Protein subassemblies of the Helicobacter pylori Cag type IV secretion system revealed by localization and interaction studies

Abstract

Type IV secretion systems are possibly the most versatile protein transport systems in gram-negative bacteria, with substrates ranging from small proteins to large nucleoprotein complexes. In many cases, such as the cag pathogenicity island of Helicobacter pylori, genes encoding components of a type IV secretion system have been identified due to their sequence similarities to prototypical systems such as the VirB system of Agrobacterium tumefaciens. The Cag type IV secretion system contains at least 14 essential apparatus components and several substrate translocation and auxiliary factors, but the functions of most components cannot be inferred from their sequences due to the lack of similarities. In this study, we have performed a comprehensive sequence analysis of all essential or auxiliary Cag components, and we have used antisera raised against a subset of components to determine their subcellular localization. The results suggest that the Cag system contains functional analogues to all VirB components except VirB5. Moreover, we have characterized mutual stabilization effects and performed a comprehensive yeast two-hybrid screening for potential protein-protein interactions. Immunoprecipitation studies resulted in identification of a secretion apparatus subassembly at the outer membrane. Combining these data, we provide a first low-resolution model of the Cag type IV secretion apparatus.

FIG. 1.

Graphical representation of potential membrane-spanning segments and sequence similarity regions of Cag secretion apparatus components examined in this study. The proteins are drawn to scale as bars, with the number of amino acid residues deduced from the published genome sequence of strain 26695 (63) indicated at their C-terminal ends. The regions of similarity to the corresponding A. tumefaciens VirB proteins are shown in blue. Potential transmembrane helices predicted by the PHDhtm or TMPred algorithms are marked as orange boxes with the position of the first transmembrane amino acid indicated above; red boxes at the N termini indicate Sec-dependent signal sequences. A surplus of positively charged amino acids, which predicts the cytoplasmic orientation of a transmembrane helix according to the positive-inside rule, is indicated by a plus sign. Coiled coils predicted by the COILS algorithm are indicated by green boxes.

FIG. 2.

Sequence analysis and characterization of the CagC protein. (A) VirB2 and VirB2-like proteins such as TrbC of plasmid RP4 are characterized by unusually long N-terminal signal sequences (dark gray boxes with the respective numbers of amino acids indicated above) and the presence of two transmembrane helices (black boxes). Conserved motifs have been identified at the signal peptidase cleavage site and a C-terminal processing site that is used for protein cyclization (indicated by arrows in the TrbC protein). A similar arrangement of transmembrane helices and weakly conserved motifs is also found in the CagC protein. Black bars below the CagC protein indicate regions from which peptides were derived for the generation of CagC antisera. (B) Immunoblot showing CagC production in the wild-type strain 26695 in comparison to the ΔPAI mutant lacking the cag pathogenicity island (PAI). Two different antisera recognizing an internal and a C-terminal peptide of CagC were used. α, anti; aa, amino acids.

FIG. 3.

Immunoblot analysis of the localization of Cag secretion apparatus proteins. (A) H. pylori bacterial cell lysates (BL) were separated by ultracentrifugation into a total membrane fraction (TM) and a soluble fraction containing cytoplasmic and periplasmic proteins (C/P). The fractions were probed with antisera raised against the indicated proteins. As controls, antisera against the integral cytoplasmic membrane protein ComB10 and the cytoplasmic protein RecA were used. (B) Differential extraction of an H. pylori total membrane (TM) fraction. For an evaluation of the extraction procedure, antisera against the outer membrane protein AlpB and against ComB10 were used. The distribution of Cag secretion apparatus proteins in the outer membrane (OM) and cytoplasmic membrane (CM) fractions was determined by immunoblotting the fractions with the indicated antisera.

FIG. 4.

Stabilization effects among Cag secretion apparatus proteins. The production of the indicated proteins was determined in adjusted cell lysates of isogenic H. pylori mutants in genes encoding essential secretion apparatus proteins. Production of the urease B subunit (UreB) was determined as a loading control. Representative immunoblots are shown. Pronounced stabilization effects, which were also confirmed by densitometry, occur for the CagX protein (in the cagY mutant) and for the CagT protein (in the cagM, cagX, and cagY mutants). The weak production of Cagα in the cagY mutant and of CagT in the cagU mutant is possibly due to a polar effect.

FIG. 5.

Protein-protein interactions among Cag proteins identified by a yeast two-hybrid screen. Diploid yeast cells containing the indicated plasmid pairs (given as bait + prey plasmid), all of which were selected for growth on triple-selective medium (SD medium lacking tryptophan, leucine, and histidine), and yeast cells containing positive control (+) or negative control (−) plasmids were assayed for β-galactosidase activity, as described in Materials and Methods. The values shown are mean values of three independent experiments including standard deviations. The activities shown were classified into three categories, as described in Table 2.

FIG. 6.

Identification of a secretion apparatus subcomplex composed of CagT, CagM, CagX, and CagY in H. pylori cells. Bacteria were lysed in RIPA buffer, and the lysates (starting extracts) were incubated with protein G-agarose to remove nonspecifically binding proteins and subsequently immunoprecipitated (IP) with antisera against CagX (A), CagY (B), and CagM (C). Coprecipitating proteins were identified by Western blotting (WB) with the indicated antisera. Control immunoprecipitations from the corresponding isogenic mutants show the specificity of the coprecipitations, except for CagM, which coprecipitated nonspecifically in a CagY immunoprecipitation.

FIG. 7.

Determination of direct and indirect interactions. Immunoprecipitations with antisera against CagX, CagY, and CagM were performed as described in the legend of Fig. 6, except that the indicated isogenic mutants were used. (A) The CagX-CagM interaction is independent of CagT and CagY. (B) The CagX-CagY interaction is independent of CagT and CagM. (C) The CagM-CagY interaction is independent of CagT but dependent on CagX. (D) The CagM-CagT interaction is independent of CagX and CagY. (E) The CagX-CagT interaction depends on CagM but is independent of CagY. (F) The CagY-CagT interaction depends on both CagX and CagM. IP, immunoprecipitation; WB, Western blotting.

FIG. 8.

Schematic model of the composition of the Cag type IV secretion apparatus. All components of the secretion apparatus are shown in their putative localizations as suggested by computer prediction, fractionation, and interaction data. Protein names are abbreviated such that X represents CagX, for example. The putative localization of (parts of) CagY (52) and CagL (38) on extracellular appendages produced by the Cag system is not depicted. Protein-protein interactions that were identified or confirmed in this study are indicated by double arrows. The direct interactions between components at the outer membrane (CagY-CagX, CagX-CagM, and CagM-CagT) were confirmed by coimmunoprecipitation; all other interactions were shown only by yeast two-hybrid data. The putative assembly of a secretin-like heterooligomeric complex at the outer membrane is represented by green boxes; a subassembly at the cytoplasmic membrane is shown by dark orange boxes. Components for which no interactions were identified are shown in light orange. See text for further details. CM, cytoplasmic membrane; PG, peptidoglycan; OM, outer membrane.