The hippocampus is well known as a central site for memory processing-critical for storing and later retrieving the experiences events of daily life so they can be used to shape future behavior. Much of what we know about the physiology underlying hippocampal function comes from spatial navigation studies in rodents, which have allowed great strides in understanding how the hippocampus represents experience at the cellular level. However, it remains a challenge to reconcile our knowledge of spatial encoding in the hippocampus with its demonstrated role in memory-dependent tasks in both humans and other animals. Moreover, our understanding of how networks of neurons coordinate their activity within and across hippocampal subregions to enable the encoding, consolidation, and retrieval of memories is incomplete. In this chapter, we explore how information may be represented at the cellular level and processed via coordinated patterns of activity throughout the subregions of the hippocampal network.

The hippocampus is well known as a central site for memory processing-critical for storing and later retrieving the experiences events of daily life so they can be used to shape future behavior. Much of what we know about the physiology underlying hippocampal function comes from spatial navigation studies in rodents, which have allowed great strides in understanding how the hippocampus represents experience at the cellular level. However, it remains a challenge to reconcile our knowledge of spatial encoding in the hippocampus with its demonstrated role in memory-dependent tasks in both humans and other animals. Moreover, our understanding of how networks of neurons coordinate their activity within and across hippocampal subregions to enable the encoding, consolidation, and retrieval of memories is incomplete. In this chapter, we explore how information may be represented at the cellular level and processed via coordinated patterns of activity throughout the subregions of the hippocampal network.

New recording technologies and the potential for closed-loop experiments have led to an increasing demand for computationally efficient and accurate algorithms to decode population spiking activity in multi-dimensional spaces. Exact point process filters can accurately decode low-dimensional signals, but are computationally intractable for high-dimensional signals. Approximate Gaussian filters are computationally efficient, but are inaccurate when the signals have complex distributions and nonlinear dynamics. Even particle filter methods tend to become inefficient and inaccurate when the filter distribution has multiple peaks. Here, we develop a new point process filter algorithm that combines the computational efficiency of approximate Gaussian methods with a numerical accuracy that exceeds standard particle filters. We use a mixture of Gaussian model for the posterior at each time step, allowing for an analytic solution to the computationally expensive filter integration step. During non-spike intervals, the filter needs only to update the mean, covariance, and mixture weight of each component. At spike times, a sampling procedure is used to update the filtering distribution and find the number of Gaussian mixture components necessary to maintain an accurate approximation. We illustrate the application of this algorithm to the problem of decoding a rat's position and velocity in a maze from hippocampal place cell data using both 2-D and 4-D decoders.

New recording technologies and the potential for closed-loop experiments have led to an increasing demand for computationally efficient and accurate algorithms to decode population spiking activity in multi-dimensional spaces. Exact point process filters can accurately decode low-dimensional signals, but are computationally intractable for high-dimensional signals. Approximate Gaussian filters are computationally efficient, but are inaccurate when the signals have complex distributions and nonlinear dynamics. Even particle filter methods tend to become inefficient and inaccurate when the filter distribution has multiple peaks. Here, we develop a new point process filter algorithm that combines the computational efficiency of approximate Gaussian methods with a numerical accuracy that exceeds standard particle filters. We use a mixture of Gaussian model for the posterior at each time step, allowing for an analytic solution to the computationally expensive filter integration step. During non-spike intervals, the filter needs only to update the mean, covariance, and mixture weight of each component. At spike times, a sampling procedure is used to update the filtering distribution and find the number of Gaussian mixture components necessary to maintain an accurate approximation. We illustrate the application of this algorithm to the problem of decoding a rat's position and velocity in a maze from hippocampal place cell data using both 2-D and 4-D decoders.

Alzheimer's disease (AD) is characterized by progressive memory loss, and there is a pressing need to identify early pathophysiological alterations that predict subsequent memory impairment. Hippocampal sharp-wave ripples (SWRs)-electrophysiological signatures of memory reactivation in the hippocampus-are a compelling candidate for this purpose. Mouse models of AD show reductions in both SWR abundance and associated slow gamma (SG) power during aging, but these alterations have yet to be directly linked to memory impairments. In aged apolipoprotein E4 knockin (apoE4-KI) mice-a model of the major genetic risk factor for AD-we find that reduced SWR abundance and associated CA3 SG power predicted spatial memory impairments measured 1-2 months later. Importantly, SWR-associated CA3 SG power reduction in young apoE4-KI mice also predicted spatial memory deficits measured 10 months later. These results establish features of SWRs as potential functional biomarkers of memory impairment in AD.

