Functional characterization of four novel PAX8 mutations causing congenital hypothyroidism: new evidence for haploinsufficiency as a disease mechanism

in European Journal of Endocrinology

(Correspondence should be addressed to S Narumi; Email: sat_naru@hotmail.com)

Background

Individuals carrying a heterozygous inactivating PAX8 mutation are affected by congenital hypothyroidism (CH), although heterozygous Pax8 knockout mice are not. It has remained unclear whether CH in PAX8 mutation carriers is caused by haploinsufficiency or a dominant negative mechanism.

Objective

To report clinical and molecular findings of four novel PAX8 mutations, including one early-truncating frameshift mutation.

Subjects and methods

Four probands were CH patients. Two had family history of congenital or childhood hypothyroidism. Three probands were diagnosed in the frame of newborn screening for CH, while one had a negative result in screening but was diagnosed subsequently. Three had thyroid hypoplasia and one had a slightly small thyroid with low echogenicity. For these probands and their family members, we sequenced PAX8 using a standard PCR-based method. Pathogenicity of identified mutations was verified in vitro.

Results

We found four novel heterozygous PAX8 mutations in the four probands: L16P, F20S, D46SfsX24, and R133Q. Family studies showed four additional mutation carriers, who were confirmed to have high serum TSH levels. Expression experiments revealed that three mutations (L16P, F20S, and R133Q) had defects in target DNA binding, while D46fs had protein instability that was rescued by the proteasome inhibitor MG132. All four mutations had reduced transactivation on the thyroglobulin promoter, supporting that they were inactivating mutations.

Conclusion

D46fs is the first PAX8 mutation with confirmed protein instability. Our clinical and in vitro findings together suggest that pure PAX8 haploinsufficiency can cause CH in humans.

Abstract

Background

Individuals carrying a heterozygous inactivating PAX8 mutation are affected by congenital hypothyroidism (CH), although heterozygous Pax8 knockout mice are not. It has remained unclear whether CH in PAX8 mutation carriers is caused by haploinsufficiency or a dominant negative mechanism.

Objective

To report clinical and molecular findings of four novel PAX8 mutations, including one early-truncating frameshift mutation.

Subjects and methods

Four probands were CH patients. Two had family history of congenital or childhood hypothyroidism. Three probands were diagnosed in the frame of newborn screening for CH, while one had a negative result in screening but was diagnosed subsequently. Three had thyroid hypoplasia and one had a slightly small thyroid with low echogenicity. For these probands and their family members, we sequenced PAX8 using a standard PCR-based method. Pathogenicity of identified mutations was verified in vitro.

Results

We found four novel heterozygous PAX8 mutations in the four probands: L16P, F20S, D46SfsX24, and R133Q. Family studies showed four additional mutation carriers, who were confirmed to have high serum TSH levels. Expression experiments revealed that three mutations (L16P, F20S, and R133Q) had defects in target DNA binding, while D46fs had protein instability that was rescued by the proteasome inhibitor MG132. All four mutations had reduced transactivation on the thyroglobulin promoter, supporting that they were inactivating mutations.

Conclusion

D46fs is the first PAX8 mutation with confirmed protein instability. Our clinical and in vitro findings together suggest that pure PAX8 haploinsufficiency can cause CH in humans.

Keywords:

Introduction

PAX8, a member of the Pax gene family, plays pivotal roles in thyroid development and physiology. PAX8 is expressed in the developing thyroid to adulthood (1). PAX8 directly regulates transcription of thyroid-specific genes, such as thyroglobulin (Tg), in cultured cell lines (2). Pax8 knockout mice have thyroid aplasia due to defective proliferation and survival of thyroid precursor cells (3). In humans, a heterozygous PAX8 mutation causes congenital hypothyroidism (CH). To date, 32 mutation carriers belonging to 13 families have been described (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14). Clinical phenotypes of mutation carriers are variable, ranging from overt CH with severe thyroid hypoplasia to subclinical CH with a morphologically normal gland. Detected mutations include ten amino acid-altering mutations in the DNA-binding paired domain (R31C, R31H, Q40P, S48F, R52P, S54G, H55Q, C57Y, L62R, and K80_A84dup) and two protein-truncating mutations (R108X and T277X). No consistent genotype–phenotype correlation has been suggested.

Several PAX genes other than PAX8 have been implicated in Mendelian disorders, including PAX2 (papillorenal syndrome; OMIM*167409), PAX3 (Waardenburg syndrome; OMIM *606597), PAX6 (aniridia; OMIM *607108), and PAX9 (tooth agenesis; OMIM *167416). All four genes, when mutated, are assumed to cause a human disease via haploinsufficiency because both entire gene deletion(s) and nucleotide-level mutation(s) with early protein truncation produce disease phenotypes (Supplementary Table 1, see section on supplementary data given at the end of this article). This assumption is also supported by observations of mutant mice with inactivating Pax allele(s), showing the gene dosage effect (e.g. disease phenotype seen in heterozygotes; Supplementary Table 1). Contrastingly, the mechanism linking heterozygous PAX8 mutations and CH in humans has remained obscure. To date, neither entire PAX8 deletion nor early truncation mutation has been reported. Moreover, heterozygous Pax8-knockout mice are not affected by CH (15). Based on the phenotypic difference between mutation-carrying human patients and mice, a dominant negative effect in the former has been proposed (7), although most previous in vitro studies have failed to recapitulate the effect.

