The shift from a response strategy to object-in-place strategy during learning is accompanied by a matching shift in neural firing correlates in the hippocampus

Temporal context and latent state inference in the hippocampal splitter signal

The hippocampus is thought to enable the encoding and retrieval of ongoing experience, the organization of that experience into structured representations like contexts, maps, and schemas, and the use of these structures to plan for the future. A central goal is to understand what the core computations supporting these functions are, and how these computations are realized in the collective action of single neurons. A potential access point into this issue is provided by 'splitter cells', hippocampal neurons that fire differentially on the overlapping segment of trajectories that differ in their past and/or future. However, the literature on splitter cells has been fragmented and confusing, owing to differences in terminology, behavioral tasks, and analysis methods across studies. In this review, we synthesize consistent findings from this literature, establish a common set of terms, and translate between single-cell and ensemble perspectives. Most importantly, we examine the combined findings through the lens of two major theoretical ideas about hippocampal function: representation of temporal context and latent state inference. We find that unique signature properties of each of these models are necessary to account for the data, but neither theory, by itself, explains all of its features. Specifically, the temporal gradedness of the splitter signal is strong support for temporal context, but is hard to explain using state models, while its flexibility and task-dependence is naturally accounted for using state inference, but poses a challenge otherwise. These theories suggest a number of avenues for future work, and we believe their application to splitter cells is a timely and informative domain for testing and refining theoretical ideas about hippocampal function.

Subtle visual change in a virtual environment induces heterogeneous remapping systematically in CA1, but not CA3

Environmental change may lead to new memories or modify old ones, but the underlying neural mechanisms are largely unclear. We recorded hippocampal place cells simultaneously from CA1 and CA3 in a virtual reality environment. Compared with CA1, place cells in CA3 are more tolerant of individual landmark changes but undergo orthogonal changes to code distinctively different environments. As visual noise (virtual fog) is introduced to a visually enriched environment, place cells in CA1 split into two subpopulations: in one, place cells maintain their field locations while changing their firing rates to reflect sensory changes; in the other, place cells exhibit global remapping in response to the contextual change. In contrast, place cells in CA3 exhibit mainly rate remapping under the same conditions. Our results suggest that CA1 may simultaneously represent heterogeneous maps of the same environment when subtle visual noise induces both sensory and contextual changes.

Significance of visual scene-based learning in the hippocampal systems across mammalian species

The hippocampus and its associated cortical regions in the medial temporal lobe play essential roles when animals form a cognitive map and use it to achieve their goals. As the nature of map-making involves sampling different local views of the environment and putting them together in a spatially cohesive way, visual scenes are essential ingredients in the formative process of cognitive maps. Visual scenes also serve as important cues during information retrieval from the cognitive map. Research in humans has shown that there are regions in the brain that selectively process scenes and that the hippocampus is involved in scene-based memory tasks. The neurophysiological correlates of scene-based information processing in the hippocampus have been reported as "spatial view cells" in nonhuman primates. Like primates, it is widely accepted that rodents also use visual scenes in their background for spatial navigation and other kinds of problems. However, in rodents, it is not until recently that researchers examined the neural correlates of the hippocampus from the perspective of visual scene-based information processing. With the advent of virtual reality (VR) systems, it has been demonstrated that place cells in the hippocampus exhibit remarkably similar firing correlates in the VR environment compared with that of the real-world environment. Despite some limitations, the new trend of studying hippocampal functions in a visually controlled environment has the potential to allow investigation of the input-output relationships of network functions and experimental testing of traditional computational predictions more rigorously by providing well-defined visual stimuli. As scenes are essential for navigation and episodic memory in humans, further investigation of the rodents' hippocampal systems in scene-based tasks will provide a critical functional link across different mammalian species.

The hearing hippocampus

The hippocampus has a well-established role in spatial and episodic memory but a broader function has been proposed including aspects of perception and relational processing. Neural bases of sound analysis have been described in the pathway to auditory cortex, but wider networks supporting auditory cognition are still being established. We review what is known about the role of the hippocampus in processing auditory information, and how the hippocampus itself is shaped by sound. In examining imaging, recording, and lesion studies in species from rodents to humans, we uncover a hierarchy of hippocampal responses to sound including during passive exposure, active listening, and the learning of associations between sounds and other stimuli. We describe how the hippocampus' connectivity and computational architecture allow it to track and manipulate auditory information - whether in the form of speech, music, or environmental, emotional, or phantom sounds. Functional and structural correlates of auditory experience are also identified. The extent of auditory-hippocampal interactions is consistent with the view that the hippocampus makes broad contributions to perception and cognition, beyond spatial and episodic memory. More deeply understanding these interactions may unlock applications including entraining hippocampal rhythms to support cognition, and intervening in links between hearing loss and dementia.

