Isoliquiritigenin Enhances the Beige Adipocyte Potential of Adipose-Derived Stem Cells by JNK Inhibition

Human adipose-derived stem cells (hASCs) can be isolated from fat tissue and have attracted interest for their potential therapeutic applications in metabolic disease. hASCs can be induced to undergo adipogenic differentiation in vitro by exposure to chemical agents or inductive growth factors. We investigated the effects and mechanism of differentiating hASC-derived white adipocytes into functional beige and brown adipocytes with isoliquiritigenin (ILG) treatment. Here, we showed that hASC-derived white adipocytes could promote brown adipogenesis by expressing both uncoupling protein 1 (UCP1) and PR/SET Domain 16 (PRDM16) following low-dose ILG treatments. ILG treatment of white adipocytes enhanced the expression of brown fat-specific markers, while the expression levels of c-Jun N-terminal kinase (JNK) signaling pathway proteins were downregulated. Furthermore, we showed that the inhibition of JNK phosphorylation contributed to white adipocyte differentiation into beige adipocytes, which was validated by the use of SP600125. We identified distinct regulatory effects of ILG dose responses and suggested that low-dose ILG induced the beige adipocyte potential of hASCs via JNK inhibition.


Introduction
Adipocytes play key roles in energy homeostasis throughout the body and can be classified as white adipocytes (WACs) and brown adipocytes (BACs). WACs are mainly lipid storage cells, whereas BACs are mitochondria-rich multilocular cells that specialize in dissipating energy by generating heat through the action of uncoupling protein 1 (UCP1) [1]. Moreover, beige adipocytes that exist in white adipose tissue (WAT) can be induced under specific conditions such as β3-adrenergic receptor signaling stimulation, exercise, cold exposure, or treatment with small molecules [2]. The beige adipocytes originate from a distinct lineage similar to that of mesenchymal stem cells, which are similar to WACs and BACs, although BACs are derived from Myf5 + precursor cells and white and beige adipocytes are derived from Myf5 − precursor cells [2,3]. However, there is also evidence that mature white adipocytes can transdifferentiate to beige adipocytes by specific factors [4,5]. There are two types of WAT: subcutaneous and visceral adipose tissues. The adipose tissue under the skin is subcutaneous , and accumulated lipids were quantified by measuring the absorbance. (E) Gene expression was analyzed by real-time PCR. Expression levels of UCP1 in low-dose ILG-treated adipocytes differentiated from ASCs. GAPDH was used as an internal control to normalize the expression of the target genes; n = 3 independent experiments; * p < 0.05 and ** p < 0.01.

The Small Molecule ILG Induces the Transdifferentiation of White Adipocytes to Beige Adipocytes
To investigate the effect of ILG on lipid metabolism, we evaluated the expression of thermogenic pathway markers. As shown in Figure 2, lipid drop quantification by ORO staining showed increased levels in WACs and BACs treated with 0.5 μM and 0.25 μM ILG, respectively. In addition, the expression of UCP1, UCP2, and UCP3 was analyzed to evaluate the activation of the heat-generating pathway and the futile cycle of proton pumping through the activation of UCPs. These results showed that ILG significantly increased the expression levels of UCP1, UCP2, and UCP3 in both WACs and BACs ( Figure 2B). , and accumulated lipids were quantified by measuring the absorbance. (E) Gene expression was analyzed by real-time PCR. Expression levels of UCP1 in low-dose ILG-treated adipocytes differentiated from ASCs. GAPDH was used as an internal control to normalize the expression of the target genes; n = 3 independent experiments; * p < 0.05 and ** p < 0.01.

The Small Molecule ILG Induces the Transdifferentiation of White Adipocytes to Beige Adipocytes
To investigate the effect of ILG on lipid metabolism, we evaluated the expression of thermogenic pathway markers. As shown in Figure 2, lipid drop quantification by ORO staining showed increased levels in WACs and BACs treated with 0.5 µM and 0.25 µM ILG, respectively. In addition, the expression of UCP1, UCP2, and UCP3 was analyzed to evaluate the activation of the heat-generating pathway and the futile cycle of proton pumping through the activation of UCPs. These results showed that ILG significantly increased the expression levels of UCP1, UCP2, and UCP3 in both WACs and BACs ( Figure 2B).     Effect of ILG treatment on the mRNA expression of browning markers in WACs and BACs. Quantitative real-time PCR analysis of the PPARGC1A, PPARG, PPARD, CEBPB, FABP4, BMP2, and PARK7 genes in cells treated with ILG for 24 h after adipogenic differentiation. GAPDH was used as an internal control to normalize the expression of the target genes; n = 3 independent experiments; * p < 0.05 and ** p < 0.01.

