Development of Legionella pneumophila serogroups classification by FT-IR spectroscopy with IR – Biotyper system

Question: For routine purposes, the ideal method for microbial characterization requires minimum sample preparation and should be rapid, automated and inexpensive. It is well known that culture and phenotyping methods to recognize Legionella spp. are often inadequate, while genotypic methods are expensive and laborious. Recently, mass spectrometry and infrared spectroscopy have proved to be rapid and successful approach for the identification of several pathogens. This technique, enables the creation of classification systems based on the intra-species identification features typical of each microorganism. The aim of the study is the use of Fourier Transform Infrared Spectroscopy technology using the IR Biotyper system (Bruker Daltonik, Germany) for the classification of Legionella pneumophila (Lp) at serogroup (Sg) level, for diagnostic purposes as well as in outbreak events. Lp inhabits fresh water environments and it represents a serious risk for human health, since it is the causative agent of a severe community-acquired pneumonia. It consists of 16 Sgs, most of these typed by monoclonal or polyclonal antibodies (e.g. agglutination test or indirect immunofluorescence assay). Methods: IR Biotyper was used to classify Lp with an Artificial Neural Network (ANN) available in the instrument software. A dataset of well characterized isolates (n=43), belonging to Sg1-15, including both ATCC strains and environmental strains collected from different sources, was tested with a classifier implemented by the manufacturer. All strains were cultivated onto BCYE agar at 35 ± 2°C for 48 h, with 2.5% CO2 and typed by latex agglutination test (Thermo Fisher Diagnostic, Basingstoke, UK). All the isolates were classified with the Bruker classifier that at moment is able to discriminate between Sg1 and Sg 2-15. Results: The preliminary results showed that IR Biotyper classifier was able to correctly discriminate Sg1 from 2-15, showing accuracy=95.9% (41/43) and error rate=4.1% (2/43). An exception was observed for the two Sg7 isolates, that were mis-classified as Sg1 (likely due to the absence of this Sg in the training set of the Bruker classifier). Conclusions: Further studies on a larger number of isolates could be useful to implement the ANN system and to obtain a robust and reliable tool for the identification of Lp at the Sg level. In conclusion, the results of this study suggest that IR Biotyper could be a powerful and easy-to-use tool for characterizing Lp Sgs especially during cluster/outbreak investigations, to promptly undertake preventive strategies and trace the source of the infection.