Nr4a1 enhances Wnt4 transcription to promote mesenchymal stem cell osteogenesis and alleviates inflammation-inhibited bone regeneration

Abstract

Intense inflammatory response impairs bone marrow mesenchymal stem cell (BMSC)-mediated bone regeneration, with transforming growth factor (TGF)-β1 being the most highly expressed cytokine. However, how to find effective and safe means to improve bone formation impaired by excessive TGF-β1 remains unclear. In this study, we found that the expression of orphan nuclear receptor Nr4a1, an endogenous repressor of TGF-β1, was suppressed directly by TGF-β1-induced Smad3 and indirectly by Hdac4, respectively. Importantly, Nr4a1 overexpression promoted BMSC osteogenesis and reversed TGF-β1-mediated osteogenic inhibition and pro-fibrotic effects. Transcriptomic and histologic analyses confirmed that upregulation of Nr4a1 increased the transcription of Wnt family member 4 (Wnt4) and activated Wnt pathway. Mechanistically, Nr4a1 bound to the promoter of Wnt4 and regulated its expression, thereby enhancing the osteogenic capacity of BMSCs. Moreover, treatment with Nr4a1 gene therapy or Nr4a1 agonist Csn-B could promote ectopic bone formation, defect repair, and fracture healing. Finally, we demonstrated the correlation of NR4A1 with osteogenesis and the activation of the WNT4/β-catenin pathway in human BMSCs and fracture samples. Taken together, these findings uncover the critical role of Nr4a1 in bone formation and alleviation of inflammation-induced bone regeneration disorders, and suggest that Nr4a1 has the potential to be a therapeutic target for accelerating bone healing.

Keywords: Csn-B; Nr4a1; TGF-β1; Wnt4/β-catenin; bone regeneration; mesenchymal stem cells; molecular therapy; osteogenesis.

Graphical abstract

Figure 1

Nr4a1 is directly and indirectly inhibited by TGF-β1/Smad3/Hdac4 signaling axis (A) Relative expression of Nr4a1 in different tissues of C57BL/6J mice. n = 3 per group. (B) Relative expression of Nr4a1 during osteogenic differentiation. (C and D) TGF-β1 (50 ng/mL) suppressed Nr4a1 expression, as assessed by qPCR (C) and western blot (D). (E) TGF-β1 and Nr4a1 expression levels of implants in C57BL/6J and nude mice as shown by immunohistochemistry. n = 5 per group. Scale bars, 50 μm. (F) Knockdown of Smad3 rescued TGF-β1-inhibited Nr4a1 transcription in BMSCs. (G) SB431542 (1 μM) treatment abolished TGF-β1-mediated inhibition of Nr4a1 expression. (H) Upregulation of Hdac4 levels after TGF-β1 (50 ng/mL) stimulation of BMSCs. (I) Knockdown of Hdac4 increased Nr4a1 transcription in BMSCs. CM, control medium; OM, osteogenic medium; NC, negative control sequence. Each dot represents an individual sample or animal, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 2

Nr4a1 enhances the osteogenic capacity of BMSCs and rescues high-dose TGF-β1-inhibited osteogenesis (A) Overexpression efficiency of Nr4a1 in BMSCs as assessed by qPCR. (B) Overexpression of Nr4a1 potentiated BMSC osteogenesis and rescued TGF-β1-induced reduction of mineralized nodule formation. (C–F) qPCR detection of the expression of osteoblast marker genes in BMSCs, including (C) Runx2, (D) Alp, (E) Sp7, and (F) Bglap. (G) Validation of Nr4a1 siRNA knockdown efficiency by qPCR. (H) Knockdown of Nr4a1 enhanced TGF-β1-induced inhibition of BMSC osteogenesis, as assessed by alizarin red staining. (I–L) Osteogenesis-related gene expression by qPCR. NC, negative control sequence. Each dot represents an individual sample, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 3

