Overview
Wu and colleagues at Lanzhou University (China) investigated whether antimony and copper induce liver inflammation through the enterohepatic circulation by examining gut microbiota composition and targeted bile acid (BA) metabolomics in C57 BL/6 mice. Twenty-eight male mice were gavaged every two days for six weeks with antimony (10 mg/kg body weight as potassium antimony tartrate), copper (100 mg/kg as copper sulfate), both together (SC group), or sterilized water as control (n = 7 per group). Antimony and/or copper treatment disrupted the gut microbiota, reducing probiotic bacteria such as Lactobacillus spp. and increasing harmful bacteria, while significantly elevating taurine-beta-muricholic acid (T-beta-MCA) in serum and liver. T-beta-MCA acts as a farnesoid X receptor (FXR) antagonist and inhibited the FXR-SHP signaling pathway in liver and the FXR-FGF15 pathway in ileum, which promoted cholesterol 7-alpha-hydroxylase (CYP7A1) transcription, increased total bile acid concentrations, and resulted in liver inflammation. The study is a mechanistic animal study; it reports no food occurrence values and does not support any direct claim about consumer-level antimony or copper exposure.
Study design
Species and housing. Male C57 BL/6 mice approximately 20 g in body weight and approximately 6 weeks of age were sourced from GemPharmatech Co., Ltd. (Nanjing, China). Animals were maintained under specific-pathogen-free (SPF) conditions with 40 to 60 percent relative humidity, 22 plus or minus 1 degrees C, and a 12 hour light/dark cycle. Mice were acclimatized for one week before treatment began.
Groups and dosing. Twenty-eight mice were randomly assigned to four groups (n = 7 each):
| Group | Treatment | Dose |
|---|---|---|
| Control | Sterilized water | n/a |
| Antimony | Potassium antimony tartrate solution | 10 mg/kg body weight |
| Copper | Copper sulfate solution | 100 mg/kg body weight |
| SC | Potassium antimony tartrate plus copper sulfate | 10 mg/kg antimony plus 100 mg/kg copper |
Each treatment solution was administered by oral gavage every two days for six weeks on a standard diet. Body weight and food intake were recorded weekly.
Tissue collection. At the end of six weeks, mice were euthanized under isoflurane anesthesia. Serum, liver, ileum, and cecal contents were collected. The intestine was rinsed with sterile saline before the ileum was excised. Cecal contents were extracted from the cecum. All samples except those for immediate analysis were preserved at minus 80 degrees C. 16S rRNA sequencing was performed on cecal contents. Ileum and liver tissues were used for qPCR. Serum and liver tissue were used for ELISA and UHPLC-QQQ-MS bile acid quantification.
Ethical approval. Ethics committee of the College of Pastoral Agriculture Science and Technology at Lanzhou University; approval number CY20230301.
Analytes measured
Serum biochemical markers (ELISA and clinical biochemical assays): alkaline phosphatase (ALP), blood urea nitrogen (BUN/UREA), total bile acids (TBA), cholesterol (CHO), alanine transaminase (ALT), aspartate aminotransferase (AST), glucose (GLU), triglycerides (TG), creatinine (CREA), high-density lipoprotein cholesterol (HDL-C), total bilirubin (T-BIL), direct bilirubin (D-BIL).
Pro-inflammatory cytokines (ELISA, liver and serum): interleukin-1 beta (IL-1 beta), interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-alpha). ELISA kits supplied by Biogoethe Biotechnology Co., Ltd., Wuhan, China.
Bile acid metabolites (UHPLC-QQQ-MS, serum and liver): 28 BAs including conjugated forms (TCA, TDCA, TCDCA, TUDCA, T-alpha-MCA, T-beta-MCA) and unconjugated forms (CA, DCA, CDCA, UDCA, alpha-MCA, beta-MCA, omega-MCA, HDCA, LCA). Concentrations expressed as nmol/L (serum) and nmol/g (liver).
Gene expression (qPCR, ileum and liver): ileal Fxr, ileal Fgf15; hepatic Fxr, Fgfr4, Shp, Cyp7a1. Beta-actin was the endogenous reference gene (ΔΔCt method).
Gut microbiota (16S rRNA amplicon sequencing, cecal contents): alpha-diversity (ACE index, richness index), beta-diversity (Bray-Curtis dissimilarity via constrained principal coordinate analysis, CPCoA), genus-level relative abundances.
