Bone 45 (2009) 534544

Contents lists available at ScienceDirect

Bone
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b o n e

Epimedium-derived avonoids promote osteoblastogenesis and suppress adipogenesis in bone marrow stromal cells while exerting an anabolic effect on osteoporotic bone
Peng Songlin a,b,1, Zhang Ge a,1, He Yixin a, Wang Xinluan a, Leung Pingchung a, Leung Kwoksui a, Qin Ling a,
a b

Musculoskeletal Research Laboratory, Department of Orthopaedics and Traumatology, the Chinese University of Hong Kong, Shatin, N.T. Hong Kong SAR, China Department of Orthopaedics and Traumatology, the University of Hong Kong, Hong Kong SAR, China

a r t i c l e

i n f o

a b s t r a c t
Background: Epimedium-derived avonoids (EFs) have been reported to prevent bone loss in ovariectomized (OVX) rats and late postmenopausal women but the underlying mechanism of the anabolic effect is unknown. Objective: This study aimed to investigate the effect of EFs on osteoporotic bone using histomorphometry and on osteoblastogenesis/adipogenesis of bone marrow stromal cells (BMSCs). Methods: 11-month-old female Wistar rats were divided into Sham, OVX, Sham + soluble vehicle (Sham + SV), OVX + SV and OVX + EFs (10 mg/kg/day) groups. 3 months after surgery, rats from the rst two groups were euthanized to verify the establishment of OVX-induced osteoporosis. Other groups were orally treated with either daily SV or EFs for 4 months. At sacrice, serum was analyzed for the levels of osteocalcin and TRACP 5b, BMD in the proximal femur was measured by pQCT. Static and dynamic bone histomorphometry was performed in proximal tibiae with microCT and undecalcied sections, respectively. The effect of EF treatment on differentiation of rat BMSCs was assessed by colony formation assays and gene expression analysis, respectively. Gene expression, ALP activity and adipocyte numbers were determined in differentiating human BMSCs after exposure to conditioned serum from SV- or EFs-treated OVX rats. Results: The serum level of osteocalcin was higher and TRACP 5b was lower in EFs versus SV-treated OVX rats. BMD, BV/TV, Tb.N and Conn.D in EFs-treated OVX rats were signicantly greater than those of SV-treated OVX rats. Bone histomorphometric parameters OS/BS, MAR, and BFR/BS were signicantly higher in EFs versus SVtreated OVX rats. EFs signicantly increased osteogenesis and decreased adipogenesis of BMSCs, as evidenced by CFU-ALP and CFU-Adipo assays, respectively. The mRNA level of Runx2 and bone sailoprotein was signicantly higher while PPAR2 was signicantly lower in BMSCs from EFs-treated versus SV-treated OVX rats. ALP activity and Runx2 mRNA was signicantly higher while adipocyte number and PPAR2 mRNA was signicantly lower in human BMSCs after exposure to conditioned serum from EFs versus SV-treated OVX rats. Conclusion: EFs exerted anabolic effect on osteoporotic bone by concomitantly promoting osteogenic and suppressing adipogenic differentiation of BMSCs. 2009 Elsevier Inc. All rights reserved.

Article history: Received 21 February 2009 Revised 13 May 2009 Accepted 21 May 2009 Available online 6 June 2009 Edited by: B. Olsen Keywords: Osteoporosis Phytoestrogen Histomorphometry Stem cell Anabolic effect

Introduction Postmenopausal osteoporosis is a disorder in which there is net bone loss and an increased risk of bone fracture because of an imbalance between osteoclast-mediated bone resorption and osteoblast-mediated bone formation [1]. The ideal strategy for treating postmenopausal osteoporosis is to inhibit bone resorption by osteoclasts and/or increase bone formation by osteoblasts. However, most of the current therapies for treating postmenopausal osteoporosis focus on inhibiting bone resorption, and there are only few agents
Corresponding author. Fax: +852 26324618. E-mail address: Lingqin@cuhk.edu.hk (Q. Ling). 1 Co-rst author. 8756-3282/$ see front matter 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2009.05.022

available that promote bone formation. Therefore, to identify agents with an anabolic effect on osteoporotic bone is desirable because such an agent would create a new alternative for treating osteoporosis. Osteoblasts originate from bone marrow stromal cells (BMSCs), which have the potential to differentiate into several different lineages, including osteoblasts, chondroblasts, adipocytes and myoblasts; and of these lineages, the osteogenic and adipogenic lineages are closely related [2,3]. Clinical observations have shown that aging and postmenopausal osteoporotic women have a concomitant increase in the amount of adipose tissue and decrease of trabecular bone [46]. It has been reported that the osteogenic potential of BMSCs of postmenopausal women is substantially lower than that of healthy control women, and that gene expression during apidogenic differentiation of BMSCs is increased [7,8]. The results of several in

P. Songlin et al. / Bone 45 (2009) 534544

535

vitro studies in cultured rat and human BMSCs suggested a reciprocal relationship between the differentiation of osteogenic cells and adipocytes [9,10]. Since BMSCs are the progenitor cells of both osteoblasts and adipocytes, the decreased number of osteoblasts in postmenopausal osteoporosis may be due to increased differentiation of the BMSCs towards the adipogenic lineage instead of the osteogenic lineage. Therefore, modulating the differentiation of BMSCs from an adipogenic lineage to an osteogenic lineage may be a potential strategy to treat postmenopausal osteoporosis. Of the currently available anti-osteoporosis agents, only few have been reported to prevent bone loss by modulating the differentiation of the osteogenic and adipogenic lineages [1113]. Epimedium-derived avonoids (EFs) are a group of phytoestrogenic components [14,15]. Recently, we reported a randomized, double-blind and placebo-controlled trial in which 85 late postmenopausal women were given EFs for 24 months. BMD levels in the femoral neck and lumbar spine at the end point were signicant greater in EFs-treated compared with placebo-treated subjects. A signicant difference in urine deoxypyridinoline between the two groups was found at both 12 and 24 months [16]. In another rat study, EF treatment started 4 days after OVX and lasted for 12 weeks where we found that EFs prevented OVX-induced reduction in failure force and bone microarchitecture deterioration [17]. The results of these two studies implied that EFs could be a potential agent for treating postmenopausal osteoporosis. However, two major issues remain unresolved. Firstly, histomorphometric information on the effect of EFs on trabecular bone at the tissue level is still lacking as previous studies mainly assessed bone turnover only using biomarkers to support bone densitometry data. Bone histomorphometry is a valuable method to assess bone remodeling at the cellular and tissue level, and this method has been used commonly to evaluate the underlying mechanisms associated with long-term effects of therapeutic agents on bone remodeling and quality [1820]. Therefore, the anabolic effect of EFs on trabecular bone in osteoporotic rats should be investigated histomorphometrically to verify the observed efcacy. Secondly, the exact mechanism of the anabolic effect of EFs on bone is not systemically evaluated. In our previous study, we demonstrated that the preventive effect of EFs on bone loss in OVX rats might be a direct effect on bone cells, independent of enhanced intestinal calcium absorption, yet without analyzing detailed cellular and molecular mechanism about the effect of EFs on osteoblast [17]. Recent in vitro studies using similar phytoestrogenic avonoids demonstrated the potential to promote osteoblastogenesis at the progenitor cell level [2123]. In view of the reported reciprocal relationship between osteogenesis and adipogenesis in BMSCs cultures, we hypothesized that EFs exerted their benecial effect on bone in OVX-induced osteoporosis by modulating adipogenic and osteogenic differentiation of BMSCs. Accordingly, the present experimental study was designed to (a) validate the anabolic effect of EFs on osteoporotic bone and its associated cellular and molecular mechanism(s) by using bone densitometry, both 2-D and 3-D bone histomorphometries, as well as molecular biological approaches.

