Metagenomic sequencing of bile from gallstone patients to identify different microbial community patterns and novel biliary bacteria

Abstract

Despite the high worldwide prevalence of gallstone disease, the role of the biliary microbiota in gallstone pathogenesis remains obscure. Next-generation sequencing offers advantages for systematically understanding the human microbiota; however, there have been few such investigations of the biliary microbiome. Here, we performed whole-metagenome shotgun (WMS) sequencing and 16S rRNA sequencing on bile samples from 15 Chinese patients with gallstone disease. Microbial communities of most individuals were clustered into two types, according to the relative enrichment of different intestinal bacterial species. In the bile samples, oral cavity/respiratory tract inhabitants were more prevalent than intestinal inhabitants and existed in both community types. Unexpectedly, the two types were not associated with fever status or surgical history, and many bacteria were patient-specific. We identified 13 novel biliary bacteria based on WMS sequencing, as well as genes encoding putative proteins related to gallstone formation and bile resistance (e.g., β-glucuronidase and multidrug efflux pumps). Bile samples from gallstone patients had reduced microbial diversity compared to healthy faecal samples. Patient samples were enriched in pathways related to oxidative stress and flagellar assembly, whereas carbohydrate metabolic pathways showed varying behaviours. As the first biliary WMS survey, our study reveals the complexity and specificity of biliary microecology.

Figure 1. Distributions of WMS sequencing reads and microbial communities within bile samples.

(a) Ratios of WMS sequencing reads after removing host background (hollow black-edged bars) and ratios of reads aligned to bacterial database (inner bars). All ratios were calculated by dividing by the total clean read number after quality control. Sample IDs are labelled on the x-axis, and ranked by their clustering relationships. Bacteria-aligned read ratios of samples from groups A, B and C are indicated by light blue, dark blue and green bars, respectively. (b) Hierarchical clustering of samples based on microbial community, distribution of bacterial origin and heatmap of species abundances. Individuals are denoted by coloured blocks, as in (a). Bacterial origins were classified as oral (referring to the oral cavity/respiratory tract), intestinal, both oral and intestinal, and other (environmental or unknown). For bacteria with ≥0.1% abundance in at least three individuals, the distribution of their origin in each individual is reflected by the histogram above the heatmap. The heatmap colour scale quantifies the log₁₀ relative abundance of species, from grey (none or low abundance) to dark red (high abundance).

Figure 2. Microbial community characteristics generated by WMS and 16S sequencing.

(a–b) Relative abundances at the genus level by WMS (a) and 16S sequencing (b). Circle sizes represent abundance, and circles are coloured by genus. Sample IDs are shown on the x-axis and are ranked as in Fig. 1. Common genera were those with a relative abundance ≥1% in at least three samples. The genera names for (a,b) are shown on the y-axis of (a). (c–d) Distributions of log₂ alpha (c) and beta diversities (d). Orange and red boxplots denote distributions of WMS and 16S sequencing of bile samples respectively, accompanied with the distributions of faecal samples from HMP (purple boxplots). Log₂ alpha and beta diversities are presented as scatter plots in insets (orange, bile WMS; red, bile 16S), with scales identical to the corresponding boxplot y-axis. (e–f) Species and OTU counts by WMS (e) and 16S sequencing (f) with increasing sample number. At a given sample number, a maximum 100-time random sampling of these bile samples was performed to calculate species or OTU counts. Boxes represent the interquartile range (IQR) between first and third quartiles (25th and 75th percentiles, respectively). Lines inside denote the median, and whiskers denote the most extreme values within 1.5 times IQR from the first and third quartiles, respectively. Outlier values are represented as circles.

Figure 3. Genome coverage (≥1%) of bacteria newly identified in bile samples by WMS.

Reference genomes of 13 newly identified bacteria are represented by circles generated by BLAST Ring Image Generator v0.95 (BRIG), with genome sizes labelled inside. Highly confident WMS read alignments (unique mapping and ≥99% identity) to these genome are illustrated by purple bars around circles. Bar height represents site coverage depth, with the corresponding scale indicated inside the circles. Overall coverage of the reference genome is labelled below the genome size. Newly identified bacteria are sorted by genome coverage. P. pallens ATCC 700821 appears twice because it was observed in two samples with ≥1% coverage (B5 and C4). E. coli UMN026, which has the highest WMS coverage (44%, in sample B5), is also shown at the end for reference.

Figure 4. Distributions of bacterial species harbouring genes involved in gallstone formation (a) or bile resistance (b).

Numbers of species within individuals are indicated by bars. Red bars denote genes that were also identified in metagenomic assemblies by gene prediction, or existed within metabolic reconstruction results. The gene encoding phospholipase C within individual B2 (blue bar with an arrow on top) was identified by metabolic reconstruction, but not by genome annotation. The term ‘urease’ on the left panel stands for genes encoding urease, which include ureA, ureB, ureC, ureD, ureE, ureF and ureG. The term ‘acrA/acrB’ stands for genes acrA and acrB, and ‘emrA/emrB’ stands for genes emrA and emrB.

Figure 5. Metabolic reconstruction of the biliary microbiota.

Bold grey lines represent abundant pathways that existed in at least five bile samples with ≥1% abundance. Pathways that are enriched or depleted in the biliary microbiota compared to healthy faecal microbiota from HMP are highlighted by green and blue lines, respectively.