The medial temporal lobe and prefrontal cortex are known to play a critical role in tasks involving learned associations, such as between an object and a location. However, the exact neural mechanisms underlying the learning and retrieval of high-fidelity spatial associations are unknown. To address this gap, we tested presurgical epilepsy patients with bilateral depth electrodes implanted in the medial temporal lobe and prefrontal cortex on two variants of an object-location spatial learning task. In the first version of this task, subjects were shown a series of objects at random positions along the circumference of an invisible circle. At test, the same objects were shown at the top of the circle, and subjects used a dial to move the object to the location shown during encoding. Angular error between the correct location and the indicated location was recorded as a continuous measure of performance. By registering pre- and post-implantation MRI scans, we were able to localize electrodes to specific hippocampal subfields. We found a correlation between increased high-frequency gamma power (40-100 Hz), thought to reflect local excitatory activity, and the precision of spatial memory retrieval in hippocampal CA1 and dorsolateral prefrontal electrodes. Additionally, we found a directional relationship between these regions, suggesting that the dorsolateral prefrontal cortex is involved in post-retrieval processing. In order to examine the neural activity underlying spatial learning, we then tested presurgical epilepsy patients on a variant of this task in which subjects attempted to learn object-location associations over the course of three training blocks with feedback. At retrieval, we found increased medial temporal and dorsolateral prefrontal gamma power for low error trials, consistent with the results described above. At feedback, we found the opposite pattern of activity, with increased medial temporal and dorsolateral prefrontal gamma power for high error trials. Increased medial temporal gamma activity at feedback also predicted greater decreases in error from one training block the next, indicating that these error signals are involved in updating memory representations or modifying incorrect associations during learning. Finally, we examined the contributions of low frequency oscillatory power to performance on this spatial learning task as well as the relationship between the 1/f aperiodic slope and spatial memory retrieval and learning. We found that the aperiodic slope, thought to reflect the ratio of excitation to inhibition, decreased across training blocks, but did not predict error within training blocks, suggesting that decreased excitation and/or increased inhibition is associated with increased familiarity and/or decreased novelty. Low frequency oscillatory power did not predict error within blocks and did not change across blocks. Overall, these data suggest putative mechanisms for the learning and retrieval of high-fidelity spatial associations.

An estimated 6.1 million Americans currently suffer with Alzheimer’s disease (AD) and in the absence of effective treatment or a cure, this number could increase to 13.8 million by 2060. One of the most common changes that occurs in aging and Alzheimer’s disease is decline in memory for everyday experiences (episodic memory). Computational models suggest that the brain uses a neural computation known as pattern separation to distinguish among highly similar inputs during memory recall. Furthermore, episodic memory has several components: what happened, where it happened, and when it happened. The posteromedial cortices are regions that are thought to be involved in spatial memory and are early sites that are affected by AD pathology. Neuropathological hallmarks of AD include the presence of beta-amyloid (Aβ) plaques, and the formation of neurofibrillary tangles (NFT) and neuropil threads. Eventually, this is followed by neurodegeneration. In vivo candidate biomarkers for AD include pathological aggregations of misfolded proteins (e.g. amyloid plaques, [A], and hyperphosphorylated tau tangles, as measured by CSF assays, and PET imaging), neuronal dysfunction or atrophy (as measured by functional or structural MRI, and total tau levels as measured by CSF assays), and cognitive impairment (as measured by neuropsychological tests). However, the relationship between the pathological, neuroimaging, behavioral, and cognitive markers in the aging brain and in AD remains unclear. The goal of this dissertation was to understand how regions of the brain that support memory for where events occur (spatial memory) are altered in the context of Alzheimer’s disease. We gave participants a spatial pattern separation task, in which we showed participants a series of objects on different locations on a screen and tested participants on their ability to remember the spatial location of the objects after a delay. We found that cortical regions known as the posteromedial network may support spatial pattern separation. Spatial pattern separation task performance was associated with word list delayed recall test commonly used in a clinical setting, a relationship that is conditional upon the level of amyloid burden. Spatial pattern separation task performance was also associated with AD pathology, and cognitive decline. These findings suggest that the pattern separation framework may provide an account for understanding mechanistic changes that occur in the progression of AD. Future studies of spatial memory should investigate associations between regionally specific effects of AD pathology, vascular and immune contributions to AD, and include community-based samples of ethnically diverse populations.

