Male C57BL6/J mice (6 weeks old) were purchased from the Nanjing Biomedical Research Institute of Nanjing University (Jiangsu, China). The mice were housed in plastic cages in which the humidity and temperature were controlled at 50–60% and 20 ± 2°C. Fenofibrate was dissolved in 0.5% sodium carboxymethyl cellulose (CMC-Na). Mice in the control group were fed a normal diet, while mice in the other group were fed HFD (HFD-fed group), which consisted of 20% carbohydrate, 20% protein, and 60% fat (total 25.07 kJ/g), for 14 weeks. The HFD-fed mice were administered CMC-Na (0.5%) daily via oral gavage for the last 4 weeks of treatment. For the fenofibrate + HFD-fed group, HFD-fed mice were administered fenofibrate (40 mg/kg) via oral gavage daily for the last 4 weeks of treatment. Body weight was measured once a week throughout the study. The animal experiment complied with the ARRIVE guidelines and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All protocols were approved by the Institutional Animal Care and Use Committee of Kunming Medical University (Approval no. kmmu2020248).

Mouse livers were lysed in lysis buffer on ice for 1 h. Protein lysates were loaded into each well (20 μg/10 μl) and separated on SDS-PAGE (7.5, 10, or 12.5%). Proteins were transferred onto PVDF membranes (Millipore, Bedford, MA, United States) and blocked with 5% skim milk-TBST for 2 h at 20°C. The primary antibodies used were anti-ATP2A2/SERCA2b (#4388, Cell Signaling Technology, Beverly, MA, United States), anti-BiP (#3183, Cell Signaling Technology), anti-CCAAT/enhancer-binding homologous protein (CHOP) (#2895, Cell Signaling Technology), and anti-GAPDH (#8884, Cell Signaling Technology). Immunoreactive signals were detected using ECL substrate reagents (#32109, Thermo Scientific Science, Waltham, MA, United States), followed by image analysis (Amersham Imager 600). Quantification was performed using ImageJ (NIH).

Total miRNA in tissue was extracted with the miRcute miRNA Isolation kit (DP501, Transgen, Being, China). cDNA was generated using the miRcute Plus miRNA First-Strand cDNA kit (KR211-01, Trangen, Being, China). qRT-PCR was performed using the miRcute Plus miRNA qPCR kit (FP411-01-01, Trangen) on a Roche LightCycler 480 System. The primers used for the qRT-PCR assay are listed in Supplementary Table S1. All qRT-PCR analyses were performed in triplicate. For each gene, mRNA expression was normalized to that of U6, and the fold change in mRNA expression was determined using the 2^−ΔΔCT method.

After sevoflurane anesthesia, blood was collected from the heart, and fasting blood glucose level was measured. The supernatant was collected, and the serum insulin content was detected using a Mouse INS (Insulin) ELISA Kit (E-EL-M1382c, Elabscience, Wuhan, China). The insulin resistance (IR) and sensitivity indices were calculated using the following formulae: IR index = [fasting blood glucose (mmol/L) × serum insulin (mIU/L)]/22.5 and insulin sensitivity index = 1/[fasting blood glucose (mmol/L) × serum insulin (mIU/L)] (Fang et al., 2010). The supernatant of liver tissue homogenate and serum were used to measure TG content using a TG assay Kit (E-BC-K261, Elabscience).

We use the public prediction platform TargetScan (http://www.targetscan.org/) and MicroRNA Target Prediction Database (miRDB, URL: http://mirdb.org/) to search potential 3′ UTR of SERCA2b that let-7 family members bind.

The 3′ UTR of SERCA2b mRNA was amplified and cloned into the psiCHECK2 vector. HepG2 human hepatocarcinoma cells were seeded onto 24-well plates and cultured in DMEM (Hyclone, Logan, UT) supplemented with 10% fetal bovine serum (Hyclone) at 37°C and 5% CO₂. At 60–80% confluence, the cells were transfected using Lipofectamine 2000® Reagent (Thermo Fisher Scientific) with the 3′ UTR of SERCA2b reporter plasmids with miR-let-7 mimic (5′-AUG​UUG​GAU​GAU​GGA​GUC​UUC-3′) or negative control (NC) mimic (5′-CAG​UAC​UUU​UGU​GUA​GUA​CAA-3′). At 48 h after transfection, the Dual Luciferase Assay System (Promega, Beijing, China) was used to determine luciferase activity. Renilla luciferase activity was used for normalization in each well.

Fresh liver tissues were fixed in 10% formalin, embedded in paraffin, and sectioned at a thickness of 5 μm. To measure hepatic macrosteatosis, the sections were stained with hematoxylin and eosin. To determine fat deposition in each group, the sections were stained with Oil Red O staining solution for 10 min at 60°C. The paraffin-embedded liver tissues were also sliced into 4-μm-thick sections and stained with Masson’s trichrome to observe the degree of liver tissue fibrosis. For immunohistochemical analysis, the sections were stained with anti-ATP2A2/SERCA2b (1:800, #4388, Cell Signaling Technology) or anti-F4/80 antibody (Proteintech, Wuhan, China) per manufacturer’s instructions. Images were acquired using a Leica Aperio CS2 system. Immunohistochemical staining was used to determine the expression of SERCA2b at the protein level, and the expression was quantified using the Image-Pro Plus 6.0 software. The stained sections were assessed at ×400 magnification, and five representative staining fields of view in each section were analyzed. The total area of staining was divided by the total area of the slide (mm²) and multiplied by 100 to obtain the percentage of the stained area.

