’mer-remwam‘mr ,- ~ - nggns i R . E r.,. fi 4 . t- ‘. .- . r .. {Kg-3d '- ' " t f‘ . " ‘5 ‘ .. .5 -", '. .- 5v 1. , l v c . "EL'Kurxzrr~~ - v This is to certify that the thesis entitled Studies of the Functional Role and Partial Charac- terization of a UDP-Galactose Glycoprotein Galactosyltransferase presented by CHRISTOPHER CHARLES MARVEL has been accepted towards fulfillment of the requirements for PhD. degree in Biochemistry 6/ 6 {mew at W71 Date—41W 0-7639 MSU LIBRARIES u \— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. STUDIES OF THE FUNCTIONAL ROLE AND PARTIAL CHARACTERIZATION OF A UDP-GALACTOSE:GLYCOPROTEIN GALACTOSYLTRANSFERASE By Christopher Charles Marvel A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1980 cm IaOOa? ABSTRACT STUDIES OF THE FUNCTIONAL ROLE AND PARTIAL CHARACTERIZATION OF A UDP-GALACTOSE: GLYCOPROTEIN GALACTOSYLTRANSFERASE BY Christopher Charles Marvel A UDP-galactose: glyc0protein galactosyltransferase, utilizing desialylzed degalactosylated fetuin as an exogenous acceptor, was partially characterized in the rat pancreas. The enzyme activity was dependent upon manganese and detergent (Triton x~1ao ). No evidence for a lipid intermediate between the nucleotide sugar and the acceptor protein was detected. Galactosyltransferase activity was shown to be localized almost exclusively in the smooth microsomal fraction, with a 3% fold enrichment in specific activity over the crude homogenate. No detectable galactosyltransferase activity was found in purified zymogen granule membranes. Galactosyltransferase activity was monitored during the embryonic and neonatal development of the rat pancreas. From day 14 of embryonic development to parturition specific activity declined from 16 to 4 nmoles/mg protein/hr. A large increase (6 fold) in galactosyltransferase activity CHRISTOPHER CHARLES MARVEL was observed in day 3 and d neonates but activity declined to adult levels by day 7. Pyrophosphatase activity rose concomitantly with galactosyltransferase activity during the neonatal period. The neonatal, adult, and embryonic galac- tosyltransferases could not be distinguished by polyacrylamide gel electrophoresis, isoelectric focusing, or molecular sieving chromatography. Galactosyltransferase 'activity was characterized in several cultured cell lines (Nil-8, Nil-BHSV, CHO, V-79, and KB) and found to have similar if not identical properties to the rat pancreatic enzyme» When cells in culture were exposed to phorbol esters, butyrate, or retinoic acid, alterations in cellular morphology were observed. Morphological changes in these cell linesmay be analogous to changes occurring during development, therefore the effects of these agents were further characterized. Phorbol esters consistently elevated galactosyltransferase activity in Nil-888V and KB cell lines while having little effect on the Nil-8 cell line.‘ This elevation was independent of cell density and was not related to the stage of the cell cycle. CHO cells and a cell surface mutant cell line (CHO-M) resistant to the cytotoxicity of WGA differed markedly in their response‘ to "chemical agents. CHO~M cells exhibited elevated galactosyltransferase activity for all tumor promotors tested. CHO cells were less sensitive. In CHRISTOPHER CHARLES MARVEL contrast, butyric acid treatment significantly increased galactosyltransferase activity in both CH0 and CHO-M cells. Most tumor promotor induced alterations in galactosyl- transferase activity correlated with endogenous membrane phosphorylation in CHO and CHO-M cells. CHO cells treated with phorbol esters tumor promoters responded with little change in either endogenous phosphorylation or galactosyl- transferase activity. However, CHO~M cells exhibited a significant increase in galactosyltransferase activity and a decrease in endogenous phosphorylation. ‘ A possible relationship has thus been shown between cell surface glycoconjugates, phosphorylation and galactosyltransferase activity. ACKNOWLEDGEMENTS Completion of this dissertation has been possible only through the assistance and support of many individuals over the years. A special appreciation to my earliest mentor, Dr. Francis X. Fellers, for introducing me to biochemistry and allowing me the freedom to explore it, and whose influence will be felt thoughout my scientific career. I wish to express my sincere gratitude to Professor's Robert A. Ronzio and Charles C. Sweeley for their guidance and financial support during the course of my graduate work. I want to express my appreciation to my many co-workers whose interactions were invaluable inside and outside the laboratory. A special thanks to Dr. Joseph Moskal, whose energy and enthusiasm led to many enriching experiences in and out of science and with whom "gonzo biochemistry" took on real meaning. I wish to express my appreciation to Sally Camper whose assistance with the figures allowed me to meet several deadlines. Finally I wish to thank my wife Pamela without whose encouragement and patience this dissertation could not have been written. Her understanding helped me to get through the roughest times. ii TABLE OF CONTENTS . Page LIST OF FIGURES ....................................... x LIST or TABLES ....................................... . x111 LIST OF ABBREVIATIONS ................................. xiv INTRODUCTION p .......................................... 1 REVIEW OF THE LITERATURE ................ . ........... . . 3 Overview of Glycoconjugates ......................... 3 Roles of the Oligosaccharide Moiety ................. 4 Protection from Proteolytic Degradation .......... 5 Recognition Signal ............................. .. 5 Binding Sites for Viruses ................... ..... 6 Blood Group Substances ..... ...................... 7 Intracellular Transport ..... . ...... .............. 7 Antifreeze Glyc0proteins ...... .......... ....... .. 8 Miscellaneous Roles ...... . .................... ... 8 Biosynthesis of Glycoconjugates .... ............. .... 9 Glycolipid Biosynthesis .... ....... ............... ll Glycoprotein Biosynthesis ...... . ........ ......... ll Synthesis of O-glycosidic Oligosaccharides .... 12 Synthesis of N-glycosidic Asparagine- 1inked Oligosaccharides .......... ....... 12 Inner-Core Glycosylation ....... ° ....... ..... l3 Processing Reactions ....................... 15 Terminal Glycosylation ... .................. l9 Alterations of Glycoconjugates in Transformation .... 20 iii Regulation of Glycoconjugate Synthesis by Glycosyltransferases ............ . ............. 21 Terminal Glycosylation as a Regulatory Point ..... 22 Diversity of Terminal Structures ......... ..... 22 Alterations of Glycosyltransferases upon Cellular Transformation ............................. 25 Cell Lines as Models .......................... 25 Glycolipid Glycosyltransferases ............ 26 Glycoprotein Glycosyltransferases .......... 27 Cancer Associated Alterations in Glycosyltransferase Activity ............ 28 Alterations of Glycosyltransferases in Development 30 Rat Pancreas as a Developmental Model ......... 33 Embryonic Differentiation .................. 34 Glycosyltransferases as a Regulatory Mechanism in Protein secretion 0.0.0.0.........OOOOOOOOOO 36 Rat Pancreas as a Secretory Model ............. 37 secretory Process OOOOOOOOOOOOOOOOOO0.00.... 38 Alterations of Glycosyltransferases by Chemical Agents OOOOOOOOOOOO...OOOOIOOOOOOQIIOOIOOOOO 43 Butyrate .................. ...... . ......... .... 43 Phorbol Ester Tumor Promoters ................. 44 Miscellaneous Agents .......................... 45 MATERIALS AND METHODS ............. ....... .i........... 46 MATERIALS 1""""""°"‘°"°"'°"‘°°"'°"“""' 46 Electrophoresis Reagents ......................... 46 Radiochemicals ..................... ............ .. 46 iv Liquid Scintillation Counting ............. . ...... Tissue Sources ....... _ ............................ Cell Lines . .................................... .. Tissue Culture ........................ ..... ...... Miscellaneous ....................... ............. METHODS ...................... Rat Pancreatic Tissue as a Galactosyltransferase source ......OOOOOOOOO Dissection and Homogenization of Adult Rat Pancreas ......OOOOOOOOOOOO......OOOOOOOO Dissection and Homogenization of Embryonic and Neonatal Rat Pancreas ................... Subcellular Fractionation of Rat Pancreatic Tissue ..... ...... Cultured Cells as a Galactosyltransferase Source . Growth conditions ......OOOOOOOOOO‘O0.0.0.0....- Addition of Chemical Effectors ................ Synchronization of Growth ...... ........ . ..... . DNA Labeling ....... O I O O O O O O O OOOOOOOOOOOOOOOOO O Harvesting and Homogenization ........ ........ . Galactosyfitransferase Assays ..................... Assays with Exogenous Protein Substrates ...... Assays Involving Endogenous Lipid Acceptors ... Assays with Free Glucose and Free N~acetylglucosamine as Substrates ....... Preparation of Desialylzed and Degalactosylated Protein Acceptors .. 49 49 SD 51 52 53 53 S4 54 55 55 56 56 57 Physical Characterization of Microsomal Galactosyltransferase Activity ............. 58 Isoelectric Focusing .. ....... ... .............. '58 SDS Polyacrylamide Gel Electrophoresis ....... . 59 Molecular Exclusion Chromatography ... ........ . 6D Pyrophosphatase Assays ............... . ........ ... 61. Phosphorylation Assays .......... . ................ 61 Purification of Membrane Fractions ....... ..... 61 Assay Conditions ............ . ................ . 62 Miscellaneous Methods ....J ................... .... 63 Paper Chromatography .......................... 63 Lowry Protein Assay ........................... 63 Purification of Alpha Lactalbumin ............. 63 Scintillation Counting ... ......... .... ....... . 64 Autoradiography ............. .......... .. ..... . 64 Electron Microsc0py ..... . ......... .........:.. 64 RESULTS . ....... 65 Requirements of UDP-Galactosyltransferase from Pancreatic Rat Microsomes ........... ....... 65 Cation REQUirementS oooooooo00000000066060.0000 65 Effect of Detergent and Dithiothreitol ........ 65 Effect of Enzyme Concentration, Incubation Time, and pH ......IOOOOOOOOOOOOO......OOOOOOOOOOIO 71 Acceptor Specificity .......................... 78 Exogenous Acceptor Concentration ............. . 73 Stability .............. . ..... . ........ ... ..... 84 vi Identification of Labeled Galactose Incorporated into Protein ............................ ... 84 Subcellular Localization of Galactosyltransferase Activity .. ...... .... ................ - ....... 84 Molecular Exclusion Chromatography of Solubilized Galactosyltransferase ..... ................. 89 Polyacrylamide Gels of Solubilized Galactosyltransferase ...................... 89 Isoelectric Focusing of Solubilized Galactosyltransferase ..... ................. 92 Pyrophosphate Activity in Microsomal Preparations 92 Galactosyltransferase Levels in Rat Embryo Pancreatic Homogenates ..................... 105 Galactosyltransferase Levels in Rat Neonate Pancreatic Homogenates .............. ....... 108 Pyrophosphatase Activity in the Neonate Pancreas ................ ....... . ........ ll] UDP~Galactosy1transferase Activity Assayed in Cultured Cells ...... ..... .......... ........... ll4 Assay Requirements‘ .............. ......... ........ 114 Cell Cycle Dependence ............... ............. 116 Monosaccharides as Exogenous Acceptors .... ....... llG Effect of Alpha-Lactalbumin on Acceptor Specificity ..................... ........... 122 Effect of Phorbol Ester Tumor Promoter on Galactosyltransferase Activity ................ 122 Nil-8 Cells ....' ....... . .............. . ........... . 122 Nil~8HSV C8118 oooooo 0.. oooooooo o. oooooooooo out... 128 KBcells ou‘ooooooooooooooocoo-...... ......... no... 128 Effect of Cell Density on Tumor Promoter Alterations in Glycoprotein vii Galactosyltransferase Activity ............. 133 Effect of Tumor Promoters on Glycoprotein Galactosyltransferase Activity in CH0 and a Membrane Mutant CHO Cell Lines .......... l33 Graded Phorbol Ester Series ...... ................ 138 Phosphorylation Assays in CH0 and CHO-M Plasma Membrane Fractions ..... . ............ . ......... 138 Purification of Plasma Membrane Fractions ........ 133 Assay System .......... . ......................... . 138 Endogenous Phosphorylation ...... .. ........... .... 147 CHO Plasma Membrane Phosphorylation ...... ..... 147 CHO~M Plasma Membrane Phosphorylation ...... ... l47 KB Plasma Membrane Phosphorylation .... ........ 147 Effect of Butyrate on Endogenous Membrane Phosphorylation ................... ...... T47 Autoradiograms of Assay Mixtures Run on Polyacrylamide Gels ... ..... ........ ........ 156 DISCUSSION ....................... ..... ..... ........... 163 Characterization of a DSG-Fetuin Glyc0protein Galactosyltransferase in the Rat Pancreas ..... 163 Adult Pancreas .. ....................... .. ........ 163 Embryonic Rat Pancreas ................... . ....... 172 Neonatal Rat Pancreas ..... .......... ............. 174 Galactosyltransferase Activity in Cultured Cell Systems ... ..... .............., ............ 175 Nil-8 and Nil-BHSV Cells ................. . ....... 176 CHO and CHO-M Cells ...... ....... ................. 177 Endogenous Phosphorylation Studies in CEO and CHO-M Cells .................. 179 viii KB Cells ......................................... 181 Autoradiography . ..... _ ............................ 181 Cell Surface Localization of Galactosyltransferase ...................... 182 Conclusions .............................. . ........... 184 Prospects - .......................................... 184 BIBLIOGRAPHY ........................ ‘ .................. 186 ix Figure LIST OF FIGURES Biosynthesis of Asparagine-Linked Complex- Type Carbohydrates Part 1 Biosynthesis of Asparagine-Linked Complex- Type Carbohydrates Part 2 Transmission Electron Micrograph of a Dispersed Pancreatic Exocrine Cell UDP-Galactose Glyc0protein: Galactosyl- transferase and UDP~Ga1actose Dyrophdsphatase Reactions The Effect of Magnesium Chloride Concentration on the Incorporation of [Carbon-14lGalactose into . Desialylzed and Degalactosylated Fetuin 10. 11. 12. 13. The Effect of pH on the Incorporation of [Carbon-141Galactose into Desialylzed and Degalactosylated Fetuin Time Course for Incorporation of [Carbon-14] Galactose into Acid Insoluble and Chloroform- Methanol (2:1) Soluble Material ‘ Linearity of Galactosyltransferase Assay with Respect to Enzyme Concentration Electrophoretic Analysis of Fetuin (GIBCO) and Bovine Submaxillary Mucin (Sigma) The Effect of DSG-Fetuin Concentration on Galactosyltransferase Assay Linearity Sub-Cellular Fractionation of Rat Pancreatic Tissue Recovery of Galactosyltransferase Activity from Polyacrylamide Gels Isoelectric Focusing of Solubilized Galactosyltransferase Page 15 17 4D 67 78 73 75 77 81 83 86 91 94 14. 15. .16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Radiochromatographic Scans for the UDG-Galactose Pyrophosphatase Assay ~ Controls Radiochromatographic Scans of the UDP~Ga1actose Pyrophosphatase Assay - Time Course for Enzymatic Hydrolysis of UDP-Gal Radiochromatographic Scans for the UDP~Ga1actose Pyrophosphatase Assay ~ Effect of AMP and Excess Unlabeled UDP-Gal on Pyrophosphatase Activity Inhibition of Perphophatase Hydrolysis of UDP—Galactose by AMP and ATP Galactosyltransferase Activity in the Embryonic Rat Pancreas Glycoprotein Galactosyltransferase Activity in the Embryonic and Neonatal Rat Pancreas Glycoprotein Galactosyltransferase Activity in Sub-Cellular Fractions of the Late Embryonic and Neonatal Rat Pancreas Glycoprotein Galactosyltransferase Activity in Synchronized KB Cells Lactose Synthetase (UDP-Galactose: D~Glucose 1-galactosy1transferase) Structures of Phorbol EsterTumor Promoters 96 99 102 104 107 110 113 118 120 124 The Effect of TPA and Retinoic Acid on Glycoprotein Galactosyltransferase Activity in Nil-8 Cells 127 The Effect of TPA and Retinoic Acid on Glycoprotein Galactosyltransferase Activity in Nil~BHSV Cells The Effect of TPA and Retinioc Acid on Glycoprotein Galactosyltransferase Activity in KB Cells ‘ The Effect of Cell Density on Tumor Promoter Induced Alterations in a Glycoprotein Galactosyltransferase The Effect of TPA, Retinoate, and Butyric Acid on a Glyc0protein Galactosyltransferase Activity in CH0 and xi 130 132 135 29. 30. 31. 32. 33. 34. 35. 35. 37. 38. 39. CHO-M Cells The Effect of Phorbol Ester Tumor Promoters on a Glycoprotein Galactosyltransferase Activity in CHO and CHO-M Cells Purification of a Plasma Membrane Fraction from Cultured Cells The Rate of Endogenous Phosphorylation as a Function of Temperature Linearity of the Phosphorylation Reaction with Respect to Enzyme Concentration Alteration of Endogenous Phosphorylation in CHO Plasma Membrane Preparatidhs by Phorbol Ester Tumor Promoters Alteration of Endogenous Phosphorylation in CHO-M Plasma Membrane Preparations by Phorbol Ester Tumor Promoters Alteration of Endogenous Phosphorylation in KB Plasma Membrane Preparations by Phorbol Ester Tumor Promoters Alteration of Endogenous Phosphorylation in CHO, CHO~M, and KB Plasma Membrane Preparations by Butyrate SDS-Polyacrylamide Gel of CHO and CHO-M Cell Homogenates Exposed to Tumor Promoters and Anti-Promoters Autorads of Endogenous Phosphorylation Assays ~ CHO Cells Autorads of Endogenous Phosphorylation Assays — CHO-M Cells xii 137 140 142 144 146 149 '151 153 155 158 160 162 LIST OF TABLES Table - ' Page 1. Requirements for the Incorporation of [Carbonel4] Galactose into Desialylated Degalactosylated Fetuin by Pancreatic Microsomes 68 2. Protein Substrate Specificity of Pancreatic Microsomal Galactoslyltransferase 79 3. Distribution of DSG-Fetuin: Galactosyltransferase in Pancreatic Subcellular Fractions 88 4. Effect of AMP and ATP on Pyrophosphatase Activity 99 5. Effect of AMP and ATP on Gala¢tosy1transferase Activity 100 6. Hydrolysis of UDG-Galactose by Neonatal Rat Pancreatic Microsomes . 