DISC ELECTROFHORETIC AND EMMUNQDEFFUSION STUDIES OF MYCOBACTERIAL CULTURE FELTRATES The“: for 1510 Degree 09 pk. D. MICHLGAN STATE UNIVERSITY Thomas L. Roszman 1966 .xv_ " ,.e w..- . . 'M-y); '. .3413 ‘ ”LA . t \- a,“ \ X r V‘ ' LiBiUi“ WV . - ate THESIb . [\"IhChlgan St UniVCrsuy S s. ;-_ -'* This is to certify that the thesis entitled Disc Electrophoretic and lmmunodiffusion Studies of Mycobacterial Culture Filtrates presented by Thomas L. Roszman has been accepted towards fulfillment of the requirements for Ph.D. degree in Microbiology and Public Health / //L¢¢~ /¢/’( h’KQifl / Major professén' Date‘ZZ‘f' [(6 /7%/‘ O~169 ABSTRACT DISC ELECTROPHORETIC AND IMMUNODIFFUSION STUDIES OF MYCOBACTERIAL CULTURE FILTRATES by Thomas L. Roszman Unheated, pervaporated culture filtrates were pre- pared from four mycobacteria, Mycobacterium bovis (strain 310, virulent, bovine origin), g. avium (strain 132, viru- lent, avian origin) and two strains of atypical mycobacteria. The atypical strains, strain 50 and strain 68, were of bovine origin and classified as Group III (slow growing, nonphotochromogens). Disc electrophoresis was the best technique to sep— arate the components in the mycobacterial culture filtrates. It was facile, economical and versatile and the results were highly reproducible. The culture filtrates contained be- tween 18 and 24 protein components and from five to eight polysaccharide or glycoprotein components. Antigenic anal- yses of the culture filtrates with homologous and heterolo- gous antisera (rabbit origin) were made with Ouchterlony immunodiffusion, immunoelectrophoretic and disc immunoelec- trophoretic tests. With the Ouchterlony immunodiffusion technique, six to ten antigens were detected in the culture Thomas L. Roszman filtrates. An antigen was present in each of the culture filtrates not detected in the other filtrates. The two Group III atypical mycobacteria were more closely related to .g. avium than to g. bovis but were not identical to g. avium. No common antigen was detected. The number of antigens detectable in the four cul- ture filtrates was increased by immunoelectrophoresis and disc immunoelectrophoresis. Disc immunoelectrophoresis was better than immunoelectrophoresis. The greatest number of antigens occurred in the g. avium culture filtrates. The components in culture filtrates were separated by preparative disc electrophoresis. Separation of the com- ponents in the culture filtrates by the preparative disc electrophoresis was as good or better than by analytical disc electrophoresis. The gels were sectioned and the com- ponents eluted. The elutes were dialyzed to remove tetra— methylethylenediamine, a constituent used in preparing the gels, which reacts in the Lowry method for the determination of protein. The per cent of protein recovered from the eluates ranged from 50 to 60%. Re-electrophoresis of cer- tain eluates in 10% analytical gels indicated that one amido black-staining component in 7% gel could be separated into as many as six components. Examination of 20 different eluates by re-electro- horesis and with four reference antisera specific for the Thomas L. Roszman culture filtrates indicated the reliability of the section- ing technique. One or more antigens were obtained from each filtrate which were not detected in the other filtrates. Examination of the antigens in the 20 eluates with 11 anti— sera specific for culture filtrate and beta-propriolactone- inactivated cells of g. avium, g, bovis, BCG and Group I, II and III atypical mycobacteria, indicated that group and species antigens were obtained by disc electrophoresis. DISC ELECTROPHORETIC AND IMMUNODIFFUSION STUDIES OF MYCOBACTERIAL CULTURE FILTRATES BY Thomas L. Roszman 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 ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation and thanks to Dr. V. H. Mallmann for her guidance and support throughout this study. Appreciation is also extended to Dr. W. L. Mallmann for his assistance and counsel. A special appreciation goes to the author's wife, Sharon, the silent partner in the pursuit for intellectual enlightenment. Preparation of this thesis has been greatly facil- itated by financial assistance given to the author by the Michigan Tuberculosis and Respiratory Disease Association. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . Cultures . . . . . . . . . . . . . . Culture filtrates . . . . . . . . . Production of antisera . . . . . . . Ouchterlony immunodiffusion . . . . Immunoelectrophoresis . . . . . . . Analytical disc electrophoresis . . Preparative disc electrophoresis . . Disc immunoelectrophoresis . . . . . RESULTS . . . . . . . . . . . . . . . . . Analytical disc electrophoresis . . Immunodiffusion analyses of the four filtrates . . . . . . . . . . . . Immunoelectrophoresis . . . . . . . Disc immunoelectrophoresis . . . . . Preparative disc electrophoresis . . culture Immunodiffusion analyses of the eluates . . DISCUSSION . . . . . . . . . . . . . . . . SUMMARY s. . . . . . . . . . . . . . . . . LITERATURE CITED . . . . . . . . . . . . . iii Page 22 22 22 23 26 27 29 29 31 33 33 38 46 50 51 56 63 78 81 LIST OF TABLES Table Page 1. Eleven mycobacteria from which cells and culture filtrates were prepared for inocula to obtain antisera . . . . . . . . . . 25 2. Amounts of protein and carbohydrate in the unheated, ten fold concentrated mycobacterial culture filtrates . . . . . . . . 34 3. Number of immunoprecipitates in Ouchterlony gel diffusion tests of mycobacterial cul- ture filtrates with the reference antisera . . 39 4. Number of immunoprecipitates in immuno- electrophoresis (IE) and disc immunoelec- trophoresis (DIE) tests of the four myco- bacterial culture filtrates with the reference antisera . . . . . . . . . . . . . . 48 5. Number of protein-staining components detected by re-electrophoresis and pro- tein concentrations of the 20 eluates obtained from preparative disc electrophoresis . . . . . . . . . . . . . . . . 55 6. Number of immunoprecipitates in Ouchterlony gel diffusion tests of 20 eluates obtained by preparative disc electrophoresis of mycobacterial culture filtrates with four reference antisera . . . . . . . . . . . . . . 58 7. Number of immunOprecipitates in Ouchterlony gel diffusion tests of 20 eluates obtained by preparative disc electrophoresis of mycobacterial culture filtrates with 11 cell-culture filtrate antisera . . . . . . . . 61 iv LIST OF FIGURES Figure Page 1. Analytical disc electrophoretic patterns of culture filtrates 68, 50, Mycobacterium bovis (bov), and M. avium (av) stained with amido black . . . . . . . . . . . . . . . 35 2. Schematic representation of the protein (A) and PAS positive (B) components in culture filtrates 68 and 50. The letters S and L refer to the Spacer and lower gel areas, respectively . . . . . . . . . . . . . . . . . 36 3. Schematic representation of the protein (A) and PAS positive (B) components in culture filtrates Mycobacterium bovis and M. avium. The letters S and L refer to the spacer and lower gel areas, respectively . . . . . . . . 37 4. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 50 (center well) . . . . . . . . . . . . . . . 40 5. Ouchterlony immunodiffusion of culture filtrates 50, Mycobacterium avium (av) and M, bovis (bov) with reference antiserum 50 (center well) . . . . . . . . . . . . . . . . 41 6. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 68 (center well) . . . . . . . . . . . . . . . 43 7. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 68 (center well) . . . . . . . . . . . . . . . 44 8. Ouchterlony immunodiffusion of culture fil- trates 50, 68 and Mycobacterium avium (av) with the M. avium reference antiserum (center well) . . . . . . . . . . . . . . . . 45 Figure Page 9. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with M. bovis reference antiserum (center well) . . . . . . 47 10. Immunoelectrophoresis of Mycobacterium avium culture filtrate with homologous reference antiserum . . . . . . . . . . . . . 49 11. Disc immunoelectrophoresis of disc gels of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with M. avium reference antiserum (long, narrow trough). The letters S and L refer to the spacer and lower gel areas, respectively . . . . . . . . 52 12. Analytical disc electrophoretic patterns of culture filtrate 50 (extreme left) and five components re-electrophoresed after section- ing from preparative gels stained with amido black . . . . . . . . . . . . . . . . . 54 13. Analytical disc electrophoretic patterns of eluate 50-18 obtained from culture filtrate 50 preparative gels on re-electrophoresis in 7 and 10% gels. Stained with amido black . . . . . . . . . . . . . . . . . . . . 