Isolation of a cDNA clone for β1–4 galactosyltransferase from embryonic chicken brain and comparison to its mammalian homologs

Vol. 189, No. 2, 1992 December 15, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS Pages 1215-1222

ISOLATION OF A cDNA CLONE FOR Pl-4 GALACTOSYLTRANSFERASE FROM EMBRYONIC CHICKEN BRAIN AND COMPARISON TO ITS MAMMAWAN HOMOLOGS Sujoy Ghosh, Shib Sankar Basu and Subhash Basul p2 Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, lN 46556 Received

November

5,

1992

Based on the observed biochemical and immunological similarities between bovine /31-4 galactosyltransferase (GalT) and a developmentally regulated galactosyltmnsferase (GalT-4; UDP-GahLc3 pl-4 GalT) from embryonic chicken brain, a genetic similarity between the two enzymes has been postulated. To test our hypothesis, we have employed a reverse transcriptionpolymerase chain reaction (RT-PCR)based approach and isolated a -600 bp cDNA clone from embryonic chicken brain mRNA. Our results indicate that, within the -600 bp fragment, the avian cDNA nucleotide sequence is 74% homologous to the human 01-4 galactosyltransferase cDNA. Similarity and identity between the predicted amino acid sequences of the two galactosyllransferases are 75% and 61%, respectively. Similar to its mammalian counterparts, the embryonic chicken brain galactosyltransferase gene appears to encode multiple mRNA transcripts (2.3 and 1.4 kb) and shows multiple bands on a Southern blot (18.6,12.9,10.5 and 3.7 kb) indicating that the avian gene is either polyexonic and/or it belongs to a multiple gene family. 0 1992 Academic press, I*=.

The glycosyltransferases constitute a family of highly specialized enzymes whose coordinate action results in the expression of a vast and diverse array of lipid and protein linked oligosaccharides both within the cell and on the cell surface. Changes in the pattern and distribution of these oligosaccharide residues, which most likely result from an altered genetic and/or functional expression of the glycosyltransferases, have been widely associated with critical physiological processes such as ontogeny, ontogeny, signal transduction, and the immune response, to name a few (l-4). One of the more widely studied glycosyltransferases is the group of enzymes collectively known as Bl-4 galactosyltransferase,which catalyzes the transfer of galactose from UDP-galactose to form a pl-4 neoglycosidic bond in glycolipids and glycoproteins. The pl-4 galactosyltransferase is an important enzyme since it is involved in the biosynthesis of the neolacto-core structure (Gal014GlcNAc-R) of carcinoembryonic antigens, blood-group antigens, and polylactosamine epitopes on glycolipids and glycoproteins (5-7). The embryonic chicken brain GalT-4 (UDP-Gal:Lc3 pl-4 GalT) is also a developmentally regulated enzyme (8) and exhibits molecular weight heterogeneity (9,l l), which is indicative of its transcriptional and/or 1Jacob Javits Award recipient. 2To whom reprint requests should be addressed. Abbreviations used: Lc3, GlcNAcpl-3Gal@-4Glc-Ceramide; brain.

1215

ECB, embryonic chicken

0006-291X/92 $4.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol.

189, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

post-translational regulation. In order to understand the molecular bases for the regulation of its gene expression, cDNA clones for pl-4 galactosyl-transferase have been isolated from bovine, murine, and human tissues (12- 15). All three mammalian cDNAs exhibit a high level of nucleotide homology (80-90%) and considerable similarity and identity in their predicted amino acid sequences ( >90% similarity; >80% identity). This observation raises the question as to whether or not this PI-4 galactosyltransferase gene is further conserved through evolution. Based on biochemical and immunological characterization, we have isolated and purified GalT-4 from human colon carcinoma, Co10 205 cells (lOO,OOO-fold,10) and embryonic chicken brain (22,000fold,1 1). In the present work, we have extended the comparison, for the first time, to the embryonic chicken brain pl-4 galactosyltransferase. Based on the regions of high sequence homology between bovine, murine and human pl-4 galactosyltransferase cDNAs, we have designed PCR primers to amplify a putative pl-4 galactosyltransferase cDNA fragment from embryonic chicken brain. The amplified fragment has been used as a probe in Southern and Northern blots. It has also been cloned into a vector and sequenced to determine its level of homology with mammalian pl-4 galactosyltransferase cDNAs. Materials

