Portal de Periódicos da CAPES

Substrate recognition of type III secretion machines -testing the RNA signal hypothesis

2005; Wiley; Volume: 7; Issue: 9 Linguagem: Inglês

10.1111/j.1462-5822.2005.00563.x

1462-5822

Joseph A. Sorg, Nathan C. Miller, Olaf Schneewind,

Vibrio bacteria research studies

Cellular MicrobiologyVolume 7, Issue 9 p. 1217-1225 Free Access Substrate recognition of type III secretion machines –testing the RNA signal hypothesis Joseph A. Sorg, Joseph A. Sorg Department of Microbiology, University of Chicago, Chicago, IL 60637, USA.Search for more papers by this authorNathan C. Miller, Nathan C. Miller Department of Microbiology, University of Chicago, Chicago, IL 60637, USA.Search for more papers by this authorOlaf Schneewind, Corresponding Author Olaf Schneewind Department of Microbiology, University of Chicago, Chicago, IL 60637, USA. For correspondence. E-mail [email protected]; Tel. (+1) 773 834 9060; Fax (+1) 773 834 8150.Search for more papers by this author Joseph A. Sorg, Joseph A. Sorg Department of Microbiology, University of Chicago, Chicago, IL 60637, USA.Search for more papers by this authorNathan C. Miller, Nathan C. Miller Department of Microbiology, University of Chicago, Chicago, IL 60637, USA.Search for more papers by this authorOlaf Schneewind, Corresponding Author Olaf Schneewind Department of Microbiology, University of Chicago, Chicago, IL 60637, USA. For correspondence. E-mail [email protected]; Tel. (+1) 773 834 9060; Fax (+1) 773 834 8150.Search for more papers by this author First published: 20 July 2005 https://doi.org/10.1111/j.1462-5822.2005.00563.xCitations: 36AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Summary Secretion by the type III pathway of Gram-negative microbes transports polypeptides into the extracellular medium or into the cytoplasm of host cells during infection. In pathogenic Yersinia spp., type III machines recognize 14 different Yop protein substrates via discrete signals genetically encoded in 7–15 codons at the 5′ portion of yop genes. Although the signals necessary and sufficient for substrate recognition of Yop proteins have been mapped, a clear mechanism on how proteins are recognized by the machinery and then initiated into the transport pathway has not yet emerged. As synonymous substitutions, mutations that alter mRNA sequence but not codon specificity, affect the function of some secretion signals, recent work with several different microbes tested the hypothesis of an RNA-encoded secretion signal for polypeptides that travel the type III pathway. This review summarizes experimental observations and mechanistic models for substrate recognition in this field. Introduction Appreciation of the simple fact that proteins require active transport across biological membranes gave birth to the field of protein topogenesis (Blobel and Dobberstein, 1975; Walter and Blobel, 1980). A vast amount of experimental work over the past three decades can be summarized in reviewing the principles of the signal peptide hypothesis (Blobel, 1980). Much like a subway slot machine separating token-bearing passengers from accidental bystanders, proteins transported across membranes carry signal sequences that dictate their fate by recognition of translocation machines and simultaneous movement across lipid bilayers (Blobel, 2000). Although modifications of this hypothesis tinkered with the possibility of mRNA signals determining the destiny of proteins in cells, overwhelming experimental evidence quickly demonstrated that peptide signals represent the unifying principle of protein topogenesis (Sabatini et al., 1972). This view is now generally accepted for all three kingdoms of life (Blobel, 2000). In fact, one can draw a cellular map of membrane-enclosed organelles just like a map of ZIP codes, each organelle addressed by unique signal peptides and their cognate membrane translocators (Mellman and Warren, 2000). Type III secretion is unique in that the secretion machinery transports polypeptides simultaneously across the double membrane envelope and cell wall of Gram-negative bacteria (Michiels et al., 1990). Indeed, type III needle complexes extend 50–100 nm or more beyond the lipopolysaccharide and protein-covered bacterial surface, a phenomenon that permits interaction of secretion machines with host cells (Roine et al., 1997; Kubori et al., 1998; Mota et al., 2005; West et al., 2005). Thus, as substrates of the type III pathway are thought to travel the narrow lumen of needle complexes, these polypeptides are transported a far greater distance than those encountered by other membrane translocation machines (Jin and He, 2001; Crepin et al., 2005). Type III machines fulfil accessory roles in that bacterial survival under laboratory conditions does not depend on their function; however, they are required when host target cells are being encountered (Rosqvist et al., 1994). Not surprisingly then, expression