Veterinary Immunology and Immunopathology, 32 ( 1992 ) 1-11
1
Elsevier Science Publishers B.V., Amsterdam
Local immunity in the mammary gland C.S. Lee, E. Meeusen and M.R. Brandon Centrefor Animal Biotechnology, School of Veterinary Science, The University of Melbourne, Parkville 3052, Vic., Australia (Accepted 2 May 1991 )
ABSTRACT Lee, C.S., Meeusen, E. and Brandon, M.R., 1992. Local immunity in the mammary gland. Vet. lmmunol. Immunopathol., 32:1-11. The mammary glands of pregnant and non-pregnant sheep were stimulated by infusion of killed
Staphylococcus aureus, and the lymphoid cell response delineated with a panel of monoclonal antibodies. Seven days after antigen infusion, the mammary glands of both pregnant and non-pregnant sheep displayed a striking feature, characterised by the presence of numerous CD45R + MHC class II + B cells in the periductal connective tissues. These cells were seen to be clustering around blood capillaries with very prominent endothelial cell lining. Some CD5 +CD4 + lymphocytes were scattered among the B-cell clusters, whereas a few CD8 + lymphocytes were seen mainly at the periphery of the B-cell clusters. Fourteen days after antigen infusion, numerous plasma cells were observed, most of them being of the IgA isotype. Seven days after parturition (approximately 40 days after antigen infusion) the number of lymphocytes and plasma cells in the infused glands had declined dramatically. These data indicate that B cell and helper T-ceU interaction can take place at the local sites of antigen stimulation in the mammary gland. ABBREVIATIONS MAb, monoclonal antibody; PBS, phosphate-buffered saline.
INTRODUCTION
Earlier work by Lascelles and his colleagues convincingly demonstrated that infusion of killed bacteria or flagellar antigens into the mammary glands of sheep around 3-4 weeks before parturition gave rise to local production of antibody which persisted into the following lactation (Lascelles, 1963; Outteridge et al., 1965; Lascelles et al., 1966; McDowell and Lascelles, 1971 ). Immunohistological studies of mammary glands of sheep revealed that the infused glands had significantly more lymphocytes and plasma cells than the non-infused glands (Lee and Lascelles, 1970; Watson and Lee, 1978). In a subsequent study (Scheldrake et al., 1985 ), specific antibody-containing cells were enumerated in mammary tissue and lymph nodes after intramammary © 1992 Elsevier Science Publishers B.V. All fights reserved 0165-2427/92/$05.00
2
C.S. LEE ET A L
infusion of antigen. The predominant immunoglobulin isotype of the antibody-containing cells was IgGl followed by IgG2. Attempts to increase the number of IgA-containing cells by previous intraperitoneal immunisation in Freund's complete adjuvant were unsuccessful. No studies have been carried out in sheep to define the distribution of Tlymphocyte subpopulations in the mammary gland following infusion of antigen despite the role T lymphocytes play in B-cell activation and isotype switching (Mosmann and Coffman, 1989 ). We have recently studied the cellular basis of local immunity in the unstimulated mammary gland of nonpregnant and pregnant sheep (Lee et al., 1989). Immunocytochemical studies of frozen sections revealed that in glands from pregnant and non-pregnant sheep the most predominant lymphocyte subpopulation bears the CD8 ÷ phenotype followed by CD4 ÷ and CD45R ÷ MHC Class II- cells. As a sequel of the previous study of unstimulated mammary gland (Lee et al., 1989 ), this paper reports on the dramatic cellular changes observed in the sheep mammary gland following local challenge with killed Staphylococcus aureus. MATERIALS AND METHODS
Animals
Fifteen primiparous sheep, of which 9 were about 120 days pregnant and the remaining 6 were non-pregnant, were used. The mammary glands of all the sheep were free from abnormality when examined prior to the commencement of the experiment. A n tigen
A formalin-killed, low virulence beta-hemolytic S. aureus suspension (from Dr. D.L. Watson, CSIRO, Armidale, Australia) was used for intramammary infusion. Each 2 ml of sterile saline contained 5 X 109 killed bacteria. Experimental procedure
Pregnant sheep were randomly allocated to three groups of three and the non-pregnant sheep were randomly allocated to two groups of three. The left gland (infused gland) of all the sheep was infused with 2 ml of the S. aureus suspension, whereas the right gland (control gland) of all the sheep was infused with 2 ml of sterile saline. The first and second groups of pregnant sheep were killed by an overdose of pentobarbitone sodium (Nembutal, Ceva Chemicals, Vic., Australia) at 7 and 14 days respectively after infusion, and the third group at 7 days after parturition (approximately 40 days after infusion ). The last two groups of non-pregnant sheep were killed at 7 and 14 days
MAMMARYGLANDIMMUNITY
3
respectively after infusion. Immediately after death, tissues were collected from the teat, the ventral, middle and dorsal regions of each gland. For both frozen and conventional paraffin sections, at least three tissue blocks, each measuring approximately 1.5 × 2.5 cm and 0.4 cm in thickness were sampled from each region.
