ETV5 Regulates Hepatic Fatty Acid Metabolism Through PPAR Signaling Pathway

Hepatic lipid metabolism, which is critical for cellular and systemic homeostasis, is tightly regulated. Dysregulation of hepatic lipid metabolism is related to multiple metabolic disorders, including nonalchoholic fatty liver disease (NAFLD), type 2 diabetes, and cardiovascular disease (1). A number of transcription factors have been identified to control hepatic lipid metabolism, such as SREBP1c, hepatocyte nuclear factor-4α, liver X receptor, and peroxisome proliferator–activated receptor-α (PPARα) and -γ (PPARγ) (2). By regulating the expression of genes encoding hepatic lipid metabolic enzymes, these transcription factors may act synergistically or antagonistically to either activate or repress the hepatic metabolic process (3).

Ets variant 5 (ETV5), also known as ERM, is an E twenty-six (ETS) transcription factor that belongs to the polyoma enhancer activator 3 (PEA3) family. This transcriptional factor is critical for organ development, tumorigenesis, and immune functions (4). ETV5 is expressed ubiquitously. In the brain, it is mainly expressed in the arcuate nucleus, ventromedial hypothalamus, substantia nigra, and the ventral tegmental area (5). Its transcriptional expression in the arcuate nucleus is altered by diet and food availability, suggesting a role in the control of energy homeostasis (5). In Drosophila, loss of neuronal Ets96B, the homolog of ETV5, increases lipid storage (6). Mice with ETV5 global knockout exhibit decreased body weight and less adipose tissue mass (7). ETV5 is also involved in insulin secretion, contributing to the modulation of systemic glucose and lipid homeostasis (8). Moreover, several genomic association studies have indicated that ETV5 is associated with human obesity and diabetes (9–12). All this evidence indicates that ETV5 may be involved in the control of systemic metabolic homeostasis. However, the physiological mechanism by which ETV5 regulates lipid homeostasis remains poorly understood. It is currently unknown whether ETV5 directly acts on metabolic organs, such as liver and adipose tissue, to alter lipid metabolism.

In the current study, we aimed to examine whether ETV5 regulates hepatic lipid metabolism. Using the mouse models of viral-mediated acute depletion and genetically mediated chronic deletion of the Etv5 gene together with in vitro HepG2 cell and primary hepatocyte culture, we demonstrated that ETV5 deficiency led to an increased lipid deposition in the liver. This occurred mainly through the reduction of fatty acid oxidation as a result of defective PPAR signaling activity. ETV5 enhanced the PPAR response element (PPRE) transactivity, leading to subsequent alteration in the expression of PPAR target genes. In addition, ETV5 could respond to the extracellular nutrient status primarily in a mammalian target of rapamycin complex 1 (mTORC1)–dependent manner. Taken together, our data demonstrate that ETV5 plays a novel role in hepatic lipid metabolism, which is required for optimal PPAR-mediated lipid utilization.

ETV5 conditional knockout mice (ETV5^fl/fl) were generated by Shanghai Model Organisms Center, Inc. (Shanghai, China). To generate the Etv5 targeting vector, two loxP sites flanking exon 10 and exon 11 were introduced. To generate hepatocyte-specific Etv5 knockout (ETV5-LKO) mice, ETV5^fl/fl mice were cross-bred with Alb-Cre transgenic mice (The Jackson Laboratory). Cre-mediated recombination at the loxP sites resulted in deletion of exon 10 and 11, causing a frameshift mutation and a premature stop codon in exon 12 (Fig. 4A). Littermate ETV5^fl/fl mice were used as the control mice. Knockout mice were identified by PCR using primers listed in Supplementary Table 1.

For the viral injection experiment, C57BL/6 or db/db male mice of 6–8 weeks of age were injected through the tail vein with lentivirus containing shRNA against Etv5 or scramble shRNA. For the high-fat diet (HFD) experiment, mice were fed with normal chow diet (ND) or HFD containing 45% fat (Mediscience, Jiangsu, China). Animals were fed an HFD for 4 weeks and sacrificed for subsequent analysis. All animal experiments were undertaken in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, with approval of the scientific investigation board of Shenzhen University Health Science Center.