of genes encoding type III machines and their substrates is regulated in all bacterial species and there is mounting evidence that each microbial species not only employs unique sets of genes encoding type III effector proteins, but that the order in which these polypeptides are transported is developmentally determined (Petterson et al., 1996). If so, one can safely predict an encounter of different kinds of type III secretion signals of which only some, but certainly not all, will be shared with those of other bacterial species (Cornelis, 2003). Yersinia spp. Yersinia enterocolitica and Yersinia pseudotuberculosis are food-borne pathogens that invade intestinal epithelia and replicate in associated lymphoid tissues. Yersinia pestis, the causative agent of bubonic plague, is transmitted from rodents via flea bite; however, these microbes can also replicate in lung tissues and then be transmitted directly between humans via respiratory droplets, precipitating the invariably fatal disease pneumonic plague. All three pathogenic Yersinia species depend on the type III secretion system encoded by their 70 kb virulence plasmid and transport Yops during host infection to escape innate immune responses such as phagocytosis (Cornelis et al., 1998). When these bacteria are incubated with cultured tissue cells, several type III substrates (effectors) are injected into the eukaryotic cytosol (YopE, YopH, YopM, YopN, YopO, YopP, YopT, YscM1 and YscM2) (Cornelis, 2002). Moreover, four proteins are primarily secreted by the type III pathway into the extracellular medium (LcrV, YopB, YopD and YopR) (Lee et al., 1998). Thus, one can view type III machines as a regulated pathway for tightly controlled secretion and injection reactions (Ramamurthi and Schneewind, 2002a). Yersinia spp. can be induced for type III secretion in culture medium even without host target cells (Michiels et al., 1990). When yersiniae are grown in rich laboratory media at 37°C in the presence of calcium ions, bacteria synthesize type III machines which remain largely non-functional. Depletion of calcium ions at elevated temperatures causes massive secretion of Yops into the culture medium via the type III pathway, which can be detected on Coomassie-stained SDS-PAGE. Because massive induction of type III secretion is associated with a reduction in growth, this 'low calcium response' (LCR) phenomenon has also been exploited for the identification of genes involved in promoting or regulating type III secretion (Goguen et al., 1984). As the concentration of free calcium ions in the eukaryotic cytosol is well below the threshold for the in vitro induction of Yop secretion, it is thought that the low calcium signal may serve as a physiological inducer for the type III pathway during infection (Lee et al., 2001). yop signals for type III secretion and injection All mapping of topogenic signals (signal peptides) aims at the identification of amino acid sequence elements that are both necessary for secretion of a polypeptide substrate and sufficient to mediate a similar transport reaction when fused to a reporter protein that otherwise cannot travel this pathway (Milstein et al., 1972). To map type III signals, Michiels and Cornelis (1991) first employed fusions of YopE and YopQ to Escherichia coliβ-galactosidase α-fragment (LacZ′) or the mature portion of alkaline phosphatase (PhoA) and demonstrated type III secretion of hybrids. A similar result was obtained for fusions between YopE and YopH to the N-terminus of Bordetella pertussis adenylate cyclase (Cya) (Sory et al., 1995). Because Cya requires binding to host cell calmodulin for activity, Cya activity determinations can be exploited as a measure for type III injection into host cells (Sory and Cornelis, 1994). Several other techniques including immunofluoresence miscroscopy, detergent extraction and immunoblotting, substrate phosphorylation and fluorescent detection now permit similar successes in measuring type III injection (Lee et al., 1998). No matter what technology is employed, the products of full-length translational fusions between yop genes and a reporter are secreted or injected with greater efficiency by the type III pathway than shorter fusions, consistent with the view that large segments of yop genes or Yop proteins contribute substrate or signal properties (Anderson and Schneewind, 1997; Ramamurthi and Schneewind, 2002b). The first 7–15 amino acids of YopE, YopH, YopN and YopQ are sufficient to promote type III secretion of reporter proteins under low calcium conditions (Sory et al., 1995; Goss et al., 2004). Fusions of larger amino acid segments, the first 70–100 amino acids of YopE, YopH, YopM or YopN, are required for type III injection into host cells (Sory et al., 1995; Boland et al., 1996; Cheng et al., 2001). For YopE, YopH and YopN these amino acid sequences encompass binding sites for small, cytoplasmic binding chaperones (SycE, SycH, SycN