Isopsoralen-mediated suppression of bone marrow adiposity and attenuation of the adipogenic commitment of bone marrow-derived mesenchymal stem cells.

Osteoporosis (OP) increases the risk of bone fractures and other complications, and is thus a major clinical problem. In this study, we examined the effect of isopsoralen on the differentiation of bone-derived marrow mesenchymal stem cells (BMSCs) into osteoblasts and adipocytes, as well as bone formation under osteoporotic conditions. Primary femoral BMSCs isolated from C57BL/6 mice were used to evaluate the isopsoralen-mediated regulation of the expression of alkaline phosphatase (ALP), osteocalcin (OCN) and runt-related transcription factor 2 (RUNX2) during osteogenesis 2 weeks. We also examined the expression of peroxisome proliferator-activated receptor γ (PPARγ) and CCAAT/enhancer binding protein β (C/EBPβ) under adipogenic conditions for 1 and 2 weeks. In addition, ovariectomized (OVX) mice were used to examine the effects of isopsoralen on bone formation for 2 months. Finally, mammalian target of rapamycin complex 1 (mTORC1) signaling was examined under osteogenic and adipogenic conditions. We found that following treatment with isopsoralen, the expression levels of ALP, OCN and RUNX2 were upregulated, whereas those of PPARγ and C/EBPβ were downregulated. mTORC1 signaling was also inhibited in vitro and in vivo. In the OVX mice that were intragastrically administered isopsoralen, bone parameters (trabecular thickness, bone volume/total volume and trabecular number) in the distal femoral metaphysis were significantly increased and the adipocyte number was decreased. On the whole, our findings demonstrate that isopsoralen promoted BMSC differentiation into osteoblasts and suppressed differentiation into adipocytes.

Introduction

Osteoporosis is a bone disease which affects bone strength. It is known to increase the risk of bone fractures and other complications, and has become a major clinical problem that affects over 200 million individuals worldwide (1). Current strategies for the treatment of osteoporosis, include the administration of teriparatide, strontium and bisphosphonates, mainly aimed at reducing bone resorption (which is regulated by osteoclasts) and increasing bone formation (which is regulated by osteoblasts). However, these synthetic medicines have several side-effects that reduce their efficacy, such as jaw bone osteonecrosis induced by large doses of bisphosphonates (which act as inhibitors of bone resorption) or an increased risk of osteosarcoma induced by prolonged treament with teriparatide (2,3). Thus, healthier and safer agents that can be used in the prevention or treatment of bone-related disorders are highly desired.

Previous studies have indicated that the decreased bone volume in osteoporosis is associated with increased adipose tissue in bone marrow (4–6). Osteoblasts and marrow adipocytes differentiate from a common precursor, namely bone marrow-derived mesenchymal stem cells (BMSCs) (7). It has been previously demonstrated that compared with the osteogenic capability of mesenchymal stem cells (MSCs) derived from healthy women, those derived from post-menopausal women with osteoporosis have a much lower osteogenic potential (8,9).

Previous studies have demonstrated a large degree of plasticity of osteoblasts and adipocytes. Specifically, fully-differentiated osteoblasts derived from huam MSCs (hMSCs) are capable of dedifferentiation and transdifferentiation into adipocytes and vice versa (10,11). There is a reciprocal association and balance between the differentiation of adipocytes and osteoblasts (12).

The majority of previous data suggest that bone marrow adipose tissue simply plays the role of filling the marrow cavity and lacks hematopoietic function (13,14). However, Elbaz et al reported a lipotoxic effect from marrow adipocytes on osteoblast differentiation and function (15). Additionally, free-fatty acids released by adipocytes can inhibit osteoblast proliferation and induce osteoblast apoptosis (16).

Existing data demonstrate that an excessive amount of bone marrow adipocytes may be a significant negative risk factor for skeletal health (17). Accordingly, inhibiting adipocyte differentiation in bone marrow, while simultaneously accelerating osteogenesis may be a therapeutic approach for age-related osteoporosis.

The fruit of Psoralea corylifolia L. (P. corylifolia) is a widely used Chinese herbal medicine and is specifically used in the treatment of fractures, and bone and joint diseases. It has also used in the treatment of other disorders, such as skin diseases, cardiovascular diseases, tumors and asthma (18,19). It has been found that psoralen and isopsoralen (ISO) have growth inhibitory effects on transplanted tumors in nude rats with osteosarcoma; however, following the administration of high doses of psoralen and ISO, toxic reactions such as writhing, lassitude and hypo-activity were observed (19). Thus, the efficacy and safety of any agent in the treatment of osteoprosis should also be consided.