Materials and methods Animal grouping and treatments Fifty 11-month-old female Wistar rats (SIPPR-BK Experimental Animal Ltd., China) were used and housed in individual cages in an animal room that was maintained at 22 C with a 12 h light and 12 h dark cycle. The rats were fed a standard rat chow (SIPPR-BK Experimental Animal Ltd., China) that contained 0.7% phosphorus and 0.9% calcium, and had free access to tap water. The rats were randomly divided equally into ve groups: (1) sham-operated group (Sham), (2) ovariectomized group (OVX), (3) sham-operated group treated with soluble vehicle (Sham + SV), (4) ovariectomized group treated with soluble vehicle (OVX + SV), and (5) ovariectomized group treated with EFs (10 mg/kg body weight/day) (OVX + EFs). EFs were obtained by a series of standard extraction and isolation procedures [14,15] and the biological active components were quantied by high performance liquid chromatography (HPLC) prole (Table 1, Fig. 1). Animals were euthanized an overdose of sodium pentobarbital (35 mg/kg) (Sigma-Aldrich, Saint Louis, MO, USA). Both femora and cardiac blood samples were collected to conrm bone loss in the OVX group. Treatment groups with either SV or EFs were then commenced for 4 months, where the body weight of rats was measured weekly in order to adjust the dose of the SV and EFs. Calcein (5 mg/kg) (Sigma-Aldrich) was injected intraperitoneally 14 and 4 days before sacricing the rats for studying the bone remodeling. At the end of the 4-month treatment period, the animals were euthanized for collecting tibiae, femora, and cardiac blood samples. The blood samples were centrifuged for serum isolation and then stored at 80 C before biochemical analysis. Part of the conditioned serum from SV-treated and EFstreated OVX rats was used in an in vitro study to assess the effect of conditioned serum on osteogenic and adipogenic differentiation of human BMSCs as described below. The left proximal tibiae were prepared for measurement of the trabecular bone architecture by microCT and analyzed histomorphometrically. Bone marrow was harvested from the right tibia and femora for primary culture of the BMSCs. All procedures were reviewed and approved by the Animal Research Committee of the Chinese University of Hong Kong. Bone turnover assessment by biochemical analysis Bone turnover was assessed by determining serum osteocalcin and tartrate-resistant acid phosphatase 5b (TRACP 5b) levels. Serum osteocalcin levels were measured by an enzyme-linked immunosorbent assay (Rat Mid Osteocalcin ELISA kit, IDS Inc., Fountain Hills, AZ, USA) according to the manufacturer's instructions. Serum TRACP 5b levels were measured by a solid phase immunoxed enzyme activity assay (Rat TRAP Assay, IDS Inc., Fountain Hills, AZ, USA) according to the manufacturer's instructions.

Table 1 Seven main avonoid molecules identied in epimedium-derived avonoids. Peak no. 1 2 3 4 5 6 7 Retention time (min) 7.9 16.0 18.4 20.3 22.7 24.5 55.7 Name Epimedoside A Hexandraside F Epimendin A Epimendin B Epimendin C Icariin Baohuoside-I Molecule form C32H38O15 C39H50O20 C39H50O20 C38H48O19 C39H50O19 C33H40O15 C27H30O10 Common structure R1 -L-Rha 4-O-(-D-Glc)--L-Rha 2-O-(-D-Glc)--L-Rha 2-O-(-D-Xyl)--L-Rha 2-O-(-L-Rha)--L-Rha -L-Rha -L-Rha R2 Glc Glc Glc Glc Glc Glc H R3 H CH3 CH3 CH3 CH3 CH3 CH3 1.8 1.2 1.4 1.5 6.8 82.9 2.1 Area (%)

536

P. Songlin et al. / Bone 45 (2009) 534544

Fig. 1. HPLC prole of epimedium-derived phytoestrogenic avonoid compounds of 1 to 7 that is specied in Table 1.

Bone mineral density (BMD) measurement by peripheral quantitative computed tomography (pQCT) BMD of the proximal femora was measured by a multilayer highresolution pQCT (Densiscan 2000, Scanco Medical, Brttisellen, Switzerland). The region of scanning was the base of femoral neck perpendicular to the axis of the femoral neck, and each slice was 1-cm thick. Measurements of cortical volumetric BMD (iBMD, g/cm3) and trabecular volumetric BMD (tBMD, g/cm3) in the scanned region were analyzed using a previously published protocol [24]. Trabecular bone microarchitecture measurement by microCT The trabecular bone microarchitecture of the left proximal tibia in each rat was measured with a microCT scanner (CT-40, Scanco Medical, Brttisellen, Switzerland) with an isotropic resolution of 15 m in all three spatial dimensions. The region of scanning was 1 mm below the growth plate, and extended distally for 1.5 mm. This region was selected for scanning in order to avoid potential bias from the primary spongiosa [25]. A threshold value of 224 was used to separate the trabecular bone from cortical bone in the analysis system in order to construct a three-dimensional (3-D) image of the original trabecular bone. The scanning procedure and reconstruction of 3-D images of the trabecular bone were performed according to previously described protocols [26,27]. The bone volume (BV, mm3) and the tissue volume (TV, mm3) were measured directly from the original 3-D images, and the trabecular bone volume (BV/TV, %) was normalized in order to compare samples of different sizes. The study parameters of trabecular structure were trabecular thickness (Tb. Th, m), trabecular number (Tb. N, 1/mm), trabecular separation (Tb. Sp, m), and connectivity density (Conn. D, 1/mm3). The structure model index (SMI), a parameter to quantify the characteristic form in terms of the amount of plates and rods, was calculated from the 3D images. Bone histomorphometry After microCT scanning, the proximal tibiae were xed in 70% ethanol, dehydrated in increasing concentrations of ethanol, and then embedded in methyl methacrylate. After polymerization, 100m thick sections were prepared from the blocks using a saw microtome (Leitz, Wetzlar, Germany). Undecalcied 4-m or 8-m thick bone sections were prepared using a Polycut E microtome

(Reichert-Jung, Nusslock, Germany). The 4-m sections were stained with Goldner's Trichrome, and used to determine the static histomorphometric parameters. The 8-m unstained sections were examined under a uorescence microscope in order to determine the dynamic histomorphometric parameters. The bone histomorphometric analysis was performed using OsteoMeasure image analysis software (OsteoMetrics, Atlanta, GA, USA). Bone histomorphometric parameters were determined according to the report of the American Society of Bone and Mineral Research Nomenclature Committee [28]. The measurement area consisted of an area of secondary spongiosa that was at least 1 mm distal from growth plate and 0.5 mm from the endocortical surface. The static parameters were bone perimeter (B. Pm), osteoclastic perimeter (Os. Pm), osteoid perimeter (O. Pm) and erosive perimeter (E. Pm). The dynamic parameters were single-labeled perimeter (sL. Pm), double-labeled perimeter (dL, Pm), and interlabel width (Ir. L. W). These parameters were used to calculate the following parameters of bone formation: osteoid surface (OS/BS), mineralizing surface (MS/BS), mineral apposition rate (MAR), and bone formation rate (BFR/BS). The eroded surface (ES/BS) and osteoclast number (N. Oc/BS) were used as parameters of bone resorption. BMSC culture and differentiation assay BMSCs were obtained from the bone marrow of the right tibia and femur of SV- and EFs-treated OVX rats. The marrow cavity was ushed with -MEM (Invitrogen, San Diego, CA, USA), and the marrow was centrifuged at 1000 rpm for 10 min. The cell pellet was suspended in a basal culture medium that contained 10% fetal bovine serum, 100 U/ml penicillin, and 100 mg/ml streptomycin, and the cells were then seeded into each well (13 105 cells/well) of a 6-well plate (Becton, Dickson and Company, San Jose, CA, USA). Bone marrow cells were incubated at 37 C in a 5% CO2 humidied atmosphere, and the medium was replaced every 3 days. The nonadherent cells were removed when the medium was replaced. The effect of EFs on the stromal cells in the bone marrow was evaluated by a broblast colony-forming unit (CFU-F) assay. The effect of EFs on the osteogenic and adipogenic differentiation potential of BMSCs was assessed by an alkaline phosphatase positive colony-forming unit assay (CFU-ALP) and an adipocyte colony-forming unit assay (CFU-Adipo), respectively. For the CFU-F assay, BMSCs were grown in the basal culture medium for 14 days. The cells were then xed with 10% formalin, stained with Giemsa, and then examined under a light microscope.