57 vi INTRODUCTION The mycobacteria are the oldest and most extensively studied bacteria, yet the basis of the pathogenicity of the tubercle bacilli is not understood. The tuberculin test aids in the detection of sensitivity induced during past or pres- ent infections. There is no serologic test which is conclu- sively, diagnostic for the disease. The final proof of active disease still depends on the isolation and identifica- tion of the causative agent. Numerous factors have contributed to the impressive decrease in the mortality rate of man due to tuberculosis, but the decrease in the morbidity rate has been considerably less impressive. As a result, the problem of the atypical mycobacteria as possible infectious and sensitizing agents has been recognized and studied during the past fifteen years (89, 116). An analogous situation exists in the field of bovine tuberculosis (39, 81). The number of cattle, which are tuberculin positive and have gross-lesions at slaughter, has been reduced to a very low number. The number of tuberculin positive cattle which have no-gross-lesions at the time of slaughter has remained relatively constant. Atypical mycobacteria similar if not identical to those of human origin, have been isolated from both categories of cattle, lesions from swine, and inanimate sources (63). Attempts have been made to isolate and characterize mycobacterial fractions which would be of value as specific sensitins or sero-diagnostic agents. Usually tuberculopro- teins and tuberculopolysaccharides are obtained by chemical precipitation from culture filtrates. The antigenic com- plexities are readily detected by immunodiffusion tests. Newer physical techniques such as column chromatography and zone electrophoresis have been used. While the data ob- tained added considerably to the understanding of the chem- ical and physical properties of the mycobacteria, no spe- cific antigen or sensitin has been isolated. This is a report of the characterization and isola- tion of mycobacterial antigens from four concentrated, un- heated culture filtrates. Ouchterlony immunodiffusion, immunoelectrophoresis and disc immunoelectrophoresis were used to study the antigenic composition and relatedness of the four culture filtrates. Analytical and preparative disc electrophoretic separation of the culture filtrate antigens afford a means of obtaining antigens from preparative disc gels for further study which are relatively pure and have promise of differential Specificity. LITERATURE REVIEW At the present time, there is no conclusive test by which disease or sensitivity due to the classical pathogen can be differentiated from that due to atypical mycobacteria (89). There is no way by which pathogenic, atypical myco— bacteria can be differentiated from saprophytic, atypical mycobacteria. A need exists for antigens with which the causative agent of disease or sensitivity can be identified while still in the host, or in pure culture. The terms atypical, unclassified or anonymous, are used currently to indicate those mycobacteria other than the well defined species. Runyon (88) placed the atypical myco- bacteria of human origin into four groups according to the number of days required for isolated colonies to grow on Lowenstein-Jensen medium at 37C and production of pigment. Group I (photochromogen): slow growth, no pigment produced in the dark, dramatic production of yellow pigment in the dark after a short exPosure to white light; one species, Mycobacterium kansasii. Group II (scotochromogen): slow growth, yellow pigment produced in light or dark; many strains. Group III (nonphotochromogen): slow growth, little or no pigment produced in light or dark; many strains. Group IV (rapid growers): rapid growth; isolated colonies visible in five to six days; many strains. Group I and some strains of Group III mycobacteria can produce pulmonary disease in man; some Group II and IV mycobacteria may cause disease in man (14, 88, 89, 110, 116). Some Group III mycobacteria of bovine origin can cause dis- ease in cattle (62). It is generally reported that little or no transmis- sion of the atypical mycobacteria occurs among humans. It may be that transmission occurs frequently but that man has little susceptibility. Transmission has occurred from eXper- mentally infected cows to their calves, and to swine housed with experimentally infected swine (J. A. Ray, Ph.D. Thesis, Michigan State Univ., East Lansing, 1966). The chemistry of the mycobacteria and the biological preparations derived from them has been studied extensively since the use of concentrated culture filtrate, tuberculin, by Koch in 1891 (22). The pioneering studies by Seibert are classical (54, 92, 93, 94, 96). Chemical fractionation of culture filtrate indicated that the active component of tuberculin was protein (54, 91). Seibert (92, 93, 94, 97) develOped the methods for preparation and standardization of the Purified Protein Derivative (PPD-S), an international standard. The PPD was fractionated into three protein com- ponents, A, B, and C and two polysaccharides, I and II (95). Although further attempts were made to purify PPD (100), one of the paradoxes of protein chemistry is the change of defi- nitions of purity. Purity is a dynamic state. Seibert recognized the dilemma and utilized new techniques as they were developed such as Tiseluis electrophoresis (95, 98), ultracentrifugation (98), gel immunodiffusion (99) and paper chromotography (100). As the prevalence of tuberculosis declined, more dependence was placed on the tuberculin test as a means of detecting tuberculin sensitive individuals without apparent disease. Initially, the assumption was made that a positive test indicated tuberculosis due to M. tuberculosis, infre- quently M. bovis, and less frequently M. avium. Veterinar- ians are credited with recognizing the high rate of false positive reactions, and the no-gross-lesion tuberculin posi- tive animals, ten years before the physicians recognized the problem (22). The sensitivity was attributed to increased prevalence of M. avium and atypical mycobacteria (22, 26). Numerous studies with PPD preparations including PPD's prepared from atypical mycobacteria proved that cross sensitivity occurred (1, 22, 24, 26, 40, 58, 59, 60). How- ever, the homologous sensitins usually elicited a larger response than heterologous sensitins when injected into sensitized animals. This was presumptive evidence that Specific antigen fractions isolated from PPD's or crude cul- ture filtrate would be of value. In theory, the specificity of the antigen fractions should be inversely proportional to the cross reactivity. Unquestionably, recent advances in physical and biological techniques for the separation, isolation and characterization of proteins have been a stimulus for those interested in tuberculoproteins. Unfortunately, many re- searchers interested in the chemical and biologic properties of mycobacteria are traditionalists. The techniques and methods set forth by Seibert forty years ago which utilized chemical precipitation of crude culture filtrate are cur- rently in use. These methods have served a purpose but no protein has been isolated which approaches the degree of purity or specificity sought. Researchers have been late in applying such techniques as immunodiffusion, immunoelectro- phoresis, starch gel electrophoresis, disc electrophoresis, gel filtration and DEAE chromatography to mycobacteria. Immunodiffusion was established as an analytical technique by Oudin (71). Until this time, multiple precip- itate bands were regarded as rhythmic precipitates formed by one antigen-antibody system rather than reactions of differ- ent antigens with their specific antibodies. The immunodiffusion technique which Oudin initially perfected (71) was the Simple diffusion tube test. He estab— lished the basic tenets of immunodiffusion which were ampli- fied by Elek (25) and Ouchterlony (69) by their indepen- dently developed double diffusion techniques. There is one basic difference between the simple and double diffusion techniques. In the simple diffusion test, one reactant (usually the antiserum) is mixed with the gel and after solidification, the other reactant is added to diffuse into the gel mixture. In the double diffusion test, both the antigen and antibody diffuse simultaneously into a neutral agar area where precipitation occurs due to the formation of antigen-antibody complexes. While the theories developed initially by both Elek and Ouchterlony to eXplain their double diffusion systems were compatible, Ouchterlony's method is the technique of choice today. Elek's technique utilized paper strips impregnated with the reactants which were placed at right angles to each other on agar plates. In the Ouchterlony technique, wells were cut in the agar, and the well arrangements were varied so that many antigenic preparations and antisera could be compared. This