Summary: Current Alzheimer’s tests typically measure the quantity of amyloid beta and phosphorylated tau proteins in blood or cerebrospinal fluid. A new study from Scripps Research suggests a different approach: examining the three-dimensional shape, or folding, of circulating proteins. The researchers found that as Alzheimer’s progresses, specific plasma proteins adopt a less “open” structural conformation—a change that can be detected and used to classify disease stage.
By profiling structural features of three plasma proteins—C1QA, clusterin (CLUS) and apolipoprotein B (ApoB)—the team developed a proteostasis-based signature that distinguishes cognitively normal adults, people with mild cognitive impairment (MCI), and individuals with Alzheimer’s disease (AD). The structural panel produced high accuracy in both three-way and binary classifications and tracked changes over time, offering a promising blood-based biomarker strategy for earlier detection and monitoring of therapeutic response.
Key Facts
- Shape over concentration: The test detects structural folding changes in plasma proteins rather than relying solely on protein abundance, revealing early biological shifts linked to neurodegeneration.
- Progressive closing: Across patient groups, proteins tended to become less structurally exposed—or more “closed”—as disease advanced, a pattern that outperformed conventional protein-level measures.
- Three-protein panel: The most informative markers were C1QA (immune signaling), clusterin (protein folding and amyloid clearance) and apolipoprotein B (lipid transport and vessel health).
- Strong accuracy: The structural model achieved about 83% accuracy for three-way classification (healthy vs MCI vs AD) and exceeded 93% accuracy in distinguishing healthy adults from those with MCI in binary comparisons.
- Longitudinal utility: The panel classified follow-up samples with roughly 86% accuracy and correlated with cognitive test scores and MRI measures of brain atrophy, supporting use for disease tracking.
Source: Scripps Research Institute
Background: Alzheimer’s disease affects millions of older adults and current biomarker efforts often focus on measuring amyloid beta (Aβ) and phosphorylated tau (p‑tau). While those protein levels are informative, they may not capture earlier disturbances in protein homeostasis—proteostasis—that lead to misfolding and aggregation. Proteostasis declines with age, increasing the risk that newly made or maintained proteins fold incorrectly. The Scripps team reasoned that if proteostasis breaks down in the brain, corresponding folding changes might also be detectable in blood.
The study, published in Nature Aging on February 27, 2026, analyzed plasma from 520 participants across three groups: cognitively normal individuals, people with MCI and patients with clinically diagnosed Alzheimer’s disease. Using mass spectrometry to quantify site-specific solvent exposure and machine-learning models to identify disease-associated patterns, the authors searched for conformational biomarkers linked to AD status.
Among hundreds of candidate sites, three lysine-containing peptide sites on C1QA, clusterin and ApoB emerged as a robust diagnostic panel. These structural alterations yielded an overall three-way classification accuracy of about 83.4%. For binary comparisons, the model reached area under the ROC curve values of 0.9343 (healthy vs MCI) and 0.9325 (MCI vs AD), with healthy-versus-MCI accuracy exceeding 93% in some analyses.
Importantly, the three-marker structural signature validated across independent cohorts and in repeat samples collected months apart. The structural score tracked with cognitive assessments and showed a moderate correlation with MRI-measured brain atrophy, indicating potential utility for both early detection and longitudinal monitoring of disease progression or therapeutic response.
“Many neurodegenerative disorders reflect disrupted protein structure and clearance,” said senior author John R. Yates III. “Our results indicate that specific conformational changes in circulating proteins carry a detectable signal that parallels cognitive decline and brain atrophy.” Co-author Casimir Bamberger noted the striking correlation of three lysine sites across different proteins with disease state.
The approach complements existing amyloid and tau assays by targeting proteome-wide folding changes linked to underlying biology. If validated in larger, longer-term studies, structural profiling of plasma proteins could enable earlier diagnosis, better staging and dynamic measurement of treatment effects in clinical trials.
Limitations remain: the test requires broader validation across diverse populations and extended follow-up before clinical adoption. The research team is also exploring whether similar structural profiling can inform other conditions where proteostasis is disturbed, such as Parkinson’s disease and certain cancers.
Funding: This work was supported by the National Institutes of Health (grants RF1AG061846-01, 5R01AG075862, P30AG072973 and P30-AG066530).
Key Questions Answered:
A: Proteins must fold into precise three-dimensional structures to carry out their functions. When the cellular systems that fold and clear proteins—collectively called proteostasis—fail, proteins can misfold. Misfolded proteins may lose normal function and can become toxic. Detecting folding abnormalities in blood proteins can therefore reveal early, biologically meaningful changes linked to neurodegeneration.
A: Potentially. Because the test measures early breakdown in proteostasis, it may detect initial warning signs before classic amyloid plaque accumulation. In this study, the signature was especially accurate at identifying people with mild cognitive impairment, a condition that often precedes Alzheimer’s disease.
A: Drug reports about GLP-1 analogs focus on potential treatments for neurodegeneration. The Scripps study is focused on detection: a blood-based structural assay could be used to monitor whether candidate treatments, including GLP-1 drugs, help restore healthy protein conformations in patients.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The original journal paper was reviewed in full by staff editors.
- Additional context was added by the editorial team for clarity.
About this Alzheimer’s disease research news
Author: Press Office
Source: Scripps Research Institute
Contact: Press Office – Scripps Research Institute
Image credit: Neuroscience News
Original Research (open access):
“Structural signature of plasma proteins classifies the status of Alzheimer’s disease” by Ahrum Son, Hyunsoo Kim, Jolene K. Diedrich, Casimir Bamberger, Heather M. Wilkins, Jeffrey M. Burns, Jill K. Morris, Robert A. Rissman, Russell H. Swerdlow & John R. Yates III. Nature Aging
DOI: 10.1038/s43587-026-01078-2
Abstract
Structural signature of plasma proteins classifies the status of Alzheimer’s disease
Alzheimer’s disease involves dysregulation of proteostasis that can lead to protein misfolding, but whether these conformational changes are detectable as plasma biomarkers has been unclear. We profiled plasma protein structures from 520 participants, including clinically diagnosed AD patients, individuals with mild cognitive impairment and cognitively normal controls. Using mass spectrometry and machine learning, we characterized structural changes across the plasma proteome and identified AD-associated signatures while accounting for ApoE genotype and neuropsychiatric variables.
A diagnostic panel based on peptides from C1QA, clusterin and apolipoprotein B represented disease-associated structural alterations. This three-marker panel achieved 83.44% accuracy in three-way classification (healthy versus MCI versus AD). Binary classification produced area under the receiver operating characteristic curves of 0.9343 for healthy versus MCI and 0.9325 for MCI versus AD. Longitudinal samples were classified with 86.0% accuracy. These findings indicate that plasma protein structural profiling may improve early AD detection and support clinical trial evaluation of therapeutic interventions.