Effect of Human Wnt10b Transgene Overexpression on Peri-Implant Osteogenesis in Ovariectomized Rats

Abstract

This study aimed to investigate the efficacy of human Wnt10b (hWnt10b) transgene expression in ovariectomized (OVX) rats to accelerate osseointegration around titanium implants, and to provide a new strategy for treating osteoporosis with implants. An in vivo osteoporosis model was generated via bilateral ovariectomy in rats, and changes in expression of Wnt pathway-related genes were investigated. In OVX rats with a femur defect, hWnt10b expressed from an adenovirus vector was locally delivered to the defect site prior to implant placement. Surrounding femur tissues were collected 1 and 3 weeks after implantation for imaging, biomechanical testing, and molecular and histological analyses. In an in vitro model, bone-marrow stromal cells (BMSCs) transfected with adenovirus containing hWnt10b (Ad-hWnt10b) were cultured for 2 weeks in adipogenic medium followed by 2 weeks in osteogenic induction medium. Alizarin Red staining and Oil Red O staining, as well as reverse transcription polymerase chain reaction and Western blot analyses, were performed to assess the effect of hWnt10b expression on BMSC differentiation. Expression of Wnt pathway genes was significantly downregulated in OVX rats. OVX rats treated with Ad-hWnt10b prior to induction of a femur defect showed markedly increased ALP, Runx-2, and osteocalcin expression and decreased cathepsin K expression. Histological and imaging analysis showed increases in the number of osteocalcin-positive cells and the density of newly formed bone surrounding the implant in the Ad-hWnt10b group relative to the untreated control. Meanwhile, Ad-hWnt10b-BMSCs showed significantly increased osteogenesis and decreased adipogenesis. hWnt10b may accelerate osseointegration around implants and subsequently enhance bone regeneration and implant stabilization under OVX conditions.

Introduction

Progress in the theory and technique of oral implants has expanded the application of endosteal implants in prosthodontic and orthopedics therapies. Osseointegration plays a critical role in the fixation and anchorage of these implants following placement. For dental implants, osseointegration involves intimate contact between the bone and implant that enables the implant to improve intraoral function. Although dental implants have a high success rate, several local and systemic conditions such as osteoporosis, periodontitis, smoking, diabetes, and chemoradiation therapy are associated with implant failure.¹ Thus, new strategies to promote osseointegration between implants and surrounding bones, particularly in the presence of these conditions, are needed.

A meta-analysis showed that osteoporosis and dental implant failure are directly related.² Several methods have been examined to improve the success rate of dental implants under osteoporotic conditions, including implant surface modification, as well as topical or systemic administration of agents that can promote the recruitment and differentiation of osteoprogenitor cells (e.g., bone-marrow stromal cells [BMSCs]) or inhibit osteoblast formation, which in turn can accelerate osseointegration. Specifically, Gabet et al. applied human parathyroid hormone (PTH) 1-34 to promote bone remolding surrounding dental implants,³ whereas Duarte et al. administered alendronate to prevent bone loss around titanium implants that were inserted in estrogen-deficient rats.⁴

Bone formation and absorption depends on the balance between osteoblasts and osteoclasts, which is governed by several cytokines. The Wnts gene family member Wnt10b can specifically activate the canonical Wnt/β-catenin signaling pathway by binding to frizzled and low-density lipoprotein receptor-related protein-5 or -6 receptors (LRP-5/6) that regulate the growth and development of several organs and tissues, including bones.⁵ Christina et al.⁶ demonstrated that Wnt10b could stimulate osteogenic differentiation of BMSCs and inhibit expression of the adipogenesis-related transcription factor PPAR-γ, whereas Wnt10b silencing promotes adipogenesis of osteoblast-like MC3T3-L1 cells.⁷ Results from William et al. indicated that the downstream β-catenin was required for Wnt10b to influence these pathways.⁸ Moreover, mice overexpressing Wnt10b showed significantly higher bone mass and reduced bone loss resulting from estrogen deficiency, while Wnt10b^–/– mice showed notably decreased bone trabecula.⁹ This result is consistent with the finding that local delivery of a lentivirus vector encoding the Wnt10b gene (LV–Wnt10b) enhanced fracture healing, even in a more challenging atrophic non-union model.¹⁰ Meanwhile, Terauchi et al.¹¹ and Amelio et al.¹² reported that osteoporosis treated with immunoreactive parathyroid hormone (iPTH) increased Wnt10b production, which subsequently activated the Wnt/β-catenin signaling pathway and increased bone strength. Taken together, these findings suggest that Wnt10b plays a key role in both bone metabolism and development of osteoporosis.

