Genomic Characterization of Selected Escherichia coli Strains from Catfish (Clarias gariepinus) in Nigeria

Abstract

According to a report by the World Health Organization (WHO), each year, over 550 million individuals worldwide suffer from and 230,000 die from diarrheal illnesses, which accounts for more than half of the global foodborne disease burden. Among them, children face a heightened vulnerability, with approximately 220 million falling ill and 96,000 succumbing to these diseases annually. This work aimed to study the genomic characterization of selected E. coli strains from catfish (Clarias (C.) gariepinus) caught from the Onitsha North axis of the River Niger in Anambra state, Nigeria. A total of 50 fish were randomly purchased from different fishermen over a period of four months. Samples that comprised six different organs (skin, flesh, gills, gonads, guts, and liver) were screened for E. coli strains using cultural and biochemical methods. Multilocus sequence typing (MLST) and core genome (cg)MLST were performed using Ridom SeqSphere+ software. The aerobic plate count (APC) and coliform count ranged from 0.5 × 10⁴ to 3.7 × 10⁴ cfu/g and 0 to 3.0 × 10⁴ cfu/g, respectively. Whole-genome sequencing (WGS) confirmed the presence of E. coli and Klebsiella quasipneumoniae isolates in our samples. We could identify only two serotypes (O102:H7 and O40:H4) of E. coli. Antimicrobial resistance genes (ARGs) and point mutations that conferred antibiotic resistance were extracted from the genome assemblies. Good hygiene is recommended to avoid the cross-contamination of raw C. gariepinus with ready-to-eat food.

1. Introduction

According to the World Health Organization (WHO), Africa and Southeast Asia constitute the greatest burden of foodborne diseases, with the highest incidence and mortality rates among children under five years of age. Every year, over 125,000 children die from foodborne illnesses in these regions [1]. Escherichia coli (E. coli) is a Gram-negative bacillus commonly found in the intestinal flora of humans and animals and is often benign to the host. Some E. coli strains can, however, act as intestinal or extraintestinal pathogens, leading to a spectrum of illnesses that range from mild gastroenteritis to severe conditions, like renal failure and septic shock [2]. For instance, O157:H7 is one of the major STEC (Shigatoxigenic E. coli) strains and it is associated with severe foodborne illnesses transmitted through contaminated food or water. This serotype has the potential to cause life-threatening complications, such as hemolytic uremic syndrome (HUS) [2]. Similarly, O104:H4, which is a hybrid STEC–enteroaggregative (EAEC) strain that gained prominence during the 2011 outbreak in Europe, caused episodes of bloody diarrhea and HUS [3]. In addition to these two E. coli pathotypes, enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), and enteroinvasive E. coli (EIEC) can also cause diarrheal illnesses. Extraintestinal pathogenic E. coli (ExPEC) do not cause intestinal disease but are associated with urinary tract infections, bloodstream infections, and meningitis [2,3].

Humans contract E. coli most frequently via eating contaminated foods, like seafood, raw or undercooked ground meat, raw milk, contaminated raw vegetables, and sprouts. E. coli infection can also result from the fecal contamination of water and other foods, as well as cross-contamination while food is being prepared (with beef and other animal-product-infected surfaces and contaminated kitchen equipment) [4]. Both daycare facilities and nursing homes have documented cases of person-to-person transmission [5].

The virulence of E. coli enables this bacterium to evade the host immune response. Coupled with the acquisition of resistance to commonly used antibiotics, infections caused by this bacterium become notably difficult to treat [2]. Through the acquisition of virulence genes, E. coli increases its pathogenicity, and hence, the severity of infection, with the great possibility of therapy failure [6]. In addition, the misuse and overuse of antibiotics for human and therapeutic purposes and as growth promotors in livestock can lead to the emergence of resistance to many antimicrobial classes, such as penicillins, cephalosporines, tetracyclines, sulfonamides, macrolides, and polymyxins [7].