Tissues For immunohistochemical studies, mammary tissues were embedded in OTC compound (Miles Scientific, Naperville, IL), snap-frozen using liquid nitrogen and stored at - 7 0 ° C. Tissues were also fixed in Bouin's fluid and processed for conventional paraffin embedding and staining, and for indirect immunoperoxidase staining to reveal IgG~, IgG2, IgM and IgA producing cells.
Antibodies The monoclonal antibodies (MAb) used were similar to those in our previous study (Lee et al., 1989). They were produced in our laboratories, and their characteristics are detailed in Table 1. Monoclonal antibodies against sheep IgG~, IgG2, IgM and IgA were the kind gift of Dr. K. Beh (CSIRO, Division of Animal Health, Sydney).
Immunocytochemical staining For light microscopy studies, 6 gm frozen sections were cut, fixed for 10 rain in cold ethanol, air-dried and then stained, using the indirect immunoTABLE 1 Reactivity of monoclonal antibodies to sheep lymphocytes ~ Antigen
MAb clone number
Antigen distribution
CD45 (SBU-LCA) MHC class II
1-28-124 49-1
CD5 (SBU-T1)
25-91
CD4 (SBU-T4)
44-38 and 44-97
CD8 (SBU-T8)
38-65
SBU-T192
19-19
All leukocytes Present on B lymphocytes and activated T lymphocytes; 49-1 is monomorphic for all four subsets of sheep class II molecules Present on all T lymphocytes, absent from B lymphocytes Present on subset o f T lymphocytes that are CD8-, SBU-T 19-; absent from B-lymphocytes Present on subset ofT lymphocytes that are CD4-, SBU-T 19-; absent from B-lymphocytes Present on subset ofT lymphocytes that are CD4-, CD8-; absent from B-lymphocytes Present mainly on B-lymphocytes and small subset of T lymphocytes
CD45R (SBU-p220) 20-96
'References for MAb are found in a previous report (Lee et al., 1988 ). 2Mackay et al. (1986).
4
C.S. LEE ET AL.
peroxidase technique as previously described (Lee et al., 1985). Briefly, tissue sections were incubated with the first antibody for 45 min at 20 ° C. Slides were washed with phosphate-buffered saline (PBS) and then incubated with peroxidase-conjugated rabbit anti-mouse Ig (DAKO-immunoglobulins, Glostrup, Denmark) diluted 1:50 in PBS containing 5% normal sheep serum. After thorough washing in PBS, sections were incubated in 0.06% ( w / v ) diaminobenzidine tetrahydrochloride (Sigma Chemicals, St. Louis, MO) in PBS containing 0.05% ( v / v ) hydrogen peroxide for 10 min at room temperature. For controls either PBS or 1% normal sheep serum in PBS was substituted for the first antibody. For identification of antibody-containing cells, 6-/~m paraffin sections were stained by the indirect immunoperoxidase technique in the same manner as described above. RESULTS
Pregnant sheep Histological observations Seven days after S. aureus infusion, the most striking feature observed was an extensive infiltration of lymphocytes into the lamina propria of the teat cistern, the periductal connective tissues of the duct system, and the epithelial lining of the teat cistern and the duct system (Figs. 1, 2). The lymphocytic infiltration into the lamina propria and the periductal areas were patchy. Closer examination revealed that the patches of lymphocytes were usually
Fig. 1. Numerous lymphocytes have infiltrated into the epithelium (small arrows) of a large duct (D) and the periductal connective tissue. Numerous lymphocytes are also seen clustering around blood vessels (large arrows) in the periductal connective tissue. Seven days after infusion of killed S. aureus. Haematoxylin and eosin, × 140.
MAMMARY GLAND IMMUNITY
5
Fig. 2. Two blood capillaries with very prominent endothelial cells (arrows) surrounded by numerous lymphocytes. It should be noted that each capillary is seen with a lymphocyte in its lumen. D, lumen of a ductule. Seven days after infusion of killed S. aureus. Haematoxylin and eosin, X 400.
Fig. 3. Staining of frozen sections of mammary gland tissue from a pregnant sheep 7 days after infusion of killed S. aureus, by immunoperoxidase using MAb against CD45R sheep lymphocyte antigen. Numerous CD45R + cells are seen surrounding the blood capillaries (arrows) in the connective tissue around a small duct (D). Indirect immunoperoxidase method + haematoxylin, X 160.
6
C S LEE ET AL.