Based on the hypothesis that Wnt10b can promote osteoblastic differentiation and is involved in the development of osteoporosis via the Wnt/β-catenin pathway, this study used ovariectomized (OVX) rats to investigate changes in expression levels of Wnt/β-catenin signaling components following estrogen depletion, as well as the efficacy of locally administered adenovirus-mediated human Wnt10b (hWnt10b) to improve osseointegration around titanium implants. To gain insight into the molecular mechanisms by which hWnt10b might regulate bone formation, the influence of Ad-hwnt10b expression on the osteogenic and adipogenic differentiation of BMSCs in vitro was also investigated.

Methods

Reagents

Construction of hWnt10b adenovirus

Adenovirus (Ad) containing hWnt10b was obtained from the Tufts University School of Dental Medicine (Boston, MA). This vector was produced following previously described sub-cloning steps.¹³ In brief, complementary DNA (cDNA) of full-length hWnt10b was cloned into a shuttle plasmid to generate pDC316-hWnt10b-GFP. Another fragment from the adenoviral construct derivative pBHGlox_E1, 3Cre was then co-transfected with the individual shuttle plasmid into 293 cells. This reconstructed adenoviral contained both green fluorescent protein (GFP) and hWnt10b cDNA segments. Ad-GFP was used as a control.

Implants

Cylindrical, threadless titanium implants (10 mm in length and 1 mm in diameter) were obtained from the Institute of Engineering Research Center in Biomaterials, Sichuan University (Chengdu, P.R. China). No surface treatments (e.g., porous coating or etching) were applied to the implants.

Animals

All animals were purchased from the West China Center of Medical Sciences, Laboratory Animal Center, Sichuan University. Sprague–Dawley rats (SD rats; 12 weeks old) weighing 270–300 g were used in this study. The animal use and care protocol complied with Institutional Animal Use and Care Committee guidelines.

Establishment of the OVX rat model

After a 1-week acclimation period, 30 SD rats were assigned randomly to three groups: OVX group, sham group, and normal group. OVX group rats were subjected to bilateral ovariectomy; sham group rats underwent sham surgery in which the ovaries were not removed; and normal group rats received no surgical treatment. Six months later, determination of serum estradiol level, HE staining, and micro computed tomography (micro-CT) examination were performed to confirm the presence of osteoporosis. Reverse transcription polymerase chain reaction (RT-PCR) and Western blot analysis were conducted to detect changes in the expression of Wnt signaling pathway-related genes.

Placement of dental implant in the OVX models and gene transfer in vivo

Forty-eight OVX SD rats were prepared, as described above, and 6 months after the first surgery, femoral implants were placed in each rat using a previously described implant surgery.¹⁴ Briefly, after intraperitoneal injections with chloral hydrate, all rats were treated by drilling a passage 1 mm in diameter through the intercondylar area in the distal femur metaphysis that penetrated the cortical bone monolayer to reach the marrow cavity. All procedures were performed with cooled sterile saline irrigation. Prior to implant placement, the rats were separated into three groups according to different treatments: 5 × 10⁷ plaque-forming units (pfu) of Ad-hWnt10b-GFP or Ad-GFP were injected into the femur defect in the Ad-hWnt10b-GFP group and the Ad-GFP group, respectively, while the virus-free group received equal amounts of sterilized phosphate-buffered saline (PBS). Then, the implant was slowly pressed into the slightly undersized bone defect before the incision was closed and prophylactic antibiotic administration was applied.

The animals were sacrificed 1 or 3 weeks after implantation, and the isolated femurs were prepared for micro-CT, confocal microscope scanning, histological evaluation, real-time RT-PCR, and immunohistochemical analysis. A total of five animals were examined for each time point in each group, and an additional six animals from each group were used for biochemical testing. The region of interest for examination was 1–4 mm below the metaphyseal growth plate in the distal femur.

Primary cell culture and gene transfer

BMSCs were harvested and cultured, as described before.¹⁵ Briefly, femurs were isolated from 4-week-old healthy rats immediately after euthanasia, and soft tissues were removed. Both ends of the long bones were excised, and low-glucose α-minimal essential medium (α-MEM) containing 10% fetal bovine serum (FBS) was used to flush the marrow. Single-cell suspensions were seeded into a flask and incubated at 37°C under a humidified atmosphere consisting of 95% air and 5% CO₂. Cells within four passages were used for the experiments.

