Distinct gut microbiota and metabolomic profiles in HBV-related liver cirrhosis: insights into disease progression

Abstract

Background: Hepatitis B virus (HBV)-related liver cirrhosis (HBV-LC) is a significant global health issue, affecting gut microbiota (GM) composition and metabolic processes. This study aimed to explore the associations between intestinal microbiota, metabolic profiles, and disease progression in patients with HBV-LC.

Methods: Fecal samples were collected prospectively from 40 healthy controls (HC) and 83 HBV-LC patients between December 2022 and August 2023. Gut microbiota alterations at various stages of liver function were analyzed using 16S rRNA gene sequencing. Untargeted metabolomics was employed to identify potential biomarkers and metabolic pathways associated with early cirrhosis. Additionally, correlations between bacterial genera, inflammatory markers, and metabolites were investigated.

Results: HBV-LC patients demonstrated a significant reduction in bacterial diversity and relative abundance compared to the HC group. Genera such as Alistipes and Lachnospira were notably depleted, while Fusobacterium and Enterococcus were enriched in patients with Model for End-Stage Liver Disease (MELD) scores ≥ 21 or Child-Turcotte-Pugh C grade. Correlation analyses revealed strong associations between intestinal flora, clinical indicators of disease severity, and inflammatory factors. Metabolic analysis showed decreased levels of tocopherol and 21-hydroxypregnenolone, which were strongly linked to the reduced abundance of Alistipes and Lachnospira. Biosynthesis of unsaturated fatty acids and linoleic acid metabolism emerged as critical enrichment pathways.

Conclusions: HBV-LC patients displayed significant alterations in gut microbiota and fecal metabolites, which correlated closely with disease severity and inflammatory status. These findings provide new insights into cirrhosis pathogenesis and suggest potential biomarkers for early diagnosis and disease monitoring.

Keywords: 16S rRNA sequencing; cirrhosis; gut microbiota; hepatitis B virus; metabolomics.

Figure 1

Research design and flow chart.

Figure 2

Gut microbiota diversity and composition in healthy controls (HC) and patients with HBV-related liver cirrhosis (HBV-LC). (a, b) Chao 1 and Shannon diversity indices comparing HC and HBV-LC; (c) Principal Coordinate Analysis (PCoA) plot based on weighted UniFrac distances, showing significant clustering differences between HC and HBV-LC groups;(permutation test, P = 0.001). (d, e) Relative abundance of bacterial taxa at the phylum level (*P < 0.05, **P < 0.01, ***P < 0.001). (f) Differential abundance of gut microbiota at the genus level.

Figure 3

Genus-level differences in gut microbiota between healthy controls (HC) and patients with HBV-related liver cirrhosis (HBV-LC). (a) Comparison of significantly altered bacterial genera between HC and HBV-LC groups (*P < 0.05, **P < 0.01, ***P < 0.001). (b) Linear discriminant analysis (LDA) highlighting bacterial communities with LDA scores > 4.0.

Figure 4

Correlation between bacterial genera abundance and clinical parameters in patients with HBV-related liver cirrhosis. Heatmaps display Spearman’s correlations between bacterial genera and clinical parameters. Red indicates positive correlations, while blue represents negative correlations (*P < 0.05, **P < 0.01).

Figure 5

Patients with cirrhosis were grouped according to a MELD score lower or higher/equal to 21. (a, b) Chao 1 and Shannon indices in patients with MELD < 21 and ≥ 21; (c) Negative correlation between the Shannon index and MELD scores using Spearman’s analysis; (d) PCoA plot based on weighted UniFrac distances, demonstrating significant clustering differences between the two groups (permutation test, P = 0.001); (e, f) Relative abundance of gut microbiota at the phylum and genus levels; (g) LDA identifying microbial communities with scores > 4.0. LDA, linear discriminant analysis; MELD: Model for End-Stage Liver Disease. *P < 0.05.

Figure 6

Association of CTP grades with gut microbiota. (a, b) Chao 1 and ACE diversity indices across CTP grades A, B, and C. (c) PCoA plot based on weighted UniFrac distances showing clustering differences among the three grades (permutation test, P = 0.002); (d) Genus-level differences in microbial abundance; (e) LDA identifying microbial communities with scores > 3.0. ACE, abundance-based coverage estimator; CTP: Child-Turcotte-Pugh; LDA, linear discriminant analysis. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7

Cytokine/Chemokine expression and correlation with gut microbiota. (a) Cytokine/chemokine expression levels in healthy controls and patients with HBV-related cirrhosis. (b) Correlation analysis between intestinal bacteria and cytokine/chemokine levels in patients with cirrhosis. Heatmaps depict Spearman’s correlations, with red indicating positive correlations and blue negative correlations (*P < 0.05; **P < 0.01; ***P < 0.001).

Figure 8

Metabolite Profiling in healthy controls (HC) and patients with HBV-related cirrhosis (LC). (a) Distribution of metabolites in LC and HC groups. (b) Categorical and quantitative metabolite distributions across samples. (c) Individual plot of PCA between HC and LC groups. (d) OPLS-DA scores plot of metabolites from HC and LC groups. (e) Univariate statistical analysis of differential metabolites: red dots indicate upregulated metabolites in the LC group, and blue dots represent downregulated ones. (f) Multivariate statistical analysis of differential metabolites. The x-axis shows Pearson’s correlation coefficients, and the y-axis represents metabolite contributions to model discrimination. Green dots indicate metabolites with VIP scores > 1. (g) Boxplots of the top nine differential metabolites.

Figure 9

Metabolic pathway enrichment and correlations between gut microbiota and metabolites. (a) Pathway enrichment analysis of predicted metabolite sets, with bar lengths indicating fold enrichment and color representing p-values. (b) Correlation analysis between differential bacterial genera and metabolites, where red indicates positive correlations and blue negative correlations (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001,).