Helicobacter pylori cholesterol-α-glucosyltransferase manipulates cholesterol for bacterial adherence to gastric epithelial cells.

Helicobacter pylori infection is associated with several gastrointestinal diseases, including gastritis, peptic ulcers, and gastric cancer. Infection of cells with H. pylori is dependent on lipid rafts, which are cholesterol-rich microdomains located in the cell membrane. H. pylori cholesterol-α-glucosyltransferase (CGT) catalyzes the conversion of membrane cholesterol to cholesteryl glucosides, which can be incorporated into the bacterial cell wall, facilitating evasion from immune defense and colonization in the host. However, the detailed mechanisms underlying this process remain to be explored. In this study, we discovered for the first time that H. pylori CGT could promote adherence to gastric epithelial cells in a cholesterol-dependent manner. Externalization of cell membrane phosphatidylserine (PS) is crucial for enhancement of binding of H. pylori to cells by CGT and for cytotoxin-associated gene A (CagA)-induced pathogenesis. Furthermore, exogenous cholesterol interferes with the actions of H. pylori CGT to catalyze cellular cholesterol, which impedes bacterial binding to cells and attenuates subsequent inflammation, indicating that the initial attachment of H. pylori to cells is closely dependent on host cholesterol. These results provide evidence that CGT contributes to H. pylori infectivity and it may serve as a key target for the treatment of H. pylori-associated diseases.

Introduction

Helicobacter pylori, a gram-negative microaerophilic bacterium, usually colonizes the human stomach. H. pylori infection is associated with a high incidence of gastrointestinal diseases, including gastritis, peptic ulcers, and gastric cancer [1]. Several important virulence factors have been found to contribute to H. pylori pathogenesis, including vacuolating cytotoxin A (VacA) and cytotoxin-associated gene A (CagA) [2]. VacA is classified as a pore-forming toxin that possesses the capacity to stimulate intracellular acidic vacuole formation and disrupt cellular homeostasis, leading to apoptosis [3]. CagA can be translocated by the H. pylori type IV secretion system (TFSS) upon attachment to cells, causing chronic inflammation and oncogenesis in gastric epithelial cells [4].

Other adhesion molecules have also been discovered in H. pylori, such as Leb blood group antigen binding adhesin (BabA) [5] and sialic acid–binding adhesin (SabA) [6], which bind to Lewis antigens and gangliosides that are expressed in human gastric epithelial cells [7,8]. In addition to adhesion molecules, cholesterol-α-glucosyltransferase (CGT) has been found to consolidate bacterial colonization in mouse models [9]. CGT is mainly located on the bacterial inner membrane and is encoded by hp0421 [10]. H. pylori takes up host cholesterol, which is then converted into cholesterol-α-D-glucopyranoside (αCG), which is further modified to cholesteryl-6-O-tetradecanoyl-α-D-glucopyranoside (αCAG) and cholesteryl-6-O-phosphatidyl-α- D-glucopyranoside (αCPG) [9-13]. The cholesteryl glucosides can be recruited to the H. pylori cell wall, which is closely associated with immune evasion [9,14].

Lipid rafts are membrane cholesterol-rich microdomains that provide entry portals for many bacterial pathogens or their virulence factors [15]. Both CagA translocation and VacA delivery into host cells following H. pylori infection require lipid rafts [16-18]. H. pylori exploits host externalized phosphatidylserine (PS) for CagA delivery via TFSS and it subsequently induces pathogenesis [19]. These results are in accordance with the finding that cgt-deficient H. pylori show a reduced ability to introduce CagA by TFSS [20]. However, detailed information on the relationship between CGT and membrane PS in response to H. pylori colonization of host cells is required for further studies. This study investigated how H. pylori CGT manipulates cellular cholesterol and membrane PS, which contribute to the bacterial pathogenesis of the host. Furthermore, H. pylori CGT-enhanced bacterial adherence could be dampened by the addition of soluble cholesterol, which elucidated the different strategies for controlling H. pylori infection in the clinic.

