Lipoprotein Processing and Sorting in Helicobacter pylori

Abstract

Our current understanding of lipoprotein synthesis and localization in Gram-negative bacteria is based primarily on studies of Escherichia coli Newly synthesized E. coli prolipoproteins undergo posttranslational modifications catalyzed by three essential enzymes (Lgt, LspA, and Lnt). The mature lipoproteins are then sorted to the inner or outer membrane via the Lol system (LolABCDE). Recent studies suggested that this paradigm may not be universally applicable among different classes of proteobacteria. In this study, we conducted a systematic analysis of lipoprotein processing and sorting in Helicobacter pylori, a member of the Epsilonproteobacteria that colonizes the human stomach. We show that H. pylorilgt, lspA, and lnt homologs can complement conditionally lethal E. coli mutant strains in which expression of these genes is conditionally regulated. Mutagenesis studies and analyses of conditionally lethal H. pylori mutant strains indicate that lgt and lspA are essential for H. pylori growth but lnt is dispensable. H. pylorilolA and the single lolC (or lolE) homolog are also essential genes. We then explored the role of lipoproteins in H. pylori Cag type IV secretion system (Cag T4SS) activity. Comparative analysis of the putative VirB7 homolog CagT in wild-type and lnt mutant H. pylori strains indicates that CagT undergoes amino-terminal modifications consistent with lipidation, and we show that CagT lipidation is essential for CagT stability and Cag T4SS function. This work demonstrates that lipoprotein synthesis and localization in H. pylori diverge from the canonical pathways and that lipidation of a T4SS component is necessary for H. pylori Cag T4SS activity.IMPORTANCE Bacterial lipoproteins have diverse roles in multiple aspects of bacterial physiology, antimicrobial resistance, and pathogenesis. Dedicated pathways direct the posttranslational lipidation and localization of lipoproteins, but there is considerable variation in these pathways among the proteobacteria. In this study, we characterized the proteins responsible for lipoprotein synthesis and localization in Helicobacter pylori, a member of the Epsilonproteobacteria that contributes to stomach cancer pathogenesis. We also provide evidence suggesting that lipidation of CagT, a component of the H. pylori Cag T4SS, is required for delivery of the H. pylori CagA oncoprotein into human gastric cells. Overall, these results constitute the first systematic analysis of H. pylori lipoprotein production and localization pathways and reveal how these processes in H. pylori differ from corresponding pathways in model proteobacteria.

Keywords: Helicobacter pylori; Toll-like receptor 2; lipoproteins; posttranslational protein modification; type IV secretion systems.

FIG 1

Conservation among lipoprotein synthetic enzymes Lgt, LspA, and Lnt. Ribbon diagrams representing E. coli Lgt (A) (PDB 5AZC [56]), P. aeruginosa LspA (B) (PDB 5DIR [59]), and E. coli Lnt (C) (PDB 5N6H [61]) and predicted structures of H. pylori Lgt (D), Lsp (E), and Lnt (F) (generated by submitting the H. pylori sequences to Phyre2 [92]) are shown. Superimposed structures of Lgt (G), LspA (H), and Lnt (I) were generated using Chimera (93). The amino acid side chains of important conserved residues are shown as green spheres.

FIG 2

Functional properties of H. pylori lgt, lsp, and lnt expressed in E. coli. E. coli PAP9403, YX238, and KA349 contain lgt, lspA, and lnt, respectively, under the control of an arabinose-inducible promoter (58, 63, 64). Since expression of lgt, lsp, and lnt is essential for bacterial growth, these strains do not grow on medium containing glucose but do grow on medium containing arabinose. Introduction of plasmids containing the homologous H. pylori genes under the control of the lacUV5 or Trc promoter supported growth on glucose. Vector-only control plasmids (C) did not support growth on glucose. Results are representative of 3 experiments.

FIG 3

Growth of H. pylori conditional mutants. H. pylori strains engineered to express lgt (VM176), lspA (VM183), lolA (VM189), or lolF (VM191) under the control of a TetR-regulated promoter were inoculated onto media in the presence or absence of anhydrotetracycline (ATc). Representative results from three independent cultures of each mutant are shown.

FIG 4

TLR2 activation by H. pylori lipoproteins. Protein extracts enriched in lipoproteins were prepared from H. pylori strain 26695 or the lnt mutant strain VM211 and were added (150 ng per ml) to 293-mTLR1/2 or 293-hTLR2/6 cell lines. As controls, cells also were incubated with triacylated Pam3CSK4 or diacylated Pam2CSK4 (30 ng per ml). Following 24 h of incubation, culture supernatants were recovered and subjected to ELISA to determine IL-8 concentrations. Results are expressed as the ratio of the level of IL-8 produced by 293-hTLR2/6 cells divided by the level of IL-8 produced by 293-hTLR1/2 cells and represent means and standard deviations of three independent experiments, each with triplicate samples. Asterisks denote results that were significantly different from the control results (Student's t test, P < 0.0003).

