C57BL/6J mice were obtained and raised in the animal center of Sichuan Academy of Medical Science. Experimental mice were kept in SPF conditions with standard housing conditions in a temperature-controlled environment with 12-h light/dark cycles and received chow diet (CD) or high-fat diet (HFD, from 8-week age) and water ad libitum for 12 weeks (15). Animals studies were conducted in accordance with institutional guidelines. For primary hepatocyte culture, we conducted a two-step collagenase perfusion method as described (16). Isolated hepatocytes were cultured in Williams E medium (Sigma) including HepExtend™ Supplement (Gibco), plus penicillin/streptomycin and gentamycin in 37°C under 5% CO₂.

For oil red O staining to validate the hepatosteatosis in mice with HFD, fresh liver samples were fixed in 4% paraformaldehyde (PFA) solution, and embedded with OCT for subsequent staining with Oil Red O, according to standard protocols (17).

To detect the profile of miRNAs in mouse livers under HFD and CD, a total of miRNAs were extracted using miRNeasy mini kit (QIAGEN). The microarray hybridization of miRNAs was performed using GeneChip miRNA 3.0 Array (Affymetrix). Microarray datasets were presented in volcano plot and GO biological process classification, and miRNAs with 2-fold or greater fold change with P-value < 0.05 were considered differentially expressed.

To confirm that miR-124-3p directly targets Pref-1, a luciferase reporter assay was performed. Briefly, the pmirGL-control luciferase vector, pmirGLO dual-Luciferase miRNA target expression vector and luciferase reporter assay system were purchased from Youbia (China). The pmirGLO-control vector was used to construct pGL-pref-1, which contained the 3’UTR of mouse pref-1 (Genbank accession NM003836). DNA fragment of Pref-1 was constructed into vector (18). For luciferase reporter assay, control vector or miR-124-3p plasmid and reporter plasmids were co-transfected into cells. Luciferase activities were measured at 48 h post-transfection by using the dual-luciferase assay system (Turner BioSystem, USA).

For western blots, cultured cells were lysed in 2% SDS buffer plus protease and phosphatase inhibitors (Thermo Scientific, USA). The lysates of equivalent total proteins were separated on SDS-PAGE. Then proteins were transferred into PVDF membrane (Millipore) for incubation with primary/secondary antibodies. The primary antibodies used were as follows, Pref-1 (Cell Signaling Technology, Cat#2069) and GAPDH (Cell Signaling Technology, Cat#2118).

Total RNA was extracted by Animal miRNA Isolation Kit (Foregene, China) and Cell Total RNA Isolation Kit (Foregene, China) followed by standard procedure. The PrimeScriptTM RT reagent Kit (Takara) was used to reverse transcript to cDNA, and SYBR Premix Ex TaqTM II (Takara) was used to qPCR assay in Bio-rad system. The qPCR primers were list as follows: miR-124-3p, F: 5’-TAAGGCACGCGGTGAATGCC-3’, R: 5’-GATTGAATCGAGCACCAGTTAC-3’; TNFα, F: 5’-CAGGCGGTGCCTATGTCTC-3’, R: 5’-CGATCACCCCGAAGTTCAGTAG-3’; Pref-1, F: 5’-TTCGGCCACAGCACCTATG-3’, R: 5’-GGGGCAGTTACACACTTGTCA-3’.

For Pref-1 knockdown, targeted siRNA were transfected into cells using Lipofectamine RNAiMAX reagent (Invitrogen, USA) in Opti-MEM (Invitrogen, USA) over 72 h. The siRNA sequences were as follows, siRNA negative-control: sense-UCCGGAACUGUUACGUGAA; antisense-UUCACGUAACAGUUCCGGA. siRNA Pref-1: sense: UCCUGAAGGUGUCCAUGAA; antisense: UUCAUGGACACCUUCAGGA.

Statistical analysis was performed in GraphPad Prism. Data were expressed as mean ± SD from at least three independent experiments. Statistical differences between two groups were analyzed by Student’s t test, and multi-group comparisons by one-way ANOVA followed by Tukey post hoc tests. P<0.05 was considered statistically significant; **indicates p<0.01, and *** p<0.001.

