Osthole stimulates bone formation, drives vascularization and retards adipogenesis to alleviate alcohol-induced osteonecrosis of the femoral head.

Abstract Characteristic pathological changes in osteonecrosis of the femoral head (ONFH) include reduced osteogenic differentiation of bone mesenchymal stem cells (BMSCs), impaired osseous circulation and increased intramedullary adipocytes deposition. Osthole is a bioactive derivative from coumarin with a wide range of pharmacotherapeutic effects. The aim of this study was to unveil the potential protective role of osthole in alcohol‐induced ONFH. In vitro, ethanol (50 mmol/L) remarkably decreased the proliferation and osteogenic differentiation of BMSCs and impaired the proliferation and tube formation capacity of human umbilical vein endothelial cell (HUVECs), whereas it substantially promoted the adipogenic differentiation of BMSCs. However, osthole could reverse the effects of ethanol on osteogenesis via modulating Wnt/β‐catenin pathway, stimulate vasculogenesis and counteract adipogenesis. In vivo, the protective role of osthole was confirmed in the well‐constructed rat model of ethanol‐induced ONFH, demonstrated by a cascade of radiographical and pathological investigations including micro‐CT scanning, haematoxylin‐eosin staining, TdT‐mediated dUTP nick end labelling, immunohistochemical staining and fluorochrome labelling. Taken together, for the first time, osthole was demonstrated to rescue the ethanol‐induced ONFH via promoting bone formation, driving vascularization and retarding adipogenesis.

several autoimmune diseases. 1 Actually, alcoholism is one of the leading risk factors for ONFH worldwide, [3][4][5] and epidemiologic studies indicated that 20%-45% of ONFH patients were due to alcohol overuse. 6 Bone mesenchymal stem cells (BMSCs) are predominant precursor cells for bone regeneration and remodelling. 7,8 Accumulating evidence indicates that alcohol impairs bone homoeostasis. 9 Alcohol inhibits BMSCs differentiation into osteoblast; on the contrary, it promotes adipogenesis of BMSCs. 10,11 Additionally, alcohol critically restrains DNA synthesis and proliferation of osteoprogenitor cells. 12 Moreover, high concentration of alcohol has a detrimental effect on neovascularization. 13,14 Therefore, alcohol may cause ONFH via inhibiting osteogenic activity of BMSCs directly and impairing vascularization to alter bone homoeostasis indirectly.
Recently, traditional Chinese medicine (TCM) is used to tackle bone diseases with increasing interest. Previous studies have found several herbs, and their monomers could ameliorate alcohol-induced ONFH in animal models. 15,16 Osthole, a coumarin bioactive derivative, obtained from many medical plants, is commonly used as ingredients in herbal medicine and functional foods. 17,18 With the in-depth investigations, osthole is revealed to possess a wide range of different pharmacological effects, including hepatoprotective, vasorelaxant, antileishmanial, neuroprotective, spasmolytic and antimicrobial properties. [19][20][21][22][23][24][25][26] Notably, it was demonstrated that osthole significantly promoted the proliferation of osteoblast-like cells. 27 Extracellular signal-regulated kinase and β-catenin/BMP signalling pathway might be involved in the osteogenic differentiation stimulated by osthole. 28,29 Intriguingly, osthole could be as effective as 17β-estradiol in rescuing bone loss in ovariectomy-induced osteoporosis in rats. 30 Moreover, osthole could drive fracture healing via promoting endochondral ossification. 31 The beneficial effects of osthole imply that it might yield protective results for alcohol-induced ONFH. Herein, we demonstrated that osthole could stimulate bone formation, drive vascularization and retard adipogenesis, thus to alleviate alcohol-induced ONFH in the rat model.

| Cell culture
Bone mesenchymal stem cells were obtained from the femurs and tibias of 4-week-old Sprague-Dawley (SD) rats according to the previous method. 32 BMSCs were cultured in α minimum essential medium (α-MEM) (Gibco) with 10% foetal bovine serum (FBS) (Invitrogen), 1% penicillin/streptomycin (Invitrogen) at 37°C in a humidified atmosphere of 5% CO 2. The BMSCs used in all experiments were between three and seven passages. Human umbilical vein endothelial cells (HUVECs) were purchased from Procell and cultured in endothelial cell medium (ECM) (Gibco).

