Isolation and characterization of E. coli strains causing intramammary infections from dairy animals and wild birds

ABSTRACT The study was conducted to estimate the prevalence of Escherichia coli (E. coli) in sub-clinically mastitic (SCM) animals, and in wild and migratory birds which may act as reservoir disseminating such pathogen. Farm hygiene, management and milking procedures were listed through a questionnaire. Thirty lactating cows and 15 lactating buffaloes from five small-scale dairy farms were randomly selected and screened for subclinical mastitis (SCM) using California Mastitis Test (CMT) and somatic cell count (SCC). In addition, 80 teat skin swabs, 5 drinking water samples and 38 wild and migratory bird faecal matter were also collected. All samples were processed for E. coli isolation by culturing on Levine’s Eosin Methylene Blue (L-EMB) agar, followed by purification and biochemical identification. Positive samples were subjected to molecular identification and serotyping. In addition, the presence of extended-spectrum beta-lactamase (ESBL) and carbapenemase-producing E. coli have been reported by antimicrobial sensitivity testing. Escherichia coli were isolated from 7.7%, 50% and 50% of the positive CMT cows’ quarters, cows’ composite and buffaloes’ composite milk samples, respectively. In addition, 14% of cows’ teats, 20% of water samples, 70% of faecal matter from wild bird, and 33.3% of faecal matter from migratory waterfowls were carrying E. coli. Serotyping, antibiotic-resistant pattern and phylogenetic analysis have pointed the bearable implication of milking hygiene and wild birds in disseminating E. coli strains causing intramammary infections.


Introduction
Bovine mastitis continues to be a major infection that affects a high proportion of dairy populations worldwide with a negative economic impact on dairy farmers and the dairy industry [1][2][3]. Furthermore, it possesses a public health concern [4], besides it is considered the single most prevalent cause for antibacterial use in lactating dairy animals [5]. The types of mastitis pathogens are either contagious or environmental organisms [6]. Coliform bacteria are a common aetiology of bovine mastitis. Until recently, Escherichia coli (E. coli) is the most predominant species and is responsible for more than 80% of coliform mastitis cases [3,7,8] and categorized as a Gramnegative opportunistic environmental bacterium. Mastitis caused by E. coli is usually sporadic with clinical signs range from mild to very severe or even fatal forms [9]. Properties and virulence factors specific to E. coli strains causing mastitis are still ill-defined, although a high prevalence of phylogroup A was observed in mastitis caused by E. coli [10][11][12][13]. E. coli is mainly a commensal bacterium in the intestinal tract of animals, birds and humans [14]. Although the establishment of biosecurity measures to prevent the entry of infectious agents into dairy farms, dairy herd health still faces several serious threats, regarding open space, which allows uncontrolled access of wild birds into yards, facilities and even feed storage areas [15]. However, few studies focused on wild birds rooming nearby dairy farms [16]. Migratory and nonmigratory wild birds serve as reservoirs of coliform bacteria, like E. coli, inclosing antimicrobial-resistance genes [17,18]. Water and food contact were suggested to be the primary routes of transmission of resistant bacteria between human or animal and wild birds [19]. Antibiotics are very remarkable for the treatment of bacterial infections in animals and humans [20]. The FDA reported that resistance in E. coli is steadily the highest for antimicrobial agents used in human and veterinary medicine [21]. Extended-spectrum betalactamase (ESBL) producing E. coli is mostly insensitive to lots of commonly used antibiotics causing an increase in the use of last-resort antimicrobial drugs (i.e. carbapenems) during treatment.
The aim of this study was to understand the possible role of wild and migratory birds in carrying and spreading mastitis-causing E. coli strains to the dairy farm environment which could represent potential veterinary and public health hazards through infecting the animals and contaminating the milk with antimicrobial-resistant avian strains, in a coincidence of poor milking hygiene.

Teat swabs
A total of 80 teat skin swabs were collected from 25 cows and 15 buffaloes (40 Hind and 40 fore teat swabs) by a cotton-tipped stick through rotating the cotton-tip on the teat barrel, as described by Piccinini et al. [24].

