Metformin regulates the LIN28B‑mediated JNK/STAT3 signaling pathway through miR‑140‑3p in subretinal fibrosis

Introduction

Age-related macular degeneration (AMD) is a prime cause of visual impairment and severe vision loss and is one of the top three eye diseases that can result in blindness (1). Subretinal fibrosis (SF) is a pivotal pathologic trait of neovascular AMD (nAMD) (2). The progression of SF can lead to vision loss (3). SF is primarily characterized by an excess of extracellular matrix (ECM) proteins, including collagen and fibronectin (4). SF and ECM components are mainly produced by activated myofibroblasts (5). In retinal diseases, oxidative stress can lead to neurodegeneration and cell loss, which is associated with early disease progression (6). Thus, it is urgently necessary to develop novel therapies for the treatment of SF.

Metformin is a biguanide that has been widely used to treat diabetes since the 1950s (7). In addition to being an antidiabetic drug with low cost and high safety, metformin has also been shown to inhibit the proliferation and migration of human retinal vascular endothelial cells and reduce oxidative stress in cells and various organs (8,9). Furthermore, metformin has been shown to reverse existing pulmonary fibrosis in mice (10) and inhibit the migration and invasion of cancer cells in vitro (11). It has also been shown to have a protective effect against several age-associated diseases, and may prevent the development of AMD (12). Currently, metformin is of interest as a candidate for the treatment of AMD, as it may reduce the progression of AMD via its antioxidant, anti-inflammatory and antifibrotic effects (13). However, the role and mechanism of metformin are intricate, and it is important to determine the mechanism of metformin in the treatment of SF.

MicroRNAs (miRNAs) are small noncoding RNAs that play critical roles in gene regulation and numerous biological processes (14). The development of retinal diseases, including AMD, is associated with the dysregulation of miRNAs and alterations in the mechanisms of miRNA biogenesis (15). Yi et al (16) showed that the overexpression of miR-140-3p significantly alleviated inflammation, oxidative stress and apoptosis in oxygen-glucose deprivation and reperfusion models, and exerted a protective effect on cells. In addition, Al-Modawi et al (17) showed that miR-140-3p prevented oxidative stress by upregulating the enzyme N(G), N(G)-dimethylarginine dimethylaminohydrolase 1. Furthermore, another study showed that the upregulation of miR-140-3p ameliorated pathological changes and inhibited inflammation, oxidative stress and fibrosis in rats with rheumatoid arthritis (18). These findings indicate that miR-140-3p improves various functions of the body by alleviating oxidative stress and fibrosis. However, the regulatory mechanism of miR-140-3p in SF is unknown.

LIN28 is a reprogramming factor and conserved RNA-binding protein, and the only homolog of LIN28 in humans (19). LIN28B regulates cell migration, invasion, apoptosis and fusion (20). In a study by Liang et al (21), LIN28B was shown to induce epithelial-mesenchymal transition (EMT) via the inhibition of let-7d, and the inhibition of LIN28B alleviated TGF-β1-induced fibrosis. In another study, Zhang and Sui (22) showed that the knockdown of circBPTF mediated the miR-384/LIN28B axis in human umbilical vein endothelial cells, thereby preventing the inflammatory injury and oxidative stress induced by high glucose. Thus, as LIN28B is indicated to regulate fibrosis, the present study investigated the effect of LIN28B on SF.

JNK, which is a Ser/Thr kinase, is a member of the mitogen-activated protein kinase family in mammals. JNK influences multiple cellular processes, including cell proliferation, survival and malignant transformation (23). STAT3 is a cytoplasmic transcription factor that plays a pivotal role in gene expression and is involved in proliferation and survival (24). Yang et al (25) showed that JNK enhanced the transcription of TGF-β1 and connective tissue growth factor and promoted fibrosis in models of acute kidney injury, while the inhibition of JNK activity protected the kidney from the development of fibrosis. Du et al (26) demonstrated that the knockdown of atypical protein kinase C-i inhibited the EMT, migration and invasion of colorectal cancer cells by inhibiting the Rac1-JNK pathway. In addition, Zhao et al (27) found that STAT3 was intimately associated with the occurrence and development of liver fibrosis induced by multiple factors, and suggested that STAT3 can play an anti-inflammatory or proinflammatory role in the pathogenesis of liver fibrosis. Zhao et al (28) also showed that the JNK/STAT3 signaling pathway is involved in the oxidative stress and apoptosis induced by excessive fluoride in female mice, which reduced the development of potential oocytes. These studies offer important information to support exploration of the role of the JNK/STAT3 pathway in fibrotic diseases and serve as a reference for further study of the mechanism of the JNK/STAT3 pathway in fibrotic diseases.

Therefore, the present study examined the role and effect of metformin in SF, and investigated whether its mechanism involves regulation of the JNK/STAT3 signaling pathway mediated by LIN28B through miR-140-3p. The study may constitute an academic reference for the clinical treatment of SF.

Materials and methods

Establishment and grouping of the animal models

A total of 40 male SPF C57BL/6J mice (18-22 g, 6-8 weeks old) were obtained from the Experimental Animal Center of Yunnan University. The license number for the use of experimental animals was SCXK (Dian) K2021-0002. All procedures were approved by the Experimental Animal Welfare Ethics Committee of the Experimental Animal Center of Yunnan University (Kunming, China; ethics no. YNU20220291). Mice were housed individually using a 12-h light/dark cycle at 22˚C with 50% humidity, with food and water available at will, and were domesticated and housed for 1 week prior to the experiment. The mice were randomly divided into 4 groups (n=10/group): Normal control group (mice without laser irradiation), SF group, SF + metformin (SF + Met) group and SF + metformin + miR-140-3p inhibitor group. The mice were anesthetized with pentobarbital sodium (30 mg/kg, intraperitoneal injection), and the laser-induced SF model was examined for 35 days as previously described (29). Briefly, on day 0, the mice were anesthetized with pentobarbital, and 4-6 laser spots (532 nm, 200 mW, 100 msec, 75 µm; VISULAS® 532s; Zeiss AG) were created in each fundus around the optic disc. Immediately after laser irradiation, a subretinal bubble formed, confirming rupture of the Bruch's membrane. Mice in the two metformin treatment groups were treated with 300 mg/kg/day metformin (Beijing Solarbio Science & Technology Co., Ltd.) by gavage from day 0 to day 35. In addition, the mice in the SF + metformin + miR-140-3p inhibitor group were injected with 5 µl miR-140-3p inhibitor lentivirus (Hunan Fenghui Biotechnology Co., Ltd., China) through the vitreous cavity on day 0 of laser-induced injury. To administer the intravitreal injection, a 30-gauge needle was first inserted behind the eye margin to make a cavity. The homologous compounds were injected into the incipient well cavities with a 34-gauge Hamilton syringe, and all injections were performed under an operating microscope. Animal health and behavior were monitored every 2 days.

The mice were euthanized on day 35 by decapitation using scissors, and were confirmed dead after decapitation by the loss of heartbeat and breath cessation. Eyeball mouse tissues were collected and fixed in 5% paraformaldehyde for 1 h. The conditions for euthanasia were as follows: i) the end of the animal experiment; ii) the animals were in pain beyond the prespecified mercy end point (such as marked reduction in pre-test body weight, markedly coarse fur, accompanied by unresponsive and behavioral abnormalities, and persistent dyspnea). In addition, meloxicam (1 mg/kg, intragastric administration) was used as an analgesic to minimize the pain and stress of the animals during the experiment.

Cell culture and model construction

The ARPE-19 adult retinal pigment epithelium (RPE) cell line was purchased from Shenzhen Otwo Biotechnology Co., Ltd. and cultured in DMEM/F-12 (MilliporeSigma) containing 10% FBS (MilliporeSigma), 100 U/ml penicillin and 100 µg/ml streptomycin at 37˚C with 5% CO2. The cultured ARPE-19 cells were treated with 10 ng/ml TGF-β1 for 24 and 48 h to establish the fibrotic cell model (TGF-β1 group), and the cells were also cultured with 10 ng/ml TGF-β1 and 2 mM metformin (TGF-β1 + Met group), or 0 ng/ml TGF-β1, 2 mM metformin and 4 µM anisomycin (co-treatment) (TGF-β1 + Met + anisomycin group) for 48 h in subsequent experiments.

Cell transfection

The negative control (NC) mimic, green fluorescent protein miR-140a-3p mimic, miR-140a-3p inhibitor and NC inhibitor were obtained from Sangon Biotech Co., Ltd. ARPE-19 cells were transfected with Lipofectamine™ 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) and the aforementioned nucleic acids (300 ng/µl) for 24 h at 37˚C. The transfection efficiency of the miR-140a-3p inhibitor was determined by reverse transcription-quantitative PCR (RT-qPCR) after 36 h. The transfection efficiency of the miR-140a-3p mimic was determined by immunofluorescence using a fluorescence microscope (Leica Microsystems, GmbH) after 36 h.

The lentiviral LIN28B (3rd generation) was obtained from Shanghai GenePharma Co., Ltd. The 293T cell (The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences Cell Bank) used to generate the virions, Packaging plasmids (pMDL, VSVG, REV) were obtained from Shanghai GenePharma Co., Ltd. A mixture comprising 1.5 µg LIN28B plasmid and 1.5 µg packaging plasmid (pMDL:VSVG:REV=5:3:2) was prepared and then added to the culture dish containing 293T cells for 4-6 h (37˚C, 5% CO2). The transfection solution was absorbed and discarded, and fresh culture solution was added for 72 h. Following filtration through a 0.45-µm filter, the virus was collected by ultracentrifugation (4˚C, 6,000 x g, 2 h), and the titers were determined by a limited dilution method. One day prior to the experiment, 4x103 ARPE19 cells were inoculated into each well of a 96-well culture plate, and 10 µl 1x108 TU/ml virus and 5 µg/ml Polybrene were added to each well (multiplicity of infection, 10) for incubation at 37˚C for 24 h. Fluorescence was observed after 72 h of infection. Empty vector was also transfected as the control. The reagents used for lentiviral transfection were provided by Shanghai GenePharma Co., Ltd. The sequences of the NC mimic, miR-140-3p mimic, miR-140-3p inhibitor, NC inhibitor and LIN28B are presented in Table SI.

Cell Counting Kit-8 (CCK-8) analysis of cell proliferation

ARPE-19 cells (5x103 cells/well) were inoculated in a 96-well plate and cultured for 24 h at 37˚C in a 5% CO2 incubator. Following the relevant treatment, 20 µl CCK-8 reagent (Beijing Solarbio Science & Technology Co., Ltd.) was added to the cells in each well. After 2 h of incubation, the absorbance was measured at 450 nm with a microplate reader.

Reactive oxygen species (ROS) detection

ARPE-19 cells were seeded in 6-well plates at 2x105 cells/well and incubated with the ROS-specific fluorescent dye dichloro-dihydro-fluorescein diacetate at 37˚C for 0.5 h in the dark. The cells were washed with PBS to remove the unbound dye, digested and then resuspended in 0.5 ml PBS. Finally, ROS levels were determined by FACSCalibur flow cytometry (BD Biosciences) and analyzed with FlowJ software (v10.8.1; BD Biosciences).

The retinal and choroidal tissues of C57BL/6J mice were immediately placed into precooled PBS solution to remove the blood and other pollutants. The tissue was cut into 1-mm3 pieces with ophthalmic scissors and rinsed in precooled PBS to remove the cell fragments. Enzyme digestion solution was added, and digestion was performed in a 37˚C constant temperature water bath for 30 min with intermittent oscillation. Digestion was then stopped with precooled PBS, tissue pellets were removed with a 300-mesh nylon mesh filter, and the filtered cells were collected to examine the ROS levels using the aforementioned method.

Western blot analysis

Proteins were isolated from cells or retinochoroidal tissue with RIPA buffer (Beyotime Institute of Biotechnology) containing 1% protease inhibitors. Protein concentrations were determined using a bicinchoninic acid assay (Beijing Solarbio Science & Technology Co., Ltd.,) according to the specifications of the kit. A 50-µg quantity of total protein was loaded per lane, and the proteins were separated by SDS-PAGE (5% stacking gel and 10% separating gel), transferred to PVDF membranes (MilliporeSigma) and then blocked with 5% non-fat milk powder for 1.5 h at room temperature. The following diluted primary antibodies were added and incubated overnight at 4˚C: E-cadherin (1:1,000; ab231303), vimentin (1:2,000; ab92547), fibronectin (1:3,000; ab2413), N-cadherin (1:5,000; ab76011), JNK (1:1,000; ab76125), phosphorylated (p)-JNK (1:1,000; ab124965), STAT3 (1:1,000; ab68153), p-STAT3 (1:1,000; ab267373), collagen I (1:5,000; ab138492), collagen III (1:1,000; ab184993), LIN28B (1:2,000; ab191881) and β-actin (1:1,000; ab8226) from Abcam and α-smooth muscle actin (SMA; 1:1,000; #192455) Cell Signaling Technology, Inc.), Next, the membranes were incubated with goat anti-rabbit IgG H&L HRP-conjugated secondary antibodies (1:4,000; ab97051; Abcam) for 1 h at room temperature and developed with an ECL kit (Abcam). Finally, the bands were semiquantitatively analyzed using ImageJ 1.52a software (National Institutes of Health).

Superoxide dismutase (SOD), malondialdehyde (MDA), glutathione peroxidase (GSH-PX) and catalase (CAT) assays

An SOD kit (cat. no. BC0170), GSH-PX kit (cat. no. BC1195) and CAT kit (cat. no. BC0200) all from Beijing Solarbio Science & Technology Co., Ltd., and MDA kit (Nanjing Jiancheng Bioengineering Institute), were used to detect the levels of SOD, MDA, GSH-PX and CAT in cells and retinal choroid tissue after processing according to the instructions of the kit manufacturers.

Detection of apoptosis by flow cytometry

Cells in each group were digested with trypsin and rinsed twice with precooled PBS. The apoptosis rate was then determined using an Annexin-V-FITC/PI apoptosis kit (Absin Bioscience, Inc.).

Scratch test

ARPE-19 cells in the logarithmic growth phase were inoculated on 6-well plates after normal digestion and passage. The cells were scratched with a pipette tip when the cell density reached 90%. The cells were not serum starved, as 1% FBS medium was used to culture with the cells. After 0 and 24 h of culture, cell migration was observed under a light microscope and photographic images were captured.

Transwell assay of cell invasion

Cell invasion experiments were performed using a Transwell chamber (Corning, Inc.). The cell concentration was adjusted to 1x105 cells/ml with serum-free DMEM. Then, 200 µl cell suspension was added to the upper chamber of the Transwell chamber, which was coated with Matrigel (3 h, 37˚C) and 600 µl DMEM containing 10% FBS was added to the lower chamber of the 24-well plate. The cells in the lower chamber were immobilized with 4% paraformaldehyde and stained with crystal violet (Beijing Solarbio Science & Technology Co., Ltd.) after 24 h (37˚C) of incubation. The cells were placed under an inverted light microscope, the number of cells at a fixed position in each well were observed, and 5 visual fields were selected for photography and counting.

Coimmunoprecipitation

Cultured cells were homogenized and lysed for 15 min on ice with a mixture of protease inhibitors (Beyotime Institute of Biotechnology) in hypotonic lysis buffer (Beyotime Institute of Biotechnology). The cell lysates were centrifuged at 4,000 x g for 10 min at 4˚C. Following the removal of a small amount of lysate for input western blot analysis, 1 µg corresponding antibody, anti-Lin28B (1:100; cat. no. ab191881; Abcam) was added to the rest, and the sample was incubated overnight at 4˚C with slow shaking. Cell lysates containing 10 µl protein A agarose beads (Beyotime Institute of Biotechnology) and antibodies were incubated overnight at 4˚C. The beads were collected by centrifugation (1,000 x g, 3 min, 4˚C) and washed three times with lysis buffer (Beyotime Institute of Biotechnology). Protein elution was followed by immunoblot analysis.