ISO is the main active ingredient extracted from the seeds of P. corylifolia (psoralen is its isomer). Previous studies have proposed a positive role for psoralen in promoting osteoblast differentiation (20), as well as its stimulatory effects on bone formation (21,22). However, only a few studies have evaluated the beneficial effects of ISO on bone development (23,24), and to the best of our knowledge, none have examined its underlying mechanisms of action.

In addition, to date, at least to the best of our knowledge, there are no studies available on the role and mechanisms of action of ISO in bone marrow adipogenesis. Thus, in the present study, by employing a mouse model of osteoporosis induced by ovariectomy along with ISO treatment, we aimed to examine the effects of ISO on bone marrow adipogenesis in vitro. Furthermore, we evaluated the effects of ISO on the differentiation of osteoblasts and adipocytes derived from BMSCs isolated from C57BL/6 mice in an effort to clarify the probable underlying cellular and molecular mechanisms that occur during this period. Our results may contribute to the development of novel therapeutic approaches for the treatment of bone-related diseases.

Materials and methods

Ethics approval

Ethics approval was provided by the Medical Ethics Committee of Southern Medical University, Guangzhou, China.

Materials and reagents

ISO (chemical structure shown in Fig. 1A) was obtained from Sigma (St. Louis, MO, USA; purity, 99%; molecular weight, 186.1635). Stock solutions of ISO were prepared in dimethyl sulfoxide (DMSO; Sigma) and stored at −20°C.

Figure 1

Isopsoralen (ISO) promotes bone-derived marrow mesenchymal stem cell (BMSC) differentiation into osteoblasts. (A) Molecular structure of ISO. (B) CCK8 assays showed that ISO did not significantly affect cell growth at the concentrations used (0.1-100 µM) following treatment for 2 days. (C–E) Alkaline phosphatase (ALP) staining showed that ISO treatment at concentrations of 5-20 µM increased ALP activity in BMSCs under osteogenic differentiation conditions for 2 weeks. *P<0.05, **P<0.01 and ***P<0.001 compared to the group without ISO (Con, control). Columns represent the means ± SD from 9 wells/group (B and E).

Animal model and animal feeding

Female C57BL/6 mice (aged 2 months; n=18) were provided by the Experimental Animal Center of Southern Medical University. The mice were randomly divided into the sham-operated, ovariectomized (OVX) or OVX plus ISO (n=6/group) groups. The mice in the OVX plus ISO group were intragastrically administered ISO at a dose of 20 mg/kg/day for 5 days prior to being subjected to ovariectomy and this was maintained for 2 months after the mice were subjected to ovariectomy. A previous study found that a significant amount of bone loss in vertebrae and femurs was observed in C57BL/6 mice at only 4 weeks following ovariectomy (25). Thus, ISO treatment was maintained for 2 months in this study. The mice were anesthetized with 1% pentobarbital sodium. The operation area was disinfected by iodophor. The back skin was then longitudinally cut step by step at the second lumbar level by one incision (1.5 cm in length) and 2 sides of the ovaries were removed. Ovary peripheral vascular flow was blocked by ligature using sutures (4-0#). The incision was then closed and the mice were allowed to recover from the anesthesia at room temperature. The mice in the sham-operated group only had some fat tissue around the ovaries removed. The mice were sacrificed by cervical dislocation. The femurs were then removed using ophthalmic scissors and fixed in 10% paraformaldehyde solution.

Bone marrow adiposity analyses

For bone marrow adiposity analyses, the distal portion of the femurs was fixed, decalcified and sectioned into 2-µm-thick sections, and subjected to hematoxylin and eosin staining according to standard histological protocols. To quantify proximal metaphyseal adipocyte parameters, we calculated the adipocyte number (AD#, per mm2). A uniform number of fields was screened in all sections by 3 authors (Jian Wang, Sheng-Fa Li and Ting Wang), starting 3 fields from the left end and 3 fields from the top endocortical surface, excluding adipocytes with disruption in the fields. To avoid any bias in the final analyses, all sections were interpreted while blinded, without knowledge of the groups (sham-operated, CON or CON + ISO). Images were obtained at ×20 magnification using an Olympus BX51RF stereomicroscope (Olympus).

