Feng J, Wang JP, Hu JR, Li P, Lv P, He HC, Cheng XW, Cao Z, Han JJ, Wang Q, Su Q, Liu LX. Multi-omics reveals the associations among the fecal metabolome, intestinal bacteria, and serum indicators in patients with hepatocellular carcinoma. World J Gastroenterol 2025; 31(15): 104996 [DOI: 10.3748/wjg.v31.i15.104996]
Corresponding Author of This Article
Li-Xin Liu, Department of Gastroenterology, The First Hospital of Shanxi Medical University, No. 85 Jiefang South Road, Yingze District, Taiyuan 030001, Shanxi Province, China. lixinliu6@hotmail.com
Research Domain of This Article
Gastroenterology & Hepatology
Article-Type of This Article
Basic Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Jing Feng, Li-Xin Liu, Department of Gastroenterology, The First Hospital of Shanxi Medical University, Taiyuan 030001, Shanxi Province, China
Jing Feng, Qiang Wang, Qian Su, Department of Infectious Diseases and Hepatology, Shanxi Provincial People’s Hospital, Affiliated to Shanxi Medical University, Taiyuan 030012, Shanxi Province, China
Jun-Ping Wang, Pin Lv, Hu-Cheng He, Zheng Cao, Jia-Jing Han, Department of Gastroenterology, Shanxi Provincial People’s Hospital, Affiliated to Shanxi Medical University, Taiyuan 030012, Shanxi Province, China
Jian-Ran Hu, Ping Li, Department of Biological Science and Technology, Jinzhong University, Jinzhong 030619, Shanxi Province, China
Xiao-Wei Cheng, Department of Interventional Therapy, Shanxi Provincial People’s Hospital, Affiliated to Shanxi Medical University, Taiyuan 030012, Shanxi Province, China
Author contributions: Feng J conceived and designed the experiments, performed the experiments, and analyzed the data; Wang JP performed the experiments, and analyzed the data; Hu JR and Li P analyzed the data and prepared figures and/or tables; Lv P, He HC, Cheng XW, Cao Z, Han JJ, Wang Q, and Su Q performed the experiments; Liu LX conceived and designed the experiments and authored or reviewed drafts of the article; All authors read and approved the final manuscript.
Supported by the Department of Science and Technology of Shanxi Province, No. 20210302124369; the Health Commission of Shanxi Province, No. 2021116; and Shanxi Administration of Traditional Chinese Medicine, No. 2024ZYY2C054.
Institutional review board statement: The study protocol was approved by the medical ethics committee of Shanxi Provincial People’s Hospital (No. 2021-45).
Institutional animal care and use committee statement: This study does not involve any animal-related experiments.
Conflict-of-interest statement: The authors declare that they have no conflict of interest.
Data sharing statement: Raw 16S rDNA gene sequencing data were deposited at NCBI Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra/PRJNA1127013). The data are available as of the date of publication.
Open Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Li-Xin Liu, Department of Gastroenterology, The First Hospital of Shanxi Medical University, No. 85 Jiefang South Road, Yingze District, Taiyuan 030001, Shanxi Province, China. lixinliu6@hotmail.com
Received: January 9, 2025 Revised: February 18, 2025 Accepted: March 24, 2025 Published online: April 21, 2025 Processing time: 100 Days and 5 Hours
Abstract
BACKGROUND
Hepatocellular carcinoma (HCC), the predominant form of primary liver cancer, is a key contributor to cancer-related deaths globally. However, HCC diagnosis solely based on blood biochemical markers lacks both sensitivity and specificity.
AIM
To investigate alterations of the fecal metabolome and intestinal bacteria and reveal the correlations among differential metabolites, distinct bacteria, and serum indicators.
METHODS
To uncover potentially effective therapeutic targets for HCC, we utilized non-targeted liquid chromatography-mass spectrometry and high-throughput DNA sequencing targeting the 16S rRNA gene. This comprehensive approach allowed us to investigate the metabolome and microbial community structure of feces samples obtained from patients with HCC. Furthermore, we conducted an analysis to assess the interplay between the fecal metabolome and intestinal bacterial population.
RESULTS
In comparison to healthy controls, a notable overlap of 161 differential metabolites and 3 enriched Kyoto Encyclopedia of Genes and Genomes pathways was observed in the HCC12 (comprising patients with stage I and II HCC) and HCC34 groups (comprising patients with stage III and IV HCC). Lachnospira, Streptococcus, and Veillonella had significant differences in abundance in patients with HCC. Notably, Streptococcus and Veillonella exhibited significant correlations with serum indicators such as alpha-fetoprotein (AFP). Meanwhile, several differential metabolites [e.g., 4-keto-2-undecylpyrroline, dihydrojasmonic acid, 1,8-heptadecadiene-4,6-diyne-3,10-diol, 9(S)-HOTrE] also exhibited significant correlations with serum indicators such as γ-glutamyl transferase, total bilirubin, AFP, aspartate aminotransferase, and albumin. Additionally, these two genera also had significant associations with differential metabolites such as 1,2-Dipentadecanoyl-rac-glycerol (15:0/20:0/0:0), arachidoyl ethanolamide, and 4-keto-2-undecylpyrroline.
CONCLUSION
Our results suggest that the metabolome of fecal samples and the composition of intestinal bacteria hold promise as potential biomarkers for HCC diagnosis.
Core Tip: This study investigated the alterations in the fecal metabolome and intestinal bacteria, and elucidated the correlations among differential metabolites, distinct bacterial taxa, and serum indicators. By employing non-targeted liquid chromatography-mass spectrometry and high-throughput DNA sequencing technologies, the researchers discovered that three Kyoto Encyclopedia of Genes and Genomes pathways, namely retinol metabolism, steroid hormone biosynthesis, and ascorbate and aldarate metabolism, were significantly enriched by differential metabolites, along with three representative bacterial genera: Streptococcus, Veillonella, and Lachnospira. Notably, Streptococcus and Veillonella exhibited evident correlations with serum indicators and differential metabolites. The findings suggest that the fecal metabolome and the composition of intestinal bacteria hold considerable potential as biomarkers for the diagnosis of hepatocellular carcinoma.
Citation: Feng J, Wang JP, Hu JR, Li P, Lv P, He HC, Cheng XW, Cao Z, Han JJ, Wang Q, Su Q, Liu LX. Multi-omics reveals the associations among the fecal metabolome, intestinal bacteria, and serum indicators in patients with hepatocellular carcinoma. World J Gastroenterol 2025; 31(15): 104996
Primary liver cancer ranked as the third most common cause of cancer-related mortality globally in 2020. Hepatocellular carcinoma (HCC) is the major form of liver cancer, accounting for 75%-85% of liver cancer cases, and it carries a 5-year survival rate lower than 7%[1,2]. To date, early diagnosis is a prerequisite for the effectiveness of various therapies (e.g., surgical resection, and radiation). To improve the prognosis of HCC, it will be important to decode the effects of key clinical and biochemical features on disease duration and treatment response. Consequently, predictive biomarkers are essential for clinical diagnosis and therapy strategy selection in HCC.
Several indicators in blood have been used clinically. For example, high levels of serum alpha-fetoprotein (AFP) are closely associated with elevated mortality, and they have been used to evaluate the risk of tumor relapse after surgical removal and liver transplantation. High levels of vascular endothelial growth factor-A and angiopoietin-2 are associated with an adverse prognosis in HCC, but neither can predict treatment response. Cytokines, such as interleukin-6, interferon alpha, and transforming growth factor-β, have garnered attention as potential predictive biomarkers of the treatment response in HCC. However, definitive conclusions regarding their efficacy, clinical relevance, and specificity in prognosticating treatment outcomes necessitate additional research endeavors and larger-scale prospective studies[3].
