Chondrogenesis of Human Bone Marrow-Derived Mesenchymal Stem Cells Is Modulated by Complex Mechanical Stimulation and Adenoviral-Mediated Overexpression of Bone Morphogenetic Protein 2

Abstract

Currently available methods to treat articular cartilage defects still fail to demonstrate satisfactory outcomes for many patients. Functional tissue engineering using human bone marrow-derived mesenchymal stem cells (hMSCs) is a promising alternative approach for the treatment of these defects. This study strived to investigate the combined effect of complex mechanical stimulation and adenoviral-mediated overexpression of bone morphogenetic protein 2 (BMP-2) on hMSC chondrogenesis. hMSCs were encapsulated in a fibrin hydrogel and seeded into biodegradable polyurethane (PU) scaffolds. A novel three-dimensional transduction protocol was used to transduce cells with an adenovirus encoding for BMP-2 (Ad.BMP-2). Control cells were left untransduced. Cells were cultured for 7 or 28 days in a chondropermessive medium, which lacks any exogenous growth factors. Thereby, the in vivo situation is mimicked more precisely. hMSCs in fibrin-PU composite scaffolds were either left as free-swelling controls or mechanically stimulated using a custom-built bioreactor system that is able to generate joint-like forces. Outcome parameters measured were BMP-2 concentration within the culture medium, and biochemical and gene expression analysis. Mechanical stimulation resulted in an upregulation of chondrogenic genes. Further, glycosaminoglycan (GAG)/DNA ratios were elevated in mechanically stimulated groups. Transduction with Ad.BMP-2 led to a pronounced upregulation of the gene aggrecan and an upregulation of Sox9 message after 7 days. Furthermore, a synergistic effect in combination with mechanical stimulation on collagen 2 message was detected after 7 days. This synergistic increase was more than 8-fold if compared to the additive effect of the application of each stimulus on its own. However, BMP-2 overexpression consistently resulted in a trend toward decreased GAG/DNA ratios in both mechanical stimulated and unloaded groups.

Introduction

Hyaline, articular cartilage covers the osseous ends in diarthroidal joints. Its function is to absorb forces generated through loads, and thereby allowing for a smooth, near-frictionless motion between the articulating surfaces. Once articular cartilage is damaged, it possesses a limited intrinsic repair capacity. Currently, several different approaches are applied in clinical practice in order to treat articular cartilage defects. Examples are microfracture,¹ soft-tissue grafts such as periosteum or perichondrium,² mosaicplasty,³ and autologous chondrocyte transplantation.⁴ Unfortunately, no procedure can yet demonstrate an ability to reproducibly regenerate articular cartilage and thereby provide a long-term solution for every patient.⁵

Tissue engineering (TE) is believed to be a promising alternative for the treatment of articular cartilage defects. It consists of three major building blocks: cells, a scaffold, and stimulating factors. Human bone marrow-derived mesenchymal stem cells (hMSCs) are a favorable cell source for cartilage TE. They can be isolated from several different tissues (e.g., adipose tissue, bone marrow, muscle, articular cartilage, or synovium^6–10), and it is possible to differentiate them into different cell types, for example, chondrocytes.¹¹

A scaffold should mimic the three-dimensional environment in which cells normally reside. In TE of articular cartilage, a plethora of different scaffold systems are applied (reviewed, e.g., in Refs.^12–14). Within our group, we are focusing on a composite scaffold, consisting of a fibrin hydrogel and a polyurethane (PU) porous scaffold.^15–17 This system combines the advantages of a natural hydrogel cell carrier (e.g., homogenous cell encapsulation, preservation of the cellular phenotype, and nutrient exchange) and a synthetic porous sponge (resilience and mechanical stability).

In TE of articular cartilage, stimulating factors should preserve the chondrocyte phenotype or induce and, respectively, enhance the chondrogenesis of hMSCs. The most commonly applied stimuli are bioactive factors and mechanical forces. Bioactive factors, such as growth hormones, transcription factors, and differentiation factors, play pivotal roles in embryogenesis, and they influence a variety of functions within the human body. For cell-based cartilage TE, the most routinely applied factors include insulin-like growth factor 1,¹⁸ fibroblast growth factors,¹⁹ SRY-related HMG-box gene (SOX) 9,²⁰ and members of the transforming growth factor family such as transforming growth factor beta (TGF-β) 1,²¹ 2,²² or 3^23,24 and different bone morphogenetic proteins (BMP).^25–27

Gene transfer is an attractive alternative to the exogenous application of recombinant bioactive factors. The concept is based on the delivery of cDNA (encoding a specific transgene) to a target cell. This procedure enables the cell to produce the desired transgene. As the half-life of most bioactive factors is short, usually their repeated administration is necessary. This feature is especially obstructive in a clinical environment and can be circumvented by the application of gene transfer. Generally, gene transfer vectors can be separated into two major classes: viral and nonviral vectors. Nonviral vectors are relatively safe to apply. However, they are considered less efficient if compared to viral vectors.^28–30 Viral vectors, on the other hand, are much more efficient, but their application raises safety concerns.^29,30 Adenoviral vectors, which do not integrate into the host genome, can be used as effective tools to provide a local, although transient, production of a bioactive factor. This might be particularly helpful for the treatment of musculoskeletal defects, where only a localized and short-term expression (days to weeks) of a therapeutic factor is needed, in order to initiate a repair response or to enhance the endogenous repair. Although adenovirally transduced cells have the potential to be immunogenic, adenovirus has many advantages for proof-of-principle studies; these include low production costs, high transduction efficiency, their ability to infect both dividing and nondividing cells, high levels of transgene product, and increased safety if compared to integrating viruses (e.g., retroviruses). Last, but not least, adenoviral vectors have already been approved for clinical trials.^31,32

Healthy articular cartilage is subjected to mechanical forces on a daily basis, and it is widely accepted that they play a pivotal role for both the maintenance and the development of this tissue. Therefore, the concept of functional TE implies mechanical stimulation (e.g., compression,^33,34 fluid-flow,³⁵ shear stress,³⁶ or hydrostatic pressure²⁴) as additional stimuli. The physiological movement of the joint is kinematically very complex. Thus, the application of a single stimulus will not adequately reflect the complex in vivo situation. For that reason, a bioreactor that is able reproduce the kinematics of the joint more precisely was designed.³⁷ With this bioreactor system, several different stimuli (compression, shear, and articular fluid transport) can be applied simultaneously. This innovative system has been used in several studies that showed promising results.^20,38–40 We have also demonstrated that retroviral-mediated overexpression of BMP-2, in combination with mechanical stimulation, synergistically enhances the re-differentiation of de-differentiated bovine chondrocytes in vitro.²⁶

