Runx2 Overexpression Enhances Osteoblastic Differentiation and ...

Calcif Tissue Int (2006) 79:169 178 DOI: 10.1007/s00223-006-0083-6

Laboratory Investigations Runx2 Overexpression Enhances Osteoblastic Differentiation and Mineralization in Adipose - Derived Stem Cells in vitro and in vivo X. Zhang,1 M. Yang,3 L. Lin,1 P. Chen,2 K. T. Ma,2 C. Y. Zhou,2 Y. F. Ao1 1

Department of Sports Medicine, Peking University Third Hospital, Beijing 100083, China Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Peking University Health Science Center, Beijing 100083, China 3 Department of Orthopedics, Wannan Medical College Yijishan Hospital, Wuhu 241001, China 2

Received: 22 March 2006 / Accepted: 22 May 2006 / Online publication: 11 September 2006

Abstract. Like bone marrow stromal cells, adipose tissue-derived stem cells (ADSCs) possess multilineage potential, a capacity for self-renewal and long-term viability. To confirm whether ADSCs represent a promising source of cells for gene-enhanced bone tissueengineering, the osteogenic potential of ADSCs under the control of certain osteoinductive genes has been evaluated. Runx2, a transcription factor at the downstream end of bone morphogenetic protein (BMP) signaling pathways, is essential for osteoblast differentiation and bone formation. In this study we used adenovirus vector to deliver Runx2 to ADSCs and then examined the enhancement of osteogenic activity. Overexpression of Runx2 inhibited adipogenesis, as demonstrated by suppression of LPL and PPARc expression at the mRNA level and reduced lipid droplet formation. Moreover, ADSCs transduced with AdRunx2 underwent rapid and marked osteoblast differentiation as determined by osteoblastic gene expression, alkaline phosphatase activity and mineral deposition. Additionally, histological examination revealed that implantation of Runx2 modified ADSCs could induce mineral deposition and bone-like tissue formation in vivo. These results confirmed, firstly, the ability of Runx2 to promote osteogenesis and cell differentiation and, secondly, the competence of ADSCs as target cells for bone tissue engineering. Our work demonstrates a potential new approach for bone repair using Runx2modified ADSCs for bone tissue engineering. Key words: Adipose derived stem cells — Runx2 — Gene therapy — Differentiation — Osteogenesis

Currently, there are no completely effective treatments for massive segmental bone defects and non-healing fractures. Conventional autograft or allograft trans-

X. Zhang and M. Yang contributed equally to this work. Correspondence to: C. Y. Zhou; E-mail: chunyanzhou@bjmu. edu.cn or Y. F. Ao; E-mail: [email protected]

plantation is severely limited by the availability of donor tissue sources [1]. Therapy with recombinant bone morphogenetic proteins (rBMPs) appears promising [2 4]. However, without optimal matrices for controlled and sustained BMP delivery, the short biological halflife and immunogenicity of rBMPs limit the use of this approach [1]. With advances in cell and molecular biology, a new therapy termed Ôgene-enhanced bone tissue engineeringÕ that combines bone tissue engineering with gene therapy has emerged [5 7]. The basis of this novel technique involves target cells, osteo-inductive genes and specific scaffolds. This approach overcomes some deficiencies of previous methods by providing increased quantity of bone-like grafts and keeps delivering of regenerative molecules at reconstruction sites. Bone marrow stromal cells (BMSCs) have been considered an important target cell source for tissue engineering because of their multilineage potential, selfrenewal capacity and long-term viability. However, some problems limit their clinical application. These include a painful bone marrow biopsy procedure, the low frequency of BMSCs in healthy marrow and low cell numbers upon harvest [8]. Adipose tissue, like bone marrow, is derived from the embryonic mesoderm. Recently, it has been reported that stem cells can be isolated from adipose tissue. These are termed adipose derived stem cells (ADSCs). Like BMSCs, ADSCs have the potential to differentiate into multiple mesenchymal cell types such as adipocytes, chondrocytes, osteoblasts and myoblasts [9]. Moreover, this cell population can be isolated from adipose tissue in significant numbers and exhibits stable growth and high proliferation kinetics in culture [8]. They can overcome some of the disadvantages of BMSCs mentioned above and, therefore, may be a promising candidate cell source for gene-enhanced tissue engineering. It has been reported that the osteo-inductive bone morphogenetic proteins (BMPs) enhance the osteogenic

170

X. Zhang et al.: Runx2 Induced Osteoblastic Differentiation of Adipose Derived Stem Cells

potential of BMSCs in vitro and in vivo [10, 11]. However, the inaccurate targeting and ossification of neighboring nonosseous tissues and the need for cell surface receptors and signaling cofactors have limited the application of BMPs in bone tissue engineering. To avoid such limitation, we have focused on an alternative gene, Runx2/ Cbfa1, which encodes an osteogenic transcriptional activator at the downstream end of BMP signaling pathways. The regulation of osteogenesis by BMPs is complicated. As a downstream transcriptional activator, Runx2 has more direct osteogenic effects and plays an important role in osteoblast differentiation during embryonic development [12]. Humans with heterozygous mutations or deletions of Runx2/Cbfa1 developed cleidocranial dysplasia (CCD) [13]. Overexpression of Runx2 can induce and upregulate expression of multiple osteoblast specific genes in nonosteogenic cells [14]. Another advantage of Runx2 protein is that, unlike BMPs, it is not secreted. Therefore only targeted cells should be induced, avoiding the inaccurate targeting and ossification of neighboring cells that occurs with the use of BMPs. In order to test the hypothesis that Runx2 modified ADSCs might be a promising source for bone tissue engineering, this study examines whether the overexpression of Runx2 can enhance the osteogenic potential of ADSCs in terms of osteogenic differentiation and mineralization in vitro and in vivo.

