Genetic dissection of EphA receptor signaling dynamics during retinotopic mapping

Abstract

Retinal ganglion cells (RGCs) project axons from their cell bodies in the eye to targets in the superior colliculus of the midbrain. The wiring of these axons to their synaptic targets creates an ordered representation, or "map," of retinal space within the brain. Many lines of experiments have demonstrated that the development of this map requires complementary gradients of EphA receptor tyrosine kinases and their ephrin-A ligands, yet basic features of EphA signaling during mapping remain to be resolved. These include the individual roles played by the multiple EphA receptors that make up the retinal EphA gradient. We have developed a set of ratiometric "relative signaling" (RS) rules that quantitatively predict how the composite low-nasal-to-high-temporal EphA gradient is translated into topographic order among RGCs. A key feature of these rules is that the component receptors of the gradient--in the mouse, EphA4, EphA5, and EphA6--must be functionally equivalent and interchangeable. To test this aspect of the model, we generated compound mutant mice in which the periodicity, slope, and receptor composition of the gradient are systematically altered with respect to the levels of EphA4, EphA5, and a closely related receptor, EphA3, that we ectopically express. Analysis of the retinotopic maps of these new mouse mutants establishes the general utility of the RS rules for predicting retinocollicular topography, and demonstrates that individual EphA gene products are approximately equivalent with respect to axon guidance and target selection.

Figure 1.

Properties of the Isl2-EphA3 knock-in mouse. A, EphA expression pattern in the retina of Isl2-EphA3 knock-in mice. Approximately every other RGC is either Isl2 expressing (purple circles) or non-Isl2 expressing (white circles) along the nasal-temporal axis of the retina (x-axis). All RGCs express an amount of total EphA (ΣEphA—black line) that increases from nasal to temporal position. Isl2⁺ RGCs express an additional amount of EphA3 (purple line) that increases the total EphA expressed in these cells. B, The effect of ectopic EphA3 expression on retinocollicular mapping. Left, A focal injection of a tracer dye (DiI) into the retina of a wild-type mouse (bracket) labels both Isl2⁺ and Isl2⁻ RGCs. Adjacent Isl2⁺ and Isl2⁻ RGCs expressing the same amount of EphA have the same sensitivity to the high-caudal-to-low-rostral gradient of ephrin-As present in the SC (gray gradient), and map to the same position (TZ) in the SC (white asterisk). Right, A focal DiI injection into the retina of an Isl2-EphA3 knock-in mouse labels both Isl2⁺ and Isl2⁻ RGCs. Isl2⁺ RGCs express ectopic EphA3, making these RGCs more sensitive to the ephrin-A gradient in the SC. These Isl2⁺ RGCs therefore map to a more rostral location in the SC (purple asterisk) than their Isl2⁻ neighbors (white asterisk).

Figure 2.

Calculation and predictions of the local relative signaling ratio (R_lrs). A, The EphA retinal gradient of a wild-type mouse is composed of EphA4 (red), EphA5 (blue), and EphA6 (green). Each EphA mRNA is plotted as a function of position on the NT axis of the retina (x), based on curve fitting of the measured expression levels of each mRNA in the study by Reber et al. (2004). EphA4 is ungraded along the NT axis and is plotted as a constant (1.05 relative units; red term). EphA5 is graded from low nasal to high temporal with a function of 0.14e^0.018x relative units (blue). EphA6 is similarly graded, with a function of 0.09e^0.029x relative units (green). ΣEphA is the total amount of EphA expressed by an RGC at position x on the NT axis of the retina and is plotted as a linear summation of the functions for each of the endogenously expressed EphAs. B, In Isl2-EphA3 knock-in mice, Isl2⁺ RGCs express a constant amount of ectopic EphA3 regardless of their position in the retina; thus, the ΣEphA function for Isl2⁺ RGCs (ΣEphA^ki) is simply the addition of an additional constant, corresponding to the measured EphA3 expression level, to the endogenous ΣEphA function (Reber et al., 2004). The R_lrs function is calculated by dividing ΣEphA^ki at position x by ΣEphA for the same position x. C, The ΣEphA^ki and ΣEphA functions for all possible compound Isl2-EphA3 knock-in/EphA5 knock-out mutants. D, Plots of the R_lrs functions of all possible Isl2-EphA3 knock-in/EphA5 knock-out mutants. Those genotypes for which the R_lrs function falls below the discrimination limit ratio (dashed horizontal line) (Reber et al., 2004) are predicted to display collapsed maps, with collapse occurring where the R_lrs function crosses the discrimination limit ratio (arrows). The Isl2⁻-EphA3^ki/+/EphA5^+/− and Isl2-EphA3^ki/+/EphA5^−/− maps are predicted to collapse at 85% and 95%, respectively. The Isl2-EphA3^ki/ki/EphA5^+/− and Isl2-EphA3^ki/ki/EphA5^−/− maps are predicted to be fully duplicated.

Figure 3.

