STUDIES ON THE RESOLUTEON 0F MECROSOMAL MEMBRANE PROTEINS Thesis for the Degree of M. S. MECHIGAN STATE UNWERSITY FREDRIOK OLIVER O’NEAL 1969 ‘ : . V“ *Wfi‘»! L I P "” Micki}; , E Unlvumt) THESIS Ii ‘7 mail!“ av HMS & SONS’ 300K BINDERY '"V 1, IIIRARY B'Nl' I mm ; a I l ABSTRACT STUDIES ON THE RESOLUTION OF MICROSOMAL MEMBRANE PROTEINS BY Fredrick Oliver O'Neal The effects of treatments, such as organic solvent and salt extractions, lysis in the presence of 2.76% glycerol, and washing with buffer containing 0.01 M EDTA and 0.15bdKCl, on the resolution of microsomal membrane proteins were investigated. The proteins extracted by these treatments were identified by polyacrylamide disc gel electrOphoresis and found to be involved in either hydrophobic or electrostatic (e.g. ribosomal proteins) interactions with the microsomal membrane. A relationship between the extent of protein solubilization and the A polarity of the organic solvents used in the extractions was observed. Similarly, a relationship was observed be- tween protein solubilization and the extent to which salts containing chaotropic ions (i.e., those which favor the transfer of apolar groups to water such as SCN-, Br—, NQ3_, and Cl-) changed the structure and lipophilicity of water. The harshness of these treatments (lysis, EDTA-KCl wash, salt extractions, and freezing in the presence of 50% Fredrick Oliver O'Neal glycerol and 0.25bdsucrose) was evaluated by determining their effects on NADPH-cytochrome C reductase and amino— pyrine demethylase activity. Protein fractions were isolated from microsomes which may correspond in function to the apparently non- catalytic, structural proteins of mitochondria. One fraction is that isolated by detergent treatments and classically termed "Structural Protein"; the other, identi- fied here by its being the predominate protein species in the electrophoresis profile of microsomal membranes, termed "Core protein." The former is believed to be functionally identical to the structure determining components of the headpieces and the latter that of the basepieces of mito- chondrial membranes. These findings give support to more recently accepted membrane model in which membranes are thought to be composed of lipoprotein repeating subunits. STUDIES ON THE RESOLUTION OF MICROSOMAL MEMBRANE PROTEINS BY Fredrick Oliver O'Neal A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biochemistry 1969 457373 /Z).22 éf7 ACKNOWLEDGMENTS The author wishes to thank Dr. Steven D. Aust for his guidance and encouragement throughout the completion of this work. Also the author would like to thank Drs. Deal, Suelter, and Wilson and members of the Aust Lab for their helpful suggestions and/or use of equipment. ii TABLE OF CONTENTS Page ACKNOWLEDGMENTS . . . . . . . . . . . ii LIST OF TABLES . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . viii LIST OF ABBREVIATIONS . . . . . . . . . ix INTRODUCTION 0 O O O O O O O O O O O 1 LITERATURE REVIEW . . . . . . . . . . 5 Membranes . . . . . . . . . . . . 5 Structural Protein . . . . . . . . . 9 METHODS AND MATERIALS . . . . . . . . . 15 Chemicals . . . . . . . . . . . . 15 Animals. . . . . . . . . . . . . 15 Isolation of Microsomes . . . . . . . l6 MethOd 1 O O O O O O O O O 0 0 16 Method 2 . . . . . . . . . . . l7 Lipid and RNA Extraction . . . . . . . 1? Isolation of Structural Protein. . . . . 18 Microsomal Fractionation . . . . . . . l8 Fractionation of Microsomes with 0.26% Sodium Deoxycholate (DOC). . . . . 18 Fractionation of Microsomes on a Dis- continuous Sucrose Density Gradient in the Presence of CsCl and MgCl2 . . l9 Isolation of Ribosomes. . . . . . . 20 Fractionation of Microsomes with Tertiary-Amyl Alcohol--Isolation of Electron Transport Membranes. . . . 21 iii Page Microsomal Extraction Procedures . . . . . 21 1.4% Acetic Acid Extraction. . . . . . 21 Lysis of Microsomal Vesicles . . . . . 22 Salt Extractions of Microsomes. . . . . 23 Organic Solvent Extractions. . . . . . 23 Sonic Oscillation of Microsomes . . . . 24 Enzyme Assays . . . . . . . . . . . 24 AminOpyrine Demethylase . . . . . . . 24 NADPH-Cytochrome c Reductase . . . . . 25 Analytical Methods. . . . . . . . . . 25 Polyacrylamide Gel Electrophoresis . . . . 26 Structural Protein Assay. . . . . . . . 27 RESULTS 0 O O O O O O O O O O O O O 2 8 Fractionation of Microsomes. . . . . . . 28 Fractionation with 0.26% DOC . . . . . 28 Fractionation on Discontinuous Sucrose Density Gradient in the Presence of Cs andng++ Ions . . . . . . . . . y 29 Tertiary-Amyl Alcohol Treatment of Microsomes . . . . . . . . . . 33 Microsome Extraction Procedures . . . . . 33 Organic Solvent Extractions. . . . . . 33 Lysis of Rat Liver Microsomes . . . . . 36 Sonic Oscillation of Beef Liver Microsomes. 37 Acetic Acid (1.4% v/v) Extraction. . . . 37 Salt Extractions of Beef and Rat Liver Microsomes . . . . . . . . . . 43 EDTA-KCl Extraction of Microsomes. . . . 55 Resolution of Rat Liver Microsomal Membrane Proteins by Combinations of Treatments . . 61 Isolation of Structural Protein from Control and Phenobarbital Treated Rats. . . . . 63 Effects of Various Treatments and Extraction Procedures on the Enzymatic Activities of Microsomes . . . . . . . . . . . 66 iv Page Effects of Freezing and Lysis on NADPH- Cytochrome c Reductase and Amino- pyrine Demethylase Activity of Micro- SOmeS o o o o o o o o o O o 0 66 EDTA-KCl Wash: Effects on Enzymatic Activities of Microsomes. . . . . . 67 Salt Extractions: Effects on Enzymatic Activities of Microsomes. . . . . . 70 DISCUSSION . . . . . . . . . . . . . 75 Fractionation of Microsomes. . . . . . . 75 Fractionation on a Discontinuous Sucrose Gradient in the Presence of Cs+ and Mg++ Ions . . . . . . . . . . . . 75 Tertiary-Amyl Alcohol Treatment . . . . 76 Extraction Procedures. . . . . . . . . 77 Organic Solvent Extraction . . . . . . 77 Lysis of Rat Liver Microsomes . . . . . 79 Acetic Acid Extraction and Structural Protein (SP) Isolation . . . . . . 82 Salt Extractions . . . . . . . . . 86 EDTA-KCl Extraction . . . . . . . . 92 Combined Treatments of Microsomes. . . . 94 Effects of Various Treatments and Extraction Procedures on the Enzymatic Activities of Microsomes . . . . . . . . . . . 95 Effects of Freezing, Lysis, and EDTA-KCl waShing O O O O O O O O O O O 95 Salt Extractions: Effects on Enzymatic Activities of Microsomes. . . . . . 97 SUMMARY 0 O O O O O O 0 O O O O O O 101 BIBLIOGRAPHY. O O O O O O O O O O O O 106 Table 10. 11. LIST OF TABLES Analyses of Smooth and Rough Endoplasmic Reticulum (SER and RER) from the Livers of Phenobarbital Treated Rats . . . . Solubilization of Beef Liver Microsomal Protein by Organic Solvents in the Pre- sence of 0.1 M MgC12 . . . . . . . Quantitation of the Effects of Sonication on Beef Liver Microsomal Protein Solu- bilization . . . . . . . . . . Analyses of Acetic Acid and EDTA Extracted Rat Liver Microsomal Membranes. . . . Analyses of Beef Liver Microsomes Following Extractions with Various Salts. . . . Analyses of Beef Liver Microsomal Membranes Extracted with High Concentrations of Various Salts . . . . . . . . . Analysis of Rat Liver Microsomal Membranes Extracted with Various Concentrations of salts O O O O O O O O O O O 0 Analysis of Tris-HCl Buffer, KCl, and EDTA Extracted Rat Liver Microsomal Mem- branes. . . . . . . . . . . . Analysis of Rat Liver Microsomes During Membrane Purifications . . . . . . Effects of Freezing and Lysis on the Activities of Microsomal Enzymes . . . The Effects of EDTA-KCl Washing on Microsomal Enzymes. . . . . . . . vi Page 29 35 41 42 44 45 53 57 62 68 69 Table Page 12. The Effects of Salts on NADPH—Cytochrome c Reductase Activity . . . . . . . 72 13. The Effects of Salts on Aminopyrine Demethylase Activity . . . . . . . 73 vii LIST OF FIGURES Figure Page 1. Fractionation of Microsomes from Pheno- barbital Treated Rats with Mg++ and Cs+ Ions . . . . . . . . . . . 32 2. EDTA-KCl and 1.4% Acetic Acid Extractions of Rat Liver Microsomes . . . . . . 40 3. Salt Extractions of Beef Liver Microsomes . 48 4. Effect of Increasing Salt Concentrations on the Extraction of Proteins from Micro- somes of Phenobarbital Treated Rats . . 50 5. Extraction of Liver Microsomes from Control Rats with Salts Containing Polyvalent Anions. O O O O O O O O O O O 52 6. EDTA-KCl Wash of Rat Liver Microsomes (Phenobarbital Treated Rats) . . . . 59 7. Isolation of SP from Control and Pheno- barbital Induced Rat Liver Microsomes . 65 viii ATPase CP DOC EDTA ER HBHM HBLM NADPH RER RNA-ribose SER SER I SER II SP Tris-HCl buffer LIST OF ABBREVIATIONS Adenosine Triphosphatase Core protein Sodium deoxycholate Ethylenediaminetetraacetic acid Endoplasmic reticulum Heavy beef heart mitochondria Heavy beef liver mitochondria Nicotine adenine dinucleotide phosphate, reduced Rough endoplasmic reticulum Ribose determined from the hydrolysate of RNA Smooth endOplasmic reticulum Mg++ binding subfraction of SER Non-Mg binding subfractions of SER Structural protein 0.05 M Tris-hydroxymethylaminomethane, pH 7.5 ix INTRODUCTION The study of biological membranes has been an intense field of research for a number of years. Attempts to define these membranes with respect to function and morphology have led scientists through the years to propose models, each of which was representative of the data available. One such model which has been widely accepted was that proposed by Danielli and Davson (l) and later extended by Robertson (2), which emphasized the function of biological membranes as being passive barriers to free diffusion and electrical insulators. This model was pictured as having an interior bimolecular leaflet of phospholipid, held together by Van der Waals attractions between apolar regions of phospholipid, with proteins relegated to the exterior interacting with the polar groups of phospholipid. Evidence, however, has been accumulating within the past few years which disproves the assumptions upon which the classical model was based, sug- gesting a model in which membranes are represented as expressions of macromolecular lipOprotein subunits (3, 4). These evidences have been reviewed (5, 6, 7, 8) and will not be enumerated here; however, mention should be made of the pioneering work of Fernandez-Moran (9) whose negative staining techniques for electron microscopy provided con- crete evidence for the subunit structure in membranes. It is interesting that Green gg_al. (10) were able to isolate a protein fraction from mitochondria by the use of bile salts and the detergent sodium dodecyl sul- fate with properties (10, ll, 12) which makes it plausible to designate it as being the repeating unit in membranes. Specifically, this protein possessed the following proper- ties: insolubility at physiological pH; the ability to form stable complexes with itself, with enzymes indigenous to the mitochondria, e.g. cytochromes a, b, and c, and with lipid; an apparent noncatalytic function; and it is the predominate protein species in the mitochondrion. In like manner proteins with similar properties have been isolated from other membrane systems (l3, l4, 15), indicating the existence of a universal protein or class of proteins which function as the structural proteins of membranes. Attempts to study the properties of structural proteins have led to the discovery that the isolation procedures of Green et_al. (10) or its more pOpular modification by Richardson (13) yield a grossly heterogenous protein frac- tion, as was determined by the electrophoresis system of Takayama (16). Recently, however, purifications of this crude structural protein has been achieved by Lenaz et a1. (17) with acidic methanol extractions followed by extrac- tions with 8 M urea (pH 5.5). Though effective, these methods were shown to yield a damaged protein species (18). It is unfortunate that the technology for resolving mem- brane proteins has not advanced to the level where such proteins can be rendered pure and in a state analogous to that in_zizg. Accurate characterization of the physio- logical and functional properties of membrane proteins, whether their function is enzymatic or structural, demands the use of milder techniques for membrane resolution. It was with this premise in mind that the present study on the resolution of liver endoplasmic reticulum (ER) mem- branes into structural and other protein fractions was undertaken. Attempts have been made during this study to resolve liver ER membranes by the use of organic solvent extractions and salt extractions. Of these methods, salt extractions of ER vesicles, following lysis to remove their soluble con- tents, were more promising because of their relative mildness and in some cases specificity with respect to protein solu- bilization. The structural protein content of the various membrane protein fractions was determined by its electro- phoretic mobility on the polyacrylamide gel electrophoresis system of Takayama (16). Identification of structural protein as being the most abundant membrane protein