In Nigeria, pathogenic and resistant E. coli strains were identified in various foods intended for human consumption, including raw milk [8], edible snails [9], seafood [10], and fish [11]. In the latter, EPEC and EIEC, as well as multidrug-resistant E. coli strains carrying bla_TEM and tetA genes, were detected [10].

The River Niger is the principal river of west Africa, with a length of 4200 km. Numerous laws and regulations have been adopted by the Niger Basin Authority (NBA) and its member countries against the pollution of the River Niger from runoff, waste disposal, and sewage discharges. Yet, legal instruments for control are neither defined nor implemented [12]. The River Niger therefore continues to receive large quantities of various wastes, including human feces and agricultural and industrial wastes. These could be sources of not just pathogenic E coli but also antibiotic-resistant strains. Since fish ingest a significant number of microorganisms from their diet and water sediment, it is likely that they are contaminated with pathogenic and resistant E. coli strains [13]. The aim of our study was to elucidate whether catfish from the River Niger could be a source of pathogenic and antibiotic-resistant E coli strains.

2. Materials and Methods

2.1. Sample Collection and Processing

The catfish (C. gariepinus) were randomly purchased from different fishermen over a period of four months from February to May 2023. The fish were caught from the River Niger at Onitsha, Anambra State, in Southeastern Nigeria. The fish were quickly transported to the Biotech Research Centre Nnamdi Azikiwe University, Awka, using sterile polythene bags. The fish species were confirmed with the help of a staff member of the Department of Zoology, Nnamdi Azikiwe University, Awka, at the Biotech Research Centre. The fish were rinsed with distilled water and the work surface was cleaned and swabbed using 70% ethanol. The fish were dissected, the internal organs exposed, and the desired parts were identified by the zoology staff. The samples that comprised six different organs (skin, flesh, gills, gut, liver, and gonad) were further analyzed.

Skin samples were collected first using a sterile lancet. The abdominal part was then swabbed with 70% ethanol and dissected before the internal organs were harvested using a sterile lancet for each part. The gills, gut, skin, liver, flesh, and gonads were aseptically excised, and one gram from each was weighed from each fish. The organs were labelled accordingly as Sk (skin), Fl (flesh), Gi (gill), Go (gonad), Gu (gut), and Li (liver).

2.2. Microbiological Analysis

For the bacterial counts, a slight modification of the method described by Adesoji et al. [14] was used. Each of the samples were aseptically transferred into test tubes containing 10 mL of sterile peptone water (Titan Biotech, New Delhi, India) and subsequently homogenized by vigorous shaking. Afterward, each sample was again serially ten-fold diluted using sterile peptone water. Then, 0.1 mL of the appropriate dilutions were used to determine the bacterial counts.

For the aerobic counts, 0.1 mL from both the previous 10⁻³ and the 10⁻⁴ dilutions were plated in duplicate on nutrient agar (Titan Biotech, New Delhi, India). The plates were then incubated at 37 °C for two days, and those containing 30–300 colonies were counted. Similarly, for the coliform counts, a 0.1 mL aliquot of the appropriate dilutions were plated in duplicates on MacConkey agar (Titan Biotech, New Delhi, India). The plates were then incubated at 37 °C for two days, and those containing 30–300 colonies were counted.

2.3. Isolation of E. coli

The method described by Yohans et al. with slight modifications was used. From the homogenized samples from the fish organs, 1 mL of each was cultured on Eosin methylene blue (EMB) agar (Titan Biotech, New Delhi, India) plates and incubated aerobically at 37 °C for 24 h. Distinct colonies that displayed green metallic sheen with dark centers, which is indicative of E. coli, were subcultured on MacConkey agar plates [15]. These plates were then incubated aerobically at 37 °C for 24 h. Distinct pink colonies were picked and stored on nutrient agar slants. Biochemical tests were done using the methods described by Bhutia et al. [16], along with indole [17], methyl red, Voges–Proskauer [18], citrate [19] (IMViC) [16,17,18,19], and glucose fermentation tests [20]. Simmon’s citrate agar (Titan Biotech, New Delhi, India) was used.

