Alterations in the gut microbiota in individuals with infantile cholestasis: a comparison of high- and low-γ-glutamyltransferase subtypes

Introduction

Infantile cholestasis (IC) is a pathophysiological condition stemming from hepatocyte/biliary tract disorders causing obstruction of bile excretion, with an incidence rate of approximately 1/2,500 (1). Severe cases can lead to cirrhosis and even death. Currently, in addition to treatment of the underlying causes of IC, ursodeoxycholic acid is administered to alleviate symptoms to a certain extent. However, more targeted treatment options are lacking for infants with different types of cholestasis (1).

Multiple studies have revealed dysbiosis in the gut microbiota of IC patients (2,3). Studies have assessed the use of probiotics as adjunctive therapies for IC (4,5), but the lack of personalized probiotic selection strategies has led to suboptimal treatment outcomes. A more in-depth study of the unique mechanisms by which the gut microbiota promotes IC is essential for the development of more precise treatment strategies. Metabolites released in response to gut microbiota disturbances can enter the liver through the mesenteric venous system and directly or indirectly participate in the regulation of hepatocyte inflammation and the immune response, thus affecting liver function (2), including by causing IC. The interaction between gut microbiota and metabolites provides a new perspective and possible therapeutic angle for diagnosing, managing, and preventing liver disorders. Although studies (2,3,6) have explored the correlations of gut microbiota with fecal metabolites and fecal bile acid profiles, few significant breakthroughs have been made. Limited research has explored the associations between gut microbiota and metabolite profiles in blood and urine within the context of IC.

There are two subtypes of cholestasis, and the subtype is determined on the basis of the serum γ-glutamyltransferase (GGT) level, namely, high-GGT (HG) (7,8) levels and low-GGT (LG) levels (9-11). These two subtypes of cholestasis may have different pathogeneses. IC can be classified into intrahepatic cholestasis and extrahepatic cholestasis (biliary obstructive cholestasis) (12). In intrahepatic cholestasis, bile becomes trapped within the liver cells due to an inability to expel it, often indicating an underlying hereditary liver disease (11). Whereas biliary obstructive cholestasis occurs when bile, having already exited the liver cells, cannot be excreted into the intestines due to an obstruction in the biliary tract, frequently associated with biliary duct malformations. GGT is synthesized by liver cells and binds to the canalicular membrane via a glycosyl phosphatidyl inositol anchor. Bile acids, acting as detergents, normally liberate GGT (9). In cases of intrahepatic cholestasis, the retention of bile within liver cells hinders the liberation of GGT, leading to LG levels. Conversely, in biliary obstructive cholestasis, bile can flow out of the liver and liberate GGT. This liberated GGT then enters the bloodstream through the damaged bile duct wall, resulting in elevated GGT levels. Therefore, HG levels in individuals with cholestasis frequently suggest an extrahepatic etiology, whereas low GGT levels often indicate intrahepatic cholestasis. However, this biochemical correlation does not constitute an absolute diagnostic criterion. Given that the high GGT and low GGT subtypes of IC may have different pathogeneses, the gut microbiota may play distinct roles in these two groups. Thus, a thorough investigation into the specific manifestations of gut microbiota in different types of IC may be crucial for elucidating disease mechanisms and developing novel treatment strategies.

Therefore, the aim of this study was to explore the variations in gut microbiota composition between infants with cholestasis and healthy infants and the correlations between these microbial changes and metabolite profiles in HG and LG subtypes of IC. Our research is expected to provide new insights into the relationships between gut microbiota and different subtypes of IC. We present this article in accordance with the STROBE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-183/rc).

Methods

Study participants

We gathered data for infants with cholestasis (the Cho group) admitted to Fujian Children’s Hospital between January 2023 and June 2024 as well as a control group (the Con group) comprising healthy infants who underwent physical examination at the hospital in the same period. Cholestasis usually occurs within 6 months after birth, and the introduction of supplementary foods can influence the gut microbial composition. Consequently, our study population comprised individuals under the age of 6 months. The inclusion criteria for cholestasis (13) were defined as a conjugated or direct serum bilirubin level >17 mmol/L (1 mg/dL) when the total bilirubin was <85.5 µmol/L (5 mg/dL) or >20% of the total bilirubin if the total bilirubin was ≥85.5 mmol/L. The inclusion criteria for the Cho group were as follows: patents (I) aged <6 months (II) who met all the established diagnostic criteria for cholestasis and (III) who did not receive antibiotics or probiotics in the 14 days prior to the study. Among the infants with cholestasis, after multiple liver function tests, the GGT level of each test in the HG group was greater than 100 U/L, and the GGT level of each test in the LG group was less than 100 U/L (10). The inclusion criteria for the Con group were as follows: (I) individuals aged <6 months who were deemed healthy on the basis of their physical exams, without yellowing of the skin or sclera to the naked eye, and whose stool color was golden yellow without any additional conditions such as cholestasis or congenital growth issues; (II) multiple noninvasive transcutaneous bilirubin determinations via a jaundice monitor at the time of enrollment that revealed no jaundice; and (III) who did not receive any antibiotics or probiotics in the 14 days prior to the study. The exclusion criteria were as follows: (I) infants whose mothers had diabetes, high blood pressure, or chronic liver disease during pregnancy; (II) infants whose mothers were continually exposed to drugs or probiotics during pregnancy or lactation; or (III) infants who had allergic diseases (e.g., food allergy, eczema, and allergic gastroenteritis). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics committee of Fujian Children’s Hospital (No. 2023ETKLR05086), and the guardians of all participants provided written informed consent. We have completed dual registration in both the Chinese Clinical Trial Registry (www.chictr.org.cn, ChiCTR2300072387) and the National Medical Research Registration & Filing Information System (www.medicalresearch.org.cn, MR-35-24-056613).

