![Open Access](/resources/images/iconopenaccess.png)
Weighted gene co‑expression network analysis of the association between upregulated AMD1, EN1 and VGLL1 and the progression and poor prognosis of breast cancer
- Authors:
- Published online on: July 18, 2021 https://doi.org/10.3892/etm.2021.10462
- Article Number: 1030
-
Copyright: © Yang et al. This is an open access article distributed under the terms of Creative Commons Attribution License.
Abstract
Introduction
According to Global Cancer Statistics 2018, newly diagnosed cases of breast cancer account for ~25% of all cancers in females. Female breast cancer is the most frequently diagnosed malignancy worldwide (in 154 of 185 countries) and is the primary cause of cancer-associated death in over 100 countries (1). Marusyk and Polyak (2) revealed that breast cancer is a heterogeneous disease on a clinicopathological, cellular and molecular level. According to a molecular study, breast cancer may be classified into ≥4 subtypes, including luminal, human epidermal growth factor receptor 2 (HER2)-enriched, basal-like and normal-like breast cancer (3). Among these subtypes, basal-like breast cancer is the most highly malignant type, accounting for 75% of triple-negative breast cancers (TNBCs) that lack estrogen receptor (ER), progesterone receptor and HER2 expression (4). TNBC is a highly aggressive and heterogenic disease with an earlier age of onset and greater metastatic potential than non-TNBC (5). Evidence suggests that patients with non-TNBC have improved survival rates compared with those with TNBC and that these patients benefit from targeted therapy. Due to a lack of available targeted therapies, chemotherapy is currently the standard treatment for TNBC (6,7). However, patients frequently experience drug resistance, which results in tumor recurrence and disease progression (8). Therefore, it is critical to identify novel potential therapeutic targets for breast cancer, particularly TNBC.
Weighted gene co-expression network analysis (WGCNA) uses systems biology to identify modules of highly related genes and associate these modules with clinical traits. Therefore, WGCNA is widely used to identify and screen for biomarkers (9), and has been successfully used to discover therapeutic targets for a variety of cancer types, including, but not limited to, laryngeal cancer (10), leiomyosarcoma (11) and advanced gastric cancer (12).
In the present study, the breast cancer microarray dataset GSE76275 was downloaded from the Gene Expression Omnibus (GEO) database and WGCNA was used to select target gene candidates, which were then validated using alterative datasets and in vitro experimentation.
Materials and methods
Data preprocessing
Gene expression data and clinical information from patients with breast cancer were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The gene expression profiles included GSE76275(13) and GSE42568(14) (platform, GPL570 (HG-U133_Plus_2); Affymetrix Human Genome U133 Plus 2.0 Array; Thermo Fisher Scientific, Inc.), as well as GSE25055(15) and GSE25065(15) (platform, GPL96 (HG-U133A); Affymetrix Human Genome U133A Array; Thermo Fisher Scientific, Inc.). The inclusion criteria were as follows: i) Patients with a diagnosis of breast cancer; ii) in GSE76275, patients with complete clinical data, including age, tumor stage, tumor size, lymph node status, metastasis and tumor grade; iii) in GSE25055, patients with complete clinical data on tumor size, lymph node status, tumor stage, tumor grade, breast cancer subtype, status (dead or alive) and specific follow-up time; iv) in GSE25065, patients with complete clinical data on tumor size, lymph node status, tumor stage, tumor grade and breast cancer subtype; and v) in GSE42568, patients with complete clinical information on tumor size, lymph node status, tumor grade and ER status.
Weighted gene co-expression network construction
The top 25% most variable genes in GSE76275 were selected for further analysis using the WGCNA package (https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/; version 1.69). First, the samples were clustered to construct the sampleTree and detected outliers were selected based on cut height. ‘Sample dendrogram’ and ‘trait heatmap’ were used to develop each network in order to investigate the relationship between the corresponding sample gene expression data and clinical phenotypes. The value of the soft-thresholding parameter used to construct the adjacency matrix was set as β=6. Furthermore, the adjacency matrix was transformed into the topological overlap matrix (TOM). According to the TOM-based dissimilarity measure, genes with absolute correlation values were clustered into the same module to generate a cluster dendrogram (deep-split, 2; minimum cluster size, 30; cut height, 0.25). In an effort to visually represent the relationships between modules and the clinical features of breast cancer, Pearson's correlation coefficient was calculated and plotted in a heatmap. Modules were determined to have a significant correlation to clinical traits when P<0.05. The highest correlating module was selected as the key module for further analysis.