Key numbers
Serum biochemical markers (Figure 1, Figure S2; mean plus or minus SD, n = 7, one-way ANOVA, different letters = p < 0.05):
| Endpoint | Antimony vs. control | Copper vs. control | SC vs. control |
|---|---|---|---|
| Serum ALP (U/L) | n.s. | +1.3-fold (significant) | n.s. |
| Serum UREA (U/L) | n.s. | n.s. | -1.6-fold (significant) |
| Serum GLU | n.s. | n.s. | remarkably elevated |
| Serum CREA | n.s. | n.s. | significantly reduced |
| Serum D-BIL | n.s. | significantly increased | n.s. |
| Serum T-BIL | n.s. | n.s. | significantly increased (vs. antimony group) |
| Hepatic IL-1 beta (pg/mL) | n.s. | n.s. | +1.4-fold (significant) |
| Serum TNF-alpha, IL-6 | no significant difference across all groups | n/a | n/a |
ALT, AST, HDL-C, and TG had no significant differences in pairwise comparisons.
Bile acid pool: cholesterol and total bile acids (Figure 2; ELISA; mean plus or minus SD, n = 7):
| Endpoint | Antimony and copper vs. control | SC vs. control |
|---|---|---|
| Serum CHO (U/L) | -1.3-fold (significant) | -1.7-fold (significant) |
| Hepatic CHO (micromol/g) | -1.3-fold (significant) | -1.7-fold (significant) |
| Serum TBA (U/L) | +1.6-fold (significant) | +2.0-fold (significant) |
| Hepatic TBA (nmol/g) | n.s. for antimony and copper individually | significantly elevated in SC |
| Hepatic Cyp7a1 mRNA (fold) | +2-fold in all three treatment groups | +2-fold |
Dominant BAs in control group serum: TCA (20.2%), beta-MCA (15.2%), DCA (15.2%), UDCA (9.3%) (Figure 3A). Dominant BAs in control group liver: T-alpha-MCA (45.4%), TCA (24.8%), T-beta-MCA (12.9%) (Figure 3B).
Taurine-beta-muricholic acid (T-beta-MCA; Figures 3C, 3D, 4A, 4B; mean plus or minus SD, n = 7, Student’s t test):
| Measurement | Treatment | Change vs. control |
|---|---|---|
| Serum T-beta-MCA (nmol/L) | Copper | +1.8-fold (significant, p < 0.05) |
| Serum T-beta-MCA (nmol/L) | SC | +1.8-fold (significant, p < 0.05) |
| Hepatic T-beta-MCA (nmol/g) | Antimony | +1.5-fold (significant, p < 0.01) |
| Hepatic T-beta-MCA (nmol/g) | Copper | +1.5-fold (significant, p < 0.01) |
| Hepatic T-beta-MCA (nmol/g) | SC | +1.8-fold (significant, p < 0.001) |
Serum TCDCA was significantly increased in all three treatment groups. Liver T-beta-MCA was significantly elevated in all three treatment groups.
FXR-SHP and FXR-FGF15 pathway gene expression (Figures 4C to 4G; qPCR, fold change vs. control; mean plus or minus SD, n = 7, one-way ANOVA):
| Gene | Tissue | Antimony | Copper | SC |
|---|---|---|---|---|
| Fxr | Ileum | n.s. | n.s. | n.s. |
| Fgf15 | Ileum | -2.2-fold (significant) | -1.7-fold (significant) | -1.7-fold (significant) |
| Fxr | Liver | -2-fold (significant) | -2-fold (significant) | -2-fold (significant) |
| Fgfr4 | Liver | n.s. | n.s. | n.s. |
| Shp | Liver | n.s. | n.s. | significantly reduced |
Gut microbiota diversity (Figure 5; n = 7 per group):
Alpha-diversity: ACE index and richness index were significantly elevated in the antimony group compared to the control group. No significant differences in alpha-diversity were detected in pairwise comparisons overall (stated in Discussion). Beta-diversity: CPCoA based on the Bray-Curtis dissimilarity matrix revealed significant inter-group variation (p = 0.001, 44.31% of variance on CPCoA1; Figure 5D).
Dominant phyla across all groups: Bacteroidetes, Firmicutes, Verrucomicrobia, and Proteobacteria (Figure 6A). Leading classes: Clostridia, Bacteroidia, Verrucomicrobiae, and Deltaproteobacteria (Figure 6B).
