Experimental autoimmune thyroiditis in nonobese diabetic mice lacking interferon regulatory factor-1

Clinical Immunology 113 (2004) 187 – 192 www.elsevier.com/locate/yclim

Experimental autoimmune thyroiditis in nonobese diabetic mice lacking interferon regulatory factor-1 Zhongtian Jina, Kouki Morib,*, Keisei Fujimoria, Saeko Hoshikawab, Jun-ichi Tanib, Jo Satohc, Sadayoshi Itob, Susumu Satomia, Katsumi Yoshidad b

a Division of Advanced Surgery and Surgical Oncology, Tohoku University Graduate School of Medicine, Sendai, Japan Division of Nephrology, Endocrinology and Vascular Medicine, Tohoku University Graduate School of Medicine, Sendai, Japan c Division of Molecular Metabolism and Diabetes, Tohoku University Graduate School of Medicine, Sendai, Japan d Division of Rheumatology and Hematology, Tohoku University Graduate School of Medicine, Sendai, Japan

Received 7 May 2004; accepted 22 June 2004 Available online 8 September 2004

Abstract Interferon regulatory factor-1 (IRF-1) is pivotal in the regulation of interferon (IFN)-mediated immune reactions, and studies suggest that IRF-1 is involved in the development of autoimmune diseases. IRF-1+/+, +/ , and / nonobese diabetic (NOD) mice were immunized with mouse thyroglobulin (mTg) to determine whether IRF-1 is required in experimental autoimmune thyroiditis (EAT), a murine model for Hashimoto’s thyroiditis (HT). IRF-1-deficient mice developed EAT and anti-mTg antibodies comparable to IRF-1+/+ and +/ mice. Whereas both CD4+ and CD8+ T cells were found in thyroids of IRF-1+/+ mice, the latter was not in IRF-1 / mice. Major histocompatibility complex class II antigen was comparably expressed in thyroids of IRF-1+/+ and / mice. Lack of IRF-1 resulted in decreased CD8+ T cell number in the spleen and reduced IFNg production by splenocytes. Our results suggest that IRF-1 is not pivotal in EAT in NOD mice. D 2004 Elsevier Inc. All rights reserved. Keywords: IRF-1; NOD mice; Thyroiditis

Introduction Experimental autoimmune thyroiditis (EAT), a murine model for Hashimoto’s thyroiditis (HT) in the human, can be induced by immunization with thyroglobulin (Tg) and adjuvant [1]. In addition, spleen or lymph node cells from mice primed in vivo with Tg and adjuvant and restimulated in vitro with Tg can transfer EAT to unimmunized mice [2]. Susceptibility to murine EAT induction is controlled by major histocompatibility complex (MHC) class II locus. In fact, I-Ak-, I-Aq-, or I-As-bearing strains predispose to EAT [1]. Prominent features of the disease include mononuclear * Corresponding author. Division of Nephrology, Endocrinology and Vascular Medicine, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan. Fax: +81 22 717 7168. E-mail address: [email protected] (K. Mori). 1521-6616/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.clim.2004.06.008

cell infiltration in the thyroid and production of autoantibody to Tg [1]. Thyroid infiltrates consist mainly of CD4+ and CD8+ T cells [3]. CD4+ T cells play a pivotal role in EAT development since depletion of CD4+ T cells prevents intrathyroidal lymphocytic infiltration and anti-mouse Tg antibody (anti-mTg Ab) production [4,5]. In contrast, the role of CD8+ T cells in EAT remains poorly understood. Depletion of CD8+ T cells has little effect on the development of EAT [4], and h2-microglobulin knockout mice lacking CD8+ T cells develop EAT [6]. Thus, studies suggest that CD8+ T cells may not be essential in EAT development. On the other hand, CD8+ T cells are involved in resolution of granulomatous EAT [7]. Recently, we have shown that iodine treatment fails to induce lymphocytic thyroiditis (LT) in interferon regulatory factor-1 (IRF-1) knockout nonobese diabetic (NOD) mice [8], in which the number of CD8+ T cells is markedly

