Two Novel Cases of Resistance to Thyroid Hormone Due to THRA Mutation

Abstract

Resistance to thyroid hormone alpha (RTHα) is a rare and under-recognized genetic disease caused by mutations of THRA, the gene encoding thyroid hormone receptor α1 (TRα1). We report here two novel THRA missense mutations (M259T, T273A) in patients with RTHα. We combined biochemical and cellular assays with in silico modeling to assess the capacity of mutant TRα1 to bind triiodothyronine (T3), to heterodimerize with RXR, to interact with transcriptional coregulators, and to transduce a T3 transcriptional response. M259T, and to a lower extent T273A, reduces the affinity of TRα1 for T3. Their negative influence is only reverted by large excess of T3. The severity of the two novel RTHα cases originates from a reduction in the binding affinity of TRα1 mutants to T3 and thus correlates with the incapacity of corepressors to dissociate from TRα1 mutants in the presence of T3.

Introduction

The currently known germline heterozygous mutations in THRA, which encodes thyroid hormone receptor α1 (TRα1), result in many signs of hypothyroidism. The genetic disease, called resistance to thyroid hormone alpha (RTHα), is notably associated with altered skeletal growth and cognitive functions, with near-normal thyroid function tests (1). However, based on the report of only 24 cases, its symptoms and severity seem to be quite variable (2). We report here two patients with germline heterozygous novel missense mutations in THRA causing RTHα.

Results

Two novel cases of RTHα

Two cases of RTHα were diagnosed in two hospitals in France (Table 1; Methods section in Supplementary Data). Case A was a girl born prematurely, with a twin brother. The mother had medically assisted reproduction. The girl was immediately hospitalized due to transient respiratory distress, severe constipation with megacolon, and malaises due to reflex vagal hypertonia. Macroglossia was first noted at 4 months. She had an operation for bilateral ovarian hernia. Her motor development was delayed, and she was unable to sit for the first year. She could walk only at three years old. In contrast, language acquisition was normal from 1 to 3 years old. Mild hypotonia and a coarse voice were noticed at two years old.

Table 1.

Clinical and Biochemical Parameters of Patients

	Patient A, M259T	Patient B, T273A
Birth and early development
	Cesarean, premature	Cesarean
Birth weight (kg)	1.75	3.2
Birth height (cm)	46	50
Head circumference (cm)	33.5	38^a
Development milestones	Sitting 1 year	Walking 3 years
Development milestones	Walking 3 years	Walking 3 years
Physical and biochemical parameters at age 2 years (with reference values)
Weight (kg)	11.5	15.4 (+3 SD)
Height (cm)	80	86 (+0.5 SD)
BMI (kg/m²)	18	21 (+2.3 SD)
Head circumference (cm)	49	54.5 (+4 SD)
fT4 (pmol/L)	NA	6.8 (7.5–21)
fT3 (pmol/L)	NA	9.6 (3.6–10.4)
TSH (mUI/L)	NA	2.09 (0.7–64)
Hemoglobin (g/dL)	NA	9.3
Mean global volume (fL)	NA	70.6
Hematocrit (%)	NA	29
Physical and biochemical parameters at age 12–14 years
Age (years)	14	14
Weight (kg)	35 (−1.8 SD)	91
Height (cm)	141 (−3 SD)	167
BMI (kg/m²)	17.6	32.6 (+3 SD)
Head circumference (cm)	55.5 (+1 SD)	62.5 (+5 SD)
fT4 (pmol/L)	14.0 (12–22)	15.5 (11.5–22.7)^b
fT4 (pmol/L)	10.8 (11.5–22.7)	12.5 (10.7–18.4)^b
fT3 (pmol/L)	6.0 (3.1–6.7)	8.2 (3.5–6.5)^b
fT3 (pmol/L)	6.6 (3.5–6.5)	6.1 (4.7–7.2)^b
TSH (mUI/L)	1.5 (0.27–4.2)	0.4 (0.5–4.9)^b
TSH (mUI/L)	1.6 (0.51–4.9)	4.8 (0.5–4.2)^b
T4/T3 ratio	3.4; 1.6	1.9; 2.0^b
Red blood cells ( × 10¹²)	3.5 (4.0–5.2)	3.1 (3.8–5.4)
Hemoglobin (g/dL)	11.2 (12.0–16.0)	10.1 (11.5–14.5)
Mean glob vol (fL)	95.1 (80–100)	102 (77–91)
Serum amino acids (μM)
Asparagine	75 (32–58)
Proline	290 (117–230)
Citrulline	51 (22–34)
Phenylalanine	82 (44–66)
Methyl-histidine	10 (<6)
Serum organic acids (μmole/mmole creatinine)
Lactic	38 (<36)
Adipic	12 (<6)
Hydroxy-3-methylglutaric	19 (<5)
2-Cetoglutaric	74 (<41)
Cis-aconitic	85 (<83)

Also present in unaffected family members.

