Neural Modeling Is Capable of Predicting Memory Retention

Memory research has made great advances in recent years, especially with respect to the biochemistry of the brain. Both observations and modeling methods have led to significant findings, and opened up even more questions for investigation.

For example, it is now known that in the early stages of storage of longer-term memories, proteins are rapidly produced to assist in the storage process. The proteins are located on brain cells that are activated in association with an event. Before being broken down, some of these proteins are retained at particular locations on specific neurons for several hours, thus providing a "location" of the memory of the event.

The interesting follow-up question is why only a portion of the details is retained within this window of 1-2 hours, instead of a comprehensive recall.

A group of researchers from the Salk Institute addressed that very question through modeling, and found that subsequent actions play a significant role. Their work was published recently in Neuron.

By applying computational modeling methods to available research data from past studies, the research team discovered a common thread. The proteins that are produced are available to multiple neurons within any particular circuit - but for an event to be retained as a memory, the activated neurons also have to be activated later by another event.

Since follow-up activation does not occur with all of the previously activated neurons, some of the details are retained while others are lost.

There is also a physical location component to the storage process. The team found that the relative locations of the proteins to specific neurons and nearby areas was an important component of recording memories, and that whether a memory was recorded or not could be predicted using this model.

The Salk team is also considering the applicability of this model, or a similar model,to memory storage during sleep. Previous research has shown that our memories of the day are processed through the brain while we sleep, and they take a similar leap from short-term storage in the hippocampus to longer-term storage in the cortex.

While other research has suggested that most of the formation of memories take place in the non-dreaming phase of sleep, the work of the Salk group indicates that a portion of the memory retaining process does take place during the dream phase.

These findings are different in an interesting way - they provide a potential bridge between the reactions that are going on at a cellular level and the more structured memory retention process going on at a system level during the important two-hour time window before the proteins degrade.

Through better understanding of the cellular mechanics during the two-hour retention window and the links between the cellular and system levels, this work may lead to breakthroughs in treatment of Alzheimer's disease, or other forms of cognitive disorders.

We're working on re-activating the neurons you used while reading this article. Let's see how much of this article you remember a few hours from now