Bacterial Adhesins and Their Assembly
Chapter
150
SCOTT J. HULTGREN, C. HAL JONES, and STAFFAN NORMARK
Organelle development is a fundamental process in all living cells, even in pathogenic bacteria. Pathogenic bacteria assemble virulence structures on their surfaces that serve as organelles of attachment (23, 24, 48, 55, 97, 99, 127, 134, 183, 186, 215, 216, 224, 225, 256, 283). The sticky attributes of attachment organelles are determined by proteins called adhesins. Adhesins have a structure that allows them to recognize molecules present on host cells, called receptors, with stereochemical specificity. The stereochemical fit between adhesins and receptors plays a significant role in determining the diverse host and tissue tropisms of pathogenic bacteria. Knowledge of the three-dimensional structure of a bacterial adhesin will facilitate the understanding of the molecular basis of host-pathogen interactions, but to date no known crystal structures have been determined. Adhesins are often assembled into hairlike fibers called pili (Fig. 1) (23, 24, 55, 99, 136). The fiber of the pilus is often composed of structural proteins that have the apparent function of serving as a pedestal for adhesin presentation. Such tip-associated adhesins are often components of specialized structures called tip fibrillae that are joined to the distal ends of thicker pilus rods (124, 145). Some proteins serve both as the structural component of the pilus fiber and as the adhesin. Adhesive pili range in diameter from 2 nm to 10 nm. Other adhesive organelles are very thin (<2 nm). Thin fibers tend to coil up into a fuzzy adhesive mass on the surface of the bacterium that is sometimes called a capsular antigen, thin aggregative pili, or curli. There are also many examples of adhesins that are not part of any known oligomeric structures. These adhesins are typically referred to as non-pilus-associated adhesins.
Escherichia coli is a diverse organism causing both gastrointestinal and extraintestinal diseases. E. coli produces a large number of different adhesive proteins and organelles. There are three known molecular machineries used by gram-negative bacteria, including E. coli and Salmonella spp., to assemble adhesive organelles, namely the chaperone/usher pathway, the general secretion pathway, and the extracellular nucleation/precipitation pathway (Table 1). The chaperone/usher pathway is common to all strains of E. coli and most Salmonella spp. and is probably the best understood of all of the assembly machineries. It will be discussed in detail throughout much of this chapter. The general secretion pathway is required for the assembly of adhesive fibers called type 4 pili. Type 4 pili have not yet been discovered in Salmonella. In E. coli, type 4 pili have been found only in enteropathogenic E. coli (EPEC) and enterotoxigenic E. coli (ETEC), and the molecular basis of these systems was being elucidated during the writing of this chapter. The extracellular nucleation/precipitation pathway was originally discovered in E.coli. Organelles assembled by this pathway have been termed curli, and they mediate the binding to fibronectin.
Table 1Strategies for adhesin presentation |
In addition to the three assembly pathways mentioned above, new adhesive mechanisms have also been discovered in various diarrheagenic E. coli, of which there are six groups (127). EPEC is a common cause of infant diarrhea. ETEC is a common cause of dehydrating diarrhea in children in developing countries and of traveler’s diarrhea. Enteroinvasive E. coli is very similar to Shigella spp. Enterohemorrhagic E. coli is a cause of bloody and nonbloody diarrhea as well as hemolytic uremic syndrome. Enteroaggregative E. coli is associated with persistent diarrhea and was so named because these strains adhere to HeLa cells in an aggregative pattern. Finally, diffusely adherent E. coli has also been associated with diarrhea in several studies, and its name comes from its ability to adhere to HeLa cells in a diffuse pattern. New adhesive mechanisms discovered in these E. coli strains will be discussed in the section below.
In this chapter, we used the assembly pathways as a guide to group the diverse adhesins that exist in E. coli and Salmonella as summarized in Table 1. Instead of discussing every attribute of every adhesive system known or listed in Table 1, we chose to discuss some of the more well-characterized systems as prototypes for each assembly classification. The best-understood system is the chaperone/usher system, which consequently is covered the most extensively in this chapter. The reader who is interested in the intricacies of a particular system is advised to consult the appropriate references listed in Table 1. Finally, this chapter focuses on the molecular biology of adhesins and their assembly. Readers who are interested in the cell biology or other aspects of pathogenesis are advised to consult the other excellent chapters in this book.
There are at least 40 architecturally diverse adhesive organelles that are assembled via the chaperone/usher pathway in gram-negative bacteria including Salmonella spp. and E. coli (136). Table 1 lists the adhesive organelles assembled via this pathway for which the chaperone genes have already been sequenced. These attachment organelles contain diverse adhesins with different receptor-binding specificities that contribute to the various host and tissue tropisms exhibited by the different pathogens (Table 1). This pathway is used to assemble both pilus and non-pilus adhesins as well as fibers with atypical morphologies. The atypical fibers tend to be very thin and to coil up into an amorphous mass on the surface of the bacterium.
Pili biogenesis is being used as a model to investigate one of the most basic problems in molecular biology: how proteins fold into domains that can serve as assembly modules for building up organelles (99, 103). The formation of pili represents a well-understood organelle development system (38, 41, 103, 136, 222). The assembly of adhesive pili, via the chaperone/usher pathway, requires immunoglobulin-like periplasmic chaperones (16, 92, 93, 103, 123, 135, 217). The immunoglobulin-like structures of periplasmic chaperones are utilized by pilus subunits like castings or templates to achieve their correct three-dimensional structures so that they can be assembled into pili (103, 147, 262). The analysis of the mechanism of action of periplasmic chaperones is providing insight into basic biological processes of chaperone-assisted import, folding, and targeting of proteins to specific assembly sites. The role of outer membrane proteins, called ushers, in uncapping chaperones from their target proteins and allowing the ordered assembly of the subunits into a surface structure is also beginning to provide basic knowledge about organelle development (52, 109). The chaperone/usher pathway will be discussed throughout this chapter to exemplify general principles in pilus biogenesis and the molecular basis and consequences of adhesin-receptor interactions.
The extracellular nucleation/precipitation pathway (Table 1) is required to produce a type of adhesive fiber found in E. coli and Salmonella that has been termed curli (9, 43, 213). Curli differ from pili in that they are thought to be formed extracellularly by precipitation of a secreted, soluble subunit protein that acts as a specific nucleator, apparently in the absence of an immunoglobulin-like chaperone (nucleated-precipitation assembly) (9, 213; M. Hammar, Z. Bian, and S. Normark, unpublished data); however, the requirement for another type of specialized chaperone has not been ruled out. The formation of curli fibers is therefore similar to the mechanism of amyloid deposition in human disease such as in Alzheimer’s disease and other neurodegenerative diseases (106, 222, 228). Curli will be discussed further in the last section of this chapter.
Type 4 pilus biogenesis also seems to bypass the need for immunoglobulin-like chaperones. Type 4 pili are produced by organisms such as Pseudomonas aeruginosa, pathogenic Neisseria, Moraxella bovis, Dichelobacter nodosus, and Vibrio cholerae (53, 72, 130, 176, 264, 283). They were first recognized and classified as a group based upon their polar location on the cell and their association with a phenomenon known as twitching motility, which appears to mediate cell movement and may be critical for other functions in pathogenesis (221). Type 4 pili are thought to be organelles of attachment important in pathogenesis, although the receptors that they recognize have not yet been elucidated. Unlike adhesive organelles assembled by the chaperone/usher pathway, which are usually composed of different structural subunit types that interact to form a heteropolymeric fiber, type 4 pili are thought to be mostly homopolymeric fibers. This has turned out to be a misperception, however, at least in the case of Neisseria type 4 pili. The adhesive component of these pili was discovered to be PilC and was localized at the tips of the pili (242).
Type 4 pilin subunits have several distinctive features, including a short (6 to 7 amino acids), positively charged leader sequence, a modified amino acid (N-methylphenylalanine) at the amino terminus of the mature pilin, and a highly conserved and hydrophobic amino-terminal domain (18, 91, 233, 234). Another conserved feature of type 4 pilins is a short disulfide-bonded loop at the COOH terminus (283). Insight into the function of some of the different gene products required for type 4 pilus assembly came from studies done by Lory and colleagues in P. aeruginosa (205, 206, 207, 271, 272). These investigators identified PilD as the leader peptidase for the pilin subunit, PilA. PilB encodes a nucleotide-binding protein, and PilC encodes a putative polytopic cytoplasmic membrane protein. Both PilB and PilC were also found to be required for pilin secretion and assembly, but their functions are not well understood (205). Other genes involved in type 4 pilus biogenesis are pilS, pilR, pilT, pilU, pilQ, and pilG. These genes are not linked to the pilA-pilD gene cluster in P. aeruginosa. PilS and PilR are part of a two-component sensor-regulator system which controls transcription of pilA (90). pilT and pilU are adjacent genes encoding nucleotide-binding proteins homologous to PilB; however, mutations in these genes confer a hyperpiliated phenotype (283, 295). pilQ encodes an outer membrane protein, and pilG encodes a homolog of CheY, a protein that is involved in transducing chemotactic responses (46, 177). The pilT, pilU, pilQ, and pilG genes were discovered based on an analysis of mutations that affected fimbrial-associated functions such as phage sensitivity and/or twitching motility. The organization of these genes varies greatly between different species, and the molecular basis of their functions is currently being elucidated.
It appears that conserved features of type 4 pilins and their assembly machinery have been adapted for the assembly of a secretion apparatus (233, 234) as well as for the assembly of type 4 pili (Table 1). For this reason we have termed type 4 pilus assembly part of the general secretion pathway, which is also known as the type II secretion pathway (Table 1). Proteins with similar leader peptides and hydrophobic amino termini have been identified that form part of a secretion apparatus necessary for the secretion of a variety of proteins across the outer membrane (233, 234). These pilin-like proteins are thought to assemble into a pilus-like secretion tube used in the secretion of a variety of proteases, toxins, and other extracellular factors (233). The assembly of this secretion machinery has in many cases been found to depend on homologs of PilB, PilC, and PilD (233).
Type 4 pili have not yet been reported in Salmonella typhimurium (official designation, Salmonella enterica serovar Typhimurium) but have recently been discovered in EPEC and ETEC by Giron et al. (72, 74), who named these structures bundle-forming pili (BFP) and longus pili, respectively. The structural gene encoding the BFP has been identified and characterized (53, 264). Donnenberg and colleagues have recently cloned a gene cluster from EPEC containing 14 genes required for the assembly of type 4 pili (M. Donnenberg, personal communication). Although many of the genes are homologs to the ones described above for P. aeruginosa, others represent new genes. The formation of BFP requires the periplasmic disulfide isomerase DsbA (127). DsbA (19) is also required for the formation of type 4 pili in V. cholerae (226), called toxin-coregulated pili, and it is required for the formation of pili in the chaperone/usher pathway as described later in this chapter (110). A newly identified set of genes necessary for type 4 pilus assembly in E. coli has also been described by Whitchurch and Mattick (296). Longus pili are up to 20 μm long (74). They are composed of a repeating subunit, LngA, that contains the characteristic features of type 4 pilins described above. The presence of lngA is restricted to human ETEC strains. LngA is highly conserved among all ETEC strains and may be an important colonization factor for ETEC. For more information on type 4 pili and related systems, refer to the excellent reviews that have been written on this subject (233, 234, 283).
E. coli can cause gastrointestinal disease by a variety of mechanisms and possesses diverse adherence mechanisms. Adhesive organelles assembled via the chaperone/usher pathway are present throughout all E. coli groups; however, new adhesive mechanisms have also been discovered that apparently do not utilize any of the three pathways mentioned above and will be discussed below.
Histopathological examination of intestinal epithelial cells infected with EPEC reveals the effacement of microvilli and intimate adherence between the bacterium and the epithelial cell membrane (118, 193). Attaching and effacing lesions are characterized by degenerated microvilli and "pedestals" of densely clustered cytoskeletal proteins, including polymerized actin, that protrude from the apical membrane and intimately cup individual bacteria (118, 193). Donnenberg and Kaper have proposed a three-stage model for the formation of attaching and effacing lesions (54, 127). In the first stage, the nonintimate adherence of the bacterium to the epithelial cell is mediated by the BFP and probably other pilus types. The BFP may be most important for bacterial-bacterial interactions that facilitate the formation of a microcolony. The interactions with the host cell may be mediated by attachment organelles assembled via the chaperone/usher pathway including type 1 pili and P pili (these organelles will be discussed extensively later in the chapter). This multifactorial pattern of adherence has been termed localized adherence (73, 248) and is associated with a 60-MDa plasmid, pMAR2 (17). In the second stage, molecular cross-talk between the adherent organism and the epithelial cell leads to a signal transduction event resulting in the effacement of microvilli (127). In the third stage, intimate adherence of the bacterium to the epithelium is mediated by intimin, a 94-kDa protein encoded by a locus that has been termed the eaeA locus (117, 118). The intimate adherence of the bacterium results in the accumulation of filamentous actin and other cytoskeletal proteins within the epithelial cell (62, 140). EaeA was found to be 31% identical to the invasin protein of Yersinia pseudotuberculosis (118). The receptor-binding motifs of both intimin and invasin have been localized to the COOH-terminal regions of the proteins, which are strikingly divergent (64, 156). The homology between the two proteins is mostly in the NH2 terminus (302). Invasin binding to epithelial cells could not be blocked by the addition of intimin but could be blocked by the addition of invasin, suggesting that intimin and invasin bind to different receptors. The eaeA locus is highly conserved in EPEC and is also found in enterohemorrhagic E. coli, which produces very similar attaching and effacing lesions (116, 118). eaeA is part of a 35-kb region of DNA that encodes all of the determinants necessary to produce the attaching and effacing phenotype (127). This region encodes gene products that have striking homology to a recently recognized family of proteins that are involved in the secretion and translocation of virulence determinants in plant and animal pathogens (67, 291). This family includes the products of the Salmonella invA locus, which is described in a separate chapter of this book (67; chapter 151).
ETEC produces at least 12 different organelles of attachment such as CFA/I and CS1 through CS6 (47). These adhesins have different receptor-binding specificities and are associated with different serogroups of ETEC and different patterns of toxin production. Most of these organelles are assembled via the chaperone/ usher pathway. However, the assembly of CS1 pili produced by human enterotoxigenic strains of E. coli may represent a distinct biogenesis system different from the others described above (66, 158, 227, 257) (Table 1). The assembly of CS1 pili involves four gene products, CooA through CooD. CooA is the major subunit, while CooB, CooC, and CooD are all required for the assembly of pili but are not needed for stability or transport of CooA (66, 227, 257). None of these assembly genes has homology with any of the immunoglobulin-like chaperones or outer membrane ushers, and thus they may represent a distinct assembly and transport system for pili. Since CooA is efficiently transported outside the cell but fails to assemble in the absence of CooB, CooC, and CooD, it is possible that the CS1 pilus assembles by a nucleation/precipitation pathway.
Diffusely adherent E. coli derives its name from the ability to adhere to HeLa cells in a diffuse pattern. The adhesin mediating the diffuse adherent pattern has been termed AIDA-I (adhesin involved in diffuse adherence) (27). The AIDA-I adhesin represents another novel biogenesis system of bacterial adhesins (25, 26, 27). Only two genes seem to be required for the expression of the mature adhesin, aidaA and aidaB. aidaB apparently encodes the adhesin and is expressed as a preproprotein. After processing of the signal peptide, it is further processed at a site in its COOH terminus. It is possible that AidaA acts as a chaperone allowing the correct folding and processing of AIDA-I, since alterations of the aidaA gene lead to incorrectly processed or degraded products of the aidaB gene product (25, 26, 27).
In summary, bacteria need to be able to interact with host cells just as people need to be able to interact with their environment. Commensal bacteria have worked out a "relationship" that is beneficial or at least neutral to both the bacterium and the host and can be said to live in symbiosis with the host. Pathogenic bacteria, on the other hand, produce factors which result in damage to the host. Bacterial pathogens produce adhesins that allow bacteria to bind to specific receptors present in the host (23, 136, 155, 208, 209) (also see Table 1). Bacterial adhesins are not invariably assembled into pili, but may be directly associated with the microbial cell surface (non-pilus adhesins; see Table 1) (250) or be assembled into structures that have an amorphous capsular-like morphology. Irrespective of the mode by which these adhesins are presented, their recognition of host receptor structures is extremely fine-tuned, allowing for very selective interactions with the host. Many microbes in nature can thrive in a variety of ecological niches, while others are restricted to a specific microenvironment. The gram-negative organisms E. coli and Salmonella spp. can colonize the intestinal tracts of many mammals and other vertebrates and cause a number of different infections in these hosts (23, 55, 57, 136, 184). Host range, tissue tropism, and target cell specificity demonstrated by a particular microbe are determined at least in part by a stereochemical fit between microbial adhesins and complementary receptor architectures on host cell surfaces (55, 63, 99, 196, 209, 258, 273, 274, 275, 277, 278). To fully appreciate the interplay between a bacterial pathogen and its host, the fine molecular details of each of the cell-type-specific binding events must be dissected. In addition, the consequences of adhesin-receptor interactions, such as the elicitation of host-derived effector molecules, must be elucidated. This chapter contains knowledge that extends from the fine molecular details of pilus-associated adhesin assembly in the microbe to the consequences of receptor recognition at the host-pathogen interface. LT2 strains of S. typhimurium and E. coli K-12 both produce chaperone/usher-dependent fibers called type 1 pili. The extracellular nucleation/precipitation pathway is used by both E. coli K-12 and Salmonella enteritidis (official designation, S. enterica serovar Enteritidis) to produce curli. E. coli does not produce the vast array of adhesive fibers (chaperone dependent and independent) that have been identified in clinical isolates of both E. coli and Salmonella as well as in other gram-negative bacteria (see Table 1). The remainder of this chapter discusses the chaperone/ usher-dependent systems as prototypes to explore the function of bacterial adhesins and their assembly into pili. In addition, the recently discovered nucleation/precipitation pathway used in curli biogenesis is discussed in the last section of this chapter. For a more exhaustive survey of other pilus systems in Salmonella and E. coli, the reader is advised to consult the following reviews and books (23, 48, 99, 103, 127, 136, 215, 223).
Uropathogenic strains of E. coli associated with acute pyelonephritis often express P pili which use the chaperone/usher pathway for their assembly. Genes involved in the biosynthesis and expression of functional P pili are clustered in an operon in the chromosome of only 5 to 10% of human fecal E. coli isolates, but in up to 90% of strains isolated from the urinary tract of children with acute pyelonephritis (98, 119, 172, 229, 255, 276). The two pap gene clusters from the human urinary tract E. coli isolate J96 have been cloned (96) and extensively characterized (Fig. 2) (102, 103, 122, 175, 201, 202). The prs (Pap-related sequence) element was positioned at 94 min on the E. coli chromosome and is part of a so-called "pathogenicity island," which is a large chromosomal element carrying many virulence genes (30). For example, it has recently been shown that the prs-containing element also carries the determinants for α-hemolysin (hly) and the cytotoxic necrotizing factor 1 (cnf1) (30). The pap (pyelonephritis-associated pili) gene cluster was mapped to 64 min and may also be part of a pathogenicity island (D. Berg, personal communication). These islands may represent a mechanism for spread of pathogenicity determinants among and between species of bacteria. This particular region of the E. coli chromosome seems to be a hot spot for the insertion of so-called pathogenicity islands. For example, the locus for effacing E. coli (LEE) is inserted into the E. coli K-12 chromosome at the same site as the pathogenicity island carrying the prs locus and hemolysin genes (127). In each case the insertion was linked to the tRNA gene for selenocysteine (selC) (127; D. Berg, personal communication). Hacker’s group made similar observations for another uropathogenic isolate, strain 536 (81). The two pathogenicity islands in this strain, consisting of 70 kb and 190 kb, carried hemolysin determinants and mapped at 82 and 97 min, respectively, on the chromosome (31). The pathogenicity islands in this strain were also linked to tRNA genes (selC and leuX, respectively). Similar pathogenicity islands may carry pap pilus operons in other clinical strains of E. coli.