115 7. Galactosyltransferase Activity with Glucose and N~Acetylglucosamine as Acceptors in the Presence and Absence of A1pha~Lacta1bumin 121 8. Phorbol Ester Tumor Promoters 125 xiii ADP Ala .AMP ATP BBpr BrdUrd CDP DNA DSG EDTA EGF ER Fuc Gal GalNac GDP GIBCO GlcNac GMP HEPES Man MES LIST OF ABBREVIATIONS adenosine diphosphate alanine adenosine monophosphate adenosine triphosphate polyoma virus transformed baby hamster kidney cells 5~bromodeoxyuridine cytidine diphosphate deoxyribonucleic acid desialized, degalactosylated ethylenediamine tetraacetate epidermal growth factor endOplasmic reticulum fucose galactose N-acetylgalactosamine guanosine diphosphate Grand Island Biological Company N-acetylglucosamine guanosine monophosphate N-2 hydroxyethyl piperazine-N’-2-ethanesulfonic acid mannose 2-(N~morpholino) ethane sulfonic acid xiv NANA NeuAc PAS PBS PDA PDB PDD PMML PTA RER RNA SDS TCA Thr TPA UDP UDP~Glc Ve V0 WGA ZG Sialic acid N-acetyl neuraminic acid periodic acid-Schiff reagent phosphate buffered saline phorbol 12,13-diacetate phorbol 12,13~dibutyrate phorbol 12,13~didecanoate phorbol 12-myristate 13-acetate phosphotungstic acid rough endoplasmic reticulum ribonucleic acid sodium dodecyl sulfate trichloroacetic acid threonine phorbol 12-tetradecanoate 13-acetate uridine diphosphate uridine diphosphate glucose included volume void volume wheat germ agglutinin zymogen granule XV INTRODUCTION Research on complex -carbohydrates has undergone an immense resurgence in the last fifteen years. Carbohydrates .are no longer delegated simply to structural roles or as energy stores, but are now known to be intimately involved in multiple aspects of cellular existence. Complex carbohydrates have been implicated in such phenomena as tumorigenicity, growth control, intercellular adhesion, and morphology. The rapid progress in this field has been due to the introduction of new and improved materials and techniques that were unavailable to our predecessors. An outstanding current research area in complex carbohydrates is the integration of empirical biochemical data with their functional roles in the cell. This thesis describes the partial characterization of a membrane~bound UDP-galactose: glycoprotein galactosyltrans- ferase that is involved in the synthesis of glycoproteins. Utilizing this enzyme activity, a role for galactosyltrans- ferase activity has been examined in systems that have shOwn potential for involvement of a glycoprotein galactosyltrans- ferase in a defined phenomenon. The systems examined were: 1) Adult Rat Pancreas ~ The ‘role of galactosyl- transferases in the secretion process. 2) Embryonic Rat Pancreas ~ The role of galacto- syltransferase in a developing tissue. 3) Normal and Transformed Cell Lines - Alterations of galactosyltransferase activity in cultured cells upon transformation. 1 2 4) Cell Surface Mutant Cells - Alterations Of galactosyltransferase activity in cell lines that are resistant to specific lectin toxicity. 5) Cultured Cells ~ Alterations of galactosyl- transferase activity upon exposure of cultured cells to tumor promotors and anti-promoters. 6) Cultured Cells ~ The relationship of galacto- syltransferase activity to endogenous phos- phorylation. The literature review discusses glycoconjugate function and biosynthesis, the general concepts Of each system, and the specific findings that have made each suitable as a model system in which to study galactosyltransferase activity. NO attempt has been made to. provide a comprehensive literature review for each individual system. REVIEW OF THE LITERATURE OVERVIEW OF GLYCOCONJUGATES Glycoconjugates comprise a diverse class of biological compounds which contain as a defining feature covalently attached carbohydrate. These carbohydrate moietes may be as Short as one unit or as large as 50 monosaccharides in some glycolipids. Of the biologically occurring macromolecules the glyco- proteins are the most diverse with the carbohydrate component ranging from 1 percent to over 80 percent of the total molecular weight. GlyCOproteins are composed of a polypeptide backbone to which oligosaccharide chains are covalently attached to specific amino acid residues. Oligosaccharide chains on glyCOproteins are classified according to their amino acid linkage as well as their inner-core structures. Classified separately from other glycoproteins are the proteoglycans which contain closely spaced heteropolysaccharide chains which may contain uronic acids and sulfated amino sugars. The linkage sugar in proteoglycans is commonly D-xylose attached to a seryl residue. The proteoglycans include the chondroitin sulfates, heparin, and the dermatan sulfates. Glycolipids are generally classified into two major divisions: the glycosphingolipids and the glycoglycer~ olipids. The glycosphingolipids are composed of a 4 sphingosine base, fatty acids, and carbohydrate while the glycoglycerolipids contain glycerol, fatty acids and carbohydrate. In many instances the inner oligosaccharide cores Of glycolipids are related and may be classified into 'families according to their tetrasaccharide core structures (1). Glycoconjugates have been implicated as being involved in a variety of cellular phenomena including: intercellular adhesion, transformation, morphology, contact inhibition, regulation of cell growth, antigenic determination, as receptors for virus and bacterial toxins, tumorigenicity, blood group substances, etc. Present thinking delegates these specific phenomena as being mediated more through glycoproteins and glycolipids than the mucopolysaccharides, lipopolysaccharides, and proteoglycans which probably play less specific and more general roles. Subsequent discussions will emphasize the roles and synthesis of glycoproteins over other glycoconjugates. ROLES OF THE OLIGOSACCHARIDE MOIETY 'The exact role of the carbohydrate moiety is still unclear in most instances. The complexity of the glycosyl- ation prOcess (and the considerable energy cost to the cell) argue for highly specific and important roles for the carbohydrate moiety. A Simple general answer for their role is probably not possible. An added complexity is that 5 segregation of the Oligosaccharide's role from other molecular structures is not always possible. For example, specificity df M or N blood group types appears to require both the presence of Sialic acid as well as an involvement ’of amino acid residues to distinguish the M type glyco- protein. from the N type glchprotein (2). Some current ideas on the roles of oligosaccharide moieties are presented below. Protection from Proteolytic Degradation Glycosylation has been shown to decrease the sensitivity of certain proteins to proteolytic degradation notably adrenocorticotropic hormone (ACTH), beta~1ipoprotein (3) and fibronectin (4). External beta-turns of polypeptides fare quite susceptible . to proteolytic degradation and a proposed general functiOn of glycosylation may be the masking of the turn configuration (4). In one case reported, dopamine beta~hydroxy1ase, the removal of the terminal Sialic acid residue alone resulted in increased proteolytic degradation (5). lRecognition Signal In some cases the oligosaccharide of a glycoprotein functions as a signal for the recognition and uptake of the glycoprotein by cells of the reticuloendothelial system. A generalized role for the terminal Sialic acid residues in circulating glycoproteins appears to be as a masking agent 6 for penultimate galactose residues. The work of Ashwell and his collaborators (reviewed in ref. 6) have demonstrated that terminal galactose residues, exposed by removal of Sialic acid, constitute a configuration recognizable by hepatic cells which bind and catabolize such glycoproteins. Most long 1iVed circulating mammalian plasma glycoproteins contain the complex type Of oligosaccharide moiety ending in Sialic acid. An analogous binding system has been found to operate in avian Species. Many avian circulating glyco~ proteins have a terminal galactose residue, and its removal, which exposes the penultimate N-acetylglucosamine, results in their rapid clearance from the circulatory System. The uptake of lysosomal glycoprotein glycosidases by cultured fibroblasts can be abolished by periodate oxidation of the glycosidases. This uptake can also be blocked by exogenous addition of a variety of carbohydrates, one of the most potent being mannose 6-phosphate. Other work has implicated a phosphorylated mannose-residue of the oligosac- charide chain as the recognition signal for' uptake of a variety of lysosomal enzymes (7). Binding Sites for Viruses_ Virus infection is initiated by the attachment of the virus to cell-surface receptors. Sendai and other myxo- and paramyxoviruses absorb to erythrocytes but their receptor activity is destroyed by sialidase (8). Holmgren et a1. (9) have recently determined that specific binding of the Sendai 7 virus occurs to gangliosides containing the terminal structure: NeuAc alpha 2~8-NeuAc alpha 2-3 Gal beta 1-3 GalNAc. It is quite conceivable that similar structures in glycoproteins also function as receptors. Blood Group Substances The blood group substances are immunochemical specific surface oligosaccharide antigens. The defining Oligosac- charides of the major blood groups are: Group A, GalNAc alpha 1~3[Fuc alpha l-2]~Gal beta 1-3 or 4 GlcNac; Group B, Gal alpha 1~3-[Fuc alpha 1-2]~Gal beta 1-3 or 4 GlaNac; Group H, Fuc alpha 1-2 Gal beta 1-3 or 4 GlcNac. These Oligosaccharides are found in saliva, gastric juice, etc. as well as on erythrocytes and may be present on glycosphingolipids or glycoproteins. Intracellular Transport Glycophorin A, the major integral sialoglycoprotein of erythrocytes, 'is incorporated normally into cell membranes when synthesis of its N-glycosidic Oligosaccharide is inhibited by tunicamycin (10). In addition to its one N-glycosidic oligosaccharide it contains 15~O~glycosidic Oligosaccharides 'whose synthesis is not inhibited by tunicamycin. Fibronectin, a major cell surface associated glycoprotein of fibroblasts, is normally transported to the cell surface in a nonglycosylated form (11). In contrast to glycophorin A fibronectin contains only aspargine linked Oligosaccharides. Studies with a viral system, vesicular .stomatitis virus, have shown that transfer of membrane glycoprotein from the rough to the smooth endoplasmic reticulum is blocked by tunicamycin (12). Any analogies to a mammalian system are not known at this time. Antifreeze Glycoproteins Specific .glyCOproteins in the blood of several antarctic fish function as antifreeze agents (13). Both the carbohydrate and peptide chain are necessary for antifreeze activity, the structure being a repeated tripeptide, Ala-Ala-Thr, with a disaccharide beta~galacto~ pyanosyl-(1-3)~N-acetyl-D-galactosamine attached to each threonine (14). The mechanism for lowering the freezing temperature may involve an interaction by the glycoprotein at the ice-water interface (15). Miscellaneous Roles Release of enveloped viruses is severely depressed when viral membrane protein glycosylation is inhibited by tunicamycin (16). Carbohydrates of immunoglobulin molecules have been shown to be required for complement-induced cytotoxicity (17). 9 The involvement of carbohydrate moieties in enzymatic activity is. varied. The enzymatic removal of Sialic acid from chorionic gonadotropin (18), follic1e~stimulating hormone (l9), and erythroprotein (20) leads to a loss of 'activity. However, incubation of pig kidney aminopeptidase with glycolytic enzymes led to no differences in catalytic activity or substrate specificity (21). Many physio-chemical properties of glycoproteins can be related to their carbohydrate moieties. The high degree of viscosity of mucous secretions can be attributed to the acylneuraminic acids of glyCOproteins (22). Similarly, solubility of many glycoconjugates is largely. attributable to their carbohydrate moieties. In one instance reported, removal of carbohydrate from glucamylase I (23) resulted in reduced cold stability. The oligosaccharide may in some instances affect the three~dimensiona1 structure of the glycoprotein, placing it into a preferred_configuration. BIOSYNTHESIS OF GLYCOCONJUGATES _Oligosaccharide chains are not primary gene products and the biosynthetic pathways for glycoconjugates proceed through a complex series of step—wise additions .of carbohydrate units regulated by the specificity of enzymes acting in concert. Although the multiplicity of oligosac~ charides contained in glycoconjugates is large, many share common structual features and similar biosynthetic pathways. 10 In nearly all.instances the sugars are transferred from the appropriate nucleotide derivative. These "activated" sugars are present as: uridine diphosphate glucose (UDP~G1u), uridine diphosphate galactose (UDP~Ga1), uridine diphosphate IN-acetylglucosamine (UDP~G1cNac),. uridine diphosphate N~acetylgalactosamine (UDP~GalNac), guanosine diphosphate fucose (GDP-Fuc), and cytidine monophosphate N-acetylneura- minic acid (CMP-NANA). The mechanism of regulation of the biosynthesis of glycoproteins is not clear. In some systems steroid hormones appear to exert a regulatory effect on the patterns of glycoproteins synthesized by tissue culture cells (24). In one example, following partial hepatectomy, injection of hydrocortisone led to a 3-fold elevation in liver sialyltransferase (25). No elevation was found in the corresponding enzyme in the serum. Other aspects that would contribute to regulation would be availability of nucleotide sugars and compartmentalization of the glycosyltransferases. In constrast to the precise enzymatic control that_ forms the classic glycoconjugates, certain proteins may undergo nonenzymatic- glycosylation. Hemoglobin may be glycosylated under physiologic conditiOns at specific protein sites (26). This has received interest, as a two- to three-fold increase in this glycosylated hemoglobin occurs in patients with diabetes mellitus (27). Other oligopeptides can be similarly nonenzymatically glycosylated (28). ll Glycolipid Biosynthesis The majOr glycolipids are derivatives of .the base sphingosine. Their‘ biosynthesis begins at the C-1 hydroxy -group of ceramide with the transfer of glucose or galactose from UDP~Glc or UDP~Gal. The carbohydrate chain is then extended by the step-wise addition of sugars from the appropiate nucleotide sugars without involvement of carrier lipids (29). The biosynthetic enzymes are membrane-bound, and in the liver are good markers for the Golgi apparatus (30). It must be emphasized that these addition reactions are enzymatically controlled through the Specificities of the transferases present. Although large numbers of carbohydrate moieties from glycolipids have been described structual similarities indicate the presence of defined carbohydrate sequences. Glycoprotein Biosynthesis The biosynthesis of carbohydrate moieties on proteins involves more than the sequential addition of sugars from their nucleotide derivatives. These additional biosynthetic reactions will be described in detail below. The common covalent linkages between carbohydrate and protein occurs with five amino acids. Two of them, 5-hydroxy-L-lysine and 4-hydroxy-L-proline, occur rarely. Three of the amino acids constitute the majority of the amino acid linkages with the biosynthesis of the carbohydrate chain being dependent upon 12 the amino acid present. Synthesis of O-Glycosidic Linked Oligosaccharides The O-glycosidic linkages in glyOOproteins are in the 'vast majority of cases to the amino acids serine or threonine. This bond is commonly characterized by its lability to mild alkali resulting in cleavage through beta-elimination. The biosynthetic route of serine and threonine linked carbohydrates is probably similar to the synthesis of glycolipids in that only sequential addition of sugars frOm their nucleotide derivatives is thought to occur (31). These Oligosaccharides have less complex structures than those found in N-glycosidic linkages. The major a1kali~labile oligosaccharide found in human erythrocyte membranes is: N-acetylneuraminyl-(2-3)~beta-D-ga1actopyra- nosyl-(l-B)-[N~acety1neuraminyl-(2-6)]~D-N~acetylgalactosa~ mine (32). Similar structures are found in erythrocytes of other species with some variance in sialylation but with the common core unit beta-D~galactopyranosyl (1~3)-D-N- acetylgalactosamine (33). In addition to erythrocytes identical Oligosaccharides are found in ‘fetuin (34) and canine submaxillary mucin (35). Synthesis of N-glycosidic Asparagine-Linked Oligosaccharides The N-glycosidically linked Oligosaccharides have complex synthetic routes which are described in detail below. l3 Inner-Core Glycosylation - The biosynthesis of the asparagine linked Oligosaccharides is initiated by the transfer of N-acetylglucosamine from UDP-GlcNac to the polyisoprenoid lipid dolichol phosphate. Sequential ”addition of sugars from their nucleotide derivatives to the N-acetylglucoSaminylpyrophosphoryldolicholi occurs to form the unit Man(beta 1-4)GlcNac(betal-4)G1cNac P-P~Dolichol. Additional mannose residues and three glucose residues are added to this structure to form the completed "inner core" dolichol linked intermediate. 