Integrated multi-omics analysis identifies TLR4-mediated mechanisms in ATBC-induced ovarian dysfunction and female infertility: A network toxicology, transcriptomic, and Mendelian randomization study.

Our research began by pinpointing ATBC-associated targets in the ChEMBL, Super-PRED, and STITCH databases, respectively. We then refined our search to 2076 ovarian damage-related and 3236 infertility-related targets using GeneCards, and identified an additional 125 ovarian damage-related and 4 infertility-related targets from the OMIM database. Following data consolidation and deduplication, we focused on 319 targets linked to ATBC, 2110 targets related to ovarian damage, and 3236 targets associated with infertility. The intersection of these target sets revealed 137 common targets for ATBC and ovarian damage, and 143 for infertility, suggesting their potential influence on ATBC-induced conditions (Fig.  1 A-B). Additionally, the intersection among ATBC, ovarian damage, and infertility revealed 104 shared targets (Figure S1 ). Fig. 1 Venn diagram of the targets of ATBC and ovarian damage and infertility. A Ovarian damage; B Infertility Venn diagram of the targets of ATBC and ovarian damage and infertility. A Ovarian damage; B Infertility We employed the STRING database to map the PPI networks related to ovarian damage and infertility. The ovarian damage network included 137 nodes interconnected by 1530 edges, averaging a node connectivity of 22.3. The infertility network featured 143 nodes with 1345 edges, averaging a node connectivity of 18.8. Cytoscape software was then used to scrutinize the topological attributes of the nodes, with a particular emphasis on degree and betweenness centrality (Fig.  2 A-D). Fig. 2 PPI networks involving ATBC and its common and core targets in ovarian damage and infertility. A PPI network for ATBC-associated ovarian damage, derived from STRING database data, with 137 nodes, 1530 edges, and an average node connectivity of 22.3. B PPI network for ATBC-associated infertility, featuring 143 nodes, 1345 edges, and an average node connectivity of 18.8. C , D PPI network diagrams for ATBC-related ovarian damage and infertility, respectively, highlighting nodes with a degree greater than 10, as generated by Cytoscape software. E , F Pivotal nodes within the ATBC-associated ovarian damage and infertility networks, respectively, with color intensity reflecting the ascending degree values of the nodes, as determined by the cytoHubba plugin PPI networks involving ATBC and its common and core targets in ovarian damage and infertility. A PPI network for ATBC-associated ovarian damage, derived from STRING database data, with 137 nodes, 1530 edges, and an average node connectivity of 22.3. B PPI network for ATBC-associated infertility, featuring 143 nodes, 1345 edges, and an average node connectivity of 18.8. C , D PPI network diagrams for ATBC-related ovarian damage and infertility, respectively, highlighting nodes with a degree greater than 10, as generated by Cytoscape software. E , F Pivotal nodes within the ATBC-associated ovarian damage and infertility networks, respectively, with color intensity reflecting the ascending degree values of the nodes, as determined by the cytoHubba plugin Network analysis pinpointed 19 core targets for ATBC-induced ovarian damage (Fig.  2 E), with TNF , SRC , and BCL2 being the most connected targets. For ATBC-induced infertility, 30 core targets were identified (Fig.  2 F), headed by TNF , SRC , and CASP3 . These genes encode proteins that are integral to a spectrum of cellular processes, such as inflammatory response, regulation of cell proliferation and cycle, apoptosis mediation, and signal transduction pathways. We then performed a Gene Ontology (GO) analysis on 137 ATBC-related targets for ovarian damage, concentrating on human biological features. This analysis yielded 1514 notable GO entries, covering 1282 biological processes, 92 cellular components, and 140 molecular functions. A parallel analysis for infertility's 143 targets identified 1384 significant entries, encompassing 1193 biological processes, 73 cellular components, and 118 molecular functions. We prioritized these findings by False Discovery Rate (FDR), highlighting the top 10 terms with the lowest FDR values across categories for visual representation in Figs.  3 and 4 . Fig. 3 GO Enrichment analysis for targets associated with ATBC-induced ovarian damage. A A bubble chart depicting the top 10 enriched biological processes; B A chord diagram linking these processes (BPs) to their respective genes. C and D Top 10 enriched cellular components and their interconnections with associated genes. E and F Top 10 enriched molecular functions and genes related to these functions (MFs). In this figure, the Gene Ontology (GO) terms are represented by bubbles, where size corresponds to the gene count within a pathway, and color depth indicates the level of