Detection of antibiotic resistance genes of Escherichia coli from domestic livestock in south east Nigeria with DNA microarray

DNA microarray was developed for detection of up to 90 antibiotic resistance genes in Escherichia coli by hybridization. Each antibiotic resistance gene was represented by two specific oligonucleotides chosen from consensus sequences of gene families. A total of 203 oligonucleotides (50-100 base) were spotted onto the microarray. The sequence identity of each gene was compared with GenBank sequences, biotin was used as the positive control and 16s rRNA as orientation. Of the 40 E. coli isolates analyzed in this study, 37 were identified as having, at least, one antibiotic resistance gene. Among the different antibiotic resistance genes detected, bla-CMY-2 and strA were the most prevalent occurring in 28 (70%) of the isolates, respectively. Other common genes included were TEM1 11(27.5%), Sul2 14 (35%) and TetA 21(52.5%). The microarray genotyping corresponded with the phenotype of the strains. The disposable microarray presents the advantage of rapidly screening bacteria for the pre-sence of known antibiotic resistance genes. This technology has a large potential for applications in basic research, food safety, and surveillance programs for antimicrobial resistance.


INTRODUCTION
During the past decades, the worldwide use of antibiotics in animal husbandry for purposes of prophylaxis, chemotherapy and growth promotion has created enormous pressure for the selection of antibiotic resistance among bacteria (Vincent et al., 2005).Today, there is increasing concern about the severity of antibiotics resistance in Escherichia coli, which is an important reservoir of antibiotic resistance genes; many other enteric pathogens and commensal bacteria may also play a role as reservoirs for antibiotics genes (Greg et al., 2010;Ma et al., 2007).It is therefore important to follow the evolution of antibiotic resistance in the bacterial population in order to prevent and repress the emergence of multidrug-resistant strains of those bacteria that can still be treated with antibiotics.
The disc diffusion assay technique is commonly used to determine the resistance of pathogenic or commensal bacteria because of its simplicity and because it provides information that is useful in prescribing appropriate antibiotics.Phenotypic testing such as disc diffusion assay technique, however, will not detect ''silent'' antibiotics resistance genes that might be expressed in vivo or disseminated to other bacteria (Frye et al., 2006;Nsofor andIroegbu 2012, 2013).Molecular testing methods offer similar information more quickly and provides for more discriminatory information.*Corresponding author.E-mail: nsoforac@gmail.com.
Because of the large number of recognized antibiotics resistance genes, parallel detection systems such as microarray are well suited to this task (Call et al., 2003).
Presently, PCR and hybridization analysis are common methods used to detect antibiotic resistance genes in bacteria.However, the detection of specific resistance genes remains a tremendous amount of work if every possible resistance gene has to be assessed, and therefore microarray technology is most suitable for resistance gene analysis (Holzman, 2003).A few microarrays have been developed for identifying antibiotics resistance genes (Call et al., 2003;Frye et al., 2006;Moneeke et al., 2003).This study describes a microarray technique for detecting the genes that confer resistance to aminoglycosides, beta-lactam, chloramphenicol, sulfonamide and tetracycline.

Specimen collection, cultivation and identification of Escherichia coli
Fresh fecal droppings were randomly collected from goats, cattle, pigs and chicken; and care was taken to avoid collecting more than one fecal sample per individual animal.One gram of each animal's feces was homogenized in 9 ml of sterile saline solution, then the volume of the homogenate was made up to 10 ml to get a 10% suspension.The contents were mixed thoroughly and 10-fold serially diluted and 0.2 ml inoculums from each dilution plated out on Eosin Methylene Blue agar (EMB) (Oxoid, England).No antibiotic was included in the EMB agar plates used for the cultivation.The inoculated plates were incubated overnight at 37°C.A single colony on EMB with green metallic sheen taken to be E. coli was selected from an individual fecal sample for further characterization.E. coli was fully identified using conventional microbiological tests-Indole positive, methyl red positive and citrate negative (Cheesbrough, 2000).The cattle and goat specimens came from the herd at Obinze Owerri, Imo State while the Madonna University Poultry Okija, Anambra State was the source of poultry specimens.The specimens from swine came from a farm located at the Ogborhil area of Aba, Abia state.

Preparation of microarray slides
Multiple DNA microarrays were printed on glass slides so that independent arrays were contained within ten individual wells defined by Teflon masking slides (Erie Scientific, Portsmouth, N.H.USA); the hydrophobic nature of the masking permitted independent samples to be hybridized within each well.Slides were derivatized with epoxysilane (3-glycidoxypropyltrimethoxysilane; (Sigma-Aldrich, Milwaukee, WS, USA) as described by Call et al. (2001).Prior to printing, the slides were soaked in 2.5% Contrad 70 detergent (Fisher Scientific, Pittsburgh, PA, USA.) for 2 min, rinsed three times with distilled water, and dried using compressed air.Slides were then soaked for 1 h in 3 N HCl, rinsed three times with deionized water, and dried with compressed air.

