This study included 23 HCV post-SVR patients and 25 patients without liver disease who underwent total colonoscopy at our hospital between, 2019 and, 2021. Tables 1 and 2 summarize the clinical characteristics of the patients. The LC group included patients with stage F4 fibrosis on liver biopsy at the time of treatment and those with shear wave velocity (SWV) > 1.86 m/s (LOGIQ S8, GE Healthcare, Milwaukee, WI, USA) in whom biopsy was not performed. The following potential enrollees were excluded: minors, patients with gastrointestinal infections, individuals with a predisposition to bleeding, patients with concurrent malignant tumors, including HCC, habitual alcohol consumption, and those with a history of antibiotic usage within the past two months ( Table 3 ). All participants provided written informed consent. This study was conducted in accordance with the Declaration of Helsinki. The study design was approved by the Research Ethics Committee of the University of Fukui (registration number: 20190167).

Total colonoscopy was performed following the usual colonoscopy preparation using polyethylene glycol. The mucosal surface was gently scraped using a sterile cytology brush (TeleMed Systems Inc., MA, USA) and for each respective case, samples were collected from terminal ileum, ascending colon, and sigmoid colon. Additionally, stool samples were collected on the day of examination. In cases with obtained consent, mucosal and liver biopsy tissues were collected during HCV treatment and preserved in RNA later (Thermo Fisher Scientific, MA, USA) at −20°C.

Double-stranded DNA (dsDNA) was extracted from each sample using the NucleoSpin DNA Stool Kit (MACHEREY-NAGEL, Düren, Germany). The measurements of DNA concentration were performed using Qubit dsDNA HS assay (Thermo Fisher Scientific), and 5 ng of DNA was used for the first PCR targeting the V3-V4 region of the 16S rRNA gene with Universal primers 341F (5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′) and 805R (5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC-3′) (Thermo Fisher Scientific). The PCR products were confirmed using electrophoresis, purified using AMPure XP beads (Beckman Coulter, CA, USA), and subjected to index PCR using the Nextera XT index kit (Illumina, San Diego, CA, USA). Subsequent bead purification was conducted and the final library concentration was diluted to 4 nM for normalization. The DNA was denatured by adjusting the final denatured DNA concentration to 6 pM using 25% PhiX spike-in. Sequencing was performed using a MiSeq Reagent v2 kit with a 250-bp paired-end strategy according to the manufacturer’s instructions (Illumina) (Muyzer et al., 1993; Inoue et al., 2016).

Paired-end FASTQ files obtained from sequencing were uploaded from the Basespace hub, an Illumina cloud database, to the Ezbiocloud (ChunLab, Inc., Seoul, Republic of Korea). Using the EzBioCloud pipeline, all raw sequencing data were re-analyzed for each 16S-based microbial taxonomic profile (MTP). Microbiome profiling was conducted using the MTP platform of the application, with the 16S database version PKSSU.4.0 (Yoon et al., 2017). The MTP sets were constructed by grouping the individual MTPs. The MTP sets were comparatively analyzed after normalizing the gene copy numbers. Beta (β) diversity was evaluated using Jensen–Shannon divergence and generalized UniFrac distances, and the results were visualized using principal coordinate analysis (PCoA). We subsequently performed taxonomic biomarker analysis based on linear discriminant analysis (LDA) effect size (LEfSe) to identify differentially distributed taxa between MTP sets that could potentially be used as microbial biomarkers. Taxa that demonstrated significant results in the relative abundance (p < 0.05) and LEfSe analysis (LDA score > 3.00) were considered significant biomarkers (Kim et al., 2023).

The EzBioCloud 16S-based MTP pipeline employed the phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt) algorithm to estimate the functional profiles of the microbiome identified using 16S rRNA sequencing. Raw sequencing reads were processed using the EzBioCloud 16S microbiome pipeline with default parameters and discriminating reads encountered in the reference database. Functional abundance profiles of MAM were annotated using bioinformatics analyses. This specifically involved multiplying the vector of gene counts for each operational taxonomic unit (OTU) by the abundance of that OTU in each sample, using the Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology and pathway database. The predicted metagenomic profiles were categorized into clusters of KEGG pathways and compared between the post-SVR CH and LC groups (Ferreira et al., 2018; Tango et al., 2020).

Intestinal mucosa or liver biopsy tissues were homogenized using Micro Smash (TOMY, Tokyo, Japan), and total RNA was extracted using an RNeasy Mini Kit (Qiagen, Hilden, Germany). The cDNA was synthesized from RNA using a High Capacity RNA-to-cDNA Kit (Applied Biosystems, Foster City, CA, USA) after quantification and purity assessment using a Nanodrop (Thermo Fisher). Quantitative real-time PCR (qRT-PCR) was conducted using a StepOnePlus real-time PCR system (Applied Biosystems) with a TaqMan probe ( Table 4 ) (Thermo Fisher). The ΔΔCt comparative threshold method was used for analyzing the expression levels of the target genes. GAPDH was considered an internal control.

Permutational multivariate analysis of variance (PERMANOVA) and the Kruskal–Wallis test were used to analyze the statistical differences in beta diversity and for functional analysis, respectively. Both analyses were performed using Ezbiocloud. The Mann–Whitney and Spearman’s correlation tests were conducted using Prism (version 9; GraphPad, GraphPad Software Inc.). Statistical significance was set at p-values < 0.05.

