CHRQMATQGRAFHZC, ELECTROF‘HGRETK. A2419 EMMUNQELECTRQPHOREHC STUDIES OF SERUMS FROM NORMAL AND TUBERCULQUS GUINEA. PIGS Them for flu chm of M. 5. MICHEGAN STATE UNIVERSITY Terry Jay Dardas 1964 was LIBRARY Michigan State University CHROMATOGRAPHIC, ELECTROPI—IORETIC, AND IMMUNOELECTROPHORETIC STUDIES OF SERUMS FROM NORMAL AND TUBERCULOUS GUINEA PIGS BY Terry Jay Dardas A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1964 AC KNOW LEDGEMENTS The author would like to express his sincere thanks to Dr. V. H. Mallmann for her interest and assistance during the research and preparation of the manuscript“ The assistance and suggestions offered by Dr. R. B. Dardas and Mr. Go F0 Dardas are also appreciated. A particular thanks goes to my family who gave purpose to the desire for an intellectual pursuit. >1: >:< >:< >:< >:< >:< ::< >1: 3}: :1: >:< >:< >:< >:< 3:: >;< ii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . l HISTORICAL REVIEW. . . . . . . . . . . . . . . . . . . . . 3 Changes in plasma constituents during inflammation. . 3 Changes in the blood during tuberculosis in man . . . . 5 Changes in the serum proteins of guinea pigs during experimental tuberculosis . . . . ..... . . . . 9 Gel filtration . . . . ..... . . . . . . . . . . . . 10 Ion- exchange Column Chromatography. . . . . . . . . 11 Zone electrophoresis . . . . . . . . . . . . . . . . . . 14 ImmunoelectIOphoresis . . . . . . . . . . . . . . . . . 18 MATERIALS AND METHODS . . ..... . . . . . . . . . . 29 Collection of serums . . . . . . . . . . . . . . . . . 29 Rabbit antiserums specific for normal guinea pig serum 29 Mycobacterium bovis . . . . . . . . . . . ..... . . 3O Guinea pigs infected with M. bovis . . . . . . . . . . . 31 Lung fluid . . . . . ..... . . . . 31 Guinea pigs sensitized with heat killed M. bovis. . . . 31 Gel filtration. . . . . . . ....... . . . . . . . . 32 Ion- exchange Column Chromatography. . . . . . . . . 33 Starch Block Electrophoresis. . . . . . . . . . . . . . 35 Agar Gel ElectrOphoresis. . . . . . . . . . . . . . . 36 Cellulose Acetate Membrane Electr0phoresis . . . . . 37 Protein, Lipoprotein, and Glycoprotein Stains . . . . . 38 Ouchterlony Immunodiffusion . . . . . . . . . . . . . . 39 Irnmunoelectrophoresis . . . . . . . . . . . . . . . . . 40 RESULTS 43 Examination of Individual Rabbit Antiserums . . . . . 43 Gel Filtration of Normal Guinea Pig Serums in Sephadex G-100 and G-ZOO . . . . . . . . . . . . . 51 iii TABLE OF CONTENTS - Continued Page Ion- exchange Column Chromatography of Normal Guinea Pig Serums on DEAE- cellulose . . . . . 55 Starch Block Electrophoresis of Normal Guinea Pig Serums . . ..... . . . . 60 Agar Gel Electrophoresis of Normal Guinea Pig Serums ..... . . ..... . . . . . . . . 65 Cellulose Acetate Membrane Electrophoresis of Normal Guinea Pig Serums ..... . . . . . . 65 Freezing and Filtration of Normal Guinea Pig Serums 68 Immunoelectrophoresis of Normal Guinea Pig Serums 68 Identification of a Lipoprotein and a Major Glyco- protein in Normal Guinea Pig Serums ....... 72 Ouchterlony Immunodiffusion of Serums from Normal and Tuberculous Guinea Pigs . . . ..... . . 72 Agar Gel Electrophoresis of Serums from Tuberculous Guinea Pigs . . . . . . . . . . . . . . . . . 74 Cellulose Acetate Membrane Electrophoresis of Serums from Tuberculous Guinea Pigs . . . . . . 74 Immunoelectrophoresis of Serums from Tuberculous GuineaPigs..................... 76 Lungfluid. .. 88 Analyses of Serums from Guinea Pigs Sensitized with Heat Killed M. bovis . . . . . . . . . . . . . . . . 88 DISCUSSION.......................... 96 Examination of Individual Rabbit Antiserums . . . . . 96 Analyses of Serums from Normal Guinea Pigs. . . . . 99 Analyses of Serums from Tuberculous Guinea Pigs . . 107 SUMMARY...........................115 LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . 117 iv TABLE LIST OF TA BLES "Results of immunoelectrophoretic analyses of ten normal human serums with four different anti- serums. " From Hirschfeld (1963). . . Results of reciprocal interfacial precipitation tests with individual rabbit antiserums Results of reciprocal interfacial precipitation tests with individual normal rabbit serums Page 27 48 48 LIST OF FIGURES FIGURE Page 1. IInmunoelectrophorogram of normal guinea pig serum developed with antiserum collected from rabbit number 7 ten days after the last injection of guineapigserum................... 44 Z. Immunoelectrophorogram of normal guinea pig serum developed with antiserum from rabbit number6....................... 45 3. Immunoelectrophorogram of normal guinea pig serum developed with antiserums collected from rabbits numbers 6 and 7 five days after the last injection of guinea pig serum . . . . . . . . . . . . 46 4. IInmunoelectrophorogram of normal guinea pig serum developed with antiserum from rabbit number 6 (doubled course of immunization) . . . . 47 5. Ouchterlony immunodiffusion of normal rabbit serums and antiserums from rabbits numbers 1 and6..... ..... 49 6. Ouchterlony immunodiffusion of normal rabbit serums and antiserums from rabbits numbers 2, 4,5,and6..................... 50 7. Immunoelectrophorogram of antiserums collected from rabbit number 6, 5 and 160 days after the last injection of guinea pig serum . . . . . . . . . . . . 52 8. Serum protein distribution in gel filtration (Sephadex G-100 and G-ZOO) effluent fractions . . . 53 9. Cellulose acetate electrophorogram of gel filtration (Sephadex G-100) effluent serum fractions . . . . . 54 vi LIST OF FIGURES - Continued FIGURE 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Ouchterlony immunodiffusion of gel filtration (Sephadex G-100) effluent serum fractions. Serum protein distribution in chromatographic (DEAE-cellulose) effluent serum fractions Cellulose acetate electrophorogram of chromato- graphic (DEAE-cellulose) effluent serum fractions Ouchterlony immunodiffusion of chromatographic (DEAE-cellulose) effluent serum fractions hnmunoelectrophorogram of chromatographic (DEAE-cellulose) effluent serum fractions Serum protein distribution in eluates of starch block segments after electrophoresis of normal guinea pig serum . . Cellulose acetate electrophorogram of serum fractions in eluates of starch block segments after electrophoresis. Ouchterlony immunodiffusion of eluates of starch block segments after electrophoresis Electrophorogram of normal guinea pig serum in agar gel with the Hirschfeld buffer solution (circular origin) Electrophorogram of normal guinea pig serum in agar gel with the Owen buffer solution (slot shaped origin). . . A composite immunoelectIOphorogram of normal guinea pig serums Ouchterlony immunodiffusion of serums from normal and tuberculous guinea pigs vii Page 56 57 58 59 61 62 63 64 66 67 69 73 LIST OF FIGURES - Continued FIGURE 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Typical densitometric recordings of cellulose acetate membranes after electrophoresis of serums from normal and tuberculous guinea pigs . IrnmunoelectIOphorograms of serums from guinea pigs pre-inoculation and 7 days post-inoculation with Mycobacterium bovis . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 14 days post-inoculation with Mycobacterium bovis . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 21 days post-inoculation with Mycobacterium bovis . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 28 days post-inoculation with Mycobacterium bovis . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 33 days post-inoculation with Mycobacterium bovis . . . . . . . . . . . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 41 days post-inoculation with Mycobacterium bovis . A composite immunoelectrophorogram of serums from guinea pigs collected during the terminal stages of tuberculosis . Composite immunoelectrophorograms of normal serums and serums from guinea pigs collected dur- ing the terminal stages of tuberculosis Comparative cellulose acetate densitometrograms and immunoelectrophorograms of pre-inoculation serum, and serum and lung fluid from a guinea pig post-inoculation with Mycobacterium bovis . viii Page 75 77 78 80 82 84 85 87 90 91 LIST OF FIGURES - Continued FIGURE 32. 33. 34. Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 7 days post-inoculation with heat killed Mycobacterium bovis. Innmunoelectrophorograms of serums from guinea pigs pre-inoculation and 14 days post-inoculation with heat killed Mycobacterium bovis. . Immunoelectrophorograms of serums from guinea pigs pre-inoculation and 21 days post-inoculation with heat killed Mycobacterium bovis. ix Page 93 94 95 INTRODUCTION The ultimate aim of most studies of disease in experimental animals is to disclose pertinent features of pathogenesis common to the natural disease in man and animals. The common features can then be studied in greater detail under carefully controlled conditions to better understand the disease processes. From these studies evolve improved preventive, diagnostic and therapeutic procedures. Many attempts have been made to find specific antigens, antibodies or other components in the blood plasma of diagnostic or prognostic value for tuberculosis. Although some success has been achieved, most of the tests are nonspecific. When techniques to fractionate blood plasma were developed and refined, some were adapted for clinical use, particularly zone electrophoresis. These tests also suffer from nonspecificity because plasma constituents usually reflect the overall physiologic state of the body rather than a specific disease process. Immunoelectrophoresis is one of the most sensitive tools for the analysis of the antigenic composition of blood plasma. Since its introduction in 1953 (Grabar and Williams), it has been used in a wide variety of research and clinical problems. Quantitative and qualitative changes occur in the serum proteins during tuberculosis. It can be anticipated that these and perhaps additional changes can be detected by immunoelectrophoresis. Simplicity and availability to modern clinical laboratories make it ideally suited for clinical application if immunoelectrophoretically detectable and significant changes occur during tuberculosis . This is a report of studies of normal guinea pig serum fractionated by (1) zone electrophoresis in insoluble potato starch, agar gel, and cellulose acetate, (2) gel filtration in Sephadex G-100 and G-200, and (3) by anion- exchange column chromatography on DEAE-cellulose. The immunoelectrophoretic pattern of normal serum is described and several specific proteins identified. A study was made of the electro- phoretically and immunoelectrophoretically detectable serial changes in the serum proteins, lipoproteins, and glycoproteins of guinea pigs infected with viable Mycobacterium bovis, and guinea pigs sensitized to tuberculin with heat-killed M. bovis. HISTORICAL REVIEW To evaluate changes in plasma proteins during tuberculosis, it is necessary to be cognizant of the