Construction of recombinant adenovirus encoding Runx2/Cbfa1 The AdEasy system (provided by Dr. T. He, Howard Hughes Medical Institute, Baltimore, MD) was used to construct recombinant adenovirus. Briefly, the pBS KS plasmid containing full-length Runx2 cDNA (provided by Professor G. Karsenty, Department of Molecular and Human Genetics, Baylor College of Medicine, Huston, TX) was digested with Xho I and Xba I, resulting in a 3000 bp fragment containing Runx2 cDNA. The target fragment was inserted into the pAdTrack-CMV vector, which was linearized by digestion with Pme I, and subsequently cotransformed into E. coli. BJ5183 with pAdEasy-1. Recombinants were selected for kanamycin resistance and recombination was confirmed by digestion with Pac I and EcoR I. Finally, the linearized recombinant plasmid was transfected into 293A cells using lipofectamine (Invitrogen, Carlsbad, CA, USA) for adenovirus packaging. Recombinant adenovirus carrying the Runx2 gene (called Ad-Runx2) was propagated by reinfecting 293A cells and purified by CsCl gradient ultracentrifugation. Purified viral particles were stored in 10% glycerol/ PBS until use at )80C at a concentration of 1 · 1011 plaque-forming units (pfu)/ml. A control recombinant adenovirus carrying the enhanced green fluorescence protein gene, Ad-EGFP, was constructed using the same method (1 · 1010 pfu/ml). In vitro infection of ADSCs with Ad-Runx2 and Ad-EGFP Infection of ADSCs was performed when cells reached 85% confluence. The medium was removed and the cells were incubated with Ad-Runx2 or Ad-EGFP at multiplicity of infection (MOI) = 250 in serum-free DMEM at 37C for 6 h, with agitation every 15 min. An equal volume of growth medium was then added. Cells were recovered with complete medium 24 h later and the medium was changed every 2 days. Cells were collected at different time points for in vitro assays. For transplantation in vivo, cells were harvested 24 h posttransduction.

Materials and Methods Real-time RT-PCR Animals Four-week-old male SD rats weighing about 150g and fourweek-old male nude mouse were purchased from Beijing Animal Administration Center. All animal experimental protocols were approved by the Animal Care and Use Committee of Peking University and are in compliance with the ‘‘Guide for the Care and Use of Laboratory Animals’’ (National Academy Press, NIH Publication No. 85-23, revised 1996). Isolation of Adipose derived Stem Cells Adipose derived stem cells were harvested from four-weekold male SD rats. The rats were killed by cervical dislocation and adipose tissue in the inguinal groove was isolated and washed extensively with equal volumes of phosphate-buffered saline (PBS) to remove blood cells. The isolated adipose tissue was then digested with 0.1% collagenase type I (Sigma) with intermittent shaking at 37C for 30 min. Enzyme activity was terminated by dilution with DulbeccoÕs modified Eagle medium (DMEM), containing 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA). The floating adipocytes were separated from the stromal cell fraction by centrifugation (800rpm) for 5 min. The pellets were filtered through a 200 lm nylon mesh to remove cellular debris and incubated overnight at 37C with 5% CO2 in culture medium (DMEM, 10% FBS, and 100u/ml penicillin/streptomycin). The primary cells were cultured for 4 5 days until they reached confluence and were defined as passage ‘‘0’’. The cells were passaged at a ratio of 1: 3. The passages of cells used in experiments were between passage 3 and 10. Adipogenic differentiation was induced according to a method reported previously [9].

Total RNA was isolated at 1, 3 and 7 days posttransduction with Trizol reagent according to the manufacturerÕs instructions (Life Technologies, Gaithersburg, MD, USA) and was quantified by ultraviolet spectroscopy. cDNA synthesis was performed using DNase I treated (27 Kunitz units/sample) total RNA (1 lg) as a template by oligo(dT) priming using the Superscript First Strand Synthesis System for RT-PCR (Invitrogen). Real-time PCR was performed with an opticon continuous fluorescence detection system (MJ research, MA, USA). 1 ll of reverse transcribed product and 1 · SYBR green (Molecular Probes, Eugene, USA) were included in 25 ll reaction mixture (10 mM Tris-HCL, pH 8.3, 50 mM KCL, 1.5 mM MgCl2, 200 lM of dNTP mix, 0.2 lM of each primer and 1 unit of Taq DNA polymerase). Real-time PCR oligonucleotide primers (Table 1) were designed using Oligo 6 primer analysis software. Each cycle consisted of 1 min denaturation at 94C, 1 min annealing at 57C and 1 min extension at 72C. mRNA levels were normalized to GAPDH using the comparative cycle threshold (CT) method [15]. Indirect Immunofluorescence A total of 3 · 104 ADSCs were seeded on each glass coverslip, cultured for 24 h and then transfected with Ad-Runx2, AdEGFP or PBS. Forty eight hours after transfection, cells were fixed with 4% paraformaldehyde for 15 min, washed with 0.05% Triton X-100 in PBS, and blocked with 5% rabbit serum for 1 h. The cells were incubated with a polyclonal antibody against mouse Runx2 (Santa Cruz, CA, USA) for 1 h at a dilution of 1:50, followed by TRITC-conjugated rabbit antigoat IgG (Zhongshan Biochemical, Beijing, China) at a dilution of 1:200 for another 1 h. After rinsing three times with