The exploration of potential biomarkers for early diagnosis, accurate prognosis, and precision stratification purposes is vital to reducing the disease burden of HCC. Mass spectrometry (MS) is a powerful metabolomics analysis technique that has been applied to different experimental methods and instrument platforms at different stages of biomarker discovery. Currently, MS-based metabolomics analysis has been applied to different samples (such as blood), and it has revealed a large number of HCC biomarkers[4]. In a study of the sera of 68 patients with HCC, 33 patients with liver cirrhosis, and 34 healthy controls (HCs), Li et al[5] identified five metabolites with significant alterations in abundance in HCC: Taurochenodeoxycholic acid, glycochenodeoxycholate, ouabain, theophylline, and xanthine. Based on an analysis of 52 patients with HCC and 59 HCs by liquid chromatography–MS, Liu et al[6] reported that the levels of DL-3-phenyllactic acid, L-tryptophan, glycocholic acid, and 1-methylnicotinamide were increased in tissue and portal vein serum samples in HCC, whereas the levels of linoleic acid and phenol were decreased in portal vein and stool samples. Zhang et al[7] found that the levels of 15 short-chain fatty acids such as propionate, butyrate, isobutyrate, and 2-methylbutyrate were extremely altered in the feces of patients with HCC. However, the research on the differential metabolites in feces of patients with HCC is insufficient.
The link between the gut microbiota and HCC was recently depicted by numerous researchers. High serum levels of lipopolysaccharide have been observed in patients with chronic liver disease (e.g., HCC, cirrhosis, alcoholic hepatitis)[8], suggesting potential alterations in the ecological structure of the gut microbiota. For example, probiotics can regulate gut microbiota and T-cell differentiation and then interfere with the progression of HCC[9]. Ma et al[10] reported that Bacteroides thetaiotaomicron and its metabolite acetic acid effectively disturbed the recurrence of HCC and improved clinical outcomes. Therefore, compared with HCs, there are differences in the gut microbiota structure of patients with HCC, but current research has not revealed a consistent pattern. In this study, we investigated the alterations and correlations of the fecal metabolome and gut microbiota in individuals with different stages of HCC to provide novel predictive biotargets for the selection of treatment strategies.
MATERIALS AND METHODS
Patient specimens
Stool and blood samples were gathered from patients with HCC undergoing treatment and subsequent follow-ups at Shanxi Provincial People’s Hospital (Taiyuan, Shanxi Province, China) from 2021 to 2022. The study protocol was approved by the medical ethics committee of Shanxi Provincial People’s Hospital affiliated to Shanxi Medical University (No. 2021-45). Written informed consent was obtained from all participants in this study. For the HC cohort, the specimens were procured from individuals who were deemed medically fit and who had abstained from antibiotic consumption for at least 1 month preceding their enrollment in the study. Table 1 provides an overview of the demographic details and clinical characteristics of the study participants.
Table 1 Demographic profiles and clinical features of the participants in the study, mean ± SD/n (%).
All patients and HCs underwent assessment of the following blood indicators after 12 hours of fasting: Alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), γ-glutamyl transferase (GGT), and AFP.
Untargeted metabolomic analysis
Fecal samples were analyzed using an ACQUITY UPLC I-Class PLUS System (Waters, Milford, MA, United States) by Biomarker Technologies (Beijing, China). The column used in this study was the Waters Acquity UPLC HSS T3 column (1.8 μm × 2.1 mm × 100 mm). Mobile phase A consisted of 0.1% formic acid aqueous solution, and 0.1% formic acid acetonitrile was used as mobile phase B. Primary and secondary MS data were collected in the MSe mode using a Xevo G2-XS QTof high-resolution mass spectrometer (Waters) under the control of acquisition software (MassLynx V4.2, Waters).
Following normalization of the original peak area data relative to the total peak area, subsequent analyses were performed. To assess the reproducibility within a group of samples and the performance of quality control samples, both principal component analysis (PCoA) and Spearman’s correlation analysis were employed. The classified compounds were queried for their pathway associations and categorization details within the Kyoto Encyclopedia of Genes and Genomes (KEGG), Human Metabolome Database, and Lipid Maps repositories. Utilizing the grouping information, the fold changes (FCs) were computed and compared, and a t-test was employed to assess the statistical significance (P value) of the differences observed for each compound. R software facilitated orthogonal partial least squares discriminant analysis (OPLS-DA) modeling and validated model reliability via 200 permutation tests. Employing multiple cross-validation techniques, we computed the variable importance in projection (VIP) score for the model. Differential metabolites were screened using a combined approach incorporating FCs, P values, and VIP scores from the OPLS-DA model. KEGG pathway enrichment significance for differential metabolites was assessed via a hypergeometric distribution test.
Fecal microbiota analysis
DNA was extracted from fecal samples. The V3-V4 region of 16S rRNA gene was amplified using the following primers: F, 5’-ACTCCTACGGGAGGCAGCA-3’; and R, 5’-GGACTACHVGGGTWTCTAAT-3’. Following polymerase chain reaction amplification with sequencing adapters, the products underwent purification, quantification, and homogenization to create a sequencing library. This library, once verified, was subjected to sequencing on the Illumina Novaseq 6000 platform by Biomarker Technologies. The raw data were analyzed using the base-calling method, transformed into the original sequenced reads, and stored in the FASTQ file format. Clean reads were obtained after filtration of the raw reads by Trimmomatic v0.33 and the removal of primer sequences by cutadapt 1.9.1. To obtain non-chimeric reads, the clean reads were concatenated using Usearch v10 and denoised, and chimeric sequences were removed using the dada2 method in QIIME2 2020.6. Utilizing species annotation and abundance analysis, deeper exploration was conducted into the microbial community’s composition, α- and β-diversity indices, dominant bacterial taxa, and functional prediction capabilities. The raw 16S rDNA gene sequencing data have been submitted to the NCBI Sequence Read Archive for access and further analysis (https://www.ncbi.nlm.nih.gov/sra/PRJNA1127013).
Correlation analysis
Correlation analysis between fecal differential metabolites and dominant bacterial taxa or serum indicators was performed using Spearman’s correlation analysis. Visualized correlations are denoted by red (positive) and blue (negative) hues, with an ellipse shape indicating the strength of the absolute correlation.
Statistical analysis
Statistical evaluations were conducted in R software (The R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism 5 (GraphPad Software, Boston, MA, United States). To discern differential bacterial taxa across groups, permutational multivariate analysis of variance was applied. For multiple comparisons, false discovery rate-adjusted P values were used, with P < 0.05 denoting statistical significance and P < 0.01 indicating high significance.
RESULTS
Clinical characteristics of the study population
Our cohort of 38 patients included 32 men (84.21%). Twenty-three patients (60.53%) were older than 60 years of age, and 37 patients (97.37%) had cirrhosis. The primary risk factor was hepatitis B virus infection in 30 patients (78.95%). According to the tumor node metastasis staging criteria, 7 (18.42%), 11 (28.95%), 11 (28.95%), and 9 patients (23.68%) had stage I, II, III, and IV HCC, respectively. The mean serum levels of ALT, AST, albumin (ALB), alkaline phosphatase (ALP), GGT, TBIL, and AFP are listed in Table 1, and the detailed values of each sample are presented in Supplementary Table 1. In contrast to the HC group, patients with HCC exhibited markedly elevated serum concentrations of AST, ALB, GGT, TBIL, and AFP. Notably, the mean ALP level in patients with stage I-II HCC was higher than that in HCs, but the difference was not significant. However, substantial differences in ALP levels were noted among patients with HCC. In addition, the serum level of AFP in patients with HCC was remarkably increased (P < 0.05), and its levels were significantly higher in advanced HCC stages (III and IV) than in earlier stages (I and II, P < 0.05). Receiver operating characteristic (ROC) curve analysis indicated that body mass index, ALT, ALB, ALP, GGT, TBIL, and AFP were significantly associated with the occurrence of HCC. However, ALT [HC vs HCC12 (comprising patients with stage I and II HCC), area under the curve (AUC) = 0.6550, P = 0.1074; HC vs HCC34 (comprising patients with stage III and IV HCC), AUC = 0.5684, P = 0.4651] were not associated with HCC (Figure 1A and B, Table 2). Furthermore, only 13 patients with HCC (34.21%) had AFP levels exceeding 400 ng/mL (Supplementary Table 1). Therefore, the sensitivity and specialty of serum indicators in the diagnosis of HCC appeared limited.
