Molecular Characterization of Antimicrobial Resistance and Virulence Genes of Bacterial Pathogens from Bovine and Caprine Mastitis in Northern Lebanon

Mastitis is an infectious disease encountered in dairy animals worldwide that is currently a growing concern in Lebanon. This study aimed at investigating the etiology of the main mastitis-causing pathogens in Northern Lebanon, determining their antimicrobial susceptibility profiles, and identifying their antimicrobial resistance (AMR) genes. A total of 101 quarter milk samples were collected from 77 cows and 11 goats presenting symptoms of mastitis on 45 dairy farms. Bacterial identification was carried out through matrix-assisted laser desorption/ionization-time of flight mass spectrometry. Antimicrobial susceptibility was tested by disc diffusion and broth microdilution methods. Molecular characterization included polymerase chain reaction (PCR) screening for genes encoding extended-spectrum beta-lactamases (ESBLs) and plasmid-mediated AmpC among Enterobacterales isolates, and virulence factors among Staphylococcus isolates. Escherichia coli isolates were subjected to phylogenetic typing by a quadruplex PCR method. The most frequently identified species were Streptococcus uberis (19.2%), Streptococcus agalactiae (15.1%), E. coli (12.3%), and Staphylococcus aureus (10.96%). Gram-positive bacteria were resistant to macrolides and tetracycline, whereas gram-negative bacteria displayed resistance to ampicillin and tetracycline. Two ESBL genes, blaTEM (83.3%) and blaOXA (16.7%), and one AmpC beta-lactamase gene, blaCMY-II (16.7%), were detected among six E. coli isolates, which mainly belonged to phylogenetic group B1. Among Staphylococcus spp., the mecA gene was present in three isolates. Furthermore, four isolates contained at least one toxin gene, and all S. aureus isolates carried the ica operon. These findings revealed the alarming risk of AMR in the Lebanese dairy chain and the importance of monitoring antimicrobial usage.


Introduction
Mastitis is by far the most widespread and costly disease in dairy cattle worldwide. It has severe and significant economic consequences that include losses in milk production, adverse health effects, early, mandatory culling, costly veterinary services, and additional labor for animal care costs [1][2][3]. It is a well-established disease that affects a high percentage of dairy cows and is of particular concern for farmers in both developed and developing countries [4,5]. Goat farms also suffer economic losses when faced with the burden of mastitis, which is mainly caused by staphylococcal intramammary infections [6,7].
North Lebanon is home to extensive dairy farming and production of milk and its derivatives, which constitute a fundamental part of the Mediterranean diet. Dairy products are consumed daily by different age categories. The dairy sector in Lebanon is also of important economic significance. It represents a leading source of income for producers and families in rural communities, as well as a beneficial opportunity for labor. Approximately 60% of livestock farmers rely on dairy as their main source of income [8]. The development of the dairy sector in Lebanon has led to excessive use of antimicrobials to improve animal health by treating clinical signs of infectious diseases, such as mastitis, and therefore increase their productivity [9]. Several antimicrobial compounds are also currently used as growth promoters in animal production [10].
Mastitis was reported as the most common cause for antimicrobial administration in farming [11]. Because it is so widespread, mastitis is a concerning issue in Lebanon, especially since local farmers and even veterinarians treat infected animals without the necessary support from laboratory services. Conventional antimicrobials used for mastitis therapy are consequently showing reduced effectiveness in dairy cattle. The prolonged misusage of antimicrobial compounds for the treatment of mastitis has contributed to the emergence of antimicrobial resistance (AMR) [12].
Unfortunately, there is neither enough data on the distribution and circulation of pathogens involved in mastitis and their susceptibility patterns to antimicrobials nor on antimicrobial usage in livestock treatment in Lebanon. However, the limited available reports showed high levels of coliform bacteria contamination in traditional Lebanese dairy products and many of these isolates are associated with multi-drug resistance patterns [13,14]. A cross-sectional study also indicated that 27.9% of raw milk samples in Lebanon are contaminated with extended-spectrum beta-lactamase (ESBL)-and/or carbapenemase-producing Enterobacterales [14]. Moreover, a high prevalence of ESBLproducing Escherichia coli isolates (67.5%) was reported among healthy cattle as well as a remarkable prevalence of ESBL-positive cattle farms (84%) [15]. Similarly, Zouhairi and colleagues stated that 98.7% of Staphylococcus isolates from Lebanese dairy-based food products were resistant to at least one antimicrobial drug, with 94% being methicillinresistant [16]. Indeed, contaminated dairy products may harbor these antimicrobial resistant microorganisms and may pose a threat to public health by acting as vehicles of transmission of these pathogens, which will later be responsible for food-borne diseases. On the other hand, a recent investigation showed the poor knowledge and irresponsibility of the Lebanese population regarding antimicrobial misuse and resistance [17]. What makes the situation more alarming is the wide availability of antimicrobials in the Lebanese market without the requirement of a veterinarian's prescription [10,18,19].
In this context, the purpose of the current study was to isolate and identify the main pathogenic bacteria that cause clinical mastitis in dairy farms in Northern Lebanon and to determine their susceptibility patterns to the most frequent classes of antimicrobials utilized in veterinary management. This study also investigated the presence of genes related to AMR, and identified the phylogenetic groups of ESBL-and AmpC-positive Escherichia coli isolates. We also determined the mosaicism of specific virulence genes among Staphylococcus isolates.