Construction of DNA microarray
Oligonucleotide probes of known antibiotics resistance genes were reconstituted in TE buffer, diluted to 60 µm in print buffer (0.1 M Na2HPO4, 0.2 M NaCl, 0.01% sodium dodecyl sulfate) with a pH of 11 and transferred to 384-microwell plates for printing.Arbitrary biotinylated oligonucleotides (70-mer; 5 µM) were included with every array.These biotin pseudoprobes served as positive controls for the detection chemistry and to orient the array for image processing.All probes were deposited as four replicates at a fixed location within each masked well using a Robotic Microgrid II arrayer (Bio-Robotics, Woburn, Mass.USA) with humidity held at 45%.Printing parameters included washing the pins in a recirculating bath (four pins washed twice for 4 s each time), followed by 0.5 s of flushing and 6 s of drying.This washing procedure was repeated twice between probes to minimize possible probe carryover.Printed slides were baked under vacuum (22 Hg/mm) for 1 h (130°C) and stored away from light at room temperature until used.

Genomic DNA extraction
The bacterial total DNA was extracted using the Qiagen DNeasy silica-gel adsorption method (Qiagen, Valencia, CA USA).
A 1.0-ml volume of overnight broth culture of the test isolate was pelleted in a 1.5 ml microcentrifuge at 10000 rpm for 10 min and resuspended in180 μl of buffer ATL from the Qiagen DNeasy kit.Then 20 μl of Qiagen proteinase K solution was added, mixed by vortexing and the cell was incubated for 3 h in a 55°C shaker water bath for lysis.After the lysis, 20 μl of RNase A (100mg/mL) (Qiagen, Valencia, CA USA) was added to each tube (to degrade RNA) and the tubes were incubated at room temperature for two minutes.This was followed by the addition 200 μl of buffer AL, vortexing, and incubation at 70°C for 10 minutes.Then, the genomic DNA (gDNA) was concentrated by the addition of 200 μl of 100% ethanol.To separate the DNA from other cellular contaminants, the treated DNA lysate was pipetted into a DNeasy column in a collection tube, and centrifuged for 1 min at 10,000 xg.The remaining contaminants were washed out by using 500 μl each of buffer AW1 and AW2 in a new collection tube at each time.The purified gDNA was eluted in a fresh1.5 ml micro-centrifuge tube by using 200 μl AE buffer and centrifugation for 1 min at 10,000 xg.Finally, the nanodrop spectrophotometer was used to quantify the DNA.DNA was quantified to properly scale the subsequent nick translation and any sample that failed to reach the value of A260/A280 ratio of 1.7 to 2 or below 25 ng/μl was re-extracted.All the buffers, enzymes and columns used in this extraction came from the Qiagen DNeasy kit (Qiagen, Valencia, CA USA; Cat.No. 69504).

Nick translation: Biotinylation and fragmentation of DNA
This reaction is designed to generate small (50-100 base) biotinlabeled DNA probes by nick translation which are important for successful in situ hybridization.
Approximately 1.0 μg (up to 40ul) of the quantified gDNA, 5 μl of 10X dNTP mix [(0.2 mM each of dCTP, dGTP, dTTP; 0.1 mM of dATP; 0.1mM of biotin-14-dATP; 500mM of Tris-HCl, pH 7.8; 100mM of βmercaptoethanol and 100 μg/ml of nuclease-free BSA) (Invitrogen, USA)] and 5 μl of 10X enzyme mix [0.5U/μl of DNA polymerase 1, 0.007 U/μl of DNase 1, 50 mM of Tris-HCl pH 7.5, 5 mM of magnesium chloride, 0.1 mM of phenylmethylsulfonyl fluoride, 5% (v/v) of glycerol and 100 μg/ml of nuclease-free BSA) (Invitrogen, USA)] were combined in 0.2 ml PCR tubes on ice.The total volume was brought to 50 μl with PCR water.The mixture was incubated at 16°C in a thermal cycler for 2 h and then held at 4°C for nick translation of DNA.To precipitate the nick translated DNA, the samples were transferred to 1.5 ml micro-centrifuge tubes followed by the addition of 5 µl of 3 M sodium acetate, (pH 5.2), 110 μl of 100% ethanol and incubation at -80°C for 30 min.After the incubation, the DNA was pelleted by centrifugation at 14000 rpm for 30 min at 4°C.Then, the pellets were resuspended with 400 μl of 70% ethanol.For more purification, the above steps were repeated once and the pellets were dried with a vacuum centrifuge for 10 min.Finally, the purified nick-translated DNA was resuspended with 100 μl 1x hybridization buffer.

