John HR Maunsell - US grants

In the macaque monkey the parvocellular and magnocellular layers of the lateral geniculate nucleus (LGN) relay qualitatively different types of information. These two channels of visual information have been extensively studied and are known to have separate projections to striate cortex, and to remain largely segregated within that area. Although many lines of evidence indicate that these channels retain their distinctions in later stages of cortical processing, as yet there is no conclusive evidence regarding their fate in extrastriate cortex. Given the longterm objective of understanding information processing in primate visual cortex, it is important to know if these channels contribute directly to the pronounced physiological differences seen among extrastriate visual areas. Experiments will be carried out to determine the relative contributions of the parvocellular and magnocellular channels and degree of mixing in two extrastriate visual areas in the macaque: V4 and MT. These areas have been selected because they are the best candidates for receiving differential inputs from the parvocellular and magnocellular channels, they are relatively well characterized in terms of other properties, they are at similar stages in cortical processing, and they are readily accessible. Several approaches will be employed to study these areas: 1) measurement of cortical responses following reversible or irreversible inactivation of selected layers of the LGN, 2) determination of the subcortical conduction speed of the axons which indirectly feed the two areas, and 3) comparing the contrast sensitivity and visual response latency and transience of neurons in the two areas. A further objective is to examine V2, which projects to both V4 and MT, in order to examine the role it plays in determining the degree of segregation of these channels in extrastriate cortex. In addition to answering a fundamental question about information processing primate visual cortex, these experiments could provide a link between channels which exist in the early stages of the visual system and differences in physiological properties which are seen among the various areas in cortex. Such a link could provide valuable information for psychological studies of the contributions of various cortical areas to specific visual capabilities.

The idea that the primate visual system contains parallel streams of processing that arise in the retina remain relatively independent throughout visual cortex has recently commanded considerable attention, and has led to far-reaching speculation about its functional consequences, Parallel organization is obvious at all levels of the geniculocortical system, including the sub-cortical pathway leading to cortex, the early stages of cortex, and the highest levels of cortical processing. However, the mapping of parallel components in one level onto those at the next has not been established conclusively. It is not known whether components in each stage map directly onto those at the next level, creating truly parallel pathways, or if instead a gradual intermixing eliminates any exclusive relationship between components in the different stages. While many lines of evidence suggest that the geniculocortical system in fact does contains largely independent pathways. there is also clear evidence of intermixing. The proposed experiments will continue on-going work that is aimed at evaluating the degree of parallel organization in monkey visual cortex. Conclusive answers cannot be provided either by anatomical studies or by those that examine similarities in response properties at different levels. Instead, the proposed experiments will record from cortical cells while reversibly inactivating individual magnocellular or parvocellular layers in the lateral geniculate nucleus (LGN). This is a method that the investigators have applied successfully to demonstrate that neurons in the middle temporal visual area (MT) depend primarily on the magnocellular layers on the LGN for their excitatory drive. The proposed experiments will address several specific questions about segregation of magnocellular (M channel) and parvocellular (P channel) contributions to visual cortex. 1) Does the degree of P and M channel segregation seen at the level of MT persist throughout visual cortex, or does it gradually erode in later stages? 2) Given that the P and M channels have been shown to mix to some extent, what are the special properties conferred by mixing? 3) How are the (two) P and M channels distributed across the three pathways that have been identified in the early stages of visual cortex using cytochrome oxidase stain?

