May 1, 2014 (Vol. 34, No. 9)

Human Neurons on MEA

The advent of induced pluripotent stem cell (iPSC) technology is enabling rapid advancements in neuroscience research by providing access to unlimited quantities of biologically relevant human neurons. Neuronal network communication is a key functional characteristic of cultured neurons that is central to the study of psychiatric and neurodegenerative diseases.

Multi-Electrode Array (MEA) is a well-established technology that measures native cellular electrical signaling and holds the promise to understand fundamental aspects of learning, memory, neuroplasticity, and neuronal disease. Here, we describe the use of MEA to perform real-time, label-free measurements of network-level activity in human iPSC-derived neurons (iCell® Neurons) in vitro.

iCell Neurons are a highly pure population of GABAergic and glutamatergic neurons, absent of any other cell types that might impact network-level activity. Using a defined assay medium and a 768-channel, 48-well MEA platform (Axion BioSystems), iCell Neurons exhibit measurable action potentials within three days post-plating, and high levels of signaling are observed throughout the cell population after just one week in culture. Indeed, the activity profile of iCell Neurons is observed earlier than other commonly used cell models (e.g., primary rodent cultures), and these iPSC-derived cells more accurately reflect normal human physiology.

iCell NeuroAnalyzer—MATLAB® App

Several methods utilizing MEA have been published that detect signal responses and measure spike train synchrony. However, to elucidate the global impact of a drug on the network communication occurring within iCell Neurons’ neuronal cultures, Cellular Dynamics International has developed a MATLAB-based app called iCell NeuroAnalyzer. This analysis tool is applied directly downstream of MEA spike detection and is tailored around a dosing scheme that monitors activity changes before, during, and after compound treatment. MEA recordings are made with the AxIS instrument software and the resulting spikes files are imported into the app for offline analysis.

The iCell NeuroAnalyzer provides the option to present the spike times recorded from all electrodes in a familiar raster plot format (Figure 1A). Additionally, these data can be collapsed into one concatenated list, binned into 500 msec segments, and displayed as a running average of the firing rate for that single well over time (Figure 1B). A custom “peak detection” algorithm is applied to this average firing rate, and punctate increases in the recorded activity level (≥theta rhythm) are marked and captured. These captured events represent “bursts” in neuronal activity.

As noted in Figure 1B, the point at which the firing rate increases and the burst begins (indicated with a blue dot) is defined as the threshold frequency (measured in Hz). Immediately following this point, the rising signal reaches a maximum, called the peak frequency (measured in Hz), and is designated with red circles. This new analysis enables the detection of agents that influence these parameters.

Figure 1. Example outputs from the iCell NeuroAnalyzer. (A) Raster plot depicting action potential times (ticks) on each of the 16 electrodes per well. Note the high percentage of active electrodes on the MEA. Green arrow indicates application of Gabazine (20 µM) to the neuronal culture; (B) Running average firing rate (Hz) for all electrodes per bin (500 msecs) for the same well as presented above. Note that both the number and frequency of red circles increases following treatment with Gabazine; (C) Histograms represent the MFR (left), bursting rate (middle), and intensity within the bursts (right). Each graph includes average values from six wells before (B) and after (A) compound treatment and the within-subject’s difference (Δ). Means and SEMs are presented; (D) Dose-response plot for Gabazine displaying an increase in the IoE and IoC. Data points represent the mean percentage changes from baseline levels.

Key Parameters Measured

A key intended application of the iCell NeuroAnalyzer is the measurement of pharmacological perturbation of neuronal cultures. In this context, three main outputs can be derived: 1) mean firing rate (MFR), 2) bursting rate or number of bursts per minute (BPM), and 3) intensity within the bursts, which can be determined by the amplitude difference between the threshold frequency and peak frequency (as measured in Hz).

Three histograms are illustrated in Figure 1C, each of which show the data before (B) and after (A) compound treatment, as well as the change or delta (Δ) per well for each condition. On the left, the graph depicts the change in MFR and is labeled “Influence on Inhibition” or IoI. The change in IoI is reported as the inverse of the difference between pre- and post-treatment, as an increase in MFR results in a decrease in IoI. The middle graph shows the change in the bursting rate, “Influence on Excitation” or IoE. Lastly, the graph on the right calculates the change in the intensity within the bursts, labeled “Influence on Connectivity” or IoC. The iCell NeuroAnalyzer integrates each of these measures into a single dataset for comparison. These data can then be plotted in dose-response format as shown in Figure 1D.

Pharmaco-influences Tested

To demonstrate the use of the iCell NeuroAnalyzer to measure the effects of various pharmacological agents on the iCell Neurons’ network activity, the cells were treated with compound on day 8 post-plating in a dose-dependent manner. Comparing the three measures (IoI, IoE, and IoC) for each compound delineates the network impact of a given pharmacological agent.

For example, first blocking all NMDA-receptor activity with a fixed concentration of the selective receptor antagonist, 2 amino-5-phosphopentanoic acid (APV), followed by disruption of AMPA signaling with a titration of 6,7-dinitroquinoxaline-2,3-dione (DNQX) illustrates a significant and dose-dependent decrease in the IoE. The combination of these two compounds are traditionally used to “shut down” excitatory synaptic activity, and a marked decrease in excitation is observed (Figure 2A).

Gabazine, a global antagonist of the GABAA receptor, shows a strong influence on connectivity (Figure 2B). Conversely, a more selective inverse agonist for some GABAA receptors, L-655,708, elicits a distinct bi-phasic response (Figure 2C). Lastly, WIN 55,212-2 is a presynaptic cannabinoid receptor agonist that affects neuronal networks differently than these other molecules (Figure 2D).

In addition to the above examples, several other compounds and combinations therein (e.g., synaptic modulators, inhibitors/blockers, and excitability agents) have been shown to display the correct pharmacological response, with influences headed in the expected direction (data not shown). Collectively, these data demonstrate iCell Neurons form a functional network in just over a week in culture that can be perturbed and modulated. We have developed the iCell NeuroAnalyzer to parse out these pharmaco-influences. End-users can now take advantage of the latest technology in electrophysiology (MEA) and leverage the power of iPSC technology to extrapolate their discoveries to the human condition. 

Figure 2. Tool compounds illustrate dose-dependent effects on network activity. (A) Titration of DNQX in the presence of a fixed concentration of APV (25 µM) shows a dose-dependent decrease in the IoE; (B) Titration of Gabazine illustrates a strong increase in the IoC and slight but significant increase in the IoE. These are the normalized data from Figure 1C; (C) Dose-response of L-655,708 suggests that lower levels of compound increase the IoE and IoC, whereas higher concentrations result in an increase in the IoI and a decrease in the IoE. These results suggest that both neuronal types (GABAergic and glutamatergic) express α5 subunit-containing GABAA receptors; (D) Treatment with WIN 55,212-2 results in a dose-dependent increase in the IoE and IoC. All values are normalized to and presented as fold changes from control levels. *Denotes p

Kile P. Mangan, Ph.D. ([email protected]), is technical application scientist, Coby B. Carlson, Ph.D. ([email protected]), is application development manager, and Susan DeLaura  ([email protected]) is product manager at CDI.

Previous articleGEN’s “Top 10” Session Picks for the 2014 BIO International Convention
Next articleCRISPR—Fast, Easy … and Increasingly Accurate