Two Novel SLC5A5 Variants (Q263L and G350D) Causing Congenital Hypothyroidism

Introduction

The sodium-iodide symporter (NIS), encoded by SLC5A5, is essential for transporting iodide into thyroid follicular cells. Biallelic loss-of-function SLC5A5 variants cause iodide transport impairment, leading to thyroid dyshormonogenesis called iodide transport defect (ITD). More than 15 SLC5A5 missense variants have been identified among ITD patients, and at least 14 have been shown to cause loss of iodide transport function in vitro. ¹ Here we report two novel SLC5A5 variants (p. [Gln263Leu]; [Gly350Asp]) causing ITD and congenital hypothyroidism (CH).

Methods

This study was approved by the Ethics Committee of Keio University School of Medicine (Approval No.2010-263). Written informed consent for molecular studies was obtained from the parents of the patients. Detail methods are shown in Supplementary Data S1. Briefly, we sequenced SLC5A5 (NM_000453.3) using the conventional polymerase chain reaction (PCR)-based method. The three-dimensional structure of the NIS protein was modeled based on the rat NIS structure (PDB accession number, 7UV0) and visualized with PyMOL v0.99. Each NIS-expressing vector (empty vector, Q263L, G350D, or wildtype) was transiently transfected to COS-7 cells with the lipofection method. Western blotting, subcellular localization analysis, and iodide uptake assays were performed.

Patients

Patient 1 (II-1; Fig. 1A), a Japanese boy, was born at term with a birth weight of 3,710 g to nonconsanguineous Japanese parents. He was diagnosed with CH in the frame of newborn screening (NBS) and received levothyroxine therapy soon after the diagnosis. Patient 2, a younger brother of Patient 1 (II-2; Fig. 1A), was born at term with a birth weight of 3,060 g. He had a negative result in NBS but developed CH-related symptoms, including macroglossia, umbilical hernia, and constipation, at age 3 months. He was found to have a high serum TSH level (737 mU/L) and was treated thereafter. Their mother breastfed both brothers in the same way. The sibling was reevaluated at age 9 years (Patient 1) and 5 years (Patient 2) with stopping levothyroxine therapy. They showed similar clinical phenotypes with overt hypothyroidism, goiter (Fig. 1B), low thyroidal ¹²³I uptake, and low saliva-to-plasma ¹²³I ratio (2.3 and 1.4; reference interval >20). They were diagnosed as CH due to ITD, and levothyroxine replacement therapy was resumed.

OPEN IN VIEWER

FIG. 1.

Identification and functional characterization of the SLC5A5 variants. (A) The pedigree and clinical information of the patients and their family. LT4, Levothyroxine; NE, not evaluated. *Data of the probands was obtained after levothyroxine withdrawal. (B) Transverse ultrasound images of the thyroid for patients. (C) Partial sequence chromatograms of the PCR products derived from patients and a control. (D) Predicted three-dimensional structure of each NIS variant. (E) Western blotting of Q263L-NIS and G350D-NIS. The band at 90 kDa (black arrowhead) was likely to represent the fully glycosylated, while the band at 55 kDa (white arrowhead) is supposed to indicate the nonglycosylated NIS monomers. (F) Immunofluorescent staining of cell surface NIS in transfected COS7 cells with Q263L-NIS and G350D-NIS. 4′,6-diamidino-2-phenylindole (DAPI) was used to stain nuclei. (G) Cellular uptake of ¹²⁵I was assessed using COS-7 cells expressing each NIS protein (WT, Q263L, G350D). The uptake of WT-NIS expressing cells was sensitive to KClO₄.

Results

We identified two SLC5A5 variants in the patients: c.788A>T, p. Q263L transmitted from the father, and c.1049G>A, p.G350D transmitted from the mother (Fig. 1A, C). These two variants were absent in 54,301 healthy Japanese individuals in 54KJPN (https://jmorp.megabank.tohoku.ac.jp/). Structure modeling indicated that the Gln263 residue forms a hydrogen bond with Ser116, which is likely disrupted by the substitution of Gln with Leu. Additionally, replacing the Gly350 residue with the Asp was predicted to create abnormal contact with helix 6 (Fig. 1D).

When each variant NIS protein was expressed in COS-7 cells, the expression of mature glycosylated proteins (90 kDa) was likely to decrease, and nonglycosylated proteins (55 kDa) were supposed to be increased (Fig. 1E). ² Subcellular localization analyses revealed that cell surface expression of each variant was attenuated (Fig. 1F). Uptake of radioiodide by COS-7 cells expressing each NIS protein was examined and was found to be negligible in cells expressing Q263L-NIS or G350D-NIS (Fig. 1G).

Discussion

In this study, we described the clinical picture of a sibling case of ITD, one of which was positive for NBS, while the other was negative, and the diagnosis was made with the subsequent development of CH-related symptoms. Previous studies have shown that individuals with identical SLC5A5 variant(s) could have variable clinical phenotypes; ³ however, differing NBS results among siblings are uncommon. It is known that the expression of the ITD phenotype is influenced by the amount of iodine ingested, but in our case, this could not be verified because iodine levels were not measured.

The substitution of Gln²⁶³ to Leu was predicted to disrupt the hydrogen bond between Gln²⁶³ and Ser¹¹⁶. The loss of hydrogen bonds would inhibit protein folding and transport from the endoplasmic reticulum to the Golgi apparatus. The results of Western blot and fluorescent immunostaining of NIS-expressing COS-7 cells were consistent with this assumption.

Transmembrane segment IX, which contains Gly³⁵⁰, is involved in the binding and transport of Na⁺. ⁴ Variants located in this region have been demonstrated to cause CH. The region around Gly³⁵⁰ is enriched by hydrophobic amino acids. We speculate that the G350D variant decreased cell surface expression and may impair membrane transport of Na⁺ due to the alterations in local hydrophobicity, leading to loss of function.

In conclusion, we report on a sibling with ITD having experimentally verified loss-of-function SLC5A5 variants (p.Q263L, p.G350D). Our findings provide unique evidence for the structure-function relationship of the NIS protein.