Friday, February 7, 2014

Guest Blog Post on Helicobacter pylori: Friend or Foe?

We're not too far into 2014 and we already have a contender for "favorite paper of the year". This treat is the story behind that paper - one that remarkably interconnects phylosymbiosis with human health. 

Kodaman et al, PNAS 2014 

Contributed by Carrie L. Shaffer, PhD
Twitter @CLShafferPhD


Electron micrograph of H. pylori (Source)

Global dissemination of Helicobacter by modern humans

One of the most successful human pathogens in recorded history is Helicobacter pylori, a Gram-negative bacterium that has colonized the gastric niche of modern humans since our migration out of Africa approximately 60,000 years ago.  Since leaving Africa, H. pylori has diversified in parallel with its human host.  Because of the predominantly vertical transmission of H. pylori within family units and the chronic carriage of H. pylori for nearly the entire lifetime of an individual, we have the remarkable ability to trace human migrations by analyzing H. pylori genetic signatures.  Recent evidence suggests that H. pylori has colonized humans for at least 100,000 years, and points to a second exodus out of Africa 52,000 years ago that resulted in generation of a hybrid European lineage of H. pylori strains (Moodley et al.).

Today, H. pylori colonizes the gastric mucosa of at least 3 billion people worldwide. Gastric colonization by H. pylori results in the development of asymptomatic chronic non-atrophic gastritis in all cases.  However, in a small percentage of cases, severe gastric disease can manifest, including emergence of peptic ulcers and gastric adenocarcinoma, which is currently the second leading cause of all cancer-related deaths worldwide. In contrast to the United States, where the prevalence of gastric cancer is low, the incidence of gastric adenocarcinoma varies globally with the highest rates occurring in Japan, China, South America, and Eastern European nations.

Because of its diverse and segregated geographic and ethnic landscape, the South American country of Colombia serves as an ideal natural laboratory for the study of H. pylori-related disease and its association with gastric adenocarcinoma. Separated by only 200 kilometers, the Colombian towns of Tuquerres and Tumaco have sharply contrasting rates of gastric cancer.  Tuquerres, located in the altitude of the Andes Mountains, has one of the highest rates of gastric cancer in the world (approximately 150 cases/100,000 people), whereas the coastal town of Tumaco has a gastric cancer rate of just 6 cases per 100,000 people, despite a nearly universal rate of H. pylori colonization throughout Colombia.  


Health outcomes echoing a continued legacy of European colonization

A rich history of settlement by native Amerindians coupled with both colonization and conquest by Spanish invaders and the introduction of African slave trades led to racial and ethnic segregation within Colombia.  People residing in modern day Tuquerres are mainly of mestizo (Spanish and Amerindian) descent, while the residents of Tumaco are primarily of mulatto (African and Spanish) descent.  These observations led us to the central question of our recently published paper in PNAS (Kodaman et al.): Are the rates of gastric cancer in Colombia governed by human or Helicobacter ancestry? We reasoned that a 25-fold difference in the rate of stomach cancer between the two populations could be attributed to differences in the host genetic composition; differences in host immune responses to chronic infection; phylogeographic origin and/or virulence factors harbored by H. pylori; environmental factors such as diet and smoking status; or a combination thereof. 

Our previous work (de Sablet et al.) demonstrated that European H. pylori ancestry is strongly predictive of increased premalignant lesions and epithelial DNA damage in both Colombian populations, while African H. pylori phylogeographic origin is associated with reduced severity in histologic disease parameters.  These findings are significant because we were able to show that in this region, the genetic ancestry of H. pylori is both a risk factor for development of gastric disease and a determinant of gastric disease severity, irrespective of the presence of the major virulence factors CagA (an oncoprotein present in some strains of H. pylori) and VacA (a secreted toxin present in all strains of H. pylori).  However, this study was constrained to a small subset of our larger cohort, and thus, we wanted to further investigate how Helicobacter phylogeographic origin contributes to development of gastric cancer.

In our follow-up study, we expanded the number of patients analyzed to 242 (122 from Tumaco and 120 from Tuquerres), and focused our efforts on analyzing human ancestry in parallel with the phylogeographic origin of the H. pylori cultured from patient-matched gastric biopsies.  Using Immunochip analysis to genotype nearly 200,000 SNPs in each patient’s blood sample, we determined that human ancestry closely followed expected patterns based on historical events: the people of coastal Tumaco trace the majority of their ancestry to Africa (58%), while the mountain population of Tuquerres are largely of Amerindian (67%) and European (31%) descent. We used multilocus sequence typing (MLST) to determine the matched H. pylori phylogeographic origin from each patient, and classified each isolate into an ancestral haplotype based on analysis of housekeeping gene sequences using STRUCTURE.  Helicobacter phylogeography recapitulated our previous report indicating that the bacteria derived from ancestral European, African, and East Asian (Amerindian) lineages.  The matched human and Helicobacter ancestries were then correlated to histopathology scores from gastric biopsies taken from each patient in order to determine if either human or H. pylori genetic ancestry was associated with disease severity.