Bulk RNA sequencing among control and Nr4a1-overexpressed BMSCs (A) Principal-component analysis (PCA) showed major segregation among the control and Nr4a1 OE groups. (B) Hierarchical clustering analysis showed the major differences between the control and Nr4a1 OE groups. (C) Volcano plot of all transcripts. Red dots indicate >2-fold change upregulation (p < 0.05) among Nr4a1-overexpressed BMSCs, while blue dots indicate >2-fold change downregulation (p < 0.05) among Nr4a1-overexpressed BMSCs. (D) Heatmap of Nr4a1 and the top 10 significantly differentially expressed genes among Nr4a1 OE group. (E) Gene Ontology (GO) enrichment analysis of significantly upregulated genes among Nr4a1 OE group. (F) Expression of canonical osteogenesis-related genes, shown in heatmap. (G) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of significantly upregulated genes among Nr4a1 OE group. (H) Gene set enrichment analysis (GSEA) showed that the Wnt gene set was enriched in Nr4a1-overexpressed BMSCs. (I) Heatmap demonstrating mRNA expression levels of Wnt signaling related genes among the control and Nr4a1 OE groups. (J) Fold changes of Wnt family protein expression among the control and N4 OE groups are shown. (K) The expression levels of the candidate differential expression genes among the control and N4 OE groups as assessed by qPCR. OE, overexpression. Results are shown as mean ± SD of three independent experiments. ∗p < 0.05; ∗∗∗p < 0.001.

Figure 4

Nr4a1 increases Wnt4 transcription and activates Wnt signaling pathway to promote BMSC osteogenesis (A–E) qPCR detection of the expression of Wnt and downstream genes in BMSCs, including (A) Wnt4, (B) β-catenin (Ctnnb1), (C) Cyclin D1 (Ccnd1), (D) Lef1, and (E) Dkk1. (F) Western blot analysis of Nr4a1 and β-catenin expression in control and Nr4a1-overexpressed BMSCs. (G) Binding of Nr4a1 to its binding sites at the Wnt4 promoter in BMSCs, as assessed by CUT&Tag assay. (H) Luciferase assay was performed on control or Nr4a1-overexpressed 293T cells after transfection with pGL3-luciferase empty vector or pGL3-Wnt4 luciferase vector. (I) Site-directed mutagenesis analyses of Nr4a1 binding site in the Wnt4 promotor. Promotor activity was analyzed using luciferase assay in 293T cells treated with or without Nr4a1 overexpression. The activities of Wnt4 promoter were suppressed after mutation at sites 1 (Mut1) and 2 (Mut2). (J) Validation of Wnt4 siRNA knockdown efficiency and Wnt downstream gene Ctnnb1 expression by qPCR. (K) Alizarin red staining of BMSCs after Nr4a1 overexpression and siRNA-mediated inhibition of Wnt4. (L) qPCR detection of the expression of Ctnnb1 and Dkk1 in BMSCs treated with Wnt signaling inhibitor LF3 (40 μm) or vehicle. (M) Alizarin red staining of BMSCs after Nr4a1 overexpression and treatment with LF3 or vehicle. NC, negative control sequence; OE, overexpression. Each dot represents an individual sample, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 5

Nr4a1 overexpression promotes BMSC-mediated bone formation (A) Schematics of ectopic implantation in C57BL/6J mice. A mixture of β-TCP and BMSCs was implanted subcutaneously and harvested after 8 weeks. (B) Micro-CT reconstructions of implants containing control or Nr4a1-overexpressed BMSCs. (C–G) Quantitative analysis of micro-CT images among the control and Nr4a1-overexpressed BMSCs, including bone mineral density (BMD, C), bone volume fraction (BV/TV, D), bone surface density (BS/TV, E), trabecular number (Tb.N, F), and trabecular separation (Tb.Sp, G). (H) H&E staining of the implants. New bone formation was quantified using ImageJ software. B, bone; CT, connective tissue. (I) Representative ALP staining images and quantitative analysis of ALP activity. (J) GFP fluorescence showed the persistence of exogenous MSCs in the implants. The nuclei were stained with DAPI (blue). (K–N) Immunofluorescence staining and semi-quantitative analysis of Nr4a1 (K), Wnt4 (L), OCN (M), and β-catenin (N) within the implants. OE, overexpression. n = 6 per group. Scale bars, 100 μm. Each dot represents an individual sample, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 6