Collection of clinical data

Fecal samples were collected from all participants on admission or at the time of visit and were kept frozen at −80 ℃, and the samples were collectively dispatched for analysis. Blood and urine samples were collected from the Cho group for measuring clinical indicators such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), total serum bilirubin (Tbil), direct bilirubin (Dbil), indirect bilirubin (Ibil) and total bile acid (TBA) levels. Owing to ethical considerations, no blood samples were obtained from the Con group.

Fecal microbiota DNA extraction and 16S rRNA sequencing

The bacterial sample genomic DNA was isolated via the Fast DNA SPIN Kit supplied by MP Biomedicals (Santa Ana, CA, USA). Polymerase chain reaction (PCR) was conducted using the 338F (ACTCCTACGGGAGGCAGCA) forward primer and the 806R (GGACTACHVGGGTWTCTAAT) reverse primer. The PCR amplification products were purified via the E.Z.N.A.^® Gel Extraction Kit (Omega, USA). Library construction was subsequently carried out according to the standard protocol of the NEBNext^® Ultra™ II DNA Library Prep Kit for Illumina^® (New England Biolabs, USA). The assembled amplicon library was analyzed via paired-end 250-bp sequencing on an Illumina Nova 6000 sequencer.

Tandem mass spectrometry (MS/MS) and gas chromatography-MS (GC-MS) analysis

In the Cho group, fasting peripheral blood was collected and dripped onto special blood-sampling filter paper for screening, with a blood spot diameter ≥8 mm. Each filter paper was subsequently punched into a U-96-well plate to prepare the DBS cards. The samples were processed with a non-derivatized MS/MS kit from PerkinElmer (Waltham, USA), bearing the product code 3040-0010, and then analyzed with an API3200 MS/MS system (Concord, Canada). The freshly collected urine was poured onto three specialized filter papers of 3 cm × 6 cm, which were allowed to dry naturally and then dissolved in 3 mL of distilled water. The mixture was subsequently centrifuged to retrieve the urine. The urine samples were processed with urease, hydroxylamine hydrochloride, sodium hydroxide, and hydrochloric acid to eliminate urea and proteins, with 17-alkanoic acid serving as the internal control. The samples were subsequently re-extracted with acetoacetate, desiccated with nitrogen, and then mixed with a derivatization agent composed of trifluoroacetamide and trimethylchlorosilane. After derivatization, the samples were detected via a GS/MS system (QP 2010, Kyoto, Japan).

Bioinformatics and statistical analysis

The dada2 algorithm (QIIME2 version 2020.11.0) (14) was employed for denoising the sequencing data. Operational taxonomic units (OTUs) were consolidated via a 97% minimum similarity criterion through the application of UPARSE software, specifically version 7.0.1090, to group similar taxonomic entities. The clustering process adhered to the standard parameters set within the UPARSE OTU analysis pipeline. The usearch-sintax tool was utilized for sequencing analysis of the clustered OTU sequences. The OTU sequences were compared against the Silva database (15) (version 132) to achieve taxonomic annotation, thereby revealing the species composition of the microbial community. All the statistical analyses in this study were completed via Wekemo Bioincloud (https://www.bioincloud.tech) (16). To assess alpha diversity, the Chao1, Richness, Shannon and Simpson indices were calculated. For beta diversity analysis, we used principal coordinate analysis (PCoA) and nonmetric multidimensional scaling (NMDS) with Bray-Curtis distance. Distinct microbiota features were pinpointed using the linear discriminant analysis (LDA) effect size (LEfSe) method, with alpha and effect size thresholds of 0.05 and 2.5, respectively. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was utilized for pathway enrichment via PICRUSt. Orthogonal projections to latent structure discriminant analysis (OPLS-DA), accompanied by a variable importance in projection (VIP) plot (17), was executed to pinpoint potential discrepancies in metabolites between the HG and LG groups. Genera with VIP scores exceeding 1.5 and P values below 0.05 were considered important contributors to the model. Spearman’s rank correlation coefficients were calculated to assess the relationships between metabolites and clinical indicators. Categorical variables were described as proportions and were compared using the χ² test, whereas continuous variables were presented as the means, standard deviations, medians, and interquartile ranges and were compared via Student’s t-test or the Mann-Whitney U (or Kruskal-Wallis) test for continuous variables, depending on the distributions and variances of the data. P<0.05 was considered to indicate statistical significance.

