- Email: [email protected]
Induction of Ia antigens on thyroid cultures by syngeneic T lymphocytes: A model of autoimmune disease
Immunology Letters,
9 (1985) 285 289
Elsevier lmlet 573
INDUCTION OF Ia ANTIGENS ON THYROID CULTURES BY SYNGENEIC T LYMPHOCYTES: A MODEL OF A U T O I M M U N E DISEASE Jean S A L A M E R O and Jeannine C H A R R E I R E I N S E R M U. 283, H6pital Cochin, 27, rue du Faubourg Saint-Jacques, 75674 Paris Cedex 14, France
(Received9 October 1984) (Modified version received20 December1984) (Accepted 21 December1984)
1. Summary We previously reported that the expression of Ia antigens on cultured monolayers of murine thyroid epithelial cells (TEC) occurred with a particular distribution exclusively on the basal part of the cultured thyroid cells, while class I antigens of the major histocompatibility complex (MHC) are only detected on the apical surface. It appears that deposition of syngeneic lymphocytes induces, 24 h later, Ia expression on the apical side of cultured TEC, the surface that is in direct contact with the responder lymphocytes during syngeneic sensitization of T lymphocytes. We hypothesized that this phenomenon could represent, in syngeneic situations, the restriction process in antigen recognition by T cells, as demonstrated by la restricted primary syngeneic sensitization (PSS) on murine TEC. 2. Introduction In the past few years we have developed an "in vitro" model of primary syngeneic sensitization (PSS) on monolayers of thyroid epithelial cells (TEC). We demonstrated that sensitization was directed against structures borne by syngeneic TEC only [1], that specifically sensitized T lymphocytes [2, 3] could induce thyroid disorders when injected into intact syngeneic recipients [4], and that this proliferation was Key words:
thyroid - autoimmunity la antigens - t lym-
phocytes J. Charreire, INSERM U. 283, HSpital Cochin, Pavilion Hardy A, 27, rue du Faubourg SaintJacques, 75674 Paris Cedex 14, France Address correspondence to:
under the same genetic control [5] as experimental autoimmune thyroiditis (EAT) [6]. Using congenic strains of mice, we found that optimal PSS was obtained when compatibility between stimulator TEC and responder lymphocytes occurs at the I-A subregion of the major histocompatibility complex (MHC) [7]. Simultaneously, we investigated the expression of the M H C antigens (K end, D end and Ia) and thyroglobulin (Tg) on these in vitro cultured TEC. Results obtained by three different direct and indirect methods [8, 9] showed that class I MHC antigens (K and D ends) and Tg are expressed exclusively on the apical cell surface of TEC while class II M H C antigens (Ia) are detected on the opposing basal face of the monolayer. The dichotomy of distribution of MHC antigens on cultured TEC, and especially the existence of la antigens on the basal part of TEC which is not accessible to responder lymphocytes was in total opposition to the strict compatibility at the I-A subregion required to obtain optimal PSS of syngeneic lymphocytes on cultured TEC. Since the role of macrophages in PSS was fully ruled out [10], we envisaged two hypotheses to explain these mutually exclusive data. Either lymphocytes after deposition cross the TEC monolayer to reach the basal side of the thyroid cells which bears the Ia antigens, or lymphocyte deposition is followed by an apparition of Ia antigens on the apical side of the thyroid monolayer which is in contact with the responder syngeneic lymphocytes. We investigated the second hypothesis by comparing Ia antigen expression on monolayers of TEC from CBA (H-2 k) mice before and at different times after the deposition of syngeneic lymphocytes using a different method.
0165 2478 / 85 / $ 3.30 © 1985 ElsevierScience Publishers B.V.(BiomedicalDivision)
285
3. Materials and Methods
3.1. Animals CBA (H-2 k) mice were provided by the C.S.E.A.L. Orl6ans-la-Source (France). 3.2. Lymphoid cell suspension Spleen cells were aseptically removed and homogenized in a glass tube with a loose fitting Teflon pestle in Hanks' balanced salt solution (HBSS). The cell suspension was then washed twice in HBSS supplemented with penicillin (100 U/ml, streptomycin (100 #g/ml) and fungizone (2.5 #g/ml). Cells were adjusted to a final concentration of 2.5× 106/ml in R P M I 1640 (Gibco, Paisley, Scotland) supplemented with penicillin (100 U/ml), streptomycin (I00 #g/ml), fungizone (2.5/zg/ml) and 1% L-glutamine (complete medium). 3.3. Mouse thyroid cell cultures These were performed as previously reported [1-5]. Briefly, thyroids were carefully dissected and minced into small pieces at 4°C in complete medium. The suspension was incubated with 1.5 mg/ml of collagenase (Boehringer, Mannheim, F.R.G.) under shaking at 37°C for 30 min. The cells were then washed and resuspended in complete medium with 5% fetal calf serum (FCS). About 8X 105, 1.5X 105 and 2X 104 viable thyroid cells were distributed respectively into 35 mm (Linbro Chemical Co. FB-6-TC), 16 mm (Falcon 3047, Oxnard, CA, U.S.A.) and microtest II (Falcon 3042) flat bottomed culture plates. These cultures were used on day+ 11 because of optimal TEC expression of the membrane antigens tested [8, 9] and minimal fibroblast or macrophage contamination [1 ~1].
