Citation:

Lynch, B., C. Grant, C. Mack, AND Tim Shafer. Measurements of in vitro neural network activity are influenced by the number of electrodes in MEAs. SOT 2018, San Antonio, Texas, March 11 - 15, 2018.

Impact/Purpose:

Measurements of in vitro neural network activity are influenced by the number of electrodes in MEAs Brittany S. Lynch1, Christopher M. Grant2, Cina M. Mack3 and Timothy J. Shafer4 1Tandon School of Engineering, New York University, Brooklyn, NY, 11201 2Institute for Advanced Analytics, North Carolina State University, Raleigh, NC, 27695 3Toxicity Assessment Division, NHEERL, United States Environmental Protection Agency, Durham NC, 27711 4Integrated Systems Toxicology Division, NHEERL, United States Environmental Protection Agency, Durham NC, 27711 Multi-well micro-electrode array (mwMEA) recordings are used to screen compounds for the potential to disrupt neural network function. Conventional MEA formats use 60-64 electrodes to detect and record activity within a neural network, but mwMEA plates have different well configurations containing varying numbers of electrodes (i.e. 12-well@64 electrodes/well, 48-well@16 electrodes/well, 96-well@8 electrodes/well). It is unclear how measures of neural network activity may be influenced by the number of electrodes used to collect the data. Therefore, the goal of this study was to determine how network activity parameters changed depending on the number of electrodes used to measure network function. The present study used 12-well mwMEA plates (n=2) to measure three network parameters: mean firing rate (MFR) and network spikes (NwSpk). Rat primary cortical cells express stable spontaneous electrical activity after 12-15 days in vitro (DIV) in the form of individual spikes and bursts. On DIV 14 or 34 initial baseline (BL) data was recorded and MFR, #AE, and NwSpk were calculated and analyzed using an in-house analysis program. Next, BL recordings were further analyzed by using an R-script to randomly eliminate data from four to sixty electrodes, in sets of four. This R-script analysis was repeated ten times, in efforts to increase accuracy of the computational analysis. For BL recordings, MFR was not influenced by the number of electrodes used to determine network parameters. By contrast, although NwSpk was also constant across different numbers of electrodes, it become inconsistent once data from more than 48 electrodes was eliminated. Post BL, cortical cells received either lindane (1uM) or verapamil (10uM) to determine if fewer electrodes alters the detection of chemically-induced network changes. Although lindane increased and verapamil decreased MFR as expected, there was no correlation with numbers of electrodes. However, NwSpk showed a similar response as untreated cultures. These results indicate that investigators should carefully consider the mwMEA format selected when designing experiments, particularly if network parameters will be evaluated. (This abstract does not reflect the policies of the United States Environmental Protection Agency)

Description:

Measurements of in vitro neural network activity are influenced by the number of electrodes in MEAs Brittany S. Lynch1, Christopher M. Grant2, Cina M. Mack3 and Timothy J. Shafer4 1Tandon School of Engineering, New York University, Brooklyn, NY, 11201 2Institute for Advanced Analytics, North Carolina State University, Raleigh, NC, 27695 3Toxicity Assessment Division, NHEERL, United States Environmental Protection Agency, Durham NC, 27711 4Integrated Systems Toxicology Division, NHEERL, United States Environmental Protection Agency, Durham NC, 27711 Multi-well micro-electrode array (mwMEA) recordings are used to screen compounds for the potential to disrupt neural network function. Conventional MEA formats use 60-64 electrodes to detect and record activity within a neural network, but mwMEA plates have different well configurations containing varying numbers of electrodes (i.e. 12-well@64 electrodes/well, 48-well@16 electrodes/well, 96-well@8 electrodes/well). It is unclear how measures of neural network activity may be influenced by the number of electrodes used to collect the data. Therefore, the goal of this study was to determine how network activity parameters changed depending on the number of electrodes used to measure network function. The present study used 12-well mwMEA plates (n=2) to measure three network parameters: mean firing rate (MFR) and network spikes (NwSpk). Rat primary cortical cells express stable spontaneous electrical activity after 12-15 days in vitro (DIV) in the form of individual spikes and bursts. On DIV 14 or 34 initial baseline (BL) data was recorded and MFR, #AE, and NwSpk were calculated and analyzed using an in-house analysis program. Next, BL recordings were further analyzed by using an R-script to randomly eliminate data from four to sixty electrodes, in sets of four. This R-script analysis was repeated ten times, in efforts to increase accuracy of the computational analysis. For BL recordings, MFR was not influenced by the number of electrodes used to determine network parameters. By contrast, although NwSpk was also constant across different numbers of electrodes, it become inconsistent once data from more than 48 electrodes was eliminated. Post BL, cortical cells received either lindane (1uM) or verapamil (10uM) to determine if fewer electrodes alters the detection of chemically-induced network changes. Although lindane increased and verapamil decreased MFR as expected, there was no correlation with numbers of electrodes. However, NwSpk showed a similar response as untreated cultures. These results indicate that investigators should carefully consider the mwMEA format selected when designing experiments, particularly if network parameters will be evaluated. (This abstract does not reflect the policies of the United States Environmental Protection Agency)

Record Details: