Vet. Pathol. 13: 90-97 (1976)
The Respiratory Tract Immune System in the Pig 11. Associated Lymphoid Tissues
PATRICIA A. BRADLEY,F. J. BOURNEand P. J. BROWN Department of Animal Husbandry, University of Bristol, Bristol
Abstract. The distribution of immunoglobulins A, G and M were assessed in the bronchial lymph node and palatine tonsil of pigs from birth to maturity, using the immunoperoxidase technique. Immunoglobulin-containing cells appeared earlier in the tonsil than in the lymph node; this probably reflects the exposure of the tonsil to environmental antigens. IgG-containing cells predominated after the first 3 weeks of life, but the 1gA:lgG ratio was much higher in the tonsil than in the lymph node. Extracellular, reticular staining was visible in the central region of some follicles in addition to a few immunoglobulincontaining cells that resembled large lymphocytes or blast cells.
Studies in local immunity have led to increasing interest in the interrelationships of local and systemic immune systems. It has been shown that local administration of antigen may result in systemic as well as local antibody production. This may be caused by the dissemination either of antigen or of immunocompetent cells from the site of immunisation or lymphoid tissues in that area to other sites [6,9, 171. We decided to complement our study of respiratory tract immunity in the pig [7] with an examination of respiratorytract-associated lymphoid tissues. Two lymphoid tissues were examined : the bronchial lymph node, which is one of the major draining nodes of the bronchial tree, and the palatine tonsil, which is situated in a strategic position in the oropharynx at the entrance to the two most extensive mucous surfaces of the body - the intestinal and respiratory tracts. Materials and Methods Tissue specinlens. Thirty-four Large White pigs were used from a closed herd free of enzootic pneumonia. Twenty were killed at intervals from birth to 10 days and thereafter
240 -
200 160
-
120
-
IgG IgA IgM
91
BRADLEY/BOURNE/BROWN
-
Bronchial lymph node
...........A
0------a
ul
0 n
80-
Tonsil
120
-
80 -
1
3
4
5
6
10
12
15 16
20
Age, weeks
Fig. 1. Pig respiratory-tract-associated lymphoid tissues; mean of cell counts for IgA, IgG and IgM.
weekly in groups of two or three until 6 weeks. At 10 weeks two pigs were killed, two at 16, two at 20, one at 12, and one at 15 weeks. Two sows also were included in the study. Tissues were taken promptly after death from the palatine tonsil and the bronchial lymph node, fixed in 95% methanol, and processed according to the method Of SAINTE-MARIE [I 61. Preparation ofantisera. Antisera to porcine immunoglobulins A, G and M were prepared as described [4, 51. Immunoperoxidase technique. Preparation and use of peroxidase-antiserum conjugates and the quantitation of immunoglobulin-containing cells were carried out as described [7].
Results Bronchial Lymph Node Immunoglobulin-containing cells were first seen in 6-day-old pigs; these were initially IgA- and IgM-containing cells with a lag of 1 o r 2 days before
sow
92
BRADLEY/BOURNE/BROWN
Fig. 2. Subtrabecular sinus in lymph node of 10-week-old pig, incubated with horseradish peroxidase-anti IgG and stained for peroxidase activity. There are large numbers of immunoglobulin-containing cells resembling large lymphocytes. A number of macrophages also show specific staining and can be detected by their irregular shape, with cytoplasmic processes spanning the sinus (m). Fig. 3. Lymph node cortex of 10-week-old pig, incubated with horseradish peroxidaseanti IgM and stained for peroxidase activity. Specific staining is associated with the cytoplasm of a few cells resembling lymphoblasts, and is also extracellular in the central region of the follicle, where it forms a continuous, reticular pattern.