Results

Demographic and clinical characteristics of the subjects

Sixty subjects were included in the study, including 37 infants in the Cho group (18 males and 19 females; median age: 2.10 months) and 23 infants in the Con group (11 males and 12 females; median age: 1.50 months); the screening process for the Cho group is shown in Figure S1. No significant differences were observed between the groups in terms of age, sex, height, weight or feeding. The demographic characteristics of the subjects are shown in Table 1. Among the 37 infants with cholestasis, 21 infants were in the HG group (10 males and 11 females; median age: 2.10 months; median GGT: 254.0 U/L), including 5 infants diagnosed with biliary atresia, 4 infants with cytomegalovirus hepatitis, 4 infants with idiopathic hepatitis, 3 infants with citrin protein deficiency, 2 infants with parenteral nutrition-related cholestasis, 2 infants with Alagille syndrome, an 1 infant with a choledochal cyst, and 16 infants were in the LG group (8 males and 8 females; median age: 2.10 months; median GGT: 81.5 U/L), including 4 infants diagnosed with idiopathic hepatitis, 3 infants with familial intrahepatic cholestasis type 2, 2 infants with sodium taurocholate cotransporting polypeptide deficiency, 2 infants with familial intrahepatic cholestasis type 1, 2 infants with familial intrahepatic cholestasis type 6, 1 infant with parenteral nutrition-related cholestasis, 1 infant with USP53 deficiency, and 1 infant with arthrogryposis-renal tubular dysfunction-cholestasis syndrome. There were no significant differences in sex and age between the two groups. Compared with LG group, in addition to differences in GGT levels, infants in the HG group had lower albumin levels (P=0.02) and higher vitamin D levels (P=0.03). There was no significant difference in other demographic and clinical indicators (Table 2).

Abundance and structural changes in gut microbiota at the phylum and genus levels

The Cho and Con groups included a total of 721 OTUs, encompassing 30 phyla, 79 classes, 146 orders, 206 families, and 390 genera. The distributions of the top 5 species at the phylum level in the Cho and Con groups, including Acidobacteria, Firmicutes, Proteobacteria, Actinobacteria and Bacteroidetes, are illustrated in Figure 1A. And the distributions of the top 10 species at the genus level in the Cho and Con groups are displayed in Figure 1B.

Figure 1 Abundance and structural changes in gut microbiota at the phylum and genus levels. (A) Distribution of the top 5 species at the phylum level in the Cho and Con groups. (B) Distributions of the top 10 species at the genus level in the Cho and Con groups. (C) Distribution of the top 5 species at the phylum level in the HG and LG groups. (D) Distribution of the top 10 species at the genus level in the HG and LG groups. Cho, cholestasis; Con, control; GGT, γ-glutamyltransferase; HG, high-GGT; LG, low-GGT.

The HG and LG groups collectively harbored 630 OTUs, spanning 30 phyla, 79 classes, 145 orders, 201 families, and 331 genera. The distribution of the top 5 species at the phylum level are shown in Figure 1C, and the distributions of the top 10 species at the genus level is shown in Figure 1D.

When comparing the Cho and the Con group at the phylum level, the abundance of Acidobacteria within the Cho group was greater than that of the Con group (P<0.05), whereas the abundances of Firmicutes and Actinobacteria were lower in the Cho group than in the Con group (P<0.05, Figure 2A-2E). At the genus level, the abundances of Lactobacillus (P<0.01), Streptococcus (P<0.01), Bacteroides (P<0.05), and Lactococcus (P<0.01) were greater in the Cho group than in the Con group. Conversely, the abundance of Lachnoclostridium (P<0.05) was lower in the Cho group than in the Con group. See Figure 2F-2O for additional information.

Figure 2 Comparison of the abundances of the gut microbiota in the Cho and Con groups at the phylum and genus levels. The Mann-Whitney U test was used to compare the abundances of the gut microbiota between the Cho and Con groups. *, P<0.05, **, P<0.01, and the y-axis represents the abundance (log₂). (A-E) Violin plots showing the abundances of Acidobacteria, Actinobacteria, Bacteroidetes, Firmicutes and Proteobacteria in the Cho and Con groups at the phylum level. (F-O) Violin plots comparing the abundances of Bacteroides, Bifidobacterium, Clostridium_sensu_stricto_1, Enterococcus, Escherichia-Shigella, Lachnoclostridium, Lactobacillus, Lactococcus, Streptococcus and Veillonella in the Cho and Con groups at the genus level. Cho, cholestasis; Con, control.

When comparing the HG and the LG group at the phylum level, there was no significant difference between the two groups (Figure S2A-S2E). At the genus level, in pairwise comparisons among the HG, LG, and Con groups, no significant difference in the abundance of Bacteroides was found between the LG and Con groups (P=0.32), whereas the abundance of Bacteroides in the HG group was greater than that in both the Con group (P=0.005) and the LG group (P=0.04), as illustrated in Figure S2F. The abundances of the remaining microbial taxa did not significantly differ at the genus level, as detailed in Figure S2G-S2O.

