Yang Dan - US grants

Computational Analysis of Visual Processing in the Mammalian Brain

A major challenge in studying mammalian brain functions is to understand how information is encoded in the electrical activity of neurons. The current project aims to understand processing of visual information with a combination of experimental and computational approaches. Novel features of this project include the use of multielectrode recording of neural activity and natural scenes as visual inputs. Several computational analyses will be carried out concurrently with the physiological experiments to address information coding in the early visual pathway.
In the first part of the project, the investigators will use several decoding techniques to reconstruct natural scenes from recorded ensemble neural responses. These decoding studies will provide a critical test of various computational models of visual coding. Because of the relatively simple response properties of their visual neurons, the decoding study will first be carried out in the visual thalamus. To understand the mechanisms of visual coding, it is also crucial to understand what features of visual inputs are encoded at different stages of the visual pathway. Therefore in the second part of the project, the investigators will systematically characterize the features of visual inputs that are represented by the responses of primary visual cortical neurons. Such studies may reveal nonlinear coding properties of cortical neurons that have eluded earlier analyses. Taken together, the results from the proposed studies are likely to provide new insights into the general principles of sensory coding. Successful implementation of these research plans will also create an exciting environment for encouraging early involvement of undergraduate students into basic research, for in depth training of graduate students in neuroscience, and for preparing postdoctoral fellows for an independent research career.

A major challenge in studying sensory processing is to understand the meanings of the neural messages encrypted in the spiking activity of neurons. The ability to "read the neural code" is essential for our understanding of the neural mechanisms underlying brain functions. The proposed project aims to understand the neural code in the early visual pathway, i.e., the lateral geniculate nucleus (LGN) and the primary visual cortex. We will use a combination of experimental and computational approaches to address this problem from two directions. In the forward direction, we will use a computational method to characterize more systematically features of visual inputs that are encoded in neuronal responses. In the reverse direction, we will reconstruct visual inputs from the recorded neuronal activity. In part 1, we will use decoding techniques to reconstruct spatiotemporal natural scenes from ensemble responses in the LGN. In part 1, we will use decoding techniques to reconstruct spatiotemporal natural scenes from ensemble response in the LGN. The optimal linear decoding technique, which has been used successfully in a preliminary study, will be applied to further explore the functions of precisely correlated spiking in the LGN in visual coding. We will also explore the use of a gradient descent learning algorithm to train artificial neural networks to perform optimal input reconstruction. In part 2, we will use a covariance matrix analysis to systematically characterize the features of visual inputs that are represented by the responses of primary visual cortical neurons. Since most of the cortical neurons are complex cells with highly non-linear responses, a complete description of their coding properties is difficult to obtain with conventional methods. The covariance matrix analysis may help to reveal previously unknown coding properties and may provide new insights into the nature of the neural code in the primary visual cortex. Finally, in part 3, we will apply decoding techniques to reconstruct natural scenes using properties identified by both conventional techniques and by the covariance matrix analysis. This will provide a critical test of our computational model of cortical visual coding developed in part 2. The large amount of information on the anatomy and physiology of the early visual pathway makes it an idea model for computational analysis for neural coding. The results from the proposed studies are likely to provide new insights into the general principles of sensory coding and may advance our understanding of the neuronal mechanisms of higher brain functions under normal and pathological conditions.

