Reduction in inhibitory control is sufficient to generate hyperalgesia in a spiking model of nociceptive integration in the superficial dorsal horn

Here we present a large network spiking model for hyperalgesia which includes significant functional components of the nociceptive circuit in the SDH. The network architecture (Figure 1A) was defined according to key experimental results, namely: the somatotopic mapping between body surface and the superficial dorsal horn [1]; the presence of an inhibitory control which, under normal conditions, leads to a proper perception of pain [2]; the existence of divergent/convergent fiber inputs to lamina I (LI) neurons [3]; and the presence of inhibitory interneurons, i.e. islet cells, with inputs/outputs within lamina II (LII) [4]. In the model, the connectivity profile between LII interneurons (excitatory, Exc and inhibitory, Inh) and LI projection neurons creates a mexican-hat input profile with a strong centered excitation flanked by a surrounding inhibition. Activity level control was provided by feed-forward inhibition from the inhibitory interneurons in LII. The network model was created and simulated using the simulation environment NeuralSyns [http://sourceforge.net/projects/neuralsyns/]. Neurons were represented as integrate-and-fire units with conductance based synapses. The full model was comprised of 2.600 neurons and 320.000 synapses. The analysis was performed by assessing the changes produced in the receptive fields of LI nociceptive specific projection neurons (NS), both in normal (control) conditions and after reduction in inhibitory control from LII inhibitory interneurons (Figure 1B). Fiber activity (Aδ, C) was modelled using spatially modulated Poisson activations. The receptive field associated with reduced inhibitory control occupies a larger skin area and involves higher mean firing rates. These two results correlate respectively with secondary and primary hyperalgesia.