3 —_——-—————- 1W“ . -,~v v0 u I 'r- {—IV 'WJ‘QI h I Iuyv, varrv't‘f' ' '3 33E???” "HWHPJHT ,',, 3, m, 1" Ow :33: ’ " " " '3'” I.” I $1,; Vi 3‘3' 33'333303 ,' I,,' ' 3 '33,, 3 l I “S .‘," 3...} 33",. hazy , 4 .' 333' . ' ., lI ,,' , .‘,, 3. . d, ' """" 3- 3'-" I 3 3'1. .-- I * Izrgr: . . 1 , ». ,3... '. .,' 33:33- .‘333': 313333,. (3 "I.- 3,333,. ,,I,,'3,-,:, 33,3 MM , ' ’ 3 3 ‘ ' '33 I3333 3'3l33 3'3 .‘33 if," :' ‘;"".3"§.:i: :. 3" 3'3“.3 3.", 63,1' ., '3_,I3""3 ""' "' ' ----- .-,_3, 3 '3: 333... 33 ‘ "fl 33 -3 '3 I3' ""' 33"3 . "II-:3 '-';.,I,3..-. ,.,, .3,'3','33 '13:; .3. ‘ 33 333 3333! 33. l3 3 ,II- 33533 ,,3, .',,33 33". , .,3I 33:] ‘3‘,‘3~'"3'3 ."' ..3, “33, .333” 1'-'.:.- '- ,.,l,":33,‘ 31,333 I I ,v', H; , ,‘ "33L,” «3. 3' “'3' 3., ’ 313"” 33:, l‘» “I ”333' 33333333 333,33 ' 3,31‘1'1'. ,3 '3'3'3- r33. '53,, 3,? 3 3 3,. a - 4-. -..-. .5:— W 3 '. " ~'3 3 ". I I' _' H" "'I 3 . ' """ " "' '-'"'3 - II I""' 3' I" '3: 3 33 '.1‘,3,3 33,3, I'33('. ),., [I 333,-,” 3;“, , \ . , .. ' -‘ " "3" ,‘I "3 3'3. ‘ '3';-.31 2’3. ' 3" ""a' ,3, ”33'3“ '3'333 '3' 3"4'3 3 :I‘sgig'u3 :3 gI.-:.I .‘ .',; 5.5.5.33. 3,13,, 3.33133" :-"-" .1: ".~I.9'I"‘33-‘: 3 ,3. 3' ‘33333' 333.3 ("V ,. I'I. ',' . ,.. , 3. I' . - I . 333-3.... :3 33% I?.'-;, 3 3, -,.. .34., .'.‘,..'3'3. .-',-; .' " 33333., ‘ "'13333'3'33333, "’""I"I'33-..-:3I'..' .. 33"3",33’ "3 c,3‘|',' H3, l'3',‘,'3" 333,33 I .3 3., 3 ,, 3.33 3,33 3.333 , , ,3, 33",,3313 3,3, ,,,.,',', ”31,333,3,3,33,33,333.333,,33‘3',I ,3 'J‘géy', .3 I _ I 3 ,3'8'3'3' ‘,333, 3“, ,',‘,-,,,3 3, , ,3", J ,. '. .- .', (33 3333 3- 3'} 333333'33"'333,'3‘,3.3 33 .‘, " 333:3?“ ', J "‘ .- ‘- ’ ‘ I 33.3 433'”. ...'I:-t3I:3 3.3333 3" .I LIBRARY hddchigan15¢aba University This is to certify that the thesis entitled Glycoprotein Synthesis in the Rat Exocrine Pancreas During the Secondary Transition Period of Pancreatic Differentiation presented by Beverly Blum DaGue has been accepted towards fulfillment of the requirements for M.S. Biochemistry degree in Major professor Date September 23, 1977 0-7 639 GLYCOPROTEIN SYNTHESIS IN THE RAT EXOCRINE PANCREAS DURING THE SECONDARY TRANSITION PERIOD OF PANCREATIC DIFFERENTIATION BY Beverly Blum DaGue A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE. Department of Biochemistry 1977 .\ . 5‘” @1053??? ABSTRACT GLYCOPROTEIN SYNTHESIS IN THE RAT EXOCRINE PANCREAS DURING THE SECONDARY TRANSITION PERIOD OF PANCREATIC DIFFERENTIATION BY Beverly Blum DaGue Induction of glycoprotein synthesis in the rat exocrine pancreas during the major secretory protein synthetic period of pancreatic differentiation (the secondary transition) was studied using embryonic pancreatic rudiments cultured in the absence or presence of the differentiation inhibitor 5-bromodeoxyuridine from day 14 of gestation. Rudiments labeled with L[3H]fucose, D[3H]glucosamine, or [3H]N—acetyl-D-mannosamine at various times during the secondary transition period were analyzed by SDS-polyacrylamide slab gel electrophoresis with gel autofluorography and/or Coomassie blue staining. Radioimmunoassays were used to analyze labeled rudiments for "secretory protein-like" and "zymogen granule membrane-like" glycoproteins. Results suggest that only six or seven of the embryonic pancreas glycoproteins are induced during the secondary transition period. One of these, a high molecular weight zymogen granule membrane-like glycoprotein, appears to be induced only with respect to glycosyla- tion. The major zymogen granule membrane glycoprotein (GP-2) does not appear to be induced during the secondary transition period. Other induced glycoproteins are soluble species, two of which probably correspond to secretory proteins. to Michael Garrison DaGue ii There was a child went forth everyday, And the first object he look'd upon, that object he became, And that object became part of him for the day or a certain part of the day, Or for many years or stretching cycles of years. --Walt Whitman-- iii ACKNOWLE DGMENTS My sincerest thanks to those who have helped me most along the way, especially to Drs. Boyd O'Dell, Wayne Gonnerman, and Edward Harris, who provided a positive initiation into research and much encouragement; to Dr. Robert Ronzio, for his guidance as my thesis advisor and for his efforts to develop in me a scholarly approach to research; to Dr. Eugene Giroux, of Merrell International, Strasbourg, France, for the opportunity to work in his laboratory; to my parents and the DaGue family for their continuing support; and, most of all, to my husband, Michael, without whom I might have taken this work all too seriously. iv TABLE OF CONTENTS INTRODUCTION AND HISTORICAL . . . . . . . . . . . . . . . . . . Differentiation of the Rat Exocrine Pancreas . . . . . . Histodifferentiation of the Rat Exocrine Pancreas. . . . . . . . . . . . . . . . . . . . Cytodifferentiation of the Rat Exocrine Pancreas. . . . . . . . .‘. . . . . . . . . . . Biochemical Differentiation of the Rat Exocrine Pancreas . . . . . . . . . . . . . . . Changes in Glycoproteins During Cellular Differentiation. . . . . . . . . . . . . . . . . . . . Glycoproteins of the Adult Rat Pancreas. . . . . . . . . Glycoprotein Synthesis in the Differentiating Rat Pancreas: A Statement of the Problem. . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . . . . . . . . . . Materials 0 O O O O O O O O C O O O O O O O O O O O O O O Electrophoresis Reagents. . . . . . . . . . . . . Radiochemicals. . . . . . . . . . . . . . . . . . Reagents for Liquid Scintillation Counting. . . . Immunochemicals . . . . . . . . . . . . . . . . . Tissue Culture Media. . . . . . . . . . . . . . . Miscellaneous Reagents and Materials. . . . . . . Laboratory Animals. . . . . . . . . . . . . . . . Me thOdS o o o o o o o o o o o o o o o o o p o o o o o o o In vitro Culture of Embryonic Rat Pancreas. . . . Procedures for SDS-Polyacrylamide Slab Gel Electrophoresis and Gel Autofluorography. . . . In vitro Labeling of Adult Rat Pancreas with [3H]Leucine and [3H]Fucose. . . . . . . . . . . Preparation of Zymogen Granule Membranes and Zymogen Granule Content Proteins from Adult Rat Pancreas. . . . . . . . . . . . . . . . . . Preparation of Rabbit Antisera to Bovine Serum Albumin and to Rat Pancreas Zymogen Granule Membranes and Zymogen Granule Content Proteins. Preparation of Anti Zymogen Granule Membrane (IgG) . . . . . . . . . . . . . . . . . . . . . Preparation of Particulate and Soluble Fractions from Pancreatic Rudiments for Radioimmunoassay. Radioimmunoassay Procedures . . . . . . . . . . . Acid Precipitation of Radioactive Protein . . . . Immunodiffusion Analysis. . . . . . . . . . . . . Page 13 15 15 15 15 15 16 16 17 17 17 17 19 24 24 26 27 28 28 31 32 Page Liquid scintillation Counting . . . . . . . . . . 32 General Assay Procedures. . . . . . . . . . . . . 33 RESULTS 0 O O O O O O I O O O O O O I O 0 O O O O O O O O O O O 34 S-Bromodeoxyuridine Inhibition of in vitro Pancreatic Differentiation. . . . . . . . . . . . . . . . . . . 34 SDS-Polyacrylamide Slab Gel Electrophoresis Analysis of Protein and Glycoprotein Synthesis in Differen- tiating Pancreatic Rudiments . . . . . . . . . . . . . 35 Coomassie Blue Staining Profiles for Cultured Pancreatic Rudiments. . . . . . . . . . . . . . 36 Incorporation of [3H1Leucine into Cultured Pancreatic Rudiments. . . . . . . . . . . . . . 41 Incorporation of Radioactive Sugar Precursors into Glycoproteins of Cultured Pancreatic Rudiments . . . . . . . . . . . . . . . . . . . 42 Elution of Radioactive Proteins and Glyco- proteins from Gels During Staining and Destainin Procedures . . . . . . . . . . . . . 47 Incorporation of [ HlLeucine and [3H1Fucose into.Zymogen Granule Content and Zymogen Granule Membrane Constituents of Adult Rat Pancreas . . . . . . . . . . . . . . . . . . . 48 Fractionation of [3H]Fucose-Labeled Differentiated Pancreatic Rudiments . . . . . . . . . . . . . . . . 55 Incorporation of [ 3HlFucose and [3H]Glucosamine into Acid-Precipitable Protein During Pancreatic Differentiation. . . . . . . . . . . . . . . . . . . . 62 Radioimmunoassay for Zymogen Granule Membrane Glyco- proteins and Secretory Glycoproteins in Differen- tiating Rat Pancreas . . . . . . . . . . . . . . . . . 65 Characterization of the Anti Zymogen Granule Content (Antiserum) . . . . . . . . . . . . . . 66 Purification of the Anti Zymogen Granule Membrane (Gamma Globulin) . . . . . . . . . . . 75 Characterization of the Anti Zymogen Granule Membrane (Gamma Globulin) . . . . . . . . . . . 79 Accumulation of Antibody-Precipitable Glyco- proteins During the Secondary Transition Period of Pancreatic Differentiation. . . . . . 82 Identification of Antibody-Complexed Glycopro- teins of Differentiating Rat Pancreas . . . . . 89 Competition of Adult Rat Pancreas Zymogen Granule Membrane and Zymogen Granule Content Proteins with Glycoproteins of Differentiated Embryonic Pancreatic Rudiments for Antibody Binding . . . . . . . . . . . . . . . . . . . . 95 Radioimmunoassay of Pancreatic Rudiments Cultured in the Presence of S-Bromodeoxy- uridine for Zymogen Granule Content and Zymogen Granule Membrane Glycoproteins. . . . . 99 vi Page DISCUSSION. 0 O O O O O O O O O 0 O O O O O O O I O O O O O O 0 101 Summary of Results . . . . . . . . . . . . . . . . . . . 101 Glycosylation of Pancreatic Secretory Glycoproteins. . . 105 Metabolism of Glycoprotein Precursors. . . . . . . . . . 108 Biogenesis of Zymogen Granule Membranes. . . . . . . . . 109 Glycosyltransferase Induction During Pancreatic Differentiation. . . . . . . . . . . . . . . . . . . . 110 Pancreatic Glycoprotein Synthesis and the Model for Pancreatic Differentiation . . . . . . . . . . . . . . 112 Suggestions for Future Investigations. . . . . . . . . . 114 LI ST OF REFERENCES 0 O O O O O O O O 0 O O O O O O 0 O O O O O O 116 vii Table LIST OF TABLES Page Relative mobilities of standard proteins on 0.1% SDS-8% acrylamide slab gels. . . . . . . . . . . . . . . 22 Coomassie blue-stained proteins observed on SDS- polyacrylamide slab gels to be induced during the secondary transition period of rat exocrine pancreas differentiation. 0 O O O O O O O O O ‘ I O O O O O O O O O 39 Glycoproteins induced in the rat exocrine pancreas during the secondary transition period of pancreatic differentiation. . . . . . . . . . . . . . . . . . . . . 46 [3H]Leucine- and [3H]fucose-labeled components of adult rat pancreas zymogen granule membranes and zymogen granule contents . . . . . . . . . . . . . . . . . . . . 51 Fractionation of [3H]fucose-1abe1ed differentiated (in vitro) pancreatic rudiments. . . . . . . . . . . . . 56 Incorporation of radioactive sugars into acid- precipitable protein by pancreatic rudiments during the secondary transition period of differentiation . . . 63 viii Figure BA BB LIST OF FIGURES A model for the differentiation of the rat exocrine pancreas as proposed by Rutter and colleagues (Rutter et al. I 1968) O 0 O O O O O O O O I O O O O O I O O O O O SDS-polyacrylamide gel electrophoresis of [3H]fucose—, [3H]glucosamine- and [3H]1eucine-labeled pancreatic rudiments. . . . . . . . . . . . . . . . . . . . . . . . SDS-polyacrylamide gel electrophoresis of pancreatic rudiments labeled with [3H]N-acetylmannosamine at various stages of differentiation . . . . . . . . . . . . SDS-polyacrylamide gel electrophoresis of adult rat pancreas zymogen granule membrane proteins and zymogen granule content proteins labeled in vitro with [3H]leucine and [3H]fucose. . . . . . . . . . . . . SDS-polyacrylamide gel electrophoresis of subcellular fractions from cultured pancreatic rudiments labeled on the equivalent of 19-20 days gestation with [3H]fucose . . . . . . . . . . . . . . . . . . . . . . Electrophoretic comparison of fucoproteins from cultured pancreatic rudiments with the fucoproteins of zymogen granule membranes and the zymogen granule contents from adult rat pancreas . . . . . . . . . . . . Titration of anti zymogen granule content (antiserum) with [3H]leucine-labe1ed zymogen granule content proteins from adult rat pancreas. . . . . . . . . . . . . . . . . Ouchterlony double diffusion test of anti zymogen granule content (antiserum) and anti zymogen granule membrane (IgG) versus zymogen granule content proteins and zymogen granule membranes prepared from adult rat pancreas . . . . . . . . . . . . . . . . . . . . . . Microheterogeneity of anti zymogen granule content (antiserum). Double diffusion test of anti zymogen granule content (antiserum) versus various dilutions of a zymogen granule content protein solution. . . . . . ix Page 38 44 50 59 61 68 71 71 Figure 10 11 12 13 14 Page Cross-reactivity of anti zymogen granule content (antiserum) with smooth microsomal membranes and NaBr-washed rough microsomal membranes from adult rat pancreas . . . . . . . . . . . . . . . . . . . . . . 74 Purification of anti zymogen granule membrane (IgG). Removal of gamma globulin fraction cross-reactive with zymogen granule content proteins. . . . . . . . . . 78 Cross-reactivity of anti zymogen granule membrane (IgG) with smooth microsomal membranes and NaBr- washed rough microsomal membranes from adult rat pancreas . . . . . . . . . . . . . . . . . . . . . . . . 81 Accumulation of "zymogen granule content-like" glycoproteins and "zymogen granule membrane-like" glycoproteins in differentiating rat exocrine pancreas . . . . . . . . . . . . . . . . . . . . . . . . 85 Specific radioactivities of the "zymogen granule content-like" glycoproteins and "zymogen granule membrane-like" glycoproteins which accumulate in the rat exocrine pancreas during the secondary transition period of pancreatic differentiation. . . . . 88 SDS-polyacrylamide gel electrophoresis analysis of radioimmunoprecipitates of soluble and particulate glycoproteins from differentiating pancreas. . . . . . . 91 LIST OF ABBREVIATIONS 54BrdU 5-bromodeoxyuridine BSA bovine serum albumin Con A Concanavalin A EDTA ethylenediaminetetraacetate GP-l glycoprotein-l GP-2 glycoprotein-2 GP-3 glyc0protein-3 IgG immunoglobulin G SDS sodium dodecylsulfate xi INTRODUCT ION AN D H I STORI CAL Differentiation of the Rat Exocrine Pancreas Differentiation of the rat exocrine pancreas is considered to be an excellent model for mammalian secretory tissue differentiation. Differentiation of the rat exocrine pancreas may be viewed as the cumulative events of three major processes. These are l) histo- genesis, in which presumptive pancreatic cells divide and orient themselves with respect to each other and surrounding non-pancreatic cells; 2) cytodifferentiation, during which the generation of specific organelles involved in the matured functioning of the cell occurs; and 3) biochemical differentiation, during which the cell begins to synthesize and package those specific secretory proteins which identify it as an exocrine pancreas cell. The histogenesis and cytodifferentiation of the rat exocrine pancreas have been examined thoroughly and documented (Spooner et al., 1970; Pictet et al., 1972). The molecular events underlying these obvious structural changes and via which the exocrine pancreas cell acquires functional maturity still remain largely unknown. Substantial information is available regarding the accumulation of secretory proteins during differentiation. Post-translational modification of these secretory proteins during differentiation has not been investigated, however. Furthermore, the effects of differentiation on protein and glycoprotein constituents of cellular membranes, in particular those of the zymogen granule membrane, have remained l 2 essentially unstudied. Such information is certain to contribute to an understanding of the regulatory mechanisms involved in initiating and coordinating the molecular events of the differentia- tion process. Histodifferentiation of the Rat Exocrine Pancreas The pancreatic tissue arises from the epithelial cells of the embryonic endoderm. These cells are organized in a single layer with a continuous basal lamina and are connected to one another at their apices by junctional complexes. The dorsal pancreatic bud first appears by the twenty-somite stage (ten days gestation) and consists of a single layer of cells arranged in a bulbous protrusion from the dorsal side of the endodermal tube. The apices of the cells form a lumen. Continued cell divisioneand an increase in the height of the cells produce a thickening of the pancreatic bud which by the thirty-five somite stage (day twelve of gestation) consists of approximately three thousand cells. Though the tissue begins to appear stratified, the orientation of the cells remains the same. They continue to be connected to each other by apical junctional complexes, the apical cell surfaces forming a lumen. It is the orientation of cell division which maintains this relationship among the cells. The axis of the mitotic spindle is always parallel to the apical cell surface and apical junctional complexes are formed between daughter cells before cytokinesis is complete. By day 12 of embryonic development, lobulation of the pancreatic tissue is apparent. The luminal canal branches throughout the tissue; however, it is small and difficult to detect. The pancreatic epithelium remains surrounded by basal lamina which separates it 3 from the mesenchyme. In the days which follow, the branching pancreatic tissue gradually penetrates into the mesenchyme. Day 15 marks the obvious appearance of the pancreatic lumen. With the development of the lumen, histodifferentiation is complete. Cytodifferentiation of the Rat Exocrine Pancreas The development of the mature subcellular structure of the exocrine pancreas parallels the biochemical differentiation of the cell. Between the twenty-four somite stage (middle of day 11 of gestation) and day 14 of embryonic development, the cell resembles other endodermal cells of the gut. This stage is called the "protodifferentiated state." Cells at this stage are characterized by 1) apical junctional complexes and luminal microvilli, 2) both cytoplasmic ribosomal clusters and ribosomes extensively bound to rough endoplasmic reticulum, 3) a rough endoplasmic reticulum consisting of cisternae which are not oriented with respect to the cell nucleus or the cell surface membrane, 4) a golgi apparatus consisting of flattened saccuoles located between the cell nucleus and the apical cell surface, and 5) the presence of mitotic struc- tures as cells undergo division. On day 15 of embryonic development, when histodifferentiation of the pancreas is essentially complete, the cells begin to acquire the structural characteristics of mature cells. This is considered the beginning of the transition to the "differentiated state." Not all exocrine cells of the pancreas undergo this transition simul- taneously; however, the same maturation process occurs for all exocrine pancreas cells. 4 The transition requires about three days. Several noticeable changes occur in cellular organization and organelle structure. The rough endoplasmic reticulum increases in size and the cisternae become oriented around the nucleus. Condensing vacuoles may be found within the golgi stacks. By day 17 in most exocrine pancreas cells, zymogen granules have begun accumulating between the golgi and the apical surface membrane. Cytodifferentiation is essentially complete by day 17 of gesta- tion. Only a continued accumulation of zymogen granules is apparent during the remainder of the transition. Biochemical Differentiation of the Rat Exocrine Pancreas Concomitant with the histogenesis and cytodifferentiation of the rat exocrine pancreas, occurs the "functional" differentiation of this cell type. Three steady states for the biochemical differen- tiation of the exocrine pancreas have been proposed by Rutter and his colleagues (Rutter et al., 1968; Kemp et al., 1972). The transi- tions from one steady state to the next are determined by the levels of secretory proteins present in the acinar cells. Figure 1 illustrates schematically the biochemical differentiation process proposed by Rutter and colleagues. At about the twenty-four somite stage, presumptive exocrine pancreas cells undergo a transition from the "predifferentiated state" in which none of the specific secretory products can be detected in the cells, to the "protodifferentiated state" in which the cells are characterized by low but detectable levels of the secretory enzymes and proenzymes. Exactly when this transition occurs is not known. The levels of the secretory proteins in the .umu mason we :m: “you cuonso: ma :mz: .Amoma ..Hm um nouusmv mwsmmoaaoo can umuusm >2 venomoum mm moonocmm ocfluooxm new era mo coflumwucmuommwp new you Hmpoe m .H onsmflm H muswwm 369 and. oEofoEw 3:602 305 count-.2035 < . mz m. o. o a c2329.... toEtdlv 22m 6222282095 c0539.... Concooowlv 32m 02229325295 1 :03.»th 32:8. 225.. ._‘ :34 Jetooioqo engiogiuaiamq ;o uouonuaouog 7 protodifferentiated pancreas range from 103 (for lipase A) to 105 (for amylase and chymotrypsinogen) molecules per cell (Rutter et al., 1968). The protodifferentiated state lasts until day 14 or 15 of gestation, at which time the second regulatory transition occurs and the rates of synthesis of the secretory proteins increase dramatically. While the rate of total protein synthesis increases negligibly on a per cell basis in vitro, the rates of synthesis of amylase and chymotrypsinogen increase lOO-fold during the secondary transition period (Kemp et al., 1972). Accumulation curves for the various secretory proteins indicate a synchronous rather than coordinate induction since not all of the secretory proteins are synthesized at an increasing rate at exactly the same time. The differential effect of actinomycin D on the accumulation of the various secretory proteins in cultures of developing pancreatic rudiments also suggests independent regulation of the synthesis of the individual secretory proteins. This effect of actinomycin D, furthermore, suggests that messenger RNAs coding for the secretory proteins are sequentially synthesized during the secondary transition. Secretory proteins fail to accumulate in pancreatic tissue cultured in the presence of the thymidine analogue 5-bromodeoxy- uridine (S-BrdU) if it is administered beginning with day 14 or day 15 of embryonic development. 5—BrdU is incorporated into DNA and subsequently alters RNA synthesis (Walther et al., 1974). That 5-BrdU completely blocks differentiation of the exocrine pancreas is further evidence for the dependence of the secondary transition on specific RNA synthesis. 