Skip to main content
NCBI home page
International Journal of Clinical and Experimental Pathology logoLink to International Journal of Clinical and Experimental Pathology
. 2015 Jun 1;8(6):6847–6858.

Analysis of microarray-identified genes and microRNAs associated with drug resistance in ovarian cancer

Jing Zou 1,*, Fuqiang Yin 2,*, Qi Wang 1, Wei Zhang 1, Li Li 1
PMCID: PMC4525906  PMID: 26261572

Abstract

The aim of this study was to identify potential microRNAs and genes associated with drug resistance in ovarian cancer through web-available microarrays. The drug resistant-related microRNA microarray dataset GS54665 and mRNA dataset GSE33482, GSE28646, and GSE15372 were downloaded from the Gene Expression Omnibus database. Dysregulated microRNAs/genes were screened with GEO2R and were further identified in SKOV3 (SKOV3/DDP) and A2780 (A2780/DDP) cells by real-time quantitative PCR (qRT-PCR), and then their associations with drug resistance was analyzed by comprehensive bioinformatic analyses. Nine microRNAs (microRNA-199a-5p, microRNA-199a-3p, microRNA-199b-3p, microRNA-215, microRNA-335, microRNA-18b, microRNA-363, microRNA-645 and microRNA-141) and 38 genes were identified to be differentially expressed in drug-resistant ovarian cancer cells, with seven genes (NHSL1, EPHA3, USP51, ZSCAN4, EPHA7, SNCA and PI15) exhibited exactly the same expression trends in all three microarrays. Biological process annotation and pathway enrichment analysis of the 9 microRNAs and 38 genes identified several drug resistant-related signaling pathways, and the microRNA-mRNA interaction revealed the existence of a targeted regulatory relationship between the 9 microRNAs and most of the 38 genes. The expression of 9 microRNAs and the 7 genes by qRT-PCR in SKOV3/DDP and A2780/DDP cells indicating a consistent expression profile with the microarrays. Among those, the expression of EPHA7 and PI15 were negatively correlated with that of microRNA-141, and they were also identified as potential targets of this microRNA via microRNA-mRNA interaction. We thus concluded that microRNA-141, EPHA7, and PI15 might jointly participate in the regulation of drug resistance in ovarian cancer and serve as potential targets in targeted therapies.

Keywords: Ovarian cancer, drug resistance, expression profiling microarray, gene, microRNA, bioinformatics

Introduction

Ovarian cancer (epithelial carcinoma) is one of the three major malignancies of the female reproductive system. The mortality rate of ovarian cancer is the highest among all gynecological malignancies, and the prognosis of ovarian cancer is generally poor [1]. Currently, surgery is the primary treatment for ovarian cancer, and this treatment is usually supplemented with chemotherapy to deliver a comprehensive therapy. Platinum-based combination chemotherapy improves the overall response rate, the clinical remission rate, and the median survival rate in patients with ovarian cancer. However, primary and acquired multidrug resistance (MDR) remains the main obstacles to the success of chemotherapeutic treatment in clinical practice [2]. Therefore, an understanding of the mechanisms underlying the development and reversal of drug resistance in ovarian cancer patients is of great clinical significance for enhancing the survival rates and improving the prognosis of these patients.

In recent years, data mining has been conducted at the molecular level using bioinformatic approaches such as biochip data extraction, sequence alignment, statistical analysis, visual mapping, biological data clustering, pathway analysis, and promoter prediction, and these analyses provide novel research ideas for studying the molecular pathogenesis of various diseases, including tumors. For example, through a comprehensive bioinformatic analysis, Yin et al. [3] identified 25 genes (including AKT1/2D) that are associated with drug resistance in ovarian cancer. In the present study, microarray data detailing the messenger RNA (mRNA) and microRNA expression profiles in drug-sensitive and drug-resistant ovarian cancer cells were retrieved from the Gene Expression Omnibus (GEO) database [4] and were analyzed to identify differentially expressed genes and microRNAs. The correlation between differentially expressed genes and microRNAs in drug-resistant and drug-sensitive ovarian cancer was examined through comprehensive bioinformatics analysis including biological process annotation, biological pathway enrichment, protein/gene interaction analysis, and mRNA-microRNA interaction analysis. In addition, text mining was conducted to obtain clues and ideas for further experimental research regarding the mechanisms underlying drug resistance in ovarian cancer.

