ISOLATION AND CHARACTERIZATION OF BOVINE LACTEAL IMMUNOGLOBULINS Thesis for the Degree of ~Ph.._ D. MICHIGAN STATE UNIVERSITY ROGER W. FRANZEN, IR. 1971 "_ ‘A .4— .A‘ A I, 9 . | . IIIIIIIIIIII I IIIIIIIIIIIII III 3 1293 10419 vulVCIDILy ' “than“ , . “Mag ‘2...“ . I. . This is to certify that the thesisentitled ISOLATION 4ND QNHRMTEIEI'LAWWON 04.3 BOVINE Lina-EAL ImmuuosroBUL/Ng presented by R0691 LU. Fran/oven )Te. has been accepted towards fulfillment of the requirements for PhD (193133111504 Science. r— Huma r) IULCI‘PH-I'OIJ Q fleets—e73” Mo: professor Date u] 10 I?! 0-7639 ABSTRACT ISOLATION AND CHARACTERIZATION OF BOVINE LACTEAL IMMUNOGLOBULINS By Roger W. Franzen, Jr. The immunoglobulins present in bovine milk are heterogeneous, high molecular weight proteins possessing antibody activity. Three anti- genically distinct classes of bovine immunoglobulins have been described. They are designated IgM, IgA, and IgG, the latter being divided into two subclasses, IgGl and 1362. These proteins comprise between 1.4-and 2.82 of the skim milk proteins. The purpose of this investigation was to determine which immune- globulin was responsible.for clustering milk fat globules and to better elucidate its mode of action in the creaming phenomenon. The immuno- globulins were isolated from whey by.ammonium sulfate precipitation, then fractionated by preparative ultracentrifugation and Sepharose 63 column chromatography into enriched molecular size classes. The.identi- fication of components within fractions was accomplished with poly- acrylamide disc gel electrophoresis and immunoelectrOphoresis, as well as other physical and chemical procedures. The 1008 pellet was observed to be very Opaque, and contained approximately 292 lipid and 122 carbohydrate. A ISS-species of IgG1 appeared to be loosely associated with the major 1908 component when this fraction was centrifuged in 0.15 M.NaCl at 20°C. A similar Roger W. Franzen, Jr. association was observed when this fraction was centrifuged at 4°C in a simulated milk salt buffer; however, the major component sedimented as a 1508 species. The eluate corresponding to the void peak of Sepharose 68 column chromatography, fraction 1, also exhibited some opacity. It contained approximately 73% lipid and 14% carbohydrate and was observed to be heterogeneous in sedimentation velocity determinations. A majority of the components in this fraction sedimented in the 20‘708 range. A lipoprotein-like boundary gradient was observed when this fraction was centrifuged in 20°C 0.15 M NaCl. In 4°C simulated milk salt buffer, this lipoprotein gradient was not apparent. Fraction 2 contained approximately 8.5% lipid and 10% carbohydrate and comprised principally IgM when analyzed by acid polyacrylamide disc gel electrophoresis and immunoelectrophoresis. The sedimentation velocity patterns of this fraction in 20°C 0.15 M NaCl showed a 208 major component which aggregated to a 258 component in 4°C simulated milk salt buffer. Fraction 3 was lipid free and contained 2.9% carbohydrate. Immuno- electrophoresis indicated that this fraction consisted of IgG; however, a minor quantity of IgA was also present. The sedimentation velocity studies revealed that a 7.18 species predominated in 0.15 M NaCl at 20°C, which aggregated to an 8.78 species in 4°C simulated milk salt buffer. Serine was found to be the most abundant amino acid in all fractions. The fractions were all mildly acidic and contained more hydrophobic than hydrophilic residues. Roger W. Franzen, Jr. A qualitative lipid analysis indicated that cholesterol and sphingo- myelins make up a majority of the lipids present in the fractions. When a heat-inactivated, recombined creaming system was used to evaluate the creaming ability of the unheated immunoglobulin fractions, only fraction 2 exhibited a normal creaming function. This fraction con- tained mainly IgM. When reduced to monomers with 2~mercaptoethanol, its creaming ability was-destroyed. ‘The cryoaggregation of IgM from a 208 to a 258 molecule in the presence of simulated milk salt buffer was postulated as being of fundamental importance to mechanisms explaining the creaming phenomenon. ISOLATION AND CHARACTERIZATION OF BOVINE LACTEAL IMMUNOGLOBULINS By ,J Id“ Roger W's-I Franzen , Jr . A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Dapartment of Food Science and Human Nutrition 1971 DEDICATION This manuscript is dedicated to the memory of my father, Roger W. Franzen, Sr. April 30, 1920-April 8, 1969 ii ACKNOWLEDGMENTS The author wishes to express his profound appreciation to Dr. J. R. Brunner for his most valuable advice and never-ending patience, under- standing and encouragement throughout the course of this study. Grateful appreciation is also expressed to the other members of. his guidance committee, Dr. L. R. Dugan, Dr. H. A. Lillevik, Dr. P. Markakis, and Dr. A. M. Pearson, for their helpful suggestions. Special thanks are also extended to Dr. J. E. Butler (USDA) for providing his excellent disc gel electrophoretic technique as well as anti-sera which were invaluable to this study. The author is also indebted to Miss Ursula Koch for providing technical assistance with the. amino acid and analytical ultracentrifugation analyses used in this study. Grateful acknowledgment is also due Dr. B. S. Schweigert and Dr. G. A. Leveille for providing the research facilities and to the Food and.Drug Administration, Grant No. FD00210-04 BAC for providing the funds necessary for this research. Finally, the author wishes to express most grateful thanks to his- wife, Linda, for her steadfast love and devotion, as well as her valuable assistance throughout his graduate program and in the preparation of this manuscript. iii TABLE OF CONTENTS INTWDUCTION O O . O O - O O O O I. O O O O O I O I O O O O O O O O 0 REVIEW OF LITERATURE. o o o o o '0 o o o .0 o o o '0 o o o .0 o o o . Immun0310bl111n8. o o o o o s o o o I. o o o As 0 o o I. a Physicochemical Properties of the Three Bovine Immuno- globulin Classes . . . . . . . . . . . . . . .'. . . . Isolation of Immunoglobulins . . . . . . .‘. . . .-. . . Creaming . . . . . . . . . . . . . .-. . . .‘. . . . . . EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . . .'. . Apparatus-and Equipment. . . . . . . . . . . .'. . . .'. Chemicals and Materials. . . . . .'. . . .v. . . .~. . Preparative Procedures . . .-. . . .'. . . .'. . . . x . Preparation of immunoglobulins. .~. . . . . . . . Precipitation of immunoglobulins with ammonium sulfate . . . . . . . . . . . Separation of immunoglobulin fractiOns by gel filtration . . .-. . . . . . . . . Preparation of cream and skim milk for creaming experiments . . . . . . . . . . . . ._. . . . . Chemical MethOds O C O O O O I. O O O ' O O O O C O O O I. O Nitrogen. . . . . . . . . .‘. . . . . . . .'. . . Hexose. . . . . . . . . . .'. . . .-. . . .'. . . Hexosamine. . . . . . . . . . . . . . . . . . .-. Sialic acid . . . . . . . . . . . . . . . . . . . Fucose. . . . . . . . . . . . . . . . . . . . . . Amino acid. . . . . . . . . . . . . . .‘. . . .'. Folin-Lowry nitrogen. . . . . . . . . . . . . . . Total lipids. . . . . . . . . . . . . . . . . . . Babcock fat determination . . . . .'. . . . . . . iv Page l4 14 15 - 17 '17 17 ' 21 22 - 23 23 23 24 25 26 27 28 ‘ 29 30 Page Physical MethOds C I O O O I O O O O O I O 0 O O O I O O O O 30 Alkaline Polyacrylamide Disc Gel Electrophoresis in a Discontinuous Buffer System. . . . . . . . . 30 Acid Polyacrylamide Disc Gel Electrophoresis in a Discontinuous Buffer System . . . . . . . . . . . 31 ImmunoelectrOphoresis . . . . . . . . . .-. . . .,. . 32. Thin-Layer Chromatography . . . . . . . . . . . . . . 34 Sedimentation Coefficient . . . . . . . . . . . . . . 34 creaming StUdies I O O O O O O O O O O O O O O O O O I O I. O 37 cream; VOlum O I I O O O I O I 0 O I. O O O O O O O O O 37 RESULTS AND DISCUSSION. 0 O O O O O O V I O O O O O O O .0 O O O O O O 39 Preparative Procedures . . . . . . . . . . . . . .‘. . . . . 39 Preparation of Immunoglobulins. . .‘. . . .'. . . .'. 39 Component Assessment by Polyacrylamide Disc Gel ElectrophoreSis O O « O O O O O O O O O O O O O O O O 42 . Component Assessment by Immunoelectrophoresis . . . . 46 Chemical Analyses O O O O O O O O '0 O O O '- 0 O O O O O O O '0 O 49 ' Lipids .‘ O O O O O O O O O I. O O O ‘ O O O O O O O O O O 53 Physical Analysis. . . . . . . . . . . . . . . . . . . .-. . 57 Sedimentation Velocity. . . . . . . . . . .'. . . . . 57 Creaming Studies . . . . . . . . . . . . . . . . . . . . . . 67 The Effects of Temperature on Creammng. . . . . . . . 67 The Effects of the Isolated Immunoglobulin Fractions on Creaming in a Model System . . . . . . 68 The Effects of Heating on a Recombined Creaming system. 0 O O O O O O . O O I O I. O O O .0 O O O I O 69 The Effects of the Isolated Immunoglobulin Fractions on a Heated Recombined Creaming System. . 70 The Effects of Neuraminidase on Creaming. . .'. . . .~ 74 The Effect of Disulfide Reducing Agent on Creaming. . 75 SUMY O I O O O O I O O O O O O O O O O O O O O . I O O O ' O O O O O C 79 BIBLIOGRAPMO o o o o o o o '0 o o o o o o o .0 o s o ‘0 o o o '0 o o o . 82 APPENDIX 0 O O 0 O O O O O O I O O O O O O O O O O O I O O O O O O O I I 86 LIST OF TABLES Table Page 1 Chemical composition of immunoglobulin fractions . . . . . . 50 2 Amino acid composition of immunoglobulin fractions . . . . . 51 3 The apparent sedimentation coefficients for separated immunoglobulin fractions centrifuged at 4°C and 20°C in different solvent systems. '. . . .'. . . . . . . . . . . . . 58 4 The effects of temperature on the creaming of raw milk .mlest O .O O O O I. O O I O I O O I O I O O O I O O O O O O 67 5 The effects of immunoglobulin fractions on.a model cream? ing system of Jenness and K00ps buffer plus cream. . . . . . 68 6 The effects of heat on the creaming factor in skim milk. . . 7O 7 The effects of isolated immunoglobulin fractions on the raw cream and heated skim model system . . ._. . . . . . . . 71 8 The effects of incorporating centrifuged heated skim.milk fractions into a creaming system of centrifuged whey, raw cream, and 2 mg fraction 2/25 m1 . . . . . . . .‘. . . .p. . 73 9 Creaming as affected by neuraminidase treatment of fraction 2, in a system of raw cream plus heated skim . . . . . . . . 75 10 Creaming as affected by incorporating mercaptoethanol into raw milk;.heated skim model system containing fraction 2; and fraction 2 followed by alkylation and incorporation into the heated skim model system. . . . . . . . . . . . . . 76 vi LIST OF FIGURES Figure Page 1 Schematic for the fractionation of immunoglobulins from cow. 8 milk I I I I I I I I I I I I I I I .I I I I I I I I II I 20 2 Elution chromatogram of crude immunoglobulin preparation through Sépharose 6B column with 0.1M Tris-HCl, 1 M NaCl + 0.02% sodium azide, pH 7.2. A-Crude Ig fraction; B-Crude Ig fraction minus 1008 pellet. . . . . . . . . . . . .‘. . . 41 3 Polyacrylamide disc gel electrOphoretic patterns under A, alkaline and B, acid conditions: l-acid whey; 2-crude Ig prep; 3-1003 pellet, 4-fraction 1; 5- fraction 2; 6- fraction 3. Twenty ul of 2% solutions were applied except in alkaline gel 1 which contained 30 pl . . . . . . . . . . . . 44 4 Immunoelectrophoretic patterns of isolated fractions: Plate A'- crude Ig prep; B - 1008 pellet; C — fraction 1; D - fraction 2; E - fraction 3 . . . . . . . . .'. . . .'. . 48 5 Thin layer chromatographic patterns of A, neutral lipids (silica gel G) and B, phospholipids (silica gel H): Spot 1-1008 pellet; 2-fraction 1; 3-fraction 2; 4-fraction 3; 5-crude Ig prep; 6-cholesterol std.; 7-tributyrin std. . .'. 55 6 Sedimentation velocity patterns of isolated immunoglobulin fractions: Arin 0.15 M NaCl + 0.02% sodium.azide pH 6.8 at 20°C; B-in Jenness and Koops buffer + 0.02% sodium azide pH 6.6 at 4°C. Row 1-1008 pellet; 2-fraction 1;‘3-fraction, 2; 4-fraction 3. Rows 1-3 0.75% protein, Row 4é1.1% protein 60 7 Sedimentation velocity patterns of.fraction.2: Arin Jenness and Koops buffer at 20°C; B-in 0.15 M NaCl at 4°C. 9 I 60°; 0.75% protein; 59,780 rpm . . . . . . . . . . . . . 