Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Elucidating the potential carcinogenic molecular mechanisms of parabens in head and neck squamous cell carcinoma through network toxicology and molecular docking

  • Lei Zhao,

    Roles Conceptualization, Formal analysis, Methodology, Project administration, Resources, Software, Supervision, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯
  • Jianwang Yang,

    Roles Methodology, Resources, Software, Writing – review & editing

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯
  • Tao Liu,

    Roles Data curation, Formal analysis, Software, Writing – review & editing

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯
  • Huan Cao,

    Roles Data curation, Investigation, Writing – review & editing

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯
  • Miaomiao Yu,

    Roles Data curation, Visualization, Writing – review & editing

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯
  • Baoshan Wang

    Roles Conceptualization, Funding acquisition, Project administration, Supervision

    wangbaoshan@hebmu.edu.cn

    Affiliation Department of Otolaryngology-Head and Neck Surgery, The Second Hospital of Hebei Medical University, Shijiazhuang, Hebei, China

    ⨯

Abstract

Objective

This study aims to systematically investigate the molecular mechanisms through which parabens may contribute to head and neck squamous cell carcinoma (HNSCC) carcinogenesis using integrated network toxicology and molecular docking.

Materials and methods

Six commonly used parabens (ethyl-, propyl-, methyl-, heptyl-, butyl-, and benzylparaben) were selected for toxicity prediction via ProTox 3.0 and ADMETlab2.0. Their potential targets were retrieved from Swiss Target Prediction, ChEMBL, and the Similarity Ensemble Approach (SEA). HNSCC-related targets were collected from GeneCards, OMIM, and CTD, while differentially expressed genes (DEGs) were identified using TCGA-HNSC data. Functional enrichment, protein-protein interaction (PPI) network construction, hub gene identification, molecular docking, and immune infiltration analyses were performed using DAVID, STRING, Cytoscape, CB-DOCK2, and TIMER2.0.

Results

We identified 80 common targets through which parabens may exert toxic effects in HNSCC. Among these, 12 hub genes (CCNB1, CDK1, CCNA2, CDK2, CDK4, TYMS, AURKA, CCNA1, CHEK1, CCNB2, PLK1, CDC25A) were significantly overexpressed in HNSCC tissues and were primarily enriched in cell cycle regulation, p53 signaling, and viral carcinogenesis pathways. CCNA1 was further validated as an independent prognostic factor associated with poor survival. Molecular docking revealed strong binding affinities between parabens and hub proteins. Immune infiltration analysis indicated a negative correlation between CCNA1 expression and CD8 + T cell and B cell infiltration.

Conclusions

Parabens may promote HNSCC progression by disrupting cell cycle regulation and immune responses via direct interactions with key hub genes. These findings provide a novel mechanistic basis for the carcinogenic potential of parabens in HNSCC and underscore the need for further experimental and epidemiological validation.

Citation: Zhao L, Yang J, Liu T, Cao H, Yu M, Wang B (2026) Elucidating the potential carcinogenic molecular mechanisms of parabens in head and neck squamous cell carcinoma through network toxicology and molecular docking. PLoS One 21(3): e0333867. https://doi.org/10.1371/journal.pone.0333867

Editor: Amel Mohamed El Asely, Benha University, EGYPT

Received: September 20, 2025; Accepted: March 6, 2026; Published: March 23, 2026

Copyright: © 2026 Zhao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting information files.

Funding: This research was funded by the Clinical Research Center Foundation of Hebei Provincial Department of Science and Technology, grant number 20577716D, and the Natural Science Foundation of Hebei Province, grant number H2022206376.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Parabens are among the most widely used synthetic preservatives globally. Their broad-spectrum antimicrobial efficacy and low cost have led to extensive application in cosmetics, pharmaceuticals, food products, and personal care items (PCPs) [1–5]. Human exposure to parabens is continuous and nearly universal, occurring primarily through dermal absorption and dietary intake [5–7]. Growing evidence suggests potential adverse effects, including associations with cancers, diabetes, thyroid diseases, and obesity, raising public health concerns [3,6–11].

Parabens are well-established endocrine-disrupting chemicals (EDCs) with estrogen-mimicking properties. They bind to and activate estrogen receptors (ERs), potentially disrupting hormonal homeostasis, reproductive functions, and developmental processes [1,12–15]. Numerous in vitro and in vivo studies have linked parabens to developmental toxicity, reproductive disorders, and allergic skin reactions [16–20]. Of particular relevance to carcinogenesis is the ability of longer-chain parabens, such as butylparaben, to exhibit stronger estrogenic activity and promote proliferation in hormone-dependent cancer cells [12,21–24]. Although parabens are not classified as direct mutagens, their role as tumor promoters is under active investigation. Research has primarily focused on hormone-responsive cancers such as breast cancer, where intact parabens have been detected in tumor tissues and shown to drive proliferation and migration via ER-mediated pathways and oncogenic signaling cascades like MAPK/ERK and PI3K/Akt [12,21,25–27]. In contrast, their impact on non-hormonally driven cancers, including head and neck cancers, remains poorly understood [28,29].

Parabens are consistently detected in environmental matrices worldwide due to extensive consumption and incomplete removal during wastewater treatment [30–33]. They have been found in human urine, breast tissue, and adipose tissue, confirming widespread exposure [7,30,31]. Biomonitoring studies detect paraben metabolites in over 90% of urine samples from general populations [34,35], highlighting the need for comprehensive toxicological risk assessment.

Head and neck squamous cell carcinoma (HNSCC) ranks as the sixth most common cancer globally, with a five-year survival rate that has remained stagnant for decades [36,37]. GLOBOCAN estimates indicate a significant global impact from HNSCC in 2025, with projected figures of 946,456 new cases and a mortality count of 482,001 [38]. Established risk factors include tobacco use, alcohol consumption, and high-risk human papillomavirus (HPV) infection [36,37]. However, many cases, especially among young adults without traditional risk factors, suggest a role for other environmental or occupational exposures [36,39]. Despite advances in treatment, the molecular pathogenesis of HNSCC remains complex and heterogeneous, frequently involving aberrant activation of EGFR, NF-κB, and PI3K/Akt pathways, as well as dysregulation of cell cycle and apoptosis mechanisms [40–46].

Recent evidence indicates that parabens can modulate chemotaxis, adhesion, and phagocytosis in macrophages [2], which play a dual regulatory role in HNSCC tumor immunology [47]. The potential contribution of parabens to HNSCC initiation or progression represents a compelling yet underexplored research area.

Thus, a critical gap exists between the well-documented widespread human exposure to parabens, their suggested tumor-promoting capabilities in other cancers, and the scarcity of research on their biological effects in HNSCC. It remains unknown whether parabens can influence hallmark traits of HNSCC, such as proliferation, invasion, metastasis, or therapy resistance, or what underlying molecular mechanisms might be involved. Addressing this gap is essential for a comprehensive understanding of HNSCC etiology and for informing public health policies on the safety of these common preservatives. Focusing on six commonly used parabens (ethyl-, propyl-, methyl-, heptyl-, butyl-, benzylparaben) [1,2], this study aims to bridge this gap by employing network toxicology integrated with molecular docking to systematically predict and elucidate the molecular mechanisms through which parabens may influence HNSCC development and progression.

Materials and methods

Toxicity prediction and analysis of common parabens

Canonical SMILES sequences of the six parabens were obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov/) [48] and submitted to ProTox 3.0 (https://tox.charite.de/protox3/) [49] and ADMETlab2.0 (https://admetmesh.scbdd.com/service/evaluation/cal) [50] for toxicity prediction.

Retrieval of paraben targets

Potential protein targets of the six parabens were identified using Swiss Target Prediction (http://www.swisstargetprediction.ch/) [51] (Homo sapiens, probability > 0), ChEMBL (https://www.ebi.ac.uk/chembl/) [52,53] (Homo sapiens), and the Similarity Ensemble Approach (SEA) (https://sea.bkslab.org/) [54] (Homo sapiens, probability > 0). SMILES sequences were used as input for Swiss Target Prediction and SEA, while PubChem names were used for ChEMBL queries. Targets were normalized using UniProt (https://www.uniprot.org/) [55,56].