and YscB) in homodimeric (SycE and SycH) or heterodimeric complexes (SycN/YscB) that wrap polypeptide substrates around their folded surface (Wattiau and Cornelis, 1993; Day and Plano, 1998; Lee et al., 1998; Birtalan et al., 2002; Schubot et al., 2005). Although essential for type III injection of wild-type yersiniae, variants lacking several different yop effector genes promote type III injection with minimal secretion signals and without the requirement for Syc chaperones (Boyd et al., 2000). It is difficult to pin down the contribution of Syc chaperones to the secretion or injection of Yops. Recent discoveries in E. coli and Salmonella spp. point to YscN (or its homologues in other bacteria) as a machinery component that interacts with chaperones and bound substrates (Gauthier and Finlay, 2003; Akeda and Galan, 2004). A review of type III secretion chaperones has recently been published and interested readers are referred to this publication for additional information (Ghosh, 2004). As YopE, YopH, YopM and YopN all travel the type III pathway into host cells, are there discrete unifying features in their secretion signals? Figure 1A lists the first 15 amino acids of these polypeptides. Unlike classical signal peptides or topogenic sequences for membrane translocation, a universal feature in the form of conserved residues or a shared physical property of these elements could not be revealed (Anderson and Schneewind, 1997). Amino acids 1–17 of YopH form an amphipathic α-helix in the structure of a fragment of YopH (residues 1–130), but do not appear as an independent structural element as the α-helix folds against the YopH globule (Evdokimov et al., 2001a). Amino acids 1–15 of YopE display a propensity to form amphipathic β-sheets, whereas the N-terminal secretion signal of YopM assumes a disordered structure (Evdokimov et al., 2001b). These data of course cannot preclude the possibility that secretion signals assume a distinctive three-dimensional fold when interacting with type III machines. Figure 1Open in figure viewerPowerPoint A. Comparison of the first 15 amino acids of 14 secretion substrates for type III machines in Y. enterocolitica. Isoleucine residues (I) are highlighted in red. See text for details. B. Comparison of the yopE, yopQ and yopN secretion signals in the first 15 codons. Nucleotides that comprise the minimal secretion signals are located upstream, while those encompassing the suppressor region are located downstream of the arrow (printed in blue). Amino acids specified by each codon are indicated using single-letter abbreviations. Nucleotides that are sensitive to mutagenesis are highlighted in red. The mRNA signal hypothesis In an effort to generate secretion signal variants that fail to initiate fused neomycin phosphotransferase into the type III secretion pathway, the first 15 codons of yopE were mutated by the insertion or deletion of nucleotides immediately following the start codon (Anderson and Schneewind, 1997). The resulting frameshift mutations were suppressed by a reciprocal nucleotide change at the fusion site with the npt gene. Surprisingly, such frameshift mutations did not abrogate the secretion signal function of yopE (Anderson and Schneewind, 1997). Similar tests for the first 15 codons of yopN, yopQ, invJ, avrPto and avrBs2 from four different organisms (Yersinia spp., Salmonella enterica, Pseudomonas syringae and Xanthomonas campestri) produced similar results (Anderson and Schneewind, 1997). One interpretation of these experiments is that relatively minor alterations of mRNA sequence precipitated fundamental changes in amino acid sequence and structure that must destroy signal function if the recognition elements were proteinaceous in nature. Even the tolerance of secretion signals to frameshift mutations has limits. Mapping the minimal secretion signal of yopE and yopQ identified the first seven or 10 codons as functional elements (Ramamurthi and Schneewind, 2005). However, frameshift mutations of these minimal secretion signals abrogated function, although that initiation of the encoded hybrids into the type III pathway could be restored by fusion of downstream yopE or yopQ codons (7–15 codons). This suppressor region harbours important signal information and its function is not abolished by frameshift mutations (Fig. 1B). Systematic mutagenesis of every nucleotide in the minimal secretion signals of yopE and yopQ identified six or 10 positions, respectively, that were sensitive to substitution of purine nucleotides with pyrimidines and vice versa (Ramamurthi and Schneewind, 2003). Remarkably, two synonymous substitutions, each at isoleucine codon 3, abrogated the function of yopE and yopQ minimal secretion signals. When introduced into full-length yopE, such single nucleotide substitutions significantly reduce YopE injection into tissue culture cells, although substrate recognition was not completely blocked (Ramamurthi