Fig. 4. Serial section of the area shown in Fig. 3 revealed that some of the cells within the cluster are CD4 +. D, duct. Indirect immunoperoxidase method using MAb against CD4 sheep antigen + haematoxylin, X 160.
clustered around blood vessels with a very prominent endothelial cell lining (Fig. 2 ), which was never observed in the control glands. Lymphocytes were frequently found in the lumen and the wall of these vessels (Fig. 2 ). Lymphocytes were also found in the alveolar epithelium. In alveoli located farther away from the teat, some neutrophils were seen in the lumen and a number of lymphocytes were observed in the interalveolar areas. By 14 days after infusion a reduction in the number of lymphocytes was observed. However, plasma cells were more frequently seen at this time in the periductal connective tissue, sometimes forming a cluster around blood vessels in the lamina propria of the teat and the vessels in the periductal connective tissues. Plasma cells were also found in interalveolar areas in close proximity to the alveolar epithelium. At 7 days after birth (approximately 40 days after infusion) active secretory activity was evident in both the infused and control glands as the lumen of ducts and alveoli were filled with milk. In the infused glands, lymphocytes were still found in the alveolar and ductal epithelium and the teat epithelium, though markedly reduced in numbers compared to glands 7 days after infusion. Clusters of 2-3 plasma cells were seen at random in the interalveolar areas and small clusters of lymphocytes were occasionally seen in the periductal connective tissue of large ducts.
MAMMARYGLANDIMMUNITY
7
Fig. 5. Serial section of the area shown in Fig. 3 revealed that very few cells are CD8 +. D, duct. Indirect immunoperoxidase method using MAb against CD8 sheep antigen+haematoxylin, × 160.
In the control glands of all sheep, lymphocytes were observed in the lamina propria of the teat, the periductal connective tissues and the epithelial lining of ducts and glands, but they were far fewer in number than the infused glands.
ImmunocytochemicaI staining of lymphocyte subpopulations The pattern of distribution of lymphocyte subsets in the mammary glandular tissue was similar at 7 and 14 days after infusion. The most prominent feature was that most of the lymphocyte clusters found in the lamina propria of the teat and the periductal connective tissues were CD45R +, MHC class II ÷ (Fig. 3). Some of the cells were CD5 ÷ and CD4 ÷ (Fig. 4) and few were CD8 ÷ (Fig. 5 ), the latter being mainly located at the periphery of what were obviously B-cell clusters. There were very few T19 ÷ cells. The endothelial cells of blood vessels surrounded by these patches of lymphocytes were very prominent and were not stained by any of the MAb used (Figs. 3-5 ). In both infused and control glands a large proportion of the lymphocytes in the ductal epithelium were CD8 + followed by some CD5 ÷ and CD4 ÷ lymphocytes and a few CD45R + lymphocytes.
Antibody-producing cells Bouin-fixed paraffin sections from infused and control glands were stained with the indirect immunoperoxidase technique to reveal IgG~, IgG2, IgM and
8
C.S. L E E E T AL.
.J =
I
a
~( ,.!1' 6 Fig. 6. Paraffin section of mammary gland tissue of pregnant sheep 14 days after infusion of killed S. aureus stained with the indirect immunoperoxide technique for plasma cells of the IgA isotype (arrows). Note the presence of numerous positive cells located in close proximity to the alveoli (A). X450.
;ii i~i¸
~
i~~
7 Fig. 7. A near-serial paraffin section of the area shown in Fig. 6 stained with the indirect immunoperoxidase technique for plasma cells of the IgGl isotype (arrows). Note that only a few positive cells are present. A, alveoli. X450.
IgA producing cells. Only the occasional positive cell was observed in the control glands and this was a consistent observation at all stages o f these studies. In contrast, large numbers o f Ig-positive cells were seen in the infused glands,
MAMMARY GLAND IMMUNITY
9
reaching maximum numbers at 14 days after infusion, but declining dramatically by 40 days after infusion. A large proportion of the plasma cells were leA positive (Fig. 6) with the next frequent plasma cells being IgG~ positive (Fig. 7 ). Much smaller numbers of IgM, and IgG2 cells were observed (not shown). Most of the antibody-producing cells were seen in interalveolar areas in close proximity to the alveolar epithelial cells. Some plasma cells were also found among lymphocytes around blood vessels in the periductal connective tissues.
Non-pregnant sheep The glands of non-pregnant sheep were poorly developed, being composed mainly of lobules rich in adipose tissue and with a rudimentary duct system. The lymphocyte infiltration patterns and the plasma cell response of infused glands of non-pregnant sheep were basically similar to that of the infused glands of pregnant sheep. DISCUSSION
Antigen infusion of the sheep mammary gland resulted in a marked increase in lymphocytes and plasma cells and the difference between infused and control glands was still evident during the first week of lactation. At 7 and 14 days after infusion the most dramatic response was along the duct system. This is to be expected as these are the sites which first come into direct contact with the infused antigens. Another feature observed in the stimulated gland was the appearance of blood capillaries with prominent endothelial cell lining. Surrounding these vessels were large collections of lymphocytes, most of which were CD45R + MHC Class II +, indicating that they were B cells. These B-cell clusters were not observed in our previous studies of mammary glands from non-pregnant and pregnant sheep (Lee et al., 1989), or in the contralateral control glands in the present study. It is therefore evident that the appearance of B-cell perivascular cuffings in the infused glands is in response to local antigenic stimulation. The frequent presence of lymphocytes within the lumen and the wall of these vessels suggests that these vessels serve as the sites for B-cell migration. The distribution of T lymphocytes in respect to these B-cell clusters is very similar to their distribution within the germinal centres of activated lymph nodes in which CD4 + T cells are present within the B-cell dusters, while CD8 + T cells are mainly restricted to the periphery. It is likely that these T cells play a similar regulatory role in local B-cell development and antibody production within the mammary tissue. The observation that more plasma cells were seen by 14 days after infusion suggests that most of the B cells, after having migrated from the blood vessels,
10
C.S. LEE ET AL.