Respiratory Lymphoid Tissues
93
IgG cells appeared. Numbers of these cells were low until the pigs were 3 weeks old when a marked increase, mainly in the IgG class, occurred (fig. 1). Thereafter, IgG cells predominated, while IgA and IgM cell numbers remained low. IgA counts were consistently less than half the IgM values. Relatively few immunoglobulin-containing cells were identified morphologically as mature plasma cells, and those that were found were situated around the periphery of follicles and in the medulla. A large number of immunoglobulin-containing cells were found in or bordering subtrabecular sinuses. Sometimes it was difficult to distinguish cell types because they were closely packed. Apart from macrophages, which stained for immunoglobulin and were distinguished by their irregular shape, the predominant type of cell resembled the large lymphocyte, having a large nucleus and scant cytoplasm (fig. 2). Lymphoid follicles, appearing as circular to oval, densely packed accumulations of cells, were visible in a 10-day old pig; these increased in size as the pig aged, and a continuous reticular pattern of staining that appeared to be extracellular was a feature of many (fig. 3). Few immunoglobulin-containing cells were associated with the follicles, and these were mainly cells with a large, slightly eccentric nucleus and a narrow rim of cytoplasm (fig. 3). Follicles containing such cellular activity and possessing an outer collar of pale stained, non-immunoglobulin-containingsmall lymphocytes were considered to be secondary follicles with germinal centres. Staining of the same follicle for more than one immunoglobulin class revealed a number of follicles that showed multiple affinity, although others stained for one class only.
Tonsil Immunoglobulin-containing cells appeared when the pig was relatively young - 3 days old. As in the lymph node, IgG-containing cells appeared a day or two later than IgA and IgM. Until the pig was 3 weeks old, numbers of IgA- and IgM-containing cells were higher than IgG, but after this time IgG cells predominated (fig. 1). The cells varied in type, most resembling
Fig. 4. Tonsil of 8-week-old pig, incubated with horseradish peroxidase-anti IgM and stained for peroxidase activity. Several immunoglobulincontaining cells lie in the crypt epithelium between the epithelial cells. Fig. 5. Tonsillar crypt epithelium of 4-week-old pig, incubated with horseradish peroxidase-anti IgA. The crypt epithelium (e) is narrow and pseudostratified columnar. Several cells are specifically stained for IgA. A few immunoglobulincontaining cells are in the tonsillar parenchyma (p), but there is no invasion of this type of epithelium.
94
BRADLEY/BOURNE/BROWN
Table I . Pig respiratory-tract-associated lymphoid tissues: mean of ratios between immunoglobulin classes for all age groups
Mean of ratios for all age groups
Tissue
Bronchial lymph node Tonsil
IgA: IgG
IgA: IgM
0.1 0.6
0.6 I .4
plasma cells or large lymphocytes; initially they were scattered throughout the lymphoid tissue, but by the time the pig was a week or so old IgA cells were for the most part immediately beneath or within the crypt epithelium, whereas IgG and IgM cells were more scattered and more frequently found in association with germinal centres. Germinal centres were numerous and showed staining patterns similar to those in the lymph node. The crypts were usually lined with a stratified, squamous epithelium extensively invaded by all classes of small lymphocytes and immunoglobulincontaining cells that appeared to lie between the epithelial cells (fig. 4). This infiltration was especially marked in areas adjacent to a germinal centre, and in such cases it was often difficult to determine the limits of the epithelium. In addition, a second type of crypt epithelium was seen, generally near the neck of the crypt; this was a fairly shallow, pseudostratified columnar epithelium that showed very little cellular infiltration. Many of the epithelial cells showed distinct, finely granular staining for IgA and IgM, but not for IgG (fig. 5). Comparing the proportions of the different classes in bronchial lymph node and tonsil (table I ) it can be seen that although IgG predominates in both sites, the IgA: IgG ratio is much higher in the tonsil than in the lymph node. Discussion
Most of the immunoglobulin-containing cells in the bronchial lymph node resembled large lymphocytes and were located for the most part in the subtrabecular sinuses. These sinuses are the terminations of the afferent lymphatics [ I ] and as such drain sites of antigen stimulation [2, 31. It is possible that the large lymphocytes in the subtrabecular sinuses of the bronchial lymph node were derived from mucosal lymphocytes after antigenic stimulation, especially as the rise in cell numbers in the lymph node corresponded
Respiratory Lymphoid Tissues
95
