scientific reports
A
Check for updates
OPEN Adrenocortical carcinoma survival gene HMMR was identified as being targeted by fluorouracil and epirubicin using a gene coexpression network-based drug repositioning strategy
Zahra Jafari, Seyed-Morteza Javadirad& & Seyede Elmira Yazdi Rouholamini
Adrenocortical carcinoma (ACC), with poor prognosis, is one of the most aggressive endocrine cancers. Surgery is the mainstay of treatment; nevertheless, chemotherapy is usually needed. Mitotane, the medication licenced for ACC, has had mixed results. Drug repositioning based on gene coexpression networks has proven to be an effective method of discovering potential cancer treatments. A total of 139 human specimens were examined, comprising 106 metastatic ACCs, 14 benign adrenocortical adenomas (ACAs), and 19 normal adrenal cortex tissues. Hub genes were identified using weighted correlation network analysis (WGCNA), and their differential expression was confirmed by microarrays and RNA sequencing. Hub genes were analyzed in more depth for enrichment, coexpression, and protein-protein interaction. Two phases of survival analysis were conducted considering sex. Hub gene expression was associated with TP53 mutations and cancer stage. Hub genes were evaluated using regression analysis. Hub genes were associated with immune cell infiltration. Network analysis identified transcription factors that regulate hub genes. Drug-gene interaction network analyses were performed to reposition existing drugs. WGCNA-PPI analysis revealed 31 hub genes, 11 of which were overexpressed by more than fourfold, as well as HMMR and UBE2T, which were novel genes. All hubs showed significant correlations with survival, tumor staging, and TP53 mutation status in ACC tissues. Women overexpressing CDK1, UBE2C, PCLAF, and CCNB1 were more likely to suffer catastrophic deaths than men. In terms of ACC diagnostic capacity, all hub genes had AUCs greater than 0.90, with TOP2A, CDK1, and CCNB1 having the highest values. All hub genes, except for THYMS and RACGAP1, were negatively correlated with M2 macrophages and CD8 +T cells infiltration. Hub genes were co- expressed and regulated by 21 DE-TFs expressed differently between ACC and normal tissues. Novel pharmaceuticals are represented by fifteen out of thirty-four medications directed at hub genes and DE-TFs. Paclitaxel and cisplatin were the central nodes within the drug-gene network. Multiple drugs target TYMS and TOP2A, making them both promising targets for ACC. Fluorouracil and epirubicin target HMMR novel hub gene. HMMR as a novel ACC hub gene targets by fluorouracil and epirubicin, but fluorouracil has advantageous. The reason for this is that M2 macrophages promote fluorouracil resistance, whereas tumors overexpressing the HMMR gene demonstrate a negative infiltration of macrophages. Hub genes, associated with ACC development and progression, have negative impacts on survival rates, particularly in women. Network analysis shows fluorouracil and epirubicin target the ACC novel hub gene, HMMR. Fluorouracil was more effective due to its impact on M2 macrophage infiltration.
Keywords Adrenocortical carcinoma, WGCNA, Survival, HMMR, Fluorouracil
Department of Cell and Molecular Biology and Microbiology, Faculty of Biological Science and Technology, University of Isfahan, Isfahan 81746-73441, Iran. ~ email: javadirad@yahoo.com; sm.javadirad@sci.ui.ac.ir
| Abbreviations | |
|---|---|
| ACC | Adrenocortical carcinoma |
| ACA | Adrenocortical adenoma |
| WGCNA | Weighted gene co-expression network analysis |
| DEGs | Differentially expressed genes |
| DE-TFs | Differentially expressed transcription factors |
| KEGG | Kyoto encyclopedia of genes and genomes |
| PPI network | Protein-protein interaction network |
| GO | Gene ontology |
| BP | Biological process |
| MF | Molecular function |
| CC | Cellular component |
| CCNB1 | Cyclin B1 |
| UBE2T | Ubiquitin conjugating enzyme E2 T |
| CDK1 | Cyclin dependent kinase 1 |
| ZWINT | ZW10 interacting kinetochore protein |
| RACGAP1 | RAC GTPase activating protein 1 |
| TOP2A | DNA topoisomerase II alpha |
| UBE2C | Ubiquitin conjugating enzyme E2 C |
| HMMR | Hyaluronan mediated motility receptor |
| PRC1 | Protein regulator of cytokinesis 1 |
| TYMS | Thymidylate synthetase |
| PCLAF | PCNA clamp associated factor |
| HR | Hazard ratio |
Adrenocortical carcinoma (ACC) is a highly lethal neoplasm of the adrenal gland, characterized by an unfavorable prognosis due to an insufficient understanding of its pathogenesis1. The primary treatment for ACC is surgical intervention, which yields the most favorable outcomes; however, recurrences and metastases are frequently inevitable. Surgery alone may be inadequate due to the advanced stage of the disease at presentation, necessitating the use of chemotherapy2. Efforts to translate the molecular foundation of ACC into effective treatments have historically yielded limited success. Mitotane, the only medication approved for this indication since 1960, is often poorly tolerated by numerous patients3. Effective diagnosis, treatment, and prevention of ACC belligerence require the identification of central (hub) genes responsible for this aggressive phenotype, as well as the targeting of the disease with drugs that interact with these genes. Gene coexpression network-based drug repositioning has demonstrated efficacy in identifying potential pharmaceuticals for cancer therapy and precision medicine4.
Drug repositioning strategies aim to identify novel applications for both approved and investigational pharmaceuticals5. It is advantageous as it lowers drug development expenses, accelerates the approval process, and minimizes employee turnover6. Additionally, the accessibility of extensive biological data and the swift advancement of bioinformatics facilitate drug repositioning, allowing for expedited drug development processes and reduced costs7. Acknowledging that drug repositioning represents a strategic approach with minimal risks is essential, as repositioned candidates typically progress through various stages of clinical trials, leading to a comprehensive understanding of their safety and pharmacokinetic properties8. Therefore, drug repositioning may represent a more effective strategy for drug development regarding the risk-reward trade-off compared to alternative approaches.
Computational drug repositioning methods depend on structural bioinformatics techniques, such as molecular modeling, dynamic simulations, and docking9. The primary limitation of structural bioinformatics methods is the insufficient understanding of the chemical structure of both the drug and its target10. Transcriptomics-based strategies for medication repositioning have emerged to address this limitation, shifting away from the traditional emphasis on drug binding sites. Molecular-level strategies rely on disease-related transcribed genes that are integral to pathomechanisms and serve as targets for specific medications10. To our knowledge, these strategies have not been previously applied to ACC, and the identification of novel candidate drugs for ACC treatment remains a significant challenge.
Weighted gene co-expression network analysis (WGCNA) is a sophisticated method for identifying novel hub genes from omics data11,12. WGCNA applied to transcriptomics for drug repositioning commences with the identification of differentially expressed genes (DEGs). The collapseRow function has enabled the adaptation of WGCNA for aggregating gene expression data. The gene sets, derived from WGCNA modules, were analyzed using correlation network-based methods13, and potential hub genes were further evaluated through the analysis of protein-protein interaction (PPI) networks focusing on centrality14. A survival analysis of hub genes was performed alongside an examination of the interactions between tumor cells and the immune system. Hub genes were integrated with their corresponding transcription factors (TFs) to investigate the interactions between drugs and genes.
WGCNA identified 11 co-expressed hub genes with differential expression levels at various stages of ACC. The hub genes exhibited a significant association with overall survival rates in ACC patients. Their significant diagnostic potential is evidenced by superior AUC values. The expression of hub genes exhibited a correlation with mutations in TP53. A negative correlation was observed between hub gene expression and immune cell populations, specifically CD8 + T cells and M2 macrophages. The analysis of the regulatory network, alongside enrichment analysis, demonstrated that hub genes and their related transcription factors play a role in
tumorigenesis via cancer-associated pathways. Fluorouracil and epirubicin were identified for the first time as agents targeting ACC through the HMMR hub gene.
Materials and methods Microarray data collection and preprocessing
This study employed the NCBI-GEO database to identify microarray datasets pertinent to ACC in Homo sapiens. Initially, fourteen datasets were identified; however, several were excluded due to insufficient phenotype information, relevance to pediatrics, inclusion of lncRNA expression, or exposure to specific treatments. Five datasets comprise adult ACC tissues: GSE143383, GSE28476, GSE12368, GSE90713, and GSE19776. Table S1 provides a comprehensive overview of the datasets. Log2 transformation and quantile normalization were employed to minimize technical variability. Outlier detection and removal were conducted using hierarchical clustering and principal component analysis (PCA). The cleaned datasets comprised of 139 human specimens, including 106 metastatic ACCs, 14 benign adrenocortical adenomas (ACAs), and 19 normal adrenal cortex samples.
WGCNA metadata analysis, probe intersection, and module detection
Cleaned microarray datasets were consolidated into a metadata set using the WGCNA package (version 1.70- 3), and hub genes were identified by our developed R-script12. Benign ACA tissues were excluded from the WGCNA analysis. Using the “MaxMean” approach, the collapseRows function aggregated identical probes to yield a single gene measurement. NCBI Gene-ID was utilized for gene annotation. Two correlation analyses were conducted to compare the datasets: one focused on average gene expression and the other on overall gene connectivity. A significant Pearson correlation was observed among the datasets. Using a power of 10, signed adjacency matrices were generated for the top 5000 genes. Subsequently, a topological overlap matrix (TOM) was computed, followed by hierarchical cluster analysis to categorize TOM-based dissimilarities13. A hybrid adaptive method, specifically cutreeHybrid, was employed to prune clusters with a cut height value of 0.99. The open-source programming language R, version 4.1.2, was utilized for statistical computing, data analysis, and data visualization (R Foundation for Statistical Computing, Vienna, Austria).
Reference dataset selection
Network modules were preserved by considering a small number of large modules. The reference dataset is characterized by the highest number of samples and modules. The Z-score was calculated using the ModulePreservation function to identify the most effective module. To assess the correlation between a gene and a module eigengene, module membership (kME) values were calculated. For the selection of eigengenes from the best module (with the highest Z score), correlation studies were conducted between each dataset and the reference dataset.
Identification of hub genes
The top 100 genes from the best module have been selected for further analysis. A PPI network was constructed using String with a minimum interaction value of 0.4, representing medium confidence. To identify the most significant nodes in the network, eigenvector centrality was determined using Gephi software (version 0.1). Genes with eigenvector values greater than 0.6 were selected as hub genes.
