ABSTRACT RELATIONSHIPS BETWEEN CHICKEN IMMUNE GLOBULINS AND A RHEUMATOID-LIKE-FACTOR PRESENT IN NORMAL CHICKEN SERUM by Richard B. Dardas The major objective of this study was to determine the effects of non—specific chicken serum co-factors on in vitro antibody-antigen reactions. A beta 2 macroeuglobulin was discovered in chicken serum which comprised 2.4% of the total serum protein. This component resembled human rheumatoid factor in its effect on aggregated avian gamma globulin. It resembled avian C'l with respect to many of its physical and chemical properties. It was absorbed with kaolin and aggregated chicken gamma globulin-and could be inhibited by 0.0025 M EDTA, 0.150 M CaCl2 or 0.1 M 2-mercaptoethanol. Since it was not possible to completely characterize this substance it is designated rheumatoid-like-factor (RLF). Attempts to isolate RLF in a pure and active form were without success. Some purification was achieved by gel filtration, but on repeating the filtration all activity was lost. Continuous flow paper curtain electrophoresis Richard B. Dardas inactivated the fraction. RLF was shown to be identical to the co-precipitating macroglobulin (Co P) reported by Mbkinodan et al. (1960). It was shown to have a titre increasing effect in PPLO and .g. pullorum tube agglutination tests if the immune serums were diluted in normal serum which had been diluted no more than 1:4 with saline. It was shown that ageing of the serum is necessary for RLF activity to become apparent and that cryoglobulin formation in chicken serum is not necessary for RLF activity. Cryoglobulin formation was shown to be inhibited by EDTA. Cryoglobulin precipitates were shown to be related to RLF precipitates by gel diffusion. Removal of RLF sufficed to protect chicken immune globulins from the effects of heating at 56 C for 30 min. Addition of fresh normal serum after heating gamma globulin absorbed serum caused a titre decrease. This was interpreted as meaning that immune globulins which were aggregated by heat were prevented from acting as participants in the PPLO agglutination reaction. Precipitin titre increase upon storage was attributed to the appearance of RLF which co—precipitated with the immune complex. Cold storage titre increase in agglutinin systems at high dilutions cannot be explained with the Richard B. Dardas RLF model since RLF is inoperative at dilutions of 1:8 or higher. An attempt at explaining differential heat sensitivity of agglutinin and HI antibodies was based on the finding of three distinct types of antibodies. These were obtained by gel filtration using Sephadex G 200 dextran polymer. High molecular weight HI antibodies were found to be different from high molecular weight agglutinins. Low molecular weight HI and agglutinating antibodies could not be separated from each other but were different from the high molecular weight types. No differences in aggregation ability of high and low molecular weight agglutinins were found. Both responded with a titre increase to excess amounts of RLF at high dilution of the antibody. No antigenic differences between the different groups and low molecular weight chicken anti-BSA precipitins were noted in gel diffusion studies. RELATIONSHIPS BETWEEN CHICKEN IMMUNE GLOBULINS AND A RHEUMATOID-LIKE-FACTOR PRESENT IN NORMAL CHICKEN SERUM BY Richard B? Dardas 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 1963 ACKNOWLEDGEMENTS Sincere thanks are extended to Dr. Delbert E. Schoenhard for his help, his encouragement and especially for his example. His gentle insistence and direction, being motivated by the concern that he feels for those who work with him were and continue to be most sincerely appreciated. The author wishes to thank each of the members of his guidance committee, Dr. Charles H. Cunningham. Dr. Harold L. Sadoff and Dr. Gordon L. Kilgour for their individual help and attention given him on special problems which arose in the course of the experiments reported herein. To Dr. Jadk J. Stockton goes the author's thanks for taking time from his very full schedule to assist in the preparation of this manuscript. The financial support given by Ciba Pharmaceutical Company. Summit. New Jersey, is also appreciated. A great many problems were solved with the unsolicited help of Mr. Irving L. Dahljelm. His assistance was always most timely and most valuable. To my family go a very special thanks, for it is the family most of all who accept the debt of advanced graduate studies. Their patience and cooperation stands as a continuing source of inspiration. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . . PPLO Antigen and Salmonella pullorum Antigen Serum Collection Antiserum Production Preparation of Avian Gamma Globulin Agglutination Tests Hemagglutination Inhibition Test Preparation of Soluble Complexes from Chicken Anti-BSA Immunodiffusion Tests Starch Slurry Electrophoresis Cellulose Acetate Zone Electrophoresis Continuous Flow Paper Curtain Zone Electrophoresis Immunoelectrophoresis Gel Filtration Density Gradient Ultracentrifugation RESULTS . . . . . . . . . . . . . . . . . . . Effect of Heat Age of Birds and the RLF Reaction Ageing of Serum and the RLF Reaction Species Specificity of the RLF Reaction Assay of RLF The Effect of Divalent Cations on the RLF Reaction Electrophoresis of RLF Euglobulin Precipitation and Kaolin Absorption of RLF Treatment of RLF with 2 Mercaptoethanol Density Gradient Ultracentrifugation of RLF Gel Filtration of RLF Summary of RLF Characteristics iii 15 15 l6 17 18 19 20 22 23 24 25 25 26 27 28 29 29 30 32 34 36 37 39 43 44 44 46 49 Page Effect of Absorption with Aggregated Gamma Globulin on Agglutinin Titre 50 Effect of EDTA and Ca++ on Anti-PPLO and Antijg. pullorum Agglutinin Titres 51 Comparison of Cryoglobulin, RLF and Soluble Complex by Immunodiffusion 54 The Effect of Kaolin Absorption, 2 Mercaptoethanol Degradation and Iodoacetate Treatment on Antijg. pullorum Agglutinin Titre 59 The Effect of Dilution on the Titre Increasing Capacity of NOrmal Serum 61 Immunodiffusion Analyses of Soluble Complexes Treated with Kaolin. 2 Mercaptoethanol. Iodoacetate Aggregated Gamma Globulin 63 Gel Filtration of Anti-PPLO Sera 66 Immunodiffusion Analysis of Immune Serums Fractionated with Sephadex G 200 76 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 83 SUMMARY . . . . . . . . . . . . . . . . . . . . . . . 92 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 95 iv Table 10. ll. 12. The The LIST OF TABLES effect of heating whole immune serum and gamma globulin at 56 C for 30 min on its agglutination activity against 'M. gallinarum . . . . . . . . . . . . . . time of appearance of the RLF reaction as related to cryoglobulin formation . . . Cross reactions between normal avian and The The The The The The mammalian serums and heated gamma glObul ins O O O O O O O O O O O O O O 0 O titration of RLF by the tube method . . . effect of EDTA on the RLF reaction . . . . effect of divalent cations on the RLF reaction . . . . . . . . . . . . . . . . . location of RLF and PPLO agglutinin in sucrose density gradients after ultracentrifugation . . . . . . . . . . . location of anti-BSA precipitin activity and RLF activity in Sephadex G 200 derived pools of 8-day post-immunization sera O O O O O O O O O O O O O O O O O O 0 effects of heating on anti-PPLO serum . . ++ effect of Ca EDTA and normal serum on PPLO and §. pullorum tube agglutination titres . . . . . . . . . . . . . . . . . . effect of kaolin, 2 mercaptoethanol and sodium iodoacetate on S. pullorum agglutination titres . . . . . . . . . . . effect of dilution on the titre increasing ability of normal chicken serum . . . . . Page 30 33 35 38 40 41 45 47 52 53 60 62 Table Page 13. Agglutination titres of serum fractions prepared by Sephadex G 200 chromato- graphy of anti—PPLO serum collected at seven different periods after inoculation of chickens with In. gallinarum . . . . . . . . . . . . . . 69 l4. Hemagglutination inhibition titres of serum fractions prepared by Sephadex G 200 chromatography of anti-PPLO serum collected at seven different periods after inoculation of chickens with M. gallinarum . . . . . . . . . . . . 70 15. RLF reactions of serum fractions prepared by Sephadex G 200 chromatography of anti-PPLO serum collected at seven different periods after inoculation of chickens with g. gallinarum . . . . . . 71 vi Figure LIST OF FIGURES Page The position of anti-PPLO agglutinin, RLF and aggregatable gamma globulin in segments from a starch slurry electrophorogram . . . . . . . . . . . . . 42 RLF and anti-BSA location in Sephadex G 200 column effluents using a chicken anti-BSA serum sample (1 column volume = 160 ml) . . . . . . . . 48 Double diffusion analysis of RLF, cryoglobulin and anti-BSA soluble complexes . . . . . . . . . . . . . . . . 58 Immunodiffusion analysis of purified gamma globulin and soluble complexes formed from serum treated with kaolin, aggregated gamma globulin and mercaptoethanol . . . . . . . . . . . . . 65 HI and agglutination response of chickens. inoculated for 17 days with viable .g. gallinarum. as shown by chromato- graphy of their serum on Sephadex G 200 (1 column volume = 160 ml) . . . . . 72 The course of light and heavy antibody production following injection with ‘g. gallinarum . . . . . . . . . . . . . . 74 Cellulose acetate electrophoresis of Sephadex prepared chicken macroglobulins . 78 Double diffusion analysis of Sephadex prepared chicken macroglobulins and anti-BSA soluble complexes . . . . . . . . . . . . 79 vii INTRODUCTION The research reported herein deals with several unique properties of chicken serum. In the pursuit of answers to problems which involve avian immunology and serology one is immediately impressed by the paucity of information in this field and the inability to accurately execute in vitro tests which are common-place when dealing with mammalian sera. Problems which consistently confound researchers in this area include titre increase upon storage at low temperatures, apparent loss of some agglutinin titres following conventional methods of serum inactivation. heat stability of hemagglutination-inhibition antibodies and the development of dense precipitates in avian serum after ageing in the cold. The lack of complete precipitation of antibody-antigen complexes at optimal proportions in precipitin systems precludes the use of the quantitative precipitin test on avian systems and makes radio-isotope tech- niques a necessity when information regarding absolute antibody protein is desired. Interpretation of results is further complicated by the so-called ”non-specific" co- precipitates which occur in avian systems and increase the apparent antibody protein present in specific aggregates. The complement fixation test is also hampered by severe interclass C'l incompatability. Within the past few years interest in chickens and embryonating chicken eggs haveincreased greatly. Chickens have been included in studies in the fields of immune tolerance, tissue transplantation, oncology-and experimental embryology. Many different human and animal vaccines are being produced from agents grown in avian eggs and avian tissue cultures. Chickens themselves are becoming more important as an inexpensive food source and as such will probably get much more attention from veterinary immunologists in the future. It