The hippocampus converts dynamic entorhinal inputs into stable spatial maps

Context and space coding in mossy cell population activity

The dentate gyrus plays a key role in the discrimination of memories by segregating and storing similar episodes. Whether hilar mossy cells, which constitute a major excitatory principal cell type in the mammalian hippocampus, contribute to this decorrelation function has remained largely unclear. Using two-photon calcium imaging of head-fixed mice performing a spatial virtual reality task, we show that mossy cell populations robustly discriminate between familiar and novel environments. The degree of discrimination depends on the extent of visual cue differences between contexts. A context decoder revealed that successful environmental classification is explained mainly by activity difference scores of mossy cells. By decoding mouse position, we reveal that in addition to place cells, the coordinated activity among active mossy cells markedly contributes to the encoding of space. Thus, by decorrelating context information according to the degree of environmental differences, mossy cell populations support pattern separation processes within the dentate gyrus.

Transformation of spatial representations along hippocampal circuits

The hippocampus is thought to provide the brain with a cognitive map of the external world by processing various types of spatial information. To understand how essential spatial variables such as direction, position, and distance are transformed along its circuits to construct this global map, we perform single-photon widefield microendoscope calcium imaging in the dentate gyrus and CA3 of mice freely navigating along a narrow corridor. We find that spatial activity maps in the dentate gyrus, but not in CA3, are correlated after aligning them to the running directions, suggesting that they represent the distance traveled along the track in egocentric coordinates. Together with population activity decoding, our data suggest that while spatial representations in the dentate gyrus and CA3 are anchored in both egocentric and allocentric coordinates, egocentric distance coding is more prevalent in the dentate gyrus than in CA3, providing insights into the assembly of the cognitive map.

Internal world models in humans, animals, and AI

How do brains-biological or artificial-respond and adapt to an ever-changing environment? In a recent meeting, experts from various fields of neuroscience and artificial intelligence met to discuss internal world models in brains and machines, arguing for an interdisciplinary approach to gain deeper insights into the underlying mechanisms.

The homogenous hippocampus: How hippocampal cells process available and potential goals

We present here a view of the firing patterns of hippocampal cells that is contrary, both functionally and anatomically, to conventional wisdom. We argue that the hippocampus responds to efference copies of goals encoded elsewhere; and that it uses these to detect and resolve conflict or interference between goals in general. While goals can involve space, hippocampal cells do not encode spatial (or other special types of) memory, as such. We also argue that the transverse circuits of the hippocampus operate in an essentially homogeneous way along its length. The apparently different functions of different parts (e.g. memory retrieval versus anxiety) result from the different (situational/motivational) inputs on which those parts perform the same fundamental computational operations. On this view, the key role of the hippocampus is the iterative adjustment, via Papez-like circuits, of synaptic weights in cell assemblies elsewhere.

Multiphoton imaging of hippocampal neural circuits: techniques and biological insights into region-, cell-type-, and pathway-specific functions

Significance

The function of the hippocampus in behavior and cognition has long been studied primarily through electrophysiological recordings from freely moving rodents. However, the application of optical recording methods, particularly multiphoton fluorescence microscopy, in the last decade or two has dramatically advanced our understanding of hippocampal function. This article provides a comprehensive overview of techniques and biological findings obtained from multiphoton imaging of hippocampal neural circuits.

Aim

This review aims to summarize and discuss the recent technical advances in multiphoton imaging of hippocampal neural circuits and the accumulated biological knowledge gained through this technology.

Approach

First, we provide a brief overview of various techniques of multiphoton imaging of the hippocampus and discuss its advantages, drawbacks, and associated key innovations and practices. Then, we review a large body of findings obtained through multiphoton imaging by region (CA1 and dentate gyrus), cell type (pyramidal neurons, inhibitory interneurons, and glial cells), and cellular compartment (dendrite and axon).

Results

Multiphoton imaging of the hippocampus is primarily performed under head-fixed conditions and can reveal detailed mechanisms of circuit operation owing to its high spatial resolution and specificity. As the hippocampus lies deep below the cortex, its imaging requires elaborate methods. These include imaging cannula implantation, microendoscopy, and the use of long-wavelength light sources. Although many studies have focused on the dorsal CA1 pyramidal cells, studies of other local and inter-areal circuitry elements have also helped provide a more comprehensive picture of the information processing performed by the hippocampal circuits. Imaging of circuit function in mouse models of Alzheimer's disease and other brain disorders such as autism spectrum disorder has also contributed greatly to our understanding of their pathophysiology.

