Unraveling the Gut Microbiota-Metabolite-Host Gene Axis in the Pathogenesis of MASLD and Type 2 Diabetes Comorbidity

Unraveling the Gut Microbiota-Metabolite-Host Gene Axis in the Pathogenesis of MASLD and Type 2 Diabetes Comorbidity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Unraveling the Gut Microbiota-Metabolite-Host Gene Axis in the Pathogenesis of MASLD and Type 2 Diabetes Comorbidity Yang Zeng, Yu Cheng, Wentao Ma, Wenjie Ye, Zhenyu Liu, Linjing Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9357546/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose This study aimed to elucidate the common pathogenic pathways underlying the comorbidity of metabolic-associated fatty liver disease (MASLD) and type 2 diabetes mellitus (T2DM), with particular focus on identifying how gut microbiota and their metabolites regulate host gene expression in these mutually reinforcing metabolic disorders. Methods We employed an integrated multi-omics approach combining Mendelian randomization (MR) analysis with transcriptomic data from GEO datasets (GSE89632 and GSE26168) and gut microbiome GWAS data from MiBioGen. Differential gene expression analysis was performed to identify co-differentially expressed genes. MR analysis screened for disease-associated gut bacteria. Integration of host gene, metabolite, and microbiome data with databases such as gutMGene was conducted to determine key indicators. Functional enrichment analysis, protein-protein interaction network construction, and molecular docking simulations were performed. A predictive linear model was developed and validated using Hosmer-Lemeshow tests and decision curve analysis. Immunohistochemical analysis assessed correlations with immune cell infiltration, and ceRNA regulatory networks were constructed. Results The analysis identified 246 co-differentially expressed genes and 14 key gut bacteria potentially associated with both diseases. Seven key indicators were ultimately determined through data integration. These genes were enriched in lipid metabolism, inflammatory regulation, and redox processes. Four key protein nodes (PTGS2, ALOX15B, KLK3, and DHFR) were identified. Molecular docking revealed strong binding stability between core metabolites (genistein, phenylacetic acid, and N-acetylornithine) and target proteins. The predictive model demonstrated excellent performance with AUC values of 0.982 and 0.903 in the two datasets. Strong correlations were observed between key genes and immune cell infiltration rates, and multi-level ceRNA regulatory mechanisms were uncovered. Conclusion This study demonstrates that the gut microbiota-metabolite-host gene axis plays a critical role in MASLD and T2DM comorbidity. The identified biomarkers and therapeutic targets provide important foundations for accurate prevention, diagnosis, and treatment strategies for these metabolic diseases. Metabolic fatty liver disease Type 2 diabetes mellitus Gut microbiota Transcriptomics Mendelian randomization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction 1.1. Metabolic-associated fatty liver disease (MASLD) and the epidemiological status of type 2 diabetes (T2DM) Formerly known as non-alcoholic fatty liver disease (NAFLD), metabolic fatty liver disease (MASLD) is one of the world’s fastest-growing chronic liver diseases, driven by global economic development and lifestyle Westernization( 1 ). Recent epidemiological data show a global adult prevalence of 25%-30% (up to 40% in specific regions)( 2 ), reflecting high incidence and a youthification trend. In China, MASLD incidence has risen sharply amid lifestyle shifts, obesity, and metabolic syndrome, with adult prevalence in urban areas approaching 30%( 3 ). Type 2 diabetes mellitus (T2DM), the most common diabetes type, affects ~ 500 million people worldwide, growing at > 4% annually( 4 ), with insulin resistance and obesity as key pathogenesis. Notably, 70%-80% of T2DM patients have hepatic steatosis, meaning MASLD prevalence in this group is far higher than in the general population( 5 ). Conversely, large cohort studies indicate MASLD patients face a ~ 23-fold higher relative risk of developing T2DM over the next 5–10 years( 6 , 7 ). 1.2. The mutual relationship and harm between MASLD and T2DM MASLD and T2DM are not simple comorbidities but a complex metabolic network interacting via inflammatory responses, insulin resistance, and lipid metabolism disorders( 2 , 8 ). Their coexistence markedly elevates risks of hepatic fibrosis, nephropathy, and cardiovascular disease, accelerates hepatic fibrosis progression, and increases cirrhosis and hepatocellular cancer prevalence( 6 ). Additionally, insulin resistance worsens, hindering blood glucose control and augmenting therapeutic burden. Despite improved clinical diagnosis rates, the dual burden of MASLD and T2DM remains a major challenge in modern metabolic disease management. 1.3. Research Progress on Gut Microbiota and Metabolic Diseases Known as the human "second genome," the gut microbiota plays a key role in maintaining metabolic balance, with numerous studies linking its compositional and functional alterations to metabolic disorders like MASLD and T2DM( 9 ). By fermenting undigested dietary fibers, it produces metabolites (bile acid metabolites, tryptophan derivatives, short-chain fatty acids [SCFAs]) that regulate host energy metabolism, inflammation, and immunity, thereby impacting insulin sensitivity and liver fat accumulation( 10 ). Notably, MASLD patients exhibit reduced gut flora diversity, increased pathogenic bacteria (certain Bacteroides/Bacteroidetes), and decreased probiotics (Bifidobacteria, Akkermansia)( 11 ). T2DM patients show similar imbalances, plus elevated intestinal permeability, which activates metabolic inflammation, impairing pancreatic β-cell function and insulin resistance( 12 ). Thus, the gut microbiota and its metabolites act as a critical bridge between MASLD and T2DM, holding potential diagnostic and therapeutic value. 1.4. The application value of Mendelian randomization analysis in metabolic disease research Traditional observational studies struggle with confounding factors and reverse causality when exploring the gut microbiota-MASLD/T2DM association( 13 ), hampering clear causal inference. Mendelian randomization (MR), a causal inference method based on genetic variation, uses genetic variants as instrumental variables to simulate random allocation, effectively avoiding environmental and behavioral interference. It has been widely applied to verify causal relationships in complex diseases( 14 ). In gut microbiota research, MR enables investigating causality between specific bacterial abundances and metabolic disease risk( 15 ). For example, prior studies used MR to clarify the protective effect of certain probiotic genera on T2DM risk, laying a scientific foundation for the gut microbiota as a potential therapeutic target( 16 ). Additionally, integrating multi-omics data (transcriptomics, metabolomics) with MR helps reveal the molecular mechanism by which gut microbiota regulates host gene expression via metabolites, deepening understanding of the complex comorbidity network. 1.5. Research objectives and significance This study aims to clarify the gut microbiota-metabolite-host gene axis’s role in MASLD-T2DM comorbidity via transcriptomics, Mendelian randomization, and the gutMgene database, systematically analyzing key gut microbiota, their metabolites, and regulatory networks on host genes. Multidimensional data integration and bioinformatics will identify potential biomarkers/therapeutic targets, facilitating precision diagnosis/targeted therapies and advancing comorbidity prevention of T2DM and MASLD. 2. Materials and Methods 2.1. Data Acquisition The Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/ , download date: 2025/09/30) provided two datasets: Dataset GSEGSE26168, including mRNA and miRNA expression profiles of type 2 diabetes (T2DM) (experimental types: non-coding RNA expression profiling, microarray-based expression profiling). Samples consist of 5 healthy rats, 5 T2DM rats, 8 human healthy controls, 7 people with impaired fasting glucose, 9 T2DM patients, and miRNA sequencing samples (7 healthy controls, 7 IFG patients, 9 T2DM patients). This study selected 9 T2DM patients and 8 healthy controls as disease and control groups, respectively.Dataset GSE89632, focusing on mRNA expression profile of non-alcoholic fatty liver disease (NAFLD) using samples of human liver tissue (63 individuals: 20 simple fatty degeneration, 19 non-alcoholic fatty hepatitis, 24 healthy controls patients). This study designated healthy controls as the control group and NASH cases as the disease group. Gut microbiome dataset was obtained from the MiBioGen consortium ( https://mibiogen.gcc.rug.nl/ , download date: 2025/09/30), comprising fecal microbiome profiles. It includes 211 gut microbial taxonomic groups ( 131 genera, 35 families, 20 orders, 16 classes, and 9 phyla) and a total of 14,587 SNP loci. GWAS data were retrieved from the IEU OpenGWAS database ( https://gwas.mrcieu.ac.uk/ , download date: 2025/09/30), including ebi-a-GCST90091033, (778,614 total samples: 8,434 MM cases and 770,180 controls) and ukb-b-13806(2,972 TT cases and 459,961 controls). 2.2. Candidate gene identification Bulk_DEGs_T2DM and bulk_DEGs_MASLD were differential gene screening results, independently obtained via the R package "limma (3.56.2)" on disease/control samples in T2DM and MASLD training sets (threshold: P.Value 0.5). R packages "ggplot2 (3.5.2)" and "pheatmap (1.0.12)" were used to generate heatmaps and volcano plots. DEGs_MASLD and DEGs_T2DM were intersected using the "Venn Diagram (1.7.3)" package to yield DEGs_MASLD_T2DM. 2.3. Mendelian Randomization Analysis 2.3.1. Selection and Extraction of Instrumental Variables (IVs) Putative instrumental variables (SNPs) associated with the gut microbiome (exposure factor) and T2DM/MASLD (outcome variables) were extracted from public GWAS datasets using the extract_instruments function of the R package TwoSampleMR (0.6.8). Screening followed five criteria: 1) SNP-exposure association strength (P < 5×10⁻⁶); 2) linkage disequilibrium (LD) exclusion to ensure IV independence; 3) exclusion of SNPs significantly correlated with outcomes to reduce confounding bias; 4) a minimum of three independent SNPs per exposure; 5) IV strength assessment (F-statistic > 10) to avoid weak instrumental variable bias. 2.3.2. MR Analysis and Effect Size Evaluation After obtaining qualified instrumental variables, we use the mr function from the TwoSampleMR (0.6.8) package, combined with various MR methods (including the main inverse variance weighting method (IVW), weighted median method, weighted mode method, simple mode method, and MR-Egger regression) to evaluate the causal effects of the gut microbiota on T2DM and MASLD. A p-value less than 0.05 is used as the threshold for judging potential causal associations. 2.3.3. Visualization Analysis To visually present the analysis results, we adopted the following visualization methods: 1) Scatter plot (mr_scatter_plot): Used to visually demonstrate the correlation between exposure factors (gut microbiota) and outcome variables (T2DM/MASLD); 2) Forest plot (mr_forest_plot): Used to evaluate the effectiveness of each SNP locus in contributing to outcome variables and to summarize the effect sizes; 3) Funnel plot (mr_funnel_plot): Used to assess whether the effect of candidate exposure factors shows random distribution across different strengths of instrumental variables to preliminarily determine the presence of horizontal pleiotropy. 2.3.4. Sensitivity Analysis To ensure MR analysis robustness and reliability, we performed four sensitivity analyses: 1) Heterogeneity Test (mr_heterogeneity): Cochran's Q test evaluates SNP heterogeneity—fixed-effect IVW model for P > 0.05 (no significant heterogeneity) and random-effect IVW for P ≤ 0.05; 2) Pleiotropy Test (mr_pleiotropy_test): MR-Egger intercept and mr_presso detect horizontal pleiotropy (P > 0.05 = no significant bias); 3) Leave-One-Out Test (mr_leaveoneout): Sequentially remove each SNP to observe overall effect size changes, visualized via mr_leaveoneout_plot; 4) Directionality Test (directionality_test): Assess causal direction (exposure→outcome vs. outcome→exposure), retaining results with P < 0.05 and directionality judgment (TRUE). 2.4. Candidate Metabolite Targets Acquisition Candidate microbiota were retrieved from the gutMGene v3.0 database ( http://bio-annotation.cn/gutmgene/ , accessed on 2025/10/08), and SMILES (Simplified Molecular Input Line Entry System) patterns of candidate metabolites were identified via the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ , accessed on 2025/10/08). Under the "Homo sapiens" model, the Similarity Ensemble Approach (SEA, http://sea.bkslab.org/ , accessed on 2025/10/08) and SwissTargetPrediction (STP, http://www.swisstargetprediction.ch/ , accessed on 2025/10/08) were used to predict targets associated with each metabolite in the candidate microbiota. Overlapping targets between SEA and STP were further analyzed using the Venny 2.1.0 platform ( https://bioinfogp.cnb.csic.es/tools/venny/index.html , accessed on 2025/10/08). 2.5. Enrichment Analysis and PPI Network Construction In this study, candidate metabolite targets were first intersected with differentially expressed genes (DEGs) to obtain key genes. Functional enrichment analysis (Gene Ontology [GO, covering biological processes, cellular components, molecular functions] and Kyoto Encyclopedia of Genes and Genomes [KEGG] pathways) was then performed on these genes using the R package "clusterProfiler (4.15.0.003)", with P-value < 0.05 as the screening criterion to identify potential biological functions and involved pathways. Finally, the online STRING platform was used to construct a protein-protein interaction (PPI) network for candidate genes; after gene input, an interaction score threshold > 0.4 was set to filter high-interaction proteins, thereby determining the final biomarkers. 2.6. Metabolite Analysis SwissADME Metabolite Drug Similarity Analysis: SwissADME was used to assess the drug-likeness of metabolites in biomarker and signaling pathway network analysis, with screening criteria: hydrogen bond acceptor (HBA) < 10, hydrogen bond donor (HBD) ≤ 5, octanol-water partition coefficient (MLog P) ≤ 4.15, topological polar surface area (TPSA (Ų)) < 140, and Lipinski's rule ≤ 1. ADMETlab 3.0 Toxicology Property Evaluation: The online tool ADMETlab 2.0 was used to evaluate the toxicological properties of candidate core metabolites, assess toxicological risks (hERG, human liver toxicity H-HT, rat oral acute toxicity, carcinogenicity, DILI, etc.), and prioritize metabolites with negative toxicity predictions. 2.7. Molecular Docking Download the three-dimensional structure of small molecules (SDF/SMILES) from PubChem, download the receptor protein structure (PDB file) from PDB, then use OpenBabel/AutoDockTools (ADT) or UCSF Chimera for protonation, removal of ligands/water molecules, addition of hydrogen atoms, and generation of PDBQT files. Use AutoDock Vina for molecular docking, with ≤ − 5 kcal/mol as the reference threshold for good binding potential. 2.8. Construction of Nomogram and Model Evaluation A nomogram is a visualization tool based on a multi-factor regression model, which converts complex regression equations into intuitive graphical representations to facilitate individualized clinical risk assessment. In this study, a nomogram model was constructed using R package "rms" based on the expression data of selected key biomarkers, mapping each key gene’s expression level to corresponding risk scores; the total cumulative score of markers was calculated to predict patients’ disease risk. After model construction, calibration curves were used to evaluate the consistency between predicted probability and actual risk, and the Hosmer−Lemeshow test to assess model goodness-of-fit. Decision curve analysis (DCA) was further applied to measure the model’s clinical utility by evaluating its net benefit at different risk thresholds, ensuring statistical significance and practical value for clinical decision-making. This process enhanced the predictive model’s credibility and practicality, providing a scientific basis for patient disease management and treatment. 2.9. Immune Infiltration Analysis and Gene-Cell Correlation Given the immune system’s key role in metabolic diseases, this study used single-sample gene set enrichment analysis (ssGSEA) via the R package "GSVA (1.53.28)" to quantitatively assess the infiltration of 28 immune cell types in transcriptome data. Differential immune cell subsets were identified by comparing immune cell enrichment scores between disease and control groups. The Wilcoxon rank-sum test was used to statistically analyze intergroup differences in immune cell infiltration, and the "psych (2.4.6.26)" package to calculate Spearman correlations between key biomarker genes and significantly differential immune cells. Results were visualized as heatmaps and bubble plots, revealing that key genes may participate in metabolic disease pathogenesis by regulating specific immune cell activity. This indicates the immune microenvironment’s regulatory role in disease pathology and provides potential immunotherapeutic targets. 2.10. Construction of ceRNA and Regulatory Networks To explore the upstream non-coding RNA regulatory mechanism of key genes in comorbidity regulation, this study utilized comprehensive database resources to construct competitive endogenous RNA (ceRNA) regulatory networks. First, miRWalk and miRDB databases were used to screen miRNAs with experimentally validated or predicted associations with key genes, yielding multiple potential regulatory miRNAs. Subsequently, combined with the LncBase database, upstream lncRNAs of these miRNAs were predicted, and a ternary ceRNA regulatory network (lncRNA, miRNA, mRNA) was constructed. 3. Results 3.1 MASLD dataset GSE89632 vs. T2DM dataset GSE26168 differential gene analysis Differential expression analysis via limma (3.56.2) (P 1) identified 3,245 differentially expressed genes (DEGs) in T2DM dataset GSE26168 (1,564 upregulated, 1,681 downregulated) and 3,063 DEGs in MASLD dataset GSE89632 (1,316 upregulated, 1,747 downregulated). Heatmaps (Fig. 1A, C) show top 20 DEGs, while volcano plots (Fig. 1B, D) visualize expression changes and significance, with key DEGs including IRS1, INSR, SOCS3, MYC, JUN, FOS, GADD45. Intersection analysis revealed 246 common DEGs (117 upregulated, 129 downregulated; Fig. 1E, F). Transcriptomically, MASLD and T2DM exhibit disease-specific differences and shared alterations, providing a suitable gene set for functional enrichment, PPI network construction, and microbial metabolite integration. 