The initiation and successful differentiation of human mesenchymal stem cells to chondrogenic lineage is influenced by several factors in vitro, such as the initial cell seeding density, presence of a three dimensional scaffold and environmental cues such as the addition and timing of the chemical factors etc.
The potential of human mesenchymal stem cells to differentiate into chondrogenic lineages has been successfully established in vitro by using a high-cell density pellet culture system (Barry et. Al., 2001). Pellet culture of human mesenchymal stem cells were capable in initiating chondrogenic differentiation since they resembled the densely packed precursor cells formed during the stage of mesenchymal cell condensation (a pre requisite to cartilage formation), which facilitates the chemical and physical factors needed to encourage cartilage formation. Formation of cell clusters allows neighboring cells to interact with each other via the surface proteins, which triggers the signaling pathway needed to commit the multipotent stem cells to osteochondrogenic lineage. Although the pellet culture has demonstrated to be a good system for chondrogenesis, it is not the most efficient technique to form cartilage and treat cartilage defects. This is primarily due to the fact that the majority of cells injected into the body in the pellet form may become non-viable and unstable since they are not localized. Furthermore, studies conducted have shown that the mature articular chondrocytes cultured in vitro require a three-dimensional support structure in order to maintain their differentiated phenotype. These scaffolds are preferably made up of biocompatible, immunoneglible materials such as collagen, agarose, and alginate etc. Additionally, the material used will also determine the cell-matrix interactions that may take place, which is needed for chondrogenesis to progress. Hence, it is vital that the scaffold used allows such interactions to take place. Since, collagen is predominantly found in the native cartilage matrix, we have used collagen to form the three dimensional microspheres, which encapsulates the undifferentiated human mesenchymal cells.
The initial cell seeding density is believed to influence the number of cell-cell interactions that can take place and thus affects the chondrogenic development process.
In our hMSC-collagen microencapsulated spheres, the extent of contraction of the microspheres is dependent on two main parameters, namely cell seeding density and the collagen concentration used. This self-induced contraction process originates from the traction force, resulting from the cells migrating. The larger the traction force, the greater is the extent of microsphere contraction.
During the contraction process, the human mesenchymal stem cells condense to form a more compact cell cluster at the center of the microsphere, thus resembling a three-dimensional high cell-density format such as a pellet culture or a micromass culture, which have proven to lead to formation of mature chondrocytes. The condensation of the mesenchymal cells is important during the initial stages of cartilage development and is a pre-requisite to endochondral ossification.
The initial cell seeded density and collagen concentration was varied to investigate whether the contraction of the microsphere recapitulates the mesenchymal condensation process and prepares the stem cells for subsequent chondrogenesis.
Changing these two parameters affected the extent of microsphere contraction taking place during the first seven days cultured. The change in microsphere size was studied by measuring the diameter length and calculating the change in microsphere volume for each of the time-points (i.e. day 0, 1, 2,4 and 7). The cell densities used were 5e5cells/ml and 1e6cells/ml with a microsphere volume of 5ul, meaning that the number of cells seeded in each microsphere were 2500 and 5000 cells respectively. The higher cell density resulted in a larger contractile force generated for the same collagen concentration used. This is due to the fact that in the higher cell density microsphere, more number of cells, in this case double, were migrating and therefore a larger tractile force is generated.
The different collagen concentrations considered were 1mg/ml, 2mg/ml and 3mg/ml. Increasing the collagen concentration meant that the matrix became stiffer**, and hence reduced the extent of contraction that could take place.
The expression of the transcription factor sox9 and runx2 were observed through immunofluorescence staining to see if there were any presence of colocalization with the nucleus. The localization of sox9 and runx2 with the nucleus indicates that the human mesenchymal stem cells have committed to the osteochondrogenic lineage, thus preparing the cells for differentiation to osteogenic or chondrogenic lineages.
Osteochondroprogenitor cells are derived from the precursor mesenchymal stem cells, present in adult bone marrow. Upon chemical induction through the addition of signaling factors such as FGF or TGF-beta, the precursors are capable of differentiating into either chondrocytes or osteoblasts. These progenitor cells express two major transcriptional factors, namely Sox9 and Runx2.