P. Songlin et al. / Bone 45 (2009) 534544

537

Only cell aggregates that contained more than 50 cells were counted. For the CFU-ALP assay, BMSCs were grown in the basal culture medium for 3 days. The cells were then grown in an osteogenesisinducing culture medium that comprised basal culture medium supplemented with 10 8 M dexamethasone, 0.2 mM ascorbic acidphosphate, and 10 mM -glycerophosphate (Sigma-Aldrich). After 21 days, the cells were xed with 10% formalin, stained for ALP activity using a Leukocyte Alkaline Phosphatase Kit (Sigma-Aldrich), and the number of ALP-positive colonies was counted under a light microscope. For the CFU-Adipo assay, BMSCs were grown in an adipogenesis-inducing culture medium that comprised basal culture medium supplemented with 10 6 M dexamethasone, 0.50 mM isobutyl-methylxanthine, and 50 M indomethacin for 21 days. Oil Red O was then added to each well in order to stain the accumulated lipid vacuoles in the cells which were used as a marker for counting the number of cells with intracellular lipid accumulation under a light microscope. Assessment of osteogenic and adipogenic gene expression by real-time PCR The expression levels of several osteogenic and adipogenic genes were then measured in order to determine their osteogenic and adipogenic potential by real-time PCR. For this purpose, total RNA was extracted from the BMSCs of the SV- and EFs-treated OVX rats. Briey, bone marrow cells were cultured in the basal culture medium in a 25-cm culture ask at a density of 1 106 cells/cm2. RNA was extracted from the BMSCs using Trizol (Invitrogen), and was then dissolved in RNase-free distilled water. 2 g of total RNA were reverse transcribed into cDNA using Superscript( III reverse transcriptase (Promega, Madison, WI, USA) according to the manufacturer's instructions. Real-time PCR was performed in 96-well plates in a reaction volume of 20 l per well that contained 2 SYBR green master mix (Applied Biosystem, Foster City, CA, USA), diluted gene primers and cDNA. The following primers were used: Runx2 forward 5-GCG TCA ACA CCA TCA TTC TG-3, reverse 5-CAG ACC AGC AGC ACT CCA TC-3; Osterix forward 5-AGC TCT TCT GAC TGC CTG CCT A-3, reverse, 5-TGG GTG CGC TGA TGT TTG CT-3; BSP forward 5-AAA AAT GCT CAC CAT CAC TGC-3, reverse 5-AAT TGC CAC ACT GAC TTC CAC-3; GAPDH forward 5-GAG TCA ACG GAT TTG GAC GT-3, reverse 5-GAC AAG CTT CCC GTT CTC AG-3; PPAR2 forward 5-CGC TGA TGC ACT GCC TAT GA-3, reverse 5-GGG CCA GAA TGG CAT CTC T-3; P2 forward 5-GGC TTC GCC ACC AGG AA3, reverse 5-CCC TTC TAC GCT GAT GAT CAA GT-3; LPL forward 5GGG TCG CCT GGT CGA AGT-3, reverse 5-AAA GTG CCT CCA TTG GGA TAA A-3; RANKL forward 5-GAG CGA AGA CAC AGA AGC-3, reverse 5-ACT ACG AAC CTT CCA TCA TAG CTG-3; OPG forward 5GTC CCT TGC CCT GAC TAC TCT-3, reverse 5-GAC ATC TTT TGC AAA CCG TGT-3. Real-time PCR was performed in an ABI Prism 7700 sequence Detection System (Applied Biosystem) using the following thermal cycling parameters: 1 cycle of initial incubation at 50 C for 2 min, 1 cycle at 95 C for 10 min to activate DNA polymerase, 40 cycles of amplication (95 C for 45 s, 60 C for 30 s and 72 C for 30 s), and followed by 1 cycle of 72 C for 10 min. Quantitative analysis was performed according to the ABI protocol. The threshold cycle (Ct) value was calculated from amplication plots. Relative quantication of gene expression was determined using the delta delta CT (CT) method, with each sample being normalized to the expression level of GAPDH. Each sample was run in duplicate and the expression level of each gene was expressed relative to the expression level of gene in the BMSCs of the SV-treated OVX rats. Human BMSC differentiation assay To determine whether the effect of EFs on primary rat BMSCs could be reproduced in human BMSCs, the effect of conditioned

serum of OVX + SV and OVX + EFs rats was evaluated on the differentiation of human BMSCs. Human BMSCs (Lonza, Gaithersburg, MD, USA) were grown in basal culture medium in a 12-well plate (1.5 104 cells/cm2) until conuence. The BMSCs were then cultured for 14 days in an osteogenesis- or adipogenesis-inducing culture medium in which 50% of the culture medium was replaced every 2 days by conditioned serum of the SV- and EFs-treated OVX rats. Gene expression, ALP staining and Oil Red O staining RNA was extracted from human BMSCs and gene expression was analyzed with real-time PCR assay. The following primers were used: Runx2 forward 5-ACT GGG CCC TTT TTC AGA-3, reverse 5-GCG GAA GCA TTC TGG AA-3; Osterix forward 5-CGG GAC TCA ACA ACT CT-3, reverse 5-CCA TAG GGG TGT GTC AT-3; BSP forward 5-TGC CTT GAG CCT GCT TCC-3, reverse 5-GCA AAA TTA AAG CAG TCT TCA TTT TG-3; PPAR-2 forward 5-TCA GAA ATG CCT TGC AGT GG-3, reverse 5-TTC TCG GCC TGT GGC ATC-3; aP2 forward 5-CTG GCA TGG CCA AAC CCA3, reverse 5-GTA CTT GTA CCA GAG CAC C-3; LPL forward 5-CAG CAA AAC CTT CAT GGT GAT-3, reverse 5-CAA GTT TTG GCA CCC AAC TC-3; GAPDH forward 5-GGA GTC AAC GGA TTT GGT-3, reverse 5GTG ATG GGA TTT CCA TTG AT-3. The osteogenic potential of human BMSCs was measured using an ALP Kit (Sigma-Aldrich) according to the manufacturer's instructions. Briey, human BMSC cultures were maintained in conditioned serum for 14 days, and then incubated in a mixture of naphthol AS-MX phosphate alkaline solution with fast red salt for 45 min at room temperature. The staining intensity of the red granular in the BMSCs was quantied subjectively under a light microscope. For the quantitative determination of ALP activity, the human BMSCs were incubated for 30 min at room temperature with p-nitrophenol phosphate as a substrate, washed with phosphate buffered saline, and then lysed with 0.1% Triton 100 in 10 mM Tris HCl, pH 9.0. Absorbance was measured spectrophotometrically at 405 nM, and ALP activity was determined from the p-nitrophenol standard titration curve. ALP activity was normalized to the total amount of protein in the human BMSCs that was measured using the Bio-Rad protein assay (Bio-Rad, Hercules, CA, USA). The adipogenic potential of human BMSCs is characterized by intracellular lipid accumulation which can be visualized by staining the cells with Oil Red O. For this purpose, 14day-old cultured human BMSCs were xed with neutral buffered formalin (10%) for 10 min, washed with 3% isopropanol, and then incubated with Oil Red O solution (Sigma-Aldrich) for 1 h at room temperature. In order to quantify adipogenic potentials, images of Oil Red O-positive cells in each well were captured by a digital camera attached to a microscope, and then transferred to a computer monitor for counting. Six microscopic elds were selected for counting the Oil Red O-positive cells, and the cell numbers were normalized to per square millimeter. Statistical analysis All data were expressed as mean standard error of the mean (SEM). The data were analyzed by one-way analysis of variance (ANOVA) with a Dunnett's post hoc test to determine group differences between the study parameters using a computerized statistical software program (SPSS version 13.0, SPSS, Chicago, IL, USA). P b 0.05 was considered to be statistically signicant. Results Bone turnover biochemical markers 3 months after surgery, serum osteocalcin and TRAP 5b levels in the untreated OVX rats were signicantly higher than those of the untreated Sham rats (P b 0.05 for both). Signicantly elevated serum

