Bayesian decoding and its application in reading out spatial memory from neural ensembles

A Quarter of a Century of Place Cells

The hippocampus and associated cholinergic inputs have important roles in spatial memory in rodents. Muscarinic acetylcholine receptors (mAChRs) are involved in the communication of cholinergic signals and regulate spatial memory. They have been found to impact the memory encoding process, but the effect on memory retrieval is controversial. Previous studies report that scopolamine (a non-selective antagonist of mAChR) induces cognitive deficits on animals, resulting in impaired memory encoding, but the effect on memory retrieval is less certain. We tested the effects of blocking mAChRs on hippocampal network activity and neural ensembles that had previously encoded spatial information. The activity of hundreds of neurons in mouse hippocampal CA1 was recorded using calcium imaging with a miniaturised fluorescent microscope and properties of place cells and neuronal ensemble behaviour in a linear track environment were observed. We found that the decoding accuracy and the stability of spatial representation revealed by hippocampal neural ensemble were significantly reduced after the administration of scopolamine. Several other parameters, including neural firing rate, total number of active neurons, place cell number and spatial information content were affected. Similar results were also observed in a simulated hippocampal network model. This study enhances the understanding of the function of mAChRs on spatial memory impairment.

LTP takes route in the hippocampus.

Spatial working memory (SWM) requires the encoding, maintenance, and retrieval of spatially relevant information to guide decision-making. The medial prefrontal cortex (mPFC) has long been implicated in the ability of rodents to perform SWM tasks. While past studies have demonstrated that mPFC ensembles reflect past and future experiences, most findings are derived from tasks that have an experimental overlap between the encoding and retrieval of trajectory specific information. In this study, we recorded single units from the mPFC of rats as they performed a T-maze delayed non-match to position (DNMP) task. This task consists of an encoding dominant sample phase, a memory maintenance delay period, and retrieval dominant choice phase. Using a linear classifier, we investigated whether distinct ensembles collectively reflect various trajectory-dependent experiences. We find that a population of high-firing rate mPFC neurons both predict a future choice and reflect changes in trajectory-dependent behaviors. We then developed a modeling procedure that estimated the number of high and low-firing rate units required to dissociate between various experiences. We find that low firing rate ensembles weakly reflect the direction that rats were forced to turn on the sample phase. This was in contrast to the highly active population that could effectively predict both future decision-making on early stem traversals and trajectory-divergences at T-junction. Finally, we compared the ensemble size necessary to code a forced trajectory to the size required to predict a decision. We provide evidence to suggest that a larger number of highly active neurons are employed during decision-making processes when compared to rewarded forced behaviors. Together, our study provides important insight into how specific ensembles of mPFC units support upcoming choices and ongoing behavior during SWM.

Spatial learning is critical for survival and its underlying neuronal mechanisms have been studied extensively. These studies have revealed a wealth of information about the neural representations of space, such as place cells and boundary cells. While many studies have focused on how these representations emerge in the brain, their functional role in driving spatial learning and navigation has received much less attention. We extended an existing computational modeling tool-chain to study the functional role of spatial representations using closed-loop simulations of spatial learning. At the heart of the model agent was a spiking neural network that formed a ring attractor. This network received inputs from place and boundary cells and the location of the activity bump in this network was the output. This output determined the movement directions of the agent. We found that the navigation performance depended on the parameters of the place cell input, such as their number, the place field sizes, and peak firing rate, as well as, unsurprisingly, the size of the goal zone. The dependence on the place cell parameters could be accounted for by just a single variable, the overlap index, but this dependence was nonmonotonic. By contrast, performance scaled monotonically with the Fisher information of the place cell population. Our results therefore demonstrate that efficiently encoding spatial information is critical for navigation performance.

Neural activity was recorded from local groups of hippocampal single units in rats performing a spatial-memory task. The organization of functional correlates in these neural ensembles was investigated by examining the spatial relationships among the place fields of single units in each ensemble. The distance and overlap between place fields were determined together with the tuning of cellular activity to behavioral variables, including direction, speed, and turning angle during movements within place fields. The place fields of recorded neural ensembles were significantly clustered: closer in space and considerably more overlapped than chance when compared statistically with Monte Carlo simulations. Just as single units often have significant firing in more than one distinct location in the environment (subfields), the ensembles had multiple and distinct clusters of overlapping subfields. In addition, proximity and overlap between place fields were significantly, but weakly, correlated with similarity in optimal movement tuning parameters. These results suggest that the hippocampus maintains a local organization with respect to place fields despite having no apparent large-scale isomorphism with the spatial environment. The organization of multiple, clustered place fields with correlated movement tuning properties in small neural ensembles suggests the existence of functional neural ensembles serving to encode multiple sensory and behavioral aspects of a place or event. Such an organization is similar to that observed for neocortical association areas afferent to the hippocampal system.

