In this work, we studied the electrophysiological activity and dynamics of neuronal ensembles made up of C1 and B2 Helix neurons during their in vitro development. The advantage of this biological preparation is the predetermined knowledge of cell type composition, biophysical properties and connectivity. In addition, by exploiting the MEA technology (i.e., extracellular recordings), we were able to perform long-lasting measures in order to characterize the development of Helix networks.
Here, the absence of spontaneous activity [10, 30] was observed even during the formation of synaptically interconnected clusters, despite the increasing neuritic growth and arborization, and the establishment of synaptic connections. Taking into account this feature, we decided to trigger electrophysiological activity by applying two chemical treatments: KCl, to promote a general depolarization (unspecific stimulation); and 5-HT, to activate serotonergic receptors which are selectively expressed on B2 neurons. Following these stimulations, we observed similar dynamics during the development of these invertebrate circuits: the network firing early reached a value of firing rate and ISI, which remained almost unchanged during the development despite the connectivity maturation. This behavior is very peculiar, especially if compared with studies regarding the development of in vitro cortical neurons of rat embryos, [12, 13], where it was found that cortical assemblies change their electrophysiological patterns as a function of the network maturation, perhaps because 1000-fold higher synapse density.
Although we observed that both KCl and 5-HT stimulations were sufficient to trigger firing, we found that only 5-HT efficiently induced a sustained long-lasting activity after the first application.
Serotonin has been frequently associated with changes in excitability. Brief 5-HT application has often been observed to induce immediate hyperexcitability of Aplysia sensory neurons, both in ganglion preparations [31–34] and in cell cultures [35, 36]. Moreover, an increase in firing activity elicited by depolarizing current pulses has also been reported in rat cortical neurons after 5-HT exposure [37, 38], suggesting a phylogenetically-conserved modulatory action for this neurotransmitter. Therefore, we evaluated the possibility that the observed long-term maintenance of firing activity could result from an increase of cell excitability. In addition to the B2 neuron response, a rapid increase of C1 neuron excitability was observed without alterations of other parameters, including membrane potential, input resistance and AP threshold. A similar effect was described in the Aplysia metacerebral cell [20, 39], homologous to Helix C1 neuron. Even if the modulation of cell excitability may be a good explanation for triggering spontaneous activity in Helix circuits, it cannot be the only effector in promoting its maintenance. Behind the alteration of single cell properties, the emerging evidence would suggest changes in neuronal circuitry features as another putative cause, including the morphological and functional connectivity.
From a morphological point of view, our results showed an increase of neuritic density in cultures treated with KCl, while a reduction (compared to the non-stimulated networks) occurred in 5-HT stimulated networks. There is a large body of evidence that attribute to 5-HT an inhibitory role in the regulation of neurite outgrowth in invertebrate models [22–24], thus affecting both the development of arborizations and connectivity formation. In mammals, 5-HT has been implicated in shaping neuronal connectivity, e.g. decreasing neurite branching in rat cortical neurons during development  and impairing neurite density in mouse organotypic slice cultures .
From a functional point of view, Helix neurons may form both chemical and electrical connections in vitro[5, 7]. Cross-correlation and latency analysis revealed a strong association between treatment and the type of circuitry connectivity. From the first hours of development, networks treated with 5-HT display synaptic latencies of about 20 ms, compared to the ones treated with KCl (about 5 ms after 24 hours). Comparing these values with those measured from interconnected pairs of neurons, we found a striking correspondence between MEA signal latency and the synaptic delays intracellularly recorded (Figure 9F).
The selective presence of the two types of synaptic connections may reflect the different strength of functional links estimated by means of cross-correlation analysis. The occurrence and the magnitude of chemical signal transduction vary dramatically depending on culture conditions, mainly related to the presence of trophic factors . Moreover, phenomena such as neurotransmitter release modulation, synaptic delay, receptor desensitization, signal propagation along neurites and the postsynaptic refractory period may negatively affect the threshold potential to trigger an action potential in the postsynaptic cell. Conversely, the presence of gap junctions facilitate synchronous activity of coupled cells, allowing the direct transmission of electrical signal, thus explaining the 10-fold higher value of cross-correlogram peaks measured from KCl-treated cultures.
In support to our results, 5-HT has also been demonstrated to selectively prevent the formation of electrical synapses while allowing chemical synaptogenesis between identified Helisoma neurons [22, 42]. The formation of electrical coupling between Helisoma neuronal pairs is thought to require an active growth state and neurite elongation [43–45], negatively modulated by 5-HT. However, a direct action of 5-HT on gap junctions has already been demonstrated in Helisoma neurons [46–48], as well as in developing rat cortical neurons .
On the other hand, the KCl-induced increase in neurite density may contribute to enhance cell-cell contact, thus promoting a higher coupling coefficient among cells . Since gap junctions likely play a fundamental role in determining network synchronization [50, 51], signals may reverberate among neurons until they return to a silent state by switching off the circuit.
Taken together, these observations suggest that the prevalence of chemical connections in 5-HT-treated circuits may be the cause of long-term maintenance of spontaneous activity. In this view, activity may be rapidly triggered by 5-HT application, through the depolarization of B2 neurons and the increased excitability of C1 neurons, and maintained by the continuous release of neurotransmitter from different firing neurons. Accordingly, repetitive applications of serotonin have previously been implicated in the enhancement of neurotransmitter release through presynaptic mechanisms [52–54].