Sensory information is represented in cortex by cascades of excitation, the patterns of which are constrained and biased by anatomical connections between neurons. Additionally, in the living animal, functional connectivity is dynamically adjusted by internally generated background activity, which varies by arousal state and behavioral context. Therefore, to understand how excitation propagates through the cortex, it is necessary to characterize the laminar flow of signal propagation as well as spontaneous network activity, which will constrain that propagation. This thesis characterizes the nature and mechanisms of awake cortical network dynamics, as well as the sources of sensory inputs in different cortical layers of the rat somatosensory system. Mammalian brains generate internal activity independent of environmental stimuli. Internally generated states may bring about distinct cortical processing modes. To investigate how brain state impacts cortical circuitry, we recorded intracellularly from the same neurons, under anesthesia and subsequent wakefulness, in the rat barrel cortex. In every cell examined throughout layers 2-6, wakefulness produced a temporal pattern of synaptic inputs differing markedly from those under anesthesia. Recurring periods of synaptic quiescence, prominent under anesthesia, were abolished by wakefulness, which produced instead a persistently depolarized state. This switch in dynamics was unaffected by elimination of afferent synaptic input from thalamus, suggesting that arousal alters cortical dynamics by neuromodulators acting directly on cortex. Indeed, blockade of noradrenergic, but not cholinergic, pathways induced synaptic quiescence during wakefulness. This thesis shows that subcortical inputs from the locus coeruleus-noradrenergic system can switch local recurrent networks into different regimes via direct neuromodulation. Having characterized the nature of wakeful dynamics, I next sought to characterize how sensory information propagates through the cortex. The thalamocortical projection to layer 4 (L4) of primary sensory cortex is thought to be the main route by which information from sensory organs reaches the neocortex. Sensory information is believed to then propagate through the cortical column along the L4→L2/3→L5/6 pathway. This thesis shows that sensory-evoked responses of L5/6 neurons derive from direct thalamocortical synapses, rather than the intracortical pathway. A substantial proportion of L5/6 neurons exhibit sensory-evoked postsynaptic potentials and spikes with the same latencies as L4. Paired in vivo recordings from L5/6 neurons and thalamic neurons revealed significant convergence of direct thalamocortical synapses onto diverse types of infragranular neurons. Pharmacological inactivation of L4 had no effect on sensory-evoked synaptic input to L5/6 neurons, and responsive L5/6 neurons continued to discharge spikes. In contrast, inactivation of thalamus suppressed sensory-evoked responses. This thesis shows that L4 is not an obligatory distribution hub for cortical activity, contrary to long-standing belief, and that thalamus activates two separate, independent "strata" of cortex in parallel.

Alexander disease (AxD) is a rare and fatal neurological disorder caused by mutations in the gene that encodes glial fibrillary acidic protein (GFAP), an intermediate filament protein found in astrocytes in the central nervous system. The clinical presentations of AxD are diverse, ranging from onset in infancy to onset in early adulthood, and include seizures, psychomotor retardation, ataxia, and a variety of neurological signs related to abnormal brain stem function. The defining neuropathological hallmark is the presence of cytoplasmic, proteinaceous inclusions called Rosenthal fibers in astrocytes. Although GFAP expression is astrocytic, AxD patients also show de/dysmyelination and variable amounts of neuronal loss, most severely in infantile-onset patients. Astrocytes undergo severe morphological changes, beyond that of typical reactive astrocytes, and develop several forms of cell stress. However, how stressed astrocytes cause the loss of myelin in this disease is unknown. In this work I have conducted a largely immunohistological investigation of AxD patient tissue, model mice, and primary astrocytes cultured from the AxD model mice, focusing on factors that might provide insight into the pathological manifestations of AxD and paying particular attention to those factors which might contribute to de/dysmyelination. To gain insight on the morphological transformation of astrocytes in AxD, I analyzed GFAP in the hippocampus of the most severely affected AxD mouse. Astrocytes in these mice lose their star-like shape, and become hypertrophic and often multinucleated. They accumulate large amounts of GFAP. Subsequent study of primary cultured astrocytes from AxD mice revealed that these cells have perinuclear inclusions of GFAP surrounded by displaced microtubules and displaced Golgi. I next investigated another mechanism of stress that may affect astrocyte function