64 vii INTRODUCTION The immunoglobulins present in bovine milk are described as hetero- geneous, high molecular weight proteins possessing antibody activity. They are usually isolated from the whey portion of milk as a minor fraction and are also present in blood serum and other milk secretions. Three antigenically distinct classes of bovine immunoglobulins have' been described. They are designated IgM, IgA, and IgG, the latter being divided into two subclasses, IgGl-and IgG2. In the earlier dairy science literature, the immune proteins were. isolated in the "lactoglobulin" fraction which is classically defined as that portion of the whey proteins precipitated by saturation with magnesium sulfate, or by half-saturation with ammonium sulfate. Upon- exhaustive dialysis this fraction separates into two components; a. precipitate called euglobulin and a protein remaining in solution called pseudoglobulin. The euglobulin has recently been shown to contain IgA, IgM, IgG2 and electrophoretically slower IgGl globulins, while the pseudo- globulin consists mainly of IgG1 and secretory IgA. Previous studies have indicated that the euglobulin fraction is very active in promoting the fat globules in milk to cluster, which is pre- liminary to the creaming phenomenon observed in milk. However, this fraction, as described previously, is made up of different types of immunoglobulins all of which are heterogeneous and exhibit different physicochemical properties. A limited amount of information is available 2 on which immunoglobulin is responsible for creaming and even less is available concerning how it functions in the creaming phenomenon. This study was undertaken to identify which of the bovine immuno- globulins is responsible for the clustering and creaming of milk fat globules. The immunoglobulins were isolated by c1assica1.methods and were then fractionated and enriched, based on molecular size, into their respective classes. The identification of components within fractions was accomplished with polyacrylamide disc gel electrophoresis and immuno— electrophoresis, as well as other physical and chemical procedures. Sedimentation velocity studies were performed on the isolated fractions and the information obtained was considered pertinent.to mechanisms explaining the creaming phenomenon. Creaming studies were performed in natural as well as model systems to better understand the role played by immunoglobulins in the creaming phenomenon. REVIEW OF LITERATURE Immunoglobulins, The term immunoglobulins when used in reference to milk proteins describes a group of heterogeneous, large molecular weight proteins, sharing common physicochemical prOperties and antigenic determinants. The third revision of the Nomenclature of the Proteins of Cows Milk (Rose et_aZ., 1970) recommends the adOption of this term to update the earlier dairy science literature and to be consistent with the nomen- clature of more extensively studied species in accord with the Werld Health Organization Report (1964). The term immunoglobulin therefore would replace terms like "immune lactoglobulin", "gamma globulin", "euglobulin", "pseudoglobulin", and "T-globulin", terminology used in earlier dairy science literature. Butler (1969) states that three antigenically distinct classes of bovine immunoglobulins have been described. They are designated IgM, IgA and IgG, the latter being divided into two subclasses, IgG1 and IgG2. All have been found to be present in serum as well as the lacteal secretions. Murphy et'aZ. (1964) employed physicochemical as well as immunochemical methods to demonstrate that the three immunoglobulins were present in milk during lactation, dry secretion and colostrum synthesis. Butler (1969) states that immunoglobulins are found in the serum and other body fluids of animals and possess y- or slow B-electrophoretic 4 mobility, and that these include all molecules with antibody activity, astell as other chemically related normal or pathological proteins. He further reports that all immunoglobulins appear to be either monomers or polymers of a,four7chain molecule consisting of two light polypeptide chains (L-chains:20,000 molecular weight) and two heavy polypeptide chains (H-chains) with molecular weights varying from 50,000 to 70,000 daltons for the different immunoglobulin classes. Carpenter (1965) states that the principal immunoglobulin of normal serum is a 7S molecule with a molecular weight of approximately 160,000. He also reports that most animal species contain a small amount of macro- globulin in-the 178 to 208 class, with a molecular weight.of 900,000 to 1,000,000. Traces of 288 to 448 components are also found. The macroe globulins appear to be polymers held together by disulfide bridges, but they are not simple polymers of 7s globulin. They differ notably in carbohydrate content; approximately ten per cent for macroglobulins and about 2.5% for the 7S immunoglobulins. Physicochemical Properties of the Three Bovine Immunoglobulin Classesa IgM is a macroglobulin with a sedimentation constant of 198 that contains approximately 12% carbohydrate. It is reported to represent between 0.1 to 0.2% of the protein present in skim milk and has a molecular weight approaching 1,000,000. This protein is eluted in the void volume peak of Sephadex G-200 when the protein fraction from whey, insoluble- in 33% ammonium sulfate, is used as the starting material. When the same starting material is fractionated on acid-disc electrophoresis at aFrom-Rose et a2. (1970) when specific reference not indicated. 5 pH 4.3, the IgM present does not enter the separating gel but forms a dense band at the stacking gel-separating gel interface. IgM is thought to be a pentamer of the basic four chain immunoglobulin molecule which loses.its antibody activity and is reduced to a monomer in the presence of 2-mercaptoethanol. Murphy et a1. (1964) suggest that the length of the immunoelectro- phoretic are for IgM indicates the presence of a continuum of molecules differing in net charge in a manner similar to the spectrum of 7S gamma globulin molecules. Murphy et al. (1965) found bovine serum.IgM to have preperties quite similar to those of the analogous protein in human serum. This was determined by comparing results from gel filtration, immunoelectrOphoresis, anion exchange chromatography, ultracentrifugation and reduction with mercaptoethanol. The antigenically distinct immunoglobulin IgA is reported to repre- sent 0.05 to 0.10% of the protein present in bovine skim.milk. It has» a molecular weight in the range of 300,000 to 500,000 and often exists as a dimer of the basic four polypeptide immunoglobulin structure. The carbohydrate content of this molecule is between 8 and 9%. Jenness et al. (1965) report that it is sensitive to 2-mercaptoethanol and sediments as a 128 molecule. Lacteal IgA is eluted between the IgM and IgG peaks during Sephadex G-200 fractionation of whey. Groves and Gordon (1967) reported the isolation of glycoprotein-a from cows milk (48,000 daltons) and was later confirmed by Butler et a1. (1968) to be present in a free form and also bound to lacteal IgA. The bound.form is probably analogous to the "secretory IgA", described for other species. Porter (1969) found that IgA in sow colostrum was present in many polymeric forms 6 since it appeared in gel filtration fractions over a wide molecular weight range. He states that the secretory IgA in humans has a molecular weight of 390,000 compared with 170,000 for serum IgA. He also postulated that in the sow the high molecular weight form.might be synthesized in. the mammary gland while the lower molecular weight components were transferred directly from the serum. 0f the immunoglobulins present in bovine milk, IgG is by far the most abundant class. IgGl represents one to two per cent while IgG2 comprises approximately 0.2 to 0.5% of the skim milk protein. These molecules have a sedimentation coefficient of approximately 73.. Smith et a1. (1946) and Nolan and Smith (1962) reported a carbohydrate content of between 2 and 4%. Murphy et al. (1965) separated IgG into two sub— classes by chromatographic elution positions, electrophoretic migration rates and biological activities (complement fixation). They characterized these immunoglobulins by anion exchange chromatography, immunoelectro- phoresis, zone electrophoresis, ultracentrifugation and analysis of the products of papain digestion. They concluded that their properties were similar to those of analogous components in human serum. The more basic IgG molecules are_the IgG2 subclass of immunoglobulins which have a mean 820,w coefficient of 6.6. These molecules migrate most rapidly toward the cathodic electrode during electrophoresis at pH 4.3 in polyacrylamide disc gels. ‘The IgG2 immunoglobulins are not retained on DEAE-Sephadex in low ionic strength environment at pH 8.0 and thus are eluted in the breakthrough peak (Murphy et aZ., 1965). The subclass IgGl consists of the less basic molecules which are eluted at higher ionic strength than the IgG2 fraction on DEAE-Sephadex. They also appear more heterogeneous on immunoelectrophoresis than the IgGZ subclass. IgGl has a mean 820 w of 6.3. 7 Rose et a1. (1970) report that the two subclasses of IgG can be correlated with the early preparations of Smith (1948) in the following manner. Smith's pseudoglobulin fractions contain mostly IgG1 and secre- tory IgA, while his euglobulin fraction consists of IgG2-like globulins, slower IgGl globulins, IgA and IgM. Isolation of Immunoglobulins Carpenter (1965) discusses the three classical methods of fractiona- tion that have been employed for many years in the isolation of serum proteins. These methods utilize the solubility characteristics of the serum proteins in the (1) presence of neutral salts such as sodium, ammonium and magnesium sulfate; (2) in the presence.or absence of electro- lyte; and (3) in the presence of cold alcohol at various hydrogen ion concentrations. Crowther and Raistrick (1916) precipitated milk globulins with magnesium sulfate and further separated these proteins into water insoluble euglobulin and water soluble pseudoglobulin by means of exhaustive dialysis. Howe (1921, 1922) employed sodium sulfate to fractionate milk pro- teins in an analogous procedure to that used for.blood protein fractiona- tion. He found that euglobulin was precipitated at up to 14.2% salt' at 34°C and that pseudoglobulin I and II were precipitated at between. l4.2~18.4% and 18.4-21.5% salt, respectively. Smith (1946) stated that classical methods such as those of Crowther and Raistrick (1916), using repeated precipitation with half-saturated ammonium sulfate or saturated magnesium sulfate, yielded preparations which show complex electrophoretic patterns. He described a method for preparing electrOphoretically homogeneous immunoglobulin by a stepwise fractionation with ammonium sulfate. The crude globulin was precipitated from pH 6.5 whey by the addition of solid ammonium sulfate to 0.5 satura- tion. This fraction was redissolved at about 3% protein concentration, the pH was adjusted to 4.6, and ammonium sulfate added to 0.25 saturation. After removing the precipitate by centrifugation, the globulins were* filtered and the immune proteins were precipitated from the filtrate by adjusting it to 0.4 saturation with ammonium sulfate at pH 6.0. This fraction which contained 80% of the immune proteins, was reworked by dissolving in water at 1°C, adjusting to pH 4.5, and removing the insoluble residue by filtration. The filtrate was brought to 0.3 saturation with solid ammonium sulfate. The supernatant was adjusted to pH 6.0 and 0.4 saturation with ammonium sulfate and the precipitate collected. Both” immune globulin fractions were dialyzed separately against distilled water_at 2°C andresolved into water-soluble "pseudoglobulins" and water—insoluble "euglobulin" fractions. The last precipitate was.electro- phoretically homogeneous in moving boundary electrophoresis. Creaming :When normal, freshly drawn cows milk is chilled under quiescent conditions, the fat phase of the milk rises at a rapid rate and packs into a relatively loose lipid—rich layer. This phenomenon is known as creaming and usually occurs within one hour after optimum.conditions have been attained. Troy and Sharp (1928) estimated the rate