3.4. In vitro cocultures ( PSS) Twelve million, 4X 106 or 0.25X 106 spleen lymphocytes were settled onto TEC respectively grown in 35 mm, 16 mm or microtest II plates with the duration of coculture varying from 6 to 72 h. When specified, 1 #Ci/well of [3 H]thymidine was added. After 18 h the cells were harvested with a multiple automated sample harvester and the radioactivity incorporated by the cells was determined by liquid scintillation spectrometry. Cultures were performed in triplicates. The results were expressed as mean 286
cpm of thymidine incorporation of sensitized lymphocytes minus mean cpm of thymidine incorporation of control lymphocytes alone and TEC alone (A cpm). 3.5. Immunofluorescence studies (a) Fresh TEC cultures: At the end of the coculture, syngeneic lymphocytes were removed and TEC were subjected to three vigorous washes with Ca 2+ and Mg2- free phosphate buffered saline (PBS) supplemented with 1% BSA. Then, 1 ml of 1/50 diluted monoclonal anti-Ia k (2) antibody (mcda k (2)) isotype IgG 2b (Beckton Dickinson, Sunnyvale, CA, U.S.A.) was added to the TEC for 30 min at room temperature (RT °) and again three washes were performed, before a further 30 min incubation at RT ° with 1 ml of 1/5 diluted fluorescein conjugated goat anti-mouse (FITC-GAM IgG 2b, Nordic, Tilburg, The Netherlands). Again three vigorous washes were performed. TEC monolayers were then fixed for i0 min at 20°C with a mixture of acetic acid (5%) and ethanol (95%) before observation under the fluorescent microscope (Leitz, Orthomat-Orthoplan, F.R.G.). One hundred and fifty TEC were scored by experiments. Evaluation of Ia positivity on control TEC monolayers which were not used for coculture with syngeneic lymphocytes never exceed 3% of la + TEC for the different determinations. (b) Fixed TEC cultures: After removal of syngeneic lymphocytes or solutions and before anti-serum deposition, TEC were fixed and then labeled as described above. Each percentage of la + thyroid cells was obtained by counting 150 cells. The number of spontaneous Ia+ cells on control TEC monolayers (without syngeneic lymphocytes or soluble factors) was determined by staining control TEC monolayers with the above experimental protocol. 3.6. Absorption experiments Twelve million splenic lymphocytes were deposited onto 4X 106, I 1-day-old syngeneic TEC and cultured for 24 h at 37°C. Lymphocytes were then collected and TEC monolayers vigorously washed 3 times before the addition of 200 #1 (diluted 1/ 10) m a l a k (2) antibody. After one hour of incubation at RT °, the m a l a k (2) antibody was collected and its cytotoxicity against fresh splenic H-2 k lymphocytes determined.
3.7. Cytotoxicity experiments Cytotoxicity was performed by incubating 3X 105 CBA spleen cells with serial dilutions of 50 #1 of mt~Iak (2) for 15 min at RT °. Then 50 ~tl of rabbit complement (RC) (Cedarlane, Hornby, Ontario, Canada) 1/12 final dilution were added and incubated for 45 min at 37°C. Viability was evaluated by trypan blue dye exclusion. Results are expressed as cytotoxicity indexes determined according to the following formula: experimental cytotoxicity (%) 100
control cytotoxicity (%)
control cytotoxicity (%)
xl00
where control percentages of cytotoxicity were obtained using normal CBA serum or monoclonal a K k (Beckton Dickinson, Sunnyvale, CA, U.S.A.). 3.8. Blocking of PSS by antisera One hundred/~1 of m a I a k (2) antibody (diluted 1/40) were deposited for 30 min at RT ° onto 1 lday-old TEC. Free antibodies were removed by three washes with HBSS and 0.25X 106 normal syngeneic spleen cells were cocultured for three days of PSS with the TEC monolayers. When specified, excess m a l a k (2) antibody was left during PSS. In some experiments, blocking by m a l a k (2) antibody was performed after 48 h of PSS. In these cases, syngeneic lymphocytes were removed by three vigorous washes with HBSS and the m a l a k (2) antibody (or mouse IgG antibodies as controls) was added to TEC as described above before a new PSS with 0.25X 106 fresh syngeneic lymphocytes.