in AxD. Work in our lab and others' has demonstrated proteasomal inhibition in AxD astrocytes. Because the unfolded protein response in the endoplasmic reticulum (ER) can be enacted by proteasomal inhibition, I examined the immunohistochemical expression of two proteins commonly increased under conditions of ER stress. We found BIP/Grp78, an ER chaperone, increased in AxD patient astrocytes and model mice. Additionally, the CCAAT enhancer binding protein homologous protein (CHOP) was expressed by a small subset of astrocytes in the AxD mouse hippocampus, unveiling ER stress as a potential contributory factor in AxD pathology. Work in other labs has found iron in astrocytes in AxD model mice. To further elucidate mechanisms of cellular stress in AxD, I conducted an immunohistochemical analysis of iron and several regulatory proteins in AxD patients and found, by enhanced Perls' staining, Fe3+ in Rosenthal fibers and iron and ferritin accumulated in astrocytes. This finding is in marked contrast to what one sees in the normal CNS, with little staining of astrocytes, and easily detectable staining of oligodendrocytes. Finally, I examined the localization of the cell surface glycoprotein CD44, along with several related proteins, including its ligand hyaluronan. I found CD44 protein expression greatly increased in the white matter, cortex and hippocampus of AxD patients and in the hippocampus of AxD mice. Additionally, through use of a biotinylated hyaluronan binding protein, I found abnormally high levels of hyaluronan in the hippocampus of AxD mice in the same areas where increases in CD44 were found. Work elsewhere has found CD44 and hyaluronan in other disorders that affect myelination, and experiments have revealed an inhibitory effect of hyaluronan on oligodendrocyte development and myelination. The studies in this thesis contribute novel stressors to the list of those that impact astrocytes in AxD and, in particular, suggest the accumulation of iron in astrocytes as potentially important to the pathological manifestations of AxD. Additionally, my research has revealed dramatic increases in the expression of CD44 in AxD astrocytes which, in conjunction with widespread increases in hyaluronan, may be critical to understanding the mechanisms underlying the de/dysmyelination that occur in this disease.

Caenorhabditis elegans is a promising organism for trying to understand how nervous systems generate real-time behavior. Its low neuron count suggests that we may be able to observe all of the constituents of the computation of sophisticated sensorimotor behavior. However, its appropriateness as a system for quantitative dynamical study has yet to be established. We show that C. elegans chemosensory neurons can operate in a highly deterministic and low-noise mode, and they act as reliable linear filters of their input. We then use dynamical systems analysis in combination with classical genetic perturbation to uncover cellular and circuit mechanisms of temporal processing. This work should firmly establish C. elegans as a viable platform for applying quantitative dynamical systems methods to understanding how a nervous system processes sensory information, integrates it with an evolving internal state, and produces goal-directed, coordinated behavior.

Goal-directedness is a fundamental property of all living things, but it is perhaps most easily identified in the movement patterns of animals. Ethologists have divided the basic forms of animal behavior into three categories: reproductive, defensive, and ingestive, all of which depend on the complex orchestration of motor control. In this dissertation, we use the framework of optimal control theory to model goal-directed behavior and repurpose it in new ways. We demonstrate a method for creating a hierarchical control network in which higher levels of the control hierarchy deal with tasks of increased abstractness. In a two-level system, the lower-level deals with short time-scale, low-dimensional motor control, and the higher-level is charged with longer time-scale, higher-dimensional planning. Central to our approach to joining the levels is the construction of a forward model of the behavior of the lower-level by the higher-level. Thus, we extend ideas of optimal control theory from controlling a "plant" to controlling a controller. We apply our method to the example problem of guiding a semi-truck in reverse around a field of obstacles. The lower-level controller drives the truck, and the higher-level detects obstacles and plans routes around them. In other work, we consider whether it is possible for a neural system that obeys certain biological constraints to solve optimal control problems. We exhibit a simple method to train a different kind of internal model, a neural network model of the Jacobian of the plant, and we integrate the internal model in a forward-in-time computation that produces an optimal feedback controller. We apply our method to two well-known model problems in optimal control, the torque-limited pendulum and cart-pole swing-up problems.