538

P. Songlin et al. / Bone 45 (2009) 534544

microarchitecture, including BV/TV, Tb. Th, Tb. N and Conn. D, in the SV-treated OVX rats were signicantly lower than those of the SVtreated sham-operated rats. In contrast, Tb. Sp in the SV-treated OVX rats was signicantly higher than that of SV-treated shamoperated rats. SMI tended to increase in SV-treated OVX rats, and marginally signicant when compared to that of the SV-treated sham-operated rats (P = 0.054). Compared to the SV-treated OVX rats, EF treatment signicantly increased BV/TV and Conn. D by 58.5% and 33.9% (P b 0.05 for both). The EFs-treated OVX rats had greater BV/TV ratios than the SV-treated OVX rats (26.8%, P b 0.05). This increased ratio was primarily due to the increased Tb. N (26.8%, P b 0.05) as no signicant difference in Tb. Th values of the SV- and EFs-treated OVX rats was detected (P = 0.42). Though not statistically signicant, EF treatment decreased Tb. Sp and SMI by 10.8%, 16.7%, respectively in the OVX rats when compared to SV-treated OVX rats (Table 2). Collectively, these data suggested that EF treatment was able to partially reverse the trabecular bone loss and deterioration of the bone microarchitecture that was induced by estrogen deciency. Dynamic bone histomorphometry The mineralizing surface (MS/BS) index in the SV-treated OVX rats was signicantly greater than that of the SV-treated shamoperated rats. Although the rate of bone formation rate (BFR/BS) in the SV-treated OVX rats was double of that of the SV-treated shamoperated rats, yet just marginally signicant (P = 0.07). Compared to those calculated in SV-treated OVX rats, EF treatment to the OVX rats signicantly increased the values OS/BS, MAR and BFR/BS by 48.7%, 28.5% and 39.7%, respectively (P b 0.05 for all). The number of

Fig. 2. Bone turnover assessment in SV- and EFs-treated OVX rats using serum biomarkers. Compared to the Sham group, the serum levels of the bone formation marker osteocalcin (A) and the bone resorption marker TRACP 5b (B) were signicantly increased in OVX rats 3 months after ovariectomy and after 4-month treatment with either SV or EFs. Data are presented as mean SEM (n = 10). P b 0.05; P b 0.01.

osteocalcin and TRAP 5b levels were found in the SV-treated OVX and sham-operated rats at the end of the 4-month treatment period (P b 0.01 for both). The serum osteocalcin levels in the EFs-treated OVX rats were 8.9% higher than those in the SV-treated OVX rats (P b 0.05) (Fig. 2A), and the serum TRAP 5b levels in the EFs-treated OVX rats were 21.9% lower than those in the SV-treated OVX rats (P b 0.05) (Fig. 2B). BMD evaluations 3 months after surgery, both iBMD and tBMD of the proximal femur in the untreated OVX rats were signicantly smaller than those in the untreated Sham rats (P b 0.05 for both). At the end of the 4-month treatment period, the iBMD and tBMD values in the SV-treated Sham rats were also signicantly lower than that of SV-treated OVX rats (P b 0.05 for both). EF treatment signicantly increased the iBMD value by 48.7% (P b 0.05) and the tBMD value by 109.5% (P b 0.05) in the OVX rats compared to those of SVtreated OVX rats (Figs. 3A and B). Static bone histomorphometric analysis by microCT Examination of representative 3-D images showed deterioration of trabecular bone morphology in the ovariectomized rats: the morphology changed from plate-like to rod-like. This deterioration was partially rescued by EF treatment (Fig. 4). Consistent with the results on BMD measured by pQCT, indices of trabecular bone
Fig. 3. Measurement of bone mineral density (BMD) in the proximal femur by pQCT. Compared to the Sham group, the values for iBMD (A) and tBMD (B) were signicantly lower in OVX rats 3 months after ovariectomy and after 4-month treatment with EFs. Compared to the SV-treated OVX rats (OVX + SV), the values for iBMD and tBMD signicantly increased in the EFs-treated OVX rats (OVX + EFs) after 4-month EF treatment. Data are presented as mean SEM (n = 10). P b 0.05.

P. Songlin et al. / Bone 45 (2009) 534544

539

Fig. 4. Representative 3-D images of proximal tibia trabecular bone by microCT. Three-dimensional images of the trabecular bone in 4-month SV-treated sham-operated (sham + SV) rats, SV-treated OVX rats (OVX + SV), and EFs-treated OVX rats (OVX + EFs). Bone loss and trabecular bone deterioration are evident in the OVX + SV group and the progression of bone loss and bone deterioration is partially rescued by EF treatment.

osteoclasts (N. Oc/BS) was signicantly greater in the SV-treated OVX rats than in the SV-treated sham-operated rats (P b 0.05). The number of osteoclasts in the EFs-treated OVX rats was signicantly lower than that found in the SV-treated OVX rats (P b 0.05). EF treatment did not reduce the eroded surface (ES/BS), an index of bone resorption (P = 0.17) (Table 3). The results of this histomorphometric analysis indicated that EF treatment in OVX rats resulted in an increase in bone formation marker and a concomitant decrease in bone resorption. CFU-F, CFU-ALP and CFU-Adipo assays The CFU-F assay was used to evaluate the effect of EFs on the stromal cells in the bone marrow. The number of CFU-F progenitors in the BMSCs from EFs-treated and SV-treated OVX rats was not signicantly different from each other (P N 0.05) (Fig. 5A). The CFUALP and CFU-Adipo assays were done to determine whether EF treatment had any effect on the osteogenic and adipogenic differentiation of BMSCs. There were 76.8% more osteogenic progenitors in the bone marrow of the EFs-treated versus SV-treated OVX rats (P b 0.05) (Fig. 5B). In contrast, there were 59.4% fewer adipogenic progenitors in the bone marrow of EFs-treated versus SV-treated OVX rats (P b 0.05) (Fig. 5C). Expression of osteogenic and adipogenic genes in rat BMSCs To analyze the effect of EFs on the osteogenic and adipogenic potential of BMSCs from SV- and EFs-treated OVX rats at the molecular level, the expression levels of several osteogenic and adipogenic genes were determined by real-time PCR. Two osteogenic related genes, Runt transcription factor (Runx2) and bone sailoprotein (BSP) were signicantly greater in EFs-treated versus SV-treated OVX rats (p b 0.05 for all). There was no signicant difference for the Osterix gene expression between the two groups (p N 0.05) (Fig. 6A). The expression of three adipogenic differentiation genes, peroxisome proliferator-activate receptor 2 (PPAR-2), adipocyte lipid-binding protein (P2) and lipoprotein lipase (LPL) were investigated in this

