nigh-— Lxul.' (A; AL; 14141 11 MACROI’HAGE CYTOPHILIC ANTIBODY T0 SALMONELLA TYPHIMURIUM Theda Tor flue Degree of M. S. MICHIGAN STATE UNIVERSITY Charles W. Stone I976 ABSTRACT MACROPHAGE CYTOPHILIC ANTIBODY TO SALMONELLA TYPHIMURIUM BY Charles W. Stone A new and more quantitative technique to study macro- phage cytoPhilic antibody is described. It essentially consists of incubation of sensitized cells with antigen in suspension followed by millipore filtration. The filtrate and filter fractions are then counted separately giving the percent of antigen bound (filter) and unbound (filtrate). This technique was used to study the role of cytophilic antibody in Salmonella typhimurium adherence to peritoneal cells. Bacterial binding was significantly increased in the presence of immune rat serum. Studies to differentiate opsonic from cytophilic properties were of only limited success, but they did confirm previous findings that a large portion of the immunoglobulin fraction had opsonic proper-‘ ties and that adsorption with large numbers of macrophages (~1.5 X 107) did not remove a significant amount is cyto- philic activity. Experiments designed to measure the cytOphilic binding of free endotoxin by sensitized peri- toneal cells showed very low levels of binding (<0.S%). Charles W. Stone This study was the first attempt to demonstrate the binding potential of free endotoxin to cytOphilic antibody. Interspecies studies using rat serum and mouse peritoneal cells showed evidence suggestive of a higher level of bind- ing with mouse cells than rat cells. This was the first interspecies study between the mouse and the rat with regard to cytophilic activity. It was also the first interspecies study of cytOphilic activity with gram negative bacteria. Cumulatively, these data show that macrophage cytophilic antibody increases the rate of phagocytosis of gram negative bacteria by increasing adherence to macrophages. This is one of several ways in which macrophage cytophilic antibody exerts its antibacterial action. MACROPHAGE CYTOPHILIC ANTIBODY TO SALMONELLA TYPHIMURIUM B y. NOS“ Charles W.\Stone 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 1976 DEDICATION To my parents, Paul W. and Sara C. Stone for their help and encouragement. ii ACKNOWLEDGEMENTS I would like to express my appreciation for his assis- tance to Dr. Robert J. Moon, my research advisor. I would also like to extend my appreciation to Dr. Norman B. McCullough for his interest and assistance. A number of people in our research group have provided helpful sugges- tions and discussion. They are: Richard L. Friedman, Richard C. Sawyer, James Zyldaszyk, Joyce Broka and Ruth Vrable. iii TABLE OF CONTENTS INTRODUCTION ................. . ...... . ................ LITERATURE REVIEW...................... ...... ........ Historical Background ...................... ..... MATERIALS AND METHODS ................................ AnimaISOOOOO0.0.00.0.0... 00000000000000 0.0.0.... Bacteria... ...... ............................... Macrophage Preparation. ..... ...... ..... ......... Heat-killed Cells............ ...... . ....... ..... Immune Serum ........ ....... ......... ............ Absorbed Serum ..... . ................ ............ Normal Serum. ........................ .......... Radiolabeled Endotoxin Preparation... .......... 51Cr- Labeled Endotoxin- -Macrophage Experiments... Statistics. . .... ......... ...... ............ ... RESULTS ................... ... .............. . ..... .... Characterization of Peritoneal Cell Population and Serum Tube Agglutination Titers.......... Experimental Apparatus Control.................. Effects of M-199, M-199-Normal Rat Serum and M-199-Immune Rat Serum on the Ability of Rat Peritoneal Cells to Bind Viable Salmonella typhimurium After Five Minutes. . ......... Effects of M- 199- Heat killed Cell Absorbed Normal Rat and M- 199 Heat- killed Cell Absorbed Immune Rat Serum on the Ability of Peritoneal Cells to Bind Viable Salmonella typhimurium SR11....... ............... ...... Effect of Macrophage Absorbed Immune Serum on the Ability of Peritoneal Cells to Bind Viable Salmonella typhimurium SRll After Five Minutes.............................. ..... ... iv Page 47 49 49 51 53 TABLE OF CONTENTSnncontinued Effects of Mq199, M«199~Normal Rat Serum and M~199-Immune Rat Serum on the Ability of Rat Peritoneal Cells to Bind 5 Cr—labeled Salmonella typhimurium Endotoxin After Five Minutes ................................... Effects of M-199nNormal and M 199 Immune Rat Serum on the Ability of Mouse Peritoneal Cells to Bind Viable Salmonella typhimurium SRll After Five Minutes and Three Hours. ..... DISCUSSION ..................................... ...... BIBLIOGRAPHY ................... . ..................... Page 54 55 57 67 LIST OF TABLES TABLE Page 1. Tube agglutination titers ....................... 47 2. Peritoneal cell counts .......................... 48 3. Comparison of the effect of M-199, normal rat serum, and immune rat serum on the ability of rat peritoneal cells to bind viable Salmonella typhimurium after five minutes .................. 50 4. Comparison of the effect of absorbed and non- absorbed normal rat serum and immune rat serum on the ability of peritoneal cells to bind viable Salmonella typhimurium .............. ..... 52 5. Comparison of the effect of immune serum and macrophage absorbed immune serum sensitized peritoneal cells to bind Salmonella typhimurium after five minutes .............................. 53 6. Comparison of the effect of M-199, M-199-normal rat serum and M-199-immune rat serum on the ability of rat peritoneal cells to bind 51Cr- labeled endotoxin after five minutes ............ S4 7. Comparison of normal and immune rat serum on the ability of mouse peritoneal cells to bind viable Salmonella typhimurium after five minutes and three hours ......................... 55 vi INTRODUCTION MacrOphage cytOphilic antibodies have been demonstrated in an association with a number of immunologic phenomenon (85,107). However, no clear—cut role has been ascribed to them. Part of the reason for this is that there is a consid- erable amount of confusion and controversy concerning the significance of their immunologic action. Another diffi- culty is that the usual methods used to measure them are semiquantitative at best (10,18,31). Macrophage cytophilic antibody given passively or present on sensitized macrophages confers protection to Salmonella typhimurium in mice (95). A similar macrophage cytOphilic antibody has been found in antisera to Escherichia coli and Vibrio cholerae (1, 81). This thesis describes a more quantitative technique to study macrophage cytOphilic antibody in vitro using viable Salmonella typhimurium SRll and S. typhimurium endotoxin as antigens. In addition, a careful analysis of macrophage cytophilic antibody in general, its immunologic character- istics and its relation to phagocytosis and resistance to infection was made. Hopefully, this review has clarified at least some of the confusion and controversy surrounding macrophage cytophilic antibody. Using this experimental model, we first tested the ability of rat peritoneal cells sensitized with normal and immune rat serum to bind viable S. typhimurium. Second, normal and immune rat serum were absorbed with heat'killed S. typhimurium and the ability of this sera to sensitize rat peritoneal cells was examined. Third, immune rat serum absorbed with macrophages was tested for its ability to sen— sitize rat peritoneal cells to bind viable S. typhimurium. These later experiments were performed in an attempt to dif— ferentiate the opsonic and cytophilic properties of immune serum. Fourth, the ability of normal and immune rat serum to sensitive rat peritoneal cells to bind S. typhimurium endotoxin was tested. The purpose for this study was to determine the role of macrOphage cytophilic antibody in the removal of free endotoxin in vitro. Fifth, the ability of normal and immune rat serum to sensitize mouse peritoneal cells to bind viable S. typhimurium was evaluated to test the interspecies potential of macrophage cytophilic anti“ bodies to gram negative organisms. LITERATURE REVIEW Historical Background The first relatively clear indication of the existence of macrophage cytophilic antibody was reported over 20 years ago. In 1954 Girard and Murray (40) detected a large amount of anti-Salmonella antibody associated with thoracic cavity macrOphages occurring in inflammation caused by gum- arabic beef extract. Passively administered antibody appeared to be concentrated on or in macrophages. There had been some suggestion of this phenomenon in earlier work by Lurie (73) on the deveIOpment of immunity to tuberculosis. Zinsser and Mueller (119) also found evidence suggesting macrophage-associated antibody in their work on the nature of bacterial allergies. Boyden and Sorkin (18) observed that antibodies in cer- tain rabbit antisera to human serum albumin (HSA) fixed to normal rabbit spleen cells. After washing, the cells were able to specifically adsorb 131 I-HSA. Boyden (16,17) later demonstrated antibodies specifically cytophilic for macro— phages. Other investigators reported a similar phenomenon in different systems (1,10,89,113). CytOphilic antibody was defined by Boyden (16) as "a globulin component of immune serum which becomes attached in vitro to certain cells in such a way that these cells are subsequently cap- able of specifically adsorbing antigen." Investigation of macrophage cytophilic antibody was greatly facilitated by the development of several techniques to detect them in vitro. Boyden (16,17) introduced a rosette—forming reaction on macrOphage monolayers usually referred to as the passive direct technique. Macrophages are allowed to attach to glass and form monolayers and washed. The monolayers are then sensitized with antiserum for a sufficient period of time and washed again. Antigenic particles, erythrocytes, proteins, bacteria, etc., are then incubated with the sensitized cells and washed to remove unbound antigen. Antigen-macrophage rosettes can then be observed and counted by microscope. This technique has several limitations: 1) sensitized cells have to be attached to glass, 2) the counting method is laborious and only semiquantitative, and 3) it does not exclude attachment of antigen to glass. Berken and Benacerraf (10) introduced the passive indirect technique. This procedure basically consists of sensitization of antigen with antiserum. The antigen-antibody complexes are then added to a macrophage monolayer. The monolayer is observed for rosette-formation as in the passive direct technique. A major drawback with this method is that it essentially measures the opsonizing activity of antiserum, not its cytophilic activity. Jonas et al. (63) introduced an alternative assay method, the suspension-centrifugation technique. Antigen is mixed with passively sensitized cells in suspension for a sufficient period of time and then gently centrifuged. The resuspended pellet is examined microscopically for attached antigen particles. This technique does not require attachment to glass and is thought to be more sensitive. The present study uses a more quantitative modification of the suspension-centrifugation technique somewhat similar to the method described by Del Guercio et al. (31). These authors sensitized peritoneal cells which were suspended in phosphate buffer with serial ten-fold dilutions of guinea pig antisera to S. typhimurium at 20C for two hours. The cells were centrifuged, resuspended in phosphate buffer, and radiolabeled S. typhimurium was added in a ratio of 200 bacteria per macrophage. The bacteria-macrophage mixture was incubated overnight at 4C. The cells were centrifuged and the radioactivity measured in each fraction. Several other techniques have been used to measure cyto— philic activity and are of interest in this study. Brumfitt et al. (19) studied the attachment of Escherichia coli to rat peritoneal macrophages. They used low temperature (9C, 20C), antibiotics in media and iodo- acetate to selectively study attachment. Macrophages were cultured in vitro with M—199 and the fate of E. coli was studied using streptomycin or neomycin in the medium. This procedure was thought to separate extracellular from intra- cellular growth and killing. Only extracellular organisms were killed and intracellular events could be followed without the complexities introduced by extracellular growth. Both reducing ambient temperature and iodoacetate decreased phagocytosis and a greater proportion of the bacterial popu- lation was destroyed under these conditions, implying a greater extracellular population. Auzins and Rowley (6) described a technique which sep- arated free bacteria from those attached to suspensions of mouse peritoneal macrophages. The bacteria, Salmonella typhimurium C5 were preOpsonized with fresh pig serum. Macrophages were suspended in Hank's buffered saline solu- tion in a siliconized two milliter syringe and a number of preopsonized S. typhimurium C5 were added at an approximate ratio of bacteria to macrophage of 5:1. Pieces of Whatman No. 1 filter paper were fitted exactly into the hub of a 25 gauge needle. Three drOps of the macrophage—bacteria mixture were passed through the filter at each sampling at which time the filter was changed and the needle cleaned and dryed. The rate of association of opsonized and un— opsonized bacteria to peritoneal macrOphages was studied at 20C and 37C. Levenson and Braude (72) studied the cytophilic proper- ties of "late" and "early" antisera to Salmonella typhi Vi antigen using the passive indirect technique. The test antigen was S. typhi Vi adsorbed on erythrocytes. This method makes it possible to follow the temporal changes in cytophilic