allowed recognition of antigenic relationships that may exist between different preparations. Reviews by both Ouchterlony (70) and Crowle (15, 16) describe immunodiffusion applications and techniques, and discuss the interpretation of the results. Crowle's monogram (16) is considered to be the handbook of immunodiffusion. He has considered both the applied and theoretical aspects of immunodiffusion. Immunoelectrophoresis is a combination of electro- phoresis and immunodouble diffusion in a gel. The technique, as devised by Grabar and Williams (33) was performed on large glass slides to separate and analyze human serum anti- gens. Scheidegger (90) adapted the technique to microscope slides which Saves time and material. Agar is the gel most widely used for immunoelectro- phoresis. It has high gel strength, transparency, solubil- ity in an aqueous medium and little or no reaction with pro- teins (16). Other supportive materials which have been employed are starch gel (77, 78), cellulose acetate (44), and more recently, polyacrylamide gels (42). Less electro- endosmosis occurs in the polyacrylamide gels than in agar. The acrylamide gels can be cast on glass Slides for use as in agar immunoelectrophoresis (42) or as a sequential combination of disc electrophoresis and agar double diffu- sion (38, 101). After disc electrophoresis (68), the gels are placed on glass slides and covered with a melted agar solution. After solidifying, trenches are cut in the agar parallel to the full length of the gel columns. Antisera are placed in the trenches and the plates incubated at a predetermined temperature until precipitin lines form. Antigens can be cut from fresh gels for use in further studies. Early zone electrophoretic techniques used only the difference in charges on molecules for separation. In 1955 Smithies (102) reported on starch gel electrophoresis which utilized not only the charge of the molecule but also the molecular weight and size. This added dimension of molecular- Sieving increased immensely the number of separations hereto- fore obtained with other techniques. New techniques and media for zone electrophoresis continue to be reported. Bier (5) and Morris and Morris (66) have reviewed the applied and theoretical facets. Polyacrylamide electrophoresis was reported indepen- dently by Raymond (83) and Ornstein and Davis (68). The investigators differed on the theoretical principles in- volved (82) but the underlying features are well accepted. The gel matrix is prepared by polymerization of two acryla- mide compounds which form a cross-linked, transparent gel of defined pore Sizes which can be controlled by the relative concentration of the two compounds. Materials electropho- resed in the gel are separated on the basis of molecular weight, structure and charge. Ornstein and Davis have termed their technique disc electrophoresis. The name is derived from the discoid Shape of the bands obtained and also from the dependence on discon- tinuities in the electrophoretic matrix (67). A controlled gel pore size and an electrophoretic step for concentrating the protein components into narrow bands prior to electro- phoretic separation increases the resolution of the macro- molecules (17). The average pore size of the 7-l/2% lower gel is approximately 50 Angstrom units (67). 10 Disc electrophoresis offers many advantages over most of the other types of zone electrophoresis. The gel is transparent, strong and flexible, qualities necessary for handling and making visual and spectrophotometric observa- tions. Of even greater importance, is the molecular sieving properties that the gels possess by virtue of their pore size. Changing the pore size of the separation gel can easily be accomplished by either decreasing or increasing the relative per cent of the two acrylamide monomers. This contributes to the versatility and value of disc electro- phoresis. One very important attribute is the small quan- tity of biological materials required. AS little as 250 pg of serum protein can be separated by this technique (17). Conversely, disc electrophoresis gels can be used for larger quantities of material. Preparative disc electrophoresis has been employed with as much as 40 mg of protein (41). The separated components were recovered by elution from the end of a polyacrylamide column into a fraction collector (21, 37, 41, 45, 61, 79, 106), or by cutting segments from the column of gel and eluting from the segments the proteins (45, 117). Disc electrophoresis as originally described was designed for the separation of serum proteins. The buffer systems used were in an alkaline pH range above the isoelec- tric point of serum proteins. Williams and Reisfeld (108) 11 have reported on new cationic (pH 4.3) and anionic (pH 7.5) buffer systems. They have also described how the complex calculations of Ornstein'S can be utilized to develop new buffer systems. Mycobacteria have certain characteristics which dissuade many potential investigators. The pathogenic myco- bacteria grow very slowly. To obtain sufficient amounts of protein, the organisms are usually grown for two to three months. Autolytic processes are relied upon to liberate the cellular proteins into the medium, during which time denatur- ization of some of the protein occurs. Frequently, the cul— ture filtrates are heated which causes further denaturation. Heating the cellular suspensions before filtration originated as a precautionary measure and is continued as such. Heating destroys certain antigenic components (51). The high lipid content of mycobacteria causes auto agglutinations which renders them unsatisfactory for classical bacterial aggluti— nation, a technique used frequently for antigenic analyses of other bacteria (6). Among the first to employ zone electrophoresis for the separation of tuberculoproteins and polysaccharides were Rhodes, Sorkin and Boyden (86, 87, 103). Such media as glass, cellulose, and filter paper were used. By precipita— tion from a culture filtrate, a hemosensitin preparation was obtained which could be further purified by electrophoresis through glass beads (87). 12 Paper electrophoresis has been used to separate the components of PPD prepared from M. bovis (19). Four protein components were detectable. Starch—agar gel electrophoresis was less effective with the same PPD (20). Yoneda and his associates have combined chemical and physical methods of separating tuberculoproteins from H37Rv, a virulent strain of M. tuberculosis (28, 111, 112, 113, 114). A culture filtrate of H37Rv was treated with ammonium sulfate and the three precipitates which occurred at 0-30%, 30-50%, and 50-80% were collected. These precipitates were sub- jected to starch block electrophoresis and the proteins eluted. Two components each were found in the 0-30% and 30-50%.precipitates. The 50-80% precipitate contained many components. On further analyses of the four components, A, B, C, D from the 0-30% and 30-50% precipitates by the Oakley- Fulthrope immunodiffusion technique, one line of precipita- tion occurred with A and C, two with D and no lines with B. The serologic reactivity of all the antigens was sensitive to 70C for 10 minutes. The physical and chemical properties of antigens A and C were Similar, and identical by Ouchterlony analyses (112). The D antigen was not identical with the A or C antigen. Pure preparations of the three antigens were obtained by diethylaminoethyl (DEAE) cellulose chromatog- raphy. ,Chemical analyses indicated that A and C were pure protein (113). No carbohydrate or nucleic acid was detected. 