Published online 18 December 2008 | Nature | doi:10.1038/news.2008.1313


Neurons on border patrol

The limits imposed by walls and trenches are recognized by special brain cells.

Rat and mirrorRats have neurons devoted to sensing the physical limits of their surroundings.R. Skjerpeng

A newly discovered class of neurons allows rats to identify the borders of their surroundings and helps them to build mental maps.

These neurons, called border cells, join three other known classes of neurons that help us to find our way through space: place cells fire when we pass through fixed locations, letting us know where we are; head-direction cells fire when we face particular directions, acting as a compass; and grid cells fire when we're at specific points on a hexagonal grid that the brain superimposes on our surroundings.

Neuroscientist Edvard Moser from the Norwegian University of Science and Technology in Trondheim and his colleagues first came across border cells in the entorhinal cortex of rat brains1. While recording the activity from single neurons, the team kept finding neurons that were linked to mental maps but that didn't act like any of the three known classes of mapping neurons.

"We ignored them in the beginning," says Moser. Later, however, the researchers realized that they might have found border cells, a theoretically predicted2 class of neurons.

Thinking inside the box

Moser and his colleagues refocused their attention on how these neurons reacted as rats ran around small rooms. They found that the border cells would fire when rats approached walls — generally, each border cell was linked with a single wall, but some cells responded to several borders.

When the team extended the walls of the room, the neurons associated with a particular wall fired as the rat approached any point along it, showing that the cells were responding to the whole border, rather than to a localized area. And when the walls were replaced with vertical drops, the neurons reacted in the same way, showing that it was borders, rather than walls, that triggered neuronal activity.

"They really respond to borders, and apparently nothing other than that," says Moser.

"These border cells give you some sense of the structure of [your surroundings] — how big that space is," says Jeffrey Taube, a neuroscientist at Dartmouth College in Hanover, New Hampshire.

Modelling mental maps

Neil Burgess of University College London, one of the scientists who originally proposed that border cells should exist, is happy to see his predictions proven correct. "I think most people ignored my model," he says.

The next step, Burgess and Moser agree, is to determine how the four classes of neurons work together to create the mental maps that let animals know where they are.

Moser suspects, for instance, that the border cells somehow align the grid cells to the appropriate borders of the environment. And, he suggests that border cells are particularly important for helping rats to plan routes.

Crucially, findings about the neurons involved in mental mapping are not limited to rats. Researchers have found place cells in humans, grid cells in mice and head-direction cells in mice, chinchillas and monkeys. "I am completely convinced that [border] cells also exist in higher mammals, including humans," Moser says.

Luck and location

Just a few years ago, Moser surprised members of the cognitive-mapping field with his unanticipated discovery and description of grid cells.


Moser puts his success down to luck and to looking in the right place — the entorhinal cortex. This part of the brain provides the main input into the brain's hippocampus and, as we now know, plays host to head-direction, grid and border cells. "It is an area that has really been under-explored," he says.

Others are now taking note, however. "He's come up with some pretty interesting findings," says Taube. "My first impression [when I read this paper] was 'What will this group find next?'". 

  • References

    1. Solstad, T., Boccara, C. N., Kropff, E., Moser, M.-B. & Moser, E. I. Science 322, 1865–1868 (2007).
    2. Hartley, T., Burgess, N., Lever, C., Cacucci, F., O'Keefe, J. Hippocampus 10, 369–379 (2000).
Commenting is now closed.