ABSTRACT
The glucosylation of free cholesterol (FC) by Helicobacter pylori cells has various biological significances for the survival of this bacterium. H. pylori cells with glucosylated FC are capable of evading host immune systems, such as phagocytosis by macrophages and activation of antigen-specific T cells, and surviving in the gastric mucosal tissues for long periods. An additional role of cholesterol glucosylation in the survival of H. pylori which is distinct from the role of escaping the host immune system, however, has yet to be identified. This study demonstrated that 7-dehydrocholesterol (7dFC), an FC precursor, is a toxic compound fatal to H. pylori cells, but the cell membrane of H. pylori is capable of absorbing this toxic sterol via glucosylation. In contrast to the case with 7dFC, no toxicity to H. pylori cells was detected from the glucosylated 7dFC. In addition, cgt gene mutant H. pylori cells that cannot glucosylate cholesterols had higher susceptibility to the toxic action of 7dFC than wild-type H. pylori cells. These results indicate that the cgt gene product of H. pylori serves to detoxify the sterol fatal to this bacterium and to permit this toxic sterol as a cell membrane lipid component. In summary, this study defined a novel role of cholesterol glucosylation in H. pylori.
INTRODUCTION
Infection of Helicobacter pylori in the stomach of humans is responsible for gastritis and peptic ulcers and further contributes to the development of gastric cancer and marginal zone B-cell lymphoma (1–8). The majority of people infected with H. pylori, however, are asymptomatic and therefore are not aware of being infected with this pathogen unless they are tested for it. H. pylori can exist for many years in the human stomach by somehow escaping the host immune system. The glucosylation of nonesterified (or free) cholesterol that has been incorporated into the membranes of H. pylori cells is an important mechanism for evading the pathogen exclusion mechanisms of the hosts (9). Via cholesterol glucosylation, H. pylori acquires resistance against the phagocytosis of macrophages, regulates the activation of antigen-specific T cells, and survives for long periods in the gastric mucosal tissues (9, 10).
Cholesterol-α-glucosyltransferase (CGT), involved in the biosynthesis of glucosyl cholesterol, is encoded by the cgt (or hp0421) gene on the chromosomal DNA of H. pylori (11). The product of the cgt gene catalyzes the dehydration reaction between a 1α-hydroxyl (OH) group in a d-glucose molecule and a 3β-OH group in a free-cholesterol (FC) molecule. H. pylori cells synthesize at least the following three types of glucosyl cholesterols (12): cholesteryl-α-d-glucopyranoside (CGL), cholesteryl-6-O-tetradecanoyl-α-d-glucopyranoside (CAG), and cholesteryl-6-O-phosphatidyl-α-d-glucopyranoside (CPG). The transferases that attach a fatty acid molecule or phosphatidyl group to the CGL molecule have yet to be identified.
Our previous study demonstrated that H. pylori glucosylates not only FC but also 3β-OH steroids, such as pregnenolone and dehydroepiandrosterone (13). This indicates that H. pylori cells have the potent ability to interact with a number of 3β-OH steroid compounds and to assimilate those steroids into the cell membranes in order to maintain viability. Moreover, a previous study by our group revealed that progesterone and its synthetic derivative, possessing a 3-oxo group, have the capability to lyse H. pylori cells (14). In addition to those 3-oxo steroids, estrone and estradiol, possessing a 3-OH group, have been found to exhibit different effects on the H. pylori cell: estrone is efficiently assimilated without glucosylation into the membranes of H. pylori cells, but estradiol inhibits the growth of the cells with bacteriostatic action (13–15). A recent study by another group has demonstrated that some bile salts, the steroid compounds possessing a 3α-OH group, show an inhibitory effect on the growth of H. pylori (16). However, the 3β-OH steroid compounds that impair the viability of H. pylori have yet to be identified.
7-Dehydrocholesterol (7dFC) is the direct precursor of FC in the cholesterol biosynthetic pathway in mammalian tissues and also possesses the 3β-OH group. FC is synthesized from 7dFC via the catalytic action of 7dFC reductase encoded by a gene on chromosome 11 in humans (17). A mutation of the gene for 7dFC reductase causes the accumulation of 7dFC in plasma and tissues and is involved in the development of Smith-Lemli-Opitz (SLO) syndrome (17, 18). The concentration of 7dFC in the plasma of SLO patients is approximately 100 μM to 400 μM, whereas the concentration of this sterol in normal plasma is less than 10 μM (18–21). Our study in 2009 revealed that some phosphatidylcholine (PC) molecular species lyse the H. pylori cells without steroid compound, but the H. pylori cells with absorbed steroid compound (FC or estrone) acquire resistance against the bactericidal action of PC (15). The acquisition of resistance by H. pylori cells with absorbed steroids against the PC is caused regardless of glucosylation of the steroids. A number of PC molecular species exist in amounts enough to kill H. pylori in human gastric juice or gastric mucus (22, 23). Given that SLO patients have a mutation only in the 7dFC reductase gene, we can assume that PC is at the normal concentration even in the gastric mucosal tissues of SLO patients. These studies may suggest a possibility that H. pylori comes into contact with 7dFC in order to acquire resistance against the bactericidal action of PC in the stomach of SLO patients, although the amounts of 7dFC in the gastric mucosal tissues of SLO patients are unclear.
The differences between the chemical structures of FC and 7dFC are in the number of hydrogen atoms in the steroid framework (see Fig. 1). This precursor lacks the two hydrogen atoms at carbon positions 7 and 8 of the FC molecule. None of the earlier studies has, however, examined whether 7dFC is useful or harmful for the survival of H. pylori. The aims of this study were to investigate the influence of 7dFC on the growth of H. pylori and to define a novel role of cholesterol glucosylation in this bacterium that is distinct from the roles of cholesterol glucosylation reported previously (9, 10).
