Skip to main content
NCBI home page
EMBO Reports logoLink to EMBO Reports
. 2021 Apr 20;22(5):e52146. doi: 10.15252/embr.202052146

Curcumin prevents obesity by targeting TRAF4‐induced ubiquitylation in m6A‐dependent manner

Yushi Chen 1,2,3,4, Ruifan Wu 1,2,3,4, Wei Chen 1,2,3,4, Youhua Liu 1,2,3,4, Xing Liao 1,2,3,4, Botao Zeng 1,2,3,4, Guanqun Guo 1,2,3,4, Fangfang Lou 5, Yun Xiang 5, Yizhen Wang 1,2,3,4, Xinxia Wang 1,2,3,4,✉
PMCID: PMC8097347  PMID: 33880847

Abstract

Obesity has become a major health problem that has rapidly prevailed over the past several decades worldwide. Curcumin, a natural polyphenolic compound present in turmeric, has been shown to have a protective effect on against obesity and metabolic diseases. However, its underlying mechanism remains largely unknown. Here, we show that the administration of curcumin significantly prevents HFD‐induced obesity and decreases the fat mass of the subcutaneous inguinal WAT (iWAT) and visceral epididymal WAT (eWAT) in mice. Mechanistically, curcumin inhibits adipogenesis by reducing the expression of AlkB homolog 5 (ALKHB5), an m6A demethylase, which leads to higher m6A‐modified TNF receptor‐associated factor 4 (TRAF4) mRNA. TRAF4 mRNA with higher m6A level is recognized and bound by YTHDF1, leading to enhanced translation of TRAF4. TRAF4, acting as an E3 RING ubiquitin ligase, promotes degradation of adipocyte differentiation regulator PPARγ by a ubiquitin–proteasome pathway thereby inhibiting adipogenesis. Thus, m6A‐dependent TRAF4 expression upregulation by ALKBH5 and YTHDF1 contributes to curcumin‐induced obesity prevention. Our findings provide mechanistic insights into how m6A is involved in the anti‐obesity effect of curcumin.

Keywords: adipogenesis, ALKBH5, curcumin, m6A, obesity

Subject Categories: Metabolism; Post-translational Modifications, Proteolysis & Proteomics; RNA Biology

Curcumin exerts its anti‐obesity effect by increasing the ALKBH5‐mediated m6A modification of TRAF4 mRNA, and enhancing TRAF4 translation in an YTHDF1‐dependent manner. TRAF4 in turn promotes ubiquitination of PPARγ and inhibits adipogenesis.

graphic file with name EMBR-22-e52146-g006.jpg

Introduction

Obesity is a complex pathophysiology and is closely associated with various chronic metabolic disorders, and it has become a major health burden that has rapidly prevailed over the past several decades worldwide (Tsai et al, 2011; Jung et al, 2017). Thus, to develop effective therapeutic strategies to reduce obesity is urgent.

Curcumin (1,7‐bis(4‐hydroxy‐3‐methoxyphenyl)‐1,6‐heptadiene‐3,5‐dione) is a polyphenol derived from the rhizome of Curcuma longa (turmeric), which has been widely consumed daily throughout Asian countries over centuries without reported toxicity (Ejaz et al, 2009). Curcumin has attracted attention in the medical field because of its potentially diverse health‐promoting effects such as antioxidant (Wafi et al, 2019), anti‐inflammation (Ikram et al, 2019), and anti‐cancer properties (Mirzaei et al, 2018). An increasing body of evidence shows that curcumin can prevent and (or) alleviate various chronic diseases like neurodegenerative disorders (Ghosh et al, 2015), inflammatory diseases (Hanai & Sugimoto, 2009), and cardiovascular diseases (Wongcharoen & Phrommintikul, 2009). In recent years, studies demonstrated that curcumin also had beneficial effects in obesity and diabetes (Ghosh et al, 2015). For instance, dietary supplementation with curcumin to 80 mg/kg body weight decreased fasting plasma glucose and improved insulin sensitivity in obese rats (El‐Moselhy et al, 2011). Similar results also have been reported in mice (Shao et al, 2012) and humans (Ghorbani et al, 2014). It has been shown that curcumin inhibits adipogenesis by blocking mitotic clonal expansion process, upregulating adipocyte energy metabolism and apoptosis (Ejaz et al, 2009; Kim et al, 2011; Shao et al, 2012), and targeting PPARγ, CCAAT/enhancer‐binding protein alpha (C/EBPα), and KLF5, as well as some enzymes involved in fatty acid metabolism like CPT‐1, FAS, and GPAT‐1 (Ejaz et al, 2009; Kim et al, 2011; Yan et al, 2018). In addition, Wnt/β‐catenin signaling pathway has been reported to participate in curcumin‐mediated suppression of adipogenesis in 3T3‐L1 cells (Ahn et al, 2010; Tian et al, 2017). However, the underlying mechanisms of curcumin’s anti‐obesity effect are not fully understood.

Recent evidence showed EGCG (epigallocatechin gallate), also a polyphenol, inhibited adipogenesis in an mRNA m6A‐YTHDF2‐dependent manner (Wu et al, 2018b). N6‐methyladenosine (m6A) is the most plentiful internal RNA modifications in eukaryotes (Wang et al, 2015b). It has been proposed that m6A could regulate adipogenesis through mediating mRNA splicing (Zhao et al, 2014), mitotic clonal expansion (Wu et al, 2018a), JAK2‐STAT3‐C/EBPβ (Wu et al, 2019b), and autophagy pathway (Wang et al, 2020b). Curcumin has been shown to change the abundance of m6A in various tissues of mammal (Lu et al, 2018; Gan et al, 2019), which raises the question of whether the anti‐obesity effect of curcumin is also related to RNA methylation.

In the present study, we revealed that curcumin inhibited ALKBH5 expression, leading to increased m6A methylation of TRAF4 and enhanced expression of TRAF4 proteins mediated by YTHDF1. TRAF4, acting as an E3 ligase of PPARγ, induced ubiquitylation of the PPARγ protein. Consequently, PPARγ protein reduced and adipogenesis was further inhibited, suggesting a completely new way in which m6A mediated the anti‐obesity effect of curcumin.

Results

Curcumin prevents HFD‐induced obesity and related metabolic dysfunctions

To assess the effect of curcumin on obesity, the mice were fed with NCD, HFD, and HFD+ curcumin (CUR) for 12 weeks. The results showed that HFD group exhibited rapid increases in mice body weight despite eating significantly less than NCD group (Fig 1A–C). However, HFD‐induced weight gain was suppressed by CUR administration (Fig 1A and B). No difference was observed in food intake in mice between HFD and HFD+CUR groups (Fig 1C). In addition, the fat mass of the subcutaneous inguinal WAT (iWAT) and visceral epididymal WAT (eWAT) was significantly lower in HFD+CUR mice than those in HFD mice at 16 weeks of age (Fig 1D and E). In agree with this, histological analysis showed that mice in HFD+CUR group had markedly reduced lipid contents and adipocyte diameter in WAT compared with those in HFD group (Fig 1F and G), suggesting the less weight gain with administration of CUR was mainly associated with fewer fat mass. H&E staining of liver showed that the administration of curcumin dramatically reversed HFD‐induced hepatic steatosis (Fig 1H). We next examined the effect of curcumin on glucose tolerance and insulin sensitivity. HFD+CUR mice showed enhanced glucose disposal ability and increased insulin sensitivity (Fig 1L–O), suggesting curcumin preserves glucose homeostasis. The serum of HFD+CUR mice showed lower levels of glucose, total triglyceride, and free fatty acid (FFA) than those of HFD mice (Fig 1I–K). These results demonstrate that curcumin can protect against HFD‐induced obesity by regulating glucose homeostasis.

Figure 1. Curcumin prevents HFD‐induced obesity.

