Abstract

Background. Escherichia coli bacteria are Gram-negative, non-spore-forming aerobes or facultative anaerobic rods. Some strains are pathogenic in men while others are commensals in the gut. The pathogenic strains cause a wide array of diseases by virtue of virulence factors. The commensal strains are generally categorized into phylogenetic groups A and B1. The aim of this study was to determine the association between phylogeny of E. coli isolates and virulence and sociodemographic characteristics of the study subjects. Method. This study was a cross-sectional study carried out from July 2018 to June 2019. E. coli isolates obtained from different clinical specimens were subjected to polymerase chain reaction to determine their phylogenetic groupings and virulence. Results. The majority of the isolates belonged to phylogroup A 101 (74.8%), and the predominant virulent gene was fimA (88.9%). There was no significant correlation between phylogenicity and virulence, except for chuA which was found in all isolates that belonged to phylogroups clade I and D. None of the 101 isolates that belonged to group A had the chuA virulence gene. There was a significant association between patient age category and phylogenetic groups B1 and D. Conclusions. This study assessed the relationship between the phylogenetic distribution and the virulence profile of clinical isolates of E. coli. The virulence of isolates belonging to phylogroup A, which are generally considered as commensals, is alarming. Measures must therefore be put in place to control the spread of these virulent E. coli.

1. Introduction

Escherichia coli bacteria are Gram-negative facultative anaerobic rods. They are non-spore-forming and possess peritrichous flagella for motility. These bacteria are potential aetiologic agents of various disease states such as cholecystitis, pneumonia, pyelonephritis, cystitis, urinary tract infection, newborn meningitis, septicaemia, central nervous, and respiratory infections [1]. However, it is worthy of note that many Escherichia coli strains are commensal resident in the gut. The barrier between commensalism and virulence of E. coli largely depends on a balance between the host immune status and the ability of the bacteria to elicit their virulence factors [2].

Extraintestinal pathogenic E. coli is one of the leading causes of morbidity and mortality globally [3]. The degree of pathogenicity of extraintestinal E. coli isolates depends on the presence of fimbriae adhesins (fimA, sfa/foc, and yfcV) which are virulent factors responsible for adherence. Adherence is a prerequisite for initiating and colonization to specific host cells [4]. Another virulent factors are haemolysins (hylA and hylF), which are responsible for lysing red blood cells and human renal epithelial cells. Siderophores (ferric aerobactin receptor (iutA), Yersiniabactin receptor (fyuA)) also function to sequester iron from the host. Another virulence factor is the cytotoxic necrotizing factor (cnf1) which is implicated in tissue damage and dysfunction of a local immune response [5]. Further, the virulence gene encoding vacuolating toxin (vat) is accountable for delaying neutrophil infiltration of the urinary tract in response to uropathogenic E. coli by cleaving surface glycoproteins from leukocytes that are involved in neutrophil attraction and migration [6]. These virulence factors contribute significantly to the pathogenicity of E. coli strains by colonizing the key anatomical sites, reducing immune response, affecting physiology, and invading host tissues [7].

Before 2013, phylogenetic analysis categorized E. coli into four phylogenetic groups: A, B1, B2, and D. This classification was based on a combination of two genes (chuA and yjaA) and anonymous DNA fragment TSPE.C2 by use of polymerase chain reaction (PCR) [8]. The commensal E. coli strains were associated with phylogenetic groups A and B1 [9, 10]. In contrast, extraintestinal pathogenic E. coli strains that carry more virulence genes belonged to phylogenetic groups B2 and D. Extraintestinal infections are mainly due to E. coli strains belonging to phylogroup B2 and, to a lesser extent, group D [11]. Other techniques used to determine the phylogenetic groups are multilocus enzyme electrophoresis, ribotyping, random amplified polymorphic DNA analysis, fluorescent amplified-fragment length polymorphism analysis, analysis of variance at mononucleotide repeats in intergenic sequences, and multilocus sequence typing [12].

There has been advancement in the knowledge on the multilocus sequence and genome data resulting in a better understanding of E. coli phylogroup classification since 2000. That notwithstanding, the available data at the time showed that 15%–20% of the phylogroups were incorrectly assigned [13]. Further studies have demonstrated some other unclassified strains such as phylogroup E [14], phylogroup F [13], and Escherichia cryptic clade I [8]. This drawback in the classification of the E. coli phylogroup necessitated a revised grouping of E. coli isolates into eight phylogroups (A, B1, B2, C, D, E, F, and Escherichia clade I) [8].

In Ghana, however, data on the virulence potentials and phylogenetic groupings of extraintestinal E. coli isolates are nonexistent. Thus, this study is aimed at detecting the virulence encoding genes of E. coli isolates from various clinical specimens and determining their phylogenetic groups in the Ho Teaching Hospital of Ghana.

