Genetic requirements and transcriptomics of Helicobacter pylori biofilm formation on abiotic and biotic surfaces

Abstract

Biofilm growth is a widespread mechanism that protects bacteria against harsh environments, antimicrobials, and immune responses. These types of conditions challenge chronic colonizers such as Helicobacter pylori but it is not fully understood how H. pylori biofilm growth is defined and its impact on H. pylori survival. To provide insights into H. pylori biofilm growth properties, we characterized biofilm formation on abiotic and biotic surfaces, identified genes required for biofilm formation, and defined the biofilm-associated gene expression of the laboratory model H. pylori strain G27. We report that H. pylori G27 forms biofilms with a high biomass and complex flagella-filled 3D structures on both plastic and gastric epithelial cells. Using a screen for biofilm-defective mutants and transcriptomics, we discovered that biofilm cells demonstrated lower transcripts for TCA cycle enzymes but higher ones for flagellar formation, two type four secretion systems, hydrogenase, and acetone metabolism. We confirmed that biofilm formation requires flagella, hydrogenase, and acetone metabolism on both abiotic and biotic surfaces. Altogether, these data suggest that H. pylori is capable of adjusting its phenotype when grown as biofilm, changing its metabolism, and re-shaping flagella, typically locomotion organelles, into adhesive structures.

Fig. 1. Specific H. pylori strains form biofilms in rich media.

Biofilm formation of H. pylori strains was assessed using the microtiter plate crystal violet biofilm assay. Strains were grown for three days in BB10, with no shaking, under 10% CO₂, 5% O₂ and 85% N₂. Results represent the crystal violet absorbance at 595 nm, which reflects the biofilm biomass. Experiments were performed three independent times with at least six technical replicates for each. Error bars represent standard error of the mean.

Fig. 2. H. pylori G27 biofilm characterization.

H. pylori G27 was allowed to form biofilms for three days on plastic as in Fig. 1. A Confocal scanning laser microscopy (CSLM) images (top view) of 3-day old H. pylori G27 biofilms stained with LIVE (green)/DEAD (red) stain. B SEM images show cell-to-cell interactions through flagellar filaments and pili-like structures. High magnification in d is derived from the red-boxed area of c. Black arrows indicate flagella filaments and white arrows indicate the presence of pili-like structures connecting cells togethe. SEM micrographs of H. pylori biofilm at 5000x (a) and 12500x (b and c).

Fig. 3. H. pylori G27 biofilms formed on AGS cells.

SEM images showing 3-day old H. pylori G27 wild-type (WT) biofilms formed on AGS cells. Magnified images highlight bacteria-to-bacteria and bacteria-to-cell interactions through flagella.

Fig. 4. H. pylori flagella is essential for initial attachment and biofilm on AGS cells.

Attachment and biofilm formation of H. pylori G27 wild type (WT), mutant fliA (aflagellated and non-motile), and mutant motB (flagellated and non-motile) was analyzed by counting colony-forming unit (CFU/ml). A H. pylori WT and mutants initial attachment on AGS cells were measured after 1 h infection, with (solid bars) or without (striped bars) centrifugation. B H. pylori cells on the surface were measured after 72 h to monitor biofilm formation using CFUs. Experiments were performed three independent times with at least three technical replicates for each. Error bars represent standard error of the mean. Statistical analyses were performed using one-way ANOVA with Tukey post hoc test (*P < 0.01; **P < 0.001; ***P < 0.0001; n.s., no significance).

Fig. 5. Biofilm formation by H. pylori G27 WT and biofilm-defective mutants.

Biofilm formation was assessed using the microtiter plate biofilm assay and confocal laser-scanning microscopy over three days. Results represent the crystal violet absorbance at 595 nm, which reflect the biofilm biomass. Experiments were performed three independent times with at least three technical replicates for each. Error bars represent standard errors for each average value. CLSM micrographs compare biofilm formation of the wild-type G27 strain and biofilm-defective mutants. Biofilms were stained with FM 1–43, which become fluorescent once inserted in the cell membrane. HP hypothetical protein, WT Wild-type. Statistical analyses were performed using ANOVA (*P < 0.05).

Fig. 6. Differential gene expression in H. pylori biofilm-grown cells as opposed to planktonic grown cells.

H. pylori strain G27 was grown as a biofilm or planktonic culture for 3 days, after which RNA was collected and sequenced. A Principal component analysis (PCA) of gene expression obtained by RNA-seq between biofilm growing cells and planktonic ones, with three biological replicates and two technical replicates each for n = 6. B Volcano plot of gene expression data. The y-axis is the negative log10 of P-values (a higher value indicates greater significance) and the x-axis is log2 fold-change or the difference in abundance between two population (positive values represent the upregulated genes in biofilm and negative values represent downregulated genes). The dashed red line shows where P = 0.01, with points above the line having P < 0.01 and points below the line having P > 0.01.

Fig. 7. Biofilm formation by H. pylori G27 WT and its isogenic mutants.

Biofilm formation was assessed using the microtiter plate crystal violet biofilm assay after 3 days of growth. Experiments were performed two independent times with at least six technical replicates for each. Error bars represent standard errors for each average value. Statistical analyses were performed using ANOVA (*P < 0.01; **P < 0.0001).

Fig. 8. Biofilm formation by H. pylori G27 WT and its isogenic mutants on AGS cells surface.

Biofilm formation was performed in 24-well plates with 85% AGS cells confluency. Bacterial numbers were determined by plating (CFU/ml) after 1 h (A) and three days (B) of coincubation. Experiments were performed three independent times with triplicates each time. Error bars represent standard deviations for each average value. Statistical analyses were performed using one-way ANOVA with Tukey post hoc test (*P < 0.05, ****P < 0.0001).