Dual-luciferase assay

The target binding sites of miR-140-3p and LIN28B were predicted by a bioinformatics database (http://starbase.sysu.edu.cn/). Luciferase reporter plasmid (pGL3-Basic/LIN28B WT, pGL3-Basic/LIN28B MUT) was purchased form Shanghai GenePharma Co., Ltd. The LIN28B 3'-untranslated region containing the miR-140-3p binding site was cloned into a pGL3 vector (Promega Corporation) to establish the wild-type LIN28B vector (WT). A site-directed mutagenesis kit was used to generate a mutant LIN28B vector (MUT). WT or MUT was cotransfected with a vector overexpressing miR-140-3p or NC mimic into 293T cells using Lipofectamine™ 3000 (Invitrogen; Thermo Fisher Scientific, Inc.). The luciferase activity was determined using a Promega luciferase reporter system after 48 h, compared with Renilla luciferase activity. The Dual-Lumi™ dual luciferase reporter gene detection kit (cat. no. RG088S; Beyotime Institute of Biotechnology) was used, and the experiment was conducted following the manufacturer's instructions.

RT-qPCR

Total RNA was isolated from eye tissues and cells with TRIzol® reagent (cat. no. 15596026; Invitrogen; Thermo Fisher Scientific, Inc.), and cDNA was synthesized by reverse transcription using a First-Strand cDNA synthesis kit (GeneCopoeia, Inc.), with the following temperature protocol: 25˚C for 5 min, 42˚C for 15 min, 85˚C for 5 min and hold at 4˚C. RT-qPCR was performed using a SYBR Green real-time fluorescence quantitative PCR kit (Beijing Solarbio Science & Technology Co., Ltd.). The conditions of PCR were as follows: 95˚C predenaturation for 30 sec; 40 cycles of 95˚C for 15 sec and 60˚C for 30 sec; and 72˚C extension for 30 sec. The internal reference genes were U6 for miR-140-3p and GAPDH for the other RNAs. Relative expression of miRNA was analyzed by the 2-ΔΔCq method (30). The primer sequences are shown in Table I.

Table I Primer sequences.

Table I

Primer sequences.

Target	Sequences (5'-3')
miR-140-3p	F: GCGCGTACCACAGGGTAGAA
	R: AGTGCAGGGTCCGAGGTATT
hsa-LIN28B	F: TAGGAAGTGAAAGAAGACCCAA
	R: CTGAGGAAACGGTGGTGA
mmu-LIN28B	F: TGTGGACTGTGCGAGAAGAAGA
	R: CCTGTCTGAGTGCTCTGCCATT
U6	F: GCTTCGGCAGCACATATACTAAAAT
	R: CGCTTCACGAATTTGCGTGTCAT
β-actin	F: CACTGTGCCCATCTACGAGG
	R: TAATGTCACGCACGATTTCC

[i] F, forward; R, reverse; miR, microRNA; hsa, Homo sapiens; mmu, Mus musculus.

RNA pull-down

Using a Magnetic RNA Protein Pull Down Kit (Thermo Fisher Scientific, Inc.), an RNA pull-down assay was performed in accordance with the manufacturer's instructions. In brief, RNA probes and miR-140-3p were incubated with cell lysate at room temperature. Then magnetic beads were added to form a compound with the probe, which was detected by immunoprecipitation, washing, purification and western blotting.

Hematoxylin and eosin (H&E) staining

After 35 days of treatment, the eyeballs were isolated, enucleated and fixed in FAS eyeball fixative (Wuhan Servicebio Technology Co., Ltd.) for >48 h and then dehydrated and embedded in paraffin. Sections (20-µm thick) were dewaxed, stained with hematoxylin for 5 min (at room temperature), washed with tap water, stained with eosin for 1 min (at room temperature) and then dehydrated before sealing with neutral gum for observation (light microscopy) and analysis.

Masson staining

Paraffin sections (20-µm thick) were dewaxed, hydrated with gradient ethanol, stained with hematoxylin for 5 min, and fully washed with water. The sections were then washed with Masson ponceau acid fuchsin solution for 5 min, soaked in 2% glacial acetic acid water, differentiated with 1% phosphomolybdic acid aqueous solution for 3 min without washing and then directly dyed with aniline blue or light green solution for 5 min. After washing with 0.2% glacial acetic acid aqueous solution for 5 sec, 95% alcohol, absolute alcohol, xylene and transparent neutral gum were applied (at room temperature). A light microscope was used to observe and images were captured.

Statistical analysis

GraphPad Prism 8.0 (GraphPad Software; Dotmatics) was used to analyze the experimental data and plot the graphs. The data are from ≥3 replicates of all experiments. One-way ANOVA was used for analysis followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant result.

Results

Metformin inhibits laser-induced SF and oxidative stress in mice

To examine the effect of metformin on laser-induced SF in mice, the pathology of optic cup tissue was observed using H&E staining. The results showed that compared with the control group, the SF group had more subretinal fibers and fibrocytes and a thickened choroid. However, the subretinal fibers and choroidal thickness were reduced by metformin (Fig. 1A). Masson staining of the optic cup tissue showed that the amount of collagen fiber synthesis in the SF group was markedly higher than that in the control group, and metformin prominently reduced the synthesis of collagen fibers (Fig. 1B). Western blot analysis showed that the levels of α-SMA, collagen I, collagen III, vimentin, fibronectin and N-cadherin were upregulated in the SF group in comparison with those in the control group, while E-cadherin expression was downregulated. Metformin reversed the effects of SF on these proteins (Fig. 1C and D). Markers of oxidative stress were examined, and the results showed that, in comparison with those in the control group, SOD and GSH-PX activity and CAT content in the SF group were prominently reduced, while the MDA level was increased. The effects in the SF group were reversed by metformin (Fig. 1E). These results indicate that metformin inhibited EMT and fibrosis-associated protein expression, increased the levels of SOD, GSH-PX and CAT, inhibited MDA formation, and alleviated the progression of SF in the mice.

