Long noncoding RNA ERLR mediates epithelial-mesenchymal transition of retinal pigment epithelial cells and promotes experimental proliferative vitreoretinopathy

Abstract

Proliferative vitreoretinopathy (PVR) is a disease that causes severe blindness and is characterized by the formation of contractile fibrotic subretinal or epiretinal membranes. The epithelial-mesenchymal transition (EMT) of retinal pigment epithelial (RPE) cells is a hallmark of PVR. This work aims to examine the role of a long noncoding RNA (lncRNA) named EMT-related lncRNA in RPE (ERLR, LINC01705-201 (ENST00000438158.1)) in PVR and to explore the underlying mechanisms. In this study, we found that ERLR is upregulated in RPE cells stimulated with transforming growth factor (TGF)-β1 as detected by lncRNA microarray and RT-PCR. Further studies characterized full-length ERLR and confirmed that it is mainly expressed in the cytoplasm. In vitro, silencing ERLR in RPE cells attenuated TGF-β1-induced EMT, whereas overexpressing ERLR directly triggered EMT in RPE cells. In vivo, inhibiting ERLR in RPE cells reduced the ability of cells to induce experimental PVR. Mechanistically, chromatin immunoprecipitation (ChIP) assays indicated that the transcription factor TCF4 directly binds to the promoter region of ERLR and promotes its transcription. ERLR mediates EMT by directly binding to MYH9 protein and increasing its stability. TCF4 and MYH9 also mediate TGF-β1-induced EMT in RPE cells. Furthermore, ERLR is also significantly increased in RPE cells incubated with vitreous PVR samples. In clinical samples of PVR membranes, ERLR was detected through fluorescent in situ hybridization (FISH) and colocalized with the RPE marker pancytokeratin (pan-CK). These results indicated that lncRNA ERLR is involved in TGF-β1-induced EMT of human RPE cells and that it is involved in PVR. This finding provides new insights into the mechanism and treatment of PVR.

Fig. 1. Detection and characterization of ERLR.

A ARPE-19 cells were incubated with TGF-β1 (10 ng/ml) for 48 h. Total RNA was extracted and subjected to lncRNA microarray analysis (N = 3 replicates/group). Differentially expressed lncRNAs are shown. Red indicates upregulation, whereas green denotes downregulation. B ARPE-19 cells and phRPE cells were treated with TGF-β1 (10 ng/ml) for 48 h. ERLR expression was detected by RT-PCR, and the relative fold changes were normalized to the expression of GAPDH (Data are presented as means ± SEM. N = 3 independent experiments/group). **P < 0.01 by two-tailed Students’s t test. C A FISH assay was performed to observe ERLR expression in phRPE cells treated with or without TGF-β1 (10 ng/ml) for 48 h. Mixed 18S RNA (exclusively expressed in the cytoplasm), and U6(exclusively expressed in the nucleus) were included as control. One of the representative results is presented. D Total RNA from phRPE was reverse transcribed by different primers (no primer, oligo dT, and random 6-mers). The resulting cDNAs were examined by PCR to detect ERLR expression. Agarose gel electrophoresis was used to visualize the PCR product. One of the representative results is presented. E Cytoplasmic (C) and nucleic (N) RNAs were separated and reverse transcribed to generate cDNA. A PCR experiment was conducted to detect the expression of β-actin, U2, and ERLR in both cytoplasmic and nucleic cDNA. A reaction without template cDNA (−) was conducted as a negative control. Agarose gel electrophoresis was used to visualize the PCR product. One of the representative results is presented.

Fig. 2. ERLR knockdown attenuates TGF-β1-induced EMT in RPE cells.

PhRPE cells were transfected with two ERLR siRNAs (SiERLR-1 and SiERLR-2) or a negative control siRNA (Si-NC) and then treated with or without TGF-β1 (10 ng/ml) for 48 h. A Expression levels of ERLR were detected by RT-PCR, and the relative fold changes were normalized to the expression of GAPDH (N = 3 independent experiments/group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. B Expression levels of EMT-related markers (ZO-1, E-cadherin, fibronectin, and α-SMA) were detected by RT-PCR, and the relative fold changes were normalized to GAPDH (N = 3 independent experiments/group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. C EMT-related markers were detected by WB. One of the representative results is presented. D The band intensities in WB results were analyzed and normalized to β-actin (N = 3 independent experiments/group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. E Column 1 presents the appearances of the cells under a microscope. Columns 2–5 present the EMT-related markers detected by IF. Column 6 depicts the cells subjected to the Transwell migration assay. A representative field of vision is captured. Column 7 shows the cells subjected to the collagen gel contraction assay. After the mixture of the cells and collagen gel type I was solidified, culture medium was added. The gels were then cultured for another 48 h, and images were captured. 3–4 independent experiments were conducted. One of the representative results is presented. F The number of migrated cells in the Transwell migration assay was quantified by counting the average cell number in five random fields of vision (N = 3 independent experiments/group). **P < 0.01 by one-way ANOVA and post hoc Bonferroni’s test. G The area of the collagen gel was calculated by ImageJ software and was standardized to the original area (N = 3 independent experiments/group). *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. All data are presented as means ± SEM.

