Skip to main content
Frontiers in Microbiology logoLink to Frontiers in Microbiology
. 2026 Nov 4;17:1970842. doi: 10.3389/fmicb.2026.1970842

The sex-related differences in gut microbiota characteristics and keystone species identification of lipid metabolism in dyslipidemia

Yafang Xu 1,2,†, Jinglong Jin 3,†, Jiaxin An 1, Qianrui Wei 1, Oumarou Chaibou Zafir 1, Zhijian Bo 4, Mingchun Luan 4, Liang Cao 3,*, Huajun Li 1,*
PMCID: PMC13779484  PMID: 43054519

Abstract

Background

Dyslipidemia exhibits pronounced sex-related differences, yet the integrated interactions among gut microbiota, host metabolism, and immune responses underlying these sex-related differences remain poorly understood.

Methods

We investigated sex-related alterations in gut microbiota composition, lipid metabolism, and inflammatory responses by integrating the re-analysis of a clinical cohort and mouse model in dyslipidemia, alongside isolation and identification of a keystone probiotic strain.

Results

High-fat diet (HFD) induced comparable hepatic steatosis and inflammation in both sexes. However, male mice exhibited earlier body weight gain and impaired intestinal barrier integrity (downregulated ZO-1/Occludin). Metabolomics analysis revealed distinct sex-related bile acid metabolic signatures, with females showing increased levels of several bile acids and enrichment of ascorbate and aldarate metabolism, while males exhibited reduced primary bile acids together with enrichment of bile acid biosynthesis and secretion pathways. Depletion of L. johnsonii observed in both clinical cohort and HFD mice was correlated with keystone taxa (Lactobacillus) and with lipid, inflammation, and bile acid markers. Moreover, the isolated L. johnsonii DM2420 strain exhibited cholesterol-lowering activity and BSH activity, harbored multiple genes associated with stress adaptation, and its supernatant metabolomics indicated bile acid remodeling in vitro.

Conclusion

Collectively, male mice exhibited greater susceptibility than females to HFD-induced dyslipidemia, which may be associated with the depletion of the candidate keystone taxon L. johnsonii in males and with related bile acid metabolic reprogramming. The isolated strain of L. johnsonii DM2420 exhibited a multifunctional profile that underscores its probiotic potential. Our findings preliminarily suggest a potential association between gut microbiota, bile acids, and immune responses in dyslipidemia and highlight the potential of keystone species-based precision strategies for sex-informed interventions.

Keywords: bile acid metabolism, dyslipidemia, gut microbiota, keystone species, Lactobacillus johnsonii, sex differences

1. Introduction

Rapid economic growth, shifting dietary habits, and the increasing prevalence of unhealthy lifestyles have significantly elevated the risk and incidence of dyslipidemia (Pirillo et al., 2021). Dyslipidemia is pathologically defined by disturbances in blood lipid profiles—including elevated levels of total cholesterol (TC), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C), alongside reduced levels of high-density lipoprotein cholesterol (HDL-C) (Joint Committee on the Chinese Guidelines for Lipid Management, 2023; Brunham et al., 2024). It is a major contributor to atherosclerotic plaque formation and increases the risk of various comorbidities, such as fractures, kidney disease, type 2 diabetes, and certain malignancies (Brunham et al., 2024).

Substantial sex differences in obesity-related metabolism have been documented in both human and animal studies (Peng et al., 2020). Males and females exhibit distinct susceptibilities and phenotypic responses to high-fat diet (HFD)-induced dyslipidemia and its associated metabolic disorders, likely reflecting the combined effects of hormone status, gene expression, and lifestyle factors (Mulder et al., 2024). Chronic HFD feeding is primarily associated with visceral obesity in males, accompanied by enhanced lipolytic activity and elevated secretion of pro-inflammatory cytokines, predisposing them to hepatic lipid accumulation and increased free fatty acid flux. Clinical evidence indicates that under chronic HFD conditions, men are more prone to hypertriglyceridemia and hypo-α-lipoproteinemia (Ye et al., 2023), whereas premenopausal women maintain more favorable lipid profiles, characterized by higher HDL and lower TC/LDL levels (Ambikairajah et al., 2019). This protective phenotype in females is partly attributable to estrogen-mediated enhancement of lipoprotein lipase (LPL) activity in adipose tissue, which promotes TG metabolism and lipoprotein clearance (Ko and Jung, 2021; Ryczkowska et al., 2023); conversely, HFD suppresses adipose LPL activity in males, thereby exacerbating hypertriglyceridemia (Spitler and Davies, 2020). Beyond lipid metabolism, sex-related differences also manifest in immune cytokine activation and immune cell populations, including T cells, B cells, and macrophages. Preclinical studies have demonstrated that HFD-fed male mice exhibit elevated circulating levels of interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), with predominant M1 macrophage polarization (Braga Tibaes et al., 2022; Lefebvre et al., 2024); whereas female mice display M2-dominant macrophage characteristics, likely mediated by estrogen-dependent protective signaling (Liu et al., 2005; Shu et al., 2022; Peng et al., 2023; Strizova et al., 2023).

The gut microbiota has emerged as a critical nexus linking diet, host metabolism, and immune function, with diet now recognized as a major external determinant of gut microbiota composition and homeostasis. Notably, sex differences in gut microbiota composition have been consistently observed in both humans and rodents (Org et al., 2016; de la Cuesta-Zuluaga et al., 2019). Previous clinical case–control studies have reported a correlation between hyperlipidemia and gut microbiota alterations, with preliminary evidence suggesting that the gut microbial structure differs between hyperlipidemic men and women (Xia et al., 2024). Peng et al. (2020) reported that HFD-fed female mice harbored higher relative abundances of Parabacteroides, Lactobacillus, Bacteroides, and Bifidobacterium than their male counterparts. Such microbial shifts can promote the production of endotoxins such as lipopolysaccharide (LPS), thereby compromising intestinal mucosal barrier function. HFD feeding induces extensive restructuring of the gut microbiota community, characterized by increased abundances of Firmicutes and Proteobacteria, and decreased abundances of Bacteroidetes and Actinobacteria, changes that correlated significantly with metabolic parameters (Wang et al., 2015).

Within this ecosystem, keystone species—taxa with high connectivity that exert disproportionately large effects on community structure and function relative to their abundance—play indispensable roles in maintaining host intestinal homeostasis and immune competence (Banerjee et al., 2018). Through systematic characterization of intestinal microbiota in healthy mice versus those with metabolic syndrome, Wang et al. identified 37 core keystone bacteria capable of discriminating between the two groups, including taxa involved in carbohydrate utilization and lactate/acetate production, such as Lactobacillus, Bifidobacterium, Bacteroides, and Parabacteroides (Wang J. et al., 2019). Keystone microbes interact through multiple mechanisms—molecular secretion, cross-feeding, and synthesis of antimicrobial compounds—thereby modulating community structure by regulating effector population abundance and participating in niche competition (Banerjee et al., 2018). These intricate interdependencies collectively shape microbiome configuration and functional output (Banerjee et al., 2018).

Despite these advances, the interplay between diet and sex in regulating gut microbiota and host metabolism remains inadequately explored. Current research has significant gaps: most HFD-based animal studies are skewed toward male models, resulting in a conspicuous paucity of female data; moreover, sex-related differences in host metabolism and immunity, as well as their interaction networks with the gut microbiota, remain largely elusive. It remains unclear how key genera such as Lactobacillus contribute to sex-related metabolic regulation. To address these gaps, the present study employed an integrated strategy combining re-mining of clinical cohort data with an HFD-induced dyslipidemia mouse model. We systematically investigated sex-related alterations in gut microbiota structure, host phenotypic manifestations, intestinal barrier integrity, adipose tissue inflammation, and immune responses, as well as serum metabolite profiles under dyslipidemic conditions. We furthermore isolated and characterized a candidate keystone bacterial strain from male mouse feces, evaluating its lipid-lowering potential through morphological observation, in vitro assays, and whole-genome sequencing. This study not only provides novel microbiota–metabolic insights regarding the sex-related differences of dyslipidemia but also offers a preliminary rationale for developing precision microecological intervention strategies tailored to sex-related lipid regulation.

2. Materials and methods

2.1. Reanalysis of a previous clinical cohort

To investigate sex-related alterations in the gut microbiota associated with dyslipidemia, we conducted a secondary analysis of a previously established clinical cohort to provide exploratory data (Xia et al., 2024). Participants were stratified according to serum total cholesterol (TC) levels and biological sex. Individuals with serum TC > 5.7 mmol/L were classified as having dyslipidemia, resulting in four groups: normal female (NF, n = 5), female with dyslipidemia (MF, n = 8), normal male (NM, n = 5), and male with dyslipidemia (MM, n = 8).

Raw 16S rRNA gene sequencing data were re-analyzed using the Majorbio Cloud Platform.1 Operational taxonomic unit (OTU) clustering of 16S rRNA gene sequences was based on 97% similarity. The Chao, observed species (Sobs), Shannon diversity, Simpson, and Good’s coverage indices were used to reflect the alpha diversity of samples. For β-diversity, non-metric multidimensional scaling (NMDS) analysis was performed to reveal the differences between groups based on Bray-Curtis distances and PERMANOVA (Vegan v2.4.3). Linear discriminant analysis effect size (LEfSe) was performed with a linear discriminant analysis (LDA) score threshold > 2 and p = 0.05 to screen for dominant microbial communities (Xu Y. et al., 2024; Yang et al., 2026; Xia et al., 2024). The microbial dysbiosis index (MDI) was calculated as previously described by Gunathilake et al. (2020), following data re-normalization and permutation using the CCREPE method. The relative abundances of taxa at different taxonomic levels (phylum and genus) were calculated and compared between groups using the Wilcoxon rank-sum test with Benjamini-Hochberg false discovery rate (FDR) correction (Q < 0.05). Inter-microbial associations and network construction were performed using SparCC, with significant correlations (p < 0.05) determined by Spearman’s correlation analysis and visualized with Gephi. Modules within each network are represented by distinct colors. Hub nodes were identified using the Maximal Clique Centrality algorithm in cytoHubba, and nodes with hub scores in the top 20 were designated as candidate keystone taxa (Banerjee et al., 2018; Xu Y. et al., 2024). Functional profiles were predicted using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2, version 2.2.0) based on the Greengenes database, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotations were used for comparative functional analysis.

2.2. Animals, diet, and experimental design

Thirty-two specific pathogen-free (SPF) C57BL/6 mice (16 females, 16 males; 5–6 weeks old) were obtained from the Animal Experiment Centre of Dalian Medical University. After one week of acclimatization under standard housing conditions, the mice were randomly assigned to four groups (n = 8 per group) according to sex and diet: normal female group (NF), female dyslipidemia group (MF), normal male group (NM), and male dyslipidemia group (MM). The NF/NM groups received a normal diet (3.79 kcal/g), whereas the MF and MM groups received HFD (4.5 kcal/g, 3% cholesterol) for 12 weeks (diets sourced from Jiangsu Synergy Pharmaceutical Bio-engineering Co., Ltd. and Jiangsu Medissen Biomedical Co., Ltd., respectively) (Supplementary Tables S1, S2). All experimental procedures were approved by the Animal Ethics Committee of Dalian Medical University (AEE24090) and were conducted in accordance with the institution’s Ethical Guidelines for the care and use of laboratory animals.

Body weight, food intake, and water consumption were recorded weekly throughout the experimental period. After the 12-week intervention, mice were fasted for 12 h before the final body weights were measured. Mice were euthanized by CO₂ asphyxiation, and the mesenteric lymph nodes (MLNs), liver, epididymal, and brown adipose tissues were collected and weighed to calculate the organ index (Xu Y. et al., 2024). Distal ileum and proximal colon tissues were fixed in formalin for histology, whereas additional ileal tissues and cecal contents were immediately snap-frozen in liquid nitrogen and stored at −80 °C until further analysis (Gândara et al., 2012; Nestor-Kalinoski et al., 2022).

2.3. Flow cytometry of lymphocyte subsets in mesenteric lymph nodes

Single-cell suspensions were prepared from mesenteric lymph nodes (MLNs) as previously described (Li et al., 2016; Aylward et al., 2019). Cells (1 × 106/mL) were incubated with fluorescently labeled antibodies against FITC-CD19+, APC-CD3+, and FITC-CD4 + (all at 1:100 dilution) for 30 min at 4 °C. For intracellular cytokine staining, cells were stimulated with phorbol 12-myristate 13-acetate (PMA), ionomycin, and brefeldin A, followed by fixation, permeabilization, and intracellular staining for IL-17A and Foxp3 using commercial kits (BD Biosciences and eBioscience). Cell phenotypes were analyzed using flow cytometry (CytoFlex, Beckman Coulter) (Huang et al., 2023b; Xu Y. et al., 2024).