Results

H. pylori CGT promotes bacterial adherence to cells

H. pylori CGT is involved in pathogenesis through mechanisms such as host colonization [13] and CagA translocation [20]. H. pylori adherence to cells is the initial step in the infection of host cells and it is essential for CagA translocation via TFSS. To investigate whether H. pylori CGT affects H. pylori attachment to gastric epithelial cells, we constructed H. pylori isogenic mutants, including cgt knockout (∆cgt), cgt knockin (∆cgt-in), and cgt dead-mutant (E285A) (Fig. S1). AGS cells (human gastric adenocarcinoma cell line) were infected with H. pylori WT or isogenic mutants for 6 h, followed by adhesion and invasion assays. Figure 1a and 1b show that the adhesion and invasion activities of WT and ∆cgt-in cells were much higher than those of ∆cgt and cgt-dead mutant strains. These results were also supported by the results of competition experiment involving co-infection with WT and ∆cgt H. pylori (1:1) (Figure 1c), indicating that CGT-bearing H. pylori is superior in bacterial attachment and internalization to gastric epithelial cells. We further conducted a time-course analysis of the bacterial adhesion assay (Figure 1d) and internalization assay (Figure 1e). Our results showed that WT H. pylori possesses higher adherence and intracellular survival than ∆cgt in a time-dependent manner. We then investigated H. pylori localization in infected cells using immunofluorescence microscopy. As shown in Figure 2, WT H. pylori adhering to AGS cells was much greater than that of ∆cgt. These results demonstrate that CGT promotes H. pylori adhesion to gastric epithelial cells, thereby contributing to its internalization by cells.

Figure 1.

H. pylori CGT is essential for bacterial adherence to cells. AGS cells were infected with WT, ∆cgt, ∆cgt-in, and cgt dead-mutant H. pylori at a MOI of 50 for 6 h. H. pylori (a) adhesion to and (b) invasion of cells were analyzed and expressed as CFU. (c) Competition assay involving co-infection with WT and ∆cgt H. pylori (1:1) was conducted. Cells where viable WT and ∆cgt H. pylori both adhered to and invaded, or only invaded were enumerated, and plated onto chloramphenicol (34 μg/ml) supplemented blood agar plates to select ∆cgt H. pylori. Cells were infected with H. pylori at a MOI of 100 for 3, 6, and 12 h, and (d) adhesion and (e) invasion assays were performed. *, P< 0.05

Figure 2.

Mutation of CGT reduces H. pylori binding to AGS cells. (a) AGS cells were uninfected or infected with WT or ∆cgt H. pylori at a MOI of 20 for 6 h. Cells were probed with anti-H. pylori antibody (red) and stained with DAPI (blue) to visualize cell nuclei. The stained samples were analyzed using confocal microscopy. Scale bar, 5 μm. (b) Total H. pylori were enumerated and divided by the number of AGS cells in each field. The mean number of bacteria was calculated using three different fields. *, P< 0.05

Disruption of lipid rafts decreases H. pylori infection of cells

Since lipid rafts play a pivotal role in H. pylori pathogenesis in gastric epithelial cells [17], we next explored whether CGT is involved in this process. AGS cells were pretreated with a cholesterol-depleting agent, methyl-β-cyclodextrin (MβCD), prior to infection with WT or ∆cgt H. pylori. MβCD pretreatment decreased bacterial adhesion to cells in both the WT and ∆cgt strains (Figure 3a). Similarly, the intracellular survival rate was also reduced in MβCD-pretreated cells (Figure 3b). We then employed immunofluorescence microscopy to investigate whether disruption of lipid rafts by MβCD inhibits H. pylori infection in cells. As shown in Figure 4, raft clustering resulted from WT H. pylori infection, but not from ∆cgt. Notably, the number of WT H. pylori attached to the cell membrane was markedly decreased in MβCD-pretreated cells. These results demonstrate that both CGT and the integrity of lipid rafts are essential for H. pylori infection.

Figure 3.

Involvement of lipid rafts in H. pylori infection. (a) AGS cells were untreated or pretreated with 5 mM MβCD for 30 min, followed by infection with WT or ∆cgt H. pylori at a MOI of 100 for 6 h. H. pylori (A) adhesion to and (b) invasion of cells were analyzed. *, P< 0.05

Figure 4.

