In this study, we demonstrate an improved method for driving ES cells to a cartilaginous fate when stimulated with BMP-2 and TGFβ1
. In the early phase of differentiation, BMP-2 operates by directing differentiation towards the cartilage lineage and acting on chondroprogenitors. During later differentiation adding mineralizing agents can trigger hypertrophy and mineralization of the ES-derived chondrocytes. In the embryo, maturation of chondrocytes in the process of endochondral ossification follows a timely regulated developmental program, whereby cellular stages can be delimitated molecularly. During ES cell differentiation into chondrocytes, developmental processes follow the same pattern, as judged by the gene expression patterns observed. Chondrocytes, osteoblasts and adipocytes are thought to arise from the same mesenchymal progenitor. Based on our previous observations around VD3
-induced osteogenesis, we have already hypothesized that during the first 5 days of differentiation mesodermal progenitors develop, which then are susceptible to the VD3
]. Treatment of the cultures with TGFβ1
at days 3–5 may augment the number of mesenchymal progenitors, as TGFβ1
is thought to inhibit the proliferation of most cells, but to stimulate some mesenchymal cells such as osteoblasts and chondrocytes [8
]. It was not surprising to see adipocytes develop in many of our cultures, as they represent another member of the TGFβ1
promoted mesodermal lineage.
The gain of adipose characteristics in culture is hallmarked by: a) the appearance of cytoplasmic lipid droplets, b) the acquisition of insulin sensitivity with regard to glucose uptake (GLUT4) and c) the expression and secretion of numerous bioactive molecules [5
]. All of these characteristics of in vivo
adipogenesis were met in the EBs treated with BMP-2, TGFβ1
, insulin and ascorbic acid. Indeed, a future challenge for improving adipogenic cultures will be the discovery of regulatory pathways in adipogenesis. Once identified, such pathways may be antagonized in order to enhance ES differentiation into chondrocytes.
During the revision of this paper, a study was published describing that the overexpression of the sox triad, namely Sox9, Sox5 and Sox6, markedly increased chondrocyte marker gene expression (collagen type 2, aggrecan) in ES cells within 3 days [29
]. Here, TGFβ1
, BMP-2, IGF-1 and FGF-2 had no affect on the immediate regulation of these genes. In our differentiation system, Sox9 is elevated very early during differentiation at day 5 (data not shown) underlining its role as an early controller of chondrogenesis. We have shown here that BMP-2 regulates later processes in cartilage development. Marker gene expression levels reached upon overexpression of the Sox trio do not meet the levels we have observed in this study, suggesting that Sox9 may be necessary but not sufficient do direct all progenitors to the chondrocytic lineage.
Earlier this year, another study portrayed chondrogenesis in ES cells using encapsulation in alginate [26
]. The usage of a 3D culture system led to an increase in Col II and aggrecan expression of about 1.5- to 2-fold above the regular 2D system of plating EBs. Comparing those values to the 80- to 90-fold up-regulation described here, it is clear that three-dimensional signals also need to be incorporated into chondrocyte differentiations to increase differentiation efficiency.
Articular cartilage has been refractory to repair following degeneration. Despite its limited capacity to self-repair, cartilage is replete with cells capable of undergoing mitotic division [30
]. Current hypotheses suggest that cells may be constrained by their ECM, thus preventing expansion and differentiation or that there is a limit in the bioactive molecules, which support chondrogenesis [31
]. Engineering bone or cartilage usually requires the handling of autologous cells. Cells are released from the ECM using collagenase and hyaluronidase. However, the small number of available progenitors, within the tissue can be problematic [33
]. Moreover, the ability of these harvested cells to proliferate is limited in the elderly, where most degenerative joint disorders occur. The outcome of conventional surgical treatment including joint resurfacing or biological autografts has been unsatisfactory following long-term evaluation [34
]. This failure is caused by insufficient repair resulting in the formation of mechanically inadequate resident fibrocartilage. These disappointing results and the limited therapeutic opportunities have led investigators to focus on more appropriate bioregenerative tissue engineering approaches, which could be specifically tailored for a patient's needs.
Pluripotent ES cells are now being contemplated as a new cell source for tissue engineering since they are the most multifaceted cells amongst all stem cells. While their pluripotency offers a huge potential for cell therapies directed against a wide spectrum of degenerative diseases that are ineffectively treated by traditional approaches, it also poses challenge for controlling developmental fate in the recipient. Understanding the developmental pathways regulating ES cell differentiation, will enable the creation of a cell source that can be manipulated to correct for a particular defect.
This study was designed to improve upon the number of differentiated cells needed for the in vitro
development of functional cartilage as this represents a first critical step in applying ES cells clinically. Additionally, the presented in vitro
model of chondrogenesis may be useful for in vitro
embryotoxicity tests [27
] as also serve as a new tool in identifying molecular pathways that underlie chondrogenesis. Compared to other in vitro
models of chondrogenesis, this model incorporates all steps of development beginning with a pluripotent, uncommitted cell and thus might further our understanding of normally unamendable stages of development.