Detection of selected antibiotic resistance genes using multiplex PCR assay in mastitis pathogens in the Czech Republic

The aim of this research was to develop multiplex polymerase chain reaction assays for the detection of aminoglycoside (strA, strB), sulphonamide (sulI, sulII), tetracycline (tetA, tetB, tetK, tetM, tetO), macrolide and lincosamide (msrA, ermA, ermB, ermC, mefA/E) genes of resistance in mastitis pathogens (Escherichia coli, Staphylococcus aureus, Streptococcus uberis, Streptococcus agalactiae and Streptococcus dysgalactiae). Applying the established assays, we investigated the distribution of antibiotic resistance genes in the above mentioned species isolated from milk samples in the Czech Republic. Each assay consisted of seven pairs of primers. Six of them amplified fragments of antibiotic resistance genes and one pair a fragment of a species specific gene. Polymerase chain reaction conditions were optimized to amplify seven gene fragments simultaneously in one reaction. In total, 249 isolates were used, among which 111 were positive for E. coli, 52 for S. aureus and 86 for Streptococcus spp. The majority (60.2%) of bacteria carried at least one antibiotic resistance gene and 44.6% were multidrug-resistant. The designed multiplex polymerase chain reaction assays may be applied as diagnostic method to replace or complement standard techniques of antibiotic susceptibility testing in the mentioned pathogens. S. uberis, S. aureus, E. coli, S. agalactiae, S. dysgalactiae Bovine mastitis presents the biggest problem in milk farming (Cressier and Bissonnette 2011). Its worldwide spread endangers food safety, causing enormous economic losses for the dairy industry. Over 150 different contagious and environmental microorganisms can cause mastitis (Kuang et al. 2009). The most common contagious pathogens are Streptococcus agalactiae, Streptococcus dysgalactiae, Staphylococcus aureus, and Mycoplasma spp. The most common environmental pathogens are Escherichia coli, Klebsiella pneumoniae, Enterobacter aerogenes, Serratia spp., Proteus spp., Pseudomonas spp., and other gramnegative bacteria, coagulase-negative staphylococci, environmental streptococci, yeast or fungi, Prototheca spp., Arcanobacterium pyogenes, and Corynebacterium bovis (Divers and Peek 2008). Bzdil (2011) reports about the prevalence of mastitis agents in the Czech Republic. Data from 41 regions over a ten year period from 2000 to 2010 were evaluated, showing that the five most frequent pathogens were Streptococcus uberis (22.09%), Staphylococcus aureus (16.41%), Escherichia coli (7.01%), Streptococcus agalactiae (5.16%) and Streptococcus dysgalactiae (5.09%) (Bzdil 2011). In a similar study, which assessed mastitis pathogens in clinically healthy cows, results differ. There, the most prevalent pathogen was Staphylococcus sciuri (14.2%), followed by Staphylococcus xylosus (10.9%). Staphylococcus aureus was detected in 9.0% of the samples, E. coli in 6.6%. So called mastitis streptococci (S. agalactiae, S. dysgalactiae, and S. uberis) were discovered in 11.7% of all samples (Cervinkova et al. 2013). The most common treatment strategy for mastitis is the use of antimicrobial drugs (Pyorala 2009); however, their wide use in veterinary medicine plays a significant role in ACTA VET. BRNO 2017, 86: 167–174; https://doi.org/10.2754/avb201786020167 Address for correspondence: Vladimir Pyatov Department of Animal Morphology, Physiology and Genetics Faculty of AgriScience, Mendel University in Brno Zemědělská 1, 613 00 Brno, Czech Republic Phone: +420 545 133 378 E-mail: pyatov.vladimir@gmail.com http://actavet.vfu.cz/ the increasing resistance to them. As a consequence, the treatment of diseases caused by bacterial pathogens has become very difficult (Schwaiger et al. 2010). For this reason, determining the susceptibility of a microorganism to antibiotics is essential to proper treatment. Testing can demonstrate whether antimicrobial agents are useful against certain causative agents (Willey et al. 2008). For in vitro testing of antimicrobial susceptibility, methods such as disk diffusion and broth dilution are commonly used (Kalmus et al. 2011). Despite the low cost and easy use, these methods have several disadvantages, such as labour and time consumption and investigator dependence (Driscoll et al. 2013; Hombach et al. 2013). During the past few years several studies have reported about the application of polymerase chain reaction (PCR) based assays as a diagnostic tool for detecting mastitisassociated pathogens in milk (Shome et al. 2011; Shakuntala et al. 2012). The technique is highly sensitive, enables a high throughput and rapid results (Gao et al. 2011). At the same time, PCR is used to identify genes of resistance. Several multiplex PCR assays for detection of antibiotic resistant genes in E. coli (Sianglum et al. 2009), S. aureus (Strommenger et al. 2003) and streptococci (Malhotra-Kumar et al. 2005) have been developed. Since penicillins and cephalosporins are the most commonly used antibiotics for mastitis treatment (De Briyne et al. 2014), many PCR assays which detect betalactam antibiotic resistance genes have been designed (Pomba et al. 2006). There are further common genes of resistance. For E. coli they are: tetA, tetB, strA, strB, sulI, sulII (Srinivasan et al. 2006); for S. aureus: ermA, ermB, ermC, msrA, tetM, tetK (Gao et al. 2011; Parvizi et al. 2012); and for streptococci: ermA, ermB, mefA/E, tetO, tetM and tetK (Loch et al. 2005; Rato et al. 2013). These genes are responsible for bacteria not being susceptible to aminoglycoside (strA, strB), sulphonamide (sulI, sulII), tetracycline (tetA, tetB, tetK, tetM, tetO), macrolide and lincosamide (msrA, ermA, ermB, ermC, mefA/E) antibiotics. These are antibiotics widely used in the treatment of mastitis (Divers and Peek 2008). Only few PCR assays were applied for the analysis of milk samples and exclusively for detecting genes of resistance in S. aureus (Gao et al. 2011; Parvizi et al. 2012). The aim of this study was to develop multiplex PCR assays for detecting the most common aminoglycoside, sulphonamide, tetracycline, macrolide and lincosamide genes of resistance in bovine milk samples, originating from the pathogens S. uberis, S. aureus, E. coli, S. agalactiae and S. dysgalactiae. Additionally, the prevalence of these resistances in the Czech Republic was assessed. Materials and Methods A total of 218 cow milk samples were used in this study. The samples were obtained during milking. Desoxyribonucleic acid (DNA) extraction directly from milk samples and bacterial species identification were carried out by using “Thermo Scientific PathoProofTM Mastitis Complete-12 assay” (Thermo Fisher Scientific Inc. Waltham, MA USA). Additionally, 31 bacterial samples isolated from milk with tested antimicrobial susceptibility were included in the assay. For this, colonies of 5 species of interest were obtained from the State Veterinary Institute in Olomouc. Seven of them were E. coli, six were S. aureus, six were S. dysgalactiae, five were S. agalactiae, and seven were S. uberis. Milk samples had been processed at the State Veterinary Institute in Olomouc with cultivation on blood agar (Trios s.r.o. Prague, Czech Republic) at 37 ± 1 °C for 42–48 h. The isolates were tested by the disk diffusion method. The antibiotic panel included streptomycin, tetracycline, neomycin, cefoperazon, novobiocin, clindamycin, cephalothin, amoxycilin/clavulanic acid, cotrimoxazol, oxacilin, colistin, ampicilin and cephalexin. The DNA extraction from the obtained bacterial colonies was carried out using “DNeasy Blood & Tissue Kit (50) (QIAGEN®), which is suitable for such bacterial samples. The DNA concentration and purity were assessed using Nanodrop 2000 (Thermo Fisher Scientific Inc. Waltham, MA USA). PCR was used for detection of aminoglycoside (strA, strB), sulphonamide (sulI, sulII), tetracycline (tetA, tetB, tetK, tetM, tetO), macrolide and lincosamide (msrA, ermA, ermB, ermC, mefA/E) resistance genes. To identify the pathogens, the following specific genes were chosen: mrdB for E. coli, rRNA-16S (ribosomal ribonucleic acid) for S. agalactiae, rRNA-16S for S. dysgalactiae, Cpn60 for S. uberis and rRNA-23S for S. aureus. The primers 168

