S Murray Sherman - US grants

This research is directed at providing an ultrastructural description of physiologically-defined circuitry in the lateral geniculate nucleus of cats. Emphasis is placed on a description of retinogeniculate circuitry and differences among the W-, X-, and Y-cell pathways. Such information is not only fundamental to an understanding of the neural basis of vision but also can serve as a framework against which to compare neural correlates of certain forms of amblyopia. Structure and function are related at the single cell level in the following manner. Physiological recordings of a single axon or cell are accomplished intracellularly with a fine micropipette filled with horseradish peroxidase (HRP), and HRP is then iontophoresed into the physiologically defined neuronal element. The HRP completely fills a soma and its dendrites or an axon and its terminal boutons, and the HRP reaction product is used as an electrondense marker for subsequent ultrastructural analysis. Electron microscopic reconstructions from such material permit a fairly complete description of synaptic circuits related to single, physiologically identified cells. Plans are described to analyze both the geniculate terminals of various afferents (W-, X-, or Y-cell optic tract, perigeniculate, corticogeniculate) as well as the synapses formed onto geniculate W-, X-, and Y-cells.

It is now clear that the lateral geniculate nucleus (LGN) acts as a variable gateway or filter for the retina-to-cortex relay: when open, all information gets through; when closed, none does; and when partly open, some is relayed. The LGN, by thus controlling the flow of retinal information to cortex, represents a major neural substrate for many forms of visual attention. Our broad goal is to understand, at the cellular level, how this gating is controlled. We proposed to do so by complementary in vivo and in vitro intracellular recording of single neurons in the cat's LGN, the former performed in anesthetized animals and the latter from LGN slices. Two factors are key to this control of retinogeniculate gating. First, the LGN relay cells possess a number of voltage- and ligand-gated membrane conductances in addition to the conventional action potential, and the mix of these active at any time determines the gain of retinogeniculate transmission. A particularly important conductance is the low threshold Ca2+ spike, which is voltage dependent and can be self-regenerating; we shall test the hypothesis that, when active, it prevents normal retinogeniculate transmission. Other more subtle membrane properties will also be studied. Second, nonretinal inputs, which dominate synaptic input to LGN relay cells, act to control these conductances. Sources of these nonretinal inputs include: local, GABAergic, inhibitory neurons; ascending inputs from the brainstem (mostly midbrain), and descending inputs from visual cortex. We shall study the above mentioned membrane conductances, their control by various nonretinal synaptic transmission, and how they affect receptive field properties. In vivo studies of stimulation of differing membrane voltage (manipulated by current injection through the recording electrode) and activation of brainstem afferents. We shall also attend to any differences between X and Y cells, which are the LGN representatives of the two main parallel pathways from retina to cortex. In vitro studies will include analyses of the voltage- and time-dependents of the low threshold an assessment of putative neurotransmitters, their agonists and antagonists, and the postsynaptic receptor types. We shall also intracellularly label most cells in vitro to determine structure/function relationships, including any differences between interneurons and relay cells. Finally, we shall attempt simultaneous recording from two connected neurons to determine the synaptic physiology and pharmacology of these identified circuits.

It is now clear that the lateral geniculate nucleus (LGN) acts as a variable gateway or filter for the retina-to-cortex relay: when open, all information gets through; when closed, none does; and when partly open, some is relayed. The LGN, by thus controlling the flow of retinal information to cortex, represents a major neural substrate for many forms of visual attention. Our broad goal is to understand, at the cellular level, how this gating is controlled. We proposed to do so by complementary in vivo and in vitro intracellular recording of single neurons in the cat's LGN, the former performed in anesthetized animals and the latter from LGN slices. Two factors are key to this control of retinogeniculate gating. First, the LGN relay cells possess a number of voltage- and ligand-gated membrane conductances in addition to the conventional action potential, and the mix of these active at any time determines the gain of retinogeniculate transmission. A particularly important conductance is the low threshold Ca2+ spike, which is voltage dependent and can be self-regenerating; we shall test the hypothesis that, when active, it prevents normal retinogeniculate transmission. Other more subtle membrane properties will also be studied. Second, nonretinal inputs, which dominate synaptic input to LGN relay cells, act to control these conductances. Sources of these nonretinal inputs include: local, GABAergic, inhibitory neurons; ascending inputs from the brainstem (mostly midbrain), and descending inputs from visual cortex. We shall study the above mentioned membrane conductances, their control by various nonretinal synaptic transmission, and how they affect receptive field properties. In vivo studies of stimulation of differing membrane voltage (manipulated by current injection through the recording electrode) and activation of brainstem afferents. We shall also attend to any differences between X and Y cells, which are the LGN representatives of the two main parallel pathways from retina to cortex. In vitro studies will include analyses of the voltage- and time-dependents of the low threshold an assessment of putative neurotransmitters, their agonists and antagonists, and the postsynaptic receptor types. We shall also intracellularly label most cells in vitro to determine structure/function relationships, including any differences between interneurons and relay cells. Finally, we shall attempt simultaneous recording from two connected neurons to determine the synaptic physiology and pharmacology of these identified circuits.

