. N . . "I. ; 5.x m... Wm m... w... . .uru....r.¢....u.¢w 1.... hr 4 24*...» Law.” 00:...U .34... + 3-1;er 02:0 .. L ..- . . . . .. ._ x .. .... ._ .2... .. . . . . . . . .... , .. .fodf.fl...wa..r at? .x a.” m 05v In»? 95%... G v ' . ~ I. [I .1! Wm ”E ES . _ . . T..... . ...._..”.I LII... . . . . . .. .F. N. .. H. . A»... . . . . . . v .1 . , 0, m . . If; . . AU. .. . . T 154......» . . v» . . .. M ~ H . ”N . , inn... 0N. ._ Hm ..—..-.— mil... Tull; T! C. , C. .A . . _ H . m n . . .....v 2.. 31.1.... .. I... ......K.......... . . . . “.qu ‘L4...$h¢i....1h.ru’tt.hnlifi .43.... 11.5313... 3.4.32... .r.J\:...Ao. L3H ....4 .1213.” .. 3.... I514..HH_JVH...N.;H£ . Urn. TI "suit. - 5 .‘bbkuull .1: In. . LIBRA NY ‘7 Michigan 'ate Univcrs. y ’. go This is to certify that the thesis entitled -' CHARACTERIZATION OF THE PHOSVITIN FRACTION 0F AVIAN EGG YOLK __‘-‘ presented by Richard Charles Shantz has been accepted towards fulfillment of the requirements for Ph.D. degree in Food Science and Human Nutrition Date June 5; 1973 ‘K ABSTRACT CHARACTERIZATION OF THE PHOSVITIN FRACTION OF AVIAN EGG YOLK BY Richard Charles Shantz This study was undertaken to electrophoretically demonstrate the protein components of avian egg yolk, and to chemically and physically characterize the phosvitin fraction. Avian egg yolk was examined using acrylamide disc gel electrOphoresis. Concentrations of 5, 6 and 7% acrylamide with a Tris-Glycine buffer system, pH 8.3, revealed 29, 29 and 25 components, respectively, in the running gel, with an additional diffuse area throughout the spacer gel. Rf values of all the bands were determined using bromthymol blue as a reference. These values ranged from 0.121 to 0.993. No attempt was made to identify the bands observed with known egg yolk proteins; however, several bands migrated in a manner expected of lipoproteins. Phosvitin was extracted by a salt precipitation centrifugation procedure. This procedure yielded 11 bands, one of which appeared as a large diffuse bubble-like area, as well as a band at the gel interface and a diffuse area in the upper gel. Richard Charles Shantz The procedure for the extraction of phosvitin was elaborated using the Amicon ultrafiltration system. This system yielded 4 fractions. The first had components with molecular weights > 1.0x105, while the second, third and fourth fractions had molecular weights of > 5.0xlO4, > 1.0x104 and < 1.0x104, respectively. The sample size was found to have a very pronounced effect on the electrOphoretic mobilities of the phosvitin components in the acrylamide gels. Larger sample sizes increased the size of the bubble and increased the Rf values of all components with Rf values greater than that of the bubble. Similarly, compenents with Rf Values less than that of the bubble showed reduced Rf values. Electrophoresis in the presence of 4 M urea reduced the number of bands to 9, but did not otherwise alter the electrophoretic pattern of phosvitin to any great degree. Differential staining of the acrylamide gels revealed 1 lipoprotein band at the gel interaface, with the possibility of 4 additional bands being lipoproteins. The interior of the bubble stained heavily for glycoprotein, while the perim- eter of the gel stained only for protein. Several other bands also stained as glycoproteins. Staining for iron showed 7 bands, including the bubble, to be ferroproteins. Molecular weights of all the bands were determined using SDS acrylamide gels. The molecular weights ranged from 9,050 to 118,700., Richard Charles Shantz N-terminal amino acid analyses of the ll bands revealed alanine, arginine, asparagine, glutamine and threonine. CHARACTERIZATION OF THE PHOSVITIN FRACTION OF AVIAN EGG YOLK BY Richard Charles Shantz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science and Human Nutrition 1973 C. ACKNOWLEDGMENTS The author wishes to express his deepest gratitude and appreciation to Dr. L. E. Dawson for his guidance and assistance throughout his course of study and preparation of this manuscript. The author also extends a thanks to Dr. G. A. Leveille and the faculty of the Department of Food Science and Human Nutrition for the facilities and encouragement which were provided. Appreciation is also extended to Drs. P. Markakis and T. Wishnetsky of the Department of Food Science and Human Nutrition, and to Drs. R. K. Ringer and D. E. Ullrey of the Departments of Animal Science and Poultry Science, respec- tively, for serving as members of the guidance committee. A special note of thanks to Dr. J. R. Brunner of the Department of Food Science and Human Nutrition for his advice in running the N-terminal amino-acid analyses, and to Mr. Sulo J. Holkonen of the Department of Poultry Science for the photography. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . REVIEW OF LITERATURE . . . . . . . . . . . Egg Yolk Proteins . . . . . . . . . . General Composition and Structure of Avian Egg Yolk . . . . . . . . . . . . . Composition and Structure of Phosvitin . . Phosphoproteins in Animal Tissues . . . . . Composition and Structure of the Livetins . General Composition and Structure of Lipoproteins . . . . . . . . . . . . . . Composition and Structure of Lipovitellin . Composition and Structure of Lipovitellenin Differential Staining Techniques . . . Protein Staining . . . . . . . . . LipOprotein Staining . . . . . . . Glyc0protein Staining . . Hemoprotein Staining . . . Molecular Weight Determination in SDS Polyacrylamide Gels . . . . . . . . . N-Terminal Amino-Acid Analysis . . . . EXPERIMENTAL PROCEDURES . . . . . . . . . . Part One: Determination of Rf Values of Yolk Proteins . . . . . . . . . . . . Preparation of Sample . . . . . . . Preparation of Acrylamide Gels . . ElectrOphoretic Procedure . . . . . Rf Measurement . . . . . . . . . . Part Two: Extraction of Phosvitin . . Fractionation of Phosvitin by Ultrafiltration . . . . . . . . . iii Page vi WP on 10 13 16 18 22 23 24 24 25 28 30 31 31 31 33 35 36 39 Part Three: Determination of Rf Values of Phosvitin Components . . . . . . . . Effect of Sample Size on Phosvitin Rf Values Effect of Urea on Phosvitin Rf Values . . Part Four: Differential Staining of Phosvitin Lipoprotein Stain . . . . . . . . . . . . Glycoprotein Stain . . . . . . . . . . . Hemoprotein Stain . . . . . . . . . . . . Part Five: Molecular Weight Determinations . Preparation of Gels . . . . . . . . . . . Preparation of Samples . . . . . . . . . Part Six: N-Terminal Amino Acid Analysis . . Separation of the Phosvitin Components . N- -Terminal Phenylthiohydantoin Amino Acid Development . . . . . . . . . . Identification of Phenylthiohydantoin Amino Acids . . . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . . . . Part One: R Values of Egg Yolk Proteins . . Part Two: Extraction and Fractionation of Phosvitin . . . . . . . . . . . . . . . . Part Three: Determination of Rf Values of Phosvitin Components . . . . . . . . . . . Part Four: Differential Staining of Phosvitin Part Five: Molecular Weight Determinations . Part Six: N-Terminal Amino Acid Analysis . . SUMMARY 0 O O O O O O O O O O O O O O O O O O O O BIBLIOGRAPHY O O O O O O O O O O I O O O O I O I iv LIST OF TABLES Table Page 1. Membranes, Operating Pressures and Starting and Final Volumes Employed in the Ultra- filtration Fractionation of Phosvitin . . . 40 2. Marker Proteins and Their Molecular Weights Used in the Determination of the Molecular Weights of the Components of Phosvitin . . . 45 3. Rf Values of Egg Yolk Proteins Run on 5, 6 and 7% Acrylamide Gels . . . . . . . . 57 4. R Values of Phosvitin Run on 6% Acrylamide Gels at 6 Different Sample Sizes . . . . . . . . . . . . . . . . 72 5. Rf Values of Phosvitin Run on 6% Acrylamide Gels in the Presence of 4M Urea at 6 Different Sample Sizes . . . . . . 77 6. R Values, Expected Molecular Weights, Apparent Molecular Weights and Percentage Deviation from Expected Molecular Weights of the Marker Proteins . . . . . . . . . . . 85 7. Rf Values and Molecular Weights of the Proteins of the Phosvitin Fraction of Egg Yo 1k 0 O O O I I I C O O O O O O O O O O 8 9 8. N-Terminal Amino Acids of the Proteins Of PhOSVitin I O O O O O O O O O O I O O O O 93 Figure 10. 11. 12. LIST OF FIGURES Acrylamide Disc Gel Electrophoresis Apparatus . . . . . . . . . . . . . . . Outline for the Extraction of Yolk Granules From Egg Yolk . . . . . . . . Outline for the Extraction of Phosvitin From the Yolk Granules . . . . . . . . Preparatory Acrylamide Disc Gel Electrophoresis Apparatus . . . . . . . Outline for the DevelOpment of PTH Amino Acids . . . . . . . . . . . . . . Densitometer Tracing of Whole Yolk Run on 5% Acrylamide Gel . . . . . . . . . . . Densitometer Tracing of Whole Yolk Run on 6% Acrylamide Gel . . . . . . . . . . . Densitometer Tracing of Whole Yolk Run on 7% Acrylamide Gel . . . . . . . . . . . Densitometer Tracing of Phosvitin Run on 6% Acrylamide Gel . . . . . . . . . . . Acrylamide Disc Gels Showing the Effect of 6 Sample Sizes (Left to Right 338, 675, 1013, 1350, 1688 and 2025 pg) on the Electrophoretic Mobilities of the 11 Phosvitin Components . . . . . . . . . Acrylamide Disc Gels Run with 4M Urea Showing the Effect of 6 Sample Sizes (Left to Right 338, 675, 1013, 1350, 1688 and 2025 ug) on the Electrophoretic Mobilities of the 9 Phosvitin Components . . . . . Acrylamide Gels of Phosvitin Samples of 1013 and 1688 ug (Left to Right) Stained With Periodic-Acid Schiff Reagent . . . vi Page 34 ~37 38 48 52 58 59 60 67 71 76 80 13. 14. 15. 16. Page Standard Mobility Curve of Molecular Weight ‘ Versus Rf Value of the Marker Proteins . . . 84 Densitometer Tracing of Phosvitin Run on SDS Acrylamide Gels . . . . . . . . . . . . . 86 Plot of Rf Values of the Phosvitin Components Run on SDS Acrylamide Gels, on the Standard Mobility Curve . . . . . . . . . . . 88 Diagrammatic Representation of the Preparative Acrylamide Disc Gel of Phosvitin Showing the Sectioning Used to Obtain Pure Protein . 92 vii INTRODUCTION Eggs possess many as yet undefined chemical and physical characteristics which enable them to perform spe- cific functions in food products. Egg albumen is used in preparing meringues because of its ability to form a stable foam. Egg yolk is uSed in making mayonnaise because of its emulsifying ability, and is also similarly used in baked goods. Whole eggs are employed in the preparation of cus— tards because of their ability to coagulate at relatively low temperatures. In addition, eggs are known to undergo changes in functional properties when pasteurized or spray dried, as well as changes in physical characteristics when frozen and thawed. Egg albumen contains 10.9% protein which is generally considered to consist of ovalbumin, ovomucin, ovomucoid, conalbumin, lysozyme, avidin and 2 globulins. In addition, albumen contains carbohydrate, ash, fat and water (watt and Merrill, 1963). Similarly, egg yolk contains 16.0% protein consisting of alpha, beta and gamma livetins, lipovitellenin, lipovitellin and phosvitin, as well as fat, carbohydrate, ash and water (Watt and Merrill, 1963). Little is known, however, about the relationship between these various components in egg albumen and between those in egg yolk. l In 1970 Chang gt_gl. used polyacrylamide disc gel electrophoresis to demonstrate 12 bands in egg albumen and 19 bands in egg yolk. These were considerably more frac- tions than were previously thought to exist in either albumen or yolk. Preliminary research in this laboratory had demon- strated a still greater number of bands in disc gels of egg yolk, and furthermore the phosvitin fraction of yolk was found to be much more complicated than had been previously described. The present study was undertaken to determine the maximum number of bands which could be demonstrated in egg yolk using acrylamide disc gel electrophoresis. Further, it was decided to study the phosvitin fraction in detail. The existing fractionation procedure was expanded to yield sev— eral phosvitin sub-fractions. Acrylamide disc gel electro— phoresis was then used to examine the electrophoretic mobili- ties of the phosvitin components under various conditions. In addition, differential staining was used to determine the chemical nature of the phosvitin proteins. Finally, the molecular weight and N-terminal amino acid of each of the- phosvitin proteins was determined. REVIEW OF LITERATURE Egg Yolk Proteins General Composition and Structure of Avian Egg Yolk Avian egg yolk contains a variety of microscopic particles belonging to 2 main groups, the yolk "globules" which are large particles, many of them resembling oil drop- lets, and the "granules" which are smaller and more uniform in size but less regular in shape (Romanoff and Romanoff, 1949). Evans and Bandemer (1957), using paper electro- phoresis, found the protein fraction of egg yolk to contain 46.4% lipovitellin, 41.7% lipovitellenin, 8.6% livetin and 3.3% other proteins. Clegg gt_gl. (1971), using diethyl- aminoethyl (DEAE) pH gradient chromatographic techniques as well as ultracentrifugal techniques, also found 4 components in egg yolk. Egg yolk proteins have been separated on the basis of their density into a high density fraCtion (HDF), and a low density fraction (LDF). The HDF is generally considered to consist of phosvitin and lipovitellin (Joubert and Cook, 1958a; Bernardi and Cook, 1960a: Burley and Cook, 1961: Schmidt gt_gl., 1956), as well as a-, 8-, and y-livetins (Martin et_gl,, 1957; Martin and Cook, 1958). McCulley gg_gl. 3 (1959) showed by paper electrophoresis that the granule matter is made up of lipovitellin and phosvitin, but that the granular matter may be contaminated with livetins if the centrifugation procedure is extended unduly. Hurley and Cook (1961) extracted granules by centri- fuging yolk diluted with an equal volume of 0.16 M sodium chloride. They found the granules to contain 48—52% water, and stated that the granules represented 11.5% by weight of the original yolk or about 23% of the yolk solids. They also reported that the granules contained 90%. of the protein phosphorus, 41% of the total phosphorus, 95% of the iron and 70% of the yolk calcium: and consisted of 70% a- and B-lipOproteins, 16% phosvitin, and 12% low density lipoprotein. They found that washing the granules removed the soluble livetin contaminants: however, they could not remove the low density lipoprotein (LDL) by washing and therefore concluded that one or more LDL formed an integral part of the granules. In addition, they reported that the granules dis- solved in salt solutions at mildly alkaline pH's, and con- cluded that the constituents are evidently held together by ionic or secondary forces. Bernardi and Cook (1960a) reported that the HDF represented 31% of the yolk solids, slightly higher than reported by Burley and Cook (1961), while phosvitin was 15% of this HDF and the lipovitellins about 50% of the HDF, the latter being considerably lower than reported by Burley Hand Cook (1961). Using moving boundary electrophoresis to separate the components of the HDF, Bernardi and Cook (1960a) found 3 phosvitins, a-, 8-, and yhlivetins as well as a- and B- lipovitellin. In the ultracentrifuge, however, only 2 boun- daries were apparent with both containing 2 or more compon- ents. The authors suggested that a- and B-lipovitellin behaved as a reversible association-dissociation system. Warner (1954) stated that hen's egg yolk contained about 0.6% phosphorus in the form of both phospholipids and phosphOproteins. Schmidt g£_§l. (1956) reported that 70% of these phospholipids remained in the supernatant yolk fraction after removal of the granules, and were thus not associated with the phosphoproteins. Radomski and Cook (1964a), using triethylaminoethyl (TEAE) cellulose chromatography, provided evidence for the fact that granules contained a monomer lipovitellin-phosvitin unit as well as its polymers. At ionic strength of u = 0.05 and above pH 9.4 phosvitin was not evident as a separate component in the ultra-centrifuge but appeared as the ionic strength was increased. The supernatant plasma fraction, remaining after removal of the HDF fraction, consists of the soluble lipo- protein, lipovitellenin (Fevold and Lausten, 1946; Martin et a1., 1964; Schmidt et a1., 1956), and the livetin fraction, which by definition is that part remaining after removal of the lipovitellenin. No complexes have been demonstrated in this LDF between the two major protein components. Composition and Structure of Phosvitin Phosvitin was first isolated from avian egg yolks by Mecham and Olcott (1949) using a technique involving a salt precipitation and subsequent centrifugation. They reported that the protein contained 10% phosphorus