Pro-osteogenic effects of fibrin glue in treatment of avascular necrosis of the femoral head in vivo by hepatocyte growth factor-transgenic mesenchymal stem cells

Allogeneic MSCs were prepared, identified and transfected with Adenovirus vectors as described before[21]. For identification and differentiation assays of BMSCs, cells were induced to differentiate into osteoblasts, chondroblasts and adipocytes separately, by culturing cells in corresponding induction media (Cyagen Biosciences Inc., Goleta, CA). The differentiation characteristics were detected using NBT-BCIP Alkaline Phosphatase Color Development Kit (Beyotime Institute of Biotechnology, Shanghai, China) and Alizarin Red Sulfate (ARS) staining (Cyagen) for osteogenic differentiation, Alcian blue staining (Cyagen) for chondrogenic differentiation and Oil Red O staining (Cyagen) for adipogenic differentiation according to the instruction of the manufactures.

For compounding with fibrin glue, Ad-HGF or Ad-GFP-transfected MSCs were collected 48 h later. Two component solutions of fibrin glue were prepared separately. In solution I, freeze-dried human fibrinogen (Shanghai RAAS Blood Products Co. Ltd., Shanghai, China) was dissolved to 80 mg/mL with 2000 KIU/mL aprotinin in sterile water (Lanzhou Dadeli Biochemical Pharmaceutical Co. Ltd., Lanzhou, China) at 37°C. In solution II, thrombin (Jin Kang Pharmaceutical Co. Ltd., Nanjing, China) was dissolved to 400 IU/mL with 40 mmol/L of CaCl₂ in sterile water. For in vitro assays, 2 × 10⁴ cell pellet was first resuspended with 180 μL of solution I, and quickly mixed with 20 μL of solution II in one well of 24-well plates (Nunc, Thermo Fisher Scientific, Waltham, MA). The MSC-fibrin glue composite was cultured and observed for 1 week. For ANFH therapy in vivo, a duplex syringe was used to transfer 90 μL of solution I containing 1 × 10⁶ transfected MSCs and 10 μL of solution II into the lesions through the tunnel of the core decompression as described previously[21].

For SEM observation, fibrin glue composites containing MSCs were cut into small pieces and fixed with pre-cold fresh 2.5% glutaraldehyde-0.1 M PBS (pH7.4) at 4°C for 1 h. Following wash 3 times with PBS, samples were post-fixed with 1% osmium tetroxide-0.1 M PBS for 30 min and stained with 1% tannic acid-0.1 M PBS for 30 min. After dehydration through gradient concentration of ethanol, the specimens were dried at critical point and coated with gold by ion sputter, and observed with S-3000 N scanning electron microscope (Hitachi, Hitachi, JP).

To assay HGF secretion by MSCs compounded with fibrin glue, one week after compounding, fibrin glue was dissolved using 1600 U/mg of bovine pancreas trypsin (Calbiochem, La Jolla, CA) at 37°C. Cells released from fibrin glue composite were cultured for 24 h, and the HGF secretion in the supernatant was analyzed using an ELISA kit (BioSource International, Inc., Camarillo, CA) as per the manufacturer’s instruction. Results were read at 450 nm wavelength using a Varioskan Flash microplate reader (Thermo Fisher Scientific).

The proliferation activity and viability of MSCs compounding with or without fibrin glue were assayed using the Cell-Light™ EdU DNA Cell Proliferation Kit (Guangzhou Ribobio Co., Ltd, Guangzhou, China) and the Cell Counting Kit-8 (CCK-8, Dojindo Laboratorise, Tokyo, Japan) according to the manufacturer’s instruction and as described before[22].

