|Year : 2014 | Volume
| Issue : 4 | Page : 1013-1018
Bioinformatics analysis of aggressive behavior of breast cancer via an integrated gene regulatory network
Xingwang Yang1, Mingguang Jia1, Zhaodong Li2, Shiyong Lu1, Xiangjie Qi1, Bo Zhao1, Xiaoming Wang1, Yu Rong1, Jian Shi3, Zhijun Zhang1, Weizhi Xu1, Yujun Gao1, Shuliang Zhang1, Gang Yu1
1 Department of Surgery, People's Hospital of Linzi District, Affiliated to Binzhou Medical College, Zibo, China
2 Department of Radiology, People's Hospital of Linzi District, Affiliated to Binzhou Medical College, Zibo, China
3 Department of Radiation Oncology, People's Hospital of Linzi District, Affiliated to Binzhou Medical College, Zibo, China
|Date of Web Publication||9-Jan-2015|
Department of Radiation Oncology, People's Hospital of Linzi District, Affiliated to Binzhou Medical College, Zibo 255400
Source of Support: None, Conflict of Interest: None
Background: Breast cancer is one of the most frequently diagnosed cancers in women. Though death from this disease is mainly caused by the metastases of the aggressive cancer cells, few studies have expounded the aggressive behavior of breast cancer.
Materials and Methods: We downloaded the gene expression profiles of GSE40057, including four aggressive and six less-aggressive breast cancer cell lines, from Gene Expression Omnibus and identified the differentially expressed genes (DEGs) between the aggressive and less-aggressive samples. An integrated gene regulatory network was built including DEGs, microRNAs (miRNAs), and transcription factors. Then, motifs and modules of the network were identified. Modules were further analyzed at a functional level using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway to study the aggressive behavior of breast cancer.
Results: A total of 764 DEGs were found and two modules were filtered from the integrated gene regulatory network. Totally two motifs and modules for DEGs were identified. Significant GO terms associated with cell proliferation and hormone stimulus of the modules were found and the target genes identified were CAV1, CD44, and TGFβR2. The KEGG pathway analysis discovered that CAV1 and FN1 were significantly enriched in focal adhesion, extracellular matrix (ECM)-receptor interaction, and pathways in cancer.
Conclusion: Aggressive behavior of breast cancer was proved to be related to cell proliferation and hormone stimulus. Genes such as CAV1, CD44, TGFβR2, and FN1 might be potential targets to diagnose the aggressive behavior of breast cancer cells.
Keywords: Aggressive behavior, breast cancer, integrated gene regulatory network
|How to cite this article:|
Yang X, Jia M, Li Z, Lu S, Qi X, Zhao B, Wang X, Rong Y, Shi J, Zhang Z, Xu W, Gao Y, Zhang S, Yu G. Bioinformatics analysis of aggressive behavior of breast cancer via an integrated gene regulatory network. J Can Res Ther 2014;10:1013-8
|How to cite this URL:|
Yang X, Jia M, Li Z, Lu S, Qi X, Zhao B, Wang X, Rong Y, Shi J, Zhang Z, Xu W, Gao Y, Zhang S, Yu G. Bioinformatics analysis of aggressive behavior of breast cancer via an integrated gene regulatory network. J Can Res Ther [serial online] 2014 [cited 2020 Jun 5];10:1013-8. Available from: http://www.cancerjournal.net/text.asp?2014/10/4/1013/137971
| > Introduction|| |
Breast cancer is one of the most common malignancies in women and is associated with a high mortality rate. , Distant metastases are the main cause of the death. Clinical data have shown that 10-15% patients with breast cancer have an aggressive disease and develop distant metastases within 3 years after the initial detection of the primary tumor.  However, the appearance of metastases at distant sites 10 years or more after the initial diagnosis is also not unusual. Thus, the indeterminate occurrence of metastases is a serious problem of breast cancer. Aggressiveness is the major feature of the metastasis; therefore, understanding the underlying mechanism of the aggressive behavior is one way to comprehend the distant metastases of breast cancer.
