectively (39,401. Preparation of Pig Brain Microsomes-Microsomes were prepared from porcine brain gray matter essentially as described previously (41), resuspended (20-40 mg protein/ml) in 0.1 M Tris-HC1, pH 7.4, 0.25 M sucrose, 10 mM 2-mercaptoethanol, and 1 mM EDTA and stored at -20 "C until used for enzyme assays. Assay for Man-P-Do1:Oligo-P-P-Dol Mannosyltransferase Activity in Pig Brain Microsomes-The transfer of radiolabeled mannosyl groups from the various labeled Man-P-polyisoprenols to endogenous Oligo-P-P-Dol acceptors was assayed essentially as described by Harford and Waechter (42).Typical enzymatic reactions consisted of pig brain microsomes (0.5 mg of membrane protein), 2 mM CaCL 1.0% Triton X-100, 50 mM Tris-HCI, pH 7.4, and the indicated concentration of mannolipid substrate dispersed ultrasonically in 1.0% Triton X-100. Enzymatic reactions utilizing Man-P-Doll0 as mannosyl donor were terminated after10 min at 37 'C by the addition of 40 volumes of CH,OH and denatured protein was sedimented by centrifugation. The CH30H-insoluble material was re-extracted three
13113
Man-P-Dol-mediated Mannosyltransferases in Pig Brain times with 2 ml of CHIOH containing 50 p M CaC12. The insoluble residue was then sequentially extracted with 2 ml of CHCldCH3OH (2:1), CHIOH, CH30H/0.S% NaCl (1:1), and CH30H/H20 (1:l). Oligo-P-P-Dol was then extracted from the delipidated residue with CHC&/CH30H/H20(1010:3), transferred to a scintillation vial, and the solvents evaporated in a warm water bath under a stream of air. The amount of [3H]mannose transferred to endogenous 0ligo-P-PDol was determined by scintillation spectrometry following the addition of 0.5 ml of 1%SDS and 4 ml of 3a70B Complete Counting Cocktail. Churacterization of Dolichyl-bound Oligosaccharides&Zed in Vitro by Pig Brain Microsomes-Enzymatically labeled Oligo-P-P-Dol was extracted with CHCl&H3OH/H20 (10103), and the labeled oligosaccharides were released from the lipid carrier by treatment with tetrahydrofuran, 0.5 N HC1 (41) at 50 "C for 30 min (30). The size of the labeled oligosaccharides was estimated before and after treatment with Endo H (5 milliunits, 0.1 M sodium acetate, pH 5.0, 18 h, 25 "C) by gel filtration on a column of Bio-Gel P-4 (1.5 X 40 cm, 200-400 mesh) eluted with 0.1 N acetic acid. The anomeric configuration of the mannosidic linkages was established by demonstrating that all radioactivity was released as [3H]mannose after treatment with jack bean a-mannosidase (10 milliunits) in 0.1 M sodium acetate, pH 4.5, for 4 h at 25 "C. Assay for Man-P-Undec:Mannan Manmsyltransferase Activity in M.luteus Membranes-Enzymatic reactions contained 0.1% Triton X-100, 10 mM MnC12, 40 mM PIPES, pH 6.8, 13 p M [2-3H]Man-Ppolyisoprenol (250-350 cpm/pmol), and M. luteus membranes (0.22 mg of membrane protein) in a total volume of0.02 ml. Following incubation at room temperature for 4 min, the reaction was terminated by the addition of 3 ml of CH30H. M. luteus membrane suspension (1 mgof membrane protein) was then added, and the membrane residue was sedimented. The insoluble residue was washed with CHsOH (3 ml) three times to extract the unincorporated [3H] mannolipid and dissolved in 0.5 ml of 1%SDS. The amount of of [23H]mannose incorporated into bacterial mannan was determined by scintillation spectrometry after the addition of 4 ml of Complete Counting Cocktail 3a70B. Following solubilization of the final membrane residue with 1%SDS, the enzymatic product was excluded by gel filtration on a column of Sephadex G-50 eluted with 10 mM TrisHCl, pH 8.0, 0.1 M NaCl and 0.1% SDS. The product which eluted in the void volume of Sephadex G-50 was stable to mild-acid treatment under conditions which completely hydrolyze Man-P-Undec. However, a substantial amount of radioactivity in the solubilized product was released as mannose by enzymatic hydrolysis with jack bean a-mannosidase. Analytical Methods-Protein was estimated by the method of Rodriguez-Vico et al. (431, and total lipid-phosphorus was measured by the method of Bartlett (28). 'H NMR Analysis-"H NMR spectra were obtained on the citronellyl derivatives dissolved in DzO using a Varian VXR-500s NMR spectrometer operating in the Fourier mode and equipped with quadrature detection. The samples were temperature equilibrated at 25 "C for 15 min prior to acquisition. The spectra were acquired using 19,000 data points, and a sweep width of 3,247. The solvent signal was suppressed by inversion recovery. Chemical shift values were relative to the major resonance of acetone which was used as the internal 2.224 reference.