Retrieval of HNSCC-related targets

HNSCC-associated targets were collected from GeneCards (https://www.genecards.org/) [57], Online Mendelian Inheritance in Man (OMIM) (https://omim.org/) [58], and Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) [59] using the keywords: “Squamous Cell Carcinoma of Head and Neck”, “HNSCC”, “Head and Neck Squamous Cell Carcinoma”, “Oral Cavity Squamous Cell Carcinoma”, “Squamous Cell Carcinoma of the Larynx”, “Laryngeal Squamous Cell Carcinoma”, “Squamous Cell Carcinoma of the Nasal Cavity”, “Oropharyngeal Squamous Cell Carcinoma”, “Hypopharyngeal Squamous Cell Carcinoma”, and “Tongue Squamous Cell Carcinoma”.

Acquisition of HNSCC-related targets from TCGA

Differential gene expression profiles and clinical phenotype data were downloaded from The Cancer Genome Atlas (TCGA) (https://www.cancer.gov/tcga) [60,61] via the Home for Researchers platform (https://www.home-for-researchers.com/). Differentially expressed genes (DEGs) were defined as those with adjusted p-value < 0.05 and |log2(fold change)| ≥ 1. Gene Ontology (GO) (http://geneontology.org/) [62,63] and Kyoto Encyclopedia of Targets and Genomes (KEGG) pathway (https://www.kegg.jp) [64,65] enrichment analyses were also completed with the same platform.

Identification of common targets for parabens and HNSCC

Overlapping targets among paraben-related targets, HNSCC-related targets, and TCGA-derived HNSCC DEGs were identified using jvenn (https://jvenn.toulouse.inrae.fr/app/example.html) [66]. The resulting common targets were considered candidate targets for paraben-induced toxicity in HNSCC. Functional enrichment analyses (GO, KEGG, and REACTOME Pathway (https://reactome.org/) [67,68])were conducted using Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/) [69,70] with default parameters (OFFICIAL GENE SYMBOL, Homo sapiens). Results were visualized via circle enrichment plots generated with ChiPlot (https://www.chiplot.online/).

Construction of protein-protein interaction network

A protein-protein interaction (PPI) network for the candidate targets of parabens in HNSCC was constructed using STRING (v12.0; https://cn.string-db.org/) [71] under the following settings: organisms “Homo sapiens”, network type “full STRING network”, medium confidence score > 0.4, and medium FDR stringency (5%).

Identification of hub genes from the PPI network

The PPI network was imported into Cytoscape (v3.9.1) [72]. The MCODE plugin [73] was used to identify central clusters with default parameters (Degree Cutoff = 2, Haircut clustering, Node Score Cutoff = 0.2, K-Core = 2, and Max Depth = 100). Key targets from MCODE modules were compared with top-ranked targets from CytoHubba [74]. The top 20 genes from each of the 12 CytoHubba algorithms (Betweenness, BottleNeck, Closeness, ClusteringCoefficient, Degree, DMNC, EcCentricity, EPC, MCC, MNC, Radiality, and Stress) were selected [75]. An upset plot generated via ChiPlot identified the 20 most frequently occurring genes across all 12 algorithms. Final hub genes were defined as the intersection between MCODE-derived core targets and the top-ranked CytoHubba targets.

Expression validation of hub genes in HNSCC

Protein expression levels of hub genes were evaluated using the Human Protein Atlas (HPA) (https://www.proteinatlas.org/) [76] and the University of ALabama at Birmingham CANcer data analysis Portal (UALCAN) (https://ualcan.path.uab.edu/) [77,78], following established methodologies [79,80]. Data analyzed by UALCAN were generated by the National Cancer Institute Clinical Proteomic Tumor Analysis Consortium (CPTAC) [81–83]. Gene expression profiles were obtained from the Gene Expression Omnibus (GEO) Datasets (https://www.ncbi.nlm.nih.gov/gds) [84,85] under the filters: expression profiling by array, tissue attribute, Homo sapiens, and paired samples only. Statistical differences were assessed and visualized using GraphPad Prism v9.0.0 (San Diego, California USA, www.graphpad.com). Additional validation was performed using Gene Expression Profiling Interactive Analysis (GEPIA3) (https://gepia3.bioinfoliu.com/) [86] with |log2(fold change) ≥ 1, q-value < 0.05, and TCGA paired peritumor samples as normal controls.

Construction of a paraben exposure risk model for HNSCC

The association between hub gene expression and overall survival in HNSCC patients was evaluated using GEPIA3. Univariable and multivariate Cox models were applied using overall survival in months, hazard ratio (HR) calculation, median group cutoff, and HNSC Tumor datasets.

Enrichment analysis of hub genes

GO and KEGG enrichment analyses for the hub genes were performed using DAVID with default parameters as above-mentioned and results were visualized with ChiPlot.

Immune cells infiltration

The relationship between hub gene expression and tumor-infiltrating immune cell abundance was assessed using the TIMER2.0 (http://timer.cistrome.org/) [87–89]. Six computational algorithms (TIMER, XCell, MCP-COUNTER, CIBERSORT, EPIC, and QUANTISEQ) were employed to estimate immune infiltration levels. Partial Spearman’s correlation, adjusted for tumor purity, was used to evaluate correlations between gene expression and immune cell abundance. Correlations with ρ > 0 and p < 0.05 were considered significantly positive; those with ρ < 0 and p < 0.05 were deemed significantly negative; and results with p ≥ 0.05 were considered non-significant.

Molecular docking

Molecular docking was performed to evaluate binding affinities between core proteins and paraben. The three-dimensional structures of the target proteins were retrieved from the RCSB Protein Data Bank (PDB) (https://www.rcsb.org/) [90,91] for experimentally-determined structures and AlphaFold Protein Structure Database (https://alphafold.ebi.ac.uk/) [92,93] for computationally predict structures, while the paraben structures in SDF format were obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov) [48,94]. Subsequently, both the receptor and ligand files were imported into CB-DOCK2 (https://cadd.labshare.cn/cb-dock2/php/index.php) [95,96] for automated blind docking simulations.

Compound-target-pathway network

A Sankey diagram was constructed using ChiPlot to visualize the multi-layered relationships between the six parabens (source nodes), hub genes (middle nodes), and enriched KEGG pathways (target nodes). Linkage data defining the connections between these nodes were formatted into a source-target-weight matrix and uploaded to ChiPlot for visualization.

Results

Toxicity prediction and identification of paraben-related targets

Toxicity predictions for the six parabens were obtained from ProTox 3.0 and ADMETlab2.0 (retrieved 2025-7-28), revealing varying degrees of potential toxicity (S1 Table). A total of 582 unique human targets related to parabens were identified after merging and deduplicating results from Swiss Target Prediction (234 targets), ChEMBL (137 targets), and SEA (342 targets) (retrieved 2025-7-29). Heptylparaben and benzylparaben were not available in ChEMBL. A flowchart schematically illustrated the refined screening workflow (Fig 1).

thumbnail
Download:
Fig 1. Schematic flowchart of the integrated network toxicology and molecular docking study on paraben toxicity in HNSCC.

https://doi.org/10.1371/journal.pone.0333867.g001

Screening of HNSCC-related targets

HNSCC-related targets were identified from GeneCards (447 protein-coding genes after removing non-coding RNAs from 852 predicted targets), OMIM (1,352 targets), CTD (6,242 targets, including 44 with curated associations to HNSCC and 6,198 with inferred associations and inference scores above the average value of 7.11 (retrieved 2025-8-3). After merging and deduplication, 7,320 HNSCC-related targets were obtained.

Analysis of differential expressed genes in HNSCC from TCGA

Analysis of RNA-seq data from 504 HNSCC tumors and 44 matched normal samples from TCGA identified 2,250 DEGs (1,623 up-regulated, 627 down-regulated; adjusted p < 0.05, |log₂FC| ≥ 1) (retrieved 2025-8-4). The distribution of DEGs was shown in a volcano plot (S1A Fig), and the top 50 up- and down-regulated genes were presented in a heatmap (S1B Fig). In order to intuitively study the shared functions or characteristics of DEGs, GO and KEGG analyses were presented in a circle enrichment plot (S1C Fig). Furtherly, separate enrichment analyses for up- and down-regulated DEGs were presented in S1D–S1G Fig. Notably, the highest number of up-regulated DEGs (80 genes) were enriched in human papillomavirus infection pathway (hsa05165), consistent with known etiology of HPV-positive HNSCC (S1E Fig).

