STUDIES ON THE lNTRACELLULAR MEMBRANES or V _ MAMMAUAN EXOCRINE PANCREASQ Dissertationfor the Degree of Ph. D. MICHieAN STATE UNIVERSITY RAYMOND J. MacDONALD 19:74 A I LIBRARY Michigan State /I W? 7‘ W W0] Will l; W / ll/Illll/l/ll l “H ‘g93 tl@%§oog l 3 ‘l r. .. Ulllvcmw Ihis is to certify that the ‘ V thesis entitled STUDIES ON THE 'INTRACELLULAR MEMBRANES OF MAMMALIRNUEXOCRINE‘PANCREAS presented by Raymond J. MacDonald has been accepted towards fulfillment of the requirements for Ph.D. degree in Department of Biochemistry Major professor OMLX‘QJE?‘ Date mi /éi /47£( 0-7 639 ' a of ‘5 \lumeasws_ ‘ 300K BlNDERY 13w. LIBRARY SINGERS immvunmwnJ ‘8’. a. ,1 tut stunt-.2 ‘ .':~,,&::E51)l ‘_ ' ~ ' L ['37. ' 6:". " w"'.‘:"’.1'\ " “ 5i‘k'9rll t'li‘v‘uw‘ infil- an” stow 'cL u: - whiz. ~. ‘71 w ate-bound gran-altos. c: Moat-tans iu-actmn‘ .na‘luc'dx- «ohm» ‘- Janna intractfcme . I.” '_‘ ' T 1 b -- we: -.‘ u .vc ~ - ".3" swath-Ha, 33L” ”lint dur n: t u ,r . . -.‘- lulu transport, .sm, .. r, ,2]. - . \ -7,,_. :93“..ch .7 g ' ' -‘3.’. 1"“. QUIQ bl. TIL“ t!-.r4‘u-,' \r‘ l- ’4 0‘ :ja- «WM‘A‘S Cf thy ~3urY- «Ara - V-‘ . ' " ‘ "4 hf”. I ’ _' WW: of niamsoms. any.“ :0;"”‘- - ’ r k‘ ' —N " ‘u 1 5: rut pancreas hen pantie-é a; ”$4.02 ., $‘V‘-w>'n.:~:2.at M m _ ' ‘mllu {rm 9.: a we» at? fi-t‘i“ We A km ' ‘ “final gram. m 0:1“;“57‘5 a . 9‘» “$4513“ 3 'fir .5 ‘ ”We mmflg-v ‘Htw *wt‘f'f' 1 s4" V ‘W I?“ 1- ' ' .. -v . ; ‘ ‘-.' : ‘ . x' - ' \M v~~_0~*-vl-‘Q.'-n\..‘.— 7; V ‘.. '-'_‘ a. v—._ y _. — . . v: .— -_ , . , .‘s - . I r v . . , . ’ f ‘ I - — ,n- n .‘ 'f' , < ; - ' $ ’ A I. ' ' Av. v 'n i o, r .-._L>- _. . - .s - - . - .A ,1- ‘_ - ABSTRACT STUDIES ON THE INTRACELLULAR MEMBRANES 0F MAMMALIAN EXOCRINE PANCREAS BY Raymond J. MacDonald The exocrine pancreas is representative of several tissues that temporarily store secretion products within membrane—bound granules. Distinct membrane functions including membrane-membrane interactions are implicit during the processes of secretory protein synthesis, intracellular transport, concentration within storage granules, and exocytosis. The aims of this research were to characterize the intra- cellular membranes of the mammalian exocrine pancreas and to relate their structural and functional properties to the mechanism of secretion. Membranes of microsomes, mitochondria and zymogen granules from adult rat pancreas were purified by sequential extraction of the subcellular organelles with 0.2 M NaHCO3 and 0.25 M NaBr. Enzymatic analyses of the final granule membrane preparation indicated that less than 2% of the protein represented granule contents and mitochondrial membrane. The granule membrane constituted only approximately 0.5% of the total granule protein. \ ’1/ \ C/ 63} Raymond J. MacDonald Purified intracellular membranes were fractionated by poly- acrylamide gel electrophoresis in 1% sodium dodecyl sulfate and stained for protein with Coomassie blue and for carbohydrate by the periodic acid-Schiff procedure. Nine polypeptides ranging in molec- ular weight from 10,000 to 130,000 were consistently enriched during purification of the zymogen granule membrane and have been designated membrane components. The predominant granule membrane polypeptide accounted for approximately 38% of the Coomassie blue staining inten- sity and had an apparent molecular weight of 74,000. This polypeptide was also the major component which stained by the periodic acid-Schiff procedure and which contained a majority of the membrane sialic acid. A membrane component of very low molecular weight stained both by Coomassie blue and periodic acid-Schiff was identified as lipid. The characteristic zymOgen granule membrane profile was also observed for membranes from dog, beef, pig and rabbit zymogen granules. The polypeptide profiles of smooth and rough microsomal mem- branes were similar; their complexity contrasted with the characteris— tic simplicity of the granule membrane profile. The microsomal mem- branes contained approximately 35 discernible Species, only 5 to 10 of these contained carbohydrate. The glycopolypeptide composition of Smooth microsomal membranes resembled granule membranes. The 74,000 molecular weight glycopolypeptide was enriched in smooth microsomal membranes, but not in rough microsomal or mitochondrial membranes. The postmicrosomal supernatant and the granule contents, two major soluble subcellular fractions, contained only minor glycopolypeptide components. These results indicate that pancreatic glycoproteins are Raymond J. MacDonald preferentially associated with membranes, and that the granule mem- brane contains a small number of unique glycopolypeptides. It is postulated that the granule membrane glycopolypeptides are important either for the segregation of polypeptides of the granule membrane or for membrane—membrane interactions, most importantly with the luminal plasmalemma during secretion. Mg2+-dependent adenosine triphosphatase activity has been ob- served to be firmly bound to rat zymogen granule membrane. Kinetic analysis of the triphosphatase activity implies that two enzymes with distinctly different Km's are present. Possible roles of two granule membrane adenosine triphosphatases in the packaging and release of the exocrine cell products are discussed. Since cyclic AMP has been implicated as an intermediate in the secretion stimulus, it was of interest to investigate the possible involvement of cyclic AMP-dependent phosphorylation of the structures immediately involved in the secretion process. Alterations of the granule membrane surface, such as phosphorylation of specific sites by a protein kinase, could alter the rate of granule discharge. A single protein was phosphorylated when granule membrane was incubated with (Y32P)ATP. The activity required a divalent metal cation and was nearly equally active with Ca2+ or Mg2+. The phosphorylation was not stimulated by cyclic nucleotides. The lack of cyclic AMP stimulation may indicate that the association of the protein kinase with the granule membrane is induced by cyclic AMP, facilitating phosphorylation of a specific membrane-bound substrate. STUDIES ON THE INTRACELLULAR mamas OF DIWALI”! BXOCRINE PANCREAS By ’60 Raymond J . MacDonald A DISSERTATION ' . V Submitted to \ Michigan State University . , ‘ in partial fulfillment of the requirements - , . ' for the degree of - 7 “ DOCTOR OF PHILOSOPHY Department of Biochemistry ‘ 1974' “Vi:- \'. ~r'..’- . nnv'l KT‘I ' ..‘.:¢-. ': 31* n4 ”ise' ’HU) jam bytmsuift .. 3,7Je:oua r05nnfnan! c” v” .fr JfTW*I3! ni’fiaul “Lu!" bv‘ [‘m-f i?’j_ef..“' h'u :2. Mai ~ éy ’4'; ‘I its; Beware of the man who works hard to learn something, learns it, - and finds himself no wiser than before, Bokonon tells us. He is full of murderous resentment of people who are ignorant without having come by their ignorance the hard way. K. Vonnegut . I .... . - . u -'FLI".L'... :.a , - . U! -. .. ' ab «1 . . ' .auz..': ‘ -. C: . v "x .. ‘ 7 n l 'J 31.? . . ' ‘ 3;, 53.33131le .~ a . . as 1'. .* rial" 12;:1‘ *0 .1: «I .- ~ . - . ‘ ‘. _' u . ‘ r ‘ ’ ', ‘- n .‘I . . all -\ — , _ -, ‘ ‘3 A ' ACKNOWLEDGMENTS I would like to express my deep and sincere appreciation to Dr. Robert A. Ronzio for his continual help and guidance throughout my graduate studies. His concern for the professional development of this graduate student was particularly encouraging. I would like to thank the members of my doctoral committee, Drs. Charles Sweeley, Clarence Suelter, and especially Steve Aust and Walter Esselman, for their helpful discussions during the course of my research. I would also like to note special satisfaction, perhaps hidden to them, in frequent interactions with Drs. Fritz Rottman, Loren Bieber and John Boezi, in addition to numerous colleagues and members of the laboratory. I would like to express my appreciation to Joe Harlan for his invaluable assistance in the development of the preparative polyacryl- amide slab gel electrophoresis, and to Diana Ersfeld for performing numerous amino acid analyses. I especially thank my uncle, Steve F. Wronski, whose generosity and subtle faith permitted my academic endeavors. Financial assistance from the Department of Biochemistry and the National Institutes of Health is appreciated. iv TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES LIST OF ABBREVIATIONS LITERATURE REVIEW . Secretion Morphology of the Pancreatic Exocrine Cell . Intracellular Transport of Secretory Protein . Properties of Secretory Granules . . . A Model of the Secretory Granule Release Reaction: Exocytosis . Control of Exocrine Pancreas Secretion . Statement of the Problem MATERIALS AND METHODS Materials Electrophoresis Reagents . Reagents for Analytical Procedures Miscellaneous . . . Tissue Sources . Methods Preparation of Homogenates . . Isolation of Subcellular Fractions Zymogen granules Mitochondria . . Microsomes . Modifications for isolation of dog and beef zymogen granules . . . . Page Page Preparation of Membranes from Subcellular Fractions . . . . . . . . . . . . . . . . 23 ZymOgen granule membranes . . . . . . . 23 Mitochondrial and microsomal membranes . . . . . . 24 Analytical Polyacrylamide Gel Electrophoresis . . . . . 24 Preparative Polyacrylamide Slab Gel Electrophoresis . . . 27 Preparative Electroelution of Polyacrylamide Gel Slices . . . . . . . . 29 Polyacrylamide Gel Staining . . . . . . . . . . . 30 Protein staining . . . . . . 30 Carbohydrate staining by the periodic acid- Schiff procedure (PAS) . . . . . . . . . . 31 Determination of Radioactivity in Polyacrylamide Gels . . 32 Scintillation Counting . . . . . 32 In Vitro Radioactive Labeling With Formaldehyde and (3H)NaBH4 . . . . . . . . . . . 32 Enzyme Assays . . . . . . . . . . . . 33 Preparation of (Y3ZP)ATP . . . . . . . . . . . . 36 Analytical Procedures . . . . . . . . . . . . . 36 RESULTS . . . . . . . . . . . . . . . . . . . 38 Isolation and Characterization of Rat Zymogen Granule Membrane . . . . . . . . . . . . . . 38 Isolation . . . . . 38 Further Characterization of the Granule Membrane Components . . . . 50 Comparison of Membrane Polypeptides of Microsomes, Mitochondria and Zymogen Granules . . . . . . 64 Carbohydrate— -Containing Components of Pancreatic Intracellular Membranes . . . . . . . . 68 Interspecies Comparison of Zymogen Granule Membrane Components . . . . . . . . . 74 Isolation of the Major Zymogen Granule Membrane Glycopolypeptide (Component 2) . . . . . . . . . . 80 Calculations on the Surface Distribution of Rat Granule Component 2 . . . . . . . . . . . . 80 Isolation of Dog Component 2 . . . . . . . . . . 82 Purity . . . . . . . . . . . . . . . . 85 vi Zymogen Granule Mg2 -Dependent TriphOSphatase (MgZ+- -ATPase). The Mg 2+-ATPase is Tightly Bound to the Granule Membrane . Partial Characterization of the Granule Mg2+ -ATPase Activity PhOSphorylation of Zymogen Granule Membranes . . Incorporation of 32P04 Into a Single Granule Membrane Polypeptide . Characterization of the Phosphorylated Components . Characterization of the Phosphoryl- Polypeptide Bond Partial Characterization of the Endogeneous Protein Kinase Activity . Phosphorylation of Mitochondrial Membrane DISCUSSION . . . . . . . . . . . . . . . Analysis of Zymogen Granule Membrane Polypeptides . Comparison of Intracellular Membranes . . Zymogen Granule Membrane Mg2*-ATPase Zymogen Granule Membrane Protein Kinase Activity APPENDIX: A PRELIMINARY STUDY OF MEMBRANE FORMATION DURING PANCREATIC DIFFERENTIATION IN THE RAT EMBRYO Abstract . . . . . . . . . . . . Introduction . . . . . . . . . . Methods and Materials . . . . . . . . Results . . . . . . . . . . . . . . . . Estimation of the Rates of Membrane Synthesis at Different Developmental Ages by 3H-Leucine Incorporation . . 3H- Leucine Labeling of Individual Subcellular Fractions During Development--The Appearance of a Distinct Membrane Class . . . . Discussion . LIST OF REFERENCES . . . . . . . . . . Page 88 91 94 108 108 111 113 115 115 120 120 127 129 132 136 136 137 138 139 139 149 153 157 Table 10. 11. 12. 13. 14. 15. LIST OF TABLES Survey of secretory cells which store and secrete cell products via secretory granules . Survey of dissociating polyacrylamide gel electrophoresis systems Amylase and cytochrome c oxidase activities of zymogen granule subfractions Relative abundance and molecular weight of zymogen granule membrane polypeptides Detergent solubilization of zymogen granule membrane protein . . . Calculations of the component 2 content of rat zymogen granule membrane . . . . Purification of zymogen granule nucleoside triphosphatase . . NaBr extraction of the granule membrane Mg2+-ATPase Effects of cations on the granule Mg2+-ATPase Relative rates of hydrolysis of various nucleotides by zymogen granule membrane-bound activities Nucleotide inhibition of (YSZP)ATP hydrolysis Stability of the phosphorylated granule membrane component . . . . . Partial characterization of the granule membrane protein kinase activity . . . Rates of 3H- leucine incorporation into particulate fractions of pancreatic rudiments . . . Distribution of 3H—leucine incorporation into subcellular fractions of embryonic pancreas viii Page 25 40 54 65 81 95 96 97 104 105 114 116 148 152 LIST OF FIGURES Figure Page 1. Electrophoretic analysis of the intermediate fractions obtained during zymogen granule membrane isolation . . . . . . . 42 2. Comparison of mitochondrial and zymogen granule membrane polypeptides separated by electrophoresis . . 45 3. Identification of membrane-bound and adsorbed polypeptides of ZGM-2 . . . . . . . . . . . . 49 4. Estimation of the molecular weights of zymogen granule membrane components by electrophoresis in 1% SDS at pH 7. 4 . . . . . . . . . . 52 5. Subfractionation of zymogen granule membrane band 5 . . 56 6. Labeling of zymogen granule membrane components with (3H)NaBH4 . . . . . . . . . . . . . . 60 7. Analysis of zymogen granule membrane polypeptides by acetic acid-urea polyacrylamide gel electrophoresis . . 63 8. Comparison ofthepolypeptides of zymogen granule membranes, mitochondrial membranes and total microsomal membranes . . . . . . . . . . . . 67 9. Glycoprotein nature of zymogen granule membrane components . . . . . . . . . . 70 10. Comparison of the polypeptides and glycopolypeptides of mitochondrial and rough and smooth microsomal membranes . . . . . . . . . . . . . . . 73 ll. Electrophoretic analysis of the glycopolypeptide distribution in the post-microsomal supernatant and zymogen granule contents . . . . . . . . . 76 12. Comparison of zymogen membrane polypeptides and glycopolypeptides from different mammalian species . . 78 ix Figure 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Page Preparative polyacrylamide gradient slab gel separation of dog zymogen granule membrane components . . . . . . . . . . 84 Purity of isolated dog granule membrane component 2 . . 87 Estimation of the molecular weight of isolated dog granule membrane component 2 by acetic acid-urea polyacrylamide gel electrophoresis . . . . 90 Association of Mg2 -ATPase activity with the zymogen granule membrane . . . . . . . . 93 The effect of pH on the rate of ATP hydrolysis catalyzed by the zymogen granule membrane Mg2+- ATPase . . . . . . . . . . 100 Chromatographic analysis of the products of ATP hydrolysis catalyzed by enzymes associated with the zymogen granule membrane . . . . . . 102 Kinetic data for zymogen granule Mg2 —ATPase . . . . . 107 Distribution of protein stain and 32P04 in electropherograms of phOSphorylated zymogen granule membrane . . . . . . . 110 Distribution of3 2P04 in electropherograms of phosphorylated mitochondrial membrane . . . . . . . 118 Changes in the soluble and particulate protein content of cells of embryonic pancreas during development . . . . . . . . . . . . 141 3H-leucine uptake and incorporation into macro- molecules by embryonic pancreas cultured in Vitro . . . . . . . . . . . . . 144 Rates of 3H-leucine incorporation into particulate and nonparticulate (soluble) cellular fractions . . . 146 Comparison of the calculated rates of particulate protein synthesis . . . . . . . . . . lSl ATPase C2 2 Ca +-Mg2+-ATPase CbG CbR cyclic AMP (Y32P)ATP Mg2+-ATPase PAS SBTI SDS TCA TEMED ZGM-l ZGM-Z ZGM-3 ZGS-3 LIST OF ABBREVIATIONS undefined adenosine triphosPhatase zymogen granule membrane component 2 Ca2+- or Mg2+-stimu1ated adenosine triphosphatase Coomassie blue G Coomassie blue R adenosine 3'5' cyclic monophosphate adenosine triphosphate containing phOSphorus—32 in the terminal phosphate Mg2+-dependent adenosine triphosphatase periodic acid-Schiff soybean trypsin inhibitor sodium dodecyl sulfate trichloroacetic acid N,N,N'N'-tetramethylethylenediamine zymogen granule membrane fraction 1 zymogen granule membrane fraction 2 zymogen granule membrane fraction 3 NaBr extracted fraction from ZGM-Z xi LITERATURE REVIEW Secretion Secretion of specific cell products is requisite for the maintenance of multitissue organisms and involves Such diverse phenomena as endocrine mediated homeostasis and exocrine directed digestion. The most evident mechanism of secretion involves concen— tration and storage of the cell products in electron opaque secretory granules within the cytoplasm. In these cases the release reaction of the packaged material appears to be a general, basic phenomenon related to endo- and exocytosis. This section is limited to a review of the processes involved in the formation of secretory granules and the extracellular release of their contents. Table 1 presents a brief survey of secretory cells which con— centrate and store cell products in membrane-bound cytoplasmic granules prior to release. The table encompasses endocrine cell secretion of polypeptide hormones and monoamines, and characteristic exocrine cells which secrete hydrolytic enzymes. Except for chick oviduct, the secretory granules of the sources indicated have been isolated and their contents have been verified as the tissue secretory material by enzymatic and chemical analyses. In oviduct, egg white proteins have been localized within the cytoplasmic granules by Table 1. Survey of secretory cells which store and secrete cell products via secretory granules. TIZEII-type Secretion products References Pancreas u-cells glucagon Mathews (1970) B-cells insulin Howell et a1. (1969) Adenohypophysis (anterior pituitary) somatotroph growth hormone Hymer and McShan (1963) mammatroph prolactin ' thyrotroph thyroid stimulating hormone " gonatotroph follicle stimulating hormone " and luteinizing hormone " corticotroph adrenocorticotrophic hormone " Neurohypophysis (posterior pituitary) neurosecretory neurons oxytocin and vasopressin Poisner and Douglas (1968) Adrenal medulla chromaffin cells Platelets Mast cells Adrenergic neurons Cholinergic neurons Parotid acinar cell Pancreas acinar cell Chick oviduct tubular gland cells Polymorphonuclear leucocytes epinephrine and norepinephrine 5-hydroxytryptamine histamine norepinephrine acetylcholine hydrolytic enzymes, eg., amylase, DNase hydrolytic enzymes and proenzymes, eg., amylase, chymotrypsinogen egg white proteins, eg., ovalbumin, lysozyme lysosomal hydrolases, eg., B-glucuronidase, RNase Poisner and Trifaro (1967) Holmsen et a1. (1969) Stormorken (1969) Hokfelt (1968) Whittaker (1965) Schramm and Danon (1961) Amersterdam et al. (1971) Greene et a1. (1963) Hokin (1955) Palmiter (1972) Woodin and Wieneke (1970) fluorescent antibody techniques (Palmiter, 1972). Fusion of the granule membrane with the cell surface membrane to facilitate release has been observed for all cell types listed (Poste and Allison, 1974). In addition, for all those secretory systems adequately investigated, the release of granule contents after exposure to a secretion stimulus has been shown to require active energy metabolism and Ca2+, to be inhibited by excess Ca2+ or Mg2+, and to involve the displacement of intracellular membrane-bound Ca2+. The uniform occurrence of cytoplasmic storage granules, the demonstration of storage granule exocytosis, and the striking simi— larity of the biochemical requirements for secretion, strongly indicate a common mechanism. The sequence of steps outlined below and expanded in the sections which follow appear fundamental to this process. 1) Protein secretory products are synthesized on ribosomes attached to the endoplasmic reticulum, then sequestered inside the reticulum cisternae. 2) The products are transported to and concentrated into secretory granules, each bounded by a smooth-surfaced unit membrane. 3) Movement of the granule displaces it from the site of forma- tion near the Golgi complex to the cell surface. 