Aggressive cell lines are the basic and foremost materials to study the molecular biology of breast cancer metastasis. Zajchowski et al. found that keratin 19 was consistently up-regulated in less-aggressive cell lines and vimentin and fos-related antigen-1 were consistently overexpressed in the more highly aggressive cell lines.  Besides, genes such as ERBB2, EZH2, and HER2 were also analyzed and found to be overexpressed in the invasion and metastasis of the carcinomas. ,, Recent studies have shown that microRNAs (miRNAs), a class of small non-coding RNA molecules, are involved in cancer proliferation and metastasis.  For example, miR-31 was down-regulated, while miR-10b was up-regulated in metastatic breast tumors.  By using miRNA expression profiling, Luo et al. identified seven down-regulated miRNAs and four up-regulated miRNAs in invasive cell lines when compared with normal and less invasive cell lines.  They also demonstrated that exogenous miR-200c could significantly down-regulate cofilin 2 which was overexpressed in aggressive breast cancer cell lines. Meanwhile, lost or low expression of transcription factors (TFs) including RE1 silencing TF and nuclear activator protein 2 were found to be associated with the aggressiveness of breast cancer. , It has been increasingly recognized that systemic investigations of the regulation relationships among miRNAs, TFs, and target genes that regulate cell migration and aggressiveness are highly needed.
In this study, we aimed to identify the differentially expressed genes (DEGs) in the aggressive breast cancer cell lines, and to build a gene regulatory network with the miRNAs and TFs, based on the DEGs. Motifs and modules were found and investigated at a functional level according to the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, which could provide a basis for analyzing the mechanism of the aggressive behavior.
| > Materials and methods|| |
Gene expression data sources
Expression profiling by array of GSE40057 was downloaded from Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) which was contributed by Luo et al. The platform used was GPL570 [(HG-U133_Plus_2) Affymetrix Human Genome U133 Plus 2.0 Array]. In this work, the samples consisted of a total of 10 breast cancer cell lines including 4 aggressive cell lines (MDA-MB-231, HS578T, BT549, and SUM159) and six less-aggressive cell lines (BT474, MDA-MB-468, T47D, ZR-75-1, MCF7, and SK-BR3). The data of less-aggressive cell lines were used as the control.
Data preprocessing and analysis
R software (v. 2.15.3)  was utilized to deal with the data. The normalization of different array data was firstly performed with Robust Multichip Averaging (RMA) method.  Then, the Linear Models for Microarray Data (LIMMA) package was used to identify DEGs between the aggressive cell lines and less-aggressive cell lines. Bayesian analysis was applied to circumvent the multi-test problem and the raw P values were adjusted into false discovery rate (FDR) by the Benjamin and Hochberg (BH) method. , The FDR <0.05 and log FC/ >1 were used as the cut-off criteria to select the DEGs.
Construction of the gene regulatory networks
TFs and miRNAs related to the DEGs were collected from Transcriptional Regulatory Element Database and microRNA regulation databases (including miRecords, TarBase, starbase, and miR2Disease), respectively. ,,,, Then, the Cytoscape software  was used to construct the integrated gene regulatory network.
Identification of network motifs
Network motif is an important local property of networks. Network motif is the recurrent and statistically significant sub-graph or pattern, which assists researchers in the identification of functional units in the networks.  In this work, motifs of the network were identified using FANMOD software,  with the cut-off criteria of P < 0.05 and Z-score >2.
