Suppression of zinc finger protein 467 alleviates osteoporosis through promoting differentiation of adipose derived stem cells to osteoblasts

Mouse abdominal adipose tissues were obtained under approval from the Animal Research Guidelines of Shang Jiao Tong University School of Medicine. The lipoaspirate was incubated with collagenase type I solution (Worthington Biochemical, Lakewood, NJ) for 1 h at 37°C, and filtered through 250 μm filters. Following centrifugation, the stromal vascular fraction was resuspended in DMEM (HyClone, Logan, UT) supplemented with 10% FBS (HyClone), 100 U/ml penicillin, and 100 μg/ml streptomycin. ADSCs were cultured under a humidified atmosphere of 5% CO₂ at 37°C for long-term culture in vitro and were capable of differentiating into adipocytes and osteogenic cells under specific induction [19].

Differentiation in cultured ADSCs was monitored including oil red O staining of lipid droplets in terminal adipocyte differentiation and von Kossa staining of calcium deposition in the extracellular matrix in terminal osteoblast differentiation.

The lentiviral vector encoding Zfp467 shRNA1 (KD1) (5'-TTAGCTCATGAACGCAGACC-3') and shRNA2 (KD2) (5'-AGTCGTACCACAGAACTGTC-3') under the human H1 RNA polymerase III promoter has been described [20]. A si-Zfp467 expression unit was generated by inserting hybridized oligo DNAs into pBShH1 plasmid DNA. After sequence confirmation, two si-Zfp467 expression units were excised from the pBShH1 plasmid DNA by XbaI and XhoI digestion and inserted into XbaI/XhoI sites of the FG12 lentiviral vector. A scrambled si-RNA was used as a control.

We generated lentiviral vector stocks using an HIV-1-based reporter virus, packaging plasmid pCMV R8.2ΔVpr and the VSV-G envelope protein-coding plasmid by calcium phosphate-mediated transient transfection of 293T cells. Culture supernatants were harvested on 2 days after transfection. Lentiviral vector particles were concentrated 300-fold by ultracentrifugation. Concentrated virus stocks at a multiplicity of infection of 10 were titrated on 293 T-cells based on EGFP expression.

ADSCs were transduced ex vivo with the VSV-G pseudotyped shRNA vector at a multiplicity of infection of once per day for 2 days with gentamicin. Transduction efficiency was analyzed by quantifying EGFP expression by flow cytometry 48 h after transduction.

After eight-week-old female ddY mice had been ovariectomized, ADSCs (2 × 10⁶ cells/mouse) labeled with iron oxide (Ferridex, Berlex Laboratories Inc, Wayne, NJ) were injected into OVX mice via tail vein on postoperative day 4 and sacrificed at day 24 after injection. Bone tissues were fixed and decalcified with EDTA solution. After washing, paraffin-embedded specimens were sectioned, deparaffinized, incubated with 1% potassium ferrocyanide in 1% HCl for 30 min, and counterstained with nuclear fast red. Six animals were included in each group.

For detection of mRNA level, total RNA was isolated with TRIzol (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions. About 2 μg of total RNA was reverse transcribed using Moloney Murine Leukemia Virus Reverse Transcriptase (Promega, Madison, WI) with oligo dT at 42°C for 1 h. Detailed information on PCR, including primer sequences and cycles, is provided in Table 1.

For Western blotting, cells were lysed by addition of lysis buffer containing 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, 1 mM EDTA, 0.1% SDS, protease inhibitors (Complete tablets, Roche Molecular Biochemicals, Mannheim, Germany), 1 mM Na₃VO₄, and 1 mM NaF. Proteins were separated by 10% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with Zfp467 antibody (R&D Systems Inc., Minneapolis).

Urine deoxypyridinoline was assayed by competitive enzyme immunoassay using the MicroVue DPD EIA kit (Quidel Corporation, Santa Clara, CA).

ADSCs were harvested by trypsinization and centrifugation, washed twice with cold PBS, and fixed with cold 75% ethanol at 4°C overnight. The fixed cells were collected, washed twice with PBS and suspended in PBS. Flow Cytometry was using to array cultured ADSCs at day 7 with FITC-conjugated anti-CD29, FITC-conjugated anti-CD44, PE-conjugated anti-CD45 and PE-conjugated anti-HLA-DR antibody (Abcom Inc., Cambridge, Massachusetts), respectively.

