Forty-eight rat pups were handled and isolated from postnatal days 3 through 13 in order to determine whether this manipulation would alter the postnatal development of the hippocampus. Half of these animals were then reared in enriched environments from weaning until maturity to determine whether enrichment would ameliorate the expected deficits in learning ability. Beginning at 90 days of age, all animals were tested on a T-maze, rotating bar and both place and cued versions of a water maze task. The study failed to find gross deficits in learning as a result of the handling/isolation procedure, although emotional differences between groups was evident, as were sex differences. Apparently more questions have been raised than answered by this study, and possible directions for future research are discussed.

This thesis aims to explore the two dominant functional roles of the hippocampal formation, in the relational encoding of episodic memory and the neural representation of allocentric space, using a combination of pharmaceutical manipulations and single-unit recording techniques in rodents. The first part of this thesis focuses on episodic-like memory, defined by the original episodic memory triad: ‘what-where-when’ (Tulving 1972), which enables the behavioural aspects of episodic memory to be tested in non-human animals. Permanent neurotoxic lesions of the hippocampus and it’s subregions were induced to assess their role in a putative episodic-like memory task developed by Eacott and Norman (2004). In view of the difficulties encountered in successfully demonstrating the temporal component of episodic-like memory in rats, this task tested integrated memory for ‘what-where-which’, where the temporal component (when) was replaced with another event specifier: context (on ‘which’ occasion). Disruption of the hippocampal circuitry led to a specific impairment in the integration of all three event components, whereas the associative recognition of any combination of these features in isolation was left intact. These results confirm the hippocampal dependence of this episodic-like memory task and further reveals the necessity of both CA3 and CA1, hypothetically due to the underlying autoassociative role of CA3 with CA1 functioning as the vital output pathway for this associated information and/or as a mismatch detector. There has been much debate over the inclusion of the temporal component and sceptics may argue that any such interpretations of task-dependence on episodic-like memory processing are invalid considering the requirement for temporal processing is absent. Due to the proposal that a temporal framework necessarily provides the foundation on which episodic memories are built, the second chapter focuses on the development of a suitable protocol in which integrated memory for the original ‘what-where-when’ episodic memory triad can be reliably tested. The other main function attributed to the hippocampus was brought to light by the fascinating revelation that it’s neurons selectively fire in different regions of an environment, termed ‘place cells’ (O’Keefe and Dostrovsky 1971). From the numerous publications resulting from this discovery it has emerged that place cells not only respond to the spatial features of the environment but are also sensitive to a multitude of non-spatial features. These characteristics support the logical assumption that the primary firing patterns of the hippocampus should underlie it’s main purported roles, leading to speculations that they reflect episodic memory processes. The second part of this thesis aims to examine the relationship between hippocampal cells and behaviour by extending the work of Ainge et al. (2007a), in which a subset of hippocampal place cells were found to encode both current and intended destination in a double Y-maze ‘win-stay’ task. The development of these ‘goal-sensitive’ cells were initially investigated during the learning phase of this task. An exciting pattern of results showed a strong positive correlation between the emergence of goal-sensitive firing and behavioural performance on the task, tempting speculations that these firing patterns may underlie spatial learning and future planning, necessary to support performance. To ensure these firing patterns were not a mere reflection of greater experience on the maze, a second study was conducted in which the task demands changed over set periods of days. A significant increase in the proportion of cells demonstrating goal-sensitive firing was revealed when the protocol shifted to incorporate the spatial memory demands of the ‘win-stay’ task, with all other parameters of the protocol remaining constant. These results support the theory that goal-sensitive firing patterns are specifically related to the learning and memory demands of the spatial task, not a result of increased exploration of the maze. The last of this series of studies assessed hippocampal-dependence of this task and revealed that bilateral hippocampal lesions induced an impairment in spatial ‘win-stay’ performance. Collectively, these experiments demonstrate that goal-sensitive firing of hippocampal cells emerge in line with behavioural performance in a hippocampal-dependent task and the emergence of these firing patterns are specific to the learning and memory demands of a spatial ‘win-stay’ protocol. The functional role of the hippocampus in allocentric spatial processing may thus underpin it’s function in episodic memory and potentially in the imagining and planning of future events, whereby the hippocampus provides a ‘space’ in which retrieved information can be integrated in a coherent context to support the fluent and flexible use of information. This hippocampal function would necessarily require visual information to be accessed, concerning the arrangement of landmarks and cues within the environment, in association with information regarding internal orientation and direction and this leads to the question assessed in the final part of this thesis of where this integration occurs. Based on anatomical evidence and the current literature, the postsubiculum, an input structure to the hippocampus, emerged as a potential site for the convergence of sensory cues into the internally generated head direction cell and place cell networks to enable hippocampal-dependent spatial processing. Thus, the effects of temporary pharmacological blockade of AMPARs and NMDARs in the postsubiculum were assessed on the encoding of spatial memory in an object recognition paradigm. The impairment revealed in the ability to recognise novel object-place configurations demonstrates a key role for NMDAR-dependent plasticity within the postsubiculum itself in the formation of allocentric spatial memory. In summary, the experimental results reported in this thesis further elucidate the critical role the hippocampal formation plays in spatial and episodic memory by combining evidence from cellular physiology and neuroanatomy to the behaving animal and extends these findings to discuss a more general role for the hippocampus in imagining both past and future events, in order to successfully navigate, learn and enable past experience to influence our intended future plans and decisions.