For the antibodies, ETV5 (PA5-30023) was obtained from Thermo Fisher Scientific. PPARα (ab24509) and CPT1A (ab128568) antibodies were from Abcam. S6 (2313), phosphorylated S6 (pS6) (4858), mTOR (2972), phosphorylated mTOR (pmTOR) (2971), myc (9B11), flag (D6W5B), and ATGL (2439) antibodies were from Cell Signaling Technology. ETV5 (13011-1-AP), actin (60008-1-Ig), β-tubulin (10094-1-AP), and GAPDH (60004-1-Ig) antibodies were from Proteintech. For the plasmids, pCMV-ETV5-myc-flag, pCMV-PPARα-myc-flag, and pCMV-PPARγ-myc-flag were obtained from Origene. Luciferase vectors pGL4.20 and pGL4.74 were obtained from Promega.

Whole blood was collected by cardiac puncture, and serum was collected after centrifugation at 4°C. The serum was used for analysis of triglyceride (TG) (BioSino), total cholesterol (TC) (BioSino), nonesterified fatty acid (NEFA) (Wako Diagnostics), AST and ALT (Nanjing Jiancheng Bioengineering Institute), and insulin (ALPCO). All measurements were performed according to the manufacturers’ instructions. Liver or treated hepatocytes were incubated with lysis buffer for 10 min and then heated at 70°C for 10 min followed by centrifugation at 2,000 rpm for 5 min at room temperature. The supernatant was evaluated using the tissue TG assay kit (Applygen Technologies) or the TC assay kit (Applygen Technologies) according to the manufacturer’s instructions.

For the intraperitoneal glucose tolerance test, mice were first starved for 16 h followed by an intraperitoneal injection of glucose (1.5 g/kg body weight). Blood glucose levels were measured from the tail vein before and at 15, 30, 60, 90, and 120 min after injection using a glucometer (Accu-Chek; Roche). For the insulin tolerance test, mice were fasted for 4 h before the test and received an injection of human regular insulin (0.5 units/kg body weight). Using a glucometer, blood glucose levels were recorded before and at 15, 30, 60, 90, and 120 min after injection.

Tissue sections or cells were fixed with 4% paraformaldehyde, permeabilized, and stained as described earlier (13). The quantification was determined using Image J software. Images were obtained using a Nikon Eclipse Ti microscope.

Primary hepatocytes were isolated from 8-week-old ETV5^fl/fl or ETV5-LKO mice by a collagenase perfusion method described previously (14). Isolated hepatocytes were cultured in RPMI medium (11.1 mmol/L glucose) supplemented with 10% FBS and penicillin/streptomycin (100 units/mL) at 37°C and 5% CO₂. After 6 h, cells were replaced with fresh medium and cultured for subsequent analysis.

Human hepatoma cell line HepG2 cells and 293T cells were grown in DMEM (4.5 g/L glucose) supplemented with 10% FBS and penicillin/streptomycin (100 units/mL) in a humidified 5% CO₂ atmosphere at 37°C. Cells were subcultured two to three times a week. Transient transfections with plasmids or siRNA were performed using Lipofectamine 3000 or Lipofectamine RNAiMAX Reagent (Thermo Fisher Scientific) according to the manufacturer’s manual. The siRNA for mouse Etv5 was CCAUCAGAAUUCCCUAUUUTT. The siRNA for human ETV5 was GCUCUCUCCGCUAUUACUATT.

Recombinant lentivirus containing the shETV5 construct was generated by Cyagen (Guangzhou, China) using the sequence TACATGAGAGGCGGGTATTTC or CCGAAGGCTTCGCTTACTAAG targeting mouse Etv5 gene.

HepG2 cells were first transfected with siRNA against ETV5 or scramble siRNA. After 24 h, the cells were seeded to the Seahorse XFe96 cell culture microplate for another 24 h. The cells then were changed to substrate-limited medium (Seahorse XF Base Medium plus 0.5 mmol/L glucose, 1.0 mmol/L GlutaMAX, 0.5 mmol/L carnitine, and 1% FBS, pH 7.4) and incubated for overnight. The Cell Mito Stress Test was performed according to the manufacturer’s instructions.

