Extended Characterization of Human Uropathogenic Escherichia coli Isolates from Slovenia

1. Introduction

Escherichia coli (E. coli) is a very diverse bacterial species found naturally in the intestinal tract of humans and many other animal species. Even though E. coli is known to be part of the normal gut microbiota, some strains – that are pathogenic – cause a wide variety of different intestinal and extraintestinal diseases (Marrs et al., 2005). Typical extraintestinal infections due to E. coli include urinary tract infections (UTI), diverse intra-abdominal infections, pneumonia, surgical-site infection, meningitis, osteomyelitis, soft-tissue infections, bacteremia (Russo & Johnson, 2006).

UTIs are one of the most frequently acquired bacterial infections and E. coli accounts for as many as 90% of all community-acquired UTIs. Approximately 50% of all women have had a UTI by their late 20s. About 20–30% of women with first UTI will have two or more infections; while 5%, will develop chronic recurring infections which greatly disrupt a woman’s life (Marrs et al., 2005). In Slovenia E. coli is the causative agent of approximately 80% of uncomplicated UTIs (Lindič, 2005).

E. coli isolates that cause UTI exhibit a number of specific characteristics and are classified, as uropathogenic E. coli (UPEC), a subgroup of extraintestinal pathogenic E. coli (ExPEC) (Russo & Johnson, 2000). UPEC strains mainly belong to the B2 phylogenetic group and to a lesser extent to the D group, while commensal strains belong to groups A and B1 (Picard et al., 1999). Further, some O-antigens (O1, O2, O4, O6, O7, O18 and O83) are more prevalent among uropathogenic E. coli strains and are therefore associated with UTI (Moreno et al. 2006). In comparison to commensal E. coli strains, UPEC possess an array of virulence factors namely, adhesins, toxins, polysaccharide coatings, invasins, iron uptake systems and systems to evade host immune responses (Oelschlaeger et al., 2002).

Of serious concern and an increasing health problem, on a global scale, is the appearance and spread of antimicrobial resistance. One of the major health care concerns is emergence of multidrug resistant bacteria and clinical microbiologists increasingly agree that multidrug resistant Gram-negative bacteria pose the greatest risk to public health (Kumarasamy et al., 2010). Therefore, it is essential to determine susceptibility of pathogenic strains for antimicrobial agents and association of antimicrobial resistance with virulence genes.

One of the means for acquiring specific virulence factor genes and antimicrobial resistance genes, is via mobile DNA (e. g. conjugal plasmids, transposons/integrons) (Alekshun & Levy, 2007; Boerlin & Reid-Smith, 2008; Dobrindt et al., 2010). Hence, determining the prevalence of mobile elements and examining the correlation of antimicrobial resistance/virulence factor genes with mobile elements among UPEC is of great significance.

The search for alternative antibacterial agents is of great importance and colicins, toxic, narrow killing spectrum exhibiting proteins produced by colicinogenic E. coli, exhibit great potential as an alternative approach in the battle against microbes. It has been reported that colicins are effective against intestinal and UPEC strains (Rijavec et al., 2007; Schamberger et al., 2004; Stahl et al., 2004), and that they prevent colonization of urinary catheters (Trautner et al., 2005). To evaluate the potential of colicins as antimicrobial agents, studies on the prevalence of colicins and colicin resistance are needed.

In summary, to diminish the burden of UPEC, using effective preventive measures, data on phylogenetic groups, serogroups, virulence factor prevalence, antimicrobial resistance, presence of mobile DNA, colicin production and colicin resistance among E. coli isolates from different geographic regions must be assessed.

In Slovenia in 2002, we collected 110 E. coli isolates from humans with community-acquired urinary tract infections at the Institute of Microbiology and Immunology, Medical Faculty, Ljubljana, Slovenia. Isolation was performed according to standard laboratory protocols and UPEC were isolates from > 10⁵ colony-forming units (CFU) (Rijavec et al., 2006). These isolates were studied in order to obtain an extended characterisation, including phylogenetic groups, serogroups, virulence factor prevalence, antimicrobial resistance, presence of mobile DNA and colicinogeny and colicin resistance (Rijavec et al., 2006; Starčič Erjavec et al., 2006; Starčič Erjavec et al., 2007; Starčič Erjavec & Žgur-Bertok, 2008; Starčič Erjavec et al., 2008; Starčič Erjavec et al., 2009; Starčič Erjavec et al., 2010).

