Normal thyroid hormone (TH) metabolism and action require adequate cellular TH signalling. This entails proper function of TH transporters in the plasma membrane, intracellular deiodination of TH and action of the bioactive hormone T3 at its nuclear receptors (TRs). The present review summarizes the discoveries of different syndromes with reduced sensitivity at the cellular level. Mutations in the TH transporter MCT8 cause psychomotor retardation and abnormal thyroid parameters. Mutations in the SBP2 protein, which is required for normal deiodination, give rise to a multisystem disorder including abnormal thyroid function tests. Mutations in TRβ1 are a well-known cause of resistance to TH with mostly a mild phenotype, while only recently, patients with mutations in TRα1 were identified. The latter patients have slightly abnormal TH levels, growth retardation and cognitive defects. This review will describe the mechanisms of disease, clinical phenotype, diagnostic testing and suggestions for treatment strategies for each of these syndromes.
Thyroid hormone (TH) is essential for normal development and for the physiological function of virtually all tissues. As a consequence, hypothyroidism affects multiple tissues resulting in a variety of symptoms such as fatigue, cold intolerance, constipation, congestive heart failure, and depression. The importance of TH for development is illustrated by the consequences of untreated congenital hypothyroidism, resulting in severe growth failure and permanent mental retardation.
The first report of patients with a familial syndrome of reduced sensitivity to TH at the tissue level was published in 1967. These patients had high levels of TH without clinical symptoms of hormone excess or even with symptoms of TH deficiency in certain tissues. After cloning of the T3 receptor isoforms TRβ1 and TRα1, encoded by the THRB and THRA genes, respectively, it was demonstrated that this clinical syndrome of resistance to TH (RTH) was due to inactivating mutations in THRB (RTH-β). Since then, more than 1000 patients with RTH have been published (see[6, 7] for excellent reviews).
In recent years, other syndromes associated with a reduced sensitivity to TH have been recognized, involving a defect in transport of TH across the cell membrane,[8, 9] a defect in the synthesis of selenoproteins, including TH-deiodinating enzymes resulting in an abnormal TH metabolism, as well as a defect in the T3 receptor TRα1.[11, 12] Thus, normal TH action requires both adequate serum TH concentrations and TH signalling at the cellular level of the target tissues. The current review focuses on the clinical diagnosis and management of all known different causes of a reduced sensitivity to TH.
Regulation of TH bioactivity
Serum TH levels are principally regulated by the hypothalamus–pituitary–thyroid (HPT) axis. The hypothalamus produces thyrotrophin releasing hormone (TRH), which stimulates the pituitary to produce thyroid-stimulating hormone (TSH). TSH acts on the thyroid gland to synthesize TH. The thyroid secretes predominantly the prohormone T4 and to a lesser extent the bioactive hormone T3.
TH bioactivity and action are regulated at the cellular level (Fig. 1). Most actions of TH are initiated by binding of T3 to its nuclear T3 receptors (TRs), located on T3 response elements (TREs) in the promoter region of target genes. Binding of T3 results in a change in the interaction of TRs with co-activator and corepressor proteins and consequently in an altered expression of the target genes. Different TR isoforms are encoded by two genes: THRA and THRB. TRα1 is the predominant TR receptor isoform expressed in brain, bone and heart, whereas TRβ1 is considered the major isoform in liver, kidney and thyroid. Through alternative exon usage, TRβ2 differs from TRβ1 at the N-terminus and displays a more restricted expression pattern (retina, cochlea, pituitary).