Rodent mnemonic similarity task performance requires the prefrontal cortex

Mnemonic similarity task performance, in which a known target stimulus must be distinguished from similar lures, is supported by the hippocampus and perirhinal cortex. Impairments on this task are known to manifest with advancing age. Interestingly, disrupting hippocampal activity leads to mnemonic discrimination impairments when lures are novel, but not when they are familiar. This observation suggests that other brain structures support discrimination abilities as stimuli are learned. The prefrontal cortex (PFC) is critical for retrieval of remote events and executive functions, such as working memory, and is also particularly vulnerable to dysfunction in aging. Importantly, the medial PFC is reciprocally connected to the perirhinal cortex and neuron firing in this region coordinates communication between lateral entorhinal and perirhinal cortices to presumably modulate hippocampal activity. This anatomical organization and function of the medial PFC suggests that it contributes to mnemonic discrimination; however, this notion has not been empirically tested. In the current study, rats were trained on a LEGO object-based mnemonic similarity task adapted for rodents, and surgically implanted with guide cannulae targeting prelimbic and infralimbic regions of the medial PFC. Prior to mnemonic discrimination tests, rats received PFC infusions of the GABAA agonist muscimol. Analyses of expression of the neuronal activity-dependent immediate-early gene Arc in medial PFC and adjacent cortical regions confirmed muscimol infusions led to neuronal inactivation in the infralimbic and prelimbic cortices. Moreover, muscimol infusions in PFC impaired mnemonic discrimination performance relative to the vehicle control across all testing blocks when lures shared 50-90% feature overlap with the target. Thus, in contrast hippocampal infusions, PFC inactivation impaired target-lure discrimination regardless of the novelty or familiarity of the lures. These findings indicate the PFC plays a critical role in mnemonic similarity task performance, but the time course of PFC involvement is dissociable from that of the hippocampus.

The effects of developmental alcohol exposure on the neurobiology of spatial processing

The consumption of alcohol during gestation is detrimental to the developing central nervous system. One functional outcome of this exposure is impaired spatial processing, defined as sensing and integrating information pertaining to spatial navigation and spatial memory. The hippocampus, entorhinal cortex, and anterior thalamus are brain regions implicated in spatial processing and are highly susceptible to the effects of developmental alcohol exposure. Some of the observed effects of alcohol on spatial processing may be attributed to changes at the synaptic to circuit level. In this review, we first describe the impact of developmental alcohol exposure on spatial behavior followed by a summary of the development of brain areas involved in spatial processing. We then provide an examination of the consequences of prenatal and early postnatal alcohol exposure in rodents on hippocampal, anterior thalamus, and entorhinal cortex-dependent spatial processing from the cellular to behavioral level. We conclude by highlighting several unanswered questions which may provide a framework for future investigation.

Moderate prenatal alcohol exposure impairs performance by adult male rats in an object-place paired-associate task

Memory impairments, including spatial and object processing, are often observed in individuals with Fetal Alcohol Spectrum Disorders. The neurobiological basis of memory deficits after prenatal alcohol exposure (PAE) is often linked to structural and functional alterations in the medial temporal lobe, including the hippocampus. Recent evidence suggests that the medial temporal lobe plays a critical role in processing high-order sensory stimuli such as complex objects and their associated locations in space. In the first experiment, we tested male rat offspring with moderate PAE in a medial temporal-dependent object-place paired-associate (OPPA) task. The OPPA task requires a conditional discrimination between an identical pair of objects presented at two spatial locations 180° opposite arms of a radial arm maze. Food reinforcement is contingent upon selecting the correct object of the pair for a given spatial location. Adult rats were given a total of 10 trials per day over 14 consecutive days of training. PAE male rats made significantly more errors than male saccharin (SACC) control rats during acquisition of the OPPA task. In Experiment 2, rats performed an object-discrimination task in which a pair of objects were presented in a single arm of the maze. Moderate PAE and SACC control rats exhibited comparable performance. The results suggest that moderate PAE rats can learn to discriminate objects, but are impaired when required to discriminate between objects on the basis of spatial location in the environment.