Microcomputed tomography (μCT) analyses

Micro-CT (µCT 80; Scanco Medical, Bassersdorf, Switzerland) analyses were performed on the distal portion of the femurs. Distal femurs were selected for scanning and corrected for the CT value, a 70 kV scanning voltage, 30 W power, 429 µA current, and 20 µm scan thickness. Our analyses included various bone parameters: trabecular thickness (Tb.Th), bone volume (BV)/total volume (TV) and trabecular number (Tb.N). The assessment of bone microstructure was carried out according to the guidelines provided in the study by Bouxsein et al (26).

Isolation of BMSCs and cell culture

BMSCs were isolated from C57BL/6 mice (n=10; aged 4 weeks; also from the Southern Medical University). The mice were sacrificed using CO2, and the mouse femurs were dissected free of surrounding soft tissue. The bone marrow was flushed with α-MEM (Invitrogen, Carlsbad, CA, USA). The marrow content from 4 bones was plated in culture flasks containing BMSC growth medium [α-MEM containing 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 mg/ml streptomycin sulfate (Gibco, Auckland, New Zealand). We used a centrifuge (1,000 rpm for 5 min) to isolate the cells from the extra soft tissue. The cells were re-suspended and non-adherent cells were removed, and adherent BMSCs were cultured and expanded for further experiments. The cell culture medium were replaced every 3 days. The cells were seeded in 96-well plates and 6-well plates at densities of 1×104 and 1×106 cells/well, respectively, and cultured in a humidified atmosphere of 5% CO2 and 95% air at 37°C. After examing the levels of different proteins characteristic of BMSCs by western blot analysis (see below), we found that the cells were CD29+ and Sca-1+, and CD45− and CD11b−; this confirmed these cells were actually BMSCs.

For osteogenic differentiation, the BMSCs were grown to 90% confluence in 6-well-plates. The culture medium was replaced with osteogenic medium (α-MEM supplemented with 15% fetal calf serum (FCS) plus 1% penicillin/streptomycin, 100 nM dexamethasone, 50 µg/ml ascorbate-2-phosphate and 10 mM β-glycerol phosphate). The medium was changed every 3 days.

For adipocyte-induced culture, the putative BMSCs were harvested and seeded into 6-well cell culture plates. Upon attaining 90% confluency, the cells were treated with adipogenic differentiation medium: BMSC maintenance medium supplemented with 1 µM dexamethasone, 100 µM indomethacin and 500 µM 1-methyl-3-isobutylxanthine (all from Sigma). For the initial induction, 5 µg/ml bovine insulin (Sigma) were added together with adipogenic differentiation medium. The medium was changed every fourth day.

Cell proliferation assays

Primary BMSCs were seeded in 96-well plates at a density of 1×104 cells/well. Following culture for 2 days, the cells were treated for 48 h with ISO at concentrations of 0, 0.1, 1, 10, 100 and 1,000 µM. Cell proliferation assays were performed using Caspase-8 Colorimetric Assay kits (KeyGEN Biotech, Nanjing, Jiangsu, China) according to the manufacturer's instructions. The absorbance in the wells was measured at 450 nm.

Alkaline phosphatase (ALP) staining assay

The BMSCs were cultured in the presence of osteogenic inducers and ISO (0, 5, 10 and 20 µM) for 2 weeks. ALP activity was also evaluated in the cells stained using an ALP staining kit (Beyotime, Nanjing, Jiangsu, China) according to the manufacturer's instructions. For ALP staining, the cells were washed in phosphate-buffered saline (PBS), fixed in 4% paraformaldehyde for 20 min at room temperature, and rinsed in distilled water. The ALP staining mixture was added for 30 min at room temperature in the dark. The cells were rinsed in distilled water and PBS to reduce non-specific staining.

Oil Red O staining

The BMSCs were cultured in the presence of adipogenic inducers and ISO (0, 5, 10 and 20 µM) for 7 and 14 days. Fat droplets that formed in differentiated adipocytes from the BMSCs were observed using Oil Red O staining. The cells were fixed in 4% formaldehyde for 15 min at room temperature, washed in PBS, and stained with 0.6% (w/v) Oil Red O solution (60% isopropanol, 40% water) for 1 h at 37°C. The cells were washed with PBS to remove unbound dye and isopropyl alcohol (1 ml) was added to the culture plates.

