Occurrence of Virulence Genes Associated with Diarrheagenic Escherichia coli Isolated from Raw Cow’s Milk from Two Commercial Dairy Farms in the Eastern Cape Province, South Africa

Escherichia coli remains a public health concern worldwide as an organism that causes diarrhea and its reservoir in raw milk may play an important role in the survival and transport of pathogenic strains. Diarrheagenic E. coli strains are diverse food-borne pathogens and causes diarrhea with varying virulence in humans. We investigated the prevalence of pathogenic E. coli in raw milk from two commercial dairy farms. Four hundred raw milk samples, 200 from each dairy farm, were screened for the presence of fliCH7, eagR, ial, eagg, lt, and papC genes. In dairy farm A, 100 E. coli were identified based on culture, oxidase and Gram staining, while 88 isolates from dairy farm B were identified in the same manner. Gene detection showed fliCH7 27 (54%) to be the highest gene detected from farm A and lt 2 (4%) to be the lowest. The highest gene detected in dairy farm B was fliCH7 16 (43.2%) and papC 1 (2.7%) was the least. The amplification of pathogenic genes associated with diarrheagenic E. coli from cows’ raw milk demonstrates that potentially virulent E. coli strains are widely distributed in raw milk and may be a cause of concern for human health.

cells is activated by ST causing fluid secretion. E. coli strains that are capable of producing both LT and ST toxins (ETEC) can cause severe diarrhea [11].
Enterohaemorrhagic E. coli (EHEC) is associated with foodborne outbreaks in the developed world and can cause bloody diarrhoea, haemorrhagic colitis (HC) and the Haemolytic Uraemic Syndrome (HUS) due to the elaboration of Shiga toxin (Stx). In EHEC strains, stx-genes are typically found in transmissible lambdoid bacteriophages [12]. Other virulence factors such as membrane protein intimin, encoded by the eae gene, fliCH7 encoding flagella antigen H7 and enterohaemolysin, encoded by the HlyA gene [13] have been identified.
Virulence factors such as toxins (cytotoxic necrotizing factors and hemolysin), adhesins (P fimbriae, afrimbriae, type 1 fimbriae, S and F1C fimbriae), polysaccharide coatings (group II capsules) and siderophores (the aerobactin system) [14,15], associated UPEC are thought to be responsible for the pathogenic potential of the organism. Specific adhesins that mediates specific adhesion can be differentiated based on the receptor binding specificity. The most significant mannose-resistant adhesins are the P fimbriae encoded by E. coli pap (pyelonephritis associated pilus) operon. However, they are expressed by few E. coli serotypes [16]. Uropathogenic E. coli strains have been identified by detecting urovirulence factors including those mentioned earlier using conventional PCR [15].
Discriminating various E. coli pathotypes from cattle and their products have been greatly studied in the last two decades with biomolecular techniques favoured over serology in detection of E. coli pathotypes because they are more sensitive and specific [17][18][19][20][21]. It therefore became the objective of this study to find the prevalence of E. coli virulence genes from two dairy farms in the Eastern Cape, South Africa.

Study Sites
Eastern Cape Province is one of the poorest and second largest provinces in South Africa and mainly comprised of rural settlements, it is also the "livestock" province of the country carrying 46% of South Africa's goats, 28% sheep, and 21% cattle. Dairy farm A has been in operation since 2007, it has 800 dairy cows and produces about 10,000 L of milk per day which is bought by one of the biggest milk companies in South Africa. Dairy farm A has a rotary milking parlour whereas dairy farm B has an in-line milking system. Both dairy farms are located in the Eastern Cape Province. Dairy farm B has been in operation since 2008, it has 600 dairy cows and produces about 8000 L of milk per day.