HEK293T cells cultured in 96-well plates were transfected with the reporter plasmid (100 ng) in combination with other plasmids as indicated in the figures. The vector pRL-TK (10 ng) expressing Renilla luciferase was transfected in all samples and served as the internal control. Cells were lysed 24 h posttransfection, and the luciferase activity was examined by the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. The firefly luciferase activity was normalized to the Renilla luciferase activity and presented as the relative luciferase activity.

HepG2 cells were transfected with pCMV-ETV5-myc or pCMV-myc vector for 48 h, and cell lysates were collected for chromatin immunoprecipitation (ChIP) assay. The assay was performed using a ChIP kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. All enrichment changes were normalized to the input.

The electrophoretic mobility shift assay was conducted using the LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific). Briefly, 10 nmol/L biotin-labeled DNA substrates were incubated with indicated concentrations of ETV5 protein in 1× binding buffer containing 5 mmol/L MgCl₂, 2.5% glycerol, 10 mmol/L EDTA, and 0.05% NP-40 for 20 min at room temperature. Reactions were then stopped by the addition of 5 μL of gel loading buffer, resolved on a 5% native polyacrylamide gel, and transferred to a nylon membrane on ice. After ultraviolet (120 mJ/cm²) crosslinking of the DNA to membrane, biotin-labeled DNA was detected using the Chemiluminescent Nucleic Acid Detection Module Kit (Thermo Fisher Scientific). The double stranded DNA probes were end labeled with biotin and used as probes. The PPRE sequences were as follows: 5′ GTCGACAGGGGACCAGGACAAAGGTCACGTTCGGGAGTCGA; 3′ GTCGACTCCCGAACGTGACCTTTGTCCTGGTCCCCTGTCGAC.

Total RNA from tissues of mice was extracted, and cDNA synthesis and SYBR green gene expression assays were performed as described previously (13).

Western blotting of 40 μg of protein lysates was performed as described previously (13).

All data are presented as mean ± SEM. Statistical differences were analyzed by unpaired Student t test or one-way ANOVA followed by Bonferroni test. All statistical analyses were performed using GraphPad Prism software. P < 0.05 was considered statistically significant.

The data sets generated and/or analyzed during the current study are available from the corresponding authors upon reasonable request.

To investigate the potential relevance of ETV5 in hepatic lipid metabolism, we first examined the expression of hepatic ETV5 in HFD-induced obese or db/db mice, the genetic model of severe obesity and NAFLD. The protein level of hepatic ETV5 was significantly increased either under the HFD condition or in the db/db mice (Fig. 1A and B). In an in vitro model of steatotic primary hepatocytes induced by oleate-BSA, we also observed an increase of ETV5 protein level (Fig. 1C). Interestingly, the mRNA levels of ETV5 were not significantly changed under these conditions (Fig. 1A–C). These data suggest that the hepatic ETV5 is associated with liver steatosis, and its level could be altered upon long-term overnutrition by a posttranslational mechanism.

We next investigated the potential mechanism by which ETV5 is upregulated in obese livers. mTORC1 is a protein complex linking the nutrient supply with protein synthesis (15). Previously, we found that ETV5’s expression was regulated by mTORC1 signaling activity in endocrine cells through posttranslational mechanisms (16). We thus examined whether ETV5 expression was also regulated by mTORC1 signaling activity in hepatocytes. HepG2 cells and primary hepatocytes were treated with siRNA interfering the mTORC1 essential component Raptor or mTORC1 pathway inhibitor rapamycin. Both treatments efficiently inhibited mTORC1 signaling demonstrated by decrement in the pS6, the downstream target of mTORC1 signaling, and decreased the protein level of ETV5 (Fig. 1D–F). The mRNA levels were not altered under either condition (Fig. 1D–F). Furthermore, inhibition of mTORC1 signaling in vivo by injecting the db/db mice with rapamycin at a dose of 1 mg/kg body weight for 14 days blocked the increase of ETV5 in the livers (Fig. 1G). These data suggest that ETV5 is a downstream target of the mTORC1 signaling pathway in hepatocytes.

