‘— w-u— v—A - ‘ ELECTROPHORETiC AND [MWNOELECTROFHQET EC ANALYSES 6F SERUMS mom RGfiMAL CQWS AND (2ng EXPEREMERTALLY AND NATURALLY INFECTED WITH MYCQBACTERIA Th3“: {’09 EM Down cf pk. D. MECHEGAN STATE UNWERSITY George Leonard Wright, Jr. 1.966 This is to certify that the thesis entitled Electrophoretic and lmmunoelectrOphoretic Analyses of Serums from Normal Cows and Cows Experimentally and Naturally Infected with Mycobacteria presented by George Leonard Wright, Jr. has been accepted towards fulfillment of the requirements for Ph.D degreein Microbiology 4 ~;}'l\fim6m\§\mm\\\ thor professor 0-169 LIBRAR Y Michigan State Ill? I {ll lllllllllt ABSTRACT ELECTROPHORETIC AND IMMUNOELECTROPHORETIC ANALYSES OF SERUMS FROM NORMAL COWS AND COWS EXPERIMENTALLY AND NATURALLY INFECTED WITH MYCOBACTERIA by George Leonard Wright, Jr. Apparatus and techniques for starch-gel electrophore- sis and for the rapid staining and destaining of both starch and disc gels were developed. This facilitated the mechan- ical manipulation of starch-gels during the preparation and slicing procedures and eliminated handling of gels during the staining and destaining processes. Serums from normal and tuberculous cattle were analyzed by various electrophoretic procedures. A modified polyacrylamide gel electrophoretic procedure, tentatively named differential disc electrophoresis, was the most sensi- tive single procedure. Thirty to thirty-eight fractions were detected in normal bovine serum by this technique. A nomenclature was proposed for the fractions separated by cellulose acetate, electrophoresis, starch-gel electrophore- sis, and immunoelectrophoresis. .A hypergammaglobulinemia and a decrease in albumin were detected by all electrophoretic procedures in the serums from calves inoculated with g, bovis, virulent George Leonard Wright, Jr. Group III's and serums collected from naturally infected cows prior to tuberculin testing. No changes were detected in the electrophorograms of serums from calves inoculated with M. avium or avirulent Group III, or from tuberculin negative cows. Changes in the albumin/gamma globulin (A/G) ratio calculated from cellulose acetate electrophorograms were differential for cattle inoculated with virulent mycobacteria (except one virulent Group III, strain 51C). Starch-gel, differential disc electrophoresis, and immunoelectrophoresis detected an increase in an alpha2 globulin fraction in serums 7 to 28 days after injection of virulent mycobacteria (including strain 51C) and in serums collected prior to tuberculin tests from tuberculin-positive gross-lesion and no-gross-lesion cows. This fraction was unchanged in serums from calves inoculated with g, avium and avirulent Group III mycobacteria. The fraction has been tentatively named the alpha -V (virulence) fraction and was 2 most reliably detected by starch-gel electrophoresis. ELECTROPHORETIC AND IMMUNOELECTROPHORETIC ANALYSES OF SERUMS FROM NORMAL COWS AND COWS EXPERIMENTALLY AND NATURALLY INFECTED WITH MYCOBACTERIA BY George Leonard Wright, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1966 PREFACE Tuberculosis occupies a peculiar place in the history of medicine. By the end of the nineteenth century, more was known of its clinical manifesta- tions and pathological characteristics than was the case for any other important infectious dis- ase. Yet, the genesis of its lesions and the factors which influence their course constitute today one of the most obscure chapters of pathol- ogy. Similarly, while the tubercle bacillus was one of the first pathogenic microorganisms to be isolated and extensively studied both in_vivo and ig vitro, knowledge of its pathogenic determinants has hardly advanced during the past two decades. Factual information concerning its metabolism and composition, its immunological behavior, and the properties responsible for its virulence is so deficient that it contributes but little to the understanding of the disease. In other words, the study of tuberculosis, which was in the fore- _front of medical sciences fifty years ago, now lags several decades behind that of many other human infections. Rene J. Dubos (52) The paradox of tuberculosis is also apparent from the attempt to eradicate bovine tuberculosis. Enough was known about the mode of transmission, and methods of detec- tion such as the tuberculin test that the incidence of the disease was reduced in the United States from 4.5 percent in 1917 to 0.2 percent in 1940. .Since 1940 the incidence has decreased only 0.1 percent. The disease continues to exist at a surprisingly high incidence in geographic pockets in spite of the eradication program. Why does the organism ii persist in animals in these areas where the eradication program is as vigorously pursued as in areas where the dis- ease has essentially been eradicated? A logical answer to this provocative question has been made by Dr. W. L. Mallmann (119), Department of Microbiology. Michigan State University: It is apparent that our knowledge of bovine tuberculosis is inadequate. Otherwise the pro- cedures in use to detect and eliminate bovine tuberculosis would have continued to reduce the incidence and the disease should have disappeared. Certainly the uses we have made of the procedures for eradication were good. Thus, the faults lie in lack of knowledge. Because a large number of tuberculin positive c ttle failed to show any gross lesions indicative of tuberculosis upon slaughter, the United States Department of Agriculture decided that a rigorous investigation of this problem was needed. Research was initiated at the College of Veterinary Medicine, in the Department of Microbiology and Public Health and the Department of Pathology at Michigan State University in 1959 by means of a contract with Animal Dis- ease and Parasite Research Division and a Cooperative Agree- ment with the Animal Disease Eradication Division (now the Animal Health Division), United States Department of Agric- culture. The research at M.S.U. was designed primarily to investigate the so-called no-gross-lesion (NGL) cattle. The initial investigations were concerned primarily with methods iii of isolation and identification of acid-fast organisms from tissues of gross-lesion and no-gross-lesion tuberculin positive cattle. Mycobacterium bovis was isolated from some of these NGL cattle. A large number of the isolants from NGL cattle and some from gross lesion cattle were "atypical" acid-fast organisms. They were identical or very closely related to those isolated from tuberculosis-like disease in man. The isolation of the "atypical" acid-fast organisms from tuberculin positive cattle presented new problems: Do they cause tuberculosis or tuberculosis-like disease? Do they cause tuberculin sensitivity without disease? Their relative virulence in cattle, swine, guinea pigs, rabbits, and chickens, and the routes by which calves were suscepti- ble to these "atypical" organisms were investigated. Many of the atypical organisms were Runyon Group III (slow grow— ing, nonpigmented and nonphotochromogenic) and varied mark- edly in infectivity in guinea pigs and calves. Some pro- duced a generalized progressive disease similar to the disease caused by M, bovis, some were of low virulence and produced only a primary complex and some produced no disease in experimental calves. Those which were infectious usually induced a tuberculin reaction of both mammalian and avian tuberculins. The response varied depending largely upon the extent of the disease, whether the disease was progressive iv or nonprogressive, and whether it was generalized or localized. Investigations were initiated to evaluate specific sensitins and serologic tests as aids in differential diag- nosis of mycobacterial infections. Serologic tests such as the hemagglutination (HA), complement fixation, and immuno- diffusion tests were found to yield false negative and false positive reactions. In some cases, a false positive re- sponse was attributed to an anamnestic-like reaction elic- ited by tuberculin. A recent epidemiologic study indicates that on a herd basis, the cervical skin test and 640 HA titer are supportive tests for M, bovis infection. The ambitious effort at M.S.U. has contributed greatly to our knowledge of bovine tuberculosis but con- tinued study is required before the eradication program can be effectively modified. Time is a critical factor as indi— cated in the field of human tuberculosis by Dubos (53): All biologic experience indicates that we are working against time in the campaign to erad- icate tuberculosis. The period of time may be ten years or more, but it has very definite limits. Chemotherapy has a chance of fulfilling its full potential effect in tuberculosis control, only if the public health attack is made before the current irregular and disorganized treatments have swelled the number of drug—resistant bacilli to the danger point. It is for this reason that the time for a campaign of eradication of tuber- culosis is now . . . lest it be never. The need still exists to differentiate between tuberculous infection and disease, progressive or nonprogressive, that is to detect the animals which are of potential hazard to man and other animals- A single specific serodiagnostic test and/or specific sensitins for each species of myco— bacteria is needed to accurately detect bovine tuberculosis and tuberculosis-like disease in the herd. Because some- thing is now known about the kinds of mycobacteria which exist, studies of antigen—antibody reactions can make direct or indirect meaningful contributions toward the ultimate goal of tuberculosis eradication. Such studies are under investigation at M.S.U. vi ACKNOWLEDGMENTS Because of the sc0pe and team effort of the bovine tuberculosis project, many individuals have contributed in various ways to this study. The author wishes to eXpress his sincere appreciation and thanks to Drs. W. L. and V. H. Mallmann for their continued interest and guidance through- out this study. The author wishes to thank all others on the project who participated in the research which provided the supportive pathologic, bacteriologic, and serologic information. Appreciation is also extended to the U.S.D.A. personnel who worked closely with the project and provided meaningful field interpretations. To those who have partic- ipated directly or indirectly, the author is indebted. A special appreciation goes to the author's wife, Yolonda for encouragement and her faithful help in many particulars, both large and small. Preparation of this thesis has been greatly facil- itated by financial assistance given to the author by the Michigan Tuberculosis and Respiratory Disease Association. Gratitude is extended to the United States Department of Agriculture who provided necessary equipment used in this study. vii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . l HISTORICAL REVIEW . . . . . . . . . . . . . . . . . . 4 Zone Electrophoresis . . . . . . . . . . . . . . 4 Measurement of Proteins Separated in Solid Media . . . . . . . . . . . . . . . . . 12 Immunoelectrophoresis . . . . . . . . . . . . . 14 Electr0phoresis as a Diagnostic Tool . . . . . . 18 Electrophoresis of Bovine Serum . . . . . . . .- l9 Electrophoresis in Human Tuberculosis . . . . . 22 Value of Electrophoresis in EXperimental Tuberculosis . . . . . . . . . . . . . . . . . 25 MATERIALS AND METHODS . . . . . . . . . . . . . . . . 28 Cultures for Inoculation . . . . . . . . . . . . 28 Experimental Animals . . . . . . . . . . . . . . 29 Naturally Infected.Animals . . . . . . . . . . . 31 Normal Animal and Bird Serum . . . . . . . . . . 31 Collection of Serum from Experimental Calves . . 31 Rabbit Antiserum Specific for Normal Bovine Serum . . . . . . . . . . . . . . . . . 32 Cellulose Acetate Electrophoresis . . . . . . . 33 Agar gel Electrophoresis . . . . . . . . . . . . 34 Vertical Starch-gel Electrophoresis . . . . . . 34 Two Dimensional Starch-gel Electrophoresis . . . 46 Immunoelectrophoresis .. . . . . . . . . . . . . 47 Starch—gel Immunoelectrophoresis . . . . . . . . 49 Continuous Flow Electrophoresis . . . . . . . . 50 Polyacrylamide Electrophoresis . . . . . . . . . 51 Disc electrOphoresis . . . . . . . . . . . . 51 Differential disc electrophoresis . . . . . 51 Protein, Lipoprotein, and Glycoprotein Stains . . . . . . . . . . . . . . . . . . . . 55 Cellulose acetate . . . . . . . . . . . . . 55 Agar gel and immunoelectrophoresis . . . . . 56 Starch—gel . . . . . . . . . . . . . . . . . 58 Polyacrylamide gel . . . . . . . . . . . . . 58 Continuous flow filter paper curtain . . . . 59 viii Page Densitometric Evaluation of Electrophorograms . . . . . . . . . . . . . . 59 Photography . . . . . . . . . . . . . . . . . . 60 RESULTS . . . . . . . . . . . . . . . . . . . . . . . 62 Effect of Freezing on Normal Bovine Serum Proteins . . . . . . . . . . . . . . . . . . . 62 Examination of Rabbit Anti—Normal Bovine Serums . . . . . . . . . . . . . . . . . . . . 62 Fractionation of Normal Bovine Serum by Continuous Flow Electrophoresis . . . . . . . 62 Cellulose Acetate ElectrOphoresis of Normal Bovine Serum . . . . . . . . . . . . . . . . . 65 Agar Gel Electrophoresis of Normal Bovine Serums . . . . . . . . . . . . . . . . . . . . 67 Starch-gel Electrophoresis of Normal Bovine Serum . . . . . . . . . . . . . . . . . . . . 67 Immunoelectrophoresis of Normal Bovine Serum . . 71 Starch-gel Immunoelectrophoresis of Normal Bovine Serum . . . . . . . . . . . . . . . . . 74 Polyacrylamide Gel Electrophoresis of Normal Bovine Serum Proteins . . . . . . . . . . . . 76 Comparison of the Major Electrophoretic Procedures Used for the Analyses of Normal Bovine Serum . . . . . . . . . . . . . 85 Cellulose Acetate Electrophoresis of Serum from Tuberculous Cows . . . . . . . . . . . . 87 Experimentally infected animals . . . . . . 87 Naturally infectedwanimalsn+ - - . . . . . . 102 Starch-gel.Electrophoresis of Serum from Tuberculous Cows . . . . . . . . . . . . . . . 106 Differential Disc Electrophoresis of Serum from Tuberculous Cows . . . . . . . . . . . . 106 Immunoelectrophoresis of Serums from Tuberculous Cows . . . . . . . . . . . . . . . 116 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 124 Development of Techniques and Analyses of Normal Bovine Serum Proteins . . . . . . . . . 124 Analyses of Serums from Tuberculous Cattle . . . 136 SUMMARY 0 O O O O O O O O O O O O O O O O O O O O 0 O 149 LITERATURE CITED . . . o . . . . . . . . . . . . . . . 153 ix LIST OF TABLES List of animals inoculated with viable mycobacteria . . . . . . . . . . . . . . . . . Stock solution . . . . . . . . . . . . . . . . Working solution . . . . . . . . . . . . . . . Serum protein changes, hemagglutination titers (HA) and pathology of calves inoculated with mycobacteria . . . . . . . . . Serum protein changes, hemagglutination titers (HA). pathology. and bacteriology of naturally infected cows . . . . . . . . . . . . . . . . Mean values of serum protein fractions and variations due to disease activity in bovine tuberCUlOSiS I O O O O O O O O O O O O O O O 0 Page 30 52 52 97 103 105 Figure LIST OF FIGURES Exploded view of a modified mold for starch- gel electrophoresis. Removable spacers (a). (Dimensions are in inches and only critical dimensions illustrated) . . . . . . . . . . Retainer plate. A: Front view. B: Cross section of gel mold. C: Cross section of. retainer plate. Spirit level (L). (Dimen— sions in inches and only critical dimensions illustrated.) . . . . . . . . . . . . . . . Modified gel-mold holder and buffer compart- ments (electrophoretor) for starch-gel elec- trophoresis. Schematic representation show- ing only the important dimensions. Bolts (a) for clamping the retainer plate in place. Co-axial terminals (b). (Dimensions in inChes O ) O O O O O O O O O O O O O 0 O O O O Slicing device. Metal bars (a). The notches in the middle of the two metal bars are for varying the thickness of the slice. Plexi- glass gel holder (b). The gel holder is guided by (a) and by the plexiglass runners (c). 0.008 inch piano wire (W). (Dimensions are in inches and only critical dimensions illustratedo) O C O O O O O O O O O O O O 0 Plastic-screen-gel-holder. A: gel holder (H), disc gel adaptor (D) and fiber glass clips (fg) B: diagramatic illustration showing how gel holder (H) is assembled for starch—gel slabs (SG) and disc gel columns (dc). Note that the disc gel adaptor is not used for starch—gel slabs. . . . . . . . . . . . . . . . . . . . Multiple staining tank . . . . . . . . . . . xi Page 37 38 39 41 43 44 Figure 10. 11. 12° l3. 14. A: destaining assembly. Destainer (d), 11% x 10% x 2% inches i.d. (bc) 6-12 volt battery charger. The destaining solution circulates through charcoal bed (cb) by pump (p). B: Sagital section through destainer. Gel holder (H) placed in slot (5) so cathode is % and anode % inch from gel (g). C: Filter unit (13 x 5% x % inches o.d.). 8 inch charcoal bed (cb) holds 2 lbs. granular charcoal, held in place by two % inch thick glass wool filter pads (w) held in place by filter holder (fh) mounted on back of screw—on cover (t). Elec— trodes (e), inlet (1), plastic screen (n), and outlet (0) . . . . . . . . . . . . . . . . Diagramatic representation of the position of the gels in the differential disc electropho- retic COlumn O O O O O O O O O O 0 ° 0 O O O 0 Paper curtain electrophorogram of normal bovine serum after separation by continuous flow electrophoresis. The numbers at the bottom of the curtain represent fraction numbers . . . . . . . . . . . . . . . . . . . Cellulose acetate electrophorograms of con— tinuous flow electrophoretic effluent serum fractions . . . . . . . . . . . . . . . . . . Typical densitometric recordings of cellulose acetate electrophorograms after electrophore— sis with the Owen buffer system. ,A: protein. B: protein—bound carbohydrate. C: protein- bound lipid . . . . . . . . . . . . . . . . . Electrophorogram of normal bovine serum in agar gel with the Owen buffer system. (1) albumin, (2) alpha globulin, (3) beta glob- ulin, (4) gamma globulin . . . . . . . . . . . Single dimensional starch—gel, cellulose acetate, and two dimensional electrophoretic patterns of normal bovine serum. (The cross hatched areas represent lightly stained bands.) . . . . . . . . . . . . . . . . . . . A composite drawing of agar immunoelectropho— retic patterns of normal bovine serum proteins xii Page 45 54 63 64 66 68 69 72 Figure 15. 16° 17. 18. 19. 20. 21. Page Starch-gel immunoelectrophoretic pattern of normal bovine serum. (The cross hatched areas represent lightly stained bands.) . . . . . . . Comparison of the separation of human serum in standard 7% gel (left), 4.75% gel (middle), and differential gel (right). Cross hatched areas and dotted lines indicate lightly stained and fine bands, respectively. (0) origin of the 10% separation gel, (a) 4.75% gel, (b) 10% gel. Spacer gel not shown . . . . . . . . . . . . . . Comparison of the separation of normal bovine serum in standard 7% gel (left), 4.75% gel (middle), and differential gel (right). Cross hatched areas and dotted lines indicate lightly stained and fine bands, respectively. (0) ori- gin of the 10% separation gel, (a) 4.75% gel, (b) 10% gel . . . . . . . . . . . . . . . . . . Densitometric tracing of the normal bovine serum separated by the differential disc elec- trophoretic method illustrated in Figure 21. The glycoprotein fractions (c) and lipoprotein fractions (1) are also indicated. Sa2=slow alpha2 lipoprotein . . . . . . . . . . . . . . . Comparative electrophorograms of normal serum proteins. A: human serum: (1) conventional procedure, (2) 7% gel-12 cm column, (3) 4.75% gel-12 cm column, (4) and (5) differential procedure. B: bovine serum: (1) conventional procedure, (2) 4.75% gel—12 cm column, (3) differential procedure . . . . . . . . . . . . . Differential disc electrophorograms of normal human serum showing reproducibility of the method . . . . . . . . . . . . . . . . . . . . . Eight individual bovine serums from a 16 column run by the differential system, illustrating variation among serums and reproducibility as measured by the distance migrated by the. albumin in a single run . . . . . . . . . . . . xiii 75 77 78 79 81 82 84 Figure 22. 23. 24. 25. 26. 27. 28. 29. 30. Comparison of the separation by the conven- tional (left) and differential (right) systems of initial serum (1) and post-inoculation serum (P) from a cow experimentally infected with Mycobacterium bovis . . . . . . . . . . .f Diagramatic representation showing the rela- tionships among differential disc electropho- resis, cellulose acetate electrophoresis, agar immunoelectrophoresis, and starch-gel electro— phoresis of normal bovine serum (top to bottom) . . . . . . . . . . . . . . . . . . . Typical densitometric recordings of cellulose acetate electrophorograms of bovine serums. (Top) M, bovis. A: pre-inoculation. B: 132 days post-inoculation. (Bottom) 68C (viru- lent Group III) . . . . ... . . . . . . . . . Typical densitometric recordings of cellulose acetate electrophorograms of bovine serums. (Top) M. avium. A: pre-inoculation, B: 132 days post-inoculation. (Bottom) 508 (aviru- lent Group III). C: pre-inoculation, D: 172 days post—inoculation . . . . . . . . . . . . Typical densitometric recordings of cellulose acetate electrophorograms of the two types of patterns from cattle infected with 51C (Viru- lent Group III). (Top) Type I. (Below) Type II. A: and C: pre-inoculation, B: and D: 179 days post—inoculation . . . . . . . . . . . . Alterations in serum protein and protein—bound carbohydrate patterns in calves experimentally infected with M. bovis . . . . . . . . . . . . Alterations in serum protein and protein-bound carbohydrate patterns in calves experimentally infected with M. avium . . . . . . . . . . . . Alterations in serum protein and protein—bound carbohydrate patterns in calves eXperimentally infected with strain 51C (virulent Group III). Alterations in serum protein patterns in calves experimentally infected with strain 68C (viru- lent Group III) . . . . . . . . . . . . . . . xiv Page 84 86 88 89 9O 91 93 94 95 Figure 31. 32. 33. 34. 35. 36. 37. 38. 39. Alterations in serum protein and protein- bound carbohydrate patterns in calves eXperimentally infected with 50B (aviru- lent Group III). . . . . . . . . . . . . . . . Typical starch-gel electrOphorograms of serums from eXperimental calves inoculated with g. bOViS 0 0 0 O O O O O O O O O O 0 0 0 Typical starch-gel electrophorograms of serums from experimental calves inoculated with 68C (virulent Group III) . . . . . . . . Typical starch-gel electrophorograms of serums from experimental calves inoculated with 51C (virulent Group III) . . . . . . . . Typical starch-gel electrophorograms of serums from experimental calves inoculated with M. avium (top) or (below) 50B (avirulent Group III) . . . . . . . . . . . . . . . . . . Typical densitometric recordings of starch-gel electrophorograms of (A) normal bovine serum and (B) post-inoculation serum from experimen- tal cattle inoculated with M. bovis, 68C or 51C, and gross—lesion and no-gross lesion cows from naturally infected herds . . . . . . Differential disc.electrophorograms of pre— and post-inoculation serums from experimen- tal calves inoculated with M, bovis, 68C or 51C (virulent Group.III) . . . . . . . . . Typical differential disc electrOphorograms of pre- and post—inoculation serums from eXperimental calves inoculated with M, avium or 50B (avirulent Group III) . . . . . . ... . Typical differential disc electrophorograms of pre— (P) and post-tuberculin (T) serums from tuberculin negative (TB neg.) cattle, and from tuberculin positive gross-lesion (GL) and no- gross—lesion (NGL) cattle from naturally in— fected herds. Note the decrease in all pro- tein bands in the (T) samples from both GL and NGL animals but not in the (T) serum from TB neg. animals . . . . . . . . . . . . . . . . . XV Page 96 107 108 109 110 111 113 114 Figure 40. 41. 42. 43. 44. 45. Page Comparison of differential disc electrophoro- grams of normal bovine serum and serum-from tuberculous cattle with serums from other chronic inflamatory diseases of cattle . . . . . 117 Immunoelectrophorograms of pre— and post- inoculation serums from calves with M. bovis. A. pre-inoculation serum (normal), (B) 14 days post—inoculation serum, (C) 82 days post-inocu- lation serum, (D) 132 days post-inoculation serum . . . . . . . . . . . . . . . . . . . . . 118 Immunoelectrophorograms of pre- and post- inoculation serum from calves with 68C (viru- lent Group III). (A) pre-inoculation serum (normal), (B) 14 days post-inoculation serum, (C) 82 days post-inoculation serum, (D) 173 days post-inoculation serum . ... . . . . . . . 119 Immunoelectrophorograms of pre- and post— inoculation serum from calves with 51C (viru- lent Group III). (A) pre-inoculation serum (normal), (B) 14 days post-inoculation serum, (C) 82 days post-inoculation serum, (D) 173 days post-inoculation serum . . . . . . . . . . 120 Immunoelectrophorograms of pre— and post- inoculation serums from calves with M, avium. (A) pre-inoculation serum (normal), (B) 14 days post-inoculation serum, (C) 82 days post-inocu~ lation serum, (D) 132 days post-inoculation serum .. . . . . . . . . . . . . . . . . . . . . 122 Immunoelectrophorograms of pre- and post- inoculation serums from calves with 50B (aviru- lent Group III). (A) pre-inoculation serum (normal), (B) 14 days post-inoculation serum, (C) 82 days post-inoculation serum, (D) 132 days post-inoculation serum . . . . . ... . . . 