The cDNAs retrotranscribed from an equal amount of mRNA from control (ADSCs at day 0), adipocyte and osteoblast derived from ADSCs at day 14, were labeled with Cy3 and Cy5 fluorescence as probes, respectively. Fluorescence intensity was measured for each gene spot. After background subtraction and the whole-chip data normalization, the ratios of gene expression differences between these sets of samples were obtained. BioStar H-40s microarray consisted of 4096 novel or known genes including control system and effective genes (provided by Biostar Genechip Inc). The control system consists of 96 housekeeping genes as loading control; 16 plant genes, and spotting solution (without DNA, 16 spots) as negative control spots in the array. The chips were scanned with a ScanArray 4000. The acquired images were analyzed using GenePix Pro 3.0 software.

A total of 2 × 10⁶ cells were injected into sham-operated or ovariectomized female ddY mice (8-week-old, Central lab animal, Korea) via tail vein on postoperative day 4 and sacrificed at day 24 after injection (n = 6 per group). Microcomputed tomography (μCT) and histological analysis were performed, as reported previously [7, 21]. Trabecular morphometry within the proximal tibia was quantified using high resolution microcomputed tomography (μCT, Skyscan 1076 μCT, Aartselaar, Belgium). From μCT data, bone loss indices, including bone volume/total volume (BV/TV), trabecular number (Tb.N), and bone mineral density (BMD) were assessed. For analysis of bone formation, mice were injected with calcein (10 mg/kg) on postoperative day 14 and day 21 and were sacrificed at postoperative day 24.

OC number (no.) per bone surface (NOc/BS) and OB number per bone surface (NOb/BS) were analyzed by tartrate-resistant acid phosphatase (TRAP) staining on the 7-mm thick sagittal sections, as described previously [22]. TRAP positive cells containing one or more nuclei and sitting on the surface of the trabeculae were defined as osteoclasts. One section per animal was analyzed for these parameters.

The results were expressed as mean ± SD. The data were treated by Student's t test to determine statistical significance. We used nonparametric tests (Mann-Whitney test), if appropriate, to compare differences. P < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS 12.0 software.

ADSCs were obtained from the abdomens of 3- to 6-week-old mice and primary ADSCs displayed typical fibroblast-like morphology and were maintained for long-term culture in vitro (Figure 1A-D), and were capable of differentiating into adipocytes (Figure 2A-C) and osteogenic cells (Figure 2D-F) under specific induction [19]. Cell surface marker expression by ADSCs isolated from human adipose tissue was characterized. Flow cytometric analysis of ADSCs revealed that high levels of MSC-related CD29 and CD44 were expressed (Figure 1E, F); however, hematopoiesis-related antigens CD45 and immunogenicity markers HLA-DR were not expressed (Figure 1G, H).

In additional, the ability of ADSCs to differentiate into adipocyte and osteoblast lineages at passage 3 was experimentally confirmed by Oil red O staining (Figure 2B), von Kossa staining (Figure 2E) and the mRNA levels of specific adipogenic (adiponectin) (Figure 2C) and osteogenic (alkaline phosphatase) markers (Figure 2F), respectively. These results showed that ADSCs have multilineage potential and characteristics of MSCs.

Next, we studied the difference of gene expression profile changes between adipocyte and osteoblast differentiation from ADSCs at day 14 and to screen the critical genes in the early differentiation stage by cDNA microarray. The scatter plots that were plotted with Cy3 and Cy5 fluorescent signal values displayed a quite disperses pattern in distribution. Most of the spots gathered around a 45° line, in which red spots represented the area where the signal intensities varied between 0.5 to 2-fold compared with those of the control. Some yellow spots distributed beyond or far from 45° line indicated the existence of abnormal gene expressions between adipocyte and osteoblast derived from ADSCs at day 14. Their signal intensities were 2 times more than that of the control (Figure 3A-C). As shown in Figure 3D-F, 298 genes were screened out that exhibited different expressions between ADSCs, adipocyte and osteoblast derived from ADSCs, there were 175 up-regulated and 123 down-regulated genes in the gene expression profiles of ADSCs which was 2 times of that in adipocyte and osteoblast derived from ADSCs.

We then choose Zfp467, a novel regulator of osteoblast and adipocyte commitment [18], to further study the molecule mechanism of ADSCs differentiation. The mRNA and protein levels of Zfp467 in adipocyte and osteoblast derived from ADSCs (Day 14) and control (ADSCs, Day 0) were measured using qRT-PCR and Western blot analysis, respectively. The results showed that the mRNA levels of Zfp467 in adipocyte derived from ADSCs was notably increased when compared with that in the control, however the mRNA levels of Zfp467 was significantly decrease after ADSCs differentiation to osteoblast (Figure 4A). The change of Zfp467 protein levels showed to be the same on the mRNA levels of Zfp467 in three groups (Figure 4B).