Measured without interruption of levothyroxine treatment.

BMI, body mass index; fT3, free triiodothyronine; fT4, free thyroxine; NA, not available; SD, standard deviation; T3, triiodothyronine; T4, thyroxine; TSH, thyrotropin.

Brain magnetic resonance imaging (MRI) showed periventricular high signals, compatible with a myelination defect, but no morphological anomaly. In particular, cerebellum foliation was normal. By the age of three years, she developed regression in terms of language, communication, and playing. At four years old, she had anxiety and depression, repetitive behavior, and fatigue. Chronic constipation was also noticed. Neurologic examination was normal with the exception of a discreet enlargement of polygona and mild global hypotonia. Interestingly an excess of several amino acids was measured in serum and urine at this point, but metabolic investigations were not continued (Table 1).

Between 5 and 9 years old, the patient had many stereotyps, anxiety attacks, and performed automutilation. Her gait remained slightly unstable and skeletal growth was slow. Epileptic seizures occurred between age 8 and 11 years. At age 14 years, because of frequent nausea and vomiting, lack of appetite and weight loss, she had tube feeding. Cold sensitivity was also reported.

While thyroxine (T4), triiodothyronine (T3), and thyrotropin (TSH) levels remained within normal range in the patient, the T3 level was in the upper range, and a low ratio between fT4 to free triiodothyronine (fT3) was noted (from 3.4 to 1.6), which was a clue for RTHα (2). Exome sequencing identified a THRA mutation (c.776T>C;M259T) and l-thyroxine treatment was started [2 μg/(kg·d)]. TSH was fully suppressed, indicating the persistence of pituitary sensitivity. Although a positive change in tonicity was noticed, the treatment did not improve constipation and mood disorders.

Case B was a girl of Guinean origin, born through cesarean section. She was first examined at 2 years old for psychomotor retardation. She was unable to speak, and had an unstable gait. This was associated with macrocephaly and obesity (Table 1). Brain MRI was normal. fT4/fT3 ratio was low (1.4). TSH responded normally to a thyrotropin-releasing hormone challenge (protirelin 250 μg; Ferring France). The concentrations of other pituitary hormones were normal. She did not display any delay in skeletal growth, and the insulin-like growth factor 1 (IGF1) level was within normal range. Developmental support was initiated at 2 years old, together with l-thyroxine treatment [3 μg/(kg·d)] that resulted in a full suppression of TSH (<0.1 mIU/L). Psychomotor development improved allowing her to attend school, while her body mass index increased.

Targeted sequencing was performed for genes involved in T3 signaling when the patient was 13 years old. The patient was found to have a heterozygous THRA mutation (c.817A>G; T273A) absent in her mother and three of her siblings. Other family members could not be genotyped. Except for macrocephalia, which was present in her mother and all her brothers and sisters (58–60 cm +2 to +3 standard deviation), other family members did not display any symptoms of RTHα.

Functional characterization of mutant receptors in transfected cells

An expression vector for wild-type or mutant TRα1 was transfected into HEK293T cells, together with a reporter construct with a minimal promoter containing two DR4 response elements. While the mutations did not affect TRα1 stability in transfected cells (Supplementary Fig. S1), they significantly reduced its transactivation capacity (Fig. 1A). We then mixed a fixed amount of the DNA vector for wild-type TRα1 expression, with increasing amounts of mutant TRα1, to assess the dominant negative activity of mutant TRα1 in the presence of T3 (10⁻⁸ M). In this setting, TRα1^M259T and, to a lesser extent, TRα1^T273A reduced the T3 response of the cells (Fig. 1B).

FIG. 1.