and Methods

Selection of PCR mimers: Published sequences of bovine, murine and human fil-4 galactosyltransferases were accessedfrom GenBank(Sequence Analysis Software Package of the Genetics Computer Group, Madison, WI; 16) and aligned for regions of maximum homology using different program options. Primers were selected from maximally homologous coding regions of the three cDNAs using the “PCR-Primer Selection Software” (Epicenter Software, Pasadena, CA). 5’- and 3’-end primers were selected based on minimal intra- as well as interprimer sequence complementarity, maximal sequence conservation between the bovine, murine and human sequences, similarity in the primer Tm values and a GC eontent of around 50% for both the primers. The oligonucleotide sequences of the PCR primers used for the experiments are 5’-end primer: 5’-TGTTATCAACCAGGCTGGAG-3’ 3’-end primer: 5’- CCCGATGTCCACTGTGATIT-3’. The 5’-end and 3’-end primers corresponded to nucleotides 446-465 and nucleotides 976-995 respectively, of the bovine pl-4 galactosyltransferase cDNA sequence (GenBank accession number M13214). Both primers were synthesized on an automated DNA synthesizer (Biocore Facility, University of Notre Dame) and purified by hydrophobic interaction chromatography over Oligonucleotide Purification Cartridges (Applied Biosystems,Foster City, CA), essentially following the manufacturer’s protocol. . . erase chain react-on (RT-PCR); Total RNA from 1l-day-old Re erse trm crmtion-Polv _ embryonic chicken brains ZO l.tg) was first riverse transcribed to cDNA using the RNA-PCR kit (Perkin Elmer Cetus, Norfolk, CT). The reverse trancriptase reaction was primed with random hexamers in a final volume of 0.02 ml (17). After keeping the tubes at room temperature for 5 minutes to allow the random hexamers to anneal, the reverse transcription reaction was performed at 42oC for 30 minutes, followed by incubation at 99oC for 5 minutes to inactivate reverse transcriptase. The reaction mixture was then subjected to PCR in the same tube by adding the 5’end and 3’-end PCR primers and other necessary reagents to a final volume of 0.1 ml. Beside the reverse-transcribed cDNAs,the PCR tubes contained the following components: deoxyribonucleotides, 0.2 mM each; reaction buffer, 0.05 M KCl, 0.01 M Tris-HCl, pH 8.4,O. 1 mg / ml nuclease free BSA (Perkin-Elmer Cetus, Norfolk, CT); MgC12,2.5 mM; 5’ and 3’ end primers, 0.25 @4 each; Taq polymerase (Perkin Elmer Cetus, Norfolk, CT), 2 units. The thermocycling profile, optimized for the specific cDNA and PCR primer pair, was as follows: 95oC, 3 min - (95OC, 1 min - 55oC, 1 min - 72oC, 2 min ) 30 cycles - 72OC, 10 min. At the end of the reaction, PCR products were checked on l-l.5 % agarose gels followed by staining with ethidium bromide (0.5- 1.O pg / ml). 1216

Vol.

189,

No.

