Membrane potential depolarization causes alterations in neuron arrangement and connectivity in cocultures

Neural connectivity was assessed based on the quantification of neurite outgrowth and the establishment of contact between neuron somas. One day cultures of E18 rat primary cortical neurons were assessed under normal physiological (no Ivm)- and depolarizing (after Ivm exposure, 1 μmol/L, 24 h) conditions between 0 and 24 h, with 2 h interval initially. The relative Vmem of cells was measured using di-8-ANEPPS dye (Fig.5A). Immature neurons were characterized using specific markers vimentin and GFAP (Fig.5B). Phase-contrast images were collected of the same cells before- and after Ivm exposure at given time points for quantification. Images taken from cells before and after Ivm exposure were shown here (Fig.5C,D). The ratio of the total number of soma to the number of isolated soma was manually counted by two researchers independently.

For imaging, four time points were chosen: 0, 2, 4, 6, and 24 h. The ratio of the total somas to the isolated somas in the images from each time point was obtained. Images were processed, consisting of 10 drug groups and 10 control groups. The ratio of total somas to isolated somas was used as a metric to determine whether Ivm suppressed cell connectivity. Statistical analysis using Student's t-test based on these 100 ratio values (10 × 2 × 5) indicated that there was a significant (P ≤ 0.05) difference between the drug and control groups, suggesting the drug dose suppressed cell connectivity.

Data revealed that neural connectivity was significantly suppressed (P ≤ 0.05) at each time point used in the study. In Ivm-exposed samples, the ratio of total soma number to isolated soma decreased at 0, 2, 4, 6, and 24 h significantly by 28%, 25%, 16%, 21%, and by 33%, respectively (Fig.5E).

To define synaptic connections, we labeled cells with beta-III tubulin, Synaptophysin and Glycine receptor alpha (GLRA) for axon outgrowth, pre-, and postsynaptic proteins, respectively. Ivm-treated cells had very thin, discontinuous, broken membrane projections (arrows) as labeled with beta-III tubulin,whereas control cells had continuous, thick bundles of projections indicating a well-integrated network of neurons (Fig.5F). Similarly Ivm-treated neurons did not show the presynaptic protein Synaptophysin, whereas control cells had large, dense synaptic areas that were positive for the protein (arrows, Fig.5G). Additionally, postsynaptic protein GLRA was not detected along the projections of Ivm-treated neurons but in somas only, whereas control cells had strong production of GLRA along their membrane projections as well as in their somas (arrows, Fig.5H), revealing that inhibitory synapses were eliminated in Ivm-treated cells. We conclude that membrane potential hyperpolarization caused a decrease in neuron connectivity.

Cortical size and shape have been accepted as an accurate clinical measure for functional brain disorders (Hutton et al. 2008; Abdel Razek et al. 2009; Takao et al. 2010; Prabhakaran et al. 2012; Kuhnt et al. 2013). Many of these disorders can be diagnosed using volumetric measurements, including by functional magnetic resonance imaging or positron emission tomography scanning. However, the underlying mechanisms of these anatomical abnormalities in diseased cortices cannot be gauged at the cellular level via these currently available diagnostic tools. Understanding the cell physiological basis of brain diseases is crucial to specifically characterize the nature of a given disorder. This is needed in order to establish realistic models of neurological diseases in the laboratory, which can then be employed for diagnostic and therapeutic purposes. In this study, we show that Vmem can change cortical tissue arrangement, shape, size, and neural connectivity in vitro. We observed that mature cortical neurons aggregated together and formed assemblies when their Vmem was depolarized. These assemblies varied in size ranging from 5–10 cells to 40–50 cells in vitro. Hypotheses for the molecular signatures of neuron assemblies (Yagi 2012) are mostly characterized based on firing patterns (Harris 2005). Membrane potential regulation of cell connectivity is a new mechanism in the field. A few pathways have been reported to imply the formation of neural circuit assemblies in vertebrates; Wnt and Semaphorin signaling (Salinas and Zou 2008; Yoshida 2012), gap junctional proteins (Baker and Macagno 2014), and Glycine receptor signaling (Xu and Tian 2008). We found that GLR production increased when neural Vmem was depolarized (Fig.2F) indicating that there was an enhancement in neural connectivity due to Vmem.