activity using a monospecific bacterial antigen without the complexities of viable bacteria. Hsu and Mayo (53) described a method for detecting macrophage cytophilic antibody that is different both in the phenomenon observed and the functional significance. Macrophages with cytOphilic antibody were found to clump in vitro in the presence of antigen. This phenomenon was species specific, could be elicited by either the O or H antigens of S. typhimurium SRll and was thought to represent an enhancement of antibacterial immunity. Vlaovic et al. (115) described a technique differen— tiating the leukocyte migration inhibition due to cyto— philic antibody from that due to cell-mediated immunity by a modification of the capillary tube method. There has been considerable controversy concerning the definition and mechanism of action of macrophage-cytOphilic antibodies as those antibodies that attach to macrOphages after antigen-antibody complexes form. These antibodies are detected by the passive indirect method which involves mixture of antigen and antibody and washing the antigen- antibody complexes formed. MacrOphage are then "sensitized" with the antigen-antibody complexes and examined micro- scopically. Rosette cells formed by this method are morphOu logically the same as those formed by the macrophage— monolayer techniques. Berken and Benacerraf (11) define cytophilia as "that property of opsonizing antibody which provides the receptors that permit the binding of the anti“ body to the macrOphage cell membrane in preparation for phagocytosis." However, Parish (89), Tizard (106) and Nelson (85) suggest that these "opsonic adherence anti“ bodies" are not cytophilic before combination with antigen and represent a different class of antibody than that described by Boyden (l6). Parish (89) reported a distinct yl fraction containing only cytOphilic antibody (as defined by Boyden). Bovine gamma globulin and bovine plasma albumin were the antigens used. It was possible to remove all of the cytophilic fraction by absorption with macro- phages. In addition Parish demonstrated only a slight increase in phagocytosis of protein antigen pretreaflbd (opsonized) with the cytOphilic fraction; presumably show- ing that the cytOphilic fraction had little classical opsonic activity. This fraction was found to be a 198 macroglobulin. Guinea pig antibodies responsible for passive cutaneous anaphylaxis or the sensitization of strips of ileum for re- action with antigen were found among the 73 72 fraction (85). me' at' tht net sh< orf 3C1 Cyt res the fev t0 prc pas hot the pie the ant mar lea Tizard (106) showed that mouse antinsheep erythrocyte cytophilic antibodies included IgM. The passive indirect method in contrast detects only IgG. Cytophilic antibodies attached only to macrophages. Erythrocytes opsonized by the passive indirect method attached to polymorphonuclear neutrophils, mast cells and macrophages. ElectrOphoresis showed cytophilic activity only in a region close to the origin with B and fast 7 mobility. In contrast, Opsonic activity was found widely distributed through the y region. Cytophilic activity in antisera obtained during the primary response was demonstrated to be susceptible to freeze“ thawing damage. It was possible to destroy activity by as few as two freezing cycles to -20C (Tizard, 107). In contrast Uhr (113) found that guinea pig antisera to S. paratyphi B flagella produced without adjuvant only produced macrophage rosettes when a method similar to the passive indirect technique was used (Opsonic adherence anti— body). However, complete Freund's adjuvant is necessary for the production of macrOphage cytophilic antibody in guinea pigs. Also the test antigen (whole bacteria) was different than the immunizing antigen (flagella). Apparently the antibodies which Uhr studied were not cytophilic in the manner described by Boyden (15). Cytophilic and opsonic adherence antibodies have at least one common characteristic. Both activities are lost 10 after pepsin treatment. This enzyme destroys the Fc por- tions of immunoglobulin heavy chains. Apparently both the cytophilic and opsonic adherence sites are located on the Fc portion. It has been suggested that opsonic antibody be- comes cytophilic by configurational changes in PC following binding of antigen. Also, cytophilic antibodies may al- ready have these changes preformed in the Fc portion (106). The immunoglobulin class and/or subclass containing macrophage cytophilic antibody varies by species, antigen used, route of injection and time after immunization. Nelson (85) and Tizard (107) have prepared a detailed list« ing of these characteristics. However, it should be pointed out that only IgG1 and IgG3 bind to macrophages in humans. This is significant because IgM or other classes of IgG are cytophilic in other species including those on which most experimental models are based (55). The production of macrophage cytOphilic antibody occurs independently from other antibodies and usually requires complete Freund's adjuvant, intravenous injection, or immunization with a live attenuated bacterial strain. The antibody class in the primary response is usually IgM and in the secondary response IgG (85,107). Apparently cytophilic antibodies in mice and rats can be produced by heat-killed E. coli and Salmonella typhimurium without 11 using complete Freund‘s adjuvant (94,107). Rowley et al. (95) observed a mouse 198 macroglobulin cytophilic antibody to Salmonella typhimurium. Cytophilic antibodies attach to macrophages with the Fc portion (106). The exact structure of the binding site is not known. Human gamma1 myeloma proteins (cytophilic) and gamma2 myeloma proteins (non-cytOphilic) differ only by 4 peptides (peptide mapping) (45,57). Opsonic antibodies bind to macrophages only after com- bination with antigen. Antigen-antibody complexing induces confirmational changes in the IgG molecule (58,84). It is likely that pre-existent changes in the unbound antibody molecule are responsible for cytophilic properties (107). Comparable changes are likely to occur in the Fc portion of homocytotropic antibodies (103). The macrophage membrane receptor for cytophilic anti- body in guinea pigs has been investigated by Howard and Benacerraf (50) and Davey and Asherson (30). The receptors were found to be resistant to trypsin, chymotrypsin, papain, ficin, pronase, reducing agents (mercaptoethanol) and enzyme poisons (azide fluoride and dinitrophenol). However, they were susceptible to phospholipase A, lecithinase C, Naja naja venom, reagents which react with free SH groups (iodo- acetamide and p-chloromercurybenzoate) or with both SH and NH2 groups (formaldehyde and isothionate) and oxidizing 38 re fo ma to IC 12 agents (periodate and nitrite) (85). Mouse and guinea pig receptors for 78 IgG cytOphilic antibodies in mice were found to be resistant to trypsin and pepsin (70). The macrophage receptors for 198 macroglobulin were sensitive to trypsin (86,106), and rabbit macr0phage cytophilic recep« tors were sensitive to trypsin (101). Tizard (106) found that alveolar and peritoneal macrophages did not differ significantly in their capacity to adsorb cytOphilic anti- body. However, Berken and Benacerraf (10) reported that mouse, guinea pig and rabbit alveolar macrophages bind a greater number of cytophilic antibody molecules than peri“ toneal macrophages in the same species. It is unknown whether this is a difference in the total number of receptor sites on the alveolar and peritoneal macrophage or if the "alveolar macrophages are more reactive because of having fewer sites occupied (being in an area with very low concen- trations of plasma proteins in the environment)" (85). Macrophage cytophilic antibodies bind only to macro- phages (17). It has been shown that neutrophils, eosino- phils, lymphocytes, fibroblasts, peritoneal serosal cells, alveolar epithelium and mast cells are not sensitized by macr0phage cytophilic antibody (50,63,106). Alveolar, splenic and peritoneal macroPhages, Kupffer cells, micro- glia and blood monocytes bound approximately the same amount of antibody in the mouse (106). However, Keller and Sorkin (64) reported a much higher level of cytophilic binding in 13 rat mast cells and rat liver cells (obtained by passage through a steel mesh) with rabbit antisera to human serum albumin. This very large uptake by mast and liver cells was thought to be partly due to their large size. This topic will be discussed later in more detail. These find- ings have never been confirmed. Coulson et al. (28) found "that guinea pig small lymphocytes, transformed in mixed cell culture, may also take up macrophage cytophilic anti- body." Kossard (68) showed "that rabbit anti—sheep erythrocyte antisera contained antibodies cytophilic for rabbit and guinea pig macrophages and small lymphocytes." Several factors have a profound effect on the binding of cytophilic antibodies to macrophages. The first to be investigated was ambient temperature. Apparently, the effect of temperature on the binding of macrophage cyto- philic antibody varies between species. A rise in tempera- ture inhibits most adsorptive processes possibly due to significant entrapy changes (107). It was found that rabbit anti-HSA bound better at 0C than at 37C. In addition the bound antibody could be eluted at 37C. Kossard and Nelson (69) and Boyden (16) reported better binding of cytophilic antibody at 4C. Mouse IgG and IgM macrophage cytophilic antibodies were found to bind much better at 4C than at 37C (106,107). Jonas et al.(63) found that changes in temperature did not significantly 14 affect guinea pig macrophage cytophilic antibody adsorption. It was observed by Berken and Benacerraf (10,11) that cyto- philic antibodies were bound better at 37C but were eluted rapidly. Another factor affecting cytophilic antibody binding is competition for receptors by non-specific cytophilic immunoglobulins. Other cytophilic antibodies present in serum profoundly affect the binding of a specific cytophilic antibody. Usually this causes an inhibition of antibody adsorption. It is believed that non-specific cytophilic antibodies compete for available cell receptors with specific cytophilic antibodies (10,57,63). There is an apparent heterogeneity of cell surface receptors. Adsorption of macrophage cytophilic antibodies is an example of adsorption of a solute onto a surface. This phenomenon has been expressed mathematically by the Langmuir adsorption isotherm which states that the amount of solute adsorbed by a surface is prOportional to the amount available for adsorption until the surface receptors are saturated. Adsorption of IgG by mouse macrophage was found to deviate from the expected values. However, mouse IgM adsorbed in the expected fashion (108). The deviation from expected values could be due to interaction between receptors. Low levels of non«specific mouse cytophilic antibody does in fact increase adsorption in vivo and 15 in vitro. This would suggest c00perative interaction between receptors in inflammatory foci with increased antibody levels. The mechanism of the observed increase in receptor avidity may be due to simple neutralization of macrophage charge by adsorbed charged immunoglobulin molecules, which may allow other molecules or parti- cles to be adsorbed. Alternatively it is possible that activation of high avidity receptors is brought about directly by the previous occupancy of low avid- ity receptors (108). Changeux et al. (23) have proposed a theory to account for such interaction between receptors, hypothesizing that receptors are linked to each other and that the efficiency of interaction is related to the degree of association between receptors (based on acetylcholine, postsynaptic receptors). Mouse peritoneal macrOphages may show a significantly in- creased uptake of antibody in inflammatory foci (108). Evidence is accumulating (Tizard, unpublished observa- tion) that in vivo macrophage cyt0philic antibodies exist entirely in solution. However, under conditions of raised immunoglobulin levels in the microenviron- ment of macrophages there may be a net movement of immunoglobulin onto cells. Huber and Fudenberg (55) suggest that based on the available evidence IgG receptors and cytophilic receptors are similar or identical. Their function is to mediate the binding of antigen-IgG complexes. They also pointed out that in studies by Berken and Benacerraf (10) a large por- tion of the 72 fraction in guinea pigs showed the ability to bind to macrophage receptors. However, it should be pointed out that the definition of cytophilic antibody used by 16 Berken and Benacerraf is not significantly different from the usual definition of Opsonic antibody. Interaction of dendritic macrophages with lymphocytes in lymphoid follicles leading to the production of antibody has been studied extensively (7,88). The localization of antigen on the processes of dentritic macrophages is be- lieved to be mediated by a cytophilic antibody. In fact, the binding of antigen has been correlated quite closely to antibody both in its time of appearance and physical loca- tion (7). The dendritic processes of these cells are extensive and in close physical proximity to lymphocytes. The antigen remains attached for long periods of time (2). It is be— lieved that antigen is presented to lymphocytes initiating antibody production. This process is usually known as anti« gen processing. It should be mentioned that dentritic cells are not typical macrophages and usually are not classified as such. The cytophilic antibody has not yet been charac- terized. More recent work has helped to clarify antigen process« ing (25). It was demonstrated that highly purified peri— toneal macrOphages from immune animals following exposure to antigen