CHIP Haploinsufficiency Exacerbates Hepatic Steatosis via Enhanced TXNIP Expression and Endoplasmic Reticulum Stress Responses

TXNIP is a critical regulator of glucose homeostasis, fatty acid synthesis, and cholesterol accumulation in the liver, and it has been reported that metabolic diseases, such as obesity, atherosclerosis, hyperlipidemia, type 2 diabetes, and nonalcoholic fatty liver disease (NAFLD), are associated with endoplasmic reticulum (ER) stress. Because CHIP, an E3 ligase, was known to be involved in regulating tissue injury and inflammation in liver, its role in regulating ER stress-induced NAFLD was investigated in two experimental NAFLD models, a tunicamycin (TM)-induced and other diet-induced NAFLD mice models. In the TM-induced NAFLD model, intraperitoneal injection of TM induced liver steatosis in both CHIP+/+ and CHIP+/− mice, but it was severely exacerbated in CHIP+/− mice compared to CHIP+/+ mice. Key regulators of ER stress and de novo lipogenesis were also enhanced in the livers of TM-inoculated CHIP+/− mice. Furthermore, in the diet-induced NAFLD models, CHIP+/− mice developed severely impaired glucose tolerance, insulin resistance and hepatic steatosis compared to CHIP+/+ mice. Interestingly, CHIP promoted ubiquitin-dependent degradation of TXNIP in vitro, and inhibition of TXNIP was further found to alleviate the inflammation and ER stress responses increased by CHIP inhibition. In addition, the expression of TXNIP was increased in mice deficient in CHIP in the TM- and diet-induced models. These findings suggest that CHIP modulates ER stress and inflammatory responses by inhibiting TXNIP, and that CHIP protects against TM- or HF–HS diet-induced NAFLD and serves as a potential therapeutic means for treating liver diseases.


Introduction
Non-alcoholic fatty liver disease (NAFLD) includes a variety of pathologies ranging from hepatic steatosis to non-alcoholic steatohepatitis (NASH), which can progress to liver fibrosis, cirrhosis, or hepatocellular carcinoma [1,2]. NAFLD is characterized by excessive triglyceride accumulation in the liver resulting in type 2 diabetes, hyperlipidemia, obesity, ies have demonstrated that CHIP is an E3 ubiquitin ligase for ER-associated degradation (ERAD), a process involved in cellular adaptations to ER stress [29,30]. In addition, CHIP has been reported to prevent ER stress-induced cell death in the central nervous system [31]. Furthermore, accumulating evidence indicates that CHIP is involved in metabolic pathways including cardiac dysfunction, lung inflammation, liver injury, and NASH [32][33][34][35]. It has also been reported that CHIP −/− mice have shorter lives and exhibit an accelerated aging phenotype, partial perinatal lethality, and altered protein quality control [36,37]. These findings suggest that CHIP regulates ER stress-associated metabolic disorders. Thus, we hypothesized that CHIP might regulate TXNIP expression and might be involved in the development of TM-or diet-induced NAFLD in mice. Specifically, we aimed to elucidate the role of CHIP in the regulation of TXNIP and ER stress-mediated NAFLD in mice. Kyungsan, Republic of Korea). Primary hepatocytes were isolated from 6-8-week-old male mice as previously described [38]. Briefly, the liver was flushed with perfusion buffer (HBSS with EDTA and HEPES) through the portal vein, dissociated with collagenase, and excised and ruptured with fine tip forceps. The hepatocytes were gently released, filtered through a 70 µm cell strainer into a 50 mL tube, and collected by centrifugation at 50× g for 2 min at 4 • C. The hepatocytes were seeded on collagen-coated plates in DMEM supplemented with 10% FBS, 15 mM HEPES (Welgene Inc. Kyungsan, Republic of Korea), 100 nM dexamethasone, 50 U/mL penicillin, and 50 µg/mL streptomycin. The cells were incubated in a humidified atmosphere containing 5% CO 2 at 37 • C.

