Thyroid hormone economy in mice overexpressing iodothyronine deiodinases.

In peripheral tissues, triiodothyronine (T3) production and consequent thyroid hormone actions are mainly regulated by iodothyronine deiodinases (DIOs) classified into 3 types: D1, D2, and D3. We aimed to investigate the effects of peripheral DIOs on thyroid hormone economy independent of the hypothalamus-pituitary-thyroid axis. We cloned coding sequences of human DIOs with FLAG-tag and HiBiT-tag sequences into a pcDNA3.1 vector. To obtain full-length proteins, we modified these vectors by cloning the selenocysteine insertion sequence of each DIO (SECIS vectors). Western blot analyses and HiBiT lytic assay using HEK293T cells revealed that SECIS vectors expressed full-length proteins with substantial activity. Subsequently, in vivo transfections of pLIVE-based SECIS vectors into male C57BL/6J mice were performed by hydrodynamic gene delivery to generate mice overexpressing DIOs predominantly in the liver (D1, D2, and D3 mice). After 7 days from transfections, mice were analyzed to clarify phenotypes. To summarize, serum thyroid hormone levels did not change in D1 mice but D2 mice had higher serum free T3 levels. D3 mice developed hypothyroidism with higher serum reverse T3 (rT3) levels. Transfections with levothyroxine administration suggested that thyroid hormone action was upregulated in D2 mice. Our DIO-overexpressing mice provided insights on the physiological properties of upregulated DIOs: D2 augments local thyroid hormone action and recruits T3 into the circulation: D3 decreases circulating T3 and T4 levels with elevated rT3, leading to consumptive hypothyroidism. As D3 mice are expected to be a novel hypothyroidism model, they can contribute to progress in the field of thyroid hormone economy and action.

Bezafibrate induces hypothyroidism in a patient with resistance to thyroid hormone ß due to a G347R variant.

OBJECTIVE: A unique clinical course was observed in a patient with resistance to thyroid hormone ß (RTHß) caused by a variant of the THRB gene leading to the replacement of glycine with arginine in codon 347 (p.G347R). He presented with the syndrome of inappropriate secretion of thyrotropin (TSH) (free T4 [fT4]: 32.43 pmol/L, TSH: 4.67 mIU/L), but slowly developed progressive hypothyroidism (fT4: 8.37 pmol/L, TSH: 100.90 mIU/L) that resolved after suspending bezafibrate (BZ) treatment (fT4: 32.18 pmol/L, TSH: 7.14 mIU/L). This study clinically and experimentally evaluated this interesting phenomenon. METHODS: A retrospective cohort analysis of non-RTHß patients was performed at Kyoto University Hospital. Data before BZ treatment were compared to the first data after treatment. Using reporter assays of iodothyronine deiodinases (DIO1, DIO2, DIO3) in HEK293T cells, we performed functional analyses of mutant thyroid hormone receptor ß with p.G347R (G347R TRß). Mice with G347R TRß were generated by hydrodynamic gene delivery. RESULTS: In non-RTHß patients (n = 7), BZ treatment did not change serum free T3 and TSH but significantly increased fT4 (p = .008). BZ administration increased DIO3 reporter activity in the context of G347R TRß, whereas did not change DIO1 and DIO2 reporter activity. In the livers of mice with G347R TRß, BZ administration increased reverse T3 content, which corresponded to an increase in Dio3 messenger RNA. CONCLUSIONS: While hypothyroidism associated with BZ treatment did not occur in non-RTHß patients, it was observed in a patient with RTHß due to the p.G347R variant. Liver DIO3 upregulation might involve this hypothyroidism.

Opioid-induced adrenal insufficiency in transdermal fentanyl treatment: a revisited diagnosis in clinical setting.