Conclusions

Multiphoton imaging has revealed much regarding region-, cell-type-, and pathway-specific mechanisms in hippocampal function and dysfunction in health and disease. Future technological advances will allow further illustration of the operating principle of the hippocampal circuits via the large-scale, high-resolution, multimodal, and minimally invasive imaging.

Correlation between desynchrony of hippocampal neural activity and hyperlocomotion in the model mice of schizophrenia and therapeutic effects of aripiprazole

AIMS

The hippocampus has been reported to be morphologically and neurochemically altered in schizophrenia (SZ). Hyperlocomotion is a characteristic SZ-associated behavioral phenotype, which is associated with dysregulated dopamine system function induced by hippocampal hyperactivity. However, the neural mechanism of hippocampus underlying hyperlocomotion remains largely unclear.

METHODS

Mouse pups were injected with N-methyl-D-aspartate receptor antagonist (MK-801) or vehicle twice daily on postnatal days (PND) 7-11. In the adulthood phase, one cohort of mice underwent electrode implantation in field CA1 of the hippocampus for the recording local field potentials and spike activity. A separate cohort of mice underwent surgery to allow for calcium imaging of the hippocampus while monitoring the locomotion. Lastly, the effects of atypical antipsychotic (aripiprazole, ARI) were evaluated on hippocampal neural activity.

RESULTS

We found that the hippocampal theta oscillations were enhanced in MK-801-treated mice, but the correlation coefficient between the hippocampal spiking activity and theta oscillation was reduced. Consistently, although the rate and amplitude of calcium transients of hippocampal neurons were increased, their synchrony and correlation to locomotion speed were disrupted. ARI ameliorated perturbations produced by the postnatal MK-801 treatment.

CONCLUSIONS

These results suggest that the disruption of neural coordination may underly the neuropathological mechanism for hyperlocomotion of SZ.

Spatial Coding Dysfunction and Network Instability in the Aging Medial Entorhinal Cortex

Across species, spatial memory declines with age, possibly reflecting altered hippocampal and medial entorhinal cortex (MEC) function. However, the integrity of cellular and network-level spatial coding in aged MEC is unknown. Here, we leveraged in vivo electrophysiology to assess MEC function in young, middle-aged, and aged mice navigating virtual environments. In aged grid cells, we observed impaired stabilization of context-specific spatial firing, correlated with spatial memory deficits. Additionally, aged grid networks shifted firing patterns often but with poor alignment to context changes. Aged spatial firing was also unstable in an unchanging environment. In these same mice, we identified 458 genes differentially expressed with age in MEC, 61 of which had expression correlated with spatial firing stability. These genes were enriched among interneurons and related to synaptic transmission. Together, these findings identify coordinated transcriptomic, cellular, and network changes in MEC implicated in impaired spatial memory in aging.

Dentate gyrus is needed for memory retrieval

The hippocampus is crucial for acquiring and retrieving episodic and contextual memories. In previous studies, the inactivation of dentate gyrus (DG) neurons by chemogenetic- and optogenetic-mediated hyperpolarization led to opposing conclusions about DG's role in memory retrieval. One study used Designer Receptors Exclusively Activated by Designer Drugs (DREADD)-mediated clozapine N-oxide (CNO)-induced hyperpolarization and reported that the previously formed memory was erased, thus concluding that denate gyrus is needed for memory maintenance. The other study used optogenetic with halorhodopsin induced hyperpolarization and reported and dentate gyrus is needed for memory retrieval. We hypothesized that this apparent discrepancy could be due to the length of hyperpolarization in previous studies; minutes by optogenetics and several hours by DREADD/CNO. Since hyperpolarization interferes with anterograde and retrograde neuronal signaling, it is possible that the memory engram in the dentate gyrus and the entorhinal to hippocampus trisynaptic circuit was erased by long-term, but not with short-term hyperpolarization. We developed and applied an advanced chemogenetic technology to selectively silence synaptic output by blocking neurotransmitter release without hyperpolarizing DG neurons to explore this apparent discrepancy. We performed in vivo electrophysiology during trace eyeblink in a rabbit model of associative learning. Our work shows that the DG output is required for memory retrieval. Based on previous and recent findings, we propose that the actively functional anterograde and retrograde neuronal signaling is necessary to preserve synaptic memory engrams along the entorhinal cortex to the hippocampal trisynaptic circuit.

