The Wnt/β-catenin signaling pathway inhibits osteoporosis by regulating the expression of TERT: an in vivo and in vitro study

LiCl reverses bone loss in OVX mice

To investigate whether LiCl could improve bone loss in OVX mice, an osteoporosis model was established by ovariectomy in mice. The mice were sacrificed after 8 weeks of OVX, and their femurs were taken for micro-CT and hematoxylin and eosin (H&E) staining. Micro-CT showed that the trabeculae in the femurs of OVX mice were sparse and less continuous than those in the femur of control mice, while the application of LiCl restored the number and continuity of the trabeculae (Figure 1A). H&E staining similarly showed that the application of LiCl prevented bone loss caused by OVX in mice (Figure 1B). Parameter analysis of the femoral bone in mice showed a significant decrease in BV/TV, Tb.N, BMD, Tb.Th, and BS/TV and a significant increase in Tb.Pf and Tb.Sp in OVX mice compared to the control group, and the application of LiCl reversed these changes (Figure 1C–1I). The results indicated that the GSK3β inhibitor could reverse the reduction in osteogenic differentiation ability in mice caused by OVX.

Figure 1. LiCl reverses bone loss in OVX mice. (A) Representative 3D micro-CT images of distal femurs from three groups (Sham surgery groups, OVX groups, OVX+LiCl-treated groups) were performed after eight-week surgery. (B) H&E staining was performed for evaluating the trabecular bone of distal femurs from each group. (C–I) Corresponding parameters showing the treatment of osteoporosis by LiCl therapy. Abbreviations: BV/TV: bone volume/tissue volume; Tb.N: trabecular number; Tb.Sp: trabecular spacing; BMD: bone mass density; Tb.Th: trabecular thickness; BS/BV: bone surface area/bone volume; Tb.PF: trabecular pattern factor. n = 6. All results are performed as mean ± SD. ^#P < 0.05 vs. control group, ^*P < 0.05, ^**P < 0.01 vs. OVX group.

LiCl reverses the inactivation of Wnt pathway, increases osteogenic capacity, and reduces apoptosis levels in OVX mice

The biological functions of TERT, β-catenin, and GSK3β in osteoporosis and the effects of LiCl on their expression have not been explored. The results showed that the protein expression of TERT and β-catenin and the p-GSK3β/GSK3β ratio were downregulated in the femoral tissues of the OVX mice compared to those in the control group, as measured by western blotting, while the trend of the above protein changes in mice in the LiCl gavage group was reversed (Figure 2A–2D). The expression of the osteogenic markers osteopontin (OPN), Runx2, and bone morphogenetic protein 2 (BMP2) was also decreased in the OVX group and increased in the LiCl-treated mice (Figure 2E–2H), indicating that the protein expression levels of TERT and β-catenin and the p-GSK3β/GSK3β ratio were positively related to the protein expression of osteogenic markers. The above results were also confirmed in human bone tissue samples, as detailed in Supplementary Figure 1. Then, the apoptosis level in bone tissue was determined. As shown in Figure 2I–2K, decreased expression of the anti-apoptotic protein B-cell lymphoma-2 (Bcl-2) and increased expression of the pro-apoptotic protein BAX, measured by western blotting in the femur tissue, were observed in OVX group mice, while the application of LiCl antagonized these results.

Figure 2. LiCl reverses the inactivation of Wnt pathway, increases osteogenic capacity, and reduces apoptosis levels in OVX mice. (A–K) Western blotting and quantification of TERT, β-catenin, p-GSK3β/GSK3β, OPN, Runx2, BMP2, Bcl-2, and BAX expression levels in the sham, OVX, and OVX+LiCl group mice. (L) ALP activity detected by ALP activity assay kit in three groups of mice. (M, N) The representative pictures of immunofluorescence staining (original magnification, ×400) and their analyses showed the expressions of BMP2 and TERT in mice femurs. All results are performed as mean ± SD. ^#P < 0.05, ^##P < 0.01 vs. control group, ^*P < 0.05, ^**P < 0.01, ^***P < 0.005 vs. OVX group, n = 6.

