A good possibility to explain the collicular KO phenotype is that the flattening

of the overall ephrinA gradient leaves nasal axons (nn- and n-axons) with insufficient targeting (positional) information to find their proper target zone, resulting in the formation of several TZs at various positions in the caudal SC. Finally, we analyzed the targeting behavior of axons from the temporal periphery (tt-axons), which in wild-type mice form TZs at the very rostral pole (Figure 6E). In full agreement with data published by Pfeiffenberger and colleagues (Pfeiffenberger et al., 2006), we observed (somewhat surprisingly) only very small or no targeting defects of tt-axons in either the collicular (Figure 6F; n = 15), the retinal (Figure 6G; n = 16), or the retinal+collicular ephrinA5 KO (Figure 6H; n = 3). However, in all three KO lines, we did occasionally E7080 research buy observe individual axons that extended caudally past the main TZ. Sometimes these overshooting axons even formed coarse arbors (arrows in Figures 6G and 6H), and in some instances we detected very weak eTZs caudal to the main TZ, particularly in the collicular KO (Figure 6B, arrow; 53% penetrance; see Experimental Procedures).

To further substantiate this finding, we investigated the full KO of ephrinA5 (Figure 6D; n = 4) as well as the ephrinA2/ephrinA5 double KO (DKO; data not SCH 900776 in vitro shown, n = 4). Again, we found only very weak targeting defects for tt-axons in the ephrinA5 full KO with a few axons overshooting caudally, but not forming discernible eTZs (arrows in Figure 6D). The phenotype was more pronounced in the ephrinA2/ephrinA5 DKO; the number of aberrantly projecting axons was markedly increased, but still failed to generate strong eTZs (data not shown). As indicated, these astonishing findings are in agreement with data from Pfeiffenberger et al. (2006). Here it was shown that only in the ephrinA2/ephrinA3/ephrinA5 TKO, and not in the ephrinA2/ephrinA5 DKO, axons from the temporal periphery show robust eTZs, which are confined to the rostral SC (Pfeiffenberger et al., 2006). Cell press We show here that ephrinA5 expression on nasal axons is a key component of repellent axon-axon

interactions, which prevents an intermingling of TZs of temporal and nasal axons during topographic mapping within the central SC. Our data provide in vivo evidence for a guidance principle during retinocollicular map development that is based on target-independent axon-axon interactions. EphrinAs and EphAs show complex expression patterns in the retina and the SC during development of the retinocollicular projection, involving expression of ephrinAs preferentially on nasal axons and of EphAs preferentially on temporal axons. We have revisited in vitro experiments from the Bonhoeffer lab performed in the 1980s, which showed that temporal axons are repelled from contacting nasal axons (Bonhoeffer and Huf, 1980 and Bonhoeffer and Huf, 1985).