INVESTIGATIONS ON THE RESISTANCE OF COMMENSAL SWINE ESCHERICHIA COLI TO SOME AMINOGLYCOSIDES-AMINOCYCLITOLS

Th e aim of this study was to describe the prevalence of antibiotic resistance to some aminoglycosides, streptomycin, spectinomycin and gentamicin and three aminoglycosideresistance genes in Escherichia coli isolated from feces and lagoon manure in six swine farms in Republic of Bulgaria. A total of 274 E. coli isolates from 270 fecal samples and twelve samples from lagoon manure were tested by disk diff usion method to determine resistance patterns to 11 antimicrobial agents. Aminoglycosides resistance also was determined by E-test, agar dilution method, PCR and qPCR. Th e highest resistance observed to streptomycin (70.0%) and spectinomycin (65.5%). Multi-resistance patterns in studied E. coli strains showed that the resistance to streptomycin/spectinomycin was most frequently seen together with resistance to ampicillin, tetracycline, and sulfonamides (39.6%). Th e E. coli isolates resistant to streptomycin, spectinomycin were examined for the presence of strA/strB, aadA1 genes, and resistant isolates to gentamicin were evaluated for the presence of the aacC1 gene. Th e most common gene determining resistance to aminoglycosides was aadA1 which was found in 54.0% of swine isolates and lagoon manure isolates followed by straA/strB genes (32.3%). Th e aacC1gene was not identifi ed in E. coli isolates resistant to gentamicin.


ISPITIVANJE REZISTENCIJE KOMENSALNE ESCHERICHIA COLI KOD SVINJA NA NEKE AMINOGLIKOZIDE-AMINOCIKLITOLE INTRODUCTION
Th e concept of EMA (2014) on the use of aminoglycosides in livestock and companion animals in the EU, development of resistance and public health risks is grounded on data about the increasing resistance to aminoglycosides in animal and human bacterial isolates.Data of EMA/ESVAC (2013) demonstra-te a widespread use of aminoglycosides in some animal species in particular (large and small ruminants, swine, horses, and pets) for treatment of septicaemic states, gastrointestinal infections, respiratory and urogenital infections.Some authors discussed the potential of resistant E. coli from domestic animals as a reservoir for the spread of resistance to aminoglycoside antibiotics among human population (Chaslus-Dancla et al., 1991;Jonson et al., 1994Jonson et al., , 1995)).Th e primary mechanism of aminoglycoside resistance is the production of aminoglycoside modifying enzymes.Th ree genetic determinants are associated to the expression of resistance to streptomycin in enterobacteria: ant(3 // )-Ia (synonym aadA) coding for the production of adenyltransferase ANT (3 // )-I, modifying streptomycin and spectinomycin, aph(3 // ) -Ib (synonym strA), determining the production of the phosphoryltransferase APH (3 // )-I, modifying streptomycin and aph(6)-Id (synonym strB), responsible for the production of phosphoryltransferase APH (6)-I, that also modifi es streptomycin (Heinzel et al., 1988;Hollingshead and Vapnek, 1985;Scholz et al., 1989).During the last years, reports about a new type of aminoglycoside resistance in animal bacterial isolates related to the prevalence of 16S rRNA methylases and the respective high levels of resistance are increasing (Gonzalez-Zorn et al  Deng et al., 2011).From modifying enzymes coding for resistance to gentamicin in E. coli strains from livestock, adenyltransferase ACC(3) -IV whose production is coded by the аас3-IV gene and that determines a combined resistance to gentamicin and apramycin, is of special interest.Th ere are also data, although limited, on the prevalence of аасС1 and аас3-II genes in domestic animals (Guerra et al., 2003;Sãenz et al., 2004).Th ese genes determine the production of acetyltransferases АСС (3)-I and ААС (3)-II, distinguished with their phenotype profi le, which for the latter gene includes also resistance to tobramycin apart to gentamicin (Vaculenko and Mobashery, 2003).

Sample collection
Between January 2013 and September 2014, 282 faecal swab samples were collected from diff erent age groups of pigs (suckling, weaned, fi nisher) and lagoon manure from 6 farrow-to-fi nish farms.Faecal swabs were transported in Stuart Transport Medium (BD, USA) at low temperature within 18-24 hours.

Culturing and identifi cation of E. coli isolates
Swabs were cultured on McConkey agar (Emapol, Poland) at 37 о С for 24 hours.Lactose-positive colonies were subcultured onto TSI agar (BD, USA) and submitted to preliminary biochemical typing via citrate utilisation, methyl red, Vogues Proskauer and indole production tests.Th e identifi cation of strains was performed with kits for non-fermenting and enteric bacteria (BD, USA) and the semi-automated identifi cation Crystal BBL system.
Th e streptomycin MIC was determined in the agar dilution test and Muller-Hinton agar (Emapol, Poland), by preparation of doubling dilutions of streptomycin (Sigma-Aldrich) within 0.01-256 μg/mL.MIC for gentamicin was defi ned by E test strips (AB Biodisk, Solna Sweden).MICs were interpreted according to epidemiological criteria (EUCAST, www.eucast.org).