had transformed into plasma cells by this time. By using monoclonal antibodies against the various classes of immunoglobulin it was found that the IgA plasma cells were the predominant cell type, confirming earlier studies (Lee and Lascelles, 1970; Watson and Lee, 1978). In contrast, Sheldrake et al. ( 1985 ) found predominantly IgG1 and few IgA plasma cells after infusion of Brucella abortus or ovalbumin into the m a m m a r y gland. It was claimed that their failure to enhance the numbers of IgA cells by previous intraperitoneal priming could be attributed to the adverse hormonal environment in the nonlactating gland (Weisz-Carrington et al., 1977). However, we found no difference in the immunological response between glands from pregnant and nonpregnant sheep. One possible explanation for the low numbers of IgA plasma cells found in their studies compared to our present and previous (Lee and Lascelles, 1970) studies could be due to the specificity of the polyclonal antiIgA reagents or the difference in antigens used in their studies. The importance of T-cell-derived cytokines in B-cell activation and isotype-switching has now been well established (Mosmann and Coffman, 1989 ). The selective induction of antibody isotypes could be of particular importance in protection of the m a m m a r y gland against mastitis as the bacteriocidal activity of neutrophils has been shown to be positively correlated only with titres of IgG2 but not with IgGl or IgA antibodies (Watson, 1989). The control of isotype-switching will most likely reside at the level of the T cell rather than the B cell and the study of this lymphocyte population will therefore be of crucial importance in our understanding of antibody regulation in the m a m m a r y glandular tissue and secretions. ACKNOWLEDGEMENTS The authors wish to thank S. Cuell, L. Tatarczuch, H. Jackson and G. Smith for technical assistance. This work was supported by the Australian Research Council.
REFERENCES Lascelles, A.K., 1963. A review of the literature on some aspects of immune milk. Dairy Sci. Abstr., 25: 359-364. Lascelles, A.K., Outteridge, P.M. and McKenzie, D.D.S., 1966. Local production of antibody by the lactating mammary gland followingantigenic stimulation. Aust. J. Exp. Biol. Med. Sci., 44: 169-180. Lee, C.S., Gogolin-Ewens,K. and Brandon, M.R., 1988. Identification of a unique lymphocyte subpopulation in the sheep uterus. Immunology,63: 157-164. Lee, C.S., Gogolin-Ewens,K., White, T.R. and Brandon, M.R., 1985. Studies on the distribution of binucleate cells in the placenta of the sheep with a monoclonal antibody SBU-3. J. Anat., 140: 565-576. Lee, C.S. and Lascelles,A.K., 1970. Antibody-producingcells in antigenicallystimulated mam-
MAMMARY GLAND IMMUNITY
11
mary glands and in the gastro-intestinal tract of sheep. Aust. J. Exp. Biol. Med. Sci., 48: 525535. Lee, C.S., Meeusen, E. and Brandon, M.R., 1989. Subpopulations of lymphocytes in the mammary gland of sheep. Immunology, 60: 388-393. McDoweU, G.H. and Lascelles, A.K., 1971. Local immunization of ewes with staphylococcal cell and cell-toxoid vaccines. Res. Vet. Sci., 12: 258-264. Mackay, C.R., Maddox, J.F. and Brandon, M.R., 1986. Three distinct subpopulations of sheep T lymphocytes. Eur. J. Immunol., 16: 19-25. Mosmann, T.R. and Coffman, R.L., 1989. Heterogeneity of cytokine secretion patterns and functions of helper T cells. In: F.J. Dixon (Editor), Advances in Immunology. Academic Press, Harcourt Brace Jovanovich, London, pp. 111-147. Outteridge, P.M., Rock, J.D. and Lascelles, A.K., 1965. The immune response of the mammary gland and regional lymph node following antigenic stimulation. Aust. J. Exp. Biol. Med. Sci., 43: 265-274. Sheldrake, R.F., Husband, A.J. and Watson, D.L., 1985. Specific antibody-containing cells in the mammary gland of non-lactating sheep after intraperitoneal and intramammary immunisation. Res. Vet. Sci., 38: 312-316. Watson, D.L. and Lee, C.G., 1978. Immunity to experimental staphylococcal mastitis - comparison of live and killed vaccines. Aust. Vet. J., 54: 374-378. Watson, E.D., 1989. Specific antibody in milk whey and phagocytosis of Actinomyces pyogenes by neutrophils in vivo. Res. Vet. Sci., 47: 253-256. Weisz-Carrington, P., Roux, M.E. and Lamm, M.E., 1977. Plasma cells and epithelial immunoglobulins in the mouse mammary gland during pregnancy and lactation. J. Immunol., 119: 1306-1309.