We then created a lentiviral vector encoding Zfp467 shRNA for long-term expressing Zfp467 siRNA to knockdown Zfp467 gene expression. We used two different siRNA sequences to target human Zfp467 gene, termed as Zfp467 shRNA-1 (KD1) and Zfp467 shRNA-2 (KD1). Control cells were transfected with lentiviral vector expressing scrambled siRNA. Zfp467 shRNA-1 and shRNA-2 dramatically reduced Zfp467 expression at both mRNA (Figure 4C) and protein levels (Figure 4D). The inhibition rate of Zfp467 shRNA-1 and Zfp467 shRNA-2 were 64.4% and 83.6%, respectively in ADSCs compared with the mock and control group at mRNA levels (Figure 4C). However, shRNA-2 caused more profound reduction of Zfp467 gene expression when compared to shRNA-1, reflecting a sequence-specific effect of the siRNA (Figure 4C, D).

As mentioned above, in two representative subclones, Zfp467 shRNA-2 caused more profound reduction of Zfp467 gene expression when compared to shRNA-1. Hereafter, we mainly used the Zfp467 shRNA-2 subline for further analysis. The expressions of specific osteogenic and adipogenic genes were evaluated by real-time RT-PCR at 0, 3, 7 and 14 days post-transduction. Our results indicated that specific osteogenic marker Osteocalcin (OCN), Osteopontin (OPN), Collagen I (COL I) and bone sialoprotein (BSP) were up-regulated in si-Zfp467 transduced cells (Figure 5A-D). Specific adipogenic marker lipoprotein lipase (LPL) and peroxisome proliferator-activated receptor gamma (PPARγ) were down-regulated in si-Zfp467 transduced cells (Figure 5E-F). The statistical analysis showed that the plasmid Zfp467 shRNA could have significant inhibitive effects on the mRNA levels of specific adipogenic gene at day 7 and 14 when compared with that of adipogenic gene at day 0 and 3 (p < 0.05 and p < 0.01, respectively) (Figure 5A-D), however, RNA interfere of Zfp467 significantly enhanced the mRNA levels of specific osteogenic gene at day 7 and 14 when compared with that of osteogenic gene at day 0 and 3 (p < 0.05 and p < 0.01, respectively) (Figure 5E-F), which suggested Zfp467 involved in the ADSCs differentiated into adipocytes and osteoblasts.

We performed μCT analysis to determine the impact of ADSCs and Zfp467 on OVX-induced osteoporotic mice. Systemic transplantation of ADSCs with si-Zfp467 into OVX mice prevented OVX-induced bone loss in mice. When compared with OVX or ADSCs (with scrambled siRNA)-transplanted OVX mice, bone loss indices, including bone volume fraction, trabecular number, and bone mineral density in si-Zfp467-ADSCs transplanted OVX mice were restored to normal (the level of sham-operated mice) (Figure 6A-C).

The recovery effect of ADSCs with si-Zfp467 on bone loss was confirmed by measurement of the concentration of urinary deoxypyridinoline, a useful marker of bone resorption. Compared with OVX or ADSCs (with scrambled siRNA)-transplanted OVX mice, a significantly lower level of urinary deoxypyridinoline (DPD) was exhibited by si-Zfp467-ADSCs transplanted OVX mice (Figure 6D). Mineral apposition rate and bone formation rate, as measured by calcein label-based analysis, were restored to the level of sham-operated mice in si-Zfp467-ADSCs transplanted OVX mice. Of particular interest, histologic analysis showed that the numbers of osteoblasts adhering to trabecular bone surfaces were significantly increased in si-Zfp467-ADSCs transplanted OVX mice compared with OVX or ADSCs (with scrambled siRNA)-transplanted OVX mice (Figure 6E) (p < 0.01) and the numbers of TRAP-positive osteoclasts evidently decreased, which suggested that systemic transplantation of si-Zfp467-ADSCs in OVX mice might affect proliferation or differentiation of osteoblasts and osteoclasts. From these data, we confirmed that Zfp467 plays an important role in ADSCs to alleviate OVX-induced osteoporosis through regulating ADSCs differentiation to adipocyte and osteoblast.