Research ArticleMOLECULAR BIOLOGY

Thyroid hormone receptor and ERRα coordinately regulate mitochondrial fission, mitophagy, biogenesis, and function

See allHide authors and affiliations

Science Signaling  26 Jun 2018:
Vol. 11, Issue 536, eaam5855
DOI: 10.1126/scisignal.aam5855

Two nuclear receptors for mitochondria

Mitochondrial processes such as fission, mitophagy, and biogenesis are regulated by thyroid hormone receptor (THR) and ERRα, both of which are nuclear receptors that alter gene expression when bound to ligand. Noting that many THR target genes lack response elements for this nuclear receptor, Singh et al. investigated the regulation of genes involved in mitochondrial pathways by THR. Using a human liver cell line and mice treated with TH in the absence or presence of an ERRα inhibitor, the authors found that THR mediated its effects on mitochondrial processes by increasing the expression of the gene encoding ERRα. These results suggest that agonist activation of ERRα could be used to improve mitochondrial quality in metabolic or neurodegenerative diseases or aging.

Abstract

Thyroid hormone receptor β1 (THRB1) and estrogen-related receptor α (ESRRA; also known as ERRα) both play important roles in mitochondrial activity. To understand their potential interactions, we performed transcriptome and ChIP-seq analyses and found that many genes that were co-regulated by both THRB1 and ESRRA were involved in mitochondrial metabolic pathways. These included oxidative phosphorylation (OXPHOS), the tricarboxylic acid (TCA) cycle, and β-oxidation of fatty acids. TH increased ESRRA expression and activity in a THRB1-dependent manner through the induction of the transcriptional coactivator PPARGC1A (also known as PGC1α). Moreover, TH induced mitochondrial biogenesis, fission, and mitophagy in an ESRRA-dependent manner. TH also induced the expression of the autophagy-regulating kinase ULK1 through ESRRA, which then promoted DRP1-mediated mitochondrial fission. In addition, ULK1 activated the docking receptor protein FUNDC1 and its interaction with the autophagosomal protein MAP1LC3B-II to induce mitophagy. siRNA knockdown of ESRRA, ULK1, DRP1, or FUNDC1 inhibited TH-induced autophagic clearance of mitochondria through mitophagy and decreased OXPHOS. These findings show that many of the mitochondrial actions of TH are mediated through stimulation of ESRRA expression and activity, and co-regulation of mitochondrial turnover through the PPARGC1A-ESRRA-ULK1 pathway is mediated by their regulation of mitochondrial fission and mitophagy. Hormonal or pharmacologic induction of ESRRA expression or activity could improve mitochondrial quality in metabolic disorders.

INTRODUCTION

Mitochondria have a central role in cellular energy metabolism, and their dysfunction contributes to the pathologic features of metabolic disorders and aging (1). In metabolically active tissues (such as liver, muscles, and brown fat), major mitochondrial metabolic pathways, such as fatty acid oxidation (FAO) and organic acid recycling from the tricarboxylic acid (TCA) cycle, contribute to adenosine triphosphate (ATP) generation by oxidative phosphorylation (OXPHOS) (2). However, the latter process can generate reactive oxygen species (ROS) that lead to mitochondria damage and dysfunction, necessitating their repair and replacement. New mitochondria can be generated from preexisting ones through complex and tightly regulated processes such as fission and fusion, whereas autophagy of mitochondria (a process called mitophagy) is often accompanied by the biogenesis of new mitochondria (3). In particular, the mitochondrial turnover involving mitochondrial biogenesis and mitophagy plays a substantial role in maintaining mitochondrial quality control (1, 3, 4). Currently, the molecular mechanisms that regulate mitochondrial homeostasis through balanced mitochondrial biogenesis and mitophagy under different physiological and hormonal conditions are not well understood (5).

Nuclear receptors (NRs) regulate the transcription of genes involved in mitochondrial biogenesis in a tissue-specific manner (6, 7). Estrogen-related receptor α (ESRRA or NR3B1; also known as ERRα) is an NR that regulates genes involved in mitochondrial biogenesis, FAO, TCA, OXPHOS, and other processes (7) in the presence of peroxisome proliferator–activated receptor γ coactivator 1α (PPARGC1A; or PGC1α) (8, 9). The ligand for ESRRA is not known, so it has been classified as an orphan NR; however, cholesterol can directly activate ESRRA (10). Besides ESRRA, thyroid hormone (TH) receptors (THRs) also regulate mitochondria biogenesis by inducing PPARGC1A and other key mitochondrial pathways in a tissue-specific manner (1113). THs (also known as T3 and T4, which are ligands for THRs) induce ROS-mediated autophagic clearance of mitochondria (14, 15). As transcription factors, both ESRRA and THRs bind to their cognate response elements in the promoters of target genes to regulate their transcription (13, 1618). However, many target genes regulated by liganded THRs do not appear to have bona fide TH response elements; thus, these genes may be regulated by other transcription factors that are transcriptionally regulated or posttranslationally activated by TH (1922). Therefore, coordinated regulation of THRs with other NRs or transcription factors may be an important mechanism for controlling a wider transcriptional network of target genes.

Here, we found that TH increased ESRRA expression through induction of PPARGC1A expression in a THRB1-dependent manner. We showed that major mitochondrial pathways such as OXPHOS, FAO, and TCA cycle were co-regulated by THRB1, PPARGC1A, and ESRRA. Moreover, TH activation of ESRRA was required for coordinated regulation of mitochondrial biogenesis, fission, and mitophagy. We found that ESRRA regulated the key mitophagic protein, Unc-51–like autophagy activating kinase 1 (ULK1), which then recruited DRP1 to induce fission and activate the docking receptor protein FUNDC1. This, in turn, initiated the recruitment of autophagy proteins to mitochondria to activate mitophagy. Our findings have highlighted the central role of ESRRA in mitochondrial function and turnover. Besides regulation by TH, it is possible that activation of ESRRA by agonists could be a key pathway for enhancing mitochondrial quality, particularly under conditions in which there is a decline in mitochondrial number and quality such as metabolic diseases and aging.

RESULTS

Transcriptome and ChIP-seq analyses showed that mitochondrial pathways were co-regulated by both THRB1 and ESRRA

To understand global co-regulation of genes by THRB1 and ESRRA, we performed hepatic transcriptome analysis of untreated euthyroid mice and those treated with TH alone or with 3-[4-(2,4-bis-trifluoromethylbenzyloxy)-3-methoxyphenyl]-2-cyano-N-(5-trifluoromethyl-1,3,4-thiadiazol-2-yl) acrylamide (XCT790), an ESRRA-specific inverse agonist that induces proteasomal degradation of ESRRA (fig. S1A) (9, 23, 24). Key mitochondrial pathways such as OXPHOS, Parkinson’s disease (which contains mitochondrial genes), TCA, glycosphingolipids, and Janus kinase/signal transducer and activator of transcription (JAK/STAT) showed significantly increased expression in the livers from TH-treated mice compared to those from mice treated with both TH and XCT790 (Fig. 1A). More than 50% of the genes in each pathway (except TCA cycle at 42.9%) were co-regulated by TH and ESRRA (Fig. 1B and fig. S1B). THRB1 and ESRRA chromatin immunoprecipitation sequencing (ChIP-seq) analysis showed that both THRB1 and ESRRA NRs bound to the promoters of more than 50% of the transcriptionally co-regulated genes in each pathway (Fig. 1C, fig. S1C, and data file S1).

Fig. 1 Transcriptome and ChIP-seq analyses of THRB1 and ESRRA showing co-regulation of mitochondrial pathways.

(A) Gene Set Enrichment Analysis (GSEA) pathway analysis of transcriptomics performed in the livers from euthyroid control, TH-treated (10 μg/100 g body weight per day for 3 days), and TH + XCT790 (10 mg/day per 100 g body weight for 3 days)–treated mice (n = 3 mice per group). (B) Venn diagram representing overlapping genes between groups from (A) under various pathways. (C) THRB1 and ESRRA binding on the genes from (A) and (B) (n = 4 livers per group). (D to G) Mitochondrial oxygen consumption rate (OCR) using a Seahorse extracellular flux analyzer under TH treatment (100 nm for 48 hours) with or without ESRRA siRNA in THRB1-HepG2 cells (n = 5 replicates per group). Mitochondrial inhibitors [1 μM oligomycin, carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), rotenone (R), or antimycin A (A)] from Seahorse XF Cell Mito Stress Test kit were used. Data are representative of three or more independent experiments. Statistical significance for the data presented in (A) to (C) was calculated as described in Materials and Methods, and that in (E) and (G) was determined by using Kruskal-Wallis nonparametric analysis of variance (ANOVA), followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment (*P ≤ 0.05).

Analyses of mitochondrial activities co-regulated by THRB1 and ESRRA showed that TH effects depended on ESRRA in cell culture and in vivo

Ectopic expression of THRB1 in HepG2 cells (THRB1-HepG2 cells) results in enhanced TH response upon gene activation (19). TH significantly increased mitochondrial basal oxygen consumption, ATP turnover, maximum respiratory capacity, and spare respiratory capacity in THRB1-HepG2 cells, effects that were inhibited by transfection of ESRRA small interfering RNA (siRNA) (Fig. 1, D to G, and fig. S1, D and E). Moreover, the TH-dependent increase in OXPHOS was enhanced by transient overexpression of ESRRA, an effect that was blocked by XCT790 treatment, thus demonstrating the critical role of ESRRA in TH-dependent induction of mitochondrial OXPHOS.

