Gain of Function Mutations of the cAMP Regulatory Cascade
For a hormone receptor, gain-of-function may have three different meanings: activation in the absence of ligand (constitutivity), or increased sensitivity to its normal agonist, or broadening of its specificity. When the receptor is part of a chemostat, as is the case for the TSHr, the first situation is expected to lead to tissue “autonomy,” whereas the second is expected to cause adjustment of the agonist concentration to a lower value. In the third case, inappropriate stimulation of the gland is expected to occur because the promiscuous agonist is not expected to be subjected to the normal negative feedback. If a gain-of-function mutation of the first category occurs in a single cell expressing normally the receptor (somatic mutation), it will only become symptomatic if the regulatory cascade controlled by the receptor is mitogenic in this particular cell type. Autonomous activity of the receptor will cause clonal expansion of the mutated cell. If the regulatory cascade also activates function, the resulting tumor will progressively take over the function of the normal tissue, leading ultimately to autonomous hyperfunction. If the mutation is present in all cells of an organism (germ line mutation), autonomy will be displayed by the whole tissue expressing the receptor. In cases where the regulatory cascade is both mitogenic and activates function, the expected result is hyperplasia associated with hyperfunction.
From what we know of thyroid cell physiology (see “Thyroid Physiology and Metabolism” above) it is easy to predict the phenotypes that would result from gain-of-function of the cAMP-dependent regulatory cascade. Two observations provide pertinent models of this situation. Transgenic mice made to express the adenosine A2a receptor in their thyroid glands display severe hyperthyroidism associated with thyroid hyperplasia.21 As the A2a adenosine receptor is coupled to the G protein Gαs and displays constitutive activity due to its continuous stimulation by ambient adenosine,22 this model mimics closely the situation expected for a gain-of-function germ line mutation of the TSHr. Patients with the McCune-Albright syndrome are mosaic for mutations in Gs (Gsp mutations), leading to the constitutive stimulation of adenylyl-cyclase.23 Hyperfunctioning thyroid adenomas develop in these patients from cells harboring the mutation, making them a model for gain-of-function somatic mutations of the TSHr. A transgenic model in which Gsp mutations are targeted for expression in the mouse thyroid gland has been constructed24 and exhibits a similar, but milder, phenotype.
TSHr belongs to the large family of G protein-coupled receptors (GPCRs).12,25– 27 More precisely, it is one of the three glycoprotein hormone receptors (LH/CGr, FSH receptor (FSHr), and TSHr) that are structurally and evolutionary related by their serpentine portion to the large subfamily of opsin-related GPCRs. The primary structure of the TSHr has been deduced from the sequence of the cDNA: it is composed of a serpentine C-terminal moiety typical of GPCRs that is encoded in a single exon (exon 10).28 In contrast to other GPCRs, but in common with the LH/CGr and FSHr, the TSHr has a long (398 residues) N-terminal extracellular domain encoded in multiple exons (1 to 9, for the TSHr).28 This domain has sequence similarity with proteins containing leucine repeats, the prototype of which is the ribonuclease inhibitor.29,30 The TSHr gene is located on chromosome 14q24.28 Amongst GPCRs, the TSHr has the unique characteristic to undergo posttranslational cleavage within a segment close to the border between the leucine-repeat-containing domain and the beginning of the serpentine domain.11,31,32 This segment has no counterpart in the LH/CG and FSH receptors. A still unidentified metalloprotease is involved in the cleavage, which does not necessarily involve all molecules at the cell surface. Cleavage probably takes place around Ser 314 and results in further removal of amino acid residues from the N-terminal end of the serpentine segment.32 The two domains remain connected by disulfide bonds. The functional significance of this proteolytic step is still unknown.
The modular structure of the glycoprotein hormone receptors reflects a dichotomy in their structure-function relationships. The extracellular N-terminal domains are responsible for high affinity binding of the hormones, the serpentine portion being the “effector” transducing the binding signal to the G protein.26,33– 36 For GPCRs of the opsin family, the serpentine molecule itself carries out ligand binding and activation of G proteins. For the glycoprotein hormone receptors, the observation that the N-terminal domain made of leucine repeats is enough to bind the hormone with high affinity,36 opens the question of how the binding signal is translated into a conformational change in the serpentine portion. Spontaneous mutants of the TSHr may help answering this question (see “Structure-Function Relationships,” below).
The modular structure of the glycoprotein hormone receptors appeared very early in evolution. Receptors with a similar structure and highly significant sequence similarity have been cloned from cnidarians (Anthopleura elegantissima)37 and Drosophila.38 Very recently, sequences with the potential to encode orphan GPCRs displaying extensive similarity with glycoprotein hormone receptors have been identified.39 Their functions remain totally unknown.
The detailed tridimensional structure of the GPCRs is still unknown. Models of the serpentine portion have been proposed for the glycoprotein hormone receptors.40 The recent availability of the tridimensional structure of the porcine ribonuclease inhibitor30 has provided a starting point to model the leucine repeat portion of the extracellular domain of the TSHr.41 This consists in a sector of a doughnut, with the concave and convex surfaces made of β sheets and α helices, respectively. Although they provide a potentially helpful basis for the interpretation of experimental observations, only the future will tell whether present models are close to reality.
The major cause of hyperthyroidism in adults is Graves disease in which an autoimmune reaction is mounted against the thyroid gland and thyroid stimulating antibodies are produced (TSAb) that recognize and stimulate the TSHr. The mechanisms leading to autoimmune thyroid diseases are still unknown, but genetic factors are certainly involved as familial clustering is frequently observed (see “Multigenic-Multifactorial Thyroid Diseases,” below). This may explain why the initial description by the group of Leclère of a family showing segregation of thyrotoxicosis as an autosomal dominant trait in the absence of signs of autoimmunity was met with skepticism.42 Reinvestigation of this family, together with another family from Reims (France), identified two mutations of the TSHr gene, which segregated in perfect linkage with the disease.43 A series of additional families have been studied since and, surprisingly, each showed a different mutation of the TSHr gene.44,45 Functional characterization of these mutant receptors confirms that they are constitutively stimulated (see “Structure-Function Relationships,” below). This new nosological entity, sometimes called Leclère disease, has these clinical characteristics: autosomal dominant transmission; hyperthyroidism with a variable age of onset (from infancy to adulthood, even within a given family); hyperplastic goiter of variable size, but with a steady growth; and absence of clinical or biologic stigmata of autoimmunity. An observation common to the cases described to date is the need for drastic ablative therapy (surgery or radioiodine) in order to control the disease, once the patient has become hyperthyroid. Incomplete thyroidectomies are indeed followed by recurrences. The autonomous nature of the thyroid tissue from these patients has been elegantly demonstrated by grafting it in nude mice.46 Contrary to tissue from Graves disease patients, HTTH cells continue to grow in the absence of stimulation by TSH or TSAb.
The prevalence of HTTH is difficult to estimate. It is likely that many cases have been (and still are) mistaken for Graves disease. This may be explained by the relative insensitivity and lack of specificity of TSAb assays, together with the high frequency of the other thyroid autoantibodies (antithyroglobulin, antithyroperoxidase) in the general population. It is expected that wider knowledge of the existence of the disease will lead to better diagnosis. Presymptomatic diagnosis in children of affected families may prevent the developmental or psychologic complications associated with infantile or juvenile hyperthyroidism.
Cases with toxic thyroid hyperplasia have been described in children born from unaffected parents.47– 52 Conspicuously, congenital hyperthyroidism was present in most cases and required aggressive treatment. Mutations of one TSHr allele were identified in the children, but were absent in the parents. Because mini- or microsatellite testing confirmed paternity, these cases qualify as true neomutations. When comparing the amino acid substitutions implicated in hereditary and sporadic cases, the majority of the substitutions do not overlap (Fig. 158-5). Whereas most of the sporadic cases harbor mutations that are also found in toxic adenomas, most of the hereditary cases have “private” mutations. The analysis of the functional characteristics of the individual mutant receptors in COS cells, and the clinical course of individual patients, suggest an explanation for this observation: “sporadic” mutations seem to have a much stronger activating effect than “hereditary” mutations. From their severe phenotypes, it is likely that newborns with neomutations would not have survived if not treated efficiently. On the contrary, from inspection of the available pedigrees, it seems that the milder phenotype of patients with “hereditary” mutations has only limited effect on reproductive fitness. The fact that hereditary mutations are rarely observed in toxic adenomas is compatible with the suggestion that they would cause extremely slow tissue growth and, accordingly, would rarely cause thyrotoxicosis. If this explanation holds true, one may predict that mutations of the hereditary type may be found in the older patients with toxic adenomas.
Representation of the TSH receptor. Panel A: The locations of known activating mutations are indicated. Panel B: The nature of the mutations is indicated with their origins (somatic, germ line sporadic, germline familial), effects on intracellular regulatory cascades, and references.