4. Results and Discussion
In a first set of experiments, 12X 106 lymphocytes
were cocultured for 0 to 48 h on I l-day-old monolayers of syngeneic TEC and expression of la antigens of TEC was evaluated by indirect immunofluorescence after 2, 4, 6, 8, 18, 24 and 48 h of culture. As shown on Table 1, Ia÷ antigens were never detected on control cultures (without lymphocytes) or on experimental cultures of TEC when coculture with syngeneic lymphocytes did not exceed 4 h. In contrast, as early as 6 to 8 h after lymphocyte deposition onto syngeneic TEC, 27.5% of TEC bear Ia antigens. It must be noted that Ia÷ TEC increased until 24 h of coculture with syngeneic lymphocytes, and then started to decrease when measured after 48 h of coculture. Another interesting observation was the particular localization of labeling on TEC in relationship to the duration of coculture with syngeneic lymphocytes. As early as 6 to 8 h of coculture, la antigens seem to be expressed only in the perinuclear area of the TEC. After 24 h of coculture, staining is uniformly detected in the cell cytoplasm, and found exclusively along the cytoplasmic membrane when cocultured for 48 h. This sequential labelling strongly favors an intrinsic presence of Ia antigens due to synthesis or rearrangement and rules out a passive deposition by contaminating cells. We confirmed this appearance of Ia antigens on TEC by absorption of a l a k (2) monoclonal antibodies on TEC previously cocultured for 24 h with syngeneic lymphocytes and by measuring the cytotoxicity in the presence of rabbit complement (RC) of these absorbed ala k (2) monoclonal antibodies against total CBA spleen cells. In Table 2, it is clearly shown that compared to non-absorbed cda k (2) antibodies, the TEC absorbed antibodies totally lost their cytotoxicity against total spleen cells from H-2 k mice.
Table 1 Induction of Ia antigen expression of TEC after various durations of coculture with 12X 106 of syngeneic lymphocytes ( m e a n + SEM of 3 determinations) Nature of the antibodies deposited onto TEC
mt~la k ( 2 ) + F I T C - G A M IgGzb F I T C - G A M IgG2b
Duration of lymphocyte coculture with syngeneic TEC (h) 2 4
5-8
18
24
48
0
27.5+2.5 0
43.0+9.1 2
45.6+5.3 I
19.3+3.8 0
287
Table 2 Cytotoxicity of monoclonal a l a k antibodies absorbed on TEC used previously for 24 h coculture with syngeneic lymphocytes Treatment of monoclonal otla k (2) sera
None Absorbed on TEC used for 24 h coculture with syngeneic lymphocytes
Monoclonal otla k (2) dilutions 1/1000
1/2000
1/4000
1/8000
24,02
25.0
27.0
14.0
0
0
4.0
1.7
We further confirmed these data by blocking PSS with monoclonal a l a k antibodies added either to TEC monolayers which were then extensively washed before lymphocyte deposition for PSS or simultaneously with lymphocytes, the excess cda k (2) remaining present during the 3-day PSS. The results of this experiment, given in Table 3, showed that, when a l a k (2) antibodies were previously deposited onto TEC for syngeneic PSS, 62.2% of the usual PSS is obtained. In contrast, when cda k (2) antibodies are added simultaneously and left on TEC during the PSS, blocking by a l a k (2) antibodies was much more effective with only 18.8% of the normal PSS being obtained (81.2% blocking). The fluorescence studies showed that during the first 24 h of PSS of normal lymphocytes on monolayers of TEC, induction of Ia antigens on the apical side of the TEC were detected. This expression of Ia antigens on TEC in direct contact with the responder syngeneic lymphocytes provides a satisfactory explanation for the I-A subregion restricted PSS. It must also be noted that allogeneic lymphocytes are unable to induce class II antigen expression. This result is in agreement with the absence of allogeneic stimulation by TEC we found [7]. In contrast to
others [11, 12], we previously reported [8, 9] the existence of Ia antigens on cultured murine TEC, and demonstrated their localization only on the basal part of the cultured TEC. We also demonstrated that cultured TEC "in vitro" expose the cell membrane corresponding to the in vivo apical surface normally oriented towards the inside of the follicules. It was recently reported that lectins could induce the expression of DR antigens on cultured human follicular thyroid cells [13] with an optimal expression on days 4-5 of thyroid cell culture. Similar to these results, this appearance of Ia antigens on cultured murine TEC, 24 h after the deposition of syngeneic lymphocytes, can be thought to represent by extension, the fundamental mechanism of induction of organ specific autoimmune disorders. However, several questions are raised: are the la k antigens the result of a de novo synthesis or are they preexistent antigen structures, simply elicited by the stimulation of syngeneic lymphocytes? The kinetics of Ia antigen appearance strongly favor a rearrangement mechanism rather than the detection of the new structures which form very quickly, whereas it takes 24 h to obtain this optimal expression of Ia on TEC. Moreover, these antigens appear initially in a patchy distribution around the TEC nucleus, then in the cytoplasm and lastly on the membrane surface before disappearing. Another mechanism, which can be thought to be responsible for this induction of Ia antigens during the 24 h incubation of TEC with syngeneic lymphocytes, is the largely described phenomenon of antigenic modulation (for review see [15, 16]). This phenomenon which is temperature dependent, is detected at 37°C in a few minutes for B cell immunoglobulins, but is consistently and incompletely obtained for M H C antigens. Several elements argue in
Table 3 Blocking of PSS on monolayers of TEC by monoclonal t~la k (2) antibody m a l a k (2) antibody added 0 + +
288
Time of addition
Before PSS (2) During PSS (3)
Thymidine uptake ( c p m x l0 -3)
% of response
3.827±328 (l) 2.3805:255 7215:95
100 62.2 18.8
favor of this hypothesis. The long duration was necessary for the appearance of the Ia antigen on the cultured TEC surface (no detection was possible before 6 to 8 h, with a maximum at 24 h of culture). This could explain the lack of detection at the usual induction times of antigenic modulation. Moreover, la expression is not obtained on lymphoid cell line surface but is detected on particular thyroid epithelial cells. The patchy distribution of this Ia expression and its disappearance also argue in favor of a lymphocyte-induced antigenic modulation. However, several facts can be considered to be in opposition to this hypothesis. Firstly, antigenic modulation occurs on a cell surface normally expressing the antigen and Ia antigens are not present on the apical side of the TEC. Moreover, in this particular model, expression of Ia antigens is due to normal syngeneic T lymphocytes and not to antibodies, as usually described. Another attractive hypothesis consists of comparing this lymphocyte-induced expression of Ia antigens to that obtained on rat epidermal cells during graft-versus-host disease or by local immunization stimuli, such as contact sensitizing agents applied to the skin. In these cases la antigen expression is observed 24 h later at the sites of contact with these stimuli [17]. The authors demonstrate that the local recognition of antigen by lymphocytes can result in the induction of Ia antigen expression in a patchy distribution by keratinocytes, this antigen being synthesized by the cells in which it is found and not passively acquired from other cells. Similarly to the graft-versus-host experiments reported above, in our experiments, syngeneic T cells are responsible for the Ia antigen expression observed after several hours of culture on thyroid epithelial cell surfaces which apparently do not intervene in the immune system. The role, and moreover, the significance of this appearance of Ia antigens are now under investigation in our laboratory. It seems that this mechanism of T cell restricted thyroid specific syngeneic recognition and stimulation occurs, as assessed by both the I-A subregion restricted PSS and this appearance of Ia antigens on TEC, via thyroid cells. This phenomenon of class II antigen induction on TEC by syngeneic or autologous lymphocytes, if applied to humans, could explain the appearance of thyroid disorders, such as Hashimoto's or Graves' disease. Thyroid antigens (Tg, TSH receptor . . . . )
and self Ia or DR recognition may be considered the inducer phenomenon, which could be triggered by viral infection, hapten aggression or external events. Taking into account the data reported above and this possible mechanism of cell recognition, it seems to us that thyroid cells could play the role of the antigen presenting cell.
Acknowledgements The authors are indebted to Ms. E. Lallemand for her excellent technical assistance, to Ms. J. Jacobson for editorial advice and to Mrs. J. Decaix for typing the manuscript.
References [1] Yeni, P., Michel-Bechet, M., Athouel-Haon, A. M., Fayet, G. and Charreire, J. (1980) in: Autoimmune Aspects of Endocrine Disorders (Pinchera et al., Eds.) pp. 204 207, Academic Press. [2] Yeni, P. and Charreire, J. (1981) Cell. lmmunol. 62, 313-323. [3] Charreire, J. (1982) Eur. J. immunol. 5,416~,21. [4] Charreire, J. and Michel-Bechet, M. (1982) Eur. J. Immunol. 5,421 425. [5] Salamero, J. and Charreire, J. (1983) Cell. lmmunol. 78, 387-391. [6] Vladutiu, A. O. and Rose, N. R. (1971) Science 174, 1137 1139. [7] Salamero, J. and Charreire, J. (1983) Eur. J. Immunol. 13, 948 955. [8] Salamero, J., Michel-Bechet, M. and Charreire, J. (1981) C.R. Acad. Sci. (Paris) 293,744 750. [9] Salamero, J., Michel-Bechet, M. and Charreire, J. (1983) Tissue Antigens 22, 231 238. [10] Salamero, J. and Charreire, J., submitted for publication. [11] Paar, E. L., Bowen, K. M. and Lafferty, K. J. (1980) Transplantation 30, 135. [12] Pujoll-Borrell, R., Hanafusa, T. and Doniach, D. (1982) Ann. Endocrinol. 43, 58A. [13] Pujoll-Borrell, R., Hanafusa, T., Chiovato, L. and Bottazzo G. F. (1983) Nature (London) 304, 71. [14] Hirsch, M. R., Wietzerbin, J., Pierres, M. and Goridis, C. (1983)'Neurosei. Lett. 41, 199-204. [15] Chatenoud , L. and Bach, J. F. (1984) Immunol. Today 5, 20-25. [16] Boyse, E. A., Old, L. J. and Luell, S. J. (1963) Natl. Cancer Inst. 31,987. [17] Barclay, N. and Masson, D. W. (1982) J. Exp. Med. 156, 1665-1676.
