Year in Thyroidology: Basic Science

Thyroid Gland

Hypothyroidism results from genetic, environmental, autoimmune, and surgical causes and affects 5% of the world population. Although hormone replacement is a well-established treatment, about a third of patients are inadequately treated, while poor quality-of-life and other comorbidities complicate sub-optimal therapy.

In recent years, methods to generate embryonic stem (ES) cell-derived organoids have been established to study thyroid gland development and model disease pathogenesis. Considerable progress has been made using murine cells, but generation of fully functional thyroid follicles from human ES cells has not been possible.

Romitti et al. reported a 58-day protocol for generation of human ES cell-derived thyroid organoids. ¹ Inducible over-expression of the critical transcription factors, NKX2-1 and PAX8, in embryoid bodies in vitro resulted in sequential cell commitment, expansion, differentiation, and organization. Resulting thyroid follicles expressed appropriate genes, formed organoids, and synthesized thyroid hormones.

Organoids were characterized by immunofluorescence and single cell RNA sequencing (scRNAseq), which were used to track cell differentiation and complexity until the production and storage of iodinated thyroglobulin and thyroxine were seen in the lumen of mature follicles. Organoid structure and function in vivo were confirmed by histology, immunofluorescence, gene expression, imaging (single-photon emission computed tomography [SPECT]), and measurement of thyroid hormones in peripheral blood of transplanted, ¹³¹I-ablated mice. ¹

These studies provide a model that reproduces development of human thyroid function, providing new opportunities to study thyroid morphogenesis, the pathogenesis of congenital hypothyroidism, and toxicity of drugs and endocrine disruptors. Ultimately, this work could help realize the potential of autologous human thyroid tissue for the treatment of hypothyroidism.

Iodide Uptake

The sodium/iodide symporter (NIS) mediates active transport of iodide as the first step in hormone synthesis. NIS couples transport of I⁻ into thyroid follicular cells against an electrochemical gradient along with the inward movement of Na⁺ along its gradient. NIS is crucial for physiological thyroid function and patient care. The symporter enables the treatment of thyroid cancer with ¹³¹I. It translocates ^99mTcO₄ ⁻, ¹⁸⁶ReO₄ ⁻, and ¹⁸⁸ReO₄ ⁻ for use in SPECT imaging, ¹⁸F-BF₄ ⁻ and ¹²⁴I⁻ in positron emission tomography, and transports pollutants such as ClO₄. Mutations in NIS cause congenital hypothyroidism.

Ravera et al. reported the structure and configuration of NIS at 3.46 Å resolution using single particle cryo-electron microscopy (CryoEM). ² They identified the I⁻ and Na⁺ binding pockets and showed a 2:1 Na⁺ to I⁻ binding stoichiometry at physiological concentrations, whereas Na⁺ and ReO₄ ⁻ bound with 1:1 stoichiometry. The authors proposed a mechanism for NIS transport in which two Na⁺ ions bind first and increase the affinity of I⁻ for its hydrophobic pocket by 10-fold.

Structural analysis of Na⁺ and I⁻ bound intermediates indicated that the NIS protein undergoes isomerization, resulting in transport of cargo through an inwardly facing hydrophilic tunnel that connects the substrate binding pockets to the cytosol. After intra-cellular release of cargo, isomerization of NIS is reversed, and the transport cycle continues. ²

These studies deepen fundamental understanding of the mechanism of NIS function and will facilitate the development of NIS variants with different substrate selectivity. Such approaches will enable the development of novel reporters for preclinical in vivo imaging, provide new strategies for cancer therapeutics using novel radioactive substrates, and allow investigation of mechanisms by which pollutants and endocrine disruptors enter thyroid follicular cells.

Thyroid Autoimmunity

Autoimmune toxicity occurs in 60% of oncology patients treated with immune checkpoint inhibitors and about 30% develop thyroiditis. Despite the increasing prevalence of checkpoint inhibitor thyroiditis, mechanisms of disease pathogenesis are unknown and have only been studied in peripheral blood cells.

Lechner et al. reported analysis of thyroid aspirates from patients with checkpoint inhibitor or Hashimoto's thyroiditis compared with normal thyroid tissue. ³ Intrathyroidal immune cells were characterized by scRNAseq after fluorescence-activated cell sorting and CD45⁺ selection of immune cells. Infiltrating follicular and peripheral T helper cells (T_FH, T_PH) were identified.

