Researchers have developed a blood test that detects Alzheimer’s biomarkers at single-molecule levels, enabling earlier, non-invasive diagnosis using extracellular vesicles.
(Nanowerk Spotlight) A definitive diagnosis of Alzheimer’s disease still relies on peering inside the brain—either through costly neuroimaging or through the extraction of cerebrospinal fluid via lumbar puncture. Yet the pathology begins years earlier, at a microscopic scale, in the form of misfolded proteins and degenerating neurons. These changes occur long before symptoms surface, but they leave behind molecular traces. The challenge is how to detect those early signals without invasive procedures. What if those biomarkers could be identified not from deep within the brain, but from a routine blood sample?
This idea has driven sustained efforts to find accessible and reliable blood-based markers for Alzheimer’s disease. The condition is defined by the buildup of amyloid-beta plaques and tau protein tangles, and researchers have focused on detecting fragments of these proteins in blood plasma. Several methods—including immunoassays, Raman spectroscopy, and transistor-based sensors—have been developed to detect them. But most technologies fall short on three fronts: they lack the sensitivity to detect low-abundance proteins, they struggle to analyze multiple markers at once, and they are often too slow or expensive for practical use.
A complicating factor is the complexity of blood itself. Biomarkers in circulation are vastly diluted and surrounded by a multitude of other proteins. This makes detection technically challenging, especially for diseases like Alzheimer’s, where early intervention depends on subtle changes. Extracellular vesicles (EVs)—tiny particles released by cells—offer a potential solution. These vesicles carry proteins, RNA, and other materials from their cells of origin and can cross the blood–brain barrier, carrying molecular information about the brain into the bloodstream. As such, they could serve as a minimally invasive window into the progression of Alzheimer’s.
Still, existing EV-based diagnostics have typically relied on neuron-derived vesicles, which are difficult to isolate and present in low quantities. The more abundant blood-derived vesicles secreted by multiple types of brain cells may provide a better alternative, if their contents can be analyzed precisely and rapidly. That is the gap this new study aims to address.
Researchers from the Guangdong Technion-Israel Institute of Technology and collaborating institutions have developed a bioelectronic sensing platform capable of detecting Alzheimer’s-related biomarkers in blood EVs with single-molecule sensitivity. Their method integrates two core technologies: an organic electrochemical transistor (OECT) and a microelectrode array. The combination allows for precise, rapid, and multiplexed detection of several biomarkers associated with Alzheimer’s disease, offering the potential for routine blood-based screening.
The central innovation lies in the sensor’s ability to detect four key proteins linked to Alzheimer’s: amyloid-beta 1-40, amyloid-beta 1-42, total tau, and a phosphorylated form of tau known as p-tau181. These proteins are embedded in or associated with the surface of extracellular vesicles that circulate in blood. To enrich these vesicles, the researchers used magnetic beads coated with antibodies that bind to CD63, a protein commonly found on the surface of EVs. Once captured, the vesicles are exposed to the bioelectronic sensor, which has been functionalized with antibodies against the four Alzheimer’s-related proteins.
Schematic illustration of bioelectronic label-free platform for detecting core Alzheimer’s disease (AD) biomarkers in blood extracellular vesicles (EVs). a) Clinical diagnosis of AD using blood EVs-derived biomarkers. b) Overview of the sensing process: (i) Immunofunctionalized magnetic beads for the enrichment of blood EVs by targeting external proteins on the surface; (ii) Antibody-modified microelectrode arrays for specific detection of multiple AD biomarkers; (iii) OECT for signal transduction and ultrasensitive detection. c) Workflow timeline from sample collection to result analysis. The hybridized SAM designed on the gate surface enables high-density antibody immobilization and enhances resistance to biofouling. The combination of electrode arrays and OECT technology enables ultrasensitive multi-target detection, robust performance in complex environments, and delivers results within 20 min, achieving near-perfect accuracy in differentiating AD patients from healthy individuals. (Image: Reprinted with permission by Wiley VCH Verlag) (click on image to enlarge)
The OECT works by translating the presence of a target biomolecule into an electrical signal. When a protein binds to its corresponding antibody on the sensor’s surface, it disrupts the flow of ions in the transistor’s organic polymer layer. This change alters the electrical current, which is measured and interpreted as a positive signal. The design is highly sensitive: the device can detect concentrations in the zeptomolar range, which corresponds to a few molecules in a sample. This level of precision is essential for early detection, when biomarker levels are still very low.
