Presence of virulence genes and pathogenicity islands in extraintestinal pathogenic Escherichia coli isolates from Brazil

Introduction: Extraintestinal pathogenic Escherichia coli (ExPEC) is associated with various diseases such as urinary tract infections, neonatal meningitis and septicemia. There are many virulence factors (VF) encoded by genes in ExPEC, including papC, papG, ecpA, iroN, fyuA, iutA, ompTp, tsh, hlyF, hlyA and iss. These virulence genes may be present in pathogenicity islands (PAI) or plasmids. Methodology: In this study, we analyzed the presence of VF encoding genes, PAI sequences and phylogenetic groups of 96 ExPEC strains isolated from the urine and blood of patients at the University Hospital of Londrina, and we compared them with 50 faecal commensal strains from healthy individuals. Results: The VF fyuA (65.60%) was detected in pathogenic strains and commensal strains (46%). A comparison of the distribution of ExPEC and commensal strains in the phylogenetic groups showed that more ExPEC strains belonged to group B2 whereas more of the commensal isolates belonged to group A. The distribution of the seven PAI sequences between commensal strains and ExPEC strains showed that PAI IV536 was common in both ExPEC and commensal isolates. Conclusions: These results showed that the ExPEC strains that belonged to group B2 had more PAI sequences compared to those of the other groups, especially group B1, which had virulence genes but the lowest percentage of PAI sequences, which leads us to conclude that the virulence of ExPEC strains characterized as B2 is likely attributed to PAI encoded genes, whereas the virulence of ExPEC strains belonging to phylogenetic group B1 is likely due to plasmid encoded virulence genes.


Introduction
Among Escherichia coli pathotypes, extraintestinal pathogenic E. coli (ExPEC) are important etiological agents of bacteremia and other systemic infections. ExPEC are often classified in different groups according to the anatomical site of infection, with uropathogenic E. coli (UPEC) being associated with urinary tract infections (UTIs) [1]. UTIs are one of the most common infections in the general population globally [2], and E. coli are the most commonly isolated etiological agent of UTIs, accounting for approximately 70%-85% of urinary infections.
In addition to the aforementioned virulence genes, genomic islands (GI), which include pathogenicity islands (PAIs), have been identified in ExPEC that 1069 carry blocks of genes encoding several virulence factors. These, may have been acquired by plasmids and bacteriophages and can encode the molecular arsenal that enables these strains to colonize the urinary tract and survive in extraintestinal sites. In UPEC, the most studied PAIs are: PAI I and II identified in E. coli J96 and CFT073, and PAI I and IV in E. coli, with PAI IV 536 being more common in UPEC [6,7]. The presence of PAIs in UPEC is a feature associated with clinically severe forms of infection [8].
Furthermore, E. coli can be classified phylogenetically into four main groups: A, B1, B2, and D [9][10][11]. Commensal isolates belong mainly to phylogenetic groups A and B1 [12]; intestinal pathogens are spread among groups B1, D, and less frequently belong to group A [5]; whereas human ExPEC are mostly members of group B2 and, to a lesser extent, group D [13].
The goal of this study was to investigate the presence of genes encoding virulence factors and pathogenicity island associated sequences, and to compare the phylogenetic groups of ExPEC isolated from humans to the phylogenetic groups of faecal commensal isolates in southern Brazil.

Isolation and identification of E. coli strains
In this study, we studied 96 ExPEC isolates (90 from the urine of patients with a UTI; 6 from blood of patients with bacteremia) collected between 2008-2012. As a comparative control, 50 E. coli isolates were also collected from faecal samples of healthy volunteers. The participants in this study, were all 18 years of age or older, and were seen by the outpatient service and emergency department of the University Hospital (UH) and Clinic Hospital (CH), Universidade Estadual de Londrina (UEL/PR). This study was approved by the Human Research Ethics Committee -UEL (Process 23335/08). E. coli strains were isolated and identified by the Phoenix automated method, and samples were stored in Brain Heart Infusion (BHI) (Difco, Livonia, USA) plus 20% glycerol (Sigma Aldrich, St. Louis, USA) media at -80 °C. ExPEC strains RS218, V27 [14], J96 [6] and EPEC 2348/69 [15] were used as reference strain controls.
Ten microliters of each reaction mixture were subjected to electrophoresis on a 2% agarose gel followed by staining with 1 µl Gel Red 20X (Biotium, Hayward, USA) (1:500) and visualization on a UV transilluminator. A 1 Kb DNA ladder (Invitrogen) was run on each gel.

Detecting PAI sequences
Specific sequences corresponding to seven different pathogenicity islands were analyzed by PCR: PAI I 536, PAI II 536, PAI III 536 , PAI IV 536 , PAI I J96, PAI I CFT073 and PAI II CFT073. The isolates were subjected to PCR in a thermal cycler (Applied Biosystems, Carlsbad, USA) with an annealing temperature of 55C as described by Sabaté et al. [6].

