Abstract
The identification of avian Mycoplasma spp. by conventional immunologic, phenotypic, and molecular methods can be demanding and time-consuming. We evaluated MALDI-TOF MS for its suitability to identify avian mycoplasmas at the species level. We generated a mycoplasma spectral database of 36 main spectrum profiles (MSPs) representing 23 avian Mycoplasma spp. using 23 type and reference strains, 1 live vaccine strain, and 8 clinical isolates. We then used 112 avian Mycoplasma clinical isolates of different avian mycoplasmas, 4 Mycoplasma live vaccine strains, and 1 Mycoplasma type strain, previously cultured and identified to the species level by molecular methods, to evaluate the MSP database. Protein extraction and MALDI-TOF MS analysis were performed with a maximum of 3 repetitions per isolate. MALDI-TOF MS resulted in accurate species-level identification with a score of ≥2.0 for 112 of 117 (96%) isolates. The MALDI-TOF MS analysis of 4 of 5 isolates that did not yield a score of ≥2.0 resulted in best-match identifications that were still concordant at species level with the molecular method used for previous identification. Therefore, MALDI-TOF MS is a promising tool for reliable identification of avian Mycoplasma spp.
Mycoplasmas, bacteria belonging to class Mollicutes, are a part of the commensal microbial flora of many avian species, but some of the 27 avian Mycoplasma species are important pathogens in veterinary medicine. 6 These pathogenic species can cause high economic losses in poultry production. 6 Wild birds can also be affected.5,17 Given that mixed infections with Mycoplasma spp. may occur, the detection of mycoplasmas does not necessarily lead to the diagnosis of a mycoplasmosis, and species differentiation is required.15,20
Infections with avian mycoplasmas can be detected indirectly by a serologic assay or directly by methods including conventional isolation and identification as well as genus-specific or species-specific PCRs. 11 However, many laboratory tests are only available and validated for pathogenic avian mycoplasmas in poultry. 12 Such tests can be used for diagnostic purposes if a specific Mycoplasma species is anticipated and if a test for this specific species is available. Hence, these tools can be used to detect pathogenic Mycoplasma spp. in poultry, but are of limited use for diagnostic or epidemiologic studies for other Mycoplasma spp. in other birds. Nonetheless, the role of mycoplasmas in diseases of wild birds varies among bird species or is not fully studied yet. 20 A single generic test for the identification and differentiation of multiple avian Mycoplasma spp. is needed.
Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is a fast, reliable, and cost-effective technology that is widely used for the identification of bacterial species by comparing their protein profiles to reference spectra. 3 It has been shown to identify even fastidious bacteria successfully. 2 MALDI-TOF MS has been used for identification of Mycoplasma pulmonis isolates from laboratory rodents 10 as well as for human and ruminant Mycoplasma spp.10,14 Identification of avian Mycoplasma spp. by MALDI-TOF MS has not been reported, to our knowledge. We investigated MALDI-TOF MS for the differentiation and identification of avian mycoplasmas to the species level by creating and evaluating an avian mycoplasma spectral database.
We constructed the database from 36 main spectrum profiles (MSPs) from 23 type and reference strains, 1 live vaccine strain, and 8 clinical isolates of 23 avian Mycoplasma spp., previously identified by molecular methods (Table 1). Apart from one M. gallisepticum (MG) live vaccine strain, all strains and isolates used for MSP creation were identified by genus-specific PCR and sequencing (LGC Genomics, Berlin, Germany). 18 The MG live vaccine strain was identified by species-specific PCR and restriction enzyme analysis as well as by sequencing of the PCR amplicon. 13 The 112 clinical isolates analyzed after the database construction were collected from 10 different avian species (Table 2). Additionally, 2 live MG vaccine strains (Nobilis MG 6/85, Merck, Kenilworth, NJ; F-strain) and the type strain of MG (PG31T) as well as 2 live M. synoviae vaccine strains (Vaxsafe MS, Bioproperties, Ringwood, VIC, Australia; Nobilis MS live, Merck) were analyzed. Single colony subcultures were performed 3 times to rule out mixed cultures for all 117 isolates. The subcultures were identified by genus-specific PCR followed by sequencing of PCR product or by MG species-specific PCR and restriction enzyme analysis as well as sequencing of the PCR amplicons.13,18
Details of the type strains (*), clinical isolates (†), live vaccine strains (‡), and reference strains (§), including their origin, used in the generation of the main spectra database.
