A High-Throughput Electrophysiology and Machine Learning Pipeline for ALS Pharmaceutical and Electroceutical Development
Abstract
Familial ALS (fALS), which accounts for around 10% of ALS cases, is caused by changes in a person s genetic sequence that is inherited amongst families. For fALS more than 30 genes have already been implicated in causing ALS. As the list of disease-causing genes increases, it has become apparent that these mutations impact various cell functions, such as the processes that regulate responses to stress or the production or degradation of proteins. Sporadic ALS (sALS) accounts for the remaining 90% of ALS cases, and it is thought that a combination of environmental factors and genetic predisposition causes the onset of ALS. Since ALS is a multifaceted disease, the timeframe and pathways that cause motor neuron degeneration differ substantially between patients. However, there are common pathologies and symptoms that are shared amongst ALS patients. A well-documented feature observed in ALS patients is termed neuronal hyperexcitability. Hyperexcitability describes an increased or exaggerated response to a stimulus. Neurons communicate with one another via electrical signals and these electrical signals increase in magnitude and frequency in ALS patients, even prior to the presentation of motor symptoms. This means that the processes that control how motor neurons communicate with one another are affected even before symptom onset in ALS. Neuronal excitability changes can be measured in patients in the clinic throughout the disease course and these alterations have been identified in different fALS cohorts bearing distinct genetic mutations, as well as in large cohorts of sALS patients. This suggests two key points: Firstly, that despite different genetic causes of the disease, hyperexcitability is a common pathology, and, secondly, that as one of the earliest features of the disease it could be targeted by a therapy to prevent motor neuron loss. Induced pluripotent stem cell (iPSC) technology can be used to convert a patient s skin cells into motor neurons, which offers a way forward to the development of novel therapies. This process allows us to generate large numbers of patient-specific cells for drug discovery, enabling the production of disease-specific cells from confirmed patients, even for those without a clear pattern of inheritance (i.e., sporadic cases). Importantly, these cells therefore represent the unique genetics of either fALS or sALS patients. Additionally, the motor neurons derived from ALS patients in vitro (in a dish) exhibit the same changes in neuronal excitability that are observed in patients in the clinic. The cells therefore provide a system to test different therapeutic approaches, and to assess whether they may be effective for certain subsets of patients or potentially even for many different patients. A major challenge for iPSC-based investigation is that ALS manifests within the complex three-dimensional (3D) architecture of the brain. This architecture has previously been difficult to reproduce, leading to some limitations with drug screening for novel pharmaceuticals. However, careful design of the drug screening assay parameters and detection methods should enable identification and preclinical validation of new lead compounds. Following methodological advancements to generate ALS neurons that truly represent patients and that can be tested in 3D formats, the next issue to overcome is how to identify drugs or a therapy that can reverse these neuronal excitability changes. One of the principal challenges in this step is how to make sense of vast amounts of information produced either by screening thousands of molecules or by assessing hundreds of variables in the electrical signals that are measured from ALS patient neurons. To address this, we have developed an innovative program of research that uses artificial intelligence and machine learning approaches to integrate all of this data. Our custom-designed pipeline uses a series of computer algorithms that build m
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 28, 2022
- Source ID
- W81XWH2210311
Entities
People
- Lezanne Ooi
Organizations
- United States Army
- University of Wollongong