The BMSCs were seeded and divided to three groups: the Ad-hWnt10b-GFP group, the Ad-GFP group, and the virus-free group that were treated with 10⁸ pfu Ad-hWnt10b-GFP, 10⁸ pfu Ad-GFP, and an equal volume of sterile PBS, respectively, when the cells reached 70–80% confluency. The activity of the transfected vectors was detected 3 and 7 days after transfection using inverted fluorescence microscopy. RT-PCR was used to confirm hWnt10b expression on days, 1, 3, 5, 7, 14, and 21.

For cell differentiation studies, cells were seeded in six-well plates, and upon reaching 70–80% confluence, the medium was changed to osteogenic (α-MEM with 10% FBS containing 100 nM of dexamethasone, 50 μg/ml of L-ascorbic acid phosphate magnesium, and 10 mM of β-glycerophosphate) or adipogenic (0.1 μM of dexamethasone, 10 μg/mL of insulin, and 0.5 mM of 3-isobutyl-1-methylxanthine; Sigma–Aldrich, St. Louis, MO) induction medium. Staining with Alizarin Red and Oil Red (see below) was used to assess the degree of mineralization and adipogenesis, respectively, whereas RT-PCR and Western blot analysis were used to detect expression changes of Wnt-related genes.

Alizarin Red staining and Oil Red staining

Alizarin Red staining was performed using Alizarin Red S (Sigma–Aldrich) at pH 4.0 on day 14 after treatment with osteogenic induction medium. The differentiated BMSCs were rinsed with PBS twice and fixed in 4% paraformaldehyde for 10–20 min at room temperature before staining with Alizarin Red S for 6–10 min and washing three times with PBS. If needed, 10% cetylpyridinium chloride was used to release the stain at room temperature for 15 min after observation using a microscope. The relative degree of mineralization was quantified by measuring the absorbance of the supernatants at 562 nm. These quantitative experiments were repeated three times with three duplicates each time.

Oil Red staining was performed on day 14 after treatment with adipogenic induction medium. The cells were fixed in 4% paraformaldehyde for 60 min after rinsing twice with PBS before staining with Oil Red O at 37°C for 60 min. After observation with an inverted microscope, isopropanol was used to release the stain, and the degree of adipogenesis was detected by measuring the absorbance of the supernatants at 540 nm. These quantitative experiments were repeated three times with three duplicates each time.

RT-PCR and Western blot analysis

Femoral tissues from around the implant were isolated and then ground in liquid nitrogen. The crushed bone and cultured BMSCs were dissociated using TRIZOL (Takara, Tokyo, Japan) and prepared for RNA extraction. cDNA synthesis was conducted using a PrimeScript RT Reagent Kit with a gDNA Eraser according to the manufacturer's instructions (Takara). Gene expression levels were quantified using SYBR^® Premix Ex Taq™ II (Takara) with an ABI PRISM 7300 instrument (Applied Biosystems, Foster City, CA). RT-PCR primer sequences are shown in Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/hum), and GAPDH was used as an internal standard. Results from at least five independent samples with three duplicates were used for data analysis.

Western blot analyses were performed, as previously described,¹ using samples from femurs ground and homogenized in liquid nitrogen and cultured BMSCs lysed with a Nuclear and Cytoplasmic Protein Extraction Kit (Signalway Antibody, Shanghai, China). Primary antibodies and horseradish peroxidase-conjugated secondary antibodies for Western blot were both purchased from Cell Signaling Technology (Danvers, MA). Final quantification of protein expression was performed using Image J software. Results from at least five independent samples with three duplicates were used for data analysis.

