Research Collaboration Yields New Treatment Possibilities for ALS
Research Collaboration Yields New Treatment Possibilities for ALS
Complete the form below and we will email you a PDF version of "Research Collaboration Yields New Treatment Possibilities for ALS"
In collaboration with researchers from Answer ALS, Johns Hopkins University School of Medicine, Harvard Medical School and Mayo Clinic, scientists from Insilico Medicine identified numerous unreported potential therapeutic targets of amyotrophic lateral sclerosis (ALS), using the company’s proprietary artificial intelligence (AI)-enabled biological target discovery platform PandaOmics. Among the 28 proposed candidates, 18 have been validated to moderately or strongly rescue eye degeneration in c9ALS Drosophila model. The findings underscore the power of PandaOmics to accelerate the novel target discovery process and provide new hope for future treatment options for ALS sufferers.
The need for new ALS treatments
ALS is a rare and lethal neuromuscular disease characterized by progressive loss of upper and lower motor neurons that control voluntary muscles. It is the most common motor neuron disease, with the mean survival time of patients ranging from two to five years post-diagnosis. Patients usually experience painless progression of muscle weakness and expire due to respiratory failure. Unfortunately, current FDA-approved drugs for ALS do not reverse any neurodegeneration in patients, and some treated patients even showed no clinical benefit. This calls for urgent developments of new therapeutic regimens.
“The results of this collaborative research effort shows what is possible when we bring together human expertise with AI tools to discover new targets for diseases where there is a high unmet need,” says Alex Zhavoronkov, PhD, founder and co-CEO of Insilico Medicine. “This is only the beginning.”
“From AI-powered target discovery based on massive datasets, to biological validation by multiple model systems (fly, mouse, human iPS cells), to rapid clinical testing through investigator-initiated trials (IIT), this collaboration represents a new trend that may dramatically reduce the costs and duration and more importantly the success rate of developing medicines, especially for neurodegenerative diseases,” said Bai Lu, professor at Tsinghua University and founder of 4B Technologies. “We are very happy to be part of this international team, and very excited about the subsequent efforts to clinically validate these novel targets.”
“We are truly excited to see the Answer ALS data being used to identify possible ALS disease-causing pathways and candidate drugs,” said Jeffrey D. Rothstein MD, PhD, director of the Robert Packard Center for ALS Research and Answer ALS. “The work by Insilico is exactly how this unprecedented program was envisioned to help change the course of ALS.”
“It is exciting to see the power of AI to help understand ALS biology,” said Merit Cudkowicz, MD, chief of neurology and director of the Healey & AMG Center for ALS at Mass General Hospital and Harvard Medical School and corresponding author. “Through Sean Healey and his friends, I was introduced to Dr. Zhavoronkov and the Insilico team. We immediately saw the potential to connect the Insilico team with the multidisciplinary Answer ALS team. We look forward to the next steps to turn this knowledge into new targets for treatments for people living with ALS.”
Using PandaOmics™ to find potential targets
To explore potentially actionable targets for ALS, Insilico Medicine utilized PandaOmics to analyze multiple publicly available transcriptomic datasets with post-mortem central nervous system (CNS) tissue, along with the transcriptomic and proteomic data using patient-derived iPSC-differentiated motor neuron (diMN) samples from Answer ALS. For each dataset, the ALS patients were divided into familial and sporadic subtypes. Comparisons were made between case and control samples independently for different tissues, ALS subtypes and data types. All the case-control comparisons belonging to the same comparison group were pooled into a single meta-analysis, yielding a total of six meta-analyses.
For each meta-analysis, PandaOmics prioritized the targets under two novelty settings (high-confidence and novel settings) with customized omics scores, text scores and druggability filters, yielding a total of 28 actionable targets. The targets were further validated using the Drosophila model with genomic editing in C9ORF72 (c9ALS Drosophila model) mimicking the most common genetic cause of ALS, to determine their functional correlations to the disease. The effects of the genes-of-interest (GOI) on neurodegeneration were determined by scoring the degree of degeneration of Drosophila eyes expressing RNA interference (RNAi) against GOI.
Target identification was performed with the public CNS tissue-based datasets, and diMN data from Answer ALS on PandaOmics. Targets were divided into two categories: novel targets for further investigation and targets for drug repurposing. The targets will be released onto ALS.AI. Feedback on proposed targets will be collected from ALS KOLs to select the best candidates for further validation. The identified targets will be further validated using in vivo and in vitro models.
Research yields new therapeutic targets for ALS
Seventeen high-confidence and 11 novel therapeutic targets were identified from CNS and diMN samples, which were disclosed in the paper and at ALS.AI. Researchers found that several well-characterized pathways in ALS pathology were dysregulated, including the immune system, RNA metabolism, excitotoxicity, as well as programmed cell death. CNS data mainly reflects the late-stage signatures of ALS (i.e., neuron cell death, neuroinflammation), while results from diMN comparisons are more likely to be attributed to the early-stage signatures (i.e., DNA damage, glutamate toxicity). Combining the usage of diMN and post-mortem CNS samples could provide a comprehensive understanding of ALS disease progression. Researchers validated 26 targets in the c9ALS Drosophila model, of which 18 (69%) demonstrated that their suppression rescued neurodegeneration, while the loss of RPS6KB1 resulted in an opposite effect. Representative images of fly eyes whose degenerations were strongly rescued by RNAi are shown in the figure below. This approach confirmed the power of PandaOmics to identify therapeutic targets with potential roles in ALS neurodegeneration.
Figure 1: Loss of seven unreported fly orthologs, corresponding to eight genes, strongly rescued C9orf72-mediated neurodegeneration in a Drosophila model. (A) A scale of magnitude of degeneration in fly eyes expressing (G4C2)30 scored from -4 to 2. The control flies (score 0), whose eyes expressing (G4C2)30, exhibited eye degeneration. The degree of eye degeneration rescue by RNAi of the GOI ranged from -4 to 2, where the positive or negative scores correspond to an increase or decrease in severity of eye degeneration. (B) Among the 26 targets with fly models, suppression of 18 targets (69%) using RNAi strongly or moderately (Scored ≤ -2) rescued eye degeneration. (C) More importantly, loss of 7 unreported fly orthologs, corresponding to 8 genes (shown in brackets), strongly rescued neurodegeneration.
The potential of Insilico’s AI
The current study applied PandaOmics to find novel targets and targets of drug repurposing for ALS. This is Insilico’s first paper showing the full potential of PandaOmics for target discovery with in vivo validation. Since the targets were identified from ALS patient post-mortem CNS tissues and patient-derived iPS neurons and further validated in c9ALS Drosophila model, these cross-validations strongly suggest the functional correlation between our AI-derived targets and ALS pathogenesis. Altogether, the present study offers new insights into how AI speeds up the target discovery process from years to months, and points to a potential direction to search for the treatment of ALS and other diseases.