Hsa_circ_0048179 attenuates free fatty acid-induced steatosis via hsa_circ_0048179/miR-188-3p/GPX4 signaling

Although circular RNAs (circRNAs) are known to play key roles in non-alcoholic fatty liver disease, much about their targets and mechanisms remains unknown. We therefore investigated the actions and mechanisms of hsa_circ_0048179 in an in vitro model of NAFLD. HepG2 cells were exposed to oleate/palmitate (2:1 ratio) for 24 h to induce intracellular lipid accumulation. Using CCK-8 assays, flow cytometry, fluorescence microscopy, western blotting, RT-qPCR, and Oil red O staining, we found that oleate/palmitate treatment reduced cell viability while increasing apoptosis and lipid accumulation in HepG2 cells. Levels of the antioxidant enzyme GPX4 were decreased in oleate/palmitate-treated HepG2 cells, and there were corresponding increases in reactive oxygen species and damage to mitochondrial cristae. Levels of hsa_circ_0048179 expression were also suppressed by oleate/palmitate treatment, and GPX4 levels were markedly increased in HepG2 cells following transfection with hsa_circ_0048179. Analysis of its mechanism revealed that hsa_circ_0048179 upregulated GPX4 levels by acting as a competitive “sponge” of miR-188-3p and that hsa_circ_0048179 attenuated oleate/palmitate-induced lipid accumulation in HepG2 cells by sponging miR-188-3p. Collectively, our findings suggest that hsa_circ_0048179 may play a key role in the pathogenesis of steatosis and may thus be a useful target for drug development.

liver tissue not due to alcohol abuse [6,7]. NAFLD severely affects the quality of life of patients and is a financial burden for society [8,9]. However, the pathogenesis of NAFLD has not been entirely elucidated. It is therefore important to continue to investigate the pathogenesis of NAFLD and expand the available treatment options.
Circular RNA (circRNA) is an endogenous noncoding RNA molecule ubiquitous in eukaryotes [10]. CircRNAs have a closed loop structure with no 5" cap or 3" poly(A) tail, and often originate from exons or introns within gene fragments [10,11]. These molecules are now known to play vital roles in a number of cellular processes, including cell growth, migration, and invasion, as well as carcinogenic transformation [12]. CircRNAs regulate gene expression by binding to proteins or sponging microRNAs (miRNAs) [13]. Recently, Jin et al. reported that circRNAs may function as molecular diagnostic biomarkers of NASH acting via a circRNA-miRNA-mRNA axis [14]. For example, circRNA_ 0046366 could inhibit hepatocellular steatosis in high fat-treated HepG2 cells via miR-34a/PPARα axis [15]. In addition, hsa_circ_0048179 was found to be expressed in the HepG2 human hepatoblastoma cell line, normal human epidermal keratinocytes (NHEK), and HeLa-S3 cell line [16]. However, the potential role of hsa_circ_0048179 in the progression of NAFLD and its underlying mechanism remain unclear. In the present study, therefore, we explored the roles of hsa_circ_0048179 in the pathogenesis of NAFLD.

In vitro oleate/palmitate-induced NAFLD model
HepG2 cells were exposed to oleate/palmitate for 24 h to induce intracellular lipid accumulation. CCK-8 assays were used to assess the viability of HepG2 cells after treatment with oleate/palmitate for 12, 24 and 36 h. Incubation in the presence of oleate/palmitate led to significant time-dependent decreases in HepG2 cell viability and corresponding increases in apoptosis ( Figure  1A). In addition, apoptosis rate was 12.65 ± 2.78%, 36.60 ± 3.61% and 45.97 ± 4.21% in HepG2 cells after treatment with oleate/palmitate for 12, 24 and 36 h respectively, compared with control (3.26 ± 0.35%) ( Figure 1B, 1C). Oil red O staining revealed that intracellular accumulation of lipid was significantly enhanced in oleate/palmitate-treated HepG2 cells ( Figure  1D, 1E). These data illustrate that oleate/palmitate induces the apoptosis and lipid accumulation characteristic of NAFLD in HepG2 cells. We have thus established an in vitro model of NAFLD. To maximize fat accumulation while minimizing cytotoxicity, incubation with oleate/palmitate for 24 h was selected for subsequent experiments.

