Hippocampal ripples are involved in memory consolidation, but the mechanisms underlying their generation remain unclear. Models relying on interneuron networks in the CA1 region disagree on the predominant source of excitation to interneurons: either "direct," via the Schaffer collaterals that provide feedforward input from CA3 to CA1, or "indirect," via the local pyramidal cells in CA1, which are embedded in a recurrent excitatory-inhibitory network. Here, we used physiologically constrained computational models of basket-cell networks to investigate how they respond to different conditions of transient, noisy excitation. We found that direct excitation of interneurons could evoke ripples (140-220 Hz) that exhibited intraripple frequency accommodation and were frequency-insensitive to GABA modulators, as previously shown in experiments. In addition, the indirect excitation of the basket-cell network enabled the expression of intraripple frequency accommodation in the fast-gamma range (90-140 Hz), as In our model, intraripple frequency accommodation results from a hysteresis phenomenon in which the frequency responds differentially to the rising and descending phases of the transient excitation. Such a phenomenon predicts a maximum oscillation frequency occurring several milliseconds before the peak of excitation. We confirmed this prediction for ripples in brain slices from male mice. These results suggest that ripple and fast-gamma episodes are produced by the same interneuron network that is recruited via different excitatory input pathways, which could be supported by the previously reported intralaminar connectivity bias between basket cells and functionally distinct subpopulations of pyramidal cells in CA1. Together, our findings unify competing inhibition-first models of rhythm generation in the hippocampus. The hippocampus is a part of the brain of humans and other mammals that is critical for the acquisition and consolidation of memories. During deep sleep and resting periods, the hippocampus generates high-frequency (∼200 Hz) oscillations called ripples, which are important for memory consolidation. The mechanisms underlying ripple generation are not well understood. A prominent hypothesis holds that the ripples are generated by local recurrent networks of inhibitory neurons. Using computational models and experiments in brain slices from rodents, we show that the dynamics of interneuron networks clarify several previously unexplained characteristics of ripple oscillations, which advances our understanding of hippocampus-dependent memory consolidation.

The entorhinal cortices in the temporal lobe of the brain are key structures relaying memory related information between the neocortex and the hippocampus. The medial entorhinal cortex (MEC) routes spatial information, whereas the lateral entorhinal cortex (LEC) routes predominantly olfactory information to the hippocampus. Gamma oscillations are known to coordinate information transfer between brain regions by precisely timing population activity of neuronal ensembles. Here, we studied the organization of in vitro gamma oscillations in the MEC and LEC of the transgenic (tg) amyloid precursor protein (APP)-presenilin 1 (PS1) mouse model of Alzheimer’s Disease (AD) at 4–5 months of age. In vitro gamma oscillations using the kainate model peaked between 30–50 Hz and therefore we analyzed the oscillatory properties in the 20–60 Hz range. Our results indicate that the LEC shows clear alterations in frequency and power of gamma oscillations at an early stage of AD as compared to the MEC. The gamma-frequency oscillation slows down in the LEC and also the gamma power in dorsal LEC is decreased as early as 4–5 months in the tg APP-PS1 mice. The results of this study suggest that the timing of olfactory inputs from LEC to the hippocampus might be affected at an early stage of AD, resulting in a possible erroneous integration of the information carried by the two input pathways to the hippocampal subfields.

Sharp wave-associated ∼200-Hz ripple oscillations in the hippocampus have been implicated in the consolidation of memories. However, knowledge on mechanisms underlying ripples is still scarce, in particular with respect to synaptic involvement of specific cell types. Here, we used cell-attached and whole-cell recordings in vitro to study activity of pyramidal cells and oriens-lacunosum-moleculare (O-LM) interneurons during ripples. O-LM cells received ripple-associated synaptic input that arrived delayed (3.3 ± 0.3 ms) with respect to the maximum amplitude of field ripples and was locked to the ascending phase of field oscillations (mean phase: 209 ± 6°). In line, O-LM cells episodically discharged late during ripples (∼6.5 ms after the ripple maximum), and firing was phase-locked to field oscillations (mean phase: 219 ± 9°). Our data unveil recruitment of O-LM neurons during ripples, suggesting a previously uncharacterized role of this cell type during sharp wave-associated activity.

No abstract

No abstract

The context-dependence of extinction learning has been well studied and requires the hippocampus. However, the underlying neural mechanisms are still poorly understood. Using memory-driven reinforcement learning and deep neural networks, we developed a model that learns to navigate autonomously in biologically realistic virtual reality environments based on raw camera inputs alone. Neither is context represented explicitly in our model, nor is context change signaled. We find that memory-intact agents learn distinct context representations, and develop ABA renewal, whereas memory-impaired agents do not. These findings reproduce the behavior of control and hippocampal animals, respectively. We therefore propose that the role of the hippocampus in the context-dependence of extinction learning might stem from its function in episodic-like memory and not in context-representation per se. We conclude that context-dependence can emerge from raw visual inputs.

The context-dependence of extinction learning has been well studied and requires the hippocampus. However, the underlying neural mechanisms are still poorly understood. Using memory-driven reinforcement learning and deep neural networks, we developed a model that learns to navigate autonomously in biologically realistic VR environments based on raw camera inputs alone. Neither is context represented explicitly in our model, nor is context change signaled. We find that memory-intact agents learn distinct context representations, and develop ABA renewal, whereas memory-impaired agents do not. These findings reproduce the behavior of control and hippocampal animals, respectively. We therefore propose that the role of the hippocampus in the context-dependence of extinction learning might stem from its function in episodic-like memory and not in context-representation per se. We conclude that context-dependence can emerge from raw visual inputs. 15 context A. Then the fear response is extinguished in a different context B, and 16 eventually tested again in context A in the absence of the US. Renewal means that the 17 fear response returns in the test phase despite the absence of the US [4]. ABA renewal 18 demonstrates that extinction learning does not delete the previously learned association, 19 although it might weaken the association [3]. Furthermore, ABA renewal underscores 20 that extinction learning is strongly context-dependent [4-10]. 21 April 23, 2020 1/20 Fear extinction is thought to depend mainly on three cerebral regions: The 22 amygdala, responsible for initial fear learning, the ventral medial prefrontal cortex, 23 guiding the extinction process, and the hippocampus, controlling the context-specific 24 retrieval of extinction [3] and/or encoding contextual information [11]. Here, we focus 25 on the role of the hippocampus in extinction learning tasks. Lesions or pharmacological 26 manipulations of the hippocampus have been reported to induce deficits in context 27 encoding [3, 11], which in turn impede context disambiguation during extinction 28 learning [5-8]. In particular, André and Manahan-Vaughan (2016) found that 29 intracerebral administration of a dopamine receptor agonist, which alters synaptic 30 plasticity in the hippocampus [12], reduced the context-dependent expression of renewal 31 in an appetitive ABA renewal task in a T-maze [9]. This task depends on the the 32 specific involvement of certain subfields of the hippocampus [13]. 33 However, it remains unclear how behavior becomes context-dependent and what 34 neural mechanisms underlie this learning. The majority of computational models 35 targets exclusively extinction tasks that involve classical conditioning, like the earliest 36 model of conditioning, the Rescorla-Wagner model [14] (for an overview, see [3]). 37 However, in its original form this model treats extinction learning as unlearning of 38 previously acquired associations between CS and US and so could not account for the 39 phenomenology of extinction learning, including rene...

Highlights d A fraction of sharp wave-ripples does not rely on CA2/3 generation d The subiculum can work as independent generator for sharp wave-ripples d These events can propagate back from the subiculum into CA1 and CA3