Gut feelings: enteropathogenic E. coli (EPEC) interactions with the host.

Tyrosine phosphorylation of eukaryotic proteins and translocated intimin receptor by enteropathogenic Escherichia coli

A workshop on enteropathogenic Escherichia coli (EPEC) was held on September 20th and 21st 1982 at the National Institutes of Health. 50 investigators from 8 countries discussed the epidemiology diagnosis and pathogenesis of EPEC. By consensus of the workshop participants EPEC were defined as diarrheagenic E. coli belonging to serogroups epidemiologically incriminated as pathogens but whose pathogenic mechanisms have not been proven to be related either to heat labile enterotoxins (LT) or heat stable enterotoxins (ST) or to Shigella like invasiveness. EPEC adhere in a seemingly pathognomonic way to the intestinal epithelium. Due to the fact that investigation of EPEC had been minimal for 25 years the workshop was designed to provide new research directions. Participants reevaluated the role of EPEC as human diarrheal agents and identified existing and provocative possibilities for further investigation. Dr. Roy M. Robins-Browne provided a historical overview. Other participants discussed the worldwide magnitude of the problem diagnosis by serotyping mechanisms of pathogenesis mechanisms of pathogenesis animal models for EPEC adhesiveness tissue culture models for EPEC adhesiveness and mechanisms of pathogenesis. Investigators from several developing countries showed EPEC to be the cause of many epidemic and sporadic cases of the infantile diarrhea that occurs mostly in the summer. EPEC appears to have become a less frequent cause of epidemics in both Europe and America. A spectrum of clinical illnesses due to EPEC occurs but diarrhea caused by EPEC is often severe and protracted. Both in vitro and in vivo studies clearly identify a characteristic histopathologic lesion of enteroadhesiveness as manifested by both intestinal brush border destruction and epithelial cell plasma membrane pedestal formation at the sites of EPEC attachment as seen by electron microscopy. Newly developed animal model systems in pigs and rabbits should help elucidate pathogenic mechanisms of EPEC. Recommendations for future investigations include: reinvestigate the epidemiology of EPEC in the 1980s including their geographic distribution seasonability and relative importance in relation to other more recently discovered pathogens; learn the reasons for age related susceptibility to EPEC related diarrhea and why it presents almost exclusively as a disease of infants younger than 2 years of age; develop simplified accurate diagnostic methods including gene probes or antisera to unique pathogenic factors phenotypically expressed and resolve current confusion about the results of different EPEC toxin assays and the characterizations of the toxins themselves.

The human adapter protein Nck is an important modulator of the actin cytoskeleton. It also is crucial to the infection strategy of enteropathogenic E.coli (EPEC). On contact with human intestinal cells, the EPEC type III secretion system translocates the virulence factor Tir (translocated intimin receptor) into the host cytoplasm which is subsequently inserted into the host cell membrane. Upon tyrosine phosphorylation, the cytoplasmic, C-terminal domain of Tir becomes a high-affinity ligand to the SH2 domain of Nck. Recruitment of Nck results in local remodeling of the host cell cytoskeleton leading to the formation of actin-rich pedestals on which the bacteria reside. Two isoproteins of Nck, Nck1 and Nck2, exist in vertebrates, both of which are recruited by phosphorylated Tir. The precise cellular functions of Nck1 and Nck2 presently remain unclear, though both have been reported to bind a number of human proteins through their SH2 and SH3 domains. Differences between Nck1 and Nck2 likewise remain elusive as several interaction partners are known to bind both proteins whereas others have been reported to interact exclusively with a single Nck family member. The binding properties of Nck1 and Nck2 SH2 domains were investigated both biophysically and structurally to resolve their ligand recognition specificity. Of all binding partners examined, both Nck1 and Nck2 bind a Tir-derived tyrosine phosphorylated peptide with highest affinity. Crystal structures of the Nck1 and Nck2 SH2 domains in complex with this Tir-derived phosphopeptide as well as the uncomplexed Nck1 SH2 domain reveal a common structure and indicate that ligand-binding involves regions conserved between Nck1 and Nck2. An epitope scan, based on the Tir phosphopeptide, identified the binding specificity of Nck1 and Nck2 confirming that the Tir peptide sequence is essentially optimal for high affinity peptide binding to both proteins. A search of human protein databases using the refined peptide sequence identified potentially new interaction partners for Nck SH2 domains, among them the G-protein coupled receptor kinase interacting protein GIT1. Pull-down assays with both Nck SH2 domains indicate that this predicted interaction may be relevant also in vivo.

