We have previously described the isolation of MAPCs, from human, mouse and rat bone marrow. MAPCs can differentiate into most mesodermal cell types as well as cells with neuroectodermal and endodermal features. However, culture of MAPCs remains laborious and difficult. To address the difficulties in MAPC culture initiation and maintenance, we performed multi-parameter analysis of different steps along the way. First, we evaluated whether collection of bone marrow influences isolation. We found that cells that initiate MAPC cultures appear to be present in higher frequency near growth plates, and near the endosteum. Cells from the middle part of the bone shaft generally can only be cultured short-term and rarely form MAPC lines. Hence, including the ends of the bones and very vigorous flushing of the marrow from the cavity are important. A second parameter we evaluated is cell density both during the initial phases of isolation, i.e. before depleting cells by MACS column, and afterwards. Initial plating before column depletion should be done at relatively higher density, i.e. between 2 and 3x104 cells/cm2. However, after MACS depletion, cell density needs to be kept between 2-5x102 cells/cm2. As MAPC are inclined to form clusters, and it is imperative that cell-cell contact is kept at minimum, it is not sufficient to assure global cell density, but assure that cells are distributed evenly throughout culture vessels. However, maintenance at too low cell densities significantly inhibits proliferation and eventually cells die. This suggests first that paracrine factors produced by the cells are important for their self-renewal, but that cell-cell contact factors, either alone or by secondary release of cytokines, induce cell differentiation. Initial evaluation of gene array studies comparing MAPC maintained at ideal density vs. allowed to grow at high density indicates differential expression of the main three developmental signal pathways, TGFb/BMP, Notch and wnt signaling. These are now being evaluated to determine whether relaxation of the density requirement can be done without loss of potency and self-renewal ability. We also tested the effect of trypsinization. MAPC are very lightly attached cells, whereas differentiating cells are larger and more adherent. Hence short term trypsinization (less than 30sec) and a few gentle taps loosens up the small MAPCs whereas it does not detach the larger cell population. Long-term exposure to trypsin is furthermore toxic to MAPCs; therefore immediate dilution of trypsin with a large volume of MAPC medium is required. On a practical note, as MAPC are small (9–12μm) and are present as a lose layer on top of the pellet, one needs to take care that they are not inadvertently disposed off when centrifuging. We have also found that different fetal calf serum (FCS)-lots significantly affect MAPC growth. In general, FCS that causes faster growth induce differentiation, loss of multi-lineage differentiation potential, and induce at much higher frequency cytogenetic abnormalities, including tetraploidy and aneuploidy. This is more common in mouse than rat MAPC cultures. Hence, as has been shown for hematopoietic progenitors, MSCs and ESCs, screening of serum lots is needed to optimize MAPC culture. A final vital factor is the CO2 and O2 concentration in which cells are maintained. MAPC die under alkaline conditions, hence optimally CO2 concentrations are kept between 5.5– 6%C02. Likewise, we have started to use 5% O2, as this decreases the frequency of cytogenetic abnormalities that occur. Whether this also supports better self-renewal is currently being studied.