species in the electrophoresis profile of whole membranes can be rationalized because of the fact that it was found, as isolated by Green et al. (10), to represent approximately 55% of the particulate protein of mitochondrial membranes. It must be mentioned, however, that designation of that protein fraction, possessing such prOperties, as structural protein may be erroneous in that a structural function has yet to be demonstrated. In working with endoplasmic reticulum membranes one has to be concerned about contamination by ribosomes which in themselves contain a rather heterogenous class of pro- teins. Since difficulties were encountered in obtaining sufficient quantities of smooth or ribosome free endo- plasmic reticulum membranes, precautions were taken to remove ribosomes. Chief among these methods were lysis of vesicles and an EDTA wash, which judging by the RNA content of the extracted membranes were somewhat effective. Attempts to chemically determine the extent of removal of non-structural protein from the membrane proteins was done by monitoring lipid phosphorus content of extracted mem— brane preparation. The rationale behind this procedure, which is in accord with the membrane model posed earlier, is that membranes are expressions of macromolecular lipo- protein repeating units (3, 4). LI TERATURE REVI EW Membranes Resolution of the proteins of membranes to study their physical, chemical, and functional properties pre- supposes that one has at least a working model. The classical model of membranes is the paucimolecular model of Danielli and Davson (l) picturing membranes as con— sisting of one or more bimolecular phospholipid leaflets sandwiched between two layers of protein in the extended or beta conformation. Both the lipid and protein form continuous and separate phases, the latter being electro- statically bonded to the polar groups of phospholipids. Initially the model proposed that there were two bimolecu- lar phospholipid layers separated by neutral lipids, however, this View in light of electron microscopic and X-ray diffraction studies (19, 20) of membranes, was revised to that of the unit membrane hypothesis of Robert- son (2). The major difference between the two being that the observed thickness of membranes could accommodate only one bimolecular phospholipid layer between the two protein layers. Though this model was supported also by the chemical composition of some membranes, e.g. lipid content of membranes range from 30% for mitochondria to 80% by weight for myelin (14), further experiments using improved techniques bore data contrary to the classical model. The suggested role of lipid was first shown to be a limitation of the model when it was discovered that the binding mode of lipid to protein was hydrophobic rather than electro- static (21, 22, 23). More alarming were the studies of Fleischer gt_al. (24) who were able to extract up to 85% of the lipid from mitochondrial membranes observing essentially no alterations in the trilaminar structure characteristic of electronmicrographs of membranes fixed with osomium tetroxide. The fact that the membrane did not collapse thus shifted the role of the structural determinant of membranes from lipids to proteins. Another argument against the classical membrane model is that about 20 to 50% of total membrane proteins, which were observed by Infrared Spectroscopy, Circular Dichroism, and Optical Rotatory Dispersion (22, 25, 26) were shown to be present partly in the alpha helical conformation and the rest pre- dominantly in a random coil conformation rather than the extended or beta conformation as predicted by the model. Thus the classical membrane model seems to be based upon observations which have been shown to be circumstantial and not in accord with chemical and physical observations. Over the last decade evidence has been accumulating indicating that the structural components of membranes are macromolecular lipoprotein subunits and that membranes are formed by repeating subunit layers one unit thick. The first indications of the subunit structure of membranes were suggested by the observations of Green (27) and others (28, 29, 30) that the enzymatic systems of electron transfer and the citric acid cycle in mitochondria appeared to be an ordered arrangement of macromolecular enzyme com- plexes. Similar observations have also been made for endoplasmic reticulum (31) and chloroplast membranes. That the ultrastructure of membranes consisted of macro- molecular repeating units was first shown in electron- micrographs of mitochondria (9) and later in microsomes (31), bacterial membranes (32, 33), and plasma membranes of liver cells (34) using the negative staining techniques of Fernandez-Moran (9). Other techniques of electron- microscopy, e.g. the freeze etching technique of Moor and Muchlethaler (35), have also demonstrated the repeating subunit structure of membranes thus eliminating the possi- bility of artifacts due to sample preparation. Studies on the inner mitochondrial membrane have provided evidence correlating the macromolecular subunits of this membrane with the enzymatic functions of electron transport. This membrane system has been shown to consist of tripartite repeating units (9), i.e., an invarient mem- brane forming sector or basepiece, a variant or detachable section consisting of a headpiece and stalk. The enzymatic functions of electron transport have been resolved into fcbur electron transport complexes: I, NADPH—Coenzyme Q reductase (28); II, Succinate—Coenzyme Q reductase (3); III, Reduced Coenzyme Q-cytochrome c reductase (36); and IV, cytochrome C oxidase (37). Each of these functions has been associated with the basepieces of the inner mitochondrial membrane (38). Most interesting have been the membrane reconstitution studies by Green (38) and Kopaczyk (39) using these electron transport complexes. They were able to show that the inner mitochondrial mem- brane or any of the complexes taken singly or in different combinations could, after being depolymerized with bile salts, form membrane vesicles upon removal of the bile salts. In like manner complexes II and III, when prepared by methods which remove structural protein (see the intro- duction or section II of this review for definition), possessed the same membrane forming capabilities (39), suggesting a structural role for catalytic proteins. One function of lipid which resulted from these reconstitution studies was that of regulation of the mode of membrane formation. Lipid, distributed on two faces of the cuboidal shaped basepieces, inhibits binding of other basepieces at these faces and assures that membranes are formed as two dimensional continuums one subunit thick (3). Lastly, formation of membranes in these reconstitution experiments was shown to be specific for the four electron transport complexes, other molecules tested including structural protein and the headpieces of the inner mitochondrial membrane did not Show this membrane forming ability. This latter finding warrants a reconsideration of the function of structural protein in membrane systems. Thus the repeating subunit structure is, in light of the evidence stated previously, an attractive working model. In most membranes which have been observed by negative staining techniques of electronmicroscopy, such a model seems applicable; what remains to be done is to define the repeating units of biological membranes with respect to organization, enzymatic activity, and the function of that most abundant protein fraction within these subunits, structural protein. Structural Protein Structural protein (SP) is that water insoluble and apparently noncatalytic protein which was designated by Green et_al. (10) as being the protein constituent of the lipoprotein subunits of mitochondrial membranes. Its iso— lation was first achieved by solubilization of membranes with bile salts (deoxycholate and cholate at concentrations of 2 mg/mg protein and l mg/mg protein respectively) and sodium dodecyl sulfate (0.75 mg/mg protein), collecting the protein which precipitates from 0—12% ammonium sulfate concentration, and extracting lipid and bile salts with either 75% methanol at 50°C or as modified by Richardson (13), with 90% acetone. The resulting protein fraction, which shall be designated here as "crude SP," was 10 characterized physically and chemically (10, 12, 40) and prematurely assigned the role of being solely responsible for membrane integrity. Crude mitochondrial SP was shown to have the following prOperties which gives support to its designated function as mentioned above: it was found to be the major protein constituent of the mitochondria and it has the ability to bind and form stable complexes with lipid and enzymes of the mitochondrial electron transfer chain. Other properties which further suggest a structural rather than catalytic role include the fact that it contains no enzyme cofactors such as flavin, heme, nonheme iron, or copper and was found to be soluble only in detergent solutions at extremes of pH. Criddel et_al, (40, 41) initially were able to obtain both physical and chemical evidence suggesting that crude SP as isolated by the above methods (10, 13) was homogeneous. Sedimentation studies of this crude SP fraction solubilized in a solution containing 0.1% sodium dodecyl sulfate and 0.1 M NaCl at pH 10.5 showed a single symmetrically migrating boundary with sedimentation co— efficient corresponding a molecular weight of roughly 22,000. Further studies revealed: the presence of only one carboxy terminal amino acid; that one protein band is observed upon starch gel electrOphoresis of SP dissolved in 0.3% sodium dodecyl sulfate at alkaline pH; and peptide mapping of trypsin digests of crude SP gave close corre- lations between the number of peptides observed and the 11 number of lysine plus arginine residues present. However, subsequent work by Lenaz gt_al. (17), who were able to purify SP from a crude SP preparation isolated by the con- ventional method by extractions with 0.4% TCA in methanol and 8 M urea, did not confirm the homogeneity of that protein fraction termed here as crude SP. To date the heterogeneity of this crude SP fraction has been shown by several criteria: electrophoresis of crude SP by the method of Takayama (16), in which samples are dissolved in a phenol-acetic acid solution and run on polyacrylamide gels which contain 5 M urea and 35% acetic acid, showed multiple bands; several N-terminal amino acid species were detected when dinitrophenylation of crude SP was performed in 6 M guanidine hydrochloride (18); and multi- ple peaks resulted from sedimentation velocity studies of crude SP in the presence of 6 M guanidine hydrochloride (42). These findings question the homogeneity of the crude SP fraction of Green gt_al. (10) and gives cause for critical review of some of the assumptions made regard- ing its properties and function. Purification of SP from heavy beef heart mitochondria (HBHM) has recently been achieved by Lenaz gt_gl. (17). Their method essentially involves extraction of crude SP, prepared according to Green (10), Richardson (13) or modifications thereof (17), with 0.4% TCA in methanol followed by extractions with 8 M urea at pH 5.5. Though 12 this method has been shown to damage the proteins (18), it is thus far the only method shown effective in purify- ing SP. One interesting aspect of these studies has been the correlation of the solubility properties of the crude SP with the method of isolation which indicates differences in conformational states of these crude SP preparations. Some physical properties of purified SP isolated from HBHM, heavy beef liver mitochondria (HBLM), HBHM ATPase, and HBHM outer membrane have been determined and found to be quite similar (18, 43). These properties include: the fact that the average molecular weights all fall within the range of 60,000 to 70,000; the peptide maps of tryptic digests of these SP species are similar, suggesting struc— tural similarities; amino acid compositions of these SP species were strikingly similar whereas in comparison to those for the other enzymes associated with these membranes were markedly different; and the N—terminal amino acids of these pure SP preparations were found to be either aspartate or alanine. Differences were found to exist in the number of SP species seen on polyacrylamide gel electrophoresis after oxidation of pure SP isolated from different mem- branes (17, 43). Some correlations have been made as to the specific locations of these species within the mito- chondrion, e.g. two species have been located in the outer and two in the inner mitochondrial membrane, and as to the small but consistent differences in the amino acid con- tents of the species located in the outer as compared to 13 those located in the inner mitochondrial membrane (18). These properties argue against a single structural protein species within a given membrane