8 The third steady state of the differentiation profile for exocrine pancreas is attained by day l9of gestation. At this time most secretory proteins have accumulated to the levels found in the adult rat pancreas. In the mature pancreas, the levels of secretory‘ proteins synthesized remain high but are subject to modulation in response to variation in diet. Changes in Glycoproteins During Cellular Differentiation Changes in glycoproteins, in particular cell surface membrane glyc0proteins, have been correlated with cellular differentiation. For example, alterations in the types of glycoproteins and their relative exposure to the external environment of the cell have been shown to occur during the development of embryonic chick neural retinal cells using carbohydrate-binding phytoagglutinins of different sugar specificities (Kleinschuster and Moscona, 1972). Cultured murine neuroblastoma cells possess a cell surface membrane glycoprotein and a soluble glycoprotein, each with an apparent molecular weight of 105,000 daltons, only in the differentiated state (Truding et al., 1974). Red blood cell maturation in the rat appears to involve altera- tions in the relative concentrations of sialated and nonsialated glycoproteins on the cell surface. Con A binding by surface membrane glycoproteins (nonSialated) of the mature erythrocyte is much less than that observed for the precursor erythroblasts. The opposite variation occurs for sialated surface membrane proteins (detected with positively charged ferric oxide) with the mature erythrocytes bearing a much higher concentration of surface membrane sialoproteins than the erythroblasts (Skutelsky and Farquhar, 1976). 9 Dictyostelium discoideum, 6-18 hours after the initiation of differentiation (when the amoebae have aggregated and the formation of a fruiting body has begun), exhibits a large increase in the intensity of labeling (by iodination) of a specific plasma membrane glycoprotein. This glycoprotein has a molecular weight of approxi- mately 150,000 daltons and binds Con A. At the same time that the 150,000-dalton species appears in large quantities on the plasma membrane, a plasma membrane glycoprotein of 180,000 daltons becomes virtually undetectable (Geltoskey et al., 1976). Of greatest significance to a consideration of glycoproteins in the differentiation of the exocrine pancreas are the results of studies of intestinal epithelial cells. Both cell types have their embryonic origin in the gut endoderm. The lumen of the small intestine is penetrated by a vast number of villi. A single layer of cells covers each villus, and from the tip of each villus to its base there is a gradual decrease in the degree of differentiation of the epithelial cells. Those epithelial cells at the tip of the villus are the most completely differentiated while those at the base of the villus, the crypt cells, are the least differentiated and are mitotically active. Over a period of thirty-six to seventy-two hours, crypt cells move up the villus to its distal end. During this time they undergo differentiation to the mature cell. Methods have been devised for isolating epithelial cells from various levels along the villus (Weiser, 1973) so that it is possible to assay for enzymes and markers of the cells in their various stages of differentiation. A number of significant findings have been reported. 10 First, the specific activities of the hydrolytic enzymes sucrase and alkaline hydrolase are greatest in differentiated cells at the tips of the villi while they are lowest in the non-differentiated crypt cells. Furthermore, from villus tip cells to crypt cells there is a gradual decrease in the specific activities of these enzymes. In vivo incorporation of radioactive sugars into glyco- proteins by intestinal epithelial cells also appears to parallel the degree of differentiation of the cells. D—glucosamine, L-fucose, and D-galactosamine are incorporated, in vivo, to the greatest extent by differentiated cells at the villus tip. Cells isolated from the tips of villi also incorporate much greater amounts of radioactive D—glucosamine than do isolated crypt cells during in vitro incubations with the sugar. Subcellular fractionation of the epithelial cells from the villus tip indicates that most of the glucosamine—labeled glycoproteins appear in the microvillus plasma membrane fraction and that the hydrolytic enzymes are associated with these glycoproteins (Weiser, 1973). Secondly, when cells are isolated from the various levels of the villi and assayed for cell surface glycosyltransferase:endogenous acceptor activity, completely differentiated cells are found to have the lowest transferase activity for the nucleotide sugars GDP-fucose, GDP-mannose, UDP-N-acetylglucosamine, UDP-galactose, and UDP-glucose. Only the CMP-sialic acid transferasezendogenous acceptor-activity is greater in differentiated villus tip cells than in crypt cells. By definition of the transferase activities, undifferentiated crypt cells must then not only have the highest levels of transferase activities, but in addition must possess on their cell surfaces incomplete glycoprotein acceptors for glycosylation by these various ll sugars. When fetal intestinal cells and chemically-induced intestinal tumor cells are assayed for cell surface galactosyltransferase:endogenous acceptor activity, they exhibit activities similar to those of the crypt cells. The lectin binding ability of intestinal epithelial cells is also a function of cellular differentiation. Crypt cells, fetal cells, and tumor cells are agglutinated by Con A, whereas differentiated villus cells are not (Weiser, 1972 and 1974). Etzler and Branstrator (1974), using fluorescein isothiocyanate-labeled lectins, have observed a decreasing ability of differentiated cells to bind various lectins other than wheat germ agglutinin on villi ten centimeters or more distant from the pyloris. This is contrasted with the ability of less differentiated cells to bind lectins from Lotus tetragonolobus (specific for terminal non-reducing a-L—fucose) and Ricinus communis (Specific for terminal non-reducing a— or B-D-galactose). Completion of cell surface membrane glycoproteins with B-N-acetylglucosamine (specifically bound by wheat germ agglutinin [LeVine et al., 1972]) and N-acetyl-neuraminic acid (bound nonspecifically by wheat germ agglutinin [Greenaway and LeVine, 1975]) appears to be related to the state of differentiation of intestinal epithelial cells. Glycoproteins of the Adult Rat Pancreas Previous studies in this laboratory have made it apparent that glycoproteins are important constituents of the adult rat exocrine pancreas. All of the membrane subfractions of the exocrine pancreas cell bear glycoprotein components (Ronzio et al., submitted for publication). Comparative carbohydrate analysis indicates that of all the membrane structures present in the cell, zymogen granule 12 membranes have the highest ratio of carbohydrate to protein (0.44) on a mass basis. A "plasma membrane-enriched" fraction is also found to be highly glycosylated with a carbohydrate to protein ratio of 0.23. Zymogen granule membranes and the "plasma membrane—enriched" fraction are also the most enriched in protein-bound fucose and galactose. Each subcellular fraction isolated from rat exocrine pancreas labeled in vitro with radioactive glucosamine exhibits specific glucosamine-labeled glycoproteins when analyzed with SDS-polyacrylamide gel electrophoresis. (Plasma membrane, it must be noted, has not been examined by these techniques.) Zymogen granule membranes bear three major glucosamine-labeled proteins with molecular weights of 120,000 (GP-l), 74,000 (GP-2), and 52,000 (GP—3) daltons. GP-2 is by far the most intensely labeled component, containing two-thirds of the radioactivity recovered from polyacrylamide gels. Smooth microsomal membranes derived from the Golgi complex bear a glucosamine- labeled glycoprotein of the same molecular weight as GP-2 in the zymogen granule membrane. Though the relative carbohydrate contents of the post-microsomal supernate and the zymogen granule content are low by comparison with the membrane fractions from exocrine pancreas, each contains a minimum of three molecular weight classes of glucosamine-labeled glycoproteins. Approximate molecular weights of these glycoproteins are 68,000, 55,000, and 25,000 daltons (Ronzio et al., submitted for publication). Furthermore, in vitro stimulation of secretion in pancreatic tissue pre-labeled with radioactive fucose or glucosamine has shown that a linear discharge of soluble glycoproteins occurs 13 parallel with the secretion of radioactive leucine—labeled proteins (Vblkl et al., 1976; Kronquist et al., 1977). Glycoprotein Synthesis in the Differentiating Rat Exocrine Pancreas: A Statement of the Problem It had been observed in this laboratory that embryonic pancreatic rudiments cultured to different stages of development incorporated radioactive L-fucose into a large number of glyc0proteins. Some of these glyc0proteins appeared to be induced during pancreatic differen- tiation. Others, notably a glycoprotein(s) with the electrophoretic mobility of the zymogen granule membrane glycoprotein GP-2, were present in both protodifferentiated and differentiated pancreas in relatively the same amounts (Ronzio, unpublished observations). In addition, it has been reported by Walther (1972, 1974) that l) S-BrdU, which prevents the secondary transition in developing exocrine pancreas, does not inhibit the synthesis of glycoproteins associated with membrane functions; 2) glycoprotein profiles on SDS-polyacrylamide gels for glucosamine-labeled control and 5-BrdU cultures of embryonic pancreas are similar; and 3) possibly only one secretory protein with the electrophoretic mobility of amylase or procarboxypeptidase A is glycosylated. (A minimum of three glycosylated zymogen granule content proteins were observed by Ronzio and co-workers [submitted for publication],) Though these observations are not entirely contradictory, it seemed that the answers to some questions about glycoprotein synthesis during pancreatic differentiation needed clarification. How many glycoprotein species are characteristic of exocrine pancreas at different stages of development? Can different species be observed by labeling pancreatic cultures with different glycoprotein sugar l4 precursors? Are all these glycoproteins constitutive proteins of the exocrine pancreas or is the synthesis of some of them induced during the secondary transition period of differentiation? Are there glycosylated secretory proteins? What are the relative rates of synthesis of the zymogen granule membrane glyc0proteins and the secretory proteins at different stages of pancreatic development? Most important of all, how does glycoprotein synthesis during the embryonic development of the pancreas fit into the "Concerted" mechanism for pancreatic differentiation proposed by Rutter and his colleagues(l968)? This mechanism requires only the two regulatory transitions shown in Figure l to account for the differentiation of the exocrine pancreas with respect to secretory protein synthesis. Does this mechanism account for glyc0protein synthesis in the dif— ferentiating exocrine pancreas as well? The work reported in this thesis was designed to answer these questions. MATERIALS AND METHODS Materials Electrophoresis Reagents Sodium dodecylsulfate (sequenal grade) was obtained from Pierce Chemical Co., Rockford, Illinois. Acrylamide (99%), N,N'-methylene- bisacrylamide, and N,N,N',N'-tetramethylethylenediamine were obtained from Ames Co., Elkhart, Illinois. Ammonium persulfate was from Canalco, Rockville, Maryland. Sigma Chemical Co., St. Louis, Missouri, was the source of Coomassie brilliant blue R and B-mercaptoethanol. Bromphenol blue was obtained from Nutritional Biochemicals Corp., Cleveland, Ohio. Radiochemicals D—[6-3H(N)]glucosamine hydrochloride (9 Ci/mmole), L-[6-3H1fucose (12 Ci/mmole), [3H(G)]N-acetyl-D-mannosamine (2.2 Ci/mmole), L[4,5-3H(N)]leucine (60 Ci/mmole), and [3H]H20 (2.03 x 105 dpm/ml, calibrated 9/14/74) were obtained from New England Nuclear, Boston, Massachusetts. Reagents for Liguid Scintillation Counting Triton X-100, 2,5-diphenyloxazole, and 1,4—bis[2-(4—methyl-5- phenyloxazolyl)]-benzene were obtained from Research Products International Corp., Elk Grove Village, Illinois. 15 16 Immunochemicals Goat antirabbit (gamma globulin) was obtained from Calbiochem, La Jolla, California. Rabbit antisera to bovine serum albumin (crystallized, Sigma Chemical Co.) and to subcellular fractions of adult rat pancreas had been prepared in the laboratory (Ronzio and Mohrlok, 1977). Complete Freund's adjuvant was obtained from Difco Laboratories, Detroit, Michigan. Tissue Culture Media Powdered Hanks' Balanced Salt Solution (with NaHC03), powdered Nutrient Mixture F-12 (with L-glutamine, without NaHCO3), Minimal Essential Medium Amino Acids Solution (50x, without L-glutamine), Minimal Essential Medium Non Essential Amino Acids Solution (10 mM, 100x), and Antibiotic-Antimycotic Solution (100x, containing 10,000 units Penicillin, 10,000 ug Streptomycin, and 25 ug Fungizone per ml) were obtained from Grand Island Biological Co. (GIBCO, Grand Island, New York) prepared according to their formulations. S-Bromodeoxyuridine was obtained from Sigma Chemical Co. Bovine Plasma Albumin, Fraction V, was from Miles Laboratories, Inc., Elkhart, Indiana. Dulbecco's Phosphate Buffered Saline was prepared as a 10x solution from reagent grade salts according to the formula- tion in the GIBCO catalog, except for the substitution of CaClZ-ZHZO for anhydrous CaClz and the addition of 0.1 g phenol red (Sigma Chemical Co.). All powdered media were prepared according to the package directions and supplemented as described in Methods. Media were filtered through sterile Millipore (0.22 pm) or Gelman Metricel (0.20 um) filters before use. 17 Miscellaneous Reagents and Materials Sucrose (ultrapure) was obtained from Schwarz/Mann, Orangeburg, New York. Soybean trypsin inhibitor and cyanogen bromide-activated Sepharose 48 were obtained from Sigma Chemical Co. Dimethyl sulfoxide was obtained from both Aldrich Chemical Co., Inc., Milwaukee, Wisconsin, andMallinckrodt, St. Louis, Missouri. ICN Pharmaceuticals, Inc., Plainview, New York, was the source of copper ethylenediamine- tetraacetate. Sephadex G200 was from Pharmacia Fine Chemicals, Inc., Piscataway, New Jersey, and DEAE cellulose.(CellexPD) was from Bio-Rad Laboratories, Richmond, California. Soluble starch was obtained from Mallinckrodt. Polyethylene glycol (average molecular weight 6000-7500) was obtained from Matheson Coleman and Bell, Norwood, Ohio. Kodak X-Omat R x-ray film came from Picker x-Ray, Detroit, Michigan, and Kodak RP Royal X-Omat X-ray film came from Grand X-Ray, Grand Rapids, Michigan. Laboratory Animals Sprague-Dawley rats, including timed-pregnant female rats, were obtained from Spartan Research Laboratories, Haslett, Michigan. Methods In vitro Culture of Embryonic Rat Pancreas The method used for culturing dorsal pancreatic rudiments from embryonic rats was essentially that described by Ronzio and Rutter (1973), except for the omission of chick embryo extract from the culture medium. Timed-pregnant Sprague-Dawley rats were sacrificed by decapitation on day 14 of gestation and the uteri transferred to sterile plastic petri dishes containing sterile Hanks' Balanced Salt 18 Solution. Iridectomy knives were used to dissect the pancreases from the embryos. Dissection was carried out in sterile plastic petri dishes containing sterile F-12 medium supplemented with 4% (v/v) Minimal Essential Medium Amino Acids Solution (50x). (This represented a supplement of two times the amounts of the essential amino acids present in Eagle's Medium [Eagle, 1959].) Pancreatic rudiments, with some adhering mesenchyme, were transferred to 2 nmrwide Millipore filter (type THWP) strips glued across the wells of a section of Linbro plastic tray (MVC-96) set into a sterile petri dish. Four to eight rudiments were placed on each filter strip. The culture wells were filled with approximately 0.25 ml of culture medium consisting of F-12 supplemented with 4% (v/v) Minimal Essential Medium Amino Acids Solution (50x), 1% (v/v) Antibiotic-Antimycotic Solution (100x), and L-glutamine to a final concentration of 2.0 mM. Culture dishes were kept under a humidified atmosphere of 95% air-5% C02 in a 37°C incubator. The culture medium was changed daily. When rudiments were cultured in the presence of 5-BrdU, 2 x 10"3 M S-BrdU in water was added to the culture medium to a final concentration of 2 x 10.5 M. Cultured rudiments were labeled with radioactive sugars on the equivalent of days 14, 17, and 19 of gestation. The appropriate volume of [3H]sugar was transferred to a sterile disposable plastic tube and taken to near dryness with a filtered stream of nitrogen. Sterile culture medium (plus or minus 5-BrdU) was then added to the tube to give a final [3H]sugar concentration of 25 uCi/ml. The culture medium was then replaced by approximately 0.25 ml of this radioactive medium. Rudiments were labeled for 18 hours. Incubation of rudiments with [3H]leucine was carried out in a similar manner. 19 The final leucine specific radioactivity was 12.4 x 106 dpm/ug (Kemp et al., 1972). At the end of the labeling period, cultured rudiments were harvested. The radioactive medium was aspirated from the culture wells and the wells were rinsed once with Dulbecco's Phosphate Buffered Saline (pH 7.0-7.4). The filter strips with adhering pancreatic rudiments were then removed from the wells and placed into a small amount of Dulbecco's Phosphate Buffered Saline in a petri dish. Using the iridectomy knives, the rudiments were gently scraped or lifted off the filter strips. The rudiments were then transferred to plastic microfuge tubes and centrifuged for 5-15 seconds in a Beckman model 152 microfuge. Usually one to two litters of rudiments were pooled per tube. Residual medium was aspirated from the tubes and the tissue pellets were frozen on dry ice and stored at -80°C. Procedures for SDS-Polyacrylamide Slab Gel Electrophoresis and Gel Autofluorography Gel Preparation. The Laemmli (1970) SDS-polyacrylamide gel system was used, with some modifications, for all slab gel electro- phoresis analyses. Slab gels, 1.5 mm thick, consisted of a one- centimeter stacking gel above a nine-centimeter separating gel. A 30% acrylamide-0.8% N,N'-methy1enebisacrylamide stock solution was used in preparing both the separating gel and the stacking gel. The separating gel of 8% acrylamide contained 0.373 M Tris—HCl (pH 8.8) and 0.1% SDS. Polymerization was with tetramethylethylenediamine and ammonium persulfate at final concentrations of 0.17% (v/v) and 0.08% (w/v), respectively. The stacking gel of 3% acrylamide 20 contained 0.116 M Tris-RC1 (pH 6.8) and 0.1% SDS and was polymerized by the addition of tetramethylethylenediamine and ammonium persulfate to final concentrations of 0.024% (v/v) and 0.09% (w/v), respectively. The concentrations of polymerizing agents in both the stacking and separating gels were modifications of the Laemmli system made by Gahmberg and Hakomori (personal communication). The day before use, gels were cast between glass plates rinsed with 2% Photo-flo 200 (Kodak). Sample Preparation. For electrophoresis, usually one to two litters of radioactively-labeled pancreatic rudiments were solubilized in 4% SDS and then mixed with an equal volume of sample buffer con- taining 0.124 M Tris-RC1 (pH 6.8), 20% (w/v) glycerol, 0.002% bromphenol blue, and 10% (w/v) B-mercaptoethanol. All samples contained 2% SDS, 0.062 M Tris-HCl (pH 6.8), 10% (w/v) glycerol, 0.001% bromphenol blue, and 5% (w/v) B-mercaptoethanol when applied to the gels. The prepared samples were boiled 2-5 minutes before application to the gels. Electrophoresis Conditions. The same buffer was used for both upper and lower reservoirs. This buffer (pH 8.3) consisted of 0.025 M Tris, 0.192 M glycine, and 0.1% SDS. Gels were electrophoresed at room temperature in a Bio-Rad Laboratories Model 220 slab gel electrophoresis unit. Run time was either 6-8 hours at 25 ma per gel or overnight at a low voltage setting. After overnight runs, the voltage was usually increased on the following morning to 75-100 volts. Gels were electrophoresed until the bromphenol blue tracking dye had migrated 8.5 to 9.0 cm from the top of the separating gel. 21 Gel Staining Procedure. After completion of electrophoresis, some gels were stained in a solution of 0.1% Coomassie brilliant blue R in 30% (v/v) methanol—10% (v/v) glacial acetic acid (Bonner and Laskey, 1974). Gels were stained 16-24 hours at room temperature. Destaining was accomplished in 30% (v/v) methanol-10% (v/v) glacial acetic acid at room temperature, usually within 48 hours. Stained gels appearing in Figures 2A, 28, 5A, 5B, and 14A were stained 24 hours, destained for 48 hours, and then stored at 0-4°C in 10% (v/v) glacial acetic acid until photographed. Procedures for Gel Autofluorography. Gels were prepared for autofluorography according to the procedure outlined by Bonner and Laskey (1974). Gels were first soaked in twenty times their volume of dimethyl sulfoxide for thirty minutes at room temperature. This soaking was repeated for an additional thirty minutes in fresh dimethyl sulfoxide. Gels were then immersed in four times their volume of 20% (w/w) 2,5-diphenyloxazole in dimethyl sulfoxide for three hours. Subsequently, the gels were rinsed in distilled water and then rehydrated for one hour in twenty volumes of distilled water. The gels were then dried on Whatman 3MM filter paper using a Bio—Rad slab gel drier Model 224. Both stained and non-stained gels were prepared for autofluorography by this procedure. Autofluorographs were obtained by exposure of either Kodak X-Omat R or RP Royal X-Omat X-ray film to the dried gels. In complete darkness a dried gel was sandwiched with a sheet of x-ray film between two glass plates held together by metal clamps. The gel-film sandwich, placed in a small manilla envelope, was wrapped in multiple layers of aluminum foil and stored at -80°C. X-ray films were 22 exposed for varying lengths of time depending on the radioactivity of samples applied to the gels. Films were not exposed to a hyper- sensitizing light flash before use (Laskey and Mills, 1975). X-ray films were developed in a Kodak RP x-Omat Rapid Processor. Molecular Weight Calibration of Gels. Protein standards were included on some of the gels that were stained before preparation for autofluorography. The standard proteins that were used and their molecular weights are listed in Table 1. An average relative mobility Table 1. Relative mobilities of standard proteins on 0.1% SDS-8% acrylamide slab gels Standard Molecular Average Relative Protein Weight Reference ,Mobility Thyroglobulin 165,000 DeCrombrugghe et 0.089 (porcine) al., 1966 Phosphorylase A 94,000 Weber and Osborn, 0.335 (rabbit muscle) 1969 ‘ Transferrin 74,000 Roberts et al., 0.445 (human) 1966 Serum albumin 68,000 weber and Osborn, 0.478 (bovine) 1969 Fumarase 49,000 Weber and Osborn, 0.670 (porcine heart) 1969 a-Chymotrypsinogen 25,700 Weber and Osborn, 0.91 (bovine pancreas) 1969 for each protein was obtained from measurements made on dried gels and/or photographs taken of the stained gels before the gels were prepared for autofluorography. The relative mobility of a protein 23 is defined as the ratio of the migration distance of the protein band (measured from the top of the separating gel to the center of the protein band) to the migration distance of the bromphenol blue tracking dye (similarly measured). All protein standards were run on a minimum of two gels. A plot of log molecular weight as a function of relative mobility for the standard proteins was linear. A least squares analysis supplied the slope (-0.9623) and y-intercept (5.3015) parameters for this plot. The following equation was used to calcu- late molecular weights of unidentified proteins on experimental slab gels: Molecular Weight = antilog[(-0.9623)(Relative Mobility) + (5.3015)] From this equation, it is estimated that proteins with molecular weights in the range of 22,000 to 200,000 are resolvable in this gel system. If the molecular weights of the standard proteins are in turn calculated using this equation, values within 10% of the known molecular weights are obtained. This is the degree of accuracy normally assigned to protein molecular weight determinations based on migration in SOS-polyacrylamide gels. It will be noted that ovalbumin appears as a standard in several of the figures presenting stained gel profiles in Results. Ovalbumin was not used, however, in the least squares calculation described above. The electrophoretic behavior of ovalbumin deviated measurably from the linear relationship between log molecular weight and relative mobility observed for the other standard proteins employed. 