ANICON = AniCon Labor; BB = Janet Bradbury, University of Liverpool; DSMZ = German Collection of Microorganisms and Cell Cultures (https://www.dsmz.de/); KVRAF = Clinic for Birds, Reptiles, Amphibians and Fish, Justus Liebig University Giessen; NCTC = National Collection of Type Cultures (https://www.phe-culturecollections.org.uk/collections/nctc.aspx); TMC = The Mollicutes Collection of Cultures and Antisera of the International Organization for Mycoplasmology (http://iom-online.org/node/28). All strains and isolates were identified by a 16S rRNA PCR followed by amplicon sequencing 18 except the M. gallisepticum F-strain, which was identified by a species-specific PCR and both restriction endonuclease analysis and sequencing of the amplicon. 13
Number and originating bird species of the isolates of Mycoplasma spp. identified by MALDI-TOF MS.
The samples were cultured in SP4 broth and agar medium or in AL-avian mycoplasma liquid medium and AS-mycoplasma agar & supplement (Mycoplasma Experience, Bletchingley, Surrey, UK) as described previously. 5 The broth was diluted (10-fold dilution up to 10−2), and an aliquot of 35 µL of each dilution was transferred onto agar medium. The cultures were incubated at 37°C with 5% CO2 in a humidified environment for up to 5 d. The solid medium was checked daily for mycoplasmal growth. If colony growth on the agar plate was detected, MALDI-TOF MS analysis was performed on the liquid medium. In case of discordant results between the previous molecular identification and MALDI-TOF MS, the isolates were again examined by genus-specific PCR targeting the 16S rRNA gene and the 16-23S rRNA intergenic transcribed spacer region (IRS) and sequencing of the PCR amplicons.16,18
The incubated liquid medium was centrifuged to obtain a Mycoplasma pellet. Although 1 mL of liquid medium was used for the database construction, 1 mL or 2.5 mL of liquid medium was used for the MALDI-TOF MS analysis of the clinical isolates depending on their growth capacities. The liquid medium was centrifuged (8,600 × g, 30 min) and the pellet washed twice in phosphate-buffered saline. For precipitation of proteins, 300 µL of water and 900 µL of absolute ethanol were added. After centrifugation (17,000 × g, 2 min) the supernatant was removed. The dried pellet was dissolved in an equal volume (10 µL) of 70% formic acid and acetonitrile and then centrifuged (17,000 × g, 2 min). An aliquot of 1 µL of the supernatant per target spot was placed onto a MALDI 96-target polished steel plate (Bruker, Daltonics, Bremen, Germany). For MSP creation, 8 spots per isolate were prepared. The protein extracts of the clinical isolates were spotted twice or, in case of visual non-homogeneity, up to 4 times. After air-drying at room temperature, the protein extracts were overlaid with an equal amount of matrix solution composed of 10 mg/mL of α-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 2.5% trifluoroacetic acid. Mass spectra were generated (Microflex LT MS, Biotyper operating system; Bruker Daltonics) in a positive linear mode at a laser frequency of 60 Hz with an acceleration voltage of 20 kV and a mass range of 2,000–20,000 Da. The data were analyzed in the automatic mode (Biotyper v.3.1) and compared to the MALDI Biotyper MSP database and the database generated in our study. Spectra were internally calibrated using a bacterial test standard (part 255343; Bruker Daltonics).
The Mycoplasma MSP database was constructed by assigning the reference peak lists to different reference strains. Three mass spectrum measurements of 8 different spots of protein extract for each strain were obtained. Quality control of the raw mass spectra was performed (Flex analysis v.3.4; Bruker Daltonics) and included a check for absence of flat-line spectra, intrusive peaks, and low matrix background signal. After smoothing, baseline correction, and peak picking, at least 20 spectra were selected for MSP creation. The automated MSP creation functionality of the MALDI Biotyper software calculated a MSP with information about mean peak masses, peak intensities, and peak frequencies.
For MALDI-TOF MS analysis of the isolates, each spot was measured once. Spectra were classified by matching MSPs (MALDI Biotyper real-time classification software; Bruker Daltonics) with the database mentioned previously. The degree of spectral concordance was expressed as a logarithmic identification score ranging from 0 to 3. The manufacturer’s instructions were used for the interpretation of acquired scores: ≥2.0 = species identification with a high level of confidence; ≥1.7 but <2.0 = a valid genus identification; <1.7 = no reliable identification was obtained. An isolate was considered correctly identified to the species level if at least one of the spots yielded a score of ≥2.0. Otherwise, the isolate was cultivated and analyzed by MALDI-TOF MS again, with a maximum of 2 repetitions.
