Salvianolic Acid B Prevents Bone Loss in Prednisone-Treated Rats through Stimulation of Osteogenesis and Bone Marrow Angiogenesis

Glucocorticoid (GC) induced osteoporosis (GIO) is caused by the long-term use of GC for treatment of autoimmune and inflammatory diseases. The GC related disruption of bone marrow microcirculation and increased adipogenesis contribute to GIO development. However, neither currently available anti-osteoporosis agent is completely addressed to microcirculation and bone marrow adipogenesis. Salvianolic acid B (Sal B) is a polyphenolic compound from a Chinese herbal medicine, Salvia miltiorrhiza Bunge. The aim of this study was to determine the effects of Sal B on osteoblast bone formation, angiogenesis and adipogenesis-associated GIO by performing marrow adipogenesis and microcirculation dilation and bone histomorphometry analyses. (1) In vivo study: Bone loss in GC treated rats was confirmed by significantly decreased BMD, bone strength, cancellous bone mass and architecture, osteoblast distribution, bone formation, marrow microvessel density and diameter along with down-regulation of marrow BMPs expression and increased adipogenesis. Daily treatment with Sal B (40 mg/kg/d) for 12 weeks in GC male rats prevented GC-induced cancellous bone loss and increased adipogenesis while increasing cancellous bone formation rate with improved local microcirculation by capillary dilation. Treatment with Sal B at a higher dose (80 mg/kg/d) not only prevented GC-induced osteopenia, but also increased cancellous bone mass and thickness, associated with increase of marrow BMPs expression, inhibited adipogenesis and further increased microvessel diameters. (2) In vitro study: In concentration from 10−6 mol/L to 10−7 mol/L, Sal B stimulated bone marrow stromal cell (MSC) differentiation to osteoblast and increased osteoblast activities, decreased GC associated adipogenic differentiation by down-regulation of PPARγ mRNA expression, increased Runx2 mRNA expression without osteoblast inducement, and, furthermore, Sal B decreased Dickkopf-1 and increased β-catenin mRNA expression with or without adipocyte inducement in MSC. We conclude that Sal B prevented bone loss in GC-treated rats through stimulation of osteogenesis, bone marrow angiogenesis and inhibition of adipogenesis.


Introduction
Glucocorticoid (GC) therapy is commonly used for inflammatory and autoimmune diseases. The long-term administration of GC can lead to glucorcoticoid-induced osteoporosis (GIO), which significantly increases the patients' morbidity and mortality. Due to limited treatment options, the side effects of GC often have to be tolerated during treatment [1]. Currently, the clinical management of GIO relies on medications similar to those used for treatment of post-menopausal osteoporosis, such as calcium, vitamin D, bisphosphonates, raloxifene, PTH, hormone replacement and calcitonin. These drugs do not address the multi-factor driven GIO. In particular, they do not target the detrimental effect of GC on bone marrow fat metabolism and circulatory system [2][3][4][5]. Thus more studies on these GC induced effects may lead to development of a novel therapeutic approach to prevent and treat GIO.
