David A. Feldheim - US grants

Description (provided by applicant): The experiments we propose aim to resolve two important issues in developmental neuroscience. The first is the long-standing debate as to the relative importance of mapping molecules ("nature") and activity-dependent processes ("nurture") toward the development of CNS connectivity. We have started to answer this question with respect to the development of topographic maps in the mouse visual system. In Aim 1, we will determine the developmental consequences on the retinocollicular map when ephrin-As, patterned neural activity, and both ephrin-As and neural activity are disrupted in vivo. We will determine how each of these mechanisms specifically acts to help form maps, the extent to which they can compensate for each other, and if topography is required to develop normal receptive field responses of target neurons. We will also compare these results with those obtained in cortical visual areas, to determine if different brain areas use these mechanisms differentially. A secondary goal of Aim 1 is to determine the mechanisms by which topographic maps align. The SC receives inputs from multiple regions of the brain, which are arranged such that they are in register with the visual world. We have designed experiments that will test the hypothesis that a combination of ephrin-As and neural activity will also be used to map and align the corticocollicular projection with that of the retinocollicular projection, but with a larger relative importance of activity-dependent mechanisms. These experiments will take advantage of our findings that EphA3-ki mice and ephrin-A2/A3/A5 tko mice have SC and V1 maps that differ in structure. Analysis of these mice will allow us to determine the extent to which the brain can adapt in structure or function to create a cohesive visual world when cortical and collicular maps have different topographic structures. Experiments proposed in Aim 2 will resolve mechanistically how topographic maps form. Multiple models for topographic mapping have been proposed and each is consistent with much of the published experimental in vivo and in vitro data. We plan to determine which, if any, of these models is true in two ways. First, we will determine the retinal vs. collicular contributions of ephrin-A5 in mapping, by removing ephrin-A5 specifically from the retina or SC, using conditional knock out technology. Second, we to determine the role of axon-axon competition in topographic map formation by analyzing the retinocollicular maps in mice that have reduced numbers of RGCs and, therefore, reduced competition for target space in the SC. PUBLIC HEALTH RELEVANCE The formation of precise neuronal connections is strictly required for productive communication between neurons. Understanding the basic processes that specify proper connectivity in the visual system will be directly relevant to treating neurological disorders involving aberrant neuronal connections and processing, such as generalized seizures, sleep disorders, and mental retardation. In addition, it is likely that the same mechanisms used to make neuronal connections during development can be manipulated in order to rewire the brain after damage due to injury or disease.

[unreadable] DESCRIPTION (provided by applicant): Most sensory input to the brain is mapped topographically, with nearest neighbor relationships of the projecting neurons maintained in their connections within target areas. For example, retinal ganglion cell (RGC) axons project topographically to retinal targets in the brain, allowing visual images to be transferred in a spatially intact form. It is believed that topographic maps form using gradients of axon guidance molecules expressed in both projection and target neurons. Although there is a consensus that gradients are used in topographic mapping, many questions and hypotheses about how axons read molecular gradients and branch at their appropriate sites remain unknown. This proposal describes a new method for fabricating density gradients of biomolecules on solid substrates. The approach employs a hierarchical strategy in which proteins are first attached to nanometer-sized metal particles, and the resulting protein-nanoparticle bioconjugates are subsequently assembled from solution onto a substrate surface. The resulting surface topography can be characterized by monitoring the large and distinctive visible light absorptions of metal nanoparticles by imaging the nanoparticles with atomic force microscopy. This method of nanoscale surface engineering is capable of forming molecular gradients of varying concentration and slope, in one and two- dimensions, and does not require expensive or sophisticated equipment. Moreover, because gold and silver nanoparticles absorb visible light of different energies, surface density gradients containing two different proteins on a single substrate may be assembled and characterized. Two dimensional protein patterning will increase the complexity of in vitro cell-based assays significantly. The utility of surfaces containing protein density gradients will be illustrated through studies of topographic mapping in the visual system. One and two-dimensional surface gradients of EphA receptors and ephrin-A ligands will be constructed in a manner that recapitulates their graded expression during visual system development. Surface-bound neuronal growth assays will deepen our understanding of the molecular mechanisms of axon guidance by more closely mimicking protein patterning encountered in living systems. This understanding will not only allow us to understand how neural connections are made in the brain but will also guide us to be able to rewire connections after injury. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): The goal of this co-principal investigator, interdisciplinary proposal is to develop techniques for the comprehensive functional characterization of retinal ganglion cell (RGC) types in the mouse retina, and to combine this characterization with mouse transgenic technology in order to determine the relationships between the morphology and physiology of RGC types. Our experiments take advantage of novel multi- electrode array (MEA) recording systems built in the Litke lab. These systems contain over 500 electrodes and can simultaneously record the activity from hundreds of neurons in an intact retina;this represents a 10 fold better yield over currently available technology. We find that this increase in yield is critical for unambiguous functional classification and reliable characterization of the many RGC types in the retina. Furthermore, we hypothesize that the detailed information provided by these MEA systems will make it possible to match the physiologically identified neurons to optically imaged RGCs. Such a match will create a link between function and structure in an unprecedented manner. We have two main aims to accomplish our goals. In the first aim we propose to use two types of large- scale MEAs, a 512-electrode array with 605m interelectrode spacing, and a high density 519-electrode array with 305m spacing, to characterize the receptive field and mosaic properties of RGC types of wild type mice. We have chosen to use these arrays to characterize RGC types in mice because the mouse retina has become an important model to study the role various genes/molecules play in the development of retinal circuitry. The mouse also serves as a model for studying the progression of retinal-degenerative diseases such as glaucoma. In the second aim we plan to use the large-scale MEA technology to correlate the morphological RGC types to their functional counterparts. In addition to being classified using physiological criteria, RGCs are classified using morphological criteria. However, only in rare circumstances can cells be classified by both morphological and physiological properties. Experiments proposed in this aim are designed to correlate morphologically labeled cells with their physiological properties. We will do this by matching the morphological images of GFP marked RGCs with their electrophysiological images and receptive fields. (As described in section c, the electrophysiological image is a technique, developed by the Litke lab, for imaging the spatiotemporal pattern of electrical activity generated by individual neurons.). The purpose of this grant is to obtain a comprehensive characterization of retinal ganglion cell (RGC) types in the mouse retina. This knowledge is essential in order to understand how various molecular and environmental perturbations affect the retina's development and function. Vision is a crucial component of human perception and blindness is a devastating affliction. Understanding what the retinal circuits are is the first step toward understanding how they develop and is also essential to better understand the progression of retinal degenerative diseases (Are specific circuits differentially affected in disease?). In the last 10-20 years, modern molecular techniques, in combination with powerful advances in imaging and electrophysiology, have led to an increase in the use of the mouse as a model system for studying retinal circuitry. It is likely that the knowledge obtained from experiments proposed here will be essential to all who use the mouse visual system as a model. PUBLIC HEALTH RELEVANCE: The first aim of this project is to develop techniques for the classification and functional characterization of retinal ganglion cell (RGC) types in the mouse retina. The second aim is to combine this characterization with mouse transgenic technology in order to determine the relationship between the morphology and physiology of RGC types. Upon completion of the proposed aims, we will be in a strong position to use these techniques to determine how various genetic, activity-related, and disease perturbations affect and control the development of neural circuits. This will build a solid foundation for future work aimed at developing therapies for treating retinal damage due to injury or disease.

