Department of Electrical and Computer Engineering, Virginia Tech, Blacksburg, Virginia 24061, United States.
Department of Electronic Engineering, Pukyong National University, Busan 48513, Republic of Korea.
ACS Sens. 2023 Mar 24;8(3):1132-1142. doi: 10.1021/acssensors.2c02412. Epub 2023 Mar 9.
In situ spatiotemporal biochemical characterization of the activity of living multicellular biofilms under external stimuli remains a significant challenge. Surface-enhanced Raman spectroscopy (SERS), combining the molecular fingerprint specificity of vibrational spectroscopy with the hotspot sensitivity of plasmonic nanostructures, has emerged as a promising noninvasive bioanalysis technique for living systems. However, most SERS devices do not allow reliable long-term spatiotemporal SERS measurements of multicellular systems because of challenges in producing spatially uniform and mechanically stable SERS hotspot arrays to interface with large cellular networks. Furthermore, very few studies have been conducted for multivariable analysis of spatiotemporal SERS datasets to extract spatially and temporally correlated biological information from multicellular systems. Here, we demonstrate in situ label-free spatiotemporal SERS measurements and multivariate analysis of biofilms during development and upon infection by bacteriophage virus Phi6 by employing nanolaminate plasmonic crystal SERS devices to interface mechanically stable, uniform, and spatially dense hotspot arrays with the biofilms. We exploited unsupervised multivariate machine learning methods, including principal component analysis (PCA) and hierarchical cluster analysis (HCA), to resolve the spatiotemporal evolution and Phi6 dose-dependent changes of major Raman peaks originating from biochemical components in biofilms, including cellular components, extracellular polymeric substances (EPS), metabolite molecules, and cell lysate-enriched extracellular media. We then employed supervised multivariate analysis using linear discriminant analysis (LDA) for the multiclass classification of Phi6 dose-dependent biofilm responses, demonstrating the potential for viral infection diagnosis. We envision extending the in situ spatiotemporal SERS method to monitor dynamic, heterogeneous interactions between viruses and bacterial networks for applications such as phage-based anti-biofilm therapy development and continuous pathogenic virus detection.
在外部刺激下，对活的多细胞生物膜活性进行原位时空生化特性分析仍然是一个重大挑战。表面增强拉曼光谱（SERS）结合了振动光谱的分子指纹特异性和等离子体纳米结构的热点敏感性，已成为一种有前途的用于活体系统的非侵入性生物分析技术。然而，由于在产生与大细胞网络接口的空间均匀且机械稳定的 SERS 热点阵列方面存在挑战，大多数 SERS 设备无法进行可靠的多细胞系统的长期时空 SERS 测量。此外，很少有研究用于对时空 SERS 数据集进行多变量分析，以从多细胞系统中提取空间和时间相关的生物信息。在这里，我们通过采用纳米层压等离子体晶体 SERS 器件，展示了在生物膜发育过程中和受噬菌体 Phi6 感染时的无标记原位时空 SERS 测量和多变量分析。我们利用无监督多元机器学习方法，包括主成分分析（PCA）和层次聚类分析（HCA），来解析生物膜中主要拉曼峰的时空演化和 Phi6 剂量依赖性变化，这些拉曼峰源于生物膜中的生化成分，包括细胞成分、细胞外聚合物物质（EPS）、代谢物分子和富含细胞裂解物的细胞外介质。然后，我们使用线性判别分析（LDA）进行监督多元分析，用于 Phi6 剂量依赖性生物膜反应的多类分类，证明了用于病毒感染诊断的潜力。我们设想将原位时空 SERS 方法扩展到用于监测病毒和细菌网络之间动态、异质相互作用的应用，例如基于噬菌体的抗生物膜治疗开发和连续的致病性病毒检测。