DESCRIPTION (from abstract): The lateral geniculate nucleus of the thalamus is the gateway to the striate cortex, representing the first opportunity for the visual system to regulate its own input dynamically. LGN relay cells possess a rich and complex repertoire of voltage and ligand gated conductances that are critical to this regulation. These are under the control of a variety of nonretinal inputs, chiefly a glutamatergic input from visual cortex and a cholinergic one from the parabrachial region (PBR) of the brainstem. One of the key conductances is a voltage dependent Ca current known as It and this determines response mode: burst mode during It activation and tonic mode during It inactivation. We found that these modes have profound consequences to visual information processing, and that they are controlled by the PBR input. We now focus our efforts on the corticogeniculate pathway, the largest input to LGN, but poorly understood and still an enigma. Although it provides more synaptic contacts to relay neurons than any other extrageniculate source, little is known of its effect on visual signal transmission or on the intrinsic membrane properties of LGN neurons. Prior examination of this pathway has yielded only limited insight into its role in retinogeniculate transmission: we propose a fresh attack on this vital problem. We shall not rely on a single approach, but instead we shall employ a multifaceted, parallel strategy that will examine the corticogeniculate pathway from the cellular level to intact circuitry. In vitro, we shall use intracellular recording in LGN slices to examine circuitry mediating corticogeniculate transmission, its underlying pharmacology, and its effects on intrinsic membrane properties of LGN cells. In many brain areas, different neurotransmitter systems may converge to modulate the same set of conductances, and we suspect the same to be true of the LGN. We shall supplement our brain slice experiments with whole cell patch recordings in dissociated LGN cells in order to examine the intracellular factors mediating LGN responses to corticogeniculate inputs, and the possible neurotransmitter convergence between the corticogeniculate and PBR pathways. In vitro, we shall pharmacologically activate and inactivate the corticogeniculate pathway in order to determine the spatial structure of the corticogeniculate influence and its effect on visual signal transmission. As a bonus to these studies, we shall also attempt to obtain cross-correlograms between single cortical and LGN cell pairs (among other pairs to be studied) to determine the effect of the corticogeniculate input at the single cell level. Our physiological inquiries will be complemented by anatomical studies of single corticogeniculate axons in order to determine the spatial extent of the cortical innervation of LGN, and to determine whether there are multiple subtypes of corticogeniculate projection cell. Furthermore, because the thalamic reticular nucleus is an essential link in the corticogeniculate input (this nucleus receives collateral input from corticogeniculate axons and projects to LGN, we shall study it as well. Our combined approaches will yield a comprehensive characterization of the corticogeniculate pathway and its role in visual signal transmission.