2,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Northern analvsis: Total ( 10 pg ) RNA samples were electrophoresed through denaturing 2.2 M formaldehyde - 1.0 % agarose gels in 1 x MOPS electrophoresis buffer ( 0.02 M 3morpholinopropanesulfonic acid, pH 7.0,0.005 M sodium acetate, 0.001 M EDTA ) for 5 hours,at 50 volts (constant),at room temperature. After removal of the formaldehyde by washing the gel 5-7 times in water at 65oC, the RNA was capillary-transferred to a nitrocellulose membrane in 10 X SSC buffer. RNA was linked to the membrane by baking in vucuo at 800C for 2 hours. The membrane was incubated with prehybridization/hybridization buffer (50 % formamide, 5 x SSC, 0.05 M sodium phosphate, pH 7.1,O.l % sodium pyrophosphate,lO x Denhardt’s solution, 0.1 % SDS, 0.25 mg/ml denatured herring sperm DNA ) at a ratio of 0.1 ml buffer I cm2 membrane surface at 420C for 4 hours in a rotary hybridization chamber (Robbins Scientific Corp. Sunnyvale,CA). The probe was prepared by random hexamer mediated radiolabeling of the RTPCR derived -600 bp cDNA (18) with Klenow and [a32P]dCTP (3000 Ci/mmol). Hybridization was initiated by adding the radiolabeled probe to fresh buffer at a concentration of 2.5X106 cpm probe per ml buffer. Hybridization was carried out at 420C for 36 hours. At the end of hybridization, the blot was washed three times in a low stringency wash buffer (ZXSSC, 0.1% SDS) at room temperature. A high stringency wash (O.lzXSSC, 0.1% SDS) was then performed twice, for 30 minutes each, at 55oC. The washed blots were autoradiographed at -700C for 24 hours. Southern analvsiS; Embryonic chicken brain genomic DNA (30 pg) was exhaustively digested with Eco RI, overnight at 370C. The digested DNA was separated via 0.9% TBE-agarose gel electrophoresis (0.05 M Tris base, 0.05 M boric acid, 0.001 M EDTA ) overnight, at 50 volts (constant), at room temperature. The DNA was then capillary-transferred in 0.4 M NaOH/ 0.2 M NaCl solution to a Zeta Probe nylon membrane (BioRad, Richmond, CA) for 18 hours at room temperature (19). DNA was linked to the membrane by baking in vacua at 8oOC for 45 minutes. The membrane was incubated in prehybridization/hybridization buffer (0.5 M Na2HPO4, pH 7.2, 7% SDS, 1mM EDTA) for 2 hours at 55oC,in the ratio of O.lml buffer /cm2 membrane surface. Radiolabeled probe DNA (the same as used for the Northern blot) was next added in the ratio of 5X106 cpm/ml of fresh buffer. Hybridization was allowed to proceed for 24 hours at 55OC. Following hybridization, the membrane was washed with 20 mM Na2HP04, pH 7.2,5% SDS two times for 30 minutes each at 55oC and then with 20 mM Na2HP04,pH 7.2, 1% SDS under the same conditions. The washed blots were sealed in plastic bags and autoradiographed at -7oOC for 24 hours. Cloning of the RNA-PCR derived cDNA: The thermostable Taq polymerase used in the RNAPCR reactions has a non-template dependent activity of adding a single deoxyadenosine residue to the 3’-end of all duplex DNA molecules. These A-overhangs were used to insert the PCR product into a specifically designed plasmid vector (Invitrogen, San Diego,CA) providing single 3’T overhangs at the insertion site. After PCR was performed on the reverse-transcribed cDNA obtained from 1l-day-old embryonic chicken brain RNA, an appropriate amount of PCR sample (-30 ng) was carefully withdrawn from underneath the oil overlay and ligated to the pCRTM vector (50 ng, vector:PCR insert at 1:3 molar ratio) according to manufacturer’s instruction. Competent E.cofi cells (INVoF’, Invitrogen) were transformed with the ligation reaction products also according to protocols supplied by the manufacturer. The transformed bacteria were subjected to blue-white selection on LB-agar plates supplemented with X-gal-kanamycin. The plasmid bearing the cDNA insert was named pCRECBGT- 1. Nucleotide sequencing ; DNA sequence analysis was performed by the “dideoxy” chain termination method (20) using the USB Sequenase Version 2.0 DNA sequencing kit. The cDNA was sequenced from both ends using 4 sets of primers - the Ml3 forward primer, 5’CAGGAAACAGCTA TGACCATG-3’; Ml3 reverse primer, 5’-GTMTCCCAGTCACGAC-3’; and the two primers used for the PCR. Sequencing reactions were followed exactly as described in Sequenase Version 2.0 protocol. Five pg plasmid DNA (-0.5 pmol) and 0.5 pmol of sequencing primer were used for each sequencing reaction.