2.4. Whole-Genome Sequencing

Selected presumptive E. coli isolates from the MacConkey agar subculture were sent to the Austrian Agency for Health and Food Safety in Vienna for genomic identification and characterization. E. coli isolates were cultured on Columbia blood agar plates (bioMérieux, Marcy-l’Étoile, France) upon arrival to the laboratory at 37 °C for 24 h. Genomic DNA was extracted from the overnight cultures using MagAttract HMW DNA Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The DNA concentration was measured using a Qubit 3.0 fluorometer (Thermo Fisher Scientific, Waltham, MA, USA) with the dsDNA HS assay kit (Thermo Fisher Scientific). Library preparation was performed with a Nextera XT kit (Illumina Inc., San Diego, CA, USA). Paired-end sequencing (2 × 150 bp) was conducted on an Illumina Nextseq 2000 device.

The raw reads were checked for quality using FastQC v0.11.9 [21]. Adapter sequences were trimmed with Trimmomatic v0.36 [22] using the default parameters. Reads were assembled with SPAdesv3.11.1 [23], and the contigs were filtered for a minimum coverage of 5x and a minimum length of 200 bp. Multilocus sequence typing (MLST) and core genome (cg)MLST were performed using RidomSeqSphere+ software v.9.0.8 (Ridom, Münster, Germany), which uses Enterobase schemes (https://enterobase.warwick.ac.uk/species/index/ecoli, accessed on 20 January 2024). A minimum spanning tree (MST) was generated to visualize the clusters and genetic relatedness between the isolates using the official cluster threshold of 10 alleles. Antimicrobial resistance genes (ARGs) and point mutations that conferred antibiotic resistance were extracted from the genome assemblies with AMRFinder Plus [24]. Plasmid replicons were extracted with MOB Suite v3.1.4 [25] and virulence genes were extracted with VFDB [26].

3. Results

3.1. Microbiological Analysis

The aerobic plate count results indicate similar bacterial presences in all sampled parts and organs of the fish, with the skin exhibiting the highest values (3.4 × 10⁴ CFU/g) and gonads showing the lowest values (1.62 × 10⁴ CFU/g). Only the skin, flesh, gill, and gut samples carried coliforms in similar amounts (

Table 1. Aggregated values of the aerobic and coliform counts of bacterial isolates from specific sampled organs.

Sample	Aerobic Plate Count (×104 CFU/g ± SD)	Coliform Count (×104 CFU/g ± SD)
Sk	3.4 ± 3.0	0.7 ± 2.0
Fl	5.0 ± 2.0	3.4 ± 4.0
Gi	2.68 ± 3.0	1.88 ± 10
Go	1.62 ± 6.0	0
Gu	3.3 ± 3.0	2.02 ± 9.0
Li	0.84 ± 3.0	0

SD—standard deviation. Sk—skin, FL—flesh, Gi—gill, Go—gonad, Gu—gut, and Li—liver.

).

3.2. Whole-Genome Sequencing (WGS)

The WGS confirmed the presence of E. coli isolates in our samples. One of the isolates was a mixed culture with Klebsiella quasipneumoniae and the separated strains were resequenced. Information on the STs, resistance genes, plasmids, and virulence genes of the seven isolates are shown in Table S1. Briefly, each of the six obtained E. coli strains belonged to a different ST, one of which was unknown. The cgMLST analysis showed a high diversity between the isolates with more than 1422 alleles of difference (Figure 1). We could identify only two serotypes (O102:H7 and O40:H4), while for the other four isolates, only the H-type antigen could be detected. Table S1 shows the molecular identities of the suspected E. coli isolates.