(A) Transactivation assay for the two mutations found in patients. The luciferase activity of a DR4-driven reporter construct was measured in the presence of increasing concentration of T3. The highest luciferase activity observed with the wild-type receptor was used for data normalization. (B) Dominant negative assay. Expression vectors for wild-type and mutant receptor were cotransfected with a DR4-driven reporter construct. While T3 concentration (10⁻⁸ M) and amount of wild-type receptor were constant, the amount of mutant receptor was increased (mutant to wild-type ratio from left to right: 0.5, 1, and 2). The first columns correspond to a control where increasing amounts of wild-type receptor were used. (C) Two-hybrid assay for the interaction between the TRα1 LBD and the NCoR corepressor. Constructs expressing a Gal4-NCoR CoRNR1 + CoRNR2 domains fusion protein, the TRα1 LBD fused to the VP16 transactivation domain, were cotransfected with a luciferase reporter construct responsive to Gal4. The association between TRα1 LBD and NCoR in the absence of T3 resulted in high luciferase activity (first column), while the addition of increasing concentration of T3 destabilized the interaction, resulting in decreased luciferase activity. All data were normalized to the enzymatic activity measured in the absence of T3. *indicates a significant difference, p < 0.05. (D) Evolution of thermal stability of wild-type and mutant TRα1 LBDs in the presence of increasing concentration of T3 (0, 0.5, 1, 2, and 5 molar excess of protein concentration). Three independent experiments were conducted for each condition. (E, F) The simulated structure of the mutant is superimposed on the structure of wild-type TRα1. For the mutant M259T, the dashed lines represent the steric hindrance generated by the Cβ carbon and preventing ligand binding. For the mutant T273A, disruption of a hydrogen bond network stabilizing the β-hairpin S1/S2, which is part of the ligand binding pocket. LBD, ligand binding domain; NCoR; T3, triiodothyronine; TRα1, thyroid hormone receptor α1.

As the mechanism underlying the dominant negative influence of mutant TRα1 is thought to reflect its incapacity to dissociate from corepressors in the presence of T3 (2), we used a two-hybrid assay to evaluate the influence of TRα1 mutations on its interaction with nuclear receptor corepressor (NCoR). A VP16 activation domain was appended at the N-terminus of TRα1, and the NCoR nuclear interaction domain, including CoRNR1 and CoRNR2 motifs, was fused to the DNA binding domain of the Gal4 transcription factor. This assay revealed that TRα1^M259T failed to dissociate from NCoR, even at very high concentration of T3, while the dissociation of TRα1^T373A was observed only at high T3 concentration (Fig. 1C).

Biochemical characterization of the ligand binding domain of TRα1 mutants

We produced, in Escherichia coli, recombinant proteins corresponding to wild-type and mutants TRα1 ligand binding domains (LBDs). We verified their purity and structural integrity using sodium dodecyl sulfate page and circular dichroism (Supplementary Fig. S2). The structural stability of LBDs was then assessed in thermal shift assays that monitor the fraction of unfolded protein as a function of temperature (Fig. 1D). Differential melting temperature measurements revealed a destabilization of the unliganded TRα1 LBD by the M259T (−7.2°C) and T273A (−6.4°C) substitutions. T3 addition induced only a slight stabilization of mutant LBDs, consistent with a strong decrease in the T3 affinity for the TRα1 mutants. The loss of affinity for T3 was confirmed by isothermal titration calorimetry (Supplementary Fig. S3).

Inspection of the structure of the wild-type TRα1 LBD bound to T3 (Protein Data Bank [PDB] code 2H79) (3) provided a likely explanation for these defects. This showed that the linear residue M259 makes favorable van der Waals interactions with T3, whereas the presence of a branched threonine at this position is very likely to create steric clashes with T3 and surrounding residues (C255, M256) (Fig. 1E). The structure also showed that the hydroxyl group of T273 is involved in a hydrogen bond network stabilizing the β-sheet S1–S2, which is part of the ligand binding pocket (Fig. 1F). Its substitution into an alanine breaks this network.

As predicted, the mutations did not alter the capacity of TRα1 to form heterodimers with RXRα (Supplementary Fig. S4). Fluorescence anisotropy measurements also showed that the mutation did not alter the capacity of the unliganded TRα1 LBD to recruit fluorescent peptides derived from the NCoR corepressor (CoRNR1 and CoRNR2 motifs; Supplementary Table S1; Supplementary Fig. S5) and to associate, when a large excess of T3 is present, with a peptide derived from the SRC-1 coactivator (NR2 motif; Supplementary Table S1; Supplementary Fig. S5).

