This study demonstrates for the first time that a routine bone marrow mesenchymal cell culture contains a complete hierarchy of cells that can only be assessed by applying a careful single cell analysis. Results also show that single HPP-MCFC can differentiate into multiple mesenchymal lineages and generate secondary colonies of HPP-MCFC. The single cell analysis performed also provides a quantitative view of a population of human bone marrow mesenchymal cells and shows heterogeneity based on growth characteristics. As shown for EPC [11
], bone marrow mesenchymal cell cultures contain four distinct cell populations, namely, HPP-MCFC, LPP-MCFC, MCC, and MMC. Although this hierarchy is purely based on growth characteristics, it clearly identifies the degree of heterogeneity of a routine MSC culture. Interestingly, greater than 90% of mesenchymal cells showed limited proliferative potential with nearly 40% of the clones demonstrating less than 3 population doublings prior to senescence. Considering numerous reports indicating that adult bone marrow MSC possess exceptional proliferative potential [10
], these studies indicate that a small subpopulation of cells within a routine bone marrow mesenchymal cell culture may be responsible for establishing a long-term culture. These data suggest that this subpopulation, termed HPP-MCFC, undergoes asymmetric cell division and establishes another generation of mesenchymal cells with varying levels of proliferative potential, that not all HPP-MCFC maintain self-renewal potential, and a pure population of true HPP-MCFC may not be possible to obtain using current culture methods. It is also possible that this asymmetric cell division is a result of suboptimal culture conditions, rather than an intrinsic phenomenon. However, it is important to note that the design of culture conditions used for this study represents those widely used in the field [10
Another important factor involved in self-renewal of HPP-MCFC is the contribution of mesenchymal cell feeders, which is evident by impaired MSC emergence from single plated HPP-MCFC in a feeder-free culture system. While MSC progeny did emerge in some cultures of single plated MSC, the largest colony of mesenchymal cells that emerged that were feeder free was approximately 70 cells (~6 population doublings). Thus, there is no evidence that a single HPP-MCFC alone can establish its own microenvironment for efficient self-renewal under the culture conditions used in this study. Although it is clear that the optimization of culture conditions is necessary to promote self-renewal of HPP-MCFC without the need for feeders to support the culture, the monolayer of heterogenous bone marrow mesenchymal cells appears to contain a population of cells that supports self-renewal of HPP-MCFC. Currently, there is no definitive evidence to support the existence of a distinct MSC population and mesenchymal microenvironment in bone marrow. Sacchetti et al. [14
] indicated that a single bone marrow CD146+ colony can establish both heterotopic bone and the hematopoietic microenvironment. However, our current data suggests a clear dependency of HPP-MCFC on other cell type(s) in culture for self-renewal. Thus, it is reasonable to postulate that hematopoietic and mesenchymal stem cell populations each may require a different microenvironment for self-renewal. It is also possible that asymmetric division of HPP-MCFC is contributed to by mesenchymal cell feeders. This hypothesis can further be tested once an optimal culture condition that promotes symmetric cell division is developed.
A study by Aubin et al. [15
] showed heterogeneity in osteoprogenitor cell culture when gene expression of individual cells was evaluated by PCR. This investigation provided important insights into functional properties of stem cells that may not be apparent when they are mixed with other cell types. Thus, sub-fractionation of mesenchymal cell culture is crucial in identifying biological properties of true MSC. Colony-forming units (CFUs) obtained by limiting dilution have been used to study the clonality of MSC [12
]. However, it is difficult to show conclusively that each CFU is derived from a single cell since mesenchymal cells tend to adhere to each other. One way to show that each CFU is derived from a single cell is by screening the entire growth surface and confirming that each mesenchymal cell is separate from neighboring cells. Compared to conventional CFU-F, the sorting method of mesenchymal cells used in this study allowed more stringent single cell identification and subsequent screening of individual cells expressing EGFP under the fluorescent microscope. Even under such stringent single cell sorting and culture conditions used in the current study, rare two-cell aggregates were detected microscopically after sorting. These two-cell aggregates showed a tendency for better growth characteristics when compared to single cells and could lead to misinterpretation of growth data. Although it is possible that these cells represented two identical daughter cells, it was not possible to show conclusively that these cells were, in fact, from a single cell except through use of a lentiviral or retroviral insertional analysis at the genomic level. Thus, these two-cell aggregates were excluded from this study. In addition, expression of a fluorescent marker was necessary to conclusively show that each well truly contained a single mesenchymal cell because of the optical edge effect produced by the culture plates. This optical interference occurs mainly at the edge of wells and makes visualization of cells very difficult under bright light, which occurs with CFU-F studies. After sorting, approximately 41% of cells were noted at or very close to the edge of the wells where visualization was difficult due to optical interference. It is important to note that under the optimized lentiviral transduction conditions used in this study no adverse effects such as growth arrest, VSV-G toxicity, and undesired cellular changes [6
] were observed as a result of transduction or EGFP expression. In addition to using EGFP as a marker for sorting single cells and subsequent screening, expression of a functional gene such as alkaline phosphatase has been shown to also be helpful to interrogate the hierarchy of osteoprogenitors that are undergoing differentiation [21
]. Purpura et al. used this method to assess cell heterogeneity in fetal rat calvaria cell culture and developed a method to fractionate osteogenic populations useful for investigating differentiation capacities of individual colonies [21
]. These studies highlight the importance of evaluating individual cells when attempting to identify the phenotype of true stem cells.
Single cell-derived colonies of HPP-MCFC possess tri-lineage differentiation potential toward adipogenic, chondrogenic, and osteogenic lineages. Other reports have indicated that MSC cultures also contain unipotent and bipotent cells [22
]. Although it is possible that non-HPP-MCFC identified in the study described here may have these qualities, differentiation assays could not be performed due to their limited proliferative potential. Further studies are necessary to investigate the optimal culture condition(s) to promote self-renewal of HPP-MCFC without the need for feeder cells. As more effort is made to investigate therapeutic properties of MSC in both animal models and humans, studies on the basic biology of these cells and performed at a single cell level will provide important quantitative and qualitative insights into identifying the best source of cells for transplantation.
A population of mesenchymal cells cultured under widely used conditions [10
] and commonly considered heterogeneous [13
] was shown to contain a complete hierarchy of cells displaying different levels of proliferative potential and HPP-MCFC that can both self-renew and differentiate into three mesenchymal lineages. This study highlights the importance of analyzing mesenchymal cells at a single cell level to identify the presence of true stem cells.