The DNA sequence of the entire pap gene cluster has been determined and was shown to encode 11 proteins, each of which has been studied by extensive genetic and biochemical analyses (Fig. 2) (12, 13, 103, 162, 164, 171, 200, 281). Briefly, there are six discrete structural proteins that make up the P pilus, PapA, PapE, PapF, PapG, PapH, and PapK. The pilus is made up of two distinct subassemblies, a thin tip fibrilla joined to the distal end of a thick pilus rod (103, 108, 145). The rod is made up of PapA subunits, and the tip fibrilla is made up mostly of PapE subunits (13, 145, 164). The adhesin, PapG, is joined to the distal end of the fibrilla via the PapF adaptor, and the fibrilla is joined to the rod by the PapK adaptor (108, 145). PapH links the pilus rod to the cell membrane and in doing so terminates biogenesis of the rod (12). PapD is a periplasmic chaperone and PapC is an outer membrane usher. Both are required for the assembly of pili but are not components of the final structures (52, 103, 162, 200). The role of PapJ is unclear, but it has been suggested to function as a cochaperone (282). The structure and function of each of the pap gene products will be discussed in detail throughout this chapter and are illustrated in the model in Fig. 2.
The regulatory circuitry responsible for pilus expression requires two genes (papI, papB) that are located upstream of the mapped pap promoter, as well as two global regulatory proteins, CRP and LRP (11, 33, 34, 35, 77, 78, 199, 289). Regulation of pilus expression, which involves phase variation, is discussed in further detail in chapter 11 of this book. It has been suggested that phase variation of P and type 1 pili contributes to the pathogenesis of E. coli, perhaps by allowing the bacteria to evade the immune response (56, 103).
Chaperone/usher-dependent pilus systems have gene clusters that are functionally related and possess genes that bear sequence homology and are similarly located in the operon. For example, the major pilus subunit gene is typically located immediately downstream from the regulatory genes. Genes encoding subunits that are minor components of the pilus or part of tip fibrillae are often located at the distal end of the operon (108, 137, 145, 189, 243, 251). The chaperone- and usher-encoding genes are typically located in the middle of the operon. The pap operon is presented in Fig. 2. In the Pap system, chaperone and usher functions have been defined and assigned to PapD and PapC, respectively (52, 100, 103, 146). In the type 1 system, FimC is the immunoglobulin-like chaperone and FimD is the usher (123, 135; P. Klemm, unpublished data).
In the early 1980s, several groups demonstrated that P pili bound to Galα(1, 2, 3, 4)Gal moieties present in the globoseries of glycolipids on uroepithelial cells and erythrocytes (32, 125, 155). Pretreatment of bacteria with Galα(1, 2, 3, 4)Gal containing saccharides inhibited hemagglutination of human erythrocytes and binding of P-piliated bacteria to uroepithelial tissue (32, 155). There are three known alleles of papG, referred to as classes I, II, and III. Classes I and II were cloned and sequenced from a clinical urinary tract infection (UTI) isolate, J96. The class III adhesin was isolated from two human UTI isolates as well as a canine UTI isolate. Class I PapG agglutinated only human erythrocytes, while class II PapG agglutinated human erythrocytes very well and sheep erythrocytes only poorly. The class III PapG agglutinated only sheep erythrocytes. Lund et al. demonstrated that the binding specificity of a P pilus changed from mediating the agglutination of human erythrocytes to mediating the agglutination of sheep erythrocytes when the so-called class I PapG (from the pap operon) was replaced with the class III PapG (from the prs operon) (171, 172). Complementation of a papG – P-pilus operon with prsG in trans changed the binding specificity of the class I Pap pilus to that of the class III Prs pilus. This experiment argued that the receptor-binding specificity and hence the tropism of the organism were determined by the class of PapG adhesin present at the tip of the pilus (171, 172).
The Galα(1, 2, 3, 4)Gal moiety is a component of the globoseries of glycolipids expressed on erythrocytes and many other tissues, including the uroepithelium of the kidney and urinary tract. The disaccharide moiety is linked by a β-glucose residue to a ceramide group that anchors the receptor in the membrane (83). This minimum receptor isotype is called globotriasylceramide (GbO3). The various members of this receptor family differ by the addition of sugar residues distal to the Galα(1, 2, 3, 4)Gal disaccharide present in GbO3 (Fig. 2). The addition of a single GalNAc sugar to GbO3 creates globoside, also called globotetraosylceramide (GbO4), whereas the addition of two GalNAc sugars to GbO3 creates the Forssman antigen (GbO5) (128). GbO3 and GbO4 are abundant in human kidney but not in canine kidney. In contrast, GbO5 is absent in the human kidney but is abundant in the canine kidney (273, 274, 275).
Stromberg et al. tested the ability of each of the three PapG alleles to bind to uroepithelium from humans and dogs (Table 2). They found that only the class III adhesin bound to canine (MDCK) uroepithelium, which contains GbO5. Neither class I nor class II PapG bound to MDCK cells. In contrast, Stromberg et al. found that only the class II adhesin bound well to human uroepithelium, which contains GbO3 and GbO4 (T24 cells, bladder carcinoma) (274). These results suggested that class III PapG recognizes the Forssman antigen (GbO5), whereas the class II PapG recognizes either GbO3, GbO4, or both (274). To examine the specificity of the different PapG adhesins further, Stromberg et al. (274) separated the different isoreceptors by thin-layer chromatography and tested the ability of metabolically 35S-labeled bacteria expressing the different PapG adhesins to bind to the various isoreceptors present on the chromatogram (Table 2). The class III adhesin bound only GbO5, consistent with its specificity for canine uroepithelium. Surprisingly, the class I and class II adhesins bound efficiently to all three glycolipids even though bacteria expressing these adhesins were unable to bind to canine MDCK cells containing GbO5 (275). The ability of the class II adhesin to bind GbO5 when separated on a chromatogram but not when present in an epithelial membrane could be due to a poor presentation of the GalαGal epitope in GbO5 when in the membrane. In contrast, the Galα(1, 2, 3, 4)Gal epitope is appropriately presented in GbO4 present in membranes such as in human kidney tissue. In support of this view, Striker et al. found that P-piliated E. coli containing the class II adhesin bound to human kidney tissue in situ and that soluble Galα(1, 2, 3, 4)Gal analogs blocked this binding (270). In another competition assay, binding of the class III adhesin to immobilized GbO5 could not be inhibited by free Galα(1, 2, 3, 4)Gal (275), but binding of the class I adhesin for immobilized GbO3 was completely blocked by the Galα(1, 2, 3, 4)Gal disaccharide. This result suggested that the class III adhesin recognizes a more complex epitope than the Galα(1, 2, 3, 4)Gal disaccharide. Binding and inhibitor studies with a panel of glycolipids demonstrated that, in addition to the disaccharide, a distal GalNAc β residue was required for efficient binding (275). The additional GalNAc residue present on GbO5 (compared to GbO4) may result in a change of conformation of the isoreceptor in the membrane, making the Galα(1, 2, 3, 4)Gal receptor unavailable or less available for binding by the class II adhesin. Consistent with this hypothesis, Stromberg et al. (275) showed by molecular modeling that the orientation of Galα(1, 2, 3, 4)Gal in the receptor rotated from parallel to perpendicular in relationship to the membrane as sugar groups were added to the terminal end of the disaccharide (see Fig. 2).
Table 2Binding specificities of the three classes of PapG |
The receptor-binding domain of PapG has been shown previously to reside in the amino terminus of the protein (101). Haslam et al. fused the receptor-binding domains of the class I, class II, and class III adhesins to the maltose-binding protein to investigate whether this domain was solely responsible for the fine recognition specificities observed with whole bacteria as described above (88). Haslam et al. tested the ability of each of the purified fusion proteins as well as purified pili containing each of the three PapG alleles to bind to the three receptor isotypes. This study verified the work by Hultgren et al. (101) in which it was shown that the amino terminus of PapG contained the sugar-binding domain. Both the class I fusion protein and the pilus-associated class I adhesin bound to GbO3, GbO4, and GbO5. Similarly, both the class III fusion protein and the pilus-associated class III adhesin bound only to the GbO5 receptor. This confirmed that the PapG protein endows the bacteria with the specificity for binding to a given isoreceptor in tissue. Both the class II fusion protein and the pilus-associated class II adhesin bound well to GbO4 and poorly to GbO5; however, the class II fusion protein, but not the pilus-associated class II adhesin, bound well to GbO3. This finding suggests that other determinants in the pilus, such as the minor pilins, can affect the fine specificity of binding (88).
To identify the functional groups on the Galα(1, 2, 3, 4)Gal disaccharide required for recognition by the PapG adhesin, a panel of galabioside analogs, in which the functional groups on the disaccharide were replaced with hydrogen, fluorine, or a methoxy group, were tested as inhibitors of hemagglutination (133). Derivatives containing substitutions of OH groups -6, -2', -3', -4', and -6' were found to be poor inhibitors, suggesting that the PapG adhesin recognized the galabiose portion of the globoseries of glycolipids by hydrogen bonding to five oxygen atoms situated on an edge of the disaccharide (shown in red in Fig. 2). Hultgren et al. (101) verified this finding by showing that purified class I adhesin in a complex with the PapD chaperone formed the same hydrogen bonds to the Galα(1, 2, 3, 4)Gal receptor as did PapG present in pili (chaperone-subunit complexes are discussed later in the chapter).
Striker et al. used a panel of receptor analogs in inhibition studies and found that the class II PapG uses relatively the same polar ridge as the class I PapG in interacting with the GbO4 isoreceptor (270). However, in addition to binding to the polar ridge on the digalactoside, the class II adhesin also made specific interactions with the distal GalNAc group in GbO4 and, more importantly, with the glucose residue proximal to the Galα(1, 2, 3, 4)Gal disaccharide (Fig. 3). The interaction that the class II PapG makes with the proximal glucose is required for strong binding to GbO4. This glucose is probably presented poorly by GbO3 due to its presentation with respect to the membrane plane. In addition to the the Galα(1, 2, 3, 4)Gal disaccharide, the correct orientation of the glucose residue in GbO4 is probably important in determining the tropism of the class II PapG for human kidney (270).
The in vivo significance of isoreceptor variants of the PapG adhesin comes from epidemiological data as well as from recent infection studies in cynomolgus monkeys. E. coli isolates from stool samples or urine from dogs frequently express P pili carrying a class III adhesin (175, 274). This correlates nicely with the Forssman antigen (GbO5) rather than GbO4 being the dominating Galα(1, 2, 3, 4)Gal isoreceptor in the canine intestinal wall as well as in the canine kidney. In humans, fecal isolates infrequently express either P pili with a class II or a class III adhesin (5% and 1% of the total E. coli flora, respectively). This could be due to the lack of a selection force, since Galα(1, 2, 3, 4)Gal-containing glycolipids are absent in the epithelial cells facing the gut lumen. The same situation prevails in the cynomolgus intestine, explaining why a mutant lacking the PapG class II adhesin has no effect on the presence of the organisms in the fecal flora (299a).
In humans, both class II and class III PapG adhesins are associated with UTI. Interestingly, cystitis isolates of E. coli expressing P pili with a class III PapG preferentially occur in blood group A1 secretor-positive individuals who contain globoA on their uroepithelial cells (166). GloboA is a glycolipid that, like the Forssman antigen, carries a terminal extension. Thus, in globoA, the Galα(1, 2, 3, 4)Gal(1, 2, 3, 4)Glc trisaccharide is most likely oriented unfavorably with respect to the membrane for recognition by the class II PapG. In contrast, class III PapG, which is the adhesin most commonly associated with group A1 secretor-positive individuals, recognizes an exposed surface of this receptor isotype. In the cynomolgus monkey, it has recently been shown that GbO4 is the dominating Galα((1, 2, 3, 4))Gal-containing isoreceptor in the kidney, ureter, and bladder and that both the Forssman antigen and globoA are lacking. This finding correlates well with the discovery that replacing the class II allele in the pyelonephritic strain DS17 with a class III papG allele abolished the ability of this organism to give rise to acute pyelonephritis in the primate model (see below).
The initiation and persistence of many bacterial infections such as UTI is thought to require the presentation of adhesins on the surface of the microbe in accessible configurations which promote binding events that dictate whether extracellular colonization, internalization, or other cellular responses will occur (23, 24, 55, 99, 136, 184, 209, 276, 278, 280). E. coli is the most common bacterial species causing UTI in humans. Such infections may manifest themselves as asymptomatic bacteriuria, acute cystitis, or most severely as acute pyelonephritis (173, 279, 280). Strains associated with acute pyelonephritis often express proposed virulence factors such as P, type 1 and S pili, hemolysin, and iron-chelating aerobactin and are usually resistant to serum bactericidal activity (120). P piliation is probably not the only virulence factor important in the etiology of acute pyelonephritis. However, it is the property that is epidemiologically best correlated with the disease in that it is reported in over 95% of children and 50 to 90% of adults (126, 151, 173, 279, 280, 287). Moreover, GbO4 is abundant in the human kidney and contains the Galα(1, 2, 3, 4)Gal receptor recognized by the P-pilus adhesin (37, 155).
It has been a paradigm for several years that, in order to establish itself in the human kidney during a nonobstructive ascending UTI, E. coli needs to express adhesins recognizing epithelial cell surface receptors in the ureter and kidney. Since most pyelonephritic E. coli isolates express P pili recognizing GbO4 and other glycosphingolipids that are abundant in human kidney, it has been thought that an interaction between P pili and the Galα(1, 2, 3, 4)Gal receptor is needed for acute pyelonephritis to take place (155), whereas type 1 pili appear to be more commonly expressed in E. coli causing cystitis (104, 105).
P-piliated E. coli strains have been shown to cause pyelonephritis in the normal urinary tract of cynomolgus monkeys (239, 240). A 1-bp deletion after codon 37 in papG was created in the uropathogenic strain DS17, resulting in strain DS17–8, which produced P pili lacking the PapG tip adhesin (241). This mutant strain remained P piliated, but was unable to bind the GbO4 receptor in vitro and was unable to bind to kidney tissue from humans or cynomolgus monkeys in situ. The virulence of DS17 and its papG mutant, DS17–8, was compared in the cynomolgus monkey (241). Monkeys receiving wild-type strain DS17 had a mean bacteriuria of 21 days, compared to 6.8 days for those receiving the mutant strain DS17–8. Renal clearance and renal function were also significantly reduced in the monkeys receiving the wild-type strain, but not in those receiving the isogenic papG mutant strain. Pathologic evaluation of the infected kidney confirmed the functional studies, showing significantly less inflammation and fewer pathologic changes in the mutant group as compared to those receiving the wild type. From these studies it was concluded that the PapG adhesin at the tip of P pili is required for pyelonephritis to occur in the normal urinary tract of primates (241).
Interestingly, in the experiments described above there was no evidence that the PapG adhesin was required for colonization of the lower urinary tract or for the development of acute cystitis, one of the most prevalent types of infections in females. Both strains colonized the vagina, persisted in the intestine, and caused bladder infection (241). Thus, although PapG was a critical virulence determinant in causing pyelonephritis, it did not appear to be crucial in cystitis. However, when both wild type (DS17) and mutant (DS17–8) were inoculated simultaneously into the bladder of cynomolgus monkeys, the mutant strain was outcompeted within a few days by the wild-type strain expressing PapG (299a). Thus, in the bladder the PapG adhesin provides E. coli with a competitive advantage. The wild-type strain, however, had no competitive advantage over the nonadhesive mutant in either the vaginal area or intestine, even after several weeks. The following scenario is possible: the vaginal area is colonized from the intestine by a mixture of clones leading to a mixed inoculation of the bladder. P-piliated clones of E. coli in the bladder inoculum might outcompete other, less adhesive clones, explaining the high percentage of P-piliated E. coli in first-time cystitis (160). It is possible that other E. coli adhesins, such as the sialic acid-specific S-adhesin (82) and the mannose-binding FimH adhesin of type 1 pili (see below), may also confer a competitive advantage to a pathogen in the bladder. In a CBA mouse model, a mutant unable to express P pili remained capable of causing pyelonephritis and acute pyelitis (188). It is possible that the lost interaction with the globoseries of glycolipids was compensated for by the expression of other adhesive determinants, such as S or type 1 pili, in this strain. Colonization of the urinary tract is a multifactorial process. Uropathogenic E. coli often produce a multitude of different adhesins that in some cases are probably able to compensate for each other depending on the composition of the glycolipids and glycoproteins in the uroepithelium of the host that is being colonized.
Type 1 pili bear a mannoside-binding determinant which has been shown to be encoded by the fimH (pilE) gene (144, 180, 185). Abraham et al. demonstrated that the FimH adhesin was serologically conserved throughout the Enterobacteriaceae (4). A number of host tissues have been shown to be targets for type 1-fimbriated bacteria. Mannose-sensitive binding to human buccal cells, proximal tubular cells of the kidney, epithelial cells in the bladder, lung, and intestine, and various inflammatory cells has been demonstrated (10, 132, 245, 284). More specific targets for the FimH adhesin have been identified, including a 65-kDa protein of guinea pig erythrocytes (70), the leukocyte adhesin molecules CD11 and CD18, and the Tamm-Horsfall glycoprotein of human urine (10). Type 1-piliated bacteria have been shown to interact with the extracellular matrix components laminin (148) and fibronectin (265). The FimH adhesin of the type 1 pilus seems to represent a family of proteins which bind to carbohydrate as well as proteinaceous targets (266). With regard to the role of type 1 pili in promoting E. coli extraintestinal colonization, a good clinical correlation exists between type 1 piliation and the potential for causing cystitis and urethritis (215). Type 1 pili mediate binding to vaginal mucus, which may influence the initial attachment and subsequent colonization of the vaginal and urinary tract epithelium by E. coli (292). This relationship has been supported by a variety of genetic and biochemical tests of colonization in experimental animals (104, 132). Furthermore, Abraham et al. (1) demonstrated protection of experimentally infected mice with anti-type 1 fimbriae or anti-mannose receptor antibodies. Therefore, blocking colonization of type 1-piliated organisms prevented cystitis in the murine model. Several studies have also shown that type 1 pili aid in protecting E. coli from phagocytic killing (131, 284) even though the pili promote binding to a variety of phagocytic cells (284). Bloch et al. (28) have recently shown a role for type 1 pili in oropharyngeal colonization in neonatal rats and transmission of E. coli K1 among littermates. It has been suggested that type 1 binding may disrupt or bypass the normal lytic pathway in the phagocyte, resulting in reduced killing, and that this protection is related specifically to the mannose-binding adhesin, FimH (215).
Salmonellosis in mice, however, does not seem to require the expression of type 1 pili since a deletion in the S. typhimurium fim gene cluster actually decreased the 50% lethal dose after peroral inoculation (168). However, when this fim deletion was combined with a fla mutation which abolished flagellar expression, a mutation that by itself had no effect on virulence, the resulting double mutant was markedly attenuated in mice. This result points to the multiple factors at play in bacterial pathogenesis and how difficult it is to directly demonstrate a need for a particular adhesin in animal studies.
Bacterial adhesins also seem to play important roles as modulators of inflammation through specific activation of host effector cells, but little is known about their mechanism of action. It has, for example, been demonstrated that binding of E. coli, expressing P pili, to a kidney cell line activates production of interleukins 6 and 8 (5, 49, 89). Purified P pili evoke the same response when they carry the tip adhesin, but not when the adhesin is removed by detergent (89). Cytokine secretion, in response to the binding event, may result in a global host response that is responsible for much of the pathology associated with bacterial infection in the urinary tract. The capacity of type 1 pili to activate a variety of immune and inflammatory cells including B lymphocytes and neutrophils has been demonstrated (284). The activation of these cells results in a variety of biological responses including cell proliferation and secretion of various biologically active compounds. Activation of inflammatory cells by bacterial pilus adhesins can result in the release of inflammatory mediators, some of which are harmful to the host whereas others can affect bacterial survival in the host.
It is not yet known whether these in vitro signaling effects induced by adhesins play a role in disease. Recent experiments in cynomolgus monkeys, however, suggest this may be the case. When DS17, a clinical isolate of E. coli expressing PapG-containing P pili, was given repeatedly in the bladder of monkeys, it resulted in resistance to repeated infections. As described above, the isogenic PapG mutant derivative of this strain, DS17–8, abolished the activity of the strain to cause pyelonephritis, but not cystitis (241). DS17 induced resistance to the isogenic mutant strain DS17–8 (J. Winberg, unpublished data). Interestingly, the PapG– mutant did not mediate resistance to subsequent challenge by either the mutant or the wild-type strain (Winberg, unpublished data). The basis for this PapG-mediated resistance is not known but could be due to an effect on proinflammatory cytokine production.