Identification of parabens-related targets in HNSCC and construction of PPI network

Intersection analysis identified 80 candidate targets through which parabens may exert toxic effects in HNSCC (Fig 2 and S2 Table). Then, enrichment analysis indicated that these targets were associated with 298 GO terms, comprising 181 Biological Process (BP), 56 Cellular Component (CC), 61 Molecular Function (MF). Analysis also shown that these targets enriched in 42 KEGG terms, and 120 REACTOME pathways. The top 10 terms ranked by p-value for each category were visualized in circle enrichment plots (Fig 3 and S3 Table) (retrieved 2025-8-7). The PPI network constructed from the 80 candidate targets contained 569 edges among 80 nodes, with an average node degree of 14.2 and an average local clustering coefficient of 0.634), indicating significant enrichment over random expectation (expected edges = 208, p < 1.0 × 10 −16) (S2 Fig) (retrieved 2025-8-7).

thumbnail
Download:
Fig 2. Identification of common candidate targets for paraben toxicity in HNSCC.

A three-way Venn diagram illustrates the overlap among three distinct datasets: 582 potential paraben-related targets from Swiss Target Prediction, ChEMBL, and the Similarity Ensemble Approach (SEA); 7,320 known HNSCC-associated genes from GeneCards, Online Mendelian Inheritance in Man (OMIM), and Comparative Toxicogenomics Database (CTD); and 2,250 differentially expressed genes (DEGs) between HNSCC tumors and normal tissues from The Cancer Genome Atlas (TCGA) cohort. The intersection of all three sets reveals 80 common candidate genes, which are hypothesized to be critically involved in the potential toxicological mechanisms of parabens on the development and progression of HNSCC.

https://doi.org/10.1371/journal.pone.0333867.g002

thumbnail
Download:
Fig 3. Functional enrichment analysis of differentially expressed genes (DEGs) in the TCGA-HNSCC cohort.

A circular enrichment plot illustrates the top 10 most significantly enriched terms (by p-value) from Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and REACTOME pathway analyses for the HNSCC-DEGs. In these plots: the outermost circle lists the significantly enriched term IDs, with different colors representing different categories. The second circle represents the total number of background genes that enriched in each item. The third circle represents the genes in DEGs list that enriched in the target term. The innermost circle represents the rich factor, which is the ratio of enriched genes to total background genes, reflecting the degree of enrichment of the target pathway.

https://doi.org/10.1371/journal.pone.0333867.g003

Identification of hub genes for parabens targeting HNSCC

MCODE analysis identified a core subnetwork with the highest score (19.905), comprising 22 nodes and 418 edges (Fig 4A). CytoHubba identified the top 20 genes from each of 12 algorithms. Further, an upset plot analysis revealed the top 20 most frequently occurring genes across all algorithms (Fig 4B). Among these genes, 12 genes (CCNB1, CDK1, CCNA2, CDK2, CDK4, TYMS, AURKA, CCNA1, CHEK1, CCNB2, PLK1, CDC25A) were located within the above-mentioned core network and were designated as hub genes (Fig 4).

thumbnail
Download:
Fig 4. Identification of hub genes via MCODE and CytoHubba analysis.

(A) A core protein-protein interaction (PPI) subnetwork was extracted using the MCODE algorithm, with the highest MCODE score (19.905), containing 22 nodes and 418 edges, indicating a highly interconnected protein cluster. (B) An upset plot shows the overlap of the top 20 genes ranked by each of the 12 CyroHubba algorithms. Horizontal bars represent the frequency of each gene across all methods. From this analysis, 12 hub genes (CCNB1, CDK1, CCNA2, CDK2, CDK4, TYMS, AURKA, CCNA1, CHEK1, CCNB2, PLK1, CDC25A) were not only highly ranked but were also all members of the core MCODE subnetwork, confirming their status as central hub genes for further investigation.

https://doi.org/10.1371/journal.pone.0333867.g004

Expression validation of hub genes in HNSCC

The protein expression levels and subcellular location of hub genes were assessed between oral mucosa and HNSCC tissues using immunohistochemistry in HPA database (retrieved 2025-8-8). Of the 12 hub proteins, 10 were retrieved in the HPA database, while CHEK1 and CDC25A were not included in it. Among the 10 proteins, the expression levels of CCNB1, CDK4, TYMS, and CCNA1 exhibited higher expression in HNSCC tissues compared to oral mucosa (S3 Fig). Besides, the hub proteins primarily localized in the cytoplasm and/or nucleus (S3 Fig). The protein expression levels of hub genes were further assessed by UALCAN (retrieved 2025-8-8). Results indicated that among the 12 proteins, nine proteins (CCNB1, CDK1, CCNA2, CDK2, TYMS, AURKA, CHEK1, CCNB2, PLK1) were over-expressed in HNSCC, while CDK4 showed no significant difference (S4 Fig). CCNA1 and CDC25A were not included in CPTAC data. Through search details: HNSCC [All Fields] AND “Homo sapiens”[porgn] AND (“Expression profiling by array”[Filter] AND “attribute name tissue”[Filter]) (retrieved 2025-8-9), the present study obtained 30 datasets. After reviewing the titles and abstracts of the 30 datasets, and focusing on paired-samples, GSE83519 (GPL4133, consist of 22 paired HNSCC tumor and normal tissues), GSE58911 (GPL6244, consist of 15 paired tumor and normal tissues) [97], and GSE160042 (GPL18180, consist of 10 paired tumor and normal tissues) [98] were used for validating the expression levels of the hub genes. The expression matrix of each GEO Datasets above-mentioned were downloaded, and expression values of each hub gene were extracted. Then, the difference of hub genes in HNSCC were identified with GraphPad Prism. The vast majority of hub genes were up-regulated in HNSCC (Fig 5). Additionally, GEPIA3 confirmed the overexpression of all 12 hub genes in HNSCC cancerous tissues compared to normal tissues (S5 Fig).

thumbnail
Download:
Fig 5. Validation of hub gene expression in HNSCC using independent GEO datasets.

The mRNA expression levels of the identified hub genes were analyzed in three independent HNSCC cohorts from the Gene Expression Omnibus (GEO) repository. (A) GSE83519 (22 paired tumor and adjacent normal tissues), Note: A constant value (C = 3.911890567) was added to the raw expression data prior to log₂ transformation to eliminate negative values. This constant value was calculated as |min (negative value)| + 0.001 to enable transformation of negative values; (B) GSE58911 (15 paired samples); (C) GSE160042 (10 paired samples). Differential expression between tumor and matched adjacent normal tissues is shown for each dataset. Statistical significance was determined by paired t-test (p < 0.05).

https://doi.org/10.1371/journal.pone.0333867.g005

Prognostic significance of hub genes for HNSCC

Univariable survival analysis using GEPIA3 (retrieved 2025-8-16) showed that only CCNA1 was significantly associated with patient survival (hazard ratio [HR] = 1.60, 95% confidence interval [CI]: 1.22–2.10, p = 0.001) (S6 Fig). The remaining 11 hub genes did not exhibit significant prognostic value (p > 0.05). Multivariable Cox regression model incorporating all 12 hub genes confirmed CCNA1 as an independent predictor of poor prognosis (adjusted HR = 1.097, 95% CI: 1.021–1.179, p = 0.012) (Fig 6). Notably, CCNA1 was the sole gene consistently significant in both analytical approaches, underscoring its strong and independent association with survival outcomes in this cohort.

thumbnail
Download:
Fig 6. Multivariable Cox regression analysis of the 12 hub genes for prognostic prediction in HNSCC.

A forest plot presents results of multivariable Cox proportional hazards regression model incorporating all 12 candidate hub genes. Each gene is shown with its hazard ratio (HR) and 95% confidence interval (CI). CCNA1 was identified as a significant independent predictor of poor overall survival in months (adjusted HR = 1.097, 95% CI: 1.021–1.179, p = 0.012). No other hub genes showed significant independent prognostic value.

https://doi.org/10.1371/journal.pone.0333867.g006

Enrichment analysis of the hub genes

Using DAVID database, functional enrichment analysis was performed on the 12 hub genes and yielded 67 significant GO terms (35 BP, 21 CC, 11 MF), 13 significant KEGG pathways, and 103 significant REACTOME pathways (S4 Table) (retrieved 2025-8-16). The top 10 enriched terms from each category were visualized in circular enrichment plots (Fig 7). These findings indicated that the 12 hub genes were implicated in a range of coherent and biologically relevant processes, supporting their potential functional synergy. For example, among the 12 hub genes that were over-expressed in HNSCC, 10 were enriched in cell cycle pathway (hsa04110), six in p53 signaling pathway (hsa04115), and six in viral carcinogenesis pathway (hsa05203), all of which were related to HNSCC.

thumbnail
Download:
Fig 7. Functional enrichment analysis of the 12 hub genes.