While the sensor is sensitive enough to detect individual molecules, this does not mean that the presence of a single molecule is enough to diagnose Alzheimer’s. Clinical diagnosis is not based on detecting whether a molecule exists at all—it’s based on whether the concentration of specific biomarkers rises above levels typically seen in healthy individuals. In this context, single-molecule sensitivity serves a different purpose: it allows the sensor to detect these markers much earlier, when their levels in the blood are still very low.
In practice, diagnosis relies on identifying patterns in the levels of several biomarkers and comparing them to statistical thresholds established from clinical data. For example, elevated concentrations of amyloid-beta 1-42 and phosphorylated tau181 in blood-derived extracellular vesicles were consistently observed in patients with Alzheimer’s, but not in healthy individuals. The researchers used these concentration patterns—not the presence of a single molecule—to distinguish between the two groups. So while the sensor’s sensitivity enables it to detect even a few dozen molecules in a sample, its role in diagnosis is to quantify how much of each marker is present and whether that level fits a pattern associated with disease.
To ensure specificity, the sensor uses a hybrid self-assembled monolayer on its electrodes to immobilize the antibodies in a stable and densely packed configuration. This surface modification helps prevent nonspecific binding and improves the signal-to-noise ratio. The researchers validated the sensor’s performance across all four biomarkers, demonstrating strong selectivity and stability even in complex biological fluids that simulate the composition of human blood.
Importantly, the system delivers results in under 20 minutes using just 15 microliters of blood plasma. The rapid turnaround and small sample volume make the technology compatible with clinical workflows. In tests involving 40 clinical samples, the sensor correctly classified all Alzheimer’s patients and healthy individuals when using a combined panel of the four biomarkers. By contrast, single-biomarker analysis was less accurate, highlighting the value of a multi-target approach.
Further analysis revealed that the sensor’s readings closely matched those obtained through traditional ELISA for Alzheimer’s patients. However, in healthy individuals—where biomarker levels are much lower—the ELISA readings diverged significantly from those of the new sensor, suggesting that the bioelectronic platform is better suited for detecting subtle early changes.
The study also showed that the combination of amyloid-beta 1-42 with either total tau or phosphorylated tau improved diagnostic accuracy over any single marker. A model using all four biomarkers achieved 100 percent classification accuracy, correctly identifying every sample in the test cohort. This indicates that integrating multiple signals that reflect different aspects of Alzheimer’s pathology—amyloid buildup and tau-related neuronal damage—provides a more reliable diagnostic basis.
The researchers emphasize that while the sensor was developed for Alzheimer’s detection, the platform could be adapted for other diseases. Because it targets molecular signals carried by EVs, it could in principle detect a wide range of biomarkers—such as nucleic acids, lipids, and metabolites—associated with other neurodegenerative conditions, cancers, or infectious diseases.
The new sensor marks a significant step toward practical, blood-based diagnosis of Alzheimer’s. It bypasses the invasiveness of cerebrospinal fluid sampling and the complexity of neuroimaging. Its ability to deliver highly sensitive, multi-target readings in a compact and inexpensive format positions it as a candidate for broader deployment in clinics. If validated in larger cohorts and standardized for routine use, this technology could enable earlier and more accurate identification of Alzheimer’s, supporting timely interventions and helping researchers monitor disease progression more effectively.
The findings also contribute to a growing body of work that sees EVs as key players in non-invasive diagnostics. Their ability to transport disease-related molecules from the brain into the bloodstream makes them particularly valuable for tracking neurological conditions. With an infrastructure that combines molecular biology, materials science, and electrical engineering, this diagnostic platform demonstrates how interdisciplinary approaches can address longstanding problems in medical detection.
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