Interestingly, we found that ETV5 could regulate mTORC1 signaling activity in turn. The mTORC1 signaling activity evidenced by the ratio of pS6 to total S6 was significantly reduced in HepG2 cells with ETV5 depletion (Fig. 1H). mTORC1 is composed of the protein kinase mTOR and two other essential core components, Raptor and GβL/mLST8. Examination of mTORC1 components showed that mTOR and GβL/mLST8 were significantly reduced, whereas Raptor remained unchanged in the condition of ETV5 deficiency (Supplementary Fig. 1A). Furthermore, overexpression of ETV5 protein in primary hepatocytes or HepG2 cells increased pS6 level and GβL/mLST8 (Fig. 1I and J and Supplementary Fig. 1B). These data suggest that ETV5 could in turn enhance mTORC1 signaling activity in hepatocytes.

Our preliminary observation showed that nutrition status altered ETV5’s protein level only, but not mRNA level; thus, we examined ETV5 protein stability of using cycloheximide chase assay. Treatment of HEK293T and HepG2 cells with oleic acid markedly reduced the degradation rate of newly synthesized ETV5 protein in the presence of cycloheximide (Supplementary Fig. 2A and B). These data suggest that overnutrition increases hepatic ETV5 levels by enhancing ETV5 protein stability.

To explore the effects of ETV5 on hepatic fatty acid metabolism, lentiviruses containing Etv5 shRNA or scramble control shRNA were generated and administered into C57BL/6 mice through tail vein injection. As shown in Fig. 2, the mRNA and protein levels of ETV5 in livers were significantly suppressed. Injection of ETV5 lentivirus had no effect on hepatic expression of ETV1 and ETV4 (Fig. 2A).

These mice were fed ND or HFD for 4 weeks. The body/liver weight, body fat mass, food intake, and blood glucose levels were not significantly changed after 4 weeks (Supplementary Fig. 3A–E). Although the liver weight was not changed, the hepatic total TG and TC contents were significantly increased upon acute depletion of hepatic Etv5 in mice fed HFD (Fig. 2E). These alterations were further confirmed by gross morphology, hematoxylin-eosin staining and Oil Red O staining (Fig. 2B–D). Serum TG and TC levels remained largely unaltered (Fig. 2F). Plasma levels of liver enzymes (ALT, AST) and hepatic levels of proinflammatory cytokines (tumor necrosis factor-α [TNF-α], interleukin-1β [IL-1β]) were virtually not affected (Fig. 2G and H).

The db/db mouse is the leptin receptor–deficient mouse model that is characterized by severe obesity, NAFLD, hyperglycemia, and hyperinsulinemia. Consistent with the HFD-induced obese C57BL/6 mice, db/db mice with ETV5 deficiency in liver accumulated more hepatic lipids compared with db/db mice with a normal ETV5 level (Fig. 3A–C). The insulin level was also markedly increased, indicating aggravated insulin resistance (Fig. 3D). All these data suggest that ETV5 deficiency exacerbates lipid dyshomeostasis in diet- and genetically induced obese mice.

To further verify the biological role of hepatic ETV5, we generated ETV5^fl/fl mice and crossed them with Alb-Cre mice to generate ETV5-LKO mice (Fig. 4A and B). Under the ND condition, their gross morphology was comparable to their littermates. No significant difference was observed in body/liver weight and fasting glucose between ETV5^fl/fl and ETV5-LKO mice at 8 weeks of age (Supplementary Fig. 4A–C). Although the liver weight was normal, liver TG was significantly increased as revealed by increased liver TG content and Oil Red O staining when fasting (Fig. 4C and D). The serum TG, TC, NEFA, and insulin levels were comparable between these two genotypes of mice (Fig. 4E–G). Hepatic inflammation is another hallmark and key mediator of NAFLD progress. Previous studies have demonstrated that ETV5 is implicated in the inflammatory process (17,18). Therefore, we analyzed whether the hepatic inflammation was affected by the deficiency of ETV5 in hepatocytes. The mRNA levels of proinflammatory cytokines in the liver, such as TNF-α, IL-1β, and IL-6, were all comparable between the two genotypes of mice (Fig. 4H). In addition, serum ALT and AST levels, the liver injury markers, remained unchanged (Fig. 4I). It is worth of noting that the mTORC1 signaling was significantly reduced in ETV5-LKO livers as evidenced by reduced pmTOR/mTOR and pS6/S6 levels (Fig. 4J).

Blood glucose tolerance and insulin tolerance were not affected in ETV5-LKO mice (Supplementary Fig. 4D and E). Adipose tissue weight and adipokine levels, including adiponectin and leptin, also remained unchanged (Supplementary Fig. 4F–H).

