Osthole Inhibits Osteoclast Formation and Enhances Bone Mass of Bone Marrow Mesenchymal Stem cells by Activating β-catenin-OPG Signaling Pathway

Summary Osthole has potential therapeutic applications due to its antiosteoporotic. Our study suggested that osthole attenuates osteoclast formation by stimulating the activation of β-catenin-OPG signaling and could be a potential agent to inhibit bone resorption. Introduction Osthole has potential therapeutic applications due to its antiosteoporotic. we performed study to test if OPG is the target gene of osthole-attenuated osteoclastogenesis. Methods In vivo, using 12-month-old male mice to evaluate the effect of osthole on bone mass. In vitro, Bone marrow stem cells (BMSCs) were isolated, extracted from 3-month-old C57BL/6J mice, 3-month-old β-catenin fx/fx mice, or 3-month-old OPG −/− mice and its littermates of OPG +/+ mice. Results we found that osthole signicantly increased the gene and protein levels of OPG expression in primary BMSCs dose-dependently. The deletion of the OPG gene did not affect β-catenin expression and the deletion of the β-catenin gene inhibited OPG expression in BMSCs, which indicated that osthole stimulated the expression of OPG through activation of β-catenin signaling. Conclusion Osthole attenuates osteoclast formation by stimulating the activation of β-catenin-OPG signaling and could be a potential agent to inhibit bone resorption.


Introduction
Osteoporosis is a systematic skeletal disease that thins and weakens the bones to the point that they become fragile and break easily, is one of the most disabling consequences of aging [1][2] . Hip fractures and vertebral fractures are strongly associated with reductions in BMD, and have been considered the prototypical osteoporotic fractures [3] . In 2010, an estimated 2.7 million hip fractures occurred worldwide, and half of them (51%) were considered preventable [4][5] . However, the incidence of all other fractures (non-hip, non-vertebral) is numerically much greater and collectively these fractures result in much larger economic costs for the population [6] . Fractures show symptoms of pain and an inability to bear weight, almost always require surgical xation [7] . In addition, the patient's functional status and quality of life are reduced, with a high risk for short-term mortality, as well as a lot of direct medical expenses [8][9] .
Estrogen replacement therapy are effective in increasing osteoblast activity, but resulted in the increased incidence of breast and uterine cancer [10][11] . Phytoestrogens have attracted attention to their potential impacts in the prevention and treatment for osteoporosis. Osthole, a coumarin derivative extracted from Cnidiummonnieri and Angelica of Chinese herbal medicine, has estrogenic effect in preventingagainst bone loss in ovariectomizedrat [12][13] . Numerous studies have con rmed a wide range of pharmacological activities of osthole in humans, such as anti-cancer activity, antihypertensive, antiarrhythmic, anti-in ammatory, anti-infection properties and promote osteoblasts differentiation [14][15][16] .
Bone marrow stem cells (BMSCs) have differentiation into osteogenic, fat, cartilage and nerve-like cells, and it have strong in vitro expansion capacity, as well as the potential for multi-directional differentiation [17][18] . Recent technological advances in cell labeling and tracing have facilitated the study of underlying mechanisms [19][20] . Used lineage tracing methods and proposed that bone marrow cells expressing Mx1 have all the known characteristics of BMSCs. These cells respond to tissue stress and migrate to sites of injury, supplying new osteoblasts during fracture healing [21] . In addition, Leptin Receptor (LepR) is another marker for identifying BMSC. LepR positive cells have been shown to produce osteoblasts and adipocytes in bone marrow [22] . Besides, cells expressing gremlin-1 have been isolated from bone marrow, and these cells are capable of bone formation, rather than adipogenesis [23] . This also makes BMSCs widely used in clinical research at the cellular level of bone and cartilage tissues, including cartilage repair.
Our previous study found that Osthole signi cantly stimulated osteoblast differentiation and bone formation by inducing the activation of Wnt/β-catenin-Bmp2 signaling [24] . And our previous metaanalysis of basic reports shows that BMSCs can promote bone call maturation, ossi cation and restore bone mechanical properties in osteoporotic fractures [25] . Although Osthole can promote bone formation, its effect on bone resorption and underlying mechanism remain unknown. In this study, we performed the in vivo and in vitro experiments to examine the effect of Ostholeon osteoclast formation.