The pathogenesis of GIO involves multiple factors, of which some suggest the decrease in number and functions of osteoblasts is the main contributing factor [2]. However, recently increased apoptosis of osteoblasts, osteocytes and endothelial cells, suppression of osteoblasts and osteoclasts, and endothelial cell precursor production as well as prolongation of the life span of osteoclasts have all been shown to contribute to the skeletal side effects of GC [4][5][6]. Recent studies suggested that the regulation of marrow stromal cell (MSC) differentiation into bone or fat cells [3] and the inhibition of bone marrow microvasculature play a very important role in GIO development [4,5]. GC can inhibit osteoblast production of bone morphogenetic protein 2 (BMP-2), which causes decreased MSC differentiation into bone cells [7]. GC also directly induce differentiation of marrow stromal cells into adipocytes and inhibit osteogenic differentiation [8]. Kitajima et al. showed that mature fat cells exposed to high dose of GC were larger than control cells derived from bone marrow [9]. The latter would lead to narrowing and obstruction of capillaries in bone marrow microvasculature from increased adipose tissue that results in increased intraosseous pressure and decreased blood flows [5,10]. Excessive GC treatment was also found to inhibit the growth of vascular endothelial cells that further contributes to microcirculation disturbance [11]. Marx et al. have previously demonstrated that the peroxisome proliferator-activated receptor c (PPARc) can induce apoptosis in vascular endothelial cells via caspase-3 activation, thus inhibiting vascular endothelial cell proliferation and angiogenesis [12]. GC can activate PPARc in MSCs through different pathways to promote adipogenesis, which reduces osteoblast differentiation, and eventually leads to fat tissue accumulation in bone marrow [6]. Taken together, these studies suggest that the GIO bone loss is comprised of multiple mechanisms involving the increase in bone marrow adipogenesis and decrease in marrow angiogenesis leading to a decrease in bone marrow microvasculature and consequent decrease in osteogenesis [4,5].
Salvia miltiorrhiza Bunge is a traditional Chinese medicine, called danshen, widely used in clinical practice for the prevention and treatment of cardio-cerebral vascular diseases. Pharmacological testing showed that danshen has anticoagulant, vasodilatory, increased blood flow, anti-inflammatory, free radical scavenging, mitochondrial protective and other activities [13]. Salvianolic acid B (Sal B), the aqueous bioactive component from Salvia miltiorrhiza Bunge, is a polyphenolic compound found in abundance in this plant. The structure of Sal B is shown in Figure 1 which consists of three molecules of Tanshinol, or called danshensu (D (+) b-3,4dihydroxyphenyl lactic acid) and a molecule of caffeic acid, Sal B can be converted in vivo to Tanshinol (Salvianolic acid A, or Sal A), another water-soluble bioactive ingredient of Salvia miltiorrhiza Bunge, with similar function to Sal B. Both Sal B and Tanshinol are wellknown as among the most effective natural product antioxidants [13,14].
Multiple pharmacological studies have found that Sal B can attenuate the effect of myocardial ischemia-reperfusion injury [15]. Interestingly, Sal B can also increase angiogenesis and reduce myocardial ischemia via vascular endothelial growth factor (VEGF) activation [16]. It also relieves brain injury by reducing neuronal damages in cerebral ischemia [17]. Sal B can improve cellular hypoxia-ischemia by expanding micro-arteries, improving microcirculation and increasing the blood flow velocity. Its beneficial effects on blood vessel dilation and protection are believed to be mediated by blocking calcium channels and angiotensin-converting enzyme [18,19]. Moreover, Salvia miltiorrhiza Bunge and its aqueous extract can increase the activity of superoxide dismutase, scavenge  reactive oxygen species (ROS) and therefore reduce the damage of ROS to the vascular endothelium. Thus, Sal B functions as a vasodilator, maintains red blood cell deformability and increases the function of the hematopoietic system [20]. Recently, our in vitro studies have demonstrated that Sal A can inhibit glucocorticoidinduced bone marrow stromal cells adipogenesis, promote osteoblast differentiation, bone matrix formation and bone mineralization [21]. Therefore, we hypothesize that the clinical use of Sal B will hold promise for a more effective and safe treatment for GIO.
The aim of the current study is to validate our hypothesis in a GIO rat model and additional study on in vitro.

Salvianolic acid B (Sal B) Preparation
The original herbal medicine Radix Salviae miltiorrhiza was selected according to the standard protocol of the pharmacopoeia of the People's Republic of China [22]. The aqueous bioactive component from Radix Salviae miltiorrhizae was extracted as previously reported [23]. Salvianolic acid B (Sal B) and Tanshinol found in the aqueous extract of Radix Salviae miltiorrhizae were      [24]. HPLC chromatograms of the control compounds and aqueous extraction samples are shown in Figure 2. The content of Sal B was 25 mg per gram of Radix Salviae miltiorrhiza.

