Search

The highly virulent Escherichia coli O104:H4 isolate that caused a large outbreak in 2011 carries three plasmids. Out of these, only one, the IncI plasmid pO104_90 that encodes two extended spectrum beta-lactamases, can transfer itself by conjugation. Considering its potential contribution to the emergence and virulence of the outbreak strain, we aimed to get a closer insight into pO104_90 transfer efficiency and control. We tested the host spectrum of the plasmid and observed transmission into Enterobactericeae including clinically relevant enterobacterial pathogens like Salmonella typhimurium and Shigella flexneri. However, we found that this plasmid did not transfer into E. coli strains that kill the donor strain due to bacteriocin production, e.g. the probiotic E. coli Nissle 1917. Under the same conditions, the highly transmittable control plasmid RP4 was efficiently transferred into all these recipients. Therefore we hypothesized that the failure of transfer of pO104_90 was simply due to the generally much lower transmission rates of this IncI plasmid and we decided to screen for factors that negatively affect the transfer of the plasmid by an in vivo deletion analysis. Our attempts to delete larger regions of the plasmid resulted in cells containing both a truncated plasmid (Δ50 kb and Δ75 kb) and a wild type copy of pO104_90. When used as donors in conjugation experiments, these cells transferred the wild type plasmid at dramatically increased rates. This indicated that the relatively limited region shared by both plasmids contained an activator of transfer. We therefore analyzed its transcriptional organization, dissected the candidate region by subcloning and showed that additional copies of repY/INC were sufficient to increase the transfer frequency of pO104_90 to the observed level. To our knowledge, this is the first evidence for a direct regulatory cross talk between core control elements of the vegetative replication and the transfer functions of an IncI1 plasmid.

Ventilator-associated pneumonia (VAP) remains the most frequent life-threatening nosocomial infection. Enterobacteriaceae including Escherichia coli are increasingly involved. If a cumulative effect of pathogenicity islands (PAIs) has been shown for E. coli virulence in urinary tract or systemic infections, very little is known regarding pathophysiology of E. coli pneumonia. This study aimed to determine the role of each of the 7 PAIs present in pathogenic E. coli strain 536 in pneumonia pathophysiology. We used mutant strains to screen pathophysiological role of PAI in a rat pneumonia model. We also test individual gene mutants within PAI identified to be involved in pneumonia pathogenesis. Finally, we determined the prevalence of these genes of interest in E. coli isolates from feces and airways of ventilated patients. Only PAIs I and III were significantly associated with rat pneumonia pathogenicity. Only the antigen-43 (Ag43) gene in PAI III was significantly associated with bacterial pathogenicity. The prevalence of tested genes in fecal and airway isolates of ventilated patients did not differ between isolates. In contrast, genes encoding Ag43, the F17-fimbriae subunits, HmuR and SepA were more prevalent in VAP isolates with statistical significance for hmuR when compared to airway colonizing isolates. The E. coli PAIs involved in lung pathogenicity differed from those involved in urinary tract and bloodstream infections. Overall, extraintestinal E. coli virulence seems to rely on a combination of numerous virulence genes that have a cumulative effect depending on the infection site.

Escherichia coli (E. coli) exhibits considerable physiological and metabolic versatility and includes a variety of non-pathogenic, commensal variants, which belong to the normal gut flora of humans and warm-blooded animals. Additionally, several pathogenic variants have been identified which cause various types of intestinal or extraintestinal infections in humans and animals. In contrast to intestinal pathogenic E. coli (IPEC), which are obligate pathogens, extraintestinal pathogenic E. coli (ExPEC) are facultative pathogens which belong to the normal gut flora of a certain fraction of the healthy population where they live as commensals. Comparative genomics and epidemiological studies have been applied to study genomic diversity, markers, and phenotypic traits that may support discrimination of different E. coli pathotypes. Whereas IPEC are often epidemiologically and phylogenetically distinct from ExPEC and non-pathogenic, commensal strains, many ExPEC and non-pathogenic E. coli share large genomic fractions. Furthermore, extraintestinal infections of elderly or immunocompromised patients can be caused by E. coli variants which differ in their geno- and phenotypes from archetypal ExPEC. Thus, strain typing based on the detection of a limited number of ExPEC virulence/fitness-related genes may be ambiguous. A limited number of ExPEC-dominated clonal complexes can be identified in the E. coli population by multi locus sequence typing. Nevertheless, ExPEC and non-pathogenic E. coli cannot be clearly discriminated by molecular epidemiological approaches. Increased knowledge of the phylogeny, virulence and fitness traits, and host factors contributing to host susceptibility of the different groups of ExPEC variants is required for a better understanding of the biological basis of ExPEC infections.