24 . In vitro Labeling of Adult Rat Pancreas with [3B1Leucine and [3H]Fucose Pancreatic tissue taken from adult Sprague-Dawley rats was minced into small fragments. The minced tissue was transferred to 50-ml Erlenmeyer flasks (approximately 2 g per flask) in complete Krebs-Ringer bicarbonate buffer (Krebs, 1950) containing 1% bovine plasma albumin, 1% (v/v) Antibiotic-Antimycotic Solution (100x), 2.5 mM Ca2+, 1.2 mM Mg2+, 0.01% soybean trypsin inhibitor, 14 mM glucose, and a complete Minimal Essential Medium (Eagle's) amino acid supplement (Amsterdam and Jamieson, 1974). Tissue was incubated at 37°C for 4 hours in 125 uCi L-[4,5-3H(N)]1eucine or 50 uCi L-[6-3H1fucose in a total volume of 5 ml medium per flask. Periodi- cally throughout the incubation period, the flasks were aerated with moistened 95% air-5% C02. At the conclusion of the incubation period, the labeling medium was aspirated from the flasks and the minced tissue was pooled and washed extensively with 0.3 M sucrose containing 0.25 mg/ml soybean trypsin inhibitor. Preparation of Zymogen Granule Membranes and Zymogen Granule Content Proteins from Adult Rat Pancreas Zymogen granule membranes and zymogen granule content proteins were obtained from unlabeled pancreatic tissue or washed labeled , tissue by the fractionation procedure of Ronzio and colleagues (submitted for publication) based on the procedure of Meldolesi et a1. (1971). All Operations were carried out at 4°C except where noted. The tissue was homogenized in ten volumes of 0.3 M sucrose containing 0.25 mg/ml soybean trypsin inhibitor with four strokes of a glass-teflon Potter-Elvehjem homogenizer driven at 620 revolu- tions per minute. The homogenate was filtered through several 25 layers of cheesecloth and centrifuged at 500 x g to sediment nuclei and tissue debris. To sediment zymogen granules, the 500 x g supernate was centri- fuged at 1730 x g for 30 minutes. The tops of the zymogen granule pellets were washed with 0.3 M sucrose containing 0.25 mg/ml soybean trypsin inhibitor to remove co-sedimenting mitochondria, then resuspended in the sucrose-soybean trypsin inhibitor solution and centrifuged for 30 minutes at 1730 x g. (In one preparation, the first 1730 x g supernate was centrifuged at 1730 x g for another 30 minutes. Zymogen granules from this centrifugation were washed, resuspended, and pooled with the first washed zymogen granule pellet. A final centrifugation at 1730 x g for 30 minutes was used to sediment all zymogen granules.) Final zymogen granule pellets were suspended in a small volume (0.5-1.0 m1) of 0.15 M NaCl. This suspension was added dropwise to 15 times the volume of 0.2 M NaHCO3, and the zymogen granules were allowed to lyse at room temperature for approximately ten minutes. The lysed suspension was layered over a discontinuous gradient con- sisting of 4 ml of 1.0 M sucrose-0.25 mg/ml soybean trypsin inhibitor. topped with 4 ml of 0.3 M sucrose-0.25 mg/ml soybean trypsin inhibitor in a nitrocellulose tube. The gradients were centrifuged at 196,000 x g for 1 hour. The top 4 m1 of solution containing zymogen granule content proteins were removed from the gradients, pooled, and stored at -80°C. Two-thirds of the 0.3 M sucrose-0.25 mg/ml soybean trypsin inhibitor layer were pipetted from the tubes and discarded. The zymogen granule membrane felt at the interface of the two layers of sucrose was transferred with a Pasteur pipet to a second nitrocellulose 26 tube and diluted with an equal volume of 0.25 M NaBr. The NaBr- membrane suspensions were sonicated with four to five pulSes (each approximately 5 seconds in duration) of the small probe of a Bronwill sonicator. The tubes were then filled with 0.25 M NaBr and centrifuged at 272,000 x g for 2 hours. The NaBr wash was aspirated from the tubes and the zymogen granule membrane pellets frozen on dry ice and stored at -80°C. Preparation of Rabbit Antisera to Bovine Serum Albumin and to Rat Pancreas Zymogen Granule Membranes and Zymogen Granule Content Proteins Antisera to crystalline bovine serum albumin (BSA), rat pancreas zymogen granule membranes, and zymogen granule content proteins were supplied by Ms. S. Mohrlok of this laboratory. They were obtained by immunization of New Zealand white rabbits. Immunization to BSA and zymogen granule content proteins was obtained by injections of these proteins (1-2 mg) dissolved in 0.15 M NaCl and mixed with equal volumes of complete Freund's adjuvant. Zymogen granule membranes (0.4-0.5 mg) were suspended in 0.15 M NaCl plus an equal volume of complete Freund's adjuvant for injections (Ronzio and Mohrlok, 1977). Rabbits were bled by puncture of the inner marginal vein of the ear. Collected blood was allowed to clot at room temperature or 37°C for 30 minutes and then stored overnight at 0-4°C. The serum was decanted from the clotting tube into clean centrifuge tubes and centrifuged (at 4°C) for thirty minutes at 1000 x g to sediment red blood cells. The serum.was stored at -20°C. Antisera to BSA and zymogen granule content proteins were heated at 56°C for thirty minutes to inactivate complement. 27 Preparation of Anti Zymogen Granule Membrane (IgG) The gamma globulin fraction from antiserum to zymogen granule membrane was prepared according to the procedures of Fraker, Cicurel, and Nisonoff (1974) and Levy and Sober (1960). A known volume of antiserum was precipitated with sufficient 25% NaZSOQ (at 25°C) to give a final NaZSOn concentration of 18%. The precipitation mixture was centrifuged at room temperature for 30 minutes at 1265 x g. The supernate was discarded. Water was added to the precipitate and then sufficient 25% Nazsog was added to make the final NaZSOn concen- tration 14%. The precipitation mixture was centrifuged as before and the supernate decanted. The final precipitate was dissolved in a small volume of borate-saline buffer (pH 8.0-8.2, containing 0.13 M NaCl, 0.16 M boric acid, and 0.02 N NaOH) and transferred to dialysis tubing. Dialysis was against several changes of this same buffer for 44 hours at 4°C. In preparation for DEAE cellulose chromatography, dialysis was continued against 0.018 M phosphate buffer (pH 6.9) containing 1.05 g Na2HPOu and 1.5 g NaHzPOquZO per liter. The dialyzed preparation was centrifuged to remove insoluble material. To remove possible contaminating beta globulins, the dialyzed anti zymogen granule membrane (gamma globulin) preparation was chromatographed on a column (9.5 cm x 1.5 cm, i.d.) of DEAE cellulose equilibrated in 0.018 M phosphate buffer (pH 6.9). Peak fractions (determined by absorbance readings at 280 nm) were pooled and trans- ferred to a dialysis bag. The dialysis bag was then dredged in polyethylene glycol (average MW 6000-7500) to concentrate the gamma 28 globulin preparation. The concentrated preparation was dialyzed against borate-saline buffer (pH 8.0-8.2), then stored at -20°C. Preparation of Particulate and Soluble Fractions from Pancreatic Rudiments for Radioimmunoassay Borate-saline buffer (pH 8.4-8.5, containing 5.0 mM boric acid, 1.25 mM sodium borate, and 0.142 M NaCl) was added to plastic micro- fuge tubes containing [3H1fucose- or [3H]glucosamine-1abeled pancreatic rudiments at a ratio of 75 ul buffer per 10-13 rudiments. The tubes were capped, sealed with parafilm, and taped to the bottom of a shallow plastic dish to which ice water was added. The large probe of a Bronwill sonicator, set at 80% maximum probe intensity, was run up and down the side of the tube, usually for one or two minutes,to completely suspend the tissue. The sonicates were then centrifuged at 111,000 x g for 2 hours at 4°C. The supernate, designated the "soluble fraction", was removed from each tube using a Hamilton syringe. Each pellet was resuspended by sonication in 0.25 M NaBr. The sonicates were centrifuged as before. The resultant supernates were discarded and the pellets, designated the "particulate fraction", were resuspended with sonication in NaZHPOu-detergent buffer (pH 7.4, containing 10 mM NaZHPOH, 1% (v/v) Triton X-100, 1% sodium deoxycholate, 5 mM disodium EDTA, 12 uM phenylmethylsulfonyl- fluoride, and 0.001% butylated hydroxytoluene [Ronzio and Mohrlok, 1977]). Radioimmunoassay Procedures Precipitin Reactions. Double antibody quantitative precipitin reactions were used to determine the levels of glycosylated secretory 29 proteins (glycosylated zymogen granule content proteins) and glycosylated zymogen granule membrane-like proteins present in pancreatic rudiments at different stages of development. Assays for the soluble glycoproteins were carried out in standard borate-saline buffer (pH 8.4-8.5) containing 5.0 mM boric acid, 1.25 mM sodium borate, and 0.142 M NaCl. A phosphate-detergent buffer (pH 7.4), containing 10 mM NaZHPOQ' 1% (v/v) Triton X-100, 1% sodium deoxycholate, 5 mM disodium EDTA, 12 uM phenylmethylsulfonylfluoride, and 0.001% butylated hydroxytoluene, was used for all immunoassays of particulate membrane glycoproteins. Final assay volume in all cases was 300 pl. Radioimmunoassay reactions for secretory glycoProteins in the soluble fractions prepared from pancreatic rudiments were carried out in the following manner. First buffer was aliquoted into plastic microfuge tubes and then various dilutions of the soluble fractions in borate-saline buffer were added to the tubes. Total volume was 90 ul. Ten microliters of decomplemented rabbit anti zymogen granule content (antiserum) was next added to the tubes followed by vortexing. Reaction mixes were incubated 30 minutes at 37°C. Two hundred micro- liters goat antirabbit (gamma globulin) (reconstituted in borate- saline buffer) were added to the reactions, and the reaction mixes were then incubated at 25°C for one hour. Subsequently the tubes were stored at 0-4°C for a minimum of 24 hours. Radioimmunoassays for zymogen granule membrane glycoproteins in particulate fractions prepared from pancreatic rudiments were carried out as follows. Phosphate—detergent buffer was aliquoted into plastic microfuge tubes followed by dilutions of particulate fractions (in phosphate-detergent buffer). Total volume was 90 ul. Ten microliters of anti zymogen granule membrane (IgG) (treated to 30 remove cross-reactivity with secretory proteins) were added to the tubes and the tubes were vortexed. The reactions were incubated at 37°C for 30 minutes. After vortexing, the tubes were stored at 0-4°C for a minimum of 16 hours. One hundred thirty microliters of phosphate-detergent buffer and 70 ul of goat antirabbit (gamma globulin) (reconstituted in phosphate-detergent buffer) were then added to the reaction mixtures. Following a one-hour incubation at 25°C, the reaction mixtures were stored a minimum of 25 hours at 0-4°C. Control reactions for measuring nonspecific adsorption of labeled proteins or trapping of free radioactive sugar by the immunoprecipitate were prepared in which 10 ul of decomplemented rabbit anti BSA (antiserum) were substituted for anti zymogen granule content (antiserum) and 40 ul of anti BSA (antiserum) diluted 1 to 10, were substituted for anti zymogen granule membrane (196) . Collection of Precipitates. Precipitin reaction mixtures were layered on t0p of 5 m1 of l M sucrose (dissolved in the appropriate buffer) in conical centrifuge tubes (Rosen et al., 1975). The reaction tubes were rinsed twice with 200 pl of reaction buffer, and these rinses were pooled with the precipitin reaction mixtures in the centrifuge tubes. The tubes were centrifuged at 2000 x g for 10 minutes (0-4°C). The sucrose solution was removed from the tubes and the precipitates were resuspended in one milliliter of the appropriate buffer. Centrifugation at 2000 x g for 5 minutes was used to sediment the precipitates. The precipitates were washed a second time in the same manner. After removal of the final wash, 31 the tubes were inverted and stored several hours or overnight at 0-4°C to drain. Washed and drained precipitates were solubilized in 200 pl of 1% SDS-0.1 N NaOH by heating at 60°C for one hour. Samples of the solubilized precipitates were taken for liquid scintillation counting. The results were expressed as the percent of total acid-precipitable dpm initially added to the precipitin reaction precipitated by the antibody. The percent radioactivity precipitated by the control anti BSA (antiserum) was subtracted from the experimental values. Acid Precipitation of Radioactive Protein Protein-bound radioactivity of labeled pancreatic rudiments sonicated in 4% SDS and of rudiment subfractions in borate-saline or phosphate-detergent buffer was determined by acid precipitation. One hundred microliters of carrier BSA solution (500 ug protein) was added to tubes containing known volumes of samples. Cold 20% trichloroacetic acid was then added to the tubes to a final concen- tration of 10%. Precipitation reactions were left at 0-4°C for at least one hour (usually overnight). Tubes were centrifuged at 2000 rpm for 10 minutes (0-4°C) in a Sorvall GLC-2 centrifuge. The acid soluble fraction was removed from each tube and the precipitate was washed three times by suspension in 200 pl cold 10% trichloroacetic acid followed by centrifugation. The final wash was aspirated off the precipitate. Washed acid precipitates were solubilized in 200 pl 1% SDS-0.l N NaOH with heating at 60°C for one hour. Samples of the solubilized precipitates were taken for liquid scintillation counting. Recovery of ”acid-soluble" plus "acid-precipitable" radioactivity was usually 32 approximately 80% of the total radioactivity of the original samples. Contamination of the precipitates with acid-soluble radioactivity was estimated at less than 1% of the total radioactivity based on the results of acid precipitation of known amounts of [3H]fucose and [3H]glucosamine in the presence of carrier protein. "Percent Acid-Precipitable Radioactivity" was calculated as the percent of total original dpm found in 10% trichloroacetic acid precipitates. Immunodiffusion Analysis Immunodiffusion analyses of anti zymogen granule content (antiserum) and anti zymogen granule membrane (190) were carried out according to the method of Ouchterlony (1958) in 1% agarose gels. Zymogen granule content proteins and zymogen granule membrane used to cross react with the antibodies were prepared as described. Microsomal membrane subfractions (from adult rat pancreas) used in cross-reactivity studies were supplied by Ms. S. Mohrlok of this laboratory. They were prepared according to published procedures (Jameson and Palade, 1967; Ronzio, 1973a and 1973b). The smooth microsomes had been taken through the NaCl-NaHCO3 lysis step but had not been washed with 0.25 M NaBr. Rough microsomal membranes were NaBr-washed. Liguid Scintillation Counting All radioactive samples were counted in a non-refrigerated Hewlett Packard Model 2405 scintillation counter using the tritium quick set or in a refrigerated Hewlett Packard Tri-Carb Model 3310 with a gain setting of 46%. Counting efficiencies were generally similar for the two instruments. Counting efficiency was determined 33 by the inclusion of "cold“ samples (prepared identically to the experimental samples) to which a [3H]H20 standard was added, or by first counting the experimental samples and then adding [3H1H20 standard and recounting. Samples were counted in 10 m1 of a scintillation fluid consisting of two parts toluene (containing 0.015% 1,4-bis[2-(4-methyl-5-phenyl- oxazolyl)]-benzene and 0.82% 2,5-diphenyloxazole) and one part Triton X-100 (Patterson and Greene, 1965). Usually 0.9 ml 1% SDS- 0.1 N NaOH and 0.1 ml 1 N HCl were included in the vials. General Assay Procedures Protein analyses were carried out according to a micromodifica- tion of the Lowry protein assay (Lowry, 1958; Rutter, 1967). Crystalline BSA was used as a standard. Samples containing Triton X-100 were analyzed for protein by the method of Wang and Smith (1975) using crystalline BSA as a standard and phosphate-detergent buffer (pH 7.4) for sample dilution. Volumes of all reagents were reduced by a factor of five to accommodate small sample volumes of 40 ul or less with protein values on the order of 1 ug/ul. Amylase was assayed in rudiment sonicates by a micromodification of the Bernfeld (1955) method. Volumes of assay components were reduced by a factor of ten. Maltose hydrate was used to prepare a reference standard curve. RESULTS S-Bromodeoxyuridine Inhibition of in vitro Pancreatic Differentiation To test the effectiveness of 5-bromodeoxyuridine (5-BrdU) as an inhibitor of pancreatic differentiation, equal numbers of pancreatic. rudiments from 14-day rat embryos were cultured for five days in the absence or presence of 2 x 10"5 M 5-BrdU. Harvested rudiments were sonicated in distilled water and samples were analyzed for amylase activity and protein. The level of amylase found in control rudiments was 20.1 ug eq maltose/pg protein/minute, whereas the level of amylase in S-BrdU-treated rudiments was only 1.62 ug eq maltose/pg protein/ minute. Thus, amylase activity was reduced by a factor of twelve in rudiments cultured in the presence of 5-BrdU. The amount of total protein accumulated per S-BrdU-treated rudiment, however, was essentially the same as that for control rudiments (18 and 19 ug/ rudiment, respectively). These results are not entirely consistent with the findings of Walther and co-workers (1974), who observed that S-BrdU added.to the growth medium at a concentration of 2 x 10-5 M inhibited the accumulation of amylase activity by a factor of twenty-eight and produced half-maximal inhibition of total pancreatic protein accumu- lation. However, differences in age of the embryos at the time of dissection, embryo dissection technique, number of rudiments sampled, and culture technique all may have contributed to the 34 35 apparent differences in effectiveness of 5-BrdU in inhibiting pancreatic differentiation. The results reported here do indicate that S-BrdU was inhibiting the accumulation of at least one pancreas specific protein well enough that the system could be used to study glycoprotein synthesis during pancreatic differentiation. Accumulation of zymogen granules in control rudiments during differentiation in culture gave a cobbled appearance to the rudiments when they were examined under the dissecting microscope. S-BrdU- treated rudiments did not acquire this structural feature. The absence of large numbers of zymogen granules in S-BrdU-treated rudiments was a further indication that S-BrdU was effectively inhibiting the differentiation of the rudiments. The appearance of fluid-filled vacuoles in pancreatic rudiments cultured in the presence of 5-BrdU was also an indicator of the effectiveness of S-BrdU in inhibiting exocrine cell differentiation. Rutter and colleagues (Githens et al., 1976) have shown that these vacuoles are lined by duct cells, the proportion of which increases dramatically in S-BrdU-treated pancreatic rudiments. SDS-Polyacrylamide Slab Gel Electgpphoresis Analysis of Protein and Glyggprotein Synthesis in Differentiating Pancreatic Rudiments Induction of specific protein and glycoprotein synthesis during pancreatic differentiation was first assessed by SDS-polyacrylamide gel electrophoresis of pancreatic rudiments labeled at various times during the secondary transition period of differentiation with radioactive leucine or radioactive sugars. Rudiments were cultured for 18 hours in the presence of the radioactive precursors. Radio- active sugars used to label the rudiments were L[3H]fucose, 36 [3H]N-acetyl-D-mannosamine, and D[3H]glucosamine. (Fucose and N-acetylmannosamine are normally incorporated at or near the non- reducing ends of glycoprotein carbohydrate chains, N-acetylmannosamine being incorporated as one of the N-acylneuraminic acids. Glucosamine, as a general label for the carbohydrate of glycoproteins, is incor- porated throughout the chains as N-acetylglucosamine or N-acetylgalactosamine and terminally as N-acylneuraminic acids.) Control rudiments were labeled at 14 days of gestation or after three or five days in culture (the equivalent of l7-day and l9-day rudiments, respectively). Rudiments cultured in the presence of S-BrdU were also labeled after three or five days in culture. Labeled rudiments were sonicated in 4% SDS and samples were prepared for electrophoresis on 0.1% SDS-polyacrylamide slab gels. Approximately the same amount of radioactivity for each sample was applied to the gels. Gels were stained with Coomassie brilliant blue and/or prepared for autofluorography. Coomassie Blue Staining Profiles for Cultured Pancreatic Rudiments The Coomassie blue staining profiles for fucose- and glucosamine- labeled pancreatic rudiments are shown in Figure 2A. Arrows indicate either distinct proteins or heterogeneous unresolvable bands of protein induced during differentiation. In this system, "induced" proteins are those proteins which appear in the culture equivalent of 19- to 20-day control rudiments at significantly greater levels than in 14- to 15-day control (protodifferentiated) rudiments and rudiments cultured in the presence of S-BrdU to the equivalent of 19-20 days gestation. 