3.2. Mendelian Randomization Analysis Investigating the Causal Relationship Between Gut Microbiota and T2DM and MASLD 3.2.1. Instrumental Variable Selection We found Bifidobacterium genus microbiota is significantly associated with multiple strong genetic variations (SNPs), with SNP (rs182549) standing out with the highest F-value (88.43). Selected representative SNPs (e.g., rs182549, rs6754311, rs7570971) all have F-values significantly higher than 10, confirming their potential as strong IVs to reduce analytical bias. These SNPs are mainly located on chromosome 2 (chr2) and strongly associated with Bifidobacterium at different taxonomic levels (genus, family, order). We corrected SNP effect directions and excluded bias-prone loci during screening, ensuring data accuracy and laying a solid foundation for subsequent causal analysis (Supplementary Table S1,S2 for details). 3.2.2. Causal effect estimation and visualization After screening out reliable instrumental variables, we use multiple MR methods to estimate causal effects and visualize them through graphs. 1.IVW forest map to analyze the effects of key microflora Uniform effect allele and effect size analysis were performed to analyze the above exposure factors separately and then based on IVW. As a result, factors with P < 0.05 were screened as potential related candidate exposure factors for subsequent analysis, passedThe forestploter package maps the forest and visualizes the results. The results showed that a total of 27 exposure factors and knots were obtained. Local MASLD was significantly associated with 27 exposure factors and outcome T2DM(Fig. 2A, B). 3.2.3. Sensitivity Analysis To ensure the robustness and reliability of the MR analysis results, we conducted a series of sensitivity analyses. 1.Heterogeneity Test The heterogeneity test results show that in the analysis of MASLD, certain microbial groups (such as the Melainabacteria class, Bifidobacterium genus, etc.) exhibit some heterogeneity (Cochran's Q test p < 0.05); in the analysis of T2DM, similar heterogeneity was observed in some microbial groups. However, neither of the two analyses found significant horizontal pleiotropy interference (MR-Egger intercept test p > 0.05), so the relevant analyses were all estimated using the random effects inverse variance weighting method (IVW) to ensure the reliability of the results (Supplementary Tables S3, S4). In addition, the leave-one-out sensitivity analysis indicates that in the analyses of MASLD and T2DM, the combined effect size of the remaining instrumental variables on the outcome did not show significant fluctuations after the sequential exclusion of individual SNPs, indicating that the main causal association results of this study are robust and not overly affected by individual abnormal instrumental variables (Figure 3D, H). 2.Horizontal Multicollinearity Testing The results showed that in MASLD, the p-values of the MR-Egger intercept test for most microbial communities were greater than 0.05, indicating no significant horizontal pleiotropy; however, some communities such as Actinobacteria, Bifidobacteriaceae, and Streptococcaceae had p-values less than 0.05, suggesting potential horizontal pleiotropy bias. In the T2DM analysis, the p-values of the MR-Egger intercept test for all microbial communities were greater than 0.05, indicating no significant horizontal pleiotropy detected. Overall, most exposure factors passed the horizontal pleiotropy test, supporting the validity of causal inference, but the results for a few communities in MASLD should be interpreted with caution(Supplementary Table 5,6). 3.Sensitivity analysis: leave one method for testing Visualized by mr_leaveoneout_plot (Fig. 3D,H), forest plots showed stable overlapping confidence intervals across iterations, with no significant heterogeneity. This indicates MR conclusions are not overly affected by single SNPs, supporting robust IV screening and highly reliable causal inference. 4.Comorbidity screening using core gut microbiota and core genes MR analysis first underwent intersection analysis of gut microbiota strongly linked to MASLD and T2DM, revealing 4 shared potential microbiota across taxonomic levels (genus, family, order; Figure 4A). Specifically, 27 strains were associated with MASLD-linked microbiota and 27 with T2DM-linked microbiota, forming the basis for subsequent metabolite-host gene integration. Using gutMGene, SEA, and SwissTargetPrediction databases, metabolites and their host targets of these microbiota were predicted. Intersecting these consensus targets with 246 shared differentially expressed genes from GSE89632 and GSE26168 datasets identified 7 common key genes (Figure 4B). As core nodes in the "microbiota-metabolite-host gene" regulatory axis, these genes were selected as potential biomarkers for subsequent functional enrichment and network construction. 3.3. Functional enrichment study with GO/KEGG GO enrichment analysis indicates these 7 common genes are involved in biological processes (e.g., unsaturated fatty acid/arachidonic acid metabolism, carboxylic acid biosynthesis, brown adipocyte differentiation, β-adrenergic receptor-mediated signaling) and molecular functions (oxidoreductase/dioxygenase activity), critical for energy/lipid metabolism and inflammatory mediator production. KEGG analysis (Figure 4D) identifies pathways including neurotransmitter regulation, lipid metabolism, VEGF signaling, and one-carbon metabolism (e.g., folate-mediated one-carbon pooling, folate biosynthesis). Gut microbiota metabolites may influence MASLD-T2DM comorbidity progression by regulating hepatic lipid metabolism, systemic inflammation, and energy homeostasis (Figure 4C). 3.4. Core Gene Modules and PPI Network Analysis Seven candidate genes were input into the STRING database to construct a PPI network (confidence score cutoff >0.4). Results showed these genes interact closely to form a highly interconnected module. Network analysis identifies PTGS2, ALOX15B, KLK3, DHFR as key nodes regulating redox, one-carbon metabolism, and inflammation (e.g., arachidonic acid/prostaglandin pathways). Critical for host responses to gut microbiota metabolites, PTGS2 and ALOX15B (linked to inflammatory signaling/lipid metabolism) are key targets for further validation (Figure 4E). 3.5. Association Network of Microbiome, Metabolites, and Biomarkers In the microbiome-metabolite-biomarker association network (Figure 4F), several microbial taxa (especially Bifidobacterium and its subspecies) are strongly associated with key small-molecule metabolites, including plant polyphenol derivatives (aglycone, kaempferol, genistein, equol) and amino acid derivatives (phenylacetic acid, N-acetylaspartate, leucylproline, folic acid). The network forms a "redundancy-multipoint" regulatory pattern: a single microbiome targets multiple host genes via diverse pathways, while multiple species/strains co-regulate the same metabolite, enhancing regulatory diversity and signal stability. Predictive analysis shows metabolites like equol and genistein directly/indirectly target key genes (PTGS2, ALOX15B, KLK3, DHFR), with microbes modifying fatty acid metabolism/inflammation via these bioactives to affect insulin sensitivity and liver lipid deposition. Genes like DHFR link to folate-related one-carbon metabolism, suggesting microbial regulation of host metabolic balance and epigenetics. This network highlights complex microbe-host multi-level interactions, providing potential biomarkers for precision microbial regulation and disease intervention. 3.6. SwissADME Metabolite Drug Similarity Analysis Using the SwissADME online platform (http://www.swissadme.ch/, access time:2025/10/09) and combining literature data, the drug-likeness of metabolites involved in biomarker-pathway network analysis is evaluated. Since metabolites typically have strong hydrophilicity and low bioavailability, this study analyzes their physicochemical properties (Supplementary Table S7) through computational simulation methods and sets the following screening thresholds to determine candidate core metabolites: hydrogen bond acceptor (HBA) < 10, hydrogen bond donor (HBD) ≤ 5, octanol-water partition coefficient (MLogP) ≤ 4.15, topological polar surface area (TPSA) < 140 Ų, and violation of the Lipinski rule ≤ 1. According to the above criteria, "Folic Acid" was excluded as it did not meet the conditions among the total of 9 metabolites. 3.7. ADMETox 3.0 Toxicological Property Assessment The physicochemical properties of metabolites are insufficient to fully assess their safety, and toxicological properties are crucial in the safety evaluation. Drug development failures are often attributed to adverse reactions, such as hERG block-induced cardiotoxicity, human liver toxicity (H-HT), respiratory toxicity, acute oral toxicity in rats, carcinogenicity, drug-induced liver injury (DILI), skin sensitization, and acute toxicity (LD50), among others. Therefore, this study employed the ADMETox 3.0 online platform (https://admetmesh.scbdd.com/,access time:2025/10/09) to predict the toxicological properties of candidate core metabolites. The screening criteria were that all four indicators—hERG block, human liver toxicity (H-HT), acute oral toxicity in rats, and carcinogenicity—were negative. The results showed that Genistein, N-acetylornithine, and another metabolite met all safety standards and were identified as key metabolites(Supplementary Table S8). 3.8. Core Metabolites and Biomarker Molecular Docking Based on the "microbiota—metabolite—biomarker" association network, Genistein/Phenylacetic Acid (targeting PTGS2) and N-acetylornithine (targeting DHFR) were selected as docking pairs. All three ligand-receptor complexes showed negative binding energies (Table 1), indicating spontaneous and stable binding: PTGS2-Genistein (-7.9 kcal/mol, highest affinity), PTGS2-Phenylacetic Acid (-5.8 kcal/mol), and DHFR-N-acetylornithine (-5.0 kcal/mol).Figure 5 details their conformations: PTGS2-Genistein forms stable bonds via hydrogen bonds and hydrophobic interactions at Thr-145/Val-349 (Figure 5A); PTGS2-Phenylacetic Acid binds at Leu-352/His-527 (multi-point binding, Figure 5B); DHFR-N-acetylornithine relies on ALA-523/MET-518 for stability (Figure 5C). These results confirm effective binding of core metabolites to biomarker proteins, suggesting they may directly regulate these proteins and participate in the pathogenesis of type 2 diabetes and stroke comorbidity, with potential for drug development. Table 1 . Molecular docking of core metabolites and biomarkers Compound name Target Protein Mode Binding Energy (kcal/mol) RMSD (l.b.) RMSD (u.b.) N-acetylornithine DHFR 1 -5 0 0 N-acetylornithine DHFR 2 -5 0.876 1.18 N-acetylornithine DHFR 3 -4.9 1.954 4.611 N-acetylornithine DHFR 4 -4.8 2.693 4.142 N-acetylornithine DHFR 5 -4.8 2.061 4.555 N-acetylornithine DHFR 6 -4.7 2.803 3.868 N-acetylornithine DHFR 7 -4.6 4.581 7.051 N-acetylornithine DHFR 8 -4.5 2.27 4.957 N-acetylornithine DHFR 9 -4.3 2.845 4.299 Genistein PTGS2 1 -7.9 0 0 Genistein PTGS2 2 -7 1.052 2.324 Genistein PTGS2 3 -6.8 19.582 21.764 Genistein PTGS2 4 -6.6 1.41 6.739 Genistein PTGS2 5 -6.3 19.615 21.653 Genistein PTGS2 6 -6.2 18.723 21.016 Genistein PTGS2 7 -6 19.559 21.672 Genistein PTGS2 8 -5.7 19.658 21.701 Genistein PTGS2 9 -5.5 20.121 22.076 Phenylacetic Acid PTGS2 1 -5.8 0 0 Phenylacetic Acid PTGS2 2 -5.8 1.067 1.413 Phenylacetic Acid PTGS2 3 -5.8 1.053 1.587 Phenylacetic Acid PTGS2 4 -5.7 1.13 1.663 Phenylacetic Acid PTGS2 5 -5.6 2.602 4.092 Phenylacetic Acid PTGS2 6 -5.6 2.701 4.094 Phenylacetic Acid PTGS2 7 -5.3 20.275 21.56 Phenylacetic Acid PTGS2 8 -5.2 1.935 2.811 Phenylacetic Acid PTGS2 9 -5.2 3.228 4.481 3.9. Building nomograms, assessing model performance, and analyzing the immunological microenvironment 3.9.1. Risk prediction and nomogram building In order to develop a prediction model that can rapidly and independently evaluate each patient's risk of comorbidity between MASLD and T2DM, we created a nomogram based on four important biomarkers (ALOX15B, PTGS2, KLK3, and DHFR). Each biomarker's expression level is converted by the nomogram into a matching score, which is then totaled together to determine the patient's overall risk of developing T2DM and MASLD. ALOX15B gene have comparable points of around 60 in the MASLD dataset (GSE89632), whereas PTGS2, KLK3, and DHFR genes have corresponding points of roughly 12, 8.5, and 13 points, respectively. The cumulative total score is 235 points. The patient is at high risk for MASLD, as indicated by the estimated risk of 0.835 based on this total score (Figure 6A). The nomogram construction results showed that the ALOX15B gene score for the T2DM dataset (GSE26168) was low, at about 5 points. The integrals of other genes are likewise mapped in accordance with the KLK3 gene score, which is greater at around 70 points. A moderate risk level was indicated by the cumulative total score of 123 points and the prediction risk of type 2 diabetes of 0.15 (Figure 6E). 3.9.2. Evaluate the predictive model's effectiveness 1.Receiver operating characteristic (ROC) curve analysis: We evaluated the nomogram model’s predictive performance by plotting ROC curves and calculating the area under the curve (AUC), a key discriminative metric. The model showed excellent discrimination in the MASLD dataset (GSE89632) with an AUC of 0.982 (95% CI: 0.947-1.000) (Figure 6B) and high accuracy in the T2DM dataset (GSE26168) with an AUC of 0.903 (95% CI: 0.855-0.951) (Figure 6F). 2.Decision curve analysis (DCA): To assess clinical utility, DCA showed the model yielded a net benefit over "treat all" and "treat none" strategies across most clinically acceptable high-risk thresholds in both datasets (Figure 6C, G). Benefits were more pronounced in low-to-medium thresholds, enabling precise identification of patients requiring intervention while avoiding overtreatment of low-risk groups. 3.Hosmer-Lemeshow test: Results showed p-values >0.05 in both datasets, indicating the nomogram had good goodness-of-fit and accurately predicted MASLD-T2DM comorbidity risk. Its robust fit supports reliable clinical risk assessment (Figure 6D, H). 3.9.3. We analyzed the relationship between important genes and immune cell infiltration 1.Immune cell infiltration analysis: Using immune cell infiltration analysis, we investigated the relative abundance of immune cells in T2DM (GSE26168) and MASLD (GSE89632) patient tissues. In MASLD patients (Figure 6I), regulatory T cells, monocytes, myeloid suppressor cells (MDSCs), and neutrophils showed significantly higher infiltration than the control group (p < 0.05), highlighting the role of immunomodulation and inflammation in MASLD pathophysiology; natural killer T cells (NKT cells) trended upward. In T2DM patients (Figure 6K), monocytes, neutrophils, and regulatory T cells also had significantly higher infiltration, indicating a close link between immune activation and T2DM progression. 2.Gene-immune cell correlation analysis: Spearman correlation heatmaps showed PTGS2 in MASLD samples (Figure 6J) had a significant positive correlation (p<0.05) with regulatory T cells, monocytes, and MDSCs (potential immunosuppressive regulation); KLK3 correlated negatively with natural killer cells and effector memory CD8+ T cells, while ALOX15B negatively correlated with multiple immune cells (anti-inflammatory potential). In T2DM samples (Figure 6L), nearly all core genes (PTGS2, KLK3, DHFR, ALOX15B) significantly correlated with immune cells (e.g., regulatory T cells, neutrophils); KLK3’s negative correlations with multiple immune cells suggest potential inhibition of diabetes-related immune inflammation. 3.ceRNA regulatory network construction: By integrating miRwalk and miRTarBase databases, we constructed a network including 6 key miRNAs (hsa-miR-2278, hsa-miR-588, hsa-miR-676-3p, hsa-miR-654-5p, hsa-miR-543, hsa-miR-520a-5p) and 4 core genes (KLK3, PTGS2, DHFR, ALOX15B) (Figure 6M), providing novel insights into the molecular regulatory mechanisms of T2DM-MASLD comorbidity. 4. Discussion 4.1. Overview of the research findings and their clinical significance This study systematically reveals the MASLD-T2DM comorbidity mechanism via integrating transcriptomics, gut microbiome GWAS, and Mendelian randomization (MR) analysis, highlighting the gut microbiota and its metabolites as core regulators of host gene expression. MR analysis filtered 14 gut microbiota linked to both diseases, with genus.Bifidobacterium.id.436 as the core candidate exerting disease-specific causal effects: for MASLD, OR = 0.78 (95% CI: 0.70384–0.86439, P < 0.05), suggesting increased abundance reduces MASLD risk (e.g., Bifidobacterium pseudolongum prevents NAFLD-HCC progression in mouse models; for T2DM, OR = 1.00055 (95% CI: 1.00003-1.00108, P < 0.05), indicating it promotes T2DM as a risk factor. This dual effect was confirmed by IVW and robustified via sensitivity analysis, potentially related to T2DM-related gut dysbiosis (e.g., increased Desulfovibrio) and metabolic disorders (e.g., insulin resistance)( 17 – 20 ).The gutMGene database was used to integrate microbiota-related metabolites, and finally 7 core biomarkers (PTGS2, ALOX15B, KLK3, DHFR, etc.) were identified, which were involved in lipid metabolism( 21 ), inflammation regulation( 22 ), and one-carbon metabolism( 23 ).Molecular docking showed that metabolites such as Genistein, Phenylacetic Acid, and N-acetylornithine could bind to key targetswith ≤ -5 kcal/mol, suggesting potential molecular regulatory mechanisms. The nomogram model based on these markers had AUC values of 0.982 and 0.903 in the two datasets, respectively, and the DCA confirmed its clinical net benefit, and the Hosmer-Lemeshow test (P > 0.05) showed a good fit. Immune infiltration analysis revealed that key genes were significantly related to immune cells such as regulatory T cells (Tregs), monocytes, and neutrophils, and ceRNA networks further elucidated the multi-level regulation mediated by miRNAs (e.g., hsa-miR-2278 and hsa-miR-588). The clinical significance of these findings lies in first, providing biomarkers and predictive tools for the comorbidity of MASLD and T2DM, which helps in early screening of high-risk populations and optimizes resource allocation. Second, revealing the dual role of the Bifidobacterium genus challenges the traditional view of "probiotics universally protect" and guides personalized microbial intervention strategies. Lastly, these findings promote the shift from association to causation in precision medicine, thereby reducing the global burden of metabolic diseases. According to WHO data, the risk of cardiovascular events in patients with comorbidity of MASLD and T2DM increases by 2–3 times, and this study's model can improve the screening accuracy to over 90%, which has significant public health value( 24 – 26 ). 