Ana Dolabela de Lima

The mechanisms underlying spontaneous burst activity (SBA), appearing in networks of embryonic cortical neurons at the end of the first week in vitro, remain elusive. Here we investigated the contribution of the hyperpolarization-activated cation current (I(h)) to SBA in cortical cultures of GAD67-GFP mice. I(h) current could be detected in GFP-positive large GABAergic interneurons (L-INs) and glutamatergic principal neurons (PNs) as early as DIV 5. Under current-clamp conditions, blockers of I(h) current, ZD7288 and Cs⁺, abolished the voltage sag and rebound depolarization. ZD7288 induced a hyperpolarization concomitant with an increase in the membrane input resistance in L-INs and PNs. Voltage-clamp recordings revealed I(h) as slowly activating inward current with a reversal potential close to -50 mV and a mid-activation point around -90 mV. Both, ZD7288 (1-10 μM) and Cs⁺ (1-2 mM) reduced SBA, spontaneous activity-driven Ca²⁺ transients, and frequency as well as amplitude of miniature GABAergic postsynaptic currents. Immunocytochemistry and Western blot demonstrated that HCN1 and HCN2 were the prevalent isoforms of HCN channels expressed in L-INs and PNs. These results suggest an important contribution of HCN channels to the maintenance of SBA in embryonic cortical cultures.

Unipolar source–drain voltage pulses of GaN/AlGaN-high electron mobility transistors (HEMTs) were used for stimulation of cultured neuronal networks obtained from embryonic rat cerebral cortex. The HEMT sensor was grown by metal organic vapour phase epitaxy on a 2 inch sapphire substrate consisting of 10 single HEMTs concentrically arranged around the wafer centre. Electrolytic reactions between the HEMT sensor surface and the culture medium were not detected using cyclic voltammetry. During voltage pulses and resulting neuronal excitation, capacitances were recharged giving indications of the contributions of the AlGaN and AlO _x isolation layers between the two-dimensional electron gas channel and the neuron culture. The resulting threshold current for stimulation of neuron activity strongly depended on the culture and HEMT position on the sensor surface under consideration which was caused by different impedances of each neuron culture and position within the culture. The differences of culture impedances could be explained by variations of composition, thickness and conductivity of the culture areas.

Available evidence converges to suggest that during the early development of the cerebral cortex, the emergence of the spontaneous network activity chronologically overlap with the end of the cell migration period in the developing cortex. We approached the functional regulation of neuronal migration in a culture model of neocortical networks, using time lapses to detect migratory movements, calcium‐imaging to assess the activity of migratory neurons, and immunocytochemical methods to identify the migratory cells retrospectively. In cell cultures, early physiological development and cell migration are reproduced at a local network level, thus allowing the study of the interrelationships between cell migration and network development independent of the topographical complexity. Neurons migrate at least until 12 days in vitro and GABAergic neurons migrate faster compared with non‐GABAergic neurons. A decline of migratory activity was coincident with the development of spontaneous synchronous network activity. Migrating interneurons did not participate in synchronous network activity, but interneurons that ended cell migration during observation time frequently engaged in synchronous activity within less than an hour. Application of GABA_A and ionotropic glutamate receptor antagonists significantly increased the number of migrating GABAergic neurons without changing the dynamics of the migratory movements. Thus, neurotransmitters released by early network activity might favor the termination of neuronal migration. These results reinforce the idea that network activity plays an important role in the development of late‐born GABAergic cells. © 2008 Wiley Periodicals, Inc. Develop Neurobiol, 2009

This report concerns investigations of the activation process of nerve cells cultured on large planar interdigitated electrodes. Cultured neurons from embryonic rat cerebral cortex form electrophysiologically active networks after one week in vitro. The treatment with antagonists to ionotropic glutamate and GABA_A receptors blocks synaptic neurotransmission without affecting intrinsic excitability of neurons. Standard stimulation parameters, which were previously determined to trigger a generalized network response in control networks, were ineffective for activation of networks under a synaptic activity blockade. Here we show that in these conditions a pulse sequence of ten biphasic pulses with amplitude of ±2.2 V is adequate for global network stimulation, without cell damage. In specially designed compartment cultures we analyzed further the stimulation process in control and synaptically blocked networks with cell bodies spatially separated from the electrodes. Our results suggest that the direct stimulation of few neurons trigger network-wide responses by the activation of synaptic connections. Further, the stimulation of axonal neuritic extensions is sufficient to active single neurons.

Networks of cortical neurons in vitro spontaneously develop synchronous oscillatory electrical activity at around the second week in culture. However, the underlying mechanisms and in particular the role of GABAergic interneurons in initiation and synchronization of oscillatory activity in developing cortical networks remain elusive. Here, we examined the intrinsic properties and the development of GABAergic and glutamatergic input onto presumed projection neurons (PNs) and large interneurons (L‐INs) in cortical cultures of GAD67‐GFP mice. Cultures developed spontaneous synchronous activity already at 5–7 days in vitro (DIV), as revealed by imaging transient changes in Fluo‐3 fluorescence. Concurrently, spontaneous glutamate‐mediated and GABA_A‐mediated postsynaptic currents (sPSCs) occured at 5 DIV. For both types of neurons the frequency of glutamatergic and GABAergic sPSCs increased with DIV, whereas the charge transfer of glutamatergic sPSCs increased and the charge transfer of GABAergic sPSCs decreased with cultivation time. The ratio between GABAergic and the overall charge transfer was significantly reduced with DIV for L‐INs and PNs, indicating an overall reduction in GABAergic synaptic drive with maturation of the network. In contrast, analysis of miniature PSCs (mPSCs) revealed no significant changes of charge transfer with DIV for both types of neurons, indicating that the reduction in GABAergic drive was not due to a decreased number of functional synapses. Our data suggest that the global reduction in GABAergic synaptic drive together with more synaptic input to PNs and L‐INs during maturation may enhance rhythmogenesis of the network and increase the synchronization at the level of population bursts. © 2008 Wiley Periodicals, Inc. Develop Neurobiol, 2008

Although less than one quarter of all neurons in the cerebral cortex are GABAergic, these neurons are morphologically diverse and their physiological complexity decisively moulds the network physiology. An important question is how different subpopulations of GABAergic neurons are regulated numerically during development. In rat neocortical cultures, neuronal precursors continue to divide, generating both GABAergic and non-GABAergic neurons. In vitro generated GABAergic neurons form a population of uniquely small, mostly fusiform neurons that differ in size and morphology from older, in situ generated, large stellate GABAergic neurons. In a large series of experiments we investigated the impact of neuronal activity on the development of these two subpopulations of GABA interneurons present in cortical networks during the first 2 weeks in vitro. Here we show that a moderate increase in the generation of GABAergic neurons was achieved by blocking activity with tetrodotoxin, indicating that intrinsic spontaneous activity inhibits GABAergic neurogenesis in culture. Antagonists to ionotropic glutamate receptor and/or GABA(A) receptor did not significantly alter GABAergic generation but agonists to these receptors showed a time-sensitive regulation of the size of small and large GABAergic neuronal subpopulations. Further, our results indicate that alterations of cell generation by activity manipulations might be overwritten by later activity effects on the survival of GABAergic cell populations.