37 Figure 2. SDS-polyacrylamide gel electrophoresis of [3H]fucose-, [3nglucosamine-, and [3H]1eucine-labeled pancreatic rudiments. Samples of cultured pancreatic rudiments, labeled at different stages of differentiation, were electrophoresed on 0.1% SDS-8% acrylamide slab gels as described in Methods. Gels were stained and/or prepared for autofluorography. Coomassie blue staining profiles for labeled rudiments are shown in Figure A. Figure B is the fluorograph prepared from the stained gel in Figure A. Figure C is the fluorograph of an identically prepared gel that was not stained before preparation for autofluorography. Fluorographs of [3H1fucose- and [ 3H]glucosamine-labeled samples were obtained from films (RP Royal X-Omat) exposed to the prepared gels for 22 days. Fluorographs of [3H]1eucine- labeled samples were from films (RP Royal X-Omat) exposed to prepared gels for three days. Gels were prepared so that on the basis of initial radio- activity measurements, 20,000 cpm were applied for samples B-G and 200,000 cpm were applied for samples H and I. The actual amounts of acid-precipitable radioactivity for each sample applied to the gels (dpm values in parentheses) were determined after electrophoresis. Lane A shows the protein standards, 1) porcine thyroglobulin (165,000), 2) human transferrin (74,000), 3) bovine serum albumin (66,000), 4) ovalbumin (43,000), and 5) bovine pancreatic ribo- nuclease A (13,700) at the bromphenol blue tracking dye. Lanes B-D 3show the profiles for cultured pancreatic rudiments labeled with [3H1fucose on the equivalent of B) 14-15 days gestation (28 mg protein; 52, 900 dpm), C) 19-20 days gestation (40 ug protein; 47,800 dpm), and D) 19-20 days gestation, 5-BrdU culture (33 ug protein; 42,100 dpm). Lanes E-G show the profiles for cultured pancreatic rudiments labeled with [ H]glucosamine on the equivalents of E) 14-15 days gestation (39 ug protein; 36,600 dpm), F) 19-20 days gestation (10 ug protein; 40,800 dpm), and G) 19-20 days gestation, S-BrdU culture (14 ug protein; 32, 800 dpm). Lanes H and I show the profiles for cultured pancreatic rudiments labeled with [3 Hlleucine on the equivalent of H) 19-20 days gestation (2. 9 ug protein; 450, 000 dpm) and I) 19-20 days gestation, 5-BrdU culture (2.5 ug protein; 540,000 dpm). Arrows indicate differentiation-induced protein and glycoprotein species. 38 C 1...... A 1:. IITTTTIITTTV 0 2 4. >b3303 w>.hk_.=§ w>.h<4m¢ >243: w>_h<4w¢ 39 A minimum of 29 protein species can be identified in this gel system as early as day 15 of pancreatic development. Stained protein profiles, as expected, are similar for corresponding samples of fucose- and glucosamine-labeled rudiments. The majority of the proteins that are detected appear to be constitutive. Six induced Coomassie blue-stained proteins are detected. The relative mobilities and calculated molecular weights of these proteins are summarized in Table 2. Table 2. Coomassie blue-stained proteins observed on SDS- polyacrylamide slab gels to be induced during the secondary transition period of rat exocrine pancreas differentiation Relative Mobility Molecular Weight Tentative Identification 0.47 70,000 0.47-0.49 67,600-70,700 0.59-0.61 51,800-54,200 Amylase (56,000; Sanders, 1970) 0.61-0.63 49,600—51,800 Procarboxypeptidase B (50,000; Sanders, 1970) 0.70-0.72 40,600-42,400 Lipase (37,000-43,000; Vandermeers and Christophe, 1968) 0.72-0.73 39,700-40,600 Procarboxypeptidase A (36,000; Walther et al., 1974) 0.93-0.94 24,900-25,500 Chymotrypsinogen (23,000; Sanders, 1970) or Trypsinogen (24,500 in bovine pancreas; Kay et al., 1961) Among the proteins known to be induced during differentiation of the rat exocrine pancreas are eight of the secretory proteins. (More than a dozen proteins have been resolved for pancreatic secretions from various animal species.) These proteins are amylase, 40 procarboxypeptidase B, lipases, procarboxypeptidase A, chymotrypsinogen, trypsinogen, and ribonuclease. With the exception of the proteins with relative mobility 0.47 and 0.47-0.49, the induced proteins in Figure 2A may be tentatively identified as secretory proteins on the basis of their estimated - molecular weights and relative abundance. In the order of increasing mobility, these induced proteins are probably amylase, procarboxy- peptidase B, lipase(s), procarboxypeptidase A, and chymotrypsinogen and/or trypsinogen. The induced proteins_with relative mobilities of 0.47 and 0.47-0.49 correspond in molecular weight (W 69,000) to a rat pancreas zymogen granule content glycoprotein(s) previously observed by Ronzio and co-workers (submitted for publication) and to an unidentified differentiation-induced rat pancreas protein observed by Walther and co-workers (1974). Ribonuclease is expected to have migrated with the tracking dye in this gel system as did bovine pancreatic ribonuclease A which has a molecular weight similar to that of the rat pancreas enzyme (13,000 daltons; Walther et al., 1974). Thus, the differentiation-dependent induction of ribonuclease could not be determined in this gel system since a number of low molecular weight proteins are expected to migrate with the tracking dye. The molecular weights of the induced proteins observed in Figure 2A correspond with the "known” molecular weights of the assigned secretory proteins within the 10% error range imposed upon molecular weight determinations based on protein migration in SDS-polyacrylamide gels (Weber and Osborn, 1969). Protein glycosyla- tion can result in electrophoretic mobilities which indicate 41 molecular weights which differ from those determined by other techniques (Segrest and Jackson, 1972). Incpgporation of [3H]Leucine into Cultured Pancreatic Rudiments The incorporation of [3Hlleucine into the equivalent of 19- to 20—day embryonic pancreatic rudiments cultured with and without the differentiation inhibitor 5—BrdU was also used to study general protein induction during pancreatic differentiation. As is shown by the fluorograph profiles in Figure ZB, Lanes H (control) and I (S-BrdU-treated), a large number of proteins can be detected in - embryonic pancreatic rudiments labeled with [3Hlleucine. The complexity of the profiles makes it somewhat difficult to assess which labeled proteins are "induced" proteins. Only four protein bands can be identified with any certainty as induced. The relative mobilities of these proteins are 0.47-0.50 (66,100-70,700 daltons; this band was seen as a doublet in one gel with components having relative mobilities 0.47-0.48 and 0.48-0.50), 0.58-0.62 (50,700- 55,400 daltons), 0.69-0.72 (40,600-43,400 daltons) and 0.89-0.93 (25,500-27,900 daltons). Each of these induced [3H]1eucine-1abeled proteins appears to correspond to an induced protein in the Coomassie blue-stained gel of Figure 2A, as expected on the basis of work from Rutter's group (Kemp et al., 1972). The latter three proteins are tentatively identified as amylase, lipase, and chymotrypsinogen/ trypsinogen. Even though there are slight variations in the measured ranges of mobilities from those of the induced proteins of Figure 2A, the relative positions of the induced [3H]leucine-1abe1ed proteins in the fluorograph are the same as those for the Coomassie blue-stained induced proteins. O 42 Incorporation of Radioactive Sugar Precursors into Glycoproteins of Cultured Pancreatic Rudiments Gel autofluorographs showing [3H]fucose- and [3H]glucosamine- labeled glyc0proteins from differentiating pancreatic rudiments are presented in Figure 28. Similarly, in Figure 3 is shown the gel autofluorograph of [3HJN-acetylmannosamine-labeled glycoproteins from cultured pancreatic rudiments at several stages of differentia- tion. (These fluorograph profiles for each of the radioactive sugar precursors of pancreatic glycoprotein synthesis are representative of the results obtained from analyses of a minimum of two separate pools of cultured rudiments labeled with the respective glycoprotein precursors.) In all cases, equally complex glycoprotein profiles are characteristic of pancreatic rudiments as early as 14-15 days of gestation. In addition, the fluorographic profiles for [3H]fucose- and [3H]glucosamine-1abeled glycoproteins of the protodifferentiated pancreas are very similar, exhibiting as many as fifteen glycoprotein species. Fewer [3H]N-acetylmannosamine- labeled species are detected in the protodifferentiated pancreas; however, no novel glycoprotein species are detectable using this precursor. The majority of glycoproteins present in the protodif- ferentiated pancreas appear to have molecular weights greater than 80,000 daltons. Most of these glycoproteins are found in the differentiated pancreas at somewhat reduced concentrations relative to the total glycoprotein content of the differentiated pancreas. The number of glycoproteins induced during pancreatic differentiation varies from five to seven species depending on the radioactive sugar used to label the glyc0proteins and the resolution of glycoprotein bands on specific gel autofluorographs. The 43 Figure 3. SDS-polyacrylamide gel electrophoresis of pancreatic rudiments labeled with [3H]N-acety1mannosamine at various stages of differentiation. Pancreatic rudiments were cultured plus or minus S-BrdU and labeled with [3H]N-acetylmannosamine at various times during the secondary transition period of differentiation. Samples of the labeled rudiments were electrophoresed on a 0.1% SDS-8% acrylamide slab gel, and the gel was prepared (without staining) for autofluoro- graphy as described in Methods. Exposure of x-Omat R x-ray film to the prepared gel was for six weeks. Total radioactivity (free plus protein-bound) values for the samples applied to the gel ranged from 3700 to 4000 cpm. Shown are the profiles for cultured rudiments labeled on the in vitro equivalent of A) 14-15 days gestation, B) 17-18 days gestation, C) 17-18 days gestation (cultured in the presence of S-BrdU), D) 19-20 days gestation, and B) 19-20 days gestation (cultured in the presence of S-BrdU). Arrows indicate differentiation-induced glycoproteins. 44 RELATIVE MOBILITY I/ R I Figure 3 45 relative mobilities and corresponding molecular weights of the induced glycoproteins are summarized in Table 3. Bands are designated I through VII in order of increasing mobility. In general, the same induced glycoprotein species can be detected regardless of the radioactive precursor used to label pancreatic rudiments. However, a given glycoprotein may incorporate different precursors in different amounts. For example, Band II is only labeled very slightly with [3Hlfucose, whereas it is intensely labeled when the precursors are [3H]glucosamine and [3H1N-acety1mannosamine. With the exception of Band VII, induced glycoproteins with relative mobilities greater than 0.75 are rela- tively minor constituents of the differentiated pancreas. They appear to correspond to non-induced protein species, however, so their glycosylation may be differentiation dependent. Band I, observed in [3nglucosamine- and [3H]N-acetylmannosamine- labeled pancreatic rudiments, is somewhat unique among the major induced glycoproteins. Band I is apparent in Coomassie blue-stained profiles for [3H]fucose- and [3H]glucosamine-labe1ed rudiments, fluorographs for [3H]fucose-1abeled rudiments, and fluorographs of [3H]leucine-1abeled rudiments. Yet in all of these cases, Band I does not appear to be induced. Furthermore, while this high molecular weight glycoprotein is not detected in the protodifferentiated pancreas labeled with [3H]glucosamine or [3HlN-acetylmannosamine, it is labeled to maximum intensity with these glycoprotein precursors by the middle of the secondary transition period during differentiation. Comparison of relative mobilities of protein species and glyco- protein species that appear to be induced during pancreatic differentiation suggests that at least three induced glycoproteins 46 oom.mN 0m.o . ooa.mm IOOH.®N mm.OIbm.o OOH.mN loom.®N Hm.OIhm.o HH> .o.z oom.vm loom.mm oom.oumb.o oom.vm noom.m~ om.oumh.o H> 49.2 . n.a.z oom.mm Iooo.om om.ouvh.o > oov.oe loov.vv mo.onoo.o oov.vv room.mm ov>.oumo.o oov.vv uooo.mm omh.0imo.o >H oom.oo loom.nm om.0ivm.o oom.oo nooo.mm oo.oaom.o oom.mm nooo.mm oo.onmm.o HHH oom.mh loom.mn mv.OINv.o oom.mh IOOH.mo m¢.OINv.o .oom.mh loom.ho mv.0I©v.o HH ooo.mmalooo.hva va.ouma.o ooo.hvauooo.vva ma.ou¢a.o m.H.z H lawman a (Emma g A lmdmmmm. a unasooaoz o>wumaom umasooaoz o>wumaom Hoasooaoz o>aumamm mcwououmooaao ooaonoq mcwououmoomao mcfiououmoomao noossz nocwsomoccmsamuoomIZHEmH ooaoomqimcwsomoosaonmma ooaonmqiomoosmnmma ocnm .omcwsmxm mammanHOSHm mounu no can >Hco cw oouomuooo «wsmonmouosam Honuo c“ omen omoun o no one nmnumouosam mco ca uoansoo o no uncommmo «oouoouoo uozn «pounced uozm .mucosaosu ooaoomdnocwsomoccmaaauoumnznmm_ «0 once mop ca msmmumouosam 03» can mucoswosn ooaoomalocastoosamnmmH can nomoosmnmmH no women on» ca msmmumouosam mourn scum oocashouoo pawn co>flm o How moaufiawnos o>wumaou ommuo>m mo omens osu ucomoumou mosam> zuaafiooa m>wuoaom scanneucouomuwo oeuoouocmm mo UOAHom cofluflmcmuu whmocooom on» mcwuso mcouocmm ocfiuooxo new on» Ga pounced mcfiououmoomaw .m wanna 47 correspond to induced protein species tentatively identified as secretory proteins. Band III corresponds with the induced protein tentatively identified as amylase. Band IV may correspond to lipase and/or procarboxypeptidase A. (Some fluorographs indicated the presence of two induced glycoproteins in this region of the profile.) Band II appears to correspond to the induced protein species with an approximate molecular weight of 69,000 daltons. As was stated earlier, this is probably the zymogen granule content glycoprotein previously observed by Ronzio and colleagues (submitted for publication). It too is apparently a differentiation-induced secretory glycoprotein. A fourth glycoprotein species, the most rapidly migrating induced glycoprotein, Band VII, may correspond to the induced protein tentatively identified as chymotrypsinogen and/or trypsinogen. However, due to variability in relative mObilities from one gel to the next, it is not possible to determine whether or not the two species co-migrate. In some cases, Band VII appears to migrate just behind the induced protein species (Figure 2). Ronzio and co-workers have observed a glycoprotein species of similar molecular weight (25,000 daltons) in zymogen granule content preparations from adult rat pancreas labeled in vitro with radioactive glucosamine (submitted for publication). Elution of Radioactive Proteins and Glycoproteins from Gels During Stain- ing and Destaining Procedures Figure 2C shows the fluorograph obtained for a non-stained gel prepared identically to the stained gel from which the photograph and fluorograph were obtained in Figures 2A and 28, respectively. 48 The only difference between the two gels was staining with Coomassie blue dissolved in 30% (v/v) methanol-10% (v/v) glacial acetic acid. X-ray film exposure to the gels was for the same period of time at -80°C. A general elution of radioactivity from the gel appears to (occur during staining and destaining. No selective elution of glycoproteins could be detected, though the process does seem to reduce the fuzziness (due to microheterogeneity?) associated with glycoprotein bands of relative mobility 0.2-0.4 and to have effectively eluted some of the high molecular weight [3H]leucine- labeled proteins. In addition, elution of the broad fuzzy bands migrating in front of the tracking dye in Figure 3 occurred when an identical gel was stained. Possibly these bands contain glycolipids which would be highly susceptible to solubilization in the gel staining and destaining solutions. Incorporation of [3H]Leucine and [3H]Fucose into Zymogen Granule Content and Zymogen Granule Membrane Constituents of Adult Rat Pancreas For the purpose of facilitating the identification of proteins and glycoproteins found in the embryonic pancreas, adult rat pancreatic tissue was labeled in vitro with [3H]leucine and [3H]fucose. Zymogen granule membranes and zymogen granule content proteins were subsequently obtained from the labeled tissue and samples were electrophoresed on 0.1% SDS-polyacrylamide slab gels. Fluorographs obtained from the 2,5-diphenyloxazole-impregnated dried gels are shown in Figure 4. Table 4 summarizes the relative mObilities and molecular weights of the labeled species. The-most intensely [3H]leucine- and [3H]fucose-labeled component of zymogen granule membranes is a protein with a molecular weight 49 Figure 4. SDS-polyacrylamide gel electrophoresis of adult rat pancreas zymogen granule membrane proteins and zymogen granule content proteins labeled in vitro with [3H]leucine and [3Hlfucose. Adult rat pancreas was labeled in vitro with [3H]leucine or [3H]fucose, and zymogen granule membrane and zymogen granule content subfractions of the tissue were obtained as described in Methods. Samples were electrophoresed on 0.1% SDS-8% acrylamide slab gels. Gels were prepared for autofluorography without first being stained. X-Omat R X-ray films were exposed to the dried gels for eight weeks. Fluorograph profiles are shown for A) [3Hlleucine-labeled zymogen granule membrane (4,450 dpm), B) [3H]1eucine-labeled zymogen granule content proteins (8,500 dpm), C) [3H]fucose-1abeled zymogen granule membrane (7,100 dpm), and D) [3H]fucose-1abe1ed zymogen granule content proteins (3,400 dpm). Arrows indicate labeled species present in the profiles. _ 50 >9 ‘1' ->-«-— Figure 4 51 oom.mm mm.o oom.~m. mm.o .o.z oom.mm mm.o .o.z oom.mm mm.o. ooo.mm mn.o ooo.Vm om.o oov.vv mo.o ooe.me mo.o oom.Hm H6.o ooo.mm 06.0 oom.oo vm.o .o.z almmmmmmu .Nmmmmmmm (Inmmmmmfl. .Nmmmmmmm Hmasooaoz 0>wuoaom unasooaoz o>wumaom ooaonoqiomooshamma ooaoomqlocwosoqnmma oom.ov oomcoo oom.mm ooo.oea pauses Hmasooaoz Aminosvm.o Amuaovam.o Asimovofl.o .Nmmmmmmm m>wumaom UQHOQMAIOWOUDE 2mm“- 00v.m¢ oom.mv oom.vm oom.hm oom.®m unmade Hmasooaoz 00.0 no.0 mm.o “mumwvom.o ANImUvmm.o .Q.z Nuwaflnoz o>wumHom ooHoQMQIocflosmAHmma mucoucoo oascmuu cmmosmn meanness: oascmuo comosaN mucoucoo oascmum comoshn 0cm mocmunfioa mascoum somosxn mowuocmm uon vasom mo mucocomsoo oonaniomoosmHmmH one rooflosoqmmmg .oouoouoo bozo ca :3osm maom beam opwsoamuom>aom mo mommumouosam scum oocwmuno ouo3 mofiuwawnofi o>wum~ou cwououm .e ousmwh . v manna. 