Transducin Β-Like 1, X-Linked and Nuclear Receptor Co-Repressor

doi:10.1507/endocrj.EJ17-0384 ORIGINAL

Transducin β-like 1, X-linked and nuclear receptor co-repressor cooperatively augment the ligand- independent stimulation of TRH and TSHβ gene promoters by thyroid hormone receptors

Tetsuya Takamizawa *, Tetsurou Satoh *, Tomoko Miyamoto, Yasuyo Nakajima, Takahiro Ishizuka, Takuya Tomaru, Satoshi Yoshino, Akiko Katano-Toki, Ayaka Nishikido, Santosh Sapkota, Takuya Watanabe, Takashi Okamura, Emi Ishida, Kazuhiko Horiguchi, Syunichi Matsumoto, Sumiyasu Ishii, Atsushi Ozawa, Nobuyuki Shibusawa, Shuichi Okada and Masanobu Yamada

Division of Endocrinology and Metabolism, Department of Internal Medicine, Gunma University Graduate School of Medicine, Maebashi, Japan

Abstract. Mutations in TBL1X, a component of the nuclear receptor co-repressor (N-CoR) and silencing mediator of retinoic acid and thyroid hormone receptor co-repressor complexes, have recently been implicated in isolated central hypothyroidism (CeH). However, the mechanisms by which TBL1X mutations affect negative feedback regulation in the hypothalamus- pituitary-thyroid axis remain unclear. N-CoR was previously reported to paradoxically enhance the ligand-independent stimulation of TRH and TSHβ gene promoters by thyroid hormone receptors (TR) in cell culture systems. We herein investigated whether TBL1X affects the unliganded TR-mediated stimulation of the promoter activities of genes negatively regulated by T3 in cooperation with N-CoR. In a hypothalamic neuronal cell line, the unliganded TR-mediated stimulation of the TRH gene promoter was significantly enhanced by co-transfected TBL1X, and the co-transfection of TBL1X with N-CoR further enhanced promoter activity. In contrast, the knockdown of endogenous Tbl1x using short interfering RNA significantly attenuated the N-CoR-mediated enhancement of promoter activity in the presence of unliganded TR. The co-transfection of N365Y or Y458C, TBL1X mutants identified in CeH patients, showed impaired co-activation with N-CoR for the ligand- independent stimulation of the TRH promoter by TR. In the absence of T3, similar or impaired enhancement of the TSHβ gene promoter by the wild type or TBL1X mutants, respectively, was observed in the presence of co-transfected TR and N- CoR in CV-1 cells. These results suggest that TBL1X is needed for the full activation of TRH and TSHβ gene promoters by unliganded TR. Mutations in TBL1X may cause CeH due to the impaired up-regulation of TRH and/or TSHβ gene transcription despite low T3 levels.

Key words: Central hypothyroidism, Transducin β-like 1, X-linked, Co-repressor, TRH, TSHβ

CENTRAL HYPOTHYROIDISM (CeH) is diagnosed conditions including pituitary and hypothalamic tumors, based on plasma free T4 levels less than the reference brain surgery and irradiation, granulomatous diseases, range in combination with inappropriately normal TSH vascular diseases, infection, and trauma as well as sev- levels [1]. Acquired CeH may be caused by a number of eral drugs such as glucocorticoids, dopamine, somato- statin analogs, and a retinoid X receptor agonist [1]. On Submitted Aug. 31, 2017; Accepted Apr. 20, 2018 as EJ17-0384 the other hand, isolated congenital CeH is known to be Released online in J-STAGE as advance publication May 23, 2018 caused by mutations in TSHβ, TRHR, and IGSF1 [1-3]. Correspondence to: Tetsurou Satoh, MD, PhD, Division of Endo- In a mouse model, we previously showed that the knock- crinology and Metabolism, Department of Internal Medicine, out of Trh resulted in isolated congenital CeH [4, 5]. Gunma University Graduate School of Medicine, 3-39-15 Showa- machi, Maebashi 371-8511, Japan. Missense mutations in transducin β-like 1, X-linked E-mail: [email protected] (TBL1X), which is located on human chromosome *These authors contributed equally to this manuscript. Xp22.3-22.2, were recently shown to cause isolated CeH

©The Japan Endocrine Society 2 Takamizawa et al.

