Molecular characterization of water-borne multi-drug resistant Escherichia coli

Antibiotics are very important in the fight against infectious disease caused by bacteria and other microbes for decades. Today microbes have developed ways to resist antimicrobial agents targeted at them. We sought to characterize and simultaneously detect virulence genes associated water-borne antibiotic resistant Escherichia coli, obtained from water sampled from ground and surface water sources. The Analytical Profile Index (API) was use for the identifications of E. coli isolates. The KirbyBauer disk diffusion method was employed for susceptibility testing. A DNA•STRIP molecular assay technology designed for detection of shiga toxin genes was used for the molecular characterization. E. coli isolates showed a high (32.99%), resistance to penicillin, and was highly susceptible (93.8%) to nitrofurantoin. E. coli was confirmed Polymerase Chain Reaction (PCR). None of the confirmed multidrug resistant E. coli isolate had genes for stx1 and stx2. However, the eae intimin virulence gene was found on 6% of the multidrug resistant E. coli isolates. It was concluded that E. coli has developed a very high resistance to the various antibiotics. Second, the genotype EHEC test based on the DNA•STRIP technology used in this study has proved to be efficient and reliable in the molecular characterization of the multi-drug resistant E. coli isolates. The DNA•STRIP Genotype EHEC technology test is therefore recommended for pathogenic E. coli detection and monitoring. There is also a need to revise strategies towards the multidrug resistance programme.


INTRODUCTION
Antibiotic resistance is a major challenge worldwide. It is a threat towards the attainment of the Sustainable Development Goals (SDGs) on health as well as gains made in health and development (Assembly, 2015(Assembly, , 2011. Today antibiotic-resistant genes imparting resistance to various antibiotics have been recognized in different water environments. This includes drinking water worldwide (Marathe et al., 2017;Odonkor and Addo, 2018). The major threat for health globally is the potential transmission of these genes from the environmental microbes unto human pathogens (Karkman et al., 2017). Prospect of drinking water passing on pathogens to *Corresponding author. E-mail: stodonkor@gimpa.edu.gh. Tel: +(233) 24 21550 or +(233) 23 4709702.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License people, thus causing disease, is well documented (Bengtsson-Palme and Larsson 2016; Pan and Chu, 2018).
The spread of multi-antibiotic resistant microbes in water environments is an important public health problem with stakeholders such as policy makers and physicians worried about their future capacity and ability to treat various infectious diseases Da Silva et al., 2011;Schmidt, 2002). Ultimately, the life time of an antibiotic hinges largely on the rise and dissemination of resistant bacteria. However, early interest relating to antibiotic resistant bacteria has predominantly been focused on medical settings and nosocomial infections.
The persistence of antibiotics in waterways accounts for the increase in bacteria resistance to antibiotics as observed in water environments. This ensues through "selective pressure" and horizontal gene transfer (Alonso et al., 2001). The phenomenon of selective pressure normally occurs in the presence of antibiotics owing to the heightened persistence or development of microbial strains carrying resistance genes (Economou and Gousia, 2015;Holmes et al., 2016). These genes block the mechanism of cellular destruction caused by antibiotic compounds. Resistance genes are usually found on plasmids that are susceptible to horizontal gene transfer. This explains medium through which antibiotic resistant bacteria and associated genes are spread in the water environment (Cheng et al., 2016;Gothwal and Shashidhar, 2015;Schmidt, 2002). It is important to note that these plasmids are extrachromosomal gene structures, and have the ability to move thus, allowing for plasmid encoded genes to be transferred across cell membranes. This method of gene transfers permits profusion and allows variety of resistant microbes to increase quickly in water environments contaminated with antimicrobial compounds. These waterways may therefore become reservoirs and incubators for antibiotic resistant bacteria (Alonso et al., 2001;Odonkor and Addo, 2018).
Escherichia coli is commonly recognized as the principal channel for the spreading of antibiotic resistance genes and associated virulence vectors. This is due to their profusion within such environments (Cheng et al., 2016;Gothwal and Shashidhar, 2015;Schmidt, 2002;Tauxe et al., 1997). For example, β-Lactam antibiotics are known to account for about a half of the global antimicrobial utilization (Alpay-Karaoglu, 2007). Resistance to various β-lactam antibiotics such as penicillins and cephalosporins is usually facilitated by the β-lactamase enzymes. E. coli strains have also become resistant to ampicillin by plasmid-mediated class A βlactamase enzyme such as TEM-1. They tend to mutate to extended spectrum activity (Odonkor and Addo 2011;Odonkor and Ampofo, 2013;Livermore, 1995). Epidemiological research findings show that TEM-1 is the most common plasmid mediated B-lactamase between medical Gram-negative bacteria (Bush and Jacoby, 1997).
However, there are other genes that are of public importance and are specific to E. coli. These genes include stx1 and stx2, the eae intimin gene, and the ipaH (invasion plasmid antigene H) gene (Odonkor and Ampofo, 2013). They do not only confer resistance to E. coli but they are also known to be toxin producing. Although there are several studies assessing multi drug resistance (MDR) in E. coli, not much work has been done on these genes specific to E. coli. The aim of this study was to therefore to characterize antibiotic-drug resistant water borne E. coli and simultaneously detect associated virulence genes.