system and suggests that structural protein forms a class of noncatalytic proteins. There now appears to be other species of noncatalytic proteins associated with mitochondria which have been recently designated by Green et a1. (43) as being core proteins (CP). Evidence has been accumulating which sug- gests different locations for the two types of noncatalytic proteins within the tripartite repeating units (9) of the mitochondrial membrane, i.e., SP has been shown to be the noncatalytic protein of the detachable sectors whereas CP is that associated with the basepieces. An example of such evidence is the recent finding that a protein fraction, which is electrophoretically identical to Green's SP (10, 43), could be extracted with 1.4% acetic acid or 7 M urea from mitochondria (44) and submitochondrial vesicles (45) without destroying membrane integrity as evidenced by electronmicro- scopy. These findings coupled with the fact that core proteins have been shown to account for some 50 per cent (35, 39) of the protein in basepieces (complexes III and IV) of the inner mitochondrial membrane adds further proof to Green's premise (43) regarding the location of non- catalytic proteins within the mitochondrion. Each of these noncatalytic proteins were shown to be physically similar, the only exceptions being that core proteins are 14 resistant to hydrolysis by proteolytic enzymes, e.g. trypsin, pronase, and papain (43), and that the molecular weights of CP fell within the range of 50,000 to 51,000 as opposed to 60,000 to 70,000 found for SP (18, 43). In light of these findings, there appears to exist within the mitochondrial membranes a general class of non— catalytic proteins of which SP and CP form subclasses. Their locations within the mitochondrial membrane have thus been established but their location in other mem- branes and their specific functions have yet to be demon- strated. METHODS AND MATERIALS Chemicals Most chemicals used in these experiments were of reagent grade from the usual sources and underwent no further purification, unless mentioned otherwise. Sodium dodecyl sulfate, ribonuclease-A (proteinase free), the sodium salts of cholic and deoxycholic acids, NADPH, NADP+, isocitric dehydrogenase, D,L-isocitric acid, and cyto- chrome c (horse heart) were all obtained from Sigma Chemical Co., St. Louis, Mo. The chemicals used in polyacrylamide gel preparation (acrylamide, N,Nl-methylene bisacrylamide, N,N,Nl,Nl-tetramethylethylenediamine and ammonium persulfate) were obtained from Canalco Industrial Co., Rockville, Md. Aminopyrine was obtained from K and K Laboratories, Inc., Plainview, N. Y. From Eastman Organic Chemical, Distillation Products Ind., Rochester 3, N. Y., 2,4-pentanedione was obtained. Orcinol was obtained from HARLECO, Philadelphia, Pa., and purified by dissolving in boiling benzene, decolarizing with charcoal, and recrystall- ized by cooling the solution. Animals The rats used in these studies, weighing from 200 to 350 grams, were of the Holtzman strain obtained from 15 16 Spartan Research, Haslett, Michigan. Unless mentioned otherwise, rats of either sex were used indiscriminately. In some experiments where proliferation of smooth endo- plasmic reticulum of the liver was desired, rats were treated with phenobarbital. The dosage of phenobarbital (100 mg/kg) was given daily by intraperitoneil injections five days prior to sacrifice, or was contained in the rats' drinking water at 0.1% concentration for a period of at least 14 days prior to sacrifice. Beef livers were obtained fresh from Van Alstein Packing Co., East Lansing, Mich. Isolation of Microsomes Method l.--Fresh beef or rat livers were minced, added to three volumes of cold 1.15% KCl, and homogenized with ten strokes in a Potter-Elvehjem homogenizer, clear- ance of 0.0069 inches, fitted with a motor driven teflon pestle. The resulting homogenate was centrifuged for 20 minutes at 10,0009. The supernatant was carefully decanted and centrifuged for 100 minutes at 105,000g in the number 30 rotor of a Spinco Model L preparative ultracentrifuge. The 105,000g supernatant was discarded and walls of the tubes wiped free of lipid. A small amount of 1.15% KCl was added to each tube followed by gentle shaking of the tubes to separate the microsomal pellet from glycogen. The pellets were washed by resuspending in KCl (three volumes of initial liver weight) and centrifuging as 17 before. Pellets were separated from glycogen as described above and resuspended in a small volume of 0.05 M Tris- HCl buffer, containing 50% glycerol and stored at -15° until used. Method 2.--Same as method 1 with the following exceptions: livers were homogenized in and the micro- somes stored in 0.25 M sucrose; and separation of the microsomal pellet from glycogen could not be obtained. All of the above procedures were performed at 0-4°. Rat livers were perfused, in £333, with the homogenizing solution until the livers were blanched. The livers were then excised from the rat and immediately placed on ice. Beef livers were packed in ice as soon as possible after sacrifice of the animal. Lipid and RNA Extraction Lipid and RNA were extracted from tissue homogenates essentially according to the method of Schneider (46). Acid soluble materials were removed by extraction of the protein sample with cold 10% TCA for 10 minutes. Lipid was then removed from the remaining pellet by extractions at room temperature for 20 minutes with 5 ml of 95% ethanol followed by another extraction with 3 ml of an ethanol- petroleum ether mixture (1:2). Extraction of RNA from the lipid free pellet was accomplished by hydrolysis of RNA with 1N NaOH for at least 20 hours at 37° followed by precipitation of protein and DNA by neutralization of the solution. 18 Isolation of Structural Protein Structural protein was isolated from beef and rat liver microsomes essentially by the method of Criddle §£_gl, (40) using modification "d" of Lenaz e£_al. (17). The method involves solubilization of microsomes with deoxycholate Ung/mg protein), cholate (1 mg/mg protein), and sodium dodecyl sulate (0.75 mg/mg protein) and reduc- tion of cytochrome by addition of solid sodium dithionite followed by precipitation of structural protein by bringing the solution to 12% saturation with respect to ammonium sulfate. The solution is then allowed to stand for 16 hrs at 0-4° after which it is centrifuged for 20 minutes at 40,0009. The removal of lipid and bile salts from the pellet was accomplished by butanol extraction at 0-4° followed by a 75% methanol extraction at 50°. The solu- tion was kept at pH 9.0 with 1.0 N NaOH during the addition of dithionite and ammonium sulfate. Microsomal Fractionation Fractionation of microsomes with 0.26% sodium deoxycholate (DOC).--The procedure used here is essenti- ally that of Ernster §E_gl. (47) which involves the addi- tion of 2.6% DOC (pH 8.0-8.5) stock solution to microsomes, suspended in Tris-HCl buffer to a protein concentration of approximately 6 mg/ml, contained in a 40 ml centrifuge tube. The tube's contents were gently mixed by inverting the tube several times and centrifuged at 105,000g for 19 40 minutes in a number 30 Spinco rotor. The smooth endo- plasmic reticulum (SER) is contained in the loose, reddish, protein pellet whereas the rough endoplasmic reticulum (RER) plus some detached ribosomes are contained within the tight, light brownish, protein pellet formed as a result of the centrifugation. These fractions were separated, 1y0philized, and assayed for their lipid- phosphate and RNA-ribose content. Fractionation of microsomes on a discontinuous sucrose density gradient in the presence of CsCl and M3212.--The method described here, which fractionates microsomal membrane vesicles according to their ability to bind Cs+ and Mg++ ions, is that of Dallner gt_gl. (48). To the 10,000g supernatant of rat livers homogenized in three volumes of 0.25 M sucrose (see microsome isolation Method 2 in Methods and Materials) enough of a l M CsCl stock solution is added to give a final concentration of 0.015 M, 4.5 ml of this is carefully layered over 2 ml of a 1.3 M sucrose, 0.015 M in CsCl contained within a 7 m1 centrifuge tube. The tubes were then centrifuged for 2 hours at 105,000g in a number 40.2 Spinco rotor. Three fractions resulted from this centrifugation: a clear reddish supernatant; a cloudy white infranatant, located at the 0.25 M—1.3 M sucrose interface, which contains the bulk of the SER; and a tight, Cs+ binding pellet containing RER. The SER and RER fractions were washed by suspension 20 in 1.15% KCl and centrifuged at 105,000g for 2 hrs, these fractions were designated as whole SER and RER, respect- ively. Fractionation of SER was performed by diluting the infranatant or whole SER fraction from two of the tubes mentioned above with distilled water to a volume of 4.5 ml and adding 3 mg MgClz, final concentration 0.007 M. This is then layered over 2.0 ml of a 1.15 M sucrose, 0.007 M in MgCl2 contained in a 7 ml centrifuge tube and is centri- fuged in a number 40.2 Spinco rotor for 45 minutes at 105,000g. Three fractions resulted from this centrifu- gation: a clear supernatant; a magnesium binding tight pellet designated as SER I; and a magnesium non-binding infranatant protein fraction, designated as SER II. Both fractions were washed by suspension in Tris-HCl buffer and centrifuging for 90 minutes at 105,000g. The washed pellets in all cases were resuspended in Tris-HCl buffer and stored at 0-4° until ready for use. Isolation of ribosomes.--Treatment of microsomes with 0.5% DOC according to the method of Palade and Siekevitz (49) was found to solubilize most of the membranes and constituent proteins with the exception of ribosomes. The procedure used was to add enough to a stock solution of 5% DOC to a microsomal suspension contained in a 40 ml centri- fuge tube at a protein concentration of 10-20 mg/ml, to final DOC concentration of 0.5%; the contents of the tube 21 are then mixed by inversion several times and centrifuged for 100 minutes at 105,0009. The fractions obtained by this centrifugation were a clear reddish supernatant and a small pellet which contained ribosomes and possibly membrane fragments as contaminants. The latter was washed three times by resuspension in Tris-HCl buffer and centrifuging as mentioned above. Fractionation of microsomes with tertiary-amyl alcohol--Isolation of electron transport membranes.-- Fractionation of microsomes with tertiary-amyl alcohol was performed according to MacLennan et al (31). Micro- somes suspended in 0.25 M sucrose to a protein concen- tration of 27 mg/ml were diluted with l/3 volume of 0.9% KCl and tertiary-amyl alcohol was added to a final con- centration of 10% by volume. The mixture is stirred slowly at room temperature for 10 minutes and is then centrifuged at 79,0009 for 30 minutes at 4°. Three fractions should result; a clear supernatant; a loose reddish pellet con- taining the electron transport membranes; and a tight brownish pellet. Fractions were washed once by suspension in Tris-HCl buffer and centrifuging at 105,0009. Microsomal Extraction Procedures 1.4% Acetic acid extraction.--The method of Zahler et a1. (44) was used without modification. It involved: extraction of microsomal protein by suspending microsomes in 1.4% acetic acid (pH 3.1) to a protein concentration 22 of approximately 10 mg/ml; incubating the mixture in an ice bath for 30 minutes; and centrifuging at 105,0009 for one hour. A protein fraction is then precipitated from the supernatant by adjusting the pH to 6.5 and collected by centrifuging for 30 minutes at 40,0009. Lysis of microsomal vesicles.-- Method l.--Microsomes are lysed at ice bath temper- atures by thawing microsomes, if stored frozen, and dilut- ing them to a final protein concentration of l to 3 mg/ml with cold distilled water. The mixture is then stirred in an ice bath for one hour after which the lysed vesicles are pelleted by centrifugation at 105,0009 for 90 minutes. Method 2.--This procedure is similar to that described in Method 1. The modifications were that lysis was per- formed at room temperature for 30 minutes and that glycerol was added to microsomes stored in 0.25 M sucrose to give a final concentration of 2.76% (approximately that concen- tration of glycerol obtained when microsomes stored in Tris-HCl buffer, 50% in glycerol are diluted 1 to 25 with distilled water to achieve lysis). In one experiment light scattering, as measured at 600 mu on a Coleman spectro- photometer equipped with recorder, during lysis by this method was followed with time to get an idea of the time needed for complete lysis. Care was taken to ensure that any changes in light scattering were not due to protein settling out of the suspension. 23 Salt extractions of microsomes.