Intra- and Interspecies Variability of Single-Cell Innate Fluorescence Signature of Microbial Cell

Abstract

Here we analyzed the innate fluorescence signature of the single microbial cell, within both clonal and mixed populations of microorganisms. We found that even very similarly shaped cells differ noticeably in their autofluorescence features and that the innate fluorescence signatures change dynamically with growth phases. We demonstrated that machine learning models can be trained with a data set of single-cell innate fluorescence signatures to annotate cells according to their phenotypes and physiological status, for example, distinguishing a wild-type Aspergillus nidulans cell from its nitrogen metabolism mutant counterpart and log-phase cells from stationary-phase cells of Pseudomonas putida. We developed a minimally invasive method (confocal reflection microscopy-assisted single-cell innate fluorescence [CRIF] analysis) to optically extract and catalog the innate cellular fluorescence signatures of each of the individual live microbial cells in a three-dimensional space. This technique represents a step forward from traditional techniques which analyze the innate fluorescence signatures at the population level and necessitate a clonal culture. Since the fluorescence signature is an innate property of a cell, our technique allows the prediction of the types or physiological status of intact and tag-free single cells, within a cell population distributed in a three-dimensional space. Our study presents a blueprint for a streamlined cell analysis where one can directly assess the potential phenotype of each single cell in a heterogenous population by its autofluorescence signature under a microscope, without cell tagging. IMPORTANCE A cell’s innate fluorescence signature is an assemblage of fluorescence signals emitted by diverse biomolecules within a cell. It is known that the innate fluoresce signature reflects various cellular properties and physiological statuses; thus, they can serve as a rich source of information in cell characterization as well as cell identification. However, conventional techniques focus on the analysis of the innate fluorescence signatures at the population level but not at the single-cell level and thus necessitate a clonal culture. In the present study, we developed a technique to analyze the innate fluorescence signature of a single microbial cell. Using this novel method, we found that even very similarly shaped cells differ noticeably in their autofluorescence features, and the innate fluorescence signature changes dynamically with growth phases. We also demonstrated that the different cell types can be classified accurately within a mixed population under a microscope at the resolution of a single cell, depending solely on the innate fluorescence signature information. We suggest that single-cell autofluoresce signature analysis is a promising tool to directly assess the taxonomic or physiological heterogeneity within a microbial population, without cell tagging.

Responses of Salmonella biofilms to oxidizing biocides: Evidence of spatial clustering

“…Support vector machines (SVM) are learning algorithms which aim to identify a function (hyperplane) that can separate datasets. Biomedical applications of SVM include the classification of bacterial species to distinguish between disease conditions (Yoram et al, 2018 ) or the innate fluorescence signatures of microbial cells (Yawata et al, 2019 ), prediction of biofilm‐inhibiting‐peptides (Gupta et al, 2016 ), classification of antibiotics (Jung et al, 2014 ) or differentiation of human and in vitro biofilm transcriptomes (Cornforth et al, 2018 ). Here we applied an SVM classifier with radial basis function (RBF) kernel to differentiate between the weak (SL1344) and strong (3750) biofilm forming strains using the mean number of particles from a single 2D plane of Z‐stack for each condition.…”

Host-encoded DNA methyltransferases modify the epigenome and host tropism of invading phages

“…To counter the epigenomic synchronization of phages and to avoid the extinction of the entire host population by phage prevalence, some prokaryotes change their sets of phage-defense systems heterogeneously. For example, many species of gut bacteria, such as Bacteroides fragilis and H. pylori, showed strain-level variations in gene sets involved in phage defense systems including MTases (65)(66)(67). However, because gene acquisition and deletion relies on genomic events, such as horizontal gene transfer and recombination, this occurs at low frequently.…”

Machine learning-assisted single-cell Raman fingerprinting for in situ and nondestructive classification of prokaryotes

“…In principle, all of them were used without selection. However, about 20% of S. acidocaldarius cells showed strong autofluorescence, and their spectra were rejected from the dataset because the objective of the present work was to classify cells based on Raman spectral patterns and not autofluorescence patterns ( Yawata et al., 2019 ). (1) The PBS spectrum (average of 10 spectra) was subtracted from each single-cell spectrum in which cosmic rays, if any, were manually removed in advance.…”