could cause transformation of immune lymphocytes (98). Thymocytes, LZC leukemia cells, polymorphonuclear leukocytes and column purified lymph node lymphocytes were 17 not able to cause lymphocyte proliferation. It was found that peritoneal cells from non«immune animals were not able to bind enough antigen at very low concentration (pg/m1) to cause lymphocyte proliferation. Peritoneal cells from immune animals were able to do so at ug/ml concentrations of antigen. Non-immune peritoneal cells sensitized with serum from immune animals were subsequently able to stimu- late lymphocyte proliferation. This effect could be blocked by hapten coupled with a heterologous antigen. Binding of antigen at very low concentrations could also be blocked by anti-immunoglobulin (25). At higher antigen concentrations anti-immunoglobulin had no effect on peritoneal cell-induced lymphocyte prolif- eration. Thus, a cytophilic antibody dependent mechanism is only necessary at very low antigen concentrations. At high antigen concentrations the alternative non-antibody dependent mechanism operates. Although the studies were performed in vitro, it appears likely that cytophilic anti- body concentrates sufficient antigen on macrophage surface to initiate a secondary immune response when antigen is present intermittently and in low concentration (25). al—globulins have been found to be associated with cytophilic activity. However, there are a number of diffi— culties in studying and interpreting findings with this class of serum factors. They are: l) lability to freezing 18 and thawing, 2) weak sensitization of macrophages, 3) nature of equilibrium between free and macroPhage bound state, 4) nature of producing cells, 5) stability and half-life in vivo, and 6) attachment to lymphocytes and other cell types (85). A poorly characterized al-globulin was found in mice. Studies were performed in mice to better character- ize their nature. The normal serum of a number of Swiss mice was tested for cytophilic activity. One pool showed a low titer (1:2) by the indirect test in a fraction contain- ing albumin and a fast al-globulin. Red blood cells adhered to normal nonsensitized cultured macrophages of CS7 B1/6J mice and some Swiss mice. Using both the direct and in- direct test, no cytOphilic factors were found in normal C57 Bl/6J mouse sera to Sarcoma 1 cells. Sera raised by injec- tion of complete Freund's adjuvant alone contained one pool which showed anti—sheep erythrocyte cytophilic activity. Mouse anti-sheep erythrocyte antisera contained cytophilic antibodies to sheep red cells and none to guinea pig and human erythrocytes. No hemagglutinating or complement- fixing activity has been found in the albumin-al-globulin fraction. The al-globulin factors appear to have the char- acteristics of antibodies (85). There have been a number of reports of antibody with more usual types of activity in the a-globulin fraction (4,20,26,32,104). Allen, Saba and Molnar (3) isolated and characterized an az-globulin which 19 stimulated in vitro Kupffer cell phagocytosis of test colloids and stimulated R.E.S. clearance of test colloids. Electrophoretic studies show that al-globulins are not related to conventional immunoglobulins (85). The apparent immunologic specificity of al-globulin cytophilic anti~ bodies suggest they are unrelated to acute phase proteins like C reactive protein and the Darcy al-glyc0protein that occur in serum after tissue injury. They also have differ— ent electrophoretic mobility than other unusual a-globulins such as: l) al-macroglobulin virus inhibitors in normal human, monkey, bovine and ferret sera (12), 2) slow al-globulin appearing intermittently in adjuvant induced arthritis of rats (36), and 3) al-globulins in tuberculosis in humans (99) and guinea pigs (29). Tizard (109) reported desensitization of macrophages for cytophilic antibody following intravenous injection of E. coli lipopolysaccharide. The response was described as being triphasic. First, a slight increase in adsorption occurred in the initial five minutes. This was followed by a significant depression of antibody adsorbance reaching a maximum level in 30 minutes. By fourteen hours there was a significant rebound of adsorption above normal levels. It is significant that the maximum depression occurred dur— ing the first hour after LPS injection. This is the time period in which the important pathogenic changes are 20 believed to occur in endotoxin poisoning (Moon, personal communication). Whether the preceding changes are major pathogenic events or merely a reflection of altered cell metabolism or membrane pr0perties is not known. Shortly afterwards, Tizard (110) investigated the mech' anism of action of LPS on antibody adsorption. He found no relationship between surface charge or rate of anaerobic glycolysis and the changes in antibody adsorbance. A serum factor was postulated as the mediator of the LPS related events. This factor was not present in serum or whole blood thus excluding clotting factors or leukocyte pyrogens as mediators. It was also thought unlikely that the factor was an immunoglobulin because the triphasic response occurs in gnotobiotic mice. Tizard (110) thought it possible that the factor could be a lymphokine because of its similarity of action to macrOphage inhibition factor. It is possible that interaction of the antigen (LPS) with sensitized lymphocytes would bring about release of a lymphokine. No further work on this topic has been reported. However, Kessel and Braun (65) reported a cell-bound antibody which protected macrophages from the toxic effect of endotoxins. There has been a significant degree of controversy concerning the biologic role of cytOphilic antibodies. MacroPhage cytophilic antibodies may act as either classical or cell-bound opsonins. Mouse macrophages were sensitized 21 with anti-sheep erythrocyte serum at 4C, washed and antigen added. The mixture was then incubated at 37C. Both IgM and IgG cytOphilic antibodies increased phagocytosis. The IgG fraction increased phagocytosis to a greater extent than the IgM—fraction (107). At 37C two processes are known to occur: 1) elution of macrOphage cytophilic antibody, and 2) activation of phagocytosis. Whether Tizard observed elution of cytophilic antibody, subsequent opsonization and phagocytosis of antigen—antibody complexes or simply temperature activated phagocytosis of membrane bound anti- gen-antibody complexes (cell-bound Opsonin) is not clear. Rowley, Turner and Jenkin (95) presented convincing evidence that protective immunity to Salmonella typhimurium in mice can be transferred to normal animals by immune peri- toneal macrophages. Resistance could also be transferred by sensitization of normal cells with immune serum. Trypsin digestion destroyed the ability to transfer immunity with immune cells. At 37C immune macrophages were found to elute a 2-mercaptoethanol sensitive material (macroglobulin). This eluted macroglobulin was found to protect normal mice. Tizard (107) suggests that this cyt0philic macroglobulin acted as a classical opsonin. Parish (89) reported a simi- lar macrophage cell-bound antibody for E. coli. More recently, Mittal (81) demonstrated the production of macrophage cytophilic antibodies in rabbits and mice to 22 Escherichia coli. Cytophilic antibodies were found only in antisera from rabbits immunized with a live smooth (3662) strain of E. coli. Capsular and somatic antigens of strain 3662 attached to red blood cells did not produce cytophilic antibody. It was possible to passively sensitize rabbit peritoneal macrophages in vivo only with antisera to live strain 3662 of E. coli. Antisera to other strains were not effective. In vitro passive sensitization was possible with peritoneal macrophages, lymph-node cells and spleen cells, but not with polymorphonuclear leukocytes. A rough strain of E. coli (Lilly) did not produce cytophilic antibodies. Sensitization of spleen cells, macrophages and PMNs was by the passive direct method. Lymphocytes were tested by the suspension-centrifugation technique. Several reports impli- cating a similar cell-bound antibody in immunity have appeared in the literature. Vi-coated erythrocytes formed rosettes on anti S. typhi Vi polysaccharide sensitized macrophages that were resistant to phagocytosis at 37C for long periods (72). Actor and Pitkin (1) reported the presence of a cell- adherent immune factor in immunity to Vibrio cholerae by immune serum or immune peritoneal cells and normal peri- toneal cells sensitized with immune serum. Immune spleen cells were able to transfer immunity but only if given at least fourteen days prior to challenge. The immunity 23 obtained by transfer of immune peritoneal cells was long- lasting; that transferred by sensitized normal cells was temporary. All injections were intraperitoneal. Mice were immun- ized with a killed whole cell vaccine of V. cholerae Ogawa 41. It was felt that two mechanisms of protection were operating; one, cytOphilic antibody and two, a non— cell-adherent factor. The reason for suggesting a non- cytOphilic factor was that heterologous serum was capable of protection on passive transfer (1/60 efficiency of homo- logous serum), but was unable to sensitize macrophages. Macrophage cyt0philic antibodies have been found in trypanosomiasis and schistosomiasis. Tizard and Soltys (lll) described cytophilic antibodies in the sera of rabbits infected with a mouse adapted strain of T. brucei. The titer of cytOphilic antibodies reached a peak in one week and decreased steadily afterwards. They were associated with high titers of agglutinating antibody. The antibodies were resistant to 2—mercaptoethanol and heating to 56C for 30 minutes. Adherence occurred only to mononuclear phago- cytes and not to polymorphonuclear leukocytes or small lymphocytes. They bind to cells better at 4C and 22C than at 37C. Recently Capron et al. (21) and Capron et al. (22) have reported an IgE macrophage cyt0philic antibody which 24 causes adherence of peritoneal cells to Schistosoma mansoni schistosomules in vitro. These antibodies were found in the serum of schistosome-infected rats. There was a highly significant difference in the number of adherent macro- phages between normal cells and antisera sensitized macrophages. Binding of antisera was destroyed by heating at 56C for 3 hours and could not be activated by fresh guinea pig serum. Adsorbance with S. mansoni antigens significantly decreased the ability of sensitized macrophages to bind to parasites. These were shown to be specific IgE antibodies by adsorption with goat anti-rat IgE serum specific for c. Shain or F(ab1)2. Serum containing IgE myeloma anti- bodies and reaginic antibodies failed to sensitize macro- phages for immune adherence. Fluorescent staining with goat anti-rat IgE was posi- tive in immune serum sensitized macrophages but negative in macrophages sensitized with antisera heated to 56C for 3 hours or adsorbed with F(ab1)2 anti-rat IgE. IgE binding to macr0phages was confirmed by peroxidase-labeled IgG anti-rat IgE and ultrastructure studies. It was felt that these specific IgE macrOphage cytophilic antibodies accelerate the killing of S. mansoni schistosomules and may be important in the immunity to the parasite in rats and possibly in humans (22). 25 Ivanyi (59) reported a suppression of the primary response to HSA by hyperimmune chicken anti—HSA. This sup- pression was abolished by repeated adsorption of the anti- sera with spleen cells. Apparently cytOphilic antibodies in anti-HSA serum caused a suppression of the primary response. For some time it has been thought that cell-bound anti- bodies might be at least partially responsible for delayed hypersensitivity. A number of investigators (9,14,42,48,49, 87) found no relation between macrophage cytOphilic anti- bodies and intensity of delayed hypersensitivity. However, Boyden (17) did find a relation between the presence of macrophage cytophilic antibodies and delayed hypersensitiv- ity in guinea pigs. The antigen in this case was sheep red blood cells and complete Freund's adjuvant was required for cytophilic activity. Boyden also suggested that unfraction« ated serum usually contains much more classical antibody than cytophilic antibody which could interfere with the action of cytophilic antibody. There is evidence that interference occurs in vivo between classical serum anti- bodies and the factors responsible for hypersensitivity to tuberculin and other types of cellular immunity (15,46). Another group of investigators, Dupuy et al. (33), found that a cytophilic factor could take part in delayed hypersensitivity to PPD. They were able to transfer delayed 26 hypersensitivity to PPD to normal guinea pigs with 2-3 ml of fresh plasma from X-irradiated PPD hypersensitive donors. Similar plasma from nonirradiated PPD donors or from irradi- ated normal guinea pigs did not transfer delayed skin reaction. It was possible to transfer delayed hypersensi- tivity to PPD by incubating large numbers of nucleated spleen cells with plasma from irradiated sensitized donor and injecting into normal guinea pigs. Hsu and Mayo (53) observed that immune macrophages clumped in the presence of antigen. This clumping was strong and could be obtained by both the H and O of Salmonella typhimurium SRll. They suggested this phenomenon was due to macrOphage cyt0philic antibody. Granger and Weiser (43) demonstrated a cytophilic factor on peritoneal macrophages (mouse C57Bl/Ks) which mediated binding of target cells (sarcoma I) to the peri- toneal cells. This factor could be eluted at 56C. This same group later found that the factor was able to fix complement and kill target cells (44). Trypsin-treatment was found to remove this factor (116). It was also found that trypsin-treated macrophages could be activated by treatment with immune serum and eluate from immune cells. Similar results were reported by Gordon (41) and Hoy and Nelson (52). 