FIG 5

Analyses of CagT-DDK. (A) Amino-terminal amino acid sequence of CagT-DDK and the CagT-DDK peptides following enterokinase treatment. The lipobox is highlighted in red, the DDK epitope is underlined, and the asterisk indicates the site of triacyl lipid modification in WT H. pylori or the site of diacyl lipid modification in lnt mutant H. pylori. Signal peptide cleavage occurs between the “A” and “C” within the lipobox. Enterokinase cleavage occurs after lysine at its cleavage site DDDDK. (B) Expression of CagT was assessed by immunoblotting extracts of strains 26695 (WT), BV199 (ΔcagT), BV321 (restored cagT), and BV357 (cagT-DDK₂₇) using anti-CagT and anti-HspB (as loading control). (C) H. pylori strains were cocultured with AGS cells, and the ability of each strain to induce IL-8 production was determined by ELISA. Asterisks denote results that were significantly different from the BV199 control results (analysis of variance [ANOVA] followed by Dunnett’s post hoc test, P < 0.05). (D) Protein extracts from BV357 and VM207 (each producing CagT-DDK₂₇, the latter in an lnt mutant background) were treated with enterokinase and immunoblotted using anti-DDK monoclonal antibody. Consistent with expectations, the immunoreactive peptides from BV357 and VM207 differed in molecular mass.

FIG 6

Activity of diacylated CagT. (A) Whole-cell lysates from H. pylori strains 26695, 26695 ΔPAI, and VM211 (Δlnt) were analyzed by immunoblotting with anti-CagT antisera. A representative blot is shown. (B and C) AGS cells were cultured alone or in the presence of H. pylori strains at an MOI of 100:1 for 7 h. Cell culture supernatants were analyzed for IL-8 by ELISA (B), and cell lysates were analyzed by immunoblotting (C) using an antibody recognizing phospho-Tyr (PY99) to detect phosphorylated CagA (indicated by an arrowhead) or antiserum directed against CagA (to detect total CagA). Multiple additional bands (unrelated to CagA) were detected by the anti-phospho-Tyr antibody in all samples, including AGS cells alone. Results in panel B represent means and standard deviations of three biological replicates, each analyzed in triplicate; results in panel C are representative of three biological replicates. Asterisks denote results that were significantly different from the 26695 ΔPAI control results (ANOVA followed by Dunnett’s post hoc test, P < 0.05).

FIG 7

Requirement of a lipobox cysteine residue for CagT stability. (A) Amino-terminal amino acid sequences of CagT and mutant forms of CagT analyzed in this study. The CagT lipobox is highlighted in red, and the disrupted lipobox (CagT-C21S) is highlighted in blue. (B) Expression of CagT was evaluated by immunoblotting extracts of strains 26695 (WT), BV199 (ΔcagT), BV321 (in which a WT copy of cagT was introduced into the ureA locus-restored cagT strain), BV218 (cagT-C21S), and BV260 (cagT1) using anti-CagT. Vertical line indicates cropping of an additional unreported lane from the image. (C) H. pylori strains were cocultured with AGS cells, and the ability of each strain to induce IL-8 production was determined by ELISA. Asterisks denote results that were significantly different from the BV199 control results (ANOVA followed by Dunnett’s post hoc test, P < 0.05).

FIG 8

An intact lipobox is required for CagT stability. (A and B) SignalP 5.0 predicts signal peptide cleavage of wild-type CagT and a mutant CagT protein (CagT2) in which the H. pylori VacA signal peptide was fused to CagT (72). The sites of predicted signal peptide cleavage by LspA (LIPO) or by signal peptidase I (SP) and the corresponding cleavage sites (CS) are shown. (C) Immunoblot showing steady-state levels of CagT 26695 (WT), BV199 (ΔcagT), BV357 (CagT-DDK), and VM253 (cagT2) determined using anti-CagT. Vertical line indicates cropping of an additional unreported lane from the image. (D) H. pylori strains were cocultured with AGS cells, and the ability of each strain to induce IL-8 production was determined by ELISA. Asterisks denote results that were significantly different from the BV199 control results (ANOVA followed by Dunnett’s post hoc test, P < 0.05).

FIG 9

The lipidated amino-terminal end of CagT is positioned at the interface between the Cag T4SS and the H. pylori outer membrane. (A and B) Ribbon representations of CagT, CagX, and CagY (PDB accession no. 6OEE, 6OEG, and 6OEF, respectively [31]) within the Cag T4SS outer membrane core complex are shown in side (A) and top (B) views. The amino-terminal residue resolved in the CagT structure is lysine 26 and is highlighted in red. (C) The amino-terminal amino acid sequence of CagT is shown with an asterisk indicating the site of lipidation, and lysine 26 is highlighted in red.