P155 Gallstone aetiologies and bacteria: a multi-site microbiome study of 51 Colombian patients

Introduction Gallstones that form and grow in the gallbladder often lead to disease or ultimately require surgical removal of the gallbladder (cholecystectomy). Aetiologies of gallstone formation are not completely understood, especially where causal roles or routes of inflammation and/or bacterial microbiota are suspected. Previous biliary microbiome/biota studies have been mostly single-site and have not resolved such questions; confounding with exogenous bacteria or bacterial DNA (‘contaminome’) may also have been underestimated.

Methods Samples from gallbladdder bile, gallstones, gallbladder tissue, subgingival plaque and faeces were obtained from 51 patients as well as a DNA extraction blank, at the time of elective cholecystectomy surgery at the Mederí University Hospital in Bogotá, Colombia. Illumina MiSeq sequencing was carried out from the 5 biological sites, as well as the DNA extraction blank, for each patient. Rigorous controls were also included to identify contribution of post-operation contamination by bacteria/bacterial DNA from laboratories, equipment or reagent kits. The paired-end V4 16S rDNA read sequences were analysed using (a) a standard clustering protocol (mothur), and (b) a custom highest-resolution (zero-radius/zOTU) sequence processing protocol.

Results We compared bacterial frequencies at five biological sites and a DNA extraction control for each of the 51 patients, and individually analysed each bacterial sequence-type having an abundance above a globally defined cutoff. Pseudomonas and Flavobacterium were identified as dominant contaminants, as is often reported. A significant portion of the bile samples were determined to be low biomass (less than 104 CFU bacteria/mL bile). Viewing of exact sequences’ 6x51 frequency matrices (sites x patients) allowed contrast-based recognition of consistent patterns of biliary bacteria in patients (frequencies in bile vs gallstone vs tissue). Ten patients exhibited high overall viable bacterial counts in bile (as inferred independently via sequences), allowing comparison with clinical records and identification of compatible gallstone formation/aetiology models. Also in other patients, more abundant bacterial sequence-types could often be reliably identified as originating from biliary sites.

Conclusions Our finding that a fifth of the cohort had infections or marked bacterial presence in bile agrees with previously reported percentages. The sequence-based and experimental findings of generally low levels of viable biliary bacteria even in many gallstone patients is compatible with a notion that bacteria are rare or absent in healthy individuals. Some abundant bacteria that were reliably recognised as having biliary origins could be tracked across the cohort. The design/methods and results presented here, including the extensive controls, may aid future studies investigating roles of bacteria in gallstone aetiologies.