Researchers have identified a multi-protein blood test panel containing seven distinct markers that could enable better blood tests for Alzheimer’s disease. Utilizing machine learning on a novel immunoassay platform, the diagnostic method improves upon traditional p-tau217 tests by predicting advanced tau pathology in patients with elevated amyloid plaques.
Advancing Alzheimer’s Diagnostics Through Blood Proteomics
Diagnosing Alzheimer’s disease has historically relied on cognitive changes that manifest at a later stage of the condition, alongside specialized cerebrospinal fluid sampling or PET scans. Biomarker-based diagnostics were initially limited to specialized clinics capable of performing these procedures. While newer blood-based tests like p-tau217 have improved accessibility, their ability to determine disease stages and predict progression has remained limited compared to PET scans. Addressing these diagnostic gaps, researchers at the University of Gothenburg and their colleagues investigated whether machine learning applied to blood proteomics data—defined as large-scale information on all proteins in the blood—could identify new biomarker profiles to improve the assessment of disease stage.
The investigative team utilized two independent international cohorts containing data gathered from a novel immunoassay platform. This system measures more than 120 inflammation and neuronal markers within a single blood sample. By analyzing this broader panel of markers, the researchers identified a specific combination of proteins that provides a clearer picture of the disease and allows for more accurate future treatment choices.
The Seven-Protein Panel and Tau Pathology
The study centers on the identification of seven proteins in the blood, including p-tau217, that could improve the blood test’s ability to determine when a patient exhibits a disease profile corresponding to a late stage. In Alzheimer’s disease, amyloid plaques form in the brain and are followed by thread-like structures made of the protein tau, changes that both have toxic effects on the brain’s nerve cells years before dementia symptoms appear.
Results from the study, published in the journal JAMA Neurology under the title Plasma Biomarkers for Neocortical Tau Burden
(DOI: 10.1001/jamaneurol.2026.2650), demonstrate that including additional proteins alongside p-tau217 significantly improved the ability of p-tau217 to predict advanced tau pathology in individuals who already exhibit elevated levels of amyloid plaques. This multi-marker approach yields a more detailed assessment than single-protein assays alone.
Clinical Implications for Treatment and Trials
According to the research team, a multi-protein-based approach could serve as a viable alternative to staging via tau-PET in both clinical and research environments.
“We identified a panel of seven proteins, including p-tau217, that could improve the blood test’s ability to determine when a patient exhibits a disease profile corresponding to a late stage. This could yield better screening results, which is particularly valuable for guiding treatment choices and recruitment for clinical trials of future treatments for Alzheimer’s disease.” — Guglielmo Di Molfetta, doctoral student in neurochemistry, University of Gothenburg
Parallel Innovations in Site-Specific Protein Modification
While neurological researchers refine diagnostic blood panels, advances in molecular biology continue to shape how scientists handle complex proteins for basic research and biomedical applications. Site-selective functionalization strategies are essential for these tasks, yet achieving precise transformations across protein sequences remains a significant challenge that often requires protein engineering with orthogonal handles or sequence interventions to achieve selectivity. In related biochemical research, scientists reported a modular strategy for site-selective protein functionalization that enables the selective targeting of desired Cys residues through late-stage reversible thiol caging.
This strategy involves splitting the protein sequence into two segments and transiently masking native Cys residues to enable chemoselective ligation, thereby restoring the full-length protein with a single reactive Cys for late-stage functionalization. A subsequent decaging step regenerates the native Cys residues, yielding the desired site-specifically modified protein while preserving the native sequence. Researchers demonstrated the versatility of this strategy through multiple transformations, including Cys arylation, alkylation, and elimination, and demonstrated the power of the approach by selective modification of 30 examples in high resolution. Furthermore, the team applied this approach to edit Cys-rich zinc fingers, leading to the discovery of advanced analogs with enhanced DNA-binding activity, ultimately establishing a powerful platform for rapid access to structurally diverse proteins for fundamental research and therapeutic development.