Drinking water samples
Five water samples (50 ml from each farm) were collected from the drinking troughs of the lactating animals into disposable sterilized test tubes. All samples were stored in an ice box until transported to the laboratory.
Port-Said is one of the Canal Zone governorates and considered a main stop for migratory waterfowls seeking food and rest during the long migration in the spring and fall seasons.
Bird traps were used to capture wild birds around dairy farms and migratory waterfowls during winter migration. Once trapped, faecal swabs were taken, and the birds were released. The swabs were then placed in 2 ml sterile saline (0.9% NaCl) and stored in ice box until transported to the laboratory.
Protocols for the collection of samples were conducted in accordance with the applicable legislation of the Institutional Animal Care and Use Committee, of the Faculty of Veterinary Medicine, Cairo University, Egypt (VetCU10102019087).

Isolation and identification of coliform and E. coli
Loopfuls from positive CMT milk samples were enriched into nutrient broth and incubated aerobically at 37°C for 24 h. The enriched milk samples, teat swabs and loopfuls from water samples were streaked on Violet Red Bile agar (VRB) for isolation of coliforms. Additionally, these samples beside the faecal samples from wild and migratory birds were plated on Levine's Eosin Methylene Blue plates (L-EMB) for isolation of E. coli and incubated at 37ºC for 24-48 h. Suspected colonies of E. coli which appeared as dark centred, flat colonies with metallic green lustre were picked up and cultured on agar slants and incubated at 35°C for 18 h and subjected to further identification. Biochemical identification of E. coli was carried out according to BAM and Quinn et al. [25,26].
A total of eight representative E. coli isolates from buffalo's milk, cow's milk, teat skin swabs, drinking water, Laughing dove, Hooded crow, Eurasian coot, Northern shoveler faecal swabs that were isolated from the same locality were selected and serologically identified by slide agglutination test for O-antigen group screening using E. coli antisera (Denka Seiken Co., LTD. Chuo-Ku, Tokyo, Japan) according to Kok et al. [27].

Molecular identification of E. coli
All E. coli isolates were extracted using DNA Mini Kit (Qiagen, Hombrechtikon, Switzerland) according to the manufacturer's protocol. For molecular identification of E. coli, Primers for 16S rRNA gene of E. coli were selected according to Wang et al. [28]. The reaction was carried out using a total volume of 25 μl contain 3 μl of DNA as a template, 5 pmol of each primer and 5μl of 1X PCR master mix (Jena bioscience, GmbH, Germany). The PCR mixtures were then subjected to the following cycling conditions: 50°C (2 min, 1 cycle); 95°C (5min, 1 cycle); 40 cycles of 95°C (45 s), 50°C (1 min), and 72°C (1 min); and 72°C (7 min, 1 cycle) in a thermal cycler (Perkin-Elmer, Waltham, USA).

Antibiotic sensitivity test
The antimicrobial sensitivity for the fore mentioned eight serotyped isolates was evaluated using eight antimicrobials that represent three groups: Carbapenems (Imipenem, Ertapenem, Meropenem), Cephalosporins (Ceftazidime, Cefotaxime, Ceftriaxone, Cefpodoxime) and Macrolides (Azithromycin). ESBL production was confirmed by the increase of ≥5 mm in the zone of inhibition for Cefotaxime/ Clavulanic acid (30/10 mcg) and Ceftazidime/ Clavulanic acid (30/10 mcg) discs compared to cefotaxime or ceftazidime alone. The interpretation was done according to Clinical and Laboratory Standards Institute (CLSI) [29].

Sequence analysis
The amplified 16S rRNA fragments of these eight isolates were purified using the QIAquick gel extraction kit (Qiagen, Hombrechtikon, Switzerland) according to the manufacturer's instructions and sequenced at Promega Lab Technology (Madison, USA) by using the forward and reverse primers listed in table (Table 1). The sequences of the 16S rRNA gene have been deposited in the GenBank database under the accession numbers mentioned in Table 4. Genes sequenced in this study were compared with the sequences available in public domain using NCBI BLAST server. The 16S rRNA gene sequences were aligned using CLUSTALW in BioEdit version 7.0.1.4. Phylogenetic analysis was performed with MEGA version 7, using the neighbour-joining approach. The bootstrap consensus tree was estimated from 1000 replicates.