DESCRIPTION (provided by applicant): Activity-dependent synaptic plasticity is believed to be the cellular basis for learning and memory. However, the relationship between the synaptic modification induced by simple patterns of activity in reduced preparations and functional plasticity in the intact brain remains an open question. Our goal is to bridge the gap in our understanding of activity-dependent neural plasticity between these two levels. In spike-timing-dependent synaptic plasticity (STDP), the direction and magnitude of synaptic modification depends on the relative timing of the pre- and postsynaptic spikes on the order of tens of milliseconds. Theoretical studies have highlighted this form of plasticity as a powerful learning rule that endows neural circuits with increased computational capacity. However, most of the previous studies on STDP used simple spike patterns to induce synaptic modification. These experiments may not provide sufficient information for understanding how STDP operates in the intact brain, where activity evoked by sensory stimuli exhibit complex spatiotemporal patterns. In the proposed project, we will carry out a series of experiments in cortical slices to systematically delineate the rules governing synaptic modifications induced by complex patterns of activity. The proposal is divided into two parts, addressing the temporal and the spatial properties of STDP. Regarding the temporal properties, we have developed a simple phenomenological model to predict the effects of complex spike trains in synaptic modification. In Part 1 of the proposed project, we will extend the previous study and will address several issues concerning the temporal interactions among multiple spikes in synaptic modification. Regarding the spatial properties, our preliminary studies suggest that synaptic modification may depend on dendritic location. Since synapses at different locations may serve different functions in information processing, the effect of dendritic location on synaptic modification may have important functional implications. In Part 2 of the proposed project we will investigate the effects of dendritic location on STDP. The temporal complexity of neuronal spiking and the spatial complexity of synaptic connections pose important challenges for understanding functional plasticity in the central nervous system. The proposed studies will provide important information for understanding how synaptic modifications are induced by natural stimuli and how they affect the functions of neuronal circuits in vivo.

DESCRIPTION (provided by applicant): Activity-dependent plasticity is essential for development and function of the nervous system. In the mammalian neocortex, sensory stimuli play crucial roles in shaping the circuitry and function, which may be largely mediated by activity-dependent synaptic modification. Although at each level - synaptic, circuitry, and functional - cortical plasticity has been studied extensively, the causal relationship between activity-induced modifications at different levels remains to be firmly established. Our goal is to bridge the understanding of cortical plasticity at these levels. In recent studies, we have demonstrated that asynchronous visual stimuli (1-2 min) can induce shifts in adult cortical receptive fields (RFs) and in human spatial perception in a manner consistent with spike-timing-dependent plasticity (STDP), a powerful synaptic learning rule widely observed among excitatory synapses in the brain. In the proposed study we will further examine the functional modifications of adult visual cortex mediated by STDP, using a combination of electrophysiological and psychophysical experiments and computational modeling. There are three specific aims, examining the mechanism and functional significance of the cortical plasticity. In Aim 1, we will test whether RF and perceptual modifications can be induced by brief periods (seconds) of visual conditioning, a possibility suggested by recent studies in cortical slices. Such rapid plasticity may operate more frequently under natural conditions, and this experiment will provide a basis for our subsequent studies of cortical modifications induced by natural stimuli. In Aim 2, we will investigate the neuronal circuitry underlying the functional modification. We will first measure the interocular transfer of the effect to determine whether it is cortical in origin. We will then examine systematically the dependence of the cortical modification on conditioning parameters (orientation, luminance polarity, timing, and location of conditioning stimuli) and on other properties of the recorded neuron (simple/complex classification, laminar location, binocularity, and direction selectivity) to further determine the underlying circuitry. In Aim 3, we will directly assess the functional relevance of the plasticity by measuring cortical modification induced by motion stimuli and by natural scenes containing motion signals. Together, these studies are likely to provide significant new insights into the functional implications of activity-dependent synaptic plasticity at the levels of cortical circuitry, RF properties, and visual perception.

Proposal No: 0749056
PI: Yang Dan

Award Abstract:

This award supports the preparation and sharing of computational neuroscience data as part of an exploratory activity aimed at catalyzing rapid and innovative advances in computational neuroscience and related fields. Investigators Yang Dan, Tim Blanche, Jack Gallant, and Frederic Theunissen will make several data sets available, each exploring different aspects of sensory coding: (1) cortical slice data acquired in order to examine the effects of complex spike trains in the induction of long-term synaptic modification; (2) recordings of primary visual cortical neurons made during stimulation with complex stimuli, white noise, and natural images; (3) recordings from visual area V4 during stimulation with parametrically varying bars, rings and gratings; (4) recordings from visual areas V1, V2, and V4 during stimulation with a rapid dynamic sequence of gratings; (5) recordings of neurons at three levels of the avian auditory system during stimulation with complex synthetic and natural sounds; and (6) large-scale neuronal recordings from primary visual cortex made with multi-site electrode arrays that allow simultaneous recording from more than a hundred single units at once. It is anticipated that these data will be useful for the study of spatial and temporal neural coding, nonlinear receptive field properties, learning rules, hierarchical processing strategies, and other aspects of the analysis of complex sensory information.