4 This complete structure is shown at the top of Figure 1. The entire oligosaccharide moiety is then transferred from the dolichol pyrophosphate to an aSparagine residue on the polypeptide chain (36.37.38). This transfer is thought to occur before complete translation of the nascent chain has occurred (39). Studies of glycosylation have been greatly advanced by the discovery of a method to specifically inhibit N~glycosylation. Tunicamycin, an antibiotic from Streptolmyces lsosuperficis, specifically inhibits the N-glycosylation of asparagine residues by blocking the transfer of oligosaccharide from the dolichol intermediate to the protein core (40,41). Processing Reactions - The glucose containing asparagine linked oligosaccharide serves as the precursor for both the polymannose and complex type Oligosaccharide chains of glycoproteins (42). The asparagine-linked oligo- saccharide is now processed as outlined in Figure 1,2. The 14 Figure l. Biosynthesis of Asparagine-Linked Complex- Type Carbohydrates Part l The biosynthesis of an asparagine-linked oligosac- charide is shown. Synthesis is initiated by the transfer of a preformed oligosaccharide from a dolichol intermediate. Specific monosaccharides are then cleaved off and one N-acetylglucosamine residue is added. 15 Marry-J—z-Han 01.6 Man / 31.6 cl 2 / “1'3 \ m4 Man—Man Man—gain critic—.Gl cNAc-«P—P-Ool ichol :31 , J :Glu—lL‘zGlu id-Glu LL’N‘lan MMan‘m—Zonan/ Nascent Protein 3 6111 111.2 Man ——+ Man '1.“ Man a 1‘2 /.1." \‘9. Man Man ——- Mun 31.4 GlcNAc -o GlcNA( - Asn Man 419. Mm " L2- MnnA‘a \sam M“ \\ £1 4 Man “—4. GlcNAc ... GlcNAo -- Asn “HA; UDP-GlcNac M“ «1.6 / Mun GL3 CLO M“ an GlcNAc - GIcNAc -— Asn 16 Figure 2. Biosynthesis of Asparagine-Linked Complex- Type Carbohydrates Part 2 Processing of the core oligosaccharide continues with the hydrolysis of two mannose residues followed by addition of monosaccharides from their respective nucleo- tide sugars. The completed "complex type" oligosaccharide is at the bottom of the figure. 17 M“ a 1.6 1 Man 4 km A?" \ZMan 111.4 GlcNAc-o GlcNA¢ -— Asn GlcNAc—4 Mu .... Jr \ J 31.4 GlcNAc - GlcNAo _ Asn GlcNAo—. 54.34" 2 UDP-GlcNac GlcNAc \Bl.4 Bl.2/M\l6m B GlcNAc ——-cclcNA 1.4 -——-GlcNAc -. Ann al.3 #1 VM GlcNAc L,” 3 UDP-Gal on L GlcNAc 31.4 V 31.2 a! 6 /M\ 31.4 GAIL —-—-00|cNAc —-—¢GlcNAc —-—~ GlcNAc - Am «1.3 m yM on fl; GlcNAc GDP-rue 3 GNP-NANA 02.3 31 4 SA —-*Gnl —°GlcN:c\m; 31.4 ”JV \mm 314 SA—-‘ —3"Gll -—OG|cNAc ——+GlcNAc -—~ GlcNAc _. Asn “/34 3 [01.6 1.2 02.3 31.4 B /M F‘" SA -——-* Gal —* GlcNAc 18 three glucose units are sequentially removed with the first excision being rapid (1 or,2 min) and the second and third occurring at decreasing rates. The exact role of the three glucose units is not known; however, several explanations 'have been offered. Glucose removal could serve as a control point for the rate of glycosylation of nascent proteins (43). Glucose-free and glucose-containing oligosaccharide-lipids may serve as Oligosaccharide donors for different classes of Oligosaccharides to nascent proteins (44). After removal of the glucose residues four mannose units are excised followed by the addition of GlcNac to an alpha~1~3 linked mannose. The addition Of GlcNac is required before additional processing can occur. Subsequently two additional mannose residues are removed. At this point one or two additional GlcNac units are added to the alpha 6 linked mannose. This oligosaccharide is described as the “complex type" and may be further glycosylated in the smooth endoplasmic reticulum or the Golgi apparatus. A different type of oligosaccharide is formed if processing does not occur at the addition of the GlcNac to the alpha 3 mannose. These Oligosaccharides are termed "high mannose type" and vary in structure from the simple core sequence to more extended structures formed by the addition of more mannose units. The major envelope glycoprotein of murine leukemia virus (99 70) contains two classes of N~asparagine linked Oligosaccharides, high-mannose and complex (45). At least 19 in this example the viral protein structure is the primary factor in determining the mode of glchsylation. Terminal Glycosylation - Inner core glycosylation occurs in the rough endoplasmic reticulum (RER) concurrent 'with or soon after emergence of the nascent polypeptide chain into the endoplasmic reticular luminal spaCe. The polypeptide is then translocated through the endoplasmic reticulum (ER) to the Golgi apparatus where terminal glyco- sylation occurs (46). This glycosylation involves addition of N-acetylglucosamine, galactose, fucose, and Sialic acid residues. The final glycosylation steps involve the incorporation of fucose and Sialic acid residues. Current evidence suggests that the synthesis of terminal trisaccharide units (Sialic acid-galactose-N- acetyglucosamine) of glycoproteins occurs by the step-wise addition of monosaccharide units frOm the appropiate nucleotide sugars (47). The presence or absence of specific terminal carbohydrates or variations in their modes of attachment contributes to the structural heterogeneity found in glycoproteins. These terminal glycosyltransferase activities are characteristic of the microsomal pellet and the Golgi apparatus (48). Evidence also exists for a plasma membrane localization of terminal glchsyltranferase activity (49). .This localization is not without controversy (50) and the existence of cell surface glycosyltransferases is still in question. 20 The presence of the terminal carbohydrates of oligosac- ‘ride chains may have_ active as well as passive roles. Kreisil et a1. (51) has determined that the protein and carbohydrate moieties of rat membrane glyCOproteins turn over at different rates. The half-life of the protein is between 70 and' 80 hours while L~fucose is 12.5 hours, N~acetylneuraminic acid is 33 hours, and D-galactose is 20 hours. Consequently, not only the terminal L-fucose and N-acetylneuraminic acid, but even the penultimate galactose residues turn over faster than the protein core. This .implies a late processing step of uncertain nature. After terminal glycosylation is completed the glycoprotein may be secreted or utilized by the cell as a membrane glycoprotein. ALTERATIONS OF GLYCOCONJUGATES IN TRANSFORMATION A characteristic of a tumor cell is continued replication under conditions in which normal cell growth would be retarded. Therefore it is possible that the surface membrane of transformed cells has properties that are more characteristic of growing cells and may reflect a less developed state than that of terminally differentiated non-growing cells. Fetal cell surfaces have a glycoprotein composition markedly different from adult cells and closely resembling that of transformed cells (52). 21 Quantitative differences are found in both glyco- proteins and glycolipids. from 3T3 and SV-3T3 (3T3 cells transformed with Simian virus) cells (53), with 3T3 cells containing 3 to 5 times as much Sialic acid, N-acetylgluco- .samine, and N~acetylgalactosamine as SVe3T3 cells (54). Many membrane alterations in glycoconjugates of transformants of Balb 3T3 cells have been described, but the only consistent alteration has been an enrichment in high molecular weight sialylglyc0peptides (55). A review of glycopeptide changes upon malignant transformation has recently been published (56). REGULATION OF GLYCOCONJUGATE SYNTHESIS BY GLYCOSYLTRANSFERASES A clear focal point for anabolic regulation is with the biosynthetic ‘ enzymes themselves. With protein glycosylation, several points of regulation are possible. The O-glycosidic Oligosaccharides may be regulated simply by the availability of transferases and/or substrates. Regulation of N-glycosidic glycosylation has more possible control points. Alterations in the dolichol pyrophosphate intermediate may reSult in complete absence of N-glycosidic linked carbohydrates, while regulation at the processing steps will determine the type of oligosaccharide formed (complex vs. high mannose). Addition of terminal carbohydrates as an important control point has been demonstiated by the role of terminal sugars in blood group 22 specificity. Considerable variation in terminal glycosyl- ation has been shown to exist, as described below. The meaning of this diversity is not yet clear. Terminal Glycosylation as a Regulatory Point As mentioned previously, the nonreducing termini Of the Oligosaccharides of mammalian glyCOproteins contain N-acetyglucosamine, galactose, fucose, and Sialic acid. A multiplicity of linkages is possible and experimental evidence implies a distinct glycosyltransferase for each separate linkage. The structures which can be synthesized are dictated by the substrate Specificities of the individual glycosyltransferases which are present. It has been shown that a number of the terminal transferases are mutually exclusive, with product formation being determined by the order in which they act (57). Diversity of Terminal Structures The terminal carbohydrates of bovine cold insoluble globulin (CIG) have several structurally unique features compared to many published structures for asparagine linked carbohydrates (58). The occurrence of NeuAc alpha 2-4 Gal is unusual in that most of the complex type sugar chains reported Sialic acids are linked at the C-6 positions of the terminal galactoses. Human chorionic gonadotropin has [been reported to contain Sialic acid residues linked at the C-3 position of the penultimate galactose residue (59). This 23 diversity of the sialyl linkages on the complex type sugar chains may indicate that the specific terminal transferases act as regulatory signals. A Gal beta 1-3 GlcNac group is found in bovine CIG (58). This grouping has also been found Lin the sugar residues of bovine prothrombin (60) and the mucin type sugar chains of glycoproteins (61). Takasaki (62) has pointed out that all the N-acetylglucosamine residues of this grouping are sialylated at the C-6 position. A sequential relationship between the sialyl and galactosyltransferase probably exists. The linkages observed in mammalian glycoproteins require at least 5 sialyltransferases. and three fucosyltransferases (63). Oligosaccharide biosynthesis "in vivo" is regulated both by the glycosyltransferases available and the order in which they act (63). In the case of several sialyltransferases, structural features beyond the terminal N-acetylglucosamine units are recognized as preferred sialylation can occur on different branch chains (64). In the blood group substances the nature of the terminal saccharide is responsible .for the immunological specificity. For example, the difference between blood groups A and B resides in the terminal D-galactose or D-N-acetylgalactosamine, and thus the determinant of specificity will reside in the regulation of the appropriate glycosyltransferase used for the terminal carbohydrate residue. Regulation of oligosaccharide synthesis has been studied by using purified glycosyltransferases, identifying 24 the products formed, and studying how the prior action of a particular glycosyltransferase will inhibit the action of another (57). I Fibronectin contains 4-5 percent carbohydrate of the complex-type linked via asparagine residues. Differences in carbohydrate structure exist between species with Sialic acid linked to C3 of galactose in hamster fibroblasts (65), while the linkage is to the C4 or C6 of the penultimate galactose in human fibronectin (66). Different sugar chains of bovine fibronectin have been found to contain both beta l~3 and beta 1-4 linked galactose (67). Of at least 3 galactosyltransferases Orequired in mammalian glycoprotein synthesis two have been purified (both soluble): one from milk, beta-N-acetylglucosaminide beta 1~4 galactosyltransferase (68), and the other from human plasma, designated (fucosyl alpha‘ 1-2) galactoside alpha 1-3 galactosyltransferase Gal alpha 1-3 (Fucose alpha l-2) Gal beta (63). I Penultimate galactose residues are beta-linked in most glycoproteins. Examples of alpha~linked galactose are earthworm cuticle collagen (69) and the blood group substances (7D). The existence of alpha;linked galactose has also been demonstrated in bovine prothrombin, and interestingly, both anomeric forms of galactose are present on similar oligosaccharide chains (71). 25 The penultimate galactose of glycoproteins serves as a recognition signal in the clearance of serum proteins in mammals. In addition, the only tumor associated glycosyltransferase defined in tissue culture systems has .been a galactosyltransferase (72). For these reasons, galactosyltransferases, in constrast to other glycosyltrans- ferases, may deserve Special recognition. Alterations of Glycosyltransferases Upon CelIular Transtrmations Cell Lines as Models Cultured cell systems can be utilized as .manipulatable models for studies that would be difficult or impossible to do "in situ". These systems can be exposed to hormones or other chemical agents and their response to this singular agent can be studied without complications arising from other cell or tissue types. Cell systems can easily be transformed by mutagenic or viral agents and biochemical alterations can be more easily monitored in these ip XIEEQ models. In lectin-resistant cell lines, the cells have lost the ability to bind certain lectins and thus can withstand lectin toxicity. In some cases a specific glycosyltransferase responsible for a specific carbohydrate residue is deficient and altered surface glycoconjugates result. These lectin-selected mutants have been utilized to study the involvement of surface glycoconjugates in 26 recognition processes. A specific wheat germ agglutinin (WGA) lectin-resistant CHO cell line has been utilized in studies described under results. WGA binds in a specific fashion to certain GlcNac, 'GalNac, and NANA residues in glycoconjugates. Several classes of NANA residues are present at the CH0 cell surface, not all of which are involved in WGA binding (73). The particular cell line utilized in these studies possesses a defect in sialylation. of surface glycoproteins (Pamela Stanley - unpublished data). Unlike the parental cell line, this mutant line shows extensive labeling of surface glyco- protein by treatment with galactose oxidase: [tritiated]borohydride without prior treatment with neuraminidase. A similar cell line resistant to the cytotoxicity of the phytohemagglutinin from Phaseolus vulgaris has been shown to be .deficient in a specific glycoprotein N-acetylglucosaminyltransferase activity (74). Glycolipid Glycosyltransferases - When cells in culture are transformed by oncogenic viruses, alterations in glycolipid composition may occur at the level of the complex oligosac- charide chains. Reduced activity of N-acetylgalactosaminyl- transferase has been reported in Swiss 3T3 cells infected with Murine Sarcoma virus (75). Mouse cells transformed by the RNA virus Kirsten sarcoma have lost UDP-Gal: Gm2 galac- tosyltransferase activity (76). This same galactosyltrans- 27 ferase is blocked in BALE/3T3 mouse embryo cells treated by X~irradiation and . the chemical carcinogens methylcholanthrene and benzophyrene (77). In addition the low levels of galactosyltransferase activity in the .transformed cells had kinetic properties indistinguishable from those in normal cells (77). Other ganglioside glyco- syltransferases were unaffected, i. e. an N~acetylgalactosa- minyltransferase and a - sialyltransferase. Mixing experiments did not reveal the presence of any inhibitor Of galactosyltransferase activity in these transformed cells. Nucleotide pyrophosphatase activity was unchanged between normal and transformed cells in this Study. Other investigators have observed a decrease in sugar nucleotide pyrophosphatase activity in hamster cells after transformation (78). Glycoprotein Glycosyltransferases A clone of Chinese-hamster ovary (CHO) cells, deficient in lectin binding sites, was found to have a selective deficiency of UDP-N-acetylglucosamine: glycoprotein N-acetylglucosaminyltransferase activity (79). Sialyl~ and galactosyl-transferase activities were similar in normal and clone cells. This line shows a slightly altered cell morphology and weaker adhesion to tissue culture flasks (80). 