The Escherichia coli bacterium is typically Gram-negative, characterized by its rod-shaped morphology, motility, lack of spore formation, absence of oxidase activity, production of indole, and inability to produce urease. It is capable of utilizing lactose and, as a result of glucose fermentation, generates both acid and gas. Additionally, it can thrive in environments with or without oxygen, particularly at a temperature of 37°C1,2.

The species E. coli is broadly dispersed and constitutes the major commensal of the human intestine including other warm-blooded animals and is used as a reference bacterium in many laboratory investigations1. E. coli serotypes can be isolated from various samples aside fecal material of warm-blooded animals. As such different serotypes of this organism are particularly introduced into the environment where they contaminate different foods and water sources without significant harmful consequences on human health3. However, the organism becomes opportunistic when it enters into some sensitive parts of the human body (such as urinary tract, blood, meninges, among others), especially in immunocompromised individuals or after involvement in surgery where it multiplies extensively and causes numerous illnesses3,4. Although most E. coli strains may be normal flora of the gut, other strains nonetheless may be major pathogens with an improved tendency to cause diseases. This may be as a result of the acquisition of virulent determinants. E. coli bacteria that are pathogenic can be grouped according to variable criteria which include virulence factors, pathogenicity mechanisms, clinical signs, and serotype5. The virulence factors that enhanced the pathogenicity of E. coli consist of toxins, invasins, adhesins, capsular and effacement factors6,7. Disease-causing strains of E. coli can be grouped into those that cause diseases within and outside the intestine8,9.

The pathogenic E. coli that causes disease within the intestine is called diarrheagenic E. coli (DEC) or intestinal pathogenic E. coli (IPEC)4 and is responsible for gastroenteritis8. The DEC pathotypes are categorized based on their virulence factors and phenotypic traits, and each pathotype has unique host preferences, prevalence, route of transmission, as well as disease burden4. As such, DEC is categorized as enterotoxigenic E. coli (ETEC), enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAEC), and diffusely adherent E. coli (DAEC)7. The most considerable and significant virulent determinants for the detection of STEC/EHEC are the intimin protein (eaeA), and shiga toxins (stx1 and stx2), while for EPEC are intimin protein (eaeA) and bundle-forming pilus (bfp), among others2.

The purpose of the present study was to identify the occurrence and antibiotic resistance profile of diarrheagenic E. coli from sources of water in the Adamawa North senatorial zone, Nigeria. This is particularly important because the study area lacks potable pipe-borne water supply, and the populace relies solely on alternative water sources with questionable microbiological quality10. Of utmost concern is the fact that data on the occurrence of DEC pathotypes in these water sources are lacking.


Study area

The study area was Adamawa North senatorial zone commonly known as the Mubi region (Figure 1). Mubi region comprises five Local Government Areas (LGAs): Madagali, Michika, Mubi South, Mubi North, and Maiha with a land area of 4494 km2 and a population of 682026 (NPC, 2010). The area has a tropical wet and dry climate. Dry season lasts for a minimum of six months (November–March), while the wet season spans between May and October. The mean annual rainfall ranges 700–1050 mm11.

Sampling plan

From each Local Government Area, 2 wards were chosen for hand-dug well (HDW) water sample collection. From each ward, water from four HDWs was chosen at random and sampled in duplicate for the period of sampling. A river/stream was also selected from each Local Government Area for sampling. For each river/stream, two samples were collected at random (upstream and downstream) in quadruples for the period of sampling.

Period of sampling

Water samples were collected aseptically from upstream and downstream of 4 rivers, and 32 hand-dug wells (HDWs) between June 2019 and April 2020.

Water sampling

A total of 256 water samples (comprising 128 each from HDW and river water sources) were taken from 4 local government areas of Adamawa North senatorial zone. Hand-dug well water samples were taken from 8 locations, 2 from each LGA as follows: Lokuwa and Kolere (Mubi North), Wuropatuji and Nassarawo (Mubi South), Michika and Bazza (Michika LGA), and Maiha and Pakka (Maiha LGA). River water samples were taken from 4 rivers, one from each LGA as follows: river Yadzaram (Mubi North and South), river Dilchim (Michika LGA) and river Mayonguli (Maiha LGA).

Isolation of bacteria

Bacteria were isolated by membrane filtration technique using, a sterile 47 mm, 0.45 μm mixed cellulose ester (MCE) membrane filter (Merck, Bangalore). At the end of the filtration, sterile forceps were used to pick the filter onto the surface of MacConkey agar (MCA) and replicated on Eosin methylene blue (EMB) agar. Recovered E. coli isolates were further streaked on the surface of sorbitol MacConkey agar for the presumptive detection of some strains of pathogenic E. coli that can ferment sorbitol. The plates were incubated at 35–37oC for 18–24 h. Discrete bacterial colonies were recultured and stored in nutrient agar slant for identification and further use.

