Adenoviral Mediated Expression of BMP2 by Bone Marrow Stromal Cells Cultured in 3D Copolymer Scaffolds Enhances Bone Formation

Selection of appropriate osteoinductive growth factors, suitable delivery method and proper supportive scaffold are critical for a successful outcome in bone tissue engineering using bone marrow stromal cells (BMSC). This study examined the molecular and functional effect of a combination of adenoviral mediated expression of bone morphogenetic protein-2 (BMP2) in BMSC and recently developed and characterized, biodegradable Poly(L-lactide-co-є-caprolactone){poly(LLA-co-CL)}scaffolds in osteogenic molecular changes and ectopic bone formation by using in vitro and in vivo approaches. Pathway-focused custom PCR array, validation using TaqMan based quantitative RT-PCR (qRT-PCR) and ALP staining showed significant up-regulation of several osteogenic and angiogenic molecules, including ALPL and RUNX2 in ad-BMP2 BMSC group grown in poly(LLA-co-CL) scaffolds both at 3 and 14 days. Micro CT and histological analyses of the subcutaneously implanted scaffolds in NOD/SCID mice revealed significantly increased radiopaque areas, percentage bone volume and formation of vital bone in ad-BMP2 scaffolds as compared to the control groups both at 2 and 8 weeks. The increased bone formation in the ad-BMP2 group in vivo was paralleled at the molecular level with concomitant over-expression of a number of osteogenic and angiogenic genes including ALPL, RUNX2, SPP1, ANGPT1. The increased bone formation in ad-BMP2 explants was not found to be associated with enhanced endochondral activity as evidenced by qRT-PCR (SOX9 and FGF2) and Safranin O staining. Taken together, combination of adenoviral mediated BMP-2 expression in BMSC grown in the newly developed poly(LLA-co-CL) scaffolds induced expression of osteogenic markers and enhanced bone formation in vivo.


Introduction
Proliferation Kit (StemCell Technologies, Part ID 05411) following standard culture protocol. All cell culture experiments were carried out at a humidified atmosphere of 37°C and 5% CO 2 . For validation of the osteogenic and angiogenic gene expression changes induced by adenoviral mediated BMP2 expression in BMSC, commercially available human BMSC from two additional donors were purchased and used (henceforth referred to as donor 2 and 3 BMSC) as described in S1 Text.

Preparation of BMSC seeded scaffolds
Poly(LLA-co-CL) scaffolds were fabricated, using the solvent-casting particulate-leaching method as described previously [16,21]. For in vitro experiments, scaffolds (diameter %12mm, height%1.3mm, porosity: 85% and pore size: 350μm on average) were placed on the bottom of 48-well plates, pre-wetted with the culture media and incubated overnight at a humidified atmosphere of 37°C and 5% CO 2 . BMSC were seeded at a density of 5 × 10 4 cells/scaffold.

Adenoviral expression vector construction and transduction of BMSC
Replication-deficient adenoviral expression vector carrying the coding sequences of BMP2 gene (reference sequence: NM_001200.2)(ad-BMP2)and eGFP gene coding for enhanced green fluorescent protein (eGFP) was purchased from Cyagen Biosciences Inc. Adenoviral vector carrying only eGFPcoding sequences (ad-GFP) was used as a control. Adenoviral particles were generated by transfecting HEK 293 cells (ATCC-CRL-1573) with Pac I digested constructs. Early passage (passage 2-3) BMSC were infected as monolayer culture with respective adenoviruses (multiplicity of infection, MOI = 100 to 150) to obtain 80-90% infection efficiency as examined under the fluorescent microscope. BMSC infected with ad-BMP2 and ad-GFP will henceforth be referred to as 'ad-BMP2 BMSC' and 'ad-GFP BMSC' respectively. To verify upregulation of BMP2 mRNA in ad-BMP2 BMSC in monolayer, cells were harvested after 48 hours of adenoviral infection. To determine the amount of secreted BMP2 after 48 hours of adenoviral infection, culture supernatant was collected and analyzed using commercially available ELISA kit following manufacturer's instructions. Further, BMSC were seeded at a density of 5 × 10 4 cells/scaffold after 48 hours of infection in monolayer with the respective adenoviral particles in Poly(LLA-co-CL) scaffolds. BMSC grown in scaffolds were harvested after 3 and 14 days for mRNA expression analysis. Culture supernatants were also collected on the respective time points for ELISA assay.