4.2. Interpretation of Mendelian Randomization Analysis Results: Bifidobacterium genus' dual causal effect The primary novelty of this work lies in Mendelian randomization (MR) analysis, which avoids confounding bias in observational studies by simulating randomized controlled trials using genetic variations as instrumental variables (IVs). To ensure IV strength, we selected SNPs with F statistics > 10 (e.g., rs182549, rs6754311) from 211 gut microbiome clusters derived from the MiBioGen consortium. Forest plots identified 4 core microbiomes (including Streptococcus and Bifidobacterium) at the intersection of 27 MASLD-associated and 27 T2DM-associated microbiomes. Bifidobacterium improves metabolic function via short-chain fatty acids (SCFAs, e.g., butyrate); studies show MASLD-HCC patients have significantly higher SCFA levels (acetic acid, propionic acid, butyrate), and increased Bifidobacterium longum/adolescentis correlates with reduced HbA1c and basal insulin demand( 27 – 29 ). These metabolites activate GPR43/41 receptors to promote hepatic lipid oxidation and inhibit fat synthesis, alleviating MASLD liver fat deposition. Scatter/funnel plots confirm no significant horizontal pleiotropy (MR-Egger intercept P > 0.05) and low heterogeneity (Cochran's Q P > 0.05), supporting causal inference. Conversely, Bifidobacterium exerts considerable detrimental effects in T2DM (OR > 1) via three mechanisms: 1) Inflammatory/metabolic interference: Overfermentation produces lactic acid/hydrogen, disrupting intestinal barrier integrity and inducing systemic low-grade inflammation; metabolites like N-acetylornithine inhibit GLUT4-mediated glucose uptake, while phenylacetic acid activates the TLR4 pathway to exacerbate insulin resistance; 2) Masked antioxidant protection: Bifidobacterium reduces liver oxidative stress via regulating PTGS2 in MASLD, but this effect is masked in T2DM’s systemic metabolic disorder; 3) Host environment-dependent effects: Decreased intestinal pH in T2DM promotes Bifidobacterium proliferation, with its metabolites’ negative impacts dominating—unlike MASLD (predominantly local liver inflammation), where antioxidant effects prevail. This underscores the importance of stratified analysis: European cohort studies failed to distinguish MASLD subtypes, potentially overlooking microbiome disease-specific effects( 30 – 33 ). Strain specificity, host baseline metabolic status (e.g., intestinal pH differences), and intervention duration may explain result discrepancies( 31 ). Other intersecting bacterial communities (e.g., Streptococcus) consistently act as risk factors (OR > 1). As commensal pathogens, their increased abundance exacerbates metabolic diseases via: 1) Producing endotoxins (e.g., lipopolysaccharides, LPS) that activate the TLR4/NF-κB pathway, inducing hepatic/peripheral low-grade inflammation and insulin resistance( 34 , 35 ); 2) Increasing intestinal permeability, disrupting mucosal integrity, and facilitating translocation of endotoxins/metabolites (e.g., phenylacetic acid) to the portal circulation, aggravating systemic metabolic disorders and liver fat deposition( 36 , 37 ); 3) Producing metabolites like trimethylamine N-oxide (TMAO) that directly inhibit insulin signaling (e.g., PI3K/Akt) and increase comorbidity risks (e.g., atherosclerosis)( 38 – 40 ). Gut microbiota (e.g., Bifidobacterium, Streptococcus) regulates host metabolic gene expression (e.g., PTGS2, TLR4) via metabolites (lactic acid, LPS), while host metabolic status (e.g., hyperglycemia, low pH) feedback-regulates microbiota composition, forming a positive feedback loop. Stratified MR analysis confirms microbiota effects differ between T2DM patients with/without MASLD. Environmental factors (antibiotic abuse, high-fat diet) and genetic variations (e.g., PNPLA3) synergistically amplify metabolic risks (e.g., MASLD progression to liver fibrosis) by altering microbiota structure(7, 41). 4.3. Molecular mechanism of the gut microbiota-metabolite-host gene regulatory axis This study’s gut-metabolite-biomarker network reveals a multi-level regulatory mechanism.GutMGene/SwissTargetPrediction predictions indicate Bifidobacterium produces metabolites (Genistein, Phenylacetic Acid, N-acetylornithine) that intersect with DEGs to identify 7 core genes. GO enrichment highlights arachidonic acid metabolism (P < 0.01) and oxidoreductase activity, with KEGG pathways involving VEGF signaling and folate biosynthesis (P < 0.05), reflecting integrated regulation of lipid-inflammatory-energy homeostasis. Metabolites exert disease-specific effects: Genistein binds PTGS2 (Thr-145/Val-349, binding energy − 7.9 kcal/mol) to inhibit PGE2, alleviating MASLD liver fibrosis; Phenylacetic Acid activates COX-2/JNK to promote T2DM β-cell apoptosis( 42 – 44 )ALOX15B downregulation protects MASLD livers but disrupts T2DM insulin signaling( 45 ); N-acetylornithine binds DHFR (-5.0 kcal/mol) to regulate folate cycle, inhibiting T2DM insulin signaling (SREBP-1c pathway) and promoting MASLD liver regeneration( 45 ). SwissADME/ADMETlab confirm drug-likeness: Genistein has > 30% bioavailability, no hERG/DILI toxicity, and complies with Lipinski’s rule (HBA = 5, TPSA = 90.9 Ų). Network analysis shows a "many-to-many" pattern (e.g., Bifidobacterium + Streptococcus synergy)( 46 ). The ceRNA network involves 6 miRNAs (e.g., hsa-miR-543 targeting PTGS2); hsa-miR-676-3p upregulates in high-glucose environments to enhance T2DM inflammation but downregulates ALOX15B to protect MASLD livers, consistent with the MALAT1-miR-200c-PTGS2 axis( 45 , 47 ). 4.4. Regulatory mechanisms of the immune microenvironment and comorbidities mediated by inflammation ssGSEA immunosuppression analysis identifies chronic inflammation as the key driver of MASLD-T2DM comorbidity: T2DM shows increased neutrophils/NKT cells, while MASLD has elevated Treg/monocytes/MDSC infiltration (P < 0.01). Spearman correlation links Treg to PTGS2, mediating hepatic/pancreatic fibrosis and immune suppression via PGE2( 48 , 49 ). Bifidobacterium-derived SCFAs alleviate MASLD inflammation by activating Foxp3 + Treg but induce T2DM insulin resistance via excessive neutrophil recruitment( 50 , 51 ).Dysbiosis increases intestinal permeability; LPS leakage activates the TLR4/NF-κB pathway, amplifying PTGS2 (COX-2)/ALOX15B signaling and chronic inflammation( 48 , 52 ).ALOX15B downregulation protects MASLD livers but accelerates T2DM pancreatic inflammation. SCFAs improve MASLD metabolism by inhibiting HDAC yet aggravate T2DM β-cell injury via neutrophil ROS activation( 50 , 51 , 53 ). 4.5. Comparison with Existing Research Current research often focuses on specific disease molecular mechanisms. Petta et al.’s meta-analysis identified over 500 MASLD DEGs but failed to incorporate T2DM comorbid characteristics, leaving shared pathways unexposed( 54 ). Huang et al. validated Akkermansia’s preventive effect on T2DM via Mendelian randomization (OR = 0.88)( 55 ). but did not clarify Bifidobacterium’s disease-specific benefit. Studies show Bifidobacterium metabolites (phenylacetic acid, genistein) exert a "protection-injury" double-edged sword effect by targeting PTGS2/ALOX15B, with action direction dependent on disease microenvironments (e.g., high-fat vs. high-sugar), which aligns well with this study’s findings( 56 – 58 ). 4.6. The Study's Limitations Despite improved reliability from multi-omics integration, drawbacks remain: ( 1 ) Data sources: MiBioGen focuses on European populations, with Asian bias limiting generalizability; GEO datasets have small sample sizes requiring larger confirmatory studies; ( 2 ) MR restrictions: Horizontal pleiotropy is notable in MASLD (P < 0.05); ( 3 ) Mechanism inference: Molecular docking ignores dynamic conformations (relying on crystal structures); ceRNA predictions are mostly bioinformatic, lacking ChIP/luciferase validation; ( 4 ) Clinical translation: In silico toxicity evaluation needs in vitro/in vivo validation; nomograms exclude clinical factors (BMI/HbA1c); ( 5 ) Causal direction: Longitudinal cohorts are required to confirm Bifidobacterium’s dual role( 59 , 60 ). 4.7. Future Research Directions and Clinical Application Prospects Future studies can leverage single-cell transcriptomics to explore how cellular heterogeneity modulates MASLD-T2DM comorbidity, while investigating the gut microbiota–metabolite axis—particularly the role of Bifidobacterium-derived Genistein in MASLD models—and developing CRISPR-based gene editing tools to generate optimized strains. In high-fat diet-fed mouse models, fecal microbiota transplantation can be employed to evaluate the impacts of wild-type and gene-deletion strains on hepatic and pancreatic histological changes as well as glucose tolerance. To verify the therapeutic efficacy of engineered probiotics (e.g., genistein-expressing strains) in patients with MASLD and T2DM, phase II randomized controlled trials (RCTs) are recommended, with efficacy assessments enhanced by integrating FibroScan imaging and microbiome data. By refining nomogram models through artificial intelligence and incorporating multimodal data (such as metabolomics and medical imaging), the goal is to achieve a prediction accuracy exceeding 95%. In epigenetics, miRNA knockdown assays should be performed to validate the functional significance of the ceRNA regulatory axis, alongside exploring the therapeutic potential of epigenetic drugs like histone deacetylase (HDAC) inhibitors( 61 ). Overall, future microbiome therapies will shift from "broad-spectrum probiotics" to "precision microbiota intervention," which is expected to substantially alleviate the global disease burden of T2DM and MASLD—projected to affect over one billion patients by 2030( 62 ). In summary, this study systematically clarifies the gut microbiota-driven comorbidity mechanism of MASLD-T2DM, identifies core biomarkers, therapeutic targets and predictive models, and lays a robust scientific foundation for precision medicine interventions. Declarations Author Contributions: Conceptualization, Y.Z. and Y.C.; methodology, W.M.; software, W.Y.; validation, Y.Z., W.M. and Y.C.; formal analysis, Z.L.; investigation, L.L.; resources, L.L.; data curation, Y.Z.; writing—original draft preparation, Y.Z.; writing—review and editing, L.L.; visualization, Z.L.; supervision, Z.Y.; project administration, L.L; funding acquisition, L.L. All authors have read and agreed to the published version of the manuscript. Funding: No funding was received for this work. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The exposure data in the Mendelian randomization study were obtained from the MiBioGen consortium (https://mibiogen.gcc.rug.nl/, download date: 2025/09/30), and the outcome data were sourced from the IEU OpenGWAS database (https://gwas.mrcieu.ac.uk/, download date: 2025/09/30), including ebi-a-GCST90091033 (a total of 778,614 samples: 8,434 cases of MASLD and 770,180 controls) and ukb-b-13806 (2,972 cases of T2DM and 459,961 controls). The Gene Expression Omnibus provided two datasets: dataset GSEGSE26168 and GSE89632 (https://www.ncbi.nlm.nih.gov/geo/, download date: 2025/09/30). Conflicts of Interest: The authors declare no conflicts of interest. References Ma C, Wang S, Dong B, Tian Y. Metabolic Reprogramming of Immune Cells in Mash. Hepatology (2025). Epub 20250505. doi: 10.1097/hep.0000000000001371. Huang DQ, Wong VWS, Rinella ME, Boursier J, Lazarus JV, Yki-Jarvinen H, et al. Metabolic Dysfunction-Associated Steatotic Liver Disease in Adults. Nat Rev Dis Primers (2025) 11(1):14. Epub 20250306. doi: 10.1038/s41572-025-00599-1. Younossi ZM, Kalligeros M, Henry L. Epidemiology of Metabolic Dysfunction-Associated Steatotic Liver Disease. Clin Mol Hepatol (2025) 31(Suppl):S32-S50. Epub 2024/08/20. doi: 10.3350/cmh.2024.0431. Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. Idf Diabetes Atlas: Global, Regional and Country-Level Diabetes Prevalence Estimates for 2021 and Projections for 2045. Diabetes Res Clin Pract (2022) 183:109119. Epub 20211206. doi: 10.1016/j.diabres.2021.109119. Wu JY, Chang HY, Tsung Y, Lin YM. Clinical Outcomes of Sglt2 Inhibitors among Patients with Masld and T2dm. Diabetes Res Clin Pract (2025) 229:112918. Epub 2025/09/27. doi: 10.1016/j.diabres.2025.112918. Cusi K, Abdelmalek MF, Apovian CM, Balapattabi K, Bannuru RR, Barb D, et al. Metabolic Dysfunction-Associated Steatotic Liver Disease (Masld) in People with Diabetes: The Need for Screening and Early Intervention. A Consensus Report of the American Diabetes Association. Diabetes Care (2025) 48(7):1057-82. Epub 2025/05/28. doi: 10.2337/dci24-0094. Han E, Han KD, Lee YH, Kim KS, Hong S, Park JH, et al. Association of Temporal Masld with Type 2 Diabetes, Cardiovascular Disease and Mortality. Cardiovasc Diabetol (2025) 24(1):289. Epub 2025/07/16. doi: 10.1186/s12933-025-02824-3. Kang M, Song J, Kang ES, Jang S, Kwak T, Kim Y, et al. Pathophysiology, Development, and Mortality of Major Non-Communicable Diseases in Metabolic Dysfunction-Associated Steatotic Liver Disease: A Comprehensive Review. Int J Biol Sci (2025) 21(13):5691-703. Epub 20250903. doi: 10.7150/ijbs.117211. Kim DY, Lee SY, Lee JY, Whon TW, Lee JY, Jeon CO, et al. Gut Microbiome Therapy: Fecal Microbiota Transplantation Vs Live Biotherapeutic Products. Gut Microbes (2024) 16(1):2412376. Epub 20241008. doi: 10.1080/19490976.2024.2412376. Schwenger KJP, Copeland JK, Ghorbani Y, Chen L, Comelli EM, Guttman DS, et al. Characterization of Liver, Adipose, and Fecal Microbiome in Obese Patients with Masld: Links with Disease Severity and Metabolic Dysfunction Parameters. #N/A (2025) 13(1):9. Epub 2025/01/15. doi: 10.1186/s40168-024-02004-7. Alam N, Jia L, Cheng A, Ren H, Fu Y, Ding X, et al. Global Research Trends on Gut Microbiota and Metabolic Dysfunction-Associated Steatohepatitis: Insights from Bibliometric and Scientometric Analysis. Front Pharmacol (2024) 15:1390483. Epub 2024/07/29. doi: 10.3389/fphar.2024.1390483. Garcia-Mateo S, Rondinella D, Ponziani FR, Miele L, Gasbarrini A, Cammarota G, et al. Gut Microbiome and Metabolic Dysfunction-Associated Steatotic Liver Disease: Pathogenic Role and Potential for Therapeutics. Best Pract Res Clin Gastroenterol (2024) 72:101924. Epub 2024/12/08. doi: 10.1016/j.bpg.2024.101924. Daniel N, Farinella R, Chatziioannou AC, Jenab M, Mayén AL, Rizzato C, et al. Genetically Predicted Gut Bacteria, Circulating Bacteria-Associated Metabolites and Pancreatic Ductal Adenocarcinoma: A Mendelian Randomisation Study. Sci Rep (2024) 14(1):25144. Epub 20241024. doi: 10.1038/s41598-024-77431-5. Liu T, Cao Y, Liang N, Ma X, Fang JA, Zhang X. Investigating the Causal Association between Gut Microbiota and Type 2 Diabetes: A Meta-Analysis and Mendelian Randomization. Front Public Health (2024) 12:1342313. Epub 2024/07/04. doi: 10.3389/fpubh.2024.1342313. Burgess S, Woolf B, Mason AM, Ala-Korpela M, Gill D. Addressing the Credibility Crisis in Mendelian Randomization. BMC Med (2024) 22(1):374. Epub 20240911. doi: 10.1186/s12916-024-03607-5. Zhang Y, Fan J. Impact of Gut Microbiota on Metabolic Syndrome and Its Comprising Traits: A Two-Sample Mendelian Randomization Study. Diabetol Metab Syndr (2024) 16(1):279. Epub 2024/11/23. doi: 10.1186/s13098-024-01520-8. Lin YC, Lin HF, Wu CC, Chen CL, Ni YH. Pathogenic Effects of Desulfovibrio in the Gut on Fatty Liver in Diet-Induced Obese Mice and Children with Obesity. J Gastroenterol (2022) 57(11):913-25. Epub 2022/08/18. doi: 10.1007/s00535-022-01909-0. Nychas E, Marfil-Sanchez A, Chen X, Mirhakkak M, Li H, Jia W, et al. Discovery of Robust and Highly Specific Microbiome Signatures of Non-Alcoholic Fatty Liver Disease. #N/A (2025) 13(1):10. Epub 2025/01/15. doi: 10.1186/s40168-024-01990-y. Kim G, Yoon Y, Park JH, Park JW, Noh MG, Kim H, et al. Bifidobacterial Carbohydrate/Nucleoside Metabolism Enhances Oxidative Phosphorylation in White Adipose Tissue to Protect against Diet-Induced Obesity. #N/A (2022) 10(1):188. Epub 2022/11/06. doi: 10.1186/s40168-022-01374-0. Song Q, Zhang X, Liu W, Wei H, Liang W, Zhou Y, et al. Bifidobacterium Pseudolongum-Generated Acetate Suppresses Non-Alcoholic Fatty Liver Disease-Associated Hepatocellular Carcinoma. J Hepatol (2023) 79(6):1352-65. Epub 2023/07/18. doi: 10.1016/j.jhep.2023.07.005. Benatzy Y, Palmer MA, Lütjohann D, Ohno RI, Kampschulte N, Schebb NH, et al. Alox15b Controls Macrophage Cholesterol Homeostasis Via Lipid Peroxidation, Erk1/2 and Srebp2. Redox Biol (2024) 72:103149. Epub 20240403. doi: 10.1016/j.redox.2024.103149. Simon LS. Role and Regulation of Cyclooxygenase-2 During Inflammation. Am J Med (1999) 106(5b):37s-42s. doi: 10.1016/s0002-9343(99)00115-1. Sehrawat R, Rathee P, Khatkar S, Akkol E, Khayatkashani M, Nabavi SM, et al. Dihydrofolatereductase (Dhfr) Inhibitors: A Comprehensive Review. Curr Med Chem (2023). Epub 20230310. doi: 10.2174/0929867330666230310091510. Gao C, Zhang L, Guo X, Lin X, Yang J, Wang Z, et al. 20s-O-Glc-Dm Regulates Fatty Acid Metabolism and Mitochondrial Function in the Treatment of Diabetes Mellitus-Associated Mafld. Phytomedicine (2025) 146:157136. Epub 2025/08/18. doi: 10.1016/j.phymed.2025.157136. Tian R, Li Y. Exploring the Pathogenesis of Mafld from an Immunological Perspective: From the Perspective of the Cgas/Sting/Nf-Kappab Signaling Pathway. Front Immunol (2025) 16:1674018. Epub 2025/10/15. doi: 10.3389/fimmu.2025.1674018. Zhao J, Zhao Y, Hu Y, Peng J. Targeting the Gpr119/Incretin Axis: A Promising New Therapy for Metabolic-Associated Fatty Liver Disease. Cell Mol Biol Lett (2021) 26(1):32. Epub 2021/07/09. doi: 10.1186/s11658-021-00276-7. Bell KJ, Saad S, Tillett BJ, McGuire HM, Bordbar S, Yap YA, et al. Metabolite-Based Dietary Supplementation in Human Type 1 Diabetes Is Associated with Microbiota and Immune Modulation. #N/A (2022) 10(1):9. Epub 2022/01/21. doi: 10.1186/s40168-021-01193-9. Lee PC, Wu CJ, Hung YW, Lee CJ, Mon HC, Chi CT, et al. Distinct Gut Microbiota but Common Metabolomic Signatures between Viral and Masld Hcc Contribute to Outcomes of Combination Immunotherapy. Hepatology (2025). Epub 2025/07/01. doi: 10.1097/HEP.0000000000001446. Tang J, Wei Y, Pi C, Zheng W, Zuo Y, Shi P, et al. The Therapeutic Value of Bifidobacteria in Cardiovascular Disease. NPJ Biofilms Microbiomes (2023) 9(1):82. Epub 2023/10/31. doi: 10.1038/s41522-023-00448-7. Shen X, Ma C, Yang Y, Liu X, Wang B, Wang Y, et al. The Role and Mechanism of Probiotics Supplementation in Blood Glucose Regulation: A Review. Foods (2024) 13(17). Epub 2024/09/14. doi: 10.3390/foods13172719. Fliegerova KO, Mahayri TM, Sechovcova H, Mekadim C, Mrazek J, Jarosikova R, et al. Diabetes and Gut Microbiome. Front Microbiol (2024) 15:1451054. Epub 2025/01/22. doi: 10.3389/fmicb.2024.1451054. Ayesha IE, Monson NR, Klair N, Patel U, Saxena A, Patel D, et al. Probiotics and Their Role in the Management of Type 2 Diabetes Mellitus (Short-Term Versus Long-Term Effect): A Systematic Review and Meta-Analysis. Cureus (2023) 15(10):e46741. Epub 2023/11/29. doi: 10.7759/cureus.46741. Qian X, Si Q, Lin G, Zhu M, Lu J, Zhang H, et al. Bifidobacterium Adolescentis Is Effective in Relieving Type 2 Diabetes and May Be Related to Its Dominant Core Genome and Gut Microbiota Modulation Capacity. #N/A (2022) 14(12). Epub 2022/06/25. doi: 10.3390/nu14122479. Guo W, Xiang Q, Mao B, Tang X, Cui S, Li X, et al. Protective Effects of Microbiome-Derived Inosine on Lipopolysaccharide-Induced Acute Liver Damage and Inflammation in Mice Via Mediating the Tlr4/Nf-Kappab Pathway. J Agric Food Chem (2021) 69(27):7619-28. Epub 20210622. doi: 10.1021/acs.jafc.1c01781. Wang X, Sun Z, Wang X, Li M, Zhou B, Zhang X. Solanum