Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Antimicrobial susceptibility and multilocus sequence typing of Mycoplasma capricolum subsp. capricolum

  • Juan Tatay-Dualde,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Miranda Prats-van der Ham,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Christian de la Fe ,

    cdelafe@um.es

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Ana Paterna,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Antonio Sánchez,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Juan Carlos Corrales,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Antonio Contreras,

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

  • Sebastiana Tola,

    Affiliation Istituto Zooprofilattico Sperimentale della Sardegna “G. Pegreffi”, Sassari, Italy

  • Ángel Gómez-Martin

    Affiliation Department of Animal Health, Faculty of Veterinary Sciences, Regional Campus of International Excellence “Campus Mare Nostrum”, Universidad de Murcia, Campus de Espinardo s/n., Murcia, Spain

Antimicrobial susceptibility and multilocus sequence typing of Mycoplasma capricolum subsp. capricolum

  • Juan Tatay-Dualde, 
  • Miranda Prats-van der Ham, 
  • Christian de la Fe, 
  • Ana Paterna, 
  • Antonio Sánchez, 
  • Juan Carlos Corrales, 
  • Antonio Contreras, 
  • Sebastiana Tola, 
  • Ángel Gómez-Martin
PLOS
x

Abstract

Mycoplasma capricolum subsp. capricolum is one of the causative agents of contagious agalactia (CA). Nevertheless, there is still a lack of information about its antimicrobial susceptibility and genetic characteristics. Therefore, the aim of this work was to study the antimicrobial and genetic variability of different Mycoplasma capricolum subsp. capricolum field isolates. For this purpose, the growth inhibition effect of 18 antimicrobials and a multilocus sequence typing (MLST) scheme based on five housekeeping genes (fusA, glpQ, gyrB, lepA and rpoB) were performed on 32 selected field isolates from Italy and Spain.The results showed a wide range of growth inhibitory effects for almost all the antimicrobials studied. Macrolides presented lower efficacy inhibiting Mcc growth than in previous works performed on other CA-causative mycoplasmas. Erythromycin was not able to inhibit the growth of any of the studied strains, contrary to doxycycline, which inhibited the growth of all of them from low concentrations. On the other hand, the study of the concatenated genes revealed a high genetic variability among the different Mcc isolates. Hence, these genetic variations were greater than the ones reported in prior works on other mycoplasma species.

Introduction

Mycoplasma capricolum subsp. capricolum (Mcc) is one of the causative agents of contagious agalactia (CA) [1]. In some regions such as France [2] or the Canary Islands (Spain) it is the second most frequently isolated CA-causing agent, and its presence has been related with severe outbreaks of this disease [3]. Furthermore, Mcc has been recently detected in a human case in which the patient displayed symptoms of septicaemia [4, 5].

Antimicrobial therapy is one of the main tools used to control mycoplasmoses, including CA. Thus, macrolides, fluoroquinolones and tetracyclines are widely employed in CA endemic areas [6]. In this context, it is noticeable that the frequent use of these antimicrobials has been related with an increase of in vitro antimicrobial resistance of the bovine pathogen M. bovis [7]. Regarding CA, previous studies have described antimicrobial resistances on field isolates of Mycoplasma (M.) agalactie, M. mycoides subsp. capri and M. putrefaciens [811]. Notwithstanding, there is a lack of information concerning the antimicrobial susceptibility of Mcc, which has only been assessed in 8 field isolates against 6 antimicrobials [12].

Regarding the control of mycoplasmoses, the assessment of the genetic diversity among different isolates has proved to be useful to epidemiological studies, by means of which infectious diseases are monitored. For this purpose, several typing techniques have been developed so as to compare isolates, including multilocus sequence typing (MLST). This technique utilizes sequences of internal fragments from housekeeping genes in order to characterize bacterial isolates, and it seems to be the most appropriate technique to analyze evolutionary relationships amongst mycoplasma isolates [1316]. Hence, in previous studies with other mycoplasmas, a great genetic diversity between isolates from the same area (country) [13, 17] or even from the same farm [16] was observed. As for the members of the M. mycoides cluster, to which Mcc belongs, a scheme based on five housekeeping genes (fusA, glpQ, gyrB, lepA and rpoB) has been described [18]. Nevertheless, apart from the isolates analyzed in that study [18], there are no previous works in which this technique has been used to analyze the genetic diversity of different Mcc field isolates.

