2025 Volume 48 Issue 10 Pages 1621-1633
Resistance to thyroid hormone-β (RTHβ) is caused by mutations in thyroid hormone receptors (TRs), with palpitations, tachycardia, and/or goiter being the most frequently reported clinical features. In recent years, several synthetic thyromimetics have been developed to target mutations associated with RTHβ. Resmetirom, which was granted an accelerated approval by the U.S. Food and Drug Administration, is a TRβ-agonist for the first line therapy for metabolic dysfunction-associated steatohepatitis. This study aimed to evaluate the potential mechanisms of action of resmetirom on RTHβ through co-factors and association with clinical manifestations. We selected cases from 13 clinical records based on recently reported domestic cases in Japan. Based on these clinical reports, we conducted screenings using clinical symptoms, laboratory findings, and mutation data. We attempted to unveil the interaction between 26 RTHβ mutants and resmetirom focusing on recruitment of co-factors. Among the 26 probands, dominant-negative effects (DNE) were identified in 12 mutant TRs (48.5 ± 11.8%). Resmetirom was involved in the recruitment of the co-activators, steroid receptor coactivator-1 and glucocorticoid receptor-interacting protein-1, as well as the co-repressors (CoRs), nuclear-CoR and silencing mediator of retinoic acid and TR. Co-factor recruitment by resmetirom was detected in all mutants. In eight patients with DNE, an association between transcriptional activity and clinical symptoms was observed, which were the reasons for clinical investigation. Notably, in the helix-12 mutant-P453, mild induction of DNE was associated with the recruitment of CoRs, suggesting that resmetirom may be effective in alleviating subjective symptoms in mutants with attenuated DNE located in helix-12.
Some patients with resistance to thyroid hormone-β (RTHβ) show several manifestations, such as tachycardia, palpitation, depression, hyperhidrosis, attention deficit disorders, and goiter.1,2) Among the symptoms commonly observed in RTH, palpitations and tachycardia are the most prevalent and consistent manifestations. Goiter is a characteristic phenotype of this mutation in cases where no subjective symptoms are present. These symptoms are primary reasons prompting patients to seek medical attention.
Patients with RTH are sometimes misdiagnosed with Graves’ disease, pituitary adenoma, or familial dysalbuminemic hyperthyroxinemia because they present with syndrome of inappropriate secretion of thyroid-stimulating hormone (SITSH). Persistent thyrotoxic symptoms, such as Graves’ disease, necessitate anti-thyroid drug treatment or surgery. In fact, there is currently no one-size-fits-all therapy to fully correct RTHβ. Therefore, several thyromimetics have been designed to restore the normal functions of wild-type thyroid hormone receptors (TRs) by mimicking their interactions with triiodothyronine (T3). Although some of these compounds have undergone clinical trials, every trial has been discontinued due to adverse effects.3) Currently, resmetirom is the only thyromimetic available for clinical applications.
In our previous study, we reported the following findings: (1) certain mutants exhibit partial recoverability in the presence of resmetirom; (2) the potential for recovery increases in specific mutants when resmetirom is used at the higher concentrations within those approved by the U.S. Food and Drug Administration (FDA); (3) the binding properties and characteristics of resmetirom to its receptor differ from those of T3; (4) in cases where there is a risk of thyrotoxicosis, such as in patients with concurrent Graves’ disease, caution is necessary because resmetirom may exert additive effects with triiodothyronine; and (5) the characteristic laboratory findings of RTHβ resemble those of SITSH, and resmetirom does not exert harmful effects in this context.4)
In this study, we conducted an investigation based on our previous findings. We selected cases with point mutations in Japan with detailed clinical course descriptions from published reports. Our study aimed to assess the effect of resmetirom on co-factor binding in each mutant and compare its effects with T3. We also considered the correlation between co-factor binding and typical clinical symptoms, such as palpitations and tachycardia, in the presence or absence of resmetirom.