which represented 70% of the total yolk phosphorus. When examined electrOphoretically, phosvitin appeared heterogeneous, but ultracentrifugal analysis showed it to be homogeneous. Their procedure did not establish whether phosvitin was a separate protein, or an integral part of the lipovitellin molecule. Mok gt_al. (1961) also reported avian phosvitin to be homogenous. They found the protein to contain 10.1% phosphorus, l residue of tryptOphan and 129 serine units based on a molecular weight of 4.2x104. They suggested that all the phosphorus in the protein was bound to these serine residues. Sundararajan gt_al. (1960), using a centrifugation- ether extraction preparation technique, found 3 phosvitin components on paper electrOphoresis. ~They reported 1 major fast moving band and 2 fainter slower moving components. Sugano (1957), however, found only 2 phosvitin components from yolk on paper electrophoresis. His phosvitin preparation represented 8.5% of the total yolk proteins. Simlot and Clegg (1967) used ultracentrifugation to separate yolk into 3 major fractions, followed by precipita- tion of the heaviest fraction with dextran sulfate to remove any lipid material. They obtained 1 component, as detected by moving boundary electrOphoresis, which contained 4.15% phosphorus. This was the fastest moving component of all the fractions examined. Clegg g£_§1. (1955) also reported that the fastest moving component of egg yolk was high in phosphorus. In 1958a Joubert and Cook develOped a method for the separation of phosvitin from whole avian yolk using a magnesium sulfate, water dilution technique. They obtained a fraction contain- ing 9.6% phosphorus which suggested 94 atoms of phosphorus per molecule based on a molecular weight of 3.04x104. Their ultracentrifugal studies suggested an elongated molecule with an axial ratio of 20:1 and dimensions of 280x14°A. At concentrations up to 0.05M magnesium sulfate the protein had the ability to form complexes with a molecular weight of 1.4x106, while in calcium salts it formed complexes of molecular weight of 7.5x105. Since Joubert and Cook's (1958a) procedure gave both phosvitin and lipovitellin, they established that phos- vitin was not an integral part of the lipovitellin molecule. They suggested that the protein interacted strongly with proteins of Opposite charge, and that although it was a small molecule it might be responsible for the apparent interaction between lipovitellin and the livetins resulting in the protein complex generally known as lipovitellin. Alderton and Periman (1965) subsequently compared the phosvitins of Mecham and Olcott (1949) and Joubert and Cook (1958b) and found them to have similar amounts of phosphorus, ash, nitrogen and amino acids. Connelly and Taborsky (1961) examined the phosvitin Of Mecham and Olcott (1949) and found 2 fractions with the same amount of phosphorus and nitrogen. However, they found that the one fraction contained all the tyrosine, and that this fraction was also responsible for most of the metal binding and alkaline lability prOperties of the unfractioned preparation. Mok g£_21, (1966) also found phosvitin to be hetero- geneous. Using column chromatography they demonstrated 2 phosvitin fractions which had similar nitrogen and phOs— phorus contents. They reported that at least 1 of the 2 phosvitin fractions appeared to be homogeneous (Radomski and Cook, 1964b). A general method for the fractionation of phosvitin from vertebrate eggs was developed by Wallace gt_gl. (1966). The method consisted of first isolating the phosvitin— lipovitellin complex, and then precipitating out the lipo- vitellin to leave the "crude phosvitin." This phosvitin could then be "cleaned up" using DEAF-cellulose chromatog- raphy. They found their preparation to have a molecular weight of 4.0x104 and suggested that the protein might be heterogeneous. Mok §£_31. (1966) suggested that the existence of heterogeneity, which has been observed for phosvitin, might be due to either the method of preparation, heterogeneity in the protein moiety, or to the extent of phosphorylation of the molecule. Mano and Lipmann (1966) subsequently reported evi- dence of discrete phosvitin subfractions in eggs of a number of species. The indicated that the various subfractions rep- resented a similar protein with different levels of phos- phorylation. Mecham and Olcott (1949) reported phosvitin to have alanine as an N-terminal amino acid. This was subsequently verified by Joubert and Cook (1958b). PhosphOproteins in Animal Tissues Phosphoproteins are widely distributed in animal tissues, and are thought to be involved in biological pro- cesses such as enzyme catalysis (EngstrOm, 1961), electron transfer (Boyer, 1963), metal ion transport (Ahmed g£_gl., 1963), and energy storage (Rabinowitz and Lippmann, 1960). The latter have shown that serine and phosphate are present in nearly equal amounts, strengthening the hypothesis of Mok §t_gl. (1961) that this must be the binding site of phosphate to proteins. 10 Avian phosvitin has been shown by Alderton and Perlman (1965) to contain 25% by weight of hexoses which corresponds to 6 or 7 residues of hexose per molecule based on their molecular weight value of 4.0 to 4.5x104. Rosenstein and Taborsky (1970) have shown that phosphate release occurs by P-O bond cleavage in phosvitin, and suggested that in the avian species this protein may serve as an energy source during embryonic development in the egg. Taborsky (1963) reported that when iron was added to phosvitin, the protein rearranged. Coupled with the rearrangement was a rapid oxidation of ferrous ion to the ferric ion. O-phosphoserine showed an affinity for the ferric ion, with 4 coordination sites of the ion filled by phosphate oxygen, and the other 2 probably filled by water or some anion present. This might have accounted for the fact that the ratio of phosphorus to iron was about 2 in their phosvitin. Composition and Structure of the Livetins In 1957 Martin gt_21. described a technique for the fractionation of livetins from egg yolk. The method was based on the removal of lipovitellins by dilution, and the subsequent removal of lipovitellenin with an ether extrac- tion and centrifugation to yield a preparation of water soluble a-, 8-, and y-livetins. They reported the molecular 11 weights of the a- and B-livetins to be 6.7 and 4.5x104,‘ respectively, as determined by the Archibald technique. Martin and Cook (1958) subsequently fractionated y-livetin from the original preparation with 37% saturated ammonium sulfate @4°C and determined its molecular weight to be 1.5x10S by the Archibald technique. Bernardi and Cook (1960a) fractionated livetins from the HDF of egg yolk with an extremely mild procedure using various concentrations of magnesium sulfate and centri- fugation techniques. They were able to remove the Y-livetin from solution, but could not separate a- and B-livetins. Williams (1962a) used starch gel electrophoresis in combination with immunoelectrophoresis to identify the yolk livetins with serum proteins. He found that a-livetin corresponded in mobility to serum albumin, while B-livetin corresponded to az-glycoprotein, and y-livetin to y-globulin. He suggested that there is a transfer Of proteins from the plasma of laying hens to the yolks of their eggs. Shepard and Hottle (1949) and McCulley gt_31. (1959) examined the livetin fraction of avian yolk using moving boundary and paper electrOphoresis, respectively. Both groups Of workers Observed three fractions of livetins. Mok and Common (1964), using the preparation of Bernardi and Cook (1960a), demonstrated 6 distinct fractions immunoelectrophoretical1y, 4 of which corresponded to a-, 8-, y-, and a yz-livetin as Observed on paper electrOphoresis. 12 Mandeles (1960), however, reported the separation of 10 livetin fractions using DEAE-cellulose chromatography employing the preparation of Martin gg_gl. (1957). These were more fractions than had been previously demonstrated and more than the authors were able to observe with paper electrophoresis. In 1966, Hui and Common examined the livetins of hen's egg prepared by the method of Bernardi and Cook (1960a). They demonstrated 16 zones of starch gel electrophoresis, 1 of which corresponded to a-livetin, 4 to B-livetin thereby establishing its electrophoretic heterogeneity, and 2 others which were identified as transferrins. The remaining 9 zones- were not identified. Several other groups of workers have found other proteins in the livetin fraction. Williams (1962a) reported transferrin in the livetin fraction of egg yolk, and subse- quently (Williams, 1962b), found traces of conalbumin in yolk. He reported that both transferrin and conalbumin were glyc0proteins which differed only in their carbohydrate groups, while the protein parts appeared identical. Marshall and Deutsch (1951) found traces of oval- bumin, while Lineweaver gt_21, (1948) found tributyrinase, amylase, phosphatase, catalase and peptidase activity. Several workers have reported on the N-terminal amino acids of the livetins. Martin §t_al, (1957) found y-livetin to have arginine and lysine as N-terminal amino 13 acids, while showing B-livetin to have lysine plus 2 others. Martin and Cook (1958) reported y-livetin to have alanine as an N-terminal. General Composition and Structure of LipOproteins Natural lipoproteins contain proteins, phospholipids, and neutral lipids, combined in reasonably constant, but non-stoichiometric proportions through forces weaker than the covalent bond (Cook and Martin, 1962a). Each of these com- ponents has characteristic properties that could affect lipoprotein structure. Proteins can solubilize lipid and, being large molecules, confer structural integrity. Phos- pholipid can form micelles and also solubilize neutral lipid but, being small molecules, would contribute less structural integrity. Likewise, neutral lipids are small molecules that must be solubilized by the other components and might therefore be considered "dependent" components (Cook and Martin, 1962a). Two models of lipoprotein structure have been pro- posed or implied by different investigators (Pankhurst, 1949; McFarlane, 1949; Dervichian, 1949). One is a molecular model where the protein is believed to interact with lipid in layers without losing its molecular integrity. In the micellar model, there is a lipid core surrounded by a film of protein, or protein and phospholipid. Brundorfer and Green (1967) have prOposed that in the latter model the phospholipids at 14 the interface are oriented with their polar groups in con— tact with the aqueous phase, while their hydrocarbon residues are immersed in the lipid globule. Cook and Martin (1962b) examined lipoproteins from diverse sources and found that lipOproteins with lipid con- tents extending over an essentially continuous range show strong evidence of a transition in their composition in the region where protein, phospholipid, and neutral lipid are present in essentially 1:1:1 proportions. Those with protein contents below 33% (low protein lipoproteins [LPL]) have neutral lipid-phospholipid ratios ranging from 1:1 to 10:1 but this ratio remains at about 1:1 in all those with pro- tein contents above 33% (high protein lipOproteins [HPL]). Cook and Martin (1962b) further noted that dissociation of HPL is pH dependent. Evans et_31. (1968) further strengthened the theory that LPL or LDL have a micellar structure. Using the LDL of egg yolk, which contained 86% lipid, he suggested that the hydrophobic groups were on the inside, while the hydro- philic groups were on the outside making it very soluble in water, in spite of its high lipid content. Scannu (1967) has examined HPL. He reported that when these lipoproteins were extracted with ethanol-ethyl ether, the apOprotein recombined in definite ratios with the whole phOSpholipids of the HPL suggesting that the HPL is made up of subunits held together by lipid bridges. Weinman 15 (1956) has further reported that all the lipids in egg are somehow bound to these proteins. Schneider and Tattrie (1968) examined the mutual solubility of the lipid components. They observed that in the presence of water, the lipids of the low density fraction of yolk partitioned into a hydrated phospholipid-cholesterol phase and a neutral lipid phase which was characterized by the absence of phospholipids. They reasoned that this par- titioning must have depended upon some of the physiochemical prOperties of the lipids, and that the same properties might play a role in controlling their structural arrangement in the native lipOprotein. Thus, since the Observed partition- ing was similar to that expected for the interfacial region and core interior of lipid-core particles, it was taken to support the theory that, in the native low density fraction, the spatial location of the lipids was controlled by their interactions with each other and with water. The protein then probably did not play a major role in ordering the lipids in space, but probably stabilized the lipid globule against aggregation. Cook gt_al. (1962) and Gornall and Kuksis (1971) have shown that the LDL and HDL are not artifacts of each other. Cook gt_al, (1962) have shown that the 2 lipopro- teins have very different protein moieties which cannot be interconverted merely by a gain or loss of lipid. Gornall and Kuksis (1971) found major differences in the lipid 16 composition of the egg yolk lipoproteins. The cholesterol, triglycerides and phospholipids varied to a great extent while there were lesser differences in the proportions of the various phospholipid classes. Composition and Structure of Lipovitellin Lipovitellin was first fractionated by Alderton and Fevold (1945), who reported that their preparation contained 16—18% of a phosphatide. In 1956 Vandegaer g£_21, examined lipovitellin and demonstrated the damaging effect of ethyl ether on the lipOprotein. They reported a loss of stability as well as impaired solubility, pointing out the sensitive nature of the molecule. In 1958(b), Joubert and Cook demonstrated the phosvitin-lipovitellin complex. Their preparation contained 20% lipid and had a molecular weight of 3.2x105. The behavior of lipovitellin indicated that it consisted of 2 parts held together by secondary valence forces and the 2 parts were found to be nearly identical in size. The 2 parts of lipovitellin, a- and B-lipovitellin, were shown by Sugano (1959) to be electrOphoretically homogen- eous in the pH range of 6.8 to 9.9 and 4.9 to 10.5, respec- tively, while at pH 4.2 both were heterogeneous. Ultracen- trifugal analysis at pH 9.8 and u a 0.15 indicated the heterogeneities of both lipoproteins. 17 Bernardi and Cook (1960a) also found that a- and B-lipovitellins were heterogeneous. They reported that a-lipovitellin dissociated into smaller units at pH 10.9 while B-lipovitellin dissociated at pH 9.0, and suggested that a- and B-lipovitellin behaved as a reversible association- dissociation system. Bernardi and Cook (1960b) further reported that a- and B-lipovitellin had the same molecular weight but differed slightly in volume or shape. They sug- gested that either the a-subunits or B-subunits were identical. The d- and B-lipovitellins have been shown to behave differently when chromatographed on Dowex 1. Radomski and Cook (1964b) found B-lipovitellin to be homogenous, since it eluted at 0.6 M phosphate buffer, while a-lipovitellin was eluted over a concentration range of 0.7 to 1.4 M. They also found that the phosphorus content changed as the eluant ionic strength increased, further demonstrating the hetero— geneity of the a-lipovitellin.‘ Burley and Cook (1962a) examined the association— dissociation process of a— and B-lipovitellin, and found that the phosphate groups did not take part in the process, nor did the sulphydryl groups. The phosphate groups, how- ever, affected the ease with which the lipovitellin complexed with Ca++ ions, while the sulphydryl groups affected the structure of the molecule. These factors in turn might, however, have affected the dissociation. The authors further 18 stated that the lipid components probably played the most important role in the complex. Burley and Cook (1962b) further reported that the association-dissociation systems of a-lipovitellin at pH 10.5 and B-lipovitellin at pH 7.8 were heavily dependent on ionic strength and temperature. The associated molecular weight of B-lipovitellin was reported to be 4.5x105 by Cook and Wallace (1965), while the molecular weight of the subunit was reported as 2.27x105. The authors did not state whether or not the subunits were identical. Bernardi and Cook (1960b) found lysine and arginine N-terminal amino acids in B-vitellin. This was subsequently confirmed by Neelin and Cook (1961). Lysine and arginine were also found to be the N-terminal amino acids of a-lipovitellin (Joubert and Cook, 1958a; Bernardi and Cook, 1960a). Composition and Structure of Lipovitellenin Lipovitellenin was first extracted from egg yolk by Fevold and Lausten in 1946. They diluted yolk with 2 volumes Of water and centrifuged off the lipovitellin to leave livetins and the lipovitellenin. Upon extraction with ether the lipovitellenin gelatinized and was easily removed. The authors reported that the protein contained 41% phospholipid, and represented approximately 40% of the total lipoprotein of egg yolk and 12-13% of the egg yolk solids. 