Sox9 is responsible for stimulating the expression of chondrogenic-related genes such as Sox5, Sox6, collagen type 2a1. Furthermore, it had been reported that sox9 is also important in upregulating aggrecan, the major type of proteoglycan found in cartilage matrix. During mesenchymal condensation, the expression of sox9 is upregulated. The results obtained from the RT-PCR of the sox9 gene expression (normalized to day 0) had resulted in an upward trend over the seven days, thus suggesting sox9 is up-regulated. Accordingly, as sox9 facilitates the activation of collagen 2, a similar trend was observed for collagen 2a1 gene expression as well. These observations could well support the idea that the collagen microencapsulation recapitulates mesenchymal condensations and prepares them for subsequent chondrogenesis process.
Furthermore, Bonferroni's post hoc test had shown there is significant difference in the relative sox9 gene expression between day 7 and days 0,1 and 2 (at 0.05 significance level).
Another role of Sox9 includes repressing the activity of Runx2, thus inhibiting osteogenesis. Runx2 is important in triggering osteogenic-specific genes as well as those genes responsible for promoting the formation of hypertrophic chondrocytes. From the results obtained from RT-PCR, the Runx2 gene's expression remained fairly constant, which could be explained by sox9's action in inhibiting the upregulation of Runx2, further favoring chondrogenic differentiation.
Encapsulation of the stem cells in the collagen microsphere did in fact trigger the initial stages of cartilage development since sox9 nuclear colocalization was clearly observed for all the conditions since day 0 post encapsulation. But, the extent of sox9 and runx2 expression in the nucleus varied at different time points and was sustained for different periods of time for the various conditions. However, this observation was very subjective and lacked any quantitative type of measurement for the colocalization, therefore cannot be completely relied upon. But, then again since several regions of the microspheres were imaged in order to cover atleast thirty individual cells, the results can be used to somewhat represent the population of that timepoint.
The condition that seemed most favorable for chondrogenesis was 1e6 cells/ml and 1mg/ml collagen concentration. For this condition, sox9 seems to be retained in the nucleus for a much longer period of time, i.e. throughout the seven days and colocalization of runx2 with the nucleus was absent, perhaps preparing the cell to follow the chondrogenic lineage. Combining a high cell-density with a soft, low collagen concentrated matrix, resulted in the most microsphere contraction, possibly due to the generation of a large tractile force. Subsequently, this resulted in the formation of a very compact three-dimensional cell mass (most likely facilitating numerous cell-cell interactions), and thus promoting chondrogenesis (seen by the sustained colocalization of sox9 with the nucleus for 7 days).
Comparing this to a microsphere with a stiffer matrix (1e6 cells/ml cell density and 2mg/ml collagen concentration for instance), a difference in the trend of sox9/runx2 colocalization with the nucleus is apparent. For the condition with a higher collagen concentration, sox9 colocalization seems to peak between day 0 and day 1, but then is no longer sustained in the remaining time points, during which runx2 seems to be activated in the nucleus. The reasoning behind this could be due to decreased cell-cell interactions established, which is possibly due to a relatively lower magnitude of contraction of the microsphere. Activation of runx2 could suggest the cells are going toward a hypertrophic stage.
However, since no chemical induction was carried out in this study, continuing to culture the progenitor cells will possibly result in dead or waste cells.
Throughout the chondrogenesis developmental process, matrix remodeling takes place and specific cell types characterize different stages. Staining for different cell surface markers and studying the temporal change in the activated transcription factors are few ways in categorizing the stage of chondrogenesis that the cell belongs to. Furthermore, the surrounding extracellular matrix components present is also used to determine the period of the developmental process. During the osteochondroprogenitor stage, which results from the mesenchymal condensation process, it has been proposed that these cells excrete collagen type II. From the immunofluorescent and immunohistochemical staining of collagen II for the undifferentiated 5e5 cells/ml cell density microspheres, tested positive. This suggests that collagen type II is already being synthesized and secreted by the cells (although relatively much lower compared to the chondrogenic differentiated microspheres). Therefore, it is quite reasonable to say that these cells are infact committed to the osteo- or chondrogenic lineage, although further signaling and chemical induction is necessary to direct it to the appropriate lineage.