In this study, we analyzed 26 probands with RTHβ caused by missense mutations reported in Japanese individuals, based on cases identified through the PubMed (http://www.ncbi.nlm.nih.gov/pubmed/) and Japan Medical Abstract Society (https://search.jamas.or.jp/) databases5–17) (Table 1). Nonsense, insertion, deletion, and frameshift mutations were excluded from this analysis. Reasons for the clinical investigation, such as chief complaints, were adopted as described in each original report. However, given that RTHβ is frequently associated with autoimmune thyroid disorders, cases initially diagnosed as Graves’ disease were specifically noted in Table 1.
| Mutation | TSH (μU/mL) | fT4 (ng/mL) | fT3 (pg/mL) | SITSH | Sex | Age (y.o) | Reason for investigation, chief complaints, manifestations | References |
|---|---|---|---|---|---|---|---|---|
| R243W | 5.38 | 2.87 | 4.8 | Y | M | 0 | Tachycardia (heart rate 150–160 beat/min) | 4) |
| 1.27 | 2.2 | 5.9 | Y | F | n/a | n/a | 5) | |
| 5.71 | 3.59 | n/a | Y | M | 31 | None | 6) | |
| L246V | 5.23 | 2.4 | 2.5 | Y | M | 50 | None | 5) |
| I250F | 0.80 | 2.5 | 7.4 | Y | M | 60 | None | 5) |
| 2.2 | 2.5 | 5.9 | Y | M | 46 | None | 6) | |
| 2.75 | 2.81 | 6.3 | Y | M | 48 | Palpitation, diarrhea, and malaise | 6) | |
| G251R | 0.007 | 3.0 | 9.3 | N | F | 33 | Palpitation, G | 7) |
| V264A | 8.41 | 2.1 | 6.0 | Y | F | 29 | None | 5) |
| D265G | 5.72 | 3.1 | 14.6 | Y | F | 8 | Goiter | 5) |
| 1.42 | 2.84 | 7.18 | Y | F | 14 | None | 5) | |
| F269L | 2.26 | 5.09 | 2.09 | Y | M | 22 | Atrial fibrillation and palpitation | 8) |
| R282G | 1.78 | 2.62 | 8.29 | Y | M | 7 | AD/HD, LD | 9) |
| 0.68 | 2.2 | 4.5 | Y | M | 50 | Palpitation | 5) | |
| R320C | 5.30 | 2.9 | 6.0 | Y | F | 0 | None | 5) |
| 4.95 | 2.7 | 6.6 | Y | F | 0 | None | 5) | |
| 2.4 | 4.02 | 6.22 | Y | F | 52 | Goiter | 6) | |
| 1.6 | 2.1 | 5.2 | Y | F | 27 | Episodic palpitation | 6) | |
| R320P | 1.982 | 2.76 | 4.99 | Y | M | 31 | Atrial fibrillation | 10) |
| T329S | 2.53 | 2.1 | 7.4 | Y | M | 45 | Dizziness | 5) |
| L330S | 2.44 | 4.4 | 10.8 | Y | F | 31 | None (goiter) | 5) |
| G332E | 1.81 | 3.40 | 7.91 | Y | F | 25 | Palpitation, hyperhidrosis | 5) |
| R338W | 2.70 | 3.71 | 8.5 | Y | F | 11 | None (small goiter) | 11) |
| 1.76 | 3.5 | 10.2 | Y | M | 36 | Atrial fibrillation | 5) | |
| 2.45 | 4.3 | 11.8 | Y | M | 34 | None | 6) | |
| Q340H | 0.95 | 2.6 | 6.2 | Y | M | 42 | None (mild goiter) | 5) |
| K342I | >200 | 0.7 | n/a | Y | F | 40 | n/a | 5) |
| 0.029 | 2.17 | 7.52 | Y | M | 37 | None, G | 6) | |
| G345C | 23.4 | 5.40 | n/a | Y | F | 0 | Tachycardia (with DUOX2 variants) | 12) |
| G347A* | 1.10 | 2.90 | 6.00 | Y | F | 21 | None (mild goiter + alopecia), H | 13) |
| R383C* | 6.2 | 2.4 | 8.3 | Y | M | 5 | None | 14) |
| I431M | 0.9 | 2.7 | 8.2 | Y | F | 18 | Goiter, mild tachycardia (heart rate 80–120 beat/min) | 14) |
| F439L* | 8.41 | 20.2 | 5.93 | Y | F | 64 | None | 5) |
| K443E | 3.84 | 2.8 | n/a | Y | F | 31 | n/a | 5) |
| F451L | 2.4 | 2.5 | 4.3 | Y | F | 33 | Palpitation and hyperhidrosis | 5) |
| 8.41 | 2.02 | 5.93 | Y | F | 29 | None (mild goiter) | 6) | |
| P453R | 0.56 | 3.4 | 6.9 | Y | M | 40 | Palpitation | 5) |
| P453S | 3.57 | 3.5 | 4.8 | Y | M | 47 | Palpitation | 5) |
| 7.7 | 2.9 | n/a | Y | F | 30 | None | 6) | |
| 4.62 | 2.64 | 9.1 | Y | F | 9 | Goiter | 6) | |
P453T |
1.124 | 3.39 | 5.13 | Y | F | 28 | Palpitation, H→G→H | 15,16) |
| 3.62 | 2.4 | 4.4 | Y | F | 30 | Palpitation | 5) | |