enrichment significance, with darker red representing greater significance Fig. 4 GO Enrichment analysis for targets associated with ATBC-induced ovarian infertility. A A bubble chart depicting the top 10 enriched biological processes; B A chord diagram linking these processes (BPs) to their respective genes. C and D Top 10 enriched cellular components and their interconnections with associated genes. E and G Top 10 enriched molecular functions and genes related to these functions (MFs). In this figure, the Gene Ontology (GO) terms are represented by bubbles, where size corresponds to the gene count within a pathway, and color depth indicates the level of enrichment significance, with darker red representing greater significance GO Enrichment analysis for targets associated with ATBC-induced ovarian damage. A A bubble chart depicting the top 10 enriched biological processes; B A chord diagram linking these processes (BPs) to their respective genes. C and D Top 10 enriched cellular components and their interconnections with associated genes. E and F Top 10 enriched molecular functions and genes related to these functions (MFs). In this figure, the Gene Ontology (GO) terms are represented by bubbles, where size corresponds to the gene count within a pathway, and color depth indicates the level of enrichment significance, with darker red representing greater significance GO Enrichment analysis for targets associated with ATBC-induced ovarian infertility. A A bubble chart depicting the top 10 enriched biological processes; B A chord diagram linking these processes (BPs) to their respective genes. C and D Top 10 enriched cellular components and their interconnections with associated genes. E and G Top 10 enriched molecular functions and genes related to these functions (MFs). In this figure, the Gene Ontology (GO) terms are represented by bubbles, where size corresponds to the gene count within a pathway, and color depth indicates the level of enrichment significance, with darker red representing greater significance We delved into the functional roles of these targets in biological systems and their underlying mechanisms in disease pathogenesis through a comprehensive KEGG pathway analysis. The analysis revealed 179 significantly enriched pathways linked to ovarian damage and 164 pathways associated with infertility. Utilizing an inverse ranking based on False Discovery Rate (FDR) values, we generated a significance statistical bubble chart depicted in Fig.  5 , visually representing the top 10 KEGG signaling pathways for both conditions (Table  1 , 2 ). A chord diagram complements this by detailing the genes participating in these pathways. These integrated bioinformatics analyses have provided deeper insights into the targets'roles across biological processes, cellular components, and molecular functions, as well as their engagement in crucial signaling pathways. Fig. 5 KEGG enrichment analysis of ATBC-related targets for ovarian damage and infertility. A Bubble chart of the top 10 enriched KEGG pathways related to ATBC and ovarian damage. Each bubble represents a specific pathway, with the area indicating the number of enriched genes within that pathway. The color intensity of the bubbles signifies the significance of the enrichment, with deeper red shades indicating higher significance. B Distribution of 19 targets ( NFKB1, IL2, FLT3, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PTK2, MCL1, TLR4, BCL2L1, BCL2, FGFR1, CHRM2, PRKCA, RXRA, IKBKB, NR4 A1, IGF1R ) related to ATBC and ovarian damage in the PI3 K/AKT signaling pathway. C Bubble chart of the top 10 enriched pathways related to ATBC and infertility. D : Distribution of 16 targets ( NFKB1, CTSD, PARP2, KRAS, MAP2 K2, CASP6, ATM, CASP3, CTSB, CASP8, MAP3 K5, MCL1, BCL2L1, BCL2, TNF, IKBKB ) related to ATBC and infertility in the apoptosis signaling pathway Table 1 KEGG enrichment results of the targets between ATBC and ovarian damage ID Term Count FDR Gene hsa05200 Pathways in cancer 35 1.34E-21 MMP2, NFKB1, IL2, FLT3, PTGER2, KRAS, PDGFRA, MAP2 K2, PPARG, PTGER4, CASP3, EDNRA, NQO1, MMP1, NOS2, HSP90 AA1, PTK2, ESR2, CASP8, STAT1, PTGS2, XIAP, HDAC1, AR, BCL2L1, KEAP1, NFE2L2, BCL2, FGFR1, ESR1, PRKCA, BRAF, RXRA, IKBKB, IGF1R hsa04625 C-type lectin receptor signaling pathway 16 2.05E-14 NFKB1, IL2, MAPK14, KRAS, PRKCD, NLRP3, CASP8, STAT1, MAPKAPK2, PTGS2, SRC, TNF, IKBKB, CASP1, PTPN11, MALT1 hsa04210 Apoptosis 16 5.56E-13 NFKB1, CTSD, PARP2, KRAS, MAP2 K2, ATM, CASP3, CTSB, CASP8, MAP3 K5, MCL1, XIAP, BCL2L1, BCL2, TNF, IKBKB hsa04621 NOD-like receptor signaling pathway 16 2.21E-11 NAMPT, NFKB1, MAPK14, PRKCD, HSP90 AA1, NLRP3, CTSB, CASP8, STAT1, XIAP, TLR4, BCL2L1, BCL2, TNF, IKBKB, CASP1 hsa05163 Human cytomegalovirus infection 17 3.74E-11 NFKB1, MAPK14, PTGER2, KRAS, PDGFRA, MAP2 K2, CCR5, CCR1, PTGER4, CASP3, PTK2, CASP8, PTGS2, SRC, TNF, PRKCA, IKBKB hsa05145 Toxoplasmosis 13 7.53E-11 