Prevalence of Antibiotic-Resistant Escherichia coli Isolated from Swine Faeces and Lagoons in Bulgaria

Antimicrobial resistance (AMR) is a worldwide health problem affecting humans, animals, and the environment within the framework of the “One Health” concept. The aim of our study was to evaluate the prevalence of pathogenic strains of the species Escherichia coli (E. coli), their AMR profile, and biofilm-forming potential. The isolated strains from three swine faeces and free lagoons (ISO 16654:2001/Amd 1:2017) were confirmed using Phoenix M50 and 16S rDNA PCR. The antibiotic sensitivity to 34 clinically applied antibiotics was determined by Phoenix M50 and the disc diffusion method, according to the protocols of the CLSI and EUCAST. We confirmed the presence of 16 E. coli isolates, of which 87.5% were multi-drug-resistant and 31.25% performed strong biofilms. The possibility for the carrying and transmission of antibiotic-resistance genes to quinolones (qnr), aminoglycosides (aac(3)), β-lactamase-producing plasmid genes ampC, and blaSHV/blaTEM was investigated. We confirmed the carrying of blaSHV/blaTEM in one and ampC in seven isolates. The strains were negative for the virulence genes (ETEC (LT, STa, and F4), EPEC (eae), and STEC/VTEC (stx and stx2all)). The results should contribute to the development of effective measures for limitation and control on the use of antibiotics, which is a key point in the WHO action plan.


Introduction
Antimicrobial resistance (AMR) is a global public health challenge mainly caused by the widespread use of antibiotics in human and veterinary medicine for over 60 years. It is reported that deaths caused by AMR could increase from 700,000 in 2014 to 10 million by 2050 [1].
The use of antibiotics in animal husbandry has triggered the emergence and dissemination of antibiotic-resistant bacteria and genes for antibiotic resistance (GAR) from livestock farms to the surrounding environment. Different fractions of animal waste, such as swine manure or wastewater, are routinely applied to the fertilization of agricultural land in many countries. However, this waste has become a reservoir of resistant bacteria and various antibiotic residues that remain active in the soil from 20-30 to 40-60 cm depth for long periods of time via long-term manure application and transfer into groundwater by lagoons [2,3]. Their presence exerts a selective pressure on microorganisms and changes the microbial communities through the elimination of sensitive strains and increases the chances of survival for those containing GAR [4,5]. Resistant bacteria and their genetic determinants, such as plasmids, transposons, integrons, and genetic islands, can be spread and exchanged in different ways [6]. When bacteria come into contact with others, they We isolated a total of 28 single colonies (15 with 3 lagoons CHROMagar CCA). This media detected and enumerated β-glucu (metallic blue to violet) and other coliforms (pink to red), accord them, 17 colonies were suspected for E. coli, and 11 colonies were In the present study, we used only colonies, which were su subsequent experiments (Figure 1).

Biochemical and 16S rDNA Characterization
All 17 suspected for E. coli colonies were positive for indole a biochemical identification by BD Phoenix M50. In additio characterization by PCR ( Figure 2) was carried out. We detected rDNAs from faeces and lagoons ( Figure 2A) and from single colo

Biochemical and 16S rDNA Characterization
All 17 suspected for E. coli colonies were positive for indole and were confirmed after biochemical identification by BD Phoenix M50. In addition, 16S rDNA E. coli characterization by PCR ( Figure 2) was carried out. We detected E. coli in both total 16S rDNAs from faeces and lagoons ( Figure 2A) and from single colonies ( Figure 2B,C).

Isolation of Single Bacterial Cultures
We isolated a total of 28 single colonies (15 with 3 lagoons and 13 with 3 faeces on CHROMagar CCA). This media detected and enumerated β-glucuronidase-positive E. coli (metallic blue to violet) and other coliforms (pink to red), according to ISO 9308-1. From them, 17 colonies were suspected for E. coli, and 11 colonies were suspected for coliforms. In the present study, we used only colonies, which were suspected for E. coli for subsequent experiments (Figure 1).