188

Z. Jin et al. / Clinical Immunology 113 (2004) 187–192

decreased in the thymus and peripheral lymphoid organs [8,9]. IRF-1 is a transcription factor mediating interferon (IFN) action [10] and is involved in the differentiation of CD8+ T cells [11] and in the development of Th1 immune reaction [12]. Taken together, IRF-1-deficient mice can be used to determine the role of IRF-1 and further CD8+ T cells and/or Th1 response in the pathogenesis of autoimmune diseases. In fact, targeted disruption of IRF-1 gene in mice resulted in prevention of collagen-induced arthritis [13]. Further, we have recently demonstrated that NOD mice lacking IRF-1 did not develop insulitis and subsequent diabetes [14]. NOD mice, which possess MHC class II I-Ag7, develop both iodine-induced LT and EAT [15]. Accordingly, using IRF-1 null NOD mice as well as IRF1+/+ and +/ mice, we studied the role of IRF-1 in the development of EAT.

Immunohistochemistry Cryostat sections (5-Am-thick) of thyroids, fixed in 4% phosphate-buffered paraformaldehyde solution, were incubated with 10% normal goat serum and then with either biotin-conjugated rat anti-mouse CD4 (Caltag, South San Francisco, CA), biotin-conjugated anti-mouse CD8 (Caltag), or biotin-conjugated anti-rat RT1B (OX6; PharMingen, San Diego, CA), which interacts with MHC class II I-Ag7. Endogenous peroxidase activity was quenched with methanol containing hydrogen peroxide and the sections were incubated with streptavidin-conjugated horseradish peroxidase (Caltag). The immunoreaction was visualized with 3,3V-diaminobenzidine (Dojin, Kumamoto, Japan), and then the sections were counterstained with hematoxylin. Negative controls were done as above with omission of the primary antibody, and absence of positive staining was confirmed in the controls (data not shown).

Materials and methods Anti-mouse thyroglobulin antibody Mice IRF-1-deficient NOD mice were generated and characterized as previously described [14]. All mice were genotyped by PCR analysis of tail DNA and maintained in our animal facility under specific pathogen-free conditions. This study was approved by the institutional review board of Tohoku University School of Medicine. Thyroglobulin immunization Mouse thyroglobulin (mTg) was prepared as previously described [16]. It was mixed with TiterMax Gold adjuvant (CytRx Corporation, Norcross, GA), and emulsion containing 100 Ag of mTg was administered subcutaneously to mice at 6 and 10 weeks of age (group 2). Control mice (group 1) were treated as above but mTg was replaced by saline. The mice were bred and the thyroids and spleens were removed at 14 weeks of age. Histopathology Thyroids were fixed in 4% phosphate-buffered paraformaldehyde solution and embedded in paraffin. Four-micrometer-thick sections were taken at 6–10 levels in a noncontiguous way and stained with hematoxylin and eosin. Thyroid lesions were interpreted quantitatively for EAT, defined as the percentage of thyroid infiltrated, using a scale as described previously with modification [15]: 0 = normal thyroid; 1 = less than 1% lymphocytic infiltration of the thyroid; 2 = 1–10% lymphocytic infiltration; 3 = 10– 40% lymphocytic infiltration; and 4 = greater than 40% lymphocytic infiltration. The histological specimens were interpreted by two investigators in a blind fashion. Total score for each mouse was divided by the number of observations for that mouse.