Bacterial pathogens with the ability to cause persistent infection have different strategies to withstand the induction of host immune responses and to successfully establish long-term colonization. In case of asymptomatic bacteriuria and other persistent infections, prolonged growth in the host is accompanied with genomic alterations that result in e.g., bacterial attenuation thus contributing to bacterial adaptation to their host niche and a reduced activation of host immune responses. The accumulating amount of information regarding bacterial adaptation during persistent infection helps to increase our understanding of driving forces of bacterial adaptation in vivo as well as of factors that contribute to symptomatic infection.

Antigen 43 (Ag43) represents an entire family of closely related autotransporter proteins in Escherichia coli and has been described to confer aggregation and fluffing of cells, to promote biofilm formation, uptake and survival in macrophages as well as long-term persistence of uropathogenic E. coli in the murine urinary tract. Furthermore, it has been reported that glycosylation of the Ag43 passenger domain (alpha(43)) stabilizes its conformation and increases adhesion to Hep-2 cells. We characterized the role of Ag43 as an adhesin and the impact of O-glycosylation on the function of Ag43. To analyze whether structural variations in the alpha(43) domain correlate with different functional properties, we cloned 5 different agn43 alleles from different E. coli subtypes and tested them for autoaggregation, biofilm formation, adhesion to different eukaryotic cell lines as well as to purified components of the extracellular matrix. These experiments were performed with nonglycosylated and O-glycosylated Ag43 variants. We show for the first time that Ag43 mediates bacterial adhesion in a cell line-specific manner and that structural variations of the alpha(43) domain correlate with increased adhesive properties to proteins of the extracellular matrix such as collagen and laminin. Whereas O-glycosylation of many alpha(43) domains led to impaired autoaggregation and a significantly reduced adhesion to eukaryotic cell lines, their interaction with collagen was significantly increased. These data demonstrate that O-glycosylation is not a prerequisite for Ag43 function and that the different traits mediated by Ag43, i.e., biofilm formation, autoaggregation, adhesion to eukaryotic cells and extracellular matrix proteins, rely on distinct mechanisms.

Escherichia coli represents a versatile and diverse enterobacterial species which can be subdivided into (i) nonpathogenic, commensal, (ii) intestinal pathogenic and (iii) extraintestinal pathogenic strains. This classification is mainly based on the presence or absence of DNA regions which are frequently associated with certain pathotypes. In most cases, this genetic information has been horizontally acquired and belongs to the flexible E. coli genome, such as plasmids, bacteriophages and genomic islands. These genomic regions contribute to the rapid evolution of E. coli variants as they are frequently subject to rearrangements, excision and transfer as well as further acquisition of additional DNA thus contributing to the creation of new (pathogenic) variants. Genetic diversity and genome plasticity of E. coli has been underestimated. The accumulating amount of sequence information generated in the era of "genomics" helps to increase our understanding of factors and mechanisms that are involved in diversification of this bacterial species as well as in those that may direct host specificity.

Sorbitol-fermenting (SF) Shiga toxin (Stx)-producing Escherichia coli (STEC) O157:H- strains are emerging as causes of hemorrhagic colitis and the hemolytic-uremic syndrome in Europe. Using subtractive hybridization between SF STEC O157:H- strain 493/89 and STEC O157:H7 strain EDL933, three different fragments, of approximately 700 bp in length, were identified. Each demonstrated > 99% homology to genes encoding the enterohemorrhagic E. coli factor for adherence (efa1) and lymphostatin (lifA). Therefore, a cosmid library was constructed from SF STEC O157:H- strain 493/89, and one clone containing these fragments was sequenced. This sequencing demonstrated a 9669-bp open reading frame (ORF) that had 99.9% sequence homology to efa1 of STEC O111:H- strain E45035 and to lifA of an enteropathogenic E. coli O127:H6 strain E2348/69. In STEC O157:H7 strain EDL933, only small (ca. 3 kb) initial and terminal fragments of this ORF are present. PCR analysis with primers complementary to the efa1/lifA sequence of strain 493/89 indicated that the complete sequence is present in each of 10 SF STEC O157:H- isolates but in none of 10 STEC O157:H7 strains investigated. The presence of the complete efa1/lifA also in both tested E. coli O55:H7 strains supports the hypothesis that SF STEC O157:H- are phylogenetically closer to the proposed E. coli O55:H7 ancestor than STEC O157:H7. Our data demonstrate the presence of a potential virulence gene in SF STEC O157:H- that is only rudimentarily present in STEC O157:H7.

Toxin-specific genes are often located on mobile genetic elements such as phages, plasmids and pathogenicity islands (PAIs). The uropathogenic E. coli strain 536 carries two alpha-hemolysin gene clusters, which are part of the pathogenicity islands I536 and II536, respectively. Using different genetic techniques, two additional PAIs were identified in the genome of the E. coli strain 536, and it is likely that further PAIs are located on the genome of this strain. Pathogenicity islands are often associated with tRNA genes. In the case of the E. coli strain 536, the PAI-associated tRNA gene leuX, which encodes a minor leucyl-tRNA, affects the expression of various virulence traits including alpha-hemolysin production. The exact mode of action of the tRNA5Leu-dependent gene expression has to be identified in the future.