2.4. Biochemical and inflammatory marker measurements

Blood was centrifuged (3,000 rpm, 15 min, 4 °C) to obtain serum for subsequent biochemical analyses. Hepatic tissues were homogenized in absolute ethanol (1:9 w/v) and centrifuged to collect supernatant. Serum and hepatic TC, as well as serum LDL-C and HDL-C, were quantified using commercial assay kits (Nanjing Jiancheng Bioengineering Institute). For inflammatory marker analysis, ileal tissues were homogenized in cold PBS (1:9 w/v) containing protease inhibitors and centrifuged to collect the supernatants. Serum and ileal levels of α-defensins, TNF-α, and LPS were determined using enzyme-linked immunosorbent assay (ELISA) kits (Quanzhou Ruixin Biotechnology Co., Ltd. and Jiangsu Mei Biao Bio-Technology Co., Ltd., China) (Huang et al., 2021; Xu Y. et al., 2024).

2.5. Histological and immunohistochemical analyses

Formalin-fixed, paraffin-embedded tissue sections (5 μm) from liver, ileum, and colon were processed for histological and immunohistochemical evaluation as previously described (Xu Y. et al., 2024). Liver tissues were stained with H&E (n = 4 per group; 100 × and 400×) and scored using the NAFLD Activity Score (NAS: steatosis 0–3, lobular inflammation 0–3, ballooning 0–2) according to a previous study (Kleiner et al., 2005; Huang et al., 2023a). Ileal villus height and crypt depth were measured from ≥ 20 villus-crypt units per section using Image-Pro Plus 6.0 (Wang L. et al., 2019; Liu et al., 2020; Xu Y. et al., 2024), and goblet cells were quantified by Alcian blue periodic acid Schiff (AB-PAS) staining (G1285, Solarbio) in four random fields per section (n = 3 per group) (Ren et al., 2022; Wu et al., 2022). Colon damage was scored blindly by two independent pathologists based on inflammation, crypt distortion, ulceration, immune infiltration, goblet cell loss, and mucosal thickening (Guo et al., 2023; Xu Y. et al., 2024).

For immunohistochemistry, deparaffinized ileal sections were subjected to antigen retrieval in citrate buffer, blocked, and incubated overnight at 4 °C with primary antibodies against IL-6, IL-10, TLR4, and F4/80 (1:200), followed by HRP-conjugated secondary antibodies and DAB detection. Immunofluorescence staining was performed using primary antibodies against ZO-1 and Occludin (1:200), followed by Alexa Coralite488- or Coralite594-labeled secondary antibodies (1:300, Wuhan Sanying Biotechnology Co., LTD.) and DAPI counterstaining (1:10000, Shanghai Biyuntian Biotechnology Co., LTD.). Integrated optical density (IOD) or fluorescence intensity was quantified using ImageJ (Huang et al., 2023b). All images were captured using an Olympus CX33/BX53 microscope.

2.6. Real-time fluorescence quantitative PCR reaction of ileum tissue

Total RNA was extracted from ileal tissues using TRIzol reagent, and cDNA was synthesized from 1 μg RNA according to established protocols. qPCR was performed with SYBR Green on a LineGene 9,600 Plus system (Hangzhou Bioer Technology). Gene expression was normalized to the internal reference gene and calculated using the 2-ΔΔCt method. The primer sequences are listed in Supplementary Table S3 (Huang et al., 2021; Xu Y. et al., 2024).

2.7. Serum non-targeted metabolomics

Serum untargeted metabolomics was performed using a methanol-acetonitrile-water (2:2:1) mixture and analyzed by UHPLC (Agilent 1,290) coupled with Q-TOF-MS (TripleTOF 6,600). Chromatographic separation was performed on a UPLC BEH column using mobile phases A (25 mM ammonium acetate/ammonium hydroxide) and B (acetonitrile) under both ESI + and ESI– modes. LC–MS data files were injected onto the analytical platform, and raw chromatographic and mass spectrometric data were processed using Progenesis QI software (Waters Corporation, Milford, MA, USA) for peak detection, feature extraction, retention time alignment, and peak integration. Then, the normalized data was generated for subsequent analysis. The Principal Component Analysis (PCA), orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) and other statistical analyses were performed using the ropls package in R (version 1.6.2). Using Pareto scaling, OPLS-DA was performed to assess class separation and model robustness, and this was followed by 200 permutation tests to validate the model and guard against overfitting. The R2Y and Q2 values of the true model were compared with those of 200 random models, and the p value was calculated as the proportion of random models with R2Y or Q2 equal to or greater than the true model. Based on the variable weights obtained from the OPLS-DA model (VIP) and the p-values obtained from the Student’s t-test, significant metabolites were identified. The screening criteria were set as |log2FC| ≥ 1, VIP ≥ 1, p < 0.05, with BH-FDR correction. Metabolite annotations were conducted using the human metabolome database (HMDB) and KEGG databases,2 and pathway enrichment analysis was performed using the Python software (scipy.stats package) (Xu Y. et al., 2024).

2.8. Gut microbiome analysis

Cecal microbial DNA (n = 6 per group) was extracted with the Foregene Fecal DNA Extraction kit for subsequent 16S rRNA gene sequencing and bacterial qPCR analysis. The V3-V4 hypervariable region of the 16S rRNA gene was amplified (primers 338F/806R) and sequenced on an Illumina NovaSeq 6,000 platform (Shanghai Personalbio Co., Ltd.). Raw sequencing data were processed using QIIME2 with the DADA2 pipeline for quality filtering, denoising, and amplicon sequence variant (ASV) generation. Taxonomic annotation was performed against the Greengenes database (version 13.8) using a Naive Bayes classifier. The α-diversity (Chao, Sobs, Shannon, Simpson and Good’s coverage indices), β-diversity (PCoA and NMDS based on Bray–Curtis distances, with PERMANOVA using 999 permutations), the relative abundances of taxa at different taxonomic levels (phylum and genus), LEfSe, network construction, and PICRUSt2 functional prediction were performed as described in Section 2.1 (Xu Y. et al., 2024; Yang et al., 2026; Xia et al., 2024), with the following differences: (1) ASV-level clustering was used instead of OTU-level clustering; and (2) LEfSe was performed with an LDA score threshold > 3.5. Correlations between microbial taxa were analyzed using Spearman’s correlation analysis, where |r| > 0.6 and p < 0.05 were considered significant. Network topology was analyzed and visualized using Gephi. Zipi analysis was performed to identify module hubs and connectors.

The relative abundances of Lactobacillus spp., Escherichia coli (E. coli), Desulfovibrio, and Lactobacillus johnsonii were quantified on a LineGene 9,600 Plus instrument by qPCR using genus- or species-specific primers (Supplementary Table S4). The 16S rRNA gene (amplified with primers 338F/806R) was used as the reference, and relative quantification was calculated using the 2-△△CT method (Huang et al., 2023a).

2.9. Integrated co-abundance network analysis

An integrated co-abundance network was constructed among the top 30 bacterial genera, serum metabolites, and host phenotypic parameters (serum lipids, LPS, TNF-α, liver TC, organ indices, flow cytometry data). Pairwise Spearman correlations with |r| > 0.6 and adjusted p < 0.05 (Benjamini-Hochberg) were visualized as an undirected network in Cytoscape (v3.9.1). Hub nodes were identified using the Maximal Clique Centrality algorithm in cytoHubba, and the 20 nodes with the highest MCC scores were designated as hub nodes. The resulting correlation matrix was used to construct an undirected co-abundance network, in which nodes represented microbial taxa, metabolites, or phenotypic parameters, and edges represented significant Spearman correlations.

2.10. Bacterial strain isolation and identification

Lactobacillus strains were isolated from fresh fecal samples collected from healthy C57BL/6 mice using Lactobacillus Selective Agar (LBS) plates (anaerobic conditions, 37 °C, 48 h). Colonies exhibiting typical Lactobacillus morphology were selected and purified by three successive rounds of streaking on de Man–Rogosa–Sharpe (MRS) agar to obtain pure cultures.

The isolate was identified by colony morphology, Gram staining, and 16S rRNA gene sequencing (primers 27F/1492R and 338F/806R). Quality-trimmed sequences were aligned to NCBI nucleotide reference sequences of type strains using BLASTn, with ≥ 99.0% sequence identity and an E-value < 1 × 10−5 (Supplementary Table S5). Based on concordant results (Supplementary Table S5), the isolate was identified as L. johnsonii and designated L. johnsonii DM2420, which was used in all subsequent experiments.

2.10.1. Growth and resistance to gastrointestinal environmental strains

An overnight culture of L. johnsonii DM2420 was inoculated (2.0%, v/v) into MRS broth at 37 °C. The optical density at 620 nm (OD620) was measured at 0, 1, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26 h.

The pre-activated culture of DM2420 was inoculated (5.0%, v/v) into MRS broth at pH 1.0–6.0 and incubated at 37 °C, with OD620 recorded at 0, 6, 12, and 24 h. All measurements were performed in triplicate (n = 3 per time point) (Assefa et al., 2015).

The method described by Gilliland and Walker was followed with minor modifications (Gilliland and Walker, 1990; Zhao et al., 2024). Briefly, DM2420 was cultured for 16–18 h, harvested by centrifugation at 6,000 rpm for 10 min at 4 °C, and washed three times with PBS (pH 7.0). The bacterial pellet was resuspended at 5.0% (v/v) in MRS broth containing 0.1%, 0.2 and 0.3% (w/v) bile salts and incubated at 37 °C. Samples were collected after 0, 6, and 12 h, and viable bacteria were enumerated using the standard plate count method. For simulated gastrointestinal tolerance, stationary-phase cultures were resuspended at 5.0% (v/v) in simulated gastric fluid or simulated intestinal fluid and incubated at 37 °C. Samples were collected after 0, 3, and 6 h, and bacterial survival was determined using the same plate counting procedure described for the bile salt tolerance assay. Survival rates were calculated using the following equation: Survival rate (%) = T_treatment1 / T_initial1 × 100%.

2.10.2. Qualitative direct plate BSH assay and cholesterol degradation assay

Sterile filter discs (6 mm) were placed on the surface of solidified MRS agar (supplemented with 3 g/L sodium taurodeoxycholate, 2 g/L sodium thioglycolate, 0.37 g/L CaCl₂). The plates were incubated at 37 °C for 48 h, after which the presence of opaque precipitation halos indicating deconjugated bile acids was assessed. MRS agar plates without conjugated bile acid supplementation served as controls (Kumar et al., 2011; Bhat and Bajaj, 2020).

A cholesterol stock solution (10 mg/mL in absolute ethanol) was filter-sterilized and added to sterile MRS liquid medium at a final concentration of 5% (v/v) to obtain MRS-CHOL medium (final cholesterol concentration 500 μg/mL). The activated DM2420 culture was adjusted to OD₆₂₀ = 0.5 and inoculated at 1% (v/v) into 10 mL of MRS-CHOL medium (three independent replicates). After anaerobic incubation at 37 °C for 24 h, bacterial cells were pelleted by centrifugation at 6,000 rpm for 10 min at 4 °C, and the supernatant was collected. Cholesterol concentrations in the supernatant were measured using a total cholesterol assay kit (micro-method, Sangon Biotech, Shanghai) according to the manufacturer’s instructions. Briefly, 20 μL of each standard or sample was mixed with 180 μL of working reagent in a 96-well plate, incubated at 37 °C for 15 min, and the absorbance at 500 nm was recorded. Cholesterol concentration was calculated from a standard curve using the formula: concentration (μg/mL) = concentration (μmol/mL) × 386.65 (molecular weight of cholesterol). The removal rate was calculated as (C₀ – Cₛ)/C₀ × 100% (Rudel and Morris, 1973; Madani et al., 2013; Bhat and Bajaj, 2020).

2.10.3. Antibiotic susceptibility test

The susceptibility of strain DM2420 to seven antibiotics (Penicillin, Erythromycin, Gentamicin, Furazolidone, Ceftazidime/avibactam, Clindamycin, and Amoxicillin/clavulanic acid) was determined using the Kirby–Bauer test (K-B method). Bacterial cultures were inoculated into fresh MRS broth and incubated anaerobically at 37 °C for 12–16 h. One hundred microliters of the activated bacterial suspension were evenly spread onto an MRS agar plate using a sterile cotton swab. Plates were left at room temperature for 5–10 min before sterile tweezers were used to place antibiotic discs onto the surface. Inhibition zone diameters (Supplementary Table S6) were measured after 24 h and interpreted per Clinical and Laboratory Standards Institute guidelines (Bauer et al., 1966; Duche et al., 2023).

