Using stem cells to replace inhibitory neurons in the hippocampi of adult epileptic mice reduced hyperexcitability, seizures, and behavioral deficits, according to a report in the May 5 Nature Neuroscience online. It is the first time such a treatment has proven effective in adult mice with established epileptic seizures. Though stem cell treatments for humans are still a long way off, this approach could one day help people with epilepsy and other disorders that result from overly excitable neurons, such as Alzheimer’s disease (AD), suggested senior author Scott Baraban, University of California, San Francisco. Scientists who were not involved in the study also saw promise. “This is very encouraging for future stem cell replacement therapy,” said Yadong Huang, Gladstone Institute of Neurological Disease, also in San Francisco.

In epilepsy, dysfunctional inhibitory neurons allow cortical neurons to become overexcited, leading to seizures. Drugs that enhance transmission of GABA, the major inhibitory neurotransmitter, help, but do not treat all forms of epilepsy, and they often come with debilitating side effects. While studies in mice suggest that more targeted approaches—such as turning on or off specific populations of neurons—quell seizures (see Krook-Magnuson et al., 2013, and Paz et al., 2013), these methods require genetic manipulation. Some scientists have been exploring stem cell therapies instead. Baraban and colleagues previously showed that precursors of inhibitory neurons, transplanted into neonatal mice before the onset of epilepsy, prevented seizures from occurring (see Baraban et al., 2009). However, developmental factors in young brains may have helped the transplanted neurons form synapses and integrate into established neural networks, the authors surmised. Would cells differentiate and integrate in mature adult mouse brains, too?

To find out, first author Robert Hunt and colleagues isolated precursors to GABA-producing inhibitory interneurons from the medial ganglionic eminence (MGE) of embryonic mice and injected them bilaterally into hippocampi of healthy, two-month-old adults. Cells, engineered to produce green fluorescent protein for ease of tracking, migrated up to 1,500 micrometers from the injection site. Most resembled mature interneurons both morphologically and electrophysiologically, and they expressed transcription factors and genetic markers unique to GABA-producing cells. The differentiated cells also formed synapses with excitatory cells and differentiated into a diversity of GABAergic cell types.

Satisfied that these neurons successfully entered and functioned in the adult mouse brain, the researchers next wanted to know if they prevented seizures. They turned to pilocarpine, a chemical convulsant that induces epilepsy when injected systemically. Treated mice typically seize up, behave abnormally, and have no response to anti-epileptic drugs, mimicking temporal lobe epilepsy in adults (see Gröticke et al., 2007). Hunt and colleagues found that if they also injected MGE cells bilaterally into the hippocampi, the mice seized 92 percent less frequently and were calmer when handled than untreated controls. The MGE-treated animals also performed comparably to non-epileptic controls in an open field test of locomotion. They also found the hidden platform as quickly as normal mice in the Morris water maze test of spatial navigation. Stem cells did not change performance on the rotarod test of motor coordination, the elevated plus maze test of general anxiety, or the forced swim test, a measure of depression.

Though these findings are specific for epilepsy, they could have implications for many diseases linked to hyperexcitability or interneuron dysfunction, such as autism, schizophrenia, and even AD, said Jorge Palop, also at Gladstone. Michela Gallagher, Johns Hopkins University, Baltimore, Maryland, reported last year that an anti-epilepsy drug dials back hippocampal hyperactivity in people with mild cognitive impairment and improves their cognition (see ARF related news story). Huang found that interneurons are damaged in mice with the ApoE4 allele, the biggest genetic risk factor for AD (see Andrews-Zwilling et al., 2010).

Gallagher pointed out that the biggest subpopulation of interneurons that survived the graft procedure expressed somatostatin. These neurons are hit particularly hard by aging and in people carrying the ApoE4 allele, which could mean this replacement therapy would be useful for AD-related disorders, she said. “It will be important to see if these findings are constrained by the model they used,” she told Alzforum. Baraban said he plans to test these transplants in mice that model other forms of epilepsy and to figure out how the interneurons integrate into the functional circuit. Huang plans to test if transplanted MGE cells benefit AD mouse models. He also suggested testing if the transplants work in older mice and whether all or just a few subtypes of interneurons are needed to improve seizures.

