Search

An integrated approach combining information gained by Fourier transformation, linear Markham superposition (real space) and mass-per-length measurement by scanning transmission electron microscopy was used to analyze the helical structure of the rod-like type 1 pili expressed by uropathogenic Escherichia coli strain W3110. The 3D reconstruction calculated from the experimental data showed the pili to be 6.9nm wide, right-handed helical tubes with a 19.31(+/-0.34)nm long helical repeat comprising 27 FimA monomers associated head-to-tail in eight turns of the genetic one-start helix. Adjacent turns of the genetic helix are connected via three binding sites making the pilus rod rather stiff. In situ immuno-electron microscopy experiments showed the minor subunit (FimH) mediating pilus adhesion to bladder epithelial cells to be the distal protein of the pilus tip, which had a spring-like appearance at higher magnification. The subunits FimG and FimF connect FimH to the FimA rod, the sequential orientation being FimA-FimF-FimG-FimH. The electron density map calculated at 18A resolution from an atomic model of the pilus rod (built using the pilin domain FimH together with the G1 strand of FimC as a template for FimA and applying the optimal helical parameters determined to the head-to-tail interaction model for pilus assembly) was practically identical with that of the actual 3D reconstruction.

The binding of uropathogenic Escherichia coli to the urothelial surface is a critical initial event for establishing urinary tract infection, because it prevents the bacteria from being removed by micturition and it triggers bacterial invasion as well as host cell defense. This binding is mediated by the FimH adhesin located at the tip of the bacterial type 1-fimbrium and its urothelial receptor, uroplakin Ia (UPIa). To localize the UPIa receptor on the 16 nm particles that form two-dimensional crystals of asymmetric unit membrane (AUM) covering >90 % of the apical urothelial surface, we constructed a 15 A resolution 3-D model of the mouse 16 nm AUM particle by negative staining and electron crystallography. Similar to previous lower-resolution models of bovine and pig AUM particles, the mouse 16 nm AUM particle consists of six inner and six outer domains that are interconnected to form a twisted ribbon-like structure. Treatment of urothelial plaques with 0.02-0.1 % (v/v) Triton X-100 allowed the stain to penetrate into the membrane, revealing parts of the uroplakin transmembrane moiety with an overall diameter of 14 nm, which was much bigger than the 11 nm value determined earlier by quick-freeze deep-etch. Atomic force microscopy of native, unfixed mouse and bovine urothelial plaques confirmed the overall structure of the luminal 16 nm AUM particle that was raised by 6.5 nm above the luminal membrane surface and, in addition, revealed a circular, 0.5 nm high, cytoplasmic protrusion of approximately 14 nm diameter. Finally, a difference map calculated from the mouse urothelial plaque images collected in the presence and absence of recombinant bacterial FimH/FimC complex revealed the selective binding of FimH to the six inner domains of the 16 nm AUM particle. These results indicate that the 16 nm AUM particle is anchored by a approximately 14 nm diameter transmembrane stalk, and suggest that bacterial binding to UPIa that resides within the six inner domains of the 16 nm AUM particle may preferentially trigger transmembrane signaling involved in bacterial invasion and host cell defense.

Induction of protective immune responses with recombinant antigens is a major challenge for the vaccine industry. Here we present a molecular assembly system that renders antigens of choice highly repetitive. Using this method, efficient antibody responses may be induced in the absence of adjuvants resulting in protection from viral infection and allergic reactions.

The binding of uropathogenic Escherichia coli to the urothelial surface is a crucial initial event for establishing urinary tract infection because it allows the bacteria to gain a foothold on the urothelial surface, thus preventing them from being removed by micturition. In addition, it triggers bacterial invasion as well as host urothelial defense. This binding is mediated by the FimH adhesin located at the tip of the bacterial type 1-fimbrium, a filamentous attachment apparatus, and its urothelial receptor. We have prepared a biotinylated, recombinant FimH-FimC adhesin:chaperone complex and used it to identify its mouse urothelial receptor. The FimH-FimC complex binds specifically to a single 24 kDa major mouse urothelial plaque protein, which we identified as uroplakin Ia by mass spectrometry, cDNA cloning and immunoreactivity. The terminal mannosyl moieties on Asn-169 of uroplakin Ia are responsible for FimH as well as concanavalin A binding. Although FimH binds to uroplakin Ia with only moderate strength (Kd ∼100 nM between pH 4 and 9), the binding between multiple fimbriae of a bacterium and the crystalline array of polymerized uroplakin receptors should achieve high avidity and stable bacterial attachment. The FimH-FimC complex binds preferentially to the mouse urothelial umbrella cells in a pattern similar to uroplakin staining. Our results indicate that the structurally related uroplakins Ia and Ib are glycosylated differently, that uroplakin Ia serves as the urothelial receptor for the type 1-fimbriated E. coli, and that the binding of uropathogenic bacteria to uroplakin Ia may play a key role in mediating the urothelial responses to bacterial attachment.