A consistent map in the medial entorhinal cortex supports spatial memory

The medial entorhinal cortex (MEC) is hypothesized to function as a cognitive map for memory-guided navigation. How this map develops during learning and influences memory remains unclear. By imaging MEC calcium dynamics while mice successfully learned a novel virtual environment over ten days, we discovered that the dynamics gradually became more spatially consistent and then stabilized. Additionally, grid cells in the MEC not only exhibited improved spatial tuning consistency, but also maintained stable phase relationships, suggesting a network mechanism involving synaptic plasticity and rigid recurrent connectivity to shape grid cell activity during learning. Increased c-Fos expression in the MEC in novel environments further supports the induction of synaptic plasticity. Unsuccessful learning lacked these activity features, indicating that a consistent map is specific for effective spatial memory. Finally, optogenetically disrupting spatial consistency of the map impaired memory-guided navigation in a well-learned environment. Thus, we demonstrate that the establishment of a spatially consistent MEC map across learning both correlates with, and is necessary for, successful spatial memory.

Subfield-specific interneuron circuits govern the hippocampal response to novelty in male mice

The hippocampus is the brain's center for episodic memories. Its subregions, the dentate gyrus and CA1-3, are differentially involved in memory encoding and recall. Hippocampal principal cells represent episodic features like movement, space, and context, but less is known about GABAergic interneurons. Here, we performed two-photon calcium imaging of parvalbumin- and somatostatin-expressing interneurons in the dentate gyrus and CA1-3 of male mice exploring virtual environments. Parvalbumin-interneurons increased activity with running-speed and reduced it in novel environments. Somatostatin-interneurons in CA1-3 behaved similar to parvalbumin-expressing cells, but their dentate gyrus counterparts increased activity during rest and in novel environments. Congruently, chemogenetic silencing of dentate parvalbumin-interneurons had prominent effects in familiar contexts, while silencing somatostatin-expressing cells increased similarity of granule cell representations between novel and familiar environments. Our data indicate unique roles for parvalbumin- and somatostatin-positive interneurons in the dentate gyrus that are distinct from those in CA1-3 and may support routing of novel information.

The medial entorhinal cortex encodes multisensory spatial information

Animals employ spatial information in multisensory modalities to navigate their natural environments. However, it is unclear whether the brain encodes such information in separate cognitive maps or integrates all into a single, universal map. We addressed this question in the microcircuit of the medial entorhinal cortex (MEC), a cognitive map of space. Using cellular-resolution calcium imaging, we examined the MEC of mice navigating virtual reality tracks, where visual and auditory cues provided comparable spatial information. We uncovered two cell types: "unimodality cells" and "multimodality cells". The unimodality cells specifically represent either auditory or visual spatial information. They are anatomically intermingled and maintain sensory preferences across multiple tracks and behavioral states. The multimodality cells respond to both sensory modalities with their responses shaped differentially by auditory and visual information. Thus, the MEC enables accurate spatial encoding during multisensory navigation by computing spatial information in different sensory modalities and generating distinct maps.

Direct cortical inputs to hippocampal area CA1 transmit complementary signals for goal-directed navigation

The subregions of the entorhinal cortex (EC) are conventionally thought to compute dichotomous representations for spatial processing, with the medial EC (MEC) providing a global spatial map and the lateral EC (LEC) encoding specific sensory details of experience. Yet, little is known about the specific types of information EC transmits downstream to the hippocampus. Here, we exploit in vivo sub-cellular imaging to record from EC axons in CA1 while mice perform navigational tasks in virtual reality (VR). We uncover distinct yet overlapping representations of task, location, and context in both MEC and LEC axons. MEC transmitted highly location- and context-specific codes; LEC inputs were biased by ongoing navigational goals. However, during tasks with reliable reward locations, the animals' position could be accurately decoded from either subregion. Our results revise the prevailing dogma about EC information processing, revealing novel ways spatial and non-spatial information is routed and combined upstream of the hippocampus.

