Sleep makes up about 1/3 of our lives. Decades of research have concluded that sleep is vital for learning functions and making long-term memories. However, knowing exactly how long-term memory is formed is not wholly understood. It remains a primary question in neuroscience
New research from neuroscientists out of the University of California Riverside published in the Journal of Neuroscience reported that there may now be an answer to that question. For the first time, this study will look at a mechanistic explanation for how deep sleep, or slow-wave sleep, is responsible for the consolidation of new memories.
Animal and human brains are disengaged from any sensory input during sleep. Even so, the brain continues to be highly active, with electrical activity present in the form of hippocampal sharp-wave ripples, as well as high-amplitude slow oscillations in the cortex. The hippocampus is a small brain region that is part of the limbic system. The cortex is the outer part of the cerebrum. This process reflects that there are alternating periods of silent and active neurons during deep sleep states. During sleep, the episodic memories acquired during wakefulness are transferred from their initial spot in the hippocampus to the cortex and stored there as long-term memories.
The researchers at UC Riverside used a computational model to create the link between the neuronal synaptic connections and the brain’s electrical activity during deep sleep. The model spontaneously generates patterns of slow oscillations in the cortex and shows that these patterns are directly influenced by the sharp-wave ripples in the hippocampus. Additionally, the oscillations in the cortex determine synaptic changes in the neurons. It is notable that synaptic strength is generally believed to play a role in memory storage and learning. In this model, the synaptic changes affect the slow oscillation patterns and promote a sort of reinforcement and replay of specific memories.
Lead researcher and author of the study, Dr. Yina Wei, notes that the slow oscillations are undisturbed by input from the hippocampus. This is interpreted as an explanation for the ability to consolidate specific memories during sleep because the traces of memory are formed within the cortex and then become completely independent of the hippocampus.
According to the computational model used, Dr. Wei explained that the hippocampal input goes to the cortex during deep sleep and then proceeds to influence how the slow oscillations are brought into the cortical network.
The influence of these slow oscillations and the input from the hippocampus activates memories during sleep, causing some memories to replay. During this replay, synapses that correspond to these are strengthened for long-term storage within the cortex. Dr. Wei reports that this suggests there is an important link between hippocampal sharp-wave ripples and the transfer of memories to the cortex.
As mentioned, brain activity stays high during sleep. Normal sleep consists of rapid eye movement (REM) and non-rapid eye movement (NREM) sleep. REM and NREM alternate several times throughout the sleep cycle, usually about four or five times in an eight-hour sleep period. NREM occurs first, which is then followed by REM sleep, and one cycle lasts between 90 and 110 minutes. There are three stages of NREM sleep, with the third stage being deep sleep. Deep sleep makes up approximately 20% of the cycle and generally occurs in the first third part of the night.
It is of note that even spatially localized and weaker input from the hippocampus had an effect on the slow oscillation pattern, which then led to persistent changes of synaptic communication between neurons. The model can make predictions, Wei believes, that can be experimentally tested, which include interventions that will either augment or suppress the process of memory consolidation.
Rachael Herman is a professional writer with an extensive background in medical writing, research, and language development. Her hobbies include hiking in the Rockies, cooking, and reading.