Ingrid Guilvout

Structural Insights into the Secretin PulD and Its Trypsin-resistant Core

Limited proteolysis, secondary structure and biochemical analyses, mass spectrometry, and mass measurements by scanning transmission electron microscopy were combined with cryo-electron microscopy to generate a three-dimensional model of the homomultimeric complex formed by the outer membrane secretin PulD, an essential channel-forming component of the type II secretion system from Klebsiella oxytoca. The complex is a dodecameric structure composed of two rings that sandwich a closed disc. The two rings form chambers on either side of a central plug that is part of the middle disc. The PulD polypeptide comprises two major, structurally quite distinct domains; an N domain, which forms the walls of one of the chambers, and a trypsin-resistant C domain, which contributes to the outer chamber, the central disc, and the plug. The C domain contains a lower proportion of potentially transmembrane β-structure than classical outer membrane proteins, suggesting that only a small part of it is embedded within the outer membrane. Indeed, the C domain probably extends well beyond the confines of the outer membrane bilayer, forming a centrally plugged channel that penetrates both the peptidoglycan on the periplasmic side and the lipopolysaccharide and capsule layers on the cell surface. The inner chamber is proposed to constitute a docking site for the secreted exoprotein pullulanase, whereas the outer chamber could allow displacement of the plug to open the channel and permit the exoprotein to escape.

The secretin-specific, chaperone-like protein of the general secretory pathway: separation of proteolytic protection and piloting functions

The chaperone‐like protein of the main terminal branch of the general secretory pathway from Klebsiella oxytoca, the outer membrane lipoprotein PulS, protects the multimeric secretin PulD from degradation and promotes its correct localization to the outer membrane. To determine whether these are separable functions, or whether resistance to proteolysis results simply from correct localization of PulD, we replaced the lipoprotein‐type signal peptide of PulS by the signal peptide of periplasmic maltose‐binding protein. The resulting periplasmic PulS retained its ability to protect PulD, but not its ability to localize PulD to the outer membrane and to function in pullulanase secretion. Periplasmic PulS competed with wild‐type PulS to prevent pullulanase secretion, presumably again by causing mislocalization of PulD. A hybrid protein comprising the mature part of PulS fused to the C‐terminus of full‐length maltose‐binding protein (MalE–PulS) had similar properties to the periplasmic PulS protein. Moreover, MalE–PulS was shown to associate with PulD by amylose‐affinity chromatography. The MalE–PulS hybrid was rendered completely functional (i.e. it restored pullulanase secretion in a pulS mutant) by replacing its signal peptide with a lipoprotein‐type signal peptide. However, this fatty‐acylated hybrid protein was only functional if it also carried a lipoprotein sorting signal that targeted it to the outer membrane. Thus, the two functions of PulS are separate and fully dissociable. Incorrect localization, rather than proteolysis, of PulD in the absence of PulS was shown to be the factor that causes high‐level induction of the phage shock response. The Erwinia chrysanthemi PulS homologue, OutS, can substitute for PulS, and PulS can protect the secretin OutD from proteolysis in Escherichia coli, indicating the possible existence of a family of PulS‐like chaperone proteins.

Bacterial outer membrane secretin PulD assembles and inserts into the inner membrane in the absence of its pilotin

Dodecamerization and insertion of the outer membrane secretin PulD is entirely determined by the C‐terminal half of the polypeptide (PulD‐CS). In the absence of its cognate chaperone PulS, PulD‐CS and PulD mislocalize to the inner membrane, from which they are extractable with detergents but not urea. Electron microscopy of PulD‐CS purified from the inner membrane revealed apparently normal dodecameric complexes. Electron microscopy of PulD‐CS and PulD in inner membrane vesicles revealed inserted secretin complexes. Mislocalization of PulD or PulD‐CS to this membrane induces the phage shock response, probably as a result of a decreased membrane electrochemical potential. Production of PulD in the absence of the phage shock response protein PspA and PulS caused a substantial drop in membrane potential and was lethal. Thus, PulD‐CS and PulD assemble in the inner membrane if they do not associate with PulS. We propose that PulS prevents premature multimerization of PulD and accompanies it through the periplasm to the outer membrane. PulD is the first bacterial outer membrane protein with demonstrated ability to insert efficiently into the inner membrane.

