RTHα, a Newly Recognized Phenotype of the Resistance to Thyroid Hormone (RTH) Syndrome in Patients With THRA Gene Mutations

The article by van Mullem et al in this issue of the JCEM (1) provides additional information about the clinical characteristics of 2 of the first 3 individuals recently identified to have resistance to thyroid hormone (RTH) due to a mutation in the thyroid hormone receptor α (TRα) 1 (2, 3). In this editorial discussing this newly recognized phenotype, we will use the term RTHα to distinguish it from RTH due to mutations in the thyroid hormone receptor β isoform (RTHβ).

A brief review of thyroid hormone action is helpful to understand the pathophysiology of these 2 RTH phenotypes. Thyroid hormone is essential for normal growth and development. The vast majority of its effects require its interaction with 1 of 2 nuclear thyroid hormone receptor proteins encoded by the thyroid hormone receptor α (THRA) and β (THRB) genes (4). The major physiologically active receptors are TRα1 and TRβ1 and -2. In the nucleus, the active thyroid hormone, 3,5,3′ triiodothyronine (T₃), binds tightly to these highly homologous DNA-binding proteins, either increasing or decreasing the transcriptional expression of thyroid hormone-responsive genes. These genes contain 1 or more short DNA sequences termed thyroid hormone response elements that bind TRα and TRβ with high affinity. When T₃ binds to TRs on positively regulated genes, corepressor proteins are released and replaced by coactivators leading to increased gene transcription. The negative regulation of genes by thyroid hormone is less well understood, and several mechanistic explanations are still under consideration to explain how T₃-receptor-DNA complexes repress transcription of genes such as the TSH β-subunit (5, 6).

Targeted homologous recombination technology in mice has allowed models with complete deletions or inactivating mutations of both receptors. These and subsequent clinical studies show that tissues express predominantly TRα or TRβ, explaining the physiological manifestations associated with a specific receptor defect. TRβ predominates in the hypothalamic-pituitary-thyroid axis, liver, and kidney, whereas TRα transduces much of the thyroid hormone signaling in the heart, skeletal muscle, central nervous system, bone, and intestines (4). Because mutations of TRβ block the feedback inhibition of TSH secretion, patients with RTHβ develop clinical and biochemical thyrotoxicosis of varying severity, accompanied by an inappropriately normal or slightly elevated TSH (7). This rare but clinically recognizable syndrome can be distinguished from pituitary tumor TSH-mediated thyrotoxicosis by appropriate tests and confirmed by sequencing the THRB gene. Tachycardia and other manifestations of excess T₄ occur in RTHβ because many tissues are TRα dominant and respond to elevated thyroid hormone. Interestingly, only a single copy of a mutated THRB gene is required to cause resistance because the unoccupied TR-containing complex appears to bind more tightly to the thyroid hormone response elements than does the wild-type, T₃-occupied receptor (4). Once a receptor mutation occurs in an individual, it is inherited in a dominant pattern in virtually all cases. Over 300 families with RTHβ have been identified, and its incidence is estimated at 1 in 40 000 live births (7).

Remarkably, considering both the numerous tissues in which TRα is dominant and the development of several experimental mouse models with mutant THRA genes, it was not until 2012 that the first patients with RTHα were recognized (8). As exemplified by the present report, the clinical features are due to dysfunction in TRα-regulated tissues, including short stature, impaired bone maturation and intellectual development, bradycardia, and severe constipation. In addition, in all 3 patients, there is an unusual thyroid hormone profile of low normal serum T₄, high normal serum T₃, and normal TSH. This aspect of the phenotype was not obvious from analyses of the THRA-mutant mice, accounting for some of the difficulty in its recognition (8). Further results in the present report show that the hypothalamic-pituitary-thyroid axis in these patients is normal, indicating that the unusual pattern of the serum T₄ and T₃ is secondary to alterations in peripheral thyroid hormone metabolism.

