Phospholipase C signaling activated by parathyroid hormone mediates the rapid osteoclastogenesis in the fracture healing of orchiectomized mice

One hundred and forty-seven C57BL/6 J male mice of seven-week-old were purchased from and acclimatized for 1 week before orchiectomy at the Laboratory Animal Center of Southern Medical University (Guangzhou, China). The animal research protocol was approved by the Animal Care and Use Committee of Southern Medical University. All applicable Southern Medical University guidelines for the care and use of animals were followed. All procedures performed in studies involving animals were in accordance with the ethical standards of the Southern Medical University.

Mice were anesthetized using the MATRX VMR small animal anesthesia machine (model VMR; USA) with continuous inhalation of 2% isoflurane mixed with oxygen. The sham treated animals underwent skin and scrotum incision without the testis removal, while the ORX mice had both testes entirely removed as previously described [2]. One week after surgery, a cross-sectional fracture was generated in the mid-shaft of the right femur by inserting a pin (0.45 mm in diameter) into the marrow cavity from the distal end and then, penetrating through the mid shaft of the femur. An incision was made in the middle region of the right thigh to expose the mid-shaft of the femur into the intermuscle space. A complete fracture was made by cutting the shaft of the femur, where the intramedullary pin was remained to stabilize the fracture ends [12].

hPTH(1–34) and GR(1–34) were synthesized at GL Biochem (Shanghai, China). The preparation for administration followed the protocol as described before [9]. Both peptides were injected at the 40 μg/kg subcutaneously once daily for 5 days in every week [30]. The same volume of vehicle (0.05 ml) was administrated into control group. The body weight, serum testosterone, and the bone quantity and quality in the fracture area were measured after the 2 and 4 weeks of injection.

Half of one milliliter of blood was collected from the animals after general anesthesia, placed at room temperature for 15 min and then, centrifuged at 4000 rpm to obtain the serum. Enzyme-linked immunosorbent assay (ELISA) kits were applied to detect the levels of serum testosterone (Boster Biological Technology CO. Ltd., Wuhan, China), tartrated-resistant acid phsphatase (TRAP) (Cusabio, Wuhan, China), N-terminal propeptide of type I collagen (P1NP) and C-terminal collagen-type I fragments (CTX) (Immunodiagnostic Systems, Fountain Hills, AZ). Alkaline phosphatase (ALP) was detected with a colorimetric kinetic determination kit by following the manufacturer’s instructions (Byotime, Beijing, China).

After the mice were sacrificed by carbon dioxide anesthesia and cervical dislocation, the right femur containing the growth of the callus was fixed in 4% paraformaldehyde for 48 h. Then, the soft tissues and intramedullary pin were removed. Bone mineral density (BMD) and bone mineral content (BMC) at the callus were measured by dual-energy X-ray absorptiometry (DEXA) with a densitometer (XR-36, NORLAND Inc., WI, USA). The region of interest (ROI) was a rectangular area (5.4 mm × 3.5 mm) centering on the fracture line, which contained both the newly generated callus tissues and the original bone.

Micro-Computed tomography (Micro-CT) analysis was conducted by the μCT80 (SCANCO MEDICAL Inc., Switzerland) with the software μCT Evaluation Program V6.5–1 in the specimens after DEXA measurements. Image recording was confined to the callus of the fractured femur. Images of the femur mid-shaft (5.40 mm in length centering on fracture line) were performed using an isotropic voxel (12 μm in size). The grayscale threshold was set up at 220, meaning the values of mineralized tissue are greater than 220. The ROI of measurements was localized to a cylindrical space (5.4 mm in height and 3.5 mm in diameter) focusing at the middle point of fracture line. Three-dimensional (3D) pictures were also reconstructed based on the images from the three spatial dimensions to show the fracture healing. Samples from 12 mice from each group underwent micro-CT scanning and half of these samples were then tested for the biomechanical properties.

Six mice from each group were sacrificed by carbon dioxide anesthesia and cervical dislocation for biomechanics testing. The bone was immersed in PBS for 30 min after the removal of intramedullary pin and surrounding soft tissues, and then, placed between two plates (a span of 8 mm) with the medial and anterior sides facing forward and down, respectively. The bending rigidity of the healing femur was measured on the fourth week after injection by a three-point bending procedure using an Electropuls Test System (E1000; Instron, Inc., Illinois, USA). A central load was applied at the mid-callus with a constant rate of 2 mm/min until failure occurred. The rigidity of the femur was determined from the curve determined by the maximal force and displacement applied on the tested bone.