Thirty New Zealand rabbits were maintained under specific pathogen-free conditions in The Experimental Animal Center of Nanfang Hospital (Guangzhou, China) and randomly assigned to 4 groups with the approve of the Animal Ethics Committee at Southern Medical University. Normal group included 5 animals untreated, while horse serum (Hyclone) and prednisolone acetate (Pharmacia & Upjohn Co., Kalamazoo, MI) were used to induce early stage of hormone ANFH in the remaining 25 animals as previously described[23]. Model group contained 5 rabbits with untreated ANFH. The remaining animals were separated evenly to 2 groups to receive treatment of HGF-transgenic MSC transplantation compounded with fibrin glue (HGF/MSCs + FG) or uncompounded (HGF/MSCs) as described previously[22]. To further demonstrate the activities of HGF-transgenic MSCs, the efficacy was further compared with another four treatments: FG only (FG), HGF-transgenic fibroblasts compounded with fibrin glue (HGF/Fibroblasts + FG), GFP-transgenic MSCs compounded with fibrin glue (GFP/MSCs + FG) and HGF-transgenic osteoblasts compounded with fibrin glue (HGF/OB + FG). Animals all survived during the observation periods and femoral head samples were obtained following air-euthanasia at 4 and 8 weeks post-treatment. Paraffin embedded-samples and sections were assayed with hematoxylin-eosin (HE) staining and immunohistochemistry with the following antibodies: osteocalcin (OCN) (OCG4; Abcam plc., Cambridge, UK; 1:100), CD105 (SN6h; Invitrogen co. Ltd., CA, USA; 1:75), and HGF (BOSTER Bioengineering Co. Ltd., Wuhan, China; 1:250), p-ERK1/2 (E-4; 1:250), p-Akt (D9E; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA; 1:250). Pictures were taken with microscopy (Nikon), while the number of empty lacunae and hematopoietic medullary cells as well as the integrated optical density (IOD) in 10 random fields per section was analyzed with Image-Pro Plus software version 6.0 (IPP 6.0, Media Cybernetics, Inc., MD). Three sections per animal were analyzed. Calculation of IOD was carried out according to the instruction of the software. After lining out positive staining areas in the sections, the area and the color depth were synthetically calculated by the software.

First, we isolated MSCs and determined their multi-potent abilities by differentiating into osteoblasts, chondroblasts and adipocytes (Figure 1A-D). Qualified MSCs were then transfected with Ad-HGF and compounded with fibrin glue. Pseudopods of GFP-expressing MSCs in the fibrin glue composite could be clearly observed under microscope, suggesting the healthy state of these compounded cells (Figure 1E). The MSCs extended on the surface of fibrin glue composite with normal morphology as shown by SEM, while internal cells mainly displayed a sphere phenotype with abundant surface villi commonly seen on healthy cells (Figure 1F). These results suggest that fibrin glue-compounded cells can survive and remain healthy.

To assess the effects of MSCs compounded with fibrin glue, we evaluated the proliferation and osteogenic differentiation in vitro. After dissolving the fibrin glue component of the MSC-fibrin glue complex one week later, the biological activity of compounded MSCs was determined and compared with uncompounded MSCs. The results of EdU incorporation and WST-8 assays showed that MSCs from the MSC-fibrin glue complex have a high proliferation rate similar as uncompounded MSCs resulting from the expression of HGF transgene (Figure 2A-C). The expression level of secreted HGF from transgenic MSCs was assayed using ELISA at 24 h after MSCs released from fibrin glue. The results showed that transgenic MSCs expressed similar level of secreted HGF no matter ever compounded with or without fibrin glue (Figure 2D). Moreover, MSCs from fibrin glue composite exhibited similar osteogenic differentiation ability as uncompounded MSCs, which was showed by ALP expression (Figure 2E) and alizarin red (ARS) staining (Figure 2F, G). The above results suggested that short-term compounding with fibrin glue will not change the biological activities of MSCs in osteogenic differentiation and responsive sensitivity to HGF stimulation.

