Helicobacter pylori initiates successful gastric colonization by utilizing L-lactate to promote complement resistance

Abstract

The complement system has long been appreciated for its role in bloodborne infections, but its activities in other places, including the gastrointestinal tract, remain elusive. Here, we report that complement restricts gastric infection by the pathogen Helicobacter pylori. This bacterium colonized complement-deficient mice to higher levels than wild-type counterparts, particularly in the gastric corpus region. H. pylori uses uptake of the host molecule L-lactate to create a complement-resistant state that relies on blocking the deposition of the active complement C4b component on H. pylori's surface. H. pylori mutants unable to achieve this complement-resistant state have a significant mouse colonization defect that is largely corrected by mutational removal of complement. This work highlights a previously unknown role for complement in the stomach, and has revealed an unrecognized mechanism for microbial-derived complement resistance.

Fig. 1. Complement plays a vital role in restricting H. pylori gastric colonization.

a Schematic diagram of complement pathways activation. b Schematic illustration of mouse infection analysis. C57BL/6J WT (n = 6) or C3^−/− (n = 7) mice were infected by oral gavage with GFP⁺ H. pylori PMSS1 for two weeks. Stomachs were isolated from infected mice and separated into corpus and antrum pieces and processed to obtain the total bacterial numbers (colony forming units, CFUs) by plating, to enumerate bacteria in glands using microscopy. Bacteria were visualized by GFP expression (green), while gland cells were visualized using a DNA Hoechst stain (blue). c Total bacterial number from gastric antrum (gray) and corpus (red) of infected WT or C3^−/− mice normalized to tissue weight (CFUs/g). Gray dotted line, limit of detection. d Percentage of glands occupied by H. pylori in the gastric antrum (gray) and corpus (red) of infected WT or C3^−/− mice, determined by analyzing 100 randomly selected isolated glands of each infected mouse. e Number of H. pylori per colonized gland in the gastric antrum (gray) and corpus (red) of infected WT or C3^−/− mice. f Gastric pH levels in either antrum or corpus region of both WT and C3^−/− mice, in H. pylori free or infected condition. g, h Gastric L-lactate levels were measured using samples for bacterial plating. For all dot plots, each dot represents one mouse sample. For all panels, black bars represent mean of samples. For plot d and e, the box shows the 10^th to 90^th percentiles of the dataset; the black whiskers represent the maxima and minima of the dataset. The p values were obtained by one-way ANOVA with Tukey’s multiple comparisons test (c, d, e, f), and unpaired two-tailed t-tests (g, h). The significance is indicated as * (p < 0.05); ** (p < 0.01). Source data and exact p values are provided in the Source Data file.

Fig. 2. H. pylori is sensitive to complement-mediated elimination.

Overnight cultures of H. pylori PMSS1 WT were treated with inactive and active NHS or mouse serum at the indicated final percentages for 1 h, and then viable bacterial CFUs were determined by plating. The survival percent was determined by comparing the number of CFUs from active serum divided by the total CFUs, obtained from heat-inactivated serum treatment. a Comparison of different final percentages of NHS at pH of 7. b Comparison of different final percentages of NHS at pH of 3.5. Survival ratio without NHS treatment was acquired by comparing the bacterial number at the end time point to the input number. c H. pylori cultures were treated by untreated NHS (gray bar, all pathways on); EGTA-conditioned NHS to block all complement pathways (white bar, all pathways off); EGTA + Ca²⁺ (blue bar, classical and lectin pathway on); EGTA + Mg²⁺ (green bar, alternative pathway on); C1q-depleted human serum (yellow bar, alternative and lectin pathway on); C2-depleted human serum (light orange bar, alternative pathway on only); or Factor-B-depleted human serum (light red bar, classical and lectin pathway on). All tests were carried out with 10% serum. d H. pylori cultures were treated by 10% final percentage of extracted mouse serum from C57BL/6J WT or C3^−/− mice. In all panels, results were presented as survival percent $\pm$ standard deviation (SD), and were derived from at least triplicate biological samples with a triplicate of each given sample. The p values were obtained by one-way ANOVA with Tukey’s multiple comparisons test (a, b, c), and a two-tail student t-test (d). The significance was indicated as * (p < 0.05), ** (p < 0.01), *** (p < 0.001), **** (p < 0.0001), or n.s. (not significant). Source data and exact p values are provided in the Source Data file.

Fig. 3. L-lactate protects H. pylori from complement-mediated elimination.