| 3.48 | 4.15 | 9.62 | Y | M | 8 | Tachycardia, hyperhydration, and hyperactivity | 6) | |
| 14.66 | 3.40 | 8.43 | Y | F | 0 | None | 6) |
From cases reported domestically between 2001 and 2025 in English, we generated mutant-thyroid hormone receptor expression constructs. Furthermore, based on the Japan Medical Abstract Society (JMAS) reports, we also depicted the reasons for the probands’ medical consultation and their initial examination values for probands with identical missense mutations. Cases exhibiting syndrome of inappropriate secretion of thyroid stimulating hormone (SITSH) were marked with “Y,” while those without were marked with “N.” *: dose escalation study in Fig. 3. M: male; F: female; n/a: not available; JMAS: the Japan Medical Abstract Society; G: Graves’ disease; H: Hashimoto’s thyroiditis; AD: attention-deficit; HD: hyperactivity disorder; LD: learning disorder.
T3, a thyroid hormone (TH), was obtained from Nacalai Tesque Inc. (Kyoto, Japan). Resmetirom was purchased from MedChemExpress LLC (NJ, U.S.A.) and dissolved in dimethyl sulfoxide to achieve the desired final concentration for experimental use.
Plasmid Construction Generation and Characterization of Thyroid Hormone Receptor β1 (TRβ1) MutantsWild-type human TRβ1 expression plasmid (pCMX-TRβ1) was provided by Dr. K. Umesono and Dr. R. M. Evans (Salk Institute, San Diego, CA, U.S.A.). Based on previous studies, 26 TRβ1 mutants were generated (Table 1). Specific mutation sites introduced in TRβ1 and VP16-TRβ1 and the corresponding primer sequences are detailed in Fig. 1 and Table 1. Site-directed mutagenesis was performed using PrimeSTAR® Max DNA Polymerase (TaKaRa Bio Inc., Shiga, Japan), with pCMX-TRβ1 serving as the template and primers listed in Supplementary Table 1. Fidelity of the introduced mutations was verified using DNA sequencing (FASMAC Co. Ltd., Kanagawa, Japan). For functional assays, we used thyroid hormone response element (TRE)-tk-Luc, a reporter plasmid containing a direct repeat 4 (DR4) TRE, to assess transcriptional activity.
The TRβ1 protein is depicted as a single ribbon, with amino acid residues numbered according to a widely accepted consensus nomenclature. Mutation sites are indicated by individual dots placed along the ribbon structure. The lower ribbon highlights specific regions of the TRβ1 protein that are frequently associated with mutations. These “hotspot” regions include amino acid residues (aa) 234–282, referred to as Cluster 3; aa 310–353, referred to as Cluster 2; and aa 329–460, referred to as Cluster 1. Conversely, the figure also delineates “cold regions” where mutations are infrequently reported. These include amino acid residues 282–310 (Cold Region 2) and aa 353–429 (Cold Region 1), as detailed in the lower panel of the figure.
In this study, we used TSA201 cells, a derivative of the human embryonic kidney 293 (HEK293) cell line, obtained from the Riken Cell Bank (Ibaraki, Japan). TSA201 cells were cultured in the Dulbecco’s Modified Eagle’s Medium (DMEM) (Nissui Pharmaceutical Co. Ltd., Tokyo, Japan) supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. The cells were maintained under standard culture conditions of 5% CO2 and 95% air at 37°C. The concentration of resmetirom used for TSA201 cell treatment was determined based on previous reports. To minimize the potential cytotoxic effects at higher concentrations, the dose was restricted to a maximum of 1.0 µM.
