Altered Network Communication Following a Neuroprotective Drug Treatment

Abstract

Preconditioning is defined as a range of stimuli that allow cells to withstand subsequent anaerobic and other deleterious conditions. While cell protection under preconditioning is well established, this paper investigates the influence of neuroprotective preconditioning drugs, 4-aminopyridine and bicuculline (4-AP/bic), on synaptic communication across a broad network of in vitro rat cortical neurons. Using a permutation test, we evaluated cross-correlations of extracellular spiking activity across all pairs of recording electrodes on a 64-channel multielectrode array. The resulting functional connectivity maps were analyzed in terms of their graph-theoretic properties. A small-world effect was found, characterized by a functional network with high clustering coefficient and short average path length. Twenty-four hours after exposure to 4-AP/bic, small-world properties were comparable to control cultures that were not treated with the drug. Four hours following drug washout, however, the density of functional connections increased, while path length decreased and clustering coefficient increased. These alterations in functional connectivity were maintained at four days post-washout, suggesting that 4-AP/bic preconditioning leads to long-term effects on functional networks of cortical neurons. Because of their influence on communication efficiency in neuronal networks, alterations in small-world properties hold implications for information processing in brain systems. The observed relationship between density, path length, and clustering coefficient is captured by a phenomenological model where connections are added randomly within a spatially-embedded network. Taken together, results provide information regarding functional consequences of drug therapies that are overlooked in traditional viability studies and present the first investigation of functional networks under neuroprotective preconditioning.

Detection of marine neurotoxins in food safety testing using a multielectrode array

“…All these K + channels are supposed to be present in rat cortical neurons, but these compounds possibly only block the K + channels only partially similar as what has been reported for the blockade of delayed rectifier K + channels by 1 mM thiamine which just slightly inhibited the delayed rectifier K + current in rat cerebral cortical cultures [31]. Similarly, the K + channel blocker 4-aminopyridine, in combination with bicuculline, did not have any effect on neuronal activity of rat cortical neurons [32]. The reason that amiodarone can be detected using the MEA approach may be due to the fact that in addition to blockade of K + channels, amiodarone also has an effect on Na + and Ca 2+ channels [33].…”

Assessment of Spontaneous Neuronal ActivityIn VitroUsing Multi-Well Multi-Electrode Arrays: Implications for Assay Development

“…However, a method of dynamic threshold calculation over a moving window (in the range of ϳ10 ms) was more common. The crossing threshold employed by most studies were in the range of 5-6 SD, although others have used crossing thresholds as low as 3 SD (Vincent et al, 2013) or as high as 8 SD (McConnell et al, 2012). While methods are available to detect subthreshold potentials within MEA recording (Henningson and Illes, 2017), the use of large magnitude thresholds is done to insure that the majority of the events recorded represent action potentials.…”

“…Most studies reporting on MEA experiments make no mention of this highly skewed distribution in firing rates, with the noted exception of the report by Biffi and colleagues that discusses it specifically (Biffi et al, 2013). This observation may have gone unnoticed, due to studies being performed in single-well MEAs having insufficient data for this distribution to be apparent, however a similar highly skewed distribution has been reported for firing frequencies at individual recording electrodes within MEAs (Vincent et al, 2013). This distribution shape is not surprising, given that any frequency based metric is lower bound at zero, and the only upper limit is the sampling rate at which the observations are being made.…”

Dynamical networks: finding, measuring, and tracking neural population activity using network science

“…genetic stimulation (Boyden, 2015;Deisseroth, 2015;Miesenböck, 2009); the sluggish but acute effects of drugs (Vincent, Tauskela, Mealing, & Thivierge, 2013); or the chronic effects of neurological damage (Otchy et al, 2015). All these manipulations potentially change the interactions between neurons, disrupting normal computation.…”

“…Srinivas, Jain, Saurav, and Sikdar (2007) used dynamical networks to quantify the changes to network-wide activity in hippocampus caused by the glutamate-injury model of epilepsy, suggesting a dramatic drop in network clustering in the epilepsy model. Vincent et al (2013) used dynamical networks to quantify the potential neuroprotective effects of drug pre-conditioning in rat cortex in vitro, finding increased clustering and increased efficiency in the network over days, implying the drugs enriched the synaptic connections between groups of neurons. Quantifying manipulations using network science is an under-explored application, rich in potential.…”