study. Consistent with the results of the CFU-Adipo assay, adipogenic genes were down-regulated in BMSCs after EF treatment. EF treatment decreased PPAR-2 and P2 mRNA transcripts by 53.7% and 33% (P b 0.05 for both), respectively (Fig. 6B). EF treatment signicantly decreased RANKL gene expression by 52.7% (P b 0.05), whereas the treatment increased OPG gene expression by 69.3% (P b 0.05) in BMSCs from the OVX rats. As a result, the RANKL/OPG ratio was signicantly reduced in the BMSCs of EFs-treated OVX rats (61.7%, P b 0.01) (Fig. 6C). The reduction in the RANKL/OPG ratio provides additional support to the notion that EFs have anti-resorptive effects in OVX rats, an effect that is reected also by our ndings of decreased serum TRACP 5b levels and N.Oc/BS values in EFs-treated OVX rats. Effect of conditioned serum on osteogenic and adipogenic differentiation of human BMSCs Conditioned serum from EFs-treated group resulted in signicant increase of mRNA level of three osteogenic genes, Runx2, Osterix and BSP by 88%, 66% and 51%, respectively compared with SV-treated group (p b 0.05 for all). The mRNA level of adipogenic genes PPAR-2 and LPL was signicantly reduced by 46% and 54%, respectively (p b 0.05 for both) (Figs. 7A and B). It was greater in human BMSCs exposed for 14 days to conditioned serum of EFs-treated versus SV-treated OVX rats (Figs. 7C and D). The ALP activity in differentiating human BMSCs was 85.9% higher incubated with conditioned serum of EFs-treated versus SV-treated OVX rats (P b 0.01) (Fig. 7G). This result is consistent with the results of the CFU-ALP assay where we found an increase in the number of osteogenic progenitors in the bone marrow of the EFstreated OVX rats. Intracellular lipid accumulation is a measure of adipogenic differentiation of BMSCs after adipogenic induction, and was quantied by staining the cells with Oil Red O (Figs. 7E and F). 33.3% less Oil Red O-positive human BMSCs was found following their exposure to conditioned serum of EFs-treated OVX rats as compared with that following exposure to conditioned serum of SV-treated OVX rats (P b 0.05) (Fig. 7H). This result is

Table 2 Effect of EFs on trabecular bone microarchitecture. BV/TV (%) Sham + SV OVX + SV OVX + EFs 28.05 3.72 13.01 2.29 20.66 1.49 Tb.Th (mm) 0.12 0.02 0.07 0.01 0.08 0.01 Tb.Sp (1/mm) 0.52 0.01 0.74 0.04 0.66 0.04 Tb.N (mm) 2.83 0.39 1.94 0.16 2.46 0.18 SMI 1.19 0.33 1.86 0.19 1.55 0.11 Conn.D (1/mm3) 30.11 4.63 19.14 3.85 25.62 2.03

Data are presented as mean SEM (n = 10). Signicantly different from the Sham + SV group; P b 0.05. Signicantly different from the OVX + SV group; P b 0.05.

540 Table 3 Effect of EFs on trabecular bone histomorphometric parameters. OS/BS (%) Sham + SV OVX + SV OVX + EFs 3.85 0.92 4.58 0.71 6.81 1.50 MS/BS (%) 9.46 1.11 15.75 1.34 17.21 1.01

P. Songlin et al. / Bone 45 (2009) 534544

MAR (m/d) 1.25 0.09 1.38 0.14 1.76 0.20

BFR/BS (m3/m2/year) 43.17 4.93 79.38 12.48 110.89 17.55

N. Oc/BS (1/mm) 1.72 0.28 9.47 1.48 7.38 0.69

ES/BS (%) 8.14 0.93 4.31 0.41 3.42 0.47

Data are presented as mean SEM (n = 10). Signicantly different from the Sham + SV group; P b 0.05. Signicantly different from the OVX + SV group; P b 0.05.

also consistent with the effect of EFs on intracellular lipid accumulation in rat BMSCs using the identical CFU-Adipo assay. Discussion The existence of a reciprocal relationship between osteogenic and adipogenic differentiation of BMSCs raises the possibility that modulation of differentiation of BMSCs towards the osteogenic lineage would promote bone formation, and could be used as a therapeutic strategy in bone degenerative disorders such as osteoporosis. Based on the results of a previous study in postmenopausal women with osteoporosis [16], we speculated that EFs exerted its anabolic effect on osteoporotic bone by modulating the differentiation of BMSCs towards the osteogenic lineage. In this study, we tested this hypothesis in a rat model of OVX-induced osteoporosis, and demonstrated that EFs exerted anabolic effects on osteoporotic trabecular bone based on systemic static and dynamic bone histomorphometric analysis. This bone-forming effect of EF treatment concomitantly increased the osteogenic potential and decreased the adipogenic potential of BMSCs at the cellular and molecular levels. In addition, this pro-osteogenic and anti-adipogenic effect of EFs on rat BMSCs could be simulated on the human BMSCs cultures with conditioned serum from EFs-treated OVX rats. Among many factors that have been attributed to bone loss in postmenopausal osteoporotic women, it should be noted that the reduced osteoblast number and/or activity cannot compensate adequately for the increased osteoclast-mediated bone resorption due to estrogen deciency [1]. The main cellular targets of the currently available efcacious anti-osteoporosis agents that have been approved by FDA are mature osteoblasts and osteoclasts, which are responsible for bone remodeling. A new potential therapeutic target are BMSCs

which are a primitive cell population and have the potential to produce mesodermal tissues that include bone, cartilage, fat and muscle [2,3,29]. Since BMSCs are pluripotent stem cells in the bone marrow, controlling their lineage differentiation is crucial for producing the desired cells and tissues. Therefore, the regenerative potential of BMSCs is very dependent on the understanding of the cellular and molecular mechanisms that underlie the lineage differentiation of BMSCs and their responses to exogenous stimuli. The existence of the reciprocal relationship between osteogenesis and adipogenesis in BMSCs was based on the observation that the increase in the fat content in bone marrow occurred concomitantly with a decrease in trabecular bone volume in aged and osteoporotic individuals [46]. The results of several in vitro studies reported that the increased adipogenesis of BMSCs of osteoporotic individuals was coupled with decreased osteogenic potential [7,8]. These ndings provide evidence that promoting bone formation and restoring bone volume by directing the lineage of BMSCs towards osteogenesis is a potential strategy to treat osteoporosis. Recently, Mukherjee et al. reported that bortezomib, a rst-class clinical proteasome inhibitor used to treat patients with multiple myeloma modulates the in vivo behavior of BMSCs [11]. Specically, they showed that the inhibitor promoted osteoblastic differentiation and inhibited adipocytic differentiation, and consequently increased bone volume in intact mice [11]. Oxytocin, which can modulate in vitro osteoblastic and adipogenic lineage differentiation of human adipose-derived stem cells, can reverse bone loss in OVX-induced osteoporotic rats by suppressing adipogenesis and promoting osteogenesis [13]. We recently reported that EFs prevented bone loss in late postmenopausal women and in rats with OVX-induced osteopenia [16,17]. The biochemical analysis demonstrated that EF treatment decreased the urinary levels of the bone resorption marker, deoxypyridinoline

