Every New Yorker has observed this at least once: the rodents that inhabit the subway show the remarkable ability to immediately find the closest place to hide whenever they feel in danger. When we get off of the subway at a station we have never been to before, or at a different exit than usual, we feel disoriented until we notice a well-known landmark—most of the time it’s the Empire State Building. Suddenly, everything makes sense: I know which way north is, it’s morning and the shadows are pointing westward, I’m walking to the lab, toward the East River.

It turns out that mammalian (and certain bird) brains are equipped with a highly sophisticated navigation system that integrates every piece of available information to locate the animals in their environment. This navigation system represents a major evolutionary advantage: it allows animals to return to the places where they previously found food or water, for example, sometimes over long distances. However, a recurrent question in neuroscience is to understand how this system is able to imagine new routes, future outcomes, or, generally speaking, to disengage from sensory inputs and deliberate between different options. This is a very difficult experimental task.

We chose to focus our work on a simple component of this brain’s global positioning system (GPS): the neuronal compass, also known as the head direction system, which at any time gives an animal directional orientation relative to its environment. We asked a simple question: can this compass exist without sensory inputs?

Using cutting-edge electrophysiology techniques, we monitored the activity of the head direction system in freely moving mice during waking, as they foraged for food in an open arena, and sleep. During waking, we were able to reconstruct the internal sense of direction by decoding the activity of ensembles of simultaneously recorded neurons: the decoded “needle” of the “brain compass” was indeed pointing toward the actual direction of the animal’s head.

More surprisingly, we noticed that this needle continued to exist during sleep, and we decoded the “virtual gaze” of the animals while they were resting. Thus, the brain compass does not require any external inputs to code for a direction and functions perfectly when disengaged from the external world. Even more interestingly, during rapid eye movement (REM) sleep, when dreaming activity is the most intense in humans and the electrical activity of the brain is virtually indistinguishable from being awake, this needle spontaneously moved at the same speed as observed when the animals were exploring the arena. In other words, this system is able to simulate an imaginary journey at actual speed.

Is it specific to this brain compass or is it the case of the entire navigation system? This will be the object of further studies. In particular, we will now try to understand how the information provided by this neuronal compass influences downstream structures.

—Adrien Peyrache, PhD

Read the paper “Internally organized mechanisms of the head direction sense” in Nature Neuroscience, published March 2, 2015.