H. pylori CGT coalesces lipid rafts on the cell membrane. (a) AGS cells were untreated or pretreated with 5 mM MβCD for 30 minutes, followed by H. pylori infection at a MOI of 20 for 6 h. Cells were probed with anti-H. pylori antibody (red) and FITC-CTX-B to visualize GM1 (green), and DAPI (blue) to visualize cell nuclei. The stained samples were analyzed by confocal microscopy. Scale bar, 5 μm. (b) Total number of H. pylori were counted and divided by the number of AGS cells untreated or pretreated with MβCD in each field. The mean number was calculated using three different fields. *, P< 0.05

PS externalization is the consequence of H. pylori adhesion

H. pylori infection triggers non-apoptotic externalization of PS, which interacts with CagA, resulting in CagA translocation [19]. In addition, CGT-bearing H. pylori has been shown to enhance CagA translocation [20]. Using confocal microscopy, we further analyzed whether H. pylori CGT regulates translocation of CagA by altering cellular PS. Figure 5 shows that ∆cgt adheres to cells much less than the WT strain. Externalization of membrane PS during H. pylori infection was assessed using flow cytometry. As shown in Figure 6, WT H. pylori triggered more PS externalization (6.7%) than that resulting from ∆cgt infection (1.5%). When cells were pretreated with MβCD followed by infection with WT H. pylori, a decrease in PS externalization occurred compared to that in untreated cells. Although PS externalization was reduced in MβCD-treated AGS cells infected with ∆cgt, there was no significant difference between the MβCD treatment and the untreated control. These results indicate that PS externalization is essential for the CGT-promoted adhesion of H. pylori to cells.

Figure 5.

H. pylori co-localizes with externalized PS. AGS cells were uninfected or infected with H. pylori at a MOI of 20 for 6 h. Cells were probed with anti-H. pylori antibody (red) or stained with anti-PS antibody (green), and DAPI to visualize cell nuclei (blue). The stained cells were analyzed by confocal microscopy. Arrows indicated H. pylori attachment to cell membrane and colocalization with PS. Scale bar, 5 μm

Figure 6.

H. pylori CGT triggers membrane PS externalization. (a) AGS cells were untreated or pretreated with 5 mM MβCD for 30 min, followed by infection with WT or ∆cgt H. pylori at a MOI of 100 for 6 h. The cells were probed with anti-PS antibody, and analyzed via flow cytometry. Cells were treated with H2O2 (10 or 20 mM) for 4 h as a positive control. (b) The level of membrane PS externalization was determined

Cholesterol coating of H. pylori inhibits its adhesion to cells

As CGT-containing H. pylori coated with excessive cholesterol influences inflammatory signaling [13], we further tested whether exogenous cholesterol affects CGT-mediated H. pylori adhesion to cells. Our results showed that coating H. pylori with water-soluble cholesterol (5 mg/mL) did not affect the bacterial viability (Fig. S2). Treatment of bacteria with exogenous cholesterol markedly decreased WT adherence to AGS cells compared to that of non-cholesterol-coated H. pylori (Figure 7a). Similar results were obtained for the ∆cgt strain but with less reduction. When higher concentrations of cholesterol were added (0.25 and 0.5 mg/mL), a significant decrease in WT H. pylori adhesion to cells was observed compared with that in cholesterol-untreated cells (Figure 7b). Additionally, treatment with water-soluble cholesterol led to remarkable inhibition of H. pylori CagA translocation/phosphorylation (Figure 7c-e), NF-κB activation (Fig. S3A), and IL-8 production (Fig. S3B). This trend was also observed in the ∆cgt strain, but with only a slight effect. The results indicate that the initial attachment of H. pylori to cells is dependent on host cholesterol, while exogenous cholesterol competes with H. pylori CGT actions to catalyze cellular cholesterol, thereby dampening its binding to cells. Taken together, our findings demonstrate that CGT contributes to the initial binding of H. pylori to cells, which is crucial for promoting subsequent pathogenesis.