frequently complicated by hearing disturbances in Dutch mechanisms by which wild-type TBL1X and mutated patients [6]. TBL1X identified in CeH patients affect the ligand- TBL1X was identified as one of the core components independent stimulation of TRH and TSHβ gene promot- in nuclear receptor co-repressor (N-CoR)/silencing medi- ers by TR in cooperation with N-CoR. ator of retinoic acid and thyroid hormone receptors (SMRT) co-repressor complexes [7, 8]. In the absence of Materials and Methods T3, the co-repressor complexes associate with DNA- bound thyroid hormone receptors (TR) in order to Cell culture actively repress the transcription of positively regulated N-1, a mouse embryonic hypothalamic cell line genes by thyroid hormone (TH) through tightly packag- immortalized by the retroviral transfer of the SV40 large ing the surrounding chromatin structure via the deacety- T-antigen, was obtained from CELLutions BIOSYSTEMS lation of histone tails [9]. The human TBL1X protein, (Westbury, NY). CV-1 cells were kindly provided by Dr. consisting of 526 amino acids, belongs to the F-box/ Shigekazu Sasaki (Hamamatsu Medical University, WD40 domain protein family [6] and is expressed in Hamamatsu, Japan). These cells were cultured in DMEM several hypothalamic nuclei including the paraventricu- supplemented with 10% FBS (Biowest, Nuaillé, France) lar nucleus and pituitary gland [6]. TBL1X was previ- and ampicillin/streptomycin (GIBCO by Life Technology). ously reported to function as a specific adaptor that recruits the ubiquitin-conjugating/19S proteasome com- Plasmids plex, which mediates the exchange of co-repressors for Firefly luciferase reporter vectors driven by the mouse co-activators in vitro [10] and also interacts with the TRH gene promoter (–256/+83 pA3Luc), human TSHβ hypoacetylated histones H2 and H4 [8]. Two mutant gene promoter (–1,192/+38 pA3Luc), and Herpes TBL1X proteins, N365Y and H453Y identified in CeH Simplex virus thymidine kinase gene promoter patients were found to be poorly expressed when tran- (TK109-pA3Luc) were described previously [11, 13, 14]. siently transfected into culture cells, and this was sugges- Expression vectors for mouse TRα1 (pRSV-TRα1), ted to be due to aberrant protein folding or stability. human TRβ1 (pKCR2-TRβ1), mouse TRβ2 (pRSV- In addition, the Y458C mutant TBL1X protein was TRβ2), and mouse N-CoR (pCMX-FLAG-mN-CoR) expressed at a similar level to the wild-type protein in were described previously [11, 13, 15]. Expression vec- transient transfection, but showed reduced thermal stabil- tors for human TBL1X (pCMV6-Entry-TBL1X) were ity in vitro [6]. This biochemical characterization of obtained from OriGene Technologies (Rockville, MD). mutated TBL1X proteins identified in CeH patients sug- Plasmids expressing GATA2 (pcDNA2.1TOPO-GATA2) gests that the instability of mutant proteins is responsible and Pit-1 (pBKCMV-Pit-1) were described previously for impaired increases in serum TSH levels in these [14]. patients; however, the molecular mechanisms by which mutant TBL1X proteins functionally disturb the nega- Total RNA isolation and RT-PCR analysis tive feedback regulation of hypothalamic TRH and/or Total RNA was isolated from N-1 and CV-1 cells pituitary TSHβ gene transcription by TH have not yet using Isogen (NIPPON GENE, Tokyo, Japan) according been elucidated. to the manufacturer ’ s protocol. Five micrograms of Previous studies and our group demonstrated that the total RNA was subjected to first-strand cDNA synthesis promoter activities of the TRH and TSHβ genes were using Superscript III reverse transcriptase (Invitrogen). stimulated or repressed by co-transfected TR in the One microliter of cDNA was amplified by PCR using absence or presence of T3, respectively, using transient AmpliTaq DNA polymerase (Thermo Fisher transfection assays [11-13] and also that the co- SCIENTIFIC). The nucleotide sequences of the PCR transfection of N-CoR/SMRT paradoxically augmented primers used were as follows: Ncor sense GTCA the ligand-independent stimulation of TRH and TSHβ CAGCCCATTTGATCCT, Ncor antisense CCTGCATC gene promoters by TR [12, 13]. However, it currently TGCTGTGAGGTA; Tbl1x sense ATCAAGTGGAGT remains unclear whether core components other than N- CCCACAGG, Tbl1x antisense TCCACTGGCCAAA CoR/SMRT in co-repressor complexes affect the unli- TACTTCC; Glyceraldehyde-3-phosphate dehydrogenase ganded TR-mediated activation of TRH and TSHβ gene (GAPDH) sense ACCACAGTCCATGCCATCAC, anti- promoters. In the present study, we investigated the sense TCCACCACCCTGTTGCTGTA. All PCR primers

