ABSTRACT: Purpose: identify potential downstream targets of SOX9 in the cortical hem (CH) and dentate neuroepithelium (DNE) of the developing archicortex, comparing control with Sox9-conditionally deleted tissue using either Sox1Cre (active in both DNE and CH) or NestinCre (active in the DNE only) Methods: archicortex mRNA profiles of E12.5 and E13.5 control, Sox9fl/fl;Sox1Cre and Sox9fl/fl;NestinCre conditional embryos were generated by deep sequencing, in quadruplicates, using Illumina GAIIx. The sequence reads that passed quality filters were analyzed to identify differentially expressed genes (DEGs) between Sox9fl/fl;Sox1Cre and Sox9fl/fl;NestinCre conditional embryos, comnpared to their controls. DEGs where further analysed singularly or in pathway analysis. Hopx downregualtion was validated in vivo with immunostaining. Results: DEGs found in Sox9fl/fl;Sox1Cre/+ but not in Sox9fl/fl;Nestin-Cre mutants might specifically indicate genes regulated by SOX9 in the CH. Conversely, genes commonly affected in both mutants, might focus on pathways regulated by SOX9 in the DNE. We screened these lists to identify pathways in which SOX9 is already known to be involved or that are relevant to this developmental context. We identified the neurogenic factor NeuroD4 to be upregulated in Sox9fl/fl;Sox1Cre/+ mutants, and conversely the astrocytic markers (Atp1a2, Vim) to be downregulated, potentially indicative of the role of Sox9 in gliogenesis induction, which, in its absence, is delayed in favour of neurogenesis (Kang et al. 2012; Martini et al. 2013). We also observed a reduction in components of the NOTCH and WNT pathways (Notch1, Wnt7b, Fzd1), and of genes involved in cell adhesion, migration, and epithelial to mesenchymal transition (EMT; Limk2, Arhgap35, Fn1, Col2a1, Myh14). These results are consistent with SOX9 being involved in or interacting with both NOTCH and WNT pathways (Leung et al. 2016; Martini et al. 2013), and with previously identified roles of SOX9 in gliogenesis induction (Kang et al. 2012) as well as EMT during neural crest cells development (Liu et al. 2013; Lincoln et al. 2007). The gene Hopx was idneitified to be affected by absence of SOX9 specifically in the CH. We validated this result with in vivo immunostaining and show that HOPX is indeed highly expressed in the CH, and only mildly in the DNE. However, its expression is downregulated in both regions in absense of SOX9. Conclusions: Our study represents the first analysis of archicortex transcriptome at two consecutive developmental stages in both controls and Sox9-conditionally deleted embryos. The strategic comparison of two Sox9-conditional deletion allows dissection of SOX9 downstream targets in two adjacent regions of the developing archicortex, which we validated with immunostainings in vivo. Our methods allows to overcome the limitations of bulk RNA sequecing which does not provide spatial resolution and could be applied in other contexts to permit the dissection of complex biologic functions.