Diversity analysis of the microbial composition

The Richness and Chao1 indices of the alpha diversity in the Con group were lower than those in the HG or LG groups, whereas no statistically significant difference was detected between the HG and LG groups (Figure 3A,3B). The assessment of community diversity utilizing the Shannon and Simpson indices did not reveal any significant difference between the Con and HG or LG groups or between the HG and LG groups (Figure S3A,S3B). Both PCoA and NMDS analyses, leveraging the Bray-Curtis distance, consistently demonstrated pronounced segregation of the Cho group from the Con group, with statistical significance [analysis of similarities (ANOSIM), P=0.01 for both PCoA and NMDS; Figure 3C and Figure S3C]. Conversely, no notable or statistically significant differences were detected between the HG and LG groups (ANOSIM, P=0.12 for PCoA; Figure 3D; P=0.14 for NMDS; Figure S3D). To illustrate the configuration of gut microbiota across the Con, HG, and LG groups, we utilized a Circos plot to visualize the data on the top 8 most abundant samples in each group at the genus level (Figure 3E). The rarefaction curves of all the samples are shown in Figure S3E.

Figure 3 Diversity analysis and a Circos plot of the microbial composition. (A,B) Alpha diversity of the gut microbiota in the Con, HG and LG groups expressed by the Richness and Chao1 indices. The x-axis represents different groups, while the y-axis represents the Richness and Chao1 indices. The Wilcoxon test was performed for pairwise comparisons; ****, P<0.0001. Beta diversity of the gut microbiota in the Cho and Con groups (C) and in the HG and LG groups (D) expressed according to PCoA based on Bray-Curtis distance. The P value was tested with ANOSIM; each point corresponds to a sample shaped and colored by diagnosis. (E) Circos plot of the top 8 most abundant samples in the Con, HG, and LG groups. The first circle: sample grouping information; the second circle: the percentage of OTU relative abundance in each sample; the third circle: OTUs and the main block of the sample are distinguished by different colors, and the scale outside the block is the absolute abundance of OTUs; the fourth circle: OTUs and the sample subblock, corresponding to the main block (the third circle), showing the abundance of OTUs in each sample and the abundance of each OTU contained in each sample; the fifth circle: the connection of the OTUs and the sample association information, corresponding to the OTUs and the sample subblock (the fourth circle). ANOSIM, analysis of similarities; Cho, cholestasis; Con, control; GGT, γ-glutamyltransferase; HG, high-GGT; LG, low-GGT; OTU, operational taxonomic unit; PCoA, principal coordinate analysis.

LEfSe analysis of gut microbiota

LEfSe was subsequently employed to precisely delineate abundance differences through LDA with a log₁₀ score threshold exceeding 2.5. In the HG group, Bacteroides, Lactococcus, Streptococcus thermophilus TH1435, Haemophilus, Parabacteroides and Subgroup_6 were characteristic bacteria, whereas Gammaproteobacteria, Streptococcus, Blautia, Pasteurellaceae, Staphylococcus, Megamonas, Helicobacter and Bacillus were characteristic bacteria in the LG group (Figure 4A).

Figure 4 LEfSe analysis of the gut microbiota and KOs. Statistical significance was determined via LEfSe, with a P value of <0.05 (Wilcoxon test) and a LDA score (log₁₀) of >2.5 considered significant. (A) Taxonomic differences in the Con, HG and LG groups. (B) Significantly different KOs were detected in the Con, HG and LG groups. Con, control; GGT, γ-glutamyltransferase; HG, high-GGT; LG, low-GGT; LDA, linear discriminant analysis; LEfSe, linear discriminant analysis effect size; KOs, functional categories.

Metabolic pathway enrichment analysis of gut microbiota

We used PICRUSt to predict the abundance of functional categories (KOs) on the basis of LEfSe (log₁₀ score >2.5). Overall, 6 significantly different KOs (ko00195, ko00523, ko00363, ko00622, ko00903 and ko03015) were detected in the HG group (P<0.05), and 11 significantly different KOs (ko00540, ko00623, ko00253, ko00130, ko00980, ko00642, ko00591, ko00020, ko00960, ko00790 and ko00830) were detected in the LG group (P<0.05, Figure 4B). L3 KEGG pathways, such as photosynthesis, bisphenol degradation and polyketide sugar unit biosynthesis, were enriched in the HG group, whereas some pathways, such as the citrate cycle, ubiquinone and other terpenoid-quinone biosynthesis, lipopolysaccharide biosynthesis, toluene degradation and retinol metabolism, were enriched in the LG group (Table S1).

Correlation analysis of gut microbiota and metabolites or clinical indicators

Blood and urine samples were taken from 37 infants with cholestasis for metabolic profiling. A total of 234 metabolites were detected, comprising 106 metabolites from blood samples and 128 metabolites derived from urine (Table S2). Utilizing the OPLS-DA model with the criteria of VIP >1.5 and P<0.05, four differentially abundant urinary metabolites were detected between the HG and LG groups, but no variations were detected among blood metabolites. For metabolites in the urine, the HG group presented higher concentrations of alkapton-3 and lower concentrations of methylfumaric acid-2,2-hhydroxyglutaric acid-3 and malic acid-3 than did the LG group.

