JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE RESEARCH J Tissue Eng Regen Med (2014) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/term.1924

ARTICLE

hTERT- and hCTLA4Ig-expressing human bone marrow-derived mesenchymal stem cells: in vitro and in vivo characterization and osteogenic differentiation Fei Dai1†, Sisi Yang2†, Fei Zhang1, Dongwen Shi2, Zehua Zhang1, Jun Wu2* and Jianzhong Xu1* 1

National and Regional United Engineering Laboratory of Tissue Engineering, Department of Orthopaedics, Southwest Hospital, Third Military Medical University, Chongqing, People’s Republic of China 2 Institute of Burns Research, Southwest Hospital, Third Military Medical University, Chongqing, People’s Republic of China

Abstract Multipotent mesenchymal stem cells (MSCs) are commonly used as seed cells in studies of tissue engineering and regenerative medicine but their clinical application is limited, due to insufficient numbers of autogeneic MSCs, immune rejection of allogeneic MSCs and replicative senescence. We constructed two gene expression vectors for transfection of the human telomerase reverse transcriptase (hTERT) and cytotoxic T lymphocyte-associated antigen 4-Ig (CTLA4Ig) genes into human bone marrow-derived stem cells (hBMSCs). Successful transfection of both genes generated hTERT–CTLA4Ig hBMSCs that expressed both telomerase (shown by immunohistochemistry and a TRAPeze assay) and CTLA4Ig (demonstrated by immunocytochemistry and western blotting) without apparent mutual interference. Both hTERT BMSCs (92 population doublings) and hTERT–CTLA4Ig hBMSCs (60 population doublings) had an extended lifespan compared with hBMSCs (18 population doublings). Cell cycle analysis revealed that, compared with hBMSCs, a lower proportion of hTERT hBMSCs were in G0/G1 phase but a higher proportion were in S phase; compared with hTERT hBMSCs, a higher proportion of hTERT–CTLA4Ig hBMSCs were in G0/G1 phase, while a lower proportion were in S and G2/M phases. hTERT–CTLA4Ig hBMSCs retained their capacity for osteogenic differentiation in vitro, shown by the detection of hydroxyapatite mineral deposition (labelled tetracycline fluorescence staining), calcareous nodules (alizarin red S staining), alkaline phosphatase (calcium–cobalt method) and osteocalcin (immunocytochemistry). Furthermore, subcutaneous transplantation of hTERT–CTLA4Ig hBMSCs in a rat xenotransplantation model resulted in the successful generation of bone-like tissue, confirmed using radiography and histological assessment. We propose that allogeneic hTERT–CTLA4Ig hBMSCs may be ideal seed cells for bone tissue engineering. Copyright © 2014 John Wiley & Sons, Ltd. Received 29 July 2013; Revised 21 February 2014; Accepted 25 April 2014

Keywords CTLA4Ig; hTERT; bone tissue engineering; human bone marrow-derived mesenchymal stem cell (hBMSC)

*Correspondence to: Jianzhong Xu, National and Regional United Engineering Laboratory of Tissue Engineering, Department of Orthopaedics, Southwest Hospital, Third Military Medical University, Chongqing 400038, People’s Republic of China. E-mail: [email protected]; [email protected] Jun Wu, Institute of Burn Research Southwest Hospital, The Third Military Medical University, Chongqing 400038, People’s Republic of China. Email: [email protected] † These authors contributed equally to this study. Copyright © 2014 John Wiley & Sons, Ltd.

1. Introduction Large bone defects are relatively common and often associated with significant functional and aesthetic problems. The causes of such defects are numerous and include infection, trauma, tumours and congenital malformation. Surgical repair of large bone defects is highly challenging,

F. Dai et al.

hence novel approaches to treatment are needed to improve the quality of life of patients with these defects. Bone tissue engineering is well established as a promising method for repairing bone defects prior to the use of other treatments, such as autogeneic or allogeneic bone transplantation or the use of prosthetic replacements (Horwitz et al., 2002; Petite et al., 2000; Yoon et al., 2007). There are three main elements to tissue engineering: seed cells, and scaffold- and tissue-construction techniques (Wang et al., 2010; Yilgor et al., 2010). Mesenchymal stromal cells (MSCs) represent a cell population that displays adherent properties, is able to self-renew and has multilineage differentiation capabilities; for example, MSCs have the potential to differentiate into adipocytes, chondroblasts and osteoblasts in vitro (Dominici et al., 2006). Thus, MSCs have become a major focus of research into bone tissue engineering. MSCs have been used as seed cells to repair bone defects, due to their ability to proliferate and differentiate (Løken et al., 2008). However, several limitations remain; a loss of stem-like properties and spontaneous differentiation have been encountered during in vitro expansion of MSCs (Woodbury et al., 2002), and rapid expansion of autogeneic MSCs in a short period of time is currently not possible. Furthermore, these cells have been characterized as showing diminished replication, altered functionality (Beausejour, 2007), a reduced potential for differentiation (Bonab et al., 2006) and the emergence of lineage-specific markers; this has compromised the use of hMSCs in clinical applications. Moreover, allogeneic MSCs are not totally immunoprivileged, and thus could gradually elicit immune rejection during their differentiation because of allelic differences between the graft and host at polymorphic loci (Dai et al., 2006). As bone tissue engineering has developed, the therapeutic efficacy of using hBMSCs as seed cells has been the focus of much investigation. Considerable attention has been paid to the problems of immune rejection of allogeneic hBMSCs, insufficient numbers of autogeneic hBMSCs and replicative cell senescence. Numerous bone marrow aspirations are required to obtain a sufficient number of hBMSCs for clinical use, which inevitably increases the pain and discomfort experienced by the patient. To solve this problem, various attempts have been made to prolong the lifespan of the cell, including electroporation (Jonak et al., 1992), Epstein–Barr virus transformation (Sugden and Mark, 1977), Simian virus 40 (SV40)-mediated transformation (Kim et al., 1997) and integration of human papillomavirus (HPV) sequences (Woodworth et al., 1989). Since the telomeric DNA at the end of each chromosome shortens at each cell division, eventually resulting in a cessation of replication, one additional method to increase the cell lifespan that has received considerable attention is the overexpression of the human telomerase reverse transcriptase (hTERT) gene (Tao et al., 2009), which encodes the catalytic subunit of human telomerase that maintains the length of the telomere (Kassem et al., 2004). It has been reported that human fibroblasts (Morales et al., 1999), retinal Copyright © 2014 John Wiley & Sons, Ltd.