28 Depression in a glycoprotein sialyltransferase (CMP~N~ acetylneuraminate: D-galactosyl~glycoprotein N-acetylneura~ 1transferase, EC 2.4.99.1) has been reported in mouse cells transformed with a temperature-sensitive mutant of Isimian virus 40 grown at the restrictive temperature (81). Temperature sensitive alterations of glycosyltransferases have also been reported in cells transformed by Rous sarcoma virus (82). V Cancer Associated Alterations in Glycosyltransferase Activity Differences in cell-surface-derived glycoproteins from normal and tumor derived cells have been noted by many investigators (reviewed in 56). Similarly, the presence of distinctive glycoproteins in the urine or serum of patients with cancer has led to the implication that these are derived from the tumor-cell surface (83,84). Investigators have periodically reported increases in serum glycosyltranferase activity from patients with cancer. A general question is whether these tumor derived soluble glycosyltransferases are distinct from the membrane-bound forms. At least one investigator has proposed that soluble galactosyltransferases are produced by proteolytic cleavage of the membrane enzyme with both forms having similar kinetic and regulatory properties (85). The occurrence of cell-surface glycosyltransferases is still disputed but it should be noted that both increased and decreased cell-surface glycosyltransferase activities have been noted 29 in tumor cells (86,87). Increases in serum glycosyltrans~ ferase activities, presumbly sloughed off.cancerous tissue, have been demonstrated, i. e. sialyltransferase (88,89) and fucosyltransferase (90,91). 1 Equivalent activity of total serum galactosyltrans- ferase is found in normal controls and patients with cancer; however Podlsky e; 31. (92) has demonstrated the presence of an electrophoretically distinct form of galactosyltrans- ferase in Samples from cancer patients. The normal (GT-l) and cancer associated (GT-II) isoenzymes differ in molecular 'weight, kinetic properties, and carbohydrate content (92). Podolsky extended these studies to an animal mOdel system in which hamsters were inoculated with polyoma transformed baby-hamster kidney cells (BHpr). A serum isoenzyme GT-IIh could be solubilized from the resultant tumor but not from normal hamster tissue (93). In additiOn media from BHpr cells grown in tissue culture contained a galactosyltrans- ferase activity that co~electrophoresed with the GT-IIh found in tumor bearing animals. These findings suggest that the amount of GT-IIh detected was related to tumor mass. Total serum galactosyltransferase or total serum sialyl- transferase fetuin acceptor activity was identical between control and tumor-bearing hamsters. These results constrast with the elevated levels of plasma sialyltransferase found in human cancer patients. (94). 30 Although a correlation between a specific galactosyl- transferase _isoenzyme and malignancy was found, a role for this isoenzyme in cellular behavior has not been shown. A distinct glycopeptide in sera and effusions of patients with 'extensive carcinoma has been detected (95). It was found to function as an acceptor for galactosyltransferase activity with a greater affinity for the cancer-associated galacto- syltransferase (96). This acceptor caused a selective growth suppression of transformed cells in tissue culture (96) and contains 60-70 percent carbohydrate (96). The mode of growth suppression in transformed cells is unknown but it may be related to its ability to act as a substrate for tumor-associated galactosyltransferase. Alterations of Glycosyltransferases in Development Developmental changes require cellular recognition processes in that cells undergo specific orientations and rearrangements with respect to one another. Consequently many such processes may be regulated by cell surface glyco- conjugates and/or cell surface glycosyltransferases. The adhesive specificity between embryonic cells and the migration of cells during development may be regulated by glycoconjugates (49). A DSG~fetuin glyCOprotein galactosyltranferase activity has been measured as a function of embryonic age in rat liver, lungs, and brain tissues (97). Enzyme activity was high at embryonic day 16 and then decreased with gestational 31 age. NO molecular differences were found between embryonic and adult .enzyme. Carlson 32 a1. (98) measured two additional galactosyltranferase activities (acceptors were: 1) GlcNac and 2) desialized sheep submaxillary mucin) from embryonic rat pancreas, liver, and gut. The liver enzyme showed no specific trend, but the pancreatic transferases increased from 10- to 40-fold in specific activity from embryonic day 12 to birth. A direct involvement of carbohydrate in differentiation is seen with Tunicamycin induced differentiation of human (BL-60) and murine (M1) myeloid leukemia cells in culture. After treatment with Tunicamycin morphological changes occurred in these cell lines that caused them to resemble mature myeloid cells. Fc receptors were induced in the M1 line (characteristic of differentiated cells). The implication of these observations is that glycosylation of cellular proteins plays a role in maintaining these myeloid leukemia cells in an undifferentiated state in culture (99). Alterations of glycoproteins have been observed during the development of embryonic chick neural retinal cells (100). When a glycoprotein galactosyltransferase was examined in this system activity was found to decline as a function of embryonic age (101). ‘ When various glycosyltranferase activities were examined in liver cells of embryonic chicks, only a GDP-Man tranferase was found to increase with embryonic age (102). Den et al. (103) have reported ‘that several glycosyltransferases undergo a 32 transition from soluble to membrane-bound forms during the development of the embryonic chicken brain. In the slime mold, Dictypstelium discoideum, a large molecular weight glycoprotein is induced 12 hours after Linitiation of differentiation (104). Other studies with this system have implied that the carbohydrate moieties play roles as signals in cellular recognition phenomena (105). A model system to study cell-cell aggregation phenomena is’ the sponge cell system. An aggregation factor purified from the Sponge Geodia cydonium was found to contain as subunits: sialyltransferase (106), glucuronosyltransferase (107), and galactosyltransferase (107). A mOdel relating these glycosyltransferases to a mechanism of adhesion has been proposed by Muller e5 El- (108). An interesting model system for_ the study of differentiation dependent biochemical alterations is the epithelial cells ‘of the villi of the small intestine. Epithelial cells at the base of the villus are the least differentiated. As the cells move up the villus they become progressively more differentiated until they. reach the distal end of the villus where they Slough off. Alterations of cellular adhesion factors clearly occur as the mature villi cells are more easily dissociated than the crypt cells. Studies by Weiser (109) with isolated villi cells found several alterations in enzymatic activity as the cells undergo differentiation. Incorporation of labeled sugars into glycoproteins increases as differentiation occurs with 33 maximal incorporation by the fully differentiated cells at the villus .tip. Crypt .cells have’ higher endogenous glycosyltransferase activity for transfer of fucose, mannose, N-acetylglucbsamine, galactose, and glucose than the tip cells. Only CMP-sialic acid transferase activity (endogenous) is elevated in tip cells as Opposed to the crypt cells. Since endogenous activity requires the ”in situ" presence of transferase acceptors, an apparant increase in completed glycoslated products occurs as the cells differentiate and move towards the tip. The elevated CMP-Sialic acid transferase in the tip cells may imply a regulatory point for terminal Sialic acid. . Glycoproteins have frequently been reported to be altered upon transformation and have been correlated with cellular differentiation. These alterations are usually associated with altered synthesis patterns rather than glycosidic activities (110). Rat Pancreas as a Developmental Model Embryonic development is a complex series of temporally related events that starts with common genetic precursors and leads to a phenotypically diverse cellular architecture. The pancreas has served as a developmental model of differentiation at three distinct levels, which are : l) histoaenesis, in which endocrine, exocrine, and duct cells orient themselves into their organ Specific morphologies, 2) cytodifferentiation, in which cellular differentiation 34 occurs, and 3) biochemical differentiation which involves the synthesis of cell specific structual and enzymatic proteins. Extensive information has been documented on the morphology and biochemistry of this organ during development ‘(lll,112) and consequently it is well suited as a model system for differentiation. Since altered levels of various galactosyltranferases are known to occur during pancreatic development (98), it was decided to extend these observations in order to more fully elucidate the role of glycosylation in development. Glucocorticoids-have been shown to modulate the “in vitro” development of the embryonic rat pancreas (113), and similar steroids alter glycosyltransferase activities (114) in other organ systems. Consequently, a correlation may exist between the endocrine system and the level of glyco- syltransferases. Embryonic Differentiation - The pancreatic divesticulum emerges from the gut at 11 days of gestation in the rat and concurrently mesodermal cells condense around it to form the pancreatic bud. Low but significant levels of exocrine proteins are detectable at this early stages Rapid cellular proliferation and morphogenesis follows with acinar structures becoming apparent at about 16 davs. Levels of exocrine proteins do not rise significantly at this stage, which is termed the ”protodifferentiated state" (115). After 16 days a large increase in the rough endoplasmic 35 reticulum occurs and distinct zymogen granules appear. The cells have .now acquired the structural characteristics of mature cells. This secondary transition leads to the “differentiated state". At day 17 or 18 of gestation .zymogen granules have accumulated near the apical surface and cytodifferentiation is complete. An important technique in the developmental study of the pancreas was the development of the “in vitro" culture of pancreatic rudiments. It was found that under appropiate culture conditions pancreatic rudiments synthesize exocrine proteins at the same levels found in intact embryos of the equivalent age. This culture system has allowed controlled studies on the underlying mechanisms of regulation of the development of this organ. RuEter e3 31. (116) has demonstrated that the accumulation of digestive enzymes was dependent on the presence of mesenchymal tissue and that the factor involved could exert its effect without physical contact. The thymidine analogue, 5-bromodeoxyuridine (BrdUrd) has been shown to exert a selective inhibition on cytodifferentiation Of the embryonic rat pancreas cultured “in vitro" (117). Rudiments grown with BrdUrd appear to be normal but do not contain zymogen granules even though protein, RNA, ‘and DNA synthesis occur at normal rates. Synthesis and accumulation of exocrine proteins is selectively inhibited by BrdUrd. BrdUrd does not inhibit the glycoprotein galactosyltrasferase to ovine submaxillary 36 mucin and does not inhibit the synthesis Of proteins labeled with tritiated glucosamine (117). In addition DaGue (118) has shown that 5-BrdUrd has no effect on zymogen granule membrane-like glycoprotein synthesis. A general mechanism 'of BrdUrd action may involve altered binding of regulatory proteins to BrdUrd-containing DNA (117). Embryonic pancreatic rudiments grown in culture incorporate radioactive L-fucose into a large number of glycoproteins some of which are induced during pancreatic differentiation (118). Walther (117) has reported that =-ernrd, which prevents the ”in vitro" secondary transition in developing exocrine pancreas, does not. inhibit the synthesis of glycoproteins. Furthermore DaGue (118) found that .qucoprotein profiles of SDS-polyacrvlamide gels of tritiated glucosamine labeled control and S-BrdUrd treated cultured rudiments were identical. The number of glyco- proteins induced during pancreatic differentiation was found to be from five to seven depending on the sugar used to label the protein (118). These all occurred during the secondary transition period. Glycosyltransferases as a Regulatory Mechanism in Protein Secretion After synthesis on ribOsomes, specific proteins may be transported across or integrated into distinct cellular membranes. ‘Then subpopulations of these proteins need to be routed to other intracellular membranes or secreted. The role of the oligosaccharide moiety as a routing signal has 37 proved attractive. The information carried in the oligosac- charide could easily account for directing the protein to various points. The role of glycosyltransferases in secretory cells is unclear. Although many secretory proteins are glycosylated, the pancreatic secretory proteins for the most part are not. Pancreatic secretory proteins are stored in zymogen granules and the purified zymogen granule membrane contains characteristic glyCOproteins (119) which may, concerted with the action of specific glycosyltransferases, play important roles in secretion. Additional studies with rabbit antisera against highly purified zymogen granule membranes support the involvement of a galactosyltranferase with zymogen granule glycoprotein formation (120). Rat Pancreas as a Secretory Model The mammalian pancreas is a structually heterogeneous organ containing two major classes of cells: the beta-cells, which are primarily of endocrine function and the alpha-cells, which are exocrine and function in the synthesis and secretion of digestive zymogens. All subsequent discussions will be concerned only with the pancreatic acinar cells. The rat pancreas has been chosen as a system in which to study the role of glyCOprotein galactosyltransferase activity in secretion. .Previous work by Ronzio (121) has partially characterized a system with respect to 38 lycosyltransferase activity. Secretory Process - The exocrine cells of the pancreas have provided an excellent model for studies of protein (synthesis, segregation, intracellular transport, concentration, and discharge. Its role as an efficient protein synthesizer and secretor has led to its being utilized as a model cell for the study of such phenomena by numerous investigators. An electron micrograph of an isolated pancreatic acinar cell is shown in Figure 3. The sequence of events leading to secretion have been delineated through pulse-chase experiments utilizing radioactive amino acids followed by autoradiography and/or cell fractionation studies. ' Proteins made for export are synthesized on polysomes attached to the rough endoplasmic reticulum (RER). The newly made secretory proteins are then segregated into the cisternal spaces of the RER. From the RER the proteins are transported to the transitional elements of the Golgi system, then to the small peripheral vesicles on the cis side of the Golgi complex and finally to the condensing vacoles. This intracellular transport requires energy and in the absence of ATP synthesis, the secretory proteins remain in the' RER. The exact molecular nature of these steps remain to be elucidated. After reaching the condensing vacoles an increase in density occurs (as noted by electron microscopy) and the result is the conversion of 39 Figure 3. Transmission Electron Micrograph of a Dispersed Pancreatic Exocrine Cell ' Transmission electron microscopy ( xl4,000) of this pancreatic exocrine cell demonstrates the complex architecture of this cell type. cb- cytoplastic blob, CVL- loose conden- sing vaculoes, arrows- Golgi elements, Z- zymogen granule, marker- l pm ( J. Cell Biol. (l974) 63, 1049) 40 41 these vacoles to the secretory particles called zymogen granules. Studies of isolated zymogen granule membranes have shown that a major glyCOprotein characterizes granule membranes of the mammalian pancreas (122). Consequently ldistinct topographical features of the zymogen granule (ZG) membrane are present which may play roles in concentration or secretion of the zymogens. The final step in the process is discharge. The ZGS discharge their contents into the glandular lumena thorough exocytosis. This membrane fusion occurs only with the apical plasmalemma. Secretion is stimulated upon hormonal action at the basal cell surface. An outstanding current problem is the coupling of secretion to this spacially distant hormonal Stimulation. The above model of pancreatic secretion proposed by Palade and his collaborators (123) has achieved general acceptance and in part for these studies Palade was awarded the Nobel Prize. The acceptance of a paradigm by the scientific community by no means implies its correctness, only its acceptability to current scientific attitudes (124). The current experimental evidence is consistent with the Palade "secretory model" but does not imply proof of its existence nor are alternative interpretations of the data excluded. In addition, several experimental Observations are not easily accounted for by the current paradigm. Consequently an alternative model of secretion has been proposed by S. S. Rothman (125). Palade's theory assumes complete sequestration of secretory proteins from the 42 cytoplasm from the point of synthesis to secretion of zymogen granules by exocytosis. Rothman’s model, i.e. an "equilibrium system", proposes that the cytoplasm of the cell acts as both a mixing chamber for digestive enzymes as ‘well as a precursor pool for secretion. Several lines of evidence are consistent with this theory. Digestive enzymes in zymogen granules have been reported to be in equilibrium with enzyme in the surrounding medium across the granule membrane (126,127). In suspended tissue slices, labeled enzyme equilibrates across the cell membrane, through the cytoplasm, and into the zymogen granules (126,127). Upon cholinergic stimulus a depression of the specific radioactivity of secreted protein is found.