Results and Discussion RT-PCR: Since the GalT-4 enzyme activity was abundant in 1l- to 19-day-old embryonic chicken brain (lo), RT-PCR was performed on total RNA isolated from fresh 1 l-day-old 1217

Vol.

189, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

b

a

kb 1 kb kb 2.3 kb 1.4 kb

kb

02 Figure 1. RT-#CR me&atedCamDplification of a pl-4 galactosyl-trausferasefrom embryonic chicken. Total RNA (1Opg)from 11-day-oldembryonic chicken brain (laneA) and a human colon carcinomacell line, Co10205 (laneB),was reversetranscribedand then subjectedto PCR, as describedunder Materials and Methods. Lane C, h-Hi&II markers, lane D, 0X- 174 HincJl markers. Aliquots of the reaction were run on a 1% agarosegel and stainedwith ethicliumbromide. Figure 2. a)Northem blot analysisof RNA isolatedfrom embryonic chicken brain The lane contained 1Opgtotal RNA isolatedfrom fresh 1l-day-old embryonic chicken brain, resolvedon formaldehyde-agarosegel (1%) and transferredto a membrane. The blot wasprobed with [32P]-labeledRT-PCR derivedcDNA. The positionsof the mRNA transcriptsare shown. b)Southem blot analysisof embryonic chickenbrain genomic DNA. Aliquots of 3Opg 1l-day-old embryonic chickenbrain genomicDNA were digestedwith EcoRI. The blot was probed with [32P]-labeledRT-PCRderived cDNA. The sizesof the hybridizing fragmentsare indicated.

embryonic chicken brains. The reverse transcription reaction was initiated with hexameric oligonucleotides since priming with oligo (dT) was found to be unsatisfactory for the subsequent PCR, consistent with other reports in the literature (21). One major cDNA fragment was amplified from both ECB RNA as well as human RNA (included as a positive control), and the amplified fragment was similar in size (-600 bp) from both the sources (Fig. 1). When the RT-PCR was applied to RNA isolated from different tissues of 1l-day-old embryonic chicken , the same major cDNA fragment was isolated from each tissue, indicating a similarity in the RNA transcripts (data not shown). RNA & DNA analvsis; Northern blots with 1 l-day-old embryonic chicken brain RNA and [32P] labeled cDNA probe generated from RT-PCR detected the presence of two major RNA transcripts 1.4 and 2.3 kb long (Fig.2a). This result is in contrast to the results obtained with mammalian RNAs and pl-4 galactosyltransferase cDNA probes where the transcripts were in the size range of 4.3-4.8 kb (12,13). Southern analysis of EcoRI digested ECB genomic DNA with the same cDNA probe identified multiple bands at 18.6, 12.9, 10.5, and 3.7 kb (Fig.2b), suggesting that either the avian gene is polyexonic and the exons reside on separate restriction fragments or that this galactosyltransferase is part of a multi-gene family. The possibility of chicken pl-4 galactosyltransferase being a polyexonic gene may be presently favored based on the observation 1218

Vol.

189,

No.

2,

-5 1 56 19 116 39 176 59 236 79 296 99 356 119 416 139 476 159 536 179

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

TGfA VINQAGDEEFGRAKLLNV G BAmAmTGU&?iC GFTEALMEYDYTCFVFSDVD CPGATCCAAn;An;ACPGAC L I Q* *QTLQCTAQGRPDPNG TGAcAGGAAcAccAC?&XXT~TiXATAA *QEHLQVLQPTKAPFCLHG* ATI'XA~'IUXXA~lT IGFGYPTISIWSVSAEQRTI c?XXWG+~BP~~TC HEDHGFPNNYWGWGGEDDDI TACAA~~ATCXXCOB'IUXXCA~ YNRLVFKGMGISRPDAVIGK !t'XAG&mmP CRMIRHSRTRKNEPNPDRFD C T RI AHTRETMSSDGLKS TGcM;GIy3TGAcFM;TIux CWTDRFPLITKSQWTS

AGIaYlm-m

L TYG

55 la 115 33 175 53 235 78 295 93 355 118 415 138 475 158 535 178 590 194

Figure 3.