7dFC as a toxic compound fatal to H. pylori cells. H. pylori cells were cultured in the dark for 24 h with either FC or 7dFC at various concentrations (1, 10, 50, and 100 μM) in PPLO broth (1.5 ml) under microaerobic conditions at 37°C. After culturing, the CFU counts were determined. The gray bars in the graphs are the baseline CFU levels measured immediately after the cultures were started. Results are indicated as the mean CFU ± SD obtained from the three strains of H. pylori (NCTC 11638, ATCC 43504, and 26695) and are representative of one of three independent experiments.
MATERIALS AND METHODS
Bacterial strains and cultures.This study examined three Helicobacter pylori strains (NCTC 11638, ATCC 43504, and 26695) and, for the cultures, used a PPLO broth (Difco Laboratories, Detroit, MI) with or without 2,6-di-O-methyl-β-cyclodextrin (dMβCD) (Sigma-Aldrich Inc., MO) or a brain heart infusion (BHI) broth (Difco Laboratories) containing 5% horse serum (HS; Gibco, Auckland, New Zealand). H. pylori cells were cultured in the dark with continuous shaking under microaerobic conditions at 37°C in an atmosphere of 10% CO2, 5% O2, and 85% N2.
Cholesterols.FC (Wako Pure Chemical Industries Ltd., Tokyo, Japan) or 7dFC (Sigma-Aldrich Inc.) was dissolved in ethanol, adjusted to a 50 mM concentration, and stored in the dark at −80°C as a stock solution until used in experiments. The 0.2% concentration of ethanol used in this study had no influence on the viability of H. pylori.
Assay of growth of H. pylori in the presence of cholesterols.H. pylori cells (approximately 106.5 or 106 CFU) were cultured for 24 h in PPLO broth (1.5 ml) in the presence of cholesterol (FC or 7dFC) at various concentrations, and then CFU counts were determined by the conventional method (13). In addition, H. pylori cells were cultured for 24 h in PPLO broth (1.5 ml) containing 7dFC (100 μM) in the presence of dMβCD (15, 75, and 150 μM) to measure the CFU.
Assay of binding of cholesterols to H. pylori cells.H. pylori cells (approximately 109 CFU) were suspended in PPLO broth (5 ml) with or without dMβCD (3 mM) in the presence of a paper disk dotted with 100 nmol of cholesterol (FC or 7dFC) and were incubated in the dark for 4 h with continuous shaking under microaerobic conditions at 37°C (24). After incubation, the cells were recovered and washed to extract membrane lipids using a method described elsewhere (13). The amount of cholesterol incorporated into the membrane lipid compositions was quantified by the ferrous chloride-sulfuric acid method (24). In addition, thin-layer chromatography (TLC) analysis was performed to detect the glucosylated cholesterol (FC or 7dFC) in the membrane lipid composition (13). After TLC with a chloroform-methanol-water (70:30:5) solvent system, the TLC plate (Silica Gel G60 plate; Merck, Darmstadt, Germany) was sprayed with an orcinol-sulfuric acid solution (0.2% orcinol in 2 N H2SO4) and heated at 180°C to visualize the spots of glucosyl cholesterols on the plate surface.
Assay of binding of cholesterols to PE of H. pylori.Experimental procedures similar to those described previously were carried out (24). Briefly, phosphatidylethanolamine (PE) isolated from the membrane lipid composition of H. pylori cells was fixed to a paper disk in various amounts, soaked in a 50 mM Tris (pH 7.5) buffer (2 ml) containing dMβCD (3 mM) together with a paper disk dotted with 100 nmol of cholesterol (FC or 7dFC), and incubated in the dark for 4 h with continuous shaking at 25°C. After the PE-fixed paper disk was washed six times with distilled water, the amount of cholesterol in the PE-fixed paper disk was quantified using the ferrous chloride-sulfuric acid method.
Analysis of glucosylation of 7dFC by H. pylori cells pretreated with dMβCD.After H. pylori cells were precultured for 24 h in PPLO broth (100 ml) in the presence or absence of dMβCD (1.5 mM) and washed three times with phosphate-buffered saline (PBS) (10 ml) via centrifugation (10,000 × g, 5 min), the cells (approximately 109 CFU/ml) harvested were suspended in fresh PPLO broth (10 ml) containing 7dFC (30 μM) and incubated in the dark for 4 h with continuous shaking under microaerobic conditions at 37°C. The membrane lipids were then isolated from the H. pylori cells (approximately 1010 CFU) recovered via centrifugation (10,000 × g, 5 min) after incubation with 7dFC, analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, and detected on the TLC plate (Silica Gel G60; Merck) surface treated with a 60% sulfuric acid solution, by heating at 180°C. In addition, the amounts of 7dFC incorporated into H. pylori cells (109 CFU) with or without the dMβCD pretreatment described above were quantified by the ferrous chloride-sulfuric acid method.
Isolation of glucosylated 7dFC from bacterial lipid constituents.Membrane lipids were extracted from H. pylori cells that were cultured for 24 h in the presence of 7dFC (15 μg/ml) in PPLO broth (2 liters) containing 0.2% dMβCD via a method described elsewhere (24). The membrane lipids (10 mg) dissolved in chloroform (1 ml) were applied onto a column (1-cm diameter, 5-cm height) filled with chloroform-activated Iatrobeads (Iatrobeads 6RS-8060; Mitsubishi Kagaku Iatron Inc., Tokyo, Japan), and the column was washed with chloroform (15 ml). Each lipid was eluted from the column sequentially with 10 ml each of chloroform solution and chloroform-acetone solutions at ratios of 9:1, 7:3, 4:6, and 2:8. The fraction from the chloroform-acetone solution at the ratio of 2:8 was dried, acetone (3 ml) was added, and then the specimen in acetone solution was sufficiently sonicated and incubated overnight at −30°C. After the acetone solution (supernatant) was carefully removed from the precipitates, the precipitates were gently washed three times with cold acetone (1 ml) to obtain 7-dehydrocholesteryl-α-d-glucopyranoside (7dCGL).
