Regulatory T cells in the control of autoimmunity: the essential role of transforming growth factor beta and interleukin 4 in the prevention of autoimmune thyroiditis in rats by peripheral CD4(+)CD45RC- cells and CD4(+)CD8(-) thymocytes

Abstract

Previous studies have shown that induction of autoimmune diabetes by adult thymectomy and split dose irradiation of PVG.RT1(u) rats can be prevented by their reconstitution with peripheral CD4(+)CD45RC-TCR-alpha/beta+RT6(+) cells and CD4(+)CD8(-) thymocytes from normal syngeneic donors. These data provide evidence for the role of regulatory T cells in the prevention of a tissue-specific autoimmune disease but the mode of action of these cells has not been reported previously. In this study, autoimmune thyroiditis was induced in PVG.RT1(c) rats using a similar protocol of thymectomy and irradiation. Although a cell-mediated mechanism has been implicated in the pathogenesis of diabetes in PVG.RT1(u) rats, development of thyroiditis is independent of CD8(+) T cells and is characterized by high titers of immunoglobulin (Ig)G1 antithyroglobulin antibodies, indicating a major humoral component in the pathogenesis of disease. As with autoimmune diabetes in PVG. RT1(u) rats, development of thyroiditis was prevented by the transfer of CD4(+)CD45RC- and CD4(+)CD8(-) thymocytes from normal donors but not by CD4(+)CD45RC+ peripheral T cells. We now show that transforming growth factor (TGF)-beta and interleukin (IL)-4 both play essential roles in the mechanism of this protection since administration of monoclonal antibodies that block the biological activity of either of these cytokines abrogates the protective effect of the donor cells in the recipient rats. The prevention of both diabetes and thyroiditis by CD4(+)CD45RC- peripheral cells and CD4(+)CD8(-) thymocytes therefore does not support the view that the mechanism of regulation involves a switch from a T helper cell type 1 (Th1) to a Th2-like response, but rather relies upon a specific suppression of the autoimmune responses involving TGF-beta and IL-4. The observation that the same two cytokines were implicated in the protective mechanism, whether thymocytes or peripheral cells were used to prevent autoimmunity, strongly suggests that the regulatory cells from both sources act in the same way and that the thymocytes are programmed in the periphery for their protective role. The implications of this result with respect to immunological homeostasis are discussed.

Figure 1

Thyroiditis in PVG rats after their thymectomy and irradiation develops independently of CD8⁺ cells and is characterized by a Th2-like anti-Tg IgG response. Female PVG rats were thymectomized at 3 wk of age followed 1 wk later, by four doses of 275 rad ¹³⁷Cs γ-irradiation at 2-wk intervals. (A) The isotype of anti-Tg IgG antibodies in sera of TxX rats with thyroiditis (n = 25) and normal 12-wk-old female PVG rats immunized with Tg (50 μg/rat) in CFA (n = 5) was determined by specific ELISA. Data are expressed as the mean percentage of the anti-Tg response for each IgG isotype where 100% is the sum of the ODs for individual isotypes above background of normal PVG sera in 1:10 dilutions of an experimental serum. Error bars indicate SD. (B) The requirement for CD8⁺ cells in the development of thyroiditis in TxX PVG rats was determined by their injection at the time of thymectomy and 7 d later with either the anti-CD8-depleting mAb OX8 (0.5 mg/injection) or PBS as control. Development of anti-Tg IgG responses was monitored between 4 and 12 wk after the last irradiation by specific ELISA. Data represent peak anti-Tg IgG titers expressed as percentage of standard for individual TxX rats. FACS^® analysis of splenocytes from OX8-treated rats, 12 wk after the last irradiation, showed that <2% of TCR⁺ cells were CD8⁺ (data not shown). Data are representative of two independent experiments.