Circular enrichment plots visualize the top 10 most significantly enriched terms from (A) Gene Ontology (GO) and (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) and REACTOME pathway analyses for the 12 hub genes.

https://doi.org/10.1371/journal.pone.0333867.g007

Analysis of hub gene expression and immune cell infiltration

Analysis using TIMER2.0 (retrieved 2025-8-20) revealed correlations between the expression of the 12 hub genes and infiltration levels of six immune cell types (CD8 + T cells, CD4 + T cells, B cells, neutrophils macrophages, dendritic cells), which using partial Spearman’s correlation to control for potential confounding factors such as tumor purity and clinical covariates. The results demonstrated that all 12 hub genes exhibited varying degrees of positive or negative correlations with different immune cell populations (S7 Fig). Notably, multiple algorithms consistently revealed that CCNA1, which is significantly associated with overall survival of HNSCC patients, showed a negative correlation with the infiltration levels of CD8 + T cells and B cells (Fig 8).

thumbnail
Download:
Fig 8. Association between CCNA1 expression and immune cell infiltration levels in TCGA-HNSC data.

Correlation between CCNA1 expression (log2(TPM + 1)) and infiltration levels of CD8 + T cell and B cell was analyzed across TCGA-HNSC tumors (n = 522) using TIMER2.0. Infiltration levels were estimated with six algorithms (EPIC, MCP-COUNTER, TIMER, CIBERSORT, QUANTISEQ, and XCELL) and adjusted for tumor purity. The association was evaluated using partial Spearman’s correlation analysis. Spearman’s ρ > 0 and p < 0.05 indicates significant positive correlation; ρ < 0 and p < 0.05 indicates significant negative correlation; p > 0.05 indicates no significance. TPM: transcripts per million. The analysis was performed by TIMER2.0 based on TCGA-HNSC data and detailed analysis can be further obtained from TIMER2.0.

https://doi.org/10.1371/journal.pone.0333867.g008

Molecular docking between hub proteins and parabens

Molecular docking simulations using CB-DOCK2 demonstrated strong binding affinities between six paraben derivatives and the 12 hub proteins (Fig 9). Our analysis revealed that each paraben analogue engaged with multiple protein targets through specific hydrogen bonding and notable electrostatic interactions. Furthermore, the compounds consistently occupied hydrophobic regions within the target binding sites, suggesting stable complex formation. Notably, several interactions exhibited strong binding affinities, with energy values ranging between −8.1 and −9.2 kcal/mol, reflecting favorable and potentially physiologically relevant binding.

thumbnail
Download:
Fig 9. Molecular docking analysis of paraben derivatives binding to hub proteins.

A structure-based blind protein-ligand docking strategy was performed using CB-DOCK2 to predict binding affinities between six paraben derivatives and 12 hub proteins. Results are presented as a heatmap, with the binding energy scores (Vina score, in kcal/mol) for each compound-protein pair indicated numerically. More negative scores represent stronger predicted binding affinity.

https://doi.org/10.1371/journal.pone.0333867.g009

Construction of parabens-hub genes-pathways network

The Sankey diagram (Fig 10) illustrated the mechanistic flow from parabens exposure (left) through hub genes (center) to downstream functional pathways (right). The visualization confirmed the central role of the 12 hub genes as a convergent signaling layer, integrating inputs from multiple parabens and influencing critical biological pathways. For example, CCNA1 simultaneously mediated the exotic effects of benzyl-, butyl-, ethyl-, and methylparaben on pathways, including cell cycle, progesterone-mediated oocyte maturation, cellular senescence, viral carcinogenesis, human T-cell leukemia virus 1 infection, and Epstein-Barr virus infection. The Sankey diagram provides a powerful synthetic overview of the analysis, positing a coherent framework in which diverse paraben compounds modulated a core set of hub genes to perturb specific biological pathways.

thumbnail
Download:
Fig 10. Sankey diagram of paraben-hub gene-pathway interactions.

The diagram summarizes the proposed mechanistic links between paraben exposure (left), hub genes (middle), and enriched biological pathways (right), illustrating how parabens may influence HNSCC pathogenesis through core regulatory genes.

https://doi.org/10.1371/journal.pone.0333867.g010

Discussion

This study systematically elucidates the potential molecular mechanisms through which parabens may exert toxic effects in HNSCC using integrated network toxicology and molecular docking approaches. Our findings reveal that parabens target key regulatory genes involved in cell cycle progression, DNA damage response, and viral carcinogenesis pathways, suggesting potential roles of parabens in HNSCC pathogenesis.

Through scientific and rigorous screening methods, we identified 12 hub genes (CCNB1, CDK1, CCNA2, CDK2, CDK4, TYMS, AURKA, CCNA1, CHEK1, CCNB2, PLK1, CDC25A) that are central to the PPI network of paraben-related targets in HNSCC and are significantly overexpressed in HNSCC tumor tissues compared to normal controls. Notably, CCNA1 emerged as an independent prognostic factor associated with poor survival in HNSCC patients, underscoring its clinical relevance. Enrichment analyses further indicated that these hub genes are functionally clustered in critical oncogenic pathways such as cell cycle regulation, p53 signaling, and viral carcinogenesis, which are frequently dysregulated in HNSCC [99–104]. Previous studies have also demonstrated that parabens can exhibit carcinogenic effects through multiple mechanisms [1,12,21,23,25].

Molecular docking results demonstrated strong binding affinities between parabens and these hub proteins, with favorable energy values indicating stable interactions. This suggests that parabens may directly modulate the activity of cell cycle regulators and DNA damage sensors, potentially leading to dysregulated proliferation and genomic instability, which are hallmarks of cancer. Particularly, the convergence of multiple parabens on CCNA1 linked exposure to altered cell cycle control and poor patient outcomes. Additionally, our immune infiltration analysis indicated that CCNA1 expression negatively correlated with CD8 + T cell and B cell infiltration, implying a potential immunosuppressive role of parabens. This aligns with recent evidence that tumor cell-intrinsic factors can shape the immune microenvironment in HNSCC, thereby influencing disease progression and therapy response [105–107]. While the estrogenic activity of parabens has been extensively studied in hormone-dependent cancers [12,21–24], their effects in non-hormonal cancers like HNSCC remained poorly understood. Our study provides a novel framework suggesting that parabens may promote carcinogenesis through non-canonical, ER-independent mechanisms, primarily by disrupting cell cycle checkpoints and enhancing chromosomal instability.

Noticeably, a key strength of this study lies in its integrative and systematic approach, combining multiple bioinformatics platforms, toxicological prediction tools, and molecular docking simulations to propose a novel mechanism of paraben-induced carcinogenesis in HNSCC. The use of publicly available multi-omics data from TCGA, GEO, CPTAC, and HPA enhances the reproducibility and reliability of the findings. Furthermore, the convergence of results across different analytical methods, including PPI network analysis, survival modeling, immune correlation assessment, and molecular docking, strengthens the validity of the identified hub genes and pathways. However, several limitations must be acknowledged. First, HNSCC is a heterogeneous group of diseases encompassing distinct anatomical subsites (such as pharynx, larynx, oral cavity) with different etiologies and molecular profiles. Our analysis was conducted using aggregated HNSCC data from public databases, which limits the ability to perform precise subsite-specific stratification without compromising statistical power. Therefore, our findings represent common pathways across HNSCC subtypes but may not fully capture site-specific mechanisms. Future studies should utilize TCGA subsite-specific cohorts (such as laryngeal cancer data) to validate and refine these findings. Second, the predictions made herein are primarily computational and require experimental validation using in vitro and in vivo HNSCC models to confirm causal relationships. Third, the study relies on existing databases that may contain inherent biases in sample selection or annotation. Additionally, the exact physiological concentration of parabens required to elicit these effects in human tissues remains unclear, and potential confounding factors such as exposure to other environmental carcinogens were not accounted for. Future work should include dose-response experiments and prospective epidemiological studies to better quantify cancer risk associated with paraben exposure. Despite these limitations, this study provides a valuable theoretical framework and a prioritized list of candidate targets for subsequent investigations into the environmental toxicology of head and neck cancers.