These encouraging results led us to develop the present study where we opted to investigate the effect of mechanical stimulation and adenoviral-mediated overexpression of BMP-2, alone and in combination, on the differentiation of monolayer-expanded hMSCs. No exogenous growth factor was applied thus mimicking the in vivo situation more precisely. Furthermore, it makes the procedure easier to apply in clinical practice, as it does not depend on the repetitive administration of an exogenous factor or the use of an integrating viral vector. It is hypothesized that mechanical stimulation will lead to the paracrine production of TGF-β1, which will induce chondrogenesis.³⁸ Transduction with Ad.BMP-2 will lead to a temporary production of BMP-2. This production is hypothesized to further enhance chondrogenesis. hMSCs were encapsulated in a clinically relevant fibrin-PU composite scaffold. They were transduced with Ad.BMP-2 in 3D (using a novel protocol for enhanced adenoviral transduction⁴¹) or left as untransduced controls. After preculture, they were either subjected to mechanical stimulation for 7 or 28 days or left as free-swelling controls. Production of biologically relevant amounts of transgene product was validated using a quantitative BMP-2 ELISA. Biochemical analyses (glycosaminoglycan [GAG], DNA) and gene expression analysis were conducted.

Materials and Methods

The following media were used during the study:

Growth medium: Minimum essential medium alpha, supplemented with 2.2 g/L NaHCO₃, 10% human hMSC certified fetal bovine serum (Thermo Fischer Scientific, Waltham, MA), 1% penicillin/streptomycin (P/S), and 5 ng/mL basic fibroblast growth factor (Peprotech, Rocky Hill, NY).

Serum-free DMEM: Serum-free high-glucose (4.5 g/L) Dulbecco's modified Eagle's medium (DMEM) containing 1 mM sodium pyruvate, 3.7 g/L NaHCO₃, and 1% P/S.

Chondropermessive medium: Serum-free DMEM, 1% insulin–transferrin selenium premix (Cyagen, Guangzhou, China), 50 μg/mL ascorbate-2-phosphate, 10⁻⁷ M dexamethasone, and 1% MEM nonessential amino acids (Millipore, Billercia, MA). P/S was replaced with 100 μg/mL of Primocin (Invivogen, San Diego, CA). This was necessary in order to prevent possible contaminations, as the bioreactor system is not a closed system. A further 5 μM of 6-aminocaproic acid was added, in order to prevent fibrin degradation.⁴²

Biodegradable PU scaffold preparation

Biodegradable, cylindrical (8×4 mm) and porous (pore size 90–300 μm) PU scaffolds were prepared as described elsewhere.⁴³ Briefly, PU was synthesized in a one-step solution polycondensation reaction from hexamethylene diisocyanate, poly(epsilon-caprolactone) diol, and 1,4:3,6-dianhydro-D-sorbitol. Next, scaffolds were prepared using the salt-leaching phase-inverse technique with sodium phosphate heptahydrate dibasic salt as porogen.

hMSC isolation

Human bone marrow aspirates were obtained after informed consent of the patients, according to the guidelines of the Federal Ethical Commission (Zürich, Switzerland). hMSCs were isolated as described before.¹⁷ Briefly, mononucleated cells were isolated through density gradient centrifugation over a ficoll cushion. Next, mononucleated cells were plated on cell culture plastic and cultured in the growth medium. hMSCs were isolated through plastic adherence. After approximately 12.77±1.11 population doublings (PD) after passage (P) 1, cells were cryopreserved.

hMSC culture

P 1 hMSCs were plated at ∼6670 cells/cm² on cell culture plastic. They were further expanded in the growth medium for another 6.28±0.4 PD (up to P 3) with medium changes every 3–4 days. Culture conditions were always 37°C, 5% CO₂, and 90% humidity in a CO₂ incubator.

Propagation of a recombinant adenoviral vector coding for BMP-2

A first-generation, E1-, E3-deleted, serotype 5 adenoviral vector carrying the cDNA for human BMP-2 was obtained from Vector Biolabs (Philadelphia, PA). The vector was then further amplified in AD293 cells (Stratagene, Santa Clara, CA) and purified over successive caesium chloride gradients. After dialysis, viral titers were calculated using standard plaque assay on AD293 cells. One infectious viral particle per cell was considered to be one multiplicity of infection (MOI)

hMSC seeding and Ad.BMP-2 transduction in 3D

Subconfluent hMSCs (a total of 19.05±1.1 PD after isolation; P 3) were embedded in a fibrin hydrogel (final concentrations: 17 mg/mL fibrinogen and 0.5 U/mL thrombin) and seeded into sterile, biodegradable, porous, cylindrical PU scaffolds at a density of 5×10⁶ cells/scaffold as previously described.¹⁷ Cells were either left untransduced (control) or transduced with Ad.BMP-2 in 3D as using an MOI of 5 (BMP-2) as previously described.⁴¹ The decision to use untransduced hMSCs as controls was based on two observations. First, they are not different from hMSCs that have been transduced with a nonbioactive transgene such as the enhanced green fluorescent protein.²⁰ Second, hMSCs that have been transduced with a control vector do not increase expression of BMP-2, TGF-β, or IGF-1.⁴⁴

After 3 days of preculture in six-well plates with 5 mL of chondropermessive medium, cells were transferred to polyetheretherketone (PEEK) control or loading holders containing a PEEK ring.

