The Reversal Potential of Inhibitory Synapses Strongly Impacts the Dynamics of Neural Networks - Foundations on Natural and Artificial Computation - page 107

neurons. Then, FS interneurons feed forward depolarizing inputs to pyramidal
neurons thanks to the higher value of the reversal potential. In this way the
interaction of the two networks is characterized by a positive feedback loop. In
the middle panel of figure 3 the results obtained in the presence of gap-junctions
among FS interneurons are plotted. The firing activities of both networks does
not exhibit new features with respect to the previous case. However a more
detailed inspection shows that the presence of gap-junctions promotes a more
synchronous firing of the cells of both networks.
From general considerations it is expected that a dynamical systems driven
by a positive feedback loop can exhibit instability, depending on the parameter
values. This behaviour arises also in the system of coupled neural networks that
is investigated here. In fact a little increase of the value of the reversal potential
value of the inhibitory synapses leads to the dynamical behaviour shown in the
right panel of figure 3. This firing pattern can be interpreted as the emergence
of a seizure-like phenomenon.
4 Conclusions
Inhibitory interneurons exert a powerful control of the firing activities of pyra-
midal cells [1,2,3,4,5]. However recent experimental results pointed out that in
particular conditions (accumulation of intracellular Cl − ) the inhibitory synapses
can depolarize the postsynaptic cells [7]. This phenomenon occurs because the
reversal potential value of the inhibitory currents increases. Motivated by these
experimental data, we studied the dynamical behaviour of a biophysical inspired
network of two coupled populations of cells: the first population is composed by
coupled FS interneurons (coupled by inhibitory and electrical synapses), while
the second one is constituted by coupled pyramidal cells (coupled by excita-
tory synapses). In agreement to the experimental model proposed in [6,7] the
two networks are connected by excitatory synapses (from pyramidal cells to
FS interneurons) and inhibitory synapses (from FS interneurons to pyramidal
neurons).
When the coupling between the two networks is set off, the excitatory synapses
promote synchronous firing of pyramidal cells, while the inhibitory synapses
among FS interneurons lead to the decrease of their firing activity. Increasing
the conductance of the excitatory synapses determines the emergence of bursting
in the network of pyramidal cells. Moreover, it was found that the presence of the
electrical synapses among FS cells promotes the increase of their firing synchrony.
When the coupling between the two networks is set on the dynamical behaviour
of the two neural populations become richer. In particular bursting regimes also
occur in the network of FS cells. Overall, the reciprocal coupling between the
two networks promotes the increase of the synchrony among inhibitory cells.
Moreover incomplete phase locking states between excitatory and inhibitory cells
were also observed. Next we investigated how the value of the reversal potential
of the inhibitory synapses affects the dynamical behaviour of both networks. It
was found that the increase of the value of the reversal potential leads to the