How the Brain Encodes Space and Place

Some people have a great sense of direction. Even after wandering around town all day, stopping for lunch and perusing some stores and taking a few detours to compare which coffee shop looks the best, they still know exactly how to get back.

Then there are those of us who get turned around and have to consult Google Maps every time we get out of a subway station. I presume I inherited this trait from my mom, who takes ages to find her car in the mall parking lot and says she once called the cops after being convinced it must be stolen.

Even though some people have a much better sense of direction than others, this is something that we all possess. All of us have an idea of what our surroundings look like and where we are located within them.

How do our brains create this perception of space?

Maps in the brain

It turns out that our brains contain a spatial map of the world. You probably heard something about this back in 2014 when the Nobel Prize was awarded to three scientists who made fundamental discoveries in this area.

But what does that really mean? What do these “maps” in the brain actually look like?

What neuroscientists mean is that there are brain cells that fire depending on where we are and where we’re going. These cells come in a bunch of different types, each with a unique role. Together, these cells enable us to navigate the crazy world we live in.

Cells that know where you are

The first cells that form part of our brain’s internal map were discovered by John O’Keefe and colleagues in the late 1960s. They were recording the activity of cells in the hippocampus, a brain region originally known for its critical role in learning and memory.

Hoping to find brain cells related to memory, O’Keefe recorded the activity of hippocampal cells in rats while they performed various memory tasks. But unlike others at the time, he didn’t just record brain activity during strict task-related time periods; he also recorded brain activity while the rats were simply roaming around their cages and going about their daily business. This was what enabled him to make a novel observation.

Place cell firing. The grey line depicts the path of a rat exploring a square box. Red dots show where the rat was when the cell fired. (from Derdikman & Moser, 2010)

O’Keefe noticed many mysterious cells in the hippocampus that were mostly inactive, but fired profusely at sporadic intervals. Their firing wasn’t related to memory, nor to anything specific the rat was doing, like eating or sleeping. Then in a true “eureka” moment, he finally figured it out: these cells cared about the rat’s location in the cage.

Each cell fired whenever the rat was passing through a particular place, like one of those big “YOU ARE HERE” markers on a map. One cell might fire when the rat was in the far left corner, whereas other cells would fire when the rat was in the far right corner, or the center, or a bit left of center… you get the idea. These cells were named “place cells”.

Place cells are really cool because they represent an abstract idea—place—that most of us think of as a pretty high-level concept. O’Keefe and colleagues showed that place cells don’t just fire in response to simple sensory cues, like the color of the wall or the shape of the doorway. Instead, place cells synthesize information from many cues and landmarks in the environment in order to represent the conceptual notion of “place”.

The place cell map

Even though each place cell only fires when you’re in a single location, if you record enough cells you can find a place cell for any given location. That’s why these cells, together, form a map of space.

Different place cells fire in different places. This figure shows the firing of 32 different place cells relative to the rat’s location in a square box (red = higher firing, blue = lower firing). (from O’Keefe, Nobel Lecture, 2004)

Unlike O’Keefe’s rats, we don’t normally spend all our time hanging out in one room. That’s why it’s important that the hippocampus can maintain multiple maps corresponding to different environments.

We know that these are different maps because the same place cell typically responds to different locations within different environments: a cell that fires when you’re in the far right corner of your bedroom might fire when you’re in the far left corner of your office. So single cells can’t actually tell you whether you’re in your bedroom or your office, but combinations of place cells with different response properties collectively create a map of each environment.

Astute readers might be wondering where the heck place cells come from. After all, it’s not like we’re born knowing what our future homes and offices will look like. That’s why any internal map of external space clearly requires experience—sensing and exploring the environment to keep track of spatial cues and landmarks. Scientists are still trying to figure out how exactly this happens.

Another layer to the map

Experiments aimed at determining where the “place signals” of hippocampal place cells come from actually led to the discovery of a second type of “map” cell. Throughout the 1980s and 1990s, scientists believed those place signals were computed within the hippcampus itself, as opposed to being transmitted from a different brain area.

