Integrated analysis of different non-coding features across the Sox2 locus implicates a diencephalic enhancer in adult brain expression

Abstract

ABSTRACTSOX2 is a prominent member of the SOX family of transcription factors that has many different functional roles. This pleiotropy is made possible by multiple regulatory mechanisms that direct appropriate spatial and temporal patterns of expression, and therefore action. The current study concerns the mechanisms that determine Sox2 gene expression in the adult mammalian brain, where SOX2 protein is absent in general, but is selectively and abundantly expressed in a majority of neurons within a ventral diencephalic brain structure, the suprachiasmatic nucleus (SCN). In this study, a comparative bioinformatic and biochemical analysis of different adult rat brain regions was conducted in order to identify SCN-selective (immaturity-related) regulatory mechanisms. The approach incorporated an integrated analysis of Sox2 enhancers, CTCF binding sites, and also expression of the Sox2-overlapping, long non-coding (lnc)RNA, Sox2ot. Initial experiments revealed brain region-specific Sox2ot expression (including region-specific novel transcripts), indicating a significant diversity of Sox2ot expression across the adult brain. However, the pattern and abundance of Sox2ot expression in the SCN, relative to selected control areas of the brain, did not indicate an overt relationship to Sox2 gene expression. Furthermore, although multiple individual Sox2ot exon sequences were shown to overlap annotated Sox2 gene enhancers at different sites across the Sox2 locus, again there was no indication of a SCN-specific functional correlation. Further integration with an analysis of selectively-active CTCF sites within the Sox2 locus directed attention to one site with both a prominent peak of activity in immature brain, and proximity to a functionally-characterized, ventral diencephalic, Sox2 enhancer termed U6 (upstream enhancer 6). Ex vivo analysis of the U6-associated CTCF site revealed SCN-selective CTCF binding, and these sequences were both localized within a known (brain region-selective) super-enhancer. Bioinformatic analysis of the U6 enhancer sequence revealed an abundance of consensus sites for the SCN-selective transcription factor LHX1, and over-expression of this factor enhanced the activity of cloned U6 sequence in transfected cells. However, despite this compelling evidence for a molecular mechanism that underlies adult brain expression of SOX2, further analysis of LHX1-SOX2 co-expression in the SCN confounded this view, indicating the presence of other concurrent mechanisms in the different cell populations of the SCN.