This project will start to integrate what we know about the challenging task of recognizing objects from visual inputs, by drawing on the highest-performing artificial neural network systems, new models of deep belief learning from cognitive science, and new experiments on the visual cortex.

The most transformative aspect of this work is that it will aim at decisive experiments which challenge traditional assumptions about purely local feedback in the learning system as such, assumptions which are prevalent in today,s mathematical models of learning in neural circuits. Many engineers and more classical systems neuroscientists believe that these assumptions are obviously false, but a decisive set of experiments would be crucial in encouraging new types of computational models of the brain, including models which fit with what actually works in image processing in technology. On the other hand, if the experiments begin to show how such learning capabilities are actually possible within the traditional paradigm, that would be equally transformative. Brain-like capabilities in image processing are an additional goal of this work; image processing is a large and growing part of the nation's cyberinfrastructure.

DESCRIPTION (provided by applicant): The mammalian neocortex mediates a variety of cognitive functions, and understanding the circuit basis for cortical processing is a central goal in neuroscience. In the primary visual cortex (V1), the receptive field properties of individual neurons have been characterized extensively, but the underlying neuronal circuitry remains unclear. The goal of the proposed research is to dissect the excitatory and inhibitory synaptic inputs underlying the spatiotemporal receptive fields of V1 neurons. Experimentally, we will make intracellular (patch clamp) recordings in anesthetized rats and cats to measure the synaptic inputs to cortical neurons. Computationally, we will apply both linear and nonlinear analyses to determine the receptive field properties of each input. There are four specific aims. Aim 1 is to characterize the synaptic mechanisms underlying the basic receptive field and direction selectivity of simple cells. To test the anti-phase inhibition model, we will determine the extent to which the excitatory and inhibitory receptive fields are matched to each other with opposite ON/OFF polarities. The contributions of several proposed mechanisms to simple cell direction selectivity will be assessed. Aim 2 is to characterize the synaptic circuitry underlying the receptive field subunits of complex cells. The predictions of different circuit models for complex cell RFs will be tested. We will also determine the contributions of several proposed mechanisms to complex cell direction selectivity. In Aims 3 and 4 we will address two of the more advanced receptive field properties involved in processing motion stimuli. Our recent study using extracellular recordings indicated a spatial asymmetry in direction-selective inputs to V1 neurons, which gives rise to two novel RF properties that could account for two visual illusions. In Aim 3, we will examine the spatial distributions of direction-selective excitatory and inhibitory inputs in order to test and constrain the asymmetric circuit model. In Aim 4, we will measure the effects of motion adaptation on the strength and direction selectivity of excitatory and inhibitory synaptic input to determine how these inputs contribute to adaptation-induced change in V1 direction selectivity. To further test the asymmetric circuit model, we will also measure the effect of motion adaptation on the spatial distributions of the excitatory and inhibitory inputs. Such comprehensive characterization of the synaptic mechanisms underlying cortical receptive field properties is not only crucial for understanding V1 functions, but also likely to shed light on the general principles of cortical computation. PUBLIC HEALTH RELEVANCE Balance between excitation and inhibition in the cortex is critical for normal brain functions, and disruption of the balance causes a variety of mental disorders. The proposed research aims to understand the relationship between excitatory and inhibitory inputs, and how they shape the functional properties of visual cortical neurons. Such knowledge will be crucial for the development of treatment for not only deficiencies in visual functions, but also other neurological disorders such as epilepsy.