The septohippocampal pathway contains cholinergic, GABAergic, and glutamatergic projections and has an established role in learning, memory, and hippocampal theta rhythm. Both GABAergic and cholinergic neurons in the medial septum-diagonal band of Broca (MSDB) have been associated with spatial memory, but the relationship between the two neuronal populations is not fully understood. The present study investigated the effect of selective GABAergic MSDB lesions on hippocampal acetylcholine (ACh) efflux and spatial memory during tasks that varied in memory demand. Male Sprague Dawley rats were given GABAergic lesions of the MSDB using GAT1-saporin (GAT1-SAP) and examined on spontaneous exploration (Experiment 1) and non-matching to position without (NMTP; Experiment 2) and with a delay (DNMTP; Experiment 3), while concurrently using in vivo microdialysis to measure hippocampal ACh efflux. Intraseptal GAT1-SAP treatment did not alter baseline or behaviorally stimulated hippocampal ACh efflux or maze exploration (Experiment 1). Moreover, GAT1-SAP did not alter evoked hippocampal ACh efflux related to NMTP nor did it impair working memory in NMTP (Experiment 2). In contrast, both ACh efflux and performance in DNMTP were impaired by intraseptal GAT1-SAP. Thus, GABAergic MSDB neurons are important for spatial working memory and modulate hippocampal ACh efflux under conditions of high memory load. The relationship between the septohippocampal cholinergic and GABAergic systems and working memory will be discussed.

Synchronous neuronal ensembles play a pivotal role in the consolidation of long-term memory in the hippocampus. However, their organization during the acquisition of spatial memory remains less clear. In this study, we used neuronal population voltage imaging to investigate the synchronization patterns of mice CA1 pyramidal neuronal ensembles during the exploration of a new environment, a critical phase for spatial memory acquisition. We found synchronous ensembles comprising approximately 40% of CA1 pyramidal neurons, firing simultaneously in brief windows (~25ms) during immobility and locomotion in novel exploration. Notably, these synchronous ensembles were not associated with contralateral ripple oscillations but were instead phase-locked to theta waves recorded in the contralateral CA1 region. Moreover, the subthreshold membrane potentials of neurons exhibited coherent intracellular theta oscillations with a depolarizing peak at the moment of synchrony. Among newly formed place cells, pairs with more robust synchronization during locomotion displayed more distinct place-specific activities. These findings underscore the role of synchronous ensembles in coordinating place cells of different place fields.

BackgroundThe lysophosphatidic acid LPA1 receptor regulates plasticity and neurogenesis in the adult hippocampus. Here, we studied whether absence of the LPA1 receptor modulated the detrimental effects of chronic stress on hippocampal neurogenesis and spatial memory.Methodology/Principal FindingsMale LPA1-null (NULL) and wild-type (WT) mice were assigned to control or chronic stress conditions (21 days of restraint, 3 h/day). Immunohistochemistry for bromodeoxyuridine and endogenous markers was performed to examine hippocampal cell proliferation, survival, number and maturation of young neurons, hippocampal structure and apoptosis in the hippocampus. Corticosterone levels were measured in another a separate cohort of mice. Finally, the hole-board test assessed spatial reference and working memory. Under control conditions, NULL mice showed reduced cell proliferation, a defective population of young neurons, reduced hippocampal volume and moderate spatial memory deficits. However, the primary result is that chronic stress impaired hippocampal neurogenesis in NULLs more severely than in WT mice in terms of cell proliferation; apoptosis; the number and maturation of young neurons; and both the volume and neuronal density in the granular zone. Only stressed NULLs presented hypocortisolemia. Moreover, a dramatic deficit in spatial reference memory consolidation was observed in chronically stressed NULL mice, which was in contrast to the minor effect observed in stressed WT mice.Conclusions/SignificanceThese results reveal that the absence of the LPA1 receptor aggravates the chronic stress-induced impairment to hippocampal neurogenesis and its dependent functions. Thus, modulation of the LPA1 receptor pathway may be of interest with respect to the treatment of stress-induced hippocampal pathology.