Overall, this work provides a foundational hypothesis that parabens may contribute to HNSCC tumorigenesis by targeting core cell cycle genes and modulating immune responses. It highlights the need for further mechanistic studies and epidemiological investigations to clarify the role of paraben exposure in HNSCC etiology. These insights may eventually inform regulatory policies regarding the safe use of parabens in consumer products.

Conclusions

In conclusion, this study provides the first comprehensive computational evidence that parabens may promote the pathogenesis of HNSCC through direct interactions with core cell cycle regulators and key signaling pathways. By integrating network toxicology and molecular docking, we identified 12 hub genes, which mediated paraben-induced toxicity in HNSCC and involved immune suppression. These findings shift the paradigm of paraben toxicity from a primarily estrogen receptor-driven mechanism to a broader oncogenic model involving cell cycle disruption and tumor microenvironment modulation. However, these insights remain predictive and necessitate further validation through in vitro and in vivo models, as well as epidemiological studies assessing real-world exposure levels. Ultimately, this work underscores the potential public health relevance of paraben exposure in HNSCC and supports the need for more stringent regulatory evaluation of these ubiquitous environmental chemicals.

Supporting information

S1 Fig. Transcriptomic analysis of head and neck squamous cell carcinoma (HNSC) from TCGA.

(A) Volcano plot of differentially expressed genes (DEGs) between HNSCC tumors and normal tissues. Red points indicate significant DEGs (adjusted p < 0.05 and |log2 fold change| > 1); gray points indicate non-significant genes. (B) Heatmap of the top 100 most dysregulated genes (50 up- and 50 down-regulated). Columns represent samples (tumor vs. normal); rows represent genes clustered by expression. Expression Z-scores are shown from blue (low) to red (high). (C) Circular enrichment plot of DEGs. From outer to inner: term IDs; number of background genes per term; number of DEGs enriched per term; rich factor. (D, F) Top 15 significant GO terms (p < 0.05) for (D) up- and (F) down-regulated DEGs. Bar length indicates gene count; color indicates ontology. (E, G) Significant KEGG pathways (p < 0.05) for (E) up- and (G) down-regulated DEGs. Dot size indicates number of enriched genes; color indicated -log10 (adjusted p-value). All analyses were performed using the Home for Researchers platform (https://www.home-for-researchers.com/) based on TCGA-HNSC data.

https://doi.org/10.1371/journal.pone.0333867.s001

(TIF)

S2 Fig. Protein-protein interaction (PPI) network of paraben-related targets in HNSCC.

The PPI network was constructed from 80 paraben-related targets using STRING, resulting in 568 interactions. Nodes represent proteins; edges represent functional associations. Edge color indicates different known or predicted interactions. The network shows significant functional connectivity (PPI enrichment p < 1.0 × 10–16).

https://doi.org/10.1371/journal.pone.0333867.s002

(TIF)

S3 Fig. Protein expression and subcellular localization of hub genes in HNSCC.

Immunohistochemistry images and subcellular localization data for the hub genes were retrieved from the Human Protein Atlas (HPA) database. For each panel (A: CCNA1, B: CDK1, C: CCNA2, D: CDK2, E: CDK4, F: TYMS, G: AURKA, H: CCNA1, I: CCNB2, J: PLK1.), the left image shows target protein expression in normal tissue (oral mucosa), while the middle image shows expression in HNSC tissue (not detected, low, medium, or high). Subcellular localization of the corresponding core proteins in U2OS cells is displayed right each pair (the blue represents the target protein). In this analysis, CDC25A and CHEK1 expression were not included in HPA, subcellular localization of TYMS and PLK1 were also not included in it. All the detailed data can be acquired from HPA database.

https://doi.org/10.1371/journal.pone.0333867.s003

(TIF)

S4 Fig. Proteomic analysis of hub genes in HNSC via UALCAN-CPTAC.

Bar plots compare normalized protein expression levels between normal (n = 71) and HNSCC tumor (n = 108) samples from CPTAC data. Wilcoxon test was used; p < 0.05 was considered significant. CCNA1 and CDC25A were not available in the dataset.

https://doi.org/10.1371/journal.pone.0333867.s004

(TIF)

S5 Fig. mRNA expression levels of hub genes in HNSCC via GEPIA3.

The mRNA expression of the 12 hub genes was analyzed in HNSCC using the GEPIA3 database. Expression levels are displayed as log2(TPM + 1). Red box plots represent tumor tissues (n = 520), and green box plots represent normal tissues (n = 44). All 12 hub genes showed significantly elevated expression in tumor tissues compared to normal controls (p < 0.001, unpaired Student’s t-test).

https://doi.org/10.1371/journal.pone.0333867.s005

(TIF)

S6 Fig. Survival analysis of hub genes in HNSCC.

Overall survival analysis of 12 hub genes was performed using the GEPIA3 database based on TCGA-HNSC data. Gene expression was quantified as log2(TPM + 1). Cox proportional hazards regression analysis revealed that among all hub genes, only high expression of CCNA1 was significantly associated with poor overall survival (p < 0.05). Survival time is shown in months. The Kaplan-Meier curve demonstrates the stratification of patients based on CCNA1 expression levels (high vs. low). TPM: transcripts per million.

https://doi.org/10.1371/journal.pone.0333867.s006

(TIF)

S7 Fig. Correlation between hub gene expression and immune cell infiltration in HNSCC.

Analysis was performed using the TIMER2.0 database to evaluate the association between expression levels of 12 hub genes and the abundance of immune cell infiltration (CD8 + T cells, CD4 + T cells, and B cells) in HNSCC. A heatmap with numbers displays the purity-adjusted Spearman’s correlation coefficients (ρ-values) across multiple algorithms in HNSCC. Red indicates statistically significant positive correlation (ρ > 0, p < 0.05), light blue indicates statistically significant negative correlation (ρ < 0, p < 0.05), and light gray indicates no significant correlation (p > 0.05). Multiple algorithms consistently showed that CCNA1 expression was negatively correlated with CD8 + T cell and B cell infiltration levels.

https://doi.org/10.1371/journal.pone.0333867.s007

(TIF)

S1 Table. Toxicity prediction results of the selected parabens.

https://doi.org/10.1371/journal.pone.0333867.s008

(DOCX)

S2 Table. The results for Venn-diagram.

https://doi.org/10.1371/journal.pone.0333867.s009

(XLSX)

S3 Table. The top 10 terms of each category in GO and KEGG analyses.

https://doi.org/10.1371/journal.pone.0333867.s010

(XLSX)

S4 Table. Enrichment analyses of hub genes.

https://doi.org/10.1371/journal.pone.0333867.s011

(XLSX)

