The fine-tuning of synaptic transmission by astrocyte signaling is crucial to CNS physiology. However, how exactly astroglial excitability and gliotransmission are affected in several neuropathologies, including epilepsy, remains unclear. Here, using a chronic model of temporal lobe epilepsy (TLE) in rats, we found that astrocytes from astrogliotic hippocampal slices displayed an augmented incidence of TTX-insensitive spontaneous slow Ca(2+) transients (STs), suggesting a hyperexcitable pattern of astroglial activity. As a consequence, elevated glutamate-mediated gliotransmission, observed as increased slow inward current (SICs) frequency, up-regulates the probability of neurotransmitter release in CA3-CA1 synapses. Selective blockade of spontaneous astroglial Ca(2+) elevations as well as the inhibition of purinergic P2Y1 or mGluR5 receptors relieves the abnormal enhancement of synaptic strength. Moreover, mGluR5 blockade eliminates any synaptic effects induced by P2Y1R inhibition alone, suggesting that the Pr modulation via mGluR occurs downstream of P2Y1R-mediated Ca(2+)-dependent glutamate release from astrocyte. Our findings show that elevated Ca(2+)-dependent glutamate gliotransmission from hyperexcitable astrocytes up-regulates excitatory neurotransmission in epileptic hippocampus, suggesting that gliotransmission should be considered as a novel functional key in a broad spectrum of neuropathological conditions.

Kindling, one of the most used models of experimental epilepsy is based on daily electrical stimulation in several brain structures. Unlike the classic or slow kindling protocols (SK), the rapid kindling types (RK) described until now require continuous stimulation at suprathreshold intensities applied directly to the same brain structure used for subsequent electrophysiological and immunohistochemical studies, usually the hippocampus. However, the cellular changes observed in these rapid protocols, such as astrogliosis and neuronal loss, could be due to experimental manipulation more than to epileptogenesis-related alterations. Here, we developed a new RK protocol in order to generate an improved model of temporal lobe epilepsy (TLE) which allows gradual progression of the epilepsy as well as obtaining an epileptic hippocampus, thus avoiding direct surgical manipulation and electric stimulation over this structure. This new protocol consists of basolateral amygdala (BLA) stimulation with 10 trains of biphasic pulses (10 s; 50 Hz) per day with 20 min-intervals, during 3 consecutive days, using a subconvulsive and subthreshold intensity, which guarantees tissue integrity. The progression of epileptic activity was evaluated in freely moving rats through electroencephalographic (EEG) recordings from cortex and amygdala, accompanied with synchronized video recordings. Moreover, we assessed the effectiveness of RK protocol and the establishment of epilepsy by evaluating cellular alterations of hippocampal slices from kindled rats. RK protocol induced convulsive states similar to SK protocols but in 3 days, with persistently lowered threshold to seizure induction and epileptogenic-dependent cellular changes in amygdala projection areas. We concluded that this novel RK protocol introduces a new variant of the chronic epileptogenesis models in freely moving rats, which is faster, highly reproducible and causes minimum cell damage with respect to that observed in other experimental models of epilepsy.

Astrocytes from the hippocampus of chronic epileptic rats exhibit an abnormal pattern of intracellular calcium oscillations, characterized by an augmented frequency of long lasting spontaneous Ca2+ transients, which are sensitive to purinergic receptor antagonists but resistant to tetrodotoxin. The above suggests that alterations in astroglial Ca2+-dependent excitability observed in the epileptic tissue could arise from changes in astrocyte-to-astrocyte signaling, which is mainly mediated by purines in physiological and pathological conditions. In spite of that, how purinergic signaling contributes to astrocyte dysfunction in epilepsy remains unclear. Here, we assessed the possible contribution of P2Y1R as well as pannexin1 and connexin43 hemichannels—both candidates for non-vesicular ATP-release—by performing astroglial Ca2+ imaging and dye uptake experiments in hippocampal slices from control and fully kindled rats. P2Y1R blockade with MRS2179 decreased the mean duration of astroglial Ca2+ oscillations by reducing the frequency of slow Ca2+ transients, and thereby restoring the balance between slow (ST) and fast transients (FT) in the kindled group. The potential contribution of astroglial pannexin1 and connexin43 hemichannels as pathways for purine release (e.g., ATP) was assessed through dye uptake experiments. Astrocytes from kindled hippocampi exhibit three-fold more EtBr uptake than controls, whereby pannexin1 hemichannels (Panx1 HCs) accounts for almost all dye uptake with only a slight contribution from connexin43 hemichannels (Cx43 HCs). Confirming its functional involvement, Panx1 HCs inhibition decreased the mean duration of astroglial Ca2+ transients and the frequency of slow oscillations in kindled slices, but had no noticeable effects on the control group. As expected, Cx43 HCs blockade did not have any effects over the mean duration of astroglial Ca2+ oscillations. These findings suggest that P2Y1R and Panx1 HCs play a pivotal role in astroglial pathophysiology, which would explain the upregulation of glutamatergic neurotransmission in the epileptic brain and thus represents a new potential pharmacological target for the treatment of drug-refractory epilepsy.

Synapse unsilencing is an essential mechanism for experience-dependent plasticity. Here, we showed that the application of the ligand Wnt-5a converts glutamatergic silent synapses into functional ones by increasing both α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) and N-methyl-D-aspartate (NMDA) currents (IAMPA and INMDA, respectively). These effects were mimicked by the hexapeptide Foxy-5 and inhibited by secreted frizzled-related protein sFRP-2. INMDA potentiation was produced by increased synaptic potency, followed by an increase in the probability of release (Pr), even in the presence of 7-nitro-2,3-dioxo-1,4-dihydroquinoxaline-6-carbonitrile (CNQX). At a longer time of Wnt-5a exposure, the Pr increments were higher in INMDA than in IAMPA. In the presence of NMDAR inhibitors, Wnt-5a-induced conversion was fully inhibited in 69.0% of silent synapses, whereas in the remaining synapses were converted into functional one. Our study findings showed that the Wnt-5a-activated pathway triggers AMPAR insertion into mammalian glutamatergic synapses, unsilencing non-functional synapses and promoting the formation of nascent synapses during the early postnatal development of the brain circuits.