Endocrine Journal Advance Publication Endocrine Journal Advance Publication Regulation of TRH and TSHβ promoters by TBL1X 3

were designed from different exons of each gene in order using pCMV6-entry-TBL1X as the template. The proper to avoid genomic DNA amplification. Predicted PCR introduction of mutations was verified by nucleotide product sizes were 243 base pairs (bp) for Ncor, 192 bp sequencing. for Tbl1x, and 462 bp for Gapdh. PCR was performed under the following conditions for 35 cycles: denaturing Preparation of whole cell lysates (WCL) and at 94°C for 30 sec, annealing at 56°C for 30 sec, and immunoblotting extension at 72°C for 1 min. PCR products were separa- N-1 cells were split onto 100-mm dishes and trans- ted by agarose gel electrophoresis and stained with ethi- fected with the wild-type or mutant TBL1X expression dium bromide. Proper amplification was verified by the vector using Lipofectamine 2000 reagent (Thermo Fisher direct sequencing of PCR products. SCIENTIFIC). Media were changed after 16 hours of transfection and WCL were prepared using the Universal Transient transfection and luciferase assays Magnetic Co-IP kit according to the manufacturer’s pro- Cells were split 16 hours before transfection onto 6- tocol (Active Motif). Protein concentrations were meas- well plates and cultured in DMEM supplemented with ured by Bradford’s method using bovine serum albumin 10% resin-charcoal double-stripped FBS, as described as the standard. Thirty micrograms of WCL was separa- previously [11, 13]. Transient transfection was per- ted by SDS-PAGE and immunoblotting was performed formed using a calcium phosphate precipitation method using a rabbit polyclonal anti-TBL1X antibody (Ab) [17] and the total amounts of transfected plasmids were ad- and rabbit polyclonal anti-cyclophilin A Ab (Upstate, justed using an empty pSVSPORT vector (Invitrogen). Lake Placid, NY), as described previously [16]. Twenty-four hours after transfection, media were changed and 10 nM T3 (SIGMA, Saint Louis, MO) was Statistical analysis added. Luciferase assays were performed 24 hours later Statistical analyses were performed using unpaired t- and luciferase activity was normalized by protein con- test and ANOVA followed by Turkey’s comparison test. centrations and expressed as relative promoter activity, The significance of differences was set at p < 0.05. as previously described [11, 13]. Results Knockdown of endogenous Tbl1x using short interfering RNA (siRNA) followed by the luciferase In order to evaluate the roles of TBL1X in negative reporter assay feedback regulation in the hypothalamus-pituitary-thyroid siGENOME mouse Tbl1x siRNA-SMART pool axis, we performed luciferase reporter assays using a (M-046786-01-0005) and siGENOME Non-Targeting reporter vector driven by the TRH gene promoter in N-1 siRNA#3 (01210-03-20) were obtained from GE Health- cells. RT-PCR analyses identified the expression of care Dharmacon Inc. (Tokyo, Japan). Sixteen hours after endogenous N-cor and Tbl1x mRNA, but not Trh mRNA the transfection of 50 pmol of siRNA/well to N-1 cells in N-1 cells (Fig. 1A and data not shown). In this cell seeded on 6-well plates using Lipofectamine 2000 line, TRH gene promoter activity was stimulated by the reagent (Invitrogen), the TRH reporter gene was co- co-transfection of TR expression vectors and the addition transfected with expression vectors for TRβ1 and/or N- of 10 nM T3 reversed this activation (Fig. 1B), as re- CoR using a calcium phosphate precipitation method, as ported previously [11-13]. Since co-transfected TRα1, described previously [11, 13]. Twenty-four hours after TRβ1, and TRβ2 showed the indistinguishable stimula- the incubation with 10 nM of T3, the luciferase assay tion of promoter activity in the absence of T3, which is was performed. The knockdown efficiency of Tbl1x was consistent with our previous findings [11], we used the evaluated by quantitative PCR (q-PCR) using the PCR TRβ1 or β2 expression vector in subsequent experi- primers described above, as reported previously [16], ments. and immunoblot analyses. We then investigated whether the co-transfection of N- CoR and/or TBL1X affects the ligand-independent acti- Construction of mutant TBL1X expression vectors vation of the TRH gene promoter by TR. As shown in Two mutant TBL1X expression vectors (N365Y and Fig. 2A, the transfection of either N-CoR or TBL1X sig- Y458C) [6] were constructed with the GeneArt Site- nificantly enhanced the ligand-independent activation of Direct Mutagenesis System (Thermo Fisher Scientific) the TRH gene promoter by TR. The co-transfection of N-