Recent advances in the field of axon guidance have revealed complex transcription factor codes that regulate neuronal subtype identity and their corresponding axon projections. Retinal axon divergence at the optic chiasm midline is key to the establishment of binocular vision in higher vertebrates. In the visual system of binocular animals, the ipsilaterally and contralaterally projecting retinal ganglion cells are distinguished by the laterality of their axonal projections. Specific axon guidance receptors and their ligands are expressed in retinal ganglion cells (RGCs) and at the chiasm, tightly regulating the development of the ipsilateral (uncrossed) and contralateral (crossed) retinal projections. Though many factors are known, their dysfunction leads to only partial misrouting of RGC axons. Moreover, the complex transcription factor codes that regulate RGC subtype identity are only beginning to be uncovered. Numerous gaps remain in our understanding of how these guidance molecules are transcriptionally regulated and how they are induced by the patterning genes that set up the different domains in which these RGC subtypes reside. An even more elusive question within the field is how the ipsilateral and contralateral RGC subpopulations acquire their different cell fates. In this thesis, I present my work on dissecting out the molecular signatures of the ipsilateral and contralateral RGC populations during embryonic development through gene profiling followed by the functional characterization of one candidate from this screen. In Chapter 2, I developed a cell purification method based on retrograde labeling of these two cell populations from their divergent axonal projections followed by cell sorting. This method can be used in studies requiring purified populations of embryonic RGCs. In Chapter 3, I conducted a microarray screen of purified ipsilateral and contralateral RGCs using the above method. Through subsequent validation of the in vivo expression patterns of select candidates, I identified a number of genes that are differentially expressed in ipsilateral and contralateral RGCs. Subsequent functional characterization of these genes has the potential to uncover novel mechanisms for regulating axon guidance, cell differentiation, fate specification, and other regulatory pathways in ipsilateral and contralateral RGC development and function. The results of this screen also revealed that ipsilateral and contralateral RGC may have distinct developmental origins and utilize different strategies for differentiation. In Chapter 4, I demonstrate a novel role for cyclin D2, one of the above candidates, in the production of ipsilateral RGCs. The G1-active cyclin D2 is highly expressed in the ventral peripheral retina preceding and coincident with the developmental window of ipsilateral RGC genesis. I further found that ipsilateral RGC production is disrupted in the cyclin D2 null mouse. The expression of cyclin D2 in a distinct proliferative zone that has evolutionary significance in ipsilateral RGC production and its subtype-specific requirement during retinal development suggest that cyclin D2 may mark a distinct progenitor pool for ipsilateral RGCs. Thus, these studies offer an important advance in our understanding of neuronal subtype diversification within the retina.

Retinitis pigmentosa (RP) is a photoreceptor neurodegenerative disease. Patients with RP present with the loss of their peripheral visual field, and the disease will progress until there is a full loss of vision. Approximately 36,000 cases of simplex and familial RP worldwide are caused by a mutation in the rod-specific cyclic guanosine monophosphate phosphodiesterase (PDE6) complex. However, despite the need for treatment, mouse models with mutations in the alpha subunit of PDE6 have not been characterized beyond 1 month of age or used to test the pre-clinical efficacy of potential therapies for human patients with RP caused by mutations in PDE6A. We first proposed to establish the temporal progression of retinal degeneration in a mouse model with a mutation in the alpha subunit of PDE6: the Pde6anmf363 mouse. Next, we developed a surgical technique to enable us to deliver therapeutic treatments into the mouse retina. We then hypothesized that increasing PDE6a levels in the Pde6anmf363 mouse model, using an AAV2/8 gene therapy vector, could improve photoreceptor survival and retinal function when delivered before the onset of degeneration. Human RP patients typically will not visit an eye care professional until they have a loss of vision, therefore we further hypothesized that this gene therapy vector could improve photoreceptor survival and retinal function when delivered after the onset of degeneration, in a clinically relevant scenario. For each of these studies, we used histology, autofluorescence (AF) and infrared (IR) imaging to examine the appearance of the retinal cell layers and retinal pigment epithelium (RPE) that are affected in human RP patients. We also used electroretinograms (ERGs) to measure both photoreceptor-specific and global retinal visual function in the Pde6anmf363 mice. For our gene therapy experiments, we utilized a vector with the cell-type-specific rhodopsin (RHO) promoter: AAV2/8(Y733F)-Rho-Pde6a, to transduce Pde6anmf363 retinas after subretinal injection at either post-natal day (P) 5 or P21. We then monitored the effects of AAV2/8(Y733F)-Rho-Pde6a transduction over at least a quarter of the mouse lifespan. In the Pde6anmf363 mutant mouse model of RP, we found that by 2 months of age the number of photoreceptor cell nuclei is roughly halved in comparison to the 1 month time-point, and this degeneration continues until all photoreceptor cell nuclei have undergone degeneration by 4 months of age. Additionally, both loss of cone cell function and RPE atrophy are present by 5 months of age in these mice. After the development of a subretinal injection surgical procedure, we delivered the AAV2/8(Y733F)-Rho-Pde6a to the Pde6anmf363 mice at either P5 or P21. We found that a single injection enhanced survival of photoreceptors and improved retinal function. At 6 months of age, the treated eyes retained photoreceptor cell bodies, while there were no detectable photoreceptors remaining in the untreated eyes. More importantly, the treated eyes demonstrated functional visual responses even after the untreated eyes had lost all vision. Despite focal rescue of the retinal structure adjacent to the injection site, global functional rescue of the entire retina was observed. We have also determined that subretinal transduction of this rod-specific transgene at P21, when approximately half of the photoreceptor cells have undergone degeneration, has similar efficacy in rescuing cone cell function long-term as transduction before disease onset, at P5. Therefore, we concluded that the Pde6anmf363 mice mimic human RP caused by mutations in PDE6A. The establishment of the temporal and biochemical characteristics of photoreceptor neurodegeneration in the Pde6anmf363 mice allows for future studies to test therapeutic options using this animal model, since the progression of RP can be compared to the established time-course of degeneration. Additionally, the development of a standard method for performing subretinal injections allows for comparable results after this surgical technique is used to deliver gene therapy vectors into the perinatal mouse eye. Our gene therapy studies suggest that RP due to PDE6a deficiency in humans, in addition to PDE6b deficiency, is also likely to be treatable by gene therapy. Furthermore, AAV2/8(Y733F)-Rho-Pde6a is an effective gene therapy treatment that can be utilized in the clinical setting, in human patients who have lost portions of their peripheral visual field and are in the mid-stage of disease when they first present to an eye-care professional.

A large body of experimental evidence points to sophisticated signal processing taking place at the level of dendritic trees and dendritic branches of neurons. This evidence suggests that, in addition to inferring the connectivity between neurons, identifying analog dendritic processing in individual cells is fundamentally important to understanding the underlying principles of neural computation. In this thesis, we develop a novel theoretical framework for the identification of dendritic processing directly from spike times produced by spiking neurons. The problem setting of spiking neurons is necessary since such neurons make up the majority of electrically excitable cells in most nervous systems and it is often hard or even impossible to directly monitor the activity within dendrites. Thus, action potentials produced by neurons often constitute the only causal and observable correlate of dendritic processing. In order to remain true to the underlying biophysics of electrically excitable cells, we employ well-established mechanistic models of action potential generation to describe the nonlinear mapping of the aggregate current produced by the tree into an asynchronous sequence of spikes. Specific models of spike generation considered include conductance-based models such as Hodgkin-Huxley, Morris-Lecar, Fitzhugh-Nagumo, as well as simpler models of the integrate-and-fire and threshold-and-fire type. The aggregate time-varying current driving the spike generator is taken to be produced by a dendritic stimulus processor, which is a nonlinear dynamical system capable of describing arbitrary linear and nonlinear transformations performed on one or more input stimuli. In the case of multiple stimuli, it can also describe the cross-coupling, or interaction, between various stimulus features. The behavior of the dendritic stimulus processor is fully captured by one or more kernels, which provide a characterization of the signal processing that is consistent with the broader cable theory description of dendritic trees. We