Hultgren et al. (101) utilized Galα(1, 2, 3, 4)Gal, the target of the PapG adhesin (described above), linked to Sepharose beads as an affinity matrix to purify a PapD-PapG (chaperone-adhesin) complex. By this strategy, a 1:1 PapD-PapG complex was purified from the periplasmic space of E. coli (101). Deletion of the last 13 residues of PapG, the putative PapD-binding site (discussed later in this chapter), was found to destabilize the protein, resulting in the appearance of several stable PapG truncates. Similarly, expression of PapG in the absence of the PapD chaperone resulted in the proteolytic cleavage of PapG into some of the same protein truncates. The COOH-terminal region of PapG was later shown to contain a chaperone recognition motif (discussed below). One PapG truncate, 24.5 kDa in size, was found to bind the Galα(1, 2, 3, 4)Gal-Sepharose column. This truncate contained the amino-terminal 150 residues of PapG, and therefore it was postulated that the carbohydrate-binding domain was located in the amino-terminal region of the protein. The fact that PapD did not copurify with the 24.5-kDa truncate suggested that the PapD recognition site had been deleted. Therefore, the PapG adhesin seems to have a two-domain structure (Fig. 4). The amino-terminal domain contains the receptor-binding site, while the COOH-terminal domain contains the chaperone recognition motif and the characteristic disulfide bond found in all pilins assembled into rodlike fibers by immunoglobulin-like chaperones. The PapG–maltose-binding protein fusion proteins described by Haslam et al. (88) confirmed that the amino-terminal domain of PapG contains the carbohydrate-binding region of the protein.
Like PapG, the type 1 pilus adhesin, FimH, is susceptible to proteolytic cleavage when expressed in wild-type E. coli lacking a functional FimC (periplasmic chaperone) (123, 124). The FimH protein is cleaved into a ∼16-kDa truncate. Amino-terminal sequencing revealed that the truncate has the same amino terminus as full-length FimH (124). Mannose-Sepharose affinity chromatography was used to purify the FimH truncate from the periplasm (124). These results suggested that, as for PapG, the carbohydrate-binding region of FimH was located in the amino-terminal domain of the protein. FimH, like all pilins, has a conserved sequence at the COOH terminus required for an interaction with the periplasmic chaperone (this will be discussed later in the chapter).
It appears that the adhesive determinant of the enteropathogenic E. coli-associated K88 and K99 pili is the major subunit protein (112, 113). Mutagenesis studies in the K99 system, targeting positively charged residues in the major subunit protein, revealed the role of the major subunit in receptor binding (112). In addition, mutation of the penultimate tyrosine to a stop codon or to serine or glutamic acid in the major pilin protein, FanC of the K99 pilus, resulted in a bald phenotype (260). These mutations abolished the ability of the chaperone to bind to the subunits, resulting in their proteolytic degradation (260). Later in this chapter, the structural basis for chaperone-subunit interactions will be discussed.
A combination of genetics and immunogold electron microscopy originally revealed the presence of the adhesin and other minor components exclusively at the tip of the P pilus and that the sugar-binding adhesin was separable from the major pilus rod (94, 161, 201, 285). Using quick-freeze deep-etch electron microscopy, it was discovered that P pili are composite fibers consisting of flexible adhesive fibrillae joined end-to-end to pilus rods (145) (Fig. 2). The pilus rod is composed of repeating PapA subunits packed into a right-handed helical cylinder (40, 76). Tip fibrillae that extend from the distal ends of each pilus rod are composed mostly of repeating subunits of PapE arranged in an open helical conformation (145). The PapG adhesin has been localized to the tips of fibrillae that are seemingly flexible, since they often appear to be bent in various orientations on electron micrographs (145). The major component of the tip fibrilla, PapE, may facilitate interactions with extracellular matrix proteins such as fibronectin, which would make the entire fibrillar structure a multifunctional virulence determinant (294). The composite architecture of the P-pilus fiber reveals the strategy used by pyelonephritic E. coli to present the PapG adhesin to eukaryotic receptors. The rigid PapA rod distances the adhesin from lipopolysaccharide and other potentially interfering components at the bacterial cell surface, while the flexible fibrilla allows PapG steric freedom to recognize and bind to the digalactoside moiety on the uroepithelium.
A three-dimensional reconstruction of the P pilus was recently reported by Bullitt and Makowski (40). Similar to the results of previous X-ray diffraction studies (76), the pilus was found to be right-handed, 68 Å (1 Å = 0.1 nm) in diameter, and approximately 1 μm in length and contained 3.28 subunits per turn. The pilus rod also has a 15-Å helical cavity winding through the rod and communicating with the external environment by a set of radial channels. The current three-dimensional reconstruction allowed the visualization of a PapA subunit as a pair of globular domains (40; E. Bullitt, personal communication). Although the boundaries between subunits were not clearly defined, the authors inferred from the reconstruction that PapA subunits make at least four interactions in the rod. Two of the interactions are best described as head-to-tail between neighboring subunits and define the fibrillar PapA polymer. The other two interactions are involved in coiling the polymer to form the right-handed helical rod. Each subunit interacts with a residue three subunits ahead (n+3) and with a subunit three subunits behind (n–3). Therefore, the n ±3 interactions, minimally, result in the transition from the thin fibrillar conformation of the PapA polymer to the right-handed helical rod. It has been shown that P-piliated bacteria are resistant to displacement from the urinary tract by fluid flow across the uroepithelium (219). Bullitt and Makowski (40) suggest that these forces perturb the n+3 and n–3 interactions, instead of causing the pilus to break. Disruption of the n+3 and n–3 interactions would result in the pilus reverting to a linear polymer conformation. A "bungee cord" model was presented to account for this resistance to shear force as well as the observation of stretched-out pili and sharp bends in pili when viewed by electron microscopy following normal preparation protocols. Interestingly, the extended segments of the PapA polymer resembled tip fibrillae; they had an open helical structure and dimensions similar to tip fibrillae. This model suggests that the transition from helicoidal to fibrillar conformation of the PapA polymer provides a degree of flexibility to the pilus and allows bacteria expressing the pilus to better adhere and colonize tissues in the urinary tract (40).
The assembly of pili by gram-negative bacteria seems to be a highly controlled process, ensuring that every pilus has virtually the same composition and structure. The correct incorporation of the various components into the pilus is dictated by stereochemical fits between the complementary surfaces on each subunit type. In the P pilus, two minor pilus components, PapF and PapK, are crucial to ensure that the assembly of the adhesive tip fibrilla precedes that of the pilus rod and that the substructures of the pilus are firmly connected (108). PapF and PapK were shown to be adaptor proteins. PapF is required to link the PapG adhesin to the distal end of the tip fibrilla, and PapK joins the tip fibrilla to the pilus rod (108). Additionally, in an F– mutant, only 20% of the wild-type piliation level was obtained, and a papF papK double mutation abolished the ability of the cells to produce pili (108, 161, 163, 164). This result supports the hypothesis that these adaptor proteins also have crucial functions in pilus formation. Only the expression of PapK in trans was able to initiate rod polymerization in the absence of the other genes encoding tip proteins (108). Furthermore, in a K– strain, tip fibrillae grew to be five times longer than in wild-type cells (145). Apparently PapA subunits were able to associate with tip fibrillae terminated by PapE, but with a low efficiency, resulting in the incorporation of a larger number of PapE subunits before PapA polymerization was initiated. Overproduction of PapK shifted the distribution of tip fibrillar lengths so that most were significantly shorter than wild type, probably by increasing the frequency by which a PapK subunit was incorporated. PapK incorporation into the growing tip fibrilla terminates its growth and initiates the formation of the pilus rod. This dual function is consistent with its deduced adaptor position between the tip fibrilla and pilus rod. Similarly, the adaptor function of PapF may be linked to an initiation mechanism that induces the polymerization of PapE subunits into the tip fibrilla.
Type 1 pili are approximately 7 nm in width and range from 0.2 to 2.0 μm in length. The type 1 rod, composed of the FimA major subunit, is arranged in a right-handed helix with 3.125 residues per turn and a helical symmetry identical to that of the P pilus (38). Abraham et al. (3), several years prior to Bullitt and Makowski’s study with P pili, utilized 50% glycerol to alter the helical conformation of the type 1 pilus. The observed unraveling of the helical conformation of the pilus was reversible following dialysis of the glycerol. Similar to the findings with the P pilus, the treated type 1 pili had regions of localized unraveling, revealing extended regions of an approximately 2-nm polymer.
The adhesive moiety of type 1 pili was reported to be the FimH protein, a separate entity from the FimA rod (2, 137, 144, 180, 181, 185), and was purified by several groups (85, 86, 284). As shown in Fig. 2, we have demonstrated that type 1 pili, like P pili, are heteropolymeric structures (124). Mannose-Sepharose affinity chromatography of extracts from fimA–, mannose-sensitive hemagglutination-positive cells was used to purify the adhesive moiety from the surface of E. coli (124). The adhesive moiety of the fimA – bacteria consisted of a FimG-FimH complex that, when examined by electron microscopy at high resolution, had a thin fibrillar architecture. Examination of purified type 1 pili by high-resolution electron microscopy revealed fibrillar structures present at the distal ends of the pilus rods that had an architecture identical to the FimG-FimH moiety. In the absence of FimH there was a significant decrease in the length of the fibrillae present at the distal ends of the type 1 pilus rods (16 nm to 3 nm) (124), arguing that FimH is present in more than one copy in the tip fibrilla. Alternatively, the fibrilla could represent a short FimG-FimH heteropolymer, or FimH could serve as a nucleator to induce the formation of the bulk of the fibrillar structure. FimF may also be a component of the fibrilla, although this has not yet been shown directly (124). Previous genetic studies have indicated a role for FimF and FimG in biogenesis of the type 1 rod (137, 243). Analogous to the studies on P-pilus biogenesis in which initiator/adaptor proteins were found to be components of the P fibrilla (discussed above), localization of the minor type 1 pilins FimG, and possibly FimF, to the tip is consistent with a role for these molecules in type 1 pilus biogenesis (124). The above findings do not rule out additional locations of the minor subunits or the FimH adhesin in the pilus structure. A previous immunoelectron microscopic analysis localized the FimH adhesin in lateral sites along the pilus shaft (4, 231).
The proposed role of the P-pilus tip fibrilla is to present the PapG adhesin in a flexible location in order to enhance the interaction of the adhesin with host receptors (108, 145). Since the major component of the P fibrilla is PapE, which self-associates into an open helical fiber, the stubbiness and apparent lack of flexibility of the type 1 tip fibrilla suggest that it does not have a PapE homolog. Type 1 fibrillae have an architecture that resembles that of papE mutants of P pili. Coexpression of the papE gene with the type 1 tip fibrilla genes did not alter the morphology of the type 1 tip fibrillae (C. H. Jones and S. J. Hultgren, unpublished data). Incorporation of PapE into the type 1 fibrilla would require an interaction with FimC, the type 1 chaperone, and would require a PapE-subunit interaction with type 1 subunits. Apparently, PapE does not fulfill these criteria, as it was not polymerized into the type 1 tip fibrilla (unpublished).
The third pilus type commonly associated with uropathogenic E. coli is the S pilus (148), which binds α-sialyl-β-2,3-galactoside (141). The S pilus has three minor pilin genes, sfaG, sfaS, and sfaH, in addition to the major pilin gene, sfaA (82, 189, 251, 252). Remarkably, the small pilin, SfaS (12 kDa), has been identified as the adhesin of the S pilus. The purified adhesin, SfaS, was found to be in a large (>106 molecular weight) multivalent complex that would hemagglutinate human erythrocytes (189). Similar to P and type 1 pili, immunoelectron microscopy was used to show that SfaS is localized to the tip of the S-pilus rod (189). Schmoll et al. (251) constructed null mutants of the three minor pilins and tested the effect of the mutations on the assembly of adhesive pili. Mutations in all three minor pilins reduced hemagglutination; however, only sfaS mutants were completely hemagglutination negative. Both sfaS and sfaH mutants affected the level of piliation, while sfaG mutants had a minor effect on hemagglutination and no effect on the level of piliation. Therefore, SfaG and SfaH may represent linkers or adaptors for assembly of the SfaS adhesin into the pilus (251). Recent site-directed mutagenesis of the SfaS protein has revealed that a highly positively charged region of the adhesin may represent the sialic acid-binding site (194). Interestingly, the major subunit antigen, SfaA, as opposed to SfaS, seems to be important in the binding of S-piliated bacteria to brain endothelial cells (232). This result suggests that the entire S pilus is a multifunctional adherence determinant. Work in the Hultgren laboratory demonstrated that the S pilus is a heteropolymeric structure with a tip fibrilla similar to that seen on P pili (see Fig. 2) (S. J. Hultgren, unpublished data). A reexamination of the S-pilin mutations in light of the bipartite S-pilus structure will help to clarify the role of these proteins in pilus and fibrilla biogenesis.
Molecular chaperones are found, in the cytoplasm of bacteria and in various cellular compartments in eukaryotes, to maintain proteins in nonnative conformations that permit their secretion across membranes or assembly into oligomeric structures, as exemplified in other chapters of this book and discussed elsewhere (45, 58, 59, 60, 69). The pilus biogenesis pathway provides an excellent model to understand the requirement for molecular chaperones in the periplasm of gram-negative bacteria involved in postsecretional folding and the assembly of cellular organelles. The exposure of interactive surfaces of protein protomers at the wrong time during intermediate stages of postsecretional assembly could cause biologically nonproductive interactions that lead to kinetically dead-end pathways and aggregation. However, when these surfaces are protected by a chaperone, the subunits are stabilized in assembly-competent states. We will discuss PapD as the prototype member of the class of chaperones required for pilus assembly in the chaperone/usher pathway. PapD is the periplasmic chaperone required for the assembly of P pili.
The three-dimensional structure of the PapD periplasmic chaperone has been solved by Holmgren and Brändén (92). PapD consists of two globular domains oriented in the shape of a boomerang (see the ribbon model in Fig. 5). Each domain is a β-barrel structure formed by two antiparallel β-pleated sheets that have a topology similar to an immunoglobulin fold (36, 100, 299). A large family of proteins that use the immunoglobulin fold as a basic structural motif in eukaryotes are grouped into the immunoglobulin superfamily. This protein family includes antibodies, cell surface adhesion molecules, the T-cell receptors, and the human growth hormone receptor (22, 44, 50, 299).
The basic structure of an immunoglobulin domain is best described as two antiparallel β-sheets packed tightly against each other to form a hydrophobic core (230, 249, 299). This protein fold is generally referred to as the β-barrel motif (36). Immunoglobulin constant domains (heavy and light) contain seven β-strands arranged in two β-sheets pinned together by a disulfide bond. An immunoglobulin variable domain, however, differs since it contains two additional β-strands as well as a different order of the strands in the sheets. The stretches of amino acid residues that link adjacent β-strands are positioned at the ends of the β-barrel; these regions vary highly in length and sequence (36). The association of two variable domains (heavy and light chain) brings the three hypervariable regions of each domain together, forming at one end of the β-barrel a large, flat surface: the antibody combining site (7). The immunoglobulin domain structure provides a stable platform for the display of specific recognition surfaces formed by the loops connecting the β-strands as well as by sequences located on the outer faces of the β-sheets (157, 299).
The topology of each of the two domains of PapD, i.e., the order of the β-strands in the β-barrel, differs from the classic immunoglobulin fold due to strand switching at the edges of the sheets (see Fig. 5) (92). The C-terminal domain, domain 2, has structural features analogous to domain 2 of the human immunodeficiency virus receptor, CD4, which is a variation of the classic immunoglobulin fold (244). The C-terminal domain of PapD also has an extra short strand in the lower sheet which is disulfide bonded to the strand preceding it (92). PapD, however, lacks the intersheet disulfide bonds normally seen in constant domains (36). Domain 1 of PapD is most similar to the variable domain of an immunoglobulin (Fig. 5) (92).
To date, 24 different adhesive structures in gram-negative bacteria have been identified that require a PapD-like chaperone for their assembly (Table 1). Sixteen of these adhesive structures are rodlike fibers that have a diameter between 2 and 10 nm. The other eight adhesins assembled by PapD-like chaperones are either nonfimbrial structures that have not yet been visualized by electron microscopy or atypical organelles with an amorphous morphology such as proteinaceous capsules (Table 1).
All 24 periplasmic chaperones are probably very similar in structure: they are similar in length, generally positively charged, and 25 to 68% identical and 48 to 82% similar in amino acid sequence (93; D. Hung, S. Knight, and S. J. Hultgren, unpublished data). Most of the conserved amino acids are concentrated in β-strands in the cleft region between the domains and are required to maintain the overall immunoglobulin-like structure of the domains (93). Some of the invariant residues occupy critical points in loops or are involved in intramolecular interactions which serve to orient some of the loops. Another invariant feature of the chaperone superfamily is an internal salt bridge formed by Asp-196 and Arg-116 in association with Glu-83. This invariant interaction probably serves to orient the two domains towards one another to create the cleft region between the domains. Interestingly, there exists a group of surface-exposed residues that have no apparent structural function, since their side chains are oriented towards the solvent without making any specific interactions with other side chains, yet these residues are conserved or invariant in the entire chaperone superfamily (93; Hung et al., unpublished data). The side chains of two positively charged residues located deep in the chaperone cleft are invariant in every member of the chaperone family (see space-filling model of PapD in Fig. 2). Site-directed mutations in the invariant Arg-8 and Lys-112 cleft residues abolished the ability of PapD to bind pilus subunits and assemble them into pili, arguing that the invariant cleft forms part of the pilus subunit-binding pocket of PapD (147, 262). The X-ray crystal structure of the Arg-8–to–alanine mutant of PapD has now been solved, revealing that the mutation only changed the surface properties of this restricted region and did not cause any other structural alterations (S. Knight, unpublished data).
FimC (originally PilB) is the periplasmic chaperone required for the assembly of type 1 pili in the Enterobacteriaceae family (123, 135, 217). The chaperone activity of FimC was shown by demonstrating that FimC binds FimH, the type 1 adhesin, to form a periplasmic preassembly complex (123). Mannose-Sepharose affinity chromatography was used to purify a FimC-FimH complex from the periplasm. Furthermore, the heterologous chaperone, PapD, was able to complement a fimC1 mutant and orchestrate the assembly of type 1 pili (123). These results confirmed that the PapD and FimC structures are highly related since they are able to functionally substitute for one another. Interestingly, mutations of the invariant cleft residue Arg-8, to glycine, alanine, and methionine (R8G, R8A, and R8M), completely abolished the ability of PapD to assemble type 1 pili. Arg-8 and the other invariant residues in the conserved cleft probably have a common function in all pilus chaperones important in binding subunits (123). In support of this hypothesis, mutations in Arg-8 and Lys-112 of FimC have now been shown to abolish the ability of FimC to assemble type 1 pili (Jones and Hultgren, unpublished). The other members of the chaperone family (Table 1) presumably carry out similar functions in the assembly of pili and utilize a similar strategy for interaction with protein subunits, described in the next section.
Since the subunit-binding cleft of PapD is highly conserved among the chaperone family, the site recognized on the pilus subunits probably contains some underlying general features. One hallmark of pilus proteins assembled by PapD-like chaperones is a conserved carboxyl terminus consisting of a conserved penultimate tyrosine, a conserved glycine 14 residues from the COOH terminus, and an alternating pattern of hydrophobic residues between the conserved tyrosine and glycine residues (see Fig. 4). Deletion of the 14 carboxyl-terminal residues of PapG abolished PapD binding and PapD-PapG complex formation in vivo, without affecting the receptor-binding ability of the PapG truncate (101). In addition, point mutations in the carboxyl-terminal residues of PapG prevented incorporation of the adhesin into the pilus, and similar mutations in PapA completely abolished pilus assembly (E. Bullitt, C. H. Jones, R. Striker, G. Soto, and S. J. Hultgren, unpublished data). Based on this information, a comprehensive set of peptides was synthesized, derived from the carboxyl termini of the P-pilus subunits, and investigated for their ability to bind to PapD (147). PapD bound best to the peptide corresponding to the COOH terminal 19 amino acids of PapG. PapD also bound to PapF, PapE, and PapK COOH-terminal peptides, but with approximately 66 to 70% less affinity than to the PapG peptide. A totally unrelated, hydrophobic peptide was used in the assay as a control, and it did not bind to PapD (147). These results supported the hypothesis that the conserved carboxyl termini of pilus proteins formed part of the chaperone recognition motif. This motif closely resembles the motif recognized by BiP (29), the sole member of the Hsp70 family of chaperones in the lumen of the endoplasmic reticulum (69).