Taking into account the previously described facts, the aim of the present study was to characterize different Mcc isolates from a CA-endemic area by antimicrobial susceptibility testing and molecular typing in order to determine the suitability of different antimicrobial treatments against this pathogen and its genetic profile, which affects the monitoring and control of this disease.

Material and methods

Mycoplasma isolates

32 Mcc isolates were investigated, including the reference strain of Mcc, California Kid (CK) (NCTC 10154). Epidemiological data of the studied isolates and CK are shown in Table 1. All isolates were identified by previously described PCR protocols targeting the genes glk and fusA [18, 19].

Microdilution susceptibility test

For the antimicrobial susceptibility tests, Mcc isolates were grown in PPLO medium (Difco, France), supplemented with Equine Serum (HyClone, USA) without any antimicrobial [20], during 24 hours until reached the stationary growing phase. Subsequently, the concentration of the inocula was determined by the simple quantification method of viable mycoplasmas [21], and then diluted with sterile saline solution up to the desired concentration.

Antimicrobial effect was assessed by determining the minimum inhibitory concentration (MIC), following previously described recommendations [22]. We evaluated the effect of 18 antimicrobials: spectinomycin (Sigma-Aldrich, USA), kanamycin (Sigma-Aldrich, USA), gentamycin (Sigma-Aldrich, USA), neomycin (Sigma-Aldrich, USA) and streptomycin (Sigma-Aldrich, USA) in a range from 128 μg/ml to 16 μg/ml; ciprofloxacin (Fluka, USA), enrofloxacin (Fluka, USA), marbofloxacin (Tokyo Chemical Industry, Japan), danofloxacin (Fluka, USA), moxifloxacin (Cayman Chemical Company, USA), norfloxacin (Sigma-Aldrich, USA), doxycycline (Sigma-Aldrich, USA), tylosin (Fluka, USA) and tilmicosin (Molekula, UK), which were studied in a range from 12.8 μg/ml to 0.006 μg/ml; and erythromycin (Fluka, USA), spiramycin (Sigma-Aldrich, USA), clindamycin (Cayman Chemical Company, USA) and lincomycin (Sigma-Aldrich, USA), which were evaluated in a range from 12.8 μg/ml to 0.8 μg/ml. The concentration range of these antimicrobials was selected considering data published in previous studies [810]. Different concentrations of antimicrobials were obtained by serial two-fold dilutions from the highest to the lowest concentration.

96-wells microtitre plates were used to carry out this method. In each well 150 μl of PH medium with 0.007% of phenol red, 25.6 μl of each antimicrobial dilution and 25 μl of the diluted inocula were added in order to reach a concentration of 103−105 CFU/ml. Furthermore, a well without antimicrobial was used as a positive control and a well without neither antimicrobial nor inoculum as a negative control. Plates were incubated at 37°C in a humid atmosphere with 5% CO2 and they were read when a change of color appeared in the positive control due to the acidification of the medium. These analyses were performed in duplicates. MIC is defined as the minimum concentration of antimicrobial at which no growth is observed. Since there are no established breakpoints for MIC in Mcc, results were evaluated according to the breakpoints reported for other mycoplasma species [23, 24] or by the European Committee on Antimicrobial Susceptibility Testing [25] (Table 2). Furthermore, the antimicrobial concentration at which the 50% of strains were inhibited (MIC50) and the concentration at which the 90% of strains were inhibited (MIC90) were also calculated.

thumbnail
Table 2. Minimal inhibitory concentration (MIC) range, MIC50, MIC90 for Mycoplasma capricolum subsp. capricolum field isolates for 18 antimicrobials and their MIC against the reference strain.

https://doi.org/10.1371/journal.pone.0174700.t002

Multilocus sequence typing

Mycoplasmal genomic DNA was obtained using a classical phenol-chloroform extraction from 25 ml volume of broth culture of each Mcc isolate [27].