19 Martin gt_al. (1963) examined the low desntiy frac- tion of yolk and reported it to contain 89% lipid of which 27% was phospholipid, 69% was triglycerides and 4% was choles- terol. They also stated that the low density fraction con- sisted of several polypeptide-phospholipid complexes that combined through interaction with the phospholipids and neu- tral lipids. These values agreed closely with those of Turner and Cook (1958), who reported that lipovitellenin contained 84-90% lipid of which 24% was phospholipid and 65% was neu- tral lipid. Nichols gt_al, (1969) found 66% neutral lipid, 28% phospholipid, and 5% unesterified cholesterol. Nichols gt_31. (1969) further reported a size dis- tribution of lipovitellenin particles ranging from l70-390A with the maximum number being around 250-270A. He calcu- lated the molecular weight of the 250A particles to be 5.2x106. Steer gt__a_t_]_._. (1968) explained the difference in particle size using proteolytic enzymes to remove part of the surface protein from the micellular particles. They reported an increase in particle size after proteolysis and reasoned that a stable particle was obtained only when suf- ficient polar material was present to cover a critical por- tion of the surface. Thus, when protein was removed by proteolysis, an increase in particle size occurred due to aggregation reducing the surface-to-volume ratio to a level permitting phospholipid to provide the required coverage. 20 The authors further reasoned that the lipid core was enclosed in a protein cover. This was in agreement with the findings of Saari g£_gl. (1964), who found that papain could hydrolyze (the protein molecules, and the protein therefore must be on the surface of the particle. This was in agreement with Evans g£_gl. (1968). Martin §E_gl. (1959) removed the lipid portion of -lipovitellenin and reported on the structure of vitellenin, the protein moiety. They found the protein moiety to be unstable in aqueous solvents, and indicated that it appeared heterogeneous even though no significant fractionation took place during centrifugation in solvents of density equal to that of the moiety. They calculated the.molecular weight as 9.3x104, and speculated that at least 2 such chains were present in the lipovitellenin molecule. Sugano and Watanabe (1961), Martin §E_21, (1964) and Augustyniak §t_gl. (1964) reported that lipovitellenin contained 2 low density subfractions. Martin gt_gl, (1964) separated the LDF into LDFl, which represented about 20% of the total LDF with the rest being LDFZ. They reported molec- 6 6 ular weights of 10.3x10 and 3.3x10 for the LDF and LDF l 2' respectively. Since the LDF1 and LDF2 lipid contents were 86.8% and 83.2%, respectively, this meant the size of the protein moieties must have been 1.4x106 and 0.6x106, respec- tively. 21 Augustyniak et a1. (1964) extracted aqueous solu- tions of LDF, as well as its 2 fractions LDF and LDF l 2 with ethyl ether and found all 3 preparations yielded 5 components. Three of these were found to have lipid contents of so, 57 and 47% and molecular weights of 2.9x105, ll.0xlo5 and 13.0x105, respectively. They could not explain the existence of these 2 stable size distributions in a "lipid core" lipoprotein where complete heterogeneity was expected. Several groups of workers (Cook and Martin, 1962a; Martin gtggl,,l959; Augustyniak gt_al., 1964) have shown that the protein moieties of yolk LDF contained approximately 3% carbohydrate. Abraham gt_al. (1960) reported that the egg yolk LDF contained hexose, glucosamine and sialic acid in the ratio of l.0:0.5:0.3, respectively, and further that removal of these carbohydrate moieties required acid hydrol- ysis. Augustyniak and Martin (1968) found 2 glyCOpeptides resulting from pronase digestion of vitellenin. The authors indicated that the polysaccharide group was probably bound N-glycosidically to vitellenin via the B-amide group of an asparaginyl residue. Smith and Turner (1958), and Martin (1961) examined the N-terminal amino acids of vitellenin. Smith and Turner (1958) found lysine to be the predominant amino acid, but also found minor quantities of glutamic and/or aspartic acid, serine, threonine and alanine. Martin (1961) also 22 found primarily lysine with lesser amounts of alanine, serine, valine, glutamic acid and aspartic acid. The former authors suggested the minor amino acids might be the result Of proteolytic enzyme activity during preparation, but more probably were due to impurities in the preparation. Martin (1961) suggested that since the quantities of alanine, serine and valine increased as the preparation was allowed to stand, they probably came from the breakup of aggregates. Differential Staining Techniques There are several methods available for staining proteins in general, as well as lipoproteins, glyc0proteins and hemOproteins in acrylamide gels. Many of these were originally designed for staining paper or starch gels and have been more recently adapted to acrylamide gels. Protein Staining Amido black 108 has been frequently used as a gen- eral protein stain. It has been shown, however, that the staining prOperties may differ considerably depending on the batch and manufacturer (Pastewka 2E_2l-r 1966). In addition, amido black tends to stain proteins metachromatically, resulting in stained bands of different shades of blue, black or brown, which is a drawback for quantitative densitometry (Johns, 1967). Nigrosin was reported to have a high protein speci- ficity, but it gave rise to difficulties in the destaining 23 step, since the dye tended to create components with low mobilities (Neelin and Conell, 1959). In 1963, St. Groth g£_31, introduced procion bril- liant blue RS and coomassie brilliant blue R250. The procion was reported to form a nearly irreversible protein bond by the reaction of its chlorine substituted triazinyl "rest" with hydroxyl, amino, amide, and peptide groups of proteins. St. Groth gt_gl. (1963) also reported that coomassie blue, in a weakly acid medium, reacted with NH; and nonpolar protein groups, and was held by van der Waal's forces. It was shown to be three times more sensitive than procion, and ten times more sensitive than amido black. However, coomassie blue did not form as stable a bond as did procion, and also tended to deviate notably from Beer's Law at high protein concentra- tions. As a result it Could not be used to quantitate gels. Chrambach gt_§l. (1967) reported that coomassie blue gave a high sensitivity without destaining, but found that sensitivity was vastly increased with background destaining. Lipoprotein Staining In 1959, McDonald and Ribeiro employed a 1% solu- tion of sudan black B in ethylene or prOpylene glycol as a prestain for serum lipoproteins. The incubated the serum and dye solution in a 5:1 ratio for 45 minutes at room tem- perature prior to starch gel electrophoresis. 24 Later Raymond §E_31. (1966) used essentially the same method with a lipid crimson prestain, and a 3% acryla- mide gel, to stain serum lipoproteins. Ressler gt_§l. (1961) also used sudan black B as a lipOprotein prestain. They incubated the lipoprotein solu- tion with the stain for 24 hours at 3°C. Prat gt_31, (1969) dissolved sudan black B in an acetone, acetic acid, water mixture and stained serum lipo- proteins. They reported the lipOproteins appeared as blue- black bands. Glycoprotein Staining Several methods have been reported for staining glyc0proteins (Felgenhauer, 1970; Caldwell and Pigman, 1965). These are based on the periodic acid-Schiff (PAS) reagent of Kojiv and Gronwall (1952) as modified by Keyser (1965) for acrylamide gel electrophoresis, and expanded by Zacharius gt_al. (1969). The method is based on the principle that the polysaccharides are oxidized by periodic acid to give polyaldehydes which yield colored compounds with Schiff's reagent, fuchsin sulfite. Hemoprotein Staining Hemoproteins and other iron-containing proteins have been stained using a benzidine solution (Smithies, 1959; Beaton et al., 1961). The staining was reported based on the principle that hemoglobin and hemoglobin-like proteins 25 catalyzed the oxidation of leuco dyes such as benzidine and guaiacol in the presence of hydrogen peroxide. The proteins were reported visible as blue to black bands. Molecular Weight Determination in SDS Polyacrylamide Gels The first molecular weight determinations in an electrOphoresis gel system were made by Smithies in 1962. He used a system of urea, formic acid, sodium hydroxide and mercaptoethanol in starch gels as a means of determining the "relative retardations" of various proteins. He reported that these retardations were functions of size and not charge, but acknowledged that there was some scatter between actual and expected retardations, and could thus not give a definite relationship between retardation and molecular size. In 1967, Shapiro g£_gl. reported that the molecular sizes of polypeptides could be estimated from the relative electrophoretic mobilities of their sodium dodecyl sulfate (SDS) complexes on polyacrylamide gels. He did not, however, attempt to explain themechanism of this technique. Pitt-Rivers and Impiombato (1968), in studying Shapiro's §t_g1, (1967) observations, found that most pro- teins bound 90-100% of their weight in SDS. They concluded that the maximum binding of the anionic detergent depended on its ability to unfold the protein, and that this binding was restricted by the presence of disulphide groups. While glyc0proteins were found to bind only 70-100% of their weight 26 of SDS, reduction of proteins containing disulphide groups caused a rise in SDS binding to 140% of the weight of the protein. In 1969, Weber and Osborn demonstrated the relia- bility of the SDS molecular weight technique. Using 40 well- characterized proteins, they found that when the electro— phoretic mobilities were plotted against the logarithm of the known polypeptide chain molecular weights, a smooth curve was obtained. Dunker and Rueckert (1969) used an internal calibra— tion technique to construct molecular weight mobility pro- files. They reported that apparent molecular weights gen— erally fell within 5 to 6% Of accepted values, but noted the presence of a few "anomalous" proteins which considerably exceeded this margin Of error. They suggested that different gel strengths be used for different size proteins. For a molecular weight range of 10,000 to 60,000 daltons a 15% gel was recommended. For 10,000 to 100,000 daltons a 10% gel, and for 20,000 to 350,000 daltons a 5% gel was found reliable. In general they found that the effects of intrinsic molecular charge and conformation on electrOphoretic beha- vior in the presence of SDS were small, and suggested that this was due to the fact that the bound SDS was somehow organized by the protein into a micellar complex of definite size governed by the size of the molecule and by its state of foldedness. 27 In 1970(a), Reynolds and Tanford expanded on the theory of Dunker and Rueckert (1969) showing that the binding of SDS by proteins was independent of ionic strength, and was primarily hydrophobic in nature. They suggested that binding induces a conformational change, giving not a globu- lar protein, but rather an extended polypeptide chain con- taining a significant degree of order. They stated that only the monomeric form of SDS binds to the proteins, not the micellar form. Reynolds and Tanford (1970b) further showed that the- SDS binding to the protein transformed it into a uniform rod- like protein-SDS complex whose length varied uniquely with the molecular weight of the protein moiety. The binding ratio on a gram-to-gram basis was found to be identical for all proteins investigated. The protein specificity was shown to be lost and mobility in'a gel was stated to be a measure of molecular size alone. Nelson (1971), however, disagreed showing that the charge Of the protein caused a variation of 20 to 50% in the amount of SDS bound. In addition, at high ionic strength, and in the presence of a high concentration of SDS, up to 2 gms of SDS were bound per gram of protein, not the 1.4 gms previously reported (Pitt-Rivers and Impiombato, 1968). Tung and Knight (1972) suggested that some protein- SDS complexes, which have been found to move at anomalous rates in SDS gels, may do so because their electrophoretic 28 mobility may be determined by a complex interplay of molecu- lar weight, net charge, differential SDS binding and struc- tural compactness of the protein molecules. In addition, the intrinsic charge of the proteins appear not to be impor- tant per se in SDS polyacrylamide gel electrophoresis, but rather becomes a significant factor in accordance with the degree of which they affect binding of SDS and conformation of the protein. N-Terminal Amino-Acid Analysis In 1950, Edman develOped a technique to determine the amino acid sequence of peptides. The method reacted phenyli- sothiocyanate with the peptide in a pyridine water medium, to form a phenylthiocarbamyl (PTC) derivative. The excess reagent was then extracted, first with benzene, and second with a mixture of methanol—ethyl ether, to leave the PTC peptide which was subsequently evaporated to dryness. The PTC amino acid was then cyclized and cleaved off the peptide in anhydrous nitromethane saturated with HCl gas. After filtration to remove the insoluble peptide, the phenylthio- hydantoin (PTH) amino acid was identified using paper chroma- tography. This procedure proved unsatisfactory for larger proteins since they tended to precipitate in the acid media needed for the formation of phenylthiohydantions. Thus the method was modified by Fraenkel-Conrat (1954) and 29 Fraenkel-Conrat gt_§1. (1955) so that the protein was sup- ported on paper strips throughout the series of reactions. Schroeder (1967) further modified the procedure to eliminate the alcohol-ether extraction of the excess phenyl- isothiocyanate, since be determined that many PTC—peptides were soluble in this solvent, and would thus be lost from the strips. He found extraction with benzene alone to be adequate. EXPERIMENTAL PROCEDURES This study was divided into six parts. In part one, electrophoretic patterns of native hen's egg yolk were developed using polyacrylamide disc gel electrOphoresis with 3 concentrations of acrylamide. The Rf values of the bands present in the patterns were then derived using bromthymol blue as a migration reference. A modified procedure for the extraction of phosvitin from hen's egg yolk was developed in part two. The phosvitin components were then further fractionated on a molecular size basis into 3 fractions each containing 1 or more indi- vidual components as shown by disc gel electrophoresis. In part three, the electrophoretic behavior of phosvitin was investigated. The effect of using urea as a denaturing agent, as well as the effect of sample size On the Rf values of the components, was determined. Differential staining was used in part four to demonstrate the presence of lipoproteins, glyc0proteins, and hemOproteins in the electrophoretic pattern of phosvitin. In part five the molecular weights of the components of phosvitin were measured, using the sodium dodecyl sulfate (SDS) technique of Weber and Osborn (1969). 