Explore the vast range of titles available.

In the hippocampus, each memory trace is encoded by a specific subset of pyramidal cells. The other pyramidal cells must be actively excluded from the memory encoding process by inhibition, which is done by selective dendrite-targeting interneurons. Szőnyi et al. found that γ-aminobutyric acid–releasing (GABAergic) cells located in a small region in the brain stem called the nucleus incertus project to the hippocampus. The nucleus incertus again is innervated by several regions that respond to salient stimuli. Its GABAergic cells preferentially inhibit the dendrite-targeting interneurons in the hippocampus. The nucleus incertus is thus a central mediator between brain regions that are highly responsive to salient stimuli and the hippocampal circuitry involved in memory formation. Science , this issue p. eaaw0445 A brainstem regulatory mechanism for the selection of hippocampal neuronal assemblies during contextual learning is described. Hippocampal pyramidal cells encode memory engrams, which guide adaptive behavior. Selection of engram-forming cells is regulated by somatostatin-positive dendrite-targeting interneurons, which inhibit pyramidal cells that are not required for memory formation. Here, we found that γ-aminobutyric acid (GABA)–releasing neurons of the mouse nucleus incertus (NI) selectively inhibit somatostatin-positive interneurons in the hippocampus, both monosynaptically and indirectly through the inhibition of their subcortical excitatory inputs. We demonstrated that NI GABAergic neurons receive monosynaptic inputs from brain areas processing important environmental information, and their hippocampal projections are strongly activated by salient environmental inputs in vivo. Optogenetic manipulations of NI GABAergic neurons can shift hippocampal network state and bidirectionally modify the strength of contextual fear memory formation. Our results indicate that brainstem NI GABAergic cells are essential for controlling contextual memories.

Spatial memories that can last a lifetime are thought to be encoded during ‘online’ periods of exploration and subsequently consolidated into stable cognitive maps through their ‘offline’ reactivation. However, the mechanisms and computational principles by which offline reactivation stabilize long-lasting spatial representations remain poorly understood. Here, we employed simultaneous fast calcium imaging and electrophysiology to track hippocampal place cells over 2 weeks of online spatial reward learning behavior and offline resting. We describe that recruitment to persistent network-level offline reactivation of spatial experiences in mice predicts the future representational stability of place cells days in advance of their online reinstatement. Moreover, while representations of reward-adjacent locations are generally more stable across days, offline-reactivation-related stability is, conversely, most prominent for reward-distal locations. Thus, while occurring on the tens of milliseconds timescale, offline reactivation is uniquely associated with the stability of multiday representations that counterbalance the overall reward-adjacency bias, thereby predicting the stabilization of cognitive maps that comprehensively reflect entire underlying spatial contexts. These findings suggest that post-learning offline-related memory consolidation plays a complimentary and computationally distinct role in learning compared to online encoding. Grosmark et al. use simultaneous calcium imaging and electrophysiology to track the formation and long-term evolution of hippocampal memory traces in mice and uncover a role for post-learning reactivation in the formation of spatially uniform cognitive maps.

First-Order, Reduced and Controlled Error (FORCE) learning and its variants are widely used to train chaotic recurrent neural networks (RNNs), and outperform gradient methods on certain tasks. However, there is currently no standard software framework for FORCE learning. We present tension , an object-oriented, open-source Python package that implements a TensorFlow / Keras API for FORCE. We show how rate networks, spiking networks, and networks constrained by biological data can all be trained using a shared, easily extensible high-level API. With the same resources, our implementation outperforms a conventional RNN in loss and published FORCE implementations in runtime. Our work here makes FORCE training chaotic RNNs accessible and simple to iterate, and facilitates modeling of how behaviors of interest emerge from neural dynamics.