27 Gill (38) reported the presence of marcophage cyto— philic antibody in heterologous antilymphocyte serum. CytOphilic antibody was present as measured by both the passive direct and passive indirect technique. Macrophages sensitized with rabbit anti-mouse lymphocyte serum showed the presence of cell-bound antibody by immunofluorescent studies. Macrophages sensitized by normal serum did not. Alveolar macrophages obtained in vivo from rabbits producing anti-lymphocyte serum formed rosettes with thymocytes. Alveolar macrophages from normal rabbits did not. Antibody eluted at 56C from macrophages sensitized by the passive direct method was able to sensitize cells by the passive indirect method. Cytophilic activity was found in the IgG fraction by anion exchange chromatography. Pepsin treatment destroyed cytOphilic activity. It was possible to remove almost all the cytophilic activity by repeated adsorption with macrophages (38). Taub and Lance (105) found a high uptake of chromium-51 labeled lymph node cells treated with ALS in vitro in the liver. It is possible that cytophilic antibody assisted Kupffer cell clearance in that situation (38). Gill et al. (39) found a close relationship between cytophilic activity and the immunosuppressive effect of rabbit anti-mouse lymphocyte serum. In this study the titers of leukoagglutinating, cytophilic and indirect and 28 direct macrophage cytophilic antibodies were correlated with the survival time of CBA mouse skin grafts on balb/c mice. Simple linear regression analysis showed significant correlation with all types of antibody. Similarly multiple regression analysis showed significant correlation with all types of antibody. However, multiple regression analysis showed that titers of indirect and direct cyto- philic antibody were the most efficient indicators of graft survival. It was suggested that leukoagglutinating and cytotoxic antibody played only a minor role (39). Pokorna (92), KOprowski and Fernandez (67) and Gibbs et al. (37) reported suggestive evidence for cytophilic antibodies in autoimmune phenomenon. Tizard (107) suggested two general ways in which macro- phage cytOphilic antibody carries out its role. The first mechanism would require combination of antigen with cell- bound antibody leading to fixation of the antigen. Follow- ing this the immune macrophage would exert its cytotoxic effect on the antigen or target cell. The second suggested mechanism would be combination of the antigen with the cell-bound antibody followed by activation of macrophage enzyme systems and altered macrophage activity. Tizard (107) listed several examples of cell "activation" by anti- gen cell-bound antibody. Rowley et al. (95) found a macro« phage cytophilic antibody to Salmonella typhimurium capable of passive transfer of immunity. Urea extraction yielded a 29 minute amount of protein which was very active in sensi- tizing peritoneal cell. This protein was found to be a macroglobulin (95). Earlier work stated that immunity to Salmonella sp. was due to cellular factors and that anti- bodies played no role (35,96). Thus, the findings of Rowley‘s group raised the distinct possibility of the cooper— ative interaction of humoral and cellular factors in the immunity to Salmonella. Parish (89) reported an enhance- ment of phagocytosis of heat-killed E. coli with macrOphages sensitized with antiserum to E. coli. This suggests that a similar cell-bound antibody enhances resistance to E. coli. Ushiba et al. (114) reported that immunization with a live avirulent strain of S. enteritis did not produce 198 macroblogulins. However, they did find a 12.78 macroglobulin in the sera of mice immunized with the S-type killed vaccine similar to that reported by Rowley. The Japanese group used different strains of bacteria. They reasserted that immun- ity to Salmonella was due to cellular factors and not humoral factors. Mackaness and his colleagues investigated the relative roles of humoral and cellular immunity in Salmonella infec« tions. It was found that passive transfer of serum from immune animals or immunization with heat-killed cells increased clearance of bacteria but could not prevent multi« plication of intracellular organism. Cellular immunity was 30 necessary for resistance to active infection (74). It is important to point out several differences between Mackaness's studies and Rowley's. First, Mackaness used actively infected animals while Rowley challenged animals with bacteria which were prOphylactically given antisera. Second, Mackaness used intravenous injection leading to widespread dissemination rather than the intraperitoneal route which resulted in more local dissemination as used by Rowley. Significantly different phenomenon were observed in the two studies. The Mackaness group went on to demonstrate that Listeria monocytogenes-infected mice had greater resistance to growth of S. typhimurium in the liver and spleen than did mice vaccinated with heat-killed S. typhimurium. Mice actively infected with S. typhimurium were totally resis- tant to intravenous challenge with Listeria monocytogenes (13). This was interpreted to mean that enhanced micro- bicidal activity (activation) of macrophages was necessary for resistance to active infection with S. typhimurium in mice. However, there is some question as to whether the Mackaness group actually studied cytOphilic antibodies. Macrophage cytophilic antibodies to S. typhimurium can be most reliably produced with active immunization with aviru- lent organisms. Intravenous injection of heat-killed.cells as used in theiBlandenet al. (13) studies are not as 31 reliable as the other two methods. Recent evidence suggests that a shared antigen between bacterial species is responsi- ble for the apparent "activation" of macrophages postulated by Mackaness (79). More recently, Margolis and Bigley (78) have demon- strated a cytophilic macroglobulin in mice immunized with RNA-protein fractions of Salmonella typhimurium. This macroglobulin was specifically reactive with a protein com— plex comprising part of the bacterial cell wall. This same protein complex and RNA are the immunogenic agents. The cytophilic macroglobulin was eluted from peritoneal cells, washed spleen and lymph node cells by incubation at 56C for 1 hour. Peritoneal cells from animals immunized with live S. typhimurium RIA, phenol-extracted RNA-protein fraction and ethanol-extracted RNA-protein fraction all had signifi- cant amounts of cytophilic antibody. Groups of normal mice were injected intravenously with macroglobulin and normal peritoneal cells sensitized with normal globulin and then 24 hours later injected intraperi- toneally with 500 LD50 doses of S. typhimurium SRll. After 5 days there was a 100X decrease of viable bacteria in the livers and spleens of macroglobulin treated animals and 750x decrease in animals treated with sensitized normal macrophages. The survival time in both macroglobulin and sensitized macrophage treated animals was approximately 32 twice that of controls. In vitro studies showed a marked depression of bacteria extracellularly in cultures of peritoneal cells sensitized with macroglobulin. Intracellu- lar bacteria were not affected. Immune peritoneal cells showed an almost complete elimination of bacteria both intracellularly and extracellularly after three hours (78). Kurashiga et al. (71) and Mitsuhashi et al. (80) re- ported macrOphage cytophilic antibodies to Salmonella enteritidis in mice immunized with attenuated S. enteritidis and on macrophages exposed to ribonucleic acid from immune monocytes. Collins and Mackaness (27) have shown that mice deve10p a delayed hypersensitivity reaction to a protein antigen in the cell wall. Smith and Bigley (100) found a similar protein-RNA complex which enhanced resistance to Salmonella and stimulated delayed hypersensitivity. The cytophilic macroglobulin was reactive with this protein complex (78).. More recently Eisenstein (34) demonstrated that pro- tection in these ribosomal vaccines was due to contaminating O antigens. Mutants with alterations in the terminal sugars of 0 antigen were not able to confer protection. The RNA may function as an adjuvant in these vaccines. Perhaps it would be helpful to review several aspects of the phagocytosis and killing of Salmonellae by phagocytes. First, there is a heterogeneity in the ability of phagocytes 33 to kill bacteria. In the initial stage most bacteria in- gested are killed. The surviving bacteria are presumably in bacteriocidally incompetent phagocytes. Persistence of bacteria reflects the capabilities of only a fraction of the phagocytes. These bacteria multiply causing active infection. Also it is apparent that the interactions be- tween macrOphage and bacteria which occur during the initial period following introduction of bacteria are important in determining whether an active infection will result (60). Marecki et al. (77) using mice showed that Salmonella typhimurium SRll was killed by both normal and immune phang cytes intracellularly and that the early immune response of the host was important in determining whether the infection would be contained locally or disseminated. The rate of extracellular growth of bacteria and phagocytosis were then the key factors in determining the outcome of infection. Antibody was found to be important in these extracellular events (54,117). Rhodes and Hsu (93) found that the intracellular p0pu— lation of Salmonella enteriditie in guinea pigs was directly related to the extracellular p0pulation through phagocytosis. Immune serum was able to suppress the extracellular bacteria initially. Accumulation of intracellular bacteria actually reflected phagocytosis of an increasing pOpulation of extra- cellular bacteria rather than intracellular growth. 34 This emphasized again the importance of extracellular events (bacterial growth and subsequent phagocytosis) in interpret- ing intracellular events (bacteriocidal properties of macro- phages). This source of misinterpretation was pointed out by Patterson and Youmans (90) in regard to the enhanced ability of "activated" macrOphage to inhibit the growth of Mycobacterium tuberculosis. Thus, factors which influence the efficiency of the phagocytosis of bacteria are more important than previously thought in the resistance to infections with Salmonellae. MacrOphage cyt0philic antibody is one of these factors. Although there has been some work on the role of macro— phage cytophilic antibody in resistance to gram-negative bacteria, little work has been done on its role with endo- toxin. Parish (89) coated red blood cells with somatic polysaccharide and eXposed them to macrophages sensitized with antisera to E. coli. The erythrocyte polysaccharide complex adhered to the macrOphages but no phagocytosis occurred. Berken and Benacerraf (10) found cytophilic anti- body specific for E. coli polysaccharide or for the dinitro« phenol hapten. This was demonstrated by the passive indirect technique using sera of guinea pigs immunized with heat- killed E. coli or DNP proteins in complete Freund's adjuvant. Dramatic binding (+3—+4 rosettes) could only be demonstrated with undiluted sera. There was some question as to whether 35 a sufficient amount of polysaccharide had been attached to the red blood cells. Because free endotoxin was not used in these studies, their relevance to in vivo situations is tenuous. The role of Kupffer cells in the clearance of bacteria, viruses and other particulate matter from the blood stream has been well~studied (Sl,62,75,76). However, many subtle questions remain unanswered. Moon, Vrable and Broka (82) demonstrated using the perfused rat liver, a distinction between trapping and killing. Following gross removal of blood by perfusion with sterile M-199 ”over 70-80% of a 109 to 1010 dose of viable S. typhimurium was trapped after a single pass, but no significant bacterial killing was ob- served. When blood or plasma was added to the perfusion medium, over 50% of the trapped bacteria were killed in 15 to 30 minutes. Phase contrast and election micrographs of perfused livers showed extensive extracellular trapping in the sinusoid." It is likely that at least part of this sinusoidal trapping of bacteria is due to cytOphilic anti- body. However, scanning electron micrographs of similar studies using Candida albicans showed no correlation between trapping and adherence to phagocytes (97). It has been known for many years that Kupffer cells tend to occur at the junction of liver sinusoids (47). This strategic location would increase the trapping ability 36 of macrophages made ”sticky" by cytophilic antibody. Keller and Sorkin (64) found a high level of cytophilic antibody adsorption on liver cells. However, a mixed cell population was studied. In their study the adsorbance of cytophilic antibody (rabbit) was observed on rat tissue mast cells and cells from other tissues in vitro. Antisera 1311 labeled-human were raised to human serum albumin and serum albumin was used as the antigen. Mast cells, spleen cells, liver cells, mesenchymal lymph node cells and red blood cells had very low levels of adsorption when sensi- tized with normal serum (1-10 c.p.m.). The same cell types had higher levels of adsorption following sensitization with specific antisera (308-4820 c.p.m.). The three cell types with the highest levels were liver cells (4820 c.p.m), mast cells (3822 c.p.m.) and red blood cells (1622 c.p.m.). The ability of mast cells to adsorb homocytotropic antibodies is well-known and their role in allergies has been well- documented. However, liver cells with the highest levels of adsorption in