Cholera resistance in Bangladesh: combining signals of ancient, pathogen driven selection with genome wide association to understand immune response. E. K. Karlsson1,2, I. Shylakhter1,2, F. Qadri3, J. B. Harris4,5, S. B. Calderwood4,6, E. T. Ryan4,5,6, R. C. LaRocque4,6, P. C. Sabeti1,2,7 1) Broad Institute, Cambridge, MA; 2) Center for Systems Biology, Harvard Univ, Cambridge, MA; 3) iccdr,b, Dhaka, Bangladesh; 4) Div. of Infectious Diseases, MGH, Boston, MA; 5) Dept of Pediatrics, Harvard Medical School, Boston, MA; 6) Dept of Medicine, Harvard Medical School, Boston, MA; 7) Dept of Immunology and Infectious Disease, Harvard School of Public Health, Boston, MA.

   Cholera is an ancient, deadly and common pathogen in the Ganges River Delta, and has likely exerted evolutionary pressure on human populations in the region. Cholera is caused by the bacterium Vibrio cholerae, which colonizes the small intestine and causes profound diarrhea, rapid dehydration, and, without medical intervention, mortality rates as high as 50%. In prior research, we combined a genome wide selection scan with a targeted association study and showed that cholera response immune pathways are enriched for selection in the Bengali population of Bangladeshi. We developed a model of the human innate immune response wherein inflammasome activation and the NF-B signaling pathway play an integrated role in TLR4-mediated sensing of V. cholerae.
Here, we report the first genome wide association study of cholera susceptibility in Bangladesh. Our results suggest that positive selection drives protective variants of large effect to high frequency, increasing statistical power. We identified genome-wide significant associations to variants in major immune genes by comparing 94 Bengali patients hospitalized with severe cholera to just 80 unphenotyped population controls (543,832 SNPs after QC), results we replicated in a second cohort of 157 cases and 83 controls. We also completed a much denser scan for positive selection, applying our Composite of Multiple Signals approach to newly released 1000 Genomes Project (1000G) full sequence data for the Bengali population. In total, we have integrated association scores for 4.8 million polymorphic SNPs (after imputation) into the composite selection score and found specific immune signaling pathways targeted by cholera driven selection, including TLR4 mediated response to bacterial lipopolysaccharide (LPS). We are now investigating the top candidate variants using new high-throughput functional methods.
Our approach is broadly applicable to other historically prevalent infectious diseases, such as Lassa fever, tuberculosis, leishmaniasis, dengue fever and malaria, and to common diseases, such as inflammatory bowel disease, for which the associated genes may have been historically selected. Combining ancient history with modern genomics is a powerful approach for investigating genome function.

You may contact the first author (during and after the meeting) at