Outer inflammatory protein a (OipA) of Helicobacter pylori is regulated by host cell contact and mediates CagA translocation and interleukin-8 response only in the presence of a functional cag pathogenicity island type IV secretion system

Abstract

OipA is a phase-variable virulence factor of Helicobacter pylori. Mutations in oipA to turn the gene phase on in a cag pathogenicity island (PAI)-negative strain of H. pylori (J68) or phase off in a cag PAI-positive strain (26695) demonstrated that phase on oipA alleles in both strains had both increased oipA mRNA and human gastric adenocarcinoma (AGS) cell adherence compared to isogenic oipA phase off mutants. An oipA phase off mutant of H. pylori 26695 demonstrated decreased IL-8 secretion by AGS cells and failure to translocate the cag PAI effector CagA. Increased attachment by OipA expressing cag PAI-negative H. pylori J68 failed to alter secreted IL-8 levels. Thus, OipA is necessary but not sufficient for the induction of IL-8; however, it is necessary for translocation of the oncoprotein CagA. Perhaps the nearly invariant phase on status of oipA alleles among cag PAI-positive H. pylori isolates relates to the role of this outer membrane protein in effective translocation of CagA. oipA mRNA comparisons between AGS cell-adherent and non-adherent H. pylori 26695 revealed significantly greater levels in the adherent cells. This may allow H. pylori to adapt to conditions of host cell contact by altering expression of this virulence factor.

Keywords: CagA translocation; adhesin; interleukin-8; phase variation.

Figure 1.

AFLP analyses demonstrates that phase variation in oipA in vitro is infrequent. Fluorescently labeled amplicons containing the poly CT repeat tract of oipA in both cag positive (26695) and cag PAI negative (J68) strains of H. pylori demonstrate relatively minor levels of variation in length, indicating that the expression phase status of oipA is quite stable when H. pylori is grown in vitro. Bars represent the mean ± standard deviation calculated using three biological replicates. Amplification of a similar sized amplicon from a non-repeat bearing region of oipA, as a PCR control, revealed no fragment polymorphisms (data not shown).

Figure 2.

Phase off status of oipA in both cag PAI-positive and negative strains of H. pylori causes a significant decrease in the relative quantity of transcripts expressed compared to phase on status. RT-qPCR was used to determine the relative expression of oipA in the cag PAI-positive H. pylori 26695 wild-type and the isogenic rdxA⁻ control strain, possessing oipA phase on, as well as to 26695 oipA phase off and oipA null mutants (A). Additionally, wild-type cag PAI-negative H. pylori strain J68 and isogenic rdxA⁻ control, as well as to the J68 oipA phase on and J68 oipA null mutants (B). The data shown here are representative of the results obtained in three independent experiments, each conducted in technical triplicate. Error bars show standard deviation. Statistics were calculated using a Welch's unpaired t-test of unequal variance with 26695 rdxA⁻ (oipA phase on) as the control (^** = P ≤ 0.01, ^* = P ≤ 0.05, n.s. = P > 0.05).

Figure 3.

Phase on status of oipA in the cag PAI-negative H. pylori strain J68 results in expression of the protein OipA. An anti-FLAG western blot was performed using 0.4 OD₆₀₀ units of (A) wild-type H. pylori J68 oipA phase off control, (B): H. pylori J68 oipA phase on mutant, (C): H. pylori J68 oipA phase off FLAG, (D): H. pylori J68 oipA phase on FLAG. This revealed the ∼34 kDa FLAG epitope tagged OipA only when oipA is phase on. An ∼30 kDa cross-reactive protein in H. pylori J68 appeared in all strains.

Figure 4.

Helicobacter pylori exhibits increased adherence ability when oipA is phase on in both cag PAI-positive or negative strains. Attachment assays were performed to determine whether oipA phase was important in the ability of H. pylori to adhere to AGS cells, particularly in the cag PAI-negative strain, J68, in which OipA is not expressed. Panel A shows the cag PAI-positive strain 26695, with oipA in alternate expression phases and panel B shows H. pylori J68 with oipA in each expression phase. The data shown here are representative of results obtained in independent experiments, each conducted in technical triplicate. Error bars show standard deviation. Statistics were calculated using a Welch's unpaired t-test of unequal variance (^* = P ≤ 0.05).

Figure 5.

Phase on status of oipA in H. pylori mediates gastric epithelial cell IL-8 production only in the presence of the cag PAI. Cell medium from AGS cells infected with H. pylori 26695 oipA phase on (26695 rdxA⁻), oipA phase on/cagE null mutant (26695 cagE⁻), oipA phase off (26695 oipA phase off), oipA phase off/cagE null mutant (26695 oipA phase off/cagE⁻) as well as from uninfected AGS cells was collected after a 5-h infection and used in ELISA to quantify AGS cell IL-8 production. This is a representative of three independent experiments (A). The same assay was performed for the cag PAI-negative H. pylori strain J68 using oipA phase off (J68 rdxA⁻), oipA null mutant (J68 oipA⁻) and oipA phase on (J68 oipA phase on). Medium form uninfected cells after a 5-h incubation is also shown. The data shown here are representative of the results obtained in three independent experiments, each conducted in technical triplicate (B). Error bars show standard deviation. Statistics were calculated using a Welch's unpaired t-test of unequal variance (^** = P ≤ 0.01, n.s. = P > 0.05).

Figure 6.

OipA is Involved in CagA Translocation from H. pylori 26695 into AGS cells. AGS cells were infected 14 h with H. pylori 26695 rdxA⁻ (oipA phase on), H. pylori 26695/rdxA⁻/oipA phase off, H. pylori 26695/rdxA⁻/oipA phase on/cagE⁻ or left uninfected. Adherent H. pylori and AGS cells were subjected to western blotting using either anti phosphotyrosine to visualize translocated CagA in its tyrosine phosphorylated form, anti-CagA, or anti glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a loading control. CagA protein is translocated and tyrosine phosphorylated only in the presence of phase on oipA allele and a functional T4SS. Changing the expression phase of oipA to phase off or ablating the cagE gene of the T4SS prevented the translocation of this effector protein. These results are representative of the results obtained in two independent experiments.

Figure 7.

Helicobacter pylori attached to AGS cells possess higher levels of oipA transcript. Attachment assays were performed using 26695 oipA phase on (A) and 26695 oipA phase off (B). For each infection, H. pylori in the supernatant was collected separately from the bacteria attached to the AGS monolayer. RNA was extracted, and cDNA synthesized was used in RT-qPCR to determine relative quantities of oipA transcripts in attached cells compared to non-attached. The data shown here are representative of the results obtained in independent experiments, each conducted in technical triplicate. Error bars show standard deviation. Statistics were calculated using a Welch's unpaired t-test of unequal variance (^* = P ≤ 0.05, n.s. = P > 0.05).