Isolation and Identification
We collected 400 raw milk samples from individual lactating cows showing clinical and subclinical signs of mastitis from the chosen two dairy farms (200 (100 from clinical and 100 from subclinical mastitis) from farm A and 200 (100 from clinical and 100 from subclinical mastitis) from farm B). Clinical mastitis occurs when visible signs of inflammation are observed in the udder of the cow or the teats; while Subclinical mastitis is more subtle in that no signs of infection are visible this requires regular monitoring and laboratory testing. Subclinical mastitis was determined using California Mastitis Test. Samples were collected according to the in-house procedure on each farm. Briefly, each milking station has a small collection bottle with a little hole for sample collection that is open and closed using a tap. The bottles are washed after each individual cow's milking session. The samples were collected from that tap. The samples were cultured on selective medium Violet red bile-mug agar (Merck, Johannesburg, South Africa) and incubated at 37 °C for 24 h. The plates were checked under UV light at 360-370 nm after 24 h of incubation. Light blue fluorescence indicates the presence of E. coli [22,23]. Biochemical tests including Gram staining, catalase and oxidase test were performed. Those organisms that were Gram negative, catalase positive and oxidase negative were identified using API 20E (BioMerieux, Marcy I'Etiole, France) following manufacturer's instructions.

DNA Extraction
DNA was extracted from identified E. coli and from a positive control strains for E. coli (ATCC 8739, South African Bureau of Standards (SABS), No. ESC 20) purchased from South African Bureau of Standards, Pretoria, South Africa. DNA extraction was done following the method of Maugeri et al. [24] and Lopez Saucedo et al. [25]. Briefly, single colonies of presumptive E. coli grown overnight at 37 °C on EMBA plates were picked, suspended in 200 μL of sterile distilled water, vortexed for 2 min using MS2 Minishaker (Digisystem Laboratory Instruments Inc., New Taipei City, Taiwan). The cells were then centrifuged for 10 min at 13,000 rpm (ThermoFisher Scientific, Schwerte, Germany). Five hundred microliters (500 µL) distilled water was pipetted into the eppendorf tubes, vortexed and the cells were lysed using a heat on Dri-Block DB.2A (Techne, Johannesburg, South Africa) for 15 min at 100 °C. The cell debris were removed by centrifugation at 10,000 rpm for 5 min using a MiniSpin microcentrifuge (ThermoFisher Scientific). The supernatant was used as a template in the PCR assay immediately after extraction.

Detection of Virulence Genes in E. coli Isolates
Oligonucleotide primers targeting the uidA gene were used to confirm the isolates in the polymerase chain reaction with the conditions as shown in Table 1. PCR assay was carried out in a 25 μL reaction volume containing a 24 µL of 2×master mix (containing 0.05 U/µL of Taq DNA polymerase, reaction buffer, 4 mM MgCl 2 , and 0.4 mM of each dNTP) and 1µL of template DNA. Oligonucleotide primers targeting the fliCH7 gene encoding for structural flagella antigen H7 in Enterohemorrhagic E. coli, eagR gene encoding for antiaggregative protein of Enteroaggregative E. coli, ial gene encoding for invasion-associated locus of Enteroinvasive E. coli, eae gene encoding for Enteropathogenic E. coli intimin, lt gene encoding for heat-labile toxins found in Enterotoxigenic E. coli and papC gene characterizing uropathogenic E. coli were used in the polymerase chain reaction following the conditions in Table 1

Gel Electrophoresis
The PCR products (10 μL aliquots) were resolved in 1.8% agarose gel containing 5 μL ethidium bromide in 1× TAE buffer (40 mM Tris-HCl, 20 mM Na-acetate, 1 mM EDTA, pH 8) [27,31] before being visualized and photographed under the Alliance 4.7 XD-79 System (Uvitec, Cambridge, UK). A 100-bp DNA ladder was included on each gel as a molecular size standard. The electrophoresis was carried out at 100 V for 1 h.

Biochemical Confirmation
There were 100 (50%) Gram negative rods, catalase positive and oxidase negative in dairy farm A samples and 88 (44%) in dairy farm B. In dairy farm A, 72 (72%) were identified as E. coli using API 20E while dairy farm B, showed 68 (77%) E. coli (Table 2). Species identity by API ranged from 76.8% to 99.9% in dairy farm B and 76.8% to 100% in dairy farm A.