NFKB1, MAPK14, CCR5, CASP3, NOS2, CASP8, STAT1, XIAP, TLR4, BCL2L1, BCL2, TNF, IKBKB hsa05205 Proteoglycans in cancer 16 7.53E-11 MMP2, MAPK14, KRAS, MAP2 K2, CASP3, PTK2, PLAU, SRC, TLR4, FGFR1, TNF, ESR1, PRKCA, BRAF, PTPN11, IGF1R hsa04151 PI3 K-Akt signaling pathway 19 3.99E-10 NFKB1, IL2, FLT3, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PTK2, MCL1, TLR4, BCL2L1, BCL2, FGFR1, CHRM2, PRKCA, RXRA, IKBKB, NR4 A1, IGF1R hsa05215 Prostate cancer 12 4.78E-10 NFKB1, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PLAU, AR, BCL2, FGFR1, BRAF, IKBKB, IGF1R hsa05161 Hepatitis B 14 5.03E-10 NFKB1, MAPK14, KRAS, MAP2 K2, CASP3, CASP8, STAT1, SRC, TLR4, BCL2, TNF, PRKCA, BRAF, IKBKB Table 2 KEGG enrichment results of the targets between ATBC and infertility ID Term Count FDR Gene hsa05200 Pathways in cancer 35 6.36E-21 MMP2,NFKB1,IL2,FLT3,PTGER2,KRAS,PDGFRA, MAP2 K2, PPARG, PTGER1, PTGER4, CASP3, EDNRA, NQO1, MMP1, NOS2, HSP90 AA1, PTK2, ESR2, PTGER3, CASP8, STAT1, PTGS2, HDAC1, AR, BCL2L1, NFE2L2, BCL2, FGFR1, ESR1, PRKCA, BRAF, RXRA, IKBKB, IGF1R hsa04080 Neuroactive ligand-receptor interaction 29 3.91E-20 DRD4, NR3 C1, KISS1R, PTGER2, PTGER1, SSTR5, AVPR1 A, MC4R, PTGER4, ADORA2B, ADRB2, PRSS1, PLG, P2RY2, EDNRA, ADRB3, C5 AR1, PTGER3, SSTR2, DRD2, ADORA1, ADRB1, HRH2, CALCRL, DRD1, HRH1, OPRM1, RXFP1, ADORA2 A hsa04210 Apoptosis 16 1.1E-12 NFKB1, CTSD, PARP2, KRAS, MAP2 K2, CASP6, ATM, CASP3, CTSB, CASP8, MAP3 K5, MCL1, BCL2L1, BCL2, TNF, IKBKB hsa04625 C-type lectin receptor signaling pathway 14 7.92E-12 NFKB1, IL2, MAPK14, KRAS, PRKCD, NLRP3, CASP8, STAT1,PTGS2, SRC, TNF, IKBKB, CASP1, PTPN11 hsa05163 Human cytomegalovirus infection 18 7.92E-12 NFKB1, MAPK14, PTGER2, KRAS, PDGFRA, MAP2 K2, CCR5, PTGER1, PTGER4, CASP3, PTK2, PTGER3, CASP8, PTGS2,SRC,TNF,PRKCA, IKBKB hsa04020 Calcium signaling pathway 16 1.18E-10 PDGFRA, PTGER1, AVPR1 A, ADORA2B, ADRB2, EDNRA, NOS2, ADRB3, PTGER3, ADRB1, HRH2, DRD1, HRH1, PRKCA, NOS1, ADORA2 A hsa05205 Proteoglycans in cancer 16 1.26E-10 MMP2, MAPK14, KRAS, MAP2 K2, CASP3, PTK2, PLAU, SRC, TLR4, FGFR1, TNF, ESR1, PRKCA, BRAF, PTPN11, IGF1R hsa04151 PI3 K-Akt signaling pathway 19 8.58E-10 NFKB1, IL2, FLT3, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PTK2, MCL1, TLR4, BCL2L1, ITGA4, BCL2, FGFR1, PRKCA, RXRA, IKBKB, NR4 A1, IGF1R hsa05215 Prostate cancer 12 8.58E-10 NFKB1, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PLAU, AR, BCL2, FGFR1, BRAF, IKBKB, IGF1R hsa05161 Hepatitis B 14 8.99E-10 NFKB1, MAPK14, KRAS, MAP2 K2, CASP3, CASP8, STAT1, SRC, TLR4, BCL2, TNF, PRKCA, BRAF, IKBKB KEGG enrichment analysis of ATBC-related targets for ovarian damage and infertility. A Bubble chart of the top 10 enriched KEGG pathways related to ATBC and ovarian damage. Each bubble represents a specific pathway, with the area indicating the number of enriched genes within that pathway. The color intensity of the bubbles signifies the significance of the enrichment, with deeper red shades indicating higher significance. B Distribution of 19 targets ( NFKB1, IL2, FLT3, KRAS, PDGFRA, MAP2 K2, HSP90 AA1, PTK2, MCL1, TLR4, BCL2L1, BCL2, FGFR1, CHRM2, PRKCA, RXRA, IKBKB, NR4 A1, IGF1R ) related to ATBC and ovarian damage in the PI3 K/AKT signaling pathway. C Bubble chart of the top 10 enriched pathways related to ATBC and infertility. D : Distribution of 16 targets ( NFKB1, CTSD, PARP2, KRAS, MAP2 K2, CASP6, ATM, CASP3, CTSB, CASP8, MAP3 K5, MCL1, BCL2L1, BCL2, TNF, IKBKB ) related to ATBC and infertility in the apoptosis signaling pathway KEGG enrichment results of the targets between ATBC and ovarian damage KEGG enrichment results of the targets between ATBC and infertility The findings from GO and KEGG analyses highlight the extensive distribution and expression of these genes within diverse cellular structures. They are implicated in critical physiological functions, including responses to chemical stimuli, modulation of various cell death pathways, and signal transduction. Notably, the KEGG pathway enrichment identified several pathways associated with signaling mechanisms crucial to processes like apoptosis, the PI3 K-Akt signaling pathway, the NF-kappa B pathway, and pathways in cancer. The prominence of these pathways aligns with documented cases of ovarian damage and infertility, offering valuable insights into the genes'roles in pathological mechanisms. To investigate the underlying signaling pathways through which ATBC influences key genes associated with infertility and ovarian damage, KEGG pathway analysis for the core target genes was conducted. Results revealed the top 10 significantly enriched pathways based on P-value for both ATBC-induced ovarian damage and infertility (Fig.  6 A, B ), with 8 overlapping pathways. Within these pathways, 18 common core genes were identified ( MMP2 , NFKB1 , IL2 , KRAS , PPARG , CASP3 , HSP90 AA1 , STAT1 , PTGS2 , BCL2L1 , NFE2L2 , BCL2 , ESR1 , MAPK14 , SRC , TLR4 , TNF , ATM ), indicating that ATBC's