Tails of two Tirs: actin pedestal formation by enteropathogenic E. coli and enterohemorrhagic E. coli O157:H7

Identification of a Molecular Target for the Yersinia Protein Kinase A

Enteropathogenic Escherichia coli (EPEC), a common cause of infant diarrhea, is associated with high risk of mortality in developing countries. The primary niche of infecting EPEC is the apical surface of intestinal epithelial cells. EPEC employs a type three secretion system (TTSS) to inject the host cells with dozens of effector proteins, which facilitate attachment to these cells and successful colonization. Here we show that EPEC elicit strong NF-κB activation in infected host cells. Furthermore, the data indicate that active, pore-forming TTSS per se is necessary and sufficient for this NF-κB activation, regardless of any specific effector or protein translocation. Importantly, upon infection with wild type EPEC this NF-κB activation is antagonized by anti-NF-κB effectors, including NleB, NleC and NleE. Accordingly, this NF-κB activation is evident only in cells infected with EPEC mutants deleted of nleB, nleC, and nleE. The TTSS-dependent NF-κB activation involves a unique pathway, which is independent of TLRs and Nod1/2 and converges with other pathways at the level of TAK1 activation. Taken together, our results imply that epithelial cells have the capacity to sense the EPEC TTSS and activate NF-κB in response. Notably, EPEC antagonizes this capacity by delivering anti-NF-κB effectors into the infected cells.

Enteropathogenic Escherichia coli: unravelling pathogenesis

Enteropathogenic Escherichia coli (EPEC) is a virulent strain of bacteria that is a leading cause of death in children with diarrhoea, especially in countries with underdeveloped healthcare. EPEC attach to host intestinal epithelial cells (IECs) and secrete proteins into them that subvert the immune response. In particular, these proteins inhibit pro-inflammatory signalling, IEC apoptosis and phagocytosis by leukocytes. It is estimated that there are approximately 800,000 fatalities worldwide per year due to EPEC. Therefore, finding effective treatments for EPEC infections is crucial to the improvement of human health across the globe. In order to find effective treatments, a detailed understanding of EPEC and its infection strategy is needed. Researchers at the Centre for Biomolecular Sciences at the University of Nottingham have recently discovered experimentally that EPEC internalise host pro-inflammatory cytokines, granting dynamic control of protein secretion and reduced susceptibility to aminoglycosides (a class of antibiotic). However, while biological studies have uncovered vital information on many of the mechanisms of EPEC's infection strategy, much is still unknown about the infection strategy. In this thesis, we first formulate and analyse a mathematical model describing a general infection resulting from interactions between host epithelial cells and non-specific bacteria in the human gut. The model comprises a set of ODEs describing the dynamics between a bacterial infection and the innate immune response. We then extend the model to account for EPEC-specific mechanisms with analysis providing insight into the effectiveness of the various aspects of EPEC's infection strategy, allowing the effectiveness of an EPEC infection to be compared to that of a general infection. The EPEC model was then adapted into an in silico multi-scale hybrid model that includes mechanisms on the macro-, cell-, and sub-cellular scales, implemented deterministically and stochastically, with the aim of exploring the dynamics between EPEC and the host's immune system both spatially and temporally. Finally, the hybrid model is extended to explore the effects of bacterial biofilms on the success of the infection. In each of the EPEC models, we account for the effects of antibiotics and the recent experimental discoveries discussed above. Analysis of the models via steady state, stability, sensitivity and bifurcation analyses and numerical simulations allows us to make predictions on the effectiveness of antibacterial treatment strategies. Our models predict effective treatment strategies targeting bacterial protein synthesis and replication, the immune response, the application of antibiotic, the positioning of bacteria and the ability of bacteria to form biofilms. We hope our predictions will inform, and reduce the time and cost of, future experimental work.

