New protein atlas identifies distinct subtypes of neurodegenerative diseases

Neurodegenerative diseases form a tangled biological web with overlapping molecular signatures and symptoms. To decode this complexity, a multi-institute collaboration led by St. Jude Children's Research Hospital scientists developed the pan-neurodegeneration atlas (PanNDA). The atlas is a comprehensive survey of neurodegenerative disease "proteomes" containing information about protein levels, modifications and interactions. This resource, published today in Cell, provides a wide-ranging protein-based outlook to better understand the origins of neurodegenerative diseases and to aid in their diagnosis and treatment.

Neurodegenerative diseases often stem from protein misfolding or accumulation. These errors also disrupt binding partners, upstream and downstream effectors, and any connected pathways. By combining multiple proteomic strategies, co-corresponding authors Junmin Peng, PhD, St. Jude Departments of Structural Biology and Developmental Neurobiology, and Bin Zhang, PhD, Department of Genetics and Genomic Sciences at the Icahn School of Medicine at Mount Sinai, created PanNDA to understand and explore this network and how it is disrupted in these diseases.

Atlas points to new disease subtypes, biomarkers and networks

The atlas covers six major neurodegenerative diseases (Alzheimer's disease, Parkinson's disease, vascular dementia, Lewy body dementia, progressive supranuclear palsy and frontotemporal degeneration with TDP-43 pathology) and was developed from the proteomes of 2,279 people with one of these diseases. By analyzing this comprehensive dataset, the researchers identified alterations both unique to and shared between diseases, as well as distinct subtypes within individual diseases.

These diseases were often thought of as single diseases, but using PanNDA, we found three major subtypes of Alzheimer's, four in Lewy body dementia and four in frontotemporal degeneration. We also found over 20 proteins that may serve as biomarkers to separate Alzheimer's disease into its three subtypes - a significant clinical aid."

 Junmin Peng, PhD, St. Jude Departments of Structural Biology and Developmental Neurobiology

The researchers compared common and distinct proteomic "fingerprints" across diseases and found that the most dramatically changed proteins differ between them, with only a small subset shared by all. They also built networks, linking disease drivers to other affected proteins to enable the researchers to draw comparisons and distinctions among neurodegenerative diseases and their subtypes. 

"Protein network analysis revealed not only the global landscape of protein–protein interactions in each disease, but also local interaction subnetworks and key candidate driver proteins," Zhang said. "Highly predictive protein subnetworks and driver proteins may play causal roles in disease pathogenesis and therefore represent promising targets for therapeutic intervention."

"We identified all the major aggregated proteins previously known, but with our data, we could correlate these with other proteins that change alongside them," Peng said. "It's the first really large-scale, deep analysis of its kind, covering more than 10,000 proteins in the brain."

PanNDA also provides a key resource for future studies. "The project was a huge undertaking, and as such we want the results to serve the whole neurodegenerative disease research community," Peng said. "Eighty percent of the proteins we identified probably haven't been studied in the context of neurodegeneration at all. There are many new components to explore and pathways to identify."

PanNDA is available at https://penglab.shinyapps.io/pannda/

Digging deep to understand neurodegenerative disease

PanNDA is the latest result from a multi-institute collaboration that also performed another investigation, published recently in Cell and focused on Alzheimer's disease. In that study, researchers combined protein and RNA data to build networks and identify and validate drivers of Alzheimer's disease. The continued collaboration aims to expand understanding of the nuances behind each neurodegenerative disease to help doctors and researchers develop diagnostic tools and identify novel treatment routes.

"Subtype information can be combined with biomarkers to stratify patients and predict who will benefit from which treatment," Peng said. "This is just the beginning; it will take time to build up these molecular signatures. However, by following certain pathways and identifying protein functions, we hope to provide a deeper understanding of disease mechanisms and point to new treatment strategies."