OF LABELED-LINES AND MANIFOLDS: STIMULATION OF THE SAME MEDIAL ENTORHINAL CORTEX INPUTS IN DIFFERENT ENVIRONMENTS LEADS TO ORTHOGONAL SPATIAL REPRESENTATIONS IN CA1

Upon exposure to a novel environment, hippocampal place cell firing fields rearrange unpredictably to form spatial maps that are orthogonalized between different environments, a process called remapping. How this process happens, and indeed, the structure of hippocampal spatial representations, remain a subject of debate. At one “labeled-line” extreme, changes in the activity along anatomically-specified inputs generates de novo spatial firing patterns. At the alternative extreme, “manifold” models explain place field remapping by tying changed inputs to a pre-existing, internally-organized pattern of preserved coactivity. We previously showed that chemogenetic depolarization of ~25% medial entorhinal cortex layer II (MECII) stellate cells causes drastic changes in CA1 spatial firing resembling remapping, but without environmental changes. Here, we performed calcium imaging of CA1 neurons undergoing this “artificial remapping” in two different familiar environments to investigate whether changing MECII input activity determines CA1 place cell firing according to labeled-line or manifold models. Because artificial remapping is caused by depolarizing the exact same MECII neurons in both environments, labeled-line models would predict similar changes to place cells during artificial remapping in both environments, unlike manifold models. We observed no similarity in cell-specific spatial firing changes during artificial remapping in the two environments, but found that, across environments, changes of the coactivity patterns in the population activity were suppressed by inducing the same stimulation. Consistent with manifold but not labeled-line models, although artificial remapping is caused by the same functional changes in the same neurons, CA1 spatial maps are orthogonalized, while changes in coactivity patterns are reduced.