Body weight and serum markers assay
Rats were weighed every week. At the end of the experiments, rats were sacrificed by cardiac puncture under anesthesia. Soft tissues were removed and weighed. Blood and serum samples were collected for measurements of serum calcium (Ca), the serum bone biomarkers alkaline phosphatase (ALP) and tartrate-resistant acid phosphatase-5b (TRACP-5b) according to manufacturers' instructions (Immunodiagnostic Systems Inc, USA).

Immunohistochemical analysis of bone tissues
The distal femoral bone marrow cavity was exposed to prepare decalcified bone slides. Samples were decalcified at room temperature in 15% EDTA for 5 weeks. After decalcification, the sample was placed in 70% alcohol and paraffin embedded. Four micrometer paraffin slides were prepared on glass slides coated with egg white-glycerol, or polylysine, then dried for one hour at 60uC and stored at 4uC for future use.
Immunohistochemical analysis of bone marrow microcirculation factor VIII-related antigen (Von Willbrand Factor, vWF) and peroxisome proliferator-activated receptor c (PPARc) were performed following the manufacturers' instructions. Rabbit antihuman vWF polyclonal antibody was purchased from CHEMI-CON International, Inc, USA, and PPARc mouse monoclonal antibody was purchased from Santa Cruz Biotechnology, Inc, USA. Briefly, the endogenous peroxidase activity of slides were blocked by 3% hydrogen peroxide and nonspecific binding was blocked with 10% goat serum. The slides were then incubated with primary antibody at 4uC overnight in a humidified chamber. The next day the slides were incubated with biotinylated secondary antibody for 15 min, washed again with PBS (2 min63 times), then incubated with streptavidin conjugated to horseradish-peroxidase for 15 min. After a final PBS wash (5 min63 times), a drop of freshly prepared  Table 7. Effects of Sal B, GC, and GC+ Sal B on proximal tibial metaphysis cancellou bone dynamic parameters.
LGR (mm/d) 8 DAB was applied on the slide for color development. The reaction was stopped when a uniform brown color became visible on the slide by rinsing in running water. Counterstaining was done with hematoxylin for 5-10 seconds. A control experiment was performed by replacing primary antibody with PBS. This method was a modification at the manufacturers' instructions.
The evaluation of Von Willbrand Factor expression and marrow microvessel density (MVD) were determined by counting the number of endothelial cells showing positive staining Von Willbrand Factor, and the MVD was determined by the total number of vWF stained cells divided by the measured area in previous studies [25,26]. The diameter of marrow microvessels (DMV) was determined as the average diameters of microvessels in the same area used to measure MVD, using a digitizing image analysis system (Osteometrics, Inc. Decatur, GA, USA) To evaluate PPARc positive expression in distal femur, the slides were scanned by LeicaQ550CW image analyzer and the average of absorbance from 30 positive expression cells on each slide was presented as mean 6 SD (absorbance/30cells). Immunohistochemical analysis of bone marrow BMP-2 and BMP-7 expressions were performed following the manufacturers' instructions: Antigen retrieval was performed by high temperature and pressure treatment of the sections in citrate buffer (0.01 mol/L, pH 6.0). After a wash in PBS, sections were incubated with hydrogen Table 9. Percentage change of all parameters from basal (BAS), vehicle control (CON) and predisone (GC). peroxide blocking agent to quench endogenous tissue peroxidases and then washed in PBS. The slides were incubated with Ultra V Block for 30 min at room temperature. The primary antibodies were incubated according to the manufacturer's protocol (Maixin-Bio, China). Slides with serial sections were incubated with the primary antibodies (Santa Cluz Corporation) in a humidified chamber at 4uC overnight, followed by being incubated with streptavidin-perosidase avidin for 10 min at room temperature. Diaminobenzadine (DAB) staining was done by incubating the sections with coloring reagent. The sections were counterstained with Harris' haematoxylin, dehydrated through increasing concentrations of alcohol and mounted with coverslips. One section on each slide was used as a control to assess nonspecific binding. For this section, dilution buffer without antibody was used.