30 31 Part six concluded the study with the determination of the N-terminal amino acids of the components of phosvitin using the modified Edman (1950) technique of Schroeder (1967). ' Part One: Determination of Rf Values of Egg Yolk Proteins Preparation of Sample Fresh eggs were obtained from the Michigan State University Poultry Science Department flock. The egg shell was cracked and the albumen was allowed to drain out. The yolk, with the yolk membrane intact, was rolled on paper toweling to remove any adhering albumen. The yolk was then positioned near the edge of the toweling, the yolk membrane was cut, and the contents were allowed to drain into a beaker. The yolk membrane was discarded. The yolk was mixed with 7 vol of 1/10 strength Tris-glycine buffer (pH 8.3) (6.0 g trishydroxymethylaminomethane, 28.8 g glycine to a volume of l l in distilled water), and stirred for 30 min at speed 4 on a Fisher Thermix model 11-493. Prior to electrophoresis approximately 5% sucrose was added to this sample. Preparation of Acrylamide Gels The electrophoretic method of Davis (1964) was used with minor modifications. Cyanogum 41 (E. C. Apparatus 32 Company) was substituted for acrylamide and N—N'- methylenebisacrylamide, and gels of 5,~6 and 7% cyanogum were run. The following stock solutions were prepared: A. l N HCl 0 O O O I O O O O O O O O O O O O O O O Tris O O O O O O C C O O O O O O O O I O O O O N,N,N',N'-tetramethylethylenediamine (TEMED) . (Eastman Organic Chemicals) Distilled Water to 100 ml B. l N CH1 . approximately . . . . . . . . . . . . Tris . . . . . . . . . . . . . . . . . . . . . TEMED . . . . . . . . . . . . . . . . . . . . . Distilled Water to 100 ml 48 ml 36 g 0.23 ml 48 ml 5.98 ml 0.46 ml The pH was adjusted to pH 6.7 with 1N HCl C. Cyanogum for 5% gels . . . . . . . . . . . . . Cyanogum for 6% gels . Cyanogum for 7% gels . . . . . . . . . . . . . Distilled Water to 100 m1 D. Cyanogum . . . . . . . . . . . . . . . . . . . Distilled Water to 100 ml E. Riboflavin (Nutritional Biochemical Corp.) . . Distilled Water to 100 ml F. Sucrose . . . . . . . . . . . . . . . .2. . . . Distilled Water to 100 m1 G. Ammonium persulfate (Eastman Organic Chemicals) Distilled Water to 100 ml 20.53 g 24.63 g 28.74 g 12.5 g 40 g 0.14 g 33 Stock buffer was prepared as described under “Prep- aration of Sample." Electrophoretic Procedure The small pore solution was prepared by mixing 4 ml A, 8 ml C, 4 ml water and 16 m1 G in a wide mouth 125 m1 Erlen- meyer flask. The solution was degassed and placed in capped glass tubes 13 cm x 0.5 cm held in a vertical position in a supporting rack (Buchler Instruments Division, Nuclear Chicago Corporation). The tubes were filled to a depth of 11 cm with the gel solution, and the gel meniscus was overlaid with a few drOps of distilled water introduced with a 0.5 cc syringe. Care was taken to avoid mixing the gel and the overlaying water layer. The gels were then allowed to polymerize for 20 min between 2 fluorescent lights. After polymerization was complete, the water layer was drawn off with a syringe. Spacer gel, prepared by mixing 1 ml B, 2 m1 D, 1 ml E, and 4 m1 F was then added to a depth of 1 cm. The meniscus was again overlaid with water and the gel was allowed to polymerize for 20 min as above. Following polymerization the bottom caps were removed and the tubes were placed in the electrOphoretic apparatus shown in Figure 1. Five hundred millilitres of stock buffer, diluted with 9 volumes of distilled water, were added to each electrode tank. A 10 ul aliquot of sample was then layered on the surface of the spacer gel. Two drOps of a 0.01% solution of bromthymol blue (Eastman Organic Chemicals) 34 Figure l.--Acrylamide disc gel electrophoresis apparatus. 35 in methanol were added to the upper tank, the platinum electrodes were connected with the anode to the bottom tank, and a current of 2 ma/tube was applied. The power supply was a Heathkit Model lP-l7. When the blue dye front was within 1 cm of the bottom Of the tube, the current was shut off, and the gels were removed from the tubes by rimming with a 20 gauge hypodermic needle while injecting water. The gels were then fixed for 20 min in a 10% solution of acetic acid, and stained over- night in the dark in coomassie brilliant blue R 250 (Sigma Chemical COmpany). The stain was prepared by dissolving 2 g Of stain in 100 m1 of methanol, followed by the addition of 50 g trichloroacetic acid, and made up to a volume of 500 ml with distilled water. Destaining was done overnight in 10% acetic acid with the gels supported in a horizontal position in a fiberglass screen covered plexiglass holder. Rf_Measurement The gels were scanned in a Photovolt Electrophoresis Densitometer model 552. The locations of the various bands were noted on the densitometer tracing, as were the position Of the bromthymol blue front, and the position of the spacer and small pore gel interface. Migration distances of all ~bands were then measured from the gel interface and Rf values were calculated using the bromthymol blue band as Rf = 1.000. 36 Part Two: Extraction of Phosvitin Egg yolk was obtained by the method described in part one. The method of Hurley and Cook (1961) was used, with slight modifications, for the extraction of the yolk granules from the yolk (Figure 2). Yolk was diluted with an equal volume of 0.16M NaCl and centrifuged at 48,200 x G for 20 min. The cloudy supernatant was discarded and the pellet was ground in a mor- tar with twice the original yolk volume of 0.16M NaCl. The mixture was then centrifuged as before. The grinding and centrifugation steps were repeated twice more to yield yolk granules. To extract the phosvitin from the yolk granules, the method Of Wallace g£_31. (1966) was used with slight modi- fication (Figure 3). The granules were dissolved in 5 vol- umes of l M NaCl to yield a clear golden yellow colored solution. To this solution 2 vol of 100% saturated (at 0°C) ammonium sulfate were added slowly. The mixture was then centrifuged at 48,200 x G for 20 min and the floating lipo- vitellin was filtered Off. The filtrate was recentrifuged and refiltered as above to remove all traces of lipovitellin and dialyzed against distilled water until no trace of chloride was detected using silver nitrate. ' After dialysis, the phosvitin solution was con- centrated to about 5% of its original volume on the Amicon ultrafiltration apparatus model 402 (Amicon Corporation) 37 Yolk 0.16M NaCl Centrifuge Superiatant Peliet Discard 0.16MINaCl Centrifuge I Supernatant Pellet Disiard 0.16M NaCl Centrifuge i Supernatant ' Pellet Discard 0.16M'NaCl Centrifuge l l Supernatant Pellet Discard Yolk Granules Figure 2.--Outline for the extraction of yolk granules from egg yolk. 38 Granules l M NaCl 100% Satd (NH4)ZSO4 Centrifuge Filter Insoluble Floating Layer Filtrate Discard Cenirifuge FilTer Insoluble Floating Layer Filtrate Discard l Dialyze Concentrate On UM 2 Membrane Figure 3.--Outline for the extraction of phosvitin from the yolk granules. 39 using a UM2 membrane. The phosvitin solution was then fil- tered through a 0.8u millipore filter to remove any micro- organisms present, and was stored at 4°C until use. Fractionation of PhOsvitin by Ultrafiltration The phosvitin preparation was further fractionated using the Cascading System of Amicon (1970), with the idea of obtaining fractions, each of which contained as few bands as possible, as detectable on disc gel electrOphoresis (Table 1). Immediately following the dialysis step, the phosvitin containing solution was passed through a XM300 membrane using the model 402 ultrafiltration cell with a coupled model RS4 liquid reservoir (Amicon Corporation) Operating under 10 psi of nitrogen. The initial volume of phosvitin solution was approximately 1000 ml. The solution was reduced to 50 ml, at which point 50 ml of distilled water were added to the cell. The volume was again reduced to 50 ml. This process was repeated until a total of 500 ml of water had been added in 50 m1 aliquots. The final 50 ml remaining on the membrane were retained. The entire process was then repeated using the filtrate from above and passing it over progressively smaller pore size membranes as out- lined in Table l. The solution retained on each membrane was further concentrated to a final volume of 5 ml on a UM2 membrane at 45 psi. These solutions were then examined using disc 40 gel electrophoresis as outlined in part one, and the electro- phoretic patterns were compared to those of native phosvitin obtained as described in part three. This procedure gave an approximate idea of the size of the proteins present in the bands of the various fractions. Table 1.--Membranes, operating pressures and starting and final volumes employed in the ultrafiltration fractionation of phosvitin. Volume Cut-off Operating Starting Retained on Molecular Pressure Volume Membrane Membrane Weighta psi ml ml XM300 300,000 10 1000 50 XM100A 100,000 15 1450 50 XMSO 50,000 20 1900 50 PM30 30,000 35 2350 50 UM10 10,000 40 2800 50 UM2 1,000 45 3250 5 aRetention depends on molecular configuration as well as size. Part Three: Determination of R£ Values Of—Phosvitin Componenté Phosvitin was extracted as described in part two. In order to determine the concentration of the aqueous phosvitin, a measured aliquOt was 1yOphilized and the residue weighed to give a concentration in mg/ml. 41 Disc gels were prepared as described in part one using 6% acrylamide. The aqueous phosvitin sample was diluted with 9 volumes of stock electrode buffer to give a final ionic strength equal to that of the running electrode buffer. Five percent sucrose was then dissolved in the sample. Effect of Sample Size on Phosvitin Rf Values Previous experiments indicated that the electro— phoretic pattern of phosvitin could be changed by altering the sample size applied to the gel. Thus, sample sizes of 25, 50, 70, 100, 125, and 150 pl of the above phosvitin sample were used. Two drops of a 0.01% solution of brom- thymol blue were added to the upper buffer tank and electro- phoresis, staining, and destaining were conducted as described in part one. Similarly, the Rf values of all the bands of the six sample Sizes were determined as described in part one. Effect of Urea on Phosvitin Rf Values The effect Of 4 M urea as a denaturing agent on phosvitin was examined using disc gel electrophoresis. Six percent gels were prepared as described in part one, except that all gel solutions, as well as the running electrode buffer, contained 4 M urea. In addition, the sample, diluted with 9 volumes of stock buffer, had 4 M urea added. Sample sizes of aqueous phosvitin of 25, 50, 75, 100, 125 42 and 150 pl were applied to the gels. Two drops of bromthymol blue were added to the upper buffer tank, and electrophoresis, staining, destaining and Rf values were again determined as described in part one. Part Four: Differential Staining of Phosvitin To Obtain information on the composition of the var- ious bands of the phosvitin preparation, lipoprotein, glyco- protein, and hemoprotein stains were used. For all staining procedures phosvitin was prepared, and disc gel electro- phoresis was conducted as described in part one. In all procedures, the sample size was 100 ul per gel. Lipoprotein Stain Two methods were used for the detection of lipopro- teins. In the first method the gels were stained overnight in sudan black B according to the method of Prat gg_gl. (1969). The stain was prepared by dissolving 500 mg of sudan black B (Matheson, Coleman and Bell) in 20 ml of acetone. To this was added a mixture of 15 ml of acetic acid in 80 ml Of distilled water. The stain was stirred for 30 min, and centrifuged at 10,000 x G for 10 min. The gels were then destained as required in a solution Of 15% acetic acid, 20% acetone, and 65% water. Dark brown or black bands were considered to be lipoproteins. In the second method, lipoprotein prestain (Esbe Laboratory Supplies) was used. The prestain was prepared by 43 dissolving 1 g of lipid crimson in 100 ml of ethylene glycol. The mixture was allowed to stand overnight before filtering through Whatman #1 filter paper. Prior to electrophoresis, 1 part of the sample was diluted with 2 parts of dye and allowed to incubate for 1 hr at room temperature. Following electrophoresis, no destaining was necessary. The gels were scanned and then overstained with coomassie blue as des- cribed in part one. Glycoprotein Stain The method of Zacharius §E_gl. (1969) was used. Following electrophoresis, the gels were removed from the tubes and immersed in 12.5% TCA (25-50 ml/gel) for 30 min. The gels were then rinsed with distilled water, and immersed in 1% periodic acid (made in 3% acetic acid) for 50 min. The gels were then washed twice in 200 ml of distilled water per gel for 10 min, followed by a final washing overnight in 500 ml of distilled water per gel. It is important to remove all the periodic acid at this point, or a background stain will result, obscuring the bands. The gels were immersed in fuchsin-sulfite stain (Fisher) in the dark for 3-4 hr. Bright red bands were taken to indicate the presence of a glycoprotein. The gels could then be rinsed in distilled water and scanned immediately, or could be washed in freshly prepared 0.5% sodium metabisulfite (3x10 min with 25-50 ml/gel) followed by washing in distilled water, and stored in 3-7.5% acetic acid. 44 Hemoprotein Stain Hemoproteins were detected using the benzidine test of Smithies (1959). The gels were immersed in a solution containing 0.2 g benzidine, 0.2 ml 30% hydrogen peroxide, 0.5 ml glacial acetic acid, and made up to a final volume Of 100 ml with distilled water. The appearance of white bands was taken as an indication of hemOproteins. Part Five: Molecular Weight Determinations The method used was essentially that Of Weber and Osborn (1969). Marker proteins were chosen to cover a range from approximately 10,000 to 100,000 daltons, and where pos- sible, were chosen for their small deviations from expected mobilities in SDS systems (Dunker and Rueckert, 1969). Table 2 shows the marker proteins selected and their molecu- lar weights. The phosvitin was prepared as described in part two, and following dialysis was 1yOphilized to be used as required. Preparation of Gels Ten percent polyacrylamide gels were used for the molecular weight determinations. For the 10% solution 22.2 g Of acrylamide, and 0.6 g of methylenebisacrylamide were dis- solved in distilled water to yield 100 ml of solution. Before use, 13.5 ml of acrylamide solution were mixed with 15 ml of gel buffer (pH 7.0) (7.8 g NaH PO 2 4 4 - 7H20 and 2 9 SDS made up to l l in distilled . H20, 3806 9 Na HPO 2 45 water), 1.5 ml of persulfate solution (1.5 9 made up to 100 ml in distilled water), and 0.045 ml TEMED. Table 2.--Marker proteins and their molecular weights used in the determination of the molecular weights of the components of phosvitin. Molecular Protein‘ Weight Reference Cytochrome C 12,400 Dunker and Rueckert 1969 Myoglobin 17,000 White gg_al, 1968 Trypsin 24,000 White gg_gl. 1968 Carbonic Anhydrase 29,000 Weber and Osborn 1969 Pepsin 35,500 Dunker and Rueckert 1969 Ovalbumin 46,000 Neurath 1963 Catalase 60,000 Weber and Osborn 1969 Bovine Serum Albumin 66,000 Neurath 1963 Phosphorylase A 94,000 Ullmann gg_gl. 1968 Glass tubes 13 cm long were filled to a depth of 11 cm with acrylamide solution, the meniscus was overlaid with water, and the gels were allowed to polymerize for 20 min between 2 fluorescent lights. The water layer was then removed with a syringe, the tubes were placed in the electrophoresis apparatus, and gel buffer diluted 1:1 with distilled water was added to the upper and lower electrode tanks. 46 Preparation of Samples The marker proteins, from a commercial source, were dissolved in the incubation solution (0.01 M sodium phosphate buffer (pH 7.0), 1% SDS, and 1% B-mercaptoethanol) at a con- centration of 1 mg/ml. The proteins were incubated for 2 hr at 37°C and were then dialyzed for several hours at room temperature against 500 ml of 0.01 M sodium phosphate buffer (pH 7.0) containing 0.1% SDS and 0.1% B-mercaptoethanol. Following dialysis, equal portions of each protein and cytochrome C were mixed in individual vials. Five percent sucrose was added, and the proteins were applied to the electrophoresis tubes so that each tube contained cytochrome C and one other protein. A sample size of 50 ul was used, giving 25 ug of each protein/tube. Phosvitin was dissolved at a concentration of 17 mg/ml, and 3 parts were diluted with 1 part of the cytochrome C solution before use. A sample of 100 ul was applied per tube giving 1275 pg of phosvitin and 25 ug of cytochrome C/tube. The high concen- tration of phosvitin was necessary to observe the fainter bands in the gel. Following the sample application, the electrodes were connected with the anode to the bottom chamber, and a current of 8 ma/tube was applied. When the cytochrome C had migrated to within approximately 1 cm of the bottom of the tube, the current was shut off. The gels were removed frOm 47 the tubes, stained, destained, and scanned as described in part one. The Rf value of each protein was calculated using cytochrome C as Rf = 1.000. The method of Li (1964) was used to calculate the equation of the regression line of Rf values versus molecular weights of the proteins. The Rf value of each of the phosvitin bands was calculated using cytochrome C as a reference, and from the regression equation the molecular weight of each of the bands was obtained. Part Six: N-Terminal Amino Acid Analysis Separation of the Phosvitin Components Figure 4 shows the preparatory disc gel apparatus designed to separate the phosvitin components prior to N-terminal amino acid analysis. The apparatus consisted of 4 basic parts. The lower electrode tank measured 15.2 cm in diameter x 11.4 cm high. The platinum electrode was maintained in the center of the tank on the end of a hori- zontal arm fastened to the tank wall. A polyethylene cooling coil was located in the bottom of the tank. Supported on this lower section was a plexiglass section consisting of 2 horizontal platforms 14 cm square and separated by 2 vertical plexiglass supports 6.3 cm high. The lower platform was drilled with 4 equally spaced 15 mm holes, while centered above these, the upper platform was drilled with four 25 mm holes. 