The coupling protein Cagbeta and its interaction partner CagZ are required for type IV secretion of the Helicobacter pylori CagA protein

Abstract

Bacterial type IV secretion systems are macromolecule transporters with essential functions for horizontal gene transfer and for symbiotic and pathogenic interactions with eukaryotic host cells. Helicobacter pylori, the causative agent of type B gastritis, peptic ulcers, gastric adenocarcinoma, and mucosa-associated lymphoid tissue (MALT) lymphoma, uses the Cag type IV secretion system to inject its effector protein CagA into gastric cells. This protein translocation results in altered host cell gene expression profiles and cytoskeletal rearrangements, and it has been linked to cancer development. Interactions of CagA with host cell proteins have been studied in great detail, but little is known about the molecular details of CagA recognition as a type IV secretion substrate or of the translocation process. Apart from components of the secretion apparatus, we previously identified several CagA translocation factors that are either required for or support CagA translocation. To identify protein-protein interactions between these translocation factors, we used a yeast two-hybrid approach comprising all cag pathogenicity island genes. Among several other interactions involving translocation factors, we found a strong interaction between the coupling protein homologue Cagβ (HP0524) and the Cag-specific translocation factor CagZ (HP0526). We show that CagZ has a stabilizing effect on Cagβ, and we demonstrate protein-protein interactions between the cytoplasmic part of Cagβ and CagA and between CagZ and Cagβ, using immunoprecipitation and pull-down assays. Together, our data suggest that these interactions represent a substrate-translocation factor complex at the bacterial cytoplasmic membrane.

FIG. 1.

The putative coupling protein Cagβ is a CagA translocation factor. (A) AGS cells were infected for 4 h with H. pylori wild-type strains or isogenic mutants, as indicated. IL-8 concentrations in culture supernatants were determined by sandwich ELISA and are shown in relation to IL-8 levels induced by wild-type bacteria (wt). (B) Lysates of AGS cells infected with wild-type H. pylori strain 26695, its isogenic cagβ mutant, and a cagβ mutant complemented in trans with a wild-type cagβ gene (26695Δβ-β) were examined for CagA production (135 kDa) and CagA tyrosine phosphorylation. Additionally, the hummingbird phenotype of infected cells was evaluated by phase-contrast microscopy at 4 h postinfection. WB, Western blot.

FIG. 2.

Novel protein-protein interactions involving CagA translocation factors identified by yeast two-hybrid screen. Diploid yeast cells harboring the indicated plasmid pairs (bait and prey plasmids), selected for growth on SD medium lacking tryptophan, leucine, and histidine (triple-selective medium), were assayed for β-galactosidase activity as described in Materials and Methods. For comparison, yeast cells containing positive-control (+) and negative-control (−) plasmids were also assayed. The values shown are mean values for three independent experiments, with standard deviations. The activities shown were classified into three categories, as described in Table 2.

FIG. 3.

CagZ is a CagA translocation factor which stabilizes Cagβ. (A) Whole-cell lysates of wild-type strain P12 and of isogenic mutants in single cag genes were examined by immunoblotting with an anti-Cagβ antiserum. (B) The P12ΔcagZ mutant was complemented in the recA locus with a cagZ expression construct encoding an N-terminal Myc tag (P12ΔZ-mycZ). The wild-type strain, the cagZ mutant, and the complemented mutant were used for infection experiments with AGS cells. Infection lysates were tested by immunoblotting for Cagβ (90 kDa) and CagA production and for CagA tyrosine phosphorylation, and infected cells were examined by microscopy for development of the hummingbird phenotype. (C) AGS cells were infected for 4 h with H. pylori wild-type strains or isogenic mutants, as indicated. IL-8 concentrations in culture supernatants were determined by sandwich ELISA and are shown in relation to IL-8 levels induced by wild-type bacteria. Data shown are average values for at least 3 independent experiments, with standard deviations.

FIG. 4.

Cagβ interacts with CagA independent of the presence of CagZ. The indicated H. pylori strains were extracted using RIPA buffer (starting extracts), and cell extracts were subjected to pull-down (PD) experiments with GST or a GST-Cagβ fusion protein coupled to glutathione Sepharose beads. Pull-down fractions were analyzed by immunoblotting against CagA and against GST. The GST-Cagβ fusion protein has an expected molecular mass of approximately 90 kDa, and GST has a molecular mass of approximately 27 kDa.

FIG. 5.

CagZ interacts with Cagβ. (A) H. pylori wild-type strain P12 and its isogenic cagA mutant were transformed with the myc-cagZ expression construct pWS259, and extracts of the corresponding strains were subjected to GST pull-down (PD) experiments. Pulled-down CagA and CagZ-Myc (24 kDa) were monitored by immunoblotting. (B) Starting extracts of the indicated strains were subjected to immunoprecipitation (IP) using an anti-Myc antibody. Starting extracts and immunoprecipitates were examined by immunoblotting for Cagβ and Myc.

FIG. 6.

CagZ localization is independent of Cagβ. H. pylori cells grown on agar plates were fractionated into soluble and membrane components, and the fractions were assayed for their Cagβ and Myc-CagZ content. As controls, RecA (40 kDa) was used as a marker for (partially) soluble proteins and CagX (61 kDa) was used as a marker for membrane-associated proteins. Note that the myc-cagZ expression construct was inserted into the recA locus, so Myc-CagZ-producing strains are recA mutants. WCL, whole-cell lysate; Sol, soluble fraction containing cytoplasmic and periplasmic proteins; TM, total membrane fraction.