Date Published: January 26, 2017
Publisher: Public Library of Science
Author(s): Marta Pérez-Sancho, Ana I. Vela, Teresa García-Seco, Sergio González, Lucas Domínguez, Jose Francisco Fernández-Garayzábal, Yung-Fu Chang.
The application of MALDI-TOF MS for identifying streptococcal isolates recovered from clinical specimens of diseased pigs was evaluated. For this proposal, the MALDI BDAL Database (Bruker Daltoniks, Germany) was supplemented with the main spectrum profiles (MSP) of the reference strains of S. porci, S. porcorum and S. plurextorum associated with pneumonia and septicemia. Although these three species showed similar MALDI profiles, several peaks were recognized that can be useful for their differentiation: S. porci (4113, 6133, 7975 and 8228 m/z Da), S. plurextorum (3979, 4078, 4665, 6164, 6491, 6812, 7959 and 9330 m/z Da) and S. porcorum (3385, 3954, 4190, 6772, 7908, and 8381 m/z Da). After adding these MSPs, an evaluation was conducted to determine the accuracy of MALDI-TOF MS for the identification of streptococci from diseased pigs using 74 field isolates. Isolates were identified as S. suis, S. porcinus, S. dysgalactiae, S. hyovaginalis, S. porcorum, S. alactolyticus, S. hyointestinalis and S. orisratti. This is the first time that the latter three species have been reported from clinical specimens of pigs. Overall, there was good concordance (95.9%) between the results obtained from MALDI-TOF MS identification (best hint) and those from genotyping. Our results demonstrate the good performance of MALDI-TOF MS (100% sensitivity and specificity) for identifying most of the species of streptococci that can frequently be isolated from diseased pigs. However, conflicting results were observed in the correct identification of some isolates of S. dysgalactiae and S. alactolyticus.
Several species of streptococci, such as Streptococcus suis, Streptococcus dysgalactiae subsp. equisimilis and Streptococcus porcinus are well-recognized swine pathogens [1, 2]. S. suis is by far the most important agent of infectious diseases in this group. It is associated with a variety of clinical conditions including meningitis, septicemia, arthritis, endocarditis and pneumonia. It has also been isolated from cases of rhinitis and abortion. S. dysgalactiae subsp. equisimilis causes sporadic cases of septicemia and arthritis in sucking pigs, endocarditis in growing pigs and ascending infection of the uterus in sows [1, 3]. S. porcinus is the etiological agent of streptococcal lymphadenitis in growing pigs. It also causes throat abscesses and has sometimes been isolated from pneumonia [1, 4, 5]. Other Streptococcus species have been isolated from clinical specimens in swine. Streptococcus plurextorum has been isolated from lesions associated with pneumonia and septicemia  and Streptococcus porci and Streptococcus porcorum have also been associated with pneumonia in pigs [7, 8]. However, there are no data on the prevalence and clinical significance of these three latter species.
The present work does not include any experimental infection trials with pigs. Only diseased farmed pigs were used to microbiologically identify the etiological agent of the disease. We did not, therefore, consult with the Institutional Animal Care and Use Committee and no specific national regulations for these procedures are available. Diseased pigs were necropsied under strict hygienic conditions by the veterinarian in the farms immediately after the pig´s death and clinical specimens were sent to the Health Surveillance VISAVET Centre of the Universidad Complutense (Madrid, Spain) for a confirmatory microbiological diagnosis. Clinical specimens were handled following the recommendations of the OIE for the transport of specimens of animal origin (http://asforce.org/course/assets/img/module2/transport.pdf). Upon arrival, clinical samples were processed in biological safety cabinets within 12h. Samples were cultured on Columbia and Columbia-CNA agar plates that were incubated at 37°C for 24h under aerobic conditions.
Analysis of the spectra of these strains was performed by selecting a range of m/z between 4000 and 9500 Da that included the major differences between their MSPs. The spectra of S. porci and S. plurextorum were quite similar, but closer analysis of their mass peak profiles revealed some different mass patterns for each species (Fig 1).
Among the Streptococcus species associated with infections in pigs , only S. suis, S. dysgalactiae and S. porcinus were included in the Bruker MALDI BDAL database (Bruker Daltoniks) at the time of this study. However, the species S. porci, S. porcorum and S. plurextorum associated with pneumonia and septicemia in pigs [6–8] were not included, which limits the utility of this technology for the diagnosis of diseases caused by streptococci. Therefore, the first objective of this study was to construct the MSPs of S. porci, S. porcorum and S. plurextorum and insert them into the Bruker database. Analysis of the MSPs of these three species as well as those of S. suis, S. dysgalactiae and S. porcinus identified some species-specific m/z peaks that exhibited good discriminatory power based on the data of their intensity, signal-to-noise and frequency of detection (Table 1), and therefore likely useful for the differentiation of these six species of streptococci. The reproducibility of mass spectra profiles could be affected by intraspecies variability . Therefore, future analysis of the mass spectra of more isolates of these species will allow confirmation of whether these biomarkers are consistently present in the majority of their respective species. We further evaluated the efficacy of MALDI-TOF MS for the identification of streptococci associated with infections in pigs using a panel of 74 field clinical isolates (S1 Table). Overall, there was good concordance between the results of identification obtained with the MALDI-TOF MS system and those obtained by genetic identification, with both approaches giving a matching identification for 95.9% of the streptococci considering the best hint identification of MALDI-TOF MS (Table 2). However a certain dispersion in the estimation (95% CI: 20.7–100 for sensitivity and 95% CI: 90.8–100 for specificity) could be due to the limited number of isolates in some Streptococcus species here evaluated. Most field isolates (70.3%) were identified as S. suis, S. porcinus or S. dysgalactiae, a result that is in line with the epidemiological relevance of these pathogens in the swine industry [1,2]. The relatively high number of clinical isolates of the species S. hyointestinalis, S. alactolyticus and Streptococcus orisratti was also interesting (Table 2). The two former species are considered to be part of the microflora in the intestine  and S. orisratti was initially isolated from the teeth of healthy rats . Isolates of these species were recovered from lesions associated with pneumonia, endocarditis or arthritis (Table 2) and therefore, this result represents the first report of the isolation of these species from clinical specimens of pigs. However, no definitive conclusions about the clinical significance of these species as pig pathogens can be reached from the results of this study.
Despite some limitations regarding the identification of S. dysgalactiae and S. alactolyticus, the results of this study confirm that MALDI-TOF MS represents a good diagnostic tool for identifying most of the species of streptococci that are frequently isolated from diseased pigs and therefore for the diagnosis of disease caused by these bacteria.