Meanwhile, the serum alkaline phosphatase (ALP) activity assay result, which indicates high bone turnover and osteoporosis in mice, showed that the serum ALP activity of mice in the OVX group was increased, while that of mice in the LiCl gavage group was decreased (Figure 2L) [27]. Consistent with the above results, immunofluorescence staining showed similar results for the expression of TERT and BMP2 (Figure 2M, 2N). The above results suggested that the GSK3β inhibitor could upregulate TERT protein expression, reverse the inactivation of the Wnt/β-catenin pathway, reduce osteogenic capacity, and increase apoptosis levels in OVX mice.

Validation of the effects of TERT on osteogenic differentiation and apoptosis in BMSCs

To verify the effect of TERT on osteogenic differentiation and apoptosis in BMSCs, we transfected small interfering RNA of TERT into BMSCs to reduce its expression. Western blotting verified the successful knockdown of TERT in BMSCs (Figure 3A, 3B). We found that the knockdown of TERT in BMSCs resulted in the reduction of osteogenic differentiation, as evidenced by the reduction of osteogenic differentiation markers OPN and Runx2 protein expression by western blotting (Figure 3C–3E) and the reduction of ALP staining as well as ALP activity (Figure 3F, 3G) compared to those in the NC group. We also found that the knockdown of TERT in BMSCs resulted in increased levels of apoptosis, as verified by the increased apoptosis ratios of BMSCs whose nuclei were labeled red, indicating increased apoptosis by Hoechst 33342/PI staining (Figure 3H, 3I), and the increased BAX protein expression and decreased Bcl-2 protein expression by western blotting compared to those in the NC group (Figure 3J–3L).

Figure 3. Validation of the effect of TERT on osteogenic differentiation and apoptosis in BMSCs. (A, B) The siRNA-mediated depletion of TERT assessed by western blotting. (C–E) The specific proteins related to osteogenesis differentiation detected by Western blotting, including OPN and Runx2. (F) ALP staining in NC group and TERT-depleted BMSCs. (G) ALP activity detected by ALP activity assay kit in NC group and TERT-depleted BMSCs. (H, I) The Hoechst 33342/PI staining and quantification results. (J–L) The specific proteins related to apoptosis detected by Western blotting, including Bcl-2 and BAX. All results are performed as mean ± SD. ^#P < 0.05, ^##P < 0.01 vs. NC group. n = 6.

LiCl reverses H₂O₂-induced inactivation of the Wnt pathway and inhibition of osteogenic differentiation in BMSCs

H₂O₂ application to BMSCs mimics oxidative stress during the development of osteoporosis. To determine the optimal concentration of H₂O₂ acting on BMSCs, a CCK8 kit was used to detect the effect of different concentrations of H₂O₂ on the cell viability of BMSCs, and the results showed that 300 μM H₂O₂ could reduce the cell viability of BMSCs to 53.8%. From this, in the following experiments, 300 μM H₂O₂ was used as the damage concentration, and 4 hours of induction was used to cause cell apoptosis. (Figure 4A). In terms of osteogenic differentiation, ALP staining showed that the alkaline phosphatase staining of BMSCs was weaker and ALP activity was lower in the H₂O₂ group than in the control group, and the application of LiCl reversed these changes (Figure 4B, 4C). Consistent with the above results, the protein expression levels of the osteogenic-related markers OPN, Runx2, and BMP2 in BMSCs under the application of H₂O₂ were significantly reduced, as measured by western blotting, while the protein levels of the osteogenic-related markers showed the opposite trend after applying LiCl (Figure 4D–4G). We also found that TERT, β-catenin protein expression levels, and the p-GSK3β/GSK3β ratio, as measured by western blotting, were decreased in H₂O₂-induced BMSCs and that the application of LiCl reversed these changes (Figure 4H–4K).

Figure 4. LiCl reverses H₂O₂-induced inactivation of the Wnt pathway and inhibition of osteogenic differentiation in BMSCs. (A) CCK-8 analysis was conducted to evaluate cell viability after being treated with different concentrations of H₂O₂ for 6 h. (B) ALP staining in three groups of BMSCs. (C) ALP activity of BMSCs detected by ALP activity assay kit. (D–K) Protein expression levels of TERT, β-catenin, p-GSK3β/GSK3β, OPN, Runx2, BMP2 in BMSCs treated with H₂O₂ or LiCl by Western blotting. All results are performed as mean ± SD. ^#P < 0.05 vs. Control group. ^*P < 0.05, ^**P < 0.01, ^***P < 0.005 vs. H₂O₂ group, n = 6.