One of the key questions in protein folding is whether polypeptide chains require unique nucleation sites to fold to the native state. In order to identify possible essential polypeptide segments for folding, we have performed a complete circular permutation analysis of a protein in which the natural termini are in close proximity. As a model system, we used the disulfide oxidoreductase DsbA from Escherichia coli , a monomeric protein of 189 amino acid residues. To introduce new termini at all possible positions in its polypeptide chain, we generated a library of randomly circularly permuted dsbA genes and screened for active circularly permuted variants in vivo . A total of 51 different active variants were identified. The new termini were distributed over about 70% of the polypeptide chain, with the majority of them occurring within regular secondary structures. New termini were not found in approximately 30% of the DsbA sequence which essentially correspond to four α-helices of DsbA. Introduction of new termini into these “forbidden segments” by directed mutagenesis yielded proteins with altered overall folds and strongly reduced catalytic activities. In contrast, all active variants analysed so far show structural and catalytic properties comparable with those of DsbA wild-type. We suggest that random circular permutation allows identification of contiguous structural elements in a protein that are essential for folding and stability.

The activation of lymphocytes and development of adaptive immune responses is initiated by the engagement of TCRs by antigenic peptide–MHC complexes and shaped at the clonal level by both positive and negative costimulatory signals. The B7 family members are involved at several stages in this process. In this issue of the JCI, Vogt et al. show that the B7 family–related protein V-set and Ig domain–containing 4 (VSIG4) can act as an inhibitor of T cell activation (see the related article beginning on page 2817). Intriguingly, the same molecule was recently independently identified as a complement receptor of the Ig superfamily (CRIg) and was convincingly demonstrated to be a receptor for complement component 3 fragments. These findings raise interesting questions regarding the physiological roles and mechanisms of action of this molecule. Identification of dual functions of this molecule provides an additional level of complexity in T cell costimulation.

An elementary step in the assembly of adhesive type 1 pili of Escherichia coli is the folding of structural pilus subunits in the periplasm. The previously determined X-ray structure of the complex between the type 1 pilus adhesin FimH and the periplasmic pilus assembly chaperone FimC has shown that FimH consists of a N-terminal lectin domain and a C-terminal pilin domain, and that FimC exclusively interacts with the pilin domain. The pilin domain fold, which is common to all pilus subunits, is characterized by an incomplete β-sheet that is completed by a donor strand from FimC in the FimC–FimH complex. This, together with unsuccessful attempts to refold isolated, urea-denatured FimH in vitro had suggested that folding of pilin domains strictly depends on sequence information provided by FimC. We have now analyzed in detail the folding of FimH and its two isolated domains in vitro. We find that not only the lectin domain, but also the pilin domain can fold autonomously and independently of FimC. However, the thermodynamic stability of the pilin domain is very low (8–10 kJ mol−1) so that a significant fraction of the domain is unfolded even in the absence of denaturant. This explains the high tendency of structural pilus subunits to aggregate non-specifically in the absence of stoichiometric amounts of FimC. Thus, pilus chaperones prevent non-specific aggregation of pilus subunits by native state stabilization after subunit folding.

Virus like particles (VLPs) are known to induce potent B cell responses in the absence of adjuvants. Moreover, epitope-specific antibody responses may be induced by VLPs that contain peptides inserted in their immunodominant regions. However, due to steric problems, the size of the peptides capable of being incorporated into VLPs while still permitting capsid assembly, is rather limited. While peptides genetically fused to either the N- or C-terminus of VLPs present fewer assembly problems, the immune responses obtained against such epitopes are often limited, most likely because the epitopes are not optimally exposed. In addition, such particles may be less stable in vivo. Here, we show that peptides and proteins engineered to contain a free cys can be chemically coupled to VLPs formed from the hepatitis B core antigen (HBcAg) containing a lys in the immuno-dominant region. By using this approach steric hindrance of capsid assembly is abrogated. Peptides or protein coupled to VLPs in an oriented fashion are shown to induce strong and protective B cell responses even against self-epitopes in the absence of adjuvants. This molecular assembly system may be used to induce strong B cell responses against most antigens.