and Schneewind, 2005). In yopE, the AUA and AUU nucleotide triplets of Ile codon 3 promote secretion signalling, whereas AUC does not (Ramamurthi and Schneewind, 2005). The secretion signal of yopQ encompasses only one Ile codon (Ile3) with an AUU nucleotide triplet and its substitution with AUA abolishes function (Ramamurthi and Schneewind, 2003). For both yopE and yopQ synonymous substitutions promote synthesis of the predicted amino acid sequence and the mutant phenotypes can be reversed by fusion of downstream 'suppressor signals' (Ramamurthi and Schneewind, 2005). Two type III secretion substrates of Y. enterocolitica do not harbour an Ile codon in the first 15 codons, YopH and YopN (Fig. 1A). The minimal secretion signal of yopN was mapped to the first 12 codons (Goss et al., 2004). Seven positions out of 36 nucleotides were sensitive to single-nucleotide substitutions (Fig. 1B). Two synonymous substitutions at Leu codon 7, replacing the nucleotide triplet CUA with either CUU or CUC abrogated function, whereas replacement with nucleotide triplets CUG, UUG or UUA did not (Goss et al., 2004). As for yopE and yopQ, synonymous substitutions in yopN do not alter the amino acid sequence of the encoded polypeptide. There is mounting evidence that nascent polypeptides modulate the activity of translating ribosomes by interacting with the ribosomal exit tunnel, thereby generating useful biological information that instructs protein synthesis machines about the need for factors involved in protein secretion or metabolism (Gong and Yanofsky, 2002; Nakatogawa and Ito, 2002). That RNA secretion signals may be elements of translational pause and instruction of the biological fate (i.e. type III secretion) of synthesized polypeptides seems plausible (Nakatogawa and Ito, 2001); however, that translational pause permits interaction of nascent polypeptides as signal peptides with the type III secretion machinery seems highly unlikely. The ribosomal exit tunnel shields 30–40 amino acid residues, peptide segments far longer than the dimensions of type III secretion signals (Nissen et al., 2000). The signal peptide hypothesis The same arguments that were employed in favour of the mRNA signal hypothesis can also be used in support of the signal peptide hypothesis for type III secretion. The first 11 amino acids of YopE represent a functional signal for the secretion of fused reporter proteins or for the secretion of signal peptideless YopE (Lloyd et al., 2001). Synonymous replacements at 17 nucleotide positions within codons 2–11 of yopE result in a polypeptide that encodes the same amino acid sequence as protein synthesized from wild-type yopE mRNA. The mRNA signal hypothesis would predict that such dramatic changes in nucleic acid sequence (analogously to the effect of frameshift mutations on protein sequence) should abrogate the function of the secretion signal. However, the synonymous changes produced no secretion defect (Lloyd et al., 2001). Similar results have been obtained for the minimal secretion signal of Salmonella type III protein InvJ (Russmann et al., 2002). Randomly engineered peptide sequences were screened for signal peptide function of YopE (Lloyd et al., 2002). Peptides with alternating Ser–Ile motif generate amphipathic sequences that promote type III secretion. These data are consistent with the notion that discrete features of the signal peptide may interact with the secretion machinery, much like an induced fit, and thereby precipitate substrate selection for type III secretion (Lloyd et al., 2002). Computational analysis of 58 known type III substrates revealed that 66% harbour predicted secretion signals and 34% with poor signal function (Lloyd et al., 2002). Type III secretion signals in animal and plant pathogens The mapping of secretion signals in Salmonella has focused on two polypeptides, SopE and InvJ. Mutational analysis of the secretion signal of InvJ in part suggested that information required for secretion lies within the mRNA as the secretion signal of InvJ is able to tolerate frameshift mutations much like that of YopQ (Russmann et al., 2002). However, synonymous substitution mutations within codons critical for secretion (codons 4–7) failed to abolish the secretion of the full-length polypeptide, suggesting that a portion of the information required for the secretion of InvJ may reside in the polypeptide sequence (Russmann et al., 2002). Introduction of a frameshift mutation into the first 15 codons of sopE1-70 fused to a herpes simplex virus (HSV) tag prevents type III secretion of the hybrid protein (Karavolos et al., 2005). Although this would suggest that the mRNA does not play a role in the secretion of SopE, it should be considered that the first 15 codons of sopE do not represent a functional signal for the type III pathway. The secretion signals of four proteins, Cif, Map, EspF and Tir, each of which is secreted by the type III machine