2.10.4. Whole-genome sequencing and genomic analysis

Genomic DNA was extracted from 2 mL of an overnight culture using a genomic DNA purification kit (Workon, China). DNA integrity was verified by 1% agarose gel electrophoresis, and concentration and purity were assessed with a NanoDrop 2000 spectrophotometer. Purified DNA was subjected to whole-genome sequencing on the Illumina MiSeq platform. Raw reads were quality-filtered with fastp (v0.23.4) and assembled using reference genomes (Supplementary Table S7). Assembly quality was evaluated with QUAST (v5.2.0) (Boucard et al., 2022; Chen K. et al., 2023; Zhang et al., 2023; Wang K. et al., 2024). Circular genome maps were generated with BRIG (v0.95), and gene annotation was performed with Prokka (v1.14.6). Protein-coding sequences were functionally categorized using the Clusters of Orthologous Groups (COG) database via the CGView server, and KEGG pathway enrichment was analyzed with TBtools (v2.119) and visualized with ggplot2 (v3.5.1). Carbohydrate-active enzymes (CAZymes), antibiotic resistance genes, and virulence factors were annotated using dbCAN2, CARD/RGI, and VFDB, respectively. Putative probiotic-related genes were manually curated from Prokka, RAST, and KEGG annotations based on previous literature, and bacteriocin gene clusters were predicted using BAGEL4 (van Heel et al., 2018).

For phylogenetic analysis, core-genome SNPs were identified using Snippy (v4.6.0), and a maximum-likelihood tree was constructed and visualized with iTOL. ANI values were computed against reference genomes obtained from NCBI using the UPGMA clustering method (Chen K. et al., 2023).

2.10.5. Metabolomics of bacterial supernatant

The cell-free supernatants harvested from DM2420 cultures (L20 group) and blank MRS medium (C group) were subjected to untargeted metabolomics as described in 2.7 (Xu Y. et al., 2024). The data were analyzed on the online tool of Majorbio Cloud Platform.3

2.11. Statistical analysis

Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using SPSS Version 22 and GraphPad Prism Version 8. Two-group comparisons were made using Student’s t-test or Mann–Whitney U test, as appropriate. For multiple-group comparisons, one-way ANOVA followed by Tukey’s post hoc test (or Kruskal–Wallis test followed by Dunn’s test for non-parametric data) was used. For α-diversity indices, multiple t-tests with Benjamini–Hochberg FDR correction were performed in GraphPad Prism. Differences in KEGG pathway abundances were analyzed using Welch’s t-test in STAMP (v2.1.3) followed by BH-FDR correction. Microbial community differences were assessed using Adonis (PERMANOVA, 999 permutations) in R (v3.5.2). For Spearman correlation and co-abundance network analyses, BH-FDR-adjusted p < 0.05 was considered significant. A p value < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

3. Result

3.1. Differences in the structure of gut microbiota between male and female patients with dyslipidemia

Age and BMI were not significantly different between the four groups (p > 0.05). Serum lipid levels (TC, LDL) were significantly higher in the MF and MM groups than in the NF and NM groups (p < 0.05). There were no significant between-group differences in other serum parameters (TG, HDL, LPa, UA, HCY, FBG; p > 0.05, Supplementary Table S8).

At the OTU level, NMDS analysis revealed distinct clustering between male and female subgroups of elderly volunteers (Stress = 0.087, p = 0.001; Figure 1b). The Simpson index increased, while the Shannon index decreased in both HFD groups versus controls (p < 0.05; Figure 1a; Supplementary Figure S1A). The MDI index in the MF and MM groups was also significantly higher than in the NF and NM groups (p < 0.01; Supplementary Figure S1B), consistent with gut microbiota imbalance in these groups.

Figure 1.

Figure containing five panels comparing microbiome data: (a) Bar graph of Shannon index for four groups (NF, MF, NM, MM) with statistical significance; (b) NMDS plot showing clustering of microbial communities; (c) Bar graph of Firmicutes to Bacteroidota ratio; (d) Heatmap and dendrogram of bacterial genera abundance by group, color-coded with up or down regulation indicators; (e) Heatmap displaying relative abundance of Lactobacillus species among groups with significance markers.

Effect of the structure of gut microbiota between male and female patients with dyslipidemia. (A) The Shannon index at the OTU level. (B) Non-metric dimensional scaling (NMDS) with cluster. (C) Firmicutes/Bacteroidota. (D) The heatmap of the top 50 genera. (E) The heatmap at the species level of Lactobacillus spp. Data analysis is derived from clinical patients: Normal female group (NF, n = 5), female with dyslipidemia group (MF, n = 8), normal male group (NM, n = 5), and male with dyslipidemia group (MM, n = 8). *P < 0.05, **P < 0.01, ***P < 0.001.

To further explore sex-related differences in microbial composition, we analyzed taxonomic changes in the gut microbiota at the phylum and genus levels across the four groups. At the phylum level, Firmicutes, Bacteroidota, and Desulfobacterota dominated all groups (Supplementary Figure S1C). The Firmicutes/Bacteroidota ratio was significantly elevated in both MF and MM groups compared to their respective controls (p < 0.001; Figure 1c); while the relative abundance of Desulfobacterota was significantly increased in the MM group compared to the NM group (p < 0.05; Supplementary Figure S1C). Heatmap analysis of the top 50 genera preliminarily suggested reductions in the relative abundances of norank_f_Oscillospiraceae, Parabacteroides, Peptococcus, Bacteroides, and Rikenellaceae_RC9_gut_group in the MF group compared with the NF group (p < 0.05; Figure 1d; Supplementary Figure S1D), with similar trends observed in the MM group compared with the NM group (p < 0.05; Figure 1d). Conversely, the relative abundances of Lactobacillus, unclassified_f_Ruminococcaceae, Rikenella, Lachnospiraceae UCG-001, and Mucispirillum were significantly increased in both the HFD groups compared with the normal groups (p < 0.05; Figure 1d). Reductions observed in males included Blautia, norank_f_Muribaculaceae, norank_f_Lachnospiraceae, unclassified_f_Oscillospiraceae, and Lachnospiraceae_NK4A136_group (p < 0.05; Figure 1d). However, Alistipes, Odoribacter, Alloprevotella, and Desulfovibrio were significantly increased in the MM group (p < 0.05; Figure 1d).

Species-level analysis of Lactobacillus revealed that Ligilactobacillus murinus (L. murinus), Ligilactobacillus reuteri (L. reuteri), Ligilactobacillus salivarius (L. salivarius), Lactobacillus intestinalis (L. intestinalis), and Lactobacillus johnsonii (L. johnsonii) were the predominant species (Figure 1e; Supplementary Figure S1E). The relative abundance of L. murinus was significantly increased in both dyslipidemic groups compared with the NM and NF groups (Figure 1e). Moreover, the relative abundances of L. intestinalis and L. johnsonii were significantly lower in the MM group than in the NM group (p < 0.01; Figure 1e).

3.2. Identification of co-abundance networks of key microbial taxa and their potential functional implications in dyslipidemia

Using the top 50 bacterial taxa, co-occurrence networks were constructed via the SparCC algorithm (Figure 2a). For the female microbial network (33 nodes and 337 edges), module 2 hubs (Lactobacillus, Bacteroides, and Lachnospiraceae) exhibited high betweenness centrality and negative correlations with one another. A high hub value of Lachnospiraceae_NK4A136_group was positively correlated with Lachnoclostridium in module 1, possibly reflecting a shared ecological niche (Figure 2a; Supplementary Table S9). In the male network (24 nodes, 187 edges), norank_f_Muribaculaceae, norank_f_Lachnospiraceae, and Lactobacillus in module 3 displayed high hub indices, suggesting central ecological roles; norank_f_Desulfovibrionaceae, Lachnospiraceae_UCG_006, and Bacteroides in module 2 and Oscillospiraceae in module 5 also showed high hub values (Figure 2a; Supplementary Table S10). Overall, these network analyses preliminarily identified candidate keystone taxa (Lactobacillus and Lachnospiraceae) based on their high connectivity within the co-abundance networks, providing a basis for subsequent functional investigations.

Figure 2.

Panel a shows two circular network diagrams comparing microbial interactions in NF vs MF (left) and NM vs MM (right) groups, with nodes colored by module and edges representing correlation strength; Lactobacillus is highlighted in red. Panel b presents dot plots of enriched pathways with corresponding statistics for NF vs MF (top) and NM vs MM (bottom), with dot size and color indicating percent and p-values. Panel c displays scatter plots of genus-level relative abundance versus total cholesterol (TC) for NF vs MF (top) and NM vs MM (bottom), showing genus differentiation by color and shape. Panel d offers similar scatter plots for species-level Lactobacillus abundance versus TC for both cohorts, with individual species identified by different markers.

Integrated analysis of gut microbiota co-abundance networks, predicted metabolic pathways, and their correlations with serum cholesterol in dyslipidemic patients. (A) The co-abundance network. (B) STAMP analysis for the inferred level 3 metabolic pathways based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. (C) Spearman correlation analysis of the relative abundance of the top 10 genera with TC. (D) Spearman correlation analysis of Lactobacillus strains with TC at the species level. Normal female group (NF, n = 5), female with dyslipidemia group (MF, n = 8), normal male group (NM, n = 5), and male with dyslipidemia group (MM, n = 8).

Figure 3.

Scientific figure showing experimental design, growth curves, biochemical data, tissue images, and histology. Panel a diagrams mouse grouping and dietary intervention. Panel b shows body weight trends over 12 weeks with MM group gaining most weight. Panels c-h present bar graphs comparing serum, liver, and adipose tissue measurements across four groups: NF, MF, NM, MM. Panel i displays photographs of mouse livers and representative liver histology images at one hundred times and four hundred times magnification, highlighting cellular changes. Panels j-l show bar graphs quantifying liver weight, liver tissue, and NAFLD activity score, demonstrating significant differences among groups.

Effects of high-fat diet-induced dyslipidemia on apparent indicators and liver injury in mice. (A) Animal and experimental design. (B) Body weight change. (C) Serum TC level (mmol/L). (D) Serum LDL-C level (mmol/L). (E) Serum HDL-C level (mmol/L). (F) Liver TC level (mmol/L). (G) The epididymal fat index (mg/g). (H) The brown adipose tissue index (mg/g). (I) Morphological observations of the liver and H&E staining of liver tissue with 100 × and 400 × magnification. (J) Liver weight (g). (K) Liver index (mg/g). (L) NAS scores. Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), Dyslipidemia female group (MF), Normal male group (NM), and Dyslipidemia male group (MM). *P < 0.05, **P < 0.01, ***P < 0.001.

PICRUSt2-based functional prediction was performed to characterize differences in KEGG pathways between groups (Figure 2b). For KEGG level 3 analysis, in comparison with the normal groups, the abilities of MAPK signaling, lipopolysaccharide biosynthesis and glucose/lipid metabolism (GPI-anchor biosynthesis, PPAR signaling, and glycosphingolipid biosynthesis-globo and isoglobo series) were increased in both HFD groups.

3.3. Correlation between gut microbiota and lipid-related indicators

Spearman correlation analysis of serum TC with the top 10 genera and top 5 Lactobacillus species was analyzed (Figures 2c,d). In the female group, Lachnospiraceae_NK4A136_group, Alistipes, and Helicobacter were positively correlated with serum TC; while norank_f_Muribaculaceae, norank_f_Lachnospiraceae, and Lachnospiraceae_UCG-006 correlated negatively (p < 0.05; Figure 2c). In the males, Lactobacillus, Helicobacter, and Lachnospiraceae_UCG-006 were positively correlated with serum TC; whereas norank_f_Muribaculaceae, norank_f_Lachnospiraceae, Bacteroides, and norank_f_Desulfovibrionaceae were negatively correlated with serum TC (p < 0.05, Figure 2c). At the species level, L. murinus was positively correlated with serum TC in both sexes, whereas L. johnsonii was negatively correlated with serum TC only in males (p < 0.05; Figure 2d).

3.4. Effects of HFD-induced weight gain and universal dyslipidemia in male and female mice

Body weight increased significantly in the MF group starting from week 5 and in the MM group starting from week 3 (p < 0.05; Figure 3a, b), indicating that female and male mice exhibited distinct weight-gain trajectories. However, no significant differences were observed in food intake, energy intake, or water intake among groups (Supplementary Figure S2).

Both HFD groups exhibited significantly elevated serum TC and LDL-C levels, reduced serum HDL-C, and increased hepatic TC compared with controls (p < 0.05; Figures 3c–f), suggesting that HFD intervention induced dyslipidemia in both sexes.

3.5. Effects of HFD-induced dyslipidemia on hepatic injury in male and female mice

HFD-induced dyslipidemia significantly increased the epididymal fat index in both the MF and MM groups compared with the NF and NM groups (p < 0.05; Figure 3g). The brown adipose tissue index was significantly higher only in the males (p < 0.05; Figure 3h).

