Figure 7

Graphical summary. (A). Expression patterns of slit2 and slit3 in a wild-type zebrafish embryo, around the stage of retinal axon crossing at the chiasm, as determined by fluorescent in situ hybridization, and described in Chalasani et al. [19] and Davison and Zolessi [17]. (B). Model illustrating how Slit2 and Slit3 might be cooperating in restricting RGC axons to the optic pathway around the optic chiasm, based on the original model described by Hutson and Chien [22]. Slit2 expression in a narrow area tightly associated with the optic nerve and tract would generate a short but steep gradient towards the pathway center. Slit3, on the other hand, is expressed in apparent larger amounts, but at a farther distance, probably making a longer and shallower gradient around the chiasm. The summation of both gradients would result in a “U” shaped Slit gradient, and axons would be able to grow just at the bottom of this “U”. (C). Application of this model to the four experimental situations analyzed in the present report. If no Slits are present, retinal axons are free to grow, but make guidance errors that are not corrected, because there is no repulsive channel. Slit3 deficiency leaves only the Slit2 gradient, which is still strong enough to correct most errors; this deficiency would also slightly increase slit2 expression, partially restoring the wild-type U gradient (dotted line). When there is a Slit2 deficiency, the remaining Slit3 gradient also allows for some uncorrected errors; there is no compensation in this case. Finally, in the wild-type situation, the full Slit (Slit2 + Slit3) “U”-shaped gradient keeps axons on track by correcting all errors. NR: neural retina; OC: optic chiasm; ON: optic nerve; OT: optic tract.