Role of metabolism and epigenetics in forced and endogenous neurogenesis in vivo

Since the characterisation of the very first methods to perform direct reprogramming into neurons, much work has been published on the topic describing new combinations of genes to improve the conversion efficiency or to obtain subtype-specific neurons (Arlotta and Berninger, 2014; Tsunemoto et al., 2015). However, while the ultimate goal of neuronal reprogramming is being actively explored and ameliorated, key questions revolving around the underlying mechanisms of conversion remain largely unanswered. Particularly, identifying the molecular roadblocks during direct neuronal conversion not only would bring a deeper knowledge of the overall procedure, but it would also represent a necessary prerequisite to improved conversion strategies in vivo, where direct reprogramming into neurons is often inefficient (Dametti et al., 2016; Grande et al., 2013; Guo et al., 2014a; Heinrich et al., 2014). In the projects discussed here I tackled this scientific question from two diverse perspectives, one metabolic and one epigenetic. We recently described how oxidative stress is produced during direct reprogramming into neurons triggered by Ascl1 or Neurog2 and how ultimately it affects the outcome of the procedure (Gascón, Murenu et al., 2016). Reactive oxygen species would result from an unsuccessful conversion of the metabolism of the starting cell population into the typical neuronal metabolism and indeed this critical check-point was overcome when molecules alleviating oxidative stress (Bcl-2, vitamin D or E) were provided (Gascón, Murenu et al., 2016). In my first project I extended this concept in vivo in a model of cortical grey matter stab-wound injury (Buffo et al., 2005). By using a retroviral approach to over-express Neurog2 and Bcl-2 in proliferating cells within the lesion area, I was able to successfully reprogram a high proportion of these cells into neurons, as compared to the over-expression of Neurog2 only (Gascón, Murenu et al., 2016; Grande et al., 2013). Combining this method with the administration of vitamin D or vitamin E further increased the efficiency of direct neuronal conversion in vivo to previously unprecedented levels. Moreover, I could show that the reprogrammed neurons acquire a deep layer identity, but also that they can develop mature structures and survive until late time points (Gascón, Murenu et al., 2016). Additionally to the effects of metabolism in direct neuronal reprogramming, I also addressed the implications of epigenetics by investigating the role of Ring1B, the catalytic protein of the Polycomb Repressive Complex 1 (PRC1). As Ring1B was previously described as a repressor of pro-neuronal genes in the glutamatergic lineage at the onset of the gliogenic phase during corticogenesis (Hirabayashi and Gotoh, 2010a; Hirabayashi et al., 2009), we speculated that it could also act as a major roadblock in direct neuronal reprogramming. To explore this hypothesis, I chose to use the neurogenic determinant Dlx2, which is important for the generation of GABAergic neurons during embryonic and adult neurogenesis but is only moderately effective in direct neuronal reprogramming (Brill et al., 2008; Heinrich et al., 2010; Petryniak et al., 2007). By simultaneously over-expressing Dlx2 and a miRNA against Ring1B in cortical cultures from the embryonic cortex as well as in cultures from the adult sub-ependymal zone (SEZ) I could show that less differentiated neurons are obtained, thus suggesting a different role for Ring1B during neuronal differentiation in the GABAergic lineage. Following these observations, I used a genetic approach to conditionally induce the knockout of Ring1B in adult neural stem cells and address the role of this gene in adult neurogenesis. The deletion of Ring1B resulted in the accumulation of neuroblasts migrating from the SEZ in the core of the olfactory bulb. Moreover, differentiating neurons were mainly found in the superficial region of the granule cell layer but did not express markers found in other cell populations within the olfactory bulb (Calbindin, Calretinin, Tyrosine Hydroxylase, Parvalbumin), suggesting that they still maintain their granule cell identity and have switched fate from deep to superficial granule cells. Interestingly, differentiating neurons originating from the sub-granular zone (SGZ) of the hippocampus showed longer and more branched dendrites after Ring1B deletion, indicating a function of this gene also in this neurogenic niche. Thus, Ring1B plays a key role in the specification of neuronal subtypes in adult neurogenesis. Taken together, my work has unravelled mechanisms inducing neuronal subtype specification in endogenous or forced neurogenesis