Endocrine Journal Advance Publication Endocrine Journal Advance Publication 4 Takamizawa et al.

Fig. 1 (A) Expression of core components of the N-CoR corepressor complex in N-1 cells. The expression of N-cor, Tbl1x and Gapdh mRNA was investigated by RT-PCR analyses. Representative agarose gel electrophoresis is shown. DNA size markers are shown in base pairs (bp). (B) Ligand-independent and -dependent regulation of the mouse TRH gene promoter by TRα1, TRβ1, and TRβ2 in N-1 cells. A luciferase vector driven by the mouse TRH gene promoter (0.7 μg/well) was co-transfected with a vector expressing TRα1, TRβ1, or TRβ2 (0.17 μg/well). Twenty-four hours after the incubation with 10 nM T3, a luciferase assay was performed as described. Data represent the mean ± SEM of three independent experiments with triplicate determinants.

Fig. 2 (A) Effects of the co-transfection of N-CoR and TBL1X on the ligand-independent and -dependent regulation of the TRH gene promoter by TRβ1 in N-1 cells. A mouse TRH promoter construct (0.7 μg/well) was co-transfected with the TRβ1 (0.17 μg/well), N-CoR (0.33 μg/well), and/or TBL1X expression vector (0.33 μg/well). Twenty-four hours after the incubation with 10 nM T3, a luciferase assay was performed as described. Data represent the mean ± SEM of triplicate determinants. The experiment was repeated three times with similar results. (B) Effects of the co-transfection of N-CoR and TBL1X on TK109 promoter activity in N-1 cells. TK109 Luc (0.7 μg/well) was co-transfected with the TRβ1 (0.17 μg/well), N-CoR (0.33 μg/well), and/or TBL1X expression vector (0.33 μg/well). A luciferase assay was performed as described and data represent the mean ± SEM of three independent experiments with triplicate determinants.

CoR and TBL1X additively augmented promoter activity TRH gene promoter using the TK109 luciferase reporter in the absence of T3. In contrast, the transfection of vector. As shown in Fig. 2B, TK109 promoter activity TBL1X did not affect promoter activity in the absence of was not affected by co-transfected N-CoR and/or co-transfected TR. We then evaluated whether the N- TBL1X in the presence of TR (Fig. 2B). CoR and/or TBL1X-mediated enhancement of the In order to further examine whether the knockdown of ligand-independent stimulation by TR was specific to the endogenous Tbl1x affects the N-CoR-mediated enhance-

Endocrine Journal Advance Publication Endocrine Journal Advance Publication Regulation of TRH and TSHβ promoters by TBL1X 5

