Signatures of the Consolidated Response of Astrocytes to Ischemic Factors In Vitro

Abstract

Whether and under what conditions astrocytes can mount a collective network response has recently become one of the central questions in neurobiology. Here, we address this problem, investigating astrocytic reactions to different biochemical stimuli and ischemic-like conditions in vitro. Identifying an emergent astrocytic network is based on a novel mathematical approach that extracts calcium activity from time-lapse fluorescence imaging and estimates the connectivity of astrocytes. The developed algorithm represents the astrocytic network as an oriented graph in which the nodes correspond to separate astrocytes, and the edges indicate high dynamical correlations between astrocytic events. We demonstrate that ischemic-like conditions decrease network connectivity in primary cultures in vitro, although calcium events persist. Importantly, we found that stimulation under normal conditions with 10 µM ATP increases the number of long-range connections and the degree of corresponding correlations in calcium activity, apart from the frequency of calcium events. This result indicates that astrocytes can form a large functional network in response to certain stimuli. In the post-ischemic interval, the response to ATP stimulation is not manifested, which suggests a deep lesion in functional astrocytic networks. The blockade of Connexin 43 during ischemic modeling preserves the connectivity of astrocytes in the post-hypoxic period.

Neuroprotective Effect of Kinase Inhibition in Ischemic Factor Modeling In Vitro

“…The key aspect in the characterization of the dynamic interactions of neuron-glial networks in our study is the neural network parameters of calcium activity, which characterize the degree of connectivity of network elements after exposure to ischemic factors ( Figure 5 ). Application of previously developed algorithms for network analysis of imaging data [ 19 ] makes it possible to comprehensively characterize the features of network activity reorganization under damaging influences.…”

“…Calcium network analysis was performed using a previously developed algorithm [ 19 ]. The approach to the reconstruction of the network consisted of two steps: computation of pairwise correlations of the preprocessed fluctuations in the intracellular calcium level of individual cells (network nodes) and applying a threshold cutoff for the correlation level ρ > 0.3 to determine the significant network connections.…”

“…In the paper by [ 19 ], several measures were proposed to estimate network activity: the average level of correlation between adjacent cells, the average number of functional connections per cell, the percentage of correlated connections, and the average propagation speed of delays. The average level of correlation between adjacent cells shows how the strength of physical connections required for network communication.…”

“…A 100% value characterizes the maximally possible level of significant correlations in the network. An increase in the last two parameters considered above indicates the presence of a long-term significantly correlated response of the cell network, whereas their low levels characterize its absence [ 18 , 19 ].…”

“…Calcium ions play a crucial role in signaling between nerve cells and the regulation of synaptic activity, and their imaging gives a means to estimate the functional state of neuronal networks with cellular resolution. The recently developed algorithm for analyzing calcium imaging data employs correlation analysis to reconstruct and follow the dynamical architecture of neuron-glial networks [ 18 , 19 ]. It provides with a powerful tool to characterize the collective and coordinated activity of these networks on the functional level and infer its response to model conditions, and evaluate the effect of damaging and protective factors.…”