impact on ovarian function could be a primary factor contributing to infertility. Fig. 6 Drug-pathway-target-disease network of ATBC in ovarian damage ( A ) and infertility ( B ). The blue hexagon represents ATBC. The green diamond represents pathways affected by ATBC in both ovarian damage and infertility. The red circles represent targets impacted by ATBC. The orange circles represent the diseases Drug-pathway-target-disease network of ATBC in ovarian damage ( A ) and infertility ( B ). The blue hexagon represents ATBC. The green diamond represents pathways affected by ATBC in both ovarian damage and infertility. The red circles represent targets impacted by ATBC. The orange circles represent the diseases The expression of 12 core genes (10 associated with infertility and 10 with ovarian damage) were validated in the context of infertility and ovarian damage using published datasets from the GEO database. In the GSE120103 dataset, when compared to the endometrium of fertile women, 7 significantly differentially expressed genes were identified in the endometrial tissue of infertile women. These included 3 downregulated genes ( ESR1 , STAT1 , and TNF ) and 4 upregulated genes ( HSP90 AA1 , KRAS , PTGS2 , and TLR4 ) (Fig.  7 A). In the subset of women with Stage IV ovarian endometriosis, 8 significantly differentially expressed genes were found in the endometrial tissue of infertile women compared to fertile women. This included 4 downregulated genes ( HSP90 AA1 , NFKB1 , PPARG , and STAT1 ) and 4 upregulated genes ( ESR1 , KRAS , PTGS2 , and TNF ) (Fig.  7 B). Within the GSE232306 dataset, when comparing women with diminished ovarian reserve function to those with normal ovarian reserve function, 9 significantly differentially expressed genes were identified, all of which were upregulated ( CASP3 , ESR1 , HSP90 AA1, KRAS, NFKB1, PPARG, PTGS2, SRC , and STAT1 ) (Fig.  7 C). Fig. 7 Expression of 12 Key Genes in Endometrium and Ovaries of Infertile Women. A Gene expression in endometrium from fertile and infertile women without disease ( n  = 9 per group). B Gene expression in endometrium from women with Stage IV endometriosis ( n  = 9 per group). C. Gene expression in ovarian tissue from women with reduced ovarian reserve and normal older controls. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001 Expression of 12 Key Genes in Endometrium and Ovaries of Infertile Women. A Gene expression in endometrium from fertile and infertile women without disease ( n  = 9 per group). B Gene expression in endometrium from women with Stage IV endometriosis ( n  = 9 per group). C. Gene expression in ovarian tissue from women with reduced ovarian reserve and normal older controls. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001 To explore the potential impact of core genes on the genetic susceptibility to infertility and ovarian damage, an Inverse Variance Weighted Mendelian Randomization analysis was conducted. The findings revealed that for each one-unit increase in the score of the TLR4 SNPs (rs2039954 and rs5020985) (Fig.  8 ), the beta value for infertility decreased by 0.2 (OR = 0.76, 95% CI 0.62–0.99; P  = 0.049) (Fig.  8 ). There was no strong evidence to suggest a association between the core genes and ovarian function. Fig. 8 Mendelian Randomization Analysis of Toll-like Receptor 4 ( TLR4 ) Genetic Variants on Female Infertility. A Scatter plot of the causal effect estimates for TLR4 single nucleotide polymorphisms (SNPs: rs2039954 and rs5020985) on female infertility risk, analyzed using a two-stage Mendelian randomization (MR) framework. Primary estimates were derived via the Inverse Variance Weighted (IVW) method (solid line), with sensitivity analysis performed using MR Egger regression (dashed line) to evaluate directional pleiotropy Mendelian Randomization Analysis of Toll-like Receptor 4 ( TLR4 ) Genetic Variants on Female Infertility. A Scatter plot of the causal effect estimates for TLR4 single nucleotide polymorphisms (SNPs: rs2039954 and rs5020985) on female infertility risk, analyzed using a two-stage Mendelian randomization (MR) framework. Primary estimates were derived via the Inverse Variance Weighted (IVW) method (solid line), with sensitivity analysis performed using MR Egger regression (dashed line) to evaluate directional pleiotropy Utilizing molecular docking strategies, we investigated the interactions of ATBC with the top 10 target proteins (8 shared proteins) implicated in both ovarian damage and infertility ( TNF, CASP3, HSP90 AA1, ESR1, PPARG, PTGS2, TLR4, STAT1, SRC, BCL2, NFKB1, KRAS ) (Table  3 ). The analysis revealed strong affinities and negative binding energies for all 12 targets, substantiating their crucial involvement in ATBC-induced conditions. Visualization of the lowest binding energies between each target and ATBC is presented (Fig.  9 , Table  4 ). These visual representations enhance our comprehension of the interaction