Molecular Hydrogen as an Energy Source for Helicobacter pylori

See allHide authors and affiliations

Science  29 Nov 2002:
Vol. 298, Issue 5599, pp. 1788-1790
DOI: 10.1126/science.1077123


The gastric pathogen Helicobacter pylori is known to be able to use molecular hydrogen as a respiratory substrate when grown in the laboratory. We found that hydrogen is available in the gastric mucosa of mice and that its use greatly increased the stomach colonization by H. pylori. Hydrogenase activity inH. pylori is constitutive but increased fivefold upon incubation with hydrogen. Hydrogen concentrations measured in the stomachs of live mice were found to be 10 to 50 times as high as theH. pylori affinity for hydrogen. A hydrogenase mutant strain is much less efficient in its colonization of mice. Therefore, hydrogen present in animals as a consequence of normal colonic flora is an energy-yielding substrate that can facilitate the maintenance of a pathogenic bacterium.

The bacterial oxidation of molecular H2 commonly occurs in nature, as hydrogen gas released by other bacteria represents a useable high-energy reductant (1). Once H2 is bound and “split” by a membrane-associated hydrogenase, further oxidation-reduction and energy-generating steps are facilitated by a series of membrane-bound heme-containing electron carriers. Hydrogen is a by-product of colonic fermentation (2), and hydrogen has been reported to be produced (measured as excreted gas) in the gastrointestinal tract of both rodents (3) and humans (4). However, whether molecular hydrogen is used as an energy reservoir for pathogenic bacteria residing in animals is not known. To help understand the microbial communities associated with digestion, H2 levels were determined in the termite hind-gut (5) and recently from the cockroach midgut (6), but H2 levels in tissues of vertebrate animal hosts has not been assessed. Helicobacter pylori is a pathogen that solely colonizes the mucosal surfaces of the human stomach, where it gives rise to gastritis and peptic ulcers and is correlated with the development of certain types of gastric cancer (7). We previously reported that lab-grownH. pylori can express a membrane-bound “uptake-type” hydrogenase (8). H2 use by H. pyloriwas accompanied by changes to other electron-carrying cell proteins that are related to energy-producing processes within cells to carry out a myriad of cell-building functions. Here we show that the mucous lining of the stomach contains ample amounts of molecular H2. Combined with our measurements of the binding affinity of these bacteria for H2, we conclude that hydrogenase is saturated with H2 in the host tissues. A mutant H. pylori strain unable to oxidize hydrogen is severely impaired in its ability to colonize in mice. Therefore, H2 availability and its use as an energy source is a formerly unrecognized factor in understanding how a human pathogen grows and persists in an animal host.

One hallmark of the energy-conserving uptake hydrogenases is the ability to respond positively to exogenously supplied hydrogen (9, 10). Hydrogenase activity (11) inH. pylori is constitutive under all conditions we have tested, but in a chemically defined media (12) amperometrically determined hydrogenase activity (13) increases from a baseline value of 0.7 nmol H2oxidized/min/108 cells in cultures grown under micro-aerobic conditions (12% O2, 5% CO2, balance N2) to 3.1 nmol H2oxidized/min/108 cells when supplemented with 10% H2. A much milder stimulation of hydrogenase activity occurs when the cultures are grown in rich media or on blood-containing plates (BA plates) (13), in which hydrogenase activity is stimulated approximately twofold by the addition of 10% hydrogen (14). To characterize hydrogenase regulation, we used promoter fusions with the reporter gene xylE(15) from Pseudomonas putida to generate catechol 2,3-dioxygenase, which can be easily assayed spectrophotometrically (13). We assayed XylE activity in H. pyloristrains carrying plasmids with hydrogenase structural gene promoter-xylE fusion (phyd:xylE), a nonhydrogenase related promoter-xylE fusion (pHP0630:xylE), and a promoterless xylEgene (pHel:xylE). The results (Table 1) show that hydrogenase is regulated at the transcriptional level. The gene directly adjacent to hydrogenase (designated HP0630 and annotated as conserved; no known function in the sequenced strain 26695) (16) is not regulated by hydrogen, and no XylE activity was seen in the strain harboring a promoterless xylE gene (Table 1). Hydrogenase transcription was not affected by other environmental conditions such as pH or oxygen concentration (14), and proper regulation of the hydrogenase operon (as measured by phyd:xylE) is retained in the hydrogenase structural gene mutant Hyd:cm (14), indicating that hydrogenase is not self-regulated. That the enzyme expression responds to molecular hydrogen availability supports our previous proposal (8) that the role of hydrogenase is in respiratory hydrogen oxidation.

Table 1

XylE activities [expressed as XylE units/108 cells (13)] of H. pyloriharboring xylE reporter plasmids grown under different growth conditions.