Calcium Phosphate Cell Transfection and Transient Expression AssaysTransient expression studies were performed as previously reported.4) In brief, TSA201 cells were seeded in 12-well plates at a density of 2 × 105 cells per well and incubated for 16 h before transfection. Calcium phosphate transfection was used to introduce the following plasmids: TRE-tk-Luc (50 ng) as the reporter plasmid, CMX-TRβ1, or their mutant constructs CMX-mutant-TRβ1 (CMX-mTRβ1) (50 ng) and pGL4.70 (20 ng) as the internal control plasmid. Five hours post-transfection, the culture medium was replaced with phenol red-free DMEM supplemented with charcoal-stripped FBS. The cells were then treated with T3 at 0.1, 0.5, and 1.0 nM and/or resmetirom at 100, 500, and 1000 nM, as described in previous studies. Following a 20-h incubation, cells were harvested, and luciferase activity was measured using the Dual-Luciferase® Reporter Assay System (Promega, Madison, WI, U.S.A.) according to the manufacturer’s protocol.
Mammalian Two-Hybrid AssaysA mammalian two-hybrid assay was performed according to previously established protocols.4) Briefly, TSA201 cells were co-transfected with Gal4-co-factor expression plasmids and either VP16-TRβ1 or VP16-mTRβ1 construct, along with 50 ng of the UAS-E1b-TATA-Luc reporter gene. Transfections were conducted in the absence or presence of ligands to assess ligand-dependent interactions. In this study, we used steroid receptor coactivator 1 (SRC1), glucocorticoid receptor-interacting protein-1 (GRIP1), CREB-binding protein (CBP), and p300 as co-activators (CoAs), and NCoR and SMRT as co-repressors (CoRs). The assay was designed to evaluate the recruitment dynamics of these co-factors in response to wild-type-TRβ1 or mutant-TRβ1 activation.
Statistical Analysis and Data PresentationAll statistical analyses were conducted using IBM SPSS Statistics software (SPSS Inc., NY, U.S.A.). Experimental data are expressed as mean ± standard deviation or standard error, derived from at least three independent in vitro experiments. To determine statistical significance, one-way ANOVA was performed, followed by post hoc multiple comparisons using the Bonferroni method or an uncorrelated t-test, where appropriate. A p-value of <0.05 (p < 0.05) was considered statistically significant.
Figure 1 provides an overview of the amino acid positions affected by missense mutations in the 26 cases reported across 13 studies. Except for one case, all mutations were consistent with previously reported data, showing an accumulation of known hotspots, particularly at codon P453, which was the most frequently affected. Using wild-type TRβ1 as a control, transcriptional activity was evaluated for each mutant, and the results are presented in Fig. 2. In this figure, the dashed line represents the transcriptional activity of wild-type TRβ1 in the presence of 0.1 nM T3, while the solid line represents activity in the presence of 100 nM resmetirom. The open columns show the silencing effects of wild-type and mutant TRβ1, as determined by assays of each mutant TRβ without any ligands. Due to the silencing effect of TRβ1, transcriptional activity in the absence of ligand was suppressed compared with the reporter-only control. Among the 26 missense mutations, 12 exhibited statistically significant dominant-negative effects (DNEs), with a mean activity of 48.5 ± 11.8% (Supplementary Table 2). Notably, mutations in helix 3, including G251R, V264A, D265G, F269L, and R282G, resulted in pronounced DNE. In particular, R282 is situated near the ligand-binding pocket and contributes to ligand stability via hydrogen bonding with the carboxyl group of T3. Similarly, R320, which resides within the ligand-binding pocket, forms a hydrogen bond with the carboxyl group of the ligand and is critical for ligand binding stability. This explains the strong DNE observed with its mutation. Mutations at positions L246, G251, G332, Q340, K342, G347, R383, and F439 showed statistically significant differences in response to T3 versus resmetirom, indicating that these residues play a role in the interaction between resmetirom and TRβ.
Effects of thyroid hormone (TH) and resmetirom on the gene transcription mediated by the native thyroid hormone receptor β1 (TRβ1), or mutant-TRβ1. Either CMX-TRβ1 (50 ng) or CMX-mutant-TRβ1 (50 ng) was co-transfected into TSA-201 cells with 100 ng of the reporter gene TRE-tk-Luc in the presence of resmetirom (100 nM) or the lowest physiological level of T3 (0.1 nM) within the normal range as the reference. To evaluate the silencing effects of wild-type and mutant TRβ1, assays were performed with each mutant TRβ in the absence of any ligands. Results are presented as mean ± standard deviation from at least three transfections performed in triplicate. RLU, relative light units; *p < 0.05; ***p < 0.005; ****p < 0.001 denote significance compared with the corresponding wild-type in the absence of TH or resmetirom; N.S., not statistically significant. The dominant negative effects, referenced against the presence of thyroid hormone and against resmetirom are provided separately as Supplementary Materials (Supplementary Table 2).
