Two types of cells have been found in the retina that determine horizontal or vertical orientation, Northwestern Medicine scientists report. They have also demonstrated for the first time how they convey information.

The laboratory of Gregory Schwartz, Ph.D., assistant professor of Ophthalmology and Physiology, investigates the retina, a layer of tissue at the back of the eye that helps the brain process visual surroundings. In particular, the team is using mouse models to characterize each of the 50 different kinds of neurons within the retina, called ganglion cells, that carry information to the brain.

“Unlike the textbook view of it being a glorified camera, the retina is actually a very complex image processing machine, with 50 different output cells. We’re defining those different cells—learning what they do, and how they do it. This study is one example of that,"

Schwartz said.

Orientation-selective Retinal Ganglion Cells

In the study, the scientists identified that there are two types of so-called orientation-selective retinal ganglion cells (OS RGCs) within the mouse retina — one helps to recognize when an object is horizontal, while the other recognizes vertical orientation.

[caption id=“attachment_94581” align=“aligncenter” width=“680”]coplanar dendrite crossings Magnified view of an OFF OS (magenta) and amacrine cell (cyan) coupled network.
Yellow and red arrowheads indicate coplanar dendrite crossings and coplanar dendrite crossings positive for connexin 36 (Cx36).
Inset shows an example crossing with Cx36 puncta. Scale bar = 50 μm.
Credit: Amurta Nath et al. CC-BY[/caption]

The scientists also demonstrated that it is the coupling between ganglion cells and amacrine cells — through what’s known as electrical synapses — that enables these RGCs to recognize orientations.

“Electrical synapses are not typically thought of as the key drivers of feature selectivity; what the cell cares about is thought to come from its regular chemical synaptic input. This is an exception. This cell actually inherits its feature selectivity through an electrical synapse,"

Schwartz said.

Retina Prosthetics

The new insights into the retina could eventually help inform interventions for blindness, including retina prosthetics.

“Retina prosthetics are now coming out, and in order to improve them, it would be helpful to know exactly what kind of computations are going on in different RGCs. The more we learn about how the retina works, the better we can make an artificial one,"

said first author Amurta Nath, a fifth-year doctoral student in the Northwestern University Interdepartmental Neuroscience Program.

[caption id=“attachment_94580” align=“aligncenter” width=“680”]OFF OS RGC circuit model Schematic OFF OS RGC circuit model.
a) Top projection view of the OFF vOS RGC circuit. Dendrites of an OFF OS RGCs (magenta) are electrically coupled to an asymmetric amacrine cell (cyan).
b) Excitatory receptive fields for OFF vOS RGCs and coupled amacrine cells. The excitatory bipolar cell input to OFF vOS RGCs is predicted to be symmetric based on the cell’s dendrites, whereas the excitatory input to the coupled cells is predicted to be asymmetric.
c) Schematic model of neural circuit underlying the OS computation in OFF OS RGCs
Credit: Amurta Nath et al. CC-BY[/caption]

In ongoing research, Nath is now investigating the cell to which these RGCs couple to carry orientation information, which he calls “comet” amacrine cells. Within Schwartz’s laboratory, scientists are also working on genotyping each of the 50 cell types within the retina, among other projects.

Funding for the work comes from a Research to Prevent Blindness Career Development Award, National Institutes of Health Grant, and a Karl Kirchgessner Foundation Vision Research Award.

Amurta Nath & Gregory W. Schwartz Electrical synapses convey orientation selectivity in the mouse retina Nature Communications volume 8, Article number: 2025 (2017) doi:10.1038/s41467-017-01980-9

Top Image: Northwestern University

For future updates, subscribe via Newsletter here or Twitter