changes occurring in normal individuals and during disease in general. "Any damage to the tissues whether it be mechanical, thermal, or toxic, will produce some degree of the local reactions included in the term inflammation. It is not so well recognized that such damage may also produce in the plasma and cells of the blood, changes which are part of the general reaction to injury” (MacFarlane, 1962) . Charfis in Plasma Constituents During Inflammation A part of the systemic inflammatory response is the change in the relative concentration of many of the plasma constituents including antibody (Marshall, 1956). Changes that occurred simultaneously with the local signs of acute inflammation were referred to as the primary inflammatory blood protein reaction (PIR) (Odenthal, 1958!. Among these changes were hypoalbuminemia, hyperalphaz—globulinemia, and the appearance of CRP, the C-reactive protein ('Tillet and Francis, 1930L The PIR represents a generalized reaction to injury and is quite nonspecific. It was almost invariably detected in the acute stages of most bacterial infections (Belfrage, 1963). It also occurred in persons with myocardial infraction (Linko and Waris, 1955) and following mechanical trauma (Hock-Ligeti, et a1. , 1953) and pyrogen administration (Hedlund, 1961). Hypoalbuminemia regularly occurs during a variety of inflammatory states in man and animals. One of the chief functions of albumin is to 3 regulate the plasma volume (Guyton, 1961). According to Bjorneboe and Schwartz (1959), albumin plays a key roll in a homeostatic mechanism which controls the plasma colloid osmotic pressure. A major cause of hypoalbuminemia in rabbits was an increase in the plasma volume (Bjorneboe and Jarnum, 1961). However it can also result from impaired synthesis, increased catabolism, or extravascular plasmaphoreis (Knutti, et a1., 1950; Petermann, 1960). Hypoalbuminemia always accompanied hyperalphaz-globulinemia during the PIR in acute bacterial infections (Belfrage, 1963). Changes in the concentration of fibrinogen, the CRP, and several glycoproteins often called acute phase substances (Kelley, 1952) occurs as a part of the PIR. Maximum changes in the acute phase substances occurred between three to five days after the onset of acute inflamma- tion (Belfrage, 196 3). Haptoglobin, one of the acute phase glycoproteins, was found to be the chief cause of hyperalphaz-globulinemia during the PIR due to a variety of causes (Nyrnan, 1959; Hever and Kalnai -Hever, 1962). Changes in the concentrations of fibrinogen and haptoglobin were directly related during the PIR (Nyman, 1959). A secondary inflammatory blood protein reaction (SIR) occurred during acute bacterial infections of long duration and in chronic bacterial infections (Odenthal, 1958; Belfrage, 1963). Hypergammaglobulinemia was the most common finding during the SIR. Hyperbetaglobulinemia usually accompanied hypergammaglobulinemia but sometimes they occurred independently. Hypergammaglobulinemia was usually indica- tive of increased antibody production (Tiselius and Kabat, 1939; Gross, et a1. , 1959; Askonas, 1960) although some gamma globulin may be serologically "inert”l (Baldwin and Iland, 1953). Three distinct molecu- lar species have been found to possess antibody activity and have b een grouped together in the human immunoglobulin system (Heremans, 1960). They include betazA, betazM and the 7 S gamma; globulins. In most bacterial diseases, the gamma globulin concentration did not parallel resistance (Belfrage, 1963). Hyperbetaglobulinemia was usually found to be due to an increase in the concentration of betalC (Belfrage, 1963), one of the complement components (Laurell and Lundh, 1962). The systemic reactions to most bacterial diseases were similar to those that occurred during bacterial pneumonia (Belfrage, 1963). Infections of short duration caused intense PIR but little, if any, SIR. In more persistent and severe infections, the PIR was pronounced and the beta; and gamma globulin concentrations were increased. Hypo- proteinemia usually developed coincident with hyperglobulinemia. A positive correlation was commonly found between the concentration of the beta; and gamma globulins as well as with all of the other com- ponents of the PIR. Many of the disease-induced alterations that have been detected in the plasma constituents of man and animals are included in the reviews by Luetscher, 1947; Gutman, 1958; Lever, 1951; Sterling, 1951; Sobotka, 1955; Jenks, et al., 1956; Reiner, 1957; Graham, et al., 1958; Owen, 1958; Petermann, 1960; Lewis, 1960; Carson and Mathingley, 1960; Belfrage, 1963. Changes in the Blood During Tuberculosis in Man Many attempts have been made to find specific changes in the blood of individuals suspected of having tuberculosis that might be of diagnostic or prognostic significance. No single test or combination of tests is undisputably specific for tuberculosis. However the results of these studies have contributed considerably to our knowledge of its pathogenesis . Numerous cellular changes occur. Leucocytosis and mild secondary anemia were not uncommon in patients with advanced pul- monary tuberculosis (Long, 1958). The erythrocyte sedimentation rate (ESR) is widely used in clinical laboratories as a measure of disease activity. The ESR was found to increase roughly with increas- ing extent and activity of tuberculosis.(Houston, et a1. , 1949; Todd, 1953; Benson and Goodard, 1954). However most investigators agree that the ESR is of no diagnostic significance since hyperproteinemia alone can cause an increase in the ESR (Malmros and Blix, 1946). Measurements of blood volume, pH, non- nitrogenous protein content, and inorganic ion content have failed to establish any consistent relationships which are of diagnostic or prognostic value (Hochen, 1922; Kolmer, et al0 , 1948; Markowitz, et al., 1955; Long, 1958). The serum proteins have been studied extensively and consider- able controversy exists among various workers. Differences in instrumentation and procedure, and protein nomenclature undoubtedly account for some of the controversy. In spite of the intensive study, plasma protein or glycoprotein patterns of diagnostic significance for tuberculosis have not been found. Some of these changes however aid in determining the extent and activity of the disease, and the prognosis. The plasma protein changes are similar to those of the PIR and the SIR which occur during acute bacterial diseases such as pneumonia (Longsworth, et al., 1939; Bruce and Alling, 1948; Belfrage, 1963). Hypoalbuminemia and hype rglobulinemia were common in all stages of tuberculosis (Eichelberger and McCluskey, 1927; Bing, 1940; Volk, et a1. , 1953; Pilheu, et a1. , 1959). . Hyperglobulinemia was caused mainly by increases in the euglobulin and pseudoeuglobulin 1 fractions (Gutman, et al., 1941). Early studies of serum by moving boundary electrophoresis revealed changes that roughly paralleled the extent and activity of the disease. Hypoalbuminemia and a reduction in the normal albumin/ globulin ratio (A/G) were common (Luetscher, 1941; Seibert and Nelson, 1942; Seibert, et al., 1943; Seibert, et al., 1947). Hyper— globulinemia was caused by an increased concentration of all of the globulin fractions, particularly the alpha; globulins (Baldwin and Iland, 1953; Kries, et al., 1955; Harrower, et al., 1957). Gillihand, et al. (1956) found a simultaneous and consistent reduction in albumin and rise in the alpha; globulins during tuberculosis. On the other hand, Jahnke and Scholteon (1951) reported that albumin and the, globulins fluctuated independently. More recently, Belfrage (1963) found that albumin and the globulins reached their respective minimal and maximal levels simultaneously and returned to normal at approximately the same rates. Changes in the alpha globulins were more closely related to clinical progress than changes in albumin, beta, or gamma globulins (Park, 1961). The alpha globulins decreased from previously high amounts when clinical improvement was most rapid. Increased alpha; and gamma globulins were the least favorable prognostic sign (Zitrin, et a1. , 1959). A positive correlation was found between the maximum values of the alpha; and the gamma globulins during advanced pulmonary tuberculosis (Belfrage, 1963). Haptoglobin contributed most to the rise in the alpha; globulins (Nyman, 1959; Hever and Kalnai-Hever, 1962). Because of the relationship between hyperalpha;-globulinemia and disease activity, some workers believe that the albumin/alpha; globulin ratio is more valuable and specific than the A/G ratio (Shaw, 1956; Pilheu, et a1. , 1959). The alpha; globulins were not elevated in the serums of patients with tuberculosis caused by nonphotochromogenic atypical mycobacteria (McCuiston and Hudgins, 1960). There is little agreement on the relationship between clinical progress and the hypergammaglobulinemia often observed in advanced and far advanced tuberculosis. Whereas some workers have found a rising or high gamma globulin level to be a good prognostic sign (Seibert, et al., 1948; Levin, et al., 1952; Schaffner, et a1. , 1953; Zitrin, et a1. , 1956), others have failed to find a consistent relation-— ship (Small, 1950; Meyer, et a1. , 1955; Grigorieva and Linishitz, 1960; Park, 1961). Some workers believe that part of the gamma globulin may be serologically "inert" (Baldwin and Iland, 1953; Sher, et a1. , 1956; Gross, et al., 1959). The carbohydrate concentration of serum from tuberculous patients was found to be considerably above normal (Seibert, et al. , 1947; Seibert, et a1. , 1948). Turner, et a1. (1953) found the mean value for the serum mucoproteins to increase from a normal value of l. 98 mg % to 5. 