123 xvi INTRODUC TI ON Many attempts have been made to use serologic tests to diagnose tuberculosis in man and animals. Various anti- gen preparations and serologic techniques such as agglutina- tion, precipitation, complement fixation, and hemagglutina- tion with its various modifications, have been investigated. No test has reliably differentiated tuberculosis, tuberculo- sis-like disease caused by "atypical“ acid-fast organisms, or nontuberculosis. Electrophoretic studies of the serum proteins in tuberculous and nontuberculous humans and animals have been made to find specific changes that might be of diag- nostic or prognostic significance. Although some success has been achieved, most of the changes detected reflected the nonspecific physiologic change in the body rather than a change specific for tuberculosis. Moving boundary or paper electrophoretic methods were generally used. They lack the sensitivity of gellified media tests, such as starch—gel electrophoresis, acrylamide electrophoresis, and immunoelectrophoresis. It is possible that specific frac— tions, antibody or non-antibody proteins, or both, could be detected and isolated with improved fractionation precedures. Considerable emphasis has been placed on the need to isolate specific mycobacterial antigens. To utilize specific antigens in a serodiagnostic test, antiserums must be exam- ined to demonstrate the fractions that contain any diagnostic specific antibodies or antigen-antibody complexes. The failure to demonstrate specific antibodies from some tuber- culous individuals may not be the failure of the animal to produce antibody, but failure in the design of the test procedure. Also the mechanisms of antibody production against mycobacterial antigens and natural resistance mech- anisms in the animal must be investigated. According to Seibert (167), there may be little progress in the develop- ment of serologic tests for tuberculosis because of the interplay between antigens, antibodies, and lysozyme ig.yiyg. More understanding of the interactions is needed before diag- nostic procedures for the detection of tuberculosis and tuberculosis—like disease can be devised for the practicing physician and veterinarian. This is a report of the studies made of serums from normal cattle and from infected cattle. Various separatory procedures were used for the fractionation of normal bovine serum proteins: cellulose acetate electrophoresis, starch— gel electrophoresis, acrylamide electrOphoresis, and immuno- electrophoresis. A standard nomenclature is proposed for the electrophoretic and immunoelectrOphoretic patterns of normal bovine serum components. Results are presented of the electrophoretic and immunoelectrophoretic evaluations of serial changes in serum proteins from naturally infected cattle and from cattle experimentally infected with various strains and species of mycobacteria. HISTORICAL REVIEW There are numerous methods available for analyses of serum proteins. Electrophoresis, the separation of ionic mixtures in an electrical field, is one of the most impor- tant analytical techniques presently available to the research and clinical scientist. Zone Electrophoresis Zone electrophoresis refers to a number of tech- niques in which the electrophoretic separation is carried out in a solid or semisolid support medium which is a modi- fication of the liquid system employed in the classical moving boundary or free solution electrophoretic method (190). It offers the advantage that particular zones may be stabilized and stained or isolated for further investigation. Various stabilizers as solid supporting media have been employed such as: glass beads, sands, resins, powdered cellulose, asbestos, sponges, filter paper, cellulose ace- tate membranes, and various gels (103). The supporting medium is ideally inert and the separation of components results from their electrophoretic mObility. Gel filtration can be an additional separatory technique in some types of zone electrophoresis such as starch—gel and acrylamide gel. The basic design of all zone electrophoretic systems consists of two electrode vessels with the electrolytic bridge (supporting medium) between them. A potential dif- ference applied across the two vessels of electrolyte will cause a current to flow across the electrolytic bridge. The flow of electrons is toward the anode. An ionizable sub- stance separates into its different electrophoretic compo- nents because of variation in charge, molecular weight, or both (46, 54, 103). A number of factors influence the electrophoretic separation: the concentration,constituents, consistency, and pH of the buffer; isoelectric points of the sample; temperature; and electroosmotic flow. The main functions of the buffer solution are to conduct current and maintain pH of the sample. Since the majority of serum proteins have their isoelectric points between pH 4 and pH 8 the electrolyte solution is usually buffered between pH 8.2 and pH 8.6 (46). Above pH 8.0, most, if not all, serum proteins react as anions in an electrical field (33). The ionicity of the electrolyte' solution must be sufficient to provide adequate buffering capacity. The maximum permissible ionicity is determined by the amount of heat due to resistance that can be toler- ated. Because of the heating effect of currents the ionicity, temperature, voltage and pH must be carefully regulated, particularly when gellified media are used. Excessive current can cause drying of the supporting medium, pH changes as a result of electrolysis of buffer salts, and increased ionic mobilities (54, 196). To avoid depleting the buffer capacity of the electrolyte solution due to electrolysis of buffer salts during electrophoresis, a buffer solution with low ionicites between 0.075 and 0.1 is used (106). The constituents of the buffer solution may also influence the separation of an ionic mixture. A buffer con- sisting of tris (hydroxymethyl) aminomethane (tris), ethylenediaminetetracetic acid (EDTA), and boric acid was described by Aronson and Gronwal (1). Human serum proteins were separated into 9 fractions which was superior to the separation in the barbital buffer commonly used in paper electrophoresis. A discontinuous buffer system in which tris-EDTA- borate buffer is applied to paper, and barbital buffer in the buffer vessels was described by Goldberg (63). Resolu- tion of abnormal hemoglobulins was superior to that in barbital or tris-EDTA-borate single buffers. The addition of either calcium or magnesium salts to barbital buffer facilitates the separation of the beta globulins (107, 108). Water tends to move toward the cathode during elec- trophoresis in alkaline solutions because of negatively charged carboxyl groups on many of the supporting materials (97, 103, 180). The movement of the water is called electro- osmotic flow. It is greater in buffers which are less concentrated or at a pH which increases the ionization of the carboxyl group of the supporting material. Little electroosmosis occurs in filter paper and cellulose acetate relative to that which occurs in agar gel (46). These and other factors regulating the separation obtained by zone electrOphoresis have been discussed (15, 16, 43, 109, 125, 157, 196, 211). The greatest impetus to zone electrophoresis was the introduction of filter paper as a supporting medium, It afforded many advantages: complete separation of individual components; storage for a permanent record; required only a small sample volume; facile; many analyses performed simul- taneously; no prior dialysis of sample required; and analy- ses made‘atroom temperature (157). Despite its great use- fulness and many applications, interest in paper electropho- resis as a clinical tool has diminished in the last several years. Because of albumin tailing, the added background color causes the percentage of albumin to always be low while the globulin values are correspondingly increased (166). Cellulose acetate was introduced as a supporting medium for zone electrophoresis by Kohn (94). The strips are used the same as paper strips but it offers some notable advantages. The background is transparent and colorless, and the definition of the protein bands are much sharper (42, 94, 97). An investigation of the effect of buffer composition, ionicity, and pH, electrolyte volume, strip length, voltage, sample size and position, and migration time, on the sample separation obtained with this medium was reported by Brackenridge (24). A microadaptation using cellulose acetate has been developed by Graham and Grunboom (66) which permits separation of serum proteins within 15 minutes. This medium may also be used for immunoelectro- phoresis (95, 97, 165). Agar gel electrOphoresis is a widely used technique, particularly the method of Wieme (207) utilizing agar coated microscopic slides. A recent modification employs movie film leader strip (Cronar strips) which is firm yet flexible enough to fit into some of the commercial equipment orig— inally made to accommodate paper strips (38). In agar, as with cellulose acetate, there is not tailing of the albumin and the bands are sharply delineated from the background (43, 45, 64). The chief disadvantage is that considerable elec- troosmosis occurs (101, 103). Starch-gel (179, 180) and acrylamide gels (132, 153) are finely porous and act as sieves as the protein molecules migrate through them. Thus, the proteins are separated not only by differences in electrophoretic mobility but also by their molecular size (179). The substantial increase in resolving power of sieving gel matrices over that of con- ventional filter paper electrophoresis is effectively demon- strated with a complex protein mixture such as human serum. In a single electrophoretic analysis, a filter paper tech- nique commonly resolves 5 to 7 protein fractions; starch and acrylamide gels resolves 20 to 30 fractions. The separating power of these gels has been thoroughly discussed by Smithies (181, 182); Raymond (152, 154); and Ornstein (132). Starch—gels can be used in horizontal (179) or ver- tical systems (180). In either method, starch—gel is made into a block and the sample inserted into a slot in the gel. ElectrOphoresis proceeds for 18 hours. Several layers are cut from the block. One layer is usually stained for pro- tein and the others are stained for specific components of serum. The vertical method permits the sample to be inserted into the gel without any supporting substance, reduces elec- trodecantation, improved the resolution of serum proteins considerably, and reduced the difficulties of obtaining reproducible results. A discontinuous system of buffers in starch—gel medium permitted higher voltage gradients, reduced the time required for electrophoresis and increased resolution of toxins, enzymes, and human serums (141). Starch—gel electrophoresis permitted the separation of certain serum proteins, such as the haptoglobins, which had not been possible with the standard electrophoretic pro- cedures (183, 185). As a result, starch-gel has been employed extensively in investigation of serum proteins, enzymes, hemoglobins, transferrins, and haptoglobins. 10 It has also been employed to demonstrate the para- proteins found in cases of myeloma (l4), and for separation and characterization of LDH isoenzymes (87, 212). Starch—gel electrophoresis has not become popular in the routine diagnostic laboratory because it is technically quite difficult. However, Marsh et al. (121) have described a modification of the conventional technique which makes it simple to perform, flexible, more rapid,.and more adaptable to routine clinical usage. Difficulty in correlating findings of paper and starch—gel electrophoresis stimulated development of a two- dimensional electrophoretic analysis combining migration first in filter paper and second at right angles in starch gel (143). The protein components common to both the whole serum and the serum fraction obtained from filter paper electrophoresis can be identified as uninterrupted bands across the entire width of the gel. .Ashton (6) found bands not detected by agar or starch—gel singly by using agar in the first direction and starch—gel in the second direction. The chief advantage of this method is to correlate bands detected in starch with those separated in a classical sys- tem (paper, cellulose acetate or agar). The utilization of polyacrylamide gels for zone electrophoresis was reported independently by Raymond and Weintraub (155) and Davis and Ornstein (50). The vertical 11 slab procedure separates and resolves serum proteins and hemoglobin qualitatively similar to that seen in starch-gels (152, 155). The properties and preparation of acrylamide gels and applications was reviewed by Raymond (150). There is In; retrograde migration of gamma globulin due to elec— troendosmotic flow as occurs in starch-gels (149, 153). Heremans et a1. (76) described a two-dimensional technique employing paper in the first direction and cyanogum gel in the second direction. Raymond and Aurell (151) employed a low concentration acrylamide gel in the first direction followed by a high—concentration gel in the second direction to produce separations analogous to those obtained in two- dimensional paper chromatography. Disc electrophoresis (50, 132) employs polyacryla- mide gel in narrow glass cylinders. The salient character- istics of this technique reside in (1) the controlled varia- tion of the pore size for the purpose of increasing the resolution of ions based on dimensional differences and (2) an electrophoretic step for concentrating the sample ions into a narrow starting zone prior to electrophoretic separation. Human serum proteins, in particular, have been extensively examined by this method with over zp protein zones demonstrated (49, 122, 214). However, the number of protein bands are so numerous that the system is of no benefit to the routine diagnostic laboratory because satis- factory interpretation of the complex patterns is not yet 12 possible. Furthermore, the gamma globulins are not ade- quately separated. They migrate as a broad zone from point of origin to the transferrin band which makes the background gel hazy (49). The greatest use of this technique presently is with special histochemical stains particularly for the study of enzymes. Because a very small quantity of protein is required it is particularly effective for separation of proteins in body fluids with low protein content (49). Electrophoretic separation of serum proteins in starch and acrylamide gels is technically far more advanced than our knowledge of the significance and characteristics of these substances. Until these are understood, the sepa- rating power of starch and acrylamide gels will not be fully utilized. Measurement 9§_Proteins Separated ig_Solid Media After proteins are separated in solid media, the methods for measuring are generally grouped as chemical or densitometrical. For clinical purposes, scanning is a more convenient and simple procedure. The amount of dye that binds to a protein will vary with the dye and the protein. Errors in measuring protein by bound dyes are inherent since the binding capacity of many proteins is unknown. If given systems are standardized, and automatic scanners used the errors can be negligible (4, 25, 31, 35, 91, 97, 136). 13 The densitometric analyses of proteins separated in starch-gels are more difficult. Before making transmission measurements of the starch—gel, the translucent destained gel must be made transparent by heating the gel slowly in 5 to 10% acetic acid or glycerol (67, 136, 160). Glycerol is preferable because no dimensional changes occur during the prolonged period of the gel manipulation. .Heating in acetic acid makes the gels more fragile, and due to drying shrink- age may reduce reproducibility. Slicing the starch—gel introduces another error because the gel is too elastic and soft to cut at a uniform thickness (123). The use of acryla- mide gels eliminates the need of treatment before scanning because of their transparency. Densometric analyses of proteins separated in gels are discussed in the following papers (73, 136, 160, 195, 213). The elution of proteins from starch-gels is diffi- cult but can be done by: (l) digesting the starch with amylase (2) freezing and thawing the gel which breaks the structure and leaves a sponge-like substance followed by centrifugation and (3) by electrodialysis (17). The latter is the most satisfactory method and is detailed by Moretti et al. (126). Elution of stained bands from cellulose acetate strips is very satisfactory and rapid, more so than filter paper. Ponceau S stained bands can be eluted with sodium hydroxide followed by the addition of acetic acid (58, 97). l4 Immunoelectrophoresis Immunoelectrophoresis (IE) is a sequential combina- tion of electrophoretic separation and double diffusion in the same agar (65). The resolution is markedly increased. It resolved more than twenty distinct antigenic components in human serum. Scheidegger (164) used microscope slides which reduced the amount of reagents required and shortened the periods of electrophoresis and immunodiffusion. Supporting materials other than agar have been used for IE: paper (59, 140, 205); starch-gel (3, 99, 142); cellulose acetate (94, 96, 98); and polyacrylamide gels (86). Agar was the most useful because it is negatively charged, has a high gel strength, is nearly transparent, is soluble in an aqueous medium, has few ionized groups, and undergoes little nonspecific reactions with proteins (46, 208). Starch— gels were opaque so that relatively high concentrations were required which increased the possibility of nonspecific interactions with proteins (99, 142) and reduced the rate of diffusion. Highly purified agar is used to decrease electro- osmosis and extraneous salts (46, 80, 101). To insure reproducibility of results, agar should be prepared in a single batch (46). Procedure for IE essentially consists of covering glass plates or slides with 1-2% agar dissolved in an appro- priate buffer. After solidification, a well is cut in the 15 agar and filled with sample. It is connected to the two buffer vessels with filter paper, and electrophoresed as for zone electrophoresis (p. 5). Following electrophoresis, a trough is cut in the agar parallel to the axis of migra- tion. Antibodies specific for the antigen constituents of the sample are added. The plates or slides are incubated in a moist chamber at a predetermined temperature. The antigens and antibodies diffuse radially from their respec— tive centers and form visible bands of precipitates which can be stained (46). Other identification procedures can be used such as special staining techniques, enzymatic reactions, or radioimmunoelectrophoresis (45, 64, 209). The same variables influencing the separation of ionic mixtures in zone electrophoresis (described on page 5) also influence IE. In addition, the electrophoresis and immunodiffusion must be carried out in the same environment. The required buffer system for electrophoresis may not be the optimum conditions for antigen-antibody precipitation. The low ionic strength used in agar electrophoresis will not support specific precipitation by some antiserums (46). The ionicity, pH, and type of salts used in the buffer have to be determined empirically, and frequently it is a com- promise. Low ionicity increases the rate of migration of fractions’but broadens their zones and lowers resolution. 16 Increasing the ionicity causes overheating and decreases the diffusion rate of macromolecules. A pH should be selected which maintains the negative charge on the anti- genic constituents to prevent their absorption to the gell- ing medium. Barbital buffers at pH 8.2 to 8.6 with low ionicities of 0.025 to 0.1 are generally used (46). It is not without its limitations. The barbital precipitates some non-antibody components of human serum (100), probably the alpha lipoprotein (46). Optimal conditions for immunoprecipitation follow- ing electrophoresis are the same as those for gel double diffusion tests. At least three other problems may occur. First, electrophoresis can separate antigens from each other, and one may depend upon the other for their solubility in agar gels. When the protective substance, probably an alpha ~1iproprotein (194) is removed, the unprotected 2 molecules may form irregular precipitin arcs in the agar which are not due to specific antigen-antibody reactions (46). The addition of gelatin or glycine to the agar may act as protective agents (46). Second, the slopes of pre- cipitin arcs in immunoelectrophoresis are governed by more factors than those in double diffusion plates primarily: mobility, electrophoretic heterogeneity, slope of the anti- gen diffusion fronts, the diffusibility of the antigens, and the relative preparations of the reactants (46, 64, 80, 92). 17 Third, and probably most important, a molecule or complex with two different determinant groups may be split by electrophoresis or denatured during electrophoresis. The probability of two or more antigens forming superimposed precipitates is decreased by IE because the electrophoretic mobility as well as the diffusibility and reactant ratios would have to be identical or very closely similar. The probability of the simultaneous occurrence of all those factors, during IE, however, is very slight (81). One of the more important reagents and chief limit- ing factor of IE is the antiserum. Variations occur due to the differences in antigenicity of the various constituents in the test sample and the qualitative and quantitative dif- ferences in the antibodies elicited (22, 46, 147, 206). The quantity, quality, and kind of precipitating antibodies can be influenced by route of injection, physical state of anti- gens, and variability in the immunologic responsiveness of individual animals of the same and different species. Crowle (46, p. 93) states: The dynamics of single and double diffusion tests and of immunoelectrophoresis are sufficiently complex to call for a great deal of caution in interpreting results obtained from those tests, but this very complexity contributes to their ability to yield remarkably varied information involving antigen-antibody precipitations con- sequently, efforts made by users of these tests to understand their intricacies will be rewarded with a generous return of knowledge. 18 Electrophoresis a 3 Diagnostic Tool One of the most significant contributions Of elec— trophoresis has been the identification of different hemo— globulins (88) and haptoblobulins (180). However, the analyses of serum and plasma requires that normal patterns be established. Patterns have been established for several animal species but the changes in pattern due to disease and nutritional deficiency have rarely been of direct diagnostic value. They are more frequently a general index to the physiologic condition of the patient (111). To interpret changes in electrophoretic patterns it is essential to know the site of production, half life, and site of destruction of serum proteins ig_yiyg. It is gen- erally believed that albumins, fibrinogen and alpha globu- lins are produced in the liver. Gamma globulins and possi- bly some beta globulins are probably produced by plasma.. cells in splanchnic bed, bone marrow and lymph nodes (55). Some protein, primarily albumin, is filtered and reabsorbed in the kidney which maintains normal colloidal osmotic pres- sure (72, 137). Under certain pathological conditions, large amounts are lost. Significant changes in protein pattern may be encountered in liver disease, in disorders of the reticuloendothelial system, or when abnormal amounts are lost through the kidney. Non-specific, stress patterns and bacterial infections (13, 137) can cause a hypoalbumin- emia, decreased beta globulins and increased alpha and gamma 19 globulins. Alpha globulins, particularly alpha increase 2. in high fever and tissue destruction. Alpha2 increases markedly in the nephrotic syndrome. Beta globulins are often increased in conditions in which increased serum phospholipid and cholesterol are associated. Hypergamma- globulinemia occurs in chronic infection, in liver disease and collagenous disorders. Hypergammaglobulinemia was usually indicative of increased antibody production (11, 69, 190) although some gamma globulins may be serologically "inert" (12). Heremans (75) found that the gamma A, gamma M and the gamma G globulins of the human system possessed antibody activity. In most bacterial diseases, the gamma globulin concentrations did not parallel resistance (13). Specific patterns have been described for relatively few diseases. Many of the disease-induced alterations that have been detected in the plasma constituents of men and animals, are noted in the following reviews (13, 32, 36, 37, 55, 69, 71, 89, 100, 105, 110, 111, 114, 115, 134, 137, 146, 156, 166, 186, 187). Electrophoresis p£_Bovine Serum With the moving boundary electrophoretic technique of Tiselius, bovine plasma were separated into five major fractions--albumin, alpha, beta, and gamma globulin and fibrinogen (84, 162). Alpha globulin was separated into two and gamma globulin into three subfractions (26, 27, 138). 