is felt that in order to more effectively attack problems in these various fields a more thorough understanding of the basic mechanisms involved in the chicken antibody system must be obtained. It is with the problems of cold storage titre increase, heat lability of avian antibody and non-specific co~precipitation that this thesis is concerned. LITERATURE REVIEW The serological analysis of chicken serum from birds with Chronic Respiratory Disease (CRD) is a relatively new area of applied immunology. In the short time since the discovery by Delaplane and Stuart (1943) that Mycoplasma gallinarum was an etiological agent in CRD most of the common serological diagnostic procedures have been used for deter— mining the presence or absence of antibodies stimulated by this pleuropneumonia-like-organism (PPLO) in infected birds. With the discovery by Van Herick and Eaton (1945) that PPLO gave a positive hemagglutination (HA) reaction. sero—diagnosis of CRD became possible. Jungherr. Luginbuhl and Jacobs (1953) reported the development of a hemmaggluti— nation inhibition (HI) test which was followed by the HI combining power test of Crawley (1960). Jacobs et a1. (1953) and White et al. (1954) each used their own version of the HI test for the diagnosis of CRD. Jungherr et a1. (1953) and Adler (1953) have reported use of the direct agglutination test in the area of CRD diagnosis. A flourescent antibody technique for the detection of PPLO in tissue culture has recently been described by Barile and Riggs (1961). Due to the incompatability of chicken and guinea pig complements. complement fixation tests have to this point been used only for the identification of organisms derived from in vitro cultures. Fortunately, useful diagnostic tests are now conceivably possible in the light of the researches of Benson et al. (1961). No known work has been done to date with purified Mycoplasma somatic or extracellular antigens. This is probably due primarily to the difficulties encountered in growing large quantities of organisms and the virtual absence of a cell wall. The most abiding difficulty in the testing of avian sera for anti-PPLO antibodies is the apparent heat lability of these antibodies. HEating serum to 56 C for 10-30 min has been observed by a number of workers and by the author to destroy or seriously limit the ability of the serum to agglutinate suspensions of PPLO (Cover et a1.. 1960; Jungherr et a1.. 1953; Jacobs et al., 1953). Strangely enough the HI reaction is not affected by heat according to Jacobs et a1. (1953) and Jungherr et a1. (1953). This observation leads one to feel that the antibodies responsible for agglutination and HI are separate entities which differ in heat sensitivity. ConCOmitant with the effect of heat on avian immune serum is the difficulty in correlating the titre of a fresh avian immune serum and the same serum after ageing for various time periods. Cover et a1. (1960) reported an increase in PPLO agglutination titre but not in Newcastle HI titre after ageing single serum samples for various time periods at 4 C and —20 C. Wolfe (1942) has reported a similar finding using the chicken anti—bovine serum albumin (BSA) system. He also found no difference in the activity of heated and non-heated chicken anti BSA serum. Hektoen (1918) described the first of a series of differences between avian and mammalian antiserums. He observed the rapid formation of a nonspecific precipitate upon thawing tubes of avian serum. He was also first to observe the advantage in titre derived at 1.8% salt con- centration as opposed to the conventional 0.9%. WOlfe (1942) discovered the high titres attainable in short periods of time after injection of protein antigens and the apparent rise of titre upon storing the serum at low temperatures. The appearance of dense flocs upon ageing was noted and for the first time a hypothesis for the increase of titre was given. Wolfe felt that a precipitin inhibitor which denatured and came out in the flock was responsible for initial depression of precipitin titre. In a later publication Wolfe and Dilks (1946) determined that an increase in titre started as early as 11 hr and continued to rise until a peak titre was observed 5—8 days after drawing the blood. This titre held for longer than 6 months. Titres attained by dilution reached into the millions and were easily produced by a single injection of antigen after 5-8 days. Goodman et a1. (1951) propounded that 8% salt concentration was required for precipitation of all of the antibody-antigen complexes in reacted serum and that the reaction at this concentration was absolutely specific and not simply a salting out phenomenon. They attributed the increase at 8% salt concentration to a low avidity characteristic of chicken antibody. The salt aided precipitation but did not cause the antibody—antigen reaction. wolfe and Dilks (1948) determined that the chidken attained immunological maturity at 5 weeks of age, but that even new-born chicks produced low levels of anti- body in 50%,of the cases. The number of avian species shown to produce antibodies was enlarged by Wolfe and Dilks (1949) by the addition of owls. pheasants. and partridges (good producers); and ducks and turkeys (poor producers). Guinea fouls and pigeons produced no precipitins under the conditions used. Schmidt and Wolfe (1953) tested a variety of antigens and discovered that chidkens produce high antibody titres to many antigens but that plant antigens were generally inferior to the animal antigens. They found that haemoglobin. pneumococcus polysaccharide and thyroglobulin were non-antigenic in chickens. The work of Goodman et a1. (1954) supported the previous conclusion that ionic strength (0) played a great role in precipitating previously formed antibody-antigen complexes by salting them out. They added that different ionic species. especially iodide and thiocyanate. differed in their ability to cause complete precipitation at the same u. Goodman and Ramsay (1957) restated the specificity problem and again showed that 8% salt was indeed causing the salting-out of specific soluble complexes and not causing non-specific precipitation. Deutsch et a1. (1949) reported that 50%.of the reported fispecific" precipitate was made up of an alpha globulin which increased during immunization and could be absorbed out with the specific precipitate by the antigen. The first hint that a macroglobulin antibody might be present in chicken serum was also presented in this article in the form of the Observation that gamma 1 or beta 2 increased significantly during immunization. Banovitz and wolfe (1959) repeated the work of Deutsch et a1. (1949) with negative results but indicated that the alpha globulin may possibly be present in a maSked form or that it was not resolubilized in antigen excess therefore making it unavailable for study by their indirect methods. Banovitz et a1. (1959) also suggested two antibody types on the basis of two different soluble complex peaks on free boundary electrophoresis. Re— peating the procedures of Deutsch et a1. (1949) gave no indi— cation of an alpha electrophoretic component involved in specific avian antibody systems. Gengozian and Wolfe (1957) used many different combinations of fresh. heated, aged and EDTA treated antiserums which added to specific precipitates and concluded that indeed a co-precipitating factor was present in normal and immune serums. A very definite increase in titre was seen by using the above treated serums in both high and low salt concentrations. This was attributed in part to the co—precipitin. Reasoning that mammalian serum did not have enough complement to contribute the weight of the non—specific substance observed. the authors disclaimed the identity of complement and co-precipitin as a possibility. They also reasoned that since low temperatures preserve complement there should be more non- specific substance in precipitates formed from serum stored at ~20 C than from those stored at 4 C. The reverse situation was shown. This evidence further supports the non-complement hypothesis. On the basis of the evidence presented it is believed that the conclusion of Gengozian and wolfe (1957) was premature. Several other lines of evidence will be developed for the identity of the protein later called Co P by makinodan et a1. (1960). Makinodan et a1. (1960) used the soluble complex technique of Heidelberger and Pedersen (1937) in conjunction with immunodiffusion to visually demonstrate the presence of Co P in specific precipitates. The molecule proved to be a 218 macroglobulin which migrated in the beta 2 electrophoretic region. In most cases its incorporation into the complex could not be prevented with EDTA. Co P constituted approximately 20% of the weight of the precipitate. The molecule was not present in soluble complexes formed when 0.15 M NaCl and fresh serum were used but was present when fresh serum and 1.5 M NaCl were used. Use of low titred (less than 350 ug of antibody N/ml) serum which had been heated and reaction of the antigen in the presence of EDTA inhibited the incorporation of Co P. Gengozian et a1. (1962) indicated that turkey. pheasant, rat. mouse, duck and guinea pig serum contained a similar Co P component. The authors presented some experimental evidence for the specificity of species Co P for its homologous gamma globulin. Orlans et a1. (1961) disagreed with the interpre- tation of the results with Co P. Their work indicated the presence of two different antibodies which had low (180.000) 10 and high (600.000) molecular weights. They submitted that since no proof had been given that the beta 2 and the 218 component were the same, the possibility existed that the beta 2 component was a heavy antibody. They did not find the 218 component consistently in their preparations. They pointed out its occasional occurrence but did not con- sider it the cause of the increased precipitation. Their explanation for increased precipitation was the salting out of monovalent antibody-antigen complexes containing the light antibody. It was significant. however, that the light antibody did precipitate to a small degree at 0.15 M salt concentration due to co-precipitation with the heavy divalent species. This complex would take the form. heavy antibody- antigen -light antibody, in the soluble complex. The mono- 2 valent species in antigen excess constitutes an end group. They postulated that C'l may well be another factor which caused eventual precipitation of non-precipitating complexes at increased salt concentrations. Orlans (1962) reported that increased salt con- centrations alone were not sufficient to cause the precipi- tation of monovalent antibody-antigen complexes and that the macroglobulin antibody previously reported did not appear to have the ability to precipitate antigen by itself. She suggested that the macroglobulin bore great 11 resemblance to a component of complement and concluded on the basis of precipitin curves and deft manipulation of a number of immunodiffusion techniques that, "As increased precipitation by fowl antisera at high salt concentrations is due neither to a simple salting-out process of soluble complexes, nor to increased precipi- tation of a macroglobulin (as suggested by Makinodan et al., 1960) the explanation must be sought in a more compli- cated