References

  1. 1. Chatterjee S, Adhikary S, Bhattacharya S, Chakraborty A, Dutta S, Roy D, et al. Parabens as the double-edged sword: understanding the benefits and potential health risks. Sci Total Environ. 2024;954:176547. pmid:39357765
  2. 2. Tachachartvanich P, Sangsuwan R, Navasumrit P, Ruchirawat M. Assessment of immunomodulatory effects of five commonly used parabens on human THP-1 derived macrophages: implications for ecological and human health impacts. Sci Total Environ. 2024;944:173823. pmid:38851341
  3. 3. Rodrigo L, Bressa C, Larrosa M, Ramírez V, Gil-Izquierdo Á, Sánchez-Muñoz C, et al. Associations between chronic exposure to bisphenols and parabens and gut microbiota in children. Environ Res. 2025;285(Pt 5):122643. pmid:40849015
  4. 4. Dai Y, Ding J, Wang Z, Zhang B, Guo Q, Guo J, et al. Biomonitoring, dietary predictors, drinking water ingestion of 33 endocrine disrupting chemicals in primary school children in China, towards an exposome approach. Environ Int. 2025;203:109725. pmid:40816046
  5. 5. Khelfi A, Andreescu S, Azzouz M. Assessment of endocrine-disrupting chemicals in bottled water: a cross-sectional study in Algiers and risk assessment in pregnant women. Water Res. 2025;287(Pt A):124301. pmid:40752387
  6. 6. Reimann B, De Ruyter T, Sleurs H, Rasking L, Verheyen L, Giesberts N, et al. Paraben exposures and satiety hormones in preschool children: an ENVIRONAGE study. Environ Res. 2025;285(Pt 1):122300. pmid:40617576
  7. 7. Liao Y, Li J, Li X, Zhu Z, Gao Y, Wei Y, et al. Assessment of human parabens exposure and health risk based on global biomonitoring data. Environ Res. 2025;285(Pt 2):122461. pmid:40721101
  8. 8. Gálvez-Ontiveros Y, González-Palacios P, Ramírez V, Monteagudo C, Samaniego-Sánchez C, Rivas A, et al. Hair as an indicator of prolonged paraben exposure and its relation to weight gain in a sample of Spanish children: a proof-of-concept study. Nutrients. 2025;17(9):1593. pmid:40362902
  9. 9. Zhang Z, Xiang M, Yang H, Guo W, Bai T, Huang R, et al. Exposure to phenols mixture, oxidative stress, and fasting blood glucose: association and potential mediation analyses. Free Radic Biol Med. 2025;234:169–77. pmid:40253013
  10. 10. Peterson AK, Zhu Y, Feng J, Calafat AM, Hedderson MM, Alexeeff SE, et al. Urinary concentrations of early and mid-pregnancy parabens and gestational diabetes: a nested case-control study within the PETALS cohort. Sci Total Environ. 2025;974:179253. pmid:40154079
  11. 11. Zheng Y, Zhang L, Xiang Q, Li J, Yao Y, Wu Y, et al. Associations of pharmaceutical and personal care product chemical exposure with thyroid nodules and thyroid homeostasis. J Environ Sci (China). 2025;157:21–34. pmid:40602876
  12. 12. Tapia JL, McDonough JC, Cauble EL, Gonzalez CG, Teteh DK, Treviño LS. Parabens promote protumorigenic effects in luminal breast cancer cell lines with diverse genetic ancestry. J Endocr Soc. 2023;7(8):bvad080. pmid:37409182
  13. 13. Chen Y, Zhao C, Zheng J, Su N, Ji H. Discovery of the mechanism of n-propylparaben-promoting the proliferation of human breast adenocarcinoma cells by activating human estrogen receptors via metabolomics analysis. Hum Exp Toxicol. 2023;42:9603271231171648. pmid:37121592
  14. 14. Tong JH, Elmore S, Huang S-S, Tachachartvanich P, Manz K, Pennell K, et al. Chronic exposure to low levels of parabens increases mammary cancer growth and metastasis in mice. Endocrinology. 2023;164(3):bqad007. pmid:36683225
  15. 15. Yoon LS, Binder AM, Pereira A, Calafat AM, Shepherd J, Corvalán C, et al. Variability in urinary phthalates, phenols, and parabens across childhood and relation to adolescent breast composition in Chilean girls. Environ Int. 2022;170:107586. pmid:36302292
  16. 16. Mogus JP, LaPlante CD, Bansal R, Matouskova K, Schneider BR, Daniele E, et al. Exposure to propylparaben during pregnancy and lactation induces long-term alterations to the mammary gland in mice. Endocrinology. 2021;162(6):bqab041. pmid:33724348
  17. 17. Tzouma Z, Dourou P, Diamanti A, Harizopoulou V, Papalexis P, Karampas G, et al. Associations between endocrine-disrupting chemical exposure and fertility outcomes: a decade of human epidemiological evidence. Life (Basel). 2025;15(7):993. pmid:40724496
  18. 18. Inderbinen SG, Kley M, Zogg M, Sellner M, Fischer A, Kędzierski J, et al. Activation of retinoic acid-related orphan receptor γ(t) by parabens and benzophenone UV-filters. Toxicology. 2022;471:153159. pmid:35337918
  19. 19. Wang Y-F, Wan Y-T, Tian Q, Yin Y, Xu J-W, Zhou L-Q. Melatonin and spermidine protect early mouse embryos from ethylparaben toxicity by restoring histone modifications and subcellular structures. Ecotoxicol Environ Saf. 2025;300:118444. pmid:40466325
  20. 20. Kim K, Hwang I, Jeung E-B. Prenatal and postnatal exposure to butylparaben induces neurodevelopmental disorders in mice offspring. J Physiol Pharmacol. 2025;76(3):10.26402/jpp.2025.3.05. pmid:40698787
  21. 21. Lin Y, Zhang Y, She J, Zhao R, Lin S, Zhang Y, et al. Novel insights into the causal relationship between endocrine-disrupting chemicals and breast cancer mediated by circulating metabolites. Environ Pollut. 2025;375:126349. pmid:40311737
  22. 22. Xiong Y, Li Z, Xiong X, Luo Z, Zhong K, Hu J, et al. Associations between phenol and paraben exposure and the risk of developing breast cancer in adult women: a cross-sectional study. Sci Rep. 2025;15(1):4038. pmid:39900803
  23. 23. Wang L, Chen L, Schlenk D, Li F, Liu J. Parabens promotes invasive properties of multiple human cells: a potential cancer-associated adverse outcome pathway. Sci Total Environ. 2024;926:172015. pmid:38547973
  24. 24. Gonzalez TL, Moos RK, Gersch CL, Johnson MD, Richardson RJ, Koch HM, et al. Metabolites of n-Butylparaben and iso-Butylparaben exhibit estrogenic properties in MCF-7 and T47D human breast cancer cell lines. Toxicol Sci. 2018;164(1):50–9. pmid:29945225
  25. 25. Amin MM, Tabatabaeian M, Chavoshani A, Amjadi E, Hashemi M, Ebrahimpour K, et al. Paraben content in adjacent normal-malignant breast tissues from women with breast cancer. Biomed Environ Sci. 2019;32(12):893–904. pmid:31918794
  26. 26. Neri I, MacCallum J, Di Lorenzo R, Russo G, Lynen F, Grumetto L. Into the toxicity potential of an array of parabens by biomimetic liquid chromatography, cell viability assessments and in silico predictions. Sci Total Environ. 2024;917:170461. pmid:38286290
  27. 27. Wróbel AM, Gregoraszczuk EŁ. Action of methyl-, propyl- and butylparaben on GPR30 gene and protein expression, cAMP levels and activation of ERK1/2 and PI3K/Akt signaling pathways in MCF-7 breast cancer cells and MCF-10A non-transformed breast epithelial cells. Toxicol Lett. 2015;238(2):110–6. pmid:26253279
  28. 28. Kim MJ, Kim C-H, Seo Y-J, An M-J, Lee J-H, Shin G-S, et al. Transcriptome dynamics of alternative splicing events revealed early phase of apoptosis induced by methylparaben in H1299 human lung carcinoma cells. Arch Toxicol. 2020;94(1):127–40. pmid:31745603
  29. 29. Shin G-S, Park Y, Kim J-Y, Kim C-H, An M-J, Lee H-M, et al. Propylparaben-induced endoplasmic reticulum stress triggers G2/M phase cell cycle arrest and initiates caspase-3-dependent apoptosis in human lung cells. Genes Genomics. 2025;47(2):223–33. pmid:39699851
  30. 30. Akintobi OA, Adedeji AO, Ogunlaja A, Ladokun OA, Ogunlaja OO. Occurrence and exposure assessment of antibiotics, parabens, and potentially toxic elements in groundwater sources from two rural communities in Nigeria. Environ Toxicol Chem. 2025;44(11):3336–48. pmid:40674573