171

Table 1. Primers used in reverse transcription ploymerase chain reactions (RT-PCR) Genesa

Primer sequences

Cbfa1/osf2 (mouse)

5¢-AGTCCCAACTTCCTGTGCT-3¢ 5¢-GTGTCATCATCTGAAATACGC-3¢ 5¢-AACGGTGGTGCCATAGATGC-3¢ 5¢-AGGACCCTCTCTCTGCTCAC-3¢ 5¢-GAAACTCCTGGACTTTGACC-3¢ 5¢-GCCACTTGGCTGAAGCCTG-3¢ 5¢-CTGACCCTCGTAGCCTTCATAG-3¢ 5¢-CGCCTACTTTTATCCTCCTCTG-3¢ 5¢-GGAGAGAGTGCCAACTCCAG-3¢ 5¢-CCACCCCAGGGATAAAAACT-3¢ 5¢-GAGATTTCTCTGTATGGCACA-3¢ 5¢-CTGCAGATGAGAAACTTTCTC-3¢ 5¢-TGGAGCCTAAGTTTGAGTTTGC-3¢ 5¢-TGACAATCTGCCTGAGGTCTG-3¢ 5¢-GAAAAGCTGTGGCGTGATGG-3¢ 5¢-GTAGGCCATGAGGTCCACCA-3¢

Osteocalcin (OCN) Osteopontin (OPN) Bone sialoprotein (BSP) Collagen I (COL I) Lipoprotein lipase (LPL) PPAcb GAPDH a b

Primers designed from rat sequences unless otherwise indicated PPARc: peroxisome proliferator activated receptor c

PBS (5 min each time), the cells were visualized with a fluorescence confocal microscope. Images of stained cells were captured using an Olympus IX70 microscope and SPOTRT image acquisition software. Western Blot Detection of Runx2 Protein ADSCs were harvested 3 days posttransduction. Cells were lysed in lysis buffer (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 150 mM Tris HCl pH 7.2, 350 lg/ml phenylmethylsulfonyl fluoride, 10 lg/ml leupeptin, 10 lg/ml aprotinin, and 1 mM sodium orthovanadate) at 4C for 30 minutes and then centrifuged at 12,000 g for 10 min. Protein concentration was measured with the BCA protein assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA) using bovine serum albumin as a standard. Samples were fractionated on 12% SDS-PAGE gels and transferred to a nitrocellulose membrane (Schleicher & Schuel, Keene, NH, USA). Goat polyclonal antibody against Runx2 (Santa Cruz) was used at a dilution of 1:500. The rat anti-goat IgG-horseradish peroxidase secondary antibody (Zhongshan) was used at a dilution of 1:1000. All the reactions were conducted as previously described [16]. Immunoreactivity was determined using the ECL chemiluminescence reaction kit (Amersham, Arlington Heights, IL, USA)

min, rinsed with distilled water and mounted for inspection. Quantification of mineralization in 2-D cultures was performed by measuring the average photodensity of images of von Kossa-stained ADSCs and von Kossa-stained sections with a pathology image analysis system (Mike Audi Image Analysis Inc. China). Examination of osteogenic activity of Ad-Runx2-transduced ADSCs in vivo After 24 h culture, ADSCs transduced with Ad-EGFP or AdRunx2, or untranduced ADSCs, were harvested by centrifugation (800 rpm) for 5 min, resuspended in fresh medium at a density of 2 · 106 cells/ml and then seeded in polylactic acid (PLA) scaffolds (5 · 5 · 5 mm3, High Polymer inc. Shan Dong China). The scaffolds were soaked in the cell suspension and incubated for 4 hours before fresh medium was then added. After 24 h, implantation was performed on the backs of 4week-old nude mice with Ad-Runx2-ADSCs-PLA, Ad-EGFPADSCs-PLA or untransduced ADSCs-PLA. After surgery, all mice were allowed to move freely in their cages. Histological Analysis

Alkaline phosphatase (ALP) activity was measured using the ALP assay kit (Zhongsheng Biochemical, Beijing, China). Briefly, ADSCs were seeded at 2 · 105 cells/well in 6-well plates (three wells for each group). Twenty-four hours later, cells were exposed to Ad-Runx2 or Ad-EGFP at a MOI of 250, or to PBS. On days 3, 7, 10 and 14, cells were harvested and dissociated in 0.1 M Tris (pH 7.4) containing 1% Triton X-100 and 5 mM MgCl2 by sonication. ALP activity was normalized against the protein concentration (measured as described above) and expressed as U/g/min.