LiCl reverses H₂O₂-induced elevation of apoptosis in BMSCs

Hoechst 33342/PI staining results showed that the application of H₂O₂ significantly increased the apoptotic level of cells compared to that in the control BMSCs, while their apoptotic level was reduced after the application of LiCl (Figure 5A, 5B). Then, we examined the apoptotic status of BMSCs by western blotting, and the results showed that the expression of the anti-apoptotic protein Bcl-2 was decreased and the expression of the pro-apoptotic protein BAX was increased in the H₂O₂-induced group, while the application of LiCl changed the above result toward the opposite trend (Figure 5C–5E).

Figure 5. LiCl reverses H₂O₂-induced elevation of apoptosis in BMSCs. (A, B) The Hoechst 33342/PI and quantification results of apoptosis of BMSCs treated with H₂O₂ or LiCl. (C–E) Western blotting and quantification of Bcl-2 and BAX in BMSCs treated with H₂O₂ or LiCl. All results were performed as mean ± SD. ^#P < 0.05 vs. control group, ^*P < 0.05, ^**P < 0.01 vs. H₂O₂ group, n = 6. Scale bar: 100 μm.

LiCl reverses the H₂O₂-induced reduction in the nuclear translocation of β-catenin and the interaction between β-catenin and TERT in BMSCs

It is known that β-catenin nuclear translocation promotes the expression of target genes, such as Runx2, by binding to transcription factors such as TCF-4 or LEF-1. In the present study, we first verified the interaction between β-catenin and TERT in normal BMSCs by using an immunoprecipitation assay (Co-IP) (Figure 6A). Then, the immunofluorescence assay showed that β-catenin fluorescence intensity was weaker in the nuclei of BMSCs in the H₂O₂-induced group than in the control group, whereas it was darker in the nuclei after the application of LiCl, indicating that the application of LiCl could reverse the reduction in β-catenin nuclear translocation in BMSCs caused by H₂O₂ (Figure 6B). To investigate the effect of LiCl on the interaction between β-catenin and TERT, we then used immunofluorescence colocalization experiments, and the results showed that the fluorescence overlap fraction of β-catenin and TERT was reduced in BMSCs in the H₂O₂-induced group compared to that in the control group, while the application of LiCl significantly increased this fraction (Figure 6C). The above results suggest that LiCl reverses the reduction in the interaction between β-catenin and TERT caused by the application of H₂O₂ in BMSCs.

Figure 6. LiCl reverses the H₂O₂-induced reduction in the nuclear translocation of β-catenin and the interaction between β-catenin and TERT in BMSCs. (A) Co-IP showing β-catenin binding to TERT in BMSCs. (B) Immunofluorescence staining reveals nuclear translocation of β-catenin in BMSCs treated with H₂O₂ or LiCl. (C) Immunofluorescence assays were performed to identify colocalization between β-catenin and TERT in BMSCs treated with H₂O₂ or LiCl. All results were performed as mean ± SD. ^#P < 0.05 vs. control group, ^*P < 0.05 vs. H₂O₂ group, n = 6. Scale bar: 100 μm.

Confirmation of the GSK3β/β-catenin/TERT pathway in BMSCs

Next, we confirmed the GSK3β/β-catenin/TERT pathway and verified the regulation of TERT expression by the Wnt pathway. First, we used small interfering RNA to knock down GSK3β in BMSCs, and western blotting verified the successful knockdown of GSK3β (Figure 7A, 7B). Western blotting results revealed that the protein expression of β-catenin and TERT was significantly increased in the siGSK3β group compared to that in the NC group (Figure 7C–7E). To further verify the upstream and downstream relationship between β-catenin and TERT, we used small interfering RNA to knock down β-catenin in BMSCs, and the results showed that knockdown of β-catenin in BMSCs led to a reduction in TERT protein expression by western blotting (Figure 7F–7H). The immunofluorescence assay showed that β-catenin fluorescence intensity was significantly increased in the nuclei of BMSCs in the siGSK3β group compared to that in the NC group, and the results of immunofluorescence colocalization experiments showed that the interaction between β-catenin and TERT was increased in the siGSK3β group compared to that in the NC group (Figure 7I, 7J). The above results validated the interrelationship of the GSK3β/β-catenin/TERT pathway.

