To visualize the distribution

of Shh protein in the neural tube, we performed anti-Shh staining on open-book preparations ( Figures 2A and 2B). Shh protein was present in a posterior-high/anterior-low gradient. Plotting the staining intensity versus the relative position along the AP axis showed that the gradient was approximately linear ( Figures 2C and Venetoclax clinical trial 2D; correlation coefficient R2 = 0.88). To our knowledge, this is the first demonstration at the protein level of a diffusible guidance cue accumulating in a gradient along the AP axis. The Shh protein gradient we observed is consistent with observations of a Shh mRNA gradient in the developing chick spinal cord ( Bourikas et al., 2005) and demonstrates that this gradient is conserved in mammals. The presence of Shh in an AP gradient along the floorplate, together with our results showing that Smo is required

cell autonomously for postcrossing commissural axons to turn anteriorly along the AP axis, supports a model where a Shh gradient directs AP guidance of postcrossing Panobinostat commissural axons in mammals. The directionality of the Shh gradient, decreasing anteriorly, implies that Shh acts as a repellent on postcrossing commissural axons. Although Shh has been proposed to function as a guidance cue for postcrossing commissural axons in the chick (Bourikas et al., 2005), Shh gradients have not been shown to directly repel commissural axons in any species. Explant-based assays in which an explant is cultured a short distance from a source of the guidance cue, such as those performed by Bourikas et al. (2005), cannot distinguish between biased outgrowth of axons and actual turning. To test whether Shh gradients can directly cause commissural axons to turn away, we used an in vitro assay for axon guidance through based on the Dunn chamber (Yam et al., 2009). Commissural neurons were dissociated from the dorsalmost part of E13 rat spinal

cords, an age where the dorsal spinal cord is populated mostly by neural precursors and young neurons that have not yet extended long neurites (Helms and Johnson, 1998). Hence, these cells have not been in proximity to the floorplate and are floorplate naive. The neurons were grown in culture and then exposed to a gradient of Shh in the Dunn chamber after a specified numbers of days in culture. With this assay, the turning of axons can be imaged and measured in response to a defined gradient of a chemical cue over a short time period. Because the response of commissural neurons to guidance cues such as Slit changes with the age of the neurons (Stein and Tessier-Lavigne, 2001), we assayed neurons cultured from 2 to 4 DIV (days in vitro).