AI platform uncovers new molecular targets for Alzheimer's disease

Groundbreaking study published in Alzheimer's & Dementia reveals the crucial links between the NAD⁺-mitophagy axis and neurodegeneration.

Insilico Medicine ("Insilico"), a clinical-stage biotechnology company powered by generative artificial intelligence (AI), today announced the publication of a landmark study in Alzheimer's & Dementia: The Journal of the Alzheimer's Association. In close collaboration with researchers from the University of Oslo (UiO) and Akershus University Hospital (Ahus) led by Dr. Sofie Lautrup and Prof. Evandro Fei Fang, the team utilized Insilico's proprietary AI biology platform, PandaOmics, alongside extensive experimental validation to uncover new molecular targets for Alzheimer's disease (AD) and other neurodegenerative conditions.

The research provides compelling evidence that the impairment of the NAD⁺-mitophagy axis plays a central role in both healthy brain aging and neurodegeneration. By examining large-scale human data, the team demonstrated that interventions modulating mitochondrial metabolism hold profound potential for diagnosing and treating age-related cognitive decline.

Our results demonstrate the strength of combining human data and artificial intelligence with rigorous experimental validation, and offer new directions for the development of biomarkers and future interventions for Alzheimer's disease." 

Dr. Sofie Lautrup, first and co-corresponding author, University of Oslo and Akershus University Hospital

Drug discovery for neurodegenerative diseases has historically faced severe challenges. Seeking a breakthrough, the international collaborative research team examined genes involved in mitochondrial function and NAD⁺ metabolism across 12 healthy brain regions, as well as in cerebrospinal fluid and blood. They compared gene-expression patterns during healthy aging with four major neurodegenerative diseases: Alzheimer's disease, Parkinson's disease, Huntington's disease, and amyotrophic lateral sclerosis (ALS).

The analysis revealed that widespread molecular changes in these pathways are already present during the early stages of disease progression. Notably, in Alzheimer's and Parkinson's diseases, several of these changes could be detected in blood samples, pointing to their immense potential as early blood-based biomarkers.

To translate these findings into therapeutic applications, the AI-driven target discovery platform PandaOmics was utilized to evaluate over 100 candidate genes involved in NAD⁺ and mitophagy pathways. This analysis prioritized five potential therapeutic targets for Alzheimer's disease: ULK1, OPA1, LAMP2, MFN1, and ATP6V0E1.

To verify the AI predictions, prioritized targets were evaluated across preclinical models, including C. elegans, a human Tau-mutant cell line, and APOE4/4 iPSC-derived cortical neurons. Functional modulation of these target genes directly altered disease pathology. Specifically, enhancing mitochondrial fusion via OPA1 activation increased cell viability and significantly attenuated Tau phosphorylation in APOE4/4 cortical neurons. Conversely, knockdown of key mitophagy drivers, such as MFN1 and LAMP2, exacerbated Tau aggregation, underscoring the critical role of mitochondrial quality control in suppressing neurodegeneration.

This research once again demonstrates how our AI biology platform can rapidly unlock new possibilities in complex fields like neuroscience. By uniting our computational capabilities with the molecular biology expertise of Dr. Evandro Fei Fang's laboratory, we are successfully translating cutting-edge algorithms into tangible therapeutic candidates and biomarker strategies."

Said Frank Pun, PhD, Head of Insilico Hong Kong, Insilico Medicine