OR Oncology Reports 1021-335X 1791-2431 D.A. Spandidos 10.3892/or.2019.7022 or-41-04-2194 Articles In silico analysis identified miRNA-based therapeutic agents against glioblastoma multiforme XiongDan-Dan 1 XuWen-Qing 1 HeRong-Quan 2 DangYi-Wu 1 ChenGang 1 LuoDian-Zhong 1 Department of Pathology, First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi Zhuang Autonomous Region 530021, P.R. China Department of Medical Oncology, First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi Zhuang Autonomous Region 530021, P.R. China Correspondence to: Professor Dian-Zhong Luo, Department of Pathology, First Affiliated Hospital of Guangxi Medical University, 6 Shuangyong Road, Nanning, Guangxi Zhuang Autonomous Region 530021, P.R. China, E-mail: 13878802796@163.com 042019 19022019 41 4 2194 2208 22062018 12022019 Copyright: © Xiong et al. 2019 This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

MicroRNAs (miRNAs or miRs) contribute to the development of various malignant neoplasms, including glioblastoma multiforme (GBM). The present study aimed to explore the pathogenesis of GBM and to identify latent therapeutic agents for patients with GBM, based on an in silico analysis. Gene chips that provide miRNA expression profiling in GBM were obtained from the Gene Expression Omnibus (GEO) database. Differentially expressed miRNAs (DEMs) were also determined via the RobustRankAggreg algorithm. The target genes of DEMs were predicted and then intersected with GBM-associated genes that were collected from the Gene Expression Profiling Interactive Analysis. Gene Oncology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of the overlapping genes were then performed. Simultaneously, a connectivity map (CMap) analysis was performed to screen for potential therapeutic agents for GBM. A total of 10 DEMs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b, hsa-miR-542-3p, hsa-miR-129-3p, hsa-miR-1224-5p, hsa-miR-876-3p and hsa-miR-770-5p) were obtained from three GEO gene chips (GSE25631, GSE42657 and GSE61710). Then, 1,720 target genes of the 10 miRNAs and 4,185 differently expressed genes in GBM were collected. By intersecting the aforementioned gene clusters, the present study identified 390 overlapping genes. GO and KEGG analyses of the 390 genes demonstrated that these genes were involved in certain cancer-associated biological functions and pathways. Eight genes [(GTPase NRas (NRAS), calcium/calmodulin-dependent protein kinase type II subunit Gamma (CAMK2G), platelet-derived growth factor receptor alpha (PDGFRA), calmodulin 3 (CALM3), cyclin-dependent kinase 6 (CDK6), calcium/calmodulin-dependent protein kinase type II subunit beta (CAMK2B), retinoblastoma-associated protein (RB1) and protein kinase C beta type (PRKCB)] that were centralized in the glioma pathway were selected for CMap analysis. Three chemicals (W-13, gefitinib and exemestane) were identified as putative therapeutic agents for GBM. In summary, the present study identified three miRNA-based chemicals for use as a therapy for GBM. However, more experimental data are needed to verify the therapeutic properties of these latent drugs in GBM.

 glioblastoma multiforme miRNA pathway connectivity map therapeutic agent Introduction

Glioblastoma multiforme (GBM), which is classified by the World Health Organization as a grade IV glioma, exhibits a high morbidity and mortality, comprising 47.1% of all malignant tumors of the central nervous system (1,2). In total, ~13,000 people in America are diagnosed with GBM each year (3). The main treatment of GBM is surgical resection in combination with radiotherapy or chemotherapy. However, the majority of patients relapse within the 7 months following their original diagnoses (4). Furthermore, a resistance to current chemotherapy leads to a heavy tumor burden for patients with GBM. Although novel treatments, including immunotherapy and molecular targeted therapy, have been in development for several years (5,6), the 5-year survival rate is relatively low, with a median survival time of 15 months (4), indicating the urgency of determining novel therapies.

MicroRNAs (miRNAs) are a class of small non-coding RNAs comprising ~19–23 nucleotides (7). By binding to target mRNAs, miRNAs regulate gene expression at the transcriptional or posttranscriptional level, which enables them to serve pivotal roles in various biological processes, including cell growth, apoptosis, invasion and metastasis (8,9). In terms of GBM, it is widely reported that miRNAs participate in various molecular pathways associated with cancer development (10–12). miRNA-targeted therapy has been utilized in cancer treatment by developing miRNA mimetic and anti-miRNA agents (13). The present study aimed to determine latent miRNA-based therapeutic agents for GBM by employing a Connectivity Map (CMap) method. Using >7,000 expression profiles (representing 1,309 compounds), the CMap reveals connections among genes, chemicals and diseases (14). It is extensively used in the exploitation of novel drugs or existing drug applications (15,16). Drugs that are available in CMap are all licensed for human use by the Food and Drug Administration (17). It is an ideal database for probing chemicals that may be applied in the therapy of GBM.

The experimental processes of the present study were as follows (Fig. 1): i) A statistical analysis of differentially expressed miRNAs (DEMs) in GBM was performed using the expression data of gene chips collected from the Gene Expression Omnibus (GEO); ii) the target genes of DEMs were predicted and intersected with the GBM-associated genes and the resulting genes were defined as miRNA-associated differentially expressed genes (DEGs); iii) Gene Ontology (GO) functional annotations and a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed, and the genes participating in the ‘glioma pathway’ were selected for further analysis; iv) a CMap analysis was applied to probe for potential therapeutic chemicals for GBM; and v) a molecular docking approach was adopted to assess the affinity of the selected chemicals to their target genes.

 Materials and methods <sec> <title>Identification of DEMs in GBM

A search of miRNA-associated microarray datasets in GBM was performed using the GEO database (18) with the following screening criteria: i) the organism must be restricted to ‘homo sapiens’; ii) a complete expression profile determination of the miRNA had to have been performed; and iii) cancerous and noncancerous samples needed to contain at least five samples. A Bioconductor package ‘Limma’ (19) was applied to screen for DEMs in the available gene chip. The DEMs in the individual gene chips were then integrated and ranked with RobustRankAggreg (20), which is an R package for the comprehensive integration of the gene list.

 Collection of miRNA-associated DEGs

With filter conditions of [log2 (fold change)]≥1.5 and an adjusted P<0.05, DEGs in GBM were collected from Gene Expression Profiling Interactive Analysis (GEPIA) (21), an online interactive server for gene sequence analysis that provides a gene expression profile comparison between 163 GBM specimens and 207 normal brain samples from the Cancer Genome Atlas and Genotype-Tissue Expression projects.