The molecular basis of the PapD-PapG interaction was investigated by cocrystallizing PapD with a peptide corresponding to the carboxyl terminal 19 amino acids of PapG (147). The molecular interactions between PapD and the peptide illustrate a mechanism of general relevance to the chaperone binding of target proteins (see Fig. 5). The peptide bound in an extended conformation, with its carboxyl terminus anchored in the cleft via hydrogen bonding to the invariant Arg-8 and Lys-112 residues that probably function as a molecular anchor in all of the PapD-like chaperones for binding pilus subunits. Site-directed mutations in Arg-8 and Lys-112 abolished the ability of PapD to bind to PapG in vitro and to mediate pilus assembly in vivo, indicating that the PapD-peptide crystal structure is a reflection of PapD-PapG interactions (147). The positioning of the BiP-like recognition sequence along the exposed edge of PapD’s G1 β-strand was mostly the result of backbone hydrogen bonds forming a β-sheet structure between the chaperone and peptide that was defined as being a "beta zipper" (Fig. 5) (147). The strong "zippering" interactions along the length of the COOH terminus of the subunit probably allow binding to a number of polypeptides in a relatively sequence-independent manner. However, it is clear that the binding does not rely solely on backbone hydrogen bonds, since PapD does not bind randomly to peptides in vitro, or proteins in vivo. Part of the specificity in binding between PapD and pilus subunits seems to be due to the alternating pattern of hydrophobic residues in the COOH terminus of pilus subunits being in register with conserved alternating hydrophobic amino acids present in the G1 β-strand of PapD. The appropriate alignment of these hydrophobic amino acids results in the formation of hydrophobic interactions that contribute to the stability of the complex.
The PapD-peptide crystal structure essentially represents a "snapshot" of a fundamental process in bacterial pathogenesis: the interaction of an adhesin with a chaperone that is a prerequisite to adhesin presentation on the microbial surface. These data suggest a general model in which the carboxyl termini of newly translocated subunits zipper to the G1 β-strand of the chaperone. PapD may provide a platform for β-strand zippering, allowing the subunits to achieve their native-like conformations. Since chaperones are thought to cap and uncap the associative surfaces of pilus subunits (146), the carboxyl terminus is implicated as part of the surface used in subunit-subunit associations. In the case of PapG, it may also be part of the surface used to join PapG to the tip fibrilla via an interaction with PapF.
A second site of interaction between PapD and PapG has recently been identified by Xu et al. (301) in a region between residues 175 and 198 of PapG (see Fig. 4). Four overlapping peptides were synthesized, corresponding to the region between residues 159 to 195, and tested for their ability to bind PapD in an enzyme-linked immunosorbent assay. A similar strategy proved successful in elucidating the β-strand zippering between PapD and the COOH terminus of pilus subunits. The peptide corresponding to residues 175 to 190, but not the other three peptides, bound to PapD in an enzyme-linked immunosorbent assay. These data argue that PapD recognizes two surfaces on PapG. PapD forms a β-strand zippering interaction with the COOH terminus but also recognizes a region containing residues 175 to 190 of PapG. The cognate surface on PapD has not yet been identified.
Although the entire chaperone family can be grouped together as one PapD-like superfamily, there are several striking structural differences between them that are conserved within two subgroups (Hung et al., unpublished data). The structural differences that delineate the two chaperone subfamilies are confined to the region of the chaperone involved in subunit binding (Hung et al., unpublished data). Remarkably, the two chaperone subgroups assemble architecturally distinct fibers. Thus, the first subfamily assembles rodlike fibers, whereas the second subfamily comprises the chaperones that assemble either nonfimbrial adhesins or structures that have an atypical morphology. The conserved features in the COOH terminus of all subunits that are assembled into rodlike fibers are absent or variable in subunits that are assembled into nonfimbrial or atypical structures (Hung et al., unpublished data). Thus, the immunoglobulin-like chaperones that assemble nonfimbrial or atypical structures probably use a variation of the beta-zipper recognition paradigm presented in Fig. 5, or perhaps even a totally different binding scheme.
In summary, the cleft is highly conserved among the entire chaperone superfamily and may have a universal function in subunit binding in all the members of the family, although two recognition paradigms may exist. It seems that PapD utilizes the immunoglobulin fold in two linked domains that are oriented to form a binding cleft. PapD binds pilin subunits through beta-zippering and molecular anchoring interactions with conserved residues in the cleft. The β-barrel structure also stabilizes variable loop regions that surround the cleft and may impart specificity to the chaperone and play a role in delineating the two chaperone subfamilies.
PapD binds to each of the pilus subunits as they emerge from the cytoplasmic membrane. By using galabiose-Sepharose affinity chromatography, the PapG adhesin was isolated in a preassembly complex with
PapD from the periplasmic space (101). PapG mediates specific binding to Galα(1, 2, 3, 4)Gal, even when in a complex with PapD, with the same receptor-binding specificity as when it is incorporated into the tip of the pilus (101, 133, 146). The binding specificity of PapG seems likely to be a function of its tertiary structure, as is the case for most lectins (293), arguing that PapG exists in a native-like state when bound to PapD. Thus, in contrast to cytoplasmic chaperones like SecB and DnaK, which bind to and maintain polypeptides in a nonnative state (87, 153, 154), the role of periplasmic chaperones such as PapD may be to maintain the bound pilus subunits in assembly-competent native-like conformations. Supporting this hypothesis was the detection of a typical β-sheet spectrum of PapG within the complex as determined by circular dichroism (146). In addition, the four cysteine residues in PapG formed intramolecular disulfide bonds even when PapG was bound by PapD. Dissociation of the PapD-PapG complex in vitro, under reducing conditions in urea, uncapped interactive surfaces on PapG that caused the proteins to form large-molecular-weight aggregates upon dilution of the denaturant (146). However, when the denaturant was diluted in the presence of an excess amount of native PapD, the interactive surfaces on PapG were effectively capped, preventing aggregation by reforming the soluble PapD-PapG complex (146).
The interaction of PapD with PapA and PapK was investigated and compared with PapD-PapG and PapD-PapE interactions in order to gain further insight into pilus assembly (269). PapA, the major subunit of the pilus rod, formed two periplasmic complexes (DA2 and DA) with PapD that were purified and characterized. PapK, which is an adaptor protein that joins the tip fibrilla to the pilus rod, was also purified as a distinct complex with PapD (DK) (269). Only "fiber-forming" or homopolymeric subunits, PapA in the rod and PapE in the tip fibrilla, were able to form subunit-subunit interactions in vivo in the periplasm or in vitro. PapA subunits interact to form distinct multimers in the periplasm that utilize the same interactive surfaces that are used in the formation of the pilus rod. Similarly, PapE multimers have been identified in the periplasm. The efficient formation of both the PapA and PapE multimers is dependent on PapD (269; Jones and Hultgren, unpublished data). Subunits that are present in single or low copy in the pilus (PapK and PapG) did not form periplasmic intersubunit interactions. A pulse-chase analysis revealed that chaperone-subunit complexes are true periplasmic intermediates in pilus assembly (269). By comparing the stability, the stoichiometry, and the surfaces that are available for interaction with other proteins, we are refining our understanding of bacterial secretion and assembly across the outer membrane.
P-pili biogenesis provides a model system to investigate fundamental principles governing the assembly of cellular organelles. An in vitro system was developed to reconstitute pilus subassemblies, PapA rods and PapE fibrillae, by dissociating PapD from purified chaperone-subunit complexes. Dissociation of the chaperone (uncapping) from the rod-forming subunit, PapA, or the tip fibrilla- forming subunit, PapE, led to the formation of pilus rod or tip fibrilla subassemblies, respectively (E. Bullitt, C. H. Jones, R. Striker, G. Soto, F. Dubuisson, M. J. Wick, L. Makowski, and S. J. Hultgren, submitted for publication). Uncapping the adaptor protein PapK did not result in the formation of any oligomeric structures. The presence of excess purified PapD blocked the formation of the pilus-like rods, arguing that PapD caps associative surfaces on the subunit. Mutations in a conserved residue in PapA, at the leading edge of the recognition motif, abolished subunit-subunit associations both in vivo and in vitro, which indicates that this region participates in subunit-subunit interactions after chaperone uncapping. The design of an in vitro system that discriminates between the two major subassemblies of the pilus and between fiber-forming and non-fiber-forming subunits has provided a unique opportunity to investigate the development of a cellular organelle that is a virulence factor.
PapD may function as a reversible capping protein which modulates polymerization by capping and uncapping interactive surfaces of pilus subunits that are imported into the periplasmic space. When PapD is bound to the subunits, incorrect interactions are blocked, and the release of PapD at the assembly site uncaps the assembly surface and allows the subunit to be polymerized into the pilus rod. The binding and release of PapD are orchestrated to occur at distinct sites within the cell, in order to guide the protein protomers along a productive assembly pathway.
Compartment-specific chaperones in eukaryotic cells have been shown to facilitate vectorial transport of target proteins into organelles. In the mitochondria (253, 286) and chloroplast (254), hsp70 homologs have been shown to "meet" proteins as they transverse the organelle membrane. In the endoplasmic reticulum, BiP (immunoglobulin-binding protein) carries out a similar role (84, 197, 246). Elegant experiments using protein fusions to dihydrofolate reductase were used to trap intermediates in the chaperone-facilitated transport process (286). In bacteria, the Sec pathway is one mechanism for secretion of proteins across the inner membrane into the periplasm (297). The assembly of both type 1 and P pili has been shown to be dependent on Sec-mediated transport of subunits across the inner membrane (51, 109, 111). Matsuyama et al. (179) demonstrated that a monoclonal antibody directed at SecD, an inner membrane component of the Sec machinery, would block periplasmic localization of maltose-binding protein. Monoclonal antibodies directed at other components of the Sec pathway had no effect on maltose-binding protein transport. Folding of maltose-binding protein was found to be dependent on "release" from the Sec apparatus. This finding suggests that proteins are actively released into the periplasm, once folded, and not simply secreted across the membrane (179).
Immunoglobulin-like chaperones, like PapD, play a role similar to other compartmentalized chaperones (Jones and Hultgren, unpublished data). PapD recognition of nascent translocating intermediates facilitates continued vectorial translocation across a membrane. The DegP protease (267, 268) was found to be responsible for pilin subunit and/or subunit aggregate degradation when the subunits were expressed in the absence of the chaperone (16; Jones and Hultgren, unpublished data). The fate of subunits was analyzed when expressed in a degP mutant in the presence and absence of PapD. Efficient vectorial transport of P-pilus subunits into the periplasmic space was dependent on PapD (Jones and Hultgren, unpublished data). This interaction probably results in the "capping" of subunits described above. Association of the chaperone with an unfolded target protein suggests that later steps in the protein-folding pathway may occur in concert with the chaperone and possibly drive the import process.
Protease sensitivity is often a marker of an unfolded structure (87). Pilus subunits are sensitive to proteolysis in the absence of an interaction with a chaperone, suggesting that the subunits by themselves are unable to fold into a stable tertiary structure (16, 101, 102, 103, 123, 162, 217). Expression of PapG in the absence of the chaperone leads to the cleavage of PapG into a stable 24.5-kDa truncate (101). This truncate corresponds to the amino terminus of PapG and exists in a native Galα((1, 2, 3, 4)Gal receptor-binding conformation (see previous section). Apparently, the COOH-terminal domain, which seems to play a role in assembly of the subunit into the pilus, is not folded properly in the absence of the chaperone and thus is a target for proteolysis. Similarly, the type 1 adhesin, FimH, is cleaved to an active 16-kDa truncate when expressed in the absence of the FimC chaperone. As discussed in the above section, expression of PapG in a degP strain results in the failure of PapG to be efficiently imported into the periplasmic space. When this full-length PapG was isolated and tested for receptor-binding activity, it was found to be inactive (Jones and Hultgren, unpublished data). This result suggests that the presence of the COOH-terminal assembly domain of PapG leads to misfolding of the sugar-binding domain of the adhesin or somehow interferes with its receptor-binding pocket when expressed in the absence of a chaperone. The ability of PapD to block the misfolding of PapG and allow for the protein to attain a protease-resistant, sugar-binding configuration suggests that PapD somehow serves as a template for the folding of subunits.
The driving force of aggregation for "unchaperoned" subunits may be the exposure of hydrophobic surfaces or the premature exposure of the associative surfaces that are involved in polymerization of subunits into the pilus. The former situation suggests that subunit folding intermediates may occur temporally right on the immunoglobulin template provided by the chaperone. The latter situation supports a role of the chaperone in blocking off-pathway interactions at the inappropriate time and location.
The E. coli periplasmic disulfide isomerase DsbA (19, 20, 21) was initially shown to be required for Tcp (type 4 pilus) formation in V. cholerae (226). Recently, it was also shown to be an essential component of the P-pilus biogenesis pathway (110). In the Pap system DsbA was required for correct folding of PapD and the pilus subunits. PapD possesses a pair of cysteines spaced only 5 amino acids apart that form an intrasheet disulfide bond that is unusual for an immunoglobulin-like protein. The disulfide bond occurs between the H2 β-strand and the G2 β-strand in the CD4-like domain 2, as determined by X-ray crystallography (92). The H2 β-strand is not present in most PapD-like chaperones. The conversion of PapD’s cysteines into serines by site-directed mutagenesis resulted in misfolded and nonfunctional PapD in vivo, mimicking the effects of the dsbA mutation (110). Paradoxically, disulfide bond formation was not required for PapD folding or for its recognition function in vitro. Denatured reduced PapD was able to refold and achieve a subunit binding conformation in vitro in the absence of disulfide bond formation. The discrepancy between in vivo and in vitro folding of PapD in the absence of its intramolecular disulfide bond may be due to a more stringent need to control folding (prevent misfolding) of the proteins in vivo as they emerge into the periplasmic space from the cytoplasmic membrane (110). DsbA might bind to nascently translocated PapD, maintaining it in a folding-competent state until the entire protein is exported across the membrane, at which time disulfide bond formation is catalyzed and correct protein folding ensues. However, it is possible that DsbA acts merely by oxidizing PapD when the cysteines emerge from the cytoplasmic membrane and that the formation of the disulfide bond is a critical event for the in vivo folding of PapD.
A chaperone role for DsbA in addition to its oxidoreductase function has been alluded to earlier (226, 303). Other oxidoreductase proteins with a thioredoxin motif have also been suggested to have chaperone-like functions. Both thioredoxin, which is required for filamentous phage f1 assembly, and protein disulfide isomerase of the eukaryotic endoplasmic reticulum have been implicated to have chaperone functions (150, 198, 220, 235).
DsbA also acts on the pilus subunits (110). The rapid acquisition of disulfide bonds in PapG even in the absence of PapD suggested that oxidation precedes chaperone-subunit complex formation. In the absence of PapD, the oxidized PapG was subject to limited proteolytic degradation. In a dsbA, papD background, PapG was not oxidized and was completely degraded very rapidly, suggesting that oxidized PapG achieves a more compact conformation or that the interaction with DsbA somewhat protects it. In related type 1 pili, whose chaperone (FimC) has no cysteines, FimC was stable and pili were assembled, albeit less efficiently than in a wild-type background. It is possible that in the absence of DsbA, oxidation of the type 1 pilus subunits can proceed spontaneously on the FimC template, or perhaps it is mediated by another disulfide periplasmic oxidase such as DsbC, which has been shown to functionally substitute for DsbA (187).
It is possible that subunit folding in the periplasmic space is a sequential process involving at least two chaperones: DsbA binds the subunit during its translocation across the cytoplasmic membrane, and after catalyzing the formation of the subunit disulfide bond(s), it passes it on to PapD (see biogenesis model in Fig. 6). Whether folding is completed prior to the interaction with PapD or on the PapD platform remains to be determined.
Pilus biogenesis involves the conversion of the chaperone-subunit complexes into surface fibers that have a distinct architecture. This conversion process depends on an outer membrane protein called a molecular usher, of which PapC is the prototype (52). In the absence of PapC, PapD-subunit complexes accumulate in the periplasmic space and the cells are bald. In each adhesin assembly system requiring a chaperone, a corresponding usher homolog was always present (see Table 1). In vitro studies have suggested that PapC might regulate the ordered targeting of chaperone-subunit complexes to the outer membrane assembly site, where the chaperone is dissociated from the respective subunits, allowing their polymerization into pili (52). Purified chaperone and chaperone-subunit complexes and partially purified PapC were used in two different in vitro assays to investigate and quantitate complementary protein interactions between PapC and chaperone-subunit complexes (52). The complexes that PapD forms with each of the three most distal tip fibrilla subunits (PapG, PapF, and PapE) were able to bind to PapC in both assay systems. Interactions between these complexes and PapC in vivo could direct their targeting to the outer membrane assembly site. Interestingly, the complexes between PapD and the most proximal subunit of the tip fibrilla (PapK) and the major rod subunit (PapA) did not bind to PapC. The ability of PapD-PapG, PapD-PapF, and PapD-PapE and the inability of PapD-PapA and PapD-PapK to bind to PapC suggest a mechanism by which pilus tips are made before pilus rods, as explained below and in Fig. 6.
The cascade of protein-protein interactions that leads to the formation of an adhesive composite pilus fiber is proposed to proceed according to the model in Fig. 6. PapD binds to each subunit as it is translocated into the periplasmic space. The resulting preassembly complexes are then targeted to PapC. In the initial stage, PapD-PapG has the highest relative affinity for PapC and thus binds to PapC first. PapF has the only complementary surface capable of correctly linking PapG to PapE (108). PapE subunits are then able to polymerize into a tip fiber upon multiple rounds of PapD-PapE binding, chaperone uncapping, and PapE incorporation. The inability of PapD-PapK and PapD-PapA to efficiently bind to empty PapC sites would ensure that pilus rods are not made in the absence of tip fibrillae. This is supported by data showing that deletion of the four tip genes (papK, papE, papF, papG) abolished the ability of bacteria to assemble pili even though the papA gene was expressed in the presence of a chaperone and an usher (108). We propose that the binding site for PapD-PapK may be the polymerized tip fibrilla in the context of PapC. Incorporation of PapK is known to terminate the growth of the tip fibrilla, and we propose this also would create a binding site for PapD-PapA. This hypothesis is supported by the discovery that PapK has been shown to be the only tip protein which, when expressed by itself, is able to nucleate the formation of a pilus rod (108). We propose that, upon chaperone uncapping and PapK incorporation into the growing fiber, tip growth is terminated and the targeting of PapD-PapA complexes to the assembly site is initiated, allowing polymerization of pilus rods.
All available data on pilus assembly would suggest that the pilus fiber is assembled from the base, meaning that the distal end containing the adhesin is assembled first (108, 170). Such a mode of polymerization is clearly different from that of flagella, which assemble from the tip such that new subunits are added to the distal end (174). A novel class of surface organelles, termed curli in E. coli and thin aggregative fimbriae in Salmonella, have recently been identified (9, 43, 213) (Table 1). These structures are morphologically distinct from pili. The formation of these fibers occurs from the outside of the microbe by the precipitation of secreted soluble subunit proteins into thin fibers (Hammar et al., unpublished data). This precipitation requires a nucleator protein that normally is expressed by the same organism but can also be secreted by adjacent cells.
Certain strains of E. coli K-12, such as MC4100, can be induced to express curli during stationary phase when grown at temperatures below 37°C and under conditions of low osmolarity (9, 213). The major subunit of curli is a 15-kDa protein encoded by the csgA gene, located at 23.1 min on the chromosome. csgA is the second gene in an operon which also contains the csgB gene (8). Neither of these two proteins shows any sequence homology to the pili subunits belonging to the chaperone/usher-dependent assembly systems. Notably, CsgA and CsgB both lack the BiP-like COOH-terminal recognition sequence required for chaperone binding.
Much less is known about curli formation than about pilus biogenesis. The Normark lab has recently shown that a nonpolar mutation in csgB abolished curli formation completely (Hammar et al., unpublished data). Interestingly, the CsgA subunit protein is still formed and is secreted out into the medium as a soluble protein. When such a CsgB– CsgA+ mutant is grown on a plate in close proximity to a CsgB+ CsgA– noncurliated mutant, curli are precipitated to form on the side of the CsgB+ CsgA– colony facing the CsgB– CsgA+ colony (Hammar et al., unpublished data). Thus, the CsgB protein on the surface of a CsgA– strain acts as a nucleator for soluble CsgA protein secreted from the CsgA+ donor strain. This result suggests that this organelle is formed on the outside of the organism by a novel precipitation reaction.
