The Coupling Protein Cagβ 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.

The human gastric pathogen Helicobacter pylori is the principal cause of chronic active gastritis and peptic ulcer disease and is also involved in the development of mucosa-associated lymphoid tissue (MALT) lymphoma and gastric cancer (38, 47). The molecular mechanisms that lead to development of ulcers or cancer are not well understood, but it is clear that both bacterial and host factors contribute to disease development (2). Among bacterial factors, the vacuolating cytotoxin VacA and the cytotoxin-associated gene (Cag) type IV secretion system (T4SS) are considered two of the most important virulence determinants. The Cag secretion system is responsible for induction of a pronounced proinflammatory response in vitro and in vivo, mediated by secretion of chemokines such as interleukin-8 (IL-8) from gastric epithelial cells, and for translocation of the effector protein CagA into various host cells. Although the exact function of CagA translocation during infection is not known, H. pylori strains containing the cag pathogenicity island (cagPAI), a 37-kb genomic island which encodes the Cag T4SS, are clearly associated with an enhanced risk of developing peptic ulcers or adenocarcinoma, and translocation of the CagA protein itself significantly increases the risk of gastric cancer in the Mongolian gerbil model (2, 21).

For translocation of the CagA protein and induction of an IL-8 response, the Cag T4SS forms a secretion apparatus composed of at least 14 different proteins (22), including more- or less-distant relatives of most or all essential components of prototypical type IV secretion systems, such as the VirB system of Agrobacterium tumefaciens (6, 29). The putative inner membrane components CagU and CagH and the outer membrane-associated proteins CagM and Cagδ (Cag3) are additional essential components (29, 39). Furthermore, the Cag secretion system elaborates pilus-like appendages on the bacterial surface which contain several apparatus proteins, but the exact composition of these structures has not been determined (4, 42, 48). These surface structures are probably required for interaction with β1 integrins as receptors on the host cell surface, and proteins contributing to this interaction include the apparatus proteins CagI, CagL, and CagY as well as the substrate CagA (28, 30).

Accessory factors that are not required for the IL-8 response but are required for translocation of the CagA protein (CagA translocation factors) include the proteins Cagβ (also known as HP0524 or Cag5), CagZ (HP0526), CagF (HP0543) (22), and possibly CagD (HP0545) (13). CagF is the strongest interaction partner of CagA in the bacterial cell, and it has been shown to have a secretion chaperone-like function, possibly recruiting CagA to the secretion apparatus (16, 37). A functional role of CagZ is presently unclear, but cagZ mutants have a colonization advantage in mouse infection experiments (12), suggesting an in vivo relevance. Cagβ has significant homology to proteins of the coupling protein family, which are essential components in all DNA-transporting T4SS and in most protein-transporting T4SS (1). In DNA-transporting systems, interactions of coupling proteins with the cognate substrate complexes (relaxosomes) and with secretion apparatus components of the VirB10 family suggest that they act as substrate recognition proteins and mediate recruitment of T4SS substrates to the secretion apparatus (reviewed in reference 1). Relaxases and most DNA-independent T4SS-secreted effector proteins have C-terminal secretion signals that are necessary and sufficient for translocation, although recognition of these signals may be modulated by the presence of other domains (8). For the A. tumefaciens effector protein VirE2, it has been shown that this C-terminal signal sequence region interacts with the coupling protein VirD4 (5). Additionally, coupling proteins from conjugation systems may have DNA-dependent ATPase activity, suggesting that they might also act as DNA-transporting motor proteins (50). However, coupling proteins from different systems display considerable variability and might thus have divergent functions (1).

In agreement with previous studies and consistent with its putative role as a type IV substrate recognition protein, we show here that Cagβ acts as a translocation factor for transport of CagA into epithelial cells and that both proteins interact with each other. Using a yeast two-hybrid screen, we identified a novel interaction between Cagβ and the third CagA translocation factor, CagZ. We confirmed this interaction by biochemical interaction assays, and we showed that CagZ stabilizes Cagβ in H. pylori. Together with the interaction between CagA and CagF, our results suggest that a complex of several translocation factors and the substrate CagA is formed at the bacterial membrane prior to type IV translocation.

