Encyclopedia of Computational Neuroscience

Living Edition
| Editors: Dieter Jaeger, Ranu Jung

Slow Oscillations and Epilepsy: Network Models

  • Alain DestexheEmail author
Living reference work entry
DOI: https://doi.org/10.1007/978-1-4614-7320-6_19-1


Network models are very useful tools to investigate the genesis of oscillatory behavior such as epileptic seizures. During many types of seizures, the brain produces oscillatory spike-and-wave discharges, which are particularly prominent for absence seizures. It was found that the thalamocortical mechanisms leading to spindle oscillations and their large-scale synchrony can explain spike-and-wave oscillations, if the excitability of the cerebral cortex is augmented. These pathological oscillations can be reproduced by network models involving the reciprocal interaction between thalamus and cortex.

Detailed Description


Many types of epileptic seizures display very synchronized oscillations producing a typical electroencephalogram (EEG) pattern consisting of one or several sharp deflections (“spikes”) followed by a surface-positive “wave.” Spike-and-wave patterns of similar characteristics are also seen in a number of experimental models in cats, rats, mice, and monkeys, as well as in many other types of epilepsies. In some of these experimental models, it was shown that the critical regions involved are the thalamus and neocortex and that the genesis of spike-and-wave activity shares common mechanisms with sleep spindle oscillations, which are generated in the thalamus. Computational models have explored mechanisms for such spike-and-wave activity, as reviewed here.

The Cellular Bases of Spike-and-Wave Discharges in Cortex

Computational models have first explored the cellular pattern needed to generate spike-and-wave EEG or local-field potentials. Using a simple biophysical model of cortical network with excitatory and inhibitory neurons (Destexhe 1998), it was found that spike-and-wave LFP patterns can be generated if all cells fire in synchrony (“spike” component), followed by a hyperpolarization mediated by K+ conductances (“wave” component). The synchronized discharge during the “spike” was indeed found in many different experimental models of seizures (reviewed in Destexhe 2013), while the hyperpolarization during the “wave” remains poorly studied. Intracellular recordings during spike-and-wave seizures showed that the “wave” is correlated by hyperpolarization, which was interpreted as due to a mixture of disfacilitation and K+ currents (Neckelmann et al. 2000). The model suggests that this slow K+ conductance can be due to two different mechanisms: either it can be generated by the slow voltage-dependent K+ conductances responsible for spike-frequency adaptation in pyramidal neurons. In this case, the strong firing during the “spike” maximally activates these adapting currents, which then produce the slow hyperpolarization. Another possible origin is synaptic; the strong firing may massively activate the slow type of GABAergic inhibition, acting on GABA B receptors, which also cause a slow activation of K+ channels.

Hypersynchronized Slow Oscillations in the Thalamus

Many experimental evidences show that thalamic circuits are implicated in seizure generation (reviewed in Gloor and Fariello 1988; Crunelli and Leresche 2002). It was found that if convulsants are applied to the thalamus in vitro, blocking GABA A receptors, they induce the production of a slow and synchronized oscillation at around 3 Hz and are dependent on GABA B receptors (von Krosigk et al. 1993). Although it is tempting to associate the origin of the seizure and the spike-and-wave oscillation to the thalamus, it is not the case, because this slow oscillation was shown to be different from spike-and-wave oscillations in vivo (Steriade and Contreras 1998). This slow oscillation was simulated by network models of thalamic circuits (Destexhe et al. 1993, 1996; Golomb et al. 1996) and was due to the mutual interaction between thalamic relay and reticular neurons, involving GABA B receptors.

Model of Spike-and-Wave Discharges in the Thalamocortical System

Experimentally, it was shown that the cerebral cortex is the key to the induction of seizures. In cats, different experimental manipulations of cortical excitability lead to spike-and-wave seizures, but the physiologically intact thalamus is necessary (Gloor and Fariello 1988). In rat models of absence seizures, a focus of increased excitability was found in cortex (Meeren et al. 2002; Polack et al. 2007), but the thalamus was also necessary for seizure generation (Vergnes and Marescaux 1992). These results were replicated by computational models of thalamic and cortical networks, endowed with the different receptor type present as well as the intrinsic cellular properties of thalamic and cortical cells (Destexhe 1998, 1999). This model is explained below.