Functional analysis of modules
Modules are the semi-independent transcriptional units clustered of the motifs in the regulatory network. In this work, the modules of the motifs were filtered from the network using Perl script. Then, GO and KEGG pathway terms of the modules were collected and analyzed by the online tool of Database for Annotation, Visualization, and Integrated Discovery (DAVID).  The DEGs' number being larger than 2 and a P value less than 0.05 were chosen as the cut-off criteria.
| > Results|| |
DEGs and construction of the network
According to the significant threshold of /log FC/ >1 and FDR < 0.05, 764 DEGs were found between aggressive cell lines and less-aggressive cell lines. After the collection of miRNAs and TFs which were related to the DEGs, an miRNA-DEG-TF network [Figure 1] was established by Cytoscape software. There were 731 nodes and 2605 edges in the regulatory network.
|Figure 1: The integrated miRNA-DEG-TF network. The blue dots, yellow box, and green triangle represent miRNA, differentially expressed gene (DEG), and transcription factor (TF), respectively|
Click here to view
Identification of network motifs and the modules
As the network is too large to yield any interesting information, it is necessary to divide the regulatory network into connected sub-networks that might represent functional modules [Figure 2]. On analyzing the miRNA-DEG-TF network via FANMOD software, two motifs were found with the cut-off criteria of P < 0.05 and Z-score >2. The motifs are clearly shown in [Table 1] and are named as motif 1 and motif 2. The two motifs had a similar geometrical structure, but had different signs of regulation and represented distinct kinds of regulatory interactions, such as TF-gene regulation (motif 1 in [Table 1]) and miRNA-gene regulation (motif 2 in [Table 1]).
|Figure 2: The two modules in the miRNA-DEG-TF network. Module 1 includes 515 nodes and 2176 edges. Meanwhile, module 2 consists of 70 nodes and 161 edges. In module 1, the yellow and blue dots represent DEGs and miRNAs, respectively. In module 2, the yellow and green dots represent DEGs and TFs, respectively|
Click here to view
GO clustering and analysis
To investigate the effect of modules on the aggressiveness of breast cancer at a more functional level, DEGs of two modules were classified into GO terms. It is clearly seen that DEGs in module 1 were associated with 18 biological processes, and the regulation of cell proliferation (P = 7.27E-12) was one of the most significant GO terms [Table 2], while the GO clusters enrichment of module 2 was mostly correlated to terms such as response to organic substance (P = 9.66E-12) and response to hormone stimulus (P = 2.59E-08) [Table 3]. Significant DEGs such as CAV1, CD44, and Transforming Growth Factor Beta Receptor II (TGFβR2) were found in these functional terms.
Pathway enrichment analysis of the modules
To gain further insight into the function of DEGs in the modules, we used the online biological classification tool DAVID and observed significant enrichment of these genes in multiple KEGG terms [Table 4]. According to the cut-off criteria of DEGs' number being larger than 2 and P value less than 0.05, the enriched pathways of both the modules included pathway in cancer, focal adhesion, and extracellular matrix (ECM)-receptor interaction. Target genes such as CAV1and FN1 were found from these pathways. Additionally, the pathways enriched in module 1 also included ATP-binding cassette (ABC) transporters and gap junction. Furthermore, regulation relationships of miR222-Cav1-TP53 and miR200c-FN1-CREB1 could be observed from the miRNA-DEG-TF network [Figure 1].
| > Discussion|| |
Previous reports have shown that eukaryotic gene regulation is performed at multiple levels. , As the regulators mediating transcription and post-transcription, TFs and miRNAs can construct an integrated gene regulatory network with their interacting target genes.  It has been proved that network analysis is useful in unraveling the complexity of biological regulation. , In the present study, we used the gene expression profile of GSE40057 downloaded from GEO and identified 764 DEGs in aggressive breast cancer cell lines. Then, the integrated miRNA-DEG-TF regulatory network, which was the heart of our study, was constructed. Two significant modules of the network were identified and analyzed at a functional level in order to illustrate the biological regulations related to breast cancer aggressiveness.
The result of miRNA-DEG-TF network construction in aggressive breast cancer cell lines reveals that genes have been linked with a number of miRNAs and TFs by our method. CAV1, CD44, TGFβR2, and FN1 were the key genes obtained as the outcome of GO and KEGG pathway of the modules from the network.