The idea that the primate visual system contains parallel streams of processing that arise in the retina remain relatively independent throughout visual cortex has recently commanded considerable attention, and has led to far-reaching speculation about its functional consequences, Parallel organization is obvious at all levels of the geniculocortical system, including the sub-cortical pathway leading to cortex, the early stages of cortex, and the highest levels of cortical processing. However, the mapping of parallel components in one level onto those at the next has not been established conclusively. It is not known whether components in each stage map directly onto those at the next level, creating truly parallel pathways, or if instead a gradual intermixing eliminates any exclusive relationship between components in the different stages. While many lines of evidence suggest that the geniculocortical system in fact does contains largely independent pathways. there is also clear evidence of intermixing. The proposed experiments will continue on-going work that is aimed at evaluating the degree of parallel organization in monkey visual cortex. Conclusive answers cannot be provided either by anatomical studies or by those that examine similarities in response properties at different levels. Instead, the proposed experiments will record from cortical cells while reversibly inactivating individual magnocellular or parvocellular layers in the lateral geniculate nucleus (LGN). This is a method that the investigators have applied successfully to demonstrate that neurons in the middle temporal visual area (MT) depend primarily on the magnocellular layers on the LGN for their excitatory drive. The proposed experiments will address several specific questions about segregation of magnocellular (M channel) and parvocellular (P channel) contributions to visual cortex. 1) Does the degree of P and M channel segregation seen at the level of MT persist throughout visual cortex, or does it gradually erode in later stages? 2) Given that the P and M channels have been shown to mix to some extent, what are the special properties conferred by mixing? 3) How are the (two) P and M channels distributed across the three pathways that have been identified in the early stages of visual cortex using cytochrome oxidase stain?

DESCRIPTION (provided by applicant): Attention to particular stimuli greatly improves performance that depends on those stimuli, while degrading performance on other stimuli. Neurophysiological studies have shown that attention changes the responses of neurons in visual cerebral cortex, but many questions remain about the neuronal mechanisms through which attention alters behavior. The proposed experiments will address two specific questions about how spatial attention affects visual processing in the visual cortex of monkeys. The first specific aim will examine how attention affects the responses of individual neurons in visual cortex. Many studies have shown that attending to a stimulus enhances the responses of neurons that represent that stimulus, but few have examined the form of this enhancement. Measurement of the effects of attention on responses to stimuli of different orientations and different contrasts have led to different views about whether attention acts by uniformly increasing the strength of neuronal responses to all stimuli. Experiments of the first specific aim will resolve this discrepancy by examining interactions among attention, orientation and contrast in determining the responses of individual neurons. Additionally, they will examine about how attention affects the timing of visual responses. The second specific aim will examine how attention affects the relationship between neuronal responses and behavior. It has been observed that the ability of individual neurons to discriminate stimuli can approach or match the performance of the subject, suggesting a close link between neuronal and behavioral performance. Recent results show that attention alters this link for some neurons, but leave open the possibility that a close relationship between neuronal and behavioral performance persists across attentional states for those neurons that are best suited for current task. The second specific aim will test this possibility by examining how attention affects the relationship between neuronal and behavioral performance for neurons during the performance of different visual tasks. The results from these experiments will greatly extend our understanding of how attention changes visual representations in cerebral cortex and improves behavioral performance, and will provide new insight about how individual neurons contribute to visual behaviors.

ABSTRACT NOT PROVIDED

DESCRIPTION (provided by applicant): The overall goal of this project is to understand how sensory signals are translated into commands for the control of movements. More specifically, the proposed experiments study the role of the primate superior colliculus (SC) and the pontomedullary reticular formation in the control of orienting movements of the eyes and head. The first set of experiments has two goals: 1) to describe the functional properties of neurons in the tecto-recipient regions of pontomedullary reticular formation during coordinated eye-head movements and during head movements in the absence of a gaze shift; and 2) to study the effects of microstimulation of these pontomedullary regions upon eye and head movements. These studies target the brainstem areas that receive dense projections from intermediate layers of SC and also have a high density of neurons projecting to regions of the spinal cord innervating the neck muscles. In the second set of experiments neurons in the rostral paramedian pontine reticular formation will be reversibly inactivated to see if the activity of these cells is critical for the eye and head components of coordinated eye-head movements or for head movements made in the absence of a gaze shift. The third set of experiments has three goals: a) to record from neurons in the superior colliculus using the same behavioral tasks that were used to record from cells in pontomedullary regions; b) to use a search strategy that will increase the probability of finding cells in deeper layers of superior colliculus with activity related to head movements; and c) to examine the possible contribution of collicular neurons to the unusual eye velocity profiles observed during large gaze shifts. The final set of experiments uses a novel method of varying the speed and amplitude of eye and head movements during a gaze shift while recording the activity of pontomedullary neurons in an attempt to understand how gaze, eye and head amplitude and velocity is encoded in the activity of pontomedullary neurons.

