Nurture Trumps Nature in a Model of Heritable Epilepsy: Will Bumetanide be the First Antiepileptogenic Drug?

Treatment During a Vulnerable Developmental Period Rescues a Genetic Epilepsy.

Marguet SL, Le-Schulte VTQ, Merseburg A, Neu A, Eichler R, Jakovcevski I, Ivanov A, Hanganu-Opatz IL, Bernard C, Morellini F, Isbrandt D. Nat Med 2015;21(12):1436–1444.

The nervous system is vulnerable to perturbations during specific developmental periods. Insults during such susceptible time windows can have long-term consequences, including the development of neurological diseases such as epilepsy. Here we report that a pharmacological intervention timed during a vulnerable neonatal period of cortical development prevents pathology in a genetic epilepsy model. By using mice with dysfunctional Kv7 voltage-gated K⁺ channels, which are mutated in human neonatal epilepsy syndromes, we demonstrate the safety and efficacy of the sodium-potassium-chloride cotransporter NKCC1 antagonist bumetanide, which was administered during the first two postnatal weeks. In Kv7 current–deficient mice, which normally display epilepsy, hyperactivity and stereotypies as adults, transient bumetanide treatment normalized neonatal in vivo cortical network and hippocampal neuronal activity, prevented structural damage in the hippocampus and restored wild-type adult behavioral phenotypes. Furthermore, bumetanide treatment did not adversely affect control mice. These results suggest that in individuals with disease susceptibility, timing prophylactically safe interventions to specific windows during development may prevent or arrest disease progression.

Commentary

Currently available antiepileptic drugs, while life-changing for millions of patients, are far from perfect. Many patients fail to achieve full seizure remission despite trials of several drugs or experience intolerable side effects. Even when successful, treatment is often lifelong, doing little to reverse the underlying pathological changes that lead to seizure generation. These limitations are particularly salient in neonatal and pediatric epilepsy syndromes, which are often refractory to treatment and associated with devastating cognitive delay or regression (1). One of the foremost goals of epilepsy research is thus to identify the cellular and network perturbations that produce recurrent seizures, in hopes of finding new interventions to prevent or reverse the process of epileptogenesis.

Marguet and colleagues used a transgenic mouse line expressing a dominant-negative Kv7.2 subunit (KCNQ2^G279S) throughout the brain. Kv7 channels pass a noninactivating current (I_M) that regulates neuronal excitability and firing patterns (2). The KCNQ2^G279S mice have dramatic reductions in I_M and consequent hyperexcitability of hippocampal pyramidal neurons (and likely cortical neurons, as well). A prior study noted frequent focal seizures in these mice with tonic limb or neck extension, and occasional generalized tonic–clonic seizures (3). Intriguingly, suppressing KCNQ2^G279S expression during a “critical period” of the first 2 postnatal weeks (using a tetracycline-suppressible expression system engineered into these mice) prevented behavioral seizures and pathological changes in the hippocampus later in life.

Marguet and colleagues now build upon these findings with a more clinically relevant paradigm, using the loop diuretic bumetanide during the same critical period to prevent the development of seizures in adult KCNQ2^G279S mice. The rationale for using bumetanide, an inhibitor of the NKCC1 cotransporter, is that GABA exerts an excitatory rather than inhibitory effect during the first postnatal week in mice as a result of the high intracellular chloride levels established by NKCC1 (4, 5). Subsequent upregulation of the potassium-chloride cotransporter KCC2 counteracts this effect, extruding Cl⁻ and establishing the more familiar inhibitory role of GABA_A receptors. The authors hypothesized that bumetanide treatment during the first 2 postnatal weeks would reduce the excitatory action of GABA, dampen overall network excitability, and perhaps counteract the pro-excitatory effects of KCNQ2^G279S expression.

The study began by confirming that CA1 pyramidal neurons of the mutant mice indeed lack I_M during the postnatal critical period and are therefore hyperexcitable in response to depolarizing current injections. In vivo single-unit recordings also showed an increased propensity for burst firing in mutant CA1 neurons. Of interest, in vitro and in vivo recordings of hippocampal local field potentials demonstrated no changes in afferent inputs to CA1, suggesting that loss of I_M may affect CA1 pyramidal neurons selectively at this developmental stage relative to other areas of the hippocampus. Consistent with these observations, immunohistochemical staining of 2-week-old animals showed increased markers of neuronal activity and microglial activation confined to area CA1. Finally, in vivo field recordings from V1 cortex also demonstrated increased cortical activity in the mutant mice.

Having established that both the hippocampus and cortex of KCNQ2^G279S mice are hyperexcitable during the early postnatal period, the authors set out to determine the effect of bumetanide during this critical developmental period. In wild-type mice, bumetanide treatment during the first 2 postnatal weeks did not affect performance on tests of motor coordination, spatial learning, or contextual fear conditioning later in adulthood, suggesting that this treatment regimen does not induce gross neurocognitive behavioral changes. Remarkably, administration of bumetanide to KCNQ2^G279S mice prevented nearly all of the previously noted pathological changes, including CA1 burst firing, microglial activation, and cortical spindle bursts. Most important, both behavioral and electrographic evidence of seizures were nearly eliminated in bumetanide-treated mutants.

Marguet and colleagues provide a promising proof-of-concept that nurture (pharmacologic intervention during a developmental critical period) can trump nature (the brain injury and chronic epilepsy expected from a KCNQ2 mutation). Whether or not their findings will be directly translatable to human patients remains to be seen. There is an ongoing clinical trial evaluating bumetanide as an adjunctive therapy in neonatal seizures caused by multiple etiologies, including genetic epilepsies (NCT00830531, http://clinicaltrials.gov). Although rare, dominant-negative KCNQ2 mutations can cause devastating epileptic encephalopathy in children (6). Of interest, while patients with KCNQ2 encephalopathy may achieve seizure control with conventional therapies, structural brain injury and the neurocognitive outcomes caused by the mutations appear to be independent of seizure control (7). With this in mind, the Marguet study using KCNQ2^G279S mice is particularly compelling, as it suggests that bumetanide therapy initiated at the proper developmental stage in patients with KCNQ2 channelopathies would not only reduce seizures but could potentially prevent epileptogenesis and brain injury. Furthermore, the presumed mechanism of action of bumetanide—reducing the excitatory effect of GABA—is distinct from the mechanism of hyperexcitability in KCNQ2^G279S mice, mediated by increased intrinsic excitability of CA1 pyramidal neurons. Thus, testing bumetanide in other mouse models of genetic epilepsy may reveal other opportunities to mitigate developmental epileptogenesis that could inform future clinical trials.

One important limitation to this approach therapeutically is that the analogous critical period for intervention in KCNQ2 mutations in humans is likely to be in utero during the third trimester (4), raising the challenge of not only establishing drug safety and efficacy during pregnancy but also necessitating prenatal diagnosis of a disease that is usually not diagnosed until after birth. Bumetanide has also shown promise in preventing epileptogenesis in other animal models with postnatal human correlates, including febrile seizures leading to subsequent temporal lobe epilepsy (8). While this may be a more readily translatable use of bumetanide, establishing this approach as a treatment for epileptogenesis poses other challenges. Because only a minority of children who experience a febrile seizure subsequently develop epilepsy, and typically not until years later, it would be very difficult to establish the efficacy of bumetanide (or any other intervention) on epileptogenesis and determine the most appropriate timing and duration of treatment. Overall, this study is a promising addition to the growing evidence that bumetanide has antiepileptogenic effects, but much remains to be done to determine if and how to implement this treatment in human patients.