Analyzing the expression of hub genes using microarray metadata
A meta-data set comprising five previously defined microarray datasets was constructed. This part of the study incorporated ACA tissues to assess the expression profiles of benign tissues alongside malignant tissues. Differential expression analysis of the meta-data was conducted to verify the reliability of the WGCNA output. The SVA and Limma packages were utilized to mitigate batch effects following log2/quantile normalization. Hierarchical clustering and PCA analyses confirmed the removal of batch effects. Bayesian methods were employed to moderate the standard errors of estimated log-fold changes. P-values were adjusted using the Benjamini-Hochberg correction. The open-source programming language R, version 4.1.2, was utilized for statistical computing, data analysis, and data visualization (R Foundation for Statistical Computing, Vienna, Austria).
Gene expression profiling interactive analysis (GEPIA) of hub genes using RNAseq data
The GEPIA database (http://gepia.cancer-pku.cn/detail.php) was utilized to verify the differentially expressed hub genes between tumor and normal samples. Hub gene expression was evaluated using RNAseq data obtained from various sources, including the Genomic Data Commons data portal (GDC, https://portal.gdc.cancer.gov) and the Genotype-Tissue Expression databanks (GTEX, https://gtexportal.org).
Co-expression analysis of hub genes
Hub genes co-expression network analysis was conducted utilizing the Hmisc (version 5.1-1) package. Cluster analysis was conducted utilizing the complete method following the computation of Euclidean distance. The open-source programming language R, version 4.1.2, (R Foundation for Statistical Computing, Vienna, Austria), was employed for data preparation, clustering, and the generation of heatmaps.
Enrichment analysis of hub genes
The ToppGene database (https://toppgene.cchmc.org/) was utilized to elucidate the biological significance of highly co-expressed hub genes. Functional annotation was conducted using Gene Ontology (GO), while pathway detection was executed through the Kyoto Encyclopedia of Genes and Genomes (KEGG)15. Gene
Ontology (GO) encompasses three biological concepts: Biological Process (BP), Molecular Function (MF), and Cellular Component (CC). The data was graphically represented utilizing the GOplot (version 1.0.2) and data. table packages (version 1.17.2). Statistical significance is indicated by p-values that are below 0.05.
Survival analysis of hub genes
ACC gene expression data were acquired from the GDC data portal and normalized using the transcripts per million method. A total of 79 ACC tissues were collected, and patients were categorized into groups according to the median expression levels of target genes. The log-rank test was employed to evaluate statistical significance. Survival plots were presented both with and without consideration of patient sex. Kaplan-Meier co-expression curves were employed to determine the hazard ratio (HR) derived from predicted survival probabilities. The open-source programming language R, version 4.1.2, was utilized for data analysis and visualization (R Foundation for Statistical Computing, Vienna, Austria).
Analyses of hub genes based on ACC stages and TP53 mutations
Expression levels of hub genes in ACC tumors across various pathological stages were obtained from the GDC portal. A total of 9 ACC tissues were identified at stage I, 44 at stage II, 19 at stage III, and 18 at stage IV. The TP53 mutation status of ACC tissues was obtained from the GDC portal. Patients were categorized into groups according to median gene expression levels. The log-rank test was employed to evaluate statistical significance. P-values below 0.05 were considered significant. The plots were created utilizing the open-source programming language R, version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria).
Receiver operating characteristic (ROC) curve analysis of hub genes
The diagnostic potential of hub genes was assessed through ROC analyses. The Mann-Whitney test was utilized to assess the significance of the area under the ROC curve (AUC) value. Statistical analysis was conducted using GraphPad Prism version 9.5.1 (GraphPad Software, Boston, Massachusetts, USA).
Immune cell infiltration analysis of hub genes
TIMER 2.0 (http://timer.cistrome.org) was utilized to quantify the presence of immune cells across various tumor types. The association between ACC hub genes and immune cell infiltration was examined using Pearson correlation analysis. CD8+T cells and M2 macrophages were represented in the landscape of tissue cellular heterogeneity utilizing the xCell method16.
TF network analysis of hub genes
The Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST) database (https://www.grnpedia.org/trrust/) was utilized to identify TFs with the potential to regulate hub genes. Cytoscape version 3.10.0 was utilized to construct and visualize a regulatory network depicting the interactions of TFs-hub genes17.
Drug gene interaction network analysis
The drug gene interaction database (DGIdb4.0)18 was utilized to identify potential pharmaceuticals that target hub genes and their regulatory transcription factors. The database contains approximately 15,000 drugs, facilitating precise discoveries of drug-gene interactions and the identification of druggable genes. This study focused exclusively on experimentally validated drug-gene interactions.
Results
Microarray datasets were cleaned and validated
Log2 transformation and quantile normalization were implemented on the microarray row data sections that lacked log2 transformation, resulting in homogeneity of variance. Hierarchical clustering and PCA identified 92 out of 231 samples as outliers (Figure S1), which were subsequently removed based on advanced biological expertise.
WGCNA identified eigengenes
All five datasets exhibited significant Pearson correlations among themselves (Figure S2). The analysis necessitated the exclusion of ACA samples when comparing ACC samples with normal samples; consequently, GSE28476 was entirely omitted from the study. Gene-tree clusters represented genes exhibiting analogous topological overlaps or co-expression patterns across the four remaining datasets (Figure S3). A limited number of large modules were considered, with GSE90713, comprising 34 modules, identified as the reference dataset. Analysis of module preservation statistics between GSE90713 and independent datasets revealed that the purple module exhibited the highest mean Z-score of 20.73, indicating its informative nature. After qualifying the module for the reference dataset (Figure S4), quantification indicated that the purple module contained the highest gene count (Figure S5). In the final step, 100 out of 129 genes from the purple modules were selected for subsequent analyses due to their high kME values (Table S2).
Thirty-one genes were selected based on eigenvector centrality
A PPI network comprising the top 100 genes in the purple module, featuring 707 nodes and 88 edges (Fig. 1). Based on the eigenvector centrality values, 31 genes had more influence (Table S3), with CDK1 demonstrating the greatest eigenvector centrality of 1.0, whilst H2AZ1 displayed the lowest eigenvector centrality of 0.62.
TEX10
DCP2
EIF4E2
SART3
FIGNLI
KHDRBST
SNRPF
CCDC167
DROSHA
SENP2
CTDSPL2
KPNA4
ANP32B
PA2G4
TARDBP
SMC1A
SNRPAI
SNRPDI
RAMAC
PAXBPI
SAE1
SMC2
DNAJC9
RAN
FBL
ILF3
SNRPD2
CPSF6
SLC2A1
FAFI
CCT5
FEN1
SUPT16H
SAC3DI
SKA2
PHIP
ATAD2
SMC4
HMGN2
NRAS
CASP2
NASP
MZT1
JPT2
MCM7
BIRC5
TOP2A
ZWILCH
PRRC2C
STMN1
MCM3
HMGB2
TMPO
RFC4
CDK1
CSEIL
ZWINT
PTBP3
CCNB1
H2AZ1
PSMB7
GGH
TYMS
CCNB2
ANP32E
NUSAP1
PRC1
MAD2L1
GMNN
RC3H2
BDP1
UBE2C
CENPN
AURKA
MCM4
PCNA
WEE1
PCLAF
CKS1B
VNSIABP
CKS2
HMMR
MTF2
RCC2
UBE2T
RACGAPI
EED
CENPW
Eleven hub genes identified after microarray and RNAseq data analysis
Analysis of microarray meta-data identified 6,654 DEGs in ACC tissues relative to normal tissues. Malignant ACC tissues exhibited overexpression of 30 of the 31 PPI-derived genes (Table 1). Cyclin B1 (CCNB1) was identified as the gene with the highest differential expression between ACC and normal tissues, exhibiting a log fold change of 2.8443. DNA topoisomerase II alpha (TOP2A) and ubiquitin conjugating enzyme E2C (UBE2C) were positioned second and third, exhibiting Log FCs of 2.7668 and 2.4895, respectively. In benign ACA tissues, 20 of 31 PPI-derived genes exhibited upregulation relative to healthy tissues. A comparison of malignant and benign ACC tissues revealed that 12 genes were exclusively expressed in malignant tissues. The genes include CCNB1, TOP2A, UBE2C, cyclin dependent kinase 1 (CDK1), ZW10 interacting kinetochore protein (ZWINT), rac GTPase activating protein 1 (RACGAP1), ubiquitin conjugating enzyme E2 T (UBE2T), hyaluronan mediated motility receptor (HMMR), protein regulator of cytokinesis 1 (PRC1), mitotic arrest deficient 2-like protein 1 (MAD2L1), baculoviral IAP repeat containing 5 (BIRC5), and centromere protein N (CENPN).
Eleven of the thirty PPI-derived genes with logFC values exceeding two exhibited significant differences between malignant and normal ACC tissues (Fig. 2a). The list comprises CCNB1, UBE2T, CDK1, ZWINT, RACGAP1, TOP2A, UBE2C, HMMR, PRC1, thymidylate synthetase (TYMS), and PCNA clamp associated factor (PCLAF). The identified genes were designated as hub genes, and RNAseq data analysis revealed that all hub genes exhibited significant upregulation in ACC compared to normal tissues (Fig. 2b-1).
Hub genes were co-expressed in the ACC
The corrplot indicated that all hub genes exhibited correlation values exceeding 68% (Fig. 3). RACGAP1 exhibited the highest correlation value of 0.94 when coexpressed with the genes PRC1, PCLAF, UBE2C, ZWINT, and UBE2T.
