In search of the memory molecule, researchers discover key protein complex

A paper published in the June 22 issue of the Journal of Neuroscience describes the new findings.

The brain is composed of neurons that communicate with each other through structures called synapses, the contact point between neurons. Synapses convey electrical signals from the "sender" neuron to the "receiver" neuron. Importantly, a synapse can vary in strength; a strong synapse has a large effect on its target cell, a weak synapse has little effect.

New research by John Lisman, professor of biology and the Zalman Abraham Kekst chair in neuroscience, helps explain how memories are stored at synapses. His work builds on previous studies showing that changes in the strength of these synapses are critical in the process of learning and memory.

"It is now quite clear that memory is encoded not by the change in the number of cells in the brain, but rather by changes in the strength of synapses," Lisman says. "You can actually now see that when learning occurs, some synapses become stronger and others become weaker."

But what is it that controls the strength of a synapse?

Lisman and others have previously shown that a particular molecule called Ca/calmodulin-dependent protein kinase II (CaMKII) is required for synapses to change their strength. Lisman's team is now showing that synaptic strength is controlled by the complex of CaMKII with another molecule called the NMDAR-type glutamate receptor (NMDAR). His lab has discovered that the amount of this molecular complex (called the CaMKII/NMDAR complex) actually determines how strong a synapse is, and, most likely, how well a memory is stored.

"We're claiming that if you looked at a weak synapse you'd find a small number of these complexes, maybe one," says Lisman. "But at a strong synapse you might find many of these complexes."

A key finding in their experiment used a procedure that reduced the amount of this complex. When the complex was reduced, the synapse became weaker. This weakening was persistent, indicating that the memory stored at that synapse was erased.

The experiments were done using small slices of rat hippocampus, the part of the brain crucial for memory storage.

"We can artificially induce learning-like changes in the strength of synapses because we know the firing pattern that occurs during actual learning in an animal," Lisman says.

To prove their hypothesis, he explained, his team first strengthened the synapse, eventually saturating it to the point where no more learning or memory could take place. They then added a chemical called CN-19 to the synapse, which they suspected would dissolve the CaMKII/NMDAR complex. As predicted, it did in fact make the synapse weaker, suggesting the loss of memory.

A final experiment, says Lisman, was the most exciting: They started out by making the synapse so strong that it was "saturated," as indicated by the fact that no further strengthening could be induced. They then "erased" the memory with the chemical CN-19. If the "memory" was really erased, the synapse should no longer be saturated. To test this hypothesis, Lisman's team again stimulated the synapse and found that it could once again "learn." Taken together, these results demonstrated the ability of CN19 to erase the memory of a synapse -- a critical criterion for establishing that the CaMKII/NMDAR complex is the long sought memory storage molecule in the brain.

Lisman's team used CN19 due to previous studies, which indicate that the chemical could affect the CaMKII/NMDAR complex. Lisman's team wanted to show that CN19 would decrease the complex in living cells. Several key control experiments proved this to be the case.

"Most people accept that the change in the synapses that you can see under the microscope is the mechanism that actually occurs during learning," says Lisman. "So this paper will have a lot of impact -- but in science you still have to prove things, so the next step would be to try this in an actual animal and see if we can make it forget something it has previously learned."

Lisman says that if memory is understood at the biochemical level, the impact will be enormous.

"You have to understand how memory works before you can understand the diseases of memory."

Lisman assembled a large team to undertake this complex research. A key collaborator was Magdalena Sanhueza, who once worked with Dr. Lisman at Brandeis, and her student, German Fernandez-Villalobos, both now of the University of Chile, Department of Biology and Ulli Bayer of the University of Colorado Denver School of Medicine, Department of Pharmacology, who developed CN19, a particular form that could actually enter neurons.

Others involved include Nikolai Otmakhov and Peng Zhang from Brandeis and Gyulnara Kasumova, who worked in the Lisman laboratory for several years as an undergraduate. An additional group contributing to the work was that of Johannes Hell, Professor of Pharmacology at the UC Davis School of Medicine. He and his student, Ivar S. Stein, used immunoprecipitation methods to actually show that the CN19 had dissolved the CaMKII/NMDAR complex.