Mice and reagents
All animal protocols were approved by Institutional Review Board of Longhua hospital, Shanghai University of Traditional Chinese Medicine (China). C57BL/6J wild-type mice, OPG knockout (KO) mice and OPG Wild-type (WT) mice were purchased from Shanghai Biomodel Organism Science & Technology Development Co.,Ltd(China). Osthole, with 98% purity, was purchased from the Shanghai Institute for Animal study Twelve12-month-old male mice were randomized into two groups, treatment group and control group, respectively. The treatment group was intervened with Osthole (5 mg/kg/day) by intraperitoneal injection once a day for 4 weeks, and the control group was intervened with vehicle (corn oil) by intraperitoneal injection once a day for 4 weeks. After sacri ced, the lumbar vertebrae were harvested for evaluation.

Micro-computed tomography (μCT) analysis
The fth lumbar vertebrae (L5) were scanned at 18-μm voxel size using the μCT scanner (μCT80, Scanco Medical AG, Bassersdorf, Switzerland). The trabecular bone under the growth plate was segmented using a contouring tool, and the contours were morphed automatically to segment the trabecular bone on all the one-hundred slices. The 3D images were reconstructed and analyzed with the evaluation software of the μCT system.

Histological and histomorphometrical assays
The third lumbar vertebrae (L3) were xed in 4% paraformaldehyde, decalci ed, dehydrated, cleared with dimethylbenzene, and then embedded in para n. At least 3 consecutive 7-μm sectionswere obtained from the coronal planes, and performed TRAPstaining for identifying osteoclasts. Histomorphometrical assay was performed to determine the number of osteoclasts andthe percentage of osteoclast surface by using an imageauto-analysis system (Olympus BX50; Japan).

Immunohistochemical staining
The para n sections of L3were depara nized by immersing the tissue in xylene, xing it with4% paraformaldehyde for 15 minutes, and treating it with0.5% Triton for 15 minutes, followed by xation with 4%paraformaldehyde for another 5 minutes.The sections were then incubated with rabbit anti-OPG monoclonal antibody(1:50dilution) and rabbit anti-β-catenin monoclonal antibody(1:50dilution), at 4 °C over night. Afterthorough wash, the slides were then incubated with a horseradish peroxidase (HRP)conjugated secondary antibody for 30 minutes. After being mounted, slides wereexamined by using an Image Analysis System (OlympusBX50, Japan).

Bone marrow stemcells (BMSCs) culture and treatment
Primary bone marrow stem cells, extracted from 3-month-old C57BL/6J mice, or 3-month-old OPG -/mice and its littermates of OPG +/+ mice, were cultured with incubation of M-CSF and RANKL for one week, and then treated with various doses (0.5-100 μM) of Osthole for 48 hours.

TRAP staining
Primary BMCswere seeded in 96-well plate at a density of 3x10 5 /ml. The cells were treated with M-CSF (30 ng/ml) and RANKL (30 ng/ml), at the present or absence of Osthole (100 µM).The medium was changed every 3 days. After 7-day-incubation, the cells was xed and performed TRAP staining to calculate the number of multi-nuclear (>=3 nucleus) osteoclasts.

Real-time qPCR analysis
Primary BMSCs extracted from 3-month-old C57BL/6J mices, were seeded in 6-well plates at a density of 1x10 6 cells/well. After 2-day culture, Cells were treated with variation doses of Osthole (1-100 μM) or Vehicle for 48 hours. Total cellular mRNA was isolated respectively using RNeasy Mini Kit. One microgram of total RNA was reverse-transcribed separately into cDNA using the iScriptcDNA synthesis kit. Quantitative polymerase chain reaction (qPCR) analysis was carried out using Absolute QPCR SYBR Green Master Mix in a total volume of 20 μl of buffered solution containing 1μl of the diluted (1:5) reverse transcription product in the presence of sense and antisense primers of target genes listed as Table 1. βactin is internal reference gene. Conditions were 15-min polymerase activation at 95 °C followed by 45 cycles, 95 °C for 20 s, 58 °C for 20 s and 72 °C for 30 s. All reactions were performed in triplicate independently.