Immunocytochemical staining

The murine BMSCs (104 cells/well) were inoculated into 12-well assay plates with collagen-coated coverslips. Various concentrations of ISO (0, 5, 10 or 20 µM) were added at room temperature. The cells were fixed in 4% paraformaldehyde for 15 min at room temperature, rinsed with PBS, and incubated for 1 h with blocking buffer (1X PBST, 0.5% Triton X-100) supplemented with 5% serum from the same species as the secondary antibody. The blocking solution was removed and primary mouse monoclonal antibodies against osteocalcin (OCN; Cat. no. sc-365797; 1:100) and proliferator-activated receptor γ (PPARγ; sc-7273; 1:200) (both from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) were added followed by incubation overnight at 4°C. Subsequently, the secondary antibody, goat anti-mouse IgG (SAB4600004; 1:200; Sigma) was added, followed by incubation for 1 h at room temperature. The secondary antibody was aspirated and the wells were washed 3 times with PBST. DAB (Sangon Biotech, Shanghai, China) was added followed by incubation for 5 min at room temperature in the dark. The slides were restained with hematoxylin (Sigma). Images were acquired using an Olympus BX51RF stereomicroscope (Olympus, Tokyo, Japan). The mean optical density (MOD) was examed using Image-Pro Plus software (IPP; Media Cybernetics, Rockville, MD, USA).

Immunofluorescence staining and microscopic analyses

The BMSCs were grown on collagen-coated coverslips, washed with PBS, fixed for 15 min at room temperature in 4% para-formaldehyde, washed, permeabilized with 0.5% Triton X-100 (5 min, room temperature) and blocked for 30 in blocking buffer (5% FBS in PBST). The cells were sequentially probed with primary antibodies against runt-related transcription fact or 2 (RUNX2; Cat. no. 12556; 1:100) and CCAAT/enhancer binding protein β (C/EBPβ; Cat. no. 3802; 1:200) (both from Cell Signaling Technology, Danvers, MA, USA), and washed 3 times with PBST. The FITC-conjugated sheep anti-rabbi t IgG (Cat. no. F5137; 1:500) and TRITC-conjugated sheep anti-mouse IgG (Cat. no. T5393; 1:500) (both from Sigma) were added followed by incubation for 1 h at room temperature, followed by PBST washes. The cells were incubated with 4′,6-diamidino-2-phenylindole (DAPI; 1 µg/ml) for 15 min and then rinsed 3 times with PBST. Immunofluorescence staining was also performed on the femur histological sections using a standard protocol. Following ovariectomy and ISO treatment, the mice were then sacrifced. The femurs were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The fixed bones were decalcified by immersion in 10% EDTA (pH 7.0) for 14 days at room temperature and embedded in paraffin. The paraffin-embedded longitudinal bone sections (5-µmp-thick) were blocked in 5% goat serum for 1 h at room temperature, and incubated overnight with the primery antibodies. The sections were then incubated with secondary antibodies for 1 h at room temperature. Femur histological sections were imaged using a FV1000 confocal microscope and positive cells were evaluated using the IPP software program. The cells and femur histologiocal sections were imaged using a laser-scanning confocal microscope (FV1000; Olympus). Finally, MOD was measured using IPP software.

Western blot analysis

Western blot analysis was performed using a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) electrophoresis system. Protein samples (20 µg) were resuspended in reduced sample buffer, electrophoresed on a 7.5-10% Tris gel with Tris running buffer, blotted onto PVDF membranes, and sequentially probed with primary antibodies against RUNX2 (Cat. no. 12556; 1:1,000; Cell Signaling Technology), OCN (Cat. no. sc-365797; 1:1,000; Santa Cruz Biotechnology, Inc.), eukaryotic translation initiation factor 4E-binding protein 1 (4E/BP1; Thr37/46; Cat. no. 9644; 1:1,000; Cell Signaling Technology), phospho-S6 ribosomal protein (P-S6; S235/S236; Cat. no. sc-293143; 1:1,000), PPARγ (Cat. no. sc-7273; 1:1,000) (both from Santa Cruz Biotechnology, Inc.), C/EBPβ (Cat. no. 3802; 1:2,000), Sca-1 (Cat. no. 9664; 1:4,000), CD29 (Cat. no. 4706; 1:2,000), CD45 (Cat. no. 13917; 1:6,000), CD11b (Cat. no. 14271; 1:3,000) and β-actin (Cat. no. 3700; 1:2,500) (all from Cell Signaling Technology). Horseradish peroxidase-conjugated goat anti-rabbit (Cat. no. SAB4600223; 1:1,000) or anti-mouse (Cat. no. SAB4600004; 1:1,000) antibodies (both from Sigma) were added, and secondary antibodies were detected using enhanced chemiluminescence (ECL Plus; General Electric Healthcare, Milwaukee, WI, USA).

