Vagococcus sp. a porcine pathogen: molecular and phenotypic characterization of strains isolated from diseased pigs in Brazil

Introduction: Vagococcus spp. is known for its importance as a systemic and zoonotic bacterial pathogen even though it is not often reported in pigs. This is related to the pathogen misidentification due to the lack of usage of more discriminatory diagnostic techniques. Here we present the first report of Vagococcus lutrae in swine and the characterization of Vagococcus fluvialis and Vagococcus lutrae isolated from diseased animals. Methodology: Between 2012 and 2017, 11 strains with morphological characteristics similar to Streptococcus spp. were isolated from pigs presenting different clinical signs. Bacterial identification was performed by matrix assisted laser desorption ionization time of flight (MALDITOF) mass spectrometry and confirmed by 16S rRNA sequencing and biochemical profile. Strains were further genotyped by single-enzyme amplified fragment length polymorphism (SE-AFLP). Broth microdilution was used to determine the minimal inhibitory concentration of the antimicrobials of veterinary interest. Results: Ten strains were identified as V. fluvialis and one was identified as V. lutrae. The SE-AFLP analysis enabled the species differentiation with specific clustering of all V. fluvialis separately from the V. lutrae strain. Most strains presented growth in the maximum antibiotic concentration values tested for eight of the 10 analyzed antimicrobial classes. Conclusions: The observed resistance pattern can represent a problem for veterinary and producers in the treatment of diseases associated Vagococcus spp. in swine production. Vagococcus species may also be a risk for pig industry workers. The data described here will be of great value in further understanding the behavior of this pathogen in animal production.


Introduction
Vagococcus spp. is a Gram-positive, catalasenegative coccus that belongs to the family Enterococcaceae [1]. Phenotypically, this bacterium presents hemolytic colonies in blood agar incubated at 37ºC between 24 and 48 hours. These characteristics are similar to other systemic pathogens commonly associated with animal diseases [2,3].
In swine, there are few reports of V. fluvialis as disease causative. Teixeira et al. [19] isolated some strains from different body organs, suggesting that this agent may play a role in systemic and opportunistic infections in pigs. In contrast, V. lutrae has not yet been described as a disease-causing bacterium in swine, but has already been isolated from humans, evidencing its zoonotic potential [16].
Here we report the isolation and identification of V. fluvialis and, for the first time, V. lutrae from diseased pigs. Therefore, the objective of this work was to characterize strains of Vagococcus spp. isolated from diseased pigs by phenotypic and genotypic techniques, combined with epidemiological data, which are still rare information for this genus in swine production.

Bacterial strains
Eleven strains with morphological characteristics similar to Streptococcus spp. (small, pale, smooth and alpha-hemolytic colonies) were isolated between 2012 and 2017 from lung, joint, brain, and vaginal discharge samples of nine pigs from eight herds of four different states in Brazil (Paraná, Mato Grosso, Santa Catarina, and São Paulo). The clinical conditions observed in these animals were encephalitis, arthritis, pneumonia, or urinary tract infection. The isolates were plated on Columbia agar (Difco-BBL, Detroit, MI, USA) with 5% sheep blood and incubated for 24 h at 37°C. Isolated colonies were inoculated in Brain-Heat Infusion (BHI -Difco-BBL, Detroit, MI, USA) supplemented with 5% fetal bovine serum and incubated for 24 hours at 37 °C for further analysis.
Two replicates of each sample were placed in plate wells and two readings were made for each sample. The obtained spectra were compared to the manufacturer's library and the standard Bruker interpretative criteria were applied; scores ≥ 2.0 were accepted for species assignment and scores ≥ 1.7 but ≤ 2.0 for genus identification.

Identification by biochemical profile
The VITEK™ 2 automated identification system (bioMérieux, Hazelwood, MO, USA) was used to confirm bacterial strain identifications and to describe their biochemical profile. The inoculum preparation was performed according to the manufacturer's instructions and VITEK™ 2 GP ID card was used for identification of Gram-positive bacteria. Species identification was obtained using the VITEK™ 2 library system.