Macroscopically, HFD caused a pallid appearance in females and mild patchy discoloration in those of the MM group (Figure 3i). Histologically, HFD-fed mice displayed disorganized, structurally disrupted hepatocytes, accompanied by the presence of fat droplets or inflammatory infiltration. Liver weights were also significantly higher in the MF and MM groups than in the NF and NM groups, while the liver index was significantly higher in the MF group than in the NF group (p < 0.05; Figures 3j,k). Consistently, NAS scores were significantly higher in both HFD groups versus controls (p < 0.05; Figure 3l).

3.6. Effects of HFD-induced dyslipidemia on Th17/Treg dysregulation in male and female mice

In MLN tissue, HFD significantly reduced the number of CD3 + T lymphocytes and increased the number of CD19 + B lymphocytes in both sexes versus controls (p < 0.05; Figures 4a,c,d). We also examined CD4 + IL-17A + Th17 cells and CD4 + Foxp3 + Treg cells (Figure 4b). The proportion of CD4 + T lymphocytes was significantly elevated in the MF and MM groups compared with the NF and NM groups (p < 0.05; Figure 4e). The percentage of Treg cells was significantly decreased in the MF group compared with the NF group (p < 0.05), with a similar downward trend observed in male mice (p = 0.058; Figure 4f). Nonetheless, Th17 cell percentages did not differ significantly between groups (p > 0.05; Figure 4g). Consequently, the Th17/Treg ratio was significantly increased in both HFD groups (p < 0.05; Figure 4h).

Figure 4.

Multi-panel scientific figure showing flow cytometry dot plots, bar graphs, immunohistochemical staining, and violin plots comparing immune cell populations and protein expression across four groups labeled NF, MF, NM, and MM. Flow cytometry plots and corresponding quantification reveal differences in immune cell subsets, while tissue staining panels for F4/80, TLR4, IL-6, and IL-10 with quantification graphs highlight expression in intestinal tissue, marked by red arrows. Statistical significance is indicated in several graphs.

Effect of high-fat diet-induced dyslipidemic mice on inflammatory responses. (A) The expression of T lymphocytes (CD3+) and B lymphocytes (CD19+). (B) The expression of Th17 (CD4 + IL-17A+) and Treg (CD4 + Foxp3+) T lymphocytes. (C–G) The percentage of CD3 + (C), CD19 + (D), CD4 + (E), Treg (F), and Th17 (G) in the MLN (%). (H) Th17/Treg ratio. (I) Serum TNF-α level (pg/mL). (J) Serum LPS level (EU/mL). (K) Serum α-defensins level (ng/mL). (L) Ileum α-defensins level (ng/mL). (M) The mRNA expression of Mapk in the ileum. (N) The mRNA expression of NF-κB p65 in the ileum. (O) Ileum immunohistochemistry of the F4/80, TLR4, IL-6, and IL-10 proteins (400 × magnification) with the area of F4/80, TLR4, IL-6, and IL-10 (%). Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), Dyslipidemia female group (MF), Normal male group (NM), and Dyslipidemia male group (MM). *P < 0.05, **P < 0.01, ***P < 0.001.

3.7. Effects of HFD-induced dyslipidemia on inflammatory responses via MAPK/NF-κB pathway activation in male and female mice

HFD significantly increased serum TNF-α and LPS levels, while reducing α-defensins expression in both serum and ileal tissue compared with their respective control groups (NF and NM; p < 0.05; Figures 4i–l).

In ileal tissues, the positive staining area for F4/80, IL-6, and toll-like receptor 4 (TLR4) was significantly higher in both HFD groups compared with control groups (p < 0.05; Figure 4o). Compared with the NF group, the IL-10 positive expression was only significantly decreased in the MF group (p < 0.05; Figure 4o). Furthermore, the mRNA expression levels of Mapk and NF-κB p65 were significantly up-regulated in both HFD groups compared with control groups (p < 0.05; Figures 4m,n).

3.8. Effects of HFD-induced dyslipidemia on intestinal structural damage in male and female mice

HFD-fed female and male mice exhibited pathological changes, including significant thinning of the basal lamina (Figure 5a). The MF group displayed inflammatory infiltration within the lamina propria, while the MM group showed marked villus shortening and epithelial shedding. Quantitative analysis demonstrated significant reductions in villus length and crypt depth in HFD groups compared with controls (p < 0.05; Figures 5d,e). AB-PAS staining revealed a decrease in goblet cell density in the MF and MM groups compared with their respective controls (p < 0.05; Figures 5b,f). Consistent findings were observed by H&E staining of colon tissue (p < 0.05; Supplementary Figure S3). Moreover, immunofluorescence analysis of tight junction proteins (Figure 5c) showed significantly reduced ZO-1 and Occludin expression in the male HFD groups (p < 0.05); whereas no significant difference was detected in the female HFD groups (p > 0.05; Figures 5g,h).

Figure 5.

Panel a shows histological images of mouse intestinal cross-sections stained with hematoxylin and eosin at 100X and 400X magnifications for four experimental groups labeled NF, MF, NM, and MM; red boxes and arrows highlight areas of interest. Panel b presents similar sections stained with a different method, also at 100X and 400X magnifications, with notable cellular structures indicated by arrows and boxes. Panel c contains immunofluorescence images displaying ZO-1 and Occludin protein localization in intestinal tissues for all groups. Panels d through h exhibit bar and scatter plots quantifying villus height, crypt height, goblet cell area, and fluorescence intensity for ZO-1 and Occludin, including statistical significance markers.

Effect of high-fat diet-induced dyslipidemic mice on intestinal structural damage. (A) Morphologic analysis of the ileum by H&E staining (100× and 400×). (B) AB-PAS staining of the ileum (400× magnification). (C) Immunofluorescence staining for ZO-1 and Occludin protein in ileum tissue (100×). (D) Height of villi (μm). (E) Height of crypt (μm). (F) The positive area of goblet cell (%). (G, H) The relative fluorescence intensity of ZO-1 (G) and Occludin (H) (%). Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), Dyslipidemia female group (MF), Normal male group (NM), and Dyslipidemia male group (MM). *P < 0.05, **P < 0.01, ***P < 0.001.

3.9. Related alterations in gut microbiota composition in HFD-induced male and female dyslipidemic mice

High-throughput sequencing of the V3 and V4 hypervariable regions of the 16S rRNA gene revealed sufficient sequencing depth for all samples by Good’s coverage index at the ASV level (Figure 6; Supplementary Figure S4A). HFD feeding significantly reduced microbial richness and diversity, as reflected by decreased Chao1, Shannon, and Observed_species indices compared with their respective controls (p < 0.05; Supplementary Figures S4A,B; Figures 6a,b). Hierarchical clustering separated samples by diet and sex, and PCoA (Adonis R2 = 0.5699, p < 0.001) with NMDS (stress = 0.0658) revealed distinct clustering patterns among the four groups (Figure 6c; Supplementary Figures S4C,D).

Figure 6.

Scientific figure displaying microbiome data comparisons among four groups. Panels a and b show box plots of observed and Shannon index alpha diversity metrics, with significant differences indicated. Panel c presents a principal coordinates analysis (PCoA) plot showing group clustering. Panel d depicts a bar graph comparing Firmicutes/Bacteroidota ratios. Panels f, g, h, i, and j display bar charts of relative abundance for different bacterial taxa, highlighting significant differences. Panel e contains a clustered heatmap summarizing differential abundance of microbial genera across groups, annotated by genus, family, and phylum, with a color scale representing log fold change.

Effect of high-fat diet-induced dyslipidemia on the structure of gut microbiota in mice. (A) The Chao1 index at the ASV level. (B) Shannon index at the ASV level. (C) Principal coordinates analysis (PCoA) with cluster. (D) Firmicutes/Bacteroidota. (E) The heatmap of the top 50 genera. (F) The relative abundance of Escherichia coli (%) based on the 16S rRNA gene. (G–J) The relative abundance of Lactobacillus (G), Desulfovibrio (H), Escherichia coli (I), and Lactobacillus johnsonii (J) (%) based on the bacterial qPCR analysis of cecal contents. Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), Dyslipidemia female group (MF), Normal male group (NM), and Dyslipidemia male group (MM). *P < 0.05, **P < 0.01, ***P < 0.001.

Taxonomic profiling further revealed marked sex-related alterations in gut microbial composition at the phylum and genus level (Supplementary Figures S4E,G). At the phylum level, the relative abundance of Bacteroidetes was significantly reduced in the MM group compared to the NM group (p < 0.05; Supplementary Figures S4E,F). The Firmicutes/Bacteroidetes ratio and relative abundance of Desulfurovibrio_I was increased in both HFD groups compared to their respective controls (p < 0.05; Figure 6d; Supplementary Figure S4F). At the genus level, HFD consistently increased the relative abundance of Bacteroides_H and Parabacteroides_B in both sexes (p < 0.05; Figure 6e; Supplementary Figure S4G), whereas decreasing Duncaniella, Prevotella, and Ileibacterium (p < 0.05; Figure 6e). The relative abundance of Mailhella exhibited an increasing trend in the MF group (p = 0.06) and was significantly enriched in the MM group (p < 0.05; Figure 6e and Supplementary Figure S4G). Male-predominant increases were observed in levels of Sporofaciens, Dysosmobacter, and Adlercreutzia, whereas Lactobacillus, Limosilactobacillus, and Dubosiella were significantly reduced (p < 0.05; Figure 6e; Supplementary Figure S4G). qPCR analysis further validated the sequencing results, confirming the changes in Lactobacillus and Desulfovibrio abundance at the genus level (p < 0.05; Figures 6g,h). At the species level, although 16S rRNA gene sequencing indicated that Escherichia coli was significantly enriched in the MF group relative to the NF group (p < 0.05; Figure 6f), qPCR analysis revealed a more pronounced alteration in E. coli levels between the NM and MM groups (p < 0.05; Figure 6i). L. johnsonii was identified as the predominant Lactobacillus species and was significantly reduced in the male dyslipidemic group (MM vs. NM, p < 0.05; Supplementary Figure S4H), corroborated by qPCR results (p < 0.05; Figure 6j).

Linear discriminant analysis effect size (LEfSe, LDA > 3.5) revealed that Muribaculaceae, Lachnospiraceae_UBA3282, and Lactobacillus were significantly enriched in the NF group. In contrast, the MF group was characterized by Desulfovibrionaceae, Desulfobacterota, and Bacteroides. The NM group was predominantly represented by Muribaculaceae, Lactobacillaceae, Muribaculaceae_g_CAG_485, and Lactobacillus, while Mailhella, Desulfovibrionaceae, Desulfobacterota, Actinobacteriota, and Bacteroides were identified as key discriminative taxa in the MM group (Supplementary Figure S4I).

3.10. Characterization of gut microbiota co-abundance networks and their predicted functional signatures in dyslipidemic mouse cohorts

To identify sex-associated keystone species, co-abundance networks were constructed for female and male mice. In the females (Figure 7a; Supplementary Figure S5; Supplementary Table S11), Muribaculaceae_g_CAG_485 (module 3) and Bifidobacterium (module 6) were dominant and showed strong inter-module correlations. Lactobacillus (module 1) and Mailhella (module 2) exhibited high hub scores, suggesting key ecological roles within the network. In the males, major nodes were concentrated in module 6, primarily represented by Lactobacillus (module 2) and Mailhella (module 1), indicating potential ecological importance (Figure 7a; Supplementary Figure S5; Supplementary Table S12). Mailhella showed correlation coefficients greater than 0.9 with other bacterial taxa. The Zipi diagram analysis further identified L. johnsonii as a taxon with a high module connectivity index (Supplementary Figure S5). Integrating the network topology with differential abundance analysis identified Lactobacillus and Mailhella as candidate keystone taxa, with L. johnsonii acting as a dominant species.

Figure 7.

Panel a displays two microbial community correlation network diagrams comparing NF vs MF and NM vs MM, with nodes representing taxa, colored by modules, and edges indicating correlation strengths; certain genera such as Lactobacillus and Bifidobacterium are highlighted in red. Panel b shows two corresponding functional enrichment bubble plots, with metabolic pathways on the Y-axis, proportion on the X-axis, bubble size representing percentage, and color indicating p-values, illustrating significant pathway differences between groups.

Network analysis and the effect of the metabolic pathways’ prediction in gut microbiota for high-fat diet-induced dyslipidemic mice. (a) The co-abundance networks. (b) STAMP analysis for the inferred level 3 metabolic pathways based on the KEGG database. Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), dyslipidemia female group (MF), normal male group (NM), and dyslipidemia male group (MM).

