Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 May 1.
Published in final edited form as: Osteoarthritis Cartilage. 2010 Feb 6;18(5):714–723. doi: 10.1016/j.joca.2010.01.008

Spatial Regulation of Human Mesenchymal Stem Cell Differentiation in Engineered Osteochondral Constructs: Effects of Pre-Differentiation, Soluble Factors and Medium Perfusion

Warren L Grayson 1,†, Sarindr Bhumiratana 1,†, P H Grace Chao 2, Clark T Hung 1, Gordana Vunjak-Novakovic 1,*
PMCID: PMC2862865  NIHMSID: NIHMS187218  PMID: 20175974

Abstract

Objective

The objective of the study was to investigate the combined effects of three sets of regulatory factors: cell pre-differentiation, soluble factors and medium perfusion on spatial control of human mesenchymal stem cell (hMSC) differentiation into cells forming the cartilaginous and bone regions in engineered osteochondral constructs.

Design

Bone-marrow derived hMSCs were expanded in their undifferentiated state (UD) or pre-differentiated (PD) in monolayer culture, seeded into biphasic constructs by interfacing agarose gels and bone scaffolds and cultured for 5 weeks either statically (S) or in a bioreactor (BR) with perfusion of medium through the bone region. Each culture system was operated with medium containing either chondrogenic supplements (C) or a cocktail (Ck) of chondrogenic and osteogenic supplements.

Results

The formation of engineered cartilage in the gel region was most enhanced by using undifferentiated cells and chondrogenic medium, whereas the cartilaginous properties were negatively affected by using pre-differentiated cells or the combination of perfusion and cocktail medium. The formation of engineered bone in the porous scaffold region was most enhanced by using pre-differentiated cells, perfusion and cocktail medium. Perfusion also enhanced the integration of bone and cartilage regions.

Conclusions

  1. Pre-differentiation of hMSCs before seeding on scaffold was beneficial for bone but not for cartilage formation.

  2. The combination of medium perfusion and cocktail medium inhibited chondrogenesis of hMSCs.

  3. Perfusion improved the cell and matrix distribution in the bone region and augmented theintegration at the bone-cartilage interface.

  4. Osteochondral grafts can be engineered by differentially regulating the culture conditions in the two regions of the scaffold seeded with hMSCs (hydrogel for cartilage, perfused porous scaffold for bone).

INTRODUCTION

The high incidence of cartilage defects due to injury and disease, coupled with the low regenerative capacity of cartilage, often necessitates surgical treatment to repair the sites of injury. One method of the treatment of cartilage injuries is to replace the full-thickness defects with osteochondral plugs, where the bony region anchors the graft and facilitates the integration of the cartilage graft with the host tissue [1]. This approach is limited by insufficient availability of autograft material, donor site morbidity and the risks of disease-transmission associated with the allografts. To alleviate these limitations, it is possible to engineer biological replacements of cartilage and bone tissues using multi-potent human mesenchymal stem cells (hMSCs). There are however some inherent technical and scientific challenges to growing osteochondral constructs from hMSCs where bone and cartilage should ideally develop in tandem to facilitate functional integration of the two tissues. Most importantly, spatial regulation of hMSC differentiation is necessary to guide hMSCs down the chondrogenic or osteoblastic lineages in vitro and generate discreet cartilaginous and osseous regions within a single construct.

Previously, hMSCs were separately differentiate into bone and cartilage constructs before physically apposing and suturing these constructs into biphasic units [2]. This approach included considerable manipulation of cultured tissues under sterile conditions and resulted in poor integration between the two tissue constructs. An alternative approach exposed the mouse MSCs to chondrogenic or osteogenic supplements during their monolayer expansion, and sequentially encapsulated the cells in gel substrates to create two distinct but contiguous constructs [3–5]. After 12 weeks in the dorsum of nude mice, the bone and cartilage remained grossly distinct with apparent regions of integration at the interface. While these results are clearly promising, the use of in vivo mouse cultivation limits the clinical utility of these constructs. A distinct ‘pre-induction’method has also been reported for hMSCs [6]. The resulting composites were cultured in a medium containing a cocktail of osteogenic and chondrogenic supplements for another 5 weeks. Encouraging results for bone-cartilage integration were achieved in this study with the main drawback being the long (10 week) cultivation times.

Bioreactors can be utilized to provide biophysical stimulation and improved nutrient transfer to the cells on scaffolds, and enhance their functional assembly into tissues. Several bioreactor designs have been used for engineering bone [7–9] and cartilage [10, 11] from human mesenchymal stem cells. Most recently, we have utilized a perfusion bioreactor for the cultivation of bone constructs [12], as the interstitial flow facilitated cell growth and differentiation, leading to the deposition of mineral and extracellular matrix (ECM) proteins throughout the bone region. We adapted this perfusion bioreactor for use in the present study.

