During memory formation, proteins are synthesized that transform newly learned information into stable synaptic changes, such as the insertion of new receptors at the synapse. This stability, or consolidation, generally occurs in a time window which covers the first few hours after learning. But, proteins have a constant turnover; they are always being made and broken down, and thus have a limited life span. One puzzle that memory researchers have been trying to solve is how the synaptic changes that occur during memory formation are maintained despite this molecular turnover. Pedro Bekinschtein and colleagues from the Instituto de Biologia Celular y Neurociencias in Argentina have shown in a recent Neuron article that a memory trace likely undergoes several rounds of consolidation-like events in order for it to persist over time.

To find out how a memory is maintained in rats' brains, the authors inhibited protein synthesis at numerous time points before and after learning by injecting the drug anisomycin into the hippocampus, which is a brain area essential in memory formation in a variety of tasks. One such task is called inhibitory avoidance, in which a rat is placed on a platform over a wire grid. When the rat steps down onto the grid it receives a painful foot shock. Not surprisingly, the rat learns to avoid the foot shock by remaining on the platform. This type of training only requires one trial and the memory is long-lasting. When the authors inhibited protein synthesis in the hippocampus immediately before training or 3 h post-training, the rats did not remember to remain on the platform 2 days later. Inhibiting protein synthesis 9, 12 or 24 h post-training did not affect the 2-day memory: these rats remembered to remain on the platform. These results represent the classical consolidation effect, where memory is impaired if protein synthesis is inhibited within the first few hours after training.

However, when the authors looked to see if the memory was maintained 7 days after learning the avoidance task, a slightly different story emerged. Again, when they inhibited protein synthesis in rats immediately before training or 3 h post-training, the animals formed no memories. And, again, rats remembered their training when the researchers inhibited protein synthesis 9 or 24 h after training. The difference occurred when the team inhibited protein synthesis 12 h after training. The rats remembered their training 2 days later, but not 7 days later. This implies that there is a second time point 12 h after training, in addition to the time period immediately after training, in which another round of protein synthesis is required to stabilize memories and make them persist.

These results challenge the classical view that a memory undergoes consolidation only once, immediately after learning, by demonstrating that there is at least one other critical time window 12 h after learning. The authors replicated this effect using a second hippocampal-dependent memory task, suggesting that this process of repeated consolidation events may occur after different types of learning tasks. Having recurrent consolidation windows not only helps the memory to persist during protein turnover, but also allows a time window where the memory can either be strengthened, resulting in persistence, or weakened, resulting in forgetting.