Stem-cell research has had more than its share of 'highs' and 'lows'. But now, reporting in Nature, Fred Gage and co-workers show that stem cells that are derived from the nervous system can be converted into endothelium-like cells in the absence of cell fusion.

Studies indicating that tissue-specific somatic stem cells can differentiate into other lineages have caused great excitement among researchers, as they promise an attractive potential resource for future regenerative therapies. But, in many cases, these phenotypic changes reflected the donation of genetic material through the formation of cell hybrids.

In view of these results, Gage and his team re-evaluated the developmental potential of stem cells that were derived from adult mouse brain, taking care to exclude fusion events. The authors harvested neural stem cells (NSCs) from mice that constitutively expressed green fluorescent protein (GFP), isolated single cells, and expanded them in culture to derive purified clonal cell lines. As NSCs and blood-vessel endothelia are often anatomical 'neighbours', the authors co-cultured these clones with GFP-negative primary human endothelial cells to ascertain if the NSCs could be induced to develop into endothelial cells. After 2–5 days of co-culture, 6% of the NSCs had developed an endothelial phenotype, as shown by positive staining for both GFP and an endothelium-specific cell-surface antigen, CD146. A purified, expanded population of these GFP/CD146-positive cells was then subjected to PCR after reverse transcription of RNA (RT-PCR), and to immunofluorescence and functional assays (such as capillary-network formation and electron-microscopic identification of specialized secretory vesicles). The results of these studies confirmed the acquisition of endothelium-specific gene expression, protein markers and functional properties. Importantly, these changes were accompanied by the decreased expression of certain neural-lineage-specific genes.

To exclude the possibility that cell fusion had occurred, Gage and co-workers adopted various strategies. First, they prepared metaphase chromosome spreads from NSC-derived endothelial cells. If cell fusion had occurred, microscopic examination would easily distinguish human chromosomes — which are metacentric and acrocentric, from the characteristic telocentric murine chromosomes. In fact, no human chromosomes were present. Next, they excluded initial fusion events by co-culturing NSCs with paraformaldehyde-treated human endothelial cells, which are unable to undergo cell fusion, but can mediate surface receptor contact. After 5 days, 2–4% of these co-cultured GFP-containing NSCs bound to lectin — a feature of mouse endothelium. Similar endothelium-like cells were found in sections of embryonic brain after in utero transplantation of one clonal stem cell line, although in vivo conversion rates were lower than those observed in vitro. The authors propose that these results could reflect a novel mechanism of angiogenesis, in which cells that line blood vessels are derived from adjacent nervous tissue.

By showing that ectodermally-derived NSCs can differentiate into endothelial cells (which originate from mesoderm), this study revives the concept that NSCs have a relatively broad developmental potential that enables these versatile cells to adopt alternate germ cell layer affiliations.