Total RNA extraction
Total RNA from the in vitro seeded scaffolds were extracted using Maxwell 1 16 LEV sim-plyRNA Kit (Cat no: AS1270, Promega) on a Maxwell 1 16 instrument following the manufacturer's protocol. Quantity and purity of the total RNA was determined using a Nanodrop Spectrophotometer (ThermoScientific Nano Drop Technologies, Wilmington, DE, USA). Agilent 2100 Bio analyzer (Agilent Technologies) was used to examine the integrity of RNA (data not shown).

Expression analysis of osteogenesis and angiogenesis related genes in vitro
To examine the range of genes modulated by BMP2 over-expression in BMSC in the 3D-scaffold (in vitro), a custom PCR array (Cat no: 330131, SuperArray Bioscience, Frederick, MD, USA) containing primer pairs for 30 genes related to osteogenesis and angiogenesis (Table 1) was used. Total RNA from 3 biological replicates (n = 3) of both ad-GFP and ad-BMP2 groups were used for cDNA synthesis at 3 and 14 days. PCR amplification was performed using the following cycling conditions: 95°C for 10 min, (95°C for 15 sec, and 60°C for 1 min) x 40 cycles in ABI Prism Sequence Detector 7900 HT (Applied Biosystems, Foster City, USA). Pre-and post-PCR quality control measures, as recommended by the manufacturer, were strictly followed. PCR array data were analysed as described previously [25]. Briefly, threshold cycle (Ct) was used to calculate 2 -ΔCt value for each gene using PCR Array Data Analysis Web Portal (SABiosciences). 2 -ΔCt values were then exported to microarray data analysis software (J-Express 2012). For statistical analysis, unsupervised hierarchical clustering and significance analysis of microarray (SAM) tests were used. Differentially expressed genes with false discovery rate (FDR) = 0 were considered to be significantly modulated genes.  Reverse transcription and quantitative RT-PCR (qRT-PCR) using TaqMan assays Three hundred Nano grams of total RNA was converted to cDNA by reverse transcription reaction using a high capacity cDNA Archive Kit (Applied Biosystems, Carlsbad, CA, USA). BMP2 TaqMan assay (Hs00154192_m) was used to verify the expression of BMP2 mRNA in adenovirus transduced BMSC. The in vitro PCR array results were independently validated by performing qRT-PCR using TaqMan assays for selected key genes {ALPL (Hs01029144_m1), RUNX2(Hs00231692_m1), BGLAP (Hs01587814_g1) and VEGFA (Hs00900055_m1)}. GAPDH (Hs99999905_m1) was used as an endogenous control. All qRT-PCR amplifications were performed on ABI Prism Sequence Detector 7900 HT (Applied Biosystems, Foster City, USA) with standard cycling conditions. Comparative 2 -ΔΔ Ct method was used to quantify the relative mRNA expression.
Alkaline phosphatase (ALP) staining ALP staining was done to analyze the osteoblastic differentiation potential of BMSC transduced with ad-BMP2. After 48 hours of infection, cells were trypsinized and 2 × 10 4 cells were seeded in monolayer on 4-well culture dish. ALP staining was done on day 3, and 14. Briefly, cells were washed with phosphatase buffer saline and stained for alkaline phosphatase (ALP) activity using Napthol AS-TR phosphate and fast red violet B salt (Sigma-Aldrich) as described previously [26].
In vivo subcutaneous implant model in nonobese severe combined immunodeficient (NOD/SCID) mice a. Preparation of BMSC seeded scaffolds. Poly(LLA-co-CL) scaffolds (diameter % 6mm, height%1.3mm and pore size of 90-500μm) were prepared as described previously [16,21]. The scaffolds were placed on the bottom of 96-well plates, pre-wetted with the culture medium and incubated overnight at a humidified atmosphere of 37°C and 5% CO 2 . 5 × 10 5 BMSC were seeded in each scaffold after 48 hours of infection with the respective adenoviral particles, incubated overnight at a humidified atmosphere of 37°C and 5% CO 2 and implanted subcutaneously in NOD/SCID mice. Scaffolds with untransduced BMSC or without any BMSC (only scaffold) were used as additional controls in the in vivo experiments.
b. Surgical Implantation of scaffolds. All animal experiments were approved by the Norwegian Animal Research Authority and conducted according to the European Convention for the Protection of Vertebrates used for Scientific Purposes, with local approval number 4940. Sixteen NOD-SCID mice (6-8 weeks old, Taconic Farms, Denmark) (8 mice each for both 2 and 8 week time-points) were used for subcutaneous implantation of scaffolds. The animals were anesthetized by subcutaneous injection Midazolam 5mg/ml/Hyponorm solution. Two midline surgical incisions of approximately 2cm in length were made on the back of mice, which were extended laterally by blunt dissection and a subcutaneous pouch was created. Each animal received four randomly allocated scaffold implants of the following groups: i) ad-GFP (9 replicates), ii) ad-BMP2 (9 replicates), iii) untransduced (9 replicates) and iv) scaffold only group (5 replicates). The scaffolds were implanted in such a way that scaffolds from all the experimental groups are implanted in all possible locations among following four sites: upper right side, upper left side, lower left side and lower right side. Wounds were closed with Histoacryl Tissue Adhesive (n-butyl cyanoacrylate) (3M, St. Paul, MN, USA). Animals were euthanized with CO 2 inhalation and subsequent cervical dislocation at 2 and 8 weeks. Three of the randomly selected scaffolds from only scaffold group and six scaffolds each from un-transduced, ad-GFP and ad-BMP2 groups at 2 week time point were divided into two equal halves. One half of the scaffolds was stored in RNA later (Ambion) for total RNA extraction and the other half was formalin fixed (4% buffered formalin), decalcified (12.5% EDTA and 2.5% PFA in phosphate buffered saline) and paraffin embedded for histological analysis. Remaining scaffolds from all of the experimental groups were also formalin fixed and used for histological analysis. At 8 weeks, 3 randomly selected scaffolds from all of the groups were used for micro CT analysis. All remaining scaffolds were formalin fixed-paraffin embedded for histological analysis.

