Isolation and characteristics of multi-drug resistant Streptococcus porcinus from the vaginal secretions of sow with endometritis

Background Sow endometritis is a common disease in pig breeding farms after artificial insemination, which leads to gray-green vaginal secretions and decreased conception rates. It is important to perform an etiologic diagnosis for effective treatments and control of diseases. The aim of this study was to carry out a pathogenic detection in five specimens of vaginal secretions collected from sick pigs with endometritis, implement identification of the pathogens by phenotypic detection and 16 s rDNA sequence and phylogeny analysis, and determinate antibiotic susceptibility of the isolates. Results A Streptococcus strain was isolated and identified from all of the five specimens. The isolate was positive for Voges-Proskauer (V-P) and for the hydrolysis of arginine, esculin and myelin-associated glycoprotein (MAG). Acid formation was observed for sorbitol, mushroom sugar, sucrose, and glucose. The 16S rDNA sequence of the isolate possessed 99.93% similarity to that of Streptococcus porcinus. The phylogenetic analysis of 16S rDNA sequence showed that the isolate belonged to the same clade as the S. porcinus strains from humans, pigs, and other animals. The isolate exhibited multi-drug resistance to aminoglycosides, quinolones, macrolides and tetracyclines except being sensitive to some β- lactams such as penicillin G, cephalothin, cefazolin, cephradine and cefuroxime. Conclusions A S. porcinus isolate with multi-drug resistance was identified from vaginal secretions of sows with endometritis in one pig breeding farm, which suggests that the sow endometritis was caused by S. porcinus infection during artificial insemination. This study indicates that sensitive antibiotics such as penicillin G or some cephalosporins could be used for treatment of the diseases. In addition, the study hints that bacterial multi-drug resistance is a tough problem for disease treatment in pig farms.

hoof disease, and the remaining part is attributed to other pathologic conditions [4].
The species Streptococcus porcinus belonging to the genus Streptococcus is commonly associated with pyogenic infections, abortion and endocarditis of pigs, as well as genitourinary tract infection of women [7][8][9]. S. porcinus was first identified in pigs in 1984, and then isolated from bovine milk [7,8]. The organism has also been detected in the urogenital tracts of pregnant women [9][10][11]. S. porcinus is usually detected in pig tonsils. 19.5% of tonsils of the pigs in slaughterhouses were positive for the isolation of S. porcinus, which is just less than the prevalence of S. suis infections (53.7%) [12]. Swine S. porcinus causes pig lymph node abscess, throat abscess, pneumonia, arthritis, and endocarditis, as well as sow abortion [1,8,13,14], but there are rare reports on sow endometritis caused by S. porcinus.
In this study, to confirm the cause of a sow endometritis case from a pig breeding farm in Hebei Province, China, five vaginal secretion samples from the sick sows were detected in etiology by isolation and culture, phenotypic examinations and 16S rRNA sequence analysis of pathogen. In addition, antibiotic susceptibility assays of the isolate were carried out in order to instruct an effective treatment of the disease.

Results
Cultural and morphological characteristics of the isolate Bacteria with similar cultural and morphological characteristics were isolated in all vaginal secretion samples from five sick pigs. The bacterial isolates showed a poor growth on ordinary nutrient agar with fewer small tipsized colonies, but a good growth on blood agar and serum agar plates with off-white, smooth-surfaced, and regular-edged microcolonies. And transparent β-type hemolysis was present around the colonies on the blood agar plate. The isolates did not grow on the eosin methylene blue and MacConkey agar plates. The bacteria precipitated at the bottom of the tube containing serum broth after overnight incubation. The microscopy analysis revealed that the isolates appeared grampositive cocci in a single, pairs, or short-chain arrangement (Fig. 1). No similar cultural and morphological bacteria were detected from the vaginal secretion samples of all healthy pigs.

Physiological characteristics of the isolate
To differentiate whether the isolated bacteria was Streptococcus or Staphylococcus, we performed a catalase test. The bubbles observed for positive controls after dropping 3% hydrogen peroxide solution onto the staphylococcal reference strain. In contrast, no bubbles generated for the isolated bacteria and the negative control after adding 3% hydrogen peroxide onto the isolated bacteria and the streptococcal reference strain, which indicated that the isolated strains were catalase-negative. Further biochemical assays revealed that the isolates were positive for Voges-Proskauer (V-P) and for the hydrolysis of arginine, esculin and myelin-associated glycoprotein (MAG). Negative results were observed for the production of dipeptidyl aminopeptidase (DPP), polymethylgalacturonase (PMG) and for hydrolysis of temozolomide (TMZ). Acid formation exhibited for sorbitol, mushroom sugar, sucrose, and glucose, but did not for lactose, raffinose, xylose, and adonitol (Table 1).  The phylogenetic tree based on 16S rDNA sequences suggested that the isolated strain was in the same small branch with the human or animal S. porcinus strains in GenBank, which indicated a closer genetic relationship between the isolated strain and S. porcinus strains compared to other subspecies. The isolate was in different small branches with S. pseudoporcinus, but co-located in a large branch, indicating a closer genetic relationship with S. pseudoporcinus. However, it situated in distinct branches with S. uberis, S. canis, S. dysgalactiae, and S. parauberis strains, which displayed a distant genetic relationship ( Fig. 2).

