Abstract
Helicobacter pylori is a gram-negative bacterium that colonizes the human gastric mucosa and can lead to gastric inflammation, ulcers, and stomach cancer. Due to the increase in H. pylori antimicrobial resistance new methods to identify the molecular mechanisms of H. pylori-induced pathology are urgently needed. Here we utilized a computational biology approach, harnessing genome-wide association and gene expression studies to identify genes and pathways determining disease development. We mined gene expression data related to H. pylori-infection and its complications from publicly available databases to identify four human datasets as discovery datasets and used two different multi-cohort analysis pipelines to define a H. pylori-induced gene signature. An initial Helicobacter-signature was curated using the MetaIntegrator pipeline and validated in cell line model datasets. With this approach we identified cell line models that best match gene regulation in human pathology. A second analysis pipeline through NetworkAnalyst was used to refine our initial signature. This approach defined a 55-gene signature that is stably deregulated in disease conditions. The 55-gene signature was validated in datasets from human gastric adenocarcinomas and could separate tumor from normal tissue. As only a small number of H. pylori patients develop cancer, this gene-signature must interact with other host and environmental factors to initiate tumorigenesis. We tested for possible interactions between our curated gene signature and host genomic background mutations and polymorphisms by integrating genome-wide association studies (GWAS) and known oncogenes. We analyzed public databases to identify genes harboring single nucleotide polymorphisms (SNPs) associated with gastric pathologies and driver genes in gastric cancers. Using this approach, we identified 37 genes from GWA studies and 61 oncogenes, which were used with our 55-gene signature to map gene-gene interaction networks. In conclusion, our analysis defines a unique gene signature driven by H. pylori-infection at early phases and that remains relevant through different stages of pathology up to gastric cancer, a stage where H. pylori itself is rarely detectable. Furthermore, this signature elucidates many factors of host gene and pathway regulation in infection and can be used as a target for drug repurposing and testing of infection models suitability to investigate human infection.
Highlights
Helicobacter pylori colonizes the stomach of approximately half of the world’s human population
Other studies have identified an association of polymorphisms in the TLR5 gene with atrophic gastritis [5] as well as other autoimmune reactions [6]
The four human gastric biopsies datasets included in the downstream analysis were used for the discovery of gene-signature and contained samples from 98 human samples, including data from 72 H. pylori-infected/gastritis/metaplasia patients, and 26 healthy controls
Summary
Helicobacter pylori colonizes the stomach of approximately half of the world’s human population. Other studies have identified an association of polymorphisms in the TLR5 gene with atrophic gastritis [5] as well as other autoimmune reactions [6] These studies hint at a complex regulatory network for disease progression during H. pylori-infection. Further gene expression studies from human patients and experimental models have elucidated many of the molecular mechanisms relevant to H. pylori pathogenicity and the pathways related to the various disease stages Their results remain indecisive as they show a variable picture, most likely due to low sample numbers in individual studies or variations in disease stage and severity in analyzed samples [7, 8]. A powerful model to study a tissue and cell specific reaction to H. pylori especially at the different stages of the pathology is the use of cell lines or animal models Such studies have elucidated many factors that contribute to disease pathogenesis. The use of adenocarcinoma cell lines is limiting because many of the primary cell transitions would be hard to detect, and the suitability of cellular systems to imitate the host’s reaction to the infection is difficult to predict purely on the basis of such biological studies
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call