The phylogenetic tree segregates into three major groups, which probably descended from a common ancestor: the bacteria, the archaea, and the eucarya (1–3). While bacteria and archaea lack a true nucleus and intracellular organelles, eucarya possess these. The origin of the ancestor remains elusive, but it emerged about 4 billion years ago. The bacteria probably started life 3.5 billion years ago and eukaryotes emerged 1.75 billion years thereafter. Sometime in between archaea appeared. Eukaryotes began as unicellular organisms but soon also developed multicellular forms. Bacteria and archaea had sufficient time to exploit all niches of the abiotic environment, ranging from ice cold glacial lakes to fiery geysers – from the highest mountain peaks to the deepest sea beds. They also quickly developed forms of mutualism by forming colonies composed of mixed populations. With the evolution of eucarya, new environments could be exploited. Today prokaryotes are the most prevalent form of life on earth, making up the majority of our biomass. It has been estimated that more than 10 prokaryotic organisms live on our globe. Thus, archaea and bacteria are far more prevalent than eucarya. Even human beings comprise more prokaryotic cells than mammalian cells, which range on the order of 10 human cells and 10 prokaryotes mostly concentrated in our intestinal system (4–7).

The apicomplexan parasite Toxoplasma gondii recognizes, binds, and penetrates virtually any kind of mammalian cell using a repertoire of proteins released from late secretory organelles and a unique form of gliding motility (also named glideosome) that critically depends on actin filaments and myosin. How T. gondii glycosylated proteins mediate host-parasite interactions remains elusive. To date, only limited evidence is available concerning N-glycosylation in apicomplexans. Here we report comprehensive proteomics and glycomics analyses showing that several key components required for host cell-T. gondii interactions are N-glycosylated. Detailed structural characterization confirmed that N-glycans from T. gondii total protein extracts consist of oligomannosidic (Man(5-8)(GlcNAc)2) and paucimannosidic (Man(3-4)(GlcNAc)2) sugars, which are rarely present on mature eukaryotic glycoproteins. In situ fluorescence using concanavalin A and Pisum sativum agglutinin predominantly stained the entire parasite body. Visualization of Toxoplasma glycoproteins purified by affinity chromatography followed by detailed proteomics and glycan analyses identified components involved in gliding motility, moving junction, and other additional functions implicated in intracellular development. Importantly tunicamycin-treated parasites were considerably reduced in motility, host cell invasion, and growth. Collectively these results indicate that N-glycosylation probably participates in modifying key proteins that are essential for host cell invasion by T. gondii.

Apoptosis, or programmed cell death, is a well-ordered process that allows damaged or diseased cells to be removed from an organism without severe inflammatory reactions. Multiple factors, including microbial infection, can induce programmed death and trigger reactions in both host and microbial cellular pathways. Whereas an ultimate outcome is host cell death, these apoptotic triggering mechanisms may also facilitate microbial spread and prolong infection. To gain a better understanding of the complex events of host cell response to microbial infection, we investigated the molecular role of the microorganism Enteropathogenic Escherichia coli (EPEC) in programmed cell death. We report that wild type strain of EPEC, E2348/69, induced apoptosis in cultured PtK2 and Caco-2 cells, and in contrast, infections by the intracellularly localized Listeria monocytogenes did not. Fractionation and concentration of EPEC-secreted proteins demonstrated that soluble protein factors expressed by the bacteria were capable of inducing the apoptotic events in the absence of organism attachment, suggesting adherence is not required to induce host cell death. Among the known EPEC proteins secreted via the Type III secretion (TTS) system, we identified the translocated intimin receptor (Tir) in the apoptosis-inducing protein sample. In addition, host cell ectopic expression of an EPEC GFP-Tir showed mitochondrial localization of the protein and produced apoptotic effects in transfected cells. Taken together, these results suggest a potential EPEC Tir-mediated role in the apoptotic signaling cascade of infected host cells.

Many mucosal pathogens use type III secretion systems for the injection of effector proteins into target cells. The type III-secreted proteins EspB and EspD of enteropathogenic Escherichia coli (EPEC) are inserted into the target cell membrane. Together with EspA, these proteins are supposed to constitute a molecular syringe, channelling other effector proteins into the host cell. In this model, EspB and EspD would represent the tip of the needle forming a pore into target cell membranes. Although contact-dependent and Esp-mediated haemolytic activity by EPEC has already been described, the formation of a putative pore resulting in haemolysis has not been demonstrated so far. Here, we show that (i) diffusely adhering (DA)-EPEC strains exhibit a type III-dependent haemolytic activity too; (ii) this activity resides in the secreted proteins and, for DA-EPEC strains, in contrast to EPEC strains, does not require bacterial contact; and (iii) pores are introduced into the target cell membrane. Osmoprotection revealed a minimal pore size of 3-5 nm. The pores induced by type III-secreted proteins of DA-EPEC were characterized by electron microscopy techniques. Analysis by atomic force microscopy demonstrated the pores to be composed of six to eight subunits with a lateral extension of 55-65 nm and to be raised 15-20 nm above the membrane plane. We could also demonstrate an association of EspB and EspD with erythrocyte membranes and an interaction of both proteins with each other in vitro. These results, together with the homologies of EspB and EspD to proposed functional domains of other pore-forming proteins (Yop/Ipa), strongly support the idea that both proteins are directly involved in pore formation, which might represent the type III secretion system translocon.