Phylogenetic classification
The phylogenetic classification into groups was carried out according to the method developed by Clermont et al. [9] and modified by Kotlowski et al. [10]. The samples were subjected to PCR in a thermal cycler (Applied Biosystems) with an annealing temperature of 54C. Ten microliters of each reaction mixture were subjected to electrophoresis on a 2% agarose gel followed by staining with 1 µl Gel Red (Biotium) and visualization in a UV transilluminator. A 1 Kb DNA ladder (Invitrogen) was run on each gel.

Southern blot and DNA probe tests
Native plasmids from 10 isolates were extracted and analyzed as described by Kado and Liu [20]. We selected only the strains that had the genes ompTp and hlyF to check if they had the native plasmid and then tested for other genes present in what has been described as the conserved virulence plasmidic (CVP) region (having 8 genes or operons: iro, iuc, sit, ompTp, cva, hlyF, ets and iss) [21] to confirm the presence of the native plasmid. Plasmidic DNA was separated by agarose gel electrophoresis and transferred to nylon membranes, Hybond-N+membrane (Amersham, GE Healthcare Life Sciences, Cleveland, USA) by capillary blotting. The probes for Southern hybridization were synthesized by PCR using specific primers for detection of virulence genes ompTp, hlyF, sitA, and APEC strain χ7122 [22] was used as positive a control. The amplified product was labeled with DIG-High Prime DNA Labeling and Detection Starter Kit II (Roche) according to the manufacturer's instructions. The blot was hybridized, blocked, incubated with anti-digoxigenin antibodies and developed by chemiluminescence according to the manufacturer's recommendations.

Statistical analysis
Comparison of frequencies among different groups was analyzed by the Fisher exact test. The threshold for statistical significance was a P value < 0.05. The test was performed with the statistical program BioEstat version 5.3.

Analysis of the distribution of extraintestinal infections
Among the 96 patients infected with ExPEC, 90 (93.8%) were from urinary tract infections that were categorized as cystitis (91.1%), urosepsis (6.7%) and pyelonephritis (2.2%) ( Table 2). Using the records of patients, we were able to identify that 20% of the cases were recurrent infections. Six patients had bacteremia with positive blood cultures. Furthermore, UTIs were more frequent in female patients (71.9%). We noticed that the age of the patients with UTIs was most commonly between 18 and 30 years of age (30.4% of cases), while male patients were more often found in the >54 years age group, representing 57.1% of cases.

Antimicrobial susceptibility testing
In the analysis of antimicrobial susceptibility testing performed on ExPEC isolates, we observed that among the 96 tested, 42 (43.75%) were resistant to at least one antimicrobial. In addition, 27 (28%) of the ExPEC isolates showed resistance to three or more of the 15 antimicrobials tested. Antimicrobials for which ExPEC were found to be commonly resistant included sulfamethoxazole-trimethoprim (35.41%), ciprofloxacin (33.33%), levofloxacin (32.29%) and cefazolin (8.33 %) (Figure 1). Of the 50 faecal E. coli isolates used as a case control, only 6 (12%) were resistant to one or more antibiotics. As such, the ExPEC strains were significantly more resistant to antimicrobials (p <0.01) than the faecal control strains.

Analysis of presence of virulence genes among E. coli strains
Among the genes tested for by PCR, ecpA (94.8 %) and fyuA (65.6 %) were the most commonly identified, and both were more prevalent in strains belonging to phylogenetic groups B2 and D. As shown in Table 3, however, the ecpA gene was also more common in faecal isolates (96%). For the iroN and iutA genes, 35%-50% of strains were positive, and these genes were present among all phylogenetic groups. The iroN gene is more frequent in group B2, and iutA was prevalent in all groups except group A. The genes ompTp, tsh, hlyF and hlyA were less prevalent (10%-20%) and were distributed among all phylogenetic groups, though hlyA was not prevalent in group B2. The iss gene, was present in 20% of ExPEC strains and was more distributed in group B1, which is statistically significant when compared to group A. Finally, papC, and papG were detected in 20-22% of ExPEC strains, which was not significant when compared to faecal strains (p > 0.05), and they were more common in strains from groups B2 and D. These results are shown in Table 3. In faecal commensal E.coli strains, the majority of these genes were less common, with the exception of ecpA, or were not found, as is the case for the iutA gene (p < 0.01). Therefore, in commensal strains, group A was prevalent especially for the fyuA, hlyA and ecpA genes. When compared to ExPEC strains, most genes were present in 0-15% of the commensal faecal isolates.

