Mesenchymal stromal cells protect hepatocytes from lipotoxicity through alleviation of endoplasmic reticulum stress by restoring SERCA activity

Abstract The aim of this study was to investigate how mesenchymal stromal cells (MSCs) modulate metabolic balance and attenuate hepatic lipotoxicity in the context of non‐alcoholic fatty liver disease (NAFLD). In vivo, male SD rats were fed with high‐fat diet (HFD) to develop NAFLD; then, they were treated twice by intravenous injections of rat bone marrow MSCs. In vitro, HepG2 cells were cocultured with MSCs by transwell and exposed to palmitic acid (PA) for 24 hours. The endoplasmic reticulum (ER) stressor thapsigargin and sarco/ER Ca2+‐ATPase (SERCA2)–specific siRNA were used to explore the regulation of ER stress by MSCs. We found that MSC administration improved hepatic steatosis, restored systemic hepatic lipid and glucose homeostasis, and inhibited hepatic ER stress in HFD‐fed rats. In hepatocytes, MSCs effectively alleviated the cellular lipotoxicity. Particularly, MSCs remarkably ameliorated the ER stress and intracellular calcium homeostasis induced by either PA or thapsigargin in HepG2 cells. Additionally, long‐term HFD or PA stimulation would activate pyroptosis in hepatocytes, which may contribute to the cell death and liver dysfunction during the process of NAFLD, and MSC treatment effectively ameliorates these deleterious effects. SERCA2 silencing obviously abolished the ability of MSCs against the PA‐induced lipotoxicity. Conclusively, our study demonstrated that MSCs were able to ameliorate liver lipotoxicity and metabolic disturbance in the context of NAFLD, in which the regulation of ER stress and the calcium homeostasis via SERCA has played a key role.


| INTRODUC TI ON
Non-alcoholic fatty liver disease (NAFLD) is characterized by abnormal lipid accumulation in hepatocytes in the absence of alcohol abuse, and can progress from simple steatosis to non-alcoholic steatohepatitis (NASH), cirrhosis and ultimately cancer. In the last decade, the prevalence of NAFLD is strikingly increased, reaching 30% in adults in developed countries. 1,2 Although intense lifestyle change aiming at weight loss is the main therapy for hepatic steatosis, it is ineffective when NAFLD has progressed to NASH and further severe liver damage.

Mesenchymal stromal cells (MSCs) are multipotent stromal cells,
showing promising therapeutic potentials in a spectrum of clinical settings, including heart failure, graft vs host disease (GvHD) and acute or chronic liver damage. 3 Our previous studies have demonstrated that MSC transplantation ameliorated hyperlipidaemia in STZ-induced diabetic rats. 4 Nevertheless, how MSCs modulate metabolic balance and attenuate hepatic lipotoxicity in the context of NAFLD is still elusive.
The pathogenesis of NAFLD is hypothesized to begin with abnormal accumulation of lipids in the liver due to a stress condition such as saturated fatty acid (SFA) overexposure or insulin resistance, and several signalling mechanisms including the accumulation of reactive oxygen species, endoplasmic reticulum (ER) stress, and chronic inflammation are involved in this process. 1,5,6 Endoplasmic reticulum performs important functions related to the synthesis, folding and transport of proteins and plays a critical role in lipid synthesis and Ca 2+ homeostasis. 7 Prolonged SFA overuptake leads to imbalance between ER protein load and ER folding, resulting in hyperactivation of unfolded protein response (UPR) and consequently activation of cell death pathways. 8 Thus, any substance that opposes ER stress may have possible potential to treat NAFLD.
Endoplasmic reticulum calcium disequilibrium is considered to be one of the initial and pivotal events of ER stress-mediated cell death. 9 Herein, the disruption of cellular Ca 2+ homeostasis is a hallmark of many ER-related diseases and a key trigger of NAFLD/NASH as well. The sarco/endoplasmic reticulum (ER/SR) Ca 2+ ATPase (SERCA) transports Ca 2+ from the cytosol to the ER or SR lumen, maintaining the resting calcium concentration. Previous studies have demonstrated that SFAs inhibit the activity of SERCAs, which leads to calcium release from the ER lumen and a heavy calcium overload in the cytosol, ultimately resulting in the disruption of ER homeostasis and cell function. 10,11 Here, we aimed to evaluate the potential therapeutic effects of MSCs on hepatic lipotoxicity using high-fat diet (HFD)-fed rats in vivo and palmitic acid (PA)-treated hepatocytes in vitro, and explore the underlying mechanisms focusing on the regulation of ER stress and SERCA activity by MSCs.