By transposon mutagenesis it has been possible to identify other genes required for curli formation. Such mutants are easy to identify since curliated E. coli colonies avidly bind Congo red, thereby forming blue-red colonies. An operon was identified adjacent to the csgBA operon but transcribed in the opposite orientation. This operon contains the csgD, csgE, csgF, and csgG genes. Transposon insertion mutations in csgD, csgE, or csgF produce mutants unable to form curli and unable to act as donors in nucleation reactions with a CsgB+ CsgA– recipient, suggesting that this set of proteins is somehow involved in the secretion of soluble CsgA across the outer membrane. While CsgD shows significant homology in its carboxyl-terminal region to the DNA-binding region of the LuxR family of transcriptional regulators, the other predicted proteins show no significant homology to any other protein in the database (83a). Thus, the curli gene cluster lacks genes encoding an immunoglobulin-like chaperone and an outer membrane usher protein typical of the 24 adhesin systems listed in Table 1. This is a further argument that the assembly of curli may follow a different pathway than that of pili, although the possibility exists that an unlinked gene encoding a periplasmic chaperone could function in the biogenesis of curli. Such a chaperone would most likely be required to escort the curli subunits through the periplasmic space.
Salmonella enteritidis (official designation, S. enterica serovar Enteritidis) strain 27655 has been shown to express large quantities of so-called thin aggregative fimbriae (43). Amino-terminal sequencing of the major subunit protein and the nucleotide sequence of the corresponding gene showed that the Salmonella fibers and curli belong to the same class of structures. The S. enteritidis fibers are also referred to as Sef17 fimbriae (S. enteritidis fimbriae, major subunit 17 kDa). DNA hybridization with a subunit-specific probe revealed that all tested Salmonella strains carried the genetic information for expression of Sef17 fimbriae. S. enteritidis 27655 was phase variable with respect to its ability to form Sef17 fibers. Interestingly, the noncurliated phase turned out to be considerably more virulent in mice than the curliated phase (C. Ericson, A. Arnquist, S. Sukupolvi, Z. Bian, and S. Normark, unpublished data).
The exact role of curli in the biology and pathogenesis of E. coli, Salmonella, and other enteric bacteria is not known. Due to the hydrophobic nature of the polymerized fiber, curliated cells tend to autoaggregate. Whether autoaggregation is also due to intercellular fiber precipitation is not yet known. Curliated cells also bind a series of host proteins such as fibronectin, laminin, and plasminogen (9, 213, 261). The latter interaction was shown to result in the activation of plasminogen to plasmin following capture of tissue plasminogen activator (261). Due to the environmental conditions that are required for curli expression, it may well be that the formation of these fibers plays a role when enterics exist outside a eukaryotic host.
Curli regulation appears to be very complex. Studies from the Normark laboratory have revealed the following clues into this regulatory pathway. Only certain E. coli K-12 strains are able to express curli in vitro even though they contain a complete set of csg genes. Transcriptional analyses have demonstrated that both the csgBA and the csgCDEF promoter are silent in the absence of the stationary phase-specific sigma factor σ S (RpoS) (83a). Crypticity may therefore be due to rpoS mutations known to accumulate in E. coli K-12 strains preferentially during prolonged incubation in stationary phase (304). The nucleoid-associated protein H-NS acts as a negative modulator for both csg promoters. In an rpoS hns double mutant both csg promoters are activated (8). Therefore it has been suggested that the csgBA and csgDEFG promoters can be recognized by either σ S or σ 70, but that the latter sigma factor is normally excluded from interaction directly or indirectly by H-NS.
Although the csgBA promoter requires σ s for activation in H-NS+ strains, an rpoS hns double mutant transcribes from this promoter in a growth-phase-dependent manner. A second pathway for growth-phase regulation of curli has therefore been postulated. It was demonstrated that an insertion in csgD totally abolished transcription from the csgBA promoter. This result suggests that CsgD is a transcriptional activator for curli expression and that one pathway to activation could involve CsgD interaction with an effector molecule that is secreted into the medium and serves as a cell density signal (M. Hammar and S. Normark, unpublished data). A recently identified 50-bp activating sequence located immediately upstream from the csgBA promoter could represent a binding site for CsgD. This sequence was shown to be required for transcriptional activation from this promoter both in wild type and in rpoS hns double mutants (8).
Curli expression requires conditions of low osmolarity (213), and Hammar and Normark (unpublished) recently demonstrated that OmpR is required for curli expression and for transcriptional activation of both the csgBA and the csgCDEF promoters. It is not known whether OmpR binds directly to one or both of the two csg promoters. Since envZ mutants still are able to express curli, it seems that phosphorylation of OmpR via acetylphosphate or another phosphate donor is sufficient for activation of the curli system. Thus, like ompF, csg transcription is apparently activated by the low-phosphorylated form of OmpR.
Understanding the molecular events occurring at the microbe-host interface promises to provide new insights into how bacteria cause disease. It will be necessary to characterize adhesin-receptor interactions in fine molecular detail. This usually requires identifying of the adhesin and purifying it in native form. In addition, the identification of putative receptor molecules and the synthesis of receptor analogs facilitate the molecular analysis of adhesin-receptor interactions. To definitively analyze the interactive surface of the adhesin molecule as it is presented to the host, the three-dimensional structure of a bacterial adhesin in a complex with its receptor and/or chaperone will need to be determined. This information will be necessary for the rational design and chemical synthesis of synthetic carbohydrate receptor analogs in addition to peptides corresponding to the interactive surfaces on these molecules. This will lead to a better understanding of how bacteria assemble and present their repertoire of specific adhesins, both to accommodate specific receptor architectures on different host cell surfaces and to evade specific host immune responses. Elucidating the nature and impact of regulatory mediators elicited by bacterial adhesins from cells actively involved in inflammatory responses, such as neutrophils, monocytes, and mast cells, will provide important clues to understanding bacterial infections.
The ability to understand the fine molecular details of pilus assembly will continue to unveil general principles of organelle development. Understanding the mechanism of action of molecular chaperones and ushers is providing paradigms for the mechanism of action of other chaperones. The blend of a powerful genetic system with X-ray crystallography, biochemistry, carbohydrate chemistry, and high-resolution electron microscopy to investigate the structure, function, and mechanism of action of immunoglobulin-like chaperones and ushers will provide a model system for understanding underlying biological processes in organelle assembly across biological membranes. In addition, the critical and prototypic role of PapD in the assembly of a virulence organelle makes it important to investigate the molecular basis of its mechanism of action. The 1990s may come to be remembered as a decade in which infectious diseases made a dramatic worldwide resurgence. The dramatic rise in antibiotic-resistant pathogens has kindled the need to understand the "molecular logic" of virulent bacteria in order to reveal new target proteins for the design of new and better drugs and vaccines. We are striving to understand the structural basis of pilus assembly, which will provide a "blueprint" for the development of novel antimicrobial therapeutic strategies and vaccines.
References
1. Abraham, S. N., J. P. Babu, C. S. Giampapa, D. L. Hasty, W. A. Simpson, and E. H. Beachey. 1985. Protection against Escherichia coli-induced urinary tract infections with hybridoma antibodies directed against type 1 fimbriae or complementary d-mannose receptors. Infect. Immun. 48:625–628.
2. Abraham, S. N., J. D. Goguen, D. Sun, P. Klemm, and E. H. Beachey. 1987. Identification of two ancillary subunits of Escherichia coli type 1 fimbriae by using antibodies against synthetic oligopeptides of fim gene products. J. Bacteriol. 169:5530–5536.
3. Abraham, S. N., M. Land, S. Ponniah, R. Endres, D. L. Hasty, and J. P. Babu. 1992. Glycerol-induced unraveling of the tight helical conformation of Escherichia coli type 1 fimbriae. J. Bacteriol. 174:5145–5148.
4. Abraham, S. N., D. Sun, J. B. Dale, and E. H. Beachey. 1988. Conservation of the d-mannose-adhesion protein among type 1 fimbriated members of the family Enterobacteriaceae. Nature (London) 336:682–684.
5. Agace, W., S. Hedges, M. Ceska, and C. Svanborg. 1993. Interleukin-8 and the neutrophil response in mucosal Gram-negative infections. J. Clin. Invest. 92:780–785.
6. Allen, B. L., G. F. Gerlach, and S. Clegg. 1991. Nucleotide sequence and functions of mrk determinants necessary for expression of type 3 fimbriae in Klebsiella pneumoniae. J. Bacteriol. 173:916–920.
7. Amit, A. G., R. A. Marrizzua, S. E. Phillips, and R. J. Poljak. 1986. Three dimensional structure of an antibody-antigen complex at 2.8A resolution. Science 233:747–753.
8. Arnqvist, A., A. Olsen, and S. Normark. 1994. σS-dependent growth phase induction of the csgBA promoter in Escherichia coli can be achieved in vivo by σ70 in the absence of the nucleoid-associated protein H-NS. Mol. Microbiol. 13:1021–1032.
9. Arnqvist, A., A. Olsen, J. Pfieffer, D. G. Russel, and S. Normark. 1992. The Crl protein activates cryptic genes for curli formation and fibronectin binding in Escherichia coli. Mol. Microbiol. 6:2443–2453.
10. Baddour, L. M., G. D. Christensen, W. A. Simpson, and E. H. Beachey. 1990. Microbial adherence, p. 9–25. In G. L. Mandell, R. G. Douglas, Jr., and J. E. Bennet (ed.), Principles and Practice of Infectious Disease, 3rd ed., vol. 1. Churchill Livingston, New York.
11. Baga, M., M. Goransson, S. Normark, and B. E. Uhlin. 1985. Transcriptional activation of a Pap pilus virulence operon from uropathogenic Escherichia coli. EMBO J. 4:3887–3893.
12. Baga, M., M. Norgren, and S. Normark. 1987. Biogenesis of E. coli Pap pili: PapH, a minor pilin subunit involved in cell anchoring and length modulation. Cell 49:241–251.
13. Baga, M., S. Normark, J. Hardy, P. O’Hanley, D. Lark, O. Olsson, G. Schoolnik, and S. Falkow. 1984. Nucleotide sequence of the gene encoding the papA pilus subunit of human uropathogenic Escherichia coli. J. Bacteriol. 157:330–333.
14. Bahrani, F. K., S. Cook, R. Hull, G. Massad, and H. L. T. Mobley. 1993. Proteus mirabilis fimbriae: N-terminal amino acid sequence of a major fimbrial subunit and nucleotide sequences of the genes from two strains. Infect. Immun. 61:884–891.
15. Bahrani, F. K., D. E. Johnson, D. Robbins, and H. L. T. Mobley. 1991. Proteus mirabilis flagella and MR/P fimbriae: isolation, purification, N-terminal analysis, and serum antibody response following experimental urinary tract infection. Infect. Immun. 59:3574–3580.
16. Bakker, D., C. E. Vader, B. Roosendaal, F. R. Mooi, B. Oudega, and F. K. de Graaf. 1991. Structure and function of periplasmic chaperone-like proteins involved in the biosynthesis of K88 and K99 fimbriae in enterotoxigenic Escherichia coli. Mol. Microbiol. 5:875–886.
17. Baldini, M. M., J. B. Kaper, M. Levine, D. C. Candy, and H. W. Moon. 1983. Plasmid-mediated adhesion in enteropathogenic Escherichia coli. J. Pediatr. Gastroenterol. Nutrition 2:534–538.
18. Bally, M., A. Filloux, M. Akrim, G. Ball, A. Lazdunski, and J. Tommassen. 1992. Protein secretion in Pseudomonas aeruginosa: characterization of seven xcp genes and processing of secretory apparatus by prepilin peptidase. Mol. Microbiol. 6:1121–1131.
19. Bardwell, J. C., K. McGovern, and J. Beckwith. 1991. Identification of a protein required for disulfide bond formation in vivo. Cell 67:581–589.
20. Bardwell, J. C. A. 1994. Building bridges: disulphide bond formation in the cell. Mol. Microbiol. 14:199–205.
21. Bardwell, J. C. A., J.-O. Lee, G. Jander, N. Martin, D. Belin, and J. Beckwith. 1993. A pathway for disulfide bond formation in vivo. Proc. Natl. Acad. Sci. USA 90:1038–1042.
22. Bazan, J. F. 1990. Structural design and molecular evolution of a cytokine receptor superfamily. Proc. Natl. Acad. Sci. USA 87:6934–6938.
23. Beachey, E. H. (ed.). 1980. Bacterial Adherence Receptors and Recognition. Chapman & Hall, London.
24. Beachey, E. H. 1981. Bacterial adherence: adhesion-receptor interactions mediating the attachment of bacteria to mucosal surfaces. J. Infect. Dis. 143:325–345.
25. Benz, I., and M. A. Schmidt. 1989. Cloning and expression of an adhesin (AIDA-I) involved in diffuse adherence of enteropathogenic Escherichia coli. Infect. Immun. 57:1506–1511.
26. Benz, I., and M. A. Schmidt. 1992. AIDA-I, the adhesin involved in diffuse adherence of the diarrhoeagenic Escherichia coli strain 2787 (O126:H27), is synthesized via a precursor molecule. Mol. Microbiol. 6:1539–1546.
27. Benz, I., and M. A. Schmidt. 1992. Isolation and serologic characterization of AIDA-I, the adhesin mediating the diffuse adherence phenotype of the diarrhea-associated Escherichia coli strain 2787 (O126:H27). Infect. Immun. 60:13–18.
28. Bloch, C. A., B. A. Stocker, and P. E. Orndorff. 1992. A key role for type 1 pili in enterobacterial communicability. Mol. Microbiol. 6:697–701.
29. Blond-Elguindi, S., S. E. Cwirla, W. J. Dower, R. J. Lipshutz, S. R. Sprang, J. F. Sambrook, and M.-J. H. Gething. 1993. Affinity panning of a library of peptides displayed of bacteriophages reveals the binding specificty of BiP. Cell 75:717–728.
30. Blum, G., V. Falbo, A. Caprioli, and J. Hacker. 1995. Gene clusters incoding the cytotoxic necrotizing factor (cnf1), Pre fimbriae (prs) and α-hemolysin (hly) form the pathogenicity island II (PaiII) of the uropathogenic strain J96. FEMS Microbiol. Lett. 126:189–195.
31. Blum, G., M. Ott, A. Lischewski, A. Ritter, H. Imrich, H. Tschope, and J. Hacker. 1994. Excision of large DNA regions termed pathogenicity islands from tRNA-specific loci in the chromosome of an Escherichia coli wild-type pathogen. Infect. Immun. 62:606–614.
32. Bock, K., M. E. Breimer, A. Brignole, G. C. Hansson, K.-A. Karlsson, G. Larson, H. Leffler, B. E. Samuelsson, N. Strömberg, C. Svanborg-Edén, and J. Thurin. 1985. Specificity of binding of a strain of uropathogenic Escherichia coli to Galα(1-4)Gal-containing glycosphingolipids. J. Biol. Chem. 260:8545–8551.
33. Braaten, B., L. B. Blyn, B. S. Skinner, and D. A. Low. 1991. Evidence for a methylation-blocking factor (mbf) locus involved in pap pilus expression and phase variation in Escherichia coli. J. Bacteriol. 173:1789–1800.
34. Braaten, B. A., X. Nou, L. S. Kaltenbach, and D. A. Low. 1994. Methylation patterns in pap regulatory DNA control pyelonephritis-associated pili phase variation in E. coli. Cell 76:577–588.
35. Braaten, B. A., J. V. Platko, M. van der Woude, B. H. Simons, F. deGraaf, J. M. Calvo, and D. A. Low. 1992. Leucine-responsive regulatory protein controls the expression of both the pap and fan pili operons in Escherichia coli. Proc. Natl. Acad. Sci. USA 89:4250–4254.
36. Branden, C.-I., and J. Tooze. 1991. Introduction to Protein Structure. Garland Publishing, New York.
37. Breimer, M. E., G. C. Hansson, and H. Leffler. 1985. The specific glycosphingolipid composition of human ureteral epithelial cells. J. Biochem. 98:1169–1180.
38. Brinton, C. C., Jr. 1965. The structure, function, synthesis, and genetic control of bacterial pili and a model for DNA and RNA transport in gram negative bacteria. Trans. N.Y. Acad. Sci. 27:1003–1065.
39. Brinton, C. C. J., M. J. Carter, D. B. Derber, S. Kar, J. A. Kramarik, A. C.-C. To, S. C.-M. To, and S. W. Wood. 1989. Design and development of pilus vaccines for Haemophilus influenzae diseases. Pediatr. Infect. Dis. 8:554–561.
40. Bullitt, E., and L. Makowski. 1995. Structural polymorphism of bacterial adhesion pili. Nature (London) 373:164–167.
41. Clegg, S., and G. F. Gerlach. 1987. Enterobacterial fimbriae. J. Bacteriol. 169:934–938.
42. Clouthier, S. C., K. H. Muller, J. L. Doran, S. K. Collinson, and W. W. Kay. 1993. Characterization of three fimbrial genes, sefABC, of Salmonella enteriditis. J. Bacteriol. 175:2523–2533.
43. Collinson, S. K., L. Emody, K.-H. Muller, T. Trust, and W. M. Kay. 1991. Purification and characterization of thin, aggregative fimbriae from Salmonella enteriditis. J. Bacteriol. 173:4773–4781.
44. Cosman, D., S. D. Lyman, R. L. Idzerda, M. P. Beckmann, L. S. Park, R. G. Goodwin, and C. J. March. 1990. A new cytokine receptor superfamily. Trends Biochem. Sci. 15:265–270.
45. Craig, E. A., B. D. Gambill, and R. J. Nelson. 1993. Heat shock proteins: molecular chaperones of protein biognesis. Microbiol. Rev. 57:402–410.
46. Darzins, A. 1993. The pilG gene product, required for Pseudomonas aeruginosa pilus production and twitching motility, is homologous to the enteric, single-domain response regulator CheY. J. Bacteriol. 175:5934–5944.
47. de Graaf, F. K., and W. Gaastra. 1994. Fimbriae of enterotoxigenic Escherichia coli, p. 53–70. In P. Klemm (ed.), Fimbriae: Adhesion, Genetics, Biogenesis, and Vaccines. CRC Press, Ann Arbor, Mich.
48. de Graaf, F. K., and F. R. Mooi. 1986. The fimbrial adhesins of Escherichia coli. Adv. Microb. Physiol. 28:65–143.
49. de Man, P., L. Aarden, I. Engberg, H. Linder, C. Svanborg-Eden, and C. van Kooten. 1989. Interleukin-6 induced at mucosal surfaces by gram-negative bacterial infection. Infect. Immun. 57:3383–3388.
50. DeVos, A., M. Ultsch, and A. Kossiakoff. 1992. Human growth hormone and extracellular doman of its receptor: crystal structure of the complex. Science 255:306–312.
51. Dodd, D. C., P. J. J. Bassford, and B. I. Eisenstein. 1984. Dependence of secretion and assembly of type 1 fimbrial subunits of Escherichia coli on normal protein transport. J. Bacteriol. 159:1077–1079.
52. Dodson, K. W., F. Jacob-Dubuisson, R. T. Striker, and S. J. Hultgren. 1993. Outer membrane PapC usher discriminately recognizes periplasmic chaperone-pilus subunit complexes. Proc. Natl. Acad. Sci. USA 90:3670–3674.
53. Donnenberg, M., J. Giron, J. Nataro, and J. Kaper. 1992. A plasmid encoded type IV fimbrial gene of enteropathogenic Escherichia coli associated with localized adherence. Mol. Microbiol. 6:3427–3437.
54. Donnenberg, M. S., and J. B. Kaper. 1992. Enteropathogenic Escherichia coli. Infect. Immun. 60:3953–3961.
55. Duguid, J. P., and D. C. Old. 1980. Adhesive properties of Enterobacteriaceae, p. 186–217. In E. H. Beachey (ed.), Bacterial Adherence Receptors and Recognition. Chapman & Hall, London.
56. Eisenstein, B. I. 1981. Phase variation of type 1 fimbriae in Escherichia coli is under transcriptional control. Science 214:337–339.
57. Eisenstein, B. I. 1990. Enterobacteriaceae, p. 1658–1673. In G. L. Mandell, R. G. Douglas, Jr., and J. E. Bennet (ed.), Principles and Practice of Infectious Disease, 3rd ed., vol. 2. Churchill Livingston, New York.
58. Ellis, R. J., and S. M. Hemmingsen. 1989. Molecular chaperones: proteins essential for the biogenesis of some macromolecular structures. Trends Biochem. Sci. 14:339–342.