Somatic Mutations: Autonomous Toxic Adenomas.
Somatic gain-of-function mutations of the cAMP regulatory pathway have played an important role in the identification of their germinal counterparts. As such, and for their importance in delineating structure-function relationships of the TSHr, they deserve a brief mention.
Soon after mutations of Gsα had been found in adenomas of the pituitary somatotrophs, similar mutations (also called Gsp mutations) were found in some toxic thyroid adenomas and follicular carcinomas.53 The mutated residues (Arg201, Glu227) are homologous to mutations found in the ras proto-oncogenes; that is, the mutations decrease the endogenous GTPase activity of the G protein, resulting in a constitutively active molecule.
The demonstration by site-directed mutagenesis that GPCRs can be constitutively activated by amino acid substitutions in the third cytoplasmic loop54– 56 led to a search for similar mutations in the TSHr.57,58 Toxic adenomas were found to be a fruitful source of somatic mutations activating the TSHr, probably because the phenotype is very conspicuous and easy to diagnose. Most of the mutations are located in the third cytoplasmic loop or in the adjacent sixth transmembrane segment of the receptor (see Fig. 158-5). The clustering reflects the pivotal role of this portion of the molecule in the activation mechanisms.59 However amino acid substitutions were found over most of the serpentine portion of the receptor,57,58,60– 63 and even in the extracellular N-terminal domain.64
Despite disagreement about the prevalence of TSHr mutations in toxic adenomas, which may be due to the different origin of patients65,66 or to different sensitivity of the methodology, we conclude that in countries with a moderate shortage of iodine, activating mutations of the TSHr are the major cause of solitary toxic adenomas.63,67 In some patients with a multinodular goiter and two anatomically distinct areas of autonomy, a different mutation of the TSHr was identified in each nodule.64 Finally, the involvement of TSHr mutations in thyroid cancers has been implicated in a limited proportion of follicular thyroid carcinoma selected for their high basal activity of adenylylcyclase.68,69
Structure-Function Relationships of the TSHr as Deduced from Activating Mutations.
The majority of the activating mutations of the TSHr have been studied by transient expression in COS cells. By the built-in amplification of the transfected construct, it is possible to detect even slight increases in the constitutive activity of the TSHr. An important observation has been that the wild-type receptor itself displays significant constitutive activity.43,70 This characteristic is not unique to the TSHr71– 73 but, interestingly, it is not shared by its close relative the LH/CG receptor.74 The effect of activating mutations must, accordingly, be interpreted in terms of increase in constitutive activity.
Most amino acid substitutions found in toxic adenomas and/or toxic thyroid hyperplasia share common characteristics: (a) they increase the constitutive activity of the receptor towards stimulation of adenylylcyclase; (b) with a few notable exceptions,61 they do not display constitutive activity towards the inositolphosphate/diacylglycerol pathway; (c) their expression at the cell surface is decreased to a variable degree; (d) most, but not all of them, keep responding to TSH for stimulation of cAMP and inositolphosphate generation, with a tendency to do so with decreased EC50; and (e) they bind TSH with an apparent affinity higher than the wild-type receptor.
There is no simple relationship between the position of the mutations or the nature of the amino acid substitution and their functional characteristics. All mutations so far identified in transmembrane segments II, III, VI, and VII and the third cytoplasmic loop have similar phenotypes; they affect amino acids belonging to all classes (charged, polar, and hydrophobic), with substitutions not necessarily involving a shift to another class. Mutations of Ile486 and Ile568 in the first and second extracellular loops, respectively, and Pro639 in transmembrane segment VI are exceptional in that, in addition to stimulating adenylylcyclase, they cause constitutive activation of the inositolphosphate pathway.61 Three additional mutations deserve special mention because of their unexpected nature or location: the four-amino-acid deletion (residues 658 to 661) in the third extracellular loop;67 the nine-amino-acid deletion in the third intracellular loop;62 and the substitutions at Ser281 in the N-terminal extracellular domain.75
There is no direct relation between the level of cAMP achieved by different mutant TSHrs in transfected COS cells and their level of expression at the cell membrane.76 This means that individual mutants have widely different specific constitutive activity (cAMP accumulation/receptor number at the cell surface ratio). While this specific activity may tell us something about the mechanisms of receptor activation, it is not a measure of the actual phenotypic effect of the mutation in vivo, which depends on both the specific activity of individual receptors and their number. Indeed, C672Y, a relatively mild mutation, observed up to now only in an HTTH family, is among the strongest according to this measurement. It would be logical to expect the best correlation to be found between the phenotype and the actual level of cAMP achieved, irrespective of the level of receptor expression. However, differences between the effects of the mutants in COS cells and thyrocytes in vivo may render these correlations a futile exercise. According to a current model for GPCR activation, the TSHr would exist under at least two interconverting conformations: R (silent conformation) and R* (the active forms).55 The unliganded receptor would shuttle between both forms, the equilibrium being in favor of R. Binding of the ligand, to the slit between the transmembrane segments (for biogenic amines) and/or residues of the N-terminal segment or extracellular loops (for neuropeptides), is believed to stabilize the R* conformation. The resulting R to R* transition is supposed to involve a conformational change modifying the relative position of transmembrane helices. In turn, this would translate into conformational changes of the cytoplasmic domains interacting with trimeric G proteins. Seminal studies with the α1b-adrenergic receptor have shown that a variety of amino acid substitutions in the C-terminal portion of the third intracellular loop led to their constitutive activation.54 The observation that all amino acid substitutions at Ala293 were effective in activating the receptor led to the concept that the silent form of GPCRs would be submitted to a structural constraint, requiring the wild-type primary structure of the third intracellular loop. This constraint could be released by a wide spectrum of amino acid substitutions in this segment.54,55,77,78
The observation that amino acid substitutions in a large number of residues scattered over the serpentine portion of the TSHr causes an increase in its constitutive activity is fully compatible with the above model, and provides arguments for its extension. That mutations in residues distributed over most of the serpentine portion of the receptor are equally effective in activating it (which does not seem to be a general characteristic in all GPCRs) suggests that the unliganded TSHr might be less constrained than others. The readily measurable constitutive activity of the wild-type receptor is compatible with this contention. Being already “noisy,” the TSHr would be more prone to further destabilization by a variety of mutations.
The precise effects of individual mutations in structural terms are difficult to predict: the sixth transmembrane segment and its continuation in the C-terminal portion of the third cytoplasmic loop is clearly a hot region (Fig. 158-5), with at least 20 residues potentially implicated in keeping the receptor inactive. That consecutive residues in transmembrane helix VI are implicated (residues 631, 632, and 633) is against a simple model in which activation would result from the rupture of an interaction with specific residues in another transmembrane helix (e.g., TM III). Nevertheless, it is likely that the common consequence of activating mutations is the stabilization of a conformation of the serpentine with individual helices in a different relative position. The identification of activating amino acid substitutions in TM II, III, and VII also fits well with this notion. The deletion of 9 residues in the third cytoplasmic loop is believed to activate the receptor by facilitating binding of Gsa to portions of the transmembrane domains.62
The activating mutations identified in the extracellular loops and the N-terminal extracellular domain are more difficult to interpret in the lights of a simple model based on constraint involving only the serpentine portion of the receptor. They are compatible with an extension of this model, in which the unliganded N-terminal domain would contribute to the constraint keeping silent the serpentine portion. According to this model activation would result from the release of a silencing interaction between the extracellular loops and the N-terminal domain.
Familial Gestational Hyperthyroidism (MIM 603373).
Some degree of stimulation of the thyroid gland by human chorionic gonadotrophin (hCG) is commonly observed during early pregnancy. It is usually responsible for the decrease in serum TSH with relative increases in the free T4 concentration that, nevertheless, remains within the normal range.79,80 When the concentrations of hCG are abnormally high, as in molar pregnancy, true hyperthyroidism may ensue. The pathophysiological mechanism is believed to be the promiscuous stimulation of the TSHr by excess hCG, as suggested by the rough correlation between serum hCG and free T4 concentrations.81,82 A convincing rationale is provided by the close structural relationships of the glycoprotein hormones and their receptors, respectively.8
A new syndrome was described in 1998, in a family with dominant transmission of hyperthyroidism limited to pregnancy.83 The propositus and her mother had severe thyrotoxicosis together with hyperemesis gravidarum during the course of each of their pregnancies. When not pregnant they were euthyroid clinically and by laboratory tests. Both patients were heterozygous for a K183R mutation in the extracellular N-terminal domain of the TSHr. When tested by transient transfection in COS cells, the mutant receptor responded normally to TSH. However, it showed higher sensitivity to stimulation by hCG, when compared with wild-type TSHr, providing a convincing explanation to the phenotype.