To investigate the miRNAs involved in hepatosteatosis, we assessed the hepatic miRNA profile in C57BL/6J mice fed with high fat diet (HFD) for 12 weeks, in comparison to littermate controls fed with chow diet (CD). Then we extracted the total miRNA from mouse livers and performed microarray hybridization using GeneChip miRNA 3.0 Array (Affymetrix). Histochemical results confirmed the lipid accumulation by Oil Red O staining in the liver of HFD mice ( Figure 1A ). By miRNA screening, we found that altered miRNAs between HFD and CD mice were enriched in metabolic process by GO (Gene Ontology) biological process classification ( Figure 1B ). As shown in the volcano plot, we differentiated the increased and decreased miRNAs. Particularly, it’s found that miR-124-3p was dramatically increased in the liver of HFD mice ( Figure 1C ). And this increasing was further validated by quantitative evaluation from more paired HFD/CD mice ( Figure 1D ). All these data suggests that increased miR-124-3p may play a vital role in HFD-induced hepatosteatosis.

Based on the finding that miR-124-3p was induced by HFD in mouse liver, we asked whether induction of miR-124-3p has a causal effect on lipid accumulation. Therefore, we examined the lipid contents in hepatocytes after miR-124-3p inhibition. As a first step, we knocked down miR-124-3p expression by transfection of its inhibitors in primary hepatocytes, and the transfection efficiency was confirmed by RT-qPCR ( Figure 2A ). Next, we inactivated miR-124-3p expression in palmetric acid (PA) treated hepatocytes ( Figure 2B ). Although PA could increase triglycerides in control hepatocytes (19), however, miR-124-3p inhibitors dramatically reduced triglyceride contents by PA treatment ( Figure 2C ). Excessive lipid accumulation often coincides with increased inflammation (20). We found that miR-124-3p inhibitors could also decrease the level of TNFα by PA treatment, indicating that miR-124-3p inactivation ameliorates liver inflammation ( Figure 2D ). Reciprocally, we examined whether overexpression of miR-124-3p by its mimics is sufficient to increase lipid contents and inflammations in hepatocytes. Results showed that miR-124-3p overexpression increased triglyceride levels and TNFα expression, suggesting that miR-124-3p may have a causal effect on steatosis in the liver ( Figures 2E–H ). Taken together, these results suggest that miR-124-3p positively regulates lipid contents and inflammation in hepatocytes.

To clarify how miR-124-3p participates in the regulation of lipid metabolism, we next examined the potential targets of miR-124-3p. By sequence analysis, we found that Preadipocyte factor 1 (Pref-1) was a potential candidate of miR-124-3p in the liver ( Figure 3A ). Pref-1 is a transmembrane protein that could be cleaved at the extracellular domain to generate a soluble form that reduces hepatosteatosis and hyperglycemia (21, 22). To confirm that miR-124-3p directly targets Pref-1, we performed a luciferase reporter assay. Data showed that the activity of the reporter plasmid with miR-124-3p mimic was decreased, indicating that miR-124-3p binds to the 3’-UTR of Pref-1 and thus regulates Pref-1 expression ( Figure 3B ). Results of qPCR showed that the mRNA level of Pref-1 was decreased by miR-124-3p overexpression and increased by its knockdown ( Figure 3C ). Consistently, the protein levels of both full and cleaved Pref-1 were both negatively regulated by miR-124-3p ( Figures 3D–F ). Therefore, we propose that miR-124-3p directly targets Pref-1 to regulate its expression.

Pref-1 has been proved to resist HFD–induced hepatosteatosis and obesity (23, 24). To study whether miR-124-3p regulates hepatic lipid metabolism through Pref-1, we assessed the changes of Pref-1 after PA treatment in miR-124-3p inactivated hepatocytes. Western blot results showed that miR-124-3p inhibitors could induce expression of full and cleaved Pref-1 with or without PA ( Figures 4A–C ). It’s noted that Pref-1 expression was slightly increased by PA, however, dramatically increased by miR-124-3p inhibitors ( Figures 4A–C ), suggesting that miR-124-3p inactivation was necessary for Pref-1 expression under lipid stress conditions.

To strengthen the point that miR-124-3p regulates hepatosteatosis through Pref-1, we examined the effect of Pref-1 knockdown on the lipid metabolism and inflammation in hepatocytes ( Figures 4D–F ). Results showed that Pref-1 knockdown did not dramatically alter triglyceride levels or TNFα expression under basal condition. However, in the presence of PA, Pref-1 knockdown caused a more dramatic increasing of triglyceride levels or TNFα expression, indicating that Pref-1 is required for the resistance of triglyceride accumulation under lipid stress ( Figures 4G–H ). Thus, we conclude that miR-124-3p regulated Pref-1 expression is critical for the maintenance of lipid homeostasis in hepatocytes.