Identification and validation of hub genes
In the present study, hub genes were screened out based on the cut-off criteria of module membership (MM), gene significance (GS) and survival analysis. MM was defined as the Pearson's correlation coefficient between each gene in the key module and the module eigengene, where MM reflects the module connectivity of each gene. GS was defined as the correlation coefficient between each gene in the key module and its corresponding clinical trait, where GS represents the biological significance of each gene. MM>0.7 and GS>0.2 were set as cut-off criteria to screen genes in the key module with high functional significance. Gene expression and clinical prognostic information (vital status and follow-up time) from patients with breast cancer were integrated based on GSE25055. The survival package in R (https://github.com/therneau/survival; version 3.2-7) was used to perform survival analysis. In order to assess the prognostic value of these genes in patients with breast cancer, the Kaplan-Meier plotter (http://kmplot.com/) database was used to generate relapse-free survival (RFS) and overall survival (OS) curves. Genes that were indicated to be associated with RFS through both methods were designated as hub genes for deeper validation. In order to validate their reliability, the expression of the hub genes in relation to clinicopathological characteristics (such as pathological T stage, pathological N stage, tumor stage, tumor grade and breast cancer subtype) were analyzed based on the GSE25055, GSE25065 and GSE42568 datasets. The R packages ‘ggplot’ (https://ggplot2.tidyverse.org/; version 3.3.0), ‘ggpubr’ (https://rpkgs.datanovia.com/ggpubr/; version 0.2.5) and ‘ggsignif’ (https://github.com/const-ae/ggsignif; version 0.6.0) were used to perform correlation analyses between gene expression and clinical traits. To verify the protein expression levels of the hub genes in breast cancer and normal tissues, immunohistochemistry (IHC) data were downloaded from the Human Protein Atlas (HPA; http://www.proteinatlas.org). The R package ‘corrplot’ (https://github.com/taiyun/corrplot; version 0.84) was used to assess the correlation between the expression levels of each hub gene.
Gene set enrichment analysis (GSEA)
GSEA was used to predict the potential function of each hub gene. For each hub gene, a total of 267 breast cancer samples in the GSE25055 dataset were divided into high-risk and low-risk groups; c2.cp.kegg.v7.1.symbols.gmt was selected as the reference gene set. The number of permutations was set at 1,000 times for each analysis. Nominal P<0.05, false discovery rate <25% and gene size ≥50 were selected as the thresholds.
IHC
Breast cancer and adjacent normal tissues were collected from patients (age range, 52-67 years; mean age, 60 years) undergoing mastectomy and with a postoperative pathology diagnosis of breast cancer at Zhongnan Hospital (Wuhan, China) between May and September 2019. Written informed consent was obtained from each patient prior to surgery and the patient protocols were approved by the hospital's ethics committee (approval no. 2015073). IHC was used to detect the expression levels of S-adenosylmethionine decarboxylase proenzyme (AMD1) in both sets of tissues. The tissue samples were excised, fixed with formalin, dehydrated and embedded in paraffin, and subsequently cut into 5-mm sections. For IHC, the sections were incubated with primary antibodies against AMD1 (1:500 dilution; cat. no. ab65820; Abcam) at 4˚C overnight. After washing three times with PBS, the sections were incubated with an HRP-conjugated secondary antibody (1:400 dilution; cat. no. AS061; ABclonal Biotech Co., Ltd.) at room temperature for 1.5 h. After further washing, the peroxidase activity was visualized using freshly prepared diaminobenzidine (OriGene Technologies, Inc.) and the slides were then lightly counterstained with Harris' hematoxylin. The negative controls were processed in the same way, but with PBS in place of the primary antibody. Finally, the slides were observed under a light microscope (Nikon, Inc.; x200).