Identification of isolates

After the Gram stain, each discrete bacterial colony was subjected to other biochemical tests such as Simmon’s citrate test, reaction on triple sugar iron (TSI) agar, and oxidase test before they were identified with Microgen GN A kit and 16 SrRNA.

Identification of bacterial isolates using Microgen Gram negative-A (GN-A) ID kits

After Gram-staining, each bacterial isolate was identified on the Microgen A kit (Gold Standard Diagnostic, Hungary)11. Each of these test kits is a plastic strip containing 12 microwells with dehydrated constituents that could identify 12 biochemical characteristics, namely lysine, ornithine, hydrogen sulfide production, glucose, mannitol, xylose, indole, urease production, Voges Proskauer, citrate utilization, tryptophan deaminase (TDA), and orthonitrophenol-β-galactoside (ONPG). The outcomes of the 12 (GN-A kit) microwell test strips after the addition of test isolate and 24-h incubation were converted to a 4-digit octal code that was used to validate the identity of the tested bacterial isolate using Microgen ID computer software version

Molecular identification of bacterial isolates

After genomic DNA extraction, the 16S rRNA gene was amplified and purified. The amplified fragments were cleaned using ethanol, and their reliability was proven on 1% agarose gel. The sequencing of the amplified fragments was achieved on a Genetic Analyzer 3130 × l sequencer (Applied Biosystems) at Inqaba Diagnostic, South Africa. The isolates’ identity was confirmed by subjecting the sequences to analysis with the Basic Local Alignment Search Tool (BLAST)11 (available at

Specific primers for detection of diarrheagenic E. coli pathotype

Specific primer sets were used to detect stx1, stx2, and eaeA coding genes on the presumptive pathogenic E. coli isolates in separate PCR reactions as shown in Table 1. The PCR profile settings were: preliminary denaturation for 5 min at 94°C, which was attended with 30 rounds of denaturation for 30 s at 94°C, hardening for 60 s at 50°C and lengthening for 30 s at 72°C with a concluding termination for 10 min at 72°C.

Table 1

Primers used for the detection of STEC/EHEC and EPEC

Targeted genePrimer sequenceAmplicon size (bp)Reference
stx1F: (ACACTGGATGATCTCAGTGG) R: (CTGAATCCCCCTCCATTATG)614(Tahamta and Namavari12, 2014)

A volume of 5 μL of each PCR product was electrophoresed in 2% agarose gel containing 5 μL of 10 mg/mL ethidium bromide at 100V for 45 min. The molecular marker used was A 1 kb plus DNA marker. DNA amplifications were examined under an ultraviolet (UV) transilluminator and results were documented12,13.

Antimicrobial susceptibility testing of DEC isolates

The disc diffusion method, also known as the Kirby-Bauer method was employed for the antimicrobial susceptibility test as recommended by CLSI14. Each DEC isolate was used to test for its susceptibility to the following antimicrobial agents; pefloxacin (10 μg), gentamicin (10 μg), ofloxacin (10 μg), imipenem (10 μg), ceftriaxone (30 μg), ceftazidime (30 μg), streptomycin (30 μg), ciprofloxacin (10 μg), amoxillinclavulanic acid (30 μg), ampicillin (30 μg), nalidixic acid (30 μg), and cotrimoxazole (30 μg). The inhibition zone diameter was measured in millimeters and was interpreted based on the diameter of interpretative standard breakpoints14. MDR phenotype was determined when an isolate was resistant to at least one antibiotic in three classes of antibiotics11.

Statistical analysis

All the data were analyzed by descriptive statistics using simple percentages. All data were analyzed using the Statistical Package for Social Sciences (SPSS) statistics for windows (Chicago) version 17.


The results in Table 2 show the total number of E. coli isolated from the water sources. Of the 256 water samples analyzed, 63 (24.6%) E. coli were isolated. Of these, 44 (69.8%) and 19 (30.2%) were isolated from river and well water sources, respectively. The sequences of some of the E. coli isolates documented in this study have been assigned accession numbers (Table 3) and have equally been added in the NCBI GenBank.

