Normal Levels of Sox9 Expression in the Developing Mouse Testis Depend on the TES/TESCO Enhancer, but This Does Not Act Alone

During mouse sex determination, transient expression of the Y-linked gene Sry up-regulates its direct target gene Sox9, via a 3.2 kb testis specific enhancer of Sox9 (TES), which includes a core 1.4 kb element, TESCO. SOX9 activity leads to differentiation of Sertoli cells, rather than granulosa cells from the bipotential supporting cell precursor lineage. Here, we present functional analysis of TES/TESCO, using CRISPR/Cas9 genome editing in mice. Deletion of TESCO or TES reduced Sox9 expression levels in XY fetal gonads to 60 or 45% respectively relative to wild type gonads, and reduced expression of the SOX9 target Amh. Although human patients heterozygous for null mutations in SOX9, which are assumed to have 50% of normal expression, often show XY female sex reversal, mice deleted for one copy of Sox9 do not. Consistent with this, we did not observe sex reversal in either TESCO-/- or TES-/- XY embryos or adult mice. However, embryos carrying both a conditional Sox9 null allele and the TES deletion developed ovotestes. Quantitative analysis of these revealed levels of 23% expression of Sox9 compared to wild type, and a significant increase in the expression of the granulosa cell marker Foxl2. This indicates that the threshold in mice where sex reversal begins to be seen is about half that of the similar to 50% levels predicted in humans. Our results demonstrate that TES/TESCO is a crucial enhancer regulating Sox9 expression in the gonad, but point to the existence of additional enhancers that act redundantly.