Models of Epilepsy in Cell Culture
When neurons from cortex or hippocampus are grown in dissociated cell culture, they form extensive synaptic interactions and local circuits. The nature of the circuitry has not been extensively examined, but a few principles have emerged. Neurons in routine high-density culture do not tend to form synapses with neighbors in culture, despite extensive overlap of processes. Instead synapses are formed with more distant targets. When neurons are monitored by intracellular recordings, they tend to fire in bursts at either irregular, or occasionally, regular intervals. These bursts, which resemble those seen in epileptic tissue in vivo, occur despite an inability of these neurons to generate endogenous bursts in response to depolarizing pulses (Buchhalter and Dichter, 1991; Dichter, 1978) (as opposed to some endogenous bursting neurons in intact cortex and hippocampus). When pairs of neurons in culture are recorded, it is common for the neurons to exhibit synchronous bursting or synchronous activation, presumably driven by synaptic events originating elsewhere in the cultures. Thus the circuit organization and synaptic interactions among neocortical and hippocampal neurons maintained in dissociated cell cultures spontaneously re-create the kinds of circuits needed to sustain epileptiform types of activity.
Synchronous bursting in the cultures can be enhanced by many of the same stimuli that provoke epileptic activity in intact cortex (e.g., GABA receptor antagonists and NMDA receptor activators). Such epileptogenicity occurs in the absence of endogenous bursting cells, electrical interactions between the neurons, or ephaptic interactions among the neurons (as a result of large extracellular currents flowing in the media). This synchronous discharge can also occur when neurons are grown in the virtual absence of astrocytes. Surprisingly this "epileptic-type activity" and epileptogenic circuitry have been largely ignored by epilepsy researchers using cell cultures to study other aspects of cellular biophysics, physiology and pharmacology, or mechanisms of action of AEDs.
One prominent exception is the work of DeLorenzo and colleagues (Gibbs et al., 1997; Pal et al., 2001; Sombati and DeLorenzo, 1995), who have exploited the enhanced synchrony produced by low magnesium exposure to create a cell culture model of epilepsy. Reducing Mg in the media in which neurons are bathed enhances activity of NMDA receptors and produces an increase in excitability of the cultures. When cells are maintained in this condition for longer than 3 hours, the cultures remain hyperexcitable for the life of the cells, even when normal Mg is restored. The DeLorenzo laboratory uses hippocampal cultures that are plated on a 2-week-old astrocyte feeder layer. After about 2 weeks, the culture media is replaced with media containing no magnesium and a small amount of glycine, sufficient for activation of NMDA receptors. During exposure to this low-magnesium media, the neurons exhibit a continuous synchronous firing pattern reminiscent of a status epilepticus (Sombati and DeLorenzo, 1995). After 3 hours of exposure to this activating media, the cell cultures are switched back to the original media. Following this treatment, the neurons exhibit activity that has been dubbed spontaneous recurrent epileptiform discharges (SREDs) (DeLorenzo et al., 1998). In contrast to control tissue, the low Mg-exposed neurons have a much higher likelihood of synchronous depolarizations; this tendency persists for the life of the cells, typically more than 2 weeks.
DeLorenzo's group has carried out a number of experiments to support their view that this model simulates focal epileptogenesis, dependent on aberrant activation of NMDA receptors. They showed that blockade of the NMDA receptors during the low-Mg treatment of these cultures prevents the development of hyperexcitability (as evidenced by SREDs) (DeLorenzo et al., 1998). Prolonged direct activation of the NMDA receptors, using the physiologic agonist glutamate instead of low magnesium, resulted in a similar long-term hyperexcitability and disruption of calcium homeostasis, likely through alteration in calcium/calmodulin kinase II activity (Sun et al., 2001). These alterations in calcium homeostasis may also lead to increased GABAA receptor endocytosis or altered GABAA receptor composition. Low-magnesium-treated cells exhibit a lower current density for a given concentration of GABA and show a lower sensitivity to the benzodiazepine, clonazepam. Blocking clathrin-mediated endocytosis in postsynaptic cells reverses hyperexcitability caused by the low-magnesium treatment (Blair et al., 2004). The proposed mechanism of this clathrin action is by blockade of GABAA receptor
Summary: Insights Into Human Epilepsy Derived from Studies of Neuronal Cell Cultures
Was this article helpful?