We explored the associations between the characteristic bacteria identified via LEfSe analysis and differentially abundant metabolites between the HG and LG groups. The findings revealed that the abundance of Helicobacter was positively correlated with 2-hydroxyglutaric acid-3 levels (R=0.34, P=0.04), whereas the abundances of Subgroup_6 and Lactococcus were negatively correlated with 2-hydroxyglutaric acid-3 levels (R=−0.35, P=0.04, R=−0.34, P=0.04, respectively), and the abundance of Gammaproteobacteria was negatively correlated with malic acid-3 levels (R=−0.42, P=0.01, Figure 5A). Furthermore, we investigated the potential relationships between the characteristic bacteria identified via LEfSe analysis and clinical indicators. The findings revealed that the abundance of Pasteurellaceae was negatively correlated with GGT levels (R=−0.34, P=0.03), and the abundance of Parabacteroides was negatively correlated with the pediatric end-stage liver disease (PELD) score (R=−0.36, P=0.03); the remaining results are shown in Figure 5B.

Figure 5 Correlation analysis of gut microbiota and metabolites or clinical indicators. Spearman correlation analysis between the characteristic bacteria of LEfSe analysis and differentially abundant metabolites (A) and clinical indicators (B) in the HG and LG groups was employed. Red and green represent positive and negative correlations, respectively. (A) Heatmap of correlation analysis of the characteristic bacteria and differentially abundant metabolites in the HG and LG groups. (B) Heatmap of correlation analysis of the characteristic bacteria and clinical indicators in the HG and LG groups. *, P<0.05; **, P<0.01. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Dbil, direct bilirubin; GGT, γ-glutamyltransferase; HG, high-GGT; Ibil, indirect bilirubin; INR, international normalized ratio; LEfSe, linear discriminant analysis effect size; LG, low-GGT; PELD, pediatric end-stage liver disease; PT, prothrombin time; TBA, total bile acid; Tbil, total serum bilirubin.

Discussion

This study explored the variations in gut microbiota composition between infants with cholestasis and healthy infants. The Cho group presented greater bacterial richness and abundance than the Con group did, while there were no significant differences in gut microbiota diversity between the two groups, which was consistent with existing research (2,6,18), indicating that infants with cholestasis exhibit gut microbiota dysbiosis and that the changes in the gut microbiota are involved in the occurrence and development of IC. IC exists in two subtypes: HG cholestasis and LG cholestasis. On the basis of prior research (7-10,19,20), serum GGT levels frequently remain below 100 U/L in patients with LG cholestasis, whereas in patients with HG cholestasis, GGT levels tend to exceed 100 U/L. We explored the correlations between these microbial changes and metabolite profiles in the HG and LG subtypes of IC and revealed specific bacteria that played different roles in the different subtypes.

At the phylum level, compared with the Con group, the Cho group presented a lower abundance of Firmicutes, echoing findings reported in adult populations (21). Firmicutes species can produce bile salt hydrolase, an enzyme that catalyzes the hydrolysis of conjugated bile acids into free bile acids, thereby facilitating their excretion via feces (22). The decrease of Firmicutes abundance in the Cho group may have contributed to cholestasis. No significant differences were found in gut microbiota richness, abundance, or diversity between the HG and LG groups or in the phylum-level distribution of gut microbiota. This can be attributed to the fact that the infants in both the HG and LG groups suffered from cholestasis, resulting in relatively minor differences in gut microbiota composition at the phylum level.

At the genus level, the Cho group demonstrated a notably elevated abundance of Bacteroides. In previous studies (23,24), Bacteroides was reported to be used to treat necrotizing enterocolitis and alcohol-related liver disease. Thus, Bacteroides may be beneficial bacteria, which is inconsistent with our results. We found that the increase in the abundance of Bacteroides might be related to IC. However, in our study, when the LG group and the Con group were compared, no significant difference in the abundance of Bacteroides was detected, with the primary increase in abundance attributed to the HG group. Compared with both the LG and Con groups, the HG group presented a markedly greater abundance of Bacteroides (P<0.05). A previous study (25) revealed that Bacteroides fragilis played a role in biliary infection, with a significant correlation (P=0.015) between the abundance of Bacteroides fragilis and the risk of developing cholangitis. A previous study (26) has demonstrated an elevated abundance of Bacteroides in patients with gallstone disease. Bacteroides fragilis is a bile-tolerant microbe as well as an opportunistic pathogen that can migrate from the intestinal tract to the biliary tract or gallbladder under conditions of impaired immunity or gut dysbacteriosis. Bacteroides fragilis can participate in bile acid oxidation and epimerization and disrupt the enterohepatic circulation (27), inducing infection of the biliary system and promoting the formation of gallstones (28). These findings suggest that Bacteroides promotes the progression of bile tract conditions such as biliary tract infections and gallstone formation. In the HG group, Bacteroides may have promoted the development of infantile biliary obstructive cholestasis through a similar mechanism, but further verification is needed.