3.3. Virulence Associated Genes

The ARGs that confer resistance to beta-lactams (bla_TEM-1, n = 1), aminoglycosides (aph(3”)-Ib, aadA5), quinolones (qnrS1, n = 1), sulfonamides (sul2, n = 2), tetracyclines (tetA, n = 2), and trimethoprim (drfA1, n = 1; drfA17, n = 1) were found. Point mutations that confer resistance to colistin (pmrB_Y358N), fosfomycin (glbT_E448K), and quinoles (gyrA_D87N, gyrA_83L, parC_80I, parE_S458A, n = 1) were found. The results are as shown in Table S1.

Between one and three plasmids were found in five of the isolates. All isolates except one of the three plasmids detected in isolate 511673-23 were non-mobilizable and carried no resistance genes. The mobilizable plasmid from this isolate was an IncH1/IncY plasmid and harbored bla_TEM-1, qnrS1, aadA5, sul2, tetA, and dfrA17 resistance genes. Virulence genes were detected, none of which were associated with diarrheagenic or extraintestinal E. coli pathotypes. As for the K. quasipneumoniae isolate, it belonged to ST1031 and carried eight plasmids, three of which harbored ARGs. The first plasmid carried mph(A), sul1, and drfA2. The second carried tet(A) only and the third carried catA2, aph(3”)-Ib, and aph(6)-Id. The latter was classified as mobilizable.

3.4. Data Availability

Sequencing data from this study were deposited in the Sequence Read Archive (SRA) under Bioproject accession number PRJNA1088746.

4. Discussion

The findings from this work show that freshly caught C. gariepinus (Catfish) from the Onitsha axis of the River Niger were contaminated with various microorganisms, including E. coli strains. This is suggestive of fecal contamination [27] of the River Niger from municipal and other wastes. There were significant differences in the microbial counts among the fish organs/parts examined. The consumption of such fish without adequate cooking or processing may lead to foodborne diseases. The samples contained a good number of microorganisms, as indicated by the results of the aerobic plate and coliform counts carried out (Table 1). According to Sayed, 2021, fish are a major source of bacterial pathogens that can be transferred to and from humans to other animals [27]. The microbial community found in freshly caught fish mainly mirrors the microbial condition of the waters where they are caught [28]. Fish, like all living beings, live alongside various pathogenic and non-pathogenic microorganisms, with each equipped with intricate defense mechanisms crucial for their survival [29].

The number of aerobic plate counts detected here indicate a lower level of contamination of C. gariepinus (0.5–3.7 × 10⁴ CFU/g) when compared with other works; Abdelhamid et al., 2013, reported that the minimum started at 10⁵ CFU/g for all fish parts sampled, including the liver, gills, and intestine. Our CFUs for aerobic counts may have been lower than in other studies due to differences in the sampling seasons [30]. For instance, summer typically exhibits higher microbiological loads in fish. Additionally, previous studies often sampled from lakes or fish farms, where water may be somewhat more stagnant compared with rivers [30,31].

As for coliforms, which were found in all organs except the gonads and liver, they were found at levels similar to the aerobic counts (0–3.0 × 10⁴ CFU/g) (Table 1).

The WHO standard is 1.0 × 10³ CFU/100 mL [32], while the Centre for Food Safety (CFS) guidelines state that <20 CFU/g is satisfactory, 20–10² is intermediate or borderline, and >10² is unacceptable [15,33]. Therefore, the results of the aerobic plate count and the coliform count showed counts higher than the WHO and CFS set standards [15,32]. A similar work by Awe and Adejo, 2017, on selected fresh fish from the River Niger, Lokoja, had results higher than the WHO and CFS standards [34].

The findings from this work show that freshly caught C. gariepinus (Catfish) from the Onitsha axis of the River Niger were contaminated with E. coli strains. Similar research by Yohans et al., 2022, reported the presence of E. coli in fish samples [15]. The considerable commercial and industrial business activities and the discharge of waste in and around the Onitsha main market may have contributed to these figures. Given the significant discharge of human excrement into the section of the River Niger where the fish in our study were caught from, the presence of E. coli in catfish should not be uncommon, as a previous study on African catfish by Yohans et al. confirmed [15].