this study have received very little atten- tion. Part of the reason for this is that a mixed cell population was studied by Keller and Sorkin (liver fragmen- tation followed by passage through a stainless steel mesh). More importantly it has been very difficult to obtain and maintain either parenchymal cells or Kupffer cells in vitro in sufficient quantity and purity to make detailed studies 37 possible. Recently, much better techniques have been devel- oped for this purpose (83, Zlydaszyk, personal communication). After discussing this report of a very high level of cytophilic antibody on liver cells, perhaps it would be interesting to mention the stickiness of Kupffer cells ob— served by early workers. Beard and Rous (8) remarked, "Except for the immense circular membrane, no feature of the Kupffer cell is more striking than its stickiness,--which is equally remarkable whether the enveloping medium is serum or Tyrode solution. It has much interfered with the study of the cells in vitro. Whatever touches their surface tends to adhere, whether it be red or white cells or "bodies" that are floating by." "A sticky surface is evidently one of the special characters enabling the cell to perform its task of phagocytosis. The stickiness of leukocytes from the blood and from exudates, with both of which we have had much eXperience, is in comparison negligible." Knisely et al. (66) after a lengthy and extensive study of selective phagocytosis in vivo noted that, "The filaments and coatings which are precipitated from frog plasma on India Ink are very sticky to the Kupffer cells which line the frog‘s hepatic sinusoids. They are so sticky to the sinusoid-lining cells that the moment a coated particle touches the surface of one it sinks into the cytoplasm of the sinusoidal lining, apparently as quickly as a small 38 globule of mercury rolling along a laboratory table top touches and merges with a larger one." Although there has been a significant amount of re— search by well-qualified investigators there are still many unanswered questions. Much of the reason for this is the unusual nature of the antibodies themselves. Perhaps it would be helpful to list these difficulties (85): 1. Macrophages are weakly sensitized. Cytophilic anti- bodies are rapidly eluted from macrophages at 37C in most cases. It is not known with absolute certainty how the equilibrium between free and macrophage-attached antibody functions in vivo. The type of cells producing cytophilic antibody is not known with certainty. . The stability and half-life in vivo of cytophilic antibodies both in plasma and/or macrophages is not completely known. The possibility exists that at least some cyto- philic antibodies sensitize cells other than macro— phage. Uhr (113) reported a low level of sensiti- zation (~S%) of lymphocytes with antisera to Salmonella flagellar antigen. It is possible that the biological consequences of attachment of antigen to 198 cytophilic antibodies on macrOphages may differ according to the nature of the antibody and its eliciting antigen (113). "Antibodies (of whatever class) which are present in relatively small amounts on the macrophage sur- face in vivo may be released either into the medium or into the interior of the cells during manipula- tions of the cells in vitro. This would provide an almost permanent excuse for failure to demon- strate a clear biologic function for them" (85). 39 In spite of these difficulties a number of roles have been suggested for macrophage cytophilic antibody in vivo in gram-negative infections. They are: 1. Parish (89) stated "in vivo cytOphilic antibody sensitizes the macrophages lining sinusoids of lymph nodes, spleen, liver and other organs so that foreign material, for example bacteria or their toxins, are more rapidly removed from the circulation. Thus, its probable function is to act as a 'cell-bound' (or 'cell-fixed") opsonin." 2. Berken and Benacerraf (10) suggested, "It must be considered whether or not macrophages can transport and effectively deliver cytOphilic, yZ, opsonic or complement binding antibodies to sites of inflammation or tissue rejec- tion that macrophages do travel from areas of high antibody concentration in serum or lymphoid organs to areas of low antibody concentration in various tissue spaces." 3. Increasing adsorbance of cytophilic antibody in areas of high antibody concentration (inflammatory foci) due to changing receptor avidity could be important in their biologic role (108). 4. Cytophilic antibodies may coat macrOphages in all parts of the body both intra— and extra—vascularly and play a role in preventing infection. 5. The significant decrease in adsorption of cytophilic antibody by macrophages in the first hour after endotoxin 40 injection may be evidence in a negative way for a signifi- cant role for cytOphilic antibody in gram-negative infec- tion. 6. It has been suggested by Jenkin and Rowley (60) that intracellular uptake of macrophage cytOphilic antibody by phagocytosis or disruption of macrophage membranes leads to increased killing of Salmonella. 7. Macrophage cytophilic antibodies have been demon- strated in infection in many species of animals due to three different gram-negative organisms (Salmonellae, E. coli, and Vibrio cholerae) and two parasites (T. brucei and S. mansoni). It is likely that they exist in other types of infection. This would suggest evolutionary evidence for their importance in resistance to infection. 41 FLOW CHART EXPERIMENTAL METHOD Obtain peritoneal cells in M-199 I Incubate cells with serum for 30 minutes at 20C I Resuspend cells in M-199 and count I Suspend appropriate number of cells in 3 ml of M—199 add test antigen I Incubate for 5 minutes at 20C in water bath with agitation I Pass cells and test antigen through millipore vacuum filter apparatus, rinsing with 100 ml of M-199 (and 100 ml of deionized water when using labeled antigen) / \ With live bacteria as With labeled antigen add antigen do appropriate filtrate and filter to dilutions and pour counting vials and plates count MATERIALS AND METHODS Animals Sprague Dawley Rats, 300-400 g, were obtained from Spartan Research Animals Inc., Haslett, Michigan. All animals were maintained under standard laboratory conditions. Purina Laboratory Chow and water were available ad libitum. Bacteria Salmonella typhimurium, strain SR-ll, was used in all experiments. Overnight cultures of the organism, grown in brain heart infusion broth were centrifuged at 8000 rpm for 10 minutes. Cells were washed once in sterile deionized water and diluted to apprOpriate concentration. Pour plates of appropriately diluted samples were used for quantitation. Macrophage Preparation Sprague Dawley rats were anesthetized by intraperi« toneal injection of 1.5 mg of sodium pentobarbital (Haver Lockhart Laboratories, Shawnee, Kansas). Sterile M-l99 (50 ml) was injected intraperitoneally and the abdomen was gently agitated to insure mixture and freeing of peritoneal cells. After five minutes 15—25 ml of the M-199 was 42 43 withdrawn and placed in a Pyrex Plastic Centrifuge tube containing 400 u of sodium heparin (200 u/ml) and the rat serum under study. Following vortexing, the macrophage-rat serum preparation was mixed in a rotary shaker for 30 minutes at 20C. The mixture was centrifuged at 3500 rpm for five minutes and resuspended in 6 ml of M-199. The macrOphages were then counted on a standard hemocytometer using methelene blue dye and 0.2% acetic acid. The cells were adjusted to 1-2 X 106 cells/ml and resuspended in 3 ml of M-199. At this time 0.1 ml of a 1:10 dilution of the Salmonella culture was added (bacteria macrophage ratio 10:1). After vortexing this was placed in a rotary shaker bath for five minutes at 20C. The macrophage—bacteria suspension was then filtered on a 25 mm diameter millipore filter (Pore size 14 pH) and diluted with 97 ml of sterile M-199. The millipore filter was removed and blenderized in 100 ml of sterile deionized H20 for one minute, diluted appr0priately and pour plates were prepared. The free bacteria in M«199 (filtrate) were diluted in sterile deionized water and pour plates were made. The bacteria were then counted on a Digimatic Colony Counter. Heat-killed Cells Salmonella typhimurium SRll was grown for 24 hours in brain heart infusion broth and washed three times in 44 saline. The bacteria were then autoclaved for 10 minutes. The heat-killed cells were then suspended in saline to a concentration of l X log/ml. Immune Serum Sprague Dawley rats were immunized by intraperitoneal injection by the following schedule: 1 X 108 heat-killed cells the first day, l X 109 heat-killed cells every four days for three weeks. The animals were bled by cardiac puncture. The serum was obtained by centrifuging at 10,000 rpm for 10 minutes. Tube agglutination titers were per— formed on the antisera. Absorbed Serum Large amounts of heat-killed cells (m2.0 X lOlO/ml of serum) were used. Absorption was done at 45C for 4 hours and stored overnight. The antiserum heat-killed cell mix- ture was then centrifuged at 10,000 rpm for 10 minutes and agglutination titers performed. Normal Serum Sprague Dawley Rats were bled by cardiac puncture. The serum was centrifuged at 10,000 rpm for 10 minutes. Tube agglutination titers were performed. 45 Radiolabeled Endotoxin Preparation ZSICrO4 in volume of 1.5 ml Two-tenths millicurie of Na was added to 100 mg of Salmonella typhimurium endotoxin. The preparation was diluted to 10 ml with saline and incu- bated at 37C for 48 hours with constant stirring. The preparation was then dialyzed for 3-5 days at 4C against fre- quent changes of deionized water. The labeled endotoxin was centrifuged at 8,000 X g for one hour in the Sorvall Superspeed RCZ-B automatic Refrigerated Centrifuge (Ivan Sorvall Inc., Newton, Connecticut). The supernatant was recentrifuged at 100,000 X g for eight hours in the Beckman Model L Ultracentrifuge (Beckman-Dickinson Co., Rutherford, New Jersey). The pellet containing the labeled endotoxin was resuspended in 4.0 ml of deionized water and stored at -70C. This material, according to Chedid et al. (24), contained the toxic, heavily labeled, high molecular weight lipOpolysaccharide. 51Cr-Labeled Endotoxin-Macrophage Experiments Normal rat serum and immune rat serum and macrophages were obtained as described previously. Five u liter of 51Cr-labeled endotoxin was pipetted into 1 ml of non-sterile M-199 and counted on the Packard Gamma Counter for 1 minute. Macrophages (ml-2 X 106) were then added and the total volume adjusted to 3.0 ml. 46 This mixture was incubated in a rotary shaker bath at 20C for 5 minutes. It was then filtered through a Millipore membrane filter and washed with 97 ml of non-sterile M-199‘ and 150 ml of deionized water, carefully rinsing the cell bottle and filter apparatus. The membrane filter was removed and placed in a gamma counter vial. Each of the two rinses of the filter apparatus were placed in gamma counter vials. The cell bottle, filter bottle and filtrate bottles were counted on the gamma counter and the percent recovery in each fraction was determined. Statistics Statistical evaluations were made by the White Rank Order method (118). RESULTS Characterization of Peritoneal Cell Population and Serum Tube Agglutination Titers Tube agglutination titers were performed on all sera used in this study. The presence of cytophilic antibody and/or activity has been related to the presence of other types of antibody, especially agglutinating and complement fixing antibodies (18,85,111). All titers were performed as described in Materials and Methods. Table 1. Tube agglutination titers. Date Serum Type Titer 1/29/76 Immune #3 5120 1/29/76 Normal #2 40 2/20/76 Immune #4 2560 2/20/76 Immune #5 '2560 2/26/76 Normal #3 80 3/12/76 Immune #3, #4, #5 640 Absorbed with AKC 3/31/76 Immune #4 5120 4/05/76 Normal #3 40 Absorbed with AKC 7/24/76 *Immune #9 5120 a . . . , Last 1mmun1zation-subcutaneous 1n a vascular plexus. 47 48 Agglutination titers were in the range 2560-5120 for immune sera. Titers for normal sera were in the range 40-80. It was not possible to remove all of the agglutin- ating antibodies from immune sera or normal sera by absorp- tion with homologous heat-killed cells. Wright stain was done on most peritoneal cell samples used in this study. Standard Wright stain technique and reagents were used. The majority of cells present were macrophages or lymphocytes, occasionally a polymorpho- nuclear leukocyte was found (see Table 2). Table 2. Peritoneal cell counts. Percent of Cells Cell Type Macrophage Lumphocyte Mouse 66 34 Rat 54*,** 46 a Average 60 experimental determinations. an Macrophage count ~10% low because of technical difficul- ties. The ratio of macrophage to lymphocyte was approximately 55:45. The macrOphage count was about 10% low because of frequent macrophage disruption on fixed slides. 49 Experimental Apparatus Control To check the capacity of the Millipore filter to allow free passage of viable bacteria and endotoxin several . experimental determinations were performed using either bacteria or endotoxin alone. Using two different endotoxin preparations the percentage remaining on the filter was in the range of 0.36-0.88. The percentage remaining in the incubation tube was in the range 0.34-0.52. Therefore, 98-99% passed through the filter. Using viable bacteria alone, the average recovery after incubation for three hours at 20C in M-199 or deionized water was 101.3%. The effects of normal and immune serum on the ability of macrophages to bind viable gram negative bacteria and endotoxin was documented semiquantitatively by several investigators (10,19,72,89,ll3). A modification of the suspension-centrifugation technique was used in this study to observe this phenomenon in a more quantitative fashion. All experimental runs were performed as described in Materials and Methods. Effects of M-199,7M-199-Normal Rat Serum and M-l99-Immune Rat Serum on the Ability of Rat Peritoneal Cells to Bind Viable Salmonella typhimurium AfterIFive Minutes Rat peritoneal cells were treated with M-199 alone, M~199 normal rat serum or M-199-immune rat serum. 