Scientists Map Brain's Primary Memory Network

December 9, 1999

For the first time, a team of Wake Forest University investigators has mapped the functional organization of the hippocampus, the brain''s primary memory network, a step that other scientists are calling "a major breakthrough."

The researchers - Sam A, Deadwyler, Ph.D., Robert E. Hampson, Ph.D. and John D. Simeral - report in today''s (Dec. 9) Nature that they have mapped the way that a part of the brain, the dorsal hippocampus, encodes information when rats perform a short-term memory task.

The researchers, members of the Department of Physiology and Pharmacology, mapped the actions with an array of 10-16 microelectrodes. The electrodes are small enough to record the electrical impulses of individual brain neurons during the animals'' performance. Recordings from the electrodes demonstrate that different portions or segments of the hippocampus are active at different times during the task depending on the type of memory function required.

In the "News and Views" section of the same issue of Nature, Howard Eichenbaum, Ph.D., of the Laboratory of Cognitive Neurobiology at Boston University termed the breakthrough in understanding memory processes highly significant, adding that the study revealed "a functional organization for the hippocampus, one of the highest cortical processing areas in the brain."

The rats are tested in an experimental chamber with two bars or levers positioned on a single wall as left or right. At the start, only one lever is presented. It is pressed by the animal, then retracted, followed by a delay period in which the rat must engage in other

unrelated activity. The delay period can be as short as one second or as long as forty seconds -- the rat never knows. At the end of the delay, both levers appear, and the animal is supposed to press he lever it did not press at the outset of the trial. If it does, it is rewarded. If the wrong lever is pressed the chamber goes dark for five seconds and a new trial begins.

"The uniqueness of this situation is that the animal''s task is to remember one piece of information in one phase of the task, that is, which lever it pressed before the delay, and then retain and use that information to make a decision about which lever to press when both levers are available after the delay is over," said Deadwyler, professor and vice chair of the department.

The task is easy if the delay is short -- the animal will get the answer correct most of the time. However, as the delay becomes longer it is more likely that animals will not remember which lever it pressed at the start and chose the wrong lever at the end of the delay. Deadwyler noted that is similar to the rapid decrease in retention of a new telephone number after it is dialed.

Animals played the game between 100 and 150 times each day and generated very stable performance profiles.

While the animals were performing, the researchers were recording which neurons in the hippocampus were active and found different patterns, depending on which lever the animal was supposed to choose at end of the trial. "There''s a distinct separation up and down the hippocampus with respect to which groups of cells fire during the different phases of the memory task," Deadwyler said.

"The findings extend this knowledge of hippocampal encoding to an anatomic framework of overlapping ''memory networks'' in which location within hippocampus determines which cells are activated under which short-term memory demands," the team reported.

In his commentary, Eichenbaum points out that the anatomic framework described by Deadwyler and colleagues follows known functional anatomy present in other brain areas that are not specialized to encode memories, indicating that "the coding properties of hippocampal neurons ''respect'' the anatomical circuitry in which they reside."

Deadwyler said the work parallels ongoing studies in people in which scientists are trying to determine how subjects encode different types of information and how retrieval of that information occurs in different brain regions. "There might be a similar anatomic encoding scheme in the human hippocampus for categorizing and partitioning information used in short term memory as we have seen in the rat."

Hampson is associate professor and Simeral is a third-year graduate student.

###

Contact: Robert Conn, Mark Wright or Jim Steele at (336) 716-4587

Media Relations