13 Subsequently, the nomenclature of the three antigens was changed. Antigens A or C were referred to as alpha, and D as beta. The alpha and beta antigens comprised about 70% of the total protein released by the H37Rv strain into the medium (28). They indicated that these antigens might be on the surface of the cell. Antisera against the alpha and beta antigens were used to study the distribution of these antigens in 120 strains of mycobacteria using the Ouchterlony technique (114). The 120 mycobacteria could be separated into four groups on the basis of the presence or absence of both alpha and beta, or the presence of one but not the other. All strains in which both alpha and beta were present consti— tuted Group I. With the sole exception of the isoniazid- resistant strains, strains of M. tuberculosis, M. bovis and one strain of M. microti were in this group. Strains in which no extracellular beta but alpha was detectable consti- tuted Group II. All isoniazid-resistant strains of M. tuberculosis and M. bovis were in this group. All the M. avium strains, M. paratuberculosis, M. balnei, M. ulcerans, ‘M. lepraemurium and 24 out of 45 atypical strains were in Group III. This group was characterized by no beta detect- able but cross-reacting material with antigenic determinants partially in common with alpha. Group IV contained all saprophytic strains including M. fortuituim, M. phlei, 14 .M. smegmatis and 21 atypical strains. Yoneda's grouping of the atypical mycobacteria did not agree with Runyon's Groups. Gel filtration has been used to separate the compo- nents of tuberculin (13). An unheated BCG culture filtrate was separated into a dialyzable and nondialyzable fraction (4). The nondialyzable fraction was precipitated with 1.0% acetic acid at pH 4. The supernatant fluid was designated fraction G and the precipitate fraction F. The G and F fractions were separated into three and two components, respectively, following separation on Sephadex G—50 columns. This resulted in a sharper separation of the protein and polysaccharide constituents. These five fractions elicited reactions when injected intradermally into sensitized guinea pigs. The complexity of these five fractions was readily detected by gel diffusion analyses. Each were composed of from five to seven antigens. In the search for an immediate skin test antigen, Glenchur, Fossieck and Silverman (31, 32) fractionated bacillary extracts of H37Ra, an avirulent strain of M. tuber- culosis, on columns of G-25 Sephadex and DEAE cellulose. Three fractions were obtained from the G-25 Sephadex chroma- tography. One of these fractions, I, was further Separated into seven fractions by DEAE column chromatography. The results of skin testings with one of the DEAE fractions in pulmonary tuberculous and control patients suggested that 15 the procedure may have some merit from the standpoint of obtaining immediate type skin testing agents. A continuation of this study indicated the PPD's could be fractionated in the same manner (32). Three frac- tions were obtained by G-25 Sephadex chromatography of PPD. The quantities of two of these fractions was considerably less than that found in H37Ra extracts. The seven H37Ra fractions obtained from DEAE separation were not pure but contained as many as five antigens in some of the fractions. Kniker and LaBorde (43) used DEAE cellulose chroma- tography to separete the antigens of four strains of M. tuberculosis. Twelve fractions collected and pooled from each culture filtrate were concentrated, and analyzed by Ouchterlony immunodiffusion with their homologous antisera and with 14 heterologous antisera. Antisera were prepared for 12 atypical mycobacteria including Group I, II and III organisms. As many as twenty antigens were detected in the homologous immunodiffusion systems. Common antigens were found in various fractions as well as antigens specific only for the tubercle bacilli. No one antigen was isolated in pure form as evidenced by the immunodiffusion studies. Complete separation of mycobacterial antigens by column chromatography methods has not been reported. Lind (51) used carboxymethyl cellulose to separate the antigens in an unheated culture filtrate. He suggested that good 16 separation of the antigens was not obtained because the proper variations in pH and molarities may not have been used. A fraction obtained by Seibert's method (95) was used by Rhodes (85) for DEAE column chromatography. The fraction could be separated into three protein components by paper electrophoresis and into at least four components by DEAE chromatography. Re-electrophoresis of the fractions collected from the column using paper and starch indicated little separation of the antigens. All the fractions elic- ited about the same level of tuberculin type Skin reactivity when tested in sensitized animals. Immunodiffusion techniques have yielded excellent results when applied to studies on the antigenic relation- ships of mycobacteria (73, 74). Quantitative as well as qualitative data can be obtained. The Ouchterlony technique is the most favored because comparisons can be made between different antigenic preparations. The first extensive study of the antigenic relation- ships between mycobacteria employing the modified Ouchterlony technique was by Parlett and Ybumans (73). Concentrated, unheated culture filtrates were injected into rabbits to obtain the antisera, and used as antigen in immunodiffusion tests. The culture filtrates of forty-two mycobacteria, including strains of M. tuberculosis, M. bovis, M. avium 17 atypical and saprophytic mycobacteria, were analyzed for their antigenic relationships with 16 antisera. The great- est number of precipitate lines was four. On the basis of their findings they divided the 42 mycobacteria into four groups. Group I included one atypical mycobacterium, one attenuated bovine strain, one avian and all of the human strains. Group II consisted mainly of bovine strains and one avian strain. Group III consisted only of saprophytic mycobacteria. Group IV was one atypical strain which re- acted only with its homologous antiserum. The cross reactiv— ity reactions indicated the close relationship of the atypi- cal strains to the human strains. A more extensive study employed 98 mycobacteria and four fungi with 42 reference antisera (74). The methods and techniques were the same except the antigens in agar immuno- diffusion tests were cell extracts. The greatest number of precipitate bands found was Six and the number of groups expanded to eight. The major differences were that M. phlei was classed in a group separate from the other saprophytic mycobacteria, and the atypical strains were divided into five groups. The antigenic relationship reaffirmed their belief that the atypical strains were genetically related to the virulent human strains of mycobacteria. The atypical strains were more related antigenically to the avian strains. No common antigen was found for all of the 98 strains of 18 mycobacteria studies. No antigenic relationship was detected between the mycobacteria and fungi cultures. The results firmly established that immunodiffusion techniques were of value in classifying mycobacteria. Further immunodiffusion studies have been reported, primarily by investigators in other countries. A common antigenic component was detected by immunodiffusion in all of 23 strains which included 11 atypical strains (34, 104). The antisera were produced in rabbits inoculated with viable cell suspensions. The antigens in the Ouchterlony immunodif- fusion tests were unheated culture filtrates of organisms grown in Youman's medium (115) with 10% horse serum. No procedures were described to free the cell suspensions of the horse serum components before use for immunization. In addition to the common antigen, whether it was of mycobacte- rial origin or not, a close antigenic relationship existed between the mammalian tubercle bacilli and the atypical mycobacteria. More recently Tuboly (107) and Gempl (29), and Gempl and Weissfeiler (30) reported the presence of com- mon antigens in the mycobacteria studies by immunodiffusion. Other immunodiffusion studies have added materially to the information concerning mycobacterial antigenic rela— tionships. Mycobacterium phlei was found to be antigenically distinct from all other mycobacteria including the saprophy- tic strains (29, 30, 52). Studies of M. avium, M, ulcerans, 19 .M. balnei and M, marinum indicated that M. avium and M. ulcerans were closely related and that all four organisms possessed at least one common antigenic factor (53). Lind (46, 47, 48, 49, 50, 51, 52) has made the most extensive and complete studies employing immunodiffusion techniques. His work has added immeasurably to the knowl- edge of the antigenic relationships among different myco- bacteria. Equally important, he has investigated some of the variables and made a plea for the establishment of sound, uniform eXperimental techniques. He has studied extensively the methods of producing antisera, and the chemical and physical factors which affect the antigens of both culture filtrates and cell extracts. His plea is for standardiza— tion of both antisera and antigens among the various labora- tories studying mycobacteria by immunologic methods. The need for standardization of the procedures for the preparation of antigens and antisera is apparent. It is difficult to draw meaningful comparisons from reports of immunodiffusion studies of mycobacteria. The antigenic preparations differ and the antisera produced by different procedures. A difference of a few weeks in the age of the cultures and the growth medium used can change the antigenic content of the culture filtrate (12). The deleterious effect of heat and chemicals such as phenol on proteins is well known, yet phenol is used with mycobacterial antigenic preparations (107). 