Mutations in the Pendred Syndrome (PDS/SLC26A) Gene: An Increasingly Complex Phenotypic Spectrum From Goiter to Thyroid Hypoplasia

Pendred syndrome is an autosomal recessive disorder traditionally defined by sensorineural deafness, goiter, and a partial defect in the organification of iodide (OMIM 274600; http://www.ncbi.nlm.nih.gov/omim) (1). The syndromic association of goiter and deafness was first described by the British practitioner Vaughan Pendred in 1896 (2), and the partial iodide organification defect was recognized by Morgans and Trotter (3) in 1958 by submitting affected individuals to a perchlorate discharge test. The molecular cause underlying this disorder, which accounts for about 10% of hereditary hearing loss and is one of most common forms of syndromic deafness, was identified in 1997 through the demonstration of biallelic mutations in the SLC26A4 (solute carrier 26A4) gene (4), which encodes the multifunctional anion exchanger pendrin. Pendrin functions as a coupled electroneutral iodide/chloride, iodide/bicarbonate, and chloride/bicarbonate exchanger with a 1:1 stoichiometry (5). In the thyroid, pendrin is expressed at the apical membrane of follicular cells and is thought to mediate iodide efflux into the follicular lumen, although this step likely involves an additional, unidentified anion transporter or exchanger (6, 7). Clinically, the hallmark of Pendred syndrome is sensorineural hearing impairment. Most patients present with severe prelingual deafness, and others develop progressive hearing impairment later in life. Radiological studies of the inner ear reveal an enlargement of the endolymphatic system, which is most readily detected by documenting an enlarged vestibular aqueduct (EVA) (8). The thyroid enlargement is variable among affected family members, and functionally some individuals are euthyroid, whereas others have (congenital) hypothyroidism. It is thought that the variability of the thyroid phenotype is, at least in part, influenced by the nutritional iodine intake (1). In countries with a high iodine intake, goiter development and thyroid dysfunction are usually not seen in patients with biallelic mutations in the SLC26A4 gene. Without performing molecular analyses and/or a perchlorate discharge test, which is unfortunately poorly standardized and subject to other factors such as iodide intake, it is not possible to clinically distinguish Pendred syndrome from nonsyndromic familial EVA (8). The patients with familial EVA have an identical phenotype at the level of the inner ear but absent thyroid abnormalities, and they have either no mutations or monoallelic mutations in the SLC26A gene, indicating that Pendred syndrome and nonsyndromic EVA are distinct clinical and genetic entities (8). Now, the phenotypic spectrum associated with mutations in the SLC26A4 gene is unfolding with further complexity. In this issue of the JCEM, Kühnen et al (9) report on two patients from distinct families with thyroid hypoplasia and biallelic mutations in the SLC26A4 gene. A heterogeneous group of developmental abnormalities, referred to as thyroid dysgenesis, is found in approximately 85% of all patients with sporadic congenital hypothyroidism. Thyroid dysgenesis includes a spectrum of anomalies: (hemi)agenesis, ectopy, and hypoplasia. In an effort to identify novel genetic defects underlying thyroid dysgenesis, these investigators performed whole exome sequencing of the DNA obtained from a patient with thyroid hypoplasia from a consanguineous family. The investigators detected 760 biallelic sequence variants with the potential to alter the encoded proteins, but only 17 were predicted to have damaging consequences based on prediction programs assessing the functional consequences of these alterations. Surprisingly, this led to the detection of a homozygous mutation in the SLC26A4 gene in the index patient. This alteration (1790T>C; p.L597S) has been previously reported in patients with Pendred syndrome, and functional characterization in Xenopus oocytes documented a loss of function in terms of anion exchange (10). The patient was diagnosed with severe hypothyroidism at the age of 40 days (TSH, >75 mU/L [1.5–4.5]; T4, 0.45 μg/dL [4.5–12.5]). Unfortunately, there are no data on the size of his thyroid in the newborn period, but at the age of 7 years, he was found to have a hypoplastic thyroid with cystic malformations on sonographic evaluation. At the age of 12 years, there was no visible uptake on scintigraphy, and formal evaluation documented persistent profound hypothyroidism. The absence of iodide uptake on scintigraphy is indicative of a severe disruption of follicular function also affecting the function of the sodium-iodide symporter (NIS). There are unfortunately no details on the hearing tests or radiological studies of the inner ear, but his auditory function was apparently not impaired. After the surprising identification of a SLC26A4 missense mutation in a patient with thyroid hypoplasia, Kühnen et al (9) subsequently analyzed 94 patients (64 females, 30 males) with thyroid dysgenesis for mutations in the SLC26A4 gene. This led to the identification of a second patient homozygous for a SLC26A4 gene mutation (c.1238A>G; p.Q41R). This patient, the offspring of first-degree cousins, was diagnosed with severe hypothyroidism at newborn screening. Moreover, he had deafness and mental retardation. At the age of 15 years, he was found to have a hypoplastic gland; structural information about the thyroid in infancy is also unavailable for this patient. The mutation has been previously described (11). Kühnen et al (9) identified three additional patients with monoallelic mutations in the coding region or a splice site and one patient who was compound heterozygous for a missense and an intronic mutation. Taken together, the observations by Kühnen et al (9) suggest that biallelic mutations of the SLC26A4 gene can be associated with a hypoplastic thyroid gland. The authors speculate that this hypoplasia is caused by a “secondary atrophy,” rather than a primary