48 Figure 4.--Preparatory acrylamide disc gel electrophoresis apparatus. 49 Four 13 mm inside diameter glass tubes containing the acrylamide gel, and connected to 7 cm diameter powder funnels with polyethylene tubing, fitted into the holes in this section. The funnels served as upper electrode tanks. The fourth section was a flat plexiglass section measuring 12.7 cm square. Inserted through 1.6 mm holes in this section, directly over the center of the funnels, were the upper platinum electrodes. This section rested loosely on top of the funnels. The design of the apparatus permitted the running of l to 4 gels at a time. It also had an advantage in that l or more of the gels could be removed at any time without disturbing the remaining gels. To run, the capped tubes were filled to a depth of 13 cm with approximately 15 ml of a 6% acrylamide gel pre- pared as described in part one. The gel meniscus was over- laid with water and the gels were allowed to polymerize for 45 min between two fluorescent lights. Following pOlymeri- zation, spacer gel was added to a depth of 1 cm, overlaid with water and polymerized for 20 min. The tubes were then placed in the apparatus and Tris-glycine buffer (pH 8.3) as described in part one was added to the electrode tanks. A sample of 200 pl of previously 1yOphilized phosvitin at a concentration of 2 mg/ml in electrode buffer, with 5% added sucrose, was applied per tube. Two microlitres of bromthymol blue were added to each upper electrode tank, 50 and electrophoresis was carried out with a current of 5 ma per tube. When the marker was within 1 cm of the bottom Of the tube, the gels were removed, stained for 6 hr in coomassie blue, and destained electrically in 7-1/2% acetic acid. These gels served as reference gels. The electrophoretic process was repeated until 12 gels had been run. As each run was completed, the gels were compared to the reference gels and the apprOpriate sections of the gels were cut out, and identical sections from the various gels were pooled. The pooled sections were then macerated in 100 ml beakers and covered with 50 ml of distilled water. The slurry was allowed to stand for 2 hr at room temperature, and was then filtered through Whatman #1 filter paper. The gel was removed from the filter paper, replaced in the 100 m1 beaker, and again covered with 50 m1 of distilled water. This slurry was allowed to stand overnight at 4°C, was filtered as above, and the two 50 m1 filtrates were pooled. Another 25 ml of distilled water was used to wash the gel on the filter paper, this this too was added to the filtrate. The combined filtrates from the 3 runs were then dialyzed against distilled water until all traces of the bromthymol blue indicator were removed from the fraction containing the electrophoretic front. The solutions con- taining the eluted components were then pervaporated to approximately 1 m1, placed in 35 m1 serum bottles and lyo- philized. 51 N-Terminal Phenylthiohydantoin Amino Acid Develqpment The phenylisothiocyanate technique of Edman (1950) as modified by Fraenkel-Conrat (1954) and Schroeder (1967) was used to determine the N-terminal amino acid of each of the components of phosvitin, as outlined in Figure 5. Whatman #1 chromatography paper strips 1 x 7 cm were used for the support of the protein. A small hole was punched in one end of the paper strip, and the strip was hung on a glass support throughout most of the series of reactions. The 1yOphilized sample was dissolved in 0.5 m1 of distilled water and 0.2 ml of the solution were applied to duplicate strips with a 0.5 m1 syringe. After drying for 1/2 hr, each strip was thoroughly saturated with 0.2 ml of a solution of 20% phenylisothiocyanate in dioxane, again applied with a 0.5 ml syringe. The strips, still suspended on the glass support, were then placed in a 375 ml jar con- taining 15 ml of a mixture of equal volumes of pyridine, dioxane, and water. The jar was covered with aluminum foil, the lid was tightly closed, and the jars were incubated at 40°C for 3 hr to form the PTC derivative. Following incu- bation, the strips were air dried until they lost translucency, and were placed in 13 x 100 mm test tubes containing suffi- cient benzene to cover them. After extraction for 1-1/2 hr at room temperature, the benzene containing excess , 52 CGHS-NCS + HzN-CHR-CO-NH-CHR'~CO-NH--- (OH') CGHS-NH-CS-NH-CHR-CO-NH-CHR'-CO-NH--- Phenylthiocarbamyl (PTC) derivative (3+, (r H N - '- - --- - HZN CHR CO NH + C 6H5 \C'= I l S\g CHR / 5-Thiazolinone C685 N-—-CS oc NH \\ /’ C /\ R H Phenylthiohydantoin (PTH) amino acid Figure 5.--Out1ine for the development Of PTH amino acids. S3 phenylisothiocyanate, was poured off and a second 1—1/2 hr extraction was performed. Finally, a third extraction with benzene was carried out overnight at room temperature. The strips were replaced on the glass support, aerated for 1 hr, and were then transferred to a dessicator into which had been placed 15 m1 of glacial acetic acid and 15 ml of 6 N HCl in individual beakers. The pressure in the dessicator was reduced to 100 mm of mercury, and the degradation was allowed to proceed for 7 hr at room temperature, forming the PTH amino acid. The strips were then placed over Drierite (Anhydrous calcium sulfate) in a dessicator overnight. Two one-hour long extractions with acetone were used to remove the PTH amino acid from the strips. The acetone was evap- orated under a stream of nitrogen gas to 0.1 ml containing the PTH amino acid residue. The strips were re-extracted with a solution Of 95% acetone in water to remove histidine and arginine. The solution was again evaporated to 0.1 m1 and added to the above corresponding sample. Identification of Phenylthiohydantoin Amino Acids The PTH amino acids were identified using thin layer chromatography. Standard PTH amino acids (Sigma Chemical Company) were dissolved at a concentration of 1 mg/ml in acetone, except histidine and arginine, which were dissolved in methanol. Reference chromagrams were prepared by spot- ting 4 ul of a mixture of PTH-histidine, PTH-arginine, 54 PTH-glycine, and PTH-phenylalanine on a Silica gel chroma- gram sheet (Eastman Organic Chemicals) previously activated @ 100°C for 15 min. The chromagram was then developed ver- tically for 15 cm in a system of chloroform and methanol (9:1) (Brenner, 1961), dried for 5 min, and developed hori- zontally for 15 cm in a system of chloroform—formic acid (100:5) (Brenner, 1961). The chromagram was dried and sprayed with an iodine— azide solution prepared as follows: Solution 1. 1.27 g iodine + 8.3 g potassium iodide + water to 100 ml . Solution 2. 3.2 g sodium azide + water to 100 ml Solution 3. 1% soluble starch solution in water Mix solutions 1, 2 and 3 in the ratio of 1:1:6. After identification of each of the 4 individual PTH amino acid spots on the chromagram, 1 ul of each of the remaining PTH amino acids were spotted on individual chroma- gram sheets along with 4 ul of the above reference mixture. Each chromagram sheet was developed as above, dried and sprayed to yield a reference chromagram for each PTH amino acid. The PTH amino acids from each of the phosvitin com- ponents were then spotted, developed and visualized as above. Identification consisted of comparing the sample chromagram to the reference chromagram, and/or where necessary, spotting known PTH amino acids along with the 55 suspected corresponding unknown PTH amino acid on the same chromagram. RESULTS AND DISCUSSION Part One: Rf Values of Egg Yolk Proteins The Rf values of the whole yolk fractions run on the 3 acrylamide concentrations are shown in Table 3. In addi- tion to the fractions indicated, a very dense band was Observed at the interface between the spacer and running gels, while the spacer gel itself stained very heavily throughout its entirety. There was also an atypical zone which somewhat resembled a large "bubble" in the gel run— ning from Rf of approximately 0.5 to Rf of approximately 0.7. The ends of this zone were rounded off and appeared to be connected with diffuse areas running down the edges of the gel. This entire bubble was superimposed on the typical gel pattern. Excluding the heretofore mentioned zones, a total of 29 bands were Observed in the 5 and 6% gels, while only 25 bands were apparent in the 7% gel. The 7% gel showed the sharpest bands. However, with the smaller pore size of the 7% gel, the Rf values were somewhat smaller than with the larger pore gels and as a result the bands were more crowded together and more difficult to observe. The fact that fewer bands were Observed with the 7% gel is probably the result of heavier bands obscuring the fainter ones. 56 57 Table 3.--Rf values of egg yolk proteins run on 5, 6 and 7% acrylamide gels.a Band Number 5% Gel 6% Gel 7% Gel 1 0.129 0.133 0.121 2 0.163 0.147 0.129 3 0.197 0.172 0.139 4 0.281 0.209 0.200 5 0.303 0.250 0.214 6 0.322 0.266 0.222 7 0.338 0.278 0.252 8 0.379 0.332 0.283 9 0.403 0.356 0.295 10 0.419 0.370 0.309 11 0.443 0.382 0.345 12 0.461 0.409 0.375 13 0.484 0.423 0.397 14 0.511 0.442 0.414 15 0.541 0.467 0.430 16 0.569 0.489 0.450 17 0.597 0.511 0.476 18 0.622 0.535 0.510 19 0.649 0.569 0.524 20 0.668 0.585 0.547 21 0.691' 0.600 0.571 22 0.711 0.659 0.594 23 0.726 0.678 0.630 24 0.790 0.698 0.721 25 0.830 0.724 0.994 26 0.850 0.763 ... 27 0.898 0.835 ... 28 0.964 0.852 ... 29 0.991 0.993 ... Reference 1.000 1.000 1.000 aEach Rf value is the mean of the Rf values of 4 gels from each of 3 separate runs. Figures 6, 7 and 8 show the photographically reduced densitometric tracings of the whole yolk run on 5, 6 and 7% acrylamide gels, respectively. It is apparent from the tracings and from Table 3 that the Rf values were reduced 58 .Hom oofiEmHmuom wm so sou xaox OHOQB mo mcflomuu HoumEouwmcooll.o musmflm H m m h m Ha ma ma ha mHHN mm mm hm mm N e. w m 0H NH va ma ma om NN vm mm mm cflmfiuo oocoummmm 59 .HO@ moflEmawuom we so can xaom OHOEB mo mcfiomuu Hmumeouflmcmo a m m h m Hana ma ha ma Hm mm mm hm .n mesmem N v m m 0H NH «A ma ma om mm «N mm mm camauo mocmummom 60 .HO@ moflEmaxuom mp so any xaox maonz mo mcwomuu umumeoufimcwall.m musmwm camauo v w I ' m ad m 0H NH mama 5H ma Hm ea ma ma om mm mm vm mm oocmummmm 61 as the gel strength was increased. Band 1 in the 5 and 6% gels was very heavy and had a sharp trailing edge, while the leading edge was very diffuse and actually appeared as 2 diffuse areas (bands 2 and 3). These diffuse bands were less intense in the 6% gel than in the 5% gel and in fact were completely absent in the 7% gel. While Powrie et_31. (1963) and Davey et_el. (1969) found lipovitellin non-mobile on paper electrophoresis, Chang eg_el. (1970) suggested that the slowest bands in a 7.5% acrylamide gel were lipoproteins. Similarly, Ornstein (1964) reported that a lipOprotein with a molecular weight of 1.3x106 from human serum moved very slightly into a 7% gel. Since Martin eE_el. (1959) reported lipovitellenin to have a molecular weight of 4.8x106, while Bernardi and Cook (1960b) reported lipovitellin to have a molecular weight of > 10.3x105, it is logical to assume that any lipoprotein which migrates into the running gel will be the smaller lipo- protein, lipovitellin. This is especially true when one considers the fact that lipovitellin has a higher protein content than lipovitellenin and would thus be expected to have a higher charge to mass ratio to prOpel it through the gel. The loss of the diffuse areas with the higher gel strength might then be explained by the fact that lipid was being sheared off of the protein moiety to a greater extent in the higher gel strengths. In the 5% gel some lipid may 62 have been sheared off the protein as it migrated through the gel giving the diffuse areas. The leading part of the dif- fuse area would probably be that protein which had the lipid removed early in its passage through the gel, and thus migrated farther. In the 6% gel the same situation would be true. That portion of the diffuse area which migrated the farthest probably had its lipid moiety removed early in its passage. With the higher gel concentration, however, there was slightly more resistance to the passage of the protein moiety through the gel; thus the leading edge of the diffuse area had an Rf value slightly less than that in the 5% gel. The heavy trailing edge of band 1 then represented the lipo- protein which had no lipid removed. This theory is further substantiated by the fact that as bands 2 and 3 were being reduced in size at higher gel strengths, bands 4, 5 and 6 and 8, 9 and 10 were becoming more intense at higher gel strengths. Thus, at the higher gel strengths, more lipid was being sheared away from the protein moiety at the interface between the spacer and run- ning gels leaving the protein to migrate freely. The fact that bands 4, 5 and 6 all increased in intensity at approxi— mately the same rate, as did bands 8, 9 and 10, indicates that 6 basic proteins were involved in the shearing phenom- enon. With the 7% gel, no diffuse areas were observed; only the sharp trailing edge referred to as hand 2. This 63 . observation indicates that all the lipid which could be sheared off had been removed at this gel strength, leaving band 2 representing a lipoprotein with more tightly bound lipid. The fact that 1 band was apparent in the 7% gel with an Rf less than that of the band in question is inconse— quential to this discussion. This band was probably obscured by the heavier fractions at the lower gel strengths, and only became apparent because it was retarded to a greater degree at the higher gel strength. It is interesting to note that the Rf value of band 1 in the 5% gel and the corresponding band in the 6 and 7% gels did not change at the higher gel strength. This anomalous behavior might be explained if one considers that this was a large molecule, probably as large or slightly larger than the gel pore size of, for example, the 5% gel, and that some "tunnelling" occurred through the gel as described by Ornstein (1964). The fact that this was a relatively large molecule would tend to overwhelm the slight increase in pore size at the lower gel strengths, and the molecule would still, in effect, be "crowded" in the gel. The remaining peaks all seem to be retarded in a uniform manner at the higher gel strengths. In the 5% gel, peaks 14, 15, 16 and 17 correspond to 14, 15, 16 and 17 in the 6% gel and to l2, 13, 14 and 15 in the 7% gel, respec- tively. Similarly, peaks 23, 21 and 17 correspond in the 5, 6 and 7% gels, respectively. The slight changes in 64 intensity of the peaks between 13 and 22, 12 and 21 and 12 and 17 are probably due to the overlapping of some of the bands as they become more compressed at the higher gel strengths, with some of the bands in this area disappearing entirely. Chang eg_el. (1970) suggested that the intermediate bands in the densitometric tracing were livetins. This would correspond to bands 11 to 23 in the 5% gel in this work; however, no verification was provided here. Similarly, they suggested that the fastest moving bands were phosvitin, which would appear to correspond to bands 24 to 29 in the 5% gel. It will subsequently be shown that there are sev- eral fast moving bands in phosvitin, as well as some inter- mediate and some slow moving bands. The heavy band at the gel interface, and the material dispersed throughout the spacergel was probably lipovitel- lenin with a molecular weight too great to allow it to enter the small pore gel. It is also worth noting the presence of one band which migrated with the reference marker in all 3 gel con- centrations. This was probably a very small protein well below the critical