20 The Oakley-Fulthrope technique (99) and immunoelec- trophoresis (8, 9, 12, 29, 51, 107) have been used with myco- bacterial antigens. The Ouchterlony (29, 50) and tube dou- ble diffusion technique (3, 7, 84) are currently in use as sero-diagnostic tests; the latter is the more widely used (72, 75, 76). Reports vary on the reliability of the test. The number of positive human sera from tuberculous patients has ranged from 44%.(50) to 60 to 80% (75, 105). Negative results have been reported by Long and Top (56) with sera from tuberculin positive cattle. The antigenic preparations employed in the tube double diffusion sero—diagnostic studies were from different mycobacteria: H37Ra (18, 56), M. le2g and BCG,the attenu— ated Calmette—Guerin strain of M. ngis (56). As several investigators have suggested, the test will undoubtedly be more accurate when specific antigens are isolated and used in the test (18, 27). The inability to differentiate Group III'S which are saprophytes and cause no disease, and those which can cause disease in man, cattle, swine and probably other animals indicates a need for antigens specific for each of the classical pathogens and representatives of Group III mycobacteria. Edwards and associates (23) have delineated what an ideal skin test Should be. "It would never indicate that an individual had not been infected if he had, or that he had been infected if he had not. Moreover, the ideal skin test 21 would never confuse infection by one organism with infection by another." They also state that the Skin test should be able to differentiate between Single and multiple infections. No skin test is available today which can meet these stan- dards. This reaffirms the need for specific sensitins. MATERIALS AND METHODS Cultures. The four mycobacteria were isolated from animals: ‘M. bovis, strain 310 from a gross-lesion cow; M. avium, strain 132 from a gross-lesion chicken; Group III, strain 68 from a tuberculin positive gross-lesion cow and Group III, strain 50 from a tuberculin positive no-gross- lesion cow. Strain 50 was categorized as of low or no viru— lence; strain 68 was of moderate virulence (64). Culture filtrates. The organisms were grown on the surface of one liter of a modified Proskauer-Beck synthetic medium (115) in large diphtheria toxin bottles. After three months at 37C, the culture fluid was drawn off aseptically in 250 m1 plastic bottles and centrifuged three hours at 2010 x g. The supernatant fluid was passed through a Berkefeld filter (N grade), the filtrates dialyzed against 0.015M phOSphate buffer, pH 7.0, for three days at 4C and concentrated ten fold by pervaporation. The concentrate was filtered (Millipore,0.45 p pore size) merthiolate added (1:10,000) and stored at —80C in small amounts. The amount of protein in the culture filtrates was determined by the method of Lowry (57). The amount of carbo- hydrate in the culture filtrates was assayed by the method of Morris (65). 22 23 For immunodiffusion and immunoelectrophoretic studies, the four culture filtrates were concentrated fur- ther as necessary with Ficoll (Pharmacia). Dialysis tubing (Visking Corp.) was filled with culture filtrate and covered with dry Ficoll. Production 2f_antisera. Twenty—nine Dutch Belted rabbits were used as the source of the four reference anti- sera, five rabbits each for culture filtrates 68, 50, and M, gyigg, and nine rabbits for M. bovis. A modified alum precipitation procedure as described by Carpenter (11) was used to prepare the inocula. To each 10 m1 of culture fil- trate, 0.46 ml of a 10% AlCl3 was added. With constant stirring, the mixture was adjusted to pH 7 with 20%.NaOH. Each rabbit was inoculated intramuscularly with 5 m1 of the mixture containing 10 mg of protein. After 40 days, the rabbits were inoculated subcutaneously at three Sites with a total of 1.5 m1 of culture filtrate containing a total of 2 mg of protein. The rabbits were bled by cardiac puncture seven and nine days later. All antisera were tested with the homologous culture filtrate by the Ouchterlony immuno- diffusion method and satisfactory antisera were pooled, merthiolate added (1:10,000), and stored in small amounts at -80C. These antisera are referred to hereafter as reference antisera. 24 Antisera were prepared against 11 mycobacterial cul- ture filtrates (Table 1) containing homologous cells. The culture filtrates were prepared similarly to those described except that the organisms were grown two months in 50 m1 of the medium in 250 ml flasks and the supernatant fluid fil— tered only with Millipore filters (0.45 p pore size).. Cells from each culture was inactivated with beta-propriolactone (BPIJ. One mg of packed BPL inactivated cells was mixed with three m1 of homologous culture filtrate containing 0.5 mg/ml protein. The suspension was mixed with an equal amount of Freund's incomplete adjuvant (Difco). ,Two Dutch Belted rabbits were inoculated for each of the 11 organisms (Table 1). Each rabbit received intravenously 0.25 ml of the preparation without Freund's incomplete adjuvant and at the same time a total of 0.5 ml at two sites subcutaneously of the preparation mixed with the adjuvant. Seven and 14 days later they were inoculated subcutaneously with 0.25 ml of the preparation with adjuvant. Two months following the initial injections, the rabbits received 0.5 m1 subcutane- ously of only culture filtrate on two successive days. Thirty-six days later they were inoculated with 0.75 ml of only culture filtrate subcutaneously and bled Six days later by cardiac puncture. The sera were pooled and stored at -80C until used. These antisera are referred to hereafter as cell-culture filtrate antisera. 25 Table 1. Eleven mycobacteria from which cells and culture filtrates were prepared for inocula to obtain antisera Designation Classification P4 Photochromogen (M. kansasii) Group I (human origin) P15 Scotochromogen, Group II (human origin) P39 Nonphotochromogen (Battey Type) Group III (human origin) 50 Nonphotochromogen, Group III (bovine origin) 51 Nonphotochromogen, Group III (bovine origin) 62 Nonphotochromogen, Group III (bovine origin) 68 Nonphotochromogen, Group III (bovine origin) 172 Nonphotochromogen, Group III (swine origin) 131 ‘M. avium virulent (avian origin) 310 .M. bovis virulent (bovine origin) BCG ;M. bovis attenuated (bovine origin) 26 Ouchterlony immunodiffusion. Glass Slides 3-1/4 by 4 inch were washed with detergent, rinsed in distilled water and dried. The slides were covered with 10 ml of T% melted agar (Difco) in a 0.15 M phosphate-saline buffer,le7.2 con- taining merthiolate (1:10,000). After the agar had solidi- fied, the Slides were placed in a humidified chamber for at least three hr before wells were cut in the agar according to a drafted pattern placed beneath the slide. The diameter of the wells was 6 mm and the diffusion distance between wells 6 mm. After the reactants were added to the wells, the plates were incubated at 28C for two to five days in a humidified chamber and observed daily. Following incubation, the plates were washed in phosphate-saline buffer solution, pH 7.0, for 24 hr followed by a second rinse in distilled water for 24 hr to remove unreacted protein. The slides were air dried and then stained in triple stain (16) for 15 min and destained for 30 min in 2% acetic acid. The four culture filtrates were placed in peripheral wells arranged around a center antiserum well in varying concentrations to determine the optimal conditions for pre— cipitin line formation. Each culture filtrate was tested at the following concentrations: undiluted, two, five, ten and 20 fold. In addition, culture filtrates 50 and 68 were tested at 30, 50, and 100 fold and M. bgyi§_and M. gyigg at 40 fold. In tests thereafter, 30 fold concentrations of 50 and 68, 20 fold concentrations of M. bovis and ten fold 27 concentrations of M. avium were used. The amount of protein in the ten fold concentrated culture filtrate is given in Table 2. By varying the arrangement of the culture fil- trates, each culture filtrate was placed in a well adjacent to every other culture filtrate to detect bands of identity and nonidentity. Immunoelectrophoresis. Immunoelectrophoresis was performed according to the Hirschfeld procedure (35) with certain modifications. The barbital buffer,;fi18.6, con- sisted of the following: Buffer in Buffer in Constituents electrode vessel agar Diethylbarbituric acid 1.38 gm 1.66 gm Sodium barbital 8.76 gm 10.51 gm Distilled water to 1 liter 1 liter A 2% purified Bacto agar (Difco) solution was prepared as described by Hirschfeld (35). After the final washing, the agar cubes were stored at 4C in distilled water until needed. The agar buffer solution was prepared by mixing two parts of the barbital buffer with one part of distilled water, heating to 60C, and adding three parts of melted 2% purified agar. Microscope slides which had previously been cleaned, rinsed and dried were covered with 2.5 m1 of the mixture of melted agar and buffer and allowed to stand at least three hr in a humidified chamber before use. 28 In the center of the long axis the slide, an antigen well 2.5 mm in diameter was cut 25 mm from the cathode end of the Slide and the agar plug removed. At a distance of 5 mm on both sides of the antigen well, the agar was cut so that after electrophoresis, two 2 x 66 mm antiserum troughs were made by removing the agar. A Shandon migration chamber (Colab) was fitted with a 6 x 8-1/2 inch plexiglass plate to hold four agar Slides on each Side. Filter paper impregnated with the well buffer served to establish electrical contact between slides. The antigen was added to the antigen well. Strains 68 and 50 were used in 100 fold concentrations, M. ngis at 40 fold and fl-.