Bone mineral density (BMD) determination
The right femurs of the rats were wrapped with saline saturated gauze to maintain moisture and stored at 220uC. After thawing at room temperature, the femurs were moisturized by soaking in saline solution, the residual muscle removed and the length of femur measured with a ruler. The femur bone mineral density between the midpoint and the distal end of the femur was scanned with a SD-1000 single-photon bone mineral density instrument (Nuclear Industry Beijing Institute of Geology) to measure the bone mineral content (BMC, g/cm) and bone width (BW, cm). The bone mineral density (BMD) measurements were performed at the midpoint of the femur and 2 cm proximal and the BMD was calculated as BMC/BW.

Bone mechanical properties determination
After measuring BMD, the femur was used to determine the bone mechanical properties through three-point bending using Bose ElectroForce Testing system (ELF3510, Bose, USA). Bone samples were tested with a 1 mm indenter, at speed of 0.01 mm/s with a 15 mm span (L). Force (F) and deflection (D) that automatically recorded. The output parameters include elastic force (the force required to cause bone specimens to deform, N), maximum force (the maximum force the bone can resist, N), fracture force (the force required to cause bone fracture, N) and the maximum deflection (maximum degree of the bone displacement, mm). The stiffness coefficient (load-displacement curve slope, N/mm) was also calculated based on the above parameters.

Bone Histomorphometry
The right proximal tibial growth plate and metaphyses were processed for cartilaginous longitudinal growth rate and cancellous bone histomorphometric analyses. The samples were opened to expose the bone marrow cavity using an Isomet Low Speed Saw (Buehler, Lake Bluff, Illinois, USA) and fixed in 10% phosphate buffered formalin for 24 hours. They were then dehydrated in graded ethanol, defatted in xylene, and embedded undecalcified in methyl methacrylate [27]. Frontal sections were cut at thicknesses of 4-and 10-mm. The 4-mm sections were stained by Goldner's Trichrome for static histomorphometric measurements. The 10mm unstained sections were used for dynamic histomorphometric analyses [28].

Culture of rat osteoblast and marrow stromal cell (MSC)
Osteoblastic cell was isolated from newborn rat calvaria (rOB). After removing the periosteum, the bones were cut into small flaps of 1 mm 3 and subjected to digestion with 0.25% trypsin (Life Technologies Gibco-BRL) for 20 min at 37 C. After centrifugation, supernatants were discarded to remove the fibroblast population. Then the flaps were digestion with 0.2% collagenase type I and 0.1% hyaluronidase (Life Technologies Gibco-BRL) for six 20 min intervals at 37 C. The samples were then washed thoroughly with Dulbecco's modified eagle's medium (DMEM, Gibco). The precipitates and bone flaps were transferred to 25 cm 2 culture flasks and cultured in Dulbecco's modified eagle's medium, 100 U/mL penicillin, 100 mg/mL streptomycin and 10% fetal bovine serum (FBS, Gibco) at 37uC in a humidified   incubator with 5% carbon dioxide (CO 2 ). The DMEM culture medium was changed every 3 days. MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Bio-Rad, USA) assay was use for the observation of cell growth/death. Primary rat bone marrow stromal cells (rMSC) were collected from marrow of femur in 4-week-old Wistar rats (obtained from the Laboratory Animal Center of Guangdong Medical College) [21]. The rMSC then were prepared by gradient centrifugation at 9006g for 30 min on a Percoll-Paque gradient (Amersham Pharmacia, USA) at a specific gravity of 1.073 g/mL. The lowdensity mononuclear cells were washed twice in Hanks' balanced salt soltion and cultured in DMEM supplemented with 10% FBS and 1% antibiotic (10 000 U/mL penicillin G sodium, 10000 mg/ mL streptomycin sulfate) at 37 C in a humid atmosphere containing 5% CO. The medium was changed every 3 to 4 d to remove the non-adherent hematopoietic cells. The adherent cell population was expanded and passaged every 12 to 14 d. All cells used for the experiments have been through three passages. To identify the abilities of rMSC with regard to osteogenesis and adipogenesis, the following studies were performed.