The main hypothesis explored by this model was that seizure generation uses the oscillatory-generating mechanisms present in the thalamus but with an over-excitable cortex. In “control” conditions, the network model generates spindle oscillations. As shown in detail in another entry (“Spindle Oscillations: Models”), spindle oscillations are generated by thalamic circuits and can be modeled by thalamic networks. In the thalamocortical system, the cortex plays an essential role in synchronizing the thalamic-generated oscillations, as also described in detail in another entry (“Corticothalamic Feedback: Large-Scale Synchrony”). In this case, the models established that the most efficient way for the cortex to control the thalamic oscillation and organize its large-scale synchrony is to evoke strong inhibition in the thalamus (Destexhe et al. 1998).

This “inhibitory dominant” character of the corticothalamic interaction is fundamental to explain seizures. While this cortically evoked thalamic inhibition is mostly mediated by GABA A receptors in normal conditions, it may be different in pathological conditions. If the cortex is hyperexcitable, the exaggerated corticothalamic feedback will be strong enough to activate GABA B receptors (due to their nonlinear activation properties – see Destexhe and Sejnowski 1995). This activation of slow GABA B inhibition will “force” the thalamus, although physiologically intact, to oscillate at a slower and more synchronized frequency around 3 Hz (Destexhe 1998).

This mechanism was tested by simulating networks of thalamic and cortical cells, with the different receptor types present, and the system was able to generate hypersynchronized oscillations at 3 Hz when solely the cortex was affected (its excitability was augmented by diminishing intracortical inhibition). Remarkably, this model generated 2–3 Hz oscillations with spike-and-wave patterns in the calculated LFPs (Destexhe 1998).

A very similar mechanism was found for faster oscillation frequencies typically observed in rodents (Destexhe 1999). In this case, the total conductance of GABA B inhibition in the thalamus had to be weaker, preventing the thalamus to oscillate at 3 Hz. Thus, this model predicts that the ~3 Hz spike-and-wave of cats, monkey, and man and the ~8 Hz spike-and-wave seizures of rats and mice have a common cellular mechanism but different respective conductances of GABA A and GABA B receptor-mediated inhibition in the thalamus. This prediction still awaits to be tested.

Testing the Mechanism in Thalamic Slices

The main prediction of this thalamocortical model of seizures is that a too-strong corticothalamic feedback should be able to switch the intact thalamus to oscillate at 3 Hz (Destexhe 1998), a frequency which is not normally seen in the physiologically intact thalamus. This prediction was tested in thalamic slices, where the corticothalamic fibers can be stimulated electrically. It was found by two independent studies (Bal et al. 2000; Blumenfeld and McCormick 2000) that, indeed, strong stimulation of corticothalamic fibers can “force” the thalamus to oscillate at 3 Hz and that this forcing is dependent on GABA B receptors. It was also shown that the “forced” 3 Hz oscillations are more synchronized than the natural 10 Hz spindle rhythm produced by the same thalamic circuits. These findings are all in agreement with the predictions of the model and therefore support the view that 3 Hz hypersynchronized oscillations can be generated by the thalamus subject to a hyperexcitable cortex.


In conclusion, computational models have provided a plausible explanation for a number of contrasting experimental results. First, the thalamus can produce a slow oscillation around 3 Hz, where GABA B receptors are important, but this oscillation is not the primary cause for ~3 Hz spike-and-wave discharges seen in epileptic seizures. The model can reproduce the experiments showing that the disinhibited thalamus does not produce spike-and-wave in cortex.

Second, seizures can be generated by a hyperexcitable cortex, but the intact thalamus is necessary. The generation of seizures depends on an interaction between cortex and thalamus. An essential ingredient in this interaction is that the corticothalamic feedback must be very efficient in evoking inhibition in the thalamus. If this is the case, the models show that a hyperexcitable cortex will generate a feedback strong enough to recruit GABA B receptors in the thalamus and “force” the slow oscillatory mode (although the thalamus is physiologically intact).

If these ingredients are assembled in a thalamocortical model, ~3 Hz spike-and-wave oscillations can be generated if the cortex alone is made hyperexcitable, but the thalamus is kept physiologically intact (Destexhe 1998). This thalamocortical mechanism is dependent on the presence of GABA B receptors in the thalamus, which are responsible for the ~3 Hz oscillation of the entire system. If the thalamus has weak GABA B -mediated inhibition, then the hyperexcitable system oscillates at a faster frequency, around 8 Hz, as found in rodents. Thus, the models predict that a common mechanism underlies absence seizures in all mammals but that the exact oscillation frequency will be dependent on the balance between GABA A and GABA B receptors in the thalamus (Destexhe 1999).