Chronic uncontrollable stress has been shown to produce various physiological alterations and impair mnemonic functions in the rodent hippocampus. Impacts on neuronal activities, however, have not been well investigated. The present study examined dorsal CA1 place cells to elucidate the computational changes associated with chronic stress effects on cognitive behaviors. After administering chronic restraint stress (CRS; 6 hours/day for ≥21 consecutive days) to adult male mice, several hippocampal characteristics were examined; i.e., spatial learning, in vitro synaptic plasticity, in vivo place cell recording, and western blot analysis to determine protein levels related to learning and memory. Behaviorally, CRS significantly impeded spatial learning but enhanced non-spatial cue learning on the Morris water maze. Physiologically, CRS reduced long-term potentiation (LTP) of Schaffer collateral/commisural-CA1 pathway, phospho-αCaMKII (alpha Ca2+/calmodulin-dependent protein kinase II) level in the hippocampus, and stability of spatial representation and the mean firing rates (FRs) of place cells. Moreover, the local cue-dependency of place fields was increased, and the intra-burst interval (IntraBI) between consecutive spikes within a burst was prolonged following CRS. These results extend the previous findings of stress impairing LTP and spatial learning to CRS modifying physical properties of spiking in place cells that contribute to changes in navigation and synaptic plasticity.

Place cells, spatially responsive hippocampal cells, provide the neural substrate supporting navigation and spatial memory. Historically most studies of these neurons have used electrophysiological recordings from implanted electrodes but optical methods, measuring intracellular calcium, are becoming increasingly common. Several methods have been proposed as a means to identify place cells based on their calcium activity but there is no common standard and it is unclear how reliable different approaches are. Here we tested four methods that have previously been applied to two-photon hippocampal imaging or electrophysiological data, using both model datasets and real imaging data. These methods use different parameters to identify place cells, including the peak activity in the place field, compared to other locations (the Peak method); the stability of cells’ activity over repeated traversals of an environment (Stability method); a combination of these parameters with the size of the place field (Combination method); and the spatial information held by the cells (Information method). The methods performed differently from each other on both model and real data. In real datasets, vastly different numbers of place cells were identified using the four methods, with little overlap between the populations identified as place cells. Therefore, choice of place cell detection method dramatically affects the number and properties of identified cells. Ultimately, we recommend the Peak method be used in future studies to identify place cell populations, as this method is robust to moderate variations in place field within a session, and makes no inherent assumptions about the spatial information in place fields, unless there is an explicit theoretical reason for detecting cells with more narrowly defined properties.

Disrupted Place Cell Remapping and Impaired Grid Cells in a Knockin Model of Alzheimer's Disease

Repeated brief seizures induce progressive hippocampal neuron loss and memory deficits

Studies in both experimental animals and human patients have demonstrated that peripheral vestibular lesions, especially bilateral lesions, are associated with spatial memory impairment that is long-lasting and may even be permanent. Electrophysiological evidence from animals indicates that bilateral vestibular loss causes place cells and theta activity to become dysfunctional; the most recent human evidence suggests that the hippocampus may cause atrophy in patients with bilateral vestibular lesions. Taken together, these studies suggest that self-motion information provided by the vestibular system is important for the development of spatial memory by areas of the brain such as the hippocampus, and when it is lost, spatial memory is impaired. This naturally suggests the converse possibility that activation of the vestibular system may enhance memory. Surprisingly, there is some human evidence that this may be the case. This review considers the relationship between the vestibular system and memory and suggests that the evolutionary age of this primitive sensory system as well as how it detects self-motion (i.e., detection of acceleration vs. velocity) may be the reasons for its unique contribution to spatial memory.

Spatial and memory processes rely on a complex circuitry in which the hippocampal formation seems to be crucial. The subiculum (SUB), located at the core of this structure receives inputs from grid cells of the medial entorhinal cortex (MEC), place cells from the CA1 area and thalamic nucleus involved in spatial coding. In return, the SUB mediates the output to cortical and sub-cortical regions, mediating memory processes. Although previous work described a heterogeneous population of spatial neurons in the SUB, its role in spatial coding and memory processes is unclear.