Fig. 3. In RPE cells, ERLR overexpression triggers EMT, promotes migration, and enhances the capacity of mediating collagen gel contraction.

PhRPE cells were transfected with a lentivirus overexpressing ERLR (Lv-ERLR), or a negative control lentivirus (Lv-NC). The cells were then cultured for 48 h. A Expression levels of ERLR were detected by RT-PCR, and the relative fold changes were normalized to the expression of GAPDH (N = 3 independent experiments/group). **P < 0.01 by two-tailed Students’s t test. B Expression levels of EMT-related markers (ZO-1, E-cadherin, Fibronectin, and α-SMA) were detected by RT-PCR and were normalized to GAPDH (N = 3 independent experiments /group). **P < 0.01 and *P < 0.05 by two-tailed Students’s t test. C Expression levels of EMT-related markers (ZO-1, E-cadherin, Fibronectin, N-cadherin and α-SMA) were detected by western blot. One of the representative results is shown. D The band intensities in WB results were analyzed and normalized to internal standard (N = 3 independent experiments/group). *P < 0.01 and *P < 0.05 by two-tailed Students’s t test. E Cells were then subjected to Transwell migration assays. A representative field of the vision is captured. F The number of migrated cells in the Transwell migration assay was quantified by counting the average cell number in five random fields of vision (N = 3 independent experiments/group). **P < 0.01 by two-tailed Students’s t test. G Cells were subjected to the collagen gel contraction assay. After the mixture of the cells and collagen gel type I was solidified, culture medium was added. The gels were then cultured for another 48 h, and their pictures were captured. H The area of the collagen gel was calculated using ImageJ software and was standardized to the original area (N = 3 independent experiments /group). *P < 0.05 by two-tailed Students’s t test. All data are presented as means ± SEM.

Fig. 4. ERLR promotes PVR progression in an experimental PVR model.

PhRPE cells were transfected with lv-sh-NC or lv-sh-ERLR and then were injected into the vitreous body of pigmented rabbits to induce PVR. A PVR severity in each group was graded according to Fastenberg’s score at the indicated times. Data are presented as means ± SEM; N = 6 rabbits. NS not significant, *P < 0.05 by Mann–Whitney U test. B Fundus photographs in each group were captured at the indicated times by a smartphone under a microscope with the help of a Volk SuperQuad 160 fundus lens. Localized detachment of medullary ray is shown in eyes of Lv-sh-NC group at day 21 (red arrow). Total RD with retinal folds and holes (red triangle) is shown at day 28. Focal traction and detachment (red arrows) is shown in the eyes of Lv-sh-ERLR group at day 28. C Ultrasound B scanning and optical coherence tomography (OCT) scanning were performed at day 28. D H&E staining of the rabbit eye in each group was performed and analyzed. A representative image is shown. Bar = 50 μm. E α-SMA expression around the retina of the rabbits was detected by immunofluorescence. The white arrow marks the α-SMA-positive epiretinal membrane. Representative results are shown.

Fig. 5. The transcription factor TCF4 directly binds to the promoter region of ERLR and promotes its expression.