04 mg %. The total serum polysaccharide/mucoprotein ratio was reduced. Electrophoretic patterns of carbohydrate distribution in serums from tuberculous patients were of no diagnostic value (Hirsch and Cattaneo, 1957). Serum proteins and carbohydrates were restored to near normal values following clinical improvement (Seibert, et a1. , 1947; Levin, et al. , 1952; Volk, et a1. , 1953). Tremendous effort and ingenuity has been expended unsuccessfully to develop a reliable serologic test for tuberculosis. Antibodies specific for various mycobacterial fractions occur unpredictably. Furthermore, there seems to be little relationship between the kind or amount of antibodies in the serum and tuberculoresistance (Raffel, 1961). According to Seibert (1960), there may be little progress in the deveIOp- ment of serological tests for tuberculosis because of the dynamic interplay between antigens and antibodies i_n vivo. Changes in the Serum Proteins of Guinea Pigs During Experimental Tuberculosis The changes that occurred in the serum proteins of tuberculous guinea pigs were similar in three respects to those that occurred in man (Weimer, et al., 1954): (1) moderate hypoalbuminemia, (2) hyperglobulinemia, and therefore, (3) a reduced A/G ratio. In contrast to man, a moderate hyperproteinemia was common in the latter stages of the disease in the guinea pig (Weimer and Moshin, 1953). Hyperglobulinemia was due to an increase in the concentration of all the globulin fractions (Sher, et al., 1956; Hudgins and Patnode, 1957; Sher, et a1. , 1958). Within eight days after the subcutaneous inoculation of 0.1 mg of M. tuberculosis, female guinea pigs had increased alpha globulins and alpha; glycoproteins (Weimer, et a1. , 1960). Hypoalbuminemia was common after the 15th day and hyperglobulinemia after the 19th day. The gamma globulins were usually sufficiently elevated after the 60th day to cause hyperproteinemia. Maximum alteration had occurred in all of the fractions except the gamma globulins by the 15 day. The hematocrit remained fairly constant throughout the experiment which indicated that the plasma volume was unaltered. The gamma globulin fraction of the serum from tuberculous guinea pigs contained anti-tuberculopolysaccharide antibodies; anti-tuberculo- protein antibodies were found in the alpha; globulin fraction (Cole and Favour, 1955). Serum gamma globulins from tuberculous guinea pigs contained less carbohydrate than gamma globulins from normal guinea pigs (Sher, et a1. , 1956; Sher, et a1. , 1958). Hypergammaglobulinemia was not invariably associated with tuberculoresistance (Sher, et a1. , 1958). The total polysaccharides and mucoprotein polysaccharides were substantially elevated in serums from tuberculous guinea pigs (Weimer 10 and Moshin, 1953). The principle mucoprotein alteration occurred in the alpha; globulin region (Sher, et a1. , 1956). Since they constituted such a small part of this region it was suggested that even their sub- stantial elevation could not have accounted for the increase in the total alpha; globulins. The serum lipids and lipoproteins were unaltered during tubercu- losis in the guinea pig (Sher, et a1. , 1956). Immunoelectrophoresis has not been used to study the serum protein fluctuations during experimental tuberculosis in guinea pigs. Definite changes were detected in the serums of tuberculous mice by this procedure (Williams and Wemyss, 1961). Gel Filtration The use of cross linked dextran gels for molecular filtration was introduced by Porath and Flodin (1959). Subsequently, granulated agar (Polson, 1961) and cross linked polyacrylamide gels (Hjerten and Mosbach, 1962) have been described. Various grades of cross linked dextrans are commercially available under the trade name of Sephadex (A. B. Pharmacia, Uppsala). Sephadex is a modified dextran of microbial origin (Pharmacia, 1963). Variable degrees of cross linking gives the gel matrix a three dimensional network of pores. Because of its high hydroxyl group content, it is very hydrophilic and swells extensively when placed in aqueous solutions. The structure of the gel matrix is important only in that it determines the extent of swelling and thereby the permeability of the gel (Pharmacia, 1963). The extent of swelling depends on the structure of the gel and the nature of the solution (Porath and Lindner, 1961; Tiselius, et a1. , 1963). Gel filtration involves the distribution of aqueous solvents between two phases, the solvent immobilized by the reticulated polysaccharide network, and the mobile solvent outside 11 the matrix. The porosity of the gel provides the basis of its "molecular seiving" property. Solutes penetrate the gel depending on their steric relationships to the molecular structure of the gel (Tiselius, et a1. , 1963). The distribution of the solute between the two solvent phases depends on both its size and shape. The porosity of the gel therefore, not the solu- bility of the solute determines the separation obtained. Sephadex contained a small number of negatively charged carboxyl groups in alkaline solutions (Miranda, et a1. , 1962), and has been used for the concentration of dilute protein solutions (Glazer and Wellner, 1962). The influence of pH and ionicity however is considerably less important than in ion exchange chromatography (Tiselius, et al., 1963). Gel filtration of concentrated protein solutions should be per- formed in fairly high ionic strength buffer solutions (Porath and Flodin, 1962). At low ionicities, protein-protein interactions occurred which resulted in the formation of relatively stable complexes (Porath and Flodin, 1962; Tiselius, et al., 1963). Because of the wide range of molecular sizes among the serum proteins, gel filtration studies using the weakly cross-linked gel, Sephadex G-200, have been most successful. Human serum was separated into three fractions in this gel (Porath and Flodin, 1962). The first peak contained the alpha; and beta; mac roglobulins. Ceruloplasmin and 7S antibodies were found in the second peak. Prealbumins, albumin, alpha glycoproteins and transferrins were eluted in the third peak. Ion-Excharie Column Chromatography Prior to the introduction of the substituted cellulosic adsorbents by Peterson and Sober (1956) little progress had been made in adapting synthetic ion-exchange resins for use in the column chromatography of 12 the serum proteins. The chief difficulties were the large molecular size and polyionic character of the proteins as well as their instability under the conditions required for elution (Sober and Peterson, 1958}. The cellulosic exchangers have a relatively high protein capacity and adsorbed proteins are eluted under relatively mild conditions (Sober and Peterson, 1958; Peterson and Sober, 1960). The adsorbents were prepared by the attachment of acidic or basic groups through ether linkages to alpha cellulose by a reaction with the appropriate halogen derivative (Peterson and Sober, 1956). Both anion and cation exchangers have been used with serum proteins, but the most widely used adsorbent has been the diethylaminoethylether derivative,DEAE-cellulose (Peterson and Sober, 1960; Serva, 1960). Ion- exchange cellulose column chromatography involves the establishment of multiple electrostatic bonds between oppositely charged sites on the polyionic adsorbent and the protein molecule (Peterson and Sober, 1962). Separation of the proteins is based on their differential requirements for elution from the adsorbent. The affinity of the protein for the adsorbent could be reduced in at least three ways (Sober, et a1. , 1956): (1) decreasing the pH of the eluting solvent which reduces the anionic character of the protein; (2) increasing the pH which decreases the ionization of the tertiary amine group of the adsorbent and reduces its anion binding capacity; and (3) increasing the ionic strength of the eluting solvent. The major effect of pH was to change the charge on the protein and adsorbent; increasing the ionicity of the eluting solvent promoted the dissociation of electrostatic linkages between the protein and the adsorbent (Sober and Peterson, 1958; Peterson and Sober, 1960). Most elution procedures involve a decreasing pH and an increasing ionicity of the eluting solvent. Stepwise changes in the pH and ionicity of the eluting solvent were most effective as a preparatory procedure, but continuous gradient elution was most effective in separating a complex 13 mixture into the maximum number of its constituent fractions (Fahey, 1960; Peterson and Sober, 1962). A variety of continuous gradient elution devices have been described (Sober and Peterson , 1958; Peterson and Sober, 1960; Serva, 1962). Techniques and procedures were discussed by Peterson and Sober (1962). Many of the studies of serum proteins with cellulosic adsorbents have been reviewed by Peterson and Sober (1960). Human serum proteins have been studied extensively by ion- exchange cellulose chromatography. Proteins in normal human serums were eluted from DEAE-cellulose in the order of increasing electro- phoretic mobility (Sober, et al., 1956; Fahey, et a1. , 1958; Goodman, et a1. , 1960; Peterson, et a1. , 1961). Gamma globulins were the first group of proteins to be eluted from DEAE-cellulose (Fahey, et al., 1958). Three separate beta globulin peaks were eluted next followed by alpha globulins separated into seven fractions. Albumin appeared midway in the chromatogram and in most of the fractions collected thereafter. The prealbumins were eluted last. The chromatographic behavior of the protein-bound carbohydrates and lipids, the vitamin B1; binding protein, the acid and alkaline phosphatases, and the thyroxine binding protein was investigated by Fahey, et a1. (1958). The chromatographic behavior of the serum macroglobulins, ceruloplasmin, and siderophilin was elucidated by Peterson, et a1. (1961). The distribution of human serum antigens after chromatography on DEAE-cellulose was studied by IE of the effluent fractions (Goodman, et al. , 1960; Bourrillon, et al., 1962). Peterson, et a1. (1961) and Bourrillon, et a1. (1962) compared the chromatographic distribution of the normal proteins in human serums before and after fractionation with ammonium sulfate. The effluent fractions were analyzed by filter paper l4 electrophoresis (Peterson, et a1. , 1961) and by starch gel electro- phoresis and IE (Bourrillon, et al. , 1962). The molecular heterogeneity of components of the human gamma system (Heremans, 1960) was substantiated by results of DEAE-celluc lose chromatography (Sober, et a1. , 1956). Fahey and Horbett (1959) subfractionated human serum gamma globulins on DEAE-cellulose. The first four subfractions contained 6. 6S globulins with progressively increasing electrophoretic mobility and carbohydrate content (1. 1% to 1. 