All patients provided written informed consent for the preservation and analysis of their tissues for research purposes. The present study was approved by the Ethics Committee of Shanghai Tenth People's Hospital (Shanghai, China) and complied to the guidelines of the Declaration of Helsinki (version 2013) (16). The ethics approval no. was SHSY-IEC-KY-4.0/17-79/01.

ARPE-19 cells were obtained from the Eye Institute of Tongji University. ARPE-19 cells at passage 4-6 were cultured in DMEM/F12 (Invitrogen; Thermo Fisher Scientific, Inc.) with 10% fetal bovine serum (Invitrogen; Thermo Fisher Scientific, Inc.) at 37˚C in an incubator containing 5% CO₂. For TGF-β1 (HumanZyme, Inc.; ProteinTech Group, Inc.) treatment, ARPE-19 cells were subjected to serum-free medium at 37˚C for 16 h. TGF-β1 was added to the medium at 10 ng/ml. A total of 3.5 µM of the NF-κB inhibitor BAY11-7082 (Beyotime Institute of Biotechnology) was added to the medium at 37˚C for 1 h before TGF-β1 treatment.

ARPE-19 cells were transfected with 100 nM miR-29b mimic (3'-UUGUGACUAAAGUUUACCACGAU-5') and miR-negative control mimic (miR-NC mimic; 3'-AUGCAGUAUACUGUUAAUGACUC-5') (Guangzhou RiboBio Co., Ltd.) using Lipofectamine^® 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.) at room temperature for 6 h following the manufacturer's instructions. ARPE-19 cells were starved overnight in serum-free medium before the addition of TGF-β1.

The plasmids of Akt2-targeting short hairpin RNA (sh)-Akt2 and sh-control (both from Shanghai GeneChem Co., Ltd.) were transfected into ARPE-19 cells (2.5 µg plasmids or controls in 2.5x10⁵ cells) at 37˚C for 48 h. The HET kit (Biowit Technologies Ltd.) was used for the transfections according to the protocols provided by the manufacturer.

All patients with PVR (n=17; aged 45-66 years; 7 females and 10 males) were from the Shanghai Tenth People's Hospital (Shanghai, China) between January 2018 and August 2019. Patients with diabetes, acute inflammation or systemic immune diseases were excluded. After vitrectomy surgery, the epiretinal membranes were removed with internal limiting membrane (ILM) forceps. The epiretinal membranes were immediately stored at -80˚C.

The total RNA of epiretinal membranes was extracted using the RNAprep pure Micro kit (cat. no. DP 420; Tiangen Biotech Co., Ltd.) according to the manufacturer's protocol. A reverse transcription kit (cat. no. RR036A; Takara Bio, Inc.) was used to synthesize cDNA according to the manufacturer's instructions. SYBR Green-based RT-qPCR was performed with a 7500 Fast RT-PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). Specific primers (GENTEC Corporation) were used to determine the expression of EMT-related genes. GAPDH was used as an internal control. The thermocycling conditions were as follows: 94˚C for 30 sec, 40 cycles of 94˚C for 5 sec and 60˚C for 30 sec. The relative amount of each gene was measured using the 2^-ΔΔCq method as reported in a previous study (15). All experiments were repeated 3 times. Primer sequences, which were also used in a previous study (10), are presented in Table I.

Western blot analysis was performed as previously described (17). The antibodies and the dilutions were as follows: Rabbit anti-RPE-65 (1:1,000; cat. no. ab231782; Abcam), anti-Akt2 (1:500; cat. no. ab131168; Abcam), anti-p-Akt2 (ser474; 1:500; cat. no. ab38513; Abcam), anti-Snail (1:1,000; cat. no. ab216347; Abcam), anti-p65 (1:1,000; cat. no. AF1234; Beyotime Institute of Biotechnology), anti-p-p65 (ser536; 1:1,000; cat. no. AF5881; Beyotime Institute of Biotechnology), and anti-GAPDH (1:500; cat. no. AB-P-R001; Goodhere Biological Technology Group). Anti-rabbit IgG secondary antibodies (1:10,000; cat. no. W4011; Promega Corporation) were then applied and incubated at room temperature for 1 h. Image Quant LAS 4000 (GE Healthcare Life Sciences; Cytiva) and Image J analysis software (version. no. 1.52t; National Institutes of Health) were used to quantify band intensities.