2. Materials and Methods

2.1. Study Design

This was a cross-sectional study consisting of 135 E. coli organisms. The isolates were obtained from various clinical samples produced by 135 patients who visited the Ho Teaching Hospital from July 2018 to June 2019. Majority, 98 (72.6%), of these samples were urine from patients with urinary tract infection. The rest were wound swabs, 14 (10.4%); high vaginal swabs (HVS), 10 (7.4%); blood, 5 (3.7%); ear swabs, 5 (3.7%); sputum, 2 (1.5%), and pleural aspirate, 1 (0.7%).

2.2. The Bacterial Isolates

E. coli isolates were cultured from various clinical specimens, including urine, high vaginal swabs, blood on blood agar, and MacConkey agar (Oxoid, UK). Growths suspected to be E. coli were confirmed using the Gram stain reaction, triple sugar fermentation test, citrate test, urease test, indole test, Voges Proskauer, and methyl red test. The identified organisms were inoculated into 80% glycerol-Mueller Hinton broth and stored in a −80°C freezer and later used for other tests. Escherichia coli ATCC 25922 and Klebsiella pneumoniae NCTC 13442 were used as control organisms.

2.3. Molecular Detection of Virulence Factors and Phylogenetic Groups
2.3.1. Revival of Isolates and DNA Extraction

The isolates were retrieved from the freezer, and the surface was aseptically scraped and emulsified in 30 ml Luria Bertani broth (Oxoid, UK) and incubated overnight in a shaking incubator. Genomic deoxyribonucleic acid (DNA) was extracted from the overnight culture using a high-molecular weight phenol-chloroform extraction method [15], except that Tris-EDTA (TE) was used as the elution buffer. The extracted DNA was incubated at 4°C for two days to resuspend the pellet into a translucent viscous gel. The concentration of the extracted DNA was measured using a NanoDrop spectrophotometer (Thermo Scientific), and the viscous DNA was stored under −80°C.

2.3.2. Detection of Virulence Genes and Phylogenetic Groups by Polymerase Chain Reaction (PCR)

Escherichia coli isolates were characterized using primer sequence from previous studies to determine the phylogenetic characteristics (Table 1) and virulence factors (Table 2). The primers were purchased from Integrated DNA Technologies, UK. OneTaq Quick-Load 2x Master Mix with standard buffer purchased from New England Biolabs® was used for the PCR. Lyophilized forward and reverse primers were reconstituted by adding a calculated amount of molecular-grade water using the nmol of the primers to give a concentration of 100 μM. A 1 in 10 working solution was prepared by taking 10 μl of the stock and 90 μl of the molecular-grade water to give 10 μM. A total of 12.5 μl reaction volume was used, comprising 6.25 μl of OneTaq Quick-Load 2x Master Mix with standard buffer, 0.25 μl each of 10 μM forward and reverse primers, 1 μl of template DNA, and 4.75 μl of nuclease-free water.

The initial denaturation and the final elongation for both phylogenetic groups and virulence factors were carried out at 94°C for 30 seconds and 68°C for 5 minutes, respectively. For phylogenetic groups, the PCR conditions were denaturation for 30 seconds at 94°C and annealing for 45 seconds at 55°C (for chuA and arpA), 51°C (for yjaA), 47°C (for TspE4.C2 and ArpAgpE), and 53°C (for TrpAgpC).

PCR conditions for the determination of virulence genes were denaturation for 30 seconds at 94°C and annealing for 45 seconds at 49°C (for ibeA), 50°C (for iutA and yfcV), 52°C (for fimA), 54°C (for hylF and cnf 1), 55°C (for hlyA, neuC, and chuA), 57°C (for sfa/foc), and 58°C (for vat and fyuA). The initial elongation was done at 68°C for 1 minute/kb.

The annealing temperatures for the various set of primers were calculated using the NEB Tm calculator [19]. In-house DNA (stored positive DNA from previous experiments) in the laboratory were used as positive controls whereas the Ambion® Nuclease-free Water (Life Technologies, USA) was used as negative control. Thermocycling was done for 30 cycles.

2.3.3. Loading of Amplicons

Using a micropipette, the resulting PCR products were loaded into agarose wells. The first and second lanes were loaded with 6 μl of the either 50 bp, 100 bp, or 1 kb DNA ladder and 10 μl of in-house-generated positive DNA samples. The last lane was loaded with sterile nuclease-free water which served as a negative control test. The remaining lanes were loaded with samples under investigation.