Statistical analysis
Results were analysed using PASW Statistics, Version 18.0 software (SPSS Inc., Chicago, IL, USA). Somatic cell counts were compared by independent sample T-test and expressed as mean ± standard error. Frequencies of isolation were tested with the Chisquare (x 2 ) (provided that at least 80% of the cells have an expected frequency of 5 or greater, and that no cell has an expected frequency smaller than 1.0), otherwise, the Fisher's Exact test and the Fisher-Freeman-Halton Exact test (it is the Fisher's Exact test for contingency tables larger than 2 × 2) were used. A P-value <0.05 was considered statistically significant.

Field observation
Dairy farms included in this study raised from 10 to 100 animals. Animal breeds were Holstein Friesian cows and native buffaloes. Cows age ranged from 6 to 10 years, while buffaloes were around 12 years old. Animals were housed in open yards surrounded with a fence, partially covered with sheds and muddy floors and each yard contains one common water trough for drinking. Each yard is occupied by 10 to 25 lactating cows or buffaloes. All animals were automatically milked twice daily either in-stanchion (10 animals) or in milk parlours (40 to 100 animals). Pre-milking practice was restricted to udder washing and wiping without pre-milking sanitizing teat dip. Only, two farms applied post-milking teat dipping. Results demonstrated that somatic cell count (SCC) in cows' quarter milk samples was 8.5 × 10 5 (±1.9 × 10 5 , ±95% C.I.) cells/ml significantly greater than that in composite milk samples (t(69.79) = 8.838, p< 0.001) ( Table 2). While, somatic cell count (SCC) in composite milk samples of cows was 30 × 10 4 (±1.9 × 10 5 , ±95% C.I.) cells/ml numerically but not significantly greater than that of buffaloes (t(15) = 0.333, p= 0.744) ( Table 2). According to the Egyptian standards of raw milk [30], subclinical mastitis is considered when SCC exceeds 500 × 103 cells/ ml. Consequently, 88.3% of examined cow quarters recorded unacceptable SCC, as well as, 46.7% of cows' composite and 50% of buffaloes' composite milk samples (Figure 2). Subclinical mastitis in cows was significantly more likely to be detected in quarter milk samples than in composite ones (p =0.002, Fisher's exact test, FET).  (Table 2). Escherichia coli (E. coli) was significantly more likely to be isolated from composite milk samples than from quarter milk samples in cows (p =0.004, Fisher's exact test, FET).

Teat skin swabs and drinking water
Coliforms were isolated from 19 (38%) and 6 (20%) teats of cows and buffaloes, respectively; a difference that was not statistically significant, x 2 (1, N = 80) = 2.828, p= 0.093. Further examination resulted in the isolation of E. coli from 7 (14%) of cow's teat skins, but could not be isolated from buffaloes' teat skins, as listed in Table 3. Coliforms and E. coli were isolated from the one-water sample (20%).

Faecal samples of resident wild birds and migratory waterfowls
Bacteriological examination of 38 faecal samples from wild birds and migratory waterfowls resulted in isolation of E. coli from 14 (70%) of resident wild birds (laughing doves "Spilopelia senegalensis" and hooded crows "Corvus cornix") and 6 (33.3%) of migratory waterfowls (eurasian coot "Fulica atra" and northern shoveler "Anas clypeata") ( Table 3). Isolation rate of E. coli from faeces of wild birds

Percent (%)
Milk samples with unacceptable SCC Figure 2. Prevalence of unacceptable milk SCC in examined quarters and composite milk from cows and buffaloes, according to Egyptian standards of raw milk, 2010.
was significantly higher than that from migratory waterfowls, x 2 (1, N = 38) = 5.109, p= 0.024. All E. coli isolates were molecularly confirmed by PCR using Primers for 16S rRNA gene of E. coli. Table 4 revealed different E. coli serotypes. E. coli O166 serotype was isolated from both cow's milk and laughing dove's faecal swab, and O146 serotype was isolated from both buffalo's milk and teat skin. Serotypes O1, O78, O18 and O158 were isolated from drinking water, faecal swab of hooded crow, faecal swab of eurasian coot and of northern shoveler; respectively.

