Alessandra Angelucci - US grants

A prominent organizational feature of the visual cerebral cortex of primates is the existence of multiple areas interconnected by a dense network of feedforward and feedback connections, and the presence of long horizontal axons linking distant loci within any given visual cortical area. The long-term goal of the present research is to disentangle the relative roles of these different systems of neuronal connections in the generation of specific visual cortex neuron responses, and ultimately in visual perception.
A basic feature of neurons at the early stages of visual cortical processing is that they respond to the presentation of specific visual stimuli within a localized region of space, called the neuron's receptive field (RF). Presentation of similar stimuli outside the neuron's RF, typically does not evoke a response from the neuron. Feedforward connections are thought to underlie the responses of visual cortex neurons to localized imaged features (such as the specific orientation of line segments), i.e. to stimuli confined to the neuron's RF.
More recently, it has been demonstrated that visual cortex neurons respond not only to details within their RF, but also to the context in which they appear. In other words, neuronal responses are affected by the perceptual organization of the visual scene as a whole. For example, the responses of neurons in visual cortical areas V1 and V2 to stimuli within their RF can be modulated (i.e. facilitated or suppressed) by stimuli lying outside their RF, i.e. in the RF's surround, even though stimulation of the surround region alone does not evoke a response from the cells. Specifically, the response to a bar stimulus presented in the RF at the orientation preferred by the cell is enhanced by the simultaneous presentation of similarly oriented and co-aligned bars in the surround. However, surrounding bars of orthogonal orientation to that in the RF have no effect on the cell's response. This property of visual cortical neurons (defined as surround modulation) is thought to represent the neural mechanism by which fragmented line segments are integrated to form perceptually complete contours. These phenomena require fast communication between distant parts of the visual image. Horizontal and/or feedback connections are thought to carry out the long distance computations underlying the perceptual integration of oriented line segments into contours, surround modulation of RF responses, and possibly visual attention.
The goal of this application is to disentangle the relative roles of feedforward, feedback and horizontal connections in the generation of single neurons' responses within and outside the RF. As an initial step towards this goal, the present research is designed to investigate how the spatial extent and organization of the different types of connections within and between primate areas V1, V2 and V3 relate to the spatial dimensions and organization of single V1, V2 and V3 cells' RF and RF's surround. Each system of connections will be visualized using neuroanatomical labeling methods, by making small injections of a neuroanatomical tracer in macaque V1, V2 or V3. The extent of the connections in visual space will be determined and compared with the spatial dimensions of electrophysiologically measured RF and surround field of neurons at the injected cortical sites.
These studies will provide anatomical and physiological constrains necessary to rule out potential mechanisms and substrates for surround modulation and perceptual integration of contours, while also identifying potential candidates. Additionally, they will identify rules of connectivity for three different visual cortical areas. Similar rules for different areas will reveal fundamental principles of organization for these connectional systems in visual cortex and, possibly, in other sensory cortices, with important functional and developmental implications. This research will also provide a wealth of quantitative anatomical and physiological data essential to the generation of biologically-based theoretical models of interactions within and beyond the RF of visual cortex neurons. These data will be disseminated and shared with students and faculty of mathematical biology, both through weekly discussion meetings, and through active collaborations and laboratory rotations offered to the students. The P.I. is an adjunct member of an active IGERT training grant in mathematical biology funded by NSF, whose central goal is to promote interactions between mathematicians and biologists; this proposal will thus help strengthen and meet the aims of that grant. It will also help strengthen Systems Neuroscience, at the University of Utah, by promoting collaborations between neuroscientists in different departments within and outside the University of Utah. Finally, the proposal will promote training and education, and broaden the participation of underrepresented groups, by funding two graduate students and one post-doctoral fellow belonging to minority groups.

