Grid Cell Phase Precession to Cognitive Map Updating

• Grid Cells: Specialized neurons in the medial entorhinal cortex (MEC). They fire at locations forming a regular, tessellating hexagonal pattern across an environment, providing a fundamental spatial metric or "coordinate system."
• Phase Precession: A phenomenon where, as an animal moves through a grid cell's (or place cell's) firing field, the neuron's spikes occur at progressively earlier phases of the ongoing **theta oscillation** (a prominent brain rhythm, typically 6-10 Hz in rodents).
• Cognitive Map: The brain's internal, allocentric (world-centered) representation of an environment. It's a mental model that enables understanding location, efficient navigation, and memory of spatial relationships, even without direct sensory cues.

Phase precession in grid cells (and place cells) is hypothesized to serve several critical functions that contribute to the ongoing updating and refinement of cognitive maps:

• Path Integration: Grid cells are crucial for path integration – tracking one's continuous position by integrating self-motion cues. Phase precession enables this by encoding precise distance moved within a grid field through spike timing relative to the theta rhythm, allowing for continuous positional updates in the cognitive map.
• Spatial Compression and Learning: Phase precession compresses sequences of spatial locations into short timescales (less than a single theta cycle). This temporal proximity facilitates **spike-time-dependent plasticity (STDP)**, strengthening synaptic connections between neurons representing sequentially visited locations, thus refining the cognitive map's spatial relationships.
• Error Correction and Map Alignment: Path integration can accumulate errors. External landmarks and sensory cues help correct these. Phase precession, especially with place cells, is thought to align the internally generated grid map (from path integration) with external sensory inputs, enabling error correction and maintaining map accuracy.

Neuroscientists use various computational models to understand how grid cell phase precession supports cognitive map updating:

• Oscillatory Interference Models: Propose that grid cell firing patterns and phase precession arise from the interference of multiple neural oscillations, whose frequencies are modulated by velocity. The phase of the resulting interference encodes spatial position, with phase changes generating precession.
• Continuous Attractor Network (CAN) Models: Represent grid cells as a network where persistent "bumps" of activity correspond to the animal's position. Movement of these bumps, driven by velocity input, simulates navigation, and phase precession emerges as neurons within a moving bump fire at progressively earlier phases.
• Hierarchical Models: Integrate grid cells with place cells, head direction cells, and border cells, suggesting a complex hierarchy of spatial representations. Phase precession in grid cells provides metric input to place cells, contributing to the cognitive map's richness and flexibility, often via STDP-driven learning rules.