br Materials and methods br

Materials and methods
Results
Discussion Here we establish a molecular signature for unmodified SCDSs that were initiated by individual multipotent SSCs in comparison with SCDSs that were initiated by Zalcitabine that were not multipotent, based upon their differentiation capacity as determined by in vivo transplantation. Of note, all of our SCDSs were established by clonogenic cells (CFU-Fs), but only ~1 out of 5 was in fact multipotent, as has been reported previously by us, and others (Friedenstein, 1980; Gronthos et al., 2003; Kuznetsov et al., 1997; Sacchetti et al., 2007). This reinforces the notion that cultures of BMSCs should not be referred to as “stem cell” cultures (as is often the case), but as cultures in which of a subset composed of stem cells exist. Not even all CFU-Fs are stem cells, although their enumeration provides an approximation of the number of stem cells within a freshly isolated single cell suspension of BM (Bianco et al., 2008). The results show that the molecular profiles of SDSCs were very similar to one another, but no two were alike, supporting the view that BM stromal CFU-Fs are heterogeneous. It has long been noted that upon plating of cells at clonal densities, there are differences in the size and growth habit (monolayer or multilayering) of colonies. Previous studies (e.g., Satomura et al., 1998) showed a positive correlation between rate of proliferation and multipotency of murine SCDS based on in vivo transplantation. In our current series, M-SCDSs appeared to proliferate slightly faster than F-SCDSs based on the number of days it took to reach the final harvest and the total number of cells generated, but this was not statistically significant (data not shown). Furthermore, colonies are composed of cells of different shapes and sizes, ranging from extended fibroblastic cells to large flat cells (Digirolamo et al., 1999; Owen and Friedenstein, 1988; Satomura et al., 1998). However, the morphological nature of the colony was not predictive of the outcome of in vivo transplantation assays (Satomura et al., 1998). When colonies are allowed to spontaneously differentiate upon prolonged culture, varying percentages of osteogenic, adipogenic or non-differentiated colonies arise (Owen and Friedenstein, 1988). This may be indicative of commitment of a particular CFU-F to one of the stromal cell phenotypes, as a reflection of the influences exerted on that CFU-F during embryonic growth, and post-natal development and homeostasis. With passage, the size and shape of the cells become more uniform; however, heterogeneity still persists, based on the fact that not all cells retain the ability to form colonies upon re-plating at clonal density (Friedenstein, 1976). This is most likely due to the kinetics of SSC self-renewal that are not yet well understood in mammalian systems (Neumuller and Knoblich, 2009). Furthermore, the rate of proliferation of cells within an established colony (as would be harvested at 14d) is not synchronized, with cells in the periphery migrating and proliferating at a faster rate than those that are more central (Friedenstein, 1990). These differences result in cells within the colony being in different phases of the cell cycle, which can impact on gene expression. For example, alkaline phosphatase is shed from the cell during the G2+M phase, and is slowly regained during G1 and S phases (Fedarko et al., 1990). Secondly, our study showed that SCDSs did not strictly segregate transcriptionally based on their differentiation potential as determined by in vivo transplantation. The basis for this is not clear, but may relate to a lack of knowledge concerning the stages of maturation of SSCs (pericytes) to more mature phenotypes (osteoblasts, adipocytes, stromal cells). Stages of osteogenic differentiation have been marked by use of mouse reporter lines that suggest that Runx2 is expressed in SSCs/BMSCs, and committed osteoprogenitors (Yoshida et al., 2002), Osterix is expressed in immature osteogenic cells (Maes et al., 2010), the Col1a1 2.3kb promoter is active in more mature osteoblastic cells (Pavlin et al., 1992), and that Osteocalcin is expressed in very mature osteoblasts and osteocytes (Zhang et al., 2002). However, such staging for SSC/BMSC differentiation is not yet clear. Based on the hierarchical clustering (Fig. 2A), it can be speculated that a cell that initiated a B-SCDS that clustered with M-SCDSs represented a cell that was in transition from being multipotent to a committed osteogenic cell. Likewise, the individual cells that initiated the F-SCDSs that clustered with other B-SCDSs may have recently transitioned to a fibroblastic phenotype from an osteogenic phenotype. The fact that the F-SCDSs clustered into 2 distinct groups (one with B-SCDSs, the other with M-SCDS) suggests that while all of the F-SCDSs could not make bone in vivo, there may be at least two subsets of fibroblastic BMSCs. A plausible explanation for the fact that 3 F-SCDSs clustered with M-SCDSs may relate to the fact that committed osteogenic cells (B-SCDSs) have a quite different repertoire of expressed genes compared to those that do not exert an overt phenotype (M-SCDSs and F-SCDSs). Clearly, further investigation will be needed to establish the hierarchy of SSCs/BMSCs.