52 of 86,000-88,000 daltons (relative mobility 0.37-0.38). This protein has been designated GP-2 by Ronzio and co-workers (Ronzio, 1973b; MacDonald and Ronzio, 1972; Lewis et al., 1977). It is known to be highly glycosylated and has been shown to have an apparent molecular weight of 70,000-83,000 daltons in 1% SDS-9% acrylamide gels. Two other major fucoproteins are found in the zymogen granule membrane. They correspond with GP-l (120,000-130,000 daltons) and GP-3 (52,000 daltons) observed by Ronzio and colleagues. Here GP-l and GP-3 have apparent molecular weights of approximately 140,000 and 58,000-60,000 daltons, respectively. GP-l and GP-3 did not label well with [3H]leucine which is in agreement with previous findings of Ronzio and co-workers (Kronquist et al., 1977). It will be noted that in the gel system employed in these experiments, GP-l, GP-2, and GP-3 all have molecular weights estimated to be somewhat greater than the molecular weight values obtained by Ronzio and co-workers. Several factors may be operating to produce this apparent inconsistency. First, other investigators have found that proteins of high and low molecular weights migrate non-linearly in the Laemmli gel system (Wray and Perdue, 1974). Secondly, glyco- proteins, especially those containing 10% or more carbohydrate, bind less SDS on a weight basis than do nonglycosylated proteins. Glyco- proteins, therefore, exhibit a decreased charge to mass ratio which results in decreased mobility on $08 gels (Segrest and Jackson, 1972). Finally, it will be noted that the molecular weights of these zymogen granule membrane glycoproteins were previously determined in 1% SDS gels of both 9% and 12% acrylamide. The gels utilized in these experiments were of 8% acrylamide and contained only 0.1% SDS. Although gel samples contained 2% SDS, it is not unlikely that some 53 SDS bound to glycoproteins was lost as the glycoproteins migrated into the 0.1% SDS-polyacrylamide gels. Furthermore, it may be assumed that the anomalous behavior observed here for zymogen granule membrane glycoproteins extends also to soluble glycoproteins. Tentative identification of most of the [3H]leucine-labe1ed zymogen granule content proteins can be made on the basis of estimated molecular weights. The proteins in order of decreasing molecular weight are probably amylase, lipase, procarboxypeptidase A (or carboxypeptidases A and B in the event of activation of the zymogen granule content solution), unknown, chymotrypsinogen and/or trypsino- gen (or chymotrypsin and/or trypsin), and ribonuclease. Comparison of the [3H]1eucine-labeled proteins from adult rat pancreas zymogen granule content with the differentiation-induced pancreatic proteins listed in Table 2 indicates some differences in the two sets of proteins. Procarboxypeptidase B, for example, is either absent or co-migrates with amylase in the [3Hlleucine-labeled zymogen granule content protein fluorographic profile. The [3H]leucine-labeled protein of molecular weight 28,500 (relative mobility 0.88) of the zymogen granule content does not appear to be an induced protein in differentiating pancreas. One major disparity is the position of procarboxypeptidase A in the profiles. The protein band with relative mobility 0.80 in the [3H11eucine-1abeled zymogen granule content corresponds in calculated molecular weight most nearly with the known molecular weight of procarboxypeptidase A. However, the relative abundance of procarboxypeptidase A in embryonic pancreas observed by Walther and co-workers (1974) strongly suggests that the protein with relative mobility 0.72-0.73 in differentiating rat pancreas is procarboxypeptidase A. 54 The possibility of proenzyme activation in the [3H11eucine- labeled zymogen granule content proteins preparation cannot be eliminated and would provide one explanation for the variations in the mobilities of secretory proteins; that is, proenzymes of the differentiating pancreas may be present in their activated forms (with lower molecular weights) in the preparation of secretory proteins from adult pancreas. Had activation occurred, however, the presence of a large amount of low molecular weight peptides should be evident in the gel profile. It is not. A number of zymogen granule content proteins appear to label. with [3H]fucose. Only two of these proteins are significantly labeled. The approximate molecular weights of these zymogen granule content fucoproteins are 36,000 and 23,000 daltons. Rat pancreas procarboxypeptidase A, carboxypeptidases A and B, and lipase (delipidated) all have molecular weights of approximately 36,000. Procarboxypeptidase A was suspected to migrate with a somewhat slower mobility (0.72-0.73) in differentiating rat pancreas, however. The 23,000 dalton fucoprotein may be ribonuclease. Bovine and porcine pancreatic ribonucleases are known to be glycosylated (Plummer and Hirs, 1963; Reinhold et al., 1968). This low molecular weight fucoprotein migrates very near the bromphenol blue tracking dye on the gel and may, therefore, consist of a number of unresolved glycoproteins of low molecular weight. Faintly labeled fucoproteins with relative mobilities 0.61 and 0.68 may correspond with amylase and lipase, respectively. When comparing the fucosylated zymogen granule membrane glyco- proteins and zymogen granule content glycoproteins of the adult rat pancreas with fucoproteins of embryonic rat pancreas, several 55 observations may be made. First, a heterogeneous fucoprotein band of molecular weight similar to that of GP-2 appears to be present in the pancreas from as early as 14-15 days of gestation. It does not appear to be one of the glycoproteins which undergoes induction during the secondary transition period of pancreatic differentiation. Glycoprotein Band I of embryonic pancreas appears to correspond to zymogen granule membrane GP-l. Band I, though not induced in terms of fucosylation, seems to be induced in rudiments labeled with [3H]glucosamine and [3H]N-acetylmannosamine. Uncertainty in the identification of the zymogen granule content proteins confounds rather than simplifies the identification of differentiation-induced fucoproteins. It is probable that the intensely labeled zymogen granule content fucoprotein with relative mobility 0.78 corresponds with one of the differentiation-induced fucoproteins designated Band IV, Band V, or Band VI. Fractionation of [3H]Fucose-Labeled Differentiated Pancreatic Rudiments In order to determine whether or not the differentiation-induced glyc0proteins of the pancreas are primarily soluble or membrane- bound glycoproteins, [3H]fucose-labeled 19- to 20-day pancreatic rudiments (cultured from day 16 of gestation instead of day 14) were fractionated into "soluble" and "particulate" components. Fractionation was carried out essentially as described in Methods. Rudiments (a total of sixteen) were sonicated in 0.3 M sucrose containing 0.25 mg/ml soybean trypsin inhibitor and 12 uM phenyl- methylsulfonylfluoride, and the sonicates were centrifuged at 111,000 x g for two hours. The supernate was designated the "soluble“ fraction. The pellet was resuspended with sonication 56 in 0.25 M NaBr and centrifugation was repeated. The supernate from this centrifugation was designated the "first NaBr wash." A second NaBr wash was carried out, and the final pellet was designated the "particulate" fraction. Table 5 shows the protein content and specific radioactivity of each of the fractions obtained. Table 5. Fractionation of [3H1fucose-labe1ed differentiated (in vitro) pancreatic rudiments Total Protein Specific Radioactivity % of Recovered Fraction (ug) (dpm/pg x 10'5) Radioactivity Homogenate 894 2.42 "Soluble" 513 1.19 58.9 fraction First NaBr 116 0.19 9.4 wash Second NaBr 28 0.06 3.0 wash "Particulate" 180 0.58 28.7 fraction Eighty-three percent of the initial radioactivity was recovered during the fractionation. The results suggest that a single NaBr wash is sufficient to remove a large portion of the adsorbed soluble proteins and trapped non-protein bound radioactivity from the particulate fraction. Acid precipitation of the homogenate indicated that 49% of the [3H]fucose incorporated by the rudiments was protein-bound after 18 hours of incubation with this precursor. 57 Figure 5A shows the Coomassie blue staining profiles for the soluble and particulate fractions obtained from the.[3H]fucose- labeled rudiments. Figure 5B shows the corresponding fluorograph from the stained gel. The salient feature of these electrOphoretic profiles is that virtually all of the fucoproteins shown earlier to be induced during differentiation (Figure ZB) are found in the soluble fraction and the first NaBr wash of the particulate fraction. The one glycoprotein with relative mobility 0.12-0.15 (Band I) which appears to be differentiation-induced in [3H]glucosamine- and [3H]N-acetylmannosamine-labeled rudiments (though not in [3H]fucose- labeled rudiments) is primarily a particulate glycoprotein. The soluble fuc0proteins tend to have molecular weights less than 74,000 daltons. One notable exception is the fucoprotein with molecular weight 157,000-160,000 daltons (relative mobility 0.10-0.11). The bulk of the fucoproteins with mobilities of 0.2 to 0.5 are particulate (membrane-bound) glycoproteins. In an effort to identify the fucoproteins of the soluble and particulate fractions from the (in Vitro) differentiated pancreatic rudiments as zymogen granule membrane or secretory glycoproteins, the fluorographic profiles for these fractions were compared with the profiles for [3Hlfucose-labeled zymogen granule membranes and zymogen granule content proteins from adult rat pancreas. This comparison is presented in Figure 6. The zymogen granule content fucoprotein with relative mobility 0.78 probably corresponds with a soluble (induced) embryonic fucoprotein with relative mobility 0.66-0.75 (Band IV, Table 3) which is suggested to be procarboxy- peptidase A and/or lipase. It may, however, correspond to the induced embryonic fucoprotein, designated Band V, the identity of 58 .Iaanaunnaoonaneaom Ego ooo.mo oeuefiwume “Heuou Emo ooo.mh acfleuoum o: «my smez Hmez ocooem e ueume cowuoeum eueasowuumm Am oce .Aaeuou Emo ooo.an “cweuoum m: vvv coauomuw momagoauuma man no name ummz amuse In .Imanmuaaaomnaucaoa sac ooo.¢m cmumsaumm “Heuou Emo ooo.oma “cfleuoum on omv :oflueeum eandHOm AU .Aeaoeuflmfloeumlofiee Emo ooo.om «Heuou Emo ooo.mma “caeuonm m: wov euecemofion Am esu Mom ozone eue meafimoum oanmeumouosfim one meHHmonm mowcaeum esan eflwmesoou ens .e>o mcflxoenu esaa Hocenmsoun ere suHS mcwueumfie Aoo>.mav m emmeaoscooflu oaumeuocem ecw>on Am oco .Aooo.mvv cwfisoae>o Av .Aooo.oov cassaam Essen ecw>on Am .Aooo.vhv cfluuemmcenu cease Am .Aooo.moav caHsnonoumnu ecaouom AH ewe d ecea ca mcwueemme moueocmum cweuoum .Hem oeCfleum efiem esp Eoum oecfleuno ammumOHOSHm ecu ma m eHsmHm eafl£3 .mGOHueeHm HmHsHHeOQSm enema How meaamoum mcwcwmum esao eflmmmsooo esp mSOSm 4 eusmwm .mmeo mm How maem oeuemenm ecu Op oemomxe eHe3 madam >mulx ueEOIx Hewom mm .anmeumouooam louse How oeuemeum H0\oce oecfleum eue3 maew .maem seam eoflseamuoe wmlmam wa.o co oenaaece eue3 mcofluoeum ena .emoosmfimmg spas mason ma oeaeaea mean one ma moo mo uceae>wsoe esp on coaueumem «0 ma moo scum oeusuaso museEHosn oaumeuucem Eoum oecweeno eues mcofiuoeum eueaoowuuem oce eaosaom .emoosmammH news cowueumem maeo omima mo uceae>fisoe emu co oeaenea muceEwosu oaueeuecem oenouaso Bonn mGOHuoeum ueHoHHeoosm mo mwmeuoomouuoeae Hem eoflseamuommaomimam .m euomflm A1|1ISOIN EAIIV'IEU AlI1ISOIN EAIIV'IBU Figure 5 6O .lmnamasanomna-aaom sac oom.ma oeueEfiume «Heuou Emo ooo.©aav muneEHosH oeHeQeHIemOOsmmmmH EOHM Acmez nmez onooem e neumev nofluoeum eueanoflunem Am one .Asmo ooa.hv meeuenem pen panoe Bonn mnfleuonm enennEeE eanneum nemoexu oeaeoealemoonmfimma AD .Aemo oov.mv meeuonem pen panoe scum mnfleuoum uneunoo eannenm nemoskn oeaenea -mmooze_mm. no .Imnnauaanomnauenom Ede ooo.vm emumsauma tampon sac ooo.omnv muneannu oeaeoeaiemoonmmmmg scum nofluoeum eaonHom Am .Aeaneuflmwoeum leave nae ooo.om tampon emu ooo.mmnv mesmeHesn cmnmnmaummousmnmmL some euenemoson An enu eue n3onm .Hem oenfieumnn ne Eonm oenfleuoo eue3 munesflonu owueenonem Eoum mnofiuoenmonm oeaenea Mom meaflmonm oanmeumouonam .m eunmflm one o eunmflm nH oenflnomeo me oemenonmouuoeae one oenemenm enes meamsmm .meenonem pen panoe Eonm muneunoe eanneum nemofihu en» one menenoEeE easnenm nemoshe mo mnfleuoumoonw enu nuflz muneEHosH eflueeuonem oennuano Eoum mnfleuonmoonm mo nOmflHemEoo ewueuonmonuoeam .o ehsmflm 61 o eunmfim II I AlI'IISOW 3A|1V138 62 which is not known. The relative mobility of GP-l from zymogen granule membrane correlates well with the relative mobility of a fucosylated protein (Band I) in the particulate fraction from differentiated pancreatic rudiments. GP-2 migrates similarly to a broad heterogeneous band of particulate fucoproteins from differen- tiated rudiments. Incorporation of [3H]Fucose and [3H]G1ucosamine into Acid- Precipitable Protein During Pancreatic Differentiation It was not possible, because of the quantity of embryonic tissue needed, to determine the pool sizes and kinetics of incorpora- tion for the radioactive sugars used to label glycoproteins synthesized during in vitro pancreatic differentiation. The second best alternative . was to measure the amount of radioactive glycoprotein precursor incorporated into 10% trichloroacetic acid-precipitable protein at various times during the secondary transition period of pancreatic differentiation. From the acid precipitation data, it is possible to estimate the sizes of the free sugar pools (acid soluble radio- activity). The rates of glycoprotein synthesis may also be estimated if the precursor pool size is assumed not to change during the labeling period and no conversion of the precursor into other potential protein or glycoprotein precursors occurs. Table 6 summarizes the results of 10% trichloroacetic acid precipitation of samples of [3H]fucose-labe1ed and [3H]glucosamine-labeled rudiments. Due to the low level of radioactivity incorporated during labeling of pancreatic rudiments with [3HlN-acetylmannosamine (one-tenth that incorporated during [3H]fucose labeling and one-twentieth that incorporated during [3H]glucosamine labeling, on the basis of total radioactivity per Hg protein) rudiments labeled with 63 Anoumum wo eonemeum 0.0m .Q.z oamm .Q.z eueHnOHuHem enu nH oeusuanev o.mm .o.z mNhH e.a.z eHnnHom oman v.m H N.mn N.H H H.mm mmv H moma omm H omvH eueanoHHHem v.m H m.ov m.m H m.om mom H mmoa NNH H Ohm eHonHom omlmH n.H H m.Hm m.m H o.mm mom H vmma mo H nova eueHnoHHHem o.m H o.mm m.m H m.om mam H omea «AH H HNHH eannaom manna >.o H o.vm m.m H 0.5m Nmm H omna NHN H omma eueHnoHuHem N.H H m.wm o.v H >.vv va H mama oom H coma eannHOm mHIvH enHEemoonHmHmmH emoonmnmma enHEemoonHmHmmH emoonmflmma oHon OHHeoeouoHnoHHB Amunon mH\nHeuonm o:\EmoV nOHuoeHm Amxeov oeaeoeq woa he oeueuHmHoeHm suu>auomoaomm Hmuos no » suH>HuomoHemm manmuHmHomnmuenoa nenz muneEHonm mo eon OHnOMHnEm uneHe>Hnom .oeanHeueo ooze .noHueH>eo neeE H eHe menHe> .noHueumem m>eo omima mo uneHe>ane en» no oeHeneH muneEHonH oeHnano nqu use oeHHHeo eHe3 muneEHHemxe Hnom .nOHueumem m>eo wanna Ho mHIvH Ho muneae>Hnoe enu no oeHeneH munefiHonn oeHnano nqu oeuononoo eHe3 muneEHHemxe 039 HeHHHeo mo eonemeum enu nH mnoHuoeHw enu mo meamfiem .oHoe eHueoeonoHnoHHu woa nuHs oeueuHmHoeHm eues nHeuoum .moonuez nH erHHumeo me oeuemeum euez apnea IHonH UHHeeHonem oeHeoeHIenHEemoonHmmmmH one nemoonmfimma scum mnOHuoeHm eueHnoHuHem one eHnnHom nOHueHuneHeMMHo Mo oOHHem nOHuHmneHH aueonooem en» mnHHno muneEHonH oHueeHonem an nHeuoum eHneuHmHoeHmioHoe ounH muemnm e>HHoe0HoeH mo nOHueHomHoonH .o eHnee 64 [3H]N-acetylmannosamine were not examined in this and subsequent experiments. The results in Table 6 suggest, first of all, that over a constant labeling period of 18 hours, the rate of particulate glycoprotein synthesis relative to total particulate protein synthesis remains essentially unchanged during the secondary transition period. This seems to be in keeping with the observa- tion that possibly only one particulate glycoprotein is induced during the secondary transition period of differentiation. Total particulate protein per cultured embryonic pancreatic rudiment appears to increase by no more than a factor of two over this period. Soluble glycoprotein synthesis declines relative to total soluble protein synthesis during the secondary transition period. Soluble fucoprotein synthesis in the differentiated pancreas is roughly one-third that of the protodifferentiated pancreas. At the same time, the amount of soluble "glucosamine"-glycoprotein synthesis in the differentiated pancreas is sixty percent that of the proto- differentiated pancreas. It appears that synthesis of the differentiation-induced soluble glycoproteins is not commensurate with total soluble protein synthesis, which increases six- to tenfold (on a per rudiment basis) during the secondary transition period. The ratio of protein-bound radioactive sugar to total radio- active sugar for the soluble and particulate fractions remains relatively constant during in vitro differentiation of the pancreas. Approximately 50% of the labeled fucose found in the soluble fraction is protein-bound, while 85-95% of the labeled fucose in 65 the particulate fraction is acid-precipitable. The amount of protein-bound radioactivity in the soluble fraction from [3H]glucosamine-labeled rudiments remains at about 30-40% throughout the secondary transition, while the particulate fraction contains 80-85% acid-precipitable radioactivity from [3H]glucosamine labeling. (It must be noted that the acid precipitates of the particulate fractions were never extracted with organic solvents to remove glycolipid. A portion of the acid-precipitable radioactivity of the particulate fractions may, therefore, be glycolipid-bound rather than protein-bound.) Estimates of the free sugar pool sizes were made on the basis of the amount of acid-soluble radioactivity found in the soluble fractions from rudiment sonicates. The amount of free [3H]fucose per pancreatic rudiment appears to increase by no more than a factor of two during the secondary transition. The apparent size of the free radioactive sugar pool from [3H]glucosamine labeling increases by a factor of four during the secondary transition. These changes in free sugar pools could reflect an increase in cell volume with cytodifferentiation and/or an increased rate of sugar transport. Whether or not these small changes in the free radio- active sugar pools are significant with respect to glyc0protein synthesis of the differentiating pancreas is not known. 'Radioimmunoassay_for Zymogen Granule Membrane Glycoproteins and Secretory Glycoproteins in Differentiating Rat Pancreas The most apparent structural change that occurs in the exocrine pancreas during cytodifferentiation is the accumulation of zymogen granules. This occurs in conjunction with the accumulation of the 66 cell specific secretory proteins. Of considerable interest is whether the zymogen granule membrane glycoprotein constituents ’undergo differentiation-dependent induction in synchrony with the induction of secretory proteins. The best way to answer this question seemed to be to make use of an indirect radioimmunoassay technique employing rabbit antibodies developed in the laboratory against zymogen granule membranes from adult rat pancreas. Similarly, the use of an anti zymogen granule content (antiserum) was expected to provide more certain evidence for the glycosylation of pancreatic secretory proteins and the accumulation of secretory glycoproteins in vitro during the secondary transition period of pancreatic differentiation. A description of the antibodies used in the radioimmunoassays follows. Subsequently, the results of assays for zymogen granule membrane glycoproteins and secretory glycoproteins in cultured pancreases at various stages of the secondary transition are reported. Characterization of the Anti Zymogen Granule Content (Antiserum) Figure 7 shows the results of both direct and indirect titration of the anti zymogen granule content (antiserum) with a preparation of [3H]1eucine-labeled zymogen granule content proteins from adult rat pancreas. In the direct titration, 50 ul of antiserum were incubated with 50 ul of zymogen granule content proteins (concentrated and dialyzed against borate-saline buffer, pH 8.4-8.5) at various dilutions in borate-saline buffer for one hour at 25°C. The reactions were then stored at 0-4°C for 24 or 72 hours. (Precipitates were collected by centrifugation and washed three times by resuspension in borate-saline buffer followed by centrifugation. The washed 67 .eHnneHm nemosun u UN .TIIV meueuHmHoeHm enu HON oenHEHeueo mes anHemHune ere xn oeueuHmHoeHm aquHponHoeH oeooe Heuou mo uneonem enu one oeuoeHHoo eHez meueHHmHoeHm .muneEHonH eHueeHonem Scum mnHeuonmoowam eHndHom mo >emmeonnEEH0HoeH ene How moonuez nH erHHomeo enoHnnoeu noHueuHmHoeHm moonHHne eHnnoo esp mnHmn uno oeHHHeo Omae mes mnHeuonm uneunoo eHnneHm nemosmn oeaeneaienHoneHHmmH mo mnOHunHHo mSOHHe> mo H: OH nHHz AH: OHV EnHemHune enu mo noHHeHHHu HoeHHoCH .UoVIO UM A‘lv munon Nb one AIOIV muses on Heume oenmes one oeuoeHHoo meueuHmHoeHm How oenHSHeueo mes EnHemHune enu an oeueuHmHoeHm >HH>HHoe0HoeH oeooe Heuou mo uneonem era .EnuemHune enu mo noHHeHuHu HoeHHo How uxeu enu nH oenHHomeo we Am:\8mo momv mnHeuonm uneunoo eHsneHm nemoswu mo nOHueHemeHm oeHeQeH IenHeneHHmmH e mo mnOHunHHo mSOHHe> Ho an om nuHs oeuoeen eHe3 AESHemHunev uneunoe eanneum nemosxn Hone oeunefieamfiooeo mo mHeuHHouoHE >HMHm .meeuonem Hen Hanoe scum mnHeuoum uneunoo eHnneHm nemoshu oeaenealenHoneHHmmH nuHa AEnHemHunev uneunoo eHnneHm nemos>n Hune mo nOHueHHHB .h euanm 0\ CO I UUVIUU I T t TTUIIITI UUUT J l [Lilli l l 1 llllll l O 0 Eumasgiurfliuaiuog 92 giuv Kg PaIDI!d!33Jd Rumooowoa 10:01 °/. 0 (D ID .50 .IO .05 .Ol ,ug llil 26 Content Proteins Figure 7 69 precipitates were solubilized and samples taken for liquid scintil- lation counting. As seen in Figure 7, the equivalence point for the antiserum occurred when zymogen granule content proteins were added to the assay mixture at a concentration of 0.126 ug/ul. The total amount of protein precipitated by 50 ul of the antiserum at the equivalence point in the direct titration was 0.46 ug and 0.55 ug for the 24—hour and 72-hour 0-4°C incubation periods, respectively. A titration curve was also obtained for the anti zymogen granule content (antiserum) with [3H]leucine-labeled zymogen granule content proteins in a double antibody (indirect) precipitation reaction using a commercial preparation of goat antirabbit (gamma globulin) to precipitate the first antigen-antibody complex. The assay system was the same as that described in Methods for radio- immunoassay of zymogen granule content glycoproteins in the soluble fractions from labeled pancreatic rudiments. A titer of 1.5 ug zymogen granule content proteins per 10 ul antiserum was obtained in the double antibody assay. Use of the second antibody appears to increase the titer efficiency of the antiserum. To further characterize the anti zymogen granule content (antiserum), the cross-reactivity of the antiserum with various subcellular fractions prepared from adult rat pancreas was examined using double immunodiffusion in 1% agarose gels (Ouchterlony, 1958). The antiserum was tested for cross-reactivity with NaBr-washed zymogen granule membranes, smooth microsomal membranes, and NaBr- washed rough microsomal membranes. In Figure 8A, it can be seen that the anti zymogen granule content (antiserum) (wells 1 and 4), in addition to forming precipitin lines with zymogen granule content proteins (wells 5 and 6), also reacted with NaBr—washed 70 Figure 8A. Ouchterlony double diffusion test of anti zymogen granule content (antiserum) and anti zymogen granule membrane (IgG) versus zymogen granule content proteins and zymogen granule membranes prepared from adult rat pancreas. Immunodiffusion was carried out in a 90-mm glass dish filled with 10 ml 1% agarose dissolved in 10 mM NaZHPOu buffer (pH 7.4) containing 1% (v/v) Triton X-100, 1% sodium deoxycholate, 5.0 mM disodium EDTA, 12 uM phenylmethylsulfonylfluoride, 0.001% butylated hydroxytoluene, and 0.01% Trypan blue. The center well contained 20 ul rabbit anti zymogen granule membrane (IgG) (4.9 ug/ul). Wells 1 and 4 contained 20 ul rabbit anti zymogen granule content (antiserum). Wells 2 and 3 each contained 20 111 of solubilized zymogen granule membrane (0.43 ug/ul and 0.21 ug/ul, respectively). wells 5 and 6 each contained 20 ul zymogen granule content proteins in borate-saline buffer (pH 8.4) (0.6 ug/ul and 0.3 ug/ul, respectively). This plate was photographed after 65 hours at room temperature in a moist chamber. Figure 8B. Microheterogeneity of anti zymogen granule content (antiserum). Double diffusion test of anti zymogen granule content (antiserum) versus various dilutions of a zymogen granule content protein solution. Immunodiffusion was carried out in a 100-mm plastic dish filled with 8 m1 of 1% agarose dissolved in borate-saline buffer, pH 8.4-8.5. The center well contained 20 ul of undiluted anti zymogen granule content (antiserum). Wells 1-6 contained 20 ul of various dilutions of zymogen granule content protein solution containing 2.02 ug/ul. Well 1, 1 to 2 dilution; Well 2, 1 to 4; Well 3, 1 to 8; Well 4, l to 16; Well 5, 1 to 32; Well 6, undiluted. The plate was photographed after 45 hours at room temperature in a moist chamber. 