59. Ellis, R. J., and S. van der Vies. 1991. Molecular chaperones. Annu. Rev. Biochem. 60:321–347.
60. Ellis, R. J., S. M. van der Vies, and S. M. Hemmingsen. 1989. The molecular chaperone concept. Biochem. Soc. Symp. 55:145–153.
61. Erickson, A. K., J. A. Willgohs, S. Y. McFarland, D. A. Benfield, and D. H. Francis. 1992. Identification of two porcine brush border glycoproteins that bind the K88ac adhesin of Escherichia coli and correlation of these glycoproteins with the adhesive phenotype. Infect. Immun. 60:983–988.
62. Finlay, B. B., I. Rosenshine, M. S. Donnenberg, and J. B. Kaper. 1992. Cytoskeletal composition of attaching and effacing lesions associated with enteropathogenic Escherichia coli adherence to HeLa cells. Infect. Immun. 60:2541–2543.
63. Firon, N., I. Ofek, and N. Sharon. 1983. Carbohydrate specificity of the surface lectins of Escherichia coli, Klebsiella pneumoniae and Salmonella typhimurium. Carbohydr. Res. 120:235–249.
64. Frankel, G. D., C. A. Candy, P. Everest, and G. Dougan. 1994. Characterization of the C-terminal domainsof intimin-like proteins of enteropathogenic and enterohemorrhagic Escherichia coli, Citrobacter freundii, and Hafnia alvei. Infect. Immun. 62:1835–1842.
65. Friedich, M. J., N. E. Kinsey, J. Vila, and R. J. Kadner. 1993. Nucleotide sequence of a 13.9 kb segment of the 90 kb virulence plasmid of Salmonella typphimurium: the presence of a fimbrial biosynthetic gene. Mol. Microbiol. 8:543–558.
66. Froehlich, B. J., A. Karkashian, L. R. Melsen, J. C. Wakefield, and J. R. Scott. 1994. CooC and CooD are required for assembly of CS1 pili. Mol. Microbiol. 12:387–401.
67. Galan, J. E., C. Ginocchio, and P. Costeas. 1992. Molecular and functional characterization of the Salmonella invasion gene invA: homology of InvA to members of a new protein family. J. Bacteriol. 174:4338–4349.
68. Gerlach, G. F., B. L. Allen, and S. Clegg. 1988. Molecular characterization of the type 3 (MR/K) fimbriae of Klebsiella pneumoniae. J. Bacteriol. 170:3547–3553.
69. Gething, M. J., and J. Sambrook. 1992. Protein folding in the cell. Nature (London) 355:33–45.
70. Giampapa, C. S., S. N. Abraham, T. M. Chiang, and E. H. Beachey. 1988. Isolation and characterization of the receptor for type 1 fimbriae of Escherichia coli from guinea pig erythrocytes. J. Biol. Chem. 263:5362–5367.
71. Girdeau, J. P., M. D. Vartanian, J. L. Ollier, and M. Contrepois. 1988. CS31A, a new K88-related fimbrial antigen on bovine enteropathogenic and septicemic Escherichia coli strains. Infect. Immun. 56:2180–2188.
72. Giron, J. A., A. S. Y. Ho, and G. K. Schoolnik. 1991. An inducible bundle-forming pilus of enteropathogenic Escherichia coli. Science 254:710–713.
73. Giron, J. A., A. S. Y. Ho, and G. K. Schoolnik. 1993. Characterization of fimbriae produced by enteropathogenic Escherichia coli. J. Bacteriol. 175:7391–7403.
74. Giron, J. A., M. M. Levine, and J. B. Kaper. 1994. Longus: a long pilus ultrastructure produced by human enterotoxigenic Escherichia coli. Mol. Microbiol. 12:71–82.
75. Goldhar, J., R. Perry, J. R. Golecki, H. Hoschutzky, B. Jann, and K. Jann. 1987. Nonfimbrial, mannose-resistant adhesins from uropathogenic Escherichia coli O83:K1:H4 and O14:K?:H11. Infect. Immun. 55:1837–1842.
76. Gong, M., and L. Makowski. 1992. Helical structure of pap adhesion pili from Escherichia coli. J. Mol. Biol. 228:735–742.
77. Goransson, M., K. Forsman, P. Nilsson, and B. E. Uhlin. 1989. Upstream activating sequences that are shared by two divergently transcribed operons mediatae cAMP-CRP regulation of pilus-adhesin in Escherichia coli. Mol. Microbiol. 3:1557–1565.
78. Goransson, M., B. Sonden, P. Nilsson, B. Dagberg, K. Forsman, K. Emanuelsson, and B. E. Uhlin. 1990. Transcriptional silencing and thermoregulation of gene expression in Escherichia coli. Nature (London) 344:682–685.
79. Grunberg, J., R. Perry, H. Hoschutsky, B. Jann, and K. G. Jann. 1988. Nonfimbrial blood group N-specific adhesin (NFA-3) from Escherichia coli O20:KX104:H-, causing systemic infection. FEMS Microbiol. Lett. 56:241.
80. Guerina, N. G., S. Langerman, H. W. Clegg, T. W. Kessler, D. A. Goldmann, and J. R. Gilsdorf. 1982. Adherence of piliated Haemophilus influenzae to human oropharyngeal cells. J. Infect. Dis. 146:564.
81. Hacker, J., L. Bender, M. Ott, J. Wingender, B. Lund, R. Marre, and W. Goebel. 1990. Deletions and chromosomal regions coding for fimbrae and hemolysins occur in vitro and in vivo in various etxtraintestinal Escherichia coli isolates. Microb. Pathog. 8:213–225.
82. Hacker, J., G. Schmidt, C. Hughes, S. Knapp, M. Marget, and W. Goebel. 1985. Cloning and characterization of genes involved in production of mannose-resistant, neuraminidase-susceptible (X) fimbriae from a uropathogenic 06:K15:H31 Escherichia coli strain. Infect. Immun. 47:434–440.
83. Hakomori, S.-I. 1990. Bifunctional role of glycosphingolipids: modulators for transmembrane signaling and mediators for cellular interactions. J. Biol. Chem. 265:18713–18716.
83a. Hammar, M., A. Arnquist, Z. Bian, A. Olsén, and S. Normark. 1995. Expression of two csg operons is required for production of fibronectin and Congo red-bindina curli polymers in Escherichia coli K-12. Mol. Microbiol. 18:661–670.
84. Hammond, C., and A. Helenius. 1994. Folding of VSV G protein: sequential interaction with BiP and Calnexin. Science 266:456–458.
85. Hanson, M. S., and C. C. Brinton, Jr. 1988. Identification and characterization of E. coli type-1 pilus tip adhesion protein. Nature (London) 332:265–268.
86. Hanson, M. S., J. Hempel, and C. C. Brinton, Jr. 1988. Purification of the Escherichia coli type 1 pilin and minor pilus proteins and partial characterization of the adhesin protein. J. Bacteriol. 170:3350–3358.
87. Hardy, S. J., and L. L. Randall. 1991. A kinetic partitioning model of selective binding of nonnative proteins by the bacterial chaperone SecB. Science 251:439–443.
88. Haslam, D., T. Boren, P. Falk, D. Ilver, A. Chou, Z. Xu, and S. Normark. 1994. The amino-terminal domain of the P-pilus adhesin determines receptor specificity. Mol. Microbiol. 14:399–409.
89. Hedges, S., P. Anderson, G. Lidin-Janson, P. de Man, and C. Svanborg. 1991. Interleukin-6 response to deliberate colonization of the human urinary tract with gram-negative bacteria. Infect. Immun. 59:421–427.
90. Hobbs, M., E. S. R. Collie, P. D. Free, S. P. Livingston, and J. S. Mattick. 1993. PilS and PilR, a two-component transcriptional regulatory system controlling expression of type 4 fimbriae in Pseudomonas aeruginosa pili. Mol. Mircobiol. 7:669–682.
91. Hobbs, M., and J. S. Mattick. 1993. Common components in the assembly of type 4 fimbriae, DNA transfer systems, filamentous phage and protein-secretion apparatus: a general system for the formation of surface-associated protein complexes. Mol. Microbiol. 1:233–243.
92. Holmgren, A., and C. Brändén. 1989. Crystal structure of chaperone protein PapD reveals an immunoglobulin fold. Nature (London) 342:248–251.
93. Holmgren, A., M. J. Kuehn, C.-I. Brändén, and S. J. Hultgren. 1992. Conserved immunoglobulin-like features in a family of periplasmic pilus chaperones in bacteria. EMBO J. 11:1617–1622.
94. . Hoschutzky, H., F. Lottspeich, and K. Jann. 1989. Isolation and characterization of the α-galactosyl-1,4-β-galactosyl-specific adhesin (P adhesin) from fimbriated Escherichia coli. Infect. Immun. 57:76–81.
95. Hoschutzky, H., W. Nimmich, F. Lottspeich, and K. Jann. 1989. Isolation and characterization of the non-fimbrial adhesin NFA-4 from uropathogenic Escherichia coli O7:K98:H6. Microb. Pathog. 6:351–359.
96. Hull, R. A., R. E. Gill, P. Hsu, B. H. Minshaw, and S. Falkow. 1981. Construction and expression of recombinant plasmids encoding type 1 and d-mannose-resistant pili from a urinary tract infection Escherichia coli isolate. Infect. Immun. 33:933–938.
97. Hull, R. A., and S. I. Hull 1994. Adherence mechanisms in urinary tract infections, p. 79–90. In V. L. Miller, J. B. Kaper, D. A. Portnoy, and R. R. Isberg (ed.), Molecular Genetics of Bacterial Pathogenesis. ASM Press, Washington, D.C.
98. Hull, R. A., S. I. Hull, and S. Falkow. 1984. Frequency of gene sequences necessary for pyelonephritis-associated pili expression among isolates of Enterobacteriaceae from human extraintestinal infections. Infect. Immun. 43:1064–1067.
99. Hultgren, S. J., S. N. Abraham, M. G. Caparon, P. Falk, J. W. St. Geme III, and S. Normark. 1993. Pilus and non-pilus bacterial adhesins: assembly and function in cell recognition. Cell 73:887–901.
100. Hultgren, S. J., F. Jacob-Dubuisson, C. H. Jones, and C.-I. Brändén. 1993. PapD and superfamily of periplasmic immunoglobulin-like pilus chaperones. Adv. Protein Chem. 44:99–123.
101. Hultgren, S. J., F. Lindberg, G. Magnusson, J. Kihlberg, J. M. Tennent, and S. Normark. 1989. The PapG adhesin of uropathogenic Escherichia coli contains separate regions for receptor binding and for the incorporation into the pilus. Proc. Natl. Acad. Sci. USA 86:4357–4361.
102. Hultgren, S. J., and S. Normark. 1991. Biogenesis of the bacterial pilus. Curr. Opin. Gen. Dev. 1:313–318.
103. Hultgren, S. J., S. Normark, and S. N. Abraham. 1991. Chaperone-assisted assembly and molecular architecture of adhesive pili. Annu. Rev. Microbiol. 45:383–415.
104. Hultgren, S. J., T. N. Porter, A. J. Schaeffer, and J. L. Duncan. 1985. Role of type 1 pili and effects of phase variation on lower urinary tract infections produced by Escherichia coli. Infect. Immun. 50:370–377.
105. Hultgren, S. J., W. R. Schwan, A. J. Schaeffer, and J. L. Duncan. 1986. Regulation of production of type 1 pili among urinary tract isolates of Escherichia coli. Infect. Immun. 54:613–620.
106. Hurle, M. R., L. R. Helms, L. Li, W. Chan, and R. Wetsel. 1994. A role for destabilizing amino acid replacements in light-chain amyloidosis. Proc. Natl. Acad. Sci. USA 91:5446–5450.
107. Iriarte, M., C. Vanooteghem, I. Delor, R. Diaz, S. Knutton, and G. R. Cornelis. 1993. The Myf fibrillae of Yersinia enterocolitica. Mol. Microbiol. 9:507–520.
108. Jacob-Dubuisson, F., J. Heuser, K. Dodson, S. Normark, and S. J. Hultgren. 1993. Initiation of assembly and association of the structural elements of a bacterial pilus depend on two specialized tip proteins. EMBO J. 12:837–847.
109. Jacob-Dubuisson, F., M. Kuehn, and S. Hultgren. 1993. A novel secretion apparatus for the assembly of adhesive bacterial pili. Trends Microbiol. 1:50–55.
110. Jacob-Dubuisson, F., J. Pinkner, Z. Xu, R. Striker, A. Padmanhaban, and S. J. Hultgren. 1994. PapD chaperone function in pilus biogenesis depends on oxidant and chaperone-like activities of DsbA. Proc. Natl. Acad. Sci. USA 91:11552–11556.
111. Jacob-Dubuisson, F., R. Striker, and S. J. Hultgren. 1994. Chaperone-assisted self-assembly of pili independent of cellular energy. J. Biol. Chem. 269:12447–12455.
112. Jacobs, A. A. C., C. B. H. Dimons, and F. K. de Graaf. 1987. The role of lysine-132 and arginine-136 in the receptor-binding domain of the K99 fibrillar subunit. EMBO J. 6:1805–1808.
113. Jacobs, A. A. C., B. Roosendaal, J. F. L. van Breemen, and F. K. de Graaf. 1987. Role of phenylalanine 150 in the receptor binding domain of the K88 fibrillar subunit. J. Bacteriol. 169:4907–4911.
114. Jann, K., and H. Hoschutzky. 1990. Nature and organization of adhesins. Curr. Top. Microbiol. Immunol. 151:55.
115. Jann, K., and H. Hoschutzky. 1991. Characterization and surface organization of E. coli adhesins, p. 3–10. In E. Z. Ron and S. Rotten (ed.), Microbial Surface Components and Toxins in Relation to Pathogenesis. Plenum Press, New York.
116. Jerse, A. E., K. G. Gicquelais, and J. B. Kaper. 1991. Plasmid and chromosomal elements involved in the pathogenesis of attaching and effacing Escherichia coli. Infect. Immun. 59:3869–3875.
117. Jerse, A. E., and J. B. Kaper. 1991. The eae gene of enteropathogenic Escherichia coli encodes a 94-kilodalton membrane protein, the expression of which is influenced by the EAF plasmid. Infect. Immun. 59:4302–4309.
118. Jerse, A. E., J. Yu, B. D. Tall, and J. B. Kaper. 1990. A genetic locus of enteropathogenic Escherichia coli necessary for the production of attaching and effacing lesions on tissue culture cells. Proc. Natl. Acad. Sci. USA 87:7839–7843.
119. Johnson, J. R. 1991. Virulence factors in Escherichia coli urinary tract infection. Clin. Microbiol. Rev. 4:80–128.
120. Johnson, J. R., I. Orskov, F. Orskov, P. Goullet, B. Picard, S. L. Moseley, P. L. Roberts, and W. E. Stamm. 1994. O K and H antigens predict virulence factors, carboxylesterase B pattern, antimicrobial resistance and host compromise among Escherichia coli strains causing urosepsis. J. Infect. Dis. 169:119–126.
121. Jokinen, M., C. Ehnholm, V. Vaisanen-Rhen, T. K. Korhonen, R. Pipkorn, N. Kalkkinen, and C. G. Gahmberg. 1985. Identification of the major human sialylglycoprotein from red cells, glycophorin AM, as the receptor for Escherichia coli IH11165 and characterization of the receptor site. Eur. J. Biochem. 147:47–52.
122. Jones, C. H., D. F. Jacob, K. Dodson, M. Kuehn, L. Slonim, R. Striker, and S. J. Hultgren. 1992. Adhesin presentation in bacteria requires molecular chaperones and ushers. Infect. Immun. 60:4445–4451.
123. Jones, C. H., J. S. Pinkner, A. V. Nicholes, L. N. Slonim, S. N. Abraham, and S. J. Hultgren. 1993. FimC is a periplasmic PapD-like chaperone that directs assembly of type 1 pili in bacteria. Proc. Natl. Acad. Sci. USA 90:8397–8401.
124. Jones, C. H., J. S. Pinkner, R. Roth, J. Heuser, A. V. Nicholes, S. N. Abraham, and S. J. Hultgren. 1995. FimH adhesin of type 1 pili is assembled into a fibrillar tip structure in the Enterobacteriaceae. Proc. Natl. Acad. Sci. USA 92:2081–2085.
125. Kallenius, G., R. Mollby, S. B. Svenson, J. Windberg, A. Lundblud, S. Svenson, and B. Cedergen. 1980. The Pk antigen as receptor for the haemagglutinin of pyelonephritogenic Escherichia coli. FEMS Microbiol. Lett. 8:297–302.
126. Kallenius, G., S. B. Svenson, H. Hultberg, R. Molby, I. Helin, B. Cedergen, and J. Windberg. 1981. Occurence of P fimbriated Escherichia coli in urinary tract infection. Lancet ii:1369–1372.
127. Kaper, J. B. 1994. Molecular pathogenesis of enteropathogenic Escherichia coli, p. 173–195. In V. L. Miller, J. B. Kaper, D. A. Portnoy, and R. R. Isberg (ed.), Molecular Genetics of Bacterial Pathogenesis. ASM Press, Washington D.C.
128. Karlsson, K.-A. 1989. Animal glycosphingolipids as membrane attachment sites for bacteria. Annu. Rev. Biochem. 58:309–350.
129. Karlyshev, A., E. Galyov, O. Smirnov, A. Guzayev, V. Abramov, and V. Zav’yalov. 1992. A new gene of the f1 operon of Y. pestis involved in the capsule biogenesis. FEBS Lett. 297:77–80.
130. Kaufman, M. R., and R. K. Taylor 1994. Fimbriae of Vibrio cholerae, p. 157–171. In P. Klemm (ed.), Fimbriae: Adhesion, Genetics, Biogenesis and Vaccines. CRC Press, Ann Arbor, Mich.
131. Keith, B. R., S. L. Harris, P. W. Russell, and P. E. Orndorff. 1990. Effect of type 1 piliation on in vitro killing of Escherichia coli by mouse peritoneal macrophages. Infect. Immun. 58:3448–3454.
132. Keith, B. R., L. Maurer, P. A. Spears, and P. E. Orndorff. 1986. Receptor-binding function of type 1 pili effects bladder colonization by a clinical isolate of Escherichia coli. Infect. Immun. 53:693–696.
133. Kihlberg, J., S. J. Hultgren, S. Normark, and G. Magnusson. 1989. Probing of the combining site of the PapG adhesion of uropathogenic Escherichia coli bacteria by synthetic analogues of galabiose. J. Am. Chem. Soc. 111:6364–6368.
134. Klemm, P. 1985. Fimbrial adhesins of Escherichia coli. Rev. Infect. Dis. 7:321–329.
135. Klemm, P. 1992. FimC, a chaperone-like periplasmic protein of Escherichia coli involved in biogenesis of type 1 fimbriae. Res. Microbiol. 143:831–840.
136. Klemm, P. (ed.). 1994. Fimbriae: Adhesion, Genetics, Biogenesis, and Vaccines. CRC Press, Ann Arbor, Mich.
137. Klemm, P., and G. Christiansen. 1987. Three fim genes required for the regulation of length and mediation of adhesion of Escherichia coli type 1 fimbriae. Mol. Gen. Genet. 208:439–445.
138. Klemm, R., and G. Christiansen. 1990. The fimD gene required for cell surface localization of Escherichia coli type 1 fimbriae. Mol. Gen. Genet. 220:334–338.
139. Knutton, S., M. M. Baldini, J. B. Kaper, and A. S. McNeish. 1987. Role of plasmid-encoded adherence factors in adhesion of enteropathogenic Escherichia coli to HEp-2 cells. Infect. Immun. 55:78–85.
140. Knutton, S., T. Baldwin, P. H. Williams, and A. S. McNeish. 1989. Actin accumulation at sites of bacterial adhesion to tissue culture cells: basis of a new diagnostic test for enteropathogenic and enterohemorrhagic Escherichia coli. Infect. Immun. 57:1290–1298.
141. Korhonen, T., V. Vaisanen-Rhen, M. Rhen, A. Pere, J. Parkkinen, and J. Finne. 1984. Escherichia coli fimbriae recognizing sialyl galactosides. J. Bacteriol. 159:762–766.
142. Korhonen, T. K., R. Virkola, and H. Holthofer. 1986. Localization of binding sites for purified Escherichia coli P fimbriae in the human kidney. Infect. Immun. 54:328–332.