‘ This is contrary to what is expected from Palade's model but is consistent with the hypothesis that zymogen granules discharge their contents into the cytoplasm prior to secretion across the cell membrane (128). The two models are not mutually exclusive, however, and both procesSes may be occuring concurrently. Glycoproteins are important constituents in all membrane subfractions of the exocrine pancreas. The zymogen granule membranes are high in protein bound galactose and fucose and have a mass ratio of carbohydrate to protein of 0.44 (122). A Zymogen granule membranes 5 contain 3 major glycoproteins, designated GP ‘ (MW 120,000), GP-2 (MW 74,000) and GP-3 (MW 52,000), which become labeled with 43 tritiated glucosamine (122). Upon immunOprecipitation with antibodies made against purified zymogen granule membrane a labeled glucosamine-labeled glyCOprotein of the same eletrophoretic characteristics of GP-2 is detected in the fismooth microsomal Golgi-enriched fraction. GP-Z has similarly been immunologically detected in both the "protodifferentiated" and the "differentiated" pancreas (Ronzio, unpublished observations).~ Present evidence implies that each* zymogen granule contains the same complement of digestive enzymes and that sub-populations of zymogen granules do not exist (129). ~Therefore secretory transport has no selectivity and regulation of enzyme concentration must be derived from selective synthesis. A specific stimulus can alter the rate of secretion of one enzyme relative to another; however the molecular basis of this phenomenon is not clear. Ronzio and Mohrlok (120) have described a Golgi membrane associated galactosyltransferase involved in the formation of zymogen granule glycoproteins. Alterations of Glycosyltransferases by Chemical Agents Butyrate - The addition of sodium butyrate to the media of a variety of cell lines causes biochemical and morphological changes. In several cases the morphological changes induced by butyrate can be blocked by the addition of dibutyryl-cAMP 130,131,132). Butyrate has also been found to induce a glycolipid glycosyltransferase in HeLa (133) and KB (134) 44 cells. cAMP has also been found to Oppose the sialyltrans- ferase induction by butyrate in HeLa cells. Cell surface carbohydrates have been found to vary in tissue culture cells when they are grown in the presence of butyrate (135). .The mechanism of the butyrate effect is unknown but recent observations. indicate it may be related, in KB cells, to a blocking of the cell cycle at the Cl stage (136). Phorbol Ester Tumor Promoters - A tumor promoter is a type of co-carcinogen that enhances tumor formation when administered after initiating action by a carcinogen. In the two-stage model of carcinogenesis a subcarcinogenic dose of carcinogen is applied to mouse skin and this is followed by repeated applications of a promoter. If the promoter treatment is continued carcinomas develop. The initiating. event is irreversible and probably occurs at a mutagenic level. On the other hand the effects of the promoter are reversible, are not mutagenic and may involve interruptions of normal intercelluar communication (137). Studies directed towards understanding the mechanism of tumor promotion are increasingly utilizing cultured cell systems as models. When phorbl ester tumor promoters are added to cells in culture a diversity of effects have been noted. Alterations have been observed in DNA and RNA synthesis, phospholipid metabolism, prostaglandin synthesis, nutrient uptake, cellular morphology, cell division, terminal differentiation, etc. 45 Driedger and 'Blumberg (138) have described three generalizations of their. biological activities: 1) normal cells will assume a transformed phenotype upon exposure to phorbol esters (139), 2) phorbol esters cause superex- Lpression of transformation-sensitive properties (140), and 3) the phorbol esters modify differentiation and differentiated cell functions (141). These phorbol ester induced alterations have all been previously linked to involvement with glycoconjugates, and consequently their effect on several glycosyltransferases has been examined in this investigation. Miscellaneous Agents Surface glycoconjugates are also altered in cells grown in tissue culture by the agents cyclic AMP (142) or ethidium bromide (143). - Turpentine injection is rats causes a two fold increase in microsomal galactosyltransferase specific activity in the liver. This increased activity was due to a proliferation of Golgi membrane rather than a specific induction of the glycosyltransferase .(144). An inhibitory effect by cycloheximide on galactosyltransferase activity in rat liver Golgi membranes has been shown but the conclusions were that this effect is secondary to the primary action on the membrane system (145). MATERIALS AND METHODS MATERIALS .Electrophoresis Reagents Acrylamide (99 percent) N,N'-Methylenebisacrylamide N,N,N',N’-Tetramethyl- ethylenediamine (TEMED) Ammonium PefSulfate Sodium Dodecyl Sulfate Bromphenol Blue Coomassie Brilliant Blue R Beta-Mercaptoethanol Basic Fuchsin Radiochemicals [Galactose Carbon-14]Uridine Diphosphate Galactose [Gamma Phosphate-32]Adenosine 5'-Triphosphate [Carbon-14JGalactose-l- Dhosohate [Carbon-141Galactose [Tritiated]Thymidine Liquid Scintillation Counting Triton x-100 Dimethyl-POPOP (1,4- (4-Methyl-5- Phenyloxazolyl)]-Benzene “OP (2.5-Diphenyloxazole) is[2- Bio-Rad Laboratories, Richmond,Ca Bio-Rad Laboratories, Richmond,Ca Bio-Rad Laboratories, Richmond,Ca Fisher Scien. Co., Fair Lawn,NJ. Pierce Chemical Co., Rockford,Il Nutritional Biochemical Corp., Cleveland,Oh Sigma Chemical Co., St. Louis,Mo Sigma Chemical Co., St. Louis,Mo Sigma Chemical Co., St. Louis,Mo New England Nuclear,Boston,Ma New England Nuclear,Boston,Ma New England Nuclear,Boston,Ma Nuclear,Boston,Ma Nuclear,Boston,Ma England England New New Research Products Internat. Corp.,Elk Grove Village,Il Research Products Internat. Corp.,Elk Grove Village,Il Research Products Internat. Corp.,Elk Grove Village,Il 46 47 Tissue Sources Rats (Sprague-Dawley) Spartan Research,Haslett,Mi Cell Lines Hamster Fibroblast-Normal Dr. P. W. Robbins,MIT, (Nil-8) Cambridge,Ma Hamster Fibroblast Dr. P. W. Robbins,MIT, Transformed with Hamster Cambridge,Ma Sarcoma Virus (Nil-BHSV) Human Epithelial Carcinoma American Type Culture (KB) Collection,Rockville,Md Chinese Hamster Ovary (CHO) Dr. Pamela Stanley,Albert . Einstein,New York,NY Hamster Fibroblasts (V79's) Dr. James Trosko,MSU, East Lansing,Mi Tissue Culture Minimal Essential Medium Grand Island Biological Co., (MEM) Grand Island,NY Dulbecco's Modified Eagle Grand Island Biological Co., Medium (DMEM) Grand Island,NY Fetal Calf Serum Grand Island Biological Co., Grand Island,NY Calf Serum Grand Island Biological Co., Grand Island,NY Earle’s Balanced Salt Grand Island Biological Co., Solution Grand Island,NY Penicillin-G Sigma Chemical Co., St. Louis,Mo Streptomycin Sulfate Sigma Chemical Co., St. Louis,Mo Miscellaneous Fetuin (Spiro Method) Grand Island Biological Co., Grand Island,NY Bovine Submaxillary Mucin Sigma Chemical Co., St. Louis,Mo Uridine Diphosphate Galactose Sigma Chemical Co., (UDP-Gal) St. Louis,Mo Adenosine Triphosphate (ATP) Sigma Chemical Co., St. Louis,Mo Adenosine MonOphosphate (AMP) Sigma Chemical Co., St. Louis,Mo 48 2-(N-Morpholino) Ethane United States Biochem. Corp., Sulfonic Acid (MES) Cleveland,Oh Orosomucoid . . Dr. Don Carlson,Purdue University,Lafayette,In Filter Discs (0.45uM) Millipore Corp.,Bedford,Ma Retinoic Acid Sigma Chemical Co., St. Louis,Mo .Butyrate Sigma Chemical Co., St. Louis,Mo Epidermal Growth Factor (EGF) Collaborative Research,Inc., Waltham,Ma Mellitin Dr. James Trosko,MSU, . East Lansing,Mi Phorbol Ester Tumor Promotors Dr. James Trosko,MSU, East Lansing,Mi Alpha-Lactalbumin Sigma Chemical Co., St. Louis,Mo Ampholytes LKB Inc.,Chicago,Il Bio-Gel P-200 Bio-Rad Laboratories, Richmond,Ca Sephadex G-100 Sigma Chemical Co., St. Louis,Mo Trichloroacetic Acid (TCA) Fisher Scien. Co., ‘ Fair Lawn,NJ Phosphotungstic Acid (PTA) Sigma Chemical Co., St. Louis,Mo Cacodylic Acid Sigma Chemical Co., (Dimethylarsinic Acid) St. Louis,Mo Dolichol MonOphosphate Sigma Chemical Co., St. Louis,Mo Sodium Meta Periodate G.F. Smith Chemical Co., Columbus,Oh . Blue Dextran Pharmacia Fine Chemicals, Uppsala,Sweden x-Ray Film Eastman Kodak Co.,Rochester,NY All additional chemicals and materials were reagent grade. 49 METHODS Rat Pancreatic Tissue as a Galactosyltransferase Source .Dissection and Homogenization of Adult Rat Pancreas Sprague-Dawley rats weighing 100 to 300 grams were stunned by a blow to the head, decapitated, and the pancreas immediately dissected and removed into a petri dish sitting in an ice bucket. The pancreatic tissue was cleaned of fat and connective tissue, minced in 0.3 M sucrose and homogenized with 10 strokes in a Potter-Elvehjem homogenizer. The homogenate was filtered through several layers of cheese cloth and centrifuged at 500 x g to sediment debris, nuclei, and unlysed cells. This supernatant fraction was labeled the fcrude homogenate“. Microsomes were prepared from this supernatant fraction by centrifigation at 100,000 x g for 1 hour (Beckman Model LS-50 preparative ultracentrifuge). The supernatant fraction was discarded and the microsomal pellet was removed and resuspended in a small volume of 0.3 M sucrose with several strokes of. a Potter-Elvehjem homogenizer. This microsomal fraction was then assayed directly or stored frozen at -80 degrees C. S0 Dissection and Homogenization of Embryonic and Neonatal Rat Pancreas Pregnant Sprague-Dawley females of appropriate gestation age were stunned by a blow to the head, decapitated, and the uterus dissected and placed in a petri dish in an ice bucket. Embryos were removed and placed in EarleZs balanced salt solution. The pancreases were dissected in plastic petri dishes using iridectomy knives with the aid of a binocular microscope. Embryonic age was confirmed by comparing the age-specific embryonic characteristics described by Christie (146) with the dissected tissue. For homogenizaton of the pancreases the tissue was placed in a microfuge tube containing 0.100 ml of ice-cold phosphate buffered saline (PBS). The microfuge tube was taped to the bottom of a plastic tray and submerged in ice-water. A sonicator probe (Biosonik II, Bronwill Scientific, Rochester, NY) was then rubbed against the microfuge tube (setting #5 for 5 sec) to disrupt the tissue. The resulting homogenate was used directly in the studies performed. The extremely small quantity of tissue present did not allow additional fractionation of embryonic tissue to microsomes. Neonates (day l is day of birth) were dissected with the aid of a dissecting Scope and iridectomy knives and the pancreases pooled. Samples were homogenized by sonication or with a Potter-Elvehjem homogenizer in an identical manner as the adult tissue. Microsomes were prepared (when tissue 51 quantity permitted) as for adult tissue. Subcellular Fractionation of Rat Pancreatic Tissue The procedure of Jamieson and Palade (147) as modified ‘by Ronzio (148) was used to fractionate rat pancreas rapidly with minimal degradation. Tissue was processed as described above to the 500 x g supernatant stage (crude homogenate). This supernatant fraction was centrifuged for 10 minutes at 1750 x g to collect the white zymogen granule-pellet. The 1750 x g supernatant fraction was then centrifuged at 8500 x g for 15 min. The resulting pellet was called the mitochondrial fraction. The post-mitochondrial supernatant fraction was then centrifuged for 1 hour at 100,000 x g. The supernatant fraction was termed the soluble fraction and the resulting crude microsomal pellet was further fractionated into smooth and rough microsomal fractions. This crude microsomal pellet was suspended in a small volume of 0.3 M sucrose by means of a Potter-Elvehjem homogenizer and centrifuged for 195,000 x g for 1 hour on a continuous sucrose gradient (0.5-1.5M). Smooth microsomes banded near 0.70 M sucrose while the region at the bottom of the tube contained the rough microsomal fraction. . Fractions were gently removed with a Pasteur Pipet and frozen at -80 degrees C until assayed. 52 Cultured Cells as a Galactosyltransferase Source Growth Conditions Cell lines were maintained in a humidified atmosphere 'Of 5 percent carbon dioxide-95 percent air at 37 degrees C and were routinely subcultured upon reaching confluency with a 0.05 percent trypsin solution containing 0.02 percent ethylenediaminetetraacetic acid (EDTA). Cells were normally grown as monolayers in plastic tissue culture flasks (Corning Glass Works, Corning,NY). Several cell lines (human epithelial and chinese hamster ovary cells) were grown in suspension culture. Human Epithelial (KB) and hamster V79 cells were grown in minimum essential medium (MEM) containing calf serum (10 percent), penicillin (100 ug/ml), and streptomycin (100 units/ml).v KB cells grown in suspension culture had calcium free .medium. Hamster fibroblasts (Nil-8, and a transformed line, Nil-BHSV) were grown in Dulbecco's modified Eagles medium with fetal calf serum (10 percent), penicillin (100 ug/ml), and streptomycin (100 units/ml). CHO cells were grown in alpha modified Eagles medium . (MEM plus ribonucleosides and deoxyribonucleosides) with calf serum . (10 percent), penicillin (100 ug/ml), and streptomycin (100 units/ml). 53 Addition of Chemical Effectors Cell lines were nOrmally maintained until cell confluency reached 40-60 percent. At this time fresh media ,containing the appropriate chemical effector was added to the culture flasks. After an incubation period of 12-16 hours, the media was removed, the cells washed twice with PBS, and the cells harvested. Sypchronization of Growth Log phase cells were synchronized in the early part of the late 61 period by exposure to two cycles of thymidine inhibition (149). Media containing 2 mM thymidine was placed on the cells for 24 hours. This media was then removed and the cells washed three times with Sphosphate buffered saline (PBS) followed by the addition of fresh media. After 12 hours the media was removed, the cells were washed with PBS, and media containing 2mM thymidine was placed on the cells for 24 hours. After this second block the thymidine containing media was removed, the cells were washed in PBS, and fresh media put on the cells. Synchronization was confirmed by measuring radioactive thymidine incorporation into DNA. 54 DNA Labeling DNA pulse-labeling was carried out by adding tritiated thymidine (52 mCi/umole; 0.1 uCi/ml) to control flasks at 60 .minute time points after synchronization. After a 60 minute incubation with tritiated thymidine the cells were washed twice with ice cold PBS, incubated at 4 degrees C for 15 minutes in the presence of 10 ml of ice cold 5 percent trichloroacetic acid (TCA) and scraped from the flasks with a rubber policeman. The- flasks were then rinsed with an additional 10 ml of 5 percent TCA and the precipitated material pelleted by centrifugation at 100 x g. Pellets were then washed twice with PBS and solubilized with 0.1 N NaOH overnight. 'Aliquots were assayed for protein (Lowry) and for radioactive incorporation of thymidine (by liquid scintillation counting). Harvesting and Homogenization Cells were removed for assays from tissue culture flasks by scraping with a rubber policeman or by a short incubation at 37 degrees C with 0.05 percent EDTA in PBS. A small amount of PBS was then added to the cells, the cells were suspended by swirling, and the suspended cells were then poured into centrifuge tubes. The tissue culture flasks were then rinsed with a second PBS wash. Cells were then centrifuged into pellets (500 x g), the pellets were washed with PBS, and then stored frozen at ~80 degrees C. 55 Cells were disrupted by mild sonication or by means of a Potter-Elvehjem homogenizer as described previously for pancreatic tissue. Galactosyltransferase Assays Assays with Exogenous Protein Substrates The assay mixture contained in a final volume of 0.05 ml : 0.25 M sodium N-morpholino-2-ethanesulfonate (MES), pH 6.7; 10 mM manganese chloride; 0.5 percent Triton X-l00; 0.125 mg desialylated degalactosylated fetuin (DSG-fetuin): and 0.46 mM UDP-galactose (UDP-galactose was added to UDP-[carbon- l4] galactose (280 mCi/mmole)) to a final concentration of 0.46 mM (22,000 dpm). After addition of sample (1 to 50 ug protein) assays were incubated for 30 minutes at 37 degrees C. Reactions were terminated by the addition of 5 ml Of ice cold 5 percent phosphotungstic acid in 0.1 N HCl. Precipitates were collected on Millipore filters (0.45 uM pore size), washed three times with 5 ml of the precipitation solution, washed with 2 ml of diethyl ether, placed in scintillation vials and dissolved in 1 percent sodium dodecyl sulfate-0.1 N NaOH. After neutralization with 1 N HCl, 10 ml of toluene-Triton X-l00 .Scintillation fluid (147) was added and the samples counted by liquid scintillation spectrometry. 