Figure 1

Open in a new tab
  • A–C
    Appearance, body weight trajectories, and food intake of the mice.
  • D
    Representative pictures of iWAT and eWAT.
  • E
    iWAT, eWAT, and liver weights at termination of study.
  • F, G
    H&E staining and quantification of size of adipocytes of fixed iWAT and eWAT (scale bar = 200 μm).
  • H
    Changes in the morphologies of livers and H&E staining depicted steatosis as circular white gaps caused when the dehydration process leaches the fat out of fixed livers (scale bar = 200 μm).
  • I–K
    Fasting blood glucose, serum triglyceride, and FFA levels of mice after 12 weeks on HFD.
  • L–O
    The blood glucose level of mice after intraperitoneal injection of glucose or insulin for glucose (GTT) (L) and insulin tolerance tests (ITT) (N), and AUC, area under the curve (M, O).

Data information: Data are representative of the mean ± s.d. of n = 6 mice per group and were analyzed using two‐way ANOVA and Tukey’s multiple comparison test; *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

Curcumin inhibits adipogenesis through attenuating ALKBH5 expression

To gain insight into the role of curcumin on adipogenesis, we incubated primary stromal vascular fraction (SVF) or 3T3‐L1 cells with various concentrations of curcumin throughout the 8‐day‐long differentiation and processed for analysis of lipid accumulation and gene expression by ORO and qPCR, respectively. ORO analysis showed that adipogenesis was inhibited by curcumin in a concentration‐dependent manner with maximum effects observed at 20 μM (Fig 2A and C). qPCR analysis showed that the expression of adipogenesis master regulators, including PPARγ, C/EBPα, and fatty acid‐binding protein 4 (FABP4), was all attenuated by curcumin also in a dose‐dependent manner (Fig 2B and D). Because a recent study showed that the protective effect of curcumin on LPS‐induced hepatic lipid metabolism disorders might be because of the increased m6A RNA level (Lu et al, 2018), we prompted to assess whether curcumin affects m6A level in adipocytes. We measured total m6A modified mRNA levels using HPLC–QqQ–MS/MS. The results showed both iWAT and eWAT in mice from HFD+CUR group or curcumin‐treated 3T3‐L1 cells exhibited elevated m6A level upon curcumin treatment (Fig 2E and H). To further analyze which m6A regulator contributes to the enhanced m6A level, the mRNA and protein expression of methyltransferases (MTETTL3 and METTL14) and demethylases (FTO and ALKBH5) were examined. Intriguingly, we observed that the mRNA and protein expression of ALKBH5, not METTL3, METTL14 or FTO, were significantly reduced in iWAT from mice in HFD+CUR group and SVF/3T3‐L1 cells upon curcumin treatment (Figs 2F, G, I and J, and EV1A). These results indicated that the elevated m6A levels might be a result of curcumin‐induced ALKBH5 deficiency. To further verify the ALKBH5 mediated the effect of curcumin on adipogenesis, we performed rescue experiment and found that forced expression of ALKBH5 restored adipogenesis and triglyceride accumulation in cells inhibited by curcumin (Fig 2K). Consistently, the mRNA and protein expression of adipocyte key regulator genes, including PPARγ, C/EBPα, and FABP4, were remarkably reduced in curcumin‐treated cells, which could be elevated to normal level by the overexpression of ALKBH5 (Figs 2L and EV1B). Collectively, ALKBH5 acts as a mediator of curcumin‐inhibited adipogenesis.

Figure 2. Curcumin inhibits adipogenesis through attenuating ALKBH5 expression.

Figure 2

Open in a new tab
  • A
    ORO staining of SVF cells isolated from iWAT of male mice. Differentiation was induced with curcumin up to day 8 (scale bar = 500 μm).
  • B
    mRNA expression of adipogenesis‐related factors of SVF cells isolated from iWAT of male mice.
  • C
    ORO staining of 3T3‐L1 cells. Differentiation was induced with curcumin up to day 8 (scale bar = 500 μm).
  • D
    mRNA expression of PPARγ, C/EBPα, and FABP4 expression of 3T3‐L1 cells treated with curcumin.
  • E
    LC‐MS/MS quantification of the m6A/A in mRNA of iWAT of mice fed with HFD and HFD+CUR groups.
  • F, G
    mRNA and protein expression levels of m6A regulator in iWAT from mice in HFD and HFD+CUR groups.
  • H
    LC‐MS/MS quantification of the m6A/A in mRNA of 3T3‐L1 cells.
  • I, J
    mRNA and protein expression levels of m6A regulator in 3T3‐L1 cells were analyzed by qPCR.
  • K
    Oil Red O staining of vector or OE‐ALKBH5‐transfected 3T3‐L1 cells. Differentiation was induced with curcumin up to day 8 (scale bar = 500 μm).
  • L
    mRNA expression of PPARγ, C/EBPα, and FABP4 expression of cells described in (K).

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

Figure EV1. The expression levels of m6A regulator in differentiated SVFs upon curcumin treatment.

Figure EV1

Open in a new tab
  • A
    Western blot analysis of m6A regulator in primary SVFs and the quantification of protein levels normalized to β‐actin expression.
  • B
    Western blot analysis of PPARγ and C/EBPα protein levels in cells described in Fig 2K and quantification of protein levels normalized to β‐actin expression.

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

ALKBH5 modulates adipogenesis via regulating TRAF4 expression

To further explore how ALKBH5 regulates adipogenesis, we conducted bioinformatics analysis of the potential target gene of ALKBH5 from GEO database, which provides RNA profile data of 3T3‐L1 cells differentiation (GSE69313, GSE6794). TRAF4, a member of the TRAF family, attracted our attention, which are involved in numerous cellular physiological processes. A recent study showed that the downregulation of TRAF4 during adipogenesis of the mesenchymal stem cells (MSCs) harvested from human bone marrow was regulated by ALKBH5 (Cen et al, 2020). Thus, we hypothesized that ALKBH5 might regulate adipogenesis of mouse 3T3‐L1 cells via TRAF4 expression. To this end, we extracted the mRNA expression of ALBKH5 and TRAF4 from these 34 samples and carried out Pearson's correlation analysis. Analysis revealed that during the process of adipogenesis, TRAF4 was negatively correlated with ALKBH5 expression in 3T3‐L1 cells (Fig 3A). To further confirm this, we first detect TRAF4 expression in iWAT from mice in HFD+CUR group and SVF, 3T3‐L1 cells, and the experiments showed a profound increase in TRAF4 protein expression upon curcumin‐treated (Fig EV2A–C); next, we performed rescue experiment and observed that knockdown of TRAF4 could partially rescue the inhibition of adipogenesis and triglyceride accumulation caused by curcumin (Fig EV2D–F). To further substantiate the hypothesis, we conducted Western blot assays in 3T3‐L1 cells and found that ALKBH5 depletion promoted TRAF4 protein expression (Fig 3B), and, reversely, ALKBH5 overexpression reduced protein abundance of TRAF4 in 3T3‐L1 cells (Fig 3C), indicating the inverse correlation between ALKBH5 and TRAF4. To validate their upstream–downstream relationship, we silenced TRAF4 and found the protein abundance of ALKBH5 was unchanged by siTRAF4 (Fig 3D), suggesting ALKBH5 was the upstream of TRAF4. We next sought to address whether TRAF4 mediated the effect of ALKBH5 on adipogenesis. We performed rescue experiment and observed that ALKBH5 depletion inhibited lipid accumulation, whereas silencing of TRAF4 could restore the siALKBH5‐induced poor adipogenesis (Fig 3E). Consistently, the mRNA and protein expression of PPARγ and C/EBPα could be restored upon depletion of TRAF4 in ALKBH5‐deleted cells (Fig 3F and G). Taken together, these results demonstrate that ALKBH5 modulates adipogenesis via regulating TRAF4 expression.

Figure 3. ALKBH5 modulates adipogenesis via regulating TRAF4.

