Skip to content ↓

Landmark study reveals mouse memory mechanism

In a ground-breaking study, MIT scientists have demonstrated how animals form memory for places, which may directly relate to the same ability in humans.

This latest "regional gene knockout" technology-through which scientists can develop a breed of mice in which a gene is eliminated in a specific area or only in one particular type of cell-will be valuable in the study of neurological diseases such as Alzheimer's, Huntington's, Parkin-son's and drug addictions.

The study, in three related scientific papers, was published in the December 27 issue of Cell. It involves a molecular, behavioral and electrical study of how animals develop spatial memory of a new environment after just a few minutes of exposure. The scientists believe humans use the same mechanisms to remember the location of objects and to navigate between places.

The group obtained evidence showing that strengthened connections between groups or ensembles of neurons enable the formation of internal "maps" of a space which allow the animals to remember that physical environment.

This technology is much more precise than the first generation of gene-knockout technology in which animals, through cross-breeding, inherit a genetic makeup that totally eliminates a specific gene throughout the body, in whatever organ or cell type those genes function. The earlier technology, developed in 1989, limited the conclusions that scientists could definitively make, because the same gene can have different functions in different regions within organs, and it may function at different times during the maturation process.

"We have developed a method to create mice in which the deletion (knockout) of virtually any gene of interest is restricted to a sub region or a specific cell type in the brain. The brain subregion-restricted gene knockout should allow a more precise analysis of the impact of a gene mutation on animal behaviors," wrote the authors of the first paper. The nine-person scientific team at the Center for Learning and Memory was headed by Assistant Professor Matthew Wilson and Nobel laureate Susumu Tonegawa, director of the Center.

"This work proves once and for all time that this part of the brain is the crucial part for this kind of memory," said Dr. Charles F. Stevens, a Howard Hughes Medical Investigator and professor at the Salk Institute. In an interview, Dr. Stevens, who commented on the paper for Cell, said: "It is a dream of neurobiologists to understand some interesting cognitive phenomena like memory from the molecular level right up through behavior. The articles in Cell are a big step in that direction."

"The new genetic technique utilized by Tonegawa and Wilson will revolutionize the field of brain research," said Professor Emilio Bizzi, head of the Department of Brain and Cognitive Sciences. "With this technique, scientists will be able to study how the brain encodes different types of memory-that is, memory of facts, of objects and of motor skills."

As described in the first paper, the researchers created a strain of mice in which the gene for a specific neurotransmitter receptor called the NMDA receptor is knocked out only in one type of nerve cells (CA1 pyramidal cells) in the hippocampus, the brain area that plays a crucial role in some forms of memory such as "spatial memory." Earlier, other scientists had hypothesized that this and many other types of memory are stored as strengthened neuronal connections, or synapses. This strengthening of synapses is termed "synaptic plasticity." But hard evidence for this hypothesis has been difficult to find.


The scientists in the second paper discovered a critical link between synaptic strengthening and spatial memory by having the new knockout mice perform spatial memory tasks and by measuring the ability of cellular connections to change.

In order to further understand how animals and humans acquire spatial information, MIT scientists in the third paper analyzed these knockout mice as they explored an environment freely by using a highly sophisticated electrophysiological technique which precisely monitors the electrical activity of a large number of the individual nerve cells in the CA1 region of the hippocampus. The researchers discovered that, unlike the normal mice, these new knockout mice have a deficiency in establishing an internal spatial "map" among an ensemble of CA1 nerve cells because of the inability to strengthen synaptic connections.

For the first time, it was demonstrated beyond doubt that synaptic plasticity in a specific region of the brain is essential for spatial memory. In addition, the scientists could demonstrate that synaptic plasticity gives rise to spatial memory through the establishment of internal "maps" of the space. The new gene knockout technology will also be extremely valuable for medical researchers to study the precise functions of disease genes in neurological disease processes as well as in drug addictions.

On the first two papers, the lead authors were Joe Z. Tsien, a post-doctoral associate with Dr. Tonegawa, and Thomas J. McHugh, a biology graduate student in Dr. Tonegawa's lab who has been supervised principally by Dr. Wilson for this work.

The scientists noted that past experiments in which NMDAR1 (NMDA receptor 1) genes were knocked out of the entire body resulted in death within a day or so.

The scientists at the Center for Learning and Memory, however, developed a mouse strain (Paper 1) in which the deletion of the NMDAR1 gene was restricted to just the pyramidal cells in the CA1 region of the hippocampus. The knockout mice grew into adulthood without obvious abnormalities.

Mice are natural swimmers but they prefer to be on land or a solid surface. A test of memory used in Paper 2 is the Morris water maze, a round pool in milky water where there is a platform that is just below the surface. The platform can be easily spotted by all mice, including the new knockout mice, when there is a flag that extends above the water attached to the platform. But it's different when the flag is gone.

The mice were placed in the Morris water maze three or four times a day for several days. They explored the maze and swam into the platform. By the fourth day, normal mice made a beeline for the submerged platform, knowing precisely where it was. But the knockout mice swam all over the place, and showed clear deficits in their ability to find the underwater platform. Clearly, they were unable to form a map of the space in their brains for some reason.

The answer was found by monitoring the electrical activity in the hippo-campus's pyramidal cells of ordinary mice and the knockout mice. The authors of Paper 3 found that each pyramidal cell has its own region of heightened firing of neurons ("place field"). Large numbers of hippocampal place cells will tile each environment with overlapping place fields, allowing the mouse to remember and navigate the space or maze.

The firing patterns are so clear that the location of the animal can be well estimated just by looking at the firing patterns on the computer screen, rather than at the animal.

The great difference between knockout mice and ordinary mice was shown when the software developed by Dr. Wilson's laboratory tracked the coordinated firing of pairs of place cells. The place cells of knockout mice exhibited completely non-correlated firing, while the place cells of ordinary mice showed significant correlation.

Because 30 to 50 percent of the cells within the hippocampus become active within a given environment, the scientists concluded that rodents use information from ensembles of cells to calculate location, rather than mapping places to individual cells.

The research was supported by the National Institutes of Health and gifts from the Shionogi Institute of Medical Science in Japan, and Amgen, Inc., to Dr. Tonegawa; and by awards from the Seaver Institute and the Sloan Foundation to Dr. Wilson.

A version of this article appeared in MIT Tech Talk on January 8, 1997.

Related Topics

More MIT News