Liver fatty acid metabolism is determined by the balance of lipogenesis, synthesis, oxidation, lipolysis, and transport. To explore the physiological mechanism for the increase of lipid content in ETV5-deficient mice, we analyzed genes involved in fatty acid metabolism using RNA sequencing. As shown in Fig. 5A, Kyoto Encyclopedia of Genes and Genomes analysis revealed that the pathways related to PPAR signaling and fatty acid degradation/metabolism/elongation were significantly altered (Fig. 5A). Most genes involved in PPAR signaling and fatty acid metabolism pathways were downregulated (Fig. 5B). Further analysis by real-time PCR confirmed a significant decrease in the fatty acid degradation genes. Lipolysis-relevant genes, such as Atgl and Lipe (Fig. 5C), as well as β-oxidation–relevant gene Lcad (Fig. 5D) were markedly decreased in ETV5-LKO mice. In addition, most genes involved in the lipogenesis and lipid synthesis were not changed (Fig. 5E and F). Furthermore, protein levels of CPT1A and ATGL were significantly reduced compared with littermate control mice (Fig. 5G). Consistent with this observation, β-hydroxybutyrate, one of the ketone bodies indicating indirect β-oxidation level, was notably decreased in ETV5-LKO livers (Fig. 5H).

To confirm the in vivo observation, we examined the effects of ETV5 deficiency in HepG2 cells. HepG2 cells were transfected with siRNA against ETV5, and the knockdown efficiency was confirmed by real-time PCR (Fig. 6A). HepG2 cells with reduced ETV5 exhibited increased lipid accumulation under oleic acid treatment (Fig. 6B and C). The expression of genes related to β-oxidation and lipolysis were both decreased (Fig. 6D and E). The cellular β-oxidation ability was assessed by the oxygen consumption rate using the Seahorse XFe96 Flux Analyzer. Notably, ETV5 deficiency resulted in a decrease in basal, maximal, ATP production, spare capacity, and proton leak. These deficiencies were exacerbated upon oleic acid stimulation (Fig. 6F–I). Thus, hepatocytes lacking ETV5 demonstrate a lower capacity of mitochondrial respiration function under lipid-stressed conditions.

We also examined whether overexpression of ETV5 altered lipid accumulation in hepatocytes. Surprisingly, ectopic expression of ETV5 did not affect the intracellular TG deposition under either basal or oleate-stimulated conditions in HepG2 cells (Supplementary Fig. 5). Therefore, ETV5 is required for the optimal β-oxidation processes of fatty acid metabolism but not sufficient to facilitate fatty acid utilization per se.

We next sought to investigate the molecular basis responsible for downregulation of fatty acid degradation genes. Fatty acid degradation or β-oxidation is mainly regulated by PPARα. However, we did not observe any change in PPARα protein level (Fig. 5G). Furthermore, a co-immunoprecipitation experiment revealed no interaction between ETV5 and PPARα (Fig. 7A). We then speculated that PPARα might not be a direct transcriptional target of ETV5. Instead, ETV5 might be involved in the regulation of PPRE transactivity. Indeed, quantitative PCR analysis confirmed that the majority of PPAR-signaling downstream genes were significantly suppressed upon ETV5 depletion in HepG2 cells (Fig. 7B). Overexpression of ETV5 markedly enhanced the PPRE transactivity in 293T cells (Fig. 7C). The dual-luciferase assay was performed to test whether ETV5 enhances the transactivation by PPARα ligands, fibrates, or GW7647. Our results showed that ETV5 indeed further enhanced the transactivity by PPARα with the PPARα agonists (Fig. 7D and Supplementary Fig. 6A–C). ETV5 is composed of the ETS DNA-binding domain and the acidic domains at the N and C termini, which constitutes the transactivation core (Fig. 7E). We thus analyzed the transactivation activity of these different domains. ETV5-truncating fragments 1–72 amino acids (aa), 1–367 aa, and 367–510 aa almost totally abolished its transactivity. Only ETV5 (1–450 aa) lacking the C-terminal maintained the majority of transactivity. This result suggests that both the ETS DNA-binding domain (367–450 aa) and the N-terminal acidic domain are critical for the transactivation activity of ETV5 on PPRE (Fig. 7F). ETV5 has two phosphorylation sites, serine 248 for mitogen-activated protein kinase and serine 367 for cAMP-dependent protein kinase activation (19). We mutated these two sites with alanine and found that only S248A mutants decreased the effects of ETV5 on PPRE transactivity (Fig. 7G).