of rise of fat globules from Stoke's law, concluding that if the fat globules rose individually approximately fifty hours would be required before a cream layer was attained. These workers also measured the rate at which clusters of fat globules rose and concluded that clustering accounted for the rapid formation of the cream layer. The mechanism that is 9 involved in the clustering of fat globules is still poorly delineated; thus it was with this fact in mind that the present study was undertaken. In 1899 Babcock observed that the fat globules in freshly drawn cows milk tended to aggregate and form clusters. His "coagulated fibrin" theory explaining this occurrence was later shown to have been an arti- fact. However, his recognition of the significance of fat globule clustering led to further investigations and the formulation of numerous theories to explain the phenomenon. Among these early theories, which were extensively reviewed by Dunkley and Sommer (1944) were: (a) gravitation of fat globules; (b) electrokinetic potential of fat globules; (c) fat-serum interfacial tension; (d) stickiness and state of hydration of the adsorbed membrane; and (e) fat clustering considered as an agglutination process. Dunkley and Sommer concluded, from.experimentation designed to test the above theories, that (a) the variable creaming properties of normal milk cannot be explained on the basis of differences in the rates of rise of individual fat globules; (b) that the salts normally present in milk are sufficient to reduce the surface potential on the fat globules and thus eliminate the charge variability of the fat globules; (c) that the inter- facial tension at the fat-serum interface does not promote creaming; (d) that evidence concerning the importance of hydration of the adsorbed membrane on the fat globules was not sufficient to justify drawing a definite conclusion regarding the significance of this factor. However, these workers isolated a protein from coldrseparated cream which had euglobulin characteristicsand promoted creaming. Sharp and Krukovsky (1939) also isolated a similar protein and considered the clustering of, fat globules as an agglutination process. They concluded that the 10 agglutinating substance normally present in milk is adsorbed on the surface of solid fat globules but not on liquid fat globules. Dunkley and Sommer (1944) also compared fat globule clustering with bacterial agglutination, citing their many similarities. Both processes involved the aggregation of particles, require the presence of globulins, are prevented by heat denaturation, and require optimum salt concentra- tions. They used Marrack's (1938) theory of bacterial agglutination as a model and stated that if the mechanisms were similar clustering would be promoted by (a) a partial dehydration of the adsorbed membrane on the fat globules affected by a specific polar adsorption of the euglobulin; (b) an aggregation of fat globules resulting from the adsorption of a single euglobulin molecule, jointly by two fat globules; (c) maintenance of the surface potential of the fat globules below the critical level by the presence of salts. Samuelsson et a1. (1954) observed that the agglutinin responsible for normal creaming in milk was.separated at 2°C asra yellowish powder from rennet whey which had been heated to 60°C. The agglutinin formed opalescent solutions in warm.whey or water and consisted of two compon- ents, one inactivated by homogenization and the other inactivated by heating in water, whey or milk to 70-75°C for 15 seconds. In systems containing cream mixed with water, whey-or separated milk, creaming would occur if one portion of the available agglutinin had been inactivated by homogenization and the other by heating, but if all the-agglutinin, had undergone either of these treatments, no creaming resulted. Kenyon and Jenness (1958) substantiated the agglutinin inactiva- tion work of.Samuelsson at al. (1954)-and also reported that the addition of 15 pg of colostral euglobulin per 100 ml of solution restored the ll creaming ability of heated skim milk but not of homogenized skim milk. Payens at d1. (1965) performed creaming experiments using 1311- labeled euglobulin from bovine colostrum. They were able to show that euglobulin is definitely adsorbed to the fat globule membrane, and that various amounts of B-lactoglobulin and casein are also adsorbed. Their studies at various temperatures showed that adsorption is temperature. dependent with a considerable decrease at 45°C and a much greater adsorption at 5° and 10°C. They also demonstrated that at surface concentrations of about 2 mg of.euglobulin per gram of fat, clustering time became constant. These results led them to the postulation that agglutination is brought about by the formation of euglobulin bridges between the fat globules with euglobulin peptide chains projecting into the plasma phase. K00ps at al. (1966) reported that low homogenization pressures, i.e., 10 kg/cmz, significantly decreased the creaming ability of milk reconsti- tuted from homogenized skim milk and unhomogenized cream (separated at 45°C). They suggested that denaturation of the euglobulin at these low pressures would be very unlikely. Instead, the results of these experi- ments indicated that casein micelles, particularly k-casein, were the inhibitory agents of clustering and creaming. They found that euglobulin was still adsorbed to the fat globule surface after homogenization and theorized that (a) homogenization caused euglobulin and k-casein to complex and that this complex was capable of adsorbing onto the fat surface but unable to effect clustering; or that (b) homogenization induces the adsorption of k-casein which would screen off the clustering sites of the adsorbed euglobulin. 12 Stadhouders and Hup (1970) found that Smith's (1946) euglobulin F isolated from colostrum not only contained the active agglutinin responsible for clustering and creaming but also an agglutinin for many bacterial species as well as one to attach these bacteria to fat globules. They reported that all three agglutinins were different. The cryoglobulin fraction, separated after aggregation at low temperature, was found to be a minor fraction (=8%) which appeared to contain the-antibodies responsible for agglutinating the fat globules as well as those which attach the bacteria to the fat globules. The antibodies which.agg1utinate the bacteria were found to remain complétely in the.non-cryoglobulin fraction. Gammack and Gupta (1970) concentrated the active clustering agent from milk by separating the cream at low temperatures and desorbing the adsorbed material from the fat globules into a small volume of aqueous phase. After ultracentrifugation of this aqueous phase, they recovered 85% of the clustering activity in-a small opalescent layer which sedimented above the casein pellet. A precipitate which formed when the opalescent layer was diluted with milk ultrafiltrate and held at 4°C contained most of the clustering activity. On warming, it redispersed. Further, the euglobulin fraction from colostrum also exhibited the cryoprecipitation behavior, but the precipitate from the milk preparation was twenty times more active in terms of its protein. content in promoting clustering and caused a more rapid formation of cream line than did euglobulin. When the aqueous phase obtained from milk was fractionated by gel filtration on agarose, the clustering activity was localized in a high particle weight fraction, i.e., 1 x 106, which contained lipOprotein particles. However, the isolated 13 lipoprotein particles themselves showed no clustering activity. Then, they separated colostral euglobulin by gel filtration on Sephadex G-150 at 50°C and found that IgG had no clustering activity with or without added lipoprotein particles. The fraction containing IgM and IgA showed some activity with a slow formation of cream line, but, on addition of lipoProtein particles, clustering activity was markedly enhanced and cream line formation was rapid. They further state that while protein molecules such as those in euglobulin preparations from colostrum promote clustering of fat globules these are not the most effective clustering agents in milk. Lipoprotein particles which exist in the aqueous phase augment the clustering effect of certain immune protein fractions such as those present in euglobulin preparations. They concluded that certain immune proteins as well as lipoprotein.particles have an affinity for each other as well as for sites on the fat globule surface to which they adsorb as a preliminary step to cluster formation.l EXPERIMENTAL. Apparatus and Equipment The milk used in this study was collected in five- or tendgallon stainless steel cans and separated with a DeLaval disc-type.separator (Model 518).. Stainless steel, plastic or Pyrex containers were used in performing all experiments. A Beckman Model 115 or an Instrumenta- tion Laboratory Model 245 pH meter, equipped with glass electrodes, was used to measure pH values. For weighings, a top-loading, directs reading Mettler Type K—7 or a Sartorius series 2400 balance were used. 'Lowdspeed centrifugations were performed with an International clinical centrifuge or a Sorvall super-speed centrifuge with a type SS-l rotor. Intermediate-speed centrifugations were performed with a. Sorvall RC2-B refrigerated centrifuge using type GSA rotor. For high- speed centrifugation, a Beckman Model L-65 refrigerated preparative ultracentrifuge, equipped with type 30 fixed-angle rotor was used. Sephadex laboratory columns, Type K25/45, equipped with up-flow adaptors and cooling jackets, were used for packing the supporting materials. The eluates from the columns were monitored at 254 nm with a recording ultraviolet analyzer manufactured by Instrumentation specialties Company, Inc. Protein solutions were dried from the frozen state on a laboratory- constructed lyophilizer. Lyophilized protein samples as well as some chemicals were dried in a vacuum obtained with a Cenco Hyvac 2 vacuum pump and stored over P205. 14 15 Laboratory—constructed Plexiglass electrophoretic cells were used for polyacrylamide disc gel electrophoresis as well as immunoelectro- phoresis. Power was supplied by a Heathkit variable voltage regulated power supply Model IP-32. A Polaroid Land Camera (Model MP-3) was used to photograph the electr0phoretograms and immunoelectrophoretograms. A Perkin-Elmer, Model 382A Tiselius electrophoresis apparatus, using circulating refrigerated water to maintain the bath temperature at 2°C, was.used for the free-boundary electrophoretic analyses. Buffer resist- ances were determined with an Industrial Instruments, Model RC, conductiv- ity bridge. Amino acid analyses were performed with a Beckman/Spinco Model 120C amino acid analyzer. A Beckman DK—ZA ratio recording spectrOphotometer, equipped with quartz cells with a 1 cm light path, was used for all spectrophotometric determinations. A Beckman/Spinco Model E analytical ultracentrifuge equipped with a RTIC temperature control unit.and a phase plate as a schlieren diaphragm was used for sedimentation-velocity studies. A capillary-type, single- sector synthetic boundary cell was used for sedimentation studies. In all determinations, 12 mm filled Epon centerpieces with quartz windows were used.. An An-D Duralumin rotor was used for centrifuging and a General Electric AH-6 Mercury lamp served as the light source. Kodak metallographic glass plates were used for recording the schlieren patterns and a Nikon.MOdel 6 Shadowgraph microcomparator was used for measuring the plates. Chemicals and Materials The principal chemicals used in this research and their suppliers are given below: l6 Tris-hydroxymethyl aminomethane (Sigma 121 and Trisma Base) was used as a primary standard and Sigma 7-9 was used to prepare buffers. Both were obtained from Sigma Chemical Co. as was iodoacetamide and neuraminidase (from CZ. perfringens). Diethyl-barbituric acid and sodium diethylbarbi- turic acid for veronal buffer, and monobasic potassium phosphate, potassium chloride, potassium sulfate, potassium citrate, sodium citrate, anhydrous calcium chloride, magnesium citrate hydrate, potassium