The hippocampus is essential for encoding new memories for space and events. In the rodent hippocampus, a result of this process is that individual neurons fire selectively when the animal is located in a particular subregion of its environment. We show that that hippocampal area CA1, but not upstream area CA3, emits twice as many spikes in novel as compared to familiar environments, and during the initial stages of novelty many CA1 neurons are active. The overall population rate and the number of active cells decreases as the environment becomes familiar, but the decline in rate is not uniform across neurons. Instead, the activity of cells with peak spatial rates above ~12 Hz is enhanced, while the activity of cells with lower peak rates is suppressed. The result is that, after several days of experience in the environment, the active CA1 population consists of a relatively small group of cells with strong spatial tuning. This process is not evident in CA3, suggesting that a region specific selection process operates in CA1 to create a sparse, spatially informative population of neurons. We also show that there is clear and long lasting replay of experiences from one place during waking behavior in subsequently experienced places. Surprisingly, this remote replay was most robust when the animal had recently been in motion as compared to during extended periods of immobility. These results indicate that waking and sleep replay both constitute a form of memory retrieval where past experiences can be reactivated, and that patterns of activity associated with memory consolidation continue unabated during both sleep and awake states.

Stressful experiences lead to important changes in physiology and behavior, including the release of corticosteroid stress hormones and engagement of vigilance behaviors. While normally important for survival, these aspects of the stress response can exert detrimental effects if they are dysregulated and become exaggerated or blunted. The ventral hippocampus (vHPC) has long been depicted as an inhibitory regulator of the corticosteroid stress response, part of a negative feedback loop that controls the size of stress responses. Downstream, the hypothalamic CRH-expressing neurons are known to be the gatekeeper of the cortisol stress response, but recent studies also implicate a non-hormonal role for these cells in anxiety-related behavior. What is still unknown is how these two regions respond in the same animals to different stressors like approach-avoidance assays used in the study of anxiety-related behavior, and how the ventral hippocampus may be involved in regulating CRH activity in these behaviors. In Chapter 1, I introduce the hippocampal-hypothalamic circuitry that is implicated in emotion-related behaviors from the stress and anxiety fields. I discuss how this indirect projection has been modelled as a regulator of the neuroendocrine stress response, with respect to the corticosteroid response. I also introduce more recent findings that imply this pathway could also be involved in non-neuroendocrine stress responses through direct control of stress-induced behaviors. In Chapter 2, we review recent work on how vCA1 contributes to a network that associates external stimuli with internal motivational drive states to promote the selection of adaptive behavioral responses. This leads us to propose a model of vHPC function that emphasizes its role in the integration and transformation of internal and external cues to guide behavioral selection when faced with multiple potential outcomes. In Chapter 3, I present evidence that PVNCRH and vHPC cells respond independently to fear- and anxiety-related stimuli. Immediate, uncontrollable threats and freezing responses to these threats evoke activity in both regions. However, this is not true for exploratory behaviors in an approach-avoidance assay and self-controlled transitions between levels of potential threat; these events reliably evoke activity only in vHPC cells. In Chapter 4, I investigate the influence of the vHPC on PVNCRH cells and their responses to stressful stimuli. I establish the first in vivo recordings of PVNCRH cells during vHPC chemogenetic perturbation. By inhibiting vHPC activity during stressful experiences, I find that vHPC selectively modulates activity of PVNCRH cells during a subset of behavioral responses to immediate, uncontrollable threats and ambiguous, self-controlled threat. In Chapter 5, I integrate these experimental findings with knowledge in the field about the roles of vHPC and the PVNCRH cells and the regions’ contributions to engaging and controlling the stress response as a whole.