Figure 7.

Exogenous cholesterol coating of H. pylori decreases its adhesion to cells. (a) H. pylori were untreated or pretreated with water-soluble cholesterol (5 mg/ml) for 1 h before infection. AGS cells pretreated with 5 mM MβCD for 30 min were then infected with WT or Δcgt H. pylori at a MOI of 100 for 6 h. Viable H. pylori that had either adhered to or invaded AGS cells were analyzed. (b) AGS cells were infected with WT or ∆cgt H. pylori at MOI of 100 for 6 h that were untreated or pretreated with water-soluble cholesterol (0.25 and 0.5 mg/ml) for 1 h prior to infection. Viable H. pylori that had either adhered to or invaded AGS cells were analyzed. (c) AGS cells were infected with WT or ∆cgt H. pylori at MOI of 100 for 6 h. H. pylori were pretreated with water-soluble cholesterol (5 mg/ml) for 1 h before infection. The levels of CagA translocation and phosphorylation were assessed using a western blot assay. β-actin was an internal control and relative protein levels were normalized to those in the uninfected group. (d) CagA phosphorylation and (e) translocation were compared between each condition and expressed as fold changes. *, P< 0.05

Discussion

H. pylori infection exploits PS for CagA translocation [19] and CGT is involved in this process in a raft-dependent manner [20]. Although PS externalization plays a pivotal role in H. pylori CagA translocation, the effect of CGT on bacterial adherence and inflammatory responses remains to be explored. In this study, we found that H. pylori CGT elicits membrane cholesterol coalescence, which promotes bacterial adherence to cells, leading to subsequent pathogenesis. Unveiling the mechanism of bacterial virulence factors is particularly crucial, as targeting of key molecules to treat H. pylori-associated diseases has been proposed.

H. pylori CGT converts cholesterol into cholesterol-α-D-glucopyranoside (αCG), cholesteryl-6ʹ-α-O-tetradecanoyl-α-D-glucopyranoside (αCAG), and cholesteryl-6ʹ-α-O- phosphatidyl-α-D-glucopyranoside (αCPG), which are categorized as cholesteryl glucosides [10]. Glucoside derivatives have been shown to generate anti-inflammatory responses [21-24]. Notably, H. pylori utilizes cholesteryl glucosides catalyzed by CGT to evade the host immune system, which is beneficial for its survival in host cells [9,25]. Furthermore, cholesterol glucosylation by H. pylori is crucial for the manipulation of autophagy to delay the macrophage clearance of bacteria [26]. A recent study reported that cholesteryl α-D-glucopyranoside acyltransferase and CGT are packaged in outer membrane vesicles and delivered to host cells, followed by gathering raft-associated molecules, to enhance bacterial attachment to cells [27]. The present study further showed that CGT in H. pylori not only enhanced its adherence ability, but also increased its intracellular survival in gastric epithelial cells. In addition, H. pylori itself expresses flotillin-like protein and Lewis antigens that serve as an immune mimicry to the host, providing an alternative way to evade host immune surveillance [14,28,29]. Together with previous findings, our current study reveals that the mechanisms underlying H. pylori infection involve action of CGT to orchestrate immune defense via cholesterol glucosylation, resulting in successful host survival.

H. pylori is much more likely to induce host pathogenesis by utilizing CGT to sense cholesterol-rich regions such as membrane lipid rafts, which mediate many responses during H. pylori infection [9]. However, a recent study indicated that H. pylori coated with excess cholesterol prior to infection failed to block the IFNγ-mediated inflammatory pathway [13]. In our study, treatment of CGT-bearing H. pylori with water-soluble cholesterol showed a decrease in its adherence to AGS cells when compared to that of H. pylori not coated with cholesterol. The binding of H. pylori to cells is largely dependent on host cholesterol, while exogenous cholesterol-coated H. pylori competes with CGT to modify host cholesterol to generate cholesteryl glucosides, resulting in decreased bacterial adhesion to cells. These results indicate that the source of cholesterol is important for H. pylori CGT functions, which contributes to the initial binding of bacteria to cells and subsequent pathogenesis.