[unreadable] DESCRIPTION (provided by applicant): The perception of a visual "figure" often relies upon the overall spatial arrangement of its local elements, as demonstrated by the perception of occluded or illusory contours. The global attributes of a visual stimulus can affect the response of visual cortical neurons to the local attributes of the stimulus. For example, the response of neurons in cortical areas V1 and V2 to stimuli within their classical receptive field (cRF) can be modulated by contextual stimuli outside their cRF (in their "surround"). Similarly, V1 and V2 neurons respond to illusory contours (ICs) within their cRF. Neuronal cRFs and local interactions cannot account for these perceptual and neurophysiological phenomena; rather, fast interactions across distant visual field locations are needed to mediate perceptual completion. The long-term goal of this proposal is to disentangle the relative contributions of long-range inter-areal (feedforward and feedback) and intra-areal (horizontal) corticocortical connections to these global-to-local computations in early visual cortex. The neural circuitry and mechanisms involved are likely to be the cornerstone of contour integration, and figure-ground segregation. Thus, results from these studies ultimately will help understanding the neural substrates for higher visual cortical processing and perception in primate and man. [unreadable] As a first step towards the broader goal, the work described in this application is designed to investigate how the spatial extent and organization of intra-areal (Aim 1) and inter-areal, feedforward and feedback (Aims 2 and 3), connections relates to the spatial scale and organization of single V1 and V2 neurons' RF, modulatory surround field and response to ICs. The rationale is that the spatial scale of a given connectional system must be commensurate with the spatial scale of the specific neuronal response that it underlies in retinotopically-organized early visual cortex. We will map the total field of connections labeled by small tracer injected cortical points, and overlay these anatomical maps to physiologically recorded retinotopic maps from the same regions of cortex. The visuotopic scale of the connectional fields will then be related to the spatial dimensions of receptive field and modulatory surround field, and with the limits of neuronal responses to ICs, measured physiologically at the injected cortical points [unreadable] [unreadable]

An important issue in visual neuroscience is to understand how neural networks (or circuits of neurons, i.e. brain cells) in the visual cerebral cortex determine neuronal responses to visual objects and visual perception.

The goal of this project is to identify the circuitry and mechanisms for surround modulation in visual cortex. Surround modulation is the ability of visual cortical neurons to adjust their responses to an object (e.g. a vertically-oriented bar) depending on the visual context within which it is presented (e.g. a background of many vertical or horizontal bars). This phenomenon is thought to underlie ability to perceive objects in cluttered scenes. For example, a background of vertical bars suppresses the response to the single vertical bar, making it perceptually invisible. Instead, a background of horizontal bars increases the response to the vertical bar, making it perceptually to stand out from the background.

A combination of computational modeling and experimental studies will be used to identify the circuitry and mechanisms for surround modulation. Specifically, to gain insights into the underlying circuitry, the spatial, temporal and tuning properties of surround modulation will be studied using visual stimuli designed to probe the underlying pathways. These data will be used to generate computational models of visual cortical circuitry and function. Modeling will be used to understand the mechanisms generating surround modulation and its role in visual perception.

This research will uncover the neural substrates for higher visual cortical processing and perception. The interdisciplinary nature of the proposed work will create novel training opportunities for graduate students at the University of Utah. This project will support the training of several graduate students, and expose to the field of neuroscience undergraduate students from the field of bioengineering.

DESCRIPTION (provided by applicant): Understanding organizing principles for neural circuits in the cortex is necessary to understand their computational function. This principle has informed the new field of connectomics, devoted to generating wiring diagrams of the brain, or subsections of it, at different scales. In the primate visual cortex, areas V1 and V2 distribute information they receive from the retina to virtually all higher areas, sorting this information ino dorsal and ventral processing streams for spatial and object vision, respectively. The objective of this application is to uncover the rules of anatomical and functional connectivity for V1 and V2 output pathways, in order to understand how these areas may refine and re-organize retinal signals into visual processing streams, and how these pathways contribute to creating the complex receptive field (RF) properties of neurons in higher areas. Parallel pathways from V1 to V2 project to distinct cytochrome-oxidase stripes (thick, thin and pale). During the previous grant period, we discovered 4 segregated pathways between V1 & V2. We also discovered specialized functional organizations of V1 pathways related to thick and pale stripes that may underlie the responses of V2 RFs to angled and curved contours. This proposal builds upon these findings. The goal of Aim 1 is to understand how single V1 cells contribute to generating V2 RFs, by determining the axonal and dendritic layout of single V1 output cells over the V1 and V2 orientation maps. We will provide the fist comprehensive anatomical description at mesoscopic scale of V1 cells projecting to specific V2 stripes, including the intra-V1 and intra-V2 axonal arborizations of identified cell types, and their functional organization within both V1 & V2. This information will provide insights into the roles of feedforward vs. intra-V2 mechanisms in the generation of V2 RFs, and on the contribution of intra-V1 and V1-to-V2 circuits to the processing of contours. We will then investigate whether anatomical and functional segregation of the 4 pathways is maintained or lost downstream of V2. The goal of Aim2 is to determine the areal projections of the two V2 pale stripe types, which we have recently demonstrated to be distinct compartments. This study will determine each pale stripe contribution to the dorsal and ventral processing streams. V2 and V4 contain segregated representations for visual stimulus orientation and color. The goal of Aim3 is to determine whether connections between V2 & V4 occur between regions of similar featural representation, or whether cross-stream convergence occurs in V4. This study will also provide insights into the roles of feedforward vs. intra-areal mechanisms in the generation of V4 RFs and featural maps. The proposed research is significant because it will reveal anatomical and functional wiring principles for V1 and V2 output pathways that will serve as a foundation for hypothesis-driven and anatomically-constrained studies of their function. The proposed research is innovative because it combines functional imaging with high-resolution labeling of single axons, using novel methods for single axon labeling and reconstruction.