Induction of osteogenic differentiation of rMSC (OB-in): When the rMSC from passage number 2 became 80% confluent in the culture plates, the culture medium was changed to an osteoblast inducing medium containing 50 mg/mL L-ascorbic acid, 10 22 mol/L b-glycerophosphate, and 10 28 mol/L dexametha-sone. When the cells became layered and confluent up to 100%, visible symmetric colonies were observed after osteoblast induction.
Induction of adipogenic differentiation of rMSC (Ad-in): The rMSC of each group were replated at 26104 cells/cm 2 in a 25 cm 2 culture dish and maintained until 80% confluence. Then, the culture medium was switched to adipogenic medium consisting of control medium supplemented with 10 mg/ml insulin (Sigma Identification for OB-in and Ad-in of MSC had according to the methods referred to Cui. et al. [21]. Determination of alkaline phosphatase (ALP) and osteocalcin secretion ALP activity assay: Cells were seeded in 96-well plates and the confluent cells were cultured for the indicated period with or without Sal B treatment. Cells were washed with PBS and 150 mL of substrat buffer (6.7 mmol/L disodium p-nitrophenylphosphate hexahydrate, 25 mmol/L diethanolamine and 1 mmol/L MgCl) was added. After the mixtures were incubated at 37uC for 3 min, we measured the absorbance at 405 nm.
Determination of osteocalcin secretion: Cells were seeded in 6well plates at a seeding density of 2610 4 cells/cm 2 and were further

Reverse transcription polymerase chain reaction (RT-PCR) assay
Test of cultured cells: Total RNA was extracted from cultured cells using TRIZOL reagent (Invitrogen, Carlsbad, CA) and subjected to RT-PCR analysis using the 9600Gen Amp PCR system (PerkinElmer Applied Biosystems, USA) with PCR reagents (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Primers used are listed in Table 2. Amplified products were then loaded on a 2% agarose gel and subjected to electrophoresis. Digital pictures were taken and analyzed with gel image system. Each gene expression was normalized to GAPDH.
Test of rat femur bone tissue: The upper femur halves obtained were snap-frozen in liquid nitrogen and stored at 270uC. Total RNA was prepared using TRIZOL reagent and stored at 220uC in ribonuclease (RNase)-free water until use. To determine mRNA levels, a one-step reverse transcription-polymerase chain reaction (RT-PCR) procedure was performed using the 9600Gen Amp PCR system (PerkinElmer Applied Biosystems, USA) according to the manufacturer's protocol. Primers used are listed in Table 3. Amplified products were then loaded on a 2% agarose gel and subjected to electrophoresis. Digital pictures were taken and analyzed using Bandscan analysis software. Each gene expression was normalized to GAPDH.

Statistic Analysis
Data were presented as mean6 SD. The statistical differences among groups were evaluated using variance (ANOVA) with Fisher's PLSD test. Probabilities (p) less than 0.05 were considered significant.