Collected femurs were decalcified in 20% ethylenediaminetetraacetic acid (EDTA) solution for 4 weeks, dehydrated, embedded in paraffin and then, sectioned in 5um for hematoxylin and eosin (HE) staining, Masson’s Trichrome staining (Maixin Biotech. Co. Ltd., Fujian, China), TRAP staining (Sigma-Aldrich, 387A-kit, USA), or TRAP immunohistochemical staining (Biosynthesis Biotechnology Co. Ltd., Beijing, China) with the standard protocols or the manufacturer’s instructions. 24 images (2 fields per section, 2 sections per sample; for callus formation analysis, magnification is 10; for osteoclasts analysis, magnification is 200) were photographed from 6 mice in each group and analyzed with Image J analysis software. ROI was defined as a rectangular box in the center of healing zone, in which the callus, woven bone and osteoclasts were outlined manually according to the specially stained color with the software.

The eight week old male mice were orchiectomized as described above. After one week of the orchiectomy, the mice were sacrificed by carbon dioxide anesthesia and cervical dislocation for femurs dissection and bone marrow collection. The isolated cells were seeded into 24-well plates precoated with collagen type I at 1 × 10⁵ cells/well for the 24 h culture, and then, induced for differentiation as described in the following. The osteogenic medium contained α-MEM supplemented with 10% FBS, 100 U/ml penicillin (Gibco, USA), 100 mg/ml streptomycin (Gibco, USA), 10 mM β-glycerophosphate (Sigma, USA) and 50 μg/m ascorbic acid (Sigma, USA). For osteoclastogenic induction, the cells were cultured in generic (α-MEM supplemented with 10% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin).

The treatment of hPTH(1–34) (10 nM) and GR(1–34) (10 nM) were followed a 4/48 h intermittent cycling plan [31]. In brief, the cells were cultured in the medium supplemented with peptides for 4 h and changed into the fore-mentioned medium without peptide for 44 h prior to the next treatment. At the 14th days of culture, ALP and mineralized nodules were examined by histochemical staining. The cells were fixed with 4% paraformaldehyde for 30 min, gently rinsed with PBS and stained for ALP with BCIP/NBT Alkaline Phosphatase Color Development Kit (Beyotime Institute of Biotechnology, Haimen, China), or for mineralized nodule demonstration with 1% Alizarin Red S solution (Sigma, St Louis, MO, USA) for 30 min at room temperature. ALP activity was measured by incubating cell lysates (extracted with 0.2% TritonX-100) in ALP substrate buffer containing the soluble substrate p-nitrophenyl phosphate. The quantification was performed based on the absorbance at 520 nm (Jiancheng Bioengineering Institute, Nanjing, China). To determine the calcium content of the cultures, cells were washed in Ca²⁺- and Mg²⁺-free PBS and then, incubated for 3 h in 0.2 ml of 0.6 N HCl. Extracted calcium was then measured spectrophotometrically at 610 nm after the reaction with methylthymol blue (Jiancheng Bioengineering Institute, Nanjing, China). TRAP staining was performed after the 7 days of treatment by using a TRAP staining kit. 10 microscopic fields (10 × 10) were randomly selected for the TRAP positive cells counting. The percentages of TRAP positive cells to total cells were calculated for quantification.

