Thomas A. Cleland, Ph.D. - US grants

[unreadable] DESCRIPTION (provided by applicant): It is crucial to understand the mechanisms by which postsensory neural circuits process and transform incoming sensory information in order to understand the composition of secondary neural signals, such as the responses of mitral/tufted cells to odor stimuli. The dendodendritic circuitry of the olfactory bulb glomeruli has been well studied; abundant cellular, slice, in vivo, and behavioral data have provided information about the cellular biophysics, adaptive properties, synaptic pharmacologies, and centrifugal modulation of this well-defined circuit. Computational modeling of single olfactory glomeruli at the biophysical level has the potential to integrate these diverse data and offer testable hypotheses regarding the computational capacity of olfactory glomeruli and their putative functional contribution to the construction of second-order odor representations in the mitral cell ensemble.

In sensory systems, much of the task of extracting useful information about stimuli of interest falls to the neural circuitry downstream from the primary receptor complement. It is critical to understand the physiological mechanisms by which these postsensory neural circuits receive and process incoming sensory information in order to properly interpret secondary neural signals, such as the responses of mitral cells within the olfactory bulb to odor stimuli. As a well-described and delimited neural network close to the sensory periphery, the olfactory bulb is a clear candidate structure within which to develop an integrated understanding of neural coding and information processing across levels of analysis, from intracellular cascades and membrane properties through their behavioral consequences. Such a vertically integrated understanding of neural processing mechanics will be invaluable for the development of clinical psychiatric applications in which gene therapeutic or pharmacological agents must be specifically delivered to appropriate effector sites in order to effect the intended,systemic or behavioral changes. Physiologically constrained computational modeling of olfactory bulb neural circuitry, the subject of this proposal, is an important tool for achieving an integrated understanding of systemic function. Briefly, well-designed and constrained models enable exploration of the capabilities of multivariate systems that are too complex to be immediately intuitive. The specific aims of this proposal concern a set of models of the olfactory bulb glomerular layer based on the "non-topographical contrast enhancement" principle. The first aim is to implement this model mechanism in a full-scale network simulation (incorporating up to 2000 glomeruli) and challenge its predictions with glomerular imaging data, while the second is to develop a series of cellular compartmental models of glomerular layer olfactory bulb neurons (mitral, external tufted, and periglomerular cells) that can accurately reflect their intrinsic dynamical and pharmacological properties. The long-term goal of this project is to merge these two threads into a unified network model of the olfactory bulb containing sufficient cellular detail to enable the quantitative integration of existing data from different levels of analysis: for example, to interpret the effects of bulbar neuromodulators on membrane properties, field phenomena, and behavior. Quantitative, integrative results such as these are unlikely to be achieved without the use of computational modeling.

DESCRIPTION (provided by applicant): This proposal presents a multidisciplinary approach to understanding a fundamental property of sensory processing: the modulation of sensory perception by acetylcholine. The proposal addresses this question in the olfactory bulb of rats, using a combination of computational modeling, brain slice physiology, in vivo electrophysiology and behavioral pharmacology. The mammalian olfactory system offers several unique advantages for such studies. First, the olfactory system is anatomically relatively simple and this simplicity has resulted in an extensive knowledge of its synaptic connections and physiology. Second, because olfactory bulb principal neurons are a single synapse removed from sensory neurons, a relatively clear relationship between neural representations in the bulb and the perceptual properties of olfactory signals has been established. Third, this knowledge in turn has enabled the development of computational models crucial for the interpretation and integration of experimental data. A collaborative effort between three labs will investigate the interplay between muscarinic and nicotinic receptor activation and the functional consequences of their coordinated activation. Specifically, we will (1) determine the effects of muscarinic and nicotinic receptor activation on olfactory bulb neural circuits in a detailed biophysical model, (2) test the cellular and synaptic effects of such modulation using brain slice electrophysiology, (3) use in vivo electrophysiology to determine the effect of cholinergic modulation on odor responses, and (4) test the functional predictions arising from these experiments in behavioral pharmacology experiments. Taken together, the proposed simulations and experiments will elucidate how nicotinic and muscarinic receptors interact within a coordinated neural circuit to improve contrast and signal-to-noise properties in early olfactory processing.