Here, we report the identification and functional characterization of four novel CH-associated PAX8 mutations. Among the four mutations, one was a frameshift mutation (D46fs) causing protein instability in vitro. Our clinical and molecular findings about the first experimentally confirmed null PAX8 mutation provides new evidence, indicating that PAX8 haploinsufficiency can cause CH in humans.

Materials and methods

Mutation detection

This study was approved by the Institutional Review Board of Keio University School of Medicine. We obtained written informed consent for molecular studies from the study subjects or his/her parents. Leukocytic DNAs were extracted from the four probands and their family members with the Gentra Puregene Blood Kit (Qiagen). Coding exons and flanking introns of PAX8 (transcript variant PAX8A; GenBank NM_003466.3) were analyzed by standard PCR-based sequencing as described previously (12). Detected mutations were tested in 100 control Japanese individuals.

Three-dimensional modeling

Three-dimensional structures of three missense mutants (L16P, F20S, and R133Q) were modeled with 3D-JIGSAW (http://bmm.cancerresearchuk.org/∼3djigsaw/). The structure data of PAX6–DNA complex (protein data bank ID 6PAX; http://www.rcsb.org/pdb) were used as a template. The pictures of the modeled structures were produced with PyMOL (http://www.pymol.org).

Plasmids, cell culture, and transfection

Vectors encoding human PAX8 cDNA (untagged, myc tagged, or enhanced green fluorescent protein (EGFP) tagged) have been described previously (12). The four mutations (L16P, F20S, D46fs, and R133Q) were introduced into these vectors by site-directed mutagenesis (QuikChange XL Site-Directed Mutagenesis Kit; Agilent Technologies, Santa Clara, CA, USA). All final constructs were verified by direct sequencing. HeLa cells were maintained in DMEM supplemented with 50 U/ml penicillin, 50 μg/ml streptomycin, and 10% fetal bovine serum. For functional assays, cells were transfected with DNA using Lipofectamine 2000 (Life Technologies) according to the manufacturer's protocol.

Western blotting

Cells transfected with each of myc-tagged PAX8 constructs (wild type (WT) or mutant) or the empty vector were harvested at 24 h after transfection. Crude cell lysate was obtained with the M-PER protein extraction reagent (Pierce, Rockford, IL, USA). Samples containing 20 μg protein were separated on 10% SDS–PAGE, and western blotting was performed with a mouse anti-myc MAB (Life Technologies) and a HRP-conjugated rabbit anti-mouse IgG polyclonal antibody (Sigma–Aldrich) as a second antibody. Bound antibody was revealed with a chemiluminescence kit (GE Healthcare, Buckinghamshire, UK).

Cells transfected with the myc-tagged D46fs mutant or the empty vector were treated with dimethyl sulfoxide (DMSO) or DMSO containing 1 μM MG132 (Sigma–Aldrich) for an additional 12 h. Western blotting analyses were performed as described earlier.

Visualization of subcellular localization

Cells grown on sterile glass coverslips were cotransfected with each PAX8–EGFP fusion construct (WT or mutant) and the vector encoding red fluorescent protein-tagged thyroid transcription factor-1 (TTF1). Twenty-four hours after transfection, cells were fixed in 2% formaldehyde/PBS at room temperature for 10 min. Then, coverslips were mounted with Vectashield Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA, USA) and were observed under a TCS-SP5 confocal microscope (Leica Microsystems, Mannheim, Germany).

Electrophoretic mobility shift assay

The two band shift probes, oligo-CT (TGA TGC CCA CTC AAG CTT AGA CAG) and oligo-C (CAC TGC CCA GTC AAG TGT TCT TGA), were prepared by annealing of 3′-biotin-labeled complementary oligonucleotides (purchased from BEX Co., Ltd, Tokyo, Japan). Five micrograms of nuclear protein extraction (prepared with the NE-PER nuclear extraction reagent (Pierce)) were incubated at room temperature in 20 μl binding reaction mixture containing 20 fmol probe, 50 mM KCl, 5 mM MgCl2, 2.5% glycerol, 0.05% NP-40, and 1 μg poly (dI-dC) for 20 min. For competition experiments, a large excess (200×) of unlabeled competitor oligonucleotides was included in the binding reactions. The protein–DNA complexes were subject to gel electrophoresis and transferred to a nylon membrane. The biotin-labeled probe was detected with the Lightshift chemiluminescent EMSA kit (Pierce).