Fig. 5. Effect of EF treatment on BMSCs proliferation and differentiation. BMSCs were obtained from bone marrow of the right tibia and femur of the group of OVX + SV and OVX + EFs. (A) Rat BMSCs were cultured in a 6-well plate (13 105 cells/well) in basal culture medium, and the effect of SV and EFs on the stromal cells was quantied after 14 days with a broblast colony-forming unit (CFU-F) assay. The number of CFU-F colonies was not signicantly different after either SV or EF treatment. (B) Rat BMSCs were cultured in basal culture medium which was then replaced with an osteogenic-inducting culture medium. At day 21, the number of alkaline phosphatase (ALP)-positive colonies (CFU-ALP) were counted. The number of CFU-ALP colonies that was formed from the BMSCs of OVX + EFs was signicantly higher than that formed from the BMSCs of OVX + SV. (C) Rat BMSCs were cultured in basal culture medium which was then replaced with an adipogenic-inducing culture medium. At day 21, the number of Oil Red O staining-positive colonies (CFU-Adipo) was counted. The number of CFU-Adipo-positive colonies that was formed from the BMSCs of OVX + EFs was signicantly lower than that formed from the BMSCs of OVX + SV. Data are presented as mean SEM. Data presented are three independent experiments with triplicates within each experiment. P b 0.05, signicantly different from the OVX + SV group.

P. Songlin et al. / Bone 45 (2009) 534544

541

Fig. 6. Effect of EF treatment on the expression levels of osteogenic and adipogenic genes of rat BMSCs. BMSCs were harvested from the bone marrow of the right tibiae and femur in the groups of OVX + SV and OVX + EFs. 7 days after the rst medium exchange, RNA was extracted from the BMSCs, and the expression levels of osteogenic (A), adipogenic (B) and osteoclastogenic (C) genes were determined by real-time PCR. Compared to SV treatment, EF treatment: signicantly increased the expression levels of the Runx2 and BSP mRNA transcripts (A); signicantly lowered the expression levels of the PPAR-2 and aP2 mRNA transcripts (B); signicantly lowered the expression levels of the RANKL mRNA transcripts and increased the expression levels of the OPG mRNA transcripts. As a result, RANKL/OPG ratio was signicantly lower (C). Data are presented as mean SEM and the relative expression level of each gene was normalized to GAPDH. Data are presented from three independent experiments with triplicates in each experiment. P b 0.05; P b 0.01 compared to the OVX + SV group.

(DPD), and increased serum levels of the bone formation marker, osteocalcin [17]. In the present study, we found that 4-month EF treatment to osteoporotic rats that had been ovariectomized 3 months prior to the start of EF treatment signicantly increased serum osteocalcin levels and decreased TRACP 5b levels. These data conrmed the results of our previous study on the effects of EFs in OVX rats in which

EF treatment was started 4 days after ovariectomy [17]. In that study, we reported that EF treatment elevated serum levels of osteocalcin, and lowered the urinary levels of DPD. One of the limitations of our previous studies is that we did not evaluate the anabolic effect of EFs at the tissue level. This prompted us to determine the static and dynamic histomorphometric parameters of trabecular bone after treating OVX-induced osteoporotic rats with EFs. Based on the values of the static and dynamic histomorphometric parameters, we found that bone turnover in these OVX rats was increased. We also found that EF treatment attenuated the deterioration of trabecular bone in the OVX rats based on the values of the static bone histomorphometric parameters. The values of trabecular bone volume, trabecular number and connectivity density in the EF-treated OVX rats were higher than those determined in SV-treated OVX rats. The greater bone volume found in these rats might be mainly due to the trabecular number, and not trabecular thickness, because trabecular thickness did not differ signicantly after 4-month EF treatment. Accordingly, we concluded that the increase of bone volume was the result of restoring damaged trabeculae instead of their thickening existing trabeculae. This is consistent with the ndings of others who reported that the benecial effects on trabecular bone of some anabolic agents or stimuli is mainly by modulating trabecular number [3033]. We found also that 4-month EF treatment in the OVX rats signicantly increased the values of the dynamic histomorphometric indices for bone formation, namely the osteoid surface, the mineral appositional rate, and decreased bone resorption as measured by the low osteoclast number. This result suggests that EFs exert a concomitant effect on bone formation and bone resorption at the tissue level. This mechanism of action is different from that of the current therapies developed for treating postmenopausal osteoporosis. The anabolic effect of intermittently administered parathyroid hormone on osteoporotic bone is to shift the balance between bone formation and bone resorption to one of increased bone remodeling, which results in an increase in the number of individual bone multicellular units, and improved bone microarchitecture and quality [3436]. There are some reports that claim the benecial effect of antiresorptive agents such as bisphosphonates on osteoporotic bone is not a result of accelerated formation of new bone. Instead, the authors of these reports propose that their benecial effect is due to a reduction of bone resorption. The overall result is preservation of the extant bone microarchitecture and prevention of further bone loss [3739]. The results of our bone histomorphometric analysis were found to corroborate the biochemical results for assessing bone turnover. This corroboration between two sets of results suggests that the benecial action of EFs on trabecular bone is achieved by simultaneous actions on bone formation and bone resorption processes. Recently, it has been reported that strontium ralenate, a new anti-osteoporosis drug, also acts on osteoblast-mediated bone formation and osteoclastmediated bone resorption at the tissue level in estrogen-decient rats [40] and in vitro as well [41,42]. To explore the cellular and molecular mechanisms of the anabolic action of EFs on trabecular bone, the present study assessed the frequency of BMSCs that had the capacity to form the broblast-like colonies using the CFU-F assay. Results showed that the number of broblast-like colonies did not increase in BMSC cultures from EFtreated versus SV-treated OVX rats. This suggested that EF treatment had no effect on the stromal cell population in the bone marrow. Mukherjee et al. recently reported that bortezomib was able to increase the BMSC population in mice [11]. Luu et al. have reported that mechanical stimulation also caused such effect in mice [32]. In our study, the reasons for why this population of BMSCs did not change following 4-month EF treatment are not clear. It is known that the stem cell population declines with aging [43,44]. Therefore, the capacity to increase the size of the stem cell pool would be limited in elderly individuals. Accordingly, we offer the explanation that the difference among our results and those of Mukherjee et al. and Luu et

542

P. Songlin et al. / Bone 45 (2009) 534544

Fig. 7. Effect of conditioned serum on the osteogenic and adipogenic lineage differentiation of human BMSCs. Human BMSCs were cultured in basal culture medium in a 12-well plate (1.5 104 cells/cm2) until conuence. The cells were then cultured in an osteogenesic and adipogenesic-inducing medium for 714 days. On day 7, real-time PCR assay was performed to analyze osteoblast and adipogenic related gene expression (A and B). ALP staining (C and D) and Oil Red O staining (E and F) were performed, respectively. (G) ALP activity was quantied using a p-nitrophenol assay, and normalized to the total amount of protein in the BMSCs. Compared to that found using conditioned serum of SV-treated OVX rats (OVX + SV), ALP activity was signicantly increased in BMSCs that were treated with conditioned serum of EFs-treated OVX rats (OVX + EFs). (H) Compared to that found using conditioned serum of SV-treated OVX rats (OVX + SV), the number of Oil Red O-positive adipocytes was signicantly lower in BMSCs that were treated with conditioned serum of SV-treated OVX rats (OVX + EFs). Data are presented from three independent experiments with triplicates in each experiment. P b 0.05 and is the signicance of the difference from the OVX + SV group.

