Combined analysis of gene expression and genome binding profiles identified potential therapeutic targets of ciclopirox in Ewing sarcoma
Abstract
Introduction
Ewing sarcoma (ES) is the second most common malignant bone tumor, with the highest morbidity rates in children aged 10–20 years and with undistinguishable features from another Ewing family tumor, primitive neuroectodermal tumor (PNET) (1,2). Surgery, radiation and chemotherapy make up the standard treatments of ES; although there has been significant improvement in prognosis for localized ES, the survival rates of metastatic patients are low with only ~13–30% 5-year survival rate (3,4). Early diagnosis would aid in improving the prognosis of patients with ES, whereas traditional pathological diagnosis has a number of limitations, such as subjectivity and an absence of biological markers (5). Besides, there are ~200 new cases among children and 20 new cases among adults in USA per year. Therefore, it is necessary to develop novel diagnostic and treatment methods to reduce ES morbidity and mortality.
The development of genomic technologies has provided unprecedented opportunities to detail the progression of ES and biomarker identification. A common biomarker target in ES is Ewing sarcoma breakpoint region 1-Friend leukemia integration 1 (EWS-FLI1), a fusion gene induced by t(11;22) (q24;q12) chromosome translocation (6–8), which has been reported as an independent prognosis determinant in ES (9). EWS-FLI1 may directly activate or repress multi gene enhancer elements associated with ES through different chromatin remodeling mechanisms that may promote tumor cell growth and proliferation (10). In addition, as a transcription factor, EWS-FLI1 binds to various genes and determines drug response (11); and based on this information several small bioactive molecules were developed for the treatment of ES (12–14). Ciclopirox (CPX) is an antifungal drug that isoften used for the treatment of fungal-induced skin and nail diseases (15–17). A previous study reported on the antitumoral activity of CPX in human breast cancer through a series molecular level analyses, which may indicate its potential use in cancer treatment (18). Another study demonstrated the inhibitory activity of CPX on mammalian target of rapamycin, which is useful for promoting parthenolide anti leukemia activity (19). However, there are few studies concerning the roles of CPX in ES.
In the present study, the roles of CPX in ES were explored through the combined analysis of genome and transcript me datasets, and for the first time, to the best of our knowledge, combined analysis of a CPX-regulated gene expression profile and an EWS-FLI1 genome-binding profile was conducted. The transcript me datasets comprised oxicycl in induced EWS-FLI1-knockdown and wild-type ES samples (A673 cell lines without any treatment), which aided in identifying targets that may be regulated by EWS-FLI1. The genome datasets are chromatin immune precipitation coupled with high-throughput sequencing (ChIP-seq) data of EWS-FLI1 in CPX-treated A673 cell line. Through the genome datasets, we should identify CPX-specific targets mediated by EWS-FLI1. Therefore, the combined analysis of the two datasets may be able to obtain more reliable targets of CPX in ES. In addition, the systemic bioinformatics analysis performed in the present study, which included functional and network analysis, may aid our understanding of the underlying mechanisms of ES and to improve prognosis.
Materials and methods
High-throughput datasets
The present study used three high-throughput genome and transcript me datasets that were downloaded from the Gene Expression Omnibus online repository (
Gene expression profiles analysis
Raw gene expression profiles were normalized according to the method described by Yang et al (22). Probesets were translated to gene symbol through the gene microarray annotation file, and the expression values for genes corresponding to multi-probesets were summarized. The differentially expressed genes (DEGs) in EWS-FLI1-knockdown and CPX-treated A673 cell lines were identified using the linear models for microarray and RNA-sequencing data (limma;
Functional enrichment analysis
To investigate the biological processes that may be influenced by EWS-FLI1 or CPX in ES, gene ontology (GO) term functional enrichment analysis was conducted for the DEGs identified from EWS-FLI1-knockdown and CPX-treated A673 cell lines through the Web Gestalt (23) online database. In this study, only the biological processes that satisfied the criteria of P<0.05 and hits count >5 were obtained. For the interpretation of significantly enriched GO terms, we submitted them to REVIGO online tool (24) with the default thresholds, i.e. medium similarity (0.7) between GO terms.