There were non-significant reductions in PTM cancellous bone mass (BV/TV down by 19%, Tb.N down by 20%, Oc.S/BS down by 14%), and increased Tb.Sp (up by 32%). In addition, PPARc

Effect of Sal B40 in Intact Male Rats
Sal B40 treatment in intact male rats stimulated cancellous and endocortical bone formation. Compared to the vehicle (aging) control, the 40 mg of Sal B/kg/d (Sal B40) treated intact male rat showed no significant changes on body and soft tissues weights, bone mineral density (BMD), femoral cortical bone strength, PTM cancellous bone mass, architecture and endocortical bone formation. However, the Sal B40 treated rats showed significant increased select PTM cancellous bone formation indices (i.e. MS/ BS, BFR/BS and BFR/TV up by 20, 19 and 24%), increase Ob.S/BS by 24% and Ec-BFR/BS by 14%. Additionally, the Sal B40 non-significantly reduced distal femoral marrow PPARc expression by 14%, reduced marrow fatty area (F.Ar/TV) by 37% and increased DMV by 14%. (Tables 4-9; 9) Effect of GC in Intact Male Rats GC induced losses in bone mass, BMD and mechanical strength associated with decreased osteogenesis, angiogenesis and increased adipogenesis. Compared to vehicle (aging) controls, the use of GC lead to significant increased of liver weight by 9.6%, and reduction in body, thymus, adrenal gland and soleus muscle weights by 10, 25, 20 and 9%, respectively, coupled with significantly increased serum calcium by 7%, tartrate-resistant acid phosphatase-5b (TRACP-5b) by 42% and decreased alkaline phosphatase (ALP) by 52%. Proximal femur and whole femur BMDs were decreased by 20% and 11% and cortical bone strength was reduced (i.e. maximum force and deflection by 11% and 22%). GC further significantly reduced PTM cancellous BV/TV by 29%, Tb.Wi by 18%, Ob.S/BS by 74%, PT LGR by 21%, and some PTM cancellous bone formation indices (i.e. MAR and BFR/TV) by 8 and 20%, Significant reductions were seen in the endocortical bone formation indices (i.e. Ec-MS/BS, Ec-MAR and Ec-BFR by 51, 36 and 68%), and microvessels diameter (DMV, down by 59%) and density (MVD) down by 23%, while fatty marrow content was increased by 124% and PPARc expression by 48% (Table 4-9, Figure 6-9). Immunohistochemical analysis of bone marrow BMP-2 and BMP-7 expression significant decreased with decrease of total BMP-2 and BMP-7 mRNA expression ( Figure 10). The Sal B40 treated GC rats tended to exhibit a larger marrow DMV (up by 38%) versus aging control, that resulted in significantly increased DMV by 238%, when compared to GC control. The rats treated with Sal B40 showed reduced PTM cancellous bone Ob.S/ BS by 34%, while the GC alone was reduced by 74%. The results suggests that both osteoblasts and microvessel diameters were stimulated by Sal B40. (Tables 4-9; 9) Effect of Sal B80 in GC treated Male Rats Sal B80 treatment of GC rats for 12 weeks not only prevented GC-induced changes but also showed additionally stimulated osteogenesis that increase cancellous bone mass and increased marrow angiogenesis by enlargement of microvessels diameter.

Effect of Sal B40 in GC treated Male Rats
Compared to vehicle (aging) control, the GC treated rats receiving 80 mg of Sal B/kg/d (Sal B80) showed no significant differences in body and soft tissue weights, serum biomarkers, BMD, bone strength (except for stiffness), and marrow PPARc expression, BMP-2 and BMP-7 expression, fatty marrow and MVD. However, the Sal B80 treated rats significant differed from Sal B40 treated rats in term of PTM cancellous bone mass up by 33%, Tb.Wi by 17%, BFR/TVby 30% and marrow DMV by 41% versus aging control. (Tables 4, 5, 8&9;  Compared to the Sal B40 treatment there was a significantly dose response increase in serum ALP level by 40%, in proximal femur BMD by 8%, in PTM cancellous BV/TV by 25%, in Tb.Wi by 18% and significantly decrease in BFR/BV by 20%. (Tables 4, 5 Figure 13).