Animal Experiments
All the animal experiments were approved by the Institutional Animal Care and Use Committee of Yeungnam University College of Medicine (Daegu, Republic of Korea). CHIP +/− mice, generated as described previously [36], were generously provided by Prof. Cam Patterson (University of North Carolina, Chapel Hill, North Carolina). For the tunicamycin (TM, Sigma Aldrich)-induced NAFLD model, 8-week-old female CHIP +/+ , CHIP +/− , and CHIP −/− mice (129/SvEv × C57BL/6 background) were intraperitoneally (i.p.) injected with TM (2 µg/g body weight). Blood samples and liver tissues were collected 8 h or 36 h after TM injection.
For the diet-induced hepatic steatosis model, two different types of diets were used. For the high-fat-high-sucrose (HF-HS) diet-induced NAFLD model, 8-week-old male CHIP +/+ or CHIP +/− mice were fed either company-recommended standard chow (11.5% fat, 0% sucrose, DooYeol Biotech, Seoul, Republic of Korea) or HF-HS diet (36% fat, 30% sucrose, DooYeol Biotech) for 22 weeks. For the high-fat (HF) diet-induced NAFLD model, the mice were fed company-recommended standard chow (10% fat, Research Diets, New Brunswick, NJ) or HF diet (60% fat, Research Diets) for 12 weeks. Body weight was measured weekly, and livers and fat pads were collected at the end of experiments. Fat mass was determined by measuring gonadal fat pad weight.

RNA Interference
For CHIP or TXNIP silencing, cells were transiently transfected with control small interfering (siRNA) or siRNA against CHIP or TXNIP using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Mouse CHIP siRNA and non-specific control siRNA were purchased from Bioneer (Daejeon, Republic of Korea). TXNIP siR-NAs (sc-44944 for mouse, sc-44943 for human) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The mouse and rat specific CHIP target sequence was 5 -GGGAUGAUAUUCCUAGUGC-3 . Non-specific control siRNA was used as the negative control. The cells were harvested 48 to 72 h after siRNA transfection, and protein expression was determined by immunoblotting with specific antibodies.

Histological Analysis
Liver tissues were fixed in 10% buffered formalin, paraffin-embedded, and cut into 5 µm slices. The sections were subjected to hematoxylin and eosin (H&E) staining and immunohistochemistry for TXNIP expression. The sections were deparaffinized, rehydrated, blocked, and then incubated with anti-TXNIP antibody (MBL International). The immunohistochemistry was performed using the HRP/DAP (ABC) detection IHC kit (Abcam, Cambridge, UK) following the manufacturer's instructions. For Oil Red O staining, liver tissues were embedded in optimum cutting temperature compound (OCT, Sakura Finetek, Zoeterwoude, Netherlands), snap-frozen, and then cryosectioned. Frozen liver sections were stained with Oil Red O solution (Sigma Aldrich) for 30 min, rinsed in 60% isopropanol, and then washed with water. All the microscopic images of sections were obtained using an optical microscope (Nikon, Tokyo, Japan).

Intraperitoneal Glucose Tolerance Test, Intraperitoneal Insulin Tolerance Test, and Insulin Sensitivity Assessment
Intraperitoneal glucose tolerance tests (IPGTTs) and intraperitoneal insulin tolerance tests (IPITTs) were performed at the end of the diet feeding schedule (HF-HS diet for 22 weeks, HF diet for 12 weeks). For IPGTT, the mice were fasted for 6 h with free access to water, and then glucose (2 g/kg body weight i.p.) was administered. For IPITT, the mice were fasted for 4 h with free access to water, and then insulin (0.3 U/kg body weight i.p.) was administered. Blood samples were collected by tail tip puncture at 0, 15, 30, 60, 90, or 120 min after glucose or insulin injection for glucose analyses using a glucometer (Accu-Chek; Roche, Indianapolis, IN). To examine insulin sensitivity, HF dietfed mice were administered insulin (1.5 U/kg body weight i.p.) and sacrificed 10 min after insulin injection. Liver tissues were immediately harvested, homogenized, lysed, and immunoblotted using phospho-Akt and Akt antibodies.

Oxidative Status Analysis
Oxidation stress markers were analyzed in liver tissue using the ELISA method. Liver tissues were homogenized in cooled PBS/cell lysates, and then the supernatant was taken as the test sample after centrifugation. By using the assay kits (Cayman Chemical, Ann Arbor, MI, USA), we detected the levels of malondialdehyde (MDA) in the tissues. The homogenized liver tissue was centrifuged at 1600× g for 10 min at 4 • C, then the supernatant was used for biochemical analysis. The color intensity was measured at 530~540 nm. All steps were carried out strictly following the kit instructions.

