Abstract
Caturegli and Kimura have recently proposed a nonclassical model of autoimmune hypothyroidism (1) in which proinflammatory cytokines (such as interferon [IFN]-γ and tumor necrosis factor) inhibit thyroid function, leading to reversible hypothyroidism. This model differs from the classical one involving CD4+ and CD8+ T cell-mediated apoptosis of thyroid cells through Fas/FasL and cytotoxic proteins (perforin, etc.), as well as another possible classical mechanism, namely, the contribution of autoantibodies to thyroid peroxidase and thyroglobulin to thyroid damage via an antibody-dependent mechanism. The proposed model of Caturegli and Kimura is based on their previous study on transgenic mice with IFN-γ targeted to the thyroid (2). I have reservations about this hypothesis, for several reasons.
First, the concept of cytokine-induced hypothyroidism is not novel. More than 20 years ago (1987), we reported in vitro data showing that IFN-γ inhibited thyroid hormone synthesis in monolayer cultures of human thyroid cells (3). Our subsequent studies published in the late 1980s and early 1990s also demonstrated suppression by various cytokines (interleukin [IL]-1, IL-6, and IFN-γ) of thyroid-specific gene expression, including the thyrotropin receptor, thyroid peroxidase, and thyroglobulin (4 –8), indicating cytokine-mediated inhibition of thyroid function. In addition, a series of articles by Sato and colleagues (9 –11) reported inhibition of iodine incorporation, iodine organification, and thyroid hormone synthesis by various cytokines (IL-1, IL-2, tumor necrosis factor-α, IFN-α, and IFN-γ) in human thyroid cell suspension cultures. Even before publication of the above-mentioned articles, in the 1980s, clinicians and researchers became aware of hypothyroidism occurring in patients treated with type I IFNs (12,13). Presently, IFN-α induction of transient (but sometimes permanent) hypothyroidism in humans is well established (14). Thus, the idea of cytokine induction of hypothyroidism is not novel even if this hypothesis is true and the mechanisms involved are not fully understood.
Second, although thyroid cells themselves can produce some cytokines, the major source of the latter must be lymphocytes infiltrating into the thyroid glands. Indeed, it is well recognized that IFN-α enhances (or possibly induces) autoantibodies to thyroid peroxidase and/or thyroglobulin (14), which could contribute directly to thyroid damage or play a critical role in antigen presentation to T cells. Thus, the classical and nonclassical models very likely coexist and are not mutually exclusive.
Third, IFN-γ is a multifunctional cytokine (15) whose effects may vary in different autoimmune thyroiditis models (16). Such differences have been reported with neutralization, transgenic overexpression, and genetic ablation of IFN-γ, as well as of the IFN-γ receptor. The effects of IFN-γ may also be dependent on its concentration, timing, and site of production. In particular, in the authors' transgenic mice with thyroglobulin promoter-driven IFN-γ production in the thyroid (2), the thyroid cells must be generating large quantities IFN-γ from the neonatal period, a situation highly artificial and quite different from human thyroid autoimmunity and in mouse models of thyroiditis. Thus, no firm conclusion can be drawn from a single article with regard to the role played by IFN-γ in autoimmune thyroid disease.
Fourth, Caturegli and Kimura overlook another scenario proposed by Stassi et al. in 2000 (17) in which the thyroid cells from Graves' and Hashimoto's diseases both express Fas and FasL, and Fas/FasL-mediated apoptosis of the thyroid cells is promoted by the Th1 cytokine IFN-γ in Hashimoto's thyroiditis, but suppressed by the Th2 cytokines IL-4 and IL-10 in Graves' disease.
In summary, the proposal that cytokines induce hypothyroidism in the pathogenesis of Hashimoto's thyroiditis is not novel. Moreover, validation of this concept in spontaneous autoimmune thyroiditis in humans will require careful studies in the future.