“The work is very interesting and provocative. The hope for a new treatment for refractory epilepsy is exciting,” said William Mobley, University of California, San Diego. However, he noted that it is essential to remember that epilepsy is a circuit disorder. "It is possible, and even likely, that only by understanding the changes in circuit structure and function can we know whether a cell-based treatment is appropriate, how to deliver it, and how to monitor the effects on circuit and clinical function," he said. He cautioned that before this type of therapy is suitable for people, researchers will have to figure out how to define interneuron status in the hippocampus of the patient to discern what benefits might accrue. They will also need to make human-derived interneurons that are safe to inject, he added. Two papers in the May 2 Cell Stem Cell report progress on the latter front, turning human pluripotent stem cells and embryonic stem cells into MGE-like cells that functionally integrate as interneurons in rodent brains (see Nicholas et al., 2013, and Maroof et al., 2013). A recent study used these human MGE-like cells to boost cholinergic cells in mouse hippocampi and reverse memory deficits (see ARF related news story).—Gwyneth Dickey Zakaib


  1. Loss of synaptic inhibition is a well-established cause of seizures, and this new study supports previous work from this laboratory showing that transplanted interneuronal precursors can become active participants in a hyperexcitable circuit and silence seizures in a genetic mouse model of epilepsy. Here, the model employed was a healthy mouse injected with a chemical convulsant, pilocarpine, that induces a hippocampal seizure focus sharing similarities with human temporal lobe epilepsy, but different in that brain development was otherwise normal and the circuit properties, while prone to generating seizures, are vastly different. In this model, grafted precursors not only reduced seizures, but also even improved performance deficits on behavioral tests relevant to hippocampal function. The authors conclude the approach holds promise not only for intractable epilepsies, but also perhaps other disorders that include altered hippocampal function such as Alzheimer’s disease and autism.

    The groundbreaking aspects of this research are clear and mark a giant step toward a future where severe focal epilepsies might be managed by cellular repair of damaged brain tissue rather than surgical removal. However, the findings are so counterintuitive that the authors should almost be chastened for their modest restraint in the Discussion. During brain development, over 21 different specific types of interneurons are painstakingly wired to precisely modulate the timing and firing patterns of hippocampal neurons. Who would imagine, given their diverse, highly individualized “personalities,” that simple addition of inexperienced newcomers could re-stabilize a normal pattern of synaptic inhibition in a network that is so severely compromised? And that their fates and excitability, which shift dramatically in immature brain, would retain properties similar to those they are intended to replace? The epileptic circuit in this model has been well studied and displays remarkable evidence of molecular and structural remodeling. Apparently, these fresh cells receive sufficient anatomic and biological guidance from the hyperactive network to quell the seizures, and the precise positioning of GABAergic synapses and the ratio of peptide co-transmitters they release are not as important as we may have thought.

    While fresh interneurons may prove to be a panacea for lowering seizure thresholds, they may be less so for other measures of hippocampal function. An alternative view is that, whereas some behavioral measurements improved, this might be due to the reduction in seizures in these networks rather than the establishment of repaired hippocampal information processing.
    For the Alzheimer’s disease brain, the results are therefore less clear. So far, essentially all experimental mouse models of AD show seizure phenotypes, and recent data suggest that elimination of the seizures, for example, by tau removal, is accompanied by improved cognitive function. Some component of the cognitive loss may therefore actually represent an "epileptic pseudo-dementia" that may be reversible by silencing seizure activity. In the absence of seizures, it is unclear how well cellular grafting of interneurons, or any other type of cellular progenitor, will repair hippocampal function. Furthermore, the primarily neurodegenerative nature of the AD microenvironment suggests that even if temporarily effective, survival of the transplanted cells would be inexorably compromised, as they are in temporal lobe epilepsy, where cell death and hippocampal atrophy are also the hallmarks of the disease process. But if all we needed was a steady supply of fresh neurons, could an indwelling precursor brain cell reservoir supply them?

  2. This is an interesting article, but the benefits of replacing inhibitory neurons with stem cells in Alzheimer’s disease (AD) are not completely clear to me. Alterations of GABAergic neurons have been related to a variety of diseases, but in the cerebral cortex (neocortex and hippocampus), the vast majority of degenerating neurons in AD are pyramidal (glutamatergic) neurons. Since the dendritic spines on pyramidal cells represent the main postsynaptic elements of cortical excitatory synapses, and they are fundamental structures in memory, learning, and cognition, the disappearance of dendritic spines constitutes what is believed to be one of the most important early events in the pathogenesis of AD.