Based on the above results, we investigated the concentration-dependent effects of resmetirom using three TRβ mutants (G347A, R383C, and F439L), which do not exhibit DNE, but involve amino acid residues critical for resmetirom binding (Fig. 3). None of these mutants showed increased transcriptional activity in response to resmetirom across its concentration gradient, suggesting that these residues are essential for effective interaction between resmetirom and TRβ1. Given that transcription was saturated at the lowest concentration and did not increase with increasing concentrations, it is plausible that resmetirom occupies the TRβ1 docking groove, thereby preventing the approach of further resmetirom molecules. In contrast, as none of the individuals harboring these mutations exhibited subjective symptoms (Table 1), effects of the mutations may have been compensated by endogenous T3.
For mutant thyroid hormone receptor β (mTRβ1) that did not show significant dominant-negative effects when initially treated with 0.1 nM triiodothyronine, at the lower end of the normal physiological concentration range, as a positive control, we re-evaluated their transcriptional activities with resmetirom using a concentration gradient, as detailed in a previous study.4) This analysis included three specific mutants, which are also illustrated in Table 1. The dashed line represents the transcriptional activity of the wild-type TRβ1 at this initial 100 nM resmetirom concentration, serving as our reference point. Following this, we increased the resmetirom concentration to 500 and 1000 nM. All results are presented as mean ± standard deviation from at least three independent transfections, each performed in triplicate, as depicted in the figure. In these figures, RLU represents relative light units. Statistical significance is indicated by asterisks: *p < 0.05 and ****p < 0.0001 denote significance compared with the corresponding wild-type in the presence of 100 nM resmetirom. N.S:, not statistically significant.
The ability of resmetirom to recruit CoAs was examined using a mammalian two-hybrid assay, focusing on SRC1 and GRIP1, both of which are known to enhance transcriptional activity. Resmetirom showed little to no SRC1 recruitment (Fig. 4a). In contrast, GRIP1 recruitment did not differ significantly between T3 and resmetirom for R243W, G251R, R320P, T329S, G347A, and F439L mutants, indicating that these residues are involved in maintaining histone acetyltransferase (HAT) activity and modulating DNE (Fig. 4b). Notably, only the I250F mutation enhanced GRIP1 recruitment in response to resmetirom, suggesting that this residue played a critical role in ligand-specific CoA interaction.
To evaluate the potency of recruitment co-factors by resmetirom, a mammalian two-hybrid assay was performed using representative coactivators and co-repressors. The effect of resmetirom on the interaction between TRβ1 and co-factors is presented. The format of the mammalian two-hybrid experiment is indicated at the upper right of the panel. Co-factors, such as steroid receptor coactivator 1 (SRC1) (Fig. 4a), glucocorticoid receptor-interacting protein-1 (GRIP1) (Fig. 4b), CREB-binding protein (CBP) (Supplementary Fig. 1a), p300 as co-activators (Supplementary material Fig. 1b), and nuclear receptor corepressor (NCoR) (Fig. 4c), silencing mediator for retinoid, and thyroid hormone receptors (SMRT) (Fig. 4d) as co-repressors, were described in the previous study. Each was fused downstream of the Gal4 DBD in-frame in pSG424, respectively. Each Gal4-co-factor (50 ng) was co-transfected into TSA201 cells with 100 ng and either VP16-TRβ1 or VP16-mutant-TRβ1 together with 100 ng of the reporter gene, UAS-E1BTATA-Luc, in the absence or presence of resmetirom or triiodothyronine (T3, TH). 1.0 µM of resmetirom was added. T3 were input with at 0.1 nM according to the previous report. The open column represents the results of an assay performed on each mutant TRβ alone, without ligands, to show the silencing effects. In the absence of a ligand, the mutant TRβ may maintain or enhance its silencing effect by strengthening its binding to co-repressors. RLU, relative light units; *p < 0.05; **p < 0.01; ****p < 0.001 denote significance compared with the corresponding wild-type TRβ1 in the absence of TH or resmetirom; N.S:, not statistically significant. As described above, CBP and p300 were included in the supplementary materials due to their minimal changes in transcriptional activation with both T3 and resmetirom (Supplementary Materials Figs. 1a and 1b, respectively).