Figure 3

Open in a new tab
  • A
    The relationship between ALKBH5 and TRAF4 was analyzed in GEO data base. Pearson's correlation analysis indicated TRAF4 and ALKBH5 had negative relationship (Cor = −0.70).
  • B
    Western blot analysis of TRAF4 protein levels in control and ALKBH5‐depleted cells after MDI‐induced for 48 h and quantification of protein levels normalized to β‐actin expression.
  • C
    Western blot analysis of TRAF4 protein levels in control and ALKBH5‐overexpressing cells after MDI‐induced for 48 h and quantification of protein levels normalized to β‐actin expression.
  • D
    Western blot analysis of PPARγ and C/EBPα protein levels in control and TRAF4‐depleted cells after MDI‐induced for 48 h and quantification of protein levels normalized to β‐actin expression.
  • E
    Oil Red O staining of control, ALKBH5‐depleted, and ALKBH5+TRAF4‐depleted cells after induced for 8 days (scale bar = 500 μm).
  • F
    qPCR analysis of PPARγ, C/EBPα, and FABP4 expression in cell described in (E). β‐Actin was used as an internal control.
  • G
    Western blot analysis of PPARγ and C/EBPα protein levels in cells described in (E) and quantification of protein levels normalized to β‐actin expression.

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

Figure EV2. TRAF4 expression contributes to curcumin‐induced obesity prevention.

Figure EV2

Open in a new tab
  • A
    Protein expression levels of TRAF4 in iWAT of mice fed with HFD and HFD+CUR.
  • B
    Protein expression levels of TRAF4 in primary SVFs treated with curcumin.
  • C
    Protein expression levels of TRAF4 in 3T3‐L1 adipocytes treated with curcumin.
  • D
    Oil Red O staining of rescue experiment of TRAF4 on day 8 of differentiation and the quantification of ORO staining (scale bar = 500 μm).
  • E
    Western blot analysis of PPARγ and C/EBPα protein levels in cells described in (D) and the quantification of protein levels normalized to β‐actin expression.
  • F
    qPCR analysis of PPARγ, C/EBPα, and FABP4 expression in cell described in (D). β‐Actin was used as an internal control.

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

ALKBH5 decreases TRAF4 protein expression by increasing its m6A level

We next investigated how ALKBH5 regulates TRAF4 protein expression. Since ALKBH5 is one of the demethylases, we raise a question whether the m6A demethylase activity of ALKBH5 is necessary for suppressing TRAF4 expression. Wild‐type (ALKBH5‐WT) and catalytic mutant ALKBH5H204A (ALKBH5‐MUT) vectors (Zhang et al, 2017) were constructed. The impact of ectopic expression of ALKBH5‐WT or ALKBH5‐MUT on cellular m6A level was detected by LC‐MS/MS (Fig 4A). Overexpression of ALKBH5‐WT decreased TRAF4 protein expression, whereas ALKBH5‐MUT had no effect on TRAF4 expression compared to the empty vector (Fig 4B), implying that ALKBH5 modulated TRAF4 protein expression in a demethylase activity‐dependent manner. According to the published m6A‐seq data of 3T3‐L1 (Zhao et al, 2014), two m6A sites were found at CDS (site #1) and 3ʹ UTR (site #2) of TRAF4 mRNA (Fig 4C). Knockdown of ALKBH5 significantly increased the m6A levels on site #2, but not site #1, of TRAF4 mRNA, indicating m6A site #2 was the target of ALKBH5 (Fig 4D). Furthermore, as expected, compared with ALKBH5‐MUT or the empty vector, ALKBH5‐WT overexpression markedly reduced the m6A level on site #2 of TRAF4 mRNA (Fig 4E). Next, to explore whether the m6A site on TRAF4 3’UTR was essential for ALKBH5‐mediated TRAF4 expression, a dual‐luciferase reporter and site‐directed mutagenesis assay was performed. Dual‐luciferase assays showed that the overexpression of ALKBH5‐WT, but not ALKBH5‐MUT, substantially increased the luciferase activity of reporter constructs containing wild‐type 3ʹUTR fragments of TRAF4, compare to the control (Fig 4F). This increase was abrogated when the m6A sites were mutated (A was replaced with T) (Fig 4F). To sum up, these results illustrate that ALKBH5 targets TRAF4 and increases its m6A level at 3’UTR of TRAF4 transcript.

Figure 4. ALKBH5 regulates the protein expression of TRAF4 via m6A‐YTHDF1‐dependent manner.

Figure 4

Open in a new tab
  • A
    LC‐MS/MS quantification of the m6A/A in mRNA of control, wild‐type (WT), and mutant (MUT) ALKBH5‐overexpressing cells.
  • B
    Western blot analysis of ALKBH5 and TRAF4 in control, WT, and MUT ALKBH5‐overexpressing cells and quantification of protein levels normalized to β‐actin expression.
  • C
    Integrative genomics viewer (IGV) plots of m6A peaks at TRAF4 mRNAs. The y‐axis shows sequence read number, blue boxes represent exons, and blue lines represent introns.
  • D
    MeRIP‐qPCR analysis of m6A levels of TRAF4 mRNA in control and ALKBH5‐depleted cells.
  • E
    MeRIP‐qPCR analysis of m6A levels of TRAF4 mRNA in control, WT, and MUT ALKBH5‐overexpressing cells.
  • F
    Left panel: Schematic diagram of dual‐luciferase reporter constructs. Right panel: Relative luciferase activity of WT or MUT (A‐to‐T mutation) TRAF4 3ʹUTR luciferase reporter in control, WT, and MUT ALKBH5‐overexpressing cells. Firefly luciferase activity was measured and normalized to Renilla luciferase activity.
  • G
    Western blot analysis of TRAF4 protein levels in control and YTHDF1‐overexpressing cells and quantification of protein levels normalized to β‐actin expression.
  • H
    Polysome profiling assays. The fractionation of lysates from 3T3‐L1 cells transfected with or without YTHDF1 plasmid is shown on the top. RNAs in different ribosome fractions were extracted and subjected to qPCR analysis.
  • I
    RIP‐qPCR analysis of the interaction of TRAF4 with FLAG in cells overexpressing FLAG‐YTHDF1. Enrichment of TRAF4 with FLAG was measured by qPCR and normalized to input.
  • J
    Relative luciferase activity of WT or MUT TRAF4‐3ʹUTR luciferase reporter in cells transfected with control or YTHDF1 plasmid. Firefly luciferase activity was measured and normalized to Renilla luciferase activity.
  • K
    Western blot analysis of YTHDF1, ALKBH5, and TRAF4 in control, ALKBH5‐depleted, and ALKBH5+YTHDF1‐depleted cells and quantification of protein levels normalized to β‐actin expression.
  • L
    Oil Red O staining of control, ALKBH5‐depleted, and ALKBH5+YTHDF1‐depleted cells on day 8 of differentiation (scale bar = 500 μm).
  • M
    qPCR analysis of PPARγ, C/EBPα, and FABP4 expression in cell described in (L). β‐Actin was used as an internal control.

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. *P < 0.05, **P < 0.01, and ***P < 0.001 compared to control group.

m6A accelerates TRAF4 protein expression in a YTHDF1‐dependent manner

Previous studies showed that m6A methylation affects mRNA stability or translation, depending on the recognition by specific m6A‐binding proteins. YTHDF1 is a m6A reader protein which facilitates the initiation of translation of m6A‐modified mRNAs (Wang et al, 2015a). Considering the positive correlation between m6A levels and protein expression of TRAF4, we proposed that YTHDF1 might mediate TRAF4 translation via binding m6A site of the transcript. To test this hypothesis, we overexpressed YTHDF1 in 3T3‐L1 cells and found forced expression of YTHDF1 significantly increased the protein abundance of TRAF4 (Fig 4G). Polysome fraction experiment showed forced expression of YTHDF1 in 3T3‐L1 cells increased the TRAF4 mRNA level in the polysome portion, which confirmed that YTHDF1 promoted the translation of TRAF4 (Fig 4H). RIP‐qPCR assay further validated that TRAF4 mRNA interacts with YTHDF1, indicating TRAF4 is a direct target of YTHDF1 (Fig 4I). In addition, forced expression of YTHDF1 significantly upregulated luciferase activity in reporters carrying wild‐type TRAF4 3′UTR fragment. However, such an increase was abrogated when the m6A consensuses sites were mutant (Fig 4J), suggesting the regulation by YTDHF1 is m6A‐dependent. Furthermore, siYTHDF1 eliminated the increased protein abundance of TRAF4 in ALKBH5‐depleted cells (Fig 4K). Knockdown of YTHDF1 could partially rescue the inhibition of adipogenesis and triglyceride accumulation caused by curcumin (Fig 4L). Consistently, the mRNA expression of PPARγ, C/EBPα, and FABP4 could be restored upon depletion of YTHDF1 in curcumin‐treated cells (Fig 4M). Taken together, our results demonstrated that YTHDF1 enhanced the TRAF4 translation in m6A‐dependent manner.