143. Krivan, H. C., D. D. Roberts, and V. Ginsburg. 1988. Many pulmonary pathogenic bacteria bind specifically to the carbohydrate sequence GalNacβ1-4Gal found in some glycolipids. Proc. Natl. Acad. Sci. USA 85:6157–6161.
144. Krogfelt, K. A., H. Bergmans, and P. Klemm. 1990. Direct evidence that the FimH protein is the mannose specific adhesin of Escherichia coli type 1 fimbriae. Infect. Immun. 58:1995–1999.
145. Kuehn, M. J., J. Heuser, S. Normark, and S. J. Hultgren. 1992. P pili in uropathogenic E. coli are composite fibres with distinct fibrillar adhesive tips. Nature (London) 356:252–255.
146. Kuehn, M. J., S. Normark, and S. J. Hultgren. 1991. Immunoglobulin-like PapD chaperone caps and uncaps interactive surfaces of nascently translocated pilus subunits. Proc. Natl. Acad. Sci. USA 88:10586–10590.
147. Kuehn, M. J., D. J. Ogg, J. Kihlberg, L. N. Slonim, K. Flemmer, T. Bergfors, and S. J. Hultgren. 1993. Structural basis of pilus subunit recognition by the PapD chaperone. Science 262:1234–1241.
148. Kukkonen, M., T. Raunio, R. Virkola, K. Lahteenmaki, P. H. Makela, P. Klemm, S. Clegg, and T. K. Korhonen. 1993. Basement membrane carbohydrate as a target for bacterial adhesion: binding of type 1 fimbriae of Salmonella enterica and Escherichia coli to laminin. Mol. Microbiol. 7:229–237.
149. Kunin, C. M., H. H. Tong, C. Krishnan, L. Van Arsdale White, and J. Hacker. 1993. Isolation of a nicotinamide-requiring clone of Escherichia coli O18:K1:H7 from women with acute cystitis: resemblance to strains found in neonatal meningitis. Clin. Infect. Dis. 16:412–416.
150. LaMantia, M., and W. J. Lennarz. 1993. The essential function of yeast protein disulfide isomerase does not reside in its isomerase activity. Cell 74:899–908.
151. Latham, R. H., and W. E. Stamm. 1984. Role of fimbriated Escherichia coli in urinary tract infections in adult women: correlation with localization studies. J. Infect. Dis. 149:835–840.
152. Le Bouguenec, C., M. I. Garcia, V. Ouin, J.-M. Desperrier, P. Gounon, and A. Labigne. 1993. Characterization of plasmid-borne afa-3 gene clusters encoding afimbrial adhesins expressed by Escherichia coli strains associated with intestinal or urinary tract infections. Infect. Immun. 61:5106–5514.
153. Lecker, S., A. J. M. Driessen, and W. Wickner. 1990. ProOmpA contains secondary and tertiary structure prior to translocation and is shielded from aggregation by association with SecB. EMBO J. 9:2309–2314.
154. Lecker, S., R. Lill, T. Ziegelhoffer, C. Georgopoulos, P. J. Bassford, C. A. Kumamoto, and W. Wickner. 1989. Three pure chaperone proteins of Escherichia coli—SecB, trigger factor and GroEL—form soluble complexes with precursor proteins in vitro. EMBO J. 8:2703–2709.
155. Leffler, H., and C. Svanborg-Eden. 1980. Chemical identification of a glycosphingolipid receptor for Escherichia coli attaching to human urinary tract epithelial cells and agglutinating human erythrocytes. FEMS Microbiol. Lett. 8:127–134.
156. Leong, J. M., R. S. Fournier, and R. R. Isberg. 1990. Identification of the integrin binding domain of the Yersinia pseudotuberculosis invasin protein. EMBO J. 9:1979–1989.
157. Lesk, A. M., and C. Chothia. 1982. Evolution of proteins formed by b-sheets. II The core of the immunoglobulin domains. J. Mol. Biol. 160:325–342.
158. Levine, M. 1987. Escherichia coli that cause diarrhea: enterotoxigenic, enteropathogenic, enteroinvasive, enterohemorrhagic and enteroadherent. J. Infect. Dis. 155:377–389.
159. Levine, M. M., P. Ristaino, G. Marley, C. Smyth, S. Knutton, E. Boedeker, R. Black, C. Young, M. L. Clements, C. Cheney, and R. Patnaik. 1984. Coli surface antigens 1 and 3 of colonization factor antigen II-positive enteritoxigenic Escherichia coli: morphology, purification, and immune responses in humans. Infect. Immun. 44:409–429.
160. Lidefelt, K.-J., I. Bollgren, G. Kallenius, and S. B. Svenson. 1987. P-fimbriated Escherichia coli in children with cystitis. Acta Paediatr. Scand. 76:775–780.
161. Lindberg, F., B. Lund, L. Johansson, and S. Normark. 1987. Localization of the receptor-binding protein adhesin at the tip of the bacterial pilus. Nature (London) 328:84–87.
162. Lindberg, F., J. M. Tennent, S. J. Hultgren, B. Lund, and S. Normark. 1989. PapD, a periplasmic transport protein in P-pilus biogenesis. J. Bacteriol. 171:6052–6058.
163. Lindberg, F. P., B. Lund, and S. Normark. 1984. Genes of pyelonephritogenic E. coli required for digalactoside-specific agglutination of human cells. EMBO J. 3:1167–1173.
164. Lindberg, F. P., B. Lund, and S. Normark. 1986. Gene products specifying adhesion of uropathogenic Escherichia coli are minor components of pili. Proc. Natl. Acad. Sci. USA 83:1891–1895.
165. Lindler, L. E., and B. D. Tall. 1993. Yersinia pestis pH 6 antigen forms fimbriae and is induced by intracellular association with macrophages. Mol. Microbiol. 8:311–324.
166. Lindstedt, R., G. Larson, P. Falk, H. Leffler, and C. Svanborg. 1991. Receptor repertoire defining the host range for attaching Escherichia coli recognizing globo-A. Infect. Immun. 59:1086–1092.
167. Lintermans, P. F., P. Pohl, A. Bertels, G. Charlier, J. Vandekerckhove, J. Van Damme, J. Shoup, C. Schlicker, T. Korhonen, H. De Greve, and M. Van Montagu. 1988. Characterization and purification of the F17 adhesin on the surface of bovine enteropathogenic and septicemic Escherichia coli. Am. J. Vet. Res. 49:1794–1799.
168. Lochman, H. A., and R. I. Curtiss. 1992. Isolation and characterization of conditional adherent and non-type 1 fimbriated Salmonella typhimurium mutants. Mol. Microbiol. 6:933–945.
169. Locht, C., M.-C. Geoffroy, and G. Renauld. 1992. Common accessory genes for the Bordetella pertussis filamentous hemagglutinin and fimbriae share sequence similarities with the papC and papD gene families. EMBO J. 11:3175–3183.
170. Lowe, M. A., S. C. Holt, and B. I. Eisenstein. 1987. Immunoelectron microscopic analysis of elongation of type 1 fimbriae in Escherichia coli. J. Bacteriol. 169:157–163.
171. Lund, B., F. Lindberg, B. I. Marklund, and S. Normark. 1987. The PapG protein is the alpha-D-galactopyranosyl-(1-4)-beta-D-galactopyranose-binding adhesin of uropathogenic Escherichia coli. Proc. Natl. Acad. Sci. USA 84:5898–5902.
172. Lund, B., B. I. Marklund, N. Stromberg, F. Lindberg, K. A. Karlsson, and S. Normark. 1988. Uropathogenic Escherichia coli can express serologically identical pili of different receptor binding specificities. Mol. Microbiol. 2:255–263.
173. Mabeck, C. E., F. Orskov, and I. Orskov. 1971. Escherichia coli serotypes and renal involvement in urinary tract infection. Lancet i:1312–1314.
174. Macnab, R. M., and D. J. DeRosier. 1988. Bacterial flagellar structure and function. Can. J. Microbiol. 34:442–451.
175. Marklund, B. I., J. M. Tennent, E. Garcia, A. Hamers, M. Baga, F. Lindberg, W. Gaastra, and S. Normark. 1992. Horizontal gene transfer of the Escherichia coli pap and prs pili operons as a mechanism for the development of tissue-specific adhesive properties. Mol. Microbiol. 6:2225–2242.
176. Marrs, C. F. 1994. Type 4 pili in the families Moraxellaceae and Neisseriaceae, p. 127–143. In V. L. Miller, J. B. Kaper, D. A. Portnoy, and R. R. Isberg (ed.), Molecular Genetics of Bacterial Pathogenesis. ASM Press, Washington D.C.
177. Martin, P. R., M. Hobbs, P. D. Free, Y. Jeske, and J. S. Mattick. 1993. Characterization of PilQ, a new gene required for the biogenesis of type 4 fimbriae in Pseudomonas aeruginosa. Mol. Microbiol. 7:669.
178. Massad, G., C. V. Lockatell, D. E. Johnson, and H. L. T. Mobley. 1994. Proteus mirabilis fimbriae: construction of an isogenic pmfA mutant and analysis of virulence in a CBA mouse model of ascending urinary tract infection. Infect. Immun. 62:536–542.
179. Matsuyama, S., Y. Fujita, and S. Mizushima. 1993. SecD is involved in the release of translocated secretory proteins from the cytoplasmic membrane of Escherichia coli. EMBO J. 12:265–270.
180. Maurer, L., and P. E. Orndorff. 1985. A new locus, pilE, required for the binding of type 1 piliated Escherichia coli to erythrocytes. FEMS Microbiol. Lett. 30:59–66.
181. Maurer, L., and P. E. Orndorff. 1987. Identification and characterization of genes determining receptor binding and pilus length of Escherichia coli type 1 pili. J. Bacteriol. 169:640–645.
182. Metcalfe, J. W., K. A. Krogfelt, H. C. Krivan, P. S. Cohen, and D. C. Laux. 1991. Characterization and identification of a porcine small intestine mucus receptor for the K88ab fimbrial adhesin. Infect. Immun. 59:91–96.
183. Meyer, T. F., C. P. Gibbs, and R. Haas. 1990. Variation and control of protein expression in Neisseria. Annu. Rev. Microbiol. 44:451–477.
184. Miller, V. L., J. B. Kaper, D. A. Portnoy, and R. R. Isberg (ed.). 1994. Molecular Genetics of Bacterial Pathogenesis. ASM Press, Washington D.C.
185. Minion, F. C., S. N. Abraham, E. H. Beachey, and J. D. Goguen. 1989. The genetic determinant of adhesive function in type 1 fimbriae of Escherichia coli is distinct from the gene encoding the fimbrial subunit. J. Bacteriol. 165:1033–1036.
186. Mirelman, D. (ed.). 1986. Microbial Lectins and Agglutinins: Properties and Biological Activity. John Wiley & Sons, New York.
187. Missiakas, D., C. Georgopoulos, and S. Raina. 1994. The Escherichia coli dsbC(xprA) gene encodes a periplasmic protein involved in disulfide bond formation. EMBO J. 13:2013–2020.
188. Mobley, H., K. Jarvis, J. Elwood, D. Whittle, V. Lockatell, R. Russel, D. Johnson, M. Donnenberg, and J. Warren. 1993. Isogenic P-fimbriated deletion mutants of pyelonephritogenic Escherichia coli: the role of αGal(1-4)βGal binding in virulence of a wild-type strain. Mol. Microbiol. 10:143–155.
189. Moch, T., H. Hoschutzky, J. Hacker, K.-D. Kroncke, and K. Jann. 1987. Isolation and characterization of the a-sialyl-b-2,3-galactosyl-specific adhesin from fimbriated Escherichia coli. Proc. Natl. Acad. Sci. USA 84:3462–3466.
190. Mol, O., R. W. Vischers, F. K. de Graaf, and B. Oudega. 1994. Escherichia coli periplasmic chaperone FaeE is a homodimer and the chaperone-K88 subunit complex is a heterotrimer. Mol. Microbiol. 11:391–402.
191. Mooi, F. R., W. H. Jansen, H. Brunings, H. Gielen, H. G. J. van der Heide, H. C. Walvoort, and P. A. M. Guniee. 1992. Construction and analysis of Bordetella pertussis mutants defective in the production of fimbriae. Microb. Pathog. 12:127–135.
192. Mooi, F. R., H. G. J. van der Heide, H. C. Walvoort, H. Brunings, W. H. Jansen, and P. A. M. Guinee. 1991. The analysis of Bordetella pertussis fimbrial mutants in a rabbit model, p. 69–76. In E. Z. Ron and S. Rottem (ed.), Microbial Surface Components and Toxins in Relation to Pathogenesis. Plenum Press, New York.
193. Moon, H. W., S. C. Whipp, M. Argenzio, M. Levine, and R. A. Giannella. 1983. Attaching and effacing activities of rabbit and human enteropathogenic Escherichia coli in pig and rabbit intestines. Infect. Immun. 41:1340–1351.
194. Morschhauser, J., H. Hoschutzky, K. Jann, and J. Hacker. 1990. Functional analysis of the sialic acid-binding adhesin SfaS of pathogenic Escherichia coli by site-specific mutagenesis. Infect. Immun. 58:2133–2138 .
195. Muller, K. H., S. K. Collinson, T. J. Trust, and W. W. Kay. 1991. Type 1 fimbriae of Salmonella enteriditis. J. Bacteriol. 173:454–457.
196. Neeser, J. R., B. Koellreutter, and P. Wuersch. 1986. Oligomannoside-type glycopeptides inhibiting adhesion of Escherichia coli strains mediated by type 1 pili: preparation of potent inhibitors from plant glycoproteins. Infect. Immun. 52:428–436.
197. Nicchitta, C. V., and G. Blobel. 1993. Lumenal proteins of the mammalian endoplasmic reticulum are required to complete protein translocation. Cell 73:989–998.
198. Nigam, S. K., A. L. Goldberg, S. Ho, M. F. Rhode, K. T. Bush, and M. Y. Sherman. 1994. A set of endoplasmic reticulum proteins possessing properties of molecular chaperones includes Ca2+-binding protein and members of the thioredoxin superfamily. J. Biochem. 269:1744–1749.
199. Nilsson, P., and B. E. Uhlin. 1991. Differential decay of a polycistronic Escherichia coli transcript is initiated by RNaseE-dependent endonucleolytic processing. Mol. Microbiol. 5:1791–1799.
200. Norgren, M., M. Baga, J. M. Tennent, and S. Normark. 1987. Nucleotide sequence, regulation and functional analysis of the papC gene required for cell surface localization of Pap pili of uropathogenic Escherichia coli. Mol. Microbiol. 1:169–178.
201. Norgren, M., S. Normark, D. Lark, P. O’Hanley, G. Schoolnik, S. Falkow, C. Svanborg-Eden, M. Baga, and B. E. Uhlin. 1984. Mutations in E. coli cistrons affecting adhesion to human cells do not abolish Pap pili fiber formation. EMBO J. 3:1159–1165.
202. Normark, S., M. Baga, M. Goransson, F. P. Lindberg, B. Lund, M. Norgren, and B. E. Uhlin. 1986. Genetics and biogenesis of Escherichia coli adhesins, p. 113–143. In D. Mirelman (ed.), Microbial Lectins and Agglutinins: Properties and Biological Activity. John Wiley & Sons, New York.
203. Nowicki, B., A. Labigne, S. Moseley, R. Hull, S. Hull, and J. Moulds. 1990. The Dr hemagglutinin, afimbrial adhesins AFA-I and AFA-III, and F1845 fimbriae of uropathogenic and diarrhea-associated Escherichia coli belong to a family of hemagglutinins with Dr receptor recognition. Infect. Immun. 58:279–281.
204. Nowicki, B., J. Moulds, R. Hull, and S. Hull. 1988. A hemagglutinatinin of uropathogenic Escherichia coli recognizes the Dr blood group antigen. Infect. Immun. 56:1057–1060.
205. Nunn, D., S. Bergman, and S. Lory. 1990. Products of three accessory genes, pilB, pilC, and pilD, are required for biogenesis of Pseudomonas aeruginosa pili. J. Bacteriol. 172:2911–2919.
206. Nunn, D. N., and S. Lory. 1991. Product of the Pseudomonas aeroginosa gene pilD is a prepilin peptidase. Proc. Natl. Acad. Sci. USA 88:3281.
207. Nunn, D. N., and S. Lory. 1992. Components of the protein-excretion apparatus of Pseudomonas aeruginosa are processed by the type IV prepilin peptidase. Proc. Natl. Acad. Sci. USA 89:47–51.
208. Ofek, I., and E. H. Beachey 1980. General concepts and principles of bacterial adherence, p. 1–29. In E. H. Beachey (ed.), Bacterial Adherence, Receptors and Recognition. Chapman & Hall, London.
209. Ofek, I., D. Mirelman, and N. Sharon. 1977. Adherence of Escherichia coli to human mucosal cells mediated by mannose receptors. Nature (London) 265:623–625.
210. Old, D. C., and R. A. Adegbola. 1983. A new mannose-resistant haemagglutinatinin in Klebsiella. J. Appl. Bacteriol. 55:165.
211. Old, D. C., A. Tavendale, and B. W. Senior. 1985. A comparative study of the type 3 fimbriae of Klebsiella species. J. Med. Microbiol. 20:203.
212. Olsen, A., A. Arnqvist, M. Hammar, S. Sukupolvi, and S. Normark. 1993. The rpoS sigma factor relieved H-NS mediated transcriptional repression of csgA, the subunit gene for fibronectin binding curli of Escherichia coli. Mol. Microbiol. 7:523–536.
213. Olsen, A., A. Jonsson, and S. Normark. 1989. Fibronectin binding mediated by a novel class of surface organelles on Escherichia coli. Nature (London) 338:652–655.
214. Ono, E., K. Abe, M. Nakazawa, and M. Naiki. 1989. Ganglioside epitope recognized by K99 fimbriae from enterotoxigenic Escherichia coli. Infect. Immun. 57:907–911.
215. Orndorff, P. E. 1994. Escherichia coli type 1 pili, p. 91–111. In V. L. Miller, J. B. Kaper, D. A. Portnoy, and R. R. Isberg (ed.), Molecular Genetics of Bacterial Pathogenesis. ASM Press, Washington, D.C.
216. Orndorff, P. E., and C. A. Bloch. 1990. The role of type 1 pili in the pathogenesis of Escherichia coli infections: a short review and some new ideas. Microb. Pathog. 9:75–79.
217. Orndorff, P. E., and S. Falkow. 1984. Organization and expression of genes responsible for type 1 piliation in Escherichia coli. J. Bacteriol. 159:736–744.
218. Ørskov, I., A. Birch-Anderson, J. P. Duguid, J. Stenderup, and F. Ørskov. 1985. An adhesive protein capsule of Escherichia coli. Infect. Immun. 47:191–200.
219. Ørskov, I., F. Ørskov, and A. Birch-Andersen. 1980. Comparison of Escherichia coli fimbrial antigen F7 with type 1 fimbriae. Infect. Immun. 27:657–666.
220. Otsu, M., F. Omura, T. Yoshimori, and M. Kikuchi. 1994. Protein disulfide isomerase associates with misfolded human lysozyme in vivo. J. Biochem. 269:6874–6877.
221. Ottow, J. C. G. 1975. Ecology, physiology and genetics of fimbriae and pili. Annu. Rev. Microbiol. 29:79.
222. Pan, K.-M., M. Baldwin, J. Nguyen, M. Gasset, A. Serban, D. Groth, I. Mehlorn, H. Ziwei, R. J. Fletterick, F. E. Cohen, and S. B. Prusiner. 1993. Conversion of a-helices into β-sheets features in the formation of the scrapie prion proteins. Proc. Natl. Acad. Sci. USA 90:10962–10966.
223. Paranchych, W. 1990. Molecular studies of N-methylphenylalanine pili, p. 61–74. In B. H. Iglewski and V. I. Clark (ed.), Molecular Basis of Bacterial Pathogenesis, vol. 2. Academic Press, Inc., New York.
224. Paranchych, W., and L. S. Frost. 1988. The physiology and biochemistry of pili. Adv. Microb. Physiol. 29:53–114.
225. Paranchych, W., P. A. Sastry, K. Volpel, B. A. Loh, and D. Speert. 1986. Fimbriae (pili): molecular basis of Pseudomonas aeruginosa adherence. Clin. Invest. Med. 9:113–118.