These cells drive thyroid autoimmunity, recruit B and T cells via cytokine CXCL13, and produce interleukin (IL)-21 to promote B cell antibody production. Similar populations of IL-21⁺ CD4⁺ T_H cells were identified in both checkpoint inhibitor and Hashimoto's thyroiditis. By contrast, clonal expansion and accumulation of CXCR6, GZMB, and IFNG expressing CD8⁺ effector T cells occurred only in checkpoint inhibitor thyroiditis. Pseudotime analysis of scRNAseq data revealed the progression of CD8⁺ T cells along a lineage from progenitors to cytotoxic effectors that drive cell-mediated immunity, cytotoxicity, and T-cell homing to inflamed tissues. ³

In vitro studies demonstrated that CD4⁺ T_FH and T_PH cells secrete IL-21 and drive cytotoxic effector CD8⁺ T cells by stimulating expression of the chemokine receptor CXCR6, IFN-γ, and the serine protease cytotoxic mediator GZMB. In vivo studies revealed that CD8⁺ T cell activation and induction of thyroiditis by checkpoint inhibitors were prevented in IL-21 receptor knockout mice, while adoptive transfer of CD4⁺ and CD8⁺ T cells from IL-21 receptor knockout mice to wild-type mice markedly decreased induction of thyroiditis. ³

These elegant studies identify IL-21 as a key driver of autoimmune checkpoint inhibitor thyroiditis. They pave the way for the development of new drugs that block IL-21 signaling, and the identification of common pathways that may be shared in the pathogenesis of autoimmunity among different tissues.

Thyroid Cancer

Anaplastic thyroid cancer has a 5-year survival of 7% and accounts for 40% of all thyroid cancer mortality. The rarity and short survival time of anaplastic disease result in limited treatment options and poor understanding of its molecular pathogenesis.

Lu et al. reported analysis of tissue from patients with papillary and anaplastic cancer in comparison with normal thyroid. ⁴ scRNAseq and DNA analysis were used to annotate cells and characterize lineage progression, fate transition, and genomic perturbations. Bioinformatics and machine learning identified differentially expressed and enriched genes, transcription factor activation and cell–cell interaction networks, DNA copy number, mutations, and single nucleotide variants.

During carcinogenesis, thyrocytes first acquire a papillary thyroid cancer (PTC) phenotype. The PTC cells remain indolent or transform to an inflammatory anaplastic phenotype (iATC) in response to stress signals. The iATC cells acquire a mesenchymal phenotype (mATC), characterized by single nucleotide and/or copy number variations and aneuploidy, with defective mitosis. Subsequently, mATC cells reprogram, proliferate, and transform the tumor microenvironment by interacting with macrophages and fibroblasts. Anti-inflammatory, tumor-promoting M2 macrophages and inflammatory cancer-associated fibroblasts proliferate, while pro-inflammatory anti-tumor M1 macrophages and myogenic cancer-associated fibroblasts decrease. ⁴

These studies provide comprehensive understanding of events that underpin the evolution of anaplastic thyroid cancer from an indolent, differentiated state to the formation of aggressive, dedifferentiated tumors. These insights will provide new opportunities to evaluate the risk of anaplastic progression in patients with differentiated thyroid cancer, and the prospect of novel drug targets to prevent anaplastic transformation.

Three additional articles in the field of thyroid cancer were discussed. Nappi et al. studied the role of the type 2 deiodinase (DIO2) in the progression of solid tumors in general. ⁵ They found that loss of the p53 tumor suppressor activates DIO2/T3 signaling and favors genomic instability and metastasis. Thus, targeting DIO2/T3 signaling may have future applications to reduce invasiveness of p53-mutated neoplasms. Veschi et al. reported a new in vitro model of differentiated thyroid cancer transformation using CRISPR-Cas9 edited human ES cells. ⁶

These studies recapitulated in vitro the oncogenic events in tissue progenitor cells, advanced understanding of the molecular pathogenesis of differentiated thyroid cancer transformation and provide opportunities to identify novel therapeutic targets. In this context, Fagin et al. reviewed current understanding of the pathogenesis of follicular thyroid cancer. ⁷ The comprehensive discussion of major signaling pathways and how their activities determine tumor phenotype provides a detailed framework that will inform and support new approaches for drug discovery.

Thyroid Hormone Action

The spatial organization and transcriptional diversity of cone photoreceptors in the retina determine the subtlety of color vision. Opsin photopigments respond to different regions of the light spectrum. M cones in the superior retina of the mouse are activated by medium and long wavelength light, while S cones in the inferior retina react to short wavelengths.

