Isolation, Characterization, and Expression of a cDNA Encoding N ...

VOl.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1993 by The American Society for Biochemistry and Molecular Biology, Inc

No. 21, Issue of July 25, pp. 15381-15385,1993

Printed in U.S.A.

Isolation, Characterization, and Expression of a cDNA Encoding N-AcetylglucosaminyltransferaseV* (Received for publication,February 16, 1993, and in revised form, April 12, 1993)

Mohamed ShoreibahS, Guang-ShingPerngj, Beverly Adlerll, Jasminder Weinsteinn, Rita Basun, Rod Cupplesn, Duanzhi Wenn,Jeffrey K. Brownell, Phillip BuckhaultsS, Nevis Fregienj, and Michael PierceSII From the $Department of Biochemistry and Complex Carbohydrate Research Center, University of Georgia, Athens, Georgia 30602, VAmgen, Inc., Thowand Oaks, California 91320,and the §Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, Florida 33101

A cDNA clone for the complete coding sequence for TV expresses in most cells the majority of the N-linked poly-

a-1,3(6)-mannosylglycoprotein B- 1,6-N-acetylgluco-

N-acetyllactosamine (polyLacNAc) sequences (3, 9); theresaminyltransferaseV (GlcNAc-TV, EC 2.4.1.165) was fore, the regulation of GlcNAc-T V activity in large part can isolated and expressed inCOS-7 cells. Degenerate oli- control the expression of polyLacNAc on cell surfaces. This gonucleotideprimersforpolymerasechainreaction is important because polyLacNAc sequences have been imwere synthesized based on the amino acid sequence ofplicated in mediating cell adhesion (10,111. These cell surface three tryptic peptides isolated from affinity-purified changes involving increases in the /3(1,6) branch and polyGlcNAc-T V. Polymerasechainreactionamplimers LacNAc expression after transformation help to explain at were isolated from rat and mouse mRNA. A cDNA- least in part many seminal observations by Warren, Glick, encoding full-length enzyme was isolated from a rat 1 Buck, and other investigators (12, 13) who have documented cell (EJ-ras-transformed) library and sequenced. Transient expression ofthis clone in COS-7cells, fol- changes in N-linked oligosaccharides after malignant translowed by enzymatic activity assays, demonstrated that formation. GlcNAc-T V is expressed in low copy number in many this cDNA sequence encodes GlcNAc-T V. Northern analysis of rat kidney mRNA revealed a single band tissues and cultured cell lines, although its specific activity corresponding to a length ofabout 7 kilobases. Se- varies considerably between tissues (14, 15). Two lectin-varquence analysis of the cDNA clone demonstrated an iant cells are known that have specific lesions in the expresopen reading frame that encoded a typeI1 membrane sion of GlcNAc-T V activity, the mouse lymphoma BW5147 protein of740 amino acids. PHAR’.’ and Chinese hamster ovary Lec 4 variants (1, 16). Another Chinese hamster ovary variant, Lec4A, expresses activity, but its GlcNAc-T V has altered kinetic properties and is mislocalized, causing a lectin-resistant phenotype simc~-1,3(6)-Mannosylglycoprotein/3-1,6-N-acetylglucosami- ilar to that of the Lec 4 cells (17). Recently, we purified the nyltransferase V (GlcNAc-T V)’ is a Golgi enzyme that cat- enzyme over400,000-fold to homogeneity from rat kidney alyzes thetransfer of N-acetylglucosamine (GlcNAc) in acetone powder (18). /3(1,6)-linkageto the a(l,6)-linked mannose of biantennary In order to study in detail GlcNAc-T V and the regulation N-linked oligosaccharides (1).This enzyme is of particular of its activity, we have isolated a cDNA that contains the interest because its activity in many cell types has been shown entire coding sequence for GlcNAc-T V. The sequence ento significantly increase after transformationby diverse tumor codes a type I1 membrane protein of 740 amino acids. When viruses (2, 3) and isolated oncogenes, including several ras transiently expressed in COS-7 cells, the cDNA produced a genes (4-6) and v-src’c.2In addition, changes in its activity 16-fold increase in GlcNAc-T V specific activity over endogappear to correlate with changes in the metastatic potential enous levels, demonstrating that the protein it encodes is of several cell lines (7,8). The branch synthesized by GlcNAc- catalytically active. A fragment of the mouse cDNA specifically hybridized to a rat kidney mRNA with a size of about 7 * This research was supported in partby National Cancer Institute kb.

