Brain oscillations are a necessity to form new memories. They are fundamentally involved in all phases supporting successful memory formation, i.e. maintenance, encoding, transfer and consolidation. To gain a better understanding of successful memory formation, examining the functional role of different frequency bands in all phases of memory formation and across different age groups is essential, as their contribution can differ. For example, theta and gamma band activity are known for their involvement in information transfer, while spindles and delta band activity play important roles in consolidation. The same oscillatory patterns may also serve distinct functions in different phases of memory formation.
We aim to give a coherent picture about the involvement of brain oscillations across phases of successful memory formation by focusing on two experimental approaches: First, examining endogenous brain oscillations and second, manipulating brain oscillations using e.g., transcranial alternating current stimulation (tACS) in which a sinusoidal current oscillating at a specific frequency is applied.
In our symposium, Monika Schönauer will present new findings about oscillatory mechanisms underlying working memory maintenance. Anna Karlsson found that modulations in theta and alpha activity lead to different outcomes in memory formation in older compared with younger adults. Sandrine Baselgia will present a tACS-based study showing a functional role of theta in the encoding of acoustically presented word pairs, and Sven Paẞmann will present preliminary results of the same approach during sleep-dependent consolidation. Marit Petzka will present how sleep spindles track encoding patterns in favour of memory consolidation.
Decoding retrieval success and memory content during short-term memory maintenance
1University of Freiburg, Deutschland; 2University of Tübingen, Deutschland; 3LMU München, Deutschland
Apart from coding the particular content of a learning episode, a memory representation must permit successful memory retrieval. Using multivariate pattern classification, we tested whether electrical brain activity recorded during short-term memory maintenance satisfies these conditions, and where identified short-term memory representations reside. In our experiment, participants learned two short-term memory tasks, encoding either pictures of faces or houses, or sequences of digits or letters while brain activity was recorded with EEG. It was possible to decode retrieval success from electrical brain activity during the delay period of both short-term memory tasks. Moreover, we could distinguish whether participants kept pictures of faces or houses in memory, and classifier performance on this problem correlated with successful memory maintenance. Using spatial as well as frequency-based searchlight analyses, we found that distinct brain areas and frequency bands coded for the success versus the content of short-term memory. Frontal and parietal higher frequency bands and alpha activity predicted retrieval success, whereas memory content was represented in temporal and parietal higher frequency ranges, as well as theta activity. We propose that frontal cortex supports memory-related control processes, whereas temporal cortex shows a sensory reinstatement of material content and is part of the wider activated network during memory retention. Interestingly, the only overlap between electrodes coding for retrieval success and memory content was found over medial parietal regions, indicating that a dedicated short-term memory representation resides in medial posterior cortex.
Imprecise theta-gamma coupling underlies age differences in associative memory
1Max-Planck Institute for Human Development, Berlin, Humboldt University, Berlin; 2Max-Planck Institute for Human Development
Neural oscillations reflect rhythmic fluctuations in neural synchronization and supports episodic memory formation by shaping synaptic connectivity. The cross-frequency coupling of gamma power to the phase of the theta rhythm has been proposed to support the formation of item-context associations via the precise temporal regulation of synaptic activation. A hallmark of aging is the decline in memory for associative information. However, little is known about whether the associative deficit seen in older age can be associated with alterations in theta-gamma coupling during memory formation. In the current study, younger (n=59) and older (n=54) adults performed an object-scene association task while being monitored with EEG. In a subsequent memory test, old and new objects were presented whit an old or new scene and participants responded if the object was old or new and if the specific object-scene pair was old or new, thus allowing us to separate object from pair memory. We demonstrate that theta-gamma coupling supports the formation of object-scene associations in both age groups. While coupling closer to the peak of the theta rhythm was beneficial for subsequent pair memory, objects later remembered without their associated scene showed a reliable deviation from the optimal coupling phase. Furthermore, older adults’ reduced pair memory was accompanied by a shift in coupling phase in comparison to younger adults. Thus, we show that neural synchronization as regulated by theta supports associative memory formation and that senescent changes in the precision of such synchrony underlies the associative deficit seen in older age.
