A Multimodal Neuroimaging Investigation into the Modulation of Motor Memory Consolidation By Stress

Retaining newly acquired skills is a fundamental capacity that underlies the elaborate and complex set of human motor behaviours. This capacity is supported by the process of memory consolidation. Although the neural correlates of motor memory consolidation have extensively been characterized, it remains unknown whether the recruitment of the relevant neural substrates can be modulated in order to influence consolidation processes. In the current project, we investigated whether experimentally-induced stress can modulate the behavioural (Study 1) and neural (Study 2) correlates of motor sequence memory acquisition and consolidation in healthy young adults. To do so we used an extensive multimodal neuroimaging approach including magnetic resonance imaging (MRI) and spectroscopy (MRS).

The current project is based on two key findings. Firstly, evidence that in addition to the striatum, the hippocampus is involved in motor memory. More specifically, the observation that the balanced recruitment of hippocampal and striatal networks during motor memory encoding is a pre-requisite for subsequent consolidation over sleep. Our goal was therefore to modulate activity in, and interaction between, the hippocampus and the striatum during initial motor learning in order to influence the subsequent consolidation process. The second key finding is the observation that experimentally-induced stress modulates activity in the hippocampus and the striatum. In the current project, participants were exposed to a stressful (the Socially Evaluated Cold Pressor Test, SECPT) or non-stressful control intervention prior to initial task exposure. Our overarching hypothesis was that exposure to stress prior to initial memory acquisition would challenge the recruitment of the hippocampus and bias the balance between hippocampal and striatal systems during initial motor learning. Consequently, given the link between these neural signatures and sleep-dependent consolidation, we hypothesized that these stress-induced modulations in brain responses would forecast a disruption of the subsequent sleep-related consolidation process.

As a model of daily motor performance, we used a motor sequence learning (MSL) task in which participants learned bimanual sequences of finger movements by repeated practice. In each study, motor performance was assessed during initial training as well as during a retest session taking place after a period of sleep. At the behavioral level, our findings indicated that stress induced prior to initial training did not modulate the behavioural correlates of motor sequence memory acquisition nor consolidation (Chapter 2). To investigate the effect of stress on the neural correlates of motor sequence memory, we measured task-related brain activity, using functional MRI, during task practice (Chapter 3). Findings of the fMRI study indicated that as compared to the control group, stress prior to initial training induced an increase in activity in sensory and motor-related cortical regions during initial learning. Moreover, stress resulted in a progressive disengagement of the hippocampus as well as a disconnection between the hippocampus and sensorimotor cortical regions as a function of practice. While these patterns of stress-induced changes in brain responses were predictive of faster performance at the end of training, they were negatively related to subsequent changes in performance after the consolidation interval. The results of Chapter 3 thus provide the first experimental evidence that stress modulates brain responses involved in motor memory consolidation processes. Next, we used MR spectroscopy (MRS) to assess the level of the main inhibitory neurotransmitter (gamma-aminobutyric acid, GABA) in the striatum and hippocampus before the intervention as well as immediately after initial training (Chapter 4). We showed that while neither learning nor stress influenced GABA+ level in the striatum at the group level, larger learning-related decreases in hippocampal activity were related to higher striatal GABA+ levels immediately after initial learning in the stress group as compared to controls. In addition, stress altered the relationship between GABA+ measures and subsequent consolidation-related changes in hippocampal and striatal activity over sleep. The results of Chapter 4 provide the first evidence that stress influences the relationship between striatal GABA and hippocampal activity during motor sequence memory acquisition and consolidation.

Altogether, the results of my doctoral work indicate that stress did not modulate motor behaviour at the group level. However, our extensive multimodal neuroimaging approach including measures of brain activity, connectivity and neurochemicals showed that stress not only modulated brain responses in motor-memory networks but also that it altered the link between these neural signatures and subsequent motor memory consolidation.