To confirm the purity of 7dCGL, TLC analysis was performed with a chloroform-methanol-water (70:30:5) solvent system, followed by visualizing the spot of 7dCGL on a Silica Gel G60 plate (Merck) surface treated with a 60% sulfuric acid solution. The purified 7dCGL was dissolved in a chloroform-methanol (2:1) solvent, adjusted to a 9.14 mM concentration, and stored in the dark at −80°C as a stock solution until it was used in experiments.
To assay H. pylori growth in the presence of 7dCGL (100 μM), the purified 7dCGL (150 nmol) in chloroform-methanol (2:1) solvent (16.4 μl) was completely dried at room temperature in the dark to vaporize the solvent and dispersed into PPLO broth (1.5 ml) containing 0.2% ethanol via sonication. H. pylori (approximately 106 CFU) was cultured for 24 h in the 7dCGL (100 μM)-dispersed PPLO broth (1.5 ml) in the dark, and then CFU were measured.
Construction of an H. pylori mutant lacking CGT activity.H. pylori chromosomal DNA in which the cgt gene was disrupted by insertion of the cat (chloramphenicol acetyltransferase) gene cassette was prepared as reported previously (25, 26). Natural transformation was adopted to exchange the chromosomal DNA in recipient cells (26, 27). Briefly, the recipient cells of H. pylori strain 26695 grown for 1 day on an agar plate of 5% HS-BHI were spread at high density on a small area (about 10 mm in diameter) of a fresh 5% HS-BHI agar plate and incubated for 5 h under microaerobic conditions at 37°C. After incubation, cgt gene-disrupted chromosomal DNA (500 ng) in a 10 mM Tris-1 mM EDTA (pH 8.0) buffer (10 μl) was dropped on an area spread with the H. pylori cells and incubated for further 12 h under the same conditions. The H. pylori cells were then spread on an agar plate of 5% HS-BHI containing chloramphenicol (15 μg/ml) to select the chloramphenicol-resistant bacterial colonies after culturing for 10 days.
To ascertain the disappearance of CGT activity, the membrane lipids extracted from the chloramphenicol-resistant H. pylori cells cultured in 5% HS-BHI broth were analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, followed by visualization of the spots of lipid on a Silica Gel G60 plate (Merck) surface treated with a 60% sulfuric acid solution.
Observation of H. pylori cells.The H. pylori cells were observed with a differential interference mode using an AX80T microscope (Olympus Co., Ltd., Tokyo, Japan).
Measurement of OD660.H. pylori cells (approximately 108.5 CFU/ml) cultured in PPLO broth (10 ml) at the stationary growth phase and harvested via centrifugation (10,000 × g, 5 min) were resuspended in fresh PPLO broth (10 ml), and the aliquot cell suspension (1.5 ml) was incubated in the presence or absence of 7dFC (100 μM) at various time points in the dark with shaking under microaerobic conditions at 37°C. After incubation, the cell suspension (1 ml) was recovered via centrifugation (10,000 × g, 5 min) and resuspended in saline (1 ml), and then the optical density at 660 nm (OD660) of the cell suspension (200 μl) was measured using a VersaMax microplate reader (Molecular Devices Co., CA).
Detection of PE in culture supernatant.Chloroform-methanol (2:1) solvent (4 ml) was added to the culture supernatant (800 μl) of H. pylori cells (approximately 108.5 CFU/ml) incubated with 7dFC (100 μM) at various time points, vigorously shaken, and incubated for a few minutes at room temperature to separate the chloroform phase and the water phase. The solvent of chloroform phase was then vaporized under nitrogen airflow to obtain hydrophobic compounds. The hydrophobic compounds were dissolved in a chloroform-methanol (2:1) solvent (40 μl), spotted onto a Silica Gel G60 plate (Merck), and developed on the plate surface via TLC with a chloroform-methanol-water (70:30:5) solvent system. After TLC, the Silica Gel G60 plate was sprayed with ethanol solution containing 0.25% ninhydrin and heated at 120°C to detect the spot of PE on the plate surface.
RESULTS
Influence of 7dFC on the growth of H. pylori.Our first experiment was to examine the influence of 7-dehydrocholesterol (7dFC) on the growth of H. pylori cells. When H. pylori cells were cultured for 24 h in the presence of free cholesterol (FC) or 7dFC, the FC had no influence on the growth of H. pylori cells: the increases in CFU level, even in the presence of the highest concentration (100 μM) of FC examined, were comparable to the increases in CFU level in the absence of FC (Fig. 1). Incidentally, when H. pylori cells were cultured for 24 h without either FC or 7dFC, the CFU levels increased from about 106.5 CFU/ml to about 108.5 CFU/ml (data not shown). Intriguingly, 7dFC exhibited obvious antimicrobial action against H. pylori cells at concentrations of 50 μM and 100 μM, and therefore, the CFU counts for H. pylori cultured in the presence of 7dFC at the 100 μM concentration were below the limits of detection. In sum, 7dFC was demonstrated to be a toxic compound fatal to the H. pylori cells.