We performed metabolomics analysis of TCA metabolites and acylcarnitines to understand the regulation of TCA cycle flux and β-oxidation of fatty acids by TH and ESRRA (fig. S2, A to C). We observed a modest increase in TCA cycle intermediates (citrate, α-ketoglutarate, and pyruvate) in the livers of TH-treated mice compared to those of euthyroid mice. This effect was inhibited by XCT790 treatment, suggesting that the increase in TCA cycle flux in TH-treated mice depended on ESRRA activity. Carnitine palmitoyltransferase 1α (CPT1α) catalyzes the first step of long-chain fatty acid import into mitochondria, and its activity is believed to be rate-limiting for β-oxidation of fatty acids (25). Transcript levels of Cpt1A (which encodes CPT1α) were higher in the livers of TH-treated mice than in those of euthryoid mice (Fig. 2A). Consistently, serum β-hydroxybutyrate (a ketogenic end product) and short-chain acylcarnitines (which are mitochondrial FAO end products) were also increased significantly in the liver tissues of mice treated with TH (Fig. 2, B and C, and fig. S2B). XCT790 treatment significantly reduced hepatic Cpt1A mRNA expression, serum β-hydroxybutyrate levels, and hepatic short-chain acylcarnitines in TH-treated mice (Fig. 2, A to C, and fig. S2B). In addition, TH- and XCT790-treated mice showed hepatic accumulation of medium- and long-chain acylcarnitines and decreased short-chain acylcarnitines (fig. S2C), suggesting decreased acylcarnitine flux and downstream inhibition of mitochondrial β-oxidation of fatty acids compared to mice treated with TH alone. Moreover, TH also increased CPT1A mRNA expression in THRB1-HepG2 cells, an effect that was blunted significantly by genetic ablation of ESRRA (Fig. 2D). We next examined triglyceride clearance in fatty acid–loaded, 4,4-difluoro-1,3,5,7,8-pentamethyl-4-bora-3a,4a-diaza-s-indacene (BODIPY)–stained THRB1-HepG2 cells and found that TH increased triglyceride clearance in an ESRRA-dependent manner (Fig. 2, E and F).

Fig. 2 TH-mediated induction of mitochondrial FAO and fat clearance depended on ESRRA.

(A) Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis of hepatic Cpt1a transcript expression (n = 5 mice per group). (B) Measurement of serum β-hydroxybutyrate/ketone bodies (n = 5 mice per group). (C) Metabolomics analysis of short-chain acylcarnitine C4-OH (d-3-hydroxybutyrylcarnitine, a derivative of ketone bodies) in euthyroid, TH-treated (10 μg/100 g body weight per day for 3 days), XCT790-treated (10 mg/day per 100 g body weight for 3 days), and XCT790 + TH–treated mice (n = 4 mice per group). (D) RT-qPCR analysis of CPT1A transcript expression in THRB1-HepG2 cells treated with control or ESRRA siRNA in the presence or absence of TH (100 nM for 48 hours) (n = 5 biological replicates per group). Data are representative of three or more independent experiments. (E and F) BODIPY 493/503 staining (E) and its quantification by flow cytometry (F) for neutral lipids and lipid droplets in fat-loaded THRB1-HepG2 cells treated with control siRNA or ESRRA siRNA in the presence or absence of TH (20,000 counts per acquisition). Quantification of BODIPY 493/503 fluorescence by flow cytometer is plotted as mean fluorescence intensity (MFI) (n = 5 biological replicates per group). Images and data are representative of three fields per group and three independent experiments. Scale bars, 200 μm. (G) Measurement of mitochondrial fuel oxidation (fatty acids; FAO) in fat-loaded (oleate/palmitate) THRB1-HepG2 cells treated with control or ESRRA siRNA in the presence or absence of TH (n = 5 biological replicates per group). Data are representative of three independent experiments. Seahorse XF Mito Fuel Flex Test kit was used with inhibitors of the glucose oxidation pathway (UK5099), glutamine oxidation pathway (BPTES), and CPT1A (long-chain FAO pathway; etoximir). Various parameters such as the measurement of cells’ reliance on a particular fuel pathway to maintain baseline respiration (dependency), cells’ ability to increase oxidation of a particular fuel to compensate for inhibition of alternative fuel pathway(s) (flexibility), and the cells’ total capability to use a fuel pathway to meet energy demand when other fuel pathways are inhibited (capacity) were calculated. Statistical significance was determined by using Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment (*P ≤ 0.05). ns, not significant.

To determine whether the triglyceride clearance was due to increased β-oxidation of fatty acids, we used the Seahorse XF Mito Fuel Flex Test to measure mitochondrial fuel usage in live cells. TH treatment increased the cellular dependency on FAO (Fig. 2G and fig. S2D) and further increased the total capacity of FAO in fatty acid–loaded cells in THRB1-HepG2 cells (Fig. 2G). These findings were consistent with our findings showing the induction of CPT1A mRNA expression by TH (Fig. 2D). Transfection of ESRRA siRNA inhibited the cellular dependency on FAO induced by TH, and the total FAO capacity returned to the initial baseline after fatty acid loading. Together, these data suggested that TH induced many key mitochondrial activities such as OXPHOS, TCA cycle flux, and β-oxidation of fatty acids in an ESRRA-dependent and cell-autonomous manner.

TH induced ESRRA expression through THRB1

We next used in vivo and in vitro models to determine whether TH regulated ESRRA expression. We found that TH significantly increased ESRRA transcript and ESRRA protein levels in mouse liver, primary hepatocytes, and THRB1-HepG2 cells (Fig. 3, A and B, and fig. S3, A and B). There was also a time-dependent increase in ESRRA protein in the livers from mice treated with TH (Fig. 3C), starting at 24 hours after TH treatment and lasting up to 10 days, suggesting that ESRRA functioned as an important mediator of chronic TH action. Furthermore, these TH-mediated increases in Esrra mRNA and ESRRA protein were THRB-dependent because they were abolished in the livers from Thrb knockout (Thrb−/−) mice but not in those from wild-type (Thrb+/+) mice (Fig. 3, D and E). Similarly, THRB1 siRNA knockdown in HepG2 cells inhibited TH-induced ESRRA expression (fig. S3, B and C).

Fig. 3 TH increased ESRRA gene expression by stimulating PPARGC1A expression.

(A) RT-qPCR analysis of Esrra transcript expression in the livers from euthyroid control and TH-treated (10 μg/100 g body weight per day for 3 days) mice (n = 15 mice per group). (B) Western blot analysis of hepatic ESRRA protein expression in euthyroid and TH-treated mice (n = 20 mice per group). Blots are representative of three or more independent experiments. (C) Western blot analysis of hepatic ESRRA protein expression in euthyroid and TH-treated mice treated for the indicated periods of time (n = 5 mice per group). Blots are representative of three or more independent experiments. (D) RT-qPCR analysis of hepatic Esrra transcript expression in Thrb wild-type (Thrb+/+) mice and liver-specific Thrb knockout (Thrb−/−) treated with TH or propylthiouracil (PTU; to make mice hypothyroid) (n = 5 mice per group). (E) Western blot analysis of ESRRA protein expression in liver tissues of Thrb+/+ and liver-specific Thrb−/− mice treated with TH or PTU (n = 5 mice per group). Blots are representative of three or more independent experiments. (F) RT-qPCR analysis of hepatic Ppargc1a transcript expression in euthyroid and TH-treated mice (n = 15 mice per group). (G) RT-qPCR analysis of PPARGC1A and ESRRA transcript expression in untreated control and TH-treated (100 nM for 48 hours) THRB1-HepG2 cells with or without PPARGC1A knockdown (n = 5 biological replicates per group). Data are representative of three independent experiments. (H) RT-qPCR analysis of PPARGC1A and ESRRA transcript expression in untreated control and TH-treated THRB1-HepG2 cells with or without PPARGC1A overexpression (n = 5 biological replicates per group). Data are representative of three independent experiments. (I) Luciferase reporter analysis of ESRRA promoter activity in THRB1-HepG2 cells treated with TH by the conditions indicated in the panel below the bar graph (n = 6 biological replicates per group). Data are representative of three independent experiments. RLU, relative light units. (J) ESRRA and POLR2A ChIP-qPCR analysis on Esrra gene promoter in mouse liver tissue from euthyroid control, TH-treated, and TH + XCT790 (10 mg/day per 100 g body weight for 3 days)–treated mice (n = 3 mice per group). Statistical significance for (A), (B), and (F) was calculated using the nonparametric Mann-Whitney test, whereas for other panels, Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment, was performed (*P ≤ 0.05).

TH induced ESRRA expression by stimulating PPARGC1A expression

ESRRA mRNA expression is induced by PPARGC1A, which also interacts with ESRRA (26, 27) to form a heterodimer that mediates ESRRA transcriptional activity on target genes. Furthermore, the ESRRA gene is regulated by ESRRA and PPARGC1A in a feed-forward manner (27, 28). We found that TH increased PPARGC1A transcripts in mouse livers, primary hepatocytes, and THRB1-HepG2 cells (Fig. 3, F and G, and fig. S3A). PPARGC1A siRNA knockdown in HepG2 cells significantly inhibited the induction of ESRRA mRNA expression by TH (Fig. 3G and fig. S3D), suggesting that the TH-mediated induction of ESRRA was PPARGC1A-dependent. Conversely, transient overexpression of PPARGC1A increased TH-induced ESRRA mRNA expression (Fig. 3H). In addition, we found that TH significantly increased the activity of an ESRRA promoter luciferase reporter (26) transfected into THRB1-HepG2 cells, an effect that was inhibited by XCT790 (Fig. 3I). In contrast, TH did not change the activity of an ESRRA promoter reporter containing a mutation in the estrogen-related receptor response element (ERRE) consensus sequence that abrogates ESRRA binding. Finally, knockdown of PPARGC1A significantly inhibited basal and TH-induced ESRRA promoter activity, whereas transient overexpression of PPARGC1A enhanced TH-induced ESRRA promoter activity (Fig. 3I). Furthermore, TH significantly increased ESRRA and POLR2A recruitment to the Esrra gene promoter in the liver to activate its own transcription (Fig. 3J). TH-induced recruitment of ESRRA and POLR2A to the ERRE and TATA box, respectively, was abolished when ESRRA was inhibited by XCT790. We further confirmed these findings in THRB1-HepG2 cells (fig. S3E) with PPARGC1A knockdown, which significantly reduced TH-dependent ESRRA recruitment to ESRRA gene promoter. Moreover, PPARGC1A knockdown inhibited TH-induced mitochondrial basal oxygen consumption, ATP turnover, and maximum respiratory capacity to levels comparable to those observed after ESRRA knockdown (fig. S3, F and G). Collectively, these data suggested that TH stimulation of ESRRA mRNA expression and ESRRA-mediated OXPHOS occurred through induction of PPARGC1A.