Bovine mastitis presents the biggest problem in milk farming (Cressier and Bissonnette 2011).Its worldwide spread endangers food safety, causing enormous economic losses for the dairy industry.
The most common treatment strategy for mastitis is the use of antimicrobial drugs (Pyorala 2009); however, their wide use in veterinary medicine plays a significant role in the increasing resistance to them.As a consequence, the treatment of diseases caused by bacterial pathogens has become very difficult (Schwaiger et al. 2010).For this reason, determining the susceptibility of a microorganism to antibiotics is essential to proper treatment.Testing can demonstrate whether antimicrobial agents are useful against certain causative agents (Willey et al. 2008).
For in vitro testing of antimicrobial susceptibility, methods such as disk diffusion and broth dilution are commonly used (Kalmus et al. 2011).Despite the low cost and easy use, these methods have several disadvantages, such as labour and time consumption and investigator dependence (Driscoll et al. 2013;Hombach et al. 2013).
Only few PCR assays were applied for the analysis of milk samples and exclusively for detecting genes of resistance in S. aureus (Gao et al. 2011;Parvizi et al. 2012).
The aim of this study was to develop multiplex PCR assays for detecting the most common aminoglycoside, sulphonamide, tetracycline, macrolide and lincosamide genes of resistance in bovine milk samples, originating from the pathogens S. uberis, S. aureus, E. coli, S. agalactiae and S. dysgalactiae.Additionally, the prevalence of these resistances in the Czech Republic was assessed.