Determination of resistance genes to aminoglycoside in commensal E. coli
DNA extraction: For DNA extraction, 24-hour cultures incubated at 37 о С, respectively 3-4 colonies on McConkey agar were suspended in 100 μl sterile distilled water free of inhibitors for molecular diagnostics (Qiagen).Th e DNA extraction kit DNeasy Blood Tissue Kit (Qiagen) was used.
Detection of resistance genes: Th e presence of resistance genes to aminoglycoside antibiotics, aadA1 was detected by qPCR and strA/strB by PCR.Th e primers sequences for strA/strB were strA-F ATGGTGGACCCTAAAACTCT and strB-R CGTCTAGGATCGAGACAAAG (Kozak et al., 2009).PCR assays in 25 μl fi nal volume contained 12.5 μl Taq PCR Master mix (Qiagen) and 3 μl DNA template.Th e PCR reaction for strA/strB consisted of an initial activation step at 94 о С for 10 min, followed by 30 cycles of DNA denaturation at 94 0 С for 30 s, primer annealing at 63 о С for 1 min, and primer extension at 72 0 C for 30 s, and fi nal extension for 10 min at 72 0 C.All reactions were carried out in Eppendorf gradient thermal cycler.Ten μL aliquots of PCR products were analyzed by gel electrophoresis with 1.5% agarose gel (Peqlab, Germany).Gels were stained with ethidium bromide at concentration 10 μg/mL and visualized by UV transillumination.A 100 bp DNA ladder plus (Qiagen) was used as marker.As positive control strain E. coli 94.4 was used, provided by Mrs. J. Mazurec from the Department of Molecular Biology, Faculty of Biological Sciences, University of Zielona Góra, Poland.Negative controls were PCR mixtures with the addition of water in place of template DNA.
To determine the aadA1gene, Microbial DNA qPCR Assay, aadA1 (Qiagen) was used.qPCR amplifi cation was done with Stratagene Mx3000P instrument.Th e thermocycler protocol consisted of: initial PCR activation 1 cycle of 10 min at 95 o C, and 40 cycles of 2-step cycling -denaturation -15 sec 95 o C, annealing and extension 2 min 60 o C. Th e results were interpreted according to manufacturer's instructions (negative control signal at CT>35 and СТ= 22±2 for positive control).

Number of isolates:
Th e total number of E. coli isolates from examined faecal swabs obtained from the diff erent age categories and from lagoon manure at studied farms were 274.

Prevalence of antibiotic resistance by disk-diff usion method
Tables 1, 2 and 3 present results from the phenotype analysis of resistance of E. coli isolates from the 6 surveyed farms to 11 chemotherapeutics.With respect to aminoglycosides, the highest resistance percentages to streptomycin and spectinomycin (93.2% and 91.0 % respectively) were observed in fi nisher pigs.Higher resistance to gentamicin (15.7%) was established in E. coli isolates from weaned pigs than in fi nisher (7.9%) or neonatal pigs (4.5%).Data about the resistance of E. coli isolates to aminoglycosides in groups of suckling pigs showed the highest resistance to streptomycin, spectinomycin and gentamicin (60%, 46.6%, and 20% respectively) at farm 3.
Isolates resistant to streptomycin and spectinomycin were the most prevalent (100%) among weaned pigs from farms 3 and 5, followed by the occurrence of resistance to aminoglycosides in 86.6% of isolates from farms 4 and 6 and in 85.7% of isolates from farms 1 and 2. Th e percentage of isolates from weaned pigs resistant to gentamicin was the highest at farm 4 (7.8%).Th e highest resistance percentage to streptomycin and spectinomycin (100%) was demonstrated in E. coli isolates from fi nisher pigs at farms 2, 3, 4 and 5.With respect to gentamicin, the highest resistance (28.5%) was found out among isolates from farm 3.Among multiresistant isolates, the highest prevalence of 23.6% was that of the phenotype profi le including ampicillin, streptomycin, spectinomycin and tetracycline, following by the profi les of resistance to ampicillin, streptomycin, spectinomycin and sulfamethoxazole (9.7%) and to ampicillin, cephalotin, gentamicin, streptomycin and tetracycline (6.3%).