published_or_final_version / abstract / Pharmacology / Master / Master of Philosophy

published_or_final_version / Biological Sciences / Master / Master of Philosophy

published_or_final_version / Biological Sciences / Master / Master of Philosophy

A positive correlation exists between increasing age and the incidence of hippocampal-associated dysfunction and disease. Normal L-glutamate neurotransmission is absolutely critical for hippocampal function, while abnormal glutamate neurotransmission has been implicated in many neurodegenerative diseases. Previous studies, overwhelmingly utilizing ex vivo methods, have filled the literature with contradicting reports about hippocampal glutamate regulation during aging. For our studies, enzyme-based ceramic microelectrode arrays (MEA) were used for rapid (2 Hz) measurements of extracellular glutamate in the hippocampal trisynaptic pathway of young (3-6 months), late-middle aged (18 mo.) and aged (24 mo.) urethane-anesthetized Fischer 344 rats. Compared to young animals, glutamate terminals in cornu ammonis 3 (CA3) showed diminished potassium-evoked glutamate release in aged rats. In late-middle aged animals, terminals in the dentate gyrus (DG) showed increased evoked release compared to young rats. The aged DG demonstrated an increased glutamate clearance capacity, indicating a possible age-related compensation to deal with the increased glutamate release that occurred in late-middle age. To investigate the impact of changes in glutamate regulation on the expression of a disease process, we modified the MEA technology to allow recordings in unanesthetized rats. These studies permitted us to measure glutamate regulation in the hippocampal formation without anesthetic effects, which showed a significant increase in basal glutamatergic tone during aging. Status epilepticus was induced by local application of 4-aminopyridine. Realtime glutamate measurements allowed us to capture never-before-seen spontaneous and highly rhythmic glutamate release events during status epilepticus. A significant correlation between pre-status tonic glutamate and the quantity of status epilepticus-associated convulsions and glutamate release events was determined. Taken together, this body of work identifies the DG and CA3 as the loci of age-associated glutamate dysregulation in the hippocampus, and establishes elevated levels of glutamate as a key factor controlling severity of status epilepticus in aged animals. Based upon the potential ability to discriminate brain areas experiencing seizure (i.e. synchronized spontaneous glutamate release) from areas not, we have initiated the development of a MEA for human use during temporal lobe resection surgery. The final studies presented here document the development and testing of a human microelectrode array prototype in non-human primates.

