Topotecan is a semisynthetic analog of the alkaloid camptothecin and is commonly used in antineoplastic chemotherapy.1 It functions by inhibiting topoisomerase-I, which is responsible for opening single-strand breaks during DNA replication.2 Lethal DNA damage to cells occurs when DNA is replicated and topotecan suppresses topoisomerase-I.3 Topotecan is currently used for metastatic ovarian carcinoma,4 recurrent small cell lung cancer,5 and recurrent cervical cancer or stage IVB disease.6
Previous studies have demonstrated that topotecan has promising anticancer effects, but the value of topotecan has been equivocal in some studies on recurrent cervical cancer and ovarian cancer.7,8 Topotecan has also been investigated for the treatment of glioma,9 Ewing sarcoma,10 and neuroblastoma,11 but the efficacy has yet to be confirmed conclusively. This study was done to identify potentially critical genes and pathways associated with topotecan using publicly accessible bioinformatics tools. We discuss the interactions between these pivotal genes, their prognostic value, and the implications of the genes and their pathways in resistance to topotecan.
2.1. Library of Integrated Network-Based Cellular data
The National Institutes of Health Library of Integrated Network-Based Cellular Signatures (LINCS) Program (http://www.ilincs.org/ilincs/) is an integrative web platform for LINCS data and signatures based on an R analytical engine and other web-based tools. The goal of this platform is to create a collection of cellular phenotypes based on LINCS that describe how cells react to specific perturbagens. The LINCS include transcriptomic and proteomic information from cells treated with specific perturbagens and are available in the public domain. Therefore, we can use these data to understand the gene expression changes caused by drugs of interest. All 11 topotecan signatures were downloaded from the LINCS online platform to identify differentially expressed genes (DEGs) for further analysis. These signatures were obtained from the treatment of HA1E, A375, HT29, MCF7, PC3, and A549 cancer cell lines with 10 µM of topotecan for either 6 or 24 hours.
2.2. Differentially expressed gene identification
The 11 signatures for topotecan-treated cells were screened individually for DEGs. The fold change in expression was determined using averages and standard deviations for the differential gene expression values associated with the selected genes. DEGs were required to satisfy both p value and fold change criteria simultaneously.12 Most studies suggest that genes are differentially expressed if they show a fold change of at least 1.5 and p value <0.05.13 In our study, DEGs were defined by the following cutoffs: p value <0.05, fold change of 50% or |log2FC| ≥0.58, and appearance at least three times.
We added genes that appear at least three times to narrow down the targets. For a specific upregulated gene, its average plus 1 standard deviation was deemed as significant if log2FC ≥0.58. In contrast, for downregulated genes, fold change values were obtained by averages minus 1 standard deviation, and log2FC ≤−0.58 was used as a cutoff value for significance.
2.3. Construction of the protein-protein interaction network and hub gene identification
We observed the cellular mechanisms involved in the response to topotecan by using DEGs that were identified more than once to construct a protein-protein interaction (PPI) network. We performed the PPI prediction using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, https://string-db.org/).14 PPI network visualization was performed using Cytoscape software (version 3.6.1, http://www.cytoscape.org, provided by the U.S. National Institute of General Medical Sciences). Hub genes were identified by evaluating the following parameters: node degrees, betweenness, and eigenfactor scores, which were obtained using the Cytoscape plugin CentiScaPe, http://apps.cytoscape.org/apps/centiscape.
2.4. Functional and pathway enrichment analysis of hub genes and DEGs with significant log2FC values
The Database for Annotation, Visualization, and Integrated Discovery (DAVID, https://david.ncifcrf.gov/) is an online tool for functional and pathway enrichment analysis based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. Three DEG groups (frequent genes, genes with significant log2FC, and hub genes) were used to create a Venn diagram to obtain a list of DEGs that appeared in at least two of the three DEG groups. This list was used to query DAVID for functional and pathway enrichment analyses. The top-10 results from the KEGG pathway analysis were recorded. All of the recorded results had p values <0.05.
2.5. Clinical prognostic values of identified key genes in ovarian cancer
We explored the association of hub genes with the clinical prognosis using genomic datasets from ovarian cancer patients from The Cancer Genome Atlas (TCGA, http://cancergenome.nih.gov) database. Gene expression profiles and the clinical data of the ovarian cancer patients were downloaded from the Genomic Data Commons (GDC, https://gdc.cancer.gov/), which is a cancer data repository managed by the National Cancer Institute (NCI). For the analysis, we obtained gene expression profiles based on RNA sequencing and fragments per kilobase of transcript per million mapped reads upper quartiles, which were derived from the primary tumor masses of 476 ovarian cancer patients.
To test the association of gene expression with patient survival times, the 476 ovarian cancer patients were divided into two or four groups depending on either a single-gene or a two-gene combination. Each grouping was performed by taking the median expression level of the target gene in all 476 patients as a cutoff. Therefore, for each single-gene survival test, the ovarian cancer patients were divided into a high expression group and a low expression group, and a log-rank test was used to compare survival curves between the two groups.
Likewise, for each set of the two-gene survival tests (for example, genes A and B), patients were divided into four groups according to the genes’ expression: high A versus high B, high A versus low B, low A versus high B, and low A versus low B. The log-rank test was then used to assess the differences in survival between each pair of the four groups. All the log-rank survival tests were performed using the R survival package.
2.6. MTT assays
Murine ovarian (MOSEC), colorectal (CT26), and lung (LLC) cancer cells were plated in a 96-well plate (5000 cells per well) and incubated with increasing concentrations of topotecan. After 3 days of topotecan treatment, the cells were subjected to the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay by adding 25 µL of MTT solution [5 mg/mL in phosphate buffered saline (PBS)] to each well. After 2 hours of incubation, the MTT solution was removed, the cells were washed with PBS, and 0.1 mL of the extraction buffer was added (20% sodium dodecyl sulfate [SDS] in 50% dimethylformamide). After 4 hours of incubation at 37°C, the optical densities were measured at 570 nm, and the extraction buffer alone was used as a blank.
2.7. Western blotting
Murine ovarian (MOSEC), colorectal (CT26), and lung (LLC) cancer cells were treated with 20 μM of topotecan for 2–6 hours. Cells were then treated with lysis buffer and analyzed using 12% SDS polyacrylamide gel electrophoresis (PAGE). The proteins separated by SDSPAGE were then electrophoretically transferred onto polyvinylidene difluoride membranes. The membranes were first treated with blocking buffer, followed by incubation with antibodies, including anti-Ezh2, anti-p21, and anti-b-actin.
Anti-rabbit or anti-mouse horseradish-conjugated secondary antibodies diluted in Tris-buffered saline-Tween were added to the washed membranes. The amounts of individual proteins were visualized using a GE Amersham 600 gel imager (Little Chalfont, United Kingdom) after the addition of peroxidase substrate. The protein expression ratio was determined using ImageQuant TL 8.1 (GE Healthcare Life Sciences, https://www.gelifesciences.com/en/us).
2.8. Quantitative real-time polymerase chain reaction
MOSEC, CT26, and LLC cells were treated with increasing concentrations of topotecan for various time periods as those prepared for Western blottings. The total RNA of different samples was isolated using Trizol reagent (Invitrogen, Carlsbad, CA). Complementary DNA (cDNA) was synthesized from 1.0 μg of total RNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, https://www.thermofisher.com/tw/zt/home/brands/applied-biosystems.html). Fast SYBR Green Master Mix from Applied Biosystems was mixed with 2 μL of diluted cDNA for each sample. The cDNAs were amplified and quantified using the StepOne plus PCR system (Applied Biosystems).
The primers used for quantitative real-time polymerase chain reaction (qRT-PCR) were p21-F-5′-CCTGGTGATGTCCGACCTG, R-5′-CCATGAGCGCATCGCAATC, EZH2-F-5′-TGGACCACAGTGTTACCAGCA; and R-5′-TGGGCGTTTAGGTGGTGTCT. For the internal control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-F-5′-TGGTTCACACCCATCACAAACA and R-5′-GGTGAAGGTCGGTGTGAACGG were used. The relative amount of each gene in each sample was calculated using the ΔΔCT method. The GAPDH RNA level was used as an internal control for normalization of the results.
3.1. DEG identification and further grouping
The workflow of this study is illustrated in Figure 1A. Total of 11 topotecan LINCS signatures were identified and downloaded. Genes that appeared at least three times with p values <0.05 and |log2FC| values ≥0.58 were identified as DEGs. A total of 65 upregulated and 87 downregulated DEGs fulfilled these criteria (Table 1). NFKBIA and CDC25B were the most frequently identified upregulated and downregulated DEGs, respectively, and appeared seven times in the 11 signatures.