Nigrum L. Berries Extract Ameliorated the Alcoholic Liver Injury by Regulating Gut Microbiota, Lipid Metabolism, Inflammation, and Oxidative Stress. Food Res Int (2024) 188:114489. Epub 20240509. doi: 10.1016/j.foodres.2024.114489. Li J, Niu C, Ai H, Li X, Zhang L, Lang Y, et al. Tsp50 Attenuates Dss-Induced Colitis by Regulating Tgf-Beta Signaling Mediated Maintenance of Intestinal Mucosal Barrier Integrity. Adv Sci (Weinh) (2024) 11(11):e2305893. Epub 20240108. doi: 10.1002/advs.202305893. Yan J, Chen Q, Tian L, Li K, Lai W, Bian L, et al. Intestinal Toxicity of Micro- and Nano-Particles of Foodborne Titanium Dioxide in Juvenile Mice: Disorders of Gut Microbiota-Host Co-Metabolites and Intestinal Barrier Damage. Sci Total Environ (2022) 821:153279. Epub 20220121. doi: 10.1016/j.scitotenv.2022.153279. Li Y, Yang P, Ye J, Xu Q, Wu J, Wang Y. Updated Mechanisms of Masld Pathogenesis. Lipids Health Dis (2024) 23(1):117. Epub 2024/04/23. doi: 10.1186/s12944-024-02108-x. Sato S, Iino C, Furusawa K, Yoshida K, Chinda D, Sawada K, et al. Effect of Oral Microbiota Composition on Metabolic Dysfunction-Associated Steatotic Liver Disease in the General Population. #N/A (2025) 14(6). Epub 2025/03/27. doi: 10.3390/jcm14062013. Borges-Canha M, Centelles-Lodeiro J, Leite AR, Chaves J, Lourenco IM, Von-Hafe M, et al. Gut Dysbiosis Is Linked to Severe Steatosis and Enhances Its Diagnostic Performance in Masld. eGastroenterology (2025) 3(3):e100204. Epub 2025/09/08. doi: 10.1136/egastro-2025-100204. Zheng Z, Huang Y, Zhang J, Xie J, Pan A, Liao Y, et al. Antibiotic Consumption, Genetic Risk and Incidence of Metabolic Dysfunction-Associated Steatotic Liver Disease: A Prospective Cohort Study. Ann Hepatol (2025):102136. Epub 2025/10/12. doi: 10.1016/j.aohep.2025.102136. Wang J, Zhu N, Su X, Gao Y, Yang R. Gut-Microbiota-Derived Metabolites Maintain Gut and Systemic Immune Homeostasis. Cells (2023) 12(5). Epub 2023/03/12. doi: 10.3390/cells12050793. Rooks MG, Garrett WS. Gut Microbiota, Metabolites and Host Immunity. #N/A (2016) 16(6):341-52. Epub 2016/05/28. doi: 10.1038/nri.2016.42. Kim CH. Immune Regulation by Microbiome Metabolites. Immunology (2018) 154(2):220-9. Epub 2018/03/24. doi: 10.1111/imm.12930. Woo V, Alenghat T. Host-Microbiota Interactions: Epigenomic Regulation. Curr Opin Immunol (2017) 44:52-60. Epub 2017/01/20. doi: 10.1016/j.coi.2016.12.001. Kasubuchi M, Hasegawa S, Hiramatsu T, Ichimura A, Kimura I. Dietary Gut Microbial Metabolites, Short-Chain Fatty Acids, and Host Metabolic Regulation. #N/A (2015) 7(4):2839-49. Epub 2015/04/16. doi: 10.3390/nu7042839. Lei J, Wang X, Liu X. Microbiota-Derived Metabolites in the Epigenetic Regulation of the Host. Sci Bull (Beijing) (2025). Epub 2025/10/06. doi: 10.1016/j.scib.2025.09.030. Zhao H, Wu L, Yan G, Chen Y, Zhou M, Wu Y, et al. Inflammation and Tumor Progression: Signaling Pathways and Targeted Intervention. Signal Transduct Target Ther (2021) 6(1):263. Epub 2021/07/13. doi: 10.1038/s41392-021-00658-5. Petraglia F, Vannuccini S, Donati C, Jeljeli M, Bourdon M, Chapron C. Endometriosis and Comorbidities: Molecular Mechanisms and Clinical Implications. Trends Mol Med (2025). Epub 2025/10/03. doi: 10.1016/j.molmed.2025.09.002. Liu M, Chen R, Zheng Z, Xu S, Hou C, Ding Y, et al. Mechanisms of Inflammatory Microenvironment Formation in Cardiometabolic Diseases: Molecular and Cellular Perspectives. Front Cardiovasc Med (2024) 11:1529903. Epub 2025/01/29. doi: 10.3389/fcvm.2024.1529903. Hu T, Liu CH, Lei M, Zeng Q, Li L, Tang H, et al. Metabolic Regulation of the Immune System in Health and Diseases: Mechanisms and Interventions. Signal Transduct Target Ther (2024) 9(1):268. Epub 2024/10/09. doi: 10.1038/s41392-024-01954-6. Dou J, Jiang J, Xue Y, Jiang X, Jiang Y, Xiao P, et al. The Interplay of Cross-Organ Immune Regulation in Inflammation and Cancer. MedComm (2020) (2025) 6(7):e70249. Epub 2025/06/18. doi: 10.1002/mco2.70249. Hui L, Li Y, Huang MK, Jiang YM, Liu T. Cxcl13: A Common Target for Immune-Mediated Inflammatory Diseases. Clin Exp Med (2024) 24(1):244. Epub 2024/10/24. doi: 10.1007/s10238-024-01508-8. Priya S, Burns MB, Ward T, Mars RAT, Adamowicz B, Lock EF, et al. Identification of Shared and Disease-Specific Host Gene-Microbiome Associations across Human Diseases Using Multi-Omic Integration. Nat Microbiol (2022) 7(6):780-95. Epub 2022/05/17. doi: 10.1038/s41564-022-01121-z. Liwinski T, Casar C, Ruehlemann MC, Bang C, Sebode M, Hohenester S, et al. A Disease-Specific Decline of the Relative Abundance of Bifidobacterium in Patients with Autoimmune Hepatitis. Aliment Pharmacol Ther (2020) 51(12):1417-28. Epub 2020/05/10. doi: 10.1111/apt.15754. Gavzy SJ, Kensiski A, Lee ZL, Mongodin EF, Ma B, Bromberg JS. Bifidobacterium Mechanisms of Immune Modulation and Tolerance. Gut Microbes (2023) 15(2):2291164. Epub 2023/12/06. doi: 10.1080/19490976.2023.2291164. Tojo R, Suarez A, Clemente MG, de los Reyes-Gavilan CG, Margolles A, Gueimonde M, et al. Intestinal Microbiota in Health and Disease: Role of Bifidobacteria in Gut Homeostasis. World J Gastroenterol (2014) 20(41):15163-76. Epub 2014/11/12. doi: 10.3748/wjg.v20.i41.15163. Bocchio F, Mancabelli L, Milani C, Lugli GA, Tarracchini C, Longhi G, et al. Compendium of Bifidobacterium-Based Probiotics: Characteristics and Therapeutic Impact on Human Diseases. Microbiome Res Rep (2025) 4(1):2. Epub 2025/04/10. doi: 10.20517/mrr.2024.52. Yan J, Wang Z, Bao G, Xue C, Zheng W, Fu R, et al. Causal Effect between Gut Microbiota and Metabolic Syndrome in European Population: A Bidirectional Mendelian Randomization Study. Cell Biosci (2024) 14(1):67. Epub 20240528. doi: 10.1186/s13578-024-01232-6. Chai Z, Su Y, Tian X, Chen C, Lv X, Chen C. Predicting Disease Associations Based on the Higher Order Structure of Cerna Networks. Brief Bioinform (2025) 26(5). doi: 10.1093/bib/bbaf518. Raucci A, Zwergel C, Valente S, Mai A. Advancements in Hydrazide-Based Hdac Inhibitors: A Review of Recent Developments and Therapeutic Potential. J Med Chem (2025) 68(14):14171-94. Epub 20250710. doi: 10.1021/acs.jmedchem.5c01677. Saeed H, Diaz LA, Gil-Gomez A, Burton J, Bajaj JS, Romero-Gomez M, et al. Microbiome-Centered Therapies for the Management of Metabolic Dysfunction-Associated Steatotic Liver Disease. Clin Mol Hepatol (2025) 31(Suppl):S94-S111. Epub 2024/11/28. doi: 10.3350/cmh.2024.0811. Additional Declarations No competing interests reported. Supplementary Files SupplementaryFile.zip Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/doi/s1, Figure S1: title; Table S1: title; Video S1: title. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9357546","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":638096814,"identity":"ad6a8d27-32d4-468f-9483-aba73dbaa448","order_by":0,"name":"Yang Zeng","email":"","orcid":"","institution":"Lanzhou University Second Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Zeng","suffix":""},{"id":638096815,"identity":"6dae14e1-a90b-4da4-b400-6b2b50291a5f","order_by":1,"name":"Yu Cheng","email":"","orcid":"","institution":"Lanzhou University Second Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Cheng","suffix":""},{"id":638096816,"identity":"8151e6db-a5b4-4e44-ab2d-ae7c41ac36cc","order_by":2,"name":"Wentao Ma","email":"","orcid":"","institution":"Lanzhou University Second Hospital","correspondingAuthor":false,"prefix":"","firstName":"Wentao","middleName":"","lastName":"Ma","suffix":""},{"id":638096817,"identity":"a60da764-5986-4b52-b7bc-ed1a7307ceb5","order_by":3,"name":"Wenjie Ye","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Wenjie","middleName":"","lastName":"Ye","suffix":""},{"id":638096818,"identity":"7d70afb4-20ed-4e76-8510-17a6d5f6042c","order_by":4,"name":"Zhenyu Liu","email":"data:image/png;base64,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","orcid":"","institution":"Xiamen University","correspondingAuthor":true,"prefix":"","firstName":"Zhenyu","middleName":"","lastName":"Liu","suffix":""},{"id":638096819,"identity":"ebfbc986-6714-4fd2-ba61-93d99a843a4c","order_by":5,"name":"Linjing Li","email":"","orcid":"","institution":"Lanzhou University Second Hospital","correspondingAuthor":false,"prefix":"","firstName":"Linjing","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2026-04-08 13:23:47","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9357546/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9357546/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109296616,"identity":"d81ad012-e8ef-4b1f-9896-f6c343f59d85","added_by":"auto","created_at":"2026-05-15 08:48:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":274417,"visible":true,"origin":"","legend":"\u003cp\u003eDifferential gene screening in transcriptome datasets: (\u003cstrong\u003eA,B\u003c/strong\u003e)heatmaps and volcano plots from GSE89632; (\u003cstrong\u003eC,D\u003c/strong\u003e)volcano plots and heatmaps from GSE26168; (\u003cstrong\u003eE,F\u003c/strong\u003e)and the intersections of upregulated and downregulated genes between the two.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/98dc8d27c59a5a2ffe2f0b8b.png"},{"id":109405234,"identity":"eb8863f5-845f-4ba9-a589-d68301404526","added_by":"auto","created_at":"2026-05-17 13:03:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":304703,"visible":true,"origin":"","legend":"\u003cp\u003eThe figure contains 5 columns, for Exposure ID (Exposure), p-value, model type, odds ratio, and their 95% confidence intervals, as well as OR (odds ratio) values. (\u003cstrong\u003eA,B\u003c/strong\u003e) IVW forest map visualize that 27 screened exposure factors link local MASLD to T2DM significantly.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/e514f249760ac70321992b41.png"},{"id":109297999,"identity":"550a2d5e-29fb-4054-a812-4aee6bcf14ef","added_by":"auto","created_at":"2026-05-15 09:07:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":178629,"visible":true,"origin":"","legend":"\u003cp\u003eMR analysis results with an example of \"genus.Bifidobacterium.id.436\". (\u003cstrong\u003eA-D\u003c/strong\u003e) in MASLD, there are scatter plots, forest plots, funnel plots, and leave-one-out sensitivity analysis plots. (\u003cstrong\u003eE-H\u003c/strong\u003e) in T2DM,there are scatter plots, forest plots, funnel plots, and leave-one-out sensitivity analysis plots.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/11988fe61c635d3f3860d836.png"},{"id":109296606,"identity":"cccc97e3-6161-4ba5-874b-b1fc6e1f9102","added_by":"auto","created_at":"2026-05-15 08:48:31","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":242522,"visible":true,"origin":"","legend":"\u003cp\u003eDetermination of biomarkers and construction of the MMTS network map. (\u003cstrong\u003eA\u003c/strong\u003e)Intersection of candidate microbiota for MASLD and T2DM; (\u003cstrong\u003eB\u003c/strong\u003e)Intersection of target genes of the candidate microbiota determined by the gutMgene database and transcriptome DEGs; (\u003cstrong\u003eC,D\u003c/strong\u003e) GO and KEGG enrichment analysis network maps for 7 genes; \u003cstrong\u003e(E\u003c/strong\u003e) PPI network map obtained by inputting 7 genes into the STRING website (confidence \u0026gt; 0.4); (\u003cstrong\u003eF\u003c/strong\u003e) microbiota-metabolite-biomarker network map.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/d69030d7286ba1c9dda0f052.png"},{"id":109296198,"identity":"fd8bb0a7-3deb-422b-b11d-218282b4bce5","added_by":"auto","created_at":"2026-05-15 08:46:06","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":273862,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic diagram of molecular docking. (\u003cstrong\u003eA\u003c/strong\u003e) Schematic diagram of docking between DHFR and N-acetylornithine; (\u003cstrong\u003eB\u003c/strong\u003e) Schematic diagram of docking between PTGS2 and Genistein; (\u003cstrong\u003eC\u003c/strong\u003e)Schematic diagram of docking between PTGS2 and Phenylacetic Acid.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/d7220b49a3e91cea72020eff.png"},{"id":109281315,"identity":"c54b1ca8-22ec-4b76-bd1c-9235ed1d603c","added_by":"auto","created_at":"2026-05-14 17:59:29","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":647291,"visible":true,"origin":"","legend":"\u003cp\u003eIntegrated MASLD and T2DM comorbidity risk prediction model and exploration of its molecular mechanism. (\u003cstrong\u003eA,B\u003c/strong\u003e)The nomogram and corresponding ROC curve based on the GSE89632 dataset; (\u003cstrong\u003eC\u003c/strong\u003e) DCA of the GSE89632 dataset, evaluating the clinical net benefit of the model; (\u003cstrong\u003eD\u003c/strong\u003e) Hosmer-Lemeshow test plot of the GSE89632 dataset, evaluating the goodness of fit of the model; (\u003cstrong\u003eE,F\u003c/strong\u003e) The nomogram and corresponding ROC curve are based on the GSE26168 dataset; (\u003cstrong\u003eG\u003c/strong\u003e) DCA of the GSE26168 dataset, evaluating the clinical net benefit of the model; (\u003cstrong\u003eH\u003c/strong\u003e)Hosmer-Lemeshow test plot of the GSE26168 dataset, evaluating the goodness of fit of the model;(\u003cstrong\u003eI\u003c/strong\u003e)Comparison of the proportion of immune cell infiltration in MASLD patients and the control group in the GSE89632 dataset;(\u003cstrong\u003eJ\u003c/strong\u003e) Spearman correlation heat map between the core genes and different immune cell types in the GSE89632 dataset; \u003cstrong\u003e(K\u003c/strong\u003e) Comparison of the proportion of immune cell infiltration in T2DM patients and the control group in the GSE26168 dataset; \u003cstrong\u003e(L\u003c/strong\u003e) Spearman correlation heat map between the core genes and different immune cell types in the GSE26168 dataset; (\u003cstrong\u003eM\u003c/strong\u003e)ceRNA regulatory network diagram based on core genes and key miRNAs, showing the potential molecular regulatory mechanisms.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/4020083367813ddbcdc2da34.png"},{"id":109406037,"identity":"f9913749-13ab-49ea-9472-8839c31e94fb","added_by":"auto","created_at":"2026-05-17 13:24:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2061365,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/035090d0-3abc-4650-b7cc-0fb28286f9ef.pdf"},{"id":109281310,"identity":"f4090179-68ed-49e8-9dea-4e3ad11e30ec","added_by":"auto","created_at":"2026-05-14 17:59:29","extension":"zip","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":110652,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/doi/s1, Figure S1: title; Table S1: title; Video S1: title.\u003c/p\u003e","description":"","filename":"SupplementaryFile.zip","url":"https://assets-eu.researchsquare.com/files/rs-9357546/v1/97384d2fbb2f914781da0d77.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Unraveling the Gut Microbiota-Metabolite-Host Gene Axis in the Pathogenesis of MASLD and Type 2 Diabetes Comorbidity","fulltext":[{"header":"1. Introduction","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e1.1. Metabolic-associated fatty liver disease (MASLD) and the epidemiological status of type 2 diabetes (T2DM)\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFormerly known as non-alcoholic fatty liver disease (NAFLD), metabolic fatty liver disease (MASLD) is one of the world\u0026rsquo;s fastest-growing chronic liver diseases, driven by global economic development and lifestyle Westernization(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Recent epidemiological data show a global adult prevalence of 25%-30% (up to 40% in specific regions)(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), reflecting high incidence and a youthification trend. In China, MASLD incidence has risen sharply amid lifestyle shifts, obesity, and metabolic syndrome, with adult prevalence in urban areas approaching 30%(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eType 2 diabetes mellitus (T2DM), the most common diabetes type, affects\u0026thinsp;~\u0026thinsp;500\u0026nbsp;million people worldwide, growing at \u0026gt;\u0026thinsp;4% annually(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e), with insulin resistance and obesity as key pathogenesis. Notably, 70%-80% of T2DM patients have hepatic steatosis, meaning MASLD prevalence in this group is far higher than in the general population(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Conversely, large cohort studies indicate MASLD patients face a\u0026thinsp;~\u0026thinsp;23-fold higher relative risk of developing T2DM over the next 5\u0026ndash;10 years(\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e1.2. The mutual relationship and harm between MASLD and T2DM\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eMASLD and T2DM are not simple comorbidities but a complex metabolic network interacting via inflammatory responses, insulin resistance, and lipid metabolism disorders(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Their coexistence markedly elevates risks of hepatic fibrosis, nephropathy, and cardiovascular disease, accelerates hepatic fibrosis progression, and increases cirrhosis and hepatocellular cancer prevalence(\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Additionally, insulin resistance worsens, hindering blood glucose control and augmenting therapeutic burden. Despite improved clinical diagnosis rates, the dual burden of MASLD and T2DM remains a major challenge in modern metabolic disease management.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e1.3. Research Progress on Gut Microbiota and Metabolic Diseases\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eKnown as the human \"second genome,\" the gut microbiota plays a key role in maintaining metabolic balance, with numerous studies linking its compositional and functional alterations to metabolic disorders like MASLD and T2DM(\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). By fermenting undigested dietary fibers, it produces metabolites (bile acid metabolites, tryptophan derivatives, short-chain fatty acids [SCFAs]) that regulate host energy metabolism, inflammation, and immunity, thereby impacting insulin sensitivity and liver fat accumulation(\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNotably, MASLD patients exhibit reduced gut flora diversity, increased pathogenic bacteria (certain Bacteroides/Bacteroidetes), and decreased probiotics (Bifidobacteria, Akkermansia)(\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). T2DM patients show similar imbalances, plus elevated intestinal permeability, which activates metabolic inflammation, impairing pancreatic β-cell function and insulin resistance(\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Thus, the gut microbiota and its metabolites act as a critical bridge between MASLD and T2DM, holding potential diagnostic and therapeutic value.