50 No statistically significant difference was found in the ability of peritoneal cells to bind viable bacteria after five minutes when treated with M-199 alone or M-199-normal rat serum. However, peritoneal cells treated with immune rat serum demonstrated statistically significant enhancement over M-199 alone or M-199-normal rat serum in their ability to bind viable bacteria after five minutes (see Table 3). Table 3. Comparison of the effect of M-199, normal rat serum, and immune rat serum on the ability of rat peritoneal cells to bind viable Salmonella typhimurium after five minutes.a Percent of Bacteria in the: Cells treated Filtrate Filter Recovery with: M-199 97.6(1 1 19.6 12.0d : 4.61 109.6 1 14.9 b b Normal Rat Serum 80.0 [+- 24.4 20.8 1 12.4 105.5 f. 19.3 Immune Rat Serum 45.5C 1 24.6 45.9C 1 20.5 87.86 -_+_ 21.5 aAverage value from at least six experimental determinations. bP> 0.05 — M-199 vs. Normal Rat Serum cP<0.05 - Normal Rat Serum vs. Immune Rat Serum dp< 0.001 - M-199 vs. Immune Rat Serum eP;;0.05 Recovery, M-199 vs. Immune Rat Serum Peritoneal cells treated with M—199 also bound 12.0% of the bacteria while 97.6% was in the filtrate. M-199-normal rat 51 serum treated peritoneal cells bound 20.8% of the bacteria with 80.8% free in the filtrate. Immune serum treated peritoneal cells bound 43.9% of the bacteria with 45.5% free in the filtrate. There was a statistically significant difference in total recovery between M-199 alone and M-199- immune rat serum (_P__'\_a_0.05). Effects of M-lQQ-Heat-killed Cell Absorbed Normal Rat and M-I991Heat-killéd Cell Absorbed Immune Rat Serum on the Ability of Peritoneal Cells to Bind Viable Salmonella typhimurium SRll Several milliliters of normal rat serum and immune rat serum were absorbed with large amounts of heat-killed Salmonella typhimurium. The effect of this serum on the ability of peritoneal cells to bind viable bacteria was studied. When compared with the results from nonabsorbed serum, the number of bacteria bound to macrophages was not statistically significant (P>’0.05). However, there was a statistically significant difference in the number of bacteria free in the filtrate (P< 0.05, P< 0.05) (see Table 4 on the following page). Peritoneal cells treated with normal absorbed serum bound 21.1% of the bacteria with 105.4% free in the filtrate, MacrOphages treated with immune absorbed serum bound 22.6% of the bacteria with 74.7% free in the filtrate. When com« pared with normal rat serum the peritoneal cells treated 52 Table 4. Comparison of the effect of absorbed and non- absorbed normal rat serum and immune rat serum on the ability of peritoneal cells to bind viable Salmonella typhimurium.3 Percent of Bacteria in the: Treatment FiItrate Filter Recovery Normal 80.0 1 24.4 20.8 1 12.4 101.5 1 19.3 Normal 105.4b1 35.7 21.1C1 18.7 121.3 1 34.4 Absorbed Immune 45.5 1 24.6 43.9 1 20.5 87.8 1 21.5 Immune 74.7d1 23.6 22.6e1 18.2 97.8 1 11.95 Absorbed aAverage value from at least six experimental determinations. bP<0.05 Normal vs. Normal Absorbed CP>-0.05 Normal vs. Normal Absorbed dP<:0.Ol Immune vs. Immune Absorbed eP>'0.05 Immune vs. Immune Absorbed with absorbed normal rat serum had a statistically signifi« cant increase (P510.05) in the number of bacteria free in the filtrate. Peritoneal cells treated with immune absorbed serum showed a statistically significant increase (P< 0.01) in the number of bacteria free in the filtrate when compared with immune serum treated macrophages. 53 Effect of Macropha e Absorbed Immune Serum on the Ability of Per1tonea1 Cells to Bind ViaBle Salmonella typhimurium SRII After Five Minutes One-half milliliter aliquots of antisera from three different pools were absorbed with l—l.5 X 107 macrophages and the amount of bacterial binding was quantitated both before and after absorption. No statistically significant difference (P> 0.05) was found in the amount of bacterial binding between immune and macrophage absorbed immune serum (see Table 5). Table 5. Comparison of the effect of immune serum and macrophage absorbed immune serum sensitized peritoneal cells to bind viable Salmonella typhimurium after five minutes.a Percent of Bacteria in the: Cells treated Filtrate Filter Recovery with: Immune Serum 38.5 1 20.7 47.1 1 23 83.0b 1 19.8 Macrophage b b b Absorbed Serum 51.6 1 20.3 47.7 1 15.7 99.3 + 20.0 aAverage value from at least six experimental runs. b1>>'0.os. 54 Effects of M—199, M-199-Normal Rat Serum and M—I99-Immune Rat Serum on the_Abili§y of Rat Peritoneal Cells to Bind bICr-laheled— Salmonella gyphimuriumlEndotoxin After Five Minutes Experimental runs were performed to observe the effects of M-199, M-199-normal rat serum and M-199-immune rat serum 51Cr—labeled on the ability of rat peritoneal cells to bind endotoxin. There was no statistically significant differ- ence (P> 0.05) between normal rat serum and immune rat serum treated cells. No statistically significant differ« ence (P>’0.05) was found between M-199 and M«199—immune rat serum treated peritoneal cells (see Table 6). Table 6. Comparison of the effect of M—199, M-199—normal rat serum and M-199-immune rat serum on the ability of rat peritoneal cells to bind 51Cr- labeled endotoxin after five minutes.a Percent of Endotoxin in the: Treatment Filtrate 'Filter 14-199 99.51 1 0.295d 0.1797 1 0.378C1 M-199-Normal b b Rat Serum 99.11 1 0.424 0.315 1 0.236 M-199-Immune c c Rat Serum 98.45 1 0.28 0.47 1 0.59 aAverage value from at least six experimental runs. bP>>0.OS M-l99 vs. M-199-Norma1 Rat Serum cP>0.95 Normal Rat Serum vs. Immune Rat Serum dP> 0.05 14-199 vs. M-199-Immune Rat Serum 55 Effects of M-199-Normal and M—199-Immune Rat Serum on the Ability oprouse Peritoneal Cells to Bind Viable Salmonella typhimuriam SRII After Five Minutes and'ThreeiHours Experimental runs were performed using mouse peritoneal cells treated with.M—199-rat serum to determine their abil- ity to bind viable Salmonella typhimurium SRll. A statis- tically significant difference (P1:0.01) was observed between M-199nnormal and M-1991immune serum treated perin toneal cells. In addition, a statistically significant difference was found between five minutes and three hours with M-199-normal rat serum treated peritoneal cells (see Table 7). Table 7. Comparison of normal and immune rat serum on the ability of mouse peritoneal cells to bind viable Salmonella typhimurium After Five Minutes and Three Hours.a Percent of bacteria in the filtrate: Cells treated with: 5 minutes 3 hours Normal Rat Serum 99.6a 1 21.0 34.1d 1 26.1 Immune Rat Serum 22.0b 1 10.8 31.9c'e124.3 aAverage value from at least five experimental determina- tions. b P1<0.01 IRS 5 minutes vs. NRS 5 minutes cP>0.05 IRS 4 hours vs. NRS 3 hours dP<0.01 NRS 5 minutes vs. NRS 3 hours e1>> 0.05 IRS 5 minutes vs. IRS 3 hours 56 Mouse peritoneal cells treated with M-l99 NRS left 99.6% of the bacteria free in the filtrate while mouse cells treated with M-199-immune rat serum left 22% free in the filtrate after five minutes. After three hours mouse peri- toneal cells treated with M-199-normal rat serum left 34.1% free in the filtrate and mouse cells treated with M-l99« immune rat serum left 31.9% of the bacteria free in the fil- trate. There was a statistically significant difference (P<=0.01) in the number of bacteria free in the filtrate between M-199-normal rat serum and M-199—immune rat serum treated cells after five minutes. In addition, there was a statistically significant difference in the number of bacteria free in the filtrate with peritoneal cells treated with M-l99—normal rat serum between five minutes and three hours. When compared with sensitized rat peritoneal cells treated with M-199-normal rat serum, sensitized mouse cells showed no statistically significant difference (P>‘0.05) in the ability to bind viable bacteria after five minutes. However, mouse cells sensitized with M-l99-immune rat serum had a statistically significant difference (P< 0.05) in their ability to bind viable bacteria after five minutes when compared with sensitized rat cells. DISCUSSION As mentioned earlier, there are several difficulties with earlier techniques for measuring cytophilic antibody, especially the macr0phage monolayer technique. It is dif- ficult to quantitate because not only each rosette-cell but also the number of antigen particles on each rosette must be counted. In addition, the counting must be performed in a short time period (3-5 minutes) following addition of antigen. Neither binding to glass or simple mechanical lodgement of bacteria next to the rosette-forming cell can be excluded. More importantly the number of macrophages with bacteria adherent are counted and not the percentage of bacteria added. For these reasons a relatively simple, more quantita- tive method was developed using Millipore vacuum filtration of a suspension of sensitized cells and antigen. In addi- tion, the mechanical stress applied to the macrophages during the experimental procedure allows only those bacteria relatively tightly bound to be counted. It is hOped that this will alleviate some of the difficulties encountered with earlier methods. 57 58 The statistically significant difference found between the ability of normal and immune rat serum to sensitize peritoneal cells to bind viable S. typhimurium (cf. Table 1). This was not unexpected in view of the known ability of sensitized peritoneal cells to bind gram-negative bacteria in vitro by less quantitative methods (81,89). Parish re- ported phagocytosis, not binding, of heat-killed bacteria with antisera to somatic polysaccharide antigen, not whole bacteria as was done in this study. He did not report the percentage of phagocytosis. Mittal found a much higher per- centage of binding (91—95%) than was found in the present study (~44%). He used the arbitrary number of two bacteria per macrophage to indicate cytophilic binding. Again, only semiquantitative measurements of the number of macrophages with adherent bacteria were made and percentage of the total bacteria bound was not reported. Both studies used Escherichia coli as the test bacteria. Another important technical point is the route and method of immunization of the eliciting antigen. This is important because there is considerable variation in the amount of true cytOphilic antibody formed with different techniques. Careful consideration of these facts is neces— sary for accurate interpretation of data on cytophilic anti- body. Immunization with a living attenuated strain of Salmonella sp. is the best and most reliable method of 59 obtaining macrophage cytophilic antibody (85,95). This method also approximates an in vivo infection more closely than other methods. Heat-killed cells given with complete Freund's adjuvant is the second best way to obtain macro- phage cytophilic antibody. Intravenous injection of heat-killed cells is the third best method (107). The use of heat-killed cells given intraperitoneally leads to a less satisfactory cytophilic antibody response in mice and presumably also in rats. Unfortunately, this method of immunization was used in this study. It is likely that the results obtained reflect a minimal level of cytophilic activity and would be more striking if a more reliable method of immunization had been used. It is interesting to note the level of cytophilic bind- ing in each set of experimental runs. When M-199 alone was used, 12% of the bacteria were bound. It is likely that this represents non-specific attachment of bacteria to the macrophage surface due to non-specific cytOphilic antibody or specific cytophilic antibody produced in low titer by bowel flora in normal serum that reaches the peritoneal cavity by diffusion or otherwise. Peritoneal cells directly sensi- tized with M-199-normal rat serum had a slightly higher level of bacterial binding (~21%). Although not statis- tically significant it is likely that this represents 60 non-specific and/or specific cytophilic antibody found in normal serum in a greater concentration than that which is reached in the peritoneal cavity. Macrophages sensitized with immune rat serum showed a much higher level of bacteria binding (~44%). It is likely that this reflects specific macrophage cytophilic antibody for Salmonella found in immune serum. Auzins and Rowley (6) found only a slightly higher level of phagocytosis (m50%) of S. typhimur— ium C5 which had been preopsonized with fresh pig serum. Both normal and immune sera were absorbed with very large numbers of heat-killed Salmonella typhimurium. A statistically significant difference was found in the free-unbound fraction (filtrate) of bacteria with both normal (P< 0.05) and immune (P< 0.01) rat serum (cf. Table 2). The difference found in the cell-bound fraction (filter) was not statistically significant. It is likely that this lack of significance with cell-bound bacteria was due to phagocytosis of bacteria by immune sera sensitized macro- phage (Recovery--88%). Capron et al. (1975) reported a dramatic decrease (95% to 5%) in macrophage adherence to Schistosoma mansoni schistosomules following absorption of rat immune sera by soluble S. mansoni antigens. It has been reported by another group of investigators that a large portion (89-95%) of the total amount of immunoglobulin present can be removed by similar procedures (94). 