pigment epithelial cells (Bodnar et al., 1998) and endothelial cells (Yang et al., 2001) transfected with hTERT were able to survive successfully over long periods when cultured in vitro. Several studies have demonstrated that modification of human MSCs (hMSCs) with the hTERT gene generates cells with an improved ability for proliferation and cell renewal that retain their potential to differentiate into osteocytes, adipocytes and chondrocytes (Bischoff et al., 2012; Huang et al., 2008; Kobune et al., 2003; Piper et al., 2012; Yang et al., 2007). hTERT-modified hMSCs have been successfully cultured in porous polylactic glycolic acid scaffolds under perfusion culture (Yang et al., 2010), and in a crystalline scaffold of hydroxyapatite– tricalcium phosphate, resulting in three-dimensional (3D) ‘osteospheroids’ that shared certain characteristics with in vivo bone formation (Burns et al., 2010). Furthermore, when osteogenically differentiated and transplanted into mice, hTERT-modified hMSCs were able to correct a bone defect with no evident adverse events (Nakahara et al., 2009). Thus, hTERT-modified cells are considered to have great potential for use in bone tissue engineering. In a previous study, we reported that transferring the gene for the human cytotoxic T lymphocyte-associated antigen 4-Ig (CTLA4Ig) into allogeneic human bone marrow-derived MSCs (hBMSCs) was successful in alleviating the problems of immune rejection. These modified cells were able to differentiate normally into osteoblasts in vitro, elicit a reduced immune response in a mixed lymphocyte culture model and survive in vivo until bone tissue formed (Dai et al., 2006). Since modified hBMSCs show promise as seed cells for bone tissue engineering, the main objective of the present study was to improve the proliferative capacity of these cells in vitro. The aim of the present study was to avoid cell ageing and retain the potential of hBMSCs to expand and differentiate. We sought to achieve this aim by transferring the hTERT gene together with the CTLA4Ig gene into hBMSCs and successfully screening out cells (which we termed hTERT–CTLA4Ig hBMSCs) in which both genes were successfully modified. To explore the possibility that hBMSCs with modification of these two genes could act as allogeneic seed cells for bone tissue engineering, we have examined the proliferative capacity of hTERT–CTLA4Ig hBMSCs in vitro, and their potential for osteogenesis. Our results demonstrate that hTERT–CTLA4Ig hBMSCs expressed both telomerase and CTLA4Ig without mutual interference, exceeded their normal life span by 46 passages and retained the potential for osteodifferentiation, both in vitro and in vivo.

2. Materials and methods 2.1. Cell culture All experiments, including clinical procedures, carried out in this study were approved by the Ethics Committee of Southwest Hospital, Chongqing, China. Primary hBMSCs J Tissue Eng Regen Med (2014) DOI: 10.1002/term

In vitro and in vivo characterization and osteogenic differentiation

were aspirated from the bone marrow of the iliac crest of healthy adult volunteers, all of whom had provided prior informed consent. hBMSCs were obtained by density gradient centrifugation (900 × g for 20 min at 20°C) in a Percoll solution (Pharmacia Corp., USA), as previously reported (Jaiswal et al., 1997). The resulting cells were cultivated in F12–Dulbecco’s modified Eagle’s medium (F12-DMEM; Invitrogen, USA) supplemented with 10% fetal calf serum (FCS; Gibco, Invitrogen), 100 U/ml penicillin, 100 U/ml streptomycin (Gibco, Invitrogen) and 2 mM L-glutamine.

2.2. Characterization of cultured cells by flow cytometry Flow-cytometric analysis was carried out to identify specific markers on the cultured cells. CD105 is known to be present on hBMSCs, whereas CD34 is a haematopoietic lineage marker that would not be expected to be present in these cells (Djouad et al., 2005; Dominici et al., 2006); hence, hBMSCs would be expected to be CD105-positive and CD34-negative, therefore these were chosen as markers to confirm the identity of the cultured cells as hBMSCs. Cells in suspension were mixed with fluorescently-conjugated monoclonal antibodies raised against CD34, labelled with phycoerythrin (PE), and CD105, labelled with fluorescein isothiocyanate (FITC) (Sigma, USA). FITC-labelled mouse IgG1 isotype control antibody (Beijing 4A Biotech Co., Beijing, China) and PE-labelled mouse IgG1 isotype control antibody (Beijing 4A Biotech Co) were used as the isotype controls for the flow-cytometry antibodies to eliminate non-specific immunostaining. Staining was performed according to the manufacturer’s recommendations. Immunostained cells were acquired and analysed on a FACS Calibur flow cytometer (BD Biosciences, USA), using cellQUEST software (BD Biosciences). The cells used for flow-cytometry experiments were obtained from passage 3 (P3).

2.3. Construction of the hTERT retroviral vector and transfection of hBMSCs The sequence coding for hTERT was extracted from the pGRN145 plasmid, kindly provided by Professor Okarma (Geron Corp., USA), and subcloned sequentially into the pLXSN retroviral shuttle plasmid (preserved by the Institute of Burn Research, Southwest Hospital, Chongqing, China) in order to construct the recombinant plasmid, pLXSN–hTERT. Both the target and shuttle plasmids were identified by restriction endonuclease (EcoRI; Roche, USA) digestion and ligated using the Quick LigationTM kit (Roche). The recombinant expression plasmid was identified by restriction enzyme (BamI) digestion and sequencing (Sunbiotech Co. Ltd, China). The primers used for sequencing were: forward 5′-CCCTTGAACCTCCTCGTTCGACC-3′, reverse 5′-GAGCCTGGGGACTTTCCACACCC-3′. Copyright © 2014 John Wiley & Sons, Ltd.

The vectors were packaged and amplified in PT67 cells (stored by the Institute of Burn Research, Southwest Hospital, Chongqing, China) using Lipofectamine 2000 (Invitrogen), according to the manufacturer’s instructions; 72 h later, PT67 cells transduced with pLXSN–hTERT were selected for resistant cells in medium containing 0.5 mg/ml G418 (Gibco). The supernatant from the transduced cells was harvested 48 h later, supplemented with polybrene to a concentration of 4 μg/ml, passed through a 0.45 μm filter and diluted to produce a 10-fold dilution series. The titres of the neo-containing vectors were measured using the NIH/3T3 mouse embryo fibroblast cell line as G418 colony-forming units (CFU), as described previously (Salgar et al., 2004); CFU was calculated as: CFU = number of clones/(volume of virus solution × multiple of the dilution). PT67 cells transduced with pLXSN–hTERT were cloned in 24-well plates and the clones were analysed for vector production after 30 days. Retroviral supernatant was prepared at a titre of 1.2 × 104 particles/ml for infection of hBMSCs. PT67 cells treated with mitomycin (10 μg/ml) were packed into the bottom of a culture flask and covered with well-grown hBMSCs (at a density of 106 cells/ml) at P2; this was followed by the addition of DMEM containing F12 and 10% FBS (Hyclone, USA). The cells were cultured for 5 days at 37°C in air containing 5% CO2, and the selected hBMSCs were then transfected with a neo-containing vector with G418; these were termed hTERT hBMSCs.