al. may be due to use of 18-month-old estrogen-decient osteoporotic rats in the present study while young mice (7 weeks) were used in the Mukherjee and Luu studies. When BMSCs from the EFs-treated OVX rats were plated under the osteogenic and adipogenic conditions, the present study showed a colony bias towards the osteogenic lineage based on the results of the CFU-ALP and CFU-Adipo assays. This proosteogenic and anti-adipogenic effect of EF treatment on BMSCs was further corroborated after quantifying the expression levels of several osteogenic and adipogenic genes of BMSCs by real-time PCR. Specically, EF treatment increased the mRNA level of Runx2, which was a transcription factor for osteoblast differentiation [45]. In contrast, the expression levels of PPAR-2, a transcription factor that regulates adipogenic differentiation of BMSCs [46], was downregulated by EF treatment. These ndings are consistent from those of other in vitro studies which have demonstrated that genestein and daidzein, both of which are phytoestrogen avonoids, can simultaneously promote the expression of osteogenic gene expression and inhibit the expression of adipogenic genes of BMSCs [21,22]. Accordingly, we propose that the increased bone formation in the estrogen-decient osteoporotic rats following 4-month EF treatment is due to EFs promoting the transformation of undifferentiated BMSCs towards the osteogenic lineage. We have reported previously the benecial effects of EFs on prevention of bone loss in late postmenopausal women [16]. It was

essential to establish whether the pro-osteogenic and anti-adipogenic action of EFs on rat BMSCs was reproducible in human BMSC cultures. Since EFs comprised of several phytoestrogens, it was presumed that any benecial effect would be due to the absorbed compound and/or its metabolite(s) after oral administration. Indeed, compared to the SV-treated OVX rats, the conditioned serum of EFs-treated OVX rats showed greater potential to promote osteogenesis and suppress adipogenesis in differentiating human BMSCs than the conditioned serum of SV-treated OVX rats. This might suggest that preferred modulating effect of EFs on the differentiation of BMSCs into the osteogenic lineage could be due to either the absorbed intact bioactive ingredient(s) or one or more metabolites. Future studies should be therefore aiming at identication of the intestinal metabolites of EFs, their chemical structure(s) and biological properties. There are a few limitations in the current study that may need future studies to address. Firstly, though our evidence was based on histomorphometric and microCT data that implied anabolic effect of EFs on osteoporotic bone, we did not include a positive control group, such as estrogen, that is known to be able to prevent bone loss due to estrogen depletion in spite of limited understanding on its potential anabolic effect [47]. Secondly, the anabolic effect of EFs on intact bone was not conrmed as the Sham rats were only treated with vehicle and without EFs. The efcacy on intact bone shall be less detectable as OVX rats mimic the status of postmenopausal women with high bone

P. Songlin et al. / Bone 45 (2009) 534544

543

turnover and prominent trabecular bone deterioration. Even so, the anabolic effect on intact bone should be investigated in future studies even with substantial larger number of bone samples. Thirdly, our evidence suggested that the anabolic effect of EFs was associated with increased osteogenesis of BMSCs, however, we are not sure if the specic EF-induced signaling pathways are involved in promoting osteogenic differentiation of BMSCs. Prostaglandin E2 (PGE2) was reported to play an important role in mediating osteoblast differentiation and new bone formation in response to some anabolic agents or stimuli [48,49]. It is also important to determine whether PGE2 or other growth factors (bone morphometric proteins, insulin growth factors, etc.) are involved in the anabolic effect of EFs on BMSCs by means of signaling pathway array or gene chips. In conclusion, the results of the present experimental study demonstrated at the tissue, cellular and molecular levels that 4month EF treatment exerted its anabolic effects on trabecular bone in an estrogen-decient induced rat model of osteoporosis by increasing bone formation and decreasing bone resorption at the tissue level. These anabolic effects of EFs were associated with a shift in the lineage differentiation of BMSCs towards osteogenesis instead of adipogenesis. This pro-osteogenesis and anti-adipogenesis effect was also simulated in human BMSC cultures by conditioned serum of EFtreated rats. Acknowledgments This study was supported by a Hong Kong Earmarked Research Grant (Project Ref. No. 478508) and a Direct Grant for Research in the Chinese University of Hong Kong (Reference No. 2007.1.067). References

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28] [1] Raisz LG. Pathogenesis of osteoporosis: concepts, conict, and prospects. J Clin Invest 2005;115:331825. [2] Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997;276:714. [3] Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284: 1437. [4] Meunier P, Aaron J, Edouard C, Vignon G. Osteoporosis and the replacement of cell populations of the marrow by adipose tissue. A quantitative study of 84 iliac bone biopsies. Clin Orthop 1971;80:14754. [5] Burkhardt R, Kettner G, Bohm W, Schmidmeier M, Schlag R, Frisch B, et al. Changes in trabecular bone, hematopoiesis and bone marrow vessels in aplastic anemia, primary osteoporosis, and old age: a comparative histomorphometric study. Bone 1987;8:15764. [6] Verma S, Rajaratnam JH, Denton J, Hoyland JA, Byers RJ. Adipocytic proportion of bone marrow is inversely related to bone formation in osteoporosis. J Clin Pathol 2002;55:6938. [7] Rodrguez JP, Garat S, Gajardo H, Pino AM, Seitz G. Abnormal osteogenesis in osteoporotic patients is reected by altered mesenchymal stem cells dynamics. J Cell Biochem 1999;75:41423. [8] Sekiya I, Larson BL, Vuoristo JT, Cui JG, Prockop DJ. Adipogenic differentiation of human adult stem cells from bone marrow stroma (MSCs). J Bone Miner Res 2004;19:25664. [9] Beresford JN, Bennett JH, Devlin C, Leboy PS, Owen ME. Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures. J Cell Sci 1992;102:34151. [10] Rodrguez JP, Montecinos L, Ros S, Reyes P, Martnez J. Mesenchymal stem cells from osteoporotic patients produce a type I collagen-decient extracellular matrix favoring adipogenic differentiation. J Cell Biochem 2000;79:55765. [11] Mukherjee S, Raje N, Schoonmaker JA, Liu JC, Hideshima T, Wein MN, et al. Pharmacologic targeting of a stem/progenitor population in vivo is associated with enhanced bone regeneration in mice. J Clin Invest 2008; 118:491504. [12] Yamaza T, Miura Y, Bi Y, Liu Y, Akiyama K, Sonoyama W, et al. Pharmacologic stem cell based intervention as a new approach to osteoporosis treatment in rodents. PLoS ONE 2008;3:e2615. [13] Elabd C, Basillais A, Beaupied H, Breuil V, Wagner N, Scheideler M, et al. Oxytocin controls differentiation of human mesenchymal stem cells and reverses osteoporosis. Stem Cells 2008;26:2399407. [14] Lee SH, Jung BH, Kim SY, Chung BC. Determination of phytoestrogens in traditional medicinal herbs using gas chromatography-mass spectrometry. J Nutr Biochem 2004;15:45260. [15] Qin L, Zhang G, Shi Y, Lee KM, Leung PC. Prevention and treatment of osteoporosis with traditional herbal medicine. In: Deng HW, Liu YZ, Guo CY, Chen D, editors. [29] [30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