4) Extrusion of the granule contents by exocytosis (reverse pinocytosis) is initiated by fusion of the granule membrane with the surface membrane. The exocrine pancreas, representative of tissues utilizing this secretion mechanism, will be discussed to illustrate further details. Morphology of the Pancreatic Exocrine Cell The appearance of a pancreatic acinar cell is that of a cellu- lar factory structured for protein synthesis and secretion. The sub- cellular organelles are strictly polarized within the cytoplasm. Occupying approximately 60% of the cell volume, the rough endoplasmic reticulum is crowded into the basal pole of the cell. Its flattened cisternae, oriented parallel to the circumference of the nucleus, enclose a functionally and physically distinct cellular compartment. Ribosomes, although frequently seen in the cytoplasm, are largely associated with reticulum-bound polysomes arranged in whorls, rosettes or linear arrays. The nucleus occupies a region of the cell extending from the basal section to the middle of the cell, surrounded on all but one face by the rough endoplasmic reticulum. The Golgi complex generally occupies the region free of rough endoplasmic reticulum adjacent to the nucleus, with the mature face of the Golgi apparatus extending toward the lumen. Numerous small smooth-surfaced vesicles are associated with the periphery of the Golgi stacks. Clustered in the apical portion of the cell toward the luminal plasma membrane are numerous electron dense zymogen granules containing the hydrolytic digestive enzymes and proenzymes destined for secretion into the duct system of the gland. Centrally located in the Golgi complex are a small number of immature zymogen granules characterized by a scalloped profile and internal material of variable density. Mitochondria are prevalent in the middle and apical cytoplasm. Secondary lysozomes are occasionally observed. The acinar lumen is formed by laterally joining each exocrine cell to its neighbor. Contact is maintained between surface membranes by tight junctional complexes which extend completely around each cell, effectively segregating the duct lumen from the remainder of the tissue. The structure implies differentiation of the luminal plasma membrane from the rest of the surface membrane and precludes mixing of membrane components by translocation along the plane of the membrane. The morphology of the pancreatic acinar cell applies, with minor modifications, to the other secretory cell types listed earlier (Table l). Microscopy studies by Jamieson and Palade (1971) noted loss of zymogen granules and concomitant increase in the amount of luminal plasma membrane after stimulation of secretion by carbamylcholine. Morphological evidence of the secretion mechanism is limited to the static electron microscopic observations of structures which appear to be granule ghosts, judged by their distinct spherical shape, fused with the luminal plasmalemma. In such instances the granule contents are partially or completely lost to the duct space. Greene et a1. (1963) and Keller and Cohen (1961) have demon- strated that the protein complement of isolated zymogen granules is identical to that of pancreatic juice collected from the gland. Since this evidence is representative of the average granule popula- tion, it remains to be shown whether each granule contains all, one, or a limited number of the digestive enzymes. Most other secretory tissues delegate the synthesis and export of each cell product to an individual cell type (for instance, the endocrine pancreas and the adenohypophysis, Table l). Intracellular Transport of Secretory Protein Extensive evidence has accumulated in support of the proposal that exportable proteins are synthesized on polysomes attached to the endoplasmic reticulum, while the synthesis of cellular non-exportable proteins occurs on polysomes free in the cytoplasm (Hicks et al., 1969; Redman, 1969; Takagi and Ogata, 1971). During or after syn- thesis, the secretory proteins are transported into the cisternal space of the rough endoplasmic reticulum. Isolated rough microsomes with attached ribosomes active in amino acid incorporation are capable of discharging their nascent polypeptide chains vectorally into the microsomal Space after normal chain termination or premature termination by addition of puromycin (Redman and Sabatini, 1966; Redman, 1967). The direct transfer across the membrane and the resistance of the nascent polypeptides of membrane-bound ribosomes to proteolytic attack have led to a simple model to describe the mechanism involved (Sabatini and Globel, 1970). Sabatini and Blobel have proposed that the nascent polypeptide grows within a channel in 'the large subunit which is attached to the membrane; upon continued elongation the polypeptide penetrates the endoplasmic reticulum through a membrane pore. The original electron microscopic radioautography studies of guinea pig exocrine pancreas by Caro and Palade (1964) indicated transfer of exportable protein through the endoplasmic reticulum cisternae. transient association of the proteins with the peripheral A} elements of the Golgi complex, and accumulation of the proteins within zymogen granules. Subsequent refinements were developed by Jamieson and Palade (1967a,b) for a more precise analysis of the intracellular movement of secretory proteins. Pancreatic slices were labeled with (3H) leucine for three minutes, then the incorporation of radio- activity was stopped by the addition of a large excess of unlabeled leucine, and the incubation continued. The intracellular transport of newly synthesized exportable protein was monitored by electron microscopic radioautography and isolation of the subcellular elements involved. The increased resolution resulting from the short, dis- crete pulse indicated that movement of labeled protein occurred through the cisternal space of the rough endoplasmic reticulum to the transitional elements of the endoplasmic reticulum. The radio- activity accumulated in these transitional elements, which are characterized by being partly covered with ribosomes, partly free and positioned adjacent to the Golgi complex, at about 10 minutes post-pulse. Shortly thereafter the label was associated with small smooth—surfaced vesicles located at the periphery of the Golgi cisternae. At about 20 minutes post-pulse, much of the radioactivity was associated with the condensing vacuoles; the number of radio- autography grains associated with the vacuoles increased as protein accumulated within, and the vacuoles became more electron opaque. Continuation of this process led to most of the label in mature zymogen granules. After 60 to 80 minutes radioactivity can be ob- served in the duct lumen as a result of granule movement to and fusion with the luminal plasmalemma. In an elegant series of radioautography experiments Jamieson and Palade (1968b, 1971) delineated the energy requirements for intra— cellular transport of newly synthesized secretory proteins. The studies were made possible by first demonstrating that cycloheximide, at a concentration which blocks protein synthesis by more than 95%, only minimally affects transport (Jamieson and Palade, 1968a). There- fore, inhibition of transport by repressing ATP synthesis cannot be re- lated to a requirement of continued protein synthesis. In the subsequent experiments cycloheximide was present to maintain a constant level of transport, and antimycin A and sodium fluoride were added to block the synthesis of ATP by oxidative phos- phorylation and glycolysis. The tissue ATP level dropped to 50% within 8 minutes after the addition of the metabolic inhibitors, and was at 5% of the initial level after 60 minutes. Several definitive transport steps were analyzed by pulse labeling, followed by a chase incubation to allow the label to accumulate in the compartment pre- ceding the transfer step under consideration. Cycloheximide, antimycin A and sodium fluoride were then added and the transfer of radio- 'activity to the next compartment relative to a control incubation was assessed by radioautography. Briefly summarized, the results were as follows. The migration of secretory proteins from their site of synthesis to the transitional .elements of the endoplasmic reticulum was not affected by the meta- bolic inhibitors, while the transfer of radioactive proteins through the peripheral elements of the Golgi apparatus to condensing vacuoles was inhibited. The next step, the conversion of condensing vacuoles to mature zymogen granules, was not altered at 20 minutes after ex— posure to the drugs, but was inhibited 20-40% after 60 minutes exposure. The extracellular release process, assayed by the carba— mylcholine induced secretion of radioactively labeled protein, was quickly and completely inhibited by the presence of antimycin A and sodium fluoride. Therefore, of the three prominent discrete steps in the trans- port process (transfer of secretory material from transitional elements to condensing vacuoles, conversion of the vacuoles to mature zymogen granules, and exocytosis of the granule contents), the first and last strictly require an energy source, presumably ATP. The energy requirements of condensing vacuole maturation are equivocal. The low level of inhibition might be due to affects not primarily associated with the granule, and thus only become apparent after an extended period. Alternatively, the energy requirement may have a high affinity for ATP or access to a select intracellular ATP pool (Jamieson and Palade, 1971). Properties of Secretory Granules Secretory granules display surprising size uniformity when grouped into two tissue classes (Mathews, 1970; Poste and Allison, 1974). The average diameter for most of the hormone containing granules of the tissues listed in Table l is 0.15 u, while granules of exocrine tissues (acinar cells of the parotid and pancreas, oviduct and leukocytes) are approximately 1 u in diameter. Little is known of the surface properties of secretory granules. Consistent with the anionic nature of many membrane 10 components, the granules of the adrenal medulla (Banks, 1966; Mathews, et al., 1972) and the neurohypophysis (Poisner and Douglas, 1968) possess a net negative charge. A significant fraction of the charge has been attributed to external sialic acid residues (Mathews, et a1. 1972). The presence of Ca2+ neutralizes the surface charge and causes the granules to aggregate. Numerous secretory granules have been observed to be osmoti- cally insensitive. The basis of this phenomenon can be readily demonstrated for adrenal medulla chromaffin granules, which contain norepinephrine and ATP. If Ca2+ is added to a solution of norepine— phrine and ATP to yield final molar ratios of 3:1:0.l (norepine- phrine: ATP: Caz+), an insoluble complex is formed (Pletscher et al., 1970). Pancreatic zymogen granules also have unusual stability properties. The granules are stable in distilled water, 0.3 M sucrose and 0.3 M urea (Hokin, 1955; Jamieson and Palade, 1971; Burwen and Rothman, 1972). Addition of increasing amounts of cations induces lysis. These results imply that secretory vesicles function as a sink for the accumulation of secretory products, and once formed to not require a continuous energy supply to maintain their integrity. Secretory granules contain few enzymatic activities. Chromaf- fin granules from adrenal medulla and their isolated membranes possess dopamine B-hydroxylase activity. Since this activity is responsible for the intragranule conversion of dopamine to norepine— phrine, its distribution would be expected to be limited to granules involved in catecholamine secretion. CaZI- or Mg2+-dependent ATPase has been found associated with secretory vesicles from the adrenal 11 medulla (Trifaro and Warner, 1972; Trifaro and Dworkind, 1971), neurohypophysis (Poisner and Douglas, 1968), polymorphonuclear leukocytes (Wooden and Wieneke, 1963, 1970), platelets (Stormorken, 1969), synapses (Kadoto et al., 1967), and guinea pig exocrine pancreas Gfleldolesi et al., 1971c). The ATPase activity appears to be membrane-bound and has been envisaged as the energy requiring step in granule exocytosis (see below). A second ATP-requiring enzymatic activity, that of a protein kinase, is associated with secretory granules of the adenohypophysis (Trifaro and Warner, 1972) and adrenal medulla (LaBrie et al., 1971). A Model of the Secretory Granule Release Reaction: Exocytosis A series of obligatory steps may be anticipated as requisites to membrane fusion. The two membranes must become directly opposed, probably to within 10 A (Mathews, 1970; Poste and Allison, 1969). Membrane destabilization must occur at the nearest adjacent sites. Formation of a new membrane continuum at the fusion point must favor inclusion of the granule membrane, and quick restabilization of the newly formed membrane must follow. One of the original models incor- porating these main features was a proposal by Woodin (1968) for the extrusion of secretory granules by polymorphonuclear leukocytes. The model has been further refined for secretory granule release (Poisner and Trifaro, 1967; Mathews, 1970) and general membrane fusion (Poste and Allison, 1969, 1974). The proposal takes into account the negative surface charge of both the secretory granule and the interior of the plasma membrane. 12 The granule may overcome the potential energy barrier imposed by the juxtaposition of two similar charges through its kinetic energy derived from Brownian motion. Once the energy barrier has been sur- mounted, the rise of intracellular Ca2+ associated with the secretion stimulus facilitates adhesion of the membranes through salt bridges (-coo'...Ca2*...'ooc-, or ossibly -OP0 '...Ca2*...'o 90-). The P 3 3 2+-Mg2+—ATPase is postulated to promote granule membrane associated Ca membrane destabilization at the fusion site by hydrolysis of membrane bound ATP. Released orthophosphate is presumed to remove membrane Ca2+ by chelation. Fusion occurs at the sites and stabilization recurs by reassociation of Ca2+ and ATP. The release reaction is strictly limited by the concentration of Ca2+ available, since fusion is prevented in both the absence of Ca2+ and in the presence of excess Ca2+. These observations are readily incorporated in the model: The rise in intracellular 032+ during stimulation of secretion is required for adhesion of the secretory granule to the plasma membrane. Saturating levels of Ca2+, 2+ however, do not permit localized loss of Ca generated by the granule ATPase, and loss of membrane stability is prevented. Control of Exocrine Pancreas Secretion Hormonal control of the physiological state and function of many metazoan cells is mediated by intracellular concentrations of adenosine 3'5' cyclic monophosphate (cyclic AMP). For example, the catecholamine stimulation of cardiac muscle glycogenolysis and con- tracting force (Rasmussen et al., 1972), the hormonal-induced increase of lipolysis in adipose tissue (Butcher et al., 1968), and the 13 glucagon stimulation of liver gluconeogenesis (Krebs, 1972) are mediated by levels of cyclic AMP. Control of cell secretion in many instances is controlled in a similar manner: Glucagon stimulation of insulin release by B-cells of pancreatic islets (Bdolah and Schramm, 1965) and catacholamine induced secretion of amylase by the parotid gland (Schramm and Naim, 1970) have been documented. The functional relationship between extracellular hormones and cyclic AMP, originally described by Sutherland et a1. (1965) as the second messenger hypothesis, may be summarized: An extracellular messenger, generally a hormone, binds to its specific receptor on the cell surface, resulting in the activation of a plasma membrane asso— ciated adenylyl cyclase. This activation generates an increase in the intracellular concentration of cyclic AMP, the product of the adenylyl cyclase reaction. Thus the information originally con- tained in the hormone is translated intracellularly through changes in cyclic AMP levels. The cyclic AMP concentration is determined by the balance between synthesis from ATP by adenylyl cyclase and hydrolysis to S'AMP by specific phosphodiesterases. As yet, only the cyclase activity has been shown responsive to extracellular signals. Acting as a second messenger, cyclic AMP activates the target cell to perform its specific function. More recently, Kuo and Greengard (1969) and Exton et al. (1971) have postulated that most, if not all, of the affects of increased cyclic AMP levels are mediated by activation of cellular protein kinases. Sutherland and Robison (1966) developed a set of criteria as a guide for assessing the involvement of cyclic AMP as the Specific 14 messenger in cell activation by hormones. In summary these criteria are 1) an increase in intracellular cyclic AMP in response to hormone; 2) the increase accompanies or precedes the physiologic effect; 3) the hormone is able to activate adenylyl cyclase in cell homo- genates; 4) exogeneous cyclic AMP or an analog has the ability to mimic the hormone effects; and 5) phosphodiesterase inhibitors such as the methyl xanthines are able to mimic or potentiate the hormone effects. Analysis of agents which cause amylase secretion by mouse pancreas incubated in_!i££g implicates cyclic AMP participation in control of exocrine secretion (Kulka and Sternlicht, 1968). Both pancreozymin, which controls exocrine secretion in_§itu, and carbamyl- choline stimulated amylase release 3-fold over control values. Cyclic AMP, its monobutryl and dibutryl analogs, and the phosphodiesterase inhibitor theophylline also stimulated amylase secretion. Acting as a cyclic AMP antagonist, 3'AMP inhibited the stimulation induced by pancreozymin, carbamylcholine and cyclic AMP. Increasing concentra- tions of cyclic AMP reversed 3'AMP directed inhibition, indicating that the effect of this antagonist is readily reversible and rela- tively specific. Further investigations with rat pancreas indicated that the control mechanism is more complex. Baudin et al. (1971) have shown that while carbamylcholine and pancreozymin stimulate glycolysis, (14C) glucose incorporation into protein, oxygen uptake, and 32P04 incorporation into phospholipid in addition to enzyme secretion, cyclic AMP or its analogs only increase the rate of secretion without IS affecting the other parameters. Although carbamyl choline and pan- creozymin elicit a 6-fold increase in the rate of secretion, maximal cyclic AMP stimulation is limited to approximately 2-fold. Obviously cyclic AMP has a limited ability to mimic hormone action in this tissue. While methyl xanthines have little or no effect on secretion in rat exocrine pancreas (Heisler et al., 1972; Baudin, et al., 1971), they do potentiate pancreozymin—induced secretion (Baudin et al., 1971). Furthermore, the increase in the rate of secretion by optimal concen- trations of carbamylcholine plus dibutryl cyclic AMP is greater than the sum of the two alone (Heisler et al., 1972). Consequently, the increased intracellular level of cyclic AMP induced either indirectly by phOSphodiesterase inhibitors or directly by exogeneous dibutryl cyclic AMP appears to stimulate a step in secretion not maximally affected by hormones, and therefore one which appears to become rate limiting only under certain conditions. The rat exocrine pancreas requires a sufficient supply of Ca2+ to maintain secretion (Heisler et al., 1972). The presence of EDTA or repeated carbamylcholine stimulation in Ca2+-free medium inhibited further secretion. Addition of Mn2+, which competes for Ca2+ uptake, also inhibited carbamylcholine stimulation. Dibutryl cyclic AMP and theoplylline induction, however, was not strictly dependent upon uptake of extracellular Caz+, as if cyclic AMP acts to release intracellular sequestered Ca2+. With the information accumu- lated to date, it is difficult to distinguish between Ca2+ functioning as an essential cofactor in secretion, or as an intracellular second messenger responsive to hormone stimulation, as suggested by Rasmussen et al. (1972). 16 The presence of cyclic AMP in the pancreas (Johnson et al., 1970), its ability to stimulate enzyme release (Kulka and Sternlicht, 1968; Baudin et al., 1971; Ridderstap and Bonting, 1969), the ability of an analog, 3'-AMP, to inhibit induced secretion (Kulka and Stern- licht, 1968), and the induction (Heisler et al., 1972; Kulka and Sternlicht, 1968) or potentiation (Baudin et al., 1971) of release by theophylline are strong arguments for the participation of cyclic AMP in exocrine secretion. The most direct evidence, a measurable in- crease in intracellular cyclic AMP prior to or concomitant with hormone stimulated enzyme secretion, is negative (Heisler et al., 1972). Consequently, the specific function of cyclic AMP and whether it is a primary intracellular messenger are not yet fully resolved. The direct action of cyclic AMP in many tissues which respond to hormones is the activation of protein kinases (for reviews, Hittelman and Butcher, 1971; Krebs, 1972). Lambert et a1. (1973) have observed that pancreozymin and caerulin stimulation of exocrine pancreas induced a significant increase in protein phosphorylation. The maximal increase occurred with the phosphorylation of zymogen granule membrane. A similar observation is the association of protein kinase activities with secretory granule membranes of anterior pitui- tary (LaBrie et al., 1971) and adrenal medulla (Trifaro and Warner, 1972). These observations link the common intracellular mediator of hormone action to the control of a secretory granule membrane-bound enzyme activity and afford an intriguing mechanism for controlling the rate of secretion. 