micrococcal membranes with 0.5% CHAPS. Although the mannosyltransferase activity could also be effectively solubilized by extraction with Triton X-100, Tween 20, Brij 58, or Nonidet P-40 at thesame concentration, the highest specific activity of Man-P-Undec synthase activity was recovered by extraction with CHAPS. Man-P-Undec synthase was resolved from other mannosyltransferasesand endogenous acceptor lipids by ion-exchange chromatography of the CHAPS-solubilized enzyme on TSK-Gel Toyopearl DEAE-650 M. As seen in Fig. 2, the partially purified preparations of Man-P-Undec synthaseactivity were dependent on the addition of an exogenous polyprenyl phosphatefor activity. The formation of Man-P-Undec was linear with respect to the time of incubation for at least 2 h with very little, if any, turnover of the mannolipid. These mannosyltransferase preparations did not mannosylate diacylglycerols or catalyze the transfer of mannose from ManP-Undec to mannan,presumably due to depletion of endogenous diacylglycerols and the mannan acceptor and/or the pertinent mannosyltransferases. These soluble enzyme fractions efficiently catalyzed the synthesis of the various B-ManP-isoprenyl substrates used in the specificity study described below. Specificity of Man-P-Do1:Oligo-P-P-Dol Mannosyltransferme Activity in PigBrain Microsomes-To investigate the importance of saturation of the a-isoprene unit of the mannosy1 donor, Man-P-dolichols and Man-P-polyprenols containing 19 ( G 5 ) or 11 (C55) isoprene units were testedas substrates for Man-P-DolOligo-P-P-Dol mannosyltransferase in microsomes from pig brain. As seen in Fig. 3, labeled mannosyl residues were actively transferred from Man-PDo195 (panel A, 0 ) and Man-P-Dols5(panel B, 0 )to endogenous Oligo-P-P-Dol acceptors. In sharp contrast to result, this the mannosylation rates observed with Man-P-Polyg5(panel A, 0) and Man-P-PolyB5 (panel B, 0)were only slightly above background levels even though the polyprenyl moiety differs
RESULTS
Solubilization and Partial Purification of Man-P- Undec Synthase from M. luteus-Membranes from M. luteus (formerly M. lysodeikticw) containmannosyltransferases that catalyze the transfer of mannose from GDP-Man to undecaprenyl phosphate (Man-P-Undec synthase), the subsequent transfer of mannosyl units from the mannolipid to mannan, and the transfer of 1-2 mannose units from GDP-Man to mannosyl diglycerides (20, 21). To obtain a preparation of Man-P-Undec synthase that would efficiently catalyze the synthesis of a variety of P-Man-P-isoprenols, it was necessary to separate the enzyme from the mannosyltransferases synthesizing mannosyl diglycerides that would compete for GDPMan or would consume the Man-P-polyisoprenol product by transferring the lipid-bound mannose to mannan. Initially, 80% of Man-P-Undec synthasewas solubilized by extracting
0
50
100
150
[UNDECAPRENYLPHOSPHATE]
200 (pM)
FIG. 2. Dependence of partially purified Man-P-Undecsynthase preparations on exogenous undecaprenylphosphate. Incubation mixtures contained partially purified Man-P-Undec synthase (32 pg of protein), 50 mM Tris-HC1, pH 7.4, 5 mM MgC12, 0.1% CHAPS, 20 p~ GDP-[3H]mannose (12 cpm/pmol), and theindicated amount of undecaprenyl phosphate (dispersed in 1%CHAPS) in a total volume of 0.05 ml. Following incubation at room temperature for 10 min, the transfer of labeled mannose into Man-P-Undec was assayed as described under "Experimental Procedures."
