Shortcomings of the Commercial MALDI-TOF MS Database and Use of MLSA as an Arbiter in the Identification of Nocardia Species

Gema Carrasco, Juan de Dios Caballero, Noelia Garrido, Sylvia Valdezate, Rafael Cantón, Juan A. Sáez-Nieto

Nocardia species are difficult to identify, a consequence of the ever increasing number of species known and their homogeneous genetic characteristics. 16S rRNA analysis has been the gold standard for identifying these organisms, but proteomic techniques such as matrix-assisted laser desorption ionization-time of flight (MALDI-TOF MS) and housekeeping gene analysis, have also been explored. One hundred high (n = 25), intermediate (n = 20), and low (n = 55) prevalence (for Spain) Nocardia strains belonging to 30 species were identified via 16S rRNA and MALDI-TOF MS analysis. The manufacturer-provided database MALDI Biotyper library v4.0 (5.627 entries, Bruker Daltonik) was employed. In the high prevalence group (Nocardia farcinica, N. abscessus, N. cyriacigeorgica and N. nova), the 16S rRNA and MALDI-TOF MS methods provided the same identification for 76% of the strains examined. For the intermediate prevalence group (N. brasiliensis, N. carnea, N. otitidiscaviarum and N. transvalensis complex), this figure fell to 45%. In the low-prevalence group (22 species), these two methods were concordant only in six strains at the species level. Tetra-gene multi-locus sequencing analysis (MLSA) involving the concatemer gyrB-16S rRNA-hsp65-secA1 was used to arbitrate between discrepant identifications (n = 67). Overall, the MLSA confirmed the results provided at species level by 16S rRNA analysis in 34.3% of discrepancies, and those provided by MALDI-TOF MS in 13.4%. MALDI-TOF MS could be a strong candidate for the identification of Nocardia species, but only if its reference spectrum database improves, especially with respect to unusual, recently described species and species included in the described Nocardia complexes.