17 Statement of the Problem The central role of pancreatic intracellular membranes in the biosynthesis, transport, storage and release of the hydrolytic enzymes and proenzymes is evident. Immediately after synthesis the proteins destined for export are segregated into protective membrane-bound compartments, and remain sequestered until extracellular release. However, the membranes function more than as mere containers. Mem— brane interactions are implicit in intracellular transport and exocy- tosis. The rate of secretion is almost certainly controlled at the point of fusion of granule membrane with the luminal plasma membrane. In an attempt to gain further insight into the attributes of membranes which function in secretion, an intensive study of the pancreatic exo- crine secretory granule (zymogen granule) membrane was undertaken. The rationale was to analyze the structural properties and enzymatic functions unique to the granule membrane in relation to the other intracellular membranes. Characteristics which would be expected to play a role in the secretion process were emphasized. Recently polyacrylamide gel electrophoresis in SDS has been routinely employed in the analysis of membrane polypeptides. The principal advantages are that membranes can be completely dissolved in SDS buffers, and that membrane polypeptides are dissociated from lipids and can be resolved into distinct minimum molecular weight classes. The hydrophobic interaction of SDS with proteins causes denaturation and forces the polypeptides into a singular conforma- tion (Reynolds and Tanford, 1970b). The uniform conformation and charge density of protein-SDS complexes accounts for the observed 18 regular relationship between protein molecular weight and mobility during electrophoresis in SDS-polyacrylamide gels. To fully realize the potential of this analytical tool, the electrophoresis procedure must resolve microgram quantities of membrane protein into discrete classes reproducibly and without artifacts. In the study reported herein electrophoretic analysis of the number, distribution and nature of the granule membrane components was coupled with enzymatic analysis of granule membrane function. MATERIALS AND METHODS Materials Electrophoresis Reagents Proteins for molecular weight standards were obtained from several commercial sources. acrylamide, technical grade, recrystallized from chloroform before use (Loening, 1967) N,N'-methylenebisacry1amide, recrystallized from acetone before use (Loening, 1967) sodium dodecyl sulfate, sequanal grade ammonium persulfate N,N,N',N'-tetramethylethylenediamine (TEMED) Coomassie brilliant blue R Coomassie blue G (xylene brilliant cyanin G) pyronin B basic fuchsin dimethyldichlorosilane urea recrystallized from ethanol before use 19 Eastman Organic Chemicals, Rochester, N.Y. Canalco, Rockville, Ma. Pierce Chemical Co., Rockford, Ill. Canalco Bio-Rad Laboratories, Richmond, Ca. Sigma Chemical Co., St. Louis, Mo. K and K Laboratories, Plainview, N.Y. Harleco, Philadelphia, Pa. Harleco Sigma Chemical Co. Mallinckrodt, St. Louis, Mo. 20 Reagents for Analytical Procedures cytochrome c, Type V soluble starch reagent (for amylase assay) bovine serum albumin N-acetylneuraminic acid nucleotides prnitrophenylphosphate Miscellaneous (3H)sodium borohydride, 120 mCi/mmole (32P)orthophosphate, carrier free soybean trypsin inhibitor Triton X—100 Tissue Sources Sigma Chemical Co. Nutritional Biochemicals Corp., Cleveland, Ohio Sigma Chemical Co. Sigma Chemical Co. P-L Biochemicals, Milwaukee, Wis. Sigma Chemical Co. New England Nuclear, Boston, Mass. New England Nuclear Sigma Chemical Co. Rohm and Haas, Philadelphia, Pa. Sprague-Dawley rats were generally obtained from Spartan Research Animals, Haslett. Several hundred rats were the considerate gifts of Dr. Robert Cook, Department of Dairy Science, Michigan State University, and Dr. Jack Gorski, Department of Physiology, University of Illinois, Urbana. Guinea pigs and rabbits were purchased through the Center of Laboratory Animal Research, Michigan State University. Fresh dog pancreases were obtained through the courtesy of Drs. M. D. Bailee, C. C. Chou, and J. Scott, of the Department of Physiology, Michigan State University. Fresh pig and beef pancreases were pur- chased from local slaughter houses. 21 Methods Preparation of Homogenates All operations were performed at 4°. Excised pancreas was trimmed free of fat, weighed and thoroughly minced with scissors. Large amounts of dog, pig, beef and sheep tissue were frequently pro- cessed with a Harvard press (Harvard Apparatus Co., Dover, Mass.) rather than mincing. Ten volumes of homogenization medium containing 0.3 M sucrose and 0.05 to 0.25 mg/ml of soybean trypsin inhibitor (SBTI) were added per gram of minced pancreas and the fatty tissue floating to the surface was removed. Cell disruption was performed by four up and down strokes of a glass-Teflon Potter—Elvehjem homo- genizer (clearance of 0.007 inch, Kontes Glass Co.) driven at 620 rpm. Tissue debris was removed by filtering the homogenate through two layers of cheesecloth. Isolation of Subcellular Fractions (Modifications of the Procedure of Jamieson and Palade, 1967a) Zymggen granules. The filtered homogenate was centrifuged at 500xg (1600 rpm, HS-4 rotor, I. Sorval Co.) for 10 minutes to remove debris, unlysed cells and nuclei; large pieces of plasma membrane may also collect in this fraction. A zymogen granule fraction was collected by centrifuging the 500xg supernatant in conical centrifuge tubes at l600xg (2800 rpm, HS~4 rotor) for 25 minutes. The loosely packed brown layer containing mostly mitochondria overlaying the white zymogen granule pellet was removed by gently agitating and rinsing with homogenization medium using a Pasteur pipet. The zymogen granule 22 pellet was washed once by gently resuspending in homogenization medium and centrifugation at l600xg for 25 minutes. Mitochondria. The supernatant and mitochondrial layer above the zymogen granule pellet from the initial l600xg centrifugation were combined and centrifuged at 8700 xg (8500 rpm, SS-34 rotor, Sorvall) for 15 minutes. The pellet contained a small zymogen granule button overlayed with mitochondria. The mitochondrial layer was resuspended in homogenization medium without disturbing the granules and recentrifuged. Resuspension and centrifugation were re- peated approximately three times until zymogen granules were not ob- served in the pellet. Microsomes. The 8700xg supernatant was centrifuged at 93,000xg (40,000 rpm, type 40 rotor, Beckman Instruments) for one hour to pellet total microsomes. The material which did not sediment was designated as postmicrosomal supernatant. Smooth and rough microsomes were separated by sucrose density gradient centrifugation of the total microsomal pellet according to Ronzio (1973a). Smooth and rough microsomal fractions from rat (Ronzio, 1973b) and guinea pig (Jamieson and Palade, 1967a; Meldolesi et al., 1971a,b,c) pancreas have been well characterized. Modifications for isolation of dog and beef zymogen granules. Zymogen granules from dog pancreas were best isolated in two steps. A large fraction of the granules were sedimented by centrifuging the 500xg supernatant for 15 minutes at 1200xg (2500 rpm, HS-4 rotor). 23 A second fraction was then obtained by centrifuging the supernatant at l600xg (2800 rpm) for 30 minutes. After removing the mitochondrial layer, both pellets were resuspended and centrifuged. Beef pancreas zymogen granules were prepared by a procedure incorporating similar modifications as described in detail by Greene et al. (1963). Preparation of Membranes From Subcellular Fractions gymogen granule membranes. Washed zymogen granule pellets from 3 to 4 gm of rat pancreas were resuspended in one ml of 0.17 M NaCl containing 0.67 mg/ml SBTI. The protein concentration was generally 12-16 mg/ml. The granules were lysed by adding 3 ml of 0.2 M NaHCOs, pH 8.2, to each ml of suspension. Lysis generally required 1—2 hours at 4° and was monitored as clarification of the solution to an absorbance of less than 0.5 at 660 nm. Membranes were collected by centrifugation for 1 hour at 192,000xg have been desig— nated ZGM-l. ZGM-l obtained from 3 or 4 gm of pancreas was resuspended in one ml of l M sucrose, requiring approximately 10 strokes of a Dounce glass homogenizer (tight fitting pestle B, Kontes Glass Co.). Generally 5 ml of the membrane suspension was placed in an SW 41 cellulose nitrate tube, overlayed with 0.3 M sucrose, and centrifuged at 192,000xg for 1 hour. Two membrane fractions resulted. Mito- chondria, marked by high cytochrome c oxidase activity, accounted for the majority of the pellet. Zymogen granule membrane, banding at the 0.3-1.0 M sucrose interface (Meldolesi et al., 1971a), was desig- nated ZGM-2. 24 Care was taken to remove all the membrane felt at the inter- face in less than 1.5 ml of sucrose solution. The suspension was diluted with 0.25 M NaBr in an SW 41 tube, and sonicated at maximum setting for 10 seconds a Biosonik sonic oscillator equipped with a microprobe (Bronwill Scientific). The membrane which sedimented at 192,000xg for 1 hour was designated ZGM-3. Mitochondrial and microsomal membranes. Mitochondrial and microsomal pellets were resuspended in 0.2 M NaHCO pH 8.2, with a 3’ Dounce glass homogenizer. The membrane components which sedimented at 192,000xg for 1 hour were resuspended in 0.25 M NaBr, sonicated for 10 seconds (microprobe, maximum setting), and collected by centrifugation at 192,000xg. Membranes isolated in this manner could be directly compared with purified zymogen granule membrane. Analytical Polyacrylamide Gel Electrophoresis The gel electr0phoresis syste ms surveyed for effective separa- tion of membrane polypeptides are summarized in Table 2. Electro- phoresis in 0.1% SDS according to Kiehn and Holland (1970) or Hoober (1970) without modification failed to resolve the major portion of microsomal membrane polypeptides in the molecular weight range below 30,000. The acetic acid-urea system originally described by Takayama et a1. (1966) and modified by Zahler et a1. (1970) and Ray and Marinetti (1971) did not resolve the entire membrane protein applied, since much of the material remained trapped at the gel surface, pre- sumably as insoluble aggregates. This system was of service, however, since solubilization and electrophoresis without detergent permitted 25 anneal". causal. we cote—o: new cougar—e Ion-a- aen an: 3332...:— Sbg .2: no yuan-3 0503.33»: no 3336! no:- Ou cannon: ace... .3 ecuueuueeoeeu Lulocol veg-0.53 30:636.:— ans-u- Suuoeuoun venue “-03:13...— 2.308 Lou-.930! 13 mo gauge-en ueeZouuu «nos-pal! we eon—Jenn; gun—cue...» 0.5.; 03!.- on»: no 02. van coda—43.0... .3- euavegua uni—unu- sueuouaogi me 3:25! 30qu al 1o- eeueuvleueu enema-ennui".- 030355 03:33- am 30. «queue... in u- vex—no.5 —e.—u0u-l account-n... 003.5- 10 u- vet—e: 133' 0:3qu:- neon-heme- ue :03 3'30»: condenses—ea £3!— ue~=ue~el .5— ue 8336-..- he: 9:6: am :25 m 3.5; a 3:05 w 0.50; a 33505 2.3.3.3 3m 2.0.2.93 :6 .3303...— SA I38. : 10 80 t .3. E: z 8.0 .44. n.— .3300. 2 .38. x 8.0 g : No.0 .44. no .33: m :29 x 3.0 0.0 .a .85.... x 3.0 a: 263: neon-2... Sn .2 5:. .4...“ ma Juana-e...— m I523- : To m.» 38 uzoo. *2 8a 3.0 .3- fibu 30.0 .3. m an .3300- a. -zu... : 10 9.0 t .03 t 2.6 E: unnamed arm .3 a3: 80 E 8 me 8:32. 870 E 5:. 8m 3 8 .3 8:8:- HTN 3 8:. Ba 3 86 :3 7.552.. 370 3 .80 2 3d .3 3o- A>\>\D\>\a £0.03: Sum. 9.0-3-09... :30- ounce-Lena...— el.-w 3.0 .2.- can: :90 .3: 08 3.0 8 =3 8:3:- uou 9.555.— *m Jon news—oeu- $4.. “afiuguaeuulnm “37w- :R: 13.: as 320:. 03:83 88: .1 u. 3.5.3: E 8th 1: uo Iguana-h “$0: 3.03:: 9 SR: 2'»... E arm: 52:: ea- gue—3.3:- a: :5: «Lu-3.3: we. sea 85.: .7 «e Lei—IN 385 .1 an alas-sou. an .08: .188: 88: 33:3. a: 530. : 039-.- E ue 33:8 oval-Zhu- * ueuusn umebuom 33318 32.2.38 It... can goo .nleuohu 3030:3500: gen analogue-:05 2.33063: me has». .N 0.3-u. 26 an independent evaluation of the size and distribution of membrane components. Acid-SDS polyacrylamide gels were prepared and electro- phoresis was performed at pH 2.4 as described by Fairbanks and Avruch (1972). Discontinuous-SDS polyacrylamide gel electrophoresis was performed as described by Laemmli (1970) without modification. The method of most utility was electrophoresis in 0.04 M Tris-acetate containing 1% SDS, essentially as described by Fairbanks et a1. (1970). Membrane pellets were dissolved by sonication in sample solvent (0.01 M Tris-Cl, pH 8.0, containing 1% SDS and 5 mM EDTA). Samples in solutions of high ionic strength were dialyzed against sample solvent overnight. Prior to electrophoresis 2- mercaptoethanol was added to 2% (v/v), and the solution was heated at 60° for 30 minutes or at 100° for 5 minutes. After heating, one volume of sample was mixed with one-third volume of sample solvent containing 20% sucrose and 40 ug/ml pyronin 8. Nine percent acryl- amide gels (acrylamide to bisacrylamide ratio, 36:1), dimensions 0.5 X 11 cm, were polymerized in chromic acid-cleaned glass tubes coated with dimethyldichlorosilane. A flat gel surface required for high resolution was insured by carefully monitoring the polymerization process. Gels were prerun at S volts/cm. Samples were applied in a minimal volume, optimally 7 to 20 ul, to promote sharp protein bands. A potential gradient of 10 volts/cm was employed, and the current did not exceed 6 mamperes per gel. To diminish curvature of polypeptide bands a temperature of 13° was maintained during electrophoresis. A constant length of each electropherogram was obtained by allowing the tracking dye, pyronin B, to migrate 8.8 cm from the origin, requiring I’ 27 an average of 4.5 hours. Mobilities are expressed relative to the tracking dye, which was marked prior to staining with India ink. Preparative Polyacrylamide Slab Gel Electrophoresis The slab gel apparatus employed has been described by Reid and Bielski (1968). Several modifications and detailed procedures have been described by Studier (1973). Preliminary experiments demonstrated that acrylamide gradient slab gels utilizing 0.1 M sodium phosphate buffer, pH 6.65, with 0.1% SDS (Maizel, 1971) gave optimal resolution and separation of individual membrane components. 0.6 cm plexiglass spacers placed between glass plates at the sides and bottom formed the slab gel compartment approximately 11 x 15 x 0.6 cm, which re- quired 110 ml of acrylamide solution. The acrylamide gradient was prepared using a Beckman Density Gradient Former. Two sets of 50 ml syringes containing the following two solutions yielded a linear 5 to 17% acrylamide gradient: Light solution, total volume 75 ml: 3.75 ml concentrated acrylamide-bisacrylamide mixture (40 gm acrylamide plus 1.1 gm bisacrylamide per 100 ml water solution), 7.5 ml lO-fold concentrated stock phosphate buffer (81.0 gm of-NaZHPO4 and 59.3 gm NaH2P04'H20 in a final volume of 1 liter), 0.375 ml of 20% SDS (w/v), 37.5 ul TEMED, and 63.0 ml of water. 0.375 ml of a 10% ammonium persulfate solution (w/v) was added immediately before use. Heavy solution, total volume 75 m1: 37.5 ml of concentrated acrylamide-bisacrylamide mixture, 7.5 ml lO-fold concentrated phosphate buffer, 0.375 ml of 20% SDS, 37.5 ul TEMED, 22.5 ml 28 glycerol, 6.75 ml water and 0.15 ml of 10% ammonium persulfate solu- tion, just prior to use. After pouring the gradient, a plexiglass spacer 0.6 cm thick and 11 cm long was inserted into the form from the top to a depth of about 2 cm, displacing some acrylamide solution. Polymerization became apparent 20 minutes after mixing the two solutions. Removal of the top spacer after polymerization and prior to electrophoresis revealed a trough of the same dimensions 1-2 cm deep. The gradient slab gel was prerun at 35 volts for 1 hour. One to five m1 of sample solution containing up to 6.5 mg of membrane protein in 0.01 M sodium phosphate buffer, pH 6.65 with 1% SDS, 1% 2-mercaptoethanol, 5% sucrose and 5 ug/ml pyronin B were applied after heating at 100° for 5 minutes. Buffer reservoirs con- tained 0.1 M sodium phosphate, pH 6.65, with 0.1% SDS. Overheating during electrophoresis was avoided by maintaining constant current at 70-75 mamperes. Under these conditions the potential difference was approximately 35 volts, and electrophoresis was at room temperature for 18 to 30 hours, depending upon the separation desired. After electrophoresis was complete, the protein band of interest was located by quick staining three 3 mm thick vertical slices removed from the middle, left and right sections of the gel. The slices were marked for identification and stained with Coomassie blue G for 1-2 hours as described below. The observed migration distances of the top and bottom of the band visualized by staining required corrections for changes in the gel length caused by the staining procedure in order to calculate the corresponding positions 29 in the unstained slab gel sections. The regions containing the band were excised for elution of the protein from the polyacrylamide matrix. Preparative Electroelution of Polyacrylamide Gel Slices A special apparatus was constructed for large-scale electro- elution of polyacrylamide gel pieces obtained from slab gels. The apparatus consisted of two buffer reservoirs, each with a two liter capacity, containing platinum wire electrodes. The upper reservoir, supported directly above the lower reservoir by a ring stand, had a 7.5 cm length of 1.9 cm 1.0. glass tubing projecting through the bottom. The lower end of the glass tube was sealed with Saran Wrap and filled with 10 ml of 5% acrylamide solution (with 0.14% bis- acrylamide, 0.05% ammonium persulfate, 0.05% TEMED, 0.1 M sodium phosphate, pH 6.65, and 0.1% SDS) containing the gel pieces to be eluted. After polymerization the Saran Wrap was replaced with nylon screen (73 u mesh; Tobler, Ernst and Traber, Inc.) secured with a rubber band. A 6 inch length of 7/8 inch diameter dialysis tubing containing 40 to 50 ml of electrophoresis buffer was fitted over the bottom of the glass tube containing the gel plug and nylon screen. The dialysis tubing was firmly secured with two lengths of copper wire fastened around protective tygon rings. Air bubbles collecting within the system were removed. Electroelution was generally per- formed at 40 volts for 18 hours, at which time the dialysis bag was replaced and electroelution continued for an additional 18 hours. The eluate was dialyzed for 18 hours against two changes of 2 liters 30 of distilled water containing 3 gm washed Dowex AG-lx and 5 ml of toluene. The dialysate was lyophilized and subsequently stored at -20°. Polyecrylamide Gel Stainipg. Protein staining, Two methods were employed to detect protein bands within polyacrylamide gels. The technique most frequently employed utilized Coomassie blue R (CbR) and required removal of SDS for optimal results. After electrophoresis gels were soaked in 30 ml screw cap culture tubes for a minimum of two days with four changes of 10% TCA. Gels of higher cross-linkage and larger diameters re- quired a longer extraction period. TCA extracted gels were stained overnight in 0.4% CbR in 10% TCA-33% methanol (Johnson et al., 1971), then destained for 8 to 10 hours in 10% TCA-33% methanol with an apparatus designed after the Hoefer diffusion destainer (Hoefer Scientific Instruments, San Francisco). Removal of background stain was then completed in 10% TCA in 30 ml screw cap culture tubes at 37° overnight with shaking. A second protein staining technique was employed for poly- acrylamide slab gels. The procedure for the preparation of quick staining solution from treatment of Coomassie blue R described by Malik and Berrie (1972) was applied to Coomassie blue G (R. Blakesly, personal communication) as follows: 0.2 gm of CbG was dissolved in 100 ml of distilled water, then diluted with 100 ml of 2 N H2804. After vigorous stirring, the insoluble material was removed by fil- tration. The brown filtrate was titrated to dark blue by slow addition of 10 N NaOH (5 to 10 m1). 26.4 gm of TCA was added and 31 the solution was used immediately. Gels were not soaked in 10% TCA, but were rinsed in distilled water to remove excess SDS. To locate protein bands for the electroelution procedure, gel slices were in- cubated for about 1 hour in the stain solution. Excess stain was then removed from the slices by two 10 minute washes with 0.2 N H2804, and the color intensified with several water rinses. This procedure was adequate for locating bands quickly. The length of time exposed to stain was increased to 6 hours to properly stain all detectable bands. Water rinses over a period of several days intensified the stain color. Carbohydrate stainingby the periodic acid-Schiff procedure (PAS). The procedure described by Fairbanks et a1. (1970) for detecting carbohydrate within polyacrylamide gels was employed with minor modifications. Complete removal of $05 by exhaustive soaking in 10% TCA (minimum of 8 changes over 4 days) was imperative to prevent artifactual staining of non-glyc0protein species (Glossman and Neville, 1971). All operations were performed in 30 ml screw cap culture tubes with shaking. The incubation times were extended to 0.5% periodic acid, 2 hours; 0.5% sodium arsenite with 5% acetic acid, 1 hour; 0.1% sodium arsenite with 5% acetic acid, 1/2 hour repeated twice; 5% acetic acid, 1/2 hour; Schiff reagent, overnight in the dark; 0.1% sodium metabisulfite in 0.01 N HCl, several changes of one hour duration until the rinse solution failed to turn pink upon the addition of formaldehyde. PAS stained gels were either scanned at 560 nm with a Gilford linear transport or photographed with Kodak Extapan film through a yellow filter. 