Hp has long been considered a microaerophile that requires O2 for growth but is highly sensitive to atmospheric O2 levels. In the present study, however, we demonstrate that atmospheric O2 tension does not kill Hp cells but promotes growth of cells when inoculated at high density, and Hp is unique in that it absolutely requires high CO2 tension for optimal growth and long-term survival. Eliminating the need to remove O2 makes it considerably easier to culture Hp in the laboratory.
Bury-Moné et al. reported that Hp strains showed similar growth profiles under aerobic and microaerobic conditions. However, when cells were inoculated in medium containing 0.2% β-cyclodextrin to low density (107 CFU/ml), growth was not detected under 15% O2 and 6% CO2 (generated with CO2 Gen gas packs) . In contrast, we found that atmospheric O2 tension did not kill Hp cells but did prolong the lag period of cultures inoculated at low cell density (3 × 104 CFU/ml). The conflicting results may have been due to different experimental conditions. We used 10% CO2 to culture Hp, whereas the previous study used 6% CO2. Culture medium pH may increase faster under lower CO2 levels than under 10% CO2, thereby inhibiting bacterial growth, particularly under 20% O2. Further, because the lag period of low-density cultures is prolonged under 20% O2, the culture period in the previous study may have been insufficient to detect growth.
Bury-Moné et al. investigated whether growth inhibitory factors played a role in the lack of Hp growth under aerobic conditions. In one experiment, Hp cells were inoculated at high or low densities in two compartments separated by a membrane that stopped bacteria moving between them but allowed the exchange of metabolites, chemical compounds, and macromolecules, and cultured under aerobic or microaerobic conditions. In another experiment, a freshly inoculated culture was supplemented with culture medium in which a high-density or low-density culture had grown. Neither experiment revealed effects of inhibitory factors .
In the present study, we found that the effect of O2 on Hp growth was dependent on inoculum size: aerobic conditions inhibited growth in low-density cultures but induced growth in high-density cultures. Conversely, under low O2 tension, low-density cultures grew faster than high-density cultures. In the present study, HPLC analysis of Hp metabolites revealed higher levels of acetate, succinate, and lactate at lower O2 tensions. These results are consistent with previous reports that Hp utilizes aerobic respiration or fermentation, depending on environmental O2 levels, suggesting a possibility that Hp is a facultative anaerobe. On the basis of these data, we presumed that it is more efficient for a low-density culture to generate ATP by fermentation rather than by aerobic respiration. In Escherichia coli, enzymes involved in the tricarboxylic acid (TCA) cycle are significantly downregulated (2- to 10-fold) and fermentation enzymes are highly upregulated (>10-fold) when glucose is used as a carbon source under microaerobic conditions; the reverse is true under aerobic conditions . Likewise, in Hp, fermentation enzyme activity would be expected to be lower under 20% O2 than under 2% or 8% O2. In addition, we observed that Hp produced more organic acids in the absence of CO2 than in the presence of CO2 (Figure 5C), suggesting that CO2 is important for efficient aerobic respiration in Hp cells, probably for enzyme induction.
CO2 is involved in a wide range of biological processes, and the addition of CO2 has been shown to shorten the lag period of bacterial cultures . Hp requires high level of CO2 for its growth and generates a large amount of CO2 through urease activity. The shaking of cultures during incubation dissipates metabolic CO2, thus Hp growth would be greatly influenced by inoculating cell density, especially under aerobic conditions. We tested this possibility by supplementing a culture inoculated at low density (3 × 104 CFU/ml) with bicarbonate; however, bicarbonate did not increase the growth rate (data not shown).
Another possible explanation for the growth inhibiting effect of O2 is the bacterial signaling system known as quorum sensing, which monitors cell population density . Bacteria release low molecular-weight autoinducers that accumulate in the environment; at threshold concentrations, these signaling molecules induce the coordinated expression of target genes in the population. Hp has been shown to possess a quorum-sensing system , and autoinducer 2 appears to regulate motility and flagella morphogenesis . In Pseudomonas aeruginosa, expression of the quorum-sensing regulatory protein LasR is regulated by iron and O2 . It is not known whether O2 concentration serves as an environmental signal for monitoring cell density in Hp. The precise mechanism for the growth inhibition by high O2 levels is under investigation.
Numerous studies have been carried out to elucidate Hp physiology under oxidative stress, including studies of morphology, gene expression, and protein expression. However, in some of these experiments, Hp was cultured under atmospheric O2 tension without supplemental CO2 [29, 49–51]. Therefore, coccoid transformation and subsequent cellular changes may have resulted, at least in part, from CO2 deprivation rather than oxidative stress.
A unique feature of Hp is its transformation to coccoid form under stress conditions. Coccoid transformation was thought to be a passive conversion that eventually leads to cell death . However, several recent reports have suggested that coccoid transformation is an active process that allows Hp to adapt to its environment [52–54]. In the present study, CO2 deprivation induced coccoid formation, but this morphological transformation was delayed in cells cultured under high O2 tension, supporting the view that coccoid transformation of Hp is not a passive process but an active energy-consuming process.
In this study, we observed that actively growing cells, but not those at a stationary phase, produce OMVs, which are discrete, closed outer membrane blebs produced by gram-negative bacteria, especially pathogenic strains . They are believed to serve as secretory vesicles that transmit virulence factors to host cells. OMVs are released by actively growing cells, and their maximal production occurs at the end of log phase in E. coli, Vibrio cholerae, and Brucella melitensis [56–58]. Hp OMVs are involved in biofilm formation in vitro and deliver VacA cytotoxin to gastric epithelium [59, 60]. They induce growth arrest and IL-8 production by gastric epithelial cells, which have been associated with gastritis caused by Hp infections [61, 62], and also enhances the carcinogenic potential of Hp . Taken together, these reports and results obtained in the present study indicate the higher virulence of actively growing Hp cells, which are able to damage host cells through toxin delivery.
In the present study, cultivation of Hp cells in the absence of CO2 increased intracellular ppGpp levels, suggesting induction of the stringent response, which induces a global alteration in cellular transcription and indirectly activates genes involved in amino acid biosynthesis [42, 64]. Many factors induce the stringent response, but nutrient stress from amino acid starvation has been the best studied. Induction of the stringent response by CO2 deprivation has also been reported in Campylobacter jejuni, a capnophilic microaerophile that is closely related to Hp .
The bicarbonate concentration of gastric juice is approximately 25 mM . Hp generates additional CO2 via the breakdown of urea, thereby increasing bicarbonate levels. In fact, the gastric CO2 levels in Hp-positive volunteers were significantly higher than those of Hp-negative subjects . The affinity for CO2 may thus be related to its ecological niche, which may have lead to adaptation and eventually dependency on high CO2 concentrations. Hp shows chemotactic responses towards high CO2 concentration in vitro . Elevated levels of CO2/bicarbonate serve as a signal of the host environment and often increase the expression of diverse virulence factors [69, 70]; however, the association between CO2 and virulence in Hp remains to be determined.