Fig. 3 (A) Effects of the knockdown of endogenous Tbl1x on the N-CoR-mediated enhancement of the ligand-independent activation of the TRH gene promoter by TR in N-1 cells. The TRH reporter vector (0.7 μg/well) and expression vectors for TR (0.17 μg/well) and/or N-CoR (0.33 μg/well) were co-transfected 16 hours after the lipofection of siRNA targeting Tbl1x (siTbl1x) or control siRNA (siCtl) into N-1 cells. Twenty-four hours after the incubation with 10 nM of T3, a luciferase assay was performed. Data represent the mean ± SEM of triplicate determinants. The experiment was repeated twice with similar results. (B) Knockdown efficiency of Tbl1x mRNA by siRNA quantitated by q-PCR. Control siRNA (siCtl) or siRNA targeting Tbl1x1 (siTbl1x) was transfected into N-1 cells using lipofection, as described above. Forty-eight hours after transfection, total RNA was isolated and subjected to q-PCR. Data represent the mean ± SEM of triplicate PCR samples. Experiments were repeated once with similar results. (C, D) Immunoblot analyses of the Tbl1x protein in N-1 cells transfected with siCtl and siTbl1x. Forty-eight hours after transfection, whole cell lysates (WCL) were prepared as described above. Thirty micrograms of WCL were subjected to SDS/ PAGE and immunoblotted with anti-TBL1X and anti-cyclophilin antibodies. The positions of molecular size markers are indicated in kilo-Daltons (kDa). Band intensities were quantitated using an image analyzer and the relative expression of Tbl1x proteins is shown. The experiment was repeated twice with similar results.

ment of the ligand-independent stimulation of the TRH We then examined how mutant TBL1X proteins iden- promoter by TR, we performed luciferase reporter assays tified in CeH patients affect TRH and TSHβ gene pro- after the knockdown of endogenous Tbl1x using siRNA moter activities in the presence of co-transfected TR and in N-1 cells. In contrast to cells transfected with control N-CoR in N-1 cells. We selected two mutant TBL1X siRNA (siCtl), the knockdown of Tbl1x (siTbl1x) signif- proteins, N365Y and Y458C in the present study because icantly attenuated the N-CoR-mediated augmentation of N365Y has been reported to be weakly expressed in the ligand-independent activation of the TRH promoter transient transfection, whereas Y458C is known to be by TRβ1 (Fig. 3A). Under the same experimental condi- expressed at a similar level to the wild-type TBL1X tions, the knockdown efficiency of Tbl1x mRNA was protein in transient transfection, but exhibited reduced estimated to be 60% using qRT-PCR (Fig. 3B) and thermal stability in vitro [6]. Consistent with previous approximately 50% in immunoblot analyses (Fig. 3C and findings [6], we found that the expression of the N365Y 3D). protein was consistently weaker than that of wild-type

Endocrine Journal Advance Publication Endocrine Journal Advance Publication 6 Takamizawa et al.

Fig. 4 (A) Western blot analysis of mutant TBL1X proteins. Forty-eight hours after the lipofection of wild-type and mutant TBL1X (N365Y and Y458C) vectors into N-1 cells, whole cell lysates were prepared and immunoblotting was performed using anti- TBL1X and anti-cyclophilin antibodies, as described above. Molecular size markers are shown in kilo-Daltons (kDa). (B) Effects of the co-transfection of N-CoR and wild-type and mutant TBL1X. The mouse TRH promoter construct (0.7 μg/well) was co- transfected with TRβ1 (0.17 μg/well) and N-CoR (0.33 μg/well) in the presence of wild-type and mutant TBL1X expression vectors (N365Y and Y458C) (0.33 μg/well). Twenty-four hours after the incubation with 10 nM T3, the luciferase assay was performed as described. Data represent the mean ± SEM of triplicate determinants. The experiment was repeated three times with similar results. (C) Mutant TBL1X exerts dominant negative activity over wild-type TBL1X on the ligand-independent stimulation of the TRH gene promoter by TR. The mouse TRH promoter construct (0.7 μg/well) was co-transfected with TRβ1 (0.17 μg/well) and N-CoR (0.33 μg/well) in the presence of the wild type (0.33 μg/well) and increasing doses of N365Y and Y458C TBL1X expression vectors (0.33 or 0.66 μg/well). The luciferase assay was performed as described. Data represent the mean ± SEM of triplicate experiments. The experiment was repeated twice with similar results.