Bone marrow stromal cells were isolated and plated into 6 well plates at 4 × 10⁶ cells/cm² as described above. After 48 h cultured in the generic medium, cells were subjected to hPTH(1–34) (10 nM), hPTH(1–34) (10 nM) combined with 1 μmol/L U73122 (PLC inhibitor, Abcam, United Kingdom) or GR(1–34) (10 nM). Total RNA was isolated after 4 h later by using RNeasy Mini Kit (Takara BIO, Dalian, China). Expressions of Receptor Activator of Nuclear Factor κ B (RANK) and its ligand (RANKL), osteoprotegrin (OPG), nuclear factor of activated T cells (NFATC) 1, TRAP, Cathepsin K were measured by two-step real-time RT-PCR. Briefly, the first strand of cDNA was synthesized according to the manufacturer’s instructions using a PrimeScript® RT reagent Kit (Takara BIO, Dalian, China). For each gene, two specific PCR primers (RANK/fw, 5′- ACCTCCAGTCAGCAAGAAGT-3′, RANK/re, 5′-TCACAGCCCTCAGAATCCAC-3′; RANKL/fw, 5′-AGCCGAGACTACGGCAAGTA-3′, RANKL/re, 5′-GCGCTCGAAAGTACAGGAAC-3′; OPG/fw, 5′-ACCTCACCACAGAGCAGCTT-3′, OPG/re, 5’-TTGTGAAGCTGTGCAGGAAC-3′; NFATC1/fw, 5′- CCGTTGCTTCCAGAAAATAACA-3′, NFATC1/re, 5’-TGTGGGATGTGAACTCGGAA-3′; TRAP/fw, 5’-TCCTGGCTCAAAAAGCAGTT-3’;TRAP/re, 5’-ACATAGCCCACACCGTTCTC-3′; Cathepsin K/fw, 5’-CTGAAGATGCTTTCCCATATGTGGG-3′, Cathepsin K/re 5′- GCAGGCGTTGTTCTTATTCCGAGC-3′, GAPDH/fw, 5′-TGTCGTGGAGTCTACTGGTG-3′; GAPDH/re, 5′-GC ATTGCTGACAATCTTGAG-3′) were designed and synthesized by Life Technologies (Shanghai, China). The PCR reactions (94 °C, 20 s; 60 °C, 20 s; 72 °C, 20 s) were performed on an GeneAmp® PCR System 9700 (Applied Biosystems, CA, USA) using a SYBR® Premix EX Tap™ (Takara BIO, Dalian, China). Gene expression was normalized to that of GAPDH and then expressed as fold over control.

All statistical analyses were conducted using SPSS version 13.0 (SPSS Inc., Chicago, IL). The results are presented as means ± standard error of the mean (SEM). The significance of differences in BMDs, BMCs, the values of three-point bend testing and CT between treatment groups was analyzed by analysis of variance (ANOVA) with Bonferroni’s test for post hoc analysis.

Compared to the sham-operated mice, the serum testosterone in the orchiectomized mice began to decrease from the third week on until to the end of fifth week after ORX (Additional file 1: Figure S1A). Less trebacular bone and thinner cortical bone were detected in the growth plate of the tibia at the third and fifth week after ORX (Additional file 1: Figure S1B). Although the post-surgery bone volume [BV/TV (%)] was similar to that in the sham animals at the third week after ORX (p > 0.05), it was reduced dramatically at the fifth week after ORX compared to the sham animals, reflecting that the loss of bone volume was associated with the decreased testosterone (Additional file 2: Figure S1C).

To address the influence of the decreased testosterone on fracture healing, 3D reconstructions of the healing fracture after 2 weeks of fracture were performed and indicated that both the bone mass and bone volume in the fracture area were significantly decreased in ORX mice (Additional file 2: Figure S2A, B), which was coincided with the slow healing and bone shape recovery in the ORX mice after 4 weeks of fracture (Additional file 2: Figure S2A). Since the fracture healing was greatly retarded in ORX mice compared to the sham group (Additional file 2: Figure S2B), the ORX mice was an ideal model to study bone repairing during androgen deficient osteoporosis.

According to the 3D BMD and BV/TV (%) from micro-CT and the 2D BMD and BMC from DEXA, the bone callus was formed in hPTH(1–34), GR(1–34) and control groups after 2 week of fracture, but there were more bone tissues in the hPTH(1–34) and GR(1–34) treated animals than that of the vehicle controls (N = 12 in each group; Fig. 1a, left panel). Micro-CT scanning revealed that after 4 weeks of fracture, the cortical bones in the ORX mice treated with hPTH(1–34) or GR(1–34) were continuously aligned and the bone callus absorbed, indicating a completely recovered bone fracture. In contrast, the cortical bone treated with vehicle control was discontinuous and the bone callus was still prominent (N = 12 in each group; Fig. 1a, right panel). The BMD of the fracture treated with hPTH(1–34) and GR(1–34) were significantly higher than that of mice treated with the vehicle control after two weeks of fracture (Fig. 1b). Moreover, GR(1–34) exhibited a greater enhancement on 3D BMD than hPTH(1–34) (Fig. 1b). In addition to BMD, hPTH(1–34) and GR(1–34) also significantly increased the BV/TV (%) compared to the vehicle control (Fig. 1c), though the effect on BV/TV (%) of hPTH(1–34) was similar to that of GR(1–34) (Fig. 1c). Further investigations disclosed that after 4 weeks of fracture, the increasing BMD and BV/TV (%) of the healing sites in the hPTH(1–34) and GR(1–34) groups were both evidently greater than those in vehicle treated mice (Fig. 1b). Even the difference of BMD between the hPTH(1–34) and GR(1–34) treated mice after 2 weeks of fracture was diminished after 4 weeks of fracture (Fig. 1b). Consistent with the outcomes of 3D micro-CT reconstruction, the plain X-ray images confirmed that the well aligned cortical bone in the fracture ends (Fig. 2a). DEXA confirmed that hPTH(1–34) and GR(1–34) treatments significantly increased BMD and BMC after 2 weeks of fracture (Fig. 2b), as well as the greater BMC in GR(1–34) group compared with the hPTH(1–34) group (Fig. 2c). Similarly, the BMD and BMC of hPTH(1–34) were close to those of GR(1–34) treated mice after 4 weeks of fracture (Fig. 2b, c). In summary, the treatments with either hPTH(1–34) or GR(1–34) could increase the rate of bone healing, while the GR(1–34) appeared to have a greater enhancement on the early fracture healing than hPTH(1–34).