View this table:

A whole-cell Michaelis constant (apparentK M) for hydrogen was determined to be 1.8 μM, indicating a very high affinity for hydrogen, and a value similar to the whole-cell affinities of other hydrogen-oxidizing bacteria (17). The method used (17) to determine this K M uses live, intact cells with O2 available as the only terminal electron acceptor in the H2 oxidizing respiratory chain. Therefore, our measured apparent K M is for the entire hydrogen oxidizing system. We have previously shown that hydrogen oxidation in H. pylori grown in an H2-containing atmosphere is linked to cytochrome reduction, with the heme-containing components functioning as intermediate electron carriers before reduction of the terminal (O2-binding) oxidases (8).

Hydrogenase mutants in the SS1 (mouse colonization strain) background (SS1:Hyd) are deficient in their ability to colonize in mice. From two separate mouse colonization studies, only 24% (9 of 38) of the hydrogenase mutant-inoculated mice were colonized, as compared to 100% (37 of 37) colonization when inoculated with the parent strain (Fig. 1). The colonization efficiency of the mutant strain correlates with the inoculum dosage, with only 15% colonization at a inoculation dose of 2 × 108 (Fig. 1A) but 33% colonization at a inoculation dose of 1 × 109 (Fig 1B). SS1-inoculated mice were 100% colonized (the strain was able to colonize every mouse) at both inoculation doses when we used an initial “two-dose” regimen for inoculation (13).

Figure 1

Mouse colonization assay of H. pylori SS1 and Hyd:cm (SS1). Data are presented as a scatter plot of colony forming units per gram of stomach as determined by plate counts (13). Mice were considered positive if more than 1 × 103 colony-forming units (CFU) of H. pylori per gram of stomach were recovered. (A) Results from experiment 1, in which mice were inoculated two times with a dose of 2 × 108 cells. (B) Results from experiment 2, in which mice were inoculated two times with a dose of 1 × 109 cells. For both panels, open symbols represent SS1 inoculated mice, and closed symbols represent Hyd:cm inoculated mice. According to Student's t-distribution test (21), the parent strain results (both experiments) are significantly greater than the mutant at the 99% degree of confidence (α′ equals 0.01, for a one-tailed test). This conclusion was so even if the undetectable CFU's (most of the data points for the mutant) were assumed to be 1 × 103 CFU of H. pylori per gram of stomach.

We determined the average hydrogen content of the mucus layer of the mouse stomach to be 43 μM, over 20 times as much as that of the apparent whole-cell K M for hydrogen. This concentration represents the average of 31 measurements taken from different regions of stomachs from four live, anesthetized mice (13) (Table 2). These measurements were taken on different days and at different times during the day and ranged in concentrations from 17 to 93 μM, indicating that under most conditions the hydrogen oxidizing system in H. pylori would be saturated. It may be expected that the type of diet of the animal would affect the colonic flora fermentation responses; diet would then affect the hydrogen concentrations in tissues, but was not studied here.

Table 2

Hydrogen concentrations in mouse stomachs. A 50-μm size microelectrode probe was used to measure H2 in the mucus lining area of the stomach of live (anesthetized) mice. For assay details see (13).

View this table:

A wide range of characteristics attributed to infectious bacteria are classified as virulence determinants to successfully combat inherent host protection mechanisms. However, the primary sources of energy used by infectious bacteria to sustain their growth, once they are established in an animal host, remain largely unknown (18). The use of molecular H2, a high-energy, diffusible reductant produced by colonic fermentations from other host-residing bacteria, thus represents a useful tool for understanding how a human pathogen grows and persists in an animal host. Hydrogen use may play an especially important role in setting up the stable infection required for the most serious of the pathologies associated with H. pylori infection, gastric ulceration, and cancer. Blood- or serum-containing media is commonly used for routine (laboratory) culture of H. pylori, and the nature of the carbon and energy sources used in the host are unknown.Helicobacter pylori is very limited in its use of oxidizable carbon substrates (19), and the primary environment forH. pylori colonization is within the complex and viscous mixture of glycoproteins known as mucin. This is expected to provide little nutritional value for the pathogen. Fermentation reactions in the colon include the hydrogen-producing reactions accompanied by acetate and butyrate production by bacteria of the anaerobic large intestine (2). This colonic H2 must move into other tissues, presumably by a combination of cross-epithelial diffusion (6) and vascular-based transport processes (20). Indeed, it has been estimated that 14% of all the intestinal-produced hydrogen is excreted through the breath (of humans), and the authors speculate that the hydrogen is carried to the lungs via the bloodstream (4). The proportion of exhaled gas as H2 can vary considerably among individuals (2, 4), so it may be possible to correlate H. pylori infection with inherent host H2-production characteristics. From our studies, H2 use must represent a large energy boost for a bacterium living in an energy-poor environment (such as gastric mucosa). H2 is an energy substrate not used by the host, so competition for this high-energy substrate in the gastric environment is not a factor. Also, some other human pathogens contain uptake-type hydrogenases, so H2 utilization within animal hosts may extend beyond just H. pylori and gastric infections.

Supporting Online Material

Materials and Methods

  • * Present address: Department of Microbiology, North Carolina State University, Raleigh, NC 27695, USA.

  • To whom correspondence should be addressed. E-mail: rmaier{at}


View Abstract

Navigate This Article