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e1.4. The application value of Mendelian randomization analysis in metabolic disease research\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eTraditional observational studies struggle with confounding factors and reverse causality when exploring the gut microbiota-MASLD/T2DM association(\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e), hampering clear causal inference. Mendelian randomization (MR), a causal inference method based on genetic variation, uses genetic variants as instrumental variables to simulate random allocation, effectively avoiding environmental and behavioral interference. It has been widely applied to verify causal relationships in complex diseases(\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn gut microbiota research, MR enables investigating causality between specific bacterial abundances and metabolic disease risk(\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). For example, prior studies used MR to clarify the protective effect of certain probiotic genera on T2DM risk, laying a scientific foundation for the gut microbiota as a potential therapeutic target(\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). Additionally, integrating multi-omics data (transcriptomics, metabolomics) with MR helps reveal the molecular mechanism by which gut microbiota regulates host gene expression via metabolites, deepening understanding of the complex comorbidity network.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e1.5. Research objectives and significance\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study aims to clarify the gut microbiota-metabolite-host gene axis\u0026rsquo;s role in MASLD-T2DM comorbidity via transcriptomics, Mendelian randomization, and the gutMgene database, systematically analyzing key gut microbiota, their metabolites, and regulatory networks on host genes. Multidimensional data integration and bioinformatics will identify potential biomarkers/therapeutic targets, facilitating precision diagnosis/targeted therapies and advancing comorbidity prevention of T2DM and MASLD.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data Acquisition\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe Gene Expression Omnibus (GEO, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, download date: 2025/09/30) provided two datasets: Dataset GSEGSE26168, including mRNA and miRNA expression profiles of type 2 diabetes (T2DM) (experimental types: non-coding RNA expression profiling, microarray-based expression profiling). Samples consist of 5 healthy rats, 5 T2DM rats, 8 human healthy controls, 7 people with impaired fasting glucose, 9 T2DM patients, and miRNA sequencing samples (7 healthy controls, 7 IFG patients, 9 T2DM patients). This study selected 9 T2DM patients and 8 healthy controls as disease and control groups, respectively.Dataset GSE89632, focusing on mRNA expression profile of non-alcoholic fatty liver disease (NAFLD) using samples of human liver tissue (63 individuals: 20 simple fatty degeneration, 19 non-alcoholic fatty hepatitis, 24 healthy controls patients). This study designated healthy controls as the control group and NASH cases as the disease group.\u003c/p\u003e \u003cp\u003eGut microbiome dataset was obtained from the MiBioGen consortium (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://mibiogen.gcc.rug.nl/\u003c/span\u003e\u003cspan address=\"https://mibiogen.gcc.rug.nl/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, download date: 2025/09/30), comprising fecal microbiome profiles. It includes 211 gut microbial taxonomic groups ( 131 genera, 35 families, 20 orders, 16 classes, and 9 phyla) and a total of 14,587 SNP loci.\u003c/p\u003e \u003cp\u003eGWAS data were retrieved from the IEU OpenGWAS database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://gwas.mrcieu.ac.uk/\u003c/span\u003e\u003cspan address=\"https://gwas.mrcieu.ac.uk/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, download date: 2025/09/30), including ebi-a-GCST90091033, (778,614 total samples: 8,434 MM cases and 770,180 controls) and ukb-b-13806(2,972 TT cases and 459,961 controls).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Candidate gene identification\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eBulk_DEGs_T2DM and bulk_DEGs_MASLD were differential gene screening results, independently obtained via the R package \"limma (3.56.2)\" on disease/control samples in T2DM and MASLD training sets (threshold: P.Value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 \u0026amp; |logFC| \u0026gt; 0.5). R packages \"ggplot2 (3.5.2)\" and \"pheatmap (1.0.12)\" were used to generate heatmaps and volcano plots. DEGs_MASLD and DEGs_T2DM were intersected using the \"Venn Diagram (1.7.3)\" package to yield DEGs_MASLD_T2DM.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Mendelian Randomization Analysis\u003c/h2\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1. Selection and Extraction of Instrumental Variables (IVs)\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003ePutative instrumental variables (SNPs) associated with the gut microbiome (exposure factor) and T2DM/MASLD (outcome variables) were extracted from public GWAS datasets using the extract_instruments function of the R package TwoSampleMR (0.6.8). Screening followed five criteria: 1) SNP-exposure association strength (P\u0026thinsp;\u0026lt;\u0026thinsp;5\u0026times;10⁻⁶); 2) linkage disequilibrium (LD) exclusion to ensure IV independence; 3) exclusion of SNPs significantly correlated with outcomes to reduce confounding bias; 4) a minimum of three independent SNPs per exposure; 5) IV strength assessment (F-statistic\u0026thinsp;\u0026gt;\u0026thinsp;10) to avoid weak instrumental variable bias.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2. MR Analysis and Effect Size Evaluation\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAfter obtaining qualified instrumental variables, we use the mr function from the TwoSampleMR (0.6.8) package, combined with various MR methods (including the main inverse variance weighting method (IVW), weighted median method, weighted mode method, simple mode method, and MR-Egger regression) to evaluate the causal effects of the gut microbiota on T2DM and MASLD. A p-value less than 0.05 is used as the threshold for judging potential causal associations.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e2.3.3. Visualization Analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eTo visually present the analysis results, we adopted the following visualization methods: 1) Scatter plot (mr_scatter_plot): Used to visually demonstrate the correlation between exposure factors (gut microbiota) and outcome variables (T2DM/MASLD); 2) Forest plot (mr_forest_plot): Used to evaluate the effectiveness of each SNP locus in contributing to outcome variables and to summarize the effect sizes; 3) Funnel plot (mr_funnel_plot): Used to assess whether the effect of candidate exposure factors shows random distribution across different strengths of instrumental variables to preliminarily determine the presence of horizontal pleiotropy.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e2.3.4. Sensitivity Analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eTo ensure MR analysis robustness and reliability, we performed four sensitivity analyses: 1) Heterogeneity Test (mr_heterogeneity): Cochran's Q test evaluates SNP heterogeneity\u0026mdash;fixed-effect IVW model for P\u0026thinsp;\u0026gt;\u0026thinsp;0.05 (no significant heterogeneity) and random-effect IVW for P\u0026thinsp;\u0026le;\u0026thinsp;0.05; 2) Pleiotropy Test (mr_pleiotropy_test): MR-Egger intercept and mr_presso detect horizontal pleiotropy (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05\u0026thinsp;=\u0026thinsp;no significant bias); 3) Leave-One-Out Test (mr_leaveoneout): Sequentially remove each SNP to observe overall effect size changes, visualized via mr_leaveoneout_plot; 4) Directionality Test (directionality_test): Assess causal direction (exposure\u0026rarr;outcome vs. outcome\u0026rarr;exposure), retaining results with P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and directionality judgment (TRUE).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Candidate Metabolite Targets Acquisition\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eCandidate microbiota were retrieved from the gutMGene v3.0 database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://bio-annotation.cn/gutmgene/\u003c/span\u003e\u003cspan address=\"http://bio-annotation.cn/gutmgene/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed on 2025/10/08), and SMILES (Simplified Molecular Input Line Entry System) patterns of candidate metabolites were identified via the PubChem database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubchem.ncbi.nlm.nih.gov/\u003c/span\u003e\u003cspan address=\"https://pubchem.ncbi.nlm.nih.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed on 2025/10/08). Under the \"Homo sapiens\" model, the Similarity Ensemble Approach (SEA, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://sea.bkslab.org/\u003c/span\u003e\u003cspan address=\"http://sea.bkslab.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed on 2025/10/08) and SwissTargetPrediction (STP, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.swisstargetprediction.ch/\u003c/span\u003e\u003cspan address=\"http://www.swisstargetprediction.ch/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed on 2025/10/08) were used to predict targets associated with each metabolite in the candidate microbiota. Overlapping targets between SEA and STP were further analyzed using the Venny 2.1.0 platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bioinfogp.cnb.csic.es/tools/venny/index.html\u003c/span\u003e\u003cspan address=\"https://bioinfogp.cnb.csic.es/tools/venny/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed on 2025/10/08).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Enrichment Analysis and PPI Network Construction\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn this study, candidate metabolite targets were first intersected with differentially expressed genes (DEGs) to obtain key genes. Functional enrichment analysis (Gene Ontology [GO, covering biological processes, cellular components, molecular functions] and Kyoto Encyclopedia of Genes and Genomes [KEGG] pathways) was then performed on these genes using the R package \"clusterProfiler (4.15.0.003)\", with P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 as the screening criterion to identify potential biological functions and involved pathways. Finally, the online STRING platform was used to construct a protein-protein interaction (PPI) network for candidate genes; after gene input, an interaction score threshold\u0026thinsp;\u0026gt;\u0026thinsp;0.4 was set to filter high-interaction proteins, thereby determining the final biomarkers.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e2.6. Metabolite Analysis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eSwissADME Metabolite Drug Similarity Analysis: SwissADME was used to assess the drug-likeness of metabolites in biomarker and signaling pathway network analysis, with screening criteria: hydrogen bond acceptor (HBA)\u0026thinsp;\u0026lt;\u0026thinsp;10, hydrogen bond donor (HBD)\u0026thinsp;\u0026le;\u0026thinsp;5, octanol-water partition coefficient (MLog P)\u0026thinsp;\u0026le;\u0026thinsp;4.15, topological polar surface area (TPSA (\u0026Aring;\u0026sup2;))\u0026thinsp;\u0026lt;\u0026thinsp;140, and Lipinski's rule\u0026thinsp;\u0026le;\u0026thinsp;1. ADMETlab 3.0 Toxicology Property Evaluation: The online tool ADMETlab 2.0 was used to evaluate the toxicological properties of candidate core metabolites, assess toxicological risks (hERG, human liver toxicity H-HT, rat oral acute toxicity, carcinogenicity, DILI, etc.), and prioritize metabolites with negative toxicity predictions.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e2.7. Molecular Docking\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eDownload the three-dimensional structure of small molecules (SDF/SMILES) from PubChem, download the receptor protein structure (PDB file) from PDB, then use OpenBabel/AutoDockTools (ADT) or UCSF Chimera for protonation, removal of ligands/water molecules, addition of hydrogen atoms, and generation of PDBQT files. Use AutoDock Vina for molecular docking, with \u0026le;\u0026thinsp;\u0026minus;\u0026thinsp;5 kcal/mol as the reference threshold for good binding potential.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e2.8. Construction of Nomogram and Model Evaluation\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eA nomogram is a visualization tool based on a multi-factor regression model, which converts complex regression equations into intuitive graphical representations to facilitate individualized clinical risk assessment. In this study, a nomogram model was constructed using R package \"rms\" based on the expression data of selected key biomarkers, mapping each key gene\u0026rsquo;s expression level to corresponding risk scores; the total cumulative score of markers was calculated to predict patients\u0026rsquo; disease risk.\u003c/p\u003e \u003cp\u003eAfter model construction, calibration curves were used to evaluate the consistency between predicted probability and actual risk, and the Hosmer\u0026minus;Lemeshow test to assess model goodness-of-fit. Decision curve analysis (DCA) was further applied to measure the model\u0026rsquo;s clinical utility by evaluating its net benefit at different risk thresholds, ensuring statistical significance and practical value for clinical decision-making. This process enhanced the predictive model\u0026rsquo;s credibility and practicality, providing a scientific basis for patient disease management and treatment.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e2.9. Immune Infiltration Analysis and Gene-Cell Correlation\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eGiven the immune system\u0026rsquo;s key role in metabolic diseases, this study used single-sample gene set enrichment analysis (ssGSEA) via the R package \"GSVA (1.53.28)\" to quantitatively assess the infiltration of 28 immune cell types in transcriptome data. Differential immune cell subsets were identified by comparing immune cell enrichment scores between disease and control groups. The Wilcoxon rank-sum test was used to statistically analyze intergroup differences in immune cell infiltration, and the \"psych (2.4.6.26)\" package to calculate Spearman correlations between key biomarker genes and significantly differential immune cells. Results were visualized as heatmaps and bubble plots, revealing that key genes may participate in metabolic disease pathogenesis by regulating specific immune cell activity. This indicates the immune microenvironment\u0026rsquo;s regulatory role in disease pathology and provides potential immunotherapeutic targets.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e2.10. Construction of ceRNA and Regulatory Networks\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eTo explore the upstream non-coding RNA regulatory mechanism of key genes in comorbidity regulation, this study utilized comprehensive database resources to construct competitive endogenous RNA (ceRNA) regulatory networks. First, miRWalk and miRDB databases were used to screen miRNAs with experimentally validated or predicted associations with key genes, yielding multiple potential regulatory miRNAs. Subsequently, combined with the LncBase database, upstream lncRNAs of these miRNAs were predicted, and a ternary ceRNA regulatory network (lncRNA, miRNA, mRNA) was constructed.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e3.1 MASLD dataset GSE89632 vs. T2DM dataset GSE26168 differential gene analysis\u003c/p\u003e\n\u003cp\u003eDifferential expression analysis via limma (3.56.2) (P \u0026lt; 0.05, |log2FC| \u0026gt; 1) identified 3,245 differentially expressed genes (DEGs) in T2DM dataset GSE26168 (1,564 upregulated, 1,681 downregulated) and 3,063 DEGs in MASLD dataset GSE89632 (1,316 upregulated, 1,747 downregulated). Heatmaps (Fig. 1A, C) show top 20 DEGs, while volcano plots (Fig. 1B, D) visualize expression changes and significance, with key DEGs including IRS1, INSR, SOCS3, MYC, JUN, FOS, GADD45. Intersection analysis revealed 246 common DEGs (117 upregulated, 129 downregulated; Fig. 1E, F). Transcriptomically, MASLD and T2DM exhibit disease-specific differences and shared alterations, providing a suitable gene set for functional enrichment, PPI network construction, and microbial metabolite integration.\u003c/p\u003e\n\u003cp\u003e3.2. Mendelian Randomization Analysis Investigating the Causal Relationship Between Gut Microbiota and T2DM and MASLD\u003c/p\u003e\n\u003cp\u003e3.2.1. Instrumental Variable Selection\u003c/p\u003e\n\u003cp\u003eWe found Bifidobacterium genus microbiota is significantly associated with multiple strong genetic variations (SNPs), with SNP (rs182549) standing out with the highest F-value (88.43). Selected representative SNPs (e.g., rs182549, rs6754311, rs7570971) all have F-values significantly higher than 10, confirming their potential as strong IVs to reduce analytical bias. These SNPs are mainly located on chromosome 2 (chr2) and strongly associated with Bifidobacterium at different taxonomic levels (genus, family, order). We corrected SNP effect directions and excluded bias-prone loci during screening, ensuring data accuracy and laying a solid foundation for subsequent causal analysis (Supplementary Table S1,S2 for details).\u003c/p\u003e\n\u003cp\u003e3.2.2. Causal effect estimation and visualization\u003c/p\u003e\n\u003cp\u003eAfter screening out reliable instrumental variables, we use multiple MR methods to estimate causal effects and visualize them through graphs.\u003c/p\u003e\n\u003cp\u003e1.IVW forest map to analyze the effects of key microflora\u003c/p\u003e\n\u003cp\u003eUniform effect allele and effect size analysis were performed to analyze the above exposure factors separately and then based on IVW. As a result, factors with P \u0026lt; 0.05 were screened as potential related candidate exposure factors for subsequent analysis, passedThe forestploter package maps the forest and visualizes the results. The results showed that a total of 27 exposure factors and knots were obtained. Local MASLD was significantly associated with 27 exposure factors and outcome T2DM(Fig. 2A, B).\u003c/p\u003e\n\u003cp\u003e3.2.3. Sensitivity Analysis\u003c/p\u003e\n\u003cp\u003eTo ensure the robustness and reliability of the MR analysis results, we conducted a series of sensitivity analyses.\u003c/p\u003e\n\u003cp\u003e1.Heterogeneity Test\u003c/p\u003e\n\u003cp\u003eThe heterogeneity test results show that in the analysis of MASLD, certain microbial groups (such as the Melainabacteria class, Bifidobacterium genus, etc.) exhibit some heterogeneity (Cochran\u0026apos;s Q test p \u0026lt; 0.05); in the analysis of T2DM, similar heterogeneity was observed in some microbial groups. However, neither of the two analyses found significant horizontal pleiotropy interference (MR-Egger intercept test p \u0026gt; 0.05), so the relevant analyses were all estimated using the random effects inverse variance weighting method (IVW) to ensure the reliability of the results (Supplementary Tables S3, S4). In addition, the leave-one-out sensitivity analysis indicates that in the analyses of MASLD and T2DM, the combined effect size of the remaining instrumental variables on the outcome did not show significant fluctuations after the sequential exclusion of individual SNPs, indicating that the main causal association results of this study are robust and not overly affected by individual abnormal instrumental variables (Figure 3D, H).