The answer surprised us: we found that perturbation of human-Helicobacter phylosymbiosis can shift a benign infection to a potentially catastrophic relationship that results in disease progression to gastric cancer.  Rather than either the human or the bacterial ancestry acting alone, the contribution of bacterial-host genetic interplay was found to be the most significant factor for predicting disease outcome.  This result was even more captivating when we analyzed the influence of the major carcinogenic virulence factor CagA on the progression of advanced gastric disease within this cohort – human-microbe incompatibility had a greater effect on the risk for developing severe disease than the bona fide Helicobacter oncoprotein.  We went on to show that disruption of human-Helicobacter phylosymbiosis accounts almost entirely for the differences we observe in stomach cancer risk between Tumaco and Tuquerres.  To our knowledge, this study was the first report of matched genetic analysis of the host and pathogen with subsequent correlation of human-microbe phylosymbiosis to clinical outcome.


Reflections of a distant past

So what do our results mean? One theory is that invasion of Colombia by the Spanish conquistadors introduced ancestral European H. pylori to the Amerindian populations, and these bacteria replaced the East Asian H. pylori strains naturally harbored by the Tuquerres population.  It is thought that H. pylori of European ancestral origin may be more virulent than some other lineages, including H. pylori of ancestral African and Amerindian haplogroups. Since many people are often co-infected by multiple H. pylori strains, it is reasonable to hypothesize that H. pylori of European lineage could have outcompeted strains of Amerindian lineage in native Colombians. Alternatively, co-infection with multiple H. pylori strains and the natural competence of the bacterium leads to a propensity of H. pylori to uptake DNA for recombination, resulting in generation of new strains that are entirely unique at the DNA level. Another possibility is that differences in virulence factors between H. pylori phylogenetic lineages could provide a competitive advantage during co-infection. In fact, analysis of the cag pathogenicity island (which encodes components of a molecular machine used by H. pylori to inject CagA into gastric epithelial cells) in several Amerindian H. pylori isolates revealed numerous changes, including gene rearrangements, indels of approximately 11.2 kb (one quarter the size of the entire pathogenicity island), and gene inversions (Olbermann et al.).  Despite these dramatic variations in macrodiversity, the Amerindian isolates retained functional secretion system machinery, suggesting that the changes are non-deleterious to pathogenicity island function, and are likely under neutral or positive selection.  Strikingly, our study reveled that patients with a high proportion of Amerindian ancestry who are infected with H. pylori strains that have a substantial percentage of African lineage (>20%) develop the most severe gastric lesions. Perhaps the frequent differences in macrodiversity found in Amerindian H. pylori isolates are essential for maintenance of host-microbe phylosymbiosis in the native Colombian population.

The results of our study provide insight into an ongoing debate in the field regarding classification of H. pylori as a pathogen rather than a commensal.  While it is true that H. pylori colonization is associated with a significantly increased risk for development of gastric disease, H. pylori may provide some benefits to its host.  For example, it has been suggested that H. pylori colonization is protective against development of esophageal reflux, cancer of the upper stomach, and cancer of the esophagus.  H. pylori may help to prevent Mycobacterium tuberculosis infection from progressing to active tuberculosis, and may play a role in providing protection from development of asthma and other allergic diseases.  The protective advantages of this fascinating microbe are intriguing, but nonetheless, additional investigation will be required to determine the mechanisms by which H. pylori provides such advantages to its host.  Our study points towards co-analysis of host and microbe genomic variation to identify those at highest gastric cancer risk so that we can selectively target H. pylori eradication.

Finally, our study brings new perspective to the so-called ‘African Enigma,’ a theory that describes the discordant prevalence of H. pylori infection and H. pylori-related disease in continental Africa.  Our findings suggest that maintenance of a harmonious host-microbe phylosymbiotic relationship can account for the low incidence of gastric disease throughout Africa in spite of H. pylori colonization rates estimated at greater than 90%.  In future investigations, it will be important to recapitulate our findings in cohorts from diverse sites including Africa, East Asia, and additional locations within Central and South America in order to determine the extent to which human-Helicobacter phylosymbiosis dictates disease outcome in global populations.