For scaffold materials, we selected those shown to be optimal in previous studies of cartilage and bone tissue engineering: Agarose gel was used for the cartilage phase, as it has been demonstrated that it yields the best mechanical properties of engineered cartilage among all materials studied with immature chondrocytes [13]. Likewise, decellularized bone was selected as a scaffold for the bone region, as it provides osteo-inductive architecture, mechanical properties and biochemical composition [14, 15]. Additionally, in our numerous previous studies, agarose and decellularized bone have been used independently and in combination as scaffolds for cartilage, bone and osteochondral constructs [16, 17]. Therefore, we hypothesized that the agarose-bone scaffolds, alone or in combination with molecular and cellular parameters and biophysical stimuli, can provide hMSCs with differential cues necessary to induce spatially confined chondrogenesis (agarose) and osteogenesis (bone) while facilitating interfacial communication between the two developing tissues. To test this hypothesis, we explored the effects of three sets of experimental variables on hMSC differentiation in biphasic constructs: (i) supplementation of chondrogenic factors vs. a cocktail of chondrogenic and osteogenic factors to culture medium), (ii) pre-differentiation of hMSCs, (iii) medium perfusion (interstitial flow).

MATERIALS AND METHODS

Human Mesenchymal Stem Cells

Cryopreserved passage 2 bone marrow-derived hMSCs were kindly provided by Dr. Arnold Caplan, after isolation using previously described protocols [18]. Cells were expanded for one passage (P3) in control medium (low glucose DMEM supplemented with 10% FBS, 1ng/ml FGF and 1% antibiotics). Passage 4 cells were split into three groups, and cultured for one more passage (8 ± 1 days) in (i) control medium (undifferentiated hMSCs), osteogenic medium (osteo-induced hMSCs), and chondrogenic medium (chondro-induced hMSCs). Undifferentiated and osteo-induced hMSCs were plated at a density of 5,000 cells/cm2 while the chondro-induced hMSCs were plated at 60,000 cells/cm2 [19, 20]. Osteogenic medium was low glucose DMEM supplemented with 10% FBS, 1% antibiotics, 10 mM sodium-β-glycerophosphate, 100 nM dexamethasone, and 50 µg/ml ascorbic acid-2-phosphate. Chondrogenic medium was high glucose DMEM supplemented with 10 ng/ml TGF-β3, 100 nM dexamethasone, 50 µg/ml ascorbic acid-2-phosphate, 100 µg/ml sodium pyruvate, 40 µg/ml proline, 1% ITS+ mix and 1% antibiotics. Passage 4 cells were then used for experiments.

Pellet Cultures

Cells cultured under the various conditions were resuspended in their specific medium, counted, and 2.5 × 105 cells per aliquot were centrifuged to form pellets. Pellets from each group (undifferentiated, osteo-induced and chondro-induced) were cultured in control, osteogenic, and chondrogenic medium, resulting in 9 different conditions. Three pellets for each condition were cultured for 4 weeks in 1 ml of medium with medium change three times per week.

Decellularized Bone Scaffolds

Decellularized bone was obtained as previously described [12].In brief, trabecular bone plugs (4 mm in diameter) were cored from the subchondral region of carpometacarpal joints of 2 to 4 month old cows. They were washed to remove marrow and rinsed in PBS with 0.1% EDTA for 1 hr at room temperature. This was followed by sequential washes in hypotonic buffer, detergent and enzymatic solution. Deceullularized bone plugs were rinsed in PBS, freeze-dried, and cut to 5 mm lengths to yield cylinders 4 mm diameter×5 mm high. The weights and dimensions of each plug were measured and used to calculate scaffold density. Scaffolds within the range of 0.30 – 0.40 mg/cm3 were used. Scaffolds were sterilized in 70% ethanol, washed in PBS and incubated in culture medium prior to seeding cells. The distribution of bone scaffolds with different densities was randomized.

Biphasic Scaffolds

Bone scaffolds were seeded as previously described [12]. In brief, scaffolds were blot-dried and seeded with 1.8 × 106 cells suspended in 40 µl of media (45 × 106 cells/mL). The scaffolds were flipped every half hour and 10 µl media were added. After 2 hours, 5 ml of media were added and scaffolds were incubated overnight. Agarose gels were made the following day by mixing equal volumes of 4% agarose solution and cell suspension to yield 25 × 106 cells/ml in 2% agarose. This was pipetted into cylindrical wells (4 mm ∅ × 2.5 mm height) in PDMS molds. Bone scaffolds seeded on the previous day were overlaid allowing a penetration depth of 500 µm of gel into the bone scaffold. The agarose was then allowed to solidify at room temperature to complete the formation of biphasic constructs. The seeded scaffolds were cultured for 4 days under static conditions to allow cell attachment, prior to applying medium perfusion.