Discussion
The current study found that GC treatment decreased bone formation by inhibiting osteoblast activity and marrow production of BMP-2 and BMP-7, increased marrow adipocytes with elevated PPARc expression that promoted bone marrow stromal cell differentiation into adipocytes. Since osteoblasts and adipocytes share common bone stromal progenitors [31], GC treatment stimulate the differentiation of marrow stromal cells to adipocytes thereby, reducing the pool of local progenitor cells to differentiate into osteoblasts, The reduction in osteoblast number in turn contributes to decreased osteogenesis. Furthermore, we specifically labeled marrow vascular endothelial cells with Von Willbrand Factor and observed similar GC disruption of marrow microvessels as previously observed in rabbit femoral head [32]. The disruption of marrow microvasculature would reduce the source of circulating progenitor cells supporting osteogenesis [33][34][35][36][37][38][39]. The reduction in local and circulating progenitors would then lead to reduce osteogenesis. Our findings suggest that bone formation, bone   marrow fat metabolism and microcirculation are closely related to each other and confirm that the latter two factors also contribute significantly to the development of GC-induced bone loss and the decline in bone strength [4,5].
Our data found 80 mg Sal B/kg/d for 12 weeks depressed adipogenesis and stimulated angiogenesis and osteogenesis in GCtreated rats. The lower 40 mg Sal B/kg/d dose in GC treated rats only depressed adipogenesis and stimulated cancellous bone formation rate. The 40 mg dose also appeared to increase angiogenesis by enlarging marrow microvessels diameter but the observations were not statistically significant. The prevention of GC-stimulation of adipogenesis maintains marrow stromal and red marrow cell levels at control levels. Since osteogenesis and adipogenesis share common bone stromal progenitors [31], this means the pool of marrow progenitors available to differentiate into osteoblasts will be maintained and not be preferentially stimulated to form adipocytes by GC treatment. Additionally the maintenance of red bone marrow means that the bone marrow will retain more microvessels, unlike fatty bone marrow, which is not as well vascularized [40,41]. The Sal B treatment further increased marrow angiogenesis that increases blood flow to improve nutrition and source of circulating progenitors [33][34][35][36][37][38][39]. In support of osteogenesis, stimulated angiogenesis reduced intraosseous pressure that increases blood flow and in turn stimulates osteoblastic activity (i.e. mineral apposition rate) [42]. Further red bone marrow sites with better vasculature exhibit a factor of 10 higher bone remodeling and formation rates than fatty marrow bone sites [43][44][45][46]. Also the improved blood flow increases osteocyte lacunarcanaliculi and blood vessel fluid volume that increase bone strength and reduce the risk of fracture [4,47].
To clarify the mechanism by which Sal B promotes osteoblast differentiation and alleviates GC-induced impairment of bone formation, we examined the effects of Sal B on primary rat calvarial osteoblasts and rat bone marrow stromal cell by upregulating Wnt/b-catenin signaling. Our data showed that Sal B stimulates osteoblast cell growth and MSC differentiation into osteoblast maturation (secretion of ALP and osteocalcin) with a dose and time dependent manner that mimicks the action of osteoblast inducer. This study further demonstrated that Sal B activated Runx2 and b-catenin mRNA expression and declined DKK-1 and PPARc mRNA expression which had been induced by GC (i.e. adipocyte inducement). It is known that canonical Wnt/b-catenin is a key pathway for regulating bone formation and remodeling and contributes to osteoblastic differentiation [48], of which DKK-1 is one of the inhibitor in Wnt-signaling. Studies demonstrate GC to be a very strong inducer of DKK-1 protein [49] which leads to a decrease in osteoblast bone formation in GC induced osteoporosis [50]. Moreover, Runx2 was found to integrate Wnt-signaling for mediating osteogenic differentiation of MSCs [51], it was known to be involved in BMP signaling as mentioned earlier [52]. Our in vivo observation further demonstrated that Sal B stimulated femur bone BMP-2 and BMP-7 mRNA expression and protected GC-treated rats from decrease of femur bone BMP-2 and BMP-7 protein expression (IHC revealed, Figure 10). BMP/Wnt/b-catenin share a crossover mechanism for stimulation of progenitor, osteoblast and angiogenetic generation [53], thus our study support the idea that Sal B protected against GC-induced osteoblast impairment associated abnormal angiogenesis and adipogenesis. Recently Lu et al. reported that Sal B protected against oxidative stressinduced apoptosis in rat bone marrow stem cells [54] suggesting that Sal B may inhibit GC-induced bone marrow stem cell apoptosis to enlarge the local pool of osteoblast precursors available for osteogenesis.