The amino acid substitution responsible for the promiscuous stimulation of the TSHr by hCG is surprisingly conservative. Also surprising is the observation that residue 183 is a lysine in both the TSH and LH/CG receptors. When placed on the available three-dimensional model of the hormone-binding domain of the TSHr,41 residue 183 belongs to one of the beta sheets that constitute the putative surface of interaction with the hormones.84 It is likely that an arginine in position 183 would confer a slight increase in stability to the illegitimate hCG-TSHr complex.85 This would be enough to cause signal transduction by the hCG concentrations achieved in pregnancy, but not by the LH concentrations observed after menopause. Indeed, the mother of the propositus remained euthyroid after menopause. This is compatible with a relatively modest gain-of-function of the K183R mutant for stimulation by hCG.
Contrary to other mammals, human and primates rely on chorionic gonadotropin for maintenance of corpus luteum in early pregnancy.86 The frequent partial suppression of TSH observed at peak hCG levels during normal pregnancy indicates that evolution has selected physiological mechanisms operating very close to the border of thyrotoxicosis. This may provide a rationale to the observation that, in comparison to other species, the glycoprotein hormones of primates display a lower biological activity due to positive selection by evolution of specific amino acid substitutions in their alpha subunits.8 If this reasoning is correct, it is likely that further cases of hereditary gestational thyrotoxicosis will be identified with mutations in the alpha or beta subunits of hCG of the fetus.
Loss-of-Function: From Hypothyroidism with Congenital Goiter to Thyroid Agenesis
Heritable or congenital situations with loss-of-function of the thyroid gland due to intrinsic molecular alterations expand over a broad spectrum of diseases. In an attempt to classify them on a pathophysiological basis, they are usually subdivided into diseases with or without goiter (Fig. 158-6). In the former, the loss-of-function affects one of the many steps leading to thyroid hormone synthesis (hormonogenesis defects). The hypothalamic-pituitary-thyroid chemostat responds to a reduction in thyroid hormone level by increasing TSH release, which overstimulates the growth of a partially or completely nonfunctional gland. The resulting goiter may be present at birth, or develop progressively, depending on the severity of the metabolic block, the efficiency of the maternal supply of hormones during fetal life and the iodine supply from the diet. Since the introduction of routine neonatal screening for congenital hypothyroidism in the mid-1970s, most neonates with a high TSH are detected and treated early, so that a goiter may never develop. This makes sometimes difficult the distinction between hormonogenesis defects and the other category of loss-of-function thyroid diseases in which, even if not treated, a goiter would never develop. The latter are grouped under the term dysembryogenesis or dysgenesis, reflecting the belief that the underlying cause is an alteration of the program required for the normal development and function of the gland. Data collected during the last 5 years suggest that nongoitrous loss-of-function of the thyroid is a heterogenous group of diseases in which the common characteristic is diminished or absent growth of the thyroid gland or its anlage due to a faulty stimulus or abnormal responses to normal stimuli. If this happens during embryonic life, when the normal stimuli for growth and differentiation are poorly characterized, it may lead to thyroid agenesis, ectopy, or hypoplasia. If it occurs later, it results in a normally located (eutopic) gland, hyporesponsive to TSH, the physiological thyroid stimulator for growth and maintenance of differentiation during fetal and postnatal life. Absence or structural alterations of TSH itself, is dealt with under a separate heading.
Classification of the various causes of congenital hypothyroidism.
Dyshormonogenesis–Congenital Hypothyroidism with Goiter.
In agreement with the autosomal location of the pertinent genes and the functional reserve observed in the majority of metabolic pathways, all inheritable dyshormonogenic disorders of the thyroid gland segregate as autosomal recessive traits. This also means that no dominant negative mechanisms are at work, suggesting that the protein/enzymes involved are abundant and behave functionally as monomers even if, as for the thyroglobulin molecule, they may actually be homodimers. In the past, many cases presented with a goiter and variable degrees of mental retardation (cretinism). Nowadays, in developed countries, most cases are detected neonatally and, when treated adequately, do not show goiter and have normal intellectual quotients (IQ).87
The quasi-uniform presentation of neonates with congenital hypothyroidism (screened as positive on the basis of an elevated TSH) makes it sometimes difficult to achieve a correct molecular diagnosis. Although this does not affect the choice of the treatment (immediate hormone administration), or its efficacy, it does make it impossible to subsequently identify fetuses at risk in the pertinent families. In spite of the success of the neonatal screening programs, it is certainly advisable to know in advance whether a newborn is affected or not, in order to initiate treatment without any delay. Figure 158-7 proposes an algorithm which may help identifying the defect when confronted with an isolated case of congenital hypothyroidism.
Conceptual algorithm for the exploration of cases with congenital hypothyroidism. The figure should be considered a help in approaching diagnosis, rather than a rigid decision tree.
Key elements of the clinical investigation of a newborn with congenital hypothyroidism are scintigraphy, echography, and the perchlorate discharge test. Scintigraphy will indicate whether (partially) functional thyroid tissue is present or not and, if positive, will anatomically localize the tissue. If scintigraphy is negative, echography may allow identification of putative thyroid tissue, whether in place or not. It must be stressed, however, that echography of the newborn neck may be difficult to interpret and requires proper equipment and an experienced echographist. A helpful addition to these image analyses is measurement of circulating thyroglobulin (Tg). If positive, even in the absence of demonstrable thyroid tissue by scintigraphy or echography, it excludes the diagnosis of athyreosis. Negative scintigraphy with a normally located thyroid is compatible with defective iodide trapping or resistance to TSH. In case of a positive scintigraphy, the perchlorate discharge test enables differentiation between organification defects and a putative Tg or deiodinase defect. The test is based on the observation that administered radioiodide is almost quantitatively and immediately organified by being covalently linked to thyroglobulin or to other intrathyroidal proteins. If organification is defective, retention of radioiodide in the gland depends on the continuous activity of the Na-I symporter (NIS). In these cases, inhibition of NIS by administration of perchlorate 2 h after that of radioiodide results in discharge of a large proportion of the radioactivity from the gland. The test is considered positive if more than 10 percent of the radioactivity is released within a 60 min period and the defect is complete when >90 percent is discharged. While a negative perchlorate test indicates that the organification machinery is working, it does not rule out qualitative or quantitative defects in Tg production because the misfolded Tg molecule, or other proteins including serum albumin, become iodinated. A positive perchlorate discharge test is diagnostic of an organification defect; this may be due to defects in thyroperoxidase (TPO) activity, H2O2 generation, or apical transport of iodide.
If not treated adequately, patients with dyshormonogenesis disorders will develop goiters that may become extremely large. Initially a homogenous hyperplasia, the goiter evolves towards nodularity and malignant transformation of some of the nodules could occur.88,89
The structure of the protein and its gene were described in much detail in the previous edition of this work.1,2 A summary, including new data, is presented here (see also Fig. 158-8). Tg is synthesized as a homodimeric glycoprotein of 2 × 330,000 daltons.1,2 The primary structure of human Tg monomers was recently revised and a compilation of polymorphisms detectable at the cDNA level provided.90 The molecule comprises 2750 amino acids. Of 20 sites for N-glycosylation, 16 are actually used,91 resulting in a protein with about 10 percent of its weight accounted for by carbohydrates. Tg is also subjected to additional posttranslational modifications whose significance is not clear; it is sulfated and phosphorylated.92,93 In addition to its obvious role as the template for thyroid hormone synthesis, Tg has the important function of storing hormones in an inactive form (as an integral part of its polypeptide backbone) as well as iodine (as iodotyrosine residues of the protein). This is achieved within the colloid of the follicular lumen, where Tg concentration reaches extremely high values (100 to 400 mg/ml). The necessity to remain soluble at such concentrations imposes additional constrains on the molecule that may be related to some of the posttranslational modifications, including sulfation. In addition to soluble 19S Tg dimers, the follicular lumen contains molecules with a higher iodine content constituting the 27S tetramers1,2 and insoluble Tg.94 The metabolism of these insoluble, highly iodinated molecules is poorly understood, but the possibility that they can be degraded, and their iodine released by the TPO/H2O2-generating system has been proposed.95
Representation of thyroglobulin monomer, with indication (panel A) of structural repetition motifs and major hormonogenic tyrosines. Panel B: List of known mutations with references.