A The line represents the promoter region (1000 bp upstream) of ERLR. The regions of 5 different pairs of primers located at different regions of the promoter are depicted and were used in the following ChIP assay. B, C The chromatin of phRPE cells was immunoprecipitated with an anti-TCF4 antibody or with rabbit IgG. The resultant chromatin and total input acted as templates for semiquantitative PCR (B) or quantitative RT-PCR (C) to determine the interaction of TCF4 and the promoter region of ERLR. B Chromatin immunoprecipitated with anti-RP2 or rabbit IgG, acting as templates in the reaction with the primer of the GAPDH promoter, was used as a positive and negative control, respectively. C The relative expression of PCR products by chromatin was normalized to total input (N = 3 independent experiments/group). **P < 0.01 by two-tailed Students’s t test. D–G PhRPE cells were transfected with two TCF4 siRNAs (SiTCF4-1 and SiTCF4-2) or a negative control siRNA (Si-NC) and then treated with or without TGF-β1 (10 ng/ml) for 48 h. D TCF4 and ERLR expression was detected by RT-PCR and was normalized to the expression of GAPDH (N = 3/group). **P < 0.01 by one-way ANOVA and post hoc Bonferroni’s test. E The expression of TCF4 and EMT-related markers was detected by RT-PCR and was normalized to that of GAPDH (N = 3 independent experiments /group). *P < 0.05 and **P < 0.01 by one-way ANOVA and post hoc Bonferroni’s test. F The expression of TCF4 and EMT-related markers was detected by WB. One of the representative results is shown. G The band intensities in WB results were analyzed and normalized to internal standard (N = 3 independent experiments /group). *P < 0.05 and **P < 0.01 by one-way ANOVA and post hoc Bonferroni’s test. All data are presented as means ± SEM.

Fig. 6. ERLR triggers EMT by directly binding to the MYH9 protein.

A An RNA pulldown assay was performed using phRPE cells and biotinylated ERLR or antisense-ERLR. Proteins coprecipitated with biotin-labeled ERLR or antisense-ERLRs were detected by WB. One of the representative results is shown (N = 3 independent experiments /group). B–F PhRPE cells were transfected with two ERLR siRNAs (SiERLR-1 and SiERLR-2), two TCF4 siRNAs (siTCF4-1 and siTCF4-2) or a negative control siRNA (Si-NC) and then treated with or without TGF-β1 (10 ng/ml) for 48 h. B, D The expression of MYH9 protein was detected by WB. One of the representative results is shown. C, E The band intensities in the WB results were analyzed and normalized to β-actin (N = 3 independent experiments /group) in B, D, respectively. **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. F The expression of MYH9 mRNA was detected by qRT-PCR and was normalized to that of GAPDH (N = 3 independent experiments /group). No significant difference by one-way ANOVA and post hoc Bonferroni’s test. G–I: PhRPE cells were transfected with two MYH9 siRNAs (SiMYH9-1 and SiMYH9-2) or a negative control siRNA (Si-NC) and then were treated with or without TGF-β1 (10 ng/ml) for 48 h. G Expression levels of MYH9 and EMT-related markers were detected by RT-PCR, and the relative fold changes were normalized to GAPDH (N = 3 independent experiments /group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. H EMT-related markers were detected by WB. One of the representative results is shown. I The band intensities in WB results were analyzed and normalized to the intensity of β-actin (N = 3 independent experiments/group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. J–L PhRPE cells were transfected with ERLR-overexpressing lentivirus (lv-ERLR) or control lentivirus (lv-NC) and were then transfected with two MYH9 siRNAs (SiMYH9-1 and SiMYH9-2) or a negative control siRNA (Si-NC) and cultured for 48 h. J. Expression levels of MYH9 and EMT-related markers were detected by RT-PCR, and the relative fold changes were normalized to GAPDH (N = 3 independent experiments/group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. K EMT-related markers were detected by WB. One of the representative results is shown. L The band intensities in WB results were analyzed and normalized to the intensity of β-actin (N = 3 independent experiments /group). **P < 0.01 and *P < 0.05 by one-way ANOVA and post hoc Bonferroni’s test. All data are presented as means ± SEM.

Fig. 7. ERLR expression is associated with clinical PVR.

A PhRPE cells were cultured with 25% PVR vitreous (N = 8 samples) or normal vitreous samples (N = 4 samples) for 48 h. ERLR expression was detected by RT-PCR. Data are presented as means ± SEM. **P < 0.01 by two-tailed Students’s t test. B PVR subretinal membranes were obtained from surgery. ERLR expression was detected by FISH. Pan-CK was detected by IF. White arrow marks the co-localization of ERLR and pan-CK. C A diagram depicting the mechanism of ERLR in EMT of RPE cells (modified with permission from our previous published paper in Discovery Medicine [3]). TGF-β1 upregulates TCF4 expression, which activates the transcription of ERLR. ERLR binds to MYH9 and increases its stability, which leads to increased MYH9 protein expression and EMT. These cells then migrate, proliferate, aggregate and form the epiretinal or subretinal membrane, which contracts and ultimately leads to tractional retinal detachment.