We have constructed a genome-saturating mutant library of the human gastric pathogen Helicobacter pylori. Microarray tracking of transposon mutants (MATT) allowed us to map the position of 5,363 transposon mutants in our library. While we generally found insertions well distributed throughout the genome, 344 genes had no detectable transposon insertions, and this list is predicted to be highly enriched for essential genes. Comparison to the essential gene set of other bacteria revealed a surprisingly limited overlap with all organisms tested (11%), while 55% were essential in some organisms but not others. We independently verified the essentiality of several gene products, including an HtrA family serine protease, a hypothetical protein with putative phospholipase D activity, and a riboflavin specific deaminase. A limited screen for motility mutants allowed us to estimate that 4.5% of the genome is dedicated to this virulence-associated phenotype.

Chronic infection of the human stomach by Helicobacter pylori leads to a variety of pathological sequelae, including peptic ulcer and gastric cancer, resulting in significant human morbidity and mortality. Several genes have been implicated in disease related to H. pylori infection, including the vacuolating cytotoxin and the cag pathogenicity island. Other factors important for the establishment and maintenance of infection include urease enzyme production, motility, iron uptake, and stress response. We utilized a C57BL/6 mouse infection model to query a collection of 2,400 transposon mutants in two different bacterial strain backgrounds for H. pylori genetic loci contributing to colonization of the stomach. Microarray-based tracking of transposon mutants allowed us to monitor the behavior of transposon insertions in 758 different gene loci. Of the loci measured, 223 (29%) had a predicted colonization defect. These included previously described H. pylori virulence genes, genes implicated in virulence in other pathogenic bacteria, and 81 hypothetical proteins. We have retested 10 previously uncharacterized candidate colonization gene loci by making independent null alleles and have confirmed their colonization phenotypes by using competition experiments and by determining the dose required for 50% infection. Of the genetic loci retested, 60% have strain-specific colonization defects, while 40% have phenotypes in both strain backgrounds for infection, highlighting the profound effect of H. pylori strain variation on the pathogenic potential of this organism.Helicobacter pylori, a bacterial pathogen of the human stomach, infects an estimated 50% of the population worldwide. Infection by H. pylori causes gastritis initially and, if allowed to persist, can induce a range of pathologies. It is the causative agent of most peptic ulcers, and other serious outcomes such as atrophic gastritis, intestinal metaplasia, and gastric cancer are correlated with long-term infections. It is not yet known whether these outcomes are due to specific factors produced by the organism or whether they result from chronic inflammation due to efficient and persistent colonization of the gastric mucosa. Thus, colonization and persistence factors may themselves constitute virulence factors for this organism.

Gland colonization may be one crucial route for bacteria to maintain chronic gastrointestinal infection. We developed a quantitative gland isolation method to allow robust bacterial population analysis and applied it to the gastric pathobiont Helicobacter pylori. After infections in the murine model system, H. pylori populations multiply both inside and outside glands in a manner that requires the bacteria to be motile and chemotactic. H. pylori is able to achieve gland densities averaging 25 to 40 bacteria/gland after 2 to 4 weeks of infection. After 2 to 4 weeks of infection, a primary infection leads to colonization resistance for a secondary infection. Nonetheless, about ~50% of the glands remained unoccupied, suggesting there are as-yet unappreciated parameters that prevent gastric gland colonization. During chronic infections, H. pylori populations collapsed to nearly exclusive gland localization, to an average of <8 bacteria/gland, and only 10% of glands occupied. We analyzed an H. pylori chemotaxis mutant (Che−) to gain mechanistic insight into gland colonization. Che− strains had a severe inability to spread to new glands and did not protect from a secondary infection but nonetheless achieved a chronic gland colonization state numerically similar to that of the wild type. Overall, our analysis shows that bacteria undergo substantial population dynamics on the route to chronic colonization, that bacterial gland populations are maintained at a low level during chronic infection, and that established gland populations inhibit subsequent colonization. Understanding the parameters that promote chronic colonization will allow the future successful design of beneficial microbial therapeutics that are able to maintain long-term mammalian colonization.

No abstract

No abstract