Ubiquitination Assay
HepG2 cells were transfected with HA-tagged ubiquitin (Addgene #11928), GFPtagged TXNIP (Addgene #18758), Myc-tagged CHIP WT, or Myc-tagged CHIP mutant (H260Q) and incubated for 24 h (CHIP, GeneBank AF 129085.1). The cells were lysed in RIPA buffer containing 5 mM NEM (N-ethylmaleimide), 1 mM PMSF, and 0.01 mM PIC, and centrifuged at 13,000× g for 10 min. The cell lysates were incubated with mouse anti-TXNIP antibody overnight at 4 • C and then incubated with protein A-agarose beads for 1 h on a roller system at 4 • C. The beads were collected by centrifugation and washed with washing buffer. The bound proteins were released using 2× SDS sample buffer. The immunoprecipitates were separated by SDS-PAGE, and the levels of ubiquitinated forms of TXNIP were assessed by immunoblotting with anti-HA antibody.

Statistical Analysis
The results in the bar graphs were expressed as the mean ± S.D. of three independent experiments. The statistical analysis was performed using Student's t test or ANOVA followed by Bonferroni post hoc tests for multiple group comparisons using GraphPad prism 8.0 (Graph-Pad Software Inc., Boston, MA, USA). Probability values (p values) of <0.05 were considered statistically significant.

CHIP Is Involved in Unfolded Protein Responses and Apoptosis In Vitro
To determine whether CHIP mediates ER stress and apoptosis in vitro, AML12 cells were transfected with control siRNA (siControl) or CHIP siRNA (siCHIP) for 48 h and then treated with TM. CHIP knockdown significantly increased the protein levels of TM-induced UPR-related proteins GRP78, ATF6, and XBP-1s (Figure 1a), and cleaved forms of PARP and caspase-3 were also increased by CHIP knockdown (Figure 1b). The mRNA levels of ATF4, CHOP, and ATF6 were also increased by CHIP knockdown (Figure 1c). These findings were further confirmed in primary hepatocytes isolated from CHIP +/+ , CHIP +/− , and CHIP −/− mice where in the TM-induced ER stress response, the protein levels of UPR-related proteins were enhanced by deficiency of CHIP ( Figure 1d). In addition, the ER stress inducer brefeldin A also increased the protein levels of GRP78, CHOP, and cleaved PARP-1 in primary hepatocytes ( Figure 1e). These findings suggest that CHIP deficiency accelerates ER stress and apoptosis in hepatocytes.
1 h on a roller system at 4 °C. The beads were collected by centrifugation and washed with washing buffer. The bound proteins were released using 2× SDS sample buffer. The immunoprecipitates were separated by SDS-PAGE, and the levels of ubiquitinated forms of TXNIP were assessed by immunoblotting with anti-HA antibody.

Statistical Analysis
The results in the bar graphs were expressed as the mean ± S.D. of three independent experiments. The statistical analysis was performed using Student's t test or ANOVA followed by Bonferroni post hoc tests for multiple group comparisons using GraphPad prism 8.0 (Graph-Pad Software Inc., Boston, MA, USA). Probability values (p values) of < 0.05 were considered statistically significant.

CHIP Is Involved in Unfolded Protein Responses and Apoptosis In Vitro
To determine whether CHIP mediates ER stress and apoptosis in vitro, AML12 cells were transfected with control siRNA (siControl) or CHIP siRNA (siCHIP) for 48 h and then treated with TM. CHIP knockdown significantly increased the protein levels of TMinduced UPR-related proteins GRP78, ATF6, and XBP-1s (Figure 1a), and cleaved forms of PARP and caspase-3 were also increased by CHIP knockdown (Figure 1b). The mRNA levels of ATF4, CHOP, and ATF6 were also increased by CHIP knockdown (Figure 1c). These findings were further confirmed in primary hepatocytes isolated from CHIP +/+ , CHIP +/− , and CHIP −/− mice where in the TM-induced ER stress response, the protein levels of UPR-related proteins were enhanced by deficiency of CHIP ( Figure 1d). In addition, the ER stress inducer brefeldin A also increased the protein levels of GRP78, CHOP, and cleaved PARP-1 in primary hepatocytes ( Figure 1e). These findings suggest that CHIP deficiency accelerates ER stress and apoptosis in hepatocytes. were determined by immunoblotting. A-tubulin was used as a loading control. (b) AML12 cells were transfected with siControl or siCHIP for 48 h and then treated with TM (10 µM) for 24 h. Protein levels of cleaved PARP-1, cleaved caspase-3 (cleaved Casp3), and CHIP were determined by immunoblotting. α-tubulin was used as a loading control. (c) AML12 cells were transfected with siControl or siCHIP for 48 h and then treated with TM (10 µM) for 24 h. Expression of UPR-related genes were measured by qRT-PCR. Relative expression levels were normalized to GAPDH levels. ** p < 0.01 vs. siControl, ## p < 0.01 vs. siCHIP, † p < 0.05 and † † p < 0.01. (d) Primary hepatocytes from CHIP +/+ , CHIP +/− , and CHIP −/− mice were treated with TM (2 or 10 µM) for 6 h. Protein levels of GRP78, CHOP, XBP-1s, and CHIP were measured by immunoblotting. β-actin was used as a loading control. (e) Primary hepatocytes from CHIP +/+ and CHIP +/− mice were treated with brefeldin A (BFA, 1 or 2 µM) for 6 or 9 h. Protein levels of GRP78, CHOP, cleaved PARP-1, and CHIP were measured by immunoblotting. α-tubulin was used as a loading control.