A decrease in survival rate was correlated with hub gene overexpression
Survival analyses indicated that all hub genes were correlated with the overall survival rate in ACC (Fig. 4a-k, log Rank <0.05). With the exception of HMMR (Fig. 4k), the five-year survival rate for all hub genes was below
| Genes | ACC-normal | ACA-normal | ACC-ACA | |||
|---|---|---|---|---|---|---|
| Log FC | Adj p-values | Log FC | Adj p-values | Log FC | Adj p-values | |
| CCNB1 | 2.8443 | 1.56E-15 | 1.6663 | 0.0005 | 1.1779 | 0.0210 |
| TOP2A | 2.7668 | 3.31E-14 | 1.4166 | 0.0049 | 1.3502 | 0.0097 |
| UBE2C | 2.4895 | 9.34E-11 | 1.3000 | 0.0192 | 1.1896 | 0.0441 |
| UBE2T | 2.3566 | 5.39E-12 | 1.2300 | 0.0112 | 1.1269 | 0.0267 |
| PRC1 | 2.3125 | 3.57E-11 | NS | NS | 1.4490 | 0.0054 |
| CDK1 | 2.221 | 1.4E-12 | 0.9185 | 0.0425 | 1.3025 | 0.0049 |
| TYMS | 2.2161 | 1.36E-10 | 1.5162 | 0.0019 | NS | NS |
| PCLAF | 2.1731 | 1.08E-11 | 1.2500 | 0.0055 | NS | NS |
| ZWINT | 2.0612 | 2.56E-11 | 1.0186 | 0.0223 | 1.0426 | 0.0241 |
| RACGAP1 | 2.0428 | 2.28E-11 | NS | NS | 1.3984 | 0.0022 |
| HMMR | 2.0324 | 2.23E-09 | NS | NS | 1.2168 | 0.0174 |
| MAD2L1 | 1.9425 | 1.92E-14 | 1.0605 | 0.0022 | 0.8821 | 0.0153 |
| AURKA | 1.9399 | 1.24E-09 | 1.0210 | 0.0300 | NS | NS |
| GMNN | 1.9358 | 1.83E-14 | 1.3989 | 3.62E-05 | NS | NS |
| NUSAP1 | 1.9268 | 5.32E-11 | 1.4602 | 0.0003 | NS | NS |
| CCNB2 | 1.7745 | 3.12E-10 | 1.1180 | 0.0054 | NS | NS |
| SMC4 | 1.5976 | 2.82E-09 | 0.7949 | 0.0474 | NS | NS |
| ATAD2 | 1.5662 | 3.15E-09 | 0.8560 | 0.0281 | NS | NS |
| CKS2 | 1.5605 | 5.91E-09 | 1.0666 | 0.0059 | NS | NS |
| BIRC5 | 1.4851 | 8.99E-07 | NS | NS | 1.3591 | 0.0034 |
| H2AZ1 | 1.2718 | 1.19E-09 | 0.7141 | 0.0193 | NS | NS |
| SMC2 | 1.2672 | 1.39E-10 | 0.7946 | 0.0047 | NS | NS |
| CKS1B | 1.1203 | 1.16E-05 | NS | NS | NS | NS |
| CENPW | 1.1140 | 0.0005 | NS | NS | NS | NS |
| FEN1 | 1.0954 | 6.58E-07 | 0.8683 | 0.0071 | NS | NS |
| MCM3 | 0.8982 | 8.33E-08 | NS | NS | NS | NS |
| RFC4 | 0.8230 | 3.47E-05 | NS | NS | NS | NS |
| CENPN | 0.8176 | 2.08E-05 | NS | NS | 0.6548 | 0.0292 |
| MCM4 | 0.7796 | 4.24E-06 | 0.5056 | 0.0500 | NS | NS |
| PCNA | 0.4492 | 0.0138 | NS | NS | NS | NS |
| MCM7 | NS | NS | NS | NS | NS | NS |
50%. The expression of CDK1 was the most significant factor (Fig. 4a) that negatively impacted survival rates (HR=11.52). Consequently, an increase in the expression of all hub genes was correlated with a decrease in long-term survival.
Overexpression of the hub gene causes more women to die than men
A visual representation of the correlation between the longevity of ACC patients and their biological sex is shown in Fig. 5a-k. In both sexes, all hub genes were associated with a decreased ACC survival rate. In women, CDK1 (Fig. 5a) had the greatest influence on survival rates, while in men, PRC1 (Fig. 5j), ZWINT (Fig. 5b), and TYMS (Fig. 5g) had the greatest influence. Women are affected more negatively by hub genes than men, with the exception of PRC1, ZWINT, and TYMS genes. Over a five-year follow-up period, only women with overexpression of CDK1, UBE2C (Fig. 5e), PCLAF (Fig. 5h), and CCNB1 (Fig. 5f) exhibited a zero survival rate, while men did not show this outcome.
TP53 signaling pathway, chromosome segregation, and cell cycle regulation are mediated by hub genes
The significance of hub genes in various biological processes is illustrated (Fig. 6 and Table S4). The roles of CDK1 and CCNB1 in the TP53 signaling pathway (adj.p.val=9.38E-03), progesterone-mediated oocyte maturation (adj.p.val = 1.46E-02), oocyte meiosis (adj.p.val=2.56E-02), and the cell cycle (adj.p.val =3.12E-02) are elucidated in Fig. 6a. According to the BP analysis (Fig. 6b), TOP2A, CDK1, RACGAP1, UBE2C, ZWINT, and CCNB1 are linked to sister chromatid/chromosome segregation and nuclear division (adj.p.val =4.50E-07 and 1.97E-06, respectively). The MF analysis (Fig. 6c) revealed chromatin binding characteristics of TOP2A, UBE2T, CDK1, and PCLAF (adj.p.val=7.95E-03). It identified CDK1 and CCNB1 to have histone kinase
activity (adj.p.val=6.64E-03) and classified UBE2T and UBE2C as components of the ubiquitination process (adj.p.val=6.64E-03). The CC analysis (Fig. 6d) indicates that TOP2A, CDK1, PCLAF, HMMR, and CCNB1 are essential components of the microtubule cytoskeleton and play a role in microtubule organizing centers (adj.p.val=6.18E-04). Proposed cellular components for hub genes are the spindle (adj.p.val=7.34E-04) and the centrosome (adj.p.val =2.80E-03).
Hub gene expression was associated with TP53 mutations and ACC tumor stage
An association was observed between the expression levels of hub genes and the TP53 mutation status in ACC tissues (Fig. 7a-k). Expression levels of hub genes were assessed at various stages of ACC tumors (Fig. 71-v). All hub genes, with the exception of PRC1, exhibited differential expression in tissues at stages 1 or 2 compared to stage 4.
TOP2A and CDK1 were the most reliable ACC diagnostic markers
Hub genes were assessed for their diagnostic potential in ACC through ROC curve analysis (Fig. 8). All hub genes exhibited an AUC exceeding 0.93, demonstrating their strong diagnostic potential. The hub genes TOP2A and CDK1 exhibited the highest diagnostic potential, evidenced by an AUC of 0.98. The evidence suggests that all hub genes play a significant role in the diagnosis of ACC.
M2 macrophages and CD8 +T cells infiltrate ACC tumors negatively with hub gene overexpression
The infiltration of three immune cell types exhibited a negative correlation with most hub genes (Fig. 9 and Table S5). The infiltration of M2 macrophages into ACC tumors was reduced upon the overexpression of hub genes (Fig. 9a-k). CD8 +T cells represented the second category of immune cells that were unable to infiltrate ACC tissues upon overexpression of hub genes (Fig. 91-u). One exception was the THYMS gene, which did not show a correlation with CD8 + T cell infiltration.
A TF network regulating hub genes was constructed
The regulatory network visualization identified 32 TFs that regulate eight hub genes (Fig. 10). ZWINT, PCLAF, and UBE2T hub genes were absent from the TF network. The node with the highest degree in the network was CCNB1, indicating a substantial number of TFs that regulate this hub gene. The activation of the CCNB1 hub gene is promoted by FOXM1, KLF5, RELA, and PTTG1, whereas its inhibition is mediated by YBX1, KLF4, TP53, and IRF1. E2F1 serves as a TF that regulates a significant number of hub genes, including TYMS, TOP2A, CDK1, RACGAP1, and CCNB1.
Nine additional TFs were identified, although their functions remain unknown (Table 2). Of the 32 TFs, 21 exhibited significant expression variations between ACC and normal tissues, categorizing them as differentially expressed TFs (DE-TFs). FOXM1 exhibited the highest level of upregulation among DE-TFs, whereas KLF4 demonstrated the greatest downregulation.
Drug-gene interaction networks targeting hub genes and TFs were constructed
By integrating the hub genes and their regulatory TF genes, drug-gene interaction networks were constructed, with each gene being matched to approved drugs (Fig. 11). Out of 668 potential drugs, 194 were officially approved, with 34 of these approved drugs capable of targeting two or more genes. According to the authors, fifteen of the 34 pharmaceuticals were not employed as drugs against ACC tumors.
Paclitaxel and cisplatin emerged as the most significant pharmaceutical agents, serving as central nodes within the network. Paclitaxel interacts with six genes, including two hub genes (TOP2A and TYMS) and four DE-TFs (BRCA1, RELA, E2F2, and RB1). Cisplatin targets five genes, with TYMS as a hub gene and BRCA1, E2F1, MYC, and RB1 as DE-TFs. Additionally, fluorouracil, mitoxantrone, and quercetin each target four genes, while doxorubicin hydrochloride, etoposide, and resveratrol target three genes individually. Several medications concentrate on TYMS, BRCA1, and TOP2A genes, with TYMS and TOP2A being the targets of 14 and 11 drugs, respectively. BRCA1, as a TF, is targeted by 12 drugs. HMMR, functioning as an ACC hub gene, is targeted by fluorouracil and epirubicin. The UBE2T gene did not enrich the drug-gene interaction network.
Discussion
The hypothalamus-pituitary-adrenal axis is essential for maintaining internal balance during stress and optimising various physiological functions19,20. The adrenal gland plays a crucial role in cognition, cardiovascular function, immune response, metabolism, and overall viability. Therefore, it is essential to identify the specific genetic factors that may compromise ACC survival. The identification of genes influencing ACC mortality poses a considerable challenge due to the restricted information available to researchers. This study identified hub genes in ACC tissues, analysed their co-expression patterns, and characterised their associated regulatory proteins. This research sought to elucidate potential therapeutic agents that interact with ACC hub genes and their regulatory proteins to enhance disease management.
The WGCNA analysis of microarray datasets identified 11 hub genes that demonstrated at least a twofold overexpression in ACC tissues relative to adjacent normal tissues. Significant co-expression and strong correlation among hub genes suggest an expected harmonious and overlapping regulation of these genes. The co-expression patterns of hub genes suggest that they collectively contribute to synchronized processes, and the enrichment of these genes in cell cycle control and the TP53 signaling pathway clarifies our comprehension. The latter function reflects the ability of hub genes to respond to various intrinsic and extrinsic stresses that impact cellular homeostasis21. The involvement of co-expressed hub genes in ubiquitination indicates their potential role in regulating histone levels via ubiquitylation-dependent proteolysis22.
a.
NS
Log2 FC
p-value
p - value and log2 FC
40
30
-Log10 P
20
-
PCLAF
CCNB1
CDK1
TOP2A
RACGAP1
10
UBE2T
PRC1
ZWINT
TYMS
UBE2C
HMMR
0
I
-5.0
2.5
0.0
2.5
5.0
Log2 fold change
b. CDK1
c. ZWINT
d. RACGAP1
e. TOP2A
f. UBE2C
g. CCNB1
10
1
-
0
₡
0
5
0
-
8
5
A
0
2
expression level
expression level
expression level
expression level
+
10
.
+
*
expression level
expression level
€
0
.:
+
”
3
*
..
4
₪
2
₪
*
₪
4
A
₪
~
-
30
-
-
A
*
o
o
o
o
o
A
0
ACC
normal
ACC
normal
ACC
normal
ACC
normal
ACC
normal
ACC
normal
h. TYMS
i. PCLAF
j. UBE2T
k. HMMR
I. PRC1
A
0
0
®
”
.