A total of 112 of 117 (96%) isolates studied yielded a score of ≥2.0 with a best-match identification of the species accordant to the molecular identification. For 3 of 117 (3%) isolates, the result was a genus-level identification and a best-match identification of the species that matched the molecular identification. For 2 of 117 (2%) isolates, no reliable identification was obtained (Table 3). One of these 2 isolates resulted in a best-match identification of M. gypis in all MALDI-TOF MS runs. The 16S rRNA gene and ISR sequences of this isolate were 98% and 97% identical to M. gypis, respectively. The second discordant isolate resulted in different best-match identifications (e.g., Sphingomonas pseudosanguinis, Citrobacter freundii, Bacillus novalis) in each run, with a maximum score of 1.306, including no Mycoplasma species. The sequences of the 16S rRNA gene and the ISR of this isolate were 97% and 95% identical to M. gypis, respectively. Regarding the number of runs needed, the first run resulted in correct species identification with scores of ≥2.0 for 93 of 117 (80%) isolates, whereas for 15 of 117 (13%) a second run, and for 3 of 117 (3%) a third run was needed.
Best scores per isolate in the indicated MALDI-TOF runs.
DR = discordant result compared with molecular identification; NP = no peaks.
Given that Mycoplasma colonies are small and often inlaid in the agar, the direct deposition of colonies has been shown to often lead to inaccurate MALDI-TOF MS results. 14 Therefore, agar medium was used to identify the Mycoplasma growth, but we used broth medium as a source of material for MALDI-TOF MS analysis. To improve the spectral quality, multiple washing steps were performed on the Mycoplasma pellet to reduce the protein background generated by media contents (i.e., porcine serum and yeast). However, washing the pellet also decreases the quantity of protein in the subsequent protein extraction. 14 Given that the approach of generating spectra by applying material directly to the target plate without protein extraction resulted in weak scores, protein extraction was performed prior to MALDI-TOF MS, as described for other bacteria.7,14
The analysis of 3 of 117 (3%) isolates resulted in a genus-level identification (score of ≥1.7 but <2.0) with a best-match identification of the species concordant to the molecular identification. Therefore, an adjustment of the acceptable score for Mycoplasma spp. identification to ≥1.7 would result in a correct species identification of 115 of 117 (98%) and may be appropriate, as discussed in previous studies.1,4,8,14 One of the 2 isolates with a score of <1.7 resulted in a concordant best-match identification to the molecular identification. Sequencing the 16S rRNA gene and IRS of these isolates resulted in highest similarity to M. gypis. The failure to achieve a higher score or concordant species identification might be the result of a possible absence of an adequate MSP in the database. However, these isolates may also belong to a yet-to-be described Mycoplasma species with high similarities to M. gypis, given that avian Mycoplasma isolates that cannot be assigned to a described species are found regularly.9,12,13 A further explanation may be the absence of sufficient protein signal in order to create an ideal spectrum to be compared to the database, a situation that has been reported if only small amounts of microbial material can be harvested for protein extraction. 4 To improve protein density, additional processing steps are described for fastidious bacteria, 19 a process not used in our study. An optimized preparation method might improve the results for these isolates.
Our avian mycoplasma spectral database was reliable for the identification of avian Mycoplasma isolates, given that all isolates with a score of ≥2.0 were identified correctly. The results showed high concordance between the MALDI-TOF MS and the molecular species identification. Even though the MALDI-TOF MS method is based on culture rather than on direct detection within a clinical sample, it offers the possibility for species identification using just one generic test. This is especially helpful if there is no specific anticipated Mycoplasma species and in the investigation of the presence of Mycoplasma spp. in a flock or a population or in studies on free-ranging bird populations. Further expansion of the database and optimization of the pre-analytic protocols will improve the accuracy of identifications of avian Mycoplasma spp. by MALDI-TOF MS.
Footnotes
Acknowledgements
We thank Dr. Dirk Enderlein for his technical advice regarding molecular methods. We also thank Ralf Doerr and the technical staff of Institute of Hygiene and Infectious Diseases of Animals, Justus Liebig University Giessen, for technical assistance. Moreover, we thank AniCon Labor (Höltinghausen, Germany), as well as Moorgut Kartzfehn von Kameke (Bösel, Germany), and Janet M. Bradbury and her team, University of Liverpool, for providing us with Mycoplasma isolates, and the Institute for Terrestrial and Aquatic Wildlife Research (ITAW) of the University of Veterinary Medicine Hannover Foundation for sampling free-ranging pheasants for mycoplasma examination.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
