Rapid Identification of Food-Borne Bacteria Using a Vast Database of Theoretical Prokaryotic Protein Masses Predicted from Genome Sequences for MALDI-MS

Importance of the Topic

Rapid and accurate identification of food-borne microorganisms is essential for ensuring food safety and quality control. Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) offers fast microbial profiling but relies on curated spectral libraries that may lag behind taxonomic updates. A genome-derived theoretical protein mass database can overcome this limitation and expand the range of identifiable taxa in food microbiology.

Objectives and Study Overview

This work describes the development of a large‐scale Genomically Predicted Protein Mass database (GPMsDB) and a cloud-based web application, MicrobialTrack. The goals were to (1) compile theoretical masses for proteins from public prokaryotic genomes, (2) implement an automated MALDI-MS identification workflow, and (3) evaluate performance on over 20 food-borne bacterial strains and multiple Acinetobacter species by comparison with 16S rRNA gene analysis.

Methodology and Instrumentation

GPMsDB was built from approximately 402,000 genomes covering some 85,000 prokaryotic species. The database offers two modes: an “All” database with one entry per genome and a “Representative” database with one entry per species for rapid genus- and species-level screening. MicrobialTrack is delivered as a web service without local installation or update costs.

Instrumentation Used:

• MALDI-8030 (Shimadzu) mass spectrometer (Research Use Only).
• α-Cyano-4-hydroxycinnamic acid (CHCA) matrix.
• Sample preparation methods: direct smear, formic acid extraction, ethanol wash with FA/ACN extraction.

Main Results and Discussion

Most food-borne and Acinetobacter strains were correctly identified at the species level with high confidence. Acinetobacter pittii, A. baumannii and related species yielded clear spectral matches. A resampling algorithm provided reliability scores by randomly selecting subsets of observed peaks. In one case, strains identified as Lacticaseibacillus paraplantarum by 16S rRNA were assigned to L. plantarum by MicrobialTrack; both species share nearly identical ribosomal protein profiles, leading to subtle differences in non-ribosomal peak counts.

Benefits and Practical Applications

MicrobialTrack enables rapid, up-to-date microbial identification without user-side library maintenance. The cloud platform minimizes initial costs and simplifies taxonomic database updates. Food safety laboratories can integrate this workflow for routine screening, outbreak investigation, and environmental monitoring.

Future Trends and Opportunities

Advancements may include integration of whole-genome sequencing data, expanded coverage of rare and uncultured taxa, incorporation of post-translational modification profiles, and coupling with machine-learning algorithms for automated spectral interpretation. Portable MALDI-MS platforms could bring on-site microbial surveillance to production lines.

Conclusion

The combination of GPMsDB and the MicrobialTrack web service offers a comprehensive, flexible, and accurate MALDI-MS identification solution for food-borne bacteria and archaea. By leveraging genome-predicted protein masses, this approach addresses library update challenges and supports robust taxonomic assignments in food microbiology contexts.

References

• Sekiguchi Y., et al. Genome Biology, 2023.
• Genome Taxonomy Database (GTDB): https://gtdb.ecogenomic.org