Date: 23 April, 2024

Dataset Title: Data in support of the study “Sleep loss diminishes hippocampal reactivation and replay”

Dataset Creators: B. Giri, N.R. Kinsky and K. Diba

Dataset Contact: Kamran Diba kdiba@umich.edu

Funding: NIMH R01MH117964 and NINDS R01NS115233

Key Points:
- We examine and compare the effects of sleep and sleep deprivation on hippocampal sharp-wave ripples and associated firing patterns in behaving rats following exposure to a novel linear maze
- Our study reports sustained rates of sharp-wave ripples but at elevated ripple frequency and lower ripple power during sleep deprivation. 
- In neuronal firing patterns, we observed diminished reactivation and replay during sleep deprivation and the subsequent recovery sleep.

Research Overview:

This study investigates the impacts of sleep and prolonged wakefulness on neuronal activity in the CA1 region of the rat hippocampus. We conducted extracellular recordings of local field potentials and single neuronal units in the CA1 area, categorized into 754 pyramidal neurons and 96 interneurons, during different states of rest and sleep before the light cycle (PRE), active exploration in novel mazes (MAZE), and post-maze sessions which included ad-libitum sleep and rest (NSD) or sleep deprivation (SD) followed by recovery sleep (RS). Recordings commenced approximately 3.5 hours before the light cycle onset, with the animals initially resting in their homecage for around 2.5 hours. After exploring novel mazes for rewards for approximately one hour, the rats were either allowed to sleep naturally for 9 hours or were sleep-deprived through gentle handling for 5 hours before being permitted recovery sleep. We segmented these post-maze sessions into 2.5-hour intervals, synchronizing them with the onset of the light cycle (zeitgeber time = 0), and compared the neural activity during the initial ad-lib sleep (RS vs. NS1) and after prolonged wakefulness (SD2 vs. NS2). The SD and NSD sessions, from a total of 7 animals, were conducted in pseudo-random order on non-consecutive days. The study's aim is to elucidate the differential effects of sleep states and deprivation on hippocampal neuronal dynamics, which may provide insights into the role of sleep in memory and learning processes.

Methodology:
The data were obtained using 128 channel high-density silicon probes implanted uni- and bilaterally in the dorsal CA1 region of the hippocampus of freely-moving Long Evans rats. Sleep deprivation was performed at the onset of the light cycle in the home cage using a standard ‘gentle handling’ procedure. Movement data was captured using multiple overhead cameras.

Date Coverage: 2019-2021

Files contained here:

.npy files containing processed data from all rats measuring different variables analyzed in the related paper, including the rate, frequency, and power of detected sharp-wave ripples, firing rates of pyramidal cells and interneurons, reactivation and replay measured by correlation analysis, and multi-neuronal trajectory scores, etc. from before (PRE), during the maze session (MAZE) and after (POST) for both non-sleep deprived (NSD) and sleep-deprived (SD) sessions.

Files are organized into seven folders for ease of access:
1)	brainstates: contains data from all analyses regarding brainstate (Active Wake, Quiet Wake, Rapid-Eye Movement (REM) sleep, and Non-REM (NREM) Sleep).
2)	delta: contains data from all analyses regarding the delta (1-4Hz) oscillation associated with NREM sleep
3)	explained_variance: contains data from all analyses of reactivation concerning pairwise activation of neurons after (and before) maze running.
4)	firing_dynamics: contains data from all analyses of firing properties (e.g. firing rates, firing rate distributions).
5)	replay: contains data regarding sequential replay of neurons active on the MAZE during PRE and POST
6)	ripple: contains data from all analyses of SWR properties such as frequency, power, and sharp-wave amplitude.
7)	misc: contains all data that does not fit into the above size folders.  

All filenames begin with the variable of interest (e.g. ripple). The suffixes “_bootstrap” or “_bs” denote that file contains mean values of that variable after performing hierarchical bootstrapping (Saravan et al., 2020) to get an estimate of the sample mean distribution.  Files without the “_bootstrap” or “_bs” suffix contain raw data values for each variable. In many cases we performed bootstrapping only including raw data from a particular brainstate, which is denoted by either “_WAKE”, “_NREM”, “_REM”, “_QW” (quiet wake), or “_AW” (active wake) in the file name.

Code used to generate figures and access the data can be found at https://github.com/diba-lab/NeuroPy and https://github.com/diba-lab/sleep_loss-hippocampal_replay.  Note that the folder structure of the processed data is provided for ease of access. To use the processed data in conjunction with this code, you will need to move all the files into one folder and adjust the code (as documented on GitHub) to point to that folder.

Related publication(s):
Giri, B., Kinsky, N.R., Kaya, U., Maboudi, K., Abel, T., Diba, K. (2024). Sleep loss diminishes hippocampal reactivation and replay. Nature, in press.

Use and Access:
This data set is made available under a Creative Commons Public Domain license (CC BY 4.0 DEED International).

To Cite Data:
Giri, B., Kinsky, N.R., Diba, K. (2024). Data in support of the study “Sleep loss diminishes hippocampal reactivation and replay”
. University of Michigan - Deep Blue. https://doi.org/10.7302/73hn-m920