Podium Presentation in Room 4 on Thursday at 9:00 (Chair: Chris Cox)
Authors: Sydney C. Povilaitis (1), Ashish Chakraborty (1), Lindsey M. Kirkpatrick (2), Sarmistha B. Hauger (3) and Livia S. Eberlin (1)
The rapid identification of infectious agents is critical for antimicrobial stewardship in the age of antimicrobial resistance. To address this, ambient ionization MS techniques have been applied for rapid identification of microbes directly from cultural isolates. We have developed a handheld, MS-based device, named the MasSpec Pen, that allows direct and rapid molecular analysis of biological samples. The MasSpec Pen deploys a droplet of solvent to the sample surface where biomolecules are extracted into the droplet and sent to the mass spectrometer for analysis. Here, we employ the MasSpec Pen for the identification of clinically relevant microbes directly from cultural isolates with no sample preparation.
Bacterial strains including Staphylococcus aureus, Group A and B Streptococcus, and Pseudomonas aeruginosa obtained from American Type Culture Collection (ATCC) were cultured on 5% sheep’s blood nutrient agar. Colonies were transferred from the agar plate to a glass slide, where they were analyzed directly with the MasSpec Pen coupled to a Q Exactive mass spectrometer (Thermo Scientific) in the negative ion mode. Each analysis was performed in under 30 seconds, including a 10 second extraction time. Three colonies from eight strains of each species investigated were analyzed and the corresponding data was used to build a statistical classifier. The least absolute shrinkage and selection operator (lasso) statistical method was used to identify molecular features and build a multi-level classification model for prediction of Gram type, genus, and species.
Various small molecules were detected including metabolites, fatty acids, and glycerophospholipid species. The mass spectral profiles exhibited qualitative differences including the higher relative abundance of m/z 159.077, attributed to the dipeptide alanine-alanine involved in peptidoglycan synthesis, in Streptococcus compared to Staphylococcus species. Five of the eight unique strains of each species investigated were used to build a multi-level lasso statistical classifier. The model selected a sparse set of molecular features including metabolites, dipeptides, fatty acids, and lipids that are predictive of Gram type, genus, and species. The performance of this model was evaluated by predicting the classification of the three remaining strains excluded from the training model and high accuracies were achieved for differentiation at all levels.
Here, the MasSpec Pen has been applied to rapidly discriminate clinically relevant bacteria in 30 seconds. A multi-level statistical classifier was developed to distinguish microbes based on Gram type, genus, and species with high accuracies based on a variety of molecular features. Next steps include MasSpec Pen analysis of infected clinical tissue samples. These results indicate that the MasSpec Pen could be a valuable tool for rapid identification of microbes in a clinical setting.
|Grants||yes||The Moore Foundation|
|Planning to mention or discuss specific products or technology of the company(ies) listed above:||