SUMMARY

Alzheimer’s disease (AD) is characterized by progressive memory loss, and there is a pressing need to identify early pathophysiological alterations that predict subsequent memory impairment. Hippocampal sharp-wave ripples (SWRs) – electrophysiological signatures of memory reactivation in the hippocampus – are a compelling candidate for doing so. Mouse models of AD show reductions in both SWR abundance and associated slow gamma (SG) power during aging, but these alterations have yet to be directly linked to memory impairments. In aged apolipoprotein E4 knock in (apoE4-KI) mice – a model of the major genetic risk factor for AD – we found that reduced SWR abundance and associated CA3 SG power predicted spatial memory impairments measured 1–2 months later. Importantly, SWR-associated CA3 SG power reduction in young apoE4-KI mice also predicted spatial memory deficits measured 10 months later. These results establish features of SWRs as potential functional biomarkers of memory impairment in AD.

Animal behavior provides context for understanding disease models and physiology. However, that behavior is often characterized subjectively, creating opportunity for misinterpretation and misunderstanding. For example, spatial alternation tasks are treated as paradigmatic tools for examining memory; however, that link is actually an assumption. To test this assumption, we simulated a reinforcement learning (RL) agent equipped with a perfect memory process. We found that it learns a simple spatial alternation task more slowly and makes different errors than a group of male rats, illustrating that memory alone may not be sufficient to capture the behavior. We demonstrate that incorporating spatial biases permits rapid learning and enables the model to fit rodent behavior accurately. Our results suggest that even simple spatial alternation behaviors reflect multiple cognitive processes that need to be taken into account when studying animal behavior.SIGNIFICANCE STATEMENT Memory is a critical function for cognition whose impairment has significant clinical consequences. Experimental systems aimed at testing various sorts of memory are therefore also central. However, experimental designs to test memory are typically based on intuition about the underlying processes. We tested this using a popular behavioral paradigm: a spatial alternation task. Using behavioral modeling, we show that the straightforward intuition that these tasks just probe spatial memory fails to account for the speed at which rats learn or the types of errors they make. Only when memory-independent dynamic spatial preferences are added can the model learn like the rats. This highlights the importance of respecting the complexity of animal behavior to interpret neural function and validate disease models.

Apolipoprotein (apo) E4 is expressed in many types of brain cells, is associated with age-dependent decline of learning and memory in humans, and is the major genetic risk factor for AD. To determine whether the detrimental effects of apoE4 depend on its cellular sources, we generated human apoE knock-in mouse models in which the human APOE gene is conditionally deleted in astrocytes, neurons, or GABAergic interneurons. Here we report that deletion of apoE4 in astrocytes does not protect aged mice from apoE4-induced GABAergic interneuron loss and learning and memory deficits. In contrast, deletion of apoE4 in neurons does protect aged mice from both deficits. Furthermore, deletion of apoE4 in GABAergic interneurons is sufficient to gain similar protection. This study demonstrates a detrimental effect of endogenously produced apoE4 on GABAergic interneurons that leads to learning and memory deficits in mice and provides a novel target for drug development for AD related to apoE4.

Representations related to past experiences play a critical role in memory and decision-making processes. The rat hippocampus expresses these types of representations during sharp-wave ripple (SWR) events, and previous work identified a minority of SWRs that contain 'replay' of spatial trajectories at ∼20x the movement speed of the animal. Efforts to understand replay typically make multiple assumptions about which events to examine and what sorts of representations constitute replay. We therefore lack a clear understanding of both the prevalence and the range of representational dynamics associated with replay. Here, we develop a state space model that uses a combination of movement dynamics of different speeds to capture the spatial content and time evolution of replay during SWRs. Using this model, we find that the large majority of replay events contain spatially coherent, interpretable content. Furthermore, many events progress at real-world, rather than accelerated, movement speeds, consistent with actual experiences.