DESCRIPTION (provided by applicant): Our research continues to be directed at an understanding of thalamic relays, with emphasis on the visual thalamus (LGN & LP-Pul), using the in vitro slice preparation in mice and cats. Techniques include the use of simple thalamic slices to study LGN and LP-Pul in cats and the use of thalamocortical slices to study VPM and POre (somatosensory thalamus) in mice. We shall also use photostimulation with caged glutamate, partly in order to identify presynaptic units in a search for synaptically coupled pairs for corticothalamic, TRN to relay cell, and relay cell to TRN synapses. A general hypothesis is that LGN and VPM are first order relays, being the first relay of peripheral information (e.g., retinal or lemniscal) to cortex, whereas LP-Pul and POm are higher order relays, relaying information between cortical areas. In particular, we have defined 3 Aims: In Aim 1, we shall study corticothalamic inputs and how they differ between LGN (or VPM) and LP-Pul (or POm). The hypothesis is that LGN (or VPM), being a first order relay, receives only layer 6 input from cortex, and that this is modulatory, whereas LP-Pul (or POm), being mostly a higher order relay, receives cortical input from layer 5 and 6. We shall determine if the layer 6 input to LP-Pul (or POm), like that to LGN (or VPM), is modulatory, whereas the layer 5 input is driver, functioning like the retinal input to LGN. This would implicate the LP-Pul (and POm) as playing a heretofore ignored, key role in corticocortical communication and would challenge the conventional hypothesis of how visual and somatosensory cortical areas are functionally connected. Finally, we will study the efficacy of layer 6 corticothalamic inputs in controlling an important voltage gated current in relay cells, known as Iv. Aim 2 will broadly test the function of the TRN in modulating thalamic relays, and in particular will test details of TRN circuitry, challenging the conventional views that relay cell to TRN connections represent feedback inhibition and that layer 6 cortical input to TRN connections represent feedforward inhibition. We will also test hypotheses regarding the different synaptic properties of relay cell and cortical layer 6 inputs to TRN cells. As above, we will study the efficacy of TRN inputs in controlling It. Finally, Aim 3 will test the hypothesis that different cell classes can be recognized in LP-Pul on the same basis that distinguishes X and Y cells in LGN.

DESCRIPTION (provided by applicant): The broad goal of the proposal is to understand the functional organization of the early thalamocortical stages of processing in the auditory system. This involves a better functional understanding of the various pathways connecting the medial geniculate body (MGB) plus first and second auditory cortices (A1 and A2). We will use an in vitro slice preparation of the mouse brain in which different slice configurations will have the MGB to A1/A2, A1 to MGB, and A1 to A2 pathways intact. We will use uncaging of glutamate by photostimulation or minimal electrical stimulation to test the function of each of these glutamatergic pathways, and this will be complemented by the use of channelrhodopsin methodology. In particular, we shall attempt to classify each pathway as Class 1 (i.e., formerly driver), Class 2 (formerly modulator), or other, and thereby we hope to construct a functional hierarchy of information flow. We shall also test the role of layer 6 corticothalamic cells in gating thalamocortical transmission and test the complex hypothesis, first, that there exists parallel direct and corticothalamocortical pathways linking A1 with A2, and second, that these dual pathways operate in a coincidence detection manner to control information flow. This could also provide more general insights regarding cortical functioning, particularly with respect to new evidence that different cortical areas can dynamically cooperate depending on behavioral needs. PUBLIC HEALTH RELEVANCE: We must better understand auditory information flow through the first few stages of cortical processing to begin to understand how pathology in these pathways leads to hearing loss including defects in cognitive auditory functions, such as lexical-semantic processing, phonological information extraction, selective attention and object recognition.

PROJECT SUMMARY The proposed research will test related hypotheses that thalamus plays a heretofore neglected and critical role in cortical processing. In particular, many thalamic nuclei, that together comprise the majority of thalamic volume and that were previously mysterious in function, we now suggest are critically involved in information flow and functional, dynamic binding between cortical areas via cortico-thalamo-cortical pathways. We propose to study these pathways using the mouse visual system as the model involving in vitro slice preparations and in vivo behaving preparations. It appears that, in many and perhaps all cases, cortical areas are connected by both direct and these transthalamic pathways, and we wish to understand why: What is different in the information passed by each pathway? Why does one pass through thalamus with the possibility of being blocked there? Is there nonlinear summation in the target cortical area when both pathways are active, and could this be involved in dynamic linking of areas to subserve various cognitive tasks, such as attention? To begin developing answers to these questions, we propose to probe basic circuit properties of these pathways to better understand the role of the transthalamic pathways in higher cognitive functioning.