The target genes of DEMs were predicted using miRWalk (22), an integrated resource that provides predictions of miRNA- target interactions. Only the target genes that were predicted by at least eight target-predicting algorithms were considered to intersect with the DEGs gathered from GEPIA. Overlapping genes were regarded as miRNA-associated DEGs in GBM.

 GO and KEGG enrichment analyses and protein-protein interaction (PPI) network construction

GO and KEGG enrichment analyses were performed using the ClusterProfiler (23), an R package that is used for the systematic analysis of gene clusters. A PPI network was constructed using STRING (24), a database that provides functional interactions among proteins.

 Immunohistochemistry

The Human Protein Atlas (https://www.proteinatlas.org/), an interactive web tool that assesses proteins in all major tissues and organs in the human body (25) was used to determine the expression level of proteins. The main clinical characteristic of the patients are presented in Table SII.

 Drug discovery in CMap

Query genes were uploaded to the CMap web tool (26), comparing >7,000 gene expression profiles following treatment with 1,309 active chemicals in human cell lines. The link between the query genes and the 1,309 chemicals was measured via a connectivity score provided by the CMap tool, which ranged valued from −1 to 1. A positive score implied a stimulative effect, while a negative score indicated a suppressive effect of the chemical on the given signatures.

 Molecular docking analysis

Molecular docking between the proteins encoded by miRNA-associated DEGs and filtered chemicals was performed using Sybyl-X (27). Protein crystal structures were downloaded from the Research Collaboratory for Structural Bioinformatics Protein Data Bank (PDB) (28) and chemical structures were obtained from PubChem (29). First, the protein crystal structure was imported into the Sybyl-X 2.1.1 software on the Surflex-Dock interface. Following the removal of irrelevant water molecules and ions, the repair of side chains, the charging of terminal groups and the addition of polar hydrogen atoms, the proteins were prepared for docking. Protomols, which are active pockets that ligands are aligned to, were generated with the use of an automatic mode. Compounds in the mol2 format were then imported into the software on the Docking interface and protein-ligand docking was run under the surflex-dock geom mode, after which a total score was exported, with these scores being directly proportional to the binding affinity. Other parameters were set by default. The binding mode was visualized by the use of PyMOL (30).

 Results <sec> <title>Screening of 10 DEMs in GBM by the RobustRankAggreg method

Three gene chips [GSE25631 (31), GSE42657 (32) and GSE61710 (33)], were collected from GEO. The basic information of the three microarray datasets are presented in Table I and the main clinicopathological characteristics of the samples in the three gene chips are presented in supplementary Table SI with the screening criteria of [log2 (fold change)] >1 and P<0.05, 51 upregulated miRNAs and 56 downregulated miRNAs were obtained in gene chip GSE25631 (Fig. 2A); 47 upregulated miRNAs and 61 downregulated miRNAs were obtained in gene chip GSE42657 (Fig. 2B); and eight upregulated miRNAs and one downregulated miRNA were identified in gene chip GSE61710 (Fig. 2C). Following robust rank aggregation, an ordered list of all of the DEMs in the three datasets was acquired, among which the top five upregulated DEMs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b and hsa-miR-542-3p) and the top five downregulated DEMs (hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p) were determined as being miRNAs associated with GBM (Fig. 2D). The expression of the 10 DEMs in each dataset is presented in Fig. 3. Published studies focusing on the associations between the 10 DEMs and GBM are included in Table II.

 Collection of 390 miRNA-associated DEGs in GBM

A total of 1,720 genes were identified as targets of the aforementioned 10 DEMs. Additionally, 4,185 DEGs containing 2,719 upregulated genes and 1,466 downregulated genes in GBM were collected from GEPIA (Fig. 4A). An intersection between the 1,720 target genes and the 4,185 DEGs was performed, revealing 390 overlapping miRNA-associated DEGs (Fig. 4B).

 GO and KEGG enrichment analyses of 390 miRNA-associated DEGs, and PPI analysis of eight GBM-associated genes

GO functional annotations were performed to determine the potential molecular mechanisms employed by the 390 miRNA-associated DEGs. The top 10 biological processes (BP), cellular components (CC) and molecular functions (MF) are listed in Fig. 5A-C. The highly enriched BP terms were ‘cell growth’, ‘signal release’ and ‘regulation of cell growth’. The markedly enriched CC terms were ‘presynapse’, ‘axon’ and ‘synaptic membrane’. The predominantly enriched MF terms were ‘transcription factor activity, RNA polymerase II core promoter proximal region sequence-specific binding’, ‘transcriptional activator activity, RNA polymerase II transcription regulatory region sequence-specific binding’ and ‘transcriptional activator activity, RNA polymerase II core promoter proximal region sequence-specific binding’.

The KEGG pathway was utilized to further probe the underlying pathological pathways that the 390 genes are involved in during the inception and progression of GBM. With an adjusted value of P<0.05, 32 pathways were enriched by the 390 genes (Fig. 5D), many of which were tumor-associated pathways, including the ‘mitogen activated protein kinase (MAPK) signaling pathway’, the ‘gonadotropin-releasing hormone (GnRH) signaling pathway’ and the ‘oxytocin signaling pathway’. In addition, eight genes [cyclin-dependent kinase-6 (CDK6), retinoblastoma-associated protein (RB1), calcium/calmodulin-dependent protein kinase type II subunit gamma (CAMK2G), calcium/calmodulin-dependent protein kinase type II subunit beta (CAMK2B), GTPase NRas (NRAS), protein kinase C beta type (PRKCB), platelet derived growth factor receptor alpha (PDGFRA) and calmodulin 3 (CALM3)] were determined to be centralized in the glioma pathway (Fig. 6). In this pathway, the eight genes were revealed to participate in cell growth and proliferation, G1/S progression, cell migration and mitosis by interacting with upstream or downstream genes. The expression of the eight genes in GBM and normal controls were compared based on the data from GEPIA using a Student's t-test presented as scatter-box plots (Fig. 7). Among them, CDK6, RB1, NRAS, and PDGFRA were upregulated in GBM, and CAMK2G, CAMK2B, PRKCB, and CALM3 were downregulated in GBM tissues. Additionally, the expression patterns of their encoded proteins in GBM and normal brain tissues were validated using The Human Protein Atlas (https://www.proteinatlas.org/) as presented in Fig. 8. The PPI network indicated close associations among the eight genes (Fig. 9).