Identification of EWS-FLI1 specific binding sites
The present study identified EWS-FLI1-specific binding sites, which were referred as peaks, through the analysis of EWS-FLI1 ChIP-seq datasets. First, quality control was performed for the raw datasets using Fast QC (
Network analysis
Genes shared between DEGs of EWS-FLI1-knockdown and CPX-treated A673 cell lines, which were referred as DECs, were considered as common targets of EWS-FLI1 and CPX in ES. To obtain more reliable results, DECs and genes that were specifically bound by EWS-FLI1 in their promoters were intersected, and the overlaps were considered as reliable EWS-FLI1 targets (EFTs) in A673 cells. The potential interaction pairs among EFTs were obtained through the STRING database (29) with the minimum combined score >0.4. In addition, the predicted regulatory relationships between EWS-FLI1 and the EFTs, and interaction pairs among the EFTs were visualized with Cy to scape software (30).
Results
Differential expression analysis
For the GSE27524EWS-FLI1-knockdown A673 cell lines, 5 lists of DEGs were identified for the 5 time points (18, 36, 53, 72 and 96 h) when they were compared with the 0 h cells following EWS-FLI1 knockdown, and these were referred as DEG18, DEG36, DEG53, DEG72 and DEG96, respectively. Volcano plots of the corresponding fold change (log2 scale) against the transformed (-log10) Q-values for every time point are illustrated in Fig. 1A-E. All volcano plots complied with normal distribution with fold-change values ranging between-6.4 and 12.0 for both downregulated and upregulated genes. The volcano plot changes as the length of time increases post-knockdown. The 18 h plot appears similar to the 36 h plot, whereas the later time points of 53, 72 and 96 h are more similar to each other, particularly in the upregulated expression area (right side of volcano plots). This observation was probably due to the knockdown of EWS-FLI1 resulting in upregulated expression of tumor suppressor genes. Unsupervised clustering based on the expression profiles of the five lists of DEGs obtained consistent pattern with volcano plots (Fig. 1F); two main clusters are obtained: 0 and 18 h, 36, 53, 72 and 96 h. The two main clusters were each further subdivided according to different time points.
The relationships among different groups of DEGs were also examined by creating Venn diagrams with overlapping regions that represent the overlapping DEGs (Fig. 2). The gene contents were more similar among the later time points, in which 973 overlapped genes were identified among the 36, 72 and 96 h DEGs. For CPX-treated A673 cell lines, a total of 466 DEGs (referred as DEG-CPX) were obtained, which comprised 167 downregulated genes and 299 upregulated genes.
Enriched functions
The 5 lists of DEGs that were identified in the EWS-FLI1-knockdown A673 cell lines were combined and the duplicates were removed, which resulted in 4,500 unique DEGs. A total of 635 biological process (BP) GO terms with minimum hits >5 were indicated to be significantly enriched in those genes. The 466 identified DEG-CPX DEGs were closely associated with 182 BP GO terms. The GO terms present level information based on GO database (
EWS-FLI1 specific binding sites
The present study used MACS version 2 to determine the binding profiles of EWS-FLI1 in ES cell lines, in which binomial distribution is used to testing enrichment significance of the binding sites. ChIP-seq analysis identified 3,885 peaks for EWS-FLI1, most of which mapped to the distal intergenic region (Fig. 4A); however, the strongest binding signals were detected in the promoter region (Fig. 4B), which indicated more reliable targets of EWS-FLI1. DECs were intersected with genes that contained EWS-FLI1 binding peaks in their promoters and 34 overlaps, that is EFTs. Table I shows the CPX-induced fold change (log2 scale) of the 34 EFTs, of which 8of them were revealed to be downregulated by CPX treatment. To illustrate the binding profiles of EWS-FLI1 on the 34 EFTs, 3 targets were randomly selected, collagen type I α2 chain (COL1A2), N-myc proto-oncogene and transforming growth factor β1 (TGFβ1), and the EWS-FLI1 binding signals on them were visualized using the Integrative Genomics Viewer (IGV). The results demonstrated that certain parts of these three targets exhibited strong binding signals in their promoter regions (Fig. 4C). For MYCN, a strong binding signal was obtained across the whole gene body (Fig. 4C).
Network analysis
The 34 EFTs were considered as potential therapeutic targets of CPX in ES and, as such, it was hypothesized that they may interact with each other to promote or suppress the progression of ES. A total of 10 interaction pairs were identified among the 34 EFTs, and a combined network of the 10 interaction pairs and the regulatory relationships between EWS-FLI1 and the 34 EFTs was created (Fig. 5).