CHIP Protects Mice from Tunicamycin-Induced Hepatic Steatosis
To determine whether CHIP plays a role in TM-induced hepatic steatosis in vivo, CHIP +/+ and CHIP +/− mice were challenged with TM (2 µg/g body weight, i.p.) for 8 h or 36 h, and histological analyses were performed on liver tissue sections from TM-injected CHIP +/+ mice and CHIP +/− mice. Lipid accumulation was assessed by Oil Red O staining. Hepatic lipid accumulation was significantly enhanced in the livers of CHIP +/− mice than those of CHIP +/+ mice 36 h after the TM challenge ( Figure 2a). H&E staining of liver sections also showed central vein obstruction after TM treatment (Figure 2a). Furthermore, the livers of CHIP +/− mice showed enhanced mRNA levels of ATF6, ATF4, and CHOP compared with those of CHIP +/+ mice (Figure 2b). Similarly, the protein levels of CHOP and cleaved caspase-3 were higher in CHIP +/− mice than in CHIP +/+ mice at 8 h after TM injection; higher levels of GRP94, ACC, and FAS were also found in CHIP +/− mice at 36 h after TM injection (Figure 2c,d). These results suggest that CHIP haploinsufficiency increases TM-induced ER stress responses and aggravates hepatic steatosis.

CHIP Protects Mice from Diet-Induced Hepatic Steatosis
Diets-induced hepatic steatosis is known to be associated with ER stress [39], and thus, we investigated whether CHIP is involved in diet-induced hepatic steatosis. CHIP +/+ and CHIP +/− mice were fed chow (10% fat) or the HF (60% fat) diet for 12 weeks. CHIP +/− mice fed the HF diet showed enhanced body weight gains compared to CHIP +/+ mice fed the HF diet (Figure 3a), even though the HF diet-exposed mice consumed similar amounts of food as the chow diet-exposed mice (Figure 3b). Significantly enhanced impairment in glucose tolerance was also observed in CHIP +/− mice fed the HF diet compared to CHIP +/+ mice ( Figure 3c). H&E staining and Oil Red O staining revealed the accumulation of larger lipid droplets in the livers of CHIP +/− mice fed the HF diet compared with CHIP +/+ mice (Figure 3d). It has been reported that insulin-stimulated Akt phosphorylation is highly correlated with systemic insulin sensitivity [40]. The levels of Akt phosphorylation in liver tissues of CHIP +/+ and CHIP +/− mice fed the HF were measured 10 min after insulin injection. The phosphorylation level of Akt was found to be significantly lower in CHIP +/− mice (Figure 3e). Relative expression levels were normalized to GAPDH levels. * p < 0.05 and ** p < 0.01 vs. CHIP +/+ mice (vehicle), # p < 0.05 and ## p < 0.01 vs. CHIP +/− mice (vehicle). † p < 0.05. (c,d) Protein levels of CHOP, cleaved Casp3, GRP94, ACC, FAS, and CHIP in liver tissues were measured by immunoblotting. β-actin was used as a loading control.