| Flow cytometry
Flow cytometry was used to test the surface markers of mesenchymal stem cells at passage three. In brief, adherent cells were washed with phosphate buffered saline (PBS) and disassociated by 0.05% trypsin (Gibco) and centrifuged at 350 g for 5 min. Next, the cells were rinsed twice with PBS and blocked with 1% FBS for 30 min at 4°C. Then, cells were incubated with fluorescein isothiocyanate (FITC)-conjugated monoclonal antibodies for rat CD29, CD90, CD105, CD31, CD34 and CD45 (Proteintech) for 1 hour at 4°C. The same amounts of cells without incubating any antibodies were used as the negative control. Finally, the cells were washed with PBS three times and analysed by flow cytometry.

| Multilineage differentiation of BMSCs
The multilineage differentiation capacity of BMSCs was detected according to the previous methods. 16 In brief, 2 × 10 5 of BMSCs were seeded in 6-well plates with α-MEM. After 80% confluence, osteogenic medium (Cyagen) was used to induce osteogenic differentiation for 21 days, and the mineralized nodules were stained by Alizarin red. Adipogenic medium (Cyagen) was used to induce adipogenic differentiation for 21 days, and the lipid vacuoles were stained by oil red O. Chondrogenic medium (Cyagen) was used to induce chondrogenic differentiation for 28 days, and the frozen sections of pellets were stained by Alcian blue.

| Cell proliferation
The effects of ethanol (50 mmol/L) and osthole (10, 50, 100 μmol/L) on BMSCs or HUVECs proliferation were tested by Cell Counting Kit-8 assay (Beyotime). The dose of ethanol was based on the previous work. 10 Briefly, cells were seeded in 96-well plates at 5 × 10 3 per well in 100 μL culture medium. To measure cell proliferation, 10 μL CCK-8 solution and 90 μL medium were added to each well, and then, the plates were incubated in 37°C for 2 hours. The absorbance values of supernatants at 450 nm were measured.

| Mineralization assay
In order to detect the roles of ethanol, osthole and Wnt/β-catenin pathway in osteogenic differentiation, BMSCs were incubated with ethanol (50 mmol/L), different doses of osthole (10, 50, 100 μmol/L) and JW74 (1 μmol/L; MCE, Monmouth Junction, NJ) for specific time. Briefly, 2 × 10 5 of BMSCs were seeded in 6-well plates with α-MEM. After 80% confluence, osteogenic medium (Cyagen) was used to induce osteogenic differentiation, 33 and the medium was refreshed every two days. Alizarin red staining was performed at day 21. ALP activity, which was measured using a microplate test kit at 560 nm, as well as ALP staining, was obtained at day 7 and 14.
Images were captured with a LEICA microscope.

| Adipogenesis assay
To investigate the effects of ethanol and osthole on adipogenic differentiation, BMSCs were incubated with ethanol (50 mmol/L) and osthole (10, 50, 100 μmol/L) for 21 days. In brief, 2 × 10 5 of BMSCs were seeded in 6-well plates with α-MEM. After 80% confluence, adipogenic medium (Cyagen) was used to induce adipogenic differentiation, and the medium was refreshed according to the manufacturer's protocol. Oil red O staining was performed at day 21, and images were captured with a LEICA microscope.
Briefly, 50 μL/well Matrigel (BD Bioscience) was added to the 96well plate. Subsequently, 4 × 10 4 HUVECs/well were seeded on the surface of Matrigel after gelatinization in 37°C for 30 min. The tube formation images were captured by a LEICA phase contrast microscope, and the number of complete capillaries and nodes of each hole were calculated.

| Quantitative real-time polymerase chain reaction (QPCR)
BMSCs and HUVECs were cultured in 6-well plates in different dissolution for 48 hours, and then, total RNA was extracted from each hole according to the manufacturer's protocol (TaKaRa).
Reverse transcriptase reactions contained the purified total RNA and 50 nmol/L RT primer. M-MLV reverse transcriptase (TaKaRa) was used. QPCR was performed by a SYBR Premix Ex Taq protocol (TaKaRa) on an MX3005P system. GAPDH was used for the internal control gene of RNA expression. The primers were listed in Table S1.

| Western blotting
Briefly, BMSCs and HUVECs were lysed with cell lysis buffer Then, the PVDF membranes were washed three times in TBST and incubated for 1 hour with an HRP-conjugated second antibody at room temperature. Finally, the membranes were washed three times and reacted with the SuperSignal West Pico kit (Thermo Scientific, Waltham, MA). Signals were quantified using scanning densitometry.

| ELISA
Human umbilical vein endothelial cells were cultured in 6-well plates for 48 hours, and then, cells were lysed with cell lysis buffer according to the manufacturer's protocol (Beyotime). Then, the proteins were diluted for analysis of VEGF content by ELISA (Neobioscience).
The absorbance values at 450 nm were measured and used to calculate the VEGF content according to the standard curve.

