Visual space is represented by nonmatching topographies of distinct mouse retinal ganglion cell types

Abstract

The distributions of neurons in sensory circuits display ordered spatial patterns arranged to enhance or encode specific regions or features of the external environment. Indeed, visual space is not sampled uniformly across the vertebrate retina. Retinal ganglion cell (RGC) density increases and dendritic arbor size decreases toward retinal locations with higher sampling frequency, such as the fovea in primates and area centralis in carnivores [1]. In these locations, higher acuity at the level of individual cells is obtained because the receptive field center of a RGC corresponds approximately to the spatial extent of its dendritic arbor [2, 3]. For most species, structurally and functionally distinct RGC types appear to have similar topographies, collectively scaling their cell densities and arbor sizes toward the same retinal location [4]. Thus, visual space is represented across the retina in parallel by multiple distinct circuits [5]. In contrast, we find a population of mouse RGCs, known as alpha or alpha-like [6], that displays a nasal-to-temporal gradient in cell density, size, and receptive fields, which facilitates enhanced visual sampling in frontal visual fields. The distribution of alpha-like RGCs contrasts with other known mouse RGC types and suggests that, unlike most mammals, RGC topographies in mice are arranged to sample space differentially.

Figure 1. Cell density and size of A_ON-S retinal ganglion cells form inverse gradients across the mouse retina

(A) Locations of A_ON-S RGCs (yellow dots) labeled by SMI-32 immunostaining (cyan) in a whole mount retina (left eye) from a P24 mouse (V, ventral, D, dorsal, N, nasal, T, temporal). (B) Maximum intensity projections of image stacks encompassing SMI-32 labeled RGC somata and primary dendrites (grey) within the boxed regions in (A). A_ON-S RGCs are identified by their relatively brighter cell bodies and dendrite labeling (yellow dots). (C) Sparse labeling of A_ON-S RGCs in Thy1-YFPH transgenic mice enables visualization of the complete dendritic arbors of individual SMI-32 RGCs (red dots in B). Arrowhead demarcates RGC axon. (D) Dendritic arbor sizes of A_ON-S RGCs represented by diameters of circles with areas equivalent to the arbor area (see Experimental Procedures) are plotted across nasal-temporal retina. Relationships between A_ON-S RGC size (circle diameter) and retinal eccentricity from the optic nerve head (ONH) (E), total dendritic length (F), soma size (G), mean segment length (H) and number of dendritic segments (I). (J) Number of Sholl intersections as a function of the distance from the soma for each cell, and (K) Sholl intersections after normalizing the curves according to dendritic arbor size. (n = 2 retinas, 27 RGCs). See also Figure S1 and S2, Table S1, and Movie S1.

Figure 2. A_ON-S RGCs have mosaic distributions and characteristic ON sustained responses to light stimuli

(A) Example spatial distributions of A_ON-S RGCs (black dots), sampled from nasal (N) and temporal (T) retina. The density recovery profile (DRP) was calculated using a 200 μm radius (red circles). Nearest neighbor distances (grey lines) were calculated for every RGC within a region. (B) DRP plots of A_ON-S RGCs as a function of distance from the reference cell (red) compared to the DRP of cells with a random distribution (black). The effective radius of exclusion for nasal regions is greater than for temporal (grey vertical lines, N = 93.2, T = 46.5 μm, p < 0.0002), but the packing factor was not (T = 0.31, N = 0.36, p = 0.10). The average nearest neighbor distance is greater in nasal compared to temporal retina (N = 168.0, T = 88.4 μm, p < 0.0002). n = 8 retinas for all measures. (C) Left: An example A_ON-S RGC targeted for electrophysiological recording in nasal retina. Dendritic arbors were visualized upon cell-filling with Lucifer yellow (LY), and neighboring A_ON-S RGCs (red dots) were identified subsequently by SMI-32 labeling. Mean neighbor distances for the recorded cell are displayed as red in (B,right). Middle: Targeted A_ON-S RGCs showing co-labeling of LY and SMI-32. Right: Excitatory (red traces) and inhibitory (grey traces) currents show characteristic sustained responses to light stimuli (grey bar, 500 ms). (D) An example A_ON-S RGCs targeted in temporal retina show the same characteristic responses as in nasal retina. Data are shown as mean ± SEM. See also Figure S1.