Phenotypic analysis of MIC concentrations for streptomycin and gentamicin
Table 4 presents the cumulative MIC percentages to streptomycin and gentamicin.Th e MIC90 of isolates to streptomycin was 16 μg/mL, whereas to gentamicin MIC90 was 1 μg/mL.Gentamicin 58.3 82.7 85.3 88.3 98.5 98.9 98.9 100 Occurrence of resistance determinants Table 5 presents the prevalence of resistance genes strA/strB and aadA1 among E. coli isolates from diff erent categories of pigs and manure lagoons resistant to streptomycin and spectinomycin.Th e highest occurrence (54.0%) was that of aadA1 among isolates resistant to streptomycin and spectinomycin, whereas 32.3% of isolates were positive for strA/strB.Th e combination of aadA1 and strA/strB genes was determined in 3.1% of strains.Th e analysis of data on the distribution of resistance genes among the diff erent age categories, the highest prevalence of aadA1 (23.4%) was observed in fi nisher pigs, while strA/strB genes were the most frequently encountered among weaned pigs.Isolates form suckling pigs also showed a higher prevalence of aadA1 (7.8%).Higher prevalence of aadA1 (2.1%) was established in E. coli isolates from manure lagoons as compared to strA/strB positive strains (1.0%).Th e combination of aadA1 and strA/strB was observed in 3.2% of resistant strains; similar were percentages (1.5%, 1.0%) among isolates from weaned and fi nisher pigs.Th e same combination of resistance genes was not found out in isolates from suckling pigs.
None of E. coli strains resistant to gentamicin has exhibited the аасС1 gene.

DISCUSSION
Molecular characteristics of the commonest co-resistant phenotypes in commensal E. coli isolates from domestic animals are related to the presence of bla TEM-1 coding resistance to ampicillin, aadA1 and strA/strB determining streptomycin resistance, tet (A) and tet (B) in tetracycline-resistant strains, sul1 in sulfamethoxazole-resistant and dfrA1 -in trimethoprim-resistant isolates.Th e resistance to gentamicin among commensal poultry and swine E. coli isolates is outlined with an ascending trend (Szmolka et al., 2013).Th e authors presented data from the Hungarian Antimicrobial Monitoring System, showing that the resistance to gentamicin among commensal porcine E. coli isolates kept the usual low levels throughout the monitoring period from 2004 to 2008.In their view, the prevalence of strA, strB, and aadA1 genes among porcine E. coli strains was within the ranges 30%-60%, 60%-100%, and 1-30% respectively.Sundin et al. (1996) commented the wide spread of strA/strB genes in the environment and the relationship between their spread and exchange in the diff erent ecological niches, plants, livestock animal species and humans.Data about the resistance to streptomycin among commensal E. coli isolates from pigs show a substantial variability.For instance Wasyl et  Th e prevalence of resistance to streptomycin (70.0%) and spectinomycin (65.5%) among commensal porcine E. coli isolates in the present study was comparable to the results of Stannarius et al. (2009) and Mazurec et al. (2013).It should be noted that according to our data, the resistance to gentamicin (12.4%) was considerably higher that reported by Szmolka et al. (2012).
Th e percentage of strA/strB positive streptomycin-resistant strains in this study (32.3%) was the same as results presented by Szmolka and Nagy ( 2013), yet the occurrence of the aadA1 gene (54.0%) was higher.

CONCLUSION
Th e established resistance to streptomycin and spectinomycin to commensal E. coli strains from pigs was close to the highest percentages reported by diff erent research teams from the EU member states.As the genetic resistance profi le was concerned, the prevalence of the aadA1 genes was incontestable among isolates from the diff erent age categories of pigs and environmental strains (from manure lagoons).Th e lack of аасС1 genes from the genetic profi le of resistance to gentamicin was not an exception as could be seen from limited data on their prevalence among domestic animals and at present, the data for the occurrence of these genes in commensal porcine E. coli strains are predominantly from the Asia region.

Table 2 .
Prevalence of antibiotic resistance in E. coli strains from weaned pigs from 6 farrow-to fi nish farms

Table 3 .
Prevalence of antibiotic resistance in E. coli strains from fi nishers pigs from 6 farrow-to fi nish farms

Table 4 .
Distribution of MICs among commensal E. coli (n=274) isolated from pigs and lagoon manure

Table 5 .
Occurrence of resistance genes determined among commensal E. coli (n= 274) from pigs and lagoon manurе al. (2007) reported the presence of resistance in 37.0% commensal porcine E. coli isolates while streptomycin resistance among isolates from weaned animals reported by Stannarius et al. (2009) was 60.6%.Th e results of Mazurec et al. (2013) with respect to streptomycin resistance showed that it was present in 88.3% of tested commensal E. coli strains from swine.