Several molecules have been found in membrane lipid rafts that are involved in the adhesion of H. pylori to cells, including PAR1/MARK, Lewis antigens, and integrin α5β1 [30,31]. Depletion of cholesterol affects membrane composition, which has an indirect effect on the ability of H. pylori to adhere to cells. This phenomenon motivated us to further investigate whether a mutation of CGT in H. pylori could exert a direct inhibitory effect on bacterial adhesion to cells. CGT catalyzes the conversion of cholesterol to cholesteryl glucosides, which are then incorporated into the bacterial cell wall, and ∆cgt cannot catalyze this conversion [9]. Our results showed that without MβCD treatment, ∆cgt significantly decreased the binding of H. pylori to cells compared with that of the wild-type strain. These results excluded the effects of MβCD on membrane proteins and demonstrated that the abrogation of cholesterol glucosylation reduces the ability of H. pylori to adhere to cells. Our findings are supported by previous studies showing that the dependence on cholesterol is quite important [32,33]. This is because ∆cgt is unable to glucosylate cholesterol, leading to decreased colonization in murine models [13].

Bacterial virulence factors employing membrane cholesterol microdomains to gain access to host cells have been reported previously [34-36]. Cholesterol usurping/depleting agents have been used to alleviate infectious diseases by preventing microbial entry into host cells [37]. For instance, cholesterol-lowering agents (i.e., statins) potentially attenuate pathogen infectivity [38-41]. Another cholesterol-depleting agent, MβCD, is commonly employed to reduce microbial adherence to the host cell membrane [36,42,43]. Our previous studies demonstrated that statin use significantly reduced the incidence of H. pylori-related diseases [41,44,45]. Moreover, hijacking cholesterol by antagonists competed for VacA and CagA actions via lipid rafts alleviated H. pylori-induced pathogenesis [46]. Consistent with these findings, the present study showed that depletion of cholesterol interferes with H. pylori adherence and internalization into cells, which depends on CGT. These lines of evidence suggest that pharmaceutical targeting of lipid rafts should be developed to treat H. pylori infection.

Although the in vitro cell models have demonstrated that CGT contributes to H. pylori adherence to gastric epithelial cells, some limitations exist in the present study. For example, only the AGS cell line was used as a cell-based assay platform. Although AGS cells have been well studied in the H. pylori infection models, more gastric cell lines should be considered to validate the findings. In addition, the direct linkage between cholesterol, CGT, and H. pylori infectivity using long-term animal or in vivo studies deserves further investigations.

In summary, this study demonstrates that H. pylori CGT drives cholesterol glucosylation, which is pivotal for the exploitation of lipid rafts as a foothold in its initial attachment to cells. Moreover, membrane PS externalization is essential for CGT-promoting H. pylori adhesion to cells, followed by CagA translocation and phosphorylation (Figure 8). The results further show that exogenous cholesterol competes with CGT actions in modifying cellular cholesterol, which decreases bacterial adherence and inflammation of cells. Exploring the mechanism of how CGT contributes to H. pylori infectivity might provide an opportunity to develop new agents to alleviate its pathogenesis.

Figure 8.

Hypothesized model depicts that CGT is crucial for H. pylori adhesion to gastric epithelial cells and subsequent pathogenesis. (1) WT (cgt) H. pylori utilizes cholesterol to adhere to cholesterol-rich microdomains in cell membrane. (2) H. pylori coalesces lipid rafts, then (3) triggers non-apoptotic PS externalization to (4) enhance CagA translocation via TFSS, leading to inflammation and pathogenesis of host cells

Materials and methods

Cell culture

AGS cells (ATCC CRL-1739) were cultured in F12 medium (Thermo Fisher Scientific, Waltham, MA, USA) containing 10% fetal bovine serum (Hyclone, Logan, UT, USA). Antibiotics were not added to the cell culture medium in the H. pylori-infected assay. Cells were cultured at 37°C in a 5% CO2 atmosphere.