Intracellular T3 levels are governed by intracellular deiodinases and TH transporters at the plasma membrane. Three deiodinating enzymes (D1-3) have been identified, which catalyse the activation of T4 to T3 or the inactivation of T4 to 3,3′,5′-triiodothyronine (reverse T3, rT3) and of T3 to 3,3′-diiodothyronine (3,3′-T2). D1 is highly expressed in liver, kidney and thyroid and is considered important for serum T3 production as well as for clearance of serum rT3. D2 is localized particularly in brain, pituitary, brown adipose tissue, thyroid and skeletal muscle. It has been firmly established that D2 is crucial for local production of T3 in different tissues. D3 is an inactivating enzyme catalysing degradation of T3 and T4. D3 is mainly expressed in foetal tissues. In adult life, D3 expression is limited to the brain and skin, but can be reactivated in other tissues under pathological conditions.[16, 17] Recent studies have established that intracellular TH signalling can be largely modified by deiodination without affecting circulating TH levels, thereby modulating processes such as differentiation and regeneration.[16, 18, 19]
Because action and deiodination of TH take place intracellularly, transport of the hormone across the plasma membrane is required. Although many transporters accept TH as a ligand, only a few have been shown to be specific TH transporters. Monocarboxylate transporter 8 (MCT8, SLC16A2) has been shown to specifically transport the iodothyronines T4, T3, rT3 and T2.[21, 22] The highly homologous MCT10 (SLC16A10) was initially designated as a T-type aromatic amino acid transporter, but has later been shown to transport TH, with a preference for T3 over T4.[23, 24] Both MCT8 and MCT10 are widely expressed. The organic anion-transporting polypeptide 1C1 (OATP1C1) is importantly expressed in brain and transports T4. Possibly, other as-yet-unknown specific TH transporters are important for human physiology.
The last decade has witnessed the discovery of several novel syndromes of reduced sensitivity to TH, related to dysfunction in TH transport,[8, 9] deiodination and receptor function.[11, 12]
Causes of Reduced Sensitivity to TH
Defect in TH transport
The clinical importance of TH transporters was established by the discovery of mutations in the MCT8 gene, which is located on the X chromosome, as a cause of psychomotor retardation accompanied by TH abnormalities.[8, 9] Affected males display a severe delay in motor and neurological development. Soon after the description of the first patients, it was realized that the phenotype had similarities to the Allan–Herndon–Dudley syndrome (AHDS), the first X-linked mental retardation syndrome described in 1944. Genetic analysis in these families revealed that MCT8 mutations are the genetic basis of AHDS. To date, over 100 families have been reported with pathogenic mutations in MCT8.
Patients have cognitive impairments with intelligence quotient values mostly below 40. Many patients are unable to speak and are only able to communicate by nonverbal acts. Some patients have been reported to suffer from seizures. All patients have difficulties with swallowing. The consequent feeding problems are one of the reasons for the first referral. Hypotonia of the limbs in childhood progresses into spastic quadriplegia with advancing age. The severe axial hypotonia, which is manifested by a poor head control, persists into adulthood. Muscle hypoplasia, in particular of the quadriceps muscle, is observed in all patients. Most patients are unable to walk independently. At birth, height and weight are usually unremarkable. However, during childhood, weight declines below the third percentile in most patients.
Few patients reportedly have somewhat milder features. Some patients are able to walk without support and can communicate verbally. Patients with a less severe clinical phenotype typically have less abnormal thyroid parameters. MCT8 mutants of less severely affected patients also display residual activity in in vitro TH transport assays, suggesting a genotype–phenotype relationship in the AHDS. In general, female carriers do not exhibit neurological features. However, they have serum FT4 levels in between those in affected males and unaffected relatives.
Laboratory findings and differential diagnosis
Characteristic for all AHDS patients is their remarkable combination of serum TH abnormalities (Fig. 2). Serum (F)T4 concentrations are low or low-normal, while TSH levels are in the high-normal range. In contrast, serum (F)T3 levels are markedly elevated. In particular during childhood serum, T3 levels are far above the upper reference limit. Serum rT3 levels are largely reduced. Consequently, T3/rT3 ratios are strongly increased. This biochemical profile is very similar to the thyroid function tests seen in patients with THRA mutations, although it appears that serum T3 levels are less elevated than those in AHDS patients.