Figure 1 Receiver operating characteristic analysis of serum indicators and alteration analysis of fecal metabolic profiling.
A and B: Receiver operating characteristic curves of healthy control (HC) vs comprising patients with stage I and II hepatocellular carcinoma (HCC12) (A) and HC vs comprising patients with stage III and IV hepatocellular carcinoma (HCC34) (B); C: Orthogonal partial least squares discriminant analysis score plots of HC vs HCC12, HC vs HCC34, and HCC12 vs HCC34; D: Clustered heatmap of the differential metabolites in each group; E: Venn diagram of the differential metabolites in each comparison. BMI: Body mass index; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; ALB: Albumin; ALP: Alkaline phosphatase; GGT: γ-glutamyl transferase; TBIL: Total bilirubin; AFP: Alpha-fetoprotein; OPLS-DA: Orthogonal partial least squares discriminant analysis; HC: Healthy control; HCC: Hepatocellular carcinoma; HCC12: Comprising patients with stage I and II hepatocellular carcinoma; HCC34: Comprising patients with stage III and IV hepatocellular carcinoma.
Table 2 Receiver operating characteristic analysis of serum indicators.
Characteristic
HC vs HCC12
HC vs HCC34
AUC
P value
AUC
P value
BMI
0.7047
0.0335
0.7000
0.0328
ALT
0.6550
0.1074
0.5684
0.4651
AST
0.8816
< 0.0001
0.9842
< 0.0001
ALB
0.9737
< 0.0001
1
< 0.0001
ALP
0.6170
0.2242
0.8132
0.0008
GGT
0.9064
< 0.0001
0.9289
< 0.0001
TBIL
0.8553
0.0002
0.9263
< 0.0001
AFP
0.9474
< 0.0001
0.9763
< 0.0001
Differential metabolites and correlations with serum indicators
The overall differences in the three comparisons are displayed in Figure 1C. R2Y (HC vs HCC12, 0.943; HC vs HCC34, 0.937; HCC12 vs HCC34, 0.995) and Q2Y (HC vs HCC12, 0.596; HC vs HCC34, 0.42; HCC12 vs HCC34, 0.32) suggested that the models were reliable.
The screening criteria for differentially accumulated metabolites comprised P ≤ 0.05 and VIP ≥ 1. The levels of the metabolites in each sample were displayed as a heatmap in Figure 1D and as volcano plots in Figure 2A-C. As presented in Table 2, 3351 metabolites were detected. In comparison with HCs, the levels of 209 and 118 metabolites were increased in HCC12 and HCC34, respectively, whereas 177 and 297, respectively, were lower (Figure 1E and Table 3). In total, 161 differential metabolites were shared in the two comparisons (HC vs HCC12 and HC vs HCC34), and seven of them were also significantly altered in the comparison of HCC12 vs HCC34 (Figure 1E), including cholylglutamic acid, beta-D-gentiobiosyl crocetin, tryptophyl-proline, pisumionoside, PS [16:0/20:5 (6E,8Z,11Z,14Z,17Z)-OH (5)], 5-hydroxyindoleacetic acid, and lithocholate 3-O-glucuronide. Additionally, ROC analysis was conducted, with Supplementary Figure 1 (HC vs HCC12) and Supplementary Figure 2 (HC vs HCC34) depicting the ROC curves of the top 10 metabolites by AUC.
Figure 2 Volcano plots of the differential metabolites and bubble plots of the Kyoto Encyclopedia of Genes and Genomes pathways.
A-C: The differential metabolites of healthy control (HC) vs comprising patients with stage I and II hepatocellular carcinoma (HCC12) (A), HC vs comprising patients with stage III and IV hepatocellular carcinoma (HCC34) (B), and HCC12 vs HCC34 (C) were displayed as volcano plots; D and E: Kyoto Encyclopedia of Genes and Genomes enrichment analysis was separately performed on the MetaboAnalyst website (https://www.metaboanalyst.ca/) for the differential metabolites in the two comparisons, namely HC vs HCC12 (D) and HC vs HCC34 (E). HCC: Hepatocellular carcinoma; HCC12: Comprising patients with stage I and II hepatocellular carcinoma; HCC34: Comprising patients with stage III and IV hepatocellular carcinoma; VIP: Variable importance in projection; FC: Fold change.
Table 3 Numbers of differentially accumulated metabolites.
Group
Total
Differential metabolites
Up
Down
HC vs HCC34
3351
386
209
177
HC vs HCC12
3351
415
118
297
Table 4 Correlations between differential metabolites and serum indicators.
To further evaluate the biological significance of differentially accumulated metabolites, we investigated their correlations with serum indicators by Spearman’s correlation analysis. The serum indicators and differential metabolites with the strongest correlation are listed in Table 4. AST and ALB were significantly correlated with the differential metabolites in HCC34. The AST level was positively correlated with those of 25-acetyl-6,7-didehydrofevicordin F 3-glucoside and coumatetralyl, whereas the ALB level was positively correlated with those of cholylglutamic acid, propamocarb, and 3b-hydroxy-6b-angeloyloxy-7 (11)-eremophilen-12,8b-olide and negatively correlated with those of 25-acetyl-6,7-didehydrofevicordin F 3-glucoside and 5-ribosylparomamine. Interestingly, we found that AFP was negatively correlated with six differential metabolites [(1S)-1-hydroxy-23,24-didehydro-25,26,27-trinorcalciol, dihydrojasmonic acid, 3-nor-3-oxopanasinsan-6-ol, (6R,7S)-6,7-epoxy-1,3-tetradecadiyne, and 1,8-heptadecadiene-4,6-diyne-3,10-diol] in HCC12 and positively correlated with vamorolone. However, in HCC34, no differential metabolites were significantly correlated with the AFP level. GGT and TBIL were correlated with different metabolites in the HCC12 and HCC34 groups. Concerning other blood indicators, we did not identify differential metabolites significantly correlated with their levels. Notably, six metabolites [(1S)-1-hydroxy-23,24-didehydro-25,26,27-trinorcalciol, dihydrojasmonic acid, vamorolone, 3-nor-3-oxopanasinsan-6-ol, (6R,7S)-6,7-epoxy-1,3-tetradecadiyne, and 1,8-heptadecadiene-4,6-diyne-3,10-diol] were significantly correlated with both TBIL and AFP levels in HCC12 compared with HCs, and 25-acetyl-6,7-didehydrofevicordin F 3-glucoside was related to both AST and ALB in HCC34.
KEGG pathway enrichment analysis
To investigate the critical metabolic pathways altered in patients with HCC, the differentially accumulated metabolites were mapped to the KEEG database. According to the P values, number of metabolites involved, and pathway impact, the two most significant KEGG pathways in the HC vs HCC12 comparison were retinol metabolism (P < 0.01, impact value = 0.47938) and steroid hormone biosynthesis (P < 0.05, impact value = 0.15113), followed by terpenoid backbone biosynthesis (P = 0.24036, impact value = 0.18571) and glycerophospholipid metabolism (P = 0.10058, impact value = 0.15634; Figure 2D and Supplementary Table 2). Retinol metabolism (P < 0.05, impact value = 0.26289) and steroid hormone biosynthesis (P < 0.01, impact value = 0.17926) were also significantly enriched in the HC vs HCC34 comparison. Meanwhile, ascorbate and aldarate metabolism (P < 0.01, impact value = 0) was observed (Figure 2E and Supplementary Table 2). These results implied that retinol metabolism, steroid hormone biosynthesis, and ascorbate and aldarate metabolism might be critical to the pathogenesis and development of HCC.
