Curcumin Prevents Diabetic Osteoporosis through Promoting Osteogenesis and Angiogenesis Coupling via NF-κB Signaling

Diabetic osteoporosis (DOP) is a metabolic disease which is characterized by impaired bone microarchitecture and reduced bone mineral density resulting from hyperglycemia. Curcumin, an effective component extracted from Curcuma longa, exhibits antioxidation, regulation of bone metabolism and hypoglycemic effects. The BMSC-mediated osteogenesis and angiogenesis coupling seems to be important in bone formation and regeneration. We aimed to explore the effect of curcumin on BMSC-mediated osteogenesis-angiogenesis coupling in high glucose conditions and underlying mechanisms. Our results showed that high glucose impaired the osteogenic and proangiogenic ability of BMSCs and that curcumin pretreatment rescued the BMSC dysfunction induced by high-concentration glucose. Inhibition of the high glucose-activated NF-κB signaling pathway has been found to contribute to the protective effects of curcumin on high glucose-inhibited coupling of osteogenesis and angiogenesis in BMSCs. Furthermore, accelerated bone loss and decreased type H vessels were observed in diabetic osteoporosis mice models. However, curcumin treatment prevented bone loss and promoted vessel formation in diabetic osteoporosis mice. Based on these results, we concluded that curcumin ameliorated diabetic osteoporosis by recovering the osteogenesis and angiogenesis coupling of BMSCs in hyperglycemia, partly through inhibiting the high glucose-activated NF-κB signaling pathway.


Introduction
Diabetic osteoporosis (DOP) is becoming an increasing complication of diabetes, which is characterized by destructive bone microarchitecture and reduced bone mineral density (BMD) [1][2][3][4][5]. Compared with normoglycemic individuals, patients with DOP have a signifcantly increased risk of fractures, leading to high disability and mortality rates [6,7]. However, an ideal treatment for DOP is currently lacking.
Bone regeneration and remodeling are associated with the osteogenesis and angiogenesis coupling [8,9]. Meanwhile, cross-talk between BMSCs and endothelial cells seems to be vital in bone regeneration [10][11][12][13]. Endothelial cells are largely dedicated to the improvement of the recruitment and osteogenic diferentiation of BMSCs. Conversely, BMSCs secrete multiple proangiogenic growth factors to promote angiogenesis and tissue regeneration. However, much evidence has shown that the declined osteogenic diferentiation function and angiogenesis ability of BMSCs is an important mechanism of DOP [14][15][16][17]. Terefore, restoring the damaged osteogenesis and angiogenesis function of BMSCs is crucial for the treatment of DOP.
NF-κB, which is known as a master transcription factor, contributes to the development of diabetes and its complications. Inactivation of NF-κB can reduce infammatory cytokine secretion in diabetic animal models and patients [39,40]. Furthermore, studies have discovered that NF-κB signaling pathway have a signifcant infuence on osteoporosis therapy and bone regeneration [41,42]. Downregulation of NF-κB p65 expression drastically alleviates osteoporosis in OVX mice [43]. Te activated NF-κB signaling pathway inhibits BMSCs mediation of the bone regeneration and angiogenesis in a distraction osteogenesis model [44]. Hence, these fndings demonstrate that inhibiting overactivated NF-κB pathway may be a new way for the treatment of diabetes and osteoporosis.
We aimed to investigate whether curcumin could rescue high glucose-inhibited osteogenesis and angiogenesis of BMSCs. In addition, we examined whether inhibiting overactivated NF-κB signaling pathway contribute to the protective efects of curcumin on BMSCs dysfunction. Furthermore, we used a DOP model to explore the efects of curcumin on bone regeneration and vessel formation in vivo. Te current study laid the foundation for curcumin in the treatment of DOP.

Materials and Methods
All animal protocols and experiments were approved by the Institutional Committee for Animal Use and Care at Shanghai University of Traditional Chinese Medicine.

Cell
Cultures. C57BL/6 male mice of six-week-old (SLAC Laboratory Animal Co. Ltd., China) were used to obtain bone marrow samples. BMSCs were isolated from the bone marrow. Te cells were inoculated in culture dishes containing DMEM (HyClone) and 10% FBS (Gibco). BMSCs from passages 2-4 were used in our study.

Cell Viability.
BMSCs were treated with curcumin (Cur; Sigma-Aldrich) from 0.1 μM to 10 μM for 48 hours. After that, a CCK-8 kit was used to measure the cell viability.