| Animal experiment
Thirty 8-week-old male SD rats were used for animal experiment with the approval from the Animal Research Committee at Shanghai Sixth People's Hospital. All rats were randomly and equally divided into three groups: (a) control group, (b) ethanol group and (c) ethanol + osthole group. All rats were allowed to adapt to the Lieber-DeCarli liquid diet for one week. For the next 6 weeks, in ethanol group, rats had an ethanol-containing Lieber-DeCarli liquid diet, containing 8% (w/v) ethanol (36% of daily calories). 34 The rats in the ethanol + osthole group had the same ethanol diet but were co-treated with osthole (100 mg/kg/d) by intraperitoneal injection, 17,18 and the rats in control group had the normal Lieber-DeCarli liquid diet without ethanol (the ethanol was displaced by maltodextrin, with the same amount of calories). All rats were free access to the diets which were refreshed every day. Fluorescence staining was performed to monitor dynamic bone formation and mineralization. In brief, 20 mg/kg tetracycline (Aladdin), 10 mg/kg calcein-AM (Aladdin) and 30 mg/kg alizarin red S (Aladdin) were injected intraperitoneally at week 0, 2 and 4 during the experiment.

| Micro-CT scanning and analysis
Rats were killed under general anaesthesia at the end of the 6th week, and bilateral femoral heads were obtained and fixed in 4% paraformaldehyde. All samples were scanned with a 9-micron voxel size micro-CT scanner (Skyscan 1176). The acquisition conditions are 35 kV of energy and 220 mA of intensity. The images were managed by CTVol software and reconstructed. Parameters, including bone mineral density (BMD), trabecular thickness (Tb.Th), trabecular number (Tb.N) and trabecular bone volume fraction (BV/TV), were calculated from the reconstructed images.

| Histological and immunohistochemical staining
After micro-CT scanning, specimens were decalcified with 10% EDTA for one month and then embedded in paraffin. The specimens were cut into 5-μm-thick sections and stained with haematoxylin and eosin (HE). For immunohistochemical staining, sections were deparaffinized, antigen retrieved, blocked and incubated with primary antibodies of COL I and OCN (CST) and relevant biotinylated secondary antibodies. Finally, sections were stained with DAB and counterstained with haematoxylin.

| Apoptosis assay
TdT-mediated dUTP nick end labelling (TUNEL) was used to detect the apoptosis within the femoral head. After the sections were deparaffinized and antigen retrieved, the TUNEL staining was performed according to the manufacture's protocols (Beyotime). Images were captured by LEICA DM 400 microscope.

| Statistical analysis
All data were presented as means ± standard error of the mean (SEM). The differences between groups were determined by Student's t test or one-way ANOVA with Bonferroni correction in SPSS 18 (IBM). *P < .05, **P < .01 and ***P < .001 were considered to be of statistical significance.

| Identification of primary BMSCs
Flow cytometry was used to determine the purity of BMSCs. The positive rates of markers including CD29, CD90 and CD105 were 99.6%, 98.9% and 99.5%, whereas the markers including CD31, CD45 and CD34 were negligible. Next, in order to detect the capacity of multilineage differentiation, BMSCs were used to differentiate into osteogenic,

| Osthole rescued the ethanol-induced inhibition on osteogenic differentiation of BMSCs via Wnt/β-catenin pathway
Firstly, CCK-8 assay was employed to detect the proliferation of BMSCs. Ethanol (50 mmol/L) could significantly inhibit the proliferation of BMSCs at day 5 and 7. However, osthole (10, 50, 100 μmol/L) was able to rescue the inhibitory effect of ethanol in a dose-dependent manner ( Figure 1A). Figure 1B Figure 2B,C). Wnt antagonist JW74 was used to investigate the role of Wnt/β-catenin signalling pathway in the osteoprotective function of osthole. 35 As shown in Figure 2D    Results were means ± SEM of four independent experiments in duplicate. C-E, The mRNA level of COL I, OCN and OPN was decreased in BMSCs after 48 h incubation with ethanol but was substantially up-regulated when treated with osthole. GAPDH was set as a normalization control. Results were means ± SEM of four independent experiments in triplicate. F, Runx2 was decreased by ethanol but was strengthened by osthole in BMSCs. Results were means ± SEM of four independent experiments. G, Immunofluorescence staining of COL I and OCN showed osthole reversed the anti-osteogenic effect of ethanol in BMSCs. BMSCs were cultured for 48 h in Osteogenic medium supplemented with ethanol and osthole. Cytoskeletons were stained with phalloidine (red), and the nucleus was stained with DAPI (blue) group. However, osthole could dose-dependently reverse the inhibition caused by ethanol ( Figure 4E). We also detected the level of VEGF by ELISA. VEGF was significantly decreased by ethanol and substantially rescued by osthole dose-dependently ( Figure 4F). In short, these data indicated that 50 mmol/L ethanol could impair tube formation in vitro whereas osthole could dose-dependently restore vasculogenesis capacity of HUVECs.