Perfusion Bioreactor

A perfusion bioreactor developed in our laboratory and described previously [12] was used for cultivating biphasic tissue constructs. The constructs were positioned with the bone region secured in wells and agarose region in the reservoir. The bone region protruded 1 mm from the well. Culture medium was pumped axially upwards through the interstices of the bone region, and out into the medium reservoir. The agarose region was completely submerged in culture medium in the reservoir, and was not perfused (Fig. 1B). The medium flowed through the bone scaffolds at a superficial velocity of 400 µm/s [12].

Figure 1. Experimental Design.

Figure 1

Open in a new tab

(A) Biphasic scaffold made by interfacing agarose and trabecular bone scaffolds. (B) Perfusion bioreactor for cultivation of biphasic scaffolds. Enlarged view shows predicted path of medium flow through the scaffolds and through the sides into the reservoir. (C) Schematic of experimental design.

Experimental Design

The experiment was set-up to test the effects of three distinct variables: cell pre-differentiation, medium perfusion and medium supplements (Fig. 1C). To evaluate the influence of cell pre-differentiation, constructs were made from undifferentiated hMSCs (UD) or hMSCs pre-differentiated to chondrocytes and osteoblasts (PD). To evaluate the effects of perfusion, each of the cell groups was cultured either in perfused bioreactor (BR) or statically (S). To evaluate the effects of medium supplements, each of the cell-bioreactor groups was cultured in either chondrogenic (C) or cocktail (Ck) medium (high glucose DMEM containing both chondrogenic and osteogenic supplements: 10% FBS, 10 mM sodium-β-glycerophosphate, 100 nM dexamethasone, 50 µg/ml ascorbic acid-2-phosphate, 10 ng/ml TGF-β3, 100 µg/ml sodium pyruvate, 40 µg/ml proline, 1% ITS+ mix and 1% antibiotics). As a result, we obtained six experimental groups as shown in Fig. 1C. This way, the effects of cell predifferentiation were compared for two culture conditions and two sets of medium supplements (UD-BR-C/Ck vs. PD-BR-C/Ck). Likewise, the effects of biophysical stimulation via medium perfusion were determined for both differentiated and undifferentiated cells, and both media compositions (UD-S-C/Ck vs. UD-BR-C/Ck). Finally, chondrogenic medium (which already contained dexamethasone and ascorbic acid) and the cocktail medium (containing additional osteogenic supplements) were compared for both cell groups and both sets of culture conditions (UD/PD-BR-C vs. UD/PD-BR-Ck). Data were obtained in two independent series of experiments.

Micro Computerized Tomography (µCT)

µ-CT imaging was performed using a modified protocol from Liu et al [21]. Constructs were aligned along their axial direction in a 15 mL centrifuge tube and stabilized with wet gauze. The tube was clamped in the specimen holder of a vivaCT 40 system (SCANCO Medical AG, Basserdorf, Switzerland). Constructs were scanned at 21 µm isotropic resolution, and the bone volume (BV) was obtained by global thresholding technique that detected only the mineralized tissue.

Mechanical Testing

The equilibrium Young’s modulus of the cartilage region were measured in unconfined compression using a modification of an established protocol [13]. An initial tare load of 0.02 N was applied. This was followed by a stress-relaxation step where the specimens were compressed to 10% strain of the cartilage region at a ramp velocity of 0.05%/s and maintained at that position for 1800 seconds. The Young’s modulus was obtained from the equilibrium forces measured at 10% strain.

Biochemical Assays

The gel and bone regions of three constructs per group were separated along the flat surface of the bone scaffold and the wet weight of the gel and bone regions determined. The samples were stored at −20 °C until assay. For analysis, the gel regions were digested in 1mL proteinase K solution at 50 °C. The bone regions were placed in 100 µL of proteinase K digestion buffer at 50 °C. DNA content was determined using the Picogreen assay (Molecular Probes, OR). The sulfated GAG (s-GAG) content of the extracts was determined using the 1,9-dimethylmethylene blue (DMMB) dye colorimetric assay withchondroitin-6-sulfate as a standard.

Histology & Immunohistochemistry

Constructs were washed in PBS and fixed in 10% formalin, decalcified with Immunocal solution, embedded in paraffin, sectioned into 5 µm slices and stained with haematoxylin & eosin and Alcian Blue (GAG). Samples were also immunohistochemically stained for collagens I & II and bone sialoprotein (BSP) as previously described [12].

Statistical Analysis

Pair-wise comparisons of results were carried out using multi-way ANOVA followed by Tukey’s post-hoc analysis using STATISTICA software. P < 0.05 was considered as significant.