DESCRIPTION (provided by applicant): Understanding information processing in the cerebral cortex requires understanding the role of feedforward (FF) and feedback (FB) circuits between lower and higher cortical centers. Organizing principles for these circuits, that could determine how they process sensory information, remain largely unknown. This is due to the complexity of inter-areal circuits, i.e. their anatomical and functional specificity, and the lack of methodologis that can reveal the fine-scale connectivity of FF and FB circuits made by specific cell types, and relate it to the functional architecture of the cortex. Ultrastructural-scale circuit anatomy, whil useful for building wiring diagrams of local connections in mouse cortex, cannot be used to study the large cortical volumes encompassed by inter-areal axons. The latter can be studied only at mesoscopic scale. Our goal is to develop a methodology for labeling and efficiently reconstructing, single cell types and their inter-areal axons. Previous single axon studies were affected by ambiguity in the origin of the axonal label, inability to restrict label to few neurons and laborious manual reconstructions. These studies have provided only a small sample of incompletely reconstructed axons, biased towards regions of sparser labeling, with no identification of their cell types of origin. Our Specific Aims are: Aim 1. To label unambiguously at high resolution the axon of single projection neurons, and to develop a novel computational framework for semi-automated single axon reconstruction. We will extend viral-mediated expression of GFP to labeling at high resolution, and sparsely the axons of inter-areal projection neurons. We will develop a novel approach for fast serial section reconstruction of single axons, which includes 3D imaging of intact tissue blocks rendered optically-transparent, and novel computational algorithms for semi-automated axon segmentation. Aim 2. To apply these methods to resolve controversies in the literature on the functional specificity, or lack thereof, f inter-areal feedback projections to primate visual cortical area V1. Two previous studies of the layout of V2 FB projections onto the V1 orientation map have demonstrated orientation-specific, one, and unspecific FB connections, the other. Our preliminary data suggest existence of two FB systems, likely related to different cell types, which show unique relationships to the cortical functional architecture, thus providing a way to reconcile apparently contradictory data. The contribution of the proposed research is significant because it will provide new tools for studying the fine-scale connectivity and functional organization of inter-areal circuits made by specific cel types. General organizing principles for these connections will emerge that will provide an anatomical foundation for hypothesis-driven studies of their function. The proposed research is innovative because unlike previous studies: 1) the novel labeling method permits high-resolution, unambiguous identification of single inter-areal neurons, from soma to axon; 2) semi-automated mapping of 3D volumes from serial sections allows for fast reconstruction and, thus, higher yield of reconstructed axons; 3) it combines for the first time functional imaging of corticl responses with labeling of single FB axons. PUBLIC HEALTH RELEVANCE: Normal brain function depends on the orderly development of circuits in the cerebral cortex and on their intact function. Knowledge of the normal circuitry provides a foundation for understanding the causes of impaired brain function and developing corrective measures. Our proposed novel approach for studying the detailed microcircuitry of single long-distance axons can be applied to any area of the cerebral cortex and to any mammalian species, including higher mammalian species with large brains. This tool for studying the detailed inter-areal circuitry in the cerebral cortex can be used for studying the normal circuitry and how the latter is modified in altered brain function. As an example, our studies of the normal circuitry between early visual cortical areas will provide greater insight ino the causes and effects of central vision defects when these circuits are damaged by stroke or other insult. In particular, our studies on feedback circuits between different visual cortical ares will help our understanding of how these pathways mediate the influence of stimulus context on neuronal responsivity, spatial integration, and changes in visual sensitivity associated with top-down attention and learning. Understanding how these circuits operate in normal vision will provide a foundation for understanding the consequences of their dysfunction in abnormal states, such as impaired visual contextual integration in schizophrenia (Dakin et al., 2005; Yoon et al., 2010), visuo-spatial local-to-global interference in autism (Wang et al., 2007; Simmons et al., 2009), as well as disorders of visual attention and learning.