3% hexose). The four subfractions were antigenically indistinguish— able. A fifth subfraction contained all 185 globulins with a hexose con- tent of five percent and antigenic determinants not found in any of the other subfractions. The chromatographic distribution of the serum proteins from human patients with a variety of disease were studied by Fahey, et a1. (1958) and by Toombs and Maclagan (1960). The latter obtained im- proved resolution when the serums were first fractionated by half saturation with ammonium sulfate. Turcotte, et a1. (1963) used DEAE- cellulose to separate the antibodies responsible for the hemagglutinating activity of serums from healthy tuberculin positive individuals and individuals with active pulmonary tuberculosis. Zone Electrophore sis Electrophoresis is the separation of charged ions or polyions by an electrical field at a particular pH. Separation is based chiefly on the degree of ionization of the constituents of the sample and thus their electric charge. Some types of zone electrophoresis (starch gel, disc. agar gel) also involve gel filtration. Glass wool, glass beads, asbestos, silica gel, agar gel, starch gel, filter paper, and cellulose acetate have been used as supporting materials for various types of zone electrOphoresis (Kunkel, 1954; 15 Cooper, 1960). In general zone electrophoretic techniques are superior to moving boundary techniques as described by Tiselius (1939) in that they are: (1) more sensitive, (2) more adaptable for preparative pur- poses, and (3) simpler and more convenient to perform. The basic design of all zone electrophoresis systems includes a system of buffer vessels which are connected by an electolyte bridge in which the sample is placed (Kunkle, 1954; Bodman, 1960; Crowle, 1961). When an electrical potential is placed across the bridge, the current moves from the anode to the cathode. The flow of electrons is toward the anode (Efron, 1960). All ionized electrolytes migrate depending on their sign and degree of ionization. Since the resistance of the bridge is relatively fixed and determined by the conditions of the experiment, the only way to increase the current is to increase the potential difference between the electrodes (Efron, 1960). However, heat is produced in the bridge, its magnitude being a function of the square of the current. As the bridge warms, the resistance decreases and allows more current to flow. Excessive current can cause drying of the supporting medium, pH changes as a result of electrolysis of buffer salts, and increased ionic mobilities. - Either the voltage or the current must be regulated to avoid these difficulties. When the voltage is regulated, the current and the bridge resistance vary according to Ohm's law. Unless pre- cautions are taken, excessive heating of the bridge can occur. When the current is regulated, both the voltage and the resistance are reduced with time. In general when the bridge temperature must be regulated, constant current is preferred (Shandon, 1960). These and other factors controlling the separation obtained by zone electrophoresis were dis— cussed by Valrnet and Svensson, (1954), Block, et a1. (1955), Leder (1955) Wolstenholme (1956), Bier (1959), and Cooper (1960). 9 16 The main functions of the buffer solution are to conduct current and maintain the pH of the sample within permissible limits (Kohn, 1960). Since the majority of the serum proteins have their isoelectric points between pH 4 and pH 8 (Sober and Peterson, 1958), the electro- lyte solution is usually buffered between pH 8. 2 and pH 8. 6 (Crowle, 1961). Above pH 8. 0 most, if not all, serum proteins behaved as anions in an electrical field (Guyton, 1961; Campbell, et al. , 1963). The ionicity of the electrolyte solution must be sufficient to provide adequate buffering capacity but low enough so that the current does not exceed permissible limits (Valmet and Svensson, 1954). Buffer solu- tions with ionicities between 0. 075 and 0.1 were usually satisfactory (Laurell and Laurell, 1955; COOper, 1960). The constituents of the buffer solution also influence the separa- tion of the sample during electrophoresis. Human serums were separated into 12 components when electrophoresed in cellulose acetate with a buffer solution containing tris (hydroxymethyl) aminomethane (tris), ethylenediaminetetracetic acid (EDTA) and boric acid (Aaronson and Gronwal , 1958). A tris-—EDTA-boric acid buffer system, disconc- tinuous with respect to pH, was described by Goldberg (1959). Excellent separation of the serum proteins from a number of animal species was obtained by using the continuous barbital-acetate buffer system described by Owen (1956). The addition of either calcium or magnesium salts to barbital buffers facilitated the separation of the beta globulins (Laurell, et a1. , 1956). The salts formed complexes with the beta, lipoproteins and altered their mobility in an electric field (Laurell, 1960). Because of negatively charged carboxyl groups on filter paper, agar, and many other supporting materials, water tends to move toward the cathode during electrophoresis in alkaline solutions (Kunkle, 1959; Cooper 1960; Kohn, 1960). This phenomenon, known as 17 electroosmosis, affects the mobility of all of the migrants, depending on the character of the migrant and the nature of the supporting material (Kunkle, 1954; Crowle, 1961). For this reason, electrophoretic data obtained on different supporting media usually cannot be compared. Filter paper and cellulose acetate showed relatively little electroosmosis when compared to agar gel (Crowle, 1961). Cellulose acetate was introduced as a supporting material for zone electrophoresis by Kohn (1957). Electrophoresis in this medium is similar in principle and application to that in filter paper but has some notable advantages (Kohn, 1957; Kohn, 1960; Colab, 1960). The effects of buffer composition, ionicity, and pH, electrolyte volume, strip length, sample position, voltage, and migration time on the sample separation obtained in this medium were investigated by Bracken- ridge (1960). Several microelectrophoretic techniques employing cellulose acetate membranes have been described (Kohn, 1958, Grun- baum, et a1. , 1960). Cellulose acetate has been also used as a supporting material for immunodiffusion and IE (Kohn, 1957; Kohn, 1960; Scheer, 1961). Zone electrophoresis can also be done in agar gels (Kunkle, 1954; Uriel, 1958; C00per, 1960). Techniques and applications of agar gel electrophoresis were discussed by Wieme (1959) and by Weiner and Zak (1963). Electrophoresis is usually on gel-covered glass microscope slides but can be easily modified for preparative purposes. The chief advantages of agar gel over paper are the high water content of approximately 95% (Grabar, 1959), and translucency (COOper, 1960) of agar. ~ Its chief disadvantage is that considerable electroosmosis occurs during electrophoresis (Kunkle, 1954; Kreutzer and Fennis, 1964). Human serums were separated into eleven fractions in agar gel (Strickland, et al. , 1959). 18 Insoluble potato starch was an effective supporting material for preparative zone electrophoresis (Kunkle, 1954; Kunkle and Slater, 1952; Blomendal, 1959; Bodman, 1960). The resolution obtained was similar to that obtained in filter paper (Cooper, 1960). Its chief advantages are good recovery of fractions and tolerable electroosmotic flow during electrophoresis (Kunkle, 1954; C00per, 1960). Insoluble starch was ideally suited as a supporting material for the preparation of larger lipoproteins and mac roglobulins that did not penetrate starch gels (Bodman, 1960) and were adsorbed on filter paper (Kunkle, 1954). Zone electrophoresis has been used extensively in the study of serum proteins from normal and diseased individuals. Many of the studies have been discussed in reviews by Luetscher (1947), Lever (1951); Sterling (1951), Sobotka (1955), Jenks, et a1. (1956), Reiner (1957), Graham, et al. (1958), Owen (1958), Carson and Mattingley (1960), Lewis (1960), Petermann (1960), and Belfrage (1963). Immunoelectrophoresis Poulik (1952) applied the double diffusion technique to substances separated by electrophoresis in paper. Immunoelectrophoresis in agar gel was introduced by Grabar and Williams (1953). Immunoelectro- phoresis consists of two steps. The components of a complex mixture of antigens are first separated by electrophoresis into a series of spots or diffusion centers in the agar gel covering a glass plate. A long rectangular trough is then cut in the gel parallel to the axis of migration and antibodies specific for the antigenic constituents of the mixture are added to precipitate the antigens that immediately begin to diffuse radially from their respective centers. When precipitation is complete, the plate is usually washed, dried, and stained. The macroimmunoelectrophoresis technique of Grabar and Williams (1953) consisted of covering glass plates with a 1% to l. 2% l9 agar solution prepared in a barbital buffer solution, pH 8. 2, ionicity O. 025. After it had solidified, a well was cut in the agar and filled with from 0. 02 to 0.1 ml. of the sample. Electrophoresis was carried out under a potential of four to five volts per centimeter for five hours at room temperature. Following electrophoresis, a trough was cut in the agar from six to ten millimeters from the antigen well, the anti- serum added, and the plates incubated in a moist chamber at 180C for three to four days. Immunoelectrophoresis of normal human serum resolved 25 antigens (Grabar and Williams, 1953; Williams and Grabar, 1955a; Williams and Grabar, 1955b; Williams and Grabar, 1955c; Hirschfeld, 1963). The chief disadvantages of the techniques were the relatively large amounts of biological reagents required, the long period of electrophoresis and immunodiffusion, and the difficulty in washing and staining the completed plates (Grabar, 1959; Williams, 1960). To circumvent many of these difficulties Scheideggar (1955) devised a microimmunoelectrophoresis technique which used standard 1x3" glass microscope slides. ElectrOphoresis was carried out at six volts per cm. for 45 minutes on slides covered with two percent agar dissolved in the same buffer solution employed in the “macro" technique. It required approximately 0. 5% as much antigen and 1% as much anti- serum as required in the "macro" technique (Crowle, 1961). Immuno- diffusion was usually complete in 24 to 36‘ hours and because of the smaller size of the slides and thinner agar gels, they were more readily washed, dryed, and stained (Grabar, 1959). The "micro" technique was ideal for clinical studies where reagents were frequently limited (Clausen, 1963). As many as 18 tests could be performed simultaneous- ly (LKB Insts. Inc. ). Other supporting materials that have been used for IE include paper (Poulik, 1952; Gendon, 1958), starch gel (Poulik, 1959;. Allison, 1959; 20 Korngold, 1963), cellulose acetate (Kohn, 1957, Kohn, 1958, Kohn 1961), and cross linked polyacrylamide gels (Huneeus-Cox, 1964). Agar was found to be the most useful because it is negatively charged, has a high gel strength, is nearly transparent, is soluble in an aqueous medium, there are few ionized groups, and there is little nonspecific reaction with proteins (Wieme, 1959; Crowle, 1961). Starch gels were opaque and starch had to be used in relatively high concentrations which increased the possibility of nonspecific interactions with proteins (Poulik, 1959; Korngold, 1963). Dust, salts, low molecular weight substances and nitrogen- containing substances had to be removed from agar prior to its use in IE'(Hirschfeld, 1960a; Crowle, 1961; Kreutzer and Fennis, 1964). All of the tests in an experiment had to be done with agar prepared in a single large batch to insure reproducibility of results (Crowle, 1961). The choice of gel solvents and buffer composition including pH, ionic strength and electrolyte and nonelectrolyte solutes, is complicated by the fact that electrophoresis and immunodiffusion must be carried out in the same environment. Therefore the chosen conditions of the medium must compromise the optimum requirements for both opera- tions. Since proteins are macromolecular polyions, the pH of the medium in which they are dissolved determines in part their net charge and therefore their rate and direction of mobility in an electric field. All of the human serum proteins were at their isoelectric point between pH 4 and 8 (Sober and Peterson, 1958) and therefore behaved as anions in an electric field at pH 8.6 (Guyton, 1961). At pH values above 8. 