Microarray slide pre-hybridization preparation
Microarray slides were prepared by immersing them in 50 ml of 1% BSA blocking solution in a Coplin staining jar followed by incubation at room temperature for 10 min, with shaking at 80 rpm to eliminate bubbles on the slide surface.The slides were rinsed 20 times in double de-ionized after which their back and edges were wiped with a Kimwipe and spin dried with slide centrifuge for 15 s.

Sample application/hybridization
The nick translated gDNA was boiled for 3 min, chilled on ice and briefly vortexed for 15 s.Then, the microarray slides were placed on a humidified chamber (200 μl tip box and lid with de-ionized water covering the bottom of the box) and 45 μl of the gDNA sample was placed in each well (2 wells per nick translated gDNA sample) on the microarray slide.The droplets were carefully spread to fully cover the well without touching the slide surface with the pipette.Carefully, the slide was sealed (face-up and frosted end toward the cap) in a hybridization chamber (50 ml conical tube with filter paper moistened with 1x hybridization buffer).The slide was placed on top of the filter paper in the hybridization chamber without allowing the damp filter paper to touch the wells.The hybridization chamber was placed in a rack and lead weight on top of the rack, then the rack was submerged in the 55°C water bath.Finally, the sample DNA was allowed to hybridize with the probes on the array for 16 h.

Post-hybridization stringency washes
After hybridization, the slides were removed from the hybridization chamber with forceps and excess hybridization solution was aspirated off the slides.Then, the slides were completely immersed (frosted end up) in a 55°C pre-warmed low stringency array wash solution (1X SSC, 0.2% SDS) contained in a Coplin jar.The above procedure was repeated in medium stringency (0.1XSSC, 0.2% SDS) and high stringency (0.1XSSC) array wash solutions, respectively.At each time, the slides were washed for 4 min at room temperature on an Orbital shaker at 80 rpm.After the stringency washes, the slides were transferred to a horizontal staining jar that contains enough TNT buffer to cover the slide and were shaken for 1 min at 80 rpm at room temperature to remove the stringency wash buffers.This TNT buffer washing was repeated three times.

Microarray development
For the following applications, 45 μl of each solution was added directly to each well.The slides were gently tapped to distribute the reagent over the full well surface without allowing the reagents to cross over to other wells.The slides were spin-dried for 5 s using a slide centrifuge followed by incubation with 1:100 Streptvadin-Horseredish peroxidase (SA-HRP) in TNB for 30 min.After the incubation, the slides were washed 3 times for 1 min each in horizontal staining jars at 80 rpm shaking.The above procedure was repeated with 10% FES, 2XSSC; 1:50 BioT, 1xAmp Dil; and 1:500 SA-Alexa 555, 1XSSC, 5X Den.This last incubation was done for one hour in the dark.All incubation was done at room temperature in a humidified chamber (made from a covered tip box with ~10 ml PCR water in the bottom).At the end of these development reactions, the slides were spin-dried for 15 s using the slide centrifuge and were stored in the dark prior to scanning.

Scanning/imaging of slides
After hybridization and development, slides were scanned or imaged by standard DNA microarray slide scanners.The florescence marker used in this experiment (Alexa555) has an optimal excitation wavelength of 555 nm and emission wavelength of 565 nm.The scanner/imager we used (Applied Precision arrayWoRx scanner) has a white light source and an emission filter for Cy3 that functions well for Alex555.We used an excitation wavelength of 540 nm (25 nm bandwidth) and an emission wavelength of 595 nm (50 nm bandwidth).
There were five pairs of Teflon-masked wells on each slide, with each well containing a full array and our normal protocol calls for two wells to be hybridized to the same sample.Within each well there were two spots per probe so in effect there are four individual probe-target hybridizations (2 wells total).Each full array has dimensions of 22 horizontal and 20 vertical spots.The distance between spots is approximately 250 μm.Table 1 shows the oligonucluotide probes sequences used in constructing the DNA microarray.