Cell culture
Human breast cancer cell lines (MCF-7, MDA-MB-231, MDA-MB-468 and MDA-MB-157) and a mammary epithelial cell line (MCF-10A) were purchased from the Cell Bank of the Chinese Academy of Sciences. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) at 37˚C with 5% CO2.
Reverse transcription-quantitative (RT-q)PCR
Total RNA was extracted from each cell type using TRIzol® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The miScript Reverse Transcription kit (Qiagen GmbH) was used according to the manufacturer's instructions for RT with 3 µg total RNA. qPCR was performed using the SYBR-Green Master Mix (Takara Biotechnology Co., Ltd.). The qPCR was performed on a StepOnePlus Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.) and the thermocycling conditions were as follows: 95˚C for 5 min and 40 cycles of 95˚C for 10 sec and 60˚C for 30 sec. The following primer sequences were used in the present study: GAPDH forward, 5'-TGTGGGCATCAATGGATTTGG-3' and reverse, 5'-ACACCATGTATTCCGGGTCAAT-3'; and AMD1 forward, 5'-GGCCTGTACCATACAAGCCC-3' and reverse, 5'-CCACGTAGACGAGGTAGTTGTG-3'. AMD1 expression was quantified using the 2-ΔΔCq method (16).
Transfection
MDA-MB-231 cells were transfected with small inhibitory RNA targeting AMD1 (si-AMD1; 5'-CGGATGGAACTTATTGGACTA-3') and negative control siRNA (si-NC, 5'-UUCUCCGAACGUGUCAGGUTT-3'), which was designed and purchased from Shanghai GenePharma Co., Ltd. Transfection was performed using Lipofectamine® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. After transfection for 48 h, the cells were harvested for subsequent analyses.
Cellular proliferation assay
The effects of AMD1 on breast cancer cell viability were determined using the Cell Counting Kit-8 (CCK-8) as per the manufacturer's protocol. MDA-MB-231 cells were seeded into 96-well plates at a concentration of 2x103 cells/well. After culturing at 37˚C for 24 h, 10 µl CCK-reagent was added to each well and the cells were incubated for a further 2 h. To estimate cell numbers, the absorbance of each well was measured using a microplate reader at 450 nm.
Migration and invasion assays
Cellular invasion and migration assays were performed using Transwell inserts (Corning, Inc.) coated with or without Matrigel, respectively. MDA-MB-231 cells (2x105 cells/ml) were seeded into the upper chamber, and DMEM containing 20% FBS was added to the lower chamber. After culturing at 37˚C for 24 h, cells that had migrated to the lower chambers were fixed with methanol, stained with 0.5% crystal violet and counted in three randomly selected fields using ImageJ software (version 1.53; National Institutes of Health).
Apoptosis analysis
Propidium Iodide/Annexin V-APC staining and flow cytometric analysis were performed to estimate the apoptotic rates of breast cancer cells. The Annexin V/PI Cell Apoptosis kit (Sungene Biotech Co., Ltd.) was used to detect Apoptotic cells according to the manufacturer's protocol. Flow cytometric analysis was performed using a BD Accuri C6 flow cytometer (BD Biosciences) and results were evaluated with FlowJo software (version 7.6.1; FlowJo LLC).
Statistical analysis
The data were analyzed using R (version 3.6.3) and GraphPad Prism 7 software (GraphPad Software, Inc.). All experiments were performed in triplicate and the data are expressed as the mean ± standard error of the mean. Differences between two groups were assessed using unpaired Student's t-test and those among multiple groups were assessed by one-way ANOVA and Bonferroni's post-hoc test. A two-tailed P<0.05 was considered to indicate a statistically significant difference.
Results
Clinicopathological characteristics of patients with breast cancer
The present study included 120 patients with breast cancer from the GSE76275 dataset, 260 from the GSE25055 dataset, 183 from the GSE25065 dataset and 101 from the GSE42568 dataset, all with complete clinicopathological data for the following analyses. Detailed clinicopathological information for each cohort is displayed in Table I.