2 precipitation with specific antibodies was inhibited and preformed antigen-antibody complexes were dissociated (Crowle, 1961). The dis—- continuous buffer system described by Hirschfeld (1961a), pH 8.6, gave excellent resolution of human, guinea pig, bovine, porcine and avian serums (personal observations). 21 The ionic strength of the medium exerts an important effect on the mobility of proteins in solution by altering their polyionic character and thereby influencing their reactions with other proteins and with the supporting medium. The buffering capacity and conductile properties of the solution is also determined by the ionicity. The maximum permissible ionicity is determined by the extent of heating that can be tolerated. Because of the heating effect of current, the ionicity, temperature, voltage and pH must be carefully regulated, particularly when gellified media are being used. Desiccation of the gel structure alters the diffusion rate of macromolecules. When the ionicity is low, the voltage can be increased without overheating because the conductivity and intensity of electrical current is low (Hess, 1951). Under these conditions either the migration time can be reduced or the path of migration lengthened. The ionicity must be sufficient to provide adequate buffering capacity against pH changes that can result from electrolysis of buffer salts during electrophoresis (Crowle, 1961). The lower limit of the permissible ionicity is also determined by the electrolyte requirements for optimum precipitation (Boyd, 1956; Raffel, 1961; Kabat and Mayer, 1961). Adequate electrolytes may be provided by the antiserum (Crowle, 1961). A number of different buffer electrolytes have been used for IE (Crowle, 1961). The results obtained with one system were not always comparable with those obtained with another. For example, the addition of calcium or magnesium to barbital buffer systems facilitated the separation of the beta globulins (Laurell, et a1. , 1956). Most immuno- electrophoretic techniques have employed barbital buffers at pH 8. 2 to 8.6 and ionicities from 0. 025 to 0.1 (Crowle, 1961). Barbital precipi- tated some non-antibody components of human serum (Korngold and VanLeuwen, 1957), probably the alpha; lipoprotein (Crowle, 1961). 22 The separation of human serum alpha globulins was facilitated in barbital buffers (Gitlin, et a1. , 1956). The chief limitation to the potential resolving power of IE is the antiserum. This is caused primarily by two factors, differences in antigenicity of the various constituents in the sample being analyzed and unpredictable variations in the antibody response among animals of the same species (Boyd, 1956; Grabar, 1958; White 1963). Blood serum contains a large number and variety of antigens (Putnam, 1960). Ideally when a heterologous serum is injected into an animal, antibodies are elicited with specificities directed against each and every antigen in the serum. Unfortunately this is seldom realized because of the numerous conditions affecting antibody production (Boyd, 1956; Raffel, 1961; White, 1963). Moreover, considerable microheterogeneity was found to exist among the serum antigens from individuals of the same species (Goodman, et a1. , 1958; Dray, 1958; Dray, 1959; Dubiski, et al., 1959; Dubiski and Kelus, 1960; Benacerraf and Gell, 1961). The conditions and requirements for antigenicity or lack of it (gelatin and hemoglobin for example) are poorly understood. It must be assumed therefore that serum is composed of constituents of varying antigenicity. The other major difficulty arises from the variability in the im» mune response seen among individuals of the same species. A "good“ antigen in one animal can be a very "poor“ antigen in its litter mate, eliciting few or no detectable antibodies. Two out of a total of 20 rabbits that were inoculated with human serum albumin failed to produce detectable specific antibodies although identical immunization procedures were employed (Grabar, 1958). Moreover, because of the wide variations in relative amounts of the various antigens in serum, the amount of serum necessary to elicit antibodies to trace components may be sufficient to cause immunologic paralysis to other components present in greater concentrations. 23 Animal species vary with respect to the quantity and quality of the precipitating antibodies which they can produce against a given antigen (Boyd, 1956; Grabar, 1959; Crowle, 1961). Two types of pre- cipitins have been described: the R-type and the H-type. The R-type precipitin formed precipitates that were poorly soluble in excess antigen and virtually insoluble in excess antibody (Kabat and Mayer, 1961). They were usually quite resistant to temperature variations during formation. Precipitates formed by the H-type antibody were soluble in an excess of either reagent and quite sensitive to tempera- ture variations during formation (Crowle, 1961). H-type antibodies are usually preferred for IE because of the fine discrete precipitates which they form. Antiserums are composed of heterogeneous populations of anti- body molecules which differ in their physicochemical and biological properties. Brown and Graves (1959) reported significant differences in the properties of antibodies collected at different times from animals convalescing from foot-and-mouth disease. Maurer (1954) found that no more than seven percent of the antibodies produced by rabbits after one injection of bovine serum albumin (BSA) cross reacted with human serum albmnin. After repeated injections of BSA, approximately 15 percent cross reacted (Melcher, et al. , 1952). The most specific and discriminatory antibodies were produced by animals which were closely related phylogenetically to the species from which the antigens were obtained (Grabar, 1959; Crowle, 1961). Hyperimmunization favored the production of antibodies which reacted with an increasing number of closely related antigenic determinants (Dixon and Maurer, 1955; Grabar, 1958). More specific antiserums were generally elicited by a short immunization procedure (Raffel, 1961). The route of injection and the physical state of the antigen also influenced the magnitude and character of the antibody response (Leskowitz and Waksman, 1960; White, 1963). One of the better 24 immunization procedures for protein antigens was their repeated intra- muscular injection after alum precipitation (Proom, 1943). The mechanism of the adjuvant action of this preparation was investigated by White, et a1. (1955). Hawkins (1963) described the disposition of alum precipitated protein antigens after intramuscular inoculation into rabbits. Following electrophoresis, the individual fractions are distributed in asymmetric spots along the length of the gel. Their relative positions depends on their respective electrophoretic properties and the extent of electroosmosis in the medium. Since the extent of electroosmosis in agar gels was found to depend on its purity (Crowle, 1961), it is important to use the same agar preparation for all tests in an experiment. IInmediately following electrophoresis, the individual fractions diffuse radially from their diffusion centers and form fronts which correspond to the curvatures of their respective spots. Meanwhile the antiserum diffuses in a straight front toward the antigens. The diffusion rate of macromolecules through agar gels was found to depend on their respective diffusion gradients, their size, shape, molecular volume, charge, and charge distribution, and the gel porosity and temperature (Wunderly, 1960). Protein-agar interactions are minimal since in an alkaline solution both are negatively charged. Visible precipitation occurs when homologous antigens and antibodies meet depending on their respective concentration ratios and the physical conditions of the medium such as the pH and ionicity. The kinetics of antigen-antibody reactions in agar gels have been reviewed by Ouchterlony (1958), and Crowle (1961). The precipitates usually appear as curved lines with their convex side toward the antibody trough. The position of the precipitates be- tween their respective reactant diffusion centers was found to depend on the diffusibility of the reactants and their concentration ratios 25 (Hirschfeld, 1960a). Generally, the higher the molecular weight of the antigen, the shorter the distance of its precipitate from its dif- fusion center (Crowle, 1961; Hirschfeld, 1962). The shape of the precipitates was also found to depend on the shape of their respective antigen diffusion fronts, the diffusibility of the antigens, and the relative proportions of the reactants (Grabar, 1959; Hirschfeld, 1960a; Hirschfeld, 1960b; Crowle, 1961). Multilined patterns are possible because of the selective permeability of immunOprecipitates to sero- logically unrelated reactants (Hirschfeld, 1963). The apex of precipitates was found to correspond to the average electrophoretic mobility of the serologically identical molecules com- posing the precipitate (Hirschfeld, 1960b) and to the place where the antigen was present in greatest concentration (Grabar, 1959; Williams, 1960). An estimate of an antigens mobility can be obtained therefore by erecting a perpendicular from the apex of its precipitate to the anti- body trough. The minimum number of antigenic constituents in a complex mixture was indicated by the number of precipitin lines produced after IE (Grabar, 1959). The possibility of superimposition of precipitates after IE is considerably less than after Ouchterlony double diffusion since in IE, the antigens are first separated by electrophoresis. Two or more antigens can form superimposed precipitates only if their electrophoretic mobility as well as their diffusibility and reactant ratios are nearly identical. The probability of the simultaneous occurrence of all these factors is very low (Grabar, 1959; Hirschfeld, 1962). Electrophoretically identical antigens can be resolved if their precipitates occupy different positions in the agar. Maximum resolution of closely adjacent precipi- tates was obtained with maximum migration path length and arc curva- ture and minimum precipitate thickness (Crowle, 1961). 