TH stimulated mitochondrial turnover through ESRRA

Although TH and ESRRA both stimulate mitochondrial biogenesis by inducing PPARGC1A (2729), the role of ESRRA in TH-mediated mitochondrial biogenesis was not previously known, so we further analyzed this interaction. Mitochondrial DNA copy number was greater in the livers of TH-treated mice compared to those of euthyroid mice (Fig. 4A). In addition, the abundance of transcripts of key mitochondrial biogenesis regulators such as Tfam and Nrf1 was increased in the livers from TH-treated mice (fig. S4A). Similarly, TH increased mitochondrial DNA copy number and TFAM and NRF1 transcript abundance in THRB1-HepG2 cells (Fig. 4, B and C, and fig. S4B). These effects were inhibited by XCT790 cotreatment of TH-treated mice (Fig. 4A and fig. S4A) or by siRNA knockdown of ESRRA or XCT790 treatment in THRB1-HepG2 cells (Fig. 4, B and C). Together, these data demonstrate that TH increased mitochondrial biogenesis in an ESRRA-dependent manner.

Fig. 4 TH induced mitochondrial biogenesis and fission through ESRRA.

(A) Mitochondrial DNA copy number analysis in the livers from euthyroid control, TH-treated (10 μg/100 g body weight per day for 3 days), and TH + XCT790 (10 mg/day per 100 g body weight for 3 days)–treated mice (n = 5 mice per group). (B and C) Mitochondrial DNA copy number analysis in THRB1-HepG2 cells treated with ESRRA siRNA or XCT790 (2 μM) in the presence of absence of TH (100 nM) for 48 hours (n = 5 biological replicates per group). Data are representative of three independent experiments. (D to H) Transmission electron microscopy of liver tissues from euthyroid (D), TH-treated (E to G) and XCT790 + TH–treated mice (H) (n = 5 mice per group). Scale bars, 0.5 μm (D, E, and H) and 100 nm (F and G). N, nucleus; A, autophagic vesicles (white arrows). Asterisk (*) indicates mitochondria or mitochondria-like structures inside autophagic vesicles. (I and J) Mitochondrial numbers and size (I) and autophagic vesicles containing mitochondria (J) were counted. Data are representative of 10 fields per group and three independent analyses. (K) MitoTracker Red CMXRos staining was performed in THRB1-HepG2 cells to analyze mitochondrial structure. These THRB1-HepG2 cells were treated with control or ESRRA siRNA in the presence or absence of TH (100 nM for 48 hours) (n = 5 biological replicates per group). Enlarged panels represent selected digitally enlarged portions of parent images to enhance the visibility of mitochondrial structure. Images are representative of 10 fields per group and three independent experiments. Scale bars, 5 μm. Statistical significance for (I) and (J) was calculated using one-way ANOVA, followed by Tukey’s post hoc test; otherwise, Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment, was performed (*P ≤ 0.05 and #P ≤ 0.05). In (I), asterisk (*) represents the comparison between euthyroid and TH-treated mice or as indicated by the horizontal line, and number sign (#) represents the comparison between TH-treated and XCT790 + TH–treated mice.

Furthermore, the livers from TH-treated mice had significantly more mitochondria that were smaller in size (<0.5 μm) compared to those from euthyroid or TH- and XCT790-treated mice (Fig. 4, D to I). These findings suggest that TH increased mitochondrial fission in an ESRRA-dependent manner. Consistent with our previous findings (15, 30), we also observed an increased number of autophagic vesicles containing mitochondria or mitochondria-like structures in the livers from TH-treated mice compared to those in euthyroid mice (Fig. 4, D to H and J), an effect that was reduced by XCT790 treatment (Fig. 4, D to H). Moreover, MitoTracker staining of THRB1-HepG2 cells showed a network of rod-shaped mitochondria in control cells and cells cotreated with TH and ESRRA siRNA, whereas fragmented mitochondria (puncta) were seen in cells treated only with TH (Fig. 4K). TH induced the expression of genes regulating mitochondrial fission (Drp1) and mitophagy (Pink1 and Parkin) without altering that of mitochondrial fusion genes (Opa1 and Mfn2) in mouse livers (fig. S4C). In contrast, hepatic expression of these transcripts in XCT790 or TH- and XCT790-treated mice was similar to that in euthyroid mice. Together, these results show that TH increased mitochondrial fission and mitophagy in an ESRRA-dependent manner.

TH induced ULK1 mRNA and ULK1 protein through ESRRA-dependent transcription

ULK1 recognizes damaged mitochondria and facilitates mitophagic degradation in response to various stressors (3135). XCT790 significantly inhibited TH-mediated induction of Ulk1 mRNA and ULK1 protein expression in mouse liver (Fig. 5, A and B). Similarly, ESRRA siRNA significantly reduced the TH-dependent increase in ULK1 mRNA and ULK1 protein expression in THRB1-HepG2 cells (Fig. 5, C and D). Pharmacological inhibition of ESRRA in vivo and ESRRA siRNA knockdown in vitro also decreased basal ULK1 expression (Fig. 5, A and D). ChIP analysis of mouse liver and THRB1-HepG2 cells showed that TH significantly increased the recruitment of ESRRA and POLR2A to the Ulk1/ULK1 promoter (Fig. 5E and fig. S5). Furthermore, ESRRA inhibition by XCT790 in mice or in THRB1-HepG2 cells and PPARGC1A siRNA knockdown in THRB1-HepG2 cells significantly decreased ESRRA recruitment by TH on the ULK1 promoter (Fig. 5E and fig. S5). Collectively, these data suggest that ESRRA specifically mediated the induction of ULK1 expression by TH.

Fig. 5 Induction of ULK1 expression by TH was mediated by ESRRA.

(A) Western blot analysis and densitometric quantification of hepatic ULK1 and ESRRA protein expression in the livers from euthyroid control, TH-treated (10 μg/100 g body weight per day for 3 days), and TH + XCT790 (10 mg/day per 100 g body weight for 3 days)–treated mice (n = 5 mice per group). Blots are representative of three independent experiments. (B) RT-qPCR analysis of hepatic ULK1 transcript expression in euthyroid control, TH-treated, and TH + XCT790–treated mice (n = 5 mice per group). Data are representative of three independent experiments. (C) Western blot analysis and densitometric quantification of ESRRA and ULK1 expression in THRB1-HepG2 cells treated with control or ESRRA siRNA with or without TH (100 nM for 48 hours) (n = 5 biological replicates per group). Blots are representative of three or more independent experiments. (D) RT-qPCR analysis of ULK1 transcript expression in THRB1-HepG2 cells treated with control or ESRRA siRNA in the presence or absence of TH (n = 5 biological replicates per group). Data are representative of three independent experiments. (E) ESRRA and POLR2A ChIP-qPCR analysis on the Ulk1 gene promoter in euthyroid, TH-treated, and XCT790 + TH–treated mouse livers (n = 3 mice per group). Statistical significance was calculated using Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment (*P ≤ 0.05).

TH induction of DRP1 translocation to mitochondria, mitochondrial fission, and mitophagy depended on ESRRA

ULK1 regulates DRP1 activity by phosphorylating Ser616 to induce mitochondrial fission (36). Furthermore, translocation of ULK1 to damaged mitochondria activates the docking receptor protein FUNDC1 and facilitates its binding to the autophagosomal protein MAP1LC3B-II, a necessary process for mitophagy (3234). We found increased phosphorylation of DRP1 at Ser616 (Fig. 6A) that was associated with increased ULK1 and FUNDC1 abundance in the livers from TH-treated mice. XCT790 cotreatment significantly reduced the TH-induced increases in ULK1 abundance and DRP1 phosphorylation but did not change FUNDC1 abundance. Subcellular fractionation revealed that TH treatment of mice significantly increased the translocation of ULK1, DRP1, MAP1LC3B-II, and Sequestosome 1 (SQSTM1) to the mitochondrial fraction in the liver (Fig. 6B), an effect that was reduced when the mice were also treated with XCT790. Transmission electron microscopy confirmed the localization of mitochondria within autophagic vesicles (Fig. 4, D to H and J).

Fig. 6 Coordinated fission and mitophagy mediated by TH depended on ESRRA.

(A) Western blot analysis and densitometric quantification of protein abundance in the livers from euthyroid control, TH-treated (10 μg/100 g body weight per day for 3 days), and TH + XCT790 (10 mg/day per 100 g body weight for 3 days)–treated mice (n = 5 mice per group). Blots are representative of three independent experiments. (B) Western blot analysis and densitometric quantification of cytosolic (Cyto) and mitochondrial (Mito) fractions from the livers of euthyroid, TH-treated, and XCT790 + TH–treated mice livers (n = 5 mice per group). Blots are representative of three independent experiments. (C) Western blot analysis and densitometric quantification of DRP1 phosphorylation in whole-cell lysates of THRB1-HepG2 cells treated with or without TH (100 nM) and the ULK1 inhibitor (iULK1) MRT0068921 (1 μM) for 48 hours (n = 5 biological replicates per group). Blots are representative of three independent experiments. (D) Western blot analysis and densitometric quantifications of cytosolic and mitochondrial fractions from THRB1-HepG2 cells treated with or without TH and the ULK1 inhibitor MRT0068921 to observe mitochondrial translocation of proteins (n = 5 biological replicates per group). Blots are representative of three independent experiments. (E) Coimmunoprecipitation analysis was performed to detect the interaction of FUNDC1 with ULK1 and MAP1LC3B-II (n = 5 biological replicates per group). Blots are representative of three independent experiments. Statistical significance was calculated using Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment (*P ≤ 0.05). IP, immunoprecipitation.

