Abstract
The aim of this study was to determine the genetic lineages, and the frequency of antibiotic resistance and virulence determinants in methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-susceptible S. aureus (MSSA) isolates recovered from milk of cows with clinical mastitis. Three hundred milk samples from bovine with clinical mastitis were obtained from 30 dairy farms in different regions of Tunisia. Fifteen of the 300 tested samples contained S. aureus (5%), in three cases were MRSA. Isolates (one/sample) were typed (S. aureus protein A [spa], multilocus sequence typing and accessory gene regulator [agr]). The presence of resistance and virulence genes was analyzed by PCR. The three MRSA isolates contained mecA and blaZ genes (one of them also the msr(A) gene), and carried the enterotoxin gene sen; they were typed as t10381-ST4114 or t267-ST4120, and corresponded to agr type-I. Twelve MSSA isolates were recovered and harbored the blaZ (7 strains) or erm(C) genes (1 strain). The MSSA isolates presented seven different spa-types, associated to new sequence types (STs): t426-ST4118, t267-ST4120, t1773-ST4115, t509-ST4119, t529-ST4117, t2844-ST4113, and t2802-ST4112; most isolates (8/12) were typed as t267/ST4120. All S. aureus isolates were scn-negative, except one MSSA of lineage ST4119 that exhibited the immune evasion cluster type D, and harbored the seg, sei, sem, seo, and seu enterotoxin genes. Four MSSA isolates carried the toxic shock syndrome toxin 1 gene (tst). S. aureus (including MRSA) is an important cause of bovine mastitis, showing isolates with high genetic diversity and high content in virulence genes.
Introduction
M
The overuse of antimicrobial agents in human and veterinary medicine could favor the emergence and dissemination of multidrug-resistant bacteria, including MRSA. S. aureus and MRSA have been recovered from dairy cattle, and also from bovine milk. Milk could be contaminated with these microorganisms either by direct excretion from udders with clinical staphylococcal mastitis, or by indirect contamination from the environment. Indeed, these bacteria could survive for several months during handling and conservation processing in bulk tank.4–6
Methicillin resistance in S. aureus is conferred by the expression of the mecA gene, which also causes resistance to most β-lactams.7,8 β-Lactam resistance in S. aureus is an important cause of treatment failure of human and animal infections. The horizontal transfer of antimicrobial resistance genes between livestock and human-associated isolates is an increasing public health concern. 9 MRSA strains were first reported in animals in 1972, causing bovine mammary infections. Later, more MRSA strains were detected in farm animals (such as cattle, pigs, or poultry) and the zoonotic risk of transmission to humans was demonstrated. 9 Due to this zoonotic risk to humans from MRSA associated to food-producing animals, it is important to assess the importance of milk from bovine mastitis cases in the transmission of MRSA in the animal–human interface. 8 Moreover, interest in methicillin-susceptible S. aureus (MSSA) has also increased in recent years, as published reports have proven that MSSA can also be implicated in important infections and may help to explain the appearance and evolution of the different MRSA lineages. 3
S. aureus isolates can contain a diversity of virulence factors, being of special relevance the Panton-Valentine leukocidin (PVL), the toxic shock syndrome toxin 1 (TSST) and the staphylococcal enterotoxins (SE). These virulence factors could have a role in skin and soft tissue infections in humans or animals, and these could also be associated with cases of severe pneumonia and food poisoning. 10 The expression of virulence genes is under the control of a global quorum-sensing regulator system, named agr (accessory gene regulator). 10 Biofilm formation in S. aureus is due to the expression of the intercellular adhesion gene A (icaA) and it is considered as an important virulence factor in bovine mastitis. 11
To develop an effective control and treatment of bovine mastitis, it is important to study the prevalence, genetic diversity, antimicrobial resistance, and virulence of S. aureus that cause clinical mastitis. Currently, these types of data are not available for bovine mastitis in Tunisia. Thus, the goals of this study were to determine the genetic lineages, incidence of antibiotic resistance, and content in virulence determinants of S. aureus isolates from cows with clinical mastitis in Tunisia.