Western-blotting analysis
Primary BMSCs isolated from 3-month-old OPG -/homozygous mice and its littermates of OPG +/+ mice, were seeded in 6-well plates at a density of 1x10 6 cells/well. Cells were treated with Osthole (100 μM) or Vehicle for 48 hours and cells lysates were, respectively, extracted with E-PER protein extraction reagents (Thermo Scienti c, Waltham, MA). Proteins were transferred onto a PVDF membrane (Bio-Rad, Hercules, CA) and the membrane was blocked with 5% non-fat milk in PBST solution for 1 h at room temperature (RT). After incubation with the primary antibody overnight at 4 °C and the HRP-conjugated secondary antibodies (Thermo Scienti c, Waltham, MA) for 1 h at RT, the protein expression was detected using a SuperSignal West Femto Maximum Sensitivity Substrate Kit (Thermo Scienti c, Waltham, MA). Rat anti-OPG monoclonal antibody and Rabbit anti-β-catenin monoclonal antibody were used as primary antibodies. Mouse anti-β-Actin monoclonal antibody was used as secondary antibody.

In vitro deletion of the -catenin gene
In vitro deletion of the β-catenin gene was performed as previouslydescribed [24] . BMSCs isolated from 3month-old β-catenin fx/fx mice were seeded in 6-well culture plates at a density of 1 × 10 6 cells per well and cultured for6 days. Cells were infected with Ad-GFP or Ad-Cre (Titer:4 × 10 8 pfu/mL; Baylor College of Medicine, Houston, TX,USA) for 72 hours. Ad-GFP was used as a control and tomonitor infection e ciency. After recovery for 48 hours, cellswere treated with or without Osthole (100 uM) for 48 hours. Real-time qPCR assay was performed to examine the expression of β-catenin and OPG. All reactions were performed in triplicate independently.

Statistical analysis
Based on all experiments conducted independently at least three times, the date was expressed as mean ± SD and analyzed using SPSS 24.0 software and GraphPad Prism 8. We analyzed the statistically signi cant differences using Student's t test and one-way analysis of variance. ImageJ software was employed to measure the grayscale analyses. In gures,*P < 0.05 or lower was considered statistically signi cant.

Results
Bone loss was reduced in twelve-month-old mice Three-month-old and twelve-month-old C57BL/6J mice were used to evaluate bone mass. The µCT 3D image analysis on the forth lumbar vertebrae (L4) showed the loss of trabecular bone of twelve-monthold mice compared with that of three-month-old mice (Fig. 1A). The quantitative analysis showed that bone volume over total volume (BV/TV) and bone mineral density (BMD) of aged mice were signi cantly decreased (p < 0.05, Fig. 1B, 1C) as compared to those of young mice,suggesting that the aged mice displayed decreased bone mass.