Figure 7. Confirmation of the GSK3β/β-catenin/TERT pathway in BMSCs. (A, B) The siRNA-mediated depletion of GSK3β was assessed by western blotting. (C–E) The specific proteins related to the pathway protein in this experiment expressed from the NC group and siGSK3β group of BMSCs were detected by Western blotting, including β-catenin and TERT. (F–H) The siRNA-mediated depletion of β-catenin and protein levels of TERT was assessed by western blotting from the NC group and siβ-catenin of BMSCs. (I) Immunofluorescence staining reveals nuclear translocation of β-catenin in BMSCs after depletion of GSK3β. (J) Immunofluorescence assays were performed to identify colocalization between β-catenin and TERT in BMSCs after depletion of GSK3β. All results were performed as mean ± SD. Scale bar: 100 μm. ^#P < 0.05, ^##P < 0.01 vs. NC group, n = 6.

The Wnt/β-catenin pathway promotes osteogenic differentiation and inhibits the apoptosis of BMSCs through upregulating TERT expression

To further verify the role of TERT in the osteogenic differentiation and apoptosis of BMSCs under the regulation of the Wnt pathway, we used small interfering RNA to simultaneously knock down the GSK3β and TERT genes in BMSCs. The cells were divided into four groups: NC, siGSK3β, siTERT, and siGSK3β+siTERT. Western blotting results showed that the TERT expression level in BMSCs was significantly elevated in the siGSK3β group and decreased in the siTERT group compared to that in the NC group, while it was slightly elevated in the siGSK3β+siTERT group compared to that in the siTERT group but lower than that in the siGSK3β group. Moreover, the expression trends of the osteogenic markers OPN and BMP2 and the antiapoptotic protein Bcl-2 were the same as those of TERT, and the expression trend of the apoptotic protein BAX was opposite to that of TERT (Figure 8A–8F). The results of Hoechst 33342/PI staining were also consistent with the above results. Compared to that in the NC group, the rate of apoptotic cells was reduced in the siGSK3β group and increased in the siTERT group, while the rate of apoptotic cells in the siGSK3β+siTERT group was slightly higher than that in the siGSK3β group but lower than that in the siTERT group (Figure 8G, 8H). The ALP activity, which was positively correlated with the osteogenic differentiation of BMSCs, was also consistent with the above results of the ALP activity assay (Figure 8I). The above results indicated that siTERT partially counteracted the osteogenic-promoting effect of siGSK3β in BMSCs, suggesting that the Wnt pathway promoted osteogenic differentiation and inhibited apoptosis in BMSCs and that the regulatory effects acted partly through upregulation of TERT expression.

Figure 8. The Wnt/β-catenin pathway promotes osteogenic differentiation and inhibits the apoptosis of BMSCs through upregulating TERT expression. (A–F) Western blotting and quantification of TERT, OPN, BMP2, Bcl-2, and BAX in NC, siGSK3β, siTERT, and siGSK3β + siTERT groups of BMSCs. (G, H) The Hoechst 33342/PI and quantification results in four groups of BMSCs. (I) ALP activity assay kit to detect ALP activity in four groups of BMSCs. Scale bar: 100 μm. All results were performed as mean ± SD. ^#P < 0.05 vs. NC group, ^*P < 0.05 vs. siGSK3β group. n = 6.

Cell culture and treatment

The mouse bone marrow mesenchymal stem cells (BMSCs) line was obtained from Chuanqiu Biotechnology (Shanghai, China, Item No: M015). BMSCs were grown in DMEM/F12 essential medium including 1% penicillin and streptomycin (KeyGEN BioTECH, Nanjing, China) and supplemented with 10% fetal bovine serum (FBS) in 5% CO2 at 37°C according to the approved culture protocols for this cell line.