A functional comparison between the Cho and Con groups revealed differential metabolic pathways. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway in the Cho group was more enriched in “xenobiotic biodegradation and metabolism” and “metabolism of cofactors and vitamins”. Among the subtypes of IC, the LG group was involved in the toluene degradation. Toluene is the raw material for the biosynthesis of S-adenosyl-L-methionine, which can be used to treat liver injury (29). The enhanced capacity of the LG group for toluene degradation may contribute to liver injury. Pathways related to bile and lipid metabolism, such as fatty acid metabolism, retinol metabolism, and steroid hormone biosynthesis, are linked to liver inflammation, cholestasis, and dyslipidemia (30). These metabolic pathways for the infants in the LG group may be closely associated with the development of cholestasis. KEGG pathway enrichment analysis further revealed that the citrate cycle was significantly upregulated in the LG group, and it is possible that energy metabolism (citrate cycle) was also involved in the pathogenesis of cholestasis in the LG group.

The levels of four distinct metabolites differed between the HG and LG groups, with the level of malic acid notably increasing in the LG group. Malic acid is a key intermediate in the citrate cycle and is extensively involved in oxidative stress processes (31). Oxidative stress affects liver cell function, further aggravating liver injury and inflammatory response (32), resulting in cholestasis. In cholestasis, the accumulation of bile acids can lead to the inhibition of the citrate cycle, thus aggravating the occurrence of oxidative stress and leading to the formation of a vicious cycle (33). Our study revealed a correlation between the abundance of Gammaproteobacteria and malic acid levels. Gammaproteobacteria was a characteristic bacterial genus in the LG group according to the LEfSe analysis, and previous studies have demonstrated its abundant enrichment in patients with cholestasis (2) or liver cirrhosis (34). In addition, Gammaproteobacteria can produce large amounts of endogenous toxins and inflammatory mediators, thus also promoting the occurrence of oxidative stress (35,36). Moreover, KEGG analysis revealed enrichment the citrate cycle among pathways in the LG group, indicating that in the pathogenesis of LG cholestasis, Gammaproteobacteria may play an important role by affecting oxidative stress and the function of the citrate cycle. In the HG group, a significant increase in alkapton in the urine was observed. Alkapton was detected in the bile of patients with alkaptonuria, indicating that alkapton can be excreted through the biliary tract (37). The existence of biliary obstruction or injury in infants in the HG group may lead to the excretion disorder of alkapton. In addition, in metabolic disorders that can cause IC, such as citrin deficiency, tyrosinemia, and galactosemia, notable elevations in tyrosine levels have been documented (1). The tyrosine metabolic pathway is disturbed, leading to increased synthesis of alkapton. Therefore, elevated levels of alkapton in the urine may be the result of cholestasis.

By investigating the relationships between gut microbiota and clinical indicators, we discovered that the abundance of Pasteurellaceae was negatively correlated with GGT levels and that Pasteurellaceae was a characteristic bacterium of the LG group according to the LEfSe analysis, indicating that it might play an important role in the occurrence of intrahepatic cholestasis. An adult study revealed that the abundance of Pasteurellaceae was significantly higher in patients with cirrhosis than in healthy controls and was higher in patients with hepatic encephalopathy, suggesting a possible correlation with the severity of liver disease (38). Pasteurellaceae may participate in the progression of liver disease by affecting iron metabolism and fatty acid synthesis (39). Further exploration of its mechanism of action in infantile LG cholestasis could provide a target for future therapeutic strategies. In our study, Parabacteroides was negatively correlated with the PELD scores. PELD scores serve as indicators of the severity of hepatic disease, with a higher score reflecting a more critical condition of the liver. Therefore, Parabacteroides may have hepatoprotective effects. Studies have shown that Parabacteroides can hydrolyze various conjugated bile acids while converting them into multiple secondary bile acids and markedly alleviates obesity, insulin resistance, lipid metabolism disorders, and nonalcoholic fatty liver disease symptoms induced by a high-fat diet (40,41). According to LEfSe analysis, Parabacteroides was identified as a characteristic bacterial taxon in the HG group, and the utilization of Parabacteroides as a therapeutic intervention for HG cholestasis may emerge as a promising treatment strategy.

The current study is subject to certain limitations, primarily the limited sample size of the Con, HG and LG subgroups, which may have introduced biases into the findings. Second, owing to ethical concerns, the Con group did not undergo blood sample testing. Nevertheless, as the purpose of this study was to explore the disparities in gut microbiota and metabolic markers between the HG and LG groups, the study goal has been achieved. Third, in the correlation analysis of metabolites, clinical indicators and gut microbiota, conducting extensive correlation analyses may carry the risk of false-positive associations, and these findings should be verified in future studies by increasing the sample size or using more rigorous multiple comparison correction methods. Finally, although this study identified differences in the gut microbiota of each group and their relationship with relevant metabolites, the causal relationship and specific mechanisms leading to cholestasis need to be further verified.