What is the basis of the high serum T₃/T₄ ratio in RTHα, and what does it tell us about the physiological role of TRα in humans? Both the family in the present report and the other individual recognized with this syndrome have mutations that lead to deletion of, and/or alterations in, amino acids in the extreme carboxy-terminus of the TRα1 receptor protein. The insertion of a nucleotide after codon F397fs406X in the 2 individuals reported here changes 6 carboxy-terminal amino acids of TRα1 and deletes the last 5, whereas the E403X mutation in the other propositus eliminates the last 7 amino acids of TRα1 (9). Structure/function studies have identified this region of both receptors as critical for T₃-binding and subsequent corepressor dissociation. Biochemical studies and structural modeling show that both mutant TRα1 receptors are intrinsically inactive and confer dominant negative inactivation to T₃-responsive genes in vitro (1–3).

A mouse model created to explore the phenotype of mutations in TRα1 is helpful in understanding the pathophysiology associated with these mutations (10). In this particular mouse, a single nucleotide insertion found in TRβ in an individual with severe RTH from the PV-kindred was introduced into the homologous position of the mouse THRA gene (10, 11). This mutation is termed TRα1PV and disrupts the last 17 amino acids of TRα. Early studies of the TRα1PV knock-in mouse showed a slight elevation of serum T₃, no change in serum T₄, and a modest but significant increase in TSH. Even more striking was a 9-fold elevation in the hepatic type 1 deiodinase (D1) mRNA. This is a well-recognized T₃/TRβ-dependent gene, and subsequent studies of this mouse indicated that the serum T₃ was about twice normal (12). Because D1 converts T₄ to serum T₃, the elevated D1 increases serum T₃ and raises the serum T₃/T₄ ratio. This mouse model provided an additional surprise with respect to thyroid hormone metabolism. When T₃ was administered parenterally to wild-type and TRα1PV mice, the serum concentration rose 8-fold higher in the mutant than in the wild-type mouse indicating a marked impairment of T₃ clearance (12). A major mechanism for T₃ degradation is via type 3 deiodinase (D3), encoded by the T₃-stimulated Dio3 gene (13). D3 activity was reduced to 30% of normal in the cerebral cortex of the TRα1PV mice and did not increase with exogenous T₃ administration as it does in wild-type mice. This indicates that Dio3 gene transcription is TRα1-dependent, as confirmed by Barca-Mayo et al (14) and that the expected T₃ stimulation of Dio3 transcription was blocked by the dominant-negative effect of the TRα1PV mutant (12, 14). The basal high serum T₃/T₄ and T₃/rT₃ ratio found in all 3 individuals identified with RTHα (see Figures 1 and 2 in Ref. 1) are consistent with the results in the TRα1PV mouse, although it is not possible to determine the relative contributions of the increase in hepatic D1 activity and the decrease in D3-mediated T₃ inactivation to the altered thyroid hormone profile observed in these patients (1–3).

Many questions remain about the RTHα phenotype. We are not certain whether the high serum T₃/T₄ ratio is specific only for mutations that eliminate the carboxy-terminal amino acids of TRα1 or whether it will be true for others as well. A similar tendency is found in other mouse models with different mutations in TRα receptors (reviewed in Ref. 8). In addition, there may be additional effects of TRα1 mutations in the intestine on other deiodinase-regulating molecules such as sonic hedgehog or β-catenin, which could contribute further to alterations in thyroid hormone signaling (15–17). There appears to be only modest benefit of levothyroxine started at age 5 or later, although the relief of constipation is a consistent positive effect in all 3 patients (1, 3). If we could identify the RTHα patients at birth or soon thereafter by their growth retardation and the putatively characteristic serum T₃/T₄/TSH pattern, a trial of T₄ or other potential therapies could be attempted earlier. Some newborn hypothyroid screening programs quantitate T₄ in all newborns followed by a TSH on the lowest 5–10% of the samples (18). If the hormonal pattern in the 3 RTHα patients recognized to date is representative, infants with a low T₄, but normal TSH, could have the blood spot tested for T₃, a high value indicating a higher risk for RTHα and the need for further evaluation. The recognition of the RTHα phenotype provided in this and earlier studies will hopefully permit identification of affected individuals at an earlier age to allow attempts at prevention of permanent developmental dysfunction.