Effect of dMβCD on the anti-H. pylori action of 7dFC.A recent study by our group has demonstrated that 2,6-di-O-methyl-β-cyclodextrin (dMβCD), a sterol (or steroid) solubilizer, inhibits the binding of the 3β-OH steroids pregnenolone and dehydroepiandrosterone to H. pylori cells, and therefore, the bacterial cells cannot absorb those steroids into the cell membrane (24). We next examined whether the anti-H. pylori action of 7dFC is inhibited by the action of dMβCD. When H. pylori cells were cultured for 24 h with 7dFC (100 μM) in the presence of dMβCD (15, 75, and 150 μM), the anti-H. pylori action of 7dFC was completely inhibited at the highest concentration (150 μM) of dMβCD examined (Fig. 2A). The CFU levels at that concentration were comparable to the CFU levels of H. pylori cultured without 7dFC. These results told us that the binding of 7dFC to H. pylori cells might be inhibited by the action of dMβCD, as in the cases of pregnenolone and dehydroepiandrosterone.
Cell membrane assimilation of 7dFC in H. pylori. (A) H. pylori cells were cultured in the dark for 24 h with 7dFC (100 μM) in the presence of dMβCD at the concentrations indicated in the graph, in PPLO broth (1.5 ml) under microaerobic conditions at 37°C. After culturing, the CFU counts were determined. The gray bar in the graph is the baseline CFU level measured immediately after the cultures were started. Results are indicated as the mean CFU ± SD obtained from the three strains of H. pylori (NCTC 11638, ATCC 43504, and 26695) and are representative of one of three independent experiments. (B) H. pylori (strain NCTC 11638) cells (109 CFU) were incubated in the dark for 4 h with a paper disk dotted with FC (100 nmol) or 7dFC (100 nmol) in the presence or absence of dMβCD (3 mM) in PPLO broth (5 ml) under microaerobic conditions at 37°C, and then the amounts of FC and 7dFC absorbed into the membrane lipid compositions of H. pylori cells (109 CFU) were quantified by the ferrous chloride-sulfuric acid method. Results are indicated as the mean level of sterol (FC or 7dFC) ± SD obtained from three independent experiments. (C) After the same experiment was carried out as described for panel B, the membrane lipids were extracted from the H. pylori cells (109 CFU). The glucosylated FCs (CGL, CAG, and CPG) and glucosylated 7dFCs (7dCGL, 7dCAG, and 7dCPG) in the lipid compositions were then analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, followed by visualization of the spots of those glucosyl cholesterols on a TLC plate surface treated with an orcinol-sulfuric acid reagent. (D) Paper disks dotted with the PE of H. pylori (strain NCTC 11638) at the amounts indicated in the graph were incubated in the dark for 4 h with an FC (100 nmol)-fixed paper disk or a 7dFC (100 nmol)-fixed paper disk in Tris buffer (2 ml) containing dMβCD (3 mM) at 25°C and then washed six times with distilled water. The amounts of FC and 7dFC in the H. pylori PE-fixed paper disks were then quantified by the ferrous chloride-sulfuric acid method. Results are indicated as the mean concentration of sterol (FC or 7dFC) ± SD obtained from three independent experiments.
Glucosylation of 7dFC by H. pylori cells.Our recent study has also demonstrated that the cell membrane of H. pylori promotes the absorption of FC via the activity of dMβCD (24). Next, we examined the effect of dMβCD on the incorporation of 7dFC into the cell membrane of H. pylori. We quantified and compared the amounts of FC and 7dFC incorporated by the H. pylori cells (109 CFU), in the presence or absence of dMβCD (3 mM), via the ferrous chloride-sulfuric acid method (Fig. 2B). When H. pylori cells were incubated for 4 h with a paper disk dotted with FC (100 nmol) or 7dFC (100 nmol) in the absence of dMβCD, the membranes of H. pylori cells absorbed both FC and 7dFC at negligible amounts (<10 nmol). This means that FC and 7dFC are hardly eluted from the paper disks in the absence of dMβCD and that the cell membrane of H. pylori cannot directly absorb those sterols fixed to the paper disks. The preliminary experiments in this study confirmed that the amounts of FC and 7dFC eluted from the paper disks dotted with 100 nmol of the sterols incubated in the absence of dMβCD were approximately 2 nmol and 15 nmol, respectively, while both FC (100 nmol) and 7dFC (100 nmol) were almost completely eluted from the same paper disks incubated in the presence of dMβCD (3 mM) (data not shown). On this basis, we next examined the absorption of FC and 7dFC by H. pylori cells in the presence of dMβCD (3 mM). The FC amount absorbed into H. pylori cells in the presence of dMβCD increased approximately 10-fold in comparison with that absorbed into the cells in the absence of dMβCD. Surprisingly, dMβCD also induced conspicuous absorption of 7dFC into the H. pylori cells: an approximately 4-fold-larger amount (>20 nmol) of 7dFC was detected in the H. pylori cells in the presence of dMβCD (3 mM) than in the absence of dMβCD. In sum, dMβCD had no inhibitory effect on the binding of 7dFC to H. pylori cells, differing from the cases of pregnenolone and dehydroepiandrosterone (24). These results suggested that H. pylori cells somehow neutralize the toxic activity of 7dFC absorbed into the cell membrane and that the detoxification of 7dFC by H. pylori cells is promoted via certain functions of dMβCD.