| D ISCUSS I ON
Alcohol consumption is considered as a vital risk factor for nontraumatic ONFH. 6,11,39 A myriad of epidemiologic studies indicated that a substantial number of ONFH patients were due to alcohol overuse. 40,41 Prior studies showed an increased fat deposition and vascular pressure but decreased osteogenic activity in alcohol-induced ONFH. 11,42 However, the natural history of found that the level of β-catenin was remarkably reduced by ethanol, in consistent with previous reports. 16,43,44 Intriguingly, osthole could dose-dependently counteract the inhibition of ethanol on the proliferation and osteogenic differentiation of BMSCs. In F I G U R E 5 Micro-CT scanning and analysis. A, Micro-CT scanning images of the femoral head divided by group and section. Subchondral trabecular bone was dramatically impaired in the ethanol group; however, the trabecular structure was largely restored in the ethanol + osthole group. B-E, BMD, Tb.N, BV/TV and Tb.Th were calculated based on reconstructed CT images. Results were means ± SEM of five specimen. BMD, bone mineral density; BV/TV, bone volume/tissue volume; Tb.N, trabecular number; Tb.Th, trabecular thickness addition, osthole could dose-dependently abolish the inhibition of ethanol on β-catenin expression to augment osteogenesis, in line with previous reports. 29,45 In vivo, typical pathological changes of hip osteonecrosis were observed by micro-CT scanning and H & E staining in the ethanol group, whereas only mild osteonecrosis was presented in the animal with supplementary osthole administration. There were more apoptotic cells in the ethanol group, reflecting apoptosis was an essential phenomenon in the development of ONFH. However, osthole might play an antiapoptotic role to ameliorate stress-induced tissue injury. 46 Intramedullary fat deposition is revealed as a critical pathological feature in the progress of ONFH, characterized by enlarged cellular size and increased number of adipocytes. [47][48][49] The strengthened capacity of adipogenesis in ethanol-induced ONFH has been revealed in several animal models. [50][51][52] Notably, promoted adipogenic differentiation led to disordered osteogenic differentiation, and F I G U R E 6 Osthole ameliorated ethanol-induced ONFH in the rat model. A, H & E staining indicated obvious osteonecrosis in the ethanol group. Empty lacunae in the subchondral trabeculae (black arrows) were observed in the ethanol group. B, TUNEL showed increased apoptosis in the ethanol group, which was alleviated by osthole treatment. The TUNEL positive cells were indicated in the trabeculae of the femoral head (red arrows). C, Immunohistochemical staining of COL I and OCN. Fewer cells in the trabeculae (black arrows) were positive for COL I and OCN in the ethanol groups; however, the staining was substantially improved with osthole treatment the accumulation of adipocytes might lead to a high intramedullary pressure and impair brittle blood circulation in the femoral heads. 50 Moreover, fat emboli in vessels and sinusoids can also indirectly induce intravascular coagulation via the complement signalling pathways and deposition of immune complex, whereas antithrombotic reagents have protective effects to promote circulation and ameliorate ONFH. 53,54 In this work, we found that ethanol could stimulate BMSCs towards adipogenic differentiation. However, it was demonstrated that osthole could augment osteogenesis and retard adipogenesis in the scenario of ethanol administration. The result is consistent with the previous report that osthole could prevent alcohol-induced fatty liver. 19 Therefore, the beneficial effect of osthole on ethanol-induced ONFH might function via inhibiting the formation of adipocytes, decreasing the number of fat emboli and improving perfusion in the femoral heads.
Currently, it is recognized that ONFH is not merely a kind of skeletal disease, but involves a series of systemic pathologies, especially intraosseous circulation. [55][56][57] ONFH might be considered as a vascular disease with attenuated blood flow and a specific bone disease with changed marrow cell differentiation and function. 2 In addition, previous reports indicated that high dose of alcohol exerted a detrimental effect on angiogenesis by regulating the expression of angiogenic-related genes of VEGF, hypoxia-inducible factor (HIF) and fibroblast growth factor (FGF). 13,14 Notably, VEGF has a vital role in the repair process for bone regeneration. 58 In our study, we found that 50 mmol/L ethanol treatment could significantly inhibit the proliferation and tube for-

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.