The MLST study was based on the previously developed scheme for the M. mycoides cluster typing [18]. Hence, the five housekeeping genes proposed in that study (fusA, glpQ, gyrB, lepA and rpoB) were analyzed. PCRs were performed with Phusion High-Fidelity PCR Kit (Thermo Scientific, USA). They were done in a final volume of 25 μl containing 0.5 μM of each primer, 0.1 mM dNTPs, 1X Phusion Buffer and 0.02 units of Phusion DNA Polymerase. Amplifications were performed applying the following incubation conditions: one cycle of 95°C for 30 s; 35 cycles of 95°C for 10 s, variable annealing temperatures (detailed in Table 3) for 1 min, 72°C for 1 min and a final incubation at 72°C for 10 min. These reactions were performed in an i-Cycler (Biorad, USA) thermal cycler. PCR products were subjected to electrophoresis in 1% agarose gels containing 0.005% of RedSafe (iNtRON Biotechnology, Korea) DNA staining, and visualized under UV light. Subsequently, PCR products were purified using QIAquick PCR Purification Kits (Qiagen, The Netherlands). After that, PCR products were sequenced at the molecular biology service of the University of Murcia and compared in both forward and reverse directions with the same primers as used for the PCR. The sequences obtained in this study for each MLST loci were trimmed to the same size as previously described (16), and are detailed in S1 Table. Moreover, the sequences of these five loci from the reference strain for Mcc, California Kid (CP000123), were retrieved from GenBank.

thumbnail
Table 3. Multilocus sequence typing (MLST) loci, PCR details and variability of each locus.

https://doi.org/10.1371/journal.pone.0174700.t003

Sequence analyzes were conducted using MEGA6 [28]. Pairwise distances were calculated as the proportion of non-matching sites between pairs of sequences. Phylogenetic trees were built using the neighbor-joining algorithm based on pairwise dissimilarities between strains [29]. The ‘‘pairwise gap block correction” option was selected with a ‘‘minimal length for gap blocks” of 1 nucleotide so that all consecutive gaps were considered a single event. Bootstrap analyses with 1000 replicates were performed.

Results

Antimicrobial MICs

MIC range, MIC50, MIC90 and MIC values for the studied field isolates and the reference strain of Mcc are provided in Table 2. Moreover, MIC values for each of the studied strains are shown in S2 Table.

All the Mcc isolates showed high MIC values for the studied aminoglycosides. Hence, spectinomycin, kanamycin and streptomycin returned minimum MIC values of at least 16 μg/ml, whereas gentamycin and neomycin yielded 32 μg/ml. Moreover, the lowest MIC50 value obtained was 32 μg/ml and the MIC90 values were over 128 μg/ml.

On the other hand, the MIC50 values obtained for quinolones were lower than 0.4 μg/ml, except for norfloxacin which showed the highest value (0.8 μg/ml) within this group. MIC90 values were also lower than 0.5 μg/ml, except for ciprofloxacin (1.6 μg/ml) and norfloxacin (1.6 μg/ml). However, 2 field isolates presented MIC values over 1 μg/ml for all quinolones except for moxifloxacin. Besides, 2 other isolates were resistant only to ciprofloxacin (>12.8 μg/ml) and 24 of the studied isolates showed high MIC values (> 0.5 μg/ml) for norfloxacin.

Doxycycline was the antimicrobial which showed the lowest MIC results in all the studied parameters (MIC range, MIC50 and MIC90). Therefore, no resistant strains to this antimicrobial group were found.

Finally, the group of macrolides and lincosamides showed great variations between the different antimicrobials assessed. Tylosin and tilmicosin were the most effective ones within this group. However, 6 isolates returned high MIC values (> 1 μg/ml) for tylosin and 4 for tilmicosin (24). Furthermore, the MIC90 for both antimicrobials was >12 μg/ml. On the other hand, clindamycin and lincomycin showed a great diversity of values for the different isolates studied, with a MIC50 of 0.2 and 0.8 μg/ml, respectively, and a MIC90 >12.8 for both antimicrobials. Erythromycin was not effective to inhibit the growth of any of the assessed isolates of Mcc (MIC >12.8 μg/ml), except for one which MIC was 3.2 μg/ml.

MLST analysis

The sequences of five housekeeping genes were analyzed, comprising a total of 3726 base pairs. Twenty five Mcc isolates were partially sequenced and compared to the CK strain, which data were retrieved from the GenBank database (CP 000123). All these sequences were aligned and trimmed to the same size, providing gene fragments ranging from 609 to 912 bp (Table 3), and their GenBank entrance numbers are available at the S1 Table. The number of single nucleotide polymorphisms (SNP) observed within each individual locus as well as in the concatenated sequences of the five loci are indicated in Table 3. Thus, from a total of 3726 base pairs, 92 were variable sites and 71 were informative sites.