E. coli isolation and identification
E. coli isolates were obtained from drinking water samples collected from ground water and surface water sources. E. coli organisms were identified using conventional methods (such as catalase test, indole test, nitrate reduction test, etc). This was then confirmed with Analytical Profile Index (API 20E). The E. coli control strain ATTC 25922 was obtained and used as the positive control.

E. coli susceptibility testing
Antibiotic susceptibility testing was done using the Kirby-Bauer disk diffusion methods as described by the Clinical and Laboratory Standards Institute (CLSI) (Lacy et al., 2004). Confirmed E. coli isolates obtained from overnight cultures on nutrient agar brought was brought to an equivalent of 0.5 McFarland. This was done by the suspension of isolates in sterile saline until the desire turbidity (0.5 McFarland) was achieved. The suspension was then streaked on complete surface of dishes with Mueller Hinton agar. The E. coli isolates obtained were tested against 14 antibiotics that were in use at the time of this study. These antibiotics and their strength are shown (Table 1).
All plates were incubated at 37°C for 24 h. This was after the antibiotic disks were placed aseptically on the streaked Mueller-Hinton agar dishes. Clinical and Laboratory Standards Institute (CLSI) protocols were used to interpret the results (Lacy et al., 2004).

Molecular characterization
A DNA strip technology test (GenoType EHEC) was used for the molecular characterization. The GenoType EHEC allows for a combined characterization and identification of the following genes: eae intimin gene, ipaH (invasion plasmid antigene H) gene, stx1 shiga toxin gene and stx2 shiga toxin gene. Three steps were involved in this molecular characterization method as follows: DNA extraction from E. coli culture, a multiplex amplification with biotinylated primers and a reverse hybridization. A template ensuring interpretation of the banding pattern was obtained and used (Prère and Fayet, 2005) Table 2 show the number of E. coli isolates     Table 4 provide a summary of resistances. It can be observed from the table that 10 of the E. coli isolates were resistant to just one antibiotic, while one E. coli isolate was resistant to 10 out of the 14 antibiotics tested.

Results in
However, of great interest and importance as shown in Table 4 is that 48 out of the 97 (49.48%) E. coli isolates tested were resistant to 2 or more antibiotics. Thus, falls under the multi drug resistance classification (Hill et al., 2005). Figure 1 shows a diagram of amplification of bands on the agarose gel. This is shown for the first few multi-drug resistant E. coli isolates. This is an important step, prior to hybridization. The success of DNA extraction and successful amplification is indicative of the presence of bands.
Results in Table 5 summarize the molecular characterization performed on the 48 multi drug resistant E. coli isolates. The PCR confirmed all isolates as E. coli. None of the confirmed multidrug resistant E. coli isolate had genes for stx1 and stx2. However, the eae intimin virulence gene was found on 6% of the multidrug resistant E. coli isolates.

