An animal’s assessment of its position in space is represented in several interconnected brain areas. In the entorhinal cortex, grid cells are active when the animal visits one of multiple locations, which are arranged in 2D space in a periodic pattern. In each individual animal, there are several distinct groups of grid cells called modules. Cells within each module are characterized by a unique spatial periodicity, and it is thought that all cells within a module participate in a highly interconnected neural network. The work of Waaga et al. asks whether the multiple modules continue to represent a unique trajectory coherently in space, even when the animal is confused about its location. To address this question, Torgeir Waaga and his Norwegian colleagues used high-density Neuropixels silicon probes to record the neural activity in rats while the animals were foraging in a dark environment. Importantly, this is the first work to decode the representation of position based on activity recorded from multiple modules. Haggai Agmon from Yoram Burak’s lab developed and applied sophisticated methods of analysis to decode the position, and to test the hypothesis that different modules remain coordinated even when they represent a position that deviates from the actual position of the animal.

Video clip: Simultaneous decoding of position from several grid cell modules during 30 seconds of motion in a dark environment. Top-left: Decoding position from all modules together. The actual position of the animal is denoted by the black cross, and the position decoded from all modules together is denoted by the red circle. Top-right and bottom panels: Decoding position only from cells belonging to specific modules (three modules shown). Since all cells in a module respond with the same spatial periodicity, the decoded position in these panels is periodic and ambiguous. However, by pooling the modules together, it is possible to decode the unique position (as shown in the top-left panel). Because the animal is foraging in dark conditions, a mismatch develops between the cross and the circle (such a mismatch is not observed in trials performed in light conditions). Despite this mismatch, the neural representations of position in the different modules remain coordinated.