Regarding the study of the five concatenated sequences, pairwise distance analyses are shown in S1 Fig. The maximum divergence observed between the studied isolates is 0.9%, and the mean intra-group distance was 0.5%. Phylogenetic analysis using the neighbor-joining method on the concatenated sequences is displayed in Fig 1. Hence, the resultant phylogenetic tree did not show any apparent correlation between the phylogenic origin of the strains and their anatomic source (auricular swab or mastitic milk) or their isolation date. Besides, there were not reported phylogenetic differences between sheep and goat isolates. In fact, both groups of strains shared their phylogenetic origin and showed the same genetic variability. Nevertheless, the majority of field isolates coming from Italy were visibly separated from the other strains (bootstrap value of 76%). However, the 78106 and 874 isolates which were from Italy too were clearly separated from this group, being most related to isolates from the Canary Islands (bootstrap 55%). Additionally, whilst all the Spanish isolates from 2015 appeared next to each other in the resulting phylogenetic tree, this kind of correlation could not observed between the isolates from other years.

thumbnail
Fig 1. Phylogenetic tree inferred from the concatenated sequences of the five housekeeping genes, corresponding to the 31 field isolates analyzed in this study and California Kid strain.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 resamplings and values over 50% are displayed. *, **, ***: Isolates from the same herd.

https://doi.org/10.1371/journal.pone.0174700.g001

Since previous studies reported phylogenetic differences between isolates coming from the same farms [16, 17], our study included the study of three farms from which more than one isolate were retrieved. Phylogenetic differences appeared between different isolates coming from one of these farms, but only when isolates were retrieved at different years.

Concerning individual gene study, the phylogenetic trees based on single locus sequences are shown in S2S6 Figs. Thus, the tree based on fusA sequences (S2 Fig) showed three main branches, one of which grouped the Italian isolates, similarly to the concatenated tree. Evolutionary distance trees based on glpQ and gyrB trees (S3 and S4 Figs) were incongruent with the phylogeny based on concatenated sequences. The tree based on lepA sequences (S5 Fig) showed two main branches. One of them could be related to the first branches displayed on the concatenated genes tree (Fig 1) (until the California Kid branch), and the other main branch might be associated with the last divisions of this tree. Last, the study of the rpoB sequences (S6 Fig) resulted in a tree with three main branches. One of them could be related to the Italian isolates, and another branch was able to group the isolates from the Canary Islands together with strain 874 (Italy), in the same way as it occurred in the concatenated genes tree.

Discussion

Antimicrobial MIC studies are an essential tool to evaluate mycoplasmal antimicrobial resistances [22], and have also proved to be useful to monitor their evolution over time as the massive use of antimicrobials in infected flocks has been related with an increase of in vitro resistances [7]. Besides, to our knowledge, this is the first phylogenetic analysis performed on Mcc since an MLST scheme for the M. mycoides cluster was developed [18]. In this sense, epidemiological investigations, which are essential for disease surveillance and control, need suitable typing tools to carry out phylogenetic studies with enough resolution. Hence, sequenced-based typing methods have demonstrated to be technically simple, robust and portable [30].

In the present study, enrofloxacin, marbofloxacin, danofloxacin and moxifloxacin (fluoroquinolones), and doxycycline (tetracycline) were the most effective antimicrobials inhibiting the growth of the different Mcc isolates. Thus, these antimicrobials showed similar inhibitory concentrations as reported in previous studies carried out on other CA-causing mycoplasmas [8, 9, 11, 12]. Concerning the susceptibility breakpoints proposed in previous works on different mycoplasma species [23, 24] (Table 2), two of our isolates must be considered as resistant to enrofloxacin. Moreover, since the proposed breakpoint for danofloxacin was 0.25 μg/ml [24], four of the studied isolates must be also considered as resistant to this antimicrobial.

Our study exhibits a high value of MIC90 for ciprofloxacin and norfloxacin. In this sense, 4 isolates showed high MIC values for ciprofloxacin and 2 isolates (Cap20 and Cap23) yielded MIC values higher than 1 μg/ml for all the studied quinolones. Therefore, doxycycline is the only antimicrobial effective in inhibiting the growth of Mcc at low concentrations for all the assessed isolates. These data, together with the available pharmacokinetic studies [31] and successful clinical applications of this antimibrobial [5], highlight the advantages of using doxycycline to effectively treat Mcc infections.