§XEEE at ten fold. The chamber was placed at 4C. A current of 1.25 ma per Slide was applied for 110 min using a constant current DC power supply (Volkam). After the agar plugs were removed to form the antiserum troughs, the antisera were added. Incubation was at 28C in a humidified chamber for 24 to 48 hr. The slides were washed for 24 hr in a 0.15 M phos- phate-saline buffer,;fii7.0 followed by a second washing in distilled water for 24 hr. The slides were covered with a moistened bibulous paper and air dried. Triple stain (16) was used followed by destaining in 2% acetic acid. 29 Analytical disc electrophoresis. The apparatus and technique was the same as described by Ornstein and Davis (68). The preparation of the stock solutions used in prepar- ing the gels was followed as outlined by Davis (17). A spacer gel one cm in thickness containing 40% sucrose was layered over the lower gel, and no sample gel used. From 0.05 to 0.2 ml of culture filtrate containing approximately 280 pg of protein was then added on top of the spacer gel by displacement. Electrophoresis was carried out at room temperature at 5 ma per tube using a 50 ma DC power supply (Volkam) until the bromphenol blue marker dye had migrated exactly 3.6 cm into the lower gel (approximately 90 min). The gels were stained for protein or polysaccharide. A 0.5% solution of amido black in 5% acetic acid was used to stain the proteins in the gel for 30 min at room temperature. Excess stain was removed electrophoretically in 5% acetic acid. Polysaccharides and glycoproteins were stained by the periodic acid-Schiff (PAS) reaction method according to Canalco (10). Preparative disc electrophoresis. Glass columns, 1.1 x 10 cm, were filled to 7 cm with the 7% lower gel and a 1 cm spacer gel added following polymerization of the lower gel. After polymerization of the spacer gel, the columns were placed in the electrophoretic chamber and buffer was added to the buffer chambers. From 0.4 ml to 1.5 m1 of cul- ture filtrate containing about 2.0 mg of protein was layered 30 onto the Spacer gel. Electrophoresis was at 11 ma per tube until the bromphenol blue marker dye had migrated 6.5 cm into the lower gel (approximately 4-1/2 hr). An amido black stained reference gel of each culture filtrate was prepared. The staining and destaining procedure was the same as that for the analytical disc gels. Line drawings of each of the reference gels were drawn to scale to serve as templates for cutting unstained gels. Eight columns were used for each of four culture filtrates. Immediately following electrophoresis the col- umns were immersed in an ice bath for l to 5 min. Each gel was placed on a glass lantern slide with a template beneath it and the gel cut into segments with a razor blade follow- ing the predetermined pattern. For each component to be recovered, the eight similar segments from each gel were placed, four segments to a tube, in 19 x 100 mm plastic tubes. Each tube contained 2 ml of 0.05 M phosphate-saline buffer solution, pH 7.0. The segments were cut into fine pieces with a microspatula and the mixture allowed to stand at room temperature for 48 hr. The contents of each of the two tubes containing the eight Similar segments were poured into a Filterfuge tube (International Equipment Company). The Filterfuge tubes were assembled without filters and the gel separated from the eluate by centrifugation at 2010 x g for 20 min. The eluate from each segment pool was removed 31 and dialyzed for 48 hr against 0.05 M phosphate-saline buffer, pH 7.0. The protein concentration was determined with cor- rection for residual tetramethylethylenediamine (TEMED). Selected eluates from the segments cut from the gels of the four culture filtrates were examined by analytical disc electrophoresis. The sample volume was from 0.1 to 0.2 ml with enough 40% sucrose added to give the sample suffi- cient density. One eluate from each of the four culture filtrates was also examined by analytical disc electropho— resis using a 10% lower gel, prepared with 40 gm acrylamide, 0.4 gm N,N'-methylenebisacrylamide, and distilled water to 100 ml. Selected eluates obtained from the preparative disc gels were analyzed with the four reference antisera and 11 cell-culture filtrate antisera. In the case of the four reference antisera, the homologous eluates were placed in the peripheral wells in such a way that each eluate was com- pared against every other homologous eluate. The homologous reference antiserum was placed in the center well. This was repeated for the heterologous reference antisera. To test the eluates with the 11 cell-culture fil- trate antisera, the antisera were placed in the peripheral wells and the eluates in the center well. Disc immunoelectrophoresis. This is a combination of disc electrophoresis and immunodiffusion in agar. After elec- trophoresis, one of each of the four culture filtrates was 32 placed on 3-1/4 by 4 inch Slides. A buffered agar solution was prepared by mixing one part melted 2% agar prepared according to the method of Hirschfeld (35) with one part tris-glycine buffer, pH 8.3. The tris-glycine buffer was the same as that used for the reservoir buffer in the disc electrophoresis procedure but was diluted, one part distilled water to two parts stock buffer. Ten m1 of the hot agar solution at 60C containing merthiolate (1:10,000) was pipetted onto the clean 3-1/4 by 4 inch Slides with the four gels. The disc gels were immediately centered and spaced 12 mm apart. The plates were incubated at 28C for 24 hr. Four antisera trenches, 2 x 66 mm, were cut the full length of the gels and spaced 5 mm from the outer edges of the gels. These trenches were filled with the four reference antisera. The plates were incubated for five days and observed daily. The plates were washed, dried and stained in the same manner as the Ouchterlony plates. RESULTS Analytical disc electrophoresis. The amounts of protein and carbohydrate in the four culture filtrates are listed in Table 2. From 0.25 to 0.30 mg of protein in not more than 0.2 ml was the optimal concentration for disc electrophoresis. The same concentration of protein was used in the duplicate gels which were stained with PAS. Represen- tative patterns of the protein components of the culture filtrates are presented in Fig. 1. Schematic drawings of the protein and PAS positive components are presented in Fig. 2 and 3. The number of protein components in the culture fil- trates are as follows: 24 in M. bovis and 22 in M. avium (Fig. 3); 18 in both strains 68 and 50 (Fig. 2). The pro- tein bands were more distinct in the M, avium and M. bovis culture filtrates. All the culture filtrates had a protein band at the surface of the lower gel. The band nearer the anode was in approximately the same position for all fil- trates. The bands in culture filtrates 68 and 50 contained approximately twice as much of this most anodic component as culture filtrates M. bovis and M. avium. No protein compo- nents were detected in the spacer gel. 33 34 Table 2. Amounts of protein and carbohydrate in the unheated, ten fold concentrated mycobacterial culture filtrates mg/ml of Culture Filtrate Protein Carbohydrate 50 (Group III) 1.4 0.47 68 (Group III) 1.4 0.55 310 (M. bovis) 2.3 0.47 132 (M- avium) 6.0 0.30 35 Fig. 1. Analytical disc electrophoretic patterns of culture filtrates 68, 50, Mycobacterium bovis (bov), and M. avium (av) stained with amido black. 36 Oh A B 50 Fig; 2. Schematic representation of the protein (A) and PAS positive (B) components in culture filtrates 68 and 50. The letters S and L refer to the spacer and lower gel areas, respectively. Fig. 3. 37 A 'T' A 'T M. bovis M. avium Schematic representation of the protein (A) and PAS positive (B) components in culture filtrates Mycobacterium bovis and M. avium. The letters S and L refer to the spacer and lower gel areas, respectively. 