Statistical analyses

We used one-way analyses of variance (ANOVA) to analyze the data. The homogeneity of the variance tests was used to evaluate data homogeneity (IBM SPSS Statistics 19.0 software). If the variance was confirmed to be equal, least-significant difference tests were used for data analyses. If the variance was determined to be unequal, Dunnett's T3 tests were used. The results are presented as the the means ± standard deviation (SD). A value of P<0.05 was considered to indicate a statistically significant difference.

Results

We examined the effects of ISO on the proliferation of primary mouse BMSCs using CCK8 assays. As shown in Fig. 1B, ISO did not significantly affect cell growth at concentrations of 1–100 µM after 48 h. However, BMSC proliferation was significantly inhibited following treatment with 1,000 µM ISO.

ALP staining in vitro

To determine whether ISO affects BMSC differentiation into osteoblasts, ALP staining (Fig. 1C–E) was performed as increased ALP activity is an important indicator of osteoblast differentiation. We found that ISO enhanced ALP activity in the primary BMSCs in a dose-dependent manner (Fig. 1C–D). The most significant effects were observed at 2 weeks after osteogenic induction with a concentration of 20 µM ISO (Fig. 1E). These results suggested that ISO signifi-cantly stimulated BMSC differentiation into osteoblasts.

Oil Red O staining in vitro

ISO inhibited the adipocytic differentiation of BMSCs induced by adipogenic inducers in a concentration-dependent manner. As shown in Fig. 2A–F, fewer lipid droplets appeared in the cytoplasm of adipocytes that were stained with Oil Red O in the cells treated with ISO (Fig. 2E) under adipogenic differentiation conditions for 7 days (Fig. 2A, B and E). After 2 weeks of adipocytic differentiation, the adipogenic rate was at its lowest level in the presence of 20 µM ISO (Fig. 2C and D and F) compared to the controls (untreated cells; Fig. 2F). These data suggested that ISO inhibited BMSC differentiation into adipocytes.

Figure 2

Isopsoralen (ISO) inhibits bone marrow stromal cell (BMSC) differentiation into adipocytes. ISO (0, 5, 10 or 20 µM) was added to primary mouse BMSCs under adipogenic differentiation conditions. (A and B) Fewer lipid droplets appeared in the cytoplasm of adipocytes that were stained with Oil Red O in cells treated with ISO under adipogenic differentiation conditions for 7 days (E). (C and D) After 2 weeks of adipogenic differentiation, the adipogenesis rate was at its lowest level in the presence of 20 µM ISO (F). *P<0.05, **P<0.01 and ***P<0.001 compared to the group without ISO (Con, control). Columns represent the means ± SD from 9 wells/group (E and F).

OCN and PPARγ immunocytochemistry

Two weeks following osteoblast differentiation, OCN expression (brown) increased (Fig. 3A and C), with a maximum level observed in the cells treated with a concentration of 20 µM ISO (Fig. 3C). Under adipogenic conditions for 2 weeks, the generation of PPARγ (brown) was significantly reduced (Fig. 4A and D). The lowest expression of PPARγ was noted in the presence of 20 µM ISO (Fig. 4A and D).

Figure 3

Isopsoralen (ISO) stimulates bone-derived marrow mesenchymal stem cell (BMSC) differentiation into osteoblasts in vitro. ISO (0, 5, 10 or 20 µM) was added to primary mouse BMSCs under osteogenic differentiation conditions for 2 weeks. (A) Osteocalcin (OCN) immunocytochemistry and (B) runt-related transcription factor 2 (RUNX2) immunofluorescence were performed to examine the expression of osteoblast-specific proteins following treatment with ISO. (A) ISO increased OCN expression (brown) in BMSCs in a dose-dependent manner, with more pronounced effects at 20 µM ISO (C). (B) ISO dose-dependently enhanced RUNX2 (red) activity in BMSCs, particularly at a concentration of 20 µM (D). *P<0.05, **P<0.01 and ***P<0.001 compared to the group without ISO (Con, control). Columns represent the means ± SD from 9 wells/group (C and D).