| Experimental animals
Eighteen male Sprague Dawley rats (6-8 weeks) were obtained from the DASHUO animal company and housed in the Animal Center of West China Hospital, Sichuan University. Protocols for animal study were approved by the Institutional Animal Care and Use Committee (IACUC) of West China Hospital, Sichuan University. The rats were kept in standard laboratory conditions with free access to food and water and were allowed to adapt to the new environment for two weeks before experimental procedures. Rats were randomly assigned to a control group (fed with chow diet, n = 6), a HFD-fed group (2% cholesterol, 10% lard, 20% sucrose, 0.2% sodium cholate and 67% standard diet, n = 6) and a MSC-treated group (HFD + MSCs, n = 6).
The chow diet and HFD were purchased from Sichuan Academy of Agricultural Cultural Science. The energy contribution in percentual of nutrient content for each diet is listed in Table 1. After 18 weeks HFD feeding, rats of HFD + MSCs group were administered of MSCs (2 × 10 6 cell/rat, suspended in 1 mL saline) via the rat's tail vein for two times with an interval of 2 weeks, whereas controls received 1mL saline. Four weeks after the last injection, all the animals were killed for sample collection ( Figure 1A).

| Isolation and culture of rat bone marrowderived MSCs
Rat MSCs (rMSCs) were harvested from the bone marrow of healthy SD rats (80-100 g) and cultured as described previously. 12

TA B L E 1
The energy contribution in per cent of nutrient content for each diet F I G U R E 1 rMSC administration ameliorated HFD-induced dyslipidaemia and steatosis in liver. Protocol for rMSC therapy in high-fat diet-fed rats (A): After 18 wk of HFD feeding, SD rats were separated into two groups, one group received two doses of 2 × 10 6 rMSCs (HFD + MSCs group, n = 6). The other group received the vehicle (HFD group, n = 6). Both groups were fed HFD during all study period (24 weeks). A third group of rats was fed exclusively with regular diet (control group, n = 6). Changes in bodyweight (B) and food intake (C) in three groups. Macroscopic image of liver (top), H&E staining (middle) and Oil Red O staining (bottom) of liver sections of the three groups at the 24th week (D). Liver/bodyweight ratio (E), levels of liver TG and liver TC (F), serum levels of AST, ALT (G) and lipid (H) in the three groups. And the mRNA levels of lipid metabolism-relative genes in the liver tissues of the three groups (I), and the experiment was repeated 3 times independently. The data are expressed as mean ± SD, *P < .05 vs the control group; # P < .

| Statistical analysis
Experiments were performed at least three times, and quantitative data are expressed as mean ± SD All data were analysed using the SPSS software (version 17.0), and comparison between two groups of values was performed by t test. A value of P < .05 was considered to be at significant difference.