DESCRIPTION (provided by applicant): The overall goal of the proposed Vision Core is to enhance vision research at Harvard by providing vision scientists with resources and facilities that could not be supported by individual laboratories. A further objective of the Vision Core is to provide a focus that will encourage collaborations between vision scientists and encourage other researchers to do research on vision and the eye. The Vision Core will provide continuing support for four existing core modules that have been serving the needs of the vision research community during the current funding period. A neural imaging module provides resources to support experiments using a wide variety of state-of-the-art imaging techniques, including in vivo two-photon microscopy, functional imaging and quantitative microscopy. A machine shop module provides services for designing and manufacturing special equipment that cannot be obtained commercially. A laboratory computer service module provides assistance in interfacing laboratory equipment and in designing and developing code to support experiments. Finally, an electronics module provides services for designing and constructing specialized electronic circuits and devices.

DESCRIPTION (provided by applicant): Virtually all studies of the neuronal mechanisms that underlie attention compare average brain activity when subjects are asked to attend to different objects, locations or features. Implicit in this approach is the assumption that a subject's state of attention remains relatively constant when they follow a given instruction. Yet anyone who has performed an attention-demanding task knows that attention can drift rapidly and widely. Experimenters can specify attentional conditions precisely, but have limited and indirect influence on a subject's actual attentional state on a given experimental trial. This is an important limitation, because fluctuations in attention can systematically affect estimates of how attention influences neuronal activity and behavioral ability. Assessing the actual attentional state of a subject at a given moment has not been easy to accomplish. However, in recent experiments we have discovered that appropriate analysis of the signals from populations of individual neurons recorded simultaneously in visual cerebral cortex can provide a precise measure of current attentional state. These measurements confirm that attention fluctuates widely even when instructions, stimuli and rewards are fixed. Moreover, they reveal that these fluctuations in attention are associated with profound differences in behavioral performance within fixed experimental situations. The ability to measure attentional state within individual trials provides a powerful tool for exploring how attention affects sensory signals and behavioral performance. We propose experiments that will exploit this novel method to explore questions about the neuronal basis of visual attention that have previously been inaccessible. Our first specific aim is to use simultaneous recordings from different populations of individual V4 neurons to determine whether the control of attention at different retinotopic representations is coordinated or independent. Our second specific aim involves analogous experiments that will provide direct measurements of whether neuronal control of spatial and feature attention is coupled. The final specific aim is to measure attentional state as a function of time within trials to compare the dynamics of stimulus responses with those of attentional shifts produced by exogenous or endogenous cues and those of the behavioral consequences of the attention shifts. Collectively, these experiments will provide new insights into the mechanisms that allocate attention to different visual stimuli and how the neuronal modulations they cause improve behavioral performance. PUBLIC HEALTH RELEVANCE: Attention is critical to perceptual and cognitive performance, and attention deficits are the most commonly diagnosed behavioral disorder of childhood, with attention deficit hyperactivity disorder (ADHD) affecting as many as 5% of children in the United States [1]. Better understanding of basic neuronal mechanisms related to attention and their interaction with sensory signals is needed for guiding assessment, diagnosis and treatment of deficits of attention. The proposed research will investigate how attention affects visual processing in the nervous system, and in particular how rapidly and independently control signals related to attention are distributed to different regions of visual cerebral cortex.

