Activity patterns in the septal-hippocampal network predict voluntary locomotion

Abstract

In the brain of animals, locomotion is encoded and represented in multiple ways. During locomotion, the hippocampus (HC) displays characteristic activity patterns that change from asynchronous states when the animal is resting to synchronous rhythmic activity during movement. The increase in firing rates of principal neurons in CA1 and the presence of oscillations in the HC both correlate to the velocity of the animal. It has been shown previously that glutamatergic neurons in the medial septum (MS) increase their activity prior to movement onset. However, the time-course of activation of individual MS neuron types during an episode of locomotion is unknown. <br /> I investigated the MS-HC circuitry with cell-type specificity by expressing the genetically encoded calcium indicator GCaMP5G in inhibitory (PV⁺) and excitatory (VGluT2⁺) cells of the MS. I have monitored activity-dependent changes in fluorescence with a fiberoptometer coupled to an implanted fiber optic cannula in head fixed mice on a linear treadmill. In addition, I obtained CA1 local field potentials and recorded multi-unit activity in both CA1 and the MS. I aligned and correlated the recorded parameters with different phases of locomotion (onset, acceleration, deceleration, offset). My results show that there is a significant representation of locomotion in both CA1 and MS neuronal populations. I demonstrate that both glutamatergic VGluT2⁺ and GABAergic PV⁺ cells in the MS show an increase in activity several hundred milliseconds before movement.<br /> My experiments provide evidence on the single neuron activity level for CA1 and MS cellular activity that predicts movement onset. A simultaneous activation of glutamatergic and GABAergic neurons within the MS suggests the activation of an excitatory-inhibitory feedback loop controlling motion execution and HC information processing.