 Identification of three potential chemicals for GBM treatment using CMap analysis

The eight genes, including four that were upregulated (CDK6, RB1, NRAS and PDGFRA) and four that were downregulated (CAMK2G, CAMK2B, PRKCB and CALM3) were submitted to the CMap web tool as up and down tags to acquire latent drugs in the therapy for GBM. By ranking the connectivity score in descending order, the top three chemicals (W-13, gefitinib and exemestane) were identified as being potential treatment options for GBM (Table III). The chemical structures of the three chemicals are presented in Fig. 10.

 Exploration of the interactions between the three chemicals and the eight genes

To ascertain whether the three chemicals directly bind to the proteins encoded by the eight genes, a protein-ligand docking analysis was performed. As the crystal structure of protein CALM3 was not identified in the PDB database, it was removed from the molecular docking analysis. The docking results are presented in Table IV and Figs. 11–13. The total docking score ranged from 3.5943 to 6.9781, indicating interactions between the three chemicals and the seven proteins. However, further experiments are required to verify these associations.

 Discussion

Since the identification of the first miRNA, an increasing number of studies have focused on the action and clinical application of miRNAs, particularly in terms of neoplasm treatment (34). It has been demonstrated that miRNAs participate in multiple biological processes involved in tumorigenesis (35). With the advantage of targeting genes that are involved in multiple pathological pathways, drug developments based on miRNAs have received increasing attention (36).

The present study obtained 10 DEMs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b, hsa-miR-542-3p, hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p) using a robust rank aggregation, which is a recognized method for integrating genes from diverse resources that are free of outliers, noise and errors (20). Among these literatures, four studies corroborated overexpressed hsa-miR-196a in GBM and its contribution to the development of GBM (37–40). Additionally, four studies indicated that a high expression of hsa-miR-10b promoted the progression of GBM (41–44). Five studies focused on hsa-miR-196b and demonstrated its high expression in GBM (45–49) and one study, performed by Cai et al (50), focused on hsa-miR-542-3p. The authors of these studies determined that hsa-miR-542-3p was downregulated in glioblastoma cell lines, which was not consistent with the gene chip results from the present study. Given the differences in sample sources, RNA extraction and detection, more studies are required to further assess the role of hsa-miR-542-3p in GBM. Additionally, the decreased expression of hsa-miR-1224-5p (51) and hsa-miR-129-3p (52,53), as well as their inhibitory effects on GBM, have been verified in previous studies. Furthermore, the association of hsa-miR-18b, hsa-miR-219-2-3p, hsa-miR-876-3p and hsa-miR-770-5p with GBM have not yet been reported. Since miRNAs control tumor development by regulating their downstream target genes (54), the present study collected the target genes of the aforementioned 10 DEMs to elucidate how these DEMs mediate the pathophysiological processes of GBM. The DEGs in GBM were also obtained and an intersection between the target genes and the DEGs was performed to determine miRNA-associated DEGs. Gene functional and pathway enrichment analyses of the miRNA-associated DEGs revealed that these genes were involved with multiple tumor-associated biological processes and signaling pathways, including the ‘cell growth’, the ‘MAPK signaling pathway’, the ‘GnRH signaling pathway’ and the ‘oxytocin signaling pathway’. Furthermore, eight genes (CDK6, RB1, CAMK2G, CAMK2B, NRAS, PRKCB, PDGFRA and CALM3) of the miRNA-associated DEGs were enriched in the glioma pathway, indicating their important roles in GBM. In the glioma pathway, the eight genes primarily participated in cell growth and proliferation, G1/S progression, and cell migration and mitosis, indicating that these genes participate in the development of GBM by mediating cell proliferation and metastasis. Each of the eight genes was targeted by more than one miRNA and one miRNA targeted more than one gene. For example, the PDGFRA gene was targeted by hsa-miR-770-5p and hsa-miR-196a. hsa-miR-196a targeted NRAS, PDGFRA and CALM3. miRNAs primarily exert effects via destabilization or translational repression by targeting the 3′ untranslated region of mRNA transcripts in the cytoplasm (7). However, an increasing number of studies have indicated that miRNAs positively regulate gene transcription by targeting promoter elements (55–57). The present study revealed that certain miRNAs were negatively associated with their target genes (hsa-miR-1224-5p and CDK6; hsa-miR-196a and CALM3), while other miRNAs were positively associated with their target genes (hsa-miR-196a and NRAS; hsa-miR-1224-5p and PRKCB). However, further studies are required to assess the regulatory mechanisms of the 10 miRNAs and their target genes.

CMap is a practical tool for the exploration of novel drugs and for the repurposing of existing drugs, and its efficiency has been supported by numerous studies (58,59). Aramadhaka et al (58) identified Gila monster venom and Byetta® as being therapeutic drugs for the treatment of type-2 diabetes using CMap analysis. Wang et al (59) demonstrated that via cell apoptosis, prenylamine could be a candidate agent for the treatment of hepatocellular carcinoma. The present study selected the aforementioned eight genes for CMap analysis. Following this analysis, three chemicals (W-13, gefitinib and exemestane) were determined as latent therapeutic agents for GBM. As a calmodulin antagonist, W-13 has been demonstrated to inhibit cell growth (60) and to induce cell apoptosis (61). However, few studies on W-13 have assessed its anti-GBM effects. The results from the molecular docking analysis performed in the present study revealed that W-13 could bind to proteins CDK6, RB1, NRAS, PDGFRA, CAMK2G, CAMK2B and PRKCB, exhibiting high binding scores and indicating that W-13 could exert its anti-GBM effects by acting on these GBM-associated genes. The present study provides a theoretical basis for the application of W-13 in patients with GBM, but further studies are required to corroborate this conclusion.

The inhibitory effect of gefitinib on GBM, which is an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, has been demonstrated in previous studies (62,63). However, its clinical application is limited due to gefitinib resistance (62). A study by Aljohani et al (64) revealed that PDGFRA was significantly upregulated in gefitinib-resistant GBM cells and that overexpressed PDGFRA regulated gefitinib resistance. The present study identified that PDGFRA was a target gene of the 10 DEMs, suggesting the potential of the 10 DEMs in enhancing gefitinib sensitivity in patients with GBM.