Next, we performed TLC analysis to confirm the biosynthesis of glucosylated cholesterols in H. pylori cells (109 CFU) incubated for 4 h with a paper disk dotted with an FC (100 μM) or 7dFC (100 μM) in the presence or absence of dMβCD (3 mM). Three spots of cholesteryl glucosides (CGL, CAG, and CPG) were detected in H. pylori cells incubated with the FC-fixed paper disk in the presence of dMβCD but were undetectable in cells incubated with the same paper disk in the absence of dMβCD (Fig. 2C). Intriguingly, H. pylori cells induced the glucosylation of 7dFC, and three spots of 7-dehydrocholesteryl glucosides, namely, 7dCGL, 7-dehydrocholesteryl-6-O-acyl-α-d-glucopyranoside (7dCAG), and 7-dehydrocholesteryl-6-O-phosphatidyl-α-d-glucopyranoside (7dCPG), were detected in the membrane lipid compositions of cells incubated with the 7dFC-fixed paper disk in the presence of dMβCD but not in the absence of dMβCD. These results tell us that H. pylori cells glucosylate 7dFC as they do FC and assimilate the glucosylated 7dFC into the cell membrane, even though this sterol is a toxic compound that impairs the viability of H. pylori.
Binding of 7dFC to H. pylori PE.Our recent study has revealed that phosphatidylethanolamine (PE) of H. pylori is involved in the incorporation of FC into the cell membrane and suggested that the myristic acid (C14:0) molecule, the most prevalent saturated fatty acid component of H. pylori PE, plays an important role in the selective binding of the nonesterified sterol but not of the esterified sterol (24). We therefore examined whether the H. pylori PE binds 7dFC as efficiently as it binds FC. Paper disks dotted with H. pylori PE were incubated for 4 h with a 7dFC (100 nmol)-fixed paper disk in the presence of dMβCD (3 mM), and then the 7dFC in the PE-fixed paper disks was quantified by the ferrous chloride-sulfuric acid method. The amount of 7dFC detected in the paper disks increased linearly along with the H. pylori PE dotted in larger amounts onto the paper disks, although those amounts were somewhat smaller than the amounts of FC detected in the H. pylori PE-fixed paper disks (Fig. 2D). Incidentally, the amounts of both FC and 7dFC in the paper disks without H. pylori PE were below the limits of detection (data not shown). In sum, the PE was also demonstrated to contribute to the incorporation of 7dFC into the H. pylori cell membrane.
Role of dMβCD in inhibiting toxic activity of 7dFC against H. pylori.Next, we tried to clarify how dMβCD protects H. pylori cells from the toxic action of 7dFC. Our recent study has demonstrated that the cell membrane of H. pylori absorbs dMβCD molecules (24). Thus, we conducted the following experiments using H. pylori cells pretreated with dMβCD. After H. pylori cells were precultured in the presence or absence of dMβCD (1.5 mM) in PPLO broth (100 ml) and washed three times with PBS, the cells (109 CFU/ml) harvested were incubated for 4 h in fresh PPLO broth (10 ml) that directly dispersed 7dFC (30 μM) without using a paper disk to quantify the amount of 7dFC absorbed into the cell membrane and to analyze the glucosylation of its toxic sterol. The amount of 7dFC absorbed into the membrane of the dMβCD-pretreated H. pylori cells was comparable to the amount of the sterol absorbed into the membrane of the H. pylori cells not treated with dMβCD (Fig. 3A).
Enhancement of glucosylation of 7dFC by H. pylori cells pretreated with dMβCD. (A) H. pylori (strain NCTC 11638) cells were precultured for 24 h with or without dMβCD (1.5 mM) in PPLO broth (100 ml), washed three times with PBS, and harvested via centrifugation (10,000 × g, 5 min). The cells (109 CFU/ml) of H. pylori with or without dMβCD pretreatment were then incubated in the dark for 4 h with 7dFC (30 μM) in the absence of dMβCD in PPLO broth (10 ml) to extract the membrane lipids from the bacterial cells via an organic solvent distribution method. The amounts of 7dFC in the membrane lipid compositions of cells (109 CFU) were quantified by the ferrous chloride-sulfuric acid method. Results are indicated as the mean concentration of 7dFC ± SD obtained from three independent experiments. (B) The membrane lipids of H. pylori cells (approximately 1010 CFU) with or without dMβCD pretreatment were prepared via the same experimental procedures as described for panel A and analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, followed by visualization of the spots of lipids on a TLC plate surface treated with a 60% sulfuric acid solution. Lanes 1 and 2 indicate the lipid profiles of non-dMβCD-treated H. pylori cells and dMβCD-pretreated H. pylori cells, respectively. PG-CL, phosphatidylglycerol-cardiolipin.
Yet, the levels of glucosylation of 7dFC were different between the dMβCD-pretreated H. pylori cells and the non-dMβCD-treated H. pylori cells. The TLC analysis demonstrated that the spot of 7dCPG was denser in the dMβCD-pretreated H. pylori cell membrane than in the non-dMβCD-treated H. pylori cell membrane, although the spot of 7dCGL was detected at similar densities in the cell membrane lipid compositions of both the dMβCD-pretreated H. pylori and the non-dMβCD-treated H. pylori (Fig. 3B). Meanwhile, 7dCAG was detected at negligible levels in the membrane lipid compositions of H. pylori cells regardless of pretreatment with dMβCD. Incidentally, dMβCD was below the limits of detection in the TLC analysis, and therefore, we could not find out the spot of dMβCD in the membrane lipid compositions of H. pylori cells pretreated with its oligomer on the TLC plate surface (data not shown). Given that the level of total 7-dehydrocholesteryl glucosides (especially 7dCGL and 7dCPG) was higher in the dMβCD-pretreated H. pylori cells than in the H. pylori cells without dMβCD pretreatment, we can assume that dMβCD somehow promotes the glucosylation of cholesterols in H. pylori cells. Apart from this, H. pylori at high cell densities (109 CFU/ml) was found to glucosylate 7dFC regardless of the activity of dMβCD.