Mechanisms of disease
The mechanisms behind the clinical and laboratory features of AHDS are only partially understood. As shown by in vitro transport assays, TH transport is largely or completely impaired by the MCT8 mutations identified. TH transport capacity is also largely reduced in fibroblasts from MCT8 patients. Thus, abnormal handling of TH transport appears as the basis for the disease.
Several mechanisms contribute to the low T4 levels. In Mct8 knockout (KO) mice, kidney T4 levels are increased despite the low serum T4 levels, suggesting that T4 is trapped in the kidney. At the same time, kidney (and liver) D1 expression is markedly increased, which should result in a prominent increase in peripheral T4 to T3 conversion. Recently, it was shown that MCT8 expression in the thyroid gland is required for TH secretion, which is therefore disturbed in Mct8 KO mice.[32, 33] The consequent accumulation of T4 within the thyroid gland when MCT8 is mutated may thus lead to increased intrathyroidal conversion to T3. This will favour an increased T3/T4 ratio in the thyroid, resulting in a net increase in T3 secretion via other efflux pathways. The important contribution of D1 in the thyroid and peripheral tissues to the high serum T3/T4 ratio in Mct8-deficient mice (and patients) is supported by findings that Mct8/Dio1 double KO mice have normal serum TH levels. Also, block-and-replace therapy of an AHDS patient with LT4 and the thyrostatic drug propylthiouracil (PTU), which also inhibits D1, normalized serum T3 concentrations. This was not the case when methimazole was used, which does not inhibit D1. The decreased serum rT3 levels are caused by reduced availability of its substrate T4 as well as by the elevated D1 activity, for which rT3 is the preferred substrate.
The high serum T3 levels induce thyrotoxic effects on peripheral tissues, which likely explain the progressive loss of muscle mass as well as the decline in body weight during childhood. Also, SHBG levels, which are T3 dependent and reflect liver thyroid state, are markedly elevated in AHDS patients. Although FT4 levels are low, TSH levels appear inappropriately high in the context of the high serum T3 concentrations, suggesting interference in the feedback of TH at the pituitary and/or hypothalamic level.
The neurological phenotype of AHDS patients is much less understood. The current hypothesis holds that derangement of TH homeostasis in the brain likely underlies the mechanism of disease in AHDS, because neuronal differentiation and myelination are TH-dependent processes. This entails a defect of T3 entry in MCT8-expressing neurons and, thus, deprivation of TH in specific brain regions and perhaps excessive accumulation of T3 in neurons which use other transporters for their T3 supply. MCT8 is also importantly expressed in capillaries and, thus, also appears important for transport of both T3 and T4 across the blood–brain barrier. Mct8 KO mice lack neurological features, despite largely impaired T3 uptake into the brain, while T4 uptake is preserved. Apparently, these mice employ compensatory mechanisms. Increased cerebral D2 activity in Mct8 KO mice may produce sufficient T3 for normal brain development. In addition, these animals likely express a specific T4 transporter, such as Oatp1c1, which mediates T4 transport across the BBB. This hypothesis is supported by a recent study of Oatp1c1 KO mice, which revealed markedly reduced cerebral T4 levels. Expression of Oatp1c1 in the mouse, but perhaps much less so in the human BBB, may well explain the differences in brain phenotype between mice and humans deficient in MCT8.
Altogether, AHDS patients exhibit clinical features caused by a combination of hyperthyroid and hypothyroid tissues. Thus, depending on expression of MCT8 or other TH transporters, tissues of AHDS patients are either deprived of TH (brain) or exposed to excess TH (liver and muscle).
Unfortunately, no effective treatment is available for AHDS patients at present. General supportive care should be provided, including adequate feeding support to avoid aspiration and anti-epileptic drugs to prevent seizures if necessary.