Osthole Inhibited Osteoclast Formation In Aged Mice
TRAP stainingwas performedon sections of L5. Osthole decreased the TRAP-positive number of multinucleated osteoclast (Fig S1A, B), moreover, redcued the percentage of osteoclast surface in aged mice after treated with Osthole ( Figure S1C).These data showed that Osthole could inhibit osteoclast formation in aged mice. However, we found that Osthole could not inhibit osteoclast formation in OPG gene knockout mice ( Figure S1D,E), suggesting that effect of Osthole may be through the OPG signaling.
Osthole inhibited osteoclastogenesis in a dose-dependent and an OPG-dependent manner OPG/RANKL signaling has been shown to play an important role in osteoclast formation. To determine the mechanism of Osthole on suppressing osteoclast formation, we examined the effect of Osthole on the expression of OPG and RANKL. Osthole(10, 50, 100 µM) also signi cantly enhanced OPG mRNA expressionin a dose-dependent manner (P < 0.05) and Osthole at the dose of 100 µM had the maximum effect with 3.8-fold increase (Fig. 3A). In contrast, Osthole had no effect on the expression of RANKL (data not shown). We also examined the effect of Osthole on protein expression of OPG and found that Osthole signi cantly increased the protein level of OPG mRNA in a dose-dependent manner (Fig. 3B) To further determine if Osthole-inhibited-osteoclastogenesis is OPG dependent, BMCs were isolated from 3-month-old OPG −/− mice and its littermates of OPG +/+ mice, cultured with M-CSF (44 ng/ml) and RANKL (100 ng/ml), plus Osthole (100 µM) or vehicle for 7 days, then TRAP staining were performed and osteoclast number was quanti cated. As shown in Fig. 3C and 3D, Osthole signi cantly inhibited osteoclast formation in OPG +/+ mice(P < 0.05), in contrast, in OPG −/− mice it had no effect on the formation of osteoclast. These data suggested that Osthole inhibits osteoclastogenesis in an OPGdependent manner.

Osthole promoted the expression of OPG through activation of β-catenin signaling
Our recent studies have demonstrated that OPG expression could be activated by β-catenin signaling and Osthole could activate β-catenin signaling [24 , 26-27] . We rst examined OPG and β-catenin expression in vivo. The immunostaining data showed that Osthole signi cantly increased OPG protein level (Fig. 4A) and β-catenin protein level (Fig. 4B) using sections of L5 samples in aged mice. To further con rm if Osthole-induced-OPG expression is through β-catenin signaling, we performed in vitro study. Primary BMSCs were isolated from 3-month-old β-Catenin fx/fx mice, infected with Ad-Cre or Ad-GFP, and treated with or without Osthole at the dose of 100uM. After 2 days, the total mRNA was collected and the expression of OPG was detected using real-time PCR assay. We found that the deletion of β-Catenin by Ad-Cre infection signi cantly inhibited Osthole-induced expression of OPG (Fig. 4C). In contrast, we found that the deletion of OPG didn't affect Osthole-induced expression of β-Catenin protein (Fig. 4D). Taken together, these results indicated that Osthole promotes the expression of OPG through activation of βcatenin signaling.

