Neuroscience How the brain consolidates memories during sleep

It is difficult to put a number to the total knowledge the human brain is able to store. It is quite a lot. Even so, we are not able to memorise everything that we would like to remember. How is knowledge stored in the complex network of nerve cells? Why do some memories stick and others don’t?

These questions are keeping Prof Dr Nikolai Axmacher busy. Together with his team at the Department of Neuropsychology, he is trying to figure out what happens in the brain when new information is stored. His approach: he analyses phenomena that are familiar from animal studies in humans. This presents a considerable challenge.

Reactivating events during sleep

Researchers have learned in experiments with mice and rats that the animals reactivate newly acquired information during sleep. If a rat learns a new way through a maze, this information is coded in a characteristic pattern of nerve cell activity.

Specific cells in the rat brain are active while the animal is running, and they reflect its location in its surroundings. The same cells fire in the same sequence again when the rat is sleeping. It appears that the brain retraces the route. Researchers assume that this reactivation is crucial for consolidating memory contents.

Rendering nerve cells receptive

Another effect seems to be also relevant, namely so-called ripple oscillations. This term describes a specific type of brain activity: a cluster of nerve cells sends out high-frequency signals for a short period of time. In the EEG, they appear in a characteristic waveform.

Researchers believe that the ripples prepare the nerve cells for the reactivation of information. The theory: following a ripple event, a nerve cell is more receptive for storing reactivated information permanently.

Nikolai Axmacher tests if these effects also occur in humans. However, traditional measurement methods used in cognitive neuroscience are insufficient for this purpose, because long-term memory is controlled by a region located deep inside the brain, the hippocampus. The relatively weak ripple events cannot be detected using scalp EEG.

Study with epilepsy patients

The solution: Nikolai Axmacher conducted tests with epilepsy patients who, for medical reasons, had electrodes implanted in the brain. With their aid, he was able to record depth EEG inside the skull and, consequently, to perform direct measurements of the activity of various brain areas.

The researcher brought the data of 13 patients from his previous place of work, the university clinic in Bonn, to RUB. Using this database, Axmacher’s team has performed various analyses and developed more and more sophisticated methods, which have eventually allowed for a combined search for ripples and reactivation.

The experiment was conducted as follows: in the first step, the study participants viewed 80 landscape pictures with or without buildings. They had to indicate if a building was pictured in the respective image or not, while the researchers were recording EEG from the implanted electrodes. Subsequently, the patients slept for one hour; during that time, the researcher again recorded EEG activity.

Memory test after sleep

After the sleep phase, the patients viewed another 80 pictures of landscapes, and again had to state if they contained any buildings. Then came the final test: the researchers showed the 80 pictures used in the first run, the 80 pictures used in the second run, as well as 80 new images. The study participants were asked to indicate which pictures they had viewed. Here, too, the researchers recorded an EEG.

Nikolai Axmacher and his postdoc researcher Hui Zhang conducted analyses to test if reactivation occurs during sleep. For this purpose, they first compared the brain activity during the initial run with the activity during the sleep phase. The EEG from the second run served as baseline – those items could not be replayed during sleep because they had only been presented afterwards.

Does the brain generate the same activation patterns during the sleep phase that had occurred while viewing the images? In other words: can the results gained in animal studies be transferred to humans?

In human as in rats

The preliminary results suggest: yes. During sleep, the same activation patters manifested themselves as during the presentation of the landscape pictures. In the next step, Zhang and Axmacher searched for ripples; here, too, they found what they’d been looking for.

The EEG data contained short high-frequency oscillations, the same that had been described in animals. It was now necessary to ascertain if the ripples had an impact on reactivation and memory performance.

Enhanced brain activity

The team analysed the intensity of reactivation after a ripple as well as its intensity before a ripple had occurred. Again, the result matched the findings gathered in animal studies. Following a ripple, reactivation was more intense than during a reference period prior to a ripple.

“Individual stimuli, in this case landscape pictures, are reactivated during sleep, and ripples appear to enhance reactivation,” explains Nikolai Axmacher. The researchers detected that enhancement mechanism only in the reactivation of those images which were remembered in the final test.

In other words: “If a ripple enhances reactivation, the person remembers the picture later,” explains Axmacher. “We are dealing with a mechanism for learning during sleep.” The researchers assume that this approach could also be used for consolidating more complex memory contents in the brain.