Oleate/palmitate induces ROS generation and decreases mitochondrial membrane potentials (MMPs) in HepG2 cells
It was previously reported that lipid deposition in hepatocytes leads to increased ROS generation and mitochondrial dysfunction [17]. As shown in Figure 2A, oleate/palmitate markedly increased ROS generation in HepG2 cells, as indicated by the increased DCFH-DA fluorescence intensity. In addition, oleate/palmitate exposure also significantly decreased MMPs in HepG2 cells ( Figure 2B, 2C), and transmission electron microscopy (TEM) revealed mitochondrial swelling and fragmentation of the mitochondrial cristae in the oleate/palmitate-treated cells ( Figure 2D). Moreover, MTG staining results indicated that oleate/palmitate exposure decreased the mitochondrial fluorescence intensity in HepG2 cells, compared with control group ( Figure 2E). Glutathione peroxidase 4 (GPX4), a phospholipid hydroperoxidase, protects cells against membrane lipid peroxidation [18]. As shown in Figure  2F, 2G, oleate/palmitate exposure resulted in decreased expression of GPX4 in HepG2 cells in a time-dependent manner.

Hsa_circ_0048179 increases GPX4 levels in HepG2 cells
Has_circRNA_0048179 is reported to be expressed in HepG2 cells [16]. As shown in Figure 3A, hsa_circ_0048179 levels were significantly lower in oleate/palmitate-treated HepG2 cells than in untreated control cells. To determine the function of hsa_circ_ 0048179, we used a lentiviral vector to overexpress it HepG2 cells ( Figure 3B). We found that levels of GPX4 were upregulated in cells overexpressing hsa_circ_ 0048179 ( Figure 3C, 3D). Moreover, overexpression of hsa_circ_0048179 led to increased expression of GPX4 in oleate/palmitate-treated HepG2 cells in a timedependent manner (Supplementary Figure 1A, 1B). These data suggest that hsa_circ_0048179 may suppress lipid deposition and mitochondrial dysfunction by mediating enhancement of GPX4 activity. AGING Hsa_circ_0048179 functions as a ceRNA targeting miR-188-3p in NAFLD CircRNAs regulate gene expression by binding to proteins or sponging miRNA [19]. Data from the online bioinformatics tool circbase (http://www.circbase.org) and the circular RNA interactome (https:// circinteractome.nia.nih.gov/index.html) indicate that hsa_circ_0048179 is an exonic circRNA, which are known to regulate expression of target genes by sponging their miRNAs [13]. We therefore used the Circular RNA interactome, TargetScan (http://www.targetscan.org/ vert_71/) and miRDB (http://www.mirdb.org) to identify miRNAs that interact with hsa_circ_0048179 and target GPX4. The bioinformatics results predicted that miR-188-3p was a likely target of both hsa_circ_0048179 and GPX4 ( Figure 4A, 4B). Moreover, levels of miR-188-3p were unaffected by oleate/palmitate stimulation of HepG2 cells ( Figure 4C), and fluorescence in situ hybridization (FISH) assays showed that hsa_circ_0048179 and miR-188-3p partially co-localized in the cytoplasm ( Figure 4D). This suggests that hsa_circ_0048179 directly interacts with miR-188-3p and that hsa_ circ_0048179 functions as a ceRNA targeting miR-188-3p in NAFLD.