56 Assays Involving Endogenous Lipid Acceptors Incorporation of [Carbon-14lgalactose into endogenous lipid acceptors was determined by extracting the _galactosyltransferase reaction mixture with chloroform-methanol (2:1, (v/v)). The lower phase was then dried in a scintillation vial, solubilized and counted. Assays with Free Glucose and Free N-Acetylglucosamine as Substrates The assay mixture contained in a final volume of 0.05 ml: 0.25 M sodium cacodylate buffer, pH 7.5; 10 mM manganese chloride; 50 mM glucose and/or 50 mM N-acetylglucosamine; 0.6 mM UDP-galactose (UDP-galactose was added to UDP-[carbon-l4]galactose to a final concentration of 0.6 mM (5000 d.p.m.)); and 0.001 to 0.050 mg protein. Some assays contained in addition to the above 0.0075 mg of alpha-lactalbumin. After incubation for 30 minutes at 37 degrees C the assays were terminated by the addition of 0.5‘ ml of ice-cold water. The entire assay mixture was then passed through a Dowex 1 column (Cl minus, 8 percent, l00-200 mesh) made. in a Pasteur Pipet. The column was washed with 1.5 m1 of water and the elutant and wash collected into a scintillation vial. Scintillation fluid was added (toluene-Triton X-100 mix) and the samples counted by liquid scintillation spectrometry. 57 Preparation of Desialylzed and Degalactosylated - Protein Acceptors The method of Kim (150,151) was utilized to Obtain protein acceptors free of the terminal N-acetylneuraminic acid and the penultimate galactose residues. The protein solution (10 ~ 20 mg/ml) was heated for 1 hour at 80 degrees C in 0.1 N sulfuric acid to remove N-acetylneuraminic acid. The solution was then neutralized with 0.1 N NaOH and dialyzed overnight against distilled water. The nondialyzable material was treated with 0.01 M sodium meta periodate in 0.05 M sodium acetate, pH 4.5, overnight at 4 degrees C in the dark. The reaction was Stopped by the addition of glycerol and then dialyzed overnight against 4 changes of distilled water. The dialyzed material was incubated for 13 hours at 4 degrees C in a 0.15 M. potassium tetraborate buffer, pH 8.0 containing 0.15 M sodium borohydride. After 13 hours the pH was adjusted to 5.0 with 2 N acetic acid and the sample was exhaustively dialyzed against distilled water. The dialysate was made 0.05 N in sulfuric acid, heated for 1 hour at 80 degrees C, neutralized with NaOH, dialyzed against distilled water and lyophilized. The lyophilized material was Stored dessicated at -20 degrees C. 58 Physical Characterization of Microsomal Galactosyltransferase Activity Isoelectric Focusing Isoelectric focusing studies were performed in a sucrose density gradient column according to the procedure of Massey and Deal (152). ConventiOnal polyacrylamide gel electrophoresis apparatus was utilized to carry out single tube isoelectric focusing experiments in sucrose density gradients. Microsomes were dissolved in 5 percent Triton x-l00, spun at 100,000 x g for 1 hour and the supernatant fraction removed and added to both the light ampholyte solution (5 percent sucrose (w/v), 0.43 percent ampholytes) and to the heavy ampholyte solution (20 percent sucrose (w/v), 2 percent ampholytes). A gradient maker was then used to form 5.6 ml gradients ‘in each column. Electrofocusing was conducted at 200 volts for 4 to 10 hours at 4 degrees C. Fractions were collected at the top of the column by forcing a 50 percent sucrose solution (w/v) into the bottom of the column with a 20 cc syringe equipped with a 22-gauge hypodermic needle. Fractions of 0.400 ml were collected, the pH .Of each was measured with a Beckman pH meter and 0.050 ml aliquots were assayed for galactosyltransferase activity (reaction mix modified from standard to allow for final assay volume Of 0.100 ml)- 59 SDS Polyacrylamide Gel Electrophoresis SDS-polyacrylamide gel electrophoresis was carried out using the system of Laemmli (153). The separating gel was 7 .percent acrylamide and cOntained Tris-HCl (pH 8.8)' and 0.1 percent SDS. The stacking gel was 3 percent acrylamide and contained Tris-HCl (pH 6.8) and 0.1 percent SDS. Polymerization was initiated by the addition of tetramethylethylenediamine and ammonium persulfate. Samples for electrophoresis were-prepared by adding the sample to an equal volume of protein buffer containing: 0.0625 M Tris-HCl (pH 6.8), glycerol 12 percent (v/v), 1.25 percent (w/v) SDS, 1 percent 2-mercaptOethanol, and 0.001 percent (w/v) bromphenol blue. The samples were heated in a boiling water bath for 2 to 3 minutes before applying to the gel. The buffer used in the top and *bottom reservoirs consisted of 0.025 M Tris base, 0.19 M glycine, and 0.1 percent SDS (w/v) (pH 8.3). Electrophoresis was done in a Bio-Rad Model 220 slab gel electrophoresis unit at 25 mA per gel until the bromphenol blue had migrated within 1 to 2 cm of the bottom of the gel (6 to 8 hours). Protein standards used for molecular weight calibration were:.Cytochrome C ~ 12,500 daltons (Boehringer Mannheim); Chymotrypsinogen - 25,000 daltons (Boehringer Mannheim); Egg Albumin - 45,000 daltons (Boehringer Mannheim); Bovine Albumin - 67,000 daltons (Boehringer Mannheim); Phosphorylase A - 94,000 daltons (Sigma); and Beta-Galactosidase ~ 130,000 daltons 60 (P-L Biochemicals). After electrophoresis the gels were stained for protein in a solution of 0.1 percent Coomassie brilliant blue R in 10 percent (v/v) acetic acid (154). After staining Lovernight gels were destained by multiple washes with 10 percent (v/v) acetic acid. Gels were stained for carbohydrate using the periodic acid-Schiff reagent (155). It was imperative that all traces of SDS be removed from the gel by repeated washes with 10 percent acetic acid before the PAS stain was applied.. The procedure of Fairbanks e5 al. was (156) used for the PAS staining process. Molecular Exclusion Chromatography Solubilized microsomes were run on .a Bio-Gel P-200 (exclusion limit 200,000 daltons) column in order to determine if multiple peaks of activity cOuld be detected. Bio-Gel P-200 (l00-200 mesh) was hydrated in 0.125 M MES buffer plus 0.5 percent Triton X~l00 and poured into a column measuring 0.9 x 80.0 cm. Microsomes were homogenized in running buffer, centrifuged for 1 hour at 100,000 x g, and the supernatant fraction was applied to the column. Fractions of 0.250 ml were collected and .0.50 ml aliquots from each were assayed for galactosyltransferase activity. The column was calibrated with Blue Dextran (V0) and potassium ferricyanide (Ve). 61 Pyrophosphatase Assays PyrophOsphatase activity was determined with the identical assay mix for galactosyltransferase except for deletion Of the exogenous acceptor. The reaction was .terminated by the addition Of 0.100 ml ethanol and 0.050 ml of 5 percent acetic acid (v/v) . Insoluble material was removed by centrifugation at 10,000 x g for 2 minutes with a microfuge (Model 6-811, Coleman Instruments Inc, Maywood, Ill). The 'Supernatant fraction was run on a descending paper chromatographic system as described under miscellaneous methods. Radioactive standards (all carbon-14) galactose, galactose-1-phosphate and _ UDP- galactose were run to calibrate the system. Radioactivity was quantitated in each peak by cutting the. chromatograms into pieces and counting them by- liquid scintillation spectrometry. Phosphorylation Assays Purification of Membrane Fractions Membrane fractions were prepared from CH0 and KB cells by the procedure of Thom pg 31. (157). Cells were lysed by dilution into hypotonic borate/EDTA buffer, pH 10.2. After removal of debris and precipitated material by low Speed centrifugation (450 x g for 10 minutes), a crude membrane fraction was obtained by centrifugation of the supernatant for 12,000 x g for 30 minutes. The pelleted material was 62 resuspended, layered over 35 percent (w/v) sucrose and centrifuged at 40,000 x g .for 45 minutes in a swinging bucket rotor. The material layering at the sucrose-buffer interface was removed, suspended in 10 mM Hepes, pH 7.4, and Irecentrifuged at 100,000 x g for 30 minutes. The pellet was then resuspended in Hepes buffer and stored at ~80 degrees C until assayed. Assay Conditions The reaction mixture contained: Hepes buffer, 20mM, pH 7.4; manganese chloride, 1mM, gamma 32p, 600,000 cpm; ATP, 15 uM; bovine serum albumin, 7.5 ug; and membrane protein, 10-60 ug in a final volume of 00.60 ml. The standard reaction was initiated by the addition of labeled ATP and after incubation at 4 degrees C for 4 minutes, was terminated by the addition of ice cold 10 percent trichloroacetic acid (TCA) containing 0.01 M sodium pyrophosphate (NaPP). The reaction mixture was placed on Millipore filters (pore size 0.45 um), washed extensively with 10 percent TCA-0.01 M NaPP, washed with diethyl ether, dried and counted .by liquid scintillation spectrometry. Promotors were added to the reaction mix before addition of labeled ATP and incubated at 37 degrees C for 5 minutes. After the 37 degrees C incubation, the assay tubes were cooled to 4 degrees C and the assay initiated with the addition of labeled ATP. 63 Miscellaneous Methods Paper Chromatography Descending paper chromatography was used to determine the rate of breakdown of UDP-galactose in the pyrophosphatase assay as well as to check the purity of UDP-galactose used in the galactosyltransferase assays. Whatmann 3 MM paper was cut into strips 3 .x 50 cm and samples were run with a solvent system of ethanol (75 m1)~sodium acetate buffer (30 ml-lM, pH3.8) (158) for 17-18 hours. Radioactive standards, UDP-galactose, galac- tose-1-phosphate, and galactose (all New England Nuclear) were run on duplicate strips. After drying, the paper strips were scanned for radioactivity with a radiochromatogram scanner (Packard Model 7201) and the areas corresponding to radioactive peaks were cut out, placed in scintillation vials, and the radioactivity quantitated by liquid scintillation counting. Lowry Protein Assay Protein was determined by the method of Lowry SE 31. (159) using bovine serum albumin as standards. Purification of Alphalactalbumin Alpha lactalbumin was obtained from Sigma (Grade II, 90 percent pure) and purified by exclusion chromatography on Sephadex G-100. The sample (500 mg) was applied in ammonium 64 carbonate buffer to a column (2.5 x 90 cm) previously calibrated with Blue Dextran (V0) and potassium ferricynide (Ve). Absorbance of column fractions was monitored at 295 nm and the major peak fractions were pooled, dialyzed 'against distilled water and ly0phi1yzed. Liquid Scintillation Counting Radioactive samples were counted on a Beckman LS-150 or a Hewlett Packard 2405 scintillation counter. The scintillation cocktail used was that of Patterson and Green (160) and contained one part Triton X~l00 to two parts toluene (0.015 percent l,4-bis[2~(4~Methyl-5- phenyloxazolyl)]~benzene (POPOP) and 0.82 percent 2,5-Diphenyloxazole (PPO)). Samples containing up to one ml of water could be used with ten ml of this cocktail. Autoradiography Radioactively labeled proteins were run on polacrylamide slab gels. The gels were dried on a gel drier (Bio-Rad Laboratories, Model 224) and autoradiography was performed at ~80 degrees C with X-Omat R film (Kodak) for exposure periods of 1-3 weeks. Electron Microscopy Transmission electron micrOSCOpy was performed with a Philips Model 300. Membrane samples were negatively stained with 2 percent ammonium molybdate. RESULTS Requirements of UDP-Galactosyltransferase from Pancreatic Rat Microsomes The assay for membrane-bound UDP-galactose: glyco- protein galactosyltransferase is described in Methods. The transfer reaction and the competing UDP-galactose .pyrophosphatase activity are outlined in Figure 4. Requirements for the transfer of galactose to desialized and degalactosylated fetuin (DSG-fetuin) are described in Table l. Cation Requirements - The reaction is dependent upon manganese chloride, although partial activity can be restored when magnesium chloride (40 mM) is substituted for manganese chloride. The Optimal manganese chloride concentration for enzymatic activity was found to occur near 10 mM as shown in Figure 5. Addition of EDTA (5 mM) to the complete reaction mixture reduces activity by approximately 80 percent. When manganese chloride was replaced with calcium chloride, zinc chloride, barium chloride, cobalt chloride, zinc sulfate, or mercury chloride (all at 40 mM) only trace levels of galactosyltransferase activity could be detected. Effect of Detergent and Dithiothreitol ~ When the assay was run in the absence of detergent, activity was reduced to near endogenous levels. Maximum activity was found to occur in the presence of 0.5 percent Triton X-l00. Other 65 66 Figure 4. UDP-Galactose Glycoprotein: Galactosyltransferase and UDP-Galactose Pyrophosphatase Reactions UDP-galactose is utilized as a substrate for the reactions outlined in the figure. The glycoprotein galactosyltransferase transfers galactose from UDP-Gal to the N-acetylglucosamine residue at the non-reducing termini of a oligosaccharide attached to a protein though a asparagine residue. A competing pyrophosphatase reaction cleaves UDP-Gal to galactose-l-phos- phate which is then further degraded to galactose and inorganic phosphate. 67 CH 2OH H” N! O O O-P-O OP-O- CHZON O O“ UDP- Galactose OH OH CH 2OH Pyroohosphoiose 0‘34 GlyCOproteln Galactosyltransferase Galactose-l-Phosphate 5‘ Pi CHZOS UDP HO OH V OH OH CHZOH CHZOH . Galactose HO 0 0 . 0-(lnner core oligosaccharide-Am) 0“ OH OH NHCCH 3 O 68 Table l. Requirements for the Incorporation of [14CJGalactose into Desialylated Degalactosylated Fetuin by Pancreatic Microsomes Modification of Reaction Mix Activity(%) Complete3 100 Minus MnCl2 5 Minus MnClz plus MgClz 25 - Plus EDTA(5mM) 20 Minus Triton-loo 5 Minus DSG-Fetuin l0 Plus Dithiothreitol(lmM) l00 Plus AMP(3.4mM) lOD Boiled Enzyme 5 Incubation at 4‘C(one half hour) l5 aAssay for galactosyltransferase is as described under Methods. Additions to the standard assay mix were as indicated. All assays were performed in duplicate with l0 and 20 pg of pancreatic microsomal protein. 69 Figure 5. The Effect of MnClz Concentration on the Incor- poration of [14C]Galactose into Desialized and. Degalactosylated Fetuin The MnClz concentrations were varied over the range indicated. Assays were performed in duplicate with l0 and 20 pg of microsomal protein per assay. Other assay proced- ures were as described in Methods. 7O _ O 2 L _ m 5 l5r- So; \Efioa 3:338: . 4O 3O 20 IO MnClz mM 71 detergents tested (deoxycholate, Tween 20, Tween 60, and Tween 80) in place of Triton x~l00 in the reaction mix were found to give reduced levels of enzyme activity. Dithiothreitol or mercaptoethanol were neither stimulators nor inhibitors of transferase activity. Effect of Enzyme Concentration, Incubation Time, and pH The galactosyltransferase has a broad pH optima as shown in Figure 6. Activity was higher with MES (Sodium 2-(N-morpholino)ethane sulfate) buffer than with Tris-HCL buffer and MES was used in all subsequent assays. ‘ The time course for [Carbon-l4]galactose incorporation into acid-precipitable and chlorofOrm-methanol soluble material is shown in Figure 7. The low constant incorporation into the chloroform-methanol soluble fraction implies that the transfer of [Carbon-14]galactose is direct from UDP-Gal and does not involve a lipid intermediate. This possibility was further investigated by adding exogenous dolichol monophosphate to the reaction mix. After this addition no increase of radioactivity into chloroform-methanol material could be detected. Endogenous incorporation (reaction mix without DSG-fetuin) remained low at all time points. The range of microsomal protein at which the standard galactosyltransferase assay gave linear results is illustrated in Figure 8. All galactosyltransferase assays were subsequently run in this linear protein range. 72 .—u:-m_rh z —.o use upampam wcmzum Ao:v_o;acosuzvnm anuom z F.o mew: pow: mrmcwam .meogumz cw umawrummu mm meowupncou cemecmpm mcwms ammma Log cvmuogn pusOmOLovs Owummrucma to m; cm ecu o” guvz mumuwpnzc cw umsrowsma are; mxmmm< assumm umumpxmouuapmmmo can uONFPHAmoo ops? amouoapmwnuepu so cowumcoaroucH map so :a Lo upmwwu use .m meamwu 73 0d ON O.® 05.... 0.? /1\ 8-3..- _2 ..o 6 228.3 2658 Aoc_.oficoE..2-.~ .568 2 ..o o 0’ 1- Thou / ugaimd bur /sa|ouiu '2 74 .cwmuoca PwEomoco_E Ovuemcucma mo m; op gap: mucuwpaze :? meow arm: mammmm FP< .cwsuwwnwmo uaozbvz mgauxvs :ovuuamc m =_ warmums m~a=pomcvnuwom ope? copumroaroucv w>vuumowcor m? cowumroagoucv mzocmmoucm .mmasa rmzop mg» ope? scrum -Loaroucw m>Fpoao_umc mzwrammms use Apumv Pocmgumsusrowocopgu new: wrzust corpummc use m:_uumruxm an outsmame mm: Omega vpgpp may ou:_ coPDHLoaroucH .coFDHDPaFOmLQ upon an cmcvstmumu mm: vasmuma mpaspomcvuewum ope? cowuoroaroocw m>wpomowumr new umumovucw macros asap as» an umaaoum use: mxammm mmmcwmmcmLu—Xmouompmm cgcecmum .m_smuaz m—azpom A—HNV —o=m;umznsrowogopgo ace m_azpom:H ewu< can? amouumpawmuepg we copumroasouca Lou mmrzou as?» .m mrzmwm 75 00 on 4 2:5 lb]- Ell is} 5:335 so 66:. CV m on ON 0. L. a Mi 355021533.on 91¢ unocoooucm ole «39336 I ‘ O on. o in s1- 0 O CO on» ugeimd 6r! 0| / mdp 76 .czosm men mxommc mumuwpgso .mmEOmoLOws Owummgocma “as mm: worsen max~cm .mcogbmz cw ema_gumoc mm operation mew: mammmm mmmgmmmcmrppxmouompmm utmvcmum cowumgucmucdu mstcm ob puwammm cup: >amm< mmmrmmmcmgupxmouumpau to zuwimmcwa .m mrzmvm 77 2:203 9: .525 253.5 ON. Co on O? on ON 0. q q u . d a q u 00. CON con 00¢ 1 00m Inou/uldp 78 Acceptor Specificity ~ The activity of the galactosyl- transferase .when assayed with different protein substrates is summarized in Table 2. Bovine mucin minus NANA gave the highest incorporation of [Carbon-l4]galactose into acid (insoluble products followed by DSG-fetuin and unmodified bovine mucin. The activity towards fetuin minus NANA was less than towards native fetuin. DSG-orosomucoid had significantly less acceptor activity than DSG-fetuin. Exogenous acceptor purity was determined on SDS-polyacrylamide gels. Native fetuin (GIBCO) ran as a major band followed by two small secondary peaks while the bovine mucin was found to be extremely heterogeneous (Figure 9). Fetuin from Sigma (Type I and II) was found to be more heterogeneous than the GIBCO fetuin (data not shown) and consequently the GIBCO fetuin has been used in all studies. The endogenous activity (minus exogenous acceptor protein) was never greater than 10 percent of the total activity with DSG-fetuin present. Exogenous Acceptor Concentration ~ DSG-fetuin was run in the reaction mix at concentrations from 1 mg/ml to 10 mg/ml. Transfer of .