20 Some differences are reported in the number of com- ponents separated in bovine serum by paper electrophoresis but, in general, the same major fractions were Obtained with the moving boundary technique. ,Rooney (159, personal communication) could not separate the components in bovine serum as satisfactorily as those in human serum. Individual components diffused more and the alpha globulin did not separate with any of the phosphate or veronal buffer systems utilized. With paper electrophoresis bovine serum separated into more than six components (18, 19, 47, 93, 138). The beta fraction was difficult to separate by paper electropho- resis in veronal buffer (47). Tris-buffer increased the resolution of more globulin fractions (Czernicki, personal communication). The various amounts of protein found in bovine serum by electrophoretic techniques was reviewed by Nilssen (130). The results obtained by various authors are not necessarily comparable. It is essential to consider the different apparatus, buffer solutions, etc. which were used (198). Cellulose acetate has not been employed as a medium for the separation of bovine serum, and agar gel has been used only in a few cases (Cornelius, personal communication) with similar problems in separating the globulin fractions. ,Bovine serums have been studied with starch-gel electrophoresis to detect serum protein polymorphism. The polymorphism is associated in the beta—globulin region of 21 the electrophoretic pattern. Cattle serums on the basis of the variation in the beta-region, can be divided into five groups (6, 184); a sixth group was subsequently recognized (8). Whole bovine serum was separated up to 25 components (7, 104, 184) by one directional starch—gel electrophoresis and into 24 components by two-dimensional agar or starch-gel electrophoresis (7). {Brummerstedt-Hansen (30) studied transferrin types in cattle with starch—gel electrophoresis and immunoelectrophoresis. Disc electrophoresis produced more discrete and more easily analyzed transferrin bands than starch-gel electrophoresis (148). The more sophisticated gel electrophoretic proce-» dures have had limited use in evaluating various diseases of cattle. Most reports have been on the use of moving boundary or paper electrophoresis. San Clemente and Huddleson (162) analyzed serums from cows infected with brucella by the mov- ing boundary method. (Moving boundary method detected hyper— beta- and hypergammaglobulinemia one and three weeks, respec— tively, after cattle were experimentally infected with the virus of foot-and—mouth disease (26, 27). Rooney (159) studied seven bovine disorders by paper electrophoresis: Ramsey-type mucosal disease (hyperalpha-globulinemia); lymphomatosis (double alpha peak, hypoalbuminemia and hyper- gammaglobulinemia); vegetative endocarditis (hypoalbuminemia and hyperalpha- and hypergammaglobulinemia); traumatic 22 reticulitis and hydronephrosis, amyloidosis (hypoalbuminemia and hypergammaglobulinemia); foot rot (hypoalbuminemia and hypergammaglobulinemia). A hyperalpha-, hyperbeta- and hypergammaglobulinemia was found by paper electrophoretic analyses of cows eXperimentally infected with anaplasmosis (51, 158). Brown (28, 29) used agar electrophoresis, immunoelec— trophoresis and paper electrophoresis of normal serums and serums from cows and guinea pigs experimentally infected with the virus of foot—and—mouth disease. The mobility of the precipitating antibodies altered considerably during convalescence was of prognostic value. Electrophoresis ip_Human Tuberculosis Many attempts have been made to find specific anti- gens, antibodies or other components in the blood which could be of diagnostic or prognostic value for tuberculosis. Measurement of erythrocyte sedimentation rate, blood volume, pH, non-nitrogenous protein content and inorganic ion con- tent have been examined (113, 127). No single test or com— bination of tests is conclusively specific for tuberculosis. Some of the changes aid in determining the extent and activ- ity of the disease and the prognosis. The electrophoresis of the serum proteins from tuberculosis individuals has been studied extensively and with similar results. Early studies of serum by moving boundary electrophoresis revealed changes that roughly 23 parallel the extent and activity of the disease. Hypoalbu- minemia and a reduction in the normal albumin/globulin ratio (A/G) were common (168, 169, 172). Increase in the alphal, alpha and gamma globulins and decrease in albumin occurred 2. in the serum as the disease progressed (12, 102). Similar changes have been shown by paper electroPhoresis (20, 39, 41, 62, 74, 117, 124, 139, 197). Starch—gel electrophoresis detected no specific quantitative change in serums from 25 patients with active tuberculosis (90). These electrOphoretic results clearly demonstrate that there are definite departures from normal values in serum proteins in active human tuberculosis. An increase in alpha globulin occurs in all forms of disease. Nymen (131) 2 and Hever and Kalnai-Hever (77) reported that haptoglobins contributed most to the rise in the alpha globulins. 2 Bovornkitti (20) believed that the increase in alpha2 globulin occurs as a result of hypersensitivity reactions and is not due to tissue destruction as previously suggested by others. However, changes in other components are related to liver impairment. Because of the relationship between hyperalpha 2-globulenemia and disease activity, some workers believe that the albumin alpha globulin ratio is more valu- 2 able and specific than the A/G ratio (139, 174). McCuiston and Hudgins (124) found that the alpha globulins were not 2 elevated in the serums of patients with tuberculosis caused by nonphotochromogenic atypical mycobacteria. They 24 emphasized (l) the diagnostic value of the pattern rather than the quantitative values and (2) that the results obtained are important only when evaluated with the total clinical picture. Grigorieva and Livshitz (68) interpreted the change in the beta globulins as being the most significant change during tuberculosis. The rise in beta globulin was propor- tional to severity and extent of disease and to the degree of amyloidosis. There was no definite pattern of relation- ship between the character of the disease and.the gamma globulins. The relationship between clinical progress and hypergammaglobulins often observed in advanced and far advanced tuberculosis is controversial. Some workers have found a rising or high gamma globulin level to be a good prognostic sign (163, 172, 215, 216, 217). Others have failed to find consistent relationships (135, 178). Freigang et a1. (57) found as improvement occurred the albumin alpha goes up and gamma decreases. 2 Significant changes in serum glycoprotein levels in human tuberculosis (171, 172, 193) suggest that glycoprotein determinations might be useful in diagnosing tuberculosis. However, electrophoretic patterns of carbohydrate distribu- tion in serum from tuberculosis patients were of no diagnos- tic value (79). 25 Price (144) used disc electrophoresis to examine the seromucoid patterns in patients with a variety of different diseases, including tuberculosis. Intensive work on normal individuals and tuberculous patients indicated a typical pattern for normal and another for active tuberculosis. However, the significance of the seromucoid pattern and the nature of their association with different disease states is not known. Value 93 ElectrOphoresis i2. EXperimental Tuberculosis Paper has been the principal medium used in the electrophoretic analysis of the serum protein changes that occurs during experimental tuberculosis. Serum from tuberculous guinea pigs examined by paper electrophoresis showed a reaction very similar to those that occurred in man (203): moderate hypoalbuminemia and hyper— globulinemia. However, in contrast to man, a moderate hyperproteinemia accompanied the latter stages of the disease due to an increase in the concentration of all the globulin fractions (85, 175, 176). An increase in the alpha globulins and alpha2 glycoproteins was present within eight days following sub- cutaneous inoculation of guinea pigs with 0.1 mg of M, tuberculosis (200, 201, 202, 204). Hypoalbuminemia was common after the 15th day and hyperglobulinemia after the 19th day. Maximum alteration had occurred in all of the 26 fractions except the gamma globulins of the 15th day. Sher et al. (176) using paper elctr0phoresis showed that the principal mucoprotein alterations occurred in the alpha2 globulin region and that the serum lipids and lipoproteins remained unaltered during tuberculosis in guinea pigs. In spite of this intensive study plasma protein or glycoprotein patterns of diagnostic significance for tuberculosis have not been found. Besides paper electrophoresis, immunoelectrophoresis has been used to study serum protein changes in tuberculous mice and guinea pigs. Williams and Wemysis (210) found an increase in the alpha globulin which was haptoglobin. 2 After 3 weeks post-inoculation all infected mice showed a marked increase in gamma globulin levels and the presence of a low mobility gamma globulin. Hyperalpha -globulinemia was detected by cellulose 2 'acetate electrophoresis in serums from tuberculous guinea pigs 14 days after inoculation (48). Coincident with the hyperalphaglobulinemia was the detection by immunoelectro- phoresis of an additional alpha globulin which increased 2 from sub-detection amounts in normal serums to detectable amounts in serums from tuberculous guinea pigs sometime between the eighth and fourteenth day after inoculation and remained increased until death. The hyperalphaglobulinemia did not occur in guinea pigs which were tuberculin hyper- sensitive but not infected. 27 Starch—gel electrophoresis and disc electrophoresis have not been used to study the serum protein fluctuations during experimental tuberculosis. Serums from cows eXperimentally infected with tuber- culosis have not been analyzed by electrophoretic procedures. However, there was one investigation which employed paper electrophoresis to study serum protein alterations in naturally infected tuberculin positive cows (199). A mild infection produced only small changes in the fractions, but animals with generalized disease showed a hyperalbumenia and a hypogammaglobulinemia. The same fractions were also altered, but to a lesser degree, after the animal was tuberculin tested. MATERIALS AND METHODS Cultures for Inoculation The cultures used to inoculate experimental calves were isolated from cattle, swine and chickens in this labora- tory. Isolants 50B, 51C, and 68C were selected for study because they were isolated from bovine tissues, had Group III culture and biochemical characteristics, but differed in their virulence for laboratory animals. Isolant 50B was from thoracic lymph nodes of a no— gross—lesion, tuberculin positive cow. It caused no lesions in guinea pigs, rabbits or chickens except at the site of injection if injected intradermally. The virulence did not increase with indirect animal passage and it appeared to be identical with P39, a well-studied Group III of human origin. Subsequent studies indicated that all antigens of 50B and P39 were not identical (Flavin, unpublished data). Isolant 51C was from lesions in mesenteric lymph nodes from a tuberculin positive cow. It caused no gross lesions in laboratory animals except at the siteof injec— tion and the regional lymph node when 0.1 mg was injected intradermally. This isolant increased in virulence for guinea pigs and rabbits by indirect passage in guinea pigs but was not equal to that of M, bovis or M, tuberculosis. 28 29 Isolant 68C was obtained from mesenteric lymph nodes from a tuberculin positive cow and its virulence was similar to that of 51C. Besides the Group III organisms, calves were inocu- lated with four strains of M, bovis and three strains of ;M. avium (Table l). The inoculums were prepared from cultures seeded in 1.0 ml quantities of Dubos broth (Difco) containing 0.5% dextrose (no tween 80 or enrichment) and incubated at 35 C for 14-21 days. The supernatant fluid was removed and the cells resuspended in Dubos medium to 10 mg wet weight per ml. Each calf was inoculated intradermally with 0.1 ml unless indicated otherwise. Experimental Animals Serum was obtained from animals that were being used in other experiments. Inasmuch as serum was collected routinely from these animals, the selection of the animals serum to be used were made on the basis of the degree of pathogensity in each animal as judged by the gross and microscopic pathology at slaughter. Serums collected at autopsy from two cows eXperi- mentally infected with Brucella abortus were also examined (Pathology Department, College of Veterinary Medicine, Michigan State University). 30 Table 1. List of animals inoculated with viable mycobacteria Culture No. Source Mycobacterium Route3 Animal No. l . 2 130-0 Cattle M, bov1s ID 6 81-0 Swine M, bovis ID 20 SC 81 310-2 Cattle M, bOVis ID 113,116,128 278-3 Cattle“ M, bovis ID 135 Lab strain Chicken M, avium ID 21 131-4 Chicken M, avium ID 114,115,118, 138 51C-0 Cattle III ID 43,59,60,104 105,106,107, 108,109,127, 136 IU 75,77 A 84,91 68C-0 Cattle III ID 64,137 IU 70,71 A 86 50B-0 Cattle III ID 66,67,126 IU 80 A 88,89 1 The letters following the culture number refer to the tissues from which the strain was isolated: B=thoracic lymph nodes, C=mesenteric lymph nodes. The number following the letters refers to the year in which the culture was iso- lated: 0=l960, l=l961, 2=1962, 3=l963, and 4=l964. 2Where Species is not indicated, organisms are atypical mycobacteria. The Roman numberal indicates the Runyon Group. ' 3ID=intraderma1: 1.0 mg; IU=intrauterine: 1.0 mg; SC=subcutanious: 1.0 mg; and AE=aerosol: 15 m1 of culture medium containing approximately 1 x 108 organisms per ml were sprayed into the aerosol chamber during a one hr expo- sure for each calf. 31 Naturally Infected Animals Serum samples were obtained from 38 cattle in 17 infected herds. Cattle in these herds included gross- and no-gross-lesion reactors, as well as 4 tuberculin-negative cows. In most cases, a sample was obtained before and after tuberculin testing. Single serum samples were collected at autopsy from a cow with actinomycosis, one with brucellosis, one with leptospirosis, and one case of Johnes disease. Normal Animal and Bird Serum A number of normal animal and bird serums (sheep, pig, guinea pig, human, dog, horse, mouse, rabbit, duck, and chicken) were obtained from various projects under investiga- tion in the Department of Microbiology and Public Health, Michigan State University. These serums were used primarily for evaluating the differential disc electrophoresis pro— cedure. Collection 9f Serum from Experimental Calves Blood was collected by venipuncture prior to inocula- tion and at various post-inoculation times. After clotting, serum was removed, distributed in 1 ml amounts, and used immediately or stored at -20(2. Fresh frozen normal serum was examined periodically by cellulose acetate and starch- gel electrophoresis. 32 Rabbit Antiserum Specific for Normal Bovine Serum Portions of the pre—inoculation serums from each of the calves were pooled. alum as described by Proom (145). pared by mixing distilled water The mixture was and centrifuged 12.5 ml of the pooled normal serum, The mixture was precipitated with The inoculums were pre- 40 ml of and 45 m1 of 10% aluminum potassium sulfate. adjusted to pH 6.5 with 5 N sodium hydroxide at 4,000 Xg for five minutes. The s ediment was washed twice with isotonic saline solution containing 0.01% merthiolate and resuspended to 50 ml in saline solu- tion. It was stored at 4C1for no longer than 14 days. Eighteen Dutch Belted rabbits approximately 6 months of age were inoculated with alum precipitated normal bovine serum and bled according to the following schedule: Day 1 14 24 34 48 62 72 77 91 105 115 121 Operation Bled Injected Injected Bled Injected Injected Injected Bled Injected Injected Injected Bled Amount 10.0 H +- O O O O O O O O or-oxoxor-mowor-ow OOOOOOOOOOO on Route Cardiocentesis Intramuscular Intramuscular Cardiocentesis Intramuscular Intramuscular Intramuscular Cardiocentesis Intramuscular Intramuscular Intramuscular Cardiocentesis Blood was obtained from the rabbits by cardiocentesis with a 2% inch 22 gauge needle on a 12 m1 syringe. The blood was allowed to clot in slanted tubes overnight. The serums were decanted, centrifuged at approximately 4,000 X g 33 for 5 min and 3 ml amounts transferred to sterile brucella tubes and stored at -20 C. Each antiserum was tested by immunoelectrophoresis (IE) with normal bovine serum before being pooled. Equal portions of each of the antiserums were pooled. Cellulose Acetate Electrpphoresis Preliminary analyses with a range of buffer systems were used to establish the conditions for subsequent work. The buffer systems employed were: 1. Phosphate buffer pH 7.4, M = 0.15, 12.8 gm Na HPO 2 4 2 2. Tris-EDTA-boric acid buffer (Aronsson and and 2.62 gm NaH2P04° H 0 per liter. Gronwall, l). 3. Tris-EDTA-boric acid discontinuous buffer (Goldberg, 63). 4. Barbital-barbituric acid buffer (Laurell et al., 107). 5. Barbital-acetate buffer (Owen, 133). The latter was used subsequently. The Owen buffer solution pH 8.6, ionicity 0.07, has the following formula: Sodium diethylbarbiturate 5.00 gm Sodium acetate (anhydrous) 3.25 gm Hydrochloric acid (0.IN) 34.20 ml Calcium lactate 0.38 gm Distilled water g, g, 2d, 1,000.00 m1 Electrophoresis was done in a Shandon migration chamber (173) with variable voltage DC power supply (Heath- kit, model PS-3). Cellulose acetate strips (OXOlD, 2.5 X 34 12 cm) was the supporting material. Five ul of serum were applied directly over the cathode on each buffer impregnated strip. It was electrophoresed with a current of 1 ma per strip for 2 hours at 4(2. Following electrophoresis, the strips were stained, dried under weighted absorbent paper, and scanned. Agar gel Electrophoresis Glass microscope slides, l x 3 inch, were covered with 2.5 m1 of a melted one percent agar solution prepared as described under IE. Each agar—covered slide was "aged" at least three hours in a humidified diffusion chamber prior to use. Just before the samples were applied, a circular well, 7 mm in diameter, was cut in the center of the gel 3.5 mm from the cathode end of each slide. The agar plugs were removed by vacuum with a Pasteur pipette. Approximately 5 ul of serum was placed in the well and electrophoresed at 1.25 ma per slide at14<2for 2% hours. Following electropho- resis the slides were immersed in a.5% solution of glacial acetic acid for 30 minutes, dried under filter paper at 37C: and stained. Vertical Starch-gel Electrophoresis Preliminary analyses were with a wide range of buffer systems and gel concentrations were used to establish the conditions for subsequent work. The buffer systems examined were: 35 1. Phosphate buffer pH 7.6 (Ashton, 8). 2. Boric buffer (Smithies, 180). 3. Tris buffer (Raymond, 150). 4. Tris-EDTA—boric acid buffer (Aronsson and Gronwall, 1). 5. Discontinuous buffer (Poulik, 141). 6. Discontinuous buffer (Ashton and Braden, 10). The discontinuous buffer system (10) with a 15% gel concentration was used throughout the investigation. The electrolyte in the buffer compartments was a solution containing 1.2 gm of lithium hydroxide and 11.8 gm of boric acid per liter. The gels were prepared with a buffer made by adding 90 volumes of a solution containing 1.6 gm citric acid and 6.2 gm tris per liter to 10 volumes of electrolyte. The gels were prepared by adding 75 gm of hydrolysed starch (Connaught Medical Research Laboratories, Toronto, Canada) to 500 ml of gel buffer in a 5 liter vacuum flask. The mixture was heated over a flame with constant and vigor- ous swirling and became a viscous or semisolid opaque mass. Heating was continued just short of boiling until the mix- ture changed consistency and was a homogenous, viscous, translucent liquid. After the heating a negative pressure was applied to the flask until the entire contents boiled vigorously for 30 seconds. The vacuum was turned off and the gel slabs prepared by pouring the hot gel solution into 36 the mold (Fig. 1) and covering the surface with the plastic cover containing the slot formers. The cover was weighted to obtain a gel of uniform thickness. After 30 minutes the gel was equilibrated with the well buffer by placing the gel at 4(2for 2 hours. Following equilibration, the cover was removed and approximately 0.1 ml sample inserted into the slot. The exposed surface was covered with warm petroleum jelly (45 C) Which sealed the slots in position and prevented evaporation. The retainer plate (Fig. 2) was placed over the surface of mold. (The bottom edges of the mold rest on a platform pro- truding from the retainer plate, the bottom and front open to allow direct contact with the buffer.) The plate and mold were carefully lifted and placed on the bolts (a) projecting from the front of the electro- phoretor (Fig. 3) the two parts were clamped by wing-nuts on the bolts. The mold holder and buffer chambers (electropho- retor) were one unit for rigidity and to facilitate leveling the gel. A spirit level was mounted at the top center of the retainer plate (Fig. 2). Leveling in all planes was made by adjusting the 3 leveling screws (LS) mounted in the base of the electrophoretor (Fig. 3). The electrodes were 5-mill platinum wire stretched across the back of the buffer chambers. Electrical contact with the electrodes was made by connecting the wire directly to the pole of a co—axial 37 ——-—--‘ ---ua \ Fig. 1. Exploded view of a modified mold for starch gel elec- trophoresis. Removable spacers (a). (Dimensions are in inches and only critical dimensions illustrated.) 38 k-- 99‘ DH 0 O i 13% o 'J : :_L_, a A -- .w m """"""" I: V 1 I : Li M. b M 1% 3w A B c Fig. 2. Retainer plate. A: front view. B: cross section of gel mold. C: cross section of retainer plate. Spirit level (L). (Dimensions in inches and only critical dimensions illustrated.) 39 Fig. 3. Modified gel-mold holder and buffer compartments (electropho— retor) for starch gel electrophoresis. Schematic representa— tion showing only the important dimensions. Bolts (a) for clamping the retainer plate in place. Co—axial terminals (b). (Dimensions in inches.) 4O connector (b). The upper buffer chamber has a capacity of 2,500 ml. The lower chamber was slightly larger holding approximately 2,750 m1. Current was provided by a variable voltage DC power supply (Heathkit model PS-3). After completion of electrophoresis, the gel mold was removed from the holder and the petroleum jelly scraped from the gel surface. Saran wrap (Dow Chemical, Midland, Michigan) was placed on the gel surface to give support to the gel as it is lifted from the mold and placed in the slicer-holder (Fig. 4b). The base of the gel cutter (Fig. 4) consists of a heavy metal plate with two metal bars (a) accurately spaced to permit the gel holder (b) to slide between them. A fine piano wire (approx. 0.008 inch in die.) was drawn taut across notches cut in the metal bars. The width of the gel holder (b) is adjusted by two 1.5 mm thick spacers to corre- spond to the gel width. The base was clamped to the work bench for additional support. A % inch thick glass plate was placed on top of the Saran wrap. The holder (b) guided by the metal bars and plexiglass runners (c), was pushed with one quick movement into the wire (W). By moving the wire to a different notch (4.5 mm, 3 mm, and 1.5 mm), four 1.5 mm slices could be obtained from a 6 mm gel. Each slice was easily removed by aid of the Saran wrap. If the wire was clean and taut and if the gel was cut with one quick movement, all slices were smooth and of uniform thickness. 41 Fig. 4. Slicing device. Metal bars (a). The notches in the middle of the two metal bars are for varying the thickness of the slice. Plexiglass gel holder (b). The gel holder is guided by (a) and by the plexiglass runners (c). 0.008 inch piano wire (W). (Dimensions are in inches and only critical dimensions illustrated.) 42 The sliced gel was lifted with the Saran wrap as support and placed into the bottom half of a framed-plastic- screen-holder (Fig. 5). Each frame was made of % inch plexi- glass and plastic screen mounted on one side. The internal dimensions were 9 11/16 inches in length and 6 1/4 inches in width; the outer dimensions were 10 3/4 inches X 9 1/4 inches. Both halves were placed screen side down so that the space between screens was approximately % inch, and the halves were held together with fiberglass U-clamps placed in grooves at each side. The screen gel was lifted by aid of a spring clamp and placed into the compartment of the staining tank (Fig. 6) containing the protein stain, (1% Amido Black 10 B in 5% acetic acid). Each individual compartment was slightly longer and wider, in order to hold four screens. The illustrated tank with 10 compartments measures 27 3/4 inches X 12 1/2 inches X 10 1/2 inches and each compartment had a drainage valve, so that when a stain or solution became exhausted it could be removed without disturbing the other compartments. The gel holder was removed from the stain, washed with running tap water to remove excess stain and placed into the destainer (Fig. 7) that was especially designed to be coupled with the convenient procedure for staining gel matrices. It has 2 compartments so that two gel slabs could be destained at one time. Destaining was done at 4 amps. for 30-45 min with continuous circulation of the destaining solution (10% glycerol in 5% acetic acid). Fig. 5. 43 Plastic-screen-gel-holder. A: gel holder (H), disc gel adaptor (D) and fiber glass clips (fg). B: dia— gramatic illustration showing how gel holder (H) is assembled for starch-gel slabs (SG) and disc gel columns (dc). Note that the disc gel adaptor is not used for starch—gel slabs. 44 ' " ' Q;~:"‘. o'vgttt‘q‘uq’ 3w ... . _ - Qt..q_..1.-‘ . - Fig. 6. Multiple staining tank. “Tl : Fig. 7. n 179 u 1+2} l r ‘0 l— C A: destaining assembly. Destainer (d), 11% x 10% x 2% in. i.d. (bc) 6—12 volt battery charger. The destaining solu— tion circulates through charcoal bed (cb) by pump (p). B: Saggital section through destainer. Gel holder (H) placed in slot (5) so cathode is % and anode % in. from gel (9). C: Filter unit (13 x 5% x % in. o.d.). 8 in. charcoal bed (cb) holds 2 lbs granular charcoal, held in place by two % in. thick glass wool filter pads (W) held in place by filter holder (fh) mounted on back of screw-on cover (t). Electrodes (e), inlet (1), plastic screen (n), and outlet (0). 46 Two—Dimensional Starchfigel Electrophoresis Two-dimensional starch-gel electrophoresis was per- formed by the method described by Poulick and Smithies (143). A. First stage: Filter paper electrophoresis. This stage of the separation was performed on 3 x 15 cm strips of Whatman No. 3 filter paper using a buffer of the following composition: 0.048 M sodium acetate, 0.048 M sodium barbital, and 0.0073 M hydrochloric acid. The strips were impregnated in this buffer and placed in the Shandon migration chamber (Colab). Approximately 0.05-0.l m1 of serum was applied to each strip with a micropipette. Elec- trophoresis was done at 120 volts for 12 hours. After elec- trophoresis one of the strips was stained as a guide to the positions of the protein zones on the unstained strips. In a few cases, 5 cm x 15 cm cellulose acetate strips were used in place of filter paper. Owen buffer, as described for cellulose acetate electrophoresis was substi- tuted for the above buffer. B. Second stage: Starch-gel electrophoresis. A starch slab was prepared as described above, except that a cover without the slot former was used. The gel mold was adjusted to give the following dimensions: 16 cm long, 12 cm wide and 0.6 cm thick. A cut was made across the full width of the gel about 5 cm from the cathode end. A second cut was made 47 above, parallel to and 3 cm from the first and a section 1 cm wide was removed. The sample slot was formed by dis— placing the gel into the space left by the removed section. Using the stained filter paper strip as a guide a 12 cm x 6 mm section was cut from the middle of an unstained strip. This strip was inserted into the slot in the gel, the displaced gel pushed back against the strip and the removed section replaced. The gel was covered with melted petroleum jelly and electrOphoresis performed under the same conditions as for one dimensional starch-gel electrophoresis. After electrophoresis, the gel was not sliced but stained directly, and destained as described for starch—gel electrophoresis. Immunoelectrpphoresis Immunoelectrophoresis was employed according to the technique of Hirschfeld (80). The agar was prepared as fol- lows. A two percent solution of Difco Bacto agar was poured into a flat dish to a.depth of 1 cm and allowed to solidify. The gel was cut into 1 cm cubes and washed in running tap water for 72 hours followed by frequent changes of distilled water for an additional 72 hours. The agar cubes were stored in distilled water at room temperature. .All of the agar used in this investigation was prepared in a single batch. 