Molecular Characterization of E. coli
Isolates that were Gram negative, oxidase negative, and catalase positive were screened for E. coli uidA gene using PCR assays. Dairy farm B showed 50 (57.5%) uidA gene amplification while dairy farm A showed 37 (42.5%) amplification (Table 2; Figure 1). Amplification of several genes representing E. coli pathotypes is shown in Figure 1. The E. coli isolates that were positive by PCR for fliCH7, eae, lt, papC, ial, and eagR genes are summarized in Table 3. Dairy farm A had the highest amount of virulence genes compared to dairy farm B except for lt gene, usually found in ETEC which was higher (13.5%) in farm B. The target gene of E. coli O157:H7 (fliCH7) was noticed in the isolates obtained from both dairy farms in high amounts in comparison to other virulence genes. Farm A is situated next to a wastewater treatment plant and the discharged final effluent is used to clean the clusters, milking machines, milking utensils and the floor of the milking parlour this could explain the high occurrence of pathogenic E. coli in this farm. Also the milking parlour is cleaned once a day in this farm, whereas in farm B it is cleaned twice a day after every milking activity. Cattle faeces were observed around the farm floor, milking and feeding equipments. Poor hygiene of feeding equipment and contaminated feed plays an important role in milk contamination. Table 3. Occurrence of pathogenic E. coli isolates from two dairy farms as indicated by presence of the target gene marker. Mastitis in lactating dairy cows is commonly caused by E. coli strains. The severity and course of mastitis differs enormously depending on the individual cow's response. However, the strain virulence may also play a role. The ability of E. coli to pick up DNA from the environment and therefore virulence genes are well demonstrated [32]. E. coli has been regarded as an indicator of fecal contamination [33], however, in the milk industry; it is regarded as poor hygiene indicator and insufficient sanitary practices during milking [34]. Raw unpasteurized milk can be an important source of food-borne pathogens because it harbors a variety of microorganisms while bulk tank raw milk may be contaminated by udder excretion of infected animals [35]. Unpasteurized milk is used to produce several types of cheese [35,36]. Consumption of unpasteurized and pasteurized milk has been blamed for disease in humans [37].

