Evidence for an Evolutionarily Conserved Memory Coding Scheme in the Mammalian Hippocampus.

Decades of research identify the hippocampal formation as central to memory storage and recall. Events are stored via distributed population codes, the parameters of which (e.g., sparsity and overlap) determine both storage capacity and fidelity. However, it remains unclear whether the parameters governing information storage are similar between species. Because episodic memories are rooted in the space in which they are experienced, the hippocampal response to navigation is often used as a proxy to study memory. Critically, recent studies in rodents that mimic the conditions typical of navigation studies in humans and nonhuman primates (i.e., virtual reality) show that reduced sensory input alters hippocampal representations of space. The goal of this study was to quantify this effect and determine whether there are commonalities in information storage across species. Using functional molecular imaging, we observe that navigation in virtual environments elicits activity in fewer CA1 neurons relative to real-world conditions. Conversely, comparable neuronal activity is observed in hippocampus region CA3 and the dentate gyrus under both conditions. Surprisingly, we also find evidence that the absolute number of neurons used to represent an experience is relatively stable between nonhuman primates and rodents. We propose that this convergence reflects an optimal ensemble size for episodic memories.SIGNIFICANCE STATEMENT One primary factor constraining memory capacity is the sparsity of the engram, the proportion of neurons that encode a single experience. Investigating sparsity in humans is hampered by the lack of single-cell resolution and differences in behavioral protocols. Sparsity can be quantified in freely moving rodents, but extrapolating these data to humans assumes that information storage is comparable across species and is robust to restraint-induced reduction in sensory input. Here, we test these assumptions and show that species differences in brain size build memory capacity without altering the structure of the data being stored. Furthermore, sparsity in most of the hippocampus is resilient to reduced sensory information. This information is vital to integrating animal data with human imaging navigation studies.

Age Is Associated with Reduced Sharp-Wave Ripple Frequency and Altered Patterns of Neuronal Variability.

UNLABELLED: Spatial and episodic memory performance declines with age, and the neural basis for this decline is not well understood. Sharp-wave ripples are brief (∼70 ms) high-frequency oscillatory events generated in the hippocampus and are associated with the consolidation of spatial memories. Given the connection between ripple oscillations and memory consolidation, we investigated whether the structure of ripple oscillations and ripple-triggered patterns of single-unit activity are altered in aged rats. Local field and single-unit activity surrounding sharp-wave ripple events were examined in the CA1 region of the hippocampus of old (n = 5) and young (n = 6) F344 rats during periods of rest preceding and following performance on a place-dependent eyeblink-conditioning task. Neural responses in aged rats differed from responses in young rats in several ways. First, compared with young rats, the rate of ripple occurrence (ripple density) is reduced in aged rats during postbehavior rest. Second, mean ripple frequency during prebehavior and postbehavior rest is lower in aged animals (aged: 132 Hz; young: 146 Hz). Third, single neurons in aged animals responded more consistently from ripple to ripple. Fourth, variability in interspike intervals was greater in aged rats. Finally, neurons were tuned to a narrower range of phases of the ripple oscillation relative to young animals. Together, these results suggest that the CA1 network in aged animals has a reduced "vocabulary" of available representational states. SIGNIFICANCE STATEMENT: The hippocampus is a structure that is critical for the formation of episodic memories. Sharp-wave ripple events generated in the hippocampus have been implicated in memory consolidation processes critical to memory stabilization. We examine here whether these ripple oscillations are altered over the course of the life span, which could contribute to hippocampus-dependent memory deficits that occur during aging. This experiment used young and aged memory-impaired rats to examine age-related changes in ripple architecture, ripple-triggered spike variance, and spike-phase coherence. We found that there are, indeed, significant changes in characteristics of ripples in older animals that could impact consolidation processes and memory stabilization in the aged brain.

Age differences in strategy selection and risk preference during risk-based decision making.

Studies of the effects of aging on decision making suggest that choices can be altered in a variety of ways depending on the situation, the nature of the outcome and risk, or certainty levels. To better characterize how aging impacts decision making in rodents, young and aged Fischer 344 rats underwent a series of probabilistic discounting tasks in which reward magnitude and probabilities were manipulated. Young rats tended to choose 1 of 2 different strategies: (a) to press for the large/uncertain reward, regardless of the reward probability; or (b) to continually adapt their behavior according to the odds of winning. The first strategy was adopted by about half of the younger rats, the second by the remaining young animals and the entire group of aged rats. Additionally, we found that when the odds of winning were varied from uncertain to certain during a session, aged rats chose most often the lever associated with the small/certain reward. This is consistent with an interpretation of increased risk aversion. When this behavior was further characterized using a lose-shift analysis, it appears that older rats exhibited an increased sensitivity to negative feedback. In contrast, sensitivity to wins was unaltered in aged rats compared with young, suggesting that aging selectively impacts rat's behavior following losses. In line with some human aging studies, it appears that aged rats are either more risk averse or have a greater certainty bias, which may result from age differences in emotion regulation.