Serum anti-mTg antibody (Ab) levels were determined as previously described [8]. Results are expressed as absorbancy minus the reagent blank. Flow cytometry Spleen cells were stained as previously described [8] with either FITC-conjugated anti-CD3 or FITC-conjugated anti-B220 (Caltag). Cells stained with anti-CD3 were further incubated with either APC-conjugated anti-CD4 or PerCPlabeled anti-CD8 (PharMingen). Cell surface phenotypes were analyzed by FACSCalibur and the CellQuest software (Becton Dickinson, Mountain View, CA). Ten thousand cells were counted. In vitro proliferation and cytokine assay As previously described [8], spleen cells were cultured in flat-bottomed, 96-well tissue culture plates at a concentration of 5  105 cells/well in the presence or absence of mTg (40 Ag/ml) or concanavalin A (Con A; 1 Ag/ml, Sigma, St. Louis, MO). Following 66 h of incubation, 1 AM bromodeoxyuridine (BrdU; Amersham Biosciences, Piscatway, NJ) was added, and cells were further incubated for 6 h. BrdU incorporated into spleen cells was determined using an ELISA kit (Amersham Biosciences). IFNg and interleukin-10 (IL-10) levels in 72-h supernatants were measured using ELISA kits (Biosource International, Camarillo, CA). Statistical analysis The incidence of EAT was tested by Fisher’s exact probability test. The severity of EAT was compared by the Kruskal–Wallis test or Mann–Whitney’s U test. Serum antimTg Ab levels are shown as the median (range) and

Z. Jin et al. / Clinical Immunology 113 (2004) 187–192

189

Table 1 Incidence of EAT and serum levels of anti-mTgAb in NOD mice immunized with saline (group 1) or Tg (group 2) Group 1

2

No. Incidence of thyroiditis (%) mTg Ab (OD: 450 nm) No. Incidence of thyroiditis (%) mTgAb (OD: 450 nm)

Total

IRF-1+/+

IRF-1+/

IRF-1 /

18 33 (6/18)

5 20 (1/5) 0.64 (0.31–1.09) 7 71.4 (5/7) 1.72 (0.66–2.31)*

8 50 (4/8) 0.96 (0.53–1.53) 5 100 (5/5) 1.75 (1.01–2.05)*

5 20 (1/5) 0.72 (0.51–1.05) 5 100 (5/5)* 1.24 (0.58–1.91)

17 88 (15/17)*

The incidence of thyroiditis was tested by Fisher’s exact probability test. Serum mTgAb levels are shown as the median (range) and compared by the Mann– Whitney’s U test or the Kruskal–Wallis test. *P b 0.05 vs. group 1.

compared by the same tests as the severity of EAT. The other data are shown as the mean F SD and were compared by unpaired Student’s t test. A level of P b 0.05 was considered statistically significant.

Results As shown in Table 1, intrathyroidal mononuclear cell infiltration was found in 6 of 18 control mice (33%) in group 1. However, the severity of thyroid lesions was very low (Fig. 1). There was no significant difference in the incidence and severity of thyroiditis among the three IRF-1 genotypes of NOD mice in group 1 (Table 1 and Fig. 1). In contrast, Tg immunization resulted in significantly severer thyroid lesions (Figs. 1 and 2) and more frequent occurrence of EAT (88%) in group 2 (Table 1). The incidence of EAT in group 2 was evidently higher than that of group 1, especially in IRF-1-deficient mice. Again, however, IRF-1 genotype had no effect on the incidence and severity of EAT in group

Fig. 1. The severity of EAT in each IRF-1 genotype of NOD mice. The severity of EAT was classified as previously described with modification [15] and compared by the Kruskal–Wallis test or Mann–Whitney U test. Closed columns, group 1; open columns, group 2. Each bar shows +SD. *P b 0.05 vs. group 1.

2. In addition, no significant sex-associated difference was found in the development of EAT (data not shown). Tg immunization resulted in a significant increase in serum anti-mTg Ab levels in NOD mice except for IRF-1 / mice (Table 1). No significant difference was found in antimTg Ab levels among the three IRF-1 genotypes. IRF-1 plays a role in the differentiation of CD8+ T cells [9,11]. Hence, we analyzed the spleen cell population in IRF-1+/+ and / NOD mice by flow cytometry. As shown

Fig. 2. Hematoxylin and eosin staining of NOD mouse thyroids. (A) Thyroid of an IRF-1 / mouse in group 1; (B) thyroid of an IRF-1 / mouse in group 2. Original magnification 200.