in Pig Brain
Man-P-Dol-mediated Mannosyltransferases
13114
FIG. 3. Recognition of a-isoprene unit of dolichyl moiety by pig brain w Man-P-DokOligo-P-P-Dol manno- 2 syltransferase(s). Assay mixtures contained pig brain microsomes (0.5 mg of LL protein), 50 mM Tris-HC1, pH 7.4, 1% < 5 \ .E Triton X-100, 2 mM CaC12, and the inE dicated concentration of Man-P-doli- -I o \ chols (0)( C Sor ~ c66)or fully unsaturated $, a Man-P-polyprenols(0)(c, or Cas) (600 cpm/pmol)in a total volume of 0.05 ml. z Following incubation at 37 “C for 10 min, the amount of labeled mannose transferred to endogenous Oligo-P-P-Dolwas assayed as described under “Experimental Procedures.”
B
’A
-
4
-E”
-k
1
0
2
4
6
[Man-P-Prenolos]
8 (pM)
100
5
10
[Mon-P-Prenol551
15
20 bM)
TABLEI Comparison of Man-P-Polykoprenolsas substrates forMan-PUndec:Mannan mannosyltransferase in M. luteus membranes Themannolipidswerecompared as substrates for Man-P-Undec:Mannan mannosyltransferase at a final concentration of 13 p~ as described under “Experimental Procedures.” Mannolipid substrate
[3H]Mannose incorporated into M.luteus mannan
pmol/min/mg
Man-P-POly,, Man-P-DO&,a Man-P-Polys5 Man-P-Dolss
23.6 22.6 31.2 31.2
from thedolichyl group only by the presenceof an unsaturated a-isoprene unit. In contrast to these results, Man-P-Undec:mannan mannosyltransferaseactivityassociated with M . luteus mem1 branes did not discriminate between Man-P-Dolbb or Man-Pwas higher Polys6 (Table I). This mannosyltransferase activity -0.5 0.0 0.5 1 when mannosyl donors containing 19 isoprene units (Man-P1/[MANNOLIPID] (prn)” Dolg5 and Man-P-Polyg5), and therefore more hydrophobic, were compared as substrates. FIG. 4. Effect of chain length of polyisoprenyl moiety on In order to determine if the length of the polyisoprenol Man-P-Do1:Oligo-P-P-Dol mannosyltransferaseactivity in chain, and therefore the hydrophobicity, of Man-P-Dol had pig brain microsomes. Assay mixtures consisted of pig brain mia n effect on its utilization as mannosyl donor by Man-P- crosomes (0.5 mg of protein), 50 mM Tris-HC1, pH7.4, l %Triton XDo1:Oligo-P-P-Dol mannosyltransferase inpig brain, Man-P- 100,2 mM CaCI2, andthe indicated concentrationof Man-P-dolichols (CS5.e; Cs5, A; or C , , 0 ) (600 cpm/pmol) in a total volume of 0.05 dolichols containing polyisoprenol chains with either 19 (Cgb), ml. Following incubation at 37 “C for 10 min, the transfer of labeled 11 (c&), or2 (Clo)isoprene unitswere compared as substrates. mannosyl units to endogenous Oligo-P-P-Dol acceptorswas assayed Citronellol(Cl0) was chosen for the synthesis of a water- as described under “Experimental Procedures.” MPD, Man-P-Dol;n soluble analogue because it is very closely related t o dolichol = number of carbon atoms in dolichol chain. structurally, consisting onlyof the saturated a-isoprene unit TABLEI1 and the terminal unsaturated 5-carbon unit. The data illustrated inFig. 4 show that the initial rates for the Man-P-Dol-Effect of anomeric configurationof mannosyl linkngeon Man-PDo1:Oligo-P-P-Dol mannosyltransferase activity mediated mannosyltransferase activity increase as the length of the polyisoprenol chain, and therefore the hydrophobicity, The preparation of the mannolipid substrates and the assay of Man-P-Do1:Oligo-P-P-Dol mannosyltransferase activity are deincreases. Man-P-Dolgs was clearly the most preferred subMan-P-Dol~5sub, Man-P-Dols5 scribedunder“ExperimentalProcedures.”The strate with an apparent K,,, of 3 p ~ whereas strates were compared at a final mannolipid concentrationof 2 pM. had an intermediate apparent K, value of 8 PM. The brain Mannolipid [3H]Mannose incorporated into enzyme(s) exhibited thelowest affinity for Man-P-Doll0 (apsubstrate Oligo-P-P-Dol parent K, approximately 200 pM). However, identical Vmax pmollminfmg