Expression analysis of osteogenesis and angiogenesis related genes for in vivo scaffold explants
For in vivo scaffold explants (2 weeks), six hundred Nano grams of total RNA from 6 biological replicates (n = 6) of both ad-GFP and ad-BMP2 BMSC groups were converted to first strand cDNA using RT 2 First Strand Kit (C-03, SABiosciences, Frederick, MD, USA). To examine the range of genes modulated by BMP2 over-expression, a custom PCR array (Cat no: 330131, SuperArray Bioscience, Frederick, MD, USA) containing primer pairs for 30 genes related to osteogenesis and angiogenesis (Table 1) was used. PCR amplification and statistical analysis was performed as described beforehand for the in vitro scaffolds.

Histological analysis
The paraffin embedded specimens were cut into 5 μm sections. The sections were deparaffinized, rehydrated in xylene and graded ethanol stained with Hematoxylin and Eosin. Safranin O staining was performed to visualize newly synthesized cartilaginous extracellular matrix glycosaminoglycan. Briefly, sections were stained with 0.1% Safranin O and 0.05% fast green was used as counter stain. The stained sections were dehydrated, cleared and mounted using resinous medium.

Micro computed tomography analysis (μCT)
The in vivo scaffold explants harvested at 8 weeks were scanned with a SkyScan 1172 Xray μCT imaging system (Aartselaar, Belgium) at 10 μm resolution with a voltage of 60 kV with 0.5 aluminum filters. The Projection image was reconstructed using Nrecon software. The quantitative analysis of the image was performed by CTan software provided by SkyScan. A global threshold of 70-255 was applied to all the samples and the region of interest (ROI) was selected by outlining the scaffolds. The percent of binarized volume within ROI represented the percent of bone volume in the scaffolds.

Statistical analyses
Data are expressed as mean ± standard error of the mean (SEM). Statistical analysis for the comparison of means between two groups was performed with Student's t-test; whereas ANOVA test with Bonferroni post hoc analysis was used for comparison between multiple groups. Statistical analyses were performed using GraphPad Prism software version 5.00 for Windows (GraphPad Software, San Diego California USA, www.graphpad.com), with the level of significance set at 5%.