Transactivation assays

Cells grown in 96-well plates with about 80% confluence were cotransfected with 50 ng of the luciferase reporter driven by the TG promoter (TG-luc (12)), and various doses of the effector plasmids or the empty vector. The amount of transfected plasmid was kept constant by adding the empty vector. Twenty-four hours after transfection, we measured firefly luciferase activities using ONE-Glo Luciferase Assay System (Promega). Luciferase activities were represented relative to the activity obtained by transfection of empty vector and were expressed as mean±s.e.m. Welch's t-test was used for statistical comparisons with significance at P<0.05.

Results

Clinical histories

Clinical phenotypes of eight mutation carriers (Patients 1–8) belonging to four families (Families 1–4) are summarized in Table 1. The pedigrees are shown in Fig. 1.

Table 1

Clinical summary of eight PAX8 mutation carriers.

Family 1 L16PFamily 2 F20SFamily 3 D46fsFamily 4 R133Q
VariablePatient 1Patient 2Patient 3Patient 4Patient 5Patient 6Patient 7Patient 8Reference
Age (years), sex11, F8, M41, F5, F38, F10, M39, M10, F
Newborn screeningNegativeNegativeNot testedPositiveNot testedPositiveNot testedPositive
Age at diagnosis (years)0.54120.1340.1390.1
Thyroid function
 Age at evaluation (years)34120.1346395
 Serum TSH (mU/l)19.716.371018316113.85.618.00.5–5.0
 Serum-free T4 (ng/dl)NAa1.0NA1.10.30.90.8NA0.9–1.8
 Serum Tg (ng/ml)1324NANA3602832NA<30
Thyroid ultrasonography
 Age at evaluation94NA33463913
 Size (SDS)3.1−1.7NA−3.5−3.3−1.3−3.0−3.5−2 to +2
 EchogenicityNormalNormalNANormalNormalLowLowLowNormal
 Thyroid growthAbsentNANAAbsentNAPresentNANAPresent
 123I uptake at 24 h (%)12.310.8NANANA22.6NA20.68–40
KClO4 discharge rate (%)8.30.0NANANA12.8NANA<10
NA, not available; T4, thyroxine; Tg, thyroglobulin. Abnormal values/findings are given in boldface.

Total thyroxine was 7.6 μg/dl.

Figure 1
Figure 1

Pedigrees of four families with PAX8 mutations are shown. Values presented below each symbol indicate TSH levels as expressed in milliunits per liter (reference, 0.5–5.0). Family studies showed that all mutation carriers had high serum TSH levels. Squares, men; circles, women; solid symbols, affected by hypothyroidism; open symbols, unaffected.

Citation: European Journal of Endocrinology 167, 5; 10.1530/EJE-12-0410

Family 1

The proband (Patient 1; a girl) had a normal blood spot TSH level (9 mU/l; cutoff level, 10) at newborn screening for CH. Her first thyroid function test was conducted at age 6 months because her mother (Patient 3) had childhood hypothyroidism. At this point, she had a high serum TSH level (14 mU/l; reference, 0.5–5.0) with a normal free thyroxine (T4) level (1.1 ng/dl; reference, 0.9–1.8). She had no CH-related symptom and was growing and developing normally. At age 9 months, levothyroxine (l-T4) replacement was started because her serum TSH level increased to 22.3 mU/l. At age 3 years, we reassessed her thyroid status with transient discontinuation of therapy and confirmed permanent CH (TSH, 19.7 mU/l; T4, 7.6 μg/dl (reference, 9.3–17.1); Tg, 13 ng/ml). Ultrasonography showed a normoechoic hypoplastic thyroid (−2.5 s.d. (16); Supplementary Figure 1, see section on supplementary data given at the end of this article). 123I uptake at 24 h was 12.3% (reference, 8–40), and perchlorate discharge rate was 8.3% (reference, 0–10). Repeated ultrasonography at age 9 years showed thyroid hypoplasia (−3.1 s.d.; Supplementary Figure 1).

Patient 2, a younger brother of the proband, also had a negative result in newborn screening. At age 2 months, his first thyroid function test showed a slightly high serum TSH level (7.3 mU/l) with a normal T4 level (11.2 μg/dl). At age 4 years, his second thyroid function test revealed subclinical hypothyroidism (TSH, 16.3 mU/l; free T4, 1.0 ng/dl; Tg, 24 ng/ml). Ultrasonography showed a slightly hypoplastic thyroid (−1.7 s.d.; Supplementary Figure 1). 123I uptake was normal (10.8% at 24 h), and no discharge after perchlorate challenge was observed. At this point, l-T4 replacement was started.

Patient 3, a mother of the proband, was born before the implementation of newborn screening for CH. She was first evaluated for her thyroid function at age 12 years due to mild mental retardation and short stature. She had a markedly high serum TSH level (710 mU/l) with no thyroid autoantibodies. Thyroid scintigraphy showed a normally located thyroid gland with reduced radioiodine uptake (detailed data were unavailable). She has been receiving l-T4 replacement therapy after the diagnosis.