Table 2

Total Escherichia coli isolated from the water sources

Water sourceNumber of samples collectedNumber of E. coli isolates
Well128 (50.0)19 (30.2)
River128 (50.0)44 (69.8)
Total256 (100)63 (24.6)
Table 3

Identification of the selected sequenced Escherichia coli using BLAST

Sample IDScientific nameMaximum scoreTotal scoreQuery cover %E-valuePercentage identity %Accession number
97E. coli188018801000100MZ437057
107E. coli18681305100098.95MZ437064
109E. coli18681305100098.95MZ437065
131E. coli188018801000100MZ437078
132E. coli188018801000100MZ437079

Of the 63 E. coli isolates recovered from the water sources, 27 (42.9%) were non-sorbitol fermenting E. coli. Of these, shiga toxins (stx1 and stx2) genes (typical characteristics of STEC/EHEC) were detected in 24 (88.9%) isolates, with stx1 and stx2 genes detected in 13 (48.2%) and 22 (81.5%) E. coli isolates, respectively. The eaeA gene along with other virulent genes was documented in 21 (77.8%) E. coli isolates. However, eaeA gene alone (a typical characteristic of an EPEC) was detected only in two E. coli isolates. Whereas 11 (40.7%) of the STEC/EHEC harbored a combination of stx1 and stx2 genes, the combination of stx1, stx2, and eaeA genes was documented in 10 (37.0%) STEC/EHEC isolates (Table 4).

Table 4

Prevalence of DEC virulence genes

PathotypeVirulence markers

[i] NA: not applicable.

The resistant pattern of the DEC pathotypes (STEC and EPEC) is shown in Table 5. Resistance to cotrimoxazole, ampicillin, nalidixic acid, amoxicillin-clavulanic acid, and ceftriaxone was 100% for both STEC and EPEC. However, 95.5%, 87.5%, and 83.3%, of STEC were resistant to ceftazidime, ciprofloxacin, and pefloxacin, respectively. Whereas 13 (54.2%) of STEC were resistant to imipenem, all the EPEC isolates were susceptible to imipenem.

Table 5

Antimicrobial susceptibility pattern of DEC pathotype

AntibioticsSTEC (N=24)EPEC (N=2)
Susceptible n (%)Resistant n (%)Susceptible n (%)Resistant n (%)
Cotrimoxazole024 (100)02 (100)
Streptomycin5 (20.8)19 (79.2)02 (100)
Ampicillin024 (100)02 (100)
Pefloxacin4 (16.7)20 (83.3)1 (50.0)1 (50.0)
Gentamicin5 (20.8)19 (79.2)02 (100)
Ofloxacin9 (37.5)15 (62.5)2 (100)0
Nalidixic acid024 (100)02 (100)
Augmentin024 (100)02 (100)
Ciprofloxacin3 (12.5)21 (87.5)2 (100)0
Ceftriaxone024 (100)02 (100)
Ceftazidime1 (4.2)23 (95.8)02 (100)
Imipenem11 (45.8)13 (54.2)2 (100)0

The resistant profile showed that the DEC pathotypes were resistant to from 7 to 12 antibiotics with 8 (30.8%) of the 26 DEC resistant to 12 antibiotics (Table 6).

Table 6

Antibiotic resistance profile of DEC

SNNumber of antimicrobialsResistance profileNumbers observedMDR status

* EPEC isolates. sxt: cotrimoxazole. s: streptomycin. pn: ampicillin. pef: pefloxacin. cn: gentamicin. ofx: ofloxacin. na: nalidixic acid. aug: augmentin. cpx: ciprofloxacin. cro: ceftriaxone. caz: ceftazidime. ipm: imipenem. MDR: multidrug resistance.


The availability of potable drinking water is paramount to life forms, especially humans. That is why water meant for domestic purposes is expected to be devoid of microbial agents like E. coli which serves as an indicator for the presence of potential pathogens15. The results of this study, on the other hand, have recorded the existence of E. coli in water sources meant for domestic purposes. The observed presence and spread of E. coli in these water sources can be linked to a variety of factors including, but not restricted, to poor sanitation practices in the vicinity of these water sources16.

The detection of 27 (42.9%) E. coli bacteria that were sorbitol-non-fermenting suggests that they were pathogenic. The frequency, however, was lower than 75%, 68.0%, and 67.5% of pathogenic E. coli documented in water sources from Ghana15, Côte d’Ivoire17, and South Africa18, respectively.

In causing diseases, strains of E. coli that are pathogenic usually employ a series of multifaceted machinery embracing a number of virulence determinants which eventually leads to the destruction of the target host cells. As such, the expression of one or more virulent determinants in appropriate combinations determines the pathogenic capability of a particular E. coli isolate19.

The pathogenic E. coli bacteria encountered in this study belong to two diarrheagenic pathotypes; shiga toxin-producing E. coli (STEC) also known as enterohemorrhagic E. coli (EHEC) and enteropathogenic E. coli (EPEC). While the STEC/EHEC constitute the majority of the DEC pathotype, EPEC constitutes only 7.4% of the DEC pathotypes encountered in this study.