Results
BMP2 adenoviral expression vector up-regulated endogenous BMP2 mRNA and secreted protein levels BMSC transduced with ad-GFP and ad-BMP2 viral particles demonstrated similar morphology in monoculture (Fig 1A-1D). ad-BMP2 BMSC grown in monolayer expressed significantly higher levels of BMP2 mRNA and secreted BMP2 protein at day 3 as compared to the control ad-GFP BMSC (Fig 1E and 1F). In parallel, significantly higher levels of BMP2 mRNA and secreted BMP2 protein was found in ad-BMP2 BMSC grown in 3D scaffold as compared to the control ad-GFP transduced cells both at day 3 and 14 ( Fig 1G-1I).
BMP2 over-expression is associated with up-regulation of osteogenic and angiogenic markers in vitro Custom PCR array was used to examine differentially expressed osteogenesis and angiogenesis related genes with BMP2 over-expression. SAM analysis showed significant (FDR = 0) up- regulation of the early osteogenic marker ALPL at day 3 in ad-BMP2 BMSC grown in 3D-scaffols as compared to the control ad-GFP BMSC (Fig 2A). At day 14, ALPL, RUNX2, BMP6, BGLAP (osteocalcin) and BMP7, the osteogenesis related genes, were found to be overexpressed in ad-BMP2 BMSC (Fig 2B). Additionally, mRNA expression levels of angiogenic molecules, such as FLT1 (VEGFR1) and its ligands VEGFA and PGF, were significantly upregulated in ad-BMP2 BMSC respectively at 3 days and 14 days (Fig 2A and 2B). Independent validation of the differentially expressed selected genes, as identified by PCR array, was done by performing qRT-PCR using TaqMan assays for ALPL, RUNX2 and BGLAP and VEGFA. Consistent with the PCR array results, ALPL was found to be significantly up-regulated in ad-BMP2 BMSC and untransduced BMSC both at day 3 (P<0.001) and day 14 (P<0.001) (Fig 2C  and 2D). Similarly, mRNA levels of RUNX2 (P<0.001), BGLAP (P<0.001) and VEGFA (P<0.001) were significantly higher in ad-BMP2 as compared to that of ad-GFP BMSC and untransduced BMSC at day 14 ( Fig 2D). To confirm the induction of ALPL mRNA at the protein level, ALP staining was performed. ALP level was found to be significantly induced at day 3 and the elevated level was also maintained at day 14 in ad-BMP2 BMSC as compared to the barely detectable levels in ad-GFP BMSC and untransduced BMSC (Fig 2E). Further, overexpression of mRNA levels of ALPL, RUNX2 and VEGFA with adenoviral mediated BMP2 over-expression was confirmed in two additional donors as described in S1 Text and S1 Fig.

BMP2 over-expression induced bone formation in 3D-scaffold explants in NOD/SCID mice
Ability of BMP2 to induce ectopic bone formation in 3D-scaffold was examined by subcutaneously implanting 3D-scaffold seeded with ad-BMP2, ad-GFP, untransduced BMSC and only scaffold controls (Fig 3A). With tactile and visual inspection, ad-BMP2 scaffold explants were found to be harder in consistency with richer vascular networks as compared to that of the control-explants both at 2 and 8 weeks (data for 2 weeks are not shown) (Fig 3B). Micro CT analysis revealed no radiopaque bone like structures in the scaffold explants from only scaffold (n = 3), untransduced (n = 3) and ad-GFP (n = 3) groups at 8 weeks (Fig 3C-3E). In contrast, all analyzed replicates (n = 3) from ad-BMP2 explants at 8 weeks revealed formation of dense bone like structure in the periphery as well as inside of the scaffold explant (Fig 3F). Quantification of ectopic bone formation in the scaffolds showed significantly higher % volume of radiopaque mass in ad-BMP2 explants as compared to the controls (Fig 3G). Histological examination of H & E stained sections was next performed to confirm the formation of bone structures in ad-BMP2 explants. Similar to the microCT findings, no bony structures were detected in all replicates of only scaffold, untransduced and ad-GFP explants both at 2 and 8 weeks (Fig 3H-3M). However, formation of bony structures (black arrows) was detected at the periphery of the scaffold explants in all replicates (9/9) of ad-BMP2 BMSC at 2 weeks (Fig 3N). At 8 weeks, formation of bony structures (green arrows) was more extensive with bony trabeculae extending throughout the whole thickness of the scaffold explants in all of replicates (6/6) examined (Fig 3O). Bony structure consisted of numerous osteocyte like cells both at 2 (black arrowheads, inset N) and 8 (green arrowheads, inset O) weeks. No inflammatory cells, except occasional multinucleated giant cells, were seen in all groups (Fig 3H-3O).

