By combining artificial intelligence with an infrared biosensor, researchers have been able to distinguish between different types of misfolded proteins implicated in a variety of neurodegenerative diseases.
Recent studies suggest that differences in the structure of harmful sticky protein clumps in the brains of patients with Parkinson’s disease could explain why the condition varies between people.
These protein clumps, called alpha-synuclein aggregates, also appear with distinct structural signatures in other neurodegenerative conditions, such as Alzheimer’s disease.
The above explains why there is no one-size-fits-all treatment for Parkinson’s and similar diseases, said researcher Hilal Lashuel, PhD, who heads a laboratory investigating the molecular and chemical biology of neurodegeneration at the École Polytechnique Fédérale de Lausanne in Switzerland.
“Right now, we don’t have a reliable way to capture and measure these differences in patients’ brains or biofluids,” he told Inside Precision Medicine.
“We believe that applying our biosensors-coupled to AI, along with other recent advancements that allow us to amplify these aggregates from biological fluids, could help us do just that.
“This could facilitate earlier diagnosis, better ways to group patients based on their specific condition, and personalized treatments based on both the chemical makeup and structure of these harmful protein clumps.”
Current, gold-standard methods used to identify biomarkers in neurodegenerative diseases, such as mass spectrometry and enzyme-linked immunosorbent assays, focus on quantifying the level of target proteins.
However, they are insensitive to changes in structural state and are therefore not able to discriminate between different protein forms.
The ImmunoSEIRA sensor, outlined in Science Advances, uses surface-enhanced infrared absorption (SEIRA) spectroscopy that is equipped with an immunoassay that identifies and captures these biomarkers.
Using the ImmunoSEIRA sensor, the researchers were able to structurally profile three conformational forms of alpha-synuclein associated with the pathology of Parkinson’s and other neurodegenerative diseases, including monomers, oligomers, and fibrils.
They were further able to quantify different structural motifs present in each conformation, and identify differences and similarities between them.
Combining the information from the ImmunoSEIRA with neural networks enabled identification of the specific different oligomers and fibril forms.
It was also able to quantitatively predict these from mixed alpha-synuclein aggregate samples.
Deepthy Kavungal, a EPFL PhD student who is the main author of the paper, told Inside Precision Medicine: “The versatile and unique ability to detect and quantify the oligomeric and fibril protein forms when present together makes our sensor a valuable tool in the future to monitor and understand the accurate effect of novel precision therapies targeting different structural forms of the proteins implied in neurodegenerative diseases.”
Researcher Hatice Altug, a professor at the EPFL institute of bioengineering, added: “Our goal is to continue to optimize and expand the capabilities of this sensor and evaluate its diagnostic potential in Parkinson’s disease and other diseases caused by protein misfolding.”