Figure 1 Metformin inhibits the progression of laser-induced subretinal fibrosis in mice. (A) Hematoxylin and eosin staining and (B) Masson staining were used to examine optic cup tissue (scale bar, 100 µm). Western blotting was performed to examine the levels of (C) fibrosis- and (D) epithelial-mesenchymal transition-associated proteins. (E) Kits were used to examine the levels of SOD, MDA, GSH-PX and CAT. **P<0.01, ***P<0.001. Statistical analysis was by one-way ANOVA followed by Tukey's post hoc tests. SOD, superoxide dismutase; MDA, malondialdehyde; GSH-PX, glutathione peroxidase; CAT, catalase; SF, subretinal fibrosis; Met, metformin; SMA, smooth muscle actin; prot, protein.

Metformin inhibits EMT in RPE cells through the JNK/STAT3 signaling pathway

To examine if the effects of metformin are achieved through the JNK/STAT3 signaling pathway, cell proliferation was examined by CCK-8 assay and ROS levels were measured using flow cytometry. The results showed that cell proliferation and ROS levels were increased in the TGF-β1 group compared with the NC group. Metformin reversed the effect of TGF-β1, and anisomycin, an activator of the JNK/STAT3 pathway, attenuated the effect of metformin in both assays (Fig. 2A and B). The oxidative stress level was also examined, and the results showed that in comparison with those in the untreated control group, the activity of SOD and GSH-PX and the CAT content were reduced, and the levels of MDA were increased in the TGF-β1 group. The effect of TGF-β1 was reversed by metformin, while the effect of metformin was attenuated by anisomycin (Fig. 2C). In addition, the results of scratch and Transwell assays showed that the migration and invasion of cells in the TGF-β1 group were markedly elevated compared with those in the NC group, the effect of TGF-β1 was reversed by metformin, while the effect of metformin was attenuated by anisomycin (Fig. 2D and E). Furthermore, western blotting showed that TGF-β1 downregulated the level of E-cadherin and upregulated the levels of vimentin, fibronectin, N-cadherin, p-STAT3/STAT3, p-JNK/JNK, α-SMA, collagen I and collagen III. The effect of TGF-β1 was reversed by the addition of metformin, and the effect of metformin was attenuated (Fig. 2F-H). These results indicate that the effect of metformin on retinal fibrosis and oxidative stress may involve the JNK/STAT3 pathway.

Figure 2 Metformin inhibits fibrosis in retinal pigment epithelial cells through the JNK/STAT3 signaling pathway. (A) A Cell Counting Kit-8 assay was used to examine cell proliferation. (B) ROS levels were detected by flow cytometry. (C) SOD, MDA, GSH-PX and CAT levels in the cells were detected using kits. (D) Scratch tests were used to examine the migration of the cells (scale bar, 100 µm). (E) Cell invasion was measured using Transwell assays (scale bar, 100 µm). The expression of (F) epithelial-mesenchymal transition-associated proteins, (G) JNK/STAT3 pathway-associated proteins and (H) fibrosis-related proteins. *P<0.05, **P<0.01, ***P<0.001. Statistical analysis was by one-way ANOVA followed by Tukey's post hoc tests. ROS, reactive oxygen species; SOD, superoxide dismutase; MDA, malondialdehyde; GSH-PX, glutathione peroxidase; CAT, catalase; NC, negative control; Met, metformin; OD, optical density; p-, phosphorylated; SMA, smooth muscle actin.

Metformin inhibits JNK/STAT3 signaling pathway-mediated EMT in RPE cells via miR-140-3p

The levels of miR-140-3p in the RPE cells were measured by RT-qPCR, and the results showed that metformin promoted miR-140-3p expression in the presence of TGB-β1 (Fig. 3A). Furthermore, the knockdown of miR-140-3p was performed to explore the mechanism of miRNA in the cell model (Fig. 3B). In comparison with that in the TGF-β1 group, the cell proliferation activity and ROS levels were decreased in the TGF-β1 + Met group. The effect of TGF-β1 + Met was attenuated by the miR-140-3p inhibitor (Fig. 3C and D). Compared with those in the TGF-β1 group, the activity of SOD and GSH-PX and the CAT content were increased, while the MDA levels were decreased in the TGF-β1 + Met group. The addition of the miR-140-3p inhibitor attenuated the effect of metformin (Fig. 3E). The scratch assay and Transwell results showed that in comparison with those in the NC group, the migration and invasion of cells in the TGF-β1 group were markedly increased, and the effect of TGF-β1 was reversed by metformin. The effect observed in the TGF-β1 + Met group was attenuated by the miR-140-3p inhibitor (Fig. 3F and G). Western blotting showed that in comparison with those in the NC group, the levels of E-cadherin were downregulated and the levels of vimentin, fibronectin, N-cadherin, p-STAT3/STAT3, p-JNK/JNK, α-SMA, collagen I and collagen III were upregulated in the TGF-β1 group, and the effect of TGF-β1 was reversed by metformin. Following transfection of the miR-140-3p inhibitor, the effect of metformin in the TGF-β1 + Met group was reversed (Fig. 3H-J). In summary, these results indicate that metformin inhibits the JNK/STAT3 signaling pathway by promoting miR-140-3p expression, thereby inhibiting RPE cell fibrosis and oxidative stress.