to the rise in plasma cell numbers in the respiratory tract tissues [7]. If this were the case, the high proportion of IgG-containing cells in the lymph node may represent an important connection between local antigenic stimulation and systemic immunisation. However, the relatively low numbers of mature plasma cells in the bronchial lymph node suggest that the site of serum antibody formation is elsewhere, and recent evidence of ours indicates that the intestinal mucosa and, to a lesser extent, the respiratory tract mucosa [7] may contribute significant amounts of all three immunoglobulins. The distribution of extracellular reticular staining in the central region of follicles in both lymph node and tonsil resembled that described for antigen [ 131. Reticular localization of immunoglobulin has also been described in follicles [12, 14, 151, and this similar distribution of antigen and immunoglobulin suggests that antibody in close association with the processes of dendritic cells may be important as a mechanism for trapping and processing antigen. Although our findings showing that follicles stained for more than one immunoglobulin class differed from the findings in one study [8], others have reported double staining of follicles in human spleen and lymph node [12,15]. Such observations would be consistent with the theory that germinal centres represent an aggregate rather than a clone of cells. If exogenous immunoglobulins are localized in a follicle by virtue of their antibody properties, they would be expected to reflect the class heterogeneity of serum antibody. Immunoglobulin-containing cells appeared relatively early in the tonsil, when the pig was 3 days old, compared to 6 days old for bronchial lymph node and 6-7 days for respiratory tract mucosa [7]. This precocious development might be expected considering that the position of the tonsil in the oropharynx and its reticulated, permeable epithelium ensure that it undergoes constant antigenic exposure [ll]. The tonsil has no afferent lymphatics, but antigen has been shown to pass from the oral cavity via the crypt epithelium to the subepithelial lymphoid tissue where it localizes in follicles and in the cytoplasm of macrophages [18]. The tonsil of gnotobiotic piglets responded to oral administration of bacterial antigen by germinal centre formation but without proliferation of cells of the plasma cell series [3]. In the present study, probably as a result of continuous antigenic stimulus [l I], germinal centres and immunoglobulin-containing cells were conspicuous in the tonsil. Most of the IgA-containing cells were closely associated with the crypt epithelium, whereas IgM and IgG cells showed a more general distribution between the follicles, and were more frequently associated with germinal centres. This distribution of the classes resembled that reported for the human tonsil [lo].
96
BRADLEY/BOURNE/BROWN
An interesting finding was the association of IgA and IgM staining with crypt epithelial cells. This type of epithelium differed from the normal, heavily cell-infiltrated, stratified squamous type of epithelium in that it appeared to be pseudostratified and columnar in form. It may represent a route for secretion of these two immunoglobulin classes into the crypt lumen to protect the tonsil against pathogens, many of which may initially lodge in the tonsillar crypts. It is clear that the immunological role of the tonsil has yet to be fully elucidated, but it is likely to be complex, involving both local and systemic immunity. Acknowledgements The authors thank Prof. T. K. EWER,in whose department the work was carried out, and the staff of the Photographic Unit for their help. Supported by grants from the Wellcome Trust and the Agricultural Research Council. This work was part of the swine fever control program financed by the European Economic Community.
References 1 ANDERSON, J.C.: The micro-anatomy of the lymph node of germ-free piglets. Br. J. exp. Path. 53: 37-39 (1972). 2 ANDERSON, J.C.: The use of germ-free piglets in the study of lymphoid tissue and germinal centre formation. Adv. exp. med. Biol. 29: 643-649 (1973). 3 ANDERSON, J. C. : The response of the tonsil and associated lymph nodes of gnotobiotic piglets t o the presence of bacterial antigen in the oral cavity. J. Anat. 117: 191-198 (1974). 4 BOURNE,F. J.: IgA immunoglobulin from porcine serum. Biochem. biophys. Res. Commun. 36: 138-145 (1969).
5 BOURNE,F. J. and CURTIS, J.: The transfer of immunoglobulins, IgG, IgA and IgM, from serum t o colostrum and milk in the sow. Immunology 24: 157-162 (1973). 6 BOURNE, F. J.; NEWBY,T. J., and CHIDLOW, J.W.: The influence of route of vaccination on the systemic and local immune response in the pig. Res. vet. Sci. 18: 244-248 (1975). 7 BRADLEY, P.A.; BOURNE, 1. J., and BROWN,P. J.: The respiratory tract immune system in the pig. 1. Distribution of immunoglobulintontainingcells in the respiratory tract mucosa. Vet. Pathol. 13: 81-89 (1976). 8 BURTIN,P. and BUFFE,D.: Synthesis of human immunoglobulins in germinal centres; in Germinal centres in immune responses, pp. 120-125 (Springer, Berlin 1967). 9 EDDIE,D. S.; SCHULKIND, M. L., and ROBBINS, J.B. : The isolation and biologic activities of purified secretory IgA and IgG anti Salmonella fyphimrrrium‘0’ antibodies from rabbit intestinal fluid and colostrum. J. Immun. 106: 181-190 (1971).
