Kay Lab

5016 AERI

Jeremy Kay, PhD, PI

Assistant Professor of Neurobiology and Ophthalmology

The lab is located on the 5th floor of the AERI building in the Duke Eye Center
Dept: Neurobiology
Email: kay@neuro.duke.edu
Phone: 919-613-4769
Office: 5004 AERI

Kay Lab external site

We study how neural circuits devoted to specific visual processing tasks arise during development of the retina, and the consequences for circuit function when development goes wrong. The tools of mouse genetics are central to our approach, and we draw on a wide range of molecular, genetic, and imaging methods.

Wiring the retina and visual system ­-- from development to circuit function

We study how the developing nervous system builds neural circuits. We seek to identify cellular and molecular mechanisms that generate different types of neurons, that endow them with specific identities, and that wire them together into circuits with a coherent function. Further, we would like to know what happens to circuit function when these mechanisms go awry, as may happen in neurodevelopmental disorders.

Here’s what motivates our research:


Many organs do their job just fine with a handful of cell types. But each individual part of the brain may have dozens of cell types – if you add that up for the whole central nervous system the number of cell types is easily in the thousands. Moreover, each of those cell types has its own unique dendrite and axon morphology, and unique patterns of connections with specific synaptic partners – this adds tremendously to the brain’s anatomical complexity. It makes intuitive sense, of course, that the brain needs complex anatomy to do all that it does. But we don’t yet understand the specifics – how do the particular anatomical features of particular circuits facilitate the encoding and transmission of information? This question is important not only for basic neuroscience, but also because there are developmental disorders that affect circuit anatomy and cognitive function – understanding why one leads to the other could help in devising treatments for these disorders.


An ideal way to answer this question would be to study a model circuit that is easy to access and manipulate experimentally, that has a well-studied relationship between circuit structure and function, and whose synaptic connectivity is easy to measure anatomically and physiologically. That is why we study the mouse retina – it fits all these criteria, allowing us to use the tools of mouse genetics to subtly perturb circuit anatomy and make specific, testable hypotheses about the effects on circuit function. By taking advantage of all the retina has to offer as a model, we hope to pinpoint key developmental events and circuit wiring features that underlie not only visual function, but brain function more broadly.


The wiring diagram of the retina, and indeed of the brain in general, is established during development by the coordinated maturation of each type of neuron within the tissue – each adopts a type-specific morphology and makes type-specific synaptic connections. Some of these maturation processes are driven by developmental programs intrinsic to each cell, while others involve highly specific cell-cell interactions that impart patterns of synaptic connectivity. If we can identify the genes required to implement these cell-intrinsic and collaborative events, we will move towards understanding at a fundamental level how circuits are assembled. Moreover, the molecules we identify may be candidate targets for therapeutic intervention in neurodevelopmental disorders.







The retina is a convenient laboratory for discovering molecules that assemble circuits, because we can easily see details of circuit wiring that are hidden in other nervous structures. Retinal synapses are organized into parallel laminae – in the inner plexiform layer (IPL), where bipolar and amacrine interneurons synapse on retinal ganglion cells, there are as many as 10 separate sublayers (see figure at right/above). Most neurons project to a specific sublayer. Thus, a simple assay for synaptic partner choice is to ask whether a neuron alters its IPL stratification following a genetic experimental manipulation. Further, all retinal cell types show precise patterning in the plane of the visual field, which is required to prevent “blind spots” (see figure at left). By a simple screen for defects in axon, dendrite, or soma positioning across the retina, we can identify genes required for cell-type-specific neuronal interactions that assemble circuits. Moreover, because the mouse retina is uniquely amenable to genetic manipulation, we can conduct these screens in vivo – a major advantage of our system.

Recent Posts

Juan Valdez-Lopez (seated) with mentor Jeremy Kay
Monday, June 28, 2021
By Dierdre Shipman

Congratulations to postdoctoral associate Juan Valdez-Lopez!  Nominated by mentor and PI Jeremy Kay, Juan has been selected as Duke's Hartwell Fellow for 2021.

Jeremy Kay at microscope
Friday, February 26, 2021

Unexpected Findings Could Lead to Future Gene Therapy for Patients Battling Retinal Degeneration