Figure 3 Metformin inhibits JNK/STAT3 signaling pathway-mediated fibrosis in retinal pigment epithelial cells through miR-140-3p. Reverse transcription-quantitative PCR was used to examine miR-140-3p expression in (A) the NC, TBF-b1 and TBF-b1 + Met groups and (B) cells transfected with miR-140-3p inhibitor or NC inhibitor. (C) A Cell Counting Kit-8 assay was used to examine cell proliferation. (D) ROS levels were detected by flow cytometry. (E) SOD, MDA, GSH-PX and CAT levels in the cells were detected using kits. (F) Scratch tests were used to examine cell migration (scale bar, 100 µm). (G) Transwell assays were used to examine cell invasion (scale bar, 100 µm). Western blot analysis of proteins associated with (H) epithelial-mesenchymal transition, (I) the JNK/STAT3 pathway and (J) fibrosis. *P<0.05, **P<0.01, ***P<0.001 Statistical analysis was performed by one-way ANOVA followed by Tukey's post hoc tests. NC, negative control; Met, metformin; miR, microRNA; inhibitor, miR-140-3p inhibitor; ROS, reactive oxygen species; SOD, superoxide dismutase; MDA, malondialdehyde; GSH-PX, glutathione peroxidase; CAT, catalase; OD, optical density; p-, phosphorylated; SMA, smooth muscle actin.

miR-140-3p affects JNK/STAT3 signaling through LIN28B

To further investigate the downstream regulatory mechanism of miR-140-3p, the cells were transfected with miR-140-3p mimic to explore the mechanism of this miRNA in ARPE19 cells. The green fluorescence of the miR-140-3p mimic showed that miR-140-3p was successfully transfected (Fig. 4A). The results of an RNA pull-down assay show that miR-140-3p bonds with LIN28B (Fig. 4B). The binding sites of LIN2B targeted by miR-140-3p were predicted using a bioinformatics website. A dual-luciferase reporter assay confirmed that miR-140-3p downregulated LIN28B (Fig. 4C). To verify whether there was an interactive relationship between LIN28B and JNK, coimmunoprecipitation experiments were performed. The results showed that LIN28B and JNK proteins were precipitated in cells incubated with LIN28B antibody compared with those incubated with IgG antibody (Fig. 4D). RT-qPCR and western blotting results showed transfection with miR-140-3p mimic reduced Lin28B expression (Fig. 4E and F).

Figure 4 miR-140-3p affects the JNK/STAT3 signaling pathway through LIN28B. (A) Immunofluorescence was used to confirm transfection with miR-140-3p mimic (scale bar, 100 µm). (B) RNA pull-down assay confirmed the targeting relationship between miR-140-3p and LIN28B. (C) StarBase was used to predict the sequences of the binding sites of miR-140-3p and LIN28B, and a dual-luciferase assay was performed to confirm the targeting relationship between miR-140-3p and LIN28B. (D) The interaction between LIN28B and JNK was verified by coimmunoprecipitation assay. (E) Reverse transcription-quantitative PCR and (F) western blot analysis of the expression of LIN28B. **P<0.01, ***P<0.001. Statistical analysis was performed by (C) two-way ANOVA and (E and F) one-way ANOVA followed by Tukey's post hoc tests. miR, microRNA; NC, negative control; WT, wild type; MUT, mutant; IP, immunoprecipitation; IB, immunoblotting; mimic, miR-140-3p mimic.

miR-140-3p affects TGF-β1-induced fibrosis in RPE cells through the JNK/STAT3 signaling pathway via LIN28B

ARPE19 cells were transfected with LIN28B overexpression vector, and western blotting results showed that the overexpression of LIN28B was successful (Fig. 5A). In comparison with those in the TGF-β1 group, the miR-140-3p mimic inhibited cell viability and decreased ROS levels, while LIN28B overexpression attenuated the effect of the mimic in the TGF-β1 + miR-140-3p mimic group (Fig. 5B and C). In comparison with those in the TGF-β1 group, the activity of SOD and GSH-PX and the content of CAT were increased, while the levels of MDA were decreased in the cells treated with TGF-β1 and miR-140-3p mimic. The overexpression of LIN28B reversed the effect of the mimic in the TGF-β1 + miR-140-3p mimic group (Fig. 5D). As before, the scratch assay and Transwell results showed that cell migration and invasion were increased in the TGF-β1 group compared with the NC group. Transfection with miR-140-3p mimic reversed the effect of TGF-β1, while the overexpression of LIN28B attenuated the effect of the mimic in the TGF-β1 + miR-140-3p mimic group (Fig. 5E and F). Western blotting results indicated that the miR-140-3p mimic inhibited EMT, activation of the JNK/STAT3 pathway and the progression of fibrosis, and the overexpression of LIN28B attenuated these effects (Fig. 5G-I). These results show that overexpression of miR-140-3p inhibited the JNK/STAT3 pathway, thereby inhibiting EMT, the JNK/STAT3 pathway, fibrosis-associated proteins, oxidative stress, cell migration and proliferation, and these effects were inhibited by the overexpression of LIN28B.

Figure 5 miR-140-3p affects TGF-β1-induced retinal pigment epithelial cell fibrosis through the JNK/STAT3 signaling pathway via LIN28B. (A) Expression of LIN28B in transfected cells was measured by western blotting. (B) Cell Counting Kit-8 analysis of cell proliferation. (C) Flow cytometric analysis of ROS levels. (D) SOD, MDA, GSH-PX and CAT levels were detected using kits. (E) Scratch tests were used to examine cell migration (scale bar, 100 µm). (F) Transwell assays were used to examine cell invasion (scale bar, 100 µm). Western blot analysis of proteins associated with (G) epithelial-mesenchymal transition, (H) the JNK/STAT3 signaling pathway and (I) fibrosis. *P<0.05, **P<0.01, ***P<0.001. Statistical analysis was performed by one-way ANOVA followed by Tukey's post hoc tests. miR, microRNA; ROS, reactive oxygen species; SOD, superoxide dismutase; MDA, malondialdehyde; GSH-PX, glutathione peroxidase; CAT, catalase; NC, negative control; OE, overexpression; mimic, miR-140-3p mimic; OD, optical density; p-, phosphorylated; SMA, smooth muscle actin.