Microbiota composition was analyzed in the MAM of the terminal ileum, ascending and sigmoid colon, and feces of patients infected with HCV who achieved an SVR. We compared the data between patients with non-cirrhotic (chronic hepatitis; CH) and cirrhotic (liver cirrhosis; LC) liver diseases and controls. Comparing patients with CH and LC revealed no significant differences between the groups at either sampling site at the phylum level by the Mann–Whitney U test ( Figure 1A ). At the genus level, patients with LC exhibited a distinctly different composition from patients with CH and the control group in MAM in the terminal ileum, ascending colon, and sigmoid colon ( Figure 1B ; Table 5 ). These results indicate that microbiota dysbiosis was observed at the genus level in the MAM of post-SVR LC patients.

In the HCV post-SVR and the control group, microbial β-diversity at the genus level for MAM at each sampling site and feces was conducted using Principal Coordinates Analysis (PCoA) based on Jensen-Shannon divergence. PERMANOVA test revealed a substantial difference in the β-diversity of MAM of the feces and the terminal ileum, ascending colon, and sigmoid colon in the HCV post-SVR group ( Figure 2A ; Table 6 ). In the control group, the β-diversity of MAM in the terminal ileum was different from that in feces. The β-diversity of MAM in the ascending and sigmoid colon exhibited significant differences in the HCV post-SVR and control groups for each sampling site ( Figure 2B ; Table 7 ). These results indicate that the microbial β-diversity of MAM in the post-HCV SVR group, especially in the ascending colon and sigmoid colon, differed from that in the control group.

We performed a PERMANOVA test for each MAM collection site (terminal ileum, ascending, and sigmoid colon) to evaluate the CLD-related factors associated with the microbial β-diversity of MAM ( Figure 3A ). Factors related to liver fibrosis, such as type IV collagen 7S, M2BPGi, and biopsy (F1-3/F4, CH/LC), were significantly correlated in the MAM of the ascending colon ( Figures 3A, B ). The association between CLD-related factors and MAM in the terminal ileum or the sigmoid colon was statistically insignificant. Thus, the β-diversity of MAM in the ascending colon is associated with liver fibrosis progression in HCV post-SVR patients.

We performed a functional analysis of predictive metagenomics using PICRUSt in HCV post-SVR patients to evaluate the function of MAM in the ascending colon in liver fibrosis. We compared the relative proportion levels associated with the KEGG pathway in the MAM of the ascending colon in the HCV SVR-CH and LC groups ( Figures 4A, B ). Cell adhesion molecules, tight junctions, focal adhesions, and primary bile acid biosynthesis were the pathways that were proportionally downregulated in the LC group. Pathways related to the intestinal barrier function (cell adhesion molecules, tight junctions, and focal adhesion) and bile acid production (primary bile acid biosynthesis and bile secretion) were significantly decreased in the MAM of the HCV SVR-LC group compared to those of the control and HCV SVR-CH groups ( Figure 4C ). Thus, MAM in the ascending colon in the HCV post-SVR LC group may have impaired intestinal barrier and bile acid production.

We evaluated the molecular expression of bile acid signaling in intestinal tissues and fibrosis-related genes in the hepatic tissues of the HCV post-SVR LC and CH groups. FGF19 expression, a downstream signaling molecule of bile acids and farnesoid X receptor (FXR), was markedly decreased in the terminal ileum biopsy tissue of the SVR LC group compared to that in the CH group ( Figure 5A ). In the liver tissue, IL-8 and TGF-β1 expression, cytokines associated with the progression of liver fibrosis, were markedly increased in the SVR LC group compared to that in the CH group ( Figure 5B ). These findings suggest that the SVR LC group exhibits decreased expression of the intestinal bile acid-related signaling molecules and increased expression of the liver fibrosis-related cytokines.

We compared the differences in the relative composition at the taxonomic level in the MAM of the ascending colon between the HCV post-SVR LC and CH groups using LEfSe. The genera Blautia (p=0.002, LDA score 4.52), Alistipes (p=0.022, LDA score 3.75), Roseburia (p=0.013, LDA score 3.67), Agathobaculum (p=0.003, LDA score 3.44), Dorea (p=0.031, LDA score 3.42), and Pseudoflavonifractor (p=0.022, LDA score 3.36) were significantly decreased in the MAM of ascending colon in HCV post-SVR LC group ( Figure 6 ). All six genera were short-chain fatty acid (SCFA)-producing bacteria.

We analyzed the correlation between the genera composition of MAM in the ascending colon and liver fibrosis markers, molecular expression in the intestinal and liver tissues, and the function of metagenomics in HCV post-SVR patients ( Figure 7 ). Six genera (Blautia, Alistipes, Roseburia, Agathobaculum, Dorea, and Pseudoflavonifractor) of MAM in the ascending colon were negatively correlated with the degree of fibrosis of liver tissue and serum liver fibrosis markers, such as type IV collagen-7S and M2BPGi, compared with the other dominant genera. Additionally, they positively correlated with the functional gene sets related to the intestinal barrier and bile production as well as FGF19 expression in the terminal ileum tissue. Conversely, they exhibited a negative correlation with IL-8 and TGF-β1 expression in the liver tissue. Thus, in post-HCV SVR cases, the composition of the six genera in the MAM of the ascending colon is associated with the hepatic fibrosis degree and bacterial flora function related to liver fibrosis.