Enhanced performance of aged rats in contingency degradation and instrumental extinction tasks.

Normal aging in rats affects behavioral performance on a variety of associative learning tasks under Pavlovian conditions. There is little information, however, on whether aging also impacts performance of instrumental tasks. Young (9-12 months) and aged (24-27 months) Fisher 344 rats were trained to press distinct levers associated with either maltodextrin or sucrose. The rats in both age groups increased their lever press frequency at a similar rate, suggesting that the initial acquisition of this instrumental task is not affected by aging. Using a contingency degradation procedure, we then addressed whether aged rats could adapt their behavior to changes in action-outcome contingencies. We found that young and aged rats do adapt, but that a different schedule of reinforcement is necessary to optimize performance in each age group. Finally, we also addressed whether aged rats can extinguish a lever press action as well as young rats, using 2 40-min extinction sessions on consecutive days. While extinction profiles were similar in young and aged rats on the first day of training, aged rats were faster to extinguish their lever presses on the second day, in spite of their performance levels being similar at the beginning of the session. Together these data support the finding that acquisition of instrumental lever press behaviors is preserved in aged rats and suggest that they have a different threshold for switching strategies in response to changes in action-outcome associations. This pattern of result implies that age-related changes in the brain are heterogeneous and widespread across structures.

Tracking the course of hippocampal representations during learning: when is the map required?

Distinct ensembles of hippocampal cells can be active in numerous contexts, but specific "cognitive maps" tend to be retrieved on repeat visits to the same place. During aging, the reliability of map retrieval in CA1 networks is reduced; this provides a unique opportunity to investigate correlations between inconsistent activity patterns in circuits hypothesized to enable context encoding and hippocampus-dependent learning ability. Here, CA1 pyramidal cells were recorded in six young and six old rats, while memory for specific locations was probed using a place-dependent eyeblink conditioning task. Rats were conditioned twice daily for 31 days, during which a total of 8259 and 7042 cells were recorded from young and old rats, respectively. Spontaneous remapping, a change in location of the majority of place fields between two consecutive sessions in the same environment, was observed in two young rats and four old rats during this task, but only after at least 13 days of training. Under these conditions the altered network representation did not result in loss of spatial accuracy of the blink, and in fact those rats with the best place conditioning remapped the most, whereas those with the best memory in a spatial water maze task remapped the least. These results suggest that when the hippocampal representation for a particular context is weak or unstable, such as can occur in senescence, extra-hippocampal systems that mediate alternate learning strategies are more likely to dominate behavior.

Reduced gamma frequency in the medial frontal cortex of aged rats during behavior and rest: implications for age-related behavioral slowing.

Age-related cognitive and behavioral slowing may be caused by changes in the speed of neural signaling or by changes in the number of signaling steps necessary to achieve a given function. In the mammalian cortex, neural communication is organized by a 30-100 Hz "gamma" oscillation. There is a putative link between the gamma frequency and the speed of processing in a neural network: the dynamics of pyramidal neuron membrane time constants suggest that synaptic integration is framed by the gamma cycle, and pharmacological slowing of gamma also slows reaction times on behavioral tasks. The present experiments identify reductions in a robust 40-70 Hz gamma oscillation in the aged rat medial frontal cortex. The reductions were observed in the form of local field potentials, later peaks in fast-spiking neuron autocorrelations, and delays in the spiking of inhibitory neurons following local excitatory signals. Gamma frequency did not vary with movement speed, but rats with slower gamma also moved more slowly. Gamma frequency age differences were not observed in hippocampus. Hippocampal CA1 fast-spiking neurons exhibited interspike intervals consistent with a fast (70-100 Hz) gamma frequency, a pattern maintained across theta phases and theta frequencies independent of fluctuations in the average firing rates of the neurons. We propose that an average lengthening of the cortical 15-25 ms gamma cycle is one factor contributing to age-related slowing and that future attempts to offset cognitive declines will find a target in the response of fast-spiking inhibitory neurons to excitatory inputs.

Differential effects of experience on tuning properties of macaque MTL neurons in a passive viewing task.

