A new study shows that researchers examined that they provided the first glimpse of the molecular mechanism by which long-term memories are encoded in a region of the hippocampus called CA3. They found a protein called Npas4, previously identified as a master controller of gene expression triggered by neuronal activity. This study is published in Neuron.
This protein controls the strength of connections between neurons in the CA3 and those in another part of the hippocampus called the dentate gyrus. Without Npas4, long-term memories cannot form. "Our study identifies an experience-dependent synaptic mechanism for memory encoding in CA3, and provides the first evidence for a molecular pathway that selectively controls it," says Yingxi Lin.
Neurons in the CA3 region play a critical role in the formation of contextual memories, which are memories that link an event with the location where it took place, or with other contextual information such as timing or emotions.
These neurons receive synaptic inputs from three different pathways, and scientists have hypothesized that one of these inputs, from the dentate gyrus, is critical for encoding new contextual memories. However, the mechanism of how this information is encoded was not known.
When the researchers knocked out the Npas4 gene, they found that mice could not remember the fearful event. They also found the same effect when they knocked out the gene just in the CA3 region of the hippocampus. Knocking it out in other parts of the hippocampus, however, had no effect on memory.
In the new study, the researchers explored in further detail how Npas4 exerts its effects. Lin's lab had previously developed a method that makes it possible to fluorescently label CA3 neurons that are activated during this fear conditioning. Using the same fear conditioning process, the researchers showed that during learning, certain synaptic inputs to CA3 neurons are strengthened, but not others.
Further experiments revealed that this strengthening is required specifically for memory encoding, not for retrieving memories already formed. The researchers also found that Npas4 loss did not affect synaptic inputs that CA3 neurons receive from other sources.
The researchers also identified one of the genes that Npas4 controls to exert this effect on synapse strength. This gene, known as plk2, is involved in shrinking postsynaptic structures. Npas4 turns on plk2, thereby reducing synapse size and strength. This suggests that Npas4 itself does not strengthen synapses, but maintains synapses in a state that allows them to be strengthened when necessary. Without Npas4, synapses become too strong and therefore cannot be induced to encode memories by further strengthening them.
In future study, author hopes to work on the circuit connecting the dentate gyrus to CA3 interacts with other pathways required for memory retrieval. Somehow there's some crosstalk between different pathways so that once the information is stored, it can be retrieved by the other inputs.