? DESCRIPTION (provided by applicant): The Gordon Research Conference (GRC) on Thalamocortical Interactions will involve roughly 150 scientists of diverse expertise assembled to discuss and debate new findings related to the functioning of thalamus and cortex and their interactions. This will take place at the Ventura Beach Marriott in Ventura, CA during the period of 14-19 February 2016. The subtitle of this GRC is Cell and Circuit Properties. The meeting itself will involve a series of 20 minute talks followed by 10 minute discussion periods plus time each day for less formal interactions at poster sessions. This GRC will offer a unique combination of features, including: breadth of research; cutting-edge emphasis; mingling of investigators from all ranks and diverse sub-fields and locales; and intimate size and extended discussion time, allowing for close and sustained interactions. The program of this GRC emphasizes cell and circuit properties of thalamocortical interactions, and builds on these to explore their roles in cognition and neuronal disease states. This GRC represents a rare opportunity for neuroscientists interested in thalamocortical interactions to exchange new results, hypotheses, and ideas at many levels, from cellular through systems to cognitive and clinical. This is especially timely, because this field has been expanding significantly with the recent appreciation that thalamus plays an ongoing and critical role in cortical functioning, and newly described deficits in such disorders as schizophrenia, epilepsy, and autism appear to have at least partly a thalamocortical locus. We thus fully anticipate this GRC to significantly advance the field and potentially offer new insights into certain clinical conditions.

PROJECT SUMMARY The proposed research will test related hypotheses that thalamus plays a heretofore neglected and critical role in cortical processing. In particular, many thalamic nuclei, that together comprise the majority of thalamic volume and that were previously mysterious in function, we now suggest are critically involved in information flow and functional, dynamic binding between cortical areas via cortico-thalamo-cortical pathways. We propose to study these pathways using the mouse somatosensory system as the model involving both in vitro slice and in vivo whole animal preparations. It appears that, in many and perhaps all cases, cortical areas are connected by both direct and these transthalamic pathways, and we wish to understand why: What is different in the information passed by each pathway? Why does one pass through thalamus with the possibility of being blocked there? Is there nonlinear summation in the target cortical area when both pathways are active, and could this be involved in dynamic linking of areas to subserve various cognitive tasks, such as attention? To begin developing answers to these questions, we propose to probe basic circuit properties of these pathways to better understand the role of the transthalamic pathways in higher cognitive functioning.

PROJECT SUMMARY The proposed research will use state of the art electron microscopic (EM) connectomics combined with specific pathway tracing to analyze the convergent inputs involving cortical and thalamic circuitry in mice. Labeling will be via different types of engineered ascorbate peroxidase (APX) that will tag either cytoplasm for some injections or mitochondria for others. Volumes of cortex and convergent inputs to be analyzed are: thalamic VPm & POm input to S1; thalamic LGN & Pul input to V1; S1 & POm input to M1; thalamic POm & VA/VL inputs to M1; and layer 5 inputs from different cortical areas to a thalamic nucleus (those to be initially analyzed are V1 & S1 convergent input to LD and V1 & V2, to Pul). We shall also analyze in each of the cortical areas under connectomic analysis inputs from layer 4 to layers 2/3 and local lateral connections within layers 2/3. The detailed connectomics analysis will allow us to determine the different connection patterns of the various labeled convergent afferents, including the extent to which they converge onto single cells. Regarding the analysis of cortical layer 5 convergent inputs to a thalamic nucleus, we expect this to address an open question of great import regarding thalamic functioning. Limited studies so far suggest that there is little or no functional convergence of different information streams onto individual thalamic relay cells. Thus retinal input to LGN cells involves little or no convergence, and even when there is little convergence (usually ?3 retinal axons), the inputs are of the same type (e.g., X or Y retinal axons) with virtually no mixing of types. This suggests a simple relay function for thalamus. However, these analyses have been mostly limited to first order thalamic nuclei (i.e., those receiving driving input from subcortical sources, like the retina). Recent study of higher order nuclei (i.e., those receiving driving input from cortical layer 5) suggests that these might have convergent input from different cortical areas. We will test this, and if we find such convergent input onto single relay cells, this will transform our understanding of thalamic function, because it will demonstrate that, at least for some higher order nuclei, significant alteration in the nature of incoming information occurs before relay to cortex. In addition to these analyses allowing a better understanding of cortical and thalamic circuitry and their interactions, most of the pathways to be studied have independently been analyzed physiologically for characteristics of driver vs modulator properties. We thus expect to find significant correlations between these physiological data and the proposed connectomics analyses, which we believe will allow future connectomic studies to be interpreted on a functional basis heretofore unavailable.