Scaffold culture and mechanical stimulation

Samples were cultured in 2.5 mL of a chondropermessive medium. The medium was changed three times per week before the application of mechanical stimulation. Scaffolds were either kept under free-swelling conditions (unloaded), or 1 h of mechanical stimulation was applied each day for 6 days per week (loaded). Mechanical stimulation was executed for either 7 or 28 days using a custom-built bioreactor system³⁷ and a ceramic hip ball with a diameter of 32 mm. Unconfined, dynamic compression strain was generated by pressing the ceramic hip ball onto the cell-seeded fibrin-PU scaffolds. Shear stresses were generated by rotating the ball around an axis perpendicular to the scaffold axis. Both stimuli were superimposed on a static offset strain of 0.4 mm. The following loading protocol was used: compression, 1 Hz 0.4–0.8 mm; rotation, 1 Hz±25°. The experiment was independently repeated with cells from 5 different donors (♀ age 65, ♀ age 78, ♂ age 55, ♀ age 73, and ♂ age 41) with triplicates for each group (Table 1).

Table 1.

Summary of the Eight Different Experimental Groups

Group	Viral transduction	Load	Time in 3D culture (+3 days preculture)
1	—	No	7
2	—	Yes	7
3	—	No	28
4	—	Yes	28
5	Ad.BMP-2	No	7
6	Ad.BMP-2	Yes	7
7	Ad.BMP-2	No	28
8	Ad.BMP-2	Yes	28

Sample collection

The medium was pooled for days 1–7, 8–14, 15–21, and 22–28. Additionally, the medium of the preculture period was collected. After 3D culture, scaffolds were vertically cut into two halves. For RNA isolation, one half scaffold was placed into 1 mL of TRI reagent and 5 μL of Polyacrylcarrier, and stored at −80°C for subsequent analysis. The other half was placed into 1 mL of 0.5 mg/mL Proteinase K and digested for 16 h at 56°C. Proteinase K was heat-deactivated for 10 min at 96°C. Finally, scaffolds were removed and samples were stored at −20°C for subsequent analysis.

BMP-2 ELISA

The amount of BMP-2 within cell culture supernatants was quantified using the quantitative Duo Set ELISA Development System for human BMP-2 (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol.

Gene expression analysis

Total RNA was precipitated using isopropanol and a high-salt precipitation solution. Reverse transcription was performed using TaqMan reagents with 1 μg of RNA in a total volume of 20 μL. Real-time PCR was performed using the 7500 real-time PCR system (Applied Biosystems, Carlsbad, CA). Gene expression was evaluated using the comparative ΔΔC_t method with 18s RNA as internal control. All oligonucleotide primers and probes (expect 18s RNA and Sox9) have previously been designed and validated using Primer Express Oligo Design software version 1.5 (Table 2). For detection of 18s RNA and Sox9 expression, the commercial predeveloped TaqMan assay reagents (Applied Biosystems, Carlsbad, CA) were used.

Table 2.

Self-Designed Forward Primers, Reverse Primers, and Probes Used for Real Time Polymerase Chain Reaction

Gene	Primer forward	Primer reverse	Probe (5′FAM/3′TAMRA)
Human Col1A1	5′-CCC TGG AAA GAA TGG AGA TGA T-3′	5′-ACT GAA ACC TCT GTG TCC CTT CA-3′	5′-CGG GCA ATC CTC GAG CAC CCT-3′
Human Col2A1	5′-GGC AAT AGC AGG TTC ACG TAC A-3′	5′-GAT AAC AGT CTT GCC CCA CTT ACC-3′	5′-CCT GAA GGA TGG CTG CAC GAA ACA TAC-3′
Human Col10A1	5′-ACG CTG AAC GAT ACC AAA TG-3′	5′-TGC TAT ACC TTT ACT CTT TAT GGT GTA-3′	5′-ACT ACC CAA CAC CAA GAC ACA GTT CTT CAT TCC-3′
Human aggrecan	5′-AGT CCT CAA GCC TCC TGT ACT CA-3′	5′-CGG GAA GTG GCG GTA ACA-3′	5′-CCG GAA TGG AAA CGT GAA TCA GAA TCA ACT-3′
Human Runx2 (cbaf1)	5′-AGC AAG GTT CAA CGA TCT GAG AT-3′	5′-TTT GTG AAG ACG GTT ATG GTC AA-3′	5′-TGA AAC TCT TGC CTC GTC CAC TCC G-3′

Biochemical analyses

The amount of sulfated GAG in both scaffolds and culture medium was quantified using the dimethylmethylene blue (DMMB) dye binding assay.⁴⁵ Chondroitin-4-sulfate from bovine trachea was used as standard. The absorbance at 535 nm was measured using the 1420 Multilabel Counter (Perkin Elmer, Waltham, MA).

The total amount of DNA was analyzed using the Hoechst 33258 dye (bisbenzimide) assay⁴⁶ with calf thymus DNA as standard. Measurements (360 nm excitation/465 nm fluorescence emission) were conducted using the 1420 Multilabel Counter (Perkin Elmer).

Statistical analyses

Statistical analyses were conducted on the complete raw data set with the SPSS software (SPSS 19, IBM, NY). Normality of each group was tested with independent-samples Kruskal–Wallis test. Levene's test of equality of error variances was conducted to test for equal variances between groups. The significance of differences between the groups was determined with a general linear model analysis of variance with a Games Howell post hoc analysis for unequal variances. For the gene expression analysis, data transformation by natural logarithm was applied. All descriptive results (except for Fig. 2) are displayed as mean±standard deviation. In order to increase clarity of Figure 2, results are displayed as mean±standard deviation only. For the BMP-2 ELISA, triplicates of three donors (total n=9) were used. For all other analyses, triplicates of five donors (total n=15) were used. Significance was defined as p≤0.05.

Results

Transduction of hMSCs in 3D with a MOI of 5 leads to the production of biologically relevant amounts of BMP-2

The feasibility of Ad.BMP-2 transduction in 3D within the fibrin-PU composite system has already been demonstrated.⁴¹ Nevertheless, it was of fundamental interest to confirm that biologically relevant amounts of BMP-2 were generated over the course of the study. Hence, BMP-2 concentration within the medium was quantitatively determined using an ELISA kit for human BMP-2. In untransduced samples, BMP-2 medium concentrations were below 0.1 ng/mL. Thus, these groups were not included in Figure 1.

FIG. 1.

