drFRET Achieves High-Sensitivity Detection of Glycosylated RNA
On April 10, 2025, a research team led by Zhangrun Xu from Northeastern University published a groundbreaking study in Nature Communications. They developed an innovative technology called dual-recognition fluorescence resonance energy transfer (drFRET), which enables highly sensitive and specific detection of glycosylated RNA on the surface of Small Extracellular Vesicles (sEVs) in extremely small blood samples (only 10 μL), opening up a new avenue for early and precise cancer diagnosis.
Glycosylated RNA is a recently discovered class of biomolecules composed of RNA and sugar chains linked by N-glycosidic bonds, anchored to the cell membrane surface. In 2021, a research team at Stanford University first confirmed the existence of these molecules, discovering that they participate in neutrophil recruitment in inflammatory responses through interactions with proteins such as P-selectin.
However, compared to traditional biomarkers such as proteins and lipids, the scientific community's understanding of glycosylated RNA remains very limited. This is primarily due to the lack of effective detection methods—existing methods often require complex Metabolic Labeling, cumbersome sample processing procedures, and cannot perform in-situ imaging. More importantly, the existence of glycosylated RNA on extracellular vesicles has never been confirmed before.
Extracellular vesicles are nanoscale membrane structures secreted by cells, approximately 30-150 nanometers in diameter, carrying important information about the mother cell's proteins, nucleic acids, and lipids, playing a crucial role in intercellular communication, immune regulation, and tumor metastasis. Because tumor-derived vesicles have high concentrations and good stability in the blood, they are considered ideal targets for liquid biopsies. If glycosylated RNA does indeed exist on the surface of these vesicles, they could become novel diagnostic biomarkers for diseases.
The drFRET technology developed by the research team cleverly solves this problem. This technology utilizes two specific probes working synergistically:
Fig. 1 Development of drFRET for in situ imaging sEV glycoRNAs. (Ren, et al 2025)
When both probes bind simultaneously to the same glycosylated RNA molecule, the distance between Cy3 and Cy5 shrinks to within 10 nanometers, resulting in fluorescence resonance energy transfer (FRET). Cy3 absorbs light energy and transfers it to Cy5 via a non-radiative process, causing it to emit fluorescence at a specific wavelength. This dual-recognition mechanism ensures that only molecules possessing both glycans and the specific RNA sequence can trigger the signal, effectively avoiding false positives.
To efficiently capture small extracellular vesicles from biological samples such as blood, the research team also designed functionalized polystyrene microspheres (fcPS). They modified the surface of the microspheres with HER2 protein-specific aptamers, enabling them to collect vesicles from complex samples like fishing. These microspheres are optically inert and do not interfere with fluorescence signal detection.
The research team first confirmed the presence of glycosylated RNA on vesicles through metabolic labeling experiments. They cultured HeLa cells using a glycan analog (Ac4ManNAz) containing an azide group, allowing the cells to integrate this label during glycan synthesis. Subsequently, they isolated secreted vesicles, extracted RNA, and attached biotin tags via Click chemistry reactions. Gel electrophoresis and dot hybridization results clearly showed that vesicle RNA was indeed glycosylated, primarily concentrated in small RNA regions.
More importantly, the researchers achieved in-situ imaging of glycosylated RNA using drFRET technology. When vesicles were labeled with two probes, a clear FRET signal was observed under a microscope, appearing as a Ring-like fluorescence along the vesicle outline. Treatment with DNase, RNase, or glycosidase significantly reduced the signal (by more than 88%), demonstrating that the signal specificity indeed originated from glycosylated RNA, rather than non-specific binding or free probes.
This technology exhibits excellent detection performance. The linear detection range spans five orders of magnitude (105 to 109 vesicles/mL), with a detection limit as low as 7.7 × 104 vesicles/mL, and the entire detection process takes only about 70 minutes.
The research team further explored the application value of this technology in cancer diagnosis. They selected five small nuclear RNAs (U1, U3, U35a, U8) and one cytoplasmic small RNA (Y5) as target glycosylated RNAs to detect vesicles secreted by seven different cancer cell lines. The results showed significant differences in the glycosylated RNA expression profiles of different cancer cell lines, and principal component analysis clearly distinguished samples from different cell origins.
In the subsequent clinical study, the team enrolled 100 subjects, including 88 cancer patients (covering breast cancer, pancreatic cancer, liver cancer, colorectal cancer, lung cancer, and cervical cancer) and 12 non-cancer controls. Excitingly, based on the unweighted sum of the expression levels (sEVSUM) of these five glycosylated RNAs, 100% sensitivity, specificity, and accuracy (95% confidence interval) were achieved in both the discovery (or training) and independent validation cohorts, with a perfect area under the receiver operating characteristic (AUC) of 1.0.
This technology also performed exceptionally well in fine-grained classification of cancer types. Using the principal coordinate analysis (PCoA) algorithm, six cancer types were automatically classified with an overall accuracy of 89% (95% confidence interval). This means that with only a 10-microliter blood sample, this technology can not only determine whether a subject has cancer but also infer the type of cancer with high accuracy, providing important reference for clinical decision-making.
In addition to its diagnostic value, the study also revealed the important role of glycosylated RNA in vesicle function. By replacing the glycan probes in the drFRET technique with fluorescently labeled Siglec proteins (a type of Sialic Acid-binding immunoglobulin-like lectin) and P-selectin, researchers found that glycosylated RNA on the vesicle surface can specifically bind to these proteins. Flow cytometry showed that approximately 99% of the vesicle-binding probes were recognized by Siglec-10, while this proportion decreased to 78% after RNase treatment, indicating that glycosylated RNA is one of the important ligands of Siglec-10.
More importantly, functional experiments revealed that when researchers treated vesicles with glycosidases or RNases to remove the surface glycans or RNA, the ability of vesicles to be taken up by endothelial cells, hepatic stellate cells, and colonic epithelial cells significantly decreased (by 67%-83%). This suggests that glycosylated RNA, through interaction with cell surface receptors, promotes vesicle adhesion and internalization to cells, and may play a crucial role in the establishment of the pre-metastatic microenvironment in tumor-derived vesicles.
The core advantages of drFRET technology lie in its high specificity resulting from its dual-recognition strategy and the convenience of detecting natural glycosylated RNA without metabolic labeling. Compared to traditional methods, it avoids complex sample pretreatment, preserves the spatial information of molecules, and is applicable to various biological sample types.
From a clinical application perspective, glycosylated RNA has unique advantages as a novel biomarker: glycosylation modification may enhance RNA stability in the circulatory system, improving detection sensitivity; the diversity of RNA sequences provides rich information dimensions; and nucleic acid hybridization-based detection probes are simple to design and easy to implement multiplex detection.
Of course, this technology still has room for optimization. For example, the current signal intensity is relatively weak and can be further enhanced through signal amplification strategies; vesicle enrichment based on a single capture aptamer may miss certain subpopulations; and current technologies cannot detect vesicle-specific glycosylated RNAs with unknown sequences. In the future, combining super-resolution imaging, nanoflow cytometry, and other techniques is expected to achieve precise quantification at the single vesicle level; developing universal RNA recognition reagents will expand the applicability of the technology.
This study not only developed a powerful tool for detecting Glycosylated RNA, but more importantly, it revealed the widespread presence of these mysterious molecules on extracellular vesicles and their crucial functions in cancer diagnosis and cell communication. With continuous technological advancements and further clinical validation, glycosylated RNA-based liquid biopsy holds promise as a new tool for early cancer screening and precise diagnosis, bringing the vision of detecting cancer through a single drop of blood closer to reality.
Reference