--This method of extracting microsomal proteins involves addition of enough crystals of the appropriate salts to give the desired final salt concentration to 5 ml of Tris-HCl buffer at 0-4° contained in a 7 ml centrifuge tube. One milliliter of microsomal protein suspension (3 to 8 mg of the resuspended lysed pellet) is then added to each tube; the tubes are then capped and their contents mixed by inversion several times. The tubes were then incubated in an ice bath for one hour, with occasional mixing, after which they were centrifuged for two hours at 105,0009 in a Spinco number 40.2 rotor. In each experiment a control, consisting only of microsomal protein in Tris-HCl buffer, was included. Organic solvent extractions.--Microsomes which had been lysed and washed with pyrophosphate buffer (0.1 M followed by 0.02 M, pH 7.8) were lyophilized and either 5 or 100 mg samples extracted with organic solvents or solvent mixtures. Extraction of the lyophilized protein was performed by homogenizing in the solvent for three minutes with a teflon pestle fitted glass tissue homoge- nizer, followed by incubation in an ice bath for one hour with occasional mixing. The insoluble proteins were collected by centrifugation at 40,0009 for 30 minutes and their protein content determined by the method of Lowry (50). 24 Sonic oscillation of microsomes.--A microsomal sus- pension of about 15-20 mg protein/ml in Tris-HCl buffer contained in a plastic tube was sonicated with a Bronson Sonic Power sonifier, with power scale turned to 4 D.C. amps. Sonication was carried out in an ice bath for intervals not exceeding 30 seconds, allowing one minute between intervals for cooling of the microsomal suspension. The sonicated suspension is then transferred quantitatively to a 7 ml centrifuge tube and the insoluble proteins re- moved by centrifugation at 105,0009 for one hour. Enzyme Assays Aminopyrine demethylase.--Aminopyrine N—demethylase activity was assayed by measuring the rate of formaldehyde production according to the method of Nash (51). Fixed time assays of 7 minutes were used through these studies, unless mentioned otherwise. The incubation mixture used contained: MgC12(7 mM), D,L-isocitrate (2mM), NADP+ (0.1 mM), NADP-isocitrate dehydrogenase (0.05 units/m1), microsomes (0.5 to 1.2 mg protein/ml) and varying concen- 3M to 1.33X10-4M). trations of aminopyrine (from 4.0X10— The reaction is started by the addition of microsomes to the incubation mixture contained in a 20 milliliter beaker on a Dubnoff Metabolic Shaking Incubator at 37°. After a 7-minute period of incubation the reaction is stopped by addition of one milliliter aliquots of the incubation to one milliliter of 10% TCA contained in a 5 milliliter 25 centrifuge tube. Protein is allowed to precipitate for about 5 minutes and two milliliters of Nash reagent (2M NH C H 0.05 M CH 4232’ 3 anedione) are added. The mixture is then heated for 15 0 COOH; and 0.002 M 2,4-pent- minutes at 50° to allow color development, centrifuged for about 5 minutes at 10009, and the color of the supernatants read at 412 millimicrons in a Coleman Jr. Spectrophotometer equipped with a flow-cell of 1 cm. path 1 of length. An extinction coefficient of 7.08 OD. ml- assay umole-1 of HCOH is used to calculate the umoles of formaldehyde formation. NADPH-Cytochrome c reductase.--NADPH—cytochrome c reductase activity was determined by the method of Omura §5_31. (52), which measures the initial rate of reduction of horse cytochrome c at 25°. The assay system contained 0.1 M phosphate buffer, pH 7.5, 1.3X10_5M cytochrome c, 5 3.0X10- M NADPH, and microsomes (from 0.2 to 0.6 mg of protein) in a volume of 2.5 milliliters. 3 for A millimolar extinction coefficient of 27.7X10- reduced cytochrome c at 550 millimicrons was used to esti- mate reduced cytochrome c. Analytical Methods Protein was determined according to the method of Lowry et al. (50) at 750 mu with crystallized bovine serum albumin as protein standard. Inorganic phosphate was determined according to the method of Bartlett (53), at 26 825 mu. Total phosphate was determined by digestion of whole protein samples whereas lipid phosphorus was deter- mined from the digests of the lipid extracts of protein samples. Ribose in the RNA extracts was determined by the orcinol method of Schneider (46). Polyacrylamide Gel Electrophoresis The disc electrophoresis system used was patterned after the method of Takayama (16). The final gels, 7.5% acrylamide in 35% acetic acid and 5 M in urea, were pre- pared by mixing stock solutions A and B with tetramethyl- ethylenediamine in the proportions, 3:1:0.02(v/v). Stock solution A consisted of 6 grams acrylamide, 0.16 grams N,Nl-methylene bisacrylamide, 12 grams urea, 28 ml glacial acetic acid and water to make 60 m1 final volume. Stock solution B consists of 12 grams urea and 0.3 grams ammonium persulfate in 20 ml water. Stock solution A could be stored for periods up to 6 months if kept refrigerated in a brown bottle; however, stock solution B was made fresh before each experiment. The buffer system used throughout was 10% acetic acid, both the cathode and anode. Poly- merization of the gels was carried out in a water bath at 47° for 15 minutes. The gels were then covered with a solution containing 5 M urea in 75% acetic acid and pre- electrophoresed for one hour at 5 milliamps per tube to remove ammonium persulfate. The removal of persulfate from the gels by pre-electrophoresis was checked by 27 electrophoresing gels for various periods of time after which the gels were immersed in a solution containing 2% benzidine chloride in 10% acetic acid. The presence of persulfate was indicated by deep blue color formation, as observed by Bennick (54). All protein samples were dissolved (1 mg/ml) in a mixture containing phenol, acetic acid, and water (2:1:1) and from 0.1 to 0.15 mg applied to the gels, unless mentioned otherwise. Only the running gel of 7.5% acrylamide, 5.5 cm in length, was used. Electrophoresis was routinely performed at room temperature for one hour with a constant current of 5 milliamps per tube. The gels were stained for a minimum of one hour in either Coomassie blue (0.05% in 12.5% TCA) or Amido Schwartz (0.55% in 7.5% acetic acid) and destaining by diffusion in 10% TCA and 7.5% acetic acid for gels stained with Comassie blue and Amido Schwartz, respectively. Ribonuclease—A was used as an internal standard. Structural Protein Assay Throughout these studies structural protein (SP) was identified as being the major protein species, i.e., the more densely stained protein band, in the poly- acrylamide gel electrophoresis profile of untreated microsomes. The reasons for this designation are men- tioned in the Introduction and Literature Review. The above definition of structural protein will be used throughout unless otherwise mentioned. RESULTS Fractionation of Microsomes Fractionation with 0.26% DOC.--The problem of re- solving liver microsomal membranes into their constituent proteins was first approached by investigating several methods known to separate the membranous from the non- membranous fractions of microsomes. A membrane prepara- tion free of readily detachable proteins and ribosomes, which constitute a rather heterogenous protein class, was desired. Several methods were investigated, one of which was the treatment of microsomes with 0.26% DOC which, according to Ernster et a1. (47), has previously been shown to yield a membrane fraction relatively free of attached ribosomes. Treatment of beef liver microsomes, isolated according to Method 1, by this method resulted in three protein fractions upon centrifugation: a clear yellowish supernatant; a loose, reddish, membranous pellet which should be essentially free of ribosomes; and a tight, brownish, protein pellet. It was found, however, by chemical analysis of the various fractions that the RNA-ribose content of the loose, reddish, membrane pellet had a higher RNA-ribose content than did the other frac- tions or untreated microsomes. This suggests a membrane contaminated by ribosomes. 28 29 Fractionation on discontinuous sucrose density gradient in the presence of Cs+ and Mg++ ions.-—Another fractionation method investigated was that which separates microsomes into RER, SER I, and SER II, on a discontinuous sucrose density gradient according to their ability to bind the cations, Mg++ and Cs+. The chemical analysis of these three fractions and those of whole microsomes and whole SER are shown in Table 1 below. TABLE l.--Analyses of smooth and rough endoplasmic reticu- lum (SER and RER) from the livers of phenobarbital treated rats. [Separation of SER and RER and subfractionation of SER was achieved by centrifugation on a discontinuous sucrose gradient in the presence of CsCl and MgCl2 (see Methods and Materials)]. Total- Lipid- RNA- phosphate phosphate ribose Sample _3 ( umole ) ( umole ) ( mgXlO ) mg protein mg protein mg protein Whole micro- somes 3.31 2.56 15.75 RER 6.75 2.59 33.30 Whole SER 4.01 3.51 7.72 SER I 4.42 3.15 12.72 SER II 2.60 -- -- The fact that there is approximately a four-fold difference in the RNA ribose content of RER and whole SER indicates some degree of separation of these.two fractions. Also the differences in the lipid—phosphate content of 30 whole microsomes and RER as compared to that of whole SER and SER I indicate that a degree of purification of mem- branes has taken place (keeping in mind the membrane model presented in the Introduction and Literature Review sec- tions in which membranes are thought to be expressions of repeating lipoprotein subunits). In spite of the chemical differences between these membrane fractions, close simi- larities in the polyacrylamide gel electrophoresis patterns were observed. However, there were some differences in the electrophoresis profiles of the RER and SER II membrane fractions as shown in Figure 1 (tubes 2 and 5). The electro- phoresis profile of RER, tube 2, shows the presence of faster migrating protein bands not seen in the other membrane frac- tions. It is likely that these faster migrating bands represent ribosomal proteins since it is generally assumed that the only difference between RER and SER is the presence of ribosomes on RER membranes. The electrophoresis profile of SER II (Figure 1, tube 5) shows a prominent protein band which, though present in all other membrane fractions, seems to be more concentrated in this fraction. It was also observed that this membrane fraction had more of a reddish color than the other frac- tions, suggesting the identity of this band to be one of the microsomal cytochromes. Though the true identity of this protein band was not pursued, it is interesting to note the observation that this protein band is extractable 31 Fig. l.--Fractionation of microsomes from phenobarbital treated rats with Mg++ and Cs+ ions. 1. Whole microsomes. 2. Rough endoplasmic reticulum. 3. Whole smooth endoplasmic reticulum. 4. Smooth endoplasmic reticulum, subfraction I. 5. Smooth endoplasmic reticulum, subfraction II. 6. 0.01 M EDTA + 0.15 M KCl (in Tris-HCl buffer) wash of microsomes: supernatant fraction. 33 with Tris-HCl buffer containing 0.01 M EDTA and 0.15 M KCl (Figure 1, tube 6). Even though there was some success in separating SER and RER as judged by the electrophoresis and quantitation studies of the fractions, the yield of SER obtained by this method was so low that other methods of membrane purification had to be investigated. Tertiary-amyl alcohol treatment of microsomes.-- Treatment of microsomes with tert-amyl alcohol according to the procedure of MacLennan et_al. (31) has been shown to yield a membrane fraction essentially free of ribosomes and rich in the electron transfer enzymes of microsomes. The method was applied here in an attempt to obtain a reasonably pure membrane fraction free of major contami- nants such as ribosomes and other protein species loosely bound to the membrane. As used here, however, the method only yielded four somewhat crosscontaminated fractions, judging from the close similarities in the RNA-ribose and lipid phosphate contents of each. These fractions (a clear supernatant, a loose reddish pellet, a mixture of loose reddish and brownish pellets, and a light brownish pellet) were observed also to have very similar electrophoresis profiles, which would indicate very little resolution of microsomal membrane components. Microsome Extraction Procedures Organic solvent extractions.