Family 2

The proband (Patient 4; a girl) was born by cesarean section due to fetal distress. Newborn screening for CH revealed a high blood spot TSH level (79.6 mU/l). At age 9 days, she had a high serum TSH level (183 mU/l) with a normal serum-free T4 level (1.1 ng/dl). Although her primary physician noted no CH-related symptoms, her distal femoral epiphyseal ossification center was absent, suggesting her hypothyroid status. Ultrasonography showed hypoplastic thyroid (−2.5 s.d.; Supplementary Figure 1). Immediately after the evaluation, l-T4 replacement was started. At age 3 years, repeated ultrasonography showed hypoplastic thyroid (−3.5 s.d.) with normal echogenicity (Supplementary Figure 1).

Patient 5, a mother of the proband, was born before the implementation of newborn screening for CH. She had normal height (153 cm), had no obvious intellectual problem, and had no apparent symptoms of hypothyroidism. We evaluated her thyroid function at age 34 years as a family study of the mutation. She had an elevated serum TSH level (161 mU/l), a low free T4 level (0.3 ng/dl), and a high Tg level (360 ng/ml). A hypoplastic normoechoic thyroid (volume, 1.4 ml) was shown by ultrasonography (Supplementary Figure 1). Thereafter, she has been receiving l-T4 replacement therapy.

Family 3

The proband (Patient 6; a boy) had a high blood spot TSH level (17.5 mU/l) at newborn screening. At age 41 days, thyroid function test showed a slightly high serum TSH level (15.6 mU/l) with a marginally low free T4 level (0.9 ng/dl). He did not have CH-related symptoms. Ultrasonography revealed a hypoplastic hypoechoic thyroid (−2.5 s.d.; Supplementary Figure 1). l-T4 replacement was started. At age 6 years, we reevaluated his thyroid status with transient discontinuation of treatment. He had permanent CH with serum TSH 13.8 mU/l, free T4 0.9 ng/dl, and Tg 28 ng/ml. He had normal 123I uptake (19.7% at 24 h), although perchlorate discharge rate was slightly high (12.8%). Ultrasonography showed a normal-sized hypoechoic gland (−1.3 s.d.; Supplementary Figure 1).

Patient 7, a father of the proband, was born before the implementation of newborn screening. He had normal height (171 cm), had no intellectual problems, and was euthyroid physically. We evaluated his thyroid status at age 39 years as a family study of the mutation. He had a slightly high serum TSH level (5.6 mU/l), a slightly low free T4 level (0.8 ng/dl), and a slightly high Tg level (32 ng/ml). Ultrasonography showed a hypoplastic thyroid (volume, 5.4 ml) with low echogenicity (Supplementary Figure 1).

Family 4

The proband (Patient 8; a girl) had a high blood spot TSH level (40.6 mU/l) at newborn screening. She had no CH-related symptoms. The size of distal femoral epiphyseal ossification center was normal. Thyroid function test showed a high serum TSH level (33.2 mU/l) accompanied by a normal free T4 level (1.4 ng/dl). She had been followed without treatment because her free T4 levels were normal but was finally started on l-T4 replacement at age 10 months (TSH, 13.9 mU/l; free T4, 1.1 ng/dl). Reassessment of her thyroid status with discontinuing therapy at age 5 years confirmed permanent CH (TSH, 18.0 mU/l; free T4, 1.2 ng/dl). 123I uptake was normal (20.6% at 24 h). Perchlorate discharge test was not performed. Thyroid ultrasonography performed at age 13 years showed a hypoplastic gland (−3.5 s.d.) with low echogenicity (Supplementary Figure 1).

Mutation detection

We identified four novel heterozygous PAX8 mutations in the four probands, including three missense mutations (c.47T>C, p.L16P in Family 1; c.59T>C, p.F20S in Family 2; c.398G>A, p.R133Q in Family 4) and one frameshift mutation (c.134_149del, p.D46SfsX24 in Family 3) (Fig. 2A). The four mutations were located in the paired domain (Fig. 2A) and were absent in 100 control individuals. Leu16, Phe20, and Arg133 are conserved among the PAX family genes (Fig. 2B). Computational modeling of the missense mutants predicted that they cause loss of contacts between the paired domain and its target DNA (Fig. 2A). Family studies revealed that the mutation was transmitted by either father or mother in three families (Families 1–3) and occurred de novo in Family 4 (Fig. 1).

Figure 2
Figure 2

Location and impact of four novel PAX8 mutations. (A) Three-dimensional structure of the DNA-binding paired domain (colored in gold) and its target DNA (colored in silver) based on the crystal structure data of PAX6–DNA complex. The domain consists of two β-sheets and six α-helices. Side chains of residues corresponding to 11 missense mutations (previously reported ones in turquoise (n=8) and novel ones found in this study in orange (n=3)) are shown as spheres. The eight previous mutations are located among three α-helices (α1–α3), whereas the three novel ones (L16P, F20S, and R133Q) are located outside the region. The D46fs mutation deletes normal protein sequence after the α1 helix. Representative chromatograms of each novel mutation are shown, along with the results of computational modeling of mutant structures. Note that the three missense mutations are predicted to affect protein–DNA interaction (indicated by white arrows). (B) Single-letter amino acid ClustalW alignments of residues surrounding Leu16, Phe20, and Arg133. The mutated residues, which are conserved in the Pax gene family, are colored in red.