[Carbon~14]galactose displayed linear kinetics over this range when up to 34 ug of pancreatic microsomal protein was assayed (Figure 10). At higher enzyme concentrations (67 ug) the assay lost its linearity. Standard assay concentration of DSG-fetuin was set at 5 mg/ml. Concentrations of DSG-fetuin above 10 mg/ml in the assay mixture were impractical as viscosity effects limited 79 ea=_srosma Doze coccm mewsamouuaFSm—auooom oom— o:_smmouapm—xumuom coca assuam Lao=\=_muora ms\ .E.a.u mzupmom Lemzm umtrommcmch smouoapawmue_u pacwsLOP ppmcaxm mucgamaam mmmrmmmcoLDPAmoanpac paEOmOLOA: upuaosucom .2. xu_O_w—umam mungumazm :.oucga .m m—nch 80 .Am o_SN pwmoz mucwszeumca meoepvwv emccwmm _mm ueommcmeu emwcvp m saw: uwseoeema mm: mpwm man we mcmmm zuvmcwu mewuao .meozumz c? mmawewmmc mew mpemuwo .Aee_;wm -uFOm meowemmv maweuzgoaemm eo Acvmum maps mammmsooov :Pmuoem eoe cw:_wmm mew: mpmm macawpmzu new mpmm quaemcvpzu x“ co cze mew: am; copy mmpaswm Anamemv :wwaz xemppwxmsaam wcw>om mew AcumHov :quwm mo mwmxpm=< Oeumeo;moeummpm .m mezmsm 8T .8 a as. so a seam o I. \ I \o \s \ I x I s \ o s 6 I ll Q a Q \ C re :23 - Eco micron. .. mam osmoEooolu 3839 52:2 f2=onaam 2.36m IIIIII I“-‘-lll.|lu‘ll \‘PUOO'DUUUIU-----D'IIUEIII- , II \ " \ =28 - 28 6.8:... .. 3.0 o_aaoEoou..l .892 52.... Mysuoa (condo [muse (condo 82 .cmumwpmcv mm mm: st Smmmm may c? cvzummuwmo eo cowumemcmmcow wzh .mmoguwz cw umawewmmc mm umm: mew: mxmmmm wmmewemcmeupxmouwmpmm vewmzmmm zuwemwcpm zwmm< wmmemememeDPAmoummpmu co cowumeucmmcou cwsummncmo we poweeu ms» .0? me:m_m 83 L ' l I l l 1 O O 3 w v- 2oo .. l60 ,- inou / UlalOJd 6w /sa|owu I so So DSG-Fetuin (mg/ml) I0.0 4.0 2.0 84 the practicality of the assay. Stability - The transferase was remarkably stable when stored as a microsomal suspension in 0.3 M sucrose. After 24 hours at 23 degrees C over 90 percent of activity was 'retained. After two weeks at ~20 degrees C over 80 percent of activity remained. Activity remained unchanged when the microsomal suspension was stored up to one year at ~80 degrees C. Identification of Labeled Galactose Incorporated into Protein The product of the DSG-fetuin galactosyltransferase assay was analyzed to rule out the possibility that the radioactivity measured in the product was a metabolic product of the labeled UDP-galactose. Labeled DSG-fetuin was hydrolyzed in 4 M HCl at 100 degrees C for 3 hours. After hydrolysis the sample was neutralized with silver carbonate and subsequent paper chromatography revealed a single radioactive spot migrating with standard galactose (data not shown). Subcellular Localization of Galactosyltransferase Activity To determine the subcellular distribution of galacto- syltransferase activity, rat exocrine pancreatic tissue was fractionated as outlined in Figure 11. The resulting fractions were assayed for galactosyltransferase activity 85 Figure 11. Sub-Ce11u1ar Fractionation of Rat Pancreatic Tissue ' The fractionation scheme used was that of Jamieson and Pa1ade (42) as modified by Ronzio (193). Detai1s are in Methods. 86 PANCREATIC TISSUE 1. 10 strokes with a Potter-E1vehjem homogenizer 2. Fi1ter through cheesecloth 3. Centrifuge 500 x g for 10 min. Debris. Nuclei and P 11 Hho1e Ce11s ‘y (Supernatant) CRUDE HOMOGENATE 1. Centrifuge at 1750 x g for 10 min. EEEEEEN GRANULE <_ (Pe11et) —___-_- ~ NV (Supernatant) 1. Centrifuge at 8500 3 g for 15 min. $L$DRIAL é (Pe11et) (Supernatant) POST-HITOCHONDRIAL SUPERNATANT 1. Centrifuge at 100.000 x g for 1 hour SOLUBLE fiKCTTEN €r_v (Supernatant) (Pe11et) MICROSOMAL PELLET . Suspend in 0.3 M sucrose . Centrifuge 195,000 x g for smom «Icszosomas Lugggge‘grgdfiggfimws ROUGH mcnosonrs (col1ected near (0.5-1.5M) (coIIected at bottom of 0.70" sucrose) gradient) Nd 87 and the results are shown in Table 3. The debris pellet had a similar specific activity to that of the crude homogenate. This fraction was largely made up of undisrupted whole cells and connective tissue. The zymogen granule fraction icontained} very little activity and other studies with, purified zymogen granule membranes (148) have shown similar results. Galactosyltransferase had little activity in the post-mitochondrial supernatant fraction as would be expected for a membrane-bound enzyme. Clearly, most of the recoverable activity (from the crude homogenate) was found in the crude microsomal fraction with an enrichment of specific activity by a factor of almost eleven. The rough microsomal fraction had essentially the same activity as did the crude microsomal mixture. The smooth microsomal fraction contained less than 4 percent of the total microsomal activity, yet gave a relative enrichment of 30 fold in specific activity over the crude homogenate. The smooth microsomal fraction was composed of Golgi, plasma membrane, and ribosome-free endoplasmic reticulum. Terminal galactosyltransferase activity is considered to be generally localized in the Golgi fraction, however, localization in other membrane fractions could not be discounted. 88 manganese: mango mo xuw>_uo< o_$_omam\co_uomcm mo xu_>_uo< u_mpooqmm .59. can 5395 a... can notch—mach 033.2% 8.95... manganese: ounce co xu_>_uo<\:owuoocc ma zu_>_ao_uu< ~>_ua_o¢ u»u_>_uu< uwc_umam usua>_uu< pouch to x ass.>_su< _ouop co.uoagm o.m:o_uoacu cops—_moaam opuoocoeaa cw omocoumcucu—xmouoa—aa ”:_:uom-wma we :o_u:a.cum_o m o—aah 89 Molecular Exclusion Chromatography of Solubilized Galactosyltransferase Multiple forms of galactosyltransferase have been detected in other systems where the enzyme exists in a [soluble form (161). The possibility that multiple forms of the microsomal galactosyltransferase might also exist was explored by running solubilized microsomes over a Biogel P~150 .molecular exclusion column. A broad peak of activity was observed immediately after the void volume (data not shown). Polyacrylamide Gels of Solubilized Galactosyltransferase Solubilized microsomes were analyzed on 7 percent polyacrylamide gels and galactosyltransferase activity localized by slicing the gels into fractions and incubating the minced fractions with a galactosyltransferase reaction mixture. Multiple forms of serum galactosyltransferase have been distinguished on polyacrylamide gel by other investigators (162). The results of a typical gel are shown in Figure 12. One or perhaps two bands of transferase activity were seen. .The recovery of activity from the gels was extremely low (less than 10 percent) under all conditions examined. 90 m pmcmau.Am opem —mvos .mucmsngmc~ ugowppcv “commence emmc__ m ;u_3 umccmom use cwmum :Pmaoga maps mpmmmsooo gum: umcwmum memz mpmm mumowpazo .< _mcma cw exocm m? pmpcmams mpnzpomcw ovum oucv :oFumLoaeooca .mssux_s coyuommg mmmmemcmcupxmouompma msu gov: nmomnaoc? new meowuommm ope? umowpm memz mpmo .mpmm muvsmngomapoa NR co :3; Am; copy mm—asmm use oo—ux cobweb am guy: umNFPFnapom memz mmsomomovs ovummgocma mam m_mu muwsmpagomxpoa seem zuw>wpo< mmmcmwmcmcupamouompmo mo mgm>oomz .NF mcamwm 91 A6022: c0202... _mo 30: On n¢ O? on On 0N ON m. 0. n o .4 qt q q d dl Saum 02m 03005000 Isoeaum .m d J A _ d 1% fi _) q d 23:04 ogcmficozanofiofic 5.8336 .< l N -: o C n. O (W 099) aouoqmqv <1: 0 O N 0? cm 00 92 Isoelectric Focusing of Solubilized Galactosyltransferase Solubilized enzyme was run on an isoelectric focusing column to determine an isoelectric point for the enzyme as well as if multiple forms might be separated by charge. Figure 13 demonstrates that a broad though well defined band of activity ran between pH 6.0 and pH 8.0 with a maximum galactosyltransferase activity at pH 6.9 to 7.0. Pyrophosphate Activity in Microsomal Preparations Since the presence of pyrophosphatase activity would hydrolyze UDP-galactose, the extent of perphosphatase activity was measured in the microsomal preparations. AMP and ATP were examined for their inhibition of pyrophosphatase activity and their effect on galactosyl- transferase activity. Standard protocol for the galactosyl- transferase assay was used except that inhibitors were present in the concentrations indicated. The assays were stopped by the addition of ethanol-acetic acid (5 percent) rather than TCA-PTA in the galactosyltransferase assay. This alteration was found to be necessary as the TCA rapidly hydrolyzed the UDPegalactose upon addition to the reaction mix. Immediately after quenching the assay, the entire reaction mix was run on a paper chromatographic system which clearly separated UDP-galactose, galactose~l~phosphate, and galactose. Control scans of three paper chromatograms are shown in Figure 14. The rate of hydrolysis of UDP-galactose 93 .muogumz cw cczom mcm mpvmumo .coFuomcc comm to; umcweLmumu Pavemums m—aapomcw ovum cue? :owumcoacoocp new .x_s xmmmm mmmemmmcmmupzmouomme ocmucmum mg» cw nmxmmmm .csapoo mg» soc» cmuomhpoo mcmz meowpomcm.csspoo mermaoom ovguompmomp bechmcm xuwmcmu mmocoam a co can mgmz oopux cobweb um cw um~p~_napom mmsomogows ovummgocma an: mmmcmmmcmcupzmouompmo cm~vanzpom we marmzoou oweuompmom~ .mp mezmwm 94 QN 0.? 0.0 0d 0.0. ON. 39:52 cozooi n. m. __ m N 1 I I q q q - 1 d) u d u 4 3.3.04 macemwmcoczxmofigoo I IQOIO CO. CON .Oon 00¢ con man /\u‘d‘p 95 Figure 14. Radiochromatographic Scans for the UDP-ga1actose Pyrophosphatase Assay Incubation and assay conditions are described in Methods. Panel A- 10 pg pancreatic microsomal protein with UDP-[14C] galactose (22,000 dpm) Panel B- Conditions are the same as for panel A except the microsomal protein was heated in a boiling water bath for 3 minutes before addition to the assay mix. Panel C- Radioactive standards- All New England Nuclear 96 A. Enzyme B. Boiled Enzyme iC. Standards — s‘ ~ K .41 j 1' L 4. a ' .g >‘ '5 .5 >~. 0 °: 0 o o UDP-Gal- Gol-l-P- 97 is shown to be proportional to incubation time in Figure 15. The effect of 3.3mM adenosine triphosphate (ATP) and 3.4 MM adenosine (AMP) on galactosyltransferase activity is summarized in Table 4. Radiochromatographic scans of the ‘pyrophosphatase assay in the presence of AMP and unlabeled UDG~Gal is shown in Figure 16. The course of hydrolysis of UDP-galactose in the presence of 3.3 mM ATP and 3.4 mM AMP is shown in Figure 17. After scanning with a radiochromatogram scanner (Packard Model 7201), the paper strips were cut into sections and counted by liquid scintillation spectrophotometry. All assays were done in duplicate. The results indicate that neither AMP nor ATP at the concentrations tested inhibited or stimulated galacto- syltransferase activity in ‘the absence of appreciable pyrophosphatase activity. In most cases the large excess of unlabeled UDP~galactose (4 mM) was sufficient to prevent excess hydrolysis of labeled substrate (UDP~[Carbon~ l4]galactose) by pyrophosphatase activity. AMP was determined to be a better inhibitor of pyrophosphatase activity than ATP and was used in subsequent galactosyl- transferase assays -when pyrophosphatase activity was high. A summary of the pyrophosphatase data is presented in Table 5. 98 Figure 15. Radiochromatographic Scans of the UDP-galactose Pyrophosphatase Assay - Time Course for Enzymatic Hydrolysis of UDP-Gal All assays contained 10 pg of pancreatic microsomal protein plus 4.0 mM UDP-Ga1. Panel A - No incubation Panel B - 15 minute incubation( 37 degrees C) Panel C - 30 minute incubation( 37 degrees C) _ v— A s l _- 99 »A. 0 min 3. I5 min C. 30 min ..m> o ....oo nail—5 1.06. do: -525 -mxo 100 Table 4 Effect of AMP and ATP on Pyrophosphatase Activity % UDP-Gal Remaining After: % UDP-Gal Assay 1. Hydrolyzed 0 min 15 min 30 min 30 min Boiled Enzyme 95.0 94.8 93.9 1.1 Enzyme 93.7 77.8 62.7 31 Enzyme + . Enzyme + Table 5 Effect of AMP and ATP on Galactosyltransferase Activity [14CIGa1actosg Transferred (d.p.m. ) per; Assay 25 pg Protein 40 pg Protein 50 pg protein Reaction Mix 1540 2025 2400 Reaction Mix + AMP(4.0mM) _ 1550 2000 2425 Reaction Mix 101 Figure 16. Radiochromatographic Scans for the UDP-galactose Pyrophosphatase Assay - Effect of AMP and Excess Unlabeled UDP-Gal on Pyrophosphatase Activity Panel A - Standard reaction mix plus 3.4mM AMP Panel B - Standard reaction mix plus 4.0 mM UDP-Gal(unlabeled) Panel C - Standard reaction mix All assays contained 10 pg of pancreatic microsomal protein. 102 A. ’UDP-Gal , ‘AMP ,U . l B. ‘UDP-Gol J/\L jbv ‘ / C. No additions / i i Dye - Origin- 1 UDP-Gal- ) Gal-l-P- Gol- 103 .mmuacws om Lmuwe cmccaooo anpupmo co :owuoumemmo mpaevomcaam oz .mcogpmz cw mew mmeacmooea mo mppmmmo .zeumsoeuomam covumppwucvom uvzcvp Sn umumuwucmac >a_>Puoeowume use use use mom: anpipew co Pawnee: o» mcpccoammceoo memem ms“ acmEQopm>mu cmuw< .msmcmoumeoeno emama mcwccmommu co :3; new mmu:=_s cm can .mp .o um umbm:_semu mom: use :quoea Fmsomoeops opammcoema co an op vm>wmome mxmmm< .pmu ion: cmpmnopcz uaoguwz emNVFPu: memz Aaz< to ab< My mmczuxvs cowuommc neonceum ae< eee az< so omoooepeu-ao= to m_m»_oees: omeoeeamoeaoesa to eoeeeepeel .AP menace 104 3.5 25... 8:335 $300.00 wan—.... n=z< SE Of. 0.... n.._.< SE 04V. 910 32:30 on I o\oON o\o CV ex. 00 o\o Om o\o 00. 105 Galactosyltransferase Levels in Rat Embryo Pancreatic Homogenates Rat embryos were dissected from Sprague~Dawley females whose pregnancies had been timed from fertilization. The small size of the tissues made microsomal preparations impractical at early embryonic periods and consequently whole pancreatic homogenates were used in the assays. Pancreatic tissue from S to 30 embryos (depending on embryonic age) was pooled, disrupted by sonication and assayed for enzymatic activity. Specific activity of DSG-fetuin galactosyltransferase as a function of embryonic age is given in Figure lB-A. The decline in specific activity from day 14 to birth was linear from 14 nmoles/mg protein/hour (day 14) to 5 nmoles/mg protein/hour (day 21, parturition at day. 21 to day 22). This decline was identical in the presence or absence of the pyrophosphatase inhibitor AMP. Tissue differentiation during development can occur with concommitent cell division and protein synthesis or by cell division with little "de novo" protein synthesis. Since the latter is known to occur in pancreatic development the specific activity of galactosyltransferase was calculated on the basis of cell number rather than total protein. Activity of galactosyltransferase as a function of embryonic age increased over the development period as shown in Figure 18-B. This duplicates the basic pattern first observed by Carlson (98). 106 .emnssc Ppmo we cowuocaw a we xu+>Fuom mzogm m szma ecu :Pmpoma as we :owuoczw a mo umumpsopmo xuv>+uoo mzocm < pmcma .mom oecoxensm mo cowuoczm m we Aaz< zs m.m H V mmumcmmoso; memeocea an; ovcoxgnsm cw umezmmms mom: moppp>vuoe mmmemmmcecppamobum—mw memeocma pom o_:ozcnsu mgu c? xaw>vao< mmmemwmcmcupxmogompmo .mp mezmwm 107 coztatem . 5 ON ..soe. m. o. h. 34 Scorecu m. d a .54 :s @996 o:<-.i. m2< SE was”. 