Grant CA 49926 (to N. F.). The costs of publication of this article were defrayed in part by the payment of page charges. This article EXPERIMENTALPROCEDURES must therefore be hereby marked “aduertisement” in accordance with Preparative Purification of GkNAc-T V, Proteolytic Digestion, and 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequencefs) reported in thispaper has been submitted Microsequencing-GlcNAc-T V was purified essentially as described totheGenBankTM/EMBLDataBankwith accession numberfs) previously (18).Briefly, eluants from multiple purifications using the synthetic oligosaccharide affinity column containing about 30 pg of L14284. 11 Recipient of a faculty research award from the American Cancer protein in a volume of 150 mlwere concentrated by precipitation Society. To whom correspondence should be addressed.: Tel.: 706overnight in 7 volumes of methanol at -20 “C. The precipitate was 542-1702; Fax: 706-542-1759. centrifuged at 10,000X g at 4 “C for 2 h, and theresulting pellet was The abbreviations used are: GlcNAc-T V, ~~1,3(6)-mannosylgly-resuspended in 100 pl of 0.1% SDS at room temperature. This sample coprotein ~-l,6-N-acetylglucosaminyltransferase V; polyLacNAc, was then applied to anApplied Biosystems high performance electropoly-N-acetyllactosamine;PAGE, polyacrylamide gel electrophoresis; phoresis chromatography (model 230A) to purify the 75- and 69-kDa HPLC, high pressure liquid chromatography; PCR, polymerase chain bands representing putative GlcNAc-T V away from any contamireaction; bp, base pair(s); kb, kilobase(s); PBS, phosphate-buffered nants before amino acid sequencing. Fractions containing only the saline. doublet (as determined by 8-25% SDS-PAGE, Phast System, Phar* P. Buckhaults, M. Shoreibah, N. Fregien, and M. Pierce, manu- macia LKB Biotechnology Inc.)were collected, pooled, and dried script in preparation. under vacuum. The sample was then shipped on dry ice to the Harvard

’

