Chemorepellents, including BMPs and Draxin, are released by the roof plate and initially “push” commissural axons ventrally into an increasing Netrin-1 gradient. The floor plate also secretes the morphogen sonic hedgehog (Shh). Like Netrin-1, Shh is a chemoattractant for HCS assay precrossing commissural

axons. Once at the floor plate, commissural axons lose their interest in Netrin-1 and Shh and acquire responsiveness to floor-plate-derived repellents, including slits and semaphorins, allowing them to exit the floor plate and move on to the second leg of their journey (Dickson and Zou, 2010). Remarkably, precrossing spinal commissural neurons exposed to a Netrin-1-deficient floor plate in the presence of Shh signaling inhibitors show residual attraction, indicating the existence of additional, unidentified floor plate attractant(s) ( Charron et al., 2003). In the developing visual system, retinal ganglion cell (RGC) axons arriving at the chiasm face the same challenge as precrossing axons in the ventral spinal cord: to cross or not to cross the midline. As they approach the optic chiasm, RGCs segregate into ipsilaterally and contralaterally projecting fibers (Figure 1B). Proper crossing, or decussation, at the chiasm is essential for organisms with prominent binocular vision. The mouse has laterally positioned eyes and limited binocular vision.

A large population of RGC axons cross the midline, and a relatively small population does not cross and project ipsilaterally.

Seemingly quite different molecular strategies have evolved for proper growth cone navigation at the optic chiasm and spinal cord midline PD0332991 purchase Afatinib in vivo structures. Molecular gatekeepers such as Netrin-1 and Slits are either absent from the optic chiasm or do not directly participate in midline crossing of RGCs. Growth inhibitory cues, on the other hand, are abundant (Erskine and Herrera, 2007). These include the midline repellent EphrinB2, an established guidance cue at the mouse optic chiasm. EphB1 is expressed by ipsilaterally projecting RGCs and EphrinB2 is necessary for the proper formation of these projections. More recent evidence suggests that Shh repels ipsilateral RGC axons at the optic chiasm via its receptor Boc (Fabre et al., 2010). Semaphorin5A and Slits are molecules that define the boundary of the optic pathway but do not directly participate in midline crossing (Erskine and Herrera, 2007). Much less is known about the molecular mechanisms that promote midline crossing at the chiasm. In zebrafish, the secreted semaphorin Sema3D is expressed at the midline and is thought to provide inhibitory signals at the chiasm midline to help channel RGC axons to the contralateral optic tract (Sakai and Halloran, 2006). The cell adhesion molecule NrCAM is expressed at the mouse chiasm and also in a small subset of late born RGCs, and it promotes their midline crossing in vivo (Williams et al., 2006).