Gut mucosal virome alterations in ulcerative colitis

Abstract

Objective: The pathogenesis of UC relates to gut microbiota dysbiosis. We postulate that alterations in the viral community populating the intestinal mucosa play an important role in UC pathogenesis. This study aims to characterise the mucosal virome and their functions in health and UC.

Design: Deep metagenomics sequencing of virus-like particle preparations and bacterial 16S rRNA sequencing were performed on the rectal mucosa of 167 subjects from three different geographical regions in China (UC=91; healthy controls=76). Virome and bacteriome alterations in UC mucosa were assessed and correlated with patient metadata. We applied partition around medoids clustering algorithm and classified mucosa viral communities into two clusters, referred to as mucosal virome metacommunities 1 and 2.

Results: In UC, there was an expansion of mucosa viruses, particularly Caudovirales bacteriophages, and a decrease in mucosa Caudovirales diversity, richness and evenness compared with healthy controls. Altered mucosal virome correlated with intestinal inflammation. Interindividual dissimilarity between mucosal viromes was higher in UC than controls. Escherichia phage and Enterobacteria phage were more abundant in the mucosa of UC than controls. Compared with metacommunity 1, metacommunity 2 was predominated by UC subjects and displayed a significant loss of various viral species. Patients with UC showed substantial abrogation of diverse viral functions, whereas multiple viral functions, particularly functions of bacteriophages associated with host bacteria fitness and pathogenicity, were markedly enriched in UC mucosa. Intensive transkingdom correlations between mucosa viruses and bacteria were significantly depleted in UC.

Conclusion: We demonstrated for the first time that UC is characterised by substantial alterations of the mucosa virobiota with functional distortion. Enrichment of Caudovirales bacteriophages, increased phage/bacteria virulence functions and loss of viral-bacterial correlations in the UC mucosa highlight that mucosal virome may play an important role in UC pathogenesis.

Keywords: bacteria; gut mucosa; ulcerative colitis; virome; virome metacommunity.

Figure 1

Mucosal virome alterations in abundance and diversity in UC. (A) Comparison of Caudovirales abundance in the rectal mucosa of controls and UC subjects. The Caudovirales abundance was calculated as RPKM sum of Caudovirales contigs recruited reads normalised by sequence depth in each subject. The dots indicate individual values of the studied subjects. Statistical significance was determined by Mann-Whitney test, *p<0.01. (B) Comparison of the mucosal Caudovirales α diversity based on Shannon diversity, evenness, Chao1 richness index in the mucosa of controls and UC subjects. Statistical significance was determined by t-test. *P<0.05; **P<0.01. For box plots, the boxes extend from the first to the third quartile (25th to 75th percentiles), with the median depicted by a horizontal line. RPKM, reads per kilobase per million.

Figure 2

Between-group and within-group viral community structure difference in health and UC. (A) Mucosal viral community structure difference between controls and UC by NMDS (non-metric multidimensional scaling) plotting based on Bray-Curtis dissimilarities at the viral species level. (B) Between-group viral community structure difference reflected on each axis of NMDS plotting. (C) Comparison of within-group mucosal virome Bray-Curtis dissimilarities between healthy individuals and UC subjects. Statistical significance was determined by t-test. **P<0.01; ***P<0.001; ****P<0.0001. For box plots, the boxes extend from the first to the third quartile (25th to 75th percentiles), with the median depicted by a horizontal line.

Figure 3

Differential viral taxa between health and UC mucosa at the family, genus and species levels. Differentially enriched viral families (A), genera (B) and species (C) between health and UC mucosa were determined by DESeq analysis with false discovery rate (FDR) correction (only those differential taxa with adjusted p<0.05 and |Log₂ (between-group fold change)| >2 are shown). For viral taxon names, taxa colour coated by black denote prokaryotic viruses, while those colour coated by orange denote eukaryotic viruses. For viral abundance box plots, the boxes extend from the first to the third quartile (25th to 75th percentiles), with the median depicted by a horizontal line. RPKM, reads per kilobase per million mapped reads.

Figure 4

Mucosal virome metacommunities in health and UC. (A) Virome metacommunity clustering based on partition around medoids (PAM) algorithm and principal coordinates analysis (PCoA) on the viral community structures of health and UC mucosa. The inset shows the ratio of healthy individuals and UC subjects within each metacommunity population. (B) Heatmap of the presence of differential viral species contributing to clustering of the two mucosal virome metacommunities. Discriminative species were identified by concordant DESeq and Random Forest analyses. Viral species abundances are colour intensified according to Log₁₀ RPKM values. Only those species concordantly determined by DESeq and Random Forest algorithm with effect size more than 2 and FDR-adjusted p<0.05 are shown. RPKM, reads per kilobase per million mapped reads.

Figure 5

Significant altered presence of core, common and unique viral species in different UC metacommunities. (A) The proportion of core, common and unique viral species in each subject. The core species, common species and unique species correspond to viral species shared among >50%, 20%–50% and <20% of studied subjects, respectively. (B) Quantification of the presence ratio of core, common and unique viral species in the viral communities of healthy controls and UC subjects with two respective mucosal virome metacommunities. Statistical significance was determined by one-way analysis of variance (ANOVA). **P<0.01; ****P<0.0001. (C) Heatmap of the abundances of the most abundant 100 core, common and unique species in healthy controls and UC subjects with two respective mucosal virome metacommunities. Viral species abundances are colour intensified according to Log₂ reads per kilobase per million (RPKM) values.

Figure 6

Significant loss of diverse viral functions in UC mucosa with concomitant increases in bacteria-pathogenicity associated functions. (A) Presence-absence heatmap of the classified viral functions in controls and UC. Viral functions were predicted and classified via HUMANN2 pipeline, exploiting the sophisticated Gene ontology (GO) and Pfam protein family databases. Functions with reads per kilobase (RPK) >10 were considered present in individuals. The abundance distribution of classified viral functions is plotted in the line chart, with abundance values expressed as Log₂RPK. (B) Differentially enriched viral functions between health and UC mucosa. Differential viral functions were determined by DESeq analysis with FDR correction. Only those functions with adjusted p<0.05 and |Log₂ (between-group fold change)| >2 are shown. For box plots, the boxes extend from the first to the third quartile (25th to 75th percentiles), with the median depicted by a horizontal line.

Figure 7

Mucosal bacterial-viral correlation patterns in health and UC. (A) Correlations between the α diversity (diversity, evenness and richness) of mucosal bacteria and viruses in healthy controls and UC, respectively. Spearman’s correlation coefficient was calculated, while statistical significance was determined for all pairwise comparisons. Significant correlations (p<0.05) are displayed with asterisk. *P<0.05; ***P<0.001. (B) Correlations between the most abundant 30 virus species and the most abundant 20 bacteria genera in health and UC mucosa. Spearman’s correlation coefficient was calculated, while statistical significance was determined for all pairwise comparisons. Only statistically significant correlations were plotted, where blue circles indicate positive correlations and red circles indicate inverse correlations. The size and shading indicate the magnitude of the correlation where darker shades denote more intensive correlations than light ones.