Project Summary The retina transforms the visual scene into ~30 different channels of information, each mediated by a unique type of retinal ganglion cell (RGC). RGC types differ in their morphology, functional response properties, brain targets, and the behaviors in which they participate. For reasons that are not yet known, different RGC types have different susceptibilities to cell death after optic nerve injury. Several fundamental aspects of each RGC type remain unknown: how each develops and maintains its unique characteristics, how each contributes to the visual response properties of neurons in the brain, and contributes to visually induced behaviors. Filling this gap in knowledge is not only a prerequisite to understand how the visual system functions, it is also necessary if we are to either salvage or regenerate RGCs that can integrate into the retinal circuitry and maintain their function after injury or disease. Satb1 and Satb2 are expressed in mouse DS RGCs and removing Satb1 and Satb2 during development leads to the loss of direction selective retinal responses. Experiments in Aim 1 are designed to elucidate the role of Satb1 and Satb2 in RGC development and to test the hypothesis that On-Off DS RGCs contribute to the formation of direction selective cells in the brain and to motion visual behavior. In Aim 2 we will test the hypothesis that Tbr2 is necessary and sufficient for the development and maintenance of non-image forming RGCs. Tbr2 is expressed in newly differentiated RGCs and expression continues into adulthood. Mice lacking retinal Tbr2 during development fail to form non- image forming circuits and lose light-dependent reflexes. We will test the hypothesis that Tbr2 is required for maintaining a non-image forming RGC fate and by over expressing Tbr2 we will determine if Tbr2 is sufficient to change developing or adult RGCs to a non-image forming fate. In Aim 3 we will examine if Tbr2 is necessary and sufficient for neuroprotection of RGCs after injury. The optic nerve crush assay is used as a model to study the mechanisms counter to degeneration, namely RGC survival and regeneration. We find that Tbr2 expressing RGCs are selectively spared after nerve crush. Experiments in this aim will determine the role of Tbr2 in RGC survival. Upon completion of these aims we will significantly advance our understanding of how direction selectivity is generated during development, discover mechanisms used to maintain RGC health and function after its incorporation into circuits, and discover new mechanisms of neuroprotection.

Project Summary Here we propose to develop and integrate a suite of experimental and computational tools to measure the visual response and network properties of a large population of neurons in the mouse superior colliculus (SC), to determine how these properties change during locomotion, and the contribution of cortical and specific retinal inputs to these properties. The mouse SC is a subcortical area that integrates vision with touch and hearing to initiate orienting movements of the eyes and head, and is an attractive model to study how specific circuits form during development. Our development of high-density, high-channel count silicon probes to record neural activity has several significant advantages compared to alternative methods: (1) high efficiency for recording neuron spatial and temporal visual response properties; (2) the ability to rapidly study the topological/functional organization in a large neuron population over a wide field of view in a uniform way in a single animal; (3) the possibility to study correlated activity and connectivity among neurons as well as network rhythms; (4) the ability to ascertain differences in visual responses associated with behavioral state such as locomotion. Experiments proposed in Aim 1 will measure the functional and topological properties of visually- responsive neurons in the SC of mice that are awake and head-fixed on a freely-floating Styrofoam ball used as a spherical treadmill. For each neuron, the spatial receptive field (RF), the temporal filtering spike-triggered average (STA), direction and orientation selectivity, and the non-linearity of spatial summation will be determined and correlated with its location in the SC and correlated with locomotion. Aim 2 will apply the recording and data analysis tools developed in Aim 1 toward understanding the changes in circuitry in mutant mice that lack cortical inputs to the SC or lack On-Off direction selecive retinal ganglion cells (DS RGCs). This will allow us to determine the contribution of the cortex and DS RGCs toward the receptive field properties of SC neurons. Upon completion, a comprehensive classification of SC neurons, their topological organization, and their coding properties will be in hand. We will then take advantage of the ever-expanding availability of genetic tools (including optogenetics) that alter visual function, and mouse models of complex neurological disease that have altered activity patterns such as autism and schizophrenia. These same techniques will be useful to understand the circuitry of brain areas of animals with more complex visual systems and brain circuitry such as cat, ferret, and non-human primates.

