Tel.+49 30 83 8-518 40/518 43 Fax.+49 30 838 45 18 51
Enterohemorrhagic Escherichia coli (EHEC) are the cause of diarrhea, bloody diarrhea, hemorrhagic colitis (HC) and the potentially fatal hemolytic uremic syndrome (HUS) in humans. The key virulence factor of EHEC is the bacteriophage-encoded Shiga-Toxin gene (stx). In addition, EHEC can harbor the locus of enterocyte effacement (LEE), which is located on a pathogenicity island, and coffers the ability to cause attaching and effacing (A/E) lesions on epithelial cells as a result of colonization of the intestinal tract. The most common EHEC serotype isolated worldwide is O157:H7. Nevertheless, a variety of non-O157 EHEC have emerged as serious causes of HUS and diarrhea in humans worldwide. The most important non-O157 O serogroups causing one third of the EHEC infections in Germany, are O26, O103, O111 and O145. We hypothesized that these non-O157 differ in their phylogeny and aimed to determine the underlying microevolution leading to the emergence of such non-O157 EHEC.
In this thesis I provide evidence to explain the population structure of the two most-important
non-O157 EHEC, O26 and O111 EHEC, and further propose a model of the microevolution of these EHEC, including also the common ancestor of both lineages. The microevolutionary model of O26 and O111 EHEC is based on results of several analyses of the genome of a large collection of non-O157 EHEC strains and the crucial inclusion of atypical enteropathogenic E. coli (aEPEC) into this collection. Application of multilocus sequence typing (MLST) on a diverse collection of 250 non-O157 STEC and EHEC revealed a cluster formation of strains of O serogroups O26 and O111 in one single sequence type complex, STC29. The additional presence of aEPEC, that differ from STEC/ EHEC merely in their absence of the stx-converting bacteriophage, but share the same O serogroups, among the STEC/ EHEC strains in common sequence types (STs) of STC29, suggests an ongoing microevolutionary scenario, characterized by a bidirectional-conversion, in which the phage-encoded
stx-gene is transferred between aEPEC and STEC/ EHEC.
To test this hypothesis, we performed whole genome sequencing of 99 selected strains (aEPEC n=20, STEC/ EHEC n=79) designated as STC29 and analyzed the single nucleotide polymorphisms (SNPs) of the maximum common genome (MCG) of those strains in order to gain more detailed information on the underlying population structure and microevolution. The resulting minimum spanning tree (MSTree) of the MCG-based SNP-analysis revealed three distinct clusters. Cluster 1 harbored strains of O serogroup O111 also designated as ST16 with MLST. Interestingly, the distinct Cluster 2 included only O26 aEPEC strains of ST29, while the more heterogeneous Cluster 3 combined STEC/ EHEC as well as aEPEC strain of O serogroup O26 that were only roughly separated into strains of ST29 and ST21.
The analysis of the presences or absence of accessory virulence associated genes (VAGs) confirmed the results of the SNP-analysis, and suggests a parallel evolution of the MCG of those strains and the acquisition of virulence genes. Furthermore, the analysis of insertion sites for mobile genetic elements, with respect to their occupation with phage-related genes and insertion elements, resulted in a similar relation of the analyzed strains. Consequently, the common results of MLST, SNP-analysis of the MCG, the presence of VAGs and occupation of insertion sites led to the development of a microevolutionary model of STEC/ EHEC of O serogroups O26 and O111, which developed as two distinct lineages from a common aEPEC ancestor of ST29 by lysogenic conversion with stx-converting bacteriophages.
Moreover, these analyses and the development of the microevolutionary model were performed on human (n=53) and bovine (n=45) STEC/ EHEC and aEPEC strains, representing the most important hosts of such strains. None of the analyses revealed a separated grouping of strains based on host species. Hence, these strains do not appear to harbor host-specific genomic alterations, neither within the MCG nor in the acquired VAGs, and therefore do not appear to emerge by adaptation to a specific niche. These results strongly support the zoonotic nature of aEPEC and STEC/ EHEC.
In conclusion, STEC/ EHEC of O serogroups O26 and O111 originate from a common ancestor, further aEPEC and STEC/ EHEC of those serogroups share a common phylogeny and are bona fide zoonotic agents.