CHIP Protects Mice from Diet-Induced Hepatic Steatosis
Diets-induced hepatic steatosis is known to be associated with ER stress [39], and thus, we investigated whether CHIP is involved in diet-induced hepatic steatosis. CHIP +/+ and CHIP +/− mice were fed chow (10% fat) or the HF (60% fat) diet for 12 weeks. CHIP +/− mice fed the HF diet showed enhanced body weight gains compared to CHIP +/+ mice fed the HF diet (Figure 3a), even though the HF diet-exposed mice consumed similar amounts of food as the chow diet-exposed mice (Figure 3b). Significantly enhanced impairment in glucose tolerance was also observed in CHIP +/− mice fed the HF diet compared to CHIP +/+ mice (Figure 3c). H&E staining and Oil Red O staining revealed the accumulation of larger lipid droplets in the livers of CHIP +/− mice fed the HF diet compared with CHIP +/+ mice (Figure 3d). It has been reported that insulin-stimulated Akt phosphorylation is highly correlated with systemic insulin sensitivity [40]. The levels of Akt phosphorylation in liver tissues of CHIP +/+ and CHIP +/− mice fed the HF were measured 10 min after insulin injection. The phosphorylation level of Akt was found to be significantly lower in CHIP +/− mice (Figure 3e). Additionally, to induce NASH, CHIP +/+ and CHIP +/− mice were fed chow (11.5% fat, 0% sucrose) or HF-HS (36% fat, 30% sucrose) diet for 22 weeks, and body and fat weight, IPGTT, and IPITT were measured. CHIP +/− mice fed with HF-HS diet showed enhanced body weight gain and fat weight than CHIP +/+ mice (Figure 4a,c). In the high-fat diet group, CHIP +/− mice had higher food intake than CHIP +/+ mice (Figure 4b). To determine whether CHIP is involved in HF-HS-induced glucose abnormalities and insulin resistance, IPGTT and IPITT were performed. CHIP +/− mice on HF-HS diet showed significantly impaired glucose tolerance and insulin resistance compared to CHIP +/+ mice (Figure 4d,e). Histological analysis of H&E-stained liver sections revealed the pathological changes of hepatic steatosis in the livers of CHIP +/− mice fed HF-HS diet (Figure 4f, top), and Oil Red O staining showed significantly more lipid accumulation in the livers of CHIP +/− mice fed HF-HS diet than CHIP +/+ mice fed HF-HS diet (Figure 4f, bottom). Similarly, enhanced protein levels of lipogenic markers (ACC and FAS) and reduced levels of a fatty acid oxidation marker (PGC1α) were observed in CHIP +/− mice fed HF-HS diet compared with CHIP +/+ mice fed HF-HS diet (Figure 4g). Furthermore, it showed that the transcriptional alteration of mRNA levels of genes involved in lipogenesis (ACC), fatty acid oxidation (PPARα) and lipid uptake (CD36, FATP1) by the HF-HS diet was significantly reversed in liver tissues from CHIP +/− mice, but not APOB, a marker of the VLDL secretion pathway ( Figure 4h). Collectively, these results suggest that CHIP protects mice against diet-induced hepatic steatosis and insulin resistance via inhibiting metabolic dysregulation. Additionally, to induce NASH, CHIP +/+ and CHIP +/− mice were fed chow (11.5% fat, 0% sucrose) or HF-HS (36% fat, 30% sucrose) diet for 22 weeks, and body and fat weight, IPGTT, and IPITT were measured. CHIP +/− mice fed with HF-HS diet showed enhanced body weight gain and fat weight than CHIP +/+ mice (Figure 4a,c). In the high-fat diet group, CHIP +/− mice had higher food intake than CHIP +/+ mice (Figure 4b). To determine whether CHIP is involved in HF-HS-induced glucose abnormalities and insulin resistance, IPGTT and IPITT were performed. CHIP +/− mice on HF-HS diet showed significantly impaired glucose tolerance and insulin resistance compared to CHIP +/+ mice ( Figure  4d, e). Histological analysis of H&E-stained liver sections revealed the pathological changes of hepatic steatosis in the livers of CHIP +/− mice fed HF-HS diet (Figure 4f, top), and Oil Red O staining showed significantly more lipid accumulation in the livers of CHIP +/− mice fed HF-HS diet than CHIP +/+ mice fed HF-HS diet (Figure 4f, bottom). Similarly, enhanced protein levels of lipogenic markers (ACC and FAS) and reduced levels of a fatty acid oxidation marker (PGC1α) were observed in CHIP +/− mice fed HF-HS diet compared with CHIP +/+ mice fed HF-HS diet (Figure 4g). Furthermore, it showed that the transcriptional alteration of mRNA levels of genes involved in lipogenesis (ACC), fatty acid oxidation (PPARα) and lipid uptake (CD36, FATP1) by the HF-HS diet was significantly reversed in liver tissues from CHIP +/− mice, but not APOB, a marker of the VLDL secretion pathway (Figure 4h). Collectively, these results suggest that CHIP protects mice against diet-induced hepatic steatosis and insulin resistance via inhibiting metabolic dysregulation. (e) HF diet-fed CHIP +/+ and CHIP +/− mice were injected intraperitoneally with insulin (1.5 U/kg body weight) and sacrificed 10 min after insulin injection. Liver tissues were harvested immediately, proteins were extracted, and immunoblotting was conducted to measure expression levels of p-Akt, Akt, and CHIP. β-actin was used as a loading control. Relative p-Akt levels were normalized versus Akt. ** p < 0.01 vs. CHIP +/+ mice with vehicle (Veh), # p < 0.05 vs. CHIP +/− with Veh, † † p < 0.01.