₡
.
5
expression level
expression level
expression level
:
expression level
0
expression level
Y
€
%.
:
a
A
4
1
*
U
e
X
3.
Y
1
E
2
15
!
2
N
~
1
M
-
-
KA
2
$
o
o
o
o
o
ACC
normal
ACC
normal
ACC
normal
cancer
normal
ACC
normal
The survival rate of ACC was significantly influenced by co-expressed hub genes, a finding corroborated by previous studies23-26. This study marks the inaugural identification of HMMR and UBE2T as novel hub genes that influence the viability of ACC. The HMMR protein directs progenitor cell division and facilitates neural development through the PLK1-dependent pathway. In pancreatic cancer patients with TP53 mutations or loss of heterozygosity, HMMR expression was elevated, leading to negative outcomes associated with increased interactions with the anaphase-promoting complex (APC)27,28. Comparable outcomes were noted in ACC patients exhibiting HMMR overexpression, as they demonstrated a reduced survival rate. APC overexpression was observed in ACC tissues relative to adjacent normal tissues, as indicated by microarray
CCNB1
CDK1
HMMR
TOP2A
TYMS
PRC1
PCLAF
UBE2C
ZWINT
UBE2T
RACGAP1
1
CCNB1
1
0.87
0.86
0.89
0.81
0.82
0.85
0.82
0.8
0.86
0.83
0.8
CDK1
1
0.89
0.9
0.79
0.89
0.9
0.89
0.9
0.88
0.9
HMMR
1
0.86
0.68
0.84
0.81
0.8
0.8
0.85
0.87
0.6
TOP2A
1
0.85
0.86
0.88
0.91
0.85
0.89
0.88
0.4
TYMS
1
0.76
0.86
0.83
0.72
0.8
0.73
0.2
PRC1
1
0.86
0.83
0.87
0.85
0.94
0
PCLAF
1
0.83
0.82
0.88
0.94
-0.2
UBE2C
1
0.86
0.87
0.94
-0.4
ZWINT
1
0.81
0.94
-0.6
UBE2T
1
0.94
-0.8
RACGAP1
1
-1
Fig. 3. Corr plot analysis. Correlations between hub genes are greater than 68% in all cases. The highest correlation value of 0.94 is obtained when RACGAP1 is co-expressed with PRC1, PCLAF, UBE2C, ZWINT, and UBE2T.
data analysis (LogFC=1.0156, adj. p-val=6.0719E-09). UBE2T was identified as overexpressed in HCC, lung adenocarcinoma (LUAD), gastric cancer, and osteosarcoma29-32. UBE2T suppression, through the modulation of epithelial-mesenchymal transition-related factors, resulted in decreased tumour growth, invasiveness, and metastatic potential32. The knockdown of UBE2T in HCC resulted in cell cycle arrest at the G1/S phase, increased apoptosis, and decreased proliferation, migration, and invasion31. The inhibition of the PI3K/Akt signalling pathway has been suggested as a mechanism through which UBE2T knockdowns reduce osteosarcoma cell proliferation, migration, and invasion30. Additionally, the UBE2T protein acts as an E2 ubiquitin-conjugating enzyme responsible for the degradation of the BRCA1 tumour suppressor33. Our microarray analysis indicated that BRCA1 exhibited a slight overexpression (LogFC=0.4, adj. p-val =0.0057) in ACC tissues. We propose that the elevated expression of BRCA1 at the mRNA level is balanced by its degradation at the protein level.
Our observations indicate that hub genes may play a role in the sex-specific reduction of ACC survival, as only females with elevated levels of CDK1, UBE2C, CCNB1, and PCLAF experienced catastrophic death five
a. CDK1
b. ZWINT
1.00
1.0
Survival probability
0.75
Survival probability
0.75
0.50
0.50
0.25
0.25
p < 0.0001
HR=11.52, (95%CI=0.02972-0.2535)
Risk Group HIGH expression + LOW expression 5000
p < 0.0001
Risk Group + HIGH expression LOW expression 5000
0.00
0.00
HR=6.423, (95%CI=0.06228-0.3892)
0
1000
2000
Days
3000
4000
0
1000
2000
Days
8000
4000
c. RACGAP1
d. TOP2A
1.00
1.00
Survival probability
Survival probability
0.75
0.75
0.50
0.50
0.25
0.25
p < 0.0001
Risk Group · HIGH LOW expression
p < 0.0001
- Risk Group HIGH expression LOW
0.00
HR-5.817, (95%CI-0.06898-0.4284)
0.00
HR-5.776, (95%CI-0.06972-0.43)
0
1000
2000
Days
3000
4000
6000
0
1000
2000
Days
3000
4000
6000
e. UBE2C
f. CCNB1
1.00
1.00
Survival probability
0.75
Survival probability
0.75
0.50
0.50
0.26
0.25
p < 0.0001
Risk Group + FISH expression LOW expression
Risk Group 4 HIGH LOW expression 5000
0.00
HR-5.44, (95%CI-0.07687-0.4395)
p < 0,0001
0.00
HR=4.9, (95%CI=0.08576-0.4856)
0
1000
2000
Days
3000
4000
8000
0
1000
2000
Days
3000
4000
g. TYMS
h. PCLAF
1.00
1.00
Survival probability
0.76
Survival probability
0.76
0.50
0.50
0.25
0.25
Risk Group HIGH SSP LOW expression M.
Risk Group 4 LOW expression 5000
p = 0.00011
p = 0.00032
0.00
HR= 4.719, (95%CI=0.0894-0.5024)
0.00
HR=4.137, (95%CI=0.1051-0.556)
0
1000
2000
Days
9000
1000
0
1000
2000
Days
3000
4000
i. UBE2T
j. PRC1
1.00
1.000
Survival probability
0.75
Survival probability
0.75
0.50
0.50
0.25
0.25
p = 0.0012
HR=3.592, (95%CI=0.122-0.6352)
Risk Group High Expression LOW expression
p = 0.0022
Risk Group + LOW expression
0.00
0.00
HR=3.402 (95%CI=0.1282-0.6742)
0
1000
2000
Days
3000
4000
5000
0
1000
2000
Days
3000
4000
5000
k. HMMR
1.00
Survival probability
0.75
0.50
0.25
p = 0.024
Risk Group · HIGH expression LOW expression
0.00
HR=2.381, (95%CI=0.193-0.914)
0
1000
2000
Days
9000
4000
5000
a. CDK1
b. ZWINT
1.00
1.00
Survival probability
0.75
Survival probability
0.75
0.50
0.50
Risk Group
HIGH expression_female expression_male
Risk Group
0.26
0.26
HIGH expression_female
LOW expression female
LOW expression_male
How expression female
LOW maman expression_male
HR_female: 21.01 (95%CI=0.01061-0.2134)
HR_male: 5.042 (95%CI=0.04204-0.9358) p < 0.0001
HR_female: 6.491 (95%CI=0.04784-0.4961)
HR_male: 8.019 (95%CI-0.02644-0.588)
0.00
0.00
p = 0.00014
0
1000
2000
Days
3000
4000
5000
0
1000
2000
Days
3000
4000
5000
c. RACGAP1
d. TOP2A
1.00
1.00
Survival probability
0.76
0.76
Survival probability
0.50
0.50
Risk Group
Risk Group
HIGH expression_female HIGH expression_malu
HIGH expression_female
0.25
LOW expression_female
0.25
HIGH expression male
expression_male
Low expression_female
LOW expression_male
HR_female: 6.279 (95%CI-0.04499-0.5638)
HR_male: 5.533 (95%CI=0.04728-0.6908)
HR_female: 6.639 (95%CI-0.04713-0.4813)
0.00
p = 0.00035
HR_male: 6.262 (95%CI=0.03411-0.7475)
0.00
p = 0.00037
0
1000
2000
Days
3000
4000
6000
6
1000
2000
Days
3000
4000
6000
e. UBE2C
f. CCNB1
1.00
1.00
Survival probability
0.75
Survival probability
0.75
0.50
0.50
Risk Group
Risk Group HIGH expression_female
0.26
HIGH expression_female
LOW expression female
0.26
HIGH expression_male
LOW expression_male
LOW expression male
HR_female: 15.44 (95%CI=0.01673-0.2508)
HR_male: 3.196 (95%CI=0.06718-1.457) p < 0.0001
HR_female: 6.936 (95%CI-0.04507-0.4612) HR_male: 3.301 (95%CI=0.07991-1.149)
0.00
0.00
p = 0.00092
0
1000
2000
Days
3000
4000
5000
0
1000
2000
Days
3000
4000
5000
g. TYMS
h. PCLAF
1.00
1.00
Survival probability
0.76
Survival probability
0.76
0.50
0.50
Risk Group
Risk Group
HIGH expression_female
44 4
+ HIGH expression_female expression male
0.25
0.25
Low Expression_female LOW expression_male
LOW expression male
HR_female: 4.281 (95%CI=0.07478-0.7297)
HR_female: 8.937 (95%CI=0.03009-0.416)
HR_male: 6.815 (95%CI=0.0378-0.5696)
HR_male: 2.775 (95%CI=0.0951-1.366)
0.00
p = 0.0013
0.00
p = 0.0019
0
1000
2000
Days
3000
4000
6000
0
1000
2000
Days
3:000
4000
5000
i. UBEZT
j. PRC1
1.00
1.00
Survival probability
0.75
0.75
Survival probability
0.50
0,60
Risk Group
HIGH expression_female HIGH expression_male LOW expression female
Risk Group
0.95
0.25
HIGH expression_female
LOW expression_male
LOW expression_male
HR_female: 3.737 (95%CI-0.09831-0.7283)
HR_male: 3.511 (95%CI=0.06131-1.323)
HR_female: 2.6298 (95%CI-0.14433-1.0018) HR_male: 8.6318 (95%CI=0.01472- 0.9119)
0.00
p = 0.011
0.00
p = 0.018
0
1000
2000
Days
3000
4000
5000
0
1000
2000
Days
3000
4000
6000
k. HMMR
1.00
Survival probability
0.76
0.50
Risk Group
HIGH expression_female
0.25
Low expression female
LOW expression_male
HR_female: 4.099 (95%CI-0.08716-0.683)
HR_male: 1.321 (95%CI=0.2199-2.607)
0.00
p = 0.047
0
1000
2000
Days
2000
4000
5000
a. KEGG
b.BP
CCNB1
CCNB1
TOP2A
UBE2C
CDK1
CDK1
ZWINT
TYMS
RACGAP1
TP53 signaling pathway
sister chromatid segregation
progesterone mediated oocyte maturation
nuclear chromosome segregation
oocyte meioses
mitotic sister chromatid segregation
cell cycle
chromosome segregation
one carbon pool by folate
nuclear division
lepFC
logFC
2
3
2
3
c.MF
d.CC
CCNB1
CCNB1
TOP2A
TOP2A
UBE2C
CDK1
UBE2T
PCLAF
CDK1
ZWINT
TYMS
PCLAF
RACGA
HMMR
histone kinase activity
ubiquitin conjugating enzyme activity
cyclin B1-CDK1 complex
ubiquitin-like protein conjugating enzyme activity
microtubule cytoskeleton
chromatin binding
chromosomal region
thymidylate synthase activity
microtubule organizing center
spindle
logFC
logFC
2
3
Transcript per milion far CDK1
Transorign por re dion for COOK1
Transcript per milan for ZMVINT
Trenstrigt por medlon for ZWINT
Transcript per milian for RACGAP1
Transcript per million fer TOP24
Transcript per miles for TOP24
p.value=0.0000922
D.v010=0.000105
TP_Metan
Sagen
Thescript por milice for UBEIC
Transcript por million far CONB
Transcript per million far DCND:1
Phobies 0.000254
P.value= 0.00000
Stages
Avtlet & Lộ0
Transcript per million for TYMS
Transcript per million for PCLAF
Trasscript per rellice fer TYAS
Transcript per million for PCLAF
P.value= 0.00353
TPS3_Mutant
Stagan
Transcript per milian fler UBERT
Transcript per milian for PACT
Transcript per million for UBEST
Trasserat per million for PRG
P-value= 0.00875
p.value= 0.00091
TP83_Uhánná
Stages
Dingen
Transcript por milice for HMMR
Transcript per million for HMMR
p.value= 0.0172
years post-diagnosis. The investigation of potential biological mechanisms revealed that CDK1 and CCNB1 have been previously associated with sexual dimorphism, as they were found to be upregulated exclusively in females diagnosed with hepatocellular carcinoma (HCC)34. Consequently, the genes were identified as indicators of unfavourable prognosis in male patients. The activity of TYMS and CDK1 has also varied between females and males with glioblastoma35. The unisexual association of TYMS polymorphism with the incidence of adverse events in cancer is mechanistically correlated with the oestrogen receptor in females, which potentially regulates TYMS expression36. UBE2C overexpression was significantly linked to poor prognosis exclusively in HR +/ HER2- breast cancer, as estrogen upregulated UBE2C mRNA and protein through direct binding to the UBE2C promoter region37. Furthermore, except for PRC1, co-expressed hub genes showed increased expression in high- grade aggressive ACC tumors relative to low-grade tumors (stage 4 versus stages 1 and 2). Consequently, all hub genes function as indicators of ACC aggressiveness, and based on this, the potential of hub genes for ACC diagnosis was evaluated. The analysis revealed a significant potential of all genes for ACC diagnosis, with TOP2A and CDK1 identified as the most effective predictors.