Materials and Methods
Bovine milk sample collection
Three hundred milk samples were collected from cows showing symptoms of clinical mastitis (one sample/cow). Samples were obtained in 30 farms with intensive breeding across different regions in North and South of Tunisia during October 2013 to September 2014. The farms implicated in the study were involved in the production of milk for self-consumption, cheese production, and milk-bottling. Each cow was clinically diagnosed for the appearance of general clinical signs related to udder and teats and presence of any gross abnormalities like fibrosis, inflammatory swellings, pain, visible injury or lesion, atrophy of the tissue, and teat blindness. The milk sample was observed for changes regarding color, odor, and consistency. The presence of clots, flakes, blood, and other consistent changes were indicators of clinical mastitis along with udder and teat morphological changes. A milk sample of one infected quarter was obtained from the 300 tested mastitis cows by veterinarians. Before taking the milk samples, teats were washed thoroughly and dried. They were then sprayed with 70% ethanol, the first few squirts of milk were discarded, and 30 ml milk samples were collected in sterile tubes. The milk samples were then transferred to the laboratory in cooler and immediately processed. It is important to note that the studied cows did not receive any antibiotic therapy.
Isolation and identification of S. aureus isolates
An aliquot of 10 μl of each 30 ml milk sample was streaked onto Oxacillin Resistance Screening Agar Base medium (Oxoid) for MRSA recovery. In parallel, 1 ml of milk sample was added in 9 ml to sterile saline solution for serial dilutions, then they were seeded on Baird Parker (Biolife) for S. aureus recovery. Plates were incubated at 37°C for 24–48 hours. Isolates with typical S. aureus morphology were selected (one per sample) and identified by classical biochemical methods [Gram staining, oxydase, catalase, DNase, and ability to coagulate rabbit plasma (Bio-Rad)]. 12 Molecular identification was performed by species-specific PCR amplification of nuc gene. 12 S. aureus ATCC 43300 was used as control strain.
Antimicrobial susceptibility testing
Susceptibility to 15 antimicrobial agents was performed using the disk-diffusion method in accordance with the Clinical and Laboratory Standards Institute (CLSI) recommendations. Antimicrobial agents tested (charge in μg) were as follows: penicillin (10), oxacillin (1), cefoxitin (30), kanamycin (30), gentamicin (10), tobramycin (10), tetracycline (30), chloramphenicol (30), trimethoprim–sulfamethoxazol (1.25/23.75), erythromycin (15), clindamycin (2), ciprofloxacin (5), vancomycin (30), and teicoplanin (30). 13 Methodology and guidelines for streptomycin (10 μg/disk) were as recommended by the French Society for Microbiology (www.sfm.asso.fr). Moreover, erythromycin-resistant strains were tested for induction of clindamycin resistance, as described previously. 14
Detection of the mecA gene
The presence of the mecA gene was studied by PCR in all oxacillin and/or cefoxitin-resistant isolates, as previously described. 12 S. aureus ATCC 43300 was used as control strain.
Molecular typing of S. aureus isolates
S. aureus protein A (spa)-Typing was performed in all S. aureus isolates (n = 15). The polymorphic X region of spa gene was amplified by PCR and sequences were analyzed using Ridom Staph-Type software version 1.5.21 (Ridom GmbH). It automatically detects spa repeats and assigns a spa-type according to http://spaserver.ridom.de/. Identification of agr allele group (I–IV) was determined by PCR as described earlier. 15 Multilocus Sequence Typing (MLST) was performed for all S. aureus isolates, as described previously. 16 The allelic profile of each strain was obtained by sequencing internal fragments of seven unlinked housekeeping genes (carbamate kinase [arcC], shikimate dehydrogenase [aroE], glycerol kinase [glpF], guanylate kinase [gmk], phosphate acetyl transferase [pta], triosephosphate isomerase [tpi], and acetyl coenzyme A acetyltransferase [yqiL]). Alleles of the seven genes defined the allelic profile, which corresponded to a sequence type (ST), assigned by the S. aureus MLST database (https://pubmlst.org/saureus).
Detection of antimicrobial resistance genes
Detection of antimicrobial resistance genes [blaZ, erm(A), erm(B), erm(C), erm(T), msr(A), msr(B), mph(A), mph(C), tet(K), tet(M), and tet(L)] was investigated in resistant and intermediate isolates by specific PCRs. 12 Positive and negative controls from the collection of the University of La Rioja were used in each PCR assay.