15381

15382

cDNA Encoding GlcNAc-T V

MicrochemistryFacility where it was reduced withdithiothreitol, RESULTS AND DISCUSSION alkylated with iodoacetamide, resuspended in urea, and then subThe strategy employed to isolate the GlcNAc-T V cDNA jected to trypsin digestion. The sample was then chromatographed by using reverse phase HPLC (Vydac C18, 2.1 X 150 mm), and the was to obtain amino acid sequence data from peptide fragamino acid sequence of three peaks was determined. These peaks mentsgenerated by trypsin digestion of highly purified were chosen for gas-phase Edman degradation based on the magniGlcNAc-T V and then to utilize these data to design degentude of theirabsorbances at 210 and 280 nm and on their clear erate oligonucleotides to be used as primers for the reverse separation from other peaks. cDNA Libraries and Screening Procedures-Three cDNA libraries transcriptase-PCR analysis of rat and mouse mRNAs. First, were screened for the isolation of the GlcNAc-T V cDNA sequences: microgram quantities of GlcNAc-T V were purified from rat an oligo(dT)primedmousecDNAlibrary was constructedfrom kidney acetone powder by affinity chromatography as depoly(A)+ RNA from BW5147 cells by using the Librarian I kit (In scribed (18) and were further purified by preparative SDSVitrogen, Inc.) that uses the vector pcDNAl; a random primed cDNA polyacrylamide gel electrophoresis with direct elutionof samlibrary was constructedfrom mouse 3T3 cell poly(A)+ RNAand of the purification cloned into &ZAP I1 (Stratagene) (19); and a rat I-EJ cDNA library ples in order to ensure purity after scale-up procedure and to maximize recovery. Fractions containing (20) was cloned into pSPORT-1 plasmid vector (Life Technologies, 75- and 69-kDa Inc., Bethesda, MD). The libraries were screened by conventional protein species corresponding to both the filter hybridization techniques (21)by using either “P-labeled probes forms of the enzyme were pooled and digested with immobiand autoradiographyor fluorescein-labeled probes andimmunological lized trypsin;theresultingpeptides were fractionated by detection (enhanced chemiluminescence 3‘-oligolabeling and detecreverse-phase HPLC. Three peptides were selected for Ntion kit, Amersham Corp.). In some cases the libraries were divided into pools and screened for the presence of GlcNAc-T V sequences terminal amino acid sequence determination as follows: Peptide 1, Asn-Thr-Asp-Phe-Phe-Ile-Gly-Lys-Pro-Thr-Leu-Arg; by PCR or Southern analysis, prior to filter hybridization. Oligonucleotide Synthesis, Reverse Transcriptase-PCR, and DNA Peptide 2, Ala-Ile-Leu-Asn-Gln-Lys-Ile-Glu-Pro-Tyr-MetSequencing-Oligodeoxynucleotides were synthesized on an Applied Pro-Tyr-Glu-Phe-Thr; Peptide3, Val-Leu-Asp-Ser-Phe-GlyBiosystems automated DNA synthesizer. PCR reactions were done Thr-Glu-Pro-Glu-Phe-Asn. by using Therrnus aquaticus (Taq) DNA polymerase (Perkin-Elmer Oligonucleotides were designed from a region within PepCetus or Promega) utilizing manufacturer-supplied buffers and rec1 and 2, based ontheleastamount of degeneracy. tides ommended conditions. cDNAs for reverse transcriptase-PCR were synthesized from either total cytoplasmic RNA or poly(A)’ RNA by Sequences were chosen from the centralregion of the peptide Moloney murine leukemia virus-reverse transcriptase(Life Technol- sequence when possible,since thiswould permit confirmation ogies, Inc.) by using the supplied buffer and either oligo(dT) (total of any cDNAclones obtained by PCR by comparison of amino RNA) or random hexanucleotides (poly(A)+ RNA) as a primer. Seacids encoded by sequences adjacent to primer sequences with quencing was done by using Taq DyeDioxy Terminator cycle sequenc- theknown, peptide-derived amino acidsequences. Within ing kits (Applied Biosystems, Inc.) and an automated DNA sequenthese