Adaptation in cone photoreceptors contributes to an unexpected insensitivity of primate On parasol retinal ganglion cells to spatial structure in natural images
- Department of Physiology and Biophysics, University of Washington, United States
Figures
Figure 1
Differences in spatial integration for gratings and naturalistic stimuli.
(A) Standard subunit model often used to account for nonlinear spatial integration by retinal ganglion cells (RGCs). (B) (Left) On and Off parasol RGC spike responses to contrast-reversing gratings. …
-
Figure 1—source data 1
Source data for Figure 1B.
- https://cdn.elifesciences.org/articles/70611/elife-70611-fig1-data1-v2.xlsx
Figure 2
Synaptic integration leads to unexpected linearity of On parasol responses to natural images.
(A) Spike count for responses of one On parasol retinal ganglion cell (RGC) to image patches and corresponding linear-equivalent stimuli. Each stimulus was flashed for 250 ms and spikes counted …
-
Figure 2—source data 1
Source data for Figure 2A–F.
- https://cdn.elifesciences.org/articles/70611/elife-70611-fig2-data1-v2.xlsx
Figure 3 with 1 supplement
Identifying the components of inhibitory input to On parasol cells elicited by natural image patches.
(A) (Left) Circuits responsible for two components of inhibitory synaptic input to On parasol retinal ganglion cells (RGCs). (Right) The onset of a light increment (top) produces an increase in …
-
Figure 3—source data 1
Source data for Figure 3B–D.
- https://cdn.elifesciences.org/articles/70611/elife-70611-fig3-data1-v2.xlsx
Figure 3—figure supplement 1
PCA-based clustering applied to contrast–response data.
The cluster definitions from responses to natural image patches were applied to responses to contrast increments and decrements. Same cell as Figure 3.
Figure 4 with 1 supplement
Crossover inhibitory synaptic input is necessary and sufficient for linear spatial integration.
(A) Model construction. (left) Architecture of subunit models. Parallel pathways generated excitatory (blue) and inhibitory (red) synaptic input to a retinal ganglion cell (RGC). Each pathway was …
Figure 4—figure supplement 1
Contrast–response functions and fits.
(A, B) Examples of contrast–response functions for excitatory and inhibitory synaptic input to an On parasol cell. Insets show measured responses. Open circles are the integrals of the measured …
Figure 5 with 1 supplement
Time dependence of On parasol responses to flashed and contrast-reversing gratings.
(A) Excitatory and inhibitory synaptic inputs and spike response to a flashed grating. (B) Summary of responses to flashed gratings from seven On parasol retinal ganglion cells (RGCs). (left) Spike …
-
Figure 5—source data 1
Source data for Figure 5B–D.
- https://cdn.elifesciences.org/articles/70611/elife-70611-fig5-data1-v2.xlsx
Figure 5—figure supplement 1
Time dependence of Off parasol responses to contrast-reversing gratings.
The left panel shows the spike response and excitatory and inhibitory synaptic inputs from an example Off parasol cell for conditions identical to the On parasol recordings in Figure 5. The right …
Figure 6 with 2 supplements
Nonlinearities in cone responses to contrast-reversing gratings and implications for On parasol synaptic inputs.
(A) Cone phototransduction currents elicited by contrast-reversing gratings exhibit three properties that could shape the time course of retinal ganglion cell (RGC) responses: (1) a dependence on …
Figure 6—figure supplement 1
Cone model and responses to sinusoidal stimuli.
(A) Schematic of phototransduction cascade and differential equations that comprise the full (i.e., adapting) cone phototransduction model (see for details). R: receptor activity; PDE and P: …
Figure 6—figure supplement 2
Measured and predicted excitatory synaptic inputs in responses to contrast-reversing gratings across a range of frequencies.
Predictions are from the cone/subunit model in Figure 6B.
Figure 7
Minimizing cone adaptation minimizes time dependence of On parasol responses to contrast-reversing gratings.
(A) Approach to minimize nonlinearities in cone responses. Standard grating stimuli were transformed to minimize the difference between the modeled responses of an adapting cone responding to the …
-
Figure 7—source data 1
Source data for Figure 7B–D.
- https://cdn.elifesciences.org/articles/70611/elife-70611-fig7-data1-v2.xlsx
Figure 8
Cone adaptation substantially increases I/E ratio for natural images.
(A) The model from Figure 6B with either adapting or nonadapting cones was used to predict excitatory and inhibitory synaptic inputs during eye movements about a scene. Predicted cone photocurrents …
Figure 9 with 2 supplements
Joint activity of On and Off parasol retinal ganglion cells (RGCs) encodes patches with positive luminance and high spatial structure.
(A) Responses of On and Off parasol models as a function of the spatial contrast (x-axis) and change in mean luminance (y-axis) for a collection of image patches. Panels show predicted …
Figure 9—figure supplement 1
Discrimination of image patches with similar mean luminance (5000 R*/cone/s) but different spatial contrast for joint responses of On and Off parasol cells.
(left) The dashed line shows discrimination when both On and Off parasol retinal ganglion cells (RGCs) integrate linearly over space, the thin line shows discrimination when the the On parasol RGC …
Figure 9—figure supplement 2
On and Off parasol cell responses to the same collection of image patches for two On/Off pairs.
(left) Spike count of responses of Off parasol cell plotted against that of On parasol for a collection of image patches. Insets show highlighted patches. (right) Mean and spatial contrast of each …