Alterations of the gut microbiota in patients with HCC
To explore the potential function of the gut microbiota in HCC, we investigated the alterations of the ecological structure of the intestinal bacteria. α-diversity was evaluated using four indices, namely abundance-based coverage estimator, Chao1, Simpson, and Shannon (Figure 3A), and the results revealed no significant changes in richness and evenness. However, PCoA identified significant differences among the HC, HCC12, and HCC34 groups, as only a portion of the samples in the three groups overlapped (Figure 3B). Furthermore, the results of OPLS-DA also suggested that these three groups were separated from each other (Figure 3C), indicating that the community compositions of intestinal bacteria were extremely changed in patients with HCC compared with those in healthy individuals. As illustrated in Figure 3B and C, unweighted pair-group method with arithmetic means revealed a clear separation of the HC group from the HCC12 or HCC34 group (Figure 3D).
Figure 3 Alterations of the ecological structure of the intestinal bacteria in patients with hepatocellular carcinoma.
A: Assessment of α-diversity within fecal samples sourced from patients with hepatocellular carcinoma; B: Visualization of unweighted UniFrac principal component analysis scores through a scatter plot; C: Scatter plot of orthogonal partial least squares discriminant analysis of the metabolic profiling of each group; D: Construction of a hierarchical clustering dendrogram utilizing the unweighted pair group method with arithmetic mean. PCoA: Principal component analysis; HCC12: Comprising patients with stage I and II hepatocellular carcinoma; HCC34: Comprising patients with stage III and IV hepatocellular carcinoma; ACE: Abundance-based coverage estimator; PLS-DA: Partial least squares discriminant analysis; HC: Healthy control.
The dominant distinct bacterial taxa and their correlations with serum indicators
To identify the primary differential bacterial taxa in patients with HCC, an evaluation of the relative abundance of all bacterial phyla was conducted, with the top 10 most abundant phyla presented in Figure 4A. Of them, Firmicutes and Bacteroidetes were the two most abundant bacterial phyla. We further compared the relative abundance of various bacterial phyla among the three groups (HC, HCC12, and HCC34), and no distinct phyla were found (Figure 4B).
Figure 4 The phylum-level composition of the gut microbiota in patients with hepatocellular carcinoma.
A: The 10 phyla with the highest abundance in each sample; B: Variance analysis of the 10 most abundant bacterial phyla among the groups. HC: Healthy control; HCC: Hepatocellular carcinoma; HCC12: Comprising patients with stage I and II hepatocellular carcinoma; HCC34: Comprising patients with stage III and IV hepatocellular carcinoma.
Then, the relative abundance of bacterial genera was analyzed in each sample, and the top 10 genera are presented in Figure 5A. The proportions of these 10 bacterial genera varied among the samples. Therefore, the dominant taxa in each group were analyzed using linear discriminant analysis (LDA) effect size. The hierarchical relationship of the main communities in each group is displayed in a cladogram in Figure 5B. Three bacterial communities, namely Lachnospira, unclassified Lachnospira species, and Selenomonadaceae, were enriched in the HC group. Four taxa, including Proteobacteria, Gammaproteobacteria, Enterobacterales, and Enterobacteriaceae, were dominant in the HCC12 group. The communities enriched in the HCC34 group were unclassified Rothia species, Micrococcales, Rothia, Micrococcaceae, unclassified Veillonella species, Veillonella, unclassified Streptococcus species, Streptococcaceae, Streptococcus, Lactobacillales, and Bacilli. The LDA score was also utilized to depict the results as a bar chart, as presented in Figure 5C. Notably, three representative bacterial genera were observed, of which the relative abundance of Lachnospira was extremely lower in the HCC12 and HCC34 groups, whereas that of Streptococcus and Veillonella was remarkably increased in these groups (Figure 5D).
Figure 5 Screening of the dominant differential bacterial communities in patients with hepatocellular carcinoma.
A: The top 10 most dominant bacterial genera and their proportional presence in the three groups; B: Linear discriminant analysis effect size; C: Linear discriminant analysis score; D: Three representative bacterial genera with significant intergroup differences. The median is indicated by the dashed line, whereas the solid line denotes the mean. HCC: Hepatocellular carcinoma; HC: Healthy control; HCC12: Comprising patients with stage I and II hepatocellular carcinoma; HCC34: Comprising patients with stage III and IV hepatocellular carcinoma; LDA: Linear discriminant analysis.
Furthermore, the correlations between intestinal bacteria and serum indicators were investigated. Spearman’s correlation analysis was used to analyze the correlations between the top 20 differentially abundant bacterial genera and serum indicators in each comparison (Figure 6). In the HC vs HCC12 comparison, seven bacterial genera, namely Streptococcus, Veillonella, Parabacteroides, Lachnoclostridium, Lachnospira, Roseburia, and Subdoligranulum, were significantly correlated with at least one serum indicator. In the HC vs HCC34 comparison, six genera were identified, specifically Lachnoclostridium, Lachnospira, Streptococcus, Rothia, Veillonella, and Haemophilus. The common genera were Streptococcus, Veillonella, Lachnoclostridium, and Lachnospira. Of these, the abundance of Streptococcus was positively correlated with the levels of five serum indicators, including GGT, AST, ALP, TBIL, and AFP, in both comparisons and negatively correlated with the serum level of ALB in the HC vs HCC34 comparison. Veillonella had significant positive correlations with TBIL (HC vs HCC12 and HC vs HCC34), AFP (HC vs HCC12 and HC vs HCC34), GGT (HC vs HCC34), and AST levels (HC vs HCC34) and a negative correlation with ALB levels (HC vs HCC34). Lachnoclostridium exhibited negative correlations with AST and AFP levels in both comparisons. Lachnospira was negatively associated with TBIL, AFP, and GGT levels in the HC vs HCC12 comparison. Meanwhile, in the HC vs HCC34 comparison, the genus was and negatively correlated with ALP, AST, AFP, GGT, and TBIL levels and positively correlated with ALB levels.
Figure 6 Correlation analysis of intestinal microbiota and serum indicators.
The heatmap presents the correlations between the top 20 differentially abundant bacterial genera and serum indicators. A: In the comparisons healthy control (HC) vs comprising patients with stage I and II hepatocellular carcinoma (HCC12); B: HC vs comprising patients with stage III and IV hepatocellular carcinoma (HCC34). aP < 0.05. bP < 0.01. cP < 0.001. ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; ALB: Albumin; ALP: Alkaline phosphatase; GGT: γ-glutamyl transferase; TBIL: Total bilirubin; AFP: Alpha-fetoprotein.
Correlation analysis of the metabolome and intestinal bacteria in patients with HCC
The correlation between the fecal metabolome and intestinal bacteria was analyzed using Spearman’s rank correlation coefficient. The top 10 bacteria in terms of the absolute value of Log2FC were selected. Differential bacterial phyla (genera) were screened using analysis of variance with the criterion of P < 0.05. In the HC vs HCC12 comparison, four differentially abundant bacterial phyla, namely Firmicutes, Proteobacteria, Fusobacteriota, and Desulfobacterota, were associated with metabolites including 4-keto-2-undecylpyrroline, retinoic acid, and flurandrenolide (Figure 7A, upper panel). Five genera (Klebsiella, Agathobacter, Escherichia Shigella, Roseburia, and Veillonella) were significantly correlated with 4-keto-2-undecylpyrroline (Klebsiella and Agathobacter, P < 0.05; Veillonella, P < 0.01), allo-hydroxycitric acid lactone (P < 0.05), and flurandrenolide levels (P < 0.05, Figure 7A, lower panel). Of these, Klebsiella and Escherichia Shigella belong to Proteobacteria, and Agathobacter, Roseburia, and Veillonella all belong to Firmicutes. In the HC vs HCC34 comparison, three genera were associated with the metabolites flurbiprofen glucuronide and DG (15:0/20:0/0:0) (Figure 7B, upper panel), and five genera had significant correlations with eight metabolites (Figure 7B, lower panel). Notably, Streptococcus and Megamonas both belong to Firmicutes. Streptococcus exhibited positive association with the levels of DG (15:0/20:0/0:0) (P < 0.05), and Megamonas was positively correlated with PZ-peptide (P < 0.05) and glycyrrhizic acid levels (P < 0.01). In the HCC12 vs HCC34 comparison, three phyla displayed significant correlations with four metabolites (Figure 7C, upper panel). Interestingly, arachidoyl ethanolamide had a significant correlation with Streptococcus (P < 0.01, Figure 7C, lower panel). Meanwhile, Bacteroides was associated with 5-hydroxyferulate (P < 0.05) and carteolol levels (P < 0.01). Combined with the correlation analysis with serum indicators (Figure 6), Streptococcus and Veillonella appeared to be most important in the onset and development of HCC.