48 A barbital buffer system was employed (80): Well buffer Agar buffer (pH 8.6, I=0.06) (pH 8.6, I=0.093) Diethylbarbituric acid 1.38 gm 1.66 gm Sodium diethylbarbiturate 8.76 gm 10.51 gm Calcium lactate 0.38 gm 1.54 gm Distilled water q. p. pp. 1000 ml 1000 ml Two parts of the agar buffer was mixed with one part distilled water and the mixture heated in a steamer. An equal volume of freshly melted agar was added with merthio— late in a final concentration of 1:10,000. While the solu- tion was warm, 2% ml were spread evenly over the surface of thoroughly washed microscope slides and allowed to solidify. Each agar-covered slide was "aged" at least 3 hours in a humidified diffusion chamber prior to use. A l x 66 mm antiserum trought equidistant between 2 antigen wells, 2 mm in diameter, 35 mm from the cathode end and 5 mm from each edge were cut in the agar with a LKB punch, type 6808A (112). The agar plugs in the antigen wells were removed by vacuum with a Pasteur pipette. The plug in the antibody trough was removed after electrophore- sis was completed. The slides were supported in a Shandon migration cham- ber as modified for IE by Dardas and Mallman (48). A 6 x 8% inch plexiglass plate was placed across the bridge supports. Sixteen slides were used at one time; eight on each side. Electrical contact between the slides and buffer chamber, and between slides were made with filter paper impregnated 49 with well buffer. Current was supplied by a variable voltage DC power supply (Heathkit, model PS-3). Electropho- resis was carried out at 4C for 2% hours at 1.25 ma per slide. Following electrophoretic separation the agar plugs in the antiserum trough were removed and approximately 0.25 ml of antiserum added. The patterns were allowed to develop for 18 to 24 hours at 28C2in a humidified diffusion chamber. The slides were washed for 48 hours in repeated changes of saline solution and in repeated changes of dis- tilled water for 24 hours. The slides were dried at 37c: under a single moistened piece of bibulous paper and stained. Starch-gel Immunoelectrophoresis Twenty-five milliliters of buffered agar (same as for agar and IE) was spread evenly over the surface of an 8 x 18 cm glass plate forming a layer 3 mm thick. The serum specimen was separated in starch-gel as above. The gel was then sliced longitudinally into 2 parts; one was stained for protein and the other placed cut side down on the agar layer. The agar plate was then placed into a humidified diffusion chamber to preincubate for 12 hours. Then two troughs, 2 mm 'wide, were cut in the agar 1 cm from each edge of the starch- gel slice. The troughs were filled with the antiserum and the pattern allowed to develop for 3-5 days at 28c: in a humidified diffusion chamber. 50 The starch-gel slice was removed and the plate washed and dried. The immunoprecipitates were stained and recorded. Continuous Flow Electrophoresis Serum was diluted 1:4 with 0.02 M barbital buffer and poured into the feed-device of the Spinco Continuous- IFlow Paper Electrophoresis Cell. The apparatus was pre- ;pared according to the directions of the manufacturer (16, 120) and the flow rate of the sample was fixed at 4 ml per 'hour. A constant current of 50 ma requiring approximately 550 v was maintained by a Spinco Constat. The apparatus was cooled in a walk-in refrigerator. The buffer (barbital, pH 8.6, 0.02 ionic strength) was recirculated continuously and only the condensation buffer was discarded during the run. The collection of the fractions into 32 tubes was begun with the appearance of the yellowish-stained albumin fraction at the serrated edges of the lower curtain. ,At the end of each run, the lower curtain was removed immediately from the cell and, after drying in an oven at 120-130 C for 30 min, stained with Pontceau S or bromphenol blue to reveal the position of the different serum fractions. The tubes containing the same fractions were pooled, concentrated by prevaporization, and each fraction adjusted to twice its amount in the original serum. The separated fractions were examined for purity by cellulose—acetate electrophoresis. 51 Polyacrylamide Electrophoresis Disc electrophoresis. The apparatus and technique ‘were employed according to the earlier version of the disc electrophoresis method as reported in detail by Ornstein and IDaVis (132). The latest modification of the stock and work- ;ing solutions (49) were employed (Tables 2 and 3). Six ul of serum were added to 0.14 ml of sample gel, 1:he mixture layered on top of the spacer gel, photopolymer- :ized for 30 minutes and the cylinders placed in the upper lauffer compartment. Electrophoresis was done at 4 ma/tube eat room temperature until the tracking dye reached the bot- ‘tom end of the tube (approximately 1 hour). Current was :supplied by a 5,000 volt, 200 ma DC power supply (Canadian lResearch Institute Model EP-5K/200). Differential disc electrophoresis. The stock and snorking solutions for the spacer—sample gel and electropho- retic procedure. The only difference in reagents was the following two acrylamide solutions: 1. Acrylamide stock solution No. I for the 10% working solution: 40 gm acrylamide 0.4 gm BIS Dissolve and bring to 100 ml with distilled water. 2. Acrylamide stock solution No. II for the 4.75% working solution: 19 gm acrylamide 0.4 gm BIS Dissolve and bring to 100 ml with distilled water. 52 Table 2. .Stock solutions (A) (B) 1 l N HCL 48 ml 1 N HCL approximately 48 ml TRIS 36.6 gm TRIS 5.98 gm TEMED 0.23 ml TEMED 0.46 ml DHOH to 100 ml DHOH to 100 m1 (pH 8.9) (pH 6.7) (C) (D) .Acrylamide 28.0 gm Acrylamide 10.0 gm BIS 0.735 gm BIS 2.5 gm DHOH to 100 m1 DHOH to 100 m1 (E) (F) Riboflavin 4 mg Sucrose 40 gm DHOH to 100 ml DHOH to 100 ml lpH adjusted by titrating with l N HCl. TRIS=tris(hydroxymethyl)aminomethane TEMED=N,N,N',N',-tetramethylethylenediamine BIS=N,N'-methylenebisacrylamide DHOH=distilled water. Table 3. WOrking solutions Small-pore Small-pore Stock buffer Large-pore solution for solution #1 solution #2 solution reservoirs 1 part A Ammonium 1 part B TRIS 6.0 gm 2 parts C persulfate 2 parts D Glycine 28.8 gm 1 part DHOH 0.14 gm 1 part B DHOH to 1 liter pH 8.9 4 parts F pH 8.3 (8.8-9.0) pH 6.7 (6.6-6.8) lDiluted 1:10 before use. 53 Pyrex tubes (6 mm i. d. x 12 cm) were marked at 11 cm, 7.5 cm, 4 cm, and 2 cm starting from the base end (Fig. 8). Prior to use, the tubes were cleaned in a deter— gent solution. The inner walls thoroughly cleaned with a cotton tipped applicator stick, rinsed thoroughly in tap water, then in distilled water and stored in a solution consisting 1-200 parts Kodak Photo Flo solution in distilled water. The tubes were drained and allowed to dry. Then they were inserted vertically into vaccine-bottle stoppers (49), and filled to the 7.5 cm mark with the 10% acrylamide solution. The 4.75% acrylamide solution was carefully added to the 4 cm mark. A thin layer of water was carefully added by means of a 1 ml tuberculin syringe with a 25 gauge needle, and allowed to polymerize for 40 min. The water layer was shaken off and rinsed with spacer gel. Spacer gel was added to the 2 cm mark, water layered and photo-polymerized for 15 min. Sample gel, prepared by mixing 8 ul of fresh serum with 0.4 ml of spacer gel was added and photo-polymerized for 30 min. Electrophoresis was started at 2 ma/tube until the marker dye reached the separating gel (4 cm mark), approx- imately 20 min. The current was adjusted to 4 ma/tube until the albumin reached the 11 cm mark, approximately 1% hours. Following electrophoresis, the gel columns were removed from the glass tubes after placing in ice water for 5 min by rimming with a finely tapered steel needle. 54 Sample Gel 2cm Spacer Gel 4cm 4.5% Gel ’ 7.5 cm 10 % Gel llcm___. Fig. 8. Diagramatic representation of the position of the gels in the differential disc electrophoretic column. 55 The gel columns were stained and destained with a modified apparatus designed for starch and acrylamide gel slabs (Figs. 5, 6, 7). The disc gel adapter (H), which has 12 slots (Fig. 5), was placed into the bottom half of the gel holder (D) and the gels inserted into the slots. The top half of the gel holder was then connected to the bottom half by two fiberglass clips. The gel holder was placed in the compartment of the staining tank (Fig. 6) containing 0.5% Amido Black 10 B in 5% acetic acid. The gels were stained for 30 min at room temperature. The gel holder was removed from the stain, washed with running tap water to remove excess stain and placed in the destainer (Fig. 7). Sixteen to 24 gels were destained in 30-45 min at 4 amps. Following destaining the gels were photographed and scanned. Protein, Lipoprotein, and Glycoprotein Stains Cellulose acetate. Pontceau S, Nigrosin, Light green SF, Coomassie brilliant blue, Lissamine green and Procion Brilliant Blue M-RS were tested comparatively for protein staining on cellulose acetate. Pontceau S was selected for routine serum protein staining. Cellulose acetate strips were immersed for 10 min in a 0.2% solution of Pontceau S in 3% trichleroacetic and differentiated in 2% glacial acetic acid. 56 Lipoproteins were separated on cellulose acetate electrophoretically according to Gelman (58). Five parts of serum were mixed with 1 part of the following Sudan Black B staining solution: 0.1 gm of Sudan Black B was dissolved in 1 ml of ethyl acetate and 9 ml of propane glycol added, mixed thoroughly and centrifuged at 4,000 X g for 5 min. The serum-staining solution mixture was allowed to stand 3 hours and centrifuged at 4,000 X g for 5 min. Twenty-five microliters of the mixture was applied to the cellulose acetate strip directly over the cathode. Electrophoresis was done as described on pages 33 and 34. Glycoproteins and/or mucoproteins were stained by the periodic acid Schiff (PAS) method as outlined by Gelman (58). After electrophoresing a sample in the usual way, the strips were immersed in 0.5% periodic acid for 5 min, washed twice with distilled water and placed in Schiff reagent at room temperature for 10 min. The strips were washed 3 times in 0.5% sodium metabisulfite, washed in tap water 5-10 minutes and dried at room temperature under weighted absorbent filter paper. Agar gel and immunoelectrophoresis. Proteins on dried agar following electrophoresis and immunoelectropho- resis in agar gel were stained with Crowle's (46) Triple Stain: 57 Thiazine red R 0.1 gm Amidoswarz 10 B 0.1 gm Light green SF 0.1 gm Acetic acid 2.0 gm Mercuric chloride 0.1 gm Distilled water 100.0 ml The slides were immersed solution for 5 min, differentiated in 2% acetic acid, washed in tap water, and dried. Following electrophoresis and IE in agar gel, lipo- protein and glycoproteins were stained according to the methods described by Gabar (64). Lipoproteins were stained by immersing the dried agar slide in a saturated solution of Oil Red 0 (37C) in 60% ethanol. Slides were differentiated in 50% ethanol, washed in tap water, and dried. Glycopro- teins were stained by the PAS method of Uriel and Grabar (194), according to Crowle (46): A. Solutions a. Schiff reagent 1.5 gm basic fuchsin 500 ml boiling distilled water Filter at 55C and cool to 40C. Add 25 ml 2 N hydrochloric acid 3.75 gm NaZS 05 Agitate to easure rapid solution. Allow to stand stoppered in a refrigerator for 6 hours. Add 1.2 gm animal charcoal Mix vigorously for 50 seconds and filter rapidly, filtration time not exceeding 2-3 min. Store stoppered in refrigerator. b. 1.0 gm periodic acid 0.82 gm anhydrous sodium acetate 100 ml distilled water 58 c. 0.54 gm acetic acid 0.89 gm anhydrous sodium acetate 10 gm hydroxylamine hydrochloride 100 ml distilled water d. 5 ml 10% Na S O 5 ml 2 N hydrgcgloric acid 90 ml distilled water Solution should be made just before it is used. This also serves as the final wash bath when it contains; 20% glycerol. B. Procedure The pre-dried agar slides were placed in c for 15 min, washed in running water for 15 min, placed in b for 10 min, washed in running water for 10 min, and placed in Shiff reagent, diluted immediately before use with an equal volume of distilled water, for 3 min. The slides were then washed 3 times for 2 min each in d, 3 times for 1 hour each in glycerinated d, dried and recorded. Starch-gel. Nigrosin, Amidoschwarz, and Wool Fast Blue BL (Natural Aniline) were tested comparatively for staining proteins. A 0.5%,solution of amidoschwarz in 5% acetic acid was selected for routine protein staining. The gel slabs were stained for 10 min at room temperature. Polyacrylamide gel. A.0.5% solution of amidoschwarz in 5% acetic acid and filtered twice, was used routinely for protein staining. The gel columns were stained for 30 min at room temperature. The detection of lipoproteins was investigated by two methods: prestaining of the serum lipoproteins as described for cellulose acetate or by immersing the gel column, following electrophoresis in a saturated solution 59 of Oil Red 0 (37CU in 30% ethanol and differentiating in 20% alcohol. The former method gave better results and was used routinely. Glycoproteins were stained according to the PSA method described by Canalco (35). After electrophoresing the sample in the usual way the gel column was immersed in 7.5% acetic acid at room temperature for 1 hour. The column was transferred.to 0.5% solution of periodic acid (PA) solu- tion at 4 C for 1 hour and the PA moved electrophoretically in 7.5% acetic acid for 1 hour. Then the columns were placed in Schiff reagent at 4 C until bands appeared. The columns were stored indefinitely in fresh Schiff reagent. Continuous flow filter paper curtain. Immediately following electrophoresis the curtain was dried in 120-130 C oven for 30 min. The dried curtain was stained by immersing in 0.2% Pontceau S in 3% TCA, differentiated in 5%.acetic acid and dried at 37 C. Densitometric Evaluation pf Electrophorogpams All scanning was done with the N. I. L.—Joyce Chromascan, a double-beam recording and integrating micro— densitometer, incorporating facilities for scanning by transmission and reflectance. The four variables: (a) the kind of light filter used, (b) the slit width of the apera- ture, (c) the slope of the grey wedge or cam, and (d) the gear ratio of sample movement to recording pen were used to 60 obtain maximum sensitivity as follows: (1) cellulose acetate: no filter, 1:1 gear ratio, 400 cam, a slit width of 1/20 mm, and scanned by reflectance; (2) starch-gpl: (reflectance) no filter, 1:1 gear ratio, 900 cam, and a slit width of 1/24 mm; (transmission) no filter, 1:1 gear ratio, 900 cam, a 1 mm slit width in front of light source and a 1/24 mm slit width in front of transmission photo- multiplier; (3) polyacrylamide gel columns: no filter, 1:1 gear ratio, 90°cam, a 1 mm slit width in front of light source, a 1/24 mm slit width in front of the transmission photomultiplier and scanned by transmission. Photogpaphy A number of 35 mm films were tested for photography of starch—gel slabs and polyacrylamide gel columns: (Kodak High Contrast Copy, Kodak Plus X, Kodak Infrared, Kodak Panatomic X, Ansco Versapan and Adox KB-l4. ,Kodak Plus X and Ansco Versapan were used during the entire investigation. A.Lieca 35 mm camera (model M3) with a 90 mm Elmarit lens was mounted on a specially constructed camera stand. The specimen to be photographed was placed on the surface of the plate glass top. A flashed Opal glass plate was mounted 10 inches below the upper clear plate glass and twenty-four 50 watt daylight type light bulbs were mounted underneath so that the gels could be photographed by transmitted light. 61 Starch—gel slabs were placed on the glass top, all air bubbles removed, and photographed with Ansco Versapan for 0.1 and 0.2 seconds at f/ll with a Lieca red filter. The film was developed in Ansco Hyfinol for 4 min. Polyacrylamide gel columns were placed in unetched test tubes, filled with 5%,glacial acetic acid and stoppered. The tubes were arranged on the surface of the upper glass plate and photographed with Kodak Plus X for 0.125 sec at f/4. The film was developed with Kodak D-76 develOper, full strength, for 10 min. Immunoelectrophoretic slides were placed in the negative carrier of the enlarger and medalist single weight F-4 enlargement paper was exposed and developed in Kodak Dektol. RESULTS Effect pf Freezing pp_Normal Bovine Serum Proteins Freezing and thawing a serum twice did not produce any changes detectable by cellulose acetate or starch-gel electrophoresis. Examination pf Rabbit Anti-Normal Bovine Serums No detectable anti-albumin antibodies were found in the antiserums from 3 of 18 rabbits and 2 failed to pro- duce anti-gamma globulin after 3 immunization schedules. Antiserums collected 10 days after the last intraperitoneal inoculation of normal serum were unsuitable for use because a large number of anti-globulins were no longer detectable. Only those antiserums collected between 5-7 days after the last intraperitoneal inoculation and which contained all the anti-normal bovine serum components were pooled. Fractionation 9£_Normal Bovine Serum .21 Continuous Flow ElectrOphoresis Normal bovine serum proteins were separated into four main components by continuous flow electrOphoresis (Fig. 9). Fractions 9 and 10 contained gamma globulin (Fig. 10) by cellulose acetate electrophoresis of the 62 5”“? ,4 .4 ' r~.v..vu¢ ..- or a,“ v. C. b“ ”0.“..0»: ..., : mIn..“.¢ 1’10: .I av. .. o a - .3. no 19‘ oh in electrophorogram.of normal bovine serum Paper curta o 9. Fig 15. by continuous flow electrophores The numbers at the bottom of the curtain represent fraction numbers. ion after separat A. Normal serum. B. Concentrated sample from fractions 9 and 10. C. Concentrated sample from fractions 11, 12, 13, and 14. D. Concentrated sample from fraction 15. E. Concentrated sample from fraction 16. F. Concentrated sample from fractions 17 andl8. G. Concentrated sample from fraction 19. H. Concentrated sample from fractions 20. I. Concentrated sample from fractions 21 and 22. J. Concentrated sample from fractions 23 and 24. Fig. 10. Cellulose acetate electrophorograms of continuous flow electrophoretic effluent serum fractions. 65 concentrated effluent. Fractions ll, 12, l3, l4, and 15 contained both gamma and beta globulins. Fractions 16 con— tained only beta globulin. Beta and alpha globulins were in fractions 17 and 18. Albumin, and alpha and beta glob— ulins were in fraction 19. Fraction 20 contained alpha globulin in high concentration and a low concentration of albumin. A trace of alpha globulin and an increasing con- centration of albumin were in fractions 21 and 22; and fractions 23 and 24 contained only albumin. Cellulose Acetate Electrophoresis .9: Normal Bovine Serum Normal bovine serum was separated into 6 fractions with the tris-EDTA-boric acid buffer system (1): prealbumin, albumin, 2 alpha globulins, beta globulin and gamma globulin but the pattern of any given serum was not reproducible. The only buffer system with which reproducible patterns were obtained was the continuous barbital-acetate buffer system (133). Normal bovine serum was separated into 4 major major fractions by electrophoresis for 2 hours with this system. These included albumin, and alpha, beta, and gamma globulins. An additional gamma component (gammal) was separated in many serums electrophoresed under the same conditions. Five ul of human serum separated into 5 frac- tions by the system: albumin, 2 alpha globulins, beta ‘ globulin and gamma globulin (Fig. 11). 66 Alb .. Bovine L+ m Fig. 11. Typical densitometric recordings of cellulose acetate electrophorograms after electrophoresis with the Owen buffer system. A: protein. B: protein-bound carbohydrate. C: protein- bound lipid. 67 Stained normal bovine serum after cellulose acetate electrophorograms contained one lipoprotein and l glycopro— tein fraction. Protein-bound lipids were in low concentra- tion with the major component migrating in the beta region (Fig. 11). One great glyCOprotein fraction was in the beta region. Agar Gel Electrophoresis pf Normal Bovine Serums Four major fractions in normal bovine serum were detected in agar gel electrophoresis (Fig. 12) using either the continuous barbital-acetate system (133) or the discon- tinuous veronal buffer system (80). Neither separation nor resolution was better than that obtained by cellulose ace- tate electrophoresis. Starch-gel Electrophoresis pf Normal Bovine Serum Sixteen to 20 fractions were routinely separated with a 15%.gel concentration using the Ashton discontinuous buffer system (10). This system yielded the clearest and most consistent separation of the normal serum proteins. Separation of serum proteins with the other buffer systems were unsatisfactory. Figure 13 is a composite, schematic representation of the relationships between protein zones following single dimensional cellulose acetate, starch-gel, and two-dimensional electrophoresis. The zones in starch-gel were assigned 68 Fig. 12. Electrophorogram of normal bovine serum in agar gel with the Owen buffer system. (1) albumin, (2) alpha globulin, (3) beta globulin, (4) gamma globulin. Fig. 13. 69 // 7/////////// ollulose cetate Single dimensional starch gel, cellulose acetate, and two dimensional electrophoretic patterns of normal bovine serum. (The cross hatched areas represent lightly stained bands.) 70 numbers according to the mobility in the gel (with number 1 for the slowest moving). Two additional zones were detected in the bwo dimensional analyses in which the serum proteins were electrophoresed on cellulose acetate and then placed in a starch-gel for electrophoresis at a right angle to the original direction of migration. The zones resolved by this procedure were classified according to their electrophoretic mobility in cellulose acetate. Further definition was depen- dent upon the isolation and characterization of the individ- ual zones. Band 1 was a light staining zone of gamma-globulin. Some serum samples contained 2 proteins which migrated only slightly toward the anode as discrete bands. Band 2 was not consistently detected and was a part of the gamma globulin. ’ Band 3 stained intensely and corresponded to the slow moving alpha globulin described for human serum (180). 2 Band 4 migrated as a gamma globulin. Bands 5, 6a, b, and c, 7, and 8 migrated as beta globulins. Three to six beta globulins were in this zone. The number of bands varied with breed of the animal (7, 8, 9). Band 5 was present in varying concentrations in all serum samples except serums from Guernsey cattle. Bands 6a, b, and c were always present, always intensely stained and were the major beta bands in serum from Guernsey cattle. 71 Holstein or Holstein-Angus cattle serums contained band 7 in addition to 6a, b, and c. Band 8 was lightly stained and variable in serum from Jersey cattle. Bands 9 to 16 were classified as postalbumin bands. Band 9 contained alphal and alpha globulin. Bands 10, ll, 2 12, and 16 had only alphal globulins while bands 13, 14, and 15 migrated as alpha globulins. Bands 10, l3, l4, and 16 2 stained intensely. Bands ll, 12, and 15 were not always present. Band 16 was frequently marked by the albumin band. Band 17 corresponded to the albumin zone on cellulose acetate. Band 18 was the only prealbumin demonstrable and migrated as an alphal globulin. Immunoelectrpphoresis pf Normal Bovine Serum Thirty-two antigenic components were detected in normal bovine serum by IE: one pre—albumin, one albumin, 6 alphal globulins, 13 alpha globulins, 4 betal globulins, 2 5 beta globulins, 1 beta globulin and 1 gamma globulin 3 (Fig. 14). The 32 immunoprecipitates were demonstrated by the use of several antiserums. The maximum number of anti- genic components which could be observed by the use of a single antiserum was 28. These precipitin arcs have been labeled according to their mobility in the whole serum pattern and in the individual fractions obtained by 72 M5560 ml mumm (U u o (I) muom comm muom muom muom muom muom V' Ml NI HI! fil Ml NI v—lI-lv—GNNNNM .Nm .Hm .om .mN .mN .hN .ON .mN .vN .mm Hmuom mamaa manna steam madam madam erode anamrmam oummsmae mummrma< a- armaa HI maum «Hum an-” OHIN mIN m. m ha nu an —N em «a 08 @— .NN .HN .ON .ma .mH .hH .mH .mH .VH .MH .NH MI ,Nl N N N H H H mamaa mamam arma< arma< madam enamsmaa muamrdaa unamrmam H: madam caesna< Hl 0' ml GHESQHon h— 0— #— «— HmEuoc mo mcuouumm UfiuouosmouuooaoOGSEfifl ummm wo mcH3MH© ouamomfioo a .va .mflm 73 continuous flow electrophoresis (page 50). The alphal globulin with the greatest mobility was designated alphal-l (al-l) and so forth. The prealbumin and al-l and al-5 areas were often difficult to observe with the protein stains whereas the albumin and the other four a -globu1ins were readily and 1 consistently seen. The al-l arc stained poorly for lipid. The use of the concentrated effluent fractions obtained from continuous flow electrophoresis, and antiserum from rabbits which produced no detectable antialbumin antibodies facil- itated the identification of the alphal and alpha2 regions. The alpha zone had 5 precipitin arcs. Alpha2-13 2 stained for lipid. The intensity of the lipoprotein stain of this arc was markedly greater than the al-l arc. In addi- tion, there was a strong Oil Red 0 stained area lying behind the a2-13 lipoprotein and immediately adjacent to the antigen well. This lipid spot, which has none of the characteristics of a precipitation band, corresponds to a similar fraction in human and sheep serums (40). One strong glycoprotein (aZ-l) and 6 weak glycopro— -2, a -9, a —11, beta teins (albumin, -2 (bl-2), and b2-4) a2 2 2 1 were stained with the Schiff reagent. None of the beta globulins except b1-2 and b2-4 have been identified. If serum contained hemoglobin the hemo— globin would migrate in the betal zone. Except at the 74 cathodic end, the beta —5 precipitate was often obscured by 2 the broad gamma globulin precipitate and resembled a biphasic precipitin curve. The beta3 arc was not consistently detect— able. The gamma globulin precipitate formed a continuous arc from the alpha2 zone to the most cathodic end of the slide. Starch—gel Immunoelectrpphoresis p£_Normal Bovine Serum Many of the bands, especially in the post albumin region, obtained by electrophoresis in starch-gel consisted of multiple antigenic components when analyzed by immuno- diffusion in agar after electr0phoretic separation in starch— gel (Fig. 15). Many of these bands could not be identified with specific precipitin arcs. Individual bands cut from the gel and placed on agar did not resolve the exact anti- genic moeity because of trailing in the highly concentrated bands, especially gamma, beta, and albumin bands. Not as many antigenic components were detected as by agar immunoelectrophoresis. The number of immunoprecip— itates was generally the same as or less than the number of bands separated in starch-gel. It was necessary to preincu- bate the starch-gel 12—18 hours before adding the antiserum to assure maximum diffusion of the antigens into the agar gel. This did not improve on the number of arcs but pre- incubation less than 12 hours resulted in a further decrease in precipitin arcs. 75 Gamma 8:12 Beta Alphaz Ilp'lll‘ Allin in Pr lb l m1 ILILLl 0°— - 111111 0° -- cou— Fig. 15. Starch gel immunoelectrophoretic pattern of normal bovine serum. (The cross hatched areas represent lightly stained bands.) 