Neurotoxicity Testing Using Microelectrode Arrays (MEAs): A Growing Trend

Home / New Perspectives / Neurotoxicity / Neurotoxicity Testing Using Microelectrode Arrays (MEAs): A Growing Trend


Neurotoxicity Testing Using Microelectrode Arrays (MEAs): A Growing Trend

Timothy J. Shafer, U.S. Environmental Protection Agency

Published: April 4, 2011

About the Author(s)
Tim Shafer received a B.S. in Biology and Chemistry from Hope College in 1986, and a Ph.D. in Pharmacology and Toxicology at Michigan State University in 1991, where he studied the actions of methylmercury on calcium channel function. Tim joined the Neurotoxicology Division of the U.S. Environmental Protection Agency’s National Health and Environmental Effects Research Laboratory as a post-doc in 1991, and was hired as a Principle Investigator in that organization in 1994. In 2009, he became part of the newly created Integrated Systems Toxicology Division. Dr. Shafer’s research program has utilized a variety of electrophysiological and biochemical approaches to increase the understanding of mechanisms underlying the neurotoxicity of diverse environmental chemicals including metals, pesticides, solvents and PCBs. Currently, research in his lab is focused on developing in vitro assays that can be utilized to identify potential hazards of large numbers of chemicals and prioritize them for additional testing. More specifically, his research is examining functional approaches, such as microelectrode arrays, that can be used to detect potential neurotoxicants and prioritize them for additional testing.

Timothy J. Shafer, PhD
Integrated Systems Toxicology Division
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
MD B105-03

Microelectrode arrays (MEAs) are groups of extracellular electrodes that are 10-30 microns in diameter and can be utilized in vivo or in vitro. For in vitro uses, an MEA typically contains up to 64 electrodes and can be utilized to measure the activity of cells and tissues that are electrically excitable, such as neurons, slices of nervous system tissue, and cardiac cells or tissue. In the past, MEAs have been widely utilized to study network function in primary cultures of neurons or tissue slices from nervous system, as well as to characterize drug effects on nervous system tissue. A series of recent publications and presentations indicates that MEAs may also provide a useful in vitro testing system for effects of chemicals on the nervous system, including effects during development.

Johnstone et al. (2010) recently published a comprehensive review of the potential for MEAs to be used in neurotoxicity testing. The review provides basic information regarding MEAs and discusses the advantages of this approach as well as areas where further development is needed to make the approach more useful for toxicity testing purposes. Among the advantages, MEAs provide high content data, and disruption of spontaneous network activity by drugs and chemicals produces characteristic changes in the activity patterns of neurons on MEAs. This latter feature may be useful when testing compounds with unknown actions, as this may allow them to be grouped with compounds having similar actions and toxicities. Although one of the drawbacks of MEAs is that throughput is relatively low, Johnstone notes that this aspect of MEAs is also improving, as higher throughput systems are now becoming available.

Several papers and recent presentations also support the utility of MEAs for toxicity testing, including developmental neurotoxicity testing. The review above by Johnstone also contains an appendix that summarizes past use of MEAs in toxicity studies. In addition to this review, several other papers have highlighted the use of MEAs for testing with human stem cell derived neurons (Heikkilä et al., 2009; Jurga et al., 2009; Ylä-Outinen et al., 2010) or for use in developmental neurotoxicity testing (Hogberg et al., 2011; Robinette et al., 2011). The manuscript by Hogberg and co-workers demonstrated that treatment with domoic acid over the period of development of cortical cultures in vitro results in mature networks of neurons that are less sensitive to block of GABAA receptors than were networks that were not exposed to domoic acid. Robinette and co-workers demonstrated that if primary cultures of cortical neurons were exposed to a compound that is known to inhibit outgrowth of neurites, the activity of networks of neurons was decreased, presumably as the result of fewer connections being made between neurons (Robinette et al., 2011). Ylä-Outinen and colleagues (2010) demonstrated that human embryonic stem cells can be differentiated to form spontaneously active networks of neurons on MEAs and that the neurotoxicant methylmercury (at nanomolar concentrations) can alter the activity of those neuronal networks. Most recently, a multilaboratory study comparing the reproducibility and reliability of MEA data has recently been accepted for publication. In this study, 6 different laboratories compared the potency of verapamil, fluoxetine and diazepam, and demonstrated highly consistent results even though different MEA hardware, software and culture approaches were utilized (Novellino et al., 2011).