Visualization and measurement of the retinocollicular maps of Isl2-EphA3 knock-in/EphA5 knock-out compound mutant mice. A, The injection site in the retina is visualized in a whole-mount preparation, an example of which is shown here. The location of the injection site is mapped as percentage of the nasal-temporal axis of the retina, with 0% being the nasal pole and 100% being the temporal pole. The SC corresponding to this injected retina is shown in B. B, The two termination zones of labeled RGC axons in the SC are visualized in a whole-mount preparation, an example of which is shown here. The location of the termination zones is mapped as percentage of the rostral-caudal axis of the SC, with 0% being the rostral extreme and 100% being the caudal extreme. This SC corresponds to the retina of A and displays two distinct areas of labeling that are measured as two separate termination zones. Examples in A and B are from an Isl2-EphA3^ki/ki/EphA5^+/− mouse, which is predicted to have a fully duplicated retinocollicular map. C, The injection site and retina corresponding to the SC in D. D, The termination zone of the retina seen in C. A single, well defined TZ is visible. Examples in C and D are from an Isl2-EphA3^ki/+/EphA5^−/− mouse, which is predicted to have a retinocollicular map that collapses in the temporal retina. E, Data points are plotted on a Cartesian graph in which the x-axis is the NT position of the retinal DiI injection and the y-axis is the CR position of the collicular TZ(s) labeled by this injection. The points plotted in this example correspond to the labeled TZs from B (green) and A (red). Upward-pointing triangles correspond to the more caudal TZs of a duplicated pair; downward-pointing to more rostral. F, A map for any given genotype is made by plotting multiple map points, obtained from systematic DiI injections across the full NT extent of the retina, from mice of the same genotype. Example sections of the SC of multiple animals are shown at the position of the x-axis corresponding to the location of their respective DiI injections in the retina (retinas not shown). The blue line connects the data points and represents the continuity of the map in vivo. The example points shown here correspond to the Isl2-EphA3^ki/+/EphA5^+/− genotype, a complete map of which is shown in Figure 4A.

Figure 4.

Anterograde labeling of Isl2-EphA3 knock-in, EphA5 knock-out compound mutant retinocollicular maps. A, Retinocollicular map of Isl2-EphA3^ki/+/EphA5^+/− compound mutant. The position of the injection site in the retina is graphed on the x-axis. The position of TZ(s) in the SC is graphed on the y-axis. Each pair of points corresponds to a single retinal injection in a single mouse. The upward-pointing triangles correspond to the caudal TZ in an animal with two TZs. The downward-pointing triangles correspond to the rostral TZ in an animal with two TZs. The hourglass shapes correspond to a collapsed TZ in an animal showing a single TZ. The collapse point is located at ∼85% of the NT axis, which is at the predicted position. B, Retinocollicular map of Isl2-EphA3^ki/+/EphA5^−/− compound mutant. As predicted, the map collapses. The collapse point is located at ∼90% of the NT axis, which is very close to the predicted position. C, Retinocollicular map of Isl2-EphA3^ki/ki/EphA5^+/− compound mutant. As predicted, the map is fully duplicated. D, Retinocollicular map of Isl2-EphA3^ki/ki/EphA5^−/− compound mutant. As predicted, the map is fully duplicated.

Figure 5.

Population relative signaling ratio and mapping density. A, The retinocollicular maps for mice of the Isl2-EphA3^ki/+/EphA5^+/− genotype (blue lines) and Isl2-EphA3^ki/+/EphA4^+/− genotype (red lines) (Reber et al., 2004) are shown. The lower line of each map corresponds to the location of Isl2⁺ RGCs. The area covered with vertical lines shows the area of the SC where Isl2⁺ RGCs map in Isl2-EphA3^ki/+/EphA5^+/− mice (blue lines) and Isl2-EphA3^ki/+/EphA4^+/− mice (red lines). B, The total amount of EphA expressed by a population of RGCs can be calculated by integrating the function for ΣEphA for that population of RGCs for x = 0 to x = 100, which corresponds to the entire NT axis of the retina. The integral for ΣEphA^ki (purple) will always be larger than that for ΣEphA (black) in a knock-in mouse. C, The population relative signaling ratio R_prs is calculated for any given genotype by dividing the integral of the ΣEphA^ki function by the integral of the ΣEphA relevant for that genotype. D, The R_prs values for Isl2-EphA3^ki/+ compound mutants (x-axis) and the percentage of the SC occupied by Isl2⁺ RGCs (y-axis) are shown.

Figure 6.

General relative signaling prediction of the EphA5^−/− map. A, The general relative signaling ratio (R_grs) function is calculated by dividing the ΣEphA value at the temporal pole of the retina by the ΣEphA value at position x. The R_grs function of EphA5^−/− mice is shown. The value of the R_grs is plotted on the y-axis for each position along the NT (x) axis of the retina. B, The EphA5^−/− retinocollicular map, as determined by repeated DiI injections across the NT axis of the retina, is shown (blue triangles) superimposed on the R_grs function (dashed blue line).