mechanisms between ATBC and the target proteins, offering vital insights for delving into ATBC's reproductive toxicity mechanisms. Table 3 Core targets information Drug Disease Gene symbol No PDB ID Protein name Degree ATBC Infertility TNF 1 5UUK Tumor necrosis factor 86 Infertility SRC 2 8BQ3 Proto-oncogene tyrosine-protein kinase Src 77 Infertility CASP3 3 1RE1 Caspase 3 69 Infertility BCL2 4 8HTS BCL-2 66 Infertility ESR1 5 1X7E Estrogen receptor 66 Infertility NFKB1 6 1LE5 Nuclear factor NF-kappa-B p50 subunit 62 Infertility HSP90 AA1 7 5 NJX HSP90 AA1 protein 60 Infertility PPARG 8 2POB Peroxisome proliferator-activated receptor gamma 55 Infertility PTGS2 9 5 F19 Prostaglandin G/H synthase 2 54 Infertility KRAS 10 6GQT GTPase KRas 45 Ovarian damage TNF 1 1 TNF Tumor necrosis factor 95 Ovarian damage SRC 2 8BQ3 Proto-oncogene tyrosine-protein kinase Src 77 Ovarian damage BCL2 3 8HTS BCL-2 76 Ovarian damage CASP3 4 1RE1 Caspase 3 75 Ovarian damage NFKB1 5 1LE5 Nuclear factor NF-kappa-B p50 subunit 71 Ovarian damage ESR1 6 1X7E Estrogen receptor 69 Ovarian damage HSP90 AA1 7 5 NJX HSP90 AA1 protein 65 Ovarian damage PTGS2 8 5 F19 Prostaglandin G/H synthase 2 62 Ovarian damage PPARG 9 2POB Peroxisome proliferator-activated receptor gamma 61 Ovarian damage TLR4 10 2Z63 Toll-like receptor 4 53 Ovarian damage STAT1 11 3 WWT Signal transducer and activator of transcription 1-alpha/beta 52 Ovarian damage KRAS 12 6GQT GTPase KRas 52 Fig. 9 Molecular Docking Analysis of ATBC with Key Target Proteins Associated with Ovarian Damage and Infertility. A - L Structural visualization of molecular docking interactions between ATBC and 12 target proteins implicated in ovarian damage and infertility. Proteins include: ( A ) BCL2, ( B ) CASP3, ( C ) ESR1, ( D ) HSP90 AA1, ( E ) KRAS, ( F ) NFKB1, ( G ) PPARG, ( H ) PTGS2, ( I ) SRC, ( J ) STAT1, ( K ) TLR4, and ( L ) TNF. Each panel illustrates the lowest-energy binding conformation of ATBC (stick model, carbon atoms in cyan) within the active site of the respective protein (surface or ribbon representation) Table 4 Molecular docking scores (kcal/mol) Ligand Receptors Docking scores ATBC TNF −6.5 SRC −4.2 CASP3 −5.1 BCL2 −4.7 ESR1 −7.5 NFKB1 −4.9 HSP90 AA1 −4.8 PPARG −6.0 PTGS2 −6.7 KRAS −6.3 TLR4 −4.8 STAT1 −5.0 Core targets information Molecular Docking Analysis of ATBC with Key Target Proteins Associated with Ovarian Damage and Infertility. A - L Structural visualization of molecular docking interactions between ATBC and 12 target proteins implicated in ovarian damage and infertility. Proteins include: ( A ) BCL2, ( B ) CASP3, ( C ) ESR1, ( D ) HSP90 AA1, ( E ) KRAS, ( F ) NFKB1, ( G ) PPARG, ( H ) PTGS2, ( I ) SRC, ( J ) STAT1, ( K ) TLR4, and ( L ) TNF. Each panel illustrates the lowest-energy binding conformation of ATBC (stick model, carbon atoms in cyan) within the active site of the respective protein (surface or ribbon representation) Molecular docking scores (kcal/mol)

We ascertained the canonical structure and SMILES string of ATBC by querying"acetyl tributyl citrate"in public chemical databases. This foundation allowed us to extract potential ATBC targets from ChEMBL [ 38 ] ( https://www.ebi.ac.uk/chembl/ ) and STITCH database [ 30 ] ( http://stitch.embl.de/ ), filtered for Homo sapiens. The SMILES notation was further employed in the Super-PRED databases ( https://prediction.charite.de/subpages/target_prediction.php ) to uncover any additional targets. A thorough review and structural consistency check were conducted on the identified targets. After deduplication, target nomenclature was standardized via UniProt ( https://www.uniprot.org/ ), culminating in the formation of a curated ATBC target library. We systematically queried the GeneCards ( https://www.genecards.org/ ) and OMIM databases ( https://www.omim.org/ ) to identify ovarian damage-associated genes using the keywords"poor ovarian response,""premature ovarian failure,""premature ovarian insufficiency,"and"diminished ovarian reserve"(Relevance score > 10), while infertility-related targets were retrieved via the terms"infertility,""sterility,""subfertility,"and"reproductive sterility"(Relevance score > 1). To align the identified genes closely with the toxicological profiles of ATBC, a stringent"score"threshold of 10 was applied, with only genes exceeding this value being included in our reproductive toxicity gene library. Venn diagrams generated via a bioinformatics platform ( http://www.bioinformatics.com.cn/ ) analyzed intersections between ATBC-associated targets and ovarian damage/infertility-related genes, and their three-way overlap, identifying shared genes as potential critical mediators of ATBC-induced reproductive toxicity. We mapped a Protein–Protein Interaction (PPI) network to pinpoint core targets. We commenced by uploading the intersecting targets to the STRING database ( https://cn.string-db.org/ ), filtered for Homo sapiens and set with a