Under condition of normal iodine availability, the yield of T4 per dimeric Tg molecule is close to 2.1,2 The tyrosine residues implicated in the hormonogenesis reactions are the first to be iodinated at low iodide concentrations. The specificity for this choice is encoded within Tg primary structure, rather that in TPO.96 They are of two different types: the donor monoiodo- or diiodo-tyrosines, as their names indicate, give their iodinated phenolic ring to acceptor diiodotyrosine to yield the triiodo- or tetraiodo-thyronines, respectively. The location of the two kinds of tyrosine residues within Tg primary structure is still debated. In human Tg, tyrosines 130, 847, and 1488 have been proposed as donor sites and positions 5, 1291, 2554, and 2747 as acceptor sites.97 The consensus that a major acceptor site is tyrosine 5 is not disputed. The corresponding major donor tyrosine is probably tyrosine 130. The study of patients or animals with mutations identified in the Tg primary structure has helped locate the functionally important sites (see “Tg Gene Defects,” below). The other residues with putative acceptor or donor functions are distributed over the molecule, with some clustering close to the C-terminus. Additional sites have been suggested by studies with animal Tgs. The precise identification of all hormonogenic residues in human Tg will require analysis by mass spectrometry, as was recently initiated for the bovine molecule.98 When iodinated Tg is isolated from normal glands, a fraction of its subunits is found cleaved, generating N-terminal fragments that remain attached to the molecule by disulfide bonds. Arguments have been proposed that link this posttranslational modification to the hormonogenic reaction.99
The Tg gene is located in 8q24.2-24.3. It spans more than 300 kb and is subdivided in at least 48 exons.100 It is an example of a large gene made of a limited number of building blocks encoding repeated motifs or entire copies of ancestral genes. Three different motifs, encoded each in multiple exons, are repeated respectively 10 (type I), 2 (type II), and 5 times (type III) to constitute about 80 percent of the polypeptide. The type I motif is found in a large family of proteins.101 The remaining 20 percent is made of an entire copy of an acetylcholinesterase homolog, with the intron-exon junctions perfectly conserved. A putative role for the type I repeated motif has recently been proposed. It is based on the observation that a similar motif is endowed with potent cysteine-protease inhibitor (ECI) activity in a protein from salmon egg.102 Accordingly, the 10 type I motifs would protect Tg molecules with a low iodine content from being degraded, and allow their recycling.
Tg Gene Defects (MIM 188450).
Mutations in the Tg gene have been suspected in a large number of cases during the precloning era. Their detailed clinical descriptions can be found in the previous editions of this book1,2 and in a review.103 Here is a brief outline of common clinical and laboratory findings in such patients.
The patient should be hypothyroid or show a compensated hypothyroidism with elevated TSH, trap iodide excessively, have no MIT or DIT in the urine, and usually a negative perchlorate discharge test. When thyroid tissue has been analyzed, some Tg-related antigens are almost invariably found in the glands, and ultrastructural evidence for defective secretion or folding of Tg is frequently present in the form of overdistension of rough endoplasmic reticulum (RER). Findings of abnormal iodoproteins in the plasma and of abnormal low-molecular-weight iodinated compounds in the urine complete the picture.104 Urinary iodohistidine105 is diagnostic of iodination of protein material unrelated to Tg within the thyroid gland. Albumin is most likely the substrate for iodohistidine production, as well as for the abnormal iodinated protein found in the blood. However, iodination of albumin is very nonspecific, being observed in a series of situations leading to goiter formation.
In the more recent years, a limited number of cases have been described with their mutations identified (see Fig. 158-8). The phenotype varies in severity, depending on the location of the mutation in Tg primary structure, and on the iodine supply in the diet. A characteristic, common to most, but not all cases, where it has been studied, is the decrease in circulating Tg as measured by conventional immunoassays. This decrease may be absolute or relative, taking into consideration the overstimulation of the gland by TSH. In a few cases, stimulation by exogenous TSH was performed and proposed as a diagnostic procedure.103 It must be understood that the results from such kind of measurements are highly variable and depend on the nature of the mutation (missense, truncation by nonsense or frameshift, amputation by missplicing, and so forth) and the characteristics of the antibodies used in the immunoassay.
Considering that no prevalent mutation has been identified to date and that there is no indication that frequent mutations do exist, except for the preferential implication of CpG dinucleotides, there is no point describing in details the individual studies. In a few patients, however, the location of the mutations may tell us something about structure-function relationships of this large molecule. This was the case with the first mutation identified in man.106 A transversion of C to G at position −3 of the acceptor splice site of intron 3 of the Tg gene caused skipping of exon 4. Because amputation of exon 4 does not affect the reading frame, this gave a strong argument in favor of Tyr130 (encoded in exon 4) being a important hormonogenic residue.107,108 Although there is still room for debate regarding the respective roles of Tyr5 and Tyr130,109 independent biochemical110 and genetic studies have since consolidated this conclusion.
Detailed investigation of a large Brazilian kindred,111 together with studies of the cog/cog mouse112 lead to the conclusion that mutations affecting the Tg gene may frequently be the cause of endoplasmic reticulum storage disease (ERSD), where the misfolded protein remains trapped in the RER in association with molecular chaperones (GRP94, BiP).113
Animal models of Tg gene defects were available before the era of homologous recombination and knockout mice. They have been identified in a variety of species including the merino sheep, the Africander cattle, the Dutch goat, the Bongo antelope, and the cog/cog mouse. The phenotypes are quite variable. In the Africander cattle, animals are euthyroid but display a huge goiter. The mutation creates a nonsense codon at position 697 in exon 9. Detailed analysis of this model has revealed the first case of a “salvage” phenomenon by which alternative splicing removes the mutated exon from a proportion of the mature transcripts.114,115 A similar phenomenon has been demonstrated in some of the human cases.116 The Dutch goat is spontaneously hypothyroid except if supplemented with excess iodine in the diet. The mutation creates also a nonsense codon at position 296 in exon 8, the result being production of a 35-kDa N-terminal peptide. These two models provide strong support to the notion that the extreme N-terminal portion of Tg contains the major hormonogenic domain. When iodine in the diet is normal or high, the animals achieve euthyroidism despite the lower amounts of handicapped Tg present in their glands. Indeed, in both cases, the amounts of Tg mRNA, and of Tg antigens in the glands were much reduced, reflecting the destabilization associated with the presence of the nonsense mutation. The mutation of the cog/cog mouse has recently been identified.112 It involves Leucine 2263 of the acetylcholinesterase-like domain, a residue highly conserved in the family of molecules containing the “α/β hydrolase fold.”117 The resulting phenotype in homozygous mice is trapping of Tg molecules in the RER. This identifies the acetylcholinesterase-like segment of Tg as an important structural motif, which has been implicated in the normal trafficking and dimerization of this family of molecules.118
Mutation of the Tg gene has been proposed as a cause of familial euthyroid goiter transmitted in an autosomal dominant manner.119 The hypothesis is tempting and agrees with predictions from the studies of Kim and Arvan.113 Misfolded Tg molecules translated from the mutated allele would interfere with routing through the ER of both the mutated and part of the normal molecules. Considering the large functional reserve of the hormonogenic machinery, the defect would become symptomatic only in areas of limited iodine supply. This notion will require independent confirmation before gaining recognition as a cause of familial goiter.
Active iodide (I−) transport is the first step in the biosynthesis of thyroid hormone. This process is mediated through NIS, a protein located in the basolateral membrane of the thyroid follicle cell. Furthermore, the process requires energy because I− is transported into the cell against concentration and electrical gradients. NIS couples the energy that is released by the transport of sodium (Na+) downhill to its electrochemical gradient with I− transport, thereby maintaining the I− concentration inside the thyroid cell up to twenty- to fortyfold higher than in serum.17,18 Active I− transport is stimulated by TSH and inhibited, in a dose-dependent manner, by thiocyanate (SCN−), perchlorate (ClO4 −), or ouabain.120 mRNA analyses suggest that the human NIS is not only expressed in the known iodide concentrating tissues, namely, the thyroid, salivary glands, gastric mucosa, choroid plexus and mammary glands, but is also present in the colon, ovary, omentum, and gall bladder.121,122
The mechanism of I− delivery to the colloid, which is the site of thyroid peroxidase (TPO)-catalyzed I− organification, is different. The anion is transported via an I− channel at the apical membrane,123 the molecular nature of which likely corresponds to the protein encoded by the pendrin gene.124
The human NIS gene was mapped to chromosome 19p13.2-p12. It has 15 exons that encode a protein of 643 amino acids with a molecular mass of ~65 kDa.121 Of serpentine structure, NIS has a transmembrane domain believed to consist of 13 segments.125 However, it is possible that there are as few as 12 segments126 or as many as 14. The former structure is similar to the glucose transporter family127 and the latter is akin to the sodium/glucose transporter-1 protein128 with which NIS shares 30 percent homology. Thus, it is uncertain whether the N-terminus of the molecule is located inside or outside the cell. In contrast, the C-terminus, having a potential cAMP-dependent phosphorylation site,126 is located intracellularly. The protein has three potential sites for N-linked glycosylation. Based on the molecular model of a 13-segment transmembrane domain, one glycosylation site is located in the fourth and two are in the last extracellular loop125 (Fig. 158-9).
Structure of the human NIS based on a model predicting a transmembrane domain of 13 segments. The location and nature of known mutations are indicated. (From Pohlenz and Refetoff.479 Used with permission.)
NIS Gene Defects (MIM 601843 and 274400).
The failure of the thyroid gland to actively accumulate I− (“iodide-trapping defect”) under conditions of normal I− supply, produces congenital hypothyroidism. About 40 patients with a phenotype compatible with iodide-trapping defect have been described since 1958, the year this defect was first suspected.129
As with other primary thyroid defects, newborns are brought to medical attention because of a high blood TSH level detected during neonatal screening. In children, attention is drawn by growth retardation and by the appearance of neck enlargement. In adults, frank hypothyroidism, enlargement of the thyroid gland or a thyroid nodule may be the reason for further investigation. The clinical phenotype is, however, quite variable, a phenomenon not simply due to the degree of loss-of-function of the mutant NIS.
The prevalence of iodide-trapping defect is unknown. Because heterozygous individuals do not express the phenotype, NIS gene defects can be detected only when both alleles are affected. Furthermore, under conditions of high I− intake, mutations causing partial loss of NIS function may not be detected even in the homozygous individual. For these reasons, and because congenital hypothyroidism in the absence of positive radionuclide scan is often misdiagnosed as thyroid agenesis, the actual prevalence of NIS gene mutations may be higher than heretofore suspected. The defect appears to be as frequent in males as in females.
Goiters can be large or small, diffuse or nodular. The degree of hypothyroidism is also variable. Severe hypothyroidism with growth and mental retardation has been observed in several instances. More rarely, subjects have been found to be clinically and biochemically euthyroid.129 The reduction of thyroid hormone levels and increase in TSH concentration in serum are usually consistent with the degree of clinical hypothyroidism. When measured, serum thyroglobulin Tg concentrations have been high.130
Histologic examination of thyroid tissue is not helpful in the diagnosis. Specimens obtained from nodules by fine needle aspiration or biopsy are often erroneously interpreted as representing malignancies because of extreme nuclear pleomorphism associated with marked follicular hyperplasia. Such findings can lead to thyroid surgery.
The laboratory hallmarks of iodide-trapping defect are markedly reduced or absent radioactive thyroidal uptake of radioiodide or pertechnetate, reduced I− saliva to plasma ratio (S/P), and restoration of euthyroid state by treatment with pharmacologic doses of iodide (1 to 5 mg/day). Absence of thyroidal radioactive iodide uptake (RAIU) is also typical for thyroid agenesis, a diagnosis erroneously assigned to some patients with iodide-trapping defect, especially when goiter was not present. Furthermore, depending on the timing of measurements following the administration of radioiodide, RAIU values can be low in other forms of congenital thyroid defects. In contrast, a low I− S/P is pathognomonic of the iodide-trapping defect and can be carried out without interruption of thyroid hormone treatment.
The test is based on the observation that all tissues that normally concentrate iodide are affected by the iodide-trapping defect.131 Saliva is collected, preferably without stimulation, over a period of 5 to 10 min, 1 h after the oral administration of 5 μCi of Na125I. At the same time, a venous blood sample is obtained. After defrothing of the saliva and removal of the cell debris by centrifugation, the S/P ratio of radioiodide is determined by counting equal volumes of these fluids in a gamma-scintillation counter. A normal I− S/P is 25 or greater, whereas the salivary glands of affected individuals are not able to concentrate iodide and therefore cannot secrete it in the saliva, resulting in a very low S/P. An I− S/P in the vicinity of 1 is considered to be the consequence of a complete iodide-trapping defect, while an I− S/P of up to 20 is considered to represent a partial defect.
Definitive diagnosis is based on the identification of a mutation in the NIS gene and on the in vitro demonstration of the resulting functional impairment. Six mutations of the NIS gene have been so far identified (Fig. 158-9). The first mutation was reported by Fujiwara et al.,132 9 months after cloning of the human NIS was published. It occurred in an individual born to consanguineous parents that presented as a homozygous missense mutation (T354P). A Brazilian patient with congenital hypothyroidism due to iodide-trapping defect was homozygous for a nonsense mutation in the NIS gene (C272X).133 Another patient with iodide-trapping defect was compound heterozygote for two mutations in the NIS gene (Q267E and Y531X).134 The latter mutation creates a new 3′-splice acceptor site that is used preferentially with the consequence that the mutant NIS has a stop at codon 515, resulting in a protein that lacks 129 amino acids from its C-terminus. Three additional families and two new missense mutations (G93R and G543E) were reported from Japan.135,136 The G93R mutation was identified in a compound heterozygous association with T354P. The latter mutation appears to occur more frequently in the Japanese population because it has been identified in both alleles of affected individuals in five apparently unrelated families.132,135,136 The small amounts of NIS mRNA present in circulating mononuclear cell is sufficient for the study of splice variants of NIS through the synthesis of cDNA.134
When expressed in HEK-293 cells, the mutant NIS354P had no detectable I− transport activity above background, measured 60 min after the addition of I−.132 However, measurement of the initial velocity of iodide uptake, 5 min after the addition of radioiodide to COS-7 cells expressing NIS354P, demonstrated a small difference in specific I− uptake above to the negative, untransfected, control.135 Functional analysis of the NIS mutants 93R, 354P, and 543G transfected into COS-7 showed minimal transport activity, the statistical significance of which has not been determined.136 No iodide transport activity was found in the mutants 267E, 272X, and 531X. None interfered with the function of the cotransfected WT-NIS.
Thyroid tissue from a patient expressing NIS354P showed a marked increase in the mRNA.135 Cells transfected with NIS 267E, 515X, and 272X, express these mutant NIS molecules as effectively as the WT-NIS. While the mutant NIS354P appears to be targeted to the plasma membrane, even though it is virtually devoid of iodide transport activity,137 267E and 515X remain trapped intracellularly.137a
There are treatment options for NIS defects. The reversible consequences of iodide-trapping defect should, by definition, be corrected with by the administration of an excess if I− . Indeed, I− supplementation has been effective in restoring euthyroidism. The recommended dosage is 14 mg of I−/day.138 Treatment with replacement doses of thyroid hormone is also effective and more practical.
TPO is a glycosylated hemoprotein bound to the apical plasma membrane of thyroid follicular cells, with its catalytic domain facing the colloid space.19,139 It corresponds to the major antigenic component of the thyroid “microsomal antigen” implicated in thyroid autoimmunity.140– 142 The primary structure of human thyroid peroxidase has been deduced from cloned cDNA.142– 144 Two cDNAs differing by 171 nucleotides have been obtained; they encode TPO1, the complete polypeptide corresponding to the protein as extracted from the gland, and TPO2, which is translated from an mRNA missing exon 10.143 When expressed by transfection in CHO cells, TPO2 remains trapped in the ER and is devoid of enzymatic activity.145 TPO extracted from thyroid glands displays 105- and 110-kDa protein species upon analysis by western blotting. Once believed to correspond to TPO1 and TPO2, both species are clearly encoded by TPO1 cDNA146 and the functional significance of TPO2 is unknown. In addition to the major 3.2-kb TPO mRNA, shorter species have been demonstrated corresponding to misspliced primary transcripts.147 Their significance, if any, is also unknown. Analysis of hydropathy profiles of the polypeptides encoded by the major messages indicates that the protein is anchored in the membrane by a segment close to its C-terminus. The extracellular domain has similarity with myeloperoxidase and other peroxidases (42 percent identity over 745 residues with myeloperoxidase),142 indicating a common evolutionary origin. Candidate histidine residues (His 239 and 494)148 implicated in the binding of the heme moiety have tentatively been identified and placed on a molecular model elaborated on the basis of the tridimensional structure of myeloperoxidase.149
TPO gene is located on the short arm of chromosome 2 (2p25).150 It contains 17 exons and spans more than 150 kb. Exon 10 is the one subjected to alternative splicing. As is the case for Tg, the promoter of the TPO gene is the target for the three transcription factors TTF1, TTF2, and Pax8.151,151 Pax8 plays a particularly important role because it can transactivate the proximal promoter in the nonthyroid environment of the Hela cells.152 A tissue-specific enhancer, which binds TTF1, has also been identified 5.5 kb upstream.153
TSH increases both TPO enzymatic activity and steady state mRNA level in the thyrocytes of all species studied.154 This effect is mimicked by cAMP analogs and by forskolin, and is likely to be mainly transcriptional.155 In contrast to the regulation of the Tg gene, activation of TPO gene transcription by cAMP is rapid and does not require ongoing protein synthesis.155 The TPO gene thus behaves as a series of other genes under rapid control by cAMP. However, the intermediate steps and all actors of this regulation have not been elucidated.
TPO Gene Defects (MIM 274500).
A comprehensive list of organification defects described since 1950 has been published.156 The first mutation responsible for the phenotype of a patient presenting with typical congenital goiter, organification defect, and hypothyroidism was identified in 1992.104 It is worth describing because it is the only recurrent mutation described to date.104,157,158 It consists in the duplication of a GGCC tetranucleotide in exon 8 of the TPO gene, resulting in the truncation of the normal reading frame. The patient was homozygous for the mutation. Translation of the mutated mRNA would yield a polypeptide less than half the length of normal TPO and devoid of the putative proximal and distal histidine residues implicated in heme binding (see “Thyroperoxidase,” above). Analysis by RT-PCR of the RNA extracted from the goitrous tissue demonstrated accumulation of a misspliced mRNA resulting from the use of a cryptic acceptor splice site in exon 9. The normal reading frame was restored in exon 9 downstream of the alternative splicing, which means that mRNA in the goiter has the potential to encode a near full-length TPO with 51 unrelated residues in its middle. The unmasking of alternative splicing as the result of nonsense or frameshift mutations is not unusual (e.g., see “Tg Gene Defects” above).
A series of additional mutations were subsequently identified (see Fig. 158-10).157– 163 They are distributed along most of the TPO molecule and, as expected for loss-of-function mutations, they comprise missense, nonsense, and frameshift mutations. Of particular interest, two different TPO mutations have been found to segregate within a large inbred pedigree of Amish, which had been screened with the hope of identifying defects in the H2O2-generating system.164 This unusual observation illustrates the strength, as well as possible pitfalls, associated with homozygosity mapping of autosomal recessive mutations in inbred populations. A compilation of polymorphic sites in TPO cDNA has been published.157
Representation of thyroperoxidase (panel A). Panel B: List of known mutations with references.
Pendred Syndrome (PDS; MIM 274600).
Pendred syndrome is a “syndromic” hereditary goiter associated with sensorineural hearing loss and transmitted in an autosomal recessive manner. Patients are usually euthyroid or present with compensated hypothyroidism (i.e., low-normal serum thyroid hormone levels with elevated TSH). Expression is variable, both in terms of thyroid and hearing defects. Deafness is often associated with the Mondini cochlear defect (lack of the normal coiled structure)165 and may be present at birth or develop in childhood.166 As with other hormonogenesis defects, the goiter is quite variable in size ranging from virtual absence (in this case patients present with deafness only) to voluminous masses requiring surgery for tracheal decompression. Diagnosis requires demonstration of sensorineural hearing defect together with a positive perchlorate discharge test. The disease thus qualifies as an organification defect with normal iodide trapping. It is only recently that mutations in a single gene were found to account for the syndrome. The gene was named pendrin after the syndrome.167
Pendrin gene was first localized to chromosome 7q31 by homozygosity mapping in inbred pedigrees.168– 170 A thorough investigation of candidate genes and exons in the interval identified the gene on the basis that it was found mutated in the families under study.167 The pendrin gene is made of 20 exons encoding an mRNA of approximately 5 kb containing a 2343-bp open-reading frame. The conceptual protein of 780 amino acids has strong similarities with a family of sulfate transporters, amongst which is down-regulated in adenoma (DRA),171 whose gene is contiguous to pendrin. Mutations in DRA are responsible for a congenital chloride diarrhea syndrome. A proposed model for sulfate transporters and pendrin includes 11 putative transmembrane segments, with the N- and C-termini in the cytoplasmic and extracellular spaces, respectively (Fig. 158-11A). The homology with sulfate transporters together with a putative role for sulfation of Tg in hormonogenesis (see above) made it tenable that pendrin would function as a sulfate transporter. However, the “sulfate transporter signature” of pendrin (a series of residues shared by the primary structure of sulfate transporters) is not canonical, and a possible role in the transport of iodide at the apical membrane of the thyrocyte is an appealing alternative hypothesis. The first functional studies of pendrin performed in Xenopus oocyte and Sf9 cells did not show evidence for a role in sulfate transport; rather they demonstrate convincingly its ability to transport iodide and chloride.124 Further studies are required to determine whether pendrin is the apical iodide transporter that has been characterized functionally.172,173 The role of pendrin in iodide transport at the apical membrane would provide a satisfactory explanation for its role in thyroid hormone synthesis and a rationale to the positive perchlorate discharge test demonstrated by patients with pendrin gene mutations. Its role in the development of the cochlea is presently less clear. It has been repeatedly proposed that the hearing loss of Pendred patients might be secondary to intrauterine hypothyroidism.1 However, recent evidence points to a direct role of the protein in the development of the inner ear.174
Representation of pendrin according to Everett et al.167 (panel A). Panel B: List of known mutations with 194 references.
Putative loss-of-function mutations have been identified in a series of patients and kindreds.167,175– 177 As expected for loss-of-functions mutations, the mutations are quite diverse in nature (missense, nonsense, frameshift, splicing defects) and scattered along the primary structure of the protein (see Fig. 158-11B). As indicated above, the phenotypes display a broad spectrum of expressivity. Most interesting is the evidence that a significant proportion of deaf people with no clinically obvious thyroid disorder (i.e., without goiter) may have mutations in the pendrin gene. However, the number of cases analyzed in molecular terms is presently too low to draw definitive conclusion.
Defects in iodide oxidation that cannot be explained by abnormalities in thyroperoxidase expression or function could be attributed to the H2O2-generating system. On that basis, they could account for a small but definite proportion of such defects.
That defects in the similar O2 −-generating system of the leucocyte and macrophages in X-linked chronic granulomatous disease are not accompanied by thyroid insufficiency indicates the existence of a distinct system in the thyroid. One element of a system has been cloned, but no mutant has yet been characterized. In the absence of genetic tools, deficiency in H2O2 production has been suggested in cases where restoration of normal organification of iodide is observed after addition of H2O2 to the thyroid homogenate or medium of incubation of thyroid slices.178 This has been demonstrated in one congenital iodide organification defect and in one cold nodule.179
Failure of Iodotyrosine Deiodinase (MIM 274800).
Failure of iodotyrosine deiodinase activity does not interfere directly with thyroid hormone synthesis or secretion. However, it leads to leakage of iodotyrosines into the circulation, with a failure of iodine recirculation in the thyroid and loss of the iodotyrosines in the urine. The iodine in the serum iodotyrosines, of course, cannot be recovered by the thyroid iodide-trapping mechanism. The main consequence of this defect is, therefore, a great loss of iodine, which steps up a vicious circle of thyroid stimulation, hyperplasia, goiter, and increased synthesis and leakage of hormone precursors. The severity of the defect is, therefore, inversely proportional to the iodine content of the diet.
The clinical picture of patients affected by the iodotyrosine deiodinase defect is that of congenital hypothyroidism with goiter. The severity of hypothyroidism and goiter and of their possible consequences depends on the severity of the defect and on the supply of iodine. The decrease in serum T4 and T3 levels and increase in serum TSH will vary accordingly. Radiodiodide or 99TCO4 − uptakes are high and fast, and radioiodine release is rapid.
Trichloroacetic acid treatment of serum precipitates partially the iodotyrosines and thus part of the released radiodione. Chromatography of extracts of the serum and the urine demonstrates radioiodine in the iodotyrosines. Demonstration of the defect is achieved by the in vivo diiodotyrosine test. Radioiodine-labeled DIT is injected intramusculary or intravenously, and urine is collected at 1-h intervals for a few hours. Chromatography of the urine demonstrates less than 5 percent of the radioactivity in the iodotyrosines in normal controls, but up to 80 percent in patients.180– 186 Alternatively, urinary DIT may be measured by radioimmunoassay after its immunoprecipitation from urine. Its secretion is considerably increased in deficient patients (1.2 to 17.7 nM/mM creatine versus 0.108 ± 0.048 in normal individuals).187 Thyroid slices from these patients will not deiodinate labeled DIT; to validate such a finding, positive deiodination by human and/or animal thyroid slices should be demonstrated at the same time. The latter activity should be inhibited by dinitrotyrosine (mM), an inhibitor of the enzyme. Finally, demonstration of the defect and of the proposed pathogenic mechanism requires that iodine medication alone should be sufficient to reestablish euthyroidism.184,188 An isolated defect in the deiodinating capacity of peripheral tissues, but not of the thyroid, should not cause the syndrome, as iodotyrosines as such are released only as traces by the thyroid.
The genetics of this kind of goitrous cretinism was studied by Hutchison and McGirr.189,190 It behaves as a simple autosomal recessive trait. There is no sex prediclection. Carriers of the trait185 are less efficient in deiodinating the DIT. When given 20 to 25 mg of stable DIT along with labeled DIT, the best discrimination was achieved during the first 2 h of excretion of undeiodinated labeled DIT. The relatives of patients excreted an average of 20.4 percent as diiodotyrosine during this time, whereas normal subjects excreted 11.4 percent and the patients with dehalogenase deficiency excreted 52 to 79 percent.