Metformin affects SF through miR-140-3p in vivo

Experiments were performed to confirm that the effect of metformin on SF was mediated via miR-140-3p in mice. The RT-qPCR results showed that the miR-140-3p inhibitor was successfully transfected in the mouse eye, and that metformin promoted miR-140-3p expression in SF model mice. In addition, the expression of miR-140-3p was markedly decreased by the miR-140-3p inhibitor (Fig. 6A and B). Western blot analysis showed that the expression levels of α-SMA, collagen I, collagen III, vimentin, fibronectin and N-cadherin were upregulated in the SF group, while E-cadherin expression was downregulated. Metformin reversed the effect of SF on these proteins, while the knockdown of miR-140-3p attenuated the effect of metformin in the the SF + Met group (Fig. 6C and D). H&E staining was performed to observe the pathology of the optic cup tissue of the mice. The results showed that compared with the control group, the SF group had greater number of subretinal fibers and fibrocytes, and a thickened choroid. The number of subretinal fibers and choroidal thickness were reduced by metformin. Following the knockdown of miR-140-3p, the formation of subretinal fibers was increased compared with that in the SF + Met group (Fig. 6E). Masson staining of the optic cup tissue showed that the synthesis of collagen fibers in the SF group was markedly higher than that in the control and SF + Met groups, and metformin notably reduced the synthesis of collagen fibers. After miR-140-3p knockdown, collagen fiber synthesis was markedly elevated compared with that in the SF + Met group (Fig. 6F). These results suggest that metformin induced miR-140-3p expression, inhibited EMT -and fibrosis-associated protein expression, thereby alleviating SF in mice.

Figure 6 Experiments using mice verified that metformin affects subretinal fibrosis through miR-140-3p. Reverse transcription-quantitative PCR was used to measure the level of miR-140-3p to (A) confirm transfection with miR-140-3p inhibitor and (B) compare its expression in different groups. Western blot analysis of the expression levels of proteins associated with (C) fibrosis and (D) epithelial-mesenchymal transition. (E) Hematoxylin and eosin staining was used to observe pathological changes in optic cup tissue (scale bar, 100 µm). (F) Masson staining of optic cup tissue (scale bar, 100 µm). *P<0.05, **P<0.01, ***P<0.001. Statistical analysis was performed by one-way ANOVA followed by Tukey's post hoc tests. miR, microRNA; NC, negative control; inhibitor, miR-140-3p inhibitor; SF, subretinal fibrosis; Met, metformin; SMA, smooth muscle actin.

Metformin affects the JNK signaling pathway and oxidative stress in SF through miR-140-3p in vivo

Whether the effect of metformin on the JNK signaling pathway and oxidative stress in SF is mediated via miR-140-3p was examined in vivo. The RT-qPCR results showed that metformin inhibited LIN28B expression, and the expression level of LIN28B was increased by miR-140-3p inhibitor treatment (Fig. 7A). Western blot analysis showed that p-STAT3 and p-JNK were upregulated in the SF group compared with the control group, and metformin attenuated STAT3 and JNK phosphorylation levels compared with those in the SF group. In addition, the knockdown of miR-140-3p reversed the effect of metformin in the SF + Met group (Fig. 7B). Flow cytometry results showed that metformin reversed the effect of SF modeling on ROS levels, while the knockdown of miR-140-3p reversed the effect of metformin in the SF + Met group (Fig. 7C). In addition, the activity of SOD and GSH-PX and the content of CAT were reduced while the levels of MDA were increased in the SF group compared with the control group. The effects observed in the SF group were reversed by metformin, and those observed the SF +Met group were attenuated by the knockdown of miR-140-3p (Fig. 7D). These results indicate that metformin inhibited the JNK signaling pathway and oxidative stress via the inhibition of LIN28B expression, and the changes induced by metformin were reversed by knocking down miR-140-3p.

Figure 7 Experiments using mice verified that metformin affects the JNK signaling pathway and oxidative stress levels in subretinal fibrosis through miR-140-3p. (A) Reverse transcription-quantitative PCR was used to examine the level of LIN28B in the different treatment groups. (B) Western blot analysis of the levels of JNK/STAT3 pathway-associated proteins. (C) Levels of ROS were detected by flow cytometry. (D) Kits were used to examine the levels of SOD, MDA, GSH-PX and CAT. *P<0.05, ***P<0.001. Statistical analysis was performed by one-way ANOVA followed by Tukey's post hoc tests. Inhibitor, miR-140-3p inhibitor; miR, microRNA; OD, superoxide dismutase; MDA, malondialdehyde; GSH-PX, glutathione peroxidase; CAT, catalase; SF, subretinal fibrosis; Met, metformin; p-, phosphorylated; ROS, reactive oxygen species; prot, protein.