PICRUSt2 functional prediction (Figure 7b) showed that females enriched glucose/lipid metabolism pathways (pyruvate, polyketide sugar, phosphonate/phosphinate, ether lipid, linoleic acid) (p < 0.05). The gut microbiota metabolites in the male group were mainly involved in the following major signaling pathways: secondary bile acid biosynthesis, PPAR signaling, insulin signaling, fatty acid biosynthesis, and glucagon signaling (p < 0.05). In addition, the immune-related pathways (MAPK signaling pathway and CAMP antimicrobial peptide resistance) were significantly enriched in the MF and MM groups compared with the NF and NM groups (p < 0.05, Figure 7b).

3.11. Effects of serum metabolite profiles in high-fat diet-induced dyslipidemic mice

Unsupervised PCA and OPLS-DA analyses (Figures 8a,b) revealed clear metabolic separation between NF/MF and NM/MM groups. Volcano plot analysis (Figure 8c) identified 448 differentially expressed metabolites in females (355 up, 93 down) and 217 in males (60 up, 157 down). In female mice, a significant upregulation of multiple bile acid derivatives was observed, including ɑ-muricholic acid, hyocholic acid, allolithocholic acid, 7-ketocholesterol, with corresponding enrichment of ascorbate and aldarate metabolism pathways as well as nucleotide sugar biosynthesis (Figure 8d). In male mice, the significant down-regulation of multiple bile acid metabolites (e.g.,ω-muricholic acid, cholic acid, ursocholic acid, allocholic acid, and 7-ketodeoxycholic acid) was accompanied by KEGG-enriched pathways related to purine metabolism, primary bile acid biosynthesis, and bile secretion pathway (Figure 8e).

Figure 8.

Panel a and b show two sets of principal component analysis (PCA) scatterplots comparing three groups (MF, NF, MM) with ellipses indicating group separation. Panel c presents two volcano plots displaying significantly changed metabolites between groups, colored by up- and downregulation. Panel d includes a horizontal bar chart of differentially abundant metabolites and a dot plot of enriched metabolic pathways for MF vs NF. Panel e follows the same format for MM vs NM groups. Panel f and g display network diagrams illustrating relationships among microbiota, metabolites, and clinical indicators, using colored nodes and edges differentiated by correlation and relationship type.

Effect of high-fat diet-induced dyslipidemic mice on the composition of serum metabolites. (a,b) The unsupervised PCA (a) and OPLS-DA (b) analyses. (c) The volcano plot based on OPLS-DA (MF vs. NF and MM vs. NM). (d) KEGG enrichment of the top 20 serum differential metabolites (by log₂FC) in female mice. (e) KEGG enrichment of the top 20 serum differential metabolites (by log₂FC) in male mice. (f,g) Integrated co-abundance network analysis among gut microbiota, serum metabolome, and phenotypic parameters in dyslipidemic female (f) and male (g) mice, respectively. Data analysis is derived from mouse experiments (n = 8): Normal female group (NF), Dyslipidemia female group (MF), Normal male group (NM), and Dyslipidemia male group (MM).

3.12. Correlation analysis between serum metabolites, gut microbiota, and host phenotypic indicators

Spearman’s correlation analysis was performed to construct an integrated correlation network linked to the gut microbiota, serum metabolites, and host phenotypic parameters (Figures 8f,g; Supplementary Figure S6). In female mice, bile acid-related metabolites (e.g., alpha-muricholic acid, hyocholic acid, allolithocholic acid) were negatively correlated with ileal α-defensins, serum HDL-C, and Akkermansia. Conversely, these bile acid metabolites were positively correlated with serum LPS, serum TC, serum LDL-C, and liver TC, and Mailhella (Figure 8f; Supplementary Figure S6; Supplementary Table S13). In males, omega-muricholic acid, cholic acid, ursocholic acid, and allocholic acid were positively correlated with liver TC, serum TC, serum LDL-C, LPS, and Lactobacillus. However, these metabolites were negatively correlated with ileal α-defensins, serum HDL-C, CD4 + T cells, and Mailhella (Figure 8g; Supplementary Figure S6; Supplementary Table S14). Thus, sex-associated patterns strongly shaped these correlations, potentially suggesting bile acid metabolism as a potential interface between host and microbiota.

3.13. Identification and probiotic characteristics of Lactobacillus johnsonii DM2420

The morphology and cultured colonies of L. johnsonii DM2420 are consistent with the reported characteristics of L. johnsonii (Supplementary Figure S7A). The growth characteristics of L. johnsonii DM2420 were evaluated by monitoring optical density over time. The strain exhibited a short lag phase during the first 2 h of incubation, followed by rapid exponential growth between 2 and 12 h; growth reached a stationary phase after approximately 12 h (Supplementary Figure S7B).

3.13.1. Evaluation of the probiotic potential

The survival of strain DM2420 was markedly reduced at pH ≤ 3.0, suggesting limited tolerance to highly acidic conditions. In contrast, the strain exhibited good survival at pH values above 3.0 (Supplementary Figure S7C). The strain DM2420 exhibited the highest bile salt tolerance after incubation for 6 h in MRS medium containing 0.2% bile salts, with a survival rate of (86.97 ± 6.63)%. After 12 h of incubation in MRS broth containing 0.3% bile salts, DM2420 maintained a survival rate of 32.85 ± 4.49%, indicating sustained tolerance under higher bile salt concentrations (Supplementary Table S15).

Moreover, the survival rates of strain DM2420 after exposure to simulated gastric fluid for 3 and 6 h were (50.30 ± 4.93)% and (37.34 ± 1.49)%, respectively. After treatment with simulated intestinal fluid for 3 and 6 h, the survival rates were (52.29 ± 3.35)% and (50.38 ± 3.23)%, respectively (Supplementary Table S16). Strain DM2420 exhibited a cholesterol removal rate of (31.92 ± 7.58)% in MRS-CHOL medium (Supplementary Table S17) and produced precipitation on MRS agar containing sodium taurodeoxycholate, indicating bile salt hydrolase activity (Supplementary Figure S7D).

Antimicrobial susceptibility testing showed that DM2420 was sensitive or moderately sensitive to penicillin, ceftazidime/avibactam, erythromycin, furazolidone, and clindamycin, whereas it was resistant to gentamicin (Supplementary Table S18).

3.13.2. Genomic features of strain DM2420

Whole-genome sequencing revealed that DM2420 contains one chromosome (2,021,883 bp) and one plasmid (15,273 bp), with G + C contents of 34.60 and 34.29%, respectively (Figure 9a; Supplementary Table S19). We furthermore identified 80 tRNAs genes, 21 rRNAs, and 1 tmRNA (Supplementary Table S19). Strain DM2420 (DL01) showed an average nucleotide identity (ANI) of 99.998% relative to other reported L. johnsonii strains, well above the 95% species-delineation threshold, and UPGMA-based phylogenetic analysis placed it within the same major clade as 18 L. johnsonii strains (Supplementary Figures S7E,F), confirming its close relationship to this species. Screening against the CARD database predicted six resistance genes, including arlS, patB, murB, treR, valS, and tetM, whereas no virulence genes were detected by VFDB screening (Supplementary Table S20).

Figure 9.

Multipanel scientific figure showing: (a) circular genome and plasmid maps annotated with gene functions; (b) bar chart of gene counts in COG categories; (c) bar chart of gene distribution at KEGG metabolic pathway level 3; (d) linear diagrams of genomic loci with gene annotations; (e) bar chart and pie chart summarizing CAZy family gene distributions; (f) OPLS-DA scores plot comparing sample groups; (g) volcano plot highlighting significant changes in gene expression; (h) bar graph displaying Z-scores of metabolites in glycerophospholipid metabolism; (i) bar graph for cholesterol metabolism metabolites between two groups; (j) bubble plot of KEGG enrichment analysis for differential pathways.

Integrative genomic, functional, and metabolomic characterization of Lactobacillus johnsonii DM2420. (a) Circular genome and plasmid map of L. johnsonii DM2420. The chromosomal genomic map consists of five concentric circles. From the outermost to the innermost layer, Circle 1 and Circle 2 show the distribution of coding genes on the forward strand and reverse strand, respectively, including CDS (purple), tRNA (orange), rRNA (blue), and tmRNA (pink); Circle 3 corresponds to GC content; Circle 4 and Circle 5 represent GC Skew (+) and GC Skew (−), respectively. (b) Distribution of genes across COG functional categories. (c) KEGG pathway classification of the annotated genes. (d) Genomic localization of bacteriocin biosynthesis gene clusters. (e) Carbohydrate-active enzyme (CAZyme) profile. (f) OPLS-DA score plot of differential metabolites in cell-free supernatants. (g) Volcano plot. (h) Relative abundance of differential metabolites involved in phospholipid metabolism. (i) Abundance changes of cholesterol-related differential metabolites. (j) KEGG tertiary-level pathway enrichment analysis.

COG annotation predicted a total of 1,589 genes (Figure 9b), most prominently associated with translation, carbohydrate metabolic processes, transcription, carbohydrate transport and metabolism, and replication, recombination, and repair. KEGG annotation identified 1,074 genes, including those involved in carbohydrate metabolism (139 genes) and membrane transport (128 genes) (Figure 9c). In addition, 99 carbohydrate-active enzymes were identified, including 53 glycoside hydrolases and a substantial number of glycosyltransferases (Figure 9e).

The DM2420 genome also harbored genes associated with tolerance to acid, bile salts, high temperatures, and oxidative stress, as well as genes linked to anti-inflammatory activity, adhesion, and carbohydrate metabolism (Supplementary Table S20). Furthermore, several putative bacteriocin gene clusters were predicted in the DM2420 genome, including enterolysin A and Bacteriocin helveticin J (Figure 9d).

3.13.3. Bioactive metabolites in the cell-free supernatant of L. johnsonii DM2420

To characterize the metabolic activity of L. johnsonii DM2420, we performed untargeted metabolomics on its cell-free supernatant. Unsupervised PCA and OPLS-DA revealed distinct clustering between the DM2420 and control groups [OPLS-DA: R2 = (0, 0.9421), Q2 = (0, −0.0284); R2X = 0.811, R2Y = 1, Q2 = 0.991, p < 0.05], indicating significant metabolic differences (Figure 9f; Supplementary Figure S7G). A total of 820 differentially abundant metabolites were identified (|log₂FC| ≥ 1, VIP ≥ 1, p < 0.05), of which 296 were up-regulated and 524 down-regulated in the DM2420 supernatant relative to the control (Figure 9g). HMDB classification showed that these metabolites were predominantly enriched in amino acids and peptides (196 metabolites), followed by carbohydrates and fatty acids (Supplementary Figure S7I).

Pathway enrichment analysis indicated that the affected pathways were primarily involved in glycerophospholipid metabolism, glycolysis/gluconeogenesis, and the pentose phosphate pathway (Figures 9h–j). DM2420 supernatant showed an altered phospholipid profile, with low abundance of multiple lysophosphatidylcholine (LPC) and phosphatidylglycerol (PG) species (Figure 9h). KEGG pathway analysis further revealed that the identified compounds were mainly associated with lipid metabolism (67 compounds), peptides (36), and hormones/transmitters (17) (Supplementary Figure S7H). Within the bile acid biosynthesis pathway, the DM2420 supernatant substantially decreased levels of glycine-conjugated bile acids (e.g., glycodeoxycholic acid, glycochenodeoxycholic acid 3-glucuronide), whereas free bile acids, including cholic acid, taurodehydrocholic acid, and glycocholic acid 3-sulfate, were significantly increased (Figure 9i). These findings suggest that DM2420 possesses the in vitro capacity to alter bile acid and phospholipid profiles, which may contribute to its potential lipid-lowering properties; however, in vivo validation is required to confirm these effects.

4. Discussion

Diet is a well-known regulator of the gut microbiota. Beyond its direct effects on body weight and organ physiology, HFD can produce indirect effects, including gut microbiota composition and metabolic imbalance (Bourdeau-Julien et al., 2023; Wang Y. et al., 2024). In this study, we characterized sex-related alterations in gut microbiota, host metabolism, and immune responses associated with the transition from a normal to a dyslipidemic state.

HFD-feeding successfully induced dyslipidemia in both female and male mice, characterized by elevated serum TC and LDL-C, alongside decreased HDL-C. This was accompanied by histological liver damage and an increased epididymal fat index, confirming a state of metabolic disorder (Deng et al., 2019). However, the trajectory of metabolic decompensation diverged markedly between sexes. Male mice exhibited significant weight gain from week 3 onward, coupled with an increased brown adipose tissue (BAT) index, whereas female mice showed delayed weight gain (week 5) without BAT changes. This earlier and more pronounced metabolic response in males suggests heightened sensitivity to HFD-induced metabolic stress, consistent with the sexual dimorphism in adipose tissue mitochondrial function and energy expenditure reported by MacCannell et al. (2021). This early divergence in metabolic dynamics provides a temporal context for the sex-specific immune and microbial alterations discussed below.