Figure 4

Isopsoralen (ISO) inhibits bone-derived marrow mesenchymal stem cell (BMSC) differentiation into adipocytes in vitro. ISO (5, 10 or 20 µM) was added to primary mouse BMSCs following adipogenic differentiation. (A) Peroxisome proliferator-activated receptor γ (PPARγ) immunocytochemistry and (B) CCAAT/enhancer binding protein β (C/EBPβ) immunofluorescence assays were performed to examine the expression of adipocyte-specific proteins following treatment with ISO. (A) ISO increased PPARγ expression (brown) in BMSCs in a dose-dependent manner, with the most pronounced effect at 20 µM (C). (B) ISO dose-dependently enhanced C/EBPβ activity (green) in BMSCs, particularly at 20 µM (D). *P<0.05, **P<0.01 and ***P<0.001 compared to the group without ISO (Con, control). Columns represent the means ± SD from 9 wells/group (C and D).

RUNX2 and C/EBPβ immunofluorescence

After 2 weeks of osteogenic differentiation, RUNX2 expression (red) in the osteoblasts was upregulated (Fig. 3B and D), and ISO promoted RUNX2 expression in a dose-dependent manner. In the presence of 20 µM ISO, RUNX2 expression was at its maximum level (Fig. 3D). Additionally, after 2 weeks of adipogenic differentiation, C/EBPβ (green) expression decreased (Fig. 4B and C). The minimum density of C/EBPβ staining occurred in the presence of 20 µM ISO (Fig. 3C).

RUNX2 immunofluorescence in vivo

RUNX2 is very important for bone remodeling, and functions by regulating the differentiation of osteoblasts (27,37). The intensity of positive RUNX2 staining (red) was reduced in the OVX group (Fig. 5A), whereas it was significantly elevated in the OVX + ISO group (Fig. 5A and C).

Figure 5

Isopsoralen (ISO) promotes bone-derived marrow mesenchymal stem cell (BMSC) differentiation into osteoblasts and inhibits differentiation into adipocytes 2 months following ovariectomy (ovariectomized mice, OVX group). (A) Runt-related transcription factor 2 (RUNX2) immunofluorescence in the distal femur was elevated by ISO treatment. (B) Peroxisome proliferator-activated receptor γ (PPARγ) immunofluorescence in the distal femur was inhibited by ISO. *P<0.05, **P<0.01 and ***P<0.001 vs. OVX. Columns represent the means ± SD from 6 mice/group (C and D). Sham, sham-operated.

PPARγ immunofluorescence in vivo

Increased PPARγ (green) expression was observed in the OVX group compared with the OVX + ISO group (Fig. 5B and D). PPARγ expression was markedly increased in mice in the OVX group (Fig. B, panel b2 and D) compared to the mice in the OVX + ISO group (Fig. 5B, panel b3 and D). In addition, there were also statistically significant differences between the sham-operaged group (Fig. 5B, panel b1 and D) and the OVX group. PPARγ is considered indispensable for adipocyte differentiation.

Bone histomorphometries

Representative hematoxylin and eosin images of bone from the OVX and OVX + ISO groups (Fig. 6A) indicated that ISO reduced the number of adipocytes (AD#) in the bone marrow (Fig. 6A and G).

Figure 6

Isopsoralen (ISO) decreases the rate of bone loss following 2 months of treatment. (A) Hematoxylin and eosin staining indicated that ovariectomy (ovariectomized mice, OVX group) increased the number of adipocytes (AD#), whereas ISO decreased the number of adipocytes (G). (B and C) Micro-CT morphometry of murine distal femora shows that the (D) trabecular thickness (Tb.Th), (E) bone volume/total volume (BV/TV), and (F) trabecular number (Tb.N) values in the OVX group were decreased compared to the sham-operated and OVX plus ISO groups. These three indicators were significantly improved by ISO treatment compared to OVX mice. Compared to sham-operated (Sham) group the (D) Tb.Th, (E) BV/TV and (F) Tb.N levels in the OVX + ISO group were also decreased. *P<0.05, **P<0.01 and ***P<0.001 vs. OVX; +P<0.05 vs. OVX + ISO. Columns represent the means ± SD from 6 mice/group (D–G).

Micro-CT analyses

We performed µCT scans on the mouse distal femurs (Fig. 6B–F). Two months after the surgery, bone turnover was significantly decreased in the OVX mice compared with the sham-operated group (Fig. 6D–F). Treatment with ISO treatment significantly rescued this bone loss. The values for Tb.Th (Fig. 6D), BV/TV (Fig. 6E) and Tb.N (Fig. 6F) in the OVX group were markedly decreased compared with the sham-operated group. Following treatment with ISO for 2 months, these indicators were significantly improved (Fig. 6D and E). However, compared to the sham-operated group, the levels of Tb.N, Tb.Th and BV/TV were still decreased in the OVX + ISO group (Fig. 6D–F). This result reminded us of the fact that ISO also cannot be a perfect substitute for estrogen.

