In the 2001 suspense thriller Memento, the lead character, Lenny, suffers a brain injury that makes him unable to remember events for longer than a minute or so. This type of amnesia, known as anterograde amnesia, is well known to neurologists and neuropsychologists. Like Lenny, sufferers remember events from their life histories that occurred before their injuries, but they cannot form lasting memories of anything that occurs afterward. As far as they recall, their personal histories ended shortly before the onset of their disorders.
The cause of Lenny's problem was probably damage to his hippocampus, a pair of small, deep-brain structures crucial to memory--and also important to some of todays most exciting and consequential neuroscience research. Decades of research have made clear that the hippocampus and surrounding cortex do more than just place our life events in time. The hippocampus, along with a newly discovered set of cells known as grid cells in the nearby cortex, traces our movement through space as well. And by doing so, it supplies a rich array of information that provides a context in which to place our life's events. The picture that is emerging is of historic importance and more than a little beauty.
Exactly how does the brain create and store autobiographical memories? Although that question has fascinated scientists, philosophers and writers for centuries, it was only 50 years ago that scientists identified a brain area clearly necessary for this task--the hippocampus. The structure's role was made clear in 1953, when William Scoville, a Hartford, Conn., surgeon seeking to relieve the epileptic seizures that were threatening to kill a patient known as H.M., removed most of H.M.'s hippocampus and discovered he had rendered him unable to form new, conscious memories. Since then, the case of H.M., along with extensive animal research, has firmly established that the hippocampus acts as a kind of encoding mechanism for memory, recording the timeline of our lives.
In the 1970s another discovery inspired the theory that the hippocampus also encodes our movement through space. In 1971 John O'Keefe and Jonathan Dostrovsky, both then at University College London, found that neurons in the hippocampus displayed place-specific firing. That is, given "place cells," as O'Keefe dubbed these hippocampal neurons, would briskly fire action potentials (the electrical impulses neurons use to communicate) whenever a rat occupied a specific location but would remain silent when the rat was elsewhere. Thus, each place cell fired for only one location, much as would a burglar alarm tied to a tile in a hallway. Similar findings have been reported subsequently in other species, including humans.
These remarkable findings led O'Keefe and Lynn Nadel, now at the University of Arizona, to propose that the hippocampus was the neural locus of a "cognitive map" of the environment. They argued that hippocampal place cells organize the various aspects of experience within the framework of the locations and contexts in which events occur and that this contextual framework encodes relations among an events different aspects in a way that allows later retrieval from memory.
This view has been hotly debated for years. Yet a consensus is emerging that the hippocampus does somehow provide a spatial context that is vital to episodic memory. When you remember a past event, you remember not only the people, objects and other discrete components of the event but also the spatiotemporal context in which the event occurred, allowing you to distinguish this event from similar episodes with similar components.
Despite intensive study, however, the precise mechanisms by which the hippocampus creates this contextual representation of memory have eluded scientists. A primary impediment was that we knew little about the brain areas that feed the hippocampus its information. Early work suggested that the entorhinal cortex, an area of cortex next to and just in front of the hippocampus [see box on next page], might encode spatial information in a manner similar to that of the hippocampus, though with less precision.
This view has now been turned upside down with the amazing discovery of a system of grid cells in the medial entorhinal cortex, described in a series of recent papers by the Norwegian University of Science and Technologys Edvard Moser and May-Britt Moser and their colleagues. Unlike a place cell, which typically fires when a rat occupies a single, particular location, each grid cell will fire when the rat is in any one of many locations that are arranged in a stunningly uniform hexagonal grid--as if the cell were linked to a number of alarm tiles spaced at specific, regular distances. The locations that activate a given grid cell are arranged in a precise, repeating grid pattern composed of equilateral triangles that tessellate the floor of the environment [see box on next page].
Imagine arranging dozens of round dinner plates to cover a floor in their optimal packing density, such that every plate is surrounded by other, equidistant plates; this arrangement mimics the triggering pattern tied to any given grid cell. As the rat moves around the floor, a grid cell in its brain fires each time the rat steps near the center of a plate. Other grid cells, meanwhile, are associated with their own hexagonal gridworks, which overlap each other. Grids of neighboring cells are of similar dimensions but are slightly offset from one another.
These grid cells, conclude the Mosers and their co-workers, are likely to be key components of a brain mechanism that constantly updates the rats sense of its location, even in the absence of external sensory input. And they almost certainly constitute the basic spatial input that the hippocampus uses to create the highly specific, context-dependent spatial firing of its place cells.
This discovery is one of the most remarkable findings in the history of single-unit recordings of brain activity. Reading the paper announcing this discovery in my office for the first time, I realized immediately that I was reading a work of historic importance in neuroscience. No one had ever reported a neural response property that was so geometrically regular, so crystalline, so perfect. How could this even be possible? Yet the data were convincing. "This changes everything," I muttered.
My excitement rose partly from the sheer beauty of the grid-cell response pattern. But it rose, too, from a belief that this was a major step in our quest to understand how the hippocampus might form the basis of episodic memory. Grid cells give us a firm handle on what kind of information is encoded in one of the major inputs into the hippocampus. From this foundation we can start to create more realistic models of what computations occur in the hippocampus to transform these grid representations into the more complex properties that have been discovered about place cells over the past three decades. For example, different subsets of place cells are active in different environments, whereas all grid cells appear to be active in all environments. How is the general spatial map encoded by grid cells turned into the environment-specific (or context-specific) maps encoded by place cells?
Moreover, the discovery of grid cells affirms emphatically that the hippocampus and medial temporal lobe are outstanding model systems for understanding the way in which the brain constructs cognitive representations of the world "out there" that are not explicitly tied to any sensory stimulation. There is no pattern of visual landmarks, auditory cues, somatosensory input or other sensations that could possibly cause a grid cell to fire in such a crystalline fashion across any environment. This firing pattern--which is similar regardless of whether the rat is in a familiar, lit room or in a strange location that is pitch-dark--must be a pure cognitive construct. Although grid cell firing patterns are updated and calibrated by sensory input from the vestibular, visual and other sensory systems, they do not depend on external sensory cues.
Some have argued that hippocampal place cells are similarly independent. But the known influence of external landmarks on place cells, and their tendency to fire in single locations, led others to argue that place cells are driven primarily by unique combinations of sensory landmarks that exist at particular locations. This argument cannot explain the firing patterns of grid cells.
The Road Ahead
So what does account for grid cell dynamics? One possibility is that these cells allow an animal to constantly update its physical location on its internal cognitive map by keeping track of its own movements. That information is in turn conveyed to the hippocampus, which combines this spatial representation with other data about an event to create specific, context-rich memories of unique experiences--the ability that Memento's Lenny had lost.
The discovery of grid cells has generated a palpable sense of excitement. We can anticipate that further research into grid cells (along with the other major input to the hippocampus, the lateral entorhinal cortex) will reveal the neural mechanisms that let us remember our personal histories--a vital process that forms the very foundation of ones sense of identity.