Multi-drug resistance of the isolate
In order to clarify further biological characteristics of the isolate, we conducted antibiotics susceptibility assays. The results were shown in Table 2. Among tested 25 different antibiotics, the isolate was only sensitive to penicillin G, cephalothin, cefazolin, cephradine, and cefuroxime, but it was resistant to most antibiotics such as lincomycin, nitrofurantoin, rifampicin, aminoglycosides, quinolones, macrolides, and tetracyclines. The results demonstrated that the isolate is multi-drug resistant.
To improve the breeding and reproduction rate, artificial insemination is widely used in intensive pig farms. However, semen is easily contaminated with microbial pathogens carried by boar or contaminated during semen fluids collection and storage process [16,17]. In the current study, the clinical endometritis cases were likely correlated with contamination of semen and artificial insemination because symptoms of endometritis emerged just 3 to 5 days after insemination. Therefore, we carried out an etiological examination to determine its causes.
In order to ensure the accuracy of detection results, we combined phenotypic and molecular biological approaches to determine the causative agent of sow endometritis, and set up vaginal secretion samples of clinical healthy sows and gilts in the same farm as the case controls. In addition to bacteriological testing, PCR was employed for detecting classical swine fever virus (CSFV), porcine reproductive and respiratory syndrome virus (PRRSV), porcine circovirus type 2 (PCV2) and porcine circovirus type 3 (PCV3) in the sow vaginal secretions and did not find these viral nucleic acids (supplementary data). These results suggest that the sow endometritis was caused by a bacterial infection most likely due to contamination of semen and artificial insemination. Indeed, we isolated and identified a Gram-positive Streptococcus from the vaginal secretions of the sick sows based on Gram-staining, microscopy, catalase and biochemical assays. The isolated Streptococcus with βhemolysis in short chains arrangement under the microscope is negative to catalase, and positive to V-P test and to acid formation of sorbitol, mushroom sugar, sucrose and glucose, and hydrolysis of arginine and esculin. The most of these characteristics are consistent with those of Streptococcus porcinus in spite of exhibition of some variable reaction in the study [9,10].
16S rRNA gene sequences contain highly and moderately conserved sequence regions, and highly variable sequence region, which is a powerful molecular target and widely used in classification and identification of bacteria from clinical isolates [18][19][20]. 16S rRNA gene sequencing is a very useful method for differentiating S. porcinus from S. pseudoporcinus isolates, and for determining the taxonomic status of isolates from animal, human and dairy sources [21,22]. In order to further clarify the subspecies of the Streptococcus isolate in this study, we amplified the16S rRNA DNA sequence and conducted a comparative phylogenetic analysis. The results suggest that the Streptococcus isolate from vaginal secretions is likely to be Streptococcus porcinus because its 16S rRNA sequence is highly similar to those of the Streptococcus porcinus reference strains from both humans and pigs.
Bacterial resistance is an important global concern of public health especially in the last 20 years. The rapid increase in drug-resistant strains has caused a serious problem in clinical treatments of human and animal infectious diseases [23][24][25]. Resistant isolates of animal origin are very common and easily disseminated between animal and humans from contaminated food [24]. Wang et al. [25] reported that multidrug-resistant Salmonella Typhimurium isolates dramatically increased among the salmonella isolated from human, animal and retail milk between 2011 to 2016. Resistances of S. porcinus for erythromycin, minocycline, sulfamethoxazoletrimethoprim, streptomycin and tetracycline were described in previous study [26]. In this study, we observed that the S. porcinus strain had multi-drug resistance to antibiotics including aminoglycosides, quinolones, macrolides, and tetracyclines except being sensitive to some βlactams such as penicillin G, cephalothin, cefazolin, cephradine and cefuroxime. These data suggest that sensitive antibiotics such as penicillin G or some cephalosporins could be used for treating this disease through local delivery into the uterus following washing and sterilizing vulva with disinfectant fluid. Meanwhile, it is necessary to implement more strict clinical veterinary medicine practice, including disinfection procedure during semen collection and artificial insemination.
Though an experimental proof would be of high interest for demonstrating the occurrence of endometritis during artificial insemination by detection of semen and instruments for insemination or animal experiment, we did not do these tests in this study because the semen and other relevant insemination materials had not been gotten then, and animal experiment was limited in fields. Nevertheless, we estimated that the occurrence of endometritis could be associated with artificial insemination since sows unfertilized artificially were not attacked by this disease in this farm. In addition, it happened just on days 3-5 after artificial insemination.
In summary, the identification of S. porcinus provides important information for determining the cause of sow endometritis. The evaluation of antibiotic susceptibility of the isolate could contribute to the treatment of sow endometritis from bacterial infection. It should be noted that bacterial multi-drug resistance is a great challenge to the current disease control.