Materials and methods

Acquisition and analysis of datasets

Microarray data published prior to December 2014 detailing ovarian cancer chemoresistance-related microRNA and mRNA expression profiles were retrieved and downloaded from the National Center for Biotechnology Information (NCBI) GEO database (http://www.ncbi.nlm Nih.gov/geo). Queries were performed using “ovarian cancer” as a keyword. The search was restricted to the following specific fields: study type, expression profiling by array, and species: Homo sapiens. We downloaded microRNA expression microarray datasets GSE54665 and mRNA expression microarray datasets GSE33482, GSE28646 [5] and GSE15372 [6].

Inclusion criteria for differentially expressed microRNAs and genes

Differentially expressed microRNAs and genes were screened using the GEO2R tool in the GEO database. GEO2R is an R programming language-based dataset analysis tool that was launched in 2012. This tool is based on a t-test or analysis of variance (ANOVA) and is useful for comparing two groups of samples under the same experimental conditions to identify differentially expressed genes or microRNAs [7].

In the present study, microRNAs and genes that were differentially expressed between drug-resistant and drug-sensitive ovarian cancer cells were screened using an adjusted p-value (adj. P) of less than 0.05 and a fold change of at least 2 (> 2-fold change) as thresholds. MicroRNAs or genes that were common to at least two expression profile datasets were selected using the Bioinformatics & Research Computing website (http://jura.wi.mit.edu/bioc/tools/compare.php). Bioinformatic analysis of differentially expressed genes and microRNAs. Pathway enrichment analysis of differentially expressed microRNAs was performed using the DIANA miRPath (http://diana.imis.athenainnovation.gr/DianaTools/index.php? r=mirpath/index) [8]. DAVID (http://david.abcc.ncifcrf.gov/) was used to analyze the pathway enrichment of differentially expressed genes [9]. Biological process annotation of differentially expressed genes was performed using ToppGene (http://toppgene.cchmc.org/) [10]. Protein/gene interactions were analyzed using GeneMANIA (http://www.genem ania.org/) [11] and mRNA-microRNA interactions were analyzed using the miRanda (http://www.microrna. org/microrna/home.do) [12], miRBase (http://www.mirbase.org/) [13], TargetScan (http://www. targetscan.org/) [14], miRWalk (http://www.umm.uni-heidelberg.de/ apps/zmf/mirwalk) [15] and RNA22 (http://cbcsrv.watson.ibm.com/ma 22 .html) [16] tools.

Cell culture

The human ovarian cancer cell lines SKOV3 and A2780 were generated in our lab, and A2780/DDP was established from A2780 as previously described A stable cisplatin-resistant cell line, SKOV3/DDP, was established from SKOV3 by continuous exposure of the cells to increasing concentrations of DDP [17], and routinely maintained in 1640 (CORNING, USA) supplemented with 10% fetal bovine serum (FBS) (CORNING, USA), 2 μmol/L L-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco BRL, Grand Island, NY) at 37°C in a humidified atmosphere containing 5% CO2.

Real-time quantitative PCR

MicroRNA

RNA was extracted from A2780, A2780/DDP, SKOV3, and SKOV3/DDP cells using an RNeasy® Mini Kit (QIAGEN, Germany) according to the manufacturer’s instructions. cDNA was then obtained using a miScript II RT kit (QIAGEN, Germany) in accordance with the manufacturer’s instructions. Real-time quantitative PCR (qRT-PCR) was completed with FastStart Universal SYBR Green PCR Master (ROX) in a total volume of 10 µL on an ABI 7500 (Applied Biosystems, USA). The conditions were as follows: 40 cycles of three-step PCR (95°C for 15 sec, 55°C for 30 sec, and 72°C for 30 sec) following an initial denaturation at 95°C for 10 min. The microRNA sequences were analyzed using miRBase (http://microrna.sanger.ac.uk/sequences