We therefore propose that selleck chemical VEGF-A is a positive signal for RGC axons and one of the long-sought-after midline factors that promotes commissural axon crossing at the optic chiasm. Because VEGF is expressed in a broad domain around the chiasm, the VEGF164-mediated promotion of

RGC growth must be balanced by repulsive cues that refine the area of axon crossing. Consistent with this idea, the chemorepellents SLIT1 and SLIT2 define the boundaries of the corridor through which RGC axons migrate at the chiasm midline, and loss of these repellents causes RGC axons to cross the midline in an abnormally broad domain (Erskine et al., 2000 and Plump et al., 2002; Figure 8D). NrCAM modulates neuropilin signaling in response to class 3 SEMAs during commissural axon guidance in the anterior commissure (Falk et al., 2005) and spinal cord (Nawabi et al., 2010). Several lines of evidence argue against the possibility that NrCAM modulates neuropilin signaling in response to VEGF164 at the optic chiasm. First, the chiasm defects of mice lacking NrCAM (Williams et al., 2006; data not shown) versus VEGF164 and NRP1 appear distinct. Second, the temporal requirement for

NrCAM versus HDAC inhibitor VEGF164 and NRP1 in contralateral RGC axon guidance differs: defective midline crossing occurs in Nrp1 null and Vegfa120/120 mutants already at E14.0, when the first RGC axons extend through the chiasm ( Godement et al., 1987), while midline crossing in NrCAM null mutants is affected

only late in development, from E17.5 onward ( Williams et al., 2006). Finally, the retinal origin of the excess ipsilateral projections differs, as VEGF164 signaling through NRP1 promotes the contralateral projection of RGCs originating whatever throughout the retina, whereas NrCAM is essential for contralateral growth of a small subset of axons that originate exclusively in the ventrotemporal retina ( Williams et al., 2006). Based on these differences, we conclude that NRP1 and NrCAM function independently of each other to promote contralateral axon growth of RGC axons. In addition to promoting contralateral guidance of RGC axons, we found that VEGF164/NRP1 signaling promotes axon cohesion within the optic tracts. Thus, mutants lacking VEGF164 or NRP1 showed defasciculation of both the ipsilateral and contralateral tract. It is not known if VEGF164 acts as an extrinsic signal in the axonal environment to control fasciculation or, because it is also expressed by RGCs themselves, in a local autocrine fashion. Further in vivo studies, for example with tissue-specific NRP1 knockouts, will be necessary to fully understand this aspect of the phenotype. Interestingly, loss of Dicer, a protein essential for the maturation of regulatory micro RNAs that regulate Nrp1 among several other targets ( Zhou et al., 2008), leads to similar defasciculation and also increases the ipsilateral projection ( Pinter and Hindges, 2010).