carbonate, and potassium hydroxide for simulated milk salt solution were purchased from Fisher Scientific Company. Reagent grade ammonium sulfate was also purchased from Fisher Scientific Co. 2-thiobarbituric acid, N,N,N',N'- tetramethylenediamine, acrylamide, N,N'—methy1enebisacrylamide and sodium azide were obtained from Eastman Organic Chemicals. Cyanogum 41 used in polyacrylamide disc gel electrophoresis was purchased from E. C. Apparatus Company. Ammonium persulfate was obtained from the Baker Chemical Company and 2-mercaptoethanol was acquired from Fisher Scientific Co. Special Agar-Noble used in immunoelectrophoresis was obtained from Difco Laboratories. Orcinol was obtained from the Matheson Company, and phenol was acquired from Mallinckrodt and redistilled prior to use. Acetylacetone was purchased from Fisher Scientific Company. The Sepharose-6B used in this study was obtained from Pharmacia Fine Chemicals, as was the Blue Dextran 2000. Mannose was acquired from Fisher Scientific Co. and galactose was obtained from Pfanstiehl Laboratories, Inc. (Germany). Glucosamine hydrochloride, galactosamine, and fucose were purchased from Nutritional Biochemicals Corp. N-acetyl neuraminic acid and tryptophan were acquired 17 from Calbiochem.' The Buffalo Black NBR and Thiazine Red R.were.obtained from Allied Chemical Corp. The other organic and inorganic chemicals used in this research were of reagent grade. Preparative Procedures‘ The milk used in this study was obtained from the Michigan State university dairy herd which consisted of Holstein cows. Allrmilk was collected immediately after milking and before the milk reached the cooling bulk tank. The whole milk was heated to 45°C before separating in a laboratory separator, and the fresh skim milk was used as the start- ing material in this study. Preparation of immunoglobulins.--The immunoglobulin fractions were prepared by combining the ammonium sulfate precipitation method described by Smith (1946), with minor modification, followed by preparative. ultracentrifugation and column chromatographic technique using Sepharose 6B. The isolation procedure for the immunoglobulin fractions.employed- in this study is presented diagrammatically ianigure 1.. Details of this procedure are described in the following sections. Precipitation of immungglobulins with ammonium sulfate.-~After the cream was separated at 45°C, the resulting skim milk was cooled to 25°C and adjusted to pH 4.6 with l N HCl and allowed to stand for three hours. The precipitated casein was removed by filtration through multi-layered cheese cloth. The acid whey was adjusted to pH 6.5 with 1N NaOH and solid ammonium sulfate (313 g/l) was added slowly to 0.5 saturation while stirring, to salt—out the crude globulin fraction. 18 After standing overnight, the majority of the supernatant was siphoned off, and the precipitate was collected by centrifugation at 16,000 x g for 15 min at 20°C in the Sorvall GSA rotor. All further centrifugations were accomplished at 20°C in the Sorvall GSA rotor. The precipitate was redissolved to about 3% protein concentration in deionized water, the pH adjusted to 4.6 and ammonium sulfate added to 0.25 saturation (144 g/l). After centrifugation at 23,000 x g for 20 min to remove the suspended precipitate, the supernatant was-filtered through Whatman No. 42 filter paper. The filtrate was then adjusted to pH 6.0 with l N NaOH and the immunoglobulins precipitated by adding ammonium sulfate to 0.4 saturation (99 g/l). The precipitate was collected by centrifugation at 23,000 x g for 20 min and redissolved to approximately 3% protein concentration in 1°C deionized water.. It~ was then adjusted to pH 4.5 with 1 N HCl and refiltered through Whatman No. 42 filter paper. The filtrate was warmed to 20°C and ammonium sulfate was added to 0.3 saturation (176 g/l). The precipitate was removed by centrifugation at 65,000 x g for 20 min, corresponding to Smith's fraction E. The supernatant was adjusted to pH 6.0 with 1 N NaOH and was brought to 0.4-saturation with ammonium sulfate (67 g/l). The precipitate was collected by centrifugation at 23,000 x-g for 30 min at 20°C in GSA and was dissolved in deionized water to 3% protein concentration., The pH was_determined to ensure a pH of 6.0 and ammonium sulfate was added to 0.4 saturation (243 g/l). The resulting precipi- tate_corresponds to Smith's fraction F and was stored at 4°C under 0.4 saturated ammonium sulfate until further use. It was designated as "crude immunoglobulin preparation" throughout this study.7 19 use on not w x ooo.mo «an «outsource Cesar m.o on om NA mzv sea ooom on are: MHoaom azaommfi Bz<fi cots; _ wu pg; in sauoqzosqv . . 1. s v... Hi. sfiin .I...~.¢..hrl..i4 _ . . . )H... . ‘ Ila! V y’.’ i 42 The opacity of fraction 1 was greatly reduced, whereas fractions 2 and 3 appeared to be clear. The fractions obtained by centrifugation exhibited the following protein contents when analyzed by the method of Lowry et a2. (1951). 1003 pellet = 10.94% 1003 supernatant = 61.93% P] 1008 supernatant con- F1. taining low-density I components = 27-10% L Total, 99.97% “A Ef When the 1008 supernatant fraction was resolved into three components' over the Sepharose 6B column, their respective protein contents were as follows: Fraction 1 = 5.60% Fraction 2 = 8.50% Fraction 3 = 80.33% Total 94.43% Component Assessment bngolyacrylamide Disc Gel Electrophoresis- Polyacrylamide disc gel electrophoretic patterns of acid whey, the crude immunoglobulin preparation and its respective fractions, under both alkaline and acid conditions, are presented in Figures 3a and 3b. In the acid system, the immunoglobulins separate into distinct zones, whereas in the alkaline system, the immunoglobulins tend to form diffuse zones. Similar smearing was also observed in alkaline starch-urea gels_ (unreported data). ..IJIII . .0 . a. ..... .1 I ‘ . .._,10'01 ”I 43 Figure 3. Polyacrylamide disc gel electrophoretic patterns under A, alkaline and B, acid conditions: l-acid whey; 2-crude Ig prep; 3—1008 pellet; 4-fraction 1; 5-fraction 2; 6-fraction 3. Twenty ul of 2% solutions were applied except in alkaline gel 1 which contained 30 pl. a-la— p-Ls— IgM— IgA“C IgGl— IgG2— BSA" a-Lc— B-ts— 44 ' , .. 123456 Figure 3 45 In Figure 3a, gel No. l was slightly overloaded with acid whey, so that the immunoglobulins normally present in whey would be concentrated enough to be detectable. The photographic method used was not sensitive enough to pick up the separation of B-lactoglobulin and a-lactalbumin; however, the separation was evident in the original gel.‘ The acid gels (Figure 3b) showed more detail relative to the purity of the immuno- globulin fractions than did the alkaline gels. Acid gel No. 1 showed that the main immunoglobulin present in whey, the raw material for the crude immunoglobulin preparation was-indeed IgG. The crude immunoglobulin~ prep contained proteins corresponding to IgG, IgA, and IgM as described by Butler (1970) (personal communication 3/30/70). The 1008 pellet, acid gel No. 3, contained a component which migrated similarly to IgG. However, during the electrophoretic run, a major portion of this material‘ did not enter the gel and diffused through the buffer. This same fraction under alkaline electrOphoresis was concentrated at the top of the stacking gel as a narrow, dense band, not clearly evident in the. photograph. The void peak components, fraction 1, behaved similarly during acid electrophoresis as did the 1008 pellet. Only small amounts. of the protein entered the gel, possibly due to its large size or low concentration of protein. It, too, formed a dense narrow band on top of the stacking gel upon electrophoresis in the alkaline system. Fraction. 2, gel No. 5, contained a slight amount of IgG when observed in the acid system. However, it contained large amounts of IgM and some IgA. The proteins which formed dense bands in the stacking gel were considered to be large polymers of macroglobulin which, because of their size, could not migrate as the 198 species. Fraction 3, gel No. 6, contained IgG asxa major constituent. A small amount of IgA was present. The 46 original electrophoretogram (alkaline gel No. 6) showed a much lighter smear in the stacking gel than in the running gel, an observation not readily apparent in the photographic reproduction. Component Assessment by Immunoelectrophoresis The results of the immunoelectrOphoretic analyses of the crude immunoglobulin preparation and the separated fractions are represented by plates A-E in Figure 4. Plates A-D were developed with rabbit anti- sera to whole bovine serum and bovine IgM (plus 130) which were obtained from Dr. J. E. Butler (U.S.D.A.). Plate E was developed against the» above anti-whole bovine serum and, in the opposite trough, to a commercial anti-bovine IgG obtained from Cappel Laboratories. The crude immunoglobulin preparation, Plate A, showed characteristic precipitation arcs to IgG and IgM when develOped against antibodies to whole bovine serum and,when developed against the antibodies to IgM, a characteristic IgM arc and a slight arc to IgG were detected. The 1008 pellet, Plate B, revealed precipitin arcs when developed against both anti-sera; however, these were not characteristic of IgM, IgA, or IgG. Due to the short length of the arc and the migration of a portion of this pellet to the IgG region in acid disc electrophoresis (Figure 3b, gel No. 3), it is postulated that this arc represented a protein similar to the 198 IgG reported by Hammer at al. (1968). The first fraction_recovered from the Sepharose 63 column develOped broad bands which were difficult to reproduce photographically (Figure 4, Plate C). They appeared as straight rather broad bands, extending approximately 0.75 cm from the antigen well. The identity of these arcs is not certain and, because of characteristics of this fraction that will be discussed later, it is postulated that they may represent a lipoprotein component. Fraction 2 47 Figure 4. ImmunoelectrOphoretic patterns of isolated fractions: Plate A - crude Ig prep; B - lOOS pellet; C - fraction 1; D - fraction 2; E - fraction 3. ions: action 48 .- I anti 8 S —. anti IgM —>onti BS —.onti IgM —>onti B S —_>anti IgM anti IgM ._..anti BS —>anti BS __. anti IgG Figure 4 49 from the Sepharose 6B column exhibited a characteristic IgM arc when developed against the two antibodies. This fraction also contains IgA as apparent in the acid disc gels. However, because the antibodies used were made against serum proteins, of which secretory IgA is not a member, its characteristic arc was absent in the immunoelectrophoreto- gram. The immunoelectrophoretogram did not indicate the presence of . IgG in this fraction even though a small zone was detected in acid disc gels. The third fraction eluted from the column showed arcs character- istic of IgG. When develOped against antibodies to whole bovine serum as well as anti-IgG, the patterns were similar, and no apparent trace: 1r6° of IgA was detected; however, the acid gel data showed that this fraction contained a small quantity of IgA. Chemical Analyses The results of the various chemical analyses performed on the crude immunoglobulin preparation as well as the isolated fractions are sums marized in Table 1. Samples analyzed for chemical composition were lyophilized and stored under vacuum and over P205 prior to analysis. Therefore, the values given in Table l are based on dry weight. The values given for the per cent protein obtained by determining Kjeldahl nitrogen in the samples that contain large amounts of lipid may be misleading because of the substantial concentration of nitrogen— containing phospholipids present. The total carbohydrate contents of the samples analyzed were as follows: crude immunoglobulin preparation, 5.27%; 1003 pellet, 12.31%; fraction 1, 13.92%; fraction 2, 9.87%; fraction 3, 2.86%. The carboe hydrate data reported for fraction 3 compared very well to that reported by Nolan and Smith (1962) for bovine IgG (i.e., 2-3%) and also with the 50 Table 1. Chemical composition of immunoglobulin fractions Crude lg 1008‘ Fraction Fraction Fraction Constituent Preparation Pellet 1 2 3 Proteina 82.53% 57.10% 40.73% 67.38% 90.37% Hexose 2.30 5.30 5.70 4.80 1.30 Fucose 0.20 0.33 0.36 0.37 0.10 Sialic Acid 1.05 3.18 3.86 1.80: 0.26 Hexosamine 1.72 3.50 4.00 2.90 1.20 Lipid 7.05 29.00 73.00 8.50 0.00 Total 94.85 98.41 127i65- 85.75 93.23 aKjeldahl N x 6.25. 