BMP-2 concentration within the cell culture medium of Ad.BMP-2-transduced hMSCs that were cultured in fibrin-PU composite scaffolds. hMSCs were transduced with Ad.BMP-2 in 3D (5 MOI). Subsequently, they were cultured for up to 28 days. Either static culture conditions were applied (unloaded groups) or cells were cultured applying 1 h of mechanical simulation each day for 6 days per week (loaded groups). The medium was changed and collected three times per week. Additionally, the medium of the 3-day preculture period was collected. Transgene concentration within the culture medium was determined using a quantitative ELISA for human BMP-2. Results are displayed as mean±standard deviation of triplicates from three donors. hMSCs, human mesenchymal stem cells; BMP-2, bone morphogenetic protein 2; PU, polyurethane; MOI, multiplicity of infection.

In the unloaded group, a trend toward a decrease in BMP-2 release into the culture medium was monitored if compared to the loaded group (Fig. 1). However, this trend did not reach statistical significance. Furthermore, BMP-2 concentrations steadily increased until week 4 in the loaded group. In contrary, BMP-2 concentration started to plateau at week 2 in the unloaded group. The peak BMP-2 concentrations were 130±37.3 ng/mL in the unloaded group and 339.8±186 ng/mL in the loaded group. It is also worth noting that, starting from week 1, BMP-2 concentrations were above 100 ng/mL (a concentration commonly used in studies applying recombinant BMP-2 exogenously) for all groups except BMP-2 unloaded at week 1 (Fig. 1).

Mechanical stimulation leads to an upregulation of chondrogenic genes if compared to free-swelling controls. Transduction with Ad.BMP-2 is the predominant stimulus for the aggrecan gene and it leads to an upregulation of Sox9 message at the early time point

The effect of mechanical stimulation and transduction with Ad.BMP-2 on the gene expression profile of hMSCs was investigated. Gene expression analysis was conducted using the comparative ΔΔC_T method with 18 s RNA as internal control. The unloaded control group at day 7 was used as normalizer and therefore set to 1 (Fig. 2).

FIG. 2.

Relative gene expression of hMSCs that have been cultured in fibrin-PU composite scaffolds for 7 or 28 days. They have been transduced with Ad.BMP-2 in 3D with an MOI of 5 (BMP-2) or were left untreated (control). Subsequently, they were either cultured under static conditions (unloaded) or 1 h of mechanical stimulation per day was applied for 6 days per week (loaded). Gene expression analysis was conducted using the comparative ΔΔC_T method for genes Col I (a), Runx2 (b), Col X (c), Col 2 (d), aggrecan (e) and Sox9 (f). 18s RNA was used as internal control. The unloaded control group at day 7 was used as calibrator and therefore set to 1. Results are displayed as mean+standard deviations of triplicates from five donors. Significantly different in the control versus the Ad.BMP-2-transduced group; significantly different in the unloaded versus the loaded group; significantly different in the control unloaded versus the Ad.BMP-2-transduced loaded group (p≤0.05).

The genes collagen I (Col I) (Fig. 2a) and Runx2 (Fig. 2b) were almost unresponsive to both stimuli applied, with slight variation due to a small response seen in one of the five donors (♀ age 65). The only significant changes observed were an upregulation for both genes in Ad.BMP-2-transduced versus control samples within the day 7 unloaded group and an upregulation for Col I on day 7 when both stimuli were simultaneously applied. The gene Col X (Fig. 2c) was significantly upregulated through mechanical stimulation in all groups except Ad.BMP-2-transduced on day 7. Interestingly, if both stimuli were combined on day 7, Col X expression was not significantly different from the unloaded control group. This might indicate that transduction with Ad.BMP-2 counteracts the upregulation in Col X message, which was mediated through mechanical stimulation. On day 28, however, this trend did not reach statistical significance. Nevertheless, as the groups control loaded and BMP-2 loaded on day 7 were not significantly different from each other, this is the only one possible interpretation and would require further testing. The gene Col 2 (Fig. 2d) was massively upregulated through both mechanical stimulation and transduction with Ad.BMP-2. This upregulation was significant for all loaded groups except the Ad.BMP-2-transduced group on day 7. In the Ad.BMP-2-transduced groups, statistically significant upregulation was only detected in the unloaded group at day 7. Further, a synergistic effect of both stimuli was detected on day 7. The n-fold upregulation in Col 2 message was 2655±7417 (control loaded), 936±1785 (BMP unloaded), and 29878±86410 (BMP-2 loaded), respectively. The response of the gene aggrecan (Fig. 2e) toward the applied stimuli was the most complex. Both stimuli led to an upregulation of aggrecan expression. At the early time point (day 7) this was significant for both load and transduction with Ad.BMP-2. Yet, transduction with Ad.BMP-2 seems to be the predominant stimulus. Aggrecan message was significantly elevated in the BMP-2-loaded versus the control-loaded group. Further, combination of both stimuli did not led to an elevated response in comparison to the unloaded Ad.BMP-2-transduced group. At the late time point (day 28), the exact same trends were detectable. However, the application of mechanical stimulation did not significantly alter aggrecan gene expression. Further, transduction with Ad.BMP-2 only led to significantly higher aggrecan message in the unloaded groups. Finally, at the early time-point, transduction with Ad.BMP-2 led to a significant upregulation of Sox9 message (Fig. 2f). This was true in both the unloaded and the loaded groups. Even though it did not reach statistical significance, the same trend was observed after 28 days.

DNA content did not differ statistically between groups

Next, the DNA content was quantified using the Hoechst 33258 assay. Neither transduction with Ad.BMP-2 nor mechanical stimulation had an effect on total DNA content (data not shown).

The bulk amount of synthesized GAG is released into the culture medium

The cumulative amount of released GAG, over the course of 28 days, was quantified for each group. Furthermore, the amount of GAG retained within the scaffolds was measured after 28 days. Thereby, matrix production of hMSCs was tracked (Fig. 3).

FIG. 3.

Total amount of GAG synthesized after 28 days. MSCs were cultured in fibrin-PU composite scaffolds and either transduced with Ad.BMP-2 in 3D with an MOI of 5 (BMP-2) or left untreated (control). Subsequently, they were either cultured under static conditions (unloaded) or 1 h of mechanical stimulation per day was applied for 6 days per week (loaded). The medium was changed and collected three times per week. It was pooled for weeks 1–4. Additionally, the medium of the 3-day preculture period was collected. GAG content was quantified (both within the scaffold and within the culture medium) using the DMMB dye binding assay with chondroitin-4-sulfate as standard. Results are displayed as means±standard deviations of triplicates of five donors. Significantly different form the control-loaded group (p≤0.05). GAG, glycosaminoglycan.

