The challenge ahead in neurological disease research
Approximately one percent of newborns are born with paediatric epilepsy,1 yet there are no diagnostic biomarkers that can help guide treatment.2 Neurologists must use trial and error to identify an effective anti-seizure treatment, but even with the best care, around 30 percent of patients fail to respond to any treatments.2 With every delay in finding an effective treatment, children experience more seizures, which can lead to cognitive and behavioural deficits and even early death.
Neurologists also face challenges treating people with amyotrophic lateral sclerosis (ALS), a disease that usually strikes during middle age and affects approximately one in every 50,000 individuals.3 ALS causes rapid deterioration of motor neurons, leading to muscle weakness, paralysis and death. Although scientists have made significant progress in identifying ALS‑associated genetic mutations and several precision therapeutics are in the pipeline, there are still no definitive diagnostics for the disease. Furthermore, currently available treatments are limited in their ability to reduce the severity of ALS symptoms.
I believe the current lack of precision and diagnostic options in neurological medicine demands a new approach to understanding the brain. There is an urgent need to find diagnostic biomarkers and corresponding treatments for neurological diseases such as epilepsy and ALS. By building better in vitro human-based models of these diseases, researchers can better evaluate therapeutic options. However, to achieve the next breakthrough discoveries, researchers need simple, reliable, high-throughput tools.
Most electrophysiological methods, including whole-cell patch-clamp electrophysiology, are highly challenging to execute and require specialised techniques, years of training and expensive equipment. In addition, scientists using the patch-clamp technique can often only examine a handful of cells per day, limiting how many treatments may be evaluated in a given timeframe.
The next generation of neuroscience techniques
Scientists have used multielectrode arrays (MEAs) to study neural activity on the surface of the brain for many years, but researchers are now adapting this technology for use in multi-well plates to study neural circuits in a dish. There are several reasons that these ‘bioelectronic assays’ are especially useful for identifying precision treatments for neurological diseases. Patient‑specific iPSC models allow biologists to recapitulate an individual’s disease in vitro. A scientist can take a patient’s somatic cells, eg, skin or blood cells, convert them into iPSCs and differentiate them into neurons of the same genetic background. These cells may be cultured and measured on MEA plates for months. This gives the cells time to develop and form synapses, providing scientists with the opportunity to study dynamic long-term effects on neural networks.
Combined with iPSCs, bioelectronic assays allow scientists to non-invasively monitor neural network behaviour in real-time and provide comprehensive data that better reflects human brain activity. In my lab, we have used bioelectronic assays to identify disease mechanisms to target for potential therapeutic interventions for epilepsy and ALS.
Using bioelectronic assays to study epilepsy
Epilepsy is difficult to study outside of human patients. Animal models do not always reflect human-specific symptoms or clinically valid biomarkers, making it hard to test novel therapeutics. Furthermore, even when studying epilepsies in human patients, symptoms and causes can vary, making it challenging to identify effective treatments.
To study epilepsy, my team developed a disease model using patient-derived iPSCs. This approach, paired with a bioelectronic assay, allowed us to study the effectiveness of various treatments for a patient-specific subtype of the condition, KCNQ2-associated epilepsy. To test our method, we looked for a characteristic firing pattern seen in the brains of children with this form of epilepsy, called the ‘burst‑suppression activity pattern’. This pattern is characterised by intermittent periods of highly synchronised firing followed by periods of low activity.4 When we compared the electrical behaviour of our neurons with those in the brains of children with KCNQ2-associated epilepsy using electroencephalography (EEG), we found that our cells displayed a similar burst-suppression firing pattern. These results told us that our bioelectronic assay could be used to test treatments against patient-specific ‘epilepsy-in-a-dish’.
As a proof-of-concept, we measured the ability of small molecules to restore defects seen in this model. For example, when we apply a KCNQ2 channel agonist, we can rescue a number of defective neuronal features. This indicates that a bioelectronic assay has the potential to be used to study novel therapeutics in vitro and identify promising candidates. Indeed, while I was still a postdoctoral fellow at Harvard, we used a similar assay to find an epilepsy drug that treats another debilitating condition, ALS.
Identifying a diagnostic biomarker for ALS
To identify potential ALS treatments, we created an ‘ALS-in-a-dish’ model and studied it using a bioelectronic assay.5,6 To begin, we created iPSC-derived neurons from ALS patients carrying the SOD1A4V mutation, a variant that causes a particularly severe form of the disease. We grew these cells on an MEA plate to study their activity.