Citation: European Journal of Endocrinology 167, 5; 10.1530/EJE-12-0410

In vitro functional analyses

To clarify the molecular pathogenesis of each mutant PAX8, we conducted a series of in vitro functional analyses using the human cervical cancer-derived HeLa cell line. Western blotting of myc-tagged PAX8 proteins showed that the protein expression level of D46fs was negligible, whereas those of the remaining three (L16P, F20S, and R133Q) were comparable with that of WT (Fig. 3A, left panel). When we treated transfected cells with the proteasome inhibitor MG132, we could detect the myc-tagged D46fs mutant, of which molecular weight was about 10 kDa, suggesting that the mutant protein was degraded via the proteasome-dependent pathway (Fig. 3A, right panel). Visualization of subcellular localization of the EGFP-tagged PAX8 proteins revealed that L16P, F20S, and R133Q were localized in the nucleus and were colocalized normally with TTF1, another thyroid-specific transcription factor (Fig. 3B). Electrophoretic mobility shift assay (EMSA) with two PAX8 response elements (oligo-CT and oligo-C) showed abrogated DNA-binding abilities of L16P, F20S, and R133Q on the two elements (Fig. 3C).

Figure 3
Figure 3

Functional characterization of four novel PAX8 mutations. (A) Protein expression levels of myc-tagged PAX8 (wild type (WT) or mutant) were assessed by western blotting using a monoclonal anti-myc antibody. The protein expression level of D46fs was negligible, while those of the remaining three were comparable with that of WT (left panel). The D46fs mutation could be detected by western blotting with use of cell lysate extracted from MG132-treated HeLa cells (right panel), indicating that the mutation was degraded via proteasome-dependent pathway. (B) Subcellular localization analyses. Each enhanced green fluorescent protein-tagged PAX8 construct (WT or mutant) was cotransfected with red fluorescent protein-tagged thyroid transcription factor-1 (TTF1). Merged images show colocalization of TTF1 and each PAX8 protein. Bars indicate 50 μm. (C) DNA binding abilities of each PAX8 protein on two PAX8 response elements (oligo-CT and oligo-C) were tested by electrophoretic mobility shift assays. WT showed specific binding to the elements, which was competed by excess amount of cold competitors. Each mutant showed no binding on these two elements. (D and E) Transactivation activities of each PAX8 protein were assessed with use of the TG-luc reporter. HeLa cells grown in a 96-well plate were transfected with indicated amount (in nanogram) of the effector plasmid(s). Firefly luciferase activities were represented relative to the activity of the empty vector. Panel D displays the results comparing WT (black) and the four mutants (gray) in the absence of TTF1. WT transactivated the TG-luc in a dose-dependent manner. The four mutants showed negligible transactivating capacities. WT-mutant contransfection experiments showed no dominant negative effect. Panel E shows the results obtained in the presence of coexpressed TTF1 (1 ng). In this condition, three missense mutants showed various levels of transactivation, whereas D46fs remained nonfunctional. However, the transactivation levels derived from WT-mutant cotransfection (5 ng each) were comparable to that of WT only (5 ng). Data are representative of three independent experiments (each performed in quadricate) with similar results. Values are mean±s.e.m. The results of mutant-only transfection (5 ng) were compared with that of equimolar WT (5 ng), while those of WT-mutant cotransfection (total amount 10 ng) were compared with that of equimolar WT (10 ng). *P<0.05, **P<0.01.

Citation: European Journal of Endocrinology 167, 5; 10.1530/EJE-12-0410

We assessed the effect of each mutation on target gene transactivation using a luciferase reporter driven by the TG promoter. To recapitulate interaction between WT and mutant PAX8, and interaction between PAX8 and TTF1 (i.e. synergistic transactivation (17)), various patterns of cotransfection were tested: mutant PAX8 only or WT-mutant cotransfection, each with or without coexpressed TTF1. In the absence of TTF1, the four mutants alone showed negligible transactivation (Fig. 3D, gray bars). Transactivating capacities measured by WT-mutant cotransfection (WT 5 ng; mutant 5 ng) were significantly lower than that derived from 10 ng of WT and were comparable with that derived from 5 ng of WT (Fig. 3D, stripe bars). This indicates that each mutant did not interfere with the transactivation of WT in WT-mutant cotransfection. In the presence of coexpressed TTF1, L16P, F20S, and R133Q had partial transactivating capacities, while D46fs did not (Fig. 3E, gray bars). However, when the four mutants were further cotransfected with WT PAX8 (WT 5 ng; mutant 5 ng), the transactivation levels of TG-luc were significantly lower than that derived from 10 ng of WT and were again comparable with that derived from 5 ng of WT (Fig. 3E, stripe bars).