interaction either between two types of antibody. or between antibody and a non-specific constituent of serum.” Aitken and Mulligan (1962) used a labeled BSA- anti BSA system and found that complete precipitation of antigen was effected only in extreme antibody excess and that at the point of maximum precipitation of antibody only 50% of the antigen was precipitated. Of particular interest in this work was the observation that several curve anomalies occurred at 1% and 4% NaCl concentrations but variations were slight at 8%. Such anomalies were found in the extreme antigen excess region of the precipitin curve. Although these authors preferred to attribute these anomalies to different concentrations of the heavy and light anti- bodies described by Orlans et a1. (1961) they pointed out that non-specific precipitation could explain the results obtained. With these results in mind it is easy to see why 12 the quantitative precipitin test cannot be used to determine the absolute antibody content of chicken serum. Although an equivalence zone does exist in chicken serum antigen mixtures. if the proportion is right, it cannot be detected with the conventional precipitin test. Maurer and Weigle (1953) pointed out several important characteristics of aged, non—haemolytic complement. Although EDTA effectively prevents incorporation of guinea pig C'l into specific aggregates (Levine et al., 1953) of many different species it has no such effect on rabbit antiserum systems. It seems that statements about C'l inhibition with EDTA should be made with care. They also showed that loss of haemolytic activity accompanying ageing of serum was no assurance that the combining capacity of the complement was lost. They postulated that an inhibitor to complement, which prevents its addition to specific aggregates. was present in fresh normal serum but lost its inhibitory activity on ageing and on dilution. Weigle and Mauere (1957) presented evidence for the presence of this inhibitor but did not isolate or characterize it. Bushnell and Hudson (1927) and Rice (1947) reported that heating avian antiserum destroyed its already limited ability to fix guinea pig complement. Brumfield and Pomeroy 13 (1957, 1959) published the details of a direct complement fixation test using avian antiserum and guinea pig complement in which the order of addition of reagents was very important. It was later discovered by Benson, Brumfield and Pomeroy (1961) that C'l was the essential component in chicken serum that was required for the fixation of guinea pig C'4, C'2 and C'3. These workers found avian C'l to be a beta 2 heat labile euglobulin which was removed by kaolin absorption. Rheins et a1. (1957) reported that another peculiar attribute of avian serum was its ability to fix or precipi- tate latex particles. This characteristic is shared by the human rheumatoid factor (Singer and Plotz, 1956). Seeing the obvious connection between the Rheins phenomenon. the ability of rheumatoid factor to precipitate heat aggregated gamma globulin and specific antibody-antigen complexes. Franklin (1962) examined the possibility that Co P and the Rheins factor were identical. Fresh and frozen sera were found to contain the factor, but since Franklin was rather broad in his interpretation of the term "fresh," all of his observations must be considered as being made using aged sera. The activity was abolished by heating at 56 C for 30 min and by treatment with EDTA. Inactivation of C13 and C'4 did not affect the titre. A very superficial attempt 14 at absorbing out the activity with antibody-antigen com- plexes was tried with no affect on the titre. This should not be considered too meaningful since all variables were not controlled. The active fraction was located in the beta 2 region on electrophoresis and at the bottom of the density gradient tubes used in ultracentrifugation experi- ments. The fraction was 198 according to analytical ultra- centrifugation. Activities as judged by latex fixation were destroyed by mercaptoethanol treatment but not by low pH (3) or urea. Franklin concluded that since specific aggregates did not absorb out the latex agglutinating factor the two proteins probably belonged to different groups but shared many similar properties. MATERIALS AND METHODS PPLO Antigen and Salmonella pullorum Antigen The same PPLO stock antigen suspension was used for HI and agglutination tests. Bacto PPLO Broth Base was pre- pared and sterilized as directed. Thallium acetate which had been sterilized by Seitz filtration was added to a concentration of 1:10.000. Bacto PPLO Serum Fraction was then added to a final concentration of 2%. Sterile Penicillin G was added to a final concentration of 50 units/ml of broth. The inoculum of PPLO in the log phase was then added to a final concentration of 1% and the broth incubated at 37 C for 4 days. Sterility tests were performed at 3 days and 4 days using brain heart infusion agar. The antigen was harvested by centrifuging at 8000 X g for one hour and twice washing with saline under the same conditions. A final volume of 30 ml of saline was used to resuspend the cells harvested from 10 liters of culture fluid. No preservative was added nor was the stock antigen heated. It was stored in a sterile vaccine bottle at 4 C and used within 2 weeks. 15 l6 Antigen used for the agglutination test was diluted with saline to the turbidity of a McFarland #5 turbidity standard and sonicated for 30 sec in a Raytheon 10 kc Model 101 sonic oscilator to break up clumps. Antigens for the HA or HI tests‘wereused at a concentration of 3 hemagglutinating units/test. The final test volume was 1 ml. All dilutions were made in saline and the antigen was sonicated before use as a dilution of McFarland #5 turbidity for 30 sec. Salmonella pullorum antigen was grown in brain heart infusion broth for 24 hr and was harvested by centrifiguation like the PPLO antigen. Formalin at 0.3%.concentration was used as a preservative. Antigen corresponding in turbidity to a McFarland #6 standard was used in the tests. Serum Collection All serum. rabbit or avian, was collected from 20 ml of whole blood after clotting for 2 hr at 37 C in slanted acid washed 25 mm x 150 mm tubes. The serum was clarified by gentle centrifugation and frozen at -20 C. All serums were considered aged if they were used longer than 2 hr from the time of their collection from the clot. l7 Antiserum Production Antiserums were developed in 6-month-old male White Leghorn chickens against bovine serum albumin (BSA) by intravenous inoculation of 40 mg BSA/kg body weight contained in 2 ml of saline. Injection was via the wing vein. Seven to 10 days after injection the birds were bled by cardio— centesis. The serum collected from eight birds was pooled. Rabbit anti-chicken serum of high titre required a lengthy production period. Intramuscular injection of four 6-month-old male New Zealand White rabbits with alum precipitated whole chidken serum at 14 day intervals for a total of 8 injections was followed by a single intravenous injection of 1 ml of whole chicken serum. The alum precipi- tated antigen was prepared by diluting 25 ml of serum with 80 m1 of water and 90 ml of 10%KA1(SO4)2 and adjusting the solution to pH 6.5 with 5 N NaOH. The white precipi— tate was washed twice in saline containing 1:10.000 merthiolate and resuspended to 100 ml final volume. Two 5 ml injections. one in each thigh. constituted a single dose. The serum was collected 10 days after the intravenous injection. Chicken anti—PPLO serum was produced by inoculating 3-week—old White Leghorn male chicks with 1 ml of an air l8 sac homogenate made up in PPLO base broth containing 1:10.000 thallium acetate per ml and 1000 units of penicillin per ml. The inoculation was by the intratracheal route. The inoculum contained from 106 to 107 organisms/ml. After a given time blood was taken by cardiocentesis and the serum harvested and stored as previously described. Chicken anti5§, pullorum serum was produced as described for anti-BSA serum with the exception that the antigen was 1 ml of a formalin killed saline suspension adjusted to the turbidity of a McFarland #6 standard. Preparation of Avian Gamma Globulin Electrophoretically homogeneous preparations of avian gamma globulin were prepared by a combination of several standard techniques. A euglobulin precipitate was prepared from pools of whole fresh normal chicken serum by dialyzing against 100 volumes of 0.005 M KH2P04 at 4 C for 24 hr. After collecting the precipitate by slow speed centrifugation at 4 C and washing once with KH2P04 the pellet was redissolved and returned to the original serum volume with saline and precipitated with three volumes of 0.4% 2-ethoxy-6,9- diamino acridine lactate (Rivanol) and unbuffered saline. The mixture was adjusted to pH 7.6 with 5 M.sodium carbonate l9 and after 15 min was centrifuged at 4 C and 800 X g for 3 hr. The supernatant fluid was adjusted to 34% saturation with 68% saturated (NH4)2SO4 and allowed to stand at 4 C for 2 hr. The precipitate was centrifuged out at 4 C and resuspended in distilled water to the original serum volume. The ammonium sulfate precipitation was repeated three times and the final precipitate resuspended in 0.3 M NaCl to 45% the original serum volume. After extensive dialysis against 0.3 M NaCl for several days to remove all Rivanol the material was adjusted to 50% the original serum volume with 0.3 M NaCl. The majority of the gamma globulin pools were in addition cleared of trace amounts of Rivanol by chromato- graphy on Sephadex G 25 with concentration to double strength by pervaporation and dialysis against 0.3 M NaCl. Agglutination Tests Tube agglutination tests were performed by diluting test serum in a suitable saline base diluent to a final volume of 0.5 ml. Depending upon the particular test the saline base contained EDTA at various concentrations, serum or calcium chloride. In addition 0.5 ml of antigen corrected to turbidity of a McFarland #6 standard was added and mixed with the diluent by shaking. The tests were read after 2 hr incubation at 37 C and again after 24 hr at 4 C. The 20 last tube to show a 2+ reaction was taken as the end point. Slide agglutination tests were performed by adding 0.05cc of serum and 0.05 ml of antigen adjusted to a #3 McFarland turbidity scale with saline. The test plate was rotated for thorough mixing and read after 5 min. Results were recorded according to the degree of clumping observed. Hemagglutination Inhibition Test Hemagglutination inhibition tests were done on serums diluted in saline to avolume of 0.5 m1. To each tube 3 hemagglutination units (HAU) of PPLO suspension contained in 0.25 ml were added and the mixture allowed to incubate for 10 min at 37 C. Twenty-five hundredths m1 of a 0.5% suspension of chicken erythrocytes in saline containing 1:1000 normal chicken serum was added. The test was read after 2 hr at 27 C and the highest serum dilution showing inhibition was designated as the titre. A serum macroglobulin exists in the blood of people suffering with rheumatoid arthritis which precipitates solutions of aggregated gamma globulin. One of the tests used for detecting the presence of this substance is performed much like an immune precipitin reaction with the exception that aggregated gamma globulin takes the place of the antigen. Such