Nr4a1-overexpressed BMSCs improve critical size calvarial bone defect healing (A) Schematics of calvarial defect repair in C57BL/6J mice. Control or Nr4a1-overexpressed BMSCs were embedded in the hydrogel, which was implanted in the parietal bone defect. The skulls were harvested after 8 weeks for the evaluation of bone regeneration. (B) Micro-CT reconstructions of calvarial defects. The red circles indicate the edges of calvarial defects. (C and D) Quantification of bone healing, as shown by (C) BMD and (D) BV/TV. (E and F) Residual defect diameter (E) and bone formation area (F) were calculated via micro-CT images. (G) H&E-stained images of the defect site. (H and I) Representative ALP staining images (H) and quantitative analysis of ALP activity (I). (J) Co-staining of GFP, Nr4a1, and Wnt4 in the defect sites at 8 weeks post-injury. (K and L) Quantification of (K) Nr4a1 and (L) Wnt4 immunoreactivity is shown. (M) Co-staining of GFP, OCN, and β-catenin in the defect sites at 8 weeks post-injury. (N and O) Quantification of (N) OCN and (O) β-catenin immunoreactivity is shown. The nuclei were stained with DAPI (blue). White dashed lines indicate bone edges. OE, overexpression. n = 6 per group. Scale bars, 100 μm. Each dot represents an individual sample, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01.

Figure 7

Nr4a1 agonist Csn-B promotes osteogenic differentiation of BMSCs and ectopic bone regeneration (A) Alizarin red staining of BMSCs after treatment with TGF-β1 and Csn-B. (B) Quantification of osteogenic differentiation in BMSCs indicated in (A). (C) Nr4a1 expression after treatment with TGF-β1 and Csn-B. (D–I) qPCR detection of the expression of osteogenic markers and Wnt-related genes in BMSCs, including (D) Runx2, (E) Sp7, (F) Bglap, (G) Col1a1, (H) Wnt4, and (I) Ctnnb1. (J) Western blotting analyses of Nr4a1 and β-catenin expression after treatment with TGF-β1 and Csn-B. (K) Schematics of ectopic implantation in C57BL/6J mice with the intraperitoneal injection of Csn-B. A mixture of β-TCP and BMSCs was implanted subcutaneously and harvested after 8 weeks. (L) Micro-CT reconstructions of implants in C57BL/6J mice after treatment with vehicle or Csn-B. (M) Quantitative analysis of micro-CT images among the vehicle and Csn-B groups, including BMD, BV/TV, BS/TV, Tb.N, and Tb.Sp. (N) H&E images and quantification of new bone formation in the implants. B, bone; CT, connective tissue. (O) Representative ALP staining images and quantitative analysis of ALP activity. (P–T) Immunofluorescence staining and semi-quantitative analysis of Nr4a1 (P), Wnt4 (Q), OCN (R and S), and β-catenin (R and T) within the implants. The nuclei were stained with DAPI (blue). (U) Co-localization study of OCN and β-catenin in the implants. Histograms demonstrate the fluorescence intensity profiles along the white dashed lines indicated in (R). Analysis was performed using ImageJ software. n = 6 per group. Scale bars, 100 μm. Each dot represents an individual sample, while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Figure 8

Treatment with Csn-B enhances calvarial bone defect repair (A) Schematics of calvarial defect repair in C57BL/6J mice with the intraperitoneal injection of Csn-B. A 1.8-mm circular defect was created in the frontal bone. The skulls were harvested after 4 weeks for the evaluation of bone regeneration. (B) Micro-CT reconstructions of calvarial defects. The red circles indicate the edges of calvarial defects. (C and D) Quantification of bone healing, as shown by (C) BMD and (D) BV/TV. (E and F) Residual defect diameter (E) and bone formation area (F) were calculated via micro-CT images. (G) H&E-stained images of the defect site. (H and I) Representative ALP staining images (H) and quantitative analysis of ALP activity (I). (J) Immunofluorescence images of Wnt4 at the bone defect edge. The nuclei were stained with DAPI (blue). White dashed lines indicate bone edges. (K) Quantification of Wnt4 immunoreactivity is shown. (L) Immunofluorescence images of OCN and β-catenin at the bone defect edge. (M and N) Quantification of (M) OCN and (N) β-catenin immunoreactivity is shown. (O) Co-localization study of OCN and β-catenin in the defects. Histograms demonstrate the fluorescence intensity profiles along the blue dashed lines indicated in (L). Analysis was performed using ImageJ software. n = 6 per group. Black scale bars, 100 μm; white scale bars, 50 μm. Each dot represents an individual sample while whisker plots indicate mean values and 1 standard deviation. ∗p < 0.05.