3.2. PPI network construction and hub DEG identification
The DEGs were used to construct a PPI network with STRING. The PPI value was a default score greater than 0.4. The PPI network contained 450 edges with an average node degree of 5.9 and an enrichment p value of 1 × 10−16 (Fig. 2A). We identified the hub genes by performing a centrality analysis on the PPI network using the Cytoscape plugin CentiScaPe.
The parameters evaluated for the centrality analysis were the node degree, betweenness, and eigenfactor scores. The degree of a node is the number of adjacent nodes that are connected to it and interact with it. The betweenness of a node reflects the number of shortest routes between two nodes that have to pass through the target node. The eigenfactor of a node is a relative score that is determined by the number of high-degree nodes that are connected to the target node.
CentiScaPe uses the average score of the PPI network for all three parameters as a threshold to select nodes that are above or at the threshold. Three lists of genes were found to be above average for a specific parameter and entered into the Pangloss Venn diagram generator (http://www.pangloss.com/seidel/Protocols/venn.cgi). This allowed us to detect nodes that were above average for all three parameters (Fig. 2B). As such, 21 DEGs were identified as hub genes (http://links.lww.com/JCMA/A50), of which 8 were upregulated and 13 were downregulated. These 21 genes were used to construct a simpler and better-connected PPI network, and again, STRING was used to identify PPIs, while Cytoscape was used for visualization. The PPI network constructed using the hub genes identified from the centrality analysis consisted of 21 nodes and 95 edges with an enrichment p value of 8 × 10−15 (Fig. 2C).
Highly important DEGs were identified by cross-referencing 21 hub genes and the top-40 DEGs (Supplement 2, http://links.lww.com/JCMA/A50) using a Venn diagram. Eight key genes were identified in both lists: NFKBIA, IKBKB, CDK7, CDKN1A, GADD45A, EZH2, CDC20, and HIST2H2BE (Fig. 2D). The PPI network is shown in Fig. 2E. Its average node degree is 2.75 edges, and its PPI enrichment p value is 3 × 10−4.
3.3. Functional and pathway enrichment analysis
KEGG pathway mapping was performed using DAVID to gain insights into the biological significance of the 65 upregulated and 87 downregulated DEGs in Table 1. The top-10 results from the KEGG pathway analysis were recorded (Supplement 3, http://links.lww.com/JCMA/A50 and Supplement 4, http://links.lww.com/JCMA/A50). All results had p values <0.05.
The KEGG pathway analysis for the top-10 upregulated DEGs (Supplement 3, http://links.lww.com/JCMA/A50) showed the enrichment of multiple pathways. The prominent pathways were related to human T cell lymphotropic virus type 1 (HTLV-I) infection, the p53 signaling pathway, Epstein-Barr virus infection, chronic myeloid leukemia, the forkhead box transcription factors of the class O (FoxO) signaling pathway, apoptosis, the cell cycle, the mitogen-activated protein kinase (MAPK) signaling pathway, hepatitis B, and viral carcinogenesis. The KEGG pathways that were enriched for the top-10 downregulated DEGs (Supplement 4, http://links.lww.com/JCMA/A50) included cell cycle, progesterone-mediated oocyte maturation, viral carcinogenesis, adenosine monophosphate-activated protein kinase (AMPK) signaling pathway, Epstein-Barr virus infection, and FoxO signaling pathways.
For the eight key DEGs of high importance, the KEGG pathways included the cell cycle, viral carcinogenesis, HTLV-I infection, chronic myeloid leukemia, and prostate cancer. IKBKB, CDKN1A, and EZH2 were revealed in the KEGG pathway of microRNAs in cancer with pathway ID hsa05206, albeit with a p value of 0.11 (Table 2). This result indicates that perturbations of IKBKB, CDKN1A, and EZH2 gene transcription by topotecan might have some effects on pathways related to cancer microRNA.
3.4. Clinical prognostic values of identified key genes in ovarian cancer
We used the TCGA database to explore the clinical prognostic values of the key genes in ovarian cancer patients, given that topotecan is one of the standard chemotherapeutics for ovarian cancer. The TCGA database indicates the prognosis of pretreatment populations. None of the eight genes was identified as a prognostic biomarker, but interestingly, EZH2 and CDKN1A appeared to form a synthetically viable gene pair. The suppressed expression of either gene alongside the activated expression of the other gene led to significantly worse survival (Fig. 3) in comparison with the survival curves obtained with the simultaneous up- or downregulation of both genes.