Survey Design
A questionnaire-based survey was performed during March and April 2019 on 45 dairy farms located in the North of Lebanon. For each participating dairy farm, a record sheet registering general information about the farm and a short description of the herd (farm name or area/district, species, clinical signs, percentage of suspected animals, therapy history, and laboratory tests) was collected at the time of inclusion.

Sample Collection
During March and April 2019, 101 mastitis milk samples (90 cow milk samples and 11 goat milk samples) were aseptically collected from 77 dairy cows and 11 goats showing symptoms of infectious mastitis. Multiple milk samples were at times collected from different quarters of the same cow. Samples were taken from a total of 45 dairy farms located in 32 villages in the Northern region of Lebanon. The villages in which the dairy farms are located are reported in Figure 1. Field veterinarians were responsible for milk sampling. The material supplied for aseptic sampling included disposable gloves and towels, alcohol, cotton, 15 mL sterile sampling tubes, and a specific record sheet for each milk sample. Briefly, teats were cleaned and disinfected using alcohol, the first three streams were discarded, and then milk samples were collected in sterile 15 mL conical tubes. The cooled milk samples were transported on the same day in iceboxes directly to the microbiology laboratory and then processed the same day.

Sample Collection
During March and April 2019, 101 mastitis milk samples (90 cow milk samples and 11 goat milk samples) were aseptically collected from 77 dairy cows and 11 goats showing symptoms of infectious mastitis. Multiple milk samples were at times collected from different quarters of the same cow. Samples were taken from a total of 45 dairy farms located in 32 villages in the Northern region of Lebanon. The villages in which the dairy farms are located are reported in Figure 1. Field veterinarians were responsible for milk sampling. The material supplied for aseptic sampling included disposable gloves and towels, alcohol, cotton, 15 mL sterile sampling tubes, and a specific record sheet for each milk sample. Briefly, teats were cleaned and disinfected using alcohol, the first three streams were discarded, and then milk samples were collected in sterile 15 mL conical tubes. The cooled milk samples were transported on the same day in iceboxes directly to the microbiology laboratory and then processed the same day.

Culturing Techniques and Species Identification
Raw milk samples were examined microbiologically by plating 10 µL of each sample on the following media: Columbia agar with 5% blood; and Columbia CNA Agar with 5% blood, drigalski agar, and mannitol salt agar (all from BioRad, Marnes-la-Coquette, France). The plates were incubated at 37 • C for 24-48 h. Blood agar plates were incubated in 5% CO 2 . The culture was considered positive if there was a growth of individual bacteria in a concentration of more than 10 4 CFU/mL. All isolates were identified and characterized by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) using a VITEK MS instrument (bioMérieux, Marcy l'Etoile, France).

