Mechanistic insights into zearalenone-accelerated colorectal cancer in mice using integrative multi-omics approaches

**Background:** Colorectal cancer (CRC) is the third leading cause of cancer-related death worldwide. Zearalenone (ZEA) is a mycotoxin produced by Fusarium spp., commonly found in cereals and grains, and structurally resembles estrogen. While in vitro studies have suggested ZEA promotes colon cancer cell growth and metastasis, in vivo evidence and mechanistic understanding have been lacking. Given that ZEA is primarily absorbed through the gastrointestinal tract and directly exposes gut microbiota, this study aimed to comprehensively evaluate the effects of ZEA exposure on CRC development in a living animal model using integrative multi-omics approaches.

**Methods:** Five-week-old male BALB/cAnN nude mice were fed a phytoestrogen-free diet and injected subcutaneously with 1 × 10^6 SW480 human colon cancer cells. When tumors reached 70–100 mm³, mice were randomly divided into two groups (n = 9 per group): control (oral gavage of olive oil vehicle) and ZEA (0.5 mg/kg body weight in olive oil), administered thrice weekly for 28 days. Tumor volume was measured three times weekly. After sacrifice, tumors and cecal contents were collected. RNA-seq was performed on tumor tissue (n = 6 per group) using Illumina HiSeq 2500. Differentially expressed genes (DEGs) were identified using DESeq2 (adjusted p < 0.1). Pathway enrichment was performed using GAGE (KEGG) and GSEA (MSigDB). Serum metabolomics included both non-targeted (LC-MS/MS) and targeted amino acid profiling (UHPLC-QqQ-MS/MS). Gut microbiota was analyzed by 16S rDNA V3-V4 sequencing (Illumina HiSeq 2500) with ANCOM-BC for differential abundance (adjusted p < 0.05). Cecal short-chain fatty acids (SCFAs) were measured by GC-MS. Integrative analysis used Spearman correlations (FDR < 0.05, |correlation| > 0.8).

**Key Results:** ZEA significantly accelerated tumor growth. Tumor volume was significantly larger in the ZEA group (p < 0.001) by day 28, and tumor weight tripled compared to controls (p = 0.036). Ki67-positive cells increased twofold (p = 0.002), while TUNEL staining showed reduced apoptosis. RNA-seq identified 21 DEGs (12 upregulated, 9 downregulated). Notably, DGKB showed a log2 fold change of 4.37 (adjusted p = 0.039), and BEST4 showed a log2 fold change of 1.32 (adjusted p = 0.054). Upregulated KEGG pathways included calcium signaling (p_adj = 0.055), protein digestion and absorption (p_adj = 0.026), and neuroactive ligand-receptor interaction (p_adj = 0.025). Downregulated pathways included ribosome (p_adj = 3.20 × 10⁻¹²), oxidative phosphorylation (p_adj = 3.51 × 10⁻⁴), and thermogenesis (p_adj = 0.0128). GSEA revealed enrichment of oncogenic signatures including epithelial-mesenchymal transition, KRAS pathway, and invasive cancer signatures (|NES| > 1, FDR < 0.25). Serum metabolomics identified 19 significantly altered amino acids (targeted) and 21 significantly altered metabolites (non-targeted). Amino acid concentrations were generally lower in the ZEA group. Four activated metabolic pathways were identified by IPA: tRNA charging, citrulline metabolism, citrulline biosynthesis, and purine nucleotide de novo biosynthesis II. Gut microbiota analysis showed significantly lower alpha diversity in the ZEA group (Shannon index, p = 0.007). Nine genera were differentially abundant, including reduced levels of SCFA-producing bacteria (Rikenella, Rikenellaceae RC9 gut group) and anti-inflammatory bacteria (Candidatus Saccharimonas). Cecal SCFAs (total, acetic acid, propionic acid, butyric acid) were significantly lower in the ZEA group (p < 0.05). Integrative analysis revealed 33 significant correlations among DEGs, metabolites, and genera. DGKB was positively correlated with uric acid (correlation = 0.909) and p-cresol sulfate (correlation = 0.902), and negatively correlated with Tuzzerella (correlation = −0.881). Hippuric acid was positively correlated with BEST4 (correlation = 0.895), CYP4F29P (correlation = 0.888), JAK3 (correlation = 0.888), and ROBO3 (correlation = 0.881). Tuzzerella was negatively correlated with BEST4 (correlation = −0.86), DGKB (correlation = −0.881), and uric acid (correlation = −0.937).

**Clinical Implications:** This study provides the first in vivo evidence that dietary ZEA exposure, even at low doses, promotes CRC tumor growth through multiple mechanisms: upregulation of the BEST4/AKT/ERK1/2 oncogenic pathway, systemic metabolic alterations (reduced amino acids, altered purine and indolic metabolites), and disruption of gut microbiota (reduced SCFA-producing and anti-inflammatory bacteria). The strong correlations between ZEA-altered gut microbes (e.g., Tuzzerella), serum metabolites (e.g., uric acid, hippuric acid), and oncogenic gene expression (BEST4, DGKB) suggest a multi-level mechanism for ZEA-induced CRC progression. Given that ZEA is commonly found in human diets and its effects may be persistent, these findings underscore the need for further research on chronic ZEA exposure as a potential environmental risk factor for CRC. Limitations include the use of a xenograft mouse model (which may not fully replicate human physiology and gut microbiota), and the inability of 16S rDNA sequencing to resolve bacterial species-level functions.