After the ANFH model was established, the animals were given MSCs or MSCs + FG treatment. Similar as previous report, there is no significant difference could be observed between fibrin glue-compounded and uncompounded groups while obvious and similar therapeutic effects were observed in both treatment groups within 4 weeks[21]. However, the MSCs + FG group exhibited more significant improvement in recovery from ANFH when the observation period was extended to 8 weeks. Results of HE staining indicated that hematopoietic cells in the MSCs + FG group increased more significantly than that of the MSCs group and the number of empty lacuna was dramatically decreased in the MSCs + FG group compared with the MSCs group (P < 0.05) (Figure 3). The above results suggested better angiogenesis and formation of osteogenic microenvironment in MSCs + FG group during tissue repairing.

As a specific marker of bone formation during bone turnover, OCN could be used to characterize the initial osteogenic differentiation of MSCs[24, 25]. In addition, OCN was mainly expressed in perivascular vessels and part of osteocytes in normal tissues. In ANFH model, expression of OCN was almost lost completely. Such loss could be rescued by transplantation of Ad-HGF transfected MSCs regardless of the presence of fibrin glue in short-term treatment. However, the significant differences only could be clearly observed at the end of 8 weeks post treatment, OCN levels in the MSCs + FG group was significantly higher than that in the MSCs group (P < 0.05) (Figure 4A, B). As expected, transplantation of HGF transgenic MSCs restored the vascularization of the femoral head indicated by the expression of CD105, the marker of newly-formed blood vessels[26, 27]. Notably, the level of CD105 in the MSCs + FG group was also higher than that in the MSCs group (P < 0.05) at 8 weeks after MSC transplantation (Figure 4C, D). These results suggested that the duration of the MSC activity and the secretion of HGF were longer in the MSCs + FG group than that in the uncompounded MSCs group.

Significant improvement of ANFH defects compared with the FG group and the fibroblasts + FG group (P < 0.05) further demonstrated the effect of MSCs in the recovery of ANFH. Even being treated with osteoblasts combined with fibrin glue (the OB + FG group), the alleviation of ANFH symptom was also not equivalent as that in the MSCs + FG group which may be owing to the lack of vascularization activity by osteoblasts (Figures 3,4).

To explore underlying mechanism, we detected HGF and downstream signaling molecules in the local regions. Consistently, the local expression of HGF was already higher in the MSCs + FG group than in the MSCs group (P < 0.05) by the end of 4 weeks post treatment. This situation of higher expression of HGF in MSCs + FG group was maintained until the end of 8 weeks post treatment (Figure 5), which was consistent with better therapeutic effects achieved in the MSCs + FG group. ERK- and Akt-mediated signaling pathways are downstream of HGF stimulation[28] and their crosstalk contribute the fine-tune regulation of osteogenesis. Our previous results showed that dosage variation of local HGF levels could significantly affect the activation of the downstream ERK and Akt signaling pathways, and consequently regulate the proliferation and differentiation of MSCs as well as MSC-mediated osteogenic differentiation during the treatment of ANFH[22]. Therefore, the activation levels of ERK and Akt signaling pathways were compared between the MSCs group and the MSCs + FG group. IHC staining showed similar level of ERK1/2 signaling activation between the two transgenic MSC-transplanted groups by the end of 4 weeks post treatment. The difference appeared between 4 to 8 weeks. In the MSCs group, the level of ERK1/2 activation fell significantly compared with that by the end of 4 weeks (P < 0.05). However, sustained activation of ERK1/2 was observed in the MSCs + FG group at 4 weeks and 8 weeks post treatment. As expected, the activation level of Akt pathway was decreased in the MSCs group significantly at 8 weeks compared with 4 weeks, while the level of Akt activation was maintained in the MSCs + FG group. In short, although HGF-secreting MSCs in the MSCs group contributed significantly to the recovery of local ANFH lesion, fibrin glue-compounded transgenic MSCs could maintain longer survival period of MSCs and result in prolonged HGF secretion time which provide more sustainable and effective therapeutic effects in treatment of ANFH.