Overexpression of hsa_circ_0048179 attenuates oleate/palmitate-induced lipid accumulation in HepG2 cells by sponging miR-188-3p
Having verified that hsa_circ_0048179 targets miR-188-3p, the mechanism of action of hsa_circ_0048179 in oleate/palmitate-induced lipid accumulation in HepG2 cells was still unknown. Oleate/palmitate significantly inhibited HepG2 cell viability while increasing both apoptosis and ROS levels within the cells, but these effects were reversed by infection with lenti-hsa_circ_0048179 viral vector. However, when HepG2 cells were treated with hsa_circ_0048179 plus miR-188-3p agomir in the presence of oleate/palmitate, the favorable effects of hsa_circ_0048179 were AGING suppressed by the miR-188-3p overexpression ( Figure  6A-6D). As summarized above, oleate/palmitate significantly decreased MMPs in HepG2 cells. This effect too was reversed by transfection with lenti-hsa_circ_0048179, and again the beneficial effect of hsa_circ_0048179 was suppressed by miR-188-3p agomir ( Figure 6E, 6F). These results further illustrate that overexpressed hsa_circ_0048179 attenuates oleate/palmitate-induced lipid accumulation in HepG2 cells by sponging miR-188-3p.
The hallmark in NASH pathogenesis is lipid deposition within hepatocytes [17]. Ricchi et al. found that oleic and palmitic acids most effectively induced steatotic deposition in hepatocyte cultures [20]. It has been shown that apoptosis and ROS generation are associated with the progression of NASH, and overproduction of ROS may lead to mitochondrial oxidative stress [21,22]. ROS, including O2· -, H 2 O 2 and ONOO − , are produced in all types of cells [23]. In hepatocytes, O2·is normally converted to H 2 O 2 and then to H 2 O by antioxidant GPX enzymes [24]. GPX4 is able to directly reduce fatty acid and phospholipid hydroperoxide [25]. In the present study, oleate/palmitate-induced fatty acid accumulation was accompanied by significant reductions in GPX4 levels, which likely account for the increased ROS levels and reduced MMPs.
CircRNAs are reported to be differentially expressed in HepG2 cells, as compared to non-cancer cells [16]. Our study is the first to directly investigate the association between hsa_circ_0048179 levels and steatosis. We found that hsa_circ_0048179 levels are downregulated in oleate/palmitate-treated HepG2 cells and that GPX4 expression was markedly increased in HepG2 cells infected with lenti-hsa_circ_0048179 viral vector. Because exon-derived circRNAs reportedly regulate target genes by sponging miRNAs [13], we used the Circular RNA interactome, Targetscan and miRDB to identify miRNAs interacting with hsa_circ_0048179 and the GPX4 3"-UTR. We found miR-188-3p to be a potential common target for both hsa_circ_0048179 and GPX4. Consistent with that finding, dual-luciferase reporter and FISH assays indicated that miR-188-3p is the binding target of hsa_circ_0048179, while GPX4 is the target of miR-188-3p. Taken together, these findings suggest NAFLD may be related to abnormalities in hsa_circ_0048179/miR-188-3p/GPX4 signaling.
Although our findings implicate hsa_circ_0048179 in steatosis, the mechanism by which it affects lipid accumulation in oleate/palmitate-treated HepG2 cells remains unclear. We found that downregulation of miR-188-3p attenuated oleate/palmitate-induced apoptosis, ROS generation and mitochondrial damage in HepG2 cells, while overexpression of hsa_circ_0048179 reversed those adverse effects. Moreover, the beneficial effects of overexpressed hsa_circ_0048179 were, in turn, reversed by transfection of miR-188-3p agomir. These results demonstrate that overexpression of hsa_circ_0048179 attenuates oleate/palmitate-induced lipid accumulation in HepG2 cells by sponging miR-188-3p and that the effects of hsa_circ_0048179 are attributable, at least in part, to its function as a ceRNA targeting miR-188-3p.

Cell culture and treatment
HepG2 human hepatoblastoma cells were purchased from the American Type Culture Collection (ATCC, Rockville, MD, USA) and were maintained in Dulbecco's modified Eagle's medium (DMEM; Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific) and 100 units/ml penicillin/streptomycin at 37° C and 5% CO 2 . For oleate/palmitate treatment, the HepG2 cells were incubated with free fatty acids (oleate/palmitate, 2:1 molar ratio, to achieve a final concentration of 1 mM fatty acids) for 24 h, and total RNA and protein were extracted for experimentation [15,26].

Lentivirus production and cell transfection
Has_circ_004817 was synthesized by GenePharma (Shanghai, China) and subcloned into the pHBLV-ciR vector (HANABIO, Shanghai, China). Lentivirus carrying has_circ_004817 was packaged in 293T cells, after which the supernatant was concentrated. HepG2 cells were then infected for 72 h using the lenti-has_circ_004817 supernatant.
Agomir NC, miR-188-3p agomir, antagomir NC and miR-188-3p antagomir were purchased from GenePharma and respectively transfected into HepG2 cells using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer"s instructions. The culture medium containing 10% FBS was changed after 6 h of transfection, and the cells were then incubated for additional 66 h at 37˚ C.

HepG2 cell viability
Cell counting kit-8 assays (CCK-8, Dojindo, Tokyo, Japan) were used to determine the viability of HepG2 cells. HepG2 cells were plated onto 96-well plates at a density of 5 x 10 3 cells per well and incubated overnight at 37° C. This was followed by incubation with oleate/palmitate for 0, 12, 24 and 36 h, after which the cells were incubated with 10 μL CCK-8 reagent for 2 h at 37° C. The optical density (OD) value of each well at the wavelength of 450 nm was then measured using a microplate reader (Thermo Fisher Scientific).

Flow cytometry
Annexin V-FITC/PI Apoptosis detection kits (Thermo Fisher Scientific) were used to assess cell apoptosis. HepG2 cells were incubated with oleate/palmitate for 0, 12, 24 or 36 h and then fixed in 70% pre-cooled ethanol and washed twice with cold PBS. Thereafter, the cells were incubated with 5 μL of Annexin V-FITC and 5 μL of propidium iodide (PI) for 15 min at room temperature in the dark. Apoptotic cells were then detected using a flow cytometer on a BD FACSCalibur system (FACScan; BD Biosciences, Franklin Lake, NJ, USA).

