School of Chemistry, Manchester Institute of Biotechnology, University of Manchester, Manchester, UK.
Ian Potter NanoBioSensing Facility, NanoBiotechnology Research Laboratory, School of Science, RMIT University, Melbourne, Australia.
Analyst. 2016 Aug 15;141(17):5127-36. doi: 10.1039/c6an00883f.
Despite the fact that various microorganisms (e.g., bacteria, fungi, viruses, etc.) have been linked with infectious diseases, their crucial role towards sustaining life on Earth is undeniable. The huge biodiversity, combined with the wide range of biochemical capabilities of these organisms, have always been the driving force behind their large number of current, and, as of yet, undiscovered future applications. The presence of such diversity could be said to expedite the need for the development of rapid, accurate and sensitive techniques which allow for the detection, differentiation, identification and classification of such organisms. In this study, we employed Fourier transform infrared (FT-IR), Raman, and surface enhanced Raman scattering (SERS) spectroscopies, as molecular whole-organism fingerprinting techniques, combined with multivariate statistical analysis approaches for the classification of a range of industrial, environmental or clinically relevant bacteria (P. aeruginosa, P. putida, E. coli, E. faecium, S. lividans, B. subtilis, B. cereus) and yeast (S. cerevisiae). Principal components-discriminant function analysis (PC-DFA) scores plots of the spectral data collected from all three techniques allowed for the clear differentiation of all the samples down to sub-species level. The partial least squares-discriminant analysis (PLS-DA) models generated using the SERS spectral data displayed lower accuracy (74.9%) when compared to those obtained from conventional Raman (97.8%) and FT-IR (96.2%) analyses. In addition, whilst background fluorescence was detected in Raman spectra for S. cerevisiae, this fluorescence was quenched when applying SERS to the same species, and conversely SERS appeared to introduce strong fluorescence when analysing P. putida. It is also worth noting that FT-IR analysis provided spectral data of high quality and reproducibility for the whole sample set, suggesting its applicability to a wider range of samples, and perhaps the most suitable for the analysis of mixed cultures in future studies. Furthermore, our results suggest that while each of these spectroscopic approaches may favour different organisms (sample types), when combined, they would provide complementary and more in-depth knowledge (structural and/or metabolic state) of biological systems. To the best of our knowledge, this is the first time that such a comparative and combined spectroscopic study (using FT-IR, Raman and SERS) has been carried out on microbial samples.
尽管各种微生物（如细菌、真菌、病毒等）与传染病有关，但它们对地球上生命的维持所起的关键作用是不可否认的。这些生物的巨大生物多样性，加上它们广泛的生化能力，一直是它们目前大量应用的驱动力，而且到目前为止，尚未发现它们的未来应用。这种多样性的存在可以说是加速了对快速、准确和敏感技术的需求，这些技术可以用于检测、区分、识别和分类这些生物体。在这项研究中，我们采用傅里叶变换红外（FT-IR）、拉曼和表面增强拉曼散射（SERS）光谱学作为分子全生物体指纹图谱技术，结合多元统计分析方法对一系列工业、环境或临床相关细菌（铜绿假单胞菌、恶臭假单胞菌、大肠杆菌、屎肠球菌、变铅青链霉菌、枯草芽孢杆菌、蜡状芽孢杆菌）和酵母（酿酒酵母）进行分类。从所有三种技术收集的光谱数据的主成分判别分析（PC-DFA）得分图允许清晰地区分所有样本，甚至可以细分到亚种水平。使用 SERS 光谱数据生成的偏最小二乘判别分析（PLS-DA）模型的准确性较低（74.9%），与常规拉曼（97.8%）和傅里叶变换红外（96.2%）分析获得的模型相比。此外，虽然在拉曼光谱中检测到酿酒酵母的背景荧光，但当将 SERS 应用于同一物种时，这种荧光被淬灭，相反，当分析恶臭假单胞菌时，SERS 似乎引入了强烈的荧光。值得注意的是，傅里叶变换红外分析为整个样本集提供了高质量和可重复性的光谱数据，这表明它适用于更广泛的样本，也许最适合在未来的研究中分析混合培养物。此外，我们的结果表明，尽管这些光谱方法中的每一种都可能有利于不同的生物体（样本类型），但当结合使用时，它们将提供对生物系统的互补和更深入的了解（结构和/或代谢状态）。据我们所知，这是首次对微生物样本进行这种比较和综合光谱研究（使用 FT-IR、拉曼和 SERS）。