| rMSC therapy improved lipid metabolism in HFD-fed rats
The bodyweights and food intake of the HFD group showed no  Figure 1D), as well as decreased liver weight/bodyweight ratio and liver TG and TC contents ( Figure 1E,F). Additionally, the serum levels of TG, TC and LDL-c were increased in HFD rats but decreased in rMSC-treated rats, while the HDL-c level was decreased in HFD rats but increased in rMSC-treated rats ( Figure 1H). And AST was increased in HFD rats but decreased in rMSC-treated rats, while the ALT levels showed no significant changes in the three groups ( Figure 1G), suggesting that rMSCs have potential to improve liver functional abnormality.
Furthermore, the mRNA levels of lipid metabolism-related genes in the liver were determined by qPCR. The results suggested that the expressions of many metabolic genes were changed in the HFD group at the 24 th week, including FFA transport-and synthesisrelated genes such as Ppara and Scd1. Particularly, the gene expressions of Srebp1 and Srebp2, which are activated by the ER stress pathway 7,13,14 and known as key regulators of lipid and cholesterol synthesis, respectively, were promoted in the HFD rats but inhibited in rMSC-treated rats ( Figure 1I).

| rMSCs improved insulin sensitivity in HFDfed rats
To explore the effects of rMSC treatment on glucose metabolism of HFD-fed rats, we measured the fasting glucose levels and serum insulin levels at the 24 th week. The fasting glucose levels were slightly increased in HFD-fed rats but decreased in rMSC-treated rats (P > .05).
The fasting serum insulin levels were remarkably increased in HFDfed rats indicating a status of hyperinsulinaemia, while rMSC treatments reversed the insulin to normal levels (

| rMSCs inhibited ER stress in the liver of HFDfed rats
After 24 weeks, HFD strongly induced hepatic ER stress, as indicated by increased protein levels of PERK and ATF6 pathway markers including p-eIF2a, ATF4 and ATF6, as well as the chaperone protein BiP. But the sliced XBP did not show significant difference. However, these up-regulations were significantly inhibited by rat MSCs treatment ( Figure 2H). Consistently, the mRNA expressions of UPRrelated genes including Bip, Edem1 and Erp72 were increased after HFD, but decreased in the rMSC-treated group ( Figure 2I). Apoptosis is believed to be closely linked to ER stress. As we expected, the protein expression of pro-apoptotic molecule BAX was markedly increased in the HFD group but decreased in the rMSC-treated group. Conversely, the antiapoptotic molecule Bcl-2 was increased in the rMSC-treated group compared with its decrease in the HFD group ( Figure 2J).

| hMSCs ameliorated PA driving hepatic lipotoxicity in HepG2 cells
In agreement with previous reports, 15 Figure 3D). As a long-chain SFA, PA is β-oxidized in mitochondria, and PA-induced lipotoxicity is able to cause mitochondrial dysfunction, which is indicated by impaired bioenergetics. 17 So, we determined the ATP production after the cells were cultured with PA for 24 h. The results showed that the ATP levels of HepG2 cells in the PA-treated group were decreased but were restored in the PA-MSC group ( Figure 3E). In addition, the p-Akt (Ser473) and p-IRS-1 (Try896) protein levels of HepG2 were decreased in the PA group but increased in the PA-M group ( Figure 3F).

F I G U R E 2
Effects of rMSC administration on glucose metabolism, insulin pathway and liver ER stress of HFD-fed rats. Fasting blood glucose and insulin levels of the three groups were measured at the 24th week (A, C). Blood glucose levels in a glucose tolerance test (GTT) and an insulin tolerance test (ITT) (B, D). Representative images of pancreas stained with anti-insulin antibody in the three groups (E) and the quantitative analysis of the insulin-positive areas in the three groups (F). The protein levels of p-IRS-1 and p-Akt in the liver tissues from the three groups were measured by Western blotting, β-actin was used as a protein loading control (G). The protein levels of ER stress markers in liver were changed in the three groups at the 24th week (H); the mRNA expression of ER stress-relative genes in liver of the three groups (I); the Bcl-2 and BAX expressions were measured by Western blotting, and β-actin was used as protein loading control (J), and the experiment was repeated 3 times independently. The data are expressed as mean ± SD, *P < .05 vs the control group; were up-regulated in the PA group but decreased in the PA-M group.
In addition, the PA-M group showed an increase of mRNA level of CPT1A, a mitochondrial enzyme responsible for the transfer of fatty acids to mitochondria for β-oxidation ( Figure 4E).