Sayed reported finding Escherichia coli and E. tarda more frequently in the intestine than in the spleen and liver of the fish. That study attributed this fact to the existence of both bacteria as part of the normal intestinal microbiota of fish [27]. Our study showed that the skin, gills, and gut displayed higher bacterial counts when compared with the other parts (flesh, gonad, and liver) sampled. These data, however, are not in agreement with the report of a previous work stating that a small number of microorganisms in general can be found on fish skin [28]. This difference can be linked to elements like the different pollutants that are dumped in the River Niger from the numerous commercial activities that occur nearby. In addition, careless handling by fishers, including the usage of soiled containers and filthy practices, may have played a role. When immunological resistance is weakened, the contamination of fish muscles is also possible. The number of bacteria present in fish and the surrounding water may be influenced by a few variables, including human activity, contaminated water sources, sloppy handling, and shipping practices [35,36].

Six of the isolates were identified as E. coli and one Klebsiella quasipneumoniae (Table S1). None of the isolates were found to be pathogenic E. coli despite carrying virulence genes. Two isolates possessed the O (lipopolysaccharide) antigen, while all E. coli isolates had the H (flagellar) antigen. Each E. coli pathotype has its characteristic pathogenicity mechanisms and a specific profile of virulence factors encoded by specific gene clusters [36,37]. None of our isolates belong to the big seven STEC for instance, but also not to other pathotypes.

The sequence types (STs) that Warwick obtained were STs 48, 196, 226, 192, and 6186, which included a new one. Two types exhibited clonal complexes (CC Warwick) ST10 Cplx and ST226 Cplx. These were found in E. coli isolates from gill and gut samples, respectively; however, ST1031 was found in the K. quasipneumoniae isolate and had blaOKP-B-1 and tet(A) AGR genes, among others (Table S1). In this study, ST48 belonging to CC10 was found in an isolate from the gill and had the blaTEM-1 AGR gene. It has been associated with extraintestinal pathogenic E. coli (ExPEC) and has been described as one of the clones that are considered high-risk zoonotic E. coli clones that exhibit both pathogenicity and multi-drug resistance [38]. The ST10 clone was detected in human and chicken in a study done in Nigeria [39]. A similar study in Egypt by Ramadan et al., 2020, reported that both ST10 and ST48 belonged to CC10, which is one of the major CCs associated with diarrheagenic E. coli infections in humans worldwide [40]. It was frequently detected in multiple studies and regions in studies carried out on swine in Northern Europe, where it was described as the main E. coli clonal complex associated with porcine ETEC, and was recently identified as being primarily responsible for the spread of mcr-4 [41].

The sequence type ST226 belonging to clonal complex (CC) 226 was found in an isolate from gut sample and it had the glpT_E448K AGR gene against Fosfomycin. It was found in human and beef in an earlier work in Egypt [40] and has been linked to a CTX-M-producing MDR isolate from Arapaima gigas in the north of the Brazilian Amazon [42] and in human and beef isolates in Egypt [40].

Our study was not without limitations. These include the need for further research that involve a larger number of isolates and comprehensive susceptibility testing data. These additional studies are essential to fully understand the burden of E. coli in catfish from the River Niger and the potential risks it poses to human health. Although our study established the presence of E. coli in C. gariepinus from the Niger River, no pathogenic strains were recovered. The virulence genes identified, such as the esp genes, were not associated with any specific pathotype, indicating that a combination of virulence genes is necessary to classify a pathotype. The detection of E. coli in raw or under-processed fish suggests a potential risk of colonization and infection for consumers. Additionally, the presence of antimicrobial resistance genes (ARGs) in most isolates highlights the risk of these genes being transmitted to the gut microbiota, potentially complicating the therapeutic management of infections. To mitigate these risks, good hygiene practices should be implemented to avoid the cross-contamination of raw C. gariepinus with ready-to-eat food, and adequate cooking and processing are recommended to prevent foodborne infections by E. coli.