Exemestane is a widely used drug in the prevention and treatment of breast cancer due to its aromatase inhibitory role in the production of oestrogen (65–67). However, to the best of our knowledge, its antitumor effect on GBM has not yet been elucidated. A study by Kritikou et al (68) revealed that the combination of exemestane and erlotinib significantly inhibited EGFR mitochondrial translocation. The EGFR mitochondrial translocation event serves important roles in tumor progression (69) and contributes to drug resistance (70). A study by Dasari et al (71) indicated that the inhibition of the EGFR mitochondrial translocation event in GBM may be a therapeutic strategy. Additionally, the translocation of EGFR into mitochondria contributes to EGFR inhibitor drug resistance (70). The combined use of exemestane and the EGFR inhibitor, gefitinib, in patients with GBM may therefore increase gefitinib sensitivity by inhibiting the translocation of EGFR into the mitochondria. However, further in-depth in vitro and in vivo experiments are essential to verify the anti-GBM effects and the synergistic antitumor effects of these compounds.

In the present study, the identification of W-13, gefitinib and exemestane were made on the basis of the eight GBM-associated genes (CDK6, RB1, CAMK2G, CAMK2B, NRAS, PRKCB, PDGFRA and CALM3), which were the target genes of the 10 DEMs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b, hsa-miR-542-3p, hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p). Thus, the present study hypothesizes that the 10 DEMs may produce synergistic or antagonistic effects on the three chemicals by targeting these genes. However, more experiments are necessary to validate this conjecture.

In conclusion, by employing an integrated strategy of data mining and computational biology, the present study obtained 10 DEMs that may participate in the development of GBM. Furthermore, three candidate agents (gefitinib, W-13 and exemestane) in the treatment of GBM were identified following CMap analysis. Since the results are based on In silico analysis, further in-depth studies are necessary to add to the validity of these results.

 Supplementary Material Supporting Data Acknowledgements

Not applicable.

 Funding

No funding was received.

 Availability of data and materials

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

 Authors' contributions

DDX designed the present study, collected miRNA-associated datasets, screened differently expressed miRNAs and wrote the manuscript. WQX collected miRNA-associated datasets and wrote the manuscript. RQH and YWD performed GO and KEGG analyses and construed the PPI network. GC performed CMap analysis and checked all data. DZL designed the experiments and wrote manuscript. All authors read and approved the final manuscript.

 Ethics approval and consent to participate

Not applicable.

 Patient consent for publication

Not applicable.

 Competing interests

The authors declare that they have no competing interests.

