Hippocampal Circuits for Multiple Gamma Oscillations

In order to support adaptive behaviour the nervous system integrates information from different external and internal sources. This puts hierarchically organized neural sub-networks at the challenge of generating novel representations by combining the activity patterns from converging input pathways. One mechanism by which receiving networks can accomplish this task in a dynamically regulated manner is by synchronised fluctuations in the responsiveness of individual neurons; neuronal oscillations. Specifically, in the CA1 area of the dorsal hippocampus distinct mid- frequency, slow, and fast gamma oscillations modulate the spike timing of pyramidal cells encoding spatial information, and are suggested to organize information processing, by regulating the impact of convergent intra- and extra-hippocampal inputs on pyramidal cell activity. In this project we will address the cellular and network mechanisms of gamma oscillations in the CA1 area of the mouse dorsal hippocampus. We will test contributions from the two main cortical inputs to CA1 area (pathways originating the medial and the lateral entorhinal cortices) by their chemogenetic inhibition. In separate experiments, we will record spikes of CA1 axo-axonic cells identified using the juxtacellular method, and analyse their entrainment by gamma oscillations, and their influence on multiple CA1 pyramidal cells, recorded by nearby electrodes. Using this complex approach we will determine the role of axo-axonic cells in conveying gamma oscillations onto the output networks of CA1 pyramidal cells. To reveal how oscillations relate to neuronal communication, we will record distinct gamma oscillations, during two behavioural tasks that require differential engagement of hippocampal input networks processing idiothetic and allothetic information, expecting a behaviour-depend reorganization of gamma oscillations in the hippocampal CA1 area. In this project we will combine recording of distinct gamma oscillations with juxtacellular recording of identified cells, selective chemogenetic inhibition of defined input networks and different behavioural paradigms in order to understand the neuronal circuitry that supports input- selective information processing by multiple gamma oscillations in the dorsal CA1 area of the hippocampus. Potentially revealing key network motifs of hippocampal information processing, the results of this project may offer novel concepts for understand hippocampal function in health and disease.

Communication between and within neuronal networks occurs through excitatory and inhibitory synaptic contacts. Fiber tracts often displaying considerable convergence mediate long range connections between brain areas, and together they form an intricate connectivity matrix called the connectome. Convergence within the connectome enables combinatorial processing, by allowing neuronal circuits to integrate and synthesize information from multiple distinct sources. However, to achieve adequate operation and to avoid unwanted interference, such convergent networks also require adaptive and dynamic regulation of interacting input channels and local circuits. In the CA1 area of the hippocampus two pathways carrying recalled memory traces (from the CA3 area) and information on current sensory experiences (from the entorhinal cortex) converge on single pyramidal cells. These inputs and local circuit operations of the CA1 area are carried by three distinct gamma frequencies (slow, mid and fast respectively), that together with the slower theta oscillation provide the temporal framework for coordinating the integration and segregation of information from different distal and local sources. In this project we set out to understand how local inhibitory cells in the CA1 area contribute to the balancing of information flow from different sources. We found a small subpopulation of CA1 GABAergic cells with their spike timing following the temporal dynamics of the cortical input pathway (mid gamma oscillation) with remarkable selectivity and high fidelity, suggesting their involvement in coordinating extra-hippocampal information inflow. By labelling a subset of this population with neurobiotin using the juxtacellular method for post hoc histological analysis, we found that their input and dense output processes (dendritic and axonal arborisations) were precisely aligned to the termination zone of cortical fibres in the distal dendritic layer of the CA1 area (stratum lacunosum-moleculare). This, together with their protein expression pattern identified these cells as neurogliaform cells, a cell type producing prolonged inhibition in dendritic processes contained within their axonal arbour, with little selectivity for cell types. Overall the spike timing of pyramidal cells and other GABAergic cells of the CA1 area was weakly modulated by mid gamma oscillations suggesting limited but detectable contribution of extra-hippocampal inputs to the excitatory synaptic drive of the CA1 network. Conspicuously, this tuning became undetectable following neurogliaform cell discharge, suggesting a prompt disengagement of the CA1 network from extra-hippocampal excitation. Our discovery identifies the inhibitory neurogliaform cells as key players in coordinating co-active afferent information streams to neuronal circuits. Unlike other GABAergic cells which instruct synchrony in neuronal circuits, neurogliaform cells selectively and dynamically detune neurons from a subset of their afferents, allowing prompt shifts in sources of excitation. Therefore, neurogliaform cells may represent a previously unidentified target in conditions that involve erroneous coordination of information transfer between neuronal circuits, such as schizophrenia or autism spectrum disorder.