Antibiotic sensitivity of E. coli isolates
The antibiogram demonstrated that the Carbapenems group of antibiotics showed the greatest efficacy against most of the E. coli isolates. Resistance to Carbapenems group appeared from the environmental E. coli serotypes isolated from teat skin and drinking water. All the eight E. coli isolates displayed resistance against both Cefotaxime and Cefpodoxime antibiotics.

Sequence analysis
Phylogenetic analysis of 16S rRNA gene sequences of these eight E coli isolates showed100% homology between isolates from northern shoveler, eurasian coot, laughing dove, hooded crow and water as shown now in the phylogenetic tree ( Figure 3).

Discussion
The Small-scale dairy farming is a significant economic sector for the enhancement of agriculture and livelihoods in developing countries [31,32]. It begins from 2, 15 up to 20 or even 50 milking animals [33][34][35]. Mastitis is a multi-factorial production disease and possess a threat for small-scale producers as its prevalence exceeded 50% according to FAO, 2014 report [36]. In subclinical mastitis, milk and udder appear normal, but there is decreased milk production, positive reactivity with CMT, high milk somatic cell count, presence of pathogens in the milk, and altered milk composition [37][38][39]. California Mastitis Test (CMT) is a common screening test used for detection of subclinical mastitis and it revealed that 43.3% of quarter milk samples were affected with subclinical mastitis. These results agreed with that reported by Sori et al. [40] (40.6%) and Ayano et al. [41] (43.02%). A higher prevalence (60%) was reported by Ayano et al. [42], and a lower prevalence ranged from 18.14% to 23.25% were reported by Bangar et al. [43], Patel and Trivedi [44], and Singh et al. [45]. On the other hand, somatic cell count (SCC) is the gold standard for confirming subclinical mastitis [46]. Healthy udder has milk SCC ranges from 50,000 and 100,000 cells/ml, up to 200,000 cells/mL, while in subclinical mastitis the SCC exceeds 200,000 cells/ml [47]. In Egypt, subclinical mastitis is confirmed when SCC exceeds 500,000 cells/ml [30]. In Table 2, the average SCC of quarter milk samples and composite milk samples were similar to the average value (9.24 ± 2.0 × 105 cells/ml) reported by Skrzypek et al. [48] and average (44.8 ± 0.26 x 104 cell/ml) reported by Mohamed et al. [49], respectively. Coliforms are environmental mastitis-causing pathogens, they are primarily present in moisture, mud, faeces and other organic matter present around the animals. Escherichia coli is the most commonly isolated coliform species from intramammary infections and clinical mastitis; besides it has a significant public health importance as it causes diarrhoea in human [40,50,51]. Prevalence of coliform infections (Table 2) disagreed with Hawari and Al-Dabbas [52] who detected coliforms in the milk of 31.9% of the mastitic quarters. On the other side, El-Khodery and Osman [53] detected coliforms in 80.36% of the affected buffaloes, but Sayed [54] recorded lower coliform prevalence (21.78%) in the examined milk samples. Escherichia coli was isolated form 7.7% quarter milk samples which is lower than that reported by Verma et al. [55] (21.28%). Escherichia coli has been detected in 50% of composite milk of buffaloes and cows which agreed with Mansour et al. [56] who found E. coli in 47% of machine-milked samples.
Poor hygiene, husbandry and milking technique could predispose to environmental mastitis as well as milk contamination [57]. Gram-negative bacteria pass into the mammary gland through teat canal [50]. Previous studies stated that regular teat dipping for controlling of mastitis is not a common practice at small-scale farms, which agrees with the findings of our study [58]. Escherichia coli was detected on the skin of 14% of cows' teats as shown in Table 3, which poses a critical threat to the animals as well as consumer health as concluded by Galiero [59]. Teat skin act as the primary reservoir of microbes that can infect milk during milking [60]. Moreover, Jones [61] showed that the incidence of clinical mastitis was related to bacterial populations on teat ends.
As E. coli is a ubiquitous organism, it colonizes the gut of birds as well as mammals [62]. Recent studies assumed that wild birds play a crucial role in spreading of foodborne bacteria among dairy farms [63,64]. The practice of feeding dairy animals in open yards and accumulation of their manure lead to the attraction of wild birds into dairy farms [65]. Results in Table 3 showed that faecal matter of 70% of wild resident birds and 33.3% of migratory waterfowls contained E. coli. Similar studies isolated identical E. coli strains from excreta of wild birds collected from two dairy farms with 32.5 km distance apart, on the same sampling time [64,66].