confidence score threshold of 0.4 to ensure interaction reliability. The derived PPI data were then integrated into Cytoscape (version 3.8.2), a network biology tool adept at visualizing and analyzing molecular interactions. With Cytoscape, topological analysis of the network's nodes and edges was performed, generating a comprehensive PPI map. For the core target selection related to ATBC-induced ovarian damage and infertility, we adopted criteria that included nodes exceeding the median in betweenness and closeness centrality, as well as those with degree values twice the median and average shortest path lengths below the median. To substantiate key genes within the PPI network, the cytoHubba plugin was applied. This synergistic approach, combining stringent criteria with analytical tools, facilitated the precise identification of targets linked to ATBC's impact on ovarian damage and infertility. The targets identified from the PPI network were subjected to GO and KEGG enrichment analyses using DAVID ( https://david-d.ncifcrf.gov/ ), a comprehensive bioinformatics resource. Significance was determined with a p -value threshold of < 0.05, focusing on the 10 most impactful pathways. The GO analysis spanned biological processes, cellular components, and molecular functions. Visualization of these enriched pathways was performed using an online bioinformatics platform ( http://www.bioinformatics.com.cn/ ), providing a clear representation of the potential biological significance. In this section, we employed Cytoscape to build a comprehensive network linking drug interactions, pathways, targets, and diseases, aiming to uncover how ATBC influences ovarian damage and infertility. By gathering targets implicated in drug-disease interactions, we were able to visually represent the potential mechanisms by which ATBC may lead to these conditions. This network provides insights into the complex interactions at play and the pathways potentially disrupted by ATBC. The GSE120103 dataset, utilizing the GPL24676 platform, comprises endometrial samples from women who are fertile and infertile, with and without Stage IV Ovarian Endometriosis, organized into four groups, each consisting of nine participants. This dataset was analyzed to validate differential expression of core genes in endometrial tissues from women with infertility, with statistical significance defined as adjusted p -values < 0.05. Additionally, the GSE232306 dataset, based on the GPL6480 platform, includes samples from women of advanced maternal age characterized by either normal ovarian reserve (NOR, with six participants) or diminished ovarian reserve (DOR, also with six participants). This dataset was analyzed to identify differential expression of core genes associated with ovarian dysfunction, with statistical significance defined as adjusted p -values < 0.05. For differential expression analysis, the GEO2R online tool, offered by the GEO database, was employed. The tool can be accessed at http://www.ncbi.nlm.nih.gov/geo/geo2r . The two-sample Mendelian Randomization (2SMR) analysis was conducted using the MRanalysis platform ( https://www.mranalysis.cn/ ). Genetic variants of 12 key genes were selected from the IEU GWAS database as instrumental variables for the exposure, and LD clumping was performed to ensure their independence. The outcomes of interest were derived from GWAS summary statistics for European women with infertility (finn-b-N14_female infertility, including 6,481 cases and 68,969 controls) and ovarian dysregulation (finn-b-E4_ovarian dysregulation, including 440 participants). The effect sizes for the instruments on both exposures and outcomes were harmonized to reference the same allele to facilitate accurate interpretation. MR analysis and sensitivity analyses were then conducted to visualize the results. To validate the MR assumptions, SNPs with F-statistics greater than 10 were selected to ensure instrument strength, horizontal pleiotropy was assessed using MR-Egger regression (intercept p  = 0.23) and leave-one-out analyses (Supplementary Figure S4), and exclusion restriction was confirmed through LD pruning (r 2  < 0.001, 10,000 kb window) and PhenoScanner screening for confounding traits. We utilized CB-Dock2 ( https://cadd.labshare.cn/cb-dock2/index.php ) for molecular docking to estimate the binding affinity of ATBC with its potential core targets. The initial step was ligand preparation, with the ATBC structure obtained in SDF format from the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ). For protein preparation, we accessed the crystallographic structures of core proteins via gene names at the Protein Data Bank ( https://www.rcsb.org/ ). PyMOL 2.4.1 was applied to eliminate water and the original ligands. Subsequently, CB-Dock2's Structure-based Blind Docking algorithm facilitated the analysis and visualization of the interactions between ATBC and these core target proteins.