The Small Molecule ILG Induces Beige Adipocyte Differentiation by JNK Inhibition
We showed that hASC-derived WACs were able to transdifferentiate to a beige or brown adipocyte by expressing both UCP1 and PRDM16 in response to low-dose ILG treatments ( Figure 4A,B). Previously, UCP1 and PRDM16 were shown to be master regulators for brown adipogenesis, and their overexpression could induce browning in stem cells. Our results confirmed that the expression levels of UCP1 and PRDM16 increased in beige and brown adipocytes after ILG treatment of WACs. However, it is thought that WACs contain beige-like adipocytes, because UCP-1 is expressed in WACs even if it is a small amount. Importantly, JNK phosphorylation was downregulated during increasing beige and brown adipocyte differentiation of WACs. Consistently, immunofluorescence staining confirmed a significant increase in UCP1, while p-JNK expression was downregulated in WACs after ILG treatment ( Figure 4C,D). To explore the regulatory effect of JNK on beige and brown adipocyte differentiation, a JNK inhibitor (SP600125) was applied to WACs during adipocyte differentiation ( Figure 5A,B). Immunoblotting showed that JNK phosphorylation was reduced by the inhibitor, while expression of UCP1 and PRDM16 was dose-dependently increased by the inhibitor.

The Small Molecule ILG Induces Beige Adipocyte Differentiation by JNK Inhibition
We showed that hASC-derived WACs were able to transdifferentiate to a beige or brown adipocyte by expressing both UCP1 and PRDM16 in response to low-dose ILG treatments ( Figure  4A,B). Previously, UCP1 and PRDM16 were shown to be master regulators for brown adipogenesis, and their overexpression could induce browning in stem cells. Our results confirmed that the expression levels of UCP1 and PRDM16 increased in beige and brown adipocytes after ILG treatment of WACs. However, it is thought that WACs contain beige-like adipocytes, because UCP-1 is expressed in WACs even if it is a small amount. Importantly, JNK phosphorylation was downregulated during increasing beige and brown adipocyte differentiation of WACs. Consistently, immunofluorescence staining confirmed a significant increase in UCP1, while p-JNK expression was downregulated in WACs after ILG treatment ( Figure 4C,D). To explore the regulatory effect of JNK on beige and brown adipocyte differentiation, a JNK inhibitor (SP600125) was applied to WACs during adipocyte differentiation ( Figure 5A,B). Immunoblotting showed that JNK phosphorylation was reduced by the inhibitor, while expression of UCP1 and PRDM16 was dose-dependently increased by the inhibitor.  The expression levels of the indicated BAC-specific proteins were quantified by Western blot analysis. Expression changes of UCP-1 and JNK by ILG treatment. Immunofluorescent staining of (C) UCP1 and (D) p-JNK. Scale bar = 50 μm. The data were normalized to β-actin antibody and analyzed using ImageJ software; n = 3 independent experiments; * p < 0.05 and ** p < 0.01. The data were normalized to β-actin antibody and analyzed using ImageJ software; n = 3 independent experiments; * p < 0.05 and ** p < 0.01.

Discussion
hASCs are isolated from human lipoaspirate tissue and show functional properties that are of interest in cell therapy and regenerative medicine. Our results revealed that hASC-derived WACs were able to promote brown adipogenesis by expressing both UCP1 and PRDM16 in response to low-dose ILG treatments by inhibiting the JNK signaling pathway.
A large number of studies have reported the beneficial effects of flavonoids on human health, and the bioactive component ILG has been well documented as a potent antioxidant with anti-inflammatory, antiatherosclerotic, cardioprotective, and cancer-preventing properties, but the roles of ILG in stem cell differentiation remain unclear [24,26,34,35]. Moreover, ILG prevents insulin-induced ROS generation in 3T3-L1 cells and suppresses lipid accumulation [32]. Another report showed that ILG (10 μM) treatment had little effect on fat production by differentiated 3T3-L1 cells [36]. Researchers have suggested that ILG can inhibit adipose tissue inflammation in both inflammasome-dependent and inflammasome-independent manners and attenuate adipose tissue fibrosis by targeting innate immune sensors. Recent studies have shown that ILG not only strongly inhibits activation of the NLRP3 inflammasome, but also improves diet-induced adipose tissue inflammation [31]. ILG reduces phosphorylated JNK expression in palmitic acid-induced macrophages, and the induced inflammatory changes were suppressed by inhibiting NFκB activation [36]. In addition, ILG inhibits LXRα-dependent hepatic steatosis through JNK1 inhibition and protects hepatocytes from oxidative injury caused by fat accumulation [33]. Furthermore, other researchers have suggested that ILG induces dose-dependent developmental toxicity and oxidative The data were normalized to β-actin antibody and analyzed using ImageJ software; n = 3 independent experiments; * p < 0.05 and ** p < 0.01.