This approach will hopefully enable scientists to gain a deeper understanding of neurological disease”
The assay revealed that these neurons were hyperexcitable.6 When we corrected the SOD1 mutation in our neurons, the cells were no longer hyperexcitable, indicating that hyperexcitability could serve as a biomarker of SOD1A4V ALS.
We also used classic patch-clamp techniques to identify functional abnormalities among ion channels that contribute to this hyperexcitability. Finally, using both patch clamp and a bioelectronic assay, we found that treatment with the anticonvulsant drug ezogabine, which activates Kv7 channels, restored neuronal activity to normal and improved cell survival.5,6 We, alongside other researchers, have also shown that neurons derived from several other genetic subtypes of ALS exhibit excitability defects.6,7 These data demonstrate the value of combining patch-clamp and MEA techniques to develop a comprehensive understanding of treatment effectiveness.
Based on these findings, researchers treated people with ALS with different doses of ezogabine in a Phase II clinical trial and observed a dose-dependent decrease in neuronal excitability.8 However, it still remains to be determined whether ezogabine or other classes of drugs that modulate Kv7 can slow ALS disease progression meaningfully, as a Phase III efficacy clinical trial has yet to be performed.
The future of neurology
Neurological diseases are notoriously hard to study and treat, in part due to a lack of reliable biomarkers that enable scientists to define disease subtypes accurately. The trial-and-error approach to treating most neurological diseases underscores the need for better diagnostics and personalised treatments. Definitions of many neurological diseases may also be too broad; what we define as one disease may in fact be a variety of diseases that each require different treatments. Scientists need better models to understand the underlying mechanisms and potential therapeutic targets of neurological disease so that patients can be matched with the right drugs more efficiently.
iPSC-derived models of diseases like KCNQ2-associated epilepsy and ALS give scientists greater precision and accuracy when testing potential treatments for specific genetic biomarkers. Furthermore, a bioelectronic assay enables scientists to study these disease subtypes over long periods, creating opportunities to see how these mutations impact physiology over time and respond to treatment. This approach will hopefully enable scientists to gain a deeper understanding of neurological disease, enabling drug manufacturers to develop more precise treatments so patients can live longer, fuller lives.
About the author
Dr Evangelos Kiskinis is Assistant Professor of Neurology and Neuroscience, a New York Stem Cell Foundation Robertson Investigator and Scientific Director of the Stem Cell Core Facility at Northwestern University Feinberg School of Medicine. His lab focuses on studying neurological diseases using stem cell-based approaches.
- Aaberg KM, Gunnes N, Bakken IJ, Søraas CL, et al. Incidence and prevalence of childhood epilepsy: a nationwide cohort study. Pediatrics. 2017;139(5):e20163908.
- Couzin-Frankel, J. As epilepsy drugs fail nearly one-third of patients, scientists seek root causes of seizures. Science. 2019 [cited 8 August 2022]. Available from: https://www.sciencemag.org/news/2019/12/epilepsy-drugs-fail-nearly-one-third-patients-scientists-seek-root-causes-seizures
- Jones P. FYI: epidemiology of ALS and suspected clusters. 2020 [cited 8 August 2022]. Available from: https://www.als.org/navigating-als/resources/fyi-epidemiology-als-and-suspected-clusters
- Simkin D, Marshall KA, Vanoye G, Desai RR, et al. Dyshomeostatic modulation of Ca2+-activated K+ channels in a human neuronal model of KCNQ2 encephalopathy. eLife. 2021;10:e64434.
- Kiskinis E, Sandoe J, Boulting G, Williams LA, et al. Pathways Disrupted in Human ALS Motor Neurons Identified Through Genetic Correction of Mutant SOD1. Cell Stem Cell, 2014 Jun 5;14(6):781-95.
- Wainger BJ, Kiskinis E, Mellin C, Wiskow O, et al. Intrinsic membrane hyperexcitability of amyotrophic lateral sclerosis patient-derived motor neurons. Cell Rep. 2014;7(1):1-11.
- Devlin AC, Burr K, Borooah S, Foster JD, et al. Human iPSC-derived motoneurons harbouring TARDBP or C9ORF72 ALS mutations are dysfunctional despite maintaining viability. Nat Commun. 2015 Jan 12;6:5999.
- Wainger BJ, Macklin EA, Vucic S, McIlduff CE, et al. Effect of ezogabine on cortical and spinal motor neuron excitability in amyotrophic lateral sclerosis. JAMA Neurol. 2021;78(2):186-196.