Consistent with our in vivo findings, TH significantly increased DRP1 phosphorylation at Ser616 and increased the recruitment of DRP1 and autophagy proteins to the mitochondrial fraction in THRB1-HepG2 cells, an effect that was blocked by coapplication of the ULK1 inhibitor (Fig. 6, C and D). FUNDC1 abundance in the mitochondrial fraction did not change because it is a resident protein of the mitochondrial outer membrane. Pharmacological inhibition of DRP1 by Mdivi-1 significantly reduced the recruitment of DRP1, MAP1LC3B-II, and SQSTM1 to mitochondria in TH-treated cells (fig. S6), suggesting that DRP1 phosphorylation precedes these events. Mdivi-1 did not block TH-induced ULK1 translocation to mitochondria, suggesting that this event is upstream of DRP1 phosphorylation. Coimmunoprecipitation analysis demonstrated that TH significantly increased the interaction of FUNDC1 with ULK1 and MAP1LC3B-II, which was inhibited by XCT790 (Fig. 6E). No interaction of FUNDC1 with DRP1 was detected.

Immunofluorescence studies in THRB1-HepG2 cells showed that TH significantly increased DRP1 translocation to mitochondria (as evidenced by the DRP1-MitoTracker overlay) (Fig. 7, A and B). This translocation was inhibited by siRNA knockdown of ESRRA or pharmacological inhibition of ULK1 and DRP1, suggesting that both ESRRA and ULK1 were required. siRNA-mediated knockdown of FUNDC1 did not significantly affect TH-mediated DRP1 translocation, suggesting that FUNDC1 activation occurred downstream of DRP1 phosphorylation and translocation. FUNDC1 knockdown led to the accumulation of puncta, which represented fragmented mitochondria caused by fission. TH significantly increased MAP1LC3B recruitment to mitochondria (as evidenced by the MAP1LC3B-MitoTracker overlay) (Fig. 7, C and D). Knockdown of ESRRA or FUNDC1 or pharmacological inhibition of ULK1 and DRP1 blocked TH-induced MAP1LC3B recruitment to mitochondria (Fig. 7, C and D).

Fig. 7 Induction of DRP1 and MAP1LC3B mitochondrial localization and mitophagic flux by TH was dependent on ESRRA.

THRB1-HepG2 cells were treated with control siRNA with or without TH (100 nM for 48 hours) or with ESRRA siRNA, FUNDC1 siRNA, or inhibitors of ULK1 (iULK1; 1 μM) or DRP1 (iDRP1; 10 μM). MitoTracker Red CMXRos was used to stain mitochondria. (A) Immunofluorescence analysis for DRP1 translocation to mitochondria. DRP1 antibody (#8570, Cell Signaling Technology) with Alexa Fluor 488 was used to detect DRP1 by confocal microscopy (n = 5 biological replicates per group). Enlarged panels represent selected digitally enlarged portions of parent images to enhance detection of DRP1 translocation to mitochondria. Images are representative of three fields per group and three independent experiments. Scale bars, 5 μm. (B) Quantification of DRP1 mitochondrial translocation (as % overlay). (C) Immunofluorescence analysis for MAP1LC3B and mitochondrial colocalization. MAP1LC3B antibody (#2775, Cell Signaling Technology) with Alexa Fluor 488 was used to detect MAP1LC3B under a confocal microscope (n = 5 biological replicates per group). Enlarged panels represent selected digitally enlarged portions of parent images to enhance the visibility of MAP1LC3B translocation to mitochondria. Images are representative of three fields per group and three independent experiments. Scale bars, 5 μm. (D) Colocalization of MAP1LC3B and mitochondria (as % overlay) was quantified. (E) Analysis of mitophagic flux (red puncta compared to yellow). Cells were transfected with siRNAs and then with the plasmid pAT016 (p-mito-mRFP-EGFP) and treated with TH (n = 5 biological replicates per group). Enlarged panels represent selected digitally enlarged portions of parent images to enhance the visibility of mitochondria inside autolysosome. Images are representative of three fields per group and three independent experiments. Scale bars, 5 μm. (F) Mitophagic flux was quantified. Statistical significance was calculated using Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment (*P ≤ 0.05).

Induction of mitophagic flux and mitochondrial activity by TH and ESRRA depended on ULK1, DRP1, and FUNDC1

To analyze the progression of mitophagy from autophagosomes to lysosomes and measure mitophagic flux, we used a plasmid (pAT016) containing a mitochondrial localization sequence fused in-frame with red fluorescent protein (RFP) and enhanced green fluorescent protein (GFP) coding sequences (Fig. 7, E and F), which exploits the differential stabilities of RFP and GFP in an acidic environment (15, 37). Under normal conditions, mitochondria are visualized as yellow structures due to GFP and RFP signals from mitochondria (Fig. 7E). Because GFP signal is quenched at a lower pH, RFP can be visualized in autolysosomes containing mitochondria; thus, an increase in RFP fluorescence in the lysosomes indicates the completion of mitophagy due to mitophagic flux. We observed more red puncta in TH-treated cells, suggesting increased mitophagic flux compared to untreated cells, an effect that was significantly decreased by knockdown of ESRRA, ULK1, DRP1, or FUNDC1 (Fig. 7, E and F). TH significantly increased mitochondrial colocalization with the lyosomal marker LAMP1, thus confirming the colocalization of mitochondria with lysosomes (fig. S7, A and B). This colocalization was inhibited by knockdown of ESRRA.

We also measured mitophagic flux by quantifying mitochondrial proteins in the presence of bafilomycin A1, which inhibits lysosomal acidification and blocks autophagy. TH increased mitophagic flux (Fig. 8, A and B), and this increase was blocked by siRNA-mediated knockdown of ESRRA, ULK1 (Fig. 8, C and D), DRP1 (Fig. 8, E and F), or FUNDC1 (Fig. 8, G and H). OXPHOS measurements demonstrated that knockdown of ULK1, DRP1, or FUNDC1 inhibited TH-induced increases in mitochondrial basal oxygen consumption, ATP turnover, maximum respiratory capacity, and spare respiratory capacity (Fig. 8, I and J), similar to the effects observed with ESRRA knockdown. These findings showed that mitochondrial fission and mitophagy were necessary to sustain TH-induced mitochondrial activation.

Fig. 8 TH-induced mitophagy and mitochondrial turnover depended on ESRRA, ULK1, DRP1, and FUNDC1.

(A) Western blot analysis of ESRRA and mitochondrial proteins in bafilomycin A1 (BAF A1)–treated THRB1-HepG2 cells transfected with control siRNA or ESRRA siRNA in the presence or absence of TH (100 nM for 48 hours). Blots are representative of three independent experiments. (B) Measurement of relative mitophagic flux or of mitochondrial protein abundance in THRB1-HepG2 cells after BAF A1 treatment to block autophagy. The ratios of band densities from BAF A1–treated samples to untreated samples are shown. The relative mitophagic flux as determined by the accumulation of COX-IV,VDAC, SDHA, and PDH. after BAF A1 treatment is shown for control siRNA (Control siRNA + BAF A1)/(Control siRNA), Control siRNA + TH (Control siRNA + TH + BAF A1)/(Control siRNA + TH), and ESRRA siRNA + TH (ESRRA siRNA + TH + BAF A1)/(ESRRA siRNA + TH) (n = 5 biological replicates per group). COX-IV, cytochrome c oxidase subunit IV; VDAC, voltage-dependent anion channel; SDHA, succinate dehydrogenase complex flavoprotein subunit A; PDH, pyruvate dehydrogenase. (C and D) Representative Western blot for the mitochondrial protein SDHA (C) and relative mitophagy flux (D) based on SDHA abundance in THRB1-HepG2 cells subjected to ULK1 knockdown after BAF A1 treatment to block autophagy (n = 5 biological replicates per group). Blots are representative of three independent experiments. (E and F) Representative Western blot for SDHA (E) and relative mitophagy flux (F) based on SDHA abundance in THRB1-HepG2 cells subjected to DRP1 knockdown after BAF A1 treatment to block autophagy (n = 5 biological replicates per group). Blots are representative of three independent experiments. (G and H) Representative Western blot for SDHA (G) and relative mitophagy flux (H) based on SDHA abundance in THRB1-HepG2 cells subjected to FUNDC1 knockdown after BAF A1 treatment to block autophagy (n = 5 biological replicates per group). Blots are representative of three independent experiments. (I and J) Measurement of mitochondrial OCR using a Seahorse extracellular flux analyzer (I) in HepG2 cells treated with TH and ULK1, DRP1, or FUNDC1 siRNA in THRB1-HepG2 cells (n = 5 biological replicates per group). Data are representative of three independent experiments. Functional parameters of OXPHOS (basal OCR, ATP turnover, maximum respiratory capacity, and spare respiratory capacity) were calculated in the presence of mitochondrial inhibitors [1 μM oligomycin, 1 μM FCCP, and 1 μM rotenone and 1 μM antimycin A (R + A)] (J). (K and L) Mitochondrial turnover analysis using mitochondrial targeted fluorescent TIMER protein (mitoTIMER). mitoTIMER plasmid was transfected in THRB1-HepG2 cells with control or ESRRA siRNA with or without TH (100 nM for 48 hours). Representative images are shown (K), and fluorescence was quantified (L) (n = 5 biological replicates per group). Images are representative of five fields per group and two independent experiments. Scale bars, 5 μm. Green fluorescence shows newly synthesized mitochondria, whereas red fluorescence indicates damaged mitochondria. Statistical significance for (J) was calculated using one-way ANOVA, followed by Tukey’s post hoc test; otherwise, Kruskal-Wallis nonparametric ANOVA, followed by post hoc pairwise comparisons using the Mann-Whitney test with Bonferroni adjustment, was performed (*P ≤ 0.05).