TRAF4 inhibits adipocyte differentiation by ubiquitination of PPARγ

The last question we need to answer is how TRAF4 inhibits adipogenesis. It is known that TRAF4 is an E3 ubiquitin ligase to conduct the ubiquitination of proteins, and adipogenetic regulator PPARγ protein is post‐translationally regulated by various modifications, including ubiquitination‐mediated proteomic degradation. Therefore, we hypothesized that TRAF4 promoted ubiquitination of PPARγ. To test this, we conducted gain or loss of TRAF4 experiment and found siTRAF4 dramatically enhanced PPARγ expression (Fig 5A), while overexpression of TRAF4 significantly reduced the protein abundance of PPARγ (Fig 5B). Importantly, the decreased PPARγ led by TRAF4 was efficiently restored by MG132, a proteasome inhibitor, suggesting that TRAF4 may downregulate PPARγ in a proteasome‐dependent degradation manner (Fig 5C). Evidence indicated that both autophagy–lysosome pathway and ubiquitin–proteasome pathway were activated in the process of adipogenic differentiation. We thus compared these two degradation pathways by using 3‐MA and MG132, respectively. The proteasome inhibitor MG132, rather than autophagy inhibitor 3‐MA, rescued endogenous PPARγ degradation (Fig 5D). This result suggested that that proteasomal activation, but not autophagy, was involved in endogenous PPARγ loss. The levels of endogenous PPARγ ubiquitinated were further detected in differentiated 3T3‐L1 cells. The results showed that an augment ubiquitination of endogenous PPARγ in differentiated 3T3‐L1 cells was detected under the table overexpression of TRAF4, a (Fig 5E), implicating TRAF4 acted as an E3 ligase of PPARγ. Altogether, these results suggest that TRAF4 decreases the stability of PPARγ protein via a proteasome‐dependent pathway.

Figure 5. TRAF4 suppresses adipogenesis through ubiquitination of PPARγ.

Figure 5

Open in a new tab
  • A
    Western blot analysis of PPARγ protein levels in control and TRAF4‐depleted cells and quantification of protein levels normalized to β‐actin expression.
  • B
    Western blot analysis of PPARγ protein levels in control and TRAF4‐overexpressing cells and quantification of protein levels normalized to β‐actin expression.
  • C, D
    Western blot analysis of PPARγ protein levels in control and TRAF4‐overexpressing cells in the absence or presence of MG132 or 3MA and quantification of protein levels normalized to β‐actin expression.
  • E
    Ubiquitination of endogenous PPARγ in control and TRAF4‐overexpressing cells was differentiated by treatment with DMI. Cells were treated with MG132 for 6 h.

Data information: Data are shown as the means ± s.d. from three independent experiments and analyzed by t‐test or one‐way ANOVA with Tukey’s test. **P < 0.01, and ***P < 0.001 compared to control group.

Discussion

The development process of obesity is characterized by the growth of the size and numbers of adipocytes. A large amount of clinical and animal model‐based studies have investigated the anti‐obesity effects of curcumin (Zhao et al, 2017; Miyazawa et al, 2018; Akbari et al, 2019). It has been well documented that HFD‐induced weight gain was suppressed by curcumin administration. For example, curcumin improved body composition parameters, reducing body weight and modulating the leptin levels in clinical studies (Akbari et al, 2019). Wu et al (2019a) found that curcumin can induce preadipocyte apoptosis and suppress adipocyte differentiation, leading to inhibition of adipogenesis, and other groups indicated that curcumin inhibits adipogenesis through a suppressive effect the PPARγ‐C/EBPα pathway by BBP‐elicited activation or AMPK activation (Sakuma et al, 2017; Lee et al, 2020). However, the molecular mechanisms regarding the action of curcumin in repressing adipogenesis thereby preventing obesity are largely unknown. Interestingly, our study demonstrated that curcumin increased the m6A RNA levels in iWAT from HFD+CUR group mice and SVF/3T3‐L1 cells, which suggested that the inhibitory effect of curcumin on adipogenesis might rely on m6A‐modification. Mechanistically, ALKBH5 targets TRAF4 and increases its m6A level at 3’UTR of TRAF4 transcript. We confirmed that ALKBH5 was the upstream of TRAF4, and ALKBH5 regulates the expression of TRAF4 through m6A‐dependent mechanism.

Accumulating evidence demonstrates that m6A modification is dramatically involved in the pathogenesis of metabolic diseases including obesity (Song et al, 2020). Actually, our previous work showed that FTO, the first identified m6A demethylase (Jia et al, 2011), is strongly associated with obesity and metabolic disorders (Church et al, 2009; Fischer et al, 2009). In addition, METTL3 has been successively reported to impact the adipogenesis of porcine BMSC (Yao et al, 2019). ALKBH5 is a nucleic acid oxygenase which can catalyze the demethylation of m6A‐labeled RNA (Zheng et al, 2013). Previous studies mainly focus on the effects of ALKBH5 on various cancer like glioblastoma (Malacrida et al, 2020), pancreatic cancer (Tang et al, 2020), lung cancer (Zhu et al, 2020), and ovarian cancer (Zhu et al, 2019). ALKBH5 regulates different kinds of biological processes, such as invasion, metastasis, proliferation, and migration, thereby playing a crucial role in various cancers (Wang et al, 2020a), whereas, so far, the role of ALKBH5 in obesity and adipocyte has not been reported. The current study, to the best of our knowledge, is the first demonstration showing that ALKBH5 acts as a mediator of curcumin‐inhibited adipogenesis (but not FTO). We found that the overexpression of ALKBH5 can reverse the inhibition of adipogenesis and triglyceride accumulation in curcumin‐treated cells.

It is noteworthy that downstream targets of ALKBH5 have been frequently explored. Recently, several studies have revealed that the underlying regulatory mechanisms of ALKBH5 that relies on m6A‐dependent modification are implicated with autophagy and cancer. Since ubiquitination plays crucial roles in the onset and progression of cancer and metabolic syndromes (Popovic et al, 2014), we speculate that there might exist a correlation between ALKBH5 and ubiquitination in the regulation of physiological processes. In this regard, our current work links with reminiscent of a recent study that showed the TRAF4 expression levels were positively correlated with ALKBH5 in MSC (Cen et al, 2020). On the contrary, our result showed that the TRAF4 negatively correlated with ALKBH5 expression in 3T3‐L1 cells. This “discrepancy” might be attributed to the cell type specificity.

m6A modification exerts its functions on mRNAs mainly by recruiting proteins [18]. YTHDF is one of the first identified reader protein family. YTHDF1 was reported to promote translational efficiency of m6A‐modified mRNAs [19]. The positive correlation between m6A methylation and protein expression of TRAF4 indicates that YTHDF1 may involve in regulating TRAF4 expression. In our study, forced expression of YTHDF1 in 3T3‐L1 cells increased the TRAF4 mRNA level in the polysome portion, and RIP‐qPCR assay confirmed that YTHDF1 promoted the translation of TRAF4. Our findings supported the notion that m6A accelerates TRAF4 protein expression in a YTHDF1‐dependent manner.