Four weeks after implantation, the cell/PLA complexes were removed from the mice. Samples were fixed with 4% paraformaldehyde, dehydrated through a series of graded ethanols and xylol, and embedded in paraffin. Serial sections (5 lm) were stained with toluidine blue and von Kossa. In brief, paraffin sections were deparaffinized in xylene, hydrated in a series of graded ethanols, and pretreated with 3% acetic acid for 3 min. Sections were then stained with 1% toluidine blue at pH 2.5 for 30 min and thoroughly rinsed with tap water. Morphometric analysis of images in histological sections was carried out with an Olympus IX-70 microscope (Tokyo, Japan) equipped with a cooled CCD camera (Cool SNAP HQ, ROPER Scientific, Tucson, AZ).

Matrix Mineralization Analysis

Statistical Analysis

Von Kossa silver staining was used to measure extracellular matrix mineralization 7, 10, 14 and 21 days after infection. Cultured cells were fixed in 70% ethanol, stained with 5% AgNO3 for 30 min under sunlight, fixed with 5% Na2SO3 for 4

All experiments were repeated a minimum of three times. The paired t test was used for statistical evaluation. Differences between the experimental and control groups were regarded as statistically significant when P < 0.01.

Alkaline Phosphatase Activity

172


Fig. 1. Morphology of adipose tissue-derived stem cells (ADSCs). A, B: Passage 3 ADSCs show fibroblast-like morphology (A: original magnification ·100; B: original magnification ·400). C, D: Oil Red stain shows lipid droplets 1 week after induction of adipogenic differentiation (C: original magnification ·100; D: original magnification ·400).

Results Characteristics of ADSCs

Approximately 2 · 106 ADSCs were harvested from the inguinal groove adipose tissue of each 4-week-old rat. Some cells underwent significant spontaneous adipogenesis in primary culture and were removed with passage. The cells in the third passage were almost all fibroblast-like cells devoid of lipid (Figs. 1A, B). The doubling time of the cells was 2 days, reaching saturation in 4.5 days. Intracellular lipid droplets were observed in ADSCs after adipogenic induction by a standard method, which proved that ADSCs possessed multi-lineage differentiation potential (Figs. 1C, D). Ad-Runx2 Transduced Primary ADSCs

Recombinant adenovirus-transduced ADSCs were analyzed for EGFP expression by fluorescence microscopy and flow cytometry. High levels of EGFP were detected 48h posttransduction using fluorescence microscopy (Fig. 2A). Transduction efficiency of Ad-Runx2 in ADSCs was 66.64% as indicated by flow cytometric detection of an EGFP marker 48 h posttransduction. Nontransduced cells were used as a control (Fig. 2B). Gene Expression in ADSCs Transduced with Ad-Runx2

The expression of Runx2 mRNA in Ad-Runx2 transduced cells was assessed by real-time RT-PCR. Runx2

mRNA was detected in Runx2-tranduced ADSCs at 1, 3 and 7 days posttransduction, but not in Ad-EGFP transduced ADSCs (Fig. 2C). Western blot analysis revealed expression of Runx2 protein (a 67KD specific band) in Ad-Runx2 transduced ADSCs, which was undetectable in untransduced cells and Ad-EGFP transduced ADSCs (Fig. 2D). Immunohistochemical staining showed nuclear expression of Runx2 protein in Ad-Runx2 transduced ADSCs (Fig. 2E). In the untransduced cells or cells transduced with Ad-EGFP, no Runx2 was detected. Induction of Osteoblast Differentiation by Transduction of ADSCs with Ad-Runx2

Expression of osteogenic genes was assessed at 1, 3 and 7 days posttransduction by real-time RT-PCR. The genes assessed included osteocalcin (OCN), osteopontin (OPN), collagen I (COLI) and bone sialoprotein (BSP). OCN is an extracellular matrix protein and a marker of mature osteoblasts and BSP is an osteoblastic matrix molecule responsible for mineralized nodule nucleation. Up-regulated mRNA expression of all four osteogenic genes was observed in Ad-Runx2 transduced ADSCs and gradually increased with time (Figs. 3 A D). There was no difference between Ad-EGFP transduced ADSCs and nontransduced ADSCs. These results suggested that osteoblast differentiation was triggered in AdRunx2 transduced ADSCs.


