A spate of human genetic studies has shed light on the root causes of idiopathic focal epilepsy, a common childhood form of the disease. The studies suggest that a group of epilepsies and encephalopathies marked by seizures and language and learning disorders can result from mutations in NMDAR subunits.

Because some of the epilepsy-associated mutations appear to cause excessive NMDAR activity, the findings build a case for testing receptor subunit-selective antagonists to treat idiopathic focal epilepsy and possibly other forms of epilepsy.

At least two companies developing subunit-selective NMDA modulators-Naurex Inc. and Mnemosyne Pharmaceuticals Inc.-see childhood epilepsy as a potential new area of focus if basic mechanistic questions as well as a clinical testing strategy can be worked out.

NMDARs are a family of ion channels used by neurons to relay signals in response to glutamate, an excitatory neurotransmitter.

In the 1980s and 1990s, nonselective NMDAR channel blockers such as ketamine were tested in a range of neurology indications including epilepsy but encountered safety problems including dissociation and psychosis.

"NMDARs have been of interest in epilepsy since the 1980s, but at that time we didn't have selective compounds," said Henrik Klitgaard, VP of CNS research at UCB Group. "The first NMDA nonselective antagonist tested for epilepsy had very severe psychiatric effects, so that approach was abandoned."

Today, the modestly potent NMDAR antagonist Namenda memantine is marketed for Alzheimer's disease (AD) by Forest Laboratories Inc. and Merz GmbH & Co. KGaA.

Now, findings from three independent international teams1-3 may rekindle interest in testing newer, more selective NMDA modulators in childhood epilepsy.

The studies show that up to 20% of idiopathic focal epilepsy cases arise from mutations in an NMDAR subunit called NMDA receptor NR2A subtype (GRIN2A; NR2A).

Search and seizure

The teams converged on NR2A by studying families with a history of idiopathic focal epilepsy, a broad set of disorders including rolandic epilepsy, atypical benign partial epilepsy, Landau-Kleffner syndrome (LKS) and epileptic encephalopathy with continuous spike and waves during slow-wave sleep (CSWSS).

"Electroencephalographic abnormalities are the hallmark of idiopathic focal epilepsy, which begins in childhood," said Sarah
von Spiczak, a professor of pediatrics at Kiel University. "This is about 15%-20% of pediatric epilepsy cases."

Spiczak and coleader Holger Lerche, a professor of neurology at the University of Tuebingen, previously identified three patients with idiopathic focal epilepsy who had deletions encompassing NR2A.4

In their new study, the duo found point mutations or microdeletions in NR2A in 30 of 259 patients. Those with NR2A mutations typically had more severe forms of idiopathic focal epilepsy than patients with intact NR2A.

Independently, a team led by Pierre Szepetowski, research director at the Centre National de la Recherche Scientifique (CNRS) and the Mediterranean Institute of Neurobiology (INMED) at Aix-Marseille University, identified NR2A mutations as the likely cause of about 20% of cases among a panel of 66 patients with LKS or CSWSS.

Finally, a team led by University of Washington assistant professor of pediatrics Heather Mefford and The University of Melbourne professor of medicine and pediatrics Ingrid Scheffer sequenced NR2A in 519 pediatric patients with a broad range of epilepsy and language disorders. Among the 44 patients who had the electroencephalographic abnormalities and speech disorders that are the hallmarks of idiopathic focal epilepsy, 4 had mutations in NR2A.

No patients in the cohort with other forms of epilepsy had NR2A mutations.

Altogether, the three studies suggest that NR2A mutations are the most common individual genetic cause of idiopathic focal epilepsy. NR2A mutations cause "between 7% and 20% of cases and are the only known genetic cause to date," said Szepetowski.

What remains unclear is what role NMDARs play in the remaining 80% of idiopathic focal epilepsy cases not attributed to NR2A mutations.

