If you are at all interested in how the brain or mind works, for example if you study psychology, neuroscience or philosophy of mind, then you ought to know about place cells.
Images: Image 1 shows the path of a rat moving around a rectangular box (black line). Red squares show where a place cell fires. Image 2 shows the average firing rate of the cell at each location in the rectangular box – “hotter” colours indicate more rapid firing. Data from O’Keefe lab.
As an animal* moves freely about in its environment, cells in the hippocampal formation signal its location, heading and speed in exquisite detail. These spatial cells come in a variety of different forms, and I plan to cover some of the other types in future posts. Place cells, which are found in the hippocampus, fire when the animal is at a particular location. In an open environment their firing rates are largely independent of the animal’s heading or the presence or absence of particular cues. Place cells were discovered in the 1970s by O’Keefe and Dostrovsky.
OK, why are these Place Cells so important?
Most other classes of neuron whose properties have been so far been studied fire in response to very particular types of stimulus or drive specific actions. For example a cell in auditory cortex might fire preferentially to a particular frequency of sound. One in visual cortex to a moving patch light with a particular retinal location and orientation. A somatosensory neuron might fire to the touch of a specific part of the body and so on. A cell in motor cortex might help to control a certain actions, for example reaching in a given direction. These sensory and motor neurons seem to represent concrete properties of the outside world and concrete actions.
The remarkable thing about place cells is that their firing is not related to individual sensory cues or to concrete behaviours in any simple way. Instead it seem the simplest way to understand the spatial code in the hippocampal formation is in terms of location and heading – i.e., the physical relationship between the animal and its environment as a whole. It’s a highly abstract form of information, far removed from the type which arrives at the sense organs.
Place cell activity is not completely tied to the environment, though. When an animal stops or when it is asleep, the place cells fire in ordered sequence, apparently replaying its experience. This patterned activity is thought to play a role in learning and memory consolidation.
So place cells give us a concrete insight into how abstract information about the world around us is represented in the brain and potentially how that information is used to form memories and guide complex behaviour. Although other cells with related properties have since been discovered, place cells have been the most closely studied and are the best understood.
How do place cells know where to fire?
Images: When the shape of the environment is changed, place fields (the areas where a given cell fires) tend to “stick” to the boundaries. Image 1: Data from O’Keefe and Burgess (1996). Image 2: Data from Lever et al. (1999).
An important clue comes from experiments showing that a given place field will tend to maintain a fixed-distance to part of the boundary when the size or shape of the environment changes: place fields typically “stick” to part of the boundary.
With Neil Burgess and colleagues, I developed a model based on the idea that place cells have inputs that respond to boundaries at particular distances and compass directions. We found that patterns of place cell firing across different environments can be explained assuming inputs from a small number of these hypothetical “boundary vector cells”, which “tell” the place cells where to fire. This very simple model can also predict the firing field of a given cell in a new environment. Amazingly enough the input cells with the very properties we predicted were discovered eight years later.
Images: The pattern of place cell firing in different environments can be explained by assuming place cells receive input from “boundary vector” (or “border”) cells which respond at a fixed distance and compass direction from the animal. We predicted these input cells and their properties in 2000 (Hartley et al., 2000). They were discovered in 2008 (Solstad et al., 2008).
What do we use place cells for?
The discovery of place cells led John O’Keefe and Lyn Nadel to develop the concept of the hippocampus as a cognitive map, a representation of the environment, independent of one’s direct sensory experience which allowing one to form enduring memories and plan spatial behaviours that extend beyond the horizon of one’s current experience. For example, you could use the cognitive map to plan a direct route from your home to your place of work by a new route, avoiding an unexpected roadblock.
This is “just” a theory, but the nature of place cell firing is certainly suggestive. Cells with similar properties have been found in humans*.
Consistent with the idea of the cognitive map, people who navigate more accurately and directly (in virtual reality) show greater activation in the hippocampus when performing the task in an fMRI scanner.
Image: parts of the human brain, including hippocampus, which show greater activity in better navigators when finding new routes (compared with following a visible trail). Hartley et al., 2003.
The size and shape of the hippocampus seems to make a difference. London taxi drivers who are expert navigators have larger (or at least different shaped – the back end gets larger, the front end gets smaller) hippocampi than non-experts or bus drivers (who in their jobs follow the same set of routes repeatedly, rather than needing to find new routes for each passenger).
As with any complex behaviour many different brain mechanisms can contribute to navigation. When following a very familiar route (akin to the daily commute, or a bus driver’s route, perhaps) better navigators activate another brain region, the caudate nucleus. This is thought to be involved in learning fixed habits which become automatic, so it could provide a kind of autopilot that we use for routines which have become so familiar and well-practiced we no longer need to plan them. Next time you find yourself on the way to your school or workplace having set off for a different destination; blame the caudate nucleus. When you find your way there by a direct route despite having to take a detour – thank your hippocampus and its place cells.
Image: parts of the human brain (right caudate nucleus) which show greater activity in better navigators when following a fixed route (compared with finding new direct routes). Hartley et al., 2003.
*most of the experiments are done with rats, but some have been done with other animals including human patients who have had electrodes implanted in preparation for surgery to treat epilepsy.