Animal models and grouping

Thirty female C57BL/6J mice (8 weeks old, mean weight 18.5 g) were purchased from Chang Sheng Biological Company (Benxi, Liaoning). They were randomly divided into three groups: sham-operated group (control, n = 10), bilateral ovariectomy group (OVX, n = 10), and bilateral ovariectomy plus GSK3β inhibitor lithium chloride gavage group (OVX+LiCl, n = 10). The mice were anesthetized with 1% pentobarbital sodium intraperitoneally after two weeks of acclimatization, and the ovariectomy was started under aseptic conditions. Mice in the OVX+LiCl group were gavaged with LiCl (200 mg/kg/d, dissolved using ddH₂O) [47, 48]. The mice in the control group and OVX group were gavaged with the same volume of ddH₂O daily. All mammalian studies were approved by the Dalian Medical University Animal Care Committee (Dalian, China; Certificate of Conformity: No. AEE21084). After 8 weeks of OVX, all mice were euthanized with 1% pentobarbital sodium, and the serum and femoral tissues were taken for subsequent experiments.

Human bone tissue

Bone tissue samples were placed in a liquid nitrogen tank immediately after acquisition. This method of preservation maintains the activity of BMSCs in the tissue and maintains the integrity of the bone tissue.

Micro-computed tomography (micro-CT) analysis

The femoral tissues of three groups of mice were taken and scanned using a Micro-CT instrument (ZhongKeKaisheng, Guangzhou, China) to reconstruct 3D images of the mice femur. Images of the femoral trabecular area, femoral cortical area, and other levels were obtained based on 3D images. Trabeculae-related parameters were quantified, including bone volume fraction (BV/TV), trabecular number (Tb.N), femoral bone mineral density (BMD), trabecular thickness (Tb.Th), bone volume ratio of bone surface area (BS/BV), trabecular pattern factor (Tb.Pf), and trabecular separation (Tb.Sp).

H&E staining of mice femur

Mouse femurs were immersed in 4% paraformaldehyde fixative to deform protein coagulation and then placed in 10 % EDTA solution, decalcified at about 4 weeks, and the femurs were embedded in melted paraffin wax. Longitudinal sections 5 μm thick were cut, stained with hematoxylin-eosin, and scanned under an inverted microscope (Leica, Germany) after sectioning to derive histomorphological images of the femur.

Mouse BMSCs osteogenic differentiation induction

BMSCs were uniformly seeded in the 6-well plates, and the osteogenesis was induced by adding an osteogenic induction solution after each group was molded. The osteogenic induction solution contained dexamethasone at a concentration of 100 nM, vitamin C at a concentration of 50 μM, and sodium β-glycerophosphate at a concentration of 10 mM. The osteogenic induction solution was changed every two days for 14 days for alkaline phosphatase staining.

Western blotting

Western blotting was used to detect the expression of the relevant proteins in BMSCs or bone tissues. The protein of bone tissues was extracted as followings, firstly, full grinding of tissues by grinding apparatus, centrifugation after full soaking with protein lysis solution to obtain supernatant, and then the protein concentration was measured by BCA protein assay kit (Beyotime, China). The cellular protein is also obtained by protein lysis, and the steps of concentration measurement were the same as that of bone tissues. The obtained proteins were subjected to sampling, electrophoresis, gel cutting, membrane transfer, blocking, incubation of primary and secondary antibodies respectively, and development steps to finally obtain protein images and quantify the protein expression by Image J (National Institutes of Health, Bethesda, MA, USA) analysis software according to the manufacturer’s instructions. The primary antibodies used in this study were TERT (Bioss, China), β-catenin (Proteintech, China), GSK3β (Proteintech, China), p-GSK3β (Proteintech, China), RUNX2 (Abcam, American), BMP2 (Proteintech, China), OPN (Proteintech, China), Bcl-2 (Proteintech, China), BAX (Proteintech, China), and GAPDH (Proteintech, China). The horseradish peroxidase (HRP)-conjugated secondary antibodies (Proteintech, China) used were directed against primary antibodies.