289
9 (1985) 285 289
Elsevier lmlet 573
INDUCTION OF Ia ANTIGENS ON THYROID CULTURES BY SYNGENEIC T LYMPHOCYTES: A MODEL OF A U T O I M M U N E DISEASE Jean S A L A M E R O and Jeannine C H A R R E I R E I N S E R M U. 283, H6pital Cochin, 27, rue du Faubourg Saint-Jacques, 75674 Paris Cedex 14, France
(Received9 October 1984) (Modified version received20 December1984) (Accepted 21 December1984)
1. Summary We previously reported that the expression of Ia antigens on cultured monolayers of murine thyroid epithelial cells (TEC) occurred with a particular distribution exclusively on the basal part of the cultured thyroid cells, while class I antigens of the major histocompatibility complex (MHC) are only detected on the apical surface. It appears that deposition of syngeneic lymphocytes induces, 24 h later, Ia expression on the apical side of cultured TEC, the surface that is in direct contact with the responder lymphocytes during syngeneic sensitization of T lymphocytes. We hypothesized that this phenomenon could represent, in syngeneic situations, the restriction process in antigen recognition by T cells, as demonstrated by la restricted primary syngeneic sensitization (PSS) on murine TEC. 2. Introduction In the past few years we have developed an "in vitro" model of primary syngeneic sensitization (PSS) on monolayers of thyroid epithelial cells (TEC). We demonstrated that sensitization was directed against structures borne by syngeneic TEC only [1], that specifically sensitized T lymphocytes [2, 3] could induce thyroid disorders when injected into intact syngeneic recipients [4], and that this proliferation was Key words:
thyroid - autoimmunity la antigens - t lym-
phocytes J. Charreire, INSERM U. 283, HSpital Cochin, Pavilion Hardy A, 27, rue du Faubourg SaintJacques, 75674 Paris Cedex 14, France Address correspondence to:
under the same genetic control [5] as experimental autoimmune thyroiditis (EAT) [6]. Using congenic strains of mice, we found that optimal PSS was obtained when compatibility between stimulator TEC and responder lymphocytes occurs at the I-A subregion of the major histocompatibility complex (MHC) [7]. Simultaneously, we investigated the expression of the M H C antigens (K end, D end and Ia) and thyroglobulin (Tg) on these in vitro cultured TEC. Results obtained by three different direct and indirect methods [8, 9] showed that class I MHC antigens (K and D ends) and Tg are expressed exclusively on the apical cell surface of TEC while class II M H C antigens (Ia) are detected on the opposing basal face of the monolayer. The dichotomy of distribution of MHC antigens on cultured TEC, and especially the existence of la antigens on the basal part of TEC which is not accessible to responder lymphocytes was in total opposition to the strict compatibility at the I-A subregion required to obtain optimal PSS of syngeneic lymphocytes on cultured TEC. Since the role of macrophages in PSS was fully ruled out [10], we envisaged two hypotheses to explain these mutually exclusive data. Either lymphocytes after deposition cross the TEC monolayer to reach the basal side of the thyroid cells which bears the Ia antigens, or lymphocyte deposition is followed by an apparition of Ia antigens on the apical side of the thyroid monolayer which is in contact with the responder syngeneic lymphocytes. We investigated the second hypothesis by comparing Ia antigen expression on monolayers of TEC from CBA (H-2 k) mice before and at different times after the deposition of syngeneic lymphocytes using a different method.
0165 2478 / 85 / $ 3.30 © 1985 ElsevierScience Publishers B.V.(BiomedicalDivision)
285
3. Materials and Methods
3.1. Animals CBA (H-2 k) mice were provided by the C.S.E.A.L. Orl6ans-la-Source (France). 3.2. Lymphoid cell suspension Spleen cells were aseptically removed and homogenized in a glass tube with a loose fitting Teflon pestle in Hanks' balanced salt solution (HBSS). The cell suspension was then washed twice in HBSS supplemented with penicillin (100 U/ml, streptomycin (100 #g/ml) and fungizone (2.5 #g/ml). Cells were adjusted to a final concentration of 2.5× 106/ml in R P M I 1640 (Gibco, Paisley, Scotland) supplemented with penicillin (100 U/ml), streptomycin (I00 #g/ml), fungizone (2.5/zg/ml) and 1% L-glutamine (complete medium). 3.3. Mouse thyroid cell cultures These were performed as previously reported [1-5]. Briefly, thyroids were carefully dissected and minced into small pieces at 4°C in complete medium. The suspension was incubated with 1.5 mg/ml of collagenase (Boehringer, Mannheim, F.R.G.) under shaking at 37°C for 30 min. The cells were then washed and resuspended in complete medium with 5% fetal calf serum (FCS). About 8X 105, 1.5X 105 and 2X 104 viable thyroid cells were distributed respectively into 35 mm (Linbro Chemical Co. FB-6-TC), 16 mm (Falcon 3047, Oxnard, CA, U.S.A.) and microtest II (Falcon 3042) flat bottomed culture plates. These cultures were used on day+ 11 because of optimal TEC expression of the membrane antigens tested [8, 9] and minimal fibroblast or macrophage contamination [1 ~1].