TBL1X, whereas the expression of Y458C was similar to teins exert dominant negative activity over wild-type that of wild-type TBL1X (Fig. 4A). Since wild-type, TBL1X, we transfected increasing doses of mutant N365, and Y458C TBL1X cDNAs were inserted into the TBL1X vectors in the presence of a fixed dose of the same expression vector, the slightly larger molecular wild-type TBL1X construct. As shown in Fig. 4C, two masses of the mutant TBL1X proteins may have been mutant TBL1X significantly inhibited the TR/N-CoR- due to differences in post-translational modifications. In mediated stimulation of TRH gene promoter activities, transient transfection assays, co-transfected N365Y and even in the presence of wild-type TBL1X. Y458C significantly attenuated the TRH gene promoter Similar to the TRH gene promoter, the promoter activ- activities enhanced by the co-transfection of N-CoR and ity of the TSHβ gene was previously reported to be wild-type TBL1X in the presence of TR. N365Y showed stimulated by unliganded TR, and ligand-independent significantly stronger attenuation than Y458C (Fig. 4B). activation was further augmented by the co-transfection In order to investigate whether these mutant TBL1X pro- of N-CoR [12]. We then investigated whether TSHβ

Endocrine Journal Advance Publication Endocrine Journal Advance Publication Regulation of TRH and TSHβ promoters by TBL1X 7

Fig. 5 (A) Expression of core components of the N-CoR corepressor complex in CV-1 cells. The expression of N-cor, Tbl1x, and Gapdh mRNA was investigated by RT-PCR. Representative agarose gel electrophoresis is shown. DNA size markers are shown in base pairs (bp). (B) Effects of the co-transfection of wild-type or mutant TBL1X with N-CoR on the ligand-independent and -dependent regulation of the TSHβ gene promoter by TRβ2 in CV-1 cells. The TSHβ promoter construct (0.7 μg/well) was co-transfected with Pit-1 (0.17 µg/well), GATA2 (0.17 µg/well), TRβ2 (0.17 μg/well), or N-CoR (0.33 μg/well) in the presence of the wild-type or mutant TBL1X expression vector (N365Y and Y458C) (0.33 μg/well). Twenty-four hours after the incubation with 10 nM T3, a luciferase assay was performed as described. Data represent the mean ± SEM of triplicate determinants. The experiment was repeated three times with similar results.

gene promoter activity is affected by the wild-type or gene transcription by the sequential recruitment of multi- mutant TBL1X proteins cooperatively with TR/N-CoR ple co-activators harboring diverse enzymatic activities in CV-1 cells. The expression of N-cor, Tbl1x mRNA [9]. The ligand-independent repression of positively was confirmed in this cell line using RT-PCR (Fig. 5A). regulated genes by TH may be mediated by unliganded The activity of the TSHβ gene promoter in this cell line TR-associated N-CoR/SMRT complexes that contain was previously reported to be significantly activated by HDAC3, TBLX1/TBL1XR1, and GPS2 as core compo- co-transfection with Pit-1 and GATA2 expression vectors nents [7, 8]. In contrast, the mechanisms underlying the [14, 18]. Under this experimental setting, the co- unliganded TR-mediated stimulation of negatively regu- transfection of N365Y or Y458C significantly attenuated lated gene promoters by TH, such as the TRH and TSHβ the enhanced promoter activity of the TSHβ gene by TR genes, remain largely unknown [19]. The overexpression and N-CoR (Fig. 5B). Collectively, these results suggest of N-CoR or SMRT was previously reported to paradoxi- that mutations in TBL1X functionally affect the magni- cally enhance the ligand-independent stimulation of the tude of the ligand-independent stimulation by TR of the TRH and TSHβ gene promoters by TR [12, 13] and co- TRH and TSHβ gene promoters, and the expression lev- repressor-mediated enhancements were more promi- els of N365Y and Y458C did not correlate with the nently observed in the presence of TRβ1 than TRβ2 [12]. potency of their dominant negative effects on the TR/N- In the present study, we showed, for the first time, that CoR-mediated enhancement of negatively regulated gene the ligand-independent stimulation of TRH and TSHβ promoters by TH. gene promoters by TRβ1/β2 was enhanced not only by N-CoR, but also by TBL1X. Moreover, we demonstrated Discussion that the knockdown of endogenous Tbl1x or the overexpression of mutant TBL1X proteins attenuated the N- In cell culture systems, the activities of promoters car- CoR-mediated enhancement of ligand-independent rying positive TH-responsive elements are suppressed by activation by TR. Since the N-1 and CV-1 cells used in TR in the absence of T3, and the addition of T3 reversed the present study expressed endogenous TBL1X, these basal repression followed by the strong activation of results suggest that N-CoR in cooperation with TBL1X