At least three independent experiments were performed for each analysis. Data are presented as the mean ± standard deviation. One-way analysis of variance (ANOVA), Dunnett's t-test and Tukey's post hoc test were used to analyze the data of multiple groups. The association between miR-29b and EMT biomarkers was evaluated using Pearson's correlation analysis. P<0.05 was considered to indicate a statistically significant difference. SPSS 16.0 (SPSS, Inc.) was used for significance analysis.

Epiretinal membranes are the characteristic of the PVR process that pulls the retina and leads to retinal detachment (1). To investigate the relationship between miR-29b and EMT during the formation process of epiretinal membranes, the correlation between miR-29b and the mRNA expression of E-cadherin, α-SMA and Akt2 in epiretinal membranes of patients with PVR was assessed. As demonstrated in Fig. 1A, miR-29b was positively correlated with E-cadherin mRNA expression, while there was a negative correlation between miR-29b and α-SMA mRNA expression (Fig. 1B), and between miR-29b and Akt2 mRNA expression (Fig. 1C). These results indicated that EMT played a vital role in the development of PVR, and that miR-29b and its target Akt2 participated in this process.

The downstream pathways of miR-29b and Akt2 were further investigated. The NF-κB pathway has been widely studied in the EMT process of other cells (18,19). Snail was revealed to play a critical part in the EMT process of ARPE-19 cells in our previous study (5). It was concluded that p65, p-p65 and Snail were downstream of miR-29b and Akt2 in ARPE-19 cells. As demonstrated in Fig. 2A and B, the expression of RPE-65 did not decrease at passage 6 of ARPE-19 cells. Therefore, ARPE-19 cells at passages 4-6 were used in this study. miR-29b expression was significantly upregulated after miR-29b mimic was transfected into ARPE-19 cells (Fig. 2C). The results also revealed that TGF-β1 induced the expression of Akt2, p-Akt2, p-p65 and Snail while the addition of miR-29b significantly inhibited these expression levels (Fig. 2D and E). No significant difference in the expression of p65 was observed. Collectively, the expression levels of p-Akt2, p-p65 and Snail expression were affected by miR-29b. Additionally, p-p65 and Snail were probably involved in the EMT of ARPE-19 cells.

To further study the relationship of Akt2, p-p65 and Snail in ARPE-19 cells, the expression of Akt2 was silenced, while p-p65 and Snail expression was determined. As demonstrated in Fig. 3A, Akt2 was significantly downregulated after sh-Akt2 transfection in ARPE-19 cells. TGF-β1 induced the expression of p-p65 and Snail; however, the expression levels of p-p65 and Snail were downregulated after Akt2 expression was inhibited by sh-Akt2 (Fig. 3B and C). The aforementioned results indicated that p-p65 and Snail participated in the downstream pathway of miR-29b and Akt2.

It was revealed that Snail is an important zinc-finger transcription factor in ARPE-19 cell EMT (5). To investigate the regulatory effect of p-p65 on Snail, the NF-κB inhibitor BAY11-7082 was used to reduce p-p65 expression. TGF-β1 induced Akt2 and p-Akt2 expression and this was not affected by the addition of BAY11-7082 (Fig. 4A and B). Moreover, Snail expression was upregulated after ARPE-19 cells were treated with TGF-β1; however, this effect was inhibited by BAY11-7082. These results indicated that the function of Snail in EMT was regulated by NF-κB effector p-p65 (Fig. 4A and B).