Proteomic approach for bacteria of food interest identification: on-plate trypsinolysis followed by MALDI-MS/MS

Whatever the field (clinical or environmental), classification and identification of bacterial strains is fundamental. For fermented foods industries, bacterial identification is a key element for risk assessment. Bacteria of food interest, mainly represented by lactic acid bacteria, are a major part of many fermented products and are found in a variety of environments. Thus the development of a rapid and simple method for their identification is particularly relevant. Most of the techniques currently used are based on molecular approaches or conventional carbohydrates use tests. However, these methods are generally time consuming. This is why MALDI Mass Spectrometry approach has been developed and is now able to determine the species within a few minutes. The first approach, based on bacterial mass fingerprinting, is now well known and widespread in clinical microbiology, but it requires commercial software package. We decide here to explore an alternative approach based on proteomics and using genomic databases freely available online. One of our objectives was to develop an automated method performing high throughput identification. MALDI MS/MS experiments were performed after [i]in situ[/i] trypsinolysis, directly on MALDI plate. For this approach we used the on line UniProt database (http:/www.uniprot.org). Samples, after trypsin digestion, were analyzed in parallel with nano LC-ESI-MS/MS technique as a control. All mass spectra were recorded using a hybrid quadrupole time-of-flight mass spectrometer QStar XL. We developed the method for five bacterial genera: [i]Lactobacillus, Streptococcus, Lactococcus, Enterococcus[/i], and [i]Propionibacterium[/i]. A total of 26 species of food interest, represented by type strains, were then studied. [i]Staphylococcus aureus [/i](type strain) was also analyzed in order to have a point of comparison with a species extensively sequenced for which a very large number of proteomic data exists in databases. For a species, the higher number of strains sequenced, the more efficient the identification by proteomic approach. Thus, the limit of our method is the number of data available for each species in database. Thanks to the new generation sequencing technologies, the number of bacterial species sequenced is increasing exponentially and this would not be a limit any longer. Thus, the obtained results highlight the interest of the proteomic approach for identification by Mass Spectrometry.