--In View of the diffi- culties experienced in attempts to achieve a sufficient 34 separation and quantity of the membranes of microsomes, methods, which would specifically extract all non-membrane forming proteins (e.g. ribosomal proteins and other pro- teins loosely bound to the membrane), were investigated. One method attempted was to remove electrostatically bound protein by successive washings of microsomes with pyro- phosphate buffers (0.1 M and 0.02 M, pH 7.8) and to resolve the remaining membrane proteins, interacting primarily through hydrophobic bonds (see Literature Review), by organic solvent extractions. The pyrophOSphate buffer washed microsomes were lyophilized and 5 mg extracted with the following solvents: glycerol-butanol (1:4 v/v), butanol, dioxane-water (1:4 v/v), dioxane-water (4:1 v/v), pyridine, N,N'-dimethylformamide, N,N'—dimethylformamide-water (1:1 v/v), and 2-aminoethanol. Extraction of the protein samples was performed according to the procedure outlined in Methods and Materials. Of the solvents mentioned above, it was ob- served that the dioxane-water (4:1 v/v) gave the best re- sults with respect to protein solubilization. A rough estimate as to the extent of protein solubilization indi- cated that this solvent extracts 2 to 4 times the protein as did the other solvent mixtures. The effect of divalent cations, for example Mg++, on the solubilization of membrane proteins by some of the organic solvents mentioned above was also investigated. This investigation was initiated due to the findings of 35 Byington et al. (55) that various aliphatic alcohols, e.g. methyl, ethyl, n-butyl, n-amyl, and n-decyl alcohols, solubilized more mitochondrial membrane proteins in the presence of Mg++. The solvent mixtures used here are shown in Table 2. TABLE 2.--Solubi1ization of beef liver microsomal protein by organic solvents in the presence of 0.1 M MgClz. [Beef liver microsomes isolated according to Method 1, were washed with pyrophosphate buffer, lyophilized, and 0.1 gram samples extracted with organic solvent solutions in the presence and absence of 0.1 M MgClz using methods described in Methods and Materials]. Solvent Mixture % Protein Solubilizeda Dioxane:HZO (2:8) 72.3 Dioxane:H20 (2.8), 0.1 M in MgCl2 76.0 DioxanezHZO (8.2) 87.4 Dioxane:HZO (8.2), 0.1 M in MgCl2 88.8 N,Nl-dimethyl formamide (1:1) 71.5 N,Nl-dimethyl formamide (1:1) 0.1 M in MgClZ 79.5 aExpressed as mg protein, as determined by the method of Lowry, in the extracted pellet divided by the weight of the lyophilized membrane preparation. It is interesting that in each of the solvents used an enhancement in protein solubilization was observed in the presence of Mg++. However, due to the fact that organic solvent treatments have been demonstrated to be 36 harsh on some proteins, milder methods of membrane resolution were pursued. Lysis of rat liver microsomes.--Lysis or osmotic shocking of microsomes was a method used to release the soluble contents of microsomal vesicles. It was noted that dilution of microsomes, stored in Tris-HCl buffer containing 50% glycerol, with distilled water resulted in a time dependent clarification. Quantitation studies showed solubilization of 6 to 14% more protein by this procedure than when microsomes were stored in 0.25 M sucrose. The effect of glycerol, present at final con- centration of approximately 0.3 M, on the extent of lysis of microsomal vesicles at room temperature was followed with time by light scattering measured at 600 millimicrons according to procedure described in Methods and Materials. Lysis was shown to be completed, both in microsomes lysed in the presence of 2.76% glycerol and those lysed with only distilled water, after approximately 15 minutes using the fact that no further decreases in optical density after this period of time as criteria. The lysis procedure was also observed to be temper- ature dependent. It was shown that 7 to 11% more protein could be solubilized by lysis at room temperature and that the presence of 2.76% glycerol enhanced protein solubilization during lysis at both ice bath and room temperatures. 37 Sonic oscillation of beef liver microsomes.--Beef liver microsomes, isolated according to Method 1 and diluted 1 to 4 with 1.15% KCl, were sonicated for various time periods to determine the extent to which microsomal vesicles were ruptured and their soluble contents released. Microsomes were sonicated, see procedure in Methods and Materials, for periods of 0, 10, 40, 60, 120, and 480 seconds. There was no significant amount of protein solubilized over the control after 480 seconds of soni- cation (Table 3). Electrophoresis profiles of the proteins of the pellets and supernatants, resulting after centri- fugation of the sonicated microsomes, were observed to show little difference from those of the control. It is signifi- cant that sonication of microsomes, a procedure which should effectively rupture all microsomal vesicles, did not extract more proteins than did the lysis procedure. This indicates that the lysis procedure (Method 2) was effective in com— pletely rupturing microsomal vesicles. Acetic acid (1.4% v/v) extraction.--Extraction of microsomes with 1.4% acetic acid was one method thought to be potentially useful as a means of extracting proteins non-essential to membrane structure. This extraction procedure was used successfully by Zahler et a1. (44) to extract a protein fraction from mitochondria which was electrophoretically identical to that of structural pro- tein, as defined by Green et a1. (10). Also, Zahler was 38 able to show by electron microscopy that this extraction procedure did not destroy membrane structure. In this investigation liver microsomes from phenobarbital treated rats were lysed, according to Method 2, extracted twice with 1.4% acetic acid, and the polyacrylamide gel electro— phoresis protein profile of the various fractions deter- mined (Figure 2). Judging from the electrOphoresis patterns of the extracted pellet and proteins precipitated from the supernatant by raising the pH to 6.5 (Figure 2, tubes 2 and 3) this treatment does not extract the protein band designated as structural protein in this study. It can be seen in Table 4 that the protein extracted by this method represents approximately 40% of membrane protein, a property which for reasons previously mentioned in the Introduction and Literature Review may justify assigning to it a structural function. However, the fact that it does not appear to contain any appreciable amount of phospholipid, as inferred by the phospholipid content of membrane proteins before and after the acetic acid extrac- tion (Table 4), would tend to rule out such a function for this extractable protein fraction. It must be mentioned that the pH of the 1.4% acetic acid solution was around 3.1, a condition known to denature most proteins. Also the fact that proteins were visually seen to coagulate and precipitate out of the microsomal suspension under these conditions gave further reason to abandon this treatment as a means of 39 Fig. 2.—-EDTA-KC1 and 1.4% acetic acid extractions of rat liver microsomes. l. Lysed microsomal pellet. 2. Lysed + 1.4% acetic acid twice extracted pellet. 3. 1.4% acetic acid extraction supernatant; proteins precipitated by raising pH to 6.5. 4. Lysed + 0.01 M EDTA, 0.15 M KCl (in Tris-HCl buffer) twice extracted. 5. 0.01 M EDTA, 0.15 M KCl extraction supernatant. 41 TABLE 3.--Quantitation of the effects of sonication on beef liver microsomal protein solubilization. [Beef liver microsomes, isolated according to Method 1 (see Methods and Materials) and stored in Tris-HCl buffer, 50% in glycerol, were thawed and diluted 1 to 4 with 1.15% KCl before being used in this experiment]. Time Sonicated - . . a (Total Seconds) % Protein Solubilized 0 11.5-22.3 10 10.5—11.4 40 10.3-19.7 60 11.4-30.5 120 12.4-18.3 480 13.0-18.4 aDue to the difficulties incurred in the quanti- tation of protein solubilized, upper and lower limits of the per cent solubilization were set according to quanti- tation of the protein contents of the pellets and supernatants, respectively, after centrifugation. 42 H.mm mm.NH hv.m mw.v o.ov mm.om Hm.m mo.h m.vm m.mH vm.m hm.m IIII m.hm mo.m mv.m cofluomuuxm Am.n mm .2 mo.ov Hom nmflee .Hom z mH.o .meom :5 OH + mflqu cofluomnuxw Uflom oflumom we.a + nflnsq mmEOmouoHE comma moEOmouowa pmumouucb MHTDOHQ me cfimuoum m8 ammuoum m8 pmNHHHQsHOm A OH NTME v A wHoE: V A THOE: V eflmuone m m mmonfim manganese tfleflq BBBBEnoee Hence ucmEummHB mHmEmm .AcoHHDHOm 480m no Ufiom Deacon Monuwm nufi3 moflzu ©m£m63 mums mmamfiwm cam m venue: on mcflpuooom Ummha mnm3 mmEOmouoflz .ucmfiummuu Hmuflnumnocmnm an UTOSUGH was Edasofluou UHEmmamoccm Hm>fla mo coflumnmmaaonma .mmsmunEmE HmEOmouoflE um>fla umu Umuomuuxm deem cam Uflom oaumom mo momMHmsdll.¢ mqmda 43 obtaining desirable membrane preparation with which to work. Salt extractions of beef and rat liver microsomes.-- Since the previous methods of resolving microsomal membrane proteins resulted in varying degrees of cross contamination and/or possible denaturation of the different protein fractions obtained, methods were sought which would specifi— cally extract proteins from microsomal membranes. Extraction of microsomes with various salts in Tris-HCl buffer were performed in an attempt to separate ribosomes and other readily detachable proteins from the lipoprotein network believed to be the basis of membrane structure (3, 4). The extent to which this was accomplished was determined by monitoring RNA-ribose and lipid-phosphate content of the various protein fractions. Of the several salts used initially, Table 5, KSCN was shown to best solubilize non- lipoprotein whereas NaBr and KNO3 were most effective in extracting RNA-ribose. Higher concentrations, 2.0 M, of salts used for the extractions (Table 6) gave similar results with respect to the effectiveness with which the various salts used solubilized non-lipoprotein and RNA— ribose. Electrophoresis profiles of beef liver microsomal membranes extracted with various salts (Figure 3) show close similarities with respect to the protein species solubilized. It was also observed that most of these 44 TABLE 5.--Ana1yses of beef liver microsomes following extractions with various salts. [Microsomes, isolated according to Method 1 in Methods section, were lysed according to lysis Method 1. All samples were extracted twice, once with 0.5 M followed by 1.0 M salt concen- trations]. Total c Lipid Ribose Sample Phosphate Phosphate Treatment ( umoles ) ( umoles ) ( mgrx 103 ) mg protein mg proteIn mg protein Lysed microsomes 4.23 1.88 3.59 Control pelleta 3.12 2.09 1.67 MgCl2 extractions 3.84 4.86 4.73 NaBr extractions 3.52 4.45 0.575 KSCN extractions 5.54 7.20 3.22 KNO3 extractions 4.00 4.18 0.388 Urea extractionsb 3.98 5.20 2.04 CaCl2 extractions 2.96 4.17 3.80 aLysed microsomal sample was extracted twice with 0.05 M Tris-HCl, pH 7.5 buffer. bThe first extraction was performed in 1.0 M urea and the second extraction with 2.0 M urea. CTotal phosphate salt extracted samples is in error in that the lipid content of samples was underestimated due to the presence of relatively high salt concentrations. .Houucoo may >2 poNHHHQSHom samuoum How Umuomuuoon .HTMMSQ .m.b mm .Homlmflue z mo.o cues coco Umuomuuxm mums mmEOmouoflE pmqum 45 nm.MH mm.H om.m Hm.m nofluoonexo Hooz 2 o.m ne.AH mm.H mH.m om.m nofleoonuxm mozs s o.m no.mm ko.~ om.o mo.o nofloonnoxo zoom 2 o.~ nn.nm SH.H mo.e NH.A oofluonnoxo nmnz z o.~ m.- ma.~ ea.m mm.m nuoaaoo Honuooo llll N©.N voom wmom mmEOmOHOHE mehfi ammuoum ma ammuoum 08 OH x mE V A smououm me mace: V A THOE: V ucmfipmmna >3 A couflaflnsaom m cflmuoum m pcmfiummue mamfimm moonflm meonomoee ofloflq munnmmoeo Honoe .Amoco wouomnuxm mmamfimm Ham .coflpomm mamflnmpmz cam moocumz CH a vogue: on ocflpnooom ommwa mums mmEOmOHOAZV .muamm msoflnm> mo mc0wu Imuucmocoo Loan Spas ompomuuxo mwcmunEmE HmEomOHOHE um>fla moon mo mommamcall.m mqmda 46 salt-labile proteins were electrophoretically identical to the proteins solubilized by lysis of beef liver micro- somes in 2.76% glycerol. This observation generally holds true for most salts, containing monovalent anions (NaBr, KSCN, KNO3, NaCl), investigated. One exception was found when 2.0 M KSCN was used to extract microsomes. This salt was shown to effectively solubilize all but two pro- tein components of microsomal membranes (Figures 3 and 4). The protein band with the slower electrophoretic mobility, because of the fact that it is identical to the predomi- nating protein species of whole rat and beef liver micro- somes (compare tubes 1 and 6 in both Figures 3 and 4), has been designated in these studies as being a structural protein of microsomal membranes. Though a structural role for this protein band has not been demonstrated, further evidence have been presented which may suggest its function. The protein fraction not extractable with KSCN has a rela— tively high lipid content compared to whole microsomes, see Tables 5, 6, and 7. This property would suggest that this protein is one of the proteins in the lipoprotein repeating units thought to be the basis of membrane structure (3, 4). The proteins extracted from rat liver microsomes, isolated from rats treated with phenobarbital, with salts containing polyvalent anions (e.g. citrate, phosphate, pyrophosphate, and carbonate) are shown in Figure 5. .ucoEumwuu 47 ucmmumuop mp moEOmouoflE uo>fla moon Eoum pmpmHOmfl mm .NH .ucmumcuomsm pmuomuuxm Humz 2 o.m comma .HH .uoHHoE oopoonnxo Hooz z o.m comma .OH .ucmumcumm5m pmuomuuxm mozx z o.m comma .m .umHHmm omuoonuxm mozx z o.m ommsq .m .ucmumcuom9m cmuomuuxm 20mm 2 o.m pmqu .n .umHHmm omuonnuxm zomx z o.