Opioids are widely used for treatment of acute and chronic pain. However, opioids have several well-known clinical adverse effects such as constipation, nausea, respiratory depression and drowsiness. Endocrine dysfunctions are also opioid-induced adverse effects but remain under-diagnosed in clinical settings, especially opioid-induced adrenal insufficiency (OIAI). A 46-year-old woman was treated with transdermal fentanyl at a dose of 90-120 mg daily morphine milligram equivalent for non-malignant chronic pain for four years. Fatigue, loss of appetite and decrease in vitality began about two years after starting fentanyl. Subsequently, constipation and abdominal pain appeared and became worse, which led to suspicion of adrenal insufficiency. Clinical diagnosis of OIAI was established based on laboratory findings of secondary adrenal insufficiency, including corticotropin-releasing hormone stimulation test, clinical history of long-term fentanyl use, and exclusion of other hypothalamic-pituitary diseases. Oral corticosteroid replacement therapy was unable to relieve her abdominal pain and constipation; opioid-rotation and dose-reduction of fentanyl were not feasible because of her persistent pain and severe anxiety. While her clinical course clearly suggested that long-term, relatively high-dose transdermal fentanyl treatment may have contributed to the development of secondary adrenal insufficiency, the symptoms associated with OIAI are generally non-specific and complex. Together with under-recognition of OIAI as a clinical entity, the non-specific, wide range of symptoms can impede prompt diagnosis. Thus, vigilance for early symptoms enabling treatments including corticosteroid replacement therapy is necessary for patients taking long-term and/or high dose opioid treatment.

Overall safety and efficacy of high-dose and low-dose intravenous glucocorticoid therapy in patients with moderate-to-severe active Graves' ophthalmopathy.

The objective of this study was to compare the safety and efficacy of high-dose and low-dose intravenous (iv) glucocorticoid (GC) therapy in patients with Graves' ophthalmopathy (GO) and to investigate which factors may help determine appropriate iv GC doses. The medical records of 43 patients who received different doses of iv GCs for GO were retrospectively reviewed. Twenty patients received high-dose iv GCs (HD group, cumulative dose 9.0-12.0 g) and 18 received low-dose iv GCs (LD group, cumulative dose 4.5 g). Five patients with previous treatment for GO were excluded. Changes in ophthalmic parameters after treatment and frequencies of adverse effects due to GCs of the 2 groups were compared. We also reviewed the incidence of GO progression and hepatic dysfunction after patients were discharged. We evaluated correlations among pretreatment (before treatment) ophthalmic parameters and investigated useful predictive factors for determining iv GC doses. There were no significant differences in ophthalmic parameters reflecting treatment efficacy or overall safety between the groups. Among baseline ophthalmic parameters, corrected signal intensity ratio (cSIR) correlated well with magnetic resonance imaging findings and were more strongly associated with changes in ophthalmic parameters after treatment in the HD group than in the LD group, indicating that pretreatment cSIR might be useful for determining iv GC doses. In conclusion, there were no significant differences in overall safety and efficacy between high-dose and low-dose iv GC therapy in patients with active GO. Further randomized clinical trials with longer observation periods are required to establish the optimal treatment regimen of GO.

High-throughput Screening in Combination With a Cohort Study for Iodothyronine Deiodinases.

Regulatory mechanisms of iodothyronine deiodinases (DIOs) require further elucidation, and conventional methods for evaluating DIOs are unsuitable for high-throughput screening (HTS). Here we explored factors of transcriptional regulation of 3 types of DIOs (DIO1, DIO2, and DIO3) from a chemical library using our designed HTS. We constructed HTS based on a promoter assay and performed a screen of 2480 bioactive compounds. For compounds that were clinically approved, we validated hit compounds through a retrospective cohort study in our department that evaluated changes in thyroid function in patients using the compounds as drug therapy. Furthermore, we verified the involvement of DIOs using mice treated with the compounds. Of the hit compounds, 6 and 7 compounds transcriptionally up- and downregulated DIO1, respectively; 34 transcriptionally upregulated DIO2; and 5 and 2 compounds transcriptionally up- and downregulated DIO3, respectively. The cohort study clarified the clinical effects of some hit compounds: ritodrine increased free triiodothyronine (fT3)/free thyroxine (fT4) ratio and decreased serum thyroid-stimulating hormone (TSH) levels, tadalafil increased serum fT3 levels, and tyrosine kinase inhibitors (TKIs) decreased serum fT3 and fT4 levels and increased serum TSH levels. Following in vivo experiments using treated mice, consistent results were observed in ritodrine, which upregulated DIO2 in the thyroid gland. In conclusion, we completed HTS for DIOs and obtained attractive hit compounds. Our cohort study revealed the clinical significance of ritodrine, sildenafil, and TKIs. We hope our unique method will contribute to analyzing various targets and lists of hit compounds will promote understanding of DIOs.

Exogenous C-type natriuretic peptide therapy for impaired skeletal growth in a murine model of glucocorticoid treatment.