Respiratory Lymphoid Tissues
97
10 ISHIKAWA, T.; WICHER,K., and ARBESMAN, C.E.: Distribution of immunoglobulins in palatine and pharyngeal tonsils. Int. Archs Allergy appl. Immun. 43: 801-812 (1972). 1 1 KOBURG, E.: in Germinal centres in immune responses, p. 170 (Springer, Berlin 1967). 12 MELLORS, R.C. and KORNGOLD, L.: The cellular origin of human immunoglobulins. J. exp. Med. 118: 387-396 (1963). 13 NOSSAL, G .J.V.; ADA,G.L., and AUSTIN,C.M.: Behaviour of active bacterial antigens during the induction of the immune response. 11. Cellular distribution of flagellar antigen labelled with iodine-131. Nature, Lond. 199: 1259-1262 (1963). 14 ORTEGA, L.G. and MELLORS, R.C.: Cellular sites of formation of gamma globulins. J. exp. Med. 106: 627-640 (1957). 15 PERNIS, B. : in Germinal centres in immune responses, p. I 12 (Springer, Berlin 1967). 16 SAINTE-MARIE, G.:A paraffin embedding technique for studies employing immunofluorescence. J. Histochem. Cytochem. 10: 250-256 (1962). 17 WALDMAN, R. H.; JURCENSEN, P. F.; OLSEN,G. N.; GANGULY, R., and JOHNSON, J.E. : Immune response of the human respiratory tract. 1. Immunoglobulin levels and influenza virus vaccine antibody response. J. Immun. I l l : 38-41 (1973). D.M. and ROWLAND, A.C.: The palatine tonsils of the pig - an afferent 18 WILLIAMS, route to the lymphoid tissue. J. Anat. 113: 131-137 (1972).
Dr. F. J. BOURNE,Department of Animal Husbandry, University of Bristol, Bristol, BS18 7 D U (England)
The Respiratory Tract Immune System in the Pig 11. Associated Lymphoid Tissues
PATRICIA A. BRADLEY,F. J. BOURNEand P. J. BROWN Department of Animal Husbandry, University of Bristol, Bristol
Abstract. The distribution of immunoglobulins A, G and M were assessed in the bronchial lymph node and palatine tonsil of pigs from birth to maturity, using the immunoperoxidase technique. Immunoglobulin-containing cells appeared earlier in the tonsil than in the lymph node; this probably reflects the exposure of the tonsil to environmental antigens. IgG-containing cells predominated after the first 3 weeks of life, but the 1gA:lgG ratio was much higher in the tonsil than in the lymph node. Extracellular, reticular staining was visible in the central region of some follicles in addition to a few immunoglobulincontaining cells that resembled large lymphocytes or blast cells.
Studies in local immunity have led to increasing interest in the interrelationships of local and systemic immune systems. It has been shown that local administration of antigen may result in systemic as well as local antibody production. This may be caused by the dissemination either of antigen or of immunocompetent cells from the site of immunisation or lymphoid tissues in that area to other sites [6,9, 171. We decided to complement our study of respiratory tract immunity in the pig [7] with an examination of respiratorytract-associated lymphoid tissues. Two lymphoid tissues were examined : the bronchial lymph node, which is one of the major draining nodes of the bronchial tree, and the palatine tonsil, which is situated in a strategic position in the oropharynx at the entrance to the two most extensive mucous surfaces of the body - the intestinal and respiratory tracts. Materials and Methods Tissue specinlens. Thirty-four Large White pigs were used from a closed herd free of enzootic pneumonia. Twenty were killed at intervals from birth to 10 days and thereafter
240 -
200 160
-
120
-
IgG IgA IgM
91
BRADLEY/BOURNE/BROWN
-
Bronchial lymph node
...........A
0------a
ul
0 n
80-
Tonsil
120
-
80 -
1
3
4
5
6
10
12
15 16
20
Age, weeks
Fig. 1. Pig respiratory-tract-associated lymphoid tissues; mean of cell counts for IgA, IgG and IgM.
weekly in groups of two or three until 6 weeks. At 10 weeks two pigs were killed, two at 16, two at 20, one at 12, and one at 15 weeks. Two sows also were included in the study. Tissues were taken promptly after death from the palatine tonsil and the bronchial lymph node, fixed in 95% methanol, and processed according to the method Of SAINTE-MARIE [I 61. Preparation ofantisera. Antisera to porcine immunoglobulins A, G and M were prepared as described [4, 51. Immunoperoxidase technique. Preparation and use of peroxidase-antiserum conjugates and the quantitation of immunoglobulin-containing cells were carried out as described [7].