Given the low FT4 levels, LT4 suppletion was initiated in some patients with no beneficial effects on peripheral thyrotoxicity.[39, 40] Normalization of serum T4 and T3 levels was readily achieved in a few AHDS patients by block-and-replace therapy using PTU and LT4.[35, 41] This had some beneficial effects such as an increased body weight and reduction in SHBG levels. Hypothetically, cerebral regions that do not rely on MCT8 for their TH supply might also benefit from this treatment. However, treatment with PTU and LT4 did not result in an improvement in cognitive functions in these older patients of 16 and 37 years of age.
Effective therapy should not only normalize toxic TH effects in peripheral tissues but also normalize the disturbed TH signalling in brain. Some studies have been performed with the T3 analogues diiodothyropropionic acid (DITPA) and triiodothyroacetic acid (Triac, TA3) and the T4 analogue tetraiodothyroacetic acid (Tetrac, TA4).[42-44] TA4 is efficiently activated by D2 to TA3, and TA4 and TA3 are inactivated by D3, thus following the normal deiodination routes. It was shown in Mct8 KO mice made hypothyroid that TA4 was able to restore brain development. Studies in Mct8 KO mice with DITPA demonstrated the normalization of TH parameters and attenuation of the thyrotoxic state of peripheral tissues. Importantly, an improvement in several indices of TH action in brain was observed. These observations prompted a study of the possible beneficial effects of DITPA therapy in four AHDS patients, the results of which were published recently. The main consistent findings were a significant decrease in serum T3, with little change in serum T4 and TSH levels. The normalization of T3 levels appeared beneficial for the liver and heart as suggested by the decrease in SHBG levels and heart rate, respectively. Weight gain was noted in some patients but also a progressive weight loss in another patient. None of the patients showed improvement in psychomotor development.
Future strategies to define optimal treatment for AHDS patients may explore the use of alternative TH analogues. The above-mentioned studies in AHDS patients all have in common that these therapies were carried out in patients in whom impaired brain development is likely irreversible. Therefore, it is important to diagnose MCT8 mutations as early as possible.
Defect in TH metabolism
Metabolism of TH is importantly controlled by the iodothyronine deiodinases. Deiodinases are selenoproteins, a small group of proteins within the human proteome containing the rare amino acid selenocysteine (Sec). Sec is crucially required for normal enzyme function, because it is located in the catalytic domain of the deiodinases. An intricate system ensures the incorporation of Sec in selenoproteins and, thus, also in the deiodinases. Sec is encoded by a UGA codon that normally functions as translation termination codon. Recoding of UGA for incorporation of Sec requires the presence of a Sec insertion sequence (SECIS) located in the 3′-UTR of the deiodinase mRNA. The stem-loop structure of the SECIS element is recognized by SECIS-binding protein 2 (SBP2). Subsequently, various factors are recruited including the specific Sec tRNA, which ultimately results in the incorporation of Sec.
In 2005, a novel syndrome of delayed growth and abnormal thyroid parameters was ascribed to mutations in SBP2. Until now, a total of 8 families have been identified (see for a recent excellent overview).
The most prominent feature in all identified families is the growth retardation. In addition to delayed growth, (mild) mental and motor retardation, muscle weakness, hypoglycaemia and impaired hearing infertility are variably reported. One adult patient has been reported with primary infertility and skin photosensitivity as well. As almost all identified patients are children, the natural course of disease is presently unknown. All patients identified until now have residual SBP2 activity. Complete lack of SBP2 is thought to be lethal.
Laboratory findings and differential diagnosis
The typical biochemical findings in patients with SBP2 mutations are elevated serum (F)T4 and rT3 levels, low to low-normal serum T3 levels and normal to slightly elevated serum TSH levels. Serum total selenium levels are reduced. Similar to patients with mutations in TRα1, growth retardation is the most prominent clinical feature. However, all patients identified so far with TRα1 mutations have low FT4 and rT3 levels. The biochemical profile of elevated (F)T4 levels and slightly elevated TSH levels is also seen in patients with RTH or a TSH-producing pituitary adenoma, but these patients have elevated T3 levels as well, whereas patients with SBP2 mutations have low or low-normal T3 levels.