76 Polyacrylamide Gel Electrophoresis p§_Norma1 Bovine Serum Proteins The conventional disc electrophoretic procedure (132) yielded poor separation and resolution, and was unsatisfac- tory for the fractionation of bovine serum proteins. The differential disc electrophoretic procedure increased the separation and resolution of human and bovine serum proteins (Fig. l6, l7, 18). Human serum separated into 44-50 components with this procedure; into only 22-28 , by the conventional procedure (Fig. 16). With the conven- tional system, bovine serum separated into 14-22 bands with poor resolution; into over 30 bands by the differential pro- cedure (Fig. 17). Because of the better resolution, the corresponding densitometric tracings (Fig. 18) were markedly improved, particularly the post albumin region of bovine serum. Ten glycoprotein fractions in normal bovine serum were stained with Schiff reagent. The more deeply stained bands were in the gamma, slow alpha and beta globulins and 2 in the cathodic end of the albumin band. Only 4 glycopro- teins were detected by the conventional procedure. The pre- staining procedure for protein-bound lipids was unsatisfac- tory. Two weak lipoproteins and 3 major lipoproteins were stained with Oil Red 0. Two of the major lipoproteins were in the gamma zone and one corresponded to the slow alpha 2 band. The carbohydrate- and lipid-bound protein bands are Fig. 16° WINNIE; I§ fi g 77 Comparison of the separation of human serum in standard 7% gel (left), 4.75% gel (middle), and differential gel (right). Cross hatched areas and dotted lines indicate lightly stained and fine bands, respectively. (0) origin of the 10% separa— tion gel, (a) 4.75% gel, (b) 10% gel. Spacer gel not shown. Fig. 17. 78' _ 'IIIIII/ 'II I . is I lllfl if new Comparison of the separation of normal bovine serum in standard 7% gel (left), 4.75%.ge1 (middle), and differential gel (right). Cross hatched areas and dotted lines indicate lightly stained and fine bands, respectively. (0) origin of the 10% separation gel, (a) 4.75%.gel, (b) 10% gel. Spacer gel not shown. Fig. 18. 79 Densitometric tracing of the normal bovine serum separated by the differential disc electrophoretic method illustrated in Figure 17. The glycoprotein fractions (c) and lipoprotein fractions (1) are also indicated. (Saz) slow alpha2 lip0protein. 80 presented in Fig. 18. Densitometric analyses of these spe- cific stained bands was unsatisfactory. A 4.75% gel in a standard tube (6.5 cm long) afford— ed better separation than the standard procedure (7%.gel), especially of human serum, which increased by 5-8 more bands. Figure 16 shows 22 bands separated by the standard gel (7%), 33 by the 4.75% gel and 50 by the differential gel. The number of bands, however, did not increase if the differen- tial tubes (12 cm long) were filled only with 7%, 4.75%, or 10% separating gels and electrophoresed under the same con— ditions described for the differential system. Because of diffusion the post albumin bands were extremely fuzzy (Fig. 19). Reproducible results were obtained when 16 tubes containing the same sample were run simultaneously (Fig. 20). The tubes in Fig. 20 were picked at random from a 16 column run. Figure 21 illustrates variation in protein fractions found in 8 individual bovine serums, and also demonstrates the over-all reproducibility of the distance migrated by the albumin band in a single run. The superiority of this system was clearly demon- strated with serums collected before inoculatibn (initial) and prior to aut0psy (post-inoculation) from a cow experi- mentally infected with Mycobacterium bovis (Fig. 22). Fig. _ _ A . ' .‘.-.‘.Al..i'-4l 0‘" -"~"' . 1--.--.--.-— n -' .- N , . r _". .". o‘_ 19. .. .. .. . ~~.; gi, 33. .3 R," ;%' ' .o' ‘3‘. .... .... ‘2: .g ' 3' "I " ..f ' E ii . .l“ ,a .a; .g . } ,_._ _‘ ,3. ' ‘1’ '0“ 1.- “ '0‘. .5» -.[. I ' ‘3 i 1'. g . . - - ‘ -- , .-., ”.... .-- ,v- -'-'-vo-vo ~7v0‘nt HIM-V. r . . - . l . .' _. Comparative electrophorograms of normal serum proteins. A: human serum: (1) conventional pro— cedure, (2) 7% gel-12 cm column, (3) 4.75% gel- 12 cm columns, (4) and (5) differential procedure. B: bovine serum: (1) conventional procedure, (2) 4.75% gel—12 cm column, (3) differential procedure. 82 Orv. ..i- o . . . {vulturtt I: t. u . u . i. if 1 ty of the method. 1i lectrophorograms of normal 1 ducib ing repro ifferential disc e how D serum 5 20. Fig. 83 Fig. 21. Eight individual bovine serums from a 16 column run by the differential system, illustrating variation among serums and reproducibility as measured by the distance migrated by the albumin in a single run. Fig. 22. Comparison of the separation by the conven- tional (left) and differential (right) systems of the initial serum (I) and post-inoculation serum (P) from a cow eXperimentally infected with Mygobacterium bovis. m.-..—— ' 85 By the standard procedure, the bands were poorly resolved and only a decrease in concentration of the individual bands could be demonstrated in the post—inoculation serum. The differential system revealed a number of bands not detected in the 7% gel, particularly in the initial serum. Both the separation and resolution were markedly improved. Serums from a number of other species (sheep, rabbit, dog, horse, guinea pig, pig, mouse, chicken, and duck) were separated by the differential system and the conventional procedure. The separation of sheep serum was increased from 19 to 31 bands. Five to 8 more bands were separated in rabbit, horse, guinea pig, and dog serums. This method did not increase the separation of mouse, dog, chicken, or pig serum. The resolutions were as good as, or better than, that found in the standard gel. With all serums tested, a 4.75%.gel by the standard procedure gave better separation, better resolution and more reproducible results than the 7% gel. Comparison 9f the Major Electrophoretic Procedures Used for the Analyses 92 Normal Bovine Serum The relationships of the patterns obtained in cellu- lose acetate, starch—gel, differential disc and agar immuno- electrophoresis are presented in Figure 23. Other than the 4 zones that have been identified (gamma, Sa beta and 2! albumin zones), no nomenclature can be given for the numerous 86 s’ \ \ \ i | 'IIIIIIII IIIIIII l I l : I" I : " l I ' / I. I *7“! 7‘“;7‘-w [in 3 M f ' I .ll ’.' I I l ‘ I I'f II!“ 'I § = E 5 if E E -——_ ::==;;;=;55555 = :::: :§ 53- = = -—{- § 5‘ 5‘. : ~ : x8 Y 'Y l're-illb M Post Alb Fig. 23. Diagramatic representation showing the relationships among differential disc electrophoresis, cellulose acetate electrophoresis, agar immunoelectrophoresis, and starch gel electrophoresis of normal bovine serum (top to bottom). 87 protein bands separated by the differential disc electropho— retic procedure. Further definition is dependent upon the isolation and characterizations of each individual zone. Most of the relationship between starch-gel and agar immunoelectrophoresis is based only on presumptive evidence. Cellulose Acetate Electrophoresis “pf Serum from Tuberculous Cows Experimentally infected animals. Representative densitometric tracings of cellulose acetate electrophoro- grams of serum collected from the experimental animals are shown in Figures 24, 25, and 26. From this type of tracing it was possible to calculate the relative percent protein values for each fraction.1 A plot of these values from electrophorograms from representative animals showing the fluctuations and alterations in serum protein and protein bound carbohydrate patterns at various stages during the disease and the effect of tuberculin testing on these pat- terns are presented in Figures 27 through 31. The patho- logic changes, the pre and post hemogglutination (HA) titers, and serum protein changes determined from cellulose acetate electrophorograms of the experimental animals are presented in Table 4. lRelative percent protein was calculated by dividing the area of each fraction by the total area of the electro— phorogram and multiplying by 100. 88 c LJD L Fig. 24. Typical densitometric recordings of cellulose acetate electrophorograms of bovine serums. (Top) M, bovis. A: pre-inoculation, B: 132 days post-inoculation. (Bottom) 68C (virulent Group III). C: pre-inoculation, D: 122 days post-inoculation. Fig. 25. 89 L L Typical densitometric recordings of cellulose acetate electrophorograms of bovine serums. 132 days post-inoculation. A: pre-inoculation, B: (Bottom) SOB (avirulent Group III). D: 172 days post-inoculation. (Top) M. avium. C: pre-inoculation, Fig. 26. 9O L ” L Typical densitometric recordings of cellulose acetate electrophorograms of the two types of patterns from cattle infected with 51C (virulent Group III). (Top) Type I. (Below) Type II. A: and C: pre-inoculation, B: and D: 179 days post-inoculation. Relative percent protein 50 4O 3O 2O 10 50 4O 3O 20 10 Fig 91 Calf #20 O O\.__.%amma O . O Alb n / O\ o/.\O um: .“\-\‘V-’.‘“* h__. Alpha 0 Beta ______.__~‘. ___________ .~~i_‘.’,’. _____________ . Carbohydrate 0 10 20 30 40 50 60 Calf #113 . .__ \o '* Alb ' ___*____‘ umin .\./ .\ . ________———. Al ha ’////’.‘~§‘““‘-O-v"'”””-.‘--:r~““.“—————————___.}? ‘ . / \.‘—* -\\\\\\‘.,,-””’ Beta F...‘ "”.“~ ’I’.’. _________ . --O ........ ..—-" ‘0’ Carbohydrate 0 20 40 60 80 100 120 140 Days Post-Inoculation . 27. Alterations in serum protein and protein-bound carbohydrate patterns in calves experimentally infected with M, bovis. Relative percent protein 50 4O 3O 20 10 50 40 30 20 10 92 Calf #116 O\\\. . .-1\\““Album1n ><. /Gamma O . ‘///. -/ . "_’flfl_dfl_..0 \./:/ \ ./. Alpha ‘\‘o 0 Beta 0 20 4O 60 80 100 120 140 Calf #81 Albumin OV/”.\\\ ,,-O-————-.-—-O//‘%:>X%<;:>XC.::>C<;L“~./// ‘N\~. ' //" Gamma O O .___ o-——-o -_ ---- o--- .--—4-..- , I.“‘ //. . “ .‘\ [II // “.“’ \. ,’ Carbohydrate o 10 20 30 4o 50 60 :70 80 90 Days Post-Inoculation Fig. 27. Continued. 93 50 Calf #21 /,// 40 \\\ Gamma O ::::><::::: Albumin 3o ..a-4’-o o o Relative percent protein ..,-r” ‘/,o ... o ‘\\\\ /// ..,-—‘ Alpha 20‘ /// —‘—-O ./ \. .\. 10 ",,.o ---o\ ..... o-”' ---’.~-~.,- ,,-'.’ ‘0'”“" Carbohydrate 0 10 20 30 4O 50 6O 7O 80 9O 50 Calf #115 O /////.\\\\\ I“-o Albumin 3o ...--*""’ 20 0 A1 ha y///’.-“—‘——’. .“‘-o””’ IIIIII“--fio 10 ~\\“‘\\1. .,/’// 0 Beta 0 20 4O 6O 80 100 120 140 Days Post-Inoculation Fig. 28. Alterations in serum protein and protein bound carbohydrate patterns in calves experimentally infected with M, avium. Relative percent protein 94 5° Calf #59 o-””'.““~O 40 ___1. ° ___./ O .\ O\ ./ Albumin O 30 ' .-/’.““--O-—'“’J'"I’GBmma \ o ' o><::: o 20 Alpha ____/ I 10 , ,OL-- . ... """" .“\‘ x’ . ‘‘‘‘ . ---- ."_‘-. ””” ‘0’ Carbohydrate TB test T TB 1 f 0 20 40 60 80 100 120 140 160 180 50 Calf #43 /.\ 40 O ‘-a..___.——-—*'.*-.., . Albumin O 30 0""’”’ ' .,///o 0”’/ Gamma O ... o- o- o I. .p_ 20 . \.r . Alpha 10 0 Beta TB Lo 10 20 3o 40 so 60 70 Days Post-Inoculation Fig. 29. Alterations in serum protein and protein-bound carbohydrate patterns in calves exPerimentally infected with strain 51C (virulent Group III). Relative percent protein 50 4O 30 20 10 50 4O 3O 20 10 Fig. 95 Calf #64 O Gamma O~_.. O O \'>< o g C .,/-”’ Albumin / O o / o /.\o / \O o O\.\Oma. /0 0/'\,/ 0\./0 Beta TB test TB TB 0 20 40 60 80 100 120 140 160 180 Calf #70 Gamma ,/-””’/’.fi~§‘-§~“—. I O O/ \ .\ g / Alpha 0 (o O-*"’.““‘*‘*-'“*“"’—"'. “'“"" Beta TB test TB 0 20 40 60 80 100 p 120 140 Days Post-Inoculation 7 30. Alterations in serum protein patterns in calves experimentally infected with strain 68C (virulent Group III). Relative percent protein 96 so 0 o\ Kowo/ \././ “\o o Albumin 40 Calf #66 30 \./ “Tn/ \.\./ ...... 20 C-—-—- .__Q - . Alpha 10 Beta .“‘ . II/ “.‘\\ ’,. ~~~~~ fi’.’/ \“.-"”. . /,/ 0’ AL Carbohydrate 0 20 40T 60 80 100 120 140 160 180 TB test TB TB 50 Calf #80 Albumin 40 ——-o 0/. —————.\.7<:><:—_ . .\. O / 30 \O Gamma 20 9’,,.’w. fi’fff,.——Oi\‘. /////// . E—-._——. 10 \\\\\\.. Beta TB test TB TB ' 0 20 40 60 80 100 120 140 160 180 Days Post-Inoculation Fig. 31. Alterations in serum protein and protein—bound carbohydrate patterns in calves experimentally infected with 50B (avirulent Group III). 97 .m3oo Daspd an .Houmcflfiocm6HMmmousm OD Hoaum wouooa Iaoo mamfimm GOHumasoocHlumom “HoumuofidanamEHocv onEMm coaumasoocflloumm .mcoc Docnmz .mcflumou :HHDUHmnfip “mom mhmp gain popuoaaoo Eduomuom «awasononnu mo aofluoonCH oHOMoQ pouuoaaoo Ednomuum "mmmoudm ou woewm cmuooaaoo moamfimm mo muouflu coaumcflusammmaon afiasouonsu Doom can can N .ucommum ma coflmoa mo GOHDMHSmmmocoul "woodman woNH Iamuocomu+ "ommmmflo o>Hmmoumoumum .GOHDMHUUOGH mo ouflm um coflmma Mano Ho mCOHmoH omoumonoflfi Ho mmonm OGHI “ouflm COHDMHDUOGH mcflcflmuc econ zmfima mnu CH mCOHmmH uonmEoo wnmaflum "Om axoamfioo wumaflum on» caowofl ouo£3hcm mGOHmoH pmNHHmuocomuO .Mmmousm ou Hoaum wouooaaoo mamamm Eamon Hmcflm GOHDMHSUOQH Doom mhmpummmaa mm\m.mm ~H\¢H ma\am em\mm H.H\o.a oz oz a I mm Hm m~\om ma\m.aa om\om m.mm\mm ¢.H\m.a cm I + on me mad mm\m.mm H.mH\oH m.ma\¢.~m mm\mm m.H\a.H oe om I on we mma m.mm\mm ~H\¢H m.sa\m.oa om\~¢ o.H\m.H owe om + mma was H.0m\a.aa ma\m.HH m.oa\m.ma e.o¢\m.om m.a\m.m cam as on mma mas m.m¢\¢m, «.maxma m.mH\om,g- m~\m.mm o.o\o.a as 6H + m6 om m.aawmm m.oH\mH oa\a.ma mm\mm m.o\H.H ow . + am e .I mH>on .2 MEEMO muom MSQH¢ om mm 0 whom .02 mcflasnoao cfladnad w\¢ llmmmll ammoHonumm mamu mwGOHHMHSUOGH “mom USN mun.— mauouomfloomfi SDHB Cowmasuocfl mo>amu mo hmoHoanm pom Acmv muopfiw GOHDMQHDSHmmmmEoS .momGMSU Camuonm Esumm .v magma 98 m~\¢.e~ ma\aa o~\ma mm\mm 4.H\¢.H ow om I on em «mos ~m\m~ ma\m.ma ~.ma\m~ am\oe ~.H\G.H om I + + m Hm\~.em ma\¢a -\AH H.~m\¢.~e o.a\o.a om om + + so am m.¢m\m.m~ m.HH\o.mH e.a~\e.mfl a.~m\mm m.o\¢.a I on + + me am om\om m.aa\m.¢a ma\o~ H.mm\m.om o.a\m.o 06H oea + + on, he m~\~.em a.aa\m.ea H.Hm\~.o~ mm\ae m.a\s.a 06H ooa I + me me m.mm\¢.s~ BH\H.RH ¢.ma\m.ma o¢\~.em 6.H\a.a omm 04 I + was ow e.m~\m.H~ . ~H\ma ma\am H4\H.mm ¢.H\m.a omm 04 I + mea mm AHHHV cam m.vm\a.em mwea\m.ma o.ma\a.aa ¢.Hm\mm m.o\m.o 06H oe + + on em ~¢\mm 4H\GH aa\m.o~ m.em\e.~m e.o\o.a omma 046 + + was an m.m¢\em H.HH\m.eH H.Hm\oa mm\om e.o\m.o ooa om + + NNH on oe\m~ oa\m.ma 4H\GH Am\m.ma m.o\m.H 06H 04 I on me, so AHHHV owe, mesmw muom mnmafi om um m U mmma .02 9.3.503 agenda o}, «mm 3365mm. memo mmCOHumasuocH umom cam mum UoUCHDCOOIle canoe 99 m.m~\m.o~ ¢.ma\m.¢a m.m~\a.sa om\H.H¢ ¢.H\m.a ow I I I 46 mm «.ma\am m.ma\m.¢a ¢.mH\mH m¢\om m.~\~.a 04 I u I mma mm m.~m\¢m ~H\¢H ~.aa\m.oa «.mm\om ~.H\H.H ooa 04 I I Hma om m.s~\m.m~ «.ma\¢.ea Hm\- sm\m.mm 4.H\N.H coo cos I I am so H.om\m~ NH\m.~H aa\m.aa H.me\m¢ a.H\H.~ can con I I NEH we AHHHV mom cm\~.~m ¢.~H\e.ma ma\ma ~.mm\m.am m.H\~.H om I I I «ma mas m.mm\m.mm ma\m.ma mH\aH m.¢m\¢.mm o.a\o.a om I I I mma was a~\m.e~ ~.HH\mH m.ma\m.a~ ~¢\mm G.H\¢.H I I I I so mma m~\m.Hm mH\mH -\m.o~ mm\em ~.H\H.H I I I I mm mHH wm\m.m~ ¢.oH\m.HH m.o~\¢.ma m.mm\m.m¢ m.o\s.a om I I + em H61 chums/m .2. ~.mm\ma m.ma\ma GH\H.RH mm\s.am m.o\a.a om ow I on mum moa em\me a.ma\ma 6H\ma em\om o.a\n.o omm as I on mom Goa mm\a.mm Ha\m.aa o~\om m.om\em m.o\H.H omm as + on em «as Hm\~.mm mH\H.HH H~\4.BH om\¢.~¢ ~.H\m.a om ow + + mum .moa s~\am ma\~.ea 6H\¢.GH m¢\oa o.m\m.H 06H o4 + + How «Boa mssmo mumm armaa om- no A o msmo .oz maflasnoao :Assnaa o\a mam ammoflozumm came mmGOADMHSUOBH Doom cam mum UmDCHDQOOIle oHQMB 100 In calves inoculated with g, bovis, serum albumin was moderately reduced and gamma globulin significantly elevated resulting in a decreasing albumin/gamma globulin (A/G) ratio (Table 4). The alpha and beta globulin remained constant except for minor fluctuations. The hypergammaglob- ulinemia was detected 7-14 days post—inoculation and remained until death. When culture 81 was injected intradermally into calf 20, it caused death. The same culture did not produce any lesions in calf 81 when injected subcutaneously and the albumin and globulin fractions remained constant (Fig. 27). No significant differences were found between the normal and post—inoculation serum protein patterns of calf 128 which was injected with culture 310. Only a primary complex developed. However, other calves (113 and 135 inoculated with culture 310 and 278 respectively) develOped a primary complex and also had hypergammaglobulinemia. No consistent differences were detected in the pro— tein distribution between normal serums and serums collected from 4 calves inoculated with g, aging. The A/G ratios were constant or slightly increased. Calf 21 (inoculated with g, ayigm) had protein changes similar to calves infected with M, bovis. There was a lesion at the site of inoculation and in the head lymph nodes. In addition to a rise in gamma globulin and decrease in albumin, there was a rise in the alpha globulin fraction 101 similar to the changes observed in human tuberculosis and experimental tuberculosis in guinea pigs. Proteinébound carbohydrate also increased. Responses of animals inoculated with Group III myco- bacteria varied. Culture SOB did not produce any demonstra- ble lesions and no consistent changes in serum protein frac— tions could be detected. The A/G ratio remained constant or slightly increased which was similar to calves inoculated with g, avium. Culture 68C produced the same pathology and serum”- protein changes as described for g, bovis injected intra- dermally. Culture 51C produced the most inconsistent changes in serum proteins. Two different patterns could be detected: The first type was similar to the serum protein changes from animals inoculated with M, avium and 50B (Fig. 26 and 29: Calf #59). The second type mimicked the serum protein changes demonstrated with animals infected with g, bovis and 68C (Fig. 26 and 29:' Calf #43). This organism was patho- genic producing gross lesions but no consistent A/G ratio resulted. In general, carbohydrate and lipid staining revealed no consistent differences between serums collected from tuberculous calves before and during infection. 102 In a few cases, the gamma globulin fraction in- creased after tuberculin testing (Figs. 29, 30, 31). It was not consistent as previously reported by Mallmann et al. (118) (Table 4). flaturally infected animals. The serum protein changes, pathologic changes, pre-and post-tuberculin HA titers, and bacteriology of naturally infected cows are presented in Table 5. Generally, serum samples collected from both GL and NGL animals before and after tuberculin testing had higher A/G ratios than tuberculin negative cows. The total serum protein level dropped considerably following the tuberculin test in all GL and NGL animals. Tuberculin testing had little or no effect on the A/G ratios in tuberculin negative animals. A comparison of the mean relative percent protein values from normal and experimentally and naturally infected cattle are presented in Table 6. The mean A/G ratio from both GL and NGL (naturally infected group) was slightly higher than calves experimen- tally infected with g, bovis or 68C but sufficiently lower than the normal mean A/G ratio. This and the fact that the mean A/G ratio from tuberculin negative cows was approxi- mately a normal A/G ratio was indicative that these animals were infected with a virulent mycobacteria. 103 seasuumnsuu.mmc me “maco coflmma cflxmuqm “coamma mmoum ocuqoz “coamma mmoumuqo .coaumHOmH OGHHZ umHmEMm ocumz "mmOHOHHmuomm m N .HouMCHEocmcumHmEMm cflasoumndu “mom “nonmumescncflasoquSu mo coauomncfl mnomwn mHmEmm cfiasoumndu mum .mCOHme mmoum cam m>flummma CHHSUHmQSuquw am>flummmc a mm\mm oa\e.aa m.ma\na ¢m\m.mm m.o\m.o om\oa mz auz maoo HHH Am . ~v\~.mm ¢H\HH m.wa\¢.ma m~\¢.mm n.o\m.o ovo\om ma>on .2 qoz ommm mm\m.¢m m.mm\mwm ~H\- om\mm m.o\o.a om\o¢ mz Inc mmmm. ..mm m.HH H.o~ om m.o oa ma>on am qoz meom m.¢m\om oa\m.ma am\m.om mm\sm o.H\~.H o¢o\om mfl>on am no ammo H.mm\a.om ma\ma -\ma om\m.mm m.a\m.a om~H\o~ ma>on am qoz mama om\~.¢m «.ma\~.¢a a.¢m\om hm\mm m.o\m.o oe\om mfl>on am_ no Noaa ~.om\mm H.na\ma H.sa\ma om\¢.em m.o\o.a om\om ma>on am no mama Hm\H¢ mH\oa ma\m.ma. mm\a.om m.o\s.o om\o~ ma>on am, am mmam mEEmw wumm MSQH¢ GHESQH< 0\< «m Nhuomm mmoHonumm ..oz H HMEHQ< awasnoao mmmamamm.cflasuumnde umom cam mum mBOU owuommcfl adamuzumc mo mmoHoaumuomn m5» .mmoaonumm .Admv mumuflu coaumcflusammmam3 .mmmcmnu camuonm 85Hmm .m magma 104 ¢.sm\¢m H.HH\HH ma\oa mm\oe m.o\~.a om\o~ mz .mm: me emmm am\m.mm mH\m.¢H H.ma\m.sa «.mm\a.o¢ o.a\¢.a o¢o\om m2 .mmc me oaoo H.mm\mm m.oH\m.mH ma\m.ma mm\¢m H.H\H.H ooa\o¢ mz .mma ma omom ~m\mm ~H\~.~H «.ma\¢.oa nm\nm N.H\H.H o¢\o~ mz .mm: ma mmmm om\mm HH\m.HH -\v.ma H.~m\¢m m.o\m.o ooa\o¢ mz quz mama «.mm ma ~.sa mm o.H ooa m2 qwz mooo mm\o¢ m.oa\ma mm\mH ~.Hm\om m.o\m.o oma\ooa mz qoz momm m.om\mm v.oa\oa H.o~\ma mm\mm m.o\m.o ooa\o~ mz gm camm m.om\a.mm ~H\¢»NH m~\om mm\mm H.H\m.o ooa\oe mz am cone om.mm oH\¢H m.sa\e.na sm\om ~.H\m.o o~m\oao mz qoz ooom mm\nm ea\ma ¢.5H\Hm ~.om\om m.o\m.o o~m\oa Hz qwz omma Hm\a.mm NH\¢.~H m~\o~ mm\mm o.a\m.o ooa\o¢ mz qwz moss mm\a.om va\¢.ma mH\H.sH v.m~\m.mm s.o\m.o om\om mz qoz mafia mm\mm oa\va m~\mH mm\m~ s.o\s.o o¢\ov Hz qwz «New m~\om ma\oa m.mH\mH m.mm\am o.H\m.o ooa\om mz qoz mama magma mumm mamaa caesnaa oxa am mmuomm ammoHozamm .oz Hmsflca :HHsQon mmwamfimm aflasoumnde umom wcm mum pmscflpcoollm mHQMB 105 Table 6. Mean values of serum protein fractions and variations due to disease activity in bovine tuberculosis No. Globulins of Cases A/G Albumin. Alpha Beta Gamma Normal calves 38 1.4:p.l 37.3:8.2 l9.lil.8 14.4:l.0 29.7:§.6 Experi— mentally infected calves g, bovis 7 0.810.07 30.2:2.5 18.9il.6 12.2:l.2 38.3:1.0 fl, avium 5 1.3 38.8:5.0 16.812.4 l3.4:1.9 29.3il.3 Unclassi— fied 68C 9 0.7 27.5:l.3 l9.7:l.0 13.7:2.7 39.7:2.4 51c 14 1.30.4 38.2:3.7 20.1:2.o 13.3:1.328.8:2.8 50B 9 1.6ip.3 40.9:l.9 l9.9:l.2 13.3:p.8 26.1:l.6 Naturally infected calves ghbovisl 5 1.0:o.2 34.1:3.0 17.4:1.3 12.7:1.0 34.8-52.7 TB neg.2 4 1.2io.2 37.0il.6 17.6j-_2.4 13.9i1.9 31.7:1.4 NGL3 13 0.9:0.02 32.1:2.6 18.5:l.4 12.810.3 37.1:0.9 lTuberculin positive animals fl, bovis isolated. 2 3 Tuberculin negative animals. M, bovis isolated from 3 cases (see Table 5). which had gross lesions and Tuberculin positive but no gross lesions (NGL) found. 106 Starch-gel Electrophoresis 9: Serum from Tuberculous Cows Starch—gel electrophoresis of serums collected from calves inoculated with M, bgyis or 68C revealed, in addition_ to an increase in gamma globulin and decrease in albumin, an increase in post albumin fraction 9 (Fig. 13) 7-14 days post- inoculation and remained elevated during the disease period (Figs. 32 and 33). This band was always elevated 7-28 days post-inoculation in serum from calves inoculated with 51C regardless of whether an increase in gamma and decrease in albumin occurred (Fig. 34). It was not elevated in serums from calves inoculated with M. avium or 50B (Fig. 35). The fraction has been tentatively designated as the az—V (virulence) fraction. The increase in the a2-V frac- tion is clearly illustrated in Fig. 36. It was also ele- vated in all serums from GL and most serums from NGL cows selected from naturally infected herds and not evident in serum from tuberculin negative cows. No other consistent changes could be observed in other bands separated by starch-gel electrophoresis. 'mn-rm-E Differential Disc Electrophoresis 9§_Serum from Tuberculous Cows It was impossible to obtain reproducible densito- metric analysis of the differential columns with the chromoscan. Therefore, only visual qualitative evaluation of the numerous fractions could be made. 107 Fig. 32. Typical starch—gel electrophorograms of serums from experimental calves inoculated with g. bovis. (l) pre-inoculation serum (normal) (2) 12 days posteinoculation serum (3) 39 days post-inoculation serum (4) 82 days post-inoculation serum (5) 98 days post—inoculation serum (6) 132 days post—inoculation serum. 108 wvflmw: ‘.~:.;m-;A0.’1r ~= Fig. 33. Typical starch—gel electrophorograms of serums from experimental calves inoculated with 68C (virulent Group III). (1) pre—inoculation serum (normal) (2) 14 days post—inoculation serum (3) 39 days post—inoculation serum (4) 82 days post—inoculation serum (5) 124 days post-inoculation serum (6) 173 days post-inoculation serum. 