1
Elsevier Science Publishers B.V., Amsterdam
Local immunity in the mammary gland C.S. Lee, E. Meeusen and M.R. Brandon Centrefor Animal Biotechnology, School of Veterinary Science, The University of Melbourne, Parkville 3052, Vic., Australia (Accepted 2 May 1991 )
ABSTRACT Lee, C.S., Meeusen, E. and Brandon, M.R., 1992. Local immunity in the mammary gland. Vet. lmmunol. Immunopathol., 32:1-11. The mammary glands of pregnant and non-pregnant sheep were stimulated by infusion of killed
Staphylococcus aureus, and the lymphoid cell response delineated with a panel of monoclonal antibodies. Seven days after antigen infusion, the mammary glands of both pregnant and non-pregnant sheep displayed a striking feature, characterised by the presence of numerous CD45R + MHC class II + B cells in the periductal connective tissues. These cells were seen to be clustering around blood capillaries with very prominent endothelial cell lining. Some CD5 +CD4 + lymphocytes were scattered among the B-cell clusters, whereas a few CD8 + lymphocytes were seen mainly at the periphery of the B-cell clusters. Fourteen days after antigen infusion, numerous plasma cells were observed, most of them being of the IgA isotype. Seven days after parturition (approximately 40 days after antigen infusion) the number of lymphocytes and plasma cells in the infused glands had declined dramatically. These data indicate that B cell and helper T-ceU interaction can take place at the local sites of antigen stimulation in the mammary gland. ABBREVIATIONS MAb, monoclonal antibody; PBS, phosphate-buffered saline.
INTRODUCTION
Earlier work by Lascelles and his colleagues convincingly demonstrated that infusion of killed bacteria or flagellar antigens into the mammary glands of sheep around 3-4 weeks before parturition gave rise to local production of antibody which persisted into the following lactation (Lascelles, 1963; Outteridge et al., 1965; Lascelles et al., 1966; McDowell and Lascelles, 1971 ). Immunohistological studies of mammary glands of sheep revealed that the infused glands had significantly more lymphocytes and plasma cells than the non-infused glands (Lee and Lascelles, 1970; Watson and Lee, 1978). In a subsequent study (Scheldrake et al., 1985 ), specific antibody-containing cells were enumerated in mammary tissue and lymph nodes after intramammary © 1992 Elsevier Science Publishers B.V. All fights reserved 0165-2427/92/$05.00
2
C.S. LEE ET A L
infusion of antigen. The predominant immunoglobulin isotype of the antibody-containing cells was IgGl followed by IgG2. Attempts to increase the number of IgA-containing cells by previous intraperitoneal immunisation in Freund's complete adjuvant were unsuccessful. No studies have been carried out in sheep to define the distribution of Tlymphocyte subpopulations in the mammary gland following infusion of antigen despite the role T lymphocytes play in B-cell activation and isotype switching (Mosmann and Coffman, 1989 ). We have recently studied the cellular basis of local immunity in the unstimulated mammary gland of nonpregnant and pregnant sheep (Lee et al., 1989). Immunocytochemical studies of frozen sections revealed that in glands from pregnant and non-pregnant sheep the most predominant lymphocyte subpopulation bears the CD8 ÷ phenotype followed by CD4 ÷ and CD45R ÷ MHC Class II- cells. As a sequel of the previous study of unstimulated mammary gland (Lee et al., 1989 ), this paper reports on the dramatic cellular changes observed in the sheep mammary gland following local challenge with killed Staphylococcus aureus. MATERIALS AND METHODS
Animals
Fifteen primiparous sheep, of which 9 were about 120 days pregnant and the remaining 6 were non-pregnant, were used. The mammary glands of all the sheep were free from abnormality when examined prior to the commencement of the experiment. A n tigen
A formalin-killed, low virulence beta-hemolytic S. aureus suspension (from Dr. D.L. Watson, CSIRO, Armidale, Australia) was used for intramammary infusion. Each 2 ml of sterile saline contained 5 X 109 killed bacteria. Experimental procedure
Pregnant sheep were randomly allocated to three groups of three and the non-pregnant sheep were randomly allocated to two groups of three. The left gland (infused gland) of all the sheep was infused with 2 ml of the S. aureus suspension, whereas the right gland (control gland) of all the sheep was infused with 2 ml of sterile saline. The first and second groups of pregnant sheep were killed by an overdose of pentobarbitone sodium (Nembutal, Ceva Chemicals, Vic., Australia) at 7 and 14 days respectively after infusion, and the third group at 7 days after parturition (approximately 40 days after infusion ). The last two groups of non-pregnant sheep were killed at 7 and 14 days
MAMMARYGLANDIMMUNITY
3
respectively after infusion. Immediately after death, tissues were collected from the teat, the ventral, middle and dorsal regions of each gland. For both frozen and conventional paraffin sections, at least three tissue blocks, each measuring approximately 1.5 × 2.5 cm and 0.4 cm in thickness were sampled from each region.