2.8% value reported for human IgG by Day (1966). Approximately 10% carbohydrate was found to be present in the second fraction, which con— tained principally IgM for which Rose et al. (1970) reported a value of 12.3% carbohydrate. The value found in this study could be slightly low because of a small quantity of associated lipid, as well as the' presence of small amounts of IgA and IgG which have lower carbohydrate contents than IgM. Considering the amount of lipid present in the 1008 pellet and in fraction 1, the carbohydrate values of this preparation appear to be rather high. Thus, these fractions appear to be complex high-density, lipo-glycoproteins which are unlike any other proteins known to be present in bovine milk. It would appear from the amino acid data, to be reported later, as well as from the carbohydrate analyses, that these fractions are polymers or aggregates of macroglobulins which are very stable and 51 possibly behave as micelles due to their high lipid content. The amino acid analyses of the various immunoglobulin fractions are presented in Table 2. All fractions were found to contain large amounts Table 2. Amino acid composition of immunoglobulin fractions at 1003 Pellet Fraction 1 Fraction 2 Fraction 3 E Residue (moles/1000 moles) + Lysine 59.19 59.65 57.39 57.61 Histidine 19.20 21.73 17.80 16.50 Arginine 37.98 36.30 35.77 30.93 unknown 8.24 12.32 tr. tr. Aspartic 86.90 93.05 82.73 80.15 Threonine 83.58 80.00 84.34 100.79- Serine 105.69 100.38 124.27 . 130.65- Glutamic 86.01 82.80 87.84 79.62 Proline 67.69 59.43 71.615 80.75 Glycine 81.37 71.59 71.49 78.7l~ Alanine 63.32 66.02 58.55 56.02 Half cystine 20.31 18.23 22.72 25.63 Valine 73.31 70.42 92.23 96.33 Methionine 8.86 8.27 7.27. 2.42~ Isoleucine 46.01 50.34» 38.45 27.20 Leucine 82.33 85.95 80.18 69.33 Tyrosine 26.23 27.57 30.75 39.39 Phenylalanine 46.63 55.88 36.59- 27.96 ZN 999.85 . 999.93 999.98 999.99 of serine, which is a characteristic of immunoglobulins, that distinguishes them from many other proteins (Day, 1966). The 1008 pellet and fraction 1 are quite similar in their amino acid composition, whereas fraction 2 52 resembles fraction 3 in certain cases and the 1003 pellet and fraction 1 in others. Fraction 3 is slightly different from the other fractions, with respect to higher contents of threonine and tyrosine, and lower~ contents of methionine, isoleucine, leucine, and phenylalanine. Due to the close similarities in amino acid composition of all four fractions, it can be assumed that they all most likely belong to the class of pro- teins called immunoglobulins. The analytical data indicate that all four fractions are mildly acidic proteins as determined by the ratio of the sum of acidic amino acid residues, i.e., aspartic and glutamic acids, to that of the basic amino acid residues, i.e., arginine, histidine and lysine. This ratio was 1.5 for all fractions. For comparisons, the caseins have a ratio of approximately 2.0 to 2.5. When the ratio of the hydrophobic residues phenylalanine, proline, methionine, valine, leucine and isoleucine are compared with the hydro~ philic residues aspartic, glutamic, tyrosine, lysine, arginine, and histidine, the four fractions are found to be relatively hydrophobic. The ratios for the 1008 pellet, and fractions 1, 2 and 3, are 1.04, 1.03, 1.04 and 1.00, respectively. This could possibly explain the affinity to lipid observed in the 1008 pellet and fraction 1 and. fraction 2. The 1008 pellet and fraction 1 were also observed to have.impaired solubility characteristics following ly0philization. It was for this reason, as well as to avoid possible structural and conformational damage, that the physical characterization studies were performed on unlyophilized samples. The large numbers of hydrophobic residues coupled with the lipids present could possibly account for the large "115%: 53 aggregates encountered in the centrifuge studies as well as their low solubility following lyOphilization. Putman (1959) reports that~disulfide interaction may account for the diminished solubility of human serum macroglobulins in cold and distilled water, as well as for the observa— tion that repeated precipitation or lyophilization of euglobulin tends to induce its insolubility. Lipids A qualitative study was undertaken to determine the nature of the neutral and polar lipids present in the various immunoglobulin fractions. The results of thin-layer chromatography of the lipid-containing extracts (Folch at aZ., 1957) of these samples are presented in Figure 5. Figure 5a indicates the presence of mono- and diglycerides, cholesterol, tri- “glycerides, and cholesterol esters as the neutral lipid components. Spot-6 was a cholesterol standard whereas spot 7 was a triglycerides standard composed of tributyrin. There did not appear to be.a large quantity of triglycerides present in the samples; however, the tributyrin Spot was very light, which could indicate that the triglyceride did not char properly. The 1008 pellet, fraction 1, fraction 2, and the crude immunoglobulin fraction - spots 1, 2, 3 and 5, respectively - all exhibit a similar neutral lipid composition, with apparent high concentration of cholesterol. Fraction 3, spot 4, appeared to contain a trace of cholesterol esters, present at levels too low to detect by gross.analysis. The phospholipid components present in the samples are illustrated in Figure 5b. Again spots 6 and 7 are standards for cholesterol and triglyceride (tributyrin), respectively. The phospholipid components detected in the samples were sphingomyelins, phosphatidyl choline, fi—Ft‘ ' .8” ' 1 EV _, > ”Mu-v _' . -.'- _ . ‘- , .\ laser". 1 '1, 54 Figure 5. Thin layer chromatographic patterns of A, neutral lipids (silica gel G) and B, phospholipids (silica gel H): Spot 1- 1008 pellet; 2-fraction 1; 3-fraction 2; 4-fraction 3; 5-crude Ig prep; 6-cholesterol std.; 7-tributyrin std. Abbreviations: O-origin; SPH-sphingomyelin; PC-phosphatidyl choline; PS—phosphatidyl serine; PE-phosphatidyl ethanolamine; NL-neutral lipids; PL-phospholipids; MDG—mono-diglycerides; C-cholesterol; TG-triglycerides; CE-cholesterol esters; F-solvent front. 55 Figure 5 56 phosphatidyl serine and phosphatidyl ethanolamine. Fraction 3, spot 4, did not appear to contain any phospholipids; however, the other four samples gave positive results. The phosphatidyl choline content of these samples appears to be low, while the sphingomyelin content was very high. This fact could be significant, since many membrane bound lipoproteins have been found to contain large quantities of sphingomyelin (Patton and Keenan, 1971). It also could explain the errors found in the fikIL. "j, Kjeldahl nitrogen determination of fraction 1, since sphingomyelins contribute two moles of nitrogen per mole of lipid. The lipid nitrogen. from all of the phospholipids would be expressed as Kjeldahl nitrogen, was. ! ‘ arm-m. ' ”TEST-541"» - -8.- .. I thus giving higher values for protein content than that actually present. The high lipid and carbohydrate contents of fraction 1 may also explain. the observation that it is soluble in high concentrations of trichloro- acetic acid (15%-TCA) and also appears to be stable to boiling for pro- longed periods. No precipitation or increase in opacity was apparent upon these treatments. Interestingly, this fraction was highly water soluble, unlike most globulins. Patton and Keenan (1971) report that 42% of the lipid phosphorous in milk was found in skim milk.lipoproteins, while the remaining 58% was present in the milk fat globule'membrane. Further investigation led to the conclusion that both sources of lipid phosphorous contained the same individual phospholipids in essentially the same prOportions with similar fatty acid compositions. Both contained sphingomyelin and cere- brosides in levels characteristic of those found in plasma membranes; however, their data did not support the concept that the skim milk lipo- protein arises by disintegration of the fat globule membrane. 57 Physical Analysis Sedimentation Velocity Table 3 shows corrected sggfw values for the different immmnoglobulin fractions isolated from the crude immunoglobulin preparation. A NaCl solvent system was chosen because of the prevalence of its use by workers in the field (Richardson and Kelleher, 1970). The Jenness and Koops buffer was chosen, because of its approximation to the indigenous ionic system of milk and, thus, would be useful for comparative work when studying the creaming phenomenon. As can be seen from these data, many of the fractions were quite, heterogeneous. However, in most instances one major bOundary gradient, as indicated by an asterisk, was seen to predominate. Some representative sedimentation-velocity patterns are presented in Figure 6. The protein concentration of the unlyophilized fractions was determined by the Lowry et al. (1951) method. All fractions were. analyzed at protein concentrations of 0.75% except fraction 3, which was studied at a concentration of 1.1%. The effect of sample opacity on the sedimentation patterns can be seen in Row 1, Figure 6. These. samples were highly opaque as indicated by the dark area present at the interference boundary which sedimented to the bottom of the cell as the analysis progressed. Similarly, but not to as great an extent, fraction 1 and fraction 2 exhibited some degree of Opacity. The 1008 pellet in NaCl at 20°C resolved into two widely separated boundaries in the centrifuge. However, when this fraction was suspended a second time in Tris-NaCl column buffer and recentrifuged for a 100$ 899 S20,w supernatant portion of the cell. This indicates that the-15$ species is pellet, the slower component, with an of 15.1, was observed in the 58 Table 3. The apparent sedimentation coefficients for separated immuno- globulin fractions centrifuged at 4°C and 20°C in different solvent systems values J&K Buffer 4°C b Centrifugation Corrected speed ' a Fraction (rpm) 0.15M NaCl 20°C lOOS Pellet 19,160 15.1 190.1* 1 59,780 LipOprotein* 2005* 32.8 2 c 7.1 "5671* 34.1 3 59,780 7.1* 14.7 152.3* 8.6 24.2* 2 59,780 aCorrection factor used was 0.9840. bCorrection factor used was 1.49866. c20°C at 39,460. 4°C at 59,780. * Indicates major boundary gradient. 59 Figure 6. Sedimentation velocity patterns of isolated immuno- globulin fractions: A-in 0.15 M NaCl + 0.02% sodium azide pH 6.8 at 20°C; B-in Jenness and K00ps buffer + 0.02% sodium azide pH 6.6 at? 4°C. Row l-1OOS pellet; 2-fraction 1; 3-fraction 2; 4-fraction 3. Rows 1-3 0.75% protein, Row 4—1.l% protein. 60 8'60 10 MIN 19.160 I'M 2 4MIN e=0o° IOMIN 59.700 ”M I ll L V1 4MIN 8:65 16MIN 39.460 I'M A 4 MIN 9:65‘ 16 MIN 59.700 RPM Figure 6 e:oo° 0 MIN 59.7.0 I'M 0 MIN 8:60° 40 MIN 59.700 RPM 61 loosely associated with the larger aggregated species and that the phenomenon is not one of a typical monomer-polymer dissociating system. The twice-centrifuged lOOS pellet in NaCl at 20°C is shown in Row 1, column A of Figure 6, while the once-centrifuged lOOS pellet in Jenness and Koops buffer at 4°C is shown in Row 1, column B of Figure 6. Theu larger aggregated species decreased in S value when centrifuged at 4°C in Jenness and Koops buffer. At this time no interpretation of these data is warranted, since the effects of lipids in this fraction are- unknown. The first fraction eluted from the Sepharose column when the crude immunoglobulin preparation was chromatographed appeared to be hetero- geneous. When examined at 20°C in NaCl solvent (Row 2, column A, Figure 6), this fraction contained a gradient boundary which appeared to remain near the top of the cell, approaching an equilibrium. Con— sequently, an accurate measurement could not be made for this component. From the chemical data, which indicate a high lipid content for this fraction, it appears that this gradient could be composed of high- density lipoprotein-type-material. Substantial amounts of 208 protein were also present as were_some, undetermined faster sedimenting components.. When.the same fraction was centrifuged at 4°C in Jenness and Koops buffer, the large lipOprotein gradient disappeared from the top of the cell, possibly due to an aggre- gation of this material. However, one would expect a lipoprotein to rise even faster in the cold due to the increased buoyant density of the solvent. The 208 species apparent in the 20°C run in NaCl increased to a 243 molecule at 4°C in Jenness and_Koops buffer. This behavior may have resulted from aggregation of the molecular species causing it. to sediment at a faster rate. 62 Fraction 2 (Row 3, Figure 6), was also increased in S value at 4°C as compared to 20°C in approximately the same proportion as that of fraction 1. Aggregation of the molecular species could also explain this phenomenon. The third fraction (Row 4, Figure 6) contained one major component with an Sggfiw of 7.1 in NaCl at 20°C which increased to 8.78 in Jenness and Koops buffer at 4°C. Both of these samples were water clear and differences in background contrast were due to photographic reproduction. To determine the effect of solvent on the protein samples, fraction L 2 was centrifuged at conditions in opposition to those described above. 