2. Phylogenetic groups and subgroups

E. coli strains can be assigned to one of four main phylogenetic groups: A, B1, B2, and D. The four groups were identified on the basis of allelic variation at enzyme-encoding genes detected by multilocus enzyme electrophoresis (Ochman et al., 1983, Whittam et al., 1983). Clermont et al. (2000) established the method of rapid and simple determination of the E. coli phylogenetic groups by a triplex PCR. This genotyping method is based on the amplification of a 279 bp fragment of the chuA gene; a 211 bp fragment of the yjaA gene; and a 152 bp fragment of TSPE4.C2, a noncoding region of the genome. The presence or absence of combinations of these three amplicons is used to assign E. coli to one of the four phylogenetic groups. However, to increase the discriminative power of phylogenetic group analysis, Escobar-Paramo et al. (2004) proposed the introduction of phylogenetic subgroups. They defined (Figure 1), apart from the phylogenetic group B1 (lacking chuA and yjaA and having Tspe4.C2), the following six subgroups: in the phylogenetic group A, subgroupA₀ (lacking chuA, yjaA, and Tspe4.C2) and subgroup A₁ (lacking chuA, having yjaA, and lacking Tspe4.C2); in the phylogenetic group B2, subgroup B2₂ (having chuA and yjaA and lacking Tspe4.C2) and subgroup B2₃ ( having chuA, yjaA, and Tspe4.C2); in the phylogenetic group D, subgroup D₁ (having chuA and lacking yjaA and Tspe4.C2) and subgroup D₂ (having chuA, lacking yjaA, and having Tspe4.C2), Fig. 1.

Analysis of our collection of 110 UPEC isolates showed that 55 (50%) belonged to group B2, 28 (25%) to A, 21 (19%) to D, and 6 (5%) to the B1 group (Rijavec et al., 2006). When subgroups were considered the distribution of the isolates was: 4 (4%) of the studied isolates belonged to the subgroup A₀, 24 (22%) to A₁, 6 (5%) to B2₂, 49 (45%) to B2₃, 16 (15%) to D₁ and 5 (5%) to D₂ (our unpublished data).

The obtained distribution of the studied isolates into the phylogenetic groups, with the large majority classifying to the B2 group was expected since, it is known that ExPEC isolates mainly belong to the B2 phylogenetic group (Picard et al., 1999) however, the disproportionate distribution into the B2₂ (5%) and B2₃ (45%), and A₀ (4%) and A₁ (22%) subgroups was surprising.

3. Serogroups

Serotyping of E. coli isolates is an often used method for distinguishing possible pathogenic E. coli from commensal E. coli. It is complex since among E. coli 173 O-antigens, 80 K-antigens, and 56 H-antigens can be found. The O-, K-, and H-antigens can occur in many possible combinations therefore, the final number of E. coli serotypes is very high, 50,000-100,000 or more. However, the number of frequent pathogenic serotypes is limited. Two main groups of frequent serotypes are: (i) serotypes from diarrhoeal disease and (ii) serotypes from extraintestinal disease (Orskov & Orskov, 1992).

Serotyping of the 110 uropathogenic strains revealed that 77 (70%) were O-antigen typable, 19 (17%) were O-nontypable, and 14 (12.7%) were rough. Sixty-three (57%) of the examined strains were H-antigen typable, 41 (37%) were H-antigen negative, and 6 (5%) were H-antigen nontypable. The O-typable strains were distributed into 31 serogroups. Nevertheless, the most frequent were O2 (9 isolates) and serotype O6:H1 (10 isolates). Five common serotypes were identified in three or more strains: O2:HNT (n = 3), O2:H6 (n = 3),

O6:H1 (n = 10), O7:HNT (n = 3), and O74:H39 (n = 4) accounting for 30% of the serotypable isolates. A number of other serotypes were detected in one or two strains (Rijavec et al., 2006).

Serotype analysis of the studied strains revealed that they belonged to diverse serogroups. However, the most frequent were O2 and O6, which are well established as associated with urinary tract infections. The large majority, 96%, of the O2 and O6 isolates were assigned to the B2 phylogenetic group (Rijavec et al., 2006).