Tissues For immunohistochemical studies, mammary tissues were embedded in OTC compound (Miles Scientific, Naperville, IL), snap-frozen using liquid nitrogen and stored at - 7 0 ° C. Tissues were also fixed in Bouin's fluid and processed for conventional paraffin embedding and staining, and for indirect immunoperoxidase staining to reveal IgG~, IgG2, IgM and IgA producing cells.
Antibodies The monoclonal antibodies (MAb) used were similar to those in our previous study (Lee et al., 1989). They were produced in our laboratories, and their characteristics are detailed in Table 1. Monoclonal antibodies against sheep IgG~, IgG2, IgM and IgA were the kind gift of Dr. K. Beh (CSIRO, Division of Animal Health, Sydney).
Immunocytochemical staining For light microscopy studies, 6 gm frozen sections were cut, fixed for 10 rain in cold ethanol, air-dried and then stained, using the indirect immunoTABLE 1 Reactivity of monoclonal antibodies to sheep lymphocytes ~ Antigen
MAb clone number
Antigen distribution
CD45 (SBU-LCA) MHC class II
1-28-124 49-1
CD5 (SBU-T1)
25-91
CD4 (SBU-T4)
44-38 and 44-97
CD8 (SBU-T8)
38-65
SBU-T192
19-19
All leukocytes Present on B lymphocytes and activated T lymphocytes; 49-1 is monomorphic for all four subsets of sheep class II molecules Present on all T lymphocytes, absent from B lymphocytes Present on subset o f T lymphocytes that are CD8-, SBU-T 19-; absent from B-lymphocytes Present on subset ofT lymphocytes that are CD4-, SBU-T 19-; absent from B-lymphocytes Present on subset ofT lymphocytes that are CD4-, CD8-; absent from B-lymphocytes Present mainly on B-lymphocytes and small subset of T lymphocytes
CD45R (SBU-p220) 20-96
'References for MAb are found in a previous report (Lee et al., 1988 ). 2Mackay et al. (1986).
4
C.S. LEE ET AL.
peroxidase technique as previously described (Lee et al., 1985). Briefly, tissue sections were incubated with the first antibody for 45 min at 20 ° C. Slides were washed with phosphate-buffered saline (PBS) and then incubated with peroxidase-conjugated rabbit anti-mouse Ig (DAKO-immunoglobulins, Glostrup, Denmark) diluted 1:50 in PBS containing 5% normal sheep serum. After thorough washing in PBS, sections were incubated in 0.06% ( w / v ) diaminobenzidine tetrahydrochloride (Sigma Chemicals, St. Louis, MO) in PBS containing 0.05% ( v / v ) hydrogen peroxide for 10 min at room temperature. For controls either PBS or 1% normal sheep serum in PBS was substituted for the first antibody. For identification of antibody-containing cells, 6-/~m paraffin sections were stained by the indirect immunoperoxidase technique in the same manner as described above. RESULTS
Pregnant sheep Histological observations Seven days after S. aureus infusion, the most striking feature observed was an extensive infiltration of lymphocytes into the lamina propria of the teat cistern, the periductal connective tissues of the duct system, and the epithelial lining of the teat cistern and the duct system (Figs. 1, 2). The lymphocytic infiltration into the lamina propria and the periductal areas were patchy. Closer examination revealed that the patches of lymphocytes were usually
Fig. 1. Numerous lymphocytes have infiltrated into the epithelium (small arrows) of a large duct (D) and the periductal connective tissue. Numerous lymphocytes are also seen clustering around blood vessels (large arrows) in the periductal connective tissue. Seven days after infusion of killed S. aureus. Haematoxylin and eosin, × 140.
MAMMARY GLAND IMMUNITY
5
Fig. 2. Two blood capillaries with very prominent endothelial cells (arrows) surrounded by numerous lymphocytes. It should be noted that each capillary is seen with a lymphocyte in its lumen. D, lumen of a ductule. Seven days after infusion of killed S. aureus. Haematoxylin and eosin, X 400.
Fig. 3. Staining of frozen sections of mammary gland tissue from a pregnant sheep 7 days after infusion of killed S. aureus, by immunoperoxidase using MAb against CD45R sheep lymphocyte antigen. Numerous CD45R + cells are seen surrounding the blood capillaries (arrows) in the connective tissue around a small duct (D). Indirect immunoperoxidase method + haematoxylin, X 160.
6
C S LEE ET AL.