2.4. Statistical Analysis

All analyses were conducted using the SPSS version 25 software and GraphPad Prism 6. Variables were expressed as percentages (%), and the chi-square test was performed to assess the relationships between variables. Variables such as the virulence genes and phylogenetic groups of the E. coli strains and sociodemographic and socioeconomic characteristics of patients were considered for the chi-square analysis. values < 0.05 were considered statistically significant.

2.5. Ethical Considerations

Ethical clearance for the study was granted by the Joint Committee on Human Research, Publication and Ethics (CHRPE), School of Medical Sciences and Dentistry, and Komfo Anokye Teaching Hospital, with the protocol number CHRPE/AP/204/18.

3. Results

A total of 135 clinical isolates of E. coli were recovered from the various clinical specimens of patients who visited the Ho Teaching Hospital for medical care. Only 1 (0.7%) isolate was recovered from the pleural aspirate whereas multiple isolates 98 (72.6%) were recovered from the urine samples of patients with urinary tract infections. The proportion of isolates from the other samples was 14 (10.4%) from wound swabs, 10 (7.4%) from HVS, 5 (3.7%) each from blood and ear swabs, and 2 (1.5%) from sputum. The relationship between sociodemographic and patients’ characteristics and the phylogeny of E. coli isolates is presented in Table 3. The majority of the isolates (82.2%) came from women, and by the type of patients, the majority was from outpatients (72.6%). There was no significant difference between the proportion of isolates in each phylogroup when compared by gender. Similarly, there was no significant difference between the proportion of phylogroups when compared by the other demographic characteristics (religion, marital status, and occupation), sample type (urine or non-urine), and patient type (out- or in-patient). Age wise, there was no statistically significant difference between the proportion of isolates by phylogroups except for isolates that belonged to phylogroup B1 () and D ().

Phylogenetic analysis segregated the 135 E. coli isolates into phylogenetic groups A 101 (74.8%), B1 3(2.2%), B2 20 (14.8%), C 4(3.0%), clade I 2(1.5%), and D 5(3.7%).

A total of 12 different virulence genes were identified with the clinical isolates. The virulent gene fimA 120 (88.9%) was the most prevalent, while three (3) of the remaining were prevalent in more than 50% of the isolates (fyuA 106 (78.5%), yfcV 104 (77.0%), and iutA 88 (65.2%)). On the other hand, three of the virulence genes under investigation were identified in less than 15% of the total isolates (chuA 18 (13.3%), ibeA 16 (11.9%), and hlyF 7(5.2%)). Other virulence genes identified were vat 56 (41.5%), hylA 41 (30.4%), neuC 40 (29.6%), cnf 1 39 (28.9%), and sfa/foc 37 (27.4%).

Table 3 presents the relationship between sociodemographic and patients’ characteristics and the phylogeny of E. coli isolates. There is a significant association between patient’s age category and phylogenetic groups B1 and D.

Chi-square analysis for the correlation between virulence genes and the phylogenetic groups of isolates showed no correlation between the two except for the virulence gene ChuA () where the proportions of clade 1 and D were higher than those of the others. The distribution also presented that none of the 101 isolates belonging to group A had the chuA virulence gene. Similarly, none of the B1 and C phylogroups possessed the ChuA virulence gene. Details of these results can be found in Table 4.

4. Discussion

This study investigated the virulence profile and phylogenetic characteristics of extraintestinal pathogenic E. coli. Majority of the organisms were recovered from urine samples received from patients with urinary tract infections. Our finding is consistent with a study by Lara et al. [20] who reported that urinary tract infections are the most common extraintestinal infection caused by E. coli. Our study reported that most of the uropathogenic E. coli belongs to phylogenetic group A. The preponderance of phylogenetic group A in uropathogenic E. coli isolates which is usually associated with commensal strains suggests that the gastrointestinal tract is the main source of strains that colonize the urinary tract [21].

Due to the pathogenicity of E. coli, it is an important cause of extraintestinal infections in health facilities around the globe. The E. coli organism must first adhere to the host cell to establish infection, and this is achieved by its surface adhesins [22]. The current study investigated the presence of fimA, yfcv, and sfa/foc virulence genes responsible for adhesions. These genes were detected in 88.9%, 77.0%, and 27.4% of the E. coli isolates, respectively. Similar results were reported in other parts of the world for sfa/foc genes [3, 23] and fimA genes [24, 25]. These plasmid-encoded fimA genes were commonly found in isolates from infection at the lower urinary tract, where they adhere to the urethral mucosa epithelial cells [24].

