From Stumbling Blocks to Stepping Stones: Progress in Treating Temporal Lobe Epilepsy With Stem Cell Transplantation

Abstract

The last three decades of scientific research provided a wealth of data on the brain origins, development, and functional roles of GABAergic interneurons and new insights into GABAergic interneuron dysfunction in different types of epilepsy. A stumbling block in treating GABAergic interneuron dysfunction in acquired temporal lobe epilepsy (TLE) has been the incapacity of the adult human brain to replace interneurons through adult neurogenesis. Recent advances in the field of stem cell biology led to the development of pluripotent stem cells (iPSCs), and this technology has been used in combination with effective differentiation protocols for generating GABAergic neurons from human iPSCs. Neuroscientists have now established that transplanting human iPSC-derived GABAergic interneurons into the hippocampus in rodent models of TLE can suppress spontaneous recurrent seizures. Basic research studies in mice further showed that interneuron transplants prevent some of the neuropathological hallmarks of TLE that contribute to hyperexcitability and epileptogenesis by forming new inhibitory synaptic connections within the host hippocampus and preventing neuropathological changes from developing. These basic scientific findings paved the way for a recent clinical trial testing human neuron transplantation in patients with severe TLE that is having promising early results.

Introduction

Over the past thirty years, neurogenesis research has enhanced the understanding of brain repair. However, the adult brain has limited neuron-replacement ability, particularly in temporal lobe epilepsy (TLE). Current surgical interventions can be effective but are constrained due to the temporal lobes’ roles in learning and memory. Emerging methods, like MRI-guided laser therapies, show promise but require further study. ¹ FDA-approved neuromodulation therapies offer incremental seizure reduction over time. ² Additionally, structural brain changes like hippocampal granule cell mossy fiber sprouting and loss of inhibitory interneurons complicate treatment.^2–5 While these treatment advances are promising, more effective treatments are needed to achieve lasting seizure control in patients with medically intractable TLE. Interneuron transplants in animal models show promise in suppressing seizures by integrating into overactive brain circuits, while also improving memory and psychological health—benefits other TLE treatments do not offer. Stem cell therapy involving laser-guided stereotaxic transplantation of human gamma-aminobutyric acid (GABAergic) interneurons is highlighted as a promising approach for patients unresponsive to conventional treatments.

This novel treatment is currently being explored in the first in-human clinical trial of a US FDA-regulated human interneuron product, NRTX-1001, sponsored by Neurona Therapeutics, in collaboration with investigators at the California Institute for Regenerative Medicine (CIRM).^6,7 The multicenter Phase 1/2 clinical trial evaluates the safety and efficacy of a single dose of NRTX-1001 for drug-resistant MTLE. In the first stage, up to 10 participants will receive either a starting or higher dose. Patients treated with NRTX-1001 will be monitored for safety, tolerability, and effects on epilepsy symptoms. Recruitment is ongoing at epilepsy centers across the United States (www.clinicaltrials.gov (NCT05135091)). NRTX-1001 is a cell therapy derived from human pluripotent stem cells consisting of fully differentiated interneurons that secrete GABA. The ongoing clinical trial investigating interneuron transplantation for TLE builds on years of research studying fetal progenitor cells and embryonic neurons in rodent models.

Sources of GABAergic Interneurons for Treatment of TLE

Stem Cells From the Embryo

The first attempt to repair the adult mammalian brain with precursors for embryonic GABAergic interneurons occurred 26 years ago. ⁸ In 1999, Wichterle and colleagues demonstrated that neuronal precursors from the embryonic medial ganglionic eminence (MGE), but not from the lateral ganglionic eminence (LGE) or neocortex, could migrate and differentiate into GABAergic interneurons across developing or adult cerebral cortex in the mouse. This finding sparked hope for replacing inhibitory interneurons in damaged brain areas like the hippocampus. The MGE and LGE are embryonic zones producing inhibitory interneurons that regulate brain activity by controlling signaling between neurons. Unfortunately, sources for developing GABAergic interneurons like the MGE and LGE disappear during development as interneuron progenitors terminally differentiate while their daughter cells migrate and disperse throughout the brain.

Stem Cells and Progenitors From the Postnatal Brain

Early evidence of adult neurogenesis of excitatory neurons in rodents came from Joseph Altman and Gopal Das in the 1960s, ⁹ but widespread acceptance followed decades later, with contributions from Fernando Nottebohm, Elizabeth Gould, and Peter Eriksson.^10–12 While the phenomenon is well established in many species, technical hurdles have made it challenging to definitively show that adult neurogenesis persists in the human brain.^13–15 Studies in this area are still highly controversial; adult neurogenesis studies are typically performed using histological approaches with resected tissue or postmortem samples, and the methods of fixation are highly variable. In addition, should adult neurogenesis be established in humans, it is likely restricted to excitatory hippocampal granule cells. Currently, it seems unlikely that a patient's own brain could provide a sufficient supply of interneurons for cell transplantation to treat TLE. However, alternative sources of human inhibitory neurons for transplantation have become possible.