~ oommg .o .ucmpmcuwmsm cmuomuuxm ummz z o.~ comma .m .umaamm couomuuxm ummz z o.m comma .o .pcmumcumd5m cofluomuuxm Homlmflue + comma .m .pmaamm AHouucooV coflpomuuxm Humlmwua + comma .m .uwaaom HmEOmOAOHE omqu .H .mmEOmouoHE Hm>fla moon mo mcofluomuuxm uHmmII.m .mam 49 .uoHHmm ompoouuxm zomm z m.H .n .umaamm ompomuuxm zomx z o.H .o .pmHHmd omnoonuxm zomx z m.o .m .umaamm pmuomuuxw ummz z m.H .v .umHHmm pmuommuxo ummz z m.o .m .AHouucooV mmEOmouoHE pmnmms Homlmflue .m .mmEOmouoHE Uwummuuca .H .mumu woummnp Hmpflnumnosmnm mo mmEOmouoflE Eonm mcflmpoum mo cofiuomuuxw any so mcoflumuucmocoo pawn mcflmmmuocfi mo poommmll.v .mflm 51 Fig. 5.--Extraction of liver microsomes from control rats with salts containing polyvalent anions. l. Lysis + 0.5 M sodium citrate extraction supernatant. 2. Lysis + 0.1 M sodium phosphate extraction supernatant. 3. Lysis + 0.1 M sodium pyrophosphate extraction supernatant. 4. Lysis + 0.05 M sodium carbonate extraction supernatant. 5. Lysis + 0.5 M sodium bromide extraction supernatant. 6. Lysis + Tris-HCl extraction supernatant. 553 TABLE 7.--Ana1ysis of rat liver microsomal membranes extracted with various concentrations of salts. [Proliferation of liver ER was induced by phenobarbital treatment of rats according to Methods and Materials section. Microsomes were isolated according to Method 1 but were stored in Tris-HCl buffer, 50% in glycerol for Experiment 1 and in 0.25 M sucrose for Experiment 2. In the case of Experiment 3 microsomes are isolated according to Method 2 and are stored in Tris-HCl buffer at 0-4° and used within 8 hours after isolation. Lysis of microsomes was per— fumed according to Method 1 for Experiment 1 and Method 2 for Experiments 2 and 3). Total Lipid‘ RNA- Experiment Treatment of Phosphate Phosphate ribose 8 Protein Number Sample ( umole ) ( umole ‘ ( m x 103 Solubilized mg proteIh mg protein’ mg protein 1 Lysed microsomes 3.73 2.57 18.1 ---- 1 Control pelleta 3.91 3.25 15.3 10.8b l 0.05 M NaBr extraction 4.22 3.36 16.3 7.2b 1 0.25 M NaBr extraction 4.35 3.65 16.1 10.2b 1 0.50 M NaBr extraction 5.00 4.02 16.8 19.4b 1 1.00 M NaBr extraction 5.49 4.19 18.9 22.8b l 1.50 M NaBr extraction 6.32 4.23 18.2 23.2 1 1.0 M KSCN extraction 8.94 7.10 34.2 47.5: 1 1.5 M KSCN extraction 9.88 8.60 41.4 55.6b 1 2.0 M KSCN extraction 10.45 10.22 48.1 60.8 1 0.5 M 10:03 4.35 4.05 16.8 16.2: 1 1.0 M KNO3 extraction 3.85 3.51 16.5 15.8b 1 1.5 M KNO3 extraction 4.47 3.80 17.1 17.5 2 Untreated microsomes 5.43 3.08 2.73 -—-— 2 Lysed microsomes 5.27 3.34 19.3 24.9 2 Control pellet 6.27 3.55 19.4 19.1 2 0.05 M LiCl extraction 5.75 3.57 17.5 ---- b 2 0.50 M LiCl extraction 6.25 3.63 15.4 1.40b 2 0.70 M LiCl extraction --—- ---- ---- 8.10b 2 1.00 M LiCl extraction 5.78 3.85 19.5 5.40b 2 1.50 M LiCl extraction 5.80 3.94 21.0 9.10b 2 2.00 M LiCl extraction 6.20 3.81 20.8 7.70 2 2.50 M LiCl extraction 5.97 4.03 21.7 13.9b 3 Unfrozen, lysed microsomes ---- 2.75 14.9 7.7Cc 3 Controla ---- 2.40 16.9 7.29 3 0.10 M Na2C6HSO7 extraction ---- 2.29 10.6 ~0c c 3 0.50 M NaZCSHSO7 extraction ---— 2.88 13.3 12.2 3 0.05 M NaZHPO extraction ---- 2.15 10.5 1.64: 3 0.10 M NaZHPO extraction ---- 2.16 12.1 1.51c 3 0.30 M NaZHPO4 extraction ---- 2.74 11.6 3.11 3 0.01 M Na4P207 extraction ---- 3.22 11.6 0.94: 3 0.05 M a4P207 extraction —--- 3.20 12.1 4.25 3 0.10 M Na4P207 extraction ---- 3.42 13.5 5.41 3 0.05 M Na CO (pH 10.0 c extraction? ---- 2.66 9.04 1.07 3 0.50 M NaNO2 extraction ---- 2.95 18.6 ---- 3 0.50 M NaBr extraction ~--- 2.57 16.2 6.74c aControl, lysed microsomes extracted once with 0.05 M Tris-8C1, pH 7.5, buffer. bCorrected for protein solubilized by control. cProtein determined on supernatant fractions by determining their optical densities at 280 and 260 millimicrons, corrections made for control solubilization when necessary; the effects of salts or salt concentrations on optical density readings at 280 and 260 mu are not known. 54 The concentrations of these salts were kept relatively low, from 0.01 M to 0.50 M, because of their solubility proper- ties under the conditions used in these experiments. For comparison, a tube showing the electrophoresis profile of proteins extracted with 0.5 M NaBr is included. Some differences can be seen with respect to the amounts of the faster migrating proteins extracted with these salts when compared to those proteins typically extracted with salts containing monovalent anions such as NaBr. It was also observed in other experiments that some of these faster migrating proteins were identical to those extractable with Tris-HCl buffer containing 0.01 M EDTA and 0.15 M KC1 (Figure 2, tube 5). With the exception of these differ- ences, the electrophoretic profiles of the proteins extract- able by these salts are very similar with respect to the specific proteins extracted and only minor differences are seen in relation to the amounts of proteins extractable (Figure 5). The effects of varying salt concentrations on the extent of protein and RNA-ribose extraction from liver microsomes of phenobarbital treated rats were also investi— gated (Table 7). Optimum concentrations for maximum solubilization of non-lipOprotein seems to be around 1.5 M for NaBr, 2.0 M for KSCN, 1.5 M for KNO and 2.5 M for 3 LiCl. The other salts tested showed increase in protein extraction with salt concentrations; however, due to the 55 limited solubilities of salts such as sodium citrate, phosphate, and pyrophosphate an optimum, as such, was not determined. Although KSCN solubilizes more protein than any of the other salts used, the evidence presented in Table 7 indicates that ribosomes are not being extracted from microsomal membranes, the criteria being the rela- tively high RNA-ribose content of the extracted micro- somal proteins. Figure 4 shows the effects of increasing salt con- centrations on the extent to which proteins were extracted from microsomal membranes isolated from phenobarbital treated rats. Only KSCN showed a correlation between salt concentration used and the specificity of the proteins so removed from the electrOphoresis profiles of the extracted microsomes. Extractions with various concentrations of other salts, for example NaBr shown in this figure, showed no such correlation. EDTA-KC1 extraction of microsomes.--Ribosomes have been generally thought to be attached to the endoplasmic reticulum (ER) by a combination of electrostatic binding and magnesium complexing. It is reasonable, therefore, to expect that chemical agents which are known to complex with or remove magnesium and weaken electrostatic bonds should detach ribosomes from ER membranes. Following the above reasoning, the effects of washing microsomes with solutions containing a known magnesium complexing agent, EDTA, in the presence of 0.15 M KC1 was investigated. 56 Liver microsomes from phenobarbital treated rats, isolated according to Method 2 in the Methods and Materials section, were thawed and washed by diluting aliquots in 40 m1 centrifuge tubes to a final protein concentration of about 2.5 mg/ml with the desired buffers (Table 6) and mixing the contents by inverting the tubes several times. The tube containing the unbuffered EDTA-KC1 solution (pH 4.9) coagulated the proteins and had to be mixed by homogenizing in a Potter-Elvehjem tissue homoge- nizer, fitted with a teflon pestle, for one minute at 0-4°. The tubes were then left in an ice bath for 20 minutes and then centrifuged for 100 minutes at 105,0009. The pellets were analyzed for RNA-ribose and lipid- phosphate. Supernatants were dialysed against distilled water for about 16 hours and lyophilized before electro- phoresis was performed. Of the treatments shown in Table 8, washing microsomes with Tris—HCl buffer containing 0.01 M EDTA and 0.15 M KC1 at pH 7.5 was most effective in extract- ing ribosomes and/or non-lipoproteins. The criteria used in making this observation is the relatively low RNA-ribose content coupled with an increase in the lipid-phosphate of the extracted pellet in comparison to those values observed in the control and the other extracted pellets. The electrophoresis profiles of the extracted micro- somal fractions and their supernatants are shown in Figure 6. It is interesting that each treatment removed one of 57 TABLE 8.--Analysis of Tris-HCl buffer, KC1, and EDTA extracted rat liver microsomal membranes. [Proliferation of the endoplasmic reticulum was induced by phenobarbital treatment]. Lipid Treatment of Phosphate % Protein Samples ( umole ) ( mg x 103 ) Solubilized mg proteIn mg protein Ribose Untreated micro- somes 2.14 14.67 ____ 0.05 M Tris-HCl, pH 7.5 ex- traction 3.31 18.55 28.8 0.05 M Tris—HCl, pH 7.5 + 0.15 M KC1 extraction 4.48 25.87 52.4 pH 7.5 + 10 mM EDTA extraction 4.06 19.55 42.8 10 mM EDTA + 0.15 M KC1, pH 4.9 extraction 3.84 25.30 46.4 58 .coflumummmum TEOmOQHn mcsno .ucmumcu992m mammq .ucmumcstSm ©o£m63 HUM z mH.o + Homlmflue .umaame cosmos Hos z ma.o + Homunene .pcmuoc IMTQSm pmnmm3 Am.v mQV Hum z mH.o + fla use wo cm63 Huxndeomll.m .mflm 60 the more electrophoretically mobile proteins associated with a densely stained protein-band of a crude rat liver ribosome preparation. This particular band is not removed by lysis of microsomal vesicles, indicating that it is not due to one of the soluble proteins located within micro- somal vesicles. Also by comparing the electrophoresis profiles of the proteins solubilized with 0.15 M KC1 in Tris-HCl buffer (compare tubes 5 and 9 with tube 3, Figure 6) one can again see that those proteins which are salt labile are also labile to lysis. In Figure 2, tubes 4 and 5 show the extracted pro- tein pellet and supernatant protein fraction of rat liver microsomes which were lysed, according to Method 2, and washed twice as mentioned above with Tris—HCl buffer con- taining 0.01 M EDTA and 0.15 M KC1 at pH 7.5. This figure shows more clearly that the EDTA-KC1 washing specifically extracts proteins of greater electrophoretic mobility than those associated with the lysis supernatant shown in Figure 6. Quantitative studies of the extracted pellet (Table 4) show a relatively low RNA-ribose content, sug- gesting that the proteins extracted from the lysed micro- somes, are ribosomal. The electrophoretic profiles, shown in Figure 6, would thus indicate a possible removal of microsomal vesicle content as well as ribosomal proteins by washing microsomes with Tris-HCl buffer containing EDTA and KC1. 61 Resolution of Rat Liver Microsomal Membrane Proteins by Combinations of Treatments Microsomes from the livers of phenobarbital treated rats were extracted successively by the treatments listed in Table 9. The protein pellets from the various treat- ments were collected by centrifugation at 105,0009 for 100 minutes in the case of lysis and the EDTA-KC1 washes and for 2 hr after the salt extractions. The pellets were then analyzed for their protein, RNA-ribose, lipid and total-phOSphate contents. Lysis of microsomes was performed according to Method 2 and the salt extractions and EDTA-KC1 washes according to Methods and Materials and the Results section, respectively. Lysis of microsomes extracted some lipid as well as RNA-ribose. Washing with EDTA-KC1 in experiment 1 (Table 9) extracts a significant amount of RNA-ribose, an obser- vation not in accord with the results shown for experiment 2 where most of the readily extractable RNA-ribose is removed by lysing microsomes. One interesting observation is that KSCN extraction removes RNA-ribose only when this treatment is preceded by NaBr extraction. This fact is also contrary to the results shown in Tables 5-7 regarding the effects of KSCN treatment alone on microsomes. Accord— ing to these data KSCN extraction does not remove RNA- ribose but in fact concentrate it with respect to the control and some of the other salt extracted pellets. It has also been observed that the electrophoresis profile of the protein €52 .cmEMOMuwm msm3 xmnu news: :a umpuo one ca beamed mum ecoEwquxe some cw mucmEumousm e.mm o.oe m.HH ma.o o~.m AeV coeuotnoxo zomx z o.H N m.me m.me o.oa Ho.m eo.q Am can «V cosmos woes» Huxnaecm a u--- u--- m.HH em.e oH.e AHV manna N null IIII H.mH ~m.a hm.v meOmouofiE poumouuca ~ ~.mm o.o~ m.e ~o.o am.e AeV coeuotnuxm zone 2 o.