  31. 31. Matwiejczuk N, Galicka A, Brzóska MM. Review of the safety of application of cosmetic products containing parabens. J Appl Toxicol. 2020;40(1):176–210. pmid:31903662
  32. 32. Ramalho A, Vale A, Carvalho F, Fernandes E, Freitas M. Parabens exposure and its impact on diabesity: a review. Toxicology. 2025;515:154125. pmid:40132785
  33. 33. Gálvez-Ontiveros Y, Moscoso-Ruiz I, Rodrigo L, Aguilera M, Rivas A, Zafra-Gómez A. Presence of parabens and bisphenols in food commonly consumed in Spain. Foods. 2021;10(1):92. pmid:33466450
  34. 34. Laor P, Limpanont Y, Phuanukoonnon S, Kho Y, Choi K, Kliengchuay W, et al. Urinary paraben concentrations among children from the northernmost of Thailand: cross-sectional study for exposure and health risks. Ecotoxicol Environ Saf. 2025;295:118172. pmid:40209348
  35. 35. Hong S, Kim O-J, Jung SK, Jeon HL, Kim S, Kil J. The exposure status of environmental chemicals in South Korea: the Korean National environmental health survey 2018-2020. Toxics. 2024;12(11):829. pmid:39591007
  36. 36. Johnson DE, Burtness B, Leemans CR, Lui VWY, Bauman JE, Grandis JR. Head and neck squamous cell carcinoma. Nat Rev Dis Primers. 2020;6(1):92. pmid:33243986
  37. 37. El-Bayoumy K, Christensen N, Broach J, Meyers C, Stairs D, Machtay M, et al. Current challenges and potential opportunities for interception and prevention of head and neck cancer. Carcinogenesis. 2025;46(2):bgaf025. pmid:40382784
  38. 38. Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. pmid:38572751
  39. 39. Barsouk A, Aluru JS, Rawla P, Saginala K, Barsouk A. Epidemiology, risk factors, and prevention of head and neck squamous cell carcinoma. Med Sci. 2023;11(2):Epub 20230613. pmid:37367741
  40. 40. Yin D, Yang Q, Li S, Peng Y, Zhang J, Xie Z, et al. ACSS2-TFEB axis acts as a critical regulator of the autophagic machinery in head and neck squamous cell carcinoma. Cell Death Dis. 2025;16(1):650. pmid:40858538
  41. 41. Mai Z, Zheng J, Lu Y, Lin P, Lin Y, Zheng Y, et al. METTL1-driven epitranscriptomic enhancement of TXNDC12 boosts c-Myc stability through USP5 in HNSCC. Exp Mol Med. 2025;57(8):1743–58. pmid:40750708
  42. 42. Lin X, Chen W, Li B, Zhao Z, Yu Z, Zhao X-Y, et al. Targeting intratumoral copper inhibits tumor progression via p62-Mediated EZH2 degradation and potentiates anti-PD-1 immunotherapy in oral squamous cell carcinoma. Adv Sci (Weinh). 2025;12(40):e17795. pmid:40719074
  43. 43. Benafsha C, Cohen L, Shimonov L, Goldbart R, Traitel T, Amar-Lewis E, et al. Targeted siRNA delivery using Cetuximab-conjugated starch for epidermal growth factor receptor-driven head and neck squamous cell carcinoma. Small Sci. 2025;5(7):2500073. pmid:40666030
  44. 44. Zhang X, Xu J, Wang X, Xu L, Zhang X, Wang Y, et al. PI3K-dependent GAB1/Erk phosphorylation renders head and neck squamous cell carcinoma sensitive to PI3Kα inhibitors. Cell Death Dis. 2025;16(1):457. pmid:40533463
  45. 45. Zhou J, He M, Zhao Q, Shi E, Wang H, Ponkshe V, et al. EGFR-mediated local invasiveness and response to Cetuximab in head and neck cancer. Mol Cancer. 2025;24(1):94. pmid:40121428
  46. 46. Xu Q, Cheng A, Li B, Tian X, Han Z, Feng Z. The small extracellular vesicle-mediated intercellular transformation of CXCR1Low to CXCR1High tumour cells promotes the progression of head and neck squamous cell carcinoma. J Extracell Vesicles. 2024;13(4):e12427. pmid:38545803
  47. 47. Arriola Benítez PC, Fusco M, Amorin R, Picón CR, Piccioni F, Victoria L, et al. Unraveling the role of tumor-infiltrating immune cells in head and neck squamous cell carcinoma: implications for antitumor immune responses and immunotherapy. Int J Mol Sci. 2025;26(13):6337. pmid:40650111
  48. 48. Kim S, Chen J, Cheng T, Gindulyte A, He J, He S, et al. PubChem 2025 update. Nucleic Acids Res. 2025;53(D1):D1516–25. pmid:39558165
  49. 49. Banerjee P, Kemmler E, Dunkel M, Preissner R. ProTox 3.0: a webserver for the prediction of toxicity of chemicals. Nucleic Acids Res. 2024;52(W1):W513–20. pmid:38647086
  50. 50. Xiong G, Wu Z, Yi J, Fu L, Yang Z, Hsieh C, et al. ADMETlab 2.0: an integrated online platform for accurate and comprehensive predictions of ADMET properties. Nucleic Acids Res. 2021;49(W1):W5–14. pmid:33893803
  51. 51. Daina A, Michielin O, Zoete V. SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Res. 2019;47(W1):W357–64. pmid:31106366
  52. 52. Zdrazil B, Felix E, Hunter F, Manners EJ, Blackshaw J, Corbett S, et al. The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods. Nucleic Acids Res. 2024;52(D1):D1180–92. pmid:37933841
  53. 53. Davies M, Nowotka M, Papadatos G, Dedman N, Gaulton A, Atkinson F, et al. ChEMBL web services: streamlining access to drug discovery data and utilities. Nucleic Acids Res. 2015;43(W1):W612-20. pmid:25883136
  54. 54. Keiser MJ, Roth BL, Armbruster BN, Ernsberger P, Irwin JJ, Shoichet BK. Relating protein pharmacology by ligand chemistry. Nat Biotechnol. 2007;25(2):197–206. pmid:17287757
  55. 55. UniProt Consortium. UniProt: the Universal Protein Knowledgebase in 2025. Nucleic Acids Res. 2025;53(D1):D609–17. pmid:39552041
  56. 56. Ahmad S, Jose da Costa Gonzales L, Bowler-Barnett EH, Rice DL, Kim M, Wijerathne S, et al. The UniProt website API: facilitating programmatic access to protein knowledge. Nucleic Acids Res. 2025;53(W1):W547–53. pmid:40331428
  57. 57. Stelzer G, Rosen N, Plaschkes I, Zimmerman S, Twik M, Fishilevich S, et al. The GeneCards suite: from gene data mining to disease genome sequence analyses. Curr Protoc Bioinformatics. 2016;54:1.30.1-1.30.33. pmid:27322403
  58. 58. Hamosh A, Amberger JS, Bocchini C, Scott AF, Rasmussen SA. Online Mendelian Inheritance in Man (OMIM®): Victor McKusick’s magnum opus. Am J Med Genet A. 2021;185(11):3259–65. pmid:34169650
  59. 59. Davis AP, Wiegers TC, Sciaky D, Barkalow F, Strong M, Wyatt B. Comparative Toxicogenomics Database’s 20th anniversary: update 2025. Nucleic Acids Res. 2025;53(D1):D1328–D134. pmid:39385618
  60. 60. Gao GF, Parker JS, Reynolds SM, Silva TC, Wang L-B, Zhou W, et al. Before and after: comparison of legacy and harmonized TCGA genomic data commons’ data. Cell Syst. 2019;9(1):24-34.e10. pmid:31344359
  61. 61. Carrot-Zhang J, Chambwe N, Damrauer JS, Knijnenburg TA, Robertson AG, Yau C. Comprehensive analysis of genetic ancestry and its molecular correlates in cancer. Cancer Cell. 2020;37(5):639–54. pmid:32396860
  62. 62. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. Nat Genet. 2000;25(1):25–9. pmid:10802651
  63. 63. Aleksander SA, Balhoff J, Carbon S, Cherry JM, Drabkin HJ, Ebert D. The Gene Ontology knowledgebase in 2023. Genetics. 2023;224(1). pmid:36866529
  64. 64. Kanehisa M, Furumichi M, Sato Y, Matsuura Y, Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53(D1):D672–7. pmid:39417505
  65. 65. Jin Z, Sato Y, Kawashima M, Kanehisa M. KEGG tools for classification and analysis of viral proteins. Protein Sci. 2023;32(12):e4820. pmid:37881892
  66. 66. Bardou P, Mariette J, Escudié F, Djemiel C, Klopp C. jvenn: an interactive Venn diagram viewer. BMC Bioinformatics. 2014;15(1):293. pmid:25176396
  67. 67. Milacic M, Beavers D, Conley P, Gong C, Gillespie M, Griss J, et al. The reactome pathway knowledgebase 2024. Nucleic Acids Res. 2024;52(D1):D672–8. pmid:37941124
  68. 68. Griss J, Viteri G, Sidiropoulos K, Nguyen V, Fabregat A, Hermjakob H. ReactomeGSA - efficient multi-omics comparative pathway analysis. Mol Cell Proteomics. 2020;19(12):2115–25. pmid:32907876
  69. 69. Sherman BT, Hao M, Qiu J, Jiao X, Baseler MW, Lane HC, et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res. 2022;50(W1):W216–w21. pmid:35325185