190

Z. Jin et al. / Clinical Immunology 113 (2004) 187–192

Table 2 Splenic cell population proportions in IRF-1+/+ (n = 8) and Group

IRF-1

CD4

1

+/+ / +/+ /

78.8 91.4 69.7 90.7

2

/

(n = 5) NOD mice analyzed by flow cytometry CD8

F F F F

4.2 2.2* 16.9 2.5

17.0 4.5 16.9 4.1

F F F F

4.0 0.7* 4.9 0.7*

CD4/CD8

B220

4.8 19.5 4.2 22.7

38.6 38.0 39.6 39.9

F F F F

1.3 3.0* 0.9 3.4*

F F F F

4.9 6.1 5.6 5.6

Values are means F SD. *P b 0.05 vs. +/+ mice (unpaired Student’s t test).

in Table 2, a marked decrease in CD8+ T cells was found in IRF-1-deficient mice, leading to an increase in CD4+ T cell number and a marked increase in CD4/CD8 ratio compared with those in IRF-1+/+ mice (P b 0.05; Table 2). Absence of IRF-1 did not affect B cell (B220-positive cell) numbers. Tg immunization had no effect on the splenocyte populations in both IRF-1+/+ and / mice. Total spleen cell numbers were not different between the two IRF-1 genotypes (data not shown). We examined immunohistochemically the presence of CD4+ and CD8+ T cells in the diseased thyroid. In IRF-1+/+ mice, both CD4+ and CD8+ T cells were demonstrated in the area of lymphocytic infiltration (Fig. 3). CD4+ T cells were predominant cells in the inflamed thyroid. In contrast, CD8+ T cells were not detected in thyroid infiltrates of IRF1 / mice while accumulations of CD4+ T cells were observed in the areas of infiltration. Since aberrant MHC class II antigen expression in the thyroid is shown in patients with HT [17] and IRF-1 is involved in MHC class II expression [18], we examined intrathyroidal MHC class II expression. The antigen expression was markedly induced on both mononuclear cells and thyrocytes in the areas of lymphocytic infiltration in both IRF-1+/+ and / mice (Fig. 3). In addition, MHC class-II-expressing cells were scattered in the interstitium between thyroid follicles. Finally, we stimulated spleen cells with mTg or ConA and examined cell proliferation and production of IFNg and

IL-10. ConA, but not mTg, stimulated spleen cell proliferation in both IRF-1+/+ and / mice (data not shown). However, no difference was found between the two IRF-1 genotypes in the proliferation. MTg failed to induce IFNg or IL-10 production by spleen cells in both IRF-1+/+ and / mice (data not shown). In contrast, ConA stimulated production of IFNg by splenocytes. There was no significant difference in IFNg levels between IRF-1+/+ and / mice in group 1 (Fig. 4). Tg immunization, which resulted in a marked increase in IFNg levels in IRF-1+/+ mice, failed to induce IFNg production in IRF-1-deficient mice in group 2 (P b 0.05). While ConA stimulated IL-10 production by splenocytes of Tg-immunized mice, there was no significant difference in IL-10 levels between the two IRF-1 genotypes (data not shown).

