Recording and Analyzing Multimodal Large-Scale Neuronal Ensemble Dynamics on CMOS-Integrated High-Density Microelectrode Array

Our research explores the frontiers of neural technology by integrating high-density CMOS-based microelectrode array for decoding neural communication in large networks. We aim to answer how neural information across scales is encoded in unique detail, enhancing our understanding of brain function/dysfunction in health and disease. Navigating the complex terrain of neuronal ensemble research, we confront challenges like achieving precise signal resolution amidst brain activity, and ensuring the biocompatibility of our CMOS microelectrode areas.

These hurdles are pivotal in accurately capturing and interpreting the rich tapestry of neural interaction using multimodal recordings. Our research addresses a critical gap in the neuroscience, the lack of a comprehensive method to record and analyze the dynamics of larger-scale neuronal ensemble with high spatial and temporal resolution. This gap hinders our understanding of complex brain networks and their function in health and disease.

Our protocol enables multimodal, label-free, high-resolution recordings across the hippocampus, olfactory bulb, and human iPSC-derived neurons, providing a versatile tool for diverse experiments. This unique approach facilitates unparalleled insight into neuronal dynamics, bridging the research gap between various brain regions and model systems, significantly advancing our understanding of neural function and disorder. Future endeavors in our lab will deeply investigate neural computations and dynamics from genes to networks, aiming to bridge molecular and functional signatures in health and disease through advanced biotronics and neurotechnology.

We'll focus on neuroplasticity, olfactory coding, developing AI, and memory-enhancing strategies for novel therapeutics and brain-machine interfaces. To begin, set up the brain slice recovery and maintenance workspace. Prepare the brain slice preparation workspace and set up the vibratome.

Next, set up the brain extraction and preparation workspace. Then transfer the extracted mouse brain to a 90-millimeter plastic culture dish with chilled carbogenated cutting solution in the brain preparation workspace. Orient the brain for positioning in the agarose mold.

Apply super glue to the rostral end of the agarose mold. Use a spatula to place the brain in the mold with the dorsal side down for horizontal slicing. Slice the hippocampus and olfactory bulb tissues.

And using a glass Pasteur pipette, collect the slices after each round. Incubate the slices in the artificial cerebrospinal fluid-filled recovery chamber in a 32 degrees Celsius water bath for 45 minutes, followed by one hour at room temperature. To begin, take the HD-MEA chip.

Wipe the chip with a glass ring with tissue moistened with 96%ethanol. Place each device into a sterile Petri dish under the hood and fill the MEA reservoir with 70%ethanol. After aspirating the ethanol, wash the reservoir three times with sterile, filtered, double-distilled water.

Add one milliliter of preconditioning media and incubate overnight. The next day, remove the preconditioning media from the devices. Coat the active area of each device with one milliliter of PDLO.

The following day, aspirate the PDLO from the devices. Wash the devices three times with double-distilled water and allow them to dry under a hood. Fill a 35-by-10 millimeter Petri dish with sterile, filtered, double-distilled water and place it beside the chip.

Using the pre-warmed media, pipette the cell suspension onto the surface of the chip's active area. Incubate the chips at 37 degrees Celsius in an environment with 5%carbon dioxide. After 45 minutes, gently add two milliliters of media warmed to room temperature into the HD-MEA reservoir.

Incubate the chips at 37 degrees Celsius with 5%carbon dioxide. To begin, place all the necessary tools in their designated workspaces. Coat the HD-MEA chip with PDLO to enhance tissue-chip coupling.

Place the chip on the acquisition recording platform. Fill its reservoir with artificial cerebrospinal fluid and place the anchor in the chip reservoir to equilibrate. Remove the prepared mouse brain slice from the recovery chamber and place it in a 90-millimeter culture dish with continuous carbogenation.

After isolating the hippocampus or olfactory bulb, move the slice into the HD-MEA reservoir with a glass pipette. Using a fine brush, gently align the slice on the MEA's active area. Aspirate all solutions from the chip well.

Using forceps, gently place the anchor on top of the slice. After gently adding the solution to the chip reservoir, initiate the perfusion system. Next, launch the BrainWave software.

Select File and click New Recording Session. Set the recording parameters to a recording frequency of one hertz and a sampling frequency of 14 kilohertz per electrode. Press Record to begin the acquisition with the preset experimental conditions.

Finally, after the last recording, capture light imaging of the acute brain slice. Then move the slice back to the recovery chamber. Remove any organic material attached to the chip using a brush and proceed with the next slice.

To begin, take the HD-MEA chip containing iPSCs. Under the hood, carefully place a PDM-based based cap with reference to the HD-MEA ring. Then place the HD-MEA chip on the acquisition platform.

Launch the BrainWave software, select File, and click New Recording Session. Set the recording parameters to a recording frequency of 50 hertz and a sampling frequency of 18 kilohertz per electrode. Record the spontaneous firing activity or pharmacologically induced responses from the iPSC network on each specified day of the experimental plan.

Incubate the HD-MEAs at 37 degrees Celsius with 5%carbon dioxide over the course of the experiment. To begin, record spatiotemporal neuronal activity from ex-vivo mouse brain acute slices and in-vitro human iPSC-derived neuronal networks. Open a recorded raw data file in BrainWave software, select Analysis, and click LFP Detection or Spike Detection.

For hippocampal and olfactory bulb circuit recordings, integrate the advanced workspace option for structural light image import from the stereo microscope into the detected event file. Create structural layers for examining large-scale hippocampal cortical circuitry. Utilize a custom-written Python script to read the BXR file.

Extract spike trains from iPSC network recordings and LFP event trains from the hippocampus and olfactory bulb brain slice circuit recordings. Save the resultant event train data with spatiotemporal information in NPY file format. Calculate the cross covariance between pairs of active electrodes in the 64-by-64 array using time series data from the BRW file.

Transform each connectivity matrix into a dynamic graph file. For network connectivity maps, apply geo layout for spatial mapping and set parameter constraints on degree range and edge weight for comparison. Assign nodal color, edge size, and degree size for enhanced visualization.

Topographical spatial mapping of the mean large-scale LFP or spike firing patterns are overlaid on the respective microscope-captured optical images. The rastergram plots display synchronously detected LFP or spike events over a specified time band. Representative event traces from HD-MEA recording electrodes showed a range of recorded oscillatory frequencies in the hippocampal, cortical, and olfactory bulb circuits and multiunit spike bursting activity in the human iPSC network.

The hippocampal, cortical, and olfactory bulb circuit connectome maps revealed various nodes representing different layers, wherein the size indicated the degree of strength, and the color denoted the layer. In human iPSC networks, connection identification was refined using spatial temporal filters and latency thresholds, where nodal color indicated if the input was excitatory or inhibitory.