At the 2011 meeting of the Society of Toxicology (March 6-10), there were several presentations that further highlighted the use of MEAs for neurotoxicity testing. Lein, Pessah and co-workers reported that the organophosphorous insecticide chlorpyrifos and its active metabolite chlorpyrifos oxon, both increased spontaneous activity in sympathetic neurons (Niknam et al., 2011). By contrast, the inactive metabolite, trichlorpyridinol, was without activity (Nikam et al., 2011). Another presentation from this group reported that the ability of the chiral polychlorinated biphenyl (PCB136) to increase spontaneous oscillations in intracellular calcium as well as spontaneous activity in neurons grown on MEAs is enantiomer-specific. Andrew Johnstone presented a poster demonstrating that chemical antagonists of GABAA receptors (e.g. lindane, RDX) produce characteristic changes in the timing of spontaneous firing rates that increase the synchronization of firing across the network (Johnstone et al., 2011). Compounds that do not act as GABAA receptor antagonists do not increase synchrony. This may be one method by which compounds with different modes of action could be identified and separated. Helena Hogberg presented her data on the effects of domoic acid on development of activity in MEAs (Hogberg et al., 2011) and McNutt and Mesngon proposed using mouse embryonic stem cell derived neurons grown on MEAs as part of a drug discovery program to evaluate compounds that may be effective against botulinum toxin. Bibiana Scelfo from the European Commission Joint Research Centre presented the results from the multilaboratory “ring trial” that have just recently been published (Scelfo et al., 2011) using MEAs. Finally, Robert Landsiedel from BASF presented a poster summarizing the results of testing 21 compounds in MEAs using a concentration-response to evaluate this approach as a method for neurotoxicity testing (Vogel et al., 2011). The results demonstrated that the approach had high sensitivity and predictivity, as well as good specificity (Vogel et al., 2011).

Overall, these recent publications and presentations indicate a growing use of MEA approaches for neurotoxicity testing purposes, and demonstrate the feasibility of MEAs for mechanistic studies and neurotoxicity screening, including developmental neurotoxicity screening. The potential to utilize stem cell-derived neurons, from rodents or humans, to assess chemical effects on network activity offer the capability to use MEAs to collect meaningful data from non-transformed neurons while still achieving the goal of reductions in animal use.

To achieve the goal of using MEA approaches for neurotoxicity testing, there are some important steps that need to occur to move the area forward.

  • Development of higher throughput MEA platforms. This is an ongoing process, and all manufacturers of MEA equipment have introduced multi-well chips as well as increased the capability of systems in terms of the number of channels that can be simultaneously recorded from. Recently, a multi-well plate MEA format was introduced by one manufacturer (for review see Johnstone et al., 2010).
  • Further demonstration of the reliability and reproducibility of MEA measurements across different platforms and laboratories. The poster presentation by Scelfo et al. demonstrated excellent reproducibility of MEA data across several different labs using 3 pharmaceutical agents. This needs to be expanded to compounds of toxicological importance; these studies are currently underway.
  • Greater numbers of compounds need to be tested by more laboratories. The example set by Vogel and co-workers (Vogel et al., 2011), who tested 20 compounds, is an excellent start. They tested several classes of compounds that were known to be neurotoxic and several that were not acutely neurotoxic. To demonstrate the utility of this approach, more laboratories should examine large sets of compounds that contain “positive control” and “negative control” compounds so that the sensitivity, selectivity and predictivity of the assay can be better characterized.
  • Additional studies focused on chemical effects on network development need to be conducted. Current developmental neurotoxicity guideline studies are expensive and require large numbers of animals. One advantage of the MEAs is the ability to record multiple times from the same neuronal network. To date, only 2 compounds have been examined in the context of their effects on network development (Hogberg et al., 2011; Robinette et al., 2011). Additional known developmental neurotoxicants should be examined for their effects on network development.
  • Development of methods and data using stem cell derived neurons. While meaningful data could be collected from primary cultures of neurons, achieving a reduction of animal use is ultimately dependent on widely available methods and protocols for using stem cell derived neurons. Currently, differentiation protocols (particularly for human stem cell-derived neurons) are labor intensive and require long differentiation times (several weeks, minimum). Establishment of efficient differentiation procedures that can be widely utilized is needed to facilitate testing of large numbers of chemicals in stem cell derived neurons to evaluate these models for neurotoxicity testing using MEAs.