The presence of CD8 + T cells and M2 macrophages in ACC tissues exhibited an adverse association with the expression of hub genes. This study aligns with previous researches that identified a negative correlations between immune cell infiltration and the expression of CDK1 in pancreatic cancer38, TOP2A in non-small cell lung cancer39, UBE2T and UBE2C in LUAD40,41, HMMR in LUAD and neuroendocrine prostate cancer42,43, TYMS in acute respiratory distress44, and RACGAP1 in HCC45. The biological mechanisms underlying these observations remain unclear; however, the functions and distribution patterns of immune cells within the tumor microenvironment can be influenced by the hub genes. A UBE2T knockdown experiment demonstrated tumor suppression through PD-L1 inhibition46. They administered a glycolysis inhibitor and established that FOXA1 upregulates UBE2T-mediated glycolysis, which diminishes CD8 + T cell function and facilitates immunological evasion in LUAD. Given the established function of hub genes in immune cell infiltration, we propose them as novel targets for immunological therapy.
Although the negative infiltration of CD8+T cells appears to be a logical factor contributing to ACC immune evasion47, the reduced infiltration of M2 macrophages was unexpected given their established role in facilitating tumor progression. This contradiction prompts the exploration of novel therapeutic strategies, as multiple studies have demonstrated that M2 macrophages contribute to chemoresistance and provide protection against radiation48-50. M2 macrophages inhibit fluorouracil chemosensitivity via the activation of the CCL22/ PI3K/AKT signaling pathway, leading to a reduction in apoptosis51. M2 macrophages not only release IL-10 and
a. TOP2A
b. CDK1
c. CCNB1
100-
100
100
80-
80-
80-
60-
60-
60-
TPR
TPR
40-
40-
TPR
40-
20-
AUC=0.98±0.01
20-
AUC=0.98±0.01
20-
Pvalue<0.0001
Pvalue<0.0001
AUC=0.97±0.01
0
0
Pvalue<0.0001
0
20
40
60
80
100
0
20
40
60
80
100
0
0
20
40
60
80
100
FPR
FPR
FPR
d. PCLAF
e. ZWINT
f. UBE2T
100-
100-
100
80-
80-
80-
TPR
60
TPR
60-
TPR
60-
40-
40-
40-
20-
AUC=0.96±0.02
20-
AUC=0.95±0.02
20-
AUC=0.95±0.01
0
Pvalue<0.0001
0
Pvalue<0.0001
0
Pvalue<0.0001
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
FPR
FPR
FPR
g. PRC1
h. HMMR
i. TYMS
100
100
100
80
80-
80
TPR
60
TPR
60-
TPR
60-
40
40-
40-
20
AUC=0.94±0.02
20-
AUC=0.94±0.02
20-
AUC=0.94±0.02
Pvalue<0.0001
0
Pvalue<0.0001
Pvalue<0.0001
0
0
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
FPR
FPR
FPR
j. RACGAP1
k. UBE2C
100-
100-
80-
80-
60-
60-
TPR
TPR
40-
40
20-
AUC=0.93±0.03
20
AUC=0.93±0.03
0
Pvalue<0.0001
Pvalue<0.0001
0
0
20
40
60
80
100
0
20
40
60
80
100
FPR
FPR
Figure 8
a. CDK1
Purity
Macrophage M2_XCELL
I. CDk1
Purity
T cull CD8+_XCELL
Hne 0.333
P = 3.916-03
Aha - - 0.321
Expression Level [log2 TPM)
Expression Level (log2 TPM)
P= 6.100-03
9-
5.0
AOC
ACC
A
*
-
-
0.0
-
0.2
0.4
0.8
0.88
1.0 0.00
0.02
0.04
Purity
Infiltration Level
0.06
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
Purity
Infiltration Level
0.16
b. ZWINT
Purity
Macrophage M2_XCELL
m. ZWINT
Purity
T cell CD8+_XCELL
Expression Level (log2 TPM)
Expression Level (log2 TPM)
P- . 930-02
NE
G
ADO
8
AOC
2
2
0.2
0.4
0.6
0.0
Purity
1.0 0.00
0.02
0.04
0.06
Infiltration Level
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
Purity
Infiltration Level
0.15
C. RACGAP
Purity
Macrophage M2_XCELL
Purity
Toell CD8+_XCELL
%
Rha — 0.259 P=2.006-02
n. RACGAP1
A
Expression Level (log2 TPM)
Expression Level (log2 TPM)
4
ACC
AOG
3
0
0.2
0.4
0.6
0.8
1.0 0.00
0.02
0.04
0.06
Purity
Infiltration Level
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
Infiltration Level
0.15
Purity
d. TOP2A
Purity
Macrophage M2_XCELL
0. TOPZA
Purity
Toall CD8+_XCELL
D - 2.224-65
Aha -0.270 ₱= 1.79%-02
Expression Level (log,2 TPM)
9.
5.0
Expression Level (log2 TPM]
4
.
2.5
ACC
ACC
N
0.0
0.2
0.4
0.6
0.8
1.0 0.00
0.02
0.04
0.08
Purity
Infiltration Level
0.2
0.4
0.6
0.0
1.0 0.00
0.05
0.10
0.15
Purity
Infiltration Level
e. UBE2C
Purity
Macrophage M2_XCELL
p. UBE2C
Purity
T call CD8+_XCELL
Expression Level (log2 TPM]
P = 4.416-03
7.5
Expression Level (log2 TPM]
7.5
6.0
O
5.0
2.5
2.5
0.0
0.2
0.4
0.6
0.8
1.0 0.DO
0.02
0.04
0.06
Purity
Infiltration Level
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
Purity
Infiltration Level
0.16
f. CCNB1
Purity
Macrophage M2_XCELL
q. CCNB1
Purity
T cell CD8+_XCELL
Expression Level (log2 TPM)
2
Expression Level |log2 TPM|
ACC
A
AOC
Nu
%
0.2
0.4
0.6
0.0
1.0 0.00
0.02
0.04
Purity
Infiltration Level
0.06
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
Purity
Infiltration Level
0.15
R. PCLAF
Purity
Macrophage M2_XCELL
F. PCLAF
Purity
Toell CD8+_XCELL
P -2.906-05
Expression Level (log2 TPM)
P = 1.048-05
Expression Level (log2 TPM]
G
5.0
A
.5
ACO
AOC
10
0.0
+
O
0.2
0.4
0.6
0.8
1.0 0.00
0.02
0.04
Purity
Infiltration Level
0.06
0.2
0.4
0.6
0.8
1.0 0.00
0.05
0.10
0.15
Purity
Infiltration Level
h. UBE2T
Purity
Macrophage M2_QUANTISEQ
S. UBE2T
Purity
T cell CD8+_XCELL
Expression Level [log2 TPM|
1
+
Expression Level (log2 TPM]
·
ACC
A
ACC
0.2
0.4
0.6
0.8
Purity
1.0 0.00
0.05
0.10
Infiltration Level
0.2
0.6
0.8
1.0 0.00
0.05
0.10
0.15
Purity
Infiltration Level
I. PRC1
Purity
Macrophage M2_XCELL
E. PRC1
Purity
T cull CDI+_XCELL
A
1
Expression Level [log2 TPM)
Expression Level |log2 TPMJa
#
. .
1
AOC
ACC
1
a
0.2
0.4
0.6
0.8
1.0 0.00
0.02
0.04
0.06
Purity
Infiltration Level
0.2
0.4
0.6
0.8
Purity
1.0 0.00
0.05
0.10
Infiltration Level
0.16
j. HMMR
Purity
Macrophage M2_XCELL
u. HMMR
Purity
T cell CD8+_XCELL
Expression Level (log2 TPM)
1
Expression Level (log2 TPM]
4
0
3
ACC
:
:
S
0.2
0.4
0.6
0.0
1.0 0.00
0.02
0.04
0.06
-1
Purity
Infiltration Level
0.2
0.4
0.6
0.8
0.05
0.10
Purity
1.0 0.DO
Infiltration Level
0.15
k. TYMS
Purity
Macrophage M2_XCELL
n
8.5-65
Expression Level (log2 TPM]
1
4
.