a test forms the basis for those tests done here in 21 connection with an avian rheumatoid-like-factor (RLF). The RLF tests performed were of two types: the tube test and the interfacial test. The former is quantitative or qualitative while the latter is only qualitative. Inter- facial tests were done using rivanol prepared chicken gamma globulin (CGG) which had been heated at 65 C for whatever time was indicated by previous standardization of the pool: i.e., the time necessary to give optimal tube titres and the absence of auto-precipitation of aggregated CGG upon standing at 4 C for 12 hr. The twofold concentrated pool was diluted 1:4 with water after heating, and 0.1 m1 layered over 0.1 m1 of whole test serum. In positive samples a precipitate forms at the interface within five minutes and usually within one. The test must be read quickly as the precipitate may become so dense that it clouds the whole tube within 1/2 hr. Tube RLF tests were performed by adding an equal volume of heated CGG and test serum appropriately diluted. After mixing the tubes were incubated for 2 hr at 37 C and overnight at 4 C. They were then washed twice in cold saline and Folin tests (Kabat and Mayer, 1961) were done on the precipitate and the mg protein per test determined. In a variation of this test a precipitate is formed within 3 hr in the presence of CGG which has been heated to the degree that it autoprecipitates in the absence of serum. A control 22 which consisted of heated CGG but no test serum was run. The last tube which showed a precipitate level higher than that of the control is assumed to be the end point. This tube value, expressed in mg protein precipitate per test. is recorded and the value of the control in the same units is subtracted from it. This value multiplied by the dilution factor and corrected for the amount of diluted serum used per test gives the mg of protein/m1 of whole serum. This value can be converted to give per cent RLF protein in serum by dividing by the mg protein/ml of the serum. Preparation 9; Soluble Complexes from Chicken Anti—BSA Chicken anti-BSA pools were diluted 1:4 in saline and equal parts of BSA in saline were added. The BSA concentration varied from 0.5 mg/ml to 0.0078 mg/ml. After incubating for 2 hr at 37 C and 12 hr at 4 C the precipi- tate was centrifuged at 3200 X g for 30 min. The super- natant fluid was tested for antigen and antibody excess by the interfacial method using 0.0625 mg BSA/ml of saline to test for antibody excess and whole anti-BSA to test for antigen excess. The equivalence point is defined as the tube where neither antibody nor antigen excess occurred. These determined proportions were used to produce specific precipi— tates. 23 In the production of the complex a precipitate formed at equivalence was treated with 100 X the equivalence amount of BSA which had been dissolved in 25% the original volume of whole antiserum used to form the precipi- tate. After 6 hr incubation at 37 C with frequent agitation the precipitate practically all dissolved. The supernatant fluid was considered the soluble complex. Immunodiffusion Tests Ouchterlony gel diffusion plates were made according to the methods of Smith (1960) and Allison and Humphrey (1960). Agar was dissolved in hot buffered saline and 10 ml of this mixture were poured over the surface of a lantern slide which had been cleaned in acid. After cooling and setting for 1 hr at 4 C the desired pattern was punched out with a cork borer using a drafted template underneath the slide as a guide. The patterns were developed in a humid chamber at room temperature for at least 4 days before clearing the unreacted protein from the agar by allowing the slides to soak in saline for 24 hr and in distilled water for another 24 hr. The plates were dried under moist filter paper and stained with Crowle's triple stain (Crowle, 1961). These plates can be stored indefinitely and used as a negative for producing contact prints of the 24 patterns. The prints in this thesis were produced from original plates using Velox F-4 single weight photographic paper and developing with Dektol. Starch Slurry Zone Electrophoresis Three types of electrophoresis were employed in the various analytical and preparative procedures to be discussed. Starch slurry electrophoresis was performed in a Shandon migration chamber at 4 C for a period of 6-10 hr using a discontinuous barbital buffer as regards ionic strength (u). A 4'ml whole serum sample was applied to a trough 0.6 cm wide and 7 cm from the cathode. The matrix and well pH were 8.6. A constant potential of 110 v was applied until the intense yellow pre-albumin band had moved to within 4 cm from the anodic end of the template. Template dimensions were 165 mm X 190 mm X 5 mm. Sections of starch 6 mm wide were cut out and the protein eluted by replacement filtration using phosphate buffered saline at pH 7.4. This buffer will be designated hereafter as saline since it was used in all of the experiments to be described. The fractions were analyzed by the Folin phenol method as described by Kabat and Mayer (1961). 25 Cellulose Acetate Zone Electrophoresis Cellulose acetate membrane electrophoresis was done in the Shandon electrophorator at 275 V for 45 min. Six lambda samples were used unless otherwise indicated and were applied exactly over the cathode onto 2.5 cm X 12 cm strips of cellulose acetate which had been previously impregnated with tris-boric acid-EDTA strip buffer at pH 9.1 (Goldberg, 1959). The well buffer was a barbital-barbituric acid mixture with a pH of 8.6. It is discontinuous as regards pH. u, and components. After electrophoresis the protein was denatured by immersion in 5% trichloracetic acid. Staining was accomplished by direct transfer into 0.2% Ponceau S stain or Nigrosin made up in 2%»acetic acid. Staining time was 5 min in Ponceau S or 12 hr in Nigrosin after which the Ponceau was resolved with 2% acetic acid and the Nigrosin with water. Strips were dried between weighted towling. Continuous Flow Paper Curtain Zone Electrophoresis Continuous flow electrophoresis was performed on a Beckman CP cell using a barbital buffer of u 0.02.pH 8.6 and a flow rate of 0.77 ml/hr onto the center applicator tab. Constant current was maintained at 40 ma in a 4 C room. No plate cooling was used. The following settings 26 were used consistently and with good results to maintain the correct buffer flow; reservoir syphon-ll; wick syphons left-7, right-8. The 2 cm diameter sample holder was used and was driven at speed 3. Upon completion of the run the curtain was dried at 90 C in an electric oven and stained with Ponceau S. Samples were concentrated by pervaporation followed by dialysis. Immunoelectrophoresis Immunoelectrophoresis was performed on 3" x 19 glass slides. The slides were prepared by washing in potassium dichromate cleaning solution followed by multiple rinses in distilled water. After air drying, 2 ml of molten agar buffered at pH 8.6 with barbital—barbituric acid were poured over them and allowed to cool and set at 4 C (Hirschfield, 1960). Migration was performed at a constant current of 22 ma/l6 slides for a period of 110 min in a Shandon migration chamber which was converted for slides by placing a 6" x 8—1/2? casette across the bridge. Two rows of eight slides each were accommodated on each casette. In each slide, two side wells 1 mm x 65 mm long and separated by 13 mm were cut with a template. Similarly a 1 mm x 2 mm sample cell was cut equidistant from the sides and the ends of the slide. 27 Hyperimmune serum from a single bleeding of a single rabbit given multiple injections of alum precipitated whole normal chicken serum was used to develop the spectrum after electro- phoresis. Development was allowed to proceed at 25 C for 48 hr in a diffusion chamber saturated with water vapor. After development the slides were allowed to clear of unreacted components by submersion in buffered physiological saline and finally in water. They were then dried and stained with Ponceau S as described by Crowle (1961). Gel Filtration Sephadex gel filtration with the G 200 resin was done in a glass column 65 cm high and 4.5 cm inside diameter. The gel was swelled with saline or some suitable buffer from 30 gm of starting material and was added to the column from a constantly stirred overhead reservoir through a tube attached to the head of the column. In this way a homogeneous gravity packing without pleating was accomplished. Serum sample sizes of 10 ml were used. The sample was added and allowed to go into the gel and then washed in further with 50 m1 of the column buffer. Fraction collecting using a 5 ml volumetric syphon was started upon the addition of the sample. Protein content of each sample was determined by the Folin method before pooling into five tube samples. 28 Each pool was pervaporated to a desired concentration and dialyzed against saline or phosphate buffered saline and frozen at -20 C for future use. Density Gradient Ultracentrifugation Density gradient ultracentrifugation was done on a Spinco Model L using an SW 39 L rotor at 4 C for 14-1/4 hr and at 35,000 RPM. A 10% to 40% surcose continuous gradient in 0.15 M NaCl was established in 5 ml tubes and 0.5 ml of whole serum or of serum fraction was layered over the gradient. The interface was disturbed by gentle rotation with a sealed capillary tube before beginning the run. Individual portions were withdrawn from the tubes from the top using a blunted 22 gauge needle fitted to a tuberculin syringe. Each sample was dialyzed 24 hr at 4 C against saline and analyzed by the RLF test, cellulose acetate electrophoresis and tube agglutination. RESULTS Effect of Heat A number of experiments were performed to determine the extent to which avian antibodies were sensitive to heat. The first of these experiments is summarized in Table 1. An avian serum pool from 8 birds which had been infected with M. gallinarum for a period of five weeks was divided into two equal portions and treated in the following manner: 1. Titred against M. gallinarum antigen by the tube method. Heated at 56 C for 30 min and then titred by the tube agglutination method using M. gallinarum antigen. Heated at 56 C for 30 min followed by the addition of an equal volume of normal chicken serum prior to titration in the M. gallinarum antigen. Precipitation of gamma globulin with 34% saturated (NH4)ZSO4 and treatment of the precipitate with Rivanol. Titration of the Rivanol preparation (diluted to original serum volume) was performed against M. gallinarum by tube agglutination. 29 30 5. Preparation #4 was heated for 30 min at 56 C and titred against M. gallinarum. 6. Preparation #5 was diluted in an equal volume of whole normal serum and titred against M.Agallinarum. It will be noted that the titre dropped after heating whole immune serum but not after heating gamma globulin. Addition of normal serum to heated whole immune serum had no effect on the titre of the latter. The titre of the gamma globulin was one-half that of the whole immune serum. Table l. The effect of heating whole immune serum and gamma globulin at 56 C for 30 min on its agglutination activity against M. gallinarum. Treatment Titre* 1. Unheated Whole Immune Serum 320 2. Heated Whole Immune Serum 10 3. Heated Whole Immune Serum + NOrmal Serum 10 4. Unheated Anti-PPLO Globulin 160 5. Heated Anti-PPLO Globulin 160 6. Heated Anti-PPLO Globulin + Normal Serum 10 *Titre expressed as a reciprocal in agglutinating units/ml of undiluted antiserum. Age 9: Birds and the RLF Reaction Both the heated serum and the heated gamma globulin became slightly opalescent upon cooling to room temperature. This seemed to indicate that some modification of the gamma 31 globulin had taken place during heating. The possibility that a rheumatoid-like reaction had occurred in the whole serum with modified gamma globulin was considered. Per- formance of interfacial tests using a variety of serums of different ages from birds of different sexes, ages and degrees of infection revealed that all gave a very faint line of precipitate at the interface between the serum and the aggregated gamma globulin. The strength of this reaction was found to be greatly increased when higher heats (61 C - 65 C) were used for the aggregation step. Eight—bird serum pools from individuals 3 hr, 24 hr, 48 hr, 72 hr, 1 week, 2 weeks, 3 weeks, 4 weeks, 5 weeks, 6 weeks, 2 months, 3 months, and 4 months of age were tested for the presence of a rheumatoid-like factor (RLF). Testing was done using Rivanol treated gamma globulin which had been heated at 61 C for 10 min followed by dilution in saline to the original normal serum volume. The interfacial test was read after 10 min incubation at 27 C. It was found that a rheumatoid-like reaction occurred in birds of both sexes and at all ages. The reaction seemed to be most pronounced with serum from newly hatched chickens. Serum from chickens hatched and reared under germ-free conditions showed the RLF reaction with greater alacrity than serum from conventional birds. 32 Ageing 2f Serum and the RLF Reaction It was found early in the course of these experiments that ageing of the serum was required before the RLF reaction would occur. It was similarly noticed that concomitant with the appearance of RLF reaction capability, cryoglobulin, a dense white to cream colored floccule, formed in the serum on cold storage. This was observed only after ageing at low temperatures. An experiment was performed in an attempt to elucidate this observation. Sera obtained from four 2—month—old cockerels were pooled and aged under a variety of conditions for varying periods of time. At given intervals these samples that were stored in the cold were returned to room temperature and the presence or absence of cryoglobulin was recorded. At the same time an interfacial test was done on each sample to check for RLF activity. The results of these tests are shown in Table 2. It can be seen that the RLF reaction did not occur with fresh serum and did not occur with serum stored at room temperature or at 4 C until after three days of ageing when the reaction was read after 2 hr at 27 C. Freezing for a period of 24 hr, however, caused the appearance of both cryoglobulin and RLF reaction capability. HOlding fresh serum for 24 hr at 4 C in the presence of aggregated 33 Table 2. The time of appearance of the RLF reaction as related to cryoglobulin formation. RLF reaction Storage Storage Cryoglobulin time temperature formation 2 hr @>27 C 24 hr @34 C 0 hr - _ _ 24 hr 27 C — - + 4 C — - + 0 C + opalescent + -20 C + opalescent + 3 days 27 C - + 4 C - + + CC + + —20C + + + 5 days 27 C - + + 4C + + + CC + + + —20C + + + 7 days 27 C - + + 4 C + + + CC + + —20C + + + 2 months 27 C - + + 4C + + + CC + + + —20 C + + + 34 gamma globulin did cause an RLF reaction to occur. Certain low temperatures seem to be required for the formation of cryoglobulin since ageing for 2 months at 27 C caused no precipitate formation. It was subsequently demonstrated that upon cooling this aged sample, cryoglobulin formed within 24 hr. The RLF reaction did not appear to depend upon cryoglobulin formation as much as upon ageing since RLF activity existed in aged serum in which no visible cryoglobulin had formed. Species Specificity g; the RLF Reaction The reactions between various avian and mammalian serums and homologous and heterologous aggregated gamma globulins are recorded in Table 3. In all cases the gamma globulins were isolated by fractionation with 34% saturated ammonium sulfate following starch slurry electrophoresis at pH 8.6 using(JJM5u barbital buffer. The aggregation times at 61 C varied from preparation to preparation due to different degrees of purity and differences between the molecular species. The aggregation time was determined by visual inspection as that time at which a slight opalescence developed in the sample. Interfacial tests were used exclusively in this experiment. Chickens, turkeys and geese appeared to contain an 35 Table 3. Cross reactions between normal avian and mammalian serums and heated gamma globulins. Species serum Species gamma globulin* 1 2 3 4 5 6 Chicken (1) 4+ 4+ 2+ - _ - Turkey (2) 4+ 4+ + — _ _ Goose (3) + 3+ + + — - Duck (4) - 2+ + 2+ _ _ Horse (5) 2+ - + _ _ _ Rabbit (6) 2+ + — _ _ _ *Gamma globulins purified by starch slurry electro— phoresis followed by precipitation with 34% saturated ammonium sulphate. RLF factor in their serum which reacted with heterologous, aggregated avian gamma globulin, but which reacted only slightly with mammalian gamma globulin aggregates. Duck RLF seemed to be quite specific in its reaction. Duck gamma globulin was observed to be non-reactive with the RLF from other species. The two mammalian serums tested showed no RLF reactions but their aggregated gamma globulins were precipitated by chicken RLF and to a much lesser extent by turkey RLF (in the case of rabbit gamma globulin) and goose RLF (in the case of horse gamma globulin). 36 Assay of RLF A great number of different types of RLF assay procedures were tried with only limited success. Attempts to determine RLF titres by bentonite fixation, antibody sensitized erythrocyte agglutination, latex fixation and enhancement of S, pullorum agglutination all showed some promise but did not give the degree of quantitation desired in terms of an absolute value. It was decided that the tube RLF test could be used in such a way as to give a desirable approximation of serum RLF in absolute units. It was found by testing numerous serum samples that at a dilution of 1:8 only a very weak reaction occurred. At this serum dilution and at higher dilutions the system may be considered to be in gamma globulin excess. The antigen excess zone in agglutination theory offers a useful corollary in understanding the above statement. If the gamma globulin is heated so that in time and at 4 C it will auto- precipitate, it would be expected that little or no effect on the degree of precipitation would be observed if RLF were present in extremely low concentration. The absolute amount of RLF would be reflected in the excess weight of the autoprecipitate which occurred at points where RLF no longer was in concentration great enough to enhance precipitation but at which concentration it still combines. Subtracting 37 the weight of a control containing only autoprecipitated gamma globulin from the weight of a sample in extreme gamma globulin excess should give a value correctable to mg RLF/mg of serum and eventually to percent RLF in the serum protein. The sensitivity of this method is dependent on the sensitivity of the method used to determine protein. The sensitivity of the Folin Phenol method used is 10-100 ug of N while the error inherent in the test is.: 1 ug of N (Kabat and Mayer, 1961). The results of the experiment are shown in Table 4. As can be seen. 2.4% of the total serum protein in chicken serum (7.0 g/100 ml according to Putnam, 1960) was made up of RLF or, 1.66 mg of RLF were present per ml of whole normal chicken serum. The Effect 9: Divalent Cations 22 the REF Reaction It was consistently noticed during the course of these experiments that in order to demonstrate an RLF reaction in chicken blood products, defibrination by the glass bead technique or by whole blood clotting must be used. Collection of plasma by the citrate method yielded a product with no RLF reaction capability. It was considered possible that calcium or some other divalent cation may be required for the reaction. Experiments were designed to test this hypothesis. 38 Table 4. The titration of RLF by the tube method. mg protein precipitate Calculated mg Dilution per/0.25 ml (5 trial RLF per ml av.) undiluted serum* Undiluted 0.260 1.52 1/2 0.215 2.32 1/4 0.140 2.24 1/8 0.110 2.56 1/16 0.083 l.66** 1/32 0.070 0.00 1/64 0.070 0.00 1/128 0.070 0.00 *Formula for calculation of mg RLF/ml undiluted serum t°f'd(p-g) p total precipitate from 0.25 ml serum + 0.25 ml heated gamma globulin g autoprecipitate from 0.25 ml heated gamma globulin f correction factor for dilution of folin reaction to readable levels t correction factor for sample serum volume d correction factor for sample dilution (reciprocal of the dilution) Sample calculation for 1/16 dilution: 4°2°l6(.083-.070) = 1.66 mg RLF/ml undiluted serum **RLF content of serum = 1.66 mg/ml RLF = 2.4% of the total serum protein. 39 Disodium ethylenediamine tetraacetate (EDTA) was chosen as a suitable chelating agent. Equal volumes of saline with various EDTA concentrations were added to RLF positive whole normal chicken serum to give the final EDTA concentrations listed in Table 5. Both interfacial and tube RLF tests were done to determine the residual activity. The results indicated that RLF reactions were blocked at an EDTA molarity of 0.00250. In order to determine whether calcium alone or a number of divalent cations could relieve the inhibition, divalent cations were added to a previously inhibited system so that the final dilution of serum was 1:2. The results seen in Table 6 show that only calcium of all the ionic species tried could relieve the inhibition. It can also be seen that at high calcium levels the RLF reaction was inhibited. Electrophoresis 9f RLF Since such results indicate that complement fraction C'l and RLF are synonymous, several additional experiments were performed. Benson, et a1. (1961) reported that chicken C'l is a beta 2 euglobulin which can be absorbed with kaolin. Figure 1 represents the results of one of a number of electrophoretic determinations done with a starch 40 Table 5. The effect of EDTA on the RLF reaction. Serum dilution Final EDTA RLF reaction molarity 1:2 0.05000 - 1:2 0.02500 — 1:2 0.00500 - 1:2 0.00250 - 1:2 0.00050 + 1:2 0.00025 + 1:2 0.00005 + Heated gamma globulin + saline 0 Normal serum 1:2 + saline 0 Normal serum + unheated gamma globulin 0 41 Table 6. The effect of divalent cations on the RLF reaction.* £553.. 2:? 333:2 13:3 5352 RLF reaction 1:2 0.1500 0.00125 - 1:2 0.0500 0.00125 + 1:2 0.0300 0.00125 + 1:2 0.0250 0.00125 + 1:2 0.0150 0.00125 + 1:2 0.0050 0.00125 + 1:2 0.0030 0.00125 + 1:2 0.0025 0.00125 + 1:2 0.0015 0.00125 + 1:2 0.0005 0.00125 - *Relief of inhibition was not accomplished by Mg++, Mn++, Co++, Fe++ or Zn++. 42 1831 "IVOI90'IOU 38 +- .Emumouosmouuomao knusam summon m Eonm mucmemmm cw cflasnon mEEmm wanmummwummm paw mam .cflcflusammm OAmNIAHGM mo sofiufimom one mmmaaz man... on ON ON QN NN ON 0. 0. v. N. O. 0 0 t. +N 1.» 205mm N (hum +¢ Emboca ole 2323.3 0 33533: \\\\X .3 c :2: 2.5.5.63 311:: II“ o Z.E:o..