Steerable-filter based quantification of axonal populations at the developing optic chiasm reveal significant defects in Slit2 −/− as well as Slit1 −/− Slit2 −/− embryos
1 Institute for Adaptive and Neural Computation, University of Edinburgh, Edinburgh, UK
2 Centre for Integrative Physiology, University of Edinburgh, Edinburgh, UK
BMC Neuroscience 2013, 14:9 doi:10.1186/1471-2202-14-9Published: 15 January 2013
Previous studies have suggested that the axon guidance proteins Slit1 and Slit2 co-operate to establish the optic chiasm in its correct position at the ventral diencephalic midline. This is based on the observation that, although both Slit1 and Slit2 are expressed around the ventral midline, mice defective in either gene alone exhibit few or no axon guidance defects at the optic chiasm whereas embryos lacking both Slit1 and Slit2 develop a large additional chiasm anterior to the chiasm’s normal position. Here we used steerable-filters to quantify key properties of the population of axons at the chiasm in wild-type, Slit1−/−, Slit2−/− and Slit1−/−Slit2−/− embryos.
We applied the steerable-filter algorithm successfully to images of embryonic retinal axons labelled from a single eye shortly after they have crossed the midline. We combined data from multiple embryos of the same genotype and made statistical comparisons of axonal distributions, orientations and curvatures between genotype groups. We compared data from the analysis of axons with data on the expression of Slit1 and Slit2. The results showed a misorientation and a corresponding anterior shift in the position of many axons at the chiasm of both Slit2−/− and Slit1−/−Slit2−/− mutants. There were very few axon defects at the chiasm of Slit1−/− mutants.
We found defects of the chiasms of Slit1−/−Slit2−/− and Slit1−/− mutants similar to those reported previously. In addition, we discovered previously unreported defects resulting from loss of Slit2 alone. This indicates the value of a quantitative approach to complex pathway analysis and shows that Slit2 can act alone to control aspects of retinal axon routing across the ventral diencephalic midline.