Fig. 4. Serial section of the area shown in Fig. 3 revealed that some of the cells within the cluster are CD4 +. D, duct. Indirect immunoperoxidase method using MAb against CD4 sheep antigen + haematoxylin, X 160.
clustered around blood vessels with a very prominent endothelial cell lining (Fig. 2 ), which was never observed in the control glands. Lymphocytes were frequently found in the lumen and the wall of these vessels (Fig. 2 ). Lymphocytes were also found in the alveolar epithelium. In alveoli located farther away from the teat, some neutrophils were seen in the lumen and a number of lymphocytes were observed in the interalveolar areas. By 14 days after infusion a reduction in the number of lymphocytes was observed. However, plasma cells were more frequently seen at this time in the periductal connective tissue, sometimes forming a cluster around blood vessels in the lamina propria of the teat and the vessels in the periductal connective tissues. Plasma cells were also found in interalveolar areas in close proximity to the alveolar epithelium. At 7 days after birth (approximately 40 days after infusion) active secretory activity was evident in both the infused and control glands as the lumen of ducts and alveoli were filled with milk. In the infused glands, lymphocytes were still found in the alveolar and ductal epithelium and the teat epithelium, though markedly reduced in numbers compared to glands 7 days after infusion. Clusters of 2-3 plasma cells were seen at random in the interalveolar areas and small clusters of lymphocytes were occasionally seen in the periductal connective tissue of large ducts.
MAMMARYGLANDIMMUNITY
7
Fig. 5. Serial section of the area shown in Fig. 3 revealed that very few cells are CD8 +. D, duct. Indirect immunoperoxidase method using MAb against CD8 sheep antigen+haematoxylin, × 160.
In the control glands of all sheep, lymphocytes were observed in the lamina propria of the teat, the periductal connective tissues and the epithelial lining of ducts and glands, but they were far fewer in number than the infused glands.
ImmunocytochemicaI staining of lymphocyte subpopulations The pattern of distribution of lymphocyte subsets in the mammary glandular tissue was similar at 7 and 14 days after infusion. The most prominent feature was that most of the lymphocyte clusters found in the lamina propria of the teat and the periductal connective tissues were CD45R +, MHC class II ÷ (Fig. 3). Some of the cells were CD5 ÷ and CD4 ÷ (Fig. 4) and few were CD8 ÷ (Fig. 5 ), the latter being mainly located at the periphery of what were obviously B-cell clusters. There were very few T19 ÷ cells. The endothelial cells of blood vessels surrounded by these patches of lymphocytes were very prominent and were not stained by any of the MAb used (Figs. 3-5 ). In both infused and control glands a large proportion of the lymphocytes in the ductal epithelium were CD8 + followed by some CD5 ÷ and CD4 ÷ lymphocytes and a few CD45R + lymphocytes.
Antibody-producing cells Bouin-fixed paraffin sections from infused and control glands were stained with the indirect immunoperoxidase technique to reveal IgG~, IgG2, IgM and
8
C.S. L E E E T AL.
.J =
I
a
~( ,.!1' 6 Fig. 6. Paraffin section of mammary gland tissue of pregnant sheep 14 days after infusion of killed S. aureus stained with the indirect immunoperoxide technique for plasma cells of the IgA isotype (arrows). Note the presence of numerous positive cells located in close proximity to the alveoli (A). X450.
;ii i~i¸
~
i~~
7 Fig. 7. A near-serial paraffin section of the area shown in Fig. 6 stained with the indirect immunoperoxidase technique for plasma cells of the IgGl isotype (arrows). Note that only a few positive cells are present. A, alveoli. X450.
IgA producing cells. Only the occasional positive cell was observed in the control glands and this was a consistent observation at all stages o f these studies. In contrast, large numbers o f Ig-positive cells were seen in the infused glands,
MAMMARY GLAND IMMUNITY
9
reaching maximum numbers at 14 days after infusion, but declining dramatically by 40 days after infusion. A large proportion of the plasma cells were leA positive (Fig. 6) with the next frequent plasma cells being IgG~ positive (Fig. 7 ). Much smaller numbers of IgM, and IgG2 cells were observed (not shown). Most of the antibody-producing cells were seen in interalveolar areas in close proximity to the alveolar epithelial cells. Some plasma cells were also found among lymphocytes around blood vessels in the periductal connective tissues.
Non-pregnant sheep The glands of non-pregnant sheep were poorly developed, being composed mainly of lobules rich in adipose tissue and with a rudimentary duct system. The lymphocyte infiltration patterns and the plasma cell response of infused glands of non-pregnant sheep were basically similar to that of the infused glands of pregnant sheep. DISCUSSION
Antigen infusion of the sheep mammary gland resulted in a marked increase in lymphocytes and plasma cells and the difference between infused and control glands was still evident during the first week of lactation. At 7 and 14 days after infusion the most dramatic response was along the duct system. This is to be expected as these are the sites which first come into direct contact with the infused antigens. Another feature observed in the stimulated gland was the appearance of blood capillaries with prominent endothelial cell lining. Surrounding these vessels were large collections of lymphocytes, most of which were CD45R + MHC Class II +, indicating that they were B cells. These B-cell clusters were not observed in our previous studies of mammary glands from non-pregnant and pregnant sheep (Lee et al., 1989), or in the contralateral control glands in the present study. It is therefore evident that the appearance of B-cell perivascular cuffings in the infused glands is in response to local antigenic stimulation. The frequent presence of lymphocytes within the lumen and the wall of these vessels suggests that these vessels serve as the sites for B-cell migration. The distribution of T lymphocytes in respect to these B-cell clusters is very similar to their distribution within the germinal centres of activated lymph nodes in which CD4 + T cells are present within the B-cell dusters, while CD8 + T cells are mainly restricted to the periphery. It is likely that these T cells play a similar regulatory role in local B-cell development and antibody production within the mammary tissue. The observation that more plasma cells were seen by 14 days after infusion suggests that most of the B cells, after having migrated from the blood vessels,
10
C.S. LEE ET AL.
