Antibiotic resistance and virulence of Escherichia coli strains isolated from animal rendering plant

Processing of animal carcasses and other animal wastes in rendering plants is a significant source of antibiotic resistant microorganisms. The main goal of this study was to investigate the resistance to 18 antibacterial agents including β-lactams, fluoroquinolones, colistin and virulence factors (iss, tsh, cvaC, iutA, papC, kps and ibeA genes) in 88 Escherichia coli strains isolated from a rendering plant over 1 year period. ESBL (Extended-spectrum beta-lactamases) and plasmid-mediated Amp were screened by interpretative reading of MIC. ESBL phenotype was detected in 20.4% of samples and high level of resistance to fluoroquinolone was found in 27.2% of strains. Cephalosporinase CTX-M1, cephamycinase CMY-2, integrase 1 and transposon 3 genes were detected by PCR. Furthermore, there were found three CMY-2 producing E. coli with O25b-ST131, resistant to the high level of enrofloxacin and containing the gene encoding the ferric aerobactin receptor (iutA). One enrofloxacin resistant E. coli strain possessed iss, ibeA, kps and papC virulence genes also with CMY-2, integrase1 and Tn3. ST131 E. coli with CMY-2 has a zoonotic potential and presents a serious health risk to humans.

However, there are still some gaps in our knowledge about the role of rendering plants in the spread of antibiotic resistance.
The aim of the study was to determine the resistance to 18 antibacterial agents including β-lactams, fluoroquinolones, colistin and virulence factors (iss, tsh, cvaC, iutA, papC, kps and ibeA genes) in E. coli isolated from surface swabs collected in the processing space and from waste water produced by the investigated rendering plant.

Material and methods
A rendering plant is a processing operation where materials of animal origin are recycled. The following processes are involved: unloading of raw material brought for processing, its sorting, primary processing and sampling, sterilisation, separation of fat and feed meals of animal origin, pressing, processing of feed meals, and processing of animal fat. The operation premises are divided to a section used for common processing of materials of categories I and II and (high risk-meat-bone meal have to be burned at a temperature of 850 °C) a separate section for processing of category III materials (lower risk-meat-bone meal can be used for production of pet granules).
Destruction (crushing of material to 50 mm particles) and sterilisation of animal by-products using the temperature of 133 °C and pressure of 3 bars during 20 min ensure high level of sanitization and limit the risk of spread of microorganisms to the environment. The above mentioned parameters are critical for adequate processing of raw materials entering the rendering plant 9 . In the dryer the solid portion is separated from the liquid one (water) and the dried meat-bone meal is pressed during which process the fat is separated from the meat-bone mash. The technological procedure in the rendering plant includes processing of the wastewater and installation of a biological air washer-the components important for reducing the hygiene-epidemiological risks.
Sampling. Over 1 year period (10 times) swabbing procedure was used to obtain samples from surfaces in the processing section of the investigated rendering plant and additional samples were collected from raw wastewater in rendering plant. Altogether 88 samples were obtained and examined. Each sample from surface swabs and waste water was inoculated and multiplied in Buffered peptone water (Oxoid, Basingstoke, United Kingdom) and then were sub-cultured on Mac Conkey agar (Oxoid) at 37 °C overnight 10 .
Identification of E. coli. The suspect E. coli colonies from McConkey agar were identified by a matrixassisted laser desorption/ionization (MALDI-TOF) biotyper (Bruker Daltonics, Leipzig, Germany). Bacterial extracts for mass spectrometry measurements were prepared as recommended by the manufacturer of the MS instrument. For MALDI-TOF analysis, one colony was spotted onto a ground steel target (Bruker Daltonik GmbH, Leipzig, Germany) and air dried for 15 min.
Each sample spot was overlaid with 2 μl of matrix solution (saturated solution of α-cyano-4-hydroxy-cinnamic acid in 50% acetonitrile with 2.5% trifluoroacetic acid), and again air dried for 15 min. To identify the relevant microorganisms, the raw spectra obtained for each isolate were imported into a BioTyper software, version 2.0 (Bruker Daltonik GmbH, Leipzig, Germany), and analysed without any user intervention 10 .
Antibiotic susceptibility. Eighty eight isolates of E. coli (one sample-one strain) were analysed for their antibiotic susceptibility and for the presence of ESBLs, pAmpC and for the high level fluoroquinolone resistance.
Phenotypic confirmation of mechanisms of ESBLs and pAmpC to the β-lactams (CTR, CAZ, CAC) was carried out by reading the MIC levels 12,14 .
Phenotypic interpretation of chromosomal quinolone-resistance mechanisms was based on modification of the method by Kmet and Kmetova 15 . High-level resistance MIC for CIP (≥ 4 mg/L) and ENR (≥ 16 mg/L) involved three mutations in QRDR (gyrA and parC).

Virulence factors.
Screening of E. coli isolates for ExPEC (Extra intestinal Pathogenic E. coli) virulence genes was carried out by PCR amplification of the following: iutA-ferric aerobactin receptor; cvaC-colicin V and kpsII-capsular polysialic acid virulence factor 24 ; iss-increased serum survival 25 ; tsh-temperature sensitive haemaglutinin 26 ; papC-P fimbrial adhesin 27 ; and ibeA-invasive factor of E. coli strains responsible for neonatal meningitis in humans 28 . Ethics approval and consent to participate. This study does not qualify for review by the University of veterinary medicine and pharmacy Ethics Board.

Results
Antimicrobial susceptibility profiles. A modified microdilution method with the VetMIC panel was used to detected antimicrobial resistance in 88 E. coli strains (Fig. 1).