DESCRIPTION (provided by applicant): Twenty-one neuroscientists within Harvard's Neuroscience Program request continued funding for six pre-doctoral positions within our. Training focuses on the study of visual pathways from retina to brain, and cellular, molecular and developmental neurobiology of the visual system. The faculty are distributed throughout the university. Eleven faculty members are in basic science departments at the Medical School, six are in hospital based laboratories, and four are in the Faculty of Arts and Sciences. Over the past 15 years, Harvard University has greatly expanded the number faculty members who study the molecular, developmental, and neural-systems approaches to visual science. Students can choose laboratories among a large community of vision researchers, most of whom are affiliated with the NEI Core Grant in Vision research. The new goal of the Visual Neuroscience Training Program is to build a larger, coherent group of students who have a sense of community based in this Harvard-wide vision community, and who are trained by its faculty. The grant will support three students each in their second and third years, after they have chosen a dissertation lab, but advanced students remain actively involved with the program. This creates a large cohort of affiliated students. We train and supervise these students with courses, thesis committees, seminars, symposia, a Training Grant retreat, and our Systems-Vision journal club. Thus throughout their graduate careers, trainees interact with the faculty and with each other. Many vision scientists visit Harvard every year to give seminars; trainees at all levels interact with them over lunch and in lab visits. Through these activities, we will help train a new generation of vision scientists whose scientific careers will help us understand all aspects of the visual system: development, information processing, and disease.

? DESCRIPTION (provided by applicant): A key aspect of brain function is how the activity of neuronal populations encodes information that is used to guide behavior. A longstanding model system to understand population coding is the visual cerebral cortex, because its structure and anatomy are well understood, and because visual stimuli can be presented to subjects with high levels of temporal and spatial control. Thousands or more neurons fire action potentials in response to a single visual stimulus, and an important open question is how this population response carries information - how the detailed timing and pattern of these spikes across neurons is decoded to guide behavior. Because it is known that genetics controls the identity and morphology of neurons, and influences which other neurons they form synaptic partners with, it appears likely that the precise details of which neurons in a population fire spikes is vitally important for behavior. But surprisingly, past experimental work hints that the primary quantity governing neuronal coding is the total number of spikes or average firing rate across a population, making the precise timing and spatial distribution of those spikes less important. Theoretical work shows that either type of code can be supported by the cortex and that the type of code used may even vary from one behavioral task to the next. However, it has not been possible to definitively determine how cortical population codes are used for behavior because of the inability to change the activity of neurons in a patterned fashion. In this project, we will use two-photon ontogenetic stimulation to activate patterns of neurons in behaving animals to understand the details of how population codes control behavior. This work is made possible by the combination of optical wave front-shaping methods to control the size and shape of a two- photon optical focal volume, and psychophysical behavioral methods in mice that allow precise quantification of animals' perceptual performance when neuronal patterns are stimulated. We will use two-photon patterned stimulation to replay naturally-occurring population responses to determine if they have special meaning to the animal, perhaps because those patterns are determined by essential synaptic connections. By using patterned stimulation to vary the activity correlation between neurons, we will also test whether previously-observed pairwise correlations, which measure the relationship between the firing activities of two neurons, are an important part of the neuronal code. In achieving our goals we will produce a new technology for stimulating neurons in the brains of behaving animals with single-cell specificity that can be adapted to explore neuronal dynamics in a wide range of animal models and behaviors.