The human genome encodes about 30 selenoproteins, whose syntheses are all impaired by SBP2 mutations. It is not surprising therefore that SBP2 mutations result in a multisystem disorder where the function of many tissues is affected. The deranged TH parameters result from affected synthesis of the deiodinases. At present, it is unknown to which extent each deiodinase contributes to the phenotype. The TH abnormalities seen in patients with SBP2 mutations are reminiscent of most TH serum parameters of mice deficient in both D1 and D2. In line with this, baseline and stimulated D2 activities are much lower in fibroblasts from patients than in cells from unaffected relatives. In vivo studies demonstrated that patients required much higher LT4 doses to suppress TSH levels than unaffected subjects, whereas TSH levels showed a similar response to LT3 treatment in patients and controls. These data suggest a decreased conversion of T4 to T3 at the pituitary level in these patients. The delayed growth may partially result from low T3 levels. Thus, the most prominent features of this disease, that is, growth retardation and abnormal serum thyroid parameters, are explained by impaired function of TH deiodination.
The low serum selenium concentrations in patients with SBP2 mutations result from impaired synthesis of selenoprotein P (SePP) and glutathione peroxidase 3 (Gpx3), which are the major carriers of Se in serum. Detailed investigations suggest that some symptoms (e.g. myopathy) result from tissue-specific selenoprotein deficiency, while other features (e.g. hearing loss, impaired T-cell function) are mediated by impaired cellular antioxidant defence and, thus, increased ROS levels.
Several reports have described the effects of LT3 suppletion to children with SBP2 mutations.[45, 46] A strong initial catch-up growth was noticed on LT3 treatment, although complete normalization was not reached.
Administration of selenium to different patients raised serum SePP and consequently selenium levels.[46, 47] However, selenium supplementation for a few months failed to normalize GPx activity or serum TH abnormalities. Selenium supplementation appears not to have benefits for these patients.
As increased ROS levels underlie parts of the syndrome, antioxidant therapy may be useful to prevent and possibly revert some features. In vitro studies suggest beneficial effects of antioxidants in patient's cells. Studies are awaited to demonstrate the usefulness of antioxidant therapy in patients.
Defect in nuclear T3 receptors
Mutations in THRB (RTH-)
It is known for more than twenty years that heterozygous mutations in the ligand-binding domain (LBD) of TRβ1/2, impairing their hormone binding and/or transcriptional activity, result in RTH. The mutant TRβ interferes with the function of wild-type (WT) TRβ, resulting in a dominant-negative effect and dominant inheritance. In contrast, RTH caused by THRB gene deletions has a recessive inheritance, due to a lack of dominant-negative interference with the WT receptor. Homozygous mutations in TRβ are rare and result in a severe phenotype.
To date, more than 1000 patients with mutations in THRB have been described, belonging to more than 350 families.[6, 7, 50, 51] The mutations cover about 50 different amino acids and are located in three separate clusters.
Patients with RTH-β have a variable phenotype including goitre, tachycardia, raised energy expenditure, hyperactive behaviour, delayed bone age, and learning disabilities[50, 52] (Table 1). In general, RTH is characterized by a relative lack of symptoms despite high serum levels of T4 and T3. Symptoms are due to a combination of low TH action in predominantly TRβ-expressing tissues and TH overexposure in TRα-expressing tissues. Patients who receive treatment to normalize their TH values often develop typical symptoms of hypothyroidism.