Our current end point findings indicated that Sal B stimulated bone formation in both intact and GC-treated male rats (Table 9). Sal B40 alone in intact male rats significantly increased BFR/TV by 24%, Sal B 40 in GC-treated rats significantly increased BFR/TV by 29%, while Sal B80 in GC-treated rats significantly increased BFR/TV by 50% and additionally increased cancellous bone mass (BV/TV) by 33% and trabecular thickness by 17%. Taken together, the combination of improved marrow microenvironment (i.e. increased local and circulatory progenitors, blood flow and nutrition) andSal B stimulation of bone formation rate not only prevented GC -induced osteopenia but also increased cancellous bone mass.
The current study used doses of 40 and 80 mg/kg gavage in rats, which is similar to the dose used in humans. Salvianolic Acid B (parental form) was approved by the State Food and Drug Administration (SFDA) of China in 2007 for clinical use in the prevention and treatment of cardiovascular diseases. The recommended dose of Salvianolic Acid B for humans is 4 mg/ kg/d i.v. Our study examined the drug impact on body weight (Figure 3), organ weight (table 4), gross necropsy and histopathology (negative data not shown) both in intact rats (SalB40 group) and glucocorticoid-treated rats. No significant adverse effects on these parameters were observed. Recently Li et al. [55] reported on the toxicity of Sal A, a derivative of Sal B, in male and female dogs after a 3-month continuous intravenous infusion at doses of 17, 50, and 150 mg/kg/day. No significant cumulative toxicity was observed either during or 90 days following treatment. Importantly, the doses of Sal A used in dogs were very high as compared to clinical practice and to rats; the 17, 50, and 150 mg/ kg/day in dogs are equivalent to 61.2, 180 and 540 mg/kg in rats. Sal B possesses a similar activity to Sal A. Our data indicate that the doses of Sal B used in current study are safe and effective and produce no adverse effects.
In summary, Sal B treatment of GC-treated male rats not only prevented GC-induced osteopenia but also increased cancellous bone mass by the combination of depressed adipogenesis and stimulated angiogenesis and osteogenesis ( Figure S1). Our findings support further investigation of Sal B stimulating of osteogenesis and marrow circulation and inhibition of adipogenesis as a potential therapeutic strategy in the prevention of not only GCinduced osteopenia but other bone diseases. Figure S1 Working scheme: the mechanism of action of Salvianolic acid B on glucocorticoid induced bone loss. Decrease of osteogenesis and angiogenesis, increase of adipogenesis are considered contributions to glucocorticoid induced bone loss. Salvianolic acid B could withstand the impairment induced by glucocorticoid. (BMPs: bone morphogenetic proteins; MSCs: Figure 13. Effects of Sal B on the gene expression of Dkk-1/b-catenin pathway in rat marrow stromal cell (rMSC) differentiation. rMSC were cultured with osteoblast induction medium (OB-in), adipocyte induction medium (Ad-in, i.e. high concentration of GC), with or without Sal B (10-7 mol/L) and Sal A (5610-7 mol/L, a derivant of Sal B, Figure 1) for 7 days. Expression levels of Runx2 (A) , PPARc (B), b-catenin (C and D) and DKK-1 (E) were measured by RT-PCR. Sal B increased Runx2 and b-catenin mRNA expression and decreased DKK-1 mRNA expression which was similar to the action of OB-in. Ad-in marked increased PPARc and DKK-1 mRNA expression. When treated the Ad-in rMSC with Sal B, the PPARc and DKK-1 mRNA expression decreased obviously. Sal A had similar effects to Sal B. doi:10.1371/journal.pone.0034647.g013 marrow stromal cells; PPARc: peroxisome proliferator-activated receptor c; VEGF: vascular endothelial growth factor; Black: effect of glococorticoid; Red: effect of Salvianolic acid B; q: increase; Q: decrease). (DOC)