\u003c/p\u003e\n\u003cp\u003e2.Horizontal Multicollinearity Testing\u003c/p\u003e\n\u003cp\u003eThe results showed that in MASLD, the p-values of the MR-Egger intercept test for most microbial communities were greater than 0.05, indicating no significant horizontal pleiotropy; however, some communities such as Actinobacteria, Bifidobacteriaceae, and Streptococcaceae had p-values less than 0.05, suggesting potential horizontal pleiotropy bias. In the T2DM analysis, the p-values of the MR-Egger intercept test for all microbial communities were greater than 0.05, indicating no significant horizontal pleiotropy detected. Overall, most exposure factors passed the horizontal pleiotropy test, supporting the validity of causal inference, but the results for a few communities in MASLD should be interpreted with caution(Supplementary Table 5,6).\u003c/p\u003e\n\u003cp\u003e3.Sensitivity analysis: leave one method for testing\u003c/p\u003e\n\u003cp\u003eVisualized by mr_leaveoneout_plot (Fig. 3D,H), forest plots showed stable overlapping confidence intervals across iterations, with no significant heterogeneity. This indicates MR conclusions are not overly affected by single SNPs, supporting robust IV screening and highly reliable causal inference.\u003c/p\u003e\n\u003cp\u003e4.Comorbidity screening using core gut microbiota and core genes\u003c/p\u003e\n\u003cp\u003eMR analysis first underwent intersection analysis of gut microbiota strongly linked to MASLD and T2DM, revealing 4 shared potential microbiota across taxonomic levels (genus, family, order; Figure 4A). Specifically, 27 strains were associated with MASLD-linked microbiota and 27 with T2DM-linked microbiota, forming the basis for subsequent metabolite-host gene integration. Using gutMGene, SEA, and SwissTargetPrediction databases, metabolites and their host targets of these microbiota were predicted. Intersecting these consensus targets with 246 shared differentially expressed genes from GSE89632 and GSE26168 datasets identified 7 common key genes (Figure 4B). As core nodes in the \u0026quot;microbiota-metabolite-host gene\u0026quot; regulatory axis, these genes were selected as potential biomarkers for subsequent functional enrichment and network construction.\u003c/p\u003e\n\u003cp\u003e3.3. Functional enrichment study with GO/KEGG\u003c/p\u003e\n\u003cp\u003eGO enrichment analysis indicates these 7 common genes are involved in biological processes (e.g., unsaturated fatty acid/arachidonic acid metabolism, carboxylic acid biosynthesis, brown adipocyte differentiation, \u0026beta;-adrenergic receptor-mediated signaling) and molecular functions (oxidoreductase/dioxygenase activity), critical for energy/lipid metabolism and inflammatory mediator production. KEGG analysis (Figure 4D) identifies pathways including neurotransmitter regulation, lipid metabolism, VEGF signaling, and one-carbon metabolism (e.g., folate-mediated one-carbon pooling, folate biosynthesis). Gut microbiota metabolites may influence MASLD-T2DM comorbidity progression by regulating hepatic lipid metabolism, systemic inflammation, and energy homeostasis (Figure 4C).\u003c/p\u003e\n\u003cp\u003e3.4. Core Gene Modules and PPI Network Analysis\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSeven candidate genes were input into the STRING database to construct a PPI network (confidence score cutoff \u0026gt;0.4). Results showed these genes interact closely to form a highly interconnected module. Network analysis identifies PTGS2, ALOX15B, KLK3, DHFR as key nodes regulating redox, one-carbon metabolism, and inflammation (e.g., arachidonic acid/prostaglandin pathways). Critical for host responses to gut microbiota metabolites, PTGS2 and ALOX15B (linked to inflammatory signaling/lipid metabolism) are key targets for further validation (Figure 4E).\u003c/p\u003e\n\u003cp\u003e3.5. Association Network of Microbiome, Metabolites, and Biomarkers\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn the microbiome-metabolite-biomarker association network (Figure 4F), several microbial taxa (especially Bifidobacterium and its subspecies) are strongly associated with key small-molecule metabolites, including plant polyphenol derivatives (aglycone, kaempferol, genistein, equol) and amino acid derivatives (phenylacetic acid, N-acetylaspartate, leucylproline, folic acid). The network forms a \u0026quot;redundancy-multipoint\u0026quot; regulatory pattern: a single microbiome targets multiple host genes via diverse pathways, while multiple species/strains co-regulate the same metabolite, enhancing regulatory diversity and signal stability. Predictive analysis shows metabolites like equol and genistein directly/indirectly target key genes (PTGS2, ALOX15B, KLK3, DHFR), with microbes modifying fatty acid metabolism/inflammation via these bioactives to affect insulin sensitivity and liver lipid deposition. Genes like DHFR link to folate-related one-carbon metabolism, suggesting microbial regulation of host metabolic balance and epigenetics. This network highlights complex microbe-host multi-level interactions, providing potential biomarkers for precision microbial regulation and disease intervention.\u003c/p\u003e\n\u003cp\u003e3.6. SwissADME Metabolite Drug Similarity Analysis\u003c/p\u003e\n\u003cp\u003eUsing the SwissADME online platform (http://www.swissadme.ch/, access time:2025/10/09) and combining literature data, the drug-likeness of metabolites involved in biomarker-pathway network analysis is evaluated. Since metabolites typically have strong hydrophilicity and low bioavailability, this study analyzes their physicochemical properties (Supplementary Table S7) through computational simulation methods and sets the following screening thresholds to determine candidate core metabolites: hydrogen bond acceptor (HBA) \u0026lt; 10, hydrogen bond donor (HBD) \u0026le; 5, octanol-water partition coefficient (MLogP) \u0026le; 4.15, topological polar surface area (TPSA) \u0026lt; 140 \u0026Aring;\u0026sup2;, and violation of the Lipinski rule \u0026le; 1. According to the above criteria, \u0026quot;Folic Acid\u0026quot; was excluded as it did not meet the conditions among the total of 9 metabolites.\u003c/p\u003e\n\u003cp\u003e3.7. ADMETox 3.0 Toxicological Property Assessment\u003c/p\u003e\n\u003cp\u003eThe physicochemical properties of metabolites are insufficient to fully assess their safety, and toxicological properties are crucial in the safety evaluation. Drug development failures are often attributed to adverse reactions, such as hERG block-induced cardiotoxicity, human liver toxicity (H-HT), respiratory toxicity, acute oral toxicity in rats, carcinogenicity, drug-induced liver injury (DILI), skin sensitization, and acute toxicity (LD50), among others. Therefore, this study employed the ADMETox 3.0 online platform (https://admetmesh.scbdd.com/,access time:2025/10/09) to predict the toxicological properties of candidate core metabolites. The screening criteria were that all four indicators\u0026mdash;hERG block, human liver toxicity (H-HT), acute oral toxicity in rats, and carcinogenicity\u0026mdash;were negative. The results showed that Genistein, N-acetylornithine, and another metabolite met all safety standards and were identified as key metabolites(Supplementary Table S8).\u003c/p\u003e\n\u003cp\u003e3.8. Core Metabolites and Biomarker Molecular Docking\u003c/p\u003e\n\u003cp\u003eBased on the \u0026quot;microbiota\u0026mdash;metabolite\u0026mdash;biomarker\u0026quot; association network, Genistein/Phenylacetic Acid (targeting PTGS2) and N-acetylornithine (targeting DHFR) were selected as docking pairs. All three ligand-receptor complexes showed negative binding energies (Table 1), indicating spontaneous and stable binding: PTGS2-Genistein (-7.9 kcal/mol, highest affinity), PTGS2-Phenylacetic Acid (-5.8 kcal/mol), and DHFR-N-acetylornithine (-5.0 kcal/mol).Figure 5 details their conformations: PTGS2-Genistein forms stable bonds via hydrogen bonds and hydrophobic interactions at Thr-145/Val-349 (Figure 5A); PTGS2-Phenylacetic Acid binds at Leu-352/His-527 (multi-point binding, Figure 5B); DHFR-N-acetylornithine relies on ALA-523/MET-518 for stability (Figure 5C). These results confirm effective binding of core metabolites to biomarker proteins, suggesting they may directly regulate these proteins and participate in the pathogenesis of type 2 diabetes and stroke comorbidity, with potential for drug development.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eMolecular docking of core metabolites and biomarkers\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"673\" class=\"fr-table-selection-hover\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCompound name\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTarget Protein\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMode\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBinding Energy (kcal/mol)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSD (l.b.)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSD (u.b.)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0.876\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e1.18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.954\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.611\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.693\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.142\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.061\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.555\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.803\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e3.868\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e4.581\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e7.051\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.957\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eN-acetylornithine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003eDHFR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-4.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.845\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.299\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-7.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e2.324\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-6.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e19.582\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.764\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-6.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e6.739\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-6.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e19.615\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.653\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-6.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e18.723\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.016\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e19.559\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.672\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e19.658\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.701\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003eGenistein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e20.121\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e22.076\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.067\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e1.413\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e1.587\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e1.663\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.602\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.092\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e2.701\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.094\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e20.275\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e21.56\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e1.935\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e2.811\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003ePhenylacetic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\n \u003cp\u003ePTGS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 50px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e-5.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e3.228\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e4.481\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e3.9. Building nomograms, assessing model performance, and analyzing the immunological microenvironment\u003c/p\u003e\n\u003cp\u003e3.9.1. Risk prediction and nomogram building\u003c/p\u003e\n\u003cp\u003eIn order to develop a prediction model that can rapidly and independently evaluate each patient\u0026apos;s risk of comorbidity between MASLD and T2DM, we created a nomogram based on four important biomarkers (ALOX15B, PTGS2, KLK3, and DHFR). Each biomarker\u0026apos;s expression level is converted by the nomogram into a matching score, which is then totaled together to determine the patient\u0026apos;s overall risk of developing T2DM and MASLD. ALOX15B gene have comparable points of around 60 in the MASLD dataset (GSE89632), whereas PTGS2, KLK3, and DHFR genes have corresponding points of roughly 12, 8.5, and 13 points, respectively. The cumulative total score is 235 points. The patient is at high risk for MASLD, as indicated by the estimated risk of 0.835 based on this total score (Figure 6A). The nomogram construction results showed that the ALOX15B gene score for the T2DM dataset (GSE26168) was low, at about 5 points. The integrals of other genes are likewise mapped in accordance with the KLK3 gene score, which is greater at around 70 points. A moderate risk level was indicated by the cumulative total score of 123 points and the prediction risk of type 2 diabetes of 0.15 (Figure 6E).\u003c/p\u003e\n\u003cp\u003e3.9.2. Evaluate the predictive model\u0026apos;s effectiveness\u003c/p\u003e\n\u003cp\u003e1.Receiver operating characteristic (ROC) curve analysis: We evaluated the nomogram model\u0026rsquo;s predictive performance by plotting ROC curves and calculating the area under the curve (AUC), a key discriminative metric. The model showed excellent discrimination in the MASLD dataset (GSE89632) with an AUC of 0.982 (95% CI: 0.947-1.000) (Figure 6B) and high accuracy in the T2DM dataset (GSE26168) with an AUC of 0.903 (95% CI: 0.855-0.951) (Figure 6F).\u003c/p\u003e\n\u003cp\u003e2.Decision curve analysis (DCA): To assess clinical utility, DCA showed the model yielded a net benefit over \u0026quot;treat all\u0026quot; and \u0026quot;treat none\u0026quot; strategies across most clinically acceptable high-risk thresholds in both datasets (Figure 6C, G). Benefits were more pronounced in low-to-medium thresholds, enabling precise identification of patients requiring intervention while avoiding overtreatment of low-risk groups.\u003c/p\u003e\n\u003cp\u003e3.Hosmer-Lemeshow test: Results showed p-values \u0026gt;0.05 in both datasets, indicating the nomogram had good goodness-of-fit and accurately predicted MASLD-T2DM comorbidity risk. Its robust fit supports reliable clinical risk assessment (Figure 6D, H).\u003c/p\u003e\n\u003cp\u003e3.9.3. We analyzed the relationship between important genes and immune cell infiltration\u003c/p\u003e\n\u003cp\u003e1.Immune cell infiltration analysis: Using immune cell infiltration analysis, we investigated the relative abundance of immune cells in T2DM (GSE26168) and MASLD (GSE89632) patient tissues. In MASLD patients (Figure 6I), regulatory T cells, monocytes, myeloid suppressor cells (MDSCs), and neutrophils showed significantly higher infiltration than the control group (p \u0026lt; 0.05), highlighting the role of immunomodulation and inflammation in MASLD pathophysiology; natural killer T cells (NKT cells) trended upward. In T2DM patients (Figure 6K), monocytes, neutrophils, and regulatory T cells also had significantly higher infiltration, indicating a close link between immune activation and T2DM progression.\u003c/p\u003e\n\u003cp\u003e2.Gene-immune cell correlation analysis: Spearman correlation heatmaps showed PTGS2 in MASLD samples (Figure 6J) had a significant positive correlation (p\u0026lt;0.05) with regulatory T cells, monocytes, and MDSCs (potential immunosuppressive regulation); KLK3 correlated negatively with natural killer cells and effector memory CD8+ T cells, while ALOX15B negatively correlated with multiple immune cells (anti-inflammatory potential). In T2DM samples (Figure 6L), nearly all core genes (PTGS2, KLK3, DHFR, ALOX15B) significantly correlated with immune cells (e.g., regulatory T cells, neutrophils); KLK3\u0026rsquo;s negative correlations with multiple immune cells suggest potential inhibition of diabetes-related immune inflammation.\u003c/p\u003e\n\u003cp\u003e3.ceRNA regulatory network construction: By integrating miRwalk and miRTarBase databases, we constructed a network including 6 key miRNAs (hsa-miR-2278, hsa-miR-588, hsa-miR-676-3p, hsa-miR-654-5p, hsa-miR-543, hsa-miR-520a-5p) and 4 core genes (KLK3, PTGS2, DHFR, ALOX15B) (Figure 6M), providing novel insights into the molecular regulatory mechanisms of T2DM-MASLD comorbidity.\u0026nbsp;\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec39\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Overview of the research findings and their clinical significance\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study systematically reveals the MASLD-T2DM comorbidity mechanism via integrating transcriptomics, gut microbiome GWAS, and Mendelian randomization (MR) analysis, highlighting the gut microbiota and its metabolites as core regulators of host gene expression.\u003c/p\u003e \u003cp\u003eMR analysis filtered 14 gut microbiota linked to both diseases, with genus.Bifidobacterium.id.436 as the core candidate exerting disease-specific causal effects: for MASLD, OR\u0026thinsp;=\u0026thinsp;0.78 (95% CI: 0.70384\u0026ndash;0.86439, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), suggesting increased abundance reduces MASLD risk (e.g., Bifidobacterium pseudolongum prevents NAFLD-HCC progression in mouse models; for T2DM, OR\u0026thinsp;=\u0026thinsp;1.00055 (95% CI: 1.00003-1.00108, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), indicating it promotes T2DM as a risk factor. This dual effect was confirmed by IVW and robustified via sensitivity analysis, potentially related to T2DM-related gut dysbiosis (e.g., increased Desulfovibrio) and metabolic disorders (e.g., insulin resistance)(\u003cspan additionalcitationids=\"CR18 CR19\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e).The gutMGene database was used to integrate microbiota-related metabolites, and finally 7 core biomarkers (PTGS2, ALOX15B, KLK3, DHFR, etc.) were identified, which were involved in lipid metabolism(\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e), inflammation regulation(\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e), and one-carbon metabolism(\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e).Molecular docking showed that metabolites such as Genistein, Phenylacetic Acid, and N-acetylornithine could bind to key targetswith \u0026le; -5 kcal/mol, suggesting potential molecular regulatory mechanisms. The nomogram model based on these markers had AUC values of 0.982 and 0.903 in the two datasets, respectively, and the DCA confirmed its clinical net benefit, and the Hosmer-Lemeshow test (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05) showed a good fit. Immune infiltration analysis revealed that key genes were significantly related to immune cells such as regulatory T cells (Tregs), monocytes, and neutrophils, and ceRNA networks further elucidated the multi-level regulation mediated by miRNAs (e.g., hsa-miR-2278 and hsa-miR-588).