Inactivation of the anti-H. pylori activity of 7dFC by glucosylation.As shown in Fig. 1, 2, and 3, the membranes of H. pylori cells were capable of absorbing 7dFC as 7-dehydrocholesteryl glucosides, although the bacterium suffered from the toxic activity of 7dFC itself. From these results, we assumed that H. pylori cells detoxify the 7dFC by glucosylation. We therefore carried out the next experiments to examine whether 7dCGL affects the growth of H. pylori.
First, we isolated 7dCGL from the membrane lipid compositions of H. pylori cultured for 24 h with 7dFC (15 μg/ml) in the presence of 0.2% dMβCD via Iatrobead column chromatography and acetone precipitation, and we then confirmed the purity of 7dCGL by TLC analysis. 7dCGL was successfully isolated with high purity; only one spot of 7dCGL was detected at high density when 150 nmol of 7dCGL was applied and chromatographed (Fig. 4A). On this basis, the next experiment was conducted using this purified 7dCGL preparation.
Detoxification of 7dFC by glucosylation. (A) Purified 7dCGL was analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, followed by visualization of the spot of 7dCGL on a TLC plate surface treated with a 60% sulfuric acid solution. Lane 1 is the membrane lipid (400 μg/lane) profile of H. pylori (strain NCTC 11638) cells cultured for 24 h with 7dFC (15 μg/ml) in PPLO broth (30 ml) containing 0.2% dMβCD. Lanes 2 and 3 indicate the spots of purified 7dCGL (150 nmol/lane) and a reference 7dFC (150 nmol/lane), respectively. (B) H. pylori cells (106 CFU) were cultured in the dark for 24 h with either 7dFC (100 μM) or 7dCGL (100 μM) in the absence of dMβCD in PPLO broth (1.5 ml) under microaerobic conditions at 37°C, and then the CFU counts were determined. Results are indicated as the mean CFU ± SD obtained from the three H. pylori strains (NCTC 11638, ATCC 43504, and 26695) and are representative of one of three independent experiments.
H. pylori cells (106 CFU/ml) were cultured for 24 h in the presence of 7dFC (100 μM) or 7dCGL (100 μM) in a medium (1.5 ml) without dMβCD. The CFU levels of H. pylori cultured with 7dFC naturally fell below the limits of detection due to the toxic action of the sterol (Fig. 4B). However, 7dCGL exhibited no toxicity to H. pylori cells, and therefore, the CFU levels at that time were comparable to the CFU levels of the organisms cultured without either 7dFC or 7dCGL. In sum, 7dFC was demonstrated to lose its toxicity against H. pylori cells through glucosylation.
Toxicity of 7dFC to cgt gene mutant H. pylori cells.As shown in Fig. 4, the activity of 7dFC against H. pylori cells was inactivated by glucosylation of the sterol. Next, we examined whether an H. pylori mutant that lacks cholesterol-α-glucosyltransferase (CGT) activity showed higher susceptibility to the toxic action of 7dFC than wild-type H. pylori. We conducted the following experiments using an H. pylori mutant in which the cgt gene was disrupted via insertion of the cat gene cassette.
To ascertain the disappearance of the CGT activity, we compared the membrane lipid profiles between cells of wild-type H. pylori and the cgt gene mutant H. pylori cultured for 24 h in medium containing 5% horse serum by TLC analysis. The cell membrane of wild-type H. pylori contained three types of cholesteryl glucosides (CGL, CAG, and CPG). However, no cholesteryl glucosides were detected in the membrane lipid compositions of cells of cgt gene mutant H. pylori, although the accumulation of FC was observed in its mutant cell membrane (Fig. 5A). These results indicate that cgt gene mutant H. pylori cells can absorb cholesterol into the membrane but not glucosylate the absorbed cholesterol. On this basis, we used cells of this mutant in the next experiment.
Susceptibility of cgt gene mutant H. pylori cells to the toxicity of 7dFC. (A) Wild-type H. pylori (strain 26695) cells (WT) and cgt gene mutant cells (cgt-M) were cultured for 24 h in a brain heart infusion broth containing 5% horse serum under microaerobic conditions at 37°C. After culturing, the membrane lipids were isolated from the bacterial cells, and the aliquoted lipids (400 μg/lane) were analyzed by TLC with a chloroform-methanol-water (70:30:5) solvent system, followed by visualization of the spots of lipids on a TLC plate surface treated with a 60% sulfuric acid solution. (B) The WT cells or cgt-M cells were cultured in the dark for 24 h with 7dFC at the concentrations indicated in the graph, in the absence of dMβCD in PPLO broth (1.5 ml) under microaerobic conditions at 37°C, and then the CFU counts were determined. The gray bar in the graph is the baseline CFU level measured immediately after the cultures were started. Results are indicated as the mean CFU ± SD obtained from three independent experiments. (C) cgt-M cells (approximately 106 CFU/ml) were cultured in the dark for 24 h in the presence or absence of 7dFC (100 μM) in PPLO broth (1.5 ml) containing dMβCD (150 μM) under microaerobic conditions at 37°C, and then CFU were measured. Results are indicated as the mean CFU ± SD obtained from three independent experiments.
When wild-type H. pylori cells or cgt gene mutant H. pylori cells were cultured for 24 h with 7dFC at various concentrations, the susceptibility of the mutant cells to the toxic action of 7dFC was obviously higher than that of the wild-type cells: the CFU decline curve of the cgt gene mutant H. pylori cells incubated in the presence of 7dFC at concentrations from 10 μM to 30 μM descended more sharply than that of the wild-type H. pylori cells incubated in the presence of 7dFC at the same concentrations (Fig. 5B). These results, in combination with those in Fig. 4, indicate that H. pylori detoxifies 7dFC by glucosylation in order to add the toxic sterol as a cell membrane lipid component.
