David M. Berson - US grants

The long range goal of this research is to understand the roles played by the superior colliculus (SC) in visual function. This requires a full accounting of the functional properties of the retinal ganglion cells innervating the SC. In the cat, almost all of these cells belong to the W-cell class. W-cells comprise 50 percent of cat ganglion cells and have obvious equivalents in primates and other mammals. In addition to dominating the retinocollicular pathway, W-cells provide most or all of the retinal input to a host of other visual nuclei that subserve visuomotor reflexes. Despite the ubiquity and importance of W-cells, study of their structure and function has been severely limited by technical factors. This project will overcome these limitations by means of a novel in vitro approach, thereby shedding new light on the organization of the retinocollicular pathway. W-cells are extremely heterogeneous in morphology and function. Rather than representing a true class, W-cells apparently constitute a loose grouping of many distinct ganglion-cell types, each as different from the others as it is from the X-and Y-cell types. Work from this laboratory in past and current project periods indicates that individual W-cell subtypes may exhibit distinctive patterns of collicular projection. The retinocollicular pathway is thus far more complex that has been generally appreciated. Functionally distinct W-cell channels may be processed independently by distinct collicular microcircuits, or may undergo a highly orderly integrative recombination by collicular neurons. Clearly, a prerequisite for understanding the functional meaning of this parallel retinocollicular organization is a detailed understanding of individual W-cell types. The proposed project begins this process through an in depth analysis of a single W-cell type -- the zeta cell -- which is a dominant contributor to the retinocollicular pathway. These W-cells will be fluorescently tagged, either by retrograde transport of tracers from the colliculus or by uptake of a vital dye. The retina will then be maintained in vitro and electrodes advanced toward tagged zeta cells under visual control. The visual response properties of these cells will be studied by extracellular and intracellular recording and their morphology revealed in detail by intracellular dye injection. This approach provides the first practical method for making direct correlations between the morphology, physiology and central projections of single ganglion cells. The method will be used to analyze the stimulus selectivities of these cells, to explore the intraretinal circuitry that produces these response properties, and to characterize the anatomical mosaic through which these cells sample the visual scene. The study will expand our understanding of diversity among the output cells of the retina and the parallel channels of visual information they emit to the SC and other brain centers. It will also establish a powerful new experimental paradigm with extremely broad applicability to the study of retinal ganglion cells.

DESCRIPTION (provided by applicant): Past work under this grant documented the existence of a peculiar type of mammalian ganglion cell that functions as an autonomous photoreceptor. These intrinsically photosensitive retinal ganglion cells (ipRGCs) use the invertebrate-like photopigment melanopsin. They encode ambient light intensity and help to drive various reflexive responses to daylight, such resetting the circadian clock and adjusting pupil diameter. Much of the work in the next grant period is inspired by two surprising lines of pilot evidence. First, we now believe that there is at least one additional type of intrinsically photosensitive ganglion cell. Second, we have evidence that signals from ganglion-cell photoreceptors propagate not only to the non-image-forming visual networks of the brain, but also within the eye itself to other retinal neurons. Targets appear to include certain ganglion cells and dopaminergic amacrine (DA) cells. The influence of ipRGCs on DA cells is apparently reciprocated, and we will study the mutual interactions between these cells in detail. DA cells and ipRGCs are further linked by the fact that both stratify in the same OFF sublamina of the IPL, yet receive a paradoxical synaptically driven ON input of unknown origin. Here, we will work to identify a neural circuit that could account for this input. The specific aims of the proposal are: 1) to characterize the structure and function of a new type of ganglion-cell photoreceptor;2) to characterize the inputs to ipRGCs from dopaminergic amacrine cells and ON bipolar cells;and 3) to characterize the influences of ipRGCs on dopaminergic amacrine cells and other retinal neurons. These studies extend our understanding of the photoreceptive capacity that persists in human outer retinal degenerations such as retinitis pigmentosa. They also bear on fundamental mechanisms of circadian and adaptation modulation of retinal function.

