Interestingly, the synchronous synaptic release of GABA from interneurons of SLM promoted CA3 pyramidal cell activation. Thus, the reduction of Cx45 could lead to desynchronization of GABA release from the interneurons, consequently decreasing CA3 pyramidal cell activity. Therefore, downregulation of Cx45 in the SLM during the latent period could indicate a plastic homeostatic change of the coupled network in order to restore its activity after the epileptogenic insult, since interneuron networks within SLM could regulate the input from the entorhinal cortex to the hippocampus. Whether modulation of Cx45 expression in SO and SLM remains at later time points is not known, but it would be worth evaluating the issue. Interestingly, whereas changes in Cx45 distribution were seen in the latent period, alterations in Cx43 distribution through CA3 layers were observed in the acute period. Astroglial networks provide the supply of glucose and lactate necessary for the maintenance of hippocampal synaptic activity in an activitydependent manner, and it has been shown that this metabolic network of astrocytes is mediated by GJ composed by Cx43 and Cx30. Thus, the increase in Cx43 distribution pattern observed in the SP is probably due to the high energetic demand promoted by the sustained epileptic activity. Additionally, the coupled astrocytes could provide an important intercellular pathway for both delivery of metabolites to distant locations during episodes of epileptic discharges and buffering of ions such as potassium. Increase of Cx43 in SP could account for two opposite situations: besides providing distribution of metabolites of neurotransmitters and potassium to distant locations during episodes of epileptic discharges, a Cx43 increase could lead to the formation of GJ hemichannels, a pathway for adenosine release, which could also contribute to the local FG-4592 neuroprotective effects; by allowing the propagation of toxic metabolites and death signals to adjacent cells, the increase in astroglial network coupling could amplify the damage to more distant sites, in addition to the possible contribution to the rapid propagation of the electrical signal. Contrary to the observed in SP, we found decreased labeling of Cx43 in the CA3 SR. It has been demonstrated that the contribution of GJ in potassium buffering in the hippocampus is layer-dependent. By comparing the properties of astrocytes from mice with Cx30/Cx43 deficiencies, those authors found that single astrocytes from SR reach larger areas than those from SLM, indicating the unequal size and orientation of astrocytes from those areas. Additionally, they found no impairment of potassium buffering in SR of transgenic mice, in contrast to the observed in the SLM, evidencing the participation of other elements in the potassium redistribution in the SR. Therefore, the reduction of Cx43 labeling noticed here in the SR is unlikely to have a great impact in spatial buffering of potassium during SE; moreover, as CA3 SR is well supplied by a vascular source, which in turn accounts for the degree of astrocyte coupling, such decrease of Cx43 labeling in this layer during epileptic activity may not have a significant impact in the physiology of CA3 SR cells.