But two Norwegian scientists, Edvard and May-Britt Moser, performed experiments that cast doubt on this theory. By cutting off different inputs to the place cells, the Mosers obtained data suggesting that a nearby brain region, the entorhinal cortex, might transmit “place” information to the hippocampus. If so, the entorhinal cortex should contain place cells too.

grid-cell-firing

Grid cell firing. As above, the black line indicates the path of a rat exploring a square box and red dots show where the rat was when the cell fired. The firing pattern forms a hexagonal grid. (from Moser et al., 2015)

So the Mosers decided to record activity from the entorhinal cortex and look for place cells. At first they got excited: they saw cells that clearly fired when the rat passed through specific locations. But these cells looked different from hippocampal place cells. Mainly, they didn’t just fire in one location, like normal place cells; each cell fired when the rat was in several different locations in the cage.

Once the Mosers started letting their rats roam larger arenas, the pattern became obvious: each entorhinal cell fired in regularly spaced locations that formed a hexagonal grid. The Mosers named them “grid cells”.

Unlike place cells, grid cells fire in the same exact grid-like pattern even in totally different environments. The grids formed by different grid cells differ in multiple ways: they can be offset to one another, oriented at a different angle, or more widely or narrowly spaced. Like the place cells, then, the grid cells also cover all points in space.

Living on the grid

The Mosers suggested that grid cells form a coordinate system that helps our brain read out the spatial map in the hippocampus. As anyone who’s ever read a map should know, it’s a lot easier to calculate distances and routes between different points if the map contains an overlaying grid. More widely spaced grids might be important for reading a larger, more zoomed out map, and vice versa for more narrowly spaced maps.

1024px-hopwas_woods_sun

Grid cells may help you judge distances and routes when everything looks the same. (credit: Bs0u10e01 via Wikimedia commons)

In addition to working together, it’s also possible that place cells and grid cells allow us to navigate in different contexts. Place cells may help us recognize where we are based on landmarks, which is important when you’re walking through your neighborhood or driving your familiar route to work. Conversely, grid cells may help us navigate in the absence of recognizable cues, like in the middle of the woods or in a foreign city where everything seems to look the same.

Personally, I think grid cells are even cooler than place cells. Sure, place cells encode the super important and abstract idea of place, but since they need to be oriented to landmarks in the environment they’re sort of an extension of your sensory representation of the world. Grid cells, on the other hand, seem to be a completely internal representation of external space. They might be responsible for that intrinsic, gut sense of direction that we sometimes have even when we don’t recognize any of our surroundings.

How grid cells and place cells interact is still a subject of intense investigation. Going back to the Mosers’ original question, if the entorhinal cortex sends place information to the hippocampus, does that mean that place cells are created from grid cell signals? Or is it a different type of entorhinal input that matters?

Other spatial cells

Because the 2014 Nobel Prize was awarded to O’Keefe and the Mosers, it’s the place cells and grid cells that have recently gotten all the glory. But there are also other important cell types that help us navigate the world. The most well-studied are the head-direction cells, which were discovered by James Ranck in the 1980s. As you can probably guess, head-direction cells fire based on the direction your head is facing, thus encoding where you’re going. Clearly this is important information if you want to actually move around in the world and not just sit in one place all day.

seelig-fig-2

Head-direction cells have even been discovered in fruit flies, and are arranged into a compass-like structure called the ellipsoid body. This figure shows that different “wedges” of the compass are active when the fly is facing different directions. The top row shows the panoramic “scene” that the fly is viewing (a bunch of moving stripes); as the stripes move, the fly turns to face different directions. (from Seelig and Jayaraman, 2015)

Environmental cues orient head-direction cells and keep them properly aligned, but these cells also fire in the absence of any cues. This is why you still have a sense of which way you’re facing even in total darkness.

This is also why, as a 10-year old, I was convinced I could ride my bike around our driveway with my eyes closed. That ride did not end well. Actually, that ride reflects an important property of head direction cells: the longer you’re cut off from external cues, the more misaligned your head-direction cells get, and the more you lose your sense of direction until eventually you’re lost (or crash into the garage door).