2.4. Determination of hTERT expression and activity The presence of hTERT in hTERT hBMSCs was confirmed using immunocytochemical methods. hTERT hBMSCs screened for 45 days were plated onto coverslips; cells at P3 were also plated as a control. On reaching 70% confluence, the cells were washed three times with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde in PBS, pH 7.4, for 30 min at room temperature. The cells were permeabilized with 0.2% v/v Triton X-100 in PBS (4°C for 5 min). Endogenous peroxidase was blocked by incubation with 3% hydrogen peroxide for 10 min, and normal rabbit serum was used for non-specific protein binding. Immunostaining was conducted using the SP Kit (Zhongshan Corp., China), according to the manufacturer’s instructions. The primary antibody was goat anti-human hTERT polyclonal antibody (Santa Cruz Biotechnology, USA), used at a dilution of 1:800; PBS was used as a negative control. Cells were visualized by means of a DAB kit (Zhongshan Corp., China), and were also counterstained with haematoxylin. Imaging was carried out using an inverted microscope (Olympus, Japan). The telomerase activities of hTERT hBMSCs screened for 45 days and (as a control) hBMSCs at P2 were assessed by the PCR-based Telomerase Repeat Amplification Protocol, using the TRAPEZE® telomerase detection kit (S7700; Chemicon Int., USA), according to the manufacturer’s recommendations. Briefly, subconfluent cells were J Tissue Eng Regen Med (2014) DOI: 10.1002/term

F. Dai et al.

trypsinized, washed with PBS, pH 7.3, pelleted and resuspended in the CHAPS lysis buffer (200 μl/106 cells) supplied in the TRAPEZE® kit. The samples used for testing included cell extracts, heat-inactivated cell extracts, a positive telomerase control extract, a primer–dimer/PCR contamination control and a quantitation control template: 48 μl TRAPEZE master mix containing 1.5 μg cell extract was incubated at 30°C for 30 min, and this was followed by two-step PCR in a thermocycler at 94°C/30 s and 59°C/30 s for 33 cycles (Bio-Rad, USA). After amplification, 25 μl PCR products was loaded onto a 10% non-denaturing polyacrylamide gel for electrophoresis, and the bands visualized by ethidium bromide staining. As part of the internal control for the assay, the TRAPEZE primer mix also included PCR-amplified internal control oligonucleotides that produced a band of 36 base pairs (bp) in each lane.

2.5. Preparation of hTERT–CTLA4Ig hBMSCs and determination of CTLA4Ig expression The adenoviral vector, Adv–CTLA4Ig–EGFP (preserved by the Institute of Burn Research, Southwest Hospital, Chongqing, China) was employed in these experiments. Viral supernatant containing Adv–CTLA4Ig–EGFP was generated using HEK293 as the packing cells. The hTERT hBMSCs and hBMSCs (negative controls) were transduced with the adenoviral vector as previously described (Dai et al., 2006). In brief, upon reaching 70% confluence, the cells were exposed to 500 μl viral supernatant containing adenovirus at a titre of 1.4 × 1010 PFU/ml in the presence of 2.5 ml antibiotic-free and serum-free DMEM, and incubated at 37˚C. After 1 h, the cells were washed with PBS, and the medium replaced with complete medium (F12/DMEM containing 10% FBS). At the indicated time points, the cultures were examined microscopically for the expression of enhanced green fluorescent protein (EGFP), using an inverted microscope (Olympus), and the percentage of EGFP-positive cells was quantified by fluorescence-activated cell sorting (FACS) analysis, using a FACS Caliber cytometer (BD Biosciences). The expression of CTLA4Ig protein was determined by immunocytochemical staining, as described above. Rabbit anti-human CTLA4Ig polyclonal antibody (Santa Cruz Biotechnology) was used as the primary antibody at a dilution of 1:400, whereas PBS was used as a negative control. The expression of CTLA4Ig protein was also detected by western blot analysis. Five days after infection, the supernatants from cultured hBMSCs and hTERT–CTLA4Ig hBMSCs (80% confluence) were collected and centrifuged at 5000 × g for 15 min, and the protein concentrations were determined using a Bradford assay. Samples were matched for protein, separated by SDS–polyacrylamide gel electrophoresis on a 10% acrylamide gel, transferred to a polyvinylidene fluoride (PVDF) membrane and immunoblotted with rabbit antiCTLA4Ig polyclonal antibody (Santa Cruz Biotechnology). The bands were visualized using enhanced chemiluminescence (Pierce, USA). Copyright © 2014 John Wiley & Sons, Ltd.

2.6. Cell cycle analysis by fluorescence-activated cell sorting (FACS) In order to determine whether any possible differences in the rates of proliferation between hTERT–CTLA4Ig hBMSCs, hTERT hBMSCs and hBMSCs were reflected by differences in the proportions of cells in the dividing and non-dividing phases of the cell cycle, cell cycle analysis was performed as previously reported (Tan et al., 2010). Cells were detached in 0.25 g/l trypsin and washed twice in ice-cold PBS. Following fixation in 70% ethanol overnight at 4°C, the cells were washed twice more in PBS and then stained for 30 min at 37°C with 50 mg/ml propidium iodide (PI) solution (Sigma) containing 200 mg/ml RNase A and 0.1% Triton-X-100. The cells were then subjected to flow cytometry using a FACS Caliber cytometer (BD Biosciences). The resulting data were analysed using Modfit LT software (Verity Software House, USA), which is capable of deconvoluting DNA content into frequency histograms. Estimates of the percentages of cells in the G0/G1, S and G2/M phases were provided by analyses of the PI profiles using Modfit LT software.

2.7. Osteogenic differentiation of hTERT–CTLA4Ig hBMSCs in vitro The potential for hTERT–CTLA4Ig hBMSCs to differentiate into osteocytes was studied using differentiation induction medium (Leboy et al., 1989). Briefly, hTERT–CTLA4Ig hBMSCs were seeded at a density of 3 × 103 cells/cm2 in flasks and cultured with MSC growth medium containing 100 nM dexamethasone, 50 μM ascorbic acid 2-phosphate and 10 mM β-glycerophosphate (Sigma). The cultures were maintained for 3 weeks, with the culture medium replaced every 3 days. Osteogenic induction of cells was confirmed by serial detections, as follows: calcareous mineral deposition was determined by labelled tetracycline fluorescence staining (Jaiswal et al., 1997); calcareous nodules were stained by alizarin red S (Hung et al., 2002); alkaline phosphatase was stained using calcium–cobalt methods (Van Goor et al., 1989); and osteocalcin expression was verified by immunocytochemistry with goat antiosteocalcin antibody (1:500; Santa Cruz Biotechnology).

