Gallbladder microbiota in healthy dogs and dogs with mucocele formation

Abstract

To date studies have not investigated the culture-independent microbiome of bile from dogs, a species where aseptic collection of bile under ultrasound guidance is somewhat routine. Despite frequent collection of bile for culture-based diagnosis of bacterial cholecystitis, it is unknown whether bile from healthy dogs harbors uncultivable bacteria or a core microbiota. The answer to this question is critical to understanding the pathogenesis of biliary infection and as a baseline to exploration of other biliary diseases in dogs where uncultivable bacteria could play a pathogenic role. A pressing example of such a disease would be gallbladder mucocele formation in dogs. This prevalent and deadly condition is characterized by excessive secretion of abnormal mucus by the gallbladder epithelium that can eventually lead to rupture of the gallbladder or obstruction of bile flow. The cause of mucocele formation is unknown as is whether uncultivable, and therefore unrecognized, bacteria play any systematic role in pathogenesis. In this study we applied next-generation 16S rRNA gene sequencing to identify the culture-negative bacterial community of gallbladder bile from healthy dogs and gallbladder mucus from dogs with mucocele formation. Integral to our study was the use of 2 separate DNA isolations on each sample using different extraction methods and sequencing of negative control samples enabling recognition and curation of contaminating sequences. Microbiota findings were validated by simultaneous culture-based identification, cytological examination of bile, and fluorescence in-situ hybridization (FISH) performed on gallbladder mucosa. Using culture-dependent, cytological, FISH, and 16S rRNA sequencing approaches, results of our study do not support existence of a core microbiome in the bile of healthy dogs or gallbladder mucus from dogs with mucocele formation. Our findings further document how contaminating sequences can significantly contribute to the results of sequencing analysis when performed on samples with low bacterial biomass.

Fig 1. Number of sequencing reads and amplicon sequence variants (ASVs) identified in 13 samples of bile from the gallbladder of healthy control dogs, 13 samples of mucus from the gallbladder of dogs with mucocele formation and 4 negative extraction controls run in triplicate under each of two extraction protocols with or without introduction of a sterile swab.

Method 1 DNA extraction performed using phenol. Method 2 DNA extraction performed without use of phenol as described in Methods.

Fig 2. Percent abundance of taxa for which 16S rRNA sequences were amplified from negative extraction control samples under each of 4 DNA isolation conditions.

Fig 3. Percent abundance of phyla amplified from the bile of 13 apparently healthy adult dogs (panel A) and gallbladder mucus from 13 dogs with mucocele formation (panel B).

Data were filtered for contaminating sequences and taxa with read counts ≤10 in all samples. For each sample, abundance data are shown for sequencing results obtained using each of two different extraction methods (phenol and no-phenol).

Fig 4. Interior appearance of a gallbladder mucocele after surgical removal from a dog.

The gallbladder was opened lengthwise exposing the intraluminal mucus content which served as the source of sample used for 16S rRNA gene sequencing.

Fig 5. Results of fluorescence in-situ hybridization (F.I.S.H.) for eubacteria and Gram stain of gallbladder mucosa from 4 dogs with mucocele formation and concurrent findings of conventional culture and 16S rRNA gene sequencing of mucus content.

Bacteria are indicated by closed arrowhead and visualized by means of positive hybridization with Eub 338—6FAM (green; panel A), Gram stain (panel B and D), or DAPI staining of bacterial nucleic acid (blue; panel A, C, and E).

Fig 6. Beta diversity of amplicon sequence variants in gallbladder bile from 13 healthy dogs and gallbladder mucus from 13 dogs with mucocele formation each having DNA extracted using two different methods.

Bray-Curtis principal coordinates analysis plots showing clustering of microbial communities based on isolation method (Panel A), disease condition (Panel B), and disease condition by isolation method (Panel C). Axis 1, 8.295%; Axis 2, 12.05%; Axis 3, 14.34%. Group significance plot comparing distances from Control—No Phenol (Panel D). Brackets represent significant differences between groups (PERMANOVA **q < 0.01). Box and Tukey whiskers represent interquartile range (IQR) and ± 1.5×IQR, respectively.