32 Determination of Radioactivity in POIyacrylamide Gels Polyacrylamide gels containing tritium were fractionated with a Savant Autogeldivider (Savant Instrument, Inc.) (Maizel, 1966). Gels containing phosphorus-32 were sectioned into 2 mm transverse fractions using a stainless steel support and cutting guide. Each fraction from 3H and 32F containing gels was placed in a scintillation vial, and 0.1 N NaOH containing 1% SDS was added to bring the fraction volume to approximately 1.4 ml. After incubating for a minimum of 16 hours at 37°, the solution was neutralized by adding an equivalent amount of l N HCl. After 10 ml of a Triton X-100 based scintillation fluid (Mostafa et al., 1970) were added to each scintillation vial, the vials were capped, shaken, and monitored for radioactivity in a liquid scintillation spectrometer. Scintillation Counting_ Radioactivity was routinely measured in a Packard Tri-Carb utilizing the Triton X-lOO/toluene scintillation fluid reported by Mostafa et a1. (1970). Tritium was optimally counted at 55% gain with window settings at 50-1000; efficiency was approximately 20%. Phosphorus-32 was monitored at 2% gain with window settings at SO-infinity; appropriate corrections were made for 32F decay during the course of experiments. In Vitro Radioactive Labelin With Formaldehyde and (5H)NaBHA ' Reductive alkylation similar to the method described by Rice and Means (1971) was employed to incorporate radiolabel into standard 33 proteins and zymogen granule membrane components. Preliminary labeling experiments with bovine serum albumin in the presence of SDS demon- strated the feasibility of the modified procedure. All operations were performed in the hood at room temperature. 0.2 mg of bovine serum albumin in 0.1 M sodium borate, pH 10, containing 1% SDS was treated first with 10 ul of 0.04 M formaldehyde, then after one minute with 4 sequential 2 pl additions of (3H)NaBH4 (0.14 M, 120 mCi/mmole). 10 ul of unlabeled sodium borohydride was added to the mixture after two minutes. Unreacted sodium borohydride was hydrolyzed by adding 5 ul of glacial acetic acid. ‘Low molecular weight components of the reaction mixture were removed by chromatography on a Bio-Gel P-4 column equilibrated with 0.1 M triethylammonium bicarbonate, pH 8, containing 0.1% SDS. The void volume fractions were pooled and lyophilized. The specific radioactivity of 3H-labeled bovine serum albumin varied between 7 x 103 and 15 x 103 cpm/ug. More than 95% of the protein-bound radioactivity co-electrophoresed with bovine serum albumin in 1% SDS, 9% acrylamide gels. Labeling of purified zymogen granule membrane was performed by the same procedure. Enzyme Assays Amylase was assayed by a micromodification (Sanders, 1970) of the method of Smith and Rowe (1949), which measures the dis- appearance of starch-iodine chromophore at 540nm. The technique was extremely sensitive; samples with high activity were assayed immediately after dilution with 0.05 M Tris-Cl, pH 7.2, containing 0.02 M NaCl and 0.1 mM CaClz. Units of activity are expressed as the change in absorbance at 540nm per minute. 34 Cytochrome c oxidase was assayed as described by Wharton and Tzagoloff (1967). Units of activity are expressed as umoles of cytochrome c oxidized per minute. Hydrolysis of ATP was assayed by measuring both the release of total orth0phosphate from unlabeled ATP and the release of 32P- orthophosphate from (YSZP)ATP. For both methods 50 ul reaction mix- tures contained 3-5 mM disodium ATP,3-S mM MgCl2 and 50 mM imidazole- Cl, pH 7.1. After incubating 30 minutes at 37°, assays were termi- nated by adding 150 pl of cold 10% TCA and centrifuging briefly at 0°. For the nonradioactive assay (assay A), orthophosphate in the supernatant was measured spectrophotometrically as described by Lindberg and Ernster (1956). To assay the specific hydrolysis of the terminal phosphate of ATP (assay B), the standard Mg2+-ATPase assay was supplemented with approximately 100,000 cpm of (YSZP)ATP. After incubation a 100 pl aliquot of the acid quenched assay mixture was added to a 13 x 100 mm test tube containing 0.4 m1 of 1.5% ammonium molybdate in 0.5 N H $04 at 4°. 0.5 ml of cold isobutanol-benzene 2 (1:1, v/v) was added to each, and the samples were mixed thoroughly by vortexing for precisely 30 seconds. After standing for 20 minutes to equilibrate to room temperature, a 250 pl aliquot of the top layer of isobutanol-benzene was placed in a scintillation vial containing 10 m1 of scintillation fluid to determine the 32P content. When the assay was performed as described, the observed 32P-orthophosphate release was directly proportional to incubation time and enzyme con- centration fer all subcellular fractions assayed. Protein kinase activity was assayed at 23° in a reaction volume of 100 pl contained in a 400 pl polyethylene microfuge tube 35 (Beckman Instruments). The standard mixture, unless otherwise indi- cated, contained 20 mM imidazole-Cl, pH 7.1, 75 uM (YSZP)ATP (7.5 nmoles, approximately 4 x 107 cpm), 1 mM MgC12, and 2 mg/ml of mem- brane protein. The assay was initiated by the addition of (YSZP)ATP, and terminated after five seconds by the addition of 300 pl of 0.3 N perchloric acid containing 5 mM H P04 and 2.5 mM ATP (Avruch and 3 Fairbanks, 1972). After standing 15 minutes at 4°, the labeled membrane was sedimented by centrifugation for 10 minutes at 100,000xg (40,000 rpm, type 50 rotor; Beckman) and the supernatant discarded. The pellet was thoroughly resuspended in another 300 pl of perchloric acid solution by sonication. The capped tubes were taped to the bottom of a large plastic beaker, and covered with ice water to a depth of about 1 cm. The tip of a Biosonik sonic oscillator probe (maximum setting) was brought into contact with the wall of each tube for 10 seconds. The membrane was again collected by centrifugation at 100,000xg. Resuspension in the perchloric acid solution and centrifugation were repeated four times. A final wash with distilled water removed excess perchloric acid. The washed pellets were stored at -80°. The pellets were prepared for electrophoresis by adding to each tube 100 pl of 0.3 M sucrose and 7.5 ul of 20% SDS, mixing thoroughly, then adding 7.5 ul of sodium phosphate, pH 2.4, 3 ul of 2-mercaptoethanol, and 10 ul of 40 ug/ml pyronin B. After a 10 minute incubation at 37° the solutions were layered on 0.6 x 11 cm, 9% acrylamide gels. The gels were prepared and electrophoresis was performed in 1% SDS at pH 2.4 as described elsewhere (Avruch and Fairbanks, 1972; Fairbanks and Avruch, 1972). The distribution of 36 32F in the gels was determined as described earlier. The data have not been corrected for loss of radioactive protein during the perchloric acid washes. Protein kinase activity is defined as the transfer of 32P from (Y32P)ATP to membrane components within the polypeptide region of SDS-electropherograms. Preparation of (yszP)ATP (Y32P)ATP was prepared by both the methods of Penefsky (1967) and Richardson (1971) with comparable results. Characteristics of a typical preparation were specific radioactivity, 4 mCi/mmole; A A260, 0.81; and A280:A260’ 0.20. When analyzed by DEAE column 250‘ chromatography (Whatman microgranular DE 52), 97% of the 32? eluted coincident with carrier ATP; the remaining 3% eluted distinct from added AMP and ADP standards. Analytical Procedures Protein was determined by the method of Lowry et a1. (1951) using bovine serum albumin as standard. Samples containing inter- fering substances such as sucrose and 2-mercaptoethanol were pre- cipitated at 4° by the addition of an equal volume of 10% trichloro- acetic acid. The precipitates were collected by centrifugation for 5 minutes (Microfuge, Beckman), washed by resuspension in 10% tri- chloroacetic acid and recentrifugation at 4°. Particulate protein samples were dissolved in 0.1 N NaOH with 1% SDS prior to analysis. Sialic acid content of membrane samples and polyacrylamide gel fractions was analyzed by the thiobarbituric acid procedure of Warren (1959), using N-acetylneuraminic acid as the standard. The 37 sodium dodecyl sulfate content of protein samples was determined by extraction of the dodecyl sulfate-methylene blue complex into chloroform and measuring the absorbance of the chloroform layer at 655nm (Reynolds and Tanford, 1970a). RESULTS Isolation and Characterization of Rat Zymogen Granule Membrane Isolation The isolation of zymOgen granule membrane first requires the preparation of an enriched and intact zymogen granule fraction. Puri- fication from this subcellular fraction is readily monitored. Con- taminants group into two easily identifiable classes, granule con— tents and mitochondria, which can be monitored by enzymatic analyses and characteristic electrophoretic mobilities of specific polypeptides of each class. The purification scheme is biased to remove each group sequentially. Alkaline lability of zymogen granules was first noted by Hokin (1955)., Meldolesi et al. (1971a) exploited this observation to par- tially purify the granule membrane from guinea pig pancreas. The first two steps of Meldolesi's procedure formed the basis of the isolation technique employed in this study. Isotonicity was maintained to prevent rupture of contaminating mitochondria, and zymogen granules were lysed by the addition of 0.2 M NaHCOS, pH 8.2. Lysis was monitored as a marked decrease of turbidity. After centrifugation, the soluble fraction, representing the granule contents (Hokin, 1955; Meldolesi et al., 1971a), contained 96% of the protein and 98% of the 38 39 amylase activity (Table 3). When examined by polyacrylamide gel electrophoresis in 1% SDS, pH 7.4, nine major polypeptide species were observed (Figure 1A). The major component at 3.1 cm accounted for approximately 50% of the total stain intensity and had a molecular weight of 52,000 when compared to proteins of known size in parallel gels. This agrees well with the reported relative amount (Sanders, 1970) and molecular weight (Sanders and Rutter, 1972) of amylase from adult rat pancreas. Procarboxypeptidase 8, molecular weight 50,000, can be tentatively identified at 3.3 cm immediately adjacent to amylase by similar criteria (Sanders, 1970). Likewise, the peaks at 5.5 cm and 7.2 cm are assumed to be chymotrypsinogen (molecular weight 25,000) and ribonuclease (molecular weight 14,000), reSpectively. Amylase, procarboxypeptidase B and chymotrypsinogen correlate with bands 5, 6 and 16 in electropherograms of zymogen granule membrane fractions discussed below. Table 3 summarizes the purification of zymogen granule membrane away from amylase and cytochrome c oxidase activities. Membranes, 595:1! collected from the lysate by centrifugation accounted for 4% of the total granule protein, all of the mitochondrial cytochrome c oxidase activity, and a significant fraction of the granule contents, monitored as amylase activity. The buoyant density difference between the granule ghosts and mitochondria was exploited to separate the two structures (Meldolesi, 1971a). Mitochondria sedimented through 1 M sucrose while the granule membrane, 595:3: floated to the l M - 0.3 M sucrose interface. The high amylase specific activity of this frac- tion necessitated a final purification step. Brief sonication in 40 UMMfloomm omepfixo o esonzoouxu ohm mosfle> .cflopoum ms\muficsfiaafis am we: oceanEoE Hewupconoouwe pofimfihsm mo xuw>fiuom .momogucoamm ca co>ww mucoeflaomxo mo Hones: onu mo memos on» .mpocuoz ca ponfiaomon we poasmeoe one: mofiufi>fluom ommpwxo o osongoouxo use ommflxa< .mponuoz c“ ponflaomop me poawmoum we: cowuomum seem Ame N.o 0.0 Mme m.o H.o em.o Am-20NU :ofiuomnuxo Hmmz Houmm monounEoz Ame a.e es Ame H.NH N.H e.s fl~-20NV peeeeeem mooscflucoomfip scum mesmHQEQZ ”HS m.oH ems “my 0.0 N.N w.m “H-20NV eoaeuaem ogeenEoe oeeeu Ame o.s flooHV flee Hes hoesv flooHV masseuse canoes“ oceans :wouoam .ma . Am-onv:fiouoam e e \maseefissse . me\maaee . . xes>eeu< Aps>see< ses>soo< xas>see< uwmfloomm Hmuob ofimwoomm Hence cfiouond :ofiuomam omepwxo o oeounoouxu ommaxa< .mcowuomumnsm oaaceum comoexu mo mofiuw>fluom ommewxo o osounoouso one omeaxa< .m maneh 41 Figure l. Electrophoretic analysis of the intermediate fractions obtained during zymogen granule membrane isolation. ElectrOphoresis was performed under standard conditions (9% acrylamide, 1% SDS, pH 7.4) as described in Methods. The following amounts of protein were applied: A, granule contents, 34 pg; B, ZGM-l, 51 pg; C, ZGM-Z, 36 pg; D, ZGM-3, 42 pg. Gels were stained with Coomassie blue R (CbR) and scanned at 550 nm. TD marks the distance of migration of the tracking dye, pyronin B. ABSORBANCE . 550 n m DISTANCE (cm) Figure 1 2 '5b 9 w a” " C 2 4 90 '516 19 I ”e a” _ D - s I e 9 I 10 :3 l7:3 2° 1 l 1 l_ l l o 2 4 6 e 43 0.25 M NaBr, a procedure which did not inactivate amylase, reduced the specific activity of the final membrane preparation, 29M;§, to less than 1% of the original level in zymogen granules. The final yield of granule membrane was approximately 0.5% of the total granule protein. Mitochondrial cytochrome c oxidase activity after the rigors of this purification was 21 milliunits/mg protein. Since the cyto- chrome c oxidase activity of the purified granule membrane was 0.2 milliunits/mg protein (Table 3), mitochondrial contamination was estimated to be less than 1%. Contamination by inactivated granule contents was evaluated by mixing (14C)leucine labeled soluble granule protein with unlabeled intact zymogen granules (MacDonald and Ronzio, 1974). From the known specific radioactivity of the soluble proteins, it was estimated that less than 3% of the membrane protein isolated from the mixture was adsorbed granule contents. The removal of mitochondrial and soluble secretory proteins and the enrichment of membrane-associated polypeptides during granule membrane isolation were monitored by SDS polyacrylamide gel electro- phoresis. Several polypeptides were consistently enriched in the membrane fraction at each stage of purification. The bands repre- senting these polypeptides are designated in Figure 10 as l, 2, 5, 10, 13, IS, 17, 18, 19, and 20. The stain intensity of band 2, the major membrane polypeptide, increased six-fold during purification from the crude membrane fraction. The distributions of mitochondrial membrane and zymogen granule membrane polypeptides are compared in Figure 2. A 44 Figure 2. Comparison of mitochondrial and zymogen granule membrane polypeptides separated by electrophoresis. Polyacrylamide gels, run under standard conditions, were stained with CbR and scanned at 550 nm as described in Methods. MITO: 65 pg mitochondrial membrane protein; M denotes the major band. MITO - ZGM-3: a mixture of mitochondrial membrane (36 pg protein) and purified granule membrane (15 pg protein). ZGM-3: 39 pg granule membrane protein; the positions of bands 2 and 15 are indicated. Insert: portion of a scan of a gel containing 40 pg of granule membrane protein and a smaller amount of mito- chondrial membrane (14 pg protein). The absorbance spike at the far right marks the migration of the tracking dye. ABSORBANCE. 550 n m "g NHTO - O V I _a S hMTO + ZGNFS ‘3” (I "Is EN l J l l 2 4 6 DISTANCE (cm) Figure 2 8 46 characteristic mitochondrial band, designated M, had a molecular weight of 26,000 and accounted for 10 to 15% of the profile. Band M and band 15 of ZGM-3 had nearly identical mobilities and it was considered whether band 15 represented mitochondrial contamination. When a mix- ture of purified granule membrane and similarly treated mitochondrial membrane was subjected to electrophoresis as shown in the middle scan of Figure 2, band 15 migrated slightly faster than the M band. The difference in mobility was readily demonstrated when a smaller amount of mitochondrial membrane was added to ZGM-3 so that both components were of equal staining intensity, and when electrOphoresis was pro- longed (Insert, Figure 2). On this basis, in addition to the absence of significant cytochrome c oxidase activity, it was concluded that mitochondrial contamination of the final granule membrane preparations was negligible. However, the mitochondrial contamination of ZGM-l was high. Frames B and C of Figure 1 illustrate that mitochondria were removed by the discontinuous gradient centrifugation step. Most of the stain intensity in bands 15/16 of the crude membrane preparation disappeared at this step; apparently band M accounted for most of the stain intensity in this region. The doublet nature of band 19 in ZGM-l and the rather high stain intensity relative to the other bands suggest that this band contained both the zymogen granule component as well as low molecular weight polypeptides of mitochondrial membranes. The loss of adsorbed secretory proteins from the membrane during purification is also documented in Figure l. The majority of the stain intensity in ZGM-l is associated with identifiable secretory 47 proteins, particularly amylase. The enrichment of membrane Species was monitored by an increase of band 2. Although significant loss of soluble protein occurs at the discontinuous gradient step, the release of adsorbed contents was essentially complete after 0.25 M NaBr extraction. Strong chaotropic agents have been shown to disaggregate membrane protein complexes (Hatefi and Hanstein, 1969). 0.25 M NaBr was chosen as a mildly chaotropic reagent which might selectively elute adsorbed contaminations without affecting intrinsic membrane polypeptides (5.0. Aust, personal communication). Figure 3 illustrates that this assumption was borne out. Polypeptides with mobilities identical to components of the granule contents were selectively eluted from ZGM-Z with 0.25 M NaBr, and apparently were secretory proteins which had adhered to the membrane. Bands 8, 12, 14, and 16 ’ were quantitatively extracted, while other content components, bands 3, S, 6 and 7 were partially extracted. The only polypeptides which were preferentially solubilized, but could not be correlated with secretory proteins, were minor bands 9, 11 and 19. Figure 3 demonstrates that ZGM-3 contained several components that were masked by NaBr solubilized bands. ZGM-Z contained two minor overlapping bands, designated 9/10. By comparing mobilities of these bands in the 0.25 M NaBr solubilized fraction (ZGS-3, Figure 3) and in the final membrane fraction (ZGM-3, Figure 3), it was concluded that band 9 partitioned with 208-3, and band 10 with ZGM-3. The partially resolved bands 15/16 were similarly separated. In addition, removal of bands 12 and 14 by the NaBr wash revealed band 13 with an inter- mediate mobility. 48 Figure 3. Identification of membrane-bound and adsorbed polypeptides of ZGM-2. ElectrOphoresis of the intermediate fractions was perfOrmed under standard conditions with the following amounts of protein: ZGM-Z, 36 pg; ZGS-3, the 0.25 M NaBr solubilized fraction, 40 pg; ZGM-3, 42 pg. Gels were stained with CbR and scanned at 550 nm. The absorbance spike at the far right denotes the migration of the tracking dye. ABSORBANCE .550 nm —0 O 23 ZGM-Z l9 Figure 3 2 4 6 DlSTANCE (Cm) 50 Further Characterization of the Granule Membrane Components The remaining polypeptides of ZGM-2 which were not signifi- cantly extracted by 0.25 M NaBr and were enriched in ZGM-3 were examined further. Estimated molecular weights of the intrinsic granule membrane polypeptides are given in Figure 4. The standard curve was constructed from duplicate analyses of proteins of known molecular weights in 9% polyacrylamide, 1% SDS gels at pH 7.4. Mobilities are recorded relative to the tracking dye, pyronin B. Membrane polypeptides were analyzed in parallel gels and in gels which included standard proteins. Line placement within the linear portion of the standard curve was determined by computer least squares analysis. The non-linear regions above 70,000 and below 15,000 were not unexpected (Weber and Osborn, 1969), and increase the uncertainty of molecular weight estimates of polypeptides of unknown size with mobilities within this region. In order to limit the uncertainty of the molecular weight estimates, mobilities of granule membrane components relative to standard proteins were analyzed in two other SDS-polyacrylamide gel systems. Increasing the monomer concentrations without altering other conditions has been shown to reveal inaccurate estimates for glyco- polypeptides (Bretscher, 1971; Segrest et al., 1971). Electrophoresis in 1% SDS at pH 2.4 has also been shown to detect changes in mobility for Specific membrane polypeptides relative to proteins of known molecular weight (Fairbanks and Avruch, 1972). Molecular weight values for the major polypeptide species of ZGM-3 obtained in the 51 .000000 .00 00 0000000000000 H ooe.m0 0000: «who: u oeonnoouxo 0 oom.e0 00003 mmo oexuomx0 m oom.m0 00000009 moon 000 ommo0oscon0a 0 000.00 000000 00000000 00000 000000000 0 oov.w0 c055: :003000m00oe0nm a ooo.0m eme0xOm 0o00n0nc0 c0mmxnu o ooo.m~ m000ocmm moon awmmxau : oom.mm m000ocem ween < cowoC0mmxhuoexnu-0 E ooo.nm ammo» omecomoupxnep Honoon 0 ooo.om 000035 000000 ommcomOprnov 0000mmo0m opxgop0enoox0m x 000.00 000000 000000 00000000 0 ooo.ov 000000000 omeuegmmonm 0:00mx00 : ooo.mm 000000 mo: omep0umomo0000 0:00:00 w 000.00 00>00 0000 00000000 0 000.00 000000 00000000 000000 000000 000>0000 0 ooo.wo moon c08500e Ednom p ooo.vm 000030 000000 e ommeuosmmozm o ooo.OM0 0000 m0co0nonomm ommv0mouom0ewum 0 0000.000 000 0000000000000 0 000003 000000002 0000000 .00:M0m 000 :0 mc0ouoHQ pudendum 000 000M0ucov0 30000 00000 000 .000300000 00000 000000 .000000 00000 000 000000 000000 000 00000 .000000 0000000 paw Nuo0x an mxm>03m 500m 00:00000 0003 m0 awsounp m pope:M0mopv mC0ououm pumpcmpm on» now mosam> unw003 000300005 009 .Am 0:0 0 moH=M0m pew 0x00 o0 Homouv 0005:: m00 an 0000000:0 m0 peep anum some mo CO0umaw0E m>000000 0:9 .mnhv exp MC0xoenu 0:0 00 0>0um0oh commouaxo one mo0u000noz .00000fivcoo 00000000 00003 000m 000ee0xaom 0m co ponx0eem 0003 munM0o3 0003000oa :3ocx mo mc00000m .0.0 mm 00 wow 00 00 00000000 -0000000 00 mucocomEoo ocmhpaoe o0scmam somoexN mo m00m003 000300005 6:0 00 cowueaflumm .0 oH=me 52 OK! 1 l o 9000 m N ,0: x .LH9I3M W‘Inoa‘low RELA‘ILIVE MOBILITY Figure 4 53 three systems are compared in Table 4. Although heterogeneity of band 5 was revealed, gross differences were not observed. Component 2 was the major granule membrane polypeptide, accounting for approximately 37% of the total protein stain intensity of Coomassie blue stained gels (Table 4). The unusually broad mole- cular weight distribution (70,000 to 83,000) of band 2 suggested that it contained a heterOgeneous population of molecules. Band 5 appeared to be a mixture of two polypeptides. The level of amylase activity in ZGM-3 (Table 2) implied that amylase comprised no more than 1% of the gel profile, however, contribution by inactive adsorbed enzyme cannot be rigorously excluded. Mixtures of purified granule membrane and soluble granule contents in different ratios were subjected to prolonged electrophoresis (8 hours) in an attempt to differentiate between band 5 and the amylase band. These electro- pherograms revealed a much broadened band 5, which was not caused by overloading; however, two bands could not be completely resolved. Furthermore, amylase has a small periodic acid-Schiff (PAS) positive component (cf. Figure 11) which is near the minimum level of detection in polyacrylamide gels. As will be demonstrated (cf. Figure 9), granule membrane component 5 has a much stronger periodic acid-Schiff reaction, not accountable by the presence of amylase alone. Figure 5 illustrates Subfractionation of band 5 by electrophoresis in gels of higher acrylamide concentration and in acidic buffer. The covalent association of significant carbohydrate to one polypeptide would be expected to shift its migration toward a lower apparent molecular weight in these two electr0phoresis systems (Segrest et al., 1971; Fairbanks and Avruch, 1972). 