Antimicrobial resistance profile by minimal inhibitory concentration (MIC)
The minimal inhibitory concentration (MIC) was determined by broth microdilution technique using Sensititre™ Standard Susceptibility MIC Plate (TREK Diagnostic Systems/Thermo Fisher Scientific, Waltham, MA, USA) according to the standards defined in CLSI document VET01, fifth edition [21]. The inoculum was prepared with BHI broth supplemented with 5% fetal bovine serum and incubated at 37°C for 24 h. The culture was adjusted to a turbidity equal to the 0.5 McFarland standard solution confirmed by a spectrophotometer. This suspension was diluted in the order of 1:1000 in Mueller Hinton II broth, supplemented with 5% fetal bovine serum. From this final suspension, 50 μL were distributed into each well of the microplate and incubated at 37 ºC for 24 hours. The Streptococcus pneumoniae ATCC 49619 strain was used as quality control.
The minimum inhibitory concentrations (MIC) were assessed visually as the lowest concentration of antibiotics in the wells without growth (without button formation). As there is no breakpoint available for Vagococcus spp. in the CLSI documents, the antimicrobials resistance assessment is presented as MIC profiles with the discrimination of the respective MIC values distribution.

DNA extraction and partial sequencing of 16S rRNA gene
Genomic DNA was extracted according to the protocol described by Boom et al. [22], with previous enzymatic digestion with lysozyme (100 mg) and proteinase K (20 mg) at 37ºC for 60 minutes. Partial amplification of the 16S rRNA gene was performed following the protocol and primers described by Twomey et al. [23].
The amplicons were purified using Illustra GFX TM PCR DNA and Gel Band Purification kit (GE Healthcare) and sequenced by the Human Genome Research Center (University of São Paulo, Brazil). The obtained sequences were edited with the BioEdit Sequence Alignment Editor 7.2.6 software [24] and aligned by the ClustalW application [25] with the sequences available in the GenBank RefSeq database (National Center for Biotechnology Information; NCBI). Phylogenetic analysis was performed in the Mega 7.0.26 software [26] using the maximumlikelihood method and 500 bootstrap replicates for branch support statistical inference. The sequences obtained from this study were deposited in GenBank under accession numbers MG098233 -MG098243.

Molecular genotyping by SE-AFLP
All strains were genotyped by Single-Enzyme Amplified Fragment Length Polymorphism (SE-AFLP) using the HindIII enzyme (New England Biolabs, Beverly, MA, USA) according to McLauchlin et al. [27] protocol. Electrophoresis was performed using a 2% agarose gel at 90 V for 4 hours. The amplified products were stained with BlueGreen™ (LGC Biotecnologia) and compared to 100 bp DNA ladder (New England Biolabs).
The Bionumerics 7.6 (Applied Maths) software was used for cluster analysis. A dendrogram was generated using the Dice coefficient and the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) method. The cut-off point of 90% of genetic similarity was applied to determine and analyze the obtained clusters [28].

Bacterial identification
All studied strains were identified by MALDI-TOF MS, with score higher than 2.00, giving reliable species identification. Among studied strains, 10 were identified as V. fluvialis (V1 to V6 and V8 to V11) and one strain was identified as V. lutrae (V7). The identification was confirmed by VITEK™ 2 automated identification system and 10 strains were identified as V. fluvialis with more than 98% confidence. The V7 strain was not identified as V. lutrae by the VITEK™ system. However, the partial sequencing of the 16S rRNA gene was able to confirm both V. fluvialis and V. lutrae species (Figure 1).