Biochemical and 16S rDNA Characterization
All 17 suspected for E. coli colonies were positive for indole and were confirmed after biochemical identification by BD Phoenix M50. In addition, 16S rDNA E. coli characterization by PCR ( Figure 2) was carried out. We detected E. coli in both total 16S rDNAs from faeces and lagoons ( Figure 2A) and from single colonies ( Figure 2B,C).  Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 (Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 (Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 (Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1).  Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1). Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Antibiotic Resistance
According to the results obtained from BD Phoenix M50 ( Table 2), 11 of the total 16 isolates were resistant to ampicillin (68.75%), trimethoprim (31.25%), and even the combination between trimethoprim and sulfamethoxazole (18.75%). Only one of them (L1.3; 6.25%) was also resistant to gentamicin, amoxicillin/clavulanic acid, ciprofloxacin, and colistin. Interestingly, this strain formed a strong biofilm (Table 1).  Interestingly, isolate F2.1 was negative for E. coli by 16S rDNA detection. Therefore, our studies continued the study on other bacterial strains with proven genetic information.

Detection of Antibiotic Resistance Genes
From a total of 12 tested strains resistant to β-lactam antibiotic, only L1.3 carried blaTEM and blaSHV β-lactam-resistance genes. The presence of ampC β-lactamases gene in three isolates from pigs for fattening and in four isolates from lagoons was found ( Figure 3).   In addition, we performed an antibiotic disc diffusion test (Table 3). We found the resistances to amoxicillin (75%), tetracycline and chloramphenicol (56.25%), trimethoprim/sulfamethoxazole (43.75%), doxycycline hydrochloride (37.5%), and nalidixic acid (25%) in all the 16 isolates. The results for the resistance to ampicillin (68.75%) from BD Phoenix M50 were confirmed. Strain L1.3 was resistant also to pefloxacin. The resistances of two lagoon isolates (L1.4 and L3.4) to streptomycin were found. Isolate L1.4 formed a moderate biofilm, while isolate L3. did not (Table 1).

Detection of Antibiotic Resistance Genes
From a total of 12 tested strains resistant to β-lactam antibiotic, only L1.3 carried blaTEM and blaSHV β-lactam-resistance genes. The presence of ampC β-lactamases gene in three isolates from pigs for fattening and in four isolates from lagoons was found ( Figure 3).  In addition, we performed an antibiotic disc diffusion test (Table 3). We found the resistances to amoxicillin (75%), tetracycline and chloramphenicol (56.25%), trimethoprim/sulfamethoxazole (43.75%), doxycycline hydrochloride (37.5%), and nalidixic acid (25%) in all the 16 isolates. The results for the resistance to ampicillin (68.75%) from BD Phoenix M50 were confirmed. Strain L1.3 was resistant also to pefloxacin. The resistances of two lagoon isolates (L1.4 and L3.4) to streptomycin were found. Isolate L1.4 formed a moderate biofilm, while isolate L3. did not (Table 1).

Detection of Antibiotic Resistance Genes
From a total of 12 tested strains resistant to β-lactam antibiotic, only L1.3 carried blaTEM and blaSHV β-lactam-resistance genes. The presence of ampC β-lactamases gene in three isolates from pigs for fattening and in four isolates from lagoons was found ( Figure 3). In addition, we performed an antibiotic disc diffusion test (Table 3). We found the resistances to amoxicillin (75%), tetracycline and chloramphenicol (56.25%), trimethoprim/sulfamethoxazole (43.75%), doxycycline hydrochloride (37.5%), and nalidixic acid (25%) in all the 16 isolates. The results for the resistance to ampicillin (68.75%) from BD Phoenix M50 were confirmed. Strain L1.3 was resistant also to pefloxacin. The resistances of two lagoon isolates (L1.4 and L3.4) to streptomycin were found. Isolate L1.4 formed a moderate biofilm, while isolate L3. did not (Table 1).

Detection of Antibiotic Resistance Genes
From a total of 12 tested strains resistant to β-lactam antibiotic, only L1.3 carried blaTEM and blaSHV β-lactam-resistance genes. The presence of ampC β-lactamases gene in three isolates from pigs for fattening and in four isolates from lagoons was found (Figure 3). In addition, we performed an antibiotic disc diffusion test resistances to amoxicillin (75%), tetracycline and chlo trimethoprim/sulfamethoxazole (43.75%), doxycycline hydro nalidixic acid (25%) in all the 16 isolates. The results for the (68.75%) from BD Phoenix M50 were confirmed. Strain L1. pefloxacin. The resistances of two lagoon isolates (L1.4 and L3 found. Isolate L1.4 formed a moderate biofilm, while isolate L3.

Detection of Antibiotic Resistance Genes
From a total of 12 tested strains resistant to β-lactam anti blaTEM and blaSHV β-lactam-resistance genes. The presence of in three isolates from pigs for fattening and in four isolates f (Figure 3).