This proposal will increase our understanding of how the brain operates by revealing fundamental principles of information processing by the cerebral cortex, which is widely held to be the site of human conscious experience. Brains cannot be understood until their fundamental processing unit, the cortical column, is understood using the type of circuit level analysis proposed. In addition, the interdisciplinary nature of the proposal will offer a unique opportunity to train young undergraduate students in both neuroscience and mathematics, disciplines that increasingly are synergistic in neuroscience. Women and minority undergraduate students will be recruited to participate in the research and materials for educators of middle and high school students will be developed that deal with the neuroscience of our senses.

Cognitive functions such as thinking, perceiving and remembering, rely on the coordinated activity of neurons in the cerebral cortex. The cortical column is believed to be the functional unit of the cortex, thus understanding its circuitry and function is crucial to determine how the cortex works. The cortical column is composed of 6 interconnected layers, each with distinct function and connections. The goal of this application is to understand the role of this layered structure in cortical information processing. Primary visual cortex (V1) will serve as the model system because its anatomy and functional properties are better known than those of any other cortical area. Electrode arrays will be used to record simultaneously across all layers the responses of populations of neurons in V1 to visual stimuli that are known to drive this area of cortex. This approach will reveal how the different layers of V1 are activated in time and space by different kinds of visual stimuli. These methods will also allow an examination of the flow of visual information processing across layers of the cortex, a fundamental but heretofore understudied question. Finally, mathematical modeling will be used to understand how complex interactions between connections within a cortical column compute the neuronal responses specific to each layer.

? DESCRIPTION (provided by applicant): Hierarchical feedforward models have provided a foundation for most theories of visual processing over the past 40 years. However, in the cerebral cortex there is a dense network of feedback (FB) connections sending topdown information from higher to lower processing centers. The anatomy and function of FB circuits is still poorly understood. This is despite the fact that FB has been implicated in many important visual functions such as contextual modulation, attention and learning. The goal of this application is to understand the anatomical and functional organization of FB connections from the secondary (V2) to the primary visual cortex (V1), and the impact of V2 FB on V1 cell responses. We will test the following hypotheses: 1. Anatomically, V2 FB consists of parallel channels, related to the stripe compartments of V2, with each channel targeting specific V1 layers, V1 compartments and the same V1 cells that provide input to the specific V2 stripe (Aim 1). The different channels also show unique specificities with respect to the functional maps of visual stimulus properties in V1 (Aim 2). 2. Functionally, V2 FB contributes to surround modulation (SM) caused by visual stimulation of the surround regions near the V1 cell's receptive field (RF) (Aim 3). SM is the ability of stimuli in the RF surround to modulate the responses of V1 cells to stimuli inside the RF, a property thought to serve efficient coding of natural images and segmentation of object boundaries (Nurminen & Angelucci, 2014) Ref43. Our understanding of FB anatomy and function has been hampered by the technical limitations of previous methods used to label FB axons and manipulate their activity. Conventional tracers are either poorly sensitive or label axons bidirectional, creating a confound in the interpretation of the axon label. As a result, there are conflicting reports regarding the anatomical and functional specificity of FB connections. Inactivation of FB systems has been performed using methods that lack spatiotemporal precision and cell type selectivity, affecting cells in an entire cortical area. These approaches could not rule out that the effects of FB manipulation were mediated by other indirect pathways. To overcome these limitations, we will investigate the anatomical and functional organization of FB connections from V2 to V1, using novel viral-mediated expression of fluorescent proteins to label FB axons and their synaptic terminals unambiguously. In Aim 1 these labeling methods will be combined with staining of V1 layers and compartment (CO maps), and in Aim 2 with optical imaging of V1 functional maps. In Aim 3, manipulation of FB neuron activity will be performed using optogenetic activation of opsin-expressing FB axons, while determining the effects of these manipulations on spontaneous and visually-evoked spike activity in V1 cells, recorded using linear arrays. These data will provide an anatomical and mechanistic foundation for modeling and hypothesis-driven studies of FB function and dysfunction, and lead to a better understanding of the computations performed by this canonical cortical circuit, i.e. top down FB.

Obtaining a "connectome" or map of the wiring of the brain is crucial to understanding brain structure and function, and has been set as long-term goals of several international government-funded initiatives due to the potential benefits for improving health, treating brain diseases, and understanding development. As technologies for sample preparation and microscopy advance, it is becoming feasible to image large sections of brain tissue. However, the vast quantities of data produced with these techniques is far outpacing the ability of neuroscientists to analyze the data. This project will address the data analysis challenge by developing new computational software tools that facilitate use of advanced computing for connectomics studies, in alignment with NSF's mission to promote the progress of science and advance national health, prosperity and welfare.

