Using artificial intelligence as a discovery tool, researchers have discovered that the hormone melatonin could offer protection against both aging and Alzheimer’s disease (AD).
They say their study sets out a reproducible “AI-to-clinic” paradigm that can provide breakthroughs towards clinically effective therapeutics targeting the intertwined complexities of aging and AD.
The findings demonstrate how sophisticated computational frameworks such as machine learning, deep learning, and network-based analytics can systematically mine and interpret vast genomic, transcriptomic, and pharmacological datasets to revolutionize drug discovery.
“Our AI-driven methodological framework offers a transformative approach to overcoming traditional drug development barriers, opening new avenues for precise multitarget therapeutics and advancing translational strategies for complex aging-related neurological disorders,” reported Yuan Sun, from the Chinese Pharmaceutical University, and co-workers in Science Advances.
Projections suggest that nearly a quarter of the world’s population will be over the age of 65 years within the next two decades. Accompanying this is an unprecedented rise in age-related diseases, particularly neurodegenerative disorders such as AD.
Aging is consistently recognized as the primary risk factor for AD, and therapies simultaneously targeting both aging mechanisms and AD pathogenesis hold immense clinical promise that may revolutionize the treatment landscape for age-related neurodegenerative diseases.
However, conventional drug discovery pipelines struggle with the complexity and interconnectedness inherent to the hallmarks of aging, such as chronic inflammation, mitochondrial dysfunction, and redox dyshomeostasis.
And around 200 clinical trials have been performed aimed at developing AD therapeutics have nearly all failed.
To investigate alternative approaches, the researchers developed an AI-guided method termed the pathway and transcriptome-driven drug efficacy predictor (PTD-DEP).
The model was specifically designed for the systematic identification and optimization of small-molecule candidates capable of targeting shared pathological pathways underlying aging and AD.
Using PTD-DEP, the team identified melatonin—commonly used for insomnia—as a promising candidate that showed therapeutic potential against both aging and AD.
Guided by proteolysis targeting chimera technology combined with CB-Dock2 computational prediction, Sun and colleagues uncovered p300 as a critical molecular target of melatonin.
Further integrative analyses were then performed using Cleavage Under Targets and Tagmentation, immunoprecipitation–mass spectrometry (IP-MS), single-cell RNA sequencing, and spatial transcriptomics, complemented by rigorous pharmacological validations.
This revealed that melatonin targets the p300/specificity protein 1 transcriptional complex localized within super-enhancer regions. This targeted interaction powerfully drives transcriptional activation of brain and muscle arnt-like protein 1, a pivotal regulator of circadian rhythm.
The efficacy of melatonin was robustly confirmed in vivo and in vitro across preclinical models of both AD and cellular senescence, underscoring its ability to mitigate hallmarks of aging as well as AD pathology at the same time.
“Collectively, our findings not only provide a clinically relevant therapeutic candidate for AD prevention and treatment but also represent a notable theoretical advancement in geroscience by elucidating a novel molecular link between aging and AD,” the researchers maintained.
