Asymmetric retinal direction tuning predicts optokinetic eye movements across stimulus conditions

  1. Scott C Harris  Is a corresponding author
  2. Felice A Dunn  Is a corresponding author
  1. Department of Ophthalmology, University of California, San Francisco, United States
  2. Neuroscience Graduate Program, University of California, San Francisco, United States
11 figures, 1 table and 1 additional file

Figures

Figure 1 with 2 supplements
The superior and inferior optokinetic reflex (OKR) are asymmetric in adult mice.

(A) Schematic of behavioral setup to elicit the vertical OKR. The mouse is situated so that one eye is centered in a hemisphere. Stimuli are projected onto the hemisphere’s concave surface via …

Figure 1—figure supplement 1
Example of sinusoidal vertical optokinetic reflex (OKR) before saccade removal.

Eye position (green) is plotted across time as a full-field grating oscillates vertically (lavender). The eye trace includes saccades (i.e., ‘fast nystagmuses,’ as indicated by tick marks: magenta …

Figure 1—figure supplement 2
Baseline vertical eye movements in head-fixed mice (see also Figure 8—figure supplement 4).

Vertical eye movements were measured in response to static gratings to calculate eye drifts for baseline subtraction. (A) Example raw eye trace over 22 s of a static grating. The calculated position …

Figure 2 with 5 supplements
Superior and Inferior ON direction-selective retinal ganglion cells (oDSGCs) have asymmetric spike tuning curves.

(A) Schematic illustrating unilateral bead injections into medial terminal nucleus (MTN) to retrogradely label ganglion cells in the contralateral retina. (B) Flat-mount retina with retrogradely …

Figure 2—figure supplement 1
Two retinal ganglion cell types project to the medial terminal nucleus.

(A) Sagittal section of medial terminal nucleus (MTN) following injection of fluorescent retrobeads (scale bar = 1mm). Dotted line outlines MTN. (B) Retrogradely labeled retinal ganglion cell somas …

Figure 2—figure supplement 2
Additional metrics of ON direction-selective retinal ganglion cell (oDSGC) spike tuning curve width.

(A) Distributions of the distance (in degrees) from each cell’s preferred direction to the point at which its response magnitude first drops below 50% of the response in the preferred direction. …

Figure 2—figure supplement 3
Asymmetries between Superior and Inferior ON direction-selective retinal ganglion cells (oDSGCs) persist under two-photon targeting.

Retrogradely labeled oDSGCs were targeted for cell-attached recordings using a two-photon laser (860 nm). Spikes were measured from Superior and Inferior oDSGCs in response to the drifting bar …

Figure 2—figure supplement 4
Physiological differences between Superior and Inferior ON direction-selective retinal ganglion cells (oDSGCs) are consistent across retinal topography.

(A) Map of retinal locations of all medial terminal nucleus (MTN)-projecting retinal ganglion cells recorded during cell-attached experiments in which epifluorescence targeting was used. D, T, V, …

Figure 2—figure supplement 5
Topographic variation in direction tuning properties across the retina revealed by two-photon targeting.

(A) Map of retinal locations of all medial terminal nucleus (MTN)-projecting retinal ganglion cells recorded during cell-attached experiments in which two-photon targeting was used. D, T, V, and N …

Figure 3 with 2 supplements
Superior ON direction-selective retinal ganglion cells (oDSGCs) receive similar inhibitory inputs but greater excitatory inputs compared to Inferior oDSGCs.

(A) Inhibitory currents measured from an exemplar Superior oDSGC under voltage-clamp at +10 mV in response to a bar drifting in eight directions. Mean peak inhibitory current is presented as the …

Figure 3—figure supplement 1
Superior ON direction-selective retinal ganglion cells (oDSGCs) receive more excitatory input, but are less intrinsically excitable, than Inferior oDSGCs.

(A, B) Linear tuning curve areas of the peak (A) inhibitory and (B) excitatory current measured in voltage-clamp recordings. Horizontal line represents median, box boundaries are IQR, and whiskers …

Figure 3—figure supplement 2
Full-field light increments elicit more spikes and excitation in Superior ON direction-selective retinal ganglion cells (oDSGCs).

(A, C) Example extracellular spike rasters from (A) a Superior and (C) an Inferior oDSGC in response to a 1 s light increment (405 nm). The schematic above shows the timing of the increment relative …

Figure 4 with 1 supplement
Superior ON direction-selective retinal ganglion cells (oDSGCs) have larger dendritic fields and excitatory postsynaptic sites.

(A) Confocal images of exemplar Superior (left) and Inferior (right) oDSGCs filled with dye. Convex polygons are drawn around the tips of their dendrites. (Bottom) Side views of different Superior …

Figure 4—figure supplement 1
Intrinsic electrophysiological properties of ON direction-selective retinal ganglion cells (oDSGCs).