protein valuesforMan-P-Do1:Oligo-P-P-Dol mannosyltransferase 6-Man-P-Dol 0.29 activity were attained with each of the three mannosyl donors. a-Man-P-Dol 0.02 Thus, the catalytic efficiency (Vmax/K,) of the mannosylation reactions increases with the hydrophobicity of the manfectively than a-Man-P-Dolg5as a substrate for thepig brain nosy1 donor. The structural requirementfor the anomeric configuration mannosyltransferase activity in uitro. A similar difference i s of themannosyl moiety of Man-P-Dol was examined by seen when the water-soluble citronellyl stereoisomers (Mancomparing /3-Man-P-Dolgb with the corresponding a stereo- P-Dollo) were compared (Fig. 5, upperpanel). To exclude the possibility that the differences seen with isomer, a-Man-P-Dolg5.As can be seen in Table 11, p-Manof entry P-Dolgs is utilized as mannosyl donor considerably more ef- the citronellyl analogues were due to differential rates
.o
Man-P-Dol-mediated Mannosyltransfermes inPig Brain
30 ’ A
20
t
Control
BMan-P-Cit
k aMan-P-Cit
+ TRITON X-100
@Man-P-Cit
I f 0 0
aMan-P-Cit
s, Y
Y
Q
10
20
30
13115
peaks corresponding to ManGsGlcNAcz could also be seen. Following treatment with Endo H (Fig. 6A, 0, dotted trace), the major labeled oligosaccharide was quantitatively converted to a smaller size, and eluted from Bio-Gel P-4 in a volume expected for MangGlcNAcl.The lipid-bound oligosaccharides labeled by Man-P-DoLo, Man-P-Dolss, and Man-PDO195 were all cleaved by Endo H supporting the conclusion that they are Man9GlcNAcz or mannosylated biosynthetic intermediate oligosaccharides (Fig. 6 , A-C, dotted truces). After treatment with a-mannosidase, the radioactivity in the oligosaccharide products labeled by the three Man-Pdolichols was released as free [3H]mannose, showing that all of the mannosyl units were added in an a-configuration (Fig. 6 , D-F, 0). The size distributions of the lipid-linked oligosaccharide products varied slightly depending on the mannosyl donor. As shown in Fig. 6C (solid trace), MansGlcNAcz represents a smaller proportion of the oligosaccharides labeled by Man-PDollo. This result suggests that the mannosyltransferase(s) catalyzing the attachment of the ninth mannose residue may be more selective for the lipid moiety of the mannosyl donor than the mannosyltransferases converting Man5GlcNAcz-PP-Dol to MansGlcNAcz-P-P-Dol.However, the exact number of enzymes catalyzing these four mannosyl transfer reactions remains to be established. DISCUSSION
In the current view of Glc3MangGlcNAcz-P-P-Dolbiosyn4 a-mannosyl residues to thesis, Man-P-Dol donates TIME OF INCUBATION (min) Man5GlcNAcz-P-P-Dolon the lumenal surface of the RER FIG. 5. Comparison of a- and 6-Man-P-Cit (Man-P-Dol,o) (reactions illustratedin Fig. 1).The mechanism by which the a8 mannosyl donors for pig brain Man-P-DokOligo-P-P-Dol lipophilic mannosyl donor and theheptasaccharyl-lipid intermannosyltransferaseactivityinthe absence (panel A ) or presence (panel B ) of 1.0% Triton X-100.Incubation mixtures mediate diffuse transversely from the cytoplasmic leaflet, the consisted of pig brain microsomes (0.5 mg of protein), 50 mM Tris- site of synthesis, to the lumenal monolayer remains to be HCl, pH 7.4,2 mM CaC12,and 0.05 mM a-Man-P-Cit(O), or @-Man- determined. P-Cit (0)in the absence or presence of 1.0% Triton X-100 in a total This studyhas investigated the specificity of Man-Pvolume of 0.05 ml. After incubation at 37 “C for 10 min, the amount Do1:Oligo-P-P-Dol mannosyltransferase activity in pig brain of labeled mannose transferred to endogenous Oligo-P-P-Dolaccepby surveying a series of Man-P-isoprenyl substratesthat were tors was assayed as outlined under “Experimental Procedures.” synthesized enzymatically or chemically. The mannosubinto theER vesicles, they were