Background Small extracellular vesicles (sEVs) have emerged as pivotal mediators in liquid biopsy, carrying molecular cargo that reflects the pathological status of their origin cells; however, their clinical translation is often hindered by the gap between computational discovery and experimental validation. Current sEVs isolation methods frequently lack the necessary purity and throughput required for reliable biomarker verification. Here, we implement a sequential workflow that integrates in silico biomarker identification with our high-performance microfluidic capture system to streamline the diagnostic pipeline for sEVs. Methods Our workflow comprises two core modules: (1) Data-Driven Discovery: A machine learning-assisted meta-analysis was performed on multi-center transcriptomic datasets to establish a sEV-Related Signature (SRS), which identified Lipocalin-2 (LCN2) as a robust candidate for pancreatic cancer diagnosis. (2) Hardware-Enabled Validation: A centrifugal microfluidic platform was engineered for the standardized, high-purity recovery of sEVs. Results The workflow demonstrated high fidelity in translating bioinformatic predictions into experimental results. Targeted validation via Western blot and label-free proteomic profiling confirmed that LCN2 was significantly enriched in the isolated sEVs, echoing the initial machine learning-based predictions. Gene Ontology (GO) analysis further validated that the captured proteome was intrinsically linked to vesicle-mediated transport and tumor-specific signaling. Conclusion By coupling machine learning mining with microfluidic isolation, this stepwise workflow provides a robust, standardized framework for the discovery and rapid validation of sEV-derived biomarkers. This approach offers a scalable solution for cancer diagnostics and supports the development of precision diagnostic tools.
The widespread use of pesticides demands sensitive, rapid, and cost-effective analytical methods for environmental monitoring. In this work, the electrochemical behavior of herbicide fomesafen (FMS) was systematically investigated at a boron-doped diamond electrode (BDD). FMS exhibited distinctive electrochemical behavior at the cathodically pretreated BDD surface (CPT-BDD). Cyclic voltammetric studies revealed a previously unreported irreversible, diffusion-controlled anodic oxidation process at approximately 0.38 V that strongly depends on a preceding cathodic reduction step. Exploiting this pathway, a novel square-wave anodic stripping voltammetric (SWASV) method was developed. Experimental parameters were systematically optimized, demonstrating that cathodic pretreatment significantly enhances the analytical response, while the addition of cetyltrimethylammonium bromide (CTAB) substantially improves signal stability and repeatability. Under optimized conditions, the proposed method enabled sensitive quantification of FMS under ambient conditions, eliminating the deoxygenation step required by the previously reported cathodic electrochemical method. The method exhibited two linear concentration ranges, from 22.9 to 123 nmol L−1 and from 123 to 909 nmol L−1 (R2 > 0.995), achieving a detection limit of 2.49 nmol L−1 and satisfactory precision (RSD < 8.1%). In comparison with reported approaches, this methodology introduces a previously unexplored anodic pathway for FMS determination. These results provide new insights into FMS electrochemistry at cathodically pretreated boron-doped diamond electrode (CPT-BDD) and demonstrate that SWASV provides high analytical performance, combining high analytical sensitivity with a low LOD for ultra-trace herbicide monitoring.
Mycotoxin contamination in traditional Chinese medicines (TCMs) has attracted increasing attention due to its potential impact on drug safety. Developing highly specific and sensitive detection strategies is essential for quality control of TCMs. This review specifically centers on the complex TCM matrices and provides a comprehensive summary of recent progress in mycotoxin detection within TCMs. First, the development and characteristics of different sensing platforms (liquid chromatography, enzyme-linked immunosorbent assay, microarray chips, suspension arrays, test strips, etc.) are summarized. Subsequently, the progress in recognition elements (nanobodies, molecularly imprinted polymers, aptamers, etc.), signal output modes (optical, electrochemical, and multimodal integration, etc.), and signal amplification technologies (nanomaterial-assisted signal amplification, etc.) are introduced. In addition, the effects of different TCM matrices on various detection methods and matrix‑adaptive sample pretreatment strategies are surveyed. The critical role of nanomaterials in sample pretreatment for complex matrices is highlighted, with a particular focus on the integrated functions of magnetic nanoprobes for enrichment, cleanup, and detection. Finally, future development directions are prospected: the deep integration of artificial intelligence will promote the construction of intelligent sensing platforms; detection targets will be expanded to masked mycotoxins, animal-derived TCMs and processed TCM products; and the integration of detection and detoxification strategies deserves in-depth exploration. The realization of all these goals relies on continuous exploration of novel sensors and nanomaterials, as well as interdisciplinary collaboration. This review aims to offer a theoretical guide for the future advancement of accurate, high-throughput, multiplexed, on-site, rapid, multifunctional, intelligent techniques for mycotoxin detection in TCMs.
In this study, a novel, pH-responsive nanocomposite hydrogel framework was successfully synthesized by grafting poly(N,N-dimethylacrylamide) onto a natural gum acacia backbone in the presence of CuCoFe2O4 nanoparticles for the controlled delivery of ciprofloxacin hydrochloride (HCl). Structural and thermal characterizations via FT-IR, PXRD, FE-SEM, and TGA confirmed successful crosslinking, uniform nanoparticle integration. The hydrogel swelling behavior and in vitro drug release kinetics were systematically investigated across distinct physiological environments (pH 4.0, 7.4, and 9.2), demonstrating optimal swelling capacity and cumulative drug release under alkaline conditions. To elucidate the transport mechanism of Ciprofloxacin HCl from the hydrogel matrix, the experimental data were fitted to multiple mathematical kinetic models, including Zero-order, First-order, Korsmeyer-Peppas, Higuchi square-root, and Hixson-Crowell cube-root equations. The release profiles showed an excellent correlation with the Korsmeyer-Peppas model (R2 > 0.99) at pH 9.2. Furthermore, the calculated diffusional exponent (n) values across all tested environments consistently ranged between 0.45 and 0.89, revealing that drug transport follows a non-Fickian (anomalous) diffusion mechanism governed by a combination of polymer chain relaxation and diffusion. These findings demonstrate that the developed gum acacia-based nanocomposite hydrogel holds strong potential as an advanced platform for pH-responsive, controlled drug delivery applications.