(A) Membrane capacitance, (B) input resistance, (C) resting membrane potential, and (D) spike threshold potential were measured from Superior (magenta) and Inferior (gray) oDSGCs during whole-cell …

Figure 5 with 2 supplements
Thresholding differentiates the tuning properties of Superior and Inferior ON direction-selective retinal ganglion cells (oDSGCs).

(A, B) Exemplar Inferior oDSGC in whole-cell current-clamp during (A) depolarizing and (B) hyperpolarizing current injection in response to a bar moving in eight directions. Numbers on concentric …

Figure 5—figure supplement 1
Spike and subthreshold voltage tuning curves with directionally untuned current injections.

(A, B) Comparison of spike tuning curve metrics from cell-attached and current injection recordings. Histograms show the direction selectivity index (left) and area of the normalized tuning curve …

Figure 5—figure supplement 2
Effects of current injection on intrinsic properties of ON direction-selective retinal ganglion cells (oDSGCs).

To measure the effects of depolarizing and hyperpolarizing current injections on the intrinsic properties of oDSGCs, the (A) peak rate of voltage change and (B) spike threshold potential were …

Figure 6 with 1 supplement
A parallel conductance model demonstrates how untuned excitation contributes to direction tuning.

An exemplar ON direction-selective retinal ganglion cell (oDSGC) was modeled using parameters recorded directly from oDSGCs, including directionally tuned inhibitory conductances for each of eight …

Figure 6—figure supplement 1
The normalized area of model spike tuning curves, but not subthreshold membrane potential tuning curves, is steeply influenced by excitation gain.

Area of the normalized tuning curve for spikes (red) and underlying subthreshold membrane potentials (blue) as a function of the gain of an untuned excitatory input to a model ON direction-selective …

Figure 7 with 2 supplements
Stimulus contrast modulates the spike tuning curves of ON direction-selective retinal ganglion cells (oDSGCs).

(A) Cell-attached tuning curves from an exemplar Superior oDSGC at high (green) and low (tan, 20% relative) contrasts. Numbers on concentric circles indicate spike counts. Dashed lines represent …

Figure 7—figure supplement 1
Stimulus contrast modulates spike tuning curve width but not the ratio of excitation to inhibition.

(A) Area of the normalized tuning curve from spike responses to high-contrast (abscissa) and low-contrast (ordinate) bars drifting in eight directions. Differences between Superior (magenta) and …

Figure 7—figure supplement 2
Two-photon targeting confirms that ON direction-selective retinal ganglion cells (oDSGCs) are contrast sensitive.

(A) Tuning curve area, (B) direction selectivity index, and (C) normalized area of the spike tuning curve were measured in the cell-attached configuration in response to high- and low-contrast …

Figure 8 with 4 supplements
ON direction-selective retinal ganglion cell (oDSGC) responses predict the optokinetic reflex (OKR) across stimulus types, directions, and contrasts.

(A) Schematic of the putative computation between oDSGCs and OKR, consisting of a subtraction between Superior and Inferior oDSGC spikes and a nonlinearity. (B–H) Two separate implementations of the …

Figure 8—figure supplement 1
Behavioral prediction for the optokinetic reflex (OKR) from spike responses to the drifting bar stimulus.

Spikes of Superior and Inferior ON direction-selective retinal ganglion cells (oDSGCs) in response to the drifting bar stimulus were used to predict the magnitude of OKR gain in behaving animals. (A)…

Figure 8—figure supplement 2
ON direction-selective retinal ganglion cell (oDSGC) responses to oscillating gratings.

(A) Oscillating sinusoidal gratings used in oDSGC electrophysiology were equivalent to those used in behavioral experiments. Motion directions apply to all panels. (B) Luminance of a single point in …

Figure 8—figure supplement 3
The optokinetic reflex (OKR) at low contrast.

Eye movements were measured from head-fixed mice in response to an oscillating sinusoidal grating. All parameters of the grating were the same as under high-contrast conditions (Figure 1), except …

Figure 8—figure supplement 4
Baseline vertical eye movements to low-contrast stimuli (see also Figure 1—figure supplement 2).

Vertical eye movements were measured in response to static, low-contrast (20% relative) gratings to calculate eye drifts for baseline subtraction. (A) Example raw eye trace over 22 s of a static …

Author response image 1
Superior oDSGCs are less excitable than Inferior oDSGCs.

(A) Input resistance in Superior (magenta) and Inferior (gray) oDSGCs calculated from current injection ramps in which current increases linearly from 0 to 50 pA over 0.25 seconds. (B) Steady state …

Author response image 2
Strength of excitatory input and spike tuning curve properties.

