Introduction
Biofilms are defined as microbial derived sessile communities which attach themselves to solid surfaces or to each other by using their sticky appendages and employing a rolling motion. So they could show continuous attachments to the surface and forming micro aggregation (Taj et al., 2012). Previous investigations have shown that biofilms are resistant to antimicrobial agents (Mohammad and Shalakany, 2015). Bacteria growing in a biofilm are 1000 fold resistant than planktonic cells to antibiotics. Therefore, higher concentrations of antibiotics are needed to inactivate bacteria grown on a biofilm (Hassan et al., 2011).
Antimicrobial agents have a significant role in decreasing the percentage of infectious diseases in both animals and
humans (Tadesse et al., 2012). However the constant use of antimicrobials over a period of time is one of the important reasons leading to prevalence of antibiotic-resistant bacteria (Van den Bogaard et al., 2001). The most significant factor for the development of antibiotic resistance is the exposure of different antibiotics with different concentrations into the microflora of human and animal guts. Over a period of time under the selective pressure, resistant bacteria will have overgrowth (Sayah et al., 2005). In veterinary medicine, antimicrobial agents are commonly used as growth promoter which leads to the high resistance toward antibiotics in normal and pathogenic bacterial flora of poultry (Romanus et al., 2012). Since resistant bacteria from food animals may colonize the humans through the food chain, contact via occupational exposure or waste run off from animal production facilities, it is possible that antibiotic-resistant bacteria could be transferred from food animals to man (Schroeder et al., 2002).
E. coli is known for its ability to cause different types of infections. Among them, gastrointestinal infections leading to the diarrhea are important. Additionally, a variety of diseases outside of the gut of humans and animals such as urinary tract infections, sepsis, meningitis, abdominal infections, osteomyelitis, cellulitis, wound infections and colibacillosis are significant (Kazemnia et al., 2014). E. coli colibacillosis usually causes the considerable damages to poultry production all around the world (Rafiei Tabatabaei and Nasirian, 2003). Surveillance data indicated that E. coli have strong ability for biofilm formation and emergence of resistance in E. coli is one of the highest for antimicrobial agents that have been in use the longest time in human and animals (Tadesse et al., 2012).
In this study, E. coli strains isolated from chicken and human clinical samples were analyzed to determine their biofilm producing properties and their susceptibilities to antimicrobial agents.
Materials and Methods
Bacterial strains
A total of 100 strains of E. coli from humans and chicken samples; including 50 samples from human urine and 50 samples from chicken muscles, were isolated from September to December 2013 from Urmia, Iran.
Bacteriology
All samples were macroscopically and microscopically examined by gram staining (Abdul Rahaman Shariff et al., 2013). Then cultured on MacConkey and Eosin methylene blue agar plates (Sigma Aldrich, USA) and incubated at 37oC for 24h. The colonies suspected to be E. coli were identified by other bacteriological assays (Hammerum and Heuer, 2009).
All human and chicken isolates were frozen in Nutrient broth containing 30% glycerol at -70oC until further processing (Tadesse et al., 2012).
Molecular identification of E. coli
All human and avian E. coli isolates were sub cultured overnight in Nutrient broth media (Sigma Aldrich, USA) and genomic DNA was extracted using bacterial genomic DNA purification kit (Intron, Korea). Then the presence of E. coli 16s rRNA gene was determined by polymerase chain reaction (PCR) method using forward primer:
5'-GTATAGATACCCTGGTAGTCCA-3' and reverse primer:
5'-CCCGGGAACGTATTCACCG-3' (Sharma et al., 2013). The PCR assay was done in a total volume of 25 μl using Intron premix (Korea). The PCR was performed in a DNA thermo cycler (MWG AG Biotech Thermal Cycler, USA) and the PCR conditions were as follow: an initial template denaturation at 95oC for 3 mins, 26 cycles followed with DNA denaturation at 94oC for 1 min, annealing at 55oC for 1 min and extension at 72oC for 10 mins (Sharma et al., 2013). The PCR products were separated electrophoretically in 1% agarose gel.
Detection of biofilm formation
Biofilm formation by E. coli isolates was assayed by microtitre plate, tube and cover slip assays. In microtitre plate assay 1×107 cfu of E. coli in 100 microliter LB Broth medium was inoculated into the wells of 96 well flat bottom polystyrene plates (Sigma Aldrich, USA). After growing bacteria at 37oC for 48h, the planktonic cells in media were discarded. The plate was washed and the attached biofilms were stained with 0.1% crystal violet (20 min). After rinsing the plates with distilled water, all stains associated with the attached biofilms were dissolved with 95% ethanol and OD595 absorbance was measured (Nakao et al., 2009). An optical density of 0.240 was chosen to distinguish biofilm producers from non-biofilm formers (Van den Bogaard et al., 2001).
In the tube method, a loop full of bacteria was cultured in 10 ml trypticase soy broth with 1% glucose. After incubation at 37oC for 24h, tubes were decanted and washed with phosphate buffer saline. Then tubes were stained with 0.1% crystal violet stain. Each tube was then gently rotated to ensure uniform staining and then the contents were gently decanted. The tubes were placed upside down to dry. Biofilm formation was considered positive when a visible film lined the wall and the bottom of the tube (Hassan et al., 2011).
In coverslip method, sterile culture flasks were filled with 50 ml Brain heart infusion broth media and 18mm glass microscope cover slip added to it. Then defined volume of overnight culture of bacteria was inoculated into the flasks. After incubating the flasks at 37oC for 48h, glass cover slip containing biofilm was removed and rinsed with phosphate buffer saline then stained with 0.5% crystal violet stain for 5min. Stained biofilms were then microscopically screened (Sharma et al., 2013).
Antimicrobial drug susceptibility testing
While a single E. coli isolated and identified from each collected sample, the antimicrobial susceptibility tests were done on Muller-Hinton agar plates using Kirby Bauer disk diffusion method. A 150mm Muller- Hinton medium plate was swabbed with Nutrient broth inoculated with E. coli, and incubated to a turbidity of 0.5 McFarland standard medium. Twelve prepared antimicrobial agent disks were place on the inoculated plate. Then these plates were incubated at 35oC for 18 to 20h. The diameter zone of growth inhibition around each disk was measured (Sayah et al., 2005). Selected disks for this study were: Tetracycline(30 μg), Ceftriaxone(30 μg), Doxycycline(30 μg), Amikacin(30μg), Cefixime(5 mcg), Gentamicin(10μg), Furazolidone(100 mcg), Amikacin(30μg), Ceftazidime(30-μg), Cefotaxime(30 μg), Nalidixic
acid(30μg), Nitrofurantoin(300 mcg), Ciprofloxacin(5μg) (Padtan Teb, Iran). The diameters of the zones of inhibition were interpreted by referring to the Performance Standard for Antimicrobial Susceptibility Testing; Twenty-Third Informational Supplement – January 2013.
Results
Bacteria detection
All 100 isolated bacteria from human and avian samples had the cultural, morphological and biochemical characteristics of E. coli (Fig. 1). Moreover, molecular identification of E. coli 16s rRNA gene using PCR method confirmed the E. coli isolates. Fig. 2 shows the product of 16s rRNA gene found at 612bp using 100bp DNA marker.