MATERIALS AND METHODS

Bacterial strains, cell lines, and transformation.

H. pylori strains 26695 and P12 were grown on GC agar plates (Difco) supplemented with vitamin mix (1%), horse serum (8%), vancomycin (10 mg/liter), trimethoprim (5 mg/liter), and nystatin (1 mg/liter) (serum plates) and were incubated for 16 to 60 h in a microaerobic atmosphere (85% N₂, 10% CO₂, 5% O₂) at 37°C. Escherichia coli strains Top10 (Invitrogen) and DH5α (BRL) were grown on Luria-Bertani (LB) agar plates or in LB liquid medium (43) supplemented with ampicillin (100 mg/liter), chloramphenicol (30 mg/liter), or kanamycin (40 mg/liter), as appropriate. AGS epithelial cells were cultivated under standard conditions as described previously (35). Plasmids were introduced into H. pylori strains by transformation, as described previously (25). H. pylori transformants were selected on serum agar plates containing 6 mg/liter chloramphenicol or 8 mg/liter kanamycin.

Plasmid construction.

For the production of a His₆-tagged Cagβ fusion protein, plasmid pJP88 was constructed by cloning a PCR fragment amplified from chromosomal DNA of strain J99, using primers JP57 (5′-CGGAATTCTGCAGAAGATAAGAGTGGCGAC-3′) and JP48 (5′-ACCGCTCGAGTCACAGTTCGCTTGAACC-3′), into the PstI and SalI sites of vector pEV40a (40). Plasmid pWS254, which was used for production of the glutathione S-transferase (GST)-Cagβ fusion protein, was constructed by cloning a PCR fragment amplified from chromosomal DNA of strain 26695, using primers WS293 (5′-GCGGATCCAAATACTTAACTCGGACTAG-3′) and RB31 (5′-TGGGTCGACTCACAGTTCACTTGAACC-3′), into the BamHI and SalI sites of vector pGEX4T-3 (GE Healthcare). For complementation of the cagβ and cagZ mutants, we constructed plasmids pSP15a and pWS259, respectively. Both plasmids are based on the chromosomal recA integration vector pJP99 (41), which enables, after integration into the H. pylori recA locus, expression of cloned genes under the control of the cagA promoter. Plasmid pSP15a was generated by cloning a PCR fragment obtained with primers SP14 (5′-GGACTAGTAAACACAGGTAGTAGGCACAATG-3′) and WS300 (5′-GCGGTACCTCACAGTTCACTTGAACCC-3′) into the SpeI and KpnI sites of the pJP99 derivative pWS241 (27). Plasmid pWS259 was generated by PCR amplification of a cagZ gene encoding an N-terminal Myc tag by use of primers WS309 (5′-GGACTAGTCAATGAAAGGAAACAGCAATGGAACAAAAACTCATCTCAGAAGAGGATCTGGAACTCGGTTTCAATGAAG-3′) and WS310 (5′-CGGGTACCTTATTCCAAATTTAATTTT-3′), with cloning of the PCR product into the SpeI and KpnI sites of plasmid pWS241.

Yeast two-hybrid assay.