While the current criteria for defining Alzheimer’s disease (AD) primarily consider amyloid pathology, increasing evidence suggests that other key factors, such as cerebrovascular injury and neuroinflammation, also contribute to memory decline among older adults. We asked how white matter hyperintensities (WMH), vascular risk factors, and neuroinflammatory markers impact neurodegeneration and memory dysfunction in the context of aging and AD. Our goal is to enhance our understanding of these non-amyloid pathologies so that they can be integrated into mechanistic models of AD and influence future therapeutic interventions. My first study (Chapter 1) investigated whether WMH and medial temporal lobe (MTL) subregional volumes are linked to memory performance in older adults at risk for AD. I found that posterior WMH, in particular, were associated with hippocampal and rhinal cortical volumes, which in turn predicted memory performance. Importantly, the relationship between WMH and memory was fully mediated by perirhinal cortical volume, indicating that cerebrovascular injury may drive memory decline by exacerbating neurodegeneration in specific MTL subregions. These results were published in Rizvi et al. (2022). In my second study (Chapter 2), I extended my investigation of cerebrovascular risk and pathology to a sample of older adults with Down syndrome (DS), a genetic condition that predisposes adults to AD. Interestingly, individuals with DS are less prone to hypertension but remain vulnerable to cerebrovascular injury. In this sample, I found that higher pulse pressure was associated with increased posterior WMH. Higher pulse pressure was also associated with dementia diagnosis, which was mediated by WMH and neurodegenerative changes. This work highlights the critical role of vascular health in cognitive outcomes for individuals with DS, even in the absence of classic hypertension. These results were published in Rizvi et al. (2024b). In my third study (Chapter 3), I focused on upstream pathologies that may lead to cerebrovascular injury and dysfunction. In particular, I studied the role of neuroinflammation in AD, revealing two distinct pathways involving plasma markers YKL-40 and GFAP. YKL-40 was related to cerebrovascular injury, while GFAP was linked to amyloid-beta (Aβ) deposition. Both pathways converged on tau pathology, leading to MTL atrophy and memory deficits. These findings point to neuroinflammation as a key component of AD, and to the parallel and convergent pathways by which it impacts tau-mediated neurodegeneration and cognitive decline. These results have been submitted for publication and posted to bioRxiv (Rizvi et al., 2024a). Collectively, these findings emphasize the critical contributions of vascular injury, vascular health, and neuroinflammation to neurodegeneration and memory loss in older adults. They suggest that both WMH and inflammatory markers are important predictors of AD-related cognitive decline, with implications for interventions to preserve brain health and mitigate memory impairment in at-risk populations.

The hippocampus’s role in memory was first recognized in 1957 after the bilateral resection of patient HM’s medial temporal lobes left him with profound amnesia; HM’s case study supported hypotheses that the brain possesses multiple memory systems, and suggested the hippocampus is critical for memories of experience. Subsequent lesion cases bolstered additional support, but initially, experimental progress was limited by a lack of technology that could ethically probe hippocampal function in humans, and in animals, was confounded by flawed experimental design and sidetracked by a competing theory about the hippocampus’s role in spatial navigation. Nevertheless, research on navigation became complementary, offering new experimental paradigms also suitable for probing hippocampal memory function and revealing many noteworthy features of hippocampal function like the theta rhythm. Today these research trajectories have begun to coalesce, and now prominent theories presume a more general role of the hippocampus as an index linking arbitrary cortical activation patterns; furthermore, these index representations are thought to relate to the ability of generalizing over or discriminating between similar stimuli via pattern separation/completion processes.Brain electrophysiology is composed of a geometric progression of distinct rhythms, and the theta rhythm (3-8 Hz) has been elegantly linked to navigation processes in the hippocampus of various mammalian species. Additionally, this activity has been associated with memory in mammals, including humans, albeit, primarily with non-invasive methods incapable of attributing this activity to the hippocampus; notably, several studies have also demonstrated that pre-stimulus theta activity is correlated with recognition memory strength. Nevertheless, in the last couple of decades, direct recordings of human hippocampi have been conducted in patients implanted with intracranial EEG (iEEG) prior to brain surgery, but these studies related to memory have provided conflicting results, casting some doubt on the function of hippocampal theta. In this dissertation, I analyze iEEG data collected from human epilepsy patients while engaged in a recollection memory task. I demonstrate that post-stimulus theta during encoding is indeed associated with successful recollection memory. While the first result runs counter to the several iEEG studies on memory, it is consistent with long-standing theories about theta’s role in hippocampal memory processes. Additionally, I provide evidence that pre-stimulus theta activity at encoding is associated with poor recollection memory in the presence of mnemonic interference. Although this result also seems to be in contradiction with all prior accounts of pre-stimulus theta’s effect on memory, it might be reconciled by interpreting the role of pre-stimulus theta as facilitating generalization of memory, rather than memory in general.