An important goal of neuroscience is to establish principles that can guide comparative studies across species. If we specifically compare rodents and macaques, there are fundamental differences in motor network anatomy. From a functional perspective, there are also a growing number of studies which indicate that in rodents, extensive skill training can lead to a ‘disengagement’ of M1 from movement control; after a period of initial deficits, movements appeared to be identical and relied only on subcortical structures. However, classic studies indicate that cortical networks are essential for prehension and perhaps less important for proximal control. However, the exact amount of task training was not clear and detailed kinematics were not performed. Thus, it remains unclear precisely how gross and fine motor control might change with a M1 lesion in primates. Here we aimed to measure changes in skilled gross and fine motor control after a M1 lesion in both macaques and mice. In macaques, we ensured that animals were well trained; we also independently measured performance in a gross motor skill and a reach-to-grasp skill. In mice, we also performed long-term monitoring of layer 5 M1 projection neurons, i.e., pathways we know are more likely to reflect movement control signals from M1. Across species, we performed detailed kinematic monitoring to quantify changes in performance, kinematic variability and transitions between sub-movements. Together, our results indicate that there is a common and preserved functional principle after the loss of M1. Even in primates, there is rapid restoration of gross movement control after a period of deficits; strikingly similar to what is reported for rodents. In contrast, for a task involving prehension, we noted prolonged deficits in prehension and changes in the transition reliability of reach to grasp. Lastly, we also do not find any evidence of neural disengagement when considering layer 5 projection neurons, pathways that are known to be critical for reach to grasp skills. Interestingly, the shared trajectory post-lesion across the two species underscores a commonality in the disruption of smooth transition probabilities that is replaced by a mosaic of fragmented movements during a task that is comprised of reaching and prehension.

A major question in motor systems neuroscience is how complex actions are encoded, on the timescale of both shorter individual elements and longer sequences. Birdsong is a tractable system, where a learned complex vocal behavior combines a categorical set of shorter individual elements into longer sequences, making it well-suited to address this question. The song nucleus HVC (used as a proper name) contributes to song sequence and timing. While much has been studied about HVC in zebra finches, which sing linear, stereotyped songs through their adult lives, relatively little data has been collected from HVC in their close relatives, Bengalese finches, that sing flexible, variably sequenced songs. We built a custom microscope to record neural activity reported by the calcium indicator GCaMP from populations of neurons from HVC in awake, freely moving Bengalese finches. We analyzed how populations of neurons in HVC encode information around divergence points, where one syllable can be followed by multiple syllables, and convergence points, where one syllable can be preceded by multiple syllables. We found that HVC projection neuron bursting can encode for upcoming sequence many syllables ahead of divergence points, and prior sequence many syllables after convergence points. We also found that HVC bursting encodes variation in the acoustic structure (phonology) of the different renditions of a given syllable in different contexts. Moreover, we found that HVC has overlapping representations of distinct syllables, especially those which are acoustically similar. These results help to reveal how premotor regions can encode multiple types of sequence and phonological information simultaneously.

Studies of cortical visual processing in mammals have traditionally focused on the visual pathway that ascends from the retina to primary visual cortex (V1) via the lateral geniculate nucleus of the thalamus (LGN), known as the geniculate pathway. However, visual information can also reach cortex through an alternate “extrageniculate” visual pathway in which retinal projections to the superior colliculus (SC) are relayed through the pulvinar nucleus of the thalamus (PN; also called the lateral posterior nucleus in rodents) before radiating to various regions of the visual cortex. While this extrageniculate pathway has been extensively characterized in primates and cats, its contribution to visually evoked activity in higher-order visual cortices is generally considered of lesser consequence than that of the geniculate pathway, at least in those higher-order visual cortices studied thus far. In the mouse, however, recent experiments have demonstrated that the SC is actually the principal driver of visually evoked activity in a higher-order visual area called postrhinal cortex (POR). The mouse has about ten other higher cortical visual areas whose visual responses are otherwise considered to largely rely on the retino-geniculo-V1 pathway. In this thesis, I use anatomical and functional approaches to determine the extent to which visual evoked responses across all higher visual areas of the mouse rely on the SC, and in particular on the extrageniculate retino-tecto-pulvinar pathway. I characterize the anatomical distribution of cortical projections from the SC through PN using retrograde, anterograde, and transsynaptic tracing methods, and assess their functional properties using a variety of circuit perturbations in combination with widefield calcium imaging and electrophysiology. This work reveals a lateromedial gradient of SC-dependency across the mouse visual cortex, with several lateral visual areas inheriting most of their visual responses from the SC via PN. We further demonstrate a potential role for this SC-dependent visual cortex in distinguishing self from externally generated visual motion. Together, these lateral cortices constitute a functionally distinct tecto-thalamic visual cortical system operating in parallel with the more medial canonical geniculo-striate system.