oligonucleotides, inosines were substituted at positions ator (Applied Biosystems 3736) and following the manufacturer’s of four-base degeneracy. The selectedoligonucleotides are instructions. shown as follows: GNT-V 1, 5”AAYACIGAYTTYTTYAT Transient Expression of Rat GlcNAc-T V in COS-7 Cells-The entire cDNA insert from one rat GlcNAc-T V clone was ligated into (Y/A)GGIAARCCNAC;GNT-V l o (antisense),5’-GTIG pJT-2 plasmid expression vector (20). COS-7 cells (American Type GYTTICC(R/T)ATRAARAARTCIGTRTT; GNT-V 2, 5” Culture Collection, CRL 1651, Rockville, MD) were transfected with AT(Y/A)GARCCITAYATGCCITAYGARTTYAC;GNT-V the pJT-2 plasmid alone or with pJT-2 plasmid containing the rat 20 (antisense), 5’-TCRTAIGGCATRTAIGGYTC(R/T)AT cDNA by electroporation as follows. 4 X lo6 cells in 0.8 mlof YTTYTG. Dulbecco’s modified Eagle’s medium(LifeTechnologies, Inc.) and 7.5% fetal bovine serum(Bocknek,Toronto,Canada) were transSince the relative positions of the two peptides in protein ferred to a 0.4-cm cuvette and mixed with 10 pg of plasmid DNA in sequence were not known, oligonucleotides were synthesized 10 p1 of water. Electroporation was performed at room temperature containing both sense (GNT-V 1 and 2) and antisense (GNTa t 1600 V and 25 microfarads by using a Gene Pulser apparatus (Biol o and 20 sequences). This permitted PCR testing for the V Rad) with the pulse controller unit set a t 200 ohms. The cells were 40 ml of Dulbecco’s modified Eagle’s GlcNAc-T V mRNA in either of the possible orientations, then diluted into approximately medium, 7.5% fetal bovine serum,transferred to 100-mm culture with only one of two combinations of oligonucleotides capable dishes. After a 17-h incubation a t 37 “C, the medium was replaced, of giving an amplification product. and incubation was continued for an additional 51 or75 h. The cells The initial analysis used both total and poly(A)+ RNAs were rinsed with PBS, harvested by scraping, washed again with PBS, from two cell lines known to express GlcNAc-T V enzyme and centrifuged to pellet the cells. After the PBS had been aspirated, activity, mouse BW5147 cells (1) and rat 13762 mammary the cell pellet was subjected to quick freezing by immersion of the adenocarcinoma cells (24). TheRNAs were reverse trantube in liquid nitrogen and was kept frozen on dry ice until resusscribed into cDNAby reverse transcriptase primedby random pended for assay. GlcNAc-T V Enzymatic Assays-The pellets of electroporated hexanucleotides and then used in PCR reactions using the COS-7 cells were thawed a t 37 “C and assayed by radiochemical (22) two possible combinations of degenerate primers, GNT-V 1 and enzyme-linked immunosorbent assays (23). The radiochemical and 20 or GNT-V lo and2. These results revealed that both assay mixture contained the following reagents that were dried under poly(A)+ RNAsgive a 178-bp amplification productusing the lo6 cpm of UDP-[3H] vacuum in a1.5-mlmicrocentrifugetube: GlcNAc (25 cpm/pmol) (American Radiolabeled Chemicals), 1 mM GNT-V 1 and 20 primers but no detectable bands from the synthetic oligosaccharide acceptor (generouslyprovidedby Dr. 0. other set of primers. Bothamplification products were cloned Hindsgaul, University of Alberta), and10 mM EDTA in totalvolume and sequenced. Their sequences showed that they contained of 0.01 ml. Radiolabeled product was isolated and quantitated as a short open readingframe.Translation of the sequences described ( 2 3 ) .All assays were conducted in duplicatefor three time revealed that the amplificationproduce encoded all seven of points, and the results were averaged. The synthetic trisaccharide the amino acids predicted by the peptide sequence, but not acceptor used was octyl6-0-[2-0-(2-acetamido-2-deoxy-~-D-g~u~osy1pyranosyl)-a-~-mannopyranosyl]-P-D-g~ucopryanoside.

