Seizure activity results in calcium- and mitochondria-independent ROS production via NADPH and xanthine oxidase activation

Seizure activity has been proposed to result in the generation of reactive oxygen species (ROS), which then contribute to seizure-induced neuronal damage and eventually cell death. Although the mechanisms of seizure-induced ROS generation are unclear, mitochondria and cellular calcium overload have been proposed to have a crucial role. We aim to determine the sources of seizure-induced ROS and their contribution to seizure-induced cell death. Using live cell imaging techniques in glioneuronal cultures, we show that prolonged seizure-like activity increases ROS production in an NMDA receptor-dependent manner. Unexpectedly, however, mitochondria did not contribute to ROS production during seizure-like activity. ROS were generated primarily by NADPH oxidase and later by xanthine oxidase (XO) activity in a calcium-independent manner. This calcium-independent neuronal ROS production was accompanied by an increase in intracellular Na⁺] through NMDA receptor activation. Inhibition of NADPH or XO markedly reduced seizure-like activity-induced neuronal apoptosis. These findings demonstrate a critical role for ROS in seizure-induced neuronal cell death and identify novel therapeutic targets.Reactive oxygen species (ROS) contribute to neuronal damage and have been linked to excitotoxicity.^{1, 2, 3, 4} An increase in ROS generation has also been identified in acute neurologic disease such as stroke,^5,6 and recent evidence indicates that this may contribute to neuronal damage in seizures and epilepsy.^{7, 8, 9, 10} However, ROS measurements during seizure-like activity were predominantly performed in homogenates, extracellular fluids or brain regions with no clear demonstration of whether the ROS were of neuronal origin.^9,11,12 Moreover, these studies lacked the necessary temporal resolution to determine accurately the evolution of ROS generation during and after prolonged seizure activity. Such obstacles can be overcome by live cell imaging of ROS, which has emerged as a powerful tool to study disease mechanisms.¹³If seizure activity induces ROS production in neurons, then a critical question is which sources of ROS production are triggered by such activity. Previous studies have suggested that mitochondria are the primary source of ROS generation in seizure models.^8,14 However, there are alternative sources of ROS, in particular the enzymes NADPH oxidase and xanthine oxidase (XO). How these contribute to excitotoxicity during seizure activity is uncertain. That these enzymes may have an important role in seizure-induced ROS generation is suggested by two observations: (1) NMDA receptors have a pivotal role in seizure-induced neuronal damage¹⁵ and (2) direct pharmacologic activation of NMDA receptors can activate NADPH oxidase, increasing free radical production and consequently neuronal death.^5,16,17 There is also burgeoning evidence of a role for NADPH oxidase activation in chronic brain pathology secondary to psychosocial stress, which leads to the development of neuropathologic alterations, and also in neurodegenerative disease.^18,19Acute activation of NADPH oxidase in neurons has mainly been shown after direct pharmacologic activation of NMDA receptors via exposure to high levels of NMDA and this activation is calcium-dependent.^16,17 More recently, activation of NADPH oxidase has been shown during seizure activity.^9,20 These pathways also involved NMDA receptor activation and upregulation of NMDA receptor subunits NR1 and NR2B. Nonetheless, these studies used chemoconvulsant epilepsy models, which, in themselves, may have an impact on ROS generation. The mechanisms and relevance of activation of NADPH oxidase during seizure activity independent of chemoconvulsants is unclear, especially given the presence of alternative sources of ROS production. Moreover, XO may also represent a major potential source of ROS during periods of increased metabolism, such as that occuring during seizures. We have therefore asked whether NMDA receptor activation has a role in seizure-induced ROS production and which sources and mechanisms of ROS production are involved in its time course during seizure-like activity.Here, we demonstrate increased ROS generation during seizure-like activity. This is activity-dependent, but it is maintained by a Ca²⁺-independent pathway involving the activation of NMDA receptors, NADPH oxidase and XO at a later phase. Blocking NADPH oxidase and XO prevented seizure-induced neuronal cell death in vitro. We thus provide compelling evidence that these ROS-generating pathways are appropriate targets for preventing neuronal death in seizures.