 References LouisDNPerryAReifenbergerGvon DeimlingAFigarella- BrangerDCaveneeWKOhgakiHWiestlerODKleihuesPEllisonDWThe 2016 World health organization classification of tumors of the central nervous system: A summaryActa Neuropathol131803820201610.1007/s00401-016-1545-127157931 OstromQTGittlemanHLiaoPVecchione-KovalTWolinskyYKruchkoCBarnholtz-SloanJSCBTRUS statistical report: Primary brain and other central nervous system tumors diagnosed in the United States in 2010–2014Neuro Oncol19(Suppl_5)v1v88201710.1093/neuonc/nox158291172895693142 OmuroADeAngelisLMGlioblastoma and other malignant gliomas: A clinical reviewJAMA31018421850201310.1001/jama.2013.28031924193082 AlifierisCTrafalisDTGlioblastoma multiforme: Pathogenesis and treatmentPharmacol Ther1526382201510.1016/j.pharmthera.2015.05.00525944528 CloughesyTFCaveneeWKMischelPSGlioblastoma: From molecular pathology to targeted treatmentAnnu Rev Pathol9125201410.1146/annurev-pathol-011110-13032423937436 LimMXiaYBettegowdaCWellerMCurrent state of immunotherapy for glioblastomaNat Rev Clin Oncol15422442201810.1038/s41571-018-0003-529643471 BartelBPMicroRNAs: Genomics, biogenesis, mechanism, and functionCell116281297200410.1016/S0092-8674(04)00045-514744438 DongHLeiJDingLWenYJuHZhangXMicroRNA: Function, detection, and bioanalysisChem Rev11362076233201310.1021/cr300362f23697835 BrackenCPScottHSGoodallGJA network-biology perspective of microRNA function and dysfunction in cancerNat Rev Genet17719732201610.1038/nrg.2016.13427795564 SumazinPYangXChiuHSChungWJIyerALlobet-NavasDRajbhandariPBansalMGuarnieriPSilvaJAn extensive microRNA-mediated network of RNA-RNA interactions regulates established oncogenic pathways in glioblastomaCell147370381201110.1016/j.cell.2011.09.041220000153214599 GodlewskiJNowickiMOBroniszANuovoGPalatiniJDe LayMVan BrocklynJOstrowskiMCChioccaEALawlerSEMicroRNA-451 regulates LKB1/AMPK signaling and allows adaptation to metabolic stress in glioma cellsMol Cell37620632201010.1016/j.molcel.2010.02.018202273673125113 HuJSunTWangHChenZWangSYuanLLiuTLiHRWangPFengYMiR-215 is induced post-transcriptionally via HIF-Drosha complex and mediates glioma-initiating cell adaptation to hypoxia by targeting KDM1BCancer Cell294960201610.1016/j.ccell.2015.12.005267665904871949 Berindan-NeagoeIMonroig PdelCPasculliBCalinGAMicroRNAome genome: A treasure for cancer diagnosis and therapyCA Cancer J Clin64311336201410.3322/caac.21244251045024461198 SubramanianANarayanRCorselloSMPeckDDNatoliTELuXGouldJDavisJFTubelliAAAsieduJKA next generation connectivity map: L1000 platform and the first 1,000,000 profilesCell17114371452.e17201710.1016/j.cell.2017.10.049291950785990023 QuXARajpalDKApplications of connectivity map in drug discovery and developmentDrug Discov Today1712891298201210.1016/j.drudis.2012.07.01722889966 ChienWSunQYLeeKLDingLWWuenschePTorres-FernandezLATanSZTokatlyIZaidenNPoellingerLActivation of protein phosphatase 2A tumor suppressor as potential treatment of pancreatic cancerMol Oncol9889905201510.1016/j.molonc.2015.01.002256372834387089 LambJThe connectivity map: A new tool for biomedical researchNat Rev Cancer75460200710.1038/nrc204417186018 CloughEBarrettTThe gene expression omnibus databaseMethods Mol Biol141893110201610.1007/978-1-4939-3578-9_5270080114944384 RitchieMEPhipsonBWuDHuYLawCWShiWSmythGKlimma powers differential expression analyses for RNA-sequencing and microarray studiesNucleic Acids Res43e47201510.1093/nar/gkv007256057924402510 KoldeRLaurSAdlerPViloJRobust rank aggregation for gene list integration and meta-analysisBioinformatics28573580201210.1093/bioinformatics/btr709222472793278763 TangZLiCKangBGaoGLiCZhangZGEPIA: A web server for cancer and normal gene expression profiling and interactive analysesNucleic Acids Res45W98W102201710.1093/nar/gkx247284071455570223 DweepHGretzNStichtCmiRWalk database for miRNA- target interactionsMethods Mol Biol1182289305201410.1007/978-1-4939-1062-5_2525055920 YuGWangLGHanYHeQYclusterProfiler: An R package for comparing biological themes among gene clustersOMICS16284287201210.1089/omi.2011.0118224554633339379 SzklarczykDMorrisJHCookHKuhnMWyderSSimonovicMSantosADonchevaNTRothABorkPThe STRING database in 2017: Quality-controlled protein-protein association networks, made broadly accessibleNucleic Acids Res45D362D368201710.1093/nar/gkw93727924014 UhlénMFagerbergLHallströmBMLindskogCOksvoldPMardinogluASivertssonÅKampfCSjöstedtEAsplundAProteomics. Tissue-based map of the human proteomeScience3471260419201510.1126/science.126041925613900 MusaAGhoraieLSZhangSDGlazkoGYli-HarjaODehmerMHaibe-KainsBEmmert-StreibFA review of connectivity map and computational approaches in pharmacogenomicsBrief Bioinform18903201710.1093/bib/bbx023283341736113891 GuedesIAde MagalhaesCSDardenneLEReceptor-ligand molecular dockingBiophys Rev67587201410.1007/s12551-013-0130-228509958 RosePWPrlicABiCBluhmWFChristieCHDuttaSGreenRKGoodsellDSWestbrookJDWooJThe RCSB protein data bank: Views of structural biology for basic and applied research and educationNucleic Acids Res43D345D356201510.1093/nar/gku121425428375 KimSThiessenPABoltonEEChenJFuGGindulyteAHanLHeJHeSShoemakerBAPubChem substance and compound databasesNucleic Acids Res44D1202D1213201610.1093/nar/gkv95126400175 AlexanderNWoetzelNMeilerJbcl::Cluster: A method for clustering biological molecules coupled with visualization in the Pymol Molecular Graphics SystemIEEE Int Conf Comput Adv Bio Med Sci201113182011278188475091839 ZhangWZhangJHoadleyKKushwahaDRamakrishnanVLiSKangCYouYJiangCSongSWmiR-181d: A predictive glioblastoma biomarker that downregulates MGMT expressionNeuro Oncol14712719201210.1093/neuonc/nos089225704263367855 JonesTAJeyapalanJNForshewTTatevossianRGLawsonARPatelSNDoctorGTMuminMAPickerSRPhippsKPMolecular analysis of pediatric brain tumors identifies microRNAs in pilocytic astrocytomas that target the MAPK and NF-kB pathwaysActa Neuropathol Commun386201510.1186/s40478-015-0266-3266829104683939 PiweckaMRolleKBelterABarciszewskaAMŻywickiMMichalakMNowakSNaskret-BarciszewskaMZBarciszewskiJComprehensive analysis of microRNA expression profile in malignant glioma tissuesMol Oncol913241340201510.1016/j.molonc.2015.03.007258640395528820 KongYWFerland-McColloughDJacksonTJBushellMmicroRNAs in cancer managementLancet Onco13e249e258201210.1016/S1470-2045(12)70073-6 LingHFabbriMCalinGAMicroRNAs and other non-coding RNAs as targets for anticancer drug developmentNat Rev Drug Discov12847865201310.1038/nrd4140241723334548803 GarzonRMarcucciGCroceCMTargeting microRNAs in cancer: Rationale, strategies and challengesNat Rev Drug Discov9775789201010.1038/nrd3179208854093904431 DouTWuQChenXRibasJNiXTangCHuangFZhouLLuDA polymorphism of microRNA196a genome region was associated with decreased risk of glioma in Chinese populationJ Cancer Res Clin Oncol13618531859201010.1007/s00432-010-0844-520229273 