Antimicrobial Susceptibility Testing
According to the Clinical Laboratory Standard Institute (CLSI) recommendations, bacterial isolates were tested for their antimicrobial susceptibility by standard disk diffusion method either on Mueller-Hinton agar or Mueller-Hinton agar with 5% blood added. The inhibition zones were measured, recorded, and interpreted according to the CLSI guide-  , and enrofloxacin (ENR) (0.12-2). The plates were incubated aerobically at 37 • C for 24 h (48 h for Streptococcus spp.) and were read manually using the Thermo Scientific™ Sensititre™ Manual Viewbox.

Detection of AMR Genes
The primers and their sequences used in the PCRs are listed in Table S1 of the Supplementary Materials. DNA was extracted by a boiling technique. A loop of bacterial colonies was put in a test tube containing 1 mL of sterile 10 mM TRIS 1 mM EDTA (T.E) buffer 0.1% and heated in the Dry Bath (EuroClone, Milan, Italy) at 95 • C for 30 min. The solution was then centrifuged for five minutes at 1000 rpm. The supernatant was collected and used for the downstream process of PCR. Two multiplex PCR reactions were performed to amplify ESBL and AmpC beta-lactamase genes in E. coli and Klebsiella oxytoca isolates, as described by [20]. The first multiplex assay (named Set I) was designed to detect TEM, SHV, CTX-M IV group, and OXA beta-lactamase encoding genes, and the second assay (named Set II) was signed to detect CTX-M I group, CTX-M II group, CMY II, and DHA encoding genes. Additionally, Staphylococcus spp. oxacillin-resistant isolates were tested for the presence of the mecA and mecC genes for confirmation of methicillin resistance.

Determination of Phylogenetic Groups
The distribution of phylogenetic groups amongst ESBL-and/or AmpC-producing E. coli isolates was determined as described by Clermont and colleagues [21]. Eight phylogroups are now recognized: seven (A, B1, B2, C, D, E, F) belong to E. coli sensu stricto, whereas the eighth is the Escherichia cryptic clade I. Phylogenetic groups were identified by multiplex PCR based on the presence or absence of four DNA markers (chuA, yjaA, DNA fragment TSPE4.C2, and arpA).

Detection of Biofilm and Virulence Associated Genes of Staphylococcus Spp.
The capability of Staphylococcus spp. to form biofilm is considered an additional virulence factor in mastitis cases, due to its ability to adhere to and persist in tissues and the environment [22]. Thus, two separate reactions of PCR were performed to amplify biofilm-associated protein (bap) and fragments of the intracellular adhesion (ica) genes involved in biofilm formation in Staphylococcus spp. as previously described [23]. Moreover, multiplex PCR assays were carried out to amplify staphylococcal enterotoxins (sea-see, seg, seh, sei, sej, and sep), exfoliative toxins (eta and etb), and toxic shock syndrome toxin 1 (tsst-1) genes, as recently described in [24].

Main Findings of the Questionnaire-Based Survey
Our comprehensive questionnaire-based survey revealed that the most prevalent form of mastitis encountered in our study area was clinical mastitis (92%; 81/88), followed by subclinical mastitis (8%; 7/88). Animals had at least one of the following symptoms: abnormal milk appearance (presence of blood in milk and/or watery to viscous milk with clots varying from gray-white to yellowish) (60.2%), hardness and swelling of the udder (47.7%), and/or reduced milk production (45.4%), inappetence (3.4%), diarrhea (2%), chronic inflammation (1.1%), and/or anorexia (1.1%). Empirical antimicrobial treatment was reported in four farms using cefaclor (Cefatek ® ), amoxicillin, tylosin, colistin, gentamicin (10% intramuscular injection), and/or a combination of penicillin and kanamycin (Penikan P ® ). No laboratory tests were performed before and after antimicrobial prescription, suggesting a potential misuse of drugs.

