At times it must seem to neuroscientists that the enigma of memory reveals its secrets to them as if they were the proverbial blind men describing an elephant. “Ah yes, it has a hose, a very thick hose, so thick it’s almost like a tree trunk!” If only it were as easy to get the feel of neurons as it is for an elephant.

Philosophers and scientists have been pondering, poking and experimenting around the concept and physical reality of memory for centuries. Saying, “We’re a lot closer now,” is probably true, but like determining the properties of an elephant, what we know is probably bits and pieces. That said, as new pieces are added, something that resembles a working hypothetical framework is emerging, and science thrives on frameworks that answer questions and lead to testable results.

A new piece, and potentially a very important new piece, has been added by John Lisman and Zalman Kekst at the Lisman Laboratory at Brandeis University (Boston, Massachusetts, USA) and published in the Journal of Neuroscience [22 June 2011, paywalled, Role of the CaMKII/NMDA Receptor Complex in the Maintenance of Synaptic Strength]. In short, memory appears to be related to proteins that exist in the unique space between neurons called the synapse.

The finding, which I’ll describe in more detail in a moment, is not in itself surprising. Neuroscientists have suspected for some time that proteins are involved in the memory process. It figures, because proteins are the ‘building blocks of biology.’ They are the most flexible, adaptable, and varied of all the biochemical materials. Why wouldn’t memory, which probably requires trillions of coding possibilities, make use of proteins? Well, it hasn’t always been seen that way. Among the many models of how memory works, it was held for some time that neurons themselves, brain cells, were created, shaped and connected to create memory. That model is in the process of being superseded by findings that indicate memory is more likely created in the synapses between neurons.

I won’t get deeply into neuroscience 101, but perhaps you’re familiar with the image of a neuron with a long ‘tail,’ an axon lead out from it and coming up next to – but not touching – another axon from another neuron. The tiny space between axons is the synapse, where the most complicated of all electrobiochemistry takes place, and where a bioelectrical impulse (action potential) ‘jumps,’ is transmitted, between axons. Only the jumping isn’t always the point. Neuroscientific research is reaching consensus that the ‘strength’ of the synapse (to oversimplify, its electrical charge) in some brain neurons is an element of memory encoding. Increase the strength and a bit of information is stored. Decrease the strength below a certain point, and that information is erased.

I’m using some terminology used around computer memory because it’s familiar. Care should be taken not to assume that what happens in the synapses is like computer memory. At this point, it’s only an analogy.

This model or hypothesis about memory has been the focus of research for about two decades, and as I mentioned it has been suspected that proteins are involved. Now, according to the new findings, a specific protein complex has been identified. It has an unmemorable (!) name: Calcium/calmodulin-dependent protein kinase II or CaMKII, combined with a NMDAR glutamate receptor to make the CaMKII/NMDAR complex, which doesn’t help much. This complex determines how strong a synapse is, which translates into how well memory is stored.

A long chain of research, which revealed the presence of this complex in the material of the synapse, led to specific experiments based on simple logic: More of the complex in the synapse would indicate a strong memory; less of the complex (or none) would indicate a weak or non-existent memory. It’s kind of the put-in-take-out sort of experiment that can lead to solid results – which in this case is what happened. Using neuron segments from the hippocampus of rats, an area long associated with memory formation, the scientists used a chemical to dissolve the complex, which should lead to a loss of memory. It did. Learning, the reverse process, has been shown to increase the quantity of the complex.

So, according to this research, at least one protein complex has been associated with memory formation and loss. No one is claiming this is the end of the story. Obviously, there are other proteins or other chemicals involved, some of which may not be fully discovered. Perhaps there are aspects to the proteins, such as the configuration (folding) that play a role. The precise mechanism of protein formation in this complex is not fully understood. Of course, the 64 trillion connection question, is what, if any, is the code used for memory? The upshot is that neuroscientists don’t know yet how memory works, BUT isolating at least one of the crucial protein complexes involved is a big step in the (hopefully) right direction.

The impact of this finding is not that the elephant in the room has been fully revealed, but that the blind men attempting to figure out what it is have another piece of fact to help shape the next round of exploration. If you’ve ever seen a room full of excited blind men dancing around an elephant…neither have I, but I presume that neuroscientists celebrate their own prospects.