Discussion

The symptoms of patient A and B are consistent with previous reports (1) and emphasize mental retardation as a common trait. However, the skeletal growth of patient A was initially normal, without thyroxine supplementation, and only patient B had macrocephalia. Other frequent symptoms of RTHα, like anemia, and the presence of skin tags, were absent. The excess of amino acids in the serum of patient A may be circumstantial. However, a recent metabolomics analysis of mouse models raises the possibility that these could be serum markers of the disease (4).

The mutations found in the two patients have equivalent mutations in RTHβ patients [TRβ1^M313V (5) and TRβ1^T327A (6)]. They modify the structure of the ligand binding pocket and reduce the affinity for T3. In agreement with another recent study (7), a clear correlation exists between the severity of the symptoms and the inability of T3 to dissociate NCoR from the mutant receptors, and the resulting dominant negative activity exerted by these TRα1 variants in heterozygous cells.

Footnotes

Acknowledgments

We thank Chloé Morin for technical help in plasmid construction and transfection experiments. We thank Caroline Mas (ISBG, Grenoble) for technical assistance for ITC experiments. We acknowledge the use of OMNISEC equipment of Laboratory of Spectroscopy and Calorimetry (LEC) at Brazilian Biosciences National Laboratory (LNBio), CNPEM in Campinas.

Author Disclosure Statement

The authors have nothing to disclose.

Funding Information

M.P. was supported by the French National Research Agency (ANR-14-ACHN-0016). This study was supported by a grant from Agence Nationale de la Recherche (Thyromut2 program; ANR-15-CE14-0011-01). The CBS is a member of the France-BioImaging (FBI) and the French Infrastructure for Integrated Structural Biology (FRISBI), supported by the French National Research Agency (ANR-10-INBS-04-01 and ANR-10-INBS-05, respectively). The authors acknowledge the support and the use of resources of Instruct-ERIC and the financial support from the CNPq Programa Ciencia Sem Fronteiras (BJT 300143/2015-0), the CNPq Programa Universal (420416/2016-1).

Supplementary Material

Supplementary Data

Supplementary Table S1

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

References

van Gucht

ALM

, Moran

, Meima

, Visser

, Chatterjee

, Visser

, Peeters

. 2017. Resistance to thyroid hormone due to heterozygous mutations in thyroid hormone receptor alpha. Curr Top Dev Biol, 125:337–355.

Moran

, Agostini

, McGowan

, Schoenmakers

, Fairall

, Lyons

, Rajanayagam

, Watson

, Offiah

, Barton

, Price

, Schwabe

, Chatterjee

. 2017. Contrasting phenotypes in resistance to thyroid hormone alpha correlate with divergent properties of thyroid hormone receptor alpha1 mutant proteins. Thyroid, 27:973–982.

Nascimento

, Dias

, Nunes

, Aparicio

, Ambrosio

, Bleicher

, Figueira

, Santos

, de Oliveira Neto

, Fischer

, Togashi

, Craievich

, Garratt

, Baxter

, Webb

, Polikarpov

. 2006. Structural rearrangements in the thyroid hormone receptor hinge domain and their putative role in the receptor function. J Mol Biol, 360:586–598.

Boumaza

, Markossian

, Busi

, Rautureau

GJP

, Gauthier

, Elena-Herrmann

, Flamant

. 2019. Metabolomic profiling of body fluids in mouse models demonstrates that nuclear magnetic resonance is a putative diagnostic tool for the presence of thyroid hormone receptor alpha1 mutations. Thyroid, 29:1327–1335.

Stagi

, Manoni

, Cirello

, Covelli

, Giglio

, Chiarelli

, Seminara

, de Martino

. 2014. Diabetes mellitus in a girl with thyroid hormone resistance syndrome: a little recognized interaction between the two diseases. Hormones (Athens), 13:561–567.

Asadi

, Torjesen

, Haug

, Berg

. 2008. Biochemical characterization of four novel mutations in the thyroid hormone receptor beta gene in patients with resistance to thyroid hormone. Scand J Clin Lab Invest, 68:563–567.

Wejaphikul

, van Gucht

ALM

, Groeneweg

, Visser

, Peeters

, Meima

. 2019. The in vitro functional impairment of thyroid hormone receptor alpha 1 isoform mutants is mainly dictated by reduced ligand sensitivity. Thyroid, 29:1834–1842.