Conclusions

There were differences in the abundance and richness of gut microbiota between the Cho and Con groups at both the phylum and genus levels. Furthermore, the changes in the gut microbiota in the HG and LG groups were investigated. Bacteroides and Parabacteroides may play roles in the HG group, whereas Gammaproteobacteria and Pasteurellaceae did so in the LG group. Our research provides new insights into the relationships between gut microbiota and different subtypes of IC, but the causal relationships and specific mechanisms remain to be characterized and further verified.

Acknowledgments

We would like to express our sincere gratitude to all the patients and their families who participated in this study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-183/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-183/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-183/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-183/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics committee of Fujian Children’s Hospital (No. 2023ETKLR05086), and the guardians of all participants provided written informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Fawaz R, Baumann U, Ekong U, et al. Guideline for the Evaluation of Cholestatic Jaundice in Infants: Joint Recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2017;64:154-68. [Crossref] [PubMed]
2. Jin M, Cui J, Ning H, et al. Alterations in gut microbiota and metabolite profiles in patients with infantile cholestasis. BMC Microbiol 2023;23:357. [Crossref] [PubMed]
3. Li M, Liu S, Wang M, et al. Gut Microbiota Dysbiosis Associated with Bile Acid Metabolism in Neonatal Cholestasis Disease. Sci Rep 2020;10:7686. [Crossref] [PubMed]
4. Han S, Wang K, Shen J, et al. Probiotic Pediococcus pentosaceus Li05 Improves Cholestasis through the FXR-SHP and FXR-FGF15 Pathways. Nutrients 2023;15:4864. [Crossref] [PubMed]
5. Yu L, Liu Y, Wang S, et al. Cholestasis: exploring the triangular relationship of gut microbiota-bile acid-cholestasis and the potential probiotic strategies. Gut Microbes 2023;15:2181930. [Crossref] [PubMed]
6. Wang Y, Gao X, Zhang X, et al. Gut Microbiota Dysbiosis Is Associated with Altered Bile Acid Metabolism in Infantile Cholestasis. mSystems 2019;4:e00463-19. [Crossref] [PubMed]
7. Qiu YL, Wang L, Huang M, et al. Association of novel TMEM67 variants with mild phenotypes of high gamma-glutamyl transpeptidase cholestasis and congenital hepatic fibrosis. J Cell Physiol 2022;237:2713-23. [Crossref] [PubMed]
8. Wang NL, Chen L, Lin J, et al. Non-invasive biomarkers for identification of vanishing bile duct syndrome among children with acute cholestatic hepatitis. Transl Pediatr 2023;12:1782-90. [Crossref] [PubMed]
9. Wang NL, Li LT, Wu BB, et al. The Features of GGT in Patients with ATP8B1 or ABCB11 Deficiency Improve the Diagnostic Efficiency. PLoS One 2016;11:e0153114. [Crossref] [PubMed]
10. Zhang J, Yang Y, Gong JY, et al. Low-GGT intrahepatic cholestasis associated with biallelic USP53 variants: Clinical, histological and ultrastructural characterization. Liver Int 2020;40:1142-50. [Crossref] [PubMed]
11. Shatokhina O, Semenova N, Demina N, et al. A Two-Year Clinical Description of a Patient with a Rare Type of Low-GGT Cholestasis Caused by a Novel Variant of USP53. Genes (Basel) 2021;12:1618. [Crossref] [PubMed]
12. Luo X, Lu LG. Progress in the Management of Patients with Cholestatic Liver Disease: Where Are We and Where Are We Going? J Clin Transl Hepatol 2024;12:581-8. [Crossref] [PubMed]
13. Feldman AG, Sokol RJ. Neonatal cholestasis: emerging molecular diagnostics and potential novel therapeutics. Nat Rev Gastroenterol Hepatol 2019;16:346-60. [Crossref] [PubMed]
14. Caporaso JG, Kuczynski J, Stombaugh J, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010;7:335-6. [Crossref] [PubMed]
15. Quast C, Pruesse E, Yilmaz P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 2013;41:D590-6. [Crossref] [PubMed]
16. Gao Y, Zhang G, Jiang S, et al. Wekemo Bioincloud: A user-friendly platform for meta-omics data analyses. Imeta 2024;3:e175. [Crossref] [PubMed]
17. Xu M, Jiang Z, Huang W, et al. Altered Gut Microbiota Composition in Subjects Infected With Clonorchis sinensis. Front Microbiol 2018;9:2292. [Crossref] [PubMed]
18. Guo C, Li Y, Wang P, et al. Alterations of Gut Microbiota in Cholestatic Infants and Their Correlation With Hepatic Function. Front Microbiol 2018;9:2682. [Crossref] [PubMed]