RESULTS

Characterization of Undifferentiated and Pre-differentiated hMSCs

Undifferentiated hMSC smaintained their fibroblast-like morphologies throughout the 2D cultivation period while osteo-induced hMSCs grew faster and appeared thinner but less elongated and less aligned to each other. Despite the high seeding densities and chondro-inductive medium, hMSCs were unable to adopt a spherical morphology, instead becoming broad and flat (Fig. 2 A–C). Pre-differentiation of hMSCs by one-week monolayer cultivation in chondrogenic or osteogenic medium did not improve their subsequent 3D differentiation along the same lineage (osteo→osteo or chondro→chondro) during pellet culture (Fig. 2 D–I). Under osteogenic conditions, pellets from all three groups stained positively for mineral deposition (Fig. 2 D–F) and expressed bone sialoprotein (not shown). Under chondrogenic conditions, pellets from all three groups expressed GAG and formed lacunar structures. Both osteogenic and chondrogenic pre-differentiation appeared to decrease their subsequent chondrogenic potential in pellet culture relative to undifferentiated hMSCs. Pellets formed by undifferentiated hMSCs were approximately twice the size of those in other groups, and stained more intensely for GAG (Fig. 2 G–I) and collagen Type II (not shown).

Figure 2. Cell Differentiation Studies.

Figure 2

Open in a new tab

(A–C) Morphology of hMSCs cultured using (A) expansion medium, (B) osteogenic supplements and (C) chondrogenic supplements. (D–F) Von Kossa staining of pellets cultured under osteogenic conditions for four weeks. (G–I) Alcian blue staining of pellets cultured under chondrogenic conditions for four weeks.

Cartilage Region in Biphasic Constructs

The highest GAG contents (expressed as a fraction ofwet weight) were achieved for UD-S-C constructs (Fig. 3A). Biochemical and mechanical properties of engineered cartilage were significantly affected by the combination of cocktail medium and perfusion, but not by either stimulus alone. For example, GAG content was statistically similar when comparing UD-S-C (p = 0.61%), UD-BR-C (p = 0.59%) and UD-S-Ck (p = 0.53%) constructs but there was a decrease in GAG contents for UD-BR-Ck group (p = 0.02%) where both bioreactor cultivation and cocktail medium were used. Pre-differentiated hMSCs did not express significant quantities of GAG (< 0.1%). The equilibrium Young’s modulus values reflected the trend of the GAG content (Fig. 3B). Similar values were obtained for UD-S-C (18 kPa), UD-BR-C (22 kPa) and UD-S-Ck (15kPa) while the modulus of UD-BR-Ck, PD-BR-C and PD-BR-Ck were all less than 10 kPa. The normalized DNA values were consistent with these findings as the DNA contents of UD-S-C, UD-BR-C and UD-S-Ck were approximately 3 times higher than the other three groups (Fig. 3C). It should be pointed out that the trend for GAG content remained the same when normalized to DNA content, but the differences observed between groups for equilibrium modulus disappears. µ-CT data indicated mineral deposition in the gel regions of UD-S-Ck and UD-BR-Ck groups only (Fig. 3D (inset)).

Figure 3. Quantitative Properties of Cartilage Region.

Figure 3

Open in a new tab

(A) GAG content of gels normalized by wet weight. (B) Equilibrium modulus. (C) DNA content normalized by wet weight. (D) Bone volume (BV) of gels measured by micro-CT. (n=3; *p<0.05; **p<0.001) (E) Table of GAG expression and equilibrium modulus values normalized to DNA (n=3; * p<0.05 as compared to the experimental groups using the same culture medium).

H&E staining indicated differences in cell distributions throughout the gel regions between the experimental groups. Under static conditions (UD-S-C & UD-S-Ck), the central regions closer to the bony substrates were less densely populated than the edges, an effect that can be attributed to the diffusional limitations of nutrient transport (Fig. 4, first row). This pattern was less pronounced for the UD-BR groups (UD-BRC & UD-BRCk) where medium was perfused directly into the lower face of the gel construct. Decreased matrix deposition was again evident in PD-BR constructs (PD-BR-C & PD-BR-Ck).

Figure 4. Representative images of Immunohistochemical Properties of Cartilage Region.

Figure 4

Open in a new tab

First Row: H&E staining. Second Row: Alcian blue staining for GAG content. Third Row: Collagen II expression. Fourth Row: Collagen X expression. Fifth Row: Bone sialoprotein expression.

Qualitative analysis of GAG expression (Fig. 4, second row) correlated with the quantitative data presented in Fig. 3 for different experimental groups. The patterns of collagen II expression (Fig. 4, third row) exactly matched those of GAG expression, confirming chondrogenic differentiation. GAG and collagen type II were expressed strongly in UD-SC, UD-S-Ck and UD-BR-C groups. The UD-BR-Ck group expressed less GAG and collagen type II, but exhibited considerable mineral formation (not shown), which corresponded to the µ-CT data (Fig. 3D inset). Collagen X stains were evident in the gel regions of constructs cultured with undifferentiated hMSCs, and were particularly intense when cocktail medium was used (UD-S-Ck& UD-BR-Ck) (Fig. 4, fourth row). Further histological assessment demonstrated that cells in the UD-BR-Ck group expressed bone sialoprotein (BSP, a marker of bone differentiation) at the boundaries of the gel region. Slight BSP expression was also observed in the PD-BR-Ck-group, but all other groups were BSP negative (Fig. 4, fifth row).

