LncRNA-COX2 inhibits Fibroblast Activation and Epidural Fibrosis by Targeting EGR1

Abstract

Rationale: Epidural fibrosis is one of the contributors to failed back surgery syndrome (FBSS) with a high incidence of about 80,000 cases per year. The fibrosis spreads from the operative region to the dura mater or the nerve root and results in functional incapacity and pain after laminectomy. Our previous study showed that down-regulation of lncRNA-COX2 is involved in the epidural scar formation. However, it remains unknown whether lncRNA-COX2 participate in the fibroblast activation and epidural fibrogenesis. Methods: LncRNA-COX2 and EGR1 expression were assessed by qRT-PCR and western blotting. Fibroblasts differentiation, proliferation and migration was determined by Collagen I/ɑ-SMA, 5-ethynyl-2'-deoxyuridine (EdU) and Transwell Assay respectively. Luciferase reporter assay was performed for the verification of target of LncRNA-COX2. Laminectomy was performed to establish the model of epidural fibrosis in mice. Epidural scar was evaluated by hematoxylin and eosin (HE) staining and Masson Trichrome staining. Results: Based on the result of transcriptome profiling, we found LncRNA-COX2 was significantly decreased in epidural tissues after laminectomy and in activated fibrotic fibroblasts. In vitro, overexpression of LncRNA-COX2 suppressed epidural fibrogenesis by inhibiting fibroblasts differentiation, proliferation and migration. Mechanistically, LncRNA-COX2 functioned as competing endogenous RNA (ceRNA) of EGR1. Gain of LncRNA-COX2 significantly decreased the expression of EGR1 and showed anti-fibrotic effect while EGR1 was markedly increased after loss of LncRNA-COX2. In vivo, LncRNA-COX2 attenuated laminectomy-induced epidural fibrosis in mice. Conclusion: In summary, the results demonstrated that LncRNA-COX2 showed anti-fibrotic effect by targeting EGR1 and identified LncRNA-COX2 as therapeutic molecule for preventing aberrant epidural fibrosis.

Keywords: EGR1.; Epidural Fibrosis; Fibroblast Activation; LncRNA-COX2; laminectomy.

Figure 1

LncRNA-COX2 inhibits fibrogenesis in epidural fibroblasts. (A) Basal LncRNA-COX2 expression in fibroblasts treated with TGF-β or not and (B) epidural tissues from surgical site and normal region of mice. qRT-PCR analysis of the expression of LncRNA-COX2 (C), Collagen I (D), α-SMA (E) in fibroblasts after transfection of LncRNA-COX2. (F) Immunofluorescence analysis of collagen I and α-SMA expression in TGF-β-cultured fibroblasts treated with LncRNA-COX2. (G) EdU analysis showing proliferative fibroblasts after TGF-β or LncRNA-COX2 treatment. Scale bar = 50 μm. (H) Quantitative analysis of proliferative fibroblasts in indicated groups. (I) Transwell assay showing migrated fibroblasts after TGF-β or LncRNA-COX2 treatment. Scale bar = 100 μm. (J) Quantitative analysis of migrated fibroblasts in indicated groups. *P<0.05.

Figure 2

LncRNA-COX2 interacts with EGR1. (A) qRT-PCR analysis of the expression of EGR1 in fibroblasts treated with TGF-β or not and (B) epidural tissues from surgical site and normal region of mice. (C) qRT-PCR analysis of EGR1 level in fibroblasts after LncRNA-COX2 overexpression. (D) Western blot analysis of EGR1 in fibroblasts in treated with TGF-β or after LncRNA-COX2 overexpression. (E) Predicted binding sites of LncRNA-COX2 and EGR1. Mut-1, mutated site 1; Mut-2, mutated site 2; Mut-3, mutated both sites 1 and 2. *P<0.05, **P<0.01.

