was recording with a very thin microelectrode as it passed hundreds, probably thousands of cells through other parts of the brain before it reached the hippocampus. I sampled the electrical activity of the brain cells as the electrode passed by, and the brief burst of electrical activity registered as little “pops,” which you can hear on an audio monitor. My goal was to figure out if the pattern of this firing from a given cell had anything to do with the animal learning a new association between a picture and a reward. But there were no guarantees. There were plenty of days when I listened to the activity of many cells with the electrode, and not a one did anything much at all. It just sounded like a bunch of radio hash with no rhyme or reason to the pattern of firing. Other days, however, I got lucky and caught a nice big fish in the form of a cell that had interesting activity—for example, cells that seemed to fire only when a particular picture was shown or cells that fired a lot during the blank delay interval of the task between the presentation of the picture and when the animal made a response to one of the targets.
I kept fishing in the hippocampus with the hope of finding something interesting, and over the first few months of recording something did start to emerge. I noticed that a particular cell we were monitoring seemed to have little or no firing associated with the task early in the trial when the animal had not learned any of the associations. But then, the cell seemed to increase its firing rate later in the session when more associations were learned. I didn’t fully appreciate the pattern until we analyzed the data later. Then it was as obvious as the nose on your face.
Just as I had noticed when listening to the cells during the experiment, these cells had little or no specific firing related to the task early in the learning session when the association had not been formed. But as the animal learned a new association, certain cells would dramatically increase their firing to double or sometimes triple their earlier rate. The increase in activity didn’t happen when all associations were learned, just during particular associations. This suggested that there were particular groups of cells in the hippocampus that signaled the learning of new associations by changing their firing rate. I realized I had been listening to the birth of a new memory in the firing of these neurons! Nobody had ever characterized learning in the hippocampus in quite this way before. We were seeing exactly how hippocampal cells encoded newly learned associations, and because we know that damage to this brain region impairs the creation of such associations, the study suggested that this pattern of brain activity was the key to the new associative learning process.
This was not only exciting for my research partners and me but for the field of neuroscience in general. Ours was one of only a handful of demonstrations of brain plasticity occurring in real time and directly associated with a change in behavior, in this case, new associative learning! Diamond had shown that rat brains had more synapses in general if the animal was raised in an enriched environment relative to an impoverished environment, but her studies did not measure behavior while learning was occurring or while a memory was forming. It was kind of assumed that if your brain got bigger, this was generally a good thing for behavior or performance. The long-term implication of our work is that if we understand this functionality in the brain, we might be able to replicate it when a brain is handicapped by various neurological problems. In other words, these findings showed us how cells in a normal hippocampus work as new memories are being formed. Because this brain area was missing in H.M., he was not able to form new memories. Importantly, these results are also key first steps to developing possible therapies for the associative or episodic memory