size of the gel. Some of the diffusivity observed in the various bands, as well as the presence of minor bands, might be explained by an extrapolation of the observations of Lush (1961). He reported that minor differences were apparent in the starch 65 gel tracings of hen egg albumin from different phenotypes of hens. The differences appeared to be manifested more as a broadening of the bands rather than the appearance or disappearance of a band, although both were apparent to some degree. If one can apply these observations to egg yolk, then it is possible that the number of bands observed in the 3 different percentage gels might be reduced if the sample had been derived from eggs from only one hen. While this might tend to reduce the number of bands somewhat, or reduce the diffusivity of some bands, more bands might be demonstrated if longer gels were used to further separate the bands. It is very probable that more minor bands were present in the gels, which were obscured by the heavier more diffuse bands. , Part Two: Extraqpion and Fractionation of Phosvitin In the original method of Burley and Cook (1961), only the original dilution and one washing in 0.16M NaCl were used to prepare the granules. When this procedure was followed, the wash solution had a pronounced golden color presumably indicative of the soluble livetins. Thus 2 more washing steps were added with the last one showing no visible trace of livetins. . When the method of Wallace e5_gl. (1966) was employed to extract the phosvitin from the granules, a slight turbidity was observed in the filtrate following centrifugation and 66 filtration to remove the lipovitellin. Thus an additional centrifugation-filtration step was added to remove all the insoluble material. Wallace EE_El° (1966) labelled this fraction ”crude phosvitin." They reported that the prepara- tion could be further purified by chromotography on DEAE cellulose. This purification step was omitted because mate— rial which was part of the granule, but was not lipovitellin, might be removed by this step and thus go undetected. In the original method of Wallace e£_gl, (1966), the phosvitinmwas either precipitated at pH 1.5 with 1N HCl and air dried, or was 1yOphilized. When a 1yOphilized sample was compared to an aqueous sample, a slight difference was noticed on disc gels. This difference was manifested as an increase in diffusivity of the bands of the 1yOphilized sample. Since both 1yOphilization and air drying of the phosvitin result in the nearly complete removal of water from the sample, they are presumably treatments harsh enough to result in some denaturation of the protein. Thus if the sample was not to be used immediately, or if it required a concentration step before examination on disc gels, it was concentrated on a UM2 membrane and filtered through a milli- pore filter to remove any microorganisms present. A clear, pale golden colored solution was obtained. When this phosvitin sample was run on acrylamide disc gels, a pattern was obtained as shown in Figure 9. Eleven zones were apparent on the tracing as well as a sharp band 67 .Hmm mowemamuom we so can cfiufl>monm mo mcflomuu Houmeoufimcmoln.m whomwm .HN m cwmwuo 7.33 m \1 0H AH mocmuowom 68 at the gel interface, and a diffuse area in the lower por— tion of the spacer gel. These latter 2 zones were presumably too large to enter the running gel. Of the 11 major zones apparent, 1 of these constituted the upper limit of a "bubble" similar to that observed in the disc gels of whole yolk. This bubble had well-defined sides, and a very pale interior when stained with coomassie blue. There was no well-defined lower perimeter to this bubble. Rather, the sides ended abruptly as did the pale interior. There was a narrow clear zone and the bubble was then followed immediately by 3 sharp, well-defined bands. When the phosvitin sample was passed through the XM300 membrane, and the resulting retentate portion examined on acrylamide disc gels, the 11 major zones as well as those in the spacer gel were apparent. The XM100A membrane reten- tate showed a similar pattern, indicating that all the zones had a molecular weight of less than 3.0x105. This molecular weight determination assumed a protein with a globular con- formation. It would be possible for a protein with a longi- tudinal conformation to pass through the membrane even though its molecular weight was in excess of 3.0x105. The fact that both the XM300 and XM100A membranes retained fractions did not necessarily indicate that the fractions were too large to pass through; however, it implied that a greater number of washings were necessary to move all of the protein through the membrane. In theory 69 (Amicon, 1970), if enough washings had been employed, all the sample would have passed through the XM300 membrane, since it has been shown here that all the fractions had a molecular weight of less than 3.0x105. Thus, to be certain of which fractions were retained on a membrane, it was necessary to examine the retentate on the next smaller membrane to deter- mine if any fraction was missing. The retentate on the XMSO membrane contained only the 11 major fractions in the running gel, indicating that the 2 diffuse areas of the spacer gel had been retained by the XM100A membrane. Again, assuming a globular shape, these diffuse fractions had a molecular weight in excess of 1.0x105. Of the 11 fractions appearing in the XMSO retentate, all had the same relative intensities as were apparent in the XM100A retentate, indicating that no fraction was retained by the XM100A to even a slight degree. The PM30 retentate contained only the bubble frac— tion and all those fractions below it, indicating that the XMSO membrane had retained those fractions above the bubble. Similarly, the UM10 membrane contained the bubble and all those fractions below it. This indicated that these frac— tions had a molecular weight of less than 3.0x104, the cut- off weight of the PM30 membrane. The UM2 membrane retentate, however, contained only the fraction immediately behind the reference front, indicat- ing that this was the only fraction to pass through the UM10 70 membrane, and thus had a molecular weight of less than l.0x104. No membrane was available with a cut-off weight of less than the UM2. However, when a sample of the filtrate~ from the UM2 was 1yOphilized, no residue remained. This indicated that this smallest, fastest moving fraction was approximately l.0x103 to l.0x104 molecular weight. Part Three: Determipetion of Rf Values of Phosvitin Components The concentration of the aqueous phosvitin sample was found to be 13.5 mg/ml.. Thus with sample volumes of 25, 50, 75, 100, 125 and 150 ul the actual sample sizes were 338, 675, 1013, 1350, 1688 and 2025 ug, respectively. Figure 10 shows the pronounced effect that the sample size had on the electrOphoretic mobility of the phosvitin components, while the Rf values of these bands are given in Table 4. The change in mobility followed a definite pattern as the sample size was increased, and appeared to originate with the bubble which was apparent near the center of the gel. The upper perimeter of the bubble was band 5, while the first band immediately below the bubble was band 6, as shown in Table 4. As the sample size was increased, the Rf value of band 5 decreased slightly, while that of band 6 increased. The Rf value of band 4 remained more or less constant with increasing sample size, while bands 3, 2 and 1 all showed lowered Rf values. Bands 7, 8, 9 and 10 all showed 71 Figure 10.--Acrylamide disc gels showing the effect of 6 sample sizes (left to right 338, 675, . 1013, 1350, 1688 and 2025 pg) on the electrophoretic mobilities of the 11 phosvitin components. 72 Table 4.--Rf values of phosvitin run on 6% acrylamide gels at 6 different sample sizes.a Sample Size (ul) Band # 25 50 75 100 125 150 1 0.116 0.103 0.100 0.086 0.085 0.072 2 0.137 0.122 0.116 0.099 0.094 0.082 3 0.157 0.153 0.154 0.147 0.145 0.125 4 0.204 0.190 0.191 0.185 0.215 0.211 5 0.511 0.512 0.509 0.498 0.490 0.483 6 0.674 0.710 0.745 0.759 0.769 0.786 7 0.709 0.739 0.766 0.770 0.783 0.798 8 0.745 0.768 0.791 0.791 0.806 0.817 9 0.846 0.852: 0.851 0.864 0.876 0.872 10 0.903 0.904 0.905 0.914 0.919 0.923 11 0.989 0.991 0.991 0.990 0.990 0.991 Reference 1.000 1.000 1.000 1.000 1.000 1.000 from each of 3 separate runs. aEach Rf value is the mean of the Rf values of l gel 73 increased Rf values with increased sample size. In general, the bands above the bubble either remained constant or decreased, while all the bands below the bubble increased in Rf value with increasing sample size. Band 11 migrated immediately behind the reference front in all cases. As described previously, it was noted that when the gels were stained with coomassie blue, the perimeter areas of the gel, that is the trailing edge and the sides, stained heavily, while the center area was much less intensely stained and was completely diffuse. It is worth noting here than when the gels were stained for glycoprotein, the inte- rior diffuse area stained very heavily as did several of the bands immediately below the bubble. The perimeter of the bubble did not show any reaction to the glycoprotein stain. This subject will be discussed in detail in part four. It thus appears that 2 phenomena are responsible for the mobilities and patterns of this system. The bubble was probably a glycoprotein which was dissociated either at the gel interface, or as it moved through the gel, and affected those bands ahead of and behind it, changing their electro— phoretic mobilities. Assume that the steady-state stacking occurred in the spacer gel as described by Ornstein (1964). The bands were thus arranged in very thin starting zones in order of decreasing mobility, and moved downward through the spacer gel, encountering very little resistance, under the 74 influence of an electrical potential. The fastest bands reached the running gel and moved into it, encountering slight resistance. When the bubble fraction, at this time still a narrow starting zone, reached the running gel, very high ' resistance was encountered, either because of the shape or size of the molecule, or possibly because of some adsorption to the running gel medium. As a result, this band was abruptly slowed or stopped as were those bands behind it. Since the same electrical potential was still in effect, the bands already in the running gel now migrated faster. The bubble fraction, because of its higher resistance, started to dissociate itself of some of its carbohydrate: in effect the carbohydrate was sheared off. There was thus a mixture of protein moieties and glycoproteins mixed at the gel inter— face, with the faster, less encumbered, proteins attempting to overtake and run ahead of the slower, larger glycoproteins. As a result, the protein moieties tended to flow to the out— side Of this system, compressing the glyc0proteins into a more or less spherical shape in the centre of the gel. By this time this dissociating system had entered the running gel completely, as had the slower trailing bands. However, since the entry of the trailing bands into the gel had been delayed by the reordering of the protein-glycoprotein system, they spent less time in the running gel and thus their mobil- ities were reduced. The entire phenomenon was then exag- gerated when larger samples were applied to the gel. The 75 3 bands immediately preceding the bubble probably represented protein moieties, or possibly small glyc0proteins, as will be discussed later, which had been previously released from the main glyc0protein system, i.e. the bubble, and migrated unaffected by the dissociation and reshuffling process. A similar situation was observed when phosvitin was run in the presence of 4M urea, as shown in Figure 11. The Rf values of the various bands are given in Table 5. Band 3 represents the band immediately beneath the bottom of the bubble. Again, all the bands above the bubble showed a reduced Rf value as the sample size was increased, while everything below showed an increased Rf value. The only exception was band 9, which migrated directly behind the reference indicator at all sample sizes. A total of 9 bands were observed in the urea system. When compared to the non-urea gels, the only differences involved bands 1 to 4, presumably due to some form of denatur— ation of these proteins. It appeared that bands 1, 2, 3 and 4 in the non-urea system had been reduced to bands 1 and 2 in the urea system. More specifically, it appeared that bands 1, 2 and 3 of the non-urea system had been reduced to band 1 in the urea system, with band 4 in the non-urea gels becoming band 2 in the urea gels. Whichever is the case, several explanations seem possible. Bernardi and Cook (1960c) observed that 4M urea removed lipid material from lipovitel- lin. Thus it would seem logical that bands 1, 2 and 3 in 76 Figure 11.--Acrylamide disc gels run with 4M urea showing the effect of 6 sample sizes (left to right 338, 675, 1013, 1350, 1688 and 2025 pg) on the electrophoretic mobilities of the 9 phosvitin components. 77 the non-urea system were a protein moiety with varying degrees of lipid attached. In the urea system this lipid was removed leaving only the protein moiety as band 1. However, as will be discussed in part four, none of the bands in this area stained for lipoprotein. If complete specificity of the lipOprotein stain is assumed, then this explanation becomes invalid.’ It wasalso unlikely that such a large amount of lipovitellin would remain in this phosvitin preparation. Table 5.--Rf values of phosvitin run on 6% acrylamide gels in the presence of 4M urea at 6 different sample sizes.a Sample Size (ul) Band # 25 50 75 100 125 150 1 0.065 0.050 0.047 0.047 0.041 0.035 2 0.127 0.115 0.110 0.107 0.106 0.103 3 0.520 0.501 0.486 0.489 0.474 0.468 4 0.668 0.699 0.717 0.746 0.768 0.779 5 0.695 0.721 0.737 0.763 0.783 0.792 6 0.731 0.751 0.761 0.781 0.798 0.808 7 0.774 0.787 0.785 0.803 0.818 0.824 8 0.875 0.878 0.875 0.891 0.891 0.882 9 0.991 0.991 0.991 0.991 0.990 0.991 Reference 1.000 1.000 1.000 1.000 1.000 1.000 aEach R value is the mean of the Rf values of l gel from each of 3 separate runs. 78 Another possibility to be considered is that the denatured proteins were somehow unable to enter the running gel due to some conformational change, and were thus not detected. The most logical explanation would seem to be that bands 1, 2 and 3 in the non-urea system represented various states of "unfoldedness" of the protein molecules. In the urea system these 3 states of unfoldedness were reduced to one state representing the most denatured state and thus appeared as only one band. When compared to the non-urea gels the Rf values of the urea gel bands were noticeably smaller, the only excep- tion being band 9, which migrated behind the reference indicator in both cases. This was probably due to the increased molarity of the gel system due to the presence of the urea, since Weber (1968) has shown that increasing the gel molarity decreased the electrOphoretic migration rate of the proteins. In both the nonéurea and urea systems, 2 diffuse bands were apparent behind the fastest band. These were quite faint at low sample concentrations, and were very diffuse and difficult to locate accurately in all gels. The fact that Sugano (1957), Connelly and Taborsky ' (1961) and Mok ep_el. (1966) found phosvitin to consist of 2 fractions, while Sundararajan egpgl, (1960) and Simlot and Clegg (1967) reported phosvitin to consist of 3 fractions 79 may be explained when one considers the techniques involved. It is conceivable that, with poorer electrOphoretic tech- niques such as before the introduction of acrylamide disc gels and with stains less sensitive than coomassie blue, the phosvitin pattern observed in the non-urea disc gels could have appeared as 2 or 3 diffuse areas, namely one area consisting of bands 1 to 4 and the second area consisting of bands 5 to 11. Further, this second area could be broken down into 2 areas consisting of bands 5 to 8, and bands 9 to 11. Part Four: Differential Staining of Phovatin The phosvitin gels stained with the periodic acid- Schiff reagent are shown in Figure 12. Sample sizes of 75 and 125 pl were used giving 1013 and 1688 ug of phos- vitin, respectively. The interior of the bubble stained very heavily, while the perimeter remained unstained. The lower or leading edge of the bubble ended very sharply, and was followed immediately by a very narrow unstained zone, which was in turn followed by the stained band 6. In gels stained with coomassie blue, no band could be detected at the sharp leading edge of the bubble; rather, the entire centre of the bubble was a very lightly stained diffuse area which ended abruptly in a manner identical to that des— cribed here. Band 7 was not stained for carbohydrate, while bands 8, 9, 10 and 11 were all stained very faintly. Bands 80 Figure 12.