Protection of cgt gene mutant H. pylori cells from the toxicity of 7dFC by dMβCD.Next, we examined the effect of dMβCD on the bactericidal action of 7dFC on the cgt gene mutant H. pylori cells. When the cgt gene mutant H. pylori cells (106 CFU/ml) were cultured for 24 h in the presence of 7dFC (100 μM) in a medium (1.5 ml) containing dMβCD (150 μM), surprisingly, the CFU levels of the mutant cells cultured with 7dFC were comparable to those of the cells cultured without 7dFC (Fig. 5C). In sum, the inhibition of anti-H. pylori activity of 7dFC by dMβCD was observed even in the cgt gene mutant H. pylori cells, as with the wild-type H. pylori cells (Fig. 2A).
When the wild-type H. pylori cells and the cgt gene mutant H. pylori cells cultured for 24 h in the presence of both 7dFC (100 μM) and dMβCD (150 μM) were microscopically observed, conspicuous differences between those strains were noted. Tremendous aggregation of the cells was induced in the cgt gene mutant H. pylori but not in the wild-type H. pylori (Fig. 6). However, the aggregations were not observed in cells of the wild-type H. pylori and the cgt gene mutant H. pylori cultured with dMβCD alone (data not shown). These results, in combination with those in Fig. 5C, indicate that the cgt gene mutant H. pylori cells are protected from the bactericidal action of 7dFC by a certain mechanism of dMβCD itself that is distinct from the detoxification of 7dFC by glucosylation in H. pylori cells.
Aggregation of cgt gene mutant H. pylori cells in the presence of both 7dFC and dMβCD. After cells (approximately 106 CFU/ml) of wild-type H. pylori (WT) and cgt gene mutant H. pylori (cgt-M) were cultured for 24 h with 7dFC (100 μM) in PPLO broth (1.5 ml) containing dMβCD (150 μM), the cells were microscopically observed with a differential interference mode.
Bacteriolysis of H. pylori cells via binding of 7dFC.Next, we investigated the bactericidal mechanism of 7dFC against H. pylori cells. When the cgt gene mutant H. pylori and the wild-type H. pylori at high cell densities (108.5 CFU/ml) were incubated at various time points in the presence or absence of 7dFC (100 μM), the OD660 of the cell suspensions (1 ml) in both strains incubated without 7dFC increased as similar curves along the time axis, although the OD660s in the wild-type H. pylori cell suspension were somewhat lower than those in the cgt gene mutant H. pylori cell suspension (Fig. 7A). In contrast, the OD660 of the cgt gene mutant H. pylori cell suspension incubated with 7dFC underwent a decrease at 8 h of incubation, although the OD660 of this cell suspension incubated from 2 h to 4 h with this sterol increased. These results suggested that 7dFC induces the lysis of cgt gene mutant H. pylori cells. Meanwhile, the OD660 of the wild-type H. pylori cell suspension incubated with 7dFC increased as a gently sloping curve along the time axis, and no decrease of OD660 was observed in the incubation time from 2 h to 8 h. In sum, the wild-type H. pylori seemed to bear the toxic action of 7dFC, in contrast to the cgt gene mutant H. pylori. Though the OD660 in the cell suspensions of both the wild-type H. pylori and the cgt gene mutant H. pylori incubated in the presence of 7dFC was relatively higher than the OD660 in the cell suspensions of those strains incubated in the absence of the sterol at the starting point (0 h) of incubation, this is due to its sterol (100 μM) dispersed into the broth.
Bactericidal mechanism of 7dFC against H. pylori cells. (A) cgt gene mutant H. pylori cells (cgt-M; approximately 108.5 CFU) or wild-type H. pylori cells (WT; approximately 108.5 CFU) were incubated in the dark for the indicated times in the presence or absence of 7dFC (100 μM) in PPLO broth (1 ml) under microaerobic conditions at 37°C. After incubation, the bacterial cells were recovered via centrifugation and resuspended in saline (1 ml) to measure the OD660 of the cell suspension (200 μl). Results are indicated as the mean OD660 ± SD obtained from three independent experiments. (B) Culture supernatants (800 μl) from cgt-M or WT incubated under the same conditions as described for panel A were applied to an organic solvent distribution to isolate hydrophobic compounds. The hydrophobic compounds were then developed on a silica gel plate via TLC with a chloroform-methanol-water (70:30:5) solvent system. After TLC, the silica gel plate was sprayed with a ninhydrin reagent and heated at 120°C to detect PE on the plate surface.
To detect PE, the most prevalent cell membrane lipid, in the culture supernatant (800 μl) of the cgt gene mutant H. pylori (108.5 CFU/ml) incubated from 2 h to 8 h with or without 7dFC (100 μM), we performed TLC analysis. The levels of PE detected in the culture supernatant of the mutant cells incubated with 7dFC were consistently higher than those of PE detected in the culture supernatant of its cells incubated without the sterol (Fig. 7B). In sum, the most prevalent membrane lipid was conspicuously released from H. pylori cells via the action of 7dFC. These results, in combination with those in Fig. 7A, indicate that the binding of 7dFC to cgt gene mutant H. pylori cells promotes the instability of the cell membrane structure and leads to bacteriolysis of the cells. In contrast to the results for the cgt gene mutant strain, the levels of PE detected in the culture supernatant (800 μl) of the wild-type H. pylori cells (108.5 CFU/ml) incubated with 7dFC (100 μM) were negligible and rather lower than those of PE detected in the culture supernatant of wild-type cells incubated without the sterol. In addition, we confirmed that the wild-type H. pylori with crowded cells (108.5 CFU/ml) induces the glucosylation of 7dFC and exhibits lipid profiles of glucosyl 7dFC similar to those of glucosyl 7dFC shown in Fig. 3B in the TLC analysis (data not shown). These results indicate that the wild-type H. pylori at high cell densities evaded the bacteriolytic activity of 7dFC via the glucosylation of its sterol.