To find out, Mefford is doing more genetic studies in the hopes of identifying further risk factors of idiopathic focal epilepsy and other pediatric epilepsies. "We're still focusing on gene discovery, continuing the targeted candidate gene approach with a larger set of patients and parents," said Mefford.

Szepetowski is not convinced that more genes with as strong an effect as NR2A will turn up. "We don't know the extent of heterogeneity in the remaining 80% of cases," he said. "We don't even know if the other cases are even genetic."

He noted that idiopathic focal epilepsy previously was thought to involve autoimmune activity against NMDARs, but there was no evidence of autoimmunity in any of the patients in Szepetowski's cohort. However, it is still possible that autoimmunity plays a role in the remaining cases.

The findings of all three teams were reported in Nature Genetics.

Function disjunction

From a genetic standpoint, the disease-linked mutations are likely to lead to loss of function of NR2A protein. However, NMDARs are highly complex, with multiple points of interaction between subunits and multiple binding sites for glutamate, glycine and zinc, which fine-tune the receptors' activities.

Thus, it is hard to predict how epilepsy-associated mutations would affect NMDAR activity.

The teams' leaders were divided about whether NMDARs could make good drug targets but agreed that further mechanistic details about how the mutations lead to disease could paint a clearer picture about how to proceed with drug development.

"These studies argue for targeting NMDARs" in idiopathic focal epilepsy, said Mefford.

But Lerche and Spiczak cautioned that the functional consequences of the mutations, and thus the appropriate therapeutic strategy, are not well understood.

"The most severe mutations are total disruptions, but some of the point mutations are not entirely clear and [are] perhaps a gain of function," said Lerche.

Added Spiczak, "We don't have enough information to answer conclusively whether NR2A is a good target."

Szepetowski said that ongoing cell culture and animal studies of how the disease-linked mutations affect NMDAR activity will provide a better picture of how to treat idiopathic focal epilepsy, which may require individually tailored approaches for each type of NR2A mutation.

Each team took preliminary steps to characterize the functional consequences of some of the disease-associated NR2A mutations on NMDAR activity in cell culture.

The Spiczak and Lerche team found that one disease-linked mutation from among their cohort reduced the binding of zinc, which inhibits NMDAR activity. As a result, the mutant receptor responded normally to glutamate but could not be turned off by zinc.

Szepetowski's team tested a different disease-linked NR2A mutant in cell culture and saw a delay in the closing of the NMDAR's ion channel after a period of activation. As a result, the mutant receptor had a longer window of activity than the wild-type receptor control.

Mefford and Scheffer's team characterized a third NR2A mutation in cell culture and, like Szepetowski's team, saw a delay in channel closing.

Altogether, the functional analysis of disease-linked NR2A mutations suggests that in at least some patients with idiopathic epilepsy, NMDAR channels remain open or active for too long, allowing the influx of an excessive number of ions (see "NMDARs in epilepsy").

Loose fit

The new findings raise the possibility of repurposing existing NMDAR modulators for epilepsy. At least 20 NMDAR antagonists are marketed or in development for a broad range of neurology indications, but none are in development for epilepsy.

Indeed, the standard of care for most forms of epilepsy is anticonvulsant drugs that are thought to act downstream of the disorder's primary causes, which have been poorly understood until now.

"The pharmacopeia right now is very limited, but these studies usher in a new dawn of specific, targeted therapy" for epilepsy, said Joseph Moskal, a professor of biomedical engineering and director of the Falk Center for Molecular Therapeutics at Northwestern University. "The ideal pharmacological agent would dampen down the protein. You want a gentle dampener, not a robust antagonist. An awful lot of NR2A is needed elsewhere in the central nervous system."

Moskal is founder and CSO of Naurex, which is developing NMDAR modulators for pain and depression. The company's lead product is
, a partial agonist of the NMDAR's glycine-binding site. GLYX-13 is in Phase II testing for depression.