173

Fig. 2. A, B: EGFP expression in Ad-Runx2 transduced ADSCs 48h posttransduction. ADSCs tranduced with Ad-Runx2 were examined by light microscopy (A-a) and fluorescence microscopy (A-b). Ad-Runx2 transduction efficiency was measured at 66.64% by flow cytometric analysis 48h posttransduction (B-b) using non-transduced ADSCs as controls for autofluorescence (B-a). C. D. E: Expression of Runx2 in ADSCs. (C) Determination of Runx2 mRNA expression by real-time RT-PCR 1, 3 and 7 days posttransduction. There is a significant difference between the Ad-Runx2 group and the control group (Ad-EGFP transduced cells) in Runx2 expression (mean + SEM, n = 3, * P < 0.01 vs. the control group). (D) Analysis of Runx2 protein levels by Western blotting 48h posttransduction. Lane 1, nontransduced ADSCs; Lane 2, Ad-EGFP transduced ADSCs; Lane 3 ADSCs tansduced with Ad-Runx2 and cultured for 3 days. A specific 67 KD Runx2 protein band is presents in Runx2-transduced ADSCs but not in control groups. (E) Immunofluorescence staining for Runx2 48h posttransduction. EGFP (green) is expressed in the nucleus and cytoplasm (E-a). Runx2 expression (red) is confined to the nucleus in transduced cells (E-b). A merged image of figures E-a and E-b (E-c). Untransduced cells were used as a control (E-d).

174


Fig. 3. Up-regulation of the expression of specific osteogenic genes evaluated by realtime RT-PCR at 1, 3 and 7 days posttransduction. Osteocalcin (OCN) (A), Osteopontin (OPN) (B), Collagen I (COL I) (C) and bone sialoprotein (BSP) (D) were up-regulated in Ad-Runx2 transduced cells (mean + SEM, n = 3, * P < 0.01 vs. AdEGFP control group).

Forced Expression of Runx2 Inhibits Adipogenesis

The expressions of peroxisome proliferator activated receptor c (PPARc) and lipoprotein lipase (LPL) were analyzed at 1, 3 and 7 days posttransduction by realtime RT-PCR. PPARc, a fat-specific transcription factor, plays a role in preadipocyte commitment. LPL, a lipid exchange enzyme, is upregulated during adipogenesis. The expression of PPARc and LPL was initially detected in both Ad-Runx2 transduced ADSCs and nontransduced controls. However, their expression levels decreased dramatically in Ad-Runx2 transduced cells compared with nontransduced control cells (Figs. 4 A, B). At day 7 posttransduction, a significant down-regulation in expression was measured in ADSCs transduced with Ad-Runx2. Fourteen days posttransduction, lipid droplets were observed in Ad-EGFP transduced ADSCs by Oil Red O staining (Fig. 4 C-a), and no lipid droplets were visible in Ad-Runx2-treated group (Fig. 4 C-b). At the same time, osteogenic was analyzed by alkaline phosphatase staining at fourteen days posttransduction. Ad-EGFP transduced ADSCs showed no stained regions (Fig. 4 C-c), while Runx2 treatment produced high level of alkaline phosphatase in cytoplasm and developed heavy stains (Fig. 4C-d). ALP assay and examination of mineralization

The alkaline phosphatase (ALP) activity of ADSCs transduced with Runx2 markedly increased at day 7 with a peak at day 10 posttransduction, and remained at a high level than the control group. Intrinsic cellular ALP activity in untransduced and Ad-EGFP transduced cells remained unchanged (Fig. 5A).

Von Kossa staining for phosphate deposits was used to detect matrix mineralization of ADSCs at 7, 10, 14 and 21 days posttransduction. Ad-EGFP transduced cells exhibited no mineral deposition, while Runx2transduced cells displayed strong positive staining for mineralization (Fig. 5B-a). Densitometric analysis of von Kossa-staining was used to quantify the mineralization in the three groups at each time point. Beginning at day 10, a significant increase in the level of mineralization in Runx2-transduced ADSCs was observed. In contrast, control groups did not mineralize until day 21 (Fig. 5B-b). In vivo Osteogenic Activity of Runx2 Transduced Cells

A final series of experiments was performed to examine the osteogenic activity of Ad-Runx2-transduced ADSCs in vivo. Significant amounts of mineral deposition were observed with toluidine blue and Von Kossa staining in PLA/Ad-Runx2- ADSCs complex (Figs. 6A e, f). There was little or no mineral formation in the PLA/Ad-EGFP-ADSCs (Figs. 6A c, d) or PLA /ADSCs groups (Figs. 6A a, b). The results of gradation analysis showed that the mineral deposits of Runx2-ADSCs/PLA composites were much higher than that in the control groups. This observation suggests that Runx2 overexpression dramatically facilitates osteogenic differentiation compared with the control groups in a challenging, nonosteogenic environment such as the subcutaneous space. Densitometric analysis of von Kossa-staining was used to quantify the mineralization in three groups. The results showed a significant difference between Runx2treated group and the control groups (Fig. 6B).