TH regulation of mitochondrial turnover depended on ESRRA

To better assess mitochondrial turnover, we used the MitoTimer reporter that encodes a mitochondria-targeted protein that emits green fluorescence when newly synthesized and then irreversibly shifts from green to red fluorescence when oxidized by ROS accumulation (3840). Under basal conditions in control cells, we observed green and red mitochondrial signals with similar intensities (shown by the yellow overlay) that represented a mixture of newly synthesized and old mitochondria (Fig. 8, K and L). TH increased the green mitochondrial signals and decreased the red mitochondrial signals, suggesting increased mitochondrial biogenesis and degradation (Fig. 8, K and L). TH also significantly increased the mitochondrial membrane potential and mitochondrial ROS generation (fig. S8, A to C), consistent with the concomitant occurrence of both these processes. In particular, there were reduced old mitochondria (shown by the red signal) due to their rapid clearance by mitophagy and their replacement by newly synthesized mitochondria (shown by the green signal). In contrast, ESRRA knockdown in TH-treated cells showed reduced mitochondrial biogenesis (less green signal) and decreased mitochondrial degradation (more red signal). In addition, mitochondrial membrane potential and ROS were reduced by ESRRA knockdown during TH treatment (fig. S8, A to C). Together, these findings showed that ESRRA was necessary for the coordinated increase in mitochondrial turnover through biogenesis and mitophagy by TH.

DISCUSSION

Cells maintain proper mitochondrial function by removing or repairing damaged mitochondria and generating a renewable pool of healthy mitochondria. Various transcription factors and NRs regulate mitochondrial biogenesis, OXPHOS, and TCA cycle flux in various tissues (3, 6, 7). Notably, the NRs THRB1 and ESRRA regulate the gene expression of these mitochondrial pathways in the liver (5, 17, 28, 4145); however, little is known about whether their cell signaling pathways converge and potentially co-regulate certain mitochondrial pathways. THs (T3 and T4) are well-characterized ligands for THRB1, but the only ligand known to stimulate ESRRA activity is cholesterol (10).

PPARGC1A is a transcriptional coactivator that regulates the expression of mitochondrial genes independently and in conjunction with both NRs (5, 45, 46). We found that TH significantly increased ESRRA mRNA and protein expression in a manner dependent on THRB1 and PPARGC1A. ESRRA also may regulate THRB1 gene expression (47), suggesting that ESRRA may cross-regulate THRB transcription, similar to its previously reported actions on THRA gene expression (48). Our transcriptomic and ChIP-seq analyses of mouse livers showed co-regulation of major mitochondrial pathways such as OXPHOS, and TCA cycle flux by TH and ESRRA. Furthermore, ESRRA played an essential role in the stimulation of OXPHOS and the induction of TCA metabolites by TH. ESRRA also promoted the stimulation of fatty acid β-oxidation by TH. TH increases lipid catabolism and decreases hepatic lipid content (30); however, Esrra-null mice have similar basal hepatic free fatty acid or triglyceride levels as wild-type mice (47). We found that ESRRA had little or no effect on CPT1A transcription and triglyceride clearance during basal conditions but was required for CPT1A transcription, triglyceride clearance, and FAO induced by TH in hepatic cells and in vivo.

The synthesis of new mitochondria and the removal of damaged mitochondria are critical for maintaining mitochondrial homeostasis and response to increased demand. Although several mitochondrial proteins are expressed from the mitochondria genome, most are encoded in the nucleus (49). PPARGC1A and nuclear respiratory factors (NRF1 and NRF2) are key components of the regulatory network that control mitochondrial biogenesis. Mitochondrial DNA replication is initiated by mitochondrial transcription factor A (TFAM), which is regulated by PPARGC1A and NRF1. We observed that TH significantly increased PPARGC1A, NRF1, and TFAM mRNA expression and mitochondrial DNA copy number in an ESRRA-dependent manner. Moreover, the livers of TH-treated mice had more mitochondria than those of euthyroid mice or mice treated with both TH and XCT790. These findings suggested that stimulation of mitochondrial biogenesis by TH required the induction of ESRRA by TH.

Excessive or prolonged energy demand, oxidative stress, and aging can lead to ROS-mediated mitochondrial damage and dysfunction. Accordingly, mitochondrial biogenesis must be coupled with control mechanisms such as fusion, fission, and mitophagy to maintain mitochondrial quality. We did not observe induction of mitochondrial fusion gene expression or enlargement of mitochondria. In contrast, we found that TH regulated mitochondrial fission and mitophagy in a manner that depended on the induction of ESRRA. Mitochondrial fission is generally regulated by DRP1 (50, 51), which translocates into mitochondria after its phosphorylation at Ser616. It then oligomerizes and constricts the mitochondrial membrane to segregate damaged mitochondrial areas and then activates and recruits mitophagic proteins (such as FUNDC1, PINK1, PARKIN, NDP52, and optineurin) and autophagic proteins (such as ULK1, WIPI1, MAP1LC3B-II, and SQSTM1) (32, 36, 37, 5054). TH induces PARKIN- and PINK-mediated mitophagy in hepatocellular carcinoma due to mitochondrial damage from excessive ROS accumulation and mitochondrial depolarization caused by hepatitis B virus–encoded X protein (55). We found that TH increased the expression of several key genes involved in both mitochondrial fission (Drp1) and mitophagy (PINK1 and Parkin).

We have previously shown that ULK1 was recruited to mitochondria and selectively regulated hepatic mitophagy after TH treatment (15). In addition, ULK1 can also enhance its own activity through autophosphorylation (56, 57). We found that ESRRA regulated ULK1 expression during basal and hyperthyroid conditions in an ESRRA-dependent manner. In this regard, ESRRA binds to the Ulk1 gene promoter in mouse liver (47, 58). We found that TH increased Ulk1 gene transcription and ESRRA and POLR2A recruitment to the Ulk1 promoter in a PPARGC1A-dependent manner. This induction of Ulk1 expression is likely important to the stimulation of mitophagy by TH and may promote mitochondrial fission because ULK1 phosphorylates DRP1 (36). We found that TH induction of ULK1 through ESRRA led to increased phosphorylation of DRP1 at Ser616 and increased its translocation to mitochondria. Pharmacological inhibition of ESRRA, ULK1, or DRP1 blocked mitochondrial fission and mitophagy induced by TH.

Upon its activation, ULK1 interacts with other mitophagic receptor proteins, such as SESN2 (also known as Sestrin2), MUL1 (mitochondrial E3 ubiquitin ligase 1), or FUNDC1, on damaged parts of mitochondria and then recruits the autophagosomal proteins MAP1LC3B-II and SQSTM1 (32, 53, 59, 60). Here, we observed that TH increased the translocation of ULK1, MAP1LC3B-II, and SQSTM1 to the mitochondrial fraction in mouse liver and hepatic cells. Furthermore, TH increased the interaction of ULK1 with FUNDC1 and MAP1LC3B-II, suggesting that ULK1 may activate FUNDC1 to recruit MAP1LC3B-II to mitochondria. Pharmacological inhibition or knockdown of ESRRA prevented TH-induced recruitment of ULK1 to mitochondria and SQSTM1 and MAP1LC3B-II to autophagosomes. In addition, TH stimulation of mitochondria localization within autolysosomes was ESRRA-dependent. Finally, we demonstrated the importance of ESRRA to mitochondrial fission and mitophagy because induction of OXPHOS by TH was inhibited when ESRRA, ULK1, DRP1, or FUNDC1 were knocked down. Together, our findings showed that increased mitochondrial fission triggered by ESRRA and TH was coupled with increased mitophagy through stimulation of ULK1 gene expression and ULK1-mediated phosphorylation of DRP1, leading to DRP1 translocation into mitochondria. These events were necessary for maintaining mitochondrial homeostasis and induction of mitochondrial activity by TH.

We showed that the induction of both mitochondrial number and autophagic vesicles containing mitochondria by TH occurred in an ESRRA-dependent manner, suggesting that biogenesis and mitophagy were coupled. In hepatic cells, TH increased mitochondrial turnover, whereas ESRRA knockdown inhibited mitochondrial turnover by TH and led to an accumulation of old mitochondria and a decrease in newly synthesized mitochondria. Thus, TH causes a coordinated increase in mitochondrial biogenesis and mitophagy and, thus, mitochondrial turnover, which depended on its induction of ESRRA expression. Maintenance of higher mitochondrial turnover is essential for the chronic increase in mitochondrial activity caused by TH.

Although TH has been suggested to regulate mitochondrial protein and DNA turnover (29, 61, 62), the underlying mechanism was not known. Here, we have demonstrated several mechanisms involving THRB1 and ESRRA co-regulation of mitochondrial turnover and function (fig. S9). Although ESRRA stimulates mitochondrial biogenesis and mitochondrial metabolic pathways (12, 41, 6365), its role in mitochondrial fission and mitophagy has not been described previously. Although ESRRA was required for many of the effects of TH on mitochondrial function and turnover, we showed that TH may induce or enhance some of ESRRA’s metabolic actions in the cell through their co-regulation of common target genes. Antagonists and inverse agonists of ESRRA have been developed, but agonists have been lacking (6672). Thus, the induction of ESRRA expression by TH represents a means of activating ESRRA. Accordingly, it is possible that hormonal and/or pharmacological induction of ESRRA may be a viable therapeutic strategy to maintain and preserve mitochondrial function in metabolic disorders and aging.

