Raman-Activated, Interactive Sorting of Isotope-Labeled Bacteria.

Due to its high spatial resolution, Raman microspectroscopy allows for the analysis of single microbial cells. Since Raman spectroscopy analyzes the whole cell content, this method is phenotypic and can therefore be used to evaluate cellular changes. In particular, labeling with stable isotopes (SIPs) enables the versatile use and observation of different metabolic states in microbes. Nevertheless, static measurements can only analyze the present situation and do not allow for further downstream evaluations. Therefore, a combination of Raman analysis and cell sorting is necessary to provide the possibility for further research on selected bacteria in a sample. Here, a new microfluidic approach for Raman-activated continuous-flow sorting of bacteria using an optical setup for image-based particle sorting with synchronous acquisition and analysis of Raman spectra for making the sorting decision is demonstrated, showing that active cells can be successfully sorted by means of this microfluidic chip.

Tracking active groundwater microbes with D2 O labelling to understand their ecosystem function.

Microbial activity is key in understanding the contribution of microbial communities to ecosystem functions. Metabolic labelling with heavy water (D2 O) leads to the formation of carbon-deuterium bonds in active microorganisms. We illustrated how D2 O labelling allows monitoring of metabolic activity combined with a functional characterization of active populations in complex microbial communities. First, we demonstrated by single cell Raman microspectroscopy that all measured bacterial cells from groundwater isolates growing in complex medium with D2 O were labelled. Next, we conducted a labelling approach with the total groundwater microbiome in D2 O amended microcosms. Deuterium was incorporated in most measured cells, indicating metabolic activity in the oligotrophic groundwater. Moreover, we spiked the groundwater microbiome with organic model compounds. We discovered that heterotrophs assimilating veratric acid, a lignin derivative, showed higher labelling than heterotrophs assimilating methylamine, a degradation product of biomass. This difference can be explained by dilution of the deuterium through hydrogen from the organic compounds. Metaproteomics identified Sphingomonadaceae and Microbacteriaceae as key players in veratric acid degradation, and the metabolic pathways employed. Methylamine, in contrast, stimulated various proteobacterial genera. We propose this combined approach of Raman microspectroscopy and metaproteomics for elucidating the complex metabolic response of microbial populations to different stimuli.

Sample-Size Planning for Multivariate Data: A Raman-Spectroscopy-Based Example.

The goal of sample-size planning (SSP) is to determine the number of measurements needed for statistical analysis. This SSP is necessary to achieve robust and significant results with a minimal number of measurements that need to be collected. SSP is a common procedure for univariate measurements, whereas for multivariate measurements, like spectra or time traces, no general sample-size-planning method exists. Sample-size planning becomes more important for biospectroscopic data because the data generation is time-consuming and costly. Additionally, ethical reasons do not allow the use of unnecessary samples and the measurement of unnecessary spectra. In this paper, a general sample-size-planning algorithm is presented that is based on learning curves. The learning curve quantifies the improvement of a classifier for an increasing training-set size. These curves are fitted by the inverse-power law, and the parameters of this fit can be utilized to predict the necessary training-set size. Sample-size planning is demonstrated for a biospectroscopic task of differentiating six different bacterial species, including Escherichia coli, Klebsiella terrigena, Pseudomonas stutzeri, Listeria innocua, Staphylococcus warneri, and Staphylococcus cohnii, on the basis of their Raman spectra. Thereby, we estimate the required number of Raman spectra and biological replicates to train a classification model, which consists of principal-component analysis (PCA) combined with linear-discriminant analysis (LDA). The presented algorithm revealed that 142 Raman spectra per species and seven biological replicates are needed for the above-mentioned biospectroscopic-classification task. Even though it was not demonstrated, the learning-curve-based sample-size-planning algorithm can be applied to any multivariate data and in particular to biospectroscopic-classification tasks.