Discussion
In 1973, Dittmar and Lohaus (31) first reported the antifungal functions of CPX, and it has since been widely used to treat fungal-induced skin and nail diseases (16,32,33). In addition, a number of subsequent studies have also demonstrated the antitumoral activity of CPX, particularly in hematologic malignancies (17,34); however, only a few studies were focused the activity of CPX on solid tumors (35,36). The present study conducted combined analysis of gene expression and genome binding profiles of CPX and EWS-FLI1 to screen potential valuable therapeutic targets of CPX in ES. These results may be helpful for the comprehensive understanding of the roles of CPX and EWS-FLI1 in ES development.
EWS-FLI1 is a fusion transcript structure caused by chromosome translocation of t(11;22) (q24;q12), which was previously demonstrated to be closely associated with the initiation and progression of ES (11,12,37,38). In the present study, the combination of gene expression profiles of CPX-treated and EWS-FLI1-knockdown ES A673 cell lines was used to identify genes that were influenced by CPX treatment and EWS-FLI1, that is, DECs. To make the results more reliable, the potential binding targets of EWS-FLI1 in A673 cell lines were identified through ChIP-seq analysis and these targets were intersected with the DECs; the resulting overlaps, the EFTs, should be CPX-targeted genes in ES through EWS-FLI1. Network analysis identified COL1A2, TWIST1 and TGFβ1 as core nodes for they have more direct neighborhoods. Consistent with the present study results, these three genes were previously demonstrated to be regulated by EWS-FLI1 and interacted with each other (39–42). TGFβ1 promotes expression of COL1A2 and is involved in the progression of several diseases, including lung fibrosis (43), skin fibrosis (44) and dermal fibrosis (45), and the inhibited expression of TGFβ-type genes by EWS-FLI1 is an important cause of ES initiation (46). Treatment with CPX also resulted in the upregulation of TGFβ1, which indicated a putative role of TGFβ1 in ES.
DEGs that were influenced by EWS-FLI1-knockdown were closely associated with BPs related to cell activity and DNA replication, which may affect the progression of numerous forms of cancer. DEGs in the CPX-treated A673 cell lines were associated with tissue development, particularly vascular development. Vasculature is important for oxygen and nutrient transportation for cancer cells (47–49). TWIST1 is one of the core genes in the network and serves a role in embryonic vascular development and tumor metastasis, and it is a direct target of hypoxia-inducible factor-2α (HIF2α) (50,51). HIF2α is closely associated with oxygen metabolism in cancer, which is associated with cancer progression (52,53). In the present study, TWIST1 is a core gene of the network and it was previously reported to be a direct target of HIF2α, so it may be important for the progression of ES. Therefore, CPX may influence the progression of ES by controlling the process of vascular development. In this study, MYCN was identified to be a target of EWS-FLI1 with significantly enriched peaks across the gene body. MYCN is a member of the MYC family and encodes a protein with a basic helix-loop-helix (bHLH) domain. The amplification of MYCN is closely associated with a number of cancer types, including ES (54). The present study demonstrated that CPX could affect progression of ES by regulating the binding of EWS-FLI1 in MYCN for the enrichment of EWS-FLI1 binding sites in MYCN in CPX-treated A673 cell line.
In conclusion, the present study performed systemic analysis of gene expression and genome binding profiles in ES A673 cell lines. The results may shed light on the influence of EWS-FLI1 on ES development and progression, and may provide a novel therapeutic method, such as CPX-target therapy, which may aid in improving the prognosis of patients with ES.