MATERIALS AND METHODS

Chemicals, reagents, and antibodies

3,3′,5-Triiodothyronine (TH or T3), XCT790, carbonyl cyanide 3-chlorophenylhydrazone (CCCP), Mdivi-1, bafilomycin A1, palmitic acid, oleic acid, CelLytic M mammalian cell lysis/extraction reagent, KiCqStart SYBR Green predesigned primers for gene expression analysis, fetal bovine serum, and 4′,6-diamidino-2-phenylindole were procured from Sigma-Aldrich. MRT0068921 was a gift from B. Saxty (Medical Research Council, UK). Phenol red–free cell culture media were purchased from Invitrogen. BODIPY 493/503, MitoTracker Red CMXRos, tetramethylrhodamine ethyl ester (TMRE), MitoSOX Red reagent, and Mitochondria Isolation kit for Cultured Cells were purchased from Thermo Fisher Scientific. QIAamp DNA Mini kit and QuantiTect SYBR Green PCR kits were from Qiagen. β-Hydroxybutyrate (Ketone Body) Colorimetric Assay kit was from Cayman Chemical. Seahorse XF Cell Mito Stress Test kit and Seahorse XF Mito Fuel Flex Test kit were from Seahorse Bioscience Inc. InviTrap Spin Universal RNA kit was from Stratec Biomedical. Dual-Luciferase Reporter Assay System was procured from Promega.

Antibodies against ESRRA (anti-ESRRA antibody–ChIP Grade, ab16363), POLR2A [anti-RNA polymerase II CTD repeat YSPTSPS (phospho S5) antibody–ChIP Grade, ab5131] were purchased from Abcam. GAPDH (glyceraldehyde-3-phosphate dehydrogenase), TUBB, MAP1LC3B, SQSTM1, ULK1, DRP1, VDAC, SDHA, COX-IV, and PDH antibodies were from Cell Signaling Technology. FUNDC1 antibody (NBP1-81063) was procured from Novus Biologicals. THRB1, LAMP1, TOM20, and horseradish peroxidase–conjugated secondary antibodies recognizing mouse (sc-2954) and rabbit (sc-2955) immunoglobulin Gs (IgGs) were purchased from Santa Cruz Biotechnology.

Animals and handing

Male C57BL/6 mice (6 to 8 weeks old) were purchased and housed in hanging polycarbonate cages under a 12-hour light/12-hour dark cycle at 23°C with food and water available ad libitum. All cages contained shelters and nesting material. Hyperthyroidism was induced by injecting TH and confirmed as described previously (19). During the course of treatment, animals were monitored daily for their general health and weight. The liver-specific Thrb-null (Thrb−/−) mice, which lack both Thrb1 and Thrb2 isoforms, were in a C57BL/6:129sv mixed background. Thrb−/− mice and wild-type (Thrb+/+) mice of the same strain were treated with TH, as described previously (19). For ESRRA inhibition, mice were treated with XCT790 (intraperitoneal injection of 10 mg/day per 100 g body weight for 3 days). After the first injection of XCT790, mice were injected subcutaneously with TH (10 μg/100 g body weight per day for 3 days) along with XCT790. Animals were then euthanized, and the liver tissues were harvested for processing.

Ethics statement and study approval

All mice were maintained according to the Guide for the Care and Use of Laboratory Animals [National Institutes of Health (NIH) publication 1.0.0; revised 2011], and experiments were approved by the Institutional Animal Care and Use Committees at the University of Pennsylvania, the National Cancer Institute (NIH), and the Duke–National University of Singapore (NUS) Graduate Medical School.

Cell culture and maintenance

THRB1-HepG2 cells, which have been used to study TH action in vitro (15, 19, 20, 73), were used in this study (a gift from M. L. Privalsky, University of California, Davis). THRB1-HepG2 cells display an enhanced response to TH, confirming that the necessary auxiliary machinery exists (73). We also used, in parallel, isogenic HepG2 cell lines that express empty expression plasmid containing no receptor. All HepG2 cell lines were cultured and maintained as described previously (19, 20). For TH (T3) treatments, cells were grown at least 3 days in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% Dowex-stripped fetal bovine serum and 1× penicillin/streptomycin (normal TH-depleted DMEM) before adding TH (100 nM; unless mentioned otherwise) with or without other compounds (2 μM XCT790, 1 μM MRT0068921, 10 μM Mdivi-1, and 20 nM bafilomycin A1) for 48 hours or as indicated in the figures. Primary mouse hepatocytes were isolated from male C57BL/6 mice (8 to 10 weeks old) using a standard two-step collagenase perfusion method, as described previously (74), and cultured in normal TH-depleted DMEM, as mentioned above.

RNA isolation and relative transcripts expression using qPCR

Total RNA isolation was performed using InviTrap Spin Universal RNA kit (Stratec Biomedical), and RT-qPCR was performed as described previously (19) using QuantiTect SYBR Green PCR kit and KiCqStart SYBR Green optimized primers from Sigma-Aldrich (KSPQ12012).

Western blotting

Cultured cells or 50 mg of liver tissues was lysed using CelLytic M mammalian cell lysis/extraction reagent (C2978, Sigma-Aldrich). An aliquot was removed, and protein concentrations were measured using the BCA kit (Bio-Rad). Western blotting was performed using a standard protocol, as described previously (19).

Coimmunoprecipitation analysis

Immunoprecipitation was performed using the immunoprecipitation starter pack (GE Healthcare), as per the manufacturer’s protocol. Nondenaturing lysis buffer (Abcam) was used to lyse THRB1-HepG2 cells treated with or without TH (100 nM for 48 hours) and XCT790 (2 μM). Antibodies against rabbit IgG (sc2027, Santa Cruz Biotechnology) and FUNDC1 (NBP1-81063, Novus Biologicals) were used for pull-down assay. Western blot analysis was performed as described above.

Genetic knockdown in vitro using siRNAs

Silencer Select siRNAs against ESRRA (s4829, s4830, and s4831; Life Technologies Inc.), PPARGC1A (s21395 and s21393), ULK1 (s15963, s15965, and s15964), DRP1 (s19559 and s19560), and FUNDC1 (s44201 and s44200) were used to silence ESRRA, PPARGC1A, ULK1, DRP1, and FUNDC1 gene, respectively, in THRB1-HepG2 cells. Negative siRNA (Silencer Negative Control No. 1 siRNA; AM4611, Life Technologies Inc.) was also used in parallel as a negative control. Transfections were carried out in THRB1-HepG2 cells in a 12-well plate or four-well chambered slides or 24-well Seahorse XF plate using 30 nM of the above indicated siRNAs and negative control siRNA with Lipofectamine RNAiMAX (Invitrogen; Life Technologies Inc.) following the reverse transfection protocol, as provided by the manufacturer. After 24 hours of transfection, cells were subjected to TH (100 nM) treatment in normal TH-depleted medium. After 48 hours of treatment, total RNA or protein was isolated for further analysis.

Transient overexpression in vitro

Flag-ESRRA plasmid (Addgene plasmid #10975) was a gift from T. Finkel (Aging Institute of University of Pittsburgh Medical Center and Pitt) and described previously (75). Hemagglutinin-PPARGC1A plasmid was described elsewhere (76) and a gift from A. Kralli (The Scripps Research Institute). Plasmid pAT016 (p-mito-mRFP-EGFP) was described elsewhere (37) and a gift from A. Till (University of California). pLAMP1-mCherry plasmid that targets lysosomes (Addgene plasmid #45147) was a gift from A. Palmer (University of Colorado) and described previously (77). pMitoTimer (Addgene plasmid #52659) was a gift from Z. Yan (University of Michigan Medical School) and described previously (39). Transfections were carried out in THRB1-HepG2 cells in a 24-well plate or four-well chambered slides or 24-well Seahorse XF plate using 1 μg of the above indicated plasmids or empty vector using Lipofectamine 3000 (Invitrogen; Life Technologies Inc.) following the reverse transfection protocol, as provided by the manufacturer. After 24 hours of transfection, cells were subjected to TH (100 nM) treatment in normal TH-depleted medium, as described above. After 48 hours of treatment, total RNA or protein was isolated for further analysis, or fluorescence microscopy was performed as standard protocol.