71 Figure 8 72 zymogen granule membrane proteins. Several explanations could account for the cross-reactivity. It is possible that the zymogen granule membrane preparation used in the experiment, even after NaBr washing, may still have contained adsorbed secretory proteins (approximately 6%; Ronzio, submitted for publication). Of more interest is the possibility that similar antigenic determinants exist for one or more of the zymogen granule content proteins and zymogen granule membrane proteins. The microheterogeneity of the anti zymogen content (antiserum) can be seen in Figure 8B. This 1% agarose gel was prepared in borate-saline buffer (pH 8.4-8.5) rather than the phosphate-detergent buffer (pH 7.4) used in Figure 8A. Better resolution of the multiple precipitin lines occurs in this gel. At least five separate precipitin bands can be observed with close examination of the pattern. Such microheterogeneity of the antiserum is expected in view of the large number of secretory proteins present in the zymogen granule content preparations used to induce the formation of the antibodies. Figures 9A and 9B show the cross-reactivity of the anti zymogen granule content (antiserum) with smooth microsomal membranes and NaBr-washed rough microsomal membranes, respectively. The smooth microsomal membrane preparation had not been washed with NaBr and was, therefore, expected to exhibit reactivity with the antiserum due to adsorbed secretory proteins. In addition to reactions of identity, however, there appear to be at least two reactions of partial identity (spurs pointing toward well 1) which suggest the presence of antibodies in the antiserum which react only with a component of smooth microsomal membranes or a non-secretory protein contaminant of the membrane preparation. 73 Figure 9. Cross-reactivity of anti zymogen granule content (antiserum) with smooth microsomal membranes and NaBr-washed rough microsomal membranes from adult rat pancreas. A. Immunodiffusion was carried out as described in Figure 8A. The center well contained 20 ul of anti zymogen granule content (antiserum). Wells 1 and 4 contained 20 ul of 0.6 ug/ul and 0.3 ug/ul zymogen granule content proteins, respectively. Wells 2, 3, 5, and 6 contained 20 ul of solubilized smooth microsomal membranes. Well 2, 12 ug/ul; Well 3, 3 ug/ul; Well 5, 0.8 ug/ul; Well 6, 0.2 ug/ul. This preparation of smooth microsomal membranes had not been washed with 0.25 M NaBr and, therefore, probably contained a minimum of 10% (by mass) adsorbed soluble proteins including zymogen granule content proteins. B. Immunodiffusion was carried out as described in Figure 8A. The center well contained 20 ul of anti zymogen granule content (antiserum). Wells 1 and 4 contained 20 ul of 0.6 ug/ul and 0.3 ug/ul zymogen granule content proteins, respectively. Wells 2, 3, 5, and 6 contained 20 pl of solubilized NaBr-washed rough microsomal membranes. Well 2, 18 ug/ul; Well 3, 4.5 ug/ul; Well 5, 1.1 ug/ul; Well 6, 0.3 ug/ul. 75 Cross-reactivity of the anti zymogen granule content (antiserum) with NaBr-washed rough microsomal membranes (Figure 9B) was similar to that seen for smooth microsomal membranes. In addition to reac- tions of identity and partial identity, at least one precipitin line is present that indicates there are antibodies in the antiserum that react only with a component of rough microsomal membranes or a possible contaminant of the membranes (reaction of nonidentity between center well and well 2 of plate). Purification of the Anti Zymogen Granule Membrane (Gamma Globulin) Ouchterlony double diffusion plates showed that the anti zymogen granule membrane (gamma globulin) preparation cross reacted with zymogen granule content proteins (not shown). Passage of the gamma globulin through a column of Sepharose 4B to which zymogen granule content proteins had been bound was the method used to remove this cross-reactivity. The procedure used to couple zymogen granule content proteins to cyanogen bromide-activated Sepharose 4B was taken from the Pharmacia Fine Chemicals publication, CNBr-activated Sepharose 4B Immobilization of Biopolymers, Con A-Sepharose fer Affinity Chroma- tography of.Polysaccharides and Glycoproteins. A lyophilized preparation of adult rat pancreas zymogen granule content proteins was dissolved in and dialyzed against high ionic strength borate- saline buffer (pH 8.4-8.5) containing 5.0 mM boric acid, 1.25 mM sodium borate, and 0.46 M NaCl (calculated ionic strength was approximately 0.5). Ten milliliters of this solution of zymogen granule content proteins (55 mg protein) was added to 2 g of CNBr- activated Sepharose 4B which had been swollen in 10.3 N HCl and 76 washed with high ionic strength borate-saline buffer. After two hours at room temperature, the absorbance at 280 nm of the solution above the beads had reached a relatively constant value. The gel suspension was then washed on a scintered glass filter with high ionic strength borate-saline buffer. This was followed by washing and soaking in 1 M ethanolamine-HCl (pH 8.0) for two hours to block any unreacted binding sites on the gel. The final step in the preparation of the affinity gel consisted of alternately washing the gel with 0.1 M acetate buffer (pH 4) and 0.1 M borate buffer (pH 8.5), each containing 1 M NaCl. The gel with coupled zymogen granule content proteins was then washed with standard borate-saline buffer (pH 8.4-8.5) containing 5.0 mM boric acid, 1.25 mM sodium borate, and 0.142 M NaCl. Approximately 2.6 m1 of anti zymogen granule membrane (gamma globulin) were applied to a column packed with the affinity gel (6.5 ml bed volume in a 10-ml glass pipet). Elution of unbound gamma globulin occurred with a single bed volume of standard borate- saline buffer (pH 8.4—8.5) (Figure 10). Approximately 58% of the applied gamma globulin was recovered in the borate-saline eluted peak. Elution of the column with 0.2 M glycine (pH 2.8) in 0.5 M NaCl to remove the bound cross-reacting gamma globulin component did not yield a sufficient quantity of proteins, even after concen- tration, to react with zymogen granule content proteins on double diffusion plates. The borate-saline eluted anti zymogen granule membrane (gamma globulin) was found to not react with zymogen granule content proteins on Ouchterlony double diffusion plates. In Figure 8A, the ratio of gamma globulin to zymogen granule content proteins, on a 77 Figure 10. Purification of anti zymogen granule membrane (IgG). Removal of gamma globulin fraction cross-reactive with zymogen granule content proteins. Anti zymogen granule membrane (196) was prepared as described in Methods from the rabbit antiserum. This gamma globulin preparation (2.6 m1; approximately 7.4 mg/ml) was then chromatographed on a 6.5 ml column of Sepharose 4B to which zymogen granule content proteins (7.8 mg/ml gel, maximally) had been covalently coupled (see text). Fractions were pooled as indicated. The borate-saline eluted IgG fraction was concentrated and used in all radioimmunoassays for zymogen granule membrane proteins. A280 78 3.0" |.O" i—“FW Boroto - Saline . MGIc'o 02_ ymfi) 20 X in 0.53fiNoCl 4O 60 Elution Volume (ml) Figure 10 79 protein basis, was equivalent to that previously used to test for cross-reactivity of the untreated gamma globulin with zymogen granule content proteins. In the latter case, a positive reaction was observed. Characterization of the Anti Zymogen Granule Membrane (Gamma Globulin) Attempts were made to determine the titer of the treated anti zymogen granule membrane (IgG) for zymogen granule membrane proteins using the double antibody precipitation system described in Methods for assay of particulate proteins from labeled pancreatic rudiments. Using this procedure, 10 ul of the IgG were found to precipitate as much as 8.6 ug of [3H]1eucine-labeled adult rat zymogen granule membrane protein. Two preparations of [3H]leucine-labeled zymogen granule membrane were used to titrate the anti zymogen granule membrane (IgG). An equivalence point was never obtained in the titration curves suggesting that the value stated above was still in the antibody excess region of the titration. Between 20 and 45% of the total [3Hlleucine-labeled protein present in the zymogen granule membrane preparations was precipitated in the double antibody precipitation system. Cross-reactivity of the treated anti zymogen granule membrane (IgG) with smooth microsomal membranes and with NaBr-washed rough microsomal membranes from adult rat pancreas was tested in Ouchterlony immunodiffusion gels. Figures 11A and 118 indicate cross-reactivity of the treated anti zymogen granule membrane (IgG) with both of these membrane subfractions. The precipitin bands suggest the presence of at least two reactions of identity for the IgG with zymogen granule membranes and smooth microsomal membranes as well as two reactions 80 Figure 11. Cross-reactivity of anti zymogen granule membrane (IgG) with smooth microsomal membranes and NaBr-washed rough microsomal membranes from adult rat pancreas. A. Immunodiffusion was carried out as described in Figure 8A. The center well contained 20 ul of anti zymogen granule membrane (IgG). Wells 1 and 4 contained 20 ul of 0.4 ug/ul and 0.2 ug/ul solubilized zymogen granule membrane proteins, respec- tively. Wells 2, 3, 5, and 6 contained 20 ul of solubilized smooth microsomal membranes. Well 2, 12 ug/ul; Well 3, 3 ug/ul; Well 5, 0.8 ug/ul; Well 6, 0.2 ug/ul. This preparation of smooth microsomal membranes had not been washed with 0.25 M NaBr to remove adsorbed soluble proteins. B. Immunodiffusion was carried out as described in Figure 8A. The center well contained 20 ul of anti zymogen granule membrane (IgG). Wells 1 and 4 contained 20 ul of 0.4 ug/ul and 0.2 ug/ul solubilized zymogen granule membrane proteins, respectively. Wells 2, 3, 5, and 6 contained 20 ul of solubilized NaBr-washed rough microsomal membranes. Well 2, l8 ug/ul; Well 3, 4.5 ug/ul; Well 5, 1.1 ug/ul; Well 6, 0.3 ug/ul. 82 of identity with zymogen granule membranes and rough microsomal membranes. TheSe results suggest that common antigenic determinants may be present on these three membrane subfractions of the exocrine pancreas. In addition, they correlate well with recent evidence obtained in this laboratory for the existence of golgi membrane glycoproteins which appear similar to zymogen granule membrane glycoproteins (Ronzio and Mohrlok, 1977). Accumulation of Antibody-Precipitable Glycoproteins During the Secondary Transition Period of Pancreatic Differentiation The anti zymogen granule content (antiserum) and the treated anti zymogen granule membrane (IgG) described were used to assay protodifferentiated and differentiating pancreatic rudiments for secretory glycoproteins and zymogen granule membrane glycoproteins. Soluble and NaBr-washed particulate fractions were prepared as described in Methods from rudiments labeled with [3Hlfucose and [3H]glucosamine. The soluble fractions were analyzed for secretory glyc0proteins using the anti zymogen granule content (antiserum) in the double antibody precipitation procedure outlined in Methods. NaBr-washed particulate fractions were assayed for zymogen granule membrane glycoproteins using the treated anti zymogen granule membrane (IgG) in a similar double antibody system. Samples were titrated to equivalence with the rabbit antibodies. Final assay results were calculated as the percent of total acid-precipitable radioactivity added to the immunoassay reaction precipitated by the antibodies, minus the percent of acid-precipitable radioactivity precipitated by control antiserum (anti bovine serum albumin [antiserum]). 83 Figure 12 shows the accumulation of antibody-precipitable [3H]fucose- and [3H]glucosamine-1abeled glyc0proteins during the secondary transition period of pancreatic differentiation. Between the in vitro equivalent of 14-15 and l9-20 days gestation, a marked increase occurs in the amounts of [3H]fucose- and [3H]glucosamine- labeled soluble glycoproteins of the pancreas which are precipitable by the anti zymogen granule content (antiserum). Antibody-precipitable fucoproteins increase from 2.4 to 40.1% of the total soluble fuco- proteins during this time. A similar rise from 2.0 to 46.7% is observed for soluble [3H]glucosamine-labeled glyc0proteins. (Mean deviations for these values ranged from 1.2 to 4.6%.) It is presumed that few membrane-bound proteins were present in the soluble frac- tions prepared from labeled rudiments so that the antibody-precipitated proteins do, in fact, represent zymogen granule content glycoproteins. The results of rudiment fractionation shown in Figure 5 suggest that this is probably the case, since only a relatively small proportion of the radioactive particulate proteins appears in the profiles of the soluble fraction. . The accumulation of anti zymogen granule content (antiserum)- precipitable glycoproteins in differentiating pancreatic rudiments parallels the accumulation of pancreatic secretory proteins known to occur during differentiation. Furthermore, it supports earlier evidence that one or more of the secretory proteins is glycosylated. In contrast to the observed accumulation of soluble glycoproteins during differentiation of the rat pancreas is the apparent steady state of glycoproteins precipitated by the anti zymogen granule membrane (IgG). In vitro, during the secondary transition period of differentiation, the percent of total acid-precipitable particulate 84 Figure 12. Accumulation of "zymogen granule content-like" glycoproteins and "zymogen granule membrane-like" glycoproteins in differentiating rat exocrine pancreas. Soluble and NaBr-washed particulate fractions were obtained from cultured pancreatic rudiments labeled with [3H1fucose or [3H]glucosamine on the in vitro equivalents of 14-15, 17-18, and 19-20 days gestation. Soluble fractions were assayed for "zymogen granule content-like" glycoproteins using the anti zymogen granule content (antiserum) in the double antibody precipitin reaction described in Methods. NaBr-washed particulate fractions were assayed for "zymogen granule membrane-like" glycoproteins using the treated anti zymogen granule membrane (IgG) in a similar double antibody precipitin reaction. Results are expressed as the percent of the total acid-precipitable radio- activity precipitated from the sample by the antiserum or IgG at equivalence. The percent of acid-precipitable radioactivity precipitated by the anti BSA (antiserum) was subtracted from all values to correct for trapping of free [3H]sugar and non- specific precipitation of labeled glycoproteins in the immunoprecipitates. Results are shown for soluble fractions from [3H]fucose-labeled rudiments (-O-) and [3H]glucosamine- labeled rudiments (-O-), and for NaBr-washed particulate fractions from [3Hlfucose-1abeled rudiments (-Ar) and [3H]glucosamine-labeled rudiments (HAP). 85 50" p 0 la 3524 am ussasea Eztoosom 2323605664 IOI' o\o Embryonic Age (Days) Figure112 86 glycoproteins precipitated by this antibody ranges from 20.4 to 22.8% (mean deviation, 0.8-2.9%) for [3H]fucose-labeled glycoproteins and 20.6 to 21.9% (mean deviation, 3.2-4.0%) for [3H]glucosamine- labeled glycoproteins. The slight increase in the antibody- precipitable glycoproteins observed during this period is not statistically significant. These results suggest that the particulate glycoproteins, especially those of the zymogen granule membrane, are present in the protodifferentiated pancreas and are not induced parallel to the induction of glycosylated and non-glycosylated secretory proteins. Figure 13 shows the levels of incorporation of [3H1fucose and [3H]glucosamine into antibody-precipitable glycoproteins relative to the total "soluble" plus "particulate" protein present at a given stage of pancreatic differentiation. The specific radioactivity of [3Hlfucose-labeled zymogen granule content-like glyc0proteins increases nearly lOO—fold between the in vitro equivalent of 14-15 and l7—18 days of gestation. The specific radioactivity of [3H]glucosamine-labe1ed zymogen granule content-like glycoproteins increases by a factor of two hundred during this same period. After 17-18 days, no significant change occurs in the specific radioacti- vities of the glyc0proteins. (Mean deviations for 17-18 and 19-20 day [3H]glucosamine-labeled embryonic rudiments were 111 and 145 dpm/pg, respectively. Thus, the apparent increase is not statisti- cally significant.) Relative to total particulate plus soluble protein present in the cultured pancreatic rudiment, [3H]fucose and [3H]glucosamine incorporation into anti zymogen granule membrane (IgG)-precipitable particulate glycoproteins actually declines during differentiation. 87 Figure 13. Specific radioactivities of the "zymogen granule content-like" glycoproteins and "zymogen granule membrane-like" glycoproteins which accumulate in the rat exocrine pancreas during the secondary transition period of pancreatic differentiation. Results shown in Figure 12 for the accumulation of anti zymogen granule content (antiserum)-precipitable glycoproteins and anti zymogen granule membrane (IgG)-precipitable glyco- proteins during pancreatic differentiation were used to calculate the specific radioactivities of "zymogen granule content-like" and "zymogen granule membrane-like" glycoproteins present in differentiating pancreas. Specific radioactivities are eXpressed as the amount of acid-precipitable radioactivity precipitated by the antibody relative to total soluble plus particulate protein (dpm/pg protein). Results are shown for [3H]fucose-labeled.(-O-) and [3H]glucosamine-labe1ed (-O~) "zymogen granule content-like" glycoproteins and for [3H]fucose-labe1ed (-A-) and [3H]glucosamine-labeled (-A—) "zymogen granule membrane-like" glycoproteins. DPM/Fg Protein 300 200 IOO 88 1 1 l l7 Embryonic Age (Days) Figure 13 I9 89 There is approximately a two-fold decrease in the specific radio- activities of both [3H]fuoose- and [3H]glucosamine-labeled zymogen granule membrane-like glycoproteins during the transition from the protodifferentiated to the differentiated pancreas. Since the proportion of total particulate glycoproteins that is antibody- precipitable remains essentially constant during the secondary transition period of differentiation, this decline in the specific radioactivities of zymogen granule membrane-like glycoproteins is probably partially accounted for by the eight-fold increase per rudiment in the average soluble plus particulate protein content over this same time period. Identification of Antibody700mplexed Glycoproteins of Differentiating Rat Pancreas [3H1Fucose- and [3H]glucosamine-1abeled glycoproteins from soluble and particulate fractions of pancreatic rudiments at various stages of differentiation were precipitated with anti zymogen ‘granule content (antiserum) and anti zymogen granule membrane (IgG), respectively, in the double antibody systems described in Methods. The washed precipitates were solubilized in sample buffer and applied to SDS-polyacrylamide slab gels. After electrophoresis was complete, gels were prepared for autofluorography and, subse- quently, X-ray films were exposed to the dried gels. The resultant fluorographs are shown in Figure 148 (from [3H]fucose-labe1ed rudiments) and Figure 14C (from [3H]glucosamine-1abeled rudiments). Interpretations of the fluorographs must take into consideration two factors. First, the large quantities of goat gamma globulin present in the immunoprecipitates have distorted the electrophoretic 90 Figure 14. SDS-polyacrylamide gel electrophoresis analysis of radioimmunOprecipitates of soluble and particulate glycoproteins from differentiating pancreas. Soluble and particulate fractions were prepared from [3H]fucose- and [3H]glucosamine-labe1ed rudiments (cultured from day 14 of gestation) and analyzed for zymogen granule content glycoproteins and zymogen granule membrane glycoproteins using the described double antibody precipitation techniques. Soluble and particulate fractions and radioimmunoprecipitates of both fractions were solubilized in sample buffer and electrophoresed on 0.1% SDS-8% acrylamide slab gels. Figure A shows the Coomassie blue staining profiles for A) the soluble fraction from [3H]glucosamine-labeled 19-20 day rudiments, B) the immunOprecipitate of the soluble fraction from [3H]glucosamine- labeled 19-20 day rudiments, C) the particulate fraction from [3H]glucosaminerlabe1ed 19-20 day rudiments, D) the immunoprecipi- tate of the particulate fraction from [3H]glucosamine-1abe1ed 19-20 day rudiments, and E) protein standards, 1) porcine thyro- globulin, 2) human transferrin, 3) bovine serum albumin, 4) ovalbumin, and 5) bovine pancreatic ribonuclease A at the bromphenol blue tracking dye. Arrows indicate the 50,000 dalton subunit of gamma globulin. Figures B and C are fluorographs of unstained gels exposed to RP Royal X-Omat X-ray film for four weeks. Figure B shows the particulate and soluble fractions and the immunoprecipitates of those fractions from [3H]fucose-labe1ed pancreatic rudiments. A) Particulate fraction from 14-15 day rudiments (11,300 dpm); B) Immunoprecipitate of particulate fraction from 14-15 day rudiments (10,400 dpm); C) Particulate fraction from 17-18 day rudiments (29,200 dpm); D) Immunoprecipi- tate of particulate fraction from 17-18 day rudiments (25,700 dpm); E) Particulate fraction from 19-20 day rudiments (22,900 dpm); F) Immunoprecipitate of particulate fraction from 19-20 day rudiments (20,000 dpm); G) Soluble fraction from 17-18 day rudiments (16,000 dpm); H) Immunoprecipitate of soluble fraction from l7-18 day rudiments (13,800 dpm); I) Soluble fraction from 19-20 day rudiments (13,800 dpm); J) Immunoprecipitate of soluble fraction from 19-20 day rudiments (11,200 dpm). Figure C shows the particulate and soluble fractions and the immunoprecipitates of those fractions from [3H]glucosamine-labeled pancreatic rudiments. Array of samples is as in Figure B. A) 15,000 dpm; B) 12,800 dpm; C) 45,700 dpm; D) 48,300 dpm; E) 20,200 dpm; F) 16,900 dpm; G) 43,000 dpm; H) 36,800 dpm; I) 24,000 dpm; J) 21,400 dpm. The "dpm" values applied to gels for soluble and particulate fractions are acid-precipitable radioactivity values. 