These results indicated that GPC1 is capable of binding Shh. Based on previous studies in flies and vertebrates, GPC1 and Shh could cooperate in two different (but not necessarily exclusive) manners to mediate postcrossing commissural axon guidance: (1) GPC1 could directly promote or inhibit Shh’s interaction with its axon guidance receptors (Beckett et al., 2008, Capurro et al., 2008 and Williams et al., 2010), and/or

(2) the presence of GPC1 within a receptor complex could regulate Gli-dependent transcription and subsequent gene expression in response to Shh (Chan et al., 2009), which in turn would specify the expression of guidance receptors on commissural LY294002 in vitro axons. Here, we investigated the latter. Gene transcription has been demonstrated to regulate discrete steps in postcommissural axon guidance (Condron, learn more 2002), and Shh has been speculated to be an appropriate floorplate-derived signal that could induce such an activity (Sánchez-Camacho and Bovolenta, 2009). However, so far evidence for such a mechanism has been elusive. Our previous studies identified Hhip as a mediator of the repulsive guidance response to Shh (Bourikas et al., 2005). Hhip messenger RNA (mRNA) is detectable transiently in dI1 neurons at the time when postcrossing axons turn into the longitudinal axis ( Figure S5; Bourikas et al., 2005). Interestingly, Hhip is a transcriptional target of Shh (

Chuang and McMahon, 1999 and Buttitta et al., 2003), suggesting that commissural neurons might begin to upregulate Hhip as they encounter high levels of Shh in the floorplate. Thus, we hypothesized that transcriptional activity in response

to Shh, in a GPC1-dependent manner, could modulate the responsiveness of the commissural growth cone at this intermediate target. To investigate this idea, we analyzed Hhip mRNA expression patterns in the spinal cord following GPC1 knockdown. Strikingly, we found that embryos electroporated with βact-hrGFPII-mi7GPC1 ( Figure 4A) or βact-hrGFPII-mi4GPC1 (data not shown) displayed a specific until loss of Hhip expression in the dorsal spinal cord on the electroporated side. In contrast, ventromedial Hhip expression was unaffected by the loss of GPC1, demonstrating a cell-type-specific requirement for GPC1 in Hhip induction. Electroporation of a control plasmid, βact-hrGFPII-mi2Luc, did not affect Hhip expression ( Figure 4B). Rescue experiments, as described above, revealed that dorsal expression of Hhip could be restored by the expression of GPC1ΔmiR ( Figures 4C–4F). We quantified these effects using two methods. First, we calculated the percentage of sections in each condition displaying “symmetrical” versus “asymmetrical” Hhip levels in the dorsal spinal cord ( Figures 4C–4F, percent values indicate the number of sections with symmetrical expression).