of enteropathogenic E. coli (EPEC), were mapped using a newly developed fluorescence-based β-lactamase assay for translocation of proteins into eukaryotic cells (Charpentier and Oswald, 2004). The first 16 codons of cif or the first 20 codons of map, espF and tir could mediate the translocation of the reporter into the eukaryotic cell irrespective of the presence of a chaperone binding site (Charpentier and Oswald, 2004). Future work will need to address whether or not mRNA sequence is important for the translocation of the proteins in EPEC. The plant pathogen P. syringae transports protein substrates via the type III pathway across the cell wall and plasma membrane of plant cells (He et al., 1993). A filamentous extension of the type III machinery, the HrpA pilus, functions as conduit for the transit of effectors (Roine et al., 1997). In an effort to identify type III secretion substrates, an in vivo genetic screen involving random fusions to a reporter with Pseudomonas effector function was scored for the activation of plant hypersensitive responses (Guttman et al., 2002). Thirteen substrates were identified that displayed no sequence similarity but shared features such as a high serine content, an aliphatic amino acid (Ile, Leu or Val) at position 3 and no acidic residues in the first 12 amino acids (Guttman et al., 2002). Expression of type III effector genes of P. syringae is activated by hrpL, a transcription factor that recognizes specific promoter proximal DNA sequences and this has been exploited for bioinformatic as well as empirical searches for new effectors (Chang et al., 2005). Together with recent data generated from a search for Cya fusion proteins (Schechter et al., 2004), the tally for genome-encoded type III effectors now stands at 19 or 29 genes, depending on individual P. syringae strains (Chang et al., 2005). The type III pathway of Y. enterocolitica secretes P. syringae effectors via short N-terminal signals that tolerate frameshift mutations (Anderson et al., 1999; Rossier et al., 1999). Furthermore, the secretion signal of P. syringae HrpA was mapped to its first 15 codons and a contribution of the 42 nucleotide 5′-untranslated leader to substrate recognition by the type III pathway was discovered (Hienonen et al., 2002). Reciprocal secretion of heterologous substrates was also achieved in plant pathogens, as X. campestri selects Yersinia YopE for transport by its type III pathway (Rossier et al., 1999). The type III secretion signals of X. campestris effector AvrBs2 were mapped (Mudgett et al., 2000). A hybrid generated by fusion of the first 28 codons of avrBs2 to a reporter gene was secreted into the extracellular milieu, whereas hybrids synthesized from larger fusions (58 codons or more) were not only secreted but also injected into plant cells (Mudgett et al., 2000). Furthermore, strains expressing +1 or +2 frameshifts of avrBs2 fused to the reporter retain the ability for type III transport (Mudgett et al., 2000). Thus, at least some features of Yersinia type III secretion signals appear to be shared between plant and animal pathogens. Flagellar export and assembly via basal body hook complexes represents a type III secretion system dedicated to bacterial motility (Macnab, 1992). While many properties of this system are shared with the type III secretion machines described here, one fundamental difference is its function as an assembled motor but not as a virulence strategy for the injection of proteins across host cell membranes. Translational fusions of genes encoding flagellin, basal body/hook complex subunits or regulatory factors to reporter genes give rise to protein products that cannot travel via the basal body secretion machine unless genetic modification relaxes the specificity of transport reactions (Chilcott and Hughes, 1998; Karlinsey et al., 2000). Nevertheless, biochemical interactions as well as genetic requirements for flagellin or hook subunit protein secretion identified cytoplasmic chaperones and physical interactions between secreted substrates and machinery components, all of which are consistent with the recognition of discrete properties of folded polypeptides by the secretion machinery (Minamino and Macnab, 2000; Hirano et al., 2003). Interested readers are referred to a recently published, detailed review of this subject (Macnab, 2004). Plugging up the holes of type III machines Although most hybrids generated by translational fusion of signal peptides to reporter genes travel through membrane translocation pores, some reporter fusions get stuck and clog the secretion machinery. The first hybrids reported to plug up membrane translocation machines employed fusions between signal peptide-bearing precursors and E. coli lacZ, encoding β-galactosidase (Oliver and Beckwith, 1981). These hybrids get stuck because the rapid folding of β-galactosidase imposes a physical blockade to the movement of the polypeptide through