Taurine modulates host cell responses to Helicobacter pylori VacA toxin

Abstract

Colonization of the human stomach with Helicobacter pylori strains producing active forms of the secreted toxin VacA is associated with an increased risk of peptic ulcer disease and gastric cancer, compared with colonization with strains producing hypoactive forms of VacA. Previous studies have shown that active s1m1 forms of VacA cause cell vacuolation and mitochondrial dysfunction. In this study, we sought to define the cellular metabolic consequences of VacA intoxication. Untargeted metabolomic analyses revealed that several hundred metabolites were significantly altered in VacA-treated gastroduodenal cells (AGS and AZ-521) compared with control cells. Pathway analysis suggested that VacA caused alterations in taurine and hypotaurine metabolism. Treatment of cells with the purified active s1m1 form of VacA, but not hypoactive s2m1 or Δ6-27 VacA-mutant proteins (defective in membrane channel formation), caused reductions in intracellular taurine and hypotaurine concentrations. Supplementation of the tissue culture medium with taurine or hypotaurine protected AZ-521 cells against VacA-induced cell death. Untargeted global metabolomics of VacA-treated AZ-521 cells or AGS cells in the presence or absence of extracellular taurine showed that taurine was the main intracellular metabolite significantly altered by extracellular taurine supplementation. These results indicate that VacA causes alterations in cellular taurine metabolism and that repletion of taurine is sufficient to attenuate VacA-induced cell death. We discuss these results in the context of previous literature showing the important role of taurine in cell physiology and the pathophysiology or treatment of multiple pathologic conditions, including gastric ulcers, cardiovascular disease, malignancy, inflammatory diseases, and other aging-related disorders.

Keywords: bacterial protein toxin; cell death; gastric cancer; metabolomics; peptic ulcer disease.

Fig 1

VacA-treated AGS cells have a significantly different metabolic profile than control cells. AGS cells were treated with acidified VacA (20 µg/mL) or acidified buffer for 8 hours, followed by untargeted metabolomic analysis of the cells. Global principal component analysis (PCA) plots and volcano plots are shown. The volcano plots illustrate features exhibiting a log₂(fold-change) ≥│1│ (fold change compares VacA-treated cells with control-treated cells) and a −log₁₀(P-value) > 1.3 based on ANOVA analysis. Metabolites shown in blue are decreased in abundance in cells treated with VacA compared with buffer-treated control cells, and metabolites shown in red are increased in abundance in cells treated with VacA compared with buffer-treated control cells. (A, C) Analysis using reverse phase liquid chromatography (RPLC) positive ion mode high-resolution tandem mass spectrometry (RPLC-HRMS/MS). (B, D) Analysis using hydrophilic interaction liquid chromatography (HILIC) negative ion mode mass spectrometry (HILIC-HRMS/MS).

Fig 2

Taurine and hypotaurine metabolism are significantly altered in response to VacA treatment. AGS cells were treated with acidified VacA (20 µg/mL) or acidified buffer for 8 hours. (A) Pathway analysis of metabolomic results, comparing combined RPLC- and HILIC-HRMS/MS data from VacA-treated cells with data from buffer-treated cells (corresponding to data in Table S3.1). Enriched pathways identified in the analysis are arranged by log₁₀(P-value) determined by a hypergeometric test on the Y-axis, and pathway impact values from pathway topology analysis are shown on the X-axis. The node color is based on its P-value (darker red corresponds to a smaller P-value), and the node radius is determined based on the pathway impact values (larger nodes correspond to greater pathway impact), as described in Materials and Methods. (B) Pathway of taurine and hypotaurine metabolism, showing metabolites that were detected in RPLC- and HILIC-HRMS/MS analyses. Gray boxes correspond to metabolites not confidently identified in the analysis. ****P < 0.0001, ***P < 0.001, **P < 0.01 based on ANOVA. Arrows indicate the directionality of the change in VacA-treated cells relative to buffer-treated control cells at the 8-hour time point. (C) Fold change in abundance of the indicated metabolites over time following VacA treatment. Fold change values represent a comparison of cells treated with VacA (20 µg/mL) to control cells treated with acidified buffer.