ABSTRACT

A Mnemonic Discrimination Account for the Behavioral and Neural Correlates Underlying the Other-Race Effect

byJessica L. Yaros Doctor of Philosophy in Neurobiology & Behavior University of California, Irvine, 2021 Professor Michael A. Yassa, Chair

People often recognize and remember faces of individuals within their own race more easily than those of other races. While behavioral research has long suggested that this so-called Other-Race Effect (ORE) is due to extensive experience with one’s own race group, the neural mechanisms underlying the effect remain unclear. Prior neuroimaging research has explored differences in perceptual processing of same and other-race faces, overlooking the important contributions of mnemonic processes in shaping the ORE. The work comprised in this dissertation attempts to fill this gap, characterizing a framework in which asymmetries in memory mechanisms give rise to the ORE. In a series of experiments employing mnemonic discrimination (pattern separation) paradigms based on computational models of episodic memory, we uncovered both behavioral and neural correlates of the ORE. In an initial set of studies, the ORE was demonstrated to be driven by differences in successful memory discrimination across races as a function of degree of interference between face stimuli. In follow-up studies we characterized different functional brain network properties conducive to successful same and other-race face recognition. Together these findings suggest that the ORE may emerge in part due to tuned memory mechanisms that may enhance same-race, at the expense of other-race face detection. Furthermore, brain network connectivity during memorization that predicted successful same-race recognition was found to be maladaptive for other-race recognition, suggesting employment of different cognitive strategies may be optimal depending on the race of face being processed. Given that cross-race eyewitness identifications disproportionately contribute to wrongful convictions by the criminal justice system, this work should further motivate the development of procedures to mitigate the impact of the ORE on eyewitness testimony.

The medial temporal lobes (MTL) comprise a set of brain regions known to be critical for the formation of memories. Different regions of the MTL support distinct aspects of experience, with the perirhinal and lateral entorhinal cortices comprising a “what” information pathway supporting local features (such as objects), and the parahippocampal and medial entorhinal cortices comprising a “where” pathway supporting configurational features (such as space or context). Importantly, both pathways converge on the hippocampus, where domain-general pattern separation has been hypothesized to occur.. These regions – chiefly perirhinal and lateral entorhinal cortices – are known to be susceptible to the earliest stages of age-related neuropathology. An understanding the distinct contributions of the regions of the MTL will help to elucidate their function in the service of memory, as well as their roles in age-related cognitive decline. However, evidence for functional subdivisions of the human entorhinal cortex, their contributions to hippocampal computations such as pattern separation, and their roles in neurocognitive aging is severely lacking.

In a series of behavioral and functional magnetic resonance imaging (MRI) experience, we sought to fill these gaps in knowledge. First, in a behavioral experiment, we developed a spatial mnemonic discrimination paradigm, modeled after work in rodents studying pattern separation. This complemented prior object discrimination paradigms studying human memory, and showed age-related deficits consistent with a decline in pattern separation ability. Next, in a high-resolution functional MRI study, we used an object vs. spatial mnemonic discrimination paradigm to show dissociable functional roles of lateral and medial entorhinal cortex in human subjects. We next used the object vs. spatial paradigm to investigate whether cognitively normal aging asymmetrically affected performance on the two tasks. We found that older adults are relatively more impaired at object than spatial discrimination, consistent with selective vulnerability of perirhinal and/or lateral entorhinal cortices. Finally, in a high-resolution functional MRI study, we replicated the aforementioned selective object discrimination deficits, and furthermore demonstrated associated hypoactivity in the lateral entorhinal cortex. We additionally replicated prior reports of hyperactivity in the dentate/CA3 subfields of the hippocampus, and furthermore found that the ratio of relative activity between dentate/CA3 and lateral entorhinal cortex predicted the extent of object discrimination impairment. In sum, the studies presented here provide novel evidence for dissociable functional roles of lateral and medial entorhinal cortex in humans, and suggest that the lateral entorhinal cortex-dentate/CA3 circuit is selectively disrupted in aging.