Fig. 1. Cultural characteristics of E. coli. A) E. coli colonies have a characteristic green sheen on Eosin methylene blue agar plate. B) E. coli appeared as pink colonies on MacConkey Agar plate.

Fig. 2. Gel electrophoresis of PCR products of 16s rRNA gene. Lane M- 100bp ladder marker, Lane 1; positive control (E. coli 25922), Lanes 2-5; human E. coli 16s rRNA gene found at 612bp, Lanes 6-8; human E. coli 16s rRNA gene found at 612bp.

Biofilm formation
Among 50 human E. coli isolates, microtitre plate method detected 72% as biofilm producers, which followed by tube test (66%) and cover slip assay (56%). Thirty two percent of human E. coli strains were confirmed to have biofilm forming ability by all three methods. Thirty four percent were positive biofilm producers by two assays. Thirty two percent were detected as biofilm producers by only one method. Two percent of isolates did not have any ability in biofilm formation.
Among 50 avian E. coli isolates, microtitre plate method detected 38% biofilm producers which followed by tube test (60%) and cover slip assay (54%). Eighteen percent of human E. coli strains were confirmed to have biofilm forming ability by all three methods. Fifty four percent were positive biofilm producer by two assays and 28% were detected as biofilm producers by only one method.
Two percent of isolates did not have any ability in biofilm formation (Fig. 3, 4, 5).