To generate yeast two-hybrid bait and prey libraries comprising all cag pathogenicity island genes, full-length open reading frames (excluding N-terminal signal sequences) and partial open reading frames (Table 1) were amplified from chromosomal DNA of strain 26695, as described previously (29). Briefly, PCR fragments obtained with primers containing internal parts of attB1 and attB2 recombination sites together with the corresponding gene-specific sequences were used as templates for PCRs with primers containing external attB1 and attB2 sequences, and the resulting products were cloned into the entry vector pDONR207 (Invitrogen), using BP clonase, and subsequently subcloned into the destination vectors pDEST-GADT7 (prey vector) and pDEST-GBKT7 (bait vector), using LR clonase. Bait and prey plasmids were transformed into the haploid Saccharomyces cerevisiae strains Y187 and AH109, and yeast cells containing the different prey plasmids were cultivated in a 96-well plate in duplicate and spotted onto SD agar plates. This master plate was replicated once for each bait construct. Yeast strains carrying the bait plasmids were cultivated in OmniTrays (Nunc) and transferred to the top of one prey master plate replica each. Yeasts on these plates were then allowed to mate, and mating was selected for by transferring the yeasts to SD medium lacking tryptophan and leucine (SD/−Trp/−Leu), thus generating all possible combinations of bait and prey plasmids. After growth on SD/−Trp/−Leu medium, yeast colonies were transferred to SD/−Trp/−Leu/−His medium in order to select for interactions. Growth after 3 to 6 days indicated bait-prey interactions. Additionally, the stringency of this screen was enhanced by selection on SD/−Trp/−Leu/−His medium containing the competitive inhibitor 3-aminotriazole (5 mM).

TABLE 1.

Genes or gene fragments used for yeast two-hybrid screen^a

Gene	Protein designation(s)	Regionb
hp520	Cagζ, Cag1	1-116
hp521	Cagɛ, Cag2	1-80
hp522	Cagδ, Cag3	23-481
hp523	Cagγ, Cag4	1-169
hp524	Cagβa, Cag5a	95-146
hp524	Cagβb, Cag5b	162-748
hp525	Cagα, VirB11	1-330
hp526	CagZ	1-199
hp527	CagYa	1-345
hp527	CagYb	363-1799
hp527	CagYc	1815-1927
hp528	CagX	29-522
hp529	CagW	26-535
hp530	CagV	59-252
hp531	CagU	1-218
hp532	CagT	21-280
hp534	CagS	1-196
hp535	CagQ	1-126
hp536	CagP	1-114
hp537	CagM	18-376
hp538	CagN	25-306
hp539	CagL	21-237
hp540	CagI	24-381
hp541	CagH	52-370
hp542	CagG	28-142
hp543	CagF	1-268
hp544	CagE	82-983
hp545	CagD	31-207
hp546	CagC	1-115
hp547	CagA	1-1186
hp547	CagAa	1-613
hp547	CagAb	565-1186

^aGenes and gene fragments were cloned in parallel into bait and prey vectors pDEST-GBKT7 and pDEST-GADT7, respectively.

^bAmino acid ranges (for proteins from strain 26695) encoded by the cloned fragments are indicated.

β-Galactosidase assay.

All diploid yeast cells that grew on SD/−Trp/−Leu/−His medium were analyzed further for β-galactosidase activity, using a yeast β-galactosidase assay kit (Pierce). Briefly, yeast cells were grown for 16 to 24 h in SD/−Trp/−Leu medium, to optical densities at 660 nm of 0.6 to 0.8. One hundred microliters of each culture was transferred to microplate wells, and 100 μl of a working solution containing Y-PER yeast protein extraction reagent in β-galactosidase assay buffer was added. After incubation at 37°C for 20 to 40 min, absorbances at 420 nm were determined. The β-galactosidase activity was calculated using the formula E = 1,000 × A₄₂₀/t × V × A₆₆₀, where E is the enzyme activity, t is the time of incubation, and V is the assay mixture volume. Activities shown represent mean values for at least three independent experiments.

Antisera and immunoblotting.