BMSCs Osteogenic
Diferentiation. Cells were cultured with 5.5 mM glucose as the control group (NG), and BMSCs treated with 33 mM glucose represented the high glucose group (HG). For the curcumin groups, cells were cultured with 33 mM glucose and 1 μM curcumin (HG + Cur). After seven days of osteogenic induction, ALP staining analysis and ALP activity was detected. After 21 days of culture, calcium nodules were detected through staining with a 2% alizarin red S solution (Solarbio, China).

Conditioned Medium (CM) of BMSCs.
After 48 hours of treatment with or without curcumin, BMSCs were cultured in fresh DMEM for an additional 48 hours. Ten, the CM from each group (CM NG group, CM HG group, and CM HG + Cur group) was harvested.

Enzyme-Linked Immunosorbent Assay (ELISA).
An ELISA kit (Beyotime Biotechnology, China) was used to detect the concentration of VEGF in diferent conditioned media.

EdU Analysis.
HUVECs were cultured with diferent CM for 24 hours and incubated in EdU working solution (Beyotime Biotechnology, China) for 2 hours. After fxation for 15 minutes with 4% paraformaldehyde, cells were treated with the click additive solution for another 15 min. Finally, the nuclei were stained using a Hoechst solution.

Scratch Wound Assay.
HUVECs were inoculated in sixwell plates and grown to 100% confuence. Two parallel scratches were generated in each well using a 200 μL pipette tip, and the medium was replaced with CM. After 24 hours, the scratched areas were measured.
2.9. Tube Formation. Cells were inoculated onto Matrigel-coated96-well plates and cultured with diferent CM for 24 hours. Te tubule number and lengths were quantifed using ImageJ software. 2.11. Western Blot. Cells were lysed with RIPA lysis bufer for 30 min and proteins were extracted. Protein samples were electrophoresed on 5-12% SDS-polyacrylamide gels.

Animals and Treatment.
Fifteen six-week-old C57BL/6 male mice received intraperitoneal injections of streptozotocin (STZ, 50 mg/kg; Sigma-Aldrich) to induce DM. At three and seven days after STZ injection, the fasting blood glucose (FBG) was detected using a glucometer (OMRON, Japan). Only mice with FBG two times higher than 16.7 mmol/L were considered successful models of DM [45][46][47][48]. Twelve mice ft the criterion and were divided into two groups (n � 6 mice per group) according to the treatment: DM and DM + curcumin (DM + Cur). Te other six mice were grouped into the control group. Curcumin was administered via gavage at a dose of 100 mg/kg/day for eight weeks after the establishment of the DM model.

Micro-CT Analysis.
At the end of these experiments, the mouse femur samples were analyzed using a micro-CT system. Te region of interest (ROI) was selected and the following parameters were calculated.
2.14. Microfl Perfusion. Cardiac Microfl (FlowTech) perfusion was performed to evaluate the neovascularization. Subsequently, the perfused mice were placed at 4°C for 24 h, and then the femur samples were decalcifed for four weeks. Te vessel formation were analyzed using a micro-CT system.

Immunofuorescence Staining of Bone Tissue.
Te sections were incubated with a CD31 antibody (1 : 100; Abcam) and an endomucin (EMCN) antibody (1 : 100; Santa Cruz) together at 4°C overnight and then were incubated with secondary antibodies (1 : 200; Santa Cruz) for 1 hour. Te images were photographed with a camera. Te CD31 hi Emcn hi type vessels were measured using Image-Pro Plus software based on color recognition.

Statistical Analysis.
All experimental data are presented as mean ± SD and the statistical analysis was tested using Student's t-test and one-way ANOVA with SPSS 17.0 software.

Curcumin Rescued the High Glucose-Inhibited BMSCs
Osteogenic Diferentiation. CCK-8 results showed that curcumin ranging from 0.1 μM to 1 μM was not signifcantly infuenced on cell viability; however, 5 μM and 10 μM curcumin decreased cell viability (Figure 1(a)). Several studies have suggested that high concentrations of glucose can suppress the BMSCs osteogenic diferentiation. We, therefore, determined whether curcumin could rescue the high glucose-induced osteogenic dysfunction of BMSCs in vitro. Our results demonstrated that curcumin pretreatment signifcantly reversed the decrease in ALP staining and activity induced by high glucose (Figure 1(b)). Similarly, alizarin red staining and semiquantitative analysis results also showed that the reduced mineralization nodule formation under high glucose conditions was recovered by curcumin (Figure 1(c)). Furthermore, the decreased mRNA and protein expression of osteogenic markers in high glucose conditions was recovered by curcumin administration (Figures 1(d)-1(f )). Tese fndings revealed that curcumin promoted BMSCs osteogenic diferentiation in high glucose.