prove that the neural identification problem, stated in terms of identifying the kernels of the dendritic stimulus processor, is mathematically dual to the neural population encoding problem. Specifically, we show that the collection of spikes produced by a single neuron in multiple experimental trials can be treated as a single multidimensional spike train of a population of neurons encoding the parameters of the dendritic stimulus processor. Using the theory of sampling in reproducing kernel Hilbert spaces, we then derive precise results demonstrating that, during any experiment, the entire neural circuit is projected onto the space of input stimuli and parameters of this projection are faithfully encoded in the spike train. Spike times are shown to correspond to generalized samples, or measurements, of this projection in a system of coordinates that is not fixed but is both neuron- and stimulus-dependent. We examine the theoretical conditions under which it may be possible to reconstruct the dendritic stimulus processor from these samples and derive corresponding experimental conditions for the minimum number of spikes and stimuli that need to be used. We also provide explicit algorithms for reconstructing the kernel projection and demonstrate that, under natural conditions, this projection converges to the true kernel. The developed methodology is quite general and can be applied to a number of neural circuits. In particular, the methods discussed span all sensory modalities, including vision, audition and olfaction, in which external stimuli are typically continuous functions of time and space. The results can also be applied to circuits in higher brain centers that receive multi-dimensional spike trains as input stimuli instead of continuous signals. In addition, the modularity of the approach allows one to extend it to mixed-signal circuits processing both continuous and spiking stimuli, to circuits with extensive lateral connections and feedback, as well as to multisensory circuits concurrently processing multiple stimuli of different dimensions, such as audio and video. Another important extension of the approach can be used to estimate the phase response curves of a neuron. All of the theoretical results are accompanied by detailed examples demonstrating the performance of the proposed identification algorithms. We employ both synthetic and naturalistic stimuli such as natural video and audio to highlight the power of the approach. Finally, we consider the implication of our work on problems pertaining to neural encoding and decoding and discuss promising directions for future research.

The hippocampus is critical for storing declarative memory, our repository of knowledge of who, what, where, and when. Mnemonic information is processed and encoded in the hippocampus through several parallel routes, most notably the trisynaptic pathway, in which information proceeds from entorhinal cortex (EC) to dentate gyrus (DG) to CA3 and then to CA1, the main hippocampal output. Absent from this pathway is the CA2 subfield, a relatively small area interposed between CA3 and CA1 that has recently been shown to mediate a powerful disynaptic circuit linking EC input with CA1 output. Usually ignored or grouped together with CA3, CA2 has generally escaped exploration presumably due to its relatively small size and somewhat ill-defined borders. A few studies have proposed an important role for the CA2 subfield of the hippocampus, however, the relevance of this subfield in a behaving animal has not been explored. The function of a particular brain region may be inferred by examining the effects of a lesion of that area. Indeed, the hippocampus's role in learning and memory was elucidated following the bilateral medial temporal lobe ablation of Henry Molaison (patient H.M.). Similarly, a lesion of CA2 could be used to infer its role in learning, memory, and disease. Due to the relatively small size of CA2, physical or chemical lesions are not precise enough to ablate this region without collateral damage. To overcome this limitation, I generated a CA2-specific transgenic mouse line to enable genetic targeting of this subfield. I used this mouse line to map CA2 connectivity and explore its behavioral role. Using monosynaptic rabies tracing, CA2 axon tracing, and electrophysiology, I confirmed the disynaptic pathway and presence of septal and subcortical inputs to CA2. Genetically targeted inactivation of CA2 caused a remarkably profound loss of social memory, with no change in sociability. This impairment was not the result of a general loss of hippocampal function as CA2-inactivation did not impact performance on several other hippocampal-dependent tasks, including spatial and contextual memory. These behavioral and anatomical results thus reveal CA2 as a hub of sociocognitive processing and implicate its dysfunction in social endophenotypes of psychiatric diseases such as schizophrenia and autism.