YangGHanDChenXZhangDWangLShiCZhangWLiCChenXLiuHMiR-196a exerts its oncogenic effect in glioblastoma multiforme by inhibition of IκBα both in vitro and in vivoNeuro Oncol16652661201410.1093/neuonc/not307244633573984554 GuanYChenLBaoYQiuBPangCCuiRWangYHigh miR-196a and low miR-367 cooperatively correlate with unfavorable prognosis of high-grade gliomaInt J Clin Exp Pathol8657665882015262615394525873 YangJPYangJKLiCCuiZQLiuHJSunXFGengSMLuSKSongJGuoCYJiaoBHDownregulation of ZMYND11 induced by miR-196a-5p promotes the progression and growth of GBMBiochem Biophys Res Commun494674680201710.1016/j.bbrc.2017.10.09829066350 SasayamaTNishiharaMKondohTHosodaKKohmuraEMicroRNA-10b is overexpressed in malignant glioma and associated with tumor invasive factors, uPAR and RhoCInt J Cancer12514071413200910.1002/ijc.2452219536818 GuessousFAlvarado-VelezMMarcinkiewiczLZhangYKimJHeisterSKefasBGodlewskiJSchiffDPurowBAbounaderROncogenic effects of miR-10b in glioblastoma stem cellsJ Neurooncol112153163201310.1007/s11060-013-1047-0233073283609924 GabrielyGTeplyukNMKrichevskyAMContext effect: microRNA-10b in cancer cell proliferation, spread and deathAutophagy713841386201110.4161/auto.7.11.1737121795860 JiYWeiYWangJGongKZhangYZuoHCorrelation of microRNA-10b upregulation and poor prognosis in human gliomasTumour Biol3662496254201510.1007/s13277-015-3310-925773393 GuanYMizoguchiMYoshimotoKHataNShonoTSuzukiSOArakiYKugaDNakamizoAAmanoTMiRNA-196 is upregulated in glioblastoma but not in anaplastic astrocytoma and has prognostic significanceClin Cancer Res1642894297201010.1158/1078-0432.CCR-10-020720601442 LakomyRSanaJHankeovaSFadrusPKrenLLzicarovaESvobodaMDolezelovaHSmrckaMVyzulaRMiR-195, miR-196b, miR-181c, miR-21 expression levels and O−6-methylguanine-DNA methyltransferase methylation status are associated with clinical outcome in glioblastoma patientsCancer Sci10221862190201110.1111/j.1349-7006.2011.02092.x21895872 MaRYanWZhangGLvHLiuZFangFZhangWZhangJTaoTYouYUpregulation of miR-196b confers a poor prognosis in glioblastoma patients via inducing a proliferative phenotypePLoS One7e38096201210.1371/journal.pone.0038096227238493378534 YouGYanWZhangWWangYBaoZLiSLiSLiGSongYKangCSignificance of miR-196b in tumor-related epilepsy of patients with gliomasPLoS One7e46218201210.1371/journal.pone.0046218230499823457999 KarsyMArslanEMoyFCurrent progress on understanding MicroRNAs in glioblastoma multiformeGenes Cancer3315201210.1177/1947601912448068228937863415667 CaiJZhaoJZhangNXuXLiRYiYFangLZhangLLiMWuJMicroRNA-542-3p suppresses tumor cell invasion via targeting AKT pathway in human astrocytomaJ Biol Chem2902467824688201510.1074/jbc.M115.649004262867474598981 QianJLiRWangYYShiYLuanWKTaoTZhangJXXuYCYouYPMiR-1224-5p acts as a tumor suppressor by targeting CREB1 in malignant gliomasMol Cell Biochem4033341201510.1007/s11010-015-2334-125648114 OuyangQChenGZhouJLiLDongZYangRXuLCuiHXuMYiLNeurotensin signaling stimulates glioblastoma cell proliferation by upregulating c-Myc and inhibiting miR-29b-1 and miR-129-3pNeuro Oncol18216226201610.1093/neuonc/nov11426180082 FangDZWangYPLiuJHuiXBWangXDChenXLiuDMicroRNA-129-3p suppresses tumor growth by targeting E2F5 in glioblastomaEur Rev Med Pharmacol Sci2210441050201829509253 TutarYmiRNA and cancer; computational and experimental approachesCurr Pharm Biotechnol15429201410.2174/13892010150514082816133525189575 MatsuiMChuYZhangHGagnonKTShaikhSKuchimanchiSManoharanMCoreyDRJanowskiBAPromoter RNA links transcriptional regulation of inflammatory pathway genesNucleic Acids Res411008610109201310.1093/nar/gkt777239990913905862 MajidSDarAASainiSYamamuraSHirataHTanakaYDengGDahiyaRMicroRNA-205-directed transcriptional activation of tumor suppressor genes in prostate cancerCancer11656375649201010.1002/cncr.25488207375633940365 XiaoMLiJLiWWangYWuFXiYZhangLDingCLuoHLiYmiRNA and cancer; computational and experimental approaches. MicroRNAs activate gene transcription epigenetically as an enhancer triggerRNA Biol1413261334201710.1080/15476286.2015.111248726853707 AramadhakaLRProrockADragulevBBaoYFoxJWConnectivity maps for biosimilar drug discovery in venoms: The case of Gila monster venom and the anti-diabetes drug Byetta®Toxicon69160167201310.1016/j.toxicon.2013.03.01823602926 WangJLiMWangYLiuXIntegrating subpathway analysis to identify candidate agents for hepatocellular carcinomaOnco Targets Ther912211230201610.2147/OTT.S97211270222814788366 StroblJSPetersonVATamoxifen-resistant human breast cancer cell growth: Inhibition by thioridazine, pimozide and the calmodulin antagonist, W-13J Pharmacol Exp Ther26318619319921403784 TakaderaTOhyashikiTCalmodulin inhibitor-induced apoptosis was prevented by glycogen synthase kinase-3 inhibitors in PC12 cellsCell Mol Neurobiol27783790200710.1007/s10571-007-9172-y17647100 MuLWangTChenYTangXYuanYZhaoYβ-Elemene enhances the efficacy of gefitinib on glioblastoma multiforme cells through the inhibition of the EGFR signaling pathwayInt J Oncol4914271436201610.3892/ijo.2016.362627498706 ParkerJJDionneKRMassarwaRKlaassenMForemanNKNiswanderLCanollPKleinschmidt-DemastersBKWaziriAGefitinib selectively inhibits tumor cell migration in EGFR-amplified human glioblastomaNeuro Oncol1510481057201310.1093/neuonc/not053237497853714155 AljohaniHKoncarRFZarzourAParkBSLeeSHBahassi elMROS1 amplification mediates resistance to gefitinib in glioblastoma cellsOncotarget62038820395201510.18632/oncotarget.3981259780314653012 BartonMKExemestane is effective for the chemoprevention of breast cancerCA Cancer J Clin61363364201110.3322/caac.2013121898372 PaganiOReganMMFrancisPAExemestane with ovarian suppression in premenopausal breast cancerN Engl J Med37113581359201410.1056/NEJMoa140403725271610 Van AstenKNevenPLintermansAWildiersHParidaensRAromatase inhibitors in the breast cancer clinic: Focus on exemestaneEndocr Relat Cancer21R31R49201410.1530/ERC-13-026924434719 KritikouIGiannopoulouEKoutrasAKLabropoulouVTKalofonosHPThe combination of antitumor drugs, exemestane and erlotinib, induced resistance mechanism in H358 and A549 non-small cell lung cancer (NSCLC) cell linesPharm BiolNov52013(Epub ahead of print)24192333 CheTFLinCWWuYYChenYJHanCLChangYLWuCTHsiaoTHHongTMYangPCMitochondrial translocation of EGFR regulates mitochondria dynamics and promotes metastasis in NSCLCOncotarget63734937366201510.18632/oncotarget.5736264973684741934 CaoXZhuHAli-OsmanFLoHWEGFR and EGFRvIII undergo stress- and EGFR kinase inhibitor-induced mitochondrial translocalization: A potential mechanism of EGFR-driven antagonism of apoptosisMol Cancer1026201110.1186/1476-4598-10-26213885433063231 DasariVRVelpulaKKAlapatiKGujratiMTsungAJCord blood stem cells inhibit epidermal growth factor receptor translocation to mitochondria in glioblastomaPLoS One7e31884201210.1371/journal.pone.0031884223481363279427