Local circuit architecture facilitates the emergence of feature selectivity in the cerebral cortex 1 . In the hippocampus, it remains unknown whether local computations supported by specific connectivity motifs 2 regulate the spatial receptive fields of pyramidal cells 3 . Here we developed an in vivo electroporation method for monosynaptic retrograde tracing 4 and optogenetics manipulation at single-cell resolution to interrogate the dynamic interaction of place cells with their microcircuitry during navigation. We found a local circuit mechanism in CA1 whereby the spatial tuning of an individual place cell can propagate to a functionally recurrent subnetwork 5 to which it belongs. The emergence of place fields in individual neurons led to the development of inverse selectivity in a subset of their presynaptic interneurons, and recruited functionally coupled place cells at that location. Thus, the spatial selectivity of single CA1 neurons is amplified through local circuit plasticity to enable effective multi-neuronal representations that can flexibly scale environmental features locally without degrading the feedforward input structure. Single-cell tracing and optogenetics manipulation in mice are used to show how spatial tuning of individual pyramidal cells in CA1 can propagate to and be amplified by their local subnetwork of neurons.

The authors conduct simultaneous recording and optogenetic silencing of PV or SOM interneurons in the CA1 region of the hippocampus in head-fixed mice actively moving a treadmill belt. They report that these interneurons have distinct roles in controlling the rate, burst and timing of hippocampal pyramidal cells. A consortium of inhibitory neurons control the firing patterns of pyramidal cells, but their specific roles in the behaving animal are largely unknown. We performed simultaneous physiological recordings and optogenetic silencing of either perisomatic (parvalbumin (PV) expressing) or dendrite-targeting (somatostatin (SOM) expressing) interneurons in hippocampal area CA1 of head-fixed mice actively moving a treadmill belt rich with visual-tactile stimuli. Silencing of either PV or SOM interneurons increased the firing rates of pyramidal cells selectively in their place fields, with PV and SOM interneurons having their largest effect during the rising and decaying parts of the place field, respectively. SOM interneuron silencing powerfully increased burst firing without altering the theta phase of spikes. In contrast, PV interneuron silencing had no effect on burst firing, but instead shifted the spikes' theta phase toward the trough of theta. These findings indicate that perisomatic and dendritic inhibition have distinct roles in controlling the rate, burst and timing of hippocampal pyramidal cells.

In addition to providing well-characterized excitatory inputs, the entorhinal cortex also sends long-range inhibitory projections to the hippocampus. Basu et al. described this input in detail and characterized its role for learning and memory. Multimodal sensory stimuli activate long-range inhibitory input in vivo. This input enables precisely timed information transfer within the cortico-hippocampal circuit. In this way, long-range inhibitory projections play an important role in providing specificity of fear conditioning, and thus help prevent overgeneralization. Science , this issue p. 10.1126/science.aaa5694 Inhibitory inputs from the lateral entorhinal cortex help to make contextual memory associations specific. The cortico-hippocampal circuit is critical for storage of associational memories. Most studies have focused on the role in memory storage of the excitatory projections from entorhinal cortex to hippocampus. However, entorhinal cortex also sends inhibitory projections, whose role in memory storage and cortico-hippocampal activity remains largely unexplored. We found that these long-range inhibitory projections enhance the specificity of contextual and object memory encoding. At the circuit level, these γ-aminobutyric acid (GABA)–releasing projections target hippocampal inhibitory neurons and thus act as a disinhibitory gate that transiently promotes the excitation of hippocampal CA1 pyramidal neurons by suppressing feedforward inhibition. This enhances the ability of CA1 pyramidal neurons to fire synaptically evoked dendritic spikes and to generate a temporally precise form of heterosynaptic plasticity. Long-range inhibition from entorhinal cortex may thus increase the precision of hippocampal-based long-term memory associations by assessing the salience of mnemonic information to the immediate sensory input.