Discussion
The present study discovered that Osthole inhibited bone resorption in aged mice and attenuated osteoclast formation through stimulatingthe activation of β-catenin-OPG signaling (Fig. 5). Osteoporosis is caused by the disorder of homeostasis between bone formation and bone resorption. Our previous results showed that Osthole has an e cacy to promote osteoblastic proliferation and differentiation as well as act on the bone metabolism [24] . Whether Osthole impacts on the function of osteoclasts is the aim of this study. Osthole was rstly used to intervene with the elderly wild-type mice. The µCT analysis showed that the treatment with Osthole for 4 weeks could signi cantly increase bone mass in senile mice. One of the important reasons for the occurrence of osteoporosis is the increased activity and quantity of osteoclasts. Thus, TRAP staining was performed and the results demonstrated that the number of osteoclasts obviously was decreased in senile mice after intraperitoneal injection with Osthole. In order to further con rm this effect, the dose-dependent effect of Osthole on osteoclasts,derived from primary BMSCs upon to M-CSF and RANKL, was studied. Our results showed that Osthole could obviously decrease the number of osteoclasts in a dose-dependent manner. We also performed pits assay on bone slices and found that Osthole could attenuate the activity of functional osteoclasts. Together, Osthole could inhibit osteoclast formation and osteoclast-involved bone resorptive activity.
We further investigated the mechanism of Osthole on osteoclast formation. Under normal physiological conditions, the resorption of cartilage and bone is essential for development and regeneration of skeleton [28] . Osteoclastogenesis is a complicated process regulated by nely orchestrated interactions between osteoclast precursors and osteoblasts/stromal cells in bone marrow environment [29][30] .Osteoblasts produce OPG, which is a decoy receptor for the receptor activator of RANKL. By binding RANKL, OPG inhibits the interaction between RANKL and the receptor activator of nuclear factor-kappa B (RANK), a receptor of RANKL, therefore, OPG/RANKL/RANK plays key roles in the process of osteoclastogenesis [31][32] . Our data showed that Osthole signi cantly promoted the expression of OPG in BMSCs with a dose-dependent manner, but did notshow obvious effect on RANKL expression. RANKL is generally essential for osteoclast formation and thought to be supplied by osteoblasts or their precursors [33] . In contrast, recent evidences have revealed that osteocytes express high levels of RANKL and contribute to the coupling of bone formation and bone resorption [34][35] .Here we found that Osthole did not have any effect on RANKL expression in BMSCs, indicating that it may generate the effect on the expression of RANKL in osteocytes.
To demonstrate if Osthole-inhibited osteoclastogenesis was OPG-dependent, we next performed a rescue experiment using BMSCs from OPG +/+ and OPG −/− mice. Osthole decreased the formation of osteoclastsin OPG +/+ mice, while it had no effect in OPG −/− mice. Therefore, our result indicated that Osthole-attenuated osteoclastogenesis was through increasing the expression of OPG in BMSCs and inhibiting the binding of RANKL and RANK, but not acting on RANKL and RANK directly. The underlying mechanism of Osthole on the regulation of OPG was further studied for OPG upstream signaling. It has been reported that the canonical Wnt pathway up-regulates the expression of OPG in osteoblasts and chondrocytes [36][37] . β-catenin, a key protein of the canonical Wnt pathway, is required to induce the expression of OPG in osteoblasts [38] . In addition, it is reported that polydatin improved the osteogenic differentiation of hBMSCs and maintained the bone matrix in the OVX mouse model through the activation of β-catenin pathway [39] . Melatonin promotes the BMP9-induced osteogenic differentiation of mesenchymal stem cells by activating the β-catenin signalling pathway [40] . But inactivation of β-catenin signaling in osteoblasts causedthe increased osteoclastogenesis due to insu cient production of OPG [41] . Wnt3a was unable to induce the production of normal OPG to inhibit bone resorption, while lacking of β-catenin in osteoclasticlineage [36,41] . Cellular and molecular studies have shown that βcatenin banding with TCF proteins regulates the expression of OPG in osteoblasts [42] . Here we found Osthole could signi cantly increase the expression of β-catenin and OPG in BMSCs. To further discuss the interactions between β-catenin and OPG, we performed the rescue experiment in vitro using β-catenin oxed mice and found that the deletion of the β-catenin gene could inhibit Osthole-induced OPG expression. This result demonstrated that Osthole stimulated the expression of OPG through the βcatenin signaling.
Canonical Wnt signaling is required for the differentiation of mesenchymal progenitors into osteoblasts, as well as importance to the connection between osteoblasts and bone metabolism [43][44] . Wnt3a is an important upstream gene to regulate this pathway. The proteins complex of Wnt3a, Frizzled8 and LRP5/6 inhibits the activity of GSK3β, which makes β-catenin to be phosphorylated and degraded.Then β-catenin is released and translocated to the nucleus, where it interacts with TCF/LEF transcription factors to activate the expression of target genes. Released β-catenin may also interact with site2 and site4 on the OPG promoter directly, to increase production and secretion of OPG and inhibit the binding of RANKL and RANK, thus reduces the formation and activity of osteoclast [42] . It has been reported that BMP2 can also increase the expression of OPG by up-regulating Wnt3aexpression and promoting Samd1/4 to interact with site2 and site4 on OPG promoter [45] . In contrast, Wnt5a activates the non canonical Wnt signaling, which increases the activity of RANKL to promote osteoclastogenesis [47] . Our previous study showed that Osthole could signi cantly stimulate the expression of Wnt3a, but not Wnt5a [24] .

Conclusion
Based on the above discussions, we concluded that Osthole inhibited osteoclast formation and bone resorption through stimulating the activation of Wnt3a/β-catenin-OPG signaling (Fig. 6).