Micro-CT analysis

Femur samples were isolated and fixed in 4% paraformaldehyde. Micro-CT (u-CT80, SCANCO, Brüttisellen, Switzerland) was used for imaging tests at a working voltage of 70 kV, working current of 114 μA, integration time of 700 ms, and 18 μm resolution (2,048 × 2048 pixels, ρ = 1.2, support = 1, threshold for bone = 205, and threshold for implant = 700). Femur samples were mounted atop one another along the axis of the holder instead of horizontal to reduce artifacts arising from the metal. For further analysis, three-dimensional reconstruction was performed using the supplied software using an area of interest encompassing a region 36 μm–0.25 mm in diameter that surrounded the implant. Several parameters, including bone mineral density (BMD), bone value/tissue value (BV/TV) ratio, trabecular thickness (Tb.Th), and trabecular number (Tb.N) were quantitatively analyzed in 100 slices from 1 mm below the metaphyseal growth plate in the distal femur. Subsequent geometric rectification including “Reweighted Total Variation Algebraic Reconstruction Technique” was performed with the help of staff at the Engineering Research Center in Biomaterials at Sichuan University. Results from at least three independent samples were used for data analysis.

Histology analysis

Toluidine blue staining

Newly formed bone around the implant at 1 and 3 weeks after implantation was detected by toluidine blue staining. Three independent samples and a minimum of five sections for each sample were used to estimate the volume of aniline blue attained in the new osteoid matrix, and the relative percentage was further analyzed using Image-Pro Plus software. The degree of osteointegration (%OI) was defined as the percentage of new bone that was in direct contact with the implant surface.

Immunohistochemistry and immunofluorescence

Immunohistochemistry and immunofluorescence were performed, as previously reported.¹⁶ In brief, bone tissue samples taken from around the implant were processed by dewaxing, rehydration, and quenching endogenous peroxidase activity before the tissue was mounted on slides. The slides were incubated with osteocalcin (OCN) antibody at 4°C overnight, and the number of OCN-positive cells in 10 random fields at 100 × magnification was calculated for each paraffin section. The results from at least three independent samples and five sections for each sample were used for data analysis.

Immunofluorescence staining was applied using an affinity purified rabbit anti-β-catenin polyclonal antibody (Abcam, Cambridge, MA) to detect changes in β-catenin and hWnt10b expression in BMSCs from paraffin sections made 3 days after transfection; the nuclei were further stained with DAPI (Beyotime, Shanghai, China). hWnt10b expression was detected for all samples, and changes in β-catenin expression were detected for three independent samples, with five sections prepared for each sample. Inverted fluorescence microscopy (Olympus, Tokyo, Japan) was used to observe the staining.

Biomechanical analysis

Six femur samples from each group were used for biomechanical analyses after micro-CT scanning. Four weeks after gene delivery, samples from femur with implants were prepared, as previously described,¹⁷ for a push-out test to examine the maximum force and interfacial shear strength of the implants, which describe the biomechanical stability of the implant.¹ The femur samples were fixed in a specially designed holder, and the proximal end of each implant was exposed after epiphyseal separation, with the distal end of femur cut parallel to the distal end of the implant. The proximal side of implant was firmly fitted into the vertical channel in an adjustable holder of a commercial testing system (Instron 4302; Instron, Norwood, MA) to maintain compression along the vertical axis before a push-out test was conducted at a speed of 1 mm/min. Results from at least three independent samples were used for data analysis.

Statistical analysis

SPSS Statistics for Windows v17.0 (SPSS, Inc., Chicago, IL) was used for statistical analysis, and all quantitative data were assessed by one-way analysis of variance and expressed as mean ± standard deviation (SD). p-Values of <0.05 were considered to represent statistical difference.

Results

Establishment of osteoporosis rats and detection of changes in expression of Wnt-related genes

Detection of serum estradiol level

Six months after bilateral ovariectomy or sham surgery, serum estradiol levels in the three groups were detected. The serum estradiol level in the OVX group was significantly lower relative to animals in either the sham surgery group or the normal group (Fig. 1Ai).

Figure 1.

Confirmation of osteoporosis in ovariectomized rats and changes in Wnt expression levels. (A) Establishment of osteoporosis in rats 6 months after bilateral ovariectomy. (a–c) Hematoxylin and eosin (H&E) staining; (d–f) micro computed tomography (micro-CT) scanning of femur bones; (g) quantitative analysis of bone value/tissue value (BV/TV) by H&E staining; (h) quantitative analysis of BV/TV and connection density (Conn. D) by micro-CT scanning; (i) change in serum estradiol level. (B) Change in expression of Wnt/β-catenin signaling pathway genes 6 months after bilateral ovariectomy. (a) Reverse transcription polymerase chain reaction (RT-PCR) analysis of relative microRNA (mRNA) levels; (b and c) Western blot and corresponding quantitative analysis of protein levels. Data shown are the mean ± standard deviation (SD). Error bars indicate the SD. *p < 0.05 and **p < 0.01 versus control. Color images available online at www.liebertpub.com/hum

Hematoxylin and eosin staining

Femur bones were isolated from animals in the three groups and used for hematoxylin and eosin (H&E) staining. The results showed a significant decrease in both volume and density of cancellous bone in OVX rats. Moreover, the bone trabeculae in the OVX group displayed an irregular arrangement and fewer junction points, as well as an increased amount of adipose tissue in the bone marrow (Fig. 1Aa–c and g).