A large proportion of synthesized GAG was released into the culture medium. After 28 days, only 92.5±46.5 μg (control unloaded), 88.6±45.2 μg (control loaded), 89.1±37.6 μg (BMP-2 unloaded), or 99.5±58.9 μg (BMP-2 loaded) of GAG was retained within the scaffolds. On the other hand, during the same period of time, a total of 438.4±196 μg (control unloaded), 660±177.7 μg (control loaded), 346.5±97.2 μg (BMP-2 unloaded), or 429.4±108.7 μg (BMP-2 loaded) of GAG was released into the culture medium. In summary, the amount of synthesized GAG retained within the scaffolds was in the range between 12.1%±5.6% (control loaded) and 21.75%±3.6% (BMP-2 unloaded). A trend toward an increased GAG release in loaded versus unloaded groups was detected. This trend only reached statistical difference for the control group and not for the BMP-2 group. Further, the total amount of GAG produced in loaded control group was highest among all groups and significantly different from the other groups (p≤0.05).

The GAG/DNA ratio is higher in control versus BMP-2-transduced and in loaded versus unloaded groups

Finally, GAG/DNA ratios were calculated in order to normalize the total amount of GAG produced (Fig. 4). GAG/DNA ratios (in μg/μg) were 37.3±10.8 (control unloaded), 53.3±19.3 (control loaded), 33±12.1 (BMP-2 unloaded), and 41.4±14.1 (BMP-2 loaded). A trend toward higher GAG/DNA ratios in control versus Ad.BMP-2-transduced and loaded versus unloaded groups was detected. This trend reached statistical significance in the control loaded group only where the GAG/DNA ratio was highest among all groups investigated.

FIG. 4.

Total amount of GAG normalized to DNA content between the different groups after 28 days. MSCs were cultured in fibrin-PU composite scaffolds and either transduced with Ad.BMP-2 in 3D with an MOI of 5 (BMP-2) or left untreated (control). Subsequently, they were either cultured under static conditions (unloaded) or 1 h of mechanical stimulation per day was applied for 6 days per week (loaded). Total amount of GAG (medium+scaffold) and amount of DNA was quantified. Next, GAG production was normalized to DNA content. Results are displayed as means±standard deviations of triplicates of five donors. Significantly different form the control loaded group (p≤0.05).

Discussion

During everyday life, cartilage is subjected to mechanical forces. It is widely accepted that they play an important role in both development and maintenance of this tissue. The concept of functional TE implicates mechanical stimulation and, thereby, mimics the prevailing in vivo situation more precisely. The present study opted to investigate the combined effect of complex mechanical stimulation, using a custom-built bioreactor system,³⁷ and adenoviral-mediated overexpression of BMP-2 on chondrogenesis of monolayer-expanded hMSCs. hMSCs were encapsulated in a fibrin-PU composite scaffold system and cultured in chondropermessive medium, which lacks any exogenous growth factor, for up to 28 days. The fibrin-PU system used compares favorably to pellet culture, particularly with respect to a lower level of expression of genes associated with endochondral ossification.¹⁷ The two stimuli were applied either alone or in combination. Mechanical stimulation is known to lead to the paracrine production of TGF-β1, inducing chondrogenesis of hMSCs in this fibrin-PU composite system.³⁸ Further, it has been demonstrated that shear plays an important role during hMSC chondrogenesis.⁴⁰

After transduction with Ad.BMP-2, hMSCs are able to endogenously produce BMP-2. This bioactive factor is widely utilized for both chondrogenic induction and enhancement of chondrogenesis in hMSCs.^25,47 In the present study, BMP-2 overproduction is believed to further enhance chondrogenesis. Outcome parameters measured were concentration of BMP-2 within cell culture medium, biochemical analyses (GAG, DNA, and GAG/DNA), and gene expression analysis.

After Ad.BMP-2 transduction in 3D, hMSCs were able to generate high levels of transgene product over the course of 4 weeks. This expression is far more prolonged if compared to standard two-dimensional Ad.BMP-2 transduction.⁴⁸ Furthermore, BMP-2 concentrations of 100 ng/mL or above were generated. This concentration is commonly used when BMP-2 is applied exogenously and is considered to be biologically relevant.^25,49,50 Even though statistical significance was not reached, it is worth mentioning that the application of mechanical stimulation led to a trend toward higher BMP-2 medium concentrations within the Ad.BMP-2-transduced group. The most likely explanation is that, in the unloaded group, the synthesized BMP-2 can leave the scaffold only by means of passive diffusion. In the loaded group, however, application of mechanical forces will likely squeeze out most of the BMP-2, which is retained within the scaffold. Interestingly, a similar trend is seen for GAG concentration within the cell culture medium within this study and in previous work from our group.^39,40

The DNA content did not change significantly within each group over the time course of the experiment. Further, there was no significant change in DNA content when both stimuli were applied solely or in combination. This indicates that the total cell numbers stay relatively consistent over the course of the study and that cell numbers were not influenced by either Ad.BMP-2 transduction in 3D or mechanical stimulation. In other words, neither proliferation nor cell death dominated the cellular response toward the culture environment. These observations have already been demonstrated in different studies using this fibrin-PU composite scaffold system.^38,39,41

Gene expression analysis revealed that mechanical stimulation led to an upregulation of chondrogenic genes (Col 2 and aggrecan) but also to an upregulation of the hypertrophic gene Col X. This suggests that mechanical stimulation enhances hMSC chondrogenesis but also their progression toward hypertrophy. Interestingly, at the early time-point, this hypertrophic response might be inhibited, or at least delayed, through adenoviral-mediated overexpression of BMP-2 (as Col X message in the control unloaded group was not different from the Ad.BMP-2-transduced and loaded group). For the gene Col 2, a synergic effect was detected when both stimuli were simultaneously applied. However, this was significant only for the early time point at day 7. After 28 days, supraphysiological doses of BMP-2 do not seem to synergistically enhance Col 2 gene expression if combined with mechanical stimulation. Interestingly, for the gene aggrecan, both stimuli led to an increase in gene expression if applied solely. However, when cells were subjected to both stimuli, the increase induced by Ad.BMP-2 transduction was not further enhanced by mechanical load. This suggests that overexpression of BMP-2 is the predominant stimulus for the gene aggrecan and that aggrecan is differently regulated if compared to the other genes that were investigated in this study. The same observation was made by Kupcsik et al in 2010.²⁰ Finally, transduction with Ad.BMP-2 led to a significant increase in Sox9 (master transcription factor for chondrogenesis) expression after 7 days. The same trend was observed after 28 days, even though it did not reach significance. This indicates that Ad.BMP-2 transduction is beneficial for the initiation of chondrogenesis.

