Rapid identification of viridans streptococci by mass spectrometric discrimination

Abstract

Viridans streptococci (VS) are responsible for several systemic diseases, such as endocarditis, abscesses, and septicemia. Unfortunately, species identification by conventional methods seems to be more difficult than species identification of other groups of bacteria. The aim of the present study was to evaluate the use of cell matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS) for the rapid identification of 10 different species of VS. A total of 99 VS clinical isolates, 10 reference strains, and 20 strains from our in-house culture collection were analyzed by MALDI-TOF-MS. To evaluate the mass-spectrometric discrimination results, all strains were identified in parallel by phenotypic and genotypic methods. MALDI-TOF-MS identified 71 isolates as the mitis group, 23 as the anginosus group, and 5 as Streptococcus salivarius. Comparison of the species identification results obtained by the MALDI-TOF-MS analyses and with the phenotypic/genotypic identification systems showed 100% consistency at the species level. Thus, MALDI-TOF-MS seems to be a rapid and reliable method for the identification of species of VS from clinical samples.

FIG. 1.

MALDI-TOF mass spectra of different streptococcal reference strains: S. oralis DSM 20627, S. mitis DSM 12643, S. gordonii DSM 6777, S. sanguinis DSM 20567, S. parasanguinis DSM 6778, S. anginosus DSM 20563, S. constellatus DSM 20575, S. intermedius DSM 20573, S. salivarius DSM 20560, and S. mutans DSM 20523.

FIG. 2.

Results of similarity analysis of reference strains and additional reference strains. The dendrogram was calculated from a peak mass-based similarity matrix. G1 and G2, S. gordonii DSM 6777, respectively; G3 and G4, two different culture collection strains of S. gordonii, respectively; P1 and P2, two colonies of S. parasanguinis DSM 6778, respectively; P3 and P4, two different culture collection strains of S. parasanguinis, respectively; I1 and I2, two colonies of S. intermedius DSM 20573, respectively; I3 and I4, two different culture collection strains of S. intermedius, respectively; C1 and C2, two colonies of S. constellatus DSM 20575, respectively; C3 and C4, two different culture collection strains of S. constellatus, respectively; A1 and A2, two colonies of S. anginosus DSM 20563, respectively; A3 and A4, two different culture collection strains of S. anginosus, respectively; San1 and San2, two colonies of S. sanguinis DSM 20567, respectively; San3 and San4, two different culture collection strains of S. sanguinis, respectively; Mi1 and Mi2, two colonies of S. mitis DSM 12643, respectively; Mi3 and Mi4, two different culture collection strains of S. mitis, respectively; O1 and O2, two colonies of S. oralis DSM 20627, respectively; O3 and O4, two different culture collection strains of S. oralis, respectively; Sal1 and Sal2, two colonies of S. salivarius DSM 20560, respectively; Sal3 and Sal4, two different culture collection strains of S. salivarius, respectively; Mu1 and Mu2, two colonies of S. mutans DSM 20523, respectively; Mu3 and Mu4, two different culture collection strains of S. mutans, respectively.

FIG. 3.

Results of the identification of one clinical sample (isolate VS 162) on the basis of a similarity analysis of the MALDI-TOF mass spectra. The peak list generated from the spectrum of VS 162 was compared to the peak lists of all reference and culture collection strains (VS database). On the y axis, the distance between the spectrum of VS 162 and the spectra of the references and culture collection (all measured in triplicate) is given. On the x axis, the entries of the VS database are arranged according to the increasing dissimilarities of their MALDI-TOF-MS spectra compared to the spectrum of VS 162. For clarity, only the results for the 19 most similar entries of the VS database are shown. San1-1, San1-2, and San1-3, triplicate cultures of S. sanguinis DSM 20567; San2-1, San2-2, San3-1, San3-2, San4-1, and San4-2, duplicate cultures of three different culture collection strains of S. sanguinis; G1-2 and G1-3, duplicate cultures of S. gordonii DSM 6777 Il-2, S. intermedius DSM 20573; C2-1, C2-2, and C3-1, duplicate cultures and single culture of two different culture collection strains of S. constellatus; A1-1, culture collection strain of S. anginosus; O3-2, culture collection strain of S. oralis. Identical entries at different positions on the x axis represent duplicate MALDI-TOF mass spectra from one culture.

FIG. 4.

Examples of results of the identification of the 23 samples which could be identified as either S. mitis or S. oralis by species-specific PCR and sequence analysis of the 16S RNA gene (Table 1, S. oralis/S. mitis [one clinical sample]) on the basis of a similarity analysis of the MALDI-TOF mass spectra. The peak list generated from the spectra were compared to the peak lists of all reference and culture collection strains (VS database). On the y axis, the distance between the spectrum of the respective clinical isolate and the spectra of the reference and culture collection isolates (all measured in triplicate) is given. On the x axis, the entries of the VS database are arranged according to the increasing dissimilarities of their MALDI-TOF-MS spectrum compared to the spectrum of the respective clinical isolate. (a) Results of identification of a clinical sample which can be identified as S. oralis; (b) results of identification of a clinical sample which can be identified as S. mitis; (c) results of identification of a clinical sample (VS 23) which cannot be identified as either S. oralis or S. mitis. See the legends to Fig. 2 and 3 for an explanation of the VS database designations.