The STEC/EHEC cause several types of disease indications in humans, which include mild diarrhea to severe disease forms like hemolytic uremic syndrome (HUS), and hemorrhagic colitis (HC) mediated primarily by shiga toxins (stx1 and stx2 genes) and intimin (eaeA gene). Though each gene of the shiga toxins (stx1 or stx2) possesses the ability to cause acute diarrhea as earlier reported7, the detection of both stx1 and stx2 genes in some strains of STEC/EHEC isolates, as shown in this study, is of grave consequence. This is because E. coli carrying a mixture of shiga toxin genes was reported to cause more complicated diarrhea in humans7,12. Also, the high prevalence of the stx2 genotype in STEC/EHEC strains of this study is of public health significance. This is because stx2 is reported to cause more severe clinical outcomes than stx120. More so, studies have shown that EHEC that causes HUS expresses stx2 in more cases than the stx1 genotype21,22.

More so, in this study, the combination of stx2 and eaeA genes was more pronounced than stx1 and eaeA genes among the STEC/EHEC strains. This observation is in contrast to a similar study in Iran2 which reported the preponderance of stx1 and eaeA over stx2 and eaeA genes. Studies have shown that STEC/EHEC strains with the eaeA gene are more virulent when compared with eaeA-negative STEC/EHEC strains2. The intimin gene was reported to be accountable for the indepth adhesion of the STEC/EHEC and EPEC pathotypes to the epithelial cells of the intestinal mucosa; this subsequently gives rise to attaching and effacing lesion at the point of attachment19,23.

The two (7.4%) E. coli isolates that lacked other virulent genes, which are typical of the STEC/EHEC pathotype but harbored only the eaeA gene, are known as EPEC. This observation was similar to a previous study in South-Western Nigeria where 4.0% of E. coli isolates from river sources were reported to harbor only eaeA7. The detection of EPEC strain with only the eaeA gene is of public health concern. This is because the EPEC pathotype that carries only the eaeA gene was reported mainly to cause an outbreak of gastroenteritis globally18,24. In general, EPEC is reported to cause infantile diarrhea more often in underdeveloped countries like Nigeria25.

Quite similar to the findings of this study, the detection of DEC pathotype isolates with a characteristic that is typical of STEC/EHEC and EPEC was previously reported in Nigeria7,23 and beyond26,27. The detection of these DEC pathotypes in surface and groundwater sources meant for domestic purposes constitutes a risk of outbreaks with possibly grave consequences if left unchecked.

In this study, a high level of DEC resistant to ampicillin (100%), cotrimoxazole (100%), nalidixic acid (100%), augmentin (100%), and ceftriaxone (100%) was documented. High resistance of DEC to ampicillin and cotrimoxazole in this study was consistent with the reports of various studies in South Africa8, Peru28, and Iran29.

The high rate of DEC resistant to cotrimoxazole and ampicillin in this study could be because these antimicrobials are the most frequently used antibiotics for therapy against diarrheoa30,31. The common use of these classes of antimicrobials could be because they are relatively inexpensive, have ease of accessibility, and initially are very effective with a broad spectrum of activity against a wide range of infections, especially against diarrhoea32.

The MDR phenotype exhibited by DEC in this study was similar to those of previous studies in Southwestern Nigeria7, and South Africa31,33. Resistance to 7–12 antibiotics by DEC in this study implies excessive, indiscriminate, and inappropriate use of these antimicrobials in the study area. It could also be due to the acquisition of resistance genes through horizontal gene transfer31,34. So, the use of surface or groundwater sources for domestic activities and/or irrigation may further increase the chances for the spread of MDR DEC in the study area, and in economy-restrained countries29.

In this study also, resistance to imipenem was relatively lower than to other antimicrobials. This corroborates studies that documented low resistance to carbapenem antibiotics by DEC isolates in Nigeria35,36 and in Asian countries37. The high susceptibility of the DEC isolates to imipenem in this study might be due to the no or low prescription and usage of the antimicrobial, especially in treating diarrhea-related illness. The results obtained from this study may be valuable in building strategies that will reduce the risk associated with the spread of DEC isolates to the public through water sources.


The limitation of this study was its inability to differentiate the two EPEC isolates into typical and atypical EPEC due to other requirements which were not captured in the course of this study.

Also, the inability to screen for other virulent genes characteristics of both EHEC and EPEC was another limitation.


The occurrence of STEC/EHEC in water sources of the study area is remarkable, and it highlights the fact that these sources may serve as significant avenues through which microbial agents of diarrhea are disseminated. Because there was a paucity of data (or there were no data) that reported the presence of STEC/EHEC and EPEC from water sources in the study area, the findings of this study contribute to a better knowledge of the occurrence of pathogenic E. coli in the major water sources of the study area and may serve as reference point data for future use and epidemiological surveillance. Also, the documented and significant high MDR phenotypes of the DEC isolates in this study call for concern, and it underscores the necessity for better and robust practical measures to be put in place that will help in curbing the menace of antimicrobial resistance.