Discussion Recently, we have demonstrated that IRF-1 null NOD mice fail to develop iodine-induced LT [8]. Our results suggest that impaired CD8+ T cell development and/or Th1 immune response associated with IRF-1 deficiency may play a role in the pathogenesis of iodine-induced LT in NOD mice. In contrast, contribution of CD8+ T cells to the pathogenesis of EAT has been studied by in vivo or in vitro depletion of CD8+ T cells, showing that CD8+ T cells play a

Fig. 3. Immunohistochemical staining of inflamed thyroid of IRF-1+/+ (A, B, C, and D) and / (E, F, G, and H) NOD mice. CD4+ T cells (B and F), CD8+ T cells (C and G), and cells expressing MHC class II I-Ag7 (D and H) were stained. Hematoxylin and eosin staining are also shown (A and E). Thyroids of three IRF-1+/+ mice and those of four IRF-1 / mice were examined. Staining patterns were consistently reproducible and representative photographs are shown. Original magnification 100.

Z. Jin et al. / Clinical Immunology 113 (2004) 187–192

Fig. 4. IFNg produced by concanavalin A (ConA)-stimulated spleen cells of NOD mice (n = 3 except for IRF-1 / mice in group 2; n = 4). Cytokine levels in culture medium were determined using ELISA kits and compared by the unpaired Student’s t test. Each bar shows +SD. *P b 0.05 vs. group 1 and **P b 0.01 vs. IRF-1+/+ mice.

minor role in EAT [4,19]. Similarly, EAT can be induced in h2-microglobulin-deficient mice, which lack CD8+ T cells [6]. Consistent with these studies, we clearly demonstrate that NOD mice lacking IRF-1 did develop EAT as well as IRF-1+/+ and +/ mice. We confirmed impaired CD8+ T cell development in IRF-1 / mice as described previously [8,9], and further we demonstrated absence of CD8+ T cells in the inflamed thyroids of IRF-1 / mice. Thus, our results support the hypothesis that CD8+ T cells play only a minor role in the development of EAT. However, we do not know why IRF-1-deficient NOD mice developed EAT but not iodine-induced LT. Whereas the fundamental role of iodine excess in the development of LT remains a matter of controversy, some studies have demonstrated that T cell recognition of Tg depends on its iodine content in rodents and the human [20–22]. Our results may propose possibility that CD8+ T cells play a role in recognition of only highly iodinated Tg. Thus, further studies are clearly required to determine the role of CD8+ T cells in both iodine-induced LT and EAT. MHC class II antigens are aberrantly expressed in the thyroids of HT patients and NOD mice [17,23]. Previous studies have demonstrated that IRF-1 is involved in MHC class II induction [18]. Thus, MHC class II expression in the thyroid was examined immunohistochemically in IRF-1+/+ and / NOD mice. We show aberrant MHC class II antigen expression in the thyroids of IRF-1 / mice comparable to +/+ mice. These results suggest that enhanced MHC class II induction may contribute to the development of EAT in IRF-1-deficient mice. However, our results may simply represent intrathyroidal MHC class II expression as a result of mononuclear cell infiltration in thyroids of mice lacking IRF-1 since MHC class II induction may be an event secondary to mononuclear cell infiltration [23,24]. In addition, our results suggest that MHC class II antigen expression may be independent of IRF-1 in the thyroid. Previous studies demonstrate impaired MHC class II induction in the kidney, heart, and spleen, but not in the liver