had transformed into plasma cells by this time. By using monoclonal antibodies against the various classes of immunoglobulin it was found that the IgA plasma cells were the predominant cell type, confirming earlier studies (Lee and Lascelles, 1970; Watson and Lee, 1978). In contrast, Sheldrake et al. ( 1985 ) found predominantly IgG1 and few IgA plasma cells after infusion of Brucella abortus or ovalbumin into the m a m m a r y gland. It was claimed that their failure to enhance the numbers of IgA cells by previous intraperitoneal priming could be attributed to the adverse hormonal environment in the nonlactating gland (Weisz-Carrington et al., 1977). However, we found no difference in the immunological response between glands from pregnant and nonpregnant sheep. One possible explanation for the low numbers of IgA plasma cells found in their studies compared to our present and previous (Lee and Lascelles, 1970) studies could be due to the specificity of the polyclonal antiIgA reagents or the difference in antigens used in their studies. The importance of T-cell-derived cytokines in B-cell activation and isotype-switching has now been well established (Mosmann and Coffman, 1989 ). The selective induction of antibody isotypes could be of particular importance in protection of the m a m m a r y gland against mastitis as the bacteriocidal activity of neutrophils has been shown to be positively correlated only with titres of IgG2 but not with IgGl or IgA antibodies (Watson, 1989). The control of isotype-switching will most likely reside at the level of the T cell rather than the B cell and the study of this lymphocyte population will therefore be of crucial importance in our understanding of antibody regulation in the m a m m a r y glandular tissue and secretions. ACKNOWLEDGEMENTS The authors wish to thank S. Cuell, L. Tatarczuch, H. Jackson and G. Smith for technical assistance. This work was supported by the Australian Research Council.
REFERENCES Lascelles, A.K., 1963. A review of the literature on some aspects of immune milk. Dairy Sci. Abstr., 25: 359-364. Lascelles, A.K., Outteridge, P.M. and McKenzie, D.D.S., 1966. Local production of antibody by the lactating mammary gland followingantigenic stimulation. Aust. J. Exp. Biol. Med. Sci., 44: 169-180. Lee, C.S., Gogolin-Ewens,K. and Brandon, M.R., 1988. Identification of a unique lymphocyte subpopulation in the sheep uterus. Immunology,63: 157-164. Lee, C.S., Gogolin-Ewens,K., White, T.R. and Brandon, M.R., 1985. Studies on the distribution of binucleate cells in the placenta of the sheep with a monoclonal antibody SBU-3. J. Anat., 140: 565-576. Lee, C.S. and Lascelles,A.K., 1970. Antibody-producingcells in antigenicallystimulated mam-
MAMMARY GLAND IMMUNITY
11
mary glands and in the gastro-intestinal tract of sheep. Aust. J. Exp. Biol. Med. Sci., 48: 525535. Lee, C.S., Meeusen, E. and Brandon, M.R., 1989. Subpopulations of lymphocytes in the mammary gland of sheep. Immunology, 60: 388-393. McDoweU, G.H. and Lascelles, A.K., 1971. Local immunization of ewes with staphylococcal cell and cell-toxoid vaccines. Res. Vet. Sci., 12: 258-264. Mackay, C.R., Maddox, J.F. and Brandon, M.R., 1986. Three distinct subpopulations of sheep T lymphocytes. Eur. J. Immunol., 16: 19-25. Mosmann, T.R. and Coffman, R.L., 1989. Heterogeneity of cytokine secretion patterns and functions of helper T cells. In: F.J. Dixon (Editor), Advances in Immunology. Academic Press, Harcourt Brace Jovanovich, London, pp. 111-147. Outteridge, P.M., Rock, J.D. and Lascelles, A.K., 1965. The immune response of the mammary gland and regional lymph node following antigenic stimulation. Aust. J. Exp. Biol. Med. Sci., 43: 265-274. Sheldrake, R.F., Husband, A.J. and Watson, D.L., 1985. Specific antibody-containing cells in the mammary gland of non-lactating sheep after intraperitoneal and intramammary immunisation. Res. Vet. Sci., 38: 312-316. Watson, D.L. and Lee, C.G., 1978. Immunity to experimental staphylococcal mastitis - comparison of live and killed vaccines. Aust. Vet. J., 54: 374-378. Watson, E.D., 1989. Specific antibody in milk whey and phagocytosis of Actinomyces pyogenes by neutrophils in vivo. Res. Vet. Sci., 47: 253-256. Weisz-Carrington, P., Roux, M.E. and Lamm, M.E., 1977. Plasma cells and epithelial immunoglobulins in the mouse mammary gland during pregnancy and lactation. J. Immunol., 119: 1306-1309.