3.4. In vitro cocultures ( PSS) Twelve million, 4X 106 or 0.25X 106 spleen lymphocytes were settled onto TEC respectively grown in 35 mm, 16 mm or microtest II plates with the duration of coculture varying from 6 to 72 h. When specified, 1 #Ci/well of [3 H]thymidine was added. After 18 h the cells were harvested with a multiple automated sample harvester and the radioactivity incorporated by the cells was determined by liquid scintillation spectrometry. Cultures were performed in triplicates. The results were expressed as mean 286
cpm of thymidine incorporation of sensitized lymphocytes minus mean cpm of thymidine incorporation of control lymphocytes alone and TEC alone (A cpm). 3.5. Immunofluorescence studies (a) Fresh TEC cultures: At the end of the coculture, syngeneic lymphocytes were removed and TEC were subjected to three vigorous washes with Ca 2+ and Mg2- free phosphate buffered saline (PBS) supplemented with 1% BSA. Then, 1 ml of 1/50 diluted monoclonal anti-Ia k (2) antibody (mcda k (2)) isotype IgG 2b (Beckton Dickinson, Sunnyvale, CA, U.S.A.) was added to the TEC for 30 min at room temperature (RT °) and again three washes were performed, before a further 30 min incubation at RT ° with 1 ml of 1/5 diluted fluorescein conjugated goat anti-mouse (FITC-GAM IgG 2b, Nordic, Tilburg, The Netherlands). Again three vigorous washes were performed. TEC monolayers were then fixed for i0 min at 20°C with a mixture of acetic acid (5%) and ethanol (95%) before observation under the fluorescent microscope (Leitz, Orthomat-Orthoplan, F.R.G.). One hundred and fifty TEC were scored by experiments. Evaluation of Ia positivity on control TEC monolayers which were not used for coculture with syngeneic lymphocytes never exceed 3% of la + TEC for the different determinations. (b) Fixed TEC cultures: After removal of syngeneic lymphocytes or solutions and before anti-serum deposition, TEC were fixed and then labeled as described above. Each percentage of la + thyroid cells was obtained by counting 150 cells. The number of spontaneous Ia+ cells on control TEC monolayers (without syngeneic lymphocytes or soluble factors) was determined by staining control TEC monolayers with the above experimental protocol. 3.6. Absorption experiments Twelve million splenic lymphocytes were deposited onto 4X 106, I 1-day-old syngeneic TEC and cultured for 24 h at 37°C. Lymphocytes were then collected and TEC monolayers vigorously washed 3 times before the addition of 200 #1 (diluted 1/ 10) m a l a k (2) antibody. After one hour of incubation at RT °, the m a l a k (2) antibody was collected and its cytotoxicity against fresh splenic H-2 k lymphocytes determined.
3.7. Cytotoxicity experiments Cytotoxicity was performed by incubating 3X 105 CBA spleen cells with serial dilutions of 50 #1 of mt~Iak (2) for 15 min at RT °. Then 50 ~tl of rabbit complement (RC) (Cedarlane, Hornby, Ontario, Canada) 1/12 final dilution were added and incubated for 45 min at 37°C. Viability was evaluated by trypan blue dye exclusion. Results are expressed as cytotoxicity indexes determined according to the following formula: experimental cytotoxicity (%) 100
control cytotoxicity (%)
control cytotoxicity (%)
xl00
where control percentages of cytotoxicity were obtained using normal CBA serum or monoclonal a K k (Beckton Dickinson, Sunnyvale, CA, U.S.A.). 3.8. Blocking of PSS by antisera One hundred/~1 of m a I a k (2) antibody (diluted 1/40) were deposited for 30 min at RT ° onto 1 lday-old TEC. Free antibodies were removed by three washes with HBSS and 0.25X 106 normal syngeneic spleen cells were cocultured for three days of PSS with the TEC monolayers. When specified, excess m a l a k (2) antibody was left during PSS. In some experiments, blocking by m a l a k (2) antibody was performed after 48 h of PSS. In these cases, syngeneic lymphocytes were removed by three vigorous washes with HBSS and the m a l a k (2) antibody (or mouse IgG antibodies as controls) was added to TEC as described above before a new PSS with 0.25X 106 fresh syngeneic lymphocytes.
