Abstract
We thank Dr. Yuji Nagayama for having carefully considered our commentary (1) about the existence of a nonclassical model of autoimmune hypothyroidism. Our reply to his four criticisms is the following.
First, the concept of cytokine-induced hypothyroidism is not novel. We are aware of the effects interferon-gamma (IFNγ) and other proinflammatory cytokines exert on cultured thyroid cells. These in vitro effects show the impact of a cytokine on a particular aspect of thyroid hormonogenesis, like iodide uptake or transcription of thyroid-restricted genes, but cannot inform us on the nature of a much more complex and chronic condition like autoimmune hypothyroidism. We were the first to show in vivo the effects of chronic exposure of the thyroid gland to a cytokine. The novelty of this mouse model is not that a cytokine can induce hypothyroidism, but rather that hypothyroidism can occur without lymphocytes. This is the seminal observation that formed the basis for developing the nonclassical model of autoimmune hypothyroidism presented in the commentary.
We agree that our IFNγ transgenic model is “highly artificial and quite different” from what observed in patients. But aren't all models artificial? They rarely replicate fully or perfectly the human disease they are designed to mimic. The point is not whether a mouse model is artificial but whether it can teach us something about the pathogenesis of the human counterpart.
Second, lymphocytes infiltrating the thyroid gland are the major source of cytokines. Having studied autoimmune diseases for over two decades, we support the view that lymphocytes are the dominant cell type in autoimmune pathology and important cytokine producers. It is on the background of this notion that we compared the results obtained from our IFNγ transgenic mice. Those results showed us that a pathological process is possible without infiltrating lymphocytes, raising new questions about their role. Different infiltrating cells (e.g., macrophages) might turn out to be more important than lymphocytes in autoimmune pathogenesis. Findings from other autoimmune diseases like Sjögren syndrome reveal that the target organ is not a battleground for lymphocytes. Sjögren patients who have a severe clinical phenotype (minimal or absent salivary gland function) might show only a few infiltrating lymphocytes and numerous normal-appearing acini upon lower lip biopsy. It is difficult to argue in this setting that lymphocytes or their cytokines are the key disease mediators.
Third, IFNγ is a multifunctional cytokine whose effects may vary in different autoimmune thyroiditis models. We agree completely with this point and refer the reader to the remark made above about the faithfulness and utility of animal models.
Fourth, the authors overlook another scenario proposed by Stassiet al. (2). We have not overlooked this excellent article by our Italian countrymen. The article focuses on apoptosis, and thus we consider it part of the classical (apoptotic) model of autoimmune hypothyroidism.
Dr. Nagayama summarized our commentary (1) by writing, “The proposal that cytokines induce hypothyroidism in the pathogenesis of Hashimoto thyroiditis is not novel.” This is not the message we had hoped the readers derive. If we were to express in one sentence our message, it would be, “Hypothyroidism can occur independently of infiltrating lymphocytes and apoptotic destruction of thyrocytes.”
Overall, we thank Dr. Nagayama for his interest in our commentary (1) and Dr. Emerson for giving us the opportunity to reply. We reflected on a fascinating and still mysterious condition like autoimmune hypothyroidism, and proposed a different scenario for disease pathogenesis. This new scenario can have clinical ramifications if we consider that a model based on apoptotic destruction of thyrocytes is irreversible, whereas a model based on chronic inhibition is potentially reversible and curable. Biology is rarely an either/or proposition, and as such we concur with Dr. Nagayama when he writes, “Classical and non-classical models very likely co-exist and are not mutually exclusive.”