  70. 70. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57. pmid:19131956
  71. 71. Szklarczyk D, Kirsch R, Koutrouli M, Nastou K, Mehryary F, Hachilif R, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(D1):D638–46. pmid:36370105
  72. 72. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. pmid:14597658
  73. 73. Bader GD, Hogue CWV. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003;4:2. pmid:12525261
  74. 74. Chin C-H, Chen S-H, Wu H-H, Ho C-W, Ko M-T, Lin C-Y. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol. 2014;8 Suppl 4(Suppl 4):S11. pmid:25521941
  75. 75. Wang L, Zong L, Cao D, Guan M. Unraveling the molecular mechanisms of PFOA in clear cell renal cell carcinoma through network toxicology and molecular docking strategies. Int J Surg. 2025;111(7):4842–53. pmid:40359556
  76. 76. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, et al. Proteomics. Tissue-based map of the human proteome. Science. 2015;347(6220):1260419. pmid:25613900
  77. 77. Chandrashekar DS, Karthikeyan SK, Korla PK, Patel H, Shovon AR, Athar M, et al. UALCAN: an update to the integrated cancer data analysis platform. Neoplasia. 2022;25:18–27. pmid:35078134
  78. 78. Chandrashekar DS, Bashel B, Balasubramanya SAH, Creighton CJ, Ponce-Rodriguez I, Chakravarthi B, et al. UALCAN: a portal for facilitating tumor subgroup gene expression and survival analyses. Neoplasia. 2017;19(8):649–58. pmid:28732212
  79. 79. Song S, Huang L, Zhou X, Yu J. Exploring the toxicological network in diabetic microvascular disease. Int J Surg. 2025;111(6):3895–907. pmid:40214234
  80. 80. Chen F, Zhang Y, Chandrashekar DS, Varambally S, Creighton CJ. Global impact of somatic structural variation on the cancer proteome. Nat Commun. 2023;14(1):5637. pmid:37704602
  81. 81. Ellis MJ, Gillette M, Carr SA, Paulovich AG, Smith RD, Rodland KK, et al. Connecting genomic alterations to cancer biology with proteomics: the NCI Clinical Proteomic Tumor Analysis Consortium. Cancer Discov. 2013;3(10):1108–12. pmid:24124232
  82. 82. Huang C, Chen L, Savage SR, Eguez RV, Dou Y, Li Y, et al. Proteogenomic insights into the biology and treatment of HPV-negative head and neck squamous cell carcinoma. Cancer Cell. 2021;39(3):361-379.e16. pmid:33417831
  83. 83. Wu C-S, Li H-P, Hsieh C-H, Lin Y-T, Yi-Feng Chang I, Chung A-K, et al. Integrated multi-omics analyses of oral squamous cell carcinoma reveal precision patient stratification and personalized treatment strategies. Cancer Lett. 2025;614:217482. pmid:39842500
  84. 84. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30(1):207–10. pmid:11752295
  85. 85. Clough E, Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, et al. NCBI GEO: archive for gene expression and epigenomics data sets: 23-year update. Nucleic Acids Res. 2024;52(D1):D138–44. pmid:37933855
  86. 86. Kang Y-J, Pan L, Liu Y, Rong Z, Liu J, Liu F. GEPIA3: enhanced drug sensitivity and interaction network analysis for cancer research. Nucleic Acids Res. 2025;53(W1):W283–90. pmid:40396370
  87. 87. Li T, Fu J, Zeng Z, Cohen D, Li J, Chen Q, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509–14. pmid:32442275
  88. 88. Li T, Fan J, Wang B, Traugh N, Chen Q, Liu JS, et al. TIMER: a web server for comprehensive analysis of tumor-infiltrating immune cells. Cancer Res. 2017;77(21):e108–10. pmid:29092952
  89. 89. Li B, Severson E, Pignon J-C, Zhao H, Li T, Novak J, et al. Comprehensive analyses of tumor immunity: implications for cancer immunotherapy. Genome Biol. 2016;17(1):174. pmid:27549193
  90. 90. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The protein data bank. Nucleic Acids Res. 2000;28(1):235–42. pmid:10592235
  91. 91. Burley SK, Bhatt R, Bhikadiya C, Bi C, Biester A, Biswas P, et al. Updated resources for exploring experimentally-determined PDB structures and Computed Structure Models at the RCSB Protein Data Bank. Nucleic Acids Res. 2025;53(D1):D564–74. pmid:39607707
  92. 92. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–9. pmid:34265844
  93. 93. Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, et al. AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res. 2022;50(D1):D439–44. pmid:34791371
  94. 94. Kim S, Bolton EE, Bryant SH. PubChem3D: conformer ensemble accuracy. J Cheminform. 2013;5(1):1. pmid:23289532
  95. 95. Liu Y, Yang X, Gan J, Chen S, Xiao Z-X, Cao Y. CB-Dock2: improved protein-ligand blind docking by integrating cavity detection, docking and homologous template fitting. Nucleic Acids Res. 2022;50(W1):W159–64. pmid:35609983
  96. 96. Yang X, Liu Y, Gan J, Xiao Z-X, Cao Y. FitDock: protein-ligand docking by template fitting. Brief Bioinform. 2022;23(3):bbac087. pmid:35289358
  97. 97. Lobert S, Graichen ME, Hamilton RD, Pitman KT, Garrett MR, Hicks C, et al. Prognostic biomarkers for HNSCC using quantitative real-time PCR and microarray analysis: β-tubulin isotypes and the p53 interactome. Cytoskeleton (Hoboken). 2014;71(11):628–37. pmid:25355403
  98. 98. Zhuang Z, Huang J, Wang W, Wang C, Yu P, Hu J, et al. Down-regulation of long non-coding RNA TINCR induces cell dedifferentiation and predicts progression in oral squamous cell carcinoma. Front Oncol. 2021;10:624752. pmid:33732637
  99. 99. Hmedat ANA, Doondeea J, Ebner D, Feller SM, Lewitzky M. The Src family kinase inhibitor drug Dasatinib and glucocorticoids display synergistic activity against tongue squamous cell carcinoma and reduce MET kinase activity. Cell Commun Signal. 2025;23(1):293. pmid:40537792
  100. 100. Zhou X, Liu J, Chen X, Zhou X, Xu B, Gan G, et al. S1PR3 inhibition impairs cell cycle checkpoint via the AKT/WEE1 pathway in oral squamous cell carcinoma. J Transl Med. 2025;23(1):573. pmid:40410851
  101. 101. Funk AL, Katerji M, Afifi M, Nyswaner K, Woodroofe CC, Edwards ZC, et al. Targeting c-MYC and gain-of-function p53 through inhibition or degradation of the kinase LZK suppresses the growth of HNSCC tumors. Sci Signal. 2025;18(873):eado2857. pmid:39933019
  102. 102. Wang Y, Chang F, Li Z, Duan C, Sun X, Wang S, et al. CircTP53/USP10/p53 signaling axis as a novel regulator of progression and prognosis of head and neck squamous cell carcinoma. Adv Sci (Weinh). 2025;12(30):e14961. pmid:40464594
  103. 103. Roy S, Miyauchi S, Jones RN, Kim SS, Venuti A, Boparai AS, et al. HPV E5 mediates resistance to chemotherapy and radiation by promoting a stem cell-like phenotype in head and neck squamous cell carcinoma. Cancer Res. 2025;85(20):3875–93. pmid:40838991
  104. 104. Payaradka R, Ramesh PS, Vyas R, Patil P, Rajendra VK, Kumar M, et al. Oncogenic viruses as etiological risk factors for head and neck cancers: An overview on prevalence, mechanism of infection and clinical relevance. Arch Oral Biol. 2022;143:105526. pmid:36049431
  105. 105. Choi S-W, Kim JH, Hong J, Kwon M. Mapping immunotherapy potential: spatial transcriptomics in the unraveling of tumor-immune microenvironments in head and neck squamous cell carcinoma. Front Immunol. 2025;16:1568590. pmid:40264779
  106. 106. Jost T, Wachter M, Meidenbauer J, Fietkau R, Gaipl US. Inhibiting the DNA damage repair of HNSCC cells in combination with normo-fractionated radiotherapy influences clonogenicity, senescence and expression of NK cell activation markers. Sci Rep. 2025;15(1):31827. pmid:40883442
  107. 107. Hu Y, Li Y, Jiang C, Han W, Chen X, Wang P, et al. DUSP9 is up-regulated and promotes tumor progression in head and neck squamous cell carcinoma. J Cancer. 2025;16(11):3403–14. pmid:40861803