Bennett Danielle, Chen Xueqian, Walker Gregory J, Stelzer-Braid Sacha, Rawlinson William D, Hibbert D Brynn, Tilley Richard D, Gooding J Justin
School of Chemistry, The University of New South Wales, Sydney, New South Wales 2052, Australia.
Australian Centre for Nanomedicine, The University of New South Wales, Sydney, New South Wales 2052, Australia.
Anal Chem. 2023 Apr 25;95(16):6550-6558. doi: 10.1021/acs.analchem.2c05292. Epub 2023 Apr 10.
Plasmonic nanoparticles are finding applications within the single molecule sensing field in a "dimer" format, where interaction of the target with hairpin DNA causes a decrease in the interparticle distance, leading to a localized surface plasmon resonance shift. While this shift may be detected using spectroscopy, achieving statistical relevance requires the measurement of thousands of nanoparticle dimers and the timescales required for spectroscopic analysis are incompatible with point-of-care devices. However, using dark-field imaging of the dimer structures, simultaneous digital analysis of the plasmonic resonance shift after target interaction of thousands of dimer structures may be achieved in minutes. The main challenge of this digital analysis on the single-molecule scale was the occurrence of false signals caused by non-specifically bound clusters of nanoparticles. This effect may be reduced by digitally separating dimers from other nanoconjugate types. Variation in image intensity was observed to have a discernible impact on the color analysis of the nanoconjugate constructs and thus the accuracy of the digital separation. Color spaces wherein intensity may be uncoupled from the color information (hue, saturation, and value (HSV) and luminance, a* vector, and b* vector (LAB) were contrasted to a color space which cannot uncouple intensity (RGB) to train a classifier algorithm. Each classifier algorithm was validated to determine which color space produced the most accurate digital separation of the nanoconjugate types. The LAB-based learning classifier demonstrated the highest accuracy for digitally separating nanoparticles. Using this classifier, nanoparticle conjugates were monitored for their plasmonic color shift after interaction with a synthetic RNA target, resulting in a platform with a highly accurate yes/no response with a true positive rate of 88% and a true negative rate of 100%. The sensor response of tested single stranded RNA (ssRNA) samples was well above control responses for target concentrations in the range of 10 aM-1 pM.
等离子体纳米颗粒正以“二聚体”形式在单分子传感领域得到应用,其中靶标与发夹DNA的相互作用会导致颗粒间距离减小,从而引起局部表面等离子体共振位移。虽然可以使用光谱法检测这种位移,但要实现统计相关性则需要测量数千个纳米颗粒二聚体,而且光谱分析所需的时间尺度与即时检测设备不兼容。然而,通过对二聚体结构进行暗场成像,可以在几分钟内对数千个二聚体结构与靶标相互作用后的等离子体共振位移进行同步数字分析。在单分子尺度上进行这种数字分析的主要挑战是由非特异性结合的纳米颗粒簇引起的假信号的出现。通过将二聚体与其他纳米共轭物类型进行数字分离,可以减少这种影响。观察到图像强度的变化对纳米共轭物构建体的颜色分析有明显影响,进而影响数字分离的准确性。将强度可以与颜色信息解耦的颜色空间(色相、饱和度和明度(HSV)以及亮度、a向量和b向量(LAB))与不能解耦强度的颜色空间(RGB)进行对比,以训练分类算法。对每个分类算法进行验证,以确定哪种颜色空间能产生最准确的纳米共轭物类型数字分离。基于LAB的学习分类器在数字分离纳米颗粒方面表现出最高的准确性。使用该分类器,监测纳米颗粒共轭物与合成RNA靶标相互作用后的等离子体颜色变化,从而得到一个具有高度准确的是/否响应的平台,真阳性率为88%,真阴性率为100%。对于浓度范围在10 aM至1 pM的靶标,测试的单链RNA(ssRNA)样品的传感器响应远高于对照响应。