Northern Analysis of Rat Kidney mRNA-Ten pgof rat kidney poly(A)+ RNA (Clontech) was subjected to electrophoresis in a 1% agarose gel and blotted to a Zeta-Probe G T membrane.A BglIIHind111 fragment of murine GlcNAc-TV cDNA was radiolabeled and used as aprobe. The blot was visualized byusing aMolecular Dynamics PhosphorImager. A laser printer outputwas generated and photographed.

included in the primer sequences, as well as a LYS at the presumptive trypsin cleavage site. The PCR amplification product was used to screen two mouse cDNA libraries in order to isolate larger cDNA sequences. Screening approximatelylo6clones fromeach library resulted in the isolation of two overlapping cDNAclones. Sequence analysis of these cDNAs demonstrated that they


15383

contained an open reading frame encoding the C-terminal end of the protein, representing about 60% of the length of the polypeptidesobserved after SDS-PAGE(18).The remainingcodingsequence was isolated from a Rat 1-EJ cDNA library by anchoredprimerPCRusing a 5”senseprimer derived from the cDNA library cloning vector, pSPORT-1, and a 3”antisense primer designed from the original 178-bp PCR amplification product. The PCR priming using these two oligomers would identify the 5’-endof the cDNA clones. Using this strategy, an approximately 1.8-kb PCR product was obtained. Sequence analysis of this fragment revealed overlap with the two partial mouse cDNA clones and completed the open reading frame region. The rat PCR amplification product contained the putativeATG translation initiation codon andapproximately 300 bp of 5“untranslated region.Anoligonucleotide, 35 to 15 bp upstream from the ATG start codon, was used to screen the Rat 1-EJ cDNA library and resulted in the isolation of a plasmid containing a 4.8-kb cDNAinsert.Restriction digestion and nucleotide sequence analysisdemonstratedthattheratcDNA clone contained approximately 300 bp of 5“untranslated sequence, a2220-base openreadingframe,andapproximately 2300 bases of apparent 3”untranslated region (Fig. 1). The initiationof translation most likely occurs at theATG at 299, since this codon is 9 bases after the TAG stopcodon a t 290. The open reading frame encodesa740-aminoacid protein with a calculated molecular mass of 84,561 Da. The amino acid sequences of all three peptide fragments are located within the predicted amino acid sequence encoded in the open reading frame and were identical to the original sequences. These are Peptide1,amino acids546-557; Peptide 2, amino acids 592-607; and Peptide 3, amino acids 375-386 (Fig. 1).Hydrophobicity analysis of the amino acid sequence (25) indicated a putative transmembrane region near the N terminus, amino acids 14-30, as expected for a type I1 membrane protein similar to other lumenalGolgi enzymes. There are six consensus sites for N-linked glycosylation. Comparison of the sequences in themouse and ratclones showed that they were greater than 95% homologous at the nucleotide level with only one difference in the C-terminal 445 amino acids,which resultsin a T residue at position 679 in the mouse sequenceas compared with the I residue inrat GlcNAcT V. When the rat GlcNAc-T V sequence was compared with those of other clonedglycosyltransferases,in particular GlcNAc-T I, GlcNAc-T 111, and the “core 2” GlcNAc-T, no significant degree of similarity could be observed. The entire cDNA insert of approximately 4.8 kb was subcloned into the pJT-2 expression vector. This vector, containingthe sequence of theputativeGlcNAc-T V clone, was transfected into COS-7 cells. Identical cultures of cells were transfected with vector alone as a control. The cells were harvestedafter68 or 92 h andassayed for GlcNAc-T V activity by using the synthetic trisaccharide acceptor specific assay confor GlcNAc-TV, UDP-[3H]GlcNA~, and standard ditions (14). EDTAwas included in the incubation toprovide additional evidence for GlcNAc-T V activity since GlcNAc-T V and very few other glycosyltransferases are active in vitro in the absence of exogenous cation. The results from this experiment, shown in Fig. 2, demonstrated that the level of are presented. The 5”untranslated region consists of nucleotides 1298. The methionine-specifyinginitiation codon (ATG)startsat nucleotide 299, and the amino acid coding region spans 2220 bases, encoding a 740-aminoacid protein. The putativetransmembrane domain (amino acids 14-30) is doubly underlined. Tryptic peptides are singly underlined. Consensus sites for N-linked glycosylation are indicated by asterisks. The first103 bases of the 3’-untranslated region are also presented.


15384

GlcNAc T V cDNA

9.49

-

7.46

-

2.37

-

vector alone

2

4

6

8

incubation time (hours)

FIG. 2. GlcNAc-T V assay of COS-7 cells transiently expressing GlcNAc-T V cDNA or vector alone.Cells were assayed by using a synthetic trisaccharide acceptor as described under “Experimental Procedures.”

:.