ACC
2
*
-
:
*
=
0.2
0.4
0.6
0.8
1.0 0.00
0.02
0.04
Purity
Infiltration Level
0.06
AR
RB1
ARID3A
TP73
ETS2
IRF1
HMMR
CUX1
CDK1
KLF4
SMAD7
RELA
NFKB1
TFDP1
E2F3
RACGAP1
FOXM1
PTTG1
TP53
E2F1
CCNB1
TFAP2A
CHD8
BRCA1
PRC1
KLF5
TYMS
YBX1
MYC
TBP
SP1
USF1
E2F4
TFCP2
TOP2A
UBE2C
ATF1
ESR1
UHRF1
MED1
TGF-1 but also exhibit CD163 as a characteristic marker52-55. To substantiate our observations, we analyzed the expression levels of M2 macrophage markers in ACC versus normal tissues. Coordination was noted as IL-10 (logFC =- 0.5946, adj.p.val=0.0069), TGF-b1 (logFC =- 0.5351, adj.p.val=5.70E-05), and CD163 (logFC =- 2.7047, adj.p.val =7.91E-14) exhibited reduced expression in ACC relative to normal tissues.
We propose that our transcriptome based chemotherapeutic agents may provide benefits for ACC tumors due to the detrimental presence of M2 macrophages. Moreover, mitotane as the sole medication authorized for the treatment of ACC, shows various side effects. At higher doses, mitotane is associated with unavoidable central nervous system damage, neuromuscular manifestations, adrenal atrophy, and adrenal insufficiency56,57. It exerts detrimental effects on thyroid hormone production and testosterone metabolism58, causes hypercholesterolemia59 and increases cardiovascular risk60. Mitotane also disrupts the metabolism of several drugs, such as antihypertensives, statins, antibiotics, chemotherapeutic agents, and coumarin-like anticoagulants61,62. It may restrict tolerability and should be discontinued upon the onset of neurotoxicity and acute hepatic impairment63-65. Consequently, mitotane is contraindicated in cases of severe impairments, necessitating the consideration of alternative antineoplastic agents. Cisplatin, another commonly used ACC chemotherapeutic agents causes severe kidney problems, hypersensitivity reactions, decreased immunity to infections, gastrointestinal disorders, hemorrhage, and hearing loss66. Cisplatin induces oxidative and nitrosative damage to hepatocytes, as the increase in xanthine oxidase (XO) and nitric oxide (NO) may contribute to
| Hub gene | Regulatory TFs | Role of TF | Microarray expression of TFs |
|---|---|---|---|
| CCNB1 | BRCA1 | unknown | 0.3842 |
| E2F1 | unknown | 0.3319 | |
| E2F3 | unknown | 0.7184 | |
| E2F4 | unknown | NS | |
| FOXM1 | Activation | 2.0405 | |
| IRF1 | Repression | -0.6446 | |
| KLF4 | Repression | -1.0758 | |
| KLF5 | Activation | -0.9512 | |
| MYC | unknown | -1.0173 | |
| NFKB1 | Activation | -0.5422 | |
| NFKB1 | unknown | -0.5422 | |
| PTTG1 | Activation | NS | |
| RELA | Activation | -0.5476 | |
| RELA | unknown | -0.5476 | |
| TBP | unknown | 0.3189 | |
| TFAP2A | unknown | NS | |
| TP53 | Repression | 0.3461 | |
| USF1 | unknown | NS | |
| YBX1 | Repression | NS | |
| CDK1 | ARID3A | unknown | NS |
| E2F1 | Activation | 0.3319 | |
| E2F3 | unknown | 0.7184 | |
| ETS2 | unknown | -0.9342 | |
| IRF1 | Repression | -0.6446 | |
| PTTG1 | Activation | NS | |
| RB1 | unknown | 0.5886 | |
| SMAD7 | Repression | -0.4355 | |
| TFDP1 | Activation | 0.4655 | |
| TP53 | Repression | 0.3461 | |
| TP53 | unknown | 0.3461 | |
| TP73 | unknown | NS | |
| HMMR | AR | unknown | NS |
| PRC1 | TP53 | Repression | 0.3461 |
| RACGAP1 | CUX1 | unknown | -0.2287 |
| E2F1 | unknown | 0.3319 | |
| TOP2A | ATF1 | unknown | 0.8839 |
| E2F1 | Activation | 0.3319 | |
| UHRF1 | unknown | 0.9386 | |
| YBX1 | unknown | NS | |
| TYMS | CHD8 | unknown | NS |
| E2F1 | Activation | 0.3319 | |
| E2F1 | unknown | 0.3319 | |
| ESR1 | unknown | NS | |
| SP1 | unknown | 0.9822 | |
| TFCP2 | unknown | NS | |
| TFDP1 | Activation | 0.4655 | |
| TP53 | Repression | 0.3461 | |
| USF1 | unknown | NS | |
| YBX1 | Repression | NS | |
| UBE2C | MED1 | unknown | 0.4352 |
| MYC | Activation | -1.0173 | |
| PRC1 | TP53 | unknown | 0.3461 |
Methotrexate
TBP
Vorinostat
Tretinoin
Irinotecan hydrochloride
Etoposide
RB1
Diethylstilbestrol
Topotecan hydrochloride
Verapamil
Fluorouracil
MYC
Folic acid
TYMS
Indomethacin
HMMR
Dexamethasone
E2F1
Vincristine
Doxorubicin hydrochloride
Bortezomib
Paclitaxel
Cisplatin
Quercetin
Tamoxifen
Olaparib
RELA
TOP2A
Epirubicin
Hydrocortisone butyrate
ETS2
Indoprofen
NFKB1
Mitoxantrone
Mitomycin
Rutin
Carboplatin
Artesunate
Daunorubicin hydrochloride
BRAC1
Mitoxantrone hydrochloride
RACGAP1
Rucaparib
Gemcitabine
Resveratrol
Hydroquinone
CDK1
oxidative stress in hepatotoxicity67,68. Cisplatin is associated with several side effects, including nephrotoxicity nephrotoxicity69, neurotoxicity70, hepatotoxicity cardiotoxicity, and heart failure71, and ototoxicity72,73. Conventional and novel strategies have led to the development of various cytotoxic agents74. The substitution of mitotane single drug treatment with combination therapy and adjunct therapy has demonstrated improved response rates and progression-free survival75. Our drug repositioning strategy, disclosed the prospective therapeutic drugs that target ACC hub genes and DE-TFs in a coordinated action. Our findings indicated that TYMS and TOP2A were targeted by multiple drugs, suggesting their potential efficacy in ACC therapy. We have for the first time determined that the HMMR hub gene is the target of fluorouracil and epirubicin. As a DNA-damaging combination treatment, fluorouracil and epirubicin cause apoptosis through the TP53 signaling pathway76. We hypothesized that ACC tumors harboring TP53 mutations may not respond positively to fluorouracil and epirubicin treatment. It was a matter of speculation, as we observed a substantial overexpression of the TOP2A, TYMS, E2F1, and HMMR hub genes in TP53-depleted ACC tumors. Together, these hub genes were identified as possible mediators of fluorouracil and epirubicin.
Mechanistically, fluorouracil possesses the capacity to induce both intrinsic and extrinsic apoptotic pathways77 while simultaneously exhibiting the advantageous characteristics of an absence of cardiotoxicity and myocardial ischemia78. The cytotoxicity of fluorouracil is primarily mediated through the inhibition of the TYMS enzyme, which plays a critical role in the synthesis of thymidines essential for DNA replication79. Consequently, the suppression of TYMS by fluorouracil promotes a reduction in cell proliferation80. When used in conjunction with apigenin, the inhibitory efficacy of fluorouracil was significantly enhanced, as apigenin is known to inhibit the upregulation of TYMS in colorectal cancer cells81. It is important to recognize that this is not the sole aspect of the intricate action of fluorouracil. The efficacy of fluorouracil in anti-tumor activity is influenced by the ribosomal stress response, mediated by the release of ribosomal proteins from the ribosomal complex. Ribosomal proteins engage in interactions with MDM2, thereby stabilizing TP53 and inducing G1/S phase arrest82. The inactivation of TP53 has been shown to negate the cell cycle arrest and apoptotic effects induced by fluorouracil in colorectal and colon malignancies83-85.Furthermore, resistance to fluorouracil therapy is frequently observed in patients with relapsed colorectal cancer who harbor TP53 mutations85. The implementation of a combination therapy that included fluorouracil and a phosphoinositide 3-kinase (PI3K) inhibitor (GDC-0326) effectively reduced drug resistance and exhibited substantial anticancer efficacy without a concurrent increase in toxicity86. Our research corroborates the current study, as UBE2T, a positive regulator of the PI3K pathway, has been identified as a novel upregulated hub gene in ACC. UBE2T facilitates the proliferation and invasion of various malignancies, including breast87, ovarian88, and renal cancers89. Consequently, the combination of fluorouracil with PI3K inhibitors may provide advantages in TP53 wild type ACC tumors.
Epirubicin also exerted its cytotoxic effects through TP53, as MCF-7 drug-resistant cells did not induce TP53 expression and activity following treatment with the drug90. Epirubicin mechanistically induces TP53 to repress FOXM1 that is required for normal G1-S, G2 and M cell cycle phase transitions. Epirubicin reduces FOXM1 expression through the depletion of the E2F1 transactivator and the accumulation of the pRB/ E2F transcriptionally repressive complex at the FOXM1 promoter91. In contrast, TP53 is subject to negative phosphorylation by CDK192, as the inhibition of CDK1 promotes TP53-mediated mitochondrial apoptosis and induces G2/M-phase arrest93. Inhibitors of CDK1 showed advantageous due to the formation of a complex between CCNB1 and CDK1. The CCNB1/CDK1 complex relocates to mitochondria, phosphorylating TP53 at the ser-315 residue and inducing a pro-survival response91. Consequently, we conclude that epirubicin combined with CDK1 inhibitors demonstrates potential benefits for ACC tumors exhibiting wild-type TP53. Moreover, epirubicin showed favorable permeability and significant lipophilicity, which enhance its local effects on bladder cancer94,95. The combination of epirubicin and mindfulness intervention enhanced the drug’s cytotoxicity in patients with urinary system tumors and depression96.
