Cheetah software

After electrode implantation, the animals were allowed to recover for 7 days before being habituated to the recording setup consisting of a home-cage without the lid placed in the middle of a large Plexiglas box used as secondary containment. During 24 hour LFP and EMG recordings, the electrophysiological tethers were connected to a motorized commutator (NeuroTek Inc.). The hippocampal LFP and neck EMG signals were amplified, filtered (0.1–2000 Hz) and digitized at 4 kHz using a Digital Lynx SX System (Neuralynx, Inc.) and Cheetah software (Neuralynx, Inc.).

To characterize effect of HIFU on brain surface activities, we used µECoG arrays (Neuronexus) with 16 small recording electrodes (surface area ~0.03 mm²) arranged in a 4 × 4 grid, allowing us to sample small populations of neurons from the superficial layers of cortex at high resolution. Epidural neural signals were acquired with a NeuraLynx Digital Data Acquisition System, including a DL 4SX-M 32ch Base, a headstage pre-amplifier HS-18-CNR-MDR50, a HS-18-N2T-16 adaptor and Cheetah data acquisition software, at a sampling frequency of 16 kHz from freely moving animals in their home cages. High pass filtered (HPF; 900 to 5 kHz) and ECoG (0.1 and 300 Hz) signals were saved with Cheetah software (NeuraLynx). The animals’ behavioral state was simultaneously recorded and classified into Move or Resting status with HomeCageScan software (CleverSys) during electrophysiological recordings.
Ipsilateral and contralateral ECoG data were collected from different cohorts of animals. Contralateral animals were allowed to recover from implantation surgery for a week, then four-week baseline data were collected (Fig. 1A, n = 10 in both HIFU and sham groups). In ipsilateral cohort, animals were subjected to HIFU/sham treatment immediately before ECoG array implantation surgery, therefore, no baseline data were collected (Fig. 1A in purple, n = 4 in both HIFU and sham groups).

Tetrode signals were recorded with a Neuralynx Digital Lynx system using Neuralynx Cheetah software (Neuralynx, MT, USA) (Matsumoto et al., 2020 (link)). Spikes were detected online, and the multi-unit activity was sorted into single-unit activity offline with Spike Sort 3D software (Neuralynx, MT, USA). Cluster classification was performed using previous methods (Matsumoto et al., 2020 (link)). Briefly, clusters were considered to represent single units if the isolation distance was ≥ 20 and L-ratio ≤ 0.3. Only units with a peak-to-peak width greater than 250 μs were classified as putative excitatory neurons and were used for this study.

In all conditions, behavioral data were analyzed using custom-made Matlab scripts. Instantaneous speed (or virtual speed) was derived from video tracking data (or virtual X and Y coordinates recorded as voltage signals in Neuralynx Cheetah software) and downsampled to 10 Hz for consistency with spectrogram and coherogram data (see below). Only epochs of active movement, defined by a sustained speed above 3 cm/s for a minimum of 4 s, were selected for further analysis.
In the home-cage environment, there was one case in which video tracking was not available and the epochs of active movement were thus defined by an amplitude threshold in EMG signal and the presence of a high theta/delta ratio in the hippocampal recordings.
For linear track and virtual reality-based experiments, active movement periods (instantaneous speed >3 cm/s) were analyzed from each 12 min session (three sessions per day) for overall calculations. Each session was then subdivided in to trials (runs), that is, each time the animal traversed the track from one reward site to the other, and measurements of speed, power and coherence were averaged by distance from reward location in 1 cm bins normalized by the occupancy (time spent in each bin).