Endocrine Journal Advance Publication Endocrine Journal Advance Publication 8 Takamizawa et al.

is involved in the ligand-independent stimulation of TRH protein-DNA interaction platform [22]. Therefore, fac- and TSHβ gene promoters by TR. N-CoR and SMRT tor(s) interacting with the WD40 domain of TBL1X may were previously reported to dissociate from DNA-bound play critical roles in the ligand-independent stimulation TR in the presence of T3 in vitro and in vivo [20, 21]. In of TRH and TSHβ gene promoters by TR. Since the pro- the present study, we showed that the T3-dependent moter activities of the TRH and TSH genes were previ- repression of TRH and TSHβ promoters by TR was pre- ously reported to be activated by the interaction between served, even in the presence of co-transfected N365Y TR and GATA2 [18], mutant TBL1X may inhibit the and Y458C. These results indicate that N-CoR and function of these transcription factors, thereby attenuat- TBL1X play less important roles in the T3-dependent ing ligand-independent activation by TR. Further studies repression of negatively regulated gene promoters. are needed in order to clarify whether the N365Y mutant Since the human TBL1X gene is located on chromo- has a weaker interaction with these factor(s) than some X and TBL1X mutations have been shown to affect Y458C. both genders [6], the pathophysiological mechanism of In summary, we herein established cell-based assay central hypothyroidism caused by TBL1X mutations may systems to assess the functions of TBL1X mutant pro- differ between male and female patients. Consistent with teins in the regulation of TRH and TSHβ gene transcrip- previous findings obtained using HEK293 cells [6], the tion by TR. The present results suggest that TBL1X is N365Y protein was more weakly expressed than the necessary for the full activation of TRH and TSHβ gene wild-type TBL1X and Y458C mutant proteins when promoters by unliganded TR cooperatively with N-CoR. transiently expressed in N-1 cells. These results suggest Mutations in TBL1X may cause CeH due to the impaired that reductions in the expression of some TBL1X mutant up-regulation of TRH and/or TSHβ gene transcription proteins are responsible for the inappropriate increase in despite low TH levels. serum TSH levels in CeH patients. However, we herein showed that N365Y exerted significantly more potent Acknowledgments inhibitory effects than Y458C on the ligand-independent stimulation of TRH and TSHβ gene promoters enhanced We thank Drs. J Wong (East China Normal Univer- by N-CoR. These results suggest that the expression sity) and Shigekazu Sasaki (Hamamatsu Medical School) levels of some mutant TBL1X proteins do not correlate for providing materials. This work was supported by with their dominant negative potency on the TR/N-CoR- Kakenhi (#26461353) to TS. mediated enhancement of negatively regulated gene promoters by TH. To date, all TBL1X mutations identified Disclosure Statement in CeH patients were found in the carboxyl-terminal WD40 domain, but not in the amino-terminal domain None of the authors have any potential conflicts of interacting with N-CoR/SMRT [6, 22]. The WD40 interest associated with this research. domain is known to function as a protein-protein or

References

1. Yamada M, Mori M (2008) Mechanisms related to the 1381. pathophysiology and management of central hypothyroid- 4. Yamada M, Saga Y, Shibusawa N, Hirato J, Murakami M, ism. Nat Clin Pract Endocrinol Metab 4: 638–694. et al. (1997) Tertiary hypothyroidism and hyperglycemia 2. Collu R, Tang J, Castagne J, Lagace G, Masson N, et al. in mice with targeted disruption of the thyrotropin- (1997) A novel mechanism for isolated central hypothy- releasing hormone gene. Proc Natl Acad Sci U S A 94: roidism: inactivating mutations in the thyrotropin-releasing 10862–10867. hormone receptor gene. J Clin Endocr Metab 82: 1561– 5. Yamada M, Shibusawa N, Ishii S, Horiguchi K, Umezawa 1565. R, et al. (2006) Prolactin secretion in mice with 3. Sun Y, Bak B, Schoenmakers N, van Trotsenburg AS, thyrotropin-releasing hormone deficiency. Endocrinology Oostdijk W, et al. (2012) Loss-of-function mutations in 147: 2591–2596. IGSF1 cause an X-linked syndrome of central hypothy- 6. Heinen CA, Losekoot M, Sun Y, Watson PJ, Fairall L, et roidism and testicular enlargement. Nat Genet 44: 1375– al. (2016) Mutations in TBL1X are associated with central