Discussion
The resistance in E. coli to some of the most widely used antibiotics in medical practice for the treatment of urinary tract infections, such as fluoroquinolones and sulfonamides, is a global problem. Due to the rapid spread of GAR, the treatment of urinary tract infections is ineffective in more than 50% of patients. According to a World Health Organization (WHO) report in 2017, most European Antimicrobial Resistance Surveillance Network (EARS-Net) countries have resistance between 10% and 25%, while more than 25% are found in Bulgaria, Cyprus, Italy, and Slovakia. Among Central Asian and European Surveillance of Antimicrobial Resistance (CAESAR) countries, the reported resistance exceeds 50% (Montenegro, Russia, Northern Macedonia, and Turkey), while in Serbia it varies between 25% and 50% [21]. The EARS-Net data show statistically significant increase in resistance of E. coli isolates in the EU to third generation cephalosporins [22]. The emergence of carbapenem-resistant E. coli has recently been identified, which is a serious challenge. Although the proportions of resistance are still low in Europe (around 1% and more), there is a tendency to increase worldwide [21,22]. A major mechanism of cephalosporin resistance is the production of β-lactamases, which hydrolyzes the β-lactam ring and inactivates β-lactam chemotherapeutics. The genetic determinants of resistance demonstrated in E. coli isolates include extended-spectrum β-lactamases (ESBL) encoded by various plasmid genes (blaSHV, blaCMY-2, blaTEM, etc.), as well as a number of GARs for quinolone resistance (qnr), trimethoprim (dhf ), aminoglycosides (aac (3)), etc. [23].
Studies in Bulgaria showed that resident strains of E. coli have a clear phenotypic and genotypic resistance profile to chemotherapeutics used in animal husbandry and human medicine. A high percent of resistance among pathogenic E. coli strains isolated from pigs in 2010-2015 was observed, which is a very alarming fact. In comparison with a previous study conducted in [2000][2001][2002][2003][2004], the resistances to tetracycline antibiotics, streptomycin, spectinomycin, ampicillin, and sulfonamides, have increased by twofold. The percentage of resistance strains to ciprofloxacin has increased by tenfold during the same period. The widespread distribution of resistant E. coli strains has also been demonstrated in isolates, representative of the resident intestinal microflora of healthy pigs. The results showed the prevalence of aadA1 genes for determining streptomycin/spectinomycin resistance, tet(A) for determining tetracycline resistance, and strA/strB for determining streptomycin resistance. There is a limit information about the presence of the genes sul1 and sul2, blaTEM, and tet(B), which determine the bacterial resistances to sulfamethoxazole, aminopenicillins and cephalosporins of first generation, tetracyclines, respectively and the intl gene, which is responsible for the synthesis of the integrase enzyme from class 1 integrons. The lowest distribution is for the aacC2 gene, which is responsible for the resistance to gentamicin, kanamycin, tobramycin and netilmicin [24]. These facts support the hypothesis that the horizontal transfer of GAR in MDR commensal gut bacteria is one of the important risk factors for gene transfer, mainly through foodstuffs of animal origin or from the environment to humans.
Fluoroquinolone-resistant E. coli in China and Korea have been isolated from fecal samples [25,26]. From 171 samples isolated in 2015, from which 52 (30.4%) were from diseased swine and the other 119 (69.6%) were from healthy swine, a total of 59 E. coli isolates (34.5%) were confirmed as fluoroquinolone-resistant (21 (40.4% from diseased swine) and 38 (31.9% from healthy swine)). The researchers found plasmid-mediated quinolone-resistance (PMQR) genes in 9 isolates (15.3%) and efflux pump activity in 56 isolates (94.9%). The authors reported that the resistance to fluoroquinolones has increased significantly in swine compared to in previous studies in Korea, although fluoroquinolones have been banned as a feed additive since 2009. These authors investigated the qnrA and qnrB genes, but as in our study, they did not prove them [25]. All isolates from China showed the moderate rates of the resistance to norfloxacin (43.0%), ciprofloxacin (47.6%), ofloxacin (47.0%), and levofloxacin (38.8%). They also did not detect the presence of qnrA and qnrB genes [26]. Hu et al. (2017) suggested that the predominant PMQR genes detected in human isolates were qnrA and qnrB, whereas qnrS was detected in swine samples.