Alkaline phosphatase (ALP) staining and activity detection

To study the osteogenic ability of BMSCs, osteogenic induction was performed for 14 days after the end of cell modeling, and the assay of ALP activity in the cell lysate was performed according to the steps of the ALP activity kit (Nanjing Jiancheng Institute of Biological Engineering, China). In addition, ALP staining was also performed according to the alkaline phosphatase staining kit (Beyotime) instructions.

Cell proliferation assay

Cell Counting Kit-8 (SEVEN, Beijing) was used to detect the effect of different concentrations of H₂O₂ on the cell viability of BMSCs. The steps were to spread the cells evenly in a 96-well plate, after which the cells were treated with different concentrations (0, 50, 100, 150, 200, 250, 300, 350, 400 μmol) of H₂O₂ for 4 hours, followed by changing the DMEM/F12 culture medium containing 10 μl CCK-8 to incubate the cells for 3 hours. Finally, the absorbance values at 450 nm per well were detected in a microplate reader (Thermo-354, Thermo Fisher Scientific), and cell viability was calculated according to the formula.

Hoechst 33342/PI staining

BMSCs were evenly seeded in 24-well plates, and the culture solution was aspirated after the end of cell modeling, 0.8 to 1 ml of staining buffer (Beyotime), 5μL Hoechst staining solution and 5μL PI staining solution (Solarbio, Beijing) was added, incubated for 30 min at 4°C, and washed once with PBS solution and observed under an inverted microscope (Leica, Germany).

Immunofluorescence staining of bone tissue and cells

The paraffin sections of the mouse femur were placed in an oven at 60°C for one day, removed and immersed in different concentrations of alcohol (95%, 75%, 50%, and 25%) and water for 5 minutes each, and then placed in 0.01 M sodium citrate solution and boiled in a microwave oven. After the above steps, the sections were rinsed with phosphate buffer, and then primary antibody (diluted 1:50-1:200) was added dropwise on the surface of the sections and incubated at 4°C in a refrigerator for 12–16 hours. On the second day, the sections were rinsed with phosphate buffer, and fluorescent secondary antibody (diluted 1:300) was added dropwise on the sections, which were closed at 37°C for 1 h. Then, the sections were incubated with DAPI solution for 10 min and protected from light, and finally photographed under a fluorescence microscope (Leica, Germany).

After modeling, BMSCs were washed using PBS, and then cells were fixed with 4% paraformaldehyde and washed again. Each well was perforated by adding 0.5% Triton-X 100 for 15 min, washed and then closed with 1% BSA for 30 min, and finally, the primary antibody (diluted 1:300) was added and hybridized at 37°C for 2 h. After PBS washing, the fluorescent secondary antibody (diluted 1:300) was added and incubated in a 37°C water bath for 1 h and were incubated with DAPI solution for 10 min. After that, the cells were washed again with PBS. Finally, it was photographed using a fluorescence microscope (Leica, Germany).

Immunocoprecipitation (Co-IP)

The total protein was extracted from the cells and incubated with protein A/G Plus - agarose beads and conjugated with an anti-brief antibody. After incubation overnight at 4°C, the beads were centrifuged at 4°C, 500 × g for 5 min. The supernatant was discarded and the IP product was washed three times with PBS. The target proteins were finally detected by western blot analysis.

BMSCs transfection

GSK3β, β-catenin, and TERT sequences were constructed by chemical synthesis, purchased from Gema Biologicals (Shanghai, China), and the sequences are shown in Table 1. BMSC was spread evenly on a 6-well plate, and the transfection was started when the cell density was about 50%. 0.5 OD of si or NC was added to 62. 5 μl of DEPC water, at which point si or NC was 20 mM, and the mate was diluted with DMEM/F12 complete medium to the same concentration as si or NC, and 1 ml of si/NC+mate was added to each well. After transfection for 4 to 6 h, the medium in each well was discarded, washed twice with complete medium, and then replaced with F12 complete medium, and waited for 2 to 3 days to extract the cell proteins for assay.

Statistical analysis

Statistical data analysis was completed using GraphPad Prism 7 and SPSS Statistics applications. All experimental data were calculated as mean (Mean) ± standard error (SD), and significant differences between any two groups were assessed by one-way analysis of variance (ANOVA), and p-values < 0.05 were considered as significant differences.