Figure 1

Thyroiditis in PVG rats after their thymectomy and irradiation develops independently of CD8⁺ cells and is characterized by a Th2-like anti-Tg IgG response. Female PVG rats were thymectomized at 3 wk of age followed 1 wk later, by four doses of 275 rad ¹³⁷Cs γ-irradiation at 2-wk intervals. (A) The isotype of anti-Tg IgG antibodies in sera of TxX rats with thyroiditis (n = 25) and normal 12-wk-old female PVG rats immunized with Tg (50 μg/rat) in CFA (n = 5) was determined by specific ELISA. Data are expressed as the mean percentage of the anti-Tg response for each IgG isotype where 100% is the sum of the ODs for individual isotypes above background of normal PVG sera in 1:10 dilutions of an experimental serum. Error bars indicate SD. (B) The requirement for CD8⁺ cells in the development of thyroiditis in TxX PVG rats was determined by their injection at the time of thymectomy and 7 d later with either the anti-CD8-depleting mAb OX8 (0.5 mg/injection) or PBS as control. Development of anti-Tg IgG responses was monitored between 4 and 12 wk after the last irradiation by specific ELISA. Data represent peak anti-Tg IgG titers expressed as percentage of standard for individual TxX rats. FACS^® analysis of splenocytes from OX8-treated rats, 12 wk after the last irradiation, showed that <2% of TCR⁺ cells were CD8⁺ (data not shown). Data are representative of two independent experiments.

Figure 2

Prevention of thyroiditis development in TxX PVG rats by their reconstitution with either CD4⁺CD45RC⁻ T cells or CD4⁺CD8⁻ thymocytes. Unfractionated CD4⁺ and CD4⁺CD45RC⁻ cells were purified from TDLs of 12-wk-old normal PVG rats by Dynal bead depletion. CD4⁺CD45RC⁺ cells were purified from CD4⁺ TDLs by positive selection using MACS beads, whereas CD4⁺CD8⁻ thymocytes were purified from thymus of 6-wk-old rats by depletion of CD8⁺ cells as described in Materials and Methods. Shortly after their last irradiation, TxX PVG rats were reconstituted with 10⁷ of unfractionated CD4⁺ cells, CD4⁺ CD45RC⁻ cells, CD4⁺CD45RC⁺ cells, or CD4⁺CD8⁻ thymocytes, while control rats received no cells. Data represent peak anti-Tg IgG antibody levels in individual TxX rats from groups reconstituted with different T cell subsets. Anti-Tg titers are expressed as percentage of a standard and data are the pool of six independent experiments. Statistics versus controls: ^a P < 10⁻⁴; ^b P < 1.2 × 10⁻⁸; ^c P > 0.52; ^d P < 3 × 10⁻⁶.

Figure 3

Immunopathology of the thyroid glands from control TxX rats and those reconstituted with CD4⁺CD45RC⁻ T cells. In experiments similar to those described in Fig. 2, thyroid glands were taken at the time of peak disease, sectioned, and stained with hematoxylin and eosin. Thyroids from control TxX PVG rats (A; original magnification ×200) show extensive mononuclear cell infiltrate and loss of follicular structure. In contrast, thyroid glands from TxX PVG rats reconstituted with 10⁷ CD4⁺CD45RC⁻ cells shortly after their last irradiation are of normal morphology with no signs of infiltration (B; original magnification ×200). Similarly, thyroids of TxX PVG rats protected by their reconstitution with CD4⁺CD8⁻ thymocytes were of normal morphology with no signs of infiltration (data not shown).

Figure 3

Immunopathology of the thyroid glands from control TxX rats and those reconstituted with CD4⁺CD45RC⁻ T cells. In experiments similar to those described in Fig. 2, thyroid glands were taken at the time of peak disease, sectioned, and stained with hematoxylin and eosin. Thyroids from control TxX PVG rats (A; original magnification ×200) show extensive mononuclear cell infiltrate and loss of follicular structure. In contrast, thyroid glands from TxX PVG rats reconstituted with 10⁷ CD4⁺CD45RC⁻ cells shortly after their last irradiation are of normal morphology with no signs of infiltration (B; original magnification ×200). Similarly, thyroids of TxX PVG rats protected by their reconstitution with CD4⁺CD8⁻ thymocytes were of normal morphology with no signs of infiltration (data not shown).