Sequence information on the RT-FCR amplified cDNA fragment from embryonicchickenbrain. The nucleotidesequenceand the predictedamino acid sequence of chicken pl-4 galactosyltransferase(partial) are shown.The position of the PCR primers hasbeen underlined.

that the genomic organization of the murine pl-4 galactosyltransferase does indeed demonstrate a coding sequence that is distributed into six exons spanning 50 kilobases. Comoarison of seouencw Using the Sequence Analysis Software package of the Genetics Computer Group (GCG)and the algorithm of Pearson and Lipman (22), the deduced cDNA sequence (Fig.3) from ECB showed maximal similarity with other glycosyltransferase cDNA sequences, especially those for @l-4 galactosyltransferases. The extent of the similarity was analyzed in detail by comparing the ECB cDNA sequence with human pl-4 galactosyltransferase cDNA sequence (GenBank accession no. M13701) using the ‘local homology’ algorithm of Smith and Waterman (23). The overall nucleic acid sequence similarity between the avian and human sequences, based on this algorithm, was 74% (Fig.4). When the cDNA nucleotide sequences were translated into amino acid sequences and compared to one another using the same algorithm (23), the overall similarity was 75%, and the overall identity was 61% (Fig.5). This match was restricted to only one open-reading frame for the chicken nucleotide sequence. The predicted amino acid sequence of the embryonic chicken brain GalT-4 fragment was compared to the bovine,murine and human pl-4 galactosyltransferase amino acid sequences to establish the conservation between the potentially important amino acid residues. Based on results obtained from chemical protection studies (24,25), epitope mapping with monoclonal antibodies (26), and mutagenesis studies on the cloned cDNA for the human enzyme (26),it appears that the cDNA fragment amplified from the embryonic chicken brain would code for part of the protein that contains the UDP-galactose binding site and probably the N-acetylglucosamine binding site. The 1219

Vol.

189, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

1 ggagacgaagaatttggccgtgctaaactgctgaatgtaggattcacgga 241

IIIIII

III

IIIIlIII

II

50

II

IIIII

II

II

II

GGAGACACTATATTCAATCGTGCTAAGCTCCTCAATGTTGGCTTTCAAGA

290

51 agctttgatggagtatgactatgactgctttgtgtttagtgatgtagacc

III

III1

III

IIIIIIII

100

IIIIIIIIIIIIIIIIII

291

AGCCTTGAAGGACTATGACTACACCTGCT’~TGTGTTTAGTGA

101

tgatccaatgatgacagacactacagtgcacagcacagggtagacctgat

150

II

333

. . . . . . . . . . . . . . ..*.......*..**......... .

151

cccaatggatgacaggaacacctacaagtgctacagccaaccaaggcacc

345

IIIIIII

lllll

249

II

III

II

I IIIIIIII

II

I

248

II

I

IIII

II

I

IIIIII

II

II

III

IIII

II

I

294

gttcccaaacaattactggggctggggaggcgaagatgatgacatctac~

II

II

II

IIIII

IIIIIIIIIIIIII

ATTTCCTAATAATTATTGGGGCTGGGGAGGAGAAGATGATGACATTTTTA

344

acaggctgg;gttcaaaggcatgggcata;ctcggccagatgctgtcat~

I II

645 494

II

I IIIII

494

I

IIIIIIII

II

I IIIIIIIIIII

III

I II

III’II

Illll

lllllll

544

I

II

II

II

I IIIIII

III

I IIlllIIII

II

II

ATGGTTTGAACTCACTCACCTACCAGGTGCTGGATGTACAGAGATACCCA

594 . 443

IIIII

644 493

IIIII

TCCTCAGAGGTTTGACCGAATTGCACACACAAAGGAGGAGACAATGCTCTCTG . . . atggtttgaagtcactcaccta.cgggtgctggactgac..aggttccct