Genotyping by single-enzyme amplified fragments length polymorphism (SE-AFLP)
The SE-AFLP fingerprint technique was able to identify seven different profiles (A1 -A7) with more than 90% genetic similarity. The largest genotypic profile identified was A1 formed by three strains, followed by A4 and A6 with two strains each, while the remaining profiles comprised only one strain each. The V. fluvialis strains presented high genetic heterogeneity -strains from the same animals (animals 4 and 8) did not cluster in the same genotypic profile, respectively. Furthermore, no associations between genotypes and epidemiological data, such as a geographic area, isolation site, year, and herd, were identified. However, at a level of 75% genetic similarity, the species differentiation was enabled with specific clustering of all V. fluvialis strains, and the V. lutrae strain separately.

Discussion
The results showed a high variability of the genotypic and phenotypic characteristics of the porcine Vagococcus spp. strains. Our results of SE-AFLP genotyping found no association between strain genotypes and epidemiological data. Teixeira et al. [19] used fingerprint techniques, such as Pulsed Field Gel Electrophoresis (PFGE) and Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE), to identify profiles of different Vagococcus species. The SDS-PAGE enabled the differentiation of V. fluvialis from Vagococcus salmoninarum species at 80% genetic similarity; the authors could also differentiate these two species by PFGE profiles. In our study, the SE-AFLP technique also enabled the differentiation of the V. fluvialis and V. lutrae species, and further demonstrated high genetic heterogeneity among V. fluvialis strains. We also observed genetic heterogeneity among strains isolated from the same geographic region, which indicates the circulation of genetically different strains of the same species within a specific area.
The genetic differences can be in accordance with variations of some biochemical reactions identified among the V. fluvialis strains. Variations in the biochemical tests adonitol, L-arabinose, salicin and sucrose were described in strains of V. fluvialis isolated from a mosquito (Culex quinquefasciatus) compared to other strains of the same species [29]. Regarding the biochemical profile obtained for V. lutrae, the reaction for alpha-galactosidase was negative, which differs from the previous report, in which the reaction for both alpha-and beta-galactosidase were positive for V. lutrae and negative for V. fluvialis [8].
The growth of V. fluvialis strains in the highest concentrations tested for several of the studied antimicrobials may suggest potential resistance to these chemotherapeutic agents. Previously, data using disk diffusion technique for a V. fluvialis strain showed that, of the 18 antimicrobials tested, the strain was considered resistant to kanamycin, nafcillin, norfloxacin, clindamycin and nalidixic acid [28]. In another study, Teixeira et al. [19] also described MIC data for this species, which were described as resistant to clindamycin, lomefloxacin and ofloxacin.
More recent data on bacterial endocarditis in humans have shown that Streptococcus viridans and Enterococcus were the bacteria most frequently associated with these cases, but a third group called rare Gram-positive coccus was also identified, in which V. fluvialis is included, evidencing its zoonotic importance [18]. The decrease in the sensitivity for several of the tested antimicrobials in the present study may increase the risk in case of zoonotic infection. Garcia et al. [16] described antimicrobial susceptibility data from a human case of V. lutrae infection, and susceptibility were reported for amoxicillin, ceftriaxone, gentamicin, erythromycin, rifampicin, clindamycin, doxycycline, and vancomycin; however, the technique used to obtain such results was not clearly described in the study. In contrast, the results obtained here for V. lutrae strain demonstrate low susceptibility for several of these antibiotics. Considering these 11 strains, both Vagococcus species were not inhibited at the highest concentrations of at least seven antimicrobials classes tested, indicating a multi-resistant profile.
Clinical signs and disease described here and associated with Vagococcus species in swine, are not commonly related to this agent, and the occurrence of these species may be underestimated due to the misidentification in commercial laboratories. Thus, when the clinician encounters these cases in a herd, there may not be a good treatment response due to the antimicrobial resistance profile observed in these strains, increasing animal mortality and losses to the producers.

Conclusions
The data obtained show that V. fluvialis and V. lutrae species may represent a risk, not only for swine production, but also for pig industry workers due to the zoonotic potential of Vagococcus spp. All these data are of great help in further understanding the epidemiologic behavior of these neglected pathogens in swine production systems.