191

of IRF-1 / mice, suggesting that involvement of IRF-1 in MHC class II expression may be tissue specific [25]. While previous studies suggest that EAT is a Th1mediated disease [26,27], the relative role of Th1 and Th2 immune reaction remains to be elucidated. IRF-1 is involved in the development of Th1 immune response, and thus IRF-1deficient mice have strongly impaired Th1 reaction [12]. Consistently, we demonstrate a defective IFNg secretion by ConA-stimulated spleen cells in IRF-1 / mice. IL-10 production was not significantly different between IRF-1+/+ and / mice. Proliferation of ConA-stimulated splenocytes was not impaired, thus indicating that Th1 response is selectively suppressed in IRF-1 / NOD mice. Taken together, our results suggest that predominance of Th1 immune reaction may not be involved in EAT in NOD mice. Consistent with our observation, either IFNg or IFNg receptor-deficient mice do develop EAT [28,29]. Immunization of NOD mice with mTg resulted in enhanced production of anti-mTg Ab in IRF-1+/+, +/ , and / mice. As shown in iodine-treated NOD mice [8], there was no significant difference in the serum titers of anti-mTg Ab among the three IRF-1 genotypes. Thus, IRF-1 is not involved in anti-mTg Ab production in mTg-immunized NOD mice. In addition, our results implicate the pathogenic role of anti-mTgAb rather than that of IRF-1, CD8+ T cells, or Th1 immune reaction in the development of EAT. However, IFNg receptor-deficient mice develop EAT despite of their decreased anti-mTgAb levels [29]. Thus, as summarized previously [30], the role of antimTgAb in the development of EAT remains to be determined. It remains unclear why IRF-1 null mice develop EAT but not other autoimmune diseases such as autoimmune diabetes [14]. Previous studies suggest nitric oxide (NO) as a mediator of pancreatic h-cell destruction [31]. IRF-1 is involved in inducible NO synthase expression and thus NO production is reduced in IRF-1-deficient mice [32]. Taken together, NO may play a role in autoimmune diabetes but not in EAT. Little information is available for the role of NO in EAT and thus this issue is to be investigated. In addition, while NOD mice lacking IRF-1 are of value in elucidating the mechanism of EAT, new animal models should be established for better understanding of the pathogenesis of autoimmune thyroiditis. In conclusion, we demonstrate that IRF-1-deficient NOD mice that have defect of CD8+ T cell differentiation and Th1 immune reaction develop EAT. Our results suggest that IRF1 is not pivotal in the development of EAT and that the mechanism involved in EAT is quite different from that in iodine-induced LT in NOD mice.

Acknowledgment The authors thank Noriko Tsuchiya for preparing and staining thyroid sections of NOD mice. This study was

192

Z. Jin et al. / Clinical Immunology 113 (2004) 187–192

supported in part by Grant-in-Aid for Scientific Research (12671073 to K. Mori) from the Ministry of Education, Science and Culture of Japan.

[16]

[17]