Table 1. Comparison of the clinical phenotype and serum thyroid hormone levels in patients with reduced sensitivity to thyroid hormone
Mutated gene name
Severe mental retardation and delayed motor and neurological development
(mild) delayed mental and motor development
ADHD, mental retardation in only minority of patients
Delayed motor and mental development
Decline in body weight during childhood
Delayed bone age and growth retardation
Delayed bone age and short stature in less than half of patients
Delayed bone age and growth retardation
Feeding problems, X-linked
Immune deficiency, hypoglycaemia
Constipation, low IGF1 levels
Normal; slightly elevated
Normal or elevated
Normal or elevated
Laboratory findings and differential diagnosis
RTH-β is characterized by elevated serum TH levels and a nonsuppressed TSH. RTH-β patients secrete a form of TSH rich in sialic acid with a higher bioactivity. This explains the high prevalence of goitre, although TSH levels may be within the normal range or only slightly elevated. RTH-β patients treated by thyroidectomy and/or radioiodine therapy and substituted with different doses of LT4 show a negative log-linear TSH-FT4 relationship with a slope lower than non-RTH patients in concordance with the decreased affinity of the mutated TRβ receptor for T3. Serum thyroglobulin levels tend to be high, reflecting hyperactivity of the thyroid. Serum rT3 levels are high resulting from a decreased activity of D1 which gene is T3 dependent and under control of TRβ.
Diffuse goitre and sinus tachycardia are the most common clinical findings. The combination with high levels of TH may easily lead to the incorrect diagnosis of hyperthyroidism. The differential diagnosis of RTH-β includes all other causes of elevated TH levels in combination with a nonsuppressed TSH. First, abnormalities in serum binding of TH, such as familial dysalbuminaemic hyperthyroxinaemia, thyroid-binding globulin (TBG) excess or transthyretin excess, should be excluded by measuring FT4 and FT3 using a different method or equilibrium dialysis.[56, 57] The subsequent distinction between RTH and a TSH-producing pituitary adenoma (TSHoma) may be the most challenging, because in both conditions, the FT4 and FT3 will be elevated independent of the method used. However, where RTH-β is usually characterized by a relative lack of symptoms, most patients with a TSHoma are hyperthyroid. About 85 per cent of patients with TSHomas have high concentrations of the glycoprotein hormone alpha-subunit, with a relative greater increase than serum TSH. As a result, a high molar serum alpha-subunit to TSH ratio is almost pathognomonic for a TSHoma.[50, 58]
A mutation in the THRB gene confirms the diagnosis of RTH-β. However, in about 10-15% of patients with classical RTH, no mutation in THRB can be detected.[59, 60] Elevated TH levels and a nonsuppressed TSH in other family members may help to distinguish between RTH-β and a TSHoma. If not, additional stimulation and repression tests using TRH and/or T3 can be required to confirm the diagnosis.[6, 50, 58]
Mechanisms of disease
Mutations in THRB identified in RTH-β patients are located in the C-terminus of TRß1/2, mostly contained within three CpG-rich ‘hot spots’ in the LBD (aa 242-460 in TRβ1) and adjacent hinge domain (aa 234-243) of the receptor protein. The mutant TRß proteins have a reduced affinity for T3 and/or abnormal interaction with cofactors (decreased interaction with co-activators or increased interaction with co-repressors). Interestingly, mutations resulting in a complete lack of T3 binding in combination with a reduced affinity for co-repressors may result in a minimal phenotype.[48, 50] The severity of TH resistance varies not only among different tissues in an affected individual but also among different subjects carrying the same gene mutation, even within the same family.[63, 64] The reasons for this variability are not yet understood, but it probably results from genetic variability of cofactors involved in TH action. Despite this variability, careful analyses of patients with various THRB mutations have demonstrated a partial genotype–phenotype correlation of receptor dysfunction vs. clinical symptoms.
Mouse models replicating mutations observed in patients have been generated (such as TRβ-PV and T337Δ). Heterozygous mice show very similar characteristics as heterozygous human patients with RTH-β. In addition, homozygous mice develop metastatic thyroid cancer.