BMP2 over-expression mediated bone formation is associated with upregulation of osteogenic markers in the in vivo 3D-scaffold explants
Custom PCR array was used to examine differentially expressed osteogenesis and angiogenesis related genes in 3D-scaffold explants at 2 weeks. Unsupervised clustering using the significantly modulated genes revealed two distinct clusters for ad-GFP and ad-BMP2 explants (Fig 4). SAM   (Fig 4). Additionally, ANGPT1 (1.8 fold change), an angiogenic factor that modulates endothelial differentiation, was found to be significantly increased in ad-BMP2 explants (Fig 4). Of note, mRNA levels of SOX9, a key transcriptional factor required for the successive steps of chondrogenesis [27] was significantly down-regulated (2.6 folds) in ad-BMP2 explants (Fig 4). In parallel, mRNA levels of FGF2, an upstream positive regulator of SOX9 [28], was also under-expressed (2.3 folds) in ad-BMP2 explants (Fig 4).

BMP2 over-expression mediated bone formation is not associated with enhanced chondrogenic activity
Significant down-regulation of SOX9 and FGF2 mRNA levels in ad-BMP2 explants led us to investigate the presence of cartilaginous extracellular matrix glycosaminoglycan by Safranin in the bony structures. No positive staining for Safranin O was detected in the bony structures in ad-BMP2 scaffold explants both at 2 and 8 weeks (Fig 5A and 5B). However, strong positivity was observed in the positive control (cranial base of mouse embryo E14) (Fig 5C).

Discussion
Employing recently developed biodegradable poly(LLA-co-CL) scaffolds as a carrier, the current study examined the role of BMSC engineered to express BMP2 by adenovirus vector in bone formation. Our results demonstrated that adenoviral mediated BMP2 delivery significantly induced osteogenic differentiation of BMSC grown in poly(LLA-co-CL) scaffolds in vitro as well as bone formation ability in vivo.
Although the therapeutic importance of BMP2 in bone regeneration is well recognized, the precise and physiological delivery of BMP2 at the defect site is difficult to achieve [5]. It has been demonstrated that a short-term expression of the BMP2 is sufficient to irreversibly induce bone formation by BMSC [29], suggesting that a transient expression of BMP2 might be sufficient for cell-mediated BMP2 therapy for bone regeneration. These findings support the use of adenoviral mediated gene delivery system which has been shown to induce a transient (less than 2 weeks) gene expression in vivo [30]. In the current study, ad-BMP2 BMSC expressed significantly higher levels of BMP2 mRNA and secreted protein (Fig 1), indicating that adenoviral mediated BMP2 transduction in BMSC was functional. PCR array showed significant upregulation of several key osteogenic marker genes (ALPL, RUNX2, BGLAP, BMP6, BMP7) in ad-BMP2 at day 3 or day 14 in vitro. These results indicate that BMP2 was able to differentiate BMSC seeded in scaffolds (in vitro) towards an osteogenic pathway. Up-regulation of osteogenic marker genes was also observed in ad-BMP2 BMSC from donor 2 and 3 (S1 Fig), suggesting that osteogenic effects of BMP2 was not restricted to a limited type of BMSC strains. In addition to osteoinduction, angiogenesis is considered essential for bone regeneration [31].

