Decorin: a potential therapeutic candidate for ligamentum flavum hypertrophy by antagonizing TGF-β1

Abstract

Ligamentum flavum hypertrophy (LFH) is the main physiological and pathological mechanism of lumbar spinal canal stenosis (LSCS). The specific mechanism for LFH has not been completely clarified. In this study, bioinformatic analysis, human ligamentum flavum (LF) tissues collection and analysis, and in vitro and in vivo experiments were conducted to explore the effect of decorin (DCN) on LFH pathogenesis. Here, we found that TGF-β1, collagen I, collagen III, α-SMA and fibronectin were significantly upregulated in hypertrophic LF samples. The DCN protein expression in hypertrophic LF samples was higher than that in non-LFH samples, but the difference was not significant. DCN inhibited the expression of TGF-β1-induced fibrosis-associated proteins in human LF cells, including collagen I, collagen III, α-SMA, and fibronectin. ELISAs showed that TGF-β1 can upregulate PINP and PIIINP in the cell supernatant, and this effect was inhibited after DCN administration. Mechanistic studies revealed that DCN suppressed TGF-β1-induced fibrosis by blocking the TGF-β1/SMAD3 signaling pathway. In addition, DCN ameliorated mechanical stress-induced LFH in vivo. In summary, our findings indicated that DCN ameliorated mechanical stress-induced LFH by antagonizing the TGF-β1/SMAD3 signaling pathway in vitro and in vivo. These findings imply that DCN is a potential therapeutic candidate for ligamentum flavum hypertrophy.

Fig. 1. Bioinformatics analysis revealed TGF-β1 as a key regulator in the development of LFH.

A The volcano plot shows 570 upregulated DEGs and 457 downregulated genes. B The heatmap shows the top 200 up- and downregulated genes. C Enrichment analysis of biological processes from the GO database. D Enrichment analysis of cell components from the GO database. E Enrichment analysis of molecular function from the GO database. F GSEA of DEGs. G KEGG enrichment analysis of the key pathways involved in LFH. H PPI analysis identified two major modules. I–L The relative expression levels of TGF-β1, COL1A2, FN1, and DCN (n = 4). Data are presented as the means ± SDs. ns, no significance; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 2. Increased fibrotic degree of LF in patients with LFH.

A Coronal and sagittal MRI of LF. The LF thickness was determined on coronal MRI. B Comparison of LF thickness between the LFH group and the non-LFH group (n = 20). C Representative images of H&E staining and EVG staining of the LF samples from the two groups (n = 8). EVG staining (collagen fibers were stained pink, while elastic fibers were stained black). The scale bar indicates 100 μm. D Comparison of the percentage of collagen fibers and elastic fibers between the two groups (n = 8). E Comparison of LF fibrosis scores between the two groups (n = 8). F Western blot analysis of collagen I, collagen III, α-SMA and fibronectin protein expression in LF samples from the two groups. GAPDH was the loading control (n = 6). G Quantitative analysis of collagen I, collagen III, α-SMA and fibronectin protein expression in LF samples from the two groups (n = 6). H Representative images of immunohistochemical staining of collagen I, collagen III, α-SMA and fibronectin in LF samples from the two groups (n = 8). The scale bar indicates 100 μm. Data are presented as the means ± SDs. ns, no significance; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 3. TGF-β1 and DCN expression were upregulated in LFH.

A Western blot analysis of DCN and TGF-β1 protein expression in LF samples from the two groups. GAPDH was the loading control (n = 6). B Quantitative analysis of DCN and TGF-β1 protein expression in LF samples from the two groups (n = 6). C Representative images of immunohistochemical staining of DCN and TGF-β1 in LF samples from the two groups (n = 8). The scale bar indicates 100 μm. Data are presented as the means ± SDs. ns, no significance; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 4. DCN inhibited LF cell proliferation.

A Calcein-AM/PI double staining of LF cells after intervention with different concentrations of DCN (green, living cells; red, nuclei of dead cells). The scale bar indicates 400 μm. B Quantitative analysis of the percentage of living cells (six random fields of view for each well). C CCK-8 assay of LF cells showed that DCN suppressed the proliferation of LF cells, especially when the concentration of DCN was 100 nM and 200 nM (n = 6). Data are presented as the means ± SDs. ns no significance; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 5. TGF-β1 induced fibrosis in normal LF cells, while DCN inhibited fibrosis in hypertrophic LF cells.

A Western blot analysis and B quantitative analysis of collagen I, collagen III, α-SMA and fibronectin protein expression in normal LF cells after administration of different concentrations of TGF-β1. GAPDH was the loading control (n = 3). C ELISAs of PINP and PIIINP levels in the cell supernatant of normal LF cells after the administration of different concentrations of TGF-β1 (n = 3). D Western blot analysis and E quantitative analysis of collagen I, collagen III, α-SMA and fibronectin protein expression in hypertrophic LF cells after administration of different concentrations of DCN. GAPDH was the loading control (n = 3). F ELISAs of PINP and PIIINP levels in the cell supernatant of hypertrophic LF cells after the administration of different concentrations of DCN (n = 3). Data are presented as the means ± SDs and compared with those of the control group. ns, no significance; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 6. DCN inhibited TGF-β1-induced fibrosis-associated protein expression in LF cells.

A Western blot analysis and B quantitative analysis of the protein expression levels of collagen I, collagen III, α-SMA and fibronectin in LF cells after intervention with 10 ng/ml TGF-β1 and different concentrations of DCN for 24 h (n = 3). C–F Immunofluorescence staining of collagen I, collagen III, α-SMA and fibronectin under different interventions for 24 h (fibronectin and collagen III were stained green; collagen I and α-SMA were stained red; DAPI, blue). Scale bar: 400 µm. G The levels of PINP and PIIINP in the cell supernatant of LF cells under different interventions for 24 h (n = 3). H Western blot analysis and I–J quantitative analysis of SMAD3 signaling protein expression levels in LF cells after intervention with 10 ng/ml TGF-β1 and different concentrations of DCN for 24 h (n = 3). Data are presented as the means ± SDs. ###p < 0.001 vs. the control group; *p < 0.05 vs. the TGF-β1 group; **p < 0.01 vs. the TGF-β1 group; ***p < 0.001 vs. the TGF-β1 group.

Fig. 7. DCN ameliorates mechanical stress-induced LFH in vivo.

A Representative images of H&E- and EVG-stained LF tissues from rats in each group. Scale bar: 400 µm. B–E Quantitative analysis of the LF thickness, LF area, percentage of collagen fibers and elastic fibers, and LF fibrosis score in H&E- and EVG-stained sections (n = 10 rats in each group). F, G Representative images of immunohistochemical staining for collagen I and fibronectin in rat LF tissues under different interventions after 8 weeks. Scale bar: 400 µm (lower power lens); 200 µm (high power lens). H ELISAs of PINP and PIIINP levels in LF tissues from rats (n = 10 rats in each group). I ELISAs of PINP and PIIINP levels in serum samples from rats (n = 10 rats in each group). Data are presented as the means ± SDs. ns, no significance; *p < 0.05; **p < 0.01; ***p < 0.001.