Luciferase reporter construction and expression in vitro

The mouse Esrra promoter was cloned by PCR using mouse embryonic stem cell genomic DNA using the following primers: AACTAGTTAGAGTGAGCGGCTCAGGAGGAG containing Spe (distal 5′ side of the mouse Esrra promoter) and GCGGCCGCCTTCCCCTCCGCCCCCGGACCTTGGGCTC containing Not I (proximal 3′ side of the mouse Esrra promoter). The resulting construct (aactagttagagtgagcggctcaggaggaggttgggtctcactgcctctgttcaggcttgcagcaacttcttcagccttcttcccctccattcttcacatcgcacctaggagacccaaaattccccaaactaacacagatccaaatccagcactaaatgccacctcaaatcacctgtggttcctcaaatgggggaagccctttgctgctgttcatggccctgagctcagctcttccctctccctagcatataggtccacgattggcaggaaccacttcatgactaaagagctgcagggatacatcgaagggatcaggaagcgcaggaacaagaggctgtactttttggatcagtgaacctccagctcaagtggccgacgagacgtccgctccgtgtcctttccggtccctcgaacccaagctcaacctcatctgcatggaatttgaatcctagagaaattaaagagtctatgcatggtcccagagtcagtgcgtggctgcttgcgtgaggcggacgcacgtggccccgcctttccccgtgaccttcattcggtcaccgcagtgaccttgagttttgtccgctcgtgtctcacctctgcctttgaggaactcgcaaaccagggctacttgcatgttttaggtctgcttctccacttgagggtcctcaggggtgctgatctaccaaatctgtcaagtgtcccatcaacataatggggtggggtggggagcagcaccctgttaggccccaccccttatgcatatgcagttacataaacccgcctttcatttacatacacgttcaacgccctcacagagtgagttcactgaacagctacgcccactcttataaggcagagcaccgccctccttcgtgatgcaagggtgctccggcttctctcccgcaccagttcctagtcttcgctcctcccctaggttcaggccccgcccacaggaggccgcgcacgcgtgtgtatcgacaaggcctgcccgctgtcctccgggtacgcaggcgccgccgcctccctcggatattagcatagggcacctggccccaccccctgttttgcatgcgcaggccggccccgcctccgctgtcagctggaggaagcggagtaggaagcagccgcgatgtccttttgtgtcctacaagcagccagcggcgccgccgagtgaggggggacgcagcgcggcggggcggtgcggccggaggaggcggcccccgctcaccccggcgctccgggccgctcggcccccatgcctgcccgccagccctgccggagcccaaggtccgggggcggaggggaaggcggccgc) was cloned as a Not I (blunt)–Spe I fragment into Nhe I–Sma I sites of pGL3 basic luciferase reporter vector (E1751, Promega) and referred to as “wild-type ESRRA Luc.” Mutations were introduced in the characterized ERREs (26) by PCR using the following primers: CGCCTTTCCCCGTGCATCTCATTCGGTCAC and CGGTCACCGCAGTGCATCTGAGTTTTGTCC for ERRE 1 and GTGACCGAATGAGATGCACGGGGAAAGGCG and GGACAAAACTCAGATGCACTGCGGTGACCG for ERRE 2 and referred to as mutant ESRRA Luc. Wild-type and mutant ESRRA Luc reporter plasmids were cotransfected with Renilla luciferase control reporter plasmid (pRL; E2231, Promega) using Lipofectamine 3000 (Thermo Fisher Scientific), as per the manufacturer’s protocol. Relative ESRRA luciferase activity was analyzed using Dual-Luciferase Reporter Assay System (Promega), as per the company’s protocol, and readings were taken on a plate luminometer (Infinite 200 PRO, Tecan).

Fluorescence staining, imaging, and quantification

For neutral lipid and lipid droplet staining, cells were cultured under fat-loaded (oleic acid/palmitic acid ratio, 2:1) and ESRRA knockdown conditions in 12-well plates. Cells were subsequently incubated with BODIPY 493/503 stain [1:1000 dilution in phosphate-buffered saline (PBS) for 10 min] and rinsed twice with medium. Lipid droplets imaging in live cells were performed in a fluorescence microscope (Olympus), and images were captured at ×20 magnification using Leica Application Suite (Leica Microsystems). Cells were then immediately trypsinized, washed with PBS, and suspended in 500-μl PBS for quantification using a flow cytometer (using a fluorescein isothiocyanate green channel filter). Graphs represent MFI of BODIPY 493/503 stain after acquisition of 20,000 events, as described previously (78).

To observe production of superoxide by mitochondria, MitoSOX Red reagent was used, respectively, as per the manufacturer’s instructions. Imaging in live cells was performed in fluorescence microscope (Olympus), and images were captured at ×20 magnification using Leica Application Suite (Leica Microsystems). MitoSOX Red fluorescence was quantified on ImageJ after converting images to grayscale and measuring densitometry value.

To stain mitochondria, MitoTracker Red CMXRos was used as per the manufacturer’s instructions. For DRP1 translocation to mitochondria, MAP1LC3B-mitochondrial colocalization, mitochondrial autophagic flux, and turnover analysis, cells were first either transfected with ESRRA, ULK1, DRP1, or FUNDC1 siRNA using reverse transfection or treated with corresponding inhibitors in four-well chambered slides, stained with MitoTracker Red CMXRos, and fixed with 4% paraformaldehyde for 15 min, followed by washing, blocking, and conjugation with primary antibodies (ULK1 and DRP1) overnight at 4°C. Anti-rabbit IgG–Alexa Fluor Plus 488 (Thermo Fisher Scientific) was used as the secondary antibody. Cells were washed once and treated with Hoechst 3334 at 1:5000 dilution in PBS for 10 min. Coverslips were mounted using Vectashield antifade mounting media (Invitrogen).

For mitophagic flux and mitochondrial turnover analysis, cells were treated with compounds and/or siRNA for 24 hours and then transfected with plasmids pAT016 (p-mito-mRFP-EGFP) and pMitoTimer, respectively, along with TH treatment. After treatment, cells were washed in PBS and fixed for 15 min in 4% paraformaldehyde. Cells were then washed once and then treated with Hoechst 3334 at 1:5000 dilution in PBS for 10 min. Coverslips were mounted using Vectashield antifade mounting media (Invitrogen). Cells were then visualized, and images were captured using an LSM 710 Carl Zeiss confocal microscope at ×40 magnification. Images were pseudocolored and visualized in ZEN 2012 SP1 (black edition) (Carl Zeiss) software. Quantification of DRP1 translocation, MAP1LC3B-mitochondrial colocalization, LAMP1-mitochondrial colocalization, mitophagic flux, and mitochondrial turnover was performed using CoLocalizer Pro software (CoLocalization Research Software) and ImageJ (NIH).

For mitochondrial membrane potential (active mitochondria) analysis, we used TMRE stain that is sequestered by active mitochondria in a membrane potential–dependent manner. TH-treated cells were stained with TMRE, as per the manufacturer’s instruction, and cells were trypsinized, washed twice with PBS, and analyzed on a flow cytometer (MACSQuant Analyzer 10, Miltenyi Biotec). CCCP (5 μM for 3 hours), a mitochondrial uncoupler, was used as a positive control.

Oxygen consumption rate measurement

To determine cellular oxygen consumption in THRB1-HepG2 cells, the Seahorse XF24 Extracellular Flux analyzer (Seahorse Bioscience), which measures the oxygen consumption rate, was used as described previously (19, 79). To analyze mitochondrial ATP turnover, XF Cell Mito Stress Test kit was used as per the manufacturer’s protocol. Before the assay was performed, TH-depleted medium was removed and replaced by 500 ml of assay medium at 37°C. Oligomycin (1 μM), which inhibits the F0 proton channel of the F0F1-ATP synthase, was used to determine the oligomycin-independent leak of the oxygen consumption rate. The mitochondrial uncoupler FCCP (1 μM) was added to determine the total respiratory capacity of the mitochondrial electron transport chain. Rotenone (1 μM) and antimycin A (1 μM) were added to block complex I and complex III of electron transport chain. The following mitochondrial functional parameters were calculated as follows: (i) basal O2 consumption = baseline oxygen consumption reading per well (before compounds are injected) subtracted from oxygen consumption reading per well after R + A injection; (ii) ATP turnover = baseline oxygen consumption reading per well (before compounds are injected) subtracted from oxygen consumption reading per well after oligomycin injection; (iii) maximum respiratory capacity = oxygen consumption reading per well (after FCCP injection) subtracted from oxygen consumption reading per well after R + A injection; and (iv) spare respiratory capacity = oxygen consumption reading per well (after FCCP injection) subtracted from baseline oxygen consumption reading per well (before compounds are injected).

Mitochondrial fuel oxidation analysis

To analyze mitochondrial fuel oxidation, we used Seahorse XF Mito Fuel Flex Test kit that determines the rate of oxidation of each fuel (glucose, fatty acids, and glutamine) by measuring mitochondrial respiration (oxygen consumption rate) of cells in the presence or absence of fuel pathway inhibitors: UK5099, an inhibitor of the glucose oxidation pathway; etomoxir, an inhibitor of the long-chain FAO pathway; BPTES [bis-2-(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethyl sulfide], an inhibitor of the glutamine oxidation pathway.

Transcriptome analysis and THRB1/ESRRA ChIP-seq data generation

Gene expression microarray profiling was performed using MouseWG-6 version 2.0 Expression BeadChip kit (Illumina) by hybridizing RNA from liver tissues of euthyroid, TH-treated, and XCT790 + TH–treated mice. Complementary RNA generation, labeling, and hybridization were performed at Duke-NUS Genome Biology Facility, Duke-NUS Graduate Medical School, Singapore. Gene expression signals were quantile-normalized, and differentially expressed genes were identified by ANOVA using treatment-specific contrasts (Partek Genomics Suite software, version 6.6). Statistical significance of differentially expressed genes was ascertained in terms of the false discovery rate (FDR) (80). Complete data set was submitted to the Gene Expression Omnibus (GEO) repository (GSE81132). Principal components analysis based on gene expression demonstrated a clear separation between the three experimental groups and no outliers. Pathway enrichment analysis was conducted with the Gene Set Enrichment Analysis tool (81) and a list of KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways extracted from the Molecular Signatures Database (82). All analyses were conducted using mouse gene symbols. Significance of pathway enrichment was ascertained by permutation testing of gene sets and calculation of the FDR (q value of 0.050) and nominal P value. To identify THRB1 and ESRRA binding on the genes extracted from our transcriptome analysis, we used previously generated THRB1 and ESRRA ChIP-seq data sets (47, 83) and reanalyzed it.

ChIP-qPCR analysis for ESRRA and POLR2A binding on ESRRA and ULK1 gene promoter

The ChIP assays were performed on liver tissues using the LowCell# ChIP kit protein G (kch-maglow-G48, Diagenode) according to the manufacturer’s protocol with some modifications, as described previously (84). Briefly, 50 mg of liver tissue was fixed with 1% formaldehyde, lysed, and sonicated to shear chromatin. Samples were precleared and incubated with 4 μg of antibodies against ESRRA (anti-ESRRA ChIP Grade, ab16363), POLR2A [anti–RNA polymerase II (phospho S5) antibody–ChIP Grade, ab5131], or normal rabbit IgG (sc2027, Santa Cruz Biotechnology) overnight at 4°C. Immune complexes were pulled down with magnetic beads, reverse cross-linked, and purified using QuantiFast SYBR Green PCR kit (Qiagen).