Understanding the microarchitecture and neuronal morphologies that comprise neural circuitry in the brain is crucial to understanding brain function. This EAGER project aims to build the computational and data infrastructure that is necessary to manage and process large microscopy imaging data sets for connectomics studies, bringing High Performance Computing (HPC) resources into the neuroscience workflow. The project will employ a data model that enables scientists to visualize, interact with, and process data of any size that is stored in any remote location, from USB drives to high-performance parallel file systems. The software infrastructure will furthermore enable automatic mapping of analysis procedures designed by neuroscientists to remote HPC systems. The system will leverage state-of-the-art tools and practices developed in the HPC community, and aims to result in greatly accelerated studies of connectivity in the brain at scale.

This Early-concept Grants for Exploratory Research (EAGER) award by the CISE Division of Advanced Cyberinfrastructure is jointly supported by the CISE Division of Information and Intelligent Systems, with funds associated with the NSF Understanding the Brain, BRAIN Initiative activities, and for developing national research infrastructure for neuroscience. This project also aligns with NSF objectives under the National Strategic Computing Initiative.

Understanding the function of neural circuits in the cerebral cortex of the non-human primate (NHP), the model system closest to human, is crucial to understanding normal cortical function and the circuit-level basis of human brain disorders. Optogenetics has emerged as a powerful tool for studying neural circuit function, by using light to perturb the activity of specific cell types genetically modified to express light-activated microbial opsins, and assessing the consequences of this perturbation on network activity and behavior. While successful in mice, it has been challenging to apply optogenetics to NHPs, largely due to the lack of multifunction integrated probes for precision light delivery and electrophysiology across mm-to-cm volumes through the depth of the NHP cortex. Large volume manipulations are essential in the large NHP brain in order to observe measurable electrophysiological or behavioral effects. An interdisciplinary team of PIs proposes to develop and test in vivo integrated penetrating arrays that allow for large-volume, spatiotemporally patterned optogenetic modulation and electrical recording of neural circuits in the NHP brain. This project requires the coordinated effort of 4 teams, including experts in photonic devices and µLED development for optogenetics, materials and packaging for biocompatible devices, primate neurophysiology, and pioneers in electrode array design and commercialization. In Aim 1 we develop the technology, and in Aim 2 we test it in vivo in the NHP visual cortex. We will initially develop a 4x4 mm penetrating 10x10 optrode array in a format analogous to the well- established Utah Electrical Array (UEA), with each probe serving as a waveguide allowing visible light to reach tissue depths >1.5mm. Following initial optimization of the probe's shank diameter and tip angle to minimize tissue damage, we will perform proof-of-concept in vivo NHP optogenetic experiments in deep cortical tissue, using broad-area illumination of the entire array. In a second stage, we will develop light coupling via µLEDs, which will be integrated into a single platform and tested in vivo, consisting of a µLED located over each optical probe. Completion of stage 2 will deliver a functional multioptrode array for large-volume patterned optogenetic stimulation. Parallel engineering efforts will add electrical recording capability, by utilizing the engineering resources already in place for the UEA, and will generate two types of integrated arrays. The ?interleaved? array consists of an optrode array inserted through the back plane of a modified UEA into which a grid of through-backplane holes is made via laser ablation to accommodate the optrodes. For the ?hybrid? array, each optrode shank will be coated with an isolation layer followed by a conductive layer, in order to allow recording while preventing light attenuation and stimulation artifacts. In vivo testing will assess the recording capabilities of both devices and subsequently the ability to perform simultaneous optical stimulation and electrical recordings. This technology will allow for unprecedented optogenetic investigations of mm-to-cm scale neural circuit function and dysfunction in NHPs, and for a new generation of therapeutic interventions via cell type specific optical neural control prosthetics.

