BackgroundOn-site monitoring of vanillylmandelic acid (VMA), homovanillic acid (HVA), and dopamine (DA) as key diagnostic biomarkers for a wide range of neurological disorders holds utmost significance in clinical settings. Numerous colorimetric sensors with mechanistic approaches based on aggregation or silver metallization have been introduced for this purpose. However, these mechanisms have drawbacks, such as sensitivity to environmental factors and probe toxicity. Therefore, there is a great demand for a robust yet non-toxic colorimetric sensor that employs a novel route to monitor these biomarkers effectively. ResultsHere, we present a single-component multi-colorimetric probe based on the controllable etching suppression of gold nanorods (AuNRs) upon exposure to the mild etchant N-bromosuccinimide (NBS), designed to accurately detect and discriminate VMA, HVA, DA, and their corresponding mixtures, i.e., VMA:HVA, VMA:DA, HVA:DA, and VMA:HVA:DA. To enhance the sensitivity and automation capabilities of the designed multi-colorimetric sensor, two machine learning techniques were employed: linear discriminant analysis (LDA) for the qualitative classification and partial least-squares regression (PLSR) for the quantitative analysis of pure biomarkers and their mixtures. The outcomes revealed a high correlation between measured and predicted values, covering a linear range of 0.8–25, 1.2–25, and 2.7–100 μmol L−1, with remarkably low detection limits of 0.260, 0.397, and 0.913 μmol L−1 for VMA, HVA, and DA, respectively. Lastly, the performance of the probe was validated by successfully detecting the neuroblastoma biomarker VMA:HVA in human urine. SignificanceOur designed multi-colorimetric probe introduces a rapid, cost-effective, user-friendly, non-toxic, and non-invasive approach to detecting and discriminating not only the pure biomarkers but also their corresponding binary and ternary mixtures. The distinctive response profiles produced by the probe in the presence of different mixture ratios can indicate various disease states in patients, which is highly crucial in clinical diagnostics.
Read full abstract