Detection of staphylococcal toxin genes and biofilm production
All isolates were tested by PCR for the presence of eight genes coding for staphylococcal enterotoxins (sea, see, seg, sei, sem, sen, seo, and seu), toxic shock syndrome toxin 1 (tst), leukocidin of Panton Valentine (PVL, lukF-lukS-PV), and exfoliative ETA and ETB toxins (etA and etB). 12 Furthermore, the presence of genes coding for the formation of biofilm in human and bovine mammary gland S. aureus infections contained in the ica operon (icaA, icaB) was determined by PCR. 11
Detection of the immune evasion gene cluster
All isolates were tested by PCR for the presence of five genes (scn, chp, sak, sea, and sep) of the immune evasion cluster (IEC) system. These genes are enclosed in the φ3 bacteriophage and encode the IEC system, which helps bacteria to survive in the human host by evading the innate immune system. Detected alleles allowed the classification of seven IEC types (from A to G), as previously reported. 17 The presence of all these genes were tested using primers and conditions as previously described. 17
Results
Prevalence of S. aureus (MRSA and MSSA)
Isolates were collected from milk samples of cows with symptoms of clinical mastitis across different regions in North and South of Tunisia. Of the 300 samples tested, 83 milk samples showed the growth of staphylococci and one isolate per sample was further characterized. Fifteen of these 83 strains (18%) were identified as S. aureus since they presented the ability to coagulate bovine plasma and were confirmed by a species-specific nuc PCR strategy. The S. aureus isolates were recovered from 15 different cows of seven different farms. Only three of these isolates were MRSA, recovered from cows of two different farms (31 milk samples were tested in these two MRSA-positive farms) (Table 1). MRSA isolates contained the mecA gene, as evidenced by PCR.
ERY, erythromycin; FOX, cefoxitin; OXA, oxacillin; PEN, penicillin; STR, streptomycin; TET, tetracycline.
IEC, genes of IEC (type).
IEC, immune evasion cluster; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible S. aureus; ST, sequence type.
Molecular typing of MRSA/MSSA isolates
The characteristics of S. aureus isolates in this study are shown in Table 1.
Among MRSA, two different spa-types were detected: (1) two strains were typed as t267 with a new ST4120; (2) one strain showed a new spa-type and a new ST (t10381/ST4114). All MRSA strains belonged to agr type-I.
The 12 MSSA isolates recovered in this study were typed by MLST (one isolate of each spa-type), and seven different spa-types were detected: t267/ST4120 (six isolates), and t426/ST4118, t509/ST4119, t1773/ST4115, t529/ST4117, t2844/ST4113, and t2802/ST4112 (one isolate, each one). All STs detected revealed novel alleles or novel allele combinations (Table 2).
CC, clonal complex; DLV, double locus variant; SLV, single locus variant; TLV, triple locus variant.
Antimicrobial resistance of MRSA/MSSA isolates
Seven MSSA isolates showed penicillin resistance and harbored the blaZ gene, one isolate was erythromycin-resistant and harbored the erm(C) gene, and another isolate was tetracycline-resistant but was negative for tet(M), tet(L), and tet(K) genes. All three MRSA strains carried the mecA gene, and the blaZ gene, and one of the isolates was erythromycin-resistant and contained the msr(A) gene (Table 1). All MRSA and MSSA isolates were susceptible to tobramycin, chloramphenicol, gentamicin, ciprofloxacin, clindamycin, kanamycin, trimethoprim-sulfamethoxazole, vancomycin, and teicoplanin (Table 1).
Virulence genes and IEC profile MRSA/MSSA isolates
The presence of virulence genes in MRSA and MSSA isolates is shown in Table 1. MRSA isolates were negative for all tested toxin genes except for the enterotoxin sen gene, which was detected in all of them. MRSA isolates were negative for the genes of the IEC system. For MSSA isolates, only one strain of lineage ST4119 was positive for scn, sak, and sea genes and exhibited the human-associated IEC-type D. Besides, this strain harbored a combination of five genes encoding enterotoxins (seg, sei, sem, seo, and seu). Another strain harbored the combination of four enterotoxins genes (seg, sei, sem, and seu) was detected. The remaining 11 MSSA isolates were IEC-negative. Four MSSA isolates carried the toxic shock syndrome toxin 1 gene (tst) (26.7%), and none of them was lukF-lukS-PV, etA, or etB positive. MRSA and MSSA isolates were negative for the intercellular adhesion genes icaA and icaB.
