The reasons for this are currently unclear, although it suggests that Notch or its targets are regulated differentially depending on the type of hair cell. In this context, it is interesting to note that some adult vestibular hair cells continue to express SOX2, which is exclusively restricted to supporting cells in the cochlea. However, control of SOX2 function may be regulated by aspects other than transcriptional control, including potentially other SoxB members such as SOX21. It is interesting to note that SOX21 is also differentially expressed between the cochlear and vestibular regions in the chicken, and overexpression of SOX21 results in a bias towards the hair cell fate in vestibular regions but not in the cochlea. It is also possible that the timing of Gfi1-Cre expression is slightly different in different types of hair cells, and this may account for the differential effects. Our results indicate that activated Notch and SOX2 must be shut off in the hair cells of the cochlea for normal hair cell function, but similar downregulation may not be required for vestibular hair cells. Our results are in contrast to those of Liu et al., who used the Atoh1CreER to express NICD in developing hair cells, and found no changes in hair cell gene expression, hair cell function or morphology, despite upregulation of SOX2. Although these two Cre alleles have slightly different times of induction, with Atoh1CreER being slightly earlier than the Gfi1-Cre used in our study, we consider it more likely that these differing outcomes are due to the different endpoints examined in each study. For example Lui et al., looked at changes in hair cell gene expression at E19, whereas our study looked at later postnatal ages. Indeed, as discussed above, we demonstrated that there are only mild gene expression changes beginning around P6, which do not become pronounced until P20, along with morphological changes. In addition to the earlier induction, Lui et al., also induced Cre at later time periods, including P0 and P1, to look at the effects of later NICD re-activation in developing and adult hair cells. Similar to their results at the earlier time points, they observed no changes in hair cell gene expression, morphology, or function even at adult stages. However, it high throughput screening inhibitor appeared that very little ectopic SOX2 expression was detected in the inner hair cells, indicating that significant NICD activation was not achieved in this cell type, since in our experience SOX2 expression is an excellent readout for NICD activation. Thus, since the major effects that we observed were in the inner hair cells, it is possible that they saw few effects on the cochlea due to the low recombination rates of the NICD allele in the inner hair cell population. However, it is also possible that NICD overexpression at P0–P1 is too late to induce cell fate changes in the hair cells. Therefore, it remains an open question as to whether Notch can induce supporting cell-like changes in postnatal and adult inner hair cells.