A consistent map in the medial entorhinal cortex supports spatial memory

The medial entorhinal cortex (MEC) is hypothesized to function as a cognitive map for memory-guided navigation. How this map develops during learning and influences memory remains unclear. By imaging MEC calcium dynamics while mice successfully learned a novel virtual environment over ten days, we discovered that the dynamics gradually became more spatially consistent and then stabilized. Additionally, grid cells in the MEC not only exhibited improved spatial tuning consistency, but also maintained stable phase relationships, suggesting a network mechanism involving synaptic plasticity and rigid recurrent connectivity to shape grid cell activity during learning. Increased c-Fos expression in the MEC in novel environments further supports the induction of synaptic plasticity. Unsuccessful learning lacked these activity features, indicating that a consistent map is specific for effective spatial memory. Finally, optogenetically disrupting spatial consistency of the map impaired memory-guided navigation in a well-learned environment. Thus, we demonstrate that the establishment of a spatially consistent MEC map across learning both correlates with, and is necessary for, successful spatial memory.

Fear memory recall involves hippocampal somatostatin interneurons

Fear-related memory traces are encoded by sparse populations of hippocampal principal neurons that are recruited based on their inhibitory-excitatory balance during memory formation. Later, the reactivation of the same principal neurons can recall the memory. The details of this mechanism are still unclear. Here, we investigated whether disinhibition could play a major role in this process. Using optogenetic behavioral experiments, we found that when fear was associated with the inhibition of mouse hippocampal somatostatin positive interneurons, the re-inhibition of the same interneurons could recall fear memory. Pontine nucleus incertus neurons selectively inhibit hippocampal somatostatin cells. We also found that when fear was associated with the activity of these incertus neurons or fibers, the reactivation of the same incertus neurons or fibers could also recall fear memory. These incertus neurons showed correlated activity with hippocampal principal neurons during memory recall and were strongly innervated by memory-related neocortical centers, from which the inputs could also control hippocampal disinhibition in vivo. Nonselective inhibition of these mouse hippocampal somatostatin or incertus neurons impaired memory recall. Our data suggest a novel disinhibition-based memory mechanism in the hippocampus that is supported by local somatostatin interneurons and their pontine brainstem inputs.

Route-dependent spatial engram tagging in mouse dentate gyrus

The dentate gyrus (DG) of hippocampus is hypothesized to act as a pattern separator that distinguishes between similar input patterns during memory formation and retrieval. Sparse ensembles of DG cells associated with learning and memory, i.e. engrams, have been labeled and manipulated to recall novel context memories. Functional studies of DG cell activity have demonstrated the spatial specificity and stability of DG cells during navigation. To reconcile how the DG contributes to separating global context as well as individual navigational routes, we trained mice to perform a delayed-non-match-to-position (DNMP) T-maze task and labeled DG neurons during performance of this task on a novel T-maze. The following day, mice navigated a second environment: the same T-maze, the same T-maze with one route permanently blocked but still visible, or a novel open field. We found that the degree of engram reactivation across days differed based on the traversal of maze routes, such that mice traversing only one arm had higher ensemble overlap than chance but less overlap than mice running the full two-route task. Mice experiencing the open field had similar ensemble sizes to the other groups but only chance-level ensemble reactivation. Ensemble overlap differences could not be explained by behavioral variability across groups, nor did behavioral metrics correlate to degree of ensemble reactivation. Together, these results support the hypothesis that DG contributes to spatial navigation memory and that partially non-overlapping ensembles encode different routes within the context of an environment.

Cognitive outcomes caused by low-level lead, cadmium, and mercury mixture exposure at distinct phases of brain development

Contaminated water and food are the main sources of lead, cadmium, and mercury in the human body. Long-term and low-level ingestion of these toxic heavy metals may affect brain development and cognition. However, the neurotoxic effects of exposure to lead, cadmium, and mercury mixture (Pb + Cd + Hg) at different stages of brain development are rarely elucidated. In this study, different doses of low-level Pb + Cd + Hg were administered to Sprague-Dawley rats via drinking water during the critical stage of brain development, late stage, and after maturation, respectively. Our findings showed that Pb + Cd + Hg exposure decreased the density of memory- and learning-related dendritic spines in the hippocampus during the critical period of brain development, resulting in hippocampus-dependent spatial memory deficits. Only the density of learning-related dendritic spines was reduced during the late phase of brain development and a higher-dose of Pb + Cd + Hg exposure was required, which led to hippocampus-independent spatial memory abnormalities. Exposure to Pb + Cd + Hg after brain maturation revealed no significant change in dendritic spines or cognitive function. Further molecular analysis indicated that morphological and functional changes caused by Pb + Cd + Hg exposure during the critical phase were associated with PSD95 and GluA1 dysregulation. Collectively, the effects of Pb + Cd + Hg on cognition varied depending on the brain development stages.