Growth retardation is an important side effect of glucocorticoid (GC)-based drugs, which are widely used in various preparations to treat many pediatric diseases. We investigated the therapeutic effect of exogenous CNP-53, a stable molecular form of intrinsic CNP, on a mouse model of GC-induced growth retardation. We found that CNP-53 successfully restored GC-induced growth retardation when both dexamethasone (DEX) and CNP-53 were injected from 4 to 8 weeks old. Notably, CNP-53 was not effective during the first week. From 4 to 5 weeks old, neither CNP-53 in advance of DEX, nor high-dose CNP-53 improved the effect of CNP. Conversely, when CNP-53 was started at 5 weeks old, final body length at 8 weeks old was comparable to that when CNP-53 was started at 4 weeks old. As for the mechanism of resistance to the CNP effect, DEX did not impair the production of cGMP induced by CNP. CNP reduced Erk phosphorylation even under treatment with DEX, while CNP did not changed that of p38 or GSK3ß. Collectively, the effect of CNP-53 on GC-induced growth retardation is dependent on age in a mouse model, suggesting adequate and deliberate use of CNP would be effective for GC-induced growth retardation in clinical settings.

Live imaging analysis of the growth plate in a murine long bone explanted culture system.

Skeletal growth in mammals, which owes the growth of an individual, occurs at the growth plate and to observe and analyze its dynamic growth is of high interest. Here we performed live imaging analysis of the growth plate of a fetal murine long bone organ culture using two-photon excitation microscopy. We could observe a dynamic growth of the growth plate of explanted fetal murine ulna, as well as the resultant linear elongation of the explants. As for the factors contributing to the elongation of the growth plate, the displacement length of each chondrocyte was larger in the prehypertrophic or hypertrophic zone than in the proliferative zone. The segmented area and its extracellular component were increased in both the proliferative and prehypertrophic-hypertrophic zones, whereas an increase in cellular components was only seen in the prehypertrophic-hypertrophic zone. C-type natriuretic peptide, a known positive stimulator of endochondral bone growth mainly targeting prehypertrophic-hypertrophic zone, augmented all of the factors affecting growth plate elongation, whereas it had little effect on the proliferation of chondrocytes. Collectively, the axial trajectory of each chondrocyte mainly owes cellular or extracellular expansion especially in prehypertrophic-hypertrophic zone and results in growth plate elongation, which might finally result in endochondral bone elongation.

Effects of growth hormone on thyroid function are mediated by type 2 iodothyronine deiodinase in humans.

PURPOSE: Growth hormone (GH) therapy in adults alters thyroid function, and acromegaly often involves thyroid disease. The present study aimed to elucidate roles and mechanisms of GH in regulating thyroid function. METHODS: We performed two retrospective observational studies, which focused on consecutive patients with severe adult GH deficiency who received recombinant human GH (rhGH) therapy (n = 20) and consecutive patients with acromegaly who underwent transsphenoidal surgery (TSS) (n = 25). In both studies, serum free triiodothyronine (fT3), free thyroxine (fT4), and fT3/fT4 ratio were examined before and after the interventions. We subsequently administered GH to four human cell lines (HepG2, TSA201, MCF7, and HTC/C3) in vitro, and examined changes in mRNA levels of iodothyronine deiodinases (D1, D2, and D3). RESULTS: Median serum fT3 level significantly increased after rhGH therapy from 2.38 to 2.78 pg/mL (p < 0.001), and fT4 decreased from 1.115 to 1.065 ng/dL (p = 0.081). TSS significantly decreased median serum fT3 from 3.03 to 2.53 pg/mL (p < 0.001), and increased fT4 from 1.230 to 1.370 ng/dL (p < 0.001). In vitro, GH significantly increased D2 expression at the mRNA level in HTC/C3 cells (p < 0.01), as well as D2 protein and its activity. CONCLUSIONS: GH increased serum fT3 level and decreased serum fT4 level in humans. Our results suggest that its mechanism involves D2 upregulation. Considering this GH effect on thyroid hormone metabolism, data on thyroid function could be useful in the management of GH deficiency and acromegaly.

Rats deficient C-type natriuretic peptide suffer from impaired skeletal growth without early death.