Endocrine Journal Advance Publication Endocrine Journal Advance Publication Regulation of TRH and TSHβ promoters by TBL1X 9

hypothyroidism. J Clin Endocrinol Metab 101: 4564– A, et al. (2012) NR4A1 (Nur77) mediates thyrotropin- 4573. releasing hormone-induced stimulation of transcription of 7. Guenther MG, Lane WS, Fischle W, Verdin E, Lazar MA, the thyrotropin β gene: analysis of TRH knockout mice. et al. (2000) A core SMRT corepressor complex contain- PLoS One 7: e40437. ing HDAC3 and TBL1, a WD40-repeat protein linked to 15. Ishizuka T, Lazar MA (2005) The nuclear receptor core- deafness. Genes Dev 14: 1048–1057. pressor deacetylase activating domain is essential for 8. Yoon HG, Chan DW, Huang ZQ, Li J, Fondell JD, et al. repression by thyroid hormone receptor. Mol Endocrinol (2003) Purification and functional characterization of the 19: 1443–1451. human N-CoR complex: the roles of HDAC3, TBL1 and 16. Yoshino S, Satoh T, Yamada M, Hashimoto K, Tomaru T, TBLR1. EMBO J 22: 1336–1346. et al. (2014) Protection against high-fat diet-induced obe- 9. Glass CK, Rosenfeld MG (2000) The coregulator sity in Helz2-deficient male mice due to enhanced expres- exchange in transcriptional functions of nuclear receptors. sion of hepatic leptin receptor. Endocrinology 155: 3459– Genes Dev 14: 121–141. 3472. 10. Perissi V, Aggarwal A, Glass CK, Rose DW, Rosenfeld 17. Li J, Wang J, Wang J, Nawaz Z, Liu JM, et al. (2000) MG (2004) A corepressor/coactivator exchange complex Both corepressor proteins SMRT and N-CoR exist in large required for transcriptional activation by nuclear receptors protein complexes containing HDAC3. EMBO J 19: and other regulated transcription factors. Cell 116: 511– 4342–4350. 526. 18. Nakano K, Matsushita A, Sasaki S, Misawa H, Nishiyama 11. Satoh T, Yamada M, Iwasaki T, Mori M (1996) Negative K, et al. (2004) Thyroid-hormone-dependent negative reg- regulation of the gene for the preprothyrotropin-releasing ulation of thyrotropin β gene by thyroid hormone recep- hormone from the mouse by thyroid hormone requires tors: study with a new experimental system using CV1 additional factors in conjunction with thyroid hormone cells. Biochem J 378: 549–557. receptors. J Biol Chem 271: 27919–27926. 19. Mendoza A, Hollenberg AN (2017) New insights into thy- 12. Tagami T, Madison LD, Nagaya T, Jameson JL (1997) roid hormone action. Pharmacol Ther 173: 135–145. Nuclear receptor corepressors activate rather than suppress 20. Hörlein AJ, Näär AM, Heinzel T, Torchia J, Gloss B, et basal transcription of genes that are negatively regulated al. (1995) Ligand-independent repression by the thyroid by thyroid hormone. Mol Cell Biol 17: 2642–2648. hormone receptor mediated by a nuclear receptor co- 13. Satoh T, Monden T, Ishizuka T, Mitsuhashi T, Yamada M, repressor. Nature 377: 397–404. et al. (1999) DNA binding and interaction with the nuclear 21. Chen JD, Umesono K, Evans RM (1996) SMRT isoforms receptor corepressor of thyroid hormone receptor are mediate repression and anti-repression of nuclear receptor required for ligand-independent stimulation of the mouse heterodimers. Proc Natl Acad Sci U S A 93: 7567–7571. preprothyrotropin-releasing hormone gene. Mol Cell 22. Xu C, Min J (2011) Structure and function of WD40 Endocrinol 154: 137–149. domain proteins. Protein Cell 2: 202–214. 14. Nakajima Y, Yamada M, Taguchi R, Shibusawa N, Ozawa

Endocrine Journal Advance Publication Endocrine Journal Advance Publication