Content uploaded by Antoine Madar
Author content
All content in this area was uploaded by Antoine Madar on Sep 20, 2018
Content may be subject to copyright.
A preview of the PDF is not available
Pattern Separation in the Hippocampus, in Health and Epilepsy
Abstract
Pattern separation is the neuronal computation posited to support our ability to store similar experiences as distinct memory traces. The dentate gyrus (DG) of the hippocampus is generally thought to perform this process, transforming similar cortical input patterns into dissimilar output patterns before they are transferred and stored in downstream hippocampal networks such as CA3. Despite the centrality of this 30-year-old hypothesis to most theories of episodic memory, whether the DG network per se reduces the overlap between similar inputs, and how it performs this computation, has remained a mystery. My doctoral work aims at rigorously assessing different temporal forms of pattern separation by the DG circuitry. Additionally, because the ability of the DG to gate cortical excitation is thought to fail in temporal lobe epilepsy (TLE), leading to seizures, and because TLE patients also suffer from memory impairments that are poorly understood, I investigated the relationship between TLE and pattern separation in the DG.
Using the ability to directly control inputs and measure outputs afforded by slice electrophysiology (in mice), I first showed that the isolated DG circuitry decorrelates non-simultaneous input spiketrains at the level of single granule cells (GCs), the output neurons of the DG. Pattern decorrelation is larger in GCs than in DG interneurons (fast-spiking interneurons and hilar mossy cells) and is amplified in CA3 pyramidal cells. Analysis of the neural noise and computational modelling suggest that this form of pattern separation is not explained by simple randomness and arises from specific presynaptic dynamics.
Second, I investigated other forms of pattern separation by considering diverse neural codes that imply different definitions of the similarity between spiketrains. Results demonstrate that the DG can perform pattern separation using multiplexed coding strategies, and that different celltypes can favor pattern separation through different codes. Pharmacological decrease of inhibitory synaptic transmission showed the importance of fast inhibition in regulating these various computations of the DG network.
I finally examined how TLE affects DG pattern separation. Behavioral experiments in humans and mice first determined that TLE is characterized by deficits in mnemonic discrimination, the cognitive function purportedly supported by pattern separation. Electrophysiological experiments in brain slices showed that temporal pattern separation is also decreased in epileptic mice, notably for a subpopulation of GCs exhibiting pathological spiking patterns.
Overall, my dissertation shows that pattern separation can be performed in the DG through different neural codes, and that the impairment of these neuronal computations is correlated with mnemonic confusion, implicating the DG in TLE as a critical nexus for both seizures and cognitive deficits.
... To reduce interference, the circuit architecture within the hippocampal complex performs certain computations, making memories for overlapping events less similar, the so-called pattern separation (Baker et al., 2016;Burke et al., 2011;Chadwick, Bonnici, & Maguire, 2014;Kesner, 2013;Kirwan & Stark, 2007;Kuhl et al., 2010;Leal & Yassa, 2018;Lohnas et al., 2018;Miller & Sahay, 2019;Motley & Kirwan, 2012;Stark, Kirwan, & Stark, 2019;Xue, 2018). Accumulated evidence from computational models, and from animal and human studies point at a critical involment of CA3 and dentate gyrus (DG) pattern separation computations as the mechanism supporting the resolution of mnemonic interference (Berron et al., 2016;Kesner, 2018;Madar, https://doi.org/10.1016/j.nlm.2020.107177 Received 27 March 2019; Received in revised form 24 January 2020; Accepted 5 February 2020 Ewell, & Jones, 2019a, 2019bReagh & Yassa, 2014;. ...
... Thus, we hypothesize that memory deficits in these patients could be accounted for by an altered pattern separation process, thus affecting the ability to differentiate among similar events, and increasing the susceptibility to interference (Bekinschtein et al., 2013;Saksida & Bussey, 2010). Surprisingly, research on mnemonic discrimination in mTLE is exceedingly rare (Madar et al., 2016;Madar, 2018;Ramezani, 2017;Yim et al., 2015), and there is only one previous study with human patients (Reyes et al., 2018). In the current study, we aim to investigate whether alterations in this process could be at the base of memory difficulties in this group of patients. ...