Molecular Characterization of Clinical Isolates
Out of the E. coli isolates, five and one were positive for bla TEM and bla CMY-II genes, respectively. One bla TEM producing E. coli carried an additional bla OXA gene. However, the K. oxytoca isolate did not harbor any ESBL or AmpC beta-lactamase encoding gene. Phylogrouping on beta-lactamase-producing E. coli isolates revealed the predominance of the phylogenetic group B1(4/6; 66.7%), followed by B2 and D (1/6; each). On the other hand, methicillin resistance was confirmed by the detection of the mecA gene among the three staphylococcal isolates resistant to oxacillin (two S. aureus and one S. haemolyticus). None of the isolates harbored the mecC gene.
This report also investigated the presence of adhesin genes and showed that 40% of staphylococcal isolates were positive for at least one of the classical enterotoxin genes, and two S. aureus isolates harbored one or more enterotoxin genes. One S. aureus isolate harbored simultaneously the toxic shock syndrome toxin 1 (tsst-1) and the sec enterotoxin gene. Regarding biofilm formation genes, all the tested isolates were negative for bap gene, but all the S. aureus isolates were icaA positive ( Table 3). None of the isolates harbored the exfoliative toxin genes.

Discussion
AMR is a growing concern in both human and veterinary medicine in Lebanon [10,18,19]. The extensive use of broad-spectrum antimicrobials in farm animals, the food industry, and in human and veterinary medicine, has led to the emergence of MDR organisms in humans, animals, and the environment [9,14,15,18,[26][27][28]. The results of the present study revealed that most E. coli isolates tested were MDR. Traditional farming and poor hygiene husbandry, typical of Mediterranean countries including Lebanon, enable the quick spread of resistant micro-organisms, which represents a major challenge for developing countries, including Lebanon, as it severely affects the quantity, quality, and safety of animal and food production with significant economic and social consequences [9,10].
Bacterial infections are the predominant cause of bovine and caprine mastitis. In this study, we reported 88 cases of clinical and subclinical mastitis. Only 10% of farms admitted recent administration of antimicrobials. Although 90% of farms did not report the use of antimicrobials, a non-negligible proportion of farmers are not aware of the proper use of antimicrobials or their effects on animals and perform false practices, which are at the root of the spread of AMR [9,10,17]. Several antimicrobials are readily and legally available in the Lebanese market without the requirement of a veterinarian's prescription [10].
The analysis of bacterial isolates from bulk milk samples showed the predominance of gram-positive bacteria, particularly S. uberis, S. agalactiae, and S. aureus, followed by E. coli. The distribution of mastitis bacterial pathogens varies between different geographic areas and even countries. For example, S. aureus and coagulase-negative staphylococci (CNS) have been reported as the most common bovine mastitis in the Middle East, North Africa, and Europe [29][30][31][32][33][34][35][36][37]. In the same context, a large-scale epidemiological study conducted in Italy reported the predominance of S. agalactiae and S. aureus in dairy herds [38]. Moreover, studies performed in European Mediterranean countries highlighted a significant presence of E. coli and Staphylococcus spp. [39][40][41][42]. Moreover, a high incidence of E. coli was also observed among bovine mastitis in India and Taiwan [43,44]. In addition, one mastitis case of Streptococcus pneumoniae was documented. Pneumococcus is a commensal bacterium that normally colonizes the human nasopharyngeal cavity and is transmitted by droplets and aerosols either from infected patients or healthy carriers [45]. Although this species is rarely reported in mastitis, various epidemiological studies showed its high prevalence in the Lebanese community [46,47]. This allows us to speculate that there is a potential threat of contamination by handlers. Antimicrobial susceptibility testing showed a high percentage of resistance among most gram-positive and gram-negative isolates against both oxytetracycline and chlortetracycline. This is in line with a previous report which described that 98.4% of streptococcal isolates from dairy cows with mastitis in China were resistant to tetracycline [48]. In Lebanon, a similar percentage of resistance was observed.
Most E. coli (82.5%) isolated from healthy adult cattle showed tetracycline resistance [15]. Numerous veterinary infectious diseases among cattle are the most frequently treated with oxytetracyclines, explaining the high resistance rates observed in this study [49].