In about 15% of patients and families with RTH-β, no mutation in THRB has been identified.[59, 60] The phenotype of this so-called non-TR RTH is not different from RTH-β. Although screening of different co-activators and co-repressors for mutations has not identified a cause for non-TR RTH yet, it is likely that new mutations will be identified in the near future now that techniques such as exome and whole-genome sequencing become more widely available.
In most patients, treatment is not necessary because the resistance seems to be adequately compensated by the elevated levels of T4 and T3. Tachycardia and tremor as symptoms of hyperthyroidism can be adequately treated using beta adrenergic blockers such as atenolol. In rare cases, treatment with TA3, which has a higher affinity for TRβ than for TRα, can be used to lower serum TSH and TH levels and thereby reduce clinical symptoms of hyperthyroidism.
The treatment of RTH-β during pregnancy is beyond the scope of this review, and for this reason, the reader is referred to an excellent overview article by Weiss and colleagues.[67, 68]
Mutations in THRA
Ever since its characterization in 1987, investigators have searched for patients with mutations in THRA. As TRα1 is not involved in the negative feedback action of TH, no major changes in serum TH levels were expected. To unravel the physiological role of TRα, different mouse models were generated. Interestingly, mice devoid of all TRs have less symptoms of hypothyroidism than wild-type hypothyroid mice, consistent with a repressive effect of unliganded TRs. Unliganded TRα1 seems to play a major role in the cerebellar damage in hypothyroid mice.[70-75] Only very recently, the first human patients with heterozygous-inactivating mutations in TRα1 were identified.[11, 12]
The phenotype of these first patients with inactivating RTH-α includes growth retardation (Fig. 3a), delayed bone development (Fig. 3b), mildly delayed motor and mental development, abnormal thyroid function tests, low GH and IGF1 levels and constipation.[11, 12, 76] These findings support an important role of TRα1 in bone, brain, intestine and possible involvement in GH regulation.
In addition to the growth retardation, delayed bone development is an important part of the phenotype. Both young patients had a delayed tooth eruption, delayed closure of the skull sutures and a clearly delayed bone age.[11, 12, 76] Also, motor and mental development was delayed in both patients. One of the two patients was even admitted to the hospital at 9 months of age because of a delayed motor development. At that age, she was not able to sit by herself and she did not have full control of head movement (pers. comm. from D. Chrysis).
Mice with a TRα1 mutation have a very diverse phenotype, depending on the location and the severity of the mutation. These diverse phenotypes are probably due to different interactions of unliganded TRα1 mutants with corepressors, whose expression is tissue dependent and developmentally regulated.[62, 77] Although the phenotypes of the first patients with mutations in THRA are similar,[11, 12, 76] there are differences as well. For example, constipation is more severe in the patient described by Bochukova and colleagues, whereas the serum T3 is much more elevated in the other two patients. Similar to the different mouse models, it can be expected that mutations with a different location or a less detrimental effect on the function of TRα1 will have a different or more subtle effect on the clinical phenotype.
Laboratory findings and differential diagnosis
The abnormal thyroid function tests in patients with mutations in TRα1 include low (F)T4, high T3, low rT3, but normal TSH levels. The high T3/T4 ratio as well as the low rT3 levels in all three RTH-α patients suggests an altered expression of deiodinases, which are the most important mediators of peripheral metabolism of TH. TRα1-PV mutant mice, with a similar frameshift mutation in TRα1 as two of the three patients, have increased levels of hepatic D1. Increased D1 activity results in increased T4 to T3 conversion and degradation of rT3. In addition, TRα1−/− mice have an impaired regulation of D3, which leads to a reduced production of rT3 and degradation of T3.[78, 79] Both changes in D1 and D3 expression may contribute to the particular TH changes in patients with TRα1 mutations.
