Culture-free Antibiotic-susceptibility Determination From Single-bacterium Raman Spectra

Abstract

Raman spectrometry appears to be an opportunity to perform rapid tests in microbiological diagnostics as it provides phenotype-related information from single bacterial cells thus holding the promise of direct analysis of clinical specimens without any time-consuming growth phase. Here, we demonstrate the feasibility of a rapid antibiotic-susceptibility determination based on the use of Raman spectra acquired on single bacterial cells. After a two-hour preculture step, one susceptible and two resistant E. coli strains were incubated, for only two hours, in the presence of different bactericidal antibiotics (gentamicin, ciprofloxacin, amoxicillin) in a range of concentrations that included the clinical breakpoints used as references in microbial diagnostic. Spectra were acquired and processed to isolate spectral modifications associated with the antibiotic effect. We evidenced an “antibiotic effect signature” which is expressed with specific Raman peaks and the coexistence of three spectral populations in the presence of antibiotic. We devised an algorithm and a test procedure that overcome single-cell heterogeneities to estimate the MIC and determinate the susceptibility phenotype of the tested bacteria using only a few single-cell spectra in four hours only if including the preculture step.

Rapid identification of pathogenic bacteria using Raman spectroscopy and deep learning

“…High signal-to-noise ratios (SNRs) are thus needed to reach high identification accuracies 9 , typically requiring long measurement times that prohibit high-throughput single-cell techniques. Additionally, the large number of clinically relevant species, strains, and antibiotic resistance patterns require comprehensive datasets that are not gathered in studies that focus on differentiating between species 10,11 , isolates (typically referred to as strains in the literature) 12,13 , or antibiotic susceptibilities [14][15][16][17][18][19] . In this work, we address this challenge by training a convolutional neural network (CNN) to classify noisy bacterial spectra by isolate, empiric treatment, and antibiotic resistance.…”

Metabolic machinery encrypted in the Raman spectrum of influenza A virus‐inoculated mammalian cells

“…Its recent spectacular awakening has brought it at the forefront of modern biochemical science (Baena & Lendl, ). In reviewing recent publications on biomolecules, it becomes clear how Raman spectroscopy has proved invaluable in unveiling a variety of issues in a number of key fields of microbiology, including metabolism of cells (M. Li, Romero‐Gonzales, Banwart, & Huang, ; Notingher, ; Notingher, Verrier, Haque, Polak, & Hench, ; Smith, Wright, & Ashton, ), bacteria (Gonchukov, Sukhinina, Bakhmutov, & Minaeva, ; Lorenz, Wichmann, Stockel, Rosch, & Popp, ; Novelli‐Rousseau et al, ; Strola et al, ), and viruses (Otange, Birech, Okonda, & Rop, ; Tuma & Thomas, ). The so far reported research also holds a promise for more accurate, fully quantitative, and rapid microbiological identifications of crucial issues in clinical diagnostics.…”

Raman‐activated sorting of antibiotic‐resistant bacteria in human gut microbiota

“…While some studies demonstrated that the fingerprint region of Raman spectra (200-1800 cm −1 ) were able to analyse the bacterial phenotype under antibiotic influences, and determine the MIC and resistant mechanisms of E. coli in different antibiotics (Athamneh et al, 2014;Teng et al, 2016;Tao et al, 2017;Xu et al, 2017;Germond et al, 2018;Kirchhoff et al, 2018;Novelli-Rousseau et al, 2018), the C-D band from SCRS is a simple and distinguishable Raman biomarker to identify the MA-ARB in environmental samples (Xu et al, 2017). In this study, we demonstrated that the C-D band can be used to identify the MA-ARB in human gut microbiota.…”