Current topics of osteoporosis. Singapore: World Scientic Publisher; 2005. p. 51331. Zhang G, Qin L, Shi Y. Epimedium-derived phytoestrogen avonoids exert benecial effect on preventing bone loss in late postmenopausal women: a 24-month randomized, double-blind and placebo-controlled trial. J Bone Miner Res 2007;22:10729. Zhang G, Qin L, Hung WY, Shi YY, Leung PC, Yeung HY, et al. Flavonoids derived from herbal Epimedium brevicornum Maxim prevent OVX-induced osteoporosis in rats independent of its enhancement in intestinal calcium absorption. Bone 2006;38:81825. Shiraishi A, Higashi S, Masaki T, Saito M, Ito M, Ikeda S, et al. A comparison of alfacalcidol and menatetrenone for the treatment of bone loss in an ovariectomized rat model of osteoporosis. Calcif Tissue Int 2002;71:6979. Recker RR, Weinstein RS, Chesnut 3rd CH, Schimmer RC, Mahoney P, Hughes C, et al. Histomorphometric evaluation of daily and intermittent oral ibandronate in women with postmenopausal osteoporosis: results from the BONE study. Osteoporos Int 2004;15:2317. Ciria-Recasens M, Perez-Edo L, Blanch-Rubio J, Marinoso ML, Benito-Ruiz P, Serrano S, et al. Bone histomorphometry in 22 male patients with normocalciuric idiopathic osteoporosis. Bone 2005;36:92630. Heim M, Frank O, Kampmann G, Sochocky N, Pennimpede T, Fuchs P, et al. The phytoestrogen genistein enhances osteogenesis and represses adipogenic differentiation of human primary bone marrow stromal cells. Endocrinology 2004; 145:84859. Dang Z, Lwik CW. The balance between concurrent activation of ERs and PPARs determines daidzein-induced osteogenesis and adipogenesis. J Bone Miner Res 2004;19:85361. Chen KM, Ma HP, Ge BF, Liu XY, Ma LP, Bai MH, et al. Icariin enhances the osteogenic differentiation of bone marrow stromal cells but has no effects on the differentiation of newborn calvarial osteoblasts of rats. Pharmazie 2007;62:7859. Lai YM, Qin L, Hung VWY, Chan KM. Regional differences in cortical bone mineral density in weight-bearing long bone shafta pQCT study. Bone 2005;36:46571. Kimmel DB. Animal models for in vivo experimentation in osteoporosis research. In: Marcus R, Feldman D, Kelesy J, editors. Osteoporosis. San Diego: Academic Press; 1996. p. 67190. Siu WS, Qin L, Cheung WH, Leung KS. A study of trabecular bones in ovariectomized goats with micro-computed tomography and peripheral quantitative computed tomography. Bone 2004;35:216. Gong H, Zhang M, Yeung HY, Qin L. Regional variations in microstructural properties of vertebral trabeculae with aging. J Bone Miner Metab 2005;23: 17480. Partt AM. Bone histomorphometry: standardization of nomenclature, symbols and units. Summary of proposed system. Bone Miner 1988;4:15. Bianco P, Gehron Robey P. Marrow stromal stem cells. J Clin Invest 2000;105: 16638. Lane NE, Yao W, Kinney JH, Modin G, Balooch M, Wronski TJ. Both hPTH (134) and bFGF increase trabecular bone mass in osteopenic rats but they have different effects on trabecular bone architecture. J Bone Miner Res 2003;18:210515. Sampath TK, Simic P, Moreno S, Bukanov N, Draca N, Kufner V, et al. Sevelamer restores bone volume and improves bone microarchitecture and strength in aged ovariectomized rats. Endocrinology 2008;149:6092102. Luu Y, Capilla E, Rosen C, Gilsanz V, Pessin J, Judex S, et al. Mechanical stimulation of mesenchymal stem cell proliferation and differentiation promotes osteogenesis while preventing dietary induced obesity. J Bone Miner Res 2009;24:5061. Hirano T, Burr DB, Turner CH, Sato M, Cain RL, Hock JM. Anabolic effects of human biosynthetic parathyroid hormone fragment (134), LY333334, on remodeling and mechanical properties of cortical bone in rabbits. J Bone Miner Res 1999;14: 53645. Burr DB, Hirano T, Turner CH, Hotchkiss C, Brommage R, Hock JM. Intermittently administered human parathyroid hormone(134) treatment increases intracortical bone turnover and porosity without reducing bone strength in the humerus of ovariectomized cynomolgus monkeys. J Bone Miner Res 2001;16: 15765. Neer RM, Arnaud CD, Zanchetta JR, Prince R, Gaich GA, Reginster JY, et al. Effect of parathyroid hormone (134) on fractures and bone mineral density in postmenopausal women with osteoporosis. N Engl J Med 2001;344:143441. Reid IR, Brown JP, Burckhardt P, Horowitz Z, Richardson P, Trechsel U, et al. Intravenous zoledronic acid in postmenopausal women with low bone mineral density. N Engl J Med 2002;346:65361. Lewiecki EM, Miller PD, McClung MR, Cohen SB, Bolognese MA, Liu Y, et al, AMG 162 Bone Loss Study Group. Two-year treatment with denosumab (AMG 162) in a randomized phase 2 study of postmenopausal women with low BMD. J Bone Miner Res 2007;22:183241. Bolland MJ, Grey AB, Horne AM, Briggs SE, Thomas MG, Ellis-Pegler RB, et al. Effects of intravenous zoledronate on bone turnover and BMD persist for at least 24 months. J Bone Miner Res 2008;23:13048. Gasser JA, Ingold P, Venturiere A, Shen V, Green JR. Long-term protective effects of zoledronic acid on cancellous and cortical bone in the ovariectomized rat. J Bone Miner Res 2008;23:54451. Marie PJ, Hott M, Modrowski D, De Pollak C, Guillemain J, Deloffre P, et al. An uncoupling agent containing strontium prevents bone loss by depressing bone resorption and maintaining bone formation in estrogen-decient rats. J Bone Miner Res 1993;8:60715. Choudhary S, Halbout P, Alander C, Raisz L, Pilbeam C. Strontium ranelate promotes osteoblastic differentiation and mineralization of murine bone marrow stromal cells: involvement of prostaglandins. J Bone Miner Res 2007;22:100210.

544

P. Songlin et al. / Bone 45 (2009) 534544 [46] Tontonoz P, Hu E, Spiegelman BM. Stimulation of adipogenesis in broblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 1994;79:114756. [47] Women's Health Initiative Steering Committee. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy. JAMA 2004;291:170112. [48] Steiniche T, Hasling C, Charles P, Eriksen EF, Mosekilde L, Melsen F. A randomized study on the effects of estrogen/gestagen or high dose oral calcium on trabecular bone remodeling in postmenopausal osteoporosis. Bone 1989;10:31320. [49] Bakker AD, Klein-Nulend J, Tanck E, Albers GH, Lips P, Burger EH. Additive effects of estrogen and mechanical stress on nitric oxide and prostaglandin E2 production by bone cells from osteoporotic donors. Osteoporos Int 2005;16:9839.

[42] Bonnelye E, Chabadel A, Saltel F, Jurdic P. Dual effect of strontium ranelate: stimulation of osteoblast differentiation and inhibition of osteoclast formation and resorption in vitro. Bone 2008;42:12938. [43] Bergman RJ, Gazit D, Kahn AJ, Gruber H, McDougall S, Hahn TJ. Age-related changes in osteogenic stem cells in mice. J Bone Miner Res 1996;11:56877. [44] Tokalov SV, Grner S, Schindler S, Wolf G, Baumann M, Abolmaali N. Age-related changes in the frequency of mesenchymal stem cells in the bone marrow of rats. Stem Cells Dev 2007;16:43946. [45] Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 1997;89:74754.