Résumé
Vibrio bacteria, and particularly members of the Harveyi clade, are the causative agents of vibriosis. This disease is responsible for mass mortality events and important economic losses on aquaculture farms. Improvements in surveillance and diagnosis are needed to successfully manage vibriosis outbreaks. 16S rRNA gene sequencing is generally considered to be the gold standard for bacterial identification but the cost and long processing time make it difficult to apply for routine identification. In contrast, MALDI-TOF MS offers rapid diagnosis and is commonly used in veterinary laboratories today. The major limiting factor for using this technique is the low environmental bacterial diversity in the commonly available databases. Here, we demonstrate that the sole use of the commercially available Bruker BioTyper database is not fully adequate for identifying Vibrio bacteria isolated from aquaculture farms. We therefore developed a new in-house database named Luvibase, composed of 23 reference MALDI-TOF mass spectra profiles obtained from Vibrio collection strains, mostly belonging to the Harveyi clade. The comparison of the accuracy of MALDI-TOF MS profiling and 16S rRNA gene sequencing revealed a lack of resolution for 16S rRNA gene sequencing. In contrast, MALDI-TOF MS profiling proved to be a more reliable tool for resolving species-level variations within the Harveyi clade. Finally, combining the Luvibase with the Bruker ver.9.0.0.0 database, led to successful identification of 47 Vibrio isolates obtained from moribund abalone, seabass and oysters. Thus, the use of Luvibase allow for increased confidence in identifying Vibrio species belonging to the Harveyi clade.