TRAF4, an adapter protein, links members of the TNFR family to various signaling pathways. We found that knockdown of TRAF4 promoted adipogenesis and triglyceride accumulation, implying that TRAF4 involved in adipogenesis (Fig 3E–G). Previous studies indicated that the “E3 ubiquitin ligase” characteristics of TRAF4 are vital for its biological function (Kedinger & Rio, 2007). In this study, we found that upregulation of TRAF4 reduces PPARγ protein levels through polyubiquitination of PPARγ, thereby blocking adipocyte differentiation, suggesting that the increase of ubiquitinated PPARγ in differentiated 3T3‐L1 cells was directly regulated by the stably expressed TRAF4. Constantly with our results, previous studies showed that several E3s were also been found to target PPARγ for proteasomal degradation, such as TRIM23, NEDD4, and FBXO4. These E3 ligases target PPARγ for proteasomal degradation, thereby regulating adipogenesis (Watanabe et al, 2015; Li et al, 2016; Peng et al, 2018). In our study, we also found that mRNA level of PPARγ was reduced in cells treated with curcumin (Fig 2B and D). Since TRAF4 cannot influence PPARγ transcript level, the reduced PPARγ mRNA expression may due to other reasons. A recent study showed that curcumin downregulated PPARγ mRNA expression by suppressing Kruppel‐like factor 15 (KLF15), which could bind to PPARγ promoter, resulting in the reduced mRNA level of PPARγ (Wang et al, 2019). Overall, in the current study, TRAF4 played important roles in regulating PPARγ expression. Nevertheless, other signaling pathways could not be ruled out in this process, which should be elucidated further. We discover for the first time that TRAF4 is a direct regulator of PPARγ, catalyzing polyubiquitination and subsequent degradation of PPARγ. Thus, PPARγ is a ubiquitination substrate of TRAF4 that mediates the effect of TRAF4 on adipogenesis.

Conclusion

In conclusion, we show that curcumin, the major polyphenol found in turmeric, effectively inhibits adipogenesis through the regulation of ubiquitination in ALKBH5‐m6A‐YTHDF1 orchestrated manner. Upregulation of TRAF4 targets PPARγ for degradation or instability, which in turn inhibits adipogenesis and thereby preventing obesity (Fig 6). This study is the first to demonstrate that the m6A methylation plays a crucial role in curcumin‐induced suppression of adipogenesis. Our work proposes that curcumin could act as a good potent model in basic research regarding adipogenesis and a candidate for the use in dietary supplements to prevent obesity.

Figure 6. Working model of the mechanism of curcumin regulates adipogenesis in m6A‐dependent manner.

Figure 6

Open in a new tab

In this model, curcumin inhibits adipogenesis by reducing the expression of ALKHB5, an m6A demethylase, which leads to higher m6A‐modified TRAF4 mRNA. TRAF4 mRNA with higher m6A level is recognized and bound by YTHDF1, leading to enhanced translation of TRAF4. TRAF4, acting as an E3 RING ubiquitin ligase, promotes degradation of adipocyte differentiation regulator PPARγ by a ubiquitin–proteasome pathway thereby inhibiting adipogenesis.

Materials and Methods

Animals and diets

Four‐week‐old C57BL/6 male mice were purchased from SLAC Laboratory Animal Co. Ltd (SLAC, China) and housed in accordance with the Committee on Animal Care and Use and Committee on the Ethics of Animal Experiments of Zhejiang University approval (Hangzhou, China). Mice were kept in individual cages at 12‐h/12‐h light/dark cycles, 25 ± 2°C room temperature and 30–70% humidity with ad libitum access to diet and tap water. After the acclimation, the mice were randomly categorized into three groups consisting of 8 animals each and fed for 12 weeks as follows: normal chow diet (NCD, 10 kcal% control diet; Jiangsu Xietong Pharmaceutical Bioengineering, China, XTCON50J), high‐fat diet‐induced obese mice (HFD, 60 kcal% high‐fat diet; Jiangsu Xietong Pharmaceutical Bioengineering, China, XTHF60), and HFD with curcumin (Aladdin, China) (HFD+CUR; 100 mg/kg/day by oral gavage). The food intake and body weight recorded weekly. At the end of experiments, blood samples were collected and tissues were flash‐frozen in liquid nitrogen and stored at −80°C. For plasma collection, the blood samples were centrifuged at 3,000 rpm at 4°C for 10 min and collected the supernatant and stored at −80°C.

Glucose and insulin tolerance tests

Before the tests, animals were fasted for 8 h. l glucose tolerance test (GTT) was conducted during week 11 on the diet. The mice were challenged with 2 g/kg body weight d‐glucose (Sigma‐Aldrich, USA). Insulin tolerance test (ITT) was conducted during week 12 on the diet. For insulin tolerance test, mice were injected intraperitoneally with 0.75 U/kg body weight insulin (Sigma‐Aldrich, USA). For both tests, blood samples were taken from the tail vein and glucose levels were measured at indicated time points after administration using an AlphaTRAK glucometer (NanJing Yuyue Electronics, China).

Hematoxylin and eosin staining

The samples of iWAT and eWAT and liver tissue were harvested and fixed in 10% formalin. Fixed tissues were embedded in paraffin, sectioned, and stained with H&E, and observed by light microscopy. Relative adipocyte size was quantified using the ImageJ plug‐in Adiposoft (Galarraga et al, 2012).

Cell isolation, cloning, and transduction

Stromal vascular cell fraction (SVF) was isolated from minced inguinal adipose tissue of male C57BL/6J mice (3–4 weeks old) and digested with 0.1% type I collagenase (Sigma‐Aldrich, USA) at 37°C for 45 min. After filtration through 200 μm nylon mesh and centrifugation for 5 min at 1,200 rpm to remove floating adipocytes, cells were plated in DMEM supplemented with 12% fetal bovine serum and incubated at 37°C in a 5% CO2 incubator. Once cells reached confluence, precursor adipocytes were differentiated as described for the 3T3‐L1 cells.

Cell culture, differentiation

SVF and 3T3‐L1 preadipocytes (Zenbio, USA) were cultured in DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin–streptomycin, at 37°C, in 5% CO2 atmosphere. After 2 days post‐confluence of cells, differentiation of 3T3‐L1 cells was induced in medium containing 1 µmol/l dexamethasone (Sigma‐Aldrich, USA), 500 µmol/l IBMX (Sigma‐Aldrich, USA), and 1 μg/ml insulin (Sigma‐Aldrich, USA). After 48 h, medium was replaced by maintenance medium containing 1 μg/ml insulin. For white adipogenesis induction of SVF, cells were cultured in differentiation medium containing 1 µmol/ml dexamethasone, 500 µmol/l IBMX, 1 µg/ml insulin, and 0.5 µM rosiglitazone for 48 h, and the maintenance medium containing 1 µg/ml of insulin and 0.5 µM rosiglitazone.

Cell transfection, plasmids, and RNA knockdown

To suppress gene expression, cell transfection with siRNA oligonucleotides (20 nM) specific for ALKBH5 and TRAF4 was performed with Lipofectamine RNAiMAX (Invitrogen, USA). The transfection was repeated next day. The sequences for negative control siRNA are as follows (5′ to 3′): 5′‐UUCUCCGAACGUGUCACGUTT‐3′. ALKBH5 siRNA and TRAF4 siRNA were ordered from GenePharma, and the sequences are described in previous studies (Singh et al, 2018; Guo et al, 2020). The wild‐type ALKBH5 (ALKBH5‐WT) was generated by cloning the full‐length ORF of mouse Alkbh5 gene (NM_017758.4) and then cloned into the pLVX‐puro lentiviral vector. The mutant ALKBH5 H204A (ALKBH5‐MUT) was amplified by PCR and then cloned into the pLVX lentiviral expression vector. The mouse TRAF4 expression plasmid was generated by cloning the full‐length ORF of mouse TRAF4 gene (NM_009423.4) and then cloned into a pCDH expression vector. The FLAG‐YTHDF1 expression plasmid was cloned into pcDNA3.1 mammalian expression vector.

Oil Red O Staining (ORO)

The lipid content of mature adipocytes was evaluated by Oil Red O staining. Briefly, mature adipocytes were fixed with 10% formalin at 4°C for 2 h and stained with a filtered Oil Red O (Sigma‐Aldrich, USA), eluted the stained lipid droplets using 100% isopropanol, and measured the absorbance of the extract at 500 nm to quantify lipid content.