Serotyping of E. coli isolates from different sources in the same locality revealed that O-146 and O-166 serotypes were the most frequent strains ( Table 4). The O-146 type was isolated from both buffalo milk and teat skin and this indicated that contamination of teat skin with environmental pathogens possess a risk for occurrence of mastitis [61]. The same serotype was isolated by Wenz [12] from milk samples of cows with acute coliform mastitis and by Wang et al. [67] from ducks in China. Escherichia coli O-166 serotype was isolated from both cow milk and dove excreta which shows the probable role of wild birds in acting as reservoir and spreader of mastitis-causing pathogens among dairy farms [65][66][67]. Furthermore, the same serotype was previously reported to cause an outbreak of gastroenteritis in humans in Japan [68], which displays the significance of this serotype on the public health. Escherichia coli serotypes O-1, O-18, O-78 and O-158 which were isolated from drinking water, migratory ducks (coot), wild birds (crows) and migratory shoveler duck, respectively, were known as the most frequent Avian Pathogenic and Avian Faecal E. coli (APEC and AFEC) serotypes infecting chickens, turkeys, ducks or other birds [67,[69][70][71]. Moreover, O-18, O-78 and O-158 were previously isolated by Wenz et al. [12] from milk samples of cows suffering acute coliform mastitis with varying systemic disease severity.
The antibiotic resistance of E. coli is of a particular concern because it is the most common Gramnegative pathogen in humans [72]. In addition, resistant E. coli strains could transfer and acquire antibiotic-resistance elements to and from other bacteria [73]. In the current study, the serotyped E. coli isolates showed resistance against most of Cephalosporins group of antibiotics (Table 4). On contrary, all E. coli isolates showed considerable sensitivity to Carbapenems group of antibiotics except that isolated from teat skin and water (Table 4). Carbapenems, like imipenem or meropenem, are considered the "last line of defense" in the treatment of infections caused by multi-resistant Gram-negative bacteria. Previous studies stated that bacteria isolated from wild and migratory bird populations acquired resistance although they are not directly affected by antibiotic practices [74]. The number of studies describing the prevalence of ESBL-producing Enterobacteriaceae has increased rapidly around the world [75]. Out of the eight E. coli isolates, five ESBL producing E. coli isolates were detected (two from wild resident birds, one from migratory waterfowls, one from buffalo milk, and one from drinking water). In recent years, ESBL-producing Enterobacteriaceae have been recovered from different sources in the community, including cattle, chickens, and raw milk [76][77][78]. Interestingly, ESBL-producing E. coli in wildlife was firstly reported shortly after their isolation from livestock farming which could suggest the spread of ESBL-E. coli into the environment through manure [79], and subsequently spread to different ecological niches [80].
These results were also confirmed by sequence analysis of 16S rRNA gene of these eight isolates. The phylogenetic analysis revealed that the sequence from northern shoveler, eurasian coot, laughing dove, hooded crow and water were identical to each other suggesting the possible role of migratory birds to contaminate water sources in the farm since they were found in the same clade.

Conclusion
The study showed that faecal matter of wild and migratory birds could be considered as a potential risk factor for disseminating pathogenic and/or resistant E. coli in a dairy farm environment, leading to contaminating teat skin of lactating animals, triggering mastitis and/or milk contamination, and threatening animal and human health. Frequent removal of manure could lessen wild bird density inside dairy farms. Milking hygiene practices, as; washing worker hands, rinsing udder and teats in sanitizing solution, then drying them, followed by effective germicidal teat dip could reduce pathogenic bacterial load on the udder skin and prevent mastitis-causing pathogens from populating the environment or teat skin of the animals. Further research should be carried out to explore other environmental risk factors responsible for intramammary infections within dairy herds.

Disclosure statement
No potential conflict of interest was reported by the authors.