... We found that patients performed worse than controls in the conditions in which the studied item had to be discriminated from a physically similar new object from the same basic-level category. Previous studies also found impaired mnemonic discrimination between slightly different items in patients with structural and functional abnormalities in the hippocampus (Baker et al., 2016;Bakker et al., 2012;Bayley et al., 2008;Brock Kirwan et al., 2012;Duff et al., 2012;Hanert, Pedersen, & Bartsch, 2019;Yassa et al., 2010), and more specifically in patients with TLE (Madar, 2018;Reyes et al., 2018). Crucially, our results revealed that reliable differences between groups were specifically observable in the conditions whereby more exemplars from a category were stored in memory. ...
Mesial temporal lobe epilepsy (mTLE) is a neurological disorder associated with histopathological changes in different subfields of the hippocampus. These alterations have been associated with memory difficulties. In this study, we tested the hypothesis that these difficulties stem on mnemonic discrimination impairment due to a reduced ability to make similar representations more distinct, leading to an increased susceptibility to interference. With this aim, we used a visual mnemonic discrimination task and evaluated the ability of a group of patients with unilateral mTLE, relative to controls, to discriminate between a studied item and a new foil item, as a function of the similarity between them, and of the number of exemplars from a category stored in memory. We found that patients performed worse than controls when the studied item had to be discriminated from a physically similar new object from the same basic-level category. Crucially, reliable differences between groups were observable in the conditions in which more exemplars from a category were held in memory. In the conditions in which the studied item had to be discriminated from a foil from a different basic-level category, there were no differences between groups, with one exception. Neither a general cognitive impairment nor a general memory impairment could account for this pattern of results. Current findings indicate that patients found more difficulties in conditions with higher interference, which poses greater demands for pattern separation. A disruption of pattern separation processes resulting from hippocampal damage provides a reasonable interpretation for these results. Future studies should explore the causal relationship between hippocampal subfields integrity and mnemonic discrimination capacity in mTLE patients.
... Rather than a deterioration in the memory representation of the environment, such remapping may serve a function and has been proposed to encode distinct episodes that occur in the same environment separated by time or small changes in environment 10,37,39,41,[58][59][60][61][62] . This remapping might therefore allow the animal to create independent representations of different episodes associated with the same environment and favor mnemonic discrimination of those episodes 63,64 . Our data support this idea as the average PF correlation in CA1 across days in a novel environment was 0.49 ± 0.02 (see methods) revealing considerable remapping across days (Fig. 4). ...
When exploring new environments animals form spatial memories that are updated with experience and retrieved upon re-exposure to the same environment. The hippocampus is thought to support these memory processes, but how this is achieved by different subnetworks such as CA1 and CA3 remains unclear. To understand how hippocampal spatial representations emerge and evolve during familiarization, we performed 2-photon calcium imaging in mice running in new virtual environments and compared the trial-to-trial dynamics of place cells in CA1 and CA3 over days. We find that place fields in CA1 emerge rapidly but tend to shift backwards from trial-to-trial and remap upon re-exposure to the environment a day later. In contrast, place fields in CA3 emerge gradually but show more stable trial-to-trial and day-to-day dynamics. These results reflect different roles in CA1 and CA3 in spatial memory processing during familiarization to new environments and constrain the potential mechanisms that support them. To understand how spatial representations emerge and evolve across hippocampal subfields, we compared trial-to-trial dynamics of place cells in CA1 and CA3 in new environments and across days. CA1 place fields form early, shift backwards and partially remap across days whereas in CA3 they develop gradually and are more stable, suggesting distinct functional roles in representing space.
ResearchGate has not been able to resolve any references for this publication.