< 2330040 (Sada .H onsmflm .1831 'IONBI-Id NI'IOJ '0 '0 43 slurry in barbital buffer at pH 8.6 and at 0.05 u. Each 6 mm segment of the completed electropherogram was analyzed for the ability to give an RLF positive reaction, the ability to be aggregated and give an RLF reaction with a known positive serum and the ability to agglutinate PPLO. The sample in this case was 4.0 ml of a 7 week post- inoculation serum pool. The results show that the beta 2 region of the electrophoretic spectrum contains the peak of the RLF activity. Anti-PPLO agglutinin activity was found to peak in the gamma globulin region and the ability to be aggre— gated and to give a positive RLF test was found all the way from the gamma to the alpha 2 region. The RLF and antibody peaks were nonconcomitant. Euglobulin Precipitation and Kaolin Absorption gf RLF Upon dialysis of the same serum specimen against 0.005 MZKHZPO4 all of the RLF activity was located in the euglobulin precipitate. Absorption of a 1:4 dilution of RLF positive serum with 180 mg kaolin /m1 of whole serum yielded a product which gave no RLF reaction. 44 Treatment of RLF with 2 Mercaptoethanol The sensitivity of RLF to the action of 2 mercapto— ethanol was shown by the following experiment. One ml of an RLF positive serum was reacted for various periods of time with 0.33 ml of 0.1 M mercaptoethanol and immediately dialyzed. This material was then tested for RLF activity after the addition of calcium ions to a final molarity of 0.015 and a serum dilution of 1:2. The mercaptoethanol treated serum showed no RLF reaction after 15 min. A control sample, treated identically with the exception that mercapto- ethanol was not added, remained positive. Density Gradient Ultracentrifugation 9f RLF It was felt that some idea of the molecular weight of the RLF could be gained by density gradient ultracentri- fugation. This means of analysis was chosen because of the ease with which mixed systems can be analyzed. The results of this experiment are shown in Table 7. The two samples were: (1) normal chicken serum and (2) 7 week post- inoculation serum. Both samples were RLF positive by the interfacial test. The results indicate that the RLF is a macroglobulin which is heavier than the 7S chicken immune globulin since 45 Table 7. The location of RLF and PPLO agglutinin in sucrose density gradients after ultracentrifugation. Volume Tube agglutination RLF Tube number withdrawn titre activity Sample 1. Normal serum pool (8 birds) 1 .5 - - 2 .5 _ _ 3 .5 .. .. 4 .5 - - 5 .5 - - 6 .5 - - 7 .5 - 2+ 8 1.0 - 4+ Sample 2. 62 day post—inoculation pool (8 birds) 1 .5 - - 2 .5 - - 3 .5 — - 4 5 10 - 5 .5 40 - 6 .5 80 - 7 .5 20 2+ 8 5 - 4+ 46 the RLF penetrated deeper into the gradient and was found at the bottom of the tube upon completion of the run. It was observed from the electrophoretic patterns done on the samples removed from the density gradient tubes that the RLF activity was located in the same gradient position as the alpha 2 macroglobulin. This observation further indicates its high molecular weight. RLF activity was found in both the normal and the immune systems. Gel Filtration of RLF Due to the macroglobulin nature of the RLF, Sephadex G 200 was used in an attempt to purify it. All the work done in this connection todk place before the importance of calcium in the RLF reaction was appreciated. Failure to recover activity after extensive purification can probably be explained in retrospect by the absence of calcium in the assay system. The results obtained by running 10 m1 samples of serum through G 200 columns made from 30 gm of dry Sephadex can be seen in Table 8 and Figure 2. All of the RLF activity was found in the first protein peak. This peak contained alpha 2 macroglobulin and can therefore be considered the macroglobulin peak. The presence of 7 S molecules in the second peak can be verified from Figure 2 which represents 47 Table 8. The location of anti-BSA precipitin activity and RLF activity in Sephadex G 200 derived pools of 8 day post-immunization sera. Pool number RLF reaction InFeFfécial (interfaCial) prec1p1t1n test 1 - _. 2 3+ _ 3 4+ _ 4 3+ + 5 2+ 2+ 5 - 3+ 7 - 4+ 8 - 3+ 9 - 2+ 10 — _ ll — _ 12 - _ l3 - - 14 - - 15 - _ l6 - _ 48 .AHE 00H u wEsHo> GESHOU av mamfimm Esnmm «mmlflucm swxofino m means mucosammm GEDHOU oom w xmpmnmmm :H soflumuoH 225400 as com o xmcmsmom co snumm Hawsu mo wnmmumoumfiouno an asonm mm mmmp ha How pmumasoosfi .mcmxowso mo uncommon soaumsflusammm tam Hm .ssumcfiaamm,.m.0HAMA> £043 .m mnsmflm 5952 .50.. 2 n. v. n. m. __ o. a o 4. o o n m _ $9.52 an?» o: 00. 00 on on 9v an 4 I P . p . - P . p L p k nu .: . . kuaa . ~s fi 9 II n \\\ let u \\ v N4 IdIIIO o s .o u. 6.. .1 n.4 ..N. “u . 3 :49. —z A o .62 £3... 4 n «05‘! Av 'o At “0 zo_h(z_._.:4u0>mm£ pom unmfla mo mmnsoo one .1 FIG...— II I 2_Z.PDI_UO< >>>(o mm mm .N V. 05 Mo on NV NV .0 mnsmflm C) 5+ doom 6 m4 doom s4 Hoom o a [III/[III]. ./. ./. 0.7 QM . v v T v V v r V V V V V T V ‘ T V i V 38L”. NOIIfl'IIO 901 V v v v v v V V V 75 This could be explained on the basis of the existence of two different antibody types or it may reflect a difference in the sensitivity of the tests. The location of RLF is shown in Figure 2 and in Table 15. All of the activity was located in the macro- globulin peak. These data support previously presented data which indicated the macroglobulin nature of RLF. » 1 Figure 2 and Table 8 show the position of the anti-BSA 5 antibody and indicate that at one week post inoculation its molecular weight is below 200,000. These data are in con— flict with those presented by Orlans et a1. (1961) who reported two different molecular weight antibodies at this time. Experiments on heat inaCtivation of heavy and light agglutinins were performed. It was learned that neither of these antibodies were affected by heating at 56 C for 30 min. The titres remained the same except in two cases where a one tube drop was observed. This was not considered significant. If the heated material was diluted in 1:2 normal serum the titre dropped as before to 1:5 in cases where the original titre was high and to 0 where the original titre was 1:20 or lower. Dilution of unheated material from both peaks in 1:2 normal serum caused a great 76 increase in titre as before. Samples were not diluted past 1:640 but most reactions were still positive at this point. On the basis of these results no differential heat sensitivity or differential dependence on titre increasing factor could be shown. Immunodiffusion Analysis 9: Immune Serums Fractionated with Sephadex S 200 Although the experiments with Sephadex permitted the conclusion that at least three different types of antibody were present in serum from chickens infected with M. gallinarum more information was desired concerning these differences. A difference of opinion from that of Orlans et a1. (1961) was also posed through the finding that although both high and low molecular weight antibodies were present against PPLO only low molecular weight antibodies could be found which were specific for BSA. This latter finding could be used to good advantage in gel diffusion tests. Any specific gamma globulin present in soluble complexes would be of low molecular weight yet would probably cross react with high molecular weight antibodies of the anti—PPLO series. It should be possible then to pick out at least two types of macroglobulins which cross react with the 7S anti—BSA antibody. 77 Sephadex derived macroglobulin showed PPLO aggluti- nating activity. This material was precipitated twice by the euglobulin method. Both the concentrated supernatant fluid and the dissolved precipitate also showed PPLO agglutinating activity. Figure 7 shows the results of cellulose acetate electrophoresis done on each preparation. As can be seen, an alpha 2 M band is present in preparations a, b and c but is absent in the euglobulin preparations. The major areas of protein concentration in preparations d and e were in the beta and gamma regions. These areas also occur in preparations b and c. Although it does not appear that all of the beta and gamma globulins can be separated by euglobulin precipi- tation, alpha 2 macroglobulin is shown to be removed by this treatment. Immunodiffusion analyses performed on the soluble complexes and the Sephadex column fractions are shown in Figure 8. As a basis for comparison two soluble complexes were formed by dissolving specific BSA anti-BSA precipitates in 100 X optimal proportions of BSA. One such soluble complex, labeled SCx, was formed from the second major peak of the Sephadex preparation shown in Figure 2. By doing this it is assured that not only will the second soluble complex a. b. C. d. e. 78 Normal serum control (6 lambda sample) Macroglobulin peak (Pool # 3) (10 lambda sample) Supernatant fluid of euglobulin precipitation of Pool # 3 concentrated 4 X (10 lambda sample) First euglobulin precipitate from Pool # 3 (10 lambda sample) + .9 ‘ lI—- Second euglobulin precipitate from Pool # 3 derived euglobulin (10 lambda sample) Figure 7. Cellulose acetate electrophoresis of Sephadex prepared chicken macroglobulins. 79 Well #1 Well #2 Well #3 Well #4 Well #5 Well R Anti-BSA soluble complex Sephadex peak II anti-BSA soluble complex Anti-PPLO macroglobulin (Sephadex peak I) Macroeuglobulin precipitate Euglobulin precipitation supernatant and unnumbered wells rabbit anti-whole chicken serum Figure 8. Double diffusion analysis of Sephadex prepared chicken macroglobulins and anti-BSA soluble complexes. 80 (SSCx) be free of macroglobulin antibody not detected in the interfacial testing, but it will also be reasonably macro— globulin-free from the standpoint that the Co P of Makinodan et al. (1960) will not be present. The photographic reproductions seen in Figure 8 are contact prints made from the original plates and appear here in their original dimensions. Success in removing the majority of the alpha 2 M can be judged by observing Figure 8a, b, and c. Wells marked 3 and 5 contained the untreated pool and the supernatant fluid from the euglobulin precipitation respectively. Both contain an antigen which shows an identity reaction with an antigen in wells 1, 2 and 4. This line represents the alpha 2 M and is marked with a pointer in Figure 8a. Two other proteins of unknown origin and nature exist in the macroglobulin preparation which do not occur in the soluble complexes. These are pointed out in Figure 8c in the right #5 well. This figure also shows that the alpha 2 M has not been completely removed from #4 (pointer at bottom #4 well). The findings of Makinodan et a1. (1960) concerning the homogeneity of the coprecipitating macroglobulin line are questioned in light of the results reported herein. The soluble complexes showed at least six distinct lines rather than the reported two. SCx (well #1) formed what 81 appears to be a single line close to the antibody well. This line split (Figure 8b top well #1) to reveal a doublet of identity with SSCx (well #2) in the same figure. This is the line shown to be pure gamma globulin in Figure 4e. The answer as to whether both are specific antibodies is not available at this time. Certainly both are 78 globulins since they occur as major components in the #2 Sephadex peak and do play a role in specific precipitation since they are found in the soluble complex. Lines of identity are formed with elements of the macroglobulin peak which indi— cates that molecules in those peaks have similar antigens as