Among the target genes identified in the integrated gene regulatory network, CAV1 was regulated by a number of miRNAs and TFs. It is demonstrated that the protein encoded by CAV1 is a putative target of miR-222, which could strengthen the proliferative signals.  Recent studies have also shown that CAV1 depletion is strictly TP53 dependent through the AMP-activated protein kinase (AMPK)-TP53/p53 signaling. , The evidence confirmed the existence of the miR222-CAV1-TP53 regulation which was predicted in the network. Therefore, the regulatory relation between CAV1 and the linked miRNAs such as miR-101 and miR-128 in the network, as well as the linked TFs of E2F4 and SP1, might exist in the aggressive breast cancer cell. Moreover, it is applicable to the other genes and their linked miRNAs or TFs.
CAV1 encodes the scaffolding protein which is the main component of the caveolae plasma membranes found in most cell types.  A mutant form of CAV1 (P132L) was found in up to 16% of human breast cancers and behaved in a dominant negative manner.  Research indicated that the growth of breast cancer cells could be inhibited by the overexpression or re-expression of CAV1.  There is also accumulating evidence showing that CAV1 has an antiproliferative function, and the mutant CAV1 expression seems to alter the actin networks in the cells and promotes the aggressive activity. , CD44 is another target gene found in the integrated miRNA-DEG-TF regulatory network. The protein encoded by CD44 is a cell-surface glycoprotein involved in cell-cell interactions, cell adhesion and migration, and participates in the cellular functions of tumor metastasis.  Moreover, it was suggested that the expression of CD44 was highly correlated with the invasive potential of the breast tumor cells by degrading the hyaluronan.  Hence, CAV1 and CD44 were suggested to be the target genes associated with the aggressiveness of breast cancer.
According to the GO enrichment analysis, most of the DEGs in module 1 were enriched in the GO terms associated with regulation of cell proliferation. Interestingly, GO terms related to hormone stimulus were enriched in both module 1 and module 2. This result suggests that cell proliferation and hormone stimulus might play important roles in the aggressive activities of breast cancer. Our analysis results are in line with previous studies. TGFβR2 is proved to encode a transmembrane protein which could phosphorylate proteins related to cell proliferation.  Additionally, breast cancer frequently subverted the tumor-suppressing function of TGF-®, leading to its conversion from an inhibitor to a stimulator of breast cancer growth, invasion, and metastasis.  Meanwhile, it has been proved that ribosome biogenesis is a basic cellular process intimately linked to cell growth and proliferation, and it was up-regulated in most of the cancers, especially in aggressive cancers.  A study has indicated that hormone (estrogen and androgen) treatment in breast cancer could positively regulate rRNA synthesis of ribosome biogenesis.  Besides, steroid hormone estrogen receptor and androgen receptor have been shown to be potentiated by CAV1 in their transcriptional activities. 
In pathway analysis of the two modules, we identified a total of five dysfunctional pathways in the aggressive breast cancer cell lines. Some of the identified pathways including focal adhesion and ECM-receptor interaction are consistent with our knowledge of the aggressiveness of breast cancer. CAV1 was a significant DEG enriched in the focal adhesion pathway, which played essential roles in important biological processes including cell motility, cell proliferation, and cell survival.  Interestingly, FN1 was enriched in both the pathways of the two modules. It was pointed out that FN1 played a key role in the regulation of adhesion, migration, and metastasis of tumors, and could be inhibited by a TF of CREB1.  Recently, it was also reported that the inhibition of FN1 by miR-200c resulted in the suppression of cell migration.  These results suggest that FN1 was related to the aggressive behavior of breast cancer, and the regulation of miR200c-FN1-CREB1 existing in the network was confirmed.
In summary, the regulation of miR222-CAV1-TP53 and miR200c-FN1-CREB1 existing in our miRNA-DEG-TF network was confirmed. We also found the aggressiveness of breast cancer was associated with cell proliferation and hormone stimulus. Considering the GO and KEGG pathway analyses, it was also evident that CAV1, CD44, TGFβR2, and FN1 might be potential target genes for detecting the aggressiveness of breast cancer cells. However, there are some limitations in our study. More data on the expression of different aggressive breast cancer cell lines and different stages should be included, and further experiments are still needed to confirm our observation.
| > References|| |
Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. Cancer statistics, 2003. CA Cancer J Clin 2003;53:5-26.
