19.2 n&v 685 MH

interest because of the possibility that they could be used to replace neurons that have been damaged or lost — perhaps as a result of injury such as trauma or stroke, or through neurodegenerative disorders such as Parkinson’s disease. These stem cells can give rise to neurons and their supporting cells, glia, and it is hoped that something akin to neural stem cells in the adult human brain could be stimulated to generate replacement neurons. Non-mammalian vertebrates, such as the salamander, can regenerate large portions of their brain and spinal cord,but humans have evidently lost this capacity during evolution. Therefore,most research on neural stem cells is carried out on mammals such as rodents, which are genetically closer to humans. However, although mammalian genomes may be similar, this similarity masks vast species differences in the way the brain is organized and in its capacity for regeneration and susceptibility to environmental insults. The failure of brain repair in clinical trials based on the promising results seen after the use of similar procedures in rodents is sobering testimony to the importance of such species-specific distinctions. Human neural stem cells behave differently from their rodent equivalents in culture, but the direct study of human brain tissue by Sanai et al., described on page 740 of this issue,shows additional significant and clinically relevant species-specific differences. The authors’ investigations on a large number of postmortem and biopsy samples reveal two basic findings. First, neural stem cells that can potentially give rise to neurons, as well as to two types of glial cell (astrocytes and oligodendrocytes), are situated in a region of the forebrain known as the subventricular zone. Second, a pathway known as the rostral migratory stream — which in adult rodents contains neurons that migrate from the subventricular zone to the brain region concerned with sensing smell — is absent in humans. In adult mammals, including humans, the subventricular zone (more commonly known as the subependymal zone) contains cells that have the characteristics of glial cells and that can generate neuronal cells in culture. Sanai et al. show that in adult humans these ‘glial progenitor cells’ form a prominent layer, or ribbon, that is restricted to a specific region in the brain that lines the lateral cerebral ventricle (Fig. 1). This region is also present in non-human primates,but it is thinner and less well delineated than in humans. Based on the knowledge that about 0.7% of cells in the human subventricular zone contain a specific nuclear protein, Ki-67, that is associated with DNA synthesis and hence cell division,Sanai and colleagues infer that these cells can proliferate. However, the number of cells that are actually dividing is probably smaller than this, as Ki-67 can also be present in dying cells. Importantly, although the cells that were positive for Ki-67 also contained marker molecules for astrocytes, they did not contain markers for immature neurons. This observation is similar to that seen with subependymal cells in other parts of the adult mammalian brain, such as the spinal cord or brain stem, which also do not generate neurons in vivo. However, if these cells are cultured or are transplanted to acceptable cellular environments such news and views