\u003c/p\u003e \u003cp\u003eThe clinical significance of these findings lies in first, providing biomarkers and predictive tools for the comorbidity of MASLD and T2DM, which helps in early screening of high-risk populations and optimizes resource allocation. Second, revealing the dual role of the Bifidobacterium genus challenges the traditional view of \"probiotics universally protect\" and guides personalized microbial intervention strategies. Lastly, these findings promote the shift from association to causation in precision medicine, thereby reducing the global burden of metabolic diseases. According to WHO data, the risk of cardiovascular events in patients with comorbidity of MASLD and T2DM increases by 2\u0026ndash;3 times, and this study's model can improve the screening accuracy to over 90%, which has significant public health value(\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec40\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Interpretation of Mendelian Randomization Analysis Results: Bifidobacterium genus' dual causal effect\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe primary novelty of this work lies in Mendelian randomization (MR) analysis, which avoids confounding bias in observational studies by simulating randomized controlled trials using genetic variations as instrumental variables (IVs). To ensure IV strength, we selected SNPs with F statistics\u0026thinsp;\u0026gt;\u0026thinsp;10 (e.g., rs182549, rs6754311) from 211 gut microbiome clusters derived from the MiBioGen consortium. Forest plots identified 4 core microbiomes (including Streptococcus and Bifidobacterium) at the intersection of 27 MASLD-associated and 27 T2DM-associated microbiomes. Bifidobacterium improves metabolic function via short-chain fatty acids (SCFAs, e.g., butyrate); studies show MASLD-HCC patients have significantly higher SCFA levels (acetic acid, propionic acid, butyrate), and increased Bifidobacterium longum/adolescentis correlates with reduced HbA1c and basal insulin demand(\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). These metabolites activate GPR43/41 receptors to promote hepatic lipid oxidation and inhibit fat synthesis, alleviating MASLD liver fat deposition. Scatter/funnel plots confirm no significant horizontal pleiotropy (MR-Egger intercept P\u0026thinsp;\u0026gt;\u0026thinsp;0.05) and low heterogeneity (Cochran's Q P\u0026thinsp;\u0026gt;\u0026thinsp;0.05), supporting causal inference.\u003c/p\u003e \u003cp\u003eConversely, Bifidobacterium exerts considerable detrimental effects in T2DM (OR\u0026thinsp;\u0026gt;\u0026thinsp;1) via three mechanisms: 1) Inflammatory/metabolic interference: Overfermentation produces lactic acid/hydrogen, disrupting intestinal barrier integrity and inducing systemic low-grade inflammation; metabolites like N-acetylornithine inhibit GLUT4-mediated glucose uptake, while phenylacetic acid activates the TLR4 pathway to exacerbate insulin resistance; 2) Masked antioxidant protection: Bifidobacterium reduces liver oxidative stress via regulating PTGS2 in MASLD, but this effect is masked in T2DM\u0026rsquo;s systemic metabolic disorder; 3) Host environment-dependent effects: Decreased intestinal pH in T2DM promotes Bifidobacterium proliferation, with its metabolites\u0026rsquo; negative impacts dominating\u0026mdash;unlike MASLD (predominantly local liver inflammation), where antioxidant effects prevail. This underscores the importance of stratified analysis: European cohort studies failed to distinguish MASLD subtypes, potentially overlooking microbiome disease-specific effects(\u003cspan additionalcitationids=\"CR31 CR32\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e). Strain specificity, host baseline metabolic status (e.g., intestinal pH differences), and intervention duration may explain result discrepancies(\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOther intersecting bacterial communities (e.g., Streptococcus) consistently act as risk factors (OR\u0026thinsp;\u0026gt;\u0026thinsp;1). As commensal pathogens, their increased abundance exacerbates metabolic diseases via: 1) Producing endotoxins (e.g., lipopolysaccharides, LPS) that activate the TLR4/NF-κB pathway, inducing hepatic/peripheral low-grade inflammation and insulin resistance(\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e); 2) Increasing intestinal permeability, disrupting mucosal integrity, and facilitating translocation of endotoxins/metabolites (e.g., phenylacetic acid) to the portal circulation, aggravating systemic metabolic disorders and liver fat deposition(\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e); 3) Producing metabolites like trimethylamine N-oxide (TMAO) that directly inhibit insulin signaling (e.g., PI3K/Akt) and increase comorbidity risks (e.g., atherosclerosis)(\u003cspan additionalcitationids=\"CR39\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGut microbiota (e.g., Bifidobacterium, Streptococcus) regulates host metabolic gene expression (e.g., PTGS2, TLR4) via metabolites (lactic acid, LPS), while host metabolic status (e.g., hyperglycemia, low pH) feedback-regulates microbiota composition, forming a positive feedback loop. Stratified MR analysis confirms microbiota effects differ between T2DM patients with/without MASLD. Environmental factors (antibiotic abuse, high-fat diet) and genetic variations (e.g., PNPLA3) synergistically amplify metabolic risks (e.g., MASLD progression to liver fibrosis) by altering microbiota structure(7, 41).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec41\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Molecular mechanism of the gut microbiota-metabolite-host gene regulatory axis\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study\u0026rsquo;s gut-metabolite-biomarker network reveals a multi-level regulatory mechanism.GutMGene/SwissTargetPrediction predictions indicate Bifidobacterium produces metabolites (Genistein, Phenylacetic Acid, N-acetylornithine) that intersect with DEGs to identify 7 core genes. GO enrichment highlights arachidonic acid metabolism (P\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and oxidoreductase activity, with KEGG pathways involving VEGF signaling and folate biosynthesis (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), reflecting integrated regulation of lipid-inflammatory-energy homeostasis.\u003c/p\u003e \u003cp\u003eMetabolites exert disease-specific effects: Genistein binds PTGS2 (Thr-145/Val-349, binding energy\u0026thinsp;\u0026minus;\u0026thinsp;7.9 kcal/mol) to inhibit PGE2, alleviating MASLD liver fibrosis; Phenylacetic Acid activates COX-2/JNK to promote T2DM β-cell apoptosis(\u003cspan additionalcitationids=\"CR43\" citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e)ALOX15B downregulation protects MASLD livers but disrupts T2DM insulin signaling(\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e); N-acetylornithine binds DHFR (-5.0 kcal/mol) to regulate folate cycle, inhibiting T2DM insulin signaling (SREBP-1c pathway) and promoting MASLD liver regeneration(\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSwissADME/ADMETlab confirm drug-likeness: Genistein has \u0026gt;\u0026thinsp;30% bioavailability, no hERG/DILI toxicity, and complies with Lipinski\u0026rsquo;s rule (HBA\u0026thinsp;=\u0026thinsp;5, TPSA\u0026thinsp;=\u0026thinsp;90.9 \u0026Aring;\u0026sup2;). Network analysis shows a \"many-to-many\" pattern (e.g., Bifidobacterium\u0026thinsp;+\u0026thinsp;Streptococcus synergy)(\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe ceRNA network involves 6 miRNAs (e.g., hsa-miR-543 targeting PTGS2); hsa-miR-676-3p upregulates in high-glucose environments to enhance T2DM inflammation but downregulates ALOX15B to protect MASLD livers, consistent with the MALAT1-miR-200c-PTGS2 axis(\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec42\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Regulatory mechanisms of the immune microenvironment and comorbidities mediated by inflammation\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003essGSEA immunosuppression analysis identifies chronic inflammation as the key driver of MASLD-T2DM comorbidity: T2DM shows increased neutrophils/NKT cells, while MASLD has elevated Treg/monocytes/MDSC infiltration (P\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Spearman correlation links Treg to PTGS2, mediating hepatic/pancreatic fibrosis and immune suppression via PGE2(\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e). Bifidobacterium-derived SCFAs alleviate MASLD inflammation by activating Foxp3\u0026thinsp;+\u0026thinsp;Treg but induce T2DM insulin resistance via excessive neutrophil recruitment(\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e).Dysbiosis increases intestinal permeability; LPS leakage activates the TLR4/NF-κB pathway, amplifying PTGS2 (COX-2)/ALOX15B signaling and chronic inflammation(\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e).ALOX15B downregulation protects MASLD livers but accelerates T2DM pancreatic inflammation. SCFAs improve MASLD metabolism by inhibiting HDAC yet aggravate T2DM β-cell injury via neutrophil ROS activation(\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec43\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Comparison with Existing Research\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eCurrent research often focuses on specific disease molecular mechanisms. Petta et al.\u0026rsquo;s meta-analysis identified over 500 MASLD DEGs but failed to incorporate T2DM comorbid characteristics, leaving shared pathways unexposed(\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e). Huang et al. validated Akkermansia\u0026rsquo;s preventive effect on T2DM via Mendelian randomization (OR\u0026thinsp;=\u0026thinsp;0.88)(\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e). but did not clarify Bifidobacterium\u0026rsquo;s disease-specific benefit. Studies show Bifidobacterium metabolites (phenylacetic acid, genistein) exert a \"protection-injury\" double-edged sword effect by targeting PTGS2/ALOX15B, with action direction dependent on disease microenvironments (e.g., high-fat vs. high-sugar), which aligns well with this study\u0026rsquo;s findings(\u003cspan additionalcitationids=\"CR57\" citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec44\" class=\"Section2\"\u003e \u003ch2\u003e4.6. The Study's Limitations\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eDespite improved reliability from multi-omics integration, drawbacks remain: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) Data sources: MiBioGen focuses on European populations, with Asian bias limiting generalizability; GEO datasets have small sample sizes requiring larger confirmatory studies; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) MR restrictions: Horizontal pleiotropy is notable in MASLD (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05); (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) Mechanism inference: Molecular docking ignores dynamic conformations (relying on crystal structures); ceRNA predictions are mostly bioinformatic, lacking ChIP/luciferase validation; (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) Clinical translation: In silico toxicity evaluation needs in vitro/in vivo validation; nomograms exclude clinical factors (BMI/HbA1c); (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) Causal direction: Longitudinal cohorts are required to confirm Bifidobacterium\u0026rsquo;s dual role(\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec45\" class=\"Section2\"\u003e \u003ch2\u003e4.7. Future Research Directions and Clinical Application Prospects\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFuture studies can leverage single-cell transcriptomics to explore how cellular heterogeneity modulates MASLD-T2DM comorbidity, while investigating the gut microbiota\u0026ndash;metabolite axis\u0026mdash;particularly the role of Bifidobacterium-derived Genistein in MASLD models\u0026mdash;and developing CRISPR-based gene editing tools to generate optimized strains. In high-fat diet-fed mouse models, fecal microbiota transplantation can be employed to evaluate the impacts of wild-type and gene-deletion strains on hepatic and pancreatic histological changes as well as glucose tolerance. To verify the therapeutic efficacy of engineered probiotics (e.g., genistein-expressing strains) in patients with MASLD and T2DM, phase II randomized controlled trials (RCTs) are recommended, with efficacy assessments enhanced by integrating FibroScan imaging and microbiome data. By refining nomogram models through artificial intelligence and incorporating multimodal data (such as metabolomics and medical imaging), the goal is to achieve a prediction accuracy exceeding 95%. In epigenetics, miRNA knockdown assays should be performed to validate the functional significance of the ceRNA regulatory axis, alongside exploring the therapeutic potential of epigenetic drugs like histone deacetylase (HDAC) inhibitors(\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e). Overall, future microbiome therapies will shift from \"broad-spectrum probiotics\" to \"precision microbiota intervention,\" which is expected to substantially alleviate the global disease burden of T2DM and MASLD\u0026mdash;projected to affect over one billion patients by 2030(\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e). In summary, this study systematically clarifies the gut microbiota-driven comorbidity mechanism of MASLD-T2DM, identifies core biomarkers, therapeutic targets and predictive models, and lays a robust scientific foundation for precision medicine interventions.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003eAuthor Contributions: Conceptualization, Y.Z. and Y.C.; methodology, W.M.; software, W.Y.; validation, Y.Z., W.M. and Y.C.; formal analysis, Z.L.; investigation, L.L.; resources, L.L.; data curation, Y.Z.; writing\u0026mdash;original draft preparation, Y.Z.; writing\u0026mdash;review and editing, L.L.; visualization, Z.L.; supervision, Z.Y.; project administration, L.L; funding acquisition, L.L. All authors have read and agreed to the published version of the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFunding: No funding was received for this work.\u003c/p\u003e\n\u003cp\u003eInstitutional Review Board Statement: Not applicable.\u003c/p\u003e\n\u003cp\u003eInformed Consent Statement: Not applicable.\u003c/p\u003e\n\u003cp\u003eData Availability Statement: The exposure data in the Mendelian randomization study were obtained from the MiBioGen consortium (https://mibiogen.gcc.rug.nl/, download date: 2025/09/30), and the outcome data were sourced from the IEU OpenGWAS database (https://gwas.mrcieu.ac.uk/, download date: 2025/09/30), including ebi-a-GCST90091033 (a total of 778,614 samples: 8,434 cases of MASLD and 770,180 controls) and ukb-b-13806 (2,972 cases of T2DM and 459,961 controls). The Gene Expression Omnibus provided two datasets: dataset GSEGSE26168 and GSE89632 (https://www.ncbi.nlm.nih.gov/geo/, download date: 2025/09/30).\u003c/p\u003e\n\u003cp\u003eConflicts of Interest: The authors declare no conflicts of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMa C, Wang S, Dong B, Tian Y. Metabolic Reprogramming of Immune Cells in Mash. \u003cem\u003eHepatology\u003c/em\u003e (2025). Epub 20250505. doi: 10.1097/hep.0000000000001371.\u003c/li\u003e\n\u003cli\u003eHuang DQ, Wong VWS, Rinella ME, Boursier J, Lazarus JV, Yki-Jarvinen H, et al. Metabolic Dysfunction-Associated Steatotic Liver Disease in Adults. \u003cem\u003eNat Rev Dis Primers\u003c/em\u003e (2025) 11(1):14. Epub 20250306. doi: 10.1038/s41572-025-00599-1.\u003c/li\u003e\n\u003cli\u003eYounossi ZM, Kalligeros M, Henry L. Epidemiology of Metabolic Dysfunction-Associated Steatotic Liver Disease. \u003cem\u003eClin Mol Hepatol\u003c/em\u003e (2025) 31(Suppl):S32-S50. Epub 2024/08/20. doi: 10.3350/cmh.2024.0431.\u003c/li\u003e\n\u003cli\u003eSun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. Idf Diabetes Atlas: Global, Regional and Country-Level Diabetes Prevalence Estimates for 2021 and Projections for 2045. \u003cem\u003eDiabetes Res Clin Pract\u003c/em\u003e (2022) 183:109119. Epub 20211206. doi: 10.1016/j.diabres.2021.109119.\u003c/li\u003e\n\u003cli\u003eWu JY, Chang HY, Tsung Y, Lin YM. Clinical Outcomes of Sglt2 Inhibitors among Patients with Masld and T2dm. \u003cem\u003eDiabetes Res Clin Pract\u003c/em\u003e (2025) 229:112918. Epub 2025/09/27. doi: 10.1016/j.diabres.2025.112918.\u003c/li\u003e\n\u003cli\u003eCusi K, Abdelmalek MF, Apovian CM, Balapattabi K, Bannuru RR, Barb D, et al. Metabolic Dysfunction-Associated Steatotic Liver Disease (Masld) in People with Diabetes: The Need for Screening and Early Intervention. A Consensus Report of the American Diabetes Association. \u003cem\u003eDiabetes Care\u003c/em\u003e (2025) 48(7):1057-82. Epub 2025/05/28. doi: 10.2337/dci24-0094.\u003c/li\u003e\n\u003cli\u003eHan E, Han KD, Lee YH, Kim KS, Hong S, Park JH, et al. Association of Temporal Masld with Type 2 Diabetes, Cardiovascular Disease and Mortality. \u003cem\u003eCardiovasc Diabetol\u003c/em\u003e (2025) 24(1):289. Epub 2025/07/16. doi: 10.1186/s12933-025-02824-3.\u003c/li\u003e\n\u003cli\u003eKang M, Song J, Kang ES, Jang S, Kwak T, Kim Y, et al. Pathophysiology, Development, and Mortality of Major Non-Communicable Diseases in Metabolic Dysfunction-Associated Steatotic Liver Disease: A Comprehensive Review. \u003cem\u003eInt J Biol Sci\u003c/em\u003e (2025) 21(13):5691-703. Epub 20250903. doi: 10.7150/ijbs.117211.\u003c/li\u003e\n\u003cli\u003eKim DY, Lee SY, Lee JY, Whon TW, Lee JY, Jeon CO, et al. Gut Microbiome Therapy: Fecal Microbiota Transplantation Vs Live Biotherapeutic Products. \u003cem\u003eGut Microbes\u003c/em\u003e (2024) 16(1):2412376. Epub 20241008. doi: 10.1080/19490976.2024.2412376.\u003c/li\u003e\n\u003cli\u003eSchwenger KJP, Copeland JK, Ghorbani Y, Chen L, Comelli EM, Guttman DS, et al. Characterization of Liver, Adipose, and Fecal Microbiome in Obese Patients with Masld: Links with Disease Severity and Metabolic Dysfunction Parameters. \u003cem\u003e#N/A\u003c/em\u003e (2025) 13(1):9. Epub 2025/01/15. doi: 10.1186/s40168-024-02004-7.