2.8. Preparation and xenotransplantation of tissue-engineering bone Tissue-engineering bone was generated by combining hTERT–CTLA4Ig hBMSCs with demineralized bone matrix (DBM), which was used as a scaffold as previously described (Dai et al., 2006). The hTERT–CTLA4Ig hBMSCs were cultured for 7 days for osteoinduction. The osteoinduced cells (2 × 106) were seeded onto DBM and further cultured for 7 days in osteoinduction medium before transplantation. The number of cells adhering to DBM was 1.0–1.5 × 106 cells/g. J Tissue Eng Regen Med (2014) DOI: 10.1002/term

In vitro and in vivo characterization and osteogenic differentiation

Specific pathogen-free adult (18 week-old) male and female Wistar rats (150–200 g) were purchased from the Animal Centre, Research Institute of Surgery, Third Military Medical University, China (Certificate No. scxk (Yu) 2002–0002, grade II). The Wistar rats were divided randomly into four groups, with six animals in each group, for transplantation of either the prepared hTERT–CTLA4Ig hBMSCs, CTLA4Ig hBMSCs, hBMSCs or control implants (DBM-decalcified bone matrix). Surgery was performed under intraperitoneal anaesthesia (1% amobarbital sodium, 100 mg/kg). A 1 cm fullthickness incision was made to the skin of the flank, parallel to the spinal column. Subcutaneous pockets were created by blunt dissection, and 2 g of the prepared cells/implant (as appropriate for the group) were transplanted subcutaneously into these pockets. The wounds were closed using staples, and 80 IU penicillin was administered by intramuscular injection. All protocols were approved by the Ethics Committee of Southwest Hospital, Chongqing, China.

2.9. Radiographic examination and histopathological analysis Serial radiographs of the rats were taken at postoperative weeks 2, 6 and 12, following intra-abdominal injection of amylobarbitone sodium. The rats were positioned anteroposteriorly and laterally for X-ray exposure. Radiographs were imaged using an SF-50 IBY camera (Siemens, Germany) with a 30 s exposure at 25 kV and 3 mA. The graft samples were harvested at designated time points and fixed in 4% paraformaldehyde for 48 h. After washing twice with PBS (15 min each), the grafts were demineralized in 50% EDTA, pH 7.2, for 14 days. The demineralized samples were washed twice in PBS (15 min each) and dehydrated in a graded series of ethanols. The samples were embedded in paraffin, cut into 5 mm sections and stained with haematoxylin and eosin (H&E). The stained sections were examined under a light microscope.

Figure 1. Identification of isolated hBMSCs. (A) The morphological appearance of the hBMSCs (reaching 80% confluence) after culture for 10 days; most of the cells were spindle-shaped, although a few polygonal cells were present. (B) Flow-cytometric analysis of the expression of surface antigens on hBMSCs at passage 3. hBMSCs were immunolabelled with FITC-conjugated anti-CD105 antibody (upper panels) and PE-conjugated anti-CD34 antibody (lower panels); dead cells were eliminated by forward and side scattering; 90.8% of hBMSCs were positive for CD105, whereas 96.2% of hBMSCs were negative for CD34 Copyright © 2014 John Wiley & Sons, Ltd.

J Tissue Eng Regen Med (2014) DOI: 10.1002/term

F. Dai et al.

2.10. Statistical analysis All statistical analyses were conducted using SPSS v. 13.0 (SPSS Inc., Chicago, IL, USA). Data are presented as mean ± standard deviation (SD). Statistical comparisons of values between the various groups were carried out using analysis of variance (ANOVA) or Student’s t-test, as appropriate. For cell-cycle analysis co-variance analysis was used, due to the correlation between the cell passage and the percentage of cells in the various phases of the cell cycle, described by regression analysis (see supporting information, Table S1). p < 0.05 was considered to be indicative of statistical significance.

3. Results 3.1. Culture of hBMSCs Percoll-separated bone marrow aspirate (Percoll 1.073 g/ml) was plated in flasks and the growth medium changed 48 h later. Fusiform or polygonal cells could be seen adhering to the flask (Figure 1A). FACS analysis of the hBMSCs harvested from P3 showed that 90.8% were CD105-positive and 96.2% were CD34-negative (Figure 1B). Since CD105 is known to be present on hBMSCs, and CD34 is a haematopoietic lineage marker that would not be expected to be present on hBMSCs (Djouad et al., 2005; Dominici et al., 2006), these data imply that the harvested hBMSCs were of high purity.

3.2. Identification of the pLXSN–hTERT recombination plasmid and characteristics of hTERT hBMSCs EcoRI was used to cut the pGRN145 plasmid into an 11.5 kb vector fragment and a 3.5 kb hTERT target fragment, and the pLXSN plasmid into a 5.9 kb linear fragment (Figure 2A). The recombinant pLXSN–hTERT plasmid (8.4 kb) was successfully constructed using ligase and identified by restriction enzyme (BamI) digestion (Figure 2B). DNA sequencing revealed that 3402 bases of the pLXSN–hTERT recombinant plasmid matched with 3402 bases of the hTERT sequence in GenBank, a 100% match (data not shown). The PT67/hTERT clone obtained was able to produce virus at a titre of 1.2 × 104 CFU. The hBMSCs that survived co-culture with package PT67 cells in selection medium (G418 at a concentration of 50 μg/ml) at 30 days represented successful selection, and these were termed hTERT hBMSCs. The hTERT hBMSCs were morphologically similar to the primary hBMSCs, even at P46 (Figure 3B2), whereas hBMSCs exhibited a flattened morphology and large size, with a vacuole in the cytoplasm, even at P9 (Figure 3B1). Immunocytochemistry analysis demonstrated successful expression of hTERT protein in the nuclei of both hBMSCs and hTERT hBMSCs (Figure 3A1, A2); staining was positive Copyright © 2014 John Wiley & Sons, Ltd.