Theta's functional role in the encoding of declarative memory traces
University of Fribourg, Switzerland
The formation of declarative memory occurs in a three-stages process (encoding, consolidation and retrieval). The bi-directional interaction between the prefrontal cortex (PFC) and the hippocampus (HPC), organised by different frequency bands, is of particular importance. Specifically, a positive correlation has been shown between oscillatory activity in the theta band (4-7 Hz) during the encoding and memory performance, implying a role for theta in the communication between PFC and HPC. However, a functional relevance of theta for the encoding of memory has not been shown.
In this study we investigated whether transcranial alternating current stimulation (tACS) delivered during encoding in the range of theta frequency band is able to improve the memory performance in a Dutch-German word pair learning paradigm. For this purpose, 30 healthy subjects joined two sessions where they learned two lists of word-pairs in each. TACS was delivered via four stimulation electrodes (target: FP1 & P7; return: FP2 & P8; phase difference: intrahemispheric 0°, interhemispheric 180°; 2mA) with either individually determined theta or 15 Hz as control (one per session). The first list was encoded under tACS (online), while the second list was not (offline). Delayed recall after a 30-minutes break serves to collect measures of long-term effects of successful encoding.
We present tACS-related findings on memory performance with respect to online and offline encoding and neurophysiological measures to show changes in brain activity induced by the stimulation. The discussion will highlight implications of those findings in the light of the current knowledge on memory formation.
Theta’s functional role during nocturnal consolidation of declarative memory traces
Cognitive Biopsychology and Methods, Universität Fribourg, Schweiz
The (nocturnal) consolidation of previously encoded information is one of the important steps to stabilize memory engrams. Newly collected information are reactivated spontaneously in the Hippocampus during non-REM sleep episodes, transferred back to the neocortex and integrated into already existing memory traces. Given that a successful consolidation of those reactivated information is accompanied by increased theta activity (as it is during their encoding), this study aimed to prove the functional role of theta during this process as well. We examined whether transcranial alternating current stimulation (tACS) applied at individual theta frequency (ITF) during early nocturnal non-REM sleep episodes altogether with targeted memory reactivations is able to improve the overnight performance compared to a control stimulation (23 Hz). Note that the protocol shares many features from the study presented before, mainly the encoding of two different lists of Dutch-German word pairs and parameters of tACS (target: FP1 and P7; return: FP2 and P8; phase difference: intrahemispheric 0°, interhemispheric 180°; 2mA). We applied one stimulation condition in the first sleep cycle, while one of two individually created lists of Dutch words (one for each tACS-condition) was played under hearing threshold. To equal parts, the newly created lists contain words indicated as remembered before sleep and completely new words (2nd sleep cycle: 2nd stim condition, randomized order). Here, we present the current state of results (collected data of 16 subjects) with respect to behavioural and neurophysiological measures and discuss the outcomes in the light of the findings of the study presented before.
Sleep spindles track cortical learning patterns for memory consolidation
School of Psychology and Centre for Human Brain Health, University of Birmingham, Birmingham, UK
Memory consolidation, the transformation of labile memory traces into stable long-term representations, is facilitated by post-encoding sleep. Computational and biophysical models suggest that sleep spindles play a key mechanistic role for consolidation, igniting structural changes at cortical sites involved in prior encoding. Here we tested the resulting prediction that spindles are most pronounced over encoding-related cortical areas and that the extent of this encoding-spindle overlap predicts behavioural measures of memory consolidation. Using high-density scalp Electroencephalography (EEG) and Polysomnography (PSG) in healthy volunteers, we first identified cortical areas engaged during a temporospatial associative memory task (power decreases in the alpha/beta frequency range, 6-20Hz). Critically, we found that participant-specific topographies of post-encoding sleep spindle amplitude directly correlated with participant-specific encoding-task topographies. The extent to which spindles tracked encoding patterns predicted memory consolidation across participants. Our results provide empirical evidence for a role of post-learning sleep spindles in tracking encoding networks, thereby facilitating memory consolidation.