DISCUSSION
dMβCD is an annular structure formed by seven dimethylated d-glucose molecules, embeds various hydrophobic compounds inside its molecule, and solubilizes hydrophobic compounds in a water solvent (28, 29). In particular, FC is one of the most suitable “guest molecules” of dMβCD for forming an inclusion complex (30–34). A recent study by our group has revealed that dMβCD functions as an FC carrier molecule for H. pylori cells and mediates the absorption of FC by H. pylori cells (24). In addition, this study demonstrated that the absorption of 7dFC into the H. pylori cell membrane is also mediated by dMβCD. When H. pylori at low cell densities (106.5 CFU/ml) was cultured with 7dFC (100 μM) in the presence of dMβCD (150 μM), the 7dFC did not express its toxic activity against H. pylori and conversely was absorbed into the cell membrane of the organism as 7-dehydrocholesteryl glucosides (7dCGL, 7dCAG, and 7dCPG) that are detoxified forms of the sterol. Our recent study also revealed that the H. pylori cell membrane induces conspicuous absorption of dMβCD itself, differing from the Escherichia coli cell membrane (24). Moreover, this study demonstrated that H. pylori cells pretreated with dMβCD promote the glucosylation of 7dFC. These results suggest that 7dFC is promptly inactivated via the glucosylation before taking its toxic action against H. pylori cells when dMβCD molecules surround H. pylori cells and/or are absorbed into the cell membrane of the organism. To clarify the mechanisms by which dMβCD upregulates cholesterol glucosylation in H. pylori cells, we will need to conduct more experiments.
Our previous studies have investigated the influences of dMβCD on the interaction of H. pylori cells with various hydrophobic compounds: (i) this agent inhibits the binding of unsaturated fatty acids (C18:2 and C20:4) and lysophosphatidylcholine (LPC) to H. pylori cells and prevents the organisms from the bacteriolytic action of those fatty acids and LPC, (ii) this agent has no influence on the bacteriolytic action of phosphatidylcholine (PC) on the H. pylori cells without steroids and rather enhances the binding of PC to the cells, (iii) this agent inhibits the interaction between H. pylori cells and steroid hormones, and (iv) this agent mediates and enhances the absorption of FC into the cell membrane of H. pylori (13–15, 24). In sum, dMβCD has various functions on the contact of individual lipophilic compounds to the cell membrane of H. pylori. In this study, we demonstrated that dMβCD promotes the glucosylation of 7dFC in H. pylori cells. This may be a novel function of dMβCD in relation to H. pylori. In addition, we showed that dMβCD induces tremendous aggregations of cgt gene mutant H. pylori cells in the presence of 7dFC. This may suggest that the cgt gene mutant cells bind the 7dFC-dMβCD inclusion complexes via the mediation of dMβCD molecules, continue to retain its inclusion complexes because the cells cannot glucosylate the sterol, and thereby cause the cell aggregations via hydrophobic interaction between the 7dFC-dMβCD inclusion complexes bound to the cells. As the result of aggregation, the cgt gene mutant cells seemed to somehow evade the toxic action of 7dFC. In the future, more investigations will be necessary to elucidate the functions of dMβCD concerned with the mechanisms of the cgt gene mutant H. pylori cells for evasion of the toxicity of 7dFC.
The cell membrane of the cgt gene mutant H. pylori retained FC itself, even though the mutant cells could not glucosylate FC. A recent study by our group has revealed that PE of H. pylori is relatively rich in the outermost layer of the outer membrane, selectively binds FC but not the esterified cholesterol, and functions as a steroid-binding lipid in assimilating 3β-OH steroid compounds into the cell membrane (24). These results, in combination with this study, indicate that the cell membrane of H. pylori glucosylates 3β-OH steroid compounds via at least two pathways: one is the direct binding of 3β-OH steroid compounds to the cgt gene product, followed by glucosylation, and one is the transport of 3β-OH steroid compounds to the cgt gene product after the binding of 3β-OH steroid compounds to the PE for glucosylation. However, it is unclear whether H. pylori cells possess such a transport molecule that carries the 3β-OH steroid compounds from the PE molecules to the cgt gene product. Further investigations will therefore be necessary to examine the membrane transport system for 3β-OH steroid compounds in H. pylori.
The steroid compounds that are glucosylated by H. pylori cells are strictly limited to the 3β-OH steroid compounds (13). H. pylori cells, however, aggressively interact with various steroid compounds, even though some of them injure this bacterium (13, 14). Some steroid compounds, such as FC and estrone, serve to strengthen the membrane lipid barrier of H. pylori cells, whereas some steroid compounds, such as progesterones and bile salts, impair the viability of H. pylori cells (14–16). This means that H. pylori cells easily ingest nonesterified steroid compounds regardless of toxicity. Up to this time, the 3β-OH steroid compounds toxic to the survival of H. pylori, however, had yet to be identified. This study demonstrated that the membrane of H. pylori cells incorporates 7dFC, a 3β-OH sterol, after detoxification of this sterol via glucosylation. Given that a number of 3β-OH sterols, such as ergosterol, sitosterol, campesterol, and stigmasterol, are in a wide range of organisms from fungi to mammals and plants, we can easily assume that some of them may be toxic to H. pylori, as with 7dFC. Therefore, H. pylori needs to inactivate such toxic 3β-OH sterols in order to survive in vivo or in vitro. As a means to do this, the glucosylation of 3β-OH steroid compounds is considered to be an essential ability in H. pylori cells.
FOOTNOTES
- Received 16 August 2012.
- Accepted 5 November 2012.
- Accepted manuscript posted online 9 November 2012.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
