To determine to which extent the scattered Pax6+ cells maintain their RG identity and to examine their characteristic radial morphology, we next stained for nestin and RC2 to reveal the

radial glial processes (Figures 3I, 3J, 3M, and 3N). This also revealed scattered cell somata Abiraterone clinical trial and rather disorganized processes, which were no longer aligned and radially oriented in the cKO cerebral cortex (Figures 3N and 3N′) in contrast to controls (Figure 3M). The loss of apical anchoring of radial glial cells (see above as indicated by the scattered Pax6+ cells and below for F-actin analysis) further contributed to the disorganized arrangement of radial glia

cell somata and processes. Indeed, many cells located apically had lost their adherens junction anchoring (Figure S7) but were still able to form points of adhesion as also visible in rosette-like structures MLN8237 nmr sometimes observed between nestin+ cells in the cKO cerebral cortices with processes emanating radially and in rare cases directed to the pial or ventricular surface (Figure 3N′). As it appeared from these stainings that RG processes do no span the radial thickness of the cKO cerebral cortex, we examined this more globally by applying the lipophilic tracer DiO onto the surface of the cortices. As described previously (Malatesta et al., 2000), this label spreads along RG processes to their somata located in the ventricular zone (VZ; Figure 3K). In the cKO cortices, however, DiO-labeled processes were arranged in a very disorganized manner, and only few labeled processes reached the VZ (Figure 3L). Interestingly, the bulk of the DiO-labeled processes ended in the middle of the cerebral cortex, consistent with the idea that the upper and lower halves of the cKO else cerebral cortex are no longer connected by radial processes. The above data suggest that aberrations in radial glial cells, the main guides for migrating pyramidal neurons,

may be responsible for the failure of many neurons to reach their normal position. However, RhoA may also affect neuronal migration directly by affecting the cytoskeleton in migrating neurons as previously suggested (Besson et al., 2004, Heng et al., 2010 and Nguyen et al., 2006). To test these possibilities, we first electroporated CreGFP or GFPonly plasmids into the cerebral cortex of E14 RhoAfl/fl embryos to delete RhoA in few cells and examined RhoA levels 1 day later ( Figure S6). Consistent with the fast reduction of RhoA protein, we observed a notable reduction of RhoA-immuno-reactivity in the electroporated regions compared to neighboring parts, where no electroporated cells were located ( Figures S6A–S6C).