compared as substrates in the strates varied in the structure of the a-isoprene units, the absence (Fig. 5, upper panel) or presence of (Fig. 5, lower chain-length (and therefore the hydrophobicity of the polyipanel) of 1.0% Triton X-100. The results of this comparison soprenol moieties), and the stereoconfiguration of the manreveal that the @-stereoisomerof Man-P-Dollo is also a sig- nosy1 linkage. The first major conclusion is that the brain nificantly better substrate in thepresence of 1.0%Triton X- enzyme(s) has an impressively strict specificity for Man100,under conditions that would unseal the ER vesicles. This P-dolichols over the corresponding Man-P-polyisoprenols. result indicates that thedifferences in activity are not due to Similarly, preliminary studies have shown that Manfailure of the a-isomer to enter asealed vesicle to be accessible P-Dol:Man5GlcNAc2-P-P-Dolmannosyltransferase purified from pig aorta exhibits preference a for Man-P-Dol over Manto themannosyltransferases. It is possible that thespecificity for the 6-linked mannosyl P-Poly (44). This is a remarkable property because the two unit is even more strict than these results indicate because substrates differ only in that the a-isoprene unit of dolichol has been reduced, while the polyprenols are fully unsaturated. the low rate of mannosylation by a-Man-P-Dols couldbe These results arerelevant to theobservation that a mutant attributed in part to a small amount of the @-isomerpresent in the a-Man-P-Dolpreparation. The NMR spectrum of the Chinese hamster ovary cell line isolated by Krag and colchemically synthesized Man-P-Dollopreparation indicated leagues (13, 14) synthesizes a truncated lipid-bound oligosacthe presence of approximately 7% of the 6-stereoisomer. Since charide intermediate lacking the a-mannosyl units derived the brain mannosyltransferase(s) also prefer the &isomer of from Man-P-Dol. Biochemical studies suggest that the muthe water-soluble analogues, it is unlikely that thedifferences tant is deficient in the “reductase” that saturates the CYseen with the dolichyl stereoisomers are due simply to the a- isoprene unit of dolichol. In the absence of available Dol-P, the mutant synthesizes significant amounts of Man-P-Poly configuration forming an unreactive micellar structure. Characterization of the Dolichyl-bound Oligosaccharides La- from a “default” pool of Poly-P, but the subsequent transfer beled In Vitro-The lipid-linked oligosaccharides labeled by of mannose to Man5GlcNAcz-P-P-(Poly)occurs at an exeach mannosubstrate were released from the carrier lipid by tremely low rate, and theformation of GlcsMan5GlcNAc2-Pmild acid hydrolysis and analyzed by gelfiltration chromatog- P-(Poly) is observed. The new information reported here on raphy and by glycosidase sensitivity. The major oligosaccha- the recognition of the saturated a-isoprene unit of Man-Pride labeled by Man-P-DolO5(Fig. 6A, 0, solid trace) eluted Dol by the lumenally oriented mannosyltransferase(s) profrom Bio-Gel P-4 in the same volume as MangGlcNAcz.Minor vides a biochemical explanation for the variant lipid-linked
13116
20
40
60
20
40
60
20
40
60
80
FRACTIONNUMBER
FIG. 6. Chromatographic analysis on Bio-Gel P-4 of dolichyl-bound oligosaccharides labeledby Man-P-Dols5 (panel A ) , Man-P-Do155(panel B ) , or Man-P-Dollo(panel C).The enzymatically mannosylated oligosaccharideswere liberated from the carrier 0 )and after treatment with lipid by mild acid hydrolysis and analyzed chromatographicallyon Bio-Gel P-4 before (panels A-C, solid traces, Endo H (panels A-C, dotted truces, 0 ) or a-mannosidase(panels D-F,0) under the conditions describedunder "Experimental Procedures." The arrows indicate the elution positionsoE blue dextran (V,,);M$V = Man9GlcNAc;M = mannose.