Regarding macrolides and lincosamides, taking into consideration the proposed susceptibility breakpoints for tylosin and clindamycin [23, 32], 6 of our isolates were resistant to tylosin and 9 were resistant to clindamycin. These results are in contrast with prior data obtained for M. agalactiae)[9], where clindamycin was the most effective antimicrobial analyzed. In the present work, MIC ranges for tylosin and tilmicosin were wide and thus, their MIC90 values were >12.8 μg/ml. These values were noticeably higher than the values shown in previous studies [8, 9]. Besides, the MIC values obtained for erythromycin and spiramycin were also higher than the outcomes of previous studies on other members of the M. mycoides cluster [8, 12]. These remarkable differences could be explained by the extensive use of macrolides such as tylosin and erythromycin to treat CA [6], explaining why Mcc strains retrieved between years 2003 and 2004 were more sensitive to antimicrobials [12] than the isolates assessed in this work, as it has been proved in other mycoplasma species such as M. bovis [7]. In fact, when comparing the MIC results of CK with those reported for the rest of the studied field isolates, CK always showed equal or lower MIC values than the MIC50 value for each antimicrobial. This finding highlights an increase of MICs over time, as CK is an elder strain.

None of the assessed aminoglycosides was able to inhibit the growth of any Mcc isolate with a concentration lower than 16 μg/ml. Moreover, their MIC50 and MIC90 values showed that these antimicrobials are not effective against this pathogen. These results are in agreement with those described in previous studies performed on other mycoplasma species [8, 10, 11, 33]. Hence, our work supports the previously proposed hypothesis about an intrinsic resistance of the mycoplasmas of the M. mycoides cluster to aminoglycosides [8].

Regarding the MIC results together with the MLST analysis, as exposed in S2 Table, our study showed that epidemiologically related strains did not always present similar MIC values. In fact, some of the strains belonging to the main phylogenetic group (Cap24, Cap25, Cap23, Cap22, Cap20, Cap19, Cap18, Cap16 and Cap17), showed low MICs for macrolides, whereas other isolates with the same phylogenetic profile proved to be resistant to these antimicrobials. Hence, our results are more in concordance with those obtained in previous studies with M. agalactiae [9] than those reported for M. bovis [34], in which molecular tying results and MIC values were closely related.

One of the most remarkable results of the present phylogenetic study was the great genetic variability showed between the 32 Mcc strains assessed. Although our mean intra-group distance was lower than in previous studies conducted on other mycoplasma species [15], there were only two groups of indistinguishable strains which had not been isolated from the same herds (Cap4, Cap6 and Cap7; Cap1 and Cap2). Hence, the rest of the isolates showed significant differences in at least one of the studied genes, contrary to the genetic homogeneity reported for M. agalactiae strains also isolated in Spain [35]. In this context, the genetic variability showed by Mcc at the present work is in concordance with a prior work with Mycoplasma mycoides subsp. capri [17] which is closely related to Mcc and also reported a substantial genetic diversity. This fact could be explained by the proved genetic transmission which happens amongst different mycoplasma species when they coinfect a single host [36]. In this sense, Mcc and Mycoplasma mycoides subsp. capri are mainly isolated from goats, which may be infected by more than one CA-causing mycoplasma species, as opposed to sheep, which are mainly affected by M. agalactiae. Therefore, goats have shown to be prone to mixed mycoplasmal infections [1], where horizontal gene transfer has demonstrated to effectively contribute to individual evolutionary differences.

Moreover, our results also suggested a correlation between the genetic profile and geographical origin of the studied strains, which has not been described for other mycoplasma species such as M. agalactiae [13]. Thus, our phylogenetic tree revealed a common branch for the majority of Italian isolates, which were clearly different from the Spanish isolates. Nevertheless, although a certain evolutionary approach was observed in the isolates from the Canary Islands (except for Cap2), it was not possible to discern a clear correlation between the evolutionary distances and geographical origin between the Spanish field isolates. This could be due to the fact that Mcc strains from the mainland (Iberian Peninsula) might have evolved together, while clonal expansion to more distant territories such as the Canary Islands or Italy may be less evident, resulting in greater similarities between isolates from these areas. However, despite the relationship suggested in this study between the geographical origin (Canary Islands, Italy and the Spanish mainland) and genetic evolution, further studies analyzing more isolates are necessary to gain a better understanding of the evolution of this pathogen in CA-endemic areas.

In connection with previous studies performed on other mycoplasma species which MLST schemes showed genetic diversity between isolates coming from the same flocks [16, 17], in the present work, different strains isolated from the same herds were selected in order to compare them. In this context, when isolates were taken at different years, phylogenetic differences were shown, highlighting the genetic variability of this pathogen. However, no differences were observed in any of the genes studied among isolates which had been obtained from the same herds during the same year. This result is contrary to the outcomes reported for M. arginini [16] and Mycoplasma mycoides subsp. capri [17], which showed significant differences between the MLST analyses of different strains isolated in the same conditions. Nevertheless, those studies [16, 17] analyzed a greater number of herds with multiple isolates and therefore, further studies are necessary to determine if MLST analysis is also able to demonstrate genetic diversity in Mcc isolates within farms, as suggested by our results.