CHIP Promotes the Ubiquitin-Dependent Degradation of TXNIP
Since TXNIP is a critical regulator of ER stress, glucose homeostasis, cholesterol accumulation, and fatty acid synthesis in the liver and CHIP is a E3 ubiquitin ligase [19], we assessed the direct relationship between CHIP and TXNIP via ubiquitin-dependent protein degradation. To determine whether CHIP regulates TXNIP expression through posttranslational modifications, the effects of CHIP on TXNIP expression were measured in HepG2 cells. The protein levels of TXNIP were decreased by CHIP in a dose-dependent manner (Figure 6a), and the immunoprecipitation analysis further confirmed that TXNIP interacts with CHIP (Figure 6b). To determine whether the ubiquitin/proteasome system mediates CHIP-induced TXNIP degradation, the effect of the proteasome inhibitor MG132 on CHIP-mediated TXNIP degradation was measured. TXNIP expression was markedly decreased by CHIP, and MG132 abrogated the CHIP-induced TXNIP degradation (Figure 6c). Because protein degradation by proteasomes is largely dependent on ubiquitination of target substrates, we investigated whether TXNIP is ubiquitinated by CHIP. Overexpression of CHIP markedly increased ubiquitination of TXNIP (Figure 6d), and TXNIP ubiquitination was greatly diminished by E3 ligase activity-defective mutant CHIP H260Q (Figure 6e), which suggest that CHIP-mediated TXNIP degradation is dependent on its Ub ligase activity. Taken together, our data clearly demonstrate that TXNIP expression is regulated by CHIP-mediated ubiquitination. Figure 5. CHIP is responsible for high-fat and high-sucrose diet-induced oxidative stress and inflammasome formation. CHIP +/+ and CHIP +/− mice were fed chow (11.5% fat, 0% sucrose) or the high-fat-high-sucrose (HF-HS, 36% fat, 30% sucrose) diet for 22 weeks (n = 5-7). (a) MDA concentrations in liver tissues were measured by ELISA method. ** p < 0.01 vs. CHIP +/+ mice with Chow, ## p < 0.01 vs. CHIP +/− mice with Chow. † † p < 0.01. (b) Protein levels of NLRP3 in liver tissues were measured by immunoblotting. α-tubulin was used as a loading control. ** p < 0.01 vs. CHIP +/+ mice with Chow, ## p < 0.01 vs. CHIP +/− mice with Chow. † p < 0.05.

CHIP Promotes the Ubiquitin-Dependent Degradation of TXNIP
Since TXNIP is a critical regulator of ER stress, glucose homeostasis, cholesterol accumulation, and fatty acid synthesis in the liver and CHIP is a E3 ubiquitin ligase [19], we assessed the direct relationship between CHIP and TXNIP via ubiquitin-dependent protein degradation. To determine whether CHIP regulates TXNIP expression through posttranslational modifications, the effects of CHIP on TXNIP expression were measured in HepG2 cells. The protein levels of TXNIP were decreased by CHIP in a dose-dependent manner (Figure 6a), and the immunoprecipitation analysis further confirmed that TXNIP interacts with CHIP (Figure 6b). To determine whether the ubiquitin/proteasome system mediates CHIP-induced TXNIP degradation, the effect of the proteasome inhibitor MG132 on CHIP-mediated TXNIP degradation was measured. TXNIP expression was markedly decreased by CHIP, and MG132 abrogated the CHIP-induced TXNIP degradation (Figure 6c). Because protein degradation by proteasomes is largely dependent on ubiquitination of target substrates, we investigated whether TXNIP is ubiquitinated by CHIP. Overexpression of CHIP markedly increased ubiquitination of TXNIP (Figure 6d), and TXNIP ubiquitination was greatly diminished by E3 ligase activity-defective mutant CHIP H260Q (Figure 6e), which suggest that CHIP-mediated TXNIP degradation is dependent on its Ub ligase activity. Taken together, our data clearly demonstrate that TXNIP expression is regulated by CHIP-mediated ubiquitination.