DESCRIPTION (provided by applicant): Reciprocal interactions between mitral and granule cells in the olfactory bulb external plexiform layer (EPL) modify the timing of mitral cell action potentials and thereby influence the information that the olfactory bulb exports to its multiple targets. Theoretical and experimental results produced by each of the two collaborating PIs of this proposal have shown that this powerful and plastic EPL network is not responsible for the simple contrast-enhancement function often attributed to it (instead, this largely occurs in the glomerular layer). Rather, EPL interactions perform additional, subsequent operations on odor representations that mediate changes in odor perception and representation based in part on an individual animal's history of odor learning. Specifically, we here propose that computations in the EPL serve to render mitral cell output patterns selective for certain higher-order features of odors in the same sense that neurons in primary visual cortex are selective for higher-order visual features such as edge and orientation - that is, they reflect the co-activation of certain spatiotemporal combinations of receptors that together are characteristic of a meaningful odor. We here propose to develop detailed theoretical models of this hypothesis and its implications and to test its main critical predictions experimentally. The model in its present form predicts (1) that whereas granule cells can be excited by mitral cell lateral dendrites irrespective of their physical proximity, spike timing in mitral cells is affected only by inhibition from physically neighboring granule cells, and (2) that granule cells require the simultaneous activation of specific sets of afferent (glomerular) inputs in order to evoke a whole-cell regenerative response and thereby evoke lateral inhibition. This architecture has substantial implications for the processing of odor representations that we will develop in Aim 1. To test this model, we will determine whether granule cell effects on mitral cell activity depend on physical proximity using spatially selective optogenetic activation and silencing of granule cells (Aim 2), measure the form and specificity of the afferent activity patterns required to evoke spikes in granule cells using optical stimulation of olfactory bulb glomeruli (Aim 3), and test the model's assumptions regarding the structure of olfactory bulb plasticity by measuring the perceptual effects of competing odor representations (Aim 4). The intellectual merit of this application derives from its use of state-of-the-art computational modeling to structure the proposed experiments and interpret their results, along with the use of newly-developed experimental techniques to address the longstanding questions about EPL function and processing that the theory described herein has framed and rendered testable. The collaboration between PIs Cleland and Schaefer is essential to the success of this proposal, as PI Schaefer's experimental techniques are uniquely able to test the prediction of PI Cleland's theoretical models. The efficiency of this collaboration is enhanced by the cross-competence of the PIs: PI Schaefer is competent in both computational and behavioral approaches and utilizes both in his research, whereas PI Cleland is competent in electrophysiological approaches and utilizes them in his research. This collaboration will benefit students and postdocs at both institutions by integrating them into a genuinely interdisciplinary framework encompassing both experimental and computational approaches, and facilitating their cross-training by enabling travel between labs. Consequently, the broader impacts of this proposal include the cross-training of students from diverse backgrounds in coordinated theoretical and experimental techniques as well as exposing them to both American and German laboratories. Both PIs have a strong history of training undergraduates and women in areas in which women remain underrepresented. This proposal also provides for the substantial participation of undergraduate researchers.