This study delineates causal and associative pathways underlying ATBC-induced reproductive toxicity. Causal evidence emerged from Mendelian randomization, revealing a protective association between genetically proxied TLR4 levels and reduced infertility risk in Europeans (OR = 0.76, 95% CI 0.62–0.99). Mechanistic associations were identified through multi-omics integration: network toxicology pinpointed 137/143 targets linked to ovarian damage/infertility, while GO/KEGG analyses implicated apoptosis, oxidative stress, and PI3 K/AKT pathways. Molecular docking confirmed strong ATBC- TLR4 interactions (binding energy: − 4.8 kcal/mol), suggesting direct target engagement. Correlative observations included transcriptomic signatures (7 ovarian damage- and 9 infertility-associated core genes). Together, these stratified findings provide a causal framework for TLR4 -mediated ovarian damage while highlighting therapeutic targets to mitigate ATBC's fertility risks. Numerous studies have highlighted the significance of ESR1 in ovarian damage and infertility. Polymorphisms within the estrogen receptor α gene ( ESR1 ) have been linked to premature ovarian insufficiency (POI) in Chinese women, with corroborating evidence from studies on adult female rhesus monkeys where ESR1 suppression leads to ovarian dysfunction [ 19 ]. The interplay between CYP19 A1 and ESR1 , along with variations in genes like HK3 and BRSK1 , influences POI risk and the timing of natural menopause [ 16 , 23 ]. Animal models have shown that ESR1 loss impairs ovulation and can cause follicular atresia, with environmental factors such as polycyclic aromatic hydrocarbons impacting estradiol secretion via ESR1 [ 37 ]. In humans, the absence of ESR1 in the pituitary or within the reproductive tract can disrupt estrogen feedback, leading to infertility and a lack of uterine responsiveness [ 33 ]. Furthermore, the ESR1 rs9340799 polymorphism is tied to infertility associated with endometriosis and failures in in vitro fertilization [ 22 ]. These insights underscore ESR1 's pivotal role in reproductive health, particularly in the realms of ovarian function, follicular development, ovulation, and estrogen feedback regulation. Recent studies have intensively explored the involvement of apoptosis in premature ovarian failure and infertility, with key roles attributed to CASP3 and BCL2 . In a cohort of 1,030 Chinese POI patients, six heterozygous TP63 gene mutations were found to constitutively activate the mutant protein, thereby inducing apoptosis through upregulation of pro-apoptotic factors [ 13 ]. A Prdx4 knockout mouse model exhibited decreased ovarian weight, more atretic follicles, and a rise in apoptotic granulosa cells, along with heightened levels of apoptotic markers such as CHOP , caspase-12 , and BAX [ 18 ]. Tripterygium wilfordii polyglycoside has been shown to trigger POF and granulosa cell apoptosis by interfering with high-mobility group protein 1 [ 32 ]. A mouse model induced by the chemotherapy drug cyclophosphamide demonstrated increased apoptosis in ovarian granulosa and theca cells [ 2 ]. Bone marrow stem cell transplantation has been found to inhibit ovarian apoptosis by modulating the expression of genes like Bax, p53, caspase-3 , and Bcl-2 , and it can also stimulate Bcl2 expression to further suppress granulosa cell apoptosis [ 1 ]. Enriched pathway analysis revealed the PI3 K/AKT pathway's substantial influence on ATBC-induced ovarian damage and infertility, a mechanism frequently implicated in premature ovarian insufficiency. The environmental endocrine disruptor nonylphenol triggers a significant rise in ROS production linked to the AKT/AMPK/mTOR pathway activation [ 4 ]. A potential correlation exists between the heightened risk of POF in young female cancer patients undergoing cyclophosphamide chemotherapy and the activation of the PTEN/AKT/FoxO3a pathway [ 9 ]. PTEN 's catalytic conversion of PIP3 to PIP2 halts the PI3 K signaling cascade, facilitating mitochondria-driven apoptosis in granulosa cells. Maternal exposure to particulate matter 2.5 activates both the PI3 K/AKT/FOXO3a and NF-κB pathways, diminishing offspring mice's ovarian reserve [ 3 ]. Cadmium (Cd), a heavy metal, induces ROS and inflammatory mediators, impeding cell cycle progression and promoting apoptosis, with its involvement in the PTEN/AKT/FOXO3a pathway leading to granulosa cell apoptosis and follicular atresia in chickens [ 40 ]. Pharmacologically, growth hormone offsets ROS accumulation by activating the PI3 K/AKT pathway, reducing CASP-3 and caspase-9 activity in polycystic ovary syndrome patients. Metformin mitigates excessive granulosa cell autophagy and PCOS in rats via the PI3 K/AKT/mTOR pathway [ 35 ]. Baicalin bolsters ovarian function in aged mice, activating the PI3 K/AKT/mTOR pathway and upregulating anti-apoptotic Bcl-2 while lowering pro-apoptotic BAX and CASP-3 , resulting in decreased cell death and heightened steroid hormone secretion [ 8 ]. Enrichment analysis indicates that ATBC's reproductive toxicity may be linked to the NF-κB signaling pathway. In mouse models with NF-KREB inactivation, there's a noted decrease in P65 nuclear expression in granulosa cells and a reduction in antral follicle count [ 20 ]. Overactivation of NF-κB can trigger the expression of inflammatory mediators, including IL-1 , adhesion molecules, and immune receptors, which can worsen the impact of oxidative stress on the ovaries, leading to a detrimental cycle. Studies show that salicylic acid may mitigate radiation-induced POI by lowering the levels of NF-κB, TNF-α, iNOS , and COX-2 , and by modulating the TGF-β/MAPK pathway [ 21 ]. Furthermore, resveratrol has been demonstrated to significantly decrease TNF-α and IL-6 levels in mouse ovaries, offering potential protective effects against ovarian damage [ 14 ]. The oxidative stress pathway is notably linked to the reproductive toxicity induced by ATBC, with oxidative stress biomarkers identified across various regions of the female reproductive system, highlighting their significance in physiological processes. A plethora of research has connected reactive oxygen species (ROS) to infertility factors, including pelvic and fallopian tube issues, endometriosis, and unexplained infertility cases. The impact of oxidative stress on these conditions has been extensively studied, revealing its potential role in impeding fertilization and early embryonic development due to harmful effects on gametes and embryos within the reproductive microenvironments [ 5 ]. In endometriosis, activated macrophages contribute to the disease's pathogenesis, being responsible for the surge in ROS within the peritoneal environment, which is closely related to the condition [ 27 ]. Analysis of the functionality of core targets and pathway enrichment underscores the pivotal role of Acetyl Tributyl Citrate (ATBC) in inducing ovarian damage and infertility. Approved by the U.S. FDA, ATBC is a commercially available plasticizer allowed for use as a food additive and in food-contact materials. It has been shown to migrate into various foods from packaging at a rate significantly higher than that of the potent endocrine disruptor DEHP [ 36 ]. ATBC, even at 4 μg/ml, can activate the expression of CYP3 A4 in ovarian cells, a key enzyme in the metabolism of endogenous steroid hormones [ 31 ]. Moreover, exposure to as little as 0.01 μg/ml ATBC can cause growth stagnation and a notable increase in atretic follicles [ 25 ]. A recent study indicated that daily exposure to 10 mg/kg/day ATBC in female CD-1 mice leads to a decrease in the quantity of ovarian follicles at various stages of development [ 24 ]. This study, primarily bioinformatics-driven, has inherent limitations. While network toxicology identified TLR4 -mediated mechanisms, it cannot assess dosage effects quantitatively. Clinical validation remains essential to confirm findings. Future studies should prioritize pharmacological validation in human models and address challenges in translating TLR4 -targeted therapies due to systemic immune risks. This multi-omics study reveals TLR4 as a central mediator of ATBC-induced ovarian toxicity and infertility, driving apoptosis and oxidative stress via PI3 K-AKT signaling. Mendelian randomization supports TLR4 's protective role against infertility, corroborated by strong ATBC- TLR4 binding. Our findings implicate TLR4 -mediated inflammation in environmental ovarian dysfunction and propose its therapeutic targeting to mitigate plasticizer-related fertility decline.