Hippocampal proBDNF facilitates place learning strategy associated with neural activity in rats

Mature brain-derived neurotrophic factor has been shown to have a promotive effect on synaptic plasticity and spatial memory. The precursor of BDNF (proBDNF) has emerged as a protein against its mature form. However, it is unknown whether and how proBDNF regulates neural excitability and spatial behavior. Through infusion of cleavage-resistant proBDNF or its antibody into HPC, we sought evidence for the influences by employing multiple behavioral tests and recording hippocampal single-unit activity. Our behavioral findings showed that proBDNF induced beneficial effects on spatial learning by facilitating the use of the place strategy and inhibiting the response strategy, including (1) using more place search strategies but less response strategies, and (2) increasing the number of rats in choosing place strategies but not response strategies. Intriguingly, infusion of an anti-proBDNF antibody did not affect rats' training process but rendered the adaption to learning reversal training more difficult, indicating deficits in choosing the proper learning strategy. The training-induced increase in proBDNF promoted the firing rate of pyramidal neurons but not fast-spiking (FS) interneurons. Importantly, endogenous proBDNF facilitated the neural correlate of spatial, but not response, learning behavior. However, the anti-proBDNF antibody effectively reversed the strategy preference and inhibited neural activity. We herein propose that proBDNF exerts pivotal effects on neural excitability and the use of cognitive strategies to facilitate the spatial learning process.

Neural Firing Patterns Are More Schematic and Less Sensitive to Changes in Background Visual Scenes in the Subiculum than in the Hippocampus

Literature suggests that the hippocampus is central to processing visual scenes to remember contextual information, but the roles of its downstream structure, subiculum, remain unknown. Here, single units were recorded simultaneously in the dorsal CA1 and subiculum while male rats made spatial choices using visual scenes as cues in a T-maze. The firing fields of subicular neurons were schematically organized following the task structure, largely divided into pre-choice and post-choice epochs, whereas those of CA1 cells were more punctate and bound to specific locations. When the rats were tested with highly familiar scenes, neurons in the CA1 and subiculum were indistinguishable in coding the task-related information (e.g., scene, choice) through rate remapping. However, when the familiar scenes were blurred parametrically, the neurons in the CA1 responded sensitively to the novelty in task demand and changed its representations parametrically following the physical changes of the stimuli, whereas these functional characteristics were absent in the subiculum. These results suggest that the unique function of the hippocampus is to acquire contextual representations in association with discrete positions in space, especially when facing new and ambiguous scenes, whereas the subiculum may translate the position-bound visual contextual information of the hippocampus into schematic codes once learning is established.SIGNIFICANCE STATEMENT Although the potential functional significance has been recognized for decades for the subiculum, its exact roles in a goal-directed memory task still remain elusive. In the current study, we present experimental evidence that may indicate that the neural population in the subiculum could translate the location-bound spatial representations of the hippocampus into more schematic representations of task demands. Our findings also imply that the visual scene-based codes conveyed by the hippocampus and subiculum may be identical in a well learned task, whereas the hippocampus may be more specialized in representing altered visual scenes than the subiculum.

Still searching for the engram

For nearly a century, neurobiologists have searched for the engram-the neural representation of a memory. Early studies showed that the engram is widely distributed both within and across brain areas and is supported by interactions among large networks of neurons. Subsequent research has identified engrams that support memory within dedicated functional systems for habit learning and emotional memory, but the engram for declarative memories has been elusive. Nevertheless, recent years have brought progress from molecular biological approaches that identify neurons and networks that are necessary and sufficient to support memory, and from recording approaches and population analyses that characterize the information coded by large neural networks. These new directions offer the promise of revealing the engrams for episodic and semantic memories.

Medial prefrontal-perirhinal cortical communication is necessary for flexible response selection

The ability to use information from the physical world to update behavioral strategies is critical for survival across species. The prefrontal cortex (PFC) supports behavioral flexibility; however, exactly how this brain structure interacts with sensory association cortical areas to facilitate the adaptation of response selection remains unknown. Given the role of the perirhinal cortex (PER) in higher-order perception and associative memory, the current study evaluated whether PFC-PER circuits are critical for the ability to perform biconditional object discriminations when the rule for selecting the rewarded object shifted depending on the animal's spatial location in a 2-arm maze. Following acquisition to criterion performance on an object-place paired association task, pharmacological blockade of communication between the PFC and PER significantly disrupted performance. Specifically, the PFC-PER disconnection caused rats to regress to a response bias of selecting an object on a particular side regardless of its identity. Importantly, the PFC-PER disconnection did not interfere with the capacity to perform object-only or location-only discriminations, which do not require the animal to update a response rule across trials. These findings are consistent with a critical role for PFC-PER circuits in rule shifting and the effective updating of a response rule across spatial locations.