CBP and p300 possess HAT activity and promote transcriptional activation through chromatin remodeling. Regarding their interaction with TRβ1, CBP/p300 interact indirectly via the activation domains 1 and 2 of SRC-1. Consequently, no distinct changes suggestive of resmetirom-induced allosteric effects were observed in the current analysis (Supplementary Figs. 1a, 1b).
For the CoRs NCoR and/or SMRT, the V264A, D265G, F269L, R320P, T329S, L330S, and P453 showed robust recruitment by resmetirom, indicating that these residues are critical for DNE (Figs. 4c, 4d). Specifically, V264A mutation resulted in marked transcriptional suppression (DNE values: 38.7 ± 4.9%). The V264A mutation, located within the "omega loop" region of TRβ, selectively disrupts the release of CoRs while preserving, at least in part, the receptor’s capacity to recruit CoAs. The strong recruitment of CoRs by resmetirom in these cases may be due to substitution of the branched-chain amino acid valine with alanine at position 264, which enhances the ability of resmetirom to stabilize CoR binding. Additionally, T329S exhibited a significant DNE (41.8 ± 3.8% vs. T3 as 100%). In the wild-type receptor, T329 indirectly interacted with the carboxyl group of T3 via a water molecule; however, this interaction was not observed with resmetirom. Interestingly, the T329S variant displayed strong recruitment of SMRT in response to resmetirom, suggesting a ligand-specific alteration in the CoR interaction.
In I250F, no DNE was observed, and transcriptional activity was maintained. Resmetirom-induced transcriptional activity was preserved at 74.6% relative to that of the control. However, since no recruitment of co-factors was observed, it is likely that this residue affected the conformation of TRβ without directly impacting the interaction with either CoAs or CoRs.
In the wild-type TRβ, R282, R320, and L330 residues are reported to form interactions with the carboxyl group of T3 or the oxygen atom in the heterocyclic ring of resmetirom.18) Accordingly, mutations at these positions (R282G, R320C, R320P, and L330S) were associated with transcriptional repression and a reduced response to resmetirom. Specifically, R282G showed a DNE with transcriptional activity of 34.0 ± 3.8% (58.2 ± 5.9% vs. resmetirom 100%), R320C showed 57.6 ± 2.3% (34.3 ± 5.1%), and L330S showed 41.9 ± 10.3% (23.9 ± 3.1%), indicating that these mutations substantially impair ligand responsiveness.
Although R320C and R320P exhibited comparable DNEs, none of the mutants showed recruitment of the co-activator GRIP or CoRs under basal conditions. However, in the presence of resmetirom, both mutants demonstrated the recruitment of CoRs, with a particularly strong interaction observed with SMRT. These findings suggest that substitution of arginine with either cysteine or proline at position 320 alters the functional group of the residue in a manner that directly affects resmetirom binding. This highlights the fact that R320 is a critical site for ligand-specific modulation of co-factor recruitment.