Figure 2. Identification of the generated hTERT-based retroviral vectors. (A) Digestion of the pLXSN retroviral shuttle plasmid and the pGRN145 plasmid containing the hTERT sequence by EcoRI. An 11.5 kb vector fragment and a 3.5 kb targeted hTERT fragment were obtained from digestion of the pGRN145 plasmid, while the pLXSN shuttle plasmid was cut into a 5.9 kb linear fragment: 1, marker λDNA/HindIII; 2, plasmid pLXSN; 3, plasmid pGRN145. (B) Identification of the recombinant pLXSN–hTERT plasmid by BamHI digestion. An 8.4 kb recombinant plasmid was present in four ampicillin-resistant monocolonies and accounted for successful construction of the hTERT retroviral vector: 1, marker λDNA/HindIII; 2–5, positive colony containing the targeted recombinant plasmid; 6, marker DL2000; representative agarose gels are shown

in approximately 99% of hTERT hBMSCs and 65% of hBMSCs. In hTERT hBMSCs, the TRAPeze assay revealed a ladder of products with six-base increments, starting at 50 nucleotides (i.e. 50, 56, 62, 68, etc.), together with a 36 bp internal control band; in contrast, only a 36 bp internal control band and a 50 bp band were evident in hBMSCs. Therefore, there was a tendency toward a higher telomerase activity in hBMSCs containing the hTERT transgene than in untransfected hBMSCs (Figure 3D). Cell cycle analysis (Figure 3C1–3) suggested a significantly higher percentage of hBMSCs in G0/G1 phase ( p = 0.037; co-variance analysis) and a significantly lower percentage of hBMSCs in the S phase ( p = 0.015; co-variance analysis) as compared with hTERT hBMSCs. The smaller proportion of hBMSCs found in the dividing phase of the cell cycle, compared with hTERT hBMSCs, was consistent with hTERT hBMSCs showing a higher rate of proliferation.

3.3. Characterization of hTERT–CTLA4Ig hBMSCs After infection of the hTERT hBMSCs with Adv– CTLA4Ig–EGFP, most of the hTERT–CTLA4Ig hBMSCs were able to express EGFP, with peak expression occurring on day 5 (Figure 4A). FACS revealed that 83.75% of the hTERT–CTLA4Ig hBMSCs were CTLA4Ig-positive (Table 1; Figure 4B). Immunocytochemical staining showed that CTLA4Ig protein was expressed in the cytoplasm of the hTERT–CTLA4Ig hBMSCs, particularly around the cell nucleus (Figure 4C). Moreover, CTLA4Ig could be detected, by western blot analysis, in the supernatant obtained from cultured hTERT–CTLA4Ig hBMSCs (Figure 4D). Cell cycle analysis demonstrated that, J Tissue Eng Regen Med (2014) DOI: 10.1002/term

In vitro and in vivo characterization and osteogenic differentiation

Figure 3. Characterization of telomerized hBMSCs. (A) Immunoreactivity of hBMSCs (A1) and hTERT hBMSCs (A2) to anti-hTERT polyclonal antibody; 65% of hBMSCs were positive for hTERT, while 99% of hTERT hBMSCs were positive for hTERT (×200). (B) Morphological appearance (under the light microscope) of hBMSCs at passage 9 (B1) and hTERT hBMSCs at passage 46 (B2). (C) A comparison over 12 passages of the proportion of hBMSCs and hTERT hBMSCs in various phases of the cell cycle. A significantly higher percentage of hBMSCs were in G0/G1 phase (C1), compared with hTERT hBMSCs (co-variance analysis, F = 5.396, p = 0.026; n = 3 for each). There were no statistically significant differences between hBMSCs and hTERT hBMSCs in the percentage of cells in G2/M phase (C2; F = 0.034, p = 0.858; n = 3 for each). A significantly lower percentage of hBMSCs were in S phase (C3) compared with hTERT hBMSCs (F = 8.555, p = 0.015; n = 3 for each). **p < 0.05, Student’s t-test; error bars indicate SD. (D) Telomerase activity, measured by the TRAP protocol, in hBMSCs and their telomerized counterparts: 1, hBMSCs; 2, hBMSCs, heat-inactivated control; 3, hTERT hBMSCs; 4, hTERT hBMSCs, heat-inactivated control; 5, telomerase-positive control; 6, telomerase-positive, heat-inactivated control; 7, telomerase quantitation control template; 8, primer–dimer/PCR contamination control. The telomerized cells showed a typical six-nucleotide DNA ladder, whereas no activity was observed in hBMSCs. No TRAP activity was present in heat-treated (HT; 65°C for 10 min) samples

compared with hTERT hBMSCs, a significantly higher percentage of hTERT–CTLA4Ig hBMSCs were in G0/G1 phase (p < 0.001, Student’s t-test) and significantly lower proportions were in G2/M (p < 0.001, Student’s t-test) and S (p = 0.003, Student’s t-test) phases. This would be consistent with hTERT–CTLA4Ig hBMSCs having a slower rate of proliferation. Copyright © 2014 John Wiley & Sons, Ltd.

3.4. Osteogenic differentiation of hTERT–CTLA4Ig hBMSCs in vitro Figure 5A, B shows representative images showing the morphological characteristics of hTERT–CTLA4Ig hBMSCs after 2 and 3 weeks of culture, respectively. hTERT–CTLA4Ig hBMSCs were found to be positive for hydroxyapatite J Tissue Eng Regen Med (2014) DOI: 10.1002/term

F. Dai et al.

Figure 4. Identification of hTERT–CTLA4Ig hBMSCs. (A) hTERT hBMSCs were infected with the CTLA4Ig-based adenoviral vector, AdvCTLA4Ig–EGFP; 5 days after infection, EGFP expression was detected by fluorescence microscopy (×100). (B) Five days after infection, EGFP expression was measured by FACS analysis; 83.75% of the hTERT hBMSCs were CTLA4Ig-positive cells. (C) CTLA4Ig protein expression, detected using immunocytochemistry methods, in hTERT hBMSCs infected with AdvCTLA4Ig–EGFP (×100). (D) Detection, by western blotting, of CTLA4Ig protein in the supernatant produced by hTERT–CTLA4Ig hBMSCs: 1, CTLA4Ig protein, positive control; 2 and 3, negative control; 4 and 5, protein extracted from the supernatant of hTERT–CTLA4Ig hBMSCs. (E) A comparison of the proportions of hTERT hBMSCs (at passage 3) and hTERT–CTLA4Ig hBMSCs (5 days after infection with AdvCTLA4Ig–EGFP) in various phases of the cell cycle. There was a significantly higher percentage of hTERT–CTLA4Ig hBMSCs in G0/G1 phase ( p < 0.001) and significantly lower percentages in G2/M (p < 0.001) and S ( p = 0.003) phases, relative to hTERT hBMSCs (Student’s t-test, n = 3 for each); error bars indicate SD

mineral deposition, as detected by labelled tetracycline fluorescence staining (Figure 5C), calcareous nodules, as shown by alizarin red S (Figure 5D, E), and alkaline phosphatase, as stained by the calcium– cobalt method (Figure 5F, G). Moreover, immunocytochemistry demonstrated the presence of osteocalcin protein in the cytoplasm of the hTERT–CTLA4Ig hBMSCs (Figure 5H). Copyright © 2014 John Wiley & Sons, Ltd.