~ H m.mm e.om m.mH mm.m Ho.n AmV coeuotnuxo nmtz : m.H a om.eH o~.o m.ma vo.m mm.n ANV coauotnuxm Hoxuaaom H oe.m oe.m m.oH o~.m no.m AHV tense a null null ~.m~ vm.m m~.w moEomouoaa Deaconess H acmEummuB ucefiuomua Acawuoum mEV Asamuoum mEV Acaououm mEV ooceneoo so comm so mod x ms mace: mace: ucotutmne nonecz congeeooaom coneAAooHom moose“ outcononn outcomono t ucosanooxm camoonn a tenuono a nazm uceoeq Hence one: Hmufinumnocmnm :ufl3 cmuowuu mums Eouu mum>flAV .Au convex ou mcwpuoouo comma .mc0auoofluwu5m occunaee newest moEooOHOAE uo>wa use no man>ad§¢ll.m mqmuw>fluom camaommm mV >ua>fluo¢ mmmamcumfima mCHH>m0CHE< o meoucoou>onmmomz o.ova m.noa m.mm moEOmosoHE CmNouw mmouosm z mm.o mo mHm>A o.nma o.woa c.mHH mmEomouoHE Cmuoum Honmomam mom .Homunene no nflmsq m.mHH m.aoa m.mHH moEOmouoflE HouuCoo mo mflmmq m.mm o.Hm m.Hm mmouosm z mm.o CH mCHNmmnm ~.em e.mm m.mHH HonooSHo won .Humlmflne Cw mCflNmmum N.vca o.mcH m.mHH omouosm S mm.o Ca omccmmmsm meomouoflz o.va o.mmH v.5Ha Houmomam wom .Homlmfiua CH popcommsm mmEOmouon mmflswmomflfim mCflummomflfid Aaouucoo mo vioa o e EmloH o q mufl>fiuom camaommm wV ucmEuomHB mufl>fiu04 mmmposcmm mHQEmm Iowa umumm meson Ben C poms ccm HouuCOO .AAmHmwumumz pcm moonuon cofiumbsocfl one CH umcu qummummu mHm>mH mumuumnsm mCHummocflad .Cowuma ow um Cowman Humlmflne CH pmuoum omoce mno3 mmEOmOHOME .Amamflnmumz UCC mcocusz N vogue: on mCHUHooom pmauowsmm mm3 mmHmEmm Ham mo mflmwa ccm mmmc m Cow omHI um mmmnoum an pm>wficom was moEomouoHE mo mCHNmmum .mmflcsum omen» Ca cows mums Hmuwnumnocmcm CuAB Umumwuu mums mo mnm>fla ecu Eouw mmEOmouosz .mmEANCm Hmfiomonofie mo mmfiufl>fluom on» C0 mwmxa UCm mCfiNmmuw mo muommmmll.oa mamas 69 ¢.mm m.mHH m.OOH HUM 2 mH.o + «Ham 2 Ho.o + HomlmHHB w.ev «.mw m.wm HUM 2 mH.o + HomImHuB «.mm c.mmH H.¢OH «Ham 2 Ho.o + HomImHHB o.mm e.mm «.mm HomImHuB mCHHmmomHfid mCHummomHfifi AHouuCoo mo EvIOH o v Em|OH o v muH>Huom oHuHoomm wV mmEOmouon Cmmz wuH>Huo< mmmuoscmm ou coma mummusm AHouucoo mo muH>Huom oHuHommm wV o mEOCCooumolmmodz muH>Huom mmmHmCumeo oCHH>m0CH8¢ .AAmHoHnouoz can neocuon onCuxHE CoHquCoCH mCu CH mmocu qummnmou mHm>mH mCHHHQOCHEd .uCoEummHu HmCuusm 0C quBCmUCC COHC3 AHoumo>Hm womV HomanuB CH Umuoum mmocu mums moEOmOHOHE Honu nCOO .umuusn HomImHuB CH pocCommCmmu UCC .mmusCHE om COM moooemOH um COHummCHHHuCTU .ov um mmusCHE om How mCHuCQDOCH .mummmsn mumHquummm ecu CH .AHoumomHm womV Hum ImHuB CH omHI um cououm .mmEOmouoHE mCHCCmmmsm we pwEHOHHTQ mm3 mCHCmmz .AmHmHumumz CCC mUOCuon N cocumz ou mCHouooom mumu meE HouuCoo Eoum UmumHomH omonu mums poms mmEomOHOHzH .mmahsco HmEomouoHE Co mCHCmms Huxumeom mo muommmm mCBII.HH mqmda 70 resulting from washing microsomes with Tris-HCl buffer and Tris-HCl buffer, 0.15 M in KC1. The increased inhibition of reductase activity observed by the latter may be interpreted as a decrease in the protein extraction by the presence of KC1. It must be stressed here that washing microsomes with 0.15 M KC1 was reported not to inhibit or extract NADPH-cytochrome c reductase (52); interpretations of this data should bear this fact in mind. The apparent increase in reductase and demethylase activities resulting from washing microsomes with Tris- HCl buffer (0.01 M EDTA) and Tris-HCl buffer (0.01 M EDTA + 0.15 M KC1) may reflect an extraction of proteins, e.g. ribosomal, not associated with these activities. It must be mentioned here that in interpreting these data one should bear in mind the fact that they represent differ- ences in specific activities of the extracted pellets as compared to the control, Table 11, and that these treat- ments were found to extract different protein species (Figure 1, tube 6 and Figure 5, tube 6). Salt extractions: Effects on enzymatic activities of microsomes.--High concentrations of salts, especially those containing monovalent anions, were found effective in extracting microsomal membrane proteins (Tables 5-7). The effects these salts have on reductase and demethylase activities were determined in order to evaluate the mild- ness of these treatments. These activities were determined 71 from microsomes "incubated" and "extracted" with concen- trations of the salts listed in Tables 12 and 13. Incu- bation of microsomes with these salts was performed by suspending aliquots of rat liver microsomes (stored in Tris-HCl buffer, 50% glycerol at -15°) to a concentration of 4.6 to 4.8 mg protein/ml in Tris-HCl buffer containing various concentrations of salts. Aliquots of this was added directly to an assay mixture containing 2.5 umole NADPH instead of the NADPH-generating system described in Methods and Materials. This was done in order to avoid any effects that these salts might have on isocitrate dehydrogenase, used as part of the NADPH generating system. Extraction of microsomes with salts was performed according to the procedure outlined in Methods and Materials which involved centrifugation of the "incubation" microsomes for 2 hr at 105,0009 and resuspension of the pellet in Tris- HCl buffer. The effects of the salts investigated (KZHPO NaBr, 4t and NaCl) on reductase and demethylase activities are summarized in Tables 12 and 13. The data shown in Table 12 indicates an activation of NADPH-cytochrome c reductase by salts containing monovalent anions, e.g. NaBr and NaCl, especially at high concentrations (1.0 M). Salts contain- ing polyvalent anions, i.e., KZHPO4, had little effect on these activities. Salt extractions were found, generally, not to extract NADPH—cytochrome c reductase as deduced 72 TABLE 12.--The effects of salts on NADPH-cytochrome c reductase activity. [Microsomes used in these studies were isolated from control rats according to Method 1 (Methods and Materials). Enzymatic activities were determined on microsomes incubated for 1 hr in Tris- HCl buffer containing various levels of salts and either underwent no further treatment (salt incubation) or were centrifuged at 105,0009 for 2 hr (salt extraction)]. NADPH-cytochrome c Reductase Activity Salts and . . . . Concentrations (% speCIflc actIVIty of control) Used Salt Incubationa Salt Extractionb 0.5 M KZHPO4 102.5 116.5 1.0 M K2HPO4 100.5 113.7 0.25 M NaBr 102.2 109.4 0.5 M NaBr 104.2 115.6 1.0 M NaBr 120.0 103.5 0.25 M NaCl 111.8 87.2 0.5 M NaCl 114.3 88.0 1.0 M NaCl 120.0 111.6 aControl was prepared by dilution of microsomes with Tris-HCl buffer. bControl was prepared by extracting microsomes with Tris-HCl buffer. 73 TABLE l3.--The effects of salts on aminopyrine demethylase activity. [Microsomes used in these experiments were iso— lated from control rats according to Method 1 (Methods and Materials). Aminopyrine demethylase activity was deter- mined on microsomes incubated for 1 hr in Tris-HCl buffer containing various levels of salts and either underwent no further treatment (incub.) or were centrifuged at 105,0009 for 2 hr (ext'd.). Aminopyrine levels of the incubation mixtures (Methods and Materials) are given]. Aminopyrine Demethylase Activity (% specific activity of control) Salts and _3 _3 _4 Concen- 4.0x 10 M 1.0x 10 M 4.0x 10 M trations Aminopyrine Aminopyrine Aminopyrine Used incub. ext'd. incub. ext'd. incub. ext'd. 0.10 M K2P04 ----- 107.0 ----- 131.3 ---------- 0.50 M K2P04 110.0 121.0 ----- 165.3 212.0 134.5 1.00 M K2P04 114.0 111.5 ----- 130.3 245.0 ----- 0.05 M NaBr ----- 88.8 ----- 101.1 __________ 0.25 M NaBr 77.9 ---------- 112.1 145.5 _____ 0.50 M NaBr 105.0 50.1 ----- 65.1 357.5 65.3 1.00 M NaBr 56.3 34.7 ----- 54.1 251.0 ————— 0.05 M NaCl ----- 88.6 ----- 104.5 —————————— 0.25 M NaCl 89.4 93.2 ----- 95.4 100.0 _____ 0.50 M NaCl 91.4 66.2 ----- 71.5 227.5 42.7 1.00 M NaCl 66.6 ---------- 88.6 178.0 ————— Note: Control used for incubation was prepared by diluting microsomes to desired concentration with Tris-HCl buffer whereas extracted control was prepared by a 105,0009 centrifugation of incubation control for 2 hrs. 74 from the increase in reductase activity as compared to control. The effects of these salts on aminopyrine demethylase activity is summarized in Table 13. Offhand, these data seem to indicate an inconsistency with respect to effects of the different salts on demethylase activity at various substrate concentrations. However, if one keeps in mind the fact that aminopyrine demethylation does not follow Michaelis-Menton kinetics (73), as mentioned earlier, one may rationalize that these salts do not effect all com- ponents of demethylase activity in the same manner. For example, K2HP04 was shown generally to increase demethylase activity, with a greater effect observed at low substrate concentrations. On the other hand, NaBr and NaCl were shown to inhibit demethylase activity at high substrate concentration but increase demethylase activity at low substrate concentrations. The effects of increasing con- centration of these salts on demethylase activity are diffi- cult to explain. It appears that the salts containing monovalent anions shows an optimum salt concentration of 0.5 M for demethylase activity, whereas K2P04 shows an increase in activity in the presence of increasing salt concentrations. DISCUSSION Fractionation of Microsomes Fractionation on a discontinuous sucrose gradient in + ++ . . . the presence of Cs and Mg ions.--Rat liver microsomes were fractionated on a discontinuous sucrose density gradient in the presence of Cs+ and Mg++ ions according to the method of Dallner et a1. (48). The fractionation achieved resulted from changes in density of the micro- somal subfractions due to their ability to specifically + or Cs+, e.g. RER and SER I subfractions bind Cs+ bind Mg+ and Mg++ ions, respectively, whereas the SER II subfraction does not. That a rather effective fractionation was achieved by this method is suggested by the quantitation studies, Table 1, which show an approximate four—fold difference in RNA—ribose content of RER over that of whole SER. A further indication of the effectiveness of this fractionation pro- cedure is shown in the electrophoresis profiles of the various fractions, Figure 1. For example, the electrophoresis profile of the RER fraction (Figure 1, tube 2) contains some faster migrating protein bands which are not present in any of the SER fractions; because it is known that only the RER fraction contains ribosomes, these proteins are assumed to be ribo- somal. Support is given to this assumption by the fact that 75 76 Tris-HCl buffer containing 0.01 M EDTA and 0.15 M KC1, a mixture which should disrupt the forces thought to be responsible for the attachment of ribosomes to RER (i.e., a combination of Mg++ complexing and electrostatic bind— ing), extracted some of these proteins (Figure 1, tube 6). One puzzling observation was made with respect to the electrophoresis profile of SER II (Figure 1, tube 5). There is one protein band, migrating just ahead of SP, which is present in all of the microsomal subfractions but seems to be specifically concentrated in this fraction. No attempts at identification of this band have been made, but judging from the relatively intense reddish color of this fraction one may suggest that it is one of the micro- somal cytochromes. The recent findings by Holtzman gE_§l. (57) that SER contains more cytochrome P-450 per milligram of protein than RER, taking into account the ribosomal con- tent of this fraction, may suggest its possible identity. However, it must be mentioned that there was no correlation shown in Holtzman's data indicating the distribution of cytochrome P-450 between the subfractions of SER. Tertiaryramyl alcohol treatment.--Treatment of micro— somes with tert-amyl alcohol was reported by MacLennan §£_gl. (31) to redistribute the lipid in microsomes such that a microsomal electron transport membrane, which apparently has a higher affinity for lipid than the other microsomal components, increases its lipid content. This 77 membrane fraction becomes less dense and could be separated from the other microsomal components by centrifugation. The method as used here, however, failed to yield the three fractions reported by MacLennan (31). Instead, four frac— tions resulted one of which appears to be a mixture of two of the other fractions (i.e., the loose reddish pellet, containing the microsomal electron transport membranes, and the tight brownish pellet) indicating incomplete separation. Quantitation of lipid-phosphate and RNA- ribose content of the four fractions obtained also suggest considerable cross-contamination, possibly resulting from incomplete sedimentation of the tert—amyl alcohol treated microsomes. Extraction Procedures Organic solvent extraction.--It has been generally accepted that the forces primarily responsible for membrane integrity are hydrophobic. According to the membrane model presented in the Introduction and Literature Review of this thesis, these forces express themselves in the interactions between the lipoprotein subunits of membranes. Attempts were made in these studies to remove weakly bound proteins from beef liver microsomal membranes by successive washings with pyrophosphate buffers (0.1 M and 0.02 M, pH 7.8) and to resolve the unextracted membrane proteins by organic solvent treatments which are known to weaken the hydro- phobic bonds (58). Of the organic solvents initially 78 investigated (glycerol-butanol, 1:4 v/v; butanol; dioxane- water, 1:4 v/v; dioxane-water, 4:1 v/v; pyridine; N,N'- dimethylformamide; N,N'-dimethylformamide-water, 1:1 v/v; and 2-aminoethanol) dioxane-water (4:1 v/v) was found most effective with respect to protein solubilization. Also within this series of organic solvents a rough correlation was observed between the extent of protein solubilization and the dielectric constant of the solvents. For example it was found that the order of decreasing effectiveness of protein solubilization by these solvents was, dioxane-water (4:1 v/v)>2-aminoethanol>N,N'-dimethylformamide. Divalent cations, e.g. Mg++, were found to enhance the protein solubilization by organic solvents, Table 2. It must be mentioned that the numerical values given for the per cent protein solubilization in this table are only relative and do not represent the exact amount of protein being solubilized (see explanation under Table 2). It is interesting, however, that a similar enhancement in solu- bilization of beef heart mitochondria treated with various aliphatic alcohols and with compounds of which diethyl- stilbestrol is a prototype, was observed by Byington et_al. (55). These authors observed a divalent cation, e.g. Mg++ and Ca++, dependent solubilization of up to 30% of the mitochondrial membrane with diethylstilbestrol. Though similar treatments of microsomal membranes with this compound were investigated in these studies, essentially no such solubilization was observed. 