3.5. Effects of topotecan on viability; CDKN1A and EZH2 gene transcriptions; and encoded protein levels of MOSEC, CT26, and LLC cells
We evaluated the cytotoxic effects of topotecan on the viability of cancer cells by conducting an MTT assay. Cells were incubated with increasing concentrations of topotecan for 3 days. Topotecan displayed indistinct cytotoxic effects on the three tested cell lines, MOSEC, CT26, and LLC, with half maximal inhibitory concentration (IC50) values of 0.9, 1.4, and 9.6 μM, respectively (Fig. 4). The IC50 values of topotecan cytotoxic effects on these cell lines after 3 days of incubation were in the range of 1 to 10 µM. Therefore, we explored the effects of topotecan with contractions of 0, 10, 20, and 40 ∝M on CDKN1A and EZH2 gene transcriptions, as well as the amounts of encoded proteins, p21 and Ezh2, using qRT-PCR and Western blotting.
Topotecan treatment enhanced CDKN1A-encoded p21 expression, especially at 10 and 20 ∝M with time points of 4 and 6 hours (Fig. 5A–C). However, we observed Ezh2 suppression when the three cell lines were treated with high concentrations of topotecan at 40 μM for longer durations of up to 4 and 6 hours. Nevertheless, no generalized dose- or time-dependent changes were observed.
We validated the topotecan-induced changes in gene transcriptions of CDKN1A and EZH2 by conducting qRT-PCR to measure the CDKN1A and EZH2 mRNA levels of MOSEC, LLC, and CT26 cells. The cells had been treated with various concentrations of topotecan for 2, 4, or 6 hours (Fig. 5D–F). The upregulation of CDKN1A mRNA levels and downregulation of EZH2 mRNA levels are in line with the results derived from the LINCS database. The differential transcription of CDKN1A and EZH2 may account for the observed changes in their protein levels in topotecan-treated cells.
We identified eight overlapping DEGs: the upregulated DEGs CDKN1A, GADD45A, NFKBIA, IKBKB, and HIST2H2BE and the downregulated DEGs EZH2, CDC20, and CDK7. The KEGG pathway analysis of these eight genes yielded similar results to those for the 65 upregulated and 87 downregulated DEGs, including cell cycle-related kinase activity, viral carcinogenesis, HTLV-1 infection, chronic myeloid leukemia, and prostate cancer. We discovered that topotecan could upregulate p21 as well as downregulate Ezh2 through in vitro studies including the MTT assay, Western blotting, and qRT-PCR. Furthermore, we found that ovarian cancer patients with high CDKN1A expression and low EZH2 expression demonstrate better clinical results.
Among the eight key genes, upregulated CDKN1A was enriched in the top-five KEGG pathways. CDKN1A encodes a small 165 amino acid protein known as p21, which could bind and inhibit the activity of cyclin-CDK2 or cyclin-CDK4 complexes and regulate cell cycle arrest in the Gap 2/Mitotic phase.15 Usually, p21 acts as a tumor suppressor, and tumor proliferation has been noted in osteosarcoma and cervical cancer cells when inhibiting p21 expression.16 However, p21 sometimes behaves as an oncogene in certain conditions.17 One mouse study showed that lymphomas arising in p21 knockout mice exhibit a high rate of apoptosis, indicating a protumorigenic characteristic of p21 by preventing the apoptotic activity of tumor cells. However, it is a challenging task to target p21 for drug development due to its complex network activity.
