SUMOylation of Sox2 in neural stem cells

Sox2 plays a crucial role in the biology of neural stem cells (NSC) by regulating differentiation and keeping NSC in a slowly proliferative state which allows to maintain over time a pool of stem cells ready to differentiate in more rapidly dividing progenitor cells. Sox2 exerts these functions by transcriptionally activating and repressing target genes. At the present time, it is still not clear how these 2 opposite functions are regulated. Here, I analyse whether SUMOylation of Sox2 plays a role in the regulation of Sox2 activity and, ultimately, in the regulation of NSC biology. Several mutant forms of Sox2 were designed and cloned, including a non-SUMOylatable mutant construct, where the SUMO acceptor lysine was mutated to arginine (Sox2_K247R), and a SUMO-fusion construct where SUMO2 was fused at the carboxyl terminal of Sox2_K247R to mimic constitutive SUMOylation (Sox2_K247R-SUMO2). Comparison between the activity of these constructs and the activity of wt Sox2 revealed that over-expression of Sox2_K247R-SUMO2 causes a reduction in the protein transcriptional activation ability in vitro. When analysed in vivo, constitutively SUMOylated Sox2 repressed a set of target genes which were also repressed by over-expressed wt Sox2. These genes were enriched for cell-cycle regulators. Conversely, few genes were differentially regulated by over-expression of wt Sox and over-expression of Sox2_K247R and the majority of these few genes was up-regulated, indicating a potential loss-of-function in Sox2 ability to transcriptionally repress some of its target genes in absence of SUMOylation. Moreover, hNSC over-expressing Sox2_K247R lost the ability of proliferating slowly, while hNSC over-expressing SUMOylated Sox2 proliferated at a slower rate, similar to hNSC over-expressing wt Sox2. This evidence further promotes the hypothesis that SUMOylation of Sox2 is involved in the regulation of neural stem cells proliferation through the repression of target proliferation genes. Biochemical analysis of Sox2 SUMOylation was also performed and confirmed that, as reported in literature, SUMOylation of Sox2 is challenging to detect, supposedly because a very small portion of Sox2 is SUMOylated at any given time in the cell as reported for other SUMO targets. Taken together, the evidences reported in this study suggest a model according to which SUMOylation enhances Sox2 transcriptional repression activity, leading to the repression of proliferation genes and therefore to a slow-cycling state of neural stem cells