Fig 3

AGS cells treated with wild-type VacA have significantly different metabolic profiles than cells treated with mutant forms of VacA. AGS cells were treated with wild-type s1m1 VacA, VacA Δ6–27, the s2m1 mutant form of VacA (each 20 µg/mL), or acidified buffer for 8 hours, followed by untargeted metabolomic analysis using RPLC-HRMS/MS. (A) Global principal component analysis (PCA) shows that the results of AGS cells treated with WT VacA cluster separately from the results of the other groups. (B) Heat map clustering of results. Samples (columns) are clustered by relative feature abundance (rows) ranging from low (blue) to high (red) abundance. AGS cells treated with WT VacA toxin (20 µg/mL) for 8 hours have significantly lower levels of taurine (C) and hypotaurine (D) compared with control-treated cells. There is no significant difference in taurine (C) or hypotaurine (D) abundance in AGS cells treated with VacAΔ6–27 or s2m1 VacA compared with control cells. ****P < 0.0001, **P < 0.0005 determined by ordinary one-way ANOVA with Dunnett’s multiple comparison test. ns, not significant.

Fig 4

VacA-treated AZ-521 cells have significantly different metabolic profiles compared with control cells. AZ-521 cells and AGS cells were treated with WT VacA (20 µg/mL) or acidified buffer for 3 or 5 hours, followed by untargeted metabolomic analysis using RPLC-HRMS/MS. (A) Global principal component analysis (PCA) of results from AZ-521 cells treated with acidified WT VacA or acidified buffer for 3 or 5 hours. (B) Heat map clustering of results for AZ-521 cells. Samples (columns) are clustered by group and relative feature abundance (rows) ranging from low (blue) to high (red) abundance. (C) Global principal component analysis (PCA) of results from AGS cells treated with WT toxin or acidified-buffer control for 3 or 5 hours. (D) Pathway of taurine and hypotaurine metabolism, indicating the compounds in the pathway that exhibited a P-value < 0.05 based on ANOVA analysis when comparing both AGS and AZ-521 cells treated with VacA for 3 hours with corresponding control cells. Gray boxes correspond to metabolites not confidently detected in the analysis.

Fig 5

Supplemental taurine has a protective effect against VacA-induced cell death. AZ-521 cells were treated with VacA (20 µg/mL) or acidified buffer for 3 hours. (A) Relative taurine abundance in AZ-521 cells treated with VacA or buffer. ****P < 0.0001 determined by unpaired t-test. (B, C, D) VacA (20 µg/mL) was acid-activated and added to AZ-521 cells in the presence of varying concentrations of extracellular taurine, hypotaurine, or cysteine. The CellTiter Blue assay was used to assess cell viability after 24 hours. Fluorescence (excitation 560 nm, emission 590 nm) was quantified after 24 hours of treatment with VacA as a readout for cell viability. Data in panel B were analyzed by unpaired t-test, and data in panels C and D were analyzed by ordinary one-way ANOVA with Dunnett’s multiple comparison test. ****P < 0.0001, ***P < 0.005, *P < 0.05.

Fig 6

Extracellular taurine has a protective effect against VacA-induced cell death. VacA proteins were acid-activated and added (at 20 µg/mL) to AZ-521 cells in the presence of varying concentrations of extracellular taurine. Lactate dehydrogenase (LDH) release was measured to assess cell death induced by VacA. LDH release was quantified after 24 hours of treatment with VacA [quantified by measuring optical density at 490 nm (OD490)]. High levels of extracellular taurine have a protective effect against VacA-induced cell death. ****P < 0.0001, determined by unpaired t-test.

Fig 7

Effect of supplemental taurine on cellular metabolic profiles in VacA-treated cells. Untargeted metabolomics analysis shows that increased intracellular taurine levels are sufficient for modulating VacA-induced cell death. AGS or AZ-521 cells were treated with VacA (20 µg/mL) for 12 hours (AGS cells) or 5 hours (AZ-521 cells) in the presence or absence of supplemental taurine (10 mM) followed by untargeted metabolomics analysis of cells. (A, B) The volcano plots illustrate features exhibiting a log₂(fold-change) ≥│1│ (fold change compares VacA-treated cells in the presence of 10 mM extracellular taurine with VacA-treated cells in the absence of 10 mM extracellular taurine) and a −log₁₀(P-value) > 1.3 based on ANOVA analysis. Volcano plots show (A) AGS cells after treatment with 20 µg/mL of acid-activated VacA for 12 hours and (B) AZ-521 cells after treatment with 20 µg/mL of acid-activated VacA for 5 hours.