3.5. hTERT–CTLA4Ig hBMSCs form bone tissue in vivo After 1 week of cultivation under conditions to induce osteogenic differentiation, hBMSCs, CTLA4Ig hBMSCs or hTERT–CTLA4Ig hBMSCs (1 × 106 cells) were seeded onto 2 g decalcified bone matrix (DBM). This cell matrix was continuously cultured in the same medium for a further J Tissue Eng Regen Med (2014) DOI: 10.1002/term

In vitro and in vivo characterization and osteogenic differentiation Table 1. Immunohistochemical detection of CTLA4Ig in hTERT– CTLA4Ig hBMSCs hTERT–CTLA4Ig hBMSCs CTLA4Ig immunoreactivity

++

+

–

Culture duration 5 days 28 days

43% 31%

39% 27%

18% 42%

Immunoreactivity of hTERT–CTLA4Ig hBMSCs to goat anti-human CTLA4Ig antibody. CTLA4Ig was found to be localized to the cytoplasm of the cells, enriched around the nuclear membrane. ++, high immunoreactivity; +, moderate immunoreactivity; –, no immunoreactivity.

week and then implanted subcutaneously into rats. Before grafting, the tissue-engineered bone contained 1 × 106 cells/g DBM (Figure 6A). Postoperative radiographic findings (12 weeks after xenotransplantation) showed that isolated, dense, bone-like nodules were evident in the rats implanted with CTLA4Ig hBMSCs (Figure 6C) or hTERT–CTLA4Ig hBMSCs (Figure 6D), but that only an opaque shadow similar to that of soft tissue was detected in rats receiving hBMSCs (Figure 6B). These observations were further corroborated by histological examination (Figure 6E–G). In the DBM group (i.e. implantation of

Figure 5. Osteogenic differentiation of hTERT–CTLA4Ig hBMSCs in vitro. (A) Representative light microscopic images showing the morphology of hTERT–CTLA4Ig hBMSCs cultured in an osteogenesis-induction medium for 2 weeks (×100). (B) hTERT–CTLA4Ig hBMSCs cultured for 3 weeks (×100). (C) hTERT–CTLA4Ig hBMSCs, cultured for 3 weeks in an osteogenesis-induction medium and visualized using labelled tetracycline fluorescence staining (×100). (D) Alizarin red S staining (×200). (E) Alizarin red S staining (×40). (F) Alkaline phosphatase staining (×100). (G) Alkaline phosphatase staining (×40). (H) Immunocytochemical labelling of osteocalcin with goat anti-osteocalcin antibody (×200) Copyright © 2014 John Wiley & Sons, Ltd.

J Tissue Eng Regen Med (2014) DOI: 10.1002/term

F. Dai et al.

Figure 6. Bone formation by transplanted hTERT–CTLA4Ig hBMSCs in a xenotransplantation model. (A) Osteogenically induced engineered bone tissue was cultured for 7 days prior to implantation (×200). (B–D) Lateral radiographs demonstrating bone formation in the rat xenotransplantation model by hBMSCs (B), CTLA4Ig hBMSCs (C) and hTERT–CTLA4Ig hBMSCs (D). Isolated bone-like nodules (arrows, C, D) were formed in the CTLA4Ig hBMSCs and hTERT–CTLA4Ig hBMSCs groups 12 weeks after transplantation, whereas the hBMSCs group showed a lower radio-opaque intensity and a smaller shadow. (E–G) Sections stained with haematoxylin and eosin (H&E); representative sections are shown following the transplantation of hBMSCs (E, ×400), CTLA4Ig hBMSCs (F, ×200) or hTERT– CTLA4Ig hBMSCs (G, ×400). The bone formed by CTLA4Ig or hTERT–CTLA4Ig hBMSCs (arrows, F, G) was evident 12 weeks after transplantation, whereas a typical bone structure was less evident in the hBMSCs group (E)

DBM without cells), postoperative radiography (12 weeks after transplantation) showed no evidence of neonatal bone tissue formation (data not shown); therefore, further histological assessment in this group was not carried out.

4. Discussion The aim of the present study was to extend our previous work in CTLA4Ig hBMSCs (Dai et al., 2006) and generate a cell with multipotentiality that had the advantages of a reduced risk of immune rejection and a prolonged lifespan. hTERT–CTLA4Ig hBMSCs were successfully generated from primary hBMSCs by sequential transfection of hTERT and CTLA4Ig and found to be capable of osteogenic differentiation in vitro. Furthermore, these cells were able to form bone-like nodules following subcutaneous transplantation into rats. To our knowledge, this is the first report of the generation and characterization of hTERT–CTLA4Ig hBMSCs, and we propose that these cells show promise as seed cells for bone tissue engineering. Copyright © 2014 John Wiley & Sons, Ltd.

In the present study, we have demonstrated that hBMSCs transfected with an hTERT-retroviral vector exhibited high telomerase activity (Figure 3D), with a significantly lower proportion of cells in G0/G1 phase and a higher proportion in S phase (Figure 3C), indicating that hTERT hBMSCs have a stronger capacity for proliferation then hBMSCs. hTERT hBMSCs with an extended life span appear to have a normal phenotype that is similar to that of young cells (Figure 3B). These results are consistent with those of numerous other studies that have reported that hTERT-modification of hMSCs greatly improves the lifespan of the cell without affecting its potential to differentiate into osteocytes, adipocytes and chondrocytes (Bischoff et al., 2012; Huang et al., 2008; Kobune et al., 2003; Piper et al., 2012; Yang et al., 2007). This apparent immortalization of hMSCs has been suggested to be due to increased cell proliferation and decreased cell senescence (Bischoff et al., 2012), and may involve the upregulation of haematopoietic growth factors (Yang et al., 2007). As a step toward overcoming the obstacle of immune rejection of transplanted allogeneic hBMSCs, we have previously reported that hBMSCs modified with the CTLA4Ig gene could significantly inhibit the immune response in a mixed lymphocyte culture model, differentiate normally into osteoblasts in vitro, and survive until bone tissue had formed in vivo (Dai et al., 2006). In the present study, we attempted to modify hBMSCs through transfection with both an hTERT-retroviral vector and a CTLA4Igadenoviral vector, with the aims of extending the lifespan of the hBMSCs and inducing local immunotolerance. We successfully generated hBMSCs modified by both genes (which we termed hTERT–CTLA4Ig hBMSCs) and confirmed that these cells expressed both hTERT and CTLA4Ig proteins (Figures 3A2, 4C, D). These hTERT– CTLA4Ig hBMSCs were able to differentiate in vitro into osteoblasts with osteogenic-specific characteristics (Figure 5A–E) and form neonatal bone in a xenotransplantation model (Figure 6C–E, G), indicating that the hTERT and CTLA4Ig genes did not significantly interfere with the osteogenic potential of these hBMSCs. We further confirmed that the lifespan of the hTERT– CTLA4Ig hBMSCs was prolonged, presumably due to transfection of the hTERT gene. Although several studies (Huang et al., 2008; Jiang et al., 1999; Morales et al., 1999; Yang et al., 1999) have reported that ectopic expression of telomerase by hTERT-transfected primary cells was associated with an extended life span without malignant transformation, there nonetheless remains the risk that tumour formation could potentially occur in hBMSCs transfected with telomerase. The potential for tumourigenesis is highlighted by the observation of epigenetic events in hTERT-transduced hMSC lines (Serakinci et al., 2004). In addition, a murine MSC system has been described that showed aneuploidization, translocations and homozygous loss of the cdkn2 region during malignant transformation, and formed osteosarcomas upon grafting (Mohseny et al., 2009). Interestingly, however, there is also evidence that although murine MSCs have the potential to differentiate J Tissue Eng Regen Med (2014) DOI: 10.1002/term

