The murine ventricular-subventricular zone (V-SVZ), in the walls of the lateral ventricle facing the striatum, is a very active germinal region in which a continual process of progenitor cell replication and differentiation results in the persistent production of olfactory bulb (OB) interneurons and corpus callosum oligodendrocytes1. The lifelong generation of these cells appears to be supported by the presence in this region of neural stem cells (NSCs; also called B1 cells), which express the astrocytic antigen glial fibrillary acidic protein (GFAP) and stem cell markers such as nestin, Id1 and Sox22. GFAP-expressing B1 cells generate transit amplifying progenitor (TAP) cells (C cells), which express transcription factors Dlx2 (distal-less homeobox 2) and Ascl1 (mammalian achaete-schute homolog 1) and divide rapidly a few times before they give rise to migrating neuroblasts (A cells) or oligodendroblasts3. Newly-generated proliferative neuroblasts migrate anteriorly, forming the rostral migratory stream (RMS) to the OB, where they integrate into the granular and glomerular layers as differentiated inhibitory interneurons. Migrating young oligodendroblasts move to the CC, where they become immature NG2-positive cells that continue to divide locally or differentiate into mature myelinating oligodendrocytes1,4.
B1 cells, which derive from fetal radial glial cells, retain the elongated and polarized morphology of their predecessors and exhibit a highly specialized relationship with their niche. They span between the ependyma which lines up the ventricle and the network of blood vessels that irrigate the V-SVZ niche. The small apical process of B1 cells intercalates among multiciliated ependymocytes and ends in a single non-motile primary cilium, whereas their basal process extends long distances to approach the planar vascular plexus that irrigates this niche ending in the basal lamina of the plexus capillaries2,5-8 .
The most reliable way to distinguish B1-NSCs from non-neurogenic astrocytes, which are also GFAP+, in the intact V-SVZ niche is based on whole-mount preparations of the ventricle lateral wall and their analysis by 3-D confocal microscopy after immunostaining for GFAP to label the thin B1-NSC apical process, β-catenin to delineate cell membranes, and either γ-tubulin as a marker of cilial basal bodies or acetylated α-tubulin to label the extent of each cilium5,8. Observations of these whole-mounts from the ventricular surface have indicated that B1 and ependymal cells are arranged in "pinwheels"5, in which the uniciliated apical processes of one or several GFAP+ B1 cells are encircled by a rosette of multiciliated ependymal cells.
The characteristic morphology of B1 cells correlates with experimental evidence indicating that blood vessels/endothelial cells and ventricular cerebrospinal fluid (CSF) constitute regulated sources of soluble signals acting on NSCs2,6,9-11. At the ventricular surface, homotypic and heterotypic apico-lateral interactions involving ependymal and B1 cells include tight junctions and adherens junctions5,12. Moreover, adhesion molecules implicated in the junctional complexes between B1 and ependymal cells, such as N-cadherin and V-CAM, have been shown to regulate not only the highly organized positioning of B1 in the V-SVZ niche, but also their quiescence12,13. The ependymal-B1 cell monolayer appears to act as a diffusion barrier allowing the regulated flux of water and small molecules from the CSF, but restricting the intercellular passage of large proteins10,11. Experimental evidence indicates that the uniquely positioned B1 cell apical cilium could play a role as a sensor of signaling polypeptides present in the CSF2,5-7. Ependymal cells are, per se, also a source of soluble and membrane-bound signals with a role in the regulation of NSC behavior14,15.
Traceable nucleosides, such as bromo-deoxyuridine (BrdU), or retroviruses have been widely used to label progenitor cells, including NSCs, in vivo. However, these methods are not optimal for long-term fate tracing because BrdU signals dilute through repeated cell divisions and retroviruses appear to preferentially target transiently amplifying cells due to their requirement of cell proliferation for transduction16,17. To examine NSC physiology in vivo, including interactions with niche components, it is crucial to establish a method to label and trace rarely dividing cells, as B1-NSCs are largely quiescent and their neighboring ependymal cells never divide under physiological conditions3. Here, we show that lentiviral vectors (LVs) allow for high-efficiency gene marking and long-term modification of adult NSCs and non-dividing ependymal cells, due most reasonably to their ability to transduce and to integrate into the genome of target cells in a cell cycle-independent way. Moreover, we show how the route of delivery and viral titer help to specifically transduce ependymal cells, but not B1 cells thereby allowing the analysis of niche-dependent, ependymal effects on NSCs.