Measurement of intracellular ROS levels
ROS generation was measured using 29,79dichlorodihydrofluorescein (DCFH-DA, Sigma Aldrich, St. Louis, MO, USA) according to the manufacturer"s protocol. HepG2 cells were incubated with oleate/palmitate for 24 h, after which they were washed twice with PBS and incubated with 10 μM DCFH-DA for 20 min at 37° C in the dark. The fluorescent signals were then detected using flow cytometry (BD Biosciences)

Oil red O staining
Lipid deposition in HepG2 cells was assessed by staining with Oil red O as previously described [27]. HepG2 cells were fixed in 10% formalin for 5 min, after which the cells were washed with isopropanol, stained for 30 min with Oil Red O reagent (Sigma Aldrich), and counter-stained for 1 min with hematoxylin. The lipid deposition in cells was observed under a fluorescence microscope (Leica, Buffalo Grove, IL, USA). The Oil red O positive cell count was measured with the ImageJ 1.47 software.

Detection of MMPs
JC-1 assay kits (Thermo Fisher Scientific) were used to evaluate mitochondrial depolarization. HepG2 cells were incubated with 1 mL of JC-1 staining solution for 20 min at 37˚ C in the dark. The stained cells were then washed twice with PBS and resuspended in PBS for analysis with a flow cytometer (BD Biosciences).

Analysis of mitochondrial morphology
Changes in mitochondrial morphology were evaluated using transmission electron microscopy (TEM). HepG2 cells were incubated with oleate/palmitate (2:1 molar ratio) for 24 h. Subsequently, HepG2 cells were fixed by 2.5% glutaraldehyde (Sigma-Aldrich) in 0.1 M phosphate buffer for 2 h at 4° C and then centrifuged at 1000 rpm for 5 min. After that, HepG2 cells were fixed in 1% osmium tetroxide in 0.1 M phosphate buffer for 1 h. Later on, the fixed cells were dehydrated by a graded series of ethanol, and then embedded in epoxy resin. Finally, the ultrastructures of cells were captured using a TEM (H-600IV, Hitachi Ltd., Japan).

Western blotting
BCA protein assays (Thermo Fisher Scientific) were used to measure protein concentrations. Proteins were separated on 10% SDS-PAGE and then transferred onto a polyvinylidene fluoride membranes (PVDF, Millipore, Billerica, MA, USA). The membranes were then blocked with 5% skim milk for 1 h at room temperature before incubation overnight at 4° C with the primary antibodies, anti-GPX4 (1:1000, Abcam) and anti-GAPDH (1:1000, Abcam). The membranes were then incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody (1:5000, Abcam) for 1 h at room temperature. The bands were visualized using an enhanced Chemiluminescence Detection System (Thermo Fisher Scientific) and quantified using ImageJ.

Quantitative real-time reverse transcriptasepolymerase chain reaction (qRT-PCR)
Total RNA was extracted from cells with TRIzol reagent (Thermo Fisher Scientific), after which cDNA was synthesized using a SuperScript IV Reverse Transcriptase kit (Thermo Fisher Scientific

FISH analysis
FISH analysis was performed as previously described [28]. The oligonucleotide probes were synthesized by Geneseed Biotech (Guangzhou, China). HepG2 cells were fixed for 30 min in 4% paraformaldehyde at room temperature, washed with distilled water, and then dehydrated through an ethanol series (70%, 90%, 100%). The air-dried cells were then incubated with FISH probes in hybridization buffer at 37° C overnight in the dark. Cell nuclei were counterstained for 30 min with DAPI (Thermo Fisher Scientific). Finally, images were captured using a FV10i confocal microscope (Olympus Corp, Tokyo, Japan).

Statistical analysis
All experiments were performed in triplicate. Data are presented as the mean ± standard error (S.E.). Graphs were generated using GraphPad Prism software (version 7.0, La Jolla, CA, USA). One-way analysis of variance (ANOVA) and Tukey"s tests were carried out for AGING multiple group comparisons. Values of P < 0.05 was considered statistically significant.

AUTHOR CONTRIBUTIONS
WY made major contributions to the conception, design and manuscript drafting of this study. JZ and YZ were responsible for the conception and design of this study and performed the experiments. WL, LZ and YR were responsible for data acquisition, data analysis, data interpretation. RO and YX made substantial contributions to conception and design of the study and revised the manuscript. All authors read and approved the final manuscript.