Dyslipidemia triggered systemic inflammation, evidenced by elevated serum LPS and TNF-α, consistent with microbial translocation resulting from compromised intestinal barrier integrity (Chen X. et al., 2023). The Th17/Treg ratio showed a significant increase, driven primarily by reduced Treg proportions rather than Th17 expansion. This immune imbalance was particularly pronounced in female mice, suggesting sex-related regulation of immune tolerance that may coexist paradoxically with their otherwise metabolically resilient phenotype (Lefebvre et al., 2024; Xu J. et al., 2024). Concurrently, HFD activated the TLR4/NF-κB/MAPK pathway in ileal tissue, as demonstrated by upregulated F4/80, IL-6, and TLR4 expression, and increased Mapk and NF-κB p65 mRNA levels. Notably, ileal IL-10, a compensatory anti-inflammatory cytokine, decreased significantly only in female mice, which may indicate that females deploy distinct immunoregulatory mechanisms; however, this requires further investigation.

The intestinal mucosal barrier is reinforced by goblet cells and Paneth cell-derived antimicrobial peptides (including α-defensins) which also limit excessive TLR4/NF-κB-driven inflammation via competitive binding to TLR4 (Miani et al., 2018; Huang et al., 2019). HFD-induced villous destruction and depletion of these defensive cell populations compromised this barrier in both sexes, disrupting physical and chemical barrier integrity and potentially unleashing unchecked inflammatory signaling (Yang and Yu, 2021; Fu et al., 2023). However, a critical sex difference emerged at the physical barrier level, including ZO-1 and Occludin expression, which was significantly downregulated exclusively in male dyslipidemic mice. This sex-related defect in tight junction integrity may contribute to excessive LPS translocation, while also intensifying TLR4/NF-κB-driven inflammation in males. In contrast, female mice maintained physical barrier integrity, implicating immune cell imbalance rather than barrier breakdown as the primary driver of their inflammatory dysregulation. Estrogen-mediated upregulation of tight junction proteins is a plausible mechanism underpinning this female-protective phenotype (Wang et al., 2021).

At the community level, dyslipidemia was consistently associated with reduced α-diversity and elevated Firmicutes/Bacteroidetes ratios across both human and mouse cohorts, corroborating previous reports (Li et al., 2022; Huang et al., 2023a). Sulfate-reducing bacteria (SRB, including Desulfovibrio and Mailhella) were consistently expanded in dyslipidemic conditions, suggesting that SRB overgrowth is a microbial signature of lipid metabolic disturbance (Zhang et al., 2021; Dohrn, 2024). Although overall Lactobacillus abundance increased in dyslipidemic subjects, this was driven primarily by L. murinus, which positively correlated with TC and may exert opportunistic pathogenic effects. Overgrowth of L. murinus was linked to metabolic impairment and hair loss following antibiotic-induced dysbiosis (Hayashi et al., 2017), as well as age-related behavioral changes (Qiao et al., 2023). In sharp contrast, L. johnsonii showed the reverse trend: it was predominantly depleted in males in both mice and patients, and its abundance was strongly inversely correlated with TC, aligning with its reported cholesterol-lowering function via improving hepatic cholesterol metabolism (Yoon et al., 2023). Its male-predominant depletion was strongly associated with dyslipidemia in males, suggesting that L. johnsonii may be a candidate microbial factor relevant to sex-related lipid abnormalities. However, causality remains to be established.

Notably, sex-related discrepancies in microbial alterations were observed. Female patients with hyperlipidemia preliminarily exhibited reduced abundances of Lachnospiraceae_NK4A136_group and Blautia. High abundance of Lachnospiraceae_NK4A136_group correlates with reduced intestinal permeability and plasma LPS levels, thereby enhancing barrier function (Li et al., 2022). In addition, HFD-induced obesity-associated genera, including Rikenella and Mucispirillum (Wang et al., 2020; Zhang et al., 2021; Bai et al., 2022), were significantly increased, consistent with previous reports. Conversely, the butyrate-producing genus Oscillospira, which mediates cholesterol metabolism and reverse cholesterol transport, was significantly reduced in hyperlipidemic patients, potentially impairing cholesterol homeostasis (den Besten et al., 2015; Gophna et al., 2017; Jia et al., 2023).

Co-abundance network analysis further revealed sex-related ecological architectures within the gut microbiota, with Lactobacillus, Muribaculaceae, and Lachnospiraceae serving as highly connected hubs in females and males, respectively. PICRUSt2-based functional prediction was consistent with the observed differences, suggesting that fatty acid degradation and glycerolipid metabolism may be the potential key pathways altered in the HFD groups. Therefore, these results suggested that dysbiosis of dyslipidemia-associated gut microbiota is associated with alterations in host inflammation and energy metabolism.

Serum metabolomics analysis further complemented the microbiota findings by revealing pronounced sex-related metabolic alterations associated with high-fat diet-induced dyslipidemia. Bile acids are recognized as central mediators linking high-fat diet consumption, gut microbiota dysbiosis, and host metabolic homeostasis (Zheng et al., 2017). Accordingly, the distinct bile acid profiles observed in male and female mice suggest that bile acid metabolism may represent a potential interface connecting microbial alterations with host metabolic phenotypes. Female dyslipidemic mice exhibited increased levels of several bile acid species, including α-muricholic acid, accompanied by enrichment of the ascorbate metabolism pathway. This metabolic signature may reflect an adaptive response to high-fat diet-induced oxidative stress through enhanced antioxidant capacity. Ascorbic acid has been reported to protect tissues against oxidative damage by reducing cytochrome C release, thereby suppressing reactive oxygen species generation and lipid peroxidation (Gęgotek and Skrzydlewska, 2022). Conversely, male mice exhibited a comprehensive down-regulation of various primary bile acids including cholic acid and chenodeoxycholic acid, with the enriched pathways mainly focusing on primary bile acid synthesis and bile secretion. Given that bile acid synthesis, transformation, and recycling are tightly regulated by the gut microbiota (Zheng et al., 2017; Collins et al., 2023), these findings further support an association between sex-related microbial dysbiosis and altered bile acid metabolism. Correlation network analysis further showed that bile acid metabolites were associated with distinct microbial taxa in a sex-related manner. In male mice, several bile acids, including ω-muricholic acid, cholic acid, ursocholic acid, and allocholic acid were positively correlated with overall Lactobacillus abundance, and adverse lipid parameters and systemic inflammatory markers. This appears paradoxical because L. johnsonii was markedly depleted in dyslipidemic male mice. However, this apparent discrepancy likely reflects species-specific functional heterogeneity within the Lactobacillus species. Whereas L. johnsonii may promote cholesterol-lowering bile acid metabolism, other species including L. murinus may exert opposing metabolic associations. In female mice, specific bile acid metabolites were primarily associated with Akkermansia and Lachnospiraceae, suggesting a distinct microbiota–bile acid network that may be associated with the greater metabolic adaptability observed in females.

The depletion of L. johnsonii preliminarily was observed in the clinical cohort, and subsequently confirmed in the HFD-induced mouse model. Together with its high connectivity within the microbial co-abundance network, this depletion identifies this species as a candidate keystone taxon. We therefore isolated and comprehensively characterized L. johnsonii DM2420 as a candidate probiotic strain. Whole-genome sequencing confirmed its taxonomic identity (ANI 99.998% against reference strains) and revealed a complete repertoire of translation-related genes, indicating robust protein synthesis capacity. In addition, the genome harbored genes associated with adhesion, acid and bile tolerance, and bacteriocin production, while no known virulence determinants were detected, supporting the strain’s safety-related genomic profile and potential for probiotic application. Functionally, DM2420 exhibited strong BSH activity and a cholesterol removal rate of 31.92 ± 7.58% in vitro. These functional characteristics were further supported by untargeted metabolomic profiling of the bacterial culture supernatant, which demonstrated substantial remodeling of bile acid composition. Specifically, glycine-conjugated bile acids, including glycodeoxycholic acid and glycochenodeoxycholic acid 3-glucuronide, were markedly reduced, whereas unconjugated or secondary bile acid species, including cholic acid, taurodehydrocholic acid, and glycocholic acid 3-sulfate, were correspondingly altered, a metabolic pattern consistent with BSH-mediated deconjugation of conjugated bile acids. A separate study on another L. johnsonii CCFM1376 provided a rationale for investigating BSH activity in DM2420. CCFM1376 strain with high BSH activity has been shown to reduce serum TC and LDL-C in hypercholesterolemic mice via the FXR-FGF15-CYP7A1 signaling axis (Chen et al., 2025). However, whether DM2420 acts through the same pathway remains unknown and requires further investigation. In addition to bile acid remodeling, the DM2420 culture supernatant induced marked alterations in glycerophospholipid metabolism, characterized by increased LysoPC(20:2) and decreased PE, PG, and PS classes. Collectively, these in vitro and genomic findings characterize DM2420 as a probiotic candidate with favorable safety-related genomic features and in vitro bile acid- and phospholipid-mediated lipid-regulatory activities. However, its in vivo efficacy and safety remain to be established.

Although this study characterizes sex-related alterations in gut microbiota composition associated with dyslipidemia, several limitations warrant acknowledgment. The number of participants was relatively small and from a single region, with confounders (diet, lifestyle) potentially limiting generalizability, and the observed microbial signatures were only regarded as preliminary and exploratory. While correlation analyses revealed strong associations, they do not establish causality. Future sex-stratified FMT and targeted microbial in vivo intervention studies are needed to determine the causal role of specific taxa, particularly L. johnsonii, in dyslipidemia. Integrating microbiome, targeting bile acid metabolomics, genomic bsh expression, and host lipid metabolism will further elucidate the molecular mechanisms underlying L. johnsonii-mediated cholesterol regulation and support the development of sex-informed microbiota-based therapeutic strategies.

In conclusion, this study revealed pronounced sex-related differences in gut microbiota composition, host immune responses, and bile acid metabolism during high-fat diet-induced dyslipidemia. Male mice exhibited intestinal barrier dysfunction and systemic inflammation, accompanied by the depletion of L. johnsonii and a broad reduction in primary bile acids. In contrast, female mice displayed greater metabolic adaptability, characterized by enrichment of specific bile acid metabolites associated with Lachnospiraceae and Akkermansia. Guided by the depletion of L. johnsonii in both clinical subjects and the murine model, we isolated and characterized L. johnsonii, which exhibited in vitro favorable probiotic properties, including bile salt hydrolase activity and cholesterol removal capacity. Collectively, these findings provide insights into the association of gut microbiota, bile acids, and immune responses with the sex-related differences in dyslipidemia and provide a rationale for further investigation into sex-informed microbiota-based intervention strategies.

Acknowledgments

We are thankful to the Model Animal Research Centre of Dalian Medical University for providing the necessary facilities.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Applied Basic Research Project of Liaoning Provincial Department of Science and Technology (Grant No. 2026JH2/101300040).

Edited by: Anandharaj Marimuthu, Berkeley Lab (DOE), United States

Reviewed by: Leitao Sun, The First Affiliated Hospital of Zhejiang Chinese Medical University, China

XiaoYan Cui, Yangzhou University, China

Ramiya Baskaran, Corteva Agriscience, United States

Data availability statement

The clinical sequencing data are available at the Sequence Read Archive (SRA) of NCBI (accession PRJNA1063629: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1063629). The raw data from the non-targeted metabolomics analysis in this study have been deposited in the OMIX database (accession OMIX013088: https://ngdc.cncb.ac.cn/omix/release/OMIX013088) of the National Genomics Data Center (accession PRJCA051444: https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA051444). The code and data for reproducing this analysis are available in the Figshare repository: https://doi.org/10.6084/m9.figshare.33088025. Other data and materials are present in the main text and Supplementary materials.

Ethics statement

The clinical study was conducted in accordance with the principles of the Declaration of Helsinki, and the study protocol was approved by the First Affiliated Hospital Ethical Committee of Dalian Medical University (approval number: PJ-KS-KY-2021-90). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. All procedures were approved by the Animal Ethics Committee of Dalian Medical University (AEE24090) and conducted in strict accordance with the institution’s Ethical Guidelines for the care and use of laboratory animals. The study was conducted in accordance with the local legislation and institutional requirements.