19. Overeem AW, Li Q, Qiu YL, et al. A Molecular Mechanism Underlying Genotype-Specific Intrahepatic Cholestasis Resulting From MYO5B Mutations. Hepatology 2020;72:213-29. [Crossref] [PubMed]
20. Toh Q, Chen Y, Lee YY, et al. Simple biliary atresia score-a validated diagnostic aid for infantile cholestasis. Pediatr Surg Int 2024;40:212. [Crossref] [PubMed]
21. Li X, Xie H, Chao JJ, et al. Profiles and integration of the gut microbiome and fecal metabolites in severe intrahepatic cholestasis of pregnancy. BMC Microbiol 2023;23:282. [Crossref] [PubMed]
22. Song Z, Cai Y, Lao X, et al. Taxonomic profiling and populational patterns of bacterial bile salt hydrolase (BSH) genes based on worldwide human gut microbiome. Microbiome 2019;7:9. [Crossref] [PubMed]
23. Chen Z, Chen H, Huang W, et al. Bacteroides fragilis alleviates necrotizing enterocolitis through restoring bile acid metabolism balance using bile salt hydrolase and inhibiting FXR-NLRP3 signaling pathway. Gut Microbes 2024;16:2379566. [Crossref] [PubMed]
24. Sangineto M, Grander C, Grabherr F, et al. Recovery of Bacteroides thetaiotaomicron ameliorates hepatic steatosis in experimental alcohol-related liver disease. Gut Microbes 2022;14:2089006. [Crossref] [PubMed]
25. Effenberger M, Al-Zoairy R, Gstir R, et al. Transmission of oral microbiota to the biliary tract during endoscopic retrograde cholangiography. BMC Gastroenterol 2023;23:103. [Crossref] [PubMed]
26. Chang WT, Lee KT, Wang SR, et al. Bacteriology and antimicrobial susceptibility in biliary tract disease: an audit of 10-year's experience. Kaohsiung J Med Sci 2002;18:221-8.
27. Grigor'eva IN, Romanova TI. Gallstone Disease and Microbiome. Microorganisms 2020;8:835. [Crossref] [PubMed]
28. Ding L, Wang S, Jiang W, et al. Identification of Intestinal Microbial Community in Gallstone Patients with Metagenomic Next-Generation Sequencing. Diagnostics (Basel) 2023;13:2712. [Crossref] [PubMed]
29. Tong S, Guo J, Song H, et al. Effect of ademetionine on cytochrome P450 isoforms activity in rats. Int J Clin Exp Med 2015;8:9716-22.
30. Han J, Wu P, Wen Y, et al. The zhuyu pill relieves rat cholestasis by regulating the mRNA expression of lipid and bile metabolism associated genes. Front Pharmacol 2023;14:1280864. [Crossref] [PubMed]
31. Chen H, Cao S, Chen J, et al. BomMDH1 regulates malate-mediated oxidative stress in tobacco BY-2 suspension cells. J Plant Physiol 2024;300:154297. [Crossref] [PubMed]
32. Heidari R, Niknahad H. The Role and Study of Mitochondrial Impairment and Oxidative Stress in Cholestasis. Methods Mol Biol 2019;1981:117-32. [Crossref] [PubMed]
33. Zhang Y, Lu Y, Ji H, et al. Anti-inflammatory, anti-oxidative stress and novel therapeutic targets for cholestatic liver injury. Biosci Trends 2019;13:23-31. [Crossref] [PubMed]
34. Wu Z, Zhou H, Liu D, et al. Alterations in the gut microbiota and the efficacy of adjuvant probiotic therapy in liver cirrhosis. Front Cell Infect Microbiol 2023;13:1218552. [Crossref] [PubMed]
35. Freeman AE, Willis KA, Qiao L, et al. Microbial-induced Redox Imbalance in the Neonatal Lung Is Ameliorated by Live Biotherapeutics. Am J Respir Cell Mol Biol 2023;68:267-78. [Crossref] [PubMed]
36. Karandikar K, Bhonde G, Palav H, et al. A novel gut microbiome-immune axis influencing pathology in HCMV infected infants with neonatal cholestasis. Microbes Infect 2023;25:105165. [Crossref] [PubMed]
37. Norman BP, Sutherland H, Wilson PJM, et al. Hepatobiliary circulation and dominant urinary excretion of homogentisic acid in a mouse model of alkaptonuria. J Inherit Metab Dis 2024;47:664-73. [Crossref] [PubMed]
38. Hua X, Feng H. Changes in intestinal microbiota of HBV-associated liver cirrhosis with/without hepatic encephalopathy. Medicine (Baltimore) 2022;101:e29935. [Crossref] [PubMed]
39. Mayneris-Perxachs J, Cardellini M, Hoyles L, et al. Iron status influences non-alcoholic fatty liver disease in obesity through the gut microbiome. Microbiome 2021;9:104. [Crossref] [PubMed]
40. Wang K, Liao M, Zhou N, et al. Parabacteroides distasonis Alleviates Obesity and Metabolic Dysfunctions via Production of Succinate and Secondary Bile Acids. Cell Rep 2019;26:222-235.e5. [Crossref] [PubMed]
41. Cui Y, Zhang L, Wang X, et al. Roles of intestinal Parabacteroides in human health and diseases. FEMS Microbiol Lett 2022;369:fnac072. [Crossref] [PubMed]