IIIIIIIIIII

I

393

IlllIlll

GGGAGGTGTCGCATGATCCGCCACTCAAGAGACAAAAAAAATGAACCCAA . . . cccggacaggtttgaccgtattgctcacaccagggagacgatgagctct~

IIIIIIIIII

695

I IIIIIII

293

IIII

ACAGATTAGTTTTTAGAGGCATGTCTATATCTCGCCCAAATGCTGTGGTC . . . gggaaatgcagaatgattcgccactcgcgtgatcggaagaacgagcccaa

IIII

444

II

444

343

IIIIIIIIIIIIII

495

III1

IIIIIII

III

TTTGGAGGTGTCTCTGCTCTAAGTAAACAACAGTTTCTAACCATCAATGG

595

394

.

445

394

Illll

.

.

344 200

I

TTTCCGTTGCAATGGATAAGTTTGGATTCAGCCTACCTTATGTTCAGTAT . . ttggagtgtgtct... .gctgagcaaagaacaattcacgaagatc.atgg

II

545

II

tttctgtctccatggataaattggattt..ggttaccctacaatcagtat

IIII

395

III

CCCAATGAATGACCATAATGCGTACAGGTGTTTTTCACAGCCACGGCACA

. 201

IIIII

CGTGGACCTCAT

I IIII

II

. . . . . . . . 332

694 . 540

I III

744

. 541 745

ctg.atcacaaaatcacag;ggacatcggg

II

II

569

I IIIIIIIIIIIIIIIIIIII

TTGTATACCCAAATCACAGTGGACATCGGG

774

Figure 4. Alignment of the nucleotide sequences of chicken and human pl-4 galactosyltransferases within their region of sequence homology. The upper panel refers to the nucleotide sequence for the chicken cDNA, and the lower panel corresponds to the sequence of the human counterpart.

cr-lactalbumin

binding

site seems to be coded by sequences upstream of the amplified

cDNA

fragment. mammalian organization

To our knowledge, pCRECBGT-1 homolog of pl-4 galactosyltransferase in ECB DNA.

Isolation

comprises the first cDNA clone of a nonwhich can be used to elucidate the GalT-4

of full-length 1220

chicken

cDNA

clones and a detailed

gene

Vol.

189,

No.

2,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

1 GDEEFGRAKiLN"GFTEALtkYDYDCFVF~DVDLIQ**Q~LQCTAQGRP~ I I * I.IIIIIIIII III.:III.IIIIIIIIII 81 GDTIFNRAKLLNVGFQEALKDYDYTCFVFSDVDLI...............

50 115

51 PNG*QE"LQ;LQPTKAPFC~HG*IGFGYP~ISIW.SVSA:EQRTIHEDH~ 98 1 . :: . . . . ..: :ll:.I : .: :lll I 116 P"NDHNAYRCFSQPRHIS;KFGFSLPYVQYFGGVSAL&iQ;L& 165 99 FPNNYWGWGiEDDDIYNRL;FKGMGISRP~AV,IGKCRMIiH~RDRKNEPi 148 IIIIIIIIIIIIIII:IIIll:II:IIII:II:I:IIlllllll:IIIII 166 FPNNYWGWGGEDDDIFNRLVFRGMSISRPNAVVGRCRMIRHSRDKKNEPN 215 . . 149 PDRFDRIAHTRETMSSDGLKSLTYGCW,TDRFPLIT~ 184 I:IIIIIIII:III llll.llll..: .:l:II.I. 216 PQRFDRIAHTKETMLSDGLNSLTYQVLDVQRYPLYTQ252 Figure 5. Alignment of the predictedamino acid sequencesof chickenand human PI-4 galactosyltransferases in their region of homology. The upper panel refers to the predictedamino acid sequencefrom chickencDNA and the lower panel correspondsto that from human cDNA.

comparison of mammalian and chicken pl-4 galactosyltransferase gene expression are under

further study and may provide useful insight into the evolutionary processes which shape the mechanism of gene expression in different species. ACKNOWLEDGMENTS: Supported by NIH grant NS-18005 and grant-in-aid from Cancer Society of St. Joseph County and The Coleman Foundation, Chicago. The authors gratefully acknowledge the gift sample of bovine cDNA probe received from Professor Joel H. Shaper.