Differential bacterial genera (Figures 6C to 6H and Figure 7; Wilcoxon rank-sum test, p < 0.05, FDR < 0.1):
| Genus | Direction in treatment groups | Notes |
|---|---|---|
| Lactobacillus spp. | Down (significantly reduced 2.6-fold in treatment groups vs. control) | Correlated negatively with serum T-beta-MCA; R2 = -0.6139, p = 0.0031 (Figure 7I) |
| Akkermansia spp. | Down | Beneficial; reduced in antimony and/or copper groups |
| Anaerocolumna spp. | Down | Beneficial; reduced in antimony and/or copper groups |
| Limosilactobacillus spp. | Up | Harmful; increased in SC group most prominently |
| Butyrivibrio spp. | Up | Harmful; increased in antimony group |
| Paramuribaculum spp. | Up | Harmful; increased in SC group |
| Muribaculum spp. | Up | Harmful; increased in SC group |
| Intestinimonas spp. | Up | Harmful; increased in copper and SC groups |
Between the control and antimony groups, 11 genera exhibited significant differences (3 up, 8 down). Between the control and copper groups, 12 genera differed (5 up, 7 down). Between the control and SC groups, 20 genera differed (6 up, 14 down).
Methods
Reagents. Potassium antimony tartrate (purity 99.5 percent, cat. XW2830074501) and copper sulfate (purity 99.0 percent, cat. 10008216) from China National Pharmaceutical Group Corporation Co., Ltd. (Beijing, China). Heparin sodium (purity 99.5 percent, cat. A603251) from Sangon Biotech Co., Ltd. (Shanghai, China). All bile acid standards from Sigma-Aldrich Chemical Co., Ltd. (St. Louis, MO, USA), including isoflurane (purity 98.0 percent, cat. 792632). Acetonitrile (MeCN, purity 99.99 percent, cat. 85188), HPLC-grade water (purity 100 percent, cat. TS-51140), and methanol (MeOH, purity 99.0 percent, cat. A955-4) from Thermo Fisher Scientific Co., Ltd. (Waltham, MA, USA).
Gavage. Treatment solutions prepared by dissolving potassium antimony tartrate (antimony group, 10 mg/kg body weight) and copper sulfate (copper group, 100 mg/kg body weight) in pure water, or both combined (SC group). Control received sterilized water. Gavage administered every two days for six weeks on a standard diet.
Serum biochemical assays and histopathology. Blood samples collected for serum biochemical parameters including TBA, ALP, AST, GLU, BUN, TG, CREA, CHO, T-BIL, HDL-C, ALT, and D-BIL. Liver samples fixed and stained with hematoxylin and eosin (H&E). Clinical biochemical assays and histopathological examination performed by Wuhan Servicebio Technology Co., Ltd. (Wuhan, China). H&E images at 400 times magnification.
Bile acid extraction and UHPLC-QQQ-MS. Bile acid metabolites isolated from liver tissue (approximately 10 mg) and serum (approximately 50 microliters) following the extraction method described in the Supporting Information. Analysis conducted by ultra-high-performance liquid chromatography coupled with a triple quadrupole mass spectrometer (UHPLC-QQQ-MS) using a previously published protocol with refinements. Twenty-eight bile acid standard compounds were measured; a complete list is in Table S1 of the paper. Data expressed as nmol/L (serum) and nmol/g (liver).
ELISA. Concentrations of TBA and total cholesterol (CHO) in liver, and pro-inflammatory cytokines IL-1 beta, IL-6, and TNF-alpha in serum, determined by ELISA using kits supplied by Biogoethe Biotechnology Co., Ltd. (Wuhan, China).
qPCR. RNA extracted from 100 mg of frozen ileum and liver tissue using an RNA extraction kit from Biogoethe Biotechnology Co., Ltd. (Wuhan, China). One microgram of RNA was reverse-transcribed to cDNA using a kit from Biogoethe Biotechnology Co., Ltd. SYBR green-based qPCR with the following thermal profile: initial denaturation 95 degrees C for 20 seconds, 40 cycles of 95 degrees C for 30 seconds and 60 degrees C for 30 seconds. Beta-actin was the endogenous control gene. Data analyzed by the Applied Biosystems Comparative CT method (ΔΔCt). Specific primer sequences are listed in Supporting Information Table S2.
16S rRNA amplicon sequencing. Approximately 100 mg of cecal content used for total DNA extraction. 16S rRNA gene amplicon sequencing library prepared following Illumina Co., Ltd. (San Diego, CA, USA) guidelines. Sequencing performed on an Illumina platform. Rarefaction curves reached a plateau, confirming adequate sequencing depth. Alpha-diversity indices: ACE and richness. Beta-diversity: Bray-Curtis dissimilarity matrix, CPCoA visualization. Differential genera identified by volcano plots using Wilcoxon rank-sum test (p < 0.05, FDR < 0.1). Quality filtering, assembly, and statistical methods described in Supporting Information. 16S rRNA sequences deposited in NCBI GenBank under accession PRJNA1077802.