Our findings provide new insights into the formation and progression of ACC and suggest the repositioning of pharmaceuticals to improve treatment outcomes. We demonstrate concurrent overexpression of hub genes alongside diminished infiltration of M2 macrophages, as well as the efficacy of epirubicin and fluorouracil in addressing ACC. We hypothesized that diminished infiltration of M2 macrophages into ACC tumors confers benefits, especially in individuals with TP53 wild-type status. In addition to the reduced infiltration of M2 macrophages that is induced by the overexpression of the HMMR hub gene, the resistance of ACC tumors to fluorouracil and epirubicin could be mitigated by administering PI3K and CDK1 inhibitors. The future validation of these findings is currently underway by our team, and we remain committed to providing experimental validations to significantly enhance the credibility and impact of the study.
Conclusion
Eleven ACC hub genes exhibiting overexpression have been identified, with UBE2T and HMMR genes classified as novel. All hub genes exhibited a significant correlation with disease survival, with women demonstrating hub gene overexpression succumbing within five years post-diagnosis. The lethality of hub genes is primarily associated with their essential roles in cell cycle progression, chromosome segregation, and the TP53 signaling pathway. Hub genes demonstrated significant predictive potential for ACC, as their overexpression correlated with the progression and aggressiveness of the disease. The hub gene exhibited a negative correlation with the infiltration of CD8+T cells and M2 macrophages. Fluorouracil and epirubicin serve as novel repositioned chemotherapeutics targeting ACC, facilitating TP53-mediated apoptosis. The overexpression of HMMR in ACC tumors contributes to M2 macrophage-mediated resistance to fluorouracil, highlighting an additional advantage of the drug.
Data availability
The datasets analyzed during the current study are available in the Gene Expression Omnibus (GEO) repository, https://www.ncbi.nlm.nih.gov/geo/ and in Genomic Data Commons Data Portal, https://portal.gdc.cancer.gov /projects/TCGA-THCA.
Received: 19 June 2024; Accepted: 3 July 2025
Published online: 17 July 2025
References
1. Fassnacht, M., Libé, R., Kroiss, M. & Allolio, B. Adrenocortical carcinoma: A clinician’s update. Nat. Rev. Endocrinol. 7, 323-335 (2011).
2. Hallanger-Johnson, J. E. Systemic therapy for adrenocortical carcinoma: A review. AME Med. J. 5, 5-5 (2020).
3. Berruti, A. et al. Long-term outcomes of adjuvant mitotane therapy in patients with radically resected adrenocortical carcinoma. J. Clin. Endocrinol. Metab. 102, 1358-1365 (2017).
4. Yuan, M. et al. (2022). A gene co-expression network-based drug repositioning approach identifies candidates for treatment of hepatocellular carcinoma. Cancers (Basel). 14, 1573
5. Xu, J., Koval, A. & Katanaev, V. L. Clofazimine: A journey of a drug. Biomed. Pharmacother. 167, 115539 (2023).
6. Begley, C. G. et al. Drug repurposing: Misconceptions, challenges, and opportunities for academic researchers. Sci. Transl. Med. 13, eabd5524 (2021).
7. Xue, H., Li, J., Xie, H. & Wang, Y. Review of drug repositioning approaches and resources. Int. J. Biol. Sci. 14, 1232-1244 (2018).
8. Ashburn, T. T. & Thor, K. B. Drug repositioning: Identifying and developing new uses for existing drugs. Nat. Rev. Drug Discov. 3, 673-683 (2004).
9. Joachim Haupt, V. & Schroeder, M. Old friends in new guise: Repositioning of known drugs with structural bioinformatics. Brief. Bioinform. 12, 312-326 (2011).
10. Arakelyan, A. et al. Transcriptome-guided drug repositioning. Pharmaceutics 11, 677 (2019).
11. Can, T. Introduction to bioinformatics. Methods Mol. Biol. 1107, 51-71 (2014).
12. Langfelder, P. & Horvath, S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinform. 9, 559 (2008).
13. Yip, A. M. & Horvath, S. Gene network interconnectedness and the generalized topological overlap measure. BMC Bioinform. 8, 22 (2007).
14. Chin, C. H. et al. cytoHubba: Identifying hub objects and sub-networks from complex interactome. BMC Syst. Biol. 8, S11 (2014).
15. 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. 53, D672-D677 (2025).
16. Aran, D., Hu, Z. & Butte, A. J. xCell: Digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 18, 220 (2017).
17. Shannon, P. et al. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498-2504 (2003).
18. Freshour, S. L. et al. Integration of the drug-gene interaction database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res. 49, D1144-D1151 (2021).
19. Russell, G. & Lightman, S. The human stress response. Nat. Rev. Endocrinol. 15, 525-534 (2019).
20. Lyraki, R. & Schedl, A. Adrenal cortex renewal in health and disease. Nat. Rev. Endocrinol. 17, 421-434 (2021).
21. Harris, S. L. & Levine, A. J. The p53 pathway: Positive and negative feedback loops. Oncogene 24, 2899-2908 (2005).
22. Singh, R. K., Kabbaj, M. H. M., Paik, J. & Gunjan, A. Histone levels are regulated by phosphorylation and ubiquitylation-dependent proteolysis. Nat. Cell Biol. 11, 925-933 (2009).
23. Xu, W. H. et al. Screening and Identification of potential prognostic biomarkers in adrenocortical carcinoma. Front. Genet. 10, 821 (2019).
24. Guo, J. et al. Identification of hub genes and pathways in adrenocortical carcinoma by integrated bioinformatic analysis. J. Cell. Mol. Med. 24, 4428-4438 (2020).
25. Xiao, H. et al. Identification of five genes as a potential biomarker for predicting progress and prognosis in adrenocortical carcinoma. J. Cancer 9, 4484-4495 (2018).
26. Xing, Z., Luo, Z., Yang, H., Huang, Z. & Liang, X. Screening and identification of key biomarkers in adrenocortical carcinoma based on bioinformatics analysis. Oncol. Lett. 18, 4667-4676 (2019).
27. Lin, A. et al. High levels of truncated RHAMM cooperate with dysfunctional p53 to accelerate the progression of pancreatic cancer. Cancer Lett. 514, 79-89 (2021).
28. Connell, M. et al. HMMR acts in the PLK1-dependent spindle positioning pathway and supports neural development. Elife 6, e28672 (2017).
29. Li, Y., Yang, X. & Lu, D. Knockdown of ubiquitin-conjugating enzyme E2T (UBE2T) suppresses lung adenocarcinoma progression via targeting fibulin-5 (FBLN5). Bioengineered 13, 11867-11880 (2022).
30. Wang, Y. et al. Knockdown of UBE2T inhibits osteosarcoma cell proliferation, migration, and invasion by suppressing the PI3K/ Akt signaling pathway. Oncol. Res. 24, 361-369 (2016).
31. Guo, J., Wang, M., Wang, J. P. & Wu, C. X. Ubiquitin-conjugating enzyme E2T knockdown suppresses hepatocellular tumorigenesis via inducing cell cycle arrest and apoptosis. World J. Gastroenterol. 25, 6386-6403 (2019).
32. Luo, C. et al. Erratum: UBE2T knockdown inhibits gastric cancer progression (Oncotarget. (2017) 8 (32639-32654) https://doi.or g/10.18632/oncotarget.15947). Oncotarget 14, 188-189 (2023).
33. Ueki, T. et al. Ubiquitination and downregulation of BRCA1 by ubiquitin-conjugating enzyme E2T overexpression in human breast cancer cells. Cancer Res. 69, 8752-8760 (2009).
34. Wu, Y. et al. Identification and characterization of sexual dimorphism-linked gene expression profile in hepatocellular carcinoma. Oncol. Rep. 42, 937-952 (2019).
35. Qin, S. et al. Identification and characterization of sex-dependent gene expression profile in glioblastoma. Neuropathology 43, 72-83 (2023).
36. Ioannou, C. et al. Gender-dependent association of TYMS-TSER polymorphism with 5-fluorouracil or capecitabine-based chemotherapy toxicity. Pharmacogenomics 22, 669-680 (2021).
37. Kim, Y. J. et al. UBE2C overexpression aggravates patient outcome by promoting estrogen-dependent/independent cell proliferation in early hormone receptor-positive and HER2-negative breast cancer. Front. Oncol. 9, 1574 (2020).
38. Wang, X., Yang, C., Liu, Y. & Wang, J. SUMOylation substrate encoding genes as prognostic biomarkers in pancreatic ductal adenocarcinoma with functional assessment of SAF-B2. Front. Pharmacol. 16, 1532658 (2025).
39. Zhou, Y., Dong, Y., Sun, Q. & Fang, C. Diagnosis and prognosis of non-small cell lung cancer based on machine learning algorithms. Comb. Chem. High Throughput Screen. 26, 2170-2183 (2023).
40. Xu, F., Xiong, N., Yuan, Y. & Liu, J. Prognostic value of UBE2T and its correlation with immune infiltrates in lung adenocarcinoma. J. Oncol. 2022, 1-9 (2022).
41. Cai, D., Tian, F., Wu, M., Tu, J. & Wang, Y. UBE2C is a diagnosis and therapeutic biomarker involved in immune infiltration of cancers including lung adenocarcinoma. J. Cancer 15, 1701-1717 (2024).
42. Ma, X. et al. HMMR associates with immune infiltrates and acts as a prognostic biomaker in lung adenocarcinoma. Comput. Biol. Med. 151, 106213 (2022).
43. Tang, H. et al. Identification of BIRC5 and HMMR as prognostic biomarkers for immune infiltration in prostate cancer. Transl. Androl. Urol. 13, 2482-2497 (2024).
44. Li, G., Yan, K., Zhang, W., Pan, H. & Guo, P. ARDS and aging: TYMS emerges as a promising biomarker and therapeutic target. Front. Immunol. 15, 1365206 (2024).
45. Zhou, Y. et al. Upregulation of RACGAP1 is correlated with poor prognosis and immune infiltration in hepatocellular carcinoma. Transl. Cancer Res. 13, 847-863 (2024).
46. Pu, J. et al. FOXA1/UBE2T inhibits CD8+T cell activity by inducing mediates glycolysis in lung adenocarcinoma. Front. Biosci. - Landmark 29, 134-null (2024).
47. Raskov, H., Orhan, A., Christensen, J. P. & Gögenur, I. Cytotoxic CD8+ T cells in cancer and cancer immunotherapy. Br. J. Cancer 124, 359-367 (2021).
48. Yang, C. et al. Increased drug resistance in breast cancer by tumor-associated macrophages through IL-10/STAT3/bcl-2 signaling pathway. Med. Oncol. 32, 14 (2015).
49. Hughes, R. et al. Perivascular M2 macrophages stimulate tumor relapse after chemotherapy. Cancer Res. 75, 3479-3491 (2015).
50. Zhang, W. et al. Depletion of tumor-associated macrophages enhances the effect of sorafenib in metastatic liver cancer models by antimetastatic and antiangiogenic effects. Clin. Cancer Res. 16, 3420-3430 (2010).