0:0 m2puom mmmamcamocaoLAQ so mumeumnsm pounds: mzu we mvmzpocu»; op mac men az< H can mAmmmm :mm3umn mpm>m~ mmmcmmmcmeu namouoe—ma c? mocmemwmwo mew .az< zE o.m co mocmmmea mgu c? mco .czg mam; mmpa5om mueowpaac .muogumz cw vmnveommc mm zuv>wuom mmmememcmeapxmopoepem cemuoeqooxpm cow emxmmmm ucm um~vcmmosoc .muec Peumcomc ecu o'coxgasm soc» cm>oEme mom: memeocea momeocma pom Peaecomz use owcernsm mgu cp xpv>_uo< mmeememcmcupxmouompou :wmuoeaoozpw .mp meamwu 110 d----------------------- lll embryo (20 day) through neonates (day 6) were separated into soluble and membrane fractions. As shown in Figure 20, the activity was concentrated in the microsomal fraction with little activity being found in the microsomal supernatant. Consequently no evidence was found for a similar transition to a soluble form of the enzyme in the neonates. The possibility was then explored that an activator of galactosyltransferase was being synthesized during the neonatal rise in enzymatic activity. A series of mixing experiments were performed to determine if a soluble activator or‘inhibitor of galactosyltransferase was present during late embryonic or early neonatal .periods. No evidence was found for the presence of an activatdr or inhibitor of galactosyltransferase activity in the neonate day 3 pancreas. The possibility that a new distinct galactosyltrans- ferase was being synthesized in the neonates and that the elevated levels were the result of an additive activity between the embryonic and the adult form was explored by characterizing the neonatal activities by SDS-polyacrylamide gel electrophoresis and isoelectric focusing. Results supported one enzyme form with characteristics identical to those reported earlier for the adult form. Pyrophosphatase Activity in the Neonate Pancreas ~ Since the rise in galactosyltransferase activity at day 3 or day 4 was concommitant with a rise in apparent pyrophosphatase activity (indicated by the difference in activity between 112 .meezemz eA emnweomme mm A ez< zs o.e mape v mceAuoeee mmmsu so see memz mxemme mmmememceeeAxmeeompmw .A penancemeem a x oo.ooA V cewuoeee eceuecemeem Aesemeeows men can A emppme m x ooo.ooA v coweoeee Aesemeeors mge eucA emeeememm mem: .mAenme m>esme e» A m x coo.m v coweemaeweucmo emmem zep emeee ecu .em~_cmmeses memz mmAaEem memeocee uee Ameecem: ecu owcexeesm muep .zuA>Auoe mmeememcmequmeeoeAmm me mseee canon meeeesms ecu mAazpem cmmkeme eeeoe mceAupmcmeu eA mewsemume ee emeee :A memeecme Hem Pouncemz use ewcezeesm mew; mew ee mceweomee em_:AAmuue:m cA zep>wee< mmeemmmcmeupzmeuompec cAmueeeeozpo .cm meemwe 113 Eaten—men. oe< 38334 385$ mvnN__NoN m. m. S _ q q - - 3| d1 . m meeemoesez meeeo olo Eeeeeemeem 362.63.: .T- cezeeee 358222 I Q ‘2 N M ¢ N Jnoq / ugand 6w /se|ou.iu O Q" ll4 assays performed in the presence and absence of AMP), the samples were directly assayed for pyrophosphatase activity. The results of the pyrophosphatase activities are presented in Table 6. A distinctive rise in pyr0phosphatase activity 'was noted for the day 2 and day 3 samples with a high level of activity being noticed for day 3 as postulated. UDP-Galactosyltransferase Activity Assayed in Cultured Ceils The assay for UDP-galactose : glycoprotein galactosyl- transferase in cultured cells was identical to that used for pancreatic tissue. Quantity of material obtainable from tissue culture flasks precluded the preparation of membrane fractions and unless indicated otherwise all assays were done with whole cell homogenates. Assay Rquirements Assays were run on the homogenates of the following cell lines: KB,V-79,CHO,CHO-M,Nil~8,Nil-BHSV. In all cases examined the assay requirements were similar if not identical to those previously described for the pancreatic enzyme. 115 .meesemz :— m. exemme we eewueweommo .mmeueepaaueoz as _.N mcwcweeceo x—s :eweemme eeeeeeum e e. emeemems we: mwmxpeeezze A Ame m. eewuweeueee we mueae N o o o_=e< mp m o n has A m c N Am: N o o A ace are an :FE mp :mE D nemNAAeeea: mmeuoepewueoa w emm< maeeemz mesemeee—z ewuemeeeee an: peeecemz >5 mmeueepewneoa we mwm>peeezz e o_eee 116 Cell Cycle Dependence Glycosyltransferases have been reported to vary as a function of the cell cycle(l63,164). In order to examine glycoprotein galactosyltransferase activity as a function of the cell cycle KB cells were synchronized by a double thymidine block and galactosyltransferase activity assayed at two hour time intervals for a complete cycle ( 24 hours).- The results are shown in Figure 21. The incorporation of tritiated thymidine into a distint peak confirms synchronization. No significant variation of galactosyl- transferase activity occurred. Monosaccharides as Exogenous Acceptors Experiments were performed to determine if the membrane galactosyltransferase from cells in culture could utilize free glucose or free N~acetylglucosamine as substrate acceptors. This would constitute a difference from the pancreatic enzyme as no transfer to free N~acetylglucosamine occurred with enzyme from this tissue. This reaction is illustrated in Figure 22. The results of the experiments are summarized in Table 7. Clearly transfer to both glucose and N-acetylglucosamine occurred, with N~acetylglucosamine being the preferred acceptor by over a factor of four. 117 .meesumz :w mee mmeeemoeee we mpweemo .meae; ezu Aem>m emxemmm we; zuw>wuoe mmeemwmceeupxmeeompeo :wmueeeeoAAw .ceweeeeeeeocw mcwewexge emeewewee meweemmme a; emsewwceo we; cewue~weeecocxm .eeeAe mcwewsxcu m—eeee e we memes he em~weeezoczm memz mppme me mmezeumeA mAAmu me em~wceegoezm cw xew>weo< mmeemwmcmeuAAmeeeerw :wmeeeeeozpw .AN mezmwe 118 m 0. w- m 8 ..D d m Om m. n) ow .D. On xee_m msEEEH eceomm Ae mmem_mm .23 A885 mEfi 1 1 fi ON 9 w. v. N_ O. m w v N fi d H Idl dl |4 d 4 J 1 d d d1 d 1 d \\ // 01/! \\‘P/ \\ // Iflll‘llla‘\ // A‘ l / A 4, e V\/AAA/w + + +/ \+ / A /+\ +\\\\\+ + A AA A A , A 8:22:85 m:_e_EA£-In. , 38322230828 £29.68 + A 1 0. ON On 0? Om (—) Jnou/ugalmd bw/salowu 119 Figure 22. Lactose Synthetase ( UDP-Galactose: D- Glucose l-galactosyltransferase) This figure illustrates the reactions catalyzed by lactose synthetase, The enzyme will utilize N-acetyl- glucosamine to form N-acetyllactosamine, however in the presence of a modifier protein, alpha-lactalbumin, the Km for transfer to glucose is lowered and synthesis of lactose is favored. 120 CH 2OH H" N' 101 N H-O-P-O g-O- CH O O UDP- Galactose OH OH CHZOH Laciose Syntheiase GI N HO OH ucose UDP N-Aceiyiglucosamine ”OF 01-1on 01-12014 HO O O OH OH OH OH OH Lactose OH NHC'ECH3 N-Aceiyllaciosamine o 121 .moumcmwgo: :00 me. me: mopzom oping .3050: 5 congemmc ma 0.58095 momma 9:. 389.3 Loo m2 m4. mmz 55333031293 98 25m ohm: acmmmea cog: mcgmoogwdnemoiz ccm enemas .A .....eé ooomvmmouomamclmg o A 9.8 N55,. ...—:3 .Am.A. :3 Amaze mumazoooee E368 z mNd “3 cm 5 23200 mean—xv: amend 9:. omNN 0mm OOHH comm omH coca omfi Seamed: + mmoozg + 9535 we. mm mmoozac + mEemucm w: mm omalwzdg + mammogawgumoiz + @535 m; mm mcgmoezawgemoeqiz + gnaw mm mm emaimcag + mwoesau + 0.53% o 87932 + 255m o ecu—8803330312 + SENSE c HlxmBOQ g confide“ .z.m.o e mcofiuficcoo zoned ewseepeeee—regepe we meemme< ecu mocmmmee mzu cw meeeemoo< me. mepsemeeepm—Aumeeuz ecu mmeoepu guwz zew>wuo< mmeemwmeeeupxmeueepew A m—eew 122 Effect of Alpha-lactalbumin on Acceptor Specificity Alpha-lactalbumin interacts specifically with many galactosyltransferases to lower the Michaelus constant for glucose to make it more suitable as an acceptor (reviewed in reference 165). This same phenomenon occurred in the cellular enzyme and is illustrated in Table 7. Alpha~lactalbumin in the reaction mixture decreased transfer to N-acetylglucosamine to base line while increasing transfer to glucose by a factor of 3. Effect of Phorbol Ester Tumor Promoters on Galactosyltransferase Activity. Q The structures of the phorbol ester tumor promoters are illustrated in Figure 23 and their "in vivo" promoting activity is listed in Table 8. Nil-8 Cells After exposure to TPA, phorbol retinoic acid, and retinoic acid plus TPA for 12 hours Nil-8 cells were harvested, homogenized; and assayed for galactosyltrans- ferase activity. Results are shown in, Figure 24. The glycolipid sialyltransferase data were supplied by Dr. J. Moskal. Phorbol alone gave a rise in galactosyltransferase activity while the other agents gave activities that were similar to the control value. The glycolipid sialyltrans- ferase data deviated significantly from the glycoprotein 123 Figure 23. Structures of Phorbol Ester Tumor Promoters The chemical structures and abbreviations of the phorbol ester tumor promoters used in this investigation are illustrated. 124 PHORBOL ESTERS PRODUCT ABB. R1 R2 R3 R4 4 alpha - Phorbol 12,13- _ _ _ didecanoate dalpha PDD C9H19C0 C9H19Co H, OH Phorbol 12,13- diacetate PDA CH3C0- CHacO- H, OH Phorbol 4-O-methyl 12- _ _ myristate-lB-acetate MPMA Cut-12700 CH3C0 H, OH CH3 Phorbol 12,13- dibutyrate PDB CercO- C3H7CO H, OH Phorbol 12,13- didecanoate PDD CoHnCO- C9H19C0- H, OH Phorbol lZ-myristate-B- ‘ _ _ acetate PMML C1 3H21CO CH300 H, OH H 125 Ammo—v mac—.oom mucmpum .m.a .OJmOth sea ..u.u.m=agu ..a.s._ubo>a ++++ :22m .I. 9.5 I Cami: I Logm muwuoowlma mpmuwfigzslma Hoppozm muaocmomaficIMH.mH Hostess mpmtzusnfiaumfi.mfi Hostess mumumomlmm mumpmfitssumfi Hoatosafiszumeaons mpmumomfiaumfl.mfl Hostage muaocmomofionmfi.mfi Hoatocanazafiaa Hofioé sufi>fiuo< wcfiuosopm m rug—PH. O>H> CH casuafi>mtan< onwOHmc< Hoonozm mgouoeoca tease swumm "cases; a «Pane 126 .Apmxmoz .a .go an umnv>oga can: V gouamooa an we mcwsmrmopzmouomp um~wppuz mmmgmmmcmguxpmwm away—ooxpm ugh .mcogumz :_ nonprommu mm cmzmmmm mm: suP>Puom mmmtmmmcmgppzmouompmm coppmNpemmoeo; mcwzoppou .cowummzwwcucmo An cmuomppoo mppwo on» van mma :. comers cos» mew: mxmmpcnp one .cmEmowpoa gonna; a saw: m:_agmcom An omumw>smg cozy use: mppmo mgh .mma gap: was?» mugsu cmgmmz «to: mppmo on» use um>oemt mm: creme mg» mcamoaxm mega Louw< .mtao; Np so» Ape\m: o—V PMUFEmzo umpwwomam one mcvcpmucoo mwcms cw exogm «to: mppmo mappz mmmzaimod mppmu wnpvz :w huw>wpo< wmaLmecutepamouomgmw cwmuogaoozpo co evo< opocpumm use <¢h we uommcm use .em mtzmwm 127 (--.) Jnou/ugeiold bw/ selow u E O 00 O 03 am: Eo< T 204 eczema use? .83.: 4% .228 A q q — q emotemcezbgm 29.830 . gee? 38.625 5305820.? 1 0.. ON On 0..» on 00 ON (...) JL|/ ugaLOJd bw/selom u 128 galactosyltransferase activity in that sialic acid transfer to glycolipid was elevated by retinoic acid and elevated synergistically by retinoic acid plus TPA. -Nil~BHSV Cells Experiments identical to those performed on Nil-8 cells were done with a tranformed cell line Nil-8HSV. The results are illustrated in Figure 25. TPA and phorbol stimulated enzyme activity to approximately the same extent. Retinoic acid stimulated activity over that observed with TPA. No synergistic effect was seen with retinoic acid plus TPA. The glycolipid sialyltransferase activity .interestingly followed the same trend as did the glycoprotein galactosyl- transferase. KB Cells Similar studies-were performed with KB cells as were done with the Nil-8 and Nil-888V cells. Results are illustrated in Figure 26. Phorbol elevated galactosyltrans- ferase activity but to a lesser extent than did TPA. Retinoic acid elevated the activity to a greater extent. As in the Nil-888V cells the glycolipid 'sialyltransferase activity paralleled the glyc0protein galactosyltransferase activity. A pronounced negative synergistic effect was noticed with retinoic acid plus TPA. Addition of epidermal growth factor (EGF) gave a lower than control level of galactosyltransferase activity and EGF inhibited the 129 .Avm mgamwu mom v mppmo mupwz tom umnvgommw page ca soccae _mowycmuw cm cw magmas umpmwp mg» o» mszmoaxm gmbmm mav>wuum mmctmmmcmgppzmouom—am :vmuotaooxpm so» umxmmmm mum; mppmo >mzmupvz omeganmob n.2ao >m=m-_lz e. su.>_so< mmmtmmmcmtupxmouowpmw cpmuotaooxpw co upo< opocwumm can Puum ommgmmmcogapzmouumpmm cwmuogaouzpm go» umxmmmm wng m_pou mx wmmcaumog mPPmu mg =_ xpv>wpu< mmmgm$mcmgupxmouumpmm cwmuogaooz—o co upu< uwocwumm new (up we uummwm wch .om mgzmwm 732 ov 0.. U _ w o _ m " ON W N Om _ w w _ on a 5 E _ / d _ . w m “ ov 5 to d m. om . . m / _ on ...... m8 “ m J . U: ) _ om m ...oo. “ u; ( _ \ ON ,w _ o: x _ x \ _ x 0.0 + om. K l33 elevation seen with TPA and phorbol. A pronounced positive synergistic effect was seen with EGF plus retinoic acid. Effect of Cell Density on Tumor Promoter Alterations in Glycoprotein Galactosyitransferase Activity The effect of cell density on glycoprotein galactosyltransferase alterations by tumor promoters is shown in Figure 27. Two cell lines were used in these studies: KB and V~79 cells. The promoters were added when the cells were in either log~phase or confluent growth. Clearly no difference was noted between activities in confluent versus log phase cells. Mellitin , a polypeptide promoter, reacted similarly to TPA with no significant difference between confluent and log~phase cells. Effect of Tumor Promoters on Glycoprotein GalactosyItransferase Activity in CEO and a Membrane Mutant CHO Cell Lines The effects of TPA,retinoate, and’ butyrate on glycoprotein galactosyltransferase activity were examined in CHO and a membrane mutant CHO.cell line (CEO-M). Results are shown in Figure 28. TPA significantly elevated activity in CHO-M but not in CHO cells. Retnoate and butyrate significantly elevated activity in both cell lines with a 4 fold increase being seen in CHO-M cells and a 2-fold increase in CHO cells. 134 .mcosumz :* convtumwu mm zpv>wuom wmmgmwmcwgupxmouoopmm tom umxmmmm new umumm>tmg mew: mppmo .ucmapwcoo so Aacmzpucoo mom V mmosasmop mew: mppmu mzp can; mtmzmpocoe :F mcpzotm mppmo mg to m~u> o» umucm who: weapoEoga mmmgmmmcmgupxmouompoc cvmuogaooxpw m :_ mcopamgmup< cmoaucH gmuoeoga Leash so havmcma Fpmu we uooeem och .mm mgzmwe 135 _c::_c2 (ch .2125 .8200 q 2.3 at. ..> will}. :.3o.0 5330 a q omozm-o04 ¢:b .cv3::00 I ugasmd bfll'mdi’ 136 .uweeoeema mew: mxmmmm wmmememcmeupzmouwm_mm memmcmum cowam~ecmmoeog use acpumm>emc emuw< .muozumz cw amnveomwc mm A ZEcP.o uveom ovezmza new .Pe\m: cop -mumo:_umm ._e\a: opuesu< wmmewemcmee_emouuaemm :_wuoeqooxpu a co uvo< opezuam new .mumocpumm .wuum mmmemmmcweu -Fxmoummpmm =quoeaoo>Fm new mmumm>emz mew: mppwm meamoaxw eao; Np m emue< ._e\m: op um mppmw zuozu can ozu wmmzaumop ow vmnum mew: memaoEoea eoe=a ewamm _oaeose m~_ow z-o=u new ozw a. ewe>eau< wmmewemccep u—xmouompmc :_wuoenooz—w m co mewaoeoea emanp emamm poneoga mo powwwm wsh .mN mesmee 140 “26223. .385. .55 8e woe <5: <9. 8e 2.24, w d u 1 4 u - . , J , S E - - E. E5 - r a... c :1. J 2.8 2.95 _H. .. r. 2.8 0.6 § .. (ouuoo won aoualamo v. 141 Figure 30. Purification of a Plasma Membrane Fraction from Cultured Cells Plasma membrane fractions were purified from whole cell pellets by the procedure of Thom g3_gl,(215). The outline of the procedure is given in the figure. Details are described in Methods. 142 - PURIFICATION OF PLASMA MEMBRANE FRACTION Cell Pellet suspend in 2 volumes of: 0.05 M boric acid, 0.15 M NaCl 1 mM MgCIZ, lmM CaClz, pH 7.2 Add 100 volumes 0.02 M boric acid, 0.2 mM EDTA, pH 10.2 1. Stir 10 min 2. Centrifuge 450 x g for 10 min Whole cells, nuclei,,_ precipitated material ‘ Supernatant l. Centrifuge 12,000 x g for 30 min Soluble fraction 6 v Membrane-Rich Pellet 1. Suspend in phosphate buffered saline 2. Layer on tap of 35 percent sucrose MW in phosphate buffered saline 3. Centrifuge at 24, 000 x g for 60 min Plasma Membrane Fraction sucrose PBS interface 1. Centrifuge at 100,000 x g for 30 min v Plasma Membrane Fraction 143 .mcogpmz cw mem mpempwm xwmm< .umcveemumu Pm+emume wpnzpomcv news cue? :o_umeoaeowee men use uwpwmemce awe?» cowuaazwcw we“ we cmpm:.2emp mew: mAmmm< .mumcmmoeo; Ppmo mx mo m; cm ;u_: u mmmemmm e am can u mmmemmu um um cae mew: mammmm copumpxeogamoga mewccwum meammemQEmh we cowamcse m we cowumpaeoznmoge mzocmmoucm we mama we» .Pm we=m_e 144 “Sc: 22:. 5:33.: v. N. O. m m a. N ‘ q q J 1 q d. comm E :31 >82 0.1.0 00¢ «0 cam >0mw< I tom LOO. (3..le )owod.o 145 .mcogumz c? wem ammmm mo mpwmumo .A apFuumowcme can mmuzcee v emuem emumcpeemu mew: meowwommm .ep< eo coeueunm me» An mmumeuwcp mm: cowwwmme we» mews: span mop cm on emcesume cmcp man .u mmwemmu um um mwuacve N eow cmumnzoev cmzu mm: mezmxve xmmmm web .ah< mazes mezuxwe coeummme mumpaeow mam cu emcee mew: Ape\mc op V mewuoeoee .xmmmm ema cowuwmee meweBEme mammpa oxu we a: op new: age mew: mzmmmm copumpzeoggmoca uemucmum memuoeoea eossh emumm poneoga xn mcoeumemawee memeaemz mammpe ozu c. coeumpxeosamoze mzocmmoucm mo coeumemup< .mm mesmee 149 422a and can .22:er (2&2 (am 0018? 405.200 m. I—-l F S a 8 (010de br/‘N'a “ommodmuu a“ 5. o 9 150 .mm ccmmmp meamvu a? mem mpemuwo .Ammmm ewe coeuwmee memenEme mammpa zuozu eo m; op saw: cze mew: mammmm cowumpaeozmmoga cemmcmum memuoeoee eoszp emumu Foaeoge an mcovumemamea memeBEmz msmmpe zlozu cw covumpzeozamoze mzocwmoucm mo cavememap< .em mezmre 151 PDD PMML PDB PDA MPMA Promoter f.“ m CONTROL 4n PDD ee; weewwe meee; Np eew wueeauee as ep.o ea ewmeexm mew: mppme mmezeumee .xemme ewe ewmueee meeeBEme esme_e we a; ow new: eee mew: exemme eewue—xeesemege eeeeeeem mueeauem xe meewueeeemee meeeeewz eEmepe me use .zuozu .ozo cw eewuepxeezemeze meeemoeeem we eewueemuw< .em weemww 155 smiling KB (ouuoo + stalking PL |ouuoo CHO-Mutont as a E § l°l01d 151/we'd) palmodmouu an (3 is: g s 4» Lino Cell 156 higher endogenous level of phosphorylation than do CHO cells; however, in the presence of butyrate both had similar levels of incorporation. Autoradiograms of Assay Mixtures Run on PonacryIamide Gels To determine whether alterations of specific protein phosphorylation would occur in the presence of the different tumor promoters, whole cell reaction mixtures were run on SDS ' polyacramide gels. When stained for protein by Coomassie blue numerous bands were noted but no clear differences were noted with any promoter treatment (Figure 37). Autoradiograms of the gels revealed distinct differences between CH0 and CHO-M cells with different promoters (Figures 38,39). There was significantly greater incorporation into CHO-M cells than CHO cells (Figures 38,39 lanes 11 and 1). Cells grown in the presence of butyrate for 12 hours showed significantly greater incorporation than control cells (lane 9 vs. lane 17) In CHO cells little additional incorporation occurred when PDA was added exogenously to the 'system (lane 1). However, a large increase was obtained when cells were grown in its presence (lane 5). A similar but smaller effect was noted in CHO~M cells (lane 16 and lane 20). 157 .meoeawz ew wee mmeeemeeee we mpweuma .eweum me—m wwmmeeeeu saw: :wmaeee eew emeweam me: _wm wee mwmmeeeeeeuempm emmw< ..mu mewEe—Aeeexpeenmem as e we weep e ea ewwpeee eee .eeee emae: me—pwee e ew mwueewe m eew ewpwee .emwwee w—eEem we —;.em ea eweee me: ewmeeee we m: cm eewueNFemmeeez emuw< .mxemme ewueexNem eew meee x—peEeee me ewemw>eez eee meeeg N— eew eeuemwwm pee—Emse wee ea ewmeexm mew: m—pme mmeseumem a. wee— m weep e weep e wee— N mee— :20 a wee. N wee— m weep m mee— — wee. :lozu mzze