To determine whether ISO affects BMSC differentiation, we performed western blot analysis (Fig. 7). ISO promoted BMSC osteoblast differentiation in a dose-dependent manner, as demonstrated by the upregulation of the osteoblast-specific markers, OCN (Fig. 7A and G) and RUNX2 (Fig. 7A and H). OCN activity was increased in the cells treated with ISO for 2 weeks (Fig. 7A and G). ISO had the most significant effect on OCN expression at a concentration of 20 µM (Fig. 7G). In addition, the maximum expression of RUNX2 (Fig. 7A and H) was observed at this concentration of ISO, although ISO regulated RUNX2 expression in a dose-dependent manner (Fig. 7H).

Figure 7

Isopsoralen (ISO) supresses the adipogenic and promoted osteogenic differentiation of bone-derived marrow mesenchymal stem cells (BMSCs) by inhibiting the activation of the mTORC1 pathway in vitro. (A) ISO altered the expression of osteocalcin (OCN) (G) and runt-related transcription factor 2 (RUNX2) (H) during the osteoblastic differentiation of cultured BMSCs for 7 days. Under adipogenic differentiation conditions for 1 week, ISO inhibited peroxisome proliferator-activated receptor γ (PPARγ) (C and K) and CCAAT/enhancer binding protein β (C/EBPβ) expression (C and L). PPARγ (E and O) and C/EBPβ (E and P) expression was markedly decreased in the presence of ISO at day 14 in a dose-dependent manner (O and P). mTORC1 signaling was down-regulated in (B) the osteoblastic differentiation process and (D and F) under adipogenic conditions. (B and I) Eukaryotic translation initiation factor 4E-binding protein 1 (4E/BP1; Thr37/46) expression was upregulated and (B and J) phospho-S6 ribosomal protein (P-S6; S235/236) expression was decreased 2 weeks following osteogenic induction. The expression of (M and Q) 4E/BP1 (Thr37/46) and (N and R) P-S6 (S235/236) in adipogenic medium at (M and N) 7 days or (Q and R) 14 days was similar to that in BMSCs cultured under osteogenic conditions for 7 days. *P<0.05, **P<0.01 and ***P<0.001 compared to the group without ISO (Con, control). Columns represent the means ± SD from 9 wells/group (G–R).

PPARγ and C/EBPβ are adipocyte-specific markers. One week after adipogenic induction, PPARγ (Fig. 7C and K) and C/EBPβ (Fig. 7C and L) activity was decreased in the cells treated with ISO in a dose-dependent manner (Fig. 7K and L). We also found that the expression of PPARγ (Fig. 7E and O) and C/EBPβ (Fig. 7E and P) was markedly decreased in the presence of ISO at day 14 in a dose-dependent manner (Fig. 7O and P).

To explore the signaling pathways involved in the regulatory effects of ISO on BMSC adipogenesis, we assessed mammalian target of rapamycin complex 1 (mTORC1) signaling. It is notable that mTORC1 signaling plays an important role in PPARγ-mediated adipogenesis (28–30). We demonstrated that the phosphorylation of S6 (S235/236) was inhibited and the phosphorylation of 4E/BP1 (Thr37/46) was promoted by ISO. These proteins are direct downstream effectors of mTORC1, suggesting that ISO may prevent BMSC adipogenesis via the inhibition of mTORC1 signaling. To determine whether the ISO-induced BMSC differentiation was dependent on the inhibition of mTOR signaling (Fig. 7B, D and F), we measured the in vitro expression of 4E/BP1 (Thr37/46) (Fig. 7I, M and Q) and P-S6 (S235/236) (Fig. 7J, N and R). We observed an increased 4E/BP1 (Thr37/46) expression (Fig. 7B and I) and a decreased P-S6 (S235/236) expression (Fig. 7B and J) 2 weeks after osteogenic induction. Importantly, similar results were observed 7 days after the induction of adipocyte differentiation (Fig. 7D, M and N). In adipogenic differentiation conditions and in the presence of ISO for 2 weeks, we observed that the expression of 4E/BP1 (Thr37/46) (Fig. 7F and Q) was significantly upregulated and that of P-S6 (S235/236) remained suppressed (Fig. 7F and R). These results were observed in a dose-dependent manner (Fig. 7Q and R).