Fear memories guide adaptive behavior in contexts associated with aversive events. The hippocampus forms a neural representation of the context that predicts aversive events. Representations of context incorporate multisensory features of the environment, but must somehow exclude sensory features of the aversive event itself. We investigated this selectivity using cell type–specific imaging and inactivation in hippocampal area CA1 of behaving mice. Aversive stimuli activated CA1 dendrite-targeting interneurons via cholinergic input, leading to inhibition of pyramidal cell distal dendrites receiving aversive sensory excitation from the entorhinal cortex. Inactivating dendrite-targeting interneurons during aversive stimuli increased CA1 pyramidal cell population responses and prevented fear learning. We propose subcortical activation of dendritic inhibition as a mechanism for exclusion of aversive stimuli from hippocampal contextual representations during fear learning.

The authors use optical activation and cell type–specific pharmacogenetic silencing in vitro to show that dendritic inhibition critically regulates input-output transformations in mouse hippocampal CA1 pyramidal cells. Dendrite-targeting interneurons are themselves modulated by interneurons targeting pyramidal cell somata. Transforming synaptic input into action potential output is a fundamental function of neurons. The pattern of action potential output from principal cells of the mammalian hippocampus encodes spatial and nonspatial information, but the cellular and circuit mechanisms by which neurons transform their synaptic input into a given output are unknown. Using a combination of optical activation and cell type–specific pharmacogenetic silencing in vitro , we found that dendritic inhibition is the primary regulator of input-output transformations in mouse hippocampal CA1 pyramidal cells, and acts by gating the dendritic electrogenesis driving burst spiking. Dendrite-targeting interneurons are themselves modulated by interneurons targeting pyramidal cell somata, providing a synaptic substrate for tuning pyramidal cell output through interactions in the local inhibitory network. These results provide evidence for a division of labor in cortical circuits, where distinct computational functions are implemented by subtypes of local inhibitory neurons.

Neuronal mitochondria play important roles beyond ATP generation, including Ca 2+ uptake, and therefore have instructive roles in synaptic function and neuronal response properties. Mitochondrial morphology differs significantly between the axon and dendrites of a given neuronal subtype, but in CA1 pyramidal neurons (PNs) of the hippocampus, mitochondria within the dendritic arbor also display a remarkable degree of subcellular, layer-specific compartmentalization. In the dendrites of these neurons, mitochondria morphology ranges from highly fused and elongated in the apical tuft, to more fragmented in the apical oblique and basal dendritic compartments, and thus occupy a smaller fraction of dendritic volume than in the apical tuft. However, the molecular mechanisms underlying this striking degree of subcellular compartmentalization of mitochondria morphology are unknown, precluding the assessment of its impact on neuronal function. Here, we demonstrate that this compartment-specific morphology of dendritic mitochondria requires activity-dependent, Ca 2+ and Camkk2-dependent activation of AMPK and its ability to phosphorylate two direct effectors: the pro-fission Drp1 receptor Mff and the recently identified anti-fusion, Opa1-inhibiting protein, Mtfr1l. Our study uncovers a signaling pathway underlying the subcellular compartmentalization of mitochondrial morphology in dendrites of neurons in vivo through spatially precise and activity-dependent regulation of mitochondria fission/fusion balance. The mechanisms regulating mitochondrial architecture in neurons remain unclear. The authors report that in dendrites, mitochondria structure is specified by the CAMKK2-AMPK pathway through compartment-specific and activity-dependent levels of fission.

The authors use two-photon Ca 2+ imaging of axonal boutons in hippocampal CA1 of behaving mice to monitor the activation of septo-hippocampal GABAergic boutons. They report that some sensory inputs are more effective than locomotion in driving firing by these long-range GABAergic projections. Hippocampal interneurons receive GABAergic input from the medial septum. Using two-photon Ca 2+ imaging of axonal boutons in hippocampal CA1 of behaving mice, we found that populations of septo-hippocampal GABAergic boutons were activated during locomotion and salient sensory events; sensory responses scaled with stimulus intensity and were abolished by anesthesia. We found similar activity patterns among boutons with common putative postsynaptic targets, with low-dimensional bouton population dynamics being driven primarily by presynaptic spiking.