109 Fig. 34. Typical starch-gel electrophorograms of serums from experimental calves inoculated with 51C (virulent Group III). (1) pre—inoculation serum (normal) (2) 14 days post-inoculation serum (3) 45 days post—inoculation serum (4) 82 days post-inoculation serum (5) 124 days post-inoculation serum (6) 173 days post-inoculation serum. Fig. 35. Typical starch—gel electrophorograms of serum from experimental calves inoculated with g, avium (top) or (below) 50B (avirulent Group III). (1) pre—inoculation serum (normal) (2) 14 days post-inoculation serum (3) 45 days post—inoculation serum (4) 82 days post-inoculation serum (5) 98 days post—inoculation serum (6) 132 days post-inoculation serum. .mcumn Umuommcfl adamusum: Eoum mzoo coamma mmoum on cam cowmmalmmoum paw .Uam no 0mm .mfl>on xfi.£ufl3 pmumadoocfl mauumo HmucmEflummxm Eoum Enumm coaumasuocwlumom Amy cam Efiumm mcw>on HmEuoc Aév mo memumouozmouuomam Hmm noumum mo mosaououmu UflumumEoufimcmp Hmowmhe a A: a 111 .a;\’ .mm .mam 112 The hypergammaglobulinemia and decrease in albumin demonstrated in serums from calves infected with g, bgyis or 68C by cellulose acetate and starch-gel electrophoresis was also detected by this methdd. There were in addition to the increase in gamma and decrease in albumin, increase in the Sa2 band (see Fig. 21), increase in the first post albumin band and a marked increase in a thick zone just gj posterior to the beta zone (Figs. 37 and 38). The latter =‘ may correspond to the a -V fraction detected in starch-gel 2 electrophoresis. These bands also increased in post~inocu- a lation serum from calves eXperimentally infected with 51C but did not increase in serum from calves inoculated with g, avium or 50B (Fig. 38). These bands also increased in pre-tuberculin serum samples from GL and NGL (naturally infected) animals but not in pre or post-tuberculin samples from tuberculin negative cows. Post-tuberculin samples from GL and NGL animals could " not be evaluated qualitatively due to a marked decrease in total protein (Fig. 39). A decrease in total protein did not occur in post—tuberculin serum samples from tuberculin (I negative cows. g Of the ten bands which stained for carbohydrate, only the Sa2 and the a2-V band were increased in serums from the M, bgyis, 51C and 68C infected calves. These bands Were not increased in serum from calves inoculated with 50B and fl, avium. Fig. 37. ll3 ! i v '— I 1‘ it; -_ 1‘ '9 1 .1 r- I r I f I in ‘ W I ¢=ff E._F i "*“'-4!T4V 'Ij‘ '1 5131 an. a} .agt J a; It: ‘41,) ' ' a - F Differential disc electrophorograms of pre- and post—inoculation serums from experimental calves inoculated with g, bovis, 68C or 51C (virulent Group III). ‘ 4. G . xfiLgsim—E (A) pre—inoculation serum (normal) (B) 14 days post—inoculation serum (C) 39 days post-inoculation serum (D) 82 days post-inoculation serum (E) 98 days post-inoculation serum (F) 132 days post-inoculation serum. Fig. 38. Iva-3. 0‘ .--_~- Ir, u... . 4 5 "7'3. 5 .-,.f.g€ u-ga-zof 114 Typical differential disc electrophorograms of pre- and post-inoculation serums from experimental calves inoculated with g. avium 0r SOB (avirulent Group III). (A) pre-inoculation serum (normal) (B) '14 days (C) 45 days (D) 82 days (E) 132 days post-inoculation post-inoculation post-inoculation post—inoculation serum serum serum serum. llS .mHmEHcm .mmc m9 Eonm Esumm ARV ms» CH uoc pan mHmEHcm A02 6cm no anon Eoum mmHmEMm “Ev mzu CH mwcmn cflmuoum Ham ca mmmmuomc mnu muoz .mwnmn wmuommcfl haamnsum: Eoum mauumu Aqozv cosmmalmmoumloc cam Aqu coammalmmoum m>HuHmom :HHSUHmQSu Eoum 6cm .mauumu A.mma mac m>Hummmc cflasonmndu Eoum madumm ARV awasoumndulumom cam Amy Imam mo mfimumouonmouuomam Umflo HMHuQmHmMMHc Hmoflmwe :2. no: 5 .5 a ..._. p h + _ I I l .— «H1 _ .- 1?; .‘M “3‘ I_r ,IIasv .1. v I. _ a... «m. .mm .mam 116 The analyses for similar changes in serums from cows natually infected with other chronic disorders are presented in Figure 40. Paratuberculosis, brucellosis, and leptospi- rosis caused serum protein patterns similar to both the terminal serum protein patterns of a calf eXperimentally infected with g, bovis. Serums from cattle With paratuber- culosis had denser gamma globulin zones than the serums from tuberculous animals. A band just prior to the major beta zone (see Fig. 40) was markedly greater from a cow with brucellosis than in serums from normal and tuberculous cattle. The serum from a cow with antinomycosis was similar to a normal serum. Immunoelectrophoresis of Serums from Tuberculous Cows Generally a -2, a -3, a -4, a —5, a -6, a -7, and 2 2 2 2 2 2 a -8 precipitates were slightly increased 7-14 days post- 2 inoculation and remained constant until death. In serums from animals inoculated with g, bovis, 68C or 51C the a2—1 arc became markedly increased 28 days post-inoculation and increased in density until death (Figs. 41, 42, and 43). The immunoprecipitate was stained with protein and carbohy— drate specific stains. All evidence indicated that this immunOprecipitate is identical with the a2—V fraction separated by starch-gel electrophoresis. An increase in the gamma globulin precipitate also develoPed simultaneously 3'1 'D' I i ‘1 Tara: ‘ J Fig. 40. ,-.A.- 117 ...!“ -.‘--o A 'x ‘ I , _ . - ”'“WWfiv.1:0f' '0“: '- - Comparison of differential disc electrophorograms of normal bovine serum and serum from tuberculous cattle with serums from other chronic inflamatory diseases of cattle. (A) (B) (C) (D) (E) (F) (G) actinomycosis pattern paratuberculosis pattern (Johne's disease) naturally infected tuberculosis pattern (M. bovis) normal bovine serum pattern brucellosis pattern experimental tuberculosis pattern (fl, bovis) leptospirosis pattern. Fig. 41. 118 Immunoelectrophorograms of pre— and post-inocula- tion serums from calves with g. bovis. (A) pre— inoculation serum (normal), (B) 14 days post- inoculation serum, (C) 82 days post-inoculation serum, (D) 132 days post—inoculation serum. Fig. 42. 119 Immunoelectrophorograms of pre- and post-inocula— tion serums from calves with 68C (virulent Group III). (A) pre-inoculation serum (normal), (B) 14 days posteinoculation serum, (C) 82 days post- inoculation serum, (D) 173 days post—inoculation serum. 120 9r)! t} “ 'x‘. ._'-p..)4~fv 02 -V Fig. 43. Immunoelectrophorograms of pre- and post-inocula— tion serums from calves with 51C (virulent Group III). (A) pre—inoculation serum (normal), (B) 14 days post—inoculation serum, (C) 82 days post— inoculation serum, (D) 173 days post—inoculation serum. 121 with the increase in the a2-1 arc. An increase in the gamma or decrease in albumin was often difficult to determine quantitatively. The a2-13 lipoprotein precipitate decreased slightly during the disease process in serums from the above group of animals. No changes were detected in any of the carbohy- drate staining arcs other than the increase in the a2-1 precipitate. None of these changes were detected in serums from calves inoculated with g, avium or 50B (Figs. 44 and 45) or tuberculin negative cows. 122 WWW , i a ‘ v-n,’-- 0.. -/. 02-1 02—] Fig. 44. Immunoelectrophorograms of pre— and post-inocula- tion serums from calves with g, avium. (A) pre- inoculation serum (normal), (B) 14 days post- inoculation serum, (C) 82 days post-inoculation serum, (D) 132 days post inoculation serum. Fig. 45. 123 ‘fs-ckftrrr-f. , ~ ." a - Immunoelectrophorograms of pre— and post— inoculation serums from calves with 50B (avirulent Group III). (A) pre—inoculation serum (normal), (B) 14 days post-inoculation serum, (C) 82 days post—inoculation serum, (D) 132 days post—inocula- tion serum. DISCUSSION Development of Techniques and Analyses gf Normal Bovine Serum Proteins Relatively few studies have been concerned with the characterization of bovine serum proteins. A more complete knowledge of the chemical, physiochemical, and immunochemical properties of the serum proteins is a prerequisite to an understanding of their normal physiologic functions and changes induced during pathologic processes. The electrophoretic composition of bovine serum characterized by cellulose acetate electrophoresis was essen- tially the same as reported from studies employing a single procedure (18, 19, 26, 27, 47, 93, 138, 159). The resolu- tions of the protein bands were markedly better than that obtained with paper electrophoresis. The selection of the appropriate buffer system to be used with blood serum from any animal species is empirical because of the differences in composition and electrophoretic properties of the protein constituents. For example, the alpha component in the serums from human, swine, horse, dog, and mouse separated into two components (alpha1 and alphaz) after cellulose acetate electrophoresis with Owen buffer 124 125 system. The alpha component in bovine serum did not sepa- rate under the same conditions but was separated into two fractions with a tris-EDTA-borate system. The barbital- acetate system was employed because it gave consistent results. A buffer system that provides ideal separation of the serum constituents from one species is not necessarily satisfactory for serums from other species. Although a sys- tem may appear to give ideal separation and resolution, the results may not be consistently reproducible as with the electrophoretic separation of bovine serum. Therefore, a complete analyses of buffer systems (pH, species of salts, ionicity) for both separation and reproducibility is re- quired. It is also of paramount importance to consider the different fractions of proteins which occur when different apparatuses, buffer solutions, etc. are used.' Establishing normal conditions for each investigation is essential. Because of the simplicity of separation, lack of protein trailing, and clear background; cellulose acetate electrophorograms were the only medium from which reliable relative percent protein values could be calculated. How- ever, errors in measuring protein by measuring bound dye are inevitable especially since the binding capacity of many proteins is not known. There is a wide variation among pro— teins in lipid, carbohydrate, and nitrogen content, within these limitations, dye-binding is a convenient parameter to me asure . 126 Simple agar gel electrophoresis did not improve the separation or resolution of bovine serums that was obtained by cellulose acetate electrophoresis. Agar electrophoro- grams were unsuitable for densitometric analysis due to con- siderable protein trailing and electroosmosis. Deionized agar might decrease one or both of these difficulties. The large pore size of agar may contribute to a poorer separa- tion and resolution of serum proteins than that in starch or acrylamide gels. The use of a more highly purified agar would permit a longer period of electrophoresis by reducing the amount of electroosmosis. A better separation of the globulin zones could be accomplished. This agar can also be prepared directly in the buffer and used immediately, elimi— nating the necessity of further purification. The apparatus for starch—gel electrophoresis, and the electrophoretic destaining of starch and polyacrylamide gels was developed during this study. Most of the commercial apparatus and techniques for starch—gel electrophoresis had several inconvenient features: (a) excessive handling and exposure to harsh solvents cause an injury to the gels; (b) a separate gel mold was required to change dimensions of the gel; (c) the gel slicing operation was tedious when using either a dermatome knife (179) or fine wire (23); (d) the methods for staining and destaining the gels were cumbersome and time consuming; and (e) destaining by electrophoresis often resulted in soft and friable gels. 127 The apparatus and techniques for starch—gel electro- phoresis was modified and improved. The gel mold, which was similar to Smithies (180) incorporates two spacers (Fig. 1) to facilitate changing the depth and width of the gel. In this manner, only one gel frame was required, all changes being made with the appro- priate spacers. The cover and removable slot formers were identical to Smithies (180) but the slot formers were made to correspond with the width and depth of the spacers. The volume of sample could be adjusted by varying the length and thickness of the slot formers. The gel mold eliminated the necessity of having a separate mold for each size gel, mak- ing it usable for a variety of experiments such as two- dimensional electrophoresis. Using acrylamide gel with this apparatus was moder— ately successful. The synthetic gels are flexible and were supported by a sponge in the lower buffer compartment and saran wrap on the surface of the gel prior to placing the retainer plate in place. Preliminary studies indicated little advantage in using acrylamide gel slabs for analyses of bovine serum. The slicing operation was improved by slicing with a fine wire. This provided smooth and nearly uniform sections. Excellent results were obtained because the gel was firmly supported during the slicing operation. Furthermore, less dexterity was required on the part of the operator. It.“ 128 The improved procedure for staining and destaining starch and acrylamide gels saved considerable time and lessened the chance of breakage. In previous procedures the sliced gel was placed directly into a staining tray or sup- ported by a glass plate or heavy filter paper and the stain poured over the gel. After staining was completed, the stain was removed and the solvent which usually consisted of a mixture of methanol and acetic acid was added. Prolonged contact with methanol caused the gel to become very brittle and handling usually resulted in a broken gel. Destaining by electrophoresis (56, 78) has eliminated the use of meth- anol and decreased the time required to destain a gel. In the procedure deve10ped, the gel is supported between two framed screens (gel holder), which is placed into the staining tank (Figs. 5, 6). The destainer (Fig. 7) was designed and coupled conveniently with the staining tank which eliminated any handling of the gel during staining and destaining. The addition of 10% glycerol to the acetic acid solution for destaining starch slabs, causes a slight dehy- dration of the gel which results in a more resilient, easily handled flexible gel. The gel becomes semi-transparent from which densitometric recordings can be made. The major difference between the results of one dimensional starch—gel and those reported by Ashton (6), was in the number of components in the post albumin fractions. 129 Nine bands were detected. Only one post albumin band was reported by Ashton (6). This difference in number of post albumin bands is largely due to different buffer systems employed in the two investigations. Ashton used a phosphate buffer system (pH 7.8). A discontinuous buffer system (10) was employed in this study. Ashton (10) had used a discon— tinuous buffer system for starch—gel electrophoresis of mouse serum proteins, not bovine serum. By two-dimensional analyses, Ashton detected 26 fractions in normal bovine serum: 13 alpha globulins, 8 beta globulins, one gamma globulin zone, one albumin band and two pre albumin bands. This represents two more alpha globulins and one additional pre albumin band than was detected in the present investigation. The buffer systems differed and the breed of cow was not indicated. Calculation of relative percent protein values from starch—gel electrophorograms was unsatisfactory. If the baseline were adjusted too high for some gels (10 counts per inch), the sensitivity was too low and an incomplete scan resulted. If the baseline were too low, the sensitivity was too great. Surface defects and unevenness in the sample were represented as protein fractions. Once the baseline was established for a given sample, repeated scans of the same electrophorogram were identical. Therefore, if the baseline was not constant from gel to gel then the relative percent protein value of a band that apparently remains '1‘??? '57?“ 130 constant may appear considerably increased. This could be partly controlled if a uniform thick gel could be formed and destaining conditions were standardized. Besides some physical errors, such as drift of the densitometer, the chromoscan was very sensitive. In starch— gel media the instrument was capable of indicating each fraction that could be seen visibly. Therefore, it was use— P? ful for detecting any changes in fine bands (post albumins) ' separated in starch—gel. The instrument was unsuitable for 4 rapid scanning of polyacrylamide columns. After tedious i manipulations and scanning the column repeatedly, all bands separated by the differential disc electrophoretic procedure could be detected. However, a densitometer especially de- signed for scanning these gels, such as the model E micro- densitometer (Canalco) would perhaps have given satisfactory scans of these columns. The conventional disc electrophoretic procedure (132) was found unsatisfactory for the separation of bovine serum. The pore size of the separating gel was too small for the adequate separation of serum proteins. The simplicity with which the pore size of polyacrylamide gels can be varied (82, 116, 128, 129, 177) is'a distinct advantage over starchy gels by which differential disc electrophoresis was possible. Since human serum has been extensively employed in establishing separating conditions of most, if not all frac— tionation procedures, it served as a control in this study. 131 The separation of human serum by the standard procedure yields a large number of fine bands closely spaced that migrate only a short distance into the separating gel (gamma globulin zone). When bovine serum as electrophoresed in 7% gel, a wide, densely stained region, the "haze zone," appears a short distance from the origin. Leaving some of the pro- tein at the point of origin is a common difficulty and crowds the slower moving components together (122). The differential disc electrophoretic procedure allows the serum or any complex protein solution to migrate first in a large pore gel (4.75%) which separates the large molecules from the smaller or faster moving components. These smaller molecules, apparently "trapped" by the larger components in the 7% gel subsequently separate in the small pore size gel (10%). Similarly, the large molecular weight components which were trapped in stiffer gels because of almost identical molecular dimensions but differed slightly in charge separate further in the large pore gel. - The length of the upper separating gel (4.75%) was a critical feature. If the distance of migration was too . J short, protein molecules would pile up at the origin of the E 10% gel, which masked other bands; and the post albumin bands were poorly resolved. A length of 3.5 cm was satis- factory for separation of human and bovine serum proteins. 132 Because large quantities of sample could be applied, bands detected only by the differential procedure may not have been I'new" bands, but because of the increase in con- centration, they now could be detected. However, no new bands could be detected when the same sample size was applied to 12 cm tubes containing only 4.75% or 10% separat- hI ing gel. ( Some of the new bands may have been "created" due to 4 splitting or denaturation of certain proteins. Hjerten, Jerstedt, and Tiselius (83) hypothesized that photopolymeri- '- zation of certain ionic mixtures could result in denatura— tion of the sample. When photopolymerized samples were com- pared with samples added directly to the column (49) idene tical patterns developed. Therefore, phot0polymerization did not appear to alter the protein molecules in the serums tested. The possibility still exists that dissociation of the molecules into dimers and trimers occurred during elec— trophoresis. Although, the separation of bovine serum was markedly improved, human serum gave the most striking results. In some cases, twice as many proteins could be detected. This increased resolution becomes extremely important when analyz- ing serums from diseased individuals. The reflection of disease in tuberculous cows was best demonstrated by the new Iprocedure while the standard procedure revealed little except a change in the concentration of bands. 133 The differential system yielded greater separation of serum proteins from a number of different animals but was not suitable for all. The separation of mouse, dog, chicken and pig serum proteins was not improved. It is unlikely that any one combination of separating gels will adequately separate all the fractions of a complex mixture. The elicitation of precipitating antibodies in the rabbit against 32 constitutents of bovine serum indicated that the serum proteins of cattle are as antigenically com- plex as that of man. A nomenclature was proposed by Heremans et a1. (76) from their studies of mouse serum designating the antigenic components detected by immunoelectrophoresis anal— yses. It is based on the electric mobility of protein and the terminology employed in moving boundary electrophoresis. It is a simple, direct method by which a component can be readily identified but does not take precedence over specific naming procedures. The sensitivity of immunoelectrophoresis (IE) is dependent on a number of factors not present with cellulose acetate, starch and acrylamide gel electrophoresis. These factors have been discussed by Crowle (46). Thepotential resolving power of IE is primarily dependent on the antiserum. Therefore, the different responses of individual animals to mixtures of antigen makes it difficult if not impossible to standardize antiserums. This was demonstrated in this study when three rabbits failed to produce antibodies specific for l34 albumin and two failed to produce anti-gamma globulin anti— bodies. This was also observed by Dardas and Mallmann (48) who suggested that unless a variety of animal species is used for antiserum production, all of the antigenic constit- uents of the mixture prObably would not be detected. The multiplicity of immunoPrecipitates in the alphal and alpha regions makes analyses difficult. Absorbed anti- 2 serums would greatly facilitate such analyses. The identifi— cation of some of the components as lipoproteins and glyco- protein substantiated that many of the proteins in human serum have their counterparts in bovine serum. The lipoproteins of human and bovine serum seem to be quantitatively similar. Apart from a fraction which penetrates poorly in the agar gel and presumably corresponds to chylomicrons, there are two main lipoproteins differen- tiated in the serum from both species by IE. The slow com- ponent carries the major part of protein bound lipids and lies between the alpha and beta zones in agar gels (a2-13). The minor component, although not so consistent, is somewhat slower than albumin al-l. In sheep serum, the major lipo- protein is fast and the minor component slow, but the mobilities place these components in the same region as the human and bovine lipoproteins (40). Serums from normal calves were separated into four main fractions by continuous flow electrophoresis. Although the resolution was poor by stained curtain, the gamma 135 globulin, beta globulin and albumin fractions were of un— usually high purity as confirmed by cellulose acetate elec— trophoresis of the eluted concentrated fractions. The only disadvantage being, that only small dilute volumes could be collected. These concentrated fractions of gamma and beta globulins and albumin contributed to the identification of the bands separated by starch-gel, polyacrylamide and immunoelectrophoresis. Each fraction was not isolated and characterized. The nomenclature assigned to cellulose acetate electr0phoro— grams, immunoelectrOphorograms, and starch—gel electrophoro— grams was based only on mobility. Classification on the basis of mObility of the bands separated by differential disc electrophoresis would not be reliable due to the difficulty of correlating the numerous bands with those separated by the other three procedures. The mobility of specific protein can be quite different in these media. For example, the major serum lipoprotein and glycoprotein was a beta globulin in cellulose acetate and alpha globulin in agar immunoelectrOphoresis. This lack of 2 correlation, especially with cellulose acetate and immuno- electrophoresis was a handicap in the analyses of serums from tuberculous cattle. Not until these bands have been isolated and characterized will the method be fully utilized. 