Influence of Proximal, Distal, and Vestibular Frames of Reference in Object-Place Paired Associate Learning in the Rat

Object-place paired associate learning has been used to test hypotheses regarding the neurobiological basis of memory in rodents. Much of this work has focused on the role of limbic and hippocampal-parahippocampal regions, as well as the use of spatial information derived from allothetic visual stimuli to determine location in an environment. It has been suggested that idiothetic self-motion (vestibular) signals and internal representations of directional orientation might play an important role in disambiguating between spatial locations when forming object-place associations, but this hypothesis has not been explicitly tested. In the present study, we investigated the relationship between allothetic (i.e., distal and proximal cues) and vestibular stimuli on performance of an object-place paired-associate task. The paired-associate task was composed of learning to discriminate between an identical pair of objects presented in 180° opposite arms of a radial arm maze. Thus, animals were required to select a particular object on the basis of spatial location (i.e., maze arm). After the animals acquired the object-place rule, a series of probe tests determined that rats utilize self-generated vestibular cues to discriminate between the two maze arms. Further, when available, animals showed a strong preference for local proximal cues associated with the maze. Together, the work presented here supports the establishment of an object-place task that requires both idiothetic and allothetic stimulus sources to guide choice behavior, and which can be used to further investigate the dynamic interactions between neural systems involved in pairing sensory information with spatial locations.

Learning-dependent Changes in the Neuronal Correlates of Response Inhibition in the Prefrontal Cortex and Hippocampus

It has been suggested that the hippocampus and the prefrontal cortex (PFC) play key roles in representing contextual memory and utilizing contextual information for flexible response selection. During response selection, a correct response should be facilitated and an incorrect response should be inhibited flexibly in association with a cueing stimulus. However, it is poorly understood how the hippocampal and PFC networks behave during such flexible control of facilitation and inhibition of behavioral responses. To find neural correlates of context-cued flexible response selection, the current study employed an object-place paired-associate (OPPA) task in which object A is only rewarded in place 1 and object B is associated with reward in place 2 while recording single units simultaneously from the hippocampus and PFC. During the task, response inhibition in front of a contextually wrong object is required for successful performance and such inhibitory responses were observed before the rat learned the task. A significant proportion of neurons that fired differentially depending on the existence of inhibitory behavior in the PFC was observed during the pre-learning stage. By contrast, the proportion of such neurons in the hippocampus was significantly greater than chance during post-learning stage. The results suggest that the development of inhibitory behavior is a critical behavioral marker that foretells an upcoming acquisition of the task and the hippocampus and PFC are involved in learning contextual response selection by learning how to control the inhibition of behavior as learning progresses.

Perirhinal cortical inactivation impairs object-in-place memory and disrupts task-dependent firing in hippocampal CA1, but not in CA3

Objects and their locations can associatively define an event and a conjoint representation of object-place can form an event memory. Remembering how to respond to a certain object in a spatial context is dependent on both hippocampus and perirhinal cortex (PER). However, the relative functional contributions of the two regions are largely unknown in object-place associative memory. We investigated the PER influence on hippocampal firing in a goal-directed object-place memory task by comparing the firing patterns of CA1 and CA3 of the dorsal hippocampus between conditions of PER muscimol inactivation and vehicle control infusions. Rats were required to choose one of the two objects in a specific spatial context (regardless of the object positions in the context), which was shown to be dependent on both hippocampus and PER. Inactivation of PER with muscimol (MUS) severely disrupted performance of well-trained rats, resulting in response bias (i.e., choosing any object on a particular side). MUS did not significantly alter the baseline firing rates of hippocampal neurons. We measured the similarity in firing patterns between two trial conditions in which the same target objects were chosen on opposite sides within the same arm [object-in-place (O-P) strategy] and compared the results with the similarity in firing between two trial conditions in which the rat chose any object encountered on a particular side [response-in-place (R-P) strategy]. We found that the similarity in firing patterns for O-P trials was significantly reduced with MUS compared to control conditions (CTs). Importantly, this was largely because MUS injections affected the O-P firing patterns in CA1 neurons, but not in CA3. The results suggest that PER is critical for goal-directed organization of object-place associative memory in the hippocampus presumably by influencing how object information is associated with spatial information in CA1 according to task demand.