Conclusions
This study demonstrates that the sow endometritis was caused by S. porcinus infection during artificial insemination in the pig breeding farm. The isolate has multidrug resistance to most antibiotics applied clinically, which may be a great challenge for the control of animal infectious diseases. In addition, sensitive antibiotics such as penicillin G or some cephalosporins could be used for treating sow endometritis by S. porcinus infection in pig farm.

Case presentation
Eighteen multiparous sows and twenty-four replacement gilts had a fever and lost appetite 3 to 5 days after artificial insemination and generated large amounts of pale yellow vaginal secretions. On the second estrus, endometritis occurred, and gilts estrus conception rate was significantly decreased or breeding capacity was totally abolished. Five random sick pigs' vaginal secretions (3 from sows and 2 from gilts) were collected aseptically using 10 mL disposable syringes with needles removed and put into 15 mL sterile collection tubes for laboratory testing on day four after presentation of the clinical signs. Five of vaginal secretion samples were subsequently collected from three healthy sows and two healthy gilts, which were not fertilized by artificial insemination as the case controls. After the samples had been collected, the pigs were released but the diseased ones were separated from the healthy and got clinical treatment.

Characteristics of bacterial culture
All of the vaginal secretion samples were taken and streaked onto ordinary nutrient agar (Beijing Biotech AOB Star Co., China) and serum-agar plates with 5% sheep serum (Gibco) in ordinary nutrient agar medium. After incubation18~24 h at 37°C, colony morphology was examined, Gram stained, and the morphology of bacteria staining was observed under a microscope.
The single colony was re-streaked onto the ordinary nutrient agar, sheep serum agar, sheep blood agar, Mac-Conkey agar, and eosin methylene blue agar plates (Beijing Biotech AOB Star Co., China), respectively, and incubated overnight at 37°C. Phenotypic characteristics of colonies were further observed. Single colony was picked, inoculated into serum nutrient broth containing 5% sheep serum in the ordinary nutrient broth, and incubated for 16-18 h at 37°C with shaking at 200 rpm. The bacterial cultures were taken for Gram staining and observed under the microscope.

Catalase test
To conduct catalase examination, the isolated bacterial colonies were picked and placed on clean glass slides, and the strain ATCC25923 of Staphylococcus and the strain ATCC 27335 of Streptococcus was as the positive or the negative control, respectively; 3% hydrogen peroxide (Sigma) was dropped onto the bacteria on the slides. The appearance of large bubbles indicated that the catalase was positive, whereas no bubbles indicated that it was negative.

Biochemical examinations
To perform biochemical assays, the bacterial cells from cultures were inoculated into biochemical identification microtubes (Hangzhou Binhe Microorganism Regent Co., Ltd., China), respectively, and cultured for 24 h at 37°C according to the manufacturer's instructions.

PCR of 16S rRNA gene sequences
The bacterial 16S rRNA gene (rDNA) sequence universal primers 16SF (5′-AGAGTTTGATCCTGGCTCAG-3′) and 16SR (5′-GGTTACCTTGTTACGACTT-3′) were utilized for amplification of 16S rRNA sequence from the isolated bacterial cells. PCR reaction system 16S rRNA gene sequence analysis The 16S rDNA sequence from the isolate, designated as SS XT2018 (GenBank ID: MN148888), was compared with the reference sequences in GenBank (Table 3) using CLUSTAL W program of molecular evolutionary genetics analysis software (MEGA), version 7.0. An unrooted phylogenetic tree was constructed using the Neighbor-joining method with the maximum likelihood method and bootstrap tests of 1000 replicates.

Antibiotics susceptibility test
K-B disk diffusion method was used for testing the susceptibility of the isolate to antibiotics. Single colony was inoculated into serum Mueller-Hinton (MH) broth (BD/ Difco) with 5% sheep serum until a turbidity of culture reached 0.5 McFarland units (equivalent to 1.5 × 10 8 colony forming unit/mL) at 37°C. The bacterial broth was spread onto the surface of serum MH agar plates and antibiotics such as β-lactams, aminoglycosides, macrolides, tetracyclines, quinolones, lincomycin, macrodantin and rifampicin (Hangzhou Binho Microbial Agent Co., China) ( Table 3) were pasted onto the plates. The inhibition zone diameters on the agar plates were measured 18 to 24 h after incubation at 37°C. The isolate′s susceptibility to these antibiotics was determined according to the manufacturers' designated standard. The experiment was carried out three times, and the final inhibition zone diameters were expressed as the mean of repeats.
Additional file 1. Primers used for RT-PCR detection in this study.
Additional file 2. Viral DNA/RNA extraction and PCR.