Disclaimer: This material has been reviewed by the National Health and Environmental Effects Research Laboratory of the US Environmental Protection Agency, and is approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.
©2011 Timothy J. Shafer

Heikkilä, J., Ylä-Outinen, L., Tanskanen, J.M., Lappalainen, R.S., Skottman, H., Suuronen, R., Mikkonen, J.E., et al. (2009). Human embryonic stem cell-derived neuronal cells form spontaneously active neuronal networks in vitro. Exp. Neurol. 218, 109-116.

Hogberg, H.T., Sobanski, T., Novellino, A., Whelan, M., Weiss, D.G. & Bal-Price, A.K. (2011). Application of micro electrode arrays (MEAs) as an emerging technology for developmental neurotoxicity: evaluation of domoic acid-induced effects in primary cultures of rat cortical neurons. Neurotoxicology 32, 158-68

Johnstone, A.F.M., Gross, G.W., Weiss, D.G., Schroeder, O.H., Gramowski, A. & Shafer, T.J. (2010). Microelectrode arrays: A physiologically based neurotoxicity testing platform for the 21st century. Neurotoxicology 31, 331-50.

Johnstone, A.F.M., Turner, J., Mack, C., Burgoon, L. & Shafer, T.J. (2011). GABAA receptor antagonists increase firing, bursting and synchrony of spontaneous activity in neuronal networks grown on microelectrode arrays (MEAs): A step toward chemical “fingerprinting”. Abstract #1354. The Toxicologist CD Society of Toxicology volume 120, Supplement 2.

Jurga, M., Lipkowski, A.W., Lukomska, B., Buzanska, L., Kurzepa, K., Sobanski, T., et al. (2009). Generation of functional neural artificial tissue from human umbilical cord-blood stem cells. Tissue Eng. Part C Methods. 15, 365-72.

McNutt, P. & Mesngon, M. (2011). Development of embryonic stem cell-derived neurons for botulinum toxin research and drug discovery. Abstract #77. The Toxicologist CD Society of Toxicology volume 120, Supplement 2.

Niknam, Y., Ghogha, A., Lein, P. & Pessah, I. (2011). Chlorpyrifos and chlorpyrifos oxon alter Ryanodine receptor function. Abstract #1327. The Toxicologist CD Society of Toxicology volume 120, Supplement 2.

Novellino, A., Scelfo, B., Palosaari, T., Price, A., Sobanski, T., Shafer, T.J., et al. (2011). Development of micro-electrode array based tests for neurotoxicity: assessment of interlaboratory reproducibility with neuroactive chemicals. Front. Neuroengineering. (in press).

Robinette, B.L., Harrill, J.A., Mundy, W.R. & Shafer, T.J. (2011). In Vitro Assessment of Developmental Neurotoxicity: Use of Microelectrode Arrays to Measure Functional Changes in Neuronal Network Ontogeny. Front. Neuroengineering. doi: 10.3389/fneng.2011.00001.

Scelfo, B., Novellino, A., Price, A., Palosaari, T., Sobanski, T., Gross, G., et al. (2011). Abstract #2510. The Toxicologist CD, Society of Toxicology volume 120, Supplement 2.

Vogel, S., Novellino, A., De Franchi, E., van Ravenzwaay, V. & Landsiedel, R. (2011). Micro electrode chip assay (MEA) as a method to detect neurotoxicity in vitro. Abstract #1020. The Toxicologist CD, Society of Toxicology volume 120, Supplement 2.

Yang, D., Kania-Korwei, I., Bose, D., Ghogha, A., Pessah, I., Lehmier, H., & Lein P. (2011). Enantioselective effects of PCB136 on dendritic growth are ryanodine receptor dependent. Abstract #2624 The Toxicologist CD, Society of Toxicology volume 120, Supplement 2.

Ylä-Outinen, L., Heikkilä, J., Skottman, H., Suuronen, R., Äänismaa, R. & Narkilhati, S. (2010). Human cell-based micro electrode array platform for studying neurotoxicity. Front. Neuroengineering. doi: 10.3389/fneg.2010.0111.