DESCRIPTION (provided by applicant): The emergence of coherent local field potentials (LFPs) in the beta (15-30 Hz) and gamma (35-100 Hz) frequency bands has been associated with attention, sensorimotor integration, and other active information processing states within and among brain regions. Beta/gamma coherence is broadly associated with action potential synchronization, which in turn has been hypothesized to define and delimit neural assemblies, and further to enable multiple assemblies of neurons within a population to synchronize within each assembly (but not among different assemblies) so that these multiple assemblies can compete to determine the systems output. The olfactory system has a strong and complex complement of LFP oscillations. Several different frequency bands are routinely observed, and are associated with particular behavioral tasks or states, such as acute sensory activity, resting alertness, and respiratory phases. Some oscillations coexist in the same structure; others appear to give way to one another. Some are local; others mediate interareal coupling either via LFP coherence or via subtler spike-field coherence in which periodic activity in the OB shapes the timing of action potentials in a limited assembly of neurons in a follower structure. Moreover, the olfactory system juxtaposes bottom-up network dynamics resulting from afferent stimulation with top-down dynamics arising from behavioral state factors, both affecting sensory information exported from olfactory bulb. Overall, the olfactory system is a particularly rich fied in which to study the biophysics and ethological utility of these neuronal dynamical systems in concert and within experimentally accessible tissues. This project will establish a robust, mechanistic, biophysically-based model of oscillations and synchronization in the mammalian olfactory system. The PIs will combine multichannel unit and LFP recordings from awake/behaving rats and from acute slices of the mouse OB using planar multielectrode array, and use the results to shape the expansion of an existing, biophysically detailed model of the early olfactory system. They will determine the extent to which the OB forms competing assemblies of gamma-coupled neurons and study beta oscillations as an interregional coupling mechanism with piriform cortex (PC) that supersedes these local gamma-coupled assemblies. Additional, less well-established OB interactions with olfactory tubercle and orbitofrontal cortex during odor sampling and response decisions also will be studied. Integrating these datasets into a common Hodgkin/Huxley-based network model will explicate the construction and utility of these systemwide dynamics based on their underlying cellular and network mechanisms. The proposed work takes a fairly well-characterized network and, via computational modeling, combines studies across different levels of analysis to build a mechanistic model of a complex dynamical system. The results will enable a deeper understanding of the dynamical flexibility of cortical circuits at many levels of analysis. Behavior provides tight control over oscillatory staes and cognitive processes associated with them, enabling explication of intact functional circuits. Slice electrophysiology and computational modeling will provide greater detail on the mechanistic, synaptic, neuromodulatory, and dynamical principles involved in generating and switching among these multiple states. The collaboration will benefit students at both institutions by integrating them into an interdisciplinary framework encompassing computational and experimental approaches, exposing students from diverse backgrounds to new research techniques and interdisciplinary and computational approaches to neuroscience. Close collaboration of the two investigators will transfer knowledge and methods across laboratories. Both laboratories actively train undergraduates in research and include them on many publications, and both laboratories actively recruit and train female and minority scientists in STEM fields. PI Kay has initiated participation in Project Exploration, an outreach program that provides access to science and scientists to underrepresented minority children and girls, and PI Cleland participates in the Leadership Alliance and Cornell Biology Scholars program for underrepresented minority and first-in-family students, and has published papers with undergraduate coauthors enrolled in these programs.

? DESCRIPTION (provided by applicant): In the mammalian brain, early sensory areas are organized as stereotyped maps of stimulus qualities. The anatomical features of these sensory maps underlie the neuronal computations and information processing that are essential to generate appropriate perceptual experiences and behaviors. In the mammalian olfactory bulb (OB), the most striking anatomical feature of this sensory map is the convergent axonal input from primary olfactory sensory neurons (OSNs); specifically, each insular glomerular structure in the OB receives axons exclusively from OSNs expressing the same odorant receptor. Moreover, OB projection neurons, the mitral/tufted cells, also project their primary dendrites exclusively into a single glomerulus. However, it is not well understood how this strict pattern of convergence and segregation contributes materially to the processing of olfactory information and odor-driven behaviors. In this project, we establish a multi-disciplinary approach to determine the contribution of this strict anatomical mapping to odor representation and perception. Specifically, we will genetically manipulate the projection patterns of OSN populations to perturb the anatomical map in the olfactory bulb, and perform a battery of automated behavioral assays probing odor discrimination and recognition. We then will integrate mathematical models based on OB circuitry with electrophysiological and optical imaging studies to elucidate the basis for these response differences in wildtype and mutant mice. By combining genetics, behavior, electrophysiological, and imaging approaches, we will determine the impact of genetically altered glomerular maps on odor perception under both naïve and learned conditions. RELEVANCE: Precise neuronal connectivity in the nervous system is essential for its proper function. Cognitive deficits in many psychiatric disorders arise from disruptions in neural connectivity, including disruptions within sensory circuits. The proposed study investigates the role of a specialized, anatomical neuronal connectivity pattern in regulating sensory information processing and perceptual experience. Results from this study will reveal how brain function can deteriorate because of disruptions in the normal connection patterns among neurons.

For some tasks, modern computers vastly outperform the human brain – for example, large-scale numerical calculations, or the precisely accurate recall of organized information. But for other important tasks, the brains of humans and other animals are far superior to any computing system that has been built, both in terms of what they can do and in terms of the startlingly low energy required. For example, people can recognize familiar individuals at a distance, not only by their facial features, but by gait or other subtleties of movement that they might not even be able to articulate. People and other animals rapidly acquire information from their environments, but also are able to intelligently apply that information under novel, unforeseen circumstances. The study of brain-inspired computing is devoted to learning fundamental new ways to think about how computers work with information, so that they can perform better on such weakly-defined, open-world problems. In parallel, advances in physics have produced new optical materials and methods that can perform computations very rapidly and with extraordinarily low energy costs – if the problems of interest can be structured in ways compatible with these brain-inspired computing techniques. This project seeks to develop a brain-inspired computing network that learns rapidly and solves a set of real-world identification tasks, and to deploy this network onto portable devices as well as custom testing platforms built with these advanced physical substrates. A key goal is to show how these brain-inspired computing methods can achieve superior performance on open-world problems, most radically so when deployed on next-generation optical computer platforms.

In contrast to contemporary deep networks, the brain-inspired networks described in this proposal are based on heterogeneous elements and feedback-mediated dynamical systems, and operate based on fully localized computations that obviate the need for shared memory resources. Consequently, they learn rapidly, and when deployed on neuromorphic platforms such as Intel Loihi they exhibit increased speed and tremendously reduced energy budgets. Importantly, state of the art photonic computing substrates are directly compatible with neuromorphic computational architectures, suggesting that they will be compelling platforms for these decentralized, brain-inspired computing algorithms. The intellectual merit of this project is to develop, deploy, and benchmark an established set of decentralized, brain-inspired algorithms designed for successful sensory identification under unpredictable, open-world conditions on a range of platforms, including leading-edge photonic computational substrates. Specifically, the algorithms will be extended to incorporate higher-order brain-inspired circuit properties, deployed onto portable device platforms for use in the field, and also deployed and tested on photonic substrates to demonstrate the transformational potential of these computational platforms. Broader impacts include a continuing commitment by both PIs to supervising undergraduate research experiences for students from groups underrepresented in STEM on projects directly connected with the research proposed here, as well as the potential for development of a new generation of smart devices using neuromorphic methods. PI Cleland also intends to incorporate the concepts discussed in this application into a unit of his advanced undergraduate Neural Representations seminar course.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Project Summary Sensory signals encountered under different circumstances may have quite different implications. In the early olfactory system, preliminary evidence suggests that this (non-olfactory) contextual information is integrated into odor representations at a very early stage, potentially even the main olfactory bulb. Recent evidence indicates that the anterior olfactory nucleus (AON), a structure directly adjoining the olfactory bulb, serves to integrate afferent odor information with contextual information from the ventral hippocampus (vHC) and is necessary to solve contextually-dependent olfactory decision-making tasks. The vHC is known to relay task-relevant spatial contextual information to other brain systems. We here hypothesize that direct projections from the vHC to the AON play a dominant role in the integration of contextual and olfactory information, and that the AON embeds this multisensory contextual information into early-stage odor representations. Our preliminary data show that rodents can learn to respond differently to odors based on the spatial context in which they are encountered, and that the expression of such a rule depends on both AON and vHC, whereas a similar but odor-independent task requires vHC but not AON. We propose a multipronged approach to understanding the integration of spatial context into olfactory representations, engaging electrophysiological ensemble recordings and interareal coherence measurements in awake, behaving rodents, the optogenetic manipulation of vHC and AON circuit activities, and a double-labeling strategy for the within-subjects comparison of immediate-early gene (Fos) responses across two experimental conditions separated in time.