References [18] [1] A.O. Vladutiu, N.R. Rose, Autoimmune murine thyroiditis: relation to histocompatibility (H-1) type, Science 174 (1971) 1137 – 1139. [2] H. Braley-Mullen, M. Johnson, G.C. Sharp, M. Kyriakos, Induction of experimental autoimmune thyroiditis in mice with in vitro activated splenic T cells, Cell. Immunol. 93 (1985) 132 – 143. [3] D.H. Conaway, A.A. Giraldo, C.S. David, Y.M. Kong, In situ kinetic analysis of thyroid lymphocyte infiltrate in mice developing experimental autoimmune thyroiditis, Clin. Immunol. Immunopathol. 53 (1989) 346 – 352. [4] Y.M. Kong, H. Waldmann, S. Cobbold, A.A. Giraldo, B.E. Fuller, L.L. Simon, Pathogenic mechanisms in murine autoimmune thyroiditis: short- and long-term effects of in vivo depletion of CD4+ and CD8+ cells, Clin. Exp. Immunol. 77 (1989) 428 – 433. [5] S.J. Stull, M. Kyriakos, G.C. Sharp, H. Braley-Mullen, Prevention and reversal of experimental autoimmune thyroiditis (EAT) in mice by administration of anti-L3T4 monoclonal antibody at different stages of disease development, Cell. Immunol. 117 (1988) 188 – 198. [6] L.C. Lomo, F. Zhang, D.J. McCormick, A.A. Giraldo, C.S. David, Y.M. Kong, Flexibility of the thyroititogenic T cell repertoire for murine autoimmune thyroiditis in CD8-deficient (b2m / ) and T cell receptor Vb c congenic mice, Autoimmunity 27 (1998) 127 – 133. [7] H. Braley-Mullen, R.W. McMurray, G.C. Sharp, M. Kyriakos, Regulation of the induction and resolution of granulomatous experimental autoimmune thyroiditis in mice by CD8+ T cells, Cell. Immunol. 153 (1994) 492 – 504. [8] J.-I. Tani, K. Mori, S. Hoshikawa, T. Nakazawa, J. Satoh, Y. Nakagawa, S. Ito, K. Yoshida, Prevention of lymphocytic thyroiditis in iodide-treated non-obese diabetic mice lacking interferon regulatory factor-1, Eur. J. Endocrinol. 147 (2002) 809 – 814. [9] T. Matsuyama, T. Kimura, M. Kitagawa, K. Pfeffer, T. Kawakami, N. Watanabe, T.M. Kundig, R. Amakawa, K. Kishihara, A. Wakeham, J. Potter, C.L. Furlonger, A. Narendran, H. Suzuki, P.S. Ohashi, C.J. Paige, T. Taniguchi, T.W. Mak, Targeted disruption of IRF-1 or IRF-2 results in abnormal type I IFN gene induction and aberrant lymphocyte development, Cell 75 (1993) 83 – 97. [10] T. Taniguchi, M.S. Lamphierm, N. Tanaka, IRF-1: the transcription factor linking the interferon response and oncogenesis, Biochim. Biophys. Acta 1333 (1997) M9 – M17. [11] J.M. Penninger, C. Sirard, H.-W. Mittrucker, A. Chidgey, I. Kozieradzki, M. Ngheim, A. Hakem, T. Kimura, E. Timms, R. Boyd, T. Taniguchi, T. Matsuyama, T.W. Mak, The interferon regulatory transcription factor IRF-1 controls positive and negative selection of CD8+ thymocytes, Immunity 7 (1997) 243 – 254. [12] M. Lohoff, D. Ferrick, H.-W. Mittrucker, G.S. Duncan, S. Bischof, M. Rollinghoff, T.W. Mak, Interferon regulatory factor-1 is required for a T helper 1 immune response in vivo, Immunity 6 (1997) 681 – 689. [13] Y. Tada, A. Ho, T. Matsuyama, T.W. Mak, Reduced incidence and severity of antigen-induced autoimmune disease in mice lacking interferon regulatory factor-1, J. Exp. Med. 185 (1997) 231 – 238. [14] T. Nakazawa, J. Satoh, K. Takahashi, Y. Sakata, F. Ikehata, Y. Takizawa, S.-I. Bando, T. Housai, Y. Li, C. Chen, T. Masuda, S. Kure, I. Kato, S. Takasawa, T. Taniguchi, H. Okamoto, T. Toyota, Complete suppression of insulitis and diabetes in NOD mice lacking interferon regulatory factor-1, J. Autoimmun. 17 (2001) 119 – 125. [15] D. Damotte, E. Colomb, C. Cailleau, N. Brousse, J. Charreire, C. Carnaud, Analysis of susceptibility of NOD mice to spontaneous and