Bone Region in Biphasic Constructs

Quantitative evaluation of DNA content within the bone region of the constructs demonstrated little variation between the experimental groups (data not shown). The large variability in initial mineral content (obtained from µ-CT) made it difficult to rigorously evaluate relative changes in mineralization based on cultivation conditions. However, histological analysis showed interesting effects of cultivation conditions on the properties of bone regions. H&E staining demonstrated that constructs cultured under static conditions resulted in cell growth and matrix deposition up to approximately 1 mm from the periphery of the bone scaffolds leaving the central regions largely unpopulated. In contrast, medium perfusion resulted in more uniform cell distribution and matrix accumulation throughout the pore spaces of the bone scaffolds (Fig. 5, first row). GAG deposition was evident for undifferentiated cells cultivated in chondrogenic medium (Un-S-C and Un-BR-C), and rather low in the groups cultivated in cocktail medium. Again, the combined effects of perfusion and cocktail medium were evident: Un-S-Ck groups also showed patches of light GAG staining while the Un-BR-Ck groups showed no GAG staining (Fig. 5, second row).

Figure 5. Representative images of Immunohistochemical Properties of Bone Region.

Figure 5

Open in a new tab

First Row: H&E staining. Second Row: Alcian blue staining for GAG content. Third Row: Collagen I expression in gels. Fourth Row: Bone sialoprotein expression.

Osteogenic pre-differentiation alone was not sufficient to eliminate chondrogenic differentiation in the bone region: PD-BR-C showed light GAG staining, but this was eliminated in the presence of osteogenic supplements (PD-BR-Ck group). Undifferentiated groups cultured in chondrogenic medium (Un-S-C & Un-BR-C) did not exhibit collagen I or BSP staining. Interestingly, the pre-induced osteoblasts cultured in chondrogenic medium (PD-BR-C) expressed BSP and collagen I throughout the scaffold. Cocktail supplements without flow were sufficient to elicit minimal BSP expression (Un-S-Ck), but the expression of BSP and collagen I increased with perfusion (Un-BR-Ck & PD-BR-Ck) (Fig. 5, third & fourth rows).

Integration of Cartilage and Bone Regions

The integration in the region where agarose gel penetrated into the porous bone scaffold was assessed by examining the properties of cell and tissue matrix. Constructs cultured under static conditions demonstrated the presence of cells only at the outer edges of the integration zone (Fig. 6A) but no cells were observed in the inner regions (Fig. 6A, B). To evaluate the effect of perfusion on integration, the integration zone for the Alcian Blue-stained Un-BR-C construct is shown in Fig. 6C. Cells were present throughout the integration zone in these constructs. Interestingly, while cells throughout the entire region are expressing GAG, indicative of chondrogenic differentiation, there is evidence of morphological differences. Cells embedded completely within the agarose gel are spherical and reside within lacunae, while cells within close proximity to the bone scaffold exhibit more fibrous morphologies (Fig. 6D). It was noted that the fibrous regions did not form as a uniformly horizontal layer throughout the integration zone hence there was no evidence of a discrete layer separating bone from cartilage.

Figure 6. Integration of Bone and Cartilage Regions.

Figure 6

Open in a new tab

(A) Integration region of static constructs (UD-S-C) is mostly acellular. (B) High magnification image of region indicated by box in (A) showing different morphologies in gel and scaffold regions and minimal GAG expression. (C) Integration between cartilage and bone is enhanced under bioreactor conditions. (D) High magnification image of region indicated by box in (C) shows high matrix production in central regions of integration zone. Distinct spherical morphology in gel is indicated by * whereas the elongated, fibrous morphology is evident in regions close to bone by arrows (↑).

DISCUSSION

Early tissue engineering studies investigated the feasibility of forming osteochondral constructs from two separate regions containing differentiated chondrocytes and osteoblasts, and using the resulting composites for the repair of focal defects [22, 23]. However, differentiated cells do not provide a practical clinical option as they are limited in supply, their harvest is associated with morbidity and the danger of secondary joint disease, and it is difficult to procure healthy cells for therapeutic use. One attractive cell source for cartilage and bone tissue engineering are hMSCs because of their innate capability to make osteochondral tissues. Numerous reports have demonstrated the feasibility of growing either bone-like or cartilage-like tissues from hMSCs, although the functional outcomes for cartilage have never been comparable to those achieved using immature chondrocytes. Still, growing ‘complex’ tissues is dependent upon the ability to provide the hMSCs with the appropriate chondrogenic or osteogenic cues in a site-specific manner. Therefore, the primary objective of our study was to evaluate various experimental conditions which could enable the simultaneous development of integrated, yet distinct, bone and cartilage tissues from a homogenous hMSC population.