The Superior Colliculus (SC) plays an essential role in processing auditory information to assess saliency and promote action; however, the underlying cell types and circuitry used to encode sound source locations remain largely unknown. Work done in primates and ferrets has shown that the receptive fields (RFs) of neurons in the deep SC (dSC) are organized in a 2-dimensional map of auditory space. This has recently been shown to also be true in the mouse, an organism that already has molecular and genetic tools available that will allow us to dissect circuitry to understand how this map forms. The overall objective of this application is to determine the functional properties of auditory neurons in the mouse SC, determine how these properties are encoded, and determine which brainstem and cortical inputs influence these properties. Our central hypothesis is that a combination of interaural level differences (ILD) and two sets of spectral cues are used to compute a 2-dimensional map of sound space; these are inherited from different brainstem regions and are modulated by the cortex. The goal of Specific Aim 1 is to test the hypothesis that the 2-dimensional map of sound space is encoded by the SC using a combination of ILDs and two sets of spectral cue patterns. To achieve this we will stimulate awake head-fixed mice, allowed to freely run on a treadmill, with spatially/temporally/spectrally restricted auditory stimuli, then simultaneously record SC neuronal response properties of thousands of auditory responsive neurons. Data analysis will determine the spatiotemporal and spectral/temporal receptive fields (RFs) of auditory neurons, their locations within the SC, the dependence of their RFs on ILDs and specific frequency combinations, and if these properties are modulated by locomotion. Experiments proposed in Specific Aim 2 will test the hypothesis that the SC computes sound location by combining inputs from different brainstem nuclei. We will record the response properties of the brachium of the inferior colliculus, the external nucleus of the IC, and the nucleus of the lateral lemniscus to auditory stimuli, and compare their RF properties to those in the SC. We will also use optogenetics to selectively excite or inhibit neurons that project from these areas to the SC in order to identify their specific contributions to the SC responses. In Specific Aim 3 we test the hypothesis that the direct projection from the auditory cortex to the SC is used to modulate the response properties of dSC neurons by measuring the response properties of auditory SC neurons both in mice that lack a cortico-collicular projection, and in those that have their auditory cortico-collicular projection silenced via optogenetics. The proposed research plan is significant because the results will establish the mouse SC as a model to study auditory spatial mapping and eventually auditory/visual spatial integration. Our findings will also lead to a better understanding of the neuronal circuitry used to compute auditory scenes in the awake behaving animal, and will shine light on neurodevelopmental disorders that have deficits in the auditory system.

The superior colliculus (SC) plays a critical role in integrating visual and auditory inputs to assess saliency and promote action. However, the underlying cell types and circuitry used to encode multimodal information and the mechanisms used during development to form the circuitry remain largely unknown. The recent explosion of new technology in mouse genetics allows neurons and circuits to be manipulated and specific genes to be removed, but surprisingly, the mouse has not yet been shown to be a model to study sensory integration. The overall objective of this proposal is to determine the functional properties of visual/auditory multisensory neurons in the mouse SC, to determine how these properties change in a mouse line genetically engineered to test hypotheses about how these properties develop. The central hypothesis to be tested is that visual and auditory information converge in the mouse SC to create multimodal neurons that form a multimodal map of space, and that map alignment forms using a visual map template-matching mechanism. The goal of Specific Aim 1 is to identify, and determine the response properties of, mouse SC visual/auditory multimodal neurons. To accomplish this, awake, head-fixed mice, allowed to freely run on a treadmill, will be stimulated with spatially/temporally/spectrally restricted visual and auditory stimuli while the SC neuronal response properties are being recorded using high-density silicon probes. The SC neural activity of ~170 neurons will be simultaneously recorded from in each mouse, using high-density silicon probes. Data analysis will determine the spatiotemporal receptive fields of the visual, auditory and visual/auditory multimodal neurons, their sensory integration properties, and the spatial/temporal/spectral components of the stimulus needed to elicit integration. Innovations include the use of virtual auditory space stimuli to present localized sound, and the recording and data analysis methods used. Experiments proposed in Specific Aim 2 will test the longstanding hypothesis that the alignment and integration of the visual and auditory inputs in the SC form using the visual map as a template. The approach will be to record and analyze the auditory and visual response properties as in Aim 1 but from transgenic mice engineered to have a duplicated visual map in the SC, and determine if the auditory map rearranges to align and integrate with the duplicated visual map. The proposed research is significant because it will provide the first comprehensive analysis of the receptive field properties of visual/auditory integrative neurons in the mouse SC, and will determine the general principles of how these properties develop. The results of this work can be exploited immediately and in the future, to determine the underlying circuitry used to integrate sensory information, the specific cell types involved, and how the state of the animal modulates these properties.