\u003c/li\u003e\n\u003cli\u003eAlam N, Jia L, Cheng A, Ren H, Fu Y, Ding X, et al. Global Research Trends on Gut Microbiota and Metabolic Dysfunction-Associated Steatohepatitis: Insights from Bibliometric and Scientometric Analysis. \u003cem\u003eFront Pharmacol\u003c/em\u003e (2024) 15:1390483. Epub 2024/07/29. doi: 10.3389/fphar.2024.1390483.\u003c/li\u003e\n\u003cli\u003eGarcia-Mateo S, Rondinella D, Ponziani FR, Miele L, Gasbarrini A, Cammarota G, et al. Gut Microbiome and Metabolic Dysfunction-Associated Steatotic Liver Disease: Pathogenic Role and Potential for Therapeutics. \u003cem\u003eBest Pract Res Clin Gastroenterol\u003c/em\u003e (2024) 72:101924. Epub 2024/12/08. doi: 10.1016/j.bpg.2024.101924.\u003c/li\u003e\n\u003cli\u003eDaniel N, Farinella R, Chatziioannou AC, Jenab M, May\u0026eacute;n AL, Rizzato C, et al. Genetically Predicted Gut Bacteria, Circulating Bacteria-Associated Metabolites and Pancreatic Ductal Adenocarcinoma: A Mendelian Randomisation Study. \u003cem\u003eSci Rep\u003c/em\u003e (2024) 14(1):25144. Epub 20241024. doi: 10.1038/s41598-024-77431-5.\u003c/li\u003e\n\u003cli\u003eLiu T, Cao Y, Liang N, Ma X, Fang JA, Zhang X. Investigating the Causal Association between Gut Microbiota and Type 2 Diabetes: A Meta-Analysis and Mendelian Randomization. \u003cem\u003eFront Public Health\u003c/em\u003e (2024) 12:1342313. Epub 2024/07/04. doi: 10.3389/fpubh.2024.1342313.\u003c/li\u003e\n\u003cli\u003eBurgess S, Woolf B, Mason AM, Ala-Korpela M, Gill D. Addressing the Credibility Crisis in Mendelian Randomization. \u003cem\u003eBMC Med\u003c/em\u003e (2024) 22(1):374. Epub 20240911. doi: 10.1186/s12916-024-03607-5.\u003c/li\u003e\n\u003cli\u003eZhang Y, Fan J. Impact of Gut Microbiota on Metabolic Syndrome and Its Comprising Traits: A Two-Sample Mendelian Randomization Study. \u003cem\u003eDiabetol Metab Syndr\u003c/em\u003e (2024) 16(1):279. Epub 2024/11/23. doi: 10.1186/s13098-024-01520-8.\u003c/li\u003e\n\u003cli\u003eLin YC, Lin HF, Wu CC, Chen CL, Ni YH. Pathogenic Effects of Desulfovibrio in the Gut on Fatty Liver in Diet-Induced Obese Mice and Children with Obesity. \u003cem\u003eJ Gastroenterol\u003c/em\u003e (2022) 57(11):913-25. Epub 2022/08/18. doi: 10.1007/s00535-022-01909-0.\u003c/li\u003e\n\u003cli\u003eNychas E, Marfil-Sanchez A, Chen X, Mirhakkak M, Li H, Jia W, et al. Discovery of Robust and Highly Specific Microbiome Signatures of Non-Alcoholic Fatty Liver Disease. \u003cem\u003e#N/A\u003c/em\u003e (2025) 13(1):10. Epub 2025/01/15. doi: 10.1186/s40168-024-01990-y.\u003c/li\u003e\n\u003cli\u003eKim G, Yoon Y, Park JH, Park JW, Noh MG, Kim H, et al. Bifidobacterial Carbohydrate/Nucleoside Metabolism Enhances Oxidative Phosphorylation in White Adipose Tissue to Protect against Diet-Induced Obesity. \u003cem\u003e#N/A\u003c/em\u003e (2022) 10(1):188. Epub 2022/11/06. doi: 10.1186/s40168-022-01374-0.\u003c/li\u003e\n\u003cli\u003eSong Q, Zhang X, Liu W, Wei H, Liang W, Zhou Y, et al. Bifidobacterium Pseudolongum-Generated Acetate Suppresses Non-Alcoholic Fatty Liver Disease-Associated Hepatocellular Carcinoma. \u003cem\u003eJ Hepatol\u003c/em\u003e (2023) 79(6):1352-65. Epub 2023/07/18. doi: 10.1016/j.jhep.2023.07.005.\u003c/li\u003e\n\u003cli\u003eBenatzy Y, Palmer MA, L\u0026uuml;tjohann D, Ohno RI, Kampschulte N, Schebb NH, et al. Alox15b Controls Macrophage Cholesterol Homeostasis Via Lipid Peroxidation, Erk1/2 and Srebp2. \u003cem\u003eRedox Biol\u003c/em\u003e (2024) 72:103149. Epub 20240403. doi: 10.1016/j.redox.2024.103149.\u003c/li\u003e\n\u003cli\u003eSimon LS. Role and Regulation of Cyclooxygenase-2 During Inflammation. \u003cem\u003eAm J Med\u003c/em\u003e (1999) 106(5b):37s-42s. doi: 10.1016/s0002-9343(99)00115-1.\u003c/li\u003e\n\u003cli\u003eSehrawat R, Rathee P, Khatkar S, Akkol E, Khayatkashani M, Nabavi SM, et al. Dihydrofolatereductase (Dhfr) Inhibitors: A Comprehensive Review. \u003cem\u003eCurr Med Chem\u003c/em\u003e (2023). Epub 20230310. doi: 10.2174/0929867330666230310091510.\u003c/li\u003e\n\u003cli\u003eGao C, Zhang L, Guo X, Lin X, Yang J, Wang Z, et al. 20s-O-Glc-Dm Regulates Fatty Acid Metabolism and Mitochondrial Function in the Treatment of Diabetes Mellitus-Associated Mafld. \u003cem\u003ePhytomedicine\u003c/em\u003e (2025) 146:157136. Epub 2025/08/18. doi: 10.1016/j.phymed.2025.157136.\u003c/li\u003e\n\u003cli\u003eTian R, Li Y. Exploring the Pathogenesis of Mafld from an Immunological Perspective: From the Perspective of the Cgas/Sting/Nf-Kappab Signaling Pathway. \u003cem\u003eFront Immunol\u003c/em\u003e (2025) 16:1674018. Epub 2025/10/15. doi: 10.3389/fimmu.2025.1674018.\u003c/li\u003e\n\u003cli\u003eZhao J, Zhao Y, Hu Y, Peng J. Targeting the Gpr119/Incretin Axis: A Promising New Therapy for Metabolic-Associated Fatty Liver Disease. \u003cem\u003eCell Mol Biol Lett\u003c/em\u003e (2021) 26(1):32. Epub 2021/07/09. doi: 10.1186/s11658-021-00276-7.\u003c/li\u003e\n\u003cli\u003eBell KJ, Saad S, Tillett BJ, McGuire HM, Bordbar S, Yap YA, et al. Metabolite-Based Dietary Supplementation in Human Type 1 Diabetes Is Associated with Microbiota and Immune Modulation. \u003cem\u003e#N/A\u003c/em\u003e (2022) 10(1):9. Epub 2022/01/21. doi: 10.1186/s40168-021-01193-9.\u003c/li\u003e\n\u003cli\u003eLee PC, Wu CJ, Hung YW, Lee CJ, Mon HC, Chi CT, et al. Distinct Gut Microbiota but Common Metabolomic Signatures between Viral and Masld Hcc Contribute to Outcomes of Combination Immunotherapy. \u003cem\u003eHepatology\u003c/em\u003e (2025). Epub 2025/07/01. doi: 10.1097/HEP.0000000000001446.\u003c/li\u003e\n\u003cli\u003eTang J, Wei Y, Pi C, Zheng W, Zuo Y, Shi P, et al. The Therapeutic Value of Bifidobacteria in Cardiovascular Disease. \u003cem\u003eNPJ Biofilms Microbiomes\u003c/em\u003e (2023) 9(1):82. Epub 2023/10/31. doi: 10.1038/s41522-023-00448-7.\u003c/li\u003e\n\u003cli\u003eShen X, Ma C, Yang Y, Liu X, Wang B, Wang Y, et al. The Role and Mechanism of Probiotics Supplementation in Blood Glucose Regulation: A Review. \u003cem\u003eFoods\u003c/em\u003e (2024) 13(17). Epub 2024/09/14. doi: 10.3390/foods13172719.\u003c/li\u003e\n\u003cli\u003eFliegerova KO, Mahayri TM, Sechovcova H, Mekadim C, Mrazek J, Jarosikova R, et al. Diabetes and Gut Microbiome. \u003cem\u003eFront Microbiol\u003c/em\u003e (2024) 15:1451054. Epub 2025/01/22. doi: 10.3389/fmicb.2024.1451054.\u003c/li\u003e\n\u003cli\u003eAyesha IE, Monson NR, Klair N, Patel U, Saxena A, Patel D, et al. Probiotics and Their Role in the Management of Type 2 Diabetes Mellitus (Short-Term Versus Long-Term Effect): A Systematic Review and Meta-Analysis. \u003cem\u003eCureus\u003c/em\u003e (2023) 15(10):e46741. Epub 2023/11/29. doi: 10.7759/cureus.46741.\u003c/li\u003e\n\u003cli\u003eQian X, Si Q, Lin G, Zhu M, Lu J, Zhang H, et al. Bifidobacterium Adolescentis Is Effective in Relieving Type 2 Diabetes and May Be Related to Its Dominant Core Genome and Gut Microbiota Modulation Capacity. \u003cem\u003e#N/A\u003c/em\u003e (2022) 14(12). Epub 2022/06/25. doi: 10.3390/nu14122479.\u003c/li\u003e\n\u003cli\u003eGuo W, Xiang Q, Mao B, Tang X, Cui S, Li X, et al. Protective Effects of Microbiome-Derived Inosine on Lipopolysaccharide-Induced Acute Liver Damage and Inflammation in Mice Via Mediating the Tlr4/Nf-Kappab Pathway. \u003cem\u003eJ Agric Food Chem\u003c/em\u003e (2021) 69(27):7619-28. Epub 20210622. doi: 10.1021/acs.jafc.1c01781.\u003c/li\u003e\n\u003cli\u003eWang X, Sun Z, Wang X, Li M, Zhou B, Zhang X. Solanum Nigrum L. Berries Extract Ameliorated the Alcoholic Liver Injury by Regulating Gut Microbiota, Lipid Metabolism, Inflammation, and Oxidative Stress. \u003cem\u003eFood Res Int\u003c/em\u003e (2024) 188:114489. Epub 20240509. doi: 10.1016/j.foodres.2024.114489.\u003c/li\u003e\n\u003cli\u003eLi J, Niu C, Ai H, Li X, Zhang L, Lang Y, et al. Tsp50 Attenuates Dss-Induced Colitis by Regulating Tgf-Beta Signaling Mediated Maintenance of Intestinal Mucosal Barrier Integrity. \u003cem\u003eAdv Sci (Weinh)\u003c/em\u003e (2024) 11(11):e2305893. Epub 20240108. doi: 10.1002/advs.202305893.\u003c/li\u003e\n\u003cli\u003eYan J, Chen Q, Tian L, Li K, Lai W, Bian L, et al. Intestinal Toxicity of Micro- and Nano-Particles of Foodborne Titanium Dioxide in Juvenile Mice: Disorders of Gut Microbiota-Host Co-Metabolites and Intestinal Barrier Damage. \u003cem\u003eSci Total Environ\u003c/em\u003e (2022) 821:153279. Epub 20220121. doi: 10.1016/j.scitotenv.2022.153279.\u003c/li\u003e\n\u003cli\u003eLi Y, Yang P, Ye J, Xu Q, Wu J, Wang Y. Updated Mechanisms of Masld Pathogenesis. \u003cem\u003eLipids Health Dis\u003c/em\u003e (2024) 23(1):117. Epub 2024/04/23. doi: 10.1186/s12944-024-02108-x.\u003c/li\u003e\n\u003cli\u003eSato S, Iino C, Furusawa K, Yoshida K, Chinda D, Sawada K, et al. Effect of Oral Microbiota Composition on Metabolic Dysfunction-Associated Steatotic Liver Disease in the General Population. \u003cem\u003e#N/A\u003c/em\u003e (2025) 14(6). Epub 2025/03/27. doi: 10.3390/jcm14062013.\u003c/li\u003e\n\u003cli\u003eBorges-Canha M, Centelles-Lodeiro J, Leite AR, Chaves J, Lourenco IM, Von-Hafe M, et al. Gut Dysbiosis Is Linked to Severe Steatosis and Enhances Its Diagnostic Performance in Masld. \u003cem\u003eeGastroenterology\u003c/em\u003e (2025) 3(3):e100204. Epub 2025/09/08. doi: 10.1136/egastro-2025-100204.\u003c/li\u003e\n\u003cli\u003eZheng Z, Huang Y, Zhang J, Xie J, Pan A, Liao Y, et al. Antibiotic Consumption, Genetic Risk and Incidence of Metabolic Dysfunction-Associated Steatotic Liver Disease: A Prospective Cohort Study. \u003cem\u003eAnn Hepatol\u003c/em\u003e (2025):102136. Epub 2025/10/12. doi: 10.1016/j.aohep.2025.102136.\u003c/li\u003e\n\u003cli\u003eWang J, Zhu N, Su X, Gao Y, Yang R. Gut-Microbiota-Derived Metabolites Maintain Gut and Systemic Immune Homeostasis. \u003cem\u003eCells\u003c/em\u003e (2023) 12(5). Epub 2023/03/12. doi: 10.3390/cells12050793.\u003c/li\u003e\n\u003cli\u003eRooks MG, Garrett WS. Gut Microbiota, Metabolites and Host Immunity. \u003cem\u003e#N/A\u003c/em\u003e (2016) 16(6):341-52. Epub 2016/05/28. doi: 10.1038/nri.2016.42.\u003c/li\u003e\n\u003cli\u003eKim CH. Immune Regulation by Microbiome Metabolites. \u003cem\u003eImmunology\u003c/em\u003e (2018) 154(2):220-9. Epub 2018/03/24. doi: 10.1111/imm.12930.\u003c/li\u003e\n\u003cli\u003eWoo V, Alenghat T. Host-Microbiota Interactions: Epigenomic Regulation. \u003cem\u003eCurr Opin Immunol\u003c/em\u003e (2017) 44:52-60. Epub 2017/01/20. doi: 10.1016/j.coi.2016.12.001.\u003c/li\u003e\n\u003cli\u003eKasubuchi M, Hasegawa S, Hiramatsu T, Ichimura A, Kimura I. Dietary Gut Microbial Metabolites, Short-Chain Fatty Acids, and Host Metabolic Regulation. \u003cem\u003e#N/A\u003c/em\u003e (2015) 7(4):2839-49. Epub 2015/04/16. doi: 10.3390/nu7042839.\u003c/li\u003e\n\u003cli\u003eLei J, Wang X, Liu X. Microbiota-Derived Metabolites in the Epigenetic Regulation of the Host. \u003cem\u003eSci Bull (Beijing)\u003c/em\u003e (2025). Epub 2025/10/06. doi: 10.1016/j.scib.2025.09.030.\u003c/li\u003e\n\u003cli\u003eZhao H, Wu L, Yan G, Chen Y, Zhou M, Wu Y, et al. Inflammation and Tumor Progression: Signaling Pathways and Targeted Intervention. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e (2021) 6(1):263. Epub 2021/07/13. doi: 10.1038/s41392-021-00658-5.\u003c/li\u003e\n\u003cli\u003ePetraglia F, Vannuccini S, Donati C, Jeljeli M, Bourdon M, Chapron C. Endometriosis and Comorbidities: Molecular Mechanisms and Clinical Implications. \u003cem\u003eTrends Mol Med\u003c/em\u003e (2025). Epub 2025/10/03. doi: 10.1016/j.molmed.2025.09.002.\u003c/li\u003e\n\u003cli\u003eLiu M, Chen R, Zheng Z, Xu S, Hou C, Ding Y, et al. Mechanisms of Inflammatory Microenvironment Formation in Cardiometabolic Diseases: Molecular and Cellular Perspectives. \u003cem\u003eFront Cardiovasc Med\u003c/em\u003e (2024) 11:1529903. Epub 2025/01/29. doi: 10.3389/fcvm.2024.1529903.\u003c/li\u003e\n\u003cli\u003eHu T, Liu CH, Lei M, Zeng Q, Li L, Tang H, et al. Metabolic Regulation of the Immune System in Health and Diseases: Mechanisms and Interventions. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e (2024) 9(1):268. Epub 2024/10/09. doi: 10.1038/s41392-024-01954-6.\u003c/li\u003e\n\u003cli\u003eDou J, Jiang J, Xue Y, Jiang X, Jiang Y, Xiao P, et al. The Interplay of Cross-Organ Immune Regulation in Inflammation and Cancer. \u003cem\u003eMedComm (2020)\u003c/em\u003e (2025) 6(7):e70249. Epub 2025/06/18. doi: 10.1002/mco2.70249.\u003c/li\u003e\n\u003cli\u003eHui L, Li Y, Huang MK, Jiang YM, Liu T. Cxcl13: A Common Target for Immune-Mediated Inflammatory Diseases. \u003cem\u003eClin Exp Med\u003c/em\u003e (2024) 24(1):244. Epub 2024/10/24. doi: 10.1007/s10238-024-01508-8.\u003c/li\u003e\n\u003cli\u003ePriya S, Burns MB, Ward T, Mars RAT, Adamowicz B, Lock EF, et al. Identification of Shared and Disease-Specific Host Gene-Microbiome Associations across Human Diseases Using Multi-Omic Integration. \u003cem\u003eNat Microbiol\u003c/em\u003e (2022) 7(6):780-95. Epub 2022/05/17. doi: 10.1038/s41564-022-01121-z.\u003c/li\u003e\n\u003cli\u003eLiwinski T, Casar C, Ruehlemann MC, Bang C, Sebode M, Hohenester S, et al. A Disease-Specific Decline of the Relative Abundance of Bifidobacterium in Patients with Autoimmune Hepatitis. \u003cem\u003eAliment Pharmacol Ther\u003c/em\u003e (2020) 51(12):1417-28. Epub 2020/05/10. doi: 10.1111/apt.15754.\u003c/li\u003e\n\u003cli\u003eGavzy SJ, Kensiski A, Lee ZL, Mongodin EF, Ma B, Bromberg JS. Bifidobacterium Mechanisms of Immune Modulation and Tolerance. \u003cem\u003eGut Microbes\u003c/em\u003e (2023) 15(2):2291164. Epub 2023/12/06. doi: 10.1080/19490976.2023.2291164.\u003c/li\u003e\n\u003cli\u003eTojo R, Suarez A, Clemente MG, de los Reyes-Gavilan CG, Margolles A, Gueimonde M, et al. Intestinal Microbiota in Health and Disease: Role of Bifidobacteria in Gut Homeostasis. \u003cem\u003eWorld J Gastroenterol\u003c/em\u003e (2014) 20(41):15163-76. Epub 2014/11/12. doi: 10.3748/wjg.v20.i41.15163.\u003c/li\u003e\n\u003cli\u003eBocchio F, Mancabelli L, Milani C, Lugli GA, Tarracchini C, Longhi G, et al. Compendium of Bifidobacterium-Based Probiotics: Characteristics and Therapeutic Impact on Human Diseases. \u003cem\u003eMicrobiome Res Rep\u003c/em\u003e (2025) 4(1):2. Epub 2025/04/10. doi: 10.20517/mrr.2024.52.\u003c/li\u003e\n\u003cli\u003eYan J, Wang Z, Bao G, Xue C, Zheng W, Fu R, et al. Causal Effect between Gut Microbiota and Metabolic Syndrome in European Population: A Bidirectional Mendelian Randomization Study. \u003cem\u003eCell Biosci\u003c/em\u003e (2024) 14(1):67. Epub 20240528. doi: 10.1186/s13578-024-01232-6.\u003c/li\u003e\n\u003cli\u003eChai Z, Su Y, Tian X, Chen C, Lv X, Chen C. Predicting Disease Associations Based on the Higher Order Structure of Cerna Networks. \u003cem\u003eBrief Bioinform\u003c/em\u003e (2025) 26(5). doi: 10.1093/bib/bbaf518.\u003c/li\u003e\n\u003cli\u003eRaucci A, Zwergel C, Valente S, Mai A. Advancements in Hydrazide-Based Hdac Inhibitors: A Review of Recent Developments and Therapeutic Potential. \u003cem\u003eJ Med Chem\u003c/em\u003e (2025) 68(14):14171-94. Epub 20250710. doi: 10.1021/acs.jmedchem.5c01677.\u003c/li\u003e\n\u003cli\u003eSaeed H, Diaz LA, Gil-Gomez A, Burton J, Bajaj JS, Romero-Gomez M, et al. Microbiome-Centered Therapies for the Management of Metabolic Dysfunction-Associated Steatotic Liver Disease. \u003cem\u003eClin Mol Hepatol\u003c/em\u003e (2025) 31(Suppl):S94-S111. Epub 2024/11/28. doi: 10.3350/cmh.2024.0811.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Metabolic fatty liver disease, Type 2 diabetes mellitus, Gut microbiota, Transcriptomics, Mendelian randomization","lastPublishedDoi":"10.21203/rs.3.rs-9357546/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9357546/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e \u003cp\u003eThis study aimed to elucidate the common pathogenic pathways underlying the comorbidity of metabolic-associated fatty liver disease (MASLD) and type 2 diabetes mellitus (T2DM), with particular focus on identifying how gut microbiota and their metabolites regulate host gene expression in these mutually reinforcing metabolic disorders.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe employed an integrated multi-omics approach combining Mendelian randomization (MR) analysis with transcriptomic data from GEO datasets (GSE89632 and GSE26168) and gut microbiome GWAS data from MiBioGen. Differential gene expression analysis was performed to identify co-differentially expressed genes. MR analysis screened for disease-associated gut bacteria. Integration of host gene, metabolite, and microbiome data with databases such as gutMGene was conducted to determine key indicators. Functional enrichment analysis, protein-protein interaction network construction, and molecular docking simulations were performed. A predictive linear model was developed and validated using Hosmer-Lemeshow tests and decision curve analysis. Immunohistochemical analysis assessed correlations with immune cell infiltration, and ceRNA regulatory networks were constructed.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThe analysis identified 246 co-differentially expressed genes and 14 key gut bacteria potentially associated with both diseases. Seven key indicators were ultimately determined through data integration. These genes were enriched in lipid metabolism, inflammatory regulation, and redox processes. Four key protein nodes (PTGS2, ALOX15B, KLK3, and DHFR) were identified. Molecular docking revealed strong binding stability between core metabolites (genistein, phenylacetic acid, and N-acetylornithine) and target proteins. The predictive model demonstrated excellent performance with AUC values of 0.982 and 0.903 in the two datasets. Strong correlations were observed between key genes and immune cell infiltration rates, and multi-level ceRNA regulatory mechanisms were uncovered.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThis study demonstrates that the gut microbiota-metabolite-host gene axis plays a critical role in MASLD and T2DM comorbidity. The identified biomarkers and therapeutic targets provide important foundations for accurate prevention, diagnosis, and treatment strategies for these metabolic diseases.\u003c/p\u003e","manuscriptTitle":"Unraveling the Gut Microbiota-Metabolite-Host Gene Axis in the Pathogenesis of MASLD and Type 2 Diabetes Comorbidity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-14 17:59:25","doi":"10.21203/rs.3.rs-9357546/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e4b503cd-5ab8-4798-8b75-42f4148bc634","owner":[],"postedDate":"May 14th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewersInvited","content":"20","date":"2026-05-06T07:00:42+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-14T17:59:25+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-14 17:59:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9357546","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9357546","identity":"rs-9357546","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}