oligosaccharide formed by the Chinese hamster ovary mutant. It will now be important to understandwhy Man-P-Poly is not efficiently utilized as a mannosyl donor by the brain enzyme(s). Since Man-P-Undec is an activated mannosyl donor in micrococcal mannan biosynthesis, saturation of the a-isoprene unit is evidently not necessary for the free energy of transfer in GDP-Man to be preserved in the mannolipid. In fact, this instudy, membrane-bound Man-P-Undec:mannan mannosyltransferase activity was found to efficiently utilize either Man-P-Dol or Man-P-Poly as the mannosyl donor. It is quite possible that the double bond at the C2-C3 position restricts rotation of the phosphorylmannose group in a way that prevents it from entering or interacting with a reactive pocket in the pig brain RER enzymes. The observation that the bacterial mannosyltransferase activity will catalyze reactions containing mannolipids with either a that the saturated or an unsaturated a-isoprene unit indicates reactive sites of all mannosyltransferases are not identical. It is also interesting that the affinity of the RER mannosyltransferase(s) for Man-P-dolichols increased as the chain length,and thus the hydrophobicity, of the polyisoprenyl moiety increased. This result suggests that anydefect aborting chain elongation by the cis-isoprenyltransferases catalyzing the de novo synthesis of dolichol would have consequences in at least the Man-P-Dol-mediated stage of Glc3ManeGlcNAczP-P-Dol assembly and probably the glucosyltransferases mediated by Glc-P-Dol. Recently, Berendes and Jaenicke (45) have also found that the natural isoprenologues are better substrates than shorter chemically synthesized Dol-P species for GlcNAc-P-P-Dol synthase and GDP-Man:(GlcNAc)z-P-
P-Dol mannosyltransferase activity in calf liver microsomes. It was observed over 20 years ago that electron transport in delipidated mitochondria isolated from several sources is restored by quinones containing the same isoprenyl side chain as thenative cofactor (46). The third feature of the specificity of the brain enzymes is their preference for the @-linked mannolipid over the astereoisomer. The discrimination between the two stereoisomers was determined by comparing D-Man-P-Dolg6, the naturalintermediate,and the water-soluble analogue, pMan-P-Dollo (Man-P-Cit) with the corresponding a-linked Man-P-dolichols. Since the mannosyl units in Oligo-P-P-Dol derived from Man-P-Dol are a-linked, there is an inversion of configuration during the formation of the lipophilic mannosyl donor from GDP-Man andduring the subsequent transfer of mannose to the Oligo-P-P-Dol substrates. The partially purified preparations of Man-P-Undec synthase from M. luteus, obtained by a facile two-step procedure, synthesized the @-stereoisomersof the Man-P-isoprenols used in these experiments by transferring mannose from GDPMan to the various defined isoprenyl phosphates at a very high efficiency. In most preparations over 80% of the GDPMan could be converted to the respective Man-P-isoprenol product. The ability of the micrococcal enzyme to mannosylate a wide range of isoprenyl monophosphates makes it a very useful reagent for preparing mannosubstratesfor similar studies in otherexperimental systems. Moreover, since ManP-citronellol, a water-soluble analogue of Man-P-Dol, could serve as a substrate for the brain mannosyltransferases, it may have applications for investigating other Man-P-Dol-