The long term goal of this project is to explore the physiology and functional roles of the intrinsically photosensitive retinal ganglion cells (ipRGCs). The present proposal is to investigate the interactions of ipRGCs with key processes in the developing retina. The ipRGCs are the first functional photoreceptors of the mammalian retina, generating electrical responses to light more than a week before rod and cone photoreceptors are mature enough to affect retinal output. At this age, ganglion cell axons are already establishing and refining their central projections to the visual centers of the brain. This process is thought to be dependent on retinal activity, especially the waves of electrical activity that sweep across the inner retina. During a critical developmental stage (first postnatal week in mice), retinal waves are driven by a network of cholinergic (starburst) amacrine cells which excite each other as well as ganglion cells through nicotinic receptors. These ¿Stage II¿ retinal waves have been considered immune from photic influence due to the immaturity of the classical photoreceptors. However, our preliminary evidence shows that light does, in fact, modulate the behavior of Stage II retinal waves and this influence requires melanopsin, the photopigment of ipRGCs. In return, the waves excite ipRGCs. These bidirectional interactions between retinal waves and ipRGCs are unexpected, and have significant implications for visual system development. The central focus of this renewal application is to explore the nature, mechanisms and functional implications of the bidirectional interactions between ipRGCs and Stage II retinal waves. The specific aims of the proposal are: 1) to determine the synaptic mechanisms by which waves excite melanopsin ganglion cells and how the waves shape the central projections of ipRGCs; and 2) to assess the impact of ipRGCs on retinal waves, the mechanisms responsible for these effects, and their impact on development of retinal projections to central visual targets. Proposed studies will be conducted in wildtype and genetically modified mice and will involve in vitro recordings and pharmacological manipulation of retinal neurons; gene expression profiling; and tracing of retinofugal projections. These studies will help to document an important and novel functional role for ganglion cell photoreceptors, and will clarify mechanisms responsible for their surprising influence on other retinal neurons. They will refine our understanding of the role of lightdriven activity in visual system development and may prompt a reconsideration of the possible impact of lighting environments on visual system development in premature human infants.

? DESCRIPTION (provided by applicant): Serial blockface electron microscopy (SBFEM) is revolutionizing the mapping of neural microcircuits. Small volumes of brain can be fully reconstructed at nanometer-scale resolution, providing a complete description of the form and location of all synaptic inputs to a single cell. For very local inputs, the identity (and synaptic inputs and outputs) of these presynaptic cells can also be reconstructed. This technical breakthrough parallels the transformative impact of genetically modified animals and viruses for characterizing and manipulating molecularly defined neuronal cell types. Using viruses, cre-lox technology, optogenetics, and chemogenetics, distributed functional circuits can be imaged, mapped, activated, silenced or deleted. This proposal seeks to bridge the divide between these approaches. We propose to perfect a method we have devised exploiting molecular-genetic technologies to mark defined cell types with an electron-dense label for SBFEM analysis. Specifically, we have generated a recombinant adeno-associated virus (AAV) that delivers a Cre-dependent genetic construct to infected cells. Exclusively in Cre-expressing cells, the viral payload expresses a membrane-targeted marker protein comprising a fusion of a fluorescent protein (membrane-targeted green fluorescent protein - mGFP) to a recombinant peroxidase enzyme (APEX2). Pilot data show that infected Cre-expressing cells strongly express the fusion protein throughout the membrane (soma, dendrites, axons and terminals). Its bright fluorescence permits detailed confocal analysis; enzyme histochemical processing reveals the same structures by electron-dense marking visible after SBFEM sectioning and imaging. Already, the method has great promise for targeted `connectomic' analysis of Cre-expressing neurons in any brain region and of their output synapses in remote neural structures. Here, we aim to improve and extend the method. Our aims for this proposal are: 1) to optimize the histochemical protocols and design of viral constructs to mark specific cellular structures or compartments without masking synaptic vesicles and other organelles; and 2) to expand the potential applications of the method, by restricting the fluorescent and ultrastructural labeling t neurons that innervate specific targets; and by generating a knock-in mouse line that expresses the marker through mating to a Cre driver line or injection of Cre-expressing viruses.