defect of thyroid development reflecting thyroid “agenesis.” Thus, although SLC26A4 mutations are typically thought to result in goiter with or without thyroid dysfunction, they can result in thyroid hypoplasia in some patients, and the SLC26A4 gene has to be added to the currently known candidate genes causing thyroid dysgenesis (TSHR, PAX8, NKX2.1, NKX2.5, and FOXE1). What could explain the thyroid hypoplasia observed by Kühnen et al? It is remarkable that a similar spectrum in the thyroid phenotype has been found in rodent models with mutations in the thyroglobulin (Tg) gene. The cog/cog mouse, very similar to the many reports in human patients with TG gene mutations, has a phenotype with severe hypothyroidism associated with the development of a large goiter (12). In contrast, the recessive dwarf rdw/rdw rat displays a nongoitrous form of congenital primary hypothyroidism (13, 14). The rdw/rdw rat also has a missense mutation in the Tg gene replacing a highly conserved amino acid (G2320R). As in the cog/cog mouse, the mutated Tg is retained in the endoplasmic reticulum, but there is no goitrous enlargement of the thyroid. Although the underlying mechanism resulting in this phenomenon remains to be characterized in further detail, one is tempted to speculate that the retained misfolded protein could have a toxic effect leading to cell death, as shown for example for misfolded vasopressin-neurophysin in the case of autosomal dominant neurohypophyseal diabetes insipidus (15). Whether this does indeed apply to the thyroid hypoplasia seen in patients with SLC26A4 mutations remains uncertain because we do not have information about the thyroid phenotype during the newborn period. Alternatively, as discussed by the authors, the thyroid may be damaged by increased production of free radicals in response to sustained stimulation by TSH. In any case, the observations in the patients reported by Kühnen et al (9) and in the rdw/rdw rat indicate that mutations in genes involved in thyroid synthesis can, in selected instances, result in thyroid hypoplasia. However, given that the mutations identified by Kühnen et al (9) have also been found in patients with classic Pendred syndrome or EVA, and that hypoplasia is probably an uncommon event in patients with mutations in genes involved in thyroid hormone synthesis, one has to infer that additional modifying (genetic) factors are necessary for the development of the hypoplastic phenotype. Very recently, Ladsous et al (16) thoroughly characterized the thyroid phenotype in 19 patients with Pendred syndrome and 23 patients with nonsyndromic EVA. The 19 patients with Pendred syndrome all had biallelic mutations in the SLC26A4 gene. Fifteen (79%) presented with a goiter, and 15 (79%) had hypothyroidism; hypothyroidism was subclinical in four patients and congenital in six patients. The perchlorate test (defined as > 10% discharge of radioiodine) was positive in 10 of 16 (63%) patients. Morphological evaluation of the inner ear showed bilateral EVA in 15 of 15 patients. In the 23 patients with nonsyndromic EVA, systematic thyroid evaluation did not identify abnormalities except for slightly increased serum Tg levels in two patients. The careful characterization of the thyroid phenotype in these patients with Pendred syndrome from northern France emphasizes that the thyroid abnormalities are variable in patients with Pendred syndrome and that they include overt (congenital) hypothyroidism. These authors proposed to use the combination of three parameters to define and diagnose Pendred syndrome: 1) sensorineural deafness with bilateral EVA; 2) a thyroid abnormality such as goiter and/or hypothyroidism and/or a positive perchlorate test; and 3) biallelic mutations in the SLC26A4 gene. Because the perchlorate test is not standardized and often not readily available, molecular analysis of the SLC26A4 gene should, in our opinion, be considered the “gold standard” for establishing a definitive diagnosis of Pendred syndrome. The studies by Kühnen et al (9) and Ladsous et al (16) stimulate reflection about the controversies concerning the physiological role of pendrin in mediating iodide efflux (6, 7). It is remarkable that Ladsous et al (16) found that 15 of 19 patients had hypothyroidism and a goiter. Of note, hypothyroidism was congenital in six of 15 patients, subclinical in four of 19, and overt but not congenital in five of 15 patients. These observations provide indirect support for a physiological role of pendrin in thyroid hormone synthesis. On the other hand, the fact that four of 19 patients did not have any thyroid dysfunction illustrates again that iodide can cross the apical membrane independently of pendrin through another iodide channel or unspecific channels following the electrochemical gradient (17). Assuming that the patients studied by Kühnen et al (9) did indeed have “secondary atrophy,” as hypothesized by the authors, the loss of function of pendrin in these patients with severe hypothyroidism allows us, however, to speculate that pendrin may have a predominant role in apical iodide efflux. In this context, it is of relevance that Shcheynikov et al (5) have shown that iodide is the preferred anion of pendrin and that pendrin transports iodide in the presence of high chloride. This study also demonstrated luminal iodide secretion in parotid gland ducts, emphasizing that SLC26A4 has a preferential affinity for iodide (5). This is important because the intracellular affinity of pendrin for iodide must be substantially higher than for chloride in order to mediate iodide efflux (18). Although the emphasis of many research groups and funding agencies is currently mainly focused on common multifactorial disorders, the observations by Kühnen et al (9) illustrate that the study of rare disorders, which is now facilitated by unbiased analyses of whole exomes or genomes, continues to provide unexpected and novel pathophysiological insights (19). Disclosure Summary: The author has nothing to declare. enlarged vestibular aqueduct thyroglobulin.