hMSC Pre-differentiation Reduces Subsequent Chondrogenesis

We tested the hypothesis that pre-differentiating the cells would turn on lineage-specific genes, expedite the formation of cartilage and bone regions within the osteochondral plugs and predispose cells to respond to either osteogenic or chondrogenic stimuli in cocktail medium, but not both. The results of pellet culture experiments, which were used to determine the phenotypic stability of cells after differentiation, demonstrated that pre-differentiated hMSCs could still respond to opposing stimuli even though the effect was attenuated relative to undifferentiated hMSCs (Fig. 2). It was also found somewhat counter-intuitively that the chondrogenic pre-differentiation reduced the hMSCs’ potential to form cartilage matrix in 3D culture. A similar approach was taken – and shown to be successful – using rat MSCs [3]. This difference may reflect an inherent distinction in the capacity of rat versus human cells under similar cultivation conditions. In retrospect, this is not surprising, because the chondrogenic potential of hMSCs is heavily dependent upon their ability to maintain a spherical morphology [24]. By providing chondrogenic stimuli during monolayer culture, the cells were essentially presented with conflicting stimuli, which negatively affected subsequent development.

Chondrogenesis is Reduced by the Combination of Flow and Cocktail Medium

Undifferentiated hMSCs cultured statically in agarose gels (UD-S-C) produced 0.6% GAG and had Young’s moduli of approximately 20 kPa. These values are considerably lower than those observed for constructs grown with immature bovine chondrocytes [25], but are in the range of values reported for bovine MSCs [11]. These results underscore the inherent difficulty in elucidating and providing the appropriate cues for directing stem cells into mature chondrocytes in vitro. Yet, the scaffold-bioreactor system provided critical insights into the ability of hMSCs to integrate multiple stimuli in cell-fate decisions. Notably, relative to the UD-S-C group, changing to perfusion only (Un-BR-C) or cocktail medium only (UD-S-Ck) had no significant effect on tissue composition content and mechanical stiffness, while both parameters were negatively affected by the combination of cocktail medium and perfusion (UD-BR-Ck) (Figure 3). Collagen X stains (Figure 4) suggested hypertrophy of hMSC-derived chondrocytes in all groups where undifferentiated hMSCs were used to form the cartilage layer, with the strongest expression in the cultures where cocktail medium was used. This is consistent with the notion that hypertrophy is generally associated with MSC chondrogenesis.

Pre-Differentiation, Flow and Cocktail Medium Provide the Best Osteogenic Conditions

Our study tested whether the biochemical environment of trabecular bone matrix combined with flow-induced shear stress through the constructs may have been sufficiently osteo-inductive for new bone formation. It was evident that the soluble factors in the medium had the most potent effects on cell differentiation: extensive GAG and collagen II staining were evident throughout the bone regions of UD-S-C and UD-BR-C groups (Fig. 5) indicating that the combination of flow and the trabecular bone biochemistry was insufficient for osteogenesis. Cells in the scaffold did however appear to be predisposed to an osteogenic phenotype since cultivation in cocktail medium (UD-S-Ck and UD-BR-Ck) virtually eliminated GAG and collagen II staining while upregulating BSP expression. Perfusion considerably improved cell distribution throughout the bone regions of the constructs (Fig. 5). There was also decreased collagen II and increased BSP expression in the UD-BR-Ck relative to UD-S-Ck groups indicating flow might play a role in further stimulating osteogenesis of hMSCs at the expense of chondrogenesis. However, further investigations are required to elucidate whether it is due to biophysical stimulation or improved cell-cell communication associated with high cell densities.

Spatial Regulation of hMSC Differentiation and Bone-Cartilage Integration

In this study we have explored the combination of in vitro culture conditions that give rise to suitable chondrogenic or osteogenic differentiation of hMSCs in pre-determined regions of osteochondral constructs. Static culture of undifferentiated hMSCs in chondrogenic medium elicited the best cartilage properties while perfusion culture of pre-differentiated osteoblasts or undifferentiated hMSCs with cocktail medium provided the best osteogenic response.

The collected data support the notion that an osteochondral bioreactor should contain two discrete compartments [26, 27] which would enable cells in either region to be exposed to optimal stimuli, including different medium compositions. In comparison to the previous in vitrostudies [2, 6], we used an experimental design that allowed the cartilage and bone phases to develop in tandem. The interface formed in biphasic constructs remains different from that in a native issue where the bone-cartilage cross-talk occurs at a clearly demarcated zone of mineralized cartilage. Future studies may focus on clarifying the conditions required to recapitulate the native interface and investigating the heterogenous cell-cell communication in this region. It is also interesting to note that the gel region of this group (UD-BR-Ck) constructs strongly expressed GAG (Fig. 3A) while simultaneously depositing considerable amounts of mineral (Fig. 3D). Future studies may determine whether hMSCs are capable of expressing the osteogenic and chondrogenic phenotypes simultaneously – similar to the immature cell phenotype which exists during intramembranous ossification [28] - or whether these are two distinct populations which co-exist in the gel regions. Such studies may provide insight into mechanisms of stem cell differentiation and cellular interactions at the osteochondral interface.