Weighted co-expression network construction and key module identification
The top 25% most variable genes (n=4,052) were selected for co-expression analysis using the WGCNA package. A sample dendrogram and trait heatmap were used to split the selected samples into the appropriate clusters; the distribution map of clinical trait data is provided in Fig. 1A. When the power was equal to 6, the R2 scale was equal to 0.9 (Fig. 1B and C). Therefore, β=6 was selected as the soft threshold for breast cancer co-expression analysis. The 15 original co-expression modules were obtained using the dynamic tree cut method. After setting the cut height to 0.25, thereby merging highly similar modules (Fig. 2A), 12 modules were screened out (Fig. 2B). The association between the modules and clinical traits was then analyzed, allowing for the selection of key modules for further investigation. The red module, which contained 273 genes, was identified as the key module. The heatmap and histogram indicated that the red module was positively correlated with the tumor grade. In a scatter plot of GS vs. MM, a significant correlation was evident in the red module. The plot revealed that MM in the red module demonstrated a significant correlation with the tumor grade (r=0.62, P=2.2x10-30) (Fig. 3A-C).
Hub gene screening and validation
Hub genes were screened out using the red module. MM>0.7 and GS>0.2 were set as the cut-off criteria to screen 36 genes with high functional significance. Among them, 9 (prominin 1, γ-butyrobetaine hydroxylase 1, BAF chromatin remodeling complex subunit BCL11A, AMD1, rhophilin associated tail protein 1B, vestigial-like protein (VGLL1), tripartite motif containing 2, homeobox protein engrailed-1 (EN1) and keratin 6B) and 6 genes (Kruppel like factor 5, AMD1, EN1, desmocollin 2, VGLL1 and allograft inflammatory factor 1 like) were negatively correlated with RFS of patients with breast cancer, based on the validation dataset GSE25055 and the Kaplan-Meier plotter tool, respectively. AMD1, EN1 and VGLL1 were associated with poor RFS in both analyses and only VGLL1 was associated with a worse OS prognosis (according to Kaplan-Meier survival curves; Fig. 4A-I). Consequently, AMD1, EN1 and VGLL1 were identified as hub genes. Upon WGCNA of the dataset GSE76275, the red module was determined to be highly associated with tumor grade. Subsequently, three validation datasets were used to determine the relationship between tumor grade and the expression levels of AMD1, EN1 and VGLL1. The results revealed that higher hub gene expression levels were associated with advanced tumor grade in dataset GSE25055 (Fig. 5A-C). From dataset GSE25055, the expression levels of these genes were also determined to be increased in basal breast tumors compared to luminal and HER2-enriched breast tumors (Fig. 5D-F). However, with regard to tumor size, lymph node status and tumor stage, no significant association was observed between these clinicopathological parameters and the expression levels of the hub genes (Fig. S1). In datasets GSE25065 and GSE42568, hub gene upregulation also corresponded with advanced tumor grade and a more malignant cancer subtype (Figs. S2 and S3). IHC data from the HPA online database also demonstrated that the protein levels of AMD1 and EN1 were higher in tumor tissues than in normal tissues (Fig. 6A-F) and that the expression of each individual protein was strongly correlated with that of the other two (in both the GSE76275 and GSE25055 datasets) (Fig. 7).
GSEA
To investigate potential signaling pathways associated with the three hub genes, GSEA was used to identify Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enriched in breast cancer samples with high AMD1, EN1 and VGLL1 expression. Based on the cut-off criteria, the top 5 KEGG pathways enriched in the samples with high AMD1, EN1 and VGLL1 expression are displayed in Fig. 8. These hub genes were commonly enriched in ‘cell cycle’, ‘oocyte meiosis’, ‘pathogenic Escherichia coli infection’ and ‘pyrimidine metabolism’.