The binding of uropathogenic Escherichia coli to the urothelial surface is a critical initial event for establishing urinary tract infection, because it prevents the bacteria from being removed by micturition and it triggers bacterial invasion as well as host cell defense. This binding is mediated by the FimH adhesin located at the tip of the bacterial type 1-fimbrium and its urothelial receptor, uroplakin Ia (UPIa). To localize the UPIa receptor on the 16 nm particles that form two-dimensional crystals of asymmetric unit membrane (AUM) covering90 % of the apical urothelial surface, we constructed a 15 A resolution 3-D model of the mouse 16 nm AUM particle by negative staining and electron crystallography. Similar to previous lower-resolution models of bovine and pig AUM particles, the mouse 16 nm AUM particle consists of six inner and six outer domains that are interconnected to form a twisted ribbon-like structure. Treatment of urothelial plaques with 0.02-0.1 % (v/v) Triton X-100 allowed the stain to penetrate into the membrane, revealing parts of the uroplakin transmembrane moiety with an overall diameter of 14 nm, which was much bigger than the 11 nm value determined earlier by quick-freeze deep-etch. Atomic force microscopy of native, unfixed mouse and bovine urothelial plaques confirmed the overall structure of the luminal 16 nm AUM particle that was raised by 6.5 nm above the luminal membrane surface and, in addition, revealed a circular, 0.5 nm high, cytoplasmic protrusion of approximately 14 nm diameter. Finally, a difference map calculated from the mouse urothelial plaque images collected in the presence and absence of recombinant bacterial FimH/FimC complex revealed the selective binding of FimH to the six inner domains of the 16 nm AUM particle. These results indicate that the 16 nm AUM particle is anchored by a approximately 14 nm diameter transmembrane stalk, and suggest that bacterial binding to UPIa that resides within the six inner domains of the 16 nm AUM particle may preferentially trigger transmembrane signaling involved in bacterial invasion and host cell defense.

Assembly of type 1 pili from Escherichia coli is mediated by FimC, a periplasmic chaperone (assembly factor) consisting of two immunoglobulin-like domains. FimC is assumed to recognize the individual pilus subunits in the periplasm mainly via their conserved C-terminal segments and to deliver the subunits to an assembly platform in the outer membrane. Here we present the first biochemical characterization of a periplasmic pilus chaperone and analyze the importance of the two chaperone domains for stability and function. Comparison of the isolated C-terminal domain with wild-type FimC revealed a strongly reduced thermodynamic stability, indicating strong interdomain interactions. The affinity of FimC toward a peptide corresponding to the 11 C-terminal residues of the type 1 pilus adhesin FimH is at least 1000-fold lower compared to binding of intact FimH, confirming that bacterial pilus chaperones, unlike other chaperones, specifically interact with folded pilus subunits.

We tested the extent of emancipation of the Whistle-shake display of male shelducks, Tadorna tadorna, from causal factors controlling its presumed evolutionary precursor, the Body-shake, a comfort movement. Both motor patterns show similarities in form and alternate in a yearly rhythm. First, in an artificial rain experiment, we analysed the influence of a stimulus controlling comfort movements. In spring almost exclusively Whistle-shakes were induced even in the absence of social stimuli, while in summer the birds performed predominantly Body-shakes. This is probably related to the seasonal production of sex hormones. Second, we tested the influence of social stimuli by confronting shelducks with displaying and non-displaying males. During moult when males almost exclusively perform Body-shakes we found no influence of social stimulation on the frequency or form of shakes. However, an effect of social stimulation was found in a period when the drakes gradually replace the Body-shake by the Whistle-shake which is after moult in autumn. In this experiment, birds were confronted with either an empty cage (control), artificial rain, a non-displaying or a displaying male conspecific. In relation to the control situation the birds performed more Whistle-shakes when stimulated socially and more Body-shakes when stimulated with artificial rain. The extent to which both shaking patterns share the same causal factors is discussed. Copyright 1998 The Association for the Study of Animal Behaviour.