Figure 7 Heatmap of the correlations between differential metabolites and bacterial genera in fecal samples.
A-C: Correlation analysis of the differential metabolites and bacterial genera based on comparisons including healthy control (HC) vs comprising patients with stage I and II hepatocellular carcinoma (HCC12) (A), HC vs comprising patients with stage III and IV hepatocellular carcinoma (HCC34) (B), and HCC12 vs HCC34 (C). Correlation analysis was conducted between differential metabolites and bacteria at the phylum (upper panel) and genus level (lower panel).
DISCUSSION
Blood-based biomarkers are commonly used indicators for diagnosing liver diseases. AFP is the conventional blood diagnostic marker for HCC. The normal AFP concentration ranges from 0 to 40 ng/mL, and it is significantly elevated in the blood of 60%-70% of patients with HCC. Meanwhile, approximately 30% of patients with HCC exhibit negative AFP levels clinically[11]. Additionally, AFP levels can be changed in response to liver injury and some other liver diseases (such as steatohepatitis, fatty liver, and chronic or active hepatitis)[12], resulting in false-positive diagnostic results. Thus, the specificity and sensitivity of AFP as a diagnostic indicator for HCC are limited. To reduce misdiagnosis, additional blood indicators need to be comprehensively evaluated clinically, such as ALT and AST. However, alterations in the levels of these indicators are usually related to various factors. For example, in addition to HCC, ALT is also associated with long-term alcohol consumption, fatty liver disease, and the use of certain liver-damaging drugs (such as isoniazid tablets and rifampicin tablets)[13]. AST is one of the biomarkers of myocardial injury, and its abnormal elevation is also closely related to glucose and lipid metabolism disorders[14]. ALB is associated with kidney disease and decreased immune function[15]. Elevated ALP might be caused by jaundice, rickets, leukemia, and other diseases[16]. The significant increase in GGT levels in serum might be related to pancreatic head cancer, cirrhosis, and other diseases[17]. Abnormally elevated TBIL levels can be suggestive of conditions including hemolysis and gallstones[18]. We analyzed blood indicators in 38 patients with HCC, with 13 patients (3 in the HCC12 group and 10 in the HCC34 group) having AFP levels exceeding 400 ng/mL. Interestingly, among patients with AFP levels lower than 40 ng/mL, 19 had normal serum levels, including 13 and 6 in the HCC12 and HCC34 groups, respectively (Tables 1 and Supplementary Table 1). Although six other blood indicators were also measured in this study, their diagnostic value for HCC was limited. In addition to blood indicators, liver ultrasonography is also used clinically to screen patients with HCC. However, because of the limited specificity and sensitivity of ultrasonography, early liver cancer lesions might not be detected promptly, and false-negative results cannot be avoided. Therefore, new and more effective biomarkers urgently need to be developed.
Metabolomics studies of feces, biofluids, and tissues have been considered powerful methods for describing metabolic features during the onset of HCC. A study involving 262 patients with HCC identified three metabolites (taurocholic acid, lysophosphoethanolamine, and lysophosphatidylcholine) with significantly increased levels in the sera of patients with HCC[19]. Jee et al[20] found five serum metabolites (leucine, phenylalanine, tyrosine, arachidonic acid, and 5-hydroxyhexanoic acid) were highly correlated with the occurrence of HCC among 37 significantly differential metabolites using an ultra-performance liquid chromatography-MS method. Compared with body fluid samples, the metabolites enriched in feces are partly sourced from the gut microbiota, potentially reflecting the pathogenesis and progression of HCC and the interplay between the gut microbiota and disease progression[21]. Of note, the fecal metabolome largely reflects the community structure of the gut microbiota, and it is considered a functional indicator of the gut microbiota[22]. Research by Liu et al[6] uncovered that the concentrations of linoleic acid and phenol in feces and hepatic portal vein blood samples from patients with HCC were notably diminished in comparison to those found in HCs. Furthermore, linoleic acid and phenol exhibit suppressive effects on the viability and colony formation of HCC cells, specifically Hep3B and Huh7 cells, while inducing apoptosis in these cells. However, these compounds did not exert similar effects on normal hepatocyte cells, such as MIHA cells[6]. In this study, we identified seven low-abundance metabolites in the HCC12 group, including three (retinol, retinoate, and 9-cis-retinoic acid) involved in retinol metabolism and four (dehydroepiandrosterone, dehydroepiandrosterone sulfate, androstenediol, and androstenedione) involved in steroid hormone biosynthesis (Figure 2D). Two differential metabolites with reduced levels in the HCC34 group, (retinol and 9-cis-retinoic acid) are involved in retinol metabolism, other six metabolites with decreased levels (dehydroepiandrosterone, dehydroepiandrosterone sulfate, androstenediol, 17-alpha-hydroxyprogesterone, estrone, and 16-glucuronide-estriol) are involved in steroid hormone biosynthesis, and two metabolites (myo-inositol and beta-D-glucuronoside) are involved in ascorbate and aldarate metabolism (Figure 2E). Retinol metabolism has been observed to be inhibited in patients with metabolic dysfunction-associated fatty liver disease[23,24], and they have critical roles in the regulation of metabolic disorders attributable to carbon tetrachloride-induced liver fibrosis[25]. Furthermore, retinaldehyde storage was extremely reduced in patients with HCC, and retinol metabolism might be a potential diagnostic and prognostic marker in HCC[26]. Steroid hormones play crucial roles in multiple metabolic pathways, and the liver is responsible for maintaining their homeostasis[27]. Steroid hormone imbalance has been linked to multiple liver disorders. For instance, hypoestrogenism can trigger the onset and progression of non-alcoholic fatty liver disease in post-menopausal females[28]. Deficiency of testosterone, the primary androgen, might contribute to the development of sarcopenia in male patients with liver cirrhosis[29]. Therefore, inhibition of steroid hormone biosynthesis might lead to HCC. Ascorbate was demonstrated to affect the proliferation and viability of tumor cells[30]. It has been reported that ascorbate and aldarate metabolism is associated with the development of metabolic-associated steatotic liver disease (MASLD)[31]. Notably, several studies also reported associations of retinol metabolism or steroid hormone biosynthesis with the gut microbiota. Han et al[32] found that the interaction between the gut microbiota and retinol metabolism could modulate white adipose tissue accumulation and obesity. A recent study demonstrated that during liver regeneration, the gut microbiota and systemic metabolism undergo extensive alterations and exhibit a strong correlation. Among the gut microbes, Escherichia Shigella, Lactobacillus, Akkermansia, and Muribaculaceae emerged as the representative differential bacterial taxa. Meanwhile, steroid hormone biosynthesis was the most prominent metabolic pathway[33]. However, few reports examined the relationship between ascorbate and aldarate metabolism and the gut microbiota. In the present study, we observed that both retinol metabolism and steroid hormone biosynthesis were suppressed in patients with stage I-IV HCC (the HCC12 and HCC34 groups). Additionally, ascorbate and aldarate metabolism was dysregulated in patients with stage III-IV HCC (the HCC34 group). Meanwhile, we also identified several gut bacterial genera, including Streptococcus, Veillonella, and Lachnospira, with significant alterations. These findings imply potential associations of the gut microbiota with the aforementioned metabolic pathways.