6» 3 That is, the sample was centrifuged in NaCl solvent at 4°C and in. i Jenness and Koops buffer at 20°C. The results are presented in Figure 7 and Table 3. Both runs gave similar results, indicating that both low temperature and the ionic composition of the Jenness and Koops buffer were necessary for the aggregation which was ramified by an~ aPP increase in the 820,W' The 33 species observed in this fraction was probably the result of a slight contamination with lipoprotein. The gel.electrophoreeis data did not indicate any small molecular species that migrated faster than the 7S IgG. The 78 molecules are IgG, while the small amount of 128 material no doubt reflects the presence of IgA which was present in the acid disc gels, Figure 3. The predominant 208 species is undoubtedly IgM, whereas the small 308 boundary could possibly be a polymeric form of IgM as reported by Day (1966). Payens and Both (1970) isolated an.immunoglobulin fraction from whey by ammonium sulfate precipitation which contained both 78 and 63 Figure 7. Sedimentation velocity patterns of fraction 2: A-in Jenness and Koops buffer at 20°C; B—in 0.15 M NaCl at 4°C. 0.75% protein; 59,780 rpm. 9 - 60°; 64 Figure 7 65 macroglobulin components. The latter component possessed cryoactivity. They stated that cryoaggregation is slow and only partially reversible and that it is strongly enhanced by decreasing the ionic strength of the solution. Their cryoaggregation studies were monitored by sedimen— tation and turbidity measurements at different temperatures. A visual increase in turbidity was observed in the present study, when fraction 2 (macroglobulin fraction) was exhaustively dialyzed F? against deionized water at 4°C. However, the author attributes this i phenomenon largely to the euglobulin characteristics of IgM rather than to cryoaggregation at low ionic strength.~ The centrifugation 2 i 1 studies performed with this macroglobulin fraction showed an increase in the S value when centrifuged at 4°C in Jenness and Koops buffer. As stated previously, it was thought that this behaviOr was caused by the effects of buffer composition as well as temperature, since no noticeable change in the S value occurred when the protein was centri- fuged in cold NaCl solution. All of the Sigfiw values presented herein for the various immuno- globulins correlate well with values given in the literature (Rose at al., 1970). It should be mentioned, however, that the S values for these proteins are concentration dependent, thus one would expect a 20,w at slight change from the apparent S value when expressed as S infinite dilution. A review of the dairy science literature revealed that there is. little indication that lipOproteins are a part of Smith's (1946) classical immunoglobulin preparations, or that there are higher polymeric forms of immunoglobulins with S values greater than 20. One exception to this was found in the work of Payens (1968), who encountered a large aggregate 66 in the void—volume elution peak while separating an immunoglobulin. preparation (Smith, 1946) on‘a Sepharose 4B (exclusion 20 x 106) column. It was later confirmed in a personal communication with Dr. Payens, 1971, that this fraction was highly Opaque and, thus, would seem to correspond to the 1008 pellet fraction of the present study. It is interesting to consider the work of Phelps and_Cann.(l957) relating to the modification of bovine y-peeudoglobulin by acid.‘ They found that in nearly neutral solutions the-sedimentation behavior was independent of salt concentration over the range of 0.02 to 1.0 M NaCl but that the sedimentation behavior was strongly dependent on salt concentration at pH 3.1. In acidic solutions, the protein sedimented as a single boundary with a decreasing sedimentation rate, i.e., 5.7 to 4.48, as the ionic strength was varied from 0.1 to 0.02 respectively. At high salt concentrations (0.3 r/2), sedimentation patterns exhibited three boundaries: 6.38, 9.58 and 128 components. The more rapidly sedimenting components were attributed to aggregation of the basic protein species. They also showed that the pH effects_were not com- pletely reversible. After a one hour exposure to pH 3.1, ionic strength 0.1, followed by dialysis against pH 7.0 phosphate buffer and then 0.1 M NaCl, the protein solution had a bluish hue and yielded a precipitate when dialyzed against distilled water. This behavior was unique for a pseudoglobulin. The re-neutralized material was found to contain sedimenting components of 6.698, 9S, 138, 248, and 668. The water-soluble fraction was found to contain only the 6.518 and 118 molecular species. It is not possible to assess the effects of acid on the proteins isolated in the crude immunoglobulin preparation in the present study; 67 however, the above authors stated that y-globulin sediments at a rate independent of pH over the range of 7.4 to 4.2.- Since the lowest pH used in the preparative procedure was 4.5, the effect, if any, would probably be slight. Creaming_8tudies f The Effects of Temperature on Creaming Prior to determining the effects of the isolated immunoglobulin E fractions on creaming, an experiment was conducted with raw milk to determine what effect temperature has on the creaming phenomenon. Raw i milk samples were observed for cream volume at three different.tempera- tures. All samples were tempered at 40°C for 30 min prior to quiescent storage. The temperatures selected for storage were 4, 22, and 40°C, and the results are shown in Table 4. Table 4. The effects of temperature on the creaming of raw milk samples Cream Volumea Temperature 1 hr 4 hr 24 hr 4°C 3.20 1.70 1.30 22°C 0.00 0.20 0.40 40°C 0.00 0.10 0.30 aExpressed in ml per 10 m1 total volume. The data in Table 4 indicate that a low temperature is prerequisite for normal clustering and creamline formation. Similar findings were also reported by Dunkley and Sommer (1944). Payens et al. (1965) showed that 68 the adsorption of euglobulin to the fat globule membrane was temperature dependent, with a considerable decrease at 45°C and a much greater adsorp- tion at 5° and 10°C. Their findings agree with the above results. The Effects of the Isolated Immunoglobulin Fractions on Creaminggin a Model System A model system was used which substituted Jenness and Koops buffer for the aqueous phase of milk and cream washed with the same buffer for the non-aqueous phase. The immunoglobulin fractions that were incorpor- ated into the system were dialyzed against Jenness and Koops (1962) buffer. However, the 1008 pellet material was not removed from.the crude immunoglobulin preparation and, consequently, was present in fraction 1 as collected from the Sepharose 6B column. The data obtained from this experiment are presented in Table 5. Table 5. The effects of immunoglobulin fractions on a model creaming system of Jenness and K00ps buffer plus cream Protein a Concentration Cream Volume at 4°C Fraction mg/25 mls 1 hr 4 hr 24 hr Control 0 0.10 0.10 0.15 Fraction 1b 2 0.10 0.20 0.60 Fraction 2 2 0.10 0.40 0.85 Fraction 3 14- 0.10 0.20_ 0.20 Crude Ig Prep 20 0.10 0.30 0.60 aExpressed in ml per 10 ml total volume. bIncludes 1008 pellet. 69 The results of this experiment indicate-that fraction 2 had the greatest effect on creaming which, however, did not result in a typical creaming. The creaming observed in this model system seemed to be more. typical of a gravity rise of small fat globules than the formation of large clusters, since the cream volumes tended to increase with time. Normal creaming exhibits a_reduction in cream volume with time due to the compaction of the large clusters. 3% Because of the atypical results obtained with the above model system, an alternate system was devised which gave creaming results that agreed more closely with the classical work of Dunkley and Sommer (1944). unlrm- In this system, the natural clustering agent inherent to separated skim milk was heat inactivated, thereby rendering an almost perfect creaming media since the natural ionic and compositional creaming environment. is left intact. The Effects of Heating on.a Recombined Creaming System The results of creaming experiments showing the effects of heating the separated cream and skim milk to 70°C for 20 min are shown in. Table 6. The raw whole milk sample contained 3.6% fat, whereas the recombined samples were adjusted to 4% fat according to Dunkley and Sommer (1944). The data in Table 6 indicate that the factor_responsible for creaming is destroyed by heating at 70°C for twenty min and that most of this factor is located in the skim milk when the milk is separated at 40°C. This conclusion is evidenced by the low cream.volumes observed when heated skim milk was recombined with both raw cream and heated* cream. These results are atypical of normal creaming where a large cream layer is initially observed which with time compacts to a smaller 70 Table 6. The effects of heat on the creaming factor in skim milk Cream Volume at 4°C3 Recombined Samples 1 hr 4 hr 24 hrs Raw whole milk (3.6% fat) 2.00 1.30 1.10 Raw cream and raw skim 1.50 1.60 1.45~ Raw cream and heated skim 0.00 0.10 0.30 HA Heated cream and raw skim 1.15 1.35 1.30 Heated cream and heated skim 0.00 0.10 0.20 g aExpressed in ml per 10 m1 total volume. ' - volume as the fat clusters rise. Closer to normal results were found when raw skim was combined with either heated or raw cream. The larger cream volumes after 24 hours found with the raw skim samples as compared to raw milk are most likely due to differences in fat content between the two samples. The Effects of the Isolated Immungglobulin Fractions on a Heated Recombined Creaming System From the data in Table 6, it was concluded that the best system for determining the effects of the isolated immunoglobulin fractions on creaming would be one which incorporated raw cream, heated skim milk, and immunoglobulin which was unlyophilized and dialyzed against Jenness and Koops (1962) buffer. The immunoglobulin fractions were incorporated into the system at concentrations approximating that found in normal skim milk (Rose et aZ., 1970). Thus, for the four different immunoglobulin fractions tested, the 1008 pellet, fraction 1, and fraction 2 were added 71 at a rate of 2.0 mg per 25 ml creaming sample volume. Because IgG is the most abundant immunoglobulin present in skim milk, fraction 3 was added at a.rate of 14.0 mg to the same volume. The 2.0 mg concentration used for the first three fractions corresponded to the finding by.Payens at al. (1965), who stated that creaming became constant when 2 mg of euglobulin per gram of fat was present. In some cases, greater than 2 mg quantities of immunoglobulins were added. However, creaming volumes similar to those obtained with the 2 mg level were observed. The effect on creaming of the isolated immunoglobulin fractions are presented in‘ Table 7. Table 7. The effects of isolated immunoglobulin fractions on the raw cream and heated skim model system Cream Volume at 4°C8 Fraction 1 hr 4 hr 24 hrs 1008 pellet . 