Probably for this reason, we also failed to confirm any of these two genes (qnrA and qnrB) (Figure 3).
According to a study by the National Diagnostic Research Veterinary Medical Institute (NDRVMI) and University of Forestry in Bulgaria conducted in the period 2012-2014, the number of positive strains of E. coli from all samples isolated from different pig farms in the country ranged between 50% and 70%. Above 75% of them were resistant to amoxicillin and erythromycin and from 51% to 75% of them were resistant to ampicillin, oxytetracycline, thiamulin, streptomycin, doxycycline, tylosin, and tilmicosin. It has been found that sensitivity is most strongly established to non-use agents (amikacin, cefquin, and cefotaxime), less commonly used in practice (kanamycin), or new agents in the fluorinated quinolone groups (ciprofloxacin, enrofloxacin, and pefloxacin) and amphenicols [27].
We performed antimicrobial susceptibility tests against additional eight classes of drugs and six other antibiotics (a total of 34 antimicrobial agents). From all isolated E. coli strains, 87.5% of them are MDR (only F1.4 and L3.1 were sensitive to the antibiotics used). Only 18.75% of all isolated E. coli strains were resistant to aminoglycosides (L1.4 and L3.4 were resistant to streptomycin, and L1.3 was resistant to gentamicin), 81.25% of them were resistant to penicillins, 25% of them were resistant to fluoroquinolones, 6.25% of them (only L3.2) were resistant to macrolides, and 68.75% of them were resistant to other antibiotics. The isolated E. coli strains from swine faeces and lagoons were susceptible to monobactams, cephalosporins, and carbapenems. We found high resistance to β-lactam (amoxicillin and ampicillin) and tetracycline (tetracycline and doxycycline hydrochloride) antibiotics (Tables 2 and 3). The percentage of resistance was also high. Moreover, compared with a previous study in Bulgaria, the current research demonstrates a substantial increase in resistance to trimethoprim/sulfamethoxazole from 7.1% in the period 2012-2014 [27] to 43.75% in our study and a significant increase in resistance to nalidix acid from 11.1% in the period 2012-2014 [27] to 25% that we found. The ampicillin and amoxicillin resistance are found today (70-80%), including those to gentamicin and pefloxacin (about 4-7%) [27]. Decreased resistances to doxycycline (from 64.7% in the period 2012-2014 [27] to 37.5% in our case), streptomycin (from 63.1% in the period 2012-2014 [27] to 12.5% found by us), erythromycin (from 80% in the period 2012-2014 [27] to 6.25%) have been observed. Dimitrova et al. (2016) also documented that the isolated E. coli strains were susceptible to amikacin, cefotaxim, ciprofloxacin and norfloxacin [27]. Urumova (2016) studied the AMR in E. coli in the period 2012-2016 from different regions in Bulgaria (Shumen, Ruse, Razgrad, Yambol, and Varna). She found high resistance in growing pigs, compared with that in suckling pigs, fattening swine, and lagoons [24]. Compared with her study, it was found that the resistances to ampicillin (68.75%) and amoxicillin/clavulanic acid (from 2.2% in the period 2012-2016 to 6.25% in our study) were doubled and tripled, respectively [24]. The researcher also proved high resistance to streptomycin (69.4%) [24], as reported by previous authors [27]. A slight resistance to ciprofloxacin (5.2%) was observed, which was also confirmed by our results (6.25%). The resistance to tetracycline was almost not changed from 73.3% in the period 2012-2016 [24] to 56.25% found by us, indicating that the antibiotic continues to be used in swine farms. Only one isolate out of a total of 157 growing animals was found was to be resistant to ceftazidime and cefotaxime [24].  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I   Tetracycline