(A-B) Relationships between excitatory postsynaptic current (EPSC) tuning curve area and (A) direction selectivity index or (B) normalized area of the spike tuning curve. Data taken from cells in …

Author response image 3
(A) Comparison of the direction selectivity index (DSI) computed for inhibitory inputs (from voltage-clamp recordings) and for spikes (cell-attached recordings) for cells in which both metrics were recorded.

There is no significant relationship, indicating that inhibition is a poor predictor of spike tuning, but this may be caused by noise contributed by other circuit and cell-intrinsic processes. R and …

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Escherichia coli)WT C57BL/6J MiceThe Jackson LaboratoryN/A
AntibodyAnti-Lucifer yellow (rabbit polyclonal)InvitrogenCat #A5750; RRID:AB_1501344IF (1:500)
AntibodyAnti-gephyrin (mouse monoclonal)Synaptic SystemCat# 147111; RRID:AB_887719IF (1:500)
AntibodyPSD-95 MAGUK scaffolding protein (mouse monoclonal)NeuromabCat# 75-028; RRID:AB_2292909IF (1:500)
AntibodyAnti-choline acetyltransferase (goat polyclonal)MilliporeCat# AB144P; RRID:AB_2079751IF (1:500)
AntibodyAnti-mouse-Dylight 405 (donkey polyclonal)Jackson ImmunoResearchCat# 715-475-150; RRID:AB_2340839IF (1:1000)
AntibodyAnti-mouse-Alexa 647 (donkey polyclonal)Jackson ImmunoResearchCat# 715-605-151; RRID:AB_2340863IF (1:1000)
AntibodyAnti-mouse IgG, Fc_Subclass 1 Specific-Dylight 405 (goat polyclonal)Jackson ImmunoResearchCat# 115-475-205; RRID:AB_2338799IF (1:1000)
AntibodyAnti-mouse moncolonal IgG, Fc_Subclass 2a Specific-Alexa 647 (goat polycolonal)Jackson ImmunoResearchCat# 115-605-206; RRID:AB_2338917IF (1:1000)
AntibodyAnti-rabbit-Alexa 488 (donkey polyclonal)Jackson ImmunoResearchCat# 711-545-152; RRID:AB_2313584IF (1:1000)
AntibodyAnti-goat IgG (H+L)-Alexa 647 (donkey polyclonal)Jackson ImmunoResearchCat# 705-605-147; RRID:AB_2340437IF (1:1000)
AntibodyStreptavidin 488 conjugate antibodyMolecular ProbesCat# S32354; RRID:AB_2315383IF (1:400)
Chemical compound, drugAmes’ MediumUnited States BiologicalCat# A1372-25
Chemical compound, drugVectashieldVector LaboratoriesCat# H-1000; RRID:AB_2336789
Chemical compound, drugRed RetrobeadsLumafluorhttps://lumafluor.com/information
Chemical compound, drugLucifer yellow CH dilithium saltSigmaCat# L0259
Chemical compound, drugBiocytinInvitrogenCat# B1592
Software, algorithmAmiraThermo Fisher Scientifichttps://www.fei.com/software/amira-avizo/; RRID:SCR_014305
Software, algorithmBassoonHarris, 2022https://doi.org/10.5281/zenodo.6757605; RRID:SCR_023333
Software, algorithmIgor ProIgor ProRRID:SCR_000325
Software, algorithmImageJNIHhttps://imagej.nih.gov/ij/;
RRID::SCR_003070
Software, algorithmImarisBitplanehttp://www.bitplane.com/;
RRID:SCR_007370
Software, algorithmMATLABMathWorkshttps://www.mathworks.com/products/matlab.html;
RRID:SCR_001622
Software, algorithmMeshmapperPaul Bourkehttp://paulbourke.net/dome/meshmapper/
Software, algorithmObjectFinderDella Santina et al., 2013https://github.com/lucadellasantina/ObjectFinder; https://zenodo.org/record/4767847; RRID:SCR_023319
Software, algorithmPsychopyOpen Science Tools Ltd. Peirce, 2007https://psychopy.org/about/index.html
Software, algorithmScanImageMBF Biosciencehttps://www.mbfbioscience.com/products/scanimage
Software, algorithmStreamPixNorPixhttps://www.norpix.com/products/streampix/streampix.php
Software, algorithmSymphony and StageCafaro, 2019https://github.com/Symphony-DAS/symphony-matlab; https://github.com/Stage-VSS/stage-v1
Software, algorithmVolumeCutDella Santina et al., 2021; Della Santina, 2021https://github.com/lucadellasantina/VolumeCut; https://doi.org/10.5281/zenodo.5048331

Additional files

Download links