Positive acceleration selectively controls theta frequency in the hippocampus and entorhinal cortex

MAY-BRITT MOSER; ERIC CARMICHAEL; KROPFF, EMILIO; EDVARD I MOSER

Conferencia; Society for Neuroscience Annual Meeting 2018; 2018

Local field potential oscillations provide clues to understand the dynamics of large populations of neurons. In rats, the theta rhythm organizes neural activity across brain structures including the hippocampus and entorhinal cortex. Recent evidence points to a role of the theta cycle in packing together related information to produce an epoch of local computation. An additional role in spatial navigation is supported by decades of research reporting that theta frequency encodes running speed linearly, so that displacement can be obtained through theta frequency integration. Here we show, however, that this relationship is an artifact caused by the fact that, until the recent introduction of the bottomless car paradigm, the speed and acceleration of rats could not be systematically disentangled. Our results indicate that theta frequency is linearly related to positive acceleration alone, and not modulated by negative acceleration or speed (Kropff et al., SfN 2017). The rhythmic spiking of neurons follows a similar pattern, implying that both theta frequency and rhythmic spiking are nonholonomic and thus non-univocally related to displacement or any other kinematic variable. This control of theta frequency by acceleration does not require visual or motor cues, and is independent of phase precession-related variations of intrinsic firing frequency. Our results suggest that variations in theta frequency reflect a temporally precise mechanism for speeding up computations in the entorhinal-hippocampal circuits. We extend previous reports by speculating on the purpose of frequency increase. The entorhinal path integrator presumably uses information about speed but not about acceleration, as suggested by failed attempts to find acceleration-coding cells. During acceleration episodes, it should make errors in the estimation of the rat´s trajectory, causing an alteration in grid spacing, as suggested by preliminary evidence that we here provide and discuss for the first time. Performing a greater number of shorter computational steps during these acceleration episodes might help the system to improve statistics on one hand and to reduce the magnitude of these errors by reducing the temporal integration window on the other.