Contextual behavior and neural circuits

Animals including humans engage in goal-directed behavior flexibly in response to items and their background, which is called contextual behavior in this review. Although the concept of context has long been studied, there are differences among researchers in defining and experimenting with the concept. The current review aims to provide a categorical framework within which not only the neural mechanisms of contextual information processing but also the contextual behavior can be studied in more concrete ways. For this purpose, we categorize contextual behavior into three subcategories as follows by considering the types of interactions among context, item, and response: contextual response selection, contextual item selection, and contextual item–response selection. Contextual response selection refers to the animal emitting different types of responses to the same item depending on the context in the background. Contextual item selection occurs when there are multiple items that need to be chosen in a contextual manner. Finally, when multiple items and multiple contexts are involved, contextual item–response selection takes place whereby the animal either chooses an item or inhibits such a response depending on item–context paired association. The literature suggests that the rhinal cortical regions and the hippocampal formation play key roles in mnemonically categorizing and recognizing contextual representations and the associated items. In addition, it appears that the fronto-striatal cortical loops in connection with the contextual information-processing areas critically control the flexible deployment of adaptive action sets and motor responses for maximizing goals. We suggest that contextual information processing should be investigated in experimental settings where contextual stimuli and resulting behaviors are clearly defined and measurable, considering the dynamic top-down and bottom-up interactions among the neural systems for contextual behavior.

Dynamics of decision-related activity in hippocampus

Place-selective activity in hippocampal neurons can be modulated by the trajectory that will be taken in the immediate future ("prospective coding"), information that could be useful in neural processes elaborating choices in route planning. To determine if and how hippocampal prospective neurons participate in decision making, we measured the time course of the evolution of prospective activity by recording place responses in rats performing a T-maze alternation task. After five or seven alternation trials, the routine was unpredictably interrupted by a photodetector-triggered visual cue as the rat crossed the middle of central arm, signaling it to suddenly change its intended choice. Comparison of the delays between light cue presentation and the onset of prospective activity for neurons with firing fields at various locations after the trigger point revealed a 420 ms processing delay. This surprisingly long delay indicates that prospective activity in the hippocampus appears much too late to generate planning or decision signals. This provides yet another example of a prominent brain activity that is unlikely to play a functional role in the cognitive function that it appears to represent (planning future trajectories). Nonetheless, the hippocampus may provide other contextual information to areas active at the earliest stages of selecting future paths, which would then return signals that help establish hippocampal prospective activity.

Neural correlates of object-in-place learning in hippocampus and prefrontal cortex

Hippocampus and prefrontal cortex (PFC) process spatiotemporally discrete events while maintaining goal-directed task demands. Although some studies have reported that neural activities in the two regions are coordinated, such observations have rarely been reported in an object-place paired-associate (OPPA) task in which animals must learn an object-in-place rule. In this study, we recorded single units and local field potentials simultaneously from the CA1 subfield of the hippocampus and PFC as rats learned that Object A, but not Object B, was rewarded in Place 1, but not in Place 2 (vice versa for Object B). Both hippocampus and PFC are required for normal performance in this task. PFC neurons fired in association with the regularity of the occurrence of a certain type of event independent of space, whereas neuronal firing in CA1 was spatially localized for representing a discrete place. Importantly, the differential firing patterns were observed in tandem with common learning-related changes in both regions. Specifically, once OPPA learning occurred and rats used an object-in-place strategy, (1) both CA1 and PFC neurons exhibited spatially more similar and temporally more synchronized firing patterns, (2) spiking activities in both regions were more phase locked to theta rhythms, and (3) CA1-medial PFC coherence in theta oscillation was maximal before entering a critical place for decision making. The results demonstrate differential as well as common neural dynamics between hippocampus and PFC in acquiring the OPPA task and strongly suggest that both regions form a unified functional network for processing an episodic event.

Hippocampus is required for paired associate memory with neither delay nor trial uniqueness

Cued retrieval of memory is typically examined with delay when testing hippocampal functions, as in delayed matching-to-sample tasks. Equally emphasized in the literature, on the other hand, is the hippocampal involvement in making arbitrary associations. Paired associate memory tasks are widely used for examining this function. However, the two variables (i.e., delay and paired association) were often mixed in paired associate tasks, and this makes it difficult to localize the cognitive source of deficits with hippocampal perturbation. Specifically, a few studies have recently shown that rats can learn arbitrary paired associations between certain locations and nonspatial items (e.g., object or flavor) and later can retrieve the paired location when cued by the item remotely. Such tasks involve both (1) delay between sampling the cue and retrieving the target location and (2) arbitrary association between the cueing object and its paired location. Here, we tested whether delay was necessary in a cued paired associate task by using a task in which no delay existed between object cueing and the choice of its paired associate. Moreover, fixed associative relationships between the cueing objects and their paired locations were repeatedly used, thus involving no trial-unique association. Nevertheless, inactivations of the dorsal hippocampus with muscimol severely disrupted retrieval of paired associates, whereas the same manipulations did not affect discriminating individual objects or locations. The results powerfully demonstrate that the hippocampus is inherently required for retrieving paired associations between objects and places, and that delay and trial uniqueness of the paired associates are not necessarily required.