Materials and Methods
A total of 218 cow milk samples were used in this study.The samples were obtained during milking.Desoxyribonucleic acid (DNA) extraction directly from milk samples and bacterial species identification were carried out by using "Thermo Scientific PathoProof™ Mastitis Complete-12 assay" (Thermo Fisher Scientific Inc. Waltham, MA USA).
Additionally, 31 bacterial samples isolated from milk with tested antimicrobial susceptibility were included in the assay.For this, colonies of 5 species of interest were obtained from the State Veterinary Institute in Olomouc.Seven of them were E. coli, six were S. aureus, six were S. dysgalactiae, five were S. agalactiae, and seven were S. uberis.Milk samples had been processed at the State Veterinary Institute in Olomouc with cultivation on blood agar (Trios s.r.o.Prague, Czech Republic) at 37 ± 1 °C for 42-48 h.The isolates were tested by the disk diffusion method.The antibiotic panel included streptomycin, tetracycline, neomycin, cefoperazon, novobiocin, clindamycin, cephalothin, amoxycilin/clavulanic acid, cotrimoxazol, oxacilin, colistin, ampicilin and cephalexin.
The DNA extraction from the obtained bacterial colonies was carried out using "DNeasy Blood & Tissue Kit (50) (QIAGEN ® ), which is suitable for such bacterial samples.The DNA concentration and purity were assessed using Nanodrop 2000 (Thermo Fisher Scientific Inc. Waltham, MA USA).
PCR was performed under the following conditions: initial denaturation at 95 °C for 1 min, 35 cycles of denaturation at 95 °C for 30 s, primer annealing at 60 °C for 30 s and elongation at 72 °C for 30 s, and final elongation at 72 °C for 7 min.Reaction products were detected using a 2% agarose gel with ethidium bromide.
All PCR fragments had the correct size.The primers were validated using DNA from genotypically defined isolates which were obtained from the Veterinary Research Institute (Brno, Czech Republic).The isolates were also used as positive controls.The products of the PCRs were sequenced and data were compared to corresponding sequences in the GeneBank using the BLAST algorithm available at the National Center for Biotechnology Information website (www.ncbi.nlm.nih.gov).

Strains with tested antimicrobial susceptibility
For the 31 isolates obtained from the State Veterinary Institute in Olomouc, results of the antimicrobial susceptibility test and PCR were compared.A good correlation was observed between phenotypical resistance and detected genes.However, four isolates with the detected genes were not resistant to the corresponding antibiotics.Three E. coli isolates, one with detected strA and strB genes, the second and the third with detected sulII gene, and one S. aureus isolate with detected ermB gene did not show resistance to corresponding antibiotics.
Four isolates of E. coli were resistant to tetracycline and carried the tetA gene.Four isolates were resistant to sulphonamide, but six carried the sulII gene and two additionally sulI.Five were resistant to streptomycin (aminoglycoside), but six carried both strA and strB genes.
Four isolates of S. aureus were resistant to tetracycline, all of them carried the tetM gene and one additionally tetK.One isolate was resistant to lincosamide and carried the ermA gene.At the same time, an isolate without resistance to lincosamide carried the ermB gene.
Sixteen isolates of streptococci were tetracycline-resistant.Nine of them carried the tetM gene and seven carried tetO.Eight isolates were resistant to lincosamide.Two carried the mefA/E gene and six carried ermB.

Strains without tested antimicrobial susceptibility
In total, 218 samples which contained species of interest were used.Following numbers of species were identified: 104 E. coli, 46 S. aureus, 25 S. agalactiae, 11 S. dysgalactiae, and 32 S. uberis.
Among streptococci, the most common genes were tetM and ermB, which were detected in 32 (47.1%) and 24 (35.3%)samples, respectively.TetK was found in 16 (23.5%)samples, tetO in 11 (16.2%),mefA/E in 9 (13.2%) and ermA in 6 (8.8%).In total, two isolates carried five genes of resistance, five carried four genes, 13 carried three genes, nine carried two genes and 11 carried one gene.No sample contained all six genes.In 28 isolates none of the sought genes was found.