Developmental Defects: Thyroid Dysgenesis.
The mammalian thyroid anlage develops from a group of cells forming a thickening of the floor of the pharynx (embryonic day E8.5 in mouse).191 These cells migrate caudally while multiplying and forming a two-lobe structure to reach their destination on both sides of the trachea (by E13 to E14 in the mouse).191 A complement of cells coming laterally from the neural crest join the thyroid primordium. These cells are believed to contribute the ultimobranchial body in lower species and the C-cells of the adult thyroid. In humans, the median thyroid primordium is visible around day 16 to 17 of gestation and has reached its definitive location in the anterior neck by 6 to 7 weeks.192 Induction and migration of the thyroid rudiment do not seem to be under control of the pituitary gland. Embryos with congenital absence of the anterior pituitary develop a thyroid rudiment,192 as do mutants lacking TSH or the TSHr. Nevertheless, in the absence of stimulation by TSH, the thyroid, while in eutopic position, remains hypoplastic and virtually nonfunctional (see “Resistance to Thyrotropin,” below). Three transcription factors play a key role in the differentiation and growth of the thyroid anlage: thyroid transcription factor 1 (TTF1 or T/ebp or Nkx2.1),151,193 thyroid transcription factor 2 (TTF2 or FKHL15),194,195 and Pax8.196 Their structures and implications in hereditary/congenital hypothyroidism are presented in detail below. Figure 158-12 summarizes their expression pattern in relation with embryonic thyroid development. In the present state of knowledge, TTF1, TTF2, and Pax8 are the earliest markers of thyroid differentiation. They can already be detected at the site of the invagination of the pharynx that will form the thyroid bud (E8.5 in the mouse), and their expression continues uninterrupted thereafter.194 The other markers of thyrocyte differentiation—Tg, TPO and TSHr—are not detectable before the thyroid has reached its final position (E14 to E15 in the mouse). TTF1, together with TTF2 and Pax8 genes, is expressed in the adult thyroid, where its role as a positive regulator of thyroid gland-specific gene expression by TSH and insulin is debated.197,198
Representation of the pattern of expression of genes involved in the control the development of the thyroid during mouse embryogenesis. For each mouse embryonal developmental stage (E7.5 to E15) and newborn, the status of expression of differentiation markers (Tg, TPO, TSHr) and developmental genes (TTF1, TTF2, Pax8) is indicated (+ or −). (Adapted with permission from M. Zanini and R. Di Lauro194 (Naples, Italy).)
Thyroid dysgenesis can be classified into agenesis or athyreosis, hemiagenesis, ectopy, and hypoplasia. When a newborn screens positive based on a high TSH but is found to have no detectable thyroid tissue by both scintigraphy and echography, then the diagnosis of thyroid agenesis or athyreosis should be straightforward. Athyreosis accounts for about 15 percent of newborns with permanent congenital hypothyroidism.199 The differential diagnosis of agenesis versus hypoplasia or even a normal-sized thyroid gland that is unable to concentrate iodide or pertechnetate, depends very much on technical skill and interpretation of scintiscans and echograms. An important additional parameter is the serum level of Tg. When detectable it signs the presence of thyroid tissue. Cases initially misclassified as athyreosis have been correctly diagnosed as hypoplasia on the basis of a normal or even elevated value for serum Tg.200 An undetectable Tg, while TSH values are still high, is compatible with true thyroid agenesis or defective Tg synthesis (see “Tg Gene Defects,” above).
Thyroid ectopy is defined as the presence of thyroid tissue outside its normal location on both sides of the trachea. It accounts for about 60 percent of permanent congenital hypothyroidism.199,201 The tissue is detected by its ability to trap iodide or 99m pertechnetate, which is frequently enhanced secondary to the overstimulation by the high circulating TSH concentrations. It may be found anywhere from the foramen caecum at the base of the tongue (sublingual thyroid), under the chin (submental) to the mediastinum, on the midline, or laterally. In most cases, the tissue is both ectopic and hypoplastic, which accounts for the hypofunction. In some cases, however, ectopy is compatible with euthyroidism or compensated hypothyroidism (elevated TSH, normal thyroid hormone levels). An indication that different pathogenetic mechanisms may be implicated in agenesis, and ectopy is given by the different sex ratio observed in the two situations. A clear excess of females is found in ectopy (3:1),199 while agenesis and hypoplasia seem to affect both sexes equally.199
Thyroid hemiagenesis is a rare situation that also affects females preferentially.202 The mechanism responsible for agenesis more frequently affecting the left lobe (80 percent of cases) is presently unknown.
Thyroid hypoplasia is the diagnosis when echography detects thyroid tissue in its normal location but in insufficient amounts to achieve euthyroidism despite hyperstimulation by TSH. Hypoplasia may be associated with absent or detectable trapping of iodide. In the first instance, identification of the tissue relies solely on echography and may be problematic.
Several studies have addressed the association of congenital hypothyroidism with other developmental defects.199,203,204 Most report an overall increase of a variety of malformations. Except for the well-defined Bamforth syndrome (see “Thyroid Transcription Factor 2,” below) and for an association with Down syndrome, the most common abnormality is cardiac septal defects,205 fivefold higher than in the general population in one study.199 Patients with Down syndrome have a thirty-fivefold increased risk for congenital hypothyroidism.203
The cause of thyroid dysgenesis in the majority of cases is unknown. Most are sporadic, although familial occurrence has been occasionally described. Therefore, except for a minority of patients in who it is clearly transmitted as a mendelian trait (see next three sections), even the genetic nature of the disease may be questioned. It is conceivable, but not demonstrated, that many cases are due to new mutations and that, before the implementation of routine neonatal screening, the affected individuals would display a severe decrease in reproductive fitness. This would lead to misclassification of the disease as sporadic. The situation may soon clarify as the generation of “rescued cretins” enters the reproduction age. Another possibility is that the disease would be multigenic. The discovery that mutated alleles of the Pax8 genes are the cause of thyroid hypoplasia transmitted in an autosomal dominant manner, but with incomplete penetrance would fit with this hypothesis (see Pax8, below).
During the past few years, mutations have been identified in a small proportion of cases with thyroid dysgenesis. They were logically screened for abnormalities in the genes suspected to be implicated in the development or regulation of thyroid function: TTF1, TTF2, PAX8, and the TSHr.
Thyroid Transcription Factor 1 (TTF1, or Titf1, or Nkx2.1; MIM 600635).
TTF1 is the prototype of a subfamily of transcription factors containing a homeobox domain.193,206 It comprises the Drosophila tinman gene and the Nkx family of factors in vertebrates. TTF1, also called Nkx2.1, is expressed in the fetal brain, the pituitary, the developing and adult thyroid gland, and in lung.207,208 The gene is encoded by two exons, on chromosome 14q13.209 In the adult lung and thyroid, it acts as a tissue-specific transcription factor, controlling positively expression of the surfactant proteins in the former and Tg and TPO genes in the latter.206 Multiple target sites for TTF1 binding have been identified in the proximal promoters of the Tg and TPO genes, and in more distant upstream enhancers. Interestingly, a similar arrangement of TTF1 and Pax8 sites has been described in the promoters of both genes. TTF1 gene is probably controlled in turn by another homeobox protein that remains unidentified.206 The role of TTF1 in embryonic development can be appreciated from the analysis of knockout mice.208 While heterozygous animals are completely normal, TTF1 −/− mice die in the neonatal period. They lack thyroid tissue completely, both thyrocytes and C cells being absent. In addition, they suffer from profound hypoplasia of the lungs and display brain malformations. This phenotype is in good agreement with expression of TTF1 in the developing thyroid gland (see above), the lung buds and the fetal forebrain. From this picture, it is expected that severe loss-of-function mutations of TTF1 in humans would be lethal, and that the gross pulmonary and brain malformations would mask the thyroid phenotype. No report fitting the description of the TTF1 −/− mouse has been reported yet in humans, and no mutation of TTF1 has been reported in a series of congenital hypothyroid patients.210,211 One case of compensated hypothyroidism (elevated TSH, normal T4) with eutopic thyroid gland was described in a child with respiratory failure harboring a heterozygotous deletion spanning less than 13 cM on chromosome 14q13 and including the TTF1 locus.212 However, the large size of the monosomic segment, which certainly encompasses other genes (Pax9 gene is included in the deletion) does not allow one to conclude that hemizygosity of TTF1 alone was responsible for the phenotype.
Thyroid Transcription Factor 2 (TTF2, TITF2, or FKHL15); Bamforth Syndrome (MIM 241850).