Western blot analysis

Cells and tissues were lysed in a mixture of RIPA buffer (Beyotime Biotechnology, China) with protease and phosphatase inhibitors (Beyotime Biotechnology, China). Lysates were collected and immediately centrifuged at 12,000 rpm for 5 min, and then electrophoresed in SDS–polyacrylamide gels and transferred to PVDF membrane (Millipore, USA). The primary antibodies used for Western blot were as follows: ALKBH5 (1:2,500, 16837‐1‐AP, Proteintech, USA), PPARγ (1:2,000, 16,643–1‐AP, Proteintech, USA), C/EBPα (1:1,000, 18,311–1‐AP, Proteintech, USA), YTHDF1 (1:2,000, 17479‐1‐AP, Proteintech, USA), β‐Actin (1:2,000, ab8227, Abcam, USA), TRAF4 (1:2,000, ER65206, Huabio, China), and Ub (1:1,000, sc‐8017, Santa Cruz Biotechnology, USA). The secondary antibodies were as follows: goat anti‐rabbit IgG‐HRP (1:5,000, HA10001, Huabio, China) and goat anti‐mouse IgG‐HRP (1:2,500, HA1006, Huabio, China).

Quantitative real‐time PCR (qPCR)

Quantitative real‐time PCRs were performed on the ABI Step‐One Plus TM Real‐Time PCR System (Applied Biosystems, USA) using SYBR Master Mix (Roche, USA). The relative mRNA levels of genes were determined by the ΔΔ‐Ct method normalized to ACTB levels. The list of primers used in this study is shown in Table 1.

Table 1.

Sequences of qPCR primers

Gene Forward primer (5’‐3’) Reverse primer (5’‐3’)
qPCR primers
ACTB GACGGCCAGGTCATCACTATTG AGGAAGGCTGGAAAAGAGCC
METTL3 AGTGGCTTTTCATCTTGGCTCTA GCTGTTTCTTATGGGCCTGGA
METTL14 CGGCTTTACTCCTCGGTAGC TCACCCACCCTGAGCAAAAG
ALKBH5 GCTGTGGTGAGAGAAAGCCT AGTGGGCAAACACAAGTCCA
FTO GAGCAGCCTACAACGTGACT GAAGCTGGACTCGTCCTCAC
TRAF4 CCCGGCTTCGACTACAAGTT TCCTTCACTGAGAAACTCCTGC
PPARγ GACCACTCGCATTCCTT CCACAGACTCGGCACTC
C/EBPα GGTTTCGGGTCGCTGGATCTCTAG ACGGCCTGACTCCCTCATCTTAGAC
FABP4 GACGACAGGAAGGTGAAGAG ACATTCCACCACCAGCTTGT
Open in a new tab

Quantification of mRNA m6A

Total RNA was isolated from subcutaneous inguinal WAT (iWAT) in mice and 3T3‐L1 cells using RNAiso Plus (TAKARA, Japan). mRNA was then purified using a Dynabeads mRNA DIRECT kit (Invitrogen, USA) following the manufacturer’s protocols. Quantification of m6A in mRNA was performed as described in our previous study (Wang et al, 2020b). Briefly, 250 ng of mRNA was digested by nuclease P1 (2 U) at 42°C for 2 h, followed by the alkaline phosphatase (0.5 U) with incubation at 37°C for 2 h. The total amount of m6A in RNA was measured using LC‐MS/MS as previously reported. The ratio of m6A modification in adenine was calculated based on the determined concentrations.

Methylated RNA immunoprecipitation‐qPCR (MeRIP‐qPCR) analysis

MeRIP‐qPCR analysis was performed according to previous reports (Dominissini et al, 2013). 3T3‐L1 cells with stable overexpression of ALKBH5, ALKBH5‐MUT, and control cells transfected with an empty vector were collected. Briefly, mRNA was fragmented using RNA Fragmentation reagent (Invitrogen, USA) and immunoprecipitated with anti‐m6A antibody coupled to Dynabeads (Invitrogen, USA) in immunoprecipitation buffer, followed by elution with free m6A (Sigma‐Aldrich). m6A enrichment was determined by qPCR analysis. The specific primers used were designed according to a m6A site predictor SRAMP ((3’UTR forward: AGGGAGGCGGGACTGG; reverse: TGGTGGTGGTGGTGGTGG) and (CDS forward: AAATGCCCTGAGGACCAA; reverse: GGATAGCCAAGCCTAACACC)).

RNA immunoprecipitation‐qPCR (RIP‐qPCR) analysis

This procedure was performed according to previous reports (Peritz et al, 2006). FLAG‐YTHDF1 overexpressed cells were lysed in lysis buffer, and anti‐FLAG M2 magnetic beads (Sigma‐Aldrich, USA) were added into lysates and incubated for 8 h at 4°C. Then, the beads were eluted in wash buffer containing 0.1% SDS and proteinase K (Invitrogen, USA), followed by qPCR to detected fold enrichment.

Dual‐luciferase reporter and mutagenesis assays

This procedure was performed according to previous reports (Wang et al, 2020b). For dual‐luciferase reporter assay, cells were co‐transfected with wild‐type or mutant TRAF4‐3ʹUTR and ALKBH5‐WT (or ALKBH5‐MUT, or YTHDF1, or empty vector). 48 h after transfection, the activities of firefly luciferase and Renilla luciferase were measured with the Dual‐Luciferase® Reporter Assay System (Beyotime Biotechnology, China).

TRAF4‐3ʹUTR with wild‐type m6A sites:

  • AGGGAGGCGGGACTGGATCTGAAACTGGATTGGGGGCCGGACCTCCTTTCAGCTTCTTTTGCCTAGGCTGTTAGCCCACTCTCCCTACCACCACCACCACCACCACCACCA

TRAF4‐3ʹUTR with mutant m6A sites:

  • AGGGAGGCGGGTCTGGATCTGAAACTGGATTGGGGGCCGGACCTCCTTTCAGCTTCTTTTGCCTAGGCTGTTAGCCCACTCTCCCTACCACCACCACCACCACCACCACCA

Polysome profiling

This procedure was performed according to previous reports (Faye et al, 2014). Control and YTHDF1‐overexpressing 3T3‐L1 cells were treated with 100 μg/ml cycloheximide in culture media for 10 min at 37°C and lysed on ice. The lysates were carefully transferred onto a 10–50% (w/v) sucrose gradients and centrifuged at 39,000 rpm for 2 h at 4°C. Gradients were then fractionated from the bottom, and approximately thirty 0.3 ml samples were collected. The A260 of samples was detected by NanoDrop to determine the 40‐80S or polysome fraction, and RNA was purified from fractions and subjected to qPCR analysis.

Statistical analysis

All statistical analyses were performed using the SPSS version 22 software (SPSS Inc., Chicago, IL, USA). Differences between treatments were analyzed by t‐test or one‐way ANOVA after testing for homogeneity of variances with Levene's test. For one‐way ANOVA, post hoc comparisons were conducted using Tukey's HSD test. We presented all data as means ± SD and regarded P < 0.05 as statistically significant.

Author contributions

YC, and XW developed the concept for this work. YC, RW, WC, YL, XL, BZ, GG, FL, and YX performed experimental work. YC, XW, and YW wrote and revised the manuscript.

Conflict of interest

The authors have declared that they have no conflict of interest.

Supporting information

Expanded View Figures PDF

Click here for additional data file. (307.1KB, pdf)

Review Process File

Click here for additional data file. (521.6KB, pdf)

Acknowledgements

This work was supported by the National Key R & D Program (2018YFD0500405), the Fundamental Research Funds for ZheJiang Provincial Colleges & Universities (2019XZZX003‐13), and the Key project of Jinhua Science and Technology (No. 2019‐2‐003).

EMBO reports (2021) 22: e52146.

Data availability

Our study did not include data deposited in public repositories, all public RNA profile data (GSE69313, GSE6794) analyzed during the study could acquire from the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/).