Last, the individual study of fusA, glpQ and gyrB showed that these genes presented less informative sites than the other two loci assessed (Table 3), which differs from the results previously reported for glpQ and gyrB [18]. However, this fact could explain why the phylogenetic trees of glpQ and gyrB (S3 and S4 Figs) are less consistent with the concatenated genes tree. Accordingly to the outcomes of the MLST scheme for the M. mycoides cluster [18], in our study the fusA gene showed the least informative sites but its phylogenetic tree (S2 Fig) managed to partly differentiate the main groups displayed at the concatenated genes tree. However, since lepA and rpoB were the genes with the most informative sites, their phylogenetic trees (S5 and S6 Figs) were capable of distinguishing more branches in concordance with the joint study of the five housekeeping genes. Therefore, these genes are the ones providing more useful information when performing genetic typing analysis of Mcc. Thus, they should be chosen when only one gene is to be studied.

Conclusions

Doxycycline was the most effective antimicrobial inhibiting Mcc field isolates in vitro; therefore, it should be one of the antimicrobials of choice to treat CA caused by Mcc. Wide MIC ranges and the increment of MIC values observed in comparison with previous studies emphasize the importance of analyzing each isolate individually so as to select the most adequate antimicrobial in each case.

The MLST scheme applied in the present study confirms its suitability to assess the genetic variability of Mcc isolates from separately locations within a CA-endemic area. Moreover, the great genetic diversity observed in this work between the different isolates studied suggests that Mcc could present higher genetic diversity than other mycoplasma species.

Supporting information

S1 Fig. Pairwaise distance analysis of concatenated protein-coding sequences from 32 strains calculated as the proportion of non-matching sites between pairs of sequences.

https://doi.org/10.1371/journal.pone.0174700.s001

(PDF)

S2 Fig. Evolutionary tree based on fusA sequences.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 replications and values over 50% are displayed.

https://doi.org/10.1371/journal.pone.0174700.s002

(PDF)

S3 Fig. Evolutionary tree based on glpQ sequences.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 replications and values over 50% are displayed.

https://doi.org/10.1371/journal.pone.0174700.s003

(PDF)

S4 Fig. Evolutionary tree based on gyrB sequences.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 replications and values over 50% are displayed.

https://doi.org/10.1371/journal.pone.0174700.s004

(PDF)

S5 Fig. Evolutionary tree based on lepA sequences.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 replications and values over 50% are displayed.

https://doi.org/10.1371/journal.pone.0174700.s005

(PDF)

S6 Fig. Evolutionary tree based on rpoB sequences.

The tree was constructed using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 1000 replications and values over 50% are displayed.

https://doi.org/10.1371/journal.pone.0174700.s006

(PDF)

S1 Table. GenBank accession numbers of sequences obtained used in this study.

https://doi.org/10.1371/journal.pone.0174700.s007

(PDF)

S2 Table. MIC results for each studied strain.

https://doi.org/10.1371/journal.pone.0174700.s008

(PDF)

Author Contributions

  1. Conceptualization: JTD CDLF AGM.
  2. Funding acquisition: CDLF AC AS JCC.
  3. Investigation: JTD MPvH AP ST JCC AC AS CDLF AGM.
  4. Project administration: AC CDLF.
  5. Resources: JTD MPvH CdlF ST AGM.
  6. Supervision: CdlF AC AGM.
  7. Writing – original draft: JTD MPvD CdlF AGM.
  8. Writing – review & editing: JTD MPvH AP ST JCC AC AS CdlF AGM.