Several diseases have already been linked to the disturbance in neural assemblies such as Alzheimer's and Epilepsy. Muldoon et al. showed that multiple functional clusters of spatially localized neurons comprise epileptic networks, and that network events are composed of the coactivation of variable subsets of these clusters (Feldt Muldoon et al. 2013). It has been also shown that Alzheimer's disease affects specific and connected neural networks via synaptic terminals affecting brain connectivity well before neuronal loss and compartmental atrophy (D'Amelio and Rossini 2012). Moreover, an abnormal connectivity and impaired cognition was found in Hyperekplexia and in Stiff Person Syndrome. These diseases are caused by hereditary mutations resulting in dysfunction of the Glycine receptor channel (alpha1 and beta subunits) that form inhibitory synapses in the nervous system. Therefore, impairment in glycinergic synapse function causes an imbalance between excitatory and inhibitory function of a neural network (Brune et al. 1996; McKeon et al. 2013; Bode and Lynch 2014). Neuron assemblies were formed as a result of Vmem depolarization in this study. This conclusion is also consistent with in vivo data on the role of Vmem in developmental pattern formation (Levin et al. 2002; Lobikin et al. 2012; Chernet and Levin 2013; Tseng and Levin 2013). Additionally, it is known that neural assemblies in large circuits are regulated by neural firing in vivo (Buzsaki and Watson 2012; Huyck and Passmore 2013). Recently, a new study suggested that circuit assemblies can be guided by neurogenesis and neural migration during development (Gao et al. 2013). In this study, we quantified single somas in dense cultures of neural networks and found that the average number of projections in mature cortical neurons increased from 3 to 7 under depolarizing conditions.

This result indicates that depolarized Vmem enhanced the wiring in mature neurons. Membrane depolarization is maintained over 3 weeks of the chronic treatments in our study. Ivm is an agonist of Glycine Receptor Chloride Channel and it leads to membrane depolarization via changing the intracellular chloride concentration. However, the true effector is the changes in resting membrane potential (Vmem), not the chloride ion flows per se. In another study, we could rescue the downstream effects of Ivm depolarization such as mitochondrial transmembrane potential changes by blocking the gap junctions. This demonstrates that the Vmem itself, not specific ion flux, controls the downstream effects (Ozkucur et al., in review).

In immature neurons, however, connectivity was significantly suppressed under hyperpolarizing conditions due to decreased rates of neurite outgrowth resulting in less cell–cell contact, as assessed by calculating the ratio between total and isolated somas. This result showed that Vmem regulated the degree of connectivity and also the development of this connectivity during the first weeks of neuron growth in vitro. Whole-brain network organization relates to cell densities in rodents (French and Pavlidis 2011; French et al. 2011). In the monkey brain, cell density in distinct regions can predict the laminar-specificity of connectivity between two regions (Barbas et al. 2005). A recent report also suggested the presence of a combinatorial effect of modular architectures in network organization relates to the nonuniformity of cell densities in the monkey brain (Shimono 2013). Interestingly, depolarization of Vmem in mature neurons caused a significant increase in glia cell density along with a decrease in glial cell area in this study, whereas neural cell area increased slightly. It is known that changes in both the cell numbers and size of neuron and glia occur during postnatal brain development (Bandeira et al. 2009) as well as in neurological diseases (Cotter et al. 2001). It has been reported that during the first postnatal week in rats, brain growth relates mainly to increased numbers of neurons of larger average size (Lyck et al. 2007; Bandeira et al. 2009). In the second and third weeks, this process correlates with increased numbers of glial cells and the elimination of 60% of the neurons (Haddara 1956; Brizzee 1964; Bandeira et al. 2009).

The present data originate from 3w cocultures. Therefore, increased glial density under depolarizing conditions (Ivm) detected in vitro might be relevant to gliogenesis that occurs mostly during the second and third weeks of postnatal brain development in rats (Sauvageot and Stiles 2002). In contrast, glia cell size decreased and neuron size increased, whereas the number of both cell types increased under depolarized conditions in 3w cocultures. A parallel increase in neuronal and glial cell density has been shown, for example, in schizophrenia, due to tighter cell packing as a consequence of reduced interneuronal neuropils, rather than as an indication of increased glial cell number (Selemon et al. 1995; Rajkowska et al. 1998). In this study, however, depolarization-induced increased glial cell density was found, and a decreased glial cell area coincided with an increase in glia cell number. Changing Vmem has resulted in common phenotypes of cortical volume and neuron networking, as reported in several neurological disorders (Selemon et al. 1995; Cotter et al. 2001; Pujol et al. 2011; d'Ambrosio et al. 2014), in traumatic brain injuries (Selemon et al. 1995; Tate et al. 2014) and in aging (Csernansky et al. 2005). Therefore, we conclude that Vmem can be used as a tool to investigate the formation of neural assemblies, establish neurological disease models and to study neural connectivity in vitro. Moreover, our data shed light on a new factor – slow changes in ion channel activity – that may underlie aspects of complex patterning of the embryonic nervous system. These observations can provide a beneficial platform for drug screening studies relevant to a broad spectrum of neurological disorders and may suggest novel strategies for regenerative medicine of the nervous system.

We thank M.Tang-Schomer and J.D.White for their help and useful feedback in cortical tissue dissociation and neuron culture. The authors declare that they have no conflict of interest. This study was supported by German Research Council (DFG, OE 541/2-1), The G. Harold and Leila Y. Mathers Charitable Foundation, and the National Institutes of Health via grants AR005593, AR061988, AR061933, and EB007542.