In conclusion, the study demonstrated the feasibility of engineering biphasic tissue constructs using biphasic scaffolds and perfusion bioreactors enabling spatial regulation of hMSC differentiation. Pre-differentiation of hMSCs in monolayer culture was beneficial for bone tissue development, but not cartilage. It was shown that undifferentiated hMSCs are capable of integrating signals from biological factors and perfusion stimuli into decisions to differentiate into chondrocytes or osteoblasts, and perfusion culture considerably enhanced tissue development and improved integration of bone and cartilage tissues. Future studies will focus on the development of an osteochondral bioreactor, and facilitation of the biological communication to foster functional bone-cartilage interface.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Hangody L, Vasarhelyi G, Hangody LR, Sukosd Z, Tibay G, Bartha L, et al. Autologous osteochondral grafting-technique and long-term results. Injury-International Journal of the Care of the Injured. 2008;39:S32–S39. doi: 10.1016/j.injury.2008.01.041. [DOI] [PubMed] [Google Scholar]
  • 2.Augst A, Marolt D, Freed LE, Vepari C, Meinel L, Farley M, et al. Effects of chondrogenic and osteogenic regulatory factors on composite constructs grown using human mesenchymal stem cells, silk scaffolds and bioreactors. Journal of the Royal Society Interface. 2008;5:929–939. doi: 10.1098/rsif.2007.1302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Alhadlaq A, Elisseeff JH, Hong L, Williams CG, Caplan AI, Sharma B, et al. Adult stem cell driven genesis of human-shaped articular condyle. Annals of Biomedical Engineering. 2004;32:911–923. doi: 10.1023/b:abme.0000032454.53116.ee. [DOI] [PubMed] [Google Scholar]
  • 4.Alhadlaq A, Mao JJ. Tissue-engineered neogenesis of human-shaped mandibular condyle from rat mesenchymal stem cells. Journal of Dental Research. 2003;82:951–956. doi: 10.1177/154405910308201203. [DOI] [PubMed] [Google Scholar]
  • 5.Alhadlaq A, Mao JJ. Tissue-engineered osteochondral constructs in the shape of an articular condyle. Journal of Bone and Joint Surgery-American Volume. 2005;87A:936–944. doi: 10.2106/JBJS.D.02104. [DOI] [PubMed] [Google Scholar]
  • 6.Tuli R, Nandi S, Li WJ, Tuli S, Huang XX, Manner PA, et al. Human mesenchymal progenitor cell-based tissue engineering of a single-unit osteochondral construct. Tissue Engineering. 2004;10:1169–1179. doi: 10.1089/ten.2004.10.1169. [DOI] [PubMed] [Google Scholar]
  • 7.Marolt D, Augst A, Freed LE, Vepari C, Fajardo R, Patel N, et al. Bone and cartilage tissue constructs grown using human bone marrow stromal cells, silk scaffolds and rotating bioreactors. Biomaterials. 2006;27:6138–6149. doi: 10.1016/j.biomaterials.2006.07.015. [DOI] [PubMed] [Google Scholar]
  • 8.Meinel L, Karageorgiou V, Fajardo R, Snyder B, Shinde-Patil V, Zichner L, et al. Bone tissue engineering using human mesenchymal stem cells: Effects of scaffold material and medium flow. Annals of Biomedical Engineering. 2004;32:112–122. doi: 10.1023/b:abme.0000007796.48329.b4. [DOI] [PubMed] [Google Scholar]
  • 9.Holtorf HL, Jansen JA, Mikos AG. Flow perfusion culture induces the osteoblastic differentiation of marrow stromal cell-scaffold constructs in the absence of dexamethasone. Journal of Biomedical Materials Research Part A. 2005;72A:326–334. doi: 10.1002/jbm.a.30251. [DOI] [PubMed] [Google Scholar]
  • 10.Meinel L, Hofmann S, Karageorgiou V, Zichner L, Langer R, Kaplan D, et al. Engineering cartilage-like tissue using human mesenchymal stem cells and silk protein scaffolds. Biotechnology and Bioengineering. 2004;88:379–391. doi: 10.1002/bit.20252. [DOI] [PubMed] [Google Scholar]
  • 11.Mauck RL, Byers BA, Yuan X, Tuan RS. Regulation of cartilaginous ECM gene transcription by chondrocytes and MSCs in 3D culture in response to dynamic loading. Biomechanics and Modeling in Mechanobiology. 2007;6:113–125. doi: 10.1007/s10237-006-0042-1. [DOI] [PubMed] [Google Scholar]