0.00 0.10 0.30 Fraction 1 0.00 0.00 0.60 Fraction 2 0.00 1.10 1.20 Fraction 3 0.00 0.00 0.207 8Expressed in ml per 10 ml volume. The data in Table 7 indicate that fraction 2 contained the active creaming factor and yielded cream volumes approaching those normally found when raw cream and raw skim milk are combined. This fraction con- tained predominantly IgM as monitored by acid disc-gel electrophoresis, immunoelectrophoresis and analytical ultracentrifugation. Fraction 3, 72 which contained predominantly IgG and a small amount of IgA, as well as the 1008 pellet material, had no effect on creaming. Fraction 1 exhibited a slight effect on creaming, possibly due to the presence of a 208 component which was apparent in the sedimentation patterns. To assess the effects of a smaller quantity of fraction 2 on cream- ing, an experiment was performed using one mg instead of 2 mg of this fraction. Heated skim milk and raw cream samples were mixed with this P3 fraction and, after 1, 4 and 24 hr at 4°C, cream volumes were 0.0, 0.55, and 1.05 ml, respectively. The reduced cream volumes observed indicate that the quantity of active creaming agent present in a_given i 1 system ista limiting factor in the creaming process. An attempt was made to duplicate the work of Gammack and Gupta. (1970). They did not state the conditions of centrifugation employed for the sedimentation of the casein pellet; thus an assumptioanas made which might not have corresponded with their conditions. ~The heated skim milk used in the above creaming experiments was centrifuged in the Beckman/Spinco, Model L-65 centrifuge at 20°C at conditions. designed to produce a 508 pellet. Following centrifugation, the tubes contained a white casein pellet, a clear yellow whey supernatant, and a small cream layer at the top. The cream layer was removed, followed by the removal of the whey fraction. The top of the casein pellet was then carefully scraped out with a spatula. It was assumed that this fraction contained the active lipoprotein fraction described by Gammack and Gupta (1970). It was opalescent as described. The casein pellet was then dissolved in Jenness and KoOps buffer and stored for subsequent creaming studies. 73 The above four fractions were all incorporated into model systems for creaming studies; the centrifuged middle whey fraction was used as the aqueous phase. For example, the casein pellet plus raw cream plus fraction 2 were mixed to determine the effects of the casein fraction on creaming. Similar mixtures using the other fractions were also examined.‘ The results of this experiment are presented in Table 8. Table 8. The effects of incorporating centrifuged heated skim milk fractions into a creaming system of centrifuged whey, raw cream, and 2 mg fraction 2/25 m1 of 3.2 .. _' ... g, Quantity added Cream Volume at 4°C8 Sample in grams 1 hr 4 hr 24 hrs Control 0.0 0.00 0.10 0.50 Centrifuged whey containing upper lipid layer 5.0 0.00 0.10 0.40 LipOprotein layer above casein pellet 5.0 0.00 0.10 0.60 508 casein pellet 5.0 0.00 0.20 0.80 8Expressed in ml per 10 ml volume. The results observed in this experiment do not correspond to those found by Gammack and Gupta (1970). They are also difficult to explain since none of the centrifuged fractions appeared to give normal creaming results as expected for non—centrifuged, heated skim milk. One explana- tion for atypical behavior is that proportions of centrifuged fractions added to the system were not.correct. However, this seems unlikely, because the authors described the lipoprotein fraction above the casein 74 layer as being very active in the presence of macroglobulinq Another variable which existed between the two different experiments should be mentioned. In this study, heated skim milk was used as the source of. lipoprotein, whereas Gammack and Gupta used raw skim milk. Possibly, heating the skim milk could have altered or denatured the lipoproteins or caused the lipoprotein to complex with the casein micelles, thus rendering them unavailable to the creaming phenomenon. It is also interesting to consider the earlier findings of Hansson (1949), who found that the addition of lecithin and cephalin prepared from cow-brains greatly increased creaming in raw or low-pasteurized milk but not in high-pasteurized milk. This observation correlates with the recent findings.of the above authors concerning the creaming activity of the high density lipoprotein. The results of the present experiments seem to be comparable in certain respects (atypical creaming) with the data obtained when Jenness and Koops buffer was used as the aqueous phase, Table-5. It would. therefore seem to indicate that the intact skim milk system.is necessary before normal creaming occurs, and would substantiate the finding of the above authors that something other than IgM be present before normal creaming occurs. The Effects of Neuraminidase on Creaming. Neuraminidase is an enzyme which cleaves terminally bound N-acetyl neuraminic acid (NANA) from glchproteins which contain this carbohydrate. The objective of this experiment was to determine if NANA played a role in creaming. Ten mg of.the fraction 2 protein in Jenness and K00ps (1962) buffer was incubated with 1 mg synthetic neuraminidase at 40°C- for 2 hr. An aliquot of the resulting mixture was then added to the . A a‘.‘ iI-n’.{" .1”, 75 normal creaming system and the results are given in Table 9. As can be Table 9. Creaming as affected by neuraminidase treatment of fraction 2, in a system of raw cream plus heated skim Cream Volume at 4°C8 Sample 1 hr} * 4 hr‘ 24 hrs Fraction 2 plus Neuraminidase 2.00 1.65 1.35 Fraction 2 2.00 1.75 1.45. 3Expressed in ml per 10 ml volume. seen from the data.in Table 9, neuraminidase treatment did not affect the creaming ability of the fraction 2 protein. This does not mean that the carbohydrates present on the proteins do not contribute to their. activity in cluster formation. In order to ascertain that the enzyme was active, a control hydrolysis was performed which indicated that sialic acid was liberated. The Effect of Disulfide Reducing_égent on Creaming The objective of this experiment was to determine the effect of 2-mercaptoethanol (ME) on three different creaming systems. The first system was comprised of whole raw milk which was incubated with 0.5% and 2.5% MB for 1 hr at 40°C. The second system contained heated skim milk, raw cream and fraction 2 and was.treated in a similar manner. The third system differed from the others in that fraction 2 in Jenness and Koops (1962) buffer was incubated as above with 0.5% and 2.5% MB which served to reduce disulfide bonds. The sulfhydryl groups were then alkylated by dialyzing for 24 hr at 20°C against an excess of. WW Jammy-g l a"; 76 iodoacetamide (IAc) in Jenness and Koops (1962) buffer. The excess IAc was then removed by further dialysis against Jenness and Koops (1962) buffer for 24 hr at 20°C. The-reduced and alkylated fraction 2 protein was then incorporated into a heated skim milk-raw cream model system. The effects of similar concentrations of IAc on fraction 2 were also determined by omitting the presence of ME in the third system, serving to alkylate any free sulfhydryl inherent to this fraction. The results of ME treatment on the above creaming systems are presented in, Table 10. Table 10. Creaming as affected by incorporating mercaptoethanol into raw milk; heated skim model system containing fraction 2; and fraction 2 followed by alkylation and incoporation into the heated skim model system Cream Volume at 4°C8 Sample 1 hr 4 hr 24 hrs Raw milk 0.00 3.20 1.15- Raw milk 0.00 3.50 1.20 Raw milk 0.00 0.00 0.10 Fraction 2b 0.00 1.30 1.15 Fraction 2b 0.00 0.00 0.15 Fraction 2C 0.00 0.60. 0.80 Fraction 2c 0.00 0.00 0.10 Fraction 2d 0.00 1.50 1.30 8Expressed in ml per 10 m1 volume. b Fraction 2 in heated skim-raw cream.system.a CTreated with ME and IAc followed by incorporation into heated skim milk model system. d Treated with IAc followed by incorporation into heated skim model system. 77 The results presented in Table 10 indicate that a mild exposure to mercaptoethanol (0.5%) has no effect on the creaming ability of the raw milk or heated skim milk creaming systems. There was a loss in.effective- ness, however, when the fraction 2 protein was subjected to disulfide reduction followed by alkylation. This is probably due to the fewer numberg'f disulfide bonds present in fraction 2 as compared to the first two systems where the large number of disulfides probably diluted the reducing capacity of this reagent. Alkylation alone of fraction 2 had no detrimental effect on its creaming ability. Therefore, the reduction of cream volume observed when fraction 2 was reduced and alkylated was due to the activity of mercaptoethanol. The effect of a five-fold increase in mercaptOethanol concentra- tion (2.5%) was totally destructive to the creaming capacity of all systems investigated. It was then of interest to determine what effect the reduction and alkylation treatments had on the molecular structure of the IgM present in fraction 2. Possible alterations in molecular structure were elucidated by acid disc-gel electrophoresis. When the proteins in fraction 2 were reduced in the presence of 0.5% ME and followed by alkylation, the gels indicated the appearance of a band in the running gel approximating the migration of IgG and a corresponding disappearance in the concentration of the IgM zone. When the 2.5% reducing system was observed, the gels showed no band for IgM and all protein migrated as the mbnomer form which approximated the mobility of IgG. The findings of the above experiments indicate that the molecular size of IgM is of paramount importance with respect to its function in promoting creaming. The present study therefore considers the apparent 78 cryo-aggregation of IgM from a 208 molecule to a 258 molecule in the presence of a simulated milk salt environment (Jenness and Koops [1962] buffer) to be a prerequisite to cluster formation and creaming. w ‘1‘.“‘131M'Ktveu. 11 SUMMARY Whey prepared from raw skim milk, which had been acidified to remove the casein, was fractionated by the addition of ammonium sulfate to yield a crude immunoglobulin preparation. This preparation was further fractionated by preparative ultracentrifugation and gel filtration chroma- tography on a Sepharose 6B column into a 1008 pellet and fraction 1, 2, ‘15? and 3, respectively. The components present in the fractions were analyzed by polyacrylamide disc gel electrophoresis, immunoelectrophoresis, analytical ultracentrifugation as well as various chemical procedures.’ The 1008 pellet was observed to be very opaque, and contained approximately 29% lipid and 12% carbohydrate. It also appeared to be loosely associated with a 158 species of IgG as evidenced by sedimenta- tion velocity determinations and by acid—disc gel electrophoresis. The major component of this fraction did not enter a 3.75% spacer gel on electrophoresis and was found to sediment as a 1908 molecular species in 0.15M NaCl and as a 1508 species in Jenness and Koops (1962) simulated milk salt buffer at 4°C. The eluate corresponding to the void peak of Sepharose,6B column~ chromatography, fraction 1, also exhibited some opacity. It contained approximately 73% lipid and 14% carbohydrate and was observed to be very heterogeneous in sedimentation velocity determinations. A majority of the-components in this fraction, other than the lipoprotein gradient, were found to be in the range of 20-708. Only a minor portion of the 79 80 material was found to enter a 3.75% spacer gel in polyacrylamide disc gel electrOphoresis, and unusual precipitin arcs were observed when submitted to analysis by immunoelectrophoresis.