Interestingly, the gene expression results within this study conflict with the GAG/DNA data where a trend towards a negative effect of Ad.BMP-2 transduction was detected. An upregulation in gene expression does not always correlate to elevated protein levels. We and others have already described a lack of direct correlation between aggrecan gene expression and GAG synthesis for hMSC.^51,52 In previous studies, we have demonstrated this using DMMB assay, safranin O staining, and immunohistochemistry for aggrecan.^20,40 We have also investigated the potential of the cells to be producing negative regulators of GAG expression, such as IL-1β, but could not detect any significant changes and have proposed that some required cofactor, present in mature chondrocytes, is lacking in chondrogenically induced MSCs.²⁰ Gene expression of Col 1 (fibroblastic marker) and Runx2 (osteoblastic marker) were almost unaffected by both stimuli. The few detected statistically significant upregulations in gene expression, after transduction with Ad.BMP-2, are so small that they are most likely not to be biologically relevant.

The quantification of total GAG revealed that the bulk amount of GAG was released into the culture medium and did not stay within the fibrin-PU scaffolds. Further, a trend toward an increased GAG production and release in mechanical stimulated samples, if compared to free-swelling controls, was observed. In the control groups, this trend reached statistical significance. Again, this has already been described within this culture system.³⁹ In general, the amount of GAG that can be retained depends on the scaffold, the mechanical environment, and the pericellular matrix. The observed results suggest that, within the fibrin-PU scaffold, the pericellular matrix is not well enough developed to retain most of the synthesized GAG. This task is further complicated by the highly porous structure and relatively large pore size of the scaffold and by the application of mechanical stimulation.

By normalizing total amount of GAG to DNA, it was demonstrated that loaded groups have a trend toward higher GAG/DNA ratios than their free-swelling counterparts. However, this trend reached significance only in the control group and not in the Ad.BMP-2-transduced group. This observation is consistent with previously published data within our group using both hMSCs³⁹ and bovine chondrocytes.²⁶ As synthesis of GAG is a qualitative measurement of hMSCs matrix production this suggests that mechanical stimulation is beneficial for hMSC chondrogenesis within this scaffold system. Comparison between Ad.BMP-2-transduced groups and untransduced controls revealed an unexpected result. A trend toward higher ratios in the untransduced control groups was observed. This is somehow surprising as overexpression of BMP-2 was believed to enhance chondrogenesis of hMSCS as already demonstrated by other investigators in different culture systems with cells from different species.^25,27,47,48 However, the reason for the conflicting results for aggrecan gene expression analysis and GAG/DNA ratios will require further investigation.

This study and further experiments with the bioreactor system might help to develop a one-step protocol that can be conducted within the operating theatre, with the aim to treat cartilage defects. Theoretically, a surgeon could harvest bone marrow during operation, isolate hMSCs through centrifugation, transduce these cells with an adenovirus expressing a desired transgene, and implant the cells immediately into the defect. The feasibility if this approach has already been demonstrated in 2004 by Pascher et al. by transducing rabbit whole BM clots intraoperatively.⁵³ Further, it would require the direct transduction of hMSCs in 3D. This has been demonstrated by recent work within our group for monolayer-expanded cells.⁴¹ The suggested protocol would no longer rely on a time- and money-consuming cell culture step and remove the need for a second operation. A potential limitation to the proposed approach would involve the release of BMP-2 from modified cells within the defect into the surrounding joint compartment, which might lead to adverse effects within tissues proximal to the defect (e.g., osteophyte formation). This would have to be investigated in an in vivo model to determine the concentration and effect of any BMP-2 release.

In summary, the present study aimed to investigate the combined effect of mechanical stimulation and adenoviral-mediated overexpression of BMP-2 on hMSC chondrogenesis. It was demonstrated that mechanical stimulation led to an upregulation of chondrogenic genes. Load also led to a small increase in expression of the hypertrophic marker Col X. This hypertrophic response seemed to be diminished, or at least delayed, when cells were transduced with Ad.BMP-2. Furthermore, mechanical stimulation increased GAG/DNA ratios in this culture system, which is an indicator for improved chondrogenesis. The effect of Ad.BMP-2 transduction in 3D on hMSC chondrogenesis was not as clear. On one hand, overexpression of BMP-2 increased aggrecan and Sox9 gene expression and further decreased Col X gene expression in the loaded groups, which suggests a positive effect on chondrogenesis and progression toward hypertrophy. Further, mechanical stimulation and transduction with Ad.BMP-2 had a synergistic effect on Col 2 message at the early time point (day 7). On the other hand, transduction with Ad.BMP-2 led to a trend toward decreased GAG/DNA ratios, suggesting that its use may be limited.

Footnotes

Acknowledgments

We would like to thank Dr. D. Eglin and M. Glarner, AO Research Institute, AO Foundation, Davos, for producing the PU scaffolds, and Dr. Andreas Goessl, Baxter Biosurgery, Vienna, for providing the fibrin components.

Disclosure Statement

No competing financial interests exist.

References

Steadman

J.R.

, Rodkey

W.G.

, Briggs

K.K.

Microfracture to treat full-thickness chondral defects: surgical technique, rehabilitation, and outcomes. J Knee Surg, 15:170. 2002.

Homminga

G.N.

, Bulstra

S.K.

, Bouwmeester

P.S.

, van der Linden

A.J.

Perichondral grafting for cartilage lesions of the knee. J Bone Joint Surg Br, 72:1003. 1990.

Szerb

, Hangody

, Duska

, Kaposi

N.P.

Mosaicplasty: long-term follow-up. Bull Hosp Jt Dis, 63:54. 2005.

Brittberg

, Lindahl

, Nilsson

, Ohlsson

, Isaksson

, Peterson

Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med, 331:889. 1994.

Getgood

, Brooks

, Fortier

, Rushton

Articular cartilage tissue engineering: today's research, tomorrow's practice?

J Bone Joint Surg Br, 91:565. 2009.

De Bari

, Dell'Accio

, Tylzanowski

, Luyten

F.P.

Multipotent mesenchymal stem cells from adult human synovial membrane. Arthritis Rheum, 44:1928. 2001.

Friedenstein

A.J.

, Gorskaja

J.F.

, Kulagina

N.N.

Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol, 4:267. 1976.

Kuroda

, Usas

, Kubo

, Corsi

, Peng

, Rose

, Cummins

, Fu

F.H.

, Huard

Cartilage repair using bone morphogenetic protein 4 and muscle-derived stem cells. Arthritis Rheum, 54:433. 2006.

Williams

, Khan

I.M.

, Richardson

, Nelson

, McCarthy

H.E.

, Analbelsi

, Singhrao

S.K.

, Dowthwaite

G.P.

, Jones

R.E.

, Baird

D.M.

, Lewis

, Roberts

, Shaw

H.M.

, Dudhia

, Fairclough

, Briggs

, Archer

C.W.

Identification and clonal characterisation of a progenitor cell sub-population in normal human articular cartilage. PLoS One, 5:e13246. 2010.

10.

Zuk

P.A.

, Zhu

, Ashjian

, De Ugarte

D.A.

, Huang

J.I.

, Mizuno

, Alfonso

Z.C.

, Fraser

J.K.

, Benhaim

, Hedrick

M.H.

Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell, 13:4279. 2002.

11.

Johnstone

, Hering

T.M.

, Caplan

A.I.

, Goldberg

V.M.

, Yoo

J.U.

In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exp Cell Res, 238:265. 1998.

12.

Balakrishnan

, Banerjee

Biopolymer-based hydrogels for cartilage tissue engineering. Chem Rev, 111:4453. 2011.

13.

Frenkel

S.R.

, Di Cesare

P.E.

Scaffolds for articular cartilage repair. Ann Biomed Eng, 32:26. 2004.

14.

Stoddart

M.J.

, Grad

, Eglin

, Alini

Cells and biomaterials in cartilage tissue engineering. Regen Med, 4:81. 2009.

15.

Grad

, Kupcsik

, Gorna

, Gogolewski

, Alini

The use of biodegradable polyurethane scaffolds for cartilage tissue engineering: potential and limitations. Biomaterials, 24:5163. 2003.

16.

Lee

C.R.

, Grad

, Gorna

, Gogolewski

, Goessl

, Alini

Fibrin-polyurethane composites for articular cartilage tissue engineering: a preliminary analysis. Tissue Eng, 11:1562. 2005.

17.

, Kupcsik

, Yao

S.J.

, Alini

, Stoddart

M.J.

Chondrogenesis of human bone marrow mesenchymal stem cells in fibrin-polyurethane composites. Tissue Eng Part A, 15:1729. 2009.

18.

Gelse

, von der

M.K.

, Aigner

, Park

, Schneider

Articular cartilage repair by gene therapy using growth factor-producing mesenchymal cells. Arthritis Rheum, 48:430. 2003.

19.

Weisser

, Rahfoth

, Timmermann

, Aigner

, Brauer

, von der

M.K.

Role of growth factors in rabbit articular cartilage repair by chondrocytes in agarose. Osteoarthritis Cartilage, 9,Suppl A:S48. 2001.

20.

Kupcsik

, Stoddart

M.J.

, Li

, Benneker

L.M.

, Alini

Improving chondrogenesis: potential and limitations of SOX9 gene transfer and mechanical stimulation for cartilage tissue engineering. Tissue Eng Part A, 16:1845. 2010.

21.

Yoo

J.U.

, Barthel

T.S.

, Nishimura

, Solchaga

, Caplan

A.I.

, Goldberg

V.M.

, Johnstone

The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am, 80:1745. 1998.

22.

Jin

X.B.

, Sun

Y.S.

, Zhang

, Wang

, Shi

T.P.

, Ju

X.D.

, Lou

S.Q.

Tissue engineered cartilage from hTGF beta2 transduced human adipose derived stem cells seeded in PLGA/alginate compound in vitro and in vivo. J Biomed Mater Res A, 86:1077. 2008.

23.

Fan

, Zhang

, Li

, Bi

, Qin

, Wu

, Hu

Gelatin microspheres containing TGF-beta3 enhance the chondrogenesis of mesenchymal stem cells in modified pellet culture. Biomacromolecules, 9:927. 2008.

24.

Miyanishi

, Trindade

M.C.

, Lindsey

D.P.

, Beaupre

G.S.

, Carter

D.R.

, Goodman

S.B.

, Schurman

D.J.

, Smith

R.L.

Dose- and time-dependent effects of cyclic hydrostatic pressure on transforming growth factor-beta3-induced chondrogenesis by adult human mesenchymal stem cells in vitro. Tissue Eng, 12:2253. 2006.

25.

Majumdar

M.K.

, Wang

, Morris

E.A.

BMP-2 and BMP-9 promotes chondrogenic differentiation of human multipotential mesenchymal cells and overcomes the inhibitory effect of IL-1. J Cell Physiol, 189:275. 2001.

26.

Salzmann

G.M.

, Nuernberger

, Schmitz

, Anton

, Stoddart

M.J.

, Grad

, Milz

, Tischer

, Vogt

, Gansbacher

, Imhoff

A.B.

, Alini

Physicobiochemical synergism through gene therapy and functional tissue engineering for in vitro chondrogenesis. Tissue Eng Part A, 15:2513. 2009.

27.

Sekiya

, Larson

B.L.

, Vuoristo

J.T.

, Reger

R.L.

, Prockop

D.J.

Comparison of effect of BMP-2, -4, and -6 on in vitro cartilage formation of human adult stem cells from bone marrow stroma. Cell Tissue Res, 320:269. 2005.

28.

Dinser

, Kreppel

, Zaucke

, Blank

, Paulsson

, Kochanek

, Maurer

Comparison of long-term transgene expression after non-viral and adenoviral gene transfer into primary articular chondrocytes. Histochem Cell Biol, 116:69. 2001.