Author contributions

YX: Investigation, Writing – original draft, Formal analysis, Software, Visualization, Data curation, Conceptualization, Methodology, Writing – review & editing. JJ: Writing – original draft, Data curation, Investigation, Conceptualization, Writing – review & editing, Software, Methodology, Visualization, Formal analysis. JA: Methodology, Writing – review & editing. QW: Methodology, Writing – review & editing. OZ: Writing – review & editing. ZB: Software, Writing – review & editing, Methodology. ML: Writing – review & editing, Methodology, Software. LC: Conceptualization, Supervision, Writing – review & editing. HL: Supervision, Writing – review & editing, Conceptualization.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2026.1970842/full#supplementary-material

Supplementary_file_1.docx (22.5MB, docx)

References

  1. Ambikairajah A., Walsh E., Cherbuin N. (2019). Lipid profile differences during menopause: a review with meta-analysis. Menopause 26, 1327–1333. doi: 10.1097/GME.0000000000001403, [DOI] [PubMed] [Google Scholar]
  2. Assefa S., Ahles K., Bigelow S., Curtis J. T., Köhler G. A. (2015). Lactobacilli with probiotic potential in the prairie vole (Microtus ochrogaster). Gut Pathog 7:35. doi: 10.1186/s13099-015-0082-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aylward B. A., Clark M. L., Galileo D. S., Baernard A. M., Wilson J. R., Brannick E., et al. (2019). Immune cell populations residing in mesenteric adipose depots and mesenteric lymph nodes of lean dairy cows. J. Dairy Sci. 102, 3452–3468. doi: 10.3168/jds.2018-15156, [DOI] [PubMed] [Google Scholar]
  4. Bai R., Cui F., Li W., Wang Y., Wang Z., Gao Y., et al. (2022). Codonopsis pilosula oligosaccharides modulate the gut microbiota and change serum metabolomic profiles in high-fat diet-induced obese mice. Food Funct. 13, 8143–8157. doi: 10.1039/d2fo01119k, [DOI] [PubMed] [Google Scholar]
  5. Banerjee S., Schlaeppi K., van der Heijden M. G. A. (2018). Keystone taxa as drivers of microbiome structure and functioning. Nat. Rev. Microbiol. 16, 567–576. doi: 10.1038/s41579-018-0024-1, [DOI] [PubMed] [Google Scholar]
  6. Bauer A. W., Kirby W. M., Sherris J. C., Turck M. (1966). Antibiotic susceptibility testing by a standardized single disk method. Am. J. Clin. Pathol. 45, 493–496. doi: 10.1093/ajcp/45.4_ts.493, [DOI] [PubMed] [Google Scholar]
  7. Bhat B., Bajaj B. K. (2020). Multifarious cholesterol lowering potential of lactic acid bacteria equipped with desired probiotic functional attributes. 3 Biotech 10:200. doi: 10.1007/s13205-020-02183-8, [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Boucard A.-S., Florent I., Polack B., Langella P., Bermúdez-Humarán L. G. (2022). Genome sequence and assessment of safety and potential probiotic traits of Lactobacillus johnsonii CNCM I-4884. Microorganisms 10:273. doi: 10.3390/microorganisms10020273, [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bourdeau-Julien I., Castonguay-Paradis S., Rochefort G., Perron J., Lamarche B., Flamand N., et al. (2023). The diet rapidly and differentially affects the gut microbiota and host lipid mediators in a healthy population. Microbiome 11:26. doi: 10.1186/s40168-023-01469-2, [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Braga Tibaes J. R., Azarcoya-Barrera J., Wollin B., Veida-Silva H., Makarowski A., Vine D., et al. (2022). Sex differences distinctly impact high-fat diet-induced immune dysfunction in wistar rats. J. Nutr. 152, 1347–1357. doi: 10.1093/jn/nxac024, [DOI] [PubMed] [Google Scholar]
  11. Brunham L. R., Lonn E., Mehta S. R. (2024). Dyslipidemia and the current state of cardiovascular disease: epidemiology, risk factors, and effect of lipid lowering. Can. J. Cardiol. 40, S4–S12. doi: 10.1016/j.cjca.2024.04.017, [DOI] [PubMed] [Google Scholar]
  12. Chen K., Dang D., Li H., Ross R. P., Stanton C., Chen W., et al. (2025). Lactobacillus johnsonii CCFM1376 improves hypercholesterolemia in mice by regulating the composition of bile acids. Microbiome Res Rep 4:6. doi: 10.20517/mrr.2024.38, [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Chen X., Wu J., Fu X., Wang P., Chen C. (2023). Fructus mori polysaccharide alleviates diabetic symptoms by regulating intestinal microbiota and intestinal barrier against TLR4/NF-κB pathway. Int. J. Biol. Macromol. 249:126038. doi: 10.1016/j.ijbiomac.2023.126038, [DOI] [PubMed] [Google Scholar]
  14. Chen K., Zhou X., Zhao J., Ross R. P., Stanton C., Chen W., et al. (2023). Comparative genomics of Lactobacillus johnsonii reveals extensive intraspecific genetic variation. Food Biosci. 56:103190. doi: 10.1016/j.fbio.2023.103190 [DOI] [Google Scholar]
  15. Collins S. L., Stine J. G., Bisanz J. E., Okafor C. D., Patterson A. D. (2023). Bile acids and the gut microbiota: metabolic interactions and impacts on disease. Nat. Rev. Microbiol. 21, 236–247. doi: 10.1038/s41579-022-00805-x, [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. de la Cuesta-Zuluaga J., Kelley S. T., Chen Y., Escobar J. S., Mueller N. T., Ley R. E., et al. (2019). Age- and sex-dependent patterns of gut microbial diversity in human adults. mSystems 4:e00261-19. doi: 10.1128/mSystems.00261-19, [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. den Besten G., Bleeker A., Gerding A., van Eunen K., Havinga R., van Dijk T. H., et al. (2015). Short-chain fatty acids protect against high-fat diet-induced obesity via a PPARγ-dependent switch from lipogenesis to fat oxidation. Diabetes 64, 2398–2408. doi: 10.2337/db14-1213, [DOI] [PubMed] [Google Scholar]
  18. Deng X., Ma J., Song M., Jin Y., Ji C., Ge W., et al. (2019). Effects of products designed to modulate the gut microbiota on hyperlipidaemia. Eur. J. Nutr. 58, 2713–2729. doi: 10.1007/s00394-018-1821-z, [DOI] [PubMed] [Google Scholar]
  19. Dohrn G. (2024). Gut microbes linked to fatty diet drive tumour growth. Nature. doi: 10.1038/d41586-024-01443-4, [DOI] [PubMed] [Google Scholar]
  20. Duche R. T., Singh A., Wandhare A. G., Sangwan V., Sihag M. K., Nwagu T. N. T., et al. (2023). Antibiotic resistance in potential probiotic lactic acid bacteria of fermented foods and human origin from Nigeria. BMC Microbiol. 23:142. doi: 10.1186/s12866-023-02883-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Fu J., Zong X., Jin M., Min J., Wang F., Wang Y. (2023). Mechanisms and regulation of defensins in host defense. Signal Transduct. Target. Ther. 8:300. doi: 10.1038/s41392-023-01553-x, [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gândara R. M. C., Mahida Y. R., Potten C. S. (2012). Regional differences in stem and transit cell proliferation and apoptosis in the terminal ileum and colon of mice after 12 Gy. Int. J. Radiat. Oncol. Biol. Phys. 82, e521–e528. doi: 10.1016/j.ijrobp.2011.07.015, [DOI] [PubMed] [Google Scholar]
  23. Gęgotek A., Skrzydlewska E. (2022). Antioxidative and anti-inflammatory activity of ascorbic acid. Antioxidants 11:1993. doi: 10.3390/antiox11101993, [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gilliland S. E., Walker D. K. (1990). Factors to consider when selecting a culture of Lactobacillus acidophilus as a dietary adjunct to produce a hypocholesterolemic effect in humans. J. Dairy Sci. 73, 905–911. doi: 10.3168/jds.S0022-0302(90)78747-4, [DOI] [PubMed] [Google Scholar]
  25. Gophna U., Konikoff T., Nielsen H. B. (2017). Oscillospira and related bacteria - from metagenomic species to metabolic features. Environ. Microbiol. 19, 835–841. doi: 10.1111/1462-2920.13658, [DOI] [PubMed] [Google Scholar]
  26. Gunathilake M., Lee J., Choi I. J., Kim Y. I., Yoon J., Sul W. J., et al. (2020). Alterations in gastric microbial communities are associated with risk of gastric cancer in a Korean population: a case-control study. Cancers 12:2619. doi: 10.3390/cancers12092619, [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Guo M., Liu H., Yu Y., Zhu X., Xie H., Wei C., et al. (2023). Lactobacillus rhamnosus GG ameliorates osteoporosis in ovariectomized rats by regulating the Th17/Treg balance and gut microbiota structure. Gut Microbes 15:2190304. doi: 10.1080/19490976.2023.2190304, [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hayashi A., Mikami Y., Miyamoto K., Kamada N., Sato T., Mizuno S., et al. (2017). Intestinal dysbiosis and biotin deprivation induce alopecia through overgrowth of Lactobacillus murinus in mice. Cell Rep. 20, 1513–1524. doi: 10.1016/j.celrep.2017.07.057, [DOI] [PubMed] [Google Scholar]
  29. Huang J., Huang J., Li Y., Wang Y., Wang F., Qiu X., et al. (2021). Sodium alginate modulates immunity, intestinal mucosal barrier function, and gut microbiota in cyclophosphamide-induced immunosuppressed BALB/c mice. J. Agric. Food Chem. 69, 7064–7073. doi: 10.1021/acs.jafc.1c02294, [DOI] [PubMed] [Google Scholar]
  30. Huang J., Xu Y., Wang M., Yu S., Li Y., Tian H., et al. (2023a). Enterococcus faecium R-026 combined with Bacillus subtilis R-179 alleviate hypercholesterolemia and modulate the gut microbiota in C57BL/6 mice. FEMS Microbiol. Lett. 370:fnad118. doi: 10.1093/femsle/fnad118, [DOI] [PubMed] [Google Scholar]
  31. Huang C., Yang X., Huang J., Liu X., Yang X., Jin H., et al. (2019). Porcine beta-defensin 2 provides protection against bacterial infection by a direct bactericidal activity and alleviates inflammation via interference with the TLR4/NF-κB pathway. Front. Immunol. 10:1673. doi: 10.3389/fimmu.2019.01673, [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Huang J., Zhou H., Song T., Wang B., Ge H., Zhang D., et al. (2023b). Fecal microbiota transplantation from sodium alginate-dosed mice and normal mice mitigates intestinal barrier injury and gut dysbiosis induced by antibiotics and cyclophosphamide. Food Funct. 14, 5690–5701. doi: 10.1039/d3fo01193c, [DOI] [PubMed] [Google Scholar]
  33. Jia B., Zou Y., Han X., Bae J.-W., Jeon C. O. (2023). Gut microbiome-mediated mechanisms for reducing cholesterol levels: implications for ameliorating cardiovascular disease. Trends Microbiol. 31, 76–91. doi: 10.1016/j.tim.2022.08.003, [DOI] [PubMed] [Google Scholar]
  34. Joint Committee on the Chinese Guidelines for Lipid Management (2023). Chinese guidelines for lipid management. Zhonghua Xin Xue Guan Bing Za Zhi 51, 221–255. doi: 10.3760/cma.j.cn112148-20230119-00038 [DOI] [PubMed] [Google Scholar]
  35. Kleiner D. E., Brunt E. M., Van Natta M., Behling C., Contos M. J., Cummings O. W., et al. (2005). Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 41, 1313–1321. doi: 10.1002/hep.20701, [DOI] [PubMed] [Google Scholar]
  36. Ko S. H., Jung Y. (2021). Energy metabolism changes and dysregulated lipid metabolism in postmenopausal women. Nutrients 13:4556. doi: 10.3390/nu13124556, [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kumar R., Grover S., Batish V. K. (2011). Hypocholesterolaemic effect of dietary inclusion of two putative probiotic bile salt hydrolase-producing Lactobacillus plantarum strains in Sprague-Dawley rats. Br. J. Nutr. 105, 561–573. doi: 10.1017/S0007114510003740, [DOI] [PubMed] [Google Scholar]
  38. Lefebvre C., Tiffay A., Breemeersch C. E., Dreux V., Bôle-Feysot C., Guérin C., et al. (2024). Sex-dependent effects of a high fat diet on metabolic disorders, intestinal barrier function and gut microbiota in mouse. Sci. Rep. 14:19835. doi: 10.1038/s41598-024-70931-4, [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Li Y., Huang J., Zhang S., Yang F., Zhou H., Song Y., et al. (2022). Sodium alginate and galactooligosaccharides ameliorate metabolic disorders and alter the composition of the gut microbiota in mice with high-fat diet-induced obesity. Int. J. Biol. Macromol. 215, 113–122. doi: 10.1016/j.ijbiomac.2022.06.073, [DOI] [PubMed] [Google Scholar]
