MSACL 2024 Abstract

INTRODUCTION: The prominence of antibiotic resistant bacterial strains has raised concern for the efficacy of current available drug therapies. Point-of-care utilization of current drug therapies require strain specific identifications of pathogens, which often involve tedious sample preparation strategies and tailored analytical methodologies. The emergence of multi-omic approaches has empowered scientists to answer complex systems biology questions regarding antibiotic resistance. The genomic and proteomic information of antibiotic-resistant ESKAPE pathogens, including Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumanii, Pseudomonas aeruginosa, and Enterobacter spp. has been thoroughly explored. However, there remains a significant lack of knowledge of the downstream metabolomic and lipidomic profiles of these small organisms. Here, we have optimized a high throughput multi-omic approach that can isolate the constituent parts of the whole biological system that are not otherwise explained by the template-driven aspects of genomics and proteomics. We showcase the potential for a streamlined method for managing multi-omic experiments of diverse antibiotic-resistant microbe populations using an optimized workflow.

OBJECTIVES: Identify pathogens with strain-specificity using an updated multi-omic workflow.

METHODS: A range of ~ 96 ESKAPE pathogens with varying antibiotic susceptibilities were cultured overnight in tryptic soy broth. Internal standards of metabolites and lipids were added prior to extraction. Biphasic Bligh and Dyer and three monophasic solvent compositions of BuOH (0-60%)/ACN (20-80%)/H2O (20% constant) were evaluated for their ability to extract metabolites and lipids simultaneously from bacteria. Extracts were reconstituted in 2:2:1 ACN/MeOH/H2O which is compatible with chromatographic separation for lipid and metabolite profiling. A hydrophilic interaction liquid chromatography (HILIC) method was optimized for chromatographic separation of lipid and metabolite classes. Data was collected on a SYNAPT XS traveling-wave ion mobility-mass spectrometer. Data was collected in both positive and negative electrospray ionization modes. Retention time, accurate mass, fragmentation, and collision cross section (CCS) values were used for identification of metabolites and lipids.

RESULTS: We analyzed three different extraction methods for their abilities to simultaneously recover lipid and metabolites from Gram-positive and Gram-negative bacteria. The presence of the outer membrane in Gram-negative organisms did influence the recovery of lipids from those organisms relative to the Gram-positives. Using multivariate analyses revealed key distinguishing features between the Gram-negatives and Gram-positives. While the presence of phosphatidylethanolamines (PEs) were found only in Gram-negative, all organisms shared the presence of phosphatidylglycerols (PGs). The fatty acyl composition of the phospholipids further distinguished organisms to the genus level. Metabolite profiles and levels also varied between the organisms. We detected 2249 and 2109 features with high significance in the positive and negative mode datasets, respectively. Thus far, features identified include amino acids, quorum sensing metabolites, and other compounds related to central carbon and tricarboxylic acid cycles that are relevant for classification at the Gram-level.

CONCLUSION: We have developed analytical methods to enable classification using the combined metabolic and lipidomic profiles of bacteria. In addition to improved extraction techniques, we optimized a mass spectrometry method coupled to ion mobility for multidimensional separation with the capability to identify significant features that separate each strain based on organism-specific lipid and metabolite profiles. The multi-dimensionality of our identification strategy will allow for more rapid classification and maximize efficiency for more tailored clinical applications with further development.