Statistical analysis. R (version 4.1.0) and GraphPad Prism (version 8.0). Significance threshold p < 0.05 for pairwise comparisons by one-way ANOVA or two-tailed Student’s t test. Correlations between Lactobacillus spp. abundance and specific BAs assessed with GraphPad Prism.
Instrument
Bile acid metabolomics: Ultra-high-performance liquid chromatography coupled with a triple quadrupole mass spectrometer (UHPLC-QQQ-MS). Make and model not specified in the main paper text; full instrument parameters are referenced to Supporting Information.
Gene expression: Applied Biosystems-compatible qPCR platform (ΔΔCt data analysis via Applied Biosystems Comparative CT method); specific thermocycler not named.
Gut microbiota sequencing: Illumina sequencing platform; specific instrument not named.
ELISA reader: Not specified.
Limit of detection and limit of quantification
Limits of detection and quantification for UHPLC-QQQ-MS bile acid measurements are not reported in the main paper. The Supporting Information is referenced for the full extraction and analytical protocol but was not available in the deposited evidence unit. LOD and LOQ for serum biochemical and ELISA assays are not stated.
Recovery and reference standards
Certified reference material (CRM) use and recovery statistics for the UHPLC-QQQ-MS bile acid assay are not reported in the main paper. The paper references a “previously established protocol with some refinements” for the bile acid extraction and quantification procedure. Reagent grade purity and commercial standard compounds for all 28 BAs are stated. CRM recovery data for the gavage dose preparations are not reported.
Regulatory context
The paper does not evaluate or cite food safety regulatory limits for antimony or copper in food or consumer products. The two hepatotoxic dose levels selected — antimony at 10 mg/kg body weight per gavage and copper at 100 mg/kg body weight per gavage — are cited in the paper as levels that induce significant hepatotoxic damage (antimony, reference 9: Zhong et al. 2021, Sci. Total Environ.) and histopathological inflammation and apoptosis (copper, reference 10: Wu et al. 2020, Ecotoxicol. Environ. Saf.) in prior mouse work. These doses are orders of magnitude above typical dietary exposures and are used here as a mechanistic model, not as a representation of human dietary intake. No environmental maximum levels, food contaminant regulations, or health-based guidance values (TDI, PTWI, or similar) are discussed in the paper.
Evidence fitness
This source provides mechanistic context for how antimony and copper exposures, at high gavage doses in mice, disrupt gut microbiota and bile acid metabolism to promote liver inflammation via the FXR-SHP and FXR-FGF15 signaling pathways. It supports the following specific mechanistic claims with quantitative evidence from the paper: (1) antimony and/or copper at the stated gavage doses elevate serum and hepatic T-beta-MCA and total bile acid concentrations; (2) Lactobacillus spp. abundance is negatively correlated with serum T-beta-MCA levels (R2 = -0.6139, p = 0.0031); (3) antimony and/or copper suppress ileal Fgf15 and hepatic Fxr and Shp expression while elevating hepatic Cyp7a1 mRNA by 2-fold; (4) gut microbiota beta-diversity is significantly disrupted (p = 0.001). The source does not support any claim about food occurrence values for antimony or copper, any claim about human dietary exposure, or any certification threshold. The mouse gavage doses used are far above consumer-relevant dietary exposures.
Limitations
The gavage doses (antimony 10 mg/kg body weight, copper 100 mg/kg body weight) are designed to produce a mechanistic model of liver inflammation and are not representative of human dietary or environmental exposures to antimony or copper from food or consumer products. The study used male mice only; sex-specific differences in bile acid metabolism and gut microbiota composition are not addressed. Sample size is small (n = 7 per group). The combined SC group confounds attribution of specific microbiota changes to antimony versus copper individually. Alpha-diversity indices showed inconsistency: ACE and richness were elevated in the antimony group versus control in group comparisons, but pairwise alpha-diversity comparisons were described as not significant overall in the Discussion. The study was conducted in a single facility under a single set of housing conditions; external validity to field-relevant exposures is limited. LOD, LOQ, and CRM recovery data for the bile acid assay are not reported. The specific UHPLC-QQQ-MS instrument model is not disclosed in the main text.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.