Fig. 3. Microtitre assay plate for E. coli biofilm.

Fig. 4. Biofilm formation of E. coli by tube assay.

Fig. 5. Microscopic view of biofilm formation of E. coli by cover slip assay.
Antibiotic susceptibility results.
Fifty samples from chickens and fifty samples from humans E. coli isolates were compared for antimicrobial resistance prevalence (Table 1). Table 1 shows the percentage of susceptible, intermediate and resistant isolates to each antimicrobial agent. The most frequent antimicrobial resistance among avian isolates was with Nalidixic acid (100%), Ciprofloxacin (80%), Doxycycline (80%) and Tetracycline (76%).
Avian E. coli expressed resistance to Nitrofuran derivatives, Nitrofurantoin and Furazolidone at frequencies to 24% and 20%, respectively. Amikacin resistance occurred at a frequency of 28%. The resistance for Cephalosporins, Cefotaxime, Ceftazidime, Cefixime and Ceftriaxon among the isolates was 30%, 22%, 10% and 30%, respectively. No Gentamicin resistant E. coli was isolated from avian samples (Table 1). However, the majority of human E. coli isolates were resistant to Tetracycline (62%), Ciprofloxacin (58%), Doxycycline (58%) and Nalidixic acid (50%). The next most frequent resistance phenotypes were resistance to Cefotaxime (36%) and Ceftazidime (40%). Twenty four and 20 percent of human E. coli isolates were resistant to Ceftriaxon and Cefixime, respectively.
Resistance to aminoglycosides; Amikacin and Gentamicin was seen in 16% and 8% E. coli samples, respectively. No Nitrofurantoin resistant E. coli was isolated from human samples (Table 1).
Table 2 shows the prevalence of multiple resistance patterns in avian and human E.coli isolates. The most frequently observed resistance pattern in avian isolates was resistance to Nalidixic acid in combination with Ciprofloxacin, Tetracycline and Doxycycline (22%).
Table 1. The percentages of susceptible (S), intermediate (I) and resistant (R) E. coli isolates from human and avian sources to individual antibiotics.

The next most frequent resistant phenotypes were resistance to Nalidixic acid, Ciprofloxacin and Tetracycline (18%), and Nalidixic acid, Ciprofloxacin, Ceftazidime, Cefotaxime, Cefixime, Cefriaxon and Tetracycline (8%). Six percent of avian E. coli isolates exhibited resistance to Nalidixic acid, Ciprofloxacin, Amikacin, Tetracycline and Doxycycline. Moreover, 6% of avian E. coli isolates was resistant to Nalidixic acid, Ciprofloxacin, Nitrofurantoin, Tetracycline and Doxycycline. Six percent of avian E. coli isolates exhibited resistance to Nalidixic acid, Ciprofloxacin, Amikacin and Tetracycline while 4% of them exhibited resistance to Nalidixic acid, Ciprofloxacin, Amikacin and Tetracycline, Doxycycline and Nitrofurantoin.
Resistance to Nalidixic acid, Ceftazidime, Cefotaxime and Ceftriaxon were exhibited in 4% avian E. coli isolates. The remaining 26% of avian E. coli isolates exhibited a single unique phenotypic pattern (Table 2).
Among human E. coli isolates the most multiple resistance patterns were for Doxycycline and Tetracycline (10%).
The next resistant phenotype was Doxycycline, Tetracycline and Nalidixic acid (8%). Six percent of human E. coli isolates were resistance to Cefotaxime, Doxycycline, Tetracycline, Ciprofloxacin, Ceftazidime and Nalidixic acid. Resistance to Cefotaxime, Doxycycline, Tetracycline and Ceftazidime was seen in 4% of E. coli samples. In addition, 4% of isolates showed resistance to Cefotaxime, Doxycycline, Tetracycline, Ciprofloxacin and Ceftazidime. Fourteen percent of human E. coli samples exhibited a single unique phenotype pattern. The remaining isolates showed resistance to only one antibiotic (Table 2).
Since the biofilm forming isolates from both human and avian sources showed multidrug resistance, their antibiotic resistance patterns was identified. A total of 37.5% avian biofilm forming E. coli showed resistance to 6 antibiotics. Among them, 31.25%, 80.75% and 6.25% had resistance to 5, 3 and 1 antibiotics, respectively.
Among human biofilm formers 55% of biofilm producers showed resistance to four antibiotics, while 22% of them revealed resistance to two drugs. The same amount (22%) was evident for five antibiotics.
Table 2. Prevalence of multiple resistant patterns in avian and Human E. coli isolates