26 Antigens could be further identified and characterized after IE by the use of specific antibodies, special staining technique, enzymatic reactions, or radioimmunoelectrophoresis (Grabar, 1959; Rejnek and Bednarik, 1960; Crowle, 1960). Details of washing, drying, and stain- ing completed slides have been discussed by Crowle (1961). The reliability of IE was investigated by Hirschfeld (1963). Ten normal human serums were tested with each of four antiserums under nearly identical conditions. Twenty-six serum antigens were found. The results of these analyses are summarized in Table 1. It is clear that the reliability of the various antigens varied from very high (proteins 1, 3, 21, and 24) to very low (proteins 19, 22, 23, and 25). Additional studies of individual serums with the same antiserum showed that reproducibility also varied among antigens and with the source of the serum'and the antiserum. The necessity of using several different antiserums for the analyses of complex mixtures of antigens was emphasized. A highly reliable precipitating antigen—antibody system (immuno— system) can be expected only when the amounts of both reactants are considerably above that required for visible precipitation. Reliability also depends on the reactant ratio. "Balanced” systems are those in which the reactant ratio is in or near the equivalence zone. The relia- bility of visible precipitate formation was usually good for strong- "balanced" systems but was poor for weak-”unbalanced" systems, weak-"balanced" systems, or strong-="unbalanced" systems (Hirsch- feld, 196 3). The chemical stability of the reactants as well as the reproduci- bility of identical gel conditions from slide to slide can also affect reliability. It is very difficult if not impossible to prepare identical gels on all of the slides in any experiment since the composition of the 27 undo...“ one? moomflcm £3on 3:339: 23 £033 5 mgduom ammo mo Meagan may “Gomonmou 63m.”— ofi mo room, on“ 5 Manon—8.9a 9:. ooSooSooooooonomooooooooonoon o soonooonoommowmeasemeoonsmionon o moonoooseommmesonnommmomooggon m SmonMSSQOONseemoemmeSNSSSS a. emmmemmmmmfiomowSeemHESNHHHoowsomemN28:33:... HOQCHDZ GMGHOHQ .32.: 28.2085 Eons : 583022.... unouomfip H33 5%? mgduom Gags: ngnoc Goo mo $09:me oflouonaofioofimocapacflw mo maddmom: .H 3nt 28 buffer solution may change upon storage. The gel strength and pore size are influenced by the extent and duration of heating during prepara— tion. Therefore, irreproducibility of the test can be in part due to variations in either the sample or the gel. MATERIALS AND METHODS Collection of Serums Blood was obtained from rabbits and guinea pigs by cardio- centesis with a 2 1/2 inch 22 gauge needle on a 12 m1 syringe. It was allowed to clot in slanted tubes for two hours and overnight at 4 C. The serums were decanted, centrifuged at approximately 4140 x g for five minutes, and transferred to sterile 17 x 60 mm screw capped vials. Sodium ethylmercurithiosalicylate (merthiolate, E. Lilly Co.) was added to a final concentration of 0. 01% and the serums frozen immediately at —70 C. Serums were collected from all experimental animals prior to inoculation. Fresh normal serum was separated into four equal fractions and treated as follows: fraction one was treated with merthiolate as described and stored at 4 C; fraction two was treated with merthiolate and frozen at -70 C for two weeks; fraction three was treated with merthiolate and filtered through a Seitz pad supported on a Swinney filter; fraction four was treated with merthiolate, frozen at -70 C for two weeks and filtered. All four fractions were compared by immuno- electrophoresis (IE). Rabbit Antiserums Specific for Normal Guinea Pig Serum Portions of the serums from each of the guinea pigs to be in— fected with Mycobacterium bovis were pooled. The mixtures were precipitated with alum as described by Proom (1943). The inocula were prepared by mixing 12. 5 ml of the pooled normal serum, 40. 0 ml 29 30 of distilled water, and 45. 0 ml of 10% aluminum potassium sulfate. The pH of the mixture was adjusted to 6. 5 with 5 N sodium hydroxide and the mixture centrifuged at 4140 x g for five minutes. The white sediment was washed twice with isotonic saline solution containing 0. 01% merthiolate and resuspended to 50 ml in the washing solution. It was stored at 4 C for no longer than 14 days. Five adult Dutch Belted rabbits were injected with the alum pre- cipitated guinea pig serum and bled according to the following schedule: Day Operation Amount Route 1 Bled 10. 0 ml Cardiocentesis l Injected 6 . 0 m1 Intramuscular 14 Injected 6. 0 ml Intramuscular 24 Injected 1. 0 ml Intraperitoneal 34 Bled 5 . 0 ml Cardiocentesis 48 Injected 6. 0 ml Intramuscular 62 Injected 6 . 0 ml Intramuscular 72 Injected 1. 0 m1 Intraperitoneal 77 Bled 35. 0 ml Cardiocentesis The antiserums were harvested and stored as described. Each anti- serum was tested individually by lE with normal guinea pig serum. Equal portions of each of the antiserums were pooled. The antiserum produced by rabbit number six (R-6) was examined by Ouchterlony immunodiffusion and IE . Mycobacterium bovi s Culture 81 C-0 was isolated from tuberculous swine and identified as M. 1322.3. by morphological, cultural, and biochemical tests and infectivity for guinea pigs, rabbits, and calves. A three week old culture of organisms grown in Dubos (Difco) 0. 5% dextrose broth at 37 C was used to infect guinea pigs. A heat killed suspension, 100 C for 30 minutes, was used to sensitize guinea pigs. 31 Guinea Pigs Infected with M. bovis Thirty-six white male guinea pigs approximately six months old and 300 gms in weight were divided into nine groups of four per group. They were maintained four per cage, fed Rockland guinea pig diet, supplemented occasionally with carrots, and given water ad libitum. Each guinea pig was inoculated intraperitoneally with 0. 01 mg wet weight of M. bovis. Different groups of four guinea pigs each were bled by cardiocentesis at 7, 14. 21, 28, 33 and 41 days after inocu- lation. The other guinea pigs died during the experiment. Blood was collected and handled aseptically. Serums were fil- tered through a Seitz pad supported on a Swinney filter. They were dispensed into sterile 17 x 60 mm screw capped vials, merthiolate added to a final concentration of 0. 01%, and stored at -70 C. Cellulose acetate electrophoresis and IE were performed with serums which had been frozen and thawed only once. Lung Fluid. Lung fluid was obtained aseptically from the thoracic cavity by thoracocentesis. Formed elements and particulate matter were re- moved by centrifugation at approximately 4140 x g for five minutes. The clear amber fluid was decanted and filtered through a Seitz pad supported on a Swinney filter. It was dispensed into 17 x 60 mm screw capped vials, merthiolate added to a final concentration of 0. 01%, and stored at -70 C. Guinea Pigs Sensitized with Heat Killed M. bovis Fourteen white adult male guinea pigs approximately six months old and 300 grams in weight were divided into four groups of four each. Each guinea pig was given three intraperitoneal inoculations of 32 approximately 1. 0 mg per inoculation of heat killed M. b_c_)_\_r_i_s at three day intervals. Fifteen days after the first inoculation, two guinea pigs were tuberculin tested with 0. 1 ml of mammalian tuberculin and examined at 24 and 48 hours. Different groups of four guinea pigs per group, excluding those tuberculin tested,were bled at 15, 22, and 29 days after the first inoculation. Serums were harvested in the manner described and stored at -70 C. The serums collected prior to and after sensitization were analyzed by cellulose acetate electrophoresis and IE. Gel Filtration Gel filtration of normal serum was done with Sephadex G-100 and G-200 (Pharmacia). Dry Sephadex was mixed in a beaker with phosphate buffered saline solution, pH 7. 2, ionicity 0. 25, and allowed to stand undisturbed for 24 hours at room temperature. Excess ”fines" were removed by repeatedly decanting the supernatant fluid after allow- ing the swollen gel to settle for one hour in the buffer solution. Chromatography tubes were drawn from 1. 8 cm inside diameter glass tubing. They were plugged at the narrow end with glass wool on which was layered two cm of four mm diameter glass beads. A slurry of the swollen Sephadex was poured into the columns and allowed to settle at room temperature. Additional Sephadex was added to adjust the level slightly above the desired column height. The tubing clamp at the bottom of the column was opened and the gel washed overnight with buffer solu- tion supplied from an aspirator bottle. A filter paper pad was fitted to the top of the gel layer to prevent agitation during sample application and subsequent operations. From four to five ml of fresh normal serum were applied to the column and allowed to sink into the gel. The filter pad and sides of 33 the glass tube were washed with several three ml portions of the buffer solution and a layer of buffer five cm deep was placed above the filter paper pad before continuous flow (30 to 50 ml per hour) was started. All operations were performed at room temperature. Five ml fractions were collected in individual glass tubes contained in an auto- matic fraction collector (Research Specialties Co. , Model D-3) fitted with a volumetric siphon. The protein content of each fraction was determined by the Lowry modification of the Folin phenol method (Kabat and Meyer, 1961). Designated fractions were concentrated tenfold by pervaporation and analyzed by cellulose acetate electrophoresis and Ouchte rlony immunodiffusion . Ion- exchange Column Chromatography Dry DEAE-cellulose (Schleicher and Scheull Co.) type 403 was seived for several hours through U. S. standard seive series numbers 100 and 325 on a motor-driven shaker. The 100 to 325 fraction was separated and used for chromatography. The dry absorbent was allowed to sink into a 1N solution of sodium hydroxide and thoroughly mixed. The mixture was transferred to a 4 3/4 inch Buchner funnel fitted with filter paper (Whatman No. l) and washed repeatedly with 1N sodium hydroxide until all of the color was removed. The filter cake was resuspended in 1N sodium hydroxide and 1N hydrochloric acid added gradually with continuous stirring until the mixture was strongly acidic. The adsorbent was collected im- mediately by filtration on a Buchner funnel and washed with distilled water. It was resuspended again in IN sodium hydroxide, filtered, and washed with water. It was resuspended in three volumes of the initial buffer solution; 0. 005 N phosphate pH 8.6, 0. 04M in Tris. 34 The pH of the adsorbent suspension was adjusted to 8.6 with 0. 1M phosphoric acid. The adjusted adsorbent was washed on a filter with 500 ml of the initial buffer solution and suspended in 60 ml of the initial buffer solution per gm of dry adsorbent. Excess "fines" were removed by allowing the adsorbent to settle for about one hour and decanting the supernatant fluid. This was repeated until the super- natant fluid was clear. Chromatography tubes were drawn from 1. 