Debily MA, El Marhomy S, Boulanger V, Eveno E, Mariage-Samson R, Camarca A, et al
. A functional and regulatory network associated with PIP expression in human breast cancer. PloS One 2009;4:e4696.
Weigelt B, Peterse JL, van 't Veer LJ. Breast cancer metastasis: Markers and models. Nat Rev Cancer 2005;5:591-602.
Zajchowski DA, Bartholdi MF, Gong Y, Webster L, Liu HL, Munishkin A, et al
. Identification of gene expression profiles that predict the aggressive behavior of breast cancer cells. Cancer Res 2001;61:5168-78.
Hayes DF, Trock B, Harris AL. Assessing the clinical impact of prognostic factors: When is "statistically significant" clinically useful? Breast Cancer Res Treat 1998;52:305-19.
Kleer CG, Cao Q, Varambally S, Shen R, Ota I, Tomlins SA, et al
. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci USA 2003;100:11606-11.
Vogel CL, Cobleigh MA, Tripathy D, Gutheil JC, Harris LN, Fehrenbacher L, et al
. Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. J Clin Oncol 2002;20L719-26.
Ventura A, Jacks T. MicroRNAs and cancer: Short RNAs go a long way. Cell 2009;136:586-91.
Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature 2007;449:682-8.
Luo D, Wilson JM, Harvel N, Liu J, Pei L, Huang S, et al
. A systematic evaluation of miRNA: mRNA interactions involved in the migration and invasion of breast cancer cells. J Transl Med 2013;11:57.
Pellikainen J, Kataja V, Ropponen K, Kellokoski J, Pietiläinen T, Böhm J, et al
. Reduced nuclear expression of transcription factor AP-2 associates with aggressive breast cancer. Clin Cancer Res 2002;8:3487-95.
Wagoner MP, Gunsalus KT, Schoenike B, Richardson AL, Friedl A, Roopra A. The transcription factor REST is lost in aggressive breast cancer. Plos Genet 2010;6:e1000979.
Team R. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2007. ISBN 3-900051-07-0, 2012.
Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, et al
. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003;4:249-64.
Smyth GK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004;3:3.
Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B (Methodological) 1995;57: 289-300.
Jiang C, Xuan Z, Zhao F, Zhang MQ. TRED: A transcriptional regulatory element database, new entries and other development. Nucleic Acids Res 2007;35:D137-40.
Xiao F, Zuo Z, Cai G, Kang S, Gao X, Li T. miRecords: An integrated resource for microRNA-target interactions. Nucleic Acids Res 2009;37:D105-10.
Sethupathy P, Corda B, Hatzigeorgiou AG. TarBase: A comprehensive database of experimentally supported animal microRNA targets. RNA 2006;12:192-7.
Yang JH, Li JH, Shao P, Zhou H, Chen YQ, Qu LH. starBase: A database for exploring microRNA-mRNA interaction maps from Argonaute CLIP-Seq and Degradome-Seq data. Nucleic Acids Res 2011;39:D202-9.
Jiang Q, Wang Y, Hao Y, Juan L, Teng M, Zhang X, et al
. miR2Disease: A manually curated database for microRNA deregulation in human disease. Nucleic Acids Res 2009;37:D98-104.
Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al
. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13:2498-504.
Masoudi-Nejad A, Schreiber F, Kashani Z. Building blocks of biological networks: A review on major network motif discovery algorithms. IET Syst Biol 2012;6:164-74.
Wernicke S, Rasche F. FANMOD: A tool for fast network motif detection. Bioinformatics 2006;22:1152-3.
Da Wei Huang BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2008;4:44-57.
Chen K, Rajewsky N. The evolution of gene regulation by transcription factors and microRNAs. Nat Rev Genet 2007;8:93-103.