Figure 3

Silencing LncRNA-COX2 promotes fibroblasts activation after EGR1 is suppressed. qRT-PCR analysis of EGR1 (A), Collagen I (B), α-SMA (C) in fibroblasts after EGR1 or LncRNA-COX2 inhibition. (D) Western blot analysis of EGR1, Collagen I, α-SMA in fibroblasts after EGR1 or LncRNA-COX2 inhibition. (E) Immunofluorescence analysis of collagen I and α-SMA expression in cultured fibroblasts knockdown of EGR1 or LncRNA-COX2. EdU assay (F) and Transwell assay (H) of fibroblasts in indicated groups. (G) Quantitative analysis of proliferative cells of (F). (I) Quantitative analysis of migrated cells of (H). ^#P<0.05, **P<0.01. Difference was analyzed by one-way ANOVA.

Figure 4

Degree of epidural fibrosis is markedly decreased after LncRNA-COX2 treatment. (A) Axial image of thoracic vertebrae in mice. Scale bar: 500μm. HE staining of representative images of the vertebrae morphology showing epidural wound injury site after laminectomy in different groups at the time of 3 days (B), 1 week (D), 4 weeks (F). Scale bar: 200 μm. Grades of epidural fibrosis in different groups at the time of 3 days (C), 1 week (E), 4 weeks (G) (n = 10). *P<0.05. **P<0.01. (H) Histological assessment of fibroblast infiltration at 3 days, 1 week and four weeks post-operation by HE staining analysis. Scale bar: 50 μm. (I) Fibroblast infiltration grades in epidural tissues in control group, LV-control group and LV- LncRNA-COX2 group at three different times (n = 10). *P<0.05, **P<0.01. Difference was analyzed by one-way ANOVA.

Figure 5

LncRNA-COX2 inhibits epidural fibrogenesis initiation. (A) Schematic illustrating the timelines for in vivo introduction of LncRNA-COX2 and determination of experimental sacrifice time in the laminectomy model of epidural fibrosis. (B) Cross sections of the vertebral column stained with Masson's trichrome staining to show collagen deposition (blue) and stained with the fibrotic indications. (n = 3 mice per group) at day 3. Scale bars: 200 μm. Higher-power image scale bars: 50 μm. (C) EGR1 level of epidural tissues in mice treated with saline, LV-control or LV-LncRNA-COX2. Effect of LncRNA-COX2 on the epidural fibrogenesis initiation reflected by the mRNA expression of fibrotic markers (Collagen I, α-SMA, YAP and Fibronectin) (D-G). In C-G, each symbol represents an individual mouse. *P<0.05. Difference was analyzed by one-way ANOVA.

Figure 6

LncRNA-COX2 weakens the progression of epidural fibrosis. (A) Schematic illustrating the timelines for in vivo use of LncRNA-COX2 and decision of experimental end points in the laminectomy model of epidural fibrosis. (B) Cross sections of the vertebral column stained with Masson's trichrome staining to show collagen deposition (blue) and stained with the fibrotic indications. (n = 3 mice per group) at day 7. Scale bars: 200 μm. Higher-power image scale bars: 50 μm. Effect of LncRNA-COX2 on the epidural fibrogenesis initiation reflected by the mRNA expression of fibrotic markers Collagen I (C), α-SMA (D), YAP (E) and Fibronectin (F). In C-F, each symbol represents an individual mouse. *P<0.05. Difference was analyzed by one-way ANOVA.

Figure 7

LncRNA-COX2 decreases laminectomy-induced epidural scar formation. (A) Schematic illustrating the timelines for in vivo administration of LncRNA-COX2 and determination of experimental end points, and for the final phase in the laminectomy model of epidural fibrosis. (B) Cross sections of the vertebral column stained with Masson's trichrome staining to show collagen deposition (blue) (n = 3 mice per group) at day 28. Scale bar: 200 μm. The effect of epidural scar on spinal cord after oral administration of saline, LV-control or LV- LncRNA-COX2 reflected by (C) mean thickness of epidural scar, (D) relative hydroxyproline content, (E) conus medullaris space area, (F) subarachnoid space area and (G) epidural space area. *P<0.05. **P<0.01. NS means not significant. Difference was analyzed by one-way ANOVA.

Figure 8

Schematic diagram illustrates the role of LncRNA-COX2 in the progression of laminectomy-induced epidural scar formation.