Although patients with TRα1 mutations may have several clinical symptoms of hypothyroidism,[11, 12, 76] the diagnosis will be easily missed when only TSH and/or FT4 are measured, because both can be normal. From the patients who have been identified so far, it seems that an elevated T3/T4 ratio and a low serum rT3 are the hallmarks of the biochemical diagnosis. These changes in thyroid function tests are very similar to those seen in patients with MCT8 mutations, suffering from severe X-linked mental retardation.
It remains to be determined whether patients with more subtle mutations in TRα1 have a similar biochemical profile. The identification and detailed characterization of additional patients with RTH-α is of critical importance. Because one affected allele is sufficient to develop the disease and patients seem to be fertile, THRA mutations may be a relatively frequent cause of growth abnormalities and/or cognitive defects which may respond to LT4 treatment depending on the severity of the mutation.
Mechanisms of disease
The mutations identified in the first human patients (F397fs406X, E403X) are located in the C-terminal domain of TRα1. Figure 4 shows the lack of T3 stimulation of the TRα1-F397fs406X mutation, which has been identified in two (father and daughter) of the three patients described so far.[11, 12] The mutant TRα1 has a dominant-negative effect over WT TRα1 when transfected in a 1:1 ratio, explaining the autosomal dominant effect.
The mutations in the first three human patients are very similar to mutations in the different TRα1 mouse mutants that have been generated.[71-75] TRα1-T394fs406X and TRα1-L400R have been reported to lack affinity for T3 as well.[71, 73] Interestingly, the phenotypes of TRα1-T394fs406X (TRα1-PV) and TRα1-L400R mice are very similar to the phenotype of the first human patients, with severe growth retardation as the most prominent phenotype.[71, 73] In contrast, TRα1-R384C mutant mice exhibit a more severe cerebral phenotype, although these mice have mild growth retardation as well.[80, 81] The TRα1 mutation in these mice is more subtle, resulting in some residual T3 binding capacity of TRα1.
An initial catch-up in growth rate and bone age was observed in the index patient identified by our group when TH therapy was started at 6 years of age (Fig. 3a), although no clear effect of LT4 on growth was observed in the other patient identified.[11, 12, 76] Additional GH treatment had little effect in our patient. In addition to the initial catch-up in growth rate, we observed a normalization of the dyslipidaemia, as well as a response in serum IGF1, SHBG and creatine kinase in the index patient. Besides this biochemical response, LT4 also resulted in an improvement in the constipation,[11, 12, 76] whereas cognitive and fine motor skill defects remained. Thyroxine treatment in the other patient (50 mcg daily for 9 months) resulted in a normalization of her basal metabolic rate and circulating IGF-1 levels, whereas heart rate and blood pressure remained low. At this dosage, her growth rate and intestinal transit time did not change significantly.
Interestingly, growth retardation can be overcome in TRα1-R384C mice, whose receptor has a 10 times decreased affinity for T3, by raising serum TH levels.[74, 80, 81] Furthermore, these mice display neurological damage, which improves after T3 treatment. A delayed cerebellar development and locomotor dysfunction is prevented by postnatal T3 treatment, whereas anxiety-like behaviour and reduced recognition memory are relieved by T3 treatment in adulthood. All patients who have been identified so far have mutations resulting in a complete lack of T3 binding. Based on these studies in TRα1-R384C mice, it is tempting to speculate about the beneficial effects of LT4 treatment in human patients with milder mutations in TRα1 resulting in a decreased instead of a lack of T3 affinity.
Normal TH metabolism and action require adequate cellular TH signalling. The last decade has witnessed the identification of several novel syndromes resulting from defects in TH transport, deiodination and receptor function. Because the clinical consequences can be severe, it is of utmost important to develop adequate treatment strategies for these different clinical syndromes. These important recent discoveries may hold the promise that the route of TH signalling will turn out to be even more exciting in the near future.
WEV is supported by a Erasmus University Fellowship 2010 and RPP by a Erasmus Medical Center fellowship and a ZonMW-VENI and TOP grant from The Netherlands Organisation for Health Research and Development (016.096.017 and 921.12.044).