4. Results and Discussion
In a first set of experiments, 12X 106 lymphocytes
were cocultured for 0 to 48 h on I l-day-old monolayers of syngeneic TEC and expression of la antigens of TEC was evaluated by indirect immunofluorescence after 2, 4, 6, 8, 18, 24 and 48 h of culture. As shown on Table 1, Ia÷ antigens were never detected on control cultures (without lymphocytes) or on experimental cultures of TEC when coculture with syngeneic lymphocytes did not exceed 4 h. In contrast, as early as 6 to 8 h after lymphocyte deposition onto syngeneic TEC, 27.5% of TEC bear Ia antigens. It must be noted that Ia÷ TEC increased until 24 h of coculture with syngeneic lymphocytes, and then started to decrease when measured after 48 h of coculture. Another interesting observation was the particular localization of labeling on TEC in relationship to the duration of coculture with syngeneic lymphocytes. As early as 6 to 8 h of coculture, la antigens seem to be expressed only in the perinuclear area of the TEC. After 24 h of coculture, staining is uniformly detected in the cell cytoplasm, and found exclusively along the cytoplasmic membrane when cocultured for 48 h. This sequential labelling strongly favors an intrinsic presence of Ia antigens due to synthesis or rearrangement and rules out a passive deposition by contaminating cells. We confirmed this appearance of Ia antigens on TEC by absorption of a l a k (2) monoclonal antibodies on TEC previously cocultured for 24 h with syngeneic lymphocytes and by measuring the cytotoxicity in the presence of rabbit complement (RC) of these absorbed ala k (2) monoclonal antibodies against total CBA spleen cells. In Table 2, it is clearly shown that compared to non-absorbed cda k (2) antibodies, the TEC absorbed antibodies totally lost their cytotoxicity against total spleen cells from H-2 k mice.
Table 1 Induction of Ia antigen expression of TEC after various durations of coculture with 12X 106 of syngeneic lymphocytes ( m e a n + SEM of 3 determinations) Nature of the antibodies deposited onto TEC
mt~la k ( 2 ) + F I T C - G A M IgGzb F I T C - G A M IgG2b
Duration of lymphocyte coculture with syngeneic TEC (h) 2 4
5-8
18
24
48
0
27.5+2.5 0
43.0+9.1 2
45.6+5.3 I
19.3+3.8 0
287
Table 2 Cytotoxicity of monoclonal a l a k antibodies absorbed on TEC used previously for 24 h coculture with syngeneic lymphocytes Treatment of monoclonal otla k (2) sera
None Absorbed on TEC used for 24 h coculture with syngeneic lymphocytes
Monoclonal otla k (2) dilutions 1/1000
1/2000
1/4000
1/8000
24,02
25.0
27.0
14.0
0
0
4.0
1.7
We further confirmed these data by blocking PSS with monoclonal a l a k antibodies added either to TEC monolayers which were then extensively washed before lymphocyte deposition for PSS or simultaneously with lymphocytes, the excess cda k (2) remaining present during the 3-day PSS. The results of this experiment, given in Table 3, showed that, when a l a k (2) antibodies were previously deposited onto TEC for syngeneic PSS, 62.2% of the usual PSS is obtained. In contrast, when cda k (2) antibodies are added simultaneously and left on TEC during the PSS, blocking by a l a k (2) antibodies was much more effective with only 18.8% of the normal PSS being obtained (81.2% blocking). The fluorescence studies showed that during the first 24 h of PSS of normal lymphocytes on monolayers of TEC, induction of Ia antigens on the apical side of the TEC were detected. This expression of Ia antigens on TEC in direct contact with the responder syngeneic lymphocytes provides a satisfactory explanation for the I-A subregion restricted PSS. It must also be noted that allogeneic lymphocytes are unable to induce class II antigen expression. This result is in agreement with the absence of allogeneic stimulation by TEC we found [7]. In contrast to
others [11, 12], we previously reported [8, 9] the existence of Ia antigens on cultured murine TEC, and demonstrated their localization only on the basal part of the cultured TEC. We also demonstrated that cultured TEC "in vitro" expose the cell membrane corresponding to the in vivo apical surface normally oriented towards the inside of the follicules. It was recently reported that lectins could induce the expression of DR antigens on cultured human follicular thyroid cells [13] with an optimal expression on days 4-5 of thyroid cell culture. Similar to these results, this appearance of Ia antigens on cultured murine TEC, 24 h after the deposition of syngeneic lymphocytes, can be thought to represent by extension, the fundamental mechanism of induction of organ specific autoimmune disorders. However, several questions are raised: are the la k antigens the result of a de novo synthesis or are they preexistent antigen structures, simply elicited by the stimulation of syngeneic lymphocytes? The kinetics of Ia antigen appearance strongly favor a rearrangement mechanism rather than the detection of the new structures which form very quickly, whereas it takes 24 h to obtain this optimal expression of Ia on TEC. Moreover, these antigens appear initially in a patchy distribution around the TEC nucleus, then in the cytoplasm and lastly on the membrane surface before disappearing. Another mechanism, which can be thought to be responsible for this induction of Ia antigens during the 24 h incubation of TEC with syngeneic lymphocytes, is the largely described phenomenon of antigenic modulation (for review see [15, 16]). This phenomenon which is temperature dependent, is detected at 37°C in a few minutes for B cell immunoglobulins, but is consistently and incompletely obtained for M H C antigens. Several elements argue in