4. Virulence factors

Any component of a microbe that is required for, or potentiates its ability to cause disease is designated as a virulence factor. Many different virulence factors exist however, they can all be placed in one of the four major groups of virulence factors: adhesins, toxins, iron uptake systems and host immunity evading systems. Hence, virulence factors facilitate colonization and invasion of the host, avoidance or disruption of host defence mechanisms, injury to host tissue, and/or stimulation of a noxious host inflammatory response (Johnson and Steel, 2000).

5. Antimicrobial susceptibility

An important task in clinical microbiology is the performance of antimicrobial susceptibility testing in order to detect possible drug resistance in common pathogens and to assure susceptibility to drugs of choice for particular infections (Jorgensen & Ferraro, 2009). Therefore, several studies on the subject of antimicrobial susceptibility and E. coli isolates from UTI have been performed (e. g. Karlowsky et al., 2003; Kahlmeter & Menday, 2003; Yilmaz et al., 2009).

In the year 2002, when the studied isolates were collected, treatment of UTI in outpatients in Slovenia was as follows: a course of antibiotics (therapy of choice - trimethoprim 160 mg or trimethoprim/sulfamethoxazole 160 mg/800 mg twice daily for 3 days) and advice to consume sufficient quantities of liquids (2–3 l per day) (Car et al., 2003).

The studied uropathogenic strains were screened for susceptibility to the following antibiotics: ampicillin, chloramphenicol, kanamycin, streptomycin, tetracycline, trimethoprim, trimethoprim-sulfamethoxazole, ciprofloxacin, norfloxacin, nalidixic acid, mezlocillin, amikacin, cefotaxime, cefotiame, cefoxitin, ceftazidime, and gentamicin. Susceptibilities to the tested antibiotics ranged from 109 (99%) susceptible to amikacin, cefotaxime, ceftazidime, to 47 (43%) susceptible to tetracycline. Antibiotics with the highest prevalence of susceptibility, apart from amikacin, cefotaxime, ceftazidime, were to cefotiame, cefoxitin, and gentamicin as 108 (98%), 103 (94%), and 101 (92%) of the studied strains, respectively, were susceptible. Antibiotics with the lowest prevalence of susceptibility, apart from tetracycline, were to chloramphenicol, ampicilin, mezlocillin and nalidixic acid, as 50 (45%), 57 (52%), 59 (54%) and 64 (58%), respectively, were susceptible (Rijavec et al., 2006).

Forty-six (42%) of the studied strains were resistant to more than three classes of the tested antimicrobial agents—beta lactams, quinolone/fluoroquinolone, trimethoprim/ trimethoprim-sulfamethoxazole, tetracycline, chloramphenicol, aminoglycosides (streptomycin, kanamycin)—and were designated as multidrug resistant (MDR). Subsequently, the association between MDR and the phylogenetic group was examined. A statistically significant correlation between non-MDR and the B2 group was determined and a significant correlation between MDR and the D phylogenetic group was found. On the other hand, there were no statistically significant correlations between MDR or non-MDR strains and the A or B1 groups (Rijavec et al., 2006).

6. Mobile genetic elements

The loss and gain of mobile genetic elements has a pivotal role in shaping the genomes of pathogenic bacteria. Horizontal gene transfer is an important mechanism that rapidly disseminates new traits to recipient organisms. Acquiring these new traits is crucial in promoting the fitness and survival of a pathogen while it coevolves with its host (Croxen & Finlay, 2010).

Bacterial plasmids, self-replicating, extrachromosomal elements are key agents of change in microbial populations. They promote the dissemination of a variety of traits, including virulence, enhanced fitness, resistance to antimicrobial agents, and metabolism of rare substances (Johnson & Nolan, 2009). E. coli strains possess a variety of plasmid types, of different sizes, usually ranging in size from approximately 300 bp to 2400 kbp nevertheless, each plasmid must harbour a replication region (Kado, 1998). Plasmids are classified into incompatibility groups mostly on the basis of the replication region (Couturier et al., 1988).