54 .00 000000 .000 00000 000 0000 00000000000 .005050x0 050000000000 05000505 003500» 50m0500 00000000 mo 000535 030 00 50 .N.0N .00 050500500 00.00 .0 050500500 00.0» .0 050509500 00.0 .0 050500500 0003 500000 0.0 mano0505000000 00 0:0 50 50000000000 003500m 50m0500 00300>0050 0000 00m 0500 0000 m0 005005300 0>000000 0:50 .50>0m 000 0500000>00 00005000 005000 000 5000 00000 m5000003 050 030 0500030 00 000050000 003 005005300 0>000000 0500 000 000 050 0055000 0003 0000 0050000 0300 0000050000 .500000 00» 0000 000 0 005000 000 000000000 00 000050000 0003 00:M003 0003000020 .005005500 0>0000o0 M500050000 50 00030050 005 003 500030000500 05050000 000 000000 00 cm 0500 0050w0 . . - . 000.00 000.00 000.00 00 000 0 0 + 0 00 -000.00 -000.00 -000.00 00 - 0.0 0 0.0 000.00 000.00 000.00 00 - 0.0 0 0.0 000.00 000.00 000.00 00 . . I . 0000mm ooo.mm . 0 0 0 0 + 0 00 0000.00 0000.00 000 00 0 0.00 0.0 0 0.00 .000.00 000.00 000.00 0 00.00 0.0 0 0.0 000.0000 000.000 000.000 0 000000 0000000 0000-0000 000000 0.0 :0 0.0 =0 0.0 =0 0.0 00 0005000000 00.0 0005000000 am 0005000000 0m 0005000000 000 0005000000 am 050500500 0 0000 00000000< 0>000000 000003 000000002 0 .0000umonx0om 05000505 003500» 50mo5xu mo usm003 000300005 050 005005300 0>0u000m .0 00005 55 Figure 5. Subfractionation of zymogen granule membrane band 5. Purified granule membrane was subjected to electrophoresis under the following conditions: A) 9% acrylamide, 1% SDS, pH 7.4 (standard conditions); B) 12% acrylamide, 1% SDS, pH 7.4 (high acrylamide concentration); C) 9% acrylamide, 1% SDS, pH 2.4 (acidic conditions); details of each procedure are given in Methods. Only the area of the scan containing band 5 is shown. Relative absor- bances of the three scans are not comparable. A550nm 56 J DISTANCE Figure 5 57 Band 19, a broad band containing polypeptides of molecular weight less than 14,000, was detected in the membrane fraction at each stage of purification (Figure 1). Several lines of evidence suggest that it represents proteolytic products which remain asso- ciated with the membrane. The appearance of band 19 was extremely variable (Table 4). In several cases its contribution to the stain intensity of granule membrane electropherograms was less than 2%; generally it accounted for approximately 13%. Storage of intact zymogen granules in ice fer 17 hours prior to membrane purification resulted in a polypeptide profile with a reduction of bands 2 and 5 and a significant induction of band 19, when compared to a profile of freshly prepared membrane from the same zymogen granule preparation. Trypsin digestion of isolated granule membrane led to identical results. The isolation of cytochrome b and cytochrome b reductase 5 5 after limited proteolytic treatment of microsomes illustrates what may be an ana10gous phenomenon (Spatz and Strittmatter, 1971, 1973). Proteolysis preferentially occurs at a site linking a hydr0philic head portion to a hydrophobic tail imbedded in the membrane. Thus, band 19 appears to represent the hydrophobic portions primarily of components 2 and S which remained associated with the membrane after limited proteolysis. The material within band 20 appeared opalescent when viewed by indirect lighting. In addition, Coomassie blue R stain was preferentially lost from this band during prolonged destaining with 33% methanol-10% TCA. Comparison with the mobilities and staining properties of sphingomylin, phosphatidyl choline, and glycolipids 58 suggested that band 20 represented membrane lipids. (3H)NaBH4 reduction of Schiff base intermediates formed by the addition of formaldehyde to purified zymogen granule membrane solubilized in 1% SDS yielded a labeled preparation containing approximately 2 x 104 cpm per ug protein. Analysis of the membrane components associated with the majority of the label provided further evidence of the lipid nature of band 20. The radioactivity associated with low molecular weight material, primarily (3H)methanol, was removed by chromatography of the labeled membrane sample through a Bio Gel P-4 column equili- brated with 1% SDS. Electrophoretic analysis of the labeled membrane is presented in Figure 6A. Significant label was associated with membrane components 2, S and 19. Eighty-three percent of the label, however, migrated with band 20. When the labeled membrane was ex- tracted with chloroform-methanol (2:1), 88% of the total radioactivity was soluble and 12% was retained in a pelletable residue, presumably membrane protein. Essentially all of the radioactivity of the chloroform-methanol soluble fraction migrated on a silicic acid thin layer plate developed with chloroform-methanol-water (100:42:6) (Figure 6B). Thus, band 20 appears to be membrane lipid rather than low molecular weight polypeptides. The recalcitrant nature of many membrane proteins to dissolu- tion into individual species prompted additional, more rigorous solubilizing conditions in an attempt to simplify or alter the granule membrane polypeptide profile. Membranes were suspended to 2 mg protein/ml in 1% SDS sample solvent buffer without 2-mercaptoethanol, then treated at 23° by the following procedures: a) dialysis against 59 Figure 6. Labeling of zymogen granule membrane components with (3H)NaBH4. Purified granule membrane was treated sequentially with formaldehyde and (3H)NaBH4 as described in Methods. A) Distribution of radioactivity of 3H-labeled membrane subjected to electrophoresis under standard SDS conditions is shown; bands 2, S and 19 are indicated. B) A chloroform-methanol (2:1, v/v) extract of the 3H-labeled membrane was analyzed by silicic acid thin layer chroma- tography. Radioactivity on the thin layer plate was detected with a Berthold Radioscanner. The phospholipid standards are S, sphingo- mylin; PC, phosphatidylcholine; PE, phosphatidylethanolamine; and PS, phosphatidylserine. Lipids were detected by reaction with iodine vapor. 60 4‘ ’3' TD 2... 'P Q l4 x. 5'3. 2.4- 0- . Q 2 5 l9 .2- -. JJ .. I r l l I 1 I 20 4O 60 80 I00 FRACTION NO. 3H-LABELED ZGM-3 CHC -C OH (Zflhfiofibmnes) v EXTRACT RESIDUE 88% 101d cpm I296 total cpm v mam 0 HCII’J‘CJ L! (g {3 " MP0 - 3&1- L OWL 5 PC PS PE ”In" from Figure 6 61 sample solvent for 12 hours; b) dialysis against sample solvent con- taining 6 M urea for 12 hours; c) addition of 2 volumes of 2- chloroethanol at 0° for 2 hours followed by dialysis as in a); d) extraction with chloroform-methanol (2:1) and solubilization of the residue by dialysis as in a). The membrane samples were then re- duced by the addition of 2-mercaptoethanol to 1%, heated at 60° and subjected to electrophoresis under standard conditions. None of the treatments altered the mobilities of the major polypeptides. In another experiment, granule membranes were dissolved in phenol-acetic acid-urea-water (2:1:l:l, w/v/w/v) at a protein concen- tration of 1 mg/ml and subjected to polyacrylamide gel electrophoresis in 7.5% acrylamide gels according to the procedure of Zahler et al. (1970). After 10 hours at 14 volts/cm, one gel was briefly stained with CbR. This electropherogram was also characterized by a simple profile (Figure 7). The regions in an unstained, unfixed gel corres- ponding to the two major species were estimated by comparison to the, stained gel. After sectioning, the gel segments containing the two major species were equilibrated with 1% SDS sample solvent at room temperature for 4 hours, then placed on top of standard pH 7.4, 1% SDS polyacrylamide gels and subjected to electrOphoresis as usual. The slower migrating broad band from the acetic-acid urea gel gave rise only to band 2 of the SDS profile of ZGM-3, while the other major component contributed only band 5. Detergents other than sodium dodecyl sulfate were tested for their ability to solubilize granule membrane polypeptides. Membrane samples were suSpended in detergent solutions by brief sonication. 62 Figure 7. Analysis of zymogen granule membrane polypeptides by acetic acid-urea polyacrylamide gel electrophoresis. Details of the electrophoresis procedure are given in Methods. A_is a Coomassie blue stained profile of 44 ug of granule membrane protein separated in the acetic acid-urea gel system; the major two bands are designated I and II. Gel sections containing bands I and II from a parallel gel run as in A were subjected to electrophoresis under standard conditions: §_contains band I; §_contains band II. D, 39 ug membrane protein run simultaneously with gels B and C but without prior fractionation by acetic acid-urea electrophoresis. 63 0 Figure 7 64 The proportion of band 2, determined by SDS electrophoresis, which remained in suspension after centrifugation at 150,000 xg for 90 minutes was used to assess solubilization. Table 5 lists the effects of several detergents. Even though detergents were present in great excess (12.5 to 188:1 w/w), little solubilization was effected. Two percent Tween 20 plus 1% deoxycholate, which has been shown to solu- bilize outer nuclear membrane and fucose-containing components of HeLa cell surface membrane (Atkinson and Summers, 1971; Penman, 1966) was the most effective of the series. Comparison of Membrane Polypeptides of Microsomes, Mitochondria and zymogen Granules The total number of granule membrane polypeptides and their molecular weight distribution are clearly distinct from those of microsomal and mitochondrial membranes. Figure 8 compares the CbR stained polypeptides for all three membranes prepared similarly by NaHCO3 and NaBr extractions. Electrophoresis of membranes from total pancreatic microsomes revealed 35 discrete bands. Since the isolation procedure removed most of the microsomal RNA (R. A. Ronzio, unpub- lished observations), the contribution of ribosomal proteins to this profile was minimal. By mixing microsomal and granule membranes prior to electrophoresis, the unique mobility of component 2 was demon- strated. Less than 2% of the stain intensity of the total microsomal profile occurred in the region around 74,000 molecular weight. Band 5 migrated with a major microsomal component, which may also contain amylase. The low molecular weight polypeptides characteristic of microsomes were absent from granule membranes. 65 Table 5. Detergent solubilization of zymogen granule membrane protein. Samples of zymogen granule membrane (ZGM-Z) were added to polyethylene microfuge tubes (Beckman) containing the detergents indicated below and sonicated at maximum setting (Biosonik, Bronwill Scientific) for 10 seconds. Solubilized material is defined as that which did not sediment after centrifugation at 150,000xg for 90 minutes. Total protein was measured according to Lowry et al. (1951). The relative amount of component 2 solubilized was determined by SDS-polyacrylamide gel electrophoresis (Laemmli, 1970) of the soluble and pellet fractions. The presence of other detergents did not affect electrophoresis. Values given in the table represent solubilization above the minimum observed upon sonication in distilled water and centrifugation. Detergentzprotein % protein % component 2 D°tergent (w/w) solubilized solubilized 2% Tween 20 125:1 0 none 0.2% deoxycholate 13:1 26 12 0.2% Tween 20 plus 0.1% deoxycholate 19:1 10 none 2% Tween 20 plus 1% deoxycholate 188:1 31 16 2% lubrol 125:1 21 none 2% NP40 125:1 16 none 66 Figure 8. Comparison of the polypeptides of zymOgen granule mem- branes, mitochondrial membranes and total microsomal membranes. Membranes were prepared by extraction with NaHCO3 and NaBr as described in Methods. Electrophoresis was performed under standard conditions, and the gels were stained with CbR. Mobility is expressed relative to the dye pyronin B. The following samples are compared: a) granule membrane, ZGMa3 (39 ug protein); b) mitochondrial membrane (65 ug); c) a mixture of granule membrane (15 ug) and mitochondrial membrane (36 pg); d) membrane from total microsomes (63 ug); e) a mixture of granule membrane (28 ug) and total microsomal membrane (42 ug). 67 >55 9.: ma: .1 . . i u u u u u n n "W 2 4 6 8 mm Figure 8 68 As described earlier, characteristic mitochondrial membrane components were not detected in purified granule membrane prepara- tions. Conversely, Figures 2 and 8 demonstrate the absence of granule membrane components in mitochondrial membranes and emphasize the difference in complexity of the two membranes. A dozen polypeptide bands appeared common to mitochondrial and microsomal membranes, but the relative amounts of these bands differed, and each membrane contained unique bands. Both contained a major component with a mobility identical to band 5. The staining of the lipid region of these two membranes was consistently lower than granule membrane for comparable amounts of protein, indicating that the granule membrane may have a relatively low protein to lipid ratio. Carbohydrate-Containing_Components of Pancreatic Intracellular Membranes ElectrOpherograms of ZGM-3 stained for carbohydrate by the periodic acid-Schiff (PAS) procedure reproducibly gave the absorbance profile illustrated in the lower frame of Figure 9. One band barely penetrated the gel surface and was not evident in the scan. The intense staining immediately behind the tracking dye has been attri- buted to lipid (Fairbanks et al., 1970; Lenard, 1970a). When a mixture of glycolipids was subjected to electrophoresis and the gel stained with PAS, the only region with stain was just behind the dye marker. The two major species at 74,000 and 52,000 molecular weight correspond to bands 2 and 5 of the CbR profile (upper frame, Figure 9). Thus, 69 Figure 9. Glchprotein nature of zymOgen granule membrane components. The periodic acid-Schiff (PAS) and sialic acid profiles of zymogen granule membrane (lower frame) are compared to the CbR pro- file (upper frame) of a second polyaerylamide gel run simultaneously. Techniques are described in Methods. Upper frame: 50 ug of ZGM-3 on a 5 mm diameter polyacrylamide gel. Lower frame: PAS (A560), 50 ug of ZGM-3 on a 5mm gel; sialic acid profile, approximately 200 ug of ZGM-3 on a 6 mm gel. ABSORBANCE 70 .4»; ; A _&A_S_ :5 2. h... M 5.2 1'A_ [J] 0 “3+- 0) DISTANCE (cm) ”O 7'5 50 35 25 MOLECULAR WEIGHT X|O3 Figure 9 nmolcs SIALIC ACID H 71 the major polypeptide Species of the granule membrane appear to be extensively glycosylated. The sialic acid content of ZGM-3 was fOund to be approximately 130 nmoles per milligram protein. The distribution of sialic acid components separated by electrOphoresis is shown in the lower frame of Figure 9. The recovery of sialic acid from sectioned gels was variable, with an average of 30%. The diffuse nature of the high molecular weight zone containing 70% of the membrane sialic acid may be due to overloading the gel. Nevertheless, the peak fractions of sialic acid were coincident with components 1 and 2. Membranes were prepared from rough and smooth microsomes isolated by isopycnic sucrose gradient centrifugation. Only minor differences, most notable in the molecular weight region 35,000 to 70,000, were observed in the distribution of CbR stained polypeptides (Figure 10). These differences were not due to contribution of ribo- somal proteins, since a preparation of ribosomes analyzed separately did not contain predominate bands in this region and since the majority of ribosomes are lost from rough microsomes during NaHCO3 and NaBr extractions. In contrast, the two membranes contained markedly different molecular weight classes of carbohydrate-containing components. Aside from the intense staining of the lipid region, smooth microsomal membrane preparations consistently demonstrated a major peak of PAS stain at a polypeptide equivalent molecular weight of 75,000 (Figure 10). Rough microsomes contained a heterOgeneous distribution of ten PAS-positive bands; the presence of several bands in the low 72 Figure 10. Comparison of the polypeptides and glycopolypeptides of mitochondrial and rough and smooth microsomal membranes. Details of membrane isolation and electrophoresis on 6 mm diameter polyacrylamide gels under standard conditions are given in Methods. Samples in each frame are A) smooth microsomal membrane, B) rough microsomal membrane, and C) mitochondrial membrane. Legend: Dotted lines outline CbR profiles (A550); samples contained 50 ug protein. Solid lines outline PAS profiles (A560); samples contained 150 ug protein. 73 Ecowm. woz oeou noum< .mncmn aflououm on» oumoofi on can no“: vocfimumeofisc one: 0 one m .< voxwme meowfim moan» one .maso: ma you we: mfimonogaouuoofim .aou on» an Lmsouu on» a“ nofiammm me: He o.m cw aflououm ocmupaoa oascmnm mop mo we n.o .mpocuoz :« nonwnomou one mom wa.o a“ mfimononmouuooao use How on» we cofiumnmmoum .mucocomeoo ocanEoE ofiscmum comoexn mop mo :ofiumpamom How beam ucofiumam owfiemaxhomxaoa o>wumammonm .mH madman 84 ms ousmflm 85 combined fractions ranged from 28-44% of the protein applied for five preparations. The crystalline material accounted for between 15 and 35% of the isolated protein measured according to Lowry et a1. (1951). Purity Figure 14 compares electropherograms of the non-crystalline material with the original granule membrane preparation. When electro- phoresis was performed in the same system used for isolation, a Single discrete band was observed (Figure 14, gel B), which was expected from the simple direct approach of the isolation procedure. Electro- phoretic analysis in acetic acid-urea as an independent measure of purity was not as conclusive. Minor diffuse bands barely detectable by photography were observed in addition to the major protein band. Since these minor bands did not coincide with bands present in the whole membrane preparation, it was not clear whether the additional bands represented impurities, products of limited proteolysis, or artifacts generated by the electrophoretic technique. Isolated com- ponent 2 was difficult to solubilize, requiring large volumes of 1% SDS solution and heating. This recalcitrant nature of isolated component 2 prohibited a thorough electrophoretic analysis. The major bands in gels C and D of Figure 14 were marked with ink, then completely destained by shaking in 10% TCA-33% methanol for nine days at 37°. After staining for carbohydrate by the PAS pro- cedure, the only stain observed was located precisely at the ink mark, i.e. C2, on both gels. These results indicate that the carbohydrate moiety remains associated with the major Coomassie blue 86 Figure 14. Purity of isolated dOg granule membrane component 2. A) Center Strip of a preparative slab gel Stained with CbG illustrating the separation and distribution of deg zymogen granule membrane components. B) After excising and electroeluting the region of gel con- taining component 2 for the preparation shown in A, the sample was electrophoresed on a second preparative Slab gel. B_shows the CbG stained center slice of the gel containing isolated component 2. The duration of electrophoresis for the second slab gel was much shorter, thus the mobilities of components in the two gels are not directly comparable. C) Separation of dog zymogen granule membrane (80 ug) com- ponents by acetic acid-urea gel electrophoresis (cf. gel system II, Table l), Stained with CbR. D) Acetic acid-urea gel electrophoresis of isolated component 2 (96 ug protein). Compare with whole membrane in gel 9: E) Acetic acid-urea gel electr0phoresis of proteins of known molecular weight, from top to bottom: B-galactosidase, phOSphorylase a, bovine serum albumin, alcohol dehydrogenase and lysozyme. 