8 cm inside diameter glass tubing and plugged at the narrow end with glass wool on which was layered two cm of four mm diameter glass beads. The tubes were filled with the initial buffer solution and the dilute adsorbent suspension slowly added from an aspirator bottle. The columns were allowed to pack by gravity to approximately 20 cm above the desired column height. Columns which had a pleated appearance were repacked. The adsorbent was packed to a constant height under ten lbs of nitrogen pressure, washed with at least 500 m1 of the initial buffer solution, and fitted with a filter paper pad to prevent agitation during sample application and elution. Normal serum samples were prepared for chromatography by dialysis for 24 hours at 4 C against at least three changes of the initial buffer solution. A slight precipitate appeared and was removed after centrifugation at 32, 000 x g for one hour. Chromatography was done at room temperature. Eight to ten ml of serum were applied to the column and allowed to sink into the adsorbent. The filter paper pad and sides of the tube were washed with several three ml portions of the initial buffer solution and a layer of buffer five cm deep was placed above the filter paper pad before continuous flow was started. Proteins were eluted from the adsorbent by continuous gradient elution (Peterson, et a1. , 1961). A concave salt gradient was produced 35 by a cone-sphere buffer vessel device. This consisted of a 250 ml Erlenmeyer flask which contained 250 ml of the limit buffer solution (0. 35 M tris phosphate) and a 500 ml flat bottomed Florence flask which contained 500 ml of the initial buffer solution. The contents of the Florence flask were stirred continually by a magnetic stirrer. The buffer solution reservoirs were in hydrostatic equilibrium. Their respective levels above the column were adjusted to provide a flow rate of approximately 30 ml per hour at the beginning of the experiment. Five ml fractions were collected in individual glass tubes con- tained in the automatic fraction collector. The relative protein content of each fraction was determined by absorption measurements at 280 mp. in a Beckman DU spectrophotometer. Designated fractions were concentrated tenfold by pervaporation and analyzed by cellulose acetate electrophoresis, Ouchterlony immunodiffusion, and IE. Starch Block Electrophoresis Starch preparation and electrophoresis were done according to modifications of the procedures described by Kunkle (1954) and Campbell,et al. (196 3). One pound of insoluble potato starch (Mallin- ckrodt Chemical Works) was suspended in one liter of 0. 005 N sodium hydroxide and allowed to stand overnight. The alkali was removed by frequent changes of distilled water over a 12 hour period and the starch washed twice with the internal buffer solution (Hirschfeld, 1960a) during a 12 hour period. The washed starch was resuspended in a minimal volume of the internal buffer solution and poured into a 38 x 7. 7 x 0. 7 cm plexiglass template. Excess moisture was blotted off the starch slurry with absorbent paper and the surface leveled by scraping with a ruler. 36 A sample well, 5. 0 x 0. 5 cm was cut in the starch block 12. 5 cm from the cathode end and perpendicular to the direction of migration. The migration chamber was similar to model 1400 sold by Research Specialties Co. Current was supplied by a variable voltage DC power supply (Heathkit, model PS—3). The starch-filled template was placed in the migration chamber and cooled to'4 C prior to electro- phoresis. Electrical connections between the starch and the external buffer solution (0. 2 M phosphate, pH 7. 5) were made with buffer im- pregnated filter paper wicks (Whatman No. 1). Three ml of fresh normal serum, previously dialyzed for 24 hours at 4 C against three changes of the internal buffer solution, were applied to the sample well. The well was immediately filled with starch and the entire starch surface covered with Saran wrap (Dow Chemical). An electrical potential of 250 volts was applied across the length of the starch block for 10 hours at 4 C. When electrophoresis was completed, one cm wide segments were cut from the starch block perpendicular to the direction of migration. The starch segments were placed in tubes and the protein removed by replacement filtration in five ml of cold phosphate buffered saline solu- tion, pH 7. 2. The tubes were agitated, the starch allowed to settle, and the supernatant fluid decanted. The protein content of the individual fractions was determined by the Lowry modification of the Folin phenol method (Kabat and Meyer, 1961). Designated fractions were concen- trated tenfold by prevaporation and analyzed by cellulose acetate electro— phoresis and Ouchterlony immunodiffusion. Agar Gel ElectrOJhoresis Glass microscope slides, 1 x 3 inch, were covered with 2. 5 ml of a melted one percent agar solution prepared as described under IE. 37 After the gels had "aged“ for at least three hours, circular origins seven mm in diameter were cut in their centers. The agar plugs were removed by aspiration with a Pasteur pipette and approximately five .11 of serum placed in the wells. Electrophoresis was carried out for three hours as described under IE. After electrophoresis, the slides were immersed in a five percent solution of glacial acetic acid for 30 minutes, dried under filter paper at 37 C, and stained. A barbital-acetate buffer system (Owen, 1956) was also used as a gel solvent. The buffer solution was prepared as described under cellulose acetate electrophoresis and adjusted to pH 8. 2 with 0.1 N hydrochloric acid. It was diluted 1:1 with an equal volume of melted two percent agar. The gel covered slides were prepared as described under IE. A 1 x 11 mm slot-shaped origin was cut in the gels 31 mm from the cathode end of the slides and the agar plugs removed by aspiration with a Pasteur pipette. Approximately eight [.11 of ‘serum were placed in each well and electrophoresed for 2 1/2 hours. The proteins were fixed and stained. Cellulose Acetate Membrane Electrophoresis Electrophoresis was done in a Shandon migration chamber with a Vokam constant current DC power supply (Colab). Cellulose acetate strips 2. 5 x 12 cm were used as the supporting material. Four buffer systems were investigated, a Tris-EDTA-boric acid buffer (Aronsson and Gronwall, 1958), a Tris-EDTA-boric acid buffer system discon- tinuous with respect to pH (Goldberg, 1959), a barbital-barbituric acid buffer system (Laurell, et a1. , 1956), and a barbital-acetate buffer system (Owen, 1956). The latter buffer system was used throughout the investigation. The Owen buffer solution pH 8. 6, ionicity 0. 07, c ontained the following: 38 Sodium diethylbarbiturate 5. 00 gm Sodium acetate (anhydrous) 3.25 gm Hydrochloric acid (0.1 N) 34.2 ml Calcium lactate 0. 38 gm Distilled water, 1. _§_. ad. 1000.0 ml Two and one-half to five ul of serum were applied directly over the cathode on each buffer impregnated strip. Electrophoresis was done with a current of 1 mA per strip for two hours at 4 C. Following electro— phoresis, the strips were stained, dried under weighted absorbent paper, and examined in a Joyce double beam recording and integrating densi- tometer. Because of the relative insensitivity of the technique and the densitometer, no attempt was made to measure the areas under the respective protein peaks . Protein, Lipoprotein, and Glycoprotein Stains Two protein stains were compared for use on dried agar following electrophoresis in agar gel. Ponceau S (0. 2% in 3. 0% trichloroacetic acid) was prepared and the slides immersed for five minutes. The slides were differentiated (de-stained) in two percent glacial acetic acid. The other stain was a protein triple stain (Crowle, 1961) which contained the following: Thiazine red R 0.1 gm Amidoswarz 10B 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 in this solution for five minutes, differentiated in two percent acetic acid, and dried. The latter stain was used for staining slides after IE. Following electrophoresis in agar gel, lipoproteins and glyco— proteins were stained according to the methods described by Grabar (1959). Lipoproteins were stained by immersing the dried agar coated 39 slides in a saturated solution of Oil Red 0 (37 C) in 60 percent ethanol. Slides were differentiated in 50 percent ethanol. Glycoproteins were stained by immersing the dried agar coated slides for 15 minutes in 100 ml of a one percent periodic acid solution in 50 percent ethanol containing 1.64 gm of anhydrous sodium acetate. The slides were washed for 15 minutes in distilled water and immersed for five minutes in 110 ml of the following solution: 50 ml of a 0. 01M aqueous solution of a-napthol plus 50 ml of a 0. 01 M solution of p-phenylenediamine plus 10 ml of a 10 percent solution of hydrogen peroxide. They were thoroughly washed in running tap water, then distilled water, and dried at 37 C. Ponceau S, Nigrosin, Light green SF, Coomassie brilliant blue, and Lissamine green were investigated for protein staining on cellulose acetate membranes. Ponceau S was used routinely for serum protein staining. Cellulose acetate strips were immersed for 10 minutes in a 0. 2 percent solution of Ponceau S in three percent trichloroacetic acid and differentiated in two percent glacial acetic acid. Nigrosin (0. 001% in 2% glacial acetic acid) was used to stain strips with low protein concentration (Colab, 1960; Kohn, 1960). Dried agar coated slides on which Ouchterlony immunnodiffusion had been performed were stained with the protein triple stain (Crowle, 1961). Ouchte rlony Immunodiffusion Ouchterlony gel diffusion plates were prepared according to the methods described by Kohn (1960) and Dardas (1962). Glass lantern slides, 3 1/4 x 4 inch, were thoroughly washed with a detergent (Tide), rinsed with distilled water, and air dried. Each slide was evenly covered with 10 ml of a phosphate buffered (pH7. 2, ionicity 0.6) one percent agar solution to which merthiolate was added to a final 40 concentration of 0. 01% (Murty, 1960). After a minimum of three hours incubation in a humidified diffusion chamber, reactant wells were cut in the gel according to a drafted pattern placed beneath the plate. The agar plugs were removed by aspiration with a Pasteur pipette and the wells filled with the reactants. The plates were incubated for 36 to 48 hours at 28 C in a humidified diffusion chamber. Non-reacted protein was removed by soaking the gel covered plates in six changes of phos- phate buffered saline solution (pH 7.4) over a period of 48 hours. Inorganic salts were leached from the gels by soaking in three changes of distilled water over a period of 24 hours. The plates were overlaid with moist filter paper, dried at 37 C, and stained. Immunoelectrophoresis Immunoelectrophoresis was done on agar-covered, l x 3 inch glass microscope slides as described by Hirschfeld (1960a). A two percent solution of Difco Bacto agar was poured into a flat dish to a depth of one cm and allowed to solidify. The gel was cut into one cm cubes and washed in running tap water for 48 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 for IE in this study was prepared in a single batch. A buffer system discontinuous with respect to ionicity was used (Hirschfeld, 190-a). The internal buffer solution (pH 8. 