Discussion
hASCs are isolated from human lipoaspirate tissue and show functional properties that are of interest in cell therapy and regenerative medicine. Our results revealed that hASC-derived WACs were able to promote brown adipogenesis by expressing both UCP1 and PRDM16 in response to low-dose ILG treatments by inhibiting the JNK signaling pathway.
A large number of studies have reported the beneficial effects of flavonoids on human health, and the bioactive component ILG has been well documented as a potent antioxidant with anti-inflammatory, antiatherosclerotic, cardioprotective, and cancer-preventing properties, but the roles of ILG in stem cell differentiation remain unclear [24,26,34,35]. Moreover, ILG prevents insulin-induced ROS generation in 3T3-L1 cells and suppresses lipid accumulation [32]. Another report showed that ILG (10 µM) treatment had little effect on fat production by differentiated 3T3-L1 cells [36]. Researchers have suggested that ILG can inhibit adipose tissue inflammation in both inflammasome-dependent and inflammasome-independent manners and attenuate adipose tissue fibrosis by targeting innate immune sensors. Recent studies have shown that ILG not only strongly inhibits activation of the NLRP3 inflammasome, but also improves diet-induced adipose tissue inflammation [31]. ILG reduces phosphorylated JNK expression in palmitic acid-induced macrophages, and the induced inflammatory changes were suppressed by inhibiting NFκB activation [36]. In addition, ILG inhibits LXRα-dependent hepatic steatosis through JNK1 inhibition and protects hepatocytes from oxidative injury caused by fat accumulation [33]. Furthermore, other researchers have suggested that ILG induces dose-dependent developmental toxicity and oxidative stress-mediated apoptosis through the Nrf2-HO1/JNK-ERK/mitochondrial pathway in zebrafish embryos and larvae [25]. Similar to our observations, low-dose ILG upregulated the expression of thermogenesis-related genes (UCP1, PRDM16, and SIRT1) in the interscapular BAT of diet-induced obese C57BL/6J mice, increasing insulin sensitivity and energy expenditure and decreasing adiposity [29].
Here, we observed that a low concentration of ILG-induced beige and BAC differentiation with increased UCP1 expression (Figure 1) and increases the expression of UCP1 and PRDM16 and adipogenic differentiation markers in both WACs and BACs in the context of lipid metabolism (Figure 2). Flavonoids have a wide spectrum of biochemical and pharmacological effects and may show variaous responses according to their concentrations. A study has shown that chronic consumption of a blueberry extract was only related to extremely low concentration of anthocyanins in rat brains [37]. Low doses of luteolin increased p21 expression and high doses of luteolin decreased its expression in cancer cells [38]. These results suggest that flavonoid bioactivity does not follow a classical dose-dependent relation and it may have critical biological implications [39]. However, so far, there has been no report to accurately explain the opposite responses and their mechanisms according to the concentration of flavonoids such as ILG.
The JNK signaling pathway has been reported to be involved in the regulation of stem cells' osteogenic and adipogenic differentiation, and JNK activity is abnormally elevated in obese conditions [40][41][42]. Several studies have shown that JNK is activated in response to ER stress stimulation in high-fat-induced obesity, and these groups also investigated the effect of ER stress on the thermogenic capacity of mouse beige adipocytes [43][44][45][46]. Ucp1 expression is severely suppressed in inguinal WAT under ER stress conditions [43]. Whether JNK is activated after ER stress stimulation and regulates the mRNA levels of Ucp1 and Pparγ in vitro and ex vivo has been investigated [43]. However, the consequences of JNK inhibition on the beige and brown adipocyte differentiation of hASCs are unknown. Therefore, the purpose of this study was to investigate the role of JNK in regulating the differentiation of hASCs to beige and brown adipocytes after ILG treatment. Interestingly, we found that JNK was deactivated by the beige and BAC differentiation of hASCs in response to low-dose (<1 µM) ILG treatments, and the ratio of p-JNK to JNK in WACs dose-dependently declined in response to ILG treatment. We then blocked JNK pathway activation with different concentrations of SP600125 (a specific inhibitor of the JNK pathway). These results indicated that the JNK signaling pathway regulated the beige adipocyte differentiation of hASCs. Consequently, we demonstrated the distinct regulatory functions of the ILG dose responses and induced beige adipocyte potential of hASCs by JNK inhibition in response to low-dose ILG.