An overarching goal of the Gerhardt laboratory is the development of an implantable neural device that allows for long-term glutamate recordings in the hippocampus. Proper L-glutamate regulation is essential for hippocampal function, while glutamate dysregulation is implicated in many neurodegenerative diseases. Direct evidence for subregional glutamate regulation is lacking in previous in vivo studies because of limitations in the spatio-temporal resolution of conventional experimental techniques. We used novel enzyme-coated microelectrode arrays (MEAs) for rapid measurements (2Hz) of extracellular glutamate in urethane-anesthetized rats. Potassium-evoked glutamate release was highest in the cornu ammonis 1 (CA1) subregion and lowest in the cornu ammonis 3 (CA3). In the dentate gyrus (DG), evoked-glutamate release was diminished at a higher potassium concentration but demonstrated faster release kinetics. These studies are the first to show subregion specific regulation of glutamate release in the hippocampus. To allow for in vivo glutamate measurements in awake rats, we have adapted our MEAs for chronic use. Resting glutamate measurements were obtained up to six days post-implantation but recordings were unreliable at later time points. To determine the cause(s) for recording failure, a detailed investigation of MEA surface characteristics was conducted. Scanning electron microscopy and atomic force microscopy showed that PT sites have unique surface chemistry, a microwell geometry and nanometer-sized features, all of which appear to be favorable for high sensitivity recordings. Accordingly, studies were initiated to improve enzyme coatings using a computer-controlled microprinting system (Microfab Technologies, Plano, TX). Preliminary testing showed that microprinting allowed greater control over the coating process and produced MEAs that met our performance criteria. Our final studies investigated the effects of chronic MEA implantation. Immunohistochemical analysis showed that the MEA produced minimal damage in the hippocampus at all time points from 1 day to 6 months. Additionally, tissue attachment to the MEA surface was minimal. Taken together with previous electrophysiology data supporting that MEAs are functional up to six months, these studies established that our chronic MEAs technology is capable of maintaining a brain-device interface that is both functional and biocompatible for extended periods of time.

The conversion of a memory from an initially fragile state to an enduring representation requires cellular, molecular, and systems-level brain network changes. This reorganization is hypothesized to involve time-dependent neuroanatomical changes that may differentially support some types of remote versus recent memory, and may also influence the latency to decide and complete responses during retrieval. To quantify the timecourse of learning and retrieval after different retention durations, a paradigm is developed to measure in humans and monkeys the retrieval speed of visuomotor associations, which require an intact hippocampus for initial acquisition but not for retrieval after days or weeks. Two components of retrieval speed, a decision time to initiate movement and a velocity-dependent movement completion time to complete a motor response, are shown to change differently relative to a pre-retention baseline. Movement completion times decrease across repetitions within single learning session, and continue to decrease from the level reached at the end of learning following retention. Decision times also decrease within the learning session, but increase on the first post-retention retrieval attempt as a function of retention interval duration. Extensive practice is required for decision times to reach a level below that obtained at the end of learning, and the transition from a long- to short-latency decision depends on the number and spacing of practice trials. The findings are discussed in a framework in which post-retention processing time is influenced by the speed of visual identification, the time to retrieve the associative relationship from long-term memory, and the time to plan and execute a motor response. The creation of sparser, long-lasting visual form representations and strengthened cortico-striatal connections predict behavioral efficiency gains in visual identification and motor responses after learning. Decision times could be fast and automatic following extensive practice when the neural representation may become stored permanently in cortico-cortical and cortico-striatal linkages, or could increase after retention because of several cognitive and neural factors, including interference and frontal inhibition of the hippocampal system to prevent new learning before choice feedback. The experimental results are discussed in the context of the existing literature on memory consolidation.

We investigated licking and grooming (LG) levels received by each pup from their dams and the locomotor activity, anxiety-like behaviors, and stress reactivity in adult female offspring. We also investigated glucocorticoid receptor (GR) gene expression and its DNA methylation status in the hippocampus, comparing pups between and with-in litters. Rats that receive more LG than their siblings showed less anxiety-like behaviors and increased locomotor activity, regardless of their litter type. Higher licked pups also showed increased expression of the GR gene. Gene expression levels of the GR 17 splice variant were not significantly different as a function of dam LG or LG received, whereas DNA methylation levels at two CpG sites within GR17 promoter were significantly higher in high LG pups than low LG pups. Our results indicate that naturally occurring intra- and inter-litter differences in maternal LG have a lasting effect on the phenotypic outcomes of adult female offspring.

We investigated licking and grooming (LG) levels received by each pup from their dams and the locomotor activity, anxiety-like behaviors, and stress reactivity in adult female offspring. We also investigated glucocorticoid receptor (GR) gene expression and its DNA methylation status in the hippocampus, comparing pups between and with-in litters. Rats that receive more LG than their siblings showed less anxiety-like behaviors and increased locomotor activity, regardless of their litter type. Higher licked pups also showed increased expression of the GR gene. Gene expression levels of the GR 17 splice variant were not significantly different as a function of dam LG or LG received, whereas DNA methylation levels at two CpG sites within GR17 promoter were significantly higher in high LG pups than low LG pups. Our results indicate that naturally occurring intra- and inter-litter differences in maternal LG have a lasting effect on the phenotypic outcomes of adult female offspring.