? DESCRIPTION (provided by applicant): Attention is essential for all daily tasks, and failure to control attention can have tragic consequences. An understanding of the neuronal mechanisms that guide attention will be needed for any comprehensive treatment of attention disorders, and is moreover likely to provide important new insights into the mechanisms of sensory processing and perception. The visual system is an ideal place to study attention, not only because optimal visual performance is critical for so many activities, but also because our relatively advanced understanding of the functional organization of the visual system makes it uniquely suited for addressing cutting-edge issues. The experiments proposed here will advance our understanding of attention by identifying distinct components that make up visual spatial attention and characterizing how neurons in different structures in the brain contribute to those components. New data from our lab show that the behavioral enhancement associated with spatial attention depends on two separable mechanisms, which are related to changes in behavioral sensitivity and changes in behavioral criterion. Moreover, some attention-related changes in neuronal activity in visual cerebral cortex are specifically related to changes in sensitivity and not to changes in criterion. These results reveal that attention encapsulates multiple processes that have been conflated in previous work. The proposed experiments will build on these results to characterizing distinct mechanisms that contribute to attention and the neuronal mechanisms that support them. In doing so, they will greatly advance our understanding of precisely what attention is and how it relates to specific sensory and cognitive representations in the brain. The first specific aims it to identify which brain structures suppor attention-related changes in behavioral criterion, which are not encoded in the region of visual cortex we have studied. We will record attention-related neuronal modulations in three carefully selected brain structures: the lateral intraparietal area in visual cortex, the prefrontal cortex, nd the superior colliculus in the brainstem. These recordings will distinguish the contributions of major stations to different components of visual spatial attention. The second specific aim is to explore whether attention can be further divided into specific components. Attention is typically discussed in terms of selecting one stimulus or location over another, but attention also has a component related to intensity (high or low levels of attention). The intensive aspect of attention has been little studied and its neuronal underpinnings are poorly understood. We will separately manipulate the selective and intensive components of attention while recording the responses of individual neurons in visual cortex. The results will establish whether different brain structures contribute differentially to the selective and intensive aspects of attention.

The Administrative Core will provide central coordination in support of this distributed program. It will monitor progress and promote exchange of results and technology by organizing monthly team video-conferences and annual face-to-face meetings. To provide oversight and critical feedback, it will organize regular meetings with the Internal Advisory Committee and External Advisory Board. With guidance from the Internal Advisory Committee and External Advisory Board, it will support inter-lab visits by team scientists to distribute technical approaches and promote consistency across experiments in different labs. In addition, it will implement and support a program web site to support team operations and disseminate information, software, and data to the broader neuroscience community.

Project Summary To survive, organisms must both accurately represent stimuli in the outside world, and use that representation to generate beneficial behavioral actions. Historically, these two processes ? the mapping from stimuli to neural responses, and the mapping from neural activity to behavior ? have largely been treated separately. Of the two, the former has received the most attention. Often referred to as the ?neural coding problem,? its goal is to determine which features of neural activity carry information about external stimuli. This approach has led to many empirical and theoretical proposals about the spatial and temporal features of neural population activity, or ?neural codes,? that represent sensory information. However, there is still no consensus about the neural code for most sensory stimuli in most areas of the nervous system. The lack of consensus arises in part because, while it is established that certain features of neural population responses carry information about specific stimuli, it is unclear whether the brain uses (?reads?) the information in these features to form sensory perceptions. We have developed a theoretical framework, based on the intersection of coding and readout, to approach this problem. Experimentally informing this framework requires manipulating patterns of neuronal activity based on, and at the same spatiotemporal scale as, their natural firing patterns during sensory perception. This work must be done in behaving animals because it is essential to know which neural codes guide behavioral decisions. In the first phase of this project (funded by the BRAIN Initiative), we developed the technology necessary for realizing this goal. In the present proposal, we will extend our patterned neuronal stimulation technology and apply it to answer long-standing questions about neural coding and readout in the visual, olfactory, and auditory systems. We will pioneer the capacity to determine which neurons within a network are encoding behaviorally relevant information, and also to determine the extent to which temporal patterns of those neurons? activity are being used to guide behavior. Finally, we will study these neural coding principles across changes in behavioral state and during learning to determine how internal context and past experience shape coding and readout. The contributions of the proposed work will be three-fold. First, we will provide the neuroscience community with the tools needed to test theories of how neural populations encode and decode information throughout the brain. Second, we will reveal fundamental principles of spatiotemporal neural coding and readout in the visual, olfactory, and auditory systems of behaving animals. And third, our unifying theoretical framework for cracking neural codes will allow the broader neuroscience community to resolve ongoing debates regarding neural coding that have been previously stalemated by considering only half of the coding/readout problem.