61 They also found absorption of antibody across almost the entire immunoglobulin range. These findings are supportive of the Speculations of Berken and Benacerraf (10) on the nature of cytOphilia; that is, cytOphilic antibodies are really Opsonizing antibodies for whole bacterial cells which also have the ability to attach to macrophages either before or after combination with antigen. Because of the large number of bacteria used (1012) it is possible that the adsorption was due to mechanical or non-specific factors rather than specific immunologic adsorp- tion. Also, it should be mentioned again that the method of immunization used in this study and Berken and Benacerraf (10) is less than Optimal for the production of cytOphilic antibody. It is likely that the sera in both studies con- tained mostly classical opsonic antibody rather then cell- bound antibody. This point is very important because there is still controversy over the differentiation of cytophilic and opsonic properties. Much of the controversy appears to be due to the differences in method of immunization and subsequent antibody produced. This study is the first attempt to evaluate the residual cytOphilic binding prOper— ties of sera absorption with gram-negative bacteria. Immune serum was absorbed with peritoneal cells and the percentage of binding before and after absorption was compared. Aliquots of immune sera from three different 62 pools were used. The difference in binding was not statis- tically significant (P> 0.05). However, it is possible that phagocytosis occurred with immune—sera—sensitized macro- phages (recovery 91%) or that not enough macrophages (~1.2 X 107/run) were used. Berken and Benacerraf (10) adsorbed guinea pig Yz-globulin onto lung macrophages. They found that uptake of antibody was maximal at a serum dilution of 1:2 and quantitatively estimated at 7—8 ug/lO7 cells. The amount of antibody adsorbed each time was relav tively constant. The results in this study suggest similar findings. Their interpretation was that cytophilic activ- ity was present in a large portion of the yz-globulin frac- tion. The results found in both studies suggest a preponderence of opsonic-cytophilic antibody in the anti- sera used and have only limited value in differentiating opsonic from cytophilic antibodies. The ability to bind 51Cr-labeled endotoxin to macro- phages sensitized with M1199, normal rat serum and immune rat serum after five minutes was tested. Very little endo~ toxin was bound in any experiments. Parish (89) reported adherence of erythrocytes coated with E. coli somatic poly- saccharride to macrophages sensitized with antisera to E. coli somatic polysaccharride. This same antiserum pro- moted the phagocytosis of killed E. coli (presumably at 37C). The cytophilic antibody was found to be a 198 macro“ globulin. 63 Berken and Benacerraf (10) reported cytophilic binding of E. coli polysaccharride or its dinitrophenol hapten by the passive direct technique with guinea pig antisera pro— duced by heat-killed E. coli or DNP proteins in complete Freund's adjuvant. They found it was necessary to use undi— luted antisera to obtain 3+ to 4+ rosette formation. This phenomena was attributed to insufficient coating of the red blood cell surface with polysaccharride. It is possible that this caused the very low levels of binding found in this study. Levenson and Brands (72) reported adherence (33-40% of macrophages) of erythrocytes coated with S. typhi Vi anti— gen in the presence of early antisera (3—4 days) to Vi antigen at 37C. Normal red cells and red cells coated with S. typhi 0 antigen (Boivin antigen) did not bind to macro- phages in the presence of Vi antisera. Peritoneal macro« phages sensitized by the passive direct method failed to bind erythrocytes coated with Vi antigen. Late antisera (7n8 days) enabled macr0phages to phagocytose antigen coated red blood cells. In summary, macrophages sensitized with early sera (3-4 days) bound S. typhi Vi coated erythro- cytes. Macrophages sensitized with late antisera (7-8 days) phagocytosed S. typhi Vi coated erythrocytes. It is likely that the failure to demonstrate cytophilic binding of endotoxin in this study was due to a lack of 64 specific cytophilic antibody for endotoxin in antisera ob— tained from intraperitoneal injection of heat-killed S. typhimurium. However, Robbins et al. (94) reported IgG antiendotoxin antibody in rabbits immunized with heat- killed S. typhimurium. They used the intravenous route of injection. This study was the first to observe the cyto- philic binding of free endotoxin. Mouse peritoneal cells were sensitized with normal and immune rat serum. A comparison of the ability of mouse and rat peritoneal cells to bind viable S. typhimurium indicates that mouse cells sensitized by immune rat serum had a statistically significant increase (P<:0.05) in the ability to bind viable S. typhimurium after five minutes. A comparison of binding of bacteria to mouse cells between normal and immune rat sera indicates a significant statis- tical difference (P< 0.01). There was a significant statis- tical difference (P< 0.01) between five minutes and three hours with mouse cells sensitized with normal rat serum, but no difference was found with immune rat serum (P> 0.05) (cf. Table 7). There was no statistical difference at three hours between normal and immune rat serum sensitized mouse cells (P> 0.05). Other studies of interspecies activity of cytOphilic antibody have shown high levels of binding with homologous cells and lower levels of binding with heterologous cells 65 (10,69). The close genetic relation between the mouse and the rat would account for the significant binding found in both species. Perhaps the increased binding with mouse cells was due to the fact that the mouse is the normal host for S. typhimurium. The clearance of bacteria at three hours is due to binding and/or phagocytosis. It is interesting to note the mouse cells sensitized with both normal and immune rat serum had similar levels of free bacteria (filtrate) (34.1% and 41.9%), after three hours. This suggests that over a period of time both normal and immune serum sensitized peritoneal cells have the ability to bind and/or phagocytose 70-75% of the total bacteria present. This was the first comparison of heterologous cytophilic activity between mouse and rat peritoneal cells. This thesis represents a partial reevaluation of studies similar in some ways to the present one using more quantitative and clearly defined techniques. The ability of sensitized peritoneal cells to bind viable bacteria has been demonstrated in a way that is not only more quantitative but is based on the total bacteria used, not on the number of cells forming rosettes with an arbitrary number of bacteria or antigen particles. As men- tioned earlier the differentiation between opsonic and cytophilic antibodies is not resolved and the importance of 66 using reliable methods for producing antisera for these studies is emphasized. Furthermore, previous studies on the ability to bind endotoxin were not clearly reported and did not use free endotoxin, as would exist in vivo. The level of cytophilic activity measured in this study very likely reflects a minimal level of activity. More striking results would have been obtained by using a live attenuated strain of bacteria for immunization and using a lower titer of cytophilic antibody for sensitiza- tion. In addition, the antibacterial effect of macrophage clumping due to cytophilic antibody was not measured in these studies. BIBLIOGRAPHY BIBLIOGRAPHY Actor, P. and D. Pitkin. 1973. Immunity to Vibrio cholerae in the Mouse. II. Effect of a Cell Adherent Immune Factor. Infec. Immun. 11:35-38. Ada, G. L., C. R. Parish, G.J.V. Nossal and A. Abbot. 1967. The Tissue Localization, Immunogenic, and Tol- erance-Inducing Properties of Antigens and Antigen- Fragments. Cold Spr. Harb. Symp. Quant. Biol. ii: 381. Allen, C., Saba, T. M, and J. Molnar. 1973. Isolation, Purification and Characterization of Opsonic Protein. J. Reticuloendothelial Soc. 11:410-423. Allerhand, J. and C. M. Zitrin. 1967. Detection of Antibodies in the a-Z and y-Globulin Fractions of Human Tuberculous Sera. J. Immunol. 12:252. Amos, H. E., B. W. Gurner, R. J. Olds, and R.R.A. Coombs. 1967. Passive Sensitization of Tissue Cells. II. Ability of Cytophilic Antibody to Render the Migration of Guinea Pig Exudate Cells Inhibitable by Antigen. Int. Arch. Allergy. ggz4l7-428. Auzins I. and Rowley, D. 1963. Factors Involved in the Adherence of S. typhimurium C5 and Mouse Peritoneal Macrophages. Aust. J. EXP- Biol. Med. Sci. 111539. Balfour B. M. and J. H. Humphrey. 1967. Localization of Y'Globulin and Labeled Antigen in Germinal Centers in Relation to the Immune Response, p.80-85. In: Germinal Centers in Immune Responses. H. Cottier, N.0d0rtcfienko, R. Scheidler,'P. C. Cbnngn. Eds. Springer-Verlag. Berlin-Heidelberg. Beard, J. W. and P. Rous. 1934. The Characteristics of Kupffer Cells Living In Vitro. J. Ihqx Med. £2:593- 607. 67 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 68 Benacerraf, B. 1968. Cytophilic Immunoglobulins and Delayed Hypersensitivity. Fed. Proc. ZZF46'48- Berken, A., and B. Benacerraf. 1965. Properties of Antibodies Cytophilic for Macrophages. J. Exptl. Med. 123:119-144. Berken, A. and B. Benacerraf. 1968. Sedimentation Properties of Antibody Cytophilic for Macrophages. J. Immunol. 100:1219-1222. Biddle, F. and K. F. Shortridge. 1967. Immunological Cross-Reactions of Influenza Virus Inhibition. Brit. J. Exptl. Pathol. g§:285. Blanden, R. V., G. B. Mackaness, M. B., and F. M. Collins, Ph.D. 1966. Mechanisms of Acquired Resistance in Mouse Typhoid. J. Exptl. Med. 124:585-600. Blazkovec, A. A., L. Hulliger, and E. Sorkin. 1968. A Study of Local Cutaneous Hypersensitivity. III. Transfer of Hypersensitivity to Sheep Erythrocytes with Different Cell Types. Int. Arch. Allergy. §§:259-280. Boyden, S. V. 1957. The Effect of Previous Injection of Tuberculo-Protein on the Development of Tuberculin Sensitivity Following BCG Vaccination in Guinea Pigs. Brit. J. Expl. Path. §§:611. Boyden, S. V. 1963. Cytophilic Antibody, p7-l7. In: Cell-Bound Antibodies. B. Amos and H. Koprowski. Eds. Wistar Institute Press, Philadelphia. Boyden, S. V. 1964. Cytophilic Antibody in Guinea Pigs with Delayed Hypersensitivity. Immunology 1:474-483. Boyden, S. V., and E. Sorkin. 1960. The Adsorption of Antigen by Spleen Cells Previously Treated with Antiserum in vitro. Immunology. §:272-283. Brumfitt, W., Glynn, A. Percival, A. 1965. Factors Influencing the Phagocytosis of E. coli. Brit. J. Exp. Path. 19:215. Burrell, R. G., M. Rheins and J. M. Birkeland. 1956 Tuberculous Antibodies Demonstrated by Agar Diffusion. II. Further Characterization of These Antibodies and their Distribution in Human Tuberculous Sera. Am. Rev. Tuberc. Pulmonary Diseases. 11:239. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 69 Capron, A., M. Capron, H. Dupas, D. Bout and A. Petitprez. 1974. In vitro Study of Immunologic Phenomena in Human and Experimental Schistosomiasis: Comparative in vitro Study of Lethal Effects of Immune Sera on Immature and Adult Forms of Schistosoma mansoni. Int. J. Parasitol. 6:613-624. Capron. A., J. P. Dessaint and M. Capron. 1975. Specific IgE Antibodies In Immune Adherence of Normal Macrophages to Schistosoma mansoni Schistosomules. Nature. 253: 474-475. Changeux, J. P., J. Thiery, Y. Tung and C. Kittel. 1967. On the Cooperativity of Biological Membranes. Proc. Nat. Acad. Sci. U.S.A. g1:335. Chedid, L., R'ST' Skarnes and M. Parant. 1963. Charac- terization of Cr-labeled Endotoxin and the Identification in Plasma and Urine After Parenteral Administration. J. Exp. Med. 111:561-571. Cohen, B. E., A. S. Rosenthal and W. E. Paul. 1973. Antigen-Macrophage Interactions. II. Relative Roles of Cytophilic Antibody and Other Membrane Sites. J. Immunol. 111:820-828. Cole, L. R. and C. B. Favour. 1955. Correlations Between Plasma Protein Reactions, Antibody Titers, and the Passive Transfer of Delayed and Immediate Cutaneous Reactivity to Tuberculosis PPD and Tuberculopolysacchar- rides. J. Exptl. Med. 1ng391. Collins, F. M. and G. B. Mackaness. 1968. Delayed Hyper- sensitivity and Arthus Reactivity in Relation to Host Resistance in Salmonella-Infected Mice. J. Immunol. 101:830-845. Coulson, A. S. B. W. Gurner and R.R.A. Coombs. 1967. Macrophage-Like Properties of Some Guinea Pig Trans- formed Cells. Int. Arch. Allergy. §§:264-279. Dardas, T. J. and V. H. Mallmann. 1966. Electrophoretic and Immunoelectrophoretic Studies of Sera from Normal Tuberculous, and Noninfected Tuberculin Sensitive Guinea Pigs. J. Bacteriol. 2;:76. Davey, M. J. and G. L. Asherson. 1967. Cytophilic Antibody. 1. Nature of the Macrophage Receptor. Immunology. 11:13-20. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 70 Del Guerico, P., G. Tolone, F. Broga Andnide, G. Biozzi and R. A. Binaghi. 1969. Opsonic, Cytophilic and Agglutinating Activity of Guinea Pig 72 ande Anti- Salmonella Antibodies. Immunology 19:361-371. Deutsch, H. P., J. L. Nichol and M. Cohn. 1949. Biophysical Studies of Blood Plasma Proteins. XI. Immunological and Electrophoretic Studies of Immune Chicken Serum. J. Immunol. g1:195. Unpuy. J. M., D.Y.E. Perey, R. A. Good. 1969. Passive Transfer with Plasma of Delayed Allergy in Guinea Pigs. Lancet. 1:551-553. Eisenstein, T. K. 1975. Evidence for O Antigens as the Antigenic Determinants in "Ribosomal" Vaccines Prepared from Salmonella. Infec. Immun. 11:364-337. Furness, G. and I. Ferreira. 1959. The Role of Macrophages in Natural Immunity to Salmonellae. J. of Infect. Disease. 104:203. Gaullet, Ph. 1965. Presence d'un a -Globu1in Lente ase Cours de la Polyarthnite Immunologique Diterminee por l'Adjuvant de Freund. Experientia 11:313. Gibbs, H. , S. Hruby , E. C. Alvard, and C. Shaw. 1970. Relationship Between Antibodies and Experimental Allergic Encephalomyelitis. IV. Cytophilic Antibodies Int. Arch. Allergy. 