Key: Doxycycline=D, Amikacin=AN, Cefixime=CFM, Gentamicin= GM, Furazolidone=FR, Ceftazidime=CAZ, Cefotaxime=CTX, Nalidixic acid=NA, Nitrofurantoin=FM, Ciprofloxacin=CP, Tetracycline=TE.
Discussion
Biofilms are the community of bacteria attached to the solid surfaces by cell surface components such as flagella, type 1 fimbriae and outer membrane proteins (Wood, 2009). It demonstrated that formation of bacterial biofilms is a process that contributes to microorganism's pathogenicity (Rinser et al., 2006).
Antimicrobial resistance is one of the important problems in both human and veterinary medicine and antibiotic administration is recognized as the most significant factor for the prevalence of antimicrobial agent resistant bacteria (Sayah et al., 2005). The microbial ecosystems of humans and poultries are connected to each other, so that it is not easy to determine the antimicrobial resistance (Schroeder et al., 2002). Emergence of antimicrobial resistance may decrease the time that these drugs will be useful for effective treatment of infections (Hammerum et al., 2009).
E. coli was chosen for this study because it’s a ubiquitous bacterium both in the nature and in the intestine of both animal and human. Although, resistant strains from the intestine may contaminate chicken carcasses at slaughter causing multi resistant E. coli emergence in poultry meats. The mechanism of the spread of antibiotic resistance from food animal to humans is not clear however, the colonization of human gut with resistant E. coli from chicken has been shown and it seems that animals act as a reservoir for E. coli found in humans (Nakao et al., 2012).
In 2000, Edland and Nord examined the effect of oral antibiotics for treatment of urinary infections on the normal human microflora and observed the overgrowth of resistant Enterobacteriacea (Wood, 2009).
Sayah et al. (2005) identified antimicrobial resistance patterns in E. coli isolated from domestic and wild animal fecal samples. Multidrug resistance was seen in a variety of sources (Sayah et al., 2005).
In a study conducted by Raum et al. (2007), the changes in E. coli resistance patterns during and after antibiotic therapy were investigated concluding that there was a significant increase in the prevalence of E. coli isolates during antibiotic treatment.
In addition, Momtaz et al. (2012) studied the distribution of antibiotic resistant genes in E. coli isolated from slaughtered commercial chickens by PCR. All isolates showed one or more antibiotic resistance genes (Miles et al., 2006). Looking to the association between biofilm production and antimicrobial resistance, Sharma et al. (2013) indicated that resistant strains of E. coli isolated from sewage water had strong biofilms forming ability which is in accordance to the finding of this study.
Evaluating different biofilm detection methods, Hassan et al. (2011) concluded that microtitre plate method is reliable than the tube test for detection of biofilm forming microorganisms. Their finding is in accordance of the results of this study indicating that microtitre plate method is the superior one.
Mohamad and Shalakany (2015) also found that antibiotic resistance was higher among biofilm producers to antimicrobial agents in comparison with non-biofilm producers. They also have reported that microplate method was reliable than the tube test in biofilm producers detection, which is again in accordance to the present findings.
In another study, the biofilm production of Staphylococcus aureus was comparatively evaluated by three conventional methods (Taj et al., 2012). According to their findings, the tube test was more sensitive than coverslip assay in screening of biofilm producers which was similar to the present study. To unveil the evolution of drug resistance in E. coli, the existing strain collections of these bacteria for their resistance to a common panel of 12 antibiotics were tested in the present study. After testing 100 E. coli isolates from human and chicken, the results suggested that the resistant strains of E. coli were common among chickens and human samples. These results are consistent with the findings of previous studies. Multi-drug resistance was observed in E. coli from both animal and human sources however, it was higher in frequency and proportion in E. coli isolated form chicken samples. This is consistent with the reports of other studies performed elsewhere (Kibert et al., 2011). In conclusion, antimicrobial resistance is more common in biofilm forming E. coli and chicken can be a source of transferring antimicrobial resistant bacteria to human.
Ethics
It is declared that all ethical standards have been respected in preparation of this article.
Acknowlegement
This work was financially supported by the University of Tabriz.