AMD1 knockdown inhibits proliferation and metastatic capacity, while promoting apoptosis in breast cancer cells
A series of additional experiments were performed to further investigate the expression levels and functions of AMD1 in breast cancer. Based on the IHC results, AMD1 was indicated to be upregulated in breast cancer tissues (Fig. 9A), which was also consistent with the results from the HPA. Compared with that in MCF-10A cells, AMD1 expression was significantly upregulated in all breast cancer cell lines, but most notably in MDA-MB-231 cells (Fig. 9B). Therefore, the MDA-MB-231 cell line was selected for further analyses. The results of the CCK-8 assay indicated that AMD1-knockdown significantly inhibited the proliferation of MDA-MB-231 cells (Fig. 9C). The number of migratory and invasive MDA-MB-231 cells transfected with si-AMD1 was also significantly reduced compared with those of the NC-transfected group (Fig. 9D and E). Furthermore, flow cytometric analysis demonstrated that the apoptotic rate of MDA-MB-231 cells was significantly increased in the si-AMD1 group as compared with that in the si-NC group (Fig. 9F).
Discussion
In the present study, the top 25% most variable genes in the GSE76275 dataset were used for co-expression analysis, from which 12 modules were identified. Among these modules, the red module was highly correlated with tumor grade. Using survival analysis, AMD1, EN1 and VGLL1 were subsequently identified as hub genes within the red module and their upregulation was associated with a poorer prognosis in patients with breast cancer in both the validation dataset GSE25055 and Kaplan-Meier plotter. The expression levels of hub genes were further validated and were indicated to be highly expressed in samples with advanced tumor grade and basal-like breast cancer. IHC staining demonstrated that the protein levels of AMD1 and EN1 were higher in breast cancer tissues than in normal tissues. In addition, the expression levels of these genes were strongly correlated with each other. According to the GSEA, the hub genes were confirmed to be commonly enriched in ‘cell cycle’, ‘oocyte meiosis’, ‘pathogenic Escherichia coli infection’ and ‘pyrimidine metabolism’. Finally, in vitro experiments were used to validate the expression and function of AMD1. As relevant experimental studies of the effect of EN1 and VGLL1 on breast cancer progression have previously been published (17,18), in vitro experimentation was not performed for these genes. The present results suggested that AMD1 is upregulated in breast cancer tissues and cells, and that AMD1 knockdown decreased the proliferation, invasion and migration abilities, whilst increasing apoptosis in breast cancer cells.
AMD1 encodes an important enzyme involved in polyamine biosynthesis, in which various aliphatic amine-associated polyamines are essential for promoting cellular proliferation and tumorigenesis. AMD1 has been demonstrated to promote epidermal wound healing by regulating cellular migration (19), and has been reported to have a significant role in the pathogenesis of multiple tumor types, such as prostate (20,21), non-small cell lung (22) and gastric cancer (23). A study revealed that mammalian target of rapamycin complex 1 regulates AMD1 to sustain polyamine metabolism in prostate cancer (20). AMD1 was also indicated to be upregulated in gastric cancer samples and patients with high AMD1 expression levels exhibited poorer OS rates; furthermore, inhibiting AMD1 suppressed cellular proliferation and migration in vitro, as well as tumor growth in vivo (23). The results of the present study provided results on the carcinogenic effects of AMD1 in breast cancer, which are consistent with its effect on the progression of gastric cancer.
The EN1 gene encodes a homeodomain-containing protein that regulates pattern formation during central nervous system development. EN1 expression was reported to be significantly higher in TNBC than in other breast cancer subtypes (17). Studies have indicated that upregulation of EN1 is correlated with significantly shorter OS times and increased risk of brain metastases in patients with TNBC (24), and that EN1 protein expression is increased in adenoid cystic carcinoma, with the higher expression of EN1 being associated with a lower survival rate (25). EN1 was also specifically expressed in normal eccrine glands and focally expressed in skin tumors and sweat gland neoplasms (26). The results of the present study are consistent with those of previous studies on the carcinogenicity of EN1.
The VGLL1 gene encodes a transcriptional co-activator involved in regulating the Hippo pathway in Drosophila (27). A study revealed that VGLL1 is predominantly expressed in BRCA1-associated TNBC and serves an oncogenic role in breast cancer (18). VGLL1 is also reportedly involved in human papillomavirus (HPV) gene expression via transcriptional enhancer factor 1, and thus, is crucial to the growth of HPV-associated malignancies such as cervical cancer (28). In gastric cancer, VGLL1 promoted cancer cell proliferation and metastasis, which was regulated by PI3K/AKT/β-catenin signaling (29). Furthermore, VGLL1 has been indicated to possess oncogenic functions in pediatric neuroepithelial neoplasms (30).