CHIP Regulates ER Stress and Inflammatory Responses by Inhibiting TXNIP
To determine whether CHIP-mediated TXNIP inhibition is involved in CHIP-dependent regulation of hepatic steatosis, the effects of TXNIP knockdown was evaluated. TM-induced ER stress responses and brefeldin A-induced inflammatory responses were enhanced by CHIP inhibition, and it was alleviated by inhibition of TXNIP using siTXNIP (Figure 7a,b). In addition, TXNIP expression was increased in the livers of TM-and diet-induced NAFLD models (Figure 7c,d). In particular, TXNIP-positive signals were mainly found in hepatic stellate cell-like cells in the livers of diet-induced NAFLD models. These results suggest that CHIP inhibits ER stress, inflammatory responses, and hepatic steatosis by inhibiting hepatic TXNIP expression.

CHIP Regulates ER Stress and Inflammatory Responses by Inhibiting TXNIP
To determine whether CHIP-mediated TXNIP inhibition is involved in CHIP-d pendent regulation of hepatic steatosis, the effects of TXNIP knockdown was evaluat TM-induced ER stress responses and brefeldin A-induced inflammatory responses w enhanced by CHIP inhibition, and it was alleviated by inhibition of TXNIP using siTXN (Figure 7a,b). In addition, TXNIP expression was increased in the livers of TM-and di induced NAFLD models (Figure 7c,d). In particular, TXNIP-positive signals were main found in hepatic stellate cell-like cells in the livers of diet-induced NAFLD models. The results suggest that CHIP inhibits ER stress, inflammatory responses, and hepatic stea sis by inhibiting hepatic TXNIP expression.   Figure 7. CHIP-mediated TXNIP regulation is responsible for ER stress and inflammatory response. (a) HepG2 cells were transfected with CHIP siRNA (siCHIP) and/or TXNIP siRNA (siTXNIP) for 48 h and then treated with tunicamycin (TM, 5 µM) for 8 h. Protein levels of cleaved Casp3, ATF4, CHOP, TXNIP, and CHIP were measured by immunoblotting. α-tubulin was used as a loading control. (b) Primary hepatocytes from CHIP +/+ and CHIP +/− mice were transfected with siTXNIP for 48 h and then treated with brefeldin A (BFA, 1 µM) for 6 h. mRNA expression of inflammatory genes (IL-6, IL-1β, and TNFα) were measured by qRT-PCR. Relative expression levels were normalized to GAPDH levels. ** p < 0.01 vs. siControl, ## p < 0.01 vs. siCHIP. (c) CHIP +/+ , CHIP +/− , and CHIP −/mice were injected intraperitoneally with TM (2 µg/g body weight) and sacrificed 36 h after TM injection. TXNIP expression in liver tissue sections was determined by immunohistochemistry (IHC). (d) CHIP +/+ and CHIP +/− mice were fed chow or HF-HS diet for 22 weeks. Representative images of liver sections from CHIP +/+ and CHIP +/− mice following HF-HS diet fed. TXNIP expression in liver tissue sections was determined by IHC.