References

  1. 1. Bergonier D, Berthelot X, Poumarat F. Contagious agalactia of small ruminants: current knowledge concerning epidemiology, diagnosis and control. Revue Scientifique Et Technique De L Office International Des Epizooties. 1997;16(3):848–73.
  2. 2. Chazel M, Tardy F, Le Grand D, Calavas D, Poumarat F. Mycoplasmoses of ruminants in France: recent data from the national surveillance network. Bmc Vet. Res. 2010;6.
  3. 3. De la Fe C, Gutierrez A, Poveda JB, Assuncao P, Ramirez AS, Fabelo F. First isolation of Mycoplasma capricolum subsp capricolum, one of the causal agents of caprine contagious agalactia, on the island of Lanzarote (Spain). Vet. J. 2007;173(2):440–2. pmid:16324858
  4. 4. Seersholm FV, Fischer A, Heller M, Jores J, Sachse K, Mourier T, et al. Draft Genome Sequence of the First Human Isolate of the Ruminant Pathogen Mycoplasma capricolum subsp. capricolum. Gen. Announ. 2015;3(3).
  5. 5. Heller M, Schwarz R, Noe G, Jores J, Fischer A, Schubert E, et al. First human case of severe septicaemia associated with Mycoplasma capricolum subsp. capricolum infection JMM Case Reports; 2015.
  6. 6. Gomez-Martin A, Amores J, Paterna A, De la Fe C. Contagious agalactia due to Mycoplasma spp. in small dairy ruminants: Epidemiology and prospects for diagnosis and control. Vet. J. 2013;198(1):48–56. pmid:23759248
  7. 7. Ayling RD, Rosales RS, Barden G, Gosney FL. Changes in antimicrobial susceptibility of Mycoplasma bovis isolates from Great Britain. Vet. Rec. 2014;175(19).
  8. 8. Antunes NT, Tavio MM, Assuncao P, Rosales RS, Aquili V, de la Fe C, et al. In vitro susceptibilities of field isolates of Mycoplasma mycoides subsp mycoides large colony type to 15 antimicrobials. Vet Microbiol. 2007;119(1):72–5. pmid:16978803
  9. 9. de Garnica ML, Rosales RS, Gonzalo C, Santos JA, Nicholas RAJ. Isolation, molecular characterization and antimicrobial susceptibilities of isolates of Mycoplasma agalactiae from bulk tank milk in an endemic area of Spain. J. Appl. Microbiol. 2013;114(6):1575–81. pmid:23445345
  10. 10. Paterna A, Sanchez A, Gomez-Martin A, Corrales JC, De la Fe C, Contreras A, et al. Short communication: In vitro antimicrobial susceptibility of Mycoplasma agalactiae strains isolated from dairy goats. J. Dairy Sci. 2013;96(11):7073–6. pmid:24035026
  11. 11. Antunes NT, Tavio MA, Mercier P, Ayling RD, Al-Momani W, Assuncao P, et al. In vitro susceptibilities of Mycoplasma putrefaciens field isolates. Antimicrob. Agents Ch. 2007;51(9):3452–4.
  12. 12. Al-Momani W, Nicholas RAJ, Janakat S, Abu-Basha E, Ayling RD. The in vitro effect of six antimicrobials against Mycoplasma putrefaciens, Mycoplasma mycoides subsp mycoides LC and Mycoplasma capricolum subsp capricolum isolated from sheep and goats in jordan. Trop. Anim. Health Pro. 2006;38(1):1–7.
  13. 13. McAuliffe L, Gosney F, Hlusek M, de Garnica ML, Spergser J, Kargl M, et al. Multilocus sequence typing of Mycoplasma agalactiae. J. Med. Microbiol. 2011;60(6):803–11.
  14. 14. Fischer A, Shapiro B, Muriuki C, Heller M, Schnee C, Bongcam-Rudloff E, et al. The Origin of the 'Mycoplasma mycoides Cluster' Coincides with Domestication of Ruminants. Plos One. 2012;7(4):6.
  15. 15. Manso-Silvan L, Dupuy V, Lysnyansky I, Ozdemir U, Thiaucourt F. Phylogeny and molecular typing of Mycoplasma agalactiae and Mycoplasma bovis by multilocus sequencing. Vet. Microbiol. 2012;161(1–2):104–12. pmid:22841405
  16. 16. Olaogun OM, Kanci A, Barber SR, Tivendale KA, Markham PF, Marenda MS, et al. Genetic diversity of Mycoplasma arginini isolates based on multilocus sequence typing. Vet. Microbiol. 2015;180(1–2):123–8. pmid:26264760
  17. 17. Tatay-Dualde J, Prats-van der Ham M, de la Fe C, Gómez-Martín A, Paterna A, Corrales JC, et al. Multilocus sequence typing of Mycoplasma mycoides subsp. capri to assess its genetic variability in a contagious agalactia endemic area. Vet. Microbiol. 2016. p. 60–4.