In wild-type TRs, TRβs are primarily responsible for mediating thyroid hormone action. When thyroid hormone levels rise beyond a certain threshold, the hypothalamic–pituitary–thyroid axis exerts negative feedback, leading to a reduction in thyroid-stimulating hormone (TSH) levels. However, in a subset of patients with RTH, variations in mutation sites and individual differences potentially influenced by epigenetic factors may contribute to the development of SITSH. In patients with RTH who present with few or no symptoms, SITSH is often observed, although compensatory changes are present and therapeutic intervention is rarely required. When SITSH manifests severely, excessive levels of circulating TH can accumulate, reaching concentrations sufficient to activate TRα-mediated thyroid hormone signaling. In such cases, inappropriate TRα activation occurs in organs and tissues where TRα expression predominates, potentially leading to tachycardia and palpitations, which are the most common and distressing symptoms experienced by patients with RTH. Even in cases in which subjective symptoms are mild, diffuse goiters may develop. Although mutations arise randomly, those lacking DNEs are unlikely to be detected clinically.19)
TRα1 is the predominant subtype and mediating T3 effects in cardiac myocytes, though both TRα1 and TRβ1 are expressed in the heart.20) Miller et al. generated and analyzed a knock-in mouse harboring a human mutation that reproduced human RTH.20,21) They reported that RTHβ forms heterodimers with TRα1, resulting in suppression of the transcriptional activity of TRα1.20) Furthermore, using microarray analysis, they demonstrated that genes that are normally suppressed by TH were not adequately repressed in knock-in mice, leading to an inappropriate induction of numerous genes. Considering those previous reports, although the location and various types of genetic mutations may affect the condition, the aforementioned background is presumed to affect cardiac function. It has been suggested that moderately weak DNEs serve as preconditions for the manifestation of thyrotoxic symptoms. Accordingly, in the present study, we compared the binding affinity of mutant TRβ1 to the DR4-promoter, correlating these molecular interactions with patients’ subjective symptoms of palpitations and/or tachycardia. We further investigated whether selective TRβ-agonist resmetirom, and variations in TRβ1–TH binding affinity, modulate transcriptional activity at this promoter. In a previous study, we investigated the effect of resmetirom on the binding affinity of mutant-TRβ2 to the TSHα-promoter, but no such effect was observed. Furthermore, we examined additive and synergistic effects in both resmetirom and thyroid hormones, and an additive effect was observed.17) Even in the absence of noticeable symptoms, in the presence of resmetirom and T3, diffuse goiter may develop.
Mutation clusters tended to occur in regions that avoided the dimerization interface (Fig. 1). Mutations in TRβ located within these hotspots markedly reduce the T3-binding capacity, whereas DNA binding to TREs and dimer formation remain largely unaffected. In other words, hotspots are formed to bypass these regions where no mutants are found. This is because the artificial insertion of mutations into these areas does not produce DNE, making them clinically undetectable.22,23) Generally, transcriptional activity activated by T3 is suppressed by unliganded TRs. Therefore, mutant TRs with severely reduced T3-binding capacity are thought to act in a repressive manner similar to unliganded TRs. Most mutant TRβs lose their ability to bind to TH, yet retain the capacity to bind DNA and form heterodimers with their partner, retinoid X receptor. Furthermore, introducing mutations into the domain responsible for binding to CoRs diminishes the DNE of mutant TRβs, suggesting that CoRs play a critical role in mediating this effect. Several studies have shown that the binding affinity of mutant TRs to the CoR SMRT correlates with the degree of DNE.24) Disruption of interactions with CoRs has also been reported to be important.25) Other reports have indicated that the severity of impaired binding to CoRs, rather than to CoAs, is more closely associated with negative transcriptional repression and DNE of mutant TRβs.26) The hydrophobic groove within the ligand-binding domain (LBD) of TRβs, a region that overlaps with the CoA binding site, is also crucial for CoR binding (Fig. 5). This interaction is primarily mediated by the LXXII motif (also known as the CoRNR box), which facilitates the direct association between CoRs and the TRβ1 LBD27–29) (Figs. 1 and 5). Critical amino acid residues, including T277 in helix 3 (H3) and L454 along with E457 in helix 12 (H12), are essential for robust CoR binding. NCoR and SMRT, two closely related CoRs in terms of both structure and function, may mediate non-redundant biological actions. Their interaction with receptors fundamentally relies on the receptor CoRNR box located within helix 1 (H1) of the LBD in the hinge region. Despite the significance of the CoRNR box, the LXXII motif did not directly interact with residues in the H1 region. Instead, it establishes an interaction by docking into a distinct hydrophobic groove on the surface of the LBD, which was specifically formed by H3 and helix 4. Because this binding site is also utilized by CoAs, their binding to the LBD is mutually exclusive. However, the mutants examined in the present study did not contain these critical residues.30) Further accumulation of cases and re-evaluation are needed in future investigations.
The structure of the ligand-binding domain of TRβ is shown as a ribbon diagram based upon the X-ray crystal structure according to 3GWS from PROTEIN DATA BANK.40) The locations of various helices and TRβ mutants are shown as writing on the ribbon or small circle. The predicted region for coactivators (CoAs) and corepressor (CoR) binding is circled.41) The central linear schematic represents the predicted docking of triiodothyronine. H: helix.