87 Figure 14 88 stained band during purification and under distinctly different dissociating and electr0phoresis conditions. The acetic acid-urea system was employed to estimate the molecular weight of isolated component 2. In Figure 15 the mobility of several proteins relative to lysozyme is plotted against the logarithm of their molecular weight. Although not precisely linear, the curve could readily be used to estimate the molecular weight of ‘proteins between 94,000 and 14,000. By this method the middle of the component 2 band (Figure 14, gel D) corresponded to a molecular weight of 72,000 (Figure 15), very similar to the estimate obtained in 1% SDS. The crystalline material obtained during dialysis contained a high proportion of bound SDS, measured according to Reynolds and Tanford (1970). The mean SDS content of two preparations was 8.7 mg/mg protein. Electrophoresis of 150 ug of the particulate protein yielded only faint staining limited to two diffuse bands with higher mobilities than component 2. Less protein could not be visualized. In addition, the two bands did not Stain by the PAS procedure. By these criteria the material was not identical to component 2. ZZEOgen Granule M 2+-de endent Adenosine Triphosphatase (EgZT-ATPase) Mg2+-dependent adenosine triphosphatase activity is distributed throughout the subcellular fractions of rat (Ronzio, 1973b and unpub- lished observations) and guinea pig (Meldolesi et al., 1971c) pancreas. Although the zymogen granule accounted for only 1 percent of the total homogenate activity, the Specific activity of the granule membrane was extremely high (Ronzio, 1973b; Meldolesi et al., 1971c). 89 Figure 15. Estimation of the molecular weight of isolated dog granule membrane component 2 by acetic acid-urea polyacrylamide gel electrophoresis. Protein Standards and purified dog component 2 were dissolved in phenol-acetic acid-urea-water-2-mercaptoethanol and subjected to electrophoresis at 13 volts/cm in acetic acid-urea as described in Methods. Mobilities are expressed relative to lysozyme. The standards are a) B-galactosidase, b) phosphorylase a, c) bovine serum albumin, d) alcohol dehydrogenase and e) lysozyme. The rela- tive mobility of component 2 is noted as C2. 90 a I A m 005 MOLECULAR WEIGHT xI0'4. N l I l J I .2 .4 .6 .8 IO RELATIVE MOBILITY Figure 15 91 TheggZT-ATPase is Tightly Bound to the ranule Membrane The extent of association of the granule activity with the membrane was demonstrated by sedimentation of a lysed granule pre- paration through a 0.2 M to 1.2 M sucrose gradient. Large cellular debris and nuclei were removed from a pancreas homogenate by a brief centrifugation at 600xg. A granule fraction rich in mitochondria was then collected by centrifuging the supernatant at 1600xg for 30 minutes. The brown mitochondrial layer was removed from the pellet and the zymogen granules were lysed by resuspension in 0.17 M NaCl containing 0.05 M Tris-Cl, pH 8.2. The lysate was layered above a 0.2 M to 1.2 M sucrose gradient containing a 1 m1 cushion of 1.75 M sucrose. Intact rat zymOgen granules and mitochondria band at densities of approximately 1.219 and 1.215 (about 1.67 M sucrose), respectively (R. A. Ronzio, unpublished observations). Figure 16 illustrates the results of sucrose gradient centrifugation of the lysed granules. Mitochondria monitored by cytochrome c oxidase activity were found at the 1.2 M- 1.75 M sucrose interface (arrow, Figure 16). The low level of amylase activity detected at the inter- face indicates that lysis was essentially complete. Two peaks of MgZT-ATPase activity, containing more than 75% of the total, were coincident with two opaque granule membrane bands observed near the middle of the gradient. The separation of granule membrane into two peaks appears to be caused by adsorption of secre- tory proteins and mitochondria to a fraction of the membrane popula- tion, resulting in an increased equilibrium density of that fraction of membrane. 92 Figure 16. Association of Mg2+-ATPase activity with the zymogen granule membrane. 0.7 ml containing 0.8 mg of protein of a lysed granule sus- pension was layered over a 9.7 ml sucrose gradient (0.2 M to 1.2 M sucrose) on a 1 ml 1.75 M sucrose cushion. After centrifugation for three hours at 192,000xg (SW 41 rotor, Beckman) and 0°, 0.33 ml fractions were collected and assayed for amylase (-—o——), cytochrome c oxidase (—n—), and Mg2+-ATPase (assay A) (—A——) as described in Methods. 93 Tl zo_Sfiuu< cofiumUAmHHSQ mxufl>wuu< Hench ”much * -eHom oAonogm s :ofipoeam mfimxflouuag aev :Aoboaa .osxuco poum>wuum may on o>HumHou one wouo>ooop xpfl>fiuum w one :oflum0fimfihamnpaom new voucomoum «new one monk .xuw>wuuw ommmefluom ofinmuo>ooou Hmpou mo oceanom oumowvcfi momonuconmm a“ nonezz Asovm.o fisemve.mm 0.8m ae< :2 H was +Nmz 2e N mafia ammz z m~.o flammvw.m ASHAVS.VN 0.0m ummz z m~.o mannaom mIZUN NIEUN pcosumouh moa¢3-Spfl>Auu< .omemHflpumofieam .fi>\>\> .H mm coy so mnflpcoomop so vomoHo>op one: meanwOmeougo one .uommm nouaflm H .02 masons: co popuomm cog“ who: monouxfle xmmmm new ocfimocopm one mz< .mo< Howunmu .mouscfla om mu one .oH mm .0 m4 um <05 yea mo ossHo> numfimuoco mcfivwm so woumcfieuou who: mammm< .onm um numb Hope: a a“ wnfiomam one noun mu“ :m scum chapxfie cowuumou ozu mcw>oEon xn woumwuficfl mm: xmmmm comm .xufl>fiuom ommmhfluomoflvon ogh .mpcooom :o>om Hoseauflpuo so you voscwucou mo: cofiuopsocfi ocu use mucouom o>wm nouwm povpm mm: me< 28 m .m uuomcH .mcofipflncou :ofiuonzocfi vcooom o>fim pnoncoum ocu nova: voflonma mo: ocmHnEoz .Hoaucou .< anomaH .onsmflm on» aw pogo: ohm mofloomm wouoflxhozmmocm oonnu pom magmaoz Hoasooaoe ouauaomxaom oumeflxOHmm< .voucsoo cam voofifim soap vo3 mfiou .mponuoz a“ ponwnumop mo ¢.~ no no mom wa cw mfimouozmonuooao oo pouoom25m use nonmoz .woumaxuogmmonm ouoz ocmunaoa Hofinwconuouae voflmflusm mo m1 oom .ocmpneoe Hmflpmconooufia poumfixpocmmogm mo meohmoponmouuuoflo :« vomIm mo defipsnwhumflo .Hm ouswfim O 118 AN ousmfln AEov mozflrma o. _ o x. N o. o . _ — COOK. B. 80 m N _ m N _ 0 4H allowed 6 80.0w. é dz: ‘WdO 119 Furthermore, 80% of the 32P04 associated with the mitochondrial 120,000 molecular weight species was released when 20-fold excess unlabeled Mg2+-ATP was added after the initial five second labeling period and incubation continued for an additional seven seconds (Insert B, Figure 21). The rapid turnover of label was distinct from the stability of the phosphorylated granule membrane polypeptide. DISCUSSION Analysis of Zymogen Granule Membrane Polypeptides The SDS polyacrylamide gel electrophoresis procedure of Fairbanks et a1. (1970) fulfilled several criteria for precise analysis of membrane polypeptide components, whereas other procedures did not (Kiehn and Holland, 1970; Hoober, 1970). Electropherograms of mem- branes from zymogen granules, mitochondria, and microsomes were highly reproducible. Peaks accounting for 2% of the protein stain intensity (0.2 to 0.4 mg protein) could be accurately measured. Procedures such as lipid extraction, dissolution in organic solvents, prolonged dialysis and heating in solvent buffer, failed to alter electrOphero- grams of granule membrane polypeptides. These results confirm the earlier report that this method is relatively free of artifacts (Fairbanks et al., 1970). Since traces of secretory proteins proved difficult to remove from purified membranes, it was essential that proteolySis be inhibited during membrane solubilization and during electrOphoreSis. Dissolution in 1% SDS combined with heating at 100° are sufficient conditions to inhibit most proteases (Fairbanks et al., 1970). 120 121 The persistent problem of proteolytic degradation during membrane isolation is particularly acute in studies of pancreatic sub- fractions. Release of hydrolytic enzymes and activation of zymogens can lead to marked degradation of membrane polypeptides, resulting in a drastic alteration of membrane electrOpherograms. Control experi- ments, in which proteolytic inhibitors were not included in the homo- genization medium, showed that degradation of microsomal membrane polypeptides lead to a dimunition of major polypeptide bands and the formation of a heterogeneous pOpulation of low molecular weight species resulting in poorer resolution of bands (R. A. Ronzio, unpublished observations). Proteolysis of zymogen granule membrane was apparent in a different manner. Under conditions which promote degradation, an increase in small polypeptides with a limited size distribution (11,000 to 14,000 molecular weight) was observed. By this criteria, rat granule membrane was susceptible to a significant but variable degree, while dag granule membrane was isolated with minimal degrada- tion, since a band 19 was not generally detected. The utility of the first two steps in the purification of zymogen granule membranes has been demonstrated previously. Granule lysis in slightly alkaline solutions was first used by Hokin (1955) to prepare zymogen granule ghosts. More recently, Meldolesi et al., (1971a) separated lysed granules from the more dense contaminating mitochondria by means of discontinuous gradient centrifugation. This step effectively removed mitochondria from granule membranes; however, the remaining amylase activity was appreciable, and electr0pherograms of these fractions indicated a significant contamination by the 122 soluble secretory polypeptides representing the granule contents. Extraction by 0.25 M NaBr was incorporated to remove these con- taminants. This chaotropic anion has been employed to solubilize mito- chondrial membrane proteins at a concentration of 2 to 4 M, approxi- mately lO-fold higher than employed in this study (Hatefi and Hanstein, 1969). Since the granule membrane accounted for such a small per- centage of the zymOgen granule protein, it was essential to determine the degree of purity of zymogen granule membranes. Secretory protein contamination was negligible. Since mitochondria are Similar in size and density, they represent the most likely particulate contamination of zymogen granules. However, mitochondrial membrane polypeptides were not present in granule membrane electropherograms and purified granule membranes were essentially devoid of cytochrome c oxidase activity. Furthermore, membrane polypeptide profiles of zymOgen granules, mitochondria, and microsomal subfractions were distinct ' from each other, and distinct from the polypeptides in the post- microsomal supernatant and granule contents. Since a variety of procedures often employed to disaggregate protein complexes failed to alter the profile of the zymogen granule membrane, it was concluded that band 2 was not an aggregate. Band 2 accounted for approximately half of the PAS stain intensity. The granule membrane possesses a high content of sialic acid, associated primarily with high molecular weight polypeptides. These observations suggest that band 2 is an acidic glycoprotein species, and that its broad zone of staining may represent microheterogeneity due to variable carbohydrate content. 123 In a survey of several soluble proteins, Reynolds and Tanford (1970a) demonstrated that at neutral pH and low ionic strength pro- teins uniformly bind 1.4 gm of SDS per mg protein. This binding stoichiometry yields a relative constant charge density for protein- SDS complexes, and is partly the basis for the regular relationship between molecular weight and electrOphoretic mobility in SDS solu- tions. On the other hand, membrane proteins, particularly glyco- proteins, may not bind predictable amounts of SDS. Since the inter- action between SDS and protein is primarily hydrophobic (Rosenberg et al., 1969), membrane polypeptides containing extensive hydrophobic regions bind more SDS (Simons and Kaariainen, 1970; Spatz and Strittmatter, 1973). As a result, the electrophoretic mobility of very hydrophobic proteins increases relative to that of other pro- teins. The anomalous behavior of glycoproteins in SDS-gel electro- phoresis (Bretscher, 1971; Segrest et al., 1971; Russ and Polakova, 1973) is attributed to the lack of SDS binding by attached carbo- hydrate. More accurate molecular weight estimates are obtained from gels of higher acrylamide concentrations, since under these conditions the charge of the protein-SDS complex becomes less important relative to the sieving effect of the gel. Furthermore, the negative charge contribution by an appreciable number of Sialic acid residues associated with acidic glycoproteins would increase electrophoretic mobility. Consequently, the apparent molecular weight estimates of membrane glycoproteins are subject to considerable uncertainty. Variations of SDS polyacrylamide gel electrophoresis pro- cedures which have been shown to uncover discrepancies of glchprotein 124 mobility were applied to the analysis of component 2. Neither in- creasing the acrylamide content of the gel (Segrest et al., 1971) nor decreasing the pH of the electrophoresis solutions (Fairbanks and Avruch, 1972) altered the observed molecular weight of band 2. It is noteworthy that acetic acid-urea polyacrylamide gel electrophoretic analysis of isolated dog component 2 yielded a molecular weight of 72,000, consistent with all previous estimates. Band 5, with a molecular weight similar to rat pancreatic amylase, was a major band in ZGM-3, though present at a much reduced level than in ZGM—Z. Band 5 was 15 times greater than could be accounted for by amylase activity. Several observations suggested that band 5 was not primarily denatured amylase, but rather a membrane component. Bands 15 and 17 were sequentially enriched in the membrane during purifications and could be distinguished from bands 14 and 16, with apparent molecular weights of 25,000 and 23,000, respectively. The latter correSponded to polypeptides in the granule contents and may be trypsinogen and chymotrypsinogen, reSpectively (Sanders, 1970). Band 19 represents a major part of the ZGM—3 profile. Ribonuclease, a relatively minor product of the rat pancreas (Kemp et al., 1972), migrated somewhat less rapidly than the peak of this band. Band 19 may be a family of small membrane polypeptides. Polypeptides of this size have been observed in a variety of membrane classes (Kiehn and Holland, 1970; Swank et al., 1971). As mentioned earlier, band 19 could represent degraded membrane polypeptides retained within the membrane. This Species did not stain with PAS, and thus it does 125 not include carbohydrate containing proteolytic fragments of band 2. Band 20 Stains with both Coomassie blue and PAS and is a prominent part of the zymOgen granule membrane profile. Several lines of evidence indicate that this band represents lipid. Reduc- tive alkylation of intact zymogen granule membranes with formaldehyde and (3H)NaBH4 primarily labels lipid, and the same fraction of incor- porated radioactivity has the electrophoretic mobility of band 20. Several other investigators (Gahmberg, 1971; Lenard, 1970a; Carraway and Kobylka, 1970; Lopez and Siekevitz, 1973) have confirmed the lipid nature of this band. The high staining intensity in the lipid region of zymogen granule membrane electrOpherOgrams relative to gels of other membranes implies a high lipid: protein ratio. This indication has been verified by the observation that the phospholipid content of rat zymogen granule membrane is approximately 2 mg/mg protein. (Ronzio, unpublished data). In a related Study, Meldolesi et al. (1971b) noted that zymogen granule membrane from guinea pig contained considerably more phospholipid than other membranes. The high lipid content of membranes from chromaffin granules (Winkler et al., 1970) signifies that this may be a general phenomenon indicative of the functional simplicity of secretory granule membranes. Meldolesi and Cova (1972) have examined membrane polypeptides from guinea pig pancreas. Zymogen granule membranes from this source seem to be somewhat more complex than those of rat, though distribu- tion of glycosylated membrane components was not reported. The present evidence suggests that the zymogen granule membrane is unusually rich 126 in glycoproteins. These could simply be plasma membrane precursors, transmitted to the luminal plasmalemma during exocytosis. Alterna- tively the glycosylated components may possess specific granule functions. Preliminary experiments suggest that a portion of the Sialic acid can be cleaved from intact zymogen granules by neuramini- dase (R. Hsieh and R. A. Ronzio, unpublished observation). Conse- quently the membrane glyc0proteins are accessible to the cytoplasmic milieu, and may interact specifically with elements of the cytoplasm or with components of the luminal plasmalemma as part of the exocytosis mechanism. Storage granules have now been isolated from many cell types. In only a few cases have the membrane polypeptides been examined. Chromaffin granule membrane contains approximately 10 polypeptides as judged by SDS polyacrylamide gel electrophoresis (Hortnagl et al., 1971). One of the two major polypeptide Species is dopamine B-hydroxy- lase, an enzyme involved in synthesis of components of the granule contents. Parotid storage granule membrane has been analyzed using acetic acid-urea polyacrylamide gels (Amersterdam et al., 1971). While the level of contamination by granule contents is unclear, the results suggest that the granule membranes are much less complex than other membranes of this tissue. It appears that storage granules represent a class of highly specialized membranes, composed of relatively few polypeptide Species. It will be important to establish whether secretory granule membrane from other tissues have a small number of polypeptide components, and whether those which are present are glycosylated. 127 Comparison of Intracellular Membranes There have been relatively few attempts to define the intra- cellular distribution of glycopolypeptides in mammalian tissues. It has been noted that both rough and smooth microsomal membranes from liver contained Sialopolypeptides (Helgeland et al., 1972; Larsen, et al., 1972), and membranes of synaptic vesicles and synaptosomal plasma membranes possessed two common glyc0polypeptides (Breckenridge and Morgan, 1972). ElectrOpherograms of particulate and soluble polypeptides of adult rat pancreap suggested that glycopolypeptides, though widely distributed, were particularly enriched in zymogen granule membranes (compare Figures 9, 10 and 11). Glycopolypeptides of the postmicrosomal supernatant and zymogen granule lysate were not readily detected by the carbohydrate Stain. However (3H) glucosamine was incorporated into both classes of soluble proteins by tissue slices, although their specific radio- activities were lower than membrane fractions (R. A. Ronzio, unpub- lished observations). Neither the PAS-positive components nor the (3H) glucosamine labeled components corresponded to the major membrane glycopolypeptides. Polypeptides of rough and smooth microsomal membranes from rat pancreas were quite similar, though there were differences in the relative amounts of the common polypeptides. A similar conclusion was made for microsomal subfractions from liver (Zahler et al., 1970; Helgeland et al., 1972). The compositions of the smooth microsomal membranes and membranes of rat pancreas Golgi-rich fractions were similar to each other (Ronzio, 1973b), and it is likely that the 128 former were derived from the latter. The glycopolypeptide profiles of rough and smooth microsomes and mitochondria were considerably less complex than their Coomassie blue profiles and were distinctive for a given membrane class. The glycopolypeptides of smooth micro— somal membranes were quite Similar to zymogen granule membranes. A glycopolypeptide with the mobility of band 2 was particularly prominent. The restriction of this component to smooth microsomes exclusive of mitochondria and rough microsomes suggests a specific function. If the smooth microsomal component bears some relation to the major granule membrane glycopolypeptide, two explanations are envisaged. The hypothesis that zymogen granules form from coalescing vesicles originating from the Golgi apparatus (Jamieson and Palade, 1967b) requires a pre-existing pool of zymogen granule membrane poly- peptides within the Golgi membranes. The occurrence of the glyco- polypeptide in smooth microsomal membrane preparations, which are primarily derived from fragmented and resealed Golgi cisternae for such a pool. Alternatively, Since membranes from granules ruptured during homogenization would be expected to collect in the smooth membrane fraction, the PAS stain may represent contamination. Based on the specific radioactivity of (3H) glucosamine labeled granule membrane and smooth microsomal membrane fractions and the total amount of cellular granule membrane relative to smooth microsomes, it can be calculated that insufficient granule membrane exists to account for the level of (3H) glucosamine found in smooth microsomes (R. A. Ronzio, unpublished observations). Therefore, the latter alternative is excluded. 