Discussion
Critical control points in rendering plants with regard to high level contamination, bioaerosol production and the risk for the environment involve unloading of the raw material and wastewater treatment. In the unloading section of the rendering plant, the raw material is dumped from collecting containers of the transport vehicles to destructors. This is associated with potential aerosolisation of liquids, such as blood, intestinal contents and similar.
Very similar drugs (beta lactams, penicillin, ampicillin, cloxacillin, tetracyclines, sulphonamides and potentiated sulphonamides, cephalosporins, and fluoroquinolones) have been used in both human medicine and agriculture production 29 .
Antimicrobials used in poultry production have the potential to accumulate in poultry feathers and during the rendering process are not completely destroyed. Poultry feathers can be recycled to a feather meal and used as a fertilizer and animal feed, thereby providing a potential pathway for re-entry of drugs into the human food supply 30 .
Hofacre et al. 31 found that a high percentage of feed samples for poultry containing meat and bone meal from rendering plant were contaminated by bacteria resistant to amoxicillin, ampicillin, cephalothin or clavulanic acid. Some samples contained bacteria resistant to kanamycin, trimethoprim/sulfamethoxazole or ciprofloxacin. The presence of mobile genetic elements mediate multi-drug resistance was proved in many of the isolated bacteria.
Higher than 30% prevalence among the 3rd-generation cephalosporin-resistant E. coli was detected mostly in poultry production 32 .
ESBLs and AmpC beta-lactamases are usually responsible for the mediation of resistance to 3rd-generation cephalosporins in E. coli 33 . Similar ESBL phenotypes with high level fluoroquinolones resistance in animal E. coli isolated from a Slovak poultry slaughterhouse was described by Gregova et al. 10 . Aggregated European Community data for E. coli isolates from broilers showed that over 50% of isolates were resistant to ciprofloxacin 34 . Moreover, the fluoroquinolones-resistant E. coli typically exhibited clinically significant elevations in MIC values 35 . E. coli isolates resistant to fluoroquinolones are often resistant to other antibiotic groups and genes of virulence 36 .
The CMY-2-producing E. coli O25b-ST131 represent a clonal lineage that differs from the CTX-M-15-producing ST131-O25b cluster. ST131-O25b strains with the presence of ESBL-type CTX-M-15 and resistance to fluoroquinolones have been reported worldwide. They are frequently a cause of infections, particularly of the urinary tract of humans. In human patients in Europe, approximately 1% of the 3rd generation cephalosporinresistant E. coli produce CMY-2. However, recent studies in Asia showed higher rates and an increasing trend among the 3rd-generation cephalosporin-resistant E. coli isolates has been reported 33 .
Recently, ten multi-resistant strains of E. coli that harboured CMY-2 were observed with increasing tendency in the European livestock production. However, ST131 isolates with CMY-2 production have been reported rarely 37,38 .
CMY-2-producing E. coli isolates were also detected in products from meat, livestock animals and human patients. The predominant way of transmission of bla CMY-2 genes between animals and humans is the horizontal transfer of temporarily stable bla CMY-2 -carrying IncK2 and IncI1 plasmids 33 . This suggests a zoonotic potential of the bla CMY-2 genes and their transmission by horizontal transfer and clonal spread along the food production chain 33,37 .
Extraintestinal virulence genes encoding adhesins, iron capture systems, toxins, and protectins have been correlated with successful colonization of gut in humans and animals 39,40 .
Our study of E. coli strains from wastewater showed that virulence genes cvaC, iutA, iss, papC were the most frequently detected in them. In some E. coli samples, we detected genes kps, tsh, papC, ibeA.  41 showed presence of antimicrobial-resistant E. coli strains with virulence factors (most frequently iutA, iss, cvaC, tsh and papC) related to avian pathogenic or human uropathogenic E. coli.
The study conducted in Canada 42 revealed high prevalence of many virulence genes (ompT, traT, uidA, vat, hemF, iss and cvaC), including the genes responsible for adhesion, fimH and kpsMT KII, in ExPEC isolates from frozen poultry meat.
Bok et al. 44 observed that virulence genes (fimH, papAH, iutA, iroN, ompT, traT, and iss) were more frequently identified in isolates from piglets than from sows. E. coli from piglets constituted a substantial reservoir of extraintestinal virulence genes and could increase the potential risk of extraintestinal infections. According this study 44 the mobile genetic elements transmitted via horizontal gene transfer play an important role in the evolution of E. coli resistance. Most ExPEC virulence genes are clustered together on mobile genetic elements, usually on pathogenicity islands (PAI) or virulence plasmids, exhibiting a unique organization.
Ten out of 13 tetracycline resistant strains carried the Int1 gene, and 6 of them the iutA gene. Five from among 8 streptomycin-resistant strains carried iutA and Int1 genes, which indicate a horizontal transfer of resistant genes between bacteria. The high number of isolates resistant to streptomycin, tetracycline and cotrimoxazol can be spread by same mobile genetic elements.

Conclusion
In conclusion, the present investigations illustrated the current state of antibiotic resistance of E. coli strains in the investigated rendering plant. We detected the presence of E. coli with CTX-M, cephamycinase CMY-2 genes and high level of chromosomal resistance to fluoroquinolones. Furthermore, we found three CMY-2 producing E. coli O25b-ST131, resistant to a high level of enrofloxacin with cvaC and iutA virulence factors. The CMY-2 producing E. coli isolates have a zoonotic potential and pose a serious health risk.
Considering that rendering plant is an important source of resistant bacteria, our data highlight the importance of adequate protection of the working personnel and observation of strict hygiene measures at operation premises.