Curcumin Recovered the High Glucose-Impaired Proangiogenic Ability of BMSCs.
Based on the results of the EdU test, HUVECs cultured in CM HG showed decreased proliferation compared with cells cultured with CM NG ; however, CM HG + Cur treatment signifcantly reversed the decrease in proliferation induced by CM HG (Figures 2(a) and 2(b)). In addition, the migration capacity of HUVECs was suppressed after stimulation with CM HG , and the migration area was signifcantly increased in the CM HG + Cur group compared with the CM HG group (Figures 2(c) and 2(d)). Moreover, CM HG markedly impaired the tube formation ability of HUVECs compared to that of the CM NG group; however, CM HG + Cur treatment signifcantly rescued the tube formation of HUVECs cultured with CM HG (Figures 2(e)-2(g)).

Curcumin Promoted the Angiogenesis Potential of BMSCs by Increasing VEGF Expression and Secretion.
In the present study, we observed markedly reduced VEGF protein expression in BMSCs exposed to HG compared with cells exposed to NG, whereas the HG-induced suppression of VEGF expression was reversed by curcumin (Figures 3(a) and 3(b)). Moreover, the production of VEGF in the CM from high glucose-induced BMSCs was much lower than that in the normal glucose group and that curcumin pretreatment increased the concentration of VEGF from high glucose + curcumin group (Figure 3(c)). To further investigate whether VEGF mediated curcumin-promoted angiogenesis in high glucose conditions, VEGF neutralizing antibodies were used. As shown in Figures 3(d)-3(j), blocking VEGF remarkably suppressed the curcuminenhanced angiogenesis capacity of BMSCs in high glucose.

NF-κB Signaling is Involved in the Efects of Curcumin on BMSCs-MediatedOsteogenesis-Angiogenesis
Coupling. Western blot results showed that high-concentration glucose increased the phosphorylation levels of NF-κB p65 (p-p65) without changing the total NF-κB p65 levels (Figures 4(a) and 4(b)). However, curcumin decreased p-p65 levels promoted by high glucose (Figure 4(c)). In Evidence-Based Complementary and Alternative Medicine addition, high glucose upregulated the nuclear protein expression of NF-κB p65 in BMSCs, whereas curcumin treatment attenuated nuclear NF-κB p65 (Figure 4(d)). Immunofuorescence assays also found that the nuclear localization of the NF-κB p65 protein in curcumin-treated cells was weaker than that in high glucose-stimulated cells (Figures 4(e) and 4(f )).
We subsequently explored whether the coupling of osteogenesis-angiogenesis in high glucose could be impacted by modulating NF-κB signaling. Our results showed that cells cultured in high glucose supplemented with Bay117082, a specifc NF-κB inhibitor, exhibited decreased p-p65 activity compared with that of cells incubated with high glucose only (Figures 5(a) and 5(b)). However, ALP and      Evidence-Based Complementary and Alternative Medicine curcumin action on BMSCs-mediated osteogenesisangiogenesis couples partly through the NF-κB signaling pathway. Te microfl showed a reduced vessel network in the DM group while more vessel formation was detected in the DM + curcumin group (Figure 6(h)). Furthermore, the role of curcumin in type H vessel was investigated. Immunofuorescence staining showed that curcumin administration increased the number of type H vessels in DM mice (Figures 6(i) and 6(j)). Tese results revealed that curcumin treatment prevented diabetes-induced bone loss and promoted vessel formation in vivo.

