FIG. 2.

GFAP-GFP expressing cells proliferate to form neurospheres, but do not account for all the formed spheres. (a) A cortico-hippocampal slice from P8 GFAP-GFP mice showing fluorescence in the cortex and hippocampus (scale bar 1 mm). (b) Neurospheres can be derived from cells of P8 mouse cortex sorted by flow cytometry into GFP-positive and negative gates. Most (but not all) spheres formed from cells in the GFP-positive population were fluorescent for GFP (scale bar 100 μm). (c) Quantitative polymerase chain reaction (PCR) analysis of cells sorted by flow cytometry from both P8 hippocampus (top panel) and cortex (bottom panel) shows that GFP mRNA expression is greatly attenuated in the negative population, confirming efficient sorting (p<0.005 and p<0.001, respectively, Student t test; n=6 replicates). (d, e) Triple labeling of sections of the hippocampus (d) and cortex (e) of P8 GFP-GFAP mouse for endogenous GFP, anti-GFP antibody and GFAP (scale bar 50 μm). Cells that co-localize GFP with GFAP are represented in the orthogonal views. GFAP cells that are negative for GFP are also shown in these representative micrographs. (f, g) Comparison of the neurosphere-forming ability of flow cytometry sorted GFP-positive and negative cells in comparison to unsorted cells for both P8 (f ) and P15 (g) mice (n=3 experiments). Note that by P15, cells from the cortex do not give rise to neurospheres. *p<0.01 one-way ANOVA with post-hoc Tukey test.