PROJECT SUMMARY Hierarchical feedforward (FF) models of the visual system have provided a foundation for most theories of visual processing over the past 40 years. It is problematic that these models have relied on probabilistic representations of FF connectivity between visual areas, partly because we lack information about the wiring principles and functional organization of FF connections, even between the earliest visual areas V1 and V2, which have been studied for decades. Our goal is to uncover the rules of anatomical and functional connectivity for V1 output pathways to V2 to anatomically-constrain FF models of vision. This information is crucial to understand how V1 and V2 reorganize retinal signals into processing streams, and how V1 output pathways contribute to generating the receptive field (RF) properties and functional maps in V2. In primates, parallel pathways from V1 project to distinct V2 cytochrome-oxidase (CO) stripes [thick (Tk), thin (Tn) and pale (Pl)]. It is unknown whether the local connections of V1 output cells, and their V1 inputs integrate information across parallel streams or maintain within-stream segregation. It is also debated whether distinct V1 and V2 CO compartments show specialized or diverse visual response properties, partly because it has been difficult to record from identified V1 output cells. During prior funding period, we found that the local intra-V1 connectivity of V1 cells projecting to Tk stripes shows within-stream segregation. It is, thus, important to extend these studies to V1 cells projecting to Tn and Pl stripes. To address this goal, we will label these cells using viral vectors, reconstruct them through whole V1 and V2 blocks rendered optically transparent, and align them to CO and functional maps of V1 and V2 (Aim1). Using viral-based monosynaptic circuit tracing combined with CO-staining and optical imaging (OI) of functional maps, we will test the hypothesis that the V1 and V2 inputs to V1 cells projecting to distinct V2 stripes are also stream specific, arising from the same CO and functional compartments as the V1 output cells that they contact (Aim2). We will also test the hypothesis of functional segregation in the monosynaptic projections from V1 to V2 stripes, by characterizing the visual responses of optogenetically-identified V1 cells projecting to distinct stripes (Aim3). Finally, it is unknown how V1 inputs are combined within local V2 columns in each stripe; this information is crucial to understand how V1 inputs contribute to generating the more complex RF properties of V2 cells. We will address this question both anatomically and functionally (Aim4). In SubAim4a, we will determine how V1 inputs to single V2 columns are distributed over the V1 functional maps, by combining OI of functional maps with injections of retrograde tracers in V2. In SubAim4b we will characterize the population RF of V1 inputs to a local V2 column, using simultaneous array recordings in V1 and V2 and spike-triggered CSD analysis. Thes studies will reveal how V1-to-V2 circuits are anatomically and functionally organized, their degree of specialization, and how V1 inputs are pooled in V2 to generate V2 RFs. The results will provide a mechanistic foundation for modeling studies of FF and parallel processing in visual cortex.

PROJECT SUMMARY In the mammalian neocortex, inhibitory neurons (INs) profoundly influence cortical computations and dynamics, and their various functions are thought to be mediated by different IN types. While a large diversity of INs exists, molecular markers in mouse cortex identify three major non-overlapping classes: parvalbumin- (PV), somatostatin- (SOM), and vasoactive intestinal peptide- (VIP) INs. Studies in mouse lines expressing Cre-recombinase in each of these IN classes are rapidly revealing distinct patterns of connectivity, response properties and in vivo function for each class. However, it remains unknown whether insights gained from mouse cortex apply to cortical INs in primates and humans. IN dysfunction in humans has been implicated in several disorders, such as epilepsy, schizophrenia, anxiety and autism, therefore it is important to understand normal cortical IN connectivity and function in primates. The lack of viral tools to selectively access IN subtypes, and the difficulty of performing genetic manipulations in primates have been major impediments to studying INs in this animal model. Our goal is to leverage recent advances in the development of viral tools to express transgenes in specific INs subtypes to investigate the connectivity, response properties, and computational function of two major classes of INs, PV and SOM, in the superficial layers of the primate primary visual cortex (V1). Using IN-type specific expression of Cre-recombinase combined with rabies-virus monosynaptic circuit tracing, we will map local and brain-wide inputs to specific V1 IN classes (Aim1). Using two-photon imaging of IN-type specific targeted calcium indicators, or optogenetic identification of channelrhodopsin-tagged IN types, we will characterize the visual response properties of distinct V1 IN classes (Aim2). Finally, we will use optogenetic inactivation of distinct IN-types expressing inhibitory opsins, to understand the relative roles of IN classes in V1 computations (Aim3). We will test specific hypotheses derived from available data in mouse, the specific geometry of PV and SOM cells in primate cortex, published computational models of feature tuning and surround suppression in visual cortex, and our preliminary results. Impact. The proposed studies will provide the first account of the connectivity, visual properties and computational function of PV and SOM INs in primate cortex, paving the way for studies of IN function in this order. They will also reveal conserved principles of IN function across species, as well as fundamental inter- species differences, stressing the importance of studying cortical function in species that are evolutionarily closer to humans.