1.35 -

GlcNAc-T V in the cells transfectedwiththeputative GlcNAc-T V sequence expressedactivity at levels 16-fold higher than the cells that received vector alone. The specific activity of the populationof cells expressing the GlcNAc-TV cDNA was 624 pmol/mg/h.Similarresults were obtained from the transfected cell cultures incubated for 92 h. More0.24 over, assay of the same samples by using an enzyme-linked immunosorbent-based assay for GlcNAc-T V (23), which utiFIG. 3. Northern analysisof rat kidney mRNA. Ten pg of rat lizes specific antibody to detect GlcNAc-TV oligosaccharide kidney poly(A)’ RNA was subjected to electrophoresis ina 1%agarose product, also demonstrated a large increase in activity in the gel, blotted, and probed with a radiolabeled BglII-Hind111 fragment V cDNA. The blot was subjected to GlcNAc-T cells transfected with GlcNAc-TV cDNA (data not shown). of murine Phosphorhager analysis. Translation of theopenreadingframe encoded by the GlcNAc-T V cDNA predicts a protein of about 84.6 kDa. This is larger than thesize observed for affinity-purified ratkidney ranging from 40 to 60 kDa (27, 28). GlcNAc-T V was present GlcNAc-T V, which reveals a doublet of bands at 69 and 75 in very low copy number in thecDNA libraries that contained kDaafterSDS-PAGE, suggesting that proteolysis of the it, and its message is large with a high degree of noncoding kidney enzyme occursduring purification. Preliminary results sequences. Stable transfectants of several cell lines are now suggest that when the affinity purification of the rat kidney being prepared tobe used in experiments designed to test the enzyme is performed rapidly and in the presence of 5-fold hypothesis that changesin GlcNAc-T V activity andconcomgreater levels of a mixture of protease inhibitors, an additionalitant changes in N-linked oligosaccharide expression on the cell surface can regulate cell adhesion and metastatic potenband at about 95 kDa is observed onsilver-stainedSDSPAGE gels.* The 11-kDa difference between the calculated tial. molecular mass and the species with the largest observed Acknowledgments-We thank Dr. Kelley Moremen for advice and molecular mass is most likely due to N-linked glycosylation assistance; Irene Margitich for technical assistance; and Dr. William since there aresix of these potential sites in the sequence. An S. Laine for expertise in peptide cleavage, HPLC, and microsequencequivalent difference between calculated andobserved molec- ing at theHarvard Microchemistry Facility. ular mass hasbeen observed for GlcNAc-T I11 (26) and other REFERENCES glycosyltransferases. The cleavage of the intact enzyme to 1. Cummings, R. D., Trowbridge, I. S., and Kornfeld, S. (1982) J . Riol. Chem. soluble, lower molecular mass species may reflect a normal 257, 13421-13427 proteolytic event that allows the enzyme tobe released from 2. Yamashita. K.. Tachibana., Y... Ohkura.. T... and Kobata. A. (1985) J. Riol. Chem. 260,’3963-3969 cells, as has been observed for several other glycosyltransfer3. Pierce, M., and Arango, J. (1986) J. Biol. Chem. 2 6 1 , 10772-10777 ases (27, 28). Soluble GlcNAc-T V activity has been detected 4. Santer, U. V . , Gilhert, F., and Glick, M. C. (1984) Cancer Res. 44, 3730-R7!I!i . ,., in human serum (23). 5. Dennis, J. W., Kosh, K., Bryce, D. M., and Breitman,M. L. (1989) Oncogene A fragment from the mouse cDNA was radiolabeled and 4,853-860 6. Easton, E. W., Bolscher, J. G., and van den Eijnden, D. H. (1991) J. Biol. hybridized to a Northern blot containing poly(A)+ RNA from Chem. 266,21674-21680 rat kidney. In order togain enhanced sensitivity, the blot was 7 . Dennis, J. W., Laferte. S., Waphorne. C., Breitman, M. L., and Kerbel, R. S. (1987) Science 236,582-585 visualized with a phosphorimaging system, and the resultsof 8. Saitoh. 0.. Wane. W. C.. Lotan.. R... and Fukuda. M. (1992) J. Riol. Chem. this experiment are shown inFig. 3. The cDNA hybridizes to 267’. 5700-5771 9. Carlsson, S. R., Roth. J., Piller, F., and Fukuda, M. (1988) J. Riol. Chem. a moderately broad band about 7 kb in size, suggesting that 263, 18911-18919 there may be more than one GlcNAc-T V transcript in rat 10. Shur, R. D. (1989) Curr. Opin. Cell Riol. 1, 905-912 Q., and Cummings. R. D. (1991) in Cell Surface Carbohydrates and kidney. Similar hybridization patterns have been observed in Jl. Zhou, Cell Deuelooment. (Fukuda, M.. ed) .. pp. 99-126, CRC Press, Inc., Boca other rat and mouse tissues and cell lines. Moreover, these Raton, FL . Click, M. C.. Rabinowitz, 2.. and Sachs, L. (1974) J. Virol. 13,967-974 results demonstrate that less than 40% of this large mRNA 12. 13. Warren, L., Buck, C. A,. and Tuszynski, G. P. (1978) Biochim. Biophys. encodes protein. Acta 516,97-127 14. Pierce. M.. Araneo. J.. Tahir. S. H.. and Hindseaul. 0. (1987) Biochem. In conclusion, we have isolatedacDNA containing the Biophp’Res. Fommun. 146,679-684 entire GlcNAc-T V coding sequence that expresses catalyti- 15. Dennis, J. W., and Laferte, S. (1989) Cancer Res. 49,945-950 16. Stanley. P., Narasimhan, S., Siminovitch, L., and Schachter, H. (1975) cally active enzyme after transient transfection into COS-7 Proc. Natl. Acad. Sci. U. S. A. 72,3323-3327 cells. This enzyme is by far the largest lumenal Golgi glyco- 17. Chaney, W., Sundaram, S., Friedman, N., and Stanley, P. (1989) J . Cell Hiol. 109,2089-2096 syltransferase tobe reported since othercloned glycosyltrans- 18. Shoreibah, M., Hindsgaul, O., and Pierce, M. (1992) J. Riol. Chem. 267, 2920-2927 ferasesencode proteinswithcalculated molecular masses