Induced Pluripotent Stem Cells

In 2007, scientists led by Shinya Yamanaka made a groundbreaking discovery: they found a way to turn ordinary adult human cells into stem cells, which are special cells that can grow into almost any type of tissue in the body. They did this by adding four specific proteins (called transcription factors) to the cells. The stem cells, generated through the addition of the four transcription factors, are called induced pluripotent stem cells (iPSCs). This breakthrough opened new possibilities for personalized medical treatments for repairing damaged brain tissues, and it earned Yamanaka the Nobel Prize in 2012.^16–18

Understanding the embryonic origins, migration patterns, and transcription factors of forebrain GABAergic interneurons enabled the development of efficient protocols to generate these cells from iPSCs. The embryonic forebrain contains germinal zones—medial, lateral, and caudal ganglionic eminences—that produce most forebrain GABAergic interneurons for the cerebral cortex, hippocampus, and striatum. MGE-derived interneurons migrate from the ventral forebrain to the cortex and hippocampus, where they integrate into circuits and form inhibitory synaptic connections. Optimized in vitro protocols now replicate essential aspects of the transcription factor codes that specify distinct interneuron subtypes, such as those expressing somatostatin or parvalbumin in addition to GABA.^19,20 Using drugs like PD 0332991 (Palbociclib), which stop certain proteins from allowing cells to divide, helps ensure that lab-grown cells from stem cells don’t form tumors. However, because these cells come from a donor, patients still need immune-suppressing medications, like cyclosporine A, right after surgery to prevent the brain from rejecting the cells. The most reliable way to prevent immune rejection when implanting certain brain cells, like inhibitory interneurons, is to grow them from stem cells made from the patient's own cells. However, this method can be extremely expensive and complicated, costing hundreds of thousands of dollars per patient to create stem cells that meet clinical standards. To address these challenges, scientists are exploring new, more efficient techniques, such as simplifying the process of reprogramming and creating banks of clinical iPSC lines that can be matched to recipients for HLA-matched allografting. ²¹

Different types of interneurons have unique electrophysiological, dendritic, and axonal properties, along with specific peptides or calcium-binding proteins. For instance, parvalbumin-expressing basket cells strongly inhibit pyramidal neurons near their cell bodies, while chandelier cells target the initial axon segments of many pyramidal neurons. Similarly, somatostatin-expressing interneurons in the hippocampus provide feedback inhibition to granule and pyramidal cells. ²² Evidence from rodent models of TLE,^23,24 and human TLE postmortem samples^25,26 suggests that the density of somatostatin- and parvalbumin-expressing GABAergic interneurons are severely reduced in the epileptic hippocampus and subiculum.^25,26 Effective protocols for generating and enriching these specific types of GABAergic interneurons from human iPSCs in vitro now make it possible to generate GABAergic progenitors that differentiate into these types of interneurons after transplantation in rodent models of TLE.^27–29

Mechanistic Studies of GABAergic Interneuron Transplantation

In the last decade, studies in rodent models of temporal lobe epilepsy found that transplants of MGE-derived GABAergic progenitors included the cell types lost in TLE, such as the somatostatin or parvalbumin-expressing neurons, and were effective in suppressing spontaneous recurrent seizures.^30–35 Long-term continuous EEG recordings in mice showed that seizures were suppressed within 6 weeks of transplanting MGE-derived interneurons and that the suppression lasted for long periods. ³² Several hypotheses that have been proposed to explain how the transplanted cells exert their effects, which are not mutually exclusive, will be discussed below.

Nonsynaptic or Synaptic Release of GABA

One question is whether it is synaptic or extrasynaptic GABA release that mediates the positive effects of the transplanted cells. To distinguish between these possibilities, optogenetic stimulation can be used to excite the transplanted interneurons in brain slices while recording from nearby host brain neurons. ³⁶ With this approach, scientists found that light-induced activation of ChR2-expressing interneurons transplanted in the dentate gyrus of mice triggered hyperpolarizing postsynaptic potentials in granule cells in the vicinity of the transplants. ³³ This finding suggests that synaptic activation of the transplanted cells leading to GABA release might, in part explain how the cells suppress seizures.