Overnight cultures of H. pylori PMSS1 WT or ∆lctP mutant with or without L-lactate supplementation overnight or for 1 h, as indicated, were treated by inactive and active serum for 1 h, and then viable bacterial CFUs were determined by plating. The survival percent was determined as described in Fig. 2. a Overnight cultures of H. pylori WT or ΔlctP mutant grown with the indicated concentration of L-lactate were treated by inactive and active 10% NHS. b Overnight cultures of WT H. pylori with 5mM L-lactate supplementation was treated by indicated percentage of NHS. c WT H. pylori cultures were treated with C57BL/6J mouse serum. d H. pylori cultures were treated by NHS only (gray bar), EGTA-conditioned NHS (white bar), C1q-depleted human serum (yellow bar), C2-depleted human serum (orange bar), Factor-B-depleted human serum (red bar), EGTA + Ca²⁺ (blue bar), or EGTA + Mg²⁺ (green bar) as in Fig. 2c. Survival ratio in the absence of L-lactate is identical to Fig. 2c. All tests were applied with 10% serum. e WT H. pylori cultures were treated with L-lactate for 1 h before exposed to NHS. In all panels, results were presented as survival percent $\pm$ standard deviation (SD), and were derived from triplicate biological samples with a triplicate of each given sample. The p values were obtained by one-way ANOVA with Tukey’s multiple comparisons test (a, b, d), and a two-tail student t-test (c, e). The significance is indicated as * (p < 0.05), ** (p < 0.01), *** (p < 0.001), **** (p < 0.0001), or n.s. (not significant). Source data and exact p values are provided in the Source Data file.

Fig. 4. L-lactate mediates destabilization of C4b binding to H. pylori.

a Schematic diagram of C4 primary activation and secondary degradation, adapted from. b Western blot of C4 protein and products after activation in the presence of WT and ∆lctP H. pylori over 60 min, using non-reducing gels. A positive control of complement activation and C4c formation was triggered by heated human IgG. Positions of C4, C4b, and C4c are shown at the left, and molecular weight is indicated on the right. The result is representative of three independent experiments with triplicate biological samples. c H. pylori PMSS1 WT was grown ± L-lactate and then treated by 10% NHS for the indicated time periods. Each sample was split into three: (1) treated by sample buffer only (total C4b); (2) centrifuged at 3000 × g for 3 min before adding sample buffer (bound C4b); (3) treated by sample buffer plus the reducing agent beta-mercaptoethanol (BME) (C4 β-chain). All protein samples were run on 5-7% SDS PAGE and blotted with anti-C4 antibody. d Quantification of the experiment shown in (c), from three independent experiments. Bands were quantified by normalizing each band to the amount of total C4b or bound C4b at 3 min (leftmost band in each treatment group) using the Bio-Rad Image Lab software. Each point represents a biological sample for an independent experiment. The data were combined from the three independent experiments (represented as mean ± SD). The p values were obtained from unpaired two-tailed t-tests. The significance is indicated as * (p < 0.05), *** (p < 0.001), **** (p < 0.0001), or n.s. (not significant). Source data and exact p-values are provided in the Source Data file.

Fig. 5. Loss of lctP renders H. pylori colonization deficient in a manner that is recovered by loss of complement.

a Two-weeks infection by H. pylori strains. Group of C57BL/6J WT (n = 8) and isogenic C3^−/− (n = 8) mice were infected with the H. pylori PMSS1 WT; another group of C57BL/6J WT (n = 6) and isogenic C3^−/− (n = 6) mice were infected with the ΔlctP strain. b One-year infection by H. pylori strains. Group of C57BL/6J WT (n = 4) and isogenic C3^−/− (n = 3) mice were infected with the H. pylori PMSS1 WT; another group of C57BL/6J WT (n = 4) and isogenic C3^−/− (n = 4) mice were infected with the ΔlctP strain. The WT and C3^−/− mice were cohoused at weaning for at least 4-weeks before the infection and remained cohoused for the whole experimental period. At the indicated time points, the stomachs were collected and bacterial CFUs were determined by plating. Each point represents one mouse sample, with black bars representing the mean. Gray dotted line represents the limit of detection. The p values were obtained by Tukey’s multiple comparisons test. The significance was indicated as * (p < 0.05), ** (p < 0.01), and n.s. (no significance). Source data and exact p values are provided in the Source Data file.

Fig. 6. Proposed model for L-lactate promotion of H. pylori gastric colonization.

Model showing how complement may operate in the stomach to inhibit H. pylori colonization. The left panel shows WT H. pylori, which is able to colonize the gastric mucosal surface and glands in the face of complement (red circles). The inset details how L-lactate (blue circles) uptake into H. pylori blocks the step of C4b surface deposition, rendering H. pylori complement resistant. This mechanism of complement resistance is distinct from that used by other bacteria, acting to further block downstream C3 convertase formation and other activities. The right gray panel shows what happen when H. pylori is unable to take up L-lactate. In this case, the H. pylori is significantly less capable of destabilizing surface bound C4b, rendering the strain susceptible to complement and less able to colonize the gastric glands and surface.