GADD45A encodes the growth arrest and DNA damage-inducible alpha protein. It is another upregulated key gene that interferes in the interaction between proliferating cell nuclear antigen and cell division protein kinase complexes, preventing cells from entering the S phase.18NFKBIA was the most frequently identified upregulated gene in the 11 signatures from topotecan-treated cancer cell lines. NFKBIA encodes nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor (IKB)-alpha, which suppresses the activity of the nuclear factor (NF)-κB/Rel complex. A previous study indicated that the inappropriate activation of NF-κB is associated with inflammatory events. Furthermore, it is also responsible for tumorigenesis and promotes chemoresistance by suppressing apoptosis.19
Intriguingly, topotecan also upregulated IKBKB, which encodes IKB kinase beta. This is a catalytic subunit of the IKB kinase complex that enhances NF-κB signaling through phosphorylation of IKB-alpha and by leading to its degradation by the proteasome.20HIST2H2BE encodes histone cluster 2, which is a core component of nucleosomes that plays key roles in chromosomal stability, DNA replication, repair, and transcriptional regulation.21
The downregulated gene EZH2 is a member of the polycomb repressive complex 2 (PRC2), and its function is to promote the trimethylation of histone H3 lysine 27 (H3K27me3). This gene plays a critical role in cancer cell proliferation, tumor metastasis, and resistance after cancer treatment.22EZH2 also maintains the stemness of cells by repressing lineage-specifying differentiation factors.23EZH2 inhibition or a loss of EZH2 expression suppresses tumor growth and improves the survival in many cancer cells, including lung cancer cells,24 ovarian cancer cells,23 and glioblastoma multiforme (GBM) cells.25
CDC20 encodes cell division cycle protein 20 (Cdc20), which plays an essential role in the completion of cellular mitosis.19 Another downregulated cell cycle-related gene identified in this study is CDK7. This gene encodes a serine/threonine kinase that is critical for controlling cell cycle progression and RNA transcription mediated by RNA polymerase II.26
Given that topotecan is one of the standard regimens for ovarian cancer, we explored the prognostic values of these key genes using data from the TCGA database from 476 ovarian cancer patients. The TCGA database indicated the prognosis of pretreatment populations. None of the eight key genes could serve as a standalone prognostic biomarker, but the CDKN1A+/EZH2− group had the longest median survival, while the CDKN1A−/EZH2+ group had the shortest median survival among ovarian cancer patients. These findings are compatible with the better clinical results for cancer patients with downregulated EZH2 and upregulated CDKN1A.
We examined the p21 and Ezh2 expression levels in topotecan-treated cancer cells, including ovarian, colorectal, and lung cancer cells, with increasing concentrations of topotecan for 2, 4, and 6 hours. There was an increasing trend of p21 levels with increasing treatment time with 10, 20, and 40 µM of topotecan. In the case of Ezh2, there was a strong decreasing trend in CT26 and LCC and a noticeable decrease in expression at 6 hours in MOSEC. We also noted similar results for mRNA, although a dose-dependent effect was partially observed. Due to the role of Ezh2 as an anticancer drug target, its suppression by topotecan may be a newly identified mechanism of action.
Previous studies showed that EZH2 could bind to the promoter region of p21 and inhibit p21 expression.27 It also acts as an oncogene by regulating p21 in hepatocellular carcinoma,28 gastric cancer,29 non-small cell lung cancer,30 renal cell carcinoma cells,31 and human epithelial ovarian cancer.32EZH2 is related to the resistance of poly (adenosine diphosphate-ribose) polymerase (PARP) inhibitors through DNA replication fork protection in ovarian cancer models.33EZH2 inhibition also prevents the emergence of chemoresistance in small cell lung cancer24 and reverses resistance to cisplatin in ovarian cancer.34
EZH2 is highly expressed in many types of cancer, especially epithelial ovarian carcinomas, and EZH2 upregulation has been significantly correlated with advanced-stage and lymph node metastasis.23 In our study, topotecan could not only prevent cancer cell repair from lethal DNA damage but also downregulated EZH2 and upregulated CDKN1A. As a result, it inhibited tumor proliferation and lymph node metastasis, decreased the stemness property of cancer cells, and may improve the survival of cancer patients. EZH2 inhibitor treatment with topotecan may be a future direction to enhance tumor toxicity and prevent acquired chemoresistance.
Our study has some advantages. In light of the complex posttranslational modifications, the PPI derived from DEGs of the LINCS database using the STRING platform could reduce false positive and negative results. By using publicly accessible bioinformatics tools, the study approach is an efficient and low-cost way to find potential agents or target genes for investigation. Furthermore, we performed in vitro experiments to prove our concept, including Western blot and qRT-PCR, and we used TCGA for correlation with clinical relevance.
Our study also has several limitations. First, the interactions between DEG-encoded proteins derived from STRING are based on sources including genomic context predictions, high-throughput laboratory experiments, microarray coexpression, automated text mining, and previous knowledge from various databases.14 Therefore, it is not certain that the PPI that we obtained from the DEGs represents direct physical interactions between proteins.