Integrons are assembly platforms that incorporate exogenous open reading frames by site-specific recombination and convert them to functional genes by ensuring their correct expression. Integrons are composed of three key elements necessary for the capture of exogenous genes: a gene (intI) encoding an integrase belonging to the tyrosine-recombinase family; a primary recombination site (attI); and an outward-orientated promoter (Pc) that directs transcription of the captured genes. At present, five classes of mobile integrons are distinguished. These classes have been historically defined based on the sequence of the encoded integrases, which show 40–58% identity. All five classes are physically linked to mobile DNA elements, such as insertion sequences, transposons and conjugative plasmids, all of which can serve as vehicles for intraspecies and interspecies transmission of genetic material. Class 1 integrons are associated with functional and non-functional transposons derived from Tn402 that can be embedded in larger transposons, such as Tn21. Class 2 integrons are exclusively associated with Tn7 derivatives and class 3 integrons are thought to be located in a transposon inserted in a plasmid. The other two classes of mobile integrons, class 4 and class 5, have been associated only with Vibrio species; class 4 is a component of a subset of SXT elements found in Vibrio cholerae, and class 5 is located in a compound transposon carried on a plasmid in Vibrio salmonicida (Mazel, 2006).

The studied UTI strains were screened for replication regions of IncFI and IncFII plasmids. We found (Table 1) a high (62 strains, 56%) incidence of rep IncF sequences among the examined UPEC strains. Particularly prevalent were RepFIB sequences that were detected in 57 (52%) of the strains while RepFIA and RepFIIA were found in 20 (18%) and 24 (22%) strains, respectively. rep sequences were found in all four phylogenetic groups. Of the 62 isolates harbouring at least one of the tested IncF replicons, 20 belonged to group A, 2 to B1, 27 to B2, and 13 to group D (Rijavec et al., 2006).

Since only class 1, 2 and 3 of integrons were shown to be associated with pathogenic E. coli in our study we focused only on these three classes. Analysis (Table 1) revealed that 29 (26%) of the strains harboured a class 1 integron, 1 strain (1%) contained a class 1 and a class 2 integron, and 4 strains (4%) a class 2 integron. Analysis of the distribution of integrons with regard to the phylogenetic group showed that integron sequences were found in all four groups. Of the 34 isolates harbouring integron sequences, 12 belonged to group A, 1 to B1, 12 to B2, and 9 to group D (Rijavec et al., 2006).

7. Colicins

Colicins are bacteriocins produced by E. coli strains. As other bacteriocins, colicins are extracellular bacterial toxic proteins that are active against the same, or closely related species as the producer cell (Daw & Falkiner, 1996). The mechanism of action of these compounds involves adsorption to specific receptors located on the external surface of sensitive bacteria followed by killing via one of three primary mechanisms: i) formation of channels in the cytoplasmic membrane, ii) degradation of cellular DNA or iii) inhibition of protein synthesis (Riley & Gordon, 1999). Because of their narrow range of activity, it has been proposed that the primary role of bacteriocins is to mediate intraspecific, or population level, interactions (Riley, 1998) however, bacteriocins have also been implicated in virulence determination, since many pathogenic strains harbour plasmid-encoded bacteriocins, for example the ColV plasmid of UPEC isolates (Johnson & Nolan, 2009). Typically, 25–50% of E. coli isolates are colicinogenic and usually, the percentages are higher among pathogenic than commensal strains (Riley & Gordon, 1996). Due to high levels of colicinogenity in natural E. coli populations, high levels of colicin resistance are known to occur (Feldgarden & Riley, 1998).

Among the studied UPEC isolates 42 (38%) exhibited colicinogenic activity (Starčič Erjavec et al., 2006). Each of the 110 UPEC strains was resistant to at least 3 colicinogenic strains from Pugsley's collection of colicinogenic strains (Pugsley & Oudega, 1987), 23 UPEC strains (21%) were resistant to all 20 tested Pugsley's strains (our unpublished data). Colicinogenic strains were found in all four phylogenetic groups (Table 1) however, most colicinogenic strains, 20 (48%) belonged to the B2 group (our unpublished data).

8. Conclusion

Our investigation of UPEC isolates from Slovenia revealed a high prevalence of drug resistance and multidrug resistance. The virulence profile of the examined strains was comparable to that of strains from other geographic regions.

Acknowledgments

This work was supported by the by grant P1-0198 from the Slovenian Research Agency.