In conclusion, the present study aimed to identify hub genes involved in the pathogenesis of breast cancer using WGCNA. The results indicated that the upregulation of AMD1, EN1 and VGLL1 are correlated, and potentially detrimentally associated, with progression and prognosis in breast cancer. Therefore, inhibiting the expression of AMD1, EN1 and VGLL1 may be a potential therapeutic strategy for breast cancer.
Supplementary Material
Associations between the expression levels of hub genes and clinicopathological parameters. Association between (A) AMD1, (B) EN1 and (C) VGLL1 expression and tumor size. Association between (D) AMD1, (E) EN1 and (F) VGLL1 expression and lymph node status. Association between (G) AMD1, (H) EN1 and (I) VGLL1 expression and tumor stage. AMD1, S-adenosylmethionine decarboxylase proenzyme; EN1, homeobox protein engrailed-1; VGLL1, vestigial-like protein 1.
Associations between the expression levels of hub genes and clinicopathological parameters in dataset GSE25065. Association between (A) AMD1, (B) EN1 and (C) VGLL1 expression and tumor grade. Association between (D) AMD1, (E) EN1 and (F) VGLL1 expression and tumor subtype. Luminal means Lumina A and B, normal means normal-like breast cancer. Association between (G) AMD1, (H) EN1 and (I) VGLL1 expression and tumor size. Association between (J) AMD1, (K) EN1 and (L) VGLL1 expression and lymph node status. Association between (M) AMD1, (N) EN1 and (O) VGLL1 expression and tumor stage. *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001. ns, no significance; AMD1, S-adenosylmethionine decarboxylase proenzyme; EN1, homeobox protein engrailed-1; VGLL1, vestigial-like protein 1; Her2, human epidermal growth factor receptor 2.
Associations between the expression levels of hub genes and clinicopathological parameters in dataset GSE42568. Association between (A) AMD1, (B) EN1 and (C) VGLL1 expression and tumor grade. Association between (D) AMD1, (E) EN1 and (F) VGLL1 expression and estrogen receptor status. Association between (G) AMD1, (H) EN1 and (I) VGLL1 expression and tumor size. Association between (J) AMD1, (K) EN1 and (L) VGLL1 expression and lymph node status. **P<0.01, ***P<0.001 and ****P<0.0001. ns, no significance; AMD1, S-adenosylmethionine decarboxylase proenzyme; EN1, homeobox protein engrailed-1; VGLL1, vestigial-like protein 1.
Acknowledgements
Not applicable.
Funding
Funding: The present study was supported by the National Natural Science Foundation of China (grant nos. 81770283 and 82070302) and the Clinical Medical Research Center of Peritoneal Cancer of Wuhan (grant no. 2015060911020462). The funding agencies did not participate in the design of the study, the collection, analysis and interpretation of data or manuscript writing.
Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Authors' contributions
LY and WX conceived the study. XL, TY and HW performed data mining, acquisition and analysis. LS collected tissue samples. MF and YL performed the experiments. YL and XL drafted the manuscript and confirm the authenticity of all the raw data. WX and MF revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Written informed consent was obtained from each patient prior to surgery and the patient protocols were approved by the ethics committee of the Zhongnan Hospital of Wuhan University (approval no. 2015073).