Figure 3. Increased density of A_ON-S RGCs in temporal retina results in an enhanced sampling of the visual field, which is not paralleled by their dominant presynaptic partner

(A) Axonal terminals of nasal (N) and temporal (T) located type 6 bipolar cells visualized by expression of tdTomato in the Grm6-tdTomato mouse (see Experimental Procedures). Their axonal territory sizes (area of the polygon) are plotted here for cells in nasal and temporal retina. (n = 4 retinas, 37 N and 34 T cells). (B) Plot of the diameter of A_ON-S RGC dendritic arbors against mean nearest neighbor distances. (C) Dendritic field coverage factor (CF) of A_ON-S RGCs in nasal and temporal retina (CFn = 3.9, CFt = 5.1). Coverage factor (colored) maps were calculated from maps of A_ON-S RGCs dendritic field overlap predicted from their neighbor distances (B, see Experimental Procedures). (D) Top: Example dendritic fields from A_ON-S RGCs targeted across the retina, whose receptive field diameters were mapped using sequential bars of light (see Experimental Procedures). Middle: Average excitatory current traces from example nasal (black) and temporal (grey) cells recorded in response to light stimuli at the indicated bar positions. Bottom: Response profiles were fit to the normalized charge transfers (dots) with Gaussian functions (smooth curves). (E) Dendritic diameters of A_ON-S RGCs were reconstructed from the recorded cells and plotted against their respective receptive field diameters (see Experimental Procedures) (n = 14 cells). Receptive fields of temporal cells (grey dots) are larger than their dendritic fields (p = 0.03), (dashed red line shows unity), but the two measures appear similar for nasal cells (black dots (p = 0.4) (See Experimental Procedures). (F) Receptive field CF of A_ON-S RGCs in nasal and temporal retina (CFn = 4.1, CFt = 8.8). Coverage factor maps were calculated for the same nasal and temporal regions in (C). Data are shown as mean ± SEM.

Figure 4. Sampling of frontal visual space is enhanced in A_ON-S RGC distributions, which contrast with the distributions of known RGC types

(A) Azimuthal equilateral projections of retina space for A_ON-S RGC distributions from right (red) and left (blue) eyes shown in Figure 1 and S1, reconstructed and plotted using the retistruct package [41]. Isodensity lines demarcate 5, 25, 50, 75, and 95% contours of the peak density located at the asterisk (~180 cells / mm²). Cyan lines delineate computed sutures of the original relief cuts made for flat mount preparation. (B) Sinusoidal projection of mouse visual space for A_ON-S RGC distributions from retinas in (A) (see Experimental Procedures). Red outline represents the edge of the right retina; blue outline represents the edge of the left retina. N - nasal, D - dorsal, V - ventral, T - temporal, indicate the projection of the corresponding pole of the retina. Grey circle represents the position of the optic nerve head. Note the peak densities for right and left retinas (red and blue asterisks) and increased density (75% and 50% isodensity lines) are biased towards the vertical midline (0) corresponding to rostral frontal visual fields of mice. (C) The density of the total RGC population peaks at a location just nasal and ventral of the optic nerve head (black asterisk) (schematized from [51] see also [42, 52]). (D) In contrast, we show here that A_ON-S and likely A_OFF-S RGCs have peak densities in the temporal-dorsal retina, whereas A_OFF-T RGCs are relatively more uniformly distributed across the retina. (E) Furthermore, the distributions of previously characterized RGCs show varied or flat distributions. The density colormaps in (C and E) are schematics based on previously reported RGC densities, and changes in dendritic arbor sizes (see Experimental Procedures). Density colormaps in (D) are schematics based on the distributions of A_ON-S and A_OFF-T RGCs shown in Figure S1, and predicted from A_OFF-S dendritic arbor sizes illustrated in Figure S2. See also Figure S2.