91 >b_.:§ manta”: Q 0 0‘ O Pr 1:2! u>.h<._w¢ 4. o. 3. >b:_IOI U>_P<..WS Figure 14 92 profiles for the immunoprecipitates. Thus, the relative mobilities of proteins migrating in the region of the 50,000 dalton subunit of the gamma globulin have been altered by the presence of the gamma globulin subunit. (See, in particular, Figure 14A. Arrows indicate the 50,000 dalton subunit of gamma globulin in the immunoprecipitate samples.) Furthermore, some of the radioactive species migrating in this region may have been masked by the large amount of gamma globulin present. _Second1y, comparisons of the immunoprecipitates from fractions of pancreatic rudiments at different stages of differentiation are not valid because of the wide variations in the amounts of radioactivity in the samples applied to the gels. Most of the glycoproteins previously observed to be particulate glycoproteins (Figure 5) were present in immunoprecipitates of particulate fractions of [3H]fucose- and [3Hngucosamine-labeled rudiments at all three stages of differentiation examined (proto- differentiated, mid-secondary transition, differentiated). Noticeably absent from the immunoprecipitate of the particulate fucoproteins of the in vitro equivalent of 14- to 15-day embryonic rudiments, however, is the high molecular weight species with relative mobility 0.12-0.15 (Band I) seen in Figure 23. Since this is not an induced fucoprotein, its absence is speculated to be due to an insufficient amount of radioactivity in the gel sample to permit its detection by autofluorography over a four-week film exposure period. This high molecular weight particulate protein does appear in the fluorograph of the immunoprecipitate of the particulate fraction from protodifferentiated pancreas labeled with [3H]glucosamine, though certainly not to the extent that it is 93 present in the 19- to 20—day cultured embryonic rudiment sample containing approximately the same amount of radioactivity. This high molecular weight particulate glycoprotein is thought to be GP-l of zymogen granule membranes (see Figure 6). Zymogen granule membrane component GP-2 appears (on the basis of its relative position in the electrophoretic profile) to migrate with a relative mobility of 0.28-0.35 in the anti zymogen granule membrane (IgG) immunoprecipitates of particulate fractions from differentiating pancreas. It is present at a somewhat higher position on the gels than usual due to the obstruction of its normal migration by the 50,000 dalton subunit of gamma globulin in the samples. Not all of the particulate glycoproteins complexed by the anti zymogen granule membrane (IgG) correspond with known zymogen granule membrane glycoproteins. In view of the cross-reactivity of the IgG with smooth and rough microsomal membranes of the adult rat pancreas, these other species are thought probably to be glycoproteins of the corresponding membrane structures in differentiating pancreas. The number of glycoproteins precipitated by anti zymogen granule content (antiserum) in reactions with soluble fractions from [3H]fucose- labeled and [3H]glucosamine-1abe1ed rudiments is few. Only one [3H]fucose-labeled species and only three [3H]glucosamine-labeled species are observed in the in vitro equivalents of 17- to 189day and 19- to 20—day embryonic pancreatic rudiments. The electrophoretic migration of the one fucoprotein present in immunoprecipitates of soluble fractions from labeled rudiments is distorted by the gamma globulin subunit from the precipitin reactions. It is not clear whether this fucoprotein corresponds to one of the 94 induced fucoproteins that migrate on this gel with relative mobilities of 0.54-0.62 and 0.69-0.71 or to the non-induced fucoprotein with a relative mobility of 0.64-0.66. The results of titration of the soluble fractions of fucose-labeled rudiments with the anti zymogen granule content (antiserum) showed the complexed glycoprotein(s) to be induced during differentiation (Figure 12). This result and the curvature of the fucoprotein band in the gel strongly suggest that the anti zymogen granule content (antiserum) complexes with the induced fucoprotein (previously designated as Band III) seen to have a relative mobility of 0.54-0.62 on thisgel. Earlier, this fucoprotein was tentatively identified as amylase. The three major glycoproteins of soluble fractions from [3H]glucosamine-labeled pancreatic rudiments which are complexed by the anti zymogen granule content (antiserum) have relative mobilities of 0.08—0.11, 0.31-0.38, and 0.68—0.69 (in the immunoprecipitates of the soluble fraction from the in vitro equivalent of 19- to 20-day embryonic pancreases). The soluble glyc0protein of unusually high molecular weight was observed previously in gels of differentiating pancreatic rudiments labeled with [3H]fucose and [3H]glucosamine. It did not appear as an induced protein or glycoprotein species and is not thought to be a secretory protein because of its high molecular weight. The migrations of the other two species have been distorted by the goat gamma globulin subunit. 0n the basis of the curvature of the bands and the relative abundance of glycoproteins in the soluble fractions from which the precipitates were obtained, it is suggested that these two precipitated species correspond to the two differentiation-induced [3H]glucosamine-labeled glycoproteins previously designated as Band II and Band III. Both glycoproteins 95 have been tentatively identified as secretory proteins, the faster migrating species as amylase. The absence of the intensely labeled fucoprotein of adult rat pancreas zymogen granule contents with relative mobility 0.78 from antibody precipitates of soluble proteins from pancreatic rudiments is not understood. It may be that it simply was not detected by the x-ray film over the four-week exposure period, or that it was a zymogen granule content protein not recognized by the antiserum. Competition of Adult Rat Pancreas Zymogen Granule Membrane and Zymogen Granule Content Proteins with Glycoproteins of Differentiated Embryonic Pancreatic Rudiments for Antibody Binding Further identification of the antibody-precipitated glycopro- teins from.differentiated pancreas (14-day embryonic rudiments cultured to the equivalent of day 19 and then labeled 18 hours with precursor) was obtained by the use of adult rat pancreas zymogen granule content proteins and zymogen granule membrane proteins as competitive antigens in the double antibody precipitation reactions. To determine the pr0portion of antibody-precipitable soluble glyco- proteins that were actually zymogen granule content proteins, 10 ul of anti zymogen granule content (antiserum) were first incubated at 37°C for thirty minutes with 24 pg of zymogen granule content proteins (10 ul of 2.4 pg protein/pl) in a total reaction volume of 90 ul. Figure 7 indicates that this is well into the antigen excess region of the titration curve for the antiserum in double antibody precipitation reactions. Ten microliters of soluble fractions from [3H]fucose- or [3H]glucosamine-labeled rudiments (diluted for maximum precipitation) were then added to the reaction 96 mixtures and the tubes were incubated at 37°C for thirty minutes more. Goat antirabbit (gamma globulin) was subsequently added to the reactions which were then incubated at 25°C for one hour and finally stored at 0-4°C for 24 hours. Precipitates were collected and washed and radioactivity levels determined as described in Methods. "Percent Competition" by the added zymogen granule content proteins was calculated as follows: x 100. where A = [(gpm precipitated by anti zymogen granule content lantiseruml) (acid-precipitable dpm added to reaction) x 100%] B (dpm precipitated by anti BSA [antiserum]) x 100%] (acid-precipitable dpm added to control reaction) (dpm precipitated by anti zymogen granule content (antiserum) in the presence of competitor zymogen granule C = I content proteins) x 100%] (acid-precipitable dpm added to reaction) Results for two separate cultures of [3H]fucose-labeled rudiments and two separate cultures of [3H]glucosamine-labe1ed rudiments showed a range of competition for both types of soluble glycoproteins. Zymogen granule content proteins competed with soluble [3H]fucose- labeled glycoproteins from 10.6% to 32.6%, while competition with soluble [3H]glucosamine-labeled glycoproteins for binding by the antiserum ranged from 10.9% to 37.2%. These results suggest that possibly not more than a third of the soluble embryonic pancreatic glycoproteins complexed by the anti zymogen granule content (antiserum) were actually secretory glycoproteins. The possibility that the i 97 zymogen granule content preparation was an inadequate competitor to embryonic pancreatic secretory proteins or that an insufficient amount of competitive zymogen granule content proteins was added to the assay reaction for complete inhibition of binding by embryonic pancreatic secretory glyc0proteins cannot be excluded, however. Alternatively, the relative proportions of adult rat pancreas zymogen granule content proteins and labeled embryonic pancreas zymogen granule content proteins may have been such that an equal opportunity for combining with the antibody binding sites did not exist for the two antigens. If this was the case, the experiments are not an accurate measure of competition of the two antigens for binding. Similar experiments were carried out using zymogen granule membrane proteins to compete with particulate [3H]fucose- or [3H]glucosamine-labe1ed glycoproteins from the cultured equivalent of 19- to 20-day embryonic rudiments for binding by the anti zymogen granule membrane (196). Thirteen micrograms of zymogen granule membrane protein were mixed with appropriately diluted samples of radioactive particulate fractions in phosphate-detergent buffer (pH 7.4), and then anti zymogen granule membrane (IgG) (10 ul) was added to the mixture. The procedure for double antibody precipita- tion of particulate glycoproteins was then carried out as described in Methods. Precipitates were collected, washed, solubilized, and samples were then taken for liquid scintillation counting. Results were calculated as in the competition experiments for soluble glycoproteins. The results of experiments involving two cultures of [3H]fucose-labeled rudiments showed that 31.6% (i1.2%, mean deviation) of the particulate glycoproteins precipitated by the 98 gamma globulin represented zymogen granule membrane-like glycoproteins. Similarly, 32.0% (i8.0%, mean deviation) of the [3H]glucosamine- labeled particulate glycoproteins complexed by this antibody were symogen granule membrane-like glycoproteins. The results of these competition experiments are consistent with the cross-reactivity of the anti zymogen granule membrane (IgG) with smooth and rough microsomal membranes from adult rat pancreas and the results of the gel electrophoresis-fluorographic analysis of immunoprecipitates of particulate fractions from embryonic pancreas. They suggest that not all of the particulate glycoproteins complexed by the antibody are zymogen granule membrane glyc0proteins. It is likely, however, that the percent competition observed in these experiments does not represent the maximum amount of zymogen granule membrane or zymogen granule membrane-like species in the in vitro 19- to 20-day embryonic pancreatic rudiment. Since the equivalence point was never reached in titrations of the zymogen granule membrane (IgG) with zymogen granule membranes from adult rat pancreas, the problem was to establish the amount of competitor zymogen granule membrane proteins which would be in the antigen excess region of the precipitation curve for the double antibody assay reaction. Thus, it is proposed that a minimum of 30% of the particulate glycoproteins from embryonic pancreas that are complexed by the anti zymogen granule membrane (IgG) are zymogen granule membrane or zymogen granule membrane-like glycoproteins. (It will be noted that the radioimmunoassay and competition assay procedures used in these studies were not the standard procedures normally used. Certain technical difficulties prevented 99 the use of the standard radioimmunoassay methods. Subsequently, interpretation of the results is somewhat limited.) Radioimmunoassay of Pancreatic Rudiments Cultured in the Presence of S-Bromodeoxy- uridine for Zymogen Granule Content and Zymogen Granule Membrane Glycoproteins A further check for differentiation-dependent induction of zymogen granule membrane-like glycoproteins and secretory glycoproteins was made on pancreatic rudiments cultured in the presence of S-BrdU to inhibit differentiation. The particulate and soluble fractions from [BH]glucosamine-1abeled pancreatic rudiments that had been cultured in the presence of 5-BrdU from day 14 to the equivalent of day 20 of gestation were assayed for zymogen granule membrane glycoproteins and zymogen granule content glycoproteins using the double antibody radioimmunoassays described in Methods. Of the acid-precipitable particulate-bound radioactivity, 22.9% was precipitated by the anti zymogen granule membrane (196). Only 8.7% of the acid-precipitable soluble glycoproteins from these pancreatic rudiments was precipitated by the anti zymogen granule content (antiserum). The specific radioactivities of the particulate and soluble glycoproteins were 336 and 72 dpm/pg (particulate plus soluble) protein, respectively. Comparison of these values with those for cultured 14- to lS-day and 19- to 20-day control [3H]glucosamine-labeled embryonic rudiments indicates that, while 5-BrdU inhibits the synthesis of soluble (secretory) glycoproteins, it has no effect on particulate zymogen granule membrane or zymogen granule membrane-like glycoprotein synthesis. Precipitation of the particulate fraction from the [3H]glucosamine-labeled 5-BrdU- treated rudiments in the presence of competitor zymogen granule 100 membrane proteins showed that a minimum of 39% of the antibody- precipitable radioactivity was associated with zymogen granule membrane or zymogen granule membrane-like glycoproteins. This result agrees well with the value obtained for control [3H]glucosamine- labeled rudiments (32%) and lends additional support to the proposal that zymogen granule membrane glycoproteins follow a unique differentiation program. DISCUSSION Summary of Results The results of these studies confirm the earlier finding in this laboratory that a large number of constitutive glycoprotein species characterize the rat exocrine pancreas throughout differen- tiation. In addition, the results of these studies have shown that a minimum of six glyc0protein species undergo differentiation- dependent induction with respect to de novo protein synthesis and/or glycosylation (in contrast to a previous report by Walther and co—workers, 1974). Essentially the same glycoprotein species appear to be induced regardless of the sugar precursor used to study induction; however, specific induced glycoproteins preferentially incorporate certain sugars. The largest number of induced glyco- proteins is detected by [3H]fucose labeling; some of these glycoproteins, however, appear to be minor constituents of the pancreas. Fewer induced species are detected with [3H1N-acetyl- mannosamine labeling, this sugar being incorporated in much smaller quantities than fucose and glucosamine. Of all the induced glycoproteins observed, only one is found primarily in the particulate fraction of the tissue. The others are all soluble glycoproteins. The differentiation-induced particulate glycoprotein was estimated to have a molecular weight of 144,000-154,000 daltons. \ ) Induction of this glycoprotein is essentially complete by day 17-18 101 102 of gestation (the time when zymogen granules first appear in the acinar cell) and is apparent only in fluorograph profiles of embryonic rudiments labeled with [3nglucosamine and [3H]N-acetyl- mannosamine. Induction, thus, seems to occur only in terms of extension and completion of the carbohydrate moiety since 1) synthesis of the protein portion of this glycoprotein is not induced during the secondary transition, 2) incorporation of glucosamine is most likely to be throughout the carbohydrate chains, and 3) incorporation of N-acetylmannosamine is expected to be as carbohydrate chain- terminating sialic acid. This particulate glycoprotein does label with [3H]fucose; however, fucose incorporation does not increase during the secondary transition period of differentiation, suggesting that completion of certain carbohydrate chains of the glycoprotein with fucose is enzymatically possible in the protodifferentiated pancreas. It is suggested that, rather than being a differentiation- induced protein, this particulate glycoprotein is the substrate for one or more differentiation-induced glycosyl transferases. Comparison of radioactive profiles for [3H]fucose-labeled zymogen granule membrane from adult rat pancreas with the [3H1fucose profile for the particulate fraction from pancreatic rudiments cultured from day 14 to the equivalent of day 20 of gestation suggests that this unique high molecular weight glyc0protein corresponds to the zymogen granule membrane component GP-l. Precipitation of the glycoprotein by anti zymogen granule membrane (IgG) is further evidence that it is a zymogen granule membrane-like glycoprotein. A minimum of two of the soluble induced glycoproteins are thought to be secretory proteins. The molecular weights of these 103 glyc0proteins (calculated from their mobilities in SDS-polyacrylamide gels), their correspondence with similarly migrating induced protein species most certain to be secretory proteins, their relative abundance, and their recognition by the anti zymogen granule content (antiserum) are all evidence supporting the tentative identification of these induced glycoproteins as secretory proteins. These two induced glycoproteins have molecular weights of approximately 57,000 and 73,000 daltons. The former glycoprotein is thought to be amylase. The latter glycoprotein corresponds to a secretory protein observed by Walther and co-workers (1974) in differentiating pancreatic rudiments. It was not identified in terms of enzymatic activity, nor was any indication given that it was a glycoprotein. As was noted earlier, Ronzio and colleagues have observed a similar glycoprotein in zymogen granule contents from adult rat pancreas labeled in vitro with radioactive glucosamine. A third soluble induced glycoprotein species is tentatively identified as procarboxypeptidase A or lipase solely on the basis of molecular weight calculations from its mobility on SDS-polyacrylamide gels. It was not, however, apparent in the electrophoretic profiles of anti zymogen granule content (antiserum) precipitates of soluble proteins from labeled pancreatic rudiments. Other soluble glycoproteins present in pancreatic rudiments appear to undergo induction primarily in terms of glycosylation rather than protein synthesis. All are lower molecular weight glycoproteins. A 34,000-38,800 dalton species is labeled only by radioactive fucose. A 29,800-34,000 dalton species labels with radioactive fucose and glucosamine. A third induced glycoprotein, molecular weight 26,600-29,800 daltons, labels with radioactive 104 N-acetylmannosamine as well as the two other glycoprotein precursors. This lowest molecular weight species is thought not to correspond with the similarly migrating induced protein of pancreatic rudiments tenta- tively identified as chymotrypsinogen/trypsinogen. Slight variations in mobility prevent absolute assignment of this glycoprotein species as one of these secretory proteins, although Ronzio et al. (to be published) have observed a similar glycoprotein in adult rat pancreas zymogen granule contents. Because these three glycoproteins are induced only with respect to glycosylation, the implication is that they also must be the substrates for induced glycosyltransferase activities. The possibility that these minor induced glycoproteins are actually minor secretory proteins, the enzymatic natures of which have not been determined, cannot be eliminated. Alternatively, they are non- secretory proteins involved in the differentiation process. Data obtained by radioimmunoassay for zymogen granule membrane- like glycoproteins in differentiating pancreas suggests that zymogen granule membrane glycoproteins, in general, do not undergo induction during the secondary transition period. This is evidenced by the presence of a glycoprotein similar to and suspected to be the major zymogen granule membrane component, glycoprotein GP-2, in the proto- differentiated as well as the differentiated pancreatic rudiment and in immunoprecipitates of membrane proteins from protodifferentiated and differentiated rudiments. This seems to contradict the observation that a GP-l-like particulate glycoprotein appears to undergo differentiation-dependent induction with respect to glycosylation. SDS-polyacrylamide gel electrophoresis analysis of the anti zymogen granule membrane (IgG) immunoprecipitates of particulate fractions from embryonic pancreas indicates, however, the presence of a large 105 number of non-induced glycoprotein species, not all of which are zymogen granule membrane-like glycoproteins. Thus, it is suggested that the induced glycosylation of a single zymogen granule membrane glycoprotein species during differentiation may have remained undetected by the radioimmunoassay technique utilized. Results of radioimmunoassays for zymogen granule content-like proteins in soluble fractions from differentiating pancreatic rudiments were also indicative of the presence of one or more major differentiation-induced secretory glycoproteins. The absence of Such glycoproteins in rudiments cultured in the presence of 5-BrdU and labeled with [3H]glucosamine is further evidence for their involvement in the secondary transition events leading to the differentiated pancreas. Glycosylation of Pancreatic Secretory Glycoproteins To date, characterization of rat pancreas secretory proteins has not provided evidence for their glycosylation. Strong evidence for glycosylation of rat pancreas secretory proteins, however, is available from gas-liquid chromatographic analysis of the rat pancreas zymogen granule content proteins (Ronzio et al., submitted for publication). Evidence for secretory protein glycosylation has been provided by in vitro labeling of rat pancreas with radioactive glycoprotein precursors and subsequent observation of either glycosylated secretory granule content proteins on SDS-polyacrylamide gels (Ronzio et al., submitted for publication) or the release of glycosylated secretory products in response to stimulation of the labeled tissue with secretagogues (Volkl et al., 1976; Kronquist et al., 1977). Furthermore, evidence exists from studies of the homologous enzymes from other animal species and other tissues for 106 glycosylation of pancreatic secretory proteins. For example, amylase has been isolated from porcine pancreas and human parotid and submandibular glands as mixtures of glycosylated and non? glycosylated forms (Beaupoil-Abadie et al., 1973; Watanabe and Keller, 1974). Ribonucleases isolated from bovine and porcine pancreas have been shown to be highly glycosylated (Plummer and Hirs, 1963; Reinhold et al., 1968). Porcine pancreatic lipases have also been found to be glycosylated (Garner and Smith, 1972). Interestingly enough, in the case of porcine ribonuclease, enzyme activity does not seem to be affected significantly by the presence or absence of the carbohydrate moiety. In view of the current ideas on post-translational processing of secretory proteins, in particular as proposed in the "signal” hypothesis of Blobel and Sabatini (1971; also, Blobel and Dobberstein, 1975), one might speculate that glycosylation of secretory proteins may also provide a transient tag by which secretory proteins are recognized for packaging. Devillers-Thiery and co-workers (1975) have found that a number of dog pancreas secretory proteins, in particular trypsinogen, are synthesized in an in vitro translation system as pre-proenzymes containing an amino terminal extension of sixteen amino acid residues. This has been called the signal sequence of the secretory proteins and is speculated to serve in guiding the secretory proteins into the cisternae of the rough endoplasmic reticulum for eventual shuttling into the Golgi and finally into packaging vesicles. Secretory protein glycosylation begun in the rough endoplasmic reticulum and completed in the Golgi stacks might be seen to serve as a further signal mechanism for the sequestration and packaging of secretory proteins. 