For now, Moskal said that further physiological characterization of epilepsy-associated mutations in cell culture and animal models would help work out exactly how to target NMDARs.

"I'd like to see these mutations expressed in transgenic mice to study hippocampal slices and brain physiology," said Moskal.

Frank Menniti, cofounder and CSO of Mnemosyne, said that targeting NR2A-bearing NMDARs could yield better results than the high-potency, nonselective antagonists of the past.

"NMDARs are so broadly involved in brain function and development that it's in some way surprising that there are specific effects of these mutations" that cause disease rather than outright failure of brain development, said Menniti.

Mnemosyne has discovery stage programs for positive and negative allosteric modulators of NMDARs for schizophrenia, depression and other neurological disorders.

UCB's Klitgaard said NR2A has not been as thoroughly explored as other NMDAR components such as NR2B (GRIN2B).

"There are NR2B-selective compounds, but I don't know if NR2A-selective compounds exist," said Klitgaard. "They may exist on the shelves of pharma."

Industry's challenge is to find NR2A-selective compounds that correct the abnormal activity of the mutated protein without perturbing NMDAR signaling throughout the brain.

"The glutamatergic system, especially NMDARs, is among the most complex neurotransmitter systems in the brain," said Szepetowski. "We would not want to just turn down this system because this might have very adverse effects."

An open question is whether modulating NR2A or other NMDAR subunits will be effective in other forms of epilepsy besides idiopathic focal epilepsy.

"My guess is that NR2A antagonists would have a much broader profile of antiepileptic activity, including adult forms, for which there has not yet been any genetic association with the NMDAR," said Menniti.

Another difficulty lies in clinical trial design in which palliative anticonvulsants are the standard of care.

"From a big pharma drug development standpoint, epilepsy has been a scary thing to get involved in," said Menniti. "If you have someone whose seizures are under control with an existing drug, you can't pull them off the drug to try your new compound."

None of the findings reported in the three papers have been patented.

Osherovich, L. SciBX 6(34); doi:10.1038/scibx.2013.916
Published online Sept. 5, 2013


1.   Lemke, J.R. et al. Nat. Genet.; published online Aug. 11, 2013; doi:10.1038/ng.2728
Contact: Sarah von Spiczak, Kiel University, Kiel, Germany
e-mail: s.vonspiczak@pedneuro.uni-kiel.de
Contact: Holger Lerche, University of Tuebingen, Tuebingen, Germany
e-mail: holger.lerche@uni-tuebingen.de

2.   Carvill, G.L. et al. Nat. Genet.; published online Aug. 11, 2013; doi:10.1038/ng.2727
Contact: Heather C. Mefford, University of Washington, Seattle, Wash.
e-mail: hmefford@u.washington.edu
Contact: Ingrid E. Scheffer, The University of Melbourne, Melbourne, Victoria, Australia
e-mail: scheffer@unimelb.edu.au

3.   Lesca, G. et al. Nat. Genet.; published online Aug. 11, 2013; doi:10.1038/ng.2726
Contact: Pierre Szepetowski, Aix-Marseille University, Marseille, France
e-mail: pierre.szepetowski@inserm.fr

4.   Reutlinger, C. et al. Epilepsia 51, 1870-1873 (2010)


      Aix-Marseille University, Marseille, France

      Centre National de la Recherche Scientifique, Marseille, France

      Forest Laboratories Inc. (NYSE:FRX), New York, N.Y.

      Kiel University, Kiel, Germany

      Merz GmbH & Co. KGaA, Frankfurt, Germany

      Mnemosyne Pharmaceuticals Inc., Providence, R.I.

      Naurex Inc., Evanston, Ill.

      Northwestern University, Evanston, Ill.

      UCB Group (Euronext:UCB), Brussels, Belgium

      The University of Melbourne, Melbourne, Victoria, Australia

      University of Tuebingen, Tuebingen, Germany

      University of Washington, Seattle, Wash.