S14.1 - Tissue specific thyroid hormone signalling

Abstract

The human thyroid gland primarily synthesizes thyroxine (T4), a metabolically stable prohormone requiring mostly extrathyroidal peripheral deiodination to T3 for cellular activity. The hypothalamic-pituitary-thyroid (HPT) axis maintains a relatively stable circulating thyroid hormone (TH) level. However, TH signaling within the brain exhibits rapid, cell-type specific modulation, demonstrating a capacity for local regulation largely independent of circulating TH concentrations. T3 production in the brain is largely dependent on the type 2 deiodinase (D2), a selenoenzyme predominantly expressed in glial cells—specifically astrocytes in extrahypothalamic regions and tanycytes within the hypothalamus. Globally, hypothyroidism affects hundreds of millions of individuals, and current standard-of-care treatment predominantly utilizes L-T4 monotherapy. Although this approach effectively normalizes thyroid-stimulating hormone (TSH) levels, a significant proportion (~10%-20%) of patients experience persistent residual symptoms, notably cognitive dysfunction often described as “brain fog”, suggesting tissue hypothyroidism and substantially impacting quality of life. Utilizing our Thyroid Hormone Action Indicator (THAI) mouse model, we have recently elucidated a molecular mechanism underlying the incomplete restoration of T3-dependent brain functions despite normalized TSH following L-T4 treatment. This mechanism involves glial subtype-specific differential regulation of D2 ubiquitination. Specifically, while tanycytic D2 in the mediobasal hypothalamus with direct impact on HPT function remains resistant to T4-mediated, ubiquitination-dependent inhibition and maintains sufficient T3 production for TSH normalization, D2 activity in extrahypothalamic astrocytes throughout the brain parenchyma is impaired by excess T4. This highlights the potential need for individualized assessment of tissue-level TH action in select patient populations. Our newly developed, minimally invasive methodology for evaluating human tissue TH action will be also discussed, that incorporates novel biomarkers and a Machine Learning-based predictive system.