Figure 4

CD4⁺CD8⁻ thymocytes are more potent than CD4⁺ CD45RC⁻ cells at preventing development of thyroiditis in TxX PVG rats. PVG rats were thymectomized at 3 wk of age and given four equal doses of 275-rad ¹³⁷Cs γ-irradiation at 2-wk intervals starting 1 wk after thymectomy. CD4⁺CD8⁻ thymocytes were purified from thymus of 6-wk-old PVG rats by depletion of CD8⁺ cells, and CD4⁺CD45RC⁻ cells were purified from TDLs of 12-wk-old PVG rats by depletion of CD8⁺ and CD45RC⁺ cells. Shortly after their last irradiation, groups of rats were reconstituted either with CD4⁺CD8⁻ thymocytes at doses of 10⁵, 5 × 10⁵, 10⁶, 5 × 10⁶, or 10⁷ per rat, or with CD4⁺CD45RC⁻ cells at doses of 10⁶, 5 × 10⁶, or 10⁷ per rat. Development of anti-Tg antibodies was determined between 4 and 12 wk after the final irradiation by specific ELISA. Data are expressed as percentage of rats protected from development of disease, where the incidence of protection in the control group (42 out of 125 control rats failed to develop disease) is considered 0% protection. NP, no protection. Statistics versus controls: ^a P > 0.11; ^b P < 0.0046; ^c P < 1.6 × 10⁻⁷; ^d P < 2.3 × 10⁻⁶; ^e P < 1.8 × 10⁻⁵; ^f P > 0.4; ^g P < 8 × 10⁻⁹; ^h P < 7 × 10⁻⁸.

Figure 5

Treatment of TxX rats with neutralizing mAbs against TGF-β or IL-4 abrogates protection from thyroiditis by both CD4⁺CD45RC⁻ cells and CD4⁺CD8⁻ thymocytes. Shortly after their last irradiation, groups of TxX PVG rats were injected with either 5 × 10⁶ CD4⁺CD45RC⁻ cells purified from TDLs of 12-wk-old normal PVG rats or 10⁶ CD4⁺CD8⁻ thymocytes purified from thymus of 6-wk-old normal PVG rats. Control rats received no cells. Rats reconstituted with cells were treated the day before and the day after cell transfer and then twice weekly for 4 wk with either the anti–TGF-β–neutralizing mAb 2G.7 (mouse IgG2a; 2 mg/injection) or the anti–IL-4–neutralizing mAb OX81 (mouse IgG1; 2 mg/injection). Control rats reconstituted with cells either received no further treatment or were injected with one of the isotype-control mAbs OX21 (mouse IgG1; 2 mg/injection) or W6/32 (mouse IgG2a; 2 mg/injection) specific for human determinants. Data represent peak anti-Tg antibody levels of individual TxX rats reconstituted with either CD4⁺CD45RC⁻ cells (A) or CD4⁺CD8⁻ thymocytes (B). Anti-Tg titers are expressed as percentage of standard and data are the pool of five independent experiments. Statistics versus controls: ^a P < 2.5 × 10⁻⁴; ^b P < 3.1 × 10⁻⁴; ^c P > 0.39; ^d P > 0.07; ^e P < 0.0021; ^f P < 0.008; ^g P > 0.51; ^h P > 0.46; versus reconstituted rats treated with control mAb (b and f): ^c  P < 0.016; ^d P < 2 × 10⁻⁵; ^g P < 0.013; ^h P < 0.005.

Figure 6

IgG isotypes of anti-Tg antibodies in TxX rats is not affected by their treatment with anti–IL-4 or anti–TGF-β blocking mAbs. The relative isotype usage of anti-Tg IgG responses was determined for TxX rats reconstituted with CD4⁺CD45RC⁻ cells but treated with either anti–IL-4 (n = 9) or anti–TGF-β (n = 7) blocking mAbs and those reconstituted with CD4⁺CD8⁻ thymocytes and treated with either anti-IL-4 (n = 6) or anti-TGF-β (n = 9) blocking mAbs from the experiments described in Fig. 5. The isotype of anti-Tg IgG in sera of control rats (n = 16) from both these series of experiments was similarly determined by specific ELISA. Data are expressed as the mean percentage of the anti-Tg response for a given IgG isotype where 100% is the sum of the ODs for individual isotypes above background of normal PVG sera in 1:10 dilutions of an experimental serum. Error bars indicate SD.