In vitro and in vivo characterization and osteogenic differentiation

into an osteosarcoma, this is not the case for human MSCs (Aguilar et al., 2007). Indeed, there is no compelling evidence that spontaneous transformation occurs in long-term cultures of human MSCs (Bernardo et al., 2007; Garcia et al., 2010; Izadpanah et al., 2008; Torsvik et al., 2010), while hTERT overexpression has been reported to contribute to genetic stability (Estrada et al., 2013). In the present study, cell-cycle analysis demonstrated that transfection with CTLA4Ig resulted in a lower proliferation capacity and a slower rate of growth (Figure 4E); this characteristic of the CTLA4Ig gene could thus be advantageous in terms of reducing any potential risk of carcinogenesis. Moreover, in our in vivo study there was no evidence of tumour tissue formation (in or around the implant) in any of the experimental groups. We have also carried out additional western blot experiments to confirm that protein expression of the proto-oncogene, C-Jun, is normal in hTERT–hCTLA4 hBMSCs (data not shown). This would support a lack of tumourigenic potential in these cells, since C-Jun is thought to play a key role in the regulation of the induction, growth and invasiveness of osteosarcoma (Fromigué et al., 2008; Dass et al., 2008a, 2008b). Nonetheless, further experimental work is merited to characterize these cells in more detail and establish more definitively whether tumourigenic potential exists. Should no safety concerns be raised, we envisage that a sufficient number of cells for clinical use could be obtained by culturing hTERT hBMSCs in vitro and transfecting these with CTLA4Ig before transplantation.

One potential limitation of our study is that, although the positive expression of CD105 and negative expression of CD34 were confirmed in our cells, additional markers of MSCs were not examined. The International Society for Cellular Therapy has recommended the positive expression of CD73, CD90 and CD105 and the negative expression of CD34 as markers for MSCs (Dominici et al., 2006). Nonetheless, in view of the fact that our cells were CD105-positive and CD34-negative and showed osteogenic differentiation in vitro, we are confident that these cells were derived from MSCs. In conclusion, hTERT–CTLA4Ig hBMSCs exhibited a prolonged lifespan in vitro and good osteogenic potential in vivo. Our study indicates that hTERT–CTLA4Ig hBMSCs have promise as seed cells for use in bone tissue engineering, as they possess a long lifespan, an improved proliferative capacity, an ability to be tolerated by the immune system and normal osteodifferentiation characteristics.

Conflict of interest The authors declare no conflicts of interest.

Acknowledgement This study was supported by the National Natural Science Foundation of China in 2011 (Grant No. 31170931).

References Aguilar S, Nye E, Chan J et al. 2007; Murine but not human mesenchymal stem cells generate osteosarcoma-like lesions in the lung. Stem Cells 25: 1586–1594. Beausejour C. 2007; Bone marrow-derived cells: the influence of aging and cellular senescence. Handb Exp Pharmacol 180: 67–88. Bernardo ME, Zaffaroni N, Novara F et al. 2007; Human bone marrow derived mesenchymal stem cells do not undergo transformation after long-term in vitro culture and do not exhibit telomere maintenance mechanisms. Cancer Res 67: 9142–9149. Bischoff DS, Makhijani NS, Yamaguchi DT. 2012; Constitutive expression of human telomerase enhances the proliferation potential of human mesenchymal stem cells. BioRes Open Access 1: 273–279. Bodnar AG, Ouellette M, Frolkis M et al. 1998; Extension of life-span by introduction of telomerase into normal human cells. Science 279: 349–352. Bonab MM, Alimoghaddam K, Talebian F et al. 2006; Aging of mesenchymal stem cells in vitro. BMC Cell Biol 7: 14. Burns JS, Rasmussen PL, Larsen KH et al. 2010; Parameters in three-dimensional osteospheroids of telomerized human mesenchymal (stromal) stem cells grown on osteoconductive scaffolds that predict in vivo bone-forming potential. Tissue Eng A 16: 2331–2342. Copyright © 2014 John Wiley & Sons, Ltd.

Dai F, Shi D, He W et al. 2006; hCTLA4-gene modified human bone marrow-derived mesenchymal stem cells as allogeneic seed cells in bone tissue engineering. Tissue Eng 12: 2583–2590. Dass CR, Khachigian LM, Choong PF. 2008a; c-Jun is critical for the progression of osteosarcoma: proof in an orthotopic spontaneously metastasizing model. Mol Cancer Res 6: 1289–1292. Dass CR, Friedhuber AM, Khachigian LM et al. 2008b; Downregulation of c-jun results in apoptosis-mediated anti-osteosarcoma activity in an orthotopic model. Cancer Biol Ther 7: 1033–1036. Djouad F, Bony C, Haupl T et al. 2005; Transcriptional profiles discriminate bone marrow-derived and synovium-derived mesenchymal stem cells. Arthritis Res Ther 7: R1304–1315. Dominici M, Le Blanc K, Mueller I et al. 2006; Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy 8: 315–317. Estrada JC, Torres Y, Benguría A et al. 2013; Human mesenchymal stem cell-replicative senescence and oxidative stress are closely linked to aneuploidy. Cell Death Dis 4: e691. Fromigué O, Hamidouche Z, Marie PJ. 2008; Blockade of the RhoA-JNK-c-JunMMP2 cascade by atorvastatin reduces