  • 12.Grayson WL, Bhumiratana S, Cannizzaro C, Chao GP, Lennon D, Caplan AI, et al. Effects of initial seeding density and fluid perfusion rate on formation of tissue-engineered bone. Tissue Eng A. 2008;14:1809–1820. doi: 10.1089/ten.tea.2007.0255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Mauck RL, Soltz MA, Wang CC, Wong DD, Chao P-HG, Valhmu WB, et al. Functional tissue engineering of articular cartilage through dynamic loading of chondrocyte-seeded agarose gels. J. Biomech. Eng. 2000;122:252–260. doi: 10.1115/1.429656. [DOI] [PubMed] [Google Scholar]
  • 14.Mauney JR, Blumberg J, Pirun M, Volloch V, Vunjak-Novakovic G, Kaplan DL. Osteogenic differentiation of human bone marrow stromal cells on partially demineralized bone scaffolds in vitro. Tissue Engineering. 2004;10:81–92. doi: 10.1089/107632704322791727. [DOI] [PubMed] [Google Scholar]
  • 15.Mauney JR, Jaquiery C, Volloch V, Herberer M, Martin I, Kaplan DL. In vitro and in vivo evaluation of differentially demineralized cancellous bone scaffolds combined with human bone marrow stromal cells for tissue engineering. Biomaterials. 2005;26:3173–3185. doi: 10.1016/j.biomaterials.2004.08.020. [DOI] [PubMed] [Google Scholar]
  • 16.Lima EG, Mauck RL, Han SH, Park S, Ng KW, Ateshian GA, et al. Functional tissue engineering of chondral and osteochondral constructs. Biorheology. 2004;41:577–590. [PubMed] [Google Scholar]
  • 17.Lima EG, Chao PHG, Ateshian GA, Bal BS, Cook JL, Vunjak-Novakovic G, et al. The effect of devitalized trabecular bone on the formation of osteochondral tissue-engineered constructs. Biomaterials. 2008;29:4292–1499. doi: 10.1016/j.biomaterials.2008.07.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lennon DP, Caplan AI. Isolation of human marrow-derived mesenchymal stem cells. Experimental Hematology. 2006;34:1604–1605. doi: 10.1016/j.exphem.2006.07.014. [DOI] [PubMed] [Google Scholar]
  • 19.Chao PG, Tang ZL, Angelini E, West AC, Costa KD, Hung CT. Dynamic osmotic loading of chondrocytes using a novel microfluidic device. Journal of Biomechanics. 2005;38:1273–1281. doi: 10.1016/j.jbiomech.2004.06.016. [DOI] [PubMed] [Google Scholar]
  • 20.Chao PHG, West AC, Hung CT. Chondrocyte intracellular calcium, cytoskeletal organization, and gene expression responses to dynamic osmotic loading. American Journal of Physiology-Cell Physiology. 2006;291:C718–C725. doi: 10.1152/ajpcell.00127.2005. [DOI] [PubMed] [Google Scholar]
  • 21.Liu XWS, Sajda P, Saha PK, Wehrli FW, Guo XE. Quantification of the roles of trabecular microarchitecture and trabecular type in determining the elastic modulus of human trabecular bone. Journal of Bone and Mineral Research. 2006;21:1608–1617. doi: 10.1359/jbmr.060716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Schaefer D, Martin I, Jundt G, Seidel J, Heberer M, Grodzinsky A, et al. Tissue-engineered composites for the repair of large osteochondral defects. Arthritis and Rheumatism. 2002;46:2524–2534. doi: 10.1002/art.10493. [DOI] [PubMed] [Google Scholar]
  • 23.Schaefer D, Martin I, Shastri P, Padera RF, Langer R, Freed LE, et al. In vitro generation of osteochondral composites. Biomaterials. 2000;21:2599–2606. doi: 10.1016/s0142-9612(00)00127-7. [DOI] [PubMed] [Google Scholar]
  • 24.Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science. 1999;284:143–147. doi: 10.1126/science.284.5411.143. [DOI] [PubMed] [Google Scholar]
  • 25.Lima EG, Bian L, Ng KW, Mauck RL, Byers BA, Tuan RS, et al. The beneficial effect of delayed compressive loading on tissue-engineered cartilage constructs cultured with TGF-beta 3. Osteoarthritis and Cartilage. 2007;15:1025–1033. doi: 10.1016/j.joca.2007.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wendt D, Jakob M, Martin I. Bioreactor-based engineering of osteochondral grafts: from model systems to tissue manufacturing. Journal of Bioscience and Bioengineering. 2005;100:489–494. doi: 10.1263/jbb.100.489. [DOI] [PubMed] [Google Scholar]
  • 27.Martin I, Wendt D, Heberer M. The role of bioreactors in tissue engineering. Trends in Biotechnology. 2004;22:80–86. doi: 10.1016/j.tibtech.2003.12.001. [DOI] [PubMed] [Google Scholar]
  • 28.Abzhanov A, Rodda SJ, McMahon AP, Tabin CJ. Regulation of skeletogenic differentiation in cranial dermal bone. Development. 2007;134:3133–3144. doi: 10.1242/dev.002709. [DOI] [PubMed] [Google Scholar]

RESOURCES