, 2002); from the OPN, Sox14-positive cells extend laterally in the thin layer of cells that make up the nucleus of the optic tract (NOT) ( Figures 2C and 2D). In a more ventral location, Sox14-positive cells cluster at the thalamus-prethalamus border to form the IGL (labeled by Npy expression) with scattered cells in the vLGN ( Figures 2C, 2D, and S1). As at E12.5, all Sox14-positive clusters coexpress the GABAergic marker Gad1 ( Figure S1). GFP-positive axons of Sox14-expressing nuclei extend into the hypothalamus to reach and surround the SCN ( Figure 2C). GFP-positive axons also extend between

the IGL and the PLi and between the PLi and the OPN and CPA ( Figures 2C and 2D). Based on their anatomical location and on http://www.selleckchem.com/products/gdc-0068.html their cross-connections, we define the pretectal and thalamic domains of Sox14-expressing cells as being part of the SVS. To show that Sox14-expressing cells PI3K inhibitors in clinical trials are part of the non-image-forming circuit originating with ipRGCs, we followed the retrograde transsynaptic spread of the Bartha strain of the pseudorabies virus (PRV152tdTomato). Upon injection in the eye chamber, PRV152 spreads through the parasympathetic circuit

controlling the PLR, eventually reaching ipRGCs in the contralateral eye 72 hr after infection ( Figure 2F) ( Pickard et al., 2002; Viney et al., 2007). We have found that at P3, pups are old enough to survive the procedure and expression from the Sox14 locus is still detectable, albeit at reduced levels and in fewer cells than at P2. Colabeling of GFP and tdTomato highlighted several Sox14-positive cells that contained viral particles within the OPN, CPA, and IGL ( Figure 2E). In contrast, hypothalamic nuclei that are

also part of the PLR circuit only contained viral particles but no GFP-expressing cells (SCN and paraventricular nucleus [PVN]) ( Figure 2E). We also noticed very few and isolated viral particles in the LHa, sometimes coexpressed with the Sox14-expressing cells in the region ( Figure 2E). Examination of the Sox14gfp/+ diencephalon at E12.5 did not show GFP-expressing cells at the thalamus-pretectum border Tolmetin or next to the habenula ( Figure 3A). By contrast, at E14.5, GFP-positive cells are visible at the future PLi and extend toward the LHa ( Figure 3A). Given that no progenitor domain other than the ones we described at E12.5 arises at this location, we supposed that GFP-positive cells move to the LHa and PLi by tangential migration. To test this hypothesis, we performed live time-lapse imaging on Sox14gfp/+ diencephalic explants in culture. GFP-positive cells are first seen migrating tangentially from the r-Th toward the pretectum at E12.5 ( Figures 3B and 3C; Movie S1). Migration starts in the ventralmost part of the thalamus and moves dorsally, eventually concerning only the dorsalmost tip of the GFP-positive r-Th at E14.5 ( Figures 3B and 3C; Movie S3). By E15.