The human pathogens enteropathogenic Escherichia coli (EPEC) and vaccinia virus trigger actin assembly in host cells by activating the host adaptor Nck and the actin nucleation promoter neural Wiskott-Aldrich syndrome protein (N-WASP). EPEC translocates effector molecules into host cells via type III secretion, and the interaction between the translocated intimin receptor (Tir) and the bacterial membrane protein intimin stimulates Nck and N-WASP recruitment, leading to the formation of actin pedestals beneath adherent bacteria. Vaccinia virus also recruits Nck and N-WASP to generate actin tails that promote cell-to-cell spread of the virus. In addition to Nck and N-WASP, WASP-interacting protein (WIP) localizes to vaccinia virus tails, and inhibition of actin tail formation upon ectopic expression of WIP mutants led to the suggestion that WIP is required for this process. Similar studies of WIP mutants, however, did not affect the ability of EPEC to form actin pedestals, arguing against an essential role for WIP in EPEC-induced actin assembly. In this study, we demonstrate that Nck and N-WASP are normally recruited by vaccinia virus and EPEC in the absence of WIP, and neither WIP nor the WIP family members CR16 and WIRE/WICH are essential for pathogen induced actin assembly. In addition, although Nck binds EPEC Tir directly, N-WASP is required for its localization during pedestal formation. Overall, these data highlight similar pathogenic strategies shared by EPEC and vaccinia virus by demonstrating a requirement for both Nck and N-WASP, but not WIP or WIP family members in pathogen-induced actin assembly.

Enteropathogenic Escherichia coli strains (EPEC) carrying the eae gene encoding intimin are divided into typical strains producing bundle forming pili, encoded by the bfpA gene, and atypical strains lacking the gene. In the study typical and atypical EPEC that did not agglutinated with EPEC polyvalent antisera but carrying virulence factors characteristic to other pathogenic E. coli i.e. diffusely adhering and enteroaggregative E. coli were isolated from 24 (43.6%) of 55 children > 10 years old with persistent diarrhea. These results indicated that non-typeable typical and atypical EPEC can contribute to chronic intestinal infections in teenagers.

Typical enteropathogenic Escherichia coli (EPEC) strains produce bundle-forming pili (BFP), type IVB fimbriae that have been implicated in EPEC virulence, antigenicity, autoaggregation, and localized adherence to epithelial cells (LA). BFP are polymers of bundlin, a pilin protein that is encoded by the bfpA gene found on a large EPEC plasmid. Striking sequence variation has previously been observed among type IV pilin genes of other gram-negative bacterial pathogens (e.g., Pseudomonas and Neisseria spp.). In contrast, the established sequences of bfpA genes from two distantly related prototype EPEC strains vary by only a single base pair. To determine whether bundlin sequences vary more extensively, we used PCR to amplify the bfpA genes from 19 EPEC strains chosen for their various serotypes and sites and years of isolation. Eight different bfpA alleles were identified by sequencing of the PCR products. These alleles can be classified into two major groups. The alpha group contains three alleles derived from strains carrying O55, O86, O111, O119, O127, or O128 somatic antigens. The beta group contains five alleles derived from strains carrying O55, O110, O128ab, O142, or nontypeable antigens. Sequence comparisons show that bundlin has highly conserved and variable regions, with most of the variation occurring in the C-terminal two-thirds of the protein. The results of multilocus enzyme electrophoresis support the hypothesis that bfpA sequences have spread horizontally across distantly related clonal lineages. Strains with divergent bundlin sequences express bundlin protein, produce BFP, and carry out autoaggregation and LA. However, four strains lack most or all of these phenotypes despite having an intact bfpA gene. These results have important implications for our understanding of bundlin structure, transmission of the bfp gene cluster among EPEC strains, and the role of bundlin variation in the evasion of host immune system responses.