Patient consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
References
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA and Jemal A: Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 68:394–424. 2018.PubMed/NCBI View Article : Google Scholar | |
Marusyk A and Polyak K: Tumor heterogeneity: Causes and consequences. Biochim Biophys Acta. 1805:105–117. 2010.PubMed/NCBI View Article : Google Scholar | |
Reis-Filho JS and Pusztai L: Gene expression profiling in breast cancer: Classification, prognostication, and prediction. Lancet. 378:1812–1823. 2011.PubMed/NCBI View Article : Google Scholar | |
Bahnassy A, Mohanad M, Ismail MF, Shaarawy S, El-Bastawisy A and Zekri AR: Molecular biomarkers for prediction of response to treatment and survival in triple negative breast cancer patients from Egypt. Exp Mol Pathol. 99:303–311. 2015.PubMed/NCBI View Article : Google Scholar | |
Garrido-Castro AC, Lin NU and Polyak K: Insights into molecular classifications of triple-negative breast cancer: Improving patient selection for treatment. Cancer Discov. 9:176–198. 2019.PubMed/NCBI View Article : Google Scholar | |
Abramson VG, Lehmann BD, Ballinger TJ and Pietenpol JA: Subtyping of triple-negative breast cancer: Implications for therapy. Cancer. 121:8–16. 2015.PubMed/NCBI View Article : Google Scholar | |
Jhan JR and Andrechek ER: Triple-negative breast cancer and the potential for targeted therapy. Pharmacogenomics. 18:1595–1609. 2017.PubMed/NCBI View Article : Google Scholar | |
Damaskos C, Garmpi A, Nikolettos K, Vavourakis M, Diamantis E, Patsouras A, Farmaki P, Nonni A, Dimitroulis D, Mantas D, et al: Triple-negative breast cancer: The progress of targeted therapies and future tendencies. Anticancer Res. 39:5285–5296. 2019.PubMed/NCBI View Article : Google Scholar | |
Langfelder P and Horvath S: WGCNA: An R package for weighted correlation network analysis. BMC Bioinformatics. 9(559)2008.PubMed/NCBI View Article : Google Scholar | |
Liu H, Sun Y, Tian H, Xiao X, Zhang J, Wang Y and Yu F: Characterization of long non-coding RNA and messenger RNA profiles in laryngeal cancer by weighted gene co-expression network analysis. Aging (Albany NY). 11:10074–10099. 2019.PubMed/NCBI View Article : Google Scholar | |
Yang J, Li C, Zhou J, Liu X and Wang S: Identification of prognostic genes in leiomyosarcoma by gene co-expression network analysis. Front Genet. 10(1408)2019.PubMed/NCBI View Article : Google Scholar | |
Wang WJ, Guo CA, Li R, Xu ZP, Yu JP, Ye Y, Zhao J, Wang J, Wang WA, Zhang A, et al: Long non-coding RNA CASC19 is associated with the progression and prognosis of advanced gastric cancer. Aging (Albany NY). 11:5829–5847. 2019.PubMed/NCBI View Article : Google Scholar | |
Burstein MD, Tsimelzon A, Poage GM, Covington KR, Contreras A, Fuqua SA, Savage MI, Osborne CK, Hilsenbeck SG, Chang JC, et al: Comprehensive genomic analysis identifies novel subtypes and targets of triple-negative breast cancer. Clin Cancer Res. 21:1688–1698. 2015.PubMed/NCBI View Article : Google Scholar | |
Clarke C, Madden SF, Doolan P, Aherne ST, Joyce H, O'Driscoll L, Gallagher WM, Hennessy BT, Moriarty M, Crown J, et al: Correlating transcriptional networks to breast cancer survival: A large-scale coexpression analysis. Carcinogenesis. 34:2300–2308. 2013.PubMed/NCBI View Article : Google Scholar | |
Hatzis C, Pusztai L, Valero V, Booser DJ, Esserman L, Lluch A, Vidaurre T, Holmes F, Souchon E, Wang H, et al: A genomic predictor of response and survival following taxane-anthracycline chemotherapy for invasive breast cancer. JAMA. 305:1873–1881. 2011.PubMed/NCBI View Article : Google Scholar | |
Livak KJ and Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 25:402–408. 2001.PubMed/NCBI View Article : Google Scholar | |