Discussion
Despite the simplicity of the phenotypical susceptibility testing methods, the necessity to wait for the results for 48 h or more is a big disadvantage (Martineau et al. 2000).Multiplex PCR assay which simultaneously allows for the detection of several genes in a single reaction, has the advantage of identifying genotypic resistance to several antibiotics more rapidly and reliably (Choi et al. 2003).In this study, five multiplex PCR assays have been developed to test for the presence of S. uberis, S. aureus, E. coli, S. agalactiae and S. dysgalactiae (Table 2) and determine their antibiotic resistance by detecting the related resistance genes.The PCR assays allow highly accurate evaluation of antimicrobial resistance of mastitis pathogens and establishment of effective antibiotic therapy for cows with the disease.The assays could also be used for detection of the genes in samples of another origin.
Another aim of the study was to investigate the distribution of genetic resistance genes in the above mentioned species in milk samples in the Czech Republic.In our study, a total of 249 samples were analyzed.Among them 111 were E. coli, 52 S. aureus and 86 Streptococcus spp.
Of the E. coli isolates, 43 did not carry any gene of resistance.This contrasts with Srinivasan et al. (2007) who found at least one resistance gene in each investigated sample.On the other hand, their results show that the majority of E. coli (90.7%) was multidrug resistant.We also detected a multidrug resistance in the majority of samples, but only to 51.4%.
In a considerable number of E. coli isolates, strA (42.3%) and strB (52.3%) genes were amplified, which goes against the results of two other studies (Lanz et al. 2003;Srinivasan et al. 2007).SulI and sulII were detected in 27.9% and 28.8% of the isolates, respectively.Karczmarczyk et al. (2011) found similar results for the first gene, but the second one was found in 90% of isolates.Srinivasan et al. (2007) detected the mentioned genes only in 8% of all isolates.In their study, TetA (23.4%) and tetB (27.9%) were also detected at higher frequencies than in that of Srinivasan et al. (2007).Karczmarczyk et al. (2011) found tetB at a similar amount of E. coli as we did, but the presence of tetA was twice as high.Skockova et al. (2012) revealed the mentioned tetracycline resistant genes in highly varying amounts over a longer period.
Almost one third of S. aureus carried tetM (30.8%), whereas tetK was detected only in 15.4%.Similar results for tetracycline resistant genes have been shown in other reports (Gao et al. 2011(Gao et al. , 2012)).However, Kumar et al. (2010) observed a significantly higher prevalence of tetK compared to tetM.
Molecular analysis also revealed a high number of isolates with msrA (30.8%) and ermC (25.0%) genes, but quite few with ermB (9.6%) and ermA (5.8%).Gao et al. (2011Gao et al. ( , 2012) ) show in both their studies similar results concerning ermC, while in one study ermB was detected at a low rate, comparably to our results, and in another one ermB and ermA were not found at all.Kumar et al. (2010) revealed msrA in about a quarter of samples, while ermA and ermC could not be detected.In the work of Parvizi et al. (2012) msrA was found in 40% of the isolates and three other resistance genes in 20% each.
In 86 samples of Streptococcus spp., 57 (66.3%) carried at least one of the selected resistance genes.A study by Ruegg et al. (2015) demonstrated similar numbers, where 52.6% of samples were positive.Genes responsible for tetracycline resistance were detected in many samples.The gene tetM was present in almost half of the isolates -41 (47.7%), tetO in 18 (20.9%)and tetK in 16 (18.6%).Similar results were shown for tetM, which was also the most frequent resistance gene discovered (Ruegg et al. 2015).This contrasts with the outcome of Rato et al. (2013), who report tetK as the most often detected tetracycline resistant gene.The most common macrolide resistance determinant was ermB (34.9%), followed by mefA/E (12.8%), and ermA (7.0%).The same ranking was shown in a report of Malhotra- Kumar et al. (2005).Combining the outcome of all these studies, it can be concluded that the distribution and relative frequency of genes of resistance in the same bacterial species are highly variable in different regions.Therefore, it is essential to test locally for every region for one cannot rely on data obtained from samples elsewhere.Another thing worth mentioning is the detection of genes of resistance in strains which did not show phenotypical resistance.This phenomenon could be explained by the fact that genes are not the only factors responsible for developing resistance to antibiotics (Cengiz et al. 2015).
The multiplex PCR assays developed in this study showed high sensitivity and specificity for S. uberis, S. aureus, E. coli, S. agalactiae and S. dysgalactiae (Fig. 1).The method is likely to be helpful for the rapid screening of antibiotic resistance.This research also reported the prevalence of aminoglycoside (strA, strB), sulphonamide (sulI, sulII), tetracycline (tetA, tetB, tetK, tetM, tetO), macrolide and lincosamide (msrA, ermA, ermB, ermC, mefA/E) resistance genes in the bacteria isolated from milk samples of dairy cows in the Czech Republic.

Table 2 .
Multiplex polymerase chain reaction assays made according to targeted bacterial species.