a narrow translocation hole. Much like jamming the subway slot machine this not only prevents passage of an oversized passenger, in fact all other passengers with proper size and tokens queue up and miss the train. Using clever genetic screens and selections, one can isolate mutant machines that fail to recognize passenger molecules with deformed signals or one can physically isolate translocation machines along with the clogged passenger molecules. Both have been achieved, which for example led to isolation of sec secretion genes and purification of the mitochondrial protein import apparatus respectively (Eilers and Schatz, 1986; Schatz and Beckwith, 1990). If one assumes that enzyme substrate interactions as they occur in protein topogenesis rely on specific, irrevocable recognition events, the fate of all signal peptide-bearing polypeptides would be initiation into the transport machine, irrespective of whether the initiated polypeptide gets stuck or not (Kiino and Silhavy, 1984). Varshavsky and colleagues exploited this general phenomenon to test whether ubiquitin fusions block the transport of signal peptide-bearing precursors into the endoplasmic reticulum of yeast (Johnsson and Varshavsky, 1994). Ribosomes with nascent polypeptides carrying signal peptides bind the signal recognition particle (SRP), a ribonucleoprotein complex that first stalls translation until the complex is docked on the SRP receptor (Walter and Blobel, 1980; Meyer et al., 1982; Lipp et al., 1987). Ribosomes then resume translation in close proximity to the membrane translocation channel and permit entry of nascent polypeptides into the water-soluble pore (Beckmann et al., 1997). Ubiquitin tethered to a signal peptide blocks the secretion pathway due to the rapid folding of this module, suggesting that the hybrid had been synthesized before its entry into the secretion machine (Johnsson and Varshavsky, 1994). In contrast, ubiquitin fusions tethered 33 residues or at a greater distance downstream from the signal peptide cannot clog the machine. The simplest explanation for these findings is that pre-formed and folded ubiquitin fusions cannot be secreted, while those that are translocated in a co-translational fashion are only folded after the translocation reaction is completed (Johnsson and Varshavsky, 1994) (Fig. 2). Figure 2Open in figure viewerPowerPoint Clogged secretion channels. Signal peptide-bearing substrates enter secretion channels and are transported from the cis compartment across a biological membrane into the trans compartment. Fusions of proteins that assume tight folding to signal peptide-bearing transport substrates can clog secretion channels and impose a blockade for the entire pathway or, as is shown for type III machines, be rejected and prevented from re-entry. See text for details. Fusions of ubiquitin to yopE or yopQ generated hybrids that cannot be transported by the type III pathway (Lee and Schneewind, 2002). This phenotype is caused by assembly of ubiquitin into its three-dimensional structure, as mutations that destabilize the ubiquitin fold generate hybrid polypeptides that can be transported through the needle complex (Lee and Schneewind, 2002). In analogy to eukaryotic secretion, this of course suggests that type III substrates are not secreted in a co-translational fashion. Surprisingly, fusions of yopE and yopQ did not clog type III machines as the secretion of YopR was not affected by expression of hybrid polypeptides (Lee and Schneewind, 2002). A similar phenotype was observed for fusions between yopE or yopQ and dihydrofolate reductase (Lee and Schneewind, 2002). Using dihydrofolate reductase fusions to yopE, it was also observed that the total amount of secreted Yops in the extracellular medium of Yersinia cultures was reduced (Feldman et al., 2002). YopE-DHFR is, however, soluble in the bacterial cytoplasm and the polypeptide does not seem associated with membranes or type III machines (Lee and Schneewind, 2002). Recent studies showed that the phenotype of reduced Yop secretion into the extracellular medium is caused by downregulation of yop expression, but not by physical blockade of type III machines (J.A. Sorg, N.C. Miller and O. Schneewind, submitted). If so, another truly astonishing feature of type III machines is their ability to reject proteins that cannot travel through the pore (Lee and Schneewind, 2002). Current models for type III secretion While entertaining models on how substrate proteins may be initiated into the type III pathway, one must first consider its unique structural features. Typical membrane translocation reactions require vertical movement of 8–10 nm across a lipid bilayer (van den Berg et al., 2005). In fact, the maximum length of an unfolded 200-residue polypeptide surpasses the longitudinal diameter of typical translocation channels, a feature that permits mechanisms such as co-translational translocation (Walter et al.