We have previously investigated the physiological role of C-type natriuretic peptide (CNP) on endochondral bone growth, mainly with mutant mouse models deficient in CNP, and reported that CNP is indispensable for physiological endochondral bone growth in mice. However, the survival rate of CNP knockout (KO) mice fell to as low as about 70% until 10 weeks after birth, and we could not sufficiently analyze the phenotype at the adult stage. Herein, we generated CNP KO rats by using zinc-finger nuclease-mediated genome editing technology. We established two lines of mutant rats completely deficient in CNP (CNP KO rats) that exhibited a phenotype identical to that observed in mice deficient in CNP, namely, a short stature with severely impaired endochondral bone growth. Histological analysis revealed that the width of the growth plate, especially that of the hypertrophic chondrocyte layer, was markedly lower and the proliferation of growth plate chondrocytes tended to be reduced in CNP KO rats. Notably, CNP KO rats did not have malocclusions and survived for over one year after birth. At 33 weeks of age, CNP KO rats persisted significantly shorter than wild-type rats, with closed growth plates of the femur in all samples, which were not observed in wild-type rats. Histologically, CNP deficiency affected only bones among all body tissues studied. Thus, CNP KO rats survive over one year, and exhibit a deficit in endochondral bone growth and growth retardation throughout life.

Exogenous C-type natriuretic peptide restores normal growth and prevents early growth plate closure in its deficient rats.

Signaling by C-type natriuretic peptide (CNP) and its receptor, natriuretic peptide receptor-B, is a pivotal stimulator of endochondral bone growth. We recently developed CNP knockout (KO) rats that exhibit impaired skeletal growth with early growth plate closure. In the current study, we further characterized the phenotype and growth plate morphology in CNP-KO rats, and the effects of exogenous CNP in rats. We used CNP-53, an endogenous form of CNP consisting of 53 amino acids, and administered it for four weeks by continuous subcutaneous infusion at 0.15 or 0.5 mg/kg/day to four-week old CNP-KO and littermate wild type (WT) rats. We demonstrated that CNP-KO rats were useful as a reproducible animal model for skeletal dysplasia, due to their impairment in endochondral bone growth. There was no significant difference in plasma bone-turnover markers between the CNP-KO and WT rats. At eight weeks of age, growth plate closure was observed in the distal end of the tibia and the calcaneus of CNP-KO rats. Continuous subcutaneous infusion of CNP-53 significantly, and in a dose-dependent manner, stimulated skeletal growth in CNP-KO and WT rats, with CNP-KO rats being more sensitive to the treatment. CNP-53 also normalized the length of long bones and the growth plate thickness, and prevented growth plate closure in the CNP-KO rats. Using organ culture experiment of fetal rat tibia, gene set enrichment analysis indicated that CNP might have a negative influence on mitogen activated protein kinase signaling cascades in chondrocyte. Our results indicated that CNP-KO rats might be a valuable animal model for investigating growth plate physiology and the mechanism of growth plate closure, and that CNP-53, or its analog, may have the potential to promote growth and to prevent early growth plate closure in the short stature.

Clinical Features of Nivolumab-Induced Thyroiditis: A Case Series Study.

BACKGROUND: The programmed cell death-1 (PD-1) pathway is a novel therapeutic target in immune checkpoint therapy for cancer. It consists of the PD-1 receptor and its two ligands, programmed death-ligand 1 (PD-L1) and programmed death-ligand 2 (PD-L2). Nivolumab is an anti-PD-1 monoclonal antibody approved for malignant melanoma, advanced non-small cell lung cancer, and advanced renal cell carcinoma in Japan. Thyrotoxicosis and hypothyroidism have both been reported in international Phase 3 studies and national post-marketing surveillance of nivolumab in Japan. METHODS: This study analyzed five consecutive cases with thyroid dysfunction associated with nivolumab therapy. Second, it examined the mRNA and protein expressions of PD-L1 and PD-L2 by reverse transcription polymerase chain reaction and Western blotting. RESULTS: All patients were diagnosed with painless thyroiditis. Thyrotoxicosis developed within four weeks from the first administration of nivolumab and normalized within four weeks of onset in three of the five patients. Hypothyroidism after transient thyrotoxicosis developed in two patients, and preexisting hypothyroidism persisted in one patient. The other two patients were treated with glucocorticoids and discontinued nivolumab therapy for comorbid adverse events. One did not develop hypothyroidism, and the other developed mild, transient hypothyroidism. In addition, it was verified that normal thyroid tissue expresses PD-L1 and PD-L2 mRNA and those proteins. CONCLUSIONS: In the present cases, nivolumab-induced thyrotoxicosis seemed to be associated with painless thyroiditis, while no patient with Graves' disease was observed. A transient and rapid course with subsequent hypothyroidism was observed in nivolumab-induced thyroiditis. In addition, it was verified that PD-L1 and PD-L2 are expressed in normal thyroid tissue. This suggests that nivolumab therapy reduces immune tolerance, even in normal thyroid tissue, and leads to the development of thyroiditis. Treating thyrotoxicosis with only supportive care and considering levothyroxine replacement therapy once subsequent hypothyroidism occurs is proposed. Further investigations are required to confirm whether glucocorticoid therapy and discontinuation of nivolumab therapy prevent subsequent hypothyroidism.