Gene | FC (log2 scale) |
---|---|
GYG2 | −1.74 |
ZNF423 | −1.74 |
COL1A2 | −1.69 |
MN1 | −1.30 |
TRAF3IP2 | −1.23 |
MAP2K5 | −1.10 |
SLITRK5 | −1.03 |
LMO3 | −1.02 |
JMJD1C | 1.02 |
DKK2 | 1.06 |
TWIST1 | 1.06 |
TLE4 | 1.07 |
CLDN1 | 1.08 |
MYCN | 1.09 |
LYPD1 | 1.14 |
TGFBI | 1.21 |
SPRY2 | 1.23 |
CTH | 1.24 |
LAMB3 | 1.25 |
DNAJC12 | 1.26 |
MMP10 | 1.28 |
PLOD2 | 1.28 |
FN1 | 1.29 |
TMEM45A | 1.42 |
KLF6 | 1.44 |
TSC22D3 | 1.56 |
SLC2A3 | 1.58 |
COL11A1 | 1.78 |
NDRG1 | 1.93 |
RFTN1 | 2.01 |
ARC | 2.14 |
TNFAIP3 | 2.22 |
KLF5 | 2.42 |
STC1 | 2.50 |
References
- 1. The Ewing family of tumors. Ewing's sarcoma and primitive neuroectodermal tumorsPediatr Clin North Am4499110041997[PubMed][Google Scholar]
- 2. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: Surveillance epidemiology and end results dataJ Pediatr Hematol Oncol304254302008[PubMed][Google Scholar]
- 3. Proceedings: Disease-free survival in children with Ewing's sarcoma treated with radiation therapy and adjuvant four-drug sequential chemotherapyCancer333843931974[PubMed][Google Scholar]
- 4. The effect of surgery with radiation on pelvic Ewing sarcoma survivalJ Surg Oncol1128618652015[PubMed][Google Scholar]
- 5. The genomic landscape of the Ewing Sarcoma family of tumors reveals recurrent STAG2 mutationPLoS Genet10e10044752014[PubMed][Google Scholar]
- 6. EWS/FLI1-induced manic fringe renders NIH 3T3 cells tumorigenicNat Genet174954971997[PubMed][Google Scholar]
- 7. Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumoursNature3591621651992[PubMed][Google Scholar]
- 8. EWS-Fli1 antisense oligodeoxynucleotide inhibits proliferation of human Ewing's sarcoma and primitive neuroectodermal tumor cellsJ Clin Invest992392471997[PubMed][Google Scholar]
- 9. EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewing's sarcomaJ Clin Oncol16124812551998[PubMed][Google Scholar]
- 10. EWS-FLI1 utilizes divergent chromatin remodeling mechanisms to directly activate or repress enhancer elements in Ewing sarcomaCancer Cell266686812014[PubMed][Google Scholar]
- 11. SLFN11 Is a transcriptional target of EWS-FLI1 and a determinant of drug response in Ewing sarcomaClin Cancer Res21418441932015[PubMed][Google Scholar]
- 12. High-throughput RNAi screen in Ewing sarcoma cells identifies leucine rich repeats and WD repeat domain containing 1 (LRWD1) as a regulator of EWS-FLI1 driven cell viabilityGene5961371462017[PubMed][Google Scholar]
- 13. Suppression of FOXO1 is responsible for a growth regulatory repressive transcriptional sub-signature of EWS-FLI1 in Ewing sarcomaOncogene33392739382014[PubMed][Google Scholar]
- 14. Targeting the EWS-FLI1 transcription factor in Ewing sarcomaCancer Chemother Pharmacol75131713202015[PubMed][Google Scholar]
- 15. Ciclopirox nail lacquer topical solution 8% in the treatment of toenail onychomycosisJ Am Acad Dermatol434 SupplS70S802000[PubMed][Google Scholar]
- 16. Dermatopharmacology of ciclopirox nail lacquer topical solution 8% in the treatment of onychomycosisJ Am Acad Dermatol434 SupplS57S692000[PubMed][Google Scholar]
- 17. The antifungal drug ciclopirox inhibits deoxyhypusine and proline hydroxylation, endothelial cell growth and angiogenesis in vitroInt J Cancer1004914982002[PubMed][Google Scholar]
- 18. The antitumor activity of the fungicide ciclopiroxInt J Cancer127246724772010[PubMed][Google Scholar]
- 19. Novel mTOR inhibitory activity of ciclopirox enhances parthenolide antileukemia activityExp Hematol41799807.e42013[PubMed][Google Scholar]
- 20. Oncogenic ETS fusions deregulate E2F3 target genes in Ewing sarcoma and prostate cancerGenome Res23179718092013[PubMed][Google Scholar]
- 21. Gene expression signature based screening identifies ribonucleotide reductase as a candidate therapeutic target in Ewing sarcomaOncotarget763003630192016[PubMed][Google Scholar]
- 22. Normalization for cDNA microarray data: A robust composite method addressing single and multiple slide systematic variationNucleic Acids Res30e152002[PubMed][Google Scholar]
- 23. WebGestalt 2017: A more comprehensive, powerful, flexible and interactive gene set enrichment analysis toolkitNucleic Acids Res
5 3 2017[PubMed][Google Scholar] - 24. REVIGO summarizes and visualizes long lists of gene ontology termsPLoS One6e218002011[PubMed][Google Scholar]
- 25. Fast gapped-read alignment with Bowtie 2Nat Methods93573592012[PubMed][Google Scholar]