  18. 18. Manso-Silvan L, Perrier X, Thiaucourt F. Phylogeny of the Mycoplasma mycoides cluster based on analysis of five conserved protein-coding sequences and possible implications for the taxonomy of the group. Int. J. Syst. Evol. Micr. 2007;57:2247–58.
  19. 19. Woubit S, Manso-Silvan L, Lorenzon S, Gaurivaud P, Poumarat F, Pellet MP, et al. A PCR for the detection of mycoplasmas belonging to the Mycoplasma mycoides cluster: Application to the diagnosis of contagious agalactia. Mol. Cell. Probe. 2007;21(5–6):391–9.
  20. 20. Gomez-Martin A, De la Fe C, Amores J, Sanchez A, Contreras A, Paterna A, et al. Anatomic location of Mycoplasma mycoides subsp capri and Mycoplasma agalactiae in naturally infected goat male auricular carriers. Vet. Microbiol. 2012;157(3–4):355–62. pmid:22284907
  21. 21. Albers AC, Fletcher RD. Simple method for quantitation of viable mycoplasmas. Appl Environmen. Microbiol. 1982;43(4):958–60.
  22. 22. Hannan PCT. Guidelines and recommendations for antimicrobial minimum inhibitory concentration (MIC) testing against veterinary mycoplasma species. Vet. Res. 2000;31(4):373–95. pmid:10958240
  23. 23. Hannan PCT, Windsor GD, deJong A, Schmeer N, Stegemann M. Comparative susceptibilities of various animal-pathogenic mycoplasmas to fluoroquinolones. Antimicrob. Agents Ch. 1997;41(9):2037–40.
  24. 24. Gerchman I, Levisohn S, Mikula I, Lysnyansky I. In vitro antimicrobial susceptibility of Mycoplasma bovis isolated in Israel from local and imported cattle. Vet Microbiol. 2009;137(3–4):268–75. pmid:19250777
  25. 25. European Committee on Antimicrobial Susceptability Testing (EUCAST). Breakpoint tables for interpretation of MICs and zone diameters. Version 6.0, 2016: http://www.eucast.org.
  26. 26. Waites B, Lysnyansky I, Bébéar CM. Emerging antimicrobial resistance in mycoplasmas of human and animals. Caister Academic Press ed. Browning GF, Citti C, editors. Mollicutes. Molecular Biology and Pathogenesis. 2014.
  27. 27. Chen WP, Kuo TT. A simple and rapid method for the preparation of gram-negative bacterial genomic DNA. Nucleic Acids Research. 1993;21(9):2260–. pmid:8502576
  28. 28. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0. Mol. Biol. Evol. 2013;30(12):2725–9. pmid:24132122
  29. 29. Saitou N, Nei M. The neighbor-joining method—a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 1987;4(4):406–25. pmid:3447015
  30. 30. van Belkum A, Tassios PT, Dijkshoorn L, Haeggman S, Cookson B, Fry NK, et al. Guidelines for the validation and application of typing methods for use in bacterial epidemiology. Clin. Microbiol. Infec. 2007;13:1–46.
  31. 31. Jha VK, Jayachandran C, Singh MK, Jha HN. Pharmacokinetics of doxycycline in goat after intramuscular administration. Indian J. Anim. Sci. 1989;59(1):68–71.
  32. 32. Francoz D, Fortin M, Fecteau G, Messier SE. Determination of Mycoplasma bovis susceptibilities against six antimicrobial agents using the E test method. Vet. Microbiol. 2005;105(1):57–64. pmid:15607084
  33. 33. Ayling RD, Bisgaard-Frantzen S, March JB, Godinho K, Nicholas RAJ. Assessing the in vitro effectiveness of antimicrobials against Mycoplasma mycoides subsp mycoides small-colony type to reduce contagious bovine pleuropneumonia infection. Antimicrob. Agents Ch. 2005;49(12):5162–5.
  34. 34. Khalil D, Becker CAM, Tardy F. Monitoring the Decrease in Susceptibility to Ribosomal RNAs Targeting Antimicrobials and Its Molecular Basis in Clinical Mycoplasma bovis Isolates over Time. Microb Drug Resist. 2017.
  35. 35. Ariza-Miguel J, Rodriguez-Lazaro D, Hernandez M. Molecular Characterization of Mycoplasma agalactiae Reveals the Presence of an Endemic Clone in Spain. J. Clin. Microbiol. 2013;51(2):656–60. pmid:23224102
  36. 36. Nouvel LX, Sirand-Pugnet P, Marenda MS, Sagne E, Barbe V, Mangenot S, et al. Comparative genomic and proteomic analyses of two Mycoplasma agalactiae strains: clues to the macro- and micro-events that are shaping mycoplasma diversity. Bmc Genom. 2010;11:19.