osteosarcoma cell invasion. J Biol Chem 283: 30549–30556. Garcia S, Bernad A, Martín MC et al. 2010; Pitfalls in spontaneous in vitro transformation of human mesenchymal stem cells. Exp Cell Res 316: 1648–1650. Horwitz EM, Gordon PL, Koo WK et al. 2002; Isolated allogeneic bone marrow-derived mesenchymal cells engraft and stimulate growth in children with osteogenesis imperfecta: implications for cell therapy of bone. Proc Natl Acad Sci U S A 99: 8932–8937. Huang G, Zheng Q, Sun J et al. 2008; Stabilization of cellular properties and differentiation mutilpotential of human mesenchymal stem cells transduced with hTERT gene in a long-term culture. J Cell Biochem 103: 1256–1269. Hung SC, Chen NJ, Hsieh SL et al. 2002; Isolation and characterization of size-sieved stem cells from human bone marrow. Stem Cells 20: 249–258. Izadpanah R, Kaushal D, Kriedt C et al. 2008; Long-term in vitro expansion alters the biology of adult mesenchymal stem cells. Cancer Res 68: 4229–4238. Jaiswal N, Haynesworth SE, Caplan AI et al. 1997; Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J Cell Biochem 64: 295–312. Jiang XR, Jimenez G, Chang E et al. 1999; Telomerase expression in human somatic cells does not induce changes associated J Tissue Eng Regen Med (2014) DOI: 10.1002/term

F. Dai et al. with a transformed phenotype. Nat Genet 21: 111–114. Jonak ZL, Henri E, Trulli S et al. 1992; Manipulation of human B cells to confer immortality. Hum Antibodies Hybridomas 3: 177–185. Kassem M, Abdallah BM, Yu Z et al. 2004; The use of hTERT-immortalized cells in tissue engineering. Cytotechnology 45: 39–46 Kim S, Banga S, Jha K et al. 1997; SV40-mediated transformation and immortalization of human cells. Dev Biol Stand 94: 297–302. Kobune M, Kawano Y, Ito Y et al. 2003; Telomerized human multipotent mesenchymal cells can differentiate into hematopoietic and cobblestone area-supporting cells. Exp Hematol 31: 715–722. Løken S, Jakobsen RB, Årøen A et al. 2008; Bone marrow mesenchymal stem cells in a hyaluronan scaffold for treatment of an osteochondral defect in a rabbit model. Knee Surg Sports Traumatol Arthrosc 16: 896–903. Leboy PS, Vaias L, Uschmann B et al. 1989; Ascorbic acid induces alkaline phosphatase, type X collagen, and calcium deposition in cultured chick chondrocytes. J Biol Chem 264: 17281–17286. Mohseny AB, Szuhai K, Romeo S et al. 2009; Osteosarcoma originates from mesenchymal stem cells in consequence of aneuploidization and genomic loss of Cdkn2. J Pathol 219: 294–305. Morales CP, Holt SE, Ouellette M et al. 1999; Absence of cancer-associated changes in human fibroblasts immortalized with telomerase. Nat Genet 21: 115–118. Nakahara H, Misawa H, Hayashi T et al. 2009; Bone repair by transplantation of hTERT-immortalized human mesenchymal stem cells in mice. Transplantation 88: 346–353.

Petite H, Viateau V, Bensaid W et al. 2000; Tissue-engineered bone regeneration. Nat Biotechnol 18: 959–963. Piper SL, Wang M, Yamamoto A et al. 2012; Inducible immortality in hTERT human mesenchymal stem cells. J Orthop Res 30: 1879–1885. Salgar SK, Yang D, Ruiz P et al. 2004; Viral interleukin-10-engineered autologous hematopoietic stem cell therapy: a novel gene therapy approach to prevent graft rejection. Hum Gene Ther 15: 131–144. Serakinci N, Guldberg P, Burns JS et al. 2004; Adult human mesenchymal stem cell as a target for neoplastic transformation. Oncogene 23: 5095–5098. Sugden B, Mark W. 1977; Clonal transformation of adult human leukocytes by Epstein–Barr virus. J Virol 23: 503–508. Tan J, Peng X, Luo G et al. 2010; Investigating the role of P311 in the hypertrophic scar. PLoS One 5: e9995. Tao Q, Lv B, Qiao B et al. 2009; Immortalization of ameloblastoma cells via reactivation of telomerase function: phenotypic and molecular characteristics. Oral Oncol 45: e239–244. Torsvik A, Røsland GV, Svendsen A et al. 2010; Spontaneous malignant transformation of human mesenchymal stem cells reflects cross-contamination: putting the research field on track – letter. Cancer Res 70: 6393–6396. Van Goor H, Gerrits P, Hardonk M. 1989; Enzyme histochemical demonstration of alkaline phosphatase activity in plasticembedded tissues using a Gomori-based cerium–DAB technique. J Histochem Cytochem 37: 399–403. Wang L, Fan H, Zhang ZY et al. 2010; Osteogenesis and angiogenesis of tissueengineered bone constructed by prevascularized β-tricalcium phosphate

scaffold and mesenchymal stem cells. Biomaterials 31: 9452–9461. Woodbury D, Reynolds K, Black IB. 2002; Adult bone marrow stromal stem cells express germline, ectodermal, endodermal, and mesodermal genes prior to neurogenesis. J Neurosci Res 69: 908–917. Woodworth C, Doniger J, Di Paolo J. 1989; Immortalization of human foreskin keratinocytes by various human papillomavirus DNAs corresponds to their association with cervical carcinoma. J Virol 63: 159–164. Yang J, Cao C, Wang W et al. 2010; Proliferation and osteogenesis of immortalized bone marrow-derived mesenchymal stem cells in porous polylactic glycolic acid scaffolds under perfusion culture. J Biomed Mater Res A 92: 817–829. Yang J, Chang E, Cherry AM et al. 1999; Human endothelial cell life extension by telomerase expression. J Biol Chem 274: 26141–26148. Yang J, Nagavarapu U, Relloma K et al. 2001; Telomerized human microvasculature is functional in vivo. Nat Biotechnol 19: 219–224. Yang YX, Miao ZC, Zhang HJ et al. 2007; Establishment and characterization of a human telomerase catalytic subunittransduced fetal bone marrow-derived osteoblastic cell line. Differentiation 75: 24–34. Yilgor P, Sousa RA, Reis RL et al. 2010; Effect of scaffold architecture and BMP-2/ BMP-7 delivery on in vitro bone regeneration. J Mater Sci Mater Med 21: 2999–3008. Yoon SJ, Park KS, Kim MS et al. 2007; Repair of diaphyseal bone defects with calcitriolloaded PLGA scaffolds and marrow stromal cells. Tissue Eng 13: 1125–1133.

Supporting information Additional supporting information may be found in the online version of this article at the publisher’s web-site.

Copyright © 2014 John Wiley & Sons, Ltd.

J Tissue Eng Regen Med (2014) DOI: 10.1002/term