The C‐terminal domain of the secretin PulD contains the binding site for its cognate chaperone, PulS, and confers PulS dependence on pIVf1 function

Related outer membrane proteins, termed secretins, participate in the secretion of macromolecules across the outer membrane of many Gram‐negative bacteria. In the pullulanase‐secretion system, PulS, an outer membrane‐associated lipoprotein, is required both for the integrity and the proper outer membrane localization of the PulD secretin. Here we show that the PulS‐binding site is located within the C‐terminal 65 residues of PulD. Addition of this domain to the filamentous phage secretin, pIV, or to the unrelated maltose‐binding protein rendered both proteins dependent on PulS for stability. A chimeric protein composed of bacteriophage f1 pIV and the C‐terminal domain of PulD required properly localized PulS to support phage assembly. An in vivo complex formed between the pIV‐PulD65 chimera and PulS was detected by co‐immunoprecipitation and by affinity chromatography.

Molecular cloning, iron-regulation and mutagenesis of the irp2 gene encoding HMWP2, a protein specific for the highly pathogenic Yersinia

Under iron‐starvation, the highly pathogenic Yersinia synthesize several iron‐regulated proteins including two high‐molecular‐weight polypeptides (HMWP1 and HMWP2). From the chromosome of Yersinia enterocolitica serovar O:8 (strain Ye 8081), the genes coding for the HMWP2 (irp2) and its promoter were cloned into plasmid pUC18 (plR2) and used as a probe. We show here that the irp2 gene is present only in the highly pathogenic strains and that its promoter is iron‐regulated in Escherichia coli. After introduction of the plR2 plasmid into a fur mutant of E. coli, both the iron‐starved and the iron‐replete bacteria expressed the HMWP2. Repressibility of irp2 by iron was restored by introduction of a plasmid carrying the fur gene. These results demonstrate that the irp2 promoter is controlled by the Fur repressor in E. coli. Mutagenesis of the chromosomal irp2 gene of Yersinia pseudotuberculosis was obtained by homologous recombination with a 1 kb fragment of this gene cloned on the suicide plasmid pJM703.1. Inactivation of irp2 resulted in the non‐expression of both HMWPs, while introduction of plasmid plR2 into the mutant strain led to the synthesis of the HMWP2 only. Therefore, It is probable that the genes coding for the HMWPs constitute an operon where irp2 is upstream of irp1. When comparing the virulence of the wild‐type strain and of its irp2 mutant derivative, we found that the 50% lethality (LD50) for mice of the mutant strain was increased, whatever the route of infection, but more markedly when injected parenterally. Accordingly, these data demonstrate that a mutation in the irp2 gene alters the pathogenicity of Y. pseudotuberculosis. Since the introduction of the irp2 gene into the mutant strain did not restore its virulence, it is likely that both HMWPs are required for the expression of the high‐pathogenicity phenotype.

Secretins take shape

Secretins are a unique class of bacterial multimeric outer membrane proteins that probably differ considerably from other, less complex outer membrane proteins in their overall structure and organization, and in their requirements for outer membrane targeting and assembly factors. In this MicroCommentary, we discuss these differences with respect to the role of a specific class of lipoproteins, often referred to as pilotins, in secretin complex assembly. We compare them with other lipoproteins that play a role in Omp85/YaeT‐mediated assembly of more classical outer membrane proteins. One of the examples we have chosen is the Myxococcus Xanthus lipoprotein Tgl. Coculture of cells with and without Tgl allows secretin assembly (and, hence, type IV pilus assembly) in the cells without Tgl, indicating that it can act by cell‐to‐cell contact or can transfer between cells.

