Abstract
Zhu Q, Mishra A, Park JS, Liu D, Le DT, Gonzalez SZ, Anderson-Crannage M, Park JM, Park G-H, Tarbay L, Daneshvar K, Brandenburg M, Signoretti C, Zinski A, Gardner E-J, Zheng KL, Abani CP, Hu C, Beaudreault CP, Zhang X-L, Stanton PK, Cho J-H, Velíšek L, Velíšková J, Javed S, Leonard CS, Kim H-Y, Chung S. Neuron. 2023;111(6):807-823.e7. ISSN 0896-6273. doi:10.1016/j.neuron.2022.12.014
Previously, we demonstrated the efficacy of human pluripotent stem cell (hPSC)-derived GABAergic cortical interneuron (cIN) grafts in ameliorating seizures. However, a safe and reliable clinical translation requires a mechanistic understanding of graft function, as well as the assurance of long-term efficacy and safety. By employing hPSC-derived chemically matured migratory cINs in two models of epilepsy, we demonstrate lasting efficacy in treating seizures and comorbid deficits, as well as safety without uncontrolled growth. Host inhibition does not increase with increasing grafted cIN densities, assuring their safety without the risk of over-inhibition. Furthermore, their closed-loop optogenetic activation aborted seizure activity, revealing mechanisms of graft-mediated seizure control, and allowing graft modulation for optimal translation. Monosynaptic tracing shows their extensive and specific synaptic connections with host neurons, providing further evidence that this treatment for epilepsy could reliably help patients suffering from intractable epilepsy.
Commentary
Resected tissue from the temporal lobes of patients with severe mesial temporal lobe epilepsy (TLE) is often characterized by distinctive changes, such as mossy fiber sprouting, granule cell layer dispersion, and GABAergic interneuron degeneration. 1,2 These neuropathological changes contribute to an imbalance between excitation and inhibition in the dentate gyrus, leading to hyperexcitability and seizures. 2 Many patients with TLE develop pharmacoresistant seizures requiring surgical ablation of the seizure foci in the temporal lobes; while this is often effective, not all patients qualify for this approach.
To address this treatment gap, decades of research have focused on alternatives, including fetal neuron transplantation to repair damaged neural circuits. Societal restrictions on the use of aborted fetal tissue and limited quantities led scientists to pursue alternative tissue sources for generating human neurons for study and transplantation. One source is human embryonic stem cells that can be obtained from early-stage embryos just under a week old, collected at in vitro fertilization clinics. The inner cell mass of these early embryos contains versatile cells, called pluripotent stem cells (PSCs) that can divide to become any type of cell in the body. Although much is known about generating human neurons from human PSCs, detailed protocols exist for only a small number of cell types.
Identifying the signaling molecules required for instructing PSCs along specific pathways to generate unique classes of human neurons is challenging. Studies of rodent and human forebrain development identified the signaling pathways and transcription factors that specify hippocampal GABAergic interneurons within the medial ganglionic eminence (MGE), a transient structure in the embryonic forebrain. 3 The addition of these signaling molecules and factors to tissue cultures of PSCs produces human MGE-like GABAergic progenitors. 4 -8 This advance enabled further study of human GABAergic interneuron transplantation in rodent models of TLE, such as the intrahippocampal kainic acid model or the systemic pilocarpine model, that replicate neuropathological hallmarks of human TLE.
Among the first questions asked was whether GABAergic interneurons would survive transplantation into these rodent TLE models. Experiments in mice and rats showed that cells harvested from the rodent MGE or PSC-derived human GABAergic cells not only survived transplantation into the hippocampus but these transplanted cells also appeared to be effective in reducing seizures, at least in the short term. 4,7 -13 Until now however, it has been difficult to evaluate the long-term survival of transplanted PSC-derived human neurons in rodents with severe TLE, whether they continue to have disease-modifying effects, and the underlying mechanisms responsible for seizure suppression. Zhu and colleagues have now addressed these important questions through an approach that generates highly enriched and differentiated human GABAergic neural progenitors from PSCs. 5,6 After transplantation, they conducted video-electroencephalography (v-EEG), comparing short versus long-term survival periods, to ask whether the transplants were effective in suppressing seizures over the long run. Remarkably, the transplanted cells maintained efficacy for suppressing seizures even up to 9 months.