Globally, 10% to 15% of reproductive-age couples face infertility, highlighting its status as a major public health issue [ 26 ]. Infertility compromises physical health, precipitating conditions like osteoporosis and cardiovascular diseases, and takes a toll on mental well-being, causing depression, anxiety, and feelings of inferiority, while also adding an economic strain to individuals and society. Causes range from female-related issues such as follicular and tubal problems to male sperm issues, and encompass endocrine, immune, genetic, and environmental factors [ 39 ]. Research indicates that exposure to environmental toxins significantly contributes to infertility by potentially disrupting the hypothalamic-pituitary-ovarian axis, impairing egg quality, and leading to fertilization issues and abnormal embryonic development [ 28 ]. These toxins can also alter the uterine environment, hindering embryo implantation and increasing miscarriage risk. Importantly, they affect both genders, impacting sperm quality in men and indirectly contributing to female infertility [ 10 ]. The influence of environmental toxins on fertility and the heightened risk of infertility is an issue of growing concern. The prevalence of ovarian damage and POI is increasing worldwide, posing a major threat to women's reproductive health. These conditions are responsible for an estimated 10–15% of infertility cases [ 15 ]. Factors such as societal pressures, environmental influences, and lifestyle shifts could further raise this percentage. Among the causes of ovarian damage, environmental toxins play a significant role, impacting ovarian health through multiple pathways [ 29 ] They can act as endocrine disruptors, throwing off hormonal balance and impairing ovarian function, potentially causing ovulation issues and menstrual irregularities. Environmental toxins may also incite oxidative stress, generating free radicals that harm ovarian tissues and oocytes, diminishing egg quality and count [ 17 ]. They could further compromise the DNA of ovarian cells, elevating the risk of genetic mutations and jeopardizing the genetic integrity of eggs. Prolonged exposure might also lead to chronic inflammation within the ovaries, altering the ovarian milieu and hindering follicular maturation and oocyte development. Recent studies have highlighted the connection between endocrine disruptors and conditions like ovarian damage and infertility, with a spotlight on plasticizers known as phthalates. Ubiquitous in the environment and everyday products, phthalates have been associated with a spectrum of gynecological disorders, from polycystic ovary syndrome to endometriosis and premature ovarian failure, and even malignancies [ 6 , 7 ]. In light of these findings, the plasticizer acetyl tributyl citrate (ATBC) is gaining popularity for its perceived safety and eco-friendliness, superseding DEHP in PVC applications [ 34 ]. Yet, the rise in ATBC usage has prompted a surge in research aimed at understanding its impact on female reproductive health. The mechanisms of ATBC's influence on reproductive well-being remain largely unknown, underscoring the need for in-depth investigation into its risks and biological interactions. Since 2007, network pharmacology has emerged as a pivotal technique for delving into the molecular mechanisms of drug action, underpinned by systems biology and bioinformatics [ 11 ]. This method underscores the intricate interactions among drugs, pathways, targets, and diseases, clarifying the multi-target nature of drug therapies, particularly in psychiatry, oncology, and infectious disease treatment [ 12 ]. The efficacy of drugs that target multiple sites is well-aligned with network pharmacology principles. From this vantage point, a more profound insight into drug-disease interactions is attainable, facilitating the innovation of treatments for complex conditions. Molecular docking, as a virtual screening technique, scrutinizes receptor-small molecule interactions, offering a vital instrument for elucidating the mechanisms behind Western medicine's efficacy against specific diseases. Our research harnessed a suite of database resources for a comprehensive analysis of acetyl tributyl citrate (ATBC), encompassing target identification, analysis of protein–protein interaction networks, Gene Ontology (GO) and KEGG pathway analysis, and the construction of drug-pathogen-target-disease networks, coupled with molecular docking analysis. This multifaceted approach has shed light on the mechanisms by which ATBC induces ovarian damage and infertility, offering a deeper molecular understanding. We anticipate that our methodology will yield novel insights and strategies regarding the toxicity of ATBC on women's fertility potential.

Supplementary Material 1. Figure S1 Venn diagram of the targets of ovarian damage and infertility Supplementary Material 1. Figure S1 Venn diagram of the targets of ovarian damage and infertility