Swine Farm and Sample Collection
The study included a swine farm near Kostinbrod, which was founded in 2008 and today is a part of the Hog and Pig Farming Companies in Bulgaria. It was designed for up to 8200 breeding animals and their offspring. It was built on an area of about 70.5 decares with introduced biosecurity measures. All normative documents for protection and animal welfare were observed. Three samples from pig faeces (F1-F3) and three samples from lagoons (L1-L3) were collected in March 2020 according to ISO 5667-3:2018. Probes F1 and F2 were from pigs for fattening, and F3 was from mother pigs.

Biochemical Characterization
All isolated single colonies suspected for E. coli were tested for indole production from trypthophan deamination using Kovacs' Indole Reagent (R008, HiMedia, Mumbai, India). We used automatic BD PhoenixTM M50 system (443624, Becton, Dickinson and Company, Franklin Lakes, NJ, USA) for a full biochemical characterization of isolates by the laboratory procedure, as described by the manufacturer. Briefly, the bacterial colonies (0.

Isolation of 16S rDNA
The total rDNA was extracted from faecal and lagoon samples with the GeneMATRIX Stool DNA Purification Kit (E3575, EURx Ltd, Gdańsk, Poland). The rDNA concentration and purity were determined with NanoDrop 1000 (Thermo Fisher Scientific Inc., Waltham, MA, USA) by migration in 0.8% SeaKem ® LE Agarose gels (50004L, Lonza Group Ltd, Basel, Switzerland) in a 1× TBE buffer. The 16S rDNA from the confirmed E. coli was isolated with the GenEluteTM Bacterial Genomic DNA Kit (NA2120, Merck (Sigma-Aldrich, St. Louis, MO, USA).

PCR Analysis
The extracted total 16S rDNAs from faeces and lagoon samples were subjected to conventional PCR with gene-specific primers for E. coli. The isolated 16S rDNAs from single colonies were subjected to conventional and multiplex PCR with primers linked to virulence and antibiotic resistance genes in the isolated E. coli strains (Table 4). For PCR amplification, we used the Taq PCR Master Mix (2×) protocol (E2520, EURx Ltd, Gdańsk, Poland) being optimized in our laboratory as follows: 1 cycle of initial denaturation running at 95 • C for 5 min; total 25 cycles of denaturation (at 94 • C for 30 s), annealing (depending on the temperature of the primer for 60 s) and extension (at 72 • C for 1 min); 1 cycle of final extension (at 72 • C for 7 min) and cooling (at 4 • C). The PCR products were visualized in

Test for Biofilm Formation
The ability of the isolated single colonies from pig faeces and lagoons to form biofilms was tested in flat-bottomed 96-well plates, according to the protocol of Stepanović et al. with small modification [46]. Briefly, we used a Brain Heart Infusion broth (M210, HiMedia, India) supplemented with 2% D-(+)-glucose (G7021, Merck, Darmstadt, Germany). The bacterial inoculums were cultured at 37 • C for 18 h, and then, the supernatants were aspirated gently. The cells were washed two times with 200 µl PBS and fixed with methanol (32213-M; Sigma-Aldrich, USA) for 15 min. The plates were placed to dry; each well was stained with a 200 µl 2% gentian violet solution for 5 min and was washed under running water. As a negative control, we used blank. The test was performed sixfold. The biofilms were photodocumented on microscopic-configuration Nikon Eclipse-Ci-L (Nikon Instruments Europe BV, Netherlands) and then dissolved with a 33% glacial acetic acid solution. The optical density (OD) was measured at 570 nm by using an ELISA reader ELx800 (BioTek Instruments, Winooski, Vermont, USA). We used the following classification of Christensen et al. (Table 5) to determine the adherence potential [47]: Table 5. Correlation between the optical density of samples and bacterial adherence [47].

Conclusions
From the total of 28 single colonies, 16 isolates (100%) were confirmed with BD Phoenix M50 and 16S rDNA PCR to be E. coli. The antimicrobial tests showed that most of them (87.5%) had MDR. Moreover, 31.25% of the E. coli strains were capable of forming strong biofilms. We found high percents of resistance varying between 50% and 75% to amoxicillin, ampicillin, tetracycline, and chloramphenicol. The resistances (25%-50%) to clinically administered antibiotics, such as trimethoprim, trimethoprim/sulfamethoxazole, doxycycline hydrochloride, and nalidixic acid, were no less. We proved the β-lactamase genes blaTEM/blaSHV in one isolate from lagoon and ampC in three isolates from pigs for fattening and in four isolates from lagoons.
From the results presented here and compared with the data for the period 2012-2016, high resistance to tetracycline was found in growing pigs and fattening swine, which is a worrying fact as coliforms resistant to this antibiotic may be ingested during consumption. This also applies to the antibiotics ampicillin and amoxicillin, which continue to be used in veterinary practice. Probably, less commonly applied are streptomycin, erythromycin, and doxycycline. Considering the clinical importance of antibiotic resistance emergence in veterinary and human medicine, the prescription of antibiotics should be carefully monitored and regulated, in order to reduce AMR in food industry in Bulgaria.