Results Bronchial Lymph Node Immunoglobulin-containing cells were first seen in 6-day-old pigs; these were initially IgA- and IgM-containing cells with a lag of 1 o r 2 days before
sow
92
BRADLEY/BOURNE/BROWN
Fig. 2. Subtrabecular sinus in lymph node of 10-week-old pig, incubated with horseradish peroxidase-anti IgG and stained for peroxidase activity. There are large numbers of immunoglobulin-containing cells resembling large lymphocytes. A number of macrophages also show specific staining and can be detected by their irregular shape, with cytoplasmic processes spanning the sinus (m). Fig. 3. Lymph node cortex of 10-week-old pig, incubated with horseradish peroxidaseanti IgM and stained for peroxidase activity. Specific staining is associated with the cytoplasm of a few cells resembling lymphoblasts, and is also extracellular in the central region of the follicle, where it forms a continuous, reticular pattern.
Respiratory Lymphoid Tissues
93
IgG cells appeared. Numbers of these cells were low until the pigs were 3 weeks old when a marked increase, mainly in the IgG class, occurred (fig. 1). Thereafter, IgG cells predominated, while IgA and IgM cell numbers remained low. IgA counts were consistently less than half the IgM values. Relatively few immunoglobulin-containing cells were identified morphologically as mature plasma cells, and those that were found were situated around the periphery of follicles and in the medulla. A large number of immunoglobulin-containing cells were found in or bordering subtrabecular sinuses. Sometimes it was difficult to distinguish cell types because they were closely packed. Apart from macrophages, which stained for immunoglobulin and were distinguished by their irregular shape, the predominant type of cell resembled the large lymphocyte, having a large nucleus and scant cytoplasm (fig. 2). Lymphoid follicles, appearing as circular to oval, densely packed accumulations of cells, were visible in a 10-day old pig; these increased in size as the pig aged, and a continuous reticular pattern of staining that appeared to be extracellular was a feature of many (fig. 3). Few immunoglobulin-containing cells were associated with the follicles, and these were mainly cells with a large, slightly eccentric nucleus and a narrow rim of cytoplasm (fig. 3). Follicles containing such cellular activity and possessing an outer collar of pale stained, non-immunoglobulin-containingsmall lymphocytes were considered to be secondary follicles with germinal centres. Staining of the same follicle for more than one immunoglobulin class revealed a number of follicles that showed multiple affinity, although others stained for one class only.
Tonsil Immunoglobulin-containing cells appeared when the pig was relatively young - 3 days old. As in the lymph node, IgG-containing cells appeared a day or two later than IgA and IgM. Until the pig was 3 weeks old, numbers of IgA- and IgM-containing cells were higher than IgG, but after this time IgG cells predominated (fig. 1). The cells varied in type, most resembling
Fig. 4. Tonsil of 8-week-old pig, incubated with horseradish peroxidase-anti IgM and stained for peroxidase activity. Several immunoglobulincontaining cells lie in the crypt epithelium between the epithelial cells. Fig. 5. Tonsillar crypt epithelium of 4-week-old pig, incubated with horseradish peroxidase-anti IgA. The crypt epithelium (e) is narrow and pseudostratified columnar. Several cells are specifically stained for IgA. A few immunoglobulincontaining cells are in the tonsillar parenchyma (p), but there is no invasion of this type of epithelium.