Cell Viability Assay
To measure the toxicity of ILG, we conducted a cell viability assay (EZ-Cytox; Dogenbio, Seoul, Korea). Cells were seeded on 96-well culture plates at 5 × 10 3 cells/well, allowed to adhere, and then treated with different concentrations of ILG (1,5,10,25, and 50 µM). After 24 h, to detect cell viability, 10 µL of EZ-Cytox solution was added to each well and incubated at 37 • C for 1 h. Optical density values at 450 nm were measured with a microplate reader (Multiskan FC, Thermo Fisher Scientific, Waltham, MA, USA).

Oil Red O Staining and Measurement of Lipid Content
The hASCs and hASC-derived adipocytes were rinsed with phosphate-buffered saline (PBS), fixed with 4% formaldehyde at room temperature for 1 h and washed again three times with distilled water. An Oil Red O (ORO) solution (Sigma, St. Louis, MO, USA) was layered onto the differentiated cells and incubated for 20 min, followed by three washes with distilled water. Stained adipocytes were observed by light microscopy. Intracellular lipid levels were quantified after extracting the ORO bound to cells with 100% isopropanol (Duksan, Ansan, Korea), and the ORO was measured at 450 nm using a microplate reader (Multiskan FC, Thermo Fisher Scientific, Waltham, MA, USA).

RNA Extraction and Quantitative Real-Time PCR
Total RNA was extracted from cells using an RNA extraction kit (iNtRON Biotechnology, Seongnam, Korea) according to the manufacturer's instructions, and 1 µg of RNA was converted to cDNA using a Maxime RT PreMix kit (iNtRON Biotechnology). To quantify the expression levels of genes, power SYBR green (SYBR Premix Ex Taq (Tli RNase H Plus, ROX Plus), Takara Bio, Foster City, CA, USA), and a StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA, USA) were used. The primer sequences used for qPCR are listed in Table 1. The transcript levels of each gene were normalized to the level of GAPDH, and the relative quantification (2 −∆∆ CT) method was used to analyze the data. Table 1. Primer sequences used for quantitative real-time PCR.

Immunofluorescence Staining
Cells in 4-well side chambers (SPL, Pocheon, Korea) were fixed with 4% formaldehyde and rinsed with PBS. Permeabilization was performed with 0.1% Triton X-100 (Sigma) for 10 min. After being incubated in blocking solution (2.5% normal horse serum (Vector Laboratories, Burlingame, CA, USA) in PBS) for 1 h, the cells were incubated with rabbit anti-UCP-1 (ThermoFisher Scientific) and p-JNK (Cell Signaling) antibodies (diluted 1:500) overnight at 4 • C. After being washed 3 times with PBS, the cells were incubated with FITC-conjugated anti-rabbit IgG (1:500, EMD Millipore) and rhodamine-conjugated anti-rabbit IgG (1:500, Vector Laboratories) secondary antibodies. DAPI (Invitrogen, Carlsbad, CA, USA) was used to stain the nuclei. Fluorescence images were captured by a confocal microscope (LSM710, Carl Zeiss Microscopy GmbH, Jena, Germany). For mitochondrial staining, MitoTracker Red (1 mM, Invitrogen) was added to the DMEM/F-12 at a concentration of 200 nM for 30 min at 37 • C. After incubation, the cells were washed with PBS, fixed with 4% formaldehyde, washed again with PBS, and immunostained.

Statistical Analysis
All experimental results were compared using one-way analysis of variance (ANOVA) by the Statistical Package of Social Science (SPSS, version 17) program. The data are expressed as the mean ± SEM. A protected least-significant difference (LSD) test, which consists of single-step procedures and one-way ANOVA to analyze multiple comparisons, was used to identify significant differences between the means. A value of p < 0.05 was considered statistically significant, and statistical significance was shown as * p < 0.05 and ** p < 0.01.