In contrast to streptococcal isolates, which showed very low susceptibility results to erythromycin and clindamycin, none of the staphylococcal isolates showed resistance to these antimicrobials. Although no similar previous study was found in Lebanon, our findings are consistent with similar data reported across the world [42,50,51]. In the human field in Lebanon, a recent cross-sectional study displayed the absence of macrolide and lincosamide resistance in both MRSA and methicillin-sensitive S. aureus (MSSA) isolates among food handlers [52]. However, previous Lebanese studies on group A streptococci human isolates showed lower resistance to erythromycin and clindamycin [53,54].
MRSA was isolated twice (25%) in this study. This percentage is consistent with clinical human studies at the nationwide level showing a similar prevalence of this resistance pattern. A compilation of antimicrobial susceptibility data of S. aureus from a network of 13 Lebanese hospitals indicated that 28% were MRSA [55]. Moreover, a previous study conducted in the same geographic area reported that 23.8% of S. aureus isolates were MRSA [52]. However, MRSA prevalence in the present study is significantly higher than that of other countries such as Croatia [56] and Ukraine [57].
Regarding E. coli, 100% and 11.1% of isolates were resistant to ampicillin and ceftiofur, respectively. Despite this, only one isolate presented resistance to ceftiofur using the phenotypic method. Third-generation cephalosporins resistance was screened by two multiplex PCR reactions as described by [20]. Overall, 55.5% of E. coli isolates were positive for ESBL and plasmid-mediated AmpC beta-lactamase genes: bla TEM , bla CMY-II , and/or bla OXA . Unfortunately, due to logistical reasons, this study did not evaluate the susceptibility of isolates to other third-generation cephalosporins such as cefotaxime, ceftriaxone, ceftazidime, and cefixime. This study confirmed that food animals and foodstuffs are a well-known reservoir of ESBL-producing E. coli in Lebanon. Wide dissemination of third-generation cephalosporin resistance, coupled with resistances to carbapenems, colistin, and numerous other antimicrobial compounds, has been reported among animals in the last decade [9,14,15,18,58]. The phylogenetic typing analysis carried out on the ESBL-producing E. coli isolates revealed that most of them belong to phylogroups B1 (66.7%) followed by both B2 and D (16.7% each), which have previously been reported in clinical third-generation cephalosporin-resistant isolates in Lebanon [19]. Phylogenetic grouping is an important approach to understanding the pathogenicity and evolutionary relationships between different strains [59]. No association was identified between AMR and phylogroup due to the limited number of E. coli isolates. Thus, a larger study on a higher number of isolates is needed. At the worldwide level, the majority of E. coli mastitis strains belong to phylogenetic groups A and B1 [60]. A high frequency of phylogroup B1 was also observed in an Iranian study evaluating the phylogeny of E. coli isolated from clinical mastitis [61]. A Brazilian study revealed that most E. coli isolates from bovine mastitis belonged to phylogenetic group A (52%), followed by B1 (38%) [62]. Similarly, other studies from Ireland, Switzerland, Serbia, and China have noted that E. coli isolates were mainly assigned to the phylogenetic groups A and B1 [63][64][65][66][67].
The ability of S. aureus isolates to form biofilms is an important mechanism that reinforces pathogenicity and contributes to AMR. PCR results showed that all S. aureus isolates were positive for the intracellular adhesion gene icaA, while the bap gene was not identified in any of the isolates. Our findings follow previous studies conducted in Poland [68], New Zealand [22], and the United Kingdom [69] that detected the unique presence of the icaA gene among all S. aureus isolates from bovine mastitis. Our study also showed that 40% and 25% of staphylococcal and S. aureus isolates were positive for at least one enterotoxin gene, respectively. Interestingly, one S. aureus isolate simultaneously harbored tsst-1 and sec genes, predicting the presence of the staphylococcal pathogenicity island 1 (SaPI1), as described previously [70].

Conclusions
This study provides a consistent source of relevant data on AMR levels and trends of mastitis bacterial pathogens in Lebanese dairy bovines and caprines. Despite resistance rate variations, our findings confirmed the wide dissemination of antimicrobial-resistant bacteria in the Lebanese dairy industry. Many causative agents isolated in this study are zoonotic and can be transmitted directly between animals and humans or through the food chain. Therefore, there is a drastic need for a national strategy based on the one health approach to address the AMR issue and regulate the usage of antimicrobials in the veterinary sector in Lebanon, including bans on over-the-counter drugs and growth promoters. Additionally, the training and education of farmers is necessary to create awareness of AMR through effective communication, education, and training. Further large-scale, nationwide studies are also needed to evaluate the correlations between the use of antimicrobials in common husbandry practices and the onset of AMR in Lebanon.