BMP2 Overexpression by BMSC Enhances Bone Formation
Angiogenesis is a complex process requiring a coordinated interplay between growth factors, their receptor systems and down-stream signaling pathways [32]. In the present study, mRNA expression levels of FLT1 (VEGFR1) and its ligands VEGFA and PGF, were significantly upregulated in ad-BMP2 BMSC respectively at day 3 and day 14 (Fig 2), suggesting that BMP2 expression was able to induce pre-angiogenic transcriptional program in BMSC. In line to these findings, previous studies have demonstrated up-regulation of VEGFA and PGF in preosteoblast like mouse cells [33] and human osteoblasts [34] with rhBMP2 treatment. Taken together, adenoviral mediated BMP2 expression in BMSC grown in poly(LLA-co-CL) scaffolds in vitro was able to induce osteogenic differentiation and to program BMSC towards more angiogenic genotype.
Examination of osteogenic ability of adenoviral mediated BMP2 expression was subsequently performed by subcutaneous implantation of poly(LLA-co-CL) scaffold seeded with BMSC in NOD/SCID mice. On gross visual examination, scaffold explants from ad-BMP2 group both at 2-and 8-weeks were reddish in color as compared to the explants from control groups (Fig 3B), indicating that these scaffold explants were richer in vascular supply. Supporting these observations, mRNA levels of ANGPT1 (angiopoietin-1), which is crucial to the maturation of newly formed blood vessels, was significantly up-regulated in ad-BMP2 scaffold explants as compared to the ad-GFP control explants (Fig 4). This is in line with the in vitro results where BMP2 expression led to significant up-regulation of other pro-angiogenic molecules in ad-BMP2 BMSC (Fig 2 and S1 Fig). MicroCT analysis revealed no radiopaque bone like mass in any of the control groups, whereas a significant amount of radiopaque mass could be detected in the ad-BMP2 explants (Fig 3C-3G). Mirroring these results, histological examination of H&E stained sections showed no bony structures in the scaffold explants from any of the control groups, both at 2-and 8-weeks (Fig 3H-3M). In contrast, formation of bony structures could be identified in ad-BMP2 scaffold explant as early as 2-weeks, with more extensive bone formation by 8-weeks (Fig 3N and 3O). Numerous osteocytes like cells present in the bony structures in ad-BMP2 explants (Fig 3N and 3O, insets) suggested that the newly formed bone in the scaffolds was indeed a vital bone. Overall, these findings indicate that the adenoviral mediated expression of BMP2 in BMSC could induce formation of vital bony structures in poly(LLA-co-CL) scaffolds. Unsupervised hierarchical cluster analysis of the PCR array data showed a distinct mRNA profile of the genes related to osteogenesis and angiogenesis in ad-BMP2 explants as compared to ad-GFP explants (Fig 4), indicating a role for BMP2 in the regulation of these genes. SAM analysis further identified a number of osteogenic genes like RUNX2, ALPL, SPP1 and BMP6 to be significantly up-regulated in the ad-BMP2 explants. Collectively, these findings indicate that BMP2 enhances bone formation in poly(LLA-co-CL) scaffold seeded with ad-BMP2 BMSC by regulating the expression of key osteogenic genes such as RUNX2, ALPL, SPP1 and BMP6.
Although some studies [35,36] have suggested endochondral ossification as the mode for ectopic bone formation in the scaffold explants, the precise mechanism is currently unknown. In the current study, mRNA levels of SOX9 and its upstream regulator, FGF2, was significantly under-expressed in ad-BMP2 explants at 2 weeks (Fig 4), suggesting that BMP2 mediated bone formation in poly(LLA-co-CL) scaffold explants was not related to enhanced chondrogenic activity. Supporting this suggestion, no positivity for safranin O staining was detected in bony structures in 2-and 8-weeks ad-BMP2 scaffold explants, whereas the positive control (cranial base of mouse embryo E14) was intensely positive (Fig 5). Hence, further studies are necessary to uncover the underlying mechanism of ectopic bone formation in scaffold explants.
In conclusion, results from the current study demonstrated that adenoviral mediated BMP2 delivery significantly induced mRNA expression levels of osteogenic and pro-angiogenic molecules in vitro in BMSC grown in recently developed, biodegradable poly(LLA-co-CL) scaffolds.
Additionally, BMP2 over-expressing BMSC significantly enhanced bone formation in poly (LLA-co-CL) scaffolds in subcutaneous mouse model, with concomitant up-regulation of key osteogenic markers. Given the superior biodegradability, biocompatibility and mechanical properties of poly(LLA-co-CL) scaffolds and growing popularity of adenoviral vectors for gene therapy [37], their combination together with BMSC might, in future, be useful for bone regeneration therapy.