In Vitro Multimerization and Membrane Insertion of Bacterial Outer Membrane Secretin PulD

Synthesis of the Klebsiella oxytoca outer membrane secretin PulD, or its membrane-associated core domain, in a liposome-supplemented Escherichia coli in vitro transcription-translation system resulted in the formation of multimers that appeared as typical dodecameric secretin rings when examined by negative-stain electron microscopy. Cryo-electron microscopy of unstained liposomes and differential extraction by urea indicated that the secretin particles were inserted into the liposome membranes. When made in the presence of the detergent Brij-35, PulD and the core domain were synthesized as monomers. Both proteins caused almost immediate growth cessation when synthesized in E. coli without a signal peptide. The small amounts of PulD synthesized before cell death appeared as multimers with characteristics similar to those of the normal outer membrane secretin dodecamers. It was concluded that multimerization and membrane insertion are intrinsic properties of secretin PulD that are independent of a specific membrane environment or membrane-associated factors. The closely related Erwinia chrysanthemi secretin OutD behaved similarly to PulD in all assays, but the more distantly related Neisseria meningitidis secretin PilQ did not form multimers either when made in vitro in the presence of liposomes or when made in E. coli without its signal peptide. This is the first report of the apparently spontaneous in vitro assembly and membrane insertion of a large outer membrane protein complex. Spontaneous multimerization and insertion appear to be restricted to outer membrane proteins closely related to PulD.

Artificial Binding Proteins (Affitins) as Probes for Conformational Changes in Secretin PulD

The DNA-binding protein Sac7d was previously modified to bind with high affinity to the N domain of the outer membrane secretin PulD from the bacterium Klebsiella oxytoca. Here, we show that binding of the Sac7d derivatives (affitins) to PulD is sensitive to conformational changes caused by denaturant and by the zwitterionic detergent Zwittergent 3-14 routinely used to extract secretins from outer membranes. This sensitivity to the conformational state of PulD allowed us to use the affitins as probes for the native structure of PulD and to devise protocols for examining in vitro synthesized protein in nonionic detergent and for the affinity purification of native PulD using affitins as ligands. When fused to periplasmic PhoA, three affitins inhibited PulD multimerization in vivo and caused loss of function. In two cases, this was likely to be due to dimerization of the affitin by the bound PhoA, as the effect was absent when the affitins were fused to monomeric MalE. In the third case, the MalE and PhoA moieties probably interfered sterically with PulD protomer interactions and, thereby, inhibited multimerization. None of the affitins tested interacted with PulD at sites of protomer interaction or blocked the secretin channel through which exoproteins cross the outer membrane in the Type II secretion pathway of which PulD is a key component.

YaeT‐independent multimerization and outer membrane association of secretin PulD

Previous studies demonstrated that targeting of the dodecameric secretin PulD to the Escherichia coli outer membrane is strictly dependent on the chaperone‐like pilotin PulS. Here, we report that PulD multimerization and membrane association in strains producing PulS were unaffected when the levels of the essential outer membrane assembly factor YaeT(Omp85) were reduced by controlled expression of a paraBAD–yaeT transcriptional fusion. This behaviour contrasted markedly to that of the trimeric porin LamB, which remained monomeric under these conditions. Furthermore, resistance to extraction by the detergent Sarkosyl and by urea, and susceptibility to trypsin digestion all suggested that PulD localized to the outer membrane in YaeT‐depleted cells. We conclude that, unlike classical β‐barrel outer membrane proteins such as LamB, multimerization of PulD is largely YaeT‐independent.

Remodeling a DNA-binding protein as a specific in vivo inhibitor of bacterial secretin PulD

We engineered a class of proteins that binds selected polypeptides with high specificity and affinity. Use of the protein scaffold of Sac7d, belonging to a protein family that binds various ligands, overcomes limitations inherent in the use of antibodies as intracellular inhibitors: it lacks disulfide bridges, is small and stable, and can be produced in large amounts. An in vitro combinatorial/selection approach generated specific, high-affinity (up to 140 pM) binders against bacterial outer membrane secretin PulD. When exported to the Escherichia coli periplasm, they inhibited PulD oligomerization, thereby blocking the type II secretion pathway of which PulD is part. Thus, high-affinity inhibitors of protein function can be derived from Sac7d and can be exported to, and function in, a cell compartment other than that in which they are produced.