129 The differences in polypeptide compositions of the major membrane classes from pancreas support the proposal that these struc- tures are biochemically, hence, functionally distinct (Meldolesi and Cova, 1972; Ronzio, 1973a). The limited number of zymogen granule membrane polypeptides may be considered to mirror limited functional responsibilities. Unlike the endoplasmic reticulum which partially serves as a matrix for attached enzymes functioning in discrete pathways, thus facilitating reaction rates by limiting the enzymes to diffusion in two rather than three dimensions, the granule membrane does not appear to serve such a function. Another inference is that the mechanism for generating zymogen granule membranes requires the segregation of membrane polypeptides associated with the Golgi complex. This selection process may be linked to the glycosylation of granule membrane polypeptides. Finally, the uniqueness of each mem- brane Species indicates that membrane mixing does not occur during the intracellular transport of secretory proteins. ZymOgen Granule Membrane_Mg2+-ATPase Conceptual models of the secretory granule release reaction have been elucidated in detail by Poste and Allison (1969, 1974), Woodin and Wieneke (1970) and Mathews (1970). A single enzymatic activity, a Ca2+- or Mg2+-ATPase associated with many secretory granules, has been postulated to catalyze the fusion reaction of granule and plasma membrane. MgZT-ATPase has been previously demon- strated in zymOgen granule membranes from guinea pig pancreas (Meldolesi, 1971c). A Mg2+-ATPase of high specific activity has been found to be firmly bound to rat pancreatic granule membrane 130 (Figure 16; Table 7). The activity is slightly stimulated by the addition of high levels of Ca2+ in either the presence or absence of Mgz+ and is slightly inhibited by Mn2+. Because of the small magnitude of these effects, it is not likely that these observations reflect the basis for stimulation of granule release by Ca2+ or the inhibition by Mn2+ (Heisler et al., 1972), although more specific effects manifested in_vi!g_cannot be excluded. The most simple, but not singular, explanation of the kinetics observed for the granule MgZT-ATPase is the presence of two enzymes with similar activities. The broad substrate Specificity observed may also be indicative of more than one enzyme. However, a single enzyme displaying negative c00perativity (Levitski and Koshland, 1969; Teipel and Koshland, 1969) cannot be excluded at present. The observed kinetics do not appear compatible with models proposed for enzyme reactions which involve substrate and metal modifier inter- actions (London and Steck, 1969). A straightforward rationale for the presence of two granule membrane Mg2+-ATPase activities may be postulated from the data of Jamieson and Palade (1971) on the energy requirements of pancreatic secretion. The concentration of granule contents during condensing vacuole maturation did not demonstrate a definite requirement for continued intracellular ATP synthesis. On the basis of this ob- servation, Jamieson and Palade excluded the possibility of an exergonic ion pump, such as the plasma membrane Na+, K+-stimulated ATPase, functioning to remove ions, and thereby water, from the granule contents as a mechanism of maturation, unless the putative 131 enzyme had a high affinity for ATP. The diminishing ATP concentra— tion during exposure to antimycin A and sodium fluoride can be calculated from the data of Jamieson and Palade (1971) assuming an initial uniform cellular ATP concentration of 2 mM (Baudin et al., 1969; Tani and Ogata, 1970). Ten, 20, 40 and 60 minutes after the addition of the metabolic inhibitors, the estimated ATP levels are 0.8, 0.3, 0.2 and 0.1 mM, re5pectively. An enzyme with a Km for ATP between 0.1 and 0.2 mM would retain approximately 60% of its maximum activity under these conditions. During this 60-minute interval, condensing vacuole conversion decreased to 60% of the uninhibited level (Jamieson and Palade, 1971). The zymogen granule Mg2+-ATPase activity with high affinity for substrate had an apparent Km for ATP of 0.04 mM (Figure 19). If this activity were functional in condensing granule maturation, it would be expected to perform nearly optimally even in the presence of antimycin A and sodium fluoride during the interval investigated. Jamieson and Palade (1971) observed that the final step of granule release was quickly and effectively inhibited upon addition of the metabolic inhibitors. The energy requirement is most probably involved in fusion of the granule with the apical plasmalemma. The high Km Mg2+-ATPase activity (Figure 19) would be expected to be very sensitive to changes of ATP concentration below 2 mM. Indeed at this level it would be only partially active, and if involved in membrane fusion may require a local rise in the ATP level as part of the secretion process. The activity of the high Km enzyme would 132 be inhibited upon antimycin A and sodium fluoride treatment, parallel to the inhibition of secretion. 2 0 en Granule Membrane Protein K1nase Activity Since cyclic AMP has been indicated as an intermediate in the secretion stimulus for several cell types, including exocrine pancreas (Kulka and Sternlicht, 1968; Baudin et al., 1971; Ridderstap and Bonting, 1969), it is of obvious interest to investigate the possible involvement of cyclic AMP-dependent phOSphorylation of the structures immediately involved in the secretion process, i.e., secretory granules and plasma membranes. A most plausible site for cyclic AMP intervention is membrane fusion. Alterations of the granule or inner plasmalemma surface properties, such as the phosphorylation of specific sites by a protein kinase, could profoundly alter the rate of granule release. The initial observation that the terminal phosphate of (YSZP)ATP was transferred to a protein component of the zymogen granule membrane did not distinguish between the action of a protein kinase or the capture of an acyl-phosphate enzyme intermediate. Since trapping a phosphorylated intermediate of the granule Mg2+-ATPase was very possible, experiments to determine the phosphate linkage were conducted. The phosphorylated protein was stable to hydrolysis in dilute base and acid, while greater than 97% hydrolysis occurred in 1 N NaOH at 37° for 30 minutes. These results are characteristic of protein O-phosphoserine or O-phosphothreonine residues, products of a protein kinase reaction. 133 The initial rate of the protein kinase of purified granule membrane preparations was comparable to several other membrane-bound kinase activities (Guthrow et al., 1972; Wray et al., 1973; Korenman et al., 1974), and varied between 3 and 10 pmoles/min/mg protein. The rigorous washing procedure involved in the membrane isolation, including alkaline sodium bicarbonate extraction of the granules and a high salt wash of the membranes, indicates that both the protein kinase and the phosphorylated proteins were firmly associated with the granule membrane and were not adventitous components. Mitochondrial membrane, the most likely source of contamination, did not contain a protein kinase of sufficient activity to account for the observed granule membrane phosphorylation. Further investigations of the granule protein kinase require an improved assay. (Y32P)ATP hydrolysis by the potent granule Mg2+-ATPase severely limits the incubation time. Addition of a selective ATPase inhibitor such as sodium fluoride (Ueda et al., 1974) should permit the use of much less membrane protein per assay and yet increase the total 32 P04 incorporation. ADP, which acts as a competitive inhibitor of the granule membrane catalyzed (YSZP)ATP hydrolysis without affecting protein phOSphorylation, may also be employed. Assuming that the phosphorylated component was a single poly- peptide Species of 130,000 molecular weight constituting 2.6% of the granule membrane (Table l), 200 ug of membrane protein (the amount in each assay) contained approximately 40 picomoles. The 0.6 picomoles 32 of P04 transferred during the five-second incubations therefore 134 represents a very small fraction of the available sites. Long term assays are required to determine whether all the sites can be phos- phorylated. Alternatively, the membrane may be partially phosphorylated prior to isolation. A reciprocal relationship between the number of available phosphorylation sites and the rate of hormone-induced secre- tion would imply a direct role for the granule membrane protein kinase in the control of secretion. No significant cyclic nucleotide Stimulation of the granule membrane protein kinase was observed. The absence of cyclic AMP stimu- lation of several membrane-bound protein kinases has been recently incorporated into a scheme which accounts for alteration of membrane prOperties as an important physiologic response during hormone-induced increases in cyclic AMP levels (LaBrie et al., 1971; Lemay et al., 1974; Korenman et al., 1974). Upon stimulation of uterine adenylyl cyclase by isoproterenol, the number of unsaturated cyclic AMP binding sites were reduced, and the concentration of cyclic AMP independent protein kinase increased at the expense of cyclic AMP dependent activity. The appearance of the independent form of protein kinase was due to the well known cyclic AMP-induced dissociation of the protein kinase cyclic AMP-receptor/catalytic subunit complex (Krebs, 1972). Unlike the dis- tribution prior to stimulation, the kinase activity was not found in the soluble fraction of the homogenates, but instead was recovered bound to membrane. Korenman et al. (1974) proposed that the cyclic AMP-induced kinase translocation is responsible for the uterine response to B-adrenergic stimulation. Likewise, it is reasonable to postulate as a working hypothesis that in the stimulated exocrine 13S pancreas cyclic AMP induced dissociation of a cytoplasmic receptor- catalytic complex results in the binding of the active catalytic subunit to the zymogen granule surface. APPENDIX APPENDIX A PRELIMINARY STUDY OF MEMBRANE FORMATION DURING PANCREATIC DIFFERENTIATION IN THE RAT EMBRYO Abstract Differentiation of the exocrine pancreas in the rat embryo between 15 and 20 days of gestation is characterized by greatly in- creased rates of synthesis of secretory enzymes (Kemp et al., 1973). Proliferation of the rough endoplasmic reticulum occurs during this period. ZymOgen granules appear late and are prominent by 19 days of gestation. To compare these observations with membrane protein syn- thesis, maximal rates of amino acid incorporation into total protein and particulate protein, were determined by incubating 14 to 20 day old pancreatic rudiments in minimum essential medium containing 3H- leucine. If the specific activity of the leucine precursor pool is taken into account, the rate of particulate protein synthesis (nmoles/ mg protein/hr) varies about 2-fold during this interval. When expressed on a per cell basis, however, the rate of 3H-leucine incorporation increased nearly 4-fold from day 14 to day 20. 3H-leucine labeled subcellular components from 15, 16 and 20 day old rudiments were frac- tionated; the relative rate of incorporation into the zymogen granule fraction increased by approximately 70-fold. These results suggest that formation of Specialized membranes is accompanied by a com- paratively small increase in the rate of membrane protein synthesis. 136 137 Introduction Biochemical studies of the differentiative process, primarily concerned with relating the appearance and accumulation of tissue Specific proteins, have led to possible schemes of cellular control (Rutter et al., 1968b; Papaconstantinou, 1967; Palmiter et al., 1970). Morph010gical studies of this same process have described ultra- structural changes, primarily the elaboration of membrane systems and the appearance of storage and secretion granules common in many dif- ferentiated cell types (Flickinger, 1969; Mills and Topper, 1969). Few attempts have been made to define the biochemical events involved in elaboration of membrane systems and their functions during differ- entiation. Hence, the relationship between regulation of membrane formation and regulation of the synthesis and secretion of cell Specific products has not been established. The increase of functional membrane is an integral part of the differentiative process. In mature pancreas exocrine cells, secretory products are thought to be synthesized preferentially on polysomes attached to rough endoplasmic reticulum (Takagi et al., 1971; Redman, 1969). Protein glycosylation and "packaging" may be performed by the Golgi apparatus (Jamieson and Palade, 1967a; Morre et al., 1969; Schacter et al., 1970). Cell-specific products, e.g., amylase, chymotrypsinOgen, and ribonuclease, are stored and tranSported to the luminal plasma membrane by membrane-bound secretory vesicles, zymOgen granules (Jamieson and Palade, 1967a). Prior to differentiation, exocrine cells possess little rough endoplasmic reticulum, and no 138 granules. Therefore, quite extensive membrane development must occur to attain the differentiated State. Kallman and Grobstein (1964) have observed that the first detectable ultrastructural change in deve10ping mouse pancreas, the accumulation of cytoplasmic ribosomal aggregates, occurs at the beginning of the secondary transition. Elaboration of rough endo- plasmic reticulum begins one day later at the time of pronounced zymogen synthesis. The appearance of zymogen granules evidence the accumulation of secretory product three days after the initial ob- servation of the accumulation of cytoplasmic ribosomes. As yet only the rates of synthesis and accumulation of secretory products have been characterized (Sanders, 1970; Rutter et al., 1968b; Kemp et al., 1972). This report contains initial experiments designed to correlate the appearance of cell-Specific secretory products with the synthesis and assembly of components of intracellular membranes. The results have led to the insight that, to be fruitful, such a study requires the analysis of precursor incor- poration into specific, well-defined and isolatable membranes. Methods and Materials Methods Specific for each experiment are included in the figure and table legends. Sprague-Dawley rats (Spartan Research Animals, Haslett, MI) were paired in breeding cages; conception was noted in the morning by dropped vaginal plugs (day 0). After decapitation, placentas from pregnant rats were removed and placed in Earle's balanced salt solution (EBSS) (Earle, 1943). Dissected pancreas tissue equivalent 139 to approximately five 15-day rudiments (70 to 100 ug tissue protein) were cultured at 37° in a minimum of 0.2 ml minimum essential medium (MEM) (Eagle, 1959) buffered with 20 mM HEPES (the kind gift of N. Good), pH 7.0. The increase of the osmolality of the medium due to the addition of 20 mM HEPES Ivas measured with a clinical osmometer and found to be negligible. Sterile conditions were maintained. Twelve hour cultures contained penicillin (100 units/ml), streptomycin (100 ug/ml) and fungizone (0.25 ug/ml) (Grand Island Biological Company). Prior to use (4,5-3H)L:leucine (22 Ci/mmole; Amersham/Searle) was characterized using an amino acid analyzer. Although the total recovery of leucine applied to the amino acid analyzer column was low, 89% of the radioactivity recovered eluted with Erleucine added as carrier. The TCA soluble fraction from homOgenates of thoroughly washed tissues obtained after a 30-minute incubation with 3H-leucine was also analyzed and was found to contain 88% of the counts recovered as leucine. Hence, negligible conversion to other soluble molecules occurred, and the radioactivity recovered in the TCA precipitates was assumed to represent protein 3H-leucine. Radioactive samples were added to 10 ml of fluid containing (per liter) 667 m1 toluene, 333 ml Triton X—100, 5.5 gm PPO and 0.1 gm POPOP, and counted in a Packard Tri-Carb scintillation Spectrometer. Results Estimation of the Rates of Membrane SyntheSis at Different Developmental Ages by 3H-1eucine Incorporation Figure 22 illustrates the soluble and particulate protein accumulation of embryonic pancreas cells between day 15 of gestation 140 Figure 22. Changes in the soluble and particulate protein content of cells of embryonic pancreas during development. Pancreases from embryos of different ages were sonicated briefly in polyethylene microfuge tubes containing 300 pl of 0.2 M NaHCOs, pH 8.2, then centrifuged at 100,000xg for 1 hour. The pellet was washed once by brief sonication in 0.25 M NaBr and recentrifuged. Particulate protein refers to the final pellet; soluble protein refers to the combined supernatants. The distribution of particulate and soluble protein was calculated from the known total cellular protein content :assuming a 7 pg DNA per cell (Rutter et al., 1968a). 141 eoo - Profein Accumulation per Cell soo- 6 soluble zoo e .s\ 0 fi 0 L G. g D particulate :_ 100 - C—— r-O U/ :5 1'5 117 Inc I; 23 Embryonic Age [days] Figure 22 142 and birth. While soluble protein accumulates to a level of 20-1/2 days that is 19-fold greater than at 15-1/2 days, the particulate pro- tein contribution per cell appears to only double. The appearance of extensive endoplasmic reticulum and the accumulation of zymOgen granule membrane is masked by the large contribution of nuclei and mitochondria to the total particulate material. Therefore, the increase of indi- vidual intracellular membranes is more significant than implied in Figure 22. Further experiments were performed to determine the rate of synthesis of membrane components and their assembly into sub- cellular structures. Preliminary experiments determined the optimum 3H-leucine labeling conditions for 15 to 20 day pancreatic rudiments in culture. Using Earle's minimum essential medium with HEPES buffer, the pulse label was initiated by the addition of 3H-leucine to 10 uCi/ml. Under these conditions the soluble leucine pool equilibrated within 20 minutes and incorporation into cellular protein is linear after 10 minutes (Figure 23) and as long as 10 hours (data not shown). Figure 24 illustrates the incorporation of 3H-leucine under the conditions just described into particulate and non-particulate fractions of embryonic pancreas of various ages. The slope of each line is indicative of the rate of incorporation of precursor. The rates of 3H-leucine incorporation into particulate protein are initially high at 15-1/2 and 17-1/2 days of gestation, then decrease during develOpment. This initial elevated rate could be responsible for the increase of particulate protein between 17 and 19 days of gestation (Figure 22). Incorporation into the soluble fraction, which 143 Figure 23. 3H-leucine uptake and incorporation into macromolecules by embryonic pancreas cultured in_xi£rg: Pancreatic rudiments from 19 day rat embryos were cultured for varied intervals in 0.5 ml HEPES buffered MEM containing 0.4 mM leucine and 10 uCi/ml 3H-leucine. 3H-leucine incorporation was terminated by the addition of 3 m1 of ice cold EBSS. The rudiments were rinsed three times in 4 ml of £888, sonicated in 400 ul polyethylene microfuge tubes (Beckman) containing 200 ul NaHCOS, and precipitated with an equal volume of cold 10% TCA. The precipitates were collected on glass fiber filters (Whatman GC/C), washed with 5 ml of 10% TCA, placed in scintillation vials with 0.5 m1 1 N NaOH, and heated 2 hours at 100°C. Aliquots of the homogenate (——o——J and TCA soluble fraction (——Ar—J were counted as well as the TCA precipitate (——u-—J. 144 3- o al-i-I.oucino Incorporation I Into Total Cellular Protein 5.. i ‘2' 2 _ homogenate E 2 TCA procipitable E l - o .i 3 I O 1' «5‘ e O /'-‘ r A A 1 * b' A 5 ‘IO 20 3O 60 INCUBA'I'ION TIME [min.] Figure 23 14S .mucouoathSm nocfinsoo ozu on whomon :aououm oannflom muoafiom Hmcfim one on whomop aflouona opmHSUHuuom .nomomwnucooon one Amoz mN.o H: oom a“ cofluooACOm moans an ooco wocmmz we: uoHHom och .Aso: A How mxooo.oo~ we pomsMApucoo coca .N.m mm .mouzmz z N.o mo A: com mcficfiopcoo mono» omomouofia ocoaxzuoxaoa :fl admofiun woumofi:0m ono: mommouoqom vomcfih one nmonHom mm mcofluooym oHnsfiom one ouoHSUAuHom owed oouoaofiu Iomum xflucoscomDSm use mmmm mo HE e :A moefiu oops» oomcfiu ouo: mucoefiosa oak .mmmm vHoo oofl mo HE m mo cofiufiowo one xn pouo:AEAou one ocflusoauzm mcficflopcou 2m: wouommsn mmmmz a“ wouob:o:« oaoz cofiuoumow mo mxep N\H-om one .N\HIwH .N\HINH .N\HImA um moanEo Eoum muaoaaooa ofluoonocom .mcofluooum neaoaaoo moHQSHOmU opoHSUAuANnco: woo ouoHSUAuAmQ coca Coauouomhoocfl ocwosoHIrm mo mouem .vm opsmflm 146 2 [901‘] "I‘m M/flldp cumin-H 8 8 eN enamfld 7.35 us: 20.5332. 0333 20.38:: $203.05“— ldaaauu 05.2. 