136 Analyses 9f Serums from Tuberculous Cattle Since it was desirable to examine and compare all serum samples for each experimental calf at the same time, samples were stored at -20 C and not analyzed by electropho— resis for at least two weeks after the last sample was collected. Freezing or "aging" mouse serums caused altera- tions in the immunoelectrophoretic properties of a beta2 globulin (210). It was possible that freezing or "aging" of bovine serums might cause observable Changes in the electro- phoretic patterns. However, no changes were detected by the methods employed. If any changes did occur, they did not alter the comparative results. The polymorphism in the beta globulin region of the electrophoretic patterns of bovine serums due to breed dif- ferences has been noted. It was possible to distinguish 3—6 beta bands, after starch-gel electrophoresis, in normal serums from the different breeds used in this study: -serum from Guernsey cattle always contained 3 major beta fractions (6a, b, and c); 4-5 bands (6a, b, and c, and 7) were present in the serum from Jersey cattle; and Holstein cattle serum contained, in addition to the above, a 6 band, number 8. Therefore, it is necessary to establish the normal varia— tions before evaluating serum protein changes that occur during disease. Such variations could be erroneously inter- preted as resulting from the disease process. l37 The possibility of other genetically dependent serum variations that might be detected in immunoelectrophorograms also had to be considered. To minimize this problem, the inoculums used.for antiserum production contained a pool of serums from all of the experimental cattle. The most regular and quantitative alterations in the serum proteins in experimental bovine tuberculosis was increase in gamma globulin and decrease in albumin. These changes were as easily demonstrated by the cellulose acetate procedure which was more facile than starch-gel, differen— tial disc electrophoresis and immunoelectrophoresis. The results by all four methods were in close agreement. Since the most consistent and apparent alterations in protein fractions from tuberculous cattle were in the albumin and gamma glObulin, the albumin—gamma ratio (A/G) had practical value. In normal calves (pre-inoculation sample) the average value was 1.4, and slightly lower 1.2, for normal adult cows. The higher gamma globulin concentra- tion in the older animals is probably the result of various antigenic contacts during the longer life span, with resul- tant increased formation of gamma globulin. The A/G values for experimental animals inoculated with g, bovis or 68C and naturally infected animals was con- sistently lower than the normal value indicating an increase in gamma globulin. The only exception to this finding was 138 the increase in the A/G values of post—inoculation serums collected from calf #81. This calf was inoculated sub- cutaneously (SC) with the same fl, bovis culture (#81) that had produced classical tuberculosis in calf #20. Since tuberculosis did not develop in calf #81 as demonstrated by the absence of gross and microscopic lesions, three assump- tions can be made: (1) the virulence of the organisms was reduced, (2) SC route of inoculation does not as readily induce infection, and (3) the resistance to infection in calf #20 was lower than calf #81. The latter is one of the more probable major factors. Some of the calves purchased for the earlier studies were younger and in poor condition. Perhaps because of some metabolic deficiency, the natural resistance mechanisms in calf #20 were less effective. This calf survived only 65 days post-inoculation. In other exper— imental laboratory animals, the intradermal route has gen- erally been more favorable than the subcutaneous route for the sensitization and infection by either a low number of virulent organisms, or by less virulent organisms. (V. A. Mallmann, personal communication). This may be related to the longer retention of organisms at the site of inoculation and the development of delayed sensitivity. Animals inoculated with g, avium (with the exception of calf #21) or 50B, and tuberculin negative cows had normal values or slightly increased. Calf #21, from the same lot as calf #20 and inoculated with g, avium, developed lesions 139 and a lowered A/G ratio. This was the only calf to develop these changes following the inoculation of g, avium, and is probably additional supportive evidence that this group of animals had lowered resistant mechanisms. Animals inoculated with 51C gave inconsistent results and the cellulose acetate electrophorograms were classified into two types: Type I, the A/G ratio was normal or slightly increased and Type II, the ratio was below the normal value. The difference in response of the host to culture 51C could not be compared to the degree of infection as determined by presence and locations of lesions. These differences may be due either to host resis- tance, age, and/or hypersensitivity. Older or adult cows appeared to be more resistant to infection as determined by extent of lesions. All calves with the exception of one (#108) had generalized lesions while the three eXperimental adult cows developed a primary complex without any signs of progressive disease. Host resistance and/or age possibly explains the variation in disease but still does not explain the variation in serum patterns. A possible explanation is that tuberculin caused these variations. This seems prOb- able since an anamnestic-like response, as detected by a marked increase in the HA titer following injection of tuber— culin, was observed inconsistently in the experimental animals (118). 140 Decreasing A/G ratios calculated from cellulose acetate electrophorograms did differentiate experimental infections of g, bovis or 68C which produced gross lesions between the infections of M, avium or 50B in which no gross lesions were detected. The difficulty in using the A/G ratio from cellulose acetate electrophorograms are the incon- sistent values obtained from animals inoculated with 51C. In spite of its shortcomings, the determination of serum protein fractions is useful in the study of bovine tubercu- losis and may serve as prognostic guides. The most striking and consistent change detected by starch-gel electrophoresis was the marked increase in the a2—V fraction of serums from calves after inoculation with g, bovis, 68C, or 51C, and all GL and some NGL animals from naturally infected herds. It may be possible that tuberculin positive and false negative animals infected with virulent mycobacteria can be differentiated from animals giving a false positive response to tuberculin due to the presence of avirulent mycobacteria, by this method. If thin layer starch-gel electrophoresis can be perfected so that the gel would not have to be sliced, reliable A/G values Could be calculated. .A larger sampling of serum from tuberculin negative cows to more accurately determine normal protein values must be determined before this method can be-reliably recommended. Only four tuberculin negative cows were avail— able for this study. 141 Immunoelectrophoresis of post—inoculation serums also demonstrated the marked increase in the a2—V fraction which migrated as an aZ-l. This method was not better than starch—gel electrOphoresis for the detection of specific or nonspecific changes. Accurate interpretations of immuno- electrophoretic results are difficult. Quantitative changes have numerous limitations. There is difficulty in standard- ization of antiserums and differences in displacement of precipitin arcs may be due to changes either in mobility of antigen or antigen concentration. Recently immunoelectro- phoretic capabilities have been extended to quantitative determinations (161). Even if the quantitative values can be obtained, it would not be as practical as starch-gel electrophoresis or perhaps the differential disc electro- phoresis. A combination of polyacrylamide electrophoresis and diffusion in agar gel may be more fruitful. Only pooled rabbit-normal-bovine-antiserums were used to develop electrophoresed serums from tuberculous cattle. Presumably the same specific immunOprecipitates would be present if these serums were deve10ped with anti- serums produced against post—inoculation serum. In addition, these antiserums might detect mycobacterial antigens (either free or in the form of soluble antigen-antibody complexes) if these antigens were present in the serum used for anti- serum production. Post-inoculation serum developed after 142 electrophoresis with anti—post-inoculation serum (containing specific anti-mycobacterial antibodies) could produce a specific precipitate not found in normal serum. Post- inoculation serums could also contain altered host tissue antigens. If so, immunoprecipitates could be formed. The az-V fraction was not isolated and characterized. Certain of its properties can be inferred from its behavior during starch-gel electrophoresis and immunoelectrophoresis. The az-V fraction was a protein antigen which migrated in the alpha globulin region in both starch-gel electrOphoresis 2 and immunoelectrophoresis. It stained with carbohydrate but not with lipid specific stains. It consistently formed a clear precipitate. It was the precipitate farthest from the antiserum trough, the alpha2 (aZ-l). These data suggest that it might be macroglobulin but not a major lipoprotein and that it is present in relatively high concentrations in serums from tuberculous cattle. It also appears to be similar to the alpha -T precipitate described by Dardas and 2 Mallmann (48) found in serums from tuberculous guinea pigs. The alpha -T was not characterized as a macroglobulin and 2 hyperalpha -globulinemia always preceeded its detection. 2 Therefore, the fraction characterized in this study is 2-V and not alphaz-T. These may subsequently “be demonstrated to be essentially identical. Dardas and called alpha .Mallmann (48) also found that hypersensitivity did not stim- ulate the hyperalphaz-globulinemia nor were consistent 143 changes found in the distribution of protein-bound carbohy- drates during experimental tuberculosis. This was also found with serums collected from human tuberculosis patients (79) and fromnaturally infected cows (188). Increases in serum glycoprotein levels may result from many causes (171, 201). Therefore serum glycoprotein determinations cannot be recommended as a diagnostic supplement to tests for tuberculosis of cattle. A simultaneous increase in the slow moving alpha2 (Sa2) globulin lipOprotein with the increase in gamma glob- ulin and the a2-V fraction occurred only in calves inocu- lated with g, bgyis, 68C or 51C. A similar phenomenon was observed by Williams and Wemysis (210) in tuberculous mice, and in tuberculous guinea pigs by Dardas and Mallman (48). The role and mechanism of increased gamma globulin in tuberculosis is not known. If the gamma globulin eleva— tion is related to antibody formation, it is not associated with immunity. Tuberculoimmunity in man or animals has no significant rise in gamma globulin and the immunity can not be passively conferred with serum (44, 175). Many of the tuberculous calves which had high gamma globulin levels 14 days post-inoculation had low HA titers. An increase in gamma globulin may be induced by tuberculin as indicated by a fourfold increase in HA titer following tuberculin testing. The degree of specificity of the antibodies is not known. 144 The increase in titer was Observed in 2 of 4 tuberculin negative cows. This phenomenon is just as difficult to explain as the variation in electrophoretic patterns in animals infected with 51C. The polysaccharides in the tuberculin might affect the amount of polysaccharide- specific antibodies or it may be a stimulation of a broad range such as that elicited by endotoxins. This phenomenon did not occur in animals exposed to an aersol of the orga- nisms although the gamma globulin was markedly elevated. Because no increasing titer could be detected in these animals, immunologic paralysis may have resulted due to the large dose of organisms. Therefore, the increase in gamma globulin appears to be largely due to tissue destruction and not to the excess of circulating antibodies alone. Increased gamma globulin occurs commonly in most hepatic diseases. Liver impairment occurs in advanced tuber- culosis. Therefore the elevation in generalized disease may be due in part to liver alterations. It probably does not account for the increase in gamma globulin in animals which develop only enlargement of the cervical lymph nodes (pri— mary complex). There is increasing evidence that excess gamma globulin is formed not only in the kupffer cells of the liver but by plasma cells and other reticuloendothelial cells in response to tissue destruction (147). This more [probably accounts for the elevated gamma globulin in animals *with only a primary complex. 145 The inability of liver cells to form albumin, possibly because of damage, does not account for the hypo- albuminemia present in animals with a primary complex. It is probable that the low albumin concentration is the result of preferential formation of gamma globulin from available amino acids over the formation of albumin (5). A hyperalpha -globulinemia accompanied an increase 2 in gamma globulin and a decrease in albumin in human tuber- culosis (12, 172, 197) and tuberculous guinea pigs (48, 176). A hyperalpha -globulinemia was not detected in serums from 2 tuberculous calves by cellulose acetate electrophoresis since the alpha band did not separate into two components. Apparently the alpha globulins were masked by the albumin 2 band. Albumin constitutes the major component of serum, therefore if the alpha globulins were masked by albumin, 2 changes in this fraction would not be detected. ’It was evident by starch-gel electrophoresis, differential disc electrophoresis and immunoelectrophoresis that changes in the alpha globulin in serum from tuberculous calves did 2 occur. The increase in only two alpha globulins (Sa and 2 2 V fraction) is additional supportive evidence that these alpha globulins were in low enough concentration to be 2 Inasked by albumin in cellulose acetate electrophorograms. The most widely accepted explanation for the in- crease in alpha globulin that occurs during tuberculosis 2 in man and guinea pigs is that it is mainly a reflection of 146 tissue destruction. Bovornkitti (20) proposed that the increase in alpha globulin was due to the result of delayed 2 hypersensitivity. It did not appear that hypersensitivity induced the increase in the aZ-V fraction and agrees with the assumption that the increase was due to tissue destruc- tion. Some investigators have reported an increase in the beta globulin during tuberculosis (172, 197). This was not demonstrated in bovine tuberculosis. The drop in total serum protein concentration in serum from naturally infected animals following the injec— tion of tuberculin is an interesting phenomenon and diffi- cult to explain. It occurred inconsistently in some of the experimental animals. The decrease in total serum protein did not occur in serum from tuberculin negative cows follow— ing tuberculin testing. Apparently animals that had prior antigenic stimulation deve10ped a mild form of shock to the colloidal osmotic regulatory mechanism due to the secondary stimulation by tuberculin. As a result excess fluid from the tissues entered the circulatory system and diluted the plasma proteins. The decrease in total proteins, however, did not have any effect on the relative protein values of the individual components. This mechanism is conjectural and further investigations are required to establish the significance of this phenomenon. 147 Because other chronic inflamatory diseases may mimic the serum protein changes demonstrated for tuberculosis, electrophoresis would not be adapted to routine use as a laboratory test. The evaluation of the starch-gel electro- phoretic patterns in some problem cases would have diagnos- tic significance, particularly when evaluated with the total clinical picture. The inability to demonstrate circulating antibodies in a large number of cows with bacteriologic and clinical evidence of tuberculosis by a procedure as sensitive as the passive hemagglutination test is paradoxical. The concept that antibody activity associated with human serum protein may be inhibited by other proteins or substances present in the serum proved fruitful. Gerstl et a1. (60, 61) using ethanol fractions of serum from humans infected with tuber— culosis found antibody activity in one or several fractions which as a whole had failed to react. However the incom- plete separation of proteins accomplished by this fraction— ation procedure did not permit exact localization of anti- body activity. Fractionation of whole serum from humans infected with tuberculosis by continuous flow electrOphore- sis revealed antibody activity in the alpha fraction (2). 2 Turcotte et al. (192) used DEAE-cellulose to separate the antibodies responsible for the hemagglutinating activity of serums from healthy tuberculin positive individuals and 148 individuals with active pulmonary tuberculosis. The latter investigation revealed that false positive HA titers were possible. The high degree of separation and resolution of serum proteins by starch-gel and differential disc electro- phoresis makes it even more possible to detect mycobacterial antigens and anti-micobacterial antibodies in serums from individuals suspected of having tuberculosis. ,Such studies might detect separate fractions that contain diagnostic anti- bodies or antigen-antibody complexes. The possibility exists that the a2-V fraction contains specific anti-myco- bacterial antibodies. If any antibody is partially respon- sible for immunity in tuberculosis, it may occur in some fraction other than gamma globulin. The chance of uncover- ing these antibodies will depend largely on the sensitivity of the fractionation procedure. The more components that can be separated the greater the chance of finding a specific component. The differential disc procedure, although still eXperimental, may meet these requirements. Such studies can also be used to investigate extracts from white blood cells such as the transfer factor. Results from either or both can contribute to our understanding of the "ecology of tuberculosis," the interactions of the mycobacteria and their hosts. awn-m ’ - ...... .§ EL. S UMMA RY Modified techniques and apparatus for starch-gel electrophoresis and for the rapid staining and destaining of both starch and disc gels is described. The mechanical manipulation of starch-gels during the preparation and slicing procedures was facilitated. The need of handling either types of gels during the staining and destaining processes was eliminated. A modified polyacrylamide electrophoretic procedure is described which separated human serums into 44-50 compo— nents and bovine serum into over 30 bands.' In this proce- dure, a complex ionic mixture migrates through both a large and a small pore size gel in a single column. This proce- dure, tentatively named differential disc electrophoresis, is simple, facile, and affords improved separation and resolution of serum proteins. Serums from normal cattle were analyzed by the elec— trophoretic methods of continuous flow, cellulose acetate, starch-gel, two—dimensional starch-gel, agar gel, immuno- electrophoresis, starch-gel immunoelectrophoresis, and differential disc electrophoresis. Differential disc electrophoresis was the most sensitive single procedure. 149 150 Thirty to 38 fractions of bovine serum were observed by this technique. Thirty-two antigens were found in normal serum by immunoelectrophoresis and a standard nomenclature was proposed. The antigenic heterogeneity of the majority of the protein bands separated by starch-gel electrophoresis was demonstrated. A few of the bands separated by these procedures were identified as lipoproteins and glycoproteins. A hypergammaglobulinemia and a decrease in albumin was detected by all electrophoretic procedures in the serums from experimental calves inoculated with Mycobacterium bovis or 68C (virulent Group III) 7-14 days post-inoculation and naturally infected cows, gross lesion and no gross lesion, prior to tuberculin testing. Inconsistent changes were detected in cellulose acetate electrophorograms of serums from calves inoculated with 51C (virulent Group III). No changes in the electro— phorograms of serum from exPerimental animals inoculated with M, avium or 50B (avirulent Group III) or tuberculin negative cows were detected. The hypergammaglobulinemia and passive-hemagglutination anamnestic-like response to tuber- culin did not consistently occur simultaneously. The A/G ratio calculated from cellulose acetate electrophorograms was practical in differentiating experi- mental cattle inoculated with virulent strains or species (M, bovis, 68C) from cattle inoculated with avirulent 151 acid-fast bacilli (M, avium, 50B). However, cattle inocu- lated with 51C could not be differentiated by this method. No consistent differences occurred in the protein- bound carbohydrates. An increase in the slow moving alphaz— l3 globulin lipoprotein (detected by immunoelectrophoresis and differential disc electrophoresis) occurred only in the serum from calves infected with virulent mycobacteria. Starch—gel, differential disc electrophoresis and immunoelectrophoresis revealed an alpha2 fraction that was always increased in serums from experimental animals inocu- lated with M, bovis, 68C or 51C and pretuberculin serums from gross lesion (GL) and no-gross lesion (NGL) cows from naturally infected herds. This fraction remained unchanged in serums from calves inoculated.with M, avium or 50B and therefore has been tentatively named the a'+V (virulence) 2 fraction. This fraction was most reliably detected by the starch—gel procedure. Further definition and characteriza- tion of the individual fractions and a reliable densitomet- ric procedure is needed before the full potential of differ— ential disc electrophoresis can be fully appreciated in the analysis of serums from pathological disorders. The hypergamma- and hyperalpha -globulinemia may 2 result from tissue destruction, circulating antibodies and in a few cases from delayed hypersensitivity. 152 .A total protein decrease was marked in serums from all GL and NGL cows from naturally infected herds following the injection of tuberculin. The total protein remained constant in serums from tuberculin negative cows. It was suggested that a secondary stimulation by tuberculin caused a mild form of shock to the colloidal osmotic regulatory mechanism so that excess fluid from the tissues entered the circulatory system diluting the plasma proteins. 10. 11. LITERATURE CITED Aaronson, T., and A. Gronwal. 1958. Electrophoretic separation of serum proteins into twelve fractions. Scand. J. Clin. Lab. Invest. 10:348. Allerhand, J., and C. M. Zitrin. 1962. Detection of antibodies in the alpha and gamma globulin frac- tions of human tubercul us sera. J. Immunol. 89: 252-256. Allison, A. C. 1959. Identification of human serum proteins binding iron, copper, and thyroid hormones by starch gel electrophoresis. Experimentia (Basel) 15:281-284. Analytrol Model RB, Spinco Division, Beckman Instruments Inc., Palo Alto, Calif. Anker, H. S. 1960. The biosynthesis of plasma proteins. pp. 267-307. In F. W. Putnam, The Plasma Proteins. Vol. II. Academic Press, New YOrk. Ashton, G. C. 1957a. Serum protein differences in cattle by starch gel electrophoresis. Nature 180: 917-919. Ashton, G. C. 1957b. Zone electrophoresis of mammalian sera in starch gel. Nature 179:824-825. Ashton, G. C. 1958. Genetics of beta-globulin poly- morphism in British cattle. (Nature 182:370-372. Ashton, G. C. 1960. Beta-globulin polymorphism and economic factors in dairy cattle. J. Agric. Soc. 54:321-328. Ashton, G. C., and A. W. Braden. 1961. Serum beta- globulin polymorphism in mice. Aust. J. Biol. Sc. 14:248-253. Askonas, B. A., C. P. Farthing, and J. H. Humphrey. 1960. The significance of multiple antibody compo- nents in serum of immunized rabbits. IImmunology 3:336—351. 153 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 154 Baldwin, R. W., and C. W. Iland. 1953. Electrophoretic studies of the serum proteins in tuberculosis. Am. Rev. Tuberc. 68:372-381. Belfrage, S. 1963. Plasma protein patterns in the course of acute infectious diseases. .Acta Med. Scand., Suppl. 395. Berlin, N. E., R. Merwin, M. Potter, J. L. Fahey, P. P. Carbone, and M. J. Cline. 1963. ,Neoplastic plasma cell. Combined Clinical Staff Conference at the National Institutes of Health. Ann. Int. Med. 58: 1017-1036. Bier, M. 1959. Electrophoresis-—Theories, Methods, and Applications. Academic Press Inc., New York. Block, R. J., E. L. Durrum, and G. A. Zweig. 1955. A Manual g£_Paper Chromatography and Paper Electro— phoresis. Academic Press Inc., New York. Bodman, J. 1960. Agar gel, starch block, starch gel, and sponge rubber electrophoresis. pp. 91-157. In I. Smith (ed.) Chromatographic and Electro- phoretic Techniques. Vol. II. Interscience Publishers, Inc., New York. Boguth, W., F. Habermalz, and H. Schnappauf. 1959a. Papierelektrophoretische bestimmung der proteinge- bundemen kohlenhydrate in serum von haus— und laboratoriumstieren. I. Mitteilung. Zent. fur Vet. Med. 6:634-639. Boguth, W., F. Habermalz, and H. Schnappauf. 1959b. Papierelektrophoretische bestimmung der proteinge- bundemen kohlenhydrate in serum von haus- und laboratoriumstieren. II. Mitteilung..Zent. fur Vet. Med. 6:901-912. Bovornkitti, S. 1962. Serum protein changes in tuber- culosis in humans. With particular reference to alpha2 globulin. Am. Rev. Resp. Dis. 85:58-65. Bovornkitti, S., P. Visutakul, and P..Kangsda1. 1961. The serum proteins in tuberculin—negative and tuberculin-positive healthy persons. Am. Rev. Resp. Dis. 84:93-94. Boyd, W. C. 1956. Fundamentals g; Immunology. 3rd ed. Interscience Publishers, Inc., New York. Full 1- ... . ... ,. 1‘. “I? - 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 155 Boyer, S. H., and R. Hiner. 1963. Modified apparatus for starch gel electrophoresis. J. Lab. Clin. Med. 61:879-882. Brackenridge, C. J. 1960. Optimal fractionation conditions for the quantitative analysis of human serum protein fractions by cellulose acetate electrophoresis. Anal. Cham. 32:1357-1359. Brackenridge, C. R., and E. R. Csillag. 1962. A quantitative electrophoretic survey of serum protein fractions in health and disease. Acta Med. Scand. 172:383. ' Bradish, C. J., W. M. Henderson, and J. B. Brooksby. 1954. Electrophoretic studies of ox serum. I. The sera of normal cattle. Biochem. J. 56:329-335. Bradish, C. J., W. M. Henderson, and J. B. Brooksby. 1954. Electrophoretic studies of ox serum II. The sera of cattle infected with the virus of foot—and—mouth disease. Biochem. J. 56:335-341. Brown, F. 1960. A beta—globulin antibody in the sera of guinea pigs and cattle infected with foot-and- mouth disease. J. Immunol. 85:298-303. Brown, F., and J. H. Graves. 1959. Changes in specificity and electrophoretic mobility of the precipitating antibodies in the serum of cattle recovering from foot-and-mouth disease. .Nature 183:1688-1689. Brummerstedt-Hansen, E., J. Moustgaard, and I. Moller. 1963. Serum-transferrin—polymorphism in Danish cattle breeds. Yearbook, Royal Veterinary and ,Agril. College, Copenhagen, Denmark. pp. 13-26. Budd, R. E. 1961. Calculation of serum total protein from the integrator trace of scanned paper elec- trophoresis strips and comparison with a chemical method. Am. J. Med. Tech. 27:215—218. Butler, E. A., F. V. Flynn, H. Harris, and E. B. Robson. 1961. The laboratory diagnosis of macroglobulinemia. With special reference to starch-gel electrophoresis. Lancet 2:289-292. ' Campbell, D. H., J. S. Garvey, N. E. Cremer, and D. H. Susdorf. 1963. Methods ig Immunology. W. A. Benjamin Inc., New York. 34. 35. 36. 37. 38° 39. 40. 41. 42. 43. 44. 