Genotypic and phenotypic β-lactam resistance and presence of PVL gene in Staphylococci from dry bovine udder

Dairy cows affected with subclinical mastitis can be sources of virulent, antimicrobial-resistant Staphylococci to humans because of the excretion of the bacteria through their milk. This study focussed on the phenotypic and genotypic antibiotic resistance patterns of Staphylococci isolated from dairy cows in early dry period. Among 96 isolates of Gram positive cocci from 157 cows, 76 were identified as Coagulase Negative Staphylococci and the remaining 20 were Staphylococcus aureus. Typical amplicons of coagulase gene were obtained for all 20 samples of S. aureus with three major coagulase types being identified as giving 627 bp (40%), 910 bp (35%) and 710 bp (25%) long PCR products. The groEL gene was amplified in PCR of all 76 isolates of Coagulase Negative Staphylococci, and incubation of PCR products with restriction enzyme PvuII yielded three distinct PCR-RFLP fragment patterns bearing resemblance to S. chromogenes and S. hyicus. Highest sensitivity of Coagulase Negative Staphylococci was noted for Azithromycin (92.5%) and the least to Tetracyclines (76.3%), whereas for S. aureus, it was Cefoperazone (95%) and Azithromycin (72.2%) respectively. Phenotypic resistance to Oxacillin (25 isolates), and Cefoxitin (11 isolates) was detected by dilution method with a commercial strip (Ezy MICTM). Genotypic resistance to β-Lactam antibiotics was found in 65 (34 with mecA gene and 31 with blaZ gene) isolates. Eighteen isolates possessed both the genes, with the PVL gene for virulence being detected in five of them. Nine isolates which had mecA gene were phenotypically susceptible to oxacillin while phenotypic resistance to oxacillin was observed in seven isolates that did not have either mecA or blaZ gene. This is the first report of persistent Staphylococcal infections possessing PVL gene and high level of genotypic resistance to β-Lactam antibiotics in small- holder dairy cattle from India.

. Infection by such antibiotic resistant Staphylococci pose a major threat to human health and dairy cows are recognised as one of the potential reservoirs [2,3]. Staphylococcus aureus (SA) is a major pathogen causing subclinical mastitis in bovines, while Coagulase Negative Staphylococci (CNS) are considered minor pathogens causing increase in somatic cell count with wide species variations. Staphylococci invade and colonise the mammary gland of cows during lactation, persist in the udder parenchyma throughout the life of the animal, and are shed continuously through the milk [4]. Antibiotic-resistant strains, particularly those resistant to β-lactam group of antibiotics have emerged in Staphylococci possibly as a result of regular antimicrobial use in dairy farms to control them. Their presence in bovine milk and dairy environment poses a potential risk to consumers, farmers, farm workers, and veterinarians alike who are exposed to the affected cattle.
Recently, the Indian Network for Surveillance of Antimicrobial Resistance in India has stressed the need to study epidemiology of MRSA [5]. India is the largest producer and consumer of milk in the world, and the practice of consuming raw milk is still prevalent among many Indian communities. The potential significance of milk as a source of antibiotic resistant Staphylococci to humans has not still had the attention it warrants despite well-established risks associated with consumption of raw or unpasteurised milk. The incidence rate of CA MRSA varies from 25% in Western parts of India to 50% in South India in human patients [5] while no reports are available on LA MRSA.
The continuing emergence of newer genetic clusters in Staphylococci demands consideration of genotypic resistance together with molecular level profile to identify relevant, evolving genotypes. The coagulase gene is recognized as a suitable target for molecular level profiling of phenotypically identified S. aureus isolates [6]. PCR-RFLP of the groEL gene was found to be capable of identifying 100% of CNS species isolated from human clinical samples and from milk samples of dairy cows with intra-mammary infection, and hence is considered to be an ideal universal target for identification of CNS to the species level [7].
Resistance to penicillin as developed by staphylococcus is mediated by two mechanisms: either by the secretion of an enzyme β-lactamases which inactivates the antibiotic by hydrolysis of its β-lactam ring, and is encoded by the gene blaZ in a plasmid or chromosome [8], or by the production of a penicillin binding protein (PBP2A) encoded by a gene mecA, embedded in a mobile genetic element on the bacterial chromosome called Staphylococcal Chromosome Cassette mec or SCCmec [6]. Isolates of Staphylococcus are characterized as Methicillin Resistant (MRSA for S. aureus and MRCNS for coagulase negative Staphylococci) if they show the presence of the mecA gene and display phenotypic resistance to oxacillin/methicillin. The TetK gene found in Staphylococci codes for the energy dependant efflux of tetracycline conferring resistance.
Existence of virulence factors in MRSA is another area of importance as pathogenicity of the bacteria depends on their production. Panton-Valentine leukocidin (PVL) is a S. aureusspecific, bi-component cytotoxin that causes leucocyte destruction and tissue necrosis [9] in human patients. Inhibition of immune response by PVL also aid in bacterial persistence and extension of inflammatory process in soft tissues, bone and joints. While the strains of community-associated MRSA, the majority of which carry genes encoding Panton-Valentine leucocidin, are spreading rapidly in human populations, only sporadic cases have been reported in animals to date [10]. study. We further declare that no individuals employed or contracted by the funders (other than the named authors) played any role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
High prevalence of Staphylococcal infections were found when secretions from a large number of dry cows maintained under small holder systems in an area of Kerala State, India were screened in this study. Phenotypic and genotypic antibiogram profile of the isolates revealed resistance to a number of commonly used antibiotics in human medicine.

Materials and methods Animals
One hundred and fifty seven pregnant dairy cows within 5 days of cessation of their lactation were included in this study. Details of all the small holder dairy units (2-3 cows/holding) from an area of twenty Sq. km in the district of Thrissur in Kerala State, India, were collected and each unit was visited frequently. The animals belonged to the cross-bred category of nondescript breeds with either Jersey or Holstein-Friesian breeds. Udder secretions were collected from every one of the 157 pregnant cows within five days of cessation of their lactation. This study did not require any specific permissions since it involved only a non-invasive collection of samples from a domestic species. Administrative and technical sanction to the work was granted as per order KVASU/DAR/R2/116/2011 dated 01/09/11 of DAR, KVASU of the Director of Academics and Research, Kerala Veterinary and Animal Sciences University.

Bacterial isolates
Mammary gland secretions of cows were collected in sterile vials and sealed aseptically. Bacterial cultures were obtained by inoculating the samples on Brain Heart Infusion Agar (Hi Media, India) and incubating for 24 h at 37˚C. The plates were examined for growth and a cutoff value of 100CFU/ml of secretion was used. Pure cultures were maintained after re-inoculating individual colonies on Blood agar and Mannitol Salt agar. Selective isolation of Staphylococcus sp. was done on Mannitol salt agar and Baird-Parker medium containing one per cent Potassium tellurite. Further identification was carried out by Grams staining and biochemical tests for catalase, oxidase, oxidation-fermentation and production of acid from various sugars. Tube coagulase test was employed for initial discrimination of S. aureus and coagulase negative Staphylococcus isolates.

Phenotypic resistance to Oxacillin and Cefoxitin
Quantitative determination of susceptibility of bacteria to oxacillin and Cefoxitin was done using an Ezy MIC™ strip (Oxacillin Ezy MIC TM Strip (OXA) and Cefoxitin Ezy MIC TM 0.016-256 mcg/ml strips, Hi Media, India) according to the manufacturer's instructions. The strip comprised of a predefined quantitative gradient which was used to determine the Minimum Inhibitory Concentration (MIC) in mcg/ml of the two antimicrobial agents against Staphylococci. Mueller-Hinton Agar with 2% NaCl added was the media used and the tests were incubated at 35˚C for 24 hours. Interpretation was done using the following interpretive criteria for susceptibility categorization as per Clinical and Laboratory Standards Institute (CLSI) guidelines, 2011 [12].

Extraction of genomic DNA and PCR
Overnight cultures of samples were prepared in Brain Heart Infusion broth. DNA was isolated using the DNeasy blood and tissue DNA extraction kit (Qiagen) according to the manufacturer's instructions. Lysis of the bacterial cell wall was enhanced by adding Lysostaphin (Sigma-Aldrich) to the lysis buffer at a concentration of 10μg/ml. The primers used and protocols followed for each PCR is given in Table 1. The restriction fragment length polymorphism of groEL Chaperonin gene in CNS was studied using PvuII restriction endonuclease (Fermentas, Germany), as per the prescribed protocol [6].

Antimicrobial susceptibility
Majority of isolates was sensitive to Cefoperazone (95%), followed by Streptomycin (94.4%), while Azithromycin (72%) was the least effective molecule against S. aureus according to the disc diffusion results. Among the CNS, the isolates were most susceptible to Azithromycin (92.5%) and highest % of antimicrobial resistance was found for Tetracycline (76.3%). Percentage susceptibility of isoltes is given in Table 2. Resistance to oxacillin was found in 25 isolates by the gradient method, while 11 isolates were resistant to Cefoxitin, and three isolates were resistant to both. (S1 Fig).

Molecular characterisation by PCR and PCR-RFLP
Coagulase typing. Typical amplicons of coagulase gene were obtained for all the 20 samples of S. aureus screened. However, the genetic diversity of the isolates was assessed using the size of the amplicon. The PCR product ranged between 600bp to 900bp for different strains of S. aureus. Three major coagulase types were identified. Eight (40%) isolates gave a 627bp long PCR product. Seven (35%) isolates revealed a 910bp amplicon while twenty five % of isolates carried a 710bp long coagulase gene ( Table 3).
RFLP of groEL gene. Single distinct bands of amplicons of the desired size (550bp) were obtained for all the 76 isolates of CNS (S2 Fig). The PCR-RFLP yielded 3 distinct patterns after partial digestion with restriction endonuclease Pvu II (S3 Fig). Thirty % of samples gave intact products after digestion. Fragments of around 400bp and 150 bp were obtained for 40% of the samples and rest yielded partially digested products of size 500bp, 400bp and 250bp (Table 3).

Discussion
This is the first report of persistent infection of dry bovine udders by Staphylococci having genes for methicillin resistance and the virulence factor PVL. Routine prevalence studies based upon detection of Staphylococci in milk during lactation find it difficult to differentiate between incidental transient infection or contamination during milking procedures and the existing infections. This study avoids this error by taking only those organisms present during the dry period into consideration [18]. The percentage (27.9) of phenotypic resistance of Staphylococci to oxacillin detected in our study was similar to reports from Uruguay (22%) [19] and Argentina (27.6%) [20], but lower than for Korea (60.2%) [8]. Phenotypic resistance to Cefoxitin was 15.8%, indicating presence of mecA or mecC MRSA [21] while 3 isolates were resistant to both antibiotics.
Nine isolates which had a mecA gene were phenotypically susceptible to oxacillin. Similar findings have been reported in two isolates from human samples from a nearby referral hospital, and further investigation before initiating treatment of infections with such organisms was recommended [22]. Phenotypic resistance to oxacillin was observed in seven isolates that did not have either mecA or blaZ gene. The blaZ gene-negative, but oxacillin-resistant Staphylococcus isolates which had been reported earlier [6] were assumed to be β-lactamase hyper producing strains by the authors since all of them were susceptible to amoxicillin-clauvulanate combination. We did not use this particular antibiotic disc since our study concentrated on antibiotics which are available as intra-mammary preparations. Another observation by the same authors, that MRSA generally exhibit resistance to tetracyclines, also could not be confirmed by our findings since none of the isolates in our study had the tetK gene. This is also the first report of PVL gene in S. aureus from dairy cattle in India. Presence of the PVL gene in methicillin succeptible [23] and methicillin resistant [24] S. aureus from other animals has been previously reported, while others [25] have found presence of PVL gene to be independent of specific genetic backgrounds like mecA gene.
Reported bacterial resistance levels to different antibiotics, as determined by disc diffusion method differ between studies [3,14,15,26]. One reason for the comparatively low susceptibility of isolates in this study could be the indiscriminate use of antibiotics in the area for treatment of mastitis providing selective pressure on the bacteria for acquisition of resistance. Poor sensitivity to Chloramphenicol and B-lactam group of antibiotics have been reported by others also [26,27]. But to our knowledge, this is the first time in India that a high level (27.8%) of resistance to Azithromycin by S. aureus is reported. This is of concern since Azithromycin is a critical important drug in treatment of bacterial infections in humans. High level of resistance to erythromycin has been reported in S. aureus (33.1%) and CNS (67%) in China, probably as a result of extensive use of macrolide antibiotics in dairy cows of that region [28].
High prevalence of Staphylococcal infection in the udder of dairy cows have been reported in large dairy herds before (1) where chances of contagious transmission are high. But high prevalence rates (61.14%) in small holder systems in our study, which are isolated from each other and employ individual milkers indicate the Staphylococcal infection to be endemic among the entire cattle population of the region under study. This, coupled with the antibiotic resistance pattern and virulence of some of the isolates indicate a serious threat to human health in the area by infection with the Staphylococci as such, or by the potential transmissibility of specific resistant clones or antimicrobial resistance determinants to pathogens in humans. Environmental contamination by MRSA from veal calves was found to be a major source of infection to farmers and members of their household [3]. Genes for resistance to βlactam group of antibiotics like methicillin are assumed to have originated in strains of S. fleurettii primarily found in animals, and have spread to human pathogens from that source [29]. PVL toxin, on the other hand, is reported to be having a major role in pathogenesis of CA-MRSA in human patients, but not in animal models [30].
Currently, staphylococci other than S. aureus comprise 39 characterized species. More than 10 CNS species have been isolated from mastitis, but only a few species predominate. The groEL chaperone has been identified as a heat shock protein essential for the survival of bacteria, and PCR-RFLP of groEL gene is a valuable tool for accurate identification of Staphylococci independent of origin of the isolate [31]. The PCR-RFLP patterns of groEL gene obtained in this study have close resemblance with the patterns obtained with typed strains of S. chromogenes and S. hyicus. Double digestion with HindIII and PvuII enzymes have been recommended for species level identification by PCR-RFLP of these two species [31]. Since we have used only PvuII for digestion, further confirmation may be required for the identity of these bacteria. Detection of S. hyicus in our study is of concern since this species has been occasionally associated with severe infections, but not all the strains have the same virulence characteristics [32].
In conclusion, dairy cattle in the area of study had high rates of infection by antibiotic resistant bacteria possessing virulent genes which may hold a serious threat to human health. Our findings point to the need for conjointly considering the molecular level identity and antimicrobial profile of Staphylococci in dairy cattle as an important aspect of epidemiological studies for therapeutic management and control in humans.