Furthermore, we analyzed the promoter region of the Acsl1 gene, which is one of the PPAR pathway downstream genes. The promoter region of Acsl1 (−2000 to 200 base pairs) contains three PPREs (Fig. 7H). HepG2 cells were transfected with pCMV-ETV5-myc or pCMV-myc vector for 48 h. ChIP assay was performed using anti-myc antibody or IgG antibody as the negative control. Real-time PCR results of the ChIP assay showed the association of ETV5 with the Acsl1 promoter (Fig. 7I). We also performed the gel shift assay and confirmed the binding of ETV5 to the PPRE sequence (Fig. 7J). All these data suggest that ETV5 is a novel factor that can bind to PPRE and is required for optimal PPAR signaling through enhancing PPRE transactivity.

The current study demonstrates that ETV5 is involved in hepatic fatty acid metabolism. Hepatic ETV5 expression is increased in steatotic liver of HFD-induced or db/db obese mice as well as in steatotic hepatocytes induced by oleate. Moreover, mTORC1 signaling activity stimulates the protein expression of ETV5, which in turn enhances mTORC1 signaling activity, suggesting a feed-forward regulatory manner. Deficiency of ETV5 induced by shRNA interference or LKO increases hepatic lipid accumulation. Mechanistically, ETV5 could bind to and enhance the PPRE signaling activity and is required for the optimal β-oxidation and lipolysis processes. Collectively, we uncovered ETV5 as a novel transcriptional factor critical for the fatty acid metabolism process in liver.

ETV5 belongs to the PEA3 group of ETS transcription factors. The PEA3 family consists of three members—ETV1 (Er81), ETV4 (Pea3), and ETV5 (Erm)—which share the conserved ETS domain and transactivating domain (4). ETV5 could bind to a specific response element within the promoter and regulatory regions of its target genes. Previous studies have focused on its biological functions in the development and carcinogenesis processes. Genomic studies have indicated that ETV5 is associated with diabetes and obesity (20). Dysregulation of several ETS target genes is linked to insulin secretion defect (8), suggesting that ETS transcriptional factors may contribute to glycemic control. Our study has revealed a novel function for ETV5 in hepatic lipid metabolism. Acute depletion of ETV5 aggravates diet- or genetically induced hepatic steatosis. In mice with chronic deficiency of ETV5, hepatic lipid accumulation is also increased when fasting.

Hepatic lipid metabolism is tightly regulated by a group of transcriptional factors that determine the balance between hepatic TG anabolism, including fatty acid uptake and de novo lipogenesis, and catabolism, such as fatty acid oxidation and export as VLDL. Among these transcriptional factors, PPARα is well characterized as the master transcriptional regulator of hepatic fatty acid catabolism. Downstream targets of PPARα include genes involved in lipid oxidation and metabolism, such as fatty acid degradation, synthesis, transport, storage, lipoprotein metabolism, and ketogenesis during fasting (21,22). Mice lacking PPARα develop steatosis during fasting because of the inability to oxidize free fatty acids released from adipocytes. In ETV5-deficient mouse or cell models, there is no change in the expression of PPARα. However, the PPAR pathway is significantly altered, as evidenced by the reduced downstream PPAR target genes and impaired β-oxidation processes. These observations suggest that ETV5 may affect the PPAR pathway through modulating PPRE transactivity. PPRE, which is typically localized in gene regulatory regions, consists of direct repeats of two hexamer core sequences, AGG(A/T)CA, separated by one nucleotide (DR-1) (23). ETV5 binds to the DNA as other ETS genes through the winged-helix-turn-helix DNA-binding motif (Ets domain) that recognizes DNA sequences that contain a GGAA/T core element. They determine the binding preference through the sequence flanking the core binding sequence (4). The similarity of these two response elements indicates that ETV5 might also bind to the PPRE and regulate the transactivity of PPRE. This concept is supported by the following observations: 1) ETV5 remarkably enhances the PPRE transactivity; 2) ETV5 further enhances the transactivity by PPARα with PPARα agonists; 3) deletion of its ETS binding domain abolishes the PPRE transactivity; 4) ETV5 is physically associated with Acsl1 promoter, which contains three PPREs; and 5) gel shift assay demonstrates the binding of ETV5 to the PPRE sequence.