Discussion

The thyroid phenotypes of the eight mutation carriers were considerably variable, regarding i) thyroid function (overt to subclinical hypothyroidism), ii) gland size (small to normal), and iii) gland echogenicity (low to normal). It is noteworthy that chronological changes were observed in several cases. Patients 1 and 2 had normal blood spot TSH levels in newborn screening but developed hypothyroidism thereafter. One similar screening-negative case has been described in the literature (10). As for Patient 1, thyroid ultrasonography was performed sequentially at ages 3 and 9 years and showed that her thyroid grew only minimally. This finding is consistent with our previous clinical observation suggesting that thyroid follicular growth is sensitive to a PAX8 mutation (18). Another observation implying the chronological phenotypic changes in mutation carriers is the paradoxical clinical history of Patient 5. She presumably had compensated hypothyroidism in her childhood because she had apparently normal adult height and intelligence. Thus, her overt hypothyroidism, which was confirmed at age 34 years, should have developed in adulthood. Collectively, we presume that deleterious effects of PAX8 mutations can exacerbate over time, resulting in significant chronological changes in disease phenotypes.

In this study, two types of PAX8 mutations were found: mutations with defective DNA binding (L16P, F20S, and R133Q) and a mutation with defective protein stability (D46fs). The former type seems to be the predominant mechanism of PAX8 mutations: ten out of 12 previously reported mutations are amino acid-altering ones located in the paired domain (4, 7, 8, 9, 10, 11, 12, 13). The paired domain consists of two β-turns (β1 and β2) and six α-helices (α1–α6). All previously reported missense mutations are confined between α1 and α3 helices (i.e. mutational hotspot; Fig. 2A). L16P (between β1 and β2), F20S (β2), and R133Q (α6) are the first mutations that are located outside this hotspot. These data imply the importance of the β-turns and α6-helix of PAX8 in target DNA binding. As for the latter type, the D46fs mutation lacks 90% of normal protein sequence and acquires extra 23 frameshifted sequence. The results of western blotting with use of MG132 implied that the mutant was degraded via the proteasome-dependent pathway. Considering that the D46fs mutant consists of about 30% abnormal protein sequence, we speculate that the mutant was misfolded and was subject to endoplasmic reticulum-associated degradation. The D46fs mutation showed abrogated transactivation on the TG promoter irrespective of experimental conditions, supporting that it is actually a ‘null’ mutation. Our clinical and molecular data clearly demonstrate that one ‘null’ PAX8 allele is enough to cause CH in humans (i.e. haploinsufficiency).

L16P, F20S, and R133Q had negligible transactivating capacities on the TG promoter in the TTF1 absent condition but showed significant transactivation when TTF1 is coexpressed. This ‘rescue’, which has been observed in three mutant PAX8 (S48F (11), H55Q (13), and K80_A84dup (12)) and one mutant TTF1 (P210L (19)), is likely based on the formation of PAX8–TTF1 complex on the TG promoter (20). However, no previous reports have tested whether those mutants could form the transcriptional complex. We suppose that demonstration of formation of the transcriptional complex containing the mutants, such as co-immunoprecipitation studies, will be required to verify the model.

In conclusion, we report clinical and molecular findings of four novel PAX8 mutations, including three missense mutations located outside the previously recognized mutational hotspot and the first experimentally confirmed mutation with protein instability (D46fs). Our data imply that PAX8 haploinsufficiency is enough to cause CH in humans.

Supplementary data

This is linked to the online version of the paper at http://dx.doi.org/10.1530/EJE-12-0410.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

The present work was supported by a Grant-in-Aid for Young Scientists (B) (24791087) from the Japan Society for the Promotion of Science (JSPS), a Grant-in-Aid for Scientific Research (C) (23591517) from JSPS, and the Health Science Research Grant for Research on Applying Health Technology (Jitsuyoka (Nanbyo)-Ippan-014) from the Ministry of Health, Labour and Welfare, Japan.

Acknowledgements

The authors would like to thank Prof. Takao Takahashi for fruitful discussion.

References

  • 1

    PlachovDChowdhuryKWaltherCSimonDGuenetJLGrussP. Pax8, a murine paired box gene expressed in the developing excretory system and thyroid gland. Development1990110643651.

    • Search Google Scholar
    • Export Citation
  • 2

    ZanniniMFrancis-LangHPlachovDDi LauroR. Pax-8, a paired domain-containing protein, binds to a sequence overlapping the recognition site of a homeodomain and activates transcription from two thyroid-specific promoters. Molecular and Cellular Biology19921242304241. (doi:10.1128/MCB.12.9.4230).

    • Search Google Scholar
    • Export Citation
  • 3

    MansouriAChowdhuryKGrussP. Follicular cells of the thyroid gland require Pax8 gene function. Nature Genetics1998198790. (doi:10.1038/ng0598-87).

    • Search Google Scholar
    • Export Citation
  • 4

    CongdonTNguyenLQNogueiraCRHabibyRLMedeiros-NetoGKoppP. A novel mutation (Q40P) in PAX8 associated with congenital hypothyroidism and thyroid hypoplasia: evidence for phenotypic variability in mother and child. Journal of Clinical Endocrinology and Metabolism20018639623967. (doi:10.1210/jc.86.8.3962).