The brain’s ability to associate experiences with subsequent rewards is fundamental to learning and memory and critical for animal survival. The neural substrates of this process are only partially understood, but are thought to rely on interactions between the hippocampus and nucleus accumbens (NAc). In particular, hippocampal input to the NAc is thought to be crucial for learning and remembering links between spatial information and reward. Hippocampal projections to the NAc arise from both the ventral hippocampus (vH) and the dorsal hippocampus (dH), and studies using optogenetic interventions have demonstrated that either vH or dH input to the NAc can support behaviors dependent on spatial-reward associations. It remains unclear, however, whether dH, vH, or both coordinate memory processing of spatial-reward information in the hippocampal-NAc circuit under normal conditions. Moreover, as dH and vH are thought to encode different aspects of an experience, whether the hippocampus can compartmentalize different types of information to circuits in the NAc is unknown. Times of memory reactivation within and outside the hippocampus are marked by hippocampal sharp-wave ripples (SWRs), discrete events which facilitate investigation of inter-regional information processing. It is unknown whether dH and vH SWRs act in concert or separately to engage NAc neuronal networks, and whether either dH or vH SWRs are preferentially linked to spatial-reward representations. To address these questions, we performed simultaneous extracellular recordings using multi-tetrode arrays in the dH, vH, and NAc of rats learning and performing an appetitive spatial task and during sleep. We report that dH and vH SWRs occur asynchronously, and that individual NAc neurons activated during SWRs from one subdivision of the hippocampus are typically suppressed or unmodulated during SWRs from the other. Furthermore, NAc neurons activated during dH versus vH SWRs show markedly different task-related firing patterns, with NAc representations related to space and reward selectively activated during dH SWRs and not vH SWRs. Our findings reveal that dorsal and ventral hippocampal interactions with the NAc are temporally and anatomically separable at times of memory processing. This work suggests that the dH-NAc and vH-NAc networks provide distinct information channels, with the dH-NAc channel dedicated to linking spatial paths with reward and reward-seeking actions. More broadly, these circuit dynamics could provide a potential neural substrate for the brain’s ability to compartmentalize aspects of experience in memory.

In novel situations, animals can leverage past experiences to learn rapidly. This ability is thought to depend on abstraction: the representation of the common structure across related experiences. In mammals, the hippocampus (HPc) and the prefrontal cortex (PFC), including its medial prefrontal (mPFC) and orbitofrontal (OFC) subregions, are thought to support abstraction by expressing neuronal firing patterns that represent generalized features of experiences. Whether these firing patterns reflect a single, distributed representation of generalized features of experience, or whether each area specializes to represent particular features at particular times, remains unknown. To address this, we continuously monitored large neural ensembles in the HPc, mPFC, and OFC of freely behaving rats performing a cognitive task. We found evidence for regional specialization in the coding of generalized task features. First, HPc firing patterns were consistent with a primarily route-based coding scheme, whereas mPFC and OFC firing patterns were organized around the act of traveling between goal locations and the specific actions required to reach goals. Second, task representations in mPFC and OFC were most reliable during distinct task phases, suggesting these areas specialize to express consistent task representations in distinct behavioral periods.

The hippocampal formation is a part of the brain that is essential for everyday learning and memory, including the ability to remember places and scenes. The mnemonic function of the hippocampal formation can be experimentally investigated in rats using behavioral paradigms that engage spatial learning and memory. Such behavioral paradigms have compelling neural correlates, as neurons in the hippocampal formation fire in response to spatial locations and environmental contexts. Here I present two significant original contributions to our understanding of the cognitive function of the hippocampal formation and the neurophysiology that underlies this function.

The first part of this dissertation is a behavioral study of the effects of hippocampal lesions on learning of the W-maze continuous spatial alternation task. The W-maze task is a test of spatial working memory and rule-learning. Neurons in the hippocampal formation exhibit task-relevant activity during performance of this task. However, previous to this study, it was not known whether learning of the W-maze task really depends on the hippocampal formation. I found that rats with excitotoxic lesions of the hippocampal formation made unusual perseverative errors and were significantly slower to learn the W-maze task than sham-operated controls. This finding suggests that the hippocampal formation contributes to rapid learning of spatial trajectories that lead to reward.

The second part of this dissertation is a single-unit recording study of the subiculum. The subiculum is a region within the hippocampal formation that has received little previous investigation, even though it is a major output structure through which information from the hippocampal formation reaches the rest of the brain. I recorded spikes and local field potentials in the subiculum while rats ran in two environments. I found that neurons in the subiculum provide a highly informative representation of the animal's spatial location and environmental context, and that the sparseness of this spatial representation exhibits a gradient along the proximal-distal anatomical axis. Additionally, I discovered that neurons in the subiculum exhibit theta phase precession, an oscillatory phase coding phenomenon that is thought to be important for coordinating information transfer and spike timing-dependent plasticity.