Strains CZ1424 and CZ1443 were grouped in the same cluster with a distance level

of up to 5, as were strains CZ1429 and CZ1449. Conversely, strains CZ1523 and CZ1504 were grouped in a different cluster, at a distance level greater than 10. Strain CZ1427, which showed a 60% similarity with the other strains isolated from the same patient, was grouped at an inter-strain distance level of 10–15. Figure 4 Score-oriented dendrogram of matrix-assisted laser desorption ionization time-of-flight mass spectrometry this website profiles generated by the default setting in MALDI Biotyper software version 2.0. Discussion O. anthropi is an adaptable bacterial species, whose individual strains can thrive in different environments. Indeed, after its molecular characterization [11] human-associated clonal complex data appear to indicate it possesses a specialized opportunistic behaviour [3]. It is frequently isolated from contaminated medical materials/devices and specimens obtained from immunocompromised patients [3], and after the first recognized

check details case of human disease induced by this organism [16], O. anthropi infections causing primary or catheter-associated bacteraemia [1, 17] have been increasingly reported [4]. With this in mind, when this infection did occur in our hospital, we set out to study the identification and typing of the O. anthropi strains through the genomic and proteomic correlation. To our knowledge, this represents the first study on strain typing of O. anthropi

where the use of both rep-PCR and MALDI-TOF-MS-based fingerprinting were carried out. All patients developed infection during their stay in hospital, and in our Institution no cases of infection due to O. anthropi had been diagnosed before. Environmental and flushing solution cultures were negative for O. anthropi, therefore the source of the infection strains remained unclear. Fluoroquinolone monotherapy yielded good clinical response, however blood cultures from all patients became negative only after removal of CVC. for Our results indicate that all investigated strains were highly related and that they arose from a common ancestor, strongly providing evidence for a clonal origin of the infection. Interestingly, the strains detected early on during the outbreak showed a great variability in correlation range (90%–99%), while bacteria isolated later showed a correlation higher than 99%. We can therefore speculate that O. anthropi is able to undergo rapid modifications, allowing bacteria to adapt to a human host. The proteomic profiles, which clustered the 23 strains in a single group, unrelated to the ATCC isolates present in the database (one of which comes from leech urine), further suggest a clonal origin of the infections.