Other possible characterization of mesenchymal condensation:
Mesenchymal condensation is the earliest cellular processes that can be detected during cartilage development. There are various molecules to verify the progenitor stage including intracellular signaling proteins, cell surface markers that enable cell-cell and cell-matrix communications and other various extracellular components for instance. The two most commonly used markers during the initial stages of chondrogenesis stage are peanut agglutinin and the cell-adhesion molecule N-Cadherin.
It has been stated in many literatures and books related to bone tissue engineering that one commonly tested marker during mesenchymal condensation is a lectin molecule called peanut agglutinin (PNA). Positive staining of peanut agglutinin is characteristic of condensing mesenchymal osteochondroprogenitor cells. Therefore, staining the undifferentiated collagen-microspheres during the early time points will aid in confirming the presence of condensation process, which can further be used to support the committing step of the multipotent stem cells.
The condensation step facilitates subsequent cell-cell interactions that need to take place by bringing the cells close to each other. Neighboring cells communicate with each other through gap junctions. One such molecule that is important during mesenchymal chondrogenesis is N-Cadherin, which is a calcium-ion dependent cell-cell adhesion molecule. It responsible for regulating cell-cell interactions in aggregated mesenchymal cells during the cellular condensation process. N-Cadherin mediated communication is key for the progression of chondrogenesis. Staining positive for N-Cadherin or evaluating the temporal change in N-Cadherin gene expression will further help in supporting the theory that hMSC-collagen microencapsulation technique does in fact promote or recapitulates the condensation process.
Apart from the cell density and scaffold mateiral used, another parameter playing an important role in chondrogenic differentiation is chemical induction brought about by the addition of TGF-beta or FGF. The timing of its addition is also important in determining the extent of chondrogenic differentiation that can be achieved as well as the quality of matrix production of the chondrocytes.
Transforming growth factor beta (TGF-beta) is a type of cytokine that regulates the cellular proliferation and cellular differentiation in many cell types and also has an essential role in inducing the chondrogenic differentiation process during the early stages.
To investigate the impact the temporal exposure of the mesenchymal stem cells to chondrogenic factorshas on chondrogenesis, freshly prepared chondrogenic differentiation medium was added at different time points post collagen-microencapsulation, essentially this was at day 0, 1, 2, 4 and 7 after collagen encapsulation. The condition used for the test was 5e5 cells/ml cell density and 2 mg/ml collagen concentration. Immunofluoroscent staining of the differentiated microspheres for collagen type II showed that those induced at earlier timepoints (i.e. day 0 and day 1) seemed to be more actively synthesizing and depositing collagen fibers onto their matrix when compared to those that have been chemically stimulated later. This suggests that early induction of the mesenchymal stem cells promotes a higher expression of collagen II in their matrix by the end of 21 days culture. Increased synthesis of collagen type II proteins for early-induced microspheres could be due to early activation of the chondrogenic specific genes such as sox9 and collagen 2a1. It has been demonstrated that TGF-beta induction within the first few days of culture is vital for committing and differentiation of mesenchymal progenitor cells as well as in expressing a chondrogenic phenotype. *** Several papers have also suggested that the role of TGF-beta is vital during the initial chondrogenic developmental process, for instance by overexpressing the marker N-Cadherin. Therefore, there seems to be a time window for which the chemical stimulation needs to take place. Otherwise, inefficient chondrogenic differentiation may result as with my case for differentiating at days 4 and 6, for which very little collagen type II was expressed in the matrix.
This result is comparable with one such study that investigated the extent of induction time on maintaining chondrogenic phenotype in alginate-encapsulated mesenchymal stem cells. Basically, alginate-encapsulated mesenchymal stem cells were exposed to the transforming growth factor for 3, 6 and 14 days. It had been concluded that longer treatment of the cells with TGF-beta did not enhance chondrogenic differentiation and stability of chondrogenic-specific phenotype. From gene expression analysis, it was apparent that the effect of the transforming growth factor was greatest during the first three days of exposure and exposing it for six das did noth ave any significant difference in terms of quantitative gene expression of cartlage0specific genes such as aggrecan, COMP and collagen type II. Hence, it has been suggested that essential events responsible for chondrogenic differentiation of mesenchymal cells take place during the early phase of exposure.
One particular literature was comparing the effect of exposing human mesenchymal stem cell in hydrogels for short and prolonged periods of time to TGF-beta1. They had concluded that TGF-beta1 was necessary to trigger the initial stages of chondrogenic process and had also stated that prolonged exposure throughout the 21 days of culture had resulted in increased collagen production, but had no effect proteoglycan synthesis.