Differential metabolites have emerged as potential biomarkers in a variety of diseases. For example, 2-methyl-1-propylamine and estrone display excellent diagnostic capabilities in the early stage of cervical intraepithelial neoplasia[34]. Palmitoylcarnitine and sphingosine might be potential biomarkers of colorectal cancer[35]. Nineteen serum metabolites related to energy metabolism, lipid metabolism, amino acid metabolism, and citric acid metabolism were considered potential biomarkers of pancreatic ductal adenocarcinoma[36]. Eight differential metabolites were identified as alternative markers for predicting the formation of gastroesophageal cancer[37]. In combination with AUC analysis (Supplementary Figure 1 and 2), the correlations of differential metabolites with serum indicators (Table 4) demonstrated that several metabolites were likely biomarkers for different stages of HCC. We further investigated the associations between these metabolites and liver physiological activities by reviewing the literature and searching relevant databases. For example, 25-acetyl-6,7-didehydrofevicordin F 3-glucoside is an intermediate metabolite mainly generated from saponin, and it is usually associated with lipid peroxidation, fatty acid metabolism, cell signaling, and lipid metabolism. Cholylglutamic acid is a conjugate that mainly consists of an amino acid and a primary bile acid such as cholic acid, deoxycholic acid, and chenodeoxycholic acid[38]. Quinn et al[39] reported that these bile acid-amino acid conjugates are produced by gut microbes such as Clostridia species, and they were frequently found in patients with inflammatory bowel disease and cystic fibrosis and in infants. They appeared to activate the farnesoid X receptor and inhibit the expression of bile acid synthesis genes[39]. Trans-2-enoyl-CoA was found to be involved in the sphingosine-to-glycerolipid metabolic pathway in mammals[40], and it also participates in the fatty acid elongation cycle[41]. Vitamin A (retinol), a yellow, fat-soluble antioxidant, is vital for embryogenesis, vision, cell function, immune regulation, and metabolism. Liver diseases causing fibrosis and cirrhosis disrupt vitamin A homeostasis, often leading to deficiency. Upon liver injury, hepatic stellate cells (HSCs) are transformed into myofibroblasts. These cells overproduce extracellular matrix, causing fibrosis. As HSCs deplete retinyl ester stores during this change, vitamin A deficiency ensues. Non-alcoholic fatty liver disease, a liver-related aspect of metabolic syndrome, spans from mild steatosis to steatohepatitis, and it can progress to cirrhosis and liver cancer[42]. 9-Cis-retinoic acid is the main active form of vitamin A, and it can bind to nuclear retinoic acid receptors and retinoid X receptors. 9-Cis-retinoic acid is typically used in the combinational treatment of breast, liver, and gastric cancers[43]. 3,4-Dimethyl-5-pentyl-2-furanpropanoic acid is a furan fatty acid that might be produced by gut microbes, and it is considered a healthy biomarker[44]. 4-keto-2-undecylpyrroline is a pyrroline which was mainly produced by genus Streptococcus[45]. 3-Nor-3-oxopanasinsan-6-ol is considered a potential biomarker for the progression of intestinal pathology in mice infected with Schistosoma[46]. Interestingly, hydroxyversicolorone, a natural product newly isolated from a blocked mutant of Aspergillus parasiticus, has been synthesized in labeled form and incorporated intact into aflatoxin B1 by mycelial suspensions of wild-type Aspergillus parasiticus[47]. 2-(Ethylamino) ethanol is not a naturally occurring metabolite, and it is only found in individuals exposed to this compound or its derivatives. This compound has been identified in human blood[48]. However, the biological functions of some of the differential metabolites listed in Table 4 have not been reported.
The total number of bacterial cells residing in the human body is approximately 3.9 × 1013, and most of them survive in the gastrointestinal tract[49]. During the evolutionary process, the gut microbiota and host have co-evolved, establishing a mutually beneficial relationship[50]. The profound connection between the gut microbiota (e.g., intestinal bacteria) and liver function affects the onset and progression of HCC. Investigations revealed that the gut microbiota of individuals with HCC undergoes alterations in its composition, with the abundance of more than 30 distinct types of intestinal microorganisms potentially changing during the early stages of the disease. This underscores the pivotal role of the gut microbiota in the initiation and progression of HCC across various stages starting from a benign liver condition[51]. Metabolites generated by the gut microbiota, along with pro-inflammatory molecules secreted by intestinal epithelial cells, have the capacity to traverse the portal vein and enter the liver. Once there, they can contribute to the development and progression of liver cancer[52]. A plethora of evidence indicates that the gut–liver axis affects the progression of liver diseases, namely fibrosis, cirrhosis, and cancer. For example, Klebsiella pneumoniae is related to MASLD in humans, and it can also cause MASLD in mice. The abundance of Bacteroides species and Ruminococcaceae was much higher in patients with HCC than in healthy individuals, whereas the opposite trend was noted for the abundance of Bifidobacterium[53]. A study revealed that some intestinal microorganisms (e.g., Clostridium) transformed primary bile acids into deoxycholic acid, which stimulate liver cells to release inflammatory factors and tumor-promoting factors through enterohepatic circulation. HCC occurred once the host was exposed to chemical carcinogens [e.g., 7, 12-dimethylbenz(a)anthracene]. Inhibition of deoxycholic acid-producing intestinal bacteria by antibiotic therapy effectively reduced the risk of HCC in carcinogen-induced obese mice[54]. Therefore, the community structure of intestinal bacteria affects the pathogenesis of HCC. In this study, three representative differential bacterial genera, namely Lachnospira, Streptococcus, and Veillonella, were identified in the HCC12 and HCC34 groups. Lachnospira was reported to exhibit similar temporal patterns and causative relationships with Akkermansia and Bifidobacterium in HCC[55], and we found that the abundance of Lachnospira was extremely decreased in the HCC12 and HCC34 groups (Figure 5D). The relative abundance of Streptococcus was positively associated with AFP levels in the sera of older patients with HCC[56]. Our findings also confirmed these results (Figure 6). Streptococcus is an anaerobic bacterium, and it has been considered a diagnostic marker for gastrointestinal cancer. For example, a significant increase in the abundance of Streptococcus might lead to gastric[57] or colon cancer[58,59]. Streptococcus also appeared to serve as a potential biomarker of liver metastasis in gastric cancer[60]. Wu et al[61] believed that gut microbiota-driven metabolic alterations might affect cachexia in gastric cancer, and Streptococcus was one of the main genera with an increase in abundance. In the patients with HCC, Streptococcus and Veillonella both exhibited constant increases in abundance in fecal samples during disease progression[62]. They were identified as two cirrhosis-related genera, and they were significantly correlated with the metabolic profile[63]. Zheng et al[64] revealed that Streptococcus and Veillonella also contributed to the early recurrence of HCC, and the potential mechanism was that gut microbe-derived acetic acid could provide the energy for tumor cell proliferation. A study conducted in Thailand reported that the abundance of Veillonella was significantly higher in patients with intrahepatic cholangiocarcinoma than in those with HCC[65]. Our findings indicated that Streptococcus was positively correlated with DG (15:0/20:0/0:0) levels and negatively correlated with enoxolone and leucine-serine-histidine-asparticacid levels (HCC12 group, Figure 6). Enoxolone exerts cytotoxic effects on human gingival fibroblasts and strong inhibitory effects on bacterial growth[66], and the combination of enoxolone and silibinin can prevent prostate cancer cell growth in mice[67]. In the HCC12 vs HCC34 comparison, Streptococcus had a positive correlation with arachidoyl ethanolamide, which displayed cannabinoid-like effects in mice[68] and a clear decrease in content in plasma from an Alzheimer’s disease mouse model[69] but no association with HCC. Additionally, Lachnospira was positively associated with the levels of flurandrenolide (HC vs HCC12) and five other metabolites, including leucine-serine-histidine-asparticacid, glycyrrhizic acid, PZ-peptide, hexadecasphingosine, and flurbiprofen glucuronide (Figure 7). Based on existing research reports, glycyrrhizic acid exerts hepatoprotective effects[70] by inhibiting high-mobility group box 1[71]. PZ-peptide was reported to trigger the opening of tight junctions and enhance the transport of paracellular solutes by stimulating transepithelial sodium ion flux across the colonic segments, resulting in uptake by hepatocytes[72]. Therefore, these three bacterial genera, along with their associated metabolites, have the potential to serve as distinctive biomarkers for screening and diagnosing alterations in the gut microbiota of patients with HCC. In the context of HCC onset and progression, the fluctuations in patients’ gut bacterial community architecture and fecal metabolomic landscapes are exceedingly intricate. Diverse bacterial communities yield a rich spectrum of secondary metabolites. These metabolites engage in a web of complex interactions and impose disparate effects on the host. Furthermore, patients’ dietary patterns and pharmaceutical regimens can significantly disrupt their liver metabolic processes and reshape the composition of their gut microbiota. In light of these complexities centered around HCC, more thorough and meticulous research endeavors are imperative to elucidate the precise role of the gut microbiota in the development and advancement of HCC.