    Nevertheless, it is true that there may be a relationship between alterations of cortical circuits in Alzheimer’s disease and epilepsy or epileptiform brain activity. For example, we have observed that the membrane surfaces of pyramidal neurons in contact with plaques (in both humans and animal models) lack GABAergic perisomatic synapses, and that a large proportion of plaques are in contact with neurons (see DeFelipe comment). In experimental animals, small reductions in the efficacy of GABA-mediated inhibition can lead to synchronized epileptiform activity; hence, a loss of perisomatic GABA axon terminals may overexcite neurons in contact with plaques.

    The progressive appearance of plaques throughout the cerebral cortex would steadily increase the susceptibility of developing and propagating seizures by increasing the numbers of hyperactive neurons nearby. Thus, it is hard to see how stem cell replacement therapy may help to repair these widely spread changes in GABAergic circuits, and most importantly, what (if any) its relationship could be to the protection of pyramidal neurons, the most susceptible neurons in the cerebral cortex to damage in AD.

    See also:

    Comment by DeFelipe J. on Chicago: AD and Epilepsy—Joined at the Synapse? 9 Dec 2009.


    . Diminished perisomatic GABAergic terminals on cortical neurons adjacent to amyloid plaques. Front Neuroanat. 2009;3:28. PubMed.

    . Effects of amyloid-β plaque proximity on the axon initial segment of pyramidal cells. J Alzheimers Dis. 2012;29(4):841-52. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Epilepsy Drug Calms the Hippocampus, Aids Memory
  2. Cholinergic Neurons From Stem Cells Rescue Mouse Memories

Paper Citations

  1. . On-demand optogenetic control of spontaneous seizures in temporal lobe epilepsy. Nat Commun. 2013;4:1376. PubMed.
  2. . Closed-loop optogenetic control of thalamus as a tool for interrupting seizures after cortical injury. Nat Neurosci. 2013 Jan;16(1):64-70. PubMed.
  3. . Reduction of seizures by transplantation of cortical GABAergic interneuron precursors into Kv1.1 mutant mice. Proc Natl Acad Sci U S A. 2009 Sep 8;106(36):15472-7. PubMed.
  4. . Behavioral alterations in the pilocarpine model of temporal lobe epilepsy in mice. Exp Neurol. 2007 Oct;207(2):329-49. PubMed.
  5. . Apolipoprotein E4 causes age- and Tau-dependent impairment of GABAergic interneurons, leading to learning and memory deficits in mice. J Neurosci. 2010 Oct 13;30(41):13707-17. PubMed.
  6. . Functional Maturation of hPSC-Derived Forebrain Interneurons Requires an Extended Timeline and Mimics Human Neural Development. Cell Stem Cell. 2013 May 2;12(5):573-86. PubMed.
  7. . Directed differentiation and functional maturation of cortical interneurons from human embryonic stem cells. Cell Stem Cell. 2013 May 2;12(5):559-72. PubMed.

Further Reading


  1. . Synergistic GABA-Enhancing Therapy against Seizures in a Mouse Model of Dravet Syndrome. J Pharmacol Exp Ther. 2013 May;345(2):215-24. PubMed.
  2. . Autistic-like behaviour in Scn1a+/- mice and rescue by enhanced GABA-mediated neurotransmission. Nature. 2012 Sep 20;489(7416):385-90. PubMed.
  3. . Differentiation and functional incorporation of embryonic stem cell-derived GABAergic interneurons in the dentate gyrus of mice with temporal lobe epilepsy. J Neurosci. 2012 Jan 4;32(1):46-61. PubMed.
  4. . The promise of an interneuron-based cell therapy for epilepsy. Dev Neurobiol. 2011 Jan 1;71(1):107-17. PubMed.
  5. . Effect of neuronal precursor cells derived from medial ganglionic eminence in an acute epileptic seizure model. Epilepsia. 2010 Jul;51 Suppl 3:71-5. PubMed.
  6. . Grafting of GABAergic precursors rescues deficits in hippocampal inhibition. Epilepsia. 2010 Jul;51 Suppl 3:66-70. PubMed.
  7. . Apolipoprotein E4 causes age- and Tau-dependent impairment of GABAergic interneurons, leading to learning and memory deficits in mice. J Neurosci. 2010 Oct 13;30(41):13707-17. PubMed.

Primary Papers

  1. . GABA progenitors grafted into the adult epileptic brain control seizures and abnormal behavior. Nat Neurosci. 2013 Jun;16(6):692-7. PubMed.