11:394-402. Gill, P. G. 1969. Cytophilic Antibody in Heterlogous Antilymphocytic Serum: Its Presence, Absorption and Possible Bioligic Significance. J. Immunol. 191: 1329-1331. Gill, P. G. 1971. Production of Immunosuppressive Potency of Antilymphocyte Serum by in vitro Testing. J. Immunol. 107:244-251. Girard, K. F., and E.G.D. Murray. 1954. The Presence of Antibody in Macrophage Extracts. Canadian Journal of Biochemistry and Physiology. 11:14-19. Gordon, B. L. 1967. The Case for Cytophilic Antibodies in Cellular Immunity. Ann. Allergy. 11:1-5. Gowland, E. 1968. The Possible Role of Cytophilic Antibodies as the Specific Mediators of Delayed-Type Hypersensitivity in Guinea Pigs. Aust. J. Biol. Med. Sci. 19:83-92. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 71 Granger, G. A. and R. S. Weiser. 1964. Homograft Target Cells: Specific Destruction in vitro by Contact Interaction with Immune Macrophages. Science. 115:1427-1429. Granger, G. A. and R. S Weiser. 1966. Homograft Target Cells: Specific Contact Destruction in vitro by Immune MacrOphages. Science. 151:97-99. Grey, H. M. and H. G. Kunkel. 1967. Heavy Chain Classes of Human yG-Globulin. Peptide and Immuno- chemical Relationships. Biochemistry. 1:2326-2334. Hasek, M., Lengerova, A. and Hraba, T. 1961. Transplantation Immunity and Tolerance. pl-66 In: Advances in Immunology, Vol. 1. W. H. Taliaferro and 3. H. Humphrey, Eds. Academic Press. New York. Higgins, G. M and T. Murphy. 1928. The Phagocytic Cell (v. Kupffer) in the Liver of Common Laboratory Animals. Anatomical Record. 19:15-39. Holtzer, J. D. 1967. Experimental Delayed Type Hyper- sensitivity without Demonstrable Antibodies. II. In vitro Activity of Reticuloendothelial Cells. Immun- ology. 11:712-723. Holtzer, J. D. and K. C. Winkler. 1967. Experimental Delayed Type Allergy without Demonstrable Allergic Antibodies. 1. Absence of Cytophilic Antibodies. Immunology. 11:701-712. Howard, J. G. and B. Benacerraf. 1966. Properties of Macrophage Receptors for Cytophilic Antibodies. Brit. J. Exp. Pathol. 11:192-200. Howard, J. G. and A. C. Wardlow. 1958. The Opsonic Effect of Normal Serum on the Uptake of Bacteria by the RES. Perfusion Studies with Isolated Rat Liver. Immunology. 1:338-352. Hoy, S. E. and D. 3. Nelson. 1969. Studies on Cytophilic Antibodies. V. Alloantibodies Cytophilic for Mouse Macrophages. Aust. J. Exp. Biol. Med. Sci. 11:525-539. Hsu, H. S. and D. R. Mayo. 1973. Interactions Between Macrophages of Guinea Pigs and Salmonellae: III. Bacteriocidal and Cytophilic Antibodies of Infected Guinea Pigs. Infect. Immun. 1:165-177. S4. 55. 56. 57. 58. 59. 60. 61. 62. 63. 72 Hsu, H. S. and V. M. Piper. 1972. Acquired Resistance to and Comparative Virulence of Salmonella typhimurium Demonstrated by Cutaneous Lesions in Guinea Pigs. J. Reticuloendothelial Soc. 11:343-357. Huber, H. and H. H. Fudenberg. 1970. The Interaction of Monocytes and Macrophages with Immunoglobulins and Complement. Ser. Haemat. 1:160-175. Humphrey, J. H. and I. Mota. 1959. The Mechanism of Anaphylaxis: Specificity of Antigen Induced Mast Cell Damage in Anaphylaxis in the Guinea Pig. Immunology. 2:19-31. Inchley, C., H. M. Grey, and J. W. Uhr. 1970. The Cytophilic Activity of Human Immunoglobulins. J. Immunology. 105:362-369. Ishizaka, K. and D. H. Campbell. 1959. Biologic Activity of Soluble Antigen-Antibody Complexes. V. Change in Optical Rotation by the Formation of Skin Reactive Complexes. J. Immunology. 11:318-326. Ivanyi, J. 1970. Cytophilic Antibodies in Passive Antibody Induced Suppression or Enhancement. Nature. (London) 226:550-551. Jenkin, C. R. and D. Rowley. 1968. Basis for Immunity to Typhoid in Mice and the Question of "Cellular" Immunity. Bacterial Rev. 11:391-404. Jenkin, C. R. D. Rowley and I. Auzins. 1964. The Basis for Immunity to Mouse Typhoid. I. The Carrier State. Aust. J. Exp. Biol. Med. Sci. 11: 215-228. Jeunet, F. S. and R. A. Good. 1967. Reticuloendothelial Function in the Isolated Perfused Liver. I. Study of Rates of Clearance, Role of a Plasma Factor, and the Nature of RE Blockade. J. Reticuloendothelial Soc. 1:351-369. Jonas, W. E., B. W. Gurner, D. S. Nelson and R.R.A. Coombs. 1965. Passive Sensitization of Tissue Cells. 1. Passive Sensitization of Macrophages by Guinea Pig Cytophilic Antibody. Int. Arch. Allergy. 11: 86-104. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 73 Keller, R. and E. Sorkin. 1963. Adsorption of Rabbit Antibody by Rat Tissue Mast Cells and Other Cells in vitro. Nature. 199:709-710. Kessel, R.W.I. and Werner Braun. 1964. Endotoxin Effects on Monocytes in vitro. Bacteriological Proceedings. 1964. p.44. Knisely, M. H., E. H. Bloch and L. Warner. 1948. Selective Phagocytosis. I. Microscopic Observations Concerning the Regulation of the Blood Flow Through the Liver and Other Organs and the Mechanism and Rate of Phagocytic Removal of Particles from the Blood. K. Danske Vidensk. Sels. Skr. 1:1-93. Koprowski, H. and M. V. Fernandez. 1962. Autosensiti- zation Reaction in vitro Contactual Agglutination of Sensitized Lymph Node Cells in Brain Tissue Culture Accompanied by Destruction of Glial Elements. J. Exp. Med. 119:467-476. Kossard, S. 1966. Ph.D. Thesis. University of Sydney. Kossard, S., and D. S. Nelson. 1968a. Studies on Cytophilic Antibody III. Sensitization of Homologous and Heterologous Macrophages by Cytophilic Antibody and Inhibition by Normal Rat Serum. Aust. J. Exp. Biol. Med. Sci. 11:51-61. Kossard, S. and D. S. Nelson. 1968b. Studies on Cytophilic Antibody, IV.. Effects of Proteolytic Enzymes (Trypsin and Papain) on the Attachment of Cytophilic Antibody to Macrophages. Aust. J. Exp. Biol. Med. Sci. 11:63-71. Kurashiga, S., N. Osawa, K. Kawahami, and S. Mitsuhashi 1967. Experimental Salmonellosis. X. Cellular Immunity and Its Antibody in Mouse Mononuclear Phago- cytes. J. Bacteriol. 21:902-906. Levenson, V. I. and N. I. Brands. 1967. Different Forms of Macrophage Interaction with Foreign Antigens Provided by "Early" and "Late" Immune Sera. Folia. Biol. 11:113-121. Lurie, M. B. 1942. Studies on the Mechanism of Im- minity in Tuberculosis, the Fate of Tubercle Bacilli Ingested by Mononuclear Phagocytes Derived from Normal and Immunized Animals. J. Expt. Med. 121247-267. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 74 Mackaness, G. B., M. B., R. V. Blanden and F. M. Collins, Ph.D. 1966. Host-Parasite Relations in Mouse Typhoid. J. Exptl. Med. 124:573-583. Manwaring, W. H., and H. C. Coe. 1916. Endothelial Opsonins. J. Immunol. 1:401-408. Manwaring, W. H., and W. Fritschen. 1923. Study of Microbic- Tissue Affinity by Perfusion Methods, J. Immunol. 1: 83- 89. Marecki. N. M., H. s. Hsu, and D. R. Mayo. 1975. Cellular and Humoral Aspects of Host Resistance in Murine Salmonellosis. Brit. J. Exp. Path. 11:231-243. Margolis, J. M. and N. J. Bigley. 1972. Cytophilic Macroglobulin Reactive with Bacterial Protein in Mice Immunized with Ribonucleic Acid-Protein Fractions of Virulent Salmonella typhimurium. Infect. Immun. 1:390-399. Minden, P., J. K. McClatchy, and R. S. Farr. 1972. Shared Antigens Between Heterologous Bacterial Species. Infect. Immun. 6: 574- S82. Mitsuhoshi, S., K. Saito, N. Osawa, and S. Kurashiga. 1967. Experimental Salmonellosis. XI. Induction of Cellular Immunity and Formation of Antibody Transfer Agent of Mouse Mononuclear Phagocytes. J. Bacteriol. 11:907-913. Mittal, K. R. 1972. Production of Cytophilic Antibodies in Rabbits and Mice Against Escherichia Coli. Acta. Microbiol. Acad. Sci. Hung. 113 245-250. Moon, R. J. R. A. Vrable and J. A. Broka. 1975. In situ Separation of Bacterial Trapping and Killing 'Ffinctlons of the Perfused Liver. Infec.Immu.11: 411- 418. Munthe-Kaas, A. C., T. Berg., P. O. Seglen, and R. Seljelid. 1975. Mass Isolation and Culture of Rat Kupffer Cells. J. Exptl. Med. 141:1-10. Najjar, V. A. 1963. Some Aspects of Antibody-Antigen Reactions and Theoretical Considerations of the Im- munologic Response. Physiol. Rev. 11:243-262. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 75 Nelson, D. S. 1969. Antibodies Cytophilic for Macrophages. p63-88. In: Macrophages and Immunity. Neuberger, A. and E. L. Tatum, Eds. North Holland Publishing Co., London. Nelson, D. S. and A. V. Boyden. 1967. Macrophage Cytophilic Antibodies and Delayed Hypersensitivity. Brit. Med. Bull. 11:1:15-20. Nelson, D. S. and P. Mildenhall. 1968. Studies on Cytophilic Antibodies. II. The Production by Guinea Pigs of Macrophage Cytophilic Antibodies to Sheep Erythrocytes and H.S.A. Relationship to the Production of Other Antibodies and the Development of Delayed Type Hypersensitivity. Aust. J. Exp. Biol. Med. Sci. 46:33-49. Nossal, G.J.V., A. Abbot, J. Mitchell, and Z. Lummus. 1968. Antigens in Immunity. XV. Ultrastructural Features of Antigen Capture in Primary and Secondary Lymphoid Follicles. J. Exp. Med. 111:277. Parish, W. E. 1965. Differentiation Between Cyto- philic Antibody and Opsonin by a Macrophage Phagocytic System. Nature. 208:594-595. Patterson, R. J. and Youmans, G. P. 1970. Multiplication of Mycobacterium tuberculosis Within Normal and Immune Mouse Macrophages With and Without Streptomycin. Infect. Immu. 1:30. Pearmain, G., R. R. Lycette, and P. H. Fitzgerald. 1963. Tuberculin Induced Mitosis in Peripheral Blood. Lancet. 1 637-638. Pokorna.z. 1969. Induction of Autoimmune Aspermogenesis by Monocellular Cytophilic Material. Folia. Biol. 111173-180. Rhodes, M. w. and H. s. Hsu. 1973. Effect of Kanamycin on the Fate of Salmonella enteritidis Within Cultured Macrophages of Guinea Pigs. J. Reticuloendothelial. Soc. 15:1. Robbins, J. B., K. Kenny and E. Suter. 1965. The Isolation and Biological Activities of Rabbit YM' and yG-Anti-SaZmonella typhimurium Antibodies. J. Exptl. Med. 111:385-402. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 76 Rowley, D., K. J. Turner and C. R. Jenkin. 1964. The Basis for Immunity to Mouse Typhoid. 3. Cell-Bound Antibody. Aust. J. Exp. Biol. Med. Sci. 11:237-248. Saito, K., M. Nokano, T. Akigana and D. Ushiba. 1962. Passive Transfer of Immunity to Typhoid by Macrophages. J. Bact. 11:500. Sawyer, R., R. L. Moon, and E. S. Beneke. 1967. Infec. Immu. 11 Press. Seeger, R. C., J. J. Oppenheim. 1970. Synergistic Interaction of Macrophages and Lymphocytes in Antigen- Induced Transformation of Lymphocytes. J. Exptl. Med. 132:44. Seibert, F. R. 1960. A Theory of Immunity in Tuberculosis. Perspectives Biol. Med. 1:264. Smith, R. A. and N. J. Bigley. 1972. Detection of Delayed Hypersensitivity in Mice Injected with Ribonucleic Acid- Protein Fractions of Salmonella typhimurium. Infect. Immun. 1:384-389. Sorkin, E. 1964. On the Cellular Fixation of Cyto- philic Antibody. Int. Arch. Allergy. 11:129-144. Stanworth, D. R. 1970. Immunochemical Mechanisms of Immediate Type Hypersensitivity Reactions. Clin. Exp. Immunol. 1:1-12. Stanworth, D. R., J. H. Humphrey, H. Bennish and 8.6.0. Johansson. 1968. Inhibition of Prousnitz-Kustner Reaction by Proteolytic Cleavage Fragments of a Human Myeloma Protein of Immunoglobulin Class E. Lancet 11:17-18. Strejan, G. and I. Flechner. 1963. An Antibody with Alpha-l Globulin Mobility. Proc. Soc. Exptl. Biol. Med. 115:352. Taub, R. N. and E. M. Lance. 1968. Effects of Hetero- logous Antilymphocyte Serum on the Distribution of Cr51 Labeled Lymph Node Cells in Mice. Immunology. 11:633- 642. Tizard, I. R. 1969. Macrophage Cytophilic Antibody in Mice. Differentiation Between Antigen Adherence Due to these Antibodies and Opsonic Adherence. Int. Arch. Allergy. 11:332-346. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 77 Tizard, I. R. 1971a. Macrophage-Cytophilic Antibodies and the Functions of Macrophage-Bound Immunoglobulins. Bact. Rev. 11:365-378. Tizard, I. R. 1971b. Macrophage Cytophilic Antibodies in Mice: the Adsorption of Cytophilic Antibodies from Solution by Mouse Peritoneal Cells and Cooperative Inter- action Between Receptors for Immunoglobulin. J. Reticuloendothelial Soc. 11g449-460. Tizard, I. R. 1971c. Macrophage Cytophilic Antibody in Mice, Effect of Bacterial Lipopolysaccharide on the Uptake of Immunoglobulins by Mouse Peritoneal Cells. Infec. Immu. 1:472-477. Tizard, I. R. 197ld. Macrophage Cytophilic Antibody in Mice: Mechanism ofA£.tion of Bacterial Lipopolysaccharide on the Uptake of Immunoglobulins by Mouse Peritoneal Cells. Infec. Immu. 1:402-406. Tizard, I. R. and M. A. Soltys. 1971. Macrophage Cytophilic Antibodies to Trypansoma brucei in Rabbits Trans. Roy. Soc, Trop. Med. Hyg. 11:407-408. Turner, K. J., C. R. Jenkin and D. Rowley. 1964. The Basis for Immunity to Mouse Typhoid. 2. Antibody Formation During the Carrier State. Aust. J. Exp. Biol. Med. Sci. 11:229-236. Uhr, J. W. 1965. Passive Sensitization of Lymphocytes and Macrophages by Antigen-Antibody Complexes. Proc. Nat. Acad. Sci. 11;1599-l606. Ushiba, D. T. Nakoe, T. Akiyama, and Y. Kishimato. 1965. Characterization of "Clearance" Factor and "Cell-Bound" Antibody in Experimental Typhoid. J. Bacteriol. 11:1705-1712. Vlaovic, M. S., G. M. Buening and R. W. Loon. 1975. Capillary Tube Leukocyte Migration Inhibition as a Correlate of Cell-Mediated Immunity in the Chicken. Cellular Immunology. 11:335-341. Weiser, R. S., E. Heise, K. McIvor, S. H. Han and G. A. Granger. 1970. In vitro Activities of Immune Macro- phages. In Cellular Recognition. R. T. Smith and R. A. Good Eds. Appleton-Centruy-Crofts. New York. 117. 118. 119. 78 Wells, P. S. and H. S. Hsu. 1970. Interactions Between Macrophages of Guinea Pigs and Salmonellae. II. Phago- cytosis of Salmonella typhimurium by Macrophages of Normal Guinea Pigs. Infec. Immu. 1:145-149. White, Colin. 1952. The Use of Ranks in a Test of Significance for Comparing Two Treatments. Biometrics. 1:33-41. Zinsser, H. and J. H. Mueller. 1925. On the Nature of Bacterial Allergies. J. Exp. Med. 11:159-177.