Strategies to Suppress Early Bone Resorption Induced by High Dose Bone Morphogenetic Protein-2

Abstract

Objectives Recombinant human morphogenic protein-2 (rhBMP-2) can induce new bone formation by differentiation of mesenchymal stem cells(MSCs) into chondrocytes and osteoblasts, and it is used in clinical treatment for skeletal repair. However, there were many experimental data and clinical papers reporting that successful bone regeneration using rhBMP-2 required supra-physiological, high dosages which can induce several side effects including early bone resorption. Our goal in current study is (1) to identify factors which linkes to early bone resorption caused by supra-physiological dose of rhBMP-2, (2) to find specific factors leading suppression of early bone resorption, and (3) to establish a carrier system for the application of high dose rhBMP-2. First, in vivo and in vitro experiments were performed to investigate dose dependent effect of rhBMP-2 on early bone resorption and to evaluate the expression pattern change of early bone resorption and formation-related factors by supra-physiological dose of rhBMP-2. Nuclear factor of activated T-cells (NFAT)-c1 plays osteoclastogenesis inducer via c-fos pathway and is induced by BMP2. We investigated the effect of NFAT-c1 inhibitor on early bone resoprtion and long-term bone regeneration by high dose rhBMP-2. Sclerostin(SOST), as osteoclastogenesis inducer, is regulated by Receptor activator of nuclear factor kappa-B ligand(RANKL)/Osteoprotegerin(OPG) rate, which is up-regulated by rhBMP-2, was the second target in our study. Our study was performed to identify crosslink of estrogen and BMP2 signaling pathway in regulation of SOST expression for providing the mechanism of bone formation and resorption by high dose BMP2. The release kinetics of a carrier system can have a crucial influence on clinical effect of BMP-2. To overcome the shortcomings of absorbable collagen sponge (ACS), which shows initial burst release pattern of BMP-2 on bone regernation, we established a system for the sustained release of rhBMP-2 using heparin conjugated collagen sponge (HCS). In this carrier system, it was expected to modulate the functions of several heparin-binding molecules and we investigated the effect of high dose rhBMP-2 (40ug) on early bone resorption, long-term bone regeneration, bone erosion, and heterotopic ossification in HCS. The final goals of this study was to identify the target molecules which played key factors for osteoclast activation and to enhance the effect of rhBMP-2 treatment in short and long-term bone regeneration in bone therapy. Materials and methods To examine the local and peripheral effects of high dose BMP-2 release in vivo, 8mm critical-sized rat calvarial defect model was used, and the total region was divided into the defect area and the surrounding calvariae. New bone formation area was also divided into central original defect area and outside area of cental defect. To investigate the effect of NFAT-c1 inhibitor, the experimental groups were implanted with ACS including buffer only, rhBMP-2 (5 μg/defect), rhBMP-2 (20 μg/defect), or rhBMP-2 (20 μg/defect) + NFAT-c1 inhibitor (150 μM). To investigate the effect of HCS on early and long-term bone formation by high does rhBMP-2 (40μg/defect), HCS was used in a rat calvarial defect model. To confirm the appropriate dose of heparin in HCS for the experiments, HCS was prepared at different concentrations of heparin (0.2, 0.5 and 1.0 %) and characterized by scanning electron microscope (SEM), MTT assay, and Alizarin red s. As a result, 0.5% HCS and ACS was chosen as the control and 40 µg of rhBMP-2 was loaded in each scaffold for further animal experiments. For in vitro experiments, primary human MSCs (hMSCs), primary rat MSCs (rMSCs), and primary human osteoblasts were also cultured and used. 1 ng/µL of TOPFlash luciferase or Smad reporter plasmids, 30 pM human estrogen receptor (ER) alpha or SMAD family member 4 (Smad4) si-RNA, rhBMP-2 (50, 200 ng), estradiol (E2) (100 nM), and NFAT-c1 inhibitor (200 nM) each or together were treated in cells and the treated cells were then additionally cultured after 7 days. To evaluate mRNA relative expression of target genes, including insulin-like growth factor-1(IGF1), BMP2, and SOST, in each experiment as well as bone resorption- or formation-related marker genes, including RANKL, OPG, tartrate-resistant acid phosphatase (TRAP), cathepsin K, c-Fos, NFAT-c1, core binding factor α1 (Cbfa1), osterix, alkaline phosphatase (ALP), and vascular endothelial growth factor (VEGF), reverse transcription polymerase chain reaction (RT-PCR) and quantitative reverse transcription polymerase chain reaction(qRT-PCR) were performed. Based on mRNA expression levels, RANKL/OPG rate was also calculated. ELISA assay, Western blotting, ALP staining, TRAP staining, and immunofluorescence (IF) or immunohistochemistry (IHC) staining were performed to evaluate the expressions or activations of proteins including Sost, Beta-catenin, ER alpha, Smad4, Alp, Trap, Rankl, Opg, and Cbfa1. To evaluate the effect of rhBMP-2, NFAT-c1 inhibitor, or HCS on long-term bone regeneration, micro-CT analysis was also performed and bone volume (BV), tissue volume (TV), BV ratio (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and bone mineral density (BMD) were calculated. Masson trichrome staining (MT staining) was also performed. To determine the statistical significance of all experimental data, statistical analyses were performed using IBM SPSS 20.0 or 22.0 software. Groups were compared using a two-tailed Students t-test, a one-way ANOVA or two-way ANOVA, and Bonferroni's test. Differences with p < 0.05 were considered significant.