For in vitro experiments, ChIP assays were performed using the EZ-Magna ChIP G-Chromatin Immunoprecipitation kit (17-610, Millipore) according to the manufacturer’s protocol with minor modifications, as described previously (20). Briefly, THRB1-HepG2 cells were transfected in 150-mm culture dishes with siRNAs, as mentioned above, or cultured in normal TH-depleted medium for 3 days in 150-mm dishes before the TH treatment. Immunoprecipitation was performed with above-mentioned antibodies.

Two microliters of immunoprecipitated DNA (1% input DNA) was used with a QuantiFast SYBR Green PCR kit (Qiagen) for 40 cycles of qPCR using Rotor-Gene Q qPCR machine (Qiagen). DNA regions used to amplify were as published previously (26). For ESRRA binding on the ESRRA gene, primer pairs [ATGCATGGTCCCAGAGTCAG (forward) and CTGGTTTGCGAGTTCCTCAA (reverse); and GTGCAGTGGTGCCATCTC (forward) and CTTTGGGAGGCCGAGGT (reverse)] were used to amplify −629 to −498 and −3196 to −2984 regions on ESRRA promoters of mouse livers and THRB1-HepG2 cells, respectively. For POLR2A binding, a primer pair [ATTAGCATAGGGCACCTGGC (forward) and CGACCACCGTGGCTGAC (reverse)] was used to amplify −69 to +11 regions in mouse livers. However, on the ULK1 gene promoter, for ESRRA binding, primer pairs [AGTCTCCGTCCCCACATACAG (forward) and CTGGTCTCGAACTTGCTTTGTC (reverse); and TTGGACACATTCTCCAACTTT (forward) and GATTGGGATGTGTGTGTGTG (reverse)] were used to amplify −995 to −832 and −1200 to −993 regions on the ULK1 promoter in mouse livers and THRB1-HepG2 cells, respectively. For POLR2A binding, a primer pair [TCCTTCATTTCGGACCCGC (forward) and TGACACCACATAAGCCCGAG (reverse)] was used to amplify −115 to +12 regions in mouse livers.

Metabolic profiling of acylcarnitines and organic acids

Methods of extraction and metabolic profiling for acylcarnitines and organic acids were performed, with some modifications, as described elsewhere (79, 85). Briefly, liver tissue was homogenized in 50% acetonitrile and 0.3% formic acid. For acylcarnitine extraction, 100 μl of tissue homogenate was extracted using methanol. The acylcarnitine extracts were derivatized with 3 M hydrochloric acid in methanol, dried, and reconstituted in methanol for analysis in liquid chromatography/mass spectrometry (LC/MS). For organic acid extraction, 300 μl of tissue homogenate was extracted with ethyl acetate, dried, and derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide, with protection of the α-keto groups using ethoxyamine. Acylcarnitine measurements were made using flow injection–tandem mass spectrometry on the Agilent 6430 Triple Quadrupole LC/MS system (Agilent Technologies). The sample analysis was carried out at 0.4 ml/min of 80:20 methanol/water as mobile phase and injection of 2 μl of sample. Data acquisition and analysis were performed on Agilent MassHunter Workstation B.06.00 software. Trimethylsilyl derivatives of organic acids were separated by gas chromatography on an Agilent Technologies HP 7890A and quantified by selected ion monitoring on a 5975C mass spectrometer using stable isotope dilution. The initial gas chromatography oven temperature was set at 70°C, ramped to 300°C at a rate of 40°C/min, and held for 2 min.

Electron microscopy

Fresh liver tissue was placed in fixative [2% paraformaldehyde and 3% glutaraldehyde in cacodylate buffer (pH 7.4)] and stored at 4°C. Samples were washed once in PBS, followed by postfixation treatment with 1% osmium tetroxide. Samples were dehydrated with ascending concentrations of alcohol and then embedded in Araldite. Ultrathin sections were cut and stained with uranyl acetate and lead citrate. Images were acquired on an Olympus EM208S transmission electron microscope at ×10,000 and ×80,000 magnifications. Mean number of mitochondria or autophagic vesicles containing mitochondria per transmission electron microscopy field from untreated euthyroid, TH-treated (hyperthyroid), and XCT790-treated mouse liver samples was calculated from a total of 10 random fields per treatment.

Mitochondrial fractionation

Mitochondrial fractionation was performed using the Mitochondrial Fractionation kit (Thermo Fisher Scientific) according to the manufacturer’s instructions.

Mitochondrial DNA copy number analysis

Total DNA was extracted from 100 mg of liver tissue or cells cultured in six-well plate using QIAamp DNA Mini kit (Qiagen). Mitochondrial DNA copy number was determined by amplifying genes encoded by mitochondrial DNA and genomic DNA, as described elsewhere (8688). MT-ND6 [5′-CGAAAGGACAAGAGAAATAAGG-3′ (forward) and 5′-CTGTAAAGTTTTAAGTTTTATGCG-3′ (reverse)] and MT-RNR2/16S RNA [5′-CCCCACAAACCCCATTACTAAACCCA-3′ (forward) and 5′-TTTCATCATGCGGAGATGTTGGATGG-3′ (reverse)] genes were used as a marker for mitochondrial DNA content. Lpl [5′-GGATGGACGGTAAGAGTGATTC-3′ (forward) and 5′-ATCCAAGGGTAGCAGACAGGT-3′ (reverse)] and HBB/β-Globin [5′-CAACTTCATCCACGTTCACC-3′ (forward) and 5′-GAAGAGCCAAGGACAGGTAC-3′ (reverse)] genes were used as a marker for mouse or human nuclear DNA content, respectively. These genes were amplified from 10 ng of total DNA with a Rotor-Gene Q qPCR machine (Qiagen). Relative quantification of mitochondrial DNA copy number was calculated after using 2ΔCt as fold change.

Statistical analysis

Individual culture experiments were performed in triplicate and repeated at least three times independently using matched controls; the data were pooled, and statistical analysis was performed. Results are expressed as mean ± SD for all in vitro experiments and mean ± SE for all in vivo experiments. The statistical significance of differences (*P < 0.05) was assessed by a one-way ANOVA, followed by Tukey’s multiple-comparisons test, whereas for non-normal distributed values, the nonparametric Mann-Whitney test to compare sample medians between two independent groups was performed. Kruskal-Wallis nonparametric ANOVA was used to test a null hypothesis of no difference in medians among three or more groups. In the event of a statistically significant Kruskal-Wallis test (P ≤ 0.05), post hoc pairwise comparisons were carried out using the Mann-Whitney test with Bonferroni adjustment for multiple comparisons. All statistical tests were performed using Prism 7 for Mac OS X (GraphPad Software).

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/536/eaam5855/DC1

Fig. S1. ESRRA co-regulated pathways induced by TH.

Fig. S2. TH-ESRRA co-regulation of TCA and FAO.

Fig. S3. TH activated ESRRA and OXPHOS in a PPARGC1A-dependent manner.

Fig. S4. TH-ESRRA co-regulated the expression of genes encoding mitochondrial biogenesis factors but not those encoding mitochondrial fission, fusion, and mitophagy factors.

Fig. S5. TH increased ESRRA recruitment to the ULK1 gene promoter.

Fig. S6. DRP1 inhibition blocked the translocation of autophagic proteins but not ULK1 translocation to mitochondria.

Fig. S7. TH increased lysosome and mitochondria colocalization in an ESRRA-dependent manner.

Fig. S8. ESRRA inhibition dysregulated TH-induced mitochondrial membrane potential and ROS.

Fig. S9. Proposed model for TH-ESRRA co-regulation of mitochondrial biogenesis, fission, mitophagy, and activity.

Data file S1. THRB1 and ESRRA ChIP-seq analysis.

REFERENCES AND NOTES

Acknowledgments: We are thankful to J.-P. Kovalik, J. Ching, Z. Tan, and K. T. Fridianto for performing and analyzing acylcarnitine and organic acid metabolomics (Metabolomics Core Facility, Duke-NUS Medical School Singapore). We are also grateful to M. L. Privalsky (University of California, Davis) for the gift of THRB1-HepG2 and HepG2 control plasmid cells. We are also thankful to J. C. Allen (Centre for Quantitative Medicine, Duke-NUS Medical School Singapore) for the valuable suggestions for statistical analysis. Funding: The authors are also thankful to our funding agencies [Ministry of Health, Ministry of Education, and Ministry of Trade, Singapore, and Agency for Science, Technology and Research (A*StaR)] for the grants CIRG/1340/2012, NMRC/CSA Grant/MH95:03/1-8 (awarded to P.M.Y.), NMRC/OFYIRG/0002/2016 (awarded to B.K.S.), and Canadian Institutes for Health Research grant MOP-125885 (awarded to V.G.). Author contributions: B.K.S., R.A.S., D.P.M., and P.M.Y. were involved in conceptualization and methodology. B.K.S., R.A.S., M.T., K.O., J.Z., J.A.C.S., S.Y.X., J.P.H., A.M., and Y.W. performed the experiments. B.K.S., A.M., B.-H.B., V.G., and S.G. analyzed the data. C.-y.C., D.P.M., J.-M.V., A.N.H., and K.G. shared the resources. B.K.S. was involved in the writing of the original draft. B.K.S., R.A.S., D.P.M, K.G., S.G., and P.M.Y. were involved in the revision of the manuscript. P.M.Y. and B.K.S. were involved in funding acquisition. Competing interests: The authors declare that they have no competing interests. Data and materials availability: The complete data set for the transcriptome microarray is available from the GEO repository (GSE81132). All other data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.
View Abstract

Navigate This Article