-

~

~

cDNA Encoding GlcNAc-T V 19. Moremen, K. W. (1989) Proc. Natl. Acad.Sci. (I. S. A. 86,5276-5280 20. Wen, D., Peles, E., Cupples, R., Suggs, S. V., Bacus, S. S., Luo, Y., Trail, G., Hu, S., Silbiger, S. M., Ben Levy, R., Koski, R. A,, Lu, H. S., and Yarden, Y. (1992) Cell 69,559-572 267,16341-16346 21. Sambrook, J., Fritscb, E. F., and Maniatis, T. (1989) Molecular Cloning: A LaboratoryManual, pp. 16.56-16.65, Cold SpringHarborLaboratory, Cold Spring Harbor, NY 22. Palcic, M. M., Ripka, J., Kaur, K. J., Shoreibah, M., Hindsgaul, O., and Pierce, M. (1990) J. Biol. Chern. 265,6759-6769

15385

23. Crawley, S. C., Hindsgaul, O., Alton, G., Pierce, M., and Palcic, M. M. (1990) Anal. Bzochem. 185,112-117 24. Sheng, Z., Wu, K., Carraway, K., and Fregien, N. (1992) J. Biol. Chem. 25. Kyte, J., and Doolittle, R. F. (1982) J. Mol. Biol. 1 5 7 , 105-132 26. Niskikawa, A,, Ihara, Y., Hatakeyama, M., Kangawa, K., andTaniguchi, N. (1992) J. Biol. Chem. 267,18199-18204 27. Scbachter, H. (1991) Curr. Opin. S t r w t . Biol. 1, 755-765 28. Joziasse, D. H. (1992) Glycobiology 2 , 271-277