94
BRADLEY/BOURNE/BROWN
Table I . Pig respiratory-tract-associated lymphoid tissues: mean of ratios between immunoglobulin classes for all age groups
Mean of ratios for all age groups
Tissue
Bronchial lymph node Tonsil
IgA: IgG
IgA: IgM
0.1 0.6
0.6 I .4
plasma cells or large lymphocytes; initially they were scattered throughout the lymphoid tissue, but by the time the pig was a week or so old IgA cells were for the most part immediately beneath or within the crypt epithelium, whereas IgG and IgM cells were more scattered and more frequently found in association with germinal centres. Germinal centres were numerous and showed staining patterns similar to those in the lymph node. The crypts were usually lined with a stratified, squamous epithelium extensively invaded by all classes of small lymphocytes and immunoglobulincontaining cells that appeared to lie between the epithelial cells (fig. 4). This infiltration was especially marked in areas adjacent to a germinal centre, and in such cases it was often difficult to determine the limits of the epithelium. In addition, a second type of crypt epithelium was seen, generally near the neck of the crypt; this was a fairly shallow, pseudostratified columnar epithelium that showed very little cellular infiltration. Many of the epithelial cells showed distinct, finely granular staining for IgA and IgM, but not for IgG (fig. 5). Comparing the proportions of the different classes in bronchial lymph node and tonsil (table I ) it can be seen that although IgG predominates in both sites, the IgA: IgG ratio is much higher in the tonsil than in the lymph node. Discussion
Most of the immunoglobulin-containing cells in the bronchial lymph node resembled large lymphocytes and were located for the most part in the subtrabecular sinuses. These sinuses are the terminations of the afferent lymphatics [ I ] and as such drain sites of antigen stimulation [2, 31. It is possible that the large lymphocytes in the subtrabecular sinuses of the bronchial lymph node were derived from mucosal lymphocytes after antigenic stimulation, especially as the rise in cell numbers in the lymph node corresponded
Respiratory Lymphoid Tissues
95
to the rise in plasma cell numbers in the respiratory tract tissues [7]. If this were the case, the high proportion of IgG-containing cells in the lymph node may represent an important connection between local antigenic stimulation and systemic immunisation. However, the relatively low numbers of mature plasma cells in the bronchial lymph node suggest that the site of serum antibody formation is elsewhere, and recent evidence of ours indicates that the intestinal mucosa and, to a lesser extent, the respiratory tract mucosa [7] may contribute significant amounts of all three immunoglobulins. The distribution of extracellular reticular staining in the central region of follicles in both lymph node and tonsil resembled that described for antigen [ 131. Reticular localization of immunoglobulin has also been described in follicles [12, 14, 151, and this similar distribution of antigen and immunoglobulin suggests that antibody in close association with the processes of dendritic cells may be important as a mechanism for trapping and processing antigen. Although our findings showing that follicles stained for more than one immunoglobulin class differed from the findings in one study [8], others have reported double staining of follicles in human spleen and lymph node [12,15]. Such observations would be consistent with the theory that germinal centres represent an aggregate rather than a clone of cells. If exogenous immunoglobulins are localized in a follicle by virtue of their antibody properties, they would be expected to reflect the class heterogeneity of serum antibody. Immunoglobulin-containing cells appeared relatively early in the tonsil, when the pig was 3 days old, compared to 6 days old for bronchial lymph node and 6-7 days for respiratory tract mucosa [7]. This precocious development might be expected considering that the position of the tonsil in the oropharynx and its reticulated, permeable epithelium ensure that it undergoes constant antigenic exposure [ll]. The tonsil has no afferent lymphatics, but antigen has been shown to pass from the oral cavity via the crypt epithelium to the subepithelial lymphoid tissue where it localizes in follicles and in the cytoplasm of macrophages [18]. The tonsil of gnotobiotic piglets responded to oral administration of bacterial antigen by germinal centre formation but without proliferation of cells of the plasma cell series [3]. In the present study, probably as a result of continuous antigenic stimulus [l I], germinal centres and immunoglobulin-containing cells were conspicuous in the tonsil. Most of the IgA-containing cells were closely associated with the crypt epithelium, whereas IgM and IgG cells showed a more general distribution between the follicles, and were more frequently associated with germinal centres. This distribution of the classes resembled that reported for the human tonsil [lo].