175

Fig. 4. Inhibition of adipogenesis by Runx2. Expressions of the specific adipogenic markers LPL and PPARc were analyzed 1, 3 and 7 days posttransduction by real-time RT-PCR. LPL and PPARc mRNA levels gradually decreased in AdRunx2 treated cells (mean + SEM, n = 3, * P < 0.01, vs. the control group) (A, B). Fourteen days posttransduction, lipid droplets were observed in Ad-EGFP transduced ADSCs by Oil Red O staining (C-a), and no lipid droplets were visible in Ad-Runx2-treated group (C-b). At the same time, osteogenic was analyzed by alkaline phosphatase staining at fourteen days posttransduction. Ad-EGFP transduced ADSCs showed no stained regions (C-c) while Runx2 treatment produced high level of alkaline phosphatase in cytoplasm, which developed heavy stains (C-d). Discussion

Transdifferentiation has been defined as ‘‘the conversion of a cell of one tissue lineage into a cell of an entirely distinct lineage, with concomitant loss of the tissuespecific markers and function of the original cell type, and acquisition of markers and function of the transdifferentiated cell type’’ [17]. In the present study, the expression of osteoblast specific genes such as OCN, OPN, COL1 and BSP was strongly upregulated in adipose tissue derived primary stromal cells transduced with exogenous Runx2, coincident with the repression of the specific adipocyte and preadipocyte markers LPL and PPARc and restrained lipid droplet formation. Furthermore, increased ALP activity and matrix mineralization in Ad-Runx2 transduced ADSCs indicated that Runx2 might induce functional osteogenesis. The in vivo results also revealed osteoblastic differentiation and mineralization in an ectopic, nonosseous subcutaneous

site. Uninduced ADSCs and PLA scaffold could not ossify in the nonosteogenic environment. These results indicate that the transduction of ADSCs cells with a recombinant adenovirus encoding Runx2 stimulated osteoblastic differentiation and mineralization in vitro and in vivo. The Adipose derived Stem Cells Represent a Promising New Source for Bone Tissue-engineering.

Several factors are critical for a potential cell source to be successfully used in tissue-engineering approaches, including amenable isolation and expansion of cells, and their capacity for osteoblastic differentiation and matrix mineralization, etc. The ideal sources of cells suitable for clinical therapy should be those that can be readily harvested, rapidly expandable in sufficient number, and have a strong potential for differentiation. Several candidate cell sources have been reported, each with

176


with a logarithmic growth curve. ADSCs could be amplified to sufficient numbers by short-term culture for implantation in vivo and are capable of differentiating into osteoblasts, adipocytes, muscle cells and neural cells when treated with established lineage-specific factors [7, 9, 18, 19]. Our results demonstrate that ADSCs also have great plasticity and powerful osteogenic potential when transduced with Runx2. These data indicate that ADSCs represent a promising new cell source for gene enhanced bone tissue-engineering. Adenoviral Vectors Would be a Suitable Delivery System for Runx2 Gene Therapy

Fig. 5. (A) Analysis of ALP activity in the lysates of ADSCs transduced with Ad-Runx2 or Ad-EGFP. ALP activity in Runx2 transduced cells could be detected at day 3 and continued to increase with a peak at 10 days. It then gradually declined but remained at a high level for the duration of the experiment. The intrinsic cellular ALP activity in Ad-EGFP transduced cells remained unchanged (mean + SEM, n = 3, *P < 0.01 vs. Ad-EGFP transduced cells). (B-a) Von Kossa staining of calcified extracellular matrix at days 7, 10, 14 and 21 post-transduction. ADSCs were transduced with AdRunx2, Ad-EGFP, or left alone. Mineralized bone nodules (black) and diffuse mineralization (light brown) can be seen in the Ad-Runx2 group, but are not seen in the other groups (·100). (B-b) Densitometric analysis of von Kossa-staining with pathology image analysis system (mean + SEM, n = 3, *P < 0.01 vs. nontranduced cells, # P < 0.01 vs. Ad-EGFP transduced cells).

advantages and disadvantages. In our study, ADSCs were isolated easily from the inguinal groove adipose tissue of 4-week-old rats at about 2 · 106 cells per animal. The proliferation kinetics of ADSCs is consistent

For clinical application of cell-based gene therapy, the target cells need to be efficiently transduced with a vector and able to express the desired protein. Adenoviral vectors can mediate gene transfer to both replicating and non-replicating cells. In the present study, adenoviral vectors could be produced in high titres. We also showed that ADSCs could be efficiently infected by the adenol viral vector in vitro. In moderate doses (MOI = 250), the transduction efficiency was 66.86% as detected by flow cytometry. This was higher than the low doses tested, and higher doses did not increase transduction efficiency and instead induced apoptosis (data not shown). The adenovirus transduction efficiency in ADSCs is much higher than that previously reported in BMSCs (about 30 40%) [20]. Another concern of gene therapy is the control of gene expression. It has been reported that prolonged overexpression of Runx2 may generate adverse effects. Geoffroy has reported that transgenic mice selectively overexpressing Runx2 in osteoblasts experienced severe osteopenia caused by excessive osteoclastogenesis [21]. Unlike retroviruses and adeno-associated viruses, adenoviruses have an extrachromosomal lifecycle without the risk of insertional mutagenesis. Additionally, they will not sustain prolonged expression of Runx2 [22, 23]. Therefore we chose the adenoviruses delivery system to restrict Runx2 expression to the narrow time periods necessary for osteogenesis. Runx2 Would be an Ideal Candidate for Osteoblastic Differentiation in ADSCs