Discussion

The active components extracted from the seeds of P. corylifolia have been widely investigated, including psoralen, which has been shown to exert beneficial effects on skeletal health (20,31,32). Psoralen and ISO are the two main active ingredients of P.corylifolia fruit extracts. However, studies evaluating the protective effects of ISO against osteoporosis are limited, particularly studies that investigate its cellular and molecular mechanisms (19,23,33).

In this study, we demonstrated that 2 months of ISO treatment (20 mg/kg/day) increased the bone mass of the distal femoral metaphysis in OVX mice (Tb.Th, BV/TV and Tb.N). The above bone metabolism parameters (BV/TV, Tb.Th and Tb.N) are usually important indicators reflect the level of osteoporosis in vivo. Based on the close association between bone marrow adipogenesis and bone loss during the pathogenesis of osteoporosis, the effects of ISO on bone marrow were evaluated using histological and immunofluorescence analyses. In ISO-treated OVX mice, we observed increased trabecular bone in parallel with reduced adipose tissue in bone marrow. Moreover, an enhanced RUNX2 secretion and decreased PPARγ secretion were observed.

PPARγ and C/EBPβ are essential transcription factors for adipogenesis (34). Its expression and/or activity determines the commitment of BMSCs into the osteoblasts or adipocyte lineage (35). ALP, RUNX2 and OCN are key regulators of osteoblast differentiation and play an important role in bone formation (36). We used the above parameters as measurements of osteogenesis in the present study.

To the best of our knowledge, this is the first study to demonstrate a role for ISO in the promotion of osteogenesis and the attenuation of adipogenesis in vivo. This mechanism of action, which focuses on BMSCs, differs from the mechanisms of currently available agents for osteoporosis that target mature osteoblasts and osteoclasts (37).

We then found that the ISO-enhanced ALP activity is a dose-dependent osteogenic inducer of BMSCs that does not affect cell growth in bone marrow. This result is similar to that of previous studies on psoralen (20,38). As previously demonstrated, in patients with age-related osteoporosis, BMSCs in bone marrow preferentially differentiate into adipocytes rather than osteoblasts (8,39). Accordingly, in this study, BMSCs were cultured under adipogenic conditions, which mimics osteoporosis in humans. Our data indicated that ISO inhibited BMSC adipogenesis in a dose-dependent manner. The pro-osteogenic and anti-adipogenic effects of ISO on BMSCs were further confirmed by evaluating the expression levels of key osteogenic and adipogenic transcription factors. As we had hypothesized, our results indicated that treatment with ISO increased RUNX2 expression. By contrast, PPARγ expression was reduced by treatment with ISO. These findings were consistent with the data from our in vitro experiments.

Based on the observation that decreased numbers of osteoblastic cells occur concomitantly with increased fat content in bone marrow during aging and osteoporosis, studies have suggested a reciprocal association between the adipocyte differentiation and osteoblast differentiation of BMSCs (40–42).

As osteoblasts and adipocytes are both derived from BMSCs (43), we hypothesized that ISO may exert its anti-osteoporotic effects under low estrogen conditions by modulating the RUNX2/PPARγ balance, thereby inhibiting the differentiation of BMSCs towards adipocytes (7).

There are several limitations to the current study. First, we did not measure the serum lipid levels of mice in our in vivo experiments. However, previous studies have demonstrated that compared with other fat types, alterations in serum lipid levels do not significantly affect the role and function of fat in bone marrow (44,45). Additionally, although our results have indicated a protective effect of ISO on osteoporotic bone and bone marrow adipose tissue, we did not include a positive control group, such as strontium and estrogen, which prevent bone loss due to the inhibition of bone marrow adipogenesis (46,47).

In conclusion, the present study demonstrated that ISO attenuated bone marrow adipogenesis, which was indicated by increased RUNX2 levels and decreased PPARγ levels. This shifts BMSC lineage differentiation toward osteoblasts rather than adipocytes. Moreover, we suggest that mTORC1 signaling may be the underlying signaling pathway involved in this process (Fig. 8). Our data suggest that the naturally occurring agent, ISO, is safer and more cost-effective for use in the treatment of post-menopausal bone-related diseases.

Figure 8

Isopsoralen (ISO) supresses the adipogenic and promotes the osteogenic differentiation of bone-derived marrow mesenchymal stem cells (BMSCs) via the mTORC1 pathway.