96
BRADLEY/BOURNE/BROWN
An interesting finding was the association of IgA and IgM staining with crypt epithelial cells. This type of epithelium differed from the normal, heavily cell-infiltrated, stratified squamous type of epithelium in that it appeared to be pseudostratified and columnar in form. It may represent a route for secretion of these two immunoglobulin classes into the crypt lumen to protect the tonsil against pathogens, many of which may initially lodge in the tonsillar crypts. It is clear that the immunological role of the tonsil has yet to be fully elucidated, but it is likely to be complex, involving both local and systemic immunity. Acknowledgements The authors thank Prof. T. K. EWER,in whose department the work was carried out, and the staff of the Photographic Unit for their help. Supported by grants from the Wellcome Trust and the Agricultural Research Council. This work was part of the swine fever control program financed by the European Economic Community.
References 1 ANDERSON, J.C.: The micro-anatomy of the lymph node of germ-free piglets. Br. J. exp. Path. 53: 37-39 (1972). 2 ANDERSON, J.C.: The use of germ-free piglets in the study of lymphoid tissue and germinal centre formation. Adv. exp. med. Biol. 29: 643-649 (1973). 3 ANDERSON, J. C. : The response of the tonsil and associated lymph nodes of gnotobiotic piglets t o the presence of bacterial antigen in the oral cavity. J. Anat. 117: 191-198 (1974). 4 BOURNE,F. J.: IgA immunoglobulin from porcine serum. Biochem. biophys. Res. Commun. 36: 138-145 (1969).
5 BOURNE,F. J. and CURTIS, J.: The transfer of immunoglobulins, IgG, IgA and IgM, from serum t o colostrum and milk in the sow. Immunology 24: 157-162 (1973). 6 BOURNE, F. J.; NEWBY,T. J., and CHIDLOW, J.W.: The influence of route of vaccination on the systemic and local immune response in the pig. Res. vet. Sci. 18: 244-248 (1975). 7 BRADLEY, P.A.; BOURNE, 1. J., and BROWN,P. J.: The respiratory tract immune system in the pig. 1. Distribution of immunoglobulintontainingcells in the respiratory tract mucosa. Vet. Pathol. 13: 81-89 (1976). 8 BURTIN,P. and BUFFE,D.: Synthesis of human immunoglobulins in germinal centres; in Germinal centres in immune responses, pp. 120-125 (Springer, Berlin 1967). 9 EDDIE,D. S.; SCHULKIND, M. L., and ROBBINS, J.B. : The isolation and biologic activities of purified secretory IgA and IgG anti Salmonella fyphimrrrium‘0’ antibodies from rabbit intestinal fluid and colostrum. J. Immun. 106: 181-190 (1971).
Respiratory Lymphoid Tissues
97
10 ISHIKAWA, T.; WICHER,K., and ARBESMAN, C.E.: Distribution of immunoglobulins in palatine and pharyngeal tonsils. Int. Archs Allergy appl. Immun. 43: 801-812 (1972). 1 1 KOBURG, E.: in Germinal centres in immune responses, p. 170 (Springer, Berlin 1967). 12 MELLORS, R.C. and KORNGOLD, L.: The cellular origin of human immunoglobulins. J. exp. Med. 118: 387-396 (1963). 13 NOSSAL, G .J.V.; ADA,G.L., and AUSTIN,C.M.: Behaviour of active bacterial antigens during the induction of the immune response. 11. Cellular distribution of flagellar antigen labelled with iodine-131. Nature, Lond. 199: 1259-1262 (1963). 14 ORTEGA, L.G. and MELLORS, R.C.: Cellular sites of formation of gamma globulins. J. exp. Med. 106: 627-640 (1957). 15 PERNIS, B. : in Germinal centres in immune responses, p. I 12 (Springer, Berlin 1967). 16 SAINTE-MARIE, G.:A paraffin embedding technique for studies employing immunofluorescence. J. Histochem. Cytochem. 10: 250-256 (1962). 17 WALDMAN, R. H.; JURCENSEN, P. F.; OLSEN,G. N.; GANGULY, R., and JOHNSON, J.E. : Immune response of the human respiratory tract. 1. Immunoglobulin levels and influenza virus vaccine antibody response. J. Immun. I l l : 38-41 (1973). D.M. and ROWLAND, A.C.: The palatine tonsils of the pig - an afferent 18 WILLIAMS, route to the lymphoid tissue. J. Anat. 113: 131-137 (1972).
Dr. F. J. BOURNE,Department of Animal Husbandry, University of Bristol, Bristol, BS18 7 D U (England)