107 That isolated rat pancreas secretory proteins have gone unremarked as glycoproteins may not be too surprising. Just as the signal peptide is hypothesized to be removed from the pre-proenzyme during passage into the cisternae of the rough endoplasmic reticulum, so might a signal carbohydrate moiety be removed from the proenzymes or amylase prior to or during concentration in secretory vesicles or during secretion. However, no evidence for such a signal carbo- hydrate moiety on secretory proteins is presently available. Less hypothetical explanations may be invoked, however, for the unremarked glycosylation of rat pancreas secretory proteins. The presence of active glycosidases such as amylase in pancreatic tissue may have resulted in loss of carbohydrate from secretory proteins during secretory protein isolation. The influence of diet on the carbohydrate content of secretory proteins has not been investigated. It is known that the composition of the diet, be it high in protein or high in carbohydrate, determines the relative rates of synthesis of chymotrypsinogen and amylase (Reboud et al., 1964; Reboud et al., 1966). The low levels of fucosylation observed for zymogen granule content proteins from adult tissue labeled in vitro with [3H]fucose in contrast with fucosylation of proteins of embryonic pancreas is a puzzle. An altered glycosylation of adult pancreatic secretory proteins relative to embryonic secretory proteins is possible. The synthesis and/or turnover of the carbohydrate moieties of secretory proteins of non-stimulated adult pancreas may be slow compared to the rate of synthesis of glycosylated secretory proteins of differentiating pancreas. A reduced rate of fucose transport into the adult tissue or the presence of a large intracellular pool of 108 fucose might also explain the low levels of incorporation of radioactive fucose into secretory proteins of the adult pancreas. Metabolism of Glycoprotein Precursors One criticism.which may be made of these experiments is that sugar-labeled rudiments were not analyzed for conversion of the labeling sugars into other glyc0protein precursors or intermediates of glycolysis. It has been found that when radioactive L-fucose is administered to rats, 2% or less of the radioactivity appears in labeled C02, approximately 30% of the radioactivity appears as free fucose in the urine, and the remainder is found free as fucose, fucose-l-P, or GDP-fucose, or is bound to protein in various tissues. No conversion of fucose to other glycoprotein precursors has been observed (Coffey et al., 1964; Bekesi and Winzler, 1967). Radioactive labeling of cells and tissues with glucosamine has generally resulted in the incorporation of radioactive N—acetyl- glucosamine, N-acetylgalactosamine, or N-acylneuraminic acids into glycoprotein. In the adult rat pancreas, it has been observed that 70-94% of the protein-bound radioactivity resulting from labeling of the tissue in vitro with [3H1- or [1”C]glucosamine is recoverable as glucosamine (Ronzio, unpublished observations). One might expect the incorporation of glucosamine by embryonic pancreatic rudiments in culture to be similar to the incorporation into glycoproteins by adult tissue. The question of metabolic reutilization of the glucosamine through conversion to glycolytic intermediates, in this case, Cannot be answered. N-Acetylmannosamine is a very proximal precursor of sialic acids; it is not incorporated itself directly into glycoproteins. 109 Though mannose is a precursor in glycoprotein synthesis, the existence of enzymes for the conversion of N-acetylmannosamine to mannose has not been reported. Biogenesis of Zymogen Granule Membranes Current theories of membrane biogenesis, elaborated by Morré and colleagues (1974), involve two concepts, that of membrane flow and that of membrane differentiation. It is generally considered that membrane of the endoplasmic reticulum with its complement of proteins and glycoproteins is gradually transported to the Golgi stacks via transporting vesicles which fuse with the outermost cisternae of the cis side of the Golgi. Subsequently, membranes of secretory vesicles, zymogen granule membranes, appear to derive from the last trans cisternae of the Golgi. Each membrane compart- ment of the cell, nevertheless, maintains a unique protein- glycoprotein composition (Meldolesi and Cova, 1972), though it appears that, in keeping with the membrane flow hypothesis, certain constituents characteristic of each type of membrane are carried over to the next type of membrane in the biogenic pathway. Evidence has appeared that additional protein synthesis and insertion into membranes does occur at the Golgi apparatus through the activity of Golgi-associated polysomes in rat liver (Elder and Morré, 1976). Indications that intracellular membranes of the rat exocrine pancreas undergo a similar biogenic process are derived from the work of Volkl and co-workers (1976) and Ronzio and colleagues (work submitted for publication). The former group has shown, using electron microscopic techniques, the transport of membrane vesicles from the endoplasmic reticulum to the cis face of the Golgi and the 110 formation of condensing vacuoles (presumably zymogen granule precursors) at the trans face of the Golgi. SDS-polyacrylamide gel electrophoresis analysis of membrane subfractions from rat pancreas by Ronzio and co-workers has shown the distinct protein-glycoprotein character of each of the intracellular membranes but, in addition, has suggested the existence of at least one zymogen granule membrane- like glycoprotein in Golgi membrane. Immunochemical studies by Ronzio and Mohrlok (1977) have provided evidence for zymogen granule membrane glycoprotein precursors in the Golgi which serve as endogenous substrates for a Golgi-associated galactosyltransferase. Studies of the cross-reactivity of anti zymogen granule membrane (IgG) with rough and smooth microsomal membranes presented in this report supply further evidence for the similarity of the intracellular membranes of the rat exocrine pancreas. SDS- polyacrylamide gel electrophoresis of anti zymogen granule membrane (IgG) immunoprecipitates of particulate glycoproteins from differen- tiating rat pancreas indicate not only the presence of glycoproteins similar to the two major components of zymogen granule membranes, GP-l and GP-2, but also the presence of a number of other high molecular weight membrane-bound glycoproteins. Presumably these are glycoprotein constituents of rough and smooth microsomal membranes (including Golgi membranes), and though direct evidence is not available to the effect, it may also be that plasma membrane glycoproteins are present as well. Glycosyltransferase Induction During Pancreatic Differentiation Differentiation-dependent induction of glycoprotein synthesis in the rat pancreas appears to be of two types. In some cases 111 glycosylation is coordinate with induced protein synthesis. In other cases, induced glycosylation occurs with non-induced protein species as substrates, thus implying the induction of specific glycosyltransferase activities. Direct involvement of glycosyltransferases as differentiation- induced enzymes of the rat pancreas is supported by the work of Carlson and colleagues (1973) and Marvel and Ronzio (1976). Carlson has found that a forty-fold increase occurs in the specific activity of a pancreatic N-acetylgalactosamine-protein:galactosyltransferase during the period from 11 to 12 days in utero to birth. S-BrdU, however, does not inhibit the differentiation-dependent rise in the galactosyltransferase activity, suggesting that galactosyltransferase synthesis and secretory protein synthesis are regulated differently during pancreatic develOpment. Similarly, a two-fold increase in the specific activity of a pancreatic N-acetylglucosamine-protein: galactosyltransferase has been observed by Marvel and Ronzio to occur during the secondary transition period of rat pancreas differentiation. In the case of those soluble glyc0protein species of the embryonic pancreas, thought to be secretory proteins, it is not possible, on the basis of the experiments described in this report, to assess the induction oqupecific glycosyltransferases associated with the processing of the newly synthesized proteins. No evidence exists to indicate that simultaneous induction of glycosyltransferase activities and endogenous substrates in the form of secretory proteins does not occur. In the event that the concentrations of endogenous substrates are so low as to be rate limiting in 112 glycosylation reactions, it is an apparent induction of transferases that will be observed, however. Pancreatic Glycoprotein Synthesis and the Model for Pancreatic Differentiation Of major concern to these studies was determining whether or not induced glycoprotein synthesis occurs synchronously with secretory protein synthesis during pancreatic differentiation and can, thus, be accounted for by the current model described by Rutter and colleagues for pancreatic differentiation (Figure l). The basic problem involves the number and timing of regulatory events required for the induction of glycoprotein synthesis during the secondary transition from the protodifferentiated to the differentiated state. The results of the studies reported here suggest that several secretory proteins, among them amylase, are glycosylated. However, the experiments were not designed such that the induction of glycosyltransferase activities responsible for the glycosylation of the secretory proteins could be observed independently of the induction of protein synthesis. Thus, it cannot be ascertained whether these specific glycosyltransferases are constitutive enzymes of the pancreas, are induced in advance of the induction of secretory protein synthesis (thereby indicating independent regulation in the synthesis of glycosyltransferases and the secretory proteins), or are induced simultaneously with the occurrence of increased secretory protein synthesis. In the latter case, additional regulatory events would not be required for initiation of glycosyl- transferase synthesis during differentiation. In the case of glycoproteins which appear to be induced during differentiation solely in terms of glycosylation, the implication is 113 for induction of specific glycosyltransferases. This could occur Synchronously with the induction of secretory proteins. Thus, an additional regulatory event again is not necessitated in the differentiation scheme. This is further supported by the inhibitory effect of 5-BrdU on the induced glycosylation of these proteins. Increased synthesis of the major zymogen granule membrane component, GP-2, does not appear to occur during the secondary transition period of pancreatic differentiation. Strong evidence is accumulating for the presence of zymogen granule membrane proteins and glycoproteins or their precursors as constituents of the Golgi membrane (and possibly the rough endoplasmic reticulum as well). This suggests that zymogen granule membrane proteins are, unlike the pancreatic secretory proteins, either constitutive proteins of the cell or that the major induction of synthesis of zymogen granule membrane protein precursors occurs before the start of the secondary transition period of pancreatic differentiation. Estimates that three percent or less of the membrane proteins of smooth and rough endoplasmic reticulum can be GP-2 (Ronzio, unpublished), however, suggest that the zymogen granule membrane glycoproteins are probably not constitutive glyc0proteins of the pancreas cell. On the other_ hand, zymogen granule membrane glycoproteins (or their protein precursors) may well be induced during the primary transition period of pancreatic differentiation, the period when low but detectable amounts of the secretory proteins are present in the cell. The differentiation-induced completion of glycosylation of a zymogen granule membrane GP-l-like glycoprotein may be the differentiation event which signals the conversion of Golgi membrane into zymogen granule membranes, which then begin to bud from the 114 Golgi in the formation of zymogen granules. This differentiation of Golgi membrane to form.zymogen granule membrane potentially requires the induction of a number of glycosyltransferase activities. Once again, the induction of glycosyltransferase activity need not occur independently of the induction of the secretory proteins. The results of these studies, therefore, are not sufficient to indicate that the current model for pancreatic differentiation does not hold for glycoprotein synthesis. They do suggest that the induction of a GP-2-like zymogen granule membrane glycoprotein must occur before the secondary transition period of differentiation. Whether there is a regulatory event other than the primary transition event of pancreatic differentiation that is required for the induction of this glycoprotein, or the primary transition is actually the major induction period for GP-2 (and possibly other zymogen granule membrane glyc0proteins), remains to be determined. Suggestions for Future Investigations Now that differentiation-dependent induction of glycoprotein synthesis is known to occur in the rat exocrine pancreas, several very pertinent lines of investigation should be pursued. First, the question of glycosylation of adult rat exocrine pancreas secretory proteins remains to be answered directly. Glycosylation may be of significance with respect to the secretory proteins. Of immense importance in furthering the understanding of the differentiation process in the exocrine pancreas will be studies of the differentiation-dependent induction of glycosyltransferases, in particular fucosyl- and sialyltransferases, and the identification of their specific endogenous substrates. The possible role of 115 membrane glycosylation in the formation of zymogen granule membranes has been suggested by the results reported here and needs to be studied in detail. Finally, the synthesis of membrane proteins, especially zymogen granule membrane glycoproteins and their precursors, must be investigated in the protodifferentiated pancreas to determine whether or not they undergo a differentiation-dependent induction prior to the secondary transition that requires the inclusion of an additional regulatory event (other than the primary transition event) in the model for exocrine pancreas differentiation. LIST OF REFERENCES LIST OF REFERENCES Amsterdam, A., and Jamieson, J. D. (1974) J. Cell Biol. 63, 1037- 1056. Beaupoil—Abadie, B., Raffalli, M., Cozzone, P., and Marchis-Mouren, G. (1973) Biochim. Biophys. Acta 297, 436-440. Bekesi, J. G., and Winzler, R. J. (1967) J. Biol. Chem. 242, 3873- 3879. Bernfeld, P. (1955) In Methods in Enzymology, eds. Colowick, S. P., and Kaplan, N. 0. (Academic Press, New York), Vbl. 1, pp. 149-158. Blobel, G., and Sabatini, D. D. (1971) In Biomembranes, ed. Manson, L. A. (Plenum Publishing Corp., New York), Vol. 2, pp. 193-195. Blobel, G., and Dobberstein, B. (1975) J. Cell Biol. 67, 835-851. Bonner, W. M., and Laskey, R. A. (1974) Eur. J. Biochem. 46, 83-88. Carlson, D. M., David, J., and Rutter, W. J. (1973) Arch. Biochem. Biophys. 157, 605-612. Coffey, J. W., Miller, 0. N., and Sellinger, 0. Z. (1964) J. Biol. Chem. 239, 4011-4017. DeCrombrugghe, B., Pitt-Rivers, R., and Edelhoch, H. (1966) J. Biol. Chem. 241, 2766-2773. Devillers-Thiery, A., Kindt, T., Scheele, G., and Blobel, G. (1975) Proc. Nat. Acad. Sci. USA 72, 5016-5020. Eagle, H. (1959) Science 130, 432-437. Elder, J. H., and Morré, J. D. (1976). J. Biol. Chem. 251, 5054- 5068. Etzler, M. E., and Branstrator, M. L. (1974) J. Cell Biol. 62, 329- 343. Fraker, P. J., Cicurel, L., and Nisonoff, A. (1974) J. Immunol. 113, 791-797. 116 117 Garner, C. W., Jr., and Smith, L. C. (1972) J. Biol. Chem. 247, 561-565. Geltoskey, J. E., Sui, C., and Lerner, R. A. (1976) Cell 8, 391-396. GIBCO Catalogue (1976-1977), Grand Island Biological Company, Grand Island, New York. Githens, S., Pictet, R., Phelps, P., and Rutter, W. J. (1976) J. Cell Biol. 71, 341-356. Greenaway, P. J., and LeVine, D. (1973) Nature (London), New Biol. 241, 191-192. Jamieson, J. D., and Palade, G. E. (1967) J. Cell Biol. 34, 577-596. Kay, C. M., Smillie, L. B., and Hilderman, F. A. (1961) J. Biol. Chem. 236, 118-121. Kemp, J. D., Walther, B. T., and Rutter, W. J. (1972) J. Biol. Chem. 247, 3941-3952. Kleinschuster, S. J., and Moscona, A. A. (1972) Exp. Cell. Res. 70, 397-410. Krebs, E. (1950) Biochim. BiOphys. Acta 4, 249-269. Kronquist, K. E., Elmahdy, A., and Ronzio, R. A. (1977) Arch. Biochem. Biophys. 182, 188-196. Laemmli, U. K. (1970) Nature (London) 227, 680-685. Laskey, R. A., and Mills, A. D. (1975) Eur. J. Biochem. 56, 335-341. LeVine, D., Kaplan, M. J., and Greenaway, P. J. (1972) Biochem. J. 129, 847-856. Levy, H. B., and Sober, H. A. (1960) Proc. Soc. Exp. Biol. Med. 103, 250-252. Lewis, D. S., MacDonald, R. J., Kronquist, K. E., and Ronzio, R. A. (1977) FEBS Lett. 76, 115-119. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. MacDonald, R. J., and Ronzio, R. A. (1972) Biochem. Biophys. Res. Marvel, C., and Ronzio, R. A. (1976) Abstracts of Papers, l72nd ACS National Meeting, San Francisco, California. Biol. abstract 216. Meldolesi, J., Jamieson, J. D., and Palade, G. E. (1971) J. Cell Biol. 49, 109-129. 118 Meldolesi, J., and Cova, D. (1972) J. Cell Biol. 55, 1-18. Morré, J. D., Keenan, T. W., and Huang, C. M. (1974) In Advances in Cytopharmacology, eds. Ceccarelli, B., Clementi, F., and Meldolesi, J. (Raven Press, New York), V01. 2, pp. 107-125. Ouchterlony, 5. (1958) Progr. Allergy 5, 1-78. Patterson, M. S., and Greene, R. C. (1965) Anal. Chem. 37, 854-857. Pharmacia Fine Chemicals AB (1972) CNBr-activated Sepharose 4B fer Immobilization of Biopolymers, Con A-Sepharose for Affinity Chromatography of Polysaccharides and Glycoproteins. Uppsala, Sweden. Pictet, R. L., Clark, W. R., Williams, R. H., and Rutter, W. J. Plummer, T. H., and Hirs, C. H. W. (1963) J. Biol. Chem. 238, 1396-1401. Reboud, J. P., Pasero, L., and Desnuelle, P. (1964) Biochem. Biophys. Res. Commun. 17, 347-351. Reboud, J. P., Marchis-Mouren, G., Pasero, L., Cozzone, A., and Desnuelle, P. (1966) Biochim. Biophys. Acta 117, 351-367. Reinhold, V. N., Dunne, F. T., Wriston, J. C., Schwarz, M., Sarda, L., and Hirs, C. H. W. (1968) J. Biol. Chem. 243, 6482-6494. Roberts, R. C., Makey, D. G., and Seal, U. S. (1966) J. Biol. Chem. Ronzio, R. A. (1973a) Biochim. Biophys. Acta 313, 286-295. Ronzio, R. A. (1973b) Arch. Biochem. Biophys. 159, 777-784. Ronzio, R. A., and Mohrlok, S. H. (1977) Arch. Biochem. Biophys. 181, 128-136. Ronzio, R. A., and Rutter, W. J. (1973) Dev. Biol. 30, 307-320. Rosen, J. M., Wbo, S. L. C., and Comstock, J. P. (1975) Biochemistry 44, 2895-2903. Rutter, W. J. (1967) In Methods in Developmental Biology: eds. Wilt, F. H., and Wessells, N. K. (Thomas Y. Crowell Company, New York). pp. 671-683. Rutter, W. J., Kemp, J. D., Bradshaw, W. S., Clark, W. R., Ronzio, R. A., and Sanders, T. G. (1968) J. Cell. Physiol. 72 Sup. 1' 1-180 Sanders, T. G. (1970) Ph.D. thesis, University of Illinois. 119 Segrest, J. P., and Jackson, R. L. (1972) In Methods in Enzymology, ed. Ginsburg, V. (Academic Press, New York), vol. 28, pp. 54—63. Skutelsky, B., and Farquhar, M. G. (1976) J. Cell Biol. 71, 218-231. Spooner, B. S., Walther, B. T., and Rutter, W. J. (1970) J. Cell Biol. 47, 235-246. Truding, R., Shelanski, M. L., Daniels, M. P., and Morell, P. (1974) J. Biol. Chem. 249, 3973—3982. Vandermeers, A., and cristophe, J. (1968) Biochim. BiOphys. Acta 154, 110-129. Volkl, A., Bieger, W., and Kern, H. F. (1976) Cell Tissue Res. 175, Walther, B. T. (1972) Ph.D. thesis, University of Washington. Walther, B. T., Pictet, R. L., David, J. D., and Rutter, W. J. (1974) J. Biol. Chem. 249, 1953-1964. Wang, C., and Smith, R. L. (1975) Anal. Biochem. 63, 414-417. Watanabe, S., and Keller, P. J. (1974) Biochim. Biophys. Acta 336, 62-69. Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244, 4406-4412. Weiser, M. M. (1972) Science 177, 525-526. Weiser, M. M. (1973) J. Biol. Chem. 248, 2536-2541. Weiser, M. M. (1974) Univ. Mich. Med. Cent. J. 40, 168-175. Wray, V. P., and Perdue, J. F. (1974) J. Biol. Chem. 249, 1189-1197. m L Y .h s R E W N U El T AH MllaL 4| 3 lllllHlHllHl