Intravenously administered Resovist-labelled CD34(+) cells home to bone marrow of mice. Homing can be tracked in vivo by using clinical 1.5-T MR imaging technology.

Expression of "stemness" markers is widely used as a predictor of stem cell properties of mesenchymal stem cells (MSC). Here, we show that bone marrow-derived (BM)-MSC show stem cell-like behavior in vivo; that is, they form ossicles with formation of bone, formation of adipocytes, and establishment of the murine hematopoietic microenvironment. Multipotent umbilical vein-derived stromal cells (UVSC), on the other hand, do not form bone, nor do they give rise to adipocytes in vivo. Despite these differences in stem-cell-like behavior, BM-MSC and UVSC express the two transcripts variants of POU5F1 at a similar level. Also, we found that in BM-MSC and UVSC, POU5F1 is detectable. However, more than 89% of the POU5F1 transcripts correspond to the POU5F1P1, -P3, or -P4 pseudogene. Despite low-level expression of POU5F1, we were unable to precipitate POU5F1 protein in either BM-MSC or UVSC. These results demonstrate that MSC stemness does not correlate to expression of POU5F1 transcripts or its pseudogenes.

Umbilical cord blood (UCB) is frequently collected for the purpose of hematopoietic stem cell therapy. However, the use of these cells is infrequent. To investigate other possible applications we recently showed that UCB is an excellent source of endothelial colony-forming cells which improve vessel density and heart function ofter myocardial infarction (Ott, Keller, et al., FASEB J, 2005;19:992–4). We here investigated whether UCB or the umbilical cord would be suitable as a routine source of mesenchymal colony-forming cells (CFU-F). CD34+ and AC133+ cells were selected by MACS (Miltenyi) from mononuclear UCB cells. The resulting cells were cultured at limiting dilutions on fibronectin in medium with 2% FCS (BioWhittaker: EGM-2). CFU-F were isolated from 3 of 7 (42%) UCB donors investigated. However, the CFU-F frequency is extremely low: only 1.3 in 100 ml of UCB. Similraly to the endothelial cells we previously described, these few UCB-derived stromal cells (CBSC) could easily be expanded. In Parallel, we also investigated CFU-F from the human umbilical vein: human umbilical vein stromal cells (HuvSC). Here, CFU-F were isolated from all donors when veins were prepared within 6 hours of collection. The CFU-F frequency was estimated to be 3 to 5 per 10 cm of umblical vein. Like the UCB-derived CFU-F, HuvSC were easily expandable. To find out whether these stromal cells showed characteristics of immature cells, we investigated the presence of embryonic markers by RT- and realtime PCR. We confirmed the expression of POU5F1 (Oct4), SSEA-4 and Stella-related (DPPA3). Immature HuvSC were shown to express mRNA of many genes associated with mesenchymal lineages but did not express the hematopoietic markers CD45 or CD34. HuvSC were shown to differentiate into adipogenic, osteogenic and myogenic lineages, though their differentiation is not as pronounced as bone marrow derived MSC. HuvSC did also support long-term production of immature cobblestone area-forming cells week 6. In conclusion, UCB and the umbilical vein contain CFU-F with characteristics of immature mesenchymal stem cells. However, not all UCB collections contained CFU-F and the CFU-frequency is very low. Thus, UCB can, in our view, not be used as routine MSC source. On the other hand, CFU-F grew from small stretches of umbilical vein (HuvSC), making these cells excellent source of mesenchymal progenitors for tissue engineering purposes.