Micro-CT

Six months after the first surgery, femur bones were scanned, and three-dimensional images were reconstructed via micro-CT. The density of cancellous bone (BV/TV) and bone connection (connection density [Conn.D]) in the OVX rats was markedly lower than that in the other two groups (Fig. 1Ad–f and h).

Changes in expression of Wnt/β-catenin signaling pathway genes

RT-PCR analysis showed that Wnt10b, cyclinD1, and Lef1 expression levels were significantly downregulated in the femoral metaphysis of OVX rats 6 months after the first surgery. Each of these genes is involved in activating the classical Wnt signaling pathway. Meanwhile, expression of Wnt pathway antagonists such as DKK1 and Sfrp-1 was upregulated by more than threefold compared to untreated animals, whereas β-catenin gene expression levels were similar across all three groups. At the protein level, Western blot analysis showed a marked decrease in Wnt10b and β-catenin protein expression in OVX rats relative to untreated animals (Fig. 1B).

hWnt10b enhances osseointegration of implants in OVX rats

Successful transfection of Ad-hWnt10b-GFP vectors in peri-implant region

Green fluorescence arising from GFP was used as a marker of successful transfection of Ad-hWnt10b vectors. Both the Ad-hWnt10b-GFP and Ad-GFP groups showed GFP signals around the implant by confocal laser scanning microscopy of frozen sections of samples taken 3 days after implantation, which suggested that the hWnt10b was successfully transfected to the bone tissues surrounding the implant (Fig. 2A).

Figure 2.

Analysis of AdhWnt10b uptake. (A) (a and b) Fluorescence intensity surrounding the implant in the Ad-hWnt10b-GFP and Ad-GFP groups; (c) Few green fluorescent protein (GFP) fluorescence signals were seen in the virus-free group. (B) Histological features of the proximal femur around the implant at (a–c) 1 week and (d–f) 3 weeks after implantation (100 × ). (g) A circle with a 0.25 mm diameter around the implant was selected as the area of interest; (h) % bone implant contact (BIC) around the implants. At 1 and 3 weeks after implantation, the % BIC of the Ad-hWnt10b-GFP was significantly higher than the other two groups. Data shown are the mean ± SD. Error bars indicate SD. **p < 0.01 versus control. Color images available online at www.liebertpub.com/hum

hWnt10b transgene modification accelerated osseointegration of implants

Toluidine blue used to stain femoral specimens taken from OVX rats at 1 and 3 weeks after implantation showed areas of new bone formation around the implants. The Ad-hWnt10b group had obviously larger and thicker staining areas compared to the other two groups. Meanwhile, histomorphometric analysis showed that the bone-implant contact in the Ad-hWnt10b group was 82% or around 1.3-fold higher than the other two groups (p < 0.01; Fig. 2B).

Transverse and longitudinal three-dimensional images on micro-CT indicated higher bone volume, bone mineral density, and more trabecular bones surrounding the implant in the Ad-hWnt10b-GFP group compared to the Ad-GFP and virus-free groups. Specifically, Tb.N, BV/TV ratio, Tb.Th, and Conn.D were all about 1.3–1.5 times higher than the control groups, whereas the trabecular separation (Tb.Sp) was 26% lower in the Ad-hWnt10b group (Fig. 3).

Figure 3.