13117
Man-P-Dol-mediated Mannosyltransferases i n Pig Brain mediated reactions under detergent-free conditions. Further studies with the citronellyl derivative are in progress in this laboratory. Ultimately, it willbe important to determine the exact number of glycosyltransferases catalyzing the four mannosyl transfer reactions involving Man-P-Dol (Fig. 1). Since the enzyme purified by Elbein and his co-workers (19) appears to add only 1 mannosyl residue, there are at least two lumenally oriented mannosyltransferases operating inthis pathway. The simplest conclusion from the characterization of the lipidlinked oligosaccharides labeled inthis study is that four mannosyltransferases are active under these in vitro conditions andthat all four mannosyl transfer reactions occur very rapidly. This is consistent with the observation that when Oligo-P-P-Dol synthesis has been studied by GDP-mannose orMan-P-Dol labeling with brain microsomes in vitro, MansGlcNAcp-P-P-Dol appears to be the major product formed (42,47). Further work will berequired to determine if these mannosyltransferases exist in a multienzyme complex with their active sites oriented toward the lumenal space. Finally, some key aspects of the specificity of pig brain Man-P-DokOligo-P-P-Dol mannosyltransferaseactivity have been defined. The results emphasize the importance of understanding more about the mechanisms and regulation of the enzymes involved in the reduction of the a-isoprene unit and chain-elongation of dolichol in de rwvo biosynthesis. Defects in either of these enzymatic processes would be expected t o impair the normal assembly of Oligo-P-P-Dol intermediates and consequently protein N-glycosylation.
Vol. XVI, pp. 449-488, Academic Press, New York 6. Rearick, J. I., Fujimoto, K., and Kornfeld, S. (1981) J. Biol. Chem. 2 5 6 , 3762-3769 7. Stoll,J.,Robbins, A. R., andKrag, S. S. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,2296-2300 8. Tanner. W.. and Lehle. L. (1987) Biochim. B ~ ~ K J ~Acta v s . 906.81-99 9 . Orlean,'P. (i990) ~ o iCell.'Biol,'lO, . 5796-580'5 10. Snider, M.D., and Rogers, 0. C. (1984) Cell 3 6 , 753-761 11. Haselbeck. A,. and Tanner. W. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 1520-1524 12. Orlean, P., Albright, C., and Robbins, P. W. (1988) J. Biol. Chem. 263, 17499-17507 13. Stoll, J., and Krag,S. (1988) J. Biol. Chem. 263,10766-10773 14. Stoll, J., Rosenwald, A. G., and Krag, S. (1988) J. Biol. Chem. 263,1077410782 15. Mankowski. T.. Sasak. W.. Janczura., E.., and Choinacki.. T. (1977) . . Arch. Biochem.'Biophys. i b l , 393-401 16. Jankowski, W. J., Palamarczyk, G., Krajewsda, I., and Vogtman, T. (1989) Chem. Phys. Lipids 51,249-259 17. SzpY'r,ska:A2 Swiezewska, E., and Chojnacki, T. (1992) Int. J. Biochem.
-
.