Second, the PPI was constructed to provide a broader view on how these perturbed genes—and potentially, their encoded proteins—might account for the effects induced by topotecan. The mechanism may be different in various challenged cancer cell lines. Third, our study included in vitro experiments. In the future, additional indications of topotecan could be explored by animal studies in vivo using various tumor models based on the pathways identified in this study.
The survival data of simultaneous down- or upregulated coexpression of CDKN1A and EZH2 in ovarian cancer patients in the TCGA database suggested that the manipulation of the two proteins encoded by these two genes might enhance the survival in patients with this type of cancer. Moreover, additional indications of topotecan could be explored by animal studies in vivo using various tumor models based on the pathways identified in this study. Our findings provide insights into the potential molecular mechanisms of action of topotecan, which may lead to the innovative use of the drug alone or in combination with other anticancer therapeutics.
We would like to express our gratitude to National Yang-Ming University for supporting this study and all the members in Department of Oncology in Taipei Veterans General Hospital.
1. Klejewski A, Świerczewska M, Zaorska K, Brązert M, Nowicki M, Zabel M, et al. New and old genes
associated with topotecan
resistance development in ovarian cancer cell lines. Anticancer Res. 2017; 37:1625–36
2. Riemsma R, Simons JP, Bashir Z, Gooch CL, Kleijnen J. Systematic review of topotecan
(Hycamtin) in relapsed small cell lung cancer. BMC Cancer. 2010; 10:436
3. Sehouli J, Oskay-Ozcelik G. Current role and future aspects of topotecan
in relapsed ovarian cancer. Curr Med Res Opin. 2009; 25:639–51
4. Lorusso D, Pietragalla A, Mainenti S, Masciullo V, Di Vagno G, Scambia G. Review role of topotecan
in gynaecological cancers: current indications and perspectives. Crit Rev Oncol Hematol. 2010; 74:163–74
5. Horita N, Yamamoto M, Sato T, Tsukahara T, Nagakura H, Tashiro K, et al. Topotecan
for relapsed small-cell lung cancer: systematic review and meta-analysis of 1347 patients. Sci Rep. 2015; 5:15437
6. Musa F, Blank S, Muggia F. A pharmacokinetic evaluation of topotecan
as a cervical cancer therapy. Expert Opin Drug Metab Toxicol. 2013; 9:215–24
7. Monk BJ, Sill MW, McMeekin DS, Cohn DE, Ramondetta LM, Boardman CH, et al. Phase III trial of four cisplatin-containing doublet combinations in stage IVB, recurrent, or persistent cervical carcinoma: a Gynecologic Oncology Group study. J Clin Oncol. 2009; 27:4649–55
8. Gordon AN, Tonda M, Sun S, Rackoff W; Doxil Study 30-49 InvestigatorsLong-term survival advantage for women treated with pegylated liposomal doxorubicin compared with topotecan
in a phase 3 randomized study of recurrent and refractory epithelial ovarian cancer. Gynecol Oncol. 2004; 95:1–8
9. Porkholm M, Valanne L, Lönnqvist T, Holm S, Lannering B, Riikonen P, et al. Radiation therapy and concurrent topotecan
followed by maintenance triple anti-angiogenic therapy with thalidomide, etoposide, and celecoxib for pediatric diffuse intrinsic pontine glioma. Pediatr Blood Cancer. 2014; 61:1603–9
10. Hunold A, Weddeling N, Paulussen M, Ranft A, Liebscher C, Jürgens H. Topotecan
and cyclophosphamide in patients with refractory or relapsed Ewing tumors. Pediatr Blood Cancer. 2006; 47:795–800
11. Chaturvedi NK, McGuire TR, Coulter DW, Shukla A, McIntyre EM, Sharp JG, et al. Improved therapy for neuroblastoma using a combination approach: superior efficacy with vismodegib and topotecan
. Oncotarget. 2016; 7:15215–29
12. McCarthy DJ, Smyth GK. Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics. 2009; 25:765–71
