Universal, untargeted detection of bacteria in tissues using metabolomics workflows

Area of Science:

• Microbiology
• Analytical Chemistry
• Biomarker Discovery

Background:

• Current bacterial identification methods in clinical diagnostics are time-consuming, involving lengthy isolation and enrichment steps.
• This delay in diagnosis can postpone crucial stratified patient treatment.
• There is a critical need for faster, more direct methods for bacterial detection in complex biological samples.

Purpose of the Study:

• To develop and validate a biomarker-based strategy for the direct detection of bacteria in clinical samples.
• To utilize bacterial small molecular metabolites and lipids as specific markers for identification.
• To demonstrate the applicability of this method across different mass spectrometry techniques.

Main Methods:

• A spectral metabolic library encompassing 233 bacterial species was curated.
• Taxon-specific markers (TSMs) were identified using univariate statistical analysis.
• The TSMs were applied for in situ bacterial detection in gastrointestinal tissues and faecal samples using spatial and bulk metabolomics.

Main Results:

• 359 taxon-specific markers (TSMs) were identified, demonstrating specificity across phylogenetic levels.
• The TSM-based approach was successfully applied to detect bacteria in healthy and cancerous gastrointestinal tissues and faecal samples.
• The method showed high applicability (>90% of samples detected) across different mass spectrometry analytical approaches.

Conclusions:

• A novel biomarker-based strategy enables direct and rapid bacterial detection in complex clinical samples.
• The identified taxon-specific markers (TSMs) are broadly applicable across various mass spectrometry techniques.
• This approach has the potential to significantly accelerate clinical microbiological diagnostics and improve patient care.