Kim YJ, Sung M, Oh E, Vrancken MV, Song JY, Jung K and Choi YL: Engrailed 1 overexpression as a potential prognostic marker in quintuple-negative breast cancer. Cancer Biol Ther. 19:335–345. 2018.PubMed/NCBI View Article : Google Scholar | |
Castilla M, López-García MA, Atienza MR, Rosa-Rosa JM, Díaz-Martín J, Pecero ML, Vieites B, Romero-Pérez L, Benítez J, Calcabrini A and Palacios J: VGLL1 expression is associated with a triple-negative basal-like phenotype in breast cancer. Endocr Relat Cancer. 21:587–599. 2014.PubMed/NCBI View Article : Google Scholar | |
Lim HK, Rahim AB, Leo VI, Das S, Lim TC, Uemura T, Igarashi K, Common J and Vardy LA: Polyamine regulator AMD1 promotes cell migration in epidermal wound healing. J Invest Dermatol. 138:2653–2665. 2018.PubMed/NCBI View Article : Google Scholar | |
Zabala-Letona A, Arruabarrena-Aristorena A, Martín-Martín N, Fernandez-Ruiz S, Sutherland JD, Clasquin M, Tomas-Cortazar J, Jimenez J, Torres I, Quang P, et al: mTORC1-dependent AMD1 regulation sustains polyamine metabolism in prostate cancer. Nature. 547:109–113. 2017.PubMed/NCBI View Article : Google Scholar | |
Stone L: Prostate cancer: Mechanisms of cancer metabolism: mTORC1 mediates AMD1. Nat Rev Urol. 14(454)2017.PubMed/NCBI View Article : Google Scholar | |
Chen K, Liu H, Liu Z, Luo S, Patz EF Jr, Moorman PG, Su L, Shen S, Christiani DC and Wei Q: Genetic variants in RUNX3, AMD1 and MSRA in the methionine metabolic pathway and survival in nonsmall cell lung cancer patients. Int J Cancer. 145:621–631. 2019.PubMed/NCBI View Article : Google Scholar | |
Xu L, You X, Cao Q, Huang M, Hong LL, Chen XL, Lei L, Ling ZQ and Chen Y: Polyamine synthesis enzyme AMD1 is closely associated with tumorigenesis and prognosis of human gastric cancers. Carcinogenesis. 41:214–222. 2020.PubMed/NCBI View Article : Google Scholar | |
Peluffo G, Subedee A, Harper NW, Kingston N, Jovanović B, Flores F, Stevens LE, Beca F, Trinh A, Chilamakuri CS, et al: EN1 is a transcriptional dependency in triple-negative breast cancer associated with brain metastasis. Cancer Res. 79:4173–4183. 2019.PubMed/NCBI View Article : Google Scholar | |
Bell D, Bell A, Roberts D, Weber RS and El-Naggar AK: Developmental transcription factor EN1-a novel biomarker in human salivary gland adenoid cystic carcinoma. Cancer. 118:1288–1292. 2012.PubMed/NCBI View Article : Google Scholar | |
Miura K, Akashi T, Ando N, Ayabe S, Kayamori K, Namiki T and Eishi Y: Homeobox transcriptional factor engrailed homeobox 1 is expressed specifically in normal and neoplastic sweat gland cells. Histopathology. 72:1199–1208. 2018.PubMed/NCBI View Article : Google Scholar | |
Zecca M and Struhl G: A feed-forward circuit linking wingless, fat-dachsous signaling, and the warts-hippo pathway to drosophila wing growth. PLoS Biol. 8(e1000386)2010.PubMed/NCBI View Article : Google Scholar | |
Mori S, Takeuchi T, Ishii Y and Kukimoto I: The transcriptional cofactor VGLL1 drives transcription of human papillomavirus early genes via TEAD1. J Virol. 94:e01945–e01919. 2020.PubMed/NCBI View Article : Google Scholar | |
Kim BK, Cheong JH, Im JY, Ban HS, Kim SK, Kang MJ, Lee J, Kim SY, Park KC, Paik S and Won M: PI3K/AKT/β-catenin signaling regulates vestigial-like 1 which predicts poor prognosis and enhances malignant phenotype in gastric cancer. Cancers (Basel). 11(1923)2019.PubMed/NCBI View Article : Google Scholar | |
Kundishora AJ, Reeves BC, Nelson-Williams C, Hong CS, Gopal PP, Snuderl M, Kahle KT and Erson-Omay EZ: Novel EWSR1-VGLL1 fusion in a pediatric neuroepithelial neoplasm. Clin Genet. 97:791–792. 2020.PubMed/NCBI View Article : Google Scholar |