Runx2, a transcription factor, is essential for osteoblastic differentiation and the formation and maintenance of the bone [20, 21, 24]. Many osteoblast specific genes such as COLI and OCN (early and late osteoblastic markers) contain consensus Runx2-binding elements in their promoter regions which can be positively regulated by Runx2 [25, 26]. As the downstream effector of BMPs in certain signalling pathways, Runx2 mediates mainly positive effects on osteogenesis, which would subsequently activate


177

Fig. 6. Analysis of in vivo osteogenic differentiation. (A) Representative histological sections: group 1, nontransduced ADSCs/PLA (a, b); group 2, ADSCs transduced with Ad-EGFP/PLA (c, d); group 3, ADSCs transduced with Ad-Runx2/ PLA (e, f). Toluidine blue (a, c, e,) was performed to examine cartilage or bone formation in the scaffolds. Von Kossa staining (b, d, f,) was performed to identify mineralized bone. Significant amounts of cartilage-like or bone-like matrix formation and mineralization were observed in Adv-Runx2/ADSCs PLA (e, f). There was little or no mineralization in nontransduced ADSCs/PLA (a, b) and PLA/AdvGFP- transduced ADSCs (c, d). (B) Analysis of average photodensity of von Kossa-stained sections (mean + SEM, n = 3, * P < 0.01 vs. nontranduced cells, # P < 0.01 vs. Ad-EGFP transduced cells).

downstream genes necessary for the osteoblast phenotype [14]. BMPs have concomitant positive and negative effects on osteogenesis, resulting in a less potent signal, whereas Runx2 may serve as a more specific osteogenic stimulator than BMPs. The other advantage of using Runx2 is that, unlike BMPs, it is not secreted and therefore should not affect neighboring host cells; hence it will be safer for clinical application [27]. Although the precise mechanism by which overexpression of Runx2 restrains the expression of the specific adipocyte and preadipocyte markers PPARc and LPL is not clear, the observations in our experiments suggest that overexpression of Runx2 alters the dynamic equilibrium between osteogenesis and adipogenesis; this may be of importance in the treatment of osteoporosis. Some issues need to be addressed before this approach can be applied clinically. Firstly, the proportion of stem cells in adipose tissue derived primary stromal cells remains unknown. Unpurified primary ADSCs

may affect the differentiation potential during ex vivo experiments. In future research, it will be important to develop a convenient and harmless method of purifying stem cells from primary adipose-isolated cells. Secondly, in the in vivo experiments, we did not observe formation of marrow. This might be due to the relatively short duration of our observations and the use of polylactic acid (PLA) scaffolds. In previous work, marrow was found 14 weeks after implantation of BMSCs in gelatin sponges [28, 29]. In summary, we have shown that Runx2 induced osteoblastic differentiation and mineralization in adipose tissue-derived primary stromal cells in vitro and in vivo. This work suggests that Runx2 gene transfer selectively redirected the differentiation of multipotent ADSCs and enhanced their osteogenic potential both in vitro and in vivo, supporting the premise that ADSCs are a promising source of adult autologous stem cells for Runx2/Cbfa1 gene therapy and bone tissue engineering.

178


Acknowledgments. This work was supported by 985 Project Foundation of Peking University and 211 Project of the Ministry of Education, China. We thank Prof. Karsenty G (Department of Molecular and Human Genetics, Baylor College of Medicine, Huston, TX) for providing cDNA of mouse cbfa1/osf2, Dr. He TC (Howard Hughes Medical Institute, Baltimore, MD) for providing the AdEasy System, Dr. Wong J (University of Cambridge, UK) and Prof. Liu S (Peking University, China) for help in revising this manuscript. References 1. Franceschi RT, Yang S, Rutherford RB, Krebsbach PH, Zhao M, Wang D (2004) Gene therapy approaches for bone regeneration. Cells Tissues Organs 176:95 108 2. Einhorn TA, Majeska RJ, Mohaideen A, Kagel EM, Bouxsein ML, Turek TJ, Wozney JM (2003) A single percutaneous injection of recombinant human bone morphogenetic protein-2 accelerates fracture repair. J Bone Joint Surg Am 85-A(8):1425 1435 3. Seeherman HJ, Bouxsein M, Kim H, Li R, Li XJ, Aiolova M, Wozney JM (2004) Recombinant human bone morphogenetic protein-2 delivered in an injectable calcium phosphate paste accelerates osteotomy-site healing in a nonhuman primate model. J Bone Joint Surg Am 86A(9):1961 1972 4. Starr AJ (2003) Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures. J Bone Joint Surg Am 85-A(10):2049 2050 5. Breitbart AS, Grande DA, Mason JM, Barcia M, James T, Grant RT (1999) Gene-enhanced tissue engineering: applications for bone healing using cultured periosteal cells transduced retrovirally with the BMP-7 gene. Ann Plast Surg 42(5):488 495 6. Mason JM, Breitbart AS, Barcia M, Porti D, Pergolizzi RG, Grande DA (2000) Cartilage and bone regeneration using gene-enhanced tissue engineering. Clin Orthop Relat Res 379(Suppl):S171 S178 7. Edwards PC, Ruggiero S, Fantasia J, Burakoff R, Moorji SM, Paric E, Razzano P, Grande DA, Mason JM (2005) Sonic hedgehog gene-enhanced tissue engineering for bone regeneration. Gene Ther 12(1):75 86 8. Zuk PA, Zhu M, Mizuno H, Huang J, Futrell JW, Katz AJ, Benhaim P, Lorenz HP, Hedrick MH (2001) Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 7(2):211 228 9. Zuk PA, Zhu M, Ashjian P, De Ugarte DA, Huang JI, Mizuno H, Alfonso ZC, Fraser JK, Benhaim P, Hedrick MH (2002) Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 13(12):4279 4295 10. Blum JS, Barry MA, Mikos AG, Jansen JA (2003) In vivo evaluation of gene therapy vectors in ex vivo-derived marrow stromal cells for bone regeneration in a rat critical-size calvarial defect model. Hum Gene Ther 14(18):1689 1701 11. Cheng SL, Lou J, Wright NM, Lai CF, Avioli LV, Riew KD (2001) In vitro and in vivo induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene. Calcif Tissue Int 68(2): 87 94 12. Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, Geoffroy V, Amling M, Karsenty G (1999) A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev 13(8): 1025 1036 13. Lee B, Thirunavukkarasu K, Zhou L, Pastore L, Baldini A, Hecht J, Geoffroy V, Ducy P, Karsenty G (1997) Missense mutations abolishing DNA binding of the