Wray GA, Hahn MW, Abouheif E, Balhoff JP, Pizer M, Rockman MV, et al
. The evolution of transcriptional regulation in eukaryotes. Mol Biol Evol 2003;20:1377-419.
Cheng C, Yan KK, Hwang W, Qian J, Bhardwaj N, Rozowsky J, et al
. Construction and analysis of an integrated regulatory network derived from high-throughput sequencing data. Plos Comput Biol 2011;7:e1002190.
Barabási AL, Oltvai ZN. Network biology: Understanding the cell's functional organization. Nat Rev Genet 2004;5:101-13.
Stites EC, Trampont PC, Ma Z, Ravichandran KS. Network analysis of oncogenic Ras activation in cancer. Science 2007;318:463-7.
Chen WX, Hu Q, Qiu MT, Zhong SL, Xu JJ, Tang JH, et al
. miR-221/222: Promising biomarkers for breast cancer. Tumor Biol 2013;34:1361-70.
Ha TK, Chi SG. CAV1/caveolin 1 enhances aerobic glycolysis in colon cancer cells via activation of SLC2A3/GLUT3 transcription. Autophagy 2012;8:1684-5.
Sherif ZA, Sultan AS. Divergent control of Cav-1 expression in non-cancerous Li-Fraumeni syndrome and human cancer cell lines. Cancer Biol Ther 2013;14:29-38.
Okamoto T, Schlegel A, Scherer PE, Lisanti MP. Caveolins, a family of scaffolding proteins for organizing "preassembled signaling complexes" at the plasma membrane. J Biol Chem 1998;273:5419-22.
Hayashi K, Matsuda S, Machida K, Yamamoto T, Fukuda Y, Nimura Y, et al
. Invasion activating caveolin-1 mutation in human scirrhous breast cancers. Cancer Res 2001;61:2361-4.
Lee SW, Reimer CL, Oh P, Campbell DB, Schnitzer JE. Tumor cell growth inhibition by caveolin re-expression in human breast cancer cells. Oncogene 1998;16:1391-7.
Williams TM, Lisanti MP. The caveolin proteins. Genome Biol 2004;5:214.
Orian-Rousseau V. CD44, a therapeutic target for metastasising tumours. Eur J Cancer 2010;46:1271-7.
Culty M, Shizari M, Thompson EW, Underhill CB. Binding and degradation of hyaluronan by human breast cancer cell lines expressing different forms of CD44: Correlation with invasive potential. J Cell Physiol 1994;160:275-86.
Huang SS, Huang JS. TGF-® control of cell proliferation. J Cell Biochem 2005;96:447-62.
Galliher AJ, Schiemann WP. Src phosphorylates Tyr284 in TGF-® type II receptor and regulates TGF-® stimulation of p38 MAPK during breast cancer cell proliferation and invasion. Cancer Res 2007;67:3752-8.
Lempiäinen H, Shore D. Growth control and ribosome biogenesis. Curr Opin Cell Biol 2009;21:855-63.
Ray S, Johnston R, Campbell DC, Nugent S, McDade SS, Waugh D, et al
. Androgens and estrogens stimulate ribosome biogenesis in prostate and breast cancer cells in receptor dependent manner. Gene 2013;526:46-53.
Lu ML, Schneider MC, Zheng Y, Zhang X, Richie JP. Caveolin-1 interacts with androgen receptor A positive modulator of androgen receptor mediated transactivation. J Biol Chem 2001;276:13442-51.
Drosten M, Dhawahir A, Sum EY, Urosevic J, Lechuga CG, Esteban LM, et al
. Genetic analysis of Ras signalling pathways in cell proliferation, migration and survival. EMBO J 2010;29:1091-104.
Chen J, Xia J, Wang S, Wei Y, Wu J, Huang G, et al
. Study on transcription regulation network in rheumatoid arthritis via bioinformatics analysis. Afr J Pharm Pharmacol 2012;6:1435-43.
Howe EN, Cochrane DR, Richer JK. Targets of miR-200c mediate suppression of cell motility and anoikis resistance. Breast Cancer Res 2011;13:R45.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]