Regulation of type 1 iodothyronine deiodinase by LXRα.

The iodothyronine deiodinases are selenoenzymes that regulate the activity of thyroid hormone via specific inner- or outer-ring deiodination. In humans, type 1 deiodinase (D1) is highly expressed in the liver, but the mechanism by which its gene expression is regulated remains to be elucidated. Liver X receptor α (LXRα), a transcription factor of the nuclear receptor superfamily, is highly expressed in the liver, where it functions as a sensor for excess intracellular oxysterols. LXRα interacts with other nuclear receptors on promoters of genes that contain a binding core sequence for nuclear receptors. In addition, it is reported that the promoter of the gene encoding human D1 (hDIO1) contains the core sequence for one of nuclear receptors, thyroid hormone receptor (TR). We investigated the involvement of LXRα in the regulation of hDIO1, in the liver. We performed hDIO1 promoter-reporter assays using a synthetic LXR agonist, T0901317, and compared promoter activity between a human liver carcinoma cell line, HepG2, and a clone of human embryonic kidney cells, TSA201. We defined the region between nucleotides -131 and -114, especially nucleotides -126 and -125, of the hDIO1 promoter as critical for basal and LXRα-mediated specific transcriptional activation in HepG2 cells. An increase in hDIO1 expression was observed in LXRα-stimulated cells, but absent in cycloheximide-treated cells, indicating that new protein synthesis is required for LXRα-mediated regulation of hDIO1. On the other hand, electrophoretic mobility shift assays revealed that LXRα and RXRα bound to the hDIO1 promoter. We also demonstrated that LXRα and TRß compete with each other on this specific region of the promoter. In conclusion, our results indicated that LXRα plays a specific and important role in activation of TH by regulating D1, and that LXRα binds to and regulates the hDIO1 promoter, competing with TRß on specific sequences within the promoter.

Circulating osteocrin stimulates bone growth by limiting C-type natriuretic peptide clearance.

Although peptides are safe and useful as therapeutics, they are often easily degraded or metabolized. Dampening the clearance system for peptide ligands is a promising strategy for increasing the efficacy of peptide therapies. Natriuretic peptide receptor B (NPR-B) and its naturally occurring ligand, C-type natriuretic peptide (CNP), are potent stimulators of endochondral bone growth, and activating the CNP/NPR-B system is expected to be a powerful strategy for treating impaired skeletal growth. CNP is cleared by natriuretic peptide clearance receptor (NPR-C); therefore, we investigated the effect of reducing the rate of CNP clearance on skeletal growth by limiting the interaction between CNP and NPR-C. Specifically, we generated transgenic mice with increased circulating levels of osteocrin (OSTN) protein, a natural NPR-C ligand without natriuretic activity, and observed a dose-dependent skeletal overgrowth phenotype in these animals. Skeletal overgrowth in OSTN-transgenic mice was diminished in either CNP- or NPR-C-depleted backgrounds, confirming that CNP and NPR-C are indispensable for the bone growth-stimulating effect of OSTN. Interestingly, double-transgenic mice of CNP and OSTN had even higher levels of circulating CNP and additional increases in bone length, as compared with mice with elevated CNP alone. Together, these results support OSTN administration as an adjuvant agent for CNP therapy and provide a potential therapeutic approach for diseases with impaired skeletal growth.

C-type natriuretic peptide restores impaired skeletal growth in a murine model of glucocorticoid-induced growth retardation.