TTF2 was identified and cloned initially in the mouse on the assumption that a factor resembling HNF3 should be present in thyroid tissue. This hypothesis was based on the identification of cis-acting elements characteristic of the forkhead gene family in the promoters of thyroid-specific genes.194,206 TTF2 gene is encoded by a single exon on chromosome 9q22.195 It is expressed in the developing thyroid, the foregut endoderm, the craniopharyngeal ectoderm and Rathke's pouch.213 The human cDNA was cloned independently from a skin cDNA library and was named FKHL15.195 In addition to its putative role as a repressor of thyroid-specific genes during thyroid embryogenesis and in the development of the thyroid anlage, TTF2 plays a role in the positive control of thyroid-specific gene expression by insulin and possibly by TSH in the adult thyroid gland.214 TTF2 knockout mice provided the phenotype to look for in humans.213 The −/− mice are profoundly hypothyroid with undetectable T4 in their blood. They die shortly after birth, most probably because of an inability to suckle due to severe cleft palate. Two different thyroid phenotypes—complete absence of thyroid tissue and sublingual ectopy—are observed with about equal frequency.213 When studied in the embryo, development of the thyroid gland is completely normal in the heterozygous mice. In −/− embryos, budding of the thyroid primordium takes place normally at day E8 to 8.5. The bud expresses TTF1 and Pax8 normally. Thereafter, no migration of the bud is observed, and at day E11.5, in 50 percent of the mice, there is no more thyroid tissue to be seen. In the other half of the animals, the tissue remains in its original position. Absence of TTF2 is thus mainly characterized by absence of migration of the thyroid primordium, which is associated with hypoplasia or complete involution of the misplaced tissue. As expected, elevated TSH levels accompany hypothyroidism. This indicates that TTF2, while expressed in Rathke's pouch, is not required for normal development of the pituitary gland. In humans, Bamforth et al. described siblings with congenital hypothyroidism, cleft palate, choanal atresia, and kinky hairs.215 Ten years after the original description, a homozygous loss-of-function mutation of TTF2 (A65V) was identified in the same two patients.216 No trace of thyroid tissue could be detected by ultrasonography. The parents were healthy and euthyroid, which classifies Bamforth syndrome as an autosomal recessive cause of congenital hypothyroidism. In agreement with one of the phenotypes of TTF2 −/− mice, it remains to be seen whether cases of Bamforth syndrome will be identified with sublingual ectopic thyroid.
The Pax8 cDNA was cloned originally as a member of the paired domain transcription factors expressed in both the kidney and the thyroid gland.196 Its gene is encoded by at least 10 exons on chromosome 2q12-q14.217,218 Together with TTF1, Pax8 was shown to be implicated in the tissue-specific expression of the thyroid genes Tg and TPO.151,206 More recently, it was shown to play a key role in the embryonic development of the gland. Transgenic mice homozygous for a null Pax8 allele display profound hypoplasia of their thyroid glands, the heterozygotes being morphologically normal.219 The thyroid anlage appears at the normal time and is positive for TTF1 transcripts. Thereafter, while starting to migrate caudally, it fails to grow and endures almost complete involution. Interestingly, and contrary to what happens in TTF1 knockouts, the calcitonin-secreting cells develop normally and reach a position similar to that of the normal thyroid lobes. The thyroids of the Pax8 −/− mice are composed almost completely of C cells. They display expression of TTF1, but show no Tg or TPO.219
In a series of 145 human cases with congenital hypothyroidism and thyroid dysgenesis, 2 sporadic cases and 3 members of 1 family were found to harbor loss-of-function mutations in their Pax8 gene.220 The mutations were shown to affect Pax8 function, as evidenced from transfection experiments: the mutants had lost the ability to transactivate a luciferase reporter gene expressed in Hela cells under the control of the TPO gene promoter.220 Contrary to the situation in the mouse model, the patients were heterozygous for the mutation and in the familial cases, transmission was autosomal dominant. The two sporadic cases were neomutations. The patients displayed variable degrees of thyroid gland dysgenesis with hypoplasia as the common denominator. In the family with three affected individuals, expressivity was extremely variable, ranging from frank hypothyroidism associated with a cystic thyroid rudiment, to compensated hypothyroidism with a thyroid gland of close to normal size. Plasma Tg was detectable or even elevated in some of the cases, which may be due to leakage from an abnormally differentiated gland. A similar observation has been made in thyroid hypoplasia caused by TSHr mutations (see next section). The reason for the difference between transmission in the mouse model and man is not clear. The variability of the phenotypes observed in man, even within the same family, suggests that the genetic background plays an important role in the expression of Pax8 defects. It is likely that recessive transmission in the knockout mice is a peculiarity associated with the genetic background of the animal strain used.
TSH Receptor: Resistance to Thyrotropin (MIM 275200).
Loss-of-function mutations in the TSHr gene are expected to cause a syndrome of resistance to TSH. The expected phenotype is likely to resemble that of patients with mutations in TSH itself (see “Isolated Central Hypothyroidism” below). A mouse model of resistance to TSH is available in the hyt/hyt line. Homozygous hyt/hyt mice are hypothyroid due to a developmental anomaly of their thyroid glands, which remain hypoplastic.221 The cause has been traced to a mutation of the TSHr gene (P556L).221,222 From this information one would expect patients with two mutated alleles to exhibit a degree of hypothyroidism in relation with the extent of the loss-of-function. Heterozygous carriers are expected to be normal or to display minimal increase in plasma TSH.
Clinical cases with documented mutations in the TSH receptor have been identified (Fig. 158-13). A few patients with convincing resistance to TSH had been described before molecular genetics allowed for identification of the mutations.223,224 Another family was described more recently, but no mutation was found in the receptor gene.225 The first cases described in molecular terms were euthyroid sibs with elevated serum TSH levels.226 They were compound heterozygotes for mutations in the extracellular N-terminal portion of the receptor (maternal allele: P162A; paternal allele: I167N). The functional characteristics of the mutant receptors were studied by transient expression in COS cells; the paternal allele was almost completely nonfunctional, whereas the maternal allele displayed an increase in EC50 for stimulation of cAMP production by TSH. The I167N paternal allele is expressed in normal amounts in COS cells, but it remains trapped intracellularly and does not reach the cell surface. When both mutations are displayed on a tentative model of the extracellular domain, their location is compatible with the observed phenotype: the P162A mutation affects a residue predicted to be at the surface of the molecule, which may explain its interference with effects of TSH. The I167N mutation affects a residue protruding within the hydrophobic tunnel between the alpha helices and the beta sheets in the doughnut-shaped model.41 It is expected that a polar residue would be incompatible with such position and result in severe misfolding of the whole extracellular domain. Coexpression in COS cells of the wild-type and mutated receptors did not show evidence for dominant negative effects of the mutants.
Representation of the TSH receptor. Panel A: The locations of known loss-of-function mutations are indicated. Panel B: The natures of the mutations are indicated with the extent of their effects and references.
Familial cases with loss-of-function mutations of the TSHr have been identified in the frame of screening programs for congenital hypothyroidism (Fig. 158-13).200,226– 229 All infants have high serum TSH levels; depending upon the severity of the TSHr defect, however, some are euthyroid,226,229 while others display hypothyroidism of variable magnitude.200,227,228 Also depending upon the severity of the defect, thyroidal uptake of 99TcO4 − can be normal or absent. All thyroids were small and normally located at echography. Surprisingly, in some of the cases, serum Tg levels were high. It is likely that, similar to what is observed in hypoplasia due Pax8 mutations (see above), the hypoplastic glands would suffer a differentiation defect that may have as a consequence the leakage of Tg from abnormal follicles or its misrouting to the basolateral membrane.
As may be expected for loss-of-functions, mutations were diverse and patients were mainly compound heterozygotes, except when consanguinity was present. In one instance, the patients were sibs born from consanguineous parents and were homozygous for a mutation in transmembrane segment IV of the serpentine portion of the receptor (A553T), close to the hyt mutation of the mouse. When transiently expressed in COS cells, the mutants were barely expressed at the cell surface. However, the residual expression was compatible with some TSH binding and stimulation of cAMP production by TSH.200 When the phenotypes of these cases is known in more detail, they will provide a means to understand the role of the receptor on thyroid organogenesis. Indeed, the difference in phenotype between people with mutations abolishing the hormone or the receptor will tell us whether the mere expression of a functional receptor and of its constitutive activity on adenylyl-cyclase stimulation (see “Structure-Function Relationships of the TSHr,” above) plays a role in the development of a structurally normal thyroid gland.
Families with autosomal dominant transmission of apparent partial resistance to TSH where the phenotype does not cosegregate with the TSHr gene have been described.230 In some of these families, linkage was also excluded with candidate genes known to be implicated in the normal development of the thyroid gland (TTF1, TTF2, Pax8) (unpublished results from the authors' laboratories). The underlying pathophysiological mechanism is still unknown.