Whether this effect is synaptic or extrasynaptic was further explored by co-localizing gephyrin, a postsynaptic scaffolding protein, at sites of putative GABAergic synaptic contacts formed by the transplanted interneurons. In areas receiving dense innervation by the transplants, large clusters of gephyrin were opposed to the presynaptic terminal boutons formed by transplanted MGE cells, suggesting that some molecules required for functional inhibitory synapses were targeted to terminal boutons formed by the transplanted cells. ³³ GABA works by activating two types of GABA receptors. Some of these receptors are found at connections between neurons and produce quick, short-lasting signals to reduce brain activity (called “phasic” signals). Others are located outside these connections and create steady, long-lasting signals to restrain neuronal activity (called “tonic” signals). Only a few types of GABA_A receptors, with a specific part called the α4 subunit, are responsible for the tonic signals and these are typically found extrasynaptically. Many studies suggest a close relationship between extrasynaptic GABA_A receptors and seizure activity. To address the role of extrasynaptic GABA_A receptors in seizure suppression by MGE transplants, scientists studied seizure suppression in the brain from α4 subunit receptor-deficient mice or wild-type littermates using electrophysiology. ³⁷ They found that MGE interneuron transplantation reduced seizure power in the wild-type mice, but not the α4 knockout mice. These data suggested that tonic inhibition created by extrasynaptic GABA receptors with the α4 subunit is necessary to reduce seizure activity. Together, these two examples suggest that GABAergic interneuron transplants may suppress seizures via synaptic and/or extrasynaptic GABA_A receptors. Further studies will be needed to distinguish between these possibilities.

Human Versus Mouse-Derived Cells in Preclinical Studies

Studies that involved transplanting human pluripotent stem cell (PSC)-derived MGE-like progenitors in rodent TLE models suggest that the human cells migrate extensively but can take much longer to mature and integrate synaptically into host brain circuits than mouse MGE-derived interneurons. Human cells must be transplanted into immunodeficient mice to prevent graft rejection. In a study of purified human GABAergic progenitors derived from H7 hPSCs (H7 is a human pluripotent cell line derived from the inner cell mass of an embryo), seizure suppression was observed 3 months after transplantation, before many of the neurons exhibited mature electrophysiological properties. ³⁸ However, in another study human iPSC-derived neurons appeared to require 5 months to synaptically integrate, suppress seizures, and improve cognition. ³⁹ While it is not yet clear why such large temporal differences were found, they may be due to technical differences; for example, different PSC cell lines were used and these can vary considerably in their differentiation into neurons. Furthermore, variations in laboratory protocols for mouse models of TLE and seizure monitoring methods may have contributed to these different results. The findings do however underscore the lengthy period required for stem cell-derived human GABAergic interneurons to differentiate and integrate into host brain circuits.

As mentioned earlier, the dentate gyrus of the hippocampus contains a population of adult-born neurons called granule cells, that continue to be replaced throughout life in rodents. Adult-generated granule cells in healthy brains are but during epilepsy development, adult-generated granule cells in healthy rodent brains play an important role in encoding memories and are generated throughout life. During a critical period in their maturation, the young adult-born granule cells suppress the firing of populations of pyramidal neurons in the dentate gyrus, a phenomenon thought to create sparse hippocampal activity patterns that are more effective for encoding new information in the rodent's environment. ⁴⁰ However, many studies in animal models show that as epilepsy develops, adult-generated granule cells develop abnormal dendritic morphology, migrate abnormally to ectopic positions in the hilus, and contribute to hyperexcitability and epileptogenesis..^41,42 In patients with TLE abnormal granule cells are present, although whether such cells are adult or embryonically generated is not known. Remodeling hippocampal circuits during epileptogenesis involves rewiring new excitatory feedback loops onto dentate granule cells, and this process likely increases brain excitation and worsens seizures. ⁴³ Moreover, ablating or silencing adult-born granule cells in TLE reduces seizures in animal models.^44,45

Based on these studies implicating adult granule cell neurogenesis in driving epileptogenesis and evidence that optogenetic activation of transplanted interneurons increases inhibitory currents in adult-born granule cells, Arshad and colleagues investigated whether interneuron transplantation in early stages of epileptogenesis prevented the formation of abnormal excitatory feedback loops onto dentate gyrus granule cells, particularly adult-born granule cells. ⁴⁶ The authors used a two-step viral tracing method to study brain connections in epileptic versus naïve mice with transplants. ⁴⁷

The study tracked new brain cells’ development and connections using viral markers, comparing adult-born granule cells in mice with or without TLE after they received interneuron transplants. Epileptic mice showed increased excitatory connections from hippocampal and cortical neurons, ⁴³ but GABAergic interneuron transplants significantly lowered the excitatory-to-inhibitory connection ratio. This reduction, spanning local and distant inputs, suggests that transplants may mitigate seizure-related excitatory circuit buildup by restraining brain activity or competing for connection sites.

Conclusion

In the past three decades, our understanding of the origins, development, and important roles of GABAergic interneurons in normal brain function and in neurological conditions such as temporal lobe epilepsy has changed remarkably. In just the last decade, much emphasis has been placed in identifying how to generate these important cells from human iPSCs and comparing them to fetal sources of GABAergic interneurons. These comparative approaches have contributed to understanding how to generate human interneurons for clinical applications. Basic research in rodent models of epilepsy has established that transplanting human or mouse fetal interneurons into the hippocampus in models of TLE leads to new synapse formation in the host brain and reduces seizures. Interneuron transplantation during epileptogenesis in rodents also appears to prevent the hallmark neuropathological features of TLE from forming. Further research identifying the underlying mechanisms for these effects can lead to a better understanding of the causes of epileptogenesis and how to prevent it.