ABSTRACT
In this study we evaluated the capacity of MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) to identify clinical mould isolates. We focused on two aspects of MALDI-TOF MS identification: the sample processing and the database. Direct smearing of the sample was compared with a simplified processing consisting of mechanical lysis of the moulds followed by a protein extraction step. Both methods were applied to all isolates and the Filamentous Fungi Library 1.0 (Bruker Daltonics) was used for their identification. This approach allowed the correct species-level identification of 25/34 Fusarium spp. and 10/10 Mucor circinelloides isolates using the simplified sample processing. In addition, 7/34 Fusarium spp. and 1/21 Pseudallescheria/Scedosporium spp. isolates were correctly identified at the genus level. The remaining isolates-60-could not be identified using the commercial database, mainly because of the low number of references for some species and the absence of others. Thus, an in-house library was built with 63 local isolates previously characterized using DNA sequence analysis. Its implementation allowed the accurate identification at the species level of 94 isolates (91.3%) and the remaining nine isolates (8.7%) were correctly identified at the genus level. No misidentifications at genus level were detected. In conclusion, with improvements of both the sample preparation and the feeding of the database, MALDI-TOF MS is a reliable, ready to use method to identify moulds of clinical origin in an accurate, rapid, and cost-effective manner.