Experimental flowchart of the present study. GEO, Gene Expression Omnibus; DEMs, differently expressed miRNAs; miRNA, microRNA; GEPIA, Gene Expression Profiling Interactive Analysis; DEGs, differently expressed genes; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of genes and genomes; PPI, protein-protein interaction.

DEMs in GBM based on data from the Gene Expression Omnibus. (A) Volcano plots for DEMs in GBM based on the datasets for (A) GSE25631, (B) GSE42657 and (C) GSE61710. (D) Heatmap for the five top upregulated and five top downregulated miRNAs in GBM using the RobustRankAggreg method with a value of P<0.05. DEMs, differentially expressed miRNAs; GBM, glioblastoma multiforme; Down, downregulated; Up, upregulated; miR, microRNA.

Violin plots for the expression of the 10 DEMs in the 3 gene chips. (A) Expression of upregulated miRNAs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b and hsa-miR-542-3p) in GSE25631. (B) Expression of five downregulated miRNAs (hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p) in GSE25631. (C) Expression of upregulated miRNAs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b and hsa-miR-542-3p) in GSE42657. (D) Expression of downregulated miRNAs (hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p) in GSE42657. (E) Expression of upregulated miRNAs (hsa-miR-196a, hsa-miR-10b, hsa-miR-196b, hsa-miR-18b and hsa-miR-542-3p) in GSE61710. (F) Expression of downregulated miRNAs (hsa-miR-219-2-3p, hsa-miR-1224-5p, hsa-miR-129-3p, hsa-miR-876-3p and hsa-miR-770-5p) in GSE61710. DEMs, differentially expressed miRNAs; miR, microRNA; GBM, glioblastoma multiforme.

Identification of 390 miRNA-associated differently expressed genes. (A) Volcano plot for DEGs in GBM from the Gene Expression Profiling Interactive Analysis. The volcano plot was generated using the R package ‘ggplot2’. (B) Venn diagram of overlapping genes from the DEGs in GBM and the target genes of the 10 differently expressed miRNAs. miRNA, microRNA; DEGs, differentially expressed genes; GBM, glioblastoma multiforme.

Significantly enriched Gene Ontology enrichment annotations and KEGG pathways of the 390 miRNA-associated differently expressed genes in glioblastoma multiforme. (A) Top 10 biological process terms (B) Top 10 cellular component terms. (C) Top 10 molecular function terms. (D) Significantly enriched KEGG pathways (P<0.05). The GO and KEGG pathway analyses were performed using the ClusterProfiler package in R. The GO terms were visualized using the R package ‘GOplot’ and the KEGG terms were visualized using the enrichMap function in the ClusterProfiler package. The enriched terms color changes gradually from deep to light in ascending order according to the P-values. KEGG, Kyoto Encyclopedia of genes and genomes.

Glioma pathway from Kyoto Encyclopedia of genes and genomes pathway analysis. The glioma pathway was visualized using the R package ‘pathview’. The color of the eight glioblastoma multiforme associated genes changed gradually from green to red in ascending order of the log2 (fold change) of each gene.

Box plots for the expression of the eight genes in glioblastoma multiforme determined from the Gene Expression Profiling Interactive Analysis: (A) CDK6, (B) RB1, (C) NRAS, (D) PDGFRA, (E) CAMK2G, (F) CAMK2B, (G) PRKCB and (H) CALM3. Num, number; T, tumor; N, normal; CDK6, cyclin-dependent kinase 6; RB1, retinoblastoma-associated protein; NRAS, GTPase NRas; PDGFRA, platelet-derived growth factor receptor alpha; CAMK2G, calcium/calmodulin-dependent protein kinase type II subunit gamma; CAMK2B, calcium/calmodulin-dependent protein kinase type II subunit beta; PRKCB, protein kinase C beta type; CALM3, calmodulin 3.

Immunohistochemistry staining of the proteins encoded by the 8 miRNA-associated differently expressed genes in the glioma pathway obtained from The Human Protein Atlas. (A) Low CDK6 expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000105810-CDK6/tissue/cerebral+cortex#img). (B) High CDK6 expression in glioblastoma multiforme (GBM) tissues (https://www.proteinatlas.org/ENSG00000105810-CDK6/pathology/tissue/glioma#img). (C) Low RB1 expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000139687-RB1/tissue/cerebral+cortex#img). (D) Moderate RB1 expression in GBM tissue (https://www.proteinatlas.org/ENSG00000139687-RB1/pathology/tissue/glioma#img). (E) NRAS protein was not detected in normal brain tissues (https://www.proteinatlas.org/ENSG00000213281-NRAS/tissue/cerebral+cortex#img). (F) Moderate NRAS expression in GBM tissue (https://www.proteinatlas.org/ENSG00000213281-NRAS/pathology/tissue/glioma#img). (G) PDGFRA protein was not detected in normal brain tissues (https://www.proteinatlas.org/ENSG00000134853-PDGFRA/tissue/cerebral+cortex#img). (H) Moderate PDGFRA expression in GBM tissues (https://www.proteinatlas.org/ENSG00000134853-PDGFRA/pathology/tissue/glioma#img). (I) High CAMK2G expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000148660-CAMK2G/tissue/cerebral+cortex#img). (J) Moderate CAMK2G expression in GBM tissues (https://www.proteinatlas.org/ENSG00000148660-CAMK2G/pathology/tissue/glioma#img). (K) Moderate CAMK2B expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000058404-CAMK2B/tissue/cerebral+cortex#img). (L) Weak CAMK2B expression in GBM tissues (https://www.proteinatlas.org/ENSG00000058404-CAMK2B/pathology/tissue/glioma#img). (M) Moderate PRKCB expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000166501-PRKCB/tissue/cerebral+cortex#img). (N) Weak PRKCB expression in GBM tissues (https://www.proteinatlas.org/ENSG00000166501-PRKCB/pathology/tissue/glioma#img). (O) Moderate CALM3 expression in normal brain tissues (https://www.proteinatlas.org/ENSG00000160014-CALM3/tissue/cerebral+cortex#img). (P) Low CALM3 expression in GBM tissues (https://www.proteinatlas.org/ENSG00000160014-CALM3/pathology/tissue/glioma#img). Magnification of each, ×100. CDK6, cyclin-dependent kinase 6; RB1, retinoblastoma-associated protein; NRAS, GTPase NRas; PDGFRA, platelet-derived growth factor receptor alpha; CAMK2G, calcium/calmodulin-dependent protein kinase type II subunit gamma; CAMK2B, calcium/calmodulin-dependent protein kinase type II subunit beta; PRKCB, protein kinase C beta type; CALM3, calmodulin 3.

