The Role of Rest Co-Repressor 2 (Rcor2) in Cortical Neurogenesis and Callosal Development

Abstract

Cortical neurogenesis is the process by which neural stem cells progressively give rise to glutamatergic neurons that are organized into 6 distinct laminae that possess unique efferent and afferent connections to the central and peripheral nervous system. This developmental process is a culmination of multiple overlapping and exclusive molecular and cellular programs that are conducted in synchrony in a temporally and spatially dependent manner. What governs how these multiple programs are properly orchestrated is poorly understood. In this study, we investigate the function of the REST-corepressor 2 (Rcor2) transcriptional and epigenetic factor in cortical neurogenesis. Rcor2, like its paralog CoREST (Rcor1 ), is an epigenetic and transcriptional modulator capable of inducing both gene repression as well as activation. Studies from our lab and others have identified thousands of stage-specific and novel classes of CoREST-targeted developmental genes; however, the impact of Rcor2 in neurogenesis is not fully known. To examine Rcor2's role in cortical neurogenesis we have generated Rcor2 conditional knockout mice (Rcor2 CKO in dorsal telencephalic neural stem cells (dNSCs). Our findings reveal that Rcor2 plays seminal roles in maintaining neural stem/progenitor populations, and the development of the corpus callosum (CC). Loss of Rcor2 results in a dysgenesis of the CC presenting as either a partial callosal thinning (hypoplasia) or complete loss of axons projecting across the midline (agenesis). Anterograde tracing and immunohistochemical analysis indicate that axons originating from the medical cortex were selectively impaired in Rcor2CKO mice with CC hypoplasia. Examination of the embryonic cortical midline indicates that specialized neuronal and glial populations located within the indusium griseum (IG), and glial wedge (GW), respectively, are reduced in Rcor2CKO specimens. These transient guidepost cells predominantly secrete chemorepulsive, Slit2, or chemoattractive, Sema3c axon guidance cues. Consistent with GW and IG population reductions, expression of both guidance molecules is decreased in the Rcor2CKO embryonic cortical midline. In addition to CC abnormalities, we observed a reduction in cortical plate (CP) thickness, with a concomitant increase in germinative zone (GZ) thickness. Expression of laminae-specific neuronal markers, Tbr1, Ctip2, and Satb2, which label cortical layers VI, V, and II-IV in adult mice, respectively were all reduced in Rcor2CKO embryonic specimens. Changes in cortical thickness and the density of laminar-specific neuronal markers were most significantly reduced within the medial cortex compared to the lateral cortex. Expression of Pax6, and Tbr2 in dNSCs, and basal progenitor cells (BPCs), respectively were increased. To gain insight into the selective midline impairments of Rcor2CKO mice, we examined cortical hem development during early neurogenesis. In E11.5 embryos, Rcor2CKO mice exhibit a smaller hem structure. Furthermore, expression of hem-specific morphogens, BMP4, BMP7, and Wnt8b are all significantly altered at this age in Rcor2CKO embryos. To investigate potential Rcor2 direct targets, we performed complementary ChIp-seq and RNA-seq analyses in E14.5 cingulate cortex. Our findings reveal that Rcor2 occupies a large array of genes involved in multiple diverse molecular programs, such as Wnt pathways, Notch signaling, and axon guidance. PSCAN motif analysis of target genes indicates that Rcor2 can associate with REST, ESR1, Otx1/2, and Onecut proteins. Furthermore, motif analysis reveals an overrepresntation of the Wnt-effector, Lef1, and Sox9 on Rcor2 targets, indicating tha these factors may act downstream within Rcor2 regulatory networks. Complementary RNA-seq anaylsis reveals an upregulation of several Notch-related transcripts, and NSC maintenance factors, and a concomitant reduction in neuronal differentiation, associated factors. This data lends credence to Rcor2's essential role in callosal development, as well as proper stem cell maintenance and cortical laminations. Finally, we sought to evaluate how the paralog Rcorl may function redundantly to Rcor2 within CC development. We have generated Rcorl/Rcor2 double knockout mice, using the same Emxl-CRE driver (RcorDCKO) to compare changes to Rcor2CKO knockout specimens. Our findings demonstrate that RcorDCKO embryos also exhibit callosal dysgenesis, and its phenotypic manifestations are more severe than that observed in Rcor2 CKO specimens. Oddly enough changes to RcorDCKO guidepost cell populations, and guidance cue production do not exhibit a concomitant increase in severity compared to Rcor2CKO specimens. However, RcorDCKO embryos exhibit an expanded GZ dNSC population, and severe reductions of CP neurons compared to Rcor2CKO specimens. These findings cumulatively suggest that Rcorl may function redundantly to Rcor2 within cortical neurogenesis, but not within callosal guidepost development. These findings represent a novel role for Rcor2 during telencephalic development. The breadth of directly targeted genes identified by ChIP-seq analysis indicates that Rcor2 is capable of acting as a master regulator of multiple discrete cortical programs not previously known to be co-regulated that are necessary for orchestrating the cumulative effects involved in proper cortical neurogenesis.