  40. Li Y. H., Xiao H. T., Hu D. D., Fatima S., Lin C. Y., Mu H.-X., et al. (2016). Berberine ameliorates chronic relapsing dextran sulfate sodium-induced colitis in C57BL/6 mice by suppressing Th17 responses. Pharmacol. Res. 110, 227–239. doi: 10.1016/j.phrs.2016.02.010, [DOI] [PubMed] [Google Scholar]
  41. Liu L., Chen D., Yu B., Yin H., Huang Z., Luo Y., et al. (2020). Fructooligosaccharides improve growth performance and intestinal epithelium function in weaned pigs exposed to enterotoxigenic Escherichia coli. Food Funct. 11, 9599–9612. doi: 10.1039/d0fo01998d, [DOI] [PubMed] [Google Scholar]
  42. Liu H., Liu K., Bodenner D. L. (2005). Estrogen receptor inhibits interleukin-6 gene expression by disruption of nuclear factor kappaB transactivation. Cytokine 31, 251–257. doi: 10.1016/j.cyto.2004.12.008, [DOI] [PubMed] [Google Scholar]
  43. MacCannell A. D. V., Futers T. S., Whitehead A., Moran A., Witte K. K., Roberts L. D. (2021). Sexual dimorphism in adipose tissue mitochondrial function and metabolic flexibility in obesity. Int. J. Obes. 45, 1773–1781. doi: 10.1038/s41366-021-00843-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Madani G., Mirlohi M., Yahay M., Hassanzadeh A. (2013). How much in vitro cholesterol reducing activity of lactobacilli predicts their in vivo cholesterol function? Int. J. Prev. Med. 4, 404–413. [PMC free article] [PubMed] [Google Scholar]
  45. Miani M., Le Naour J., Waeckel-Enée E., Verma S. C., Straube M., Emond P., et al. (2018). Gut microbiota-stimulated innate lymphoid cells support β-defensin 14 expression in pancreatic endocrine cells, preventing autoimmune diabetes. Cell Metab. 28, 557–572.e6. doi: 10.1016/j.cmet.2018.06.012, [DOI] [PubMed] [Google Scholar]
  46. Mulder J. W. C. M., Tromp T. R., Al-Khnifsawi M., Blom D. J., Chlebus K., Cuchel M., et al. (2024). Sex differences in diagnosis, treatment, and cardiovascular outcomes in homozygous familial hypercholesterolemia. JAMA Cardiol. 9, 313–322. doi: 10.1001/jamacardio.2023.5597, [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Nestor-Kalinoski A., Smith-Edwards K. M., Meerschaert K., Margiotta J. F., Rajwa B., Davis B. M., et al. (2022). Unique neural circuit connectivity of mouse proximal, middle, and distal colon defines regional colonic motor patterns. Cell. Mol. Gastroenterol. Hepatol. 13, 309–337.e3. doi: 10.1016/j.jcmgh.2021.08.016, [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Org E., Mehrabian M., Parks B. W., Shipkova P., Liu X., Drake T. A., et al. (2016). Sex differences and hormonal effects on gut microbiota composition in mice. Gut Microbes 7, 313–322. doi: 10.1080/19490976.2016.1203502, [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Peng C., Xu X., Li Y., Li X., Yang X., Chen H., et al. (2020). Sex-specific association between the gut microbiome and high-fat diet-induced metabolic disorders in mice. Biol. Sex Differ. 11:5. doi: 10.1186/s13293-020-0281-3, [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Peng Y., Zhou M., Yang H., Qu R., Qiu Y., Hao J., et al. (2023). Regulatory mechanism of M1/M2 macrophage polarization in the development of autoimmune diseases. Mediat. Inflamm. 2023:8821610. doi: 10.1155/2023/8821610, [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Pirillo A., Casula M., Olmastroni E., Norata G. D., Catapano A. L. (2021). Global epidemiology of dyslipidaemias. Nat. Rev. Cardiol. 18, 689–700. doi: 10.1038/s41569-021-00541-4, [DOI] [PubMed] [Google Scholar]
  52. Qiao B., Liu J., Peng X., Cai Y., Peng M., Li X., et al. (2023). Association of short-chain fatty acids with gut microbiota and lipid metabolism in mice with diarrhea induced by high-fat diet in a fatigued state. Mol. Nutr. Food Res. 67:e2300452. doi: 10.1002/mnfr.202300452, [DOI] [PubMed] [Google Scholar]
  53. Ren S., Chen A., Tian Y., Bai Z., Wang C. (2022). Lactobacillus paracasei from koumiss ameliorates diarrhea in mice via tight junctions modulation. Nutrition 98:111584. doi: 10.1016/j.nut.2021.111584, [DOI] [PubMed] [Google Scholar]
  54. Rudel L. L., Morris M. D. (1973). Determination of cholesterol using o-phthalaldehyde. J. Lipid Res. 14, 364–366. doi: 10.1016/S0022-2275(20)36896-6, [DOI] [PubMed] [Google Scholar]
  55. Ryczkowska K., Adach W., Janikowski K., Banach M., Bielecka-Dabrowa A. (2023). Menopause and women’s cardiovascular health: is it really an obvious relationship? Arch. Med. Sci. 19, 458–466. doi: 10.5114/aoms/157308, [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Shu Z., Zhang G., Zhu X., Xiong W. (2022). Estrogen receptor α mediated M1/M2 macrophages polarization plays a critical role in NASH of female mice. Biochem. Biophys. Res. Commun. 596, 63–70. doi: 10.1016/j.bbrc.2022.01.085, [DOI] [PubMed] [Google Scholar]
  57. Spitler K. M., Davies B. S. J. (2020). Aging and plasma triglyceride metabolism. J. Lipid Res. 61, 1161–1167. doi: 10.1194/jlr.R120000922, [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Strizova Z., Benesova I., Bartolini R., Novysedlak R., Cecrdlova E., Foley L. K., et al. (2023). M1/M2 macrophages and their overlaps - myth or reality? Clin. Sci. 137, 1067–1093. doi: 10.1042/CS20220531, [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. van Heel A. J., de Jong A., Song C., Viel J. H., Kok J., Kuipers O. P. (2018). BAGEL4: a user-friendly web server to thoroughly mine RiPPs and bacteriocins. Nucleic Acids Res. 46, W278–W281. doi: 10.1093/nar/gky383, [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wang Y., Ai Z., Xing X., Fan Y., Zhang Y., Nan B., et al. (2024). The ameliorative effect of probiotics on diet-induced lipid metabolism disorders: a review. Crit. Rev. Food Sci. Nutr. 64, 3556–3572. doi: 10.1080/10408398.2022.2132377, [DOI] [PubMed] [Google Scholar]
  61. Wang B., Kong Q., Li X., Zhao J., Zhang H., Chen W., et al. (2020). A high-fat diet increases gut microbiota biodiversity and energy expenditure due to nutrient difference. Nutrients 12:3197. doi: 10.3390/nu12103197, [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wang J., Lang T., Shen J., Dai J., Tian L., Wang X. (2019). Core gut bacteria analysis of healthy mice. Front. Microbiol. 10:887. doi: 10.3389/fmicb.2019.00887, [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang Q., Li Z., Liu K., Liu J., Chai S., Chen G., et al. (2021). Activation of the G protein-coupled estrogen receptor prevented the development of acute colitis by protecting the crypt cell. J. Pharmacol. Exp. Ther. 376, 281–293. doi: 10.1124/jpet.120.000216, [DOI] [PubMed] [Google Scholar]
  64. Wang J., Tang H., Zhang C., Zhao Y., Derrien M., Rocher E., et al. (2015). Modulation of gut microbiota during probiotic-mediated attenuation of metabolic syndrome in high fat diet-fed mice. ISME J. 9, 1–15. doi: 10.1038/ismej.2014.99, [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Wang K., Wang Y., Gu L., Yu J., Liu Q., Zhang R., et al. (2024). Characterization of probiotic properties and whole-genome analysis of Lactobacillus johnsonii N5 and N7 isolated from swine. Microorganisms 12:672. doi: 10.3390/microorganisms12040672, [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wang L., Zhu F., Yang H., Li J., Li Y., Ding X., et al. (2019). Effects of dietary supplementation with epidermal growth factor on nutrient digestibility, intestinal development and expression of nutrient transporters in early-weaned piglets. J Anim Physiol Anim Nutr 103, 618–625. doi: 10.1111/jpn.13059, [DOI] [PubMed] [Google Scholar]
  67. Wu B., Tan Y., Huang H., Liu Y., Bai T., Yang L. (2022). Alleviating effect of methionine on intestinal development and intercellular junction induced by nickel. Biol. Trace Elem. Res. 200, 4007–4016. doi: 10.1007/s12011-021-02992-9, [DOI] [PubMed] [Google Scholar]
  68. Xia M., Xu Y., Li H., Huang J., Zhou H., Gao C., et al. (2024). Structural and functional alteration of the gut microbiota in elderly patients with hyperlipidemia. Front. Cell. Infect. Microbiol. 14:1333145. doi: 10.3389/fcimb.2024.1333145, [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Xu Y., Wang Y., Song T., Li X., Zhou H., Chaibou O. Z., et al. (2024). Immune-enhancing effect of Weizmannia coagulans BCG44 and its supernatant on cyclophosphamide-induced immunosuppressed mice and RAW264.7 cells via the modulation of the gut microbiota. Food Funct. 15, 10679–10697. doi: 10.1039/d4fo02452d, [DOI] [PubMed] [Google Scholar]
  70. Xu J., Xia Q., Wu T., Shao Y., Wang Y., Jin N., et al. (2024). Prophylactic treatment with Bacteroides uniformis and Bifidobacterium bifidum counteracts hepatic NK cell immune tolerance in nonalcoholic steatohepatitis induced by high fat diet. Gut Microbes 16:2302065. doi: 10.1080/19490976.2024.2302065, [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Yang C., He P., Wen Y., Zeng Q., Li J. (2026). Gut microbiota dysbiosis in infants and young children with severe pneumonia and sepsis: a matched case-control study identifying potential biomarkers for early risk stratification. Front. Immunol. 17:1796938. doi: 10.3389/fimmu.2026.1796938, [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Yang S., Yu M. (2021). Role of goblet cells in intestinal barrier and mucosal immunity. J. Inflamm. Res. 14, 3171–3183. doi: 10.2147/JIR.S318327, [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Ye Y., Zhou Q., Dai W., Peng H., Zhou S., Tian H., et al. (2023). Gender differences in metabolic syndrome and its components in southern China using a healthy lifestyle index: a cross-sectional study. BMC Public Health 23:686. doi: 10.1186/s12889-023-15584-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Yoon H., Lee Y., Park H., Kang H.-J., Ji Y., Holzapfel W. H. (2023). Lactobacillus johnsonii BFE6154 ameliorates diet-induced hypercholesterolemia. Probiotics Antimicrob Proteins 15, 451–459. doi: 10.1007/s12602-021-09859-4, [DOI] [PubMed] [Google Scholar]
  75. Zhang X., Coker O. O., Chu E. S., Fu K., Lau H. C. H., Wang Y. X., et al. (2021). Dietary cholesterol drives fatty liver-associated liver cancer by modulating gut microbiota and metabolites. Gut 70, 761–774. doi: 10.1136/gutjnl-2019-319664, [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Zhang Z., Zhao L., Wu J., Pan Y., Zhao G., Li Z., et al. (2023). The effects of Lactobacillus johnsonii on diseases and its potential applications. Microorganisms 11:2580. doi: 10.3390/microorganisms11102580, [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Zhao M., Li Y., Zhang Y., Li G. (2024). Genomic analysis and functional properties of Lactobacillus johnsonii GJ231 isolated from healthy beagles. Front. Microbiol. 15:1437036. doi: 10.3389/fmicb.2024.1437036, [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Zheng X., Huang F., Zhao A., Lei S., Zhang Y., Xie G., et al. (2017). Bile acid is a significant host factor shaping the gut microbiome of diet-induced obese mice. BMC Biol. 15:120. doi: 10.1186/s12915-017-0462-7, [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary_file_1.docx (22.5MB, docx)

Data Availability Statement

The clinical sequencing data are available at the Sequence Read Archive (SRA) of NCBI (accession PRJNA1063629: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1063629). The raw data from the non-targeted metabolomics analysis in this study have been deposited in the OMIX database (accession OMIX013088: https://ngdc.cncb.ac.cn/omix/release/OMIX013088) of the National Genomics Data Center (accession PRJCA051444: https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA051444). The code and data for reproducing this analysis are available in the Figshare repository: https://doi.org/10.6084/m9.figshare.33088025. Other data and materials are present in the main text and Supplementary materials.


Articles from Frontiers in Microbiology are provided here courtesy of Frontiers Media SA

RESOURCES