Discussion

Fibrosis is a pathologic trait of most chronic inflammatory diseases, which affects almost all tissues and can eventually lead to organ dysfunction and death (31). SF is a vascularized lesion with abundant immune cells, myofibroblasts, ECM protein sedimentation and EMT (6,32,33). SF damages photoreceptors, the RPE and choroidal capillaries, leading to irreversible loss of central vision (5). To date, no effective antifibrotic therapy to reduce SF formation in patients with nAMD has been discovered (5). Therefore, it is imperative to further study the pathogenesis of SF and identify novel and more effective treatments for SF.

Metformin, which is a potent AMP-activated protein kinase (AMPK) activator, has emerged as a promising agent for the reduction or reversal of fibrosis (34). This agent also has antioxidant and anti-inflammatory effects (35). Kheirollahi et al (36) showed that metformin accelerated the regression of fibrosis in the lung by inducing the transdifferentiation of myofibroblasts into adipofibroblasts. Metformin has also been shown to attenuate renal fibrosis induced by unilateral ureteral obstruction in AMPKα2-deficient mice (37). In the present study, a laser-induced SF model was established in mice and the optic cup tissue of the mice was examined by H&E and Masson staining. The results indicated that metformin effectively alleviated SF in mice.

JNK and STAT3 mediate cell transformation, proliferation, survival and migration (23,38). The JNK/STAT3 pathway has been reported to contribute to numerous diseases; for example, Wang et al (39) showed that the JNK/STAT3 pathway is involved in the formation of ECM in scar tissue after tendon injury. In addition, Yan et al (40) and Zhao et al (28) showed that the JNK/STAT3 pathway plays a role in the regulation of oxidative stress and apoptosis. Furthermore, Dong et al (41) showed that metformin significantly attenuated the fibrotic activity of hepatic stellate cells induced by injury in hepatocellular carcinoma cells by targeting JNK in vivo. In the present study, the results showed that JNK/STAT3 pathway proteins were highly activated in SF mouse and cell models, suggesting that the JNK/STAT3 pathway may be involved in the occurrence and development of SF. Further experiments indicated that metformin inhibited EMT and fibrosis-associated protein expression, cell migration, invasion and proliferation, and reduced ROS and MDA levels by inhibiting the JNK/STAT3 signaling pathway. The levels of SOD, GSH-PX and CAT were also increased, which may be associated with the inhibition of fibrosis in RPE cells.

miRNAs have been shown to regulate organ fibrosis by modulating the activity of relevant signaling pathways (42). For example, it has been shown that miR-144-3p and miR-328 are involved in the regulation of cardiac fibrosis, and miR-34a and miR-17-5p contribute to the development of liver fibrosis under different conditions (42). Wu et al (18) indicated that miR-140-3p is a mediator of hepatic stellate cell proliferation, and its knockdown inhibited the TGF-β1-induced proliferation and fibrosis of HSC-T6 cells. In addition, Zhang et al (43) showed that the overexpression of miR-140-3p inhibited EMT, invasion and metastasis in hepatocellular carcinoma. In the present study, it was found that miR-140-3p was expressed at low levels in SF mouse and cell models, and the expression level of miR-140-3p was markedly elevated after metformin treatment. The upregulation of miR-140-3p expression by metformin was associated with inhibition of JNK/STAT3 pathway activation and EMT- and fibrosis-associated protein expression, increases in the levels of SOD, GSH-PX and CAT, and inhibition of MDA production and SF in cells and animals.

LIN28 is a highly conserved RNA-binding protein that has two homologs: LIN28A and LIN28B. Lin28B deficiency has been shown to increase let-7a/let-7b expression and decrease hepatic stellate cell activation and liver fibrosis in mice with alcoholic liver injury (44). In addition, a study by Lu et al (45) showed that the overexpression of LIN28B activated the STAT3 signaling pathway in lung cancer, thus promoting EMT and accelerating migration and invasion. In the present study, the relationship between LIN28B and the JNK/STAT3 pathway was examined. The results demonstrated an interaction between LIN28B and JNK. Overexpression of LIN28B was also shown to upregulate the phosphorylation of JNK/STAT3 pathway-related proteins and induce activation of the JNK/STAT3 pathway. Moreover, the results of the double luciferase experiment indicated that miR-140-3p targets LIN28B.

In summary, the present study indicated that metformin inhibits SF by facilitating miR-140-3p expression and inhibiting LIN28B and the JNK/STAT3 pathway at the cellular and organismal levels. The study may be considered as a novel academic reference for those interested the treatment of SF. However, the mechanism was only examined in cell and mouse experiments, and it remains to be further studied whether there are any adverse reactions or side effects of this treatment in clinical use.

Supplementary Material

Sequences of the nucleic acids used for transfection.

Acknowledgements

Not applicable.

Funding

Funding: The study was supported by The Applied Basic Research Foundation of the Department of Science, Technology of Yunnan Province, Yunnan, China (grant no. 202201AY070001-036), the National Natural Science Foundation Project (grant no. 82260207) and Scientific Research Fund of Education Department of Yunnan Province (grant no. 2023J0036).

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

ZH, WY, YG and DL were responsible for conceptualization of the study. ZH, WY and YC were responsible for methodology. ZH and WY performed validations. ZH, WY, LS and LR carried out the formal analysis. ZH, WY, YG and LY performed investigations, wrote the original draft of the manuscript, reviewed and edited the manuscript and supervised the study. LY provided resources and acquired funding. YZ, QZ and WZ were responsible for visualization and data analysis. ZH and LY confirm the authenticity of all the raw data. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

All animal experimental protocols were approved by the Laboratory Animal Welfare Ethics Committee of Yunnan University (approval no. YNU20220291), and the animal procedures are reported according to ARRIVE guidelines 2.0.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References