Transcriptional regulation, target genes and functional roles of the SOX2 transcription factor in mouse neural stem cells maintenance and neuronal differentiation

The aims of this PhD research were: to examine molecular mechanisms underlying the transcriptional regulation of the Sox2 gene during forebrain development; to examine the role of Sox2 for the proper neuronal differentiation of neural stem cells; and to examine the role of Sox2 in controlling the maintenance of neural stem cells (in vivo and in vitro). The aim of the first work (Chapter 1) was to investigate the transcription factors and the regulatory sequences that control transcription of the Sox2 gene in the developing brain and neural stem cells. Our laboratory previously identified Sox2 regulatory sequences able to drive expression of a reporter β-geo transgene to neural stem cells of the brain in transgenic mice. I focused on two mouse forebrain-specific enhancers able to recapitulate Sox2 telencephalic expression throughout forebrain development, also active in neural stem cells of the adult and embryonic brain (Sox2 5’ and 3’ enhancers). This work showed that Emx2 acts as a direct transcriptional repressor of both Sox2 telencephalic enhancers, acting in two different ways to repress their transcriptional activity: by directly binding to a specific site within these regulatory elements, thus preventing the binding of activators, or possibly by protein to protein interaction sequestring the activators, thus antagonizing their activity. By the study of double mutant mice (expressing reduced levels of Sox2 and Emx2) we further found that Emx2 deficiency counteracts (at least in part) the deleterious effects of Sox2 deficiency on neural stem cell proliferation ability in the postnatal hippocampus, and also rescued other brain morphological abnormalities of Sox2-deficient mutants. It is likely possible that a simultaneous decrease of Emx2 levels (a Sox2 repressor) may antagonize these defects, by restoring Sox2 levels. In the second line of my research (Chapter 2) we performed in vitro differentiation studies on neural stem cells cultured from embryonic and adult brains of Sox2 “knockdown” mutants (expressing reduced levels of Sox2) where Sox2 deficiency impairs neuronal differentiation. In particular, my contribution to this work was to evaluate the in vitro differentiation defects of Sox2 mutant neurospheres by immunofluorescence staining for different glial and neuronal markers. Strikingly, I observed that mutant cells produce reduced numbers of mature neurons (in particular GABAergic neurons), but generate normal glia. Most of the cells belonging to the neuronal lineage failed to progress to mature neurons showing morphological abnormalities. To evaluate if restoration of Sox2 levels is able to rescue the differentiation defects of mutant cells, I engineered Sox2-expressing lentiviral vector, which I used to infect neural cells at early or late differentiation stages. I found that, Sox2 overexpression is able to rescue the neuronal maturation defects of mutant cells only if administered at early stages of differentiation. Further, I observed that Sox2 suppresses the endogenous GFAP gene, a marker of glial differentiation. These results suggests that Sox2 is required in early in vitro differentiating neuronal cells, for maturation and for suppression of alternative lineage markers. The third research (Chapter 3) investigated neurogenesis and neural stem cells properties in mice carrying a conditional mutation in the Sox2 gene (Sox2flox). Here, Sox2 was deleted via a nestin-Cre transgene that leads to complete Sox2 loss in the central nervous system by 12.5 dpc. These studies showed that embryonic neurogenesis was not importantly defective, however shortly after birth, NSC and neurogenesis are completely lost in the hippocampus. The expression of cytokine-encoding genes, essential for stem cell niche, is also strongly perturbed and leads to impaired stem cell maintenance (in vivo and in vitro). In vitro, NSC cultures derived from Sox2-deleted forebrain become rapidly exhausted, losing their proliferation and self-renewal properties. In Sox2-deleted neurospheres, Shh is extremely downregulated. However, the conditioned medium from wild type NSC cultures or the administration of a Shh agonist efficiently rescue the proliferation defects. These results suggest that the effect of Sox2 on neural stem cells growth and maintenance is partially mediated by Shh secretion, and that the Shh gene must be a direct target of Sox2. To confirm this hypothesis, I infected Sox2-deleted NSC with a Sox2-IRES-GFP expressing lentivirus just prior to the beginning of the growh decline, and I observed that the re-expression of Sox2 induces the ability to re-express Shh and rescues the formation of neurosphere. These findings indicate that NSC control their status, at least in part, through non cell-autonomous mechanisms (such as activation of important cytochine-encoding genes) which depend on Sox2.