| hMSCs alleviate PA-induced ER stress in hepatocytes
Our previous studies have indicated that ER stress played a key role in SFA-induced lipotoxicity both in hepatocyte and in pancreatic β cell. 18,19 To convince the effect of hMSC treatment on ER stress in HepG2 cells, we first observed the morphology of the ER using We further convinced that in vivo, the SERCA expression levels in the liver of the HFD-fed rats were significantly decreased but were increased in the HFD-MSC group ( Figure 5I,J).

| MSCs alleviate Thp-induced hepatocyte injury and ER stress
To further convince the relevance of SERCA to MSC-mediated pro-  (Table 3). Although the bodyweight showed no significant difference between the groups, the HFD group showed macrosteatosis in liver, which suggested that this animal model may mimic the progression of NAFLD in non-obesity humans. Meanwhile, the HFD rats that received rMSCs did not develop predominant macrovesicular steatosis characterized with massive ballooning, and showed a decreased liver TG and TC content, indicating a significant reverse of fatty liver. It is noteworthy that insulin resistance is also an important character when the development of NAFLD to NASH, and consistently, the impairment of glucose tolerance, hyperlipidaemia and hyperinsulinaemia levels were detected in these HFD rats. 24 Our results indicated that rMSC treatment dramatically reduced the fasting insulin level in serum and the compensatory hyperplasia in Numerous studies have demonstrated that ER stress is a major contributor to NAFLD, but its role varies depending on the cause and progression of the disease. Moreover, it is important to note that each cell type responds to ER stress and activates the UPR in a unique manner. 7,8,27,28 As previous studies mentioned, the UPR is regulated by the three master regulators, IRE1, PERK and ATF6. Activated IRE1α splices X-box binding protein 1 (XBP1) mRNA and then activates further downstream genes that function in ERAD such as ER-degradation-enhancingα-mannidose-like protein (EDEM). 28 And the UPR adjusts the capacity of the ER to fold and remove abnormally folded proteins according to need in a dynamic way to re-establish homeostasis. 29  TA B L E 3 NAFLD and NASH histological scoring system and scores of all the experimental groups XBP1 expression showed no significant difference in HFD-fed rats compared with the control group at the 24th week ( Figure 2H).
Furthermore, the XBP1s showed a decrease in PA-treated HepG2 cells ( Figure 7D), but an increase in the PA-M group. lipids and organelles to recipient cells. 38 It is reported that in vivo, EVs from MSCs with miRNA-181-5p overexpression alleviated liver fibrosis via activation of autophagy in CCl4-induced liver injury. In vitro, transfer of miNRA-181-5p from MSCs to mouse hepatic stellate cells by EV uptake was shown to be responsible for MSC-mediated protection. 39 Additionally, the hepatocyte growth factor (HGF) derived from conditioned culture media of MSCs has been demonstrated playing an important role in calcium homeostasis and ER stress in fibrotic lungs. 40 We assume that the paracrine activity of MSCs is essential for their therapeutic functions, and the exact mechanisms are warranted to be investigated in more details.
In conclusion, we established NAFLD rats and PA-induced hepatocytes to investigate the effects of MSC therapy on steatosis, insulin resistance and dyslipidaemia, and explored the potential molecular mechanisms. Our findings demonstrated that MSC treatment restored the intracellular Ca 2+ homeostasis by regulating SERCA, contributing to the alleviation of ER stress and improvement of metabolic dysfunction in NAFLD rats and PA-induced HepG2 cells.
We reaffirmed that ER stress plays a critical role in the pathology of NAFLD, and SERCA could be an interesting target for metabolic dysfunction. Our study suggested that MSC treatment may be considered as a novel and promising therapeutics in future to benefit the Ca 2+ homeostasis and ER stress-related disorders.

CO N FLI C T O F I NTE R E S T
We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work; there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, this paper.