CONCLUSION
In summary, this study revealed that the sensitivity and specificity of serum indicators were insufficient in the diagnosis of HCC. Compared with the findings in the HC group, 161 overlapped differential metabolites and 3 identical enriched KEGG metabolic pathways were identified in the HCC12 and HCC34 groups. Lachnospira, Streptococcus, and Veillonella were the representative differential bacterial genera in the feces of patients with HCC. Streptococcus and Veillonella were both significantly correlated with serum indicators. Several differential metabolites [e.g., 4-keto-2-undecylpyrroline, dihydrojasmonic acid, 1,8-heptadecadiene-4,6-diyne-3,10-diol, 9(S)-HOTrE], which displayed significant correlations with serum indicators such as GGT, TBIL, AFP, AST, and ALB, might be potential biomarkers of HCC. Furthermore, Streptococcus and Veillonella also had significant associations with differential metabolites such as DG (15:0/20:0/0:0), arachidoyl ethanolamide, and 4-keto-2-undecylpyrroline. Our findings suggested that the fecal metabolome and gut microbiota might be potential targets for the diagnosis of HCC.
ACKNOWLEDGEMENTS
We sincerely thank all patients and individuals for their participation.
Footnotes
Provenance and peer review: Unsolicited article; Externally peer reviewed.
Peer-review model: Single blind
Specialty type: Gastroenterology and hepatology
Country of origin: China
Peer-review report’s classification
Scientific Quality: Grade A, Grade B
Novelty: Grade A, Grade B
Creativity or Innovation: Grade B, Grade B
Scientific Significance: Grade A, Grade B
P-Reviewer: Qian XK; Wang C S-Editor: Fan M L-Editor: A P-Editor: Zhao S
Li X, Yi Y, Wu T, Chen N, Gu X, Xiang L, Jiang Z, Li J, Jin H. Integrated microbiome and metabolome analysis reveals the interaction between intestinal flora and serum metabolites as potential biomarkers in hepatocellular carcinoma patients.Front Cell Infect Microbiol. 2023;13:1170748.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in RCA: 6][Reference Citation Analysis (0)]
Liu J, Geng W, Sun H, Liu C, Huang F, Cao J, Xia L, Zhao H, Zhai J, Li Q, Zhang X, Kuang M, Shen S, Xia Q, Wong VW, Yu J. Integrative metabolomic characterisation identifies altered portal vein serum metabolome contributing to human hepatocellular carcinoma.Gut. 2022;71:1203-1213.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited in This Article: ][Cited by in Crossref: 30][Cited by in RCA: 50][Article Influence: 16.7][Reference Citation Analysis (0)]
Lin RS, Lee FY, Lee SD, Tsai YT, Lin HC, Lu RH, Hsu WC, Huang CC, Wang SS, Lo KJ. Endotoxemia in patients with chronic liver diseases: relationship to severity of liver diseases, presence of esophageal varices, and hyperdynamic circulation.J Hepatol. 1995;22:165-172.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 258][Cited by in RCA: 255][Article Influence: 8.5][Reference Citation Analysis (0)]
Ma H, Yang L, Liang Y, Liu F, Hu J, Zhang R, Li Y, Yuan L, Feng F. B. thetaiotaomicron-derived acetic acid modulate immune microenvironment and tumor growth in hepatocellular carcinoma.Gut Microbes. 2024;16:2297846.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 6][Reference Citation Analysis (0)]
Gil-Rojas S, Suárez M, Martínez-Blanco P, Torres AM, Martínez-García N, Blasco P, Torralba M, Mateo J. Application of Machine Learning Techniques to Assess Alpha-Fetoprotein at Diagnosis of Hepatocellular Carcinoma.Int J Mol Sci. 2024;25:1996.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Xiao G, Yang J, Yan L. Comparison of diagnostic accuracy of aspartate aminotransferase to platelet ratio index and fibrosis-4 index for detecting liver fibrosis in adult patients with chronic hepatitis B virus infection: a systemic review and meta-analysis.Hepatology. 2015;61:292-302.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 338][Cited by in RCA: 378][Article Influence: 37.8][Reference Citation Analysis (0)]
Ghosh T, Bhadra K. A mini review on human serum albumin - natural alkaloids interaction and its role as drug carrier.J Biomol Struct Dyn. 2024;1-18.
[PubMed] [DOI] [Full Text][Cited in This Article: ]
Jiang C, Hu F, Li J, Gao G, Guo X. Diagnostic value of alkaline phosphatase and bone-specific alkaline phosphatase for metastases in breast cancer: a systematic review and meta-analysis.Breast Cancer Res Treat. 2023;202:233-244.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Nano J, Muka T, Cepeda M, Voortman T, Dhana K, Brahimaj A, Dehghan A, Franco OH. Association of circulating total bilirubin with the metabolic syndrome and type 2 diabetes: A systematic review and meta-analysis of observational evidence.Diabetes Metab. 2016;42:389-397.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 36][Cited by in RCA: 48][Article Influence: 5.3][Reference Citation Analysis (0)]
Tan Y, Yin P, Tang L, Xing W, Huang Q, Cao D, Zhao X, Wang W, Lu X, Xu Z, Wang H, Xu G. Metabolomics study of stepwise hepatocarcinogenesis from the model rats to patients: potential biomarkers effective for small hepatocellular carcinoma diagnosis.Mol Cell Proteomics. 2012;11:M111.010694.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 92][Cited by in RCA: 116][Article Influence: 8.3][Reference Citation Analysis (0)]
Zhao L, Ni Y, Su M, Li H, Dong F, Chen W, Wei R, Zhang L, Guiraud SP, Martin FP, Rajani C, Xie G, Jia W. High Throughput and Quantitative Measurement of Microbial Metabolome by Gas Chromatography/Mass Spectrometry Using Automated Alkyl Chloroformate Derivatization.Anal Chem. 2017;89:5565-5577.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 76][Cited by in RCA: 102][Article Influence: 12.8][Reference Citation Analysis (0)]
Ashla AA, Hoshikawa Y, Tsuchiya H, Hashiguchi K, Enjoji M, Nakamuta M, Taketomi A, Maehara Y, Shomori K, Kurimasa A, Hisatome I, Ito H, Shiota G. Genetic analysis of expression profile involved in retinoid metabolism in non-alcoholic fatty liver disease.Hepatol Res. 2010;40:594-604.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 54][Cited by in RCA: 54][Article Influence: 3.6][Reference Citation Analysis (0)]
Sanookpan K, Chantaravisoot N, Kalpongnukul N, Chuenjit C, Wattanathamsan O, Shoaib S, Chanvorachote P, Buranasudja V. Pharmacological Ascorbate Elicits Anti-Cancer Activities against Non-Small Cell Lung Cancer through Hydrogen-Peroxide-Induced-DNA-Damage.Antioxidants (Basel). 2023;12.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 2][Reference Citation Analysis (0)]