Entorhinal grid-like codes and time-locked network dynamics track others navigating through space

Navigating through crowded, dynamically changing environments requires the ability to keep track of other individuals. Grid cells in the entorhinal cortex are a central component of self-related navigation but whether they also track others' movement is unclear. Here, we propose that entorhinal grid-like codes make an essential contribution to socio-spatial navigation. Sixty human participants underwent functional magnetic resonance imaging (fMRI) while observing and re-tracing different paths of a demonstrator that navigated a virtual reality environment. Results revealed that grid-like codes in the entorhinal cortex tracked the other individual navigating through space. The activity of grid-like codes was time-locked to increases in co-activation and entorhinal-cortical connectivity that included the striatum, the hippocampus, parahippocampal and right posterior parietal cortices. Surprisingly, the grid-related effects during observation were stronger the worse participants performed when subsequently re-tracing the demonstrator's paths. Our findings suggests that network dynamics time-locked to entorhinal grid-cell-related activity might serve to distribute information about the location of others throughout the brain.

Entorhinal cortex directs learning-related changes in CA1 representations

Learning-related changes in brain activity are thought to underlie adaptive behaviours1,2. For instance, the learning of a reward site by rodents requires the development of an over-representation of that location in the hippocampus3–6. How this learning-related change occurs remains unknown. Here we recorded hippocampal CA1 population activity as mice learned a reward location on a linear treadmill. Physiological and pharmacological evidence suggests that the adaptive over-representation required behavioural timescale synaptic plasticity (BTSP)7. BTSP is known to be driven by dendritic voltage signals that we proposed were initiated by input from entorhinal cortex layer 3 (EC3). Accordingly, the CA1 over-representation was largely removed by optogenetic inhibition of EC3 activity. Recordings from EC3 neurons revealed an activity pattern that could provide an instructive signal directing BTSP to generate the over-representation. Consistent with this function, our observations show that exposure to a second environment possessing a prominent reward-predictive cue resulted in both EC3 activity and CA1 place field density that were more elevated at the cue than at the reward. These data indicate that learning-related changes in the hippocampus are produced by synaptic plasticity directed by an instructive signal from the EC3 that seems to be specifically adapted to the behaviourally relevant features of the environment.

Hemisphere-specific spatial representation by hippocampal granule cells

The dentate gyrus (DG) output plays a key role in the emergence of spatial and contextual map representation within the hippocampus during learning. Differences in neuronal network activity have been observed between left and right CA1-3 areas, implying lateralization in spatial coding properties. Whether bilateral differences of DG granule cell (GC) assemblies encoding spatial and contextual information exist remains largely unexplored. Here, we employed two-photon calcium imaging of the left or the right DG to record the activity of GC populations over five consecutive days in head-fixed mice navigating through familiar and novel virtual environments. Imaging revealed similar mean GC activity on both sides. However, spatial tuning, context-selectivity and run-to-run place field reliability was markedly higher for DG place cells in the left than the right hemisphere. Moreover, the proportion of GCs reconfiguring their place fields between contexts was greater in the left DG. Thus, our data suggest that contextual information is differentially processed by GC populations depending on the hemisphere, with higher context discrimination in the left but a bias towards generalization in the right DG.

Closing the mechanistic gap: the value of microarchitecture in understanding cognitive networks

Cognitive neuroscience aims to provide biologically relevant accounts of cognition. Contemporary research linking spatial patterns of neural activity to psychological constructs describes 'where' hypothesised functions occur, but not 'how' these regions contribute to cognition. Technological, empirical, and conceptual advances allow this mechanistic gap to be closed by embedding patterns of functional activity in macro- and microscale descriptions of brain organisation. Recent work on the default mode network (DMN) and the multiple demand network (MDN), for example, highlights a microarchitectural landscape that may explain how activity in these networks integrates varied information, thus providing an anatomical foundation that will help to explain how these networks contribute to many different cognitive states. This perspective highlights emerging insights into how microarchitecture can constrain network accounts of human cognition.