Project Summary Optimal allocation of attention is key to achieving peak behavioral performance. A detailed understanding of the neuronal mechanisms that control attention will be essential for any comprehensive strategies to reduce attention lapses or treat attention disorders. Moreover, better understanding of attention is likely to provide valuable new insights into sensory processing and perception. The visual system is an ideal subject for the study of attention. Peak visual performance is required for many human activities, and our relatively advanced understanding of the functional organization of the visual system makes it choice for deploying state-of-the-art techniques in well-described and well-differentiated brain regions. The experiments proposed here will advance our understanding of attention by providing a comprehensive characterization of the role of the locus coeruleus in visual attention. The locus coeruleus has long been associated with attention and arousal. However, the specificity of its contributions to attention has received relatively little investigation, despite many findings that suggest it has substantial functional specialization. Recent work in our lab has been directed at identifying distinct component of visual attention. We have shown that different brain structures make distinct contributions to changes in attention associated with behavioral sensitivity, selectivity, and perceptual criterion. Recently, we have used optogenetic stimulation of the locus coeruleus in monkeys and found that it can produce a robust and selective enhancement in behavioral sensitivity. This result emphasizes that the locus coeruleus is a potent factor in controlling visual attention, and highlights how little we know about the role it plays in controlling the various components that make up visual attention. We will optogenetically activate locus coeruleus neurons to measure their influence on visual sensitivity, selectivity and perceptual criterion, measuring both the magnitude and the dynamics of attention-related behavioral enhancements that it mediates. To reveal how it enhances visual performance, we will record visual responses from key visual structures ? area V4, the frontal eye fields and the superior colliculus ? during locus coeruleus stimulation to directly assess how it contributes to the quality of central representations of behaviorally- relevant visual stimuli. We will also record neuronal responses from the locus coeruleus itself while monkeys do tasks that modulate specific components of visual attention. The results from these recordings will provide a direct, detailed appraisal of the extent of specialization that exists within the locus coeruleus. The locus coeruleus plays a major role in behavioral performance, yet its role has been largely overlooked in efforts to understand the neurophysiology of visual attention. The proposed experiments will provide a precise, comprehensive assessment of the place of the locus coeruleus in attention and will substantially advance our understanding of the interaction of different brain structures in mediating visual behaviors.

Project Summary Here, we propose to thoroughly characterize the origins of pairwise correlations in cortex using a synergistic mix of experimental methodologies, behavior, and computation in mice and macaques. We will elucidate the mechanistic underpinnings of normalization and test our hypothesis that changes in cortical pairwise correlations and other signature arise from ongoing cortical computations. In Aim 1 we will record from populations of neurons in the middle temporal visual area of trained, behaving monkeys to test the hypothesis that pairwise spike correlations, gamma oscillation and transient responses at the onset of visual stimuli arise in part from the dynamics of the circuits that normalize neuronal responses. These tests require measurements with a precision that is not feasible in mice. Conversely, the experiments in Aim 2 and 3 address questions that are not feasible in monkeys. In Aim 2 we will exploit the accessibility of mouse visual cortex by using both two-photon laser scanning microscopy and multielectrode arrays to comprehensively measure the relationship between normalization and pairwise correlations in populations of V1 neurons and measure how spatial separation within cerebral cortex affects that relationship. Finally, in Aim 3 we will establish the contributions of specific cell classes to normalization and pairwise correlations in mouse V1. We record the activity of pyramidal neurons and the three most thoroughly characterized classes of cortical interneuron (VIP, SST and PV) during normalization. We will then separately manipulate the activity of these cells classes to revealing the role that changes to the ratio of excitation and inhibition play in driving normalization. In this way, we will establish the role these neurons play in changing pairwise correlations within the excitatory pool of neurons. Results from all three Aims will be tied together using a new family of dynamic, recurrent circuit models of normalization to formalize the hypothesis that normalization imposes pairwise correlations and other activity signatures, and will use experimental data to constrain and refine these models.