In addition to the head-direction cells, there are “border cells” that tell you when you’re near a wall and “speed cells” that fire depending on the speed at which you’re moving. There have also been reports of cells encoding the direction and distance to a specific goal, such as food. I suspect that we will soon discover a vast array of cells that encode information about everything you’d ever need to navigate in the world, including information about yourself as well as other objects and cues.

What’s in a map?

But that doesn’t mean that we’ve solved the problem of how spatial navigation works. For one thing, there’s very little evidence demonstrating that any of these cell types are actually necessary for navigation, despite their intriguing firing properties. Even if they are, we still need to figure out exactly how the brain uses the information that they encode. Having a map isn’t the same as knowing how to use it.

olympicmtbtestevent11-gouldpodium

Making 1st, 2nd, and 3rd place winners stand at different heights reflects our tendency to project abstract ideas (sporting performance) onto space. (credit: usacycling.org)

Moreover, neuroscientists still argue about whether the apparent spatial map in the hippocampus truly evolved for the purpose of navigation. Some people think that this spatial map serves primarily to encode memories: events are always associated with a particular place, so linking the place to the event may be a handy way of “indexing” your memories. This indexing system seems plausible: just think about a specific place you’ve visited, and your memories of things that happened there will probably come flooding back effortlessly.

Another theory is that the hippocampus doesn’t just represent physical space but more generally encodes “cognitive space”, which is the abstract framework in which you organize your thoughts and ideas. For example, you probably have a mental “map” of your family tree or your schedule for the week. The spatial map could represent just one example of how the hippocampus and place cells organize information.

So despite all the cool “map” cells that scientists have discovered over several decades, the jury’s still out on what they really do, especially place cells. Are they a roadmap to the world? A record of our past? A framework for cognition? A coded cipher leading to the national treasure? Or something else that we haven’t even thought of yet?

 

Note: this post was largely inspired by a recent symposium honoring James Ranck, “Sense of Direction and the Cognitive Map”, on September 9, 2016.

References:

Derdikman D, Moser EI. A manifold of spatial maps in the brain. Trends Cogn Sci 14:561-569 (2010).

Eichenbaum H, Dudchenko P, Wood E, Shapiro M, Tanila H. The hippocampus, memory, and place cells: is it spatial memory or a memory space? Neuron 23:209-226 (1999).

Fyhn M, Molden S, Witter MP, Moser EI, Moser MB. Spatial representation in the entorhinal cortex. Science 305:1258-1264 (2004).

Hafting T, Fyhn M, Molden S, Moser MB, Moser EI. Microstructure of a spatial map in the entorhinal cortex. Nature 436:801-806 (2005).

Kubie J. Place cells, remapping, and memory. 5 Oct 2013. Posted on BrainFacts.org: http://blog.brainfacts.org/2013/10/place-cells-remapping-and-memory/

Kumaran D, Maguire EA. The human hippocampus: cognitive maps or relational memory? J Neurosci 25:7254-7259 (2005).

Moser EI. Grid cells and the entorhinal map of space. Nobel Lecture, December 2014.

Moser M-B. Grid cells, place cells, and memory. Nobel Lecture, December 2014.

Moser M-B, Rowland DC, Moser EI. Place cells, grid cells, and memory. Cold Spring Harb Perspect Biol 7:a021808 (2015).

O’Keefe J. Spatial cells in the hippocampal formation. Nobel Lecture, December 2014.

O’Keefe J, Dostrovsky J. The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Res 34:171-175 (1971).

Seelig JD, Jayaraman V. Neural dynamics for landmark orientation and angular path integration. Nature 521:186-191 (2015).

Taube JS, Muller RU, Ranck JB Jr. Head-direction cells recorded from the postsubiculum in freely moving rats. I. Description and quantitative analysis. J Neurosci 10:420-435 (1990).


Leave a Reply

Your email address will not be published.