2033828002— 2525?! and—gov at: N V 0 34/1“? aqunat-I-I. 2 [z—Ol‘] “INOM 147 includes both the leucine intracellular pool and all soluble labeled protein, follows a similar pattern. The apparent augmented rate of particulate protein synthesis may reflect a differential response of rudiments of different ages to culture and labeling conditions rather than true changes due to the mechanisms of differentiation. If more developed rudiments are less capable of maintaining sufficient intracellular amino acid pools in culture, the rate of protein synthesis could be limited by precursor levels. In addition, the apparent rate of protein synthesis is directly related to the specific radioactivity of the precursor pool, so that differences of specific activity in these studies could alone account for the observed differences in rate. To determine the influence of these two possibilities on the apparent rates of particulate protein synthesis, pancreatic rudiments, dissected from embryos at different gestation times, were cultured for 30 minutes in HEPES buffered MEM containing 3H-leucine, and the sizes and Specific radioactivities of the intracellular leucine pool were determined. Table 14 summarizes these data. As can be seen, the size of the pool does not change significantly during the developmental interval observed. The Specific radioactivity of the pool, however, declines at day 20 to one-third the value it was at day 14. The corrected rates of protein synthesis for l4, l7 and 20 day rudiments can be calculated from the relative rates of incorporation (Figure 24; expressed as dpm/ug cell protein) and the specific radio- activity of intracellular leucine (Table 14; expressed as dpm/uumole leucine). The rates thus calculated (last column, Table 14) indicate 148 Table 14. Rates of 3H-leucine incorporation into particulate fractions of pancreatic rudiments. Rudiments of different ages were incubated in HEPES buffered MEM containing 0.4 mM 3H-leucine (10 uCi/ml) for 30 minutes. The tissues were then collected, washed three times with EBSS, sonicated, assayed for protein content, and precipitated with 10% TCA. The TCA soluble fraction was exhaustively extracted with ether. The aqueous phase was then evaporated under vacuum to dryness, redissolved in citrate buffer, and subjected to amino acid analysis. Specific radioactivity of the leucine pool was calculated from the radioactivity applied to the column and the nmoles of leucine calculated from the chromatogram peak heights relative to prchlorophenylalanine. Eighty-eight per cent of the radioactivity detected in the eluate was confined to the leucine peak. TCA Embryonic Age soluble Specific Rate (days) leucine activity (nmole/mg (nmole per (dpm/uumole) protein/hr) mg protein) 14 7.0(i1.0)a 64 10 15 8.3(i0.1)a - _ 17 12.7b 42 13 20 10.8b 23 6 aTwo determinations. bA single determination- 149 that the rate of protein synthesis and subsequent assembly in a membrane structure does not change appreciably during or prior to the accumulation of intracellular membrane. Figure 25 contrasts the results obtained when intracellular and extracellular leucine specific radioactivities are used to cal- culate the rates. The differences between Figures 25A and 25B can be attributed to the dramatic increase in protein per cell during development (Figure 22). As differentiating acinar cells increase in size and membrane content, their rate of membrane protein synthesis increases nearly 4-fold (Figure 25B). 3H-Leucine Labeling of Individual Subcellular Fractions During Development--The Appearance of a Distinct Membrane Glass Table 15 gives the results of an experiment designed to measure the appearance of zymogen granules in developing cells. At 15 and 16 days the extent of synthesis of zymOgen granule protein is less than 0.1% of the total cell protein synthesized; this radio- activity may actually represent a small contamination by mitochondria. At 19-1/2 days, however, more than 7% of protein synthesized is incorporated into zymogen granules. This estimate of incorporation into zymOgen granules is probably low by a factor of 3, since only 25% of the total granule population was isolated in this fraction. These results corroborate electron microscope observations (Kallman and Grobstein, 1964; Pictet et al., 1972) and suggest that the appearance of zymogen granules may result in a 100-fold increase in the rate of synthesis of the zymogen granule membrane. 150 .Aewcmev .ee eo eooosm Eoew coxoe mo: momm ecouomweo mo mmonocom now :Aoeoem seemeaflee pom mHHoo mo Hones: one .ceoeoom we pom cone honemp mHAoo ooH pom vommoumxo ohm .m< :A we poeoasofioo .moeom Am .flem oeswem Eoem :oxoe moeoe o>eemeoev mocflo:oa oeoe :1\Emv mmv esflpos coeeon:oce one we xee>eeom0eeoe oemfioomm oceosoa one Eoem woeofisoeoo moemu nIIOII “AVA oaboev xefl>fieoo0ebmn oewfloomm oneoooa Hmasaaooeeece Eoem woemHSUHoo moemh nllnfill m< .memonecxm aeoeoem oeoasoeeeom mo moeee poeoaooaoo one we someeemsou .mm oesmem 151 mN cosmea Zea N04 0.2955 0N m. o9 e. o_N m. o. e. b _ sues 9QI/lq/sauoulu :3 M78 (\l _ i 1 IT) I co_:._ooo.=c_.n. m 63:30:36.0. < T q. 1 uIeIOJd bw%q/se|owu Igiva 1 SE 152 Table 15. Distribution of 3H-leucine incorporation into subcellular fractions of embryonic pancreas. Fifteen, 16 and 19 1/2 day pancreatic rudiments were incu- bated for 10 hours in HEPES buffered MEM containing 0.02 mM leucine (10 uCi/ml). After rinsing in cold EBSS, the tissues were homogenized with a small glass-teflon Potter-Elvehjem homogenizer, combined with carrier homogenate from a single adult rat pancreas, and fractionated into five subcellular fractions by differential centrifugation (Methods). Percent total dpm incorporated Fraction Embryonic age: 15 day 16 day 19-1/2 day Debris 20 18 16 Zymogen granules 0.05 0.08 7 Mitochondria ll 11 12 Microsomes 21 19 9 100,000xg supernatant 48 52 56 153 Discussion The transitions which occur within a presumptive pancreatic acinar cell during differentiation are dramatic. In order to compare the rates of protein synthesis at individual time points in a con- tinuously changing system, several criteria must be fulfilled. For all developmental Stages (a) the intracellular pool of labeled pre- cursor muSt attain a steady state level quickly, relative to the total labeling period, (b) the steady state conditions must be main- tained during the labeling period, (c) incorporation into protein must be linear throughout the labeling period, and (d) the specific activity of the intracellular pools must be monitored, and variations observed must be incorporated into calculations to obtain more accurate rates of protein synthesis. Requirements (a), (b) and (c) were fulfilled under the con- ditions employed for pancreatic rudiments of embryos from 15 through 20 days of gestation. In the absence of information on the specific activity of intracellular leucine, the relative rates of pancreatic membrane protein synthesis appeared uniformly to decrease from 15-1/2 days of gestation until term. When the rates of synthesis of par- ticulate protein were corrected from the observed relative rates using the measured Specific activity of the intracellular leucine, the rates appear more constant, fluctuating only Z-fold. The rates of membrane assembly monitored independently by 3H-choline incorporation gave Similar results (preliminary results, data not given). These results of leucine and choline incorporation studies suggest that the increase of total embryonic pancreatic membranes, 154 which is particularly Significant for rough endoplasmic reticulum and zymOgen granule membranes (Pictet et al., 1972), may involve only a slight increase in the rate of synthesis of membrane protein and lipid. It Should be noted, however, that since acinar cell develop- ment results in an approximate lO-fold increase in protein per cell, that there was up to a 4-fold increase in the rate of membrane protein synthesis when expressed on a per cell basis. The Steady state leucine pool Specific activity in cultured l4-day embryonic pancreas was approximately the same as the extra- cellular leucine of the labeling medium. The steady state specific activities of more developed pancreatic rudiments were significantly lower. This observed decrease may reflect the presence of two dis- tinct intracellular amino acid pools (Righetti et al., 1971; Mortimore et al., 1972) and either the relative changes which occur between them during development, or differential effects of culture conditions on the two pools at the different stages of development. The diffi- culties encountered in interpretation of results which reflect temporal changes in the contribution by two independent precursor pools may be experimentally overcome. Since the Size, and therefore influence, of the pool in equilibrium with extracellular leucine can be enhanced several-fold by increasing the external leucine concentration, the specific activity of the total internal leucine pool can be made to reflect that of external leucine by increasing the concentration of leucine in the medium above 2 mM (R. J. MacDonald, unpublished observation; Mortimore et al., 1972). 155 Analysis of pancreatic development by culturing l4-day pancreatic rudiments has been shown to mirror the in vivo rates of synthesis and accumulation of the tissue Specific hydrolytic enzymes (Rutter et al., 1968b; Kemp et al., 1972). Parallel changes in other important parameters of exocrine development are not observed, how— ever. The extent of cell protein accumulation is low, presumably due to enhanced mesenchymal tissue growth of cells with a low protein to DNA ratio (Kemp et al., 1972). Similarly, neither the apparent changes in the rate of 3H-leucine incorporation nor the decrease in specific activity of the leucine intracellular pool is observed for rudiments in extended culture. The utilization of freshly dissected embryonic pancreas to monitor rates of macromolecular synthesis at distinct developmental stages and the use of a continuous culture system during which development occurs must be viewed as two distinct experimental approaches to the same system, and caution must be exercised in the comparative analyses of their results. The appearance of zymOgen granules prior to 19-1/2 days gestation (Pictet et al., 1973; Kallman and Grobstein, 1964; and Table 15) clearly indicates the emergence of a functionally distinct membrane species as indicative of differentiation as the appearance of pancreatic digestive enzymes and pro-enzymes. The concept that the zymogen granule membrane represents part of the differentiated characteristics of an acinar cell (i.e., part of the mechanism of secretion of tissue specific products) has led to the premise that the rate of synthesis and assembly of zymogen granule membrane may be employed as another index of pancreatic development. An analysis 156 of the rate of assembly would be free of many of the ambiguities of investigating the synthesis of bulk intracellular membranes. Analysis of several physical and enzymatic parameters of the zymOgen granule membrane for the purpose of identifying distinctive characteristics useful in measuring rates of synthesis and assembly during development formed the basis of the studies reported in the main text of this thesis. LIST OF REFERENCES LIST OF REFERENCES Amsterdam, A., M. Schramm, I. Ohad, Y. Salomon and Z. Selinger (1971) J. Cell Biol. 59, 187. Atkinson, P. H., and D. F. Summers (1971) J. Biol. Chem. 246, 5162. Avruch, J., and G. Fairbanks (1972) Proc. Nat. Acad. Sci. USA 99, 1216. Banks, 9. (1966) Biochem. J. 101, 18c. Baudin, H., M. Colin and J. E. Dumont (1969) Biochim. BiOphys. Acta 174, 722. Baudin, H., L. Rochus, D. Vincent and J. E. Dumont (1971) Biochim. Biophys. Acta 252, 171. Bdolah, A., and M. Schramm (1965) Biochem. Biophys. Res. Commun. 18, 452. Breckenridge, W. C., and I. B. Morgan (1972) Fed. Eur. Biochem. Soc. Lett. 23, 253. Bretscher, M. S. (1971) Nature New Biol. 231, 229. Burwen, S. J. and S. S. Rothman (1972) Amer. J. Physiol. 222, 1177. Butcher, R. W., C. E. Baird and E. W. Sutherland (1968) J. Biol. Chem. 243, 1705. Caro, L. G. and G. E. Palade (1964) J. Cell Biol. 29! 473. Carraway, K. L., and D. Kobylka (1970) Biochim. Biophys. Acta 219, 238. DeCrombrugghe, B., R. Pitt-Rivers and H. Edelhoch (1966) J. Biol. Chem. 241, 2766. Eagle, H. (1959) Science 130, 432. 157 158 Earle, W. R. (1943) Natl. Cancer Inst. 4, 167. Exton, J. H., J. G. Hardman, T. F. Williams, E. W. Sutherland and C. R. Park (1971) J. Biol. Chem. 246, 2658. Fairbanks, G. (1969) Ph.D. dissertation, Massachusetts Institute of Technology, Cambridge, Mass. Fairbanks, G., and J. Avruch (1972) J. of Supramol. Struct. l, 66. Fairbanks, C., T. L. Steck and D. F. H. Wallach (1970) Biochemistry 10, 2606. Flickinger, C. J. (1969) J. Ultrastruct. Res. 32, 344. Gahmberg, C. G. (1971) Biochim. BiOphys. Acta.2£9, 81. Glossmann H., and D. M. Neville, Jr. (1971) J. Biol. Chem. 239, 6339. Greene, L. J., C.H.W. Hirs, G. E. Palade (1963) 238, 2054. Guthrow, C. E., J. E. Allen and H. Rasmussen (1972) J. Biol. Chem. 247, 8145. Hatefi, Y., and W. Hanstein (1969) Proc. Nat. Acad. Sci. USA 93, 1129. Heisler, S., D. Fast and A. Tenenhouse (1972) Biochim. Biophys. Acta 279, 561. Helgeland, L., T. B. Christensen and T. L. Janson (1972) Biochim. Biophys. Acta 286, 62. Hicks, 8. J., J. W. Drysdale and H. N. Munro (1969) Science 164, S84. Hittelman, K. J., and R. W. Butcher (1971) in "Effects of Drugs on Cellular Control Mechanisms (B. R. Rubin and R. B. Freedman, eds.) p. 153, MacMillan, London. Hokfelt, T. (1968) Z. Zellforsch 91, l. Hokin, L. E. (1955) Biochim. Biophys. Acta 18, 379. Hokin, L. E., P. S. Sastry, P. R. Galsworthy and A. Yoda (1965) Proc. Nat. Acad. Sci. USA 53, 177. Holmsen, H., H. J. Day and H. Stormorken (1969) Scand. J. Haemat., Suppl. 2. Hoober, J. K. (1970) J. Biol. Chem. 245, 4327. Hortnagl, H., H. Winkler, J. A. L. Schopf and W. Hokenwallner (1971) Biochem. J. 122, 299. 159 Howell, S. L., D. A. Young and P. E. Lacy (1969) J. Cell Biol. g}, 167. Hymer, W. C., and W. H. McShan (1963) J. Cell Biol. 12, 67. Jamieson, J. D., and G. E. Palade (1967a) J. Cell Biol. 35, 577. LA Jamieson, D., and G. E. Palade (1967b) J. Cell Biol. 34, 597. Jamieson, J. D., and G. E. Palade (1968a) J. Cell Biol. 39, 580. Jamieson, J. D., and G. E. Palade (1968b) J. Cell Biol. 32, 589. Jamieson, J. D., and G. E. Palade (1971) J. Cell Biol. 58, 503. Johnson, J. C., M. DeBacker and J. A. Boezi (1971) J. Biol. Chem. 246, 1222. Johnson, M. H. S. A. Sheratt, R. M. Case and T. Scratchard (1970) Biochem. J. 120, SP. Kadoto, K., S. Mori and R. Imaizumi (1967) J. Biochem. Tokoyo 61, 424. Kallman, F., and C. Grobstein (1964) J. Cell Biol. 22, 399. Keller, P. J., and E. Cohen (1961) J. Biol. Chem. 236, 1407. Kemp, J. D., B. T. Walther and W. J. Rutter (1972) J. Biol. Chem. 247, 3941. Kiehn, E. D., and J. J. Holland (1970) Biochemistry g, 1716. Klotz, I. M., and D. W. Darnall (1969) Science 166, 126. Korenman, S. C., R. C. Ghalla, B. M. Sanborn and R. H. Stevens (1974) Science 183, 432. Krebs, E. G. (1972) Current Topics in Cellular Regulation 5, 99. Kulka, R. C., and E. Sternlicht (1968) Proc. Nat. Acad. Sci. USA 91, 1123. Kuo, J. F., and P. Greengard (1969) Proc. Nat. Acad. Sci. USA 93, 1349. LaBrie, F., S. Lemaire, G. Poirier, G. Pelletier and R. Boucher (1971) J. Biol. Chem. 246, 7311. Laemmli, U. K. (1970) Nature 227, 680. 160 Lambert, M., J. Cama and J. CristOphe (1973) Biochem. Biophys. Res. Commun. 52, 935. Larsen, C., G. Dallner and L. Ernster (1972) Biochem. BiOphys. Res. Commun. 15, 1300. Lemay, A., M. Deschenes, S. Lemaire, L. Poulin and F. LaBrie (1974) J. Biol. Chem. 249, 323. Lenard, J. (1970) Biochemistry 5, 1129. Levitski, A., and D. E. Koshland (1969) Proc. Nat. Acad. Sci. USA E, 1121. Lindberg, 0., and L. Ernster (1956) Methods Biochem. Anal. 5, 1. Leoning, U. (1967) Biochem. J. 593, 251. London, W. P., and T. L. Steck (1969) Biochemistry 5, 1767. Lopez, R. M. Y., and P. Siekevitz (1973) Anal. Biochem. 55, 594. Lowry, O. H., N. J. Rosebrough, A. L. Farr and R. J. Randall (1951) J. Biol. Chem. 193, 265. MacDonald, R. J., and R. A. Ronzio (1974) Fed. Eur. Biochem. Soc. Lett. 52. 203- Maizel, J. V., Jr. (1966) Science 151, 988. Maizel, J. V., Jr. (1971) in "Methods in Virology (K. Maramorosch and H. Koprowski, eds.) Vol. 5, p. 179, Academic Press, New York. Malik, M., and A. Berrie (1972) Anal. Biochem. 55, 176. Mathews, E. K. (1970) in "Calcium and Cellular Function (A. W. Cuthbert, ed.) p. 103, St. Martin's Press, New York. Mathews, E. K., R. J. Evans and P. M. Dean (1972) Biochem. J. 130, 825. Meldolesi, J., and D. Cova (1972) J. Cell Biol. 55, 1. Meldolesi, J., J. D. Jamieson and G. E. Palade (1971a) J. Cell Biol. 49, 109. Meldolesi, J., J. D. Jamieson and G. E. Palade (1971b) J. Cell Biol. 49, 130. 161 Meldolesi, J., J. D. Jamieson and G. E. Palade (1971c) J. Cell Biol. 55, 150. Mills, E. S., and Y. J. Topper (1969) J. Cell Biol. 31, 310. Morre, D. J., L. M. Merlin and T. W. Keenan (1969) Biochem. BiOphys. Res. Commun. 51, 813. Mortimore, G. E., K. H. Woodside and J. E. Henry (1972) J. Biol. Chem. 247, 2776. Mostafa, I. Y., E. Fallin and S. C. Fang (1970) Anal. Biochem. 55, 238. Palmiter, R. D. (1972) J. Biol. Chem. 247, 6450. Palmiter, R. D., A. K. Christensen and R. T. Schimke (1970) J. Biol. Chem. 245, 833. Papaconstantinou, J. (1967) Science 156, 338. Penefsky, H. S. (1967) Methods Enzymol. 59, 702. Penman, S. (1966) J. Mol. Biol. 12, 117. Pictet, R. L., W. R. Clark, R. H. Williams and W. J. Rutter (1972) Devel. Biol. 35, 436. Pletscher, A., K. H. Berneis and M. DaPrada (1970) Advan. Biochem. Psychopharmacol. 3, 205. Plimmer, R. H. A. and W. M. Bayliss (1906) J. Physiol. (Lon.) .55, 439. Poisner, A. M., and W. W. Douglas (1968) Mol. Pharmacol. 5, S31. Poisner, A. M., and J. M. Trifaro (1967) Mol. Pharmacol. 5, 561. Poste, G., and A. C. Allison (1969) J. Theor. Biol. 53, 165. Poste, G., and A. C. Allison (1974) Biochim. Biophys. Acta 599, 421. Rasmussen, H., D. B. P. Goodman and A. Tenenhouse (1972) CRC Critical Reviews in Biochemistry l, 95. Ray, T. K., and G. V. Marinetti (1971) Biochim. Biophys. Acta 233, 787. Redman, C. M. (1967) J. Biol. Chem. 242, 761. Redman, c. M. (1969) J. Biol. Chem. 244, 4308. 162 Redman, C. M., and D. D. Sabatini (1966) Proc. Nat. Acad. Sci. USA 55, 608. Reid, M. S., and N. L. Bieleski (1968) Anal. Biochem. 33, 374. Reynolds, J. A., and C. Tanford (1970a) Proc. Nat. Acad. Sci. USA 55, 1002. Reynolds, J. A., and C. Tanford (1970b) J. Biol. Chem. 245, 5161. Rice, R. H., and G. E. Means (1971) J. Biol. Chem. 246, 831. Richardson, C. C. (1971) in "Procedures in Nucleic Acid Research" (G. L. Cantoni and D. R. Davies, eds.) Vol. 2, p. 815, Harper and Rowe, New York. Ridderstap, A. S. and S. L. Bonting (1969) Pflugers Arch. 313, 62. Righetti, P., E. P. Little and G. Wolf (1971) J. Biol. Chem. 246, 5724. Riley, W. D., R. D. DeLange, G. E. Bratvold and E. G. Krebs (1968) J. Biol. Chem. 243, 2209. Robinson, J. D. (1969) Biochemistry 5, 3348. Ronzio, R. A. (1973a) Biochim. Biophys. Acta 313, 286. Ronzio, R. A. (1973b) Arch. Biochem. Biopyys. 159, 777. Rosenberg, R. M., H. L. Crespi and J. J. Katz (1969) Biochim. Biophys. Acta 175, 31. Rubin, C. S., J. Erlichman and O. M. Rosen (1972) J. Biol. Chem. 247, 6135. Russ, G., and K. Polakova (1973) Biochem. Biophys. Res. Commun. 55, 666. Rutter, W. J., W. D. Ball, W. S. Bradshaw, J. D. Kemp and B. Walther (1968a) unpublished observations. Rutter, W. J., J. D. Kemp, W. S. Bradshaw, W. R. Clark, R. A. Ronzio and T. G. Sanders (1968b) J. Cell. Physiol. 23, Suppl. 1, l. Sabatini, D. D., and G. Blobel (1970) J. Cell Biol. 55, 146. Sanders, T. G. (1970) Ph.D. dissertation, University of Illinois. Sanders, T. G., and W. J. Rutter (1972) Biochemistry 11, 130. Schachter, H., I. Jabbal, R. L. Hudgin and L. Pinteric (1970) J. Biol. Chem. 245, 1090. 163 Schramm, M., and D. Danon (1961) Biochim. Biophys. Acta 59, 102. Schramm, M., and E. Naim (1970) J. Biol. Chem. 245, 3225. Segrest, J. P., R. L. Jackson, E. P. Andrews and V. T. Marchesi (1971) Biochem. Biophys. Res. Commun. 33, 390. Simons, K., and L. Kaariainen (1970) Biochem. Biophys. Res. Commun. 38, 981. Smith, B. W., and J. H. Rowe (1949) J. Biol. Chem. 179, 53. Smith, M. H. (1968) in "CRC Handbook of Biochemistry (H. A. Sober, ed.) p. C-3, Chemical Rubber Co., Cleveland. Spatz, L., and P. Strittmatter (1971) Proc. Nat. Acad. Sci. USA 55, 1042. Spatz, L., and P. Strittmatter (1973) J. Biol. Chem. 248, 793. Stormorken, H. (1969) Scand. J. Haemat., Suppl. 9. Studier, F. W. (1973) J. Mol. Biol. 15, 237. Sutherland, E. W., I. Oye and R. W. Butcher (1965) Recent Prog. Harmon. Res. 31, 623. Sutherland, E. W., and G. A. Robison (1966) Pharmacol. Rev. 15, 145. Swank, R. T., G. I. Sheir and K. D. Munkres (1971) Biochemistry 19, 3924. Takagi, M., and K. Ogata (1971) Biochem. Biophys. Res. Commun. 53, 125. Takayama, K., D. H. MacLennan, A. Tzagoloff and C. D. Stoner (1966) Arch. Biochem. Biophys. 114, 223. Tani, H., and K. Ogata (1970) Biochim. Biophys. Acta 215, 264. Teipel, J., and D. E. Koshland (1969) Biochemistry 5, 4656. Trifaro, J. M., and J. Dworkind (1971) Mol. Pharmacol. Z, 52. Trifaro, J. M., and M. Warner (1972) Mol. Pharmacol. 5, 159. Ueda, T., H., Maeno and P. Greengard (1974) J. Biol. Chem. 248, 8295. Warren, L. (1959) J. Biol. Chem. 234, 1971. 164 Weber, K., and M. Osborn (1969) J. Biol. Chem. 244, 4406. Wharton, D. C., and A. Tzagoloff (1967) Methods Enzymol. $2, 245. Whittaker, V. P. (1965) Progr. Biophys. Mol. Biol. 55, 39. Winkler, H., H. Hortnagl, H. Hortnagl and A. D. Smith (1970) Biochem. J. 118, 303. Woodin, A. M. (1968) in "The Biological Basis of Medicine" (E. E. Bittar and N. Bittar, eds.) Vol. 2, p. 373, Academic Press, New York. Weadin, A. M., and A. A. Wieneke (1963) Biochem. J. 51, 487. Woodin, A. M., and A. A. Wieneke (1970) in "Calcium and Cellular Function" (A. W. Cuthbert, ed.) p. 184, St. Martin's Press, New York. Wray, H. L., R. Gray and R. A. Olsson (1973) J. Biol. Chem. 248, 1496. Zahler, W. L., B. Fleischer and S. Fleischer (1970) Biochim. Biophys. Acta 203, 283. "111111111IIIIIIIIIIIT