29.

Steinert

A.F.

, Noth

, Tuan

R.S.

Concepts in gene therapy for cartilage repair. Injury, 39,Suppl 1:S97. 2008.

30.

Thomas

C.E.

, Ehrhardt

, Kay

M.A.

Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet, 4:346. 2003.

31.

Peng

Current status of gendicine in China: recombinant human Ad-p53 agent for treatment of cancers. Hum Gene Ther, 16:1016. 2005.

32.

Penny

W.F.

, Hammond

H.K.

Clinical use of intracoronary gene transfer of fibroblast growth factor for coronary artery disease. Curr Gene Ther, 4:225. 2004.

33.

Angele

, Schumann

, Angele

, Kinner

, Englert

, Hente

, Fuchtmeier

, Nerlich

, Neumann

, Kujat

Cyclic, mechanical compression enhances chondrogenesis of mesenchymal progenitor cells in tissue engineering scaffolds. Biorheology, 41:335. 2004.

34.

Buschmann

M.D.

, Gluzband

Y.A.

, Grodzinsky

A.J.

, Hunziker

E.B.

Mechanical compression modulates matrix biosynthesis in chondrocyte/agarose culture. J Cell Sci, 108,Pt 4:1497. 1995.

35.

Freed

L.E.

, Hollander

A.P.

, Martin

, Barry

J.R.

, Langer

, Vunjak-Novakovic

Chondrogenesis in a cell-polymer-bioreactor system. Exp Cell Res, 240:58. 1998.

36.

Frank

E.H.

, Jin

, Loening

A.M.

, Levenston

M.E.

, Grodzinsky

A.J.

A versatile shear and compression apparatus for mechanical stimulation of tissue culture explants. J Biomech, 33:1523. 2000.

37.

Wimmer

M.A.

, Grad

, Kaup

, Hanni

, Schneider

, Gogolewski

, Alini

Tribology approach to the engineering and study of articular cartilage. Tissue Eng, 10:1436. 2004.

38.

, Kupcsik

, Yao

S.J.

, Alini

, Stoddart

M.J.

Mechanical load modulates chondrogenesis of human mesenchymal stem cells through the TGF-beta pathway. J Cell Mol Med, 14:1338. 2010.

39.

, Yao

S.J.

, Alini

, Stoddart

M.J.

Chondrogenesis of human bone marrow mesenchymal stem cells in fibrin-polyurethane composites is modulated by frequency and amplitude of dynamic compression and shear stress. Tissue Eng Part A, 16:575. 2010.

40.

Schatti

, Grad

, Goldhahn

, Salzmann

, Li

, Alini

, Stoddart

M.J.

A combination of shear and dynamic compression leads to mechanically induced chondrogenesis of human mesenchymal stem cells. Eur Cell Mater, 22:214. 2011.

41.

Neumann

A.J.

, Schroeder

, Alini

, Archer

C.W.

, Stoddart

M.J.

Enhanced adenovirus transduction of hMSCs using 3D hydrogel cell carriers. Mol Biotechnol, 53:207. 2013.

42.

Kupcsik

, Alini

, Stoddart

M.J.

Epsilon-aminocaproic acid is a useful fibrin degradation inhibitor for cartilage tissue engineering. Tissue Eng Part A, 15:2309. 2009.

43.

Gorna

, Gogolewski

Biodegradable polyurethanes for implants. II. In vitro degradation and calcification of materials from poly(epsilon-caprolactone)-poly(ethylene oxide) diols and various chain extenders. J Biomed Mater Res, 60:592. 2002.

44.

Steinert

A.F.

, Palmer

G.D.

, Pilapil

, Noth

, Evans

C.H.

, Ghivizzani

S.C.

Enhanced in vitro chondrogenesis of primary mesenchymal stem cells by combined gene transfer. Tissue Eng Part A, 15:1127. 2009.

45.

Farndale

R.W.

, Buttle

D.J.

, Barrett

A.J.

Improved quantitation and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue. Biochim Biophys Acta, 883:173. 1986.

46.

Labarca

, Paigen

A simple, rapid, and sensitive DNA assay procedure. Anal Biochem, 102:344. 1980.

47.

Schmitt

, Ringe

, Haupl

, Notter

, Manz

, Burmester

G.R.

, Sittinger

, Kaps

BMP2 initiates chondrogenic lineage development of adult human mesenchymal stem cells in high-density culture. Differentiation, 71:567. 2003.

48.

Palmer

G.D.

, Steinert

, Pascher

, Gouze

J.N.

, Betz

, Johnstone

, Evans

C.H.

, Ghivizzani

S.C.

Gene-induced chondrogenesis of primary mesenchymal stem cells in vitro. Mol Ther, 12:219. 2005.

49.

Sailor

L.Z.

, Hewick

R.M.

, Morris

E.A.

Recombinant human bone morphogenetic protein-2 maintains the articular chondrocyte phenotype in long-term culture. J Orthop Res, 14:937. 1996.

50.

Toh

W.S.

, Yang

, Liu

, Heng

B.C.

, Lee

E.H.

, Cao

Effects of culture conditions and bone morphogenetic protein 2 on extent of chondrogenesis from human embryonic stem cells. Stem Cells, 25:950. 2007.

51.

Barry

, Boynton

R.E.

, Liu

, Murphy

J.M.

Chondrogenic differentiation of mesenchymal stem cells from bone marrow: differentiation-dependent gene expression of matrix components. Exp Cell Res, 268:189. 2001.

52.

Kisiday

J.D.

, Kopesky

P.W.

, Evans

C.H.

, Grodzinsky

A.J.

, McIlwraith

C.W.

, Frisbie

D.D.

Evaluation of adult equine bone marrow- and adipose-derived progenitor cell chondrogenesis in hydrogel cultures. J Orthop Res, 26:322. 2008.

53.

Pascher

, Palmer

G.D.

, Steinert

, Oligino

, Gouze

J.N.

, Betz

, Spector

, Robbins

P.D.

, Evans

C.H.

, Ghivizzani

S.C.

Gene delivery to cartilage defects using coagulated bone marrow aspirate. Gene Ther, 11:133. 2004.