79 The mechanism by which divalent cations enhance protein solubilization is not clear. Byington e£_31. (55) suggest that the combined action of diethylstilbestrol and divalent cations effectively weakens mitochondrial mem- branes by causing changes in the conformations of the lipoprotein subunits. They, however, fail to suggest a specific role for the divalent cations in enhancing protein solubilization by diethylstilbestrol or organic solvents. One may postulate a function for these ions if it is taken into account that they (e.g. Mg++, Mn++, Ba++, Ca++) are good protein denaturants (59). These ions could exert their enhancing effect on protein solubilization by de- naturing proteins, subsequently exposing their hydrophobic regions to the effects of organic solvents or molecules such as diethylstilbestrol. Though there is a degree of speculation involved in this explanation, it seems reason- able for the data observed. Lysis of rat liver microsomes.--Lysis or osmotic shocking of microsomes by dilution with distilled water was originally investigated as a method of releasing the soluble contents of microsomal vesicles. It was observed that an increase in the amount of protein solubilized (6 to 14%) resulted from lysis of microsomes stored at -15° in Tris-HCl buffer, 50% in glycerol as compared to those stored in 0.25 M sucrose. This effect was initially contributed solely to the effects of freezing on micro— somes. Microsomes stored at -15° in 0.25 M sucrose were 80 subjected to the formation of ice crystals, a condition not present when microsomes are stored in Tris-HCl buffer (50% in glycerol), which are known to disrupt hydrophobic bonds causing denaturation and/or percipitation of some proteins (60). This effect was later observed to be due to glycerol, in that an increase in protein solubilized by lysis of 0.25 M sucrose stored microsomes in 2.76% glycerol (a concentration which approximates that in the lysis experiments using microsomes stored in Tris-HCl buffer, 50% in glycerol). The observation was made in a separate experiment that lysis in the presence of 2.76% glycerol of freshly isolated microsomes and of microsomes stored at -15° for two days in Tris-HCl buffer (50% in glycerol) and in 0.25 M sucrose showed a distinct correlation between the storage of microsomes and protein solubilization. Using microsomes isolated and resuspended in Tris-HCl buffer as the control, it was found that lysis of microsomes stored in Tris-HCl buffer (50% in glycerol) solubilized approxi— mately 2% more protein than control upon lysis and that microsomes stored in 0.25 M sucrose solubilized approxi- mately 4% less than control. These observations would tend to confirm the fact that protein denaturation, possibly expressed as percipitation of some proteins, occurs upon freezing microsomes in 0.25 M sucrose and to suggest storage in glycerol at -15° may enhance protein 81 solubilization. However one must determine whether or not the 2% difference is significant enough to warrant the latter interpretation of this data. A temperature effect was also observed in the lysis treatment of microsomes. An increase in protein solubilization of 7 to 11% results when microsomes are lysed at room temperature for 30 minutes as compared to lysis at 0-5° for the same time period. In both in- stances the enhancement of protein solubilization by glycerol was observed. Such effects indicate that lysis is an energy requiring process; however, the specific role of glycerol is not known. One could postulate that it acts to weaken the bonds responsible for the integrity of the membranes of microsomal vesicles, causing a more efficient rupture of these vesicles. If this were so one should expect a shorter time requirement for lysis in the presence of 2.76% glycerol. However, the fact that there were no differences in the time required for lysis of microsomes in the presence of 2.76% glycerol and with distilled water, as determined by light scattering, would argue against this interpretation. Another possibility is that glycerol may not effect the extent to which lysis occurs but that enhancement in protein solubili- zation is due to the extraction of proteins from the microsomal membranes. Support for this is shown by a comparison of the electrophoresis profiles of proteins 82 solubilized by lysis (Figure 6, tube 10) with those solu- bilized by the various salt treatments (see Figure 5 and Figure 3, tubes 3, 5, 7, 9, and 11). It is observed that these treatments solubilize essentially the same protein species, but differ with respect to their effectiveness. Another observation made was that the proteins solubilized by sonication of beef liver microsomes in the presence of approximately 1.2 M glycerol and 0.12 M KC1 had electrOphoresis profiles similar to the proteins solubilized by lysis. Also, the extent of protein solu- bilization resulting from sonication was less than that observed by lysis of microsomes in 2.76% glycerol; that is, if one takes into account the protein solubilized by the control. These data suggest that lysis in the presence of 2.76% glycerol was effective in rupturing microsomal vesicles and that glycerol extracts a limited amount of protein from microsomal membranes. Acetic acid extraction and structural protein (SP) isolation.—-The method of extracting membrane proteins with 1.4% acetic acid (pH approximately 3.1) was investigated here because it was found by Zahler et_31. (44) to success- fully extract from mitochondrial membranes a protein fraction electrophoretically identical to the SP of Green g£_gl. (10). Since the extraction was achieved without destroying mitochondrial membrane integrity, it was sug- gested that SP originates from the headpieces (43) ob- served in the electron micrographs (9) of mitochondrial 83 membranes. These findings led to the postulation of a membrane model composed of two structural or non-catalytic classes of proteins; i.e., one representing that isolate- able with detergents (10, 13) (SP) and another, termed "Core Protein" (CP) (43), located in the basepieces of membranes. It seems reasonable that microsomal membranes may also exhibit a similar structural organization, i.e., a readily detachable component and a basepiece or membrane forming component. Evidence was obtained in these studies which could possibly support this postulate. Extraction of a protein species, electrophoretically identical to the major protein species of Crude SP isolated from control microsomes with detergents (see Methods and Materials), was achieved by treatment of microsomes with 1.4% acetic acid (Figure 3, tube 3 and 7, tube 5). It was observed that this protein species does not represent SP as defined in these studies; i.e., the major protein species in the electrophoresis profile of whole microsomal membranes. Another observation made was that the electrophoresis pro- files of microsomes isolated from the livers of rats treated with phenobarbital showed an increase of SP as defined in these studies, Figure 7. Phenobarbitol causes a prolifer- ation of the smooth endOplasmic reticulum of rat livers (56) and should likewise be expected to cause an increase in the proteins responsible for the structure of membranes; i.e., those which compose the lipoprotein repeating 84 subunits (3,4) or basepieces of membranes. These obser- vations support the definition of SP in these studies and suggests that its function is analogous to CP of mito- chondria. Further support was given for the structural function of the proteins species when it was found to be closely associated with lipid. Extraction of most of the contaminating microsomal proteins with 2.0 M KSCN (Figure 7, tube 3) gives a protein fraction with a lipid content two- to four-fold that of the control, Tables 5-7. It must be stressed that the exact function of the protein, designated as SP in these studies, has not been conclusively demonstrated neither has it been made clear as to the number of proteins involved in maintaining the structure of microsomal membranes. It is conceivable that a protein may function both enzymatically and structurally; examples of this have been shown with respect to the four electron transport complexes of mitochondria (38, 39) (see also the Literature Review section of this thesis). An example of an enzymatic function for SP, isolated accord- ing to Green et a1. (10), was demonstrated by Woodward and Munkres (61) that SP from respiratory-deficient mutants of Neurospora differs from that of the wild-type by a single amino acid replacement. Further support for the postulate, that proteins may function both structurally and enzymati- cally, is given by the data of Ernster and Orrenius (56) which show that proliferation of rat liver SER by 85 phenobarbitol treatment also induced aminopyrine demethy- lase, cytochrome P-450, and NADPH-cytochrome c reductase activities. It is possible that the increase in the pro- tein bands seen in the electrOphoresis profile of pheno- barbitol treated rat liver microsomes (Figure 7, tube 1) may represent an increased synthesis of the enzymes mentioned above. Past failures to demonstrate an enzy- matic function for SP isolated by detergents (10, 13) from different membrane systems (e.g. mitochondria, microsomes, chloroplasts, etc.) may reflect denaturation of enzymes by these isolation procedures. The recent findings of Schatz and Saltzgabor (62), that SP isolated from beef heart mitochondria contains a considerable amount of denatured ATPase supports this postulate. These data warrants a review of the identity and function of the proteins pre- viously defined by Green g£_al. (10) as being SP. The exact number of proteins involved in the structural integrity of microsomal membranes cannot be clearly pinpointed. Electrophoresis profiles of whole microsomes generally contains several protein bands with electrOphoretic mobilities very close to that of SP, Figure 7. It was usually observed that one of these bands, defined as SP in these studies, was more heavily stained than the rest. This observation was made with gels stained both with Coomassie blue and Amido Schwartz; however, visualization of this band was usually better 86 when Coomassie blue was used to stain the gels. It was impossible to determine from the electrophoresis profiles whether or not these bands represent polymers of SP or species of a class of SP, having very similar properties. The latter has been suggested for SP recently isolated and purified from mitochondrial membranes (17, 43) and it may be possible to extrapolate the concept of a class of SP to other membrane systems. Salt extractions.--Determination of the exact functions of the proteins associated with the membrane forming subunits, the basepieces free of all readily detachable membrane components, requires a pure membrane preparation. Salt extractions of microsomal membrane pro- teins were investigated as a relatively mild method of re- moving proteins loosely attached through electrostatic interactions. A variety of salts was used and their effec- tiveness in extracting non—lipoprotein, i.e., ribosomal and other proteins not essential to membrane structure, from microsomal membranes was determined, Tables 5-7. It was found, generally, that the extent of protein extraction was not a function of ionic strength but was due to specific properties of the salts used. For example the observation was made that extraction of beef liver microsomal membranes with salts containing monovalent cations (e.g. NaBr, KSCN, KNO3, and NaCl) extracted essentially the same protein species as did the control, Figure 3; however, the extent to which the speCies were extracted differed significantly, 87 Table 6. Quantitation of the per cent protein solubilized and the lipoprotein content of pellets extracted with rela- tively high salt concentrations, 2.0 M, showed an increas- ing order of efficiency with respect to non-lipoprotein solubilization with the salts: KN03