Antiserum against Cagβ/HP524 was generated by immunization of rabbits with a recombinant C-terminal part of Cagβ fused to a hexahistidine tag, which was produced from plasmid pJP88. The His₆-tagged fusion protein was overproduced in E. coli 2136 and purified from inclusion bodies, and the purified fusion protein was used to raise a polyclonal rabbit antiserum, as described previously (29). Rabbit polyclonal antisera against CagX, CagA, and RecA have been described previously (37). GST antiserum was obtained from Sigma (monoclonal anti-GST; clone GST-2), and Myc tag antiserum was obtained from Cell Signaling (monoclonal anti-Myc tag; clone 9B11). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting were performed as described previously (22). For the visualization of proteins after SDS-PAGE, gels were stained with Coomassie brilliant blue R250. For the development of immunoblots, polyvinylidene difluoride (PVDF) filters were blocked with 3% bovine serum albumin (BSA) in TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl) and incubated with the respective antisera at a dilution of 1:1,000 to 1:5,000. Alkaline phosphatase-conjugated protein A or horseradish peroxidase-conjugated anti-rabbit IgG was used to visualize bound antibody.

Tyrosine phosphorylation assay and determination of IL-8 secretion.

Standard infections of AGS cells with H. pylori strains and subsequent preparations for phosphotyrosine immunoblotting were performed as described previously (35). Briefly, cells were infected with bacteria at a multiplicity of infection of 100 for 4 h at 37°C, washed three times, and suspended in phosphate-buffered saline (PBS) containing 1 mM EDTA, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 μg/ml leupeptin, and 10 μg/ml pepstatin. Cells with adherent bacteria were collected by centrifugation and resuspended in sample solution. Tyrosine-phosphorylated proteins were analyzed by immunoblotting with the phosphotyrosine antiserum PY99 (Santa Cruz Biotechnologies). The production of IL-8 by AGS cells after infection with H. pylori strains for 4 h was determined from cell supernatants by a sandwich enzyme-linked immunosorbent assay (ELISA) as described previously (22).

Production of GST fusion proteins and GST pull-down assays.

For the production and purification of a GST-Cagβ fusion protein, overnight cultures of E. coli strain BL21(DE3) containing plasmid pWS254 or pGEX4T-3 as a control were diluted in fresh LB medium and grown for 4 h at 37°C in a shaking incubator. Expression was induced by addition of IPTG (isopropyl-β-d-thiogalactopyranoside) to a final concentration of 0.2 mM, and cells were grown for an additional 2 h. Bacterial cells were harvested by centrifugation, resuspended in 50 mM Tris-HCl, pH 7.4, containing protease inhibitors (1 mM PMSF, 10 μg/ml leupeptin, 10 μg/ml pepstatin), and lysed by 2 passages through a French pressure cell press. The lysate was centrifuged for 15 min at 15,000 × g to remove insoluble material, and the supernatant was subjected to affinity chromatography on glutathione Sepharose 4B (GE Healthcare) according to the instructions of the manufacturer. The GST fusion protein was allowed to bind to the affinity matrix for 30 min at room temperature. After triple washing of the matrix with 10 bed volumes each of PBS, the fusion protein was eluted in three fractions with 50 mM Tris-HCl, pH 8.2, and 30 mM reduced glutathione for 15 min each at room temperature.

For GST pull-down assays, glutathione Sepharose was washed with PBS, and then 200 μg of purified GST or GST-Cagβ was added to a 100-μl bed volume of glutathione Sepharose and incubated at room temperature for 1 h. Excess protein was removed by three washing steps with PBS containing protease inhibitors. Cleared bacterial lysates obtained by sonication and subsequent centrifugation were added, and the mixture was incubated overnight at 4°C. After three washings with PBS, GST fusions and bound proteins were eluted three times with 50 mM Tris-HCl, pH 8.0, containing 10 mM reduced glutathione and then analyzed by immunoblotting.

Immunoprecipitation.

Bacteria grown on agar plates were suspended in PBS and washed twice. A total of 5 × 10¹⁰ bacteria was resuspended in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM PMSF, 10 μg/ml leupeptin, 10 μg/ml pepstatin), and the cells were lysed by sonication. Unbroken cells were removed by centrifugation for 10 min at 10,000 × g. To remove unspecifically interacting proteins, the lysates were incubated with prewashed protein G-agarose (Roche Diagnostics) for 2 h at 4°C and then centrifuged. To the supernatants, 5 μl of monoclonal Myc tag antibody (Cell Signaling) was added, and samples were incubated for 3 h at 4°C. Next, 50 μl of prewashed protein G-agarose was added, and samples were incubated at 4°C for an additional 2 h. After three washing steps with RIPA buffer, proteins were eluted with 100 mM glycine, pH 2.7, or by boiling in SDS-PAGE sample solution.