Micro-CT image analysis of treated regions. (A) Longitudinal (a–c and g–i) and transverse (d–f and j–l) 3D images of the femur with the implant. (B) 3D microarchitectural indexes analyzed by micro-CT. The hWnt10b modification group of OVX animals showed a marked increase in peri-implant bone volume at weeks 1 and 3. Data for BV/TV, trabecular thickness (Tb.Th), Conn.D, trabecular separation (Tb.Sp). and trabecular number (Tb.N) are expressed as mean ± SD. Error bars indicate SD. *p < 0.05 and **p < 0.01 versus control.

hWnt10b upregulated expression of osteoblastogenesis-related genes and decreased osteoclast formation in OVX rats

RT-PCR analysis conducted at 1 and 3 weeks after implantation and hWnt10b transgene modification indicated that Wnt10b, ALP, Runx-2, and OCN gene expression in newly formed bone of the Ad-hWnt10b group was two- to fourfold higher than the other two groups, and the expression level of the osteoclast-related gene cathepsin K (CPK) was much lower in the Ad-hWnt10b group compared to the other two groups (Fig. 4h and i)

Figure 4.

Immunohistochemistry and RT-PCR analysis of gene expression changes. (A–F) Representative images of osteocalcin (OCN) immunohistochemical staining surrounding the implant at 1 and 3 weeks after implantation. Black arrow: OCN-positive staining. (G) OCN-positive cell/total cells at 1 and 3 weeks after implantation. (H and I) RT-PCR results for Wnt10b, ALP, RUNX2, OCN, and CPK expression at 1 and 3 weeks after implantation. The data shown are the mean ± SD from three independent experiments. Error bars indicate SD. **p < 0.01 versus control. Color images available online at www.liebertpub.com/hum

The immunohistochemistry results also showed that at 1 and 3 weeks after implantation, the number of OCN-positive cells around the implant was significantly higher in the Ad-hWnt10b group compared to the other two groups (Fig. 4a–g).

Biomechanical analysis

A total of six femur samples for each group were included in a push-out test performed 4 weeks after gene transfection. The Ad-hWnt-10b group exhibited a marked increase in both maximum force required to dissociate the implant and the ultimate shear strength relative to the other groups, which is indicative of higher implant stability (Table 1).

Table 1.

Biomechanical test

Groups	Maximal push-out (N)	Interfacial shear strength (N/mm2)
Ad-hWnt-GFP	58.51 ± 8.11**	5.38 ± 0.36*
Ad-GPP	21.37 ± 4.37	3.61 ± 1.13
Virus-free	28.39 ± 3.96	3.95 ± 0.98

Data are expressed as mean ± standard deviation, n = 6 femurs/group.

^* p < 0.05 Ad-hWnt10b-GFP versus control groups (Ad-EGFP and virus-free group); ^**p < 0.01 Ad-hWnt10b-GFP versus control groups (Ad-EGFP and virus-free group).

N, Newton.

hWnt10b enhanced osteogenic differentiation and blunted adipogenic differentiation of BMSCs in vitro

Persistent expression of hWnt10b in BMSCs

Successful transfection of Ad-hWnt10b to BMSCs was indicated by GFP co-expression that persisted for >21 days after a single transfection. Furthermore, RT-PCR subsequently confirmed that hWnt10b was expressed in transfected BMSCs and showed a gradual elevation, peaking 7 days after transfection. The expression levels then slowly returned to initial levels (Fig. 5A).

Figure 5.

In vitro analysis of bone-marrow stromal cells (BMSCs). (A) Gene expression of (a) BMSCs transduced with Ad-hWnt10b-GFP, (b) BMSC control groups, and (c) hWnt10b expression in Ad-hWnt10b-GFP-transfected BMSCs analyzed by RT-PCR on days 1, 3, 5, 7, 14, and 21. (d) PCR results for hWnt10b in vitro in isolated cultured BMSCs administered with Ad-hWnt10b-GFP compared to Ad-GFP and virus-free control groups. (B) Immunofluorescence staining of β-catenin expression in BMSCs (100, scale bars = 100 μm). (a–c) Ad-hWnt-10b group, (d–f) Ad-GFP group, and (g–i) virus-free group. β-catenin stained red in the cytoplasm (a, d, and g) and nuclei were stained blue (b, e, and h). Merged images are shown in panel c, f, and i. (C) (a–f) Alizarin Red staining of BMSCs cultured in osteogenic medium. (g–i) Oil Red O staining of BMSCs cultured in adipogenic medium (Alizarin Red staining: magnification 100 × , scale bars = 100 μm; Oil Red O staining, magnification 200 × , scale bars = 50 μm). More mineral deposits were found in the Ad-hWnt10b-GFP group, whereas fewer lipid droplets were observed in the Ad-hWnt10b-GFP group. (j–l) Relative mRNA levels and Western blot assay of PPAR-γ, FABP4. (m–o) Relative mRNA levels and Western blot assay of Runx-2 and OCN. The data shown are the mean ± SD from three independent experiments. Error bars indicate SD. *p < 0.05 and **p < 0.01 versus control. Color images available online at www.liebertpub.com/hum

