This phenotype was observed before TAs interacted with corticofugal axons and before any defects in cortical axons were detected in mutant embryos ( Bagri et al., 2002) ( Figure S6). Thus, in Slit2−/− embryos, in which corridor cells are misplaced, TAs undertake external alternative paths ( Figures 7L and 7P), a situation

highly reminiscent of the chicken embryo. Because TAs express Robo1 and 2 receptors and are repelled by Slit2 (Braisted et al., 2009 and Lopez-Bendito et al., 2007), these pathfinding defects could be due either to a direct effect of Slit2 on axons and/or to an indirect effect via corridor cell positioning. To AZD6244 nmr determine the relative contribution of these modes of Slit2 activity, we first tested using slice culture experiments whether the lack of Slit2/Robo signaling in TAs directly affects their pathfinding (Figure S8). When wild-type thalamic explants are grafted in Slit2−/− coronal slices, TAs grew into the mutant corridor, even though ATM Kinase Inhibitor ic50 Slit2 is lacking in host slices (ncontrol = 12, nSlit2−/− = 12; Figure S8). Thus, consistent with Nrg1 expression in mutant embryos (data not shown), Slit2 inactivation does not affect the guidance properties of corridor cells. Conversely, Robo1−/−;Robo2−/− TAs ( Grieshammer et al., 2004, Long et al., 2004 and Ma and Tessier-Lavigne, 2007) navigate into the corridor of a wild-type coronal

slice (ncontrol = 6, nRobo1−/−;Robo2−/− = 13; Figure S8). Taken together, these experiments show that Slit2 expression within the ventral telencephalon does not directly control TA internal/external pathfinding. To further test the role of Slit2 hypothalamic expression, we grafted Robo1−/−;Robo2−/− thalamic explants

in 45° angle and wild-type slices that contain the entire axonal pathway as well as the hypothalamus ( Figures 7D and 7L). Although some Robo1−/−;Robo2−/− mutant TAs abnormally entered the hypothalamus (n = 6/9; Figure S8), thereby confirming the role of Slit2 in this region ( Braisted et al., 2009), mutant axons that entered the ventral telencephalon followed an internal path similar to wild-type axons (ncontrol = 18, nRobo1−/−;Robo2−/− = 9; Figure S8). Thus, whereas Slit2 prevents TAs from entering the hypothalamus, it does not have a major direct activity in TA positioning within the ventral telencephalon. To address if Slit2 acts indirectly on TAs via corridor cell positioning, we tested whether grafting a wild-type corridor into Slit2−/− mutant slices would be sufficient to rescue TA pathfinding defects in mutant embryos ( Figures 8A–8G). To this end, we used 45° angle slices that contained the entire axonal pathway ( Figures 7J and 8A–8C), in which we performed an initial cut at the border between the ventral telencephalon and diencephalon ( Figures 8A–8C).