38 All the culture filtrates contained PAS positive material (Figs. 2 and 3). The number of PAS positive compo- nents in the culture filtrates were as follows: eight in M. ngig and five in.fl-.2X£EE (Fig. 3); seven in both 68 and 50 (Fig. 2). All the culture filtrates had a PAS positive com- ponent at the surface of the spacer gel and at the interface of the spacer gel and lower gel. Three of the culture fil— trates, 68, 50, and M. ngis had an additional PAS positive band in the spacer gel. The other PAS positive bands in the four culture filtrates were located in the upper third of the lower gel. Most of these PAS positive components could be correlated with amido black-staining components. The PAS positive material in the culture filtrates did not react strongly with Schiff's reagent. The strongest reactions occurred with the material located in the spacer gel. Immunodiffusion analyses 9f the four culture fil— trates. The results obtained with the four reference anti— sera systems are summarized in Table 3. When reference anti— serum 50 was reacted in various combinations with the four culture filtrates from two to eight immunoPrecipitates were observed (Figs. 4 and 5). The homologous system had eight immunoprecipitates. Seven were obtained with culture fil- trate 68 and two each with M. bovis and M. avium. All the immunoprecipitates of culture filtrate 68 were in common identity with those of culture filtrate 50. Only one of the ‘M. avium immunoprecipitates was in common identity with those 39 Table 3. Number of immunoprecipitates in Ouchterlony gel diffusion tests of mycobacterial culture filtrates with the reference antisera Antisera Culture Filtrate 50 68 132 310 50 (Group III) §_ 7 1 0 68 (Group III) 7 .19 3 0 132 (M. avium) 2 4 1 0 310 (M. bovis) 2 2 0 _6_ 4O Fig. 4. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av), and M. bovis (bov) with reference antiserum 50 (center well). 41 Fig. 5. Ouchterlony immunodiffusion of culture filtrates 50, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 50 (cetner well). 42 of 50 and 68. The other M. avium line was in common iden- tity with one line produced by M, bovis. None of the M. bovis immunoprecipitates were in common identity with those of culture filtrate 50. From two to ten immunoprecipitates were formed with culture filtrate 68 reference antiserum (Figs. 6 and 7). The greatest number of immunoprecipitates was obtained with the homologous system. Seven, four and two separate immuno- precipitates occurred with culture filtrates 50, M. avium and M. bovis, respectively. All culture filtrates with the exception of M. bovis had lines of common identity with the homologous system. One line of common identity occurred among culture filtrates M. avium, 68 and 50. The M. avium culture filtrate possessed one other antigen in common iden— tity with culture filtrate 68, not in common identity with 50. None of the immunoprecipitates of M. bovis were in common identity with any others. With the M. avium reference antiserum, eight immuno- precipitates occurred with the homologous culture filtrate (Fig. 8). No reaction was detected with the M. bovis cul- ture filtrate. Of the three immunoprecipitates obtained with culture filtrate 68, two were in common identity with those oko. avium. One immunoprecipitate formed with cul- ture filtrate 50. The one immunoprecipitate with culture filtrate 50 and the third immunoprecipitate with culture filtrate 68 were not in common identity with each other or with M. avium. 43 Fig. 6. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 68 (center well). 44 Fig. 7. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with reference antiserum 68 (center well). 45 Fig. 8. Ouchterlony immunodiffusion of culture filtrates 50, 68 and Mycobacterium avium (av) with the M. avium reference antiserum (center well). 46 The M. bovis reference antiserum reacted only with the homologous culture filtrate (Fig. 9). Six immunoprecip- itates were observed. Immunoelectrophoresis. Nine antigens were detected in the culture filtrates 50 and 68 by immunoelectrophoresis (Table 4). The immunoprecipitate patterns of the two cul- ture filtrates were very Similar. No immunOprecipitates were found on the cathode side of the origin. Two of the immunoprecipitates extended almost to the sample origin. At least two immunOprecipitates were close to the sample origin with slight lateral displacement. Some of the immunoprecip- itates had similar positions and curvatures, and were stained very lightly. Eight immunoprecipitates were present after electro- phoresis of the M, bovis culture filtrate (Table 4). None of the immunoprecipitates was on the cathode side of the sample origin. There was only slight lateral displacement from the center of diffusion with three of the immunoprecip- itates. Those antigens which migrated the least, stained lightly. . Fifteen antigens were detected in the M, avium cul- ture filtrate by immunoelectrophoresis (Table 4). Only one antigen was found on the cathode Side of the sample origin (Figure 10). One very darkly stained immunoprecipitate began at the cathode side of the sample origin and extended far toward the anode side. The lateral displacement of one 47 Fig. 9. Ouchterlony immunodiffusion of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with M. bovis reference antiserum (center well). 48 Table 4. Number of immunOprecipitateS in immunoelectrophore- sis (IE) and disc immunoelectrophoresis (DIE) tests of the four mycobacterial culture filtrates with the reference antisera DIE Antisera IE Homologous Culture Filtrate 50 68 132 310 Antisera 50 (Group III) .12 11 1 o 9 68 (Group III) 12 .l2 7 0 9 132 (M. avium) 4 6 _l_6_ 0 15 310 (M. bovis) 2 4 0 g 8 49 ($5.45-; ‘_I v9-1; k‘k‘gx- -‘ Fig. 10. Immunoelectrophoresis of Mycobacterium avium cul- ture filtrate with homologous reference antiserum. 50 immunoprecipitate was slight in comparison to the others. Most of the immunoprecipitates stained Sharply. Disc immunoelectrophoresis. When the reference serum 50 was tested against gels in which the four culture fil- trates had been electrophoresed, culture filtrates 50, 68, .M. avium, and M. bovis had 13, 11, four and two antigens, respectively (Table 4). The lateral displacement of the immunoprecipitates away from the gels generally was progres- sively greater for antigens farther from the top of the lower gel and was consistent for all of the culture filtrates re- acted with the four reference antisera. The immunoprecipitate patterns of culture filtrates 50 and 68 were very similar. There were at least three antigens located in approximately the same*position in the most distal part of the gel. One antigen was located at the surface of the lower gel, and the immunOprecipitate was extremely faint and formed close to the gel. The four anti- gens of M. avium and two antigens of M. ngi§_which reacted with the reference serum 50 were located in the lower half of the gels. There was no evidence of immunoprecipitates with the components found in the spacer gels with any of the four reference antisera. Culture filtrates 68, 50, M. avium and M, bovis formed 12, 11, six and four immunoprecipitates, respectively, with the reference serum 68 (Table 4). The position and lateral displacements of the immunoprecipitates were Similar 51 to those with the reference serum 50. With the exception of one, all the immunoprecipitates formed by M. avium and M. bovis antigens were in the lower half of the gel. When the 5°.22l23 culture filtrate was electropho- resed 16 immunoprecipitates formed with the homologous ref- erence antiserum (Fig. 11; Table 4). Seven and one antigens were detected in the culture filtrates 68 and 50, respective- ly, when reacted with the M. gyigg reference serum. There was no reaction with the M. bovis culture filtrate. Most of the antigens of the M. avium culture filtrate and culture filtrate 68 were located in upper two-thirds protion of the gel. Several of the M. avium immunoprecipitates extended almost the full length of the gel. Consistent with the other immunodiffusion tests, the 4M.‘ngi§ reference antiserum reacted only with the electro- phoresed M, ngig culture filtrate (Table 4). Nine immuno- precipitates were observed. Three of these immunoprecipi- tates were in the upper third of the gel with slight lateral displacement and were lightly stained. The remaining Six immunoprecipitates were in the lower half of the gel, more laterally displaced and stained sharply. Preparative disc electrophoresis. The optimal con- centration of protein for preparative disc electrophoresis was between 2.1 and 2.6 mg, in no more than 1.5 ml, approx- imately ten fold more than for analytical disc electrophore— sis. Exceeding 2.6 mg caused broadening and overlapping Fig. 11. 52. " ”x g z I; r i :5 : :‘l-t.‘ Disc immunoelectrophoresis of disc gels of culture filtrates 50, 68, Mycobacterium avium (av) and M. bovis (bov) with M. avium reference antiserum (long, narrow trough). The letters S and L refer to the spacer and lower gel areas, respectively. 