part of their total antigenic schema. It will also be noticed that a line of partial identity is formed between the component closest to the antibody well in Figure 8b bottom #3 and the inner line of the gamma doublet of adjacent #2 well. This indicates that all of the antigens of the light gamma are present in this heavy molecules plus some additional ones not present in the lighter molecules. very small amounts of the two components previously identified by Makinodan et a1. (1960) as Co P are present in the SSCx as compared to the SCx. Both components are macroglobulins as shown by their diminished concentration in the #2 wells of Figure 8. At least one of these Co P components is present in sufficient concentration in the 82 Sephadex material to be visible after much initial loss through the previously mentioned non—specific precipitation which occurred after performing the Sephadex separation. Two unidentified lines occur in SCx consistently and are shown in Figure 8 close to the antigen wells. The line nearest the well is very faint. These lines are identi— fied by pointers in Figure 8b. DISCUSSION The results of the experiments reported here suggest that a close correlation exists between specific precipi- tation and agglutination reactions and those reactions generally considered to be non-specific. It was learned that using the ammonium sulfate method for partial purification of immune globulins resulted in a decrease of the anti-PPLO agglutinin titre. This loss of titre may be due to the loss of some essential component during the purification process. It was demonstrated that a titre enhancing substance which was present in normal whole serum was lost upon dilution of the normal whole serum to dilutions greater than 1:4. It may also be that a loss of immune globulins occurred during the fractionation process. Likewise, these partially purified immune globulins were not as sensitive to heat as whole immune serum as evidenced by the failure of heat to cause a drop in titre. Reduction in titre also occurred when calcium was removed from whole immune serum with EDTA. In cases where the antibody level was initially low, EDTA decreased the titre proportionately more than when the titre was initially 83 84 high. The action of titre enhancing substance required in low titred serum is lost upon removal of calcium with EDTA since calcium is required for its action. These results, however, cannot be used to explain theloss of titre following treatment of high titred serum with EDTA since the titre enhancing substance is not effective at dilutions greater than 1:4. At present there is no satisfactory explanation for this observation. It is possible that the specific antibody-antigen reaction has a divalent cation requirement. Adding normal whole serum to immune serum which has been absorbed with aggregated gamma globulin re—establishes the original titre in unheated samples. The addition of whole normal serum reduces the titre in absorbed, heated samples. These findings allow the conclusion that aggregated gamma globulin removes the factor which is responsible for precipitating the immune globulin in heated immune serum. Heating immune globulin alone did not greatly decrease its ability to react. The same immune globulin if aggregated is precipitated by some factor in normal serum. It can be concluded that this is probably what is happening when one inactivates avian serum. Such a two step reaction can best be seen by the following hypothetical formulae: 1. Immune globulin + heat ————————> aggregated immune globulin. 85 2. Aggregated immune globulin + RLF ———————4> precipitate. In order to interpret these results with complete confidence it would be necessary to present reasonable assurance that a protein had been prepared in a high state of purity which both precipitated aggregated gamma globulin and caused an increase in the titre of a highly purified immune gamma globulin. This has not been done. Attempts to purify the RLF component of normal serum met with failure due to an inability to keep the RLF in soluble form. As was suggested earlier, the use of EDTA may offer a solution to this problem. Much indirect evidence has been accumulated, however, which points to the possibility that RLF, titre increasing substance and Co P are the same. Absorption of immune serum with aggregated normal gamma globulin both decreased the titre and abolished the RLF reaction. Calcium depletion with EDTA drastically reduced the titre of initially weak antiserum and eliminated the RLF reaction. All of the above procedures except the EDTA treatment, prevented incorporation of Co P into specific precipitates. Makinodan et al. (1960), however, have reported that incorporation of Co P can be blocked with EDTA when initial antibody titres are low. Immunodiffusion techniques have consistently showed the 86 presence of Co P in both RLF and cryoglobulin precipitates. It would appear, therefore, that Co P, RLF and titre in- creasing substance are different names for the same entity. The stated hypothesis has some points of conflict, however. Experiments reported herein have shown that titre increasing substance and Co P are active in fresh serum while RLF does not appear until after storage at lower temperatures for at least 24 hr. It is possible that the type of aggregation of gamma globulin may be quite different in the case of heating and a specific immune reaction. Specificity most likely is lost by a precursor of RLF upon storage. Heating serums to cause aggregation was an attempt at: simulating the aggregation which occurs during specific immune reactions. The factor referred to as RLF, Co P and titre in- creasing substance could also be called C'l-like-substance since it has many of the characteristics of avian C'l. Benson et a1. (1961), referred to the ability of kaolin to absorb out C'l activity and to the beta 2 euglobulin nature of avian C'l. They also reported that C'l seemed to consist of two distinct molecular types. The fact that RLF is not as labile as C'l does not detract from the possible identity of the two factors. Mauer and weigle (1953) reported that non-haemolytic complement definitely adds to immune 87 precipitates. Since the results obtained in these experiments have not been unconditionally attributed to C'l and there is a reaction similar to the rheumatoid reaction the term RLF will be retained. The total significance of the cryoglobulin reaction is not known at the present time. Immunodiffusion analysis has shown, however, that the components of this precipitate and the components of both soluble complexes and solubilized RLF are nearly identical. It is possible that changes in configuration of gamma globulin trigger this reaction since the presence of aggregated gamma globulin significantly shortens the time of appearance of the precipitate. Naturally occurring cryoglobulin contains gamma globulin and RLF in addition to another yet uncharacterized fraction. The finding that EDTA completely inhibits cryoglobulin formation at all temperatures tested also adds weight to the above statement. The presence of gamma globulin in cryoglobulin does more than merely suggest a mechanism for its formation. It constitutes additional contradictory information. If gamma globulin disappears from serum in the form of cryoglobulin then the titre should fall rather than rise upon ageing. Three answers can be suggested as possible explanations for this paradox. Gamma globulin may exist in several forms, 88 one or more of which may be incorporated in cryoglobulin while others are not. In such a case no drop in titre would necessarily be observed. The plausibility of the multiple antibody explanation has already been demonstrated. A rise in titre could be accounted for by the appearance of RLF in a more reactive form than is present in fresh serum. Another possibility is the presence of an inhibitor in fresh serum which is taken out during cryoglobulin formation or it is denatured allowing cryoglobulin formation and the titre increasing phenomenon to occur. This possibility was tested to a limited extent with the result that no such inhibitor was found in fresh serum which inhibited titre increasing factor. The data collected on this were not reported, how- ever, due to the limited number of experiments done. In view of the results obtained it would seem that for an accurate determination of agglutination titre several modifications should be made where avian serum is being evaluated. If a great majority of nonagglutinating co- factor dependent antibodies are present, then the titre will be dependent upon cofactor concentration and hence the degree of ageing of the serum. For this reason a strict protocol with regard to ageing should be observed and dilutions should be made in normal serum or a suitable dilution thereof. At no time should the immune serum be inactivated by heat or 89 by other methods. An attempt should always be made to determine the degree of dependence of the particular system on RLF. This can be done by running a control in which dilutions are made in normal serum. Should it be necessary to hold serum for any long period of time, the addition of EDTA would provide protection against the possibility that antibody would be lost in cryo- globulin. Calcium can be added upon thawing and titreing provided that the test does not require incubation at 4 C. Attempts to show differences in the HI and aggluti- nating antibodies met with only limited success. On the basis of the curves seen in Figure 6 it can be said that two different groups of antibodies exist as regard molecular weight. This would be expected in the light of similar findings in rabbit antiserum by Bauer and Stavitsky (1961). Within the high molecular weight group the HI and aggluti- nating antibodies are different as can be seen from their different slopes. The distinction is not as clear in the light weight group. No discernable difference in the shape of the curves can be seen, but this does not mean that they are not different. No difference in the light or heavy agglutinins could be seen as regards dependence on RLF or aggregation ability. The fact that heavy HI antibodies con— tinue to persist cannot be used to adequately explain the 90 difference in heat sensitivity of the two antibodies but does suggest that differences do exist which might also apply to the light species. Sephadex chromatography of anti-BSA antibody did not reveal the presence of a high molecular weight antibody to BSA in the serum sample used. This observation agrees with the most recent ideas of Orlans (1962) who retracted her original statement that heavy precipitins occur in chicken anti-BSA serum. It now seems that the macroglobulin which she reported cannot form specific precipitates by itself, but requires the reaction of specific 78 gamma globulin first. Of even greater importance are the reports by Orlans et a1. (1961) and Orlans (1962) of non-precipitating 7S gamma globulin. Two distinct types of 7S globulin do in fact exist, one of which could be responsible for HI and the other for agglutination. The immunodiffusion analysis of Sephadex fractions derived from anti-BSA and anti-PPLO serums showed that a macroglobulin did exist in anti-PPLO serum. It did cross react with the 7S anti-BSA gamma globulin. It cannot be said, however, which of the two anti-BSA components of the gamma doublet was reacting with which of the two anti-PPLO macroglobulin antibodies represented in the first major Sephadex peak. This cross reaction appears to be one of 91 partial identity in Figure 8b where #3 is compared to #1 and #2 but appears in c as a single line. If it is kept in mind that this line is really a doublet (easily seen in Figure 8a #3) it will be apparent that the line is in reality showing a reaction of identity. SUMMARY Heat aggregated chicken anti-PPLO agglutinins were precipitated from serum by a rheumatoid—like-factor (RLF) found in the serum of normal chickens of all ages. This accounted for the loss of titre witnessed after inactivating immune serum for 10 min at 56 C. The factor, which resembles human rheumatoid factor in its action and avian C'l with respect to many of its physical properties, was found to be a beta 2 macroeuglobulin which made up 2.4%. of the serum protein in chicken serum. It was absorbed out with kaolin and aggregated chicken gamma globulin and could be inhibited by 0.0025 M EDTA, 0.150 M CaCl or 0.1 M 2 mercaptoethanol. 