Han H, Zhang S, Wang M, Yi B, Zhao Y, Schroyen M, Zhang H. Retinol metabolism signaling participates in microbiota-regulated fat deposition in obese mice.J Nutr Biochem. 2025;136:109787.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Sun R, Fei F, Jin D, Yang H, Xu Z, Cao B, Li J. The integrated analysis of gut microbiota and metabolome revealed steroid hormone biosynthesis is a critical pathway in liver regeneration after 2/3 partial hepatectomy.Front Pharmacol. 2024;15:1407401.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Wang X, Guan X, Tong Y, Liang Y, Huang Z, Wen M, Luo J, Chen H, Yang S, She Z, Wei Z, Zhou Y, Qi Y, Zhu P, Nong Y, Zhang Q. UHPLC-HRMS-based Multiomics to Explore the Potential Mechanisms and Biomarkers for Colorectal Cancer.BMC Cancer. 2024;24:644.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Teng T, Shi H, Fan Y, Guo P, Zhang J, Qiu X, Feng J, Huang H. Metabolic responses to the occurrence and chemotherapy of pancreatic cancer: biomarker identification and prognosis prediction.Sci Rep. 2024;14:6938.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Che J, Zhao Y, Gu B, Li S, Li Y, Pan K, Sun T, Han X, Lv J, Zhang S, Fan B, Li C, Wang C, Wang J, Zhang T. Untargeted serum metabolomics reveals potential biomarkers and metabolic pathways associated with the progression of gastroesophageal cancer.BMC Cancer. 2023;23:1238.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in RCA: 3][Reference Citation Analysis (0)]
Quinn RA, Melnik AV, Vrbanac A, Fu T, Patras KA, Christy MP, Bodai Z, Belda-Ferre P, Tripathi A, Chung LK, Downes M, Welch RD, Quinn M, Humphrey G, Panitchpakdi M, Weldon KC, Aksenov A, da Silva R, Avila-Pacheco J, Clish C, Bae S, Mallick H, Franzosa EA, Lloyd-Price J, Bussell R, Thron T, Nelson AT, Wang M, Leszczynski E, Vargas F, Gauglitz JM, Meehan MJ, Gentry E, Arthur TD, Komor AC, Poulsen O, Boland BS, Chang JT, Sandborn WJ, Lim M, Garg N, Lumeng JC, Xavier RJ, Kazmierczak BI, Jain R, Egan M, Rhee KE, Ferguson D, Raffatellu M, Vlamakis H, Haddad GG, Siegel D, Huttenhower C, Mazmanian SK, Evans RM, Nizet V, Knight R, Dorrestein PC. Global chemical effects of the microbiome include new bile-acid conjugations.Nature. 2020;579:123-129.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 373][Cited by in RCA: 365][Article Influence: 73.0][Reference Citation Analysis (0)]
Kato R, Takenaka Y, Ohno Y, Kihara A. Catalytic mechanism of trans-2-enoyl-CoA reductases in the fatty acid elongation cycle and its cooperative action with fatty acid elongases.J Biol Chem. 2024;300:105656.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Lu Y, Liu D, Jiang R, Li Z, Gao X. Prodigiosin: unveiling the crimson wonder - a comprehensive journey from diverse bioactivity to synthesis and yield enhancement.Front Microbiol. 2024;15:1412776.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Chienwichai P, Tipthara P, Tarning J, Limpanont Y, Chusongsang P, Chusongsang Y, Kiangkoo N, Adisakwattana P, Reamtong O. Identification of trans-genus biomarkers for early diagnosis of intestinal schistosomiasis and progression of gut pathology in a mouse model using metabolomics.PLoS Negl Trop Dis. 2024;18:e0011966.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Yuan J, Chen C, Cui J, Lu J, Yan C, Wei X, Zhao X, Li N, Li S, Xue G, Cheng W, Li B, Li H, Lin W, Tian C, Zhao J, Han J, An D, Zhang Q, Wei H, Zheng M, Ma X, Li W, Chen X, Zhang Z, Zeng H, Ying S, Wu J, Yang R, Liu D. Fatty Liver Disease Caused by High-Alcohol-Producing Klebsiella pneumoniae.Cell Metab. 2019;30:1172.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 38][Cited by in RCA: 62][Article Influence: 10.3][Reference Citation Analysis (0)]
Yoshimoto S, Loo TM, Atarashi K, Kanda H, Sato S, Oyadomari S, Iwakura Y, Oshima K, Morita H, Hattori M, Honda K, Ishikawa Y, Hara E, Ohtani N. Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome.Nature. 2013;499:97-101.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 1318][Cited by in RCA: 1613][Article Influence: 134.4][Reference Citation Analysis (0)]
Ponziani FR, De Luca A, Picca A, Marzetti E, Petito V, Del Chierico F, Reddel S, Paroni Sterbini F, Sanguinetti M, Putignani L, Gasbarrini A, Pompili M. Gut Dysbiosis and Fecal Calprotectin Predict Response to Immune Checkpoint Inhibitors in Patients With Hepatocellular Carcinoma.Hepatol Commun. 2022;6:1492-1501.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited in This Article: ][Cited by in Crossref: 6][Cited by in RCA: 28][Article Influence: 9.3][Reference Citation Analysis (0)]
Zhou CB, Pan SY, Jin P, Deng JW, Xue JH, Ma XY, Xie YH, Cao H, Liu Q, Xie WF, Zou XP, Sheng JQ, Wang BM, Wang H, Ren JL, Liu SD, Sun YW, Meng XJ, Zhao G, Chen JX, Cui Y, Wang PQ, Guo HM, Yang L, Chen X, Ding J, Yang XN, Wang XK, Qian AH, Hou LD, Wang Z, Chen YX, Fang JY. Fecal Signatures of Streptococcus anginosus and Streptococcus constellatus for Noninvasive Screening and Early Warning of Gastric Cancer.Gastroenterology. 2022;162:1933-1947.e18.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 8][Cited by in RCA: 42][Article Influence: 14.0][Reference Citation Analysis (0)]
Wu J, Zhang R, Yin Z, Chen X, Mao R, Zheng X, Yuan M, Li H, Lu Y, Liu S, Gao X, Sun Q. Gut microbiota-driven metabolic alterations reveal the distinct pathogenicity of chemotherapy-induced cachexia in gastric cancer.Pharmacol Res. 2024;209:107476.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Sharma SP, Gupta H, Kwon GH, Lee SY, Song SH, Kim JS, Park JH, Kim MJ, Yang DH, Park H, Won SM, Jeong JJ, Oh KK, Eom JA, Lee KJ, Yoon SJ, Ham YL, Baik GH, Kim DJ, Suk KT. Gut microbiome and metabolome signatures in liver cirrhosis-related complications.Clin Mol Hepatol. 2024;30:845-862.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 3][Reference Citation Analysis (0)]
Zheng C, Lu F, Chen B, Yang J, Yu H, Wang D, Xie H, Chen K, Xie Y, Li J, Bo Z, Wang Y, Chen G, Deng T. Gut microbiome as a biomarker for predicting early recurrence of HBV-related hepatocellular carcinoma.Cancer Sci. 2023;114:4717-4731.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in RCA: 4][Reference Citation Analysis (0)]
Friedman CA, Saha A, Lavender Hackman G, Lu X, Lodi A, Tiziani S, DiGiovanni J. Novel two-tiered screening approach identifies synergistic combinations of natural compounds for prostate cancer prevention and treatment.Mol Carcinog. 2024;63:589-600.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
Watanabe K, Matsunaga T, Nakamura S, Kimura T, Ho IK, Yoshimura H, Yamamoto I. Pharmacological effects in mice of anandamide and its related fatty acid ethanolamides, and enhancement of cataleptogenic effect of anandamide by phenylmethylsulfonyl fluoride.Biol Pharm Bull. 1999;22:366-370.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Cited by in Crossref: 20][Cited by in RCA: 22][Article Influence: 0.8][Reference Citation Analysis (0)]
Zhou L, Chen S, Wei Y, Sun Y, Yang Y, Lin B, Li Y, Wang C. Glycyrrhizic acid restores the downregulated hepatic ACE2 signaling in the attenuation of mouse steatohepatitis.Eur J Pharmacol. 2024;967:176365.
[RCA] [PubMed] [DOI] [Full Text][Cited in This Article: ][Reference Citation Analysis (0)]