45. 156 Campbell, E. A- 1957. The use of paper electrophore— sis as an aid to diagnosis. J. Comp- Path. and Therap. 67:345-353. ,Canalco- 1963. Canal Industrial Corporation, Bethesda 14, Maryland- Carson, P. D. A-, and J. M. Mattingley. 1960. Electro— phoresis and its applications to clinical medicine. New Zealand Med. J. 59:159-164. Cawley, L. P. 1963. Changing concepts in protein electrophoresis. J. Kan- Med. Soc. 64:470-476. Cawley, L. P., and L. Eberhardt. 1962. Simplified gel electrophoresis. I. Rapid technic applicable to the clinical laboratory. Am. J. Clin. Path. 38: 539-547. Chievitz, E., and T. Thiede. 1960. Electrophoretic study of the serum proteins in patients with pulmonary tuberculosis. A series of 122 patients admitted to the Anstrop Sanatorium. Acta Tuberc. Scand. 39:270-285. Chordi, A., and I. G. Kagan. 1964. Analysis of normal sheep serum by immunoelectrophoresis. J. Immunol. 93:439-445. Choremis, C., D. Liacacos, and B. Anghelia. 1959. Serum protein electrophoretic patterns and Middle- brook—Dubos titer in tuberculosis children follow— ing BCG vaccination. Am. Rev. Tuberc. 79:522-524. Consolidated Laboratories, Inc. (Colab). 1960. Electrophoresis with Oxoid cellulose acetate strips. Chicago Heights, Illinois. Cooper, G. R. 1960. Electrophoretic and ultracentrif- ugal analysis of normal human serum. p. 51-103. In F. W. Putnam (ed.), The Plasma Proteins. Vol. II. Academic Press, New York. Crowle, A. J. 1958. Immunizing constitutents of the tubercle bacillus. Bact. Rev. 22:183-203. Crowle, A. J. 1960. Four modifications of the micro agar diffusion precipitation test. J. Lab. Clin. Med. 55:593-604. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 157 Crowle, A. J. 1961. Immunodiffusion. Academic Press, New York. Czernicki, B. 1957. Elektrophoretische untersuchungen an rinderserum. Zentralblatt fur Vet. Med. 4:588- 602. Dardas, T. J. and V. A. Mallmann. 1966. Chromato— graphic, electrophoretic, and immunoelectrophoretic studies of serums from normal and tuberculous guinea pigs. J. Bacteriol. In press. Davis, B. J. 1964. Disc electrophoresis. II. Method and application to human serum proteins. Ann. New York Acad. Sc. 121:404-427. Davis, B. J., and L. Ornstein. .A new high resolution electrophoresis method. Delivered at the Society for the Study of Blood, March 24, 1959, at the New York Academy of Medicine. Dimopoullos, G. T., G. T. Schrader, and L. E. Foote. 1960. Electrophoretic studies of bovine serum. III. Serum protein changes in anaplasmosis. Am. J. Vet. Res. 21:222-225. Dubos, R. J. 1954. Unsolved problems in tuberculosis. Am. Rev. Tuberc. 70:391-401. Dubos, R. J. 1961. Evolution of concepts in the prevention of tuberculosis. The 1961 Baker Lecture delivered before the School of Public Health, University of Michigan, December 4, 1961. Efron, M. 1960. High voltage paper electrophoresis. pp. 158-189. In I. Smith (ed.), Chromatographic and Electrgphoretic Techniques. Vol. II. Inter— science Publishers, Inc., New York. Engle, R. L., K. R. Woods, G. B. Castillo, and J. H. Pert. 1961. Starch gel electrophoresis of serum and urinary proteins from patients with multiple myeloma, macroglobulinemia and other forms of disproteinemia. J. Lab. and Clin. Med. 58:1. Ferris, T. G., R. E. Easterling, and R. E. Budd. 1963. Electrophoresis of serum proteins in acrylamide gel. II. An improved apparatus and technic for the removal of background stain, by means of an electric current. Am. J. Clin. Path. 39:193-197. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 158 Freigang, B., R. P. Boyd, and G. B. Elliott. 1963. Serum protein electrophoresis in tuberculosis. Canad. Med. Assoc. J. 88:240-242. Gelman Instrument Company, Ann Arbor, Michigan. 1964. Technical bulletin 102A and 103A. Gendon, Y. X. 1958. The agar gel diffusion precipita- tion method and its used in immunochemical analysis of antigens. Microbiol. Epiden. Immunobiol. 29: 416-423. Gerstl, B., D. Kirsch, W. E. Davis,Jr., and M. Barbieri. 1954. Absence of circulating antibodies in patients with pulmonary tuberculosis. Science 120:853-854. Gerstl, B., W. E. Davis Jr., D. Kirsch, A. G. Hollander, M. Barbieri, and S. B. Weinstein° 1955. Detection of apparently absent circulating antibodies in tuberculous sera. Am. Rev. Tuberc. 72:345-355. Gilliland, I. C., R. N. Johnston, P. Stradling, and E. M. Abdel-Wahab. 1956. Serum proteins in pulmonary tuberculosis. Brit. Med. J. 1:1460-1464. Goldberg, C. A. J. 1959. A discontinuous buffer sys- tem for paper electrophoresis of human haemoglobin. Clin. Chem. 5:446-452. Grabar, P. 1959. Immunoelectrophoretic analysis. Meth. Biochem. Analysis. 7:1-38. Grabar, P., and C. A. Williams. 1953. Methode‘- permettant l'etude conjugee des proprietes elec— trophoretiques et immunochimiques d'un melange de proteines. Application au serum sanguin. Biochim. Biophys, Acta 10:193-194. Cited from Quchterlony, O. 1962. Diffusion—in-gel methods for immunological analysis. II. Progr. ALdergYI Iv. Basel. Graham, J. L., and B. W. Grunbaum. 1963. IA rapid method for microelectrophoresis and quantitation of hemoglobins on cellulose acetate. Am. J. Clin. Path. 37:567-578. Gratzer, W. B., and G. H. Beaven. 1960. Transparent starch gels: Preparation, Optical properties and application to haemoglobin characterisation. Clin. Chim. Acta 5:577-582. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 159 Grigorieva, V. I., and N. A. Livshitz. 1960. Blood serum proteins in tuberculous patients. Lab. Delo. 1:8. Abstract in Am. Rev. Resp. Dis. 82:441. Gross, P. A. M., D. Gitlin, and C. A. Janeway. 1959. The gamma globulins and their clinical significance. New Eng. Med. 260:121-125. Grunbaum, B. W., P. L. Kirk, and W. A. Atchley. 1960. Microelectrophoresis on cellulose acetate membranes. Anal. Chem. 32:1361-1362. Gutman, A. B. 1948. The plasma proteins in disease. Adv. in Prptein Chem. 4:155-250. Guyton, A. C. 1961. Textbook 9: Medical Physiology. 2nd ed. W. B. Saunders Company, Philadelphia. Hansl, Jr., R. 1964. Review of some new applications for disc electrophoresis and of the criteria for quantitative densitometry. Ann. New York Acad. Sc. 121:391—403. Harrower, J. R., M. T. Kahn, M. L. Becker, and E. J. Shabart. 1957. Preliminary observations with the use of paper electrophoresis in studying serum protein changes as an index of progress in the tuberculous patient. Am. Rev. Tuberc. 76:792-795. Heremans, J. F. 1960. Les globulines seriques du systems gamma. Leur nature et leur pathologie. Masson (Paris). Cited from Belfrage, S. 1963. Plasma protein patterns in the course of acute infectious diseases. Acta Med..Scand., Suppl. 395. Heremans, J., J. Clausen, M. T. Heremans, and R. Rask- Nielsen. 1959. ImmunoelectrOphoretic character- istics of normal mouse serums as a basis for studying pathological changes in serums of mice carrying transplantable malignant growths. J. Nat. Cancer Inst. 22:45—55. Hever, O. and E. Kalnai-Hever. 1962. Haptoglobin values in patients with tuberculosis and their modifications by factors injurious to the liver. Prot. Biol. Fluids: 201-204. Hiller, O. 1963. Madison, Wisconsin. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 160 Hirsch, A. and Cattaneo. 1957. Protein-bound carbohy- drates of blood serum in pulmonary tuberculosis. Am. Rev. Tuberc. 75:793-806. Hirschfeld, J. 1960. Immunoelectrophoresis-procedure and application to the study of group specific variations in sera. Science Tools 7:18-25. Hirschfeld, J. 1962. The use of immunoelectrophoresis in the analysis of normal sera and in studies of the inheritance of certain serum proteins. Science Tools 8:17-24. Hjerten, S. 1962. Molecular sieve chromatography on polyacrylamide gels, prepared according to a simplified method. Arch. Biochem. Biophys. Suppl. 1:147-151. Hjerten, S., S. Jerstedt, and A. Tiselius. 1965. Some aspects of the use of "continuous" and "dis- continuous" buffer systems in polyacrylamide gel electrophoresis. Anal. Biochem. 11:219-223. Hogness, K. R., J. W. Giffee, and V. L. Koenig. 1946. Electrophoretic analysis of bovine plasma and serum. Arch. Biochem. 10:281-289., Hudgins, P. C., and R. A. Patnode. 1957. Electro- phoretic distribution of serum protein and glyco- protein in the tuberculous rat, rabbit, and guinea pig. Proc. Soc. Exptl. Biol. Med. 95:181-183. Huneeus-Cox, F. 1964. Electrophoretic and immunolog- ical studies of squid axoplasm proteins. IScience 143:1036-1037. Hunter, R. L., and C. L. Markert. 1957. Histochemical demonstration of enzymes separated by zone electro- phoresis in starch gels. Science 125:1294-1295. Itano, H. A. 1953. Human hemoglobin. Science 117: 89-94. - Jenks, W. P., E. R. Smith, and E. L. Durrum. 1956. Clinical significance of analysis of serum protein distribution by filter paper electrophoresis. Am. J. Med. 21:387-405. Johnston, R. N., and I. Steven. 1958. Zone electro- phoresis in starch gels. A report on Smithies method in normal adults and in patients with tuberculosis. Am. Rev. Tuberc. 78:932-933. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 161 Kabara, J. J., D. Zyskowski, and H. Pokornicki. 1964. A new method for scanning electrophorograms. Clin. Chem. 10:950—958. Kabat, E. Q., and M. M. Mayer. 1961. .Experimental Immunochemistry. 2nd ed. Charles C. Thomas Publishers, Springfield, Ill. Ketz, H. A. 1959. Die Bedeutung und Zusammensetzung der Eiwei B-Fraktionen in Serumxund Plasma beim Schaf. Archiv fur Exptl. Vet. 13:508-521. Kohn, J. 1957a. A cellulose acetate supporting medium for zone electrophoresis. Clin. Chem. Acta 2:297- 303. Kohn, J. 1957b. An immunoelectrophoretic technique. Nature 180:986—987. Kohn, J. 1958. A micro-electrophoretic method. Nature 181:839-840. Kohn, J. 1960. Cellulose acetate electrophoresis and immunodiffusion techniques, pp. 56-90. In I. Smith (ed.), Chromatographic and Electrophoretic Techniqges, Vol. II. Zone Electrophoresis. Inter- science Publishers, Inc., New York. Kohn, J. 1961. Immunodiffusion technique on cellulose acetate. Prot. biol. Fluids 120-122. Korngold, L. 1963. A simple micro starch gel electro- phoresis method applicable to immunodiffusion. Analytical Biochem. 6:47-53. Korngold, L., and G. VanLeeuwen.' 1957. Macroglobulin- emia. J. EXptl. Med. 106:467-476. Kreutzer, H. H., and W. H. S. Fennis. 1964. Washing of agar for agar gel electrophoresis. Clin. Chem. Acta 9:404-406. Kries, B., M. F. Jayle, P. Clerc, and H. Lutier. 1955. La pertubation des proteins du plasma au cours de la tuberculose pulmonaire. Rev. Tuberc. 19:187-196. Kunkel, H. G. 1954. Zone electrophoresis. Meth. of Biochem. Anal. I:14l-170. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 162 Latner, A. L., and A. H. Zaki. 1957. Starch-gel electrophoresis of animal sera. Nature 180:1366- 1367. Latner, A. L., and A. H. Zaki. 1960. Clinical uses of starch-gel electrophoresis with special ref- erence to leukaemia. Clin. Chim. Acta 5:22-25. Laurell, C. B., and S. Laurell. 1955. Electrophoresis of plasma proteins. Lancet 269:40-41. Laurell, C. B., S. Laurell, and N. Skoog. 1956. Buffer composition in paper electrophoresis. Clin. Chem. 2:99-111. Laurell, C. B., and B. Lundh. 1962. An electropho- retic method for estimation of some beta globulins. Scand. J. Clin. Lab. Invest. 14:490-494. Lederer, M. 1955. .53 Introduction 32 Paper Electro— phoresis and Related Methods. Elsevier Publishing Company, New York. Lever, W. F. 1951. Electrophoretic analysis of plasma proteins in various diseases. Bull. New Eng. M. Center. 13:160-174. Lewis, L. A. 1960. Electrophoresis £2 Physiology. Charles C. Thomas Publishers, Springfield, Ill. L. K. B. Instruments Inc. 1962. Washington, D. C. Long, E. R. 1958. The Chemistry and Chemotherapy 9f Tuberculosis. 3rd ed. The Williams and Wilkins Company, Baltimore, Maryland. Longsworth, L. G., T. Shedlovsky, and D. A. MacInnes. 1939. Electrophoretic patterns of normal and pathological blood serum and plasma. J. Exptl. Med. 70:399-413. Luetscher, Jr., J. A. 1947. Biological and medical applications of electrophoresis. Physiol. Rev. 27:621-642. Lytle, I. M., and J. G. Haskell. 1964. Human chrorionic gonadotropin: Isolation by disc electrophoresis. Am. J. Obstetrics and Gynecology 89:156-159. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 163 Maher, J. R., T. B. Miller, R. S. Fraser, and A. F. Lincoln. 1957. Serum paper electrophoretic patterns of tuberculous patients presenting therapeutic problems. Am. Rev. Tuberc. 75:999- 1002. Mallmann, V. A., P. Robinson, and M. D. McGavin. 1964. An anamestic-like reaction elecited by tuberculin. Am. J. Vet. Res. 25:693-698. Mallmann, W. L., V. H. Mallmann, and J. A. Ray. 1964. Bovine tuberculosis: What we know, do not know, and need to know in order to eradicate. Proc. U. S. Livestock San. Assoc. 68th Anal. Mtg. pp. 327-332. Manual for spinco continuous flow electrophoresis. Spinco Division, Beckman Instruments, Inc., Palo Alto, Calif. Marsh, C. L., C. R. Jolliff, and L. C. Payne. 1964. A rapid micromethod for starch-gel electrophoresis. Am. J. Clin. Path. 41:217-223. Matson, C. F. 1965. Pllacrylamide Gel Electrophoresis. Gel E1ectrophoresis.. A simple system using gel columns. Anal. Biochem. 13:294-304. Matsui, K., and K. Yaemo. 1963. Quantitative starch gel electrophoresis. Anal. Biochem. 6:491-503. McCuiston, C. F., and P. C. Hudgins. 1960. Serum electrophoresis in sarcoidosis, tuberculosis, and diseases due to unclassified mycobacteria. Am. Rev. Resp. Dis. 82:59-63. Moore, D. H., J. B. Roberts, M. Costello, and T. W. Schonberger° 1949. Factors influencing the electrophoretic analysis of human serum. J. Biol. Chem. 180:1147-1158. Moretti, J., G. Boussier, M. P. Hugou, and L. Hartman. 1959. Recherche des correspondances entre Les resultats des electrophoreses sur papier, a travere un gel diamiden et de L'immune-electrophorese en gelose des proteines du serum. Bull. Soc. Chim. Biol. 41:79-87. Muller, G. L. 1943. Clinical Significance gf the Blood lg Tuberculosis. The Commonwealth Fund, New York, p. 516. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 164 Namba, T., A. H. Wolintz, and D. Grob. 1963. .Acrylamide gel electrophoresis of serum and skeletal muscle extract. Fed. Proc. 22:453. Narayan, K. A., S. Narayan, and F. A. Kummerow. 1965. Disc electrOphoresis of human serum lipoproteins. Nature 205:246-248. Nilsson, T. 1958. Blood and milk proteins and protein-bound carbohydrates in bovine chronic mastitis: A paper electrophoretic study. .Acta Path. Micro. Scand. Suppl. 125:1-74. Nyman, M. 1959. .Serum Haptoglobin. Scand. J. Clin. Lab. Invest. Suppl. p. 39. Ornstein, L., and B. J. Davis. 1961. Disc electro- phoresis. Distillation Products Industries, Rochester, New York. P. 63. Owen, J. A. 1956. Determination of serum protein fractions by zone electrophoresis. Analyst° 81: 26. Owen, J. A. 1958. Paper electrOphoresis of proteins and protein-bound substances in clinical investi- gations. Adv. in Clin. Chem. 1:237-300. Park, Y. J. 1961. Studies on the serum electropho- retic patterns in pulmonary tuberculosis. Korean J. of Internal. Med. 4:463-476. Pert, J. H., and L. Pinteric. 1964. Quantitation of starch gel electrophoresis. Ann. New York Acad. Sc. 121:310-320. Petermann, M. L. 1960. Alterations in plasma protein patterns in disease, pp. 309-343. In F. W. Putnam (ed.) The Plasma Proteins. Vol. II. Academic Press, New York. Pierce, A. E. 1955. Electrophoretic and immunological studies of sera from calves from birth to weaning I. Electrophoretic studies. J. Hygiene. 53:247- 260. Pilheu, J. A., J. A. Iannello, R. B. Torrent, and J. A. Willson. 1962. Serum proteins in pulmonary tuberculosis. Dis. Chest. 41:173-179. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. isz. 165 Poulik, M. D. 1952. Filter paper electrophoresis of purified diphtheria toxoid. Canad. J. Med. Sc. 30:417-419. Poulik, M. D. 1957. Starch gel electrophoresis in a discontinuous system of buffers. Nature 180:1477- 1479. Poulik, M. D. 1959. Starch gel immunoelectrophoresis. J. Immunol. 82:502-515. Poulik, M. D., and O. Smithies. 1958. Comparison and combination of the starch gel and filter paper electrophoretic methods applied to human sera: Two-dimensional electrophoresis. Biochem. J. 68:636-643. Price, W. H., and H. Harrison. 1964. Disc electrOpho- resis on polyacrylamide gels of serum mucoids of individuals with selected chronic diseases. Ann. New York Acad. Sc. 121:460-469. Proom, H. 1943. The preparation of precipitating sera for the identification of animal species. J. Path. Bact. 55:419-426. Putnam, F. W. 1960. Abnormal serum globulins, pp. 345-406. In F. W. Putnam (ed.) The Plasma Proteins, Vol. II. Academic Press, New York. Raffel, S. 1961. Immunity. 2nd ed. Appleton- Century—Crofts, Inc., New York. Rausch, W. H., T. M. Ludwick, and D. F. Weseli. 1965. Determination of bovine transferrin types by disc electrophoresis. J. Dairy Sc. 48:720-727. Raymond, S. 19620 A convenient apparatus for gel electrophoresis. Clin. Chem. 82455-470. Raymond, S. 1964. Acrylamide gel electrophoresis. Ann. New York Acad. Sc. 121:350-365. Raymond, S., and B. Aurell. 1962. Two-dimensional gel electrophoresis. Science 138:152-153. Raymond, S., and M. Nakamichi. 1962. Electrophoresis in synthetic gels. Relation of gel structure to resolution. Anal. Biochem. 3:23-30. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 166 Raymond, S., M. Nakamichi, and A. Barbro. 1962. Acrylamide gel as an electrophoresis medium. Nature 195:697-698. Raymond, S., and Yi-Ju Wang. 1960. Preparation and properties of acrylamide gel for use in electro- phoresis. Anal. Biochem. 12391-396. Raymond, S., and L. Weintraub. 1959. Acrylamide gel as a supporting medium for zone electrophoresis. Science 130:711. Reiner, M. 1957. Role of electrophoresis in medicine. Internat. Rec. Med. 170:374-387. Riberiro, L. P., E. Mitidieri, and O. R. Affonso. 1961. Paper Electrophoresis. Elsevier Publishing Company, Amsterdam, p. 441. Rogers, T. E., and G. T. Dimopoullos. 1962. Electro- phoretic studies of bovine serum. V. Complement- fixing antibodies in serum fractions in anaplas- mosis. Proc. Soc. EXp. Biol. Med. 110:362-365. Rooney, J. R. 1957. Paper electrophoresis of bovine serum. Am. J. Vet. Res. 18:67-72. Rubinstein, H. J., J. A. Oven, and W. Larsen. 1960. Measurements of proteins separated by electropho- resis in starch gel. Clin. Chim. Acta 5:583. Russell, W. J. 1966. Some influences of antigen concentration and nonreacting additives on mobility and diffusion in immunoelectrophoresis. J. of Immunol. 95:1142-1146. San Clemente, C. L., and I. F. Huddleson. 1943. Electrophoretic studies of the proteins with respect to brucella. Mich. State College, Agric. EXper. Sta. Tech. Bull. 182:3-45. Schaffner, F., G. C. Turner, D. E. Eshbaugh, W. B. Bukingham, and H. Popper. 1953. Hypergamma- globulinemia in tuberculosis. Arch. Int. Med. 92:490-493. Scheidegger, J. J. 1955. Une micro-methode de l'immuno-electrophorese Int. Arch. AllergYI 7: 103-110. Cited from Quchterlony, O. 1962. Diffusion-in-gel methods for immunological analysis. II. AllergyI IV. Basel. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 167 Scherr, G. H. 1961. Cellulose acetate electrophoresis in microbiology and immunology. The New York Acad. of Sciences. 23:519-530. Schulz, D. M. 1964. Methods and applications of zone electrophoresis. Jour. Ind. St. Med. Assn. 57:28- 32. Seibert, F. B. 1960. A theory of immunity in tuber- culosis. Prospectives in Biology. 3:264-281. Seibert, F. B., and J. W. Nelson. 1942. Electrpho- retic study of the blood protein response in tuberculosis. J. Biol. Chem. 143:29-38. Seibert, F. B., and J. W. Nelson. 1943. Electropho- resis and serum proteins in tuberculosis and other chronic diseases. Am. Rev. Tuberc. 47:66-67. Seibert, F. B., J. W. Nelson, and M. V. Seibert. 1943. Correlation of the extent of tuberculosis with the amount of polysaccharide in serum. Proc. Soc. Exptl. Biol. Med. 52:219-222. Seibert, F. B., M. L. Pfaff, and M. V. Seibert. 1948. A serum polysaccharide in tuberculosis and carcinoma. Arch. Biochem. 18:379-395. Seibert, F. B., M. V. Seibert, A. J. Atno, and H. W. Campbell. 1947. Variation in protein, and poly- saccharide content of sera in chronic disease, tuberculosis, sarcoidosis, and carcinoma. J. Clin. Invest. 26:90-102. Shandon. 1960. Apparatus for electrophoresis. Consolidated Laboratories, Chicago Heights, Ill. Shaw, B. D. 19560 Serum proteins in pulmonary tuberculosis. Brit. Med. J. 2:47. Sher, B. C., A. Dubin, Y. Takimura, J. DeLaHurega, and H. Popper. 1958. Serum protein changes in experimental tuberculosis. Am. Rev. Tuberc. 77: 120-133. Sher, B. C., A. Dubin, Y. Takimura, and H. Popper. 1956. Serum protein-bound carbohydrates and lipids in experimental tuberculosis. Proc. Soc. EXptl. Biol. Med. 93:578-581. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 168 Slater, G. G. 1965. Electrophoresis on a density gradient of polyacrylamide gel. Fed. Proc. 24: 225. Small, M. J. 1950. Serum gamma-globulins in pulmonary tuberculosis. Am. Rev. Tuberc. 61:893-894. Smithies, O. 1955. Zone electrophoresis in starch gels: Group variations in serum proteins of normal human adults. Biochem. J. 61:629-641. Smithies, O. 1959. An improved procedure for starch- gel electrophoresis. Further variations in the serum proteins of normal individuals. Biochem. J. 71:585-587. Smithies, O. 1962. Molecular size and starch-gel electrophoresis. Arch. Biochem. Biophys. Suppl. 1:125-131. Smithies, O. 1964. Starch-gel electrophoresis. Metabolism 13:974-983. Smithies, O., G. E. Connell, and G. H. Dixon. 1962. Inheritance of haptoglobin subtypes. Am. J. Human Genetics 14:14-21. Smithies, O., and C. G. Hickman. 1958. Inherited variations in the serum proteins of cattle. Genetics 43:374-385. Smithies, O., and N. F. Walker. 1956. Notation for serum-protein groups and the genes controlling their inheritance. Nature 178:694-695. Sterling, K. 1951. Plasma proteins and their abnormalities in disease: ‘Review- Bull. New Eng. M. Center 13:121-134. Sunderman, Jr., F. W., and F. W. Sunderman. 1957. Clinical applications of the fractionation of serum proteins by paper electrophoresis. Am. J. Clin. Path. 27:125-158. Thurston, J. R., L. F. Velicer, and R. S. Merkal. 1965. Glycoprotein levels in cattle normally and eXperimentally infected with Mycobacterium bovis. Am. J. Vet. Res. 26:1328-1330. Tiselius, A. 1939. Electrophoresis of serum globulin. Biochem. J. 31:313-317. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 169 Tiselius, A., and E. A. Kabat. 1939. An electropho- retic study of immune sera and purified antibody preparations. J. Exptl. Med. 69:119-131. Tiselius, A., J. Porath, and P. Albertsson. 1963. Separation and fractionation of macromolecules and particles. Science 141:13-20. Turcotte, R., S. O. Freedman, and A. H. Sehon. 1963. A new hemagglutination procedure for the diagnosis of active tuberculosis. Am. Rev. Respir. Dis. 88: 725—728. Turner, G. C., F. Schifferin, E. E. Erhbaugh, and J. DeLaHuerga. 1953. .Serum mucoproteins in pulmonary tuberculosis. Dis. of Chest 23:154-159. Uriel, J., and P. Grabar. 1956. Emploi de colorants dans L'analyse electrophoretetique et immuno electrophoretique en millieu gelifie. Ann. Inst. Pasteur 90:472-440. Vahvaselka, E. 1962. A method for quantitative starch gel electrophoresis. Nature 193:474. Valmet, E., and H. Svensson. 1954. Some problems inherent in paper electrophoresis. Science Tools. 1:3—70 Volk, B. W., A. Saifer, L. E. Johnson, and I. Oreskes. 1953. Electrophoretic and chemical serum protein fractions in pulmonary tuberculosis. Am. Rev. Tuberc. 67:299-321. Weber, T. B. 1964. Electrophoretic analysis of bovine serum by vertical and horizontal types of apparatus. Am. J. Vet. Res. 25:386-390. Weber, W. 1955. Das electrophorogramm des blutserums tuberkuloser rinder. Schweizer Arch. Tierheilkunder 97:222-229. Weimer, H. E., R. T. Bell, S. Froman, H. Nishihara, and E. H. Rice. 1960. Electrophoretic studies of serum proteins and glycoproteins during early stages of experimental tuberculosis. Proc. Soc. Exptl. Bio. Med. 105:300-303. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 170 Weimer, H. E., and J. R. Moshin. 1953. Serum glyco— protein concentration in eXperimental tuberculosis of guinea pigs. Am. Rev. Tuberc. 68:594-602. Weimer, H. E., and J. R. Moshin. 1953. Comparative effects of intramuscular injections of ACTH, cortisone, and saline on serum glycoprotein levels. Proc. Soc. Exptl. Biol. Med. 84:34-37. Weimer, H. E., J. R. Moshin, R. A. Boak, E. Bogen, and C. M. Carpenter. 1954. Serum protein studies in experimental tuberculosis of the guinea pig. Am. Rev. Tuberc. 70:344-348. Weimer, H. E., J. R. Moshin, and E. L. Nelson. 1955. Comparative effects of infection and immunization on serum glycoprotein concentrations. J. Immunol. 74:243-248. Weiner, L. M., and B. Zak. 1963. Electrophoresis and immunoelectrOphoresis in an_agar gel—paper medium. Clin. Chim. Acta 8:440-445. White, R. G., A. H. Coons, and J. M. Connally. 1955. Studies on antibody formation. III. The alum granuloma. J. EXptl. Med. 102:73-82. Wieme, R. J. 1957. Uber einige Aspekte der electro- phorese in agar gel. Behringwerk Mitteilungen. 34:27-37. Wieme, R. J. 1959. Studies pp Agar Gel Electropho— resis--Techniqpes, Applications. Arscia Uitgaven N. V., Brussel. Williams, C. A. 1960. Immunoelectrophoresis. Scientific American 202:130-140. Williams, C. A., and C. T. Wemysis,Jr. 1961. Changes produced in mouse plasma proteins by acute bacterial infections. J. EXptl. Med. 114:311-325. Wolstenholme, G. E. W., and E. C. P. Miller. 1956. Paper electrophoresis. Ciba Foundation Symposium. Little Brown and Company, Boston. Wroblewski, F. 1961. Multiple molecular forms of enzymes. Ann. New York Acad. Sc. 94:655-1039. 213. 214. 215. 216. 217. 171 Yeoman, W. B. 1960. The examination of scanning instruments used in electrophoresis. Clin. Chim. Acta 5:279—282. Zingale, S. B., C. A. Mattioli, H. D. Bohner, and M. P. Bueno. 1963. Disc electrophoresis study of serum proteins from patients with multiple mycloma and macroglobulinemia. Blood 22:152-164. Zitrin, C. M., E. M. Lincoln, A. Saifer, and S. Leskowitz. 1956. Serum gamma globulin in child— hood tuberculosis. Am. Rev. Tuberc. 74:15-28. Zitrin, C. M., E. M. Lincoln, R. Carretero, and E. Melly. 1959. Serum proteins in childhood tuber- culosis. J. Dis. Child. 98:330-341. Zitrin, C. M. 1960. The C—reactive protein in childhood tuberculosis. Am. Rev. Resp. Dis. 81:266-270.