In the present study, haemolysin-encoding genes, hlyA and hlyF, were present in 30.4% and 5.2% of the isolates respectively. A higher percentage of hly-encoding genes (41.2%) among the E. coli isolates was observed in Japan [26]. Contrary to our finding, Agarwal et al. [27] in India reported a lower prevalence of 4.7% of E. coli isolates carrying haemolysin genes in women suffering from acute cystitis. HlyA is a pore-forming exotoxin secreted to lyse red blood cells and human renal epithelial cells by creating pores in them. The E. coli organisms utilize the iron released from the lysed erythrocytes through the siderophore system. Its production and expression are controlled by the availability of iron [28]. A high proportion of hlyA recorded in this study belonged to phylogroup D, which corroborates the finding of Zhang et al. [29]. That study reported that highly haemolytic isolates belonged to phylogroup D.

Most of the E. coli isolates in this study belonged to phylogenetic group A (74.8%). Our finding contradicted the study by Iranpour et al. [16], who reported phylogenetic group A minority (0.7%). A plausible reason for this observation could be the difference in sample characteristics; while in their study, they used E. coli isolates from patients with urinary tract infections, our study involved not only urine samples. Although isolates belonging to this phylogenetic group were generally regarded as commensals in the gastrointestinal tract, they transcended the gut and had access to the urinary tract and caused infections. This could be due to the availability of functional genes that directly contribute to pathogenesis or the presence of certain putative factors enabling successful colonization of the host that enhances fitness and adaptation of the bacteria to their surroundings [16] According to Ochman et al. [30], commensal E. coli can acquire chromosomal or extrachromosomal virulence operons and become pathogenic.

In furtherance to that, commensal strains may become virulent by a genomic deletion that enhances pathogenicity and random functional point mutations adaptive for pathogenic environments [31]. However, it is unclear whether E. coli isolates should be defined as commensals or pathogens based wholly on the source of the specimen and/or phylogenetic group they belong to since phylogroups A and B1 can cause extraintestinal infections in immunocompromised hosts at a point in time [32]. The pathogenicity of E. coli isolates belonging to phylogenetic group A is at variance with other studies that categorized pathogenic E. coli isolates into phylogenetic groups B2 and D [33, 34]. Our finding is comparable to other studies in other parts of the globe that reported phylogroup A majority [35, 36]. The prevalence of B1 and B2 phylogroups found in the current study are lower than those reported in Abidjan [37]. The three predominant phylogroups, A, B2, and D, reported in our study agree with those stated by Derakhshandeh et al. [38] but at variance with those described by Iranpour et al. [16]. The variations in the distribution of the phylogenetic groups reported in different studies may be due to the health status of the host, geographical and climatic conditions, dietary factors, antibiotic usage, host genetic factors, and the differences arising from different sampling methods [38].

In addition to the four main phylogroups (A, B1, B2, and D) previously reported, the new Clermont quadruplex PCR method of E. coli discrimination added four new phylogroups made up of C, E, F, and clade I. In this current study, 4 (3.0%) of the isolates belonged to phylogroup C and 2 (1.5%) to clade I. In total, our study reported that 6 (4.4%) of the E. coli isolates belonged to the newly described phylogroups against the 13% by Clermont et al. [8] and 25% by Iranpour et al. [16]. This difference in phylogenetic distribution could be due to socioeconomic factors of the study population, climatic conditions, hygienic status, and dietary habits of the host [39].

A significant association was observed with patients’ age and phylogenic groups B1 and D in the present study. There are divergent views by various scientists on the impact of age on the phylogenetic distribution of E. coli isolates. For instance, Escobar-Páramo et al. [40], in their study, established a link between phylogenicity and patients’ age. Their findings were consistent with other studies [41, 42]. On the other hand, a Chinese report did not find any significant difference in the phylogenetic group composition and the age of the patients [43]. The present study did not show any significant difference in the phylogenetic grouping and sociodemographic characteristics of study participants' gender, occupation, marital status, religion, and admission status) except for age. It supports the study by Iranpour et al. [16], who likewise reported no statistical difference in phylogenetic characteristics and patients’ gender.

5. Conclusions

In conclusion, this study assessed the relationship between the phylogenetic distribution and the virulence profile of clinical isolates of E. coli. Most of the isolates studied belonged to phylogenetic group A and carried essential factors responsible for virulence and pathogenicity. The virulent genes included adhesin (fimA, yfcv, and sfa/foc) and haemolysin (hlyA and hlyF). Patients’ age was significantly observed to be associated with phylogenic groups B1 and D. However, there was no association between virulence genes and phylogenetic distribution except chuA which was found in all phylogroups D and clade I isolates. The virulence of isolates belonging to phylogroup A, generally considered commensals, is alarming. Therefore, measures must be put in place to control the spread of these virulent strains E. coli.

Data Availability

Data is obtainable from the corresponding author upon satisfactory request.

Conflicts of Interest

The authors declare that there is no conflict of interest.

Acknowledgments

The authors are grateful to the staff of Duedu laboratory and the Ho Teaching Hospital laboratory for their support.