2 The RLF macroglobulin was partially purified by gel filtration with Sephadex G 200 but proved to be very unstable in this state suggesting that it may have been separated from some stabilizing factor present in another fraction of the serum. Continuous flow paper curtain electrophoresis of RLF active euglobulin precipitates yielded a beta globulin which cross reacted with one of the bands of a solubilized RLF precipi— tate in gel diffusion but showed no RLF activity. Fresh normal serum, if not diluted more than 1:4 caused an increase in the titres of anti-PPLO and antifiS. 92 93 pullorum serums. Aged normal serum showed both titre in- creasing ability and RLF activity. The latter disappeared upon dilution to 1:8. The two phenomena were considered to be related through the RLF. It was felt that aggregating gamma globulin did not mimic the specific antibody-antigen reaction closely enough when the factor responsible for precipitation of modified gamma globulin was in its native state. Upon conversion to RLF, however, some specificity was lost and the RLF reaction occurred. It was shown that aggregated gamma globulin exerted a stimulating effect on the conversion of a precursor substance to RLF and that slow aggregation of gamma globulin in the cold might have been responsible for cryoglobulin formation. Cryoglobulin formation and the RLF reaction are related as exhibited by the similarity of their precipitate components as shown in gel diffusion. EDTA was shown to completely inhibit cryoglobulin formation at all temperatures tested. It was shown that cryoglobulin did not form at room temperature. The co—precipitating beta 2 macroglobulin demon- strated by Makinodan et a1. (1960) in soluble complexes was shown to exist in RLF precipitates along with three other components not reported by these authors. Makinodan's Co P was absorbed out by aggregated gamma globulin and kaolin. 94 Co P was also prevented from being incorporated into specific precipitates by 2 mercaptoethanol. It was therefore con- cluded that Co P and RLF are identical. Gel filtration, using Sephadex G 200, of serum from chickens infected with PPLO for varying periods of time revealed that at least three distinct types of anti-PPLO antibodies were present in the serum in the course of the disease. A pair of macroglobulins, a HI antibody and an agglutinin, were different from those molecules of the light antibody group responsible for HI and agglutination. Whether the light HI antibody differs from the light agglutinin remains to be determined. Neither the light nor the heavy agglutinin differed in their dependence on RLF or their ability to be aggregated. No antigenic differences between the light and heavy antibodies could be detected with the methods used in this study. BIBLIOGRAPHY Adler, H. E. 1953. Preliminary report on the slide aggluti- nation test for pleuropneumonia-like agents associated with infectious sinusitis of turkeys and chronic respiratory disease of chickens. S. W. Vet. 'S:362—363. Aitken, D., and W. Mulligan. 1962. Quantitative precipitin studies on foul antisera to bovine serum albumin and bovine gamma globulin. Immunology S:295-305. Allison, A. C., and J. H. Humphrey. 1960. A theoretical and experimental analysis of double diffusion precip- itin reactions in gels and its application to character- ization of antigens. Immunology S:95-106. Banovitz, J., S. J. Singer, and H. R. WOlfe. 1959. Precipitin production in chickens. XVIII Physical and chemical studies on complexes of bovine serum albumin and its chicken antibodies. J. Immunol. .Sg:481—488. Banovitz, J., and H. R. Wolfe. 1959. Precipitin production in chickens. XIX The components of chicken anti- serum involved in the precipitin reaction. J. Immunol.-Sg:489-496. Barile, M. F., and D. B. Riggs. 1961. Immunofluorescence of PPLO in tissue cultures. Bact. Proc. 83. Benson, H. N., H. P. Brumfield, and B. S. Pomeroy. 1961. Requirement of avian C'l for fixation of guinea pig complement by avian antibody antigen complexes. J. Immunol. Sl:6l6—622. Brumfield, H. P., and B. S. Pomeroy. 1957. Direct complement fixation by turkey and chicken serum in viral systems. Proc. Soc. Exptl. Biol. Med. 23:146-149. Brumfield, H. P., and B. S. Pomeroy. 1959. Test based on normal serum complement implementing fixation of complement by turkey antiserum. Proc. Soc. Exptl. Biol. Med. 102:278—280. 95 96 Bushnell, L. D., and C. B. Hudson. 1927. Complement fix— ation and agglutination tests for Salmonella pullorum infection. J. Infect. Dis. 3;:388-394. Cover, M. S., W. J. Benton, and M. A, Whelan. 1960. The thermostability of chicken serum to be used in the PPLO agglutination test. Ann. N. Y. Acad. Sci. ZS: Art. 10:567-574. Crawley, J. F. 1960. Use of the hemmaglutination inhibition test in the control of chronic respiratory disease of chidkens. Ann. N. Y. Acad. Sci. ZS: Art. 10: 562-566. Crowle, A. J. 1961. Immunodiffusion. Academic Press New York and London. Delaplane, J. P., and H. 0. Stuart. 1943. The propagation of a virus in embryonated chicken eggs causing a chronic respiratory disease of chickens. Am. J. Vet. Res. ‘3:325-332. Deutsch, H. F., J. C. Nichol, and M. Cohn. 1949. Biophysical studies of blood plasma proteins. XI. Immunological and electrophoretic studies of immune chicken serum. J. Immunol. SS:195-210. Deutsch, H. F., and J. I. Morton. 1957. Dissociation of human serum macroglobulins. Science. 125:600-601. Franklin, E. C. 1962. Some physiochemical properties of high molecular weight gamma and beta globulins in chicken sera. Proc. Soc. Exptl. Biol. Med. I1QS:338—342. Gengozian, N., and H. R. Wolfe. 1957. Precipitin production in chickens XV. The effect of varying concentrations of NaCl on precipitate formation. J. Immunol. 'ZS:401-408. Gengozian, N., T. Makinodan, and R. R. Carter. 1962. Analysis of immune precipitates by serum agar diffusion. J. Immundl. .SS:426-434. Goldberg, C. A. J. 1959. A discontinuous buffer system for paper electrophoresis of human haemoglobin. Clin. Chem. S:446-452. 97 Goodman, M., H. R. Wolfe, and S. Morton. 1951. Precipitin production in chickens. VI The effect of varying concentration of NaCl on precipitate formation. J. Immunol. SS:225-236. Goodman, M., H. R. Wolfe, and R. Goldberg. 1954. Precipitin production in chickens. XIII The effects of variation in ionic species and concentration on precipitate formation. J. Immunol. “1;:440-445. Goodman, M., and D. S. Ramsay. 1957. Specificity of the reaction of chicken antiserum in high sodium chloride concentrations. Fed. Proc. .lé‘416- Heidelberger, M., and K. O. Pedersen. 1937. The molecular weight of antibodies. J. Exptl. Med. 'SS:393-4l4. Hirschield, J. 1960. Immunoelectrophoresis procedure and application to the study of group specific variations in sera. Science Tools .1:18-25. Jacobs, R. E., E. L. Jungherr, R. E. Lugenbuhl, and E. Gianforte. 1953. Serologic studies in air sac infection. Proc. 25th Annual Pullorum Conference. Amherst, Mass., 3. Jungherr, E. L., R. E. Lugenbuhl, and R. E. Jacobs. 1953. Pathology and serology of air sac infection. Proc. Kabat, E. A,, and M. M. Mayer. 1961. Experimental Immunochemsitry. Charles C. Thomas, Springfield 2nd ed. Levine, L., K. M, Cowan, A, G. Osler, and M. M. Mayer. 1953. Studies on the role of Ca and Mg in complement fixation and immune haemolysis. I. Uptake of complement nitrogen by specific precipitates and its inhibition b eth lene-diamine-tetra-acetate. J. Immunol. 1;: 59-3 6. Makinodan, T., N. Gengozian, and R. E. Canning. 1960. Demonstration of a normal serum macroglobulin co- precipitating with the bovine serum albumin (BSA)— chicken anti-BSA system. J. Immunol. SS:439-446. Maurer, P. H., and W; Weigle. 1953. The persistence of complement in aged sera. J. Immunol. 1;:284-295. 98 Orlans, E., M. E. Rose, and J. R. Marrack. 1961. Foul antibody I. Some physical and immunochemical properties. Immunology 4:262-277. Orlans, E. 1962. Foul antibody II. The composition of specific precipitates found by antisera to serum albumin, haemoglobin and myoglobin and some properties of non-precipitating antibody. Immunology .S:306-321. Rheins, M. S., E. C. McCoy, E. V. Buehler, and E. V. Burrell. 1957. Effects of animal sera and serum albumin on latex fixation test for rheumatoid arthritis. Proc. Soc. Exptl. Biol. Med. SS:67-7l. Rice, C. E. 1947. Atypical behavior of certain avian anti- sera in complement fixation tests. Canad. J. Comp. Med..ll:236-245. Schmidt, M. H., and H. R. Wolfe. 1953. Precipitin production in chickens IX A quantitative study of the antibody response to nine different purified substances. J. Immunol. 1;:214-219. Singer, J. M., and C. M. Plotz. 1956. A latex fixation test. I Application to the serologic diagnosis of rheumatoid arthritis. Am.J. Med. gl:888-892. Smith, Ivor. 1960. Chromatographic and electrophoretic techniques. V61. II Zone electrophoresis. Interscience Publishers, Inc., New York. Van Herick, W., and M. D. Eaton. 1945. An unidentified pleuropneumonia-like-organism isolated during passages in chick embryos. J. Bacterol. SQ:47-55. Weigle, W. O., and P. H. Maurer. 1957. The effect of complements on soluble antigen antibody complexes. J. Immunol. 12:211-222. White, F. H., G. I. Wallace, and J. O. Alberts. 1954. Serological and electron microscope studies of chronic respiratory disease agent of chickens and of turkey sinusitis agent. Poult. Sci. Sg:500-507. Wolfe, Wolfe, WOlfe: Wolfe, Wolfe, 99 H. R. 1942. Precipitin production in chickens I. Interfacial titres as affected by quantity of antigen injected and ageing of antisera. J. Immunol. 43:135-145. H. R., and E. Dilks. 1946. Precipitin production in chickens II. Studies on the in vitro rise of the interfacial titres and the formation of precipitins. J. Immunol. Sg:331-341. H. R., and E. Dilks. 1948. Precipitin production in chickens III. The variation in the antibody response as correlated with the age of the animal. J. Immunol. SS:245—250. H. R., and E. Dilks, 1949. Precipitin production in chickens IV. A comparison of the antibody response of eight avian species. J. Immunol. .SQ:251-257. H. R., A. P. Mueller, and J. C. Neess. 1959. Physical factors affecting maximum precipitation of the BSAranti-BSA foul system. Immunology .g:195—202. {luv-u UJL L.,:IIL‘I