51. Wei, C. et al. M2 macrophages confer resistance to 5-fluorouracil in colorectal cancer through the activation of CCL22/PI3K/AKT signaling. Onco. Targets. Ther. 12, 3051-3063 (2019).
52. Potas, J. R., Haque, F., Maclean, F. L. & Nisbet, D. R. Interleukin-10 conjugated electrospun polycaprolactone (PCL) nanofibre scaffolds for promoting alternatively activated (M2) macrophages around the peripheral nerve in vivo. J. Immunol. Methods 420, 38-49 (2015).
53. Liu, J., Geng, X., Hou, J. & Wu, G. New insights into M1/M2 macrophages: key modulators in cancer progression. Cancer Cell Int. 21, 389 (2021).
54. Mantovani, A., Allavena, P., Marchesi, F. & Garlanda, C. Macrophages as tools and targets in cancer therapy. Nat. Rev. Drug Discov. 21, 799-820 (2022).
55. Cendrowicz, E., Sas, Z., Bremer, E. & Rygiel, T. P. The role of macrophages in cancer development and therapy. Cancers (Basel). 13, 1946 (2021).
56. Waszut, U., Szyszka, P. & Dworakowska, D. Understanding mitotane mode of action. J. Physiol. Pharmacol. 68, 13-26 (2017).
57. Veytsman, I., Nieman, L. & Fojo, T. Management of endocrine manifestations and the use of mitotane as a chemotherapeutic agent for adrenocortical carcinoma. J. Clin. Oncol. 27, 4619-4629 (2009).
58. Paragliola, R. M. et al. Role of mitotane in adrenocortical carcinoma - Review and state of the art. Eur. Endocrinol. 14, 62-66 (2018).
59. Chortis, V. et al. Mitotane therapy in adrenocortical cancer induces CYP3A4 and inhibits 5a-reductase, explaining the need for personalized glucocorticoid and androgen replacement. J. Clin. Endocrinol. Metab. 98, 161-171 (2013).
60. Gagnon, N. et al. Characterization of hyperlipidemia secondary to mitotane in adrenocortical carcinoma. Endocr. Oncol. 2, 1-8 (2022).
61. Kroiss, M., Quinkler, M., Lutz, W. K., Allolio, B. & Fassnacht, M. Drug interactions with mitotane by induction of CYP3A4 metabolism in the clinical management of adrenocortical carcinoma. Clin. Endocrinol. (Oxf) 75, 585-591 (2011).
62. Cuddy, P. G. & Loftus, L. S. Influence of mitotane on the hypoprothrombinemic effect of warfarin. South. Med. J. 79, 387-388 (1986).
63. Corso, C. R. et al. Pharmacological profile and effects of mitotane in adrenocortical carcinoma. Br. J. Clin. Pharmacol. 87, 2698- 2710 (2021).
64. Else, T. et al. Adrenocortical carcinoma. Endocr. Rev. 35, 282-326 (2014).
65. Goto, T. et al. Case report: Adjuvant therapy with a high dose of mitotane for adrenocortical carcinoma in a 4-year-old boy. Clin. Pediatr. Endocrinol. 17, 71-74 (2008).
66. Dasari, S. & Bernard Tchounwou, P. Cisplatin in cancer therapy: Molecular mechanisms of action. Eur. J. Pharmacol. 740, 364-378 (2014).
67. Kart, A., Cigremis, Y., Karaman, M. & Ozen, H. Caffeic acid phenethyl ester (CAPE) ameliorates cisplatin-induced hepatotoxicity in rabbit. Exp. Toxicol. Pathol. 62, 45-52 (2010).
68. Al-Majed, A. A. Carnitine deficiency provokes cisplatin-induced hepatotoxicity in rats. Basic Clin. Pharmacol. Toxicol. 100, 145- 150 (2007).
69. De Jongh, F. E. et al. Weekly high-dose cisplatin is a feasible treatment option: Analysis on prognostic factors for toxicity in 400 patients. Br. J. Cancer 88, 1199-1206 (2003).
70. Hamers, F. P. T., Gispen, W. H. & Neijt, J. P. Neurotoxic side-effects of cisplatin. Eur. J. Cancer Clin. Oncol. 27, 372-376 (1991).
71. Hu, Y. et al. Cisplatin-induced cardiotoxicity with midrange ejection fraction: A case report and review of the literature. Med. (United States) 97, e13807 (2018).
72. Yousef, M. I., Saad, A. A. & El-Shennawy, L. K. Protective effect of grape seed proanthocyanidin extract against oxidative stress induced by cisplatin in rats. Food Chem. Toxicol. 47, 1176-1183 (2009).
73. Najafi, M., Tavakol, S., Zarrabi, A. & Ashrafizadeh, M. Dual role of quercetin in enhancing the efficacy of cisplatin in chemotherapy and protection against its side effects: A review. Arch. Physiol. Biochem. 128, 1438-1452 (2022).
74. Schteingart, D. E. Conventional and novel strategies in the treatment of adrenocortical cancer. Brazilian J. Med. Biol. Res. 33, 1197-1200 (2000).
75. Fassnacht, M. et al. Combination chemotherapy in advanced adrenocortical carcinoma. N. Engl. J. Med. 366, 2189-2197 (2012).
76. Kandioler-Eckersberger, D. et al. TP53 mutation and p53 overexpression for prediction of response to neoadjuvant treatment in breast cancer patients. Clin. Cancer Res. 6, 50-56 (2000).
77. Gielecińska, A. et al. Apoptosis, necroptosis, and pyroptosis as alternative cell death pathways induced by chemotherapeutic agents ?. Biochim. Biophys. Acta - Rev. Cancer 1878, 189024 (2023).
78. Freeman, N. J. & Costanza, M. E. 5-Fluorouracil-associated cardiotoxicity. Cancer 61, 36-45 (1988).
79. Wigmore, P. M. et al. Effects of 5-FU. in 157-164 (2010). https://doi.org/10.1007/978-1-4419-6306-2_20.
80. Zhang, N., Yin, Y., Xu, S. J. & Chen, W. S. 5-Fluorouracil: Mechanisms of resistance and reversal strategies. Molecules 13, 1551- 1569 (2008).
81. Yang, C. et al. Apigenin enhances apoptosis induction by 5-fluorouracil through regulation of thymidylate synthase in colorectal cancer cells. Redox Biol. 47, 102144 (2021).
82. Sun, X .- X., Dai, M .- S. & Lu, H. 5-fluorouracil activation of p53 involves an MDM2-ribosomal protein interaction. J. Biol. Chem. 282, 8052-8059 (2007).
83. Xiu, Z. et al. A multifunctional injectable microsphere with enhanced near-infrared photo-antibacterial, ROS scavenging, and anti-inflammatory properties for periodontitis treatment. Theranostics 15, 3750-3780 (2025).
84. Bunz, F. et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents. J. Clin. Invest. 104, 263-269 (1999).
85. Gao, X. et al. Deficient or R273H and R248W mutations of p53 promote chemoresistance to 5-FU via TCF21/CD44 axis-mediated enhanced stemness in colorectal carcinoma. Front. Cell Dev. Biol. 9, 788331 (2022).
86. Zhang, Z. et al. Gdc-0326 enhances the effects of 5-fu in colorectal cancer cells by inducing necroptotic death. Onco. Targets. Ther. 14, 2519-2530 (2021).
87. Dutta, R. et al. UBE2T promotes breast cancer tumor growth by suppressing DNA replication stress. NAR Cancer 4, zcac035 (2022).
88. Cui, P. et al. UBE2T regulates epithelial-mesenchymal transition through the PI3K-AKT pathway and plays a carcinogenic role in ovarian cancer. J. Ovarian Res. 15, 103 (2022).
89. Hao, P., Kang, B., Li, Y., Hao, W. & Ma, F. UBE2T promotes proliferation and regulates PI3K/Akt signaling in renal cell carcinoma. Mol. Med. Rep. 20, 1212-1220 (2019).
90. Millour, J. et al. ATM and p53 regulate FOXM1 expression via E2F in breast cancer epirubicin treatment and resistance. Mol. Cancer Ther. 10, 1046-1058 (2011).
91. Nantajit, D. et al. Cyclin B1/Cdk1 phosphorylation of mitochondrial p53 induces anti-apoptotic response. PLoS One 5, e12341 (2010).
92. Huang, Y., Li, D., Wang, L., Su, X. & Tang, X. CENPF/CDK1 signaling pathway enhances the progression of adrenocortical carcinoma by regulating the G2/M-phase cell cycle. J. Transl. Med. 20, 78 (2022).
93. Kojima, K., Shimanuki, M., Shikami, M., Andreeff, M. & Nakakuma, H. Cyclin-dependent kinase 1 inhibitor RO-3306 enhances p53-mediated bax activation and mitochondrial apoptosis in AML. Cancer Sci. 100, 1128-1136 (2009).
94. Lu, J. L. et al. Efficacy of intravesical therapies on the prevention of recurrence and progression of non-muscle-invasive bladder cancer: A systematic review and network meta-analysis. Cancer Med. 9, 7800-7809 (2020).
95. Suo, N. et al. Magnetic multiwalled carbon nanotubes with controlled release of epirubicin: An intravesical instillation system for bladder cancer. Int. J. Nanomedicine 14, 1241-1254 (2019).
96. Liu, J., Guo, Y .- P., Lu, Y .- M. & Wang, B .- L. Clinical value of combining epirubicin with mindfulness intervention in patients with urinary system tumors and depression. World J. psychiatry 15, 98737 (2025).
Acknowledgements
The research was carried out at Isfahan University of Iran (https://ui.ac.ir/) with support from the departments of Research, Technology, and Graduate Studies.
Author contributions
Zahra: Conception, design, collection, and/or assembly of data, data analysis, interpretation, and drafting of the manuscript Seyed-Morteza: Conception, design, assembly of data, data analysis, interpretation, financial support, drafting the manuscript, revising it critically for important intellectual content, and final approval of the manuscript. seyedeh-elmira: Conception, design, collection, and/or assembly of data, data analysis, interpre- tation, and drafting of the manuscript.
Funding
This study has been conducted in Isfahan University of Iran and was supported financially by the Departments of Research, Technology and Graduate Offices (Grant number: 1400/08/15).
Declarations
Competing interests
The authors declare no competing interests.
Ethical approval
Each author confirms that their research is supported by an institution that is primarily involved in education or research.
Consent for publication
Not applicable.
Additional information
Supplementary Information The online version contains supplementary material available at https://doi.org/1 0.1038/s41598-025-10452-w.
Correspondence and requests for materials should be addressed to S .- M.J.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommo ns.org/licenses/by-nc-nd/4.0/.
@ The Author(s) 2025, corrected publication 2025