An additional concern was whether their culturing methods eliminated residual cells that retain pluripotency and can cause tumor formation after transplantation. Their modified differentiation protocol led to 85% of the cells expressing Sox-6, a transcription factor that is a marker of human-MGE like GABAergic interneuron lineages. And, when they treated the cells with 3 chemicals dubbed “CDP” (CultureOne supplement, DAPT, and PD0332991), they obtained even more homogenous populations of post-mitotic neurons that expressed Sox-6 (89%) and GABA (86%). Importantly, they found minimal expression of Ki67 (1%), a marker of cellular proliferation, and no expression of Oct 4, a marker for undifferentiated cells. These findings show that their transplants do not contain pluripotent cells and only very low numbers of dividing cells. It will be important for future clinical applications to eliminate dividing cells entirely to prevent the possibility of tumor formation.
A further advance in this new study was to determine whether the transplanted neurons fully matured, given the protracted development of a child’s brain and slower maturation of human neurons. By recording from the transplanted interneurons many months after transplantation, they found that these cells exhibited firing patterns and membrane properties characteristic of mature inhibitory interneurons. The transplanted cells also wired up with hippocampal neurons in the host brain, resulting in increased spontaneous inhibitory postsynaptic currents in host brain granule cells in the dentate gyrus. Recordings from transplanted interneurons showed that they had spontaneous excitatory postsynaptic currents, indicating that the transplanted cells received excitatory synapses from host neurons. They also wondered whether the synaptic integration of the transplanted cells was responsible for seizure suppression or were the transplanted cells acting in a paracrine fashion? Ideally, well-integrated GABAergic transplants would increase synaptic inhibition of the host brain when levels of neural activity ratcheted up at the onset of seizures.
They investigated this question by leveraging optogenetics, to ask whether manipulation of the firing rates of the transplanted cells altered seizure severity. By closed-loop optogenetic manipulations, they discovered that activating the transplanted interneurons made seizures shorter, whereas silencing the transplanted interneurons resulted in longer seizures. These findings provide direct evidence that the transplanted human GABAergic interneurons suppress seizures in an activity-dependent manner, through the new synaptic connections they form with host brain neurons.
To identify which host brain neurons participated in the newly formed neural circuits, Zhu and colleagues used viral-mediated tract tracing, an approach that relies on the retrograde transfer of a modified rabies virus across one synaptic connection, to reveal the neurons, locally and over long distances, that synapse with the transplanted cells. This type of rabies virus travels retrogradely across the synapses that have formed with the transplanted interneurons, but the virus cannot spread further or cause widespread cell death. With this approach, they showed that the transplanted interneurons received local hippocampal inputs and long-range cortical projections, as well as inputs from the amygdala and septum; this is strong anatomical evidence that new excitatory connections are formed by host brain neurons with the transplanted GABAergic neurons, and suggests that these new connections may replace endogenous connections with hippocampal interneurons that were lost as a consequence of status epilepticus and recurrent seizures.
This ambitious study advances the field of stem cell and neural transplantation research closer to its goal of developing clinically feasible cell-based treatments to help patients with intractable seizures and TLE. By combining novel stem cell biological approaches for generating human GABAergic interneurons, with state of the art electrophysiological, anatomical, and optogenetic studies, this new work provides evidence that transplanted human GABAergic cells survive, integrate into host brain circuits, and regulate seizures for extended periods of time. Before moving into the clinic, issues with graft versus host disease need to be solved to avoid the need for long-term immunosuppression in patients with transplants. Zhu and colleagues performed xenografts in immunodeficient mice to circumvent this problem. The discovery of the Yamanaka factors, 4 factors that can induce skin fibroblasts to dedifferentiate into pluripotent stem cells (iPSCs), could provide an approach for generating a patient’s own neurons and alleviate the need for immunosuppression.