4 , llDl-ll0i
18. McLachlan, K. R., and Krag, S. S. (1992) Glycobiobgy 2,313-319 19. Sharma. C. B.. Kaushal. G. P... Pan.. Y. T.. and Elbein.. A. (1990) . . Biochemistry 29,8901-8907 ' 20. Lennarz, W. J., and Talamo, B. (1966) J. Biol. Chem. 241,2707-2719 21. Scher, M., and Lennarz, W. J. (1969) J. Bwl. Chem. 244,2777-2789 22. Warren, C. D., and Jeanloz, R. W. (1973) Biochemistry 12,5038-5045 23. Warren, C. D., and Jeanloz, R. W. (1978) Methods Enzymol. 60,122-137 24. Rush, J. S., and Waechter, C. J. (1992) Glycobiology 2 , 483 (Abstr. 9.29) 25. Braell, W. A., Tyo, M. A., Krag, S. S., and Robbins. P. W. (1976) Anal. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
Acknowledgments-We thank Drs. Dean Crick and David Frank for their helpful suggestions and encouragement during the course of these studies. We are also grateful to Rose Bouaziz (Ecole Normale Superieure, Paris, France) for her assistance in the preparation of the manuscript. We also thank the Commonwealth of Kentucky and the University of Kentucky for the use of the University of Kentucky Biological NMR Facility. 1. 2. 3. 4. 5.
REFERENCES Kornfeld, R., and Kornfeld, S. (1985) Annu. Reu. Biochem. 5 4 , 631-664 Hirachberg, C. B., and Snider, M. D. (1987) Annu. Reo. Biochem. 5 6 , 6387 Waechter, C. J. (1989) in Neurobiology of Glycoconjugates (Margolis, R. U., and,Margolis, R. K., e&) pp. 127-149, Plenum Press, New York Parod, A., and Leloir, L. F. (1979) Bzochim. Biophys. Acta 5 5 9 , l - 3 7 Presper, K. A,, and Heath, E. C. (1983) in The Enzymes (Boyer, P. D., ed)
36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.
Riochpm. .... ... ... 74.484-4R7 . , -. . -.
Rush, J. S., and Waechter, C. J. (1992) Aml. Biochem. 2 0 6 , 325-333 Glaser, L. (1966) Methods Enz mol 8,183-185 Bartlett, G. R. (1959) J. Biol. & 234,466-468 e i . Preiss, J. (1966) Methods Enzymol. 3 , 271-275 Waechter, C. J., and Scher, M. G. (1981) in Research Methods in Neuro201-231, Plenum chemistry (Marks, N., and Rodnight,R., eds) Vol. 5, pp. ~. Press, New York Danilov, L. L., and Cho'nacki, T. (1981) FEBS Lett. 1 3 1 , 310-312 Folch, J., Lees, M., andsloane-Stanley, G. A. (1957) J. Biol. Chem. 2 2 6 , 497-509 Steiner, S., and Lester R. L. (1972) J. Bacteriol. 1 0 9 , 81-88 Waechter, C. J., and tester, R. L. (1973) Arch. Biochem. Biophys. 158, 401-410 Dunphy, P. J.,Kerr, J. D., Pennock, J. F., andWhittle,K. J. (1967) Biochim. Biophys. Acta 13,136-147 Dittmer, J. C., and Lester, R. L. (1964) J. Lipid Res. 5 , 126-127 OConnor, J. V., Nunez, H. A., and Barker, R. (1979) Biochemistry 1 8 , 500-507 Rush, J. S., and Waechter, C. J. (1991) Glycobiology 1,229-235 Wedgwood, J. F., Strominger, J. L., and Warren, C. D. (1974) J. Biol. Chem. 249,6316-6324 Crick, D. C., Rush, J. S., and Waechter, C. J. (1991) J. Neurochem. 5 7 , 1354-1362 Sumbilla, C., and Waechter, C. J. (1985) Methods Enzymol. 111,471-482 Harford, J. B., and Waechter, C. J. (1979) J. Neurochen. 32,1707-1715 Rodriguez-Vico, F., Martinez-Cayuela, M., Garcia-Peregrin, E., and Ramirez, H.(1989) Anal. Biochem. 183,275-278 D'Souza-Schorey, C., Malchan, K., Krag, S., andElbein, A. D. (1992) Glycobiology 2 , 476 (Abstr. 9.01) Berendes, R., and Jaenicke, L. (1992) Biol. Chem. Hoppe-Seyler 3 7 3 , 3542 Crane, F. L. (1977) Annu. Reu. Biochem. 46,439-469 Waechter, C. J., Kennedy, J. L., and Harford, J. B. (1976) Arch. Biochem. Biophys. 174,726-737