226. Peek, J. A., and R. K. Taylor. 1992. Characterization of a periplasmic thiol:disulfide interchange protein required for the functional maturation of secreted virulence factors of Vibrio cholerae. Proc. Natl. Acad. Sci. USA 89:6210–6214.
227. Perez-Casal, J., J. S. Swartley, and J. R. Scott. 1990. Gene encoding the major subunit of CS1 pili of human enterotoxigenic Escherichia coli. Infect. Immun. 58:3594–3600.
228. Perutz, M. F., T. Johnson, M. Suzuki, and J. T. Finch. 1994. Glutamine repeats as polar zippers: their possible role in inherited neurodegenerative diseases. Proc. Natl. Acad. Sci. USA 91:5355–5358.
229. Plos, K., T. Carter, S. Hull, R. Hull, and C. Svanborg-Eden. 1990. Frequency and organization of pap homologous DNA in relation to clinical origin of uropathogenic Escherichia coli. J. Infect. Dis. 161:518–524.
230. Poljak, R. J., L. M. Amzel, H. P. Avey, B. L. Chen, R. P. Phizackerley, and F. Saul. 1973. Three dimensional structure of the Fab fragment of a human immunoglobulin at 2.8A resolution. Proc. Natl. Acad. Sci. USA 70:3305–3310.
231. Ponniah, S., R. O. Endres, D. L. Hasty, and S. N. Abraham. 1991. Fragmentation of Escherichia coli type 1 fimbriae exposes cryptic d-mannose-binding sites. J. Bacteriol. 173:4195–4202.
232. Prasadarao, N. V., C. A. Wass, J. Hacker, K. Jann, and K. S. Kim. 1993. Adhesion of S-fimbriated Escherichia coli to brain glycolipids mediated by sfaA gene encoded protein of S-fimbriae. J. Biol. Chem. 268:10356–10363.
233. Pugsley, A. 1993. The complete general secretory pathway in gram-negative bacteria. Microbiol. Rev. 57:50–108.
234. Pugsley, A. P. 1992. Protein traffic in bacteria, p. 461–479. In J. Lederberg (ed.), Encyclopedia of Microbiology, vol. 3. Academic Press Inc., New York.
235. Puig, A., and H. Gilbert. 1994. Anti-chaperone behaviour of Bip during the protein disulfide isomerase-catalyzed refolding of reduced denatured lysozyme. J. Biochem. 269:25889–25896.
236. Rhen, M., P. Klemm, and T. K. Korhonen. 1986. Identification of two new hemagglutinins of Escherichia coli, N-acetyl-d-glucosamine-specific fimbriae and a blood group M-specific agglutinin, by cloning the corresponding genes in Escherichia coli K-12. J. Bacteriol. 168:1234–1242.
237. Rhen, M., V. Vaisanen-Rhen, M. Sarasta, and T. K. Korhonen. 1986. Organization of genes expressing the blood group M-specific hemagglutinin of Escherichia coli: identification and nucleotide sequence of the M-agglutinin subunit gene. Gene 49:351.
238. Riegman, N., R. Kusters, H. van Veggel, H. Bergmans, P. Van Bergen en Henegouwen, J. Hacker, and I. Van Die. 1990. F1C fimbriae of a uropathogenic Escherichia coli strain: genetic and functional organization of the foc gene cluster and identification of minor subunits. J. Bacteriol. 172:1114–1120.
239. Roberts, J. A., K. Hardaway, B. Kaack, E. N. Fussell, and G. Baskin. 1984. Prevention of pyelonephritis by immunization with P-fimbriae. J. Urol. 131:602–607.
240. Roberts, J. A., B. Kaack, G. Kallenius, R. Mollby, J. Winberg, and S. B. Svenson. 1984. Receptors for pyelonephritogenic Escherichia coli in primates. J. Urol. 131:163–168.
241. Roberts, J. A., B.-I. Marklund, D. Ilver, D. Haslam, M. B. Kaack, G. Baskin, M. Louis, R. Mollby, J. Winberg, and S. Normark. 1994. The Gal α(1-4) Gal-specific tip adhesin of Escherichia coli P-fimbriae is needed for pyelonephritis to occur in the normal urinary tract. Proc. Natl. Acad. Sci. USA 91:11889–11893.
242. Rudel, T., I. Scheuerpflug, and T. F. Meyer. 1995. Neisseria PilC protein identified as type 4 pilus tip-located adhesin. Nature (London) 373:357–359.
243. Russell, P. W., and P. E. Orndorff. 1992. Lesions in two Escherichia coli type 1 pilus genes alter pilus number and length without affecting receptor binding. J. Bacteriol. 174:5923–5935.
244. Ryu, S. E., P. D. Kwong, A. Truneh, T. G. Porter, J. Arthos, M. Rosenberg, X. Dai, N.-H. Xuong, R. Axel, R. W. Sweet, and W. A. Hendrickson. 1990. Crystal structure of an HIV-binding recombinant fragment of human CD4. Nature (London) 348:419–426.
245. Salit, I. E., and E. C. Gotschlich. 1977. Type 1 Escherichia coli pili: characterization of binding to monkey kidney cells. J. Exp. Med. 146:1182–1194.
246. Sanders, S. L., K. M. Whitfield, J. P. Vogel, M. D. Rose, and R. W. Scheckman. 1992. Sec61p and BiP directly facilitate polypeptide translocatin into the ER. Cell 69:353–365.
247. Savarino, S., P. Fox, D. Yikang, and J. P. Nataro. 1994. Identification and characterization of a gene cluster mediating enteroaggregative Escherichia coli aggregative adherence fimbriae I biogenesis. J. Bacteriol. 176:4949–4957.
248. Scaletsky, I. C. A., M. L. M. Silva, and L. R. Trabulsi. 1984. Distinctive patterns of adherence of enteropathogenic Escherichia coli to HeLa cells. Infect. Immun. 45:534–536.
249. Schiffer, M., R. L. Girling, K. R. Ely, and A. B. Edmundson. 1973. Structure of a l-type Bence Jones protein at 3.5A resolution. Biochemistry 12:4620–4631.
250. Schmidt, M. A. 1994. Non-fimbrial adhesins of Escherichia coli, p. 85–97. In P. Klemm (ed.), Fimbriae: Adhesins, Genetics, Biogenesis, and Vaccines. CRC Press, Ann Arbor, Mich.
251. Schmoll, T., H. Hoschutzky, J. Morschhauser, F. Lottspeich, K. Jann, and J. Hacker. 1989. Analysis of genes coding for the sialic acid-binding adhesin and two other minor fimbrial subunits of the S-fimbrial adhesin determinant of Escherichia coli. Mol. Microbiol. 3:1735–1744.
252. Schmoll, T., J. Morschhauser, M. Ott, B. Ludwig, I. van Die, and J. Hacker. 1990. Complete genetic organization and functional aspects of the Escherichia coli S fimbrial adhesion determinant: nucleotide sequence of the genes sfa B, C, D, E, F. Microb. Pathog. 9:331–343.
253. Schneider, H.-C., J. Berthold, M. F. Bauer, K. Dietmeier, B. Guiard, M. Brunner, and W. Neupert. 1994. Mitochondrial Hsp70/MIM44 complex facilitates protein import. Nature (London) 371:768–774.
254. Schnell, D. J., F. Kessler, and G. Blobel. 1994. Isolation of components of the chloroplast protein import machinery. Science 266:1007–1012.
255. Schoolnik, G. 1989. How Escherichia coli infects the urinary tract. N. Engl. J. Med. 320:804–805.
256. Scott, J. R. 1990. The M protein of group A Streptococcus: evolution and regulation, p. 177–199. In B. H. Iglewski and V. I. Clark (ed.), Molecular Basis of Bacterial Pathogenesis. Academic Press, Inc., New York.
257. Scott, J. R., J. C. Wakefield, P. Russel, P. E. Orndorff, and B. J. Froehlich. 1992. CooB is required for assembly but not transport of CS1 pilin. Mol. Microbiol. 6:293–300.
258. Sharon, N. 1987. Bacterial lectins, cell-cell recognition and infectious disease. FEBS Lett. 217:145–157.
259. Shaw, C. E., and R. K. Taylor. 1990. Vibrio cholerae O395 tcpA pilin gene sequence and comparison of predicted protein structural features to those of type 4 pilins. Infect. Immun. 58:3042–3049.
260. Simons, B., P. Rathman, C. Malij, B. Oudega, and F. K. de Graaf. 1990. The penultimate tyrosine residue of the K99 fibrillar subunit is essential for the stability of the protein and its interaction with the periplasmic carrier protein. FEMS Microbiol. Lett. 67:107–112.
261. Sjobring, U., G. Pohl, and A. Olsen. 1994. Plasminogen, absorbed by Escherichia coli expressing curli or by Salmonella enteriditis expressing thin aggregative fimbriae, can be activated by simultaneously captured tissue-type plasminogen activator (t-PA). Mol. Microbiol. 14:443–452.
262. Slonim, L. N., J. S. Pinkner, C. I. Branden, and S. J. Hultgren. 1992. Interactive surface in the PapD chaperone cleft is conserved in pilus chaperone superfamily and essential in subunit recognition and assembly. EMBO J. 11:4747–4756.
263. Smit, H., W. Gaastra, J. P. Kamerling, J. F. G. Vliegenthart, and F. K. de Graaf. 1984. Isolation and structural characterization of the equine erythrocyte receptor for enterotoxigenic Escherichia coli K99 fimbrial adhesin. Infect. Immun. 46:578–584.
264. Sohel, I., J. L. Puente, W. J. Murray, J. Vuopio-Varkila, and G. K. Schnoolnik. 1993. Cloning and characterization of the bundle-forming pilin gene of enteropathogenic Escherichia coli and its distribution in Salmonella serotypes. Mol. Microbiol. 7:563.
265. Sokurenko, E. V., H. S. Courtney, S. N. Abraham, P. Klemm, and D. L. Hasty. 1992. Functional heterogeneity of type 1 fimbriae of Escherichia coli. Infect. Immun. 60:4709–4719.
266. Sokurenko, E. V., H. S. Courtney, D. E. Ohman, P. Klemm, and D. L. Hasty. 1994. FimH family of type 1 fimbrial adhesins: functional heterogeneity due to minor sequence variations among fimH genes. J. Bacteriol. 176:748–755.
267. Strauch, K. L., and J. Beckwith. 1988. An Escherichia coli mutation preventing degradation of abnormal periplasmic proteins. Proc. Natl. Acad. Sci. USA 85:1576–1580.
268. Strauch, K. L., K. Johnson, and J. Beckwith. 1989. Characterization of degP, a gene required for proteolysis in the cell envelope and essential for growth of Escherichia coli at high temperature. J. Bacteriol. 171:2689–2696.
269. Striker, R., F. Jacob-Dubuisson, C. Frieden, and S. J. Hultgren. 1994. Stable fiber forming and non-fiber forming chaperone-subunit complexes in pilus biogenesis. J. Biol. Chem. 269:12233–12239.
270. Striker, R., U. Nilsson, A. Stonecipher, G. Magnusson, and S. J. Hultgren. 1995. Structural requirements for the glycolipid receptor of human uropathogenic E. coli. Mol. Microbiol.16:1021–1029.
271. Strom, M. S., D. Nunn, and S. Lory. 1991. Multiple roles of the pilus biogenesis protein PilD: involvement of PilD in excretion of enzymes from Pseudomonas aeruginosa. J. Bacteriol. 173:1175–1180.
272. Strom, M. S., D. N. Nunn, and S. Lory. 1993. A single bifunctional enzyme, PilD, catalyzes cleavage and N-methylation of proteins belonging to the type IV pilin family. Proc. Natl. Acad. Sci. USA 90:2404.
273. Stromberg, N., S. J. Hultgren, D. G. Russell, and S. Normark. 1992. Microbial attachment, molecular mechanisms, p. 1–16. In J. Lederberg (ed.), Encyclopedia of Microbiology, vol. 3. Academic Press, Inc., New York.
274. Stromberg, N., B. I. Marklund, B. Lund, D. Ilver, A. Hamers, W. Gaastra, K. A. Karlsson, and S. Normark. 1990. Host-specificity of uropathogenic Escherichia coli depends on differences in binding specificity to Galα(1-4)Gal-containing isoreceptors. EMBO J. 9:2001–2010.
275. Stromberg, N., P.-G. Nyholm, I. Pascher, and S. Normark. 1991. Saccharide orientation at the cell surface affects glycolipid receptor function. Proc. Natl. Acad. Sci. USA 88:9340–9344.
276. Svanborg, C., F. Orskov, and I. Orskov. 1994. Fimbriae and disease, p. 239–255. In P. Klemm (ed.), Fimbriae: Adhesion, Genetics, Biogenesis and Vaccines. CRC Press, Ann Arbor, Mich.
277. Svanborg, E. C., B. Andersson, G. Aniansson, H. Leffler, H. Lomberg, J. Mestecky, and A. E. Wold. 1987. Glycoconjugate receptors for bacteria attaching to mucosal sites: examples for Escherichia coli and Streptococcus pneumoniae. Adv. Exp. Med. Biol. 216:931.
278. Svanborg-Eden, C., B. Andersson, L. Hagberg, L. A. Hanson, H. Leffler, G. Magnusson, G. Noori, J. Dahmen, and T. Soderstrom. 1983. Receptor analogues and anti-pili antibodies as inhibitors of bacterial attachment in vivo and in vitro. Ann. N.Y. Acad. Sci. 409:580–591.
279. Svanborg-Eden, C., B. Eriksson, L. A. Hanson, U. Jodal, B. Kaijser, G. Lidin-Janson, O. Lindberg, and S. Olling. 1978. Adhesion to normal human uroepithelial cells of Escherichia coli from children with various forms of urinary tract infection. J. Pediatr. 93:398.
280. Svanborg-Eden, C., L. A. Hanson, U. Jodal, U. Lindberg, and A. Sohl-Akerlund. 1976. Variable adherence to normal urinary tract epithelial cells of Escherichia coli strains associated with various forms of urinary tract infection. Lancet ii:490.
281. Tennent, J. M., S. Hultgren, K. Forsman, M. Goransson, B. I. Marklund, B. E. Uhlin, and S. Normark. 1990. Genetics of adhesin expression in Escherichia coli, p. 79–110. In B. H. Iglewski and V. I. Clark (ed.), Molecular Basis of Bacterial Pathogenesis, vol. 11. Academic Press, Inc., New York.
282. Tennent, J. M., F. Lindberg, and S. Normark. 1990. Integrity of Escherichia coli P pili during biogenesis: properties and role of PapJ. Mol. Microbiol. 4:747–758.
283. Tennent, J. M., and J. S. Mattick. 1994. Type 4 pili, p. 127–147. In P. Klemm (ed.), Fimbriae: Adhesion, Genetics, Biogenesis and Vaccines. CRC Press, Ann Arbor, Mich.
284. Tewari, R., J. I. MacGregor, T. Ikeda, J. R. Little, S. J. Hultgren, and S. N. Abraham. 1993. Neutrophil activation by nascent FimH subunits of type 1 fimbrae purified from the periplasm of Escherichia coli. J. Biol. Chem. 268:3009–3015.
285. Uhlin, B. E., M. Nogren, M. Baga, and S. Normark. 1985. Adhesion to human cells by Escherichia coli lacking the major subunit of a digalactoside-specific pilus. Proc. Natl. Acad. Sci. USA 82:1800–1804.
286. Ungermann, C., W. Neupert, and D. M. Cyr. 1994. The role of Hsp70 in conferring unidirectionality on protein translocation into mitochondria. Science 266:1250–1253.
287. Vaisanen, V., J. Elo, L. G. Tallgren, A. Siitonen, P. H. Makela, C. Svanborg-Eden, G. Kallenius, S. B. Svenson, H. Hultberg, and T. Korhonen. 1981. Mannose-resistant hemagglutination and P antigen recognition are characteristics of Escherichia coli causing primary pyelonephritis. Lancet ii:1366–1369.
288. van Alphen, L., L. Geelen-van Den Broek, L. Blaas, M. van Ham, and J. Dankert. 1991. Blocking of fimbria-mediated adherence of Haemophilus influenzae by sialyl gangliosides. Infect. Immun. 59:4473–4477.
289. van der Woude, M. W., B. A. Braaten, and D. A. Low. 1992. Evidence of global regulatory control of pilus expression in Escherichia coli by Lrp and DNA methylation: model building based on analysis of pap. Mol. Microbiol. 6:2429–2435.
290. Van Die, I., C. Krammer, J. Hacker, H. Bergmans, W. Jongen, and W. Hoekstra. 1991. Nucleotide sequence of the genes coding for the minor fimbrial subunits of the F1C fimbriae of Escherichia coli. Res. Microbiol. 142:653–658.
291. Van Gijsegem, F., S. Genin, and C. Boucher. 1993. Conservation of secretion pathways for pathogenicity determinants of plant and animal bacteria. Trends Microbiol. 1:161–163.
292. Venegas, M. F., E. N. Navas, R. A. Gaffney, J. L. Duncan, B. E. Anderson, and A. J. Schaeffer. 1995. Binding of type 1-piliated Escherichia coli to vaginal mucus. Infect. Immun. 63:416–422.
293. Weis, W., J. H. Brown, S. Cusack, J. C. Paulson, J. J. Skehel, and D. C. Wiley. 1988. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature (London) 333:426–431.
294. Westerlund, B., I. van Die, C. Kramer, P. Kuusela, H. Holthofer, A. M. Tarkkanen, R. Virkola, N. Riegman, H. Bergmans, W. Hoekstra, and T. K. Korhonen. 1991. Multifunctional nature of P fimbriae of uropathogenic Escherichia coli: mutations in fsoE and fsoF influence fimbrial binding to renal tubuli and immobilized fibronectin. Mol. Microbiol. 5:2965–2975.
295. Whitchurch, C. B., M. Hobbs, S. P. Livingston, V. Krishnapillai, and J. S. Mattick. 1991. Characterization of a Pseudomonas aeruginosa twitching motility gene and evidence for a specialized protein export system widespread in eubacteria. Gene 101:33.
296. Whitchurch, C. B., and J. S. Mattick. 1994. Escherichia coli contains a set of genes homologous to those involved in protein secretion, DNA uptake and the assembly of type-4 fimbrae in other bacteria. Gene 1504:9–15.
297. Wickner, W., A. J. M. Driessen, and F.-U. Hartl. 1991. The enzymology of protein translocation across the Escherichia coli plasma membrane. Annu. Rev. Biochem. 60:101–124.
298. Willems, R. J. L., H. G. J. van der Heide, and F. R. Mooi. 1992. Characterization of a Bordetella pertussis fimbrial gene cluster which is located directly downstream of the filamentous haemagglutinin gene. Mol. Microbiol. 6:2661–2671.
299. Williams, A. F., and A. N. Barclay. 1988. The immunoglobulin superfamily-domains for cell surface recognition. Annu. Rev. Immunol. 6:381–405.
299a. Winberg, J., R. Möllby, J. Bergström, K.-A. Karlsson, I. Leonardsson, M. Milh, S. Teneberg, D. Haslam, B.-I. Marklund, and S. Normark. 1995. The PapG-adhesin at the tip of P-fimbriae provides Escherichia coli with a competitive edge in experimental bladder infections of cynomolgus monkeys. J. Exp. Med. 182:1695–1702.
300. Wray, S. K., S. I. Hull, R. G. Cook, J. Barrish, and R. A. Hull. 1986. Identification and characterization of a uroepithelial cell adhesin from a uropathogenic isolate of Proteus mirabilis. Infect. Immun. 54:43–49.
301. Xu, Z., C. H. Jones, D. Haslam, K. Dodson, J. Kihlberg, and S. J. Hultgren. 1995. Molecular dissection of PapD interaction with PapG reveals two chaperone binding sites. Mol. Microbiol. 16:1011–1020.
302. Yu, J., and J. B. Kape. 1992. Cloning and characterization of the eae gene of enterohaemorrhagic Escherichia coli O157:H7. Mol. Microbiol. 6:411–417.
303. Yu, J., H. Webb, and T. R. Hirst. 1992. A homologue of the Escherichia coli DsbA protein involved in disulphide bond formation is required for enterotoxin biogenesis in Vibrio cholerae. Mol. Microbiol. 6:1949–1958.
304. Zambrano, M. M., D. A. Siegle, M. Almiron, A. Tormo, and R. Kolter. 1993. Microbial competition: Escherichia coli mutants that take over stationary phase cultures. Science 259:1757–1760.
Chapter150