    • Search Google Scholar
    • Export Citation
  • 5

    de SanctisLCorriasARomagnoloDDi PalmaTBiavaABorgarelloGGianinoPSilvestroLZanniniMDianzaniI. Familial PAX8 small deletion (c.989_992delACCC) associated with extreme phenotype variability. Journal of Clinical Endocrinology and Metabolism20048956695674. (doi:10.1210/jc.2004-0398).

    • Search Google Scholar
    • Export Citation
  • 6

    KomatsuMTakahashiTTakahashiINakamuraMTakahashiITakadaG. Thyroid dysgenesis caused by PAX8 mutation: the hypermutability with CpG dinucleotides at codon 31. Journal of Pediatrics2001139597599. (doi:10.1067/mpd.2001.117071).

    • Search Google Scholar
    • Export Citation
  • 7

    MacchiaPELapiPKrudeHPirroMTMisseroCChiovatoLSouabniABasergaMTassiVPincheraAFenziGGrutersABusslingerMDi LauroR. PAX8 mutations associated with congenital hypothyroidism caused by thyroid dysgenesis. Nature Genetics1998198386. (doi:10.1038/ng0598-83).

    • Search Google Scholar
    • Export Citation
  • 8

    MeeusLGilbertBRydlewskiCParmaJRoussieALAbramowiczMVilainCChristopheDCostagliolaSVassartG. Characterization of a novel loss of function mutation of PAX8 in a familial case of congenital hypothyroidism with in-place, normal-sized thyroid. Journal of Clinical Endocrinology and Metabolism20048942854291. (doi:10.1210/jc.2004-0166).

    • Search Google Scholar
    • Export Citation
  • 9

    VilainCRydlewskiCDuprezLHeinrichsCAbramowiczMMalvauxPRenneboogBParmaJCostagliolaSVassartG. Autosomal dominant transmission of congenital thyroid hypoplasia due to loss-of-function mutation of PAX8. Journal of Clinical Endocrinology and Metabolism200186234238. (doi:10.1210/jc.86.1.234).

    • Search Google Scholar
    • Export Citation
  • 10

    Al TajiEBiebermannHLimanovaZHnikovaOZikmundJDameCGrutersALeblJKrudeH. Screening for mutations in transcription factors in a Czech cohort of 170 patients with congenital and early-onset hypothyroidism: identification of a novel PAX8 mutation in dominantly inherited early-onset non-autoimmune hypothyroidism. European Journal of Endocrinology2007156521529. (doi:10.1530/EJE-06-0709).

    • Search Google Scholar
    • Export Citation
  • 11

    GrasbergerHRingkananontULefrancoisPAbramowiczMVassartGRefetoffS. Thyroid transcription factor 1 rescues PAX8/p300 synergism impaired by a natural PAX8 paired domain mutation with dominant negative activity. Molecular Endocrinology20051917791791. (doi:10.1210/me.2004-0426).

    • Search Google Scholar
    • Export Citation
  • 12

    NarumiSMuroyaKAsakuraYAdachiMHasegawaT. Transcription factor mutations and congenital hypothyroidism: systematic genetic screening of a population-based cohort of Japanese patients. Journal of Clinical Endocrinology and Metabolism20109519811985. (doi:10.1210/jc.2009-2373).

    • Search Google Scholar
    • Export Citation
  • 13

    Di PalmaTZampellaEFilipponeMGMacchiaPERis-StalpersCde VroedeMZanniniM. Characterization of a novel loss-of-function mutation of PAX8 associated with congenital hypothyroidism. Clinical Endocrinology201073808814. (doi:10.1111/j.1365-2265.2010.03851.x).

    • Search Google Scholar
    • Export Citation
  • 14

    JoWIshizuKFujiedaKTajimaT. Congenital hypothyroidism caused by a PAX8 gene mutation manifested as sodium/iodide symporter gene defect. Journal of Thyroid Research20102010619013. (doi:10.4061/2010/619013).

    • Search Google Scholar
    • Export Citation
  • 15

    AmendolaEDe LucaPMacchiaPETerraccianoDRosicaAChiappettaGKimuraSMansouriAAffusoAArraCMacchiaVDi LauroRDe FeliceM. A mouse model demonstrates a multigenic origin of congenital hypothyroidism. Endocrinology200514650385047. (doi:10.1210/en.2005-0882).

    • Search Google Scholar
    • Export Citation
  • 16

    YasumotoMInoueHOhashiIShibuyaHOnishiT. Simple new technique for sonographic measurement of the thyroid in neonates and small children. Journal of Clinical Ultrasound2004328285. (doi:10.1002/jcu.10234).

    • Search Google Scholar
    • Export Citation
  • 17

    MiccadeiSDe LeoRZammarchiENataliPGCivitarealeD. The synergistic activity of thyroid transcription factor 1 and Pax 8 relies on the promoter/enhancer interplay. Molecular Endocrinology200216837846. (doi:10.1210/me.16.4.837).