Protein-protein interaction network of the eight miRNA-associated differently expressed genes enriched in the glioma pathway. CDK6, cyclin-dependent kinase 6; RB1, retinoblastoma-associated protein; NRAS, GTPase NRas; PDGFRA, platelet-derived growth factor receptor alpha; CAMK2G, calcium/calmodulin-dependent protein kinase type II subunit gamma; CAMK2B, calcium/calmodulin-dependent protein kinase type II subunit beta; PRKCB, protein kinase C beta type; CALM3, calmodulin 3.

3D-structure of the three compounds identified by the Connective map analysis. (A) W-13, (B) gefitinib and (C) exemestane.

Binding mode of W-13 and proteins, (A) cyclin-dependent kinase 6, (B) retinoblastoma-associated protein, (C) calcium/calmodulin-dependent protein kinase type II subunit gamma, (D) calcium/calmodulin-dependent protein kinase type II subunit beta, (E) GTPase NRas, (F) protein kinase C beta type and (G) platelet-derived growth factor receptor alpha.

Binding mode of gefitinib and proteins, (A) cyclin-dependent kinase 6, (B) retinoblastoma-associated protein, (C) calcium/calmodulin-dependent protein kinase type II subunit gamma, (D) calcium/calmodulin-dependent protein kinase type II subunit beta, (E) GTPase NRas, (F) protein kinase C beta type and (G) platelet-derived growth factor receptor alpha.

Binding mode of exemestane and proteins, (A) cyclin-dependent kinase 6, (B) retinoblastoma-associated protein, (C) calcium/calmodulin-dependent protein kinase type II subunit gamma, (D) calcium/calmodulin-dependent protein kinase type II subunit beta, (E) GTPase NRas, (F) protein kinase C beta type and (G) platelet-derived growth factor receptor alpha.

Basic information of the three gene chips obtained from Gene Expression Omnibus.

First author (publication year) Country Data source Platform Sample size (T/N) (Refs.)
Zhang et al (2012) China GSE25631 GPL8179 82/5 (31)
Jones et al (2015) United Kingdom GSE42657 GPL8179 5/7 (32)
Piwecka et al (2015) Poland GSE61710 GPL10656 10/5 (33)

T, tumor; N, normal control; Ref., reference.

Five top upregulated miRNAs and five top downregulated miRNAs from Gene Expression Omnibus.

Literature retrieval
Author miRNA RobustRankAggreg Expression Function (Refs.)
Dou et al hsa-miR-196a Up Polymorphism (37)
Yang et al Up Tumor growth (38)
Guan et al Up None (39)
Yang et al Up None (40)
Sasayama et al hsa-miR-10b Up Up Cell invasion (41)
Guessous et al Up Cell invasion, cell migration, tumor growth (42)
Gabriely et al Up Cell proliferation, cell death (43)
Ji et al Up None (44)
Guan et al hsa-miR-196b Up Up None (45)
Lakomy et al Up None (46)
Ma et al Up None (47)
You et al Up None (48)
Karsy M et al Up None (49)
hsa-miR-18b Up None No ref.
Cai et al hsa-miR-543-3p Up Down Cell invasion (50)a
hsa-miR-219-2-3p Down None No ref.
Qian et al hsa-miR-1224-5p Down Down Cell proliferation, cell invasion, cell apoptosis (51)
Ouyang et al hsa-miR-129-3p Down Down Cell proliferation, tumor growth (52)
Fang et al Down Cell viability, cell growth (53)
hsa-miR-876-3p Down None No ref.
hsa-miR-770-5p Down None No ref.

The expression of hsa-miR-542-3p was inconsistent with our result. Ref., reference.

Three chemicals identified as therapeutic agents for glioblastoma multiforme from CMap analysis.

CMap name Enrichment Dose Cell lines Up score Down score
W-13 −0.989 10 µM MCF7 −0.356 0.532
Gefitinib −0.989 10 µM HL60 −0.289 0.479
Exemestane −0.981 10 nM MCF7 −0.382 0.351

CMap, connectivity map.

Results of protein-ligand docking.

Gene symbol Protein name PDB ID Chemical Total score Crash Polar
CDK6 Cyclin-dependent kinase 6 1BI7 W-13 6.9158 −0.6018 1.0852
Gefitinib 4.5461 −1.5497 1.0374
Exemestane 4.2617 −0.6835 1.2109
RB1 Retinoblastoma-associated protein 2QDJ W-13 6.1141 −0.7408 2.1589
Gefitinib 5.1361 −3.0039 1.0429
Exemestane 4.5673 −1.2303 2.2315
CAMK2G Calcium/calmodulin-dependent protein kinase type II subunit gamma 2V7O W-13 6.5899 −0.9631 2.1155
Gefitinib 5.5314 −0.8239 4.6009
Exemestane 5.5085 −2.4882 2.8500
CAMK2B Calcium/calmodulin-dependent protein kinase type II subunit beta 3BHH W-13 6.9781 −0.8004 2.9012
Gefitinib 5.4665 −1.7396 3.6928
Exemestane 4.7782 −1.8678 4.6573
NRAS GTPase NRas 3CON W-13 6.9580 −0.7966 3.5558
Gefitinib 5.1191 −1.6695 2.1113
Exemestane 3.7527 −1.3243 1.6902
PRKCB Protein kinase C beta type 3PFQ W-13 5.9087 −0.7642 4.3292
Gefitinib 5.9886 −0.9356 2.5291
Exemestane 3.5943 −0.6226 2.3138
PDGFRA Platelet-derived growth factor receptor alpha 5K5X W-13 6.7711 −1.3263 4.9966
Gefitinib 4.0157 −1.0638 1.1911
Exemestane 4.8780 −0.9005 2.6185

PDB ID: protein data bank identification.