Sustained overexpression of hWnt10b activated the Wnt pathway

At 3 days after transfection, immunofluorescence staining was performed to detect changes in nuclear and cytoplasmic β-catenin expression and to investigate the effect of Ad-hWnt10b on the canonical Wnt pathway in BMSCs. Compared to the control group and the virus-free group, the Ad-hWnt10b group showed much higher amounts of red fluorescence, indicative of β-catenin expression, in both the cytoplasm and nucleus. The accumulated β-catenin reflected activation of the canonical Wnt pathway, suggesting that hWnt10b overexpression in rat BMSCs of rat can activate the endogenous Wnt pathway in vitro (Fig. 5B).

In vitro overexpression of hWnt10b enhanced osteogenic differentiation and blunted adipogenic differentiation of BMSCs

Alizarin Red and Oil Red O staining were performed on BMSCs 14 days after transgene modification to assess the effect of hWnt10b on adipogenic or osteogenic induction, respectively, as well as its effect on BMSC differentiation. Compared to the control group and the virus-free group, a threefold increase in the number of mineralization nodules was seen for the Ad-hWnt10b group. The staining area in the Ad-hWnt10b group was also significantly larger than that of the other two groups. Meanwhile, mRNA levels of Runx-2 and OCN, which represent a major osteogenic transcription factor and a late osteogenic marker, respectively, were markedly increased in the Ad-hWnt10b-GFP group. These results persisted at the protein level, as shown by Western blot analysis.

In contrast, notably fewer Oil Red O-positive adipocytes were observed in the Ad-hWnt10b group, as evidenced by a reduction in Oil Red O staining density by one half to one third for the Ad-hWnt10b group relative to the other two groups. RT-PCR analysis also showed significantly decreased mRNA levels of PPAR-γ, a key transcription factor for adipogenesis, as well as FABP4, an adipocyte marker, in the Ad-hWnt10b group. Results from Western blot analysis further supported this finding (Fig. 5C).

Discussion

Osteoporosis is characterized by decreased bone mineral density and bone quality,¹⁸ it and affects millions of individuals worldwide. The detailed pathogenesis of osteoporosis due to decreased estrogen levels remains poorly understood. The Wnt protein family is thought to be involved in the development of bone tissue wherein binding of membrane-bound receptors activate downstream signaling pathway thatpromotes osteogenic differentiation and inhibits adipogenesis in BMSCs.^19–21

This study built a rat osteoporosis model by performing a bilateral ovariectomy. The OVX rats showed a significant decrease in both serum estradiol level and systematic bone volume and density 6 months after surgery. To detect potential mechanisms involved in osteoporosis, changes in gene expression were analyzed for the OVX group, the sham group, and the normal group. Six months after surgery, the expression of several Wnt pathway-related genes such as Wnt10b, cyclinD1, and Lef1 was significantly inhibited in the OVX group, and expression of Wnt pathway antagonists, including DKK1 and Sfrp-1, were markedly increased. Together, these results indicated that inhibition of the Wnt pathway might be involved in the occurrence and development of estrogen decline that induces osteoporosis.

With improvements in dental and orthopedic implant technology, the use of implants is increasing for orthopedic and prosthodontic therapies. The management and improvement of these therapies focuses in part on defining parameters that can accelerate osseointegration between implants and surrounding bone tissues. Osteoporosis can affect the degree of osseointegration and in turn the success of treatments for arthritis and tooth loss. For example, previous studies have shown that the presence of osteoporosis can significantly increase the risk of dental implant failure due to insufficient bone-to-implant connection.²

Based on previous results and the finding here that genes involved in the classical Wnt pathway exhibit changes in expression, the role of Wnt10b in implant osseointegration was examined further. Wnt10b is a member of the Wnts secretory protein family that has been demonstrated to play a prominent role in resisting bone resorption due to a decline in estrogen levels and promoting bone formation.^6,9 PTH has also been approved for osteoporosis therapy due to its positive effects on bone architecture and strength. Terauchi et al.¹¹ showed that the application of PTH could increase the production of Wnt10b, which subsequently activates the Wnt/β-catenin signaling pathway in osteoprogenitor cells, and enhance bone strength. Based on these findings, it was hypothesized that local administration of Wnt10b could be a promising therapeutic strategy to promote bone formation and osseointegration around implants under osteoporotic conditions.²²