Brain-wide reconstruction of inhibitory circuits after traumatic brain injury

Despite the fundamental importance of understanding the brain's wiring diagram, our knowledge of how neuronal connectivity is rewired by traumatic brain injury remains remarkably incomplete. Here we use cellular resolution whole-brain imaging to generate brain-wide maps of the input to inhibitory neurons in a mouse model of traumatic brain injury. We find that somatostatin interneurons are converted into hyperconnected hubs in multiple brain regions, with rich local network connections but diminished long-range inputs, even at areas not directly damaged. The loss of long-range input does not correlate with cell loss in distant brain regions. Interneurons transplanted into the injury site receive orthotopic local and long-range input, suggesting the machinery for establishing distant connections remains intact even after a severe injury. Our results uncover a potential strategy to sustain and optimize inhibition after traumatic brain injury that involves spatial reorganization of the direct inputs to inhibitory neurons across the brain.

Different Profiles of Spatial Navigation Deficits In Alzheimer’s Disease Biomarker-Positive Versus Biomarker-Negative Older Adults With Amnestic Mild Cognitive Impairment

Background

Spatial navigation impairment is a promising cognitive marker of Alzheimer’s disease (AD) that can reflect the underlying pathology.

Objectives

We assessed spatial navigation performance in AD biomarker positive older adults with amnestic mild cognitive impairment (AD aMCI) vs. those AD biomarker negative (non-AD aMCI), and examined associations between navigation performance, MRI measures of brain atrophy, and cerebrospinal fluid (CSF) biomarkers.

Methods

A total of 122 participants with AD aMCI (n = 33), non-AD aMCI (n = 31), mild AD dementia (n = 28), and 30 cognitively normal older adults (CN) underwent cognitive assessment, brain MRI (n = 100 had high-quality images for volumetric analysis) and three virtual navigation tasks focused on route learning (body-centered navigation), wayfinding (world-centered navigation) and perspective taking/wayfinding. Cognitively impaired participants underwent CSF biomarker assessment [amyloid-β1–42, total tau, and phosphorylated tau181 (p-tau181)] and amyloid PET imaging (n = 47 and n = 45, respectively), with a subset having both (n = 19).

Results

In route learning, AD aMCI performed worse than non-AD aMCI (p < 0.001), who performed similarly to CN. In wayfinding, aMCI participants performed worse than CN (both p ≤ 0.009) and AD aMCI performed worse than non-AD aMCI in the second task session (p = 0.032). In perspective taking/wayfinding, aMCI participants performed worse than CN (both p ≤ 0.001). AD aMCI and non-AD aMCI did not differ in conventional cognitive tests. Route learning was associated with parietal thickness and amyloid-β1–42, wayfinding was associated with posterior medial temporal lobe (MTL) volume and p-tau181 and perspective taking/wayfinding was correlated with MRI measures of several brain regions and all CSF biomarkers.

Conclusion

AD biomarker positive and negative older adults with aMCI had different profiles of spatial navigation deficits that were associated with posterior MTL and parietal atrophy and reflected AD pathology.

Recurrent Excitatory Feedback From Mossy Cells Enhances Sparsity and Pattern Separation in the Dentate Gyrus via Indirect Feedback Inhibition

It is generally appreciated that storing memories of specific events in the mammalian brain, and associating features of the environment with behavioral outcomes requires fine-tuning of the strengths of connections between neurons through synaptic plasticity. It is less understood whether the organization of neuronal circuits comprised of multiple distinct neuronal cell types provides an architectural prior that facilitates learning and memory by generating unique patterns of neuronal activity in response to different stimuli in the environment, even before plasticity and learning occur. Here we simulated a neuronal network responding to sensory stimuli, and systematically determined the effects of specific neuronal cell types and connections on three key metrics of neuronal sensory representations: sparsity, selectivity, and discriminability. We found that when the total amount of input varied considerably across stimuli, standard feedforward and feedback inhibitory circuit motifs failed to discriminate all stimuli without sacrificing sparsity or selectivity. Interestingly, networks that included dedicated excitatory feedback interneurons based on the mossy cells of the hippocampal dentate gyrus exhibited improved pattern separation, a result that depended on the indirect recruitment of feedback inhibition. These results elucidate the roles of cellular diversity and neural circuit architecture on generating neuronal representations with properties advantageous for memory storage and recall.