Virulence Genes from E. coli Isolates in Raw Cow's Milk
Different virulence genes were isolated from E. coli in diarrheic stools and water [38]. We based our selection of virulence genes for this study on their association with E. coli strains causing extra-intestinal and intestinal infections. There are no biochemical difference between bovine mastitis E. coli strains and faecal strains. Approximately 30.3% of the isolates in our study did not show amplification of examined genes. This might be indication that E. coli strains associated with mammary gland infections may use different mechanisms to cause diseases [39]. In another study, E. coli isolates (10.7%) from bovine mastitis were found to have a single virulence gene [40]. Several virulence factors have been identified in bovine mastitis associated E. coli isolates [41] but many bovine mastitis isolates lack any of the virulence factors [42].
In our study all of the isolates were stx1 and stx 2 negative. This result is similar to the results of [43,44] where stx1 and 2 were not detected from bovine mastitis. This might not be a true reflection of the results since sub-types of stx 1 and stx 2 were not tested. The fliCH7 gene is one of the genes used as a marker to identify E. coli O157:H7 [45] has been amplified from 27 (54%) samples from farm A and 16 (43.2%) from dairy farm B. This gene has also been isolated in E. coli from water by Okoh and Osode [46]. Although fliCH7 gene's associated with mastitis has not been isolated from cows with mastitis in South Africa, its association with severe clinical symptoms, including hemolytic colitis, hemolytic-uremic syndrome (HUS) and diarrhea have been well established [47,48]. Yogurt made on a farm from pasteurized milk was implicated in people affected by E. coli O157:H7, causing HUS cases among children [49]. Therefore, its presence in raw milk is a cause for concern.
The lt gene which encodes for heat-labile toxin, found in Enterotoxigenic E. coli [50] was amplified from 2 (4 %) isolates in dairy farm B and 5 (13.5%) dairy farm A. These results are different from what Frank et al. [51] reported (3.2%) of ETEC strains in milk and milk products. As far as we know heat-labile toxin has never been shown to be associated with mastitis. However, the organism that produces heat-labile toxin, Enterotoxigenic E. coli (ETEC), causes infantile and travelers' in humans from developed to underdeveloped countries [52]. Contaminated food including milk and water serves as a route of infection for ETEC [53]. The disease manifests as a severe cholera-like syndrome [53,54]. This gene has also been detected E. coli from water [38].
The ial gene (a marker for the virulence plasmid pInv) which is found in Enteroinvasive E. coli was also amplified from 4 (8%) isolates in dairy farm A and 2 (5.4%) isolates from dairy farm B. Contaminated food and water have been implicated in the transmission of EIEC [55,56] even though person-to-person cases of EIEC have been noted [57]. When this organism is ingested, invasion and adhesion of the epithelial cells in the intestine mediated by the ial gene occurs. The characteristics of the illness are the appearance of blood and mucus in a condition called colitis of infected individuals [58]. Diarrhea, abdominal cramps, fever, a generalized malaise, vomiting, chills, are some of the symptoms. Outbreaks have been associated with hamburger meat and unpasteurized milk [59]. This is the first report where EIEC has been isolated from raw milk in the Eastern Cape Province of South Africa. Enteroaggregative E. coli was isolated and eagR amplified from 5 (10%) isolates in dairy farm A and 4 (10.8%) isolates from dairy farm B. Human clinical, animal, and environmental samples are some of the samples where this organism has been increasingly isolated and characterized around the world [60,61]. However, frequencies of Enteroaggregative Escherichia coli among cattle in the Eastern Cape Province, South Africa, are not known.
The virulence factor, eae, is associated with EPEC strains, defined as Stx-negative E. coli strains able to cause "attaching and effacing" (A/E) lesions, and also characterized by bacterial adherence to the intestinal epithelium intimately [62]. The eaeA and bfp genes have been used for identification of EPEC [63]. In our study, the ratio of eae was 9 (24.3%) in dairy farm B and 9 (18%) isolates in dairy farm A. However, in our study, we did not find any evidence of bfp in all the isolates tested.
Our results inform us that the isolated organisms were atypical EPEC as they lacked the gene bfp. The presence of this pathogen in milk proved to be variable in different regions. In Brazil, Aleixo and Aver [64] reported EPEC in 25% milk samples while 26 (27.7%) from bulk milk were EPEC in Trinidad [65]. These values were higher than the 3.09% reported by Holko et al. [66] from Slovakia and that of Altalhi and Hussen [41], Saudi Arabia. EPEC strains are the oldest recognized category of diarrheagenic E. coli [67] and are well known prominent cause of diarrhea, particularly in children in less developed countries [68].
Uropathogenic E. coli (UPEC) which contain papC characterized as one of the pathogenic gene associated with the adhesion of UPEC to the upper urinary tract in humans [2] was amplified in three (6%) isolates from dairy farm A and one (2.7%) isolate in dairy farm B. The papC gene was reported to be associated with persistent mastitis by Dogan et al. [69]. Ingestion of milk contaminated with UPEC it can lead to cystitis in the bladder and acute pyelonephritis in the kidneys [3,70].
Act, No. 54 of 1972 in South Africa governs the safety of milk, and sets standards to which milk and dairy products, foodstuffs, cosmetics and disinfection must conform to. According to this Act milk may not contain pathogenic organisms, extraneous matter, any inflammatory product or any substance which may render it unfit for consumption. Bacteriologically it may not contain more than 20 coliforms or any E. coli per mL. This study indicates that raw milk samples from both dairy farm A and dairy farm B were not complying with the act and could be a course for public health concern as some pathogenic genes known to be associated with diarheagenic E. coli were detected.

Conclusions
The results obtained in this study shows that raw cow's milk available to consumers in the Eastern Cape, South Africa was contaminated with the pathogen E. coli. Regular washing of milker's hands and animal udders together with sterilization of utensils and dairy equipment, is highly recommended as strict preventive measures. This might help in suppression of mastitis disease from the herd. Teaching and training programs for those working at the dairies can possibly improve the situation.