In the current study, first a recombinant adenovirus carrying the hWnt10b gene fused to GFP was constructed, and the resulting vector, Ad-hWnt10b-GFP, was delivered to the site of the femur defect created before placement of an implant in the OVX rat model. It was shown that the hWnt10b transgene was successfully transfected to the bone tissue of rats and that it subsequently activated the classical Wnt pathway. Toluidine blue staining and micro-CT further showed that the delivered hWnt10b promoted new woven bone formation and bone implant connection, whereas biomechanical tests revealed increased implant stability in the Ad-hWnt10b group. Immunofluorescence analysis showed a higher number of OCN-positive cells surrounding the implant in the Ad-hWnt10b group. RT-PCR and Western blot analysis demonstrated that hWnt10b delivery increased gene and protein expression of ALP, Runx-2, and OCN but inhibited expression of the osteoclasis-related gene CPK. From these results, it was concluded that hWnt10b could effectively reverse the loss of bone mineral density and bone volume around an implant and accelerate osseointegration between implant and surrounding bone tissues under osteoporotic conditions by promoting and inhibiting osteogenesis and osteoclasis, respectively.

Osteoclasts do play an essential role in resorption of defective bone during bone remodeling. Although the short-term results obtained in the study could provide clues about how hWnt10b promotes bone formation and osseointegration of implants, the potential long-term negative biomechanical impact on bone tissues resulting from the inhibition of cathepsin K was not examined. Thus, further study is needed to assess how Wnt10b affects bone formation and resorption from the perspective of osteoclast function.

This study also investigated the effects of hWnt10b on adipogenesis and osteogenesis of BMSCs derived from SD rats. BMSCs are multipotent stem cells that could facilitate regeneration and remodeling of bone tissues and other mesenchymal tissues such as adipose tissue. Upon development of a bone defect or a fracture, the expression of several cytokines such as SDF-1 is upregulated, and the newly synthesized cytokines are released by injured tissues to mobilize BMSCs to the defect sites where they can proliferate and differentiate into osteoblasts.^23,24 In this study, BMSCs transfected with Ad-hWnt10b-GFP continually expressed hWnt10b over 21 days. The transfected BMSCs also showed significant upregulation of intracellular β-catenin expression, which reflects activation of the classical Wnt pathway. In addition, 14 days after adipogenic or osteogenic induction, Alizarin Red and Oil Red O staining together with RT-PCR and Western blot analyses showed that hWnt10b overexpression in the BMSCs could activate osteogenic differentiation and suppress adipogenic differentiation. This result is consistent with previous findings showing that osteoblasts and adipocytes are both derived from BMSCs and share an inverse relationship in the bone-marrow stroma, wherein increased expression of osteogenic transcription factors is associated with downregulation of adipogenic transcription factors.^25,26 Another study showed that Wnt10b could effectively promote the differentiation of BMSCs (ST2 cell line) to osteoblasts by upregulating Runx-2 and OST expression while downregulating PPAR-γ and FABP4 expression.⁵ Accordingly, the mechanism by which hWnt10b promotes formation of new woven bone around the implant in OVX rats might be related to the promotion of osteogenic differentiation by surrounding BMSCs.

Gene therapy is a new technology that treats diseases at the DNA level with long-term efficacy.²⁷ Several groups have achieved satisfactory results for applications of gene therapy to treat bone defects or to promote implant osseointegration.^1,10,14 Although the current investigations showed that hWnt modification was associated with significantly increased bone volume, bone mineral density, and bone-implant connections in OVX rats, this was a preclinical study, and additional safety, ethical, and technical issues must be considered and resolved prior to clinical application.

In summary, the present results provided reliable evidence that the Wnt pathway is involved in osteoporosis due to estrogen decline, and hWnt10b transgene modification could promote osseointegration around an implant and accelerate the bone remodeling process under OVX condition in rats. Furthermore, Wnt10b appeared to shift the balance between osteogenic/adipogenic BMSC differentiation toward osteogenesis. The findings also suggest that hWnt10b gene therapy could be an effective strategy to promote osseointegration of implants and subsequently enhance bone regeneration and implant stabilization for osteoporosis patients.

Author Disclosure

No competing financial interests exist.