Protein clumps, called amyloids, accumulate in the brains of people with Alzheimer’s disease. The current methods for detecting amyloids—positron emission tomography (PET) imaging of the brain and measurement of the clumping protein amyloid-β in cerebrospinal fluid—are expensive and invasive. Researchers want more easily accessible biomarkers that can provide the same information, but they’ve had no luck so far.
An international team led by Katsuhiko Yanagisawa of the Japanese National Center for Geriatrics & Gerontology now reports a promising set of amyloid biomarkers from blood plasma (Nature 2018, DOI: 10.1038/nature25456).
The researchers pull target peptides out of plasma using immunoprecipitation and then analyze them by mass spectrometry. Rather than use individual peptides as biomarkers, the researchers use ratios of peptides. They measure levels of three peptides derived from the amyloid precursor protein—one that is involved in amyloid formation and two that are not—and calculate the ratios.
The research team analyzed plasma samples from two independent data sets that contained samples from people who were cognitively normal, people who had mild cognitive impairment, and people who had Alzheimer’s-related dementia. The biomarker ratios were approximately 90% accurate in predicting brain amyloid-β levels measured by PET imaging.
Christian Humpel, an Alzheimer’s researcher at the Medical University of Innsbruck, calls the work an “interesting scientific finding,” but he questions whether it will make a good routine method. The method is more complex than the immunoassays typically used in clinical analyses. Furthermore, he says, it remains to be seen if the results can be reproduced in other labs.