DISCUSSION
Generally, a drug resistance refers to a phenomenon where there is a microbial resistance to at least two of the following classes of antibiotics: quinolones, aminoglycosides and lactams (Hill et al., 2005).
In this study, penicillin resistance was found to be very high. This has serious implications for public health, as the chances of these antibiotics curing infections caused by E. coli are significantly hampered. On the other hand, it is worth nothing that nearly all the 97 strains of E. coli (93.8%) found to be susceptible to nitrofurantoin, thus presenting a good choice for treatments. However, the rate at which these microbes are increasingly resistant to various antibiotics with an estimated over 10 million mortality attributable to antimicrobial-resistant infections annually by 2050 (O'Neill, 2018), the susceptibility of E. coli to just a single antibiotic as observed in this study may not be considered a significant threat. E. coli are generally apathogenic and belong to the enterobacteria group of microbes. They will not normally cause infections or disease. However, some strains such as Shiga toxin-producing E. coli (STEC), which are verotoxin-producing are known to possess pathogenic properties in humans as well as animals (Jaeger and Acheson, 2000).
Pathogenicity of a given bacteria strain such as E. coli is to a large extent influenced by virulence factors. In the case of E. coli, these factors include, capsule, adhesins, toxins and invasins, which are often present in a large genetic block on chromosome and may be horizontally transmitted between various strains. In this study, we found that 6% of the multiple resistant E. coli carried the eae virulence gene. Notwithstanding, a low percentage (6%) of virulence eae genes found on the stains of E. coli, the real danger is the potential to transfer these genes within the E. coli group and even with other members of the enterobacteria. Furthermore, presence of eae gene in water sources is of a grave public health concern, since spread of these virulence factors aside transmitting through food and water can also transfer from person to person (Karch et al., 2000).
Several disease outbreaks have been linked to the eae genes. The E. coli strains carrying eae gene have previously been isolated from adults who presented with the hemolytic-uremic syndrome. Similarly, it was also isolated in 101 children also presenting with hemolyticuremic syndrome. However, in this case, the E. coli strains lacked the intestinal adherence factor intimin, which is known to be encoded by the eae gene (Bielaszewska et al., 2006). Gerber et al. (2002) reported that 97% of E. coli that were shiga toxin producing were isolated from children in Australia and Germany, were found to carry the eae gene. It worth noting that in this particular case the estimated mean incubation period was 8 days, longer than the 3 to 4-day incubation period previously reported for shiga-toxin-producing E. coli O157: H7.13 (Mead and Griffin, 1998).
Also, in this study, we found that all multiple resistant E. coli isolates were non stx1 and stx2 genes producing. E. coli carrying shiga toxin (stx1 and stx2) genes are known to possess pathogenic properties in humans as well as animals, thus they cause infections (Jaeger and Acheson, 2000).
E. coli are also known to be significant reservoirs of several genes that code for antibiotic resistance (Bucknell et al., 1997). Thus, major risk for public health lies in the potential for the transfer of resistant genes from environmental bacteria to human pathogens. The capacity of the resistant bacteria and associated genes to move across ecosystems is well documented (Von Wintersdorff et al., 2013).

Conclusion
This study found that all E. coli was highly resistant to penicillin and highly susceptible to nitrofurantoin. Furthermore, none of the isolates was gene producing for the virulence factor Stx1 and stx2. However, 3 of the multidrug resistant E. coli isolates representing 6% were found to be eae virulence factor producing. Thus, the principal risk for public health is that resistance genes may be transmitted from environmental bacteria to human pathogens. We recommend the DNA•STRIP Genotype EHEC technology test is therefore recommended for pathogenic E. coli detection and monitoring. There is also a need to revise strategies towards the multidrug resistance programme.