The structures of the medial temporal lobe (MTL) have been shown to be causally involved in episodic and recognition memory. However, recent work in a number of species has demonstrated that impairments in recognition memory seen following lesions of the perirhinal cortex (PRh) can be accounted for by deficits in perceptual discrimination. These findings suggest that object representation, rather than explicit recognition memory signals, may be crucial to the mnemonic process. Given the large amount of visual information encountered by primates, there must be a reconsideration of the mechanisms by which the brain efficiently stores visually presented information. Previous neurophysiological recordings from MTL structures in primates have largely focused on tasks that implicitly define object familiarity (i.e., novel vs. familiar) or contain significant mnemonic demands (e.g., conditional associations between two stimuli), limiting their utility in understanding the mechanisms underlying visual object recognition and information storage. To clarify how different regions in the MTL may contribute to visual recognition, we recorded from three rhesus macaques performing a passive viewing task. The task design systematically varies the relative familiarity of different stimuli enabling an examination of how neural activity changes as a function of experience. The data collected during this passive viewing task revealed that neurons in the MTL are generally not sensitive to the relative familiarity of a stimulus. In addition, when the specificity (i.e., which images a neuron was selective for) of individual neurons was analyzed, there was a significant dissociation between different medial temporal regions, with only neurons in TF, but not CA3 or the PRh, altering their activity as stimuli became familiar. The implications of these findings are discussed in the context of how MTL structures process information during a passive viewing paradigm.

Greater running speeds result in altered hippocampal phase sequence dynamics.

Hebb (1949) described a "phase sequence" to be the sequential activation of sets of cell assemblies. Within the hippocampus, cell assemblies have been described as groups of coactive neurons whose place fields overlap. Membership of assemblies in a phase sequence changes systematically as a rat travels through an environment, serving to accelerate the temporal order that place fields are encountered during a single theta cycle. This sweeping forward of network activity ("look ahead"), results in locations in front of the animal being transiently represented. In this experiment, a population vector-based reconstruction method was used to capture the look ahead and reveals that the composition of the phase sequence changes with velocity, such that more cell assemblies are active within a theta cycle at higher running speeds. These results are consistent with hypotheses suggesting that hippocampal networks generate short time scale predictions of future events to optimize behavior.

Methodological considerations on the use of template matching to study long-lasting memory trace replay.

Replay of behaviorally induced neural activity patterns during subsequent sleep has been suggested to play an important role in memory consolidation. Many previous studies, mostly involving familiar experiences, suggest that such reactivation occurs, but decays quickly (approximately 1 h). Recently, however, long-lasting (up to approximately 48 h) "reverberation" of neural activity patterns induced by a novel experience was reported on the basis of a template-matching analysis. Because detection and quantification of memory-trace replay depends critically on analysis methods, we investigated the statistical properties of the template-matching method and analyzed rodent neural ensemble activity patterns after a novel experience. For comparison, we also analyzed the same data with an independent analysis technique, the explained variance method. Contrary to the recent report, we did not observe significant long-lasting reverberation using either the template matching or the explained variance approaches. The latter, however, did reveal short-lasting reactivation in the hippocampus and prefrontal cortex. In addition, detailed analysis of the template-matching method shows that, in the present study, coarse mean firing rate differences among neurons, but not fine temporal spike structures, dominate the results of template matching. Most importantly, it is also demonstrated that partial comparisons of template-matching correlations, such as used in the recent paper, may lead to erroneous conclusions. These investigations indicate that the outcome of template-matching analysis is very sensitive to the conditions of how it is applied, and should be interpreted cautiously, and that the existence of long-lasting reverberation after a novel experience requires additional verification.

Recent behavioral history modifies coupling between cell activity and Arc gene transcription in hippocampal CA1 neurons.

The ability of neurons to alter their transcriptional programs in response to synaptic input is of fundamental importance to the neuroplastic mechanisms underlying learning and memory. Because of technical limitations of conventional gene detection methods, the current view of activity-dependent neural transcription derives from experiments in which neurons are assumed quiescent until a signaling stimulus is given. The present study was designed to move beyond this static model by examining how earlier episodes of neural activity influence transcription of the immediate-early gene Arc. Using a sensitive FISH method that detects primary transcript at genomic alleles, the proportion of hippocampal CA1 neurons that activate transcription of Arc RNA was constant at approximately 40% in response to both a single novel exploration session and daily sessions repeated over 9 days. This proportion is similar to the percentage of active neurons defined electrophysiologically. However, this close correspondence was disrupted in rats exposed briefly, but repeatedly, to the same environment within a single day. Arc transcription in CA1 neurons declined dramatically after as few as four 5-min sessions, despite stable electrophysiological activity during all sessions. Additional experiments indicate that the decrement in Arc transcription occurred at the cellular, rather than synaptic level, and was not simply linked to habituation to novelty. Thus, the neural genomic response is governed by recent, but not remote, cell firing history in the behaving animal. This state-dependence of neuronal transcriptional coupling provides a mechanism of metaplasticity and may regulate capacity for synaptic modification in neural networks.