Analysis of the distribution of PAI sequences in ExPEC and commensal strains
Of the 96 ExPEC strains analyzed in this report, 75 (78%) were positive for PAI sequences and 26 (52%) of 50 faecal E. coli isolates contained PAI sequences. A comparison of the distribution of PAI sequences in faecal commensal strains and ExPEC strains are shown in Figure 2. PAI III 536 sequences were present in only one strain (1.04%), whereas PAI I J96 was not   1072 detected in any of the isolates. The distribution of PAIs in ExPEC and faecal commensal isolates is presented in Figure 2.
The distribution of PAI sequences in ExPEC and commensal strains by phylogenetic group shows that most ExPEC strains contain PAI sequences (78%) and belong to group B2 (42.7%). In commensal strains that have PAI sequences (52%) most of them belong to group A (84.6%). Of the 96 ExPEC isolates, three or more PAI sequences were found in 30 ExPEC strains (31.25%). Strains that are considered to be most virulent, generally belonging to group B2, had three or more PAI sequences (86.7%), whereas strains that did not have PAI sequences, were more commonly distributed in group B1 (47.6%) ( Table 4).

Southern blot and DNA probe tests
All 10 samples tested presented (1 to 5) plasmids of high molecular weight, and all the samples that had hlyF and ompTp genes were present in the higher molecular weight plasmids, as demonstrated by Southern blot using specific DNA probes labeled with digoxigenin. Of the 10 samples tested, fifty percent belonged to the B1 group and fifty percent harbored seven of eight genes presented in the CVP region (Table 5).

Discussion
There is a major concern regarding acquisition of resistance to several drugs used in the treatment of infections caused by E. coli causing UTIs [23].
Antimicrobials that are widely prescribed for treatment of UTI episodes include beta-lactams, trimethoprimsulfamethoxazole, quinolones, fluoroquinolones and aminoglycosides [24]. In Southern Brazil we have little recent data concerning the antimicrobial resistance of ExPEC. We had higher resistance profiles in ExPEC compared to commensal isolates, particularly with respect to resistance to fluoroquinolones (ciprofloxacin and levofloxacin) and trimethoprim-sulfamethoxazole (both at approximately 1/3 of strains being resistant) compared to other studies [24] (Figure 1). Among the ExPEC strains we tested, 28% were resistant to three or more antimicrobials. It is necessary to perform antimicrobial susceptibility testing because the E. coli strains show  ExPEC are known to produce several virulence factors, contributing to their survival in the host urinary tract, blood, and other tissues. Among the eleven tested virulence genes, all of them were detected in some of the ExPEC strains and faecal commensal isolates, as shown by others [4,25]. The gene most commonly found in both ExPEC and commensal samples was ecpA (94.8% and 96%, respectively). Despite potentially playing a role in UTI virulence [26], as ecpA is highly conserved between both ExPEC and faecal commensal strains, it cannot serve as a marker to potentially differentiate ExPEC from commensal E. coli. However, in commensal strains, group A was prevalent especially for genes fyuA, hlyA and ecpA. In strains belonging to group B2 a greater presence of all tested genes was expected because this group is characterized as being the most virulent [13]. In our study, the percentage of ExPEC belonging to group B2 was 34.4%. By contrast, although ExPEC group B2 predominated, ExPEC group B1 was also common (19.8%) in our study.
The comparison of the presence of these virulence genes in commensal faecal E. coli isolates and ExPEC showed that some genes are significantly more present in ExPEC (Table 3). Genes such as iroN, iutA, ompTp and hlyF present in the CVP region [21] were significantly more frequent in clinical isolates compared to commensal faecal isolates. In this study, fifty percent of the ExPEC strains containing plasmidencoded virulence-associated genes such as ompTp and hlyF belonged to phylogenetic group B1 ( Table 5).
The PAI sequences were also present in some commensal faecal isolates although less often (52%) than in ExPEC strains (78%). The distribution of PAI sequences in ExPEC samples in this study also corroborates the results published by Sabaté et al. [6]. PAI IV 536, reported as the most common in UPEC, was found most frequently in our strains ( Figure 2). This PAI is quite stable in E. coli 536 (originally isolated from a patient with acute pyelonephritis) which could explain its high frequency in our strains as well. The PAI III 536, which is classified as unstable, easily lost and not required for virulence [6], was present in a low percentage of ExPEC (1.04%). The majority of strains that had PAI sequences belonged mainly to phylogenetic group B2, and the strains that belonged to group B1 had few PAI sequences, which leads us to conclude that the virulence of ExPEC strains characterized as B2 is likely attributed to PAI encoded genes, whereas the virulence of ExPEC strains belonging to phylogenetic group B1 is likely due to plasmid encoded virulence genes ( Figure 3).

Conclusion
In summary, we have found virulence genes in our ExPEC and commensal samples, but some genes (iroN, iutA, ompTp and hlyF) are significantly more enriched in ExPEC strains. We also observed that PAI sequences and the B2 phylogroup were highly prevalent among ExPEC strains. Our results suggest that continued research on virulence factors of ExPEC is necessary because there is no consensus regarding which genes would be ideal virulence markers, as all of these genes provide the bacteria more resistance against host defense mechanisms when compared to the commensal strains. . Study of pathogenicity islands, phylogenetic classification, and virulence genes in 96 ExPEC strains from human. In blue are the samples belonging to phylogenetic group B2; in orange, the samples belonging to group D; in green, the group B1; and finally in pink, samples related to group A.