Pellet Vs. Microsphere:
An experiment regarding in vitro chondrogenesis of human mesenchymal stem cells at Johns Hopkins University aimed to investigate the effect of recapitulating the hypoxia condition during mesenchymal condensation in pellet culture system. They had proposed that exposure to hypoxia condition (2% oxygen) recapitulated mesenchymal-condensation environment. Therefore, they had conditioned human mesenchymal stem cells to either hypoxic (2% oxygen) or normoxic (20% oxygen) environment for different periods of time prior to the addition of chondrogenic differentiation medium. The stem cells were cultured for a total of 28 days. They then assessed whether priming the undifferentiated stem cells to different oxygen concentrations for different lengths of time has any effect on the chondrogenesis process. This was measured by analyzing the expression of chondrogenic-related genes such as Sox-9, Runx-2, Collagen X, Collagen I, MMP13 and N-CAM. This study had concluded that the efficiency of subsequent chondrogenesis was both time and oxygen-dependent. Up-regulation of genes such as Sox9 and N-CAM which are condensation-specific genes were observed at the end of 1 week post exposure to hypoxic conditions. But, differentiating this group and culturing them for 4 weeks did not present to be the most efficient chondrogenesis, which was evaluated by matrix deposition and accumulation. In fact, the group that exhibited the most efficient chondrogenesis at the end of the four weeks was that, which was preconditioned in normoxic conditions for 1 week and was chemically induced also under 20% oxygen.
Pellet cultures were used in this study to maximize the formation of cell aggregates and the amount of cell-cell interactions that can take place.
The Collagen-hMSC microsphere developed in our lab is quite similar to the pellet culture system in providing an environment that facilitates the initial stages of chondrogenic developmental process.
Three-dimensional microspheres are increasingly being used in cartilage tissue engineering instead of pellet cultures to mimic the highly dense cellular systems of the established pellet culture or micromass cultures.
The properties of individual microspheres can be altered to match what we want. The features that are amendable include porosity, size, surface characteristics, permeability, cell adhesion strength etc by selecting the materials and the process used to formulate them. . For example, microsphere porosity is especially important for cartilage tissue engineering to allow cell proliferation and deposition of newly synthesized matrix in the void areas. The microsphere technology further allows coencapsulation with the scaffold material and other soluble factors. Furthermore, microsphere encapsulated cells are easier to be localized within a defect tissue compared to a pellet culture system, for which the majority of the cells may become non-viable upon in vivo conditions, by means of injection. Therefore, a pellet culture is less feasible to be implanted in vivo. Additionally, microspheres provide a three dimensional scaffold for the cells to interact with to facilitate subsequent migration, proliferation and differentiation of the stem cells, and thus enhancing matrix production and ccumulation. The synthesis of the matrix leads to increased mechanical strength of the construct. This feature is especially beneficial for treating articular defects since improved mechanical properties is necessary to enable handling and implantation of the engineered cartilage without losing is shape and it should also be capable in withstanding the forces experienced in the joint until the regenerating cartilage becomes resilient enough to support the loads. The engineered constructs formed are also able to incorporate well with the native tissue.
Although pellet cultures of mesenchymal cells have exhibited to be efficient in chondrogenesis and matrix production, it has failed to maintain the chondrogenic phenotype. Whereas, hMSC encapsulated in microspheres can be expanded to the required number and induced to chondrogenic lineages. Moreover, hMSC-microspheres also aid in maintaining the chondrocyte phenotype, which is important for in vivo inclusion with the native tissue. There have been several studies that support this statement. One such study showed that injectable chondrocyte-PLGA microspheres that were fabricated showed a better incorporation (producing tissue more similar to the native tissue in terms of matrix deposition, spacing between cells and uniformity of the matrix) with the in vivo cartilage tissue in comparison to implanting the chondrocyte alone or the microsphere alone, used as a control. These injectable microspheres can therefore be used as in vivo cell delivery systems, where they scaffold material degrades over time as it is replaced by the secreted exctracellular matrix.
(Tsunematsu, 1979; Zanetti and Solursh, 1984; Bruckner et al., 1989; Solursh, 1991; Lemare et al., 1998; Liu et al., 1998)