Studying the Cortico-Thalamic Circuits Underlying Perturbational Complexity Analysis to Detect the Presence of Consciousness in Mice
INSTITUTION: MindScope Program at the Allen Institute
PROJECT LEADERS: Christof Koch, PhD, Irene Rembado, PhD, Leslie Claar, PhD, Shawn Olsen, PhD
PROJECT DURATOIN: 2019 - 2022
LAB WEBSITE: AllenInstitute.org
This project aims to convert the TMS-EEG experiments being used to probe different states of consciousness in humans (Casali, Gosseries et al., 2013; Casarotto et al., 2016; Comolatti et al., 2019) to similar experiments in mice, i.e., stimulating the cortex, recording global electrical responses across the cortex, and computing a murine equivalent of Perturbational Complexity Index, PCIST.
In this project we use two different recording modalities to record brain-wide responses to direct cortical stimulation while mice are awake and anesthetized: a mouse EEG array to record global cortical signals above the skull and multiple Neuropixels probes (high-density, linear silicon probes with 384 electrodes) to record local field potential signals and activity from hundreds of single neurons in cortical and subcortical brain areas. We show that cortical stimulation elicits a widespread, complex evoked response in the EEG signals during wakefulness, but a profoundly simpler response during anesthesia (Figure 1), in agreement with what has been shown in humans (Casali, Gosseries et al., 2013) and rats (Arena et al., 2021). PCIST captures this difference; it is higher in the awake state compared to the anesthetized state in all mice.
Figure 1: Evoked EEG responses in awake and anesthetized mice. (A) Schematic of the mouse brain overlayed with evoked responses (-0.2 to +0.8 s around stimulus onset) from all EEG electrodes. Traces are shown in their approximate location. Traces in black represent awake responses, traces in orange represent anesthetized responses, and traces in gray represent signals removed for low quality. The red star and white circles mark the approximate insertion points of the stimulating electrode (in the motor cortex) and the Neuropixels probes, respectively. (B) PCIST calculated using the EEG evoked responses (baseline window: -0.8 to -0.002 s; response window: 0.002 to 0.8 s) for the awake and anesthetized states. Individual values represented with gray circles (n=29). Mean +\- SEM represented with filled circles with error bars (black: awake; orange: anesthetized). The PCIST values for the awake state are significantly higher than those for the anesthetized state (paired t-test, p=2.8E-10).
Because we have access to cortical and subcortical activity with the Neuropixels probes, we observed that cortico-thalamo-cortical interactions coincide with the long-lasting evoked responses in the EEG elicited during the awake state (Figure 2).
Figure 2: Single electrical pulses evoke strong responses in the EEG, LFP, and in some populations of neurons in the awake state. Top: Sagittal view of a schematic of the mouse brain highlighting areas of interest: motor, anterior cingulate, somatosensory, visual, and thalamic areas. The solid black lines show the approximate locations of three Neuropixels probes acutely inserted to record single neuron activity and local field potentials. Approximate location of stimulating electrode in the deep layers of motor cortex is shown in red. Bottom: Evoked responses (-0.2 to +0.8 s around stimulus onset) for the EEG, LFP, and neuron spiking activity for the areas of interest.
Broader Impact:
With an experimental method to study consciousness in mice, we would be able to interrogate the contributions of specific brain regions and cell types to this phenomenon using a variety of techniques, such as high-resolution recordings of cortical and subcortical structures and/or the optogenetic perturbations.
Publications:
Claar LD*, Rembado* I, Kuyat J, Marks L, Olsen SR & Koch C (2022) Thalamocortical neuronal interactions modulate electrically evoked EEG responses in mice. In preparation.