RESULTS

The putative coupling protein Cagβ is a CagA translocation factor.

Previous studies have suggested that the coupling protein homolog Cagβ (HP0524) is required for CagA translocation but not for induction of IL-8 secretion from epithelial cells (22, 45). However, secondary effects of inactivation of the cagβ gene caused by introduction of the resistance marker could not be excluded. To confirm the putative role of Cagβ as a CagA translocation factor, we complemented the 26695Δcagβ mutant with a chromosomal integration vector containing the cagβ gene under the control of the cagA promoter (plasmid pSP15a). As described previously, deletion of the cagβ gene or complementation of the cagβ mutant had no influence on the capability of the strains to induce IL-8 secretion, whereas a cagY mutant, which had no functional secretion apparatus, was defective for IL-8 induction (Fig. 1A). In comparison to the 26695Δcagβ mutant, the complemented cagβ mutant had a restored capability to translocate the CagA protein into AGS epithelial cells, as shown by CagA tyrosine phosphorylation and by induction of the hummingbird phenotype (Fig. 1B). These data confirm that Cagβ is involved in CagA translocation but is not required for IL-8 induction and thus probably not required for secretion apparatus assembly.

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.

A protein-protein interaction screen reveals interactions among CagA translocation factors.

In DNA-transporting type IV secretion systems such as conjugation systems, coupling proteins physically interact with relaxases, which represent the secreted substrates within the relaxosome complexes, and with the VirB10-like secretion apparatus components (31). However, coupling proteins also interact with protein substrates of type IV secretion systems that are not associated with DNA (5). We previously performed a yeast two-hybrid screen comprising all components of the Cag type IV secretion apparatus as well as CagA, CagF, and Cagβ (29). To identify further interactions between CagA, its translocation factors, and/or nonessential proteins of the system, we extended the yeast two-hybrid screen to all cag genes, including cagG, cagI, cagP, cagQ, cagS, cagZ, cagɛ, and cagζ, and we additionally used two overlapping fragments of the cagA gene (Table 1). Growth of diploid yeast cells on triple-selective medium was obtained for only seven combinations of bait and prey plasmids containing these additional genes or gene fragments. All interactions identified were between CagA and its putative translocation factors (CagG, CagI, CagZ, and Cagβ) or among these translocation factors (Table 2), whereas no interactions were found for CagP, CagQ, CagS, Cagɛ, or Cagζ fusions. Quantification of the interactions identified by use of a lacZ gene reporter assay revealed moderate to strong expression levels for most combinations, with yeast cells containing the Cagβ_b-CagZ bait-prey combination producing the highest LacZ levels among all Cag interactions (Fig. 2).

TABLE 2.

Novel protein-protein interactions among Cag secretion system components identified in the yeast two-hybrid screen

Bait protein	Prey protein	β-Galactosidase activitya
Cagβa	CagAa	++
Cagβb	CagZb	+++
CagZ	Cagβa	++
CagAa	Cagβa	+++
CagA	Cagβb	+
CagAa	CagI	+
CagA	CagG	+

^a+++, β-galactosidase activity higher than that of the positive control (>35 Miller units); ++, β-galactosidase activity similar to that of the positive control (28 to 35 Miller units); +, β-galactosidase activity lower than that of the positive control but significantly higher than that of the negative control (21 to 28 Miller units).

^bPreviously identified as a yeast two-hybrid interaction by use of a complete cagβ gene (7).

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.

CagZ is a CagA translocation factor that stabilizes Cagβ.

