Taken together, these results indicate that VEGF chemoattracts commissural axons through Flk1. To analyze whether Flk1 also functionally regulated commissural axon guidance in vivo, we inactivated Flk1 specifically in commissural neurons by crossing Flk1lox/LacZ mice with the Wnt1-Cre driver line, which induces Cre-mediated recombination in commissural neurons in the click here dorsal spinal cord ( Charron et al., 2003). We and others previously described that intercrossing Flk1lox/lox mice with various Cre-driver lines resulted only in incomplete inactivation of Flk1 ( Maes et al., 2010 and Ruiz de Almodovar et al., 2010). In order to increase the efficiency of Flk1
excision and to obtain complete absence of Flk1 in commissural Buparlisib solubility dmso neurons, we intercrossed Wnt1-Cre mice with Flk1lox/LacZ mice that carry one floxed and one inactivated
Flk1 allele in which the LacZ expression cassette replaces the first exons of Flk1 ( Ema et al., 2006). PCR analysis confirmed that the floxed Flk1 allele was correctly inactivated in the spinal cord from E11.5 Wnt1-Cre(+);Flk1lox/LacZ embryos (referred to as Flk1CN-ko embryos) (data not shown). Spinal cord sections from E11.5 Flk1CN-ko embryos immunostained for Robo3 revealed that precrossing commissural axons exhibited abnormal pathfinding, projected to the lateral edge of the ventral spinal cord, invaded the motor columns and were defasciculated ( Figures 5A–5G). Such aberrant axon pathfinding was only very rarely observed in control E11.5 Wnt1-Cre(–);Flk1lox/LacZ (Flk1CN-wt) embryos, which still express functional Flk1 ( Figures 5A, 5D, and 5G). Morphometric analysis confirmed that the area occupied by Robo3+ axons was significantly larger and that these guidance defects were more frequent in Flk1CN-ko than Flk1CN-wt embryos ( Figure 5H). Similar to what we found in VegfFP-he mouse embryos, the pattern and level of Thalidomide expression of Netrin-1 and Shh were comparable
between Flk1CN-ko and their corresponding wild-type littermates ( Figures S3A–S3D), indicating that Flk1 cell-autonomously controls guidance of precrossing commissural axons in vivo. To assess how specific the role of VEGF and Flk1 in commissural axon guidance is, we analyzed the expression and role of additional VEGF homologs that can bind to murine Flk1 (VEGF-C) or indirectly activate Flk1 (Sema3E) (see Introduction). ISH revealed that VEGF-C was not expressed at the floor plate or ventral spinal cord at the time of commissural axon guidance (Figure S1C). In addition, VEGF-C did not induce turning of commissural axons in the Dunn chamber assay (Figure S4A). Consistent with these in vitro findings, homozygous VEGF-C deficiency did not cause commissural axon guidance defects in vivo (data not shown). Through binding Npn1/PlexinD1, which forms a signaling complex with Flk1, Sema3E is capable of activating Flk1 independently of VEGF (Bellon et al., 2010).