Regarding the interaction between TRβ1 and CoAs, SRC-1 (also known as NCOA1) and GRIP1 (NCOA2) interact with TRβ1 through an interaction motif: these CoAs bind to the LBD of TRβ1 via their LXXLL motifs (also known as NR boxes). The binding site on TRβ1 consists of a hydrophobic groove formed by H3 and H12 within the LBD, which accommodates the LXXLL motif.31) The three key amino acid residues depicted above are also critical for CoA binding. The activation function-2 domains of human TRα1 (amino acids 397–405; sequence: PLFLEVFED) and human TRβ1 (amino acids, 453–461) share an identical sequence, suggesting that they serve a common functional role. This conserved sequence is also considered crucial for maintaining the structural integrity required for interaction with CoAs.32,33) Therefore, in RTHβ mutants harboring mutations in this region, it is possible that TRα function is preserved, potentially resulting in enhanced TRα-mediated effects.33) The P453 mutation, located in H12, is associated with a higher likelihood of presenting with symptoms such as tachycardia, palpitations, and hyperhidrosis. This is because H12 is involved in interactions with co-factors, which is also known as the tau 4 activation domain.34) These findings contribute to a growing understanding of the mechanisms underlying negative transcriptional regulation mediated by mutant TRs and TH. However, the extent to which the diverse clinical manifestations observed in patients with RTH can be explained by abnormalities in known RTH receptors remains a major unresolved issue and an enduring enigma.
RTH occurs at an estimated frequency of 1 in 40000 individuals.35) According to the most recent Japanese demographic survey, the total population of Japan was 125858000 (national census on October 1, 2020).36) Based on this estimate, the expected number of patients with RTH in Japan is approximately 3154. According to Tagami,37) the total number of reported RTH cases in Japan between 1991 and 2020 was 103 cases (approximately 3.3%). Dumitrescu et al.2) reported 459 families with identified mutations, including 432 with single-nucleotide substitutions and 11 with nonsense mutations. Among 171 distinct mutations, some were shared across families; notably, R338W was found in 33 unrelated families, and 7 different substitutions occurred at codon 453 (P453T, S, A, N, Y, H, and L), which is the most common variant described in the literature. Although diagnostic triggers in affected individuals have not been described in detail, the presence of subjective symptoms and/or goiters is a possible indication. Ohba et al.6) reported that the most common trigger for diagnosis was the detection of abnormal thyroid function during evaluation for other diseases or annual health checkups, accounting for 41.2% of the cases. The most frequently reported subjective symptoms were palpitations (39.7%) and thyroid enlargement (35.3%). Given that point mutations occur at equal probabilities per nucleotide, the frequent reporting of mutations, such as R338W and those affecting P453T, S, A, N, Y, H, and L in previous studies, suggests that thyrotoxic clinical manifestations, such as palpitations, may serve as key diagnostic triggers.2) Okosieme et al.38) were the first to report reduced survival and increased cardiovascular morbidity in patients with RTHβ in Wales, UK. Campi et al.39) reached a similar conclusion in an epidemiological study conducted in Italy. These outcomes were likely due to chronic cardiac exposure to excess thyroid hormones. Targeted therapies that address hormone resistance pathways may help reduce this risk and improve long-term outcomes.
Similarly, in the present study, domestic reports on probands revealed a high prevalence of mutations in these two amino acid residues, with several individuals presenting with symptoms (Table 1). Resmetirom recruited the CoRs NCoR and SMRT to the P453 variants: NCoR (P453R, 70.9%; P453S, 81.0%; P453T, 45.8%, vs. wild-type TR as 100%) and SMRT (P453R, 123.6%; P453S, 211.3%; P453T, 111.3%, vs. wild-type TR as 100%) (Figs. 4C, 4D). These results suggest that resmetirom may exert therapeutic effects on symptomatic mutations located in H12, where co-factor interactions are critical (Table 1, Fig. 5). In this context, our findings strongly suggest that resmetirom holds promise for clinical applications. Specifically, its use warrants consideration as a therapeutic strategy to alleviate the subjective symptom burden experienced by a carefully identified subset of patients with RTH.
The authors thank Mr. Ryosuke Murakami, M.Eng. for providing the mutation primer creation software “primer_gen_v002.”
This study was supported by Grants from the Foundation for Growth Science (to M.M.) and JSPS KAKENHI (Grant No. 23K07980 to T.T.). The founders played no role in the study design, data collection and analysis, decision to publish, or manuscript preparation.
The authors declare no conflict of interest.
This article contains supplementary materials.