How ETV5 expression is regulated remains largely unknown. Previous studies have suggested that protein levels of ETV5 may be altered by signaling pathways relevant to development, oncogenesis, and metastasis, such as Ret, FGF2, and Ras-ERK signaling (4). Our study has identified the mTORC1 signaling pathway as the critical mechanism by which overnutrition stimulates hepatic expression of ETV5. This concept is supported by the following observations: 1) ETV5 protein level is significantly increased in both in vivo and in vitro models of hepatic steatosis, and 2) both pharmacological and genetic inhibition of mTORC1 signaling suppress the protein expression of ETV5. mTORC1 is well characterized as a signal integrator and master regulator of cellular anabolic processes linked to cell growth and survival (15). Activation of mTORC1 signaling increases translation and ribosome biogenesis. Consistently, our study indicates that upregulation of ETV5 by mTORC1 signaling occurs at the posttranscriptional level since neither Raptor silence nor rapamycin alter the mRNA levels of ETV5 in hepatocytes. Interestingly, ETV5 in turn stimulates mTORC1 signaling through regulating the mTORC1 components, suggesting a positive feedback loop. The role of mTOR in obesity and hepatic lipid metabolism is complicated and varies with the level of mTOR signaling. Complete suppression of mTORC1 activity promotes lipid accumulation by inhibiting lipolysis and β-oxidation and concurrently stimulating lipogenesis through SREBP1 (15). However, moderate mTORC1 activation by TSC1 or DEPTOR deletion increases oxidative metabolism in hepatocytes and adipocytes (24–26) and is also associated with defective SREBP1c activation and lipogenesis in the liver (27). In addition, inhibition of the mTOR by rapamycin could decrease mitochondrial gene expression and oxygen consumption through a YY1-PGC-1α transcriptional complex in skeletal muscle tissues (28). Therefore, the complex mTOR signaling confers multiple layers of regulation in the mitochondrial oxidative and lipid metabolism in the hepatocytes. Our present study has identified ETV5 as a new mTOR downstream signaling molecule that contributes to the intricate regulation of mitochondrial oxidation.

Regulation of mTOR signaling by ETV5 is evidenced by reduced pmTOR and pS6 levels in ETV5-LKO liver and HepG2 cells. The mTORC1 components, such as GβL/mLST8 or mTOR, were significantly altered upon change of ETV5 levels. This observation suggests a reciprocal regulation between mTORC1 and ETV5.

In summary, our findings uncover a novel function for ETV5 in hepatic fatty acid metabolism mainly through affecting the PPRE signaling activity. ETV5 binds to the PPRE sequence and is required for the optimal PPAR-mediated fatty acid degradation process (β-oxidation and lipolysis). ETV5 responds to the extracellular energy status and nutrient availability in an mTORC1-dependent manner. Demonstration of the role of ETV5 in lipid metabolism as well as in the downstream target of mTORC1 signaling may provide an alternative strategy for the intervention of mTORC1-driven metabolic disorders, including liver steatosis.

Acknowledgments. The authors acknowledge Yuan Liu and Ruolu Bao (Health Science Center, Shenzhen University) for technical support.

Funding. This work was funded by the National Natural Science Foundation of China (81730020, 81930015, 81870405), the National Key R&D Program of China (2017YFC0908900), the Guangdong Medical Science and Technology Research Fund project A2019422, the Shenzhen Science and Technology Project Innovation Commission (JCYJ20190808144203802), an SZU Top Ranking Project (86000000210), and the SZU medical young scientists program (71201-00000-1).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. Z.M. and W.Z. developed the study rationale, supervised the study, and wrote the manuscript. Z.M. and M.F. designed and performed most of the experiments. Z.L., M.Z., L.X., K.P., S.W., and W.S. performed the experiments and assisted with data analysis. Z.M. and W.Z. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the 55th Annual Meeting of the European Association for the Study of Diabetes, Barcelona, Spain, 16–20 September 2019.