Functional connectivity networks that integrate remote areas of the brain as working functional units are thought to underlie fundamental mechanisms of perception and cognition, and have emerged as an active area of investigation. However, traditional approaches of measuring functional connectivity are limited in that they rely on a priori specification of one or a few brain regions. Therefore, the development of data-driven and exploratory approaches that assess functional connectivity on a large-scale are required in order to further understand the functional network organization of these processes in both health and disease. In this thesis project, I investigate the roles of functional connectivity in visual search (Chapter 2, (Pantazatos, Yanagihara et al., 2012)) and bistable perception (Chapter 3, (Karten et al., 2013)) using traditional functional connectivity approaches, and develop and apply new approaches to characterize the large-scale networks underlying the processing of supraliminal (Chapter 4, (Pantazatos et al., 2012a)) and subliminal (Chapter 5, (Pantazatos, Talati et al., 2012b)) emotional threat signals, speech and song processing in autism (Chapter 6, (Lai et al., 2012)), and face processing in social anxiety disorder (Chapter 7, (Pantazatos et al., 2013)). Finally, I complement the latter study with an investigation of structural morphological abnormalities in social anxiety disorder (Chapter 8, (Talati et al., 2013)). Each of these chapters has been or is about to be published in peer reviewed journals and this thesis provides an overview of the entire body of investigation, based on advances in understanding the role of large-scale neural processes as fundamental organizational units that underlie behavior. In Chapter 2, Independent Components Analysis (ICA), Psychophysiological Interactions (PPI) and Dynamic Causal Modeling (DCM) analyses were used to investigate the hypothesis that expectation and attention-related interactions between ventral and medial prefrontal cortex and association visual cortex underlie visual search for an object. Results extend previous models of visual search processes to include specific frontal-occipital neuronal interactions during a natural and complex search task. In Chapter 3, PPI analyses revealed percept-dependent changes in connectivity between visual cortex, frontoparietal attention and default mode networks during bistable image perception. These findings advance neural models of bistable perception by implicating the default mode and frontoparietal networks during image segmentation. In Chapters 4 and 5, an exploratory approach based on multivariate pattern analysis of large-scale, condition-dependent functional connectivity was developed and applied in order to further understand the neural mechanisms of threat-related emotion processing. This approach was successful in extracting sufficient information to "brain-read" both unattended supraliminal (Chapter 4) and subliminal (Chapter 5) fear perception in healthy subjects. Informative features for supraliminal fear perception included functional connections between thalamus and superior temporal gyrus, angular gyrus and hippocampus, and fusiform and amygdala, while informative features for subliminal fear perception included middle temporal gyrus, cerebellum and angular gyrus. In psychiatric disorders, large-scale functional connectivity is typically assessed during resting-state (i.e. no task or stimulus). However, disorder-dependent alterations in functional network architecture may be more or less prominent during a stimulus or task that is behaviorally relevant to the disorder, as is exemplified by enhanced long-range, frontal-posterior connectivity during song (vs. speech) perception in autism (Chapter 6). In the case of social anxiety disorder (SAD), pattern analysis of large-scale, functional connectivity during neutral face perception was sensitive enough to discriminate individual subjects with SAD from both healthy controls and panic disorder (Chapter 7). The most informative feature was functional connectivity between left hippocampus and left temporal pole, which was reduced in medication-free SAD subjects, and which increased following 8-weeks SSRI treatment, with greater increases correlating with greater decreases in symptom severity. This finding parallels results from observed neuroanatomical abnormalities in SAD, which include reduced grey matter volume in the temporal pole, in addition to increased grey matter volume in cerebellum and fusiform (Chapter 8). The above findings suggest promise for emerging functional connectivity and structural-based neurobiomarkers for SAD diagnosis and treatment effects.

The goal of this dissertation is to explore the neural correlates of endogenous task-related attentional modulations. Natural fluctuations in task engagement are challenging to study, primarily because they are by nature not event related and thus cannot be controlled experimentally. Here we exploit well-accepted links between attention and various measures of neural activity while subjects perform simple target detection tasks that leave their minds free to wander. We use multimodal neuroimaging, specifically simultaneous electroencephalograpy and functional magnetic resonance imaging (EEG-fMRI) and EEG-pupillometry, with data-driven machine learning methods and study activity across the whole brain. We investigate BOLD fMRI correlates of EEG variability spanning each trial, enabling us to unravel a cascade of attention-related activations and determine their temporal ordering. We study activity during auditory and visual paradigms independently, and we also combine data to investigate supra modal attention systems. Without aiming to study known attention-related functional brain networks, we found correlates of attentional modulations in areas representative of the default mode network (DMN), ventral attention network (VAN), locus coeruleus norepinephrine (LC-NE) system, and regions implicated in generation of the extensively-studied P300 EEG response to target stimuli. Our results reveal complex interactions between known attentional systems, and do so non-invasively to study normal fluctuations of task engagement in the human brain.