Discussion
Recent studies have reported the regulation of TXNIP protein at the post-translational level [43]. However, the post-translational regulatory mechanisms of TXNIP are not well understood yet. TXNIP is a member of the α-arrestin protein family that contains two distinctive arrestin-like domains and two PPxY motifs in the C-terminal tail [44]. Zhang et al. demonstrated that the E3 ubiquitin ligase Itch mediates polyubiquitination of TXNIP. TXNIP and Itch interact via the WW domain and the PPXY motif [45]. CHIP contains three tetratricopeptide repeats at the N terminus, a middle dimerization domain, and a U-box at the C terminus [34]; the U-box domain has intrinsic ubiquitin E3 ligase activity, which promotes the ubiquitination of CHIP-bound target proteins [34,46]. CHIP H260Q mutants carry a point mutation in the U-box domain that interferes with E3 ubiquitin ligase activity [47]. In the present study, we found that CHIP promotes the ubiquitindependent degradation of TXNIP (Figure 6), which provides new insight into the molecular mechanisms underlying the regulation of ER stress and diet-induced hepatic steatosis and demonstrates the role of CHIP as a novel therapeutic gene for treating hepatic steatosis. The identification of a CHIP region that regulates TXNIP expression may reveal promising therapeutic strategies for the treatment of hepatic steatosis. Therefore, future studies will focus on investigating the binding regions of CHIP responsible for its interactions with TXNIP.
Previous reports have shown that ER stress exacerbates hepatic lipidosis [39], which has also been confirmed in the present study. In addition, the present study showed that steatosis is exacerbated in CHIP +/− mice, and provided an underlying molecular mechanism whereby CHIP regulates ER stress-mediated NAFLD in mice.
Recent studies have shown that ER stress and UPR signaling are associated with hepatic steatosis, which is due to either increased lipogenesis or decreased hepatic lipoprotein secretion [48]. Lee et al. reported that hepatic IRE1α/XBP1 controls the expression of lipogenic enzymes (SCD1, ACC2, and DGAT2), which are crucial for fatty acid and cholesterol biosynthesis [49], whereas the IRE1α and/or ATF6 play a role in preventing ER stress-dependent hepatic steatosis [50,51]. Moreover, PERK/eIF2a is required for the expression of lipogenic genes and progression of hepatic steatosis [52]. However, the underlying mechanisms linking ER stress to hepatic steatosis are not fully understood yet. Previous studies have reported that the kidneys of male mice are typically significantly more sensitive to ER stress-induced kidney damage than those of females [17]. Thus, in this study, female mice were used to focus on TM-induced hepatic steatosis. The present study showed that haploinsufficiency of CHIP accelerates TM-induced ER stress and hepatic steatosis in vivo ( Figure 2).
The current prevalence of obesity and related metabolic disorders are closely associated with excessive consumption of HF-HS foods called the western-style diet. The classic western diet is high in both saturated fat and sugar, and has been related to the development of NAFLD. Previous studies demonstrated that the effects of HF and/or HS diets on metabolic risk factors [53][54][55]. A recent report showed that the rapid onset of hepatic steatosis, adipose tissue hypertrophy, and hyperinsulinemia by ingestion of a HF-HS diet may be due to the rapid response of insulin signaling, lipogenesis, and inflammatory genes [56]. In addition, it has also been reported that high-fat diets could only induce steatosis and that the HF-HS diet generated severe steatosis with inflammation, oxidative stress, and myofibroblast and collagen deposition associated with increased serum AST and ALT levels [57]. In this study, haploinsufficiency of CHIP accelerated HF-HS diet-induced weight gain, increased fat mass, impairment of glucose tolerance and insulin resistance, and hepatic steatosis (Figure 4).
Hepatic triglyceride accumulation is a hallmark of NAFLD and results from an imbalance in lipid content between lipid acquisition (de novo lipogenesis, fatty acid absorption) and removal (fatty acid oxidation, VLDL secretion) [58]. It is not clear whether CHIP regulates the progression of NAFLD through the lipid metabolic pathway. Hepatic stellate cells (HSCs) play a critical role in fibrogenesis, and are known to contribute to pathways of inflammation and tissue injury, especially in NASH [59]. In addition, it was reported that inhibition of TXNIP expression can suppress HSC activation in the LX-2 cell line [60]. It was confirmed that the expression of TXNIP was increased in HSCs of CHIP-insufficient mouse liver tissues (Figure 7d). This suggests that CHIP could suppress hepatic fibrosis by regulating TXNIP expression in HSCs.
Further study is required to understand the underlying molecular mechanisms through which TXNIP is regulated by CHIP and how CHIP tightly controls the suppressive response to cirrhosis. Taken together, these findings indicate that CHIP protects against TM-or diet-induced NAFLD and is a potential therapeutic target for the treatment of liver diseases.

Conflicts of Interest:
The authors declare no conflict of interest.