Data are presented as the mean ± SD. Normal distribution was checked using Shapiro-Wilk test. Statistical differences between two groups were analyzed using Student’s t-test. Statistical analyses between multiple groups were performed using one-way analysis of variance with Student-Newman-Keuls test. GraphPad Prism 7 was used to analyze the data. p < 0.05 was considered statistically significant.

After 14 weeks of HFD feeding, the mice became obese, and this was accompanied by an increase in body weight and hip width (Figures 1B,C). To confirm the beneficial effects of fenofibrate, these DIO mice were administered vehicle (CMC-Na) and fenofibrate via oral gavage daily for 4 weeks (Figure 1A). The body weight and average hip width of DIO mice significantly decreased after fenofibrate administration. Compared with mice administered the vehicle, DIO mice treated with fenofibrate exhibited a significant decrease in hepatocyte ballooning (Figure 1D), macrosteatosis, and fat deposition in the liver (Figure 1D) as well as the levels of serum and hepatic TG (Figures 2B,C). We used Masson staining to detect the fibrosis of liver, and a reduction in liver fibrosis was observed after fenofibrate treatment. Additionally, we measured the expression of F4/80, a macrophage-specific marker; the proportion of F4/80-positive cells in mice in the fenofibrate + HFD-fed group was significantly decreased compared with the HFD-fed group (Figure 1E). Moreover, the expression of stearoyl-CoA desaturase (SCD-1), a key enzyme for monounsaturated fatty acid synthesis (Poloni et al., 2015), was markedly upregulated in DIO mice, and it was reduced by fenofibrate treatment (Figure 2A). Furthermore, IR is one of the most frequent complications of obesity; thus, the effects of fenofibrate on IR were determined. The results indicated that fenofibrate treatment reduced homeostasis model assessment-insulin resistance (HOMA-IR) values (Figure 3A; Supplementary Table S2) and increased homeostasis model assessment-insulin sensitivity index (HOMA-ISI) (Figure 3B; Supplementary Table S2). These data indicate that fenofibrate treatment markedly attenuates NAFLD phenotypes in DIO mice.

ER stress plays a crucial role in the pathology of hepatic steatosis (Puri et al., 2008). Therefore, we tested whether ER stress, activated by HFD feeding, was attenuated by fenofibrate treatment. The expression of two indicators of ER stress at the protein level, BiP and CHOP, was upregulated in DIO mice (Figures 4A,B) compared with mice fed a normal diet. Fenofibrate treatment markedly ameliorated ER stress, which was activated by HFD feeding (Figures 4A,B). The expression or activity of SERCA2b is significantly reduced (Park et al., 2010; Fu et al., 2011; Dai et al., 2019), while increased SERCA2b can ameliorate ER stress in obese mice (Kang et al., 2016; Ouyang et al., 2017). To identify whether the amelioration of ER stress is associated with SERCA2b, the protein levels of SERCA2b were measured by western blotting. Indeed, fenofibrate treatment significantly restored the protein levels of SERCA2b in DIO mice (Figure 4C). Similar results were confirmed by immunohistochemical analysis (Figures 4D,E). These results indicate that fenofibrate treatment ameliorates ER stress and simultaneously upregulates the expression of SERCA2b in the liver of HFD-fed mice.

Overexpression of let-7 disrupts glucose tolerance, while global knockdown of let-7 improves impaired glucose tolerance in DIO mice (Frost and Olson, 2011). Using miRNA microarrays, our recent study indicated that let-7 family miRNA levels are elevated in HFD-fed rats (Dai et al., 2019). There are 12 members in the let-7 family in mice (Papaioannou et al., 2013). Therefore, to investigate whether fenofibrate regulates the expression of let-7 family members, we measured the levels of nine let-7 family members (let-7a, let-7b, let-7c, let-7d, let-7e, let-7f, let-7g, let-7i, and miR-98) in the liver of mice using qRT-PCR and found that the levels of let-7a, let-7b, let-7c, let-7d, let-7e, let-7f, let-7g, let-7i, and miR-98 were increased in DIO mice (Figure 5A). These results indicate that fenofibrate treatment significantly inhibits the levels of let-7 family members in the liver of DIO mice.

SERCA2b sequence is highly conserved in different species, including humans and rodents (Figure 5B). To identify whether SERCA2b is regulated by let-7, the sequence alignment of SERCA2b 3′ UTR and let-7 family members was analyzed using the TargetScan algorithm. The data indicated that let-7 family members could bind to the 3′ UTR of SERCA2b (Figure 5C; Supplementary Table S3). To confirm the involvement of let-7 in the regulation of SERCA2b expression, the 3′ UTR of SERCA2b was cloned into a reporter vector, and transfection with miR-let-7 mimics markedly decreased the luciferase activity of the 3′ UTR of SERCA2b in HepG2 cells (Figure 5D). These data demonstrate that let-7 miRNA directly targets SERCA2b.