Table 3 Blocking of PSS on monolayers of TEC by monoclonal t~la k (2) antibody m a l a k (2) antibody added 0 + +
288
Time of addition
Before PSS (2) During PSS (3)
Thymidine uptake ( c p m x l0 -3)
% of response
3.827±328 (l) 2.3805:255 7215:95
100 62.2 18.8
favor of this hypothesis. The long duration was necessary for the appearance of the Ia antigen on the cultured TEC surface (no detection was possible before 6 to 8 h, with a maximum at 24 h of culture). This could explain the lack of detection at the usual induction times of antigenic modulation. Moreover, la expression is not obtained on lymphoid cell line surface but is detected on particular thyroid epithelial cells. The patchy distribution of this Ia expression and its disappearance also argue in favor of a lymphocyte-induced antigenic modulation. However, several facts can be considered to be in opposition to this hypothesis. Firstly, antigenic modulation occurs on a cell surface normally expressing the antigen and Ia antigens are not present on the apical side of the TEC. Moreover, in this particular model, expression of Ia antigens is due to normal syngeneic T lymphocytes and not to antibodies, as usually described. Another attractive hypothesis consists of comparing this lymphocyte-induced expression of Ia antigens to that obtained on rat epidermal cells during graft-versus-host disease or by local immunization stimuli, such as contact sensitizing agents applied to the skin. In these cases la antigen expression is observed 24 h later at the sites of contact with these stimuli [17]. The authors demonstrate that the local recognition of antigen by lymphocytes can result in the induction of Ia antigen expression in a patchy distribution by keratinocytes, this antigen being synthesized by the cells in which it is found and not passively acquired from other cells. Similarly to the graft-versus-host experiments reported above, in our experiments, syngeneic T cells are responsible for the Ia antigen expression observed after several hours of culture on thyroid epithelial cell surfaces which apparently do not intervene in the immune system. The role, and moreover, the significance of this appearance of Ia antigens are now under investigation in our laboratory. It seems that this mechanism of T cell restricted thyroid specific syngeneic recognition and stimulation occurs, as assessed by both the I-A subregion restricted PSS and this appearance of Ia antigens on TEC, via thyroid cells. This phenomenon of class II antigen induction on TEC by syngeneic or autologous lymphocytes, if applied to humans, could explain the appearance of thyroid disorders, such as Hashimoto's or Graves' disease. Thyroid antigens (Tg, TSH receptor . . . . )
and self Ia or DR recognition may be considered the inducer phenomenon, which could be triggered by viral infection, hapten aggression or external events. Taking into account the data reported above and this possible mechanism of cell recognition, it seems to us that thyroid cells could play the role of the antigen presenting cell.
Acknowledgements The authors are indebted to Ms. E. Lallemand for her excellent technical assistance, to Ms. J. Jacobson for editorial advice and to Mrs. J. Decaix for typing the manuscript.
References [1] Yeni, P., Michel-Bechet, M., Athouel-Haon, A. M., Fayet, G. and Charreire, J. (1980) in: Autoimmune Aspects of Endocrine Disorders (Pinchera et al., Eds.) pp. 204 207, Academic Press. [2] Yeni, P. and Charreire, J. (1981) Cell. lmmunol. 62, 313-323. [3] Charreire, J. (1982) Eur. J. immunol. 5,416~,21. [4] Charreire, J. and Michel-Bechet, M. (1982) Eur. J. Immunol. 5,421 425. [5] Salamero, J. and Charreire, J. (1983) Cell. lmmunol. 78, 387-391. [6] Vladutiu, A. O. and Rose, N. R. (1971) Science 174, 1137 1139. [7] Salamero, J. and Charreire, J. (1983) Eur. J. Immunol. 13, 948 955. [8] Salamero, J., Michel-Bechet, M. and Charreire, J. (1981) C.R. Acad. Sci. (Paris) 293,744 750. [9] Salamero, J., Michel-Bechet, M. and Charreire, J. (1983) Tissue Antigens 22, 231 238. [10] Salamero, J. and Charreire, J., submitted for publication. [11] Paar, E. L., Bowen, K. M. and Lafferty, K. J. (1980) Transplantation 30, 135. [12] Pujoll-Borrell, R., Hanafusa, T. and Doniach, D. (1982) Ann. Endocrinol. 43, 58A. [13] Pujoll-Borrell, R., Hanafusa, T., Chiovato, L. and Bottazzo G. F. (1983) Nature (London) 304, 71. [14] Hirsch, M. R., Wietzerbin, J., Pierres, M. and Goridis, C. (1983)'Neurosei. Lett. 41, 199-204. [15] Chatenoud , L. and Bach, J. F. (1984) Immunol. Today 5, 20-25. [16] Boyse, E. A., Old, L. J. and Luell, S. J. (1963) Natl. Cancer Inst. 31,987. [17] Barclay, N. and Masson, D. W. (1982) J. Exp. Med. 156, 1665-1676.
289