References

  1. Ahn J, Lee H, Kim S, Ha T (2010) Curcumin‐induced suppression of adipogenic differentiation is accompanied by activation of Wnt/beta‐catenin signaling. Am J Physiol‐Cell Ph 298: C1510–C1516 [DOI] [PubMed] [Google Scholar]
  2. Akbari M, Lankarani KB, Tabrizi R, Ghayour‐Mobarhan M, Peymani P, Ferns G, Ghaderi A, Asemi Z (2019) The effects of curcumin on weight loss among patients with metabolic syndrome and related disorders: a systematic review and meta‐analysis of randomized controlled trials. Front Pharmacol 10: 649 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cen SZ, Li JT, Cai ZP, Pan YQ, Sun ZH, Li ZF, Ye GW, Zheng G, Li M, Liu WJ et al (2020) TRAF4 acts as a fate checkpoint to regulate the adipogenic differentiation of MSCs by activating PKM2. Ebiomedicine 54: 102722 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Church C, Lee S, Bagg EA, McTaggart JS, Deacon R, Gerken T, Lee A, Moir L, Mecinovic J, Quwailid MM et al (2009) A mouse model for the metabolic effects of the human fat mass and obesity associated FTO gene. PLoS Genet 5: e1000599 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dominissini D, Moshitch‐Moshkovitz S, Salmon‐Divon M, Amariglio N, Rechavi G (2013) Transcriptome‐wide mapping of N(6)‐methyladenosine by m(6)A‐seq based on immunocapturing and massively parallel sequencing. Nat Protoc 8: 176–189 [DOI] [PubMed] [Google Scholar]
  6. Ejaz A, Wu D, Kwan P, Meydani M (2009) Curcumin inhibits adipogenesis in 3T3‐L1 adipocytes and angiogenesis and obesity in C57/BL mice. J Nutr 139: 919–925 [DOI] [PubMed] [Google Scholar]
  7. El‐Moselhy MA, Taye A, Sharkawi SS, El‐Sisi SF, Ahmed AF (2011) The antihyperglycemic effect of curcumin in high fat diet fed rats. Role of TNF‐alpha and free fatty acids. Food Chem Toxicol 49: 1129–1140 [DOI] [PubMed] [Google Scholar]
  8. Faye MD, Graber TE, Holcik M (2014) Assessment of selective mRNA translation in mammalian cells by polysome profiling. J Vis Exp 92: e52295 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fischer J, Koch L, Emmerling C, Vierkotten J, Peters T, Bruning JC, Ruther U (2009) Inactivation of the Fto gene protects from obesity. Nature 458: 894–898 [DOI] [PubMed] [Google Scholar]
  10. Galarraga M, Campion J, Munoz‐Barrutia A, Boque N, Moreno H, Martinez JA, Milagro F, Ortiz‐de‐Solorzano C (2012) Adiposoft: automated software for the analysis of white adipose tissue cellularity in histological sections. J Lipid Res 53: 2791–2796 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gan Z, Wei W, Wu J, Zhao Y, Zhang L, Wang T, Zhong X (2019) Resveratrol and curcumin improve intestinal mucosal integrity and decrease m(6)A RNA methylation in the intestine of weaning piglets. ACS Omega 4: 17438–17446 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ghorbani Z, Hekmatdoost A, Mirmiran P (2014) Anti‐hyperglycemic and insulin sensitizer effects of turmeric and its principle constituent curcumin. Int J Endocrinol Metab 12: e18081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ghosh S, Banerjee S, Sil PC (2015) The beneficial role of curcumin on inflammation, diabetes and neurodegenerative disease: a recent update. Food Chem Toxicol 83: 111–124 [DOI] [PubMed] [Google Scholar]
  14. Guo X, Li K, Jiang W, Hu Y, Xiao W, Huang Y, Feng Y, Pan Q, Wan R (2020) RNA demethylase ALKBH5 prevents pancreatic cancer progression by posttranscriptional activation of PER1 in an m6A‐YTHDF2‐dependent manner. Mol Cancer 19: 91 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hanai H, Sugimoto K (2009) Curcumin has bright prospects for the treatment of inflammatory bowel disease. Curr Pharm Design 15: 2087–2094 [DOI] [PubMed] [Google Scholar]
  16. Ikram M, Saeed K, Khan A, Muhammad T, Khan MS, Jo MG, Rehman SU, Kim MO (2019) Natural dietary supplementation of curcumin protects mice brains against ethanol‐induced oxidative stress‐mediated neurodegeneration and memory impairment via Nrf2/TLR4/RAGE signaling. Nutrients 11: 1082 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jia GF, Fu Y, Zhao X, Dai Q, Zheng GQ, Yang Y, Yi CQ, Lindahl T, Pan T, Yang YG et al (2011) N6‐Methyladenosine in nuclear RNA is a major substrate of the obesity‐associated FTO. Nat Chem Biol 7: 885–887 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jung CH, Lee WJ, Song KH (2017) Metabolically healthy obesity: a friend or foe? Korean J Intern Med 32: 611–621 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kedinger V, Rio MC (2007) TRAF4, the unique family member. Adv Exp Med Biol 597: 60–71 [DOI] [PubMed] [Google Scholar]
  20. Kim CY, Le TT, Chen CY, Cheng JX, Kim KH (2011) Curcumin inhibits adipocyte differentiation through modulation of mitotic clonal expansion. J Nutr Biochem 22: 910–920 [DOI] [PubMed] [Google Scholar]
  21. Lee ES, Kwon MH, Kim HM, Woo HB, Ahn CM, Chung CH (2020) Curcumin analog CUR5‐8 ameliorates nonalcoholic fatty liver disease in mice with high‐fat diet‐induced obesity. Metabolism 103: 5–8 [DOI] [PubMed] [Google Scholar]
  22. Li JJ, Wang R, Lama R, Wang X, Floyd ZE, Park EA, Liao FF (2016) Ubiquitin ligase NEDD4 regulates PPARgamma stability and adipocyte differentiation in 3T3‐L1 cells. Sci Rep 6: 38550 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lu N, Li X, Yu J, Li Y, Wang C, Zhang L, Wang T, Zhong X (2018) Curcumin attenuates lipopolysaccharide‐induced hepatic lipid metabolism disorder by modification of m(6) A RNA methylation in piglets. Lipids 53: 53–63 [DOI] [PubMed] [Google Scholar]
  24. Malacrida A, Rivara M, Di Domizio A, Cislaghi G, Miloso M, Zuliani V, Nicolini G (2020) 3D proteome‐wide scale screening and activity evaluation of a new ALKBH5 inhibitor in U87 glioblastoma cell line. Bioorg Med Chem 28: 115300 [DOI] [PubMed] [Google Scholar]
  25. Mirzaei H, Masoudifar A, Sahebkar A, Zare N, Sadri Nahand J, Rashidi B, Mehrabian E, Mohammadi M, Mirzaei HR, Jaafari MR (2018) MicroRNA: a novel target of curcumin in cancer therapy. J Cell Physiol 233: 3004–3015 [DOI] [PubMed] [Google Scholar]
  26. Miyazawa T, Nakagawa K, Kim SH, Thomas MJ, Paul L, Zingg JM, Dolnikowski GG, Roberts SB, Kimura F, Miyazawa T et al (2018) Curcumin and piperine supplementation of obese mice under caloric restriction modulates body fat and interleukin‐1 beta. Nutr Metab 15: 12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Peng JT, Li YT, Wang XH, Deng S, Holland J, Yates E, Chen J, Gu HT, Essandoh K, Mu XJ et al (2018) An Hsp20‐FBXO4 axis regulates adipocyte function through modulating PPAR gamma ubiquitination. Cell Rep 23: 3607–3620 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Peritz T, Zeng F, Kannanayakal TJ, Kilk K, Eiriksdottir E, Langel U, Eberwine J (2006) Immunoprecipitation of mRNA‐protein complexes. Nat Protoc 1: 577–580 [DOI] [PubMed] [Google Scholar]
  29. Popovic D, Vucic D, Dikic I (2014) Ubiquitination in disease pathogenesis and treatment. Nat Med 20: 1242–1253 [DOI] [PubMed] [Google Scholar]
  30. Sakuma S, Sumida M, Endoh Y, Kurita A, Yamaguchi A, Watanabe T, Kohda T, Tsukiyama Y, Fujimoto Y (2017) Curcumin inhibits adipogenesis induced by benzyl butyl phthalate in 3T3‐L1 cells. Toxicol Appl Pharm 329: 158–164 [DOI] [PubMed] [Google Scholar]