    • Search Google Scholar
    • Export Citation
  • 18

    NarumiSYoshidaAMuroyaKAsakuraYAdachiMFukuzawaRKameyamaKHasegawaT. PAX8 mutation disturbing thyroid follicular growth: a case report. Journal of Clinical Endocrinology and Metabolism201196E2039E2044. (doi:10.1210/jc.2011-1114).

    • Search Google Scholar
    • Export Citation
  • 19

    CarreASzinnaiGCastanetMSura-TruebaSTronEBroutin-L'HermiteIBaratPGoizetCLacombeDMoutardMLRaybaudCRaynaud-RavniCRomanaSYthierHLegerJPolakM. Five new TTF1/NKX2.1 mutations in brain–lung–thyroid syndrome: rescue by PAX8 synergism in one case. Human Molecular Genetics20091822662276. (doi:10.1093/hmg/ddp162).

    • Search Google Scholar
    • Export Citation
  • 20

    Di PalmaTNitschRMasciaANitschLDi LauroRZanniniM. The paired domain-containing factor Pax8 and the homeodomain-containing factor TTF-1 directly interact and synergistically activate transcription. Journal of Biological Chemistry200327833953402. (doi:10.1074/jbc.M205977200).

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

 

     European Society of Endocrinology

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 596 416 7
PDF Downloads 124 94 1
  • View in gallery

    Pedigrees of four families with PAX8 mutations are shown. Values presented below each symbol indicate TSH levels as expressed in milliunits per liter (reference, 0.5–5.0). Family studies showed that all mutation carriers had high serum TSH levels. Squares, men; circles, women; solid symbols, affected by hypothyroidism; open symbols, unaffected.

  • View in gallery

    Location and impact of four novel PAX8 mutations. (A) Three-dimensional structure of the DNA-binding paired domain (colored in gold) and its target DNA (colored in silver) based on the crystal structure data of PAX6–DNA complex. The domain consists of two β-sheets and six α-helices. Side chains of residues corresponding to 11 missense mutations (previously reported ones in turquoise (n=8) and novel ones found in this study in orange (n=3)) are shown as spheres. The eight previous mutations are located among three α-helices (α1–α3), whereas the three novel ones (L16P, F20S, and R133Q) are located outside the region. The D46fs mutation deletes normal protein sequence after the α1 helix. Representative chromatograms of each novel mutation are shown, along with the results of computational modeling of mutant structures. Note that the three missense mutations are predicted to affect protein–DNA interaction (indicated by white arrows). (B) Single-letter amino acid ClustalW alignments of residues surrounding Leu16, Phe20, and Arg133. The mutated residues, which are conserved in the Pax gene family, are colored in red.

  • View in gallery

    Functional characterization of four novel PAX8 mutations. (A) Protein expression levels of myc-tagged PAX8 (wild type (WT) or mutant) were assessed by western blotting using a monoclonal anti-myc antibody. The protein expression level of D46fs was negligible, while those of the remaining three were comparable with that of WT (left panel). The D46fs mutation could be detected by western blotting with use of cell lysate extracted from MG132-treated HeLa cells (right panel), indicating that the mutation was degraded via proteasome-dependent pathway. (B) Subcellular localization analyses. Each enhanced green fluorescent protein-tagged PAX8 construct (WT or mutant) was cotransfected with red fluorescent protein-tagged thyroid transcription factor-1 (TTF1). Merged images show colocalization of TTF1 and each PAX8 protein. Bars indicate 50 μm. (C) DNA binding abilities of each PAX8 protein on two PAX8 response elements (oligo-CT and oligo-C) were tested by electrophoretic mobility shift assays. WT showed specific binding to the elements, which was competed by excess amount of cold competitors. Each mutant showed no binding on these two elements. (D and E) Transactivation activities of each PAX8 protein were assessed with use of the TG-luc reporter. HeLa cells grown in a 96-well plate were transfected with indicated amount (in nanogram) of the effector plasmid(s). Firefly luciferase activities were represented relative to the activity of the empty vector. Panel D displays the results comparing WT (black) and the four mutants (gray) in the absence of TTF1. WT transactivated the TG-luc in a dose-dependent manner. The four mutants showed negligible transactivating capacities. WT-mutant contransfection experiments showed no dominant negative effect. Panel E shows the results obtained in the presence of coexpressed TTF1 (1 ng). In this condition, three missense mutants showed various levels of transactivation, whereas D46fs remained nonfunctional. However, the transactivation levels derived from WT-mutant cotransfection (5 ng each) were comparable to that of WT only (5 ng). Data are representative of three independent experiments (each performed in quadricate) with similar results. Values are mean±s.e.m. The results of mutant-only transfection (5 ng) were compared with that of equimolar WT (5 ng), while those of WT-mutant cotransfection (total amount 10 ng) were compared with that of equimolar WT (10 ng). *P<0.05, **P<0.01.