Pattern generation by two coupled time-discrete neural networks with synaptic depression

β-pompilidotoxin modulates spontaneous activity and persistent sodium currents in spinal networks

The origin of rhythm generation in mammalian spinal cord networks is still poorly understood. In a previous study, we showed that spontaneous activity in spinal networks takes its origin in the properties of certain intrinsically spiking interneurons based on the persistent sodium current (INaP). We also showed that depolarization block caused by a fast inactivation of the transient sodium current (INaT) contributes to the generation of oscillatory activity in spinal cord cultures. Recently, a toxin called beta-pompilidotoxin (β-PMTX) that slows the inactivation process of tetrodotoxin (TTX)-sensitive sodium channels has been extracted from the solitary wasp venom. In the present study, we therefore investigated the effect of β-PMTX on rhythm generation and on sodium currents in spinal networks. Using intracellular recordings and multielectrode array (MEA) recordings in dissociated spinal cord cultures from embryonic (E14) rats, we found that β-PMTX reduces the number of population bursts and increases the background asynchronous activity. We then uncoupled the network by blocking all synaptic transmission (APV, CNQX, bicuculline and strychnine) and observed that β-PMTX increases both the intrinsic activity at individual channels and the number of intrinsically activated channels. At the cellular level, we found that β-PMTX has two effects: it switches 58% of the silent interneurons into spontaneously active interneurons and increases the firing rate of intrinsically spiking cells. Finally, we investigated the effect of β-PMTX on sodium currents. We found that this toxin not only affects the inactivation of INaT but also increases the peak amplitude of the persistent sodium current (INaP). Altogether, theses findings suggest that β-PMTX acting on INaP and INaT enhances intrinsic activity leading to a profound modulation of spontaneous rhythmic activity in spinal networks.

Recurrent networks with short term synaptic depression

Cortical circuitry shows an abundance of recurrent connections. A widely used model that relies on recurrence is the ring attractor network, which has been used to describe phenomena as diverse as working memory, visual processing and head direction cells. Commonly, the synapses in these models are static. Here, we examine the behaviour of ring attractor networks when the recurrent connections are subject to short term synaptic depression, as observed in many brain regions. We find that in the presence of a uniform background current, the network activity can be in either of three states: a stationary attractor state, a uniform state, or a rotating attractor state. The rotation speed can be adjusted over a large range by changing the background current, opening the possibility to use the network as a variable frequency oscillator or pattern generator. Finally, using simulations we extend the network to two-dimensional fields and find a rich range of possible behaviours.

Patterns of spontaneous activity in unstructured and minimally structured spinal networks in culture

The rhythmic activity observed in locomotion is generated by local neuronal networks in the spinal cord. The alternating patterns are produced by reciprocal connections between these networks. Synchronous rhythmic activity, but not alternation, can be reproduced in disinhibited networks of dissociated spinal neurons of rats. This suggests that a specific network architecture is required for pattern generation but not for rhythm generation. Here we were interested in the recruitment of neurons to produce population bursts in unstructured and minimally structured cultures of rat spinal cord grown on multielectrode arrays. We tested whether two networks, connected by a small number of axons, could be functionally separated into two units and generate more complex patterns such as alternation. In the unstructured cultures, we found that the recruitment of the neurons into bursting populations is divided into two steps: the fast recruitment of a "trigger network", consisting of intrinsically firing cells connected in networks with short delays, and slow recruitment of the rest of the network. One or several trigger networks were observed in a single culture and could account for variable patterns of propagation. In the minimally structured cultures, a functional separation between loosely connected networks was achieved. Such separation led either to an independent bursting between the networks or to synchronized bursting with long and variable delays. However, no qualitatively novel pattern such as alternation could be generated. In addition, we found that the strength of reciprocal inhibitory connections was modulated by spontaneous activity.

Dissociated retinal neurons form periodically active synaptic circuits

Throughout the developing nervous system, immature circuits generate rhythmic activity patterns that influence the formation of adult networks. The cellular mechanisms underlying this spontaneous, correlated activity can be studied in dissociated neuronal cultures. Using calcium imaging and whole cell recording, we showed that cultured dissociated mammalian retinal neurons form networks that produce spontaneous, correlated, highly periodic activity. As the culture matures, the spatial correlations of the periodic calcium transients evolve from being highly synchronized across neighboring cells to propagating across the culture in a wavelike manner reminiscent of retinal waves recorded in vivo. Spontaneous calcium transients and synaptic currents were blocked either by cadmium, tetrodotoxin, or the glutamate receptor antagonist 6,7-dinitroquinoxaline, indicating that the periodic activity was driven primarily by synaptic transmission between retinal ganglion cells. Evoked responses between pairs of ganglion cells exhibited paired-pulse synaptic depression, and the time constant of recovery from this depression was similar to the interval between periodic events. These results suggest that synaptic depression may regulate the frequency of network activity. Together, these findings provide insight into how networks containing primarily excitatory connections generate highly correlated activity.

Mean-field theory of globally coupled integrate-and-fire neural oscillators with dynamic synapses

We analyze the effects of synaptic depression or facilitation on the existence and stability of the splay or asynchronous state in a population of all-to-all, pulse-coupled neural oscillators. We use mean-field techniques to derive conditions for the local stability of the splay state and determine how stability depends on the degree of synaptic depression or facilitation. We also consider the effects of noise. Extensions of the mean-field results to finite networks are developed in terms of the nonlinear firing time map.