--Acrylamide gels of phosvitin samples of 1013 and 1688 ug (left to right) stained with periodic—acid Schiff reagent. 81 8 and 11 appeared sharp as they had with the coomassie blue stain, while_bands 9 and 10 were very diffuse. Band 4 also appeared diffuse and stained very faintly. In addition, the band at the gel interface as well as the diffuse area Of the spacer gel stained very heavily. The carbohydrate in the phosvitin fraction thus appeared to be concentrated in the bubble fraction as well as in the very slow-moving components in the spacer gel, while minor amounts were observed in 6 other fractions. Carbohydrate has been shown to be widely distributed in egg albumin as glycoprotein complexes. Avidin A (Fraenkel- Conrat EE_El-r 1952), ovumucin (Donovan e£_el,, 1970), ovamacroglobulin (Donovan ep_§1., 1969), ovoinhibitor (Davis e£_31., 1969), and ovomucoid (Bragg and Hough, 1961) have all been reported as glyc0proteins. In view of these reports, it appears reasonable that glyc0proteins are promi- nent in the phosvitin preparation. Hartley and Jevons (1962) suggested that the carbohydrate is bound glycosidically to serine and/or threonine groups. Since phosvitin has been shown to be very high in serine residues (Mok e£_gl., 1961), there would obviously be ample opportunity for the attach— ment of carbohydrate to phosvitin. The 2 lipOprotein stains gave somewhat different results from the phosvitin gels. Sudan black B stained bands 5, 6, 7 and 8 with equal intensity. Only the interior portion of the bubble stained for lipoprotein and gave a 82 pattern similar to that observed with the PAS stain. The fraction at the gel interface also stained very faintly, while the spacer gel otherwise showed no lipoprotein. The lipid crimson prestain stained only the gel interface. It was believed that the prestain might have bound to other lipoproteins, thereby making them too large to migrate into the running gel, and that they remained at the gel interface. When the gel was overstained with coomassie blue, however, all the bands were present in the usual pattern. Since only the fraction at the gel interface stained positively with bOth lipOprotein stains, it is assumed that this band was a lipoprotein. Bands 5, 6, 7 and 8 may or may not have been lipoproteins. The benzidine stain for hemOproteins, and other iron- containing proteins, reacted positively with all bands below and including the bubble. Instead of blue to black bands, as suggested by Smithies (1959), very pale blue to nearly white bands were observed. Taborsky‘s (1963) description of the rearrangement of phosvitin in the presence of and sub— sequent binding of iron would seem to leave a molecule basically resembling heme. The iron, however, would be attached to 4 phosphate oxygens instead of nitrogens. It thus appears that iron was present in 6 of the 11 fractions demonstrated in phosvitin. This appears to agree with Connelly and Taborsky (1961) who, while reporting 83 phosvitin to contain 2 fractions, stated that 1 fraction was responsible for most of the metal binding properties of the unfractionated preparation. If, as discussed in part four, bands 5 to 11 appeared as 1 fraction in Connelly and Taborsky's (1961) work, and these bands all stained positive for iron in this work, then this stain was indeed specific for bound iron. Part Five: Molecular Weight Determinations The 9 marked proteins were run on the SDS disc gel system and their Rf values measured to yield the standard curve shown in Figure 13. The plotted values all fell closely along a straight line indicating that, as reported by Dunker and Rueckert (1969) and Weber and Osborn (1969), the mobility in SDS gels varied with the log of their molec- ular weight. The Rf values, molecular weights and percentage deviation from expected molecular weights are shown in Table 6. Phosphorylase A showed the greatest deviation at 12.5%. Dunker and Rueckert (1969) reported that apparent molecular weights generally fell within 5 to 6% of expected values, but reported a few anomalous proteins that showed greater deviations. The mean of the deviation for the 9 proteins considered here was 5.3%, well within the anticipated devia- tion. When the phosvitin fraction was run on the SDS system, 18 individual bands were observed (Figure 14). The 84 .mcaououm noxume may mo moam> mm m5mum> pnmaos smasomaofi mo m>uoo huaafiooe onmocmumll.ma muomfim em ... 0.. 0.0 0.0 ho 0.0 0.0 v.0 n0 «.0 ..0 a q u q u d q q u '00- mbgom :mnoouou 001 £2 85 Table 6.--Rf values, expected molecular weights, apparent molecular weights and percentage deviation from expected molecular weights of the marker proteins. Expected Apparent Molecular Molecular Rf Percentage Protein Weight Weight Valuea Deviation Cytochrome C 12,400 12,000 1.0 V 3.2 Myoglobin 17,000 15,000 .904 11.7 Trypsin 24,000 26,000 .662 8.3 Carbonic Anhydrase 29,000 28,000 .636 3.4 Pepsin 35,500 36,000 .524 1.4 Ovalbumin 46,000 46,000 .413 0.0 Catalase 60,000 64,500 .269 7.5 B.S.A. 66,000 66,000 .266 0.0 PhOSphorylase A 94,000 82,000 .165 12.5 aEach Rf value is the mean of the Rf values of 1 gel from each of 3 separate runs. 86 .mamm oofiecahuom mom :0 cos cwua>mocm mo mcwomuu umuoeouwmcmoun.va ouomwm H_m m h a HH cwmwuo ma ma NH «H 0H i .jA. U mEoucoouhu ma 87 Rf values of these bands were measured and plotted on the standard curve (Figure 15). From this curve the molecular weights were calculated and are reported in Table 7. The molecular weights fell into 2 general areas, between 9,000 and 22,000, and between 71,000 and 119,000. These results verified and elaborated the results of part two concerning the ultrafiltration fractionation of phosvitin. The UM50 membrane effectively separated those proteins with molecular weights above and below 50,000. Those 2 size classes appeared to correspond in general to the 2 size classes observed here. Counting the bubble, 7 bands, numbers 5 to 11, passed through the 50,000 molecular weight cut-off membrane. In the SDS gels considered here, 7 bands were apparent with molecular weights below 22,000. .In addition, band 11 passed through the UM10 membrane with a molecular weight cut-off of 10,000. In the SDS gels, 1 band was observed with a molecular weight of 9,050. It appears certain, then, that band 11 in the native phosvitin fraction had a molecular weight of 9,050. It is not, however, possible to directly identify the remaining 6 bands of the SDS system, with molecular weights between 11,000 and 22,000, with the bands 5 to 10 in the native phosvitin gels. In the native gels, the proteins migrated based on their size and charge, while in the SDS system mobility was based only on size. It was thus possible 88 .0>Hoo Suwawnoa osmosmum on» so .mamm moHEMHmuom wow so can mucmsomaoo cflpw>monm on» «o mmoam> mm mo uonII.mH Tasman .m N.. ... 0.. 0.0 0.0 Nd 00 0.0 #0 n0 ~.0 ..0 q q 4 q q d u - d d d 4 l T I 9 lubgem 10|n00|0w 601 0:0. 89 Table 7.--Rf values and molecular weights of the proteins of the phosvitin fraction of egg yolk. Band Number Rf Valuea Molecular Weightb 1 0.004 118,700 2 0.014 116,000 3 0.019 114,900 4 0.026 112,800 5 0.048 107,400 6 0.061 104,200 7 0.079 100,100 8 0.118 91,620 9 0.146 ’ 86,000 10 0.208 74,730 11 0.226 71,750 12 0.752 21,800 13 0.763 21,260 14 0.782 20,360 15 0.795 19,770 16 0.909 15,280 17 1.04 11,360 18 , 1.14 9,050 aEach R value is the mean of the Rf values of 3 gels from each of 3 ifferent runs. bCalculated from the regression equation of Rf value versus molecular weight of the marker proteins.. 90 that some reordering of the bands of the native phosvitin occurred when they were run on the SDS system. Weber and Osborn (1969) have shown that the SDS system had the ability to dissociate proteins of several polypeptide chains into single monomeric units. Since an identical number of bands with molecular weights of less than 50,000 occurred in both systems, the bands present in the native phosvitin tracing may represent proteins of only 1 polypeptide chain. A total of 11 bands were observed with molecular weights greater than 50,000, which was 7 more bands than were observed:hlthe native phosvitin gels. However, no band was observed at the gel interface, and no diffuse area remained in the spacer gel of the SDS system. Since the SDS system is known to dissociate proteins of several poly- peptide chains into subunits as well as to remove lipid material from proteins (Helenius and Simons, 1971), both factors would result in smaller polypeptide chains which could then conceivably move into the running gel and migrate on the basis of their molecular weights. The extra bands observed in the range above 71,000 molecular weight are probably the result of such factors acting on the proteins present at the gel interface and in the spacer gel of the native phosvitin gels. 91 Part Six: N-Terminal Amino Acid Analysis Figure 16 is a diagrammatic‘representation of the pattern achieved when phosvitin was run on the preparatory disc gels. Sections which were removed from the gels yielded the samples for the N-terminal amino acid analysis, and the band numbers of the components of the native phosvitin preparation. Section 1 represented the somewhat diffuse area of the spacer gel, while section 2 was the area at the interface of the spacer and running gels. Section 7 was the bubble observed in earlier gels. Here, however, the sample size was smaller based on the cross sectional area of the gels (3.03 ug/square mm) than in the earlier gels of native phosvitin where, for example, the sample size of 100 pl was 68.8 ug/square millimeter. This smaller sample size resulted in the bubble being reduced to the diffuse area represented in Figure 16, presumably indicating little, if any, breakup of the glycoprotein complex as discussed earlier. This was verified by the fact that band 5 stained uniformly throughout with coomassie blue. Section 8 consisted of bands 6, 7 and 8, which migrated too closely together to be individually sectioned out. Similarly, band 10 was quite diffuse and overlapped into band 11, and thus these 2 bands constituted section 10. Table 8 shows the N-terminal amino acids obtained from the various sections. No amino acid was detected in sections 1 or 2, even though disc gel electrophoresis of the 92 Band Section Numbe[_k Number H | 2 ' 3 2 4 3 5 ‘* 6 Figure 16.--Diagrammatic representation of the preparative acrylamide disc gel of phosvitin showing the sectioning used to obtain pure proteins. 93 eluant of the macerated section showed that the proteins were present. This was probably due to the fact that these 2 sections represented several proteins of high molecular weights, and as a result, the molar quantity of each N-terminal amino acid was too small to be detected. Table 8.--N-termina1 amino acids of the proteins of phosvitin. Section Number N-Terminal Amino Acid 1 None detected 2 None detected 3 Alanine 4 Glutamine 5 Asparagine, Arginine 6 Threonine 7 Asparagine, Glutamine 8 Alanine, Asparagine, Glutamine 9 Glutamine 10 , Alanine Mok eE_el. (1961), Joubert and Cook (1958b) and Mecham and Olcott (1949) reported that alanine was the N-terminal amino acid of phosvitin. In this work, alanine was detected in sections 3, 8 and 10, while arginine, asparagine, glutamine and threonine were detected throughout sections 4, 5, 6, 7, 8 and 9. Since alanine was detected 94 here in 3 sections, including section 10 which contained band 11, a comparatively small molecular weight protein present in a relatively large amount (as evidenced by the intensity of band 11), it thus was present in a comparatively high molar concentration. This may explain the ease of detection reported by previous workers who considered phos- vitin to be homogeneous. In their phosvitin preparation, the other N-terminal amino acids reported here were prob- ably present in quantities too small to detect. Section 5 yielded both asparagine and a minor amount of arginine, indicating that this diffuse area either rep- resented 2 proteins or a protein with 2 subunits. Sections 7 and 8 were not completely separated by the sectioning pro- cedure, since section 7 contained some protein from band 6, and section 8 contained part of band 5. Since alanine was present only in section 8, band 7 or 8 must have had this amino acid as an N-terminal, while bands 5, 6 and 7 had either glutamine or asparagine. SUMMARY Disc gel electrOphoresis was employed to determine the number of proteins present in whole egg yolk. Concen- trations of 5, 6 and 7% acrylamide in Tris—glycine buffer at pH 8.3 revealed 29, 29 and 25 bands, respectively, in the running gel, with an additional diffuse area throughout the spacer gel. The presence of a large, poorly defined bubble was also noted in all concentrations of acrylamide gels. The Rf values ranged more or less uniformly from 0.121 to 0.993 depending on the gel strength. A modified procedure was employed to extract and concentrate phosvitin from the whole yolk. Electrophoretic examination revealed 11 bands, in addition to a diffuse area in the spacer gel, with Rf values ranging from 0.072 to 0.991. The bubble was very apparent in the phosvitin preparation, and the electrophoretic mobility of the com- ponents of the phosvitin fraction appeared to be related to the mobility of this bubble, which in turn was found to be related to the sample size applied to the gels. Electro- phoresis, in the presence of 4M urea, reduced the number of bands to 9, plus a diffuse area in the spacer gel. The electrophoretic mobility was still found to be related to the sample size. 95 96 Ultrafiltration fractionation, using the Amicon ultrafiltration apparatus, yielded 3 sub-fractions of phos- vitin. A definite cut-off in molecular weight was observed with 4 components in 1 fraction greater than 5.0x104, 7 components in a second fraction less than 5.0x104, includ- ing 1 component in the third fraction less than 1.0x104. Differential staining of the phosvitin disc gels revealed 1 lipoprotein band which did not enter the running gel, with the possibility of 4 additional bands being lipo- protein in nature. Staining with the PAS reagent demon- strated carbohydrate to varying degrees in 7 bands, as well as in the spacer gel. The interior diffuse area of the bub- ble stained very heavily. Coomassie blue stained the perimeter of the gel very heavily, while the interior area was only lightly stained. All bands below and including the bubble stained positively for iron. The phosvitin fraction consisted of 18 bands when observed on SDS acrylamide gels. Seven of these bands had molecular weights ranging from approximately 9,000 to 22,000 and the remaining 11 ranged from 71,000 to 119,000. No dif- fuse area remained in the spacer gel in this electrophoresis ‘system. N-terminal amino acid analyses of eluted sections of preparative acrylamide gels of phosvitin, using the Edman technique, revealed alanine, arginine, asparagine, glutamine, and threonine. BIBLIOGRAPHY 97 BIBLIOGRAPHY Abraham, 5., L.A. Hillyard and I.L. Chaikoff. 1960. Compon- ents of serum and egg yolk lipOproteins: galactose, mannose, glucosamine and sialic acid. Arch. Biochem. Bi0phys. 89:74. Ahmed, K., J.D. Judah and H. Wellgren. 1963. Phosphopro- teins and ion transport of cerebral cortex slices. Biochim. Bi0phys. Acta. 69:248. Alderton, G. and H.L. Fevold. 1945. Preparation of the egg yolk lipoprotein, lipovitellin. Arch. Biochem. 8:415. Alderton, S.E. and G.E. Perlman. 1965. Chemical character- ization of the phosphoprotein phosvitin. J. Biol. Chem. 240:3892. Amicon. 1970. Catalog system selection and applications guide. Pub. #403. Amicon Corp. Lexington, Mass. Augustyniak, J., W.G. Martin and W.H. Cook. 1964. Char- acterization of lipovitellenin components and their relation to low-density lipOprotein structure. Biochim. Bi0phys. Acta. 84:721. Augustyniak, J.Z. and W.G. Martin. 1968. Characterization of glyCOpeptides derived from the low-density lipo- protein of hen's egg yolk. Can. J. Biochem. 46:983. Beaton, G.H., A.E. Selby and A.M. Wright. 1961. Starch gel electrophoresis of rat serum proteins. I. Procedure and designation of components. J. Biol. Chem. 236:2001. Bernardi, G. and W.H. Cook. 1960a. An electrophoretic and ultracentrifugal study on the proteins of the high density fraction of egg yolk. Biochim. Biophys. Acta. 44:86. Bernardi, G. and W.H. Cook. 1960b. .Separation and char- acterization of the two high density lipoproteins of egg yolk, a- and B-lipovitellin. Biochim. Biophys. Acta. 44:96. 98 99 Bernardi, G. and W.H. Cook. 1960c. Molecular weight and behavior of lipovitellin in urea solutions. Biochim. Biophys. Acta. 44:105. Boyer, P.D. 1963. Phosphohistidine. Science. 141:1147. Bragg, P.D. and L. Hough. 1961. An investigation of the egg-white mucoproteins, ovomucoid and ovalbumin. Biochem. J. 78:11.- Brenner, M. and A. Niederwieser. 1961. As reported in Techniques of Thin Layer Chromatography in Amino Acid and Peptide ChemiStry. Ed. G. Pataki. Walter de Gruyter and Co., Berlin. Brundorfer, K.R. and C. Green. 1967. The exchange of unesterified cholesterol between human low-density lipOproteins and rat erythrocyte "ghosts." Biochem. Burley, R.W. and W.H. Cook. 1961. Isolation and composi- tion of avian egg yolk granules and their con- stituent d- and B-lipovitellins. Can. J. Biochem. Physiol. 39:1295. Burley, R.W. and W.H. Cook. 1962a. The dissociation of a- and B-lipovitellin in aqueous solution. Part I. The effect of pH, temperature and other factors. Can. J. Biochem. Physiol. 40:363. Burley, R.W. and W.H. Cook. 1962b. The dissociation of a- and B-lipovitellin in aqueous solution. Part II. Influence of protein phosphate groups, sulfhydryl groups and related factors. Can. J. Biochem. Physiol. 40:373. Caldwell, R.C. and W. Pigman. 1965. Disc electrophoresis of human saliva in polyacrylamide gel. Arch. Biochem. Bi0phys. 110:91. Chang, P., W.D. Powrie, and O. Fennema. 1970. Disc gel electrOphoresis of proteins in native and heat- treated albumin, yolk, and centrifuged whole egg. J. Food Sci. 35:774. Chrambach, A., R.A. Reisfeld, M. Wyckoff and J. Zaccari. 1967. A procedure for rapid and sensitive staining of protein fractionated by polyacrylamide gel electrOphoresis. J. Anal. Biochem. 20:150. 100 Clegg, R.E., R.E. Hein, C.H. Suelter and R.E. McFarland. 1955. The distribution of radioactive phosphorus in the electrOphoretic components of egg yolk protein. P. Sci. 34:210. Clegg, R.E., M.M. Simlot and R. Janbunathan. 1971. Frac- tionation and analysis of the components in the high density fraction of whole egg yolk. P. Sci. 50:1565. Connelly, C. and G. Taborsky. 1961. Chromatographic fractionation of phosvitin. J. Biol. Chem. 236:1364. Cook, W.H., R.W. Burley, W.G. Martin and J.W. Hopkins. 1962. Amino acid compositions of the egg yolk lipoproteins and a statistical comparison of their amino acid ratios. Biochimw Biophys. Acta. 60:98. Cook, W.H. and W.G. Martin. 1962a. Composition and struc- ture of soluble lipOproteins. Biochim. Biophys. Acta. 56:362. Cook. W.H. and W.G. Martin. 1962b. Composition and prop- erties of soluble lipOproteins in relation to structure. Can. J. Biochem. Physiol. 40:1273. Cook, W.H. and R.A. Wallace.‘ 1965. Dissociation and sub- unit size of B-lipovitellin. Can. J. Biochem. 43:661. Davey, E.M., M.E. Zabik and L.E. Dawson. 1969. Fresh and frozen egg yolk protein fractions: emulsion stabil- izing power, viscosity and electrophoretic patterns. P. Sci. 48:251. Davis, B.J. 1964. Disc electrOphoresis II. Method and application to human serum proteins. Annals N.Y. Acad. Sci. 121:404. Davis, J.G., J.C. Zahnley and J.W. Donovan. 1969. Separa- tion and characterization of the ovoinhibitors from chicken egg white. Biochem. 8:2044. Dervichian, D.G. 1949. Structural aspect of lipoprotein .association. Discussions Faraday Soc. 6:7. Donovan, J.W., C.J. Mapes, J.G. Davis and R.D. Hamburg. 1969. ' Dissociation of chicken egg-white macroglobulin into subunits in acid. Hydrodynamic, spectrophotometric, and Optical rotatory measurements. Biochem. 8:4190. 101 Donovan, J.W., J.G. Davis and L.M. White. 1970. Chemical and physical characterization of ovomucin, a sul- fated glyc0protein complex from chicken eggs. Biochim. Biophys. Acta. 207: 190. Dunker, A.K. and R.R. Rueckert. 1969. Observations on molecular weight determinations on polyacrylamide gel. J. Biol. Chem. 244:5074. Edman, P. 1950. Method for determination of the amino acid sequence in peptides. Acta Chem. Scand. 4:283. Engstrom, L. 1961. Further studies on the incorporation of inorganic phosphate into calf-intestinal alkaline phosphatase. Biochim. Biophys. Acta. 54:179. Evans, R.J. and S.L. Bandemer. 1957. Separation of egg yolk proteins by paper electrophoresis. J. Agr. Food Chem. 5:868. Evans, R.J., S.L. Bandemer, J.A. Davidson, K. Heinlein and 8.8. Vaghefi. 1968. Binding of lipid to protein in the low density lipOprotein from the hen's egg. Biochim. Biophys. Acta. 164:566. Felgenhauer, K. 1970. Quantitation and specific detection methods after disc electrophoresis of serum proteins. Clin. Chim. Acta. 27:305. Fevold, H.L. and A. Lausten. 1946. Isolation of a new lipOprotein, lipovitellenin, from egg yolk. Arch. Biochem. 11:1. Fraenkel-Conrat, H., N.S. Snell and E.D. Ducay. 1952. Avidin II. Composition and mode of action-of avidin A. Arch. Biochem. Biophys. 39:97. Fraenkel-Conrat, H. 1954. A technique for stepwise degrad- ation of proteins from the amino end. J. Amer. Chem. Soc. 76:3606. Fraenkel-Conrat, H., J.I. Harris and A.L. Levy. 1955. The phenylisothiocyanate method for stepwise degradation from the amino end. Methods Biochem. Anal. 2:393. Gornall, D.A. and A. Kuksis. 1971. Molecular species of glycerOphosphatides and triglycerides of egg yolk lipOproteins. Can. J. Biochem. 49:51. Hartley, F.R. and F.R. Jevons. 1962. The attachment of carbohydrate in ovomucoid. Biochem. J. 82:1. 102 Helenius, A. and K. Simons. 1971. Removal of lipids from human plasma low density lipOprotein by detergents. Biochem. 10:2542. Hui, S.E. and R.H. Common. 1966. Studies on the livetins of egg yolk. III. Heterogeneity and immunoelectro- phoretic behaviour of beta-livetin. Can. J. Biochem. 44:1357. Johns, E.W. 1967. The electrophoresis of histones in poly- acrylamide gel and their quantitative determination. Biochem. J. 104:78. Joubert, F.J. and W.H. Cook. 1958a. Preparation and char- acterization of phosvitin from heg egg yolk. Can. J. Biochem. Physiol. 36:399. Joubert, F.J. and W.H. Cook. 1958b. Separation and char- acterization of lipovitellin.from hen egg yolk. Can. J. Biochem. Physiol. 36:389. Keyser, J.W. 1965. Staining of serum glyc0proteins after electrOphoretic separation. Anal. Biochem. 9:249. Kojiv, E. and A. Gronwall. 1952. Staining of protein—bound carbohydrates after electrophoresis of serum on filter paper. Scand. J. Clin. Lab. Invest. 4:244. Li, J.C.R. 1964. Statistical Inference I. Edwards Brothers Inc. Ann Arbor, MI. Lineweaver, H., H.J. Morris, L. Kline and R.S. Bean. 1948. Enzymes of fresh hen eggs. Arch. Biochem. 16:443. Lush, I.E. 1961. Genetic polymorphisms in the egg albumin proteins of the domestic fowl. Nature. 189:981. Mandeles, S. 1960. Use of DEAE-cellulose in the separation of protein from egg white and other biological mate- rials. J. Chromatog. 3:256. Mano, Y. and F. Lipmann. 1966. Characteristics of phospho- proteins (phosvitins) from a variety of fish roes. J. Biol. Chem. 241:3822. Marshall, M.E. and H.F. Deutsch. 1951. Distribution of egg white proteins in chicken blood serum and egg yolk. J. Biol. Chem. 189:1. 103 Martin, W.G. 1961. Molecular weight of bovine plasma albumin, vitellenin, and their oxidation products in formic acid. Can. J. Biochem. Physiol. 39:1405. Martin, W.G., J. Augustyniak and W.H. Cook. 1964. Fraction— ation and characterization of the low density lipo- proteins of hen's egg yolk. Biochim. Biophys. Acta. 84:714. Martin, W.G. and W.H. Cook. 1958. Preparation and molecular weight of gamma livetin from egg yolk. Can. J. Biochem. Physiol. 36:389. Martin, W.G., N.H. Tattrie and W.H. Cook. 1963. Lipid extraction and distribution studies of egg yolk lipoproteins. Can. J. Biochem. Physiol. 41:657. Martin, W.G., K.J. Turner and W.H. Cook. 1959. Macromolecu- lar prOperties of vitellenin from egg yolk and its parent complex with lipids. Can. J. Biochem. Physiol. 37:1194. Martin, W.G., J.E. Vandegaer and W.H. Cook. 1957. Fraction- ation of livetin and the molecular weights of the alpha and beta components. Can. J. Biochem. Physiol. 35:241. McCulley, K.A., W.A. Maw and R.H. Common. 1959. Zone elec— trOphoresis of the proteins of the fowl's serum and egg yolk. Can. J. Biochem. Physiol. 37:1457. McDonald, H.J. and L.P. Ribeiro. 1959. Ethylene and prOpylene glycol in the pre-staining of lipoproteins for electro- phoresis. Clin. Chim. Acta. 4:458. McFarlane, A.S. 1949. State of lipids in blood plasma. Discussions Faraday Soc. 6:74. Mecham, D.K. and H.S. Olcott. 1949. Phosvitin the principal phosphoprotein of egg yolk. J. Am. Chem. Soc. 71:3670. Mok, C.C. and R.H. Common. 1964. Studies on the livetin of hen's egg yolk. I. Identification of paper electro- phoretic and immunoelectrophoretic livetin fractions with serum protein antigens by immunoelectrOphoretic analysis. Can. J. Biochem. 42:871. Mok, C.C., C.T. Grant and G. Taborsky. 1966. Countercurrent distribution of phosvitin. Biochem. 5:2517. 104 Mok, C.C., W.G. Martin and R.H. Common. 1961. A comparison of phosvitins prepared from hen's serum and from hen's egg yolk. Can. J. Biochem. Physiol. 39:109. Neelin, J.M. and G.E. Conell. 1959. Zone electrophoresis of chicken—erythrocyte histone in starch gel. Biochim. Biophys. Acta. 31:539. Neelin, J.M. and W.H. Cook. 1961. Terminal amino acids of egg yolk lipoproteins. Can. J. Biochem. Physiol. 39:1075. Nelson, C.A. 1971. The binding of detergents to proteins. I. The maximum amount of dodecyl sulfate bound to proteins and the resistance to binding of several proteins. J. Biol. Chem. 246:3895. Neurath, H. 1963. The Proteins, V01. I. Academic Press. New York. Nichols, A.V., G.M. Forte and E.L. Coggiola. 1969. Electron microscopy of very low density lipoproteins from egg yolk using negative staining. Biochim. Biophys. Acta. 175:451. Ornstein, L. 1964. Disc electrophoresis — 1 Background and theory. Annals. N.Y. Acad. Sci. 121:321. Pankhurst, K.G.A. 1949. The adsorption of size of ions on the binding of amphipathic anions and cations to gelatin. Discussions Faraday Soc. 6:52. Pastewka, J.V., A.T. Ness and A.C. Peacock. 1966. Electro— phoretic patterns of normal human serums by disc electrOphoresis in polyacrylamide gel. Clin. Chim. Acta. 14:219. Pitt-Rivers, R. and F.S.A. Impiombato. 1968. The binding of sodium dodecyl sulfate to various proteins. Biochem. J. 109:825. Powrie, W.D., H. Little and A. Lopez. 1963. Gelation of egg yolk. J. Food Sci. 28:38. Prat, J.P., J.N. Lamy and J.D. Weill. 1969. Coloration des lipOproteines apres electrophorése en gel de poly- acrylamide. Bull. Soc. Chim. Biol. 51:1367. Rabinowitz, M. and F. Lippmann. 1960. Reversible phosphate transfer between yolk phosphoproteins and ATP. J. Biol. Chem. 235:1043. 105 Radomski, M.W. and W.H. Cook. 1964a. Chromatographic separa- tion of phosvitin: a— and B—lipovitellin of egg yolk granules on TEAE cellulose. J. Biol. Chem. 42:1203. Radomski, M.W. and W.H. Cook. 1964b. Fractionation and dissociation of the avian lipovitellins and their interactions with phosvitin. Can. J. Biochem. 42:395. Raymond, S., J.L. Miles and J.C.J. Lee. 1966. LipOprotein patterns in acrylamide gel electrophoresis. Science. 151:346. Ressler, N., R. Springgate and J. Saufman. 1961. A method of lipOprotein electrOphoresis. J. Chromatog. 6:409. Reynolds, J.A. and C. Tanford. 1970a. Binding of dodecyl sulfate to proteins at high binding ratios. Possible implications for the state of proteins in biological membranes. Proc. Nat. Acad. Sci. U.S. 66:1002. Reynolds, J.A. and C.J. Tanford. 1970b. The gross conforma- tion of protein sodium dodecyl sulfate complexes. J. Biol. Chem. 245:5161. Romanoff, A.L. and A.J. Romanoff. 1949. In The Avian Egg. John Wiley and Sons, Inc., New York. Rosenstein, R.W. and G. Taborsky. 1970. Mechanism of the oxidative dephosphorylation of the phosphOprotein phosvitin. Biochem. 9:649. Saari, A., W.D. Powrie and O. Fennema.' 1964. Isolation and characterization of low-density lipoproteins in native egg yolk plasma. J. Food Sci. 29:307. Scanu, A. 1967. Binding of human serum high density lipo- protein apoprotein with aqueous dispersions of phos- pholipids. J. Biol. Chem. 242:711. Schmidt, G., M.J. Bessman, M.D. Hickey and S.J. Thannhauser. 1956. The concentration of some constituents of egg yolk in its soluble phase. J. Biol. Chem. 223:1027. Schneider, H._and N.H. Tattrie. 1968. Mutual solubility of the lipid components of egg yolk low-density lipo- protein. Can. J. Biochem. 46:979. 106 Schroeder, W.A. 1967. Degradation of peptides by the Edman method with direct identification of the phenylthioe hydantoin-amino acid. In Methods in Enzymology. Vol. II. p. 445. Ed. C.H.W. Hirs. Academic Press. New York. Shapiro, A.L., E. Vinuela and J.V. Maizel Jr.- 1967. Molecu- lar weight estimation of polypeptide chains by elec- trophoresis in sodium dodecyl sulfate polyacrylamide gels. Biochem. BiOphys. Res. Comm. 28:815. Shepard, C.C. and G.A. Hottle. 1949. Studies of the compo- sition of the livetin fraction of the yolk of hen's eggs with the use of electrophoretic analysis. J. Biol. Chem. 179:349. Simlot, M.M. and R.E. Clegg. 1967. Preparation of an electrophoretically homogeneous phosphoprotein from egg yolk. P. Sci. 47:617. Smith, D.B. and K.J. Turner. 1958. N—terminal amino acids of the proteins of the floating fraction of egg yolk. Can. J. Biochem. Physiol. 36:951. Smithies, O. 1959. Zone electrophoresis in starch gels and its application to studies of serum proteins. In Advances in Protein Chemistry. Eds. M.L. Anson, K. Bailey and J.T. Edsall. Vol. 14, p. 75. Academic Press, Inc., New York. Smithies, O. 1962. Molecular size and starch gel electro— phoresis. Arch. Biochem. BiOphys. 98 Supp. 1, p. 125. Steer, D.C., W.G. Martin and W.H. Cook. 1968. Structural investigation of the low density lipoprotein of hen's egg yolk using proteolysis. Biochem. 7:3309. St. Groth, F.S., DE, R.G. Webster and A. Datyner. 1963. Two new staining procedures for quantitative estima- tion of proteins on electrophoretic strips. Biochim. Bi0phys. Acta. 71:377. Sugano, H. 1957. Studies on egg yolk proteins. II. Electrophoretic studies on phosvitin, lipovitellin rand lipovitellenin. J. Biochem. 44:205. Sugano, H. 1959. Studies on egg yolk proteins. V. Electro- phoretic and ultracentrifugal investigations on the homogeneities and some prOperties of a- and B- lipovitellin. J. Biochem. 46:417. 107 Sugano, H. and I. Watanabe. 1961. Isolations and some prOperties of native lipoproteins in egg yolk. J. Biochem. 50:473. Sundararajan, T.A., S. Kumar and P.S. Sarma. 1960. A sim— plified procedure for the preparation of phosvitin and vitellin. Biochim. BiOphys. Acta. 38:360. Taborsky, G. ‘1963. Interaction between phosvitin and iron and its effect on a rearrangement of phosvitin struc- ture. Biochem. 2:266. Tung, J.S. and C.A. Knight. 1972. Relative importance of some factors affecting the electrophoretic migration of proteins in sodium dodecyl sulfate polyacrylamide gels. Anal. Biochem. 48:153. Turner, K.J. and W.H. Cook. 1958. Molecular weight and physical properties of a lipoprotein from the float- ing fraction of egg yolk. Can. J. Biochem. Physiol. 36:937. Ullmann, A., M.E. Goldberg, D. Perrin and J. Monod. 1968. On the determination of molecular weight of proteins and protein subunits in the presence of 6M guanidine hydrochloride. Biochem. 7:261. Vandegaer, J.E., M.E. Reichmann and W.H. Cook. 1956. Preparation and molecular weight of lipovitellin from egg yolk. Arch. Biochem. Biophys. 62:328. Wallace, R.A., D.W. Jared and A.Z. Eisen. 1966. A general method for the isolation and purification of phosvitin from vertebrate eggs. Can. J. Biochem. 44:1647. Warner, R.C. 1954. In The Proteins. Vol. II. p. 435. Academic Press Inc., New York. Watt, B.K. and A.L. Merrill. 1963. Composition of Foods. Agricultural Handbook #8. Consumer and Food Economics Research Division, Agricultural Research Service. USDA. p. 6. Weber, K. 1968. New structural model of E. coli aspartate transcarbamylase and the amino-acid sequence of the regulatory polypeptide chain. Nature. 218:1116. Weber, K. and M.J. Osborn. 1969. The reliability of molec- ular weight determinations by sodium dodecyl sulfate polyacrylamide gel electrOphoresis. Biol. Chem. 244:4406. 108 Weinmann, E.O. 1956. Lipid-protein complex in egg yolk. Fed. Proc. 15:381. White, A., P. Handler and E.L. Smith. 1968. Principles of ‘ Biochemistgy. McGraw-Hill Book Co., New York. Williams, J. 1962a. Serum proteins and the livetins of hen's egg yolk. Biochem. J. 83:346. Williams, J. 1962b. A comparison of conalbumin and trans- ferrin in the domestic fowl. Biochem. J. 83:355. Zacharius, R.M., T.E. Zell, J.H. Morrison and J.J. Woodlock. 1969. Glycoprotein staining following electrophor- esis on acrylamide gels. Anal. Biochem. 30:148.