Self-motion and the hippocampal spatial metric.

Self-motion signals are sufficient for animal navigation ("path integration") and for updating hippocampal location-specific firing. The contributions of ambulatory, vestibular, and optic self-motion signals to CA1 unit activity and EEG were studied while rats either walked or drove a car between locations on a circular track (referred to as WALK and CAR, respectively) or experienced pseudomotion, in which the animal was stationary and the environment was rotated (WORLD). Fewer pyramidal cells expressed place fields during CAR and those that did exhibited substantially larger place fields. The number of theta cycles required to traverse a place field increased, whereas the slope of the theta phase of firing versus position function was reduced. The presence and/or location of place fields were not well correlated between conditions. These effects were even more accentuated during WORLD. These results are not explainable by a simple "smearing out" of place fields but, in terms of size of place fields relative to the track size, are comparable with what would be observed if the track circumference was reduced and the animal moved around it at a correspondingly slower speed. Theta (and its 14-18 Hz harmonic) power were dependent on velocity, but the gain of this function was substantially reduced during CAR and WORLD, again as if the rat were moving more slowly. The spatial scale over which the hippocampal population vector is updated appears to be derived primarily from the gain of a self-motion velocity signal with approximately equal components derived from ambulation, vestibular, and optic-flow signals.

Self-motion and the origin of differential spatial scaling along the septo-temporal axis of the hippocampus.

Spatial scaling of place specific activity in the hippocampus varies systematically from the septal pole (high resolution) to the temporal pole (low resolution). Place fields get progressively larger, and the probability of observing a field in a given environment gets progressively smaller. It was previously found that decoupling movement in space from ambulation, by having the animal actively ride on a mobile platform, results in marked enlargement of the spatial scale factor in the dorsal hippocampus and a reduction in the increase in theta rhythm power with running speed, suggesting that a self-motion signal determines the spatial scale at which the hippocampal population vector updates. These results led to the hypothesis that the gain of the self-motion signal may vary systematically along the septo-temporal axis of the hippocampus. To test this hypothesis, EEG theta rhythm and ensembles of CA1 pyramidal cells and interneurons were recorded from the extreme dorsal and middle portions of the hippocampus. Pyramidal cell population vectors representing successive locations became decorrelated over substantially shorter distances in the dorsal than in the middle hippocampus. Dorsal pyramidal cells had smaller place fields, higher mean and peak firing rates, and higher intrinsic oscillation frequencies during track running than that of middle pyramidal cells. Both dorsal pyramidal cells and interneurons had more elevated mean rates during running, compared with rest, than that of the corresponding cell classes in the middle hippocampus, and both cell classes increased their rates more as a function of speed in the dorsal hippocampus.The amplitude, but not the frequency of fissure recorded theta rhythm, increased more as a function of running speed in the dorsal than in the middle hippocampus. We conclude that variation in the neuronal response to movement speed is the likely basis for the systematic variation in spatial scaling along the septo-temporal axis of the hippocampus.

Bimodality of theta phase precession in hippocampal place cells in freely running rats.

The firing of hippocampal principal cells in freely running rats exhibits a progressive phase retardation as the animal passes through a cell's "place" field. This "phase precession" is more complex than a simple linear shift of phase with position. In the present paper, phase precession is quantitatively analyzed by fitting multiple (1-3) normal probability density functions to the phase versus position distribution of spikes in rats making repeated traversals of the place fields. The parameters were estimated by the Expectation Maximization method. Three data sets including CA1 and DG place cells were analyzed. Although the phase-position distributions vary among different cells and regions, this complexity is well described by a superposition of two normal distribution functions, suggesting that the firing behavior consists of two components. This conclusion is supported by the existence of two distinct maxima in the mean spike density in the phase versus position plane. In one component, firing phase shifts over a range of about 180 degrees. The second component, which occurs near the end of the traversal of the place field, exhibits a low correlation between phase and position and is anti-phase with the phase-shift component. The functional implications of the two components are discussed with respect to their possible contribution to learning and memory mechanisms.