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

experimentally induced thyroiditis, Eur. J. Immunol. 27 (1997) 2854 – 2862. Y.M. Kong, C.S. David, A.A. Giraldo, M. Elrehewy, N.R. Rose, Regulation of autoimmune response to mouse thyroglobulin: influence of H-2D-end genes, J. Immunol. 123 (1979) 15 – 18. G.F. Bottazzo, R. Pujol-Borrell, T. Hanafusa, M. Feldmann, Role of aberrant HLA-DR expression and antigen presentation in induction of endocrine autoimmunity, Lancet 2 (8359) (1983) 1115 – 1119. M. Hobart, V. Ramassar, N. Goes, J. Urmson, P.F. Halloran, IFN regulatory factor-1 plays a central role in the regulation of the expression of class I and II MHC genes in vivo, J. Immunol. 158 (1997) 4260 – 4269. J.C. Flynn, D.H. Conaway, S. Cobbold, H. Waldmann, Y.M. Kong, Depletion of L3T4+ and Lyt-2+ cells by rat monoclonal antibodies alters the development of adoptively transferred experimental autoimmune thyroiditis, Cell. Immunol. 122 (1989) 377 – 390. B.R. Champion, D.C. Rayner, P.G. Byfield, K.R. Page, C.T.J. Chan, I.M. Roitt, Critical role of iodination for T cell recognition of thyroglobulin in experimental murine thyroid autoimmunity, J. Immunol. 139 (1987) 3665 – 3670. S.A. Ebner, W. Lueprasitsakul, S. Alex, S.L. Fang, M.C. Appel, L.E. Braverman, Iodine content of rat thyroglobulin affects its antigenicity in inducing lymphocytic thyroiditis in the BB/Wor rat, Autoimmunity 13 (1992) 209 – 214. L. Rasooly, N.R. Rose, A.M. Saboori, P.W. Landenson, C.L. Burek, Iodine is essential for human T cell recognition of human thyroglobulin, Autoimmunity 27 (1998) 213 – 219. H. Margolese, C. O’Keefe, F. Chung, F. Ertug, J.R. Wall, N.F. Bernard, Expression of major histocompatibility complex class II antigen in NOD mouse thyroid. Autoimmunity 17 (1994) 1–11. R.E. Bonita, N.R. Rose, L. Rasooly, P. Caturegli, C.L. Burek, Kinetics of mononuclear cell infiltration and cytokine expression in iodineinduced thyroiditis in the NOD-H2h4 mouse, Exp. Mol. Pathol. 74 (2003) 1 – 12. M. Hobart, V. Ramassar, N. Goes, J. Urmson, P.F. Halloran, The induction of class I and II major histocompatibility complex by allogenic stimulation is dependent on the transcription factor interferon regulatory factor 1 (IRF-1), Transplantation 62 (1996) 1895 – 1901. H. Tang, K. Mignon-Godefroy, P.L. Meroni, G. Garotta, J. Charreire, F. Nicoletti, The effect of a monoclonal antibody to interferon-g on experimental autoimmune thyroiditis (EAT): prevention of disease and decrease of EAT-specific T cells, Eur. J. Immunol. 23 (1993) 275 – 278. P. Zaccone, P. Hutchings, F. Nicoletti, G. Penna, L. Adorini, A. Cooke, The involvement of IL-12 in murine experimentally induced autoimmune thyroiditis, Eur. J. Immunol. 29 (1999) 1933 – 1942. H. Tang, G.C. Sharp, K.P. Peterson, H. Braley-Mullen, IFN-gammadeficient mice develop severe granulomatous experimental autoimmune thyroiditis with eosinophil infiltration in thyroids, J. Immunol. 160 (1998) 5105 – 5112. E. Alimi, S. Huang, M.P. Brazillet, J. Charreire, Experimental autoimmune thyroiditis (EAT) in mice lacking the IFN-gamma receptor gene, Eur. J. Immunol. 28 (1998) 201 – 208. Y. Tomer, Anti-thyroglobulin antibodies in autoimmune thyroid diseases: cross-reactive or pathogenic? Clin. Immunol. Immunopathol. 82 (1997) 3 – 11. A. Rabinovitch, W.L. Suarez-Pinzon, O. Sorensen, R.C. Bleackley, Inducible nitric oxide synthase (iNOS) in pancreatic islets of nonobese diabetic mice: Identification of iNOS-expressing cells and relationships to cytokines expressed in the islets, Endocrinology 137 (1996) 2093 – 2099. R. Kamijo, H. Harada, T. Matsuyama, M. Bosland, J. Gerectiano, D. Shapiro, J. Le, S.I. Koh, T. Kimura, S.J. Green, T.W. Mak, T. Taniguchi, J. Vilcek, Requirement for transcription factor IRF-1 in NO synthase induction in macrophages, Science 263 (1994) 1612 – 1615.