The properties of CR cells have been controversial in part due to inconsistent criteria for their morphology and gene expression. CR cells were defined as reelin-immunoreactive cells in the marginal zone in a previous study 
. Later, P73 was used as a novel marker for CR cells derived from the cortical hem in the present study 
. These definitions reduce confusion but reflect the complexity of early corticogenesis and postnatal development.
are important in the controlling CR cell differentiation, migration and survival 
; however, the molecular mechanism underlying cortical hem-derived CR cell development has not been elucidated. In the developing cerebral cortex, the cortical hem is a major source of CR cells, and is rich in bone morphogenetic proteins (BMPs) and the wingless-Int gene (Wnt) family 
but is lacking in forkhead box G1 (Foxg1) 
. Previous work has shown that Foxg1 plays a key role in fate determination of CR cells. Cortical progenitors fail to generate later-born neurons and instead continue to produce CR cells in the absence of Foxg1. Moreover, after the conditional inactivation of Foxg1, deep-layer progenitors revert to CR cells 
. A cell ablation study with a Wnt3a
-driven DTA mouse line resulted in the loss of most of the CR cells in the developing cortex 
. By tracing hem-derived CR cells using the Frizzled 10-CreER™ transgenic line, we demonstrated that hem-derived CR cells arose from Frizzled10-positive progenitors, suggesting that the Wnt signaling pathway may be important for CR cell development. More interestingly, our results indicate the progenitor zone for later CR cells production is specified from as early as E6.5. It has been reported that Foxg1 inhibits the Wnt pathway 
, while Foxg1 is not expressed until E9.5. These data suggest that if Wnt signaling is involved in specifying CR cells, it may occur earlier than Foxg1 is expressed. Foxg1 alone may not be sufficient for the specification of CR cell fate. Recent gene expression profiling of CR cells has identified a number of genes that are expressed in these pioneer neurons 
. These studies, together with our results may provide the clues to identify the molecular pathways that control the hem-derived CR cell development.
In the developing cerebral cortex, neurons are born on a predictable schedule, and the timing mechanism is programmed with individual progenitor cells. A previous study has suggested that the subplate and cortical plate cells come from a common progenitor 
. However a common ancestry for preplate cells and CR cells has not been reported, they may be pre-fated separately during a very early developmental time point 
. Our fate mapping study indicates that CR cells may be pre-fated separately from the preplate cells as early as E6.5. Further understanding these extremely early events will be quite important for cortex development.
In the present study, we have found the hem-derived CR cells migrate along the fimbrial radial glial scaffold, and preferentially settled at the hippocampal marginal zone. In addition, disruption of the fimbrial radial glial scaffold causes the arrest of CR cells in the fimbria. Ablation of the hem does not alter neocortical layering but leads to the deletion of the hippocampus 
. However, because the mice died shortly after birth, the role of the hem-derived CR cells in hippocampal development needs to be further elucidated. Our results showed that hem-derived CR cells preferentially settled in the hippocampal MZ, indicating that they may be pivotal for the development of the hippocampus. In addition, although the number of hem-derived CR cells in the neocortex decreased rapidly after birth, many remained alive in the hippocampal MZ, suggesting that these hem-derived CR cells may play more important roles in postnatal hippocampal development.
CR cells and their secreted protein, reelin, are important for cell positioning, path-finding and branching of entorhinal afferents during hippocampus formation 
. Previous studies have indicated that the cortical hem is required for hippocampal induction and expansion, and ectopic hem cells can induce ectopic hippocampal tissue, thus providing evidence that the hem is a hippocampal organizer 
. However, understanding how the cortical hem organizes hippocampal development has been difficult. Our results showing that hem-derived CR cells preferentially settle at the marginal zone of the developing dentate gyrus, indicating that hem-derived CR cells may play more important roles in dentate gyrus development.
Previous studies have demonstrated CR cells migrate continuously to the MZ of the hippocampus and neocortex after their birth, and more than 80% of CR cell are born during the time window of E10.5–13.5, the production peak of CR cells appears at E11.5, however the portion of CR cells born later than E13.5 are very small 
. The reason that we chose this specific developmental stage of E13.5 to give tamoxifen is because most CR cells have born at this time and large amount of fimbrilar radial glial scaffold begin to extend from the hem as showing in . Thus the ablation by DT-A expression will mainly disrupt the fimbrilar radial glial scaffold, CR cells born earlier than E13.5 will remain untouched, only a very small portion of. CR cells might be ablated.
The hippocampus is crucial for higher brain functions such as learning and memory. In addition, reelin deficiency in human beings leads to lissencephaly, malformations of the hippocampus and cerebellum and severe epilepsy 
. Furthermore, decreased level of reelin expression in the hippocampus is related to schizophrenia. These findings suggest the possible involvement of hem-derived CR cells and its secreted protein reelin in the pathogenesis of certain mental disorders 
. Understanding the development and roles of hem-derived CR cells may help to shed lights on the mechanisms underlying these disorders.