Detection of antimicrobial resistance gene with microarray
Forty (40) E. coli isolates were tested for antimicrobial resistance genes with the microarray.Thirty seven isolates were identified as having at least one antimicrobial resistance gene.Three remaining isolates (CA2, cow; GO3, goat; PL18, poultry) did not hybridize to any of the resistance genes presented on the array.Multiple antimicrobial resistance genes belonging to same category of antimicrobials were detected in most isolates.Among the aminoglycosides, the most prevalent resistance genes were aadE and strA, 28 (70%) respectively, the most prevalent host were the isolates from poultry 07 (87.5%) (Table 2).The most encountered beta-lactam gene in this study was bla-CMY-2, 28(70%).However, blaCTX-M-12 and blaIMP-2 were detected only in isolates from poultry specimens (Table 3).The most prevalent chloramphenicol resistance genes observed in this study was floR, 22 (55.0%), while Integrase gene, int1 had the highest occurrence rate of 37.5% (15 isolates) (Table 4).In the trimethoprim and sulfonamide resistance gene families, the most prevalent was dhfrV, which was detected in 9 isolates (22.5%).For sulfonamide resistance genes, 14 isolates (35%) of the animal specimens harbored Sul2 at highest rate.The dhfrII gene was only detected in isolates from pigs and poultry (Table 5).Among the tetracycline resistance genes, TetA was most with 21 isolates (52.5%) of animal specimens bearing this gene (Table 6).A sample micrograph of microarrays hybridized with genomic DNAs of the E. coli isolates are shown in Figure 1.

DISCUSSION
DNA microarrays have been used previously to detect resistance genes in bacteria (Call et al., 2003;Frye et al., 2006;Moneeke et al., 2003;Van Hoek et al., 2005;Ma et al., 2007;Greg et al., 2010).Several types of DNA templates can be used to construct microarrays, depending on the intended use.For example, short oligonucleotide probes can be used to detect single nucleotide polymorphism, long oligonucleotide probes can be used to detect sequences that contain a few mismatches, and PCR probes can be used to detect moderately divergent genes.In the present study, oligonucleotide probes were used to construct microarrays that could identify up to ninety genes that confer resistance to variety of antibiotics used in combating Gram-ve bacteria like E. coli.
When compared with phenotypic testing, microarrays have the advantage of detecting the presence of antibiotic resistance genes that are not phenotypically expressed (Peterson et al., 2009).In this study, antibiotic resistance genes of 40 E. coli isolates from variety of domestic live stock viz cattle, goats, swine and poultry in south eastern states of Nigeria were detected.It was observed that microarray detected genes that were not phenotypically expressed in the following isolates, PG6, PG 11-Swine (aadE,floR,OtrB,qnrA1,strA,TetD,strA); CA 12-Cattle (Aph E) and PL 7-Poultry (aadE,aphA7,floR,IncFII/OriB,IncP / trfA2,qnrA1,strA,TetE,TetJ).Ma et al. (2007) observed that two isolates of Salmonella which did not phenotypically express resistance to aminoglycosides were harboring aadA1 and aadA2 genes, while Maynard et al. (2003) found that two E. coli isolates harboring the aph(3)-la gene, which confer resistance to Kanamycin and Neomycin, were susceptible to Kanamycin and Neomycin.Thus, our results and those of Ma et al. (2007) and Maynard et al. (2003) indicate that some antibiotic resistance genes are silent in bacteria in vitro; however, these silent genes can spread to other bacteria or turn on in vivo, especially under antibiotic pressure.
Furthermore, there were also discrepancies between the absence of the antibiotic gene test on the microarray and the phenotypic resistance (false negative).This was observed in isolates GO13-Goat (Am-C-Sxt-S-T-Amc); CA 9-Cattle (Am); and PL 18-Poultry (Am-C-Sxt-S).Resistance was phenotypically observed against the antibiotics written against each of the isolates but the genes were not detected by the microarray.This could be attributed to the non inclusion of the oligonucleotide probes encoding theses genes in the construction of the microarray or the genes encoding the resistance are novel.However, more research is needed in this area before conclusion can be established.
In conclusion, the microarray technique employed in this study proved to be an efficient method that allows for rapid detection and identification of resistance genes in E. coli isolates.

Table 2 .
The Prevalence of aminoglycosides resistance genes in E. coli Isolates.

Table 3 .
The prevalence of beta-lactam resistance genes in E. coli isolates.
N = Number of isolates hybridized.

Table 4 .
The prevalence of chloramphenicol and qinolone resistance genes in E. coli isolates.
N = Number of isolates hybridized.

Table 5 .
The prevalence of trimethoprim and sulfonamide resistance genes in E. coli isolates.
N = Number of isolates hybridized.

Table 6 .
The prevalence of tetracycline resistance genes in E. coli isolates.
N = Number of isolates hybridized.Figure 1. Microphotograph of microarrays hybridized with genomic DNAs of E. coli Isolates from cattle.