Bacterial culture

H. pylori 26,695 (ATCC 700392) wild-type (WT), cgt knockout (∆cgt), cgt knockin (∆cgt-in), and cgt dead-mutant (E285A) strains were constructed (Fig. S1). Each isogenic mutant H. pylori was generated by the insertion of an antibiotic resistance cassette via allelic replacement, as described previously [26]. H. pylori wild-type and isogenic mutants were routinely cultured on blood agar plates (Brucella agar with 10% defibrinated sheep blood) and incubated in a microaerophilic environment (10% CO2, 5% O2, 85% N2) at 37°C.

Gentamycin protection assay

AGS cells (2 × 105) were seeded in 12-well plates and cultured for 16 h. The cells were treated with 5 mM methyl-β-cyclodextrin (MβCD) (Sigma-Aldrich, St. Louis, MO, USA) for 30 min, followed by infection with H. pylori strains (WT, ∆cgt, and E285A) at an MOI of 100 for 3, 6, and 12 h. The infected cells were washed with PBS prior to incubation with 100 μg/mL gentamycin (Sigma-Aldrich) for 90 min. The cells were then lysed using sterile water. The cell lysate was serially diluted onto blood agar plates and incubated for 5 days. Viable H. pylori were counted and indicated as colony-forming units (CFUs).

Immunofluorescence staining and confocal microscopic analysis

AGS cells (2 × 105) were seeded in 6-well plates and cultured for 16 h. The cells were treated with 5 mM MβCD for 30 min prior to H. pylori infection at an MOI of 20 for 6 h. Then, the cells were fixed with 1% paraformaldehyde (Alfa Aesar, Haverhill, MA, USA) for 1 h, permeabilized with 0.1% Triton X-100 for 30 min, and blocked with 1% FBS for 1 h. The cells were probed with an anti-H. pylori antibody (Bio-Techne, Minneapolis, MN, USA) and anti-phosphatidylserine antibodies (MERCK, Darmstadt, Germany), respectively. The nuclei and membrane lipid rafts were stained with 4ʹ,6-diamidino-2-phenylindole (DAPI) and fluorescein isothiocyanate-conjugated cholera toxin subunit B (FITC-CTX-B) (Invitrogen, Carlsbad, CA, USA). The fluorescent signals were analyzed using a confocal laser scanning microscope (LSM780, Carl Zeiss, Oberkochen, Germany).

Flow cytometry

AGS cells (2 × 105) were seeded in 6-cm plates and cultured for 16 h. The cells were treated with 5 mM MβCD for 30 min, followed by WT or ∆cgt H. pylori infection at an MOI of 100 for 6 h. The infected cells were washed with PBS and probed with an anti-phosphatidylserine antibody at room temperature for 1 h. The cells were incubated with FITC-conjugated goat anti-mouse IgG1 antibody at room temperature for 30 min. The stained cells were determined using a FACSCalibur™ flow cytometer (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) and analyzed using the Cell Quest software WinMDI (Verity Software House, Topsham, ME, USA).

Western blot assay

H. pylori were pretreated with water-soluble cholesterol (5 mg/mL) for 1 h prior to infection. AGS cells were treated with 5 mM MβCD for 30 min, followed by infection with WT or Δcgt H. pylori at an MOI of 100 for 6 h. H. pylori-infected cells were analyzed by 6% SDS-PAGE followed by transfer onto polyvinylidene difluoride membranes (Millipore, Burlington, MA, USA). The membranes were blocked with 5% skim milk for 1 h prior to incubation with the primary antibody at 4°C overnight. The membrane was then incubated with a horseradish peroxidase-conjugated secondary antibody (Millipore). Protein expression levels were analyzed using ECL Western Blotting Detection Reagents (GE Healthcare, Chicago, IL, USA) and analyzed using the Azure 400 system and AzureSpot Analysis Software (Azure Biosystems, Dublin, CA, USA).

Statistical analysis

Experimental results are expressed as the mean ± standard deviation of independent triplicate experiments. Student’s t-test was used to calculate the statistical significance of differences between the two groups. Differences were considered significant at P< 0.05. Statistical analysis was performed using Prism6 (Graph Pad, San Diego, CA, USA).

Click here for additional data file.