Finally, this model is compatible with the finding that a focus in cerebral cortex is associated to absence seizures in rats (Meeren et al. 2002). It was found that deep (layers 5–6) cortical neurons are hyperexcitable and seem to lead the discharges during ictal activity (Polack et al 2007). As layer 6 neurons project to thalamus, this finding is consistent with the idea that an excessive corticothalamic feedback may be a primary cause for absence seizures.


  1. Bal T, Debay D, Destexhe A (2000) Cortical feedback controls the frequency and synchrony of oscillations in the visual thalamus. J Neurosci 20:7478–7488PubMedGoogle Scholar
  2. Blumenfeld H, McCormick DA (2000) Corticothalamic inputs control the pattern of activity generated in thalamocortical networks. J Neurosci 20:5153–5162PubMedGoogle Scholar
  3. Crunelli V, Leresche N (2002) Childhood absence epilepsy: genes, channels, neurons and networks. Nat Rev Neurosci 3:371–382PubMedCrossRefGoogle Scholar
  4. Destexhe A (1998) Spike-and-wave oscillations based on the properties of GABAB receptors. J Neurosci 18:9099–9111PubMedGoogle Scholar
  5. Destexhe A (1999) Can GABAA conductances explain the fast oscillation frequency of absence seizures in rodents? Eur J Neurosci 11:2175–2181PubMedCrossRefGoogle Scholar
  6. Destexhe A (2013) Network models of absence seizures. In: Faingold CL, Blumenfeld H (eds) Neuronal networks in brain function, CNS disorders and therapeutics. Elsevier, Amsterdam (in press)Google Scholar
  7. Destexhe A, Sejnowski TJ (1995) G-protein activation kinetics and spill-over of GABA may account for differences between inhibitory responses in the hippocampus and thalamus. Proc Natl Acad Sci U S A 92:9515–9519PubMedCentralPubMedCrossRefGoogle Scholar
  8. Destexhe A, McCormick DA, Sejnowski TJ (1993) A model for 8–10 Hz spindling in interconnected thalamic relay and reticularis neurons. Biophys J 65:2474–2478Google Scholar
  9. Destexhe A, Bal T, McCormick DA, Sejnowski TJ (1996) Ionic mechanisms underlying synchronized oscillations and propagating waves in a model of ferret thalamic slices. J Neurophysiol 76:2049–2070PubMedGoogle Scholar
  10. Destexhe A, Contreras D, Steriade M (1998) Mechanisms underlying the synchronizing action of corticothalamic feedback through inhibition of thalamic relay cells. J Neurophysiol 79:999–1016PubMedGoogle Scholar
  11. Gloor P, Fariello RG (1988) Generalized epilepsy: some of its cellular mechanisms differ from those of focal epilepsy. Trends Neurosci 11:63–68PubMedCrossRefGoogle Scholar
  12. Golomb D, Wang XJ, Rinzel J (1996) Propagation of spindle waves in a thalamic slice model. J Neurophysiol 75:750–769PubMedGoogle Scholar
  13. Meeren HK, Pijn JP, Van Luijtelaar EL, Coenen AM, Lopes da Silva FH (2002) Cortical focus drives widespread corticothalamic networks during spontaneous absence seizures in rats. J Neurosci 22:1480–1495PubMedGoogle Scholar
  14. Neckelmann D, Amzica F, Steriade M (2000) Changes in neuronal conductance during different components of cortically generated spike-wave seizures. Neuroscience 96:475–485PubMedCrossRefGoogle Scholar
  15. Polack PO, Guillemain I, Hu E, Deransart C, Depaulis A, Charpier S (2007) Deep layer somatosensory cortical neurons initiate spike-and-wave discharges in a genetic model of absence seizures. J Neurosci 27:6590–6599PubMedCrossRefGoogle Scholar
  16. Steriade M, Contreras D (1998) Spike-wave complexes and fast components of cortically generated seizures. I. Role of neocortex and thalamus. J Neurophysiol 80:1439–1455PubMedGoogle Scholar
  17. Vergnes M, Marescaux C (1992) Cortical and thalamic lesions in rats with genetic absence epilepsy. J Neural Transm 35(Suppl):71–83Google Scholar
  18. von Krosigk M, Bal T, McCormick DA (1993) Cellular mechanisms of a synchronized oscillation in the thalamus. Science 261:361–364CrossRefGoogle Scholar

Further Reading

  1. Destexhe A (2007) Spike-and-wave oscillations. Scholarpedia 2:1402. http://www.scholarpedia.org/article/Spike-and-Wave_Oscillations
  2. Destexhe A, Sejnowski TJ (2001) Thalamocortical assemblies. Oxford University Press, OxfordGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  1. 1.UNICCNRSGif-sur-YvetteFrance