Circuit Dynamics and Computational Neuroscience I.1.b Neural network models Monday AM + Wednesday AM

2460 - Drift and stabilization of hippocampal response selectivity

Topic / Sub Topic
I.1.b Neural network models
whenever - just send me an Email via
My link to connect:


Abstract Body

Place cells are a core functional characteristic of the hippocampal coding of space. Technological and methodological advances now allow for large scale and longitudinal recordings of neuronal activity. Accumulating evidence suggests that various hippocampal regions display a turnover of place cell coding with a timescale on the order of days and weeks. While the overall dynamics and its possible function has received substantial attention, the individual dynamics of the cells’ place fields remain incompletely characterized.

We analyzed two-photon activity imaging data of neurons in the hippocampal CA1-region of behaving mice unidirectionally navigating a linear, circular virtual environment in which they encountered visual, as well as reward stimuli at fixed locations, Fig. a).

We tracked neurons over up to 100 sessions and characterized their place coding behavior. Cells could switch from place encoding to non-coding, or relocate their place field position, Fig. b). Specifically we found that

1) continuously coding cells largely maintain their place field in a closely circumscribed region of space, but have a low probability of long-range relocation, Fig. c)

2) periods of non-coding facilitates random relocation, with the relocation probability increasing with the duration of non-coding episodes, Fig. d)

3) transition rates between coding and non-coding states, as well as relocation probabilities are modulated by visual salience of place-field-location

In conclusion, we find that place field dynamics in CA1 consists of localized random motion, interrupted by long-range random relocation. Periods of non-coding activity are associated with place field turnover by promoting chances of place field relocation.