Glucocorticoids are widely used for treating autoimmune conditions or inflammatory disorders. Long-term use of glucocorticoids causes impaired skeletal growth, a serious side effect when they are used in children. We have previously demonstrated that C-type natriuretic peptide (CNP) is a potent stimulator of endochondral bone growth. In this study, we investigated the effect of CNP on impaired bone growth caused by glucocorticoids by using a transgenic mouse model with an increased circulating CNP level. Daily administration of a high dose of dexamethasone (DEX) to 4-week-old male wild-type mice for 4weeks significantly shortened their naso-anal length, which was restored completely in DEX-treated CNP transgenic mice. Impaired growth of the long bones and vertebrae by DEX was restored to a large extent in the CNP transgenic background, with recovery in the narrowed growth plate by increased cell volume, whereas the decreased proliferation and increased apoptosis of the growth plate chondrocytes were unaffected. Trabecular bone volume was not changed by DEX treatment, but decreased significantly in a CNP transgenic background. In young male rats, the administration of high doses of DEX greatly decreased N-terminal proCNP concentrations, a marker of CNP production. In organ culture experiments using fetal wild-type murine tibias, longitudinal growth of tibial explants was inhibited by DEX but reversed by CNP. These findings now warrant further study of the therapeutic potency of CNP in glucocorticoid-induced bone growth impairment.

Increased Bone Turnover and Possible Accelerated Fracture Healing in a Murine Model With an Increased Circulating C-Type Natriuretic Peptide.

Recent studies have revealed that C-type natriuretic peptide (CNP) is a potent stimulator of endochondral bone growth. Nevertheless, the effect of CNP on bone turnover has not yet been well studied. To elucidate this issue, we investigated the bone phenotype of a mouse model with elevated plasma CNP concentrations (SAP-CNP-Tg mice) in the present study. Microcomputed tomography (CT) analysis revealed less bone in femurs, but not in lumber vertebrae, of young adult SAP-CNP-Tg mice than that of wild-type mice. Bone histomorphometry of the tibiae from 8-week-old SAP-CNP-Tg mice showed enhanced osteoblastic and osteoclastic activities, in accordance with elevated serum levels of osteocalcin and tartrate-resistant acid phosphatase-5b, respectively. Next we performed an open and stabilized femoral fracture using 8-week-old SAP-CNP-Tg mice and compared the healing process with age-matched wild-type mice. An immunohistochemical study revealed that CNP and its receptors, natriuretic peptide receptor-B and natriuretic peptide clearance receptor, are expressed in hard calluses of wild-type mice, suggesting a possible role of CNP/natriuretic peptide receptor-B signaling in fracture repair, especially in bone remodeling stage. On micro-CT analysis, a rapid decrease in callus volume was observed in SAP-CNP-Tg mice, followed by a generation of significantly higher new bone volume with a tendency of increased bone strength. In addition, a micro-CT analysis also showed that bone remodeling was accelerated in SAP-CNP-Tg mice, which was also evident from increased serum osteocalcin and tartrate-resistant acid phosphatase-5b levels in SAP-CNP-Tg mice at the remodeling stage of fracture repair. These results indicate that CNP activates bone turnover and remodeling in vivo and possibly accelerates fracture healing in our mouse model.

Forkhead box A1 (FOXA1) and A2 (FOXA2) oppositely regulate human type 1 iodothyronine deiodinase gene in liver.

Type 1 iodothyronine deiodinase (D1), a selenoenzyme that catalyzes the bioactivation of thyroid hormone, is expressed mainly in the liver. Its expression and activity are modulated by several factors, but the precise mechanism of its transcriptional regulation remains unclear. In the present study, we have analyzed the promoter of human D1 gene (hDIO1) to identify factors that prevalently increase D1 activity in the human liver. Deletion and mutation analyses demonstrated that a forkhead box (FOX)A binding site and an E-box site within the region between nucleotides -187 and -132 are important for hDIO1 promoter activity in the liver. EMSA demonstrated that FOXA1 and FOXA2 specifically bind to the FOXA binding site and that upstream stimulatory factor (USF) specifically binds to the E-box element. Overexpression of FOXA2 decreased hDIO1 promoter activity, and short interfering RNA-mediated knockdown of FOXA2 increased the expression of hDIO1 mRNA. In contrast, overexpression of USF1/2 increased hDIO1 promoter activity. Short interfering RNA-mediated knockdown of FOXA1 decreased the expression of hDIO1 mRNA, but knockdown of both FOXA1 and FOXA2 restored it. The response of the hDIO1 promoter to USF was greatly attenuated in the absence of FOXA1. Taken together, these results indicate that a balance of FOXA1 and FOXA2 expression modulates hDIO1 expression in the liver.