References

10. 11. 12. 13. 14. 15. 16.

Feizi, T. (1985) Nature, 314,53-57. Hakomori, S. (1985) Cancer Res., 45,2405-2414. Hakomori, S. (1990) J.Biol.Chem., 265, 18713-18716. Walz, E., Aruffo, A., Kolanus, W., Bevilacqua, M., and Seed, B. (1990) Science, 250, 1132-1135. Hakomori, S., and Strycharz, G.D. (1968) Biochemistry, 7, 1279-1286. Basu, M., and Basu, S. (1972) J. Bioi. Chem., 247, 1489-1495. Kapadia, A., Feizi, T., and Evans, M.J. (1981) Exp.Cell Res., 131, 185-195. Basu, S., and Basu, M. (1982) The Glycoconjugates, 3,265-285. Basu, S., Basu, M., Chien, J.L., and Presper, K.A. (1980) in Structure and Function of Gangliosides (L. Svennerholm, H. Dreyfus, and P-F. Urban. Eds.), pp. 213-226. Plenum Press, New York. Basu, M., Weng, S.A., Das, K.K., and Zhang, B.J. (1989) Glycoconjugate J., 6, 372. Ghosh, S. (1992) Studies on the Regulation of Two Galactosyitransferases from Embryonic Chicken Brain. Ph.D thesis, University of Notre Dame, Notre Dame, IN. Shaper, N.L., Shaper, J.H., Meuth, J.L., Fox, L., Chang, H., Kirsch, I.R., and Hollis, G.F. (1986) Proc. Natl. Acad. Sci., USA, 83, 1573-1577. Russo, R.N., Shaper, N.L., and Shaper, J.H. (1990) J. Biol. Chem., 265, 3324-3331. Nakazawa, K., Ando, T., Kimura, T., and Narimatsu, H. (1988) J. Biochem.(Tokyo), 104, 165-168. Appert, H.E., Rutherford, T.J., Tarr, G.E., Wiest, F.S., Thornford, N.R., and McCorquodale, D.J. (1986) B&hem. Biophys. Res. Comm., 139,163- 168. Devereux, J., Haeberli, P., and Smithies, 0. (1984) Nut. Acid. Res., 12, 387-395. 1221

Vol.

189, No. 2, 1992

17. 18. :::

Sk Xi: 25. 26.

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Wang, A.M., Doyle, M.V., and Mark, D.F. (1989) Proc. Natl. Acad. Sci., USA, 86,9717-9721. Feinberg, A.P., and Vogelstein, B. (1983) Anal. B&hem., 132,6-13. Reed, K.C., and Mann, D.A. (1985) Nucleic Acids Res., 13,7207-7221. Sanger, F., Nicklen, S., and Coulson, A.R. (1977) Proc. Natl. Acad. Sci., USA, 74, 5463-5467. Noonan, K.E., and Roninson, LB. (1988) Nut. Acids Res., 16, 10366. Pearson, W.R., and Lipman, D.J. (1988) Proc. Natl. Acad. Sci., USA, 85,24442448. Smith, T.F., and Waterman, M.S. (1981) Adv. Appl. Math, 2,482-489. Takase, K., and Ebner, K.E. (1981) J. Biol. Chem., 256,7269-7276. Yadav, S., and Brew, K. (1990) J. Biol. Chem., 265, 14163-14169. * Ulrich, J.T., Schenck, J.R., Rittenhouse, H.G., Shaper, N.L., and Shaper, J.H. (1986) J. Biol. Chem., 261, 7975-7981.

1222