Since the the bone is too soft to endure the mechanical analyses before the fourth week of fracture, the bending force and rigidity of the femurs from the hPTH(1–34) and GR(1–34) treated and vehicle control mice was compared at the fourth week after fracture. The healing femurs from both the hPTH(1–34) and GR(1–34) treated mice were able to sustain a much greater force (Fig. 3a) and exhibited an increased bending rigidity (Fig. 3b). However, there was no significant difference in the measured biomechaniccal characteristics between the femurs of mice treated with GR(1–34) and hPTH(1–34).

The serum levels of P1NP, CTX and TRAP in both the hPTH(1–34) and GR(1–34) groups remarkably increased at the second week and returned normal as control at the fourth week after fracture (Fig. 4a, b, d). Interestingly, the ALP levels in both the hPTH(1–34) and GR(1–34) groups kept higher than control from the second to the fourth week after fracture (Fig. 4c). Moreover, hPTH(1–34) only induced a higher TRAP level than GR(1–34) at the second week after fracture (Fig. 4d).

Since GR(1–34) increased more BMD than hPTH(1–34) in the first two weeks after fracture, the histological features in the healing areas at the two weeks after fracture were analyzed to clarify the early healing process. Although no significant difference was found among the callus sizes of the three groups (Fig. 5a, b), the amounts of bony callus were significantly increased in both peptide-injecting groups, especially in the GR(1–34) mice compared with the vehicle group (Fig. 5a, c). Both the hPTH(1–34) and GR(1–34) treatments induced more osteoclasts around the woven bones, but the number and size of the osteoclasts in the hPTH(1–34) group were also more and larger than those in GR(1–34) group (Fig. 5a, d, e).

The bone marrow cells isolated from the femurs of ORX mice were applied in the in vitro exploration of the soteoblastogenesis and osteoclastogenesis during the fracture healing of the androgen-deficient osteoporosis. After 2 weeks of the intermittent administration, both hPTH(1–34) and GR(1–34) could induce the ALP activity and calcium deposition at the similar intensity (Fig. 6a, b). On the other and, the intermittent treatments of both hPTH(1–34) and GR(1–34) significantly increased TRAP⁺ cells in bone marrow cells culture in 7 days. However, less TRAP⁺ cells were detected in the culture treated by GR(1–34) compared to that by hPTH(1–34). In the hPTH(1–34) group, 10.2 ± 3.4% of the TRAP⁺ cells were multinucleated, while only 2.2 ± 1.3% in GR(1–34) group (Fig. 6c). There was no multinucleated TRAP⁺ cells found in vehicle treatment.

Quantitative PCR revealed that both hPTH(1–34) and GR(1–34) administration increased the transcription of RANK, RANKL and OPG, but only the increase of RANKL expression by hPTH(1–34) was significantly higher than that by GR(1–34) in the bone marrow cells of the ORX mice (Additional file 3: Figure S3). Similarly, the hPTH(1–34) treatment was capable of inducing the robuster transcription of NFATC1, TRAP and Cathepsin K than GR(1–34) (Additional file 3: Figure S3). Surprisingly, the efficient activation of RANKL, NFATC1, TRAP and Cathepsin K expression by hPTH(1–34) could be neutralized by the PLC inhibitor U73122 (Additional file 3: Figure S3).