- 26. Identifying ChIP-seq enrichment using MACSNat Protoc7172817402012[PubMed][Google Scholar]
- 27. ChIPseeker: An R/Bioconductor package for ChIP peak annotation, comparison and visualizationBioinformatics31238223832015[PubMed][Google Scholar]
- 28. Integrative genomics viewer (IGV): High-performance genomics data visualization and explorationBrief Bioinform141781922013[PubMed][Google Scholar]
- 29. The STRING database in 2017: Quality-controlled protein-protein association networks, made broadly accessibleNucleic Acids Res45D362D3682017[PubMed][Google Scholar]
- 30. Cytoscape: A software environment for integrated models of biomolecular interaction networksGenome Res13249825042003[PubMed][Google Scholar]
- 31. HOE 296, a new antimycotic compound with a broad antimicrobial spectrum. Laboratory resultsArzneimittelforschung236706741973[PubMed][Google Scholar]
- 32. Ciclopirox nail lacquer 8%: In vivo penetration into and through nails and in vitro effect on pig skinSkin Pharmacol489941991[PubMed][Google Scholar]
- 33. Ciclopirox: An overviewInt J Dermatol403053102001[PubMed][Google Scholar]
- 34. Oral ciclopirox olamine displays biological activity in a phase I study in patients with advanced hematologic malignanciesAm J Hematol893633682014[PubMed][Google Scholar]
- 35. The antitumor activity of the fungicide ciclopiroxInt J Cancer10246724772010[PubMed][Google Scholar]
- 36. Chelation of intracellular iron with the antifungal agent ciclopirox olamine induces cell death in leukemia and myeloma cellsBlood114306430732009[PubMed][Google Scholar]
- 37. Dual targeting of EWS-FLI1 activity and the associated DNA damage response with trabectedin and SN38 synergistically inhibits Ewing sarcoma cell growthClin Cancer Res20119012032014[PubMed][Google Scholar]
- 38. Functional genomic screening reveals splicing of the EWS-FLI1 fusion transcript as a vulnerability in Ewing SarcomaCell Rep145986102016[PubMed][Google Scholar]
- 39. Combined experience of six independent laboratories attempting to create an Ewing sarcoma mouse modelOncotarget834141341632016[Google Scholar]
- 40. Transcription factor Fli1 regulates collagen fibrillogenesis in mouse skinMol Cell Biol294254342009[PubMed][Google Scholar]
- 41. Hypoxia-regulated target genes implicated in tumor metastasisJ Biomed Sci191022012[PubMed][Google Scholar]
- 42. Proteomic analysis of the EWS-Fli-1 interactome reveals the role of the Lysosome in EWS-Fli-1 turnoverJ Proteome Res13378337912014[PubMed][Google Scholar]
- 43. Therapies for bleomycin induced lung fibrosis through regulation of TGF-beta1 induced collagen gene expressionJ Cell Physiol2115855892007[PubMed][Google Scholar]
- 44. Stimulation of type I collagen transcription in human skin fibroblasts by TGF-beta: Involvement of Smad 3J Invest Dermatol11249571999[PubMed][Google Scholar]
- 45. Identification of novel TGF-β/Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approachJ Biol Chem27617058170622001[PubMed][Google Scholar]
- 46. Repression of the gene encoding the TGF-beta type II receptor is a major target of the EWS-FLI1 oncoproteinNat Genet234811999[PubMed][Google Scholar]
- 47. Hypoxia-inducible factor 1: Oxygen homeostasis and disease pathophysiologyTrends Mol Med73453502001[PubMed][Google Scholar]
- 48. Reactive oxygen species regulate epidermal growth factor-induced vascular endothelial growth factor and hypoxia-inducible factor-1 alpha expression through activation of AKT and P70S6K1 in human ovarian cancer cellsFree Radic Biol Med41152115332006[PubMed][Google Scholar]
- 49. The role of oxygen availability in embryonic development and stem cell functionNat Rev Mol Cell Biol92852962008[PubMed][Google Scholar]
- 50. Twist1 induces CCL2 and recruits macrophages to promote angiogenesisCancer Res736626712013[PubMed][Google Scholar]
- 51. The TWIST1 oncogene is a direct target of hypoxia-inducible factor-2alphaOncogene27150115102008[PubMed][Google Scholar]
- 52. Sensors, transmitters, and targets in mitochondrial oxygen shortage-a hypoxia-inducible factor relay storyAntioxidants Redox Signal203393522014[PubMed][Google Scholar]
- 53. Differential contribution of key metabolic substrates and cellular oxygen in HIF signallingExp Cell Res33013282015[PubMed][Google Scholar]
- 54. Expression of MYCN in pediatric synovial sarcomaModern Pathol207347412007[PubMed][Google Scholar]