Interactions among components of type IV secretion systems often manifest themselves as mutual stabilization effects (19, 26, 29). To evaluate possible interactions of Cagβ with other Cag components, we raised an antiserum against a recombinant Cagβ fusion protein comprising the cytoplasmic part (amino acids 162 to 748). This antiserum specifically recognized a 90-kDa protein in immunoblots with whole-cell lysates of strain P12 but not with the isogenic cagβ mutant (Fig. 3A). Subsequently, we tested isogenic mutants of strain P12 in each cag gene for Cagβ production. As shown in Fig. 3A, Cagβ production was not influenced in the absence of CagA or CagF. The amounts of Cagβ production were similar in all mutants except the cagα and cagY mutants, which showed reduced amounts of Cagβ, and the cagZ mutant, which produced virtually no Cagβ (Fig. 3A and data not shown). Since the cagZ, cagα, and cagβ genes are organized in an operon and cagY contains an internal transcriptional start site which might read through cagZ, cagα, and cagβ (12, 46), replacement of cagY, cagZ, or cagα with a resistance marker may result in a reduction of the expression level of cagβ, despite the use of a nonpolar chloramphenicol resistance cassette (22). To rule out that such a downstream effect caused the absence of Cagβ in the cagZ mutant, we complemented the P12ΔcagZ mutant in trans with plasmid pWS259, containing a cagZ gene encoding an N-terminal Myc tag (P12ΔZ-mycZ). In comparison to that of the cagZ mutant, Cagβ production was restored to similar levels to those found in the cagα and cagY mutants (Fig. 3B). This indicates, on the one hand, that deletion of the cagα, cagZ, or cagY gene results in a certain reduction of cagβ expression. On the other hand, since the cagZ deletion was complemented in trans, these results strongly suggest that the CagZ protein is required for stability of Cagβ and that both proteins interact in H. pylori. As shown by a phosphotyrosine assay and also by induction of the hummingbird phenotype after infection of AGS cells with the complemented cagZ mutant, production of Myc-CagZ and restoration of cagβ expression were sufficient for reconstitution of Cag secretion system function (Fig. 3B). Complementation of the cagZ mutant with the myc-cagZ expression construct also restored IL-8 induction to similar levels to those in the cagA and cagβ mutants (Fig. 3C). The observations that the cagZ mutant induced less IL-8 than the cagβ mutant and that complementation with myc-cagZ did not restore IL-8 induction to wild-type levels can be explained by the fact that Cagα production was reduced in the cagZ mutant (data not shown), probably due to the downstream effect mentioned above.

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.

Cagβ interacts with CagA and with CagZ.

To determine whether Cagβ is able to interact with CagA, we used a GST pull-down assay with recombinant GST-Cagβ coupled to glutathione Sepharose beads. As a control, GST coupled to glutathione Sepharose was used. We incubated these beads with extracts of H. pylori wild-type strain P12, its isogenic cagZ mutant, or the myc-cagZ-complemented cagZ mutant. Immunoblots of the pulled-down fractions with an anti-CagA antibody showed that GST-Cagβ bound to CagA, whereas GST alone did not bind (Fig. 4). Furthermore, pull-down experiments with the cagZ mutant and the complemented cagZ mutant showed that the interaction of GST-Cagβ with CagA is independent of the presence of CagZ.

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.

The strongly reduced amount of Cagβ in the absence of CagZ supported the yeast two-hybrid interaction between both proteins. To confirm that CagZ and Cagβ are able to interact, we applied the GST-Cagβ pull-down procedure to the P12ΔcagZ mutant complemented with the myc-cagZ expression construct pWS259. Immunoblot analysis of the pulled-down fractions showed that Myc-CagZ bound to GST-Cagβ but not to GST alone (Fig. 5A). Since GST-Cagβ resulted in CagA being pulled down as well, as shown above, we asked whether CagA would be required for the Cagβ-CagZ interaction. Therefore, we transformed pWS259 into a P12ΔcagA mutant and used this strain for the pull-down experiments. Interestingly, GST-Cagβ was able to pull down comparable amounts of Myc-CagZ from this mutant as well, demonstrating that the presence of CagA is not required for the interaction.