Aramaki et al. reported divergent differentiation of cone precursors and post-natal emergence of groups of genes expressed in regional gradients within distinct cone subtypes. ⁸ The thyroid hormone receptor TRβ2 isoform facilitated spatial organization of cone photoreceptors and controlled expression of gene gradients. Analysis of chromatin remodeling sites and DNA binding of TRβ2 in triiodothyronine (T3) regulated genes in purified cone populations was performed by integrating data from the assay for transposase-accessible chromatin with sequencing (ATACseq) with findings from chromatin affinity purification sequencing (ChAPseq).

T3-responsive genes were enriched for open chromatin regions that coincided with TRβ2 DNA binding sites in positively regulated target genes. The direct repeat +4 (DR) response element was the most highly enriched binding site for TRβ2. By contrast, specific TRβ2 binding sites were not enriched in genes whose expression was suppressed by T3.

Thus, mechanisms underlying negative regulation of gene expression by T3 remain elusive despite these elegant genome-wide approaches. scRNAseq and immunohistochemical analysis of gradient genes revealed that cones from TRβ2 knockout mice have diminished gene expression gradient bias in both superior and inferior regions. ⁸

Overall, the studies demonstrate that TRβ2 promotes spatial and functional cone diversity and the full potential for color vision by activating specific transcriptome signatures in superior and inferior cones during post-natal organ maturation, including at the time of eye opening when circulating thyroxine (T4) and T3 levels rise. These sophisticated studies used new in vivo models and integrated state-of-the-art methods to advance the field. They provide robust experimental approaches that can be adapted more generally for investigation of other T3 target tissues and cell types.

Six further articles in the field of thyroid hormone action were discussed. Vamesu et al. ⁹ and Wang et al. ¹⁰ investigated T3 action in the neonatal and adult lung. They showed that thyroid hormones promote alveolar regeneration in neonates with hypoxic lung damage and exert preventative actions in pulmonary fibrosis in adults. The studies raise the possibility of therapeutic utility for intranasal T3 in fibrotic lung disease.

Two novel approaches identified the importance of biological rhythmicity in physiological responses to thyroid hormone. de Assis et al. showed that metabolic responses to T3 in the liver differ according to time of day, ¹¹ thus identifying the impact of diurnal rhythm, while Wei et al. showed that adaptation of the retina to light versus dark cues is regulated by DIO2/T3 signaling. ¹²

Salas-Lucia et al. identified a novel pathway for T3 uptake by endosomal lysosomes in axons. ¹³ This pathway bypasses hormone degradation by type 3 deiodinase and may represent a mechanism to preserve axonal function in conditions when DIO3 activity is high.

Yan et al. demonstrated that regulation of glucose homeostasis by thyroid hormones involves crosstalk between the liver and intestine. ¹⁴ They demonstrated that T3 suppresses farnesoid X receptor (FXR) signaling in the intestine, thus potentiating GLP-1 production and insulin secretion. This is achieved by T3-mediated inhibition of Cyp8b1 (sterol 12-alpha-hydroxylase) expression in hepatocytes, resulting in increased levels of FXR antagonist bile acids.

Finally, Dore et al. studied resistance to thyroid hormone α (RTHα). ¹⁵ Supra-physiological doses of thyroid hormones have been used to treat RTHα patients with delayed skeletal development and growth retardation, but theoretical concerns about cardiac toxicity have been raised. In this study, high-dose T4 treatment of mice with RTHα did not result in adverse cardiac events. Additional studies in humans were consistent, showing that supra-physiological doses of thyroid hormone did not induce tachycardia.

Future Applications

This last year has been exciting. State-of-the-art and cross disciplinary approaches were applied to study all aspects of thyroid research. Methods included: manipulation of human ES cells, optimization of thyroid organoid culture systems; CryoEM; use of bespoke animal models generated by homologous recombination and CRISPR/Cas9 mutagenesis; adoptive cell transfer; organoid transplantation; in vivo imaging; multi-omics such as scRNAseq, ATACseq, and ChAPseq; integrated bioinformatic analysis and machine learning.

These innovative approaches have provided new reagents for the thyroid community, including functional human thyroid organoids and new animal models that can be used to dissect mechanisms of thyroid-related physiology and disease, as well as extensive new multi-omic datasets and data analysis tools with open access availability for all researchers.

This infrastructure has identified promising new areas for thyroid-related research in the fields of pulmonary fibrosis, integrated glucose homeostasis, mechanisms of cellular uptake of T3, and the rhythmic control of T3-regulated gene transcription. Ultimately such advances may result in novel targeted therapies based on newly identified molecular mechanisms of disease pathogenesis, for example, in anaplastic thyroid cancer and checkpoint inhibitor thyroiditis. The diversity of basic thyroid research never ceases to amaze, and the future appears bright as the level of basic understanding continues to expand, often in unexpected directions.