The persistence of memory

Glas, Annet (2021): The persistence of memory: the effect of spaced training on memory and neuronal ensemble stability. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)

Abstract

Distributing learning in time has the remarkable ability to enhance memory in a wide range of species and behavioral paradigms, a phenomenon termed the spacing effect. An extensive body of scientific work provides insight into the molecular and cellular processes that underlie the spacing effect. However, it is unclear how trial spacing alters the activity of the neuronal populations that store a specific memory. With my work presented in this doctoral dissertation, I explored the relationship between trial spacing, memory strength, and the pattern of in vivo neuronal activity. To achieve this aim, I executed two initial studies to address two outstanding methodological concerns. In the first study, I describe two nutritional restriction methods that balance mouse well-being and behavioral performance on an operant conditioning task. Nutritional restriction can be achieved by either food or fluid restriction and is typically necessary to ensure task engagement in mice. However, these procedures can have detrimental effects on mouse welfare if not executed diligently. I monitored the the effect of food or water restriction on mouse welfare as well as performance on a head-fixed two-choice visual discrimination task. In this study, both restriction regimen resulted in similar maximum learning performance while mouse discomfort was typically sub-threshold, providing a blueprint to the wider neuroscientific community to carry out similar experiments. In the second study, I compare a novel in vivo microscopy technique with the current golden standard for in vivo imaging of individual neurons, which is two-photon microscopy. Imaging using a miniaturized epifluorescence microscope is an efficient and effective approach to image hundreds of neurons while a mouse is engaged in a freely moving behavioral task, but does not achieve the same lateral and axial resolution as two-photon imaging. I performed in vivo calcium imaging of mouse primary visual cortex neurons expressing genetically encoded calcium indicators using both microscopy techniques while mice were presented with drifting gratings. I demonstrated that the response properties and tuning features of mouse visual cortex neurons to gratings of different orientations were quantitatively comparable in spite of qualitative differences between the two imaging methods. In the third and main study, I explore whether trial spacing affects memory strength and in vivo activity of a population of individual neocortical neurons. I addressed this question by examining two non-mutually exclusive hypotheses. Trial spacing could enhance the selective strengthening of the connections between neurons that store a preexisting memory. This would stabilize this neuronal ensemble, which would allow for more precise ensemble reactivation and thereby more effective memory retrieval. Alternatively, trial spacing could affect the recruitment of additional neurons that store new information from subsequent trials. This would increase the size of the neuronal ensemble, which would make the stored memory more resilient to destabilization. To explore these hypotheses, I trained mice on an everyday memory task, a delayed matching-to-place task that instilled episodic-like memories. Trial spacing promoted memory retrieval, yet surprisingly impaired memory encoding. Simultaneously, I measured neuronal activity using a miniaturized microscope in the dorsomedial prefrontal cortex, a neocortical structure that stored these memories as evidenced by the amnesic effect of chemogenetic inhibition of the dorsomedial prefrontal cortex during training. Trial spacing promoted reactivation of the neuronal ensemble but did not affect the size of the neuronal ensemble, thereby providing the first direct observation of the effect of trial spacing on the activity of neurons in the intact mammalian brain. In summary, the work presented in this doctoral dissertation used modern neuroscientific methods to study whether altered neuronal ensemble characteristics underlie the spacing effect, a phenomenon that was first described over a century ago.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: spacing effect, learning, memory, mouse, neuronal ensemble
• Themengebiete: 500 Naturwissenschaften und Mathematik ; 500 Naturwissenschaften und Mathematik > 570 Biowissenschaften, Biologie
• Fakultäten: Graduate School of Systemic Neurosciences (GSN)
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 26. April 2021
• 1. Bericht­erstatter:in: Goltstein, Pieter