53 adjacent components. The resolution of the components was equal to that in the analytical gels and by increasing the gel length many of the components were separated further from adjacent ones. The components in the eluates from the preparative gels were numbered as in the analytical disc electrophoretic schematic drawings (Figs. 2 and 3). The bands were numbered from the top of the gels to the bottom. The eluates were coded so that the first number indicated the culture fil- trate and the second number the number of the disc band from which the antigen was eluted. Re-electrophoresis of the eluates obtained from gel segments affirmed the accuracy of the sectioning process. A representative disc electropho- retic pattern of the eluates obtained from five of the seg- ments cut from a culture filtrate 50 disc preparative gel is presented in Figure 12. All of the culture filtrate 50 eluates appeared as single bands when re-electrophoresed in analytical disc gels. The number of amido black-staining components in each of the 20 eluates is presented in Table 5. Two of the culture filtrate 68 eluates had two amido black-staining components, while the remaining three had only one component. Of the five M. avium eluates, one had three components and another two. Three of the M. bovis eluates contained only one amido black-staining component. Two of the eluates contained two components. 54 .‘AOIN‘OO.O , J . . s .1er I‘C‘mm. n. .. ”Jo /vw4 £.¥§ . . .41.». "NADA. . n , . . . cabin»... nfi4~rwaww5 ..., 5%.“? .fiswxfinmfifimafima xx... . . 1.1.2...- .\ .51 tLWflC {fictive} 5.: 13:19.1 . 1 "£13... 51 , y . Analytical disc electrophoretic patterns of culture filtrate so (extreme left) and five componen 12. Fig. 1V6 ts re— from preparat ing electrophoresed after section gels stained with amido black. 55 Table 5. Number of protein-staining components detected by re- electrophoresis and protein concentrations of the 20 eluates obtained from preparative disc electro- phoresis Number of Amido Black- Protein Concentration Eluate # Staining Components ug/ml 50-101 1 80 50-11 1 110 50-13 1 170 50-14 1 140 50-18 1 340 68-1 1 160 68-8 2 120 68-10 2 140 68-17 1 200 68-18 1 300 132-2 1 80 132-3 3 130 132-12 1 270 132-19 2 290 132-22 1 370 310-14 2 120 310-15 1 260 310-21 2 1120 310-23 1 410 310-24 1 270 1The first number indicates the strain. 50 and 68 are Group III mycobacteria, 132 is M. avium, 310 is M. bovis.. The second number is the number of disc band from which the antigen was eluted. 56 The protein concentrations ranged from 80 ug to 1120 pg per ml (Table 5). In general, the concentration was greater for those components found progressively more anodic in the gel. Based on theoretical calculations, 20 to 60% of the protein was recovered. The values obtained for the pro— tein concentration of the eluates were corrected for resid- ual TEMED, a chemical component of the gels which reacted in the Lowry procedure for the determination of protein. The four most anodic components which were found in eluates 50-18, 68-18, 132-22 and 310-24, were re-electropho- resed in 10% disc analytical gels. In 7% analytical gels, these eluates appeared as single amido black-staining compo- nents. A representative disc electrophoretic pattern of the eluate 50-18 electrophoresed in a 7 and 10% gel is presented in Fig. 13. In the 10% gel, the single component was separ rated into six. The single components in 7% gels, in the eluates 68-18, 132-22, and 310-24 were separated into five, three and components respectively in the 10% gels. Those components separated in 10% gels which did not correspond to the original band that was cut were stained very faintly with the amido black dye. Immunodiffusion analyses 2; the eluates. The number of immunoprecipitates obtained when the 20 eluates were reacted with the four reference antisera are presented in Table 6. Eluate 50-13 had two antigenic components one of which was in identity with the Single antigen in eluate 57 3.4:, 5m t.1 «a r». “wry .«rwfyvaHtfl. v 911. ‘1 A . 'aAJ‘v .a,»~n tterns of eluate 7 and 10% gels. d from culture filtrate 50 prepara- 1ectrophoretic pa isc e 1 d ine 1C8. 18 obta tive gels on re-electrophoresis in Stained with amido black. Analyt . 13. 50 Fig 58 Table 6. Number of immunoprecipitates in Ouchterlony gel diffusion tests of 20 eluates obtained by prepara- tive disc electrophoresis of mycobacterial culture filtrates with four reference antisera Antisera Eluate 50 68 132 310 50-101 1 l 0 0- 50-11 1 1 O 0 50—13 2 2 O 0 50-14 1 l O 0 50-18 3 2 O 0 68-1 0 l l 0 68-8 1 l O 0 68-10 1 2 1 0 68-17 2 2 O 0 68-18 2 3 O 0 132-2 0 l 1 0 132-3 0 l 3 0 132-12 0 1 3 0 132-19 1 1 3 0 132-22 0 O 1 0 310-14 0 0 O 0 310-15 0 0 0 2 310-21 0 O O 1 310-23 0 O O 1 310-24 0 O O 1 l The first number indicates the strain. 50 and 68 are Group III mycobacteria, 132 is M. avium, 310 is M, bovis. The second number is the number of disc band from which the antigen was eluted. 59 50—14. One of the eluates, 50-18, electrophoresed as a single band but formed three immunoprecipitates with the homologous reference antisera. All of the culture filtrate 50 eluates reacted with the culture filtrate 68 reference serum, but none reacted with the culture filtrate M. avium or M. bovis reference antisera. Eluate 50-18 had one anti- gen more when reacted with the homologous reference anti— serum than with the culture filtrate 68 reference serum. Nine separate immunoprecipitates occurred in the five eluates of culture filtrate 68 with the homologous anti- serum. In one eluate, two amido black components were pres- ent in the disc gel but only one was antigenic. Eluate 68—18 contained three antigens but electrophoresed as a single component. Using the culture filtrate 50 and M. avium reference antisera, six and two immunoprecipitates were observed, respectively, with the five culture filtrate 68 eluates. No immunoprecipitates formed with the M. bovis reference antiserum. Two of the culture filtrate 68 eluates had at least one antigen not in common with the other cul- ture filtrates. The five M. avium eluates formed 11 separate immuno- precipitates with the homologous reference antiserum. Three of the eluates contained three antigens each. The reference antisera 50 and 68 reacted with the antigens found in the .M. avium eluates yielding one and four immunoprecipitates respectively. No reaction was observed with the M. bovis 60 reference antiserum. Four of the eluates contained specific antigens but they were mixed with antigens which cross reacted with reference antisera 50 and 68. Only eluate 132-22 had an antigen which did not cross react with the other three reference antisera. The M. bovis eluates reacted only with their homol- ogous reference antiserum. Five separate immunOprecipitates occurred. Eluate 310-14 had two amido black-staining compo- nents but neither of them reacted with the reference anti- serum. Two antigens were detected in the eluate 310-15 but only one component was observed in the disc gel after re- electrophoresis. The Opposite was true of eluate 310-23; there were two amido black-staining components but only one was antigenic. The reactions of the 20 eluates with 11 cell-culture filtrate antisera are recorded in Table 7. All eluates of culture filtrate 50 reacted with the three Group III anti- sera 50, 68, and P39. Four reacted with the Group I anti- serum P4. Each of the eluates formed two immunoprecipitates with the Group III P39 antiserum. Antigens common only to Group III mycobacteria were found in eluate 50-18. Eluates 68-1 and 68-8 reacted with four of the Group III antisera and to M, bovis and to attenuated M, bovis (BCG) antisera. An immunOprecipitate formed with the 68—10 eluate and BCG antiserum. Eluate 68-18 contained antigens specific for the Group III mycobacteria. 61 Table 7. Number of immunoprecipitates in Ouchterlony gel diffusion tests of 20 eluates obtained by prepara- tive disc electrophoresis of mycobacterial culture filtrates with 11 cell—culture filtrate antisera Antisera . 62 172 132 310 BCG P4 P15 ‘— Eluate # P392 50 0‘ (I) U1 l-‘ 50-101 50-11 50-13 50-14 50-18 68-1 68-8 68-10 68-17 68-18 132-2 132-3 132-12 132-19 132-22 310-14 310-15 310-21 310-23 310-24 0 +4 H +4 H +4ro Hacs o +4 N +4 N +4 N no N no N o c: H +4 o +4 N +4 0 C) N no N no H +4 N +4 N +4 o +4 H +4t4 Have 0 c: o no N be m +4 H +4 N be H o c: o +4+4 Ha+4 o c: H +4 o c: H +4 o c: o C) o c: o C) H +4 o +4 H C) H c: o c: H +4 o c: o c: o o c: 0 c3 0 c: H c: o c: o c: o c: o C) o c: o c3 C) o c: o +4 o as H +4 N c3 H +4 o c: o c: o c: o H +4 H +4 H c: o +4 H na<3 Ha+4 H +4 o :3 o c: o +4 H no N»+4 o c: o +4 H C) o +4 H +4 0 :3 o +4 o <3 0 +4 H c><3 H'CD o c><3 HI+4 o C) o +4 H +4 H o c: o c: o c><3 c>