13. Peart MJ, Smyth GK, van Laar RK, Bowtell DD, Richon VM, Marks PA, et al. Identification and functional significance of genes
regulated by structurally different histone deacetylase inhibitors. Proc Natl Acad Sci U S A. 2005; 102:3697–702
14. Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45D1D362–8
15. Cazzalini O, Scovassi AI, Savio M, Stivala LA, Prosperi E. Multiple roles of the cell cycle inhibitor p21(CDKN1A) in the DNA damage response. Mutat Res. 2010; 704:12–20
16. Zhang X, Li F, Zhu L. Clinical significance and functions of microRNA-93/CDKN1A axis in human cervical cancer. Life Sci. 2018; 209:242–8
17. Abbas T, Dutta A. p21 in cancer: intricate networks and multiple activities. Nat Rev Cancer. 2009; 9:400–14
18. Li D, Dai C, Yang X, Li B, Xiao X, Tang S. GADD45a regulates olaquindox-induced DNA damage and S-phase arrest in human hepatoma G2 cells via JNK/p38 pathways. Molecules. 2017; 22:124
19. Kapanidou M, Curtis NL, Bolanos-Garcia VM. Cdc20: at the crossroads between chromosome segregation and mitotic exit. Trends Biochem Sci. 2017; 42:193–205
20. Hu H, Snyder JP. Models for predicting IKKA and IKKB blockade. J Chem Inf Model. 2012; 52:3190–9
21. Tessarz P, Kouzarides T. Histone core modifications regulating nucleosome structure and dynamics. Nat Rev Mol Cell Biol. 2014; 15:703–8
22. Poirier JT, Gardner EE, Connis N, Moreira AL, de Stanchina E, Hann CL, et al. DNA methylation in small cell lung cancer defines distinct disease subtypes and correlates with high expression of EZH2. Oncogene. 2015; 34:5869–78
23. Jones BA, Varambally S, Arend RC. Histone methyltransferase EZH2: a therapeutic target for ovarian cancer. Mol Cancer Ther. 2018; 17:591–602
24. Gardner EE, Lok BH, Schneeberger VE, Desmeules P, Miles LA, Arnold PK, et al. Chemosensitive relapse in small cell lung cancer proceeds through an EZH2-SLFN11 axis. Cancer Cell. 2017; 31:286–99
25. Stazi G, Taglieri L, Nicolai A, Romanelli A, Fioravanti R, Morrone S, et al. Dissecting the role of novel EZH2 inhibitors in primary glioblastoma cell cultures: effects on proliferation, epithelial-mesenchymal transition, migration, and on the pro-inflammatory phenotype. Clin Epigenetics. 2019; 11:173
26. Fisher RP. Secrets of a double agent: CDK7 in cell-cycle control and transcription. J Cell Sci. 2005; 118Pt 225171–80
27. Mu X, Chen M, Xiao B, Yang B, Singh S, Zhang B. EZH2 confers sensitivity of breast cancer cells to taxol by attenuating p21 expression epigenetically. DNA Cell Biol. 2019; 38:651–9
28. Xu X, Gu J, Ding X, Ge G, Zang X, Ji R, et al. LINC00978 promotes the progression of hepatocellular carcinoma by regulating EZH2-mediated silencing of p21 and E-cadherin expression. Cell Death Dis. 2019; 10:752
29. Xu J, Wang Z, Lu W, Jiang H, Lu J, Qiu J, et al. EZH2 promotes gastric cancer cells proliferation by repressing p21 expression. Pathol Res Pract. 2019; 215:152374
30. Liu JQ, Feng YH, Zeng S, Zhong MZ. linc01088 promotes cell proliferation by scaffolding EZH2 and repressing p21 in human non-small cell lung cancer. Life Sci. 2020; 241:117134
31. Dong JS, Wu B, Jiang B. LncRNA SNHG7 promotes the proliferation and inhibits apoptosis of renal cell cancer cells by downregulating CDKN1A. Eur Rev Med Pharmacol Sci. 2019; 23:10241–7
32. Li Y, Jiao Y, Hao J, Xing H, Li C. Long noncoding RNA TP73-AS1 accelerates the epithelial ovarian cancer via epigenetically repressing p21. Am J Transl Res. 2019; 11:2447–54
33. Kubalanza K, Konecny GE. Mechanisms of PARP inhibitor resistance in ovarian cancer. Curr Opin Obstet Gynecol. 2020; 32:36–41
34. Sun Y, Jin L, Liu JH, Sui YX, Han LL, Shen XL. Interfering EZH2 expression reverses the cisplatin resistance in human ovarian cancer by inhibiting autophagy. Cancer Biother Radiopharm. 2016; 31:246–52