Discussion
Some studies have proven that osteogenesis and angiogenesis coupling plays a vital role in the pathogenic progression of DOP [14][15][16][17]. In the study, we demonstrated that curcumin treatment could enhance BMSC-mediated osteogenesis and angiogenesis coupling in high glucose microenvironments. Mechanistically, the efects of curcumin on BMSC osteogenic determination and BMSCmediated angiogenesis were achieved, at least partially, by inhibiting the high glucose-activated NF-κB signaling pathway. Furthermore, we confrmed that curcumin treatment promoted bone regeneration and accelerated angiogenesis in a DM model. To our knowledge, this is the frst to confrm that curcumin prevents diabetes-induced bone loss by promoting BMSC-mediated osteogenesis and angiogenesis coupling.
BMSCs are a cell type which have self-renewal and multidirectional diferentiation potential. It is well known that osteogenesis ability of BMSCs plays a crucial role in bone repair and regeneration. However, increasing evidence has shown that the decline in the osteogenic function of BMSCs is a vital mechanism for diabetic osteoporosis. BMSCs derived from diabetic patients show a decreased osteogenic diferentiation ability [49]. In a high glucose microenvironment, advanced glycation end products (AGEs) induce Wnt/LRP5/β-catenin to inhibit the BMSCs osteogenic diferentiation [50]. Tese studies suggest that hyperglycemia can lead to a change in BMSCs diferentiation function, resulting in a decrease in the self-repair and regeneration ability of the bone tissue and related pathological changes in diabetic osteoporosis. Recently, an increasing number of reports have shown that curcumin possesses favorable properties on diabetic bone metabolism [32][33][34]. Li and Zhang founded that curcumin pretreatment could promote osteogenesis of BMSCs in high glucose and enhanced bone formation in diabetic rats by regulating the Keap1/Nrf2/HO-1 signaling pathway [3]. Another study showed that curcumin enhanced osteogenesis-related gene expressions and suppressed apoptosis in osteoblasts under high glucose microenvironment [35,36]. In this study, our results showed that curcumin treatment rescued high glucose-inhibited osteogenic diferentiation ability of BMSCs in vitro. Furthermore, our in vivo results also revealed that curcumin prevented bone loss in diabetic mice.
Te osteogenesis and angiogenesis coupling is crucial in the process of bone regeneration. Tere is increasing evidence that BMSCs promote vessel formation in the skeletal system based on the coupling between osteogenesis and angiogenesis.
BMSCs not only induce diferentiation into endothelial cells to form vascular-like tissue but also produce multiple proangiogenic growth factors to promote angiogenesis [51,52]. Moreover, the proangiogenic ability of BMSCs is regulated by many factors. Platelet-derived growth factor-B (PDGF-B) gene overexpression in BMSCs can enhance angiogenesis and promote the repair of bone defects [53]. BMSCs derived from OVX rats showed a decreased angiogenesis ability [11]. Catalpol, a natural iridoid glycoside, accelerates bone regeneration through improving the angiogenesis of BMSCs [13]. However, the infuence of curcumin on the ability of BMSCs to regulate angiogenesis has been rarely reported. In this study, we verifed that curcumin could rescue the high glucose-impaired angiogenic ability of BMSCs by elevating the expression and secretion of the proangiogenic factor VEGF. Besides, we proved that curcumin treatment promoted vessel formation in the DM model. Recently, type H vessels were discovered in mouse and human bone tissues that is characterized by the high expression of CD31 and endomucin (CD31 hi Emcn hi ). Type H vessels secrete factors such as HIF-1α, VEGF, and Notch that promote vessel assembly and bone formation [16,17]. Given the important role of type H vessels in coupling osteogenesis and angiogenesis, the efect of curcumin on type H vessels in DM mice was investigated. Our results showed a decreased number of type H vessels in DM mice, but curcumin administration increased the quantity of vessels.
Activation of the NF-κB pathway is involved in diabetes and osteoporosis [38][39][40]. Inhibiting overactivated NF-κB pathway promotes the diferentiation and mineralization of MC3T3 cells [54]. NF-κB inhibitor treatment rescues high glucose-dependent infammatory cytokine expression and PDLSCs osteogenic diferentiation [55,56]. In this study, we found that the activated NF-κB signaling pathway was inhibited by curcumin treatment in BMSCs. Moreover, Bay117082, an NF-κB inhibitor, was used to investigate the role of NF-κB in high glucose-induced BMSCs dysfunction. Our results showed that Bay117082 treatment reversed the glucose-inhibited osteoblastic diferentiation and angiogenesis of BMSCs in vitro. Taken together, the above results indicated that curcumin rescued the high glucose-impaired osteogenesis and angiogenesis coupling of BMSCs via inhibiting the overactivated NF-κB signaling pathway.

Conclusions
Our fndings reveal the efects of curcumin in promoting the BMSCs-mediated osteogenesis and angiogenesis coupling in high glucose conditions. Tese impacts are preliminarily considered to be via NF-κB signaling pathway inhibition. Furthermore, curcumin may become a potential drug to prevent and treat diabetic osteoporosis through promoting bone regeneration and vessel formation.

Data Availability
Te data used to support the fndings of this study are included within the article.

Conflicts of Interest
Te authors declare that they have no conficts of interest.