DESCRIPTION (provided by applicant): The mammalian neocortex is instrumental in a variety of cognitive functions, and understanding cortical circuitry is essential for prevention and treatment of mental disorders. A prominent feature of the cortex is that neurons with similar functional properties are organized in columns. However, how the columnar structure arises during development remains poorly understood. Our recent study in mouse visual cortex has shown that sister neurons originating from a common progenitor cell during early brain development exhibit strong similarity in orientation tuning, arguably the most important functional property in the primary visual cortex. This finding demonstrated, for the first time, a direct correspondence between the ontogenetic and functional columns, and it opened new avenues for studying cortical circuitry. The proposed study aims to address several basic questions concerning the developmental origin of cortical microcircuits, using a combination of techniques including two-photon imaging, electrophysiological recording, neuronal labeling, and molecular perturbation with recombinant viral vectors. By testing whether and how the functional similarity between sister neurons depends on their lineage and physical distances, we will define the basic processing unit of the cortex (Aim 1). By manipulating the excitability and NMDA receptor expression in selected neurons and visual experience of the animal, we will elucidate the interactions between developmental lineage, electrical activity, and visual experience in shaping the functional properties of cortical neurons (Aims 2 and 3). Finally, using high-speed circuit mapping with glutamate uncaging, we will detect common inputs to clonally related sister neurons in order to understand the synaptic basis for their functional similarity. These experiments represent a major step to bridge developmental and systems neuroscience, and they are likely to provide unprecedented insights into the development and organizational principle of cortical microcircuits.

SUMMARY Animal behaviors are orchestrated by the sophisticated nervous system, which is dynamically regulated by neuromodulators including lipids and neuropeptides. Endocannabinoids (eCBs) are neurolipids exist broadly in the brain and regulate learning and memory, addiction, pain sensation, and food intake. Among neuropeptides, cholecystokinin (CCK) is involved in nutrient sensing, food intake, and sleep regulation, and oxytocin (OXT) and vasopressin (AVP) play important roles in various aspects of social behaviors. However, how and when lipid and neuropeptide transmission occur in the brain are largely unclear. Existing methods (e.g. microdialysis) that measures brain chemical content suffer from low temporal and spatial resolution. Additionally, since neurolipid and neuropeptide releases often require repetitive neuronal firing and can occur at both axonal and dendritic sites, activity of the neuromodulator- releasing neurons cannot reliably predict where and when neurolipids and neuropeptides are released. Here we propose to develop a set of new tools for long-term monitoring of neurolipids and neuropeptides. Our strategy taps into their natural receptors, human G protein-coupled receptors (GPCRs), which are coupled to GFP. In the presence of neurolipids or neuropeptides, these GPCR Activation-Based (GRAB) sensors transform ligand binding-induced conformational changes into rapid fluorescent signals. We aim to develop and optimize neurolipid and neuropeptide GRAB sensors with >500% fluorescence change (dF/F) and 10- nanomolar affinity in vitro and validate these novel tools in brain slices ex vivo and mouse behavioral paradigms in vivo. In Aim 1, we will develop GRAB sensors for endocannabinoids, CCK, vasopressin, and OXT by systematically varying key sites involved in ligand binding, conformational change, etc. In Aim 2, we will validate the performance of these sensors in brain slice following long-term expression using viral tools. In Aim 3, we will use three different imaging methods (fiber photometry, epifluorescence and 2-photon imaging coupled with GRIN lens) in different behavioral paradigms to test in vivo performance of the novel GRAB sensors in mice. Feedback from experiments in Aims 2-3 will guide iterative optimization in Aim 1. Successful completion of our proposal will yield a suite of powerful tools and technical approaches, which will greatly facilitate studies of neurolipids and neuropeptides under both physiological and pathological conditions, helping reveal disease mechanisms, providing therapeutic guidance, and eventually benefiting human health.