6, ionicity 0.093) refers to that in the agar gel; the external buffer solution (pH 8.6, ionicity 0. 06) was used in the electrode vessels. The buffer solu- tions contained the following: External buffer Internal buffer 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, 3.3.fl. 1000 ml 1000 ml 41 Two parts of the internal buffer solution were mixed with one part of distilled water and the mixture heated in flowing steam. An equal volume of freshly melted agar was added with merthiolate in a final concentration of 0. 01%. While the solution was warm, 2 1/2 ml were spread evenly over the surface of thoroughly washed and rinsed micro- scope slides and allowed to solidify. Each agar-covered slide was "aged" at least three hours in a humidified diffusion chamber prior to use. Immediately before the samples were applied, two circular antigen wells, two mm in diameter, were cut in the gel 39 mm from the anode end of the slide and five mm from each edge. The agar plugs were removed by aspiration with a blunt 12 gauge needle. A 1 x 66 mm antiserum trough was cut in the gel equidistant between the two antigen wells. The agar plug was left in place until after electrophoresis was completed. Approximately five ul of undiluted serum were placed in each antigen well with a 2 1/2 ml disposable glass syringe fitted with a blunt 5/8 inch 25 gauge needle. A Shandon migration chamber was modified for IE by placing a 6 x 8 1/2 inch plexiglass plate across the bridge supports. Eight slides were used at a time; four slides were blanks and were placed on the anode side of the plate. Electrical connections between the electrode vessels were made with filter paper wicks (Whatman No. 1) impregnated with the external buffer solution. Current was supplied by a variable voltage DC power supply (Heathkit, model PS-3). Electrophoresis was carried out at 4 C for two hours at l. 25 mA per slide (eight volts per cm). Following electrophoresis, the agar plugs in the antiserum troughs were removed by aspiration with a Pasteur pipette and approxi- mately 0. 25 ml of the antiserum added. Slides were incubated for 24 to 28 hours at 28 C in a humidified diffusion chamber. 42 Non-reacted protein was removed by soaking the gels in at least six changes of phosphate buffered saline solution (pH 7.4) over a period of 24 to 36 hours. Inorganic salts were leached from the gels by soaking in several changes of distilled water during a 24 hour period. The slides were overlaid with filter paper, dried at 37 C, and stained. RESULTS Examination of Individual Rabbit Antiserums Similar immunoelectrophoretic patterns were produced with normal guinea pig serum and the antiserums from rabbits numbers one, two, four, and five. Antiserums collected 10 days after the last intrapertoneal inoculation of normal serum were unsuitable for use in IE (Figure 1). No detectable anti-albumin antibodies were found in the antiserum from rabbit number six (R-6) after two courses of immunization (Figure 2). The immunoelectrophoretic patterns of normal serum developed with the antiserums from R-6 and R-7 collected five days after the last intraperiteneal injection of guinea pig serum are shown in Figure 3. No anti-albumin antibodies were detected by IE in the serum of R-6 after four courses of immunization (Figure 4). A precipitate formed when the five rabbit antiserums were pooled. The results of interfacial precipitation tests with normal serums and post-injection rabbit antiserums are shown in Tables 2 and 3. Precipitation occurred only when antiserum from R-6 was one of the reactants. This was confirmed by Ouchterlony immunodiffusion (Figures 5 and 6). Two precipitation lines formed between the antiserums from R-1 and R-6 (Figure 5), R-4 and R-6 (Figure 6), and R—5 and R-6 (Figure 6). Only one line was observed between the antiserums from R-2 and R-6 (Figure 6). The major precipitation line of the two line systems formed a reaction of identity with the single line between the antiserums from R-2 and R-6. The smaller of the two precipitates that formed between 43 44 Figure 1. Immunoelectrophorogram of normal guinea pig serum developed with antiserum collected from rabbit number 7 ten days after the last injection of guinea pig serum. 45 Figure 2. Immunoelectrophorogram of normal guinea pig serum developed with antiserum from rabbit number 6. Figure 3. 46 Immunoelectrophorogram of normal guinea pig serum developed with antiserums eollected from rabbits number 6 and number 7 five days after the last injection of guinea pig serum. 47 Figure 4. Immunoelectrophorogram of normal guinea pig serum developed with antiserum from rabbit number 6 (doubled course of immunization). 48 Table 2. Results of reciprocal interfacial precipitation tests with individual rabbit antiserums. R-Z R-4 R-5 R-6 R-6* R-l - - _ + - R-2 - - + _ R-4 - + - R-5 + - R-6* - antiserum collected from rabbit number 6, 160 days after the last injection. Table 3. Results of reciprocal interfacial precipitation tests with individual normal rabbit serums. R-l - - - - R-Z - _ - R-4 - - R-5 - Figure 5. 49 Ouchterlony immunodiffusion of normal rabbit serums and antiserums from rabbits number 1 and number 6. Well A. Antiserum from rabbit number 6 Well B. Antiserum from rabbit number 1 Well C. Normal serum from rabbit number 6 Well D. Normal serum from rabbit number 1 Figure 6. 50 Ouchterlony immunodiffusion of normal rabbit serums and antiserums from rabbits numbers 2, 4, 5, and 6. Well A. Antiserum from rabbit number 2 Well B. Antiserum from rabbit number 2 Well C. Antiserum from rabbit number 6 collected 160 days after the last injection of guinea pig serum Well D. Antiserum from rabbit number 6 collected five days after the last injection of guinea pig serum Well E. Antiserum from rabbit number 4 Well F. Antiserum from rabbit number 5 Well G. Antiserum from rabbit number 5 51 the antiserums from R-1, R—4, and R-6 appeared on the R-6 side of the major precipitation line. It was on the R-5 side of the major precipitation line in the R-5, R—6 system. No visible precipitation occurred between normal serum from R-6 and antiserum from R-1 or between normal serum from R-1 and antiserum from R-6 (Figure 5). There was no visible precipitation between antiserum from R-6 and normal serums from any of the other rabbits (Figure 6). All of the precipitation lines curved away from the reactant well that contained antiserum from R-6. Serum collected from R-6, 160 days after its last injection with untreated normal guinea pig serum failed to precipitate with any of the other antiserums in interfacial precipitation tests (Table 3) and in Ouchterlony immunodiffusion tests (Figure 6). A single precipitate formed in the albumin region when antiserum from R-6 was developed with antiserum from R-1 during IE (Figure 7). No visible precipitation was observed when the serum collected from R-6, 160 days after its last injection with normal serum was tested simultaneously with the same developing antiserum. Gel Filtration of Normal Guinea PigSerums in Sephadex G-100 and G-200 The serum proteins were separated into two main fractions by gel filtration in Sephadex G-100; three fractions were separated in Sephadex G-200 (Figure 8). The first effluent fractions collected from both gels were quite opalescent. Cellulose acetate electrophoresis of the concentrated effluent fractions from Sephadex G-100 columns indicated that fraction six contained beta; globulins and components that migrated very little from the sample origin (Figure 9). Albumin, alpha;, betal, beta;-l, and gamma globulins were found in fraction eight. Fraction 10 contained 52 Albumin Figure 7 . Immunoelectrophorogram of antiserums collected from rabbit number 6, 5 and 160 days after the last injection of guinea pig serum. A is antiserum collected 5 days post-immunization. B is antiserum collected 160 days post-irnmunization. ._—o G-lOO _ ' H o-zoo 1 1 1 L l 1 1 1| 1 J l 1' 2 4. 6' 8 10» 12 14- 16 18 20 22‘ 24 26 Fraction Numbe r Figure 8. Serum protein distribution in gel filtration (Sephadex G-100 and G-200) effluent fractions. 54 'Y 2'2 1324 51 C1291-11 w wwv II b. Concentrated sample from fraction number 6. a. Normal serum. c. Concentrated sample from fraction number 8. 1|~ d. Concentrated sample from fraction number 10. NH 3. Concentrated sample‘from fraction ‘riuhdber 12. “ Figure 9. Cellulose acetate electr0phorogram of gel filtration (Sephadex G-100) effluent serum fractions. 55 albumin, alpha;, betal globulins. Both beta; globulins were found in fraction 12. Protein was not detected in any of the other fractions beyond fraction 12. Analysis of the concentrated effluent fractions from Sephadex G-100 columns by Ouchterlony immunodiffusion indicated that most of the serum antigens were found in more than one fraction (Figure 10). Satisfactory separation of the serum antigens was not obtained by this procedure. Ion—exchange Column Chromatography of Normal Guinea Pig Serums on DEAE-cellulose The serum proteins were separated into four major fractions and several minor fractions by column chromatography on DEAE- cellulose (Figure 11). Most of the proteins were eluted after 300 ml of the eluting solvent had passed through the column and the pH of the effluent was below 7. 0. The pH of the effluent decreased from 8. 6 to 5. 5 after 500 ml had been collected. Gamma globulins of slightly different electrOphoretic mobilities were found in fractions 14, 31, 41, and 48 (Figure 12). The mobility of the gamma globulins in fraction 31 was less than that of the gamma globulins in the other three fractions. Beta;-2 globulins were found in fractions 41, 48, and 65. Fraction 65 also contained beta;-l globulins. Beta; globulins were found in fractions 65, 80, and 90. Fractions 72, 80, and 90 contained alpha; globulins. The alpha; globulins were found in fractions 80 and 90. Albumin was found in fractions 72, 80, and 90. Ouchterlony immunodiffusion of the effluent fractions indicated that the minimum number of antigenic components in fractions 14, 31, 41, 48, 65, 72, 80, and 90 was 3, 4, 6, 3, 6, 7, 8, and4 respectively (Figure 13). Most of the fractions contained antigens in common with one or more other fractions. 56 Figure 10. Ouchterlony immunodiffusion of gel filtration (Sephadex G-100) effluent serum fractions. Well A. Concentrated sample from fraction number 6 Well B. Concentrated sample from fraction number 8 Well C. Concentrated sample from fraction number 10 Well D. Concentrated sample from fraction number 12 Well E. Concentrated sample from fraction number 14 Well F. Concentrated sample from fraction number 16 Open Wells. Pooled rabbit anti-guinea pig serum antiserum mo 00 .mcofiome “sesame AomoHSHHoUIM