  31. Shao W, Yu Z, Chiang Y, Yang Y, Chai T, Foltz W, Lu H, Fantus IG, Jin T (2012) Curcumin prevents high fat diet induced insulin resistance and obesity via attenuating lipogenesis in liver and inflammatory pathway in adipocytes. PLoS ONE 7: e28784 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Singh R, Karri D, Shen H, Shao J, Dasgupta S, Huang S, Edwards DP, Ittmann MM, O'Malley BW, Yi P (2018) TRAF4‐mediated ubiquitination of NGF receptor TrkA regulates prostate cancer metastasis. J Clin Invest 128: 3129–3143 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Song TX, Yang Y, Jiang SW, Peng J (2020) Novel insights into adipogenesis from the perspective of transcriptional and RNA N6‐methyladenosine‐mediated post‐transcriptional regulation. Adv Sci 7: 2001563 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Tang B, Yang Y, Kang M, Wang Y, Wang Y, Bi Y, He S, Shimamoto F (2020) m(6)A demethylase ALKBH5 inhibits pancreatic cancer tumorigenesis by decreasing WIF‐1 RNA methylation and mediating Wnt signaling. Mol Cancer 19: 3 [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  35. Tian L, Song Z, Shao W, Du WW, Zhao LR, Zeng K, Yang BB, Jin T (2017) Curcumin represses mouse 3T3‐L1 cell adipogenic differentiation via inhibiting miR‐17‐5p and stimulating the Wnt signalling pathway effector Tcf7l2. Cell Death Dis 8: e2559 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tsai AG, Williamson DF, Glick HA (2011) Direct medical cost of overweight and obesity in the USA: a quantitative systematic review. Obes Rev 12: 50–61 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wafi AM, Hong J, Rudebush TL, Yu L, Hackfort B, Wang H, Schultz HD, Zucker IH, Gao L (2019) Curcumin improves exercise performance of mice with coronary artery ligation‐induced HFrEF: Nrf2 and antioxidant mechanisms in skeletal muscle. J Appl Physiol 126: 477–486 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wang X, Zhao BS, Roundtree IA, Lu Z, Han D, Ma H, Weng X, Chen K, Shi H, He C (2015a) N(6)‐methyladenosine modulates messenger RNA translation efficiency. Cell 161: 1388–1399 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wang X, Zhu L, Chen J, Wang Y (2015b) mRNA m(6)A methylation downregulates adipogenesis in porcine adipocytes. Biochem Biophys Res Comm 459: 201–207 [DOI] [PubMed] [Google Scholar]
  40. Wang T, Yan R, Xu X, Li X, Cao L, Gao L, Liu J, Zhou X, Yu H, Wang X et al (2019) Curcumin represses adipogenic differentiation of human bone marrow mesenchymal stem cells via inhibiting kruppel‐like factor 15 expression. Acta Histochem 121: 253–259 [DOI] [PubMed] [Google Scholar]
  41. Wang J, Wang J, Gu Q, Ma Y, Yang Y, Zhu J, Zhang Q (2020a) The biological function of m6A demethylase ALKBH5 and its role in human disease. Cancer Cell Int 20: 347 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wang XX, Wu RF, Liu YH, Zhao YL, Bi Z, Yao YX, Liu Q, Shi HL, Wang FQ, Wang YZ (2020b) m(6)A mRNA methylation controls autophagy and adipogenesis by targeting Atg5 and Atg7. Autophagy 16: 1221–1235 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Watanabe M, Takahashi H, Saeki Y, Ozaki T, Itoh S, Suzuki M, Mizushima W, Tanaka K, Hatakeyama S (2015) The E3 ubiquitin ligase TRIM23 regulates adipocyte differentiation via stabilization of the adipogenic activator PPARgamma. eLife 4: e05615 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wongcharoen W, Phrommintikul A (2009) The protective role of curcumin in cardiovascular diseases. Int J Cardiol 133: 145–151 [DOI] [PubMed] [Google Scholar]
  45. Wu R, Liu Y, Yao Y, Zhao Y, Bi Z, Jiang Q, Liu Q, Cai M, Wang F, Wang Y et al (2018a) FTO regulates adipogenesis by controlling cell cycle progression via m(6)A‐YTHDF2 dependent mechanism. Biochim Biophys Acta Mol Cell Biol Lipids 1863: 1323–1330 [DOI] [PubMed] [Google Scholar]
  46. Wu R, Yao Y, Jiang Q, Cai M, Liu Q, Wang Y, Wang X (2018b) Epigallocatechin gallate targets FTO and inhibits adipogenesis in an mRNA m(6)A‐YTHDF2‐dependent manner. Int J Obes (Lond) 42: 1378–1388 [DOI] [PubMed] [Google Scholar]
  47. Wu LY, Chen CW, Chen LK, Chou HY, Chang CL, Juan CC (2019a) Curcumin attenuates adipogenesis by inducing preadipocyte apoptosis and inhibiting adipocyte differentiation. Nutrients 11: 2307 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wu RF, Liu YH, Zhao YL, Bi Z, Yao YX, Liu Q, Wang FQ, Wang YZ, Wang XX (2019b) m(6)A methylation controls pluripotency of porcine induced pluripotent stem cells by targeting SOCS3/JAK2/STAT3 pathway in a YTHDF1/YTHDF2‐orchestrated manner. Cell Death Dis 10: 171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yan C, Zhang Y, Zhang X, Aa J, Wang G, Xie Y (2018) Curcumin regulates endogenous and exogenous metabolism via Nrf2‐FXR‐LXR pathway in NAFLD mice. Biomed Pharmacother 105: 274–281 [DOI] [PubMed] [Google Scholar]
  50. Yao Y, Bi Z, Wu R, Zhao Y, Liu Y, Liu Q, Wang Y, Wang X (2019) METTL3 inhibits BMSC adipogenic differentiation by targeting the JAK1/STAT5/C/EBPbeta pathway via an m(6)A‐YTHDF2‐dependent manner. FASEB J 33: 7529–7544 [DOI] [PubMed] [Google Scholar]
  51. Zhang S, Zhao BS, Zhou A, Lin K, Zheng S, Lu Z, Chen Y, Sulman EP, Xie K, Bogler O et al (2017) m(6)A Demethylase ALKBH5 maintains tumorigenicity of glioblastoma stem‐like cells by sustaining FOXM1 expression and cell proliferation program. Cancer Cell 31: 591–606 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhao X, Yang Y, Sun BF, Shi Y, Yang X, Xiao W, Hao YJ, Ping XL, Chen YS, Wang WJ et al (2014) FTO‐dependent demethylation of N6‐methyladenosine regulates mRNA splicing and is required for adipogenesis. Cell Res 24: 1403–1419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zhao Y, Chen B, Shen J, Wan L, Zhu Y, Yi T, Xiao Z (2017) The beneficial effects of quercetin, curcumin, and resveratrol in obesity. Oxid Med Cell Longev 2017: 1–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zheng GQ, Dahl JA, Niu YM, Fedorcsak P, Huang CM, Li CJ, Vagbo CB, Shi Y, Wang WL, Song SH et al (2013) ALKBH5 is a mammalian RNA Demethylase that impacts RNA metabolism and mouse fertility. Mol Cell 49: 18–29 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhu HT, Gan XL, Jiang XW, Diao S, Wu H, Hu JG (2019) ALKBH5 inhibited autophagy of epithelial ovarian cancer through miR‐7 and BCL‐2. J Exp Clin Canc Res 38: 163 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhu Z, Qian Q, Zhao X, Ma L, Chen P (2020) N(6)‐methyladenosine ALKBH5 promotes non‐small cell lung cancer progress by regulating TIMP3 stability. Gene 731: 144348 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Expanded View Figures PDF

Click here for additional data file. (307.1KB, pdf)

Review Process File

Click here for additional data file. (521.6KB, pdf)

Data Availability Statement

Our study did not include data deposited in public repositories, all public RNA profile data (GSE69313, GSE6794) analyzed during the study could acquire from the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/).

Articles from EMBO Reports are provided here courtesy of Nature Publishing Group

RESOURCES