14.

15.

16.

17. 18.

19.

20.

21.

22. 23. 24.

25.

26. 27.

28.

29.

osteoblast-specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nat Genet 16(3):307 310 Gersbach CA, Byers BA, Pavlath GK, Garcia AJ (2004) Runx2/Cbfa1 stimulates transdifferentiation of primary skeletal myoblasts into a mineralizing osteoblastic phenotype. Exp Cell Res 300(2):406 417 Byers BA, Pavlath GK, Murphy TJ, Karsenty G, Garcia AJ (2002) Cell-type-dependent up-regulation of in vitro mineralization after overexpression of the osteoblast-specific transcription factor Runx2/Cbfal. J Bone Miner Res 17(11):1931 1944 Winnard RG, Gerstenfeld LC, Toma CD, Franceschi RT (1995) Fibronectin gene expression, synthesis and accumulation during in vitro differentiation of chicken osteoblasts. J Bone Miner Res 10(12):1969 1977 Wagers AJ, Weissman IL (2004) Plasticity of adult stem cells. Cell. 116(5):639 648 Bruder SP, Jaiswal N, Ricalton NS, Mosca JD, Kraus KH, Kadiyala S (1998) Mesenchymal stem cells in osteobiology and applied bone regeneration. Clin Orthop Relat Res 355(Suppl):247 256 Kang SK, Lee DH, Bae YC, Kim HK, Baik SY, Jung JS (2003) Improvement of neurological deficits by intracerebral transplantation of human adipose tissue-derived stromal cells after cerebral ischemia in rats. Exp Neurol 183(2):355 366 Conget PA, Minguell JJ (2000) Adenoviral-mediated gene transfer into ex vivo expanded human bone marrow mesenchymal progenitor cells. Exp Hematol 28(4): 382 390 Geoffroy V, Kneissel M, Fournier B, Boyde A, Matthias P (2002) High bone resorption in adult aging transgenic mice overexpressing cbfa1/runx2 in cells of the osteoblastic lineage. Mol Cell Biol 22(17):6222 6233 Hidaka C, Khan SN, Farmer JC, Sandhu HS (2002) Gene therapy for spinal applications. Orthop Clin North Am 33(2):439 446 Hannallah D, Peterson B, Lieberman JR, Huard Fu (2003) Gene therapy in orthopaedic surgery. Instr Course Lect 52:753 768 Yang S, Wei D, Wang D, Phimphilai M, Krebsbach PH, Franceschi RT (2003) In vitro and in vivo synergistic interactions between the Runx2/Cbfa1 transcription factor and bone morphogenetic protein-2 in stimulating osteoblast differentiation. J Bone Miner Res 18(4): 705 715 Frendo JL, Xiao G, Fuchs S, Franceschi RT, Karsenty G, Ducy P (1998) Functional hierarchy between two OSE2 elements in the control of osteocalcin gene expression in vivo. J Biol Chem 273(46):30509 30516 Kern B, Shen J, Starbuck M, Karsenty G (2001) Cbfa1 contributes to the osteoblast-specific expression of type I collagen genes. J Biol Chem 276(10):7101 7107 Gazit D, Turgeman G, Kelley P, Wang E, Jalenak M, Zilberman Y, Moutsatsos I (1999) Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone: a novel cell-mediated gene therapy. J Gene Med 1(2):121 133 Krebsbach PH, Kuznetsov SA, Satomura K, Emmons RV, Rowe DW, Robey PG (1997) Bone formation in vivo: comparison of osteogenesis by transplanted mouse and human marrow stromal fibroblasts. Transplantation 63(8):1059 1069 Jaiswal N, Haynesworth SE, Caplan AI, Bruder SP (1997) Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. Cell Biochem 64(2):295 312