- From these physico- chemical data, these two fractions appeared to be complex high-density lipoglycoproteins. Fraction 2 contained approximately 8.5% lipid and 10% carbohydrate and contained principally IgM when analyzed by polyacrylamide acid disc gel electrophoresis and immunoelectrOphoresis. The sedimentation velocity patterns for this fraction indicated that the major component. present was a 208 species which aggregated to a 258 species in 4°C Jenness and Koops (1962) simulated milk salt buffer. Fraction 3 contained 2.9% carbohydrate and was not.found.to be associated with lipid. Immunoelectrophoresis indicated that this frac- tion consisted of IgG. However, polyacrylamide acid disc gel.electro- phoresis showed a minor quantity of IgA. The sedimentation velocity studies revealed that a 7.18 species predominated in ambient NaCl solvent which aggregated to an 8.78 species in 4°C Jenness and Koops (1962) simulated milk salt buffer. Amino acid analyses indicated that all fractions studied contained serine as the major amino acid constituent. The proteins were mildly acidic and contained more hydrOphobic than hydrophilic residues. The 1008 pellet and fraction 1 contained 29% and 73% lipid,- respectively. A qualitative lipid analysis of these high lipid- containing fractions indicated that the neutral lipids consisted mainly of cholesterol, while the principal phospholipid present was sphingo- myelin. Fraction 2, which contained 8.5% lipid, had a similar distri- bution of lipid components. 81 The results of creaming experiments, utilizing raw milk, indicate that a low temperature, i.e., 4°C, is prerequisite for normal clustering and creamline formation. When Jenness and K00ps (1962) simulated milk salt buffer was substituted for the skim milk aqueous phase, normal . creaming did not result. Instead, a gradual increase in cream volume with time resulted, rather than a cream volume reduction which normally occurs with time. When a heat-inactivated creaming system was used, into which the- unheated immunoglobulin fractions were incorporated, normal creaming resulted only in the case of fraction 2. This fraction was shown to contain mainly IgM. The result of the enzymatic removal of terminally bound sialic acid from fraction 2 by the action of neuraminidase had no effect on its creaming ability. Following partial reduction of disulfide bonds with 0.5% 2-mercapto- ethanol and alkylation with iodoacetamide, a decrease in creaming ability was observed. In the presence of 2.5% 2—mercaptoethano1, the creaming ability of fraction 2 was totally destroyed. This-was attributed to the IgM being reduced to its monomeric species. It was concluded from the above observations that IgM cryoaggree gates from a 208 to a 258 molecule in the presence of a cold milk salt. ionic environment. This aggregation is postulated as being prerequisite to fat globule clustering and creaming. The size characteristics of IgM are therefore considered of fundamental importance in the creaming phenomenon. 3! m .0. thPF-i‘lfi .-_ “I BIBLIOGRAPHY BIBLIOGRAPHY Babcock, S. M. 1889. The constitution of milk, and some of the con— ditions which affect the separation of cream. Wis. Agr. Expt. Sta. Bul. 18. (Original not seen. Cited by Dunkley, W. L. and Sommer, H. H. 1944. Wis. Agr. Expt. Sta. Res. Bull., No. 151). Bull. Wbrld Health Organ. 1964. Nomenclature for human immunoglobulins. 30:447. Butler, J. E. 1969. Bovine immunoglobulins: A review. J. Dairy Sci., 52:1895. Carpenter, P. L. 1965. Immunology and Serology. Serum Proteins, Ch. 4. W. B. Saunders Company, Philadelphia. Crowther, C., and Raistrick, H. 1916. A comparative study of the pro- teins of the colostrum and milk of the cow and their relation to the serum proteins. Biochem. J., 193434. Day, E. D. 1966. Fbundations of’Immunochemistry. The Williams & Wilkins Company/Baltimore. Dische, 2., and Shettles, L. B. 1948. A specific color reaction of methylpentoses and a spectrOphotometric micromethod for their determination. J. Biol. Chem., 175:595. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., and Smith, F. 1956. Colorimetric method for the determination,of sugars and related substances. Anal. Chem., 28:350.' Dunkley, W. L., and Sommer, H. H. 1944- The creaming of milk. Wis. Agr. Expt. Sta. Res. Bull., No. 151. Folch, J., Lees, M., and Sloane-Stanley,-G. H. 1957. A simple.method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem., 226:497. Franzen, R. W., Jr. 1967. The chemical composition of broiler neck tissues and the effects of phospholipids on emulsions. M.S. Thesis, University of Georgia, Athens, Georgia. Gammack, D. B., and Gupta, B. B. 1970. Clustering of milk fat globules. XVIII Intern. Dairy Congr., Proc., Melbourne, 1§:20. 82 83 Grabar, P., and Williams, C. A., 1953. Methode permettant L'etude conjuguée des propriétés electrophoretiques et immunochemiques d'um melange de protéines. Application au serum sanguin. Biochim. Biophys. Acta, 19:193. Greenberg, D. M.- 1951. Amino Acids and Proteins. pp. 338. Charles C. Thomas, Springfield, Ill. Groves, M. L., and Gordon, W. G. 1967. Isolation of a new glycoprotein—a and a yG-globulin from individual cow milks. Biochemistry, 952388. Hammer, D. K., Kickhofen, B., and Henning, G.' 1968. Mblecular classes and properties of antibodies in cattle serum and colostrum synthe- sized during the primary and secondary response to protein anti- gens. European J. Biochem., 9:443. Hansson, E. 1949. The creaming of milk. XIIth. Intern. Dairy Congr., 2:30. Wi- ..-.‘- _ Howe, P. E. 1921. The use of sodium sulfate as the globulin precipi- tant in the determination of proteins in blood. J. Biol. Chem., 49:93. Howe, P. E. 1922. The differential precipitation of the proteins of colostrum and a method for the determination of the proteins in colostrum. J. Biol. Chem., 99:51. International, Critical, Tables of; Numerical Data, Physics, Chemistry and Technology. 1928, 1929. V01. III and Vol. V. Publ. for the National Research Council by McGraw-Hill Book Co., Inc., New York. Jenness, R., and Koops, J. 1962. Preparation and properties of a salt solution which simulates milk ultrafiltrate. Neth. Milk Dairy J., 16:153. Jenness, R., Anderson, R. K., and Gough, P. M. 1965. Fractionation of bovine milk and blood agglutinins for "Brucella" by gel filtration and specific adsorption techniques. Federation Proc., 995503. Johansen, P. C., Marshall, R. D., and Neuberger, A. 1960. Carbohydrates in protein. The hexose, hexosamine, acetyl and amid-nitrogen content of hen's-egg albumin. Biochem. J., 11:239. Kenyon, A. J., and Jenness, R. 1958. Effect of proteins on agglutina- tion of fat globules. J. Dairy Sci., 993716.. Koops, J., Payens, T. A. J., and Kerkhof Mogot, M. F. 1966. The effect of homogenization on the spontaneous creaming of milk. Neth. Milk Dairy J., 995296. Lowry, O. H., Rosebrough, N. J., Farr, L.A., and Randall, R. J. 1951. Protein measurement with the Folin Phenol reagent. J. Biol. Chem., 193:265. 84 Marrack,J. R. 1938. The chemistry of antigens and antibodies. Gr. Brit. Med. Res. Council Spec. Rep. Ser. 230. (Original not seen. Cited by Dunkley, W. L., and Sommer, H. H. 1944. Wis. Agr. Expt. Sta. Res. Bull., No. 151). Melachouris, N. P. 1969. Discontinuous gel electrohporeis of whey, casein, and clotting enzymes.’ J. Dairy Sci., 29:456. Moore, 8., Spackman, D. H., and Stein, W. H. 1958. Chromatography of amino acids on sulfonated polystyrene resins. Anal. Chem., 99: 1185. Murphy, F. A., Aalund, 0., and Osebold, J. W. 1964. Physical hetero- geneity of bovine gamma globulins: Gamma-1M globulin electrophoretic {i heterogeneity. Proc. Soc. Exptl. Biol. Med., 117: 513. 3 Murphy, F. A., Aalund, 0., Osebold, J. W., and Carroll, E. J. 1964. Gamma globulins of bovine lacteal secretions. Arch. Biochem. , Biophys., 108:230. i' Murphy, F. A., Osebold, J. W., and Aalund, O. 1965. Physical hetero- geneity of bovine y-globulins: Characterization of yM and yG globulins.‘ Arch. Biochem. BiOphys., 112:126. Nolan, C., and Smith, E. L. 1962. Glycopeptides III. Isolation and prOperties of glycopeptides from a bovine globulin of colostrum and from fraction II-3 of human globulin. J. Biol. Chem., 237:453. Patton, 8., and Keenan, T. W.- 1971. The relationship of milk phospho- lipids to membranes of the secretory cell. Lipids, 9358. Payens, T. A. J.- 1968. Neuere forschungen aug dem gebiet der milcheiweisse. Milchwissenschaft, 993325. Payens, T. A. J., and Both, P. 1970. Cryoglobulins from milk and a mechanism for the cold agglutination of milk-fat globules. Immunochemistry,_z:869. Payens, T. A. J., Koops, J., and Kerkhof Mogot, M. F. 1965. Adsorption of euglobulin on agglutinating milk fat globules. Biochim. Biophys. Acta, 993576. Phelps, R. A., and Cann, J. R. 1957. On the modification of y-globulin by acid. Biochim. Biophys. Acta,_93:149. Porter, P. 1969. Transfer of immunoglobulins IgG, IgA, and IgM to lacteal secretions in the parturient sow and their absorption by the neonatal piglet. Biochim. Biophys. Acta, 181:381. Putnam, F. W. 1959. Abnormal human serum globulins. I. The nonidentity. of macroglobulins. Arch. Biochem. Biophys.,.19:67. Richardson, A. K., and Kelleher, P. C. 1970. The ll-S sow colostral immunoglobulin isolation and physicochemical properties.. Biochim. Biophys. Acta, 214:117. Rose, 85 D., Brunner, J. R., Kalan, E. B., Larson, B. L., Melnychyn, P., Swaisgood, H. E., and Waugh, D. F. 1970. Nomenclature of the Proteins of Cow's Milk: Third revision. J. Dairy Sci., 93:1. Samuelsson, E., Bengtsson, G., Nilsson, 8., and Mattsson, N. 1954. Cream rising. Svenska Mejeritidn. 46:(l3) 163; 168; 173; 179; (14) 193; 203; 209. (Original not seen. Cited from Dairy Sci. Abstr., 9931015 [1954]). Schachman, H. K. 1957. Ultracentrifugation, Diffusion, and Viscometry, ‘pp. 32-103. Vol. IV. ln_Colowick, S. P., and Kaplan, N. 0., (ed.) Methods in Enzymology. Atademic Press, New York. Sharp, P. F., and Krukovsky, V. N. 1939. Differences in adsorption of Smith, Smith, Smith, solid and liquid fat globules as influencing the surface tension and creaming of milk. J. Dairy Sci., 993743. E. L. 1946. Isolation and properties of immune lactoglobulins from bovine whey. J. Biol. Chem., 165:665. E. L., Green, R. D., and Bartner, E. 1946. Amino acid and carbo- hydrate analysis of some immune proteins. J. Biol. Chem., 146: 359. E. L., and Holm, A. 1948. The transfer of immunity to the new- born calf from colostrum. J. Biol. Chem., 175:349. Stadhouders, J., and Hup, G. 1970. Complexity and specificity of euglobulin in relation to inhibition of bacteria and to cream. rising. Neth. Milk Dairy J., 99379. Svennerholm, L. 1957. Quantitative Estimation of Sialic Acids. II. Troy, A colorimetric Resorcinol-Hydrochloric Acid Method. Biochim. Biophys. Acta, 99:604. H. C., and Sharp, P. F. 1928. Physical factors affecting the formation and fat content of gravity cream. J.-Dairy Sci.,. 11:189. Warren, L. 1959. The thiobarbituric acid assay of sialic acids.. J. Biol. Chem., 234:1971. W4 :4: i3? APPENDIX ' rears? ‘17—“?- APPENDIX Compgsition of Jenness and Koops' (1962) Simulated Milk Salt Buffer Solution 1 The following gram quantities of salts were made to 9.0 1 with deionized water: KHZPO4 15.80 K3citrate-H20 5.08 Na3citrate-2H20 17.90 K2804 1.80 Mg3citrate'H20 5.02 KZ'CO3 3.00 KCl 10.78 Solution 11 A 13.20 g of CaClz-ZHZO was dissolved in 500 ml deionized water. Solution II was slowly added to Solution I while stirring the mixture vigorously. Two grams of sodium azide was then added and the pH was adjusted to 6.6 with l N KOH.‘ The final volume was adjusted to 10 1 with deionized water. Electrophoretic Mobility The electrophoretic mobilities of the various isolated fractions,were determined in veronal buffer at pH 8.6, ionic strength 0.1. The buffer was composed of 5.6 g veronal plus 41.2 g sodium veronal made to 2 l with deionized water. The mobilities were calculated from the descending V channels and are expressed as ElectrOphoretic mobility - X10‘5 cm? volts"1 sec' . 86 a. - .cnlh ~' 4 .1. Whm 87 Two boundaries were observed when the crude immunoglobulin_prepara— tion was analyzed, a minor leading boundary migrating at.a rate of -7.15, and a major boundary with a rate of -3.68. When the void peak from Sepharose 6B, which had not been centrifuged to remove the 1008 pellet, was analyzed, one boundary was observed with a mobility of -7.6. Fraction 2 exhibited two migrating boundaries with mobilities of -6.5 and -2.75. Fraction 3 appeared as a single migrating boundary with a mobility of -3.60. The crude immunoglobulin prep and fraction 3 were analyzed at 1% protein concentration, and the fraction 1 and fraction 2 at 0.5% protein.concentration. "Illiiiii‘liiliflI