Results In 8mm critical-sized calvarial defects, 20ug or 40ug rhBMP-2 increased not only the expression of bone formation-related genes but also several bone resorption-related markers including osteoclast activation, RANKL/OPG rate, and the expression of bone resorption-related genes than 5ug rhBMP-2 did. NFAT-c1 inhibitor treatment with rhBMP-2 significantly reduced rhBMP-2 effect on those bone resorption-related markers. However, co-treatment of NFAT-c1 inhibitor also reduced the expression of bone formation-related genes. Micro-CT analysis of long term bone regeneration also reveal that co-treatment of NFAT-c1 inhibitor could lead to increased BV and BV/TV indicating effective new bone formation compared to rhBMP-2 only treatment, while BMD and other markers were similar in two groups. In hMSCs, rhBMP-2 (200ng/mL) increased SOST expression at 7 days. However, co-treatment of 100nM E2 supressed SOST induction by rhBMP-2 and increased the activation of beta catenin as key factor in Wnt signaling pathway. On the other hand, ER alpha siRNA and ICI 182,780 which was Wnt and ER alpha inhibitor could Smad-independently reduce the repressive effect of E2 on SOST induction by BMP2 and beta-catenin activation. In osteoblasts, the effect of E2 on SOST induction by rhBMP-2 differed among the human osteoblast lines. RANKL/OPG rate in two human osteoblast cell lines, which showed estrogen-mediated suppression of SOST induction by BMP2, also showed a similar pattern to SOST expression. All concentrations of heparin could lead to sustained release of rhBMP-2 for 4 weeks in HCS. 0.5% HCS showed different reponses to rhBMP-2 treatment between the defect area and the surrounding marginal bone. At the defect area in HCS, OPG expression was very low and RANKL expression, RANKL/OPG rate was increased 14.5 fold and 145 fold by rhBMP-2, and the expression of bone resorption-related genes was also increased. Especially, the increase in RANKL/OPG rate by rhBMP-2 was higher 51.9 fold and osteoclast activation was also more increased in HCS than in ACS. The expression of bone formation-related genes also showed similar pattern to those of bone resorption-related genes. In micro-CT analysis, BV and Tb.N indicating more effective long term bone regeneration at central defect was higher to a certain degree by rhBMP-2 in HCS than in ACS at 8 and 12 weeks, Histological staining was consistent with results of micro-CT analysis. On the other hand, the portion of the bone erosion region of ACS in rhBMP-2 treatment was 1.8 fold higher than that of HCS in rhBMP-2 treatment. Heterotopic ossification by rhBMP-2 was detected in ACS at 4 weeks. However, this side effect was less as time goes by and completely disappeared after 12 weeks.

Conclusions Supra-physiological dose of rhBMP-2 caused early bone resorption with bone formation by regulation of bone formation and resorption-related genes via BMP2 signaling in vivo. NFAT-c1 inhibitor reduced the effect of rhBMP-2 on early bone resorption-related factor and also led to better long term bone regeneration. However, the effect of NFAT-c1 was not showed in rMSCs, suggesting these roles of NFAT-c1 in rhBMP-2 signaling was limited via RANKL/OPG mechanism in osteoclasts. These results indicated that the limited repression of rhBMP-2 signaling in osteoclast during early phase can be an effective strategy for the reduction of early bone resorption and the induction of long-term bone regeneration. However, NFAT-c1 inhibitor also repressed early bone formation-related marker leading less effective bone regeneration. Therefore, additional studies were needed to conform suitable conditions of NFAT-c1 inhibitor. The increase in RANKL/OPG rate by rhBMP-2 also implied potential regulation of SOST as bone resorption-related marker by BMP2. In hMSCs, estrogen could smad4-independently repress the induction of SOST by BMP2 signaling cascade via Wnt signaling pathway including ER alpha and activated beta-catenin, indicating crosstalk between estrogen and BMP2 which might have a clue about the regulation of bone homeostasis by estrogen in osteoblast. Although further studies were needed to investigate estrogens effect on bone regeneration process including early bone formation and resorption, SOST regulation by estrogen might be a target in bone regeneration by rhBMP-2. Our studies showed that HCS reduced rhBMP-2-induced early bone erosion and heterotopic ossification and also enhanced long-term bone regeneration by low initial burst and sustained release of rhBMP-2 as well as heparin itself effect. Although not significant, site specific regulation of OPG by rhBMP-2 could be also observed. However, accumulation of rhBMP-2 by sustained release at defect area in HCS might be able to intensified early bone resorption, suggesting that sustained release of rhBMP-2 was considered with side effects in defect area. Our in vivo study using HCS also showed low initial burst and sustained release of rhBMP-2 could reduce rhBMP-2-induced early bone erosion and early heterotopic ossification and enhance long-term bone regeneration.