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.

Finally, we performed immunoprecipitation experiments with different H. pylori P12 mutants producing Myc-CagZ, using the anti-Myc antibody. Immunoblot analysis of the precipitated fractions showed coprecipitation of Cagβ from all strains producing Myc-CagZ, whereas Cagβ was not precipitated from the wild-type P12 strain (Fig. 5B). Notably, Cagβ coprecipitation occurred in the absence of CagA, again suggesting that CagZ and Cagβ interact with each other, independent of CagA.

Interaction with Cagβ does not influence membrane localization of CagZ.

To obtain evidence for Cagβ and CagZ localization, we fractionated bacterial cells into soluble and membrane-associated components and examined these fractions for Cagβ and Myc-CagZ content by immunoblot assays. As expected from membrane topology predictions (29), Cagβ was found in the insoluble fraction, containing membrane proteins, but not in the soluble fraction, containing cytoplasmic and periplasmic proteins (Fig. 6). A similar distribution was found for the outer membrane component CagX. In contrast, Myc-CagZ was found in both the soluble and membrane fractions, similar to RecA, which has been reported to partition between soluble and membrane-associated states (20). However, since an antibody against an exclusively cytoplasmic protein was not available, we could not estimate the relative amounts of Myc-CagZ in the soluble and membrane fractions. To find out whether CagZ might be recruited to the membrane by its interaction with Cagβ, we transformed a P12Δcagβ mutant with the myc-cagZ expression construct pWS259 and used the resulting mutant for fractionation experiments. Interestingly, the distribution of Myc-CagZ between membrane and soluble fractions was identical to that in the P12ΔcagZ mutant (Fig. 6), suggesting that membrane association of CagZ is independent of Cagβ.

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.

DISCUSSION

Type IV secretion systems are known as versatile transporters of DNA and protein substrates in both Gram-negative and Gram-positive bacteria (1, 10). Assembly of different T4SS has been studied in detail, and several structures of individual secretion apparatus components and even of multiprotein core complexes have been determined (15, 23), but the molecular mechanisms of substrate recognition and translocation are still poorly understood. In DNA-transporting T4SS, interaction between the substrate and coupling protein is the first step in a sequence of interactions with different components of the secretion apparatus (11), and this step is considered critical for substrate recognition. Consistent with this view, our data from the current study and from previous studies show that the putative coupling protein Cagβ interacts with CagA and is required for CagA translocation but not for IL-8 induction, and thus probably not for assembly of the secretion apparatus. The N-terminal part of Cagβ, which according to topology predictions contains three transmembrane domains anchoring the protein in the cytoplasmic membrane (29), was not included in the GST-Cagβ fusion protein and is thus not required for this interaction, as has been shown for the VirE2-VirD4 effector/coupling protein pair (5). The third transmembrane domain of Cagβ is preceded by a periplasmic loop of about 50 amino acids, which is a common feature of many coupling proteins (1). Although we have not yet confirmed this, it is interesting that this short region showed an interaction with CagA in the yeast two-hybrid assay. The significance of this finding is not clear, but the A. tumefaciens coupling protein VirD4 has also been shown to interact with substrate proteins via its periplasmic domain (36).

The current model of substrate recognition in T4SS postulates an interaction of the coupling protein with the C-terminal substrate regions containing the secretion signals. These C-terminal regions often seem to form unstructured tails which might be required for recognition (3, 51). Alternatively, an unstructured C-terminal tail of the coupling protein itself may be involved in recognition, as shown for TraD of the F plasmid, which binds to the relaxosome component TraM in order to initiate conjugative transfer (32). In contrast to effectors of other T4SS, the C-terminal CagA region is required but not sufficient for protein translocation, which led to the suggestion that a second recognition step takes place (27). However, the need for an intact N-terminal region might also be due to the fact that CagA binds to β1 integrins via this region and that this binding is probably necessary for translocation (28). Furthermore, a sequence motif in the CagA middle region was recently shown to be necessary for binding to phosphatidylserine in the host cell membrane, which is also required for translocation (34). Since GST pull-down experiments with strains producing C-terminally truncated CagA variants resulted in reduced and variable binding (data not shown), it is not clear at this point which part of CagA is recognized by Cagβ.

Apart from secretion apparatus components and coupling proteins, translocation of some T4SS substrates requires the presence of secretion chaperones or adaptor proteins, which might help to expose C-terminal signal sequences to the substrate receptors (9, 18, 52). For the Cag system, such a secretion chaperone-like function has been ascribed to the CagA-interacting protein CagF (37). Further accessory proteins in DNA-transporting systems are required for processing of DNA for transfer (Dtr factors), and thus for formation of the relaxosomes (17), or as spatial adaptors facilitating recruitment of relaxosomes to the secretion apparatus (1). Accessory proteins similar to CagZ have not been described for protein-transporting T4SS so far. The data shown here indicate that cagZ deletion results in a complete CagA translocation defect in strain P12, and functional complementation of the cagZ mutant showed that no secondary mutations are responsible for this defect. However, since we found very low CagA translocation levels in the cagZ mutant of a different H. pylori strain in a previous study (22), we cannot rule out that CagA translocation is still possible in the absence of CagZ. Alternatively, it might be that strain-specific differences in the requirement for CagZ exist, but in any case, cagZ deletion results in severe impairment of CagA translocation. CagZ is a protein with an acidic pI value, and both a negatively charged patch on its surface and an unstructured C-terminal tail were taken as indications of its possible interaction with (positively charged) CagA (14). Our results demonstrate that CagZ interacts directly with the coupling protein, which in turn is able to bind CagA, but they do not exclude an additional interaction between CagZ and CagA. The reduced stability of Cagβ observed in the cagZ mutant indicates that incorrectly positioned Cagβ might be removed rapidly from the membrane and degraded. Since the Cagβ-CagZ interaction was also found in the absence of CagA, it seems rather unlikely that CagZ is recruited only during active translocation.

As a possible model for recognition of CagA as a substrate of the Cag T4SS, we thus propose that CagA is bound by CagF at the bacterial cytoplasmic membrane, in an effector protein-secretion chaperone complex, and that Cagβ permanently interacts with CagZ to form a functional substrate-receptor complex. Binding of the CagA-CagF complex to the Cagβ-CagZ complex via recognition of CagA by Cagβ, possibly involving the C-terminal type IV secretion signal of CagA (27), would result in a translocation-competent state of CagA and in subsequent introduction of CagA into the secretion channel. It is conceivable that positioning of the CagZ-Cagβ complex at the bacterial cytoplasmic membrane is correlated with the locations of secretion apparatus assembly sites. Although we were unable to identify interactions between either CagZ or Cagβ and components of the secretion apparatus, interactions between GST-CagZ and recombinant CagV, CagM, and CagX proteins have been observed previously (7). Binding of the CagZ-Cagβ complex to the secretion apparatus might thus be a means to recruit CagA to the translocation channel. It is currently unclear whether coupling proteins have a role in substrate transport. In the R388 conjugation system, the coupling protein TrwB was shown to bind DNA and to exhibit a DNA-dependent ATPase activity (33, 50). Interestingly, recombinant Cagβ was also found to oligomerize and to bind DNA unspecifically, but it lacks nucleoside triphosphatase (NTPase) activity, despite containing a conserved nucleotide-binding domain (44). This would argue against Cagβ providing the energy for translocation, as suggested for conjugation system coupling proteins (24). However, it cannot be excluded that ATPase activity would have to be stimulated by interaction with other factors (49), such as CagA or CagZ.

In conclusion, the requirement of an additional translocation factor which interacts with and stabilizes the coupling protein underscores the evolutionary distance of the Cag system from prototypical type IV transporters and also their structural and functional diversity in general. It will be of great future interest to determine how this diversity translates into alternative secretion mechanisms and thus to substrate and target cell adaptation.