Discussion
To the best of our knowledge, there are no published data concerning MRSA from bovine mastitis in Tunisia, moreover, this type of data are also scarce in other African counties.18,19 S. aureus is an important etiological agent of mastitis in cattle worldwide.4,8,20 Its presence in bovine mastitis milk has been studied in Brazil, Poland, China, and several other countries.1,21,22 In addition, the emergence and spread of MRSA and MSSA is becoming a major animal and public health problem.
In this study, a low percentage of milk samples from clinical bovine mastitis carried S. aureus (5%). Other staphylococcal strains, as coagulase-negative staphylococci, were also detected in these samples (data not shown). The mecA gene was found in three of these 15 S. aureus isolates recovered (20%), representing 1% of total milk samples tested. MRSA represents a public health threat since it can be transmitted to farm workers and veterinarians after contact with infected animals. 23 It could be also accidentally transmitted after consumption of contaminated milk.
MRSA was rarely detected as a causing agent of bovine mastitis in several countries such as Finland (1.5%) and Germany (4.4%)9,24 with similar prevalence to our study. However, it was more frequently detected in Brazil (23%), Turkey (17.5%), and China (47.6%)7,25,26; differences in sample sizes, seasons, or geographical locations might explain these discrepancies. In Tunisia, few studies have covered the prevalence of MRSA and MSSA in meat products, healthy humans, or animals.12,27–29
β-Lactam antibiotics are widely used for the intramammary treatment of bovine mastitis in Tunisia, which could contribute to the dissemination of resistant strains for these agents. In our study, 10 S. aureus isolates were penicillin-resistant and harbored the blaZ gene. Furthermore, penicillin resistance of S. aureus from bovine mastitis has been increasingly reported throughout the world. The low frequency of tetracycline (one strain) or erythromycin resistances [two strains: one MSSA with erm(C), and one MRSA with msr(A)] detected among our isolates, was in agreement with previously published data from Poland. 21 While, higher frequency of resistance was reported in China and Iran with erm(C) as predominant erythromycin resistance gene, and tet(M) and tet(K) tetracycline-resistant genes.1,4 All MRSA and MSSA isolates were fully susceptible to chloramphenicol, gentamicin, ciprofloxacin, clindamycin, tobramycin sulfamethoxazole/trimethoprim, vancomycin, and teicoplanin. Similar results were found in Poland, 21 while other studies reported a multi-resistance profile for S. aureus isolated from bovine mastitis milk. 30 These low resistance rates of S. aureus in Tunisia compared to other countries could be explained by the low use of antibiotics. Similarly, in a recent Tunisian study, the analysis of bovine milk of healthy animals revealed low antimicrobial resistance among S. aureus isolates, except for penicillin, and most isolates corresponded to clonal complex 97 (CC97). 31
The virulence gene tst, which encodes toxic shock syndrome toxin, was identified in four MSSA isolates (26.7%). This super antigen has been suggested to enhance the persistence of bovine intrammamary infection; however, their role as virulence factor in bovine mastitis is still speculative.32,33
In this study genes encoding for enterotoxins, of relevance in food safety, were identified. Indeed, six of our strains (40%) harbored genes included into the egc operon (seg-sei-sem,seu or seg-sei-sem-seo,seu or sen), but no strain showed the complete operon (seg, sei, sem, sen, seu, and seo). Similar data about the absence of one or more genes in the egc cluster have been previously reported. Moreover, other studies also reported that the distribution of the SE-encoding genes might vary among S. aureus strains isolated from bovine mastitis. Thus, the prevalence of SE encoding genes may vary depending on the geographical location. 34 The presence of isolates with genes encoding toxins involved in human infections in the milk of cows with mastitis might represent a potential risk for human health. In our study, MSSA isolates harbored more virulence genes than MRSA isolates. The lukF-lukS-PV, etA, and etB genes have not been detected in S. aureus strains isolated from bovine mastitis in previous studies. 8
Interestingly, most of the MSSA isolates (11/12) and all MRSA isolates of this study lacked the IEC system; only one strain of lineage ST4119 carried the IEC gene cluster (type D). The IEC system is a set of genes that allows evasion of the human defenses that are associated with strains of human origin. Its presence or absence in S. aureus strains isolated from milk mastitis bovine is very important, since it gives us information about the possible origin of the isolates (human or animal origin). 10 As expected, most of our MSSA/MRSA isolates lacked the IEC human marker; we cannot exclude the human origin of the MSSA strain IEC-positive.
In this study, molecular typing revealed a high diversity of new genetic lineages among the S. aureus isolates in MSSA and MRSA strains that cause bovine mastitis. Eight new different STs were identified and they represented a single and double locus variants of ST97, ST700, ST351, ST124, or ST487 (Table 2). The spa-type t267 was the most prevalent among our isolates, similar data were reported in a previous report in India. 35
Among MRSA, two strains were typed as t267 with a new ST4120, and were isolated from two cows of the same farm; the other MRSA strain, typed as t10381/ST4114, was obtained from another farm. It is interesting to remark that six MSSA isolates recovered from milk samples of mastitis cows of four different farms, were also typed as t267-ST4120; It appears that this specific clone, both as MRSA and MSSA, is circulating in dairy farms causing bovine mastitis in Northern Tunisia. CC ST4120 is a single locus variant of ST124, included in CC5, being this CC frequently detected in human infections. 36 MSSA of lineages CC5-ST5/-ST8/-ST15 and MRSA of lineages CC5-ST5/-ST8 appears to occur frequently in African hospital environments, with CC5 being the most prevalent CC for MSSA and MRSA across Africa. 36
In addition, three new STs detected in one MRSA (ST4114) and two MSSA strains (ST4112/ST4113) represent single or double locus variants of ST97 related to CC97. According to a recent study, the lineage ST97 can evolve to various ST through active locus mutation; in this sense, ST71 may have originated from ST97, and both ST71 and ST97 belong to CC97, which includes mainly strains of clinical and subclinical bovine mastitis all over the world. 37
In Tunisia, few reports have described the detection of S. aureus in samples of non-human origins. For instance, in healthy animals (such as sheep, donkey, pet) the CCs identified from MRSA and MSSA isolates were CC1, CC6, CC22, CC72, CC130, CC80, or CC522, among others.12,28 Moreover, MSSA isolates of CC97 have been previously reported in healthy bovine milk in Tunisia. 30
In relation with MRSA of human origin in Tunisia, the lineage CC80 was mostly found among clinical isolates, although other lineages have been also reported, as CC5, CC8, or CC45, among others.38–41 Moreover, MRSA-ST80-t203 has been detected in healthy humans. 26
In this study, we found that similar to what was reported previously, all our MRSA isolates belonged to agr-type I.21,42 It has been previously found that S. aureus strains belonging to agr group I have a greater ability of invading epithelial cells and persistence in the mammary gland. This suggests that agr-I containing strains might be more effective at causing clinical or subclinical mastitis than strains of other groups. 43
Conclusion
To conclude, we show that S. aureus, both MSSA and MRSA, is a cause of bovine mastitis in Tunisian dairy farms. These isolates were genetically diverse, with seven new STs, and presented low resistance rates for most antibiotics, except β-lactams. Many MSSA isolates carried the tst and/or enterotoxin genes that are responsible for toxic shock syndrome and food poisoning. Our study gives important information to know the implication of S. aureus in bovine mastitis at dairy farms in the analyzed region, which is important for designing strategic plans for surveillance, control, prevention, and treatment of staphylococcal mastitis. Our findings emphasize the risk of transmission of these microorganisms to humans through direct contact with cows or through the consumption of milk with mastitis and its serious effects on food safety and public health.
Ethics Statement
The study underwent ethical review and was given approval by the Bio-Medical Ethics Animal Committee at Pasteur Institute of Tunis.
Footnotes
Acknowledgments
This work was supported by internal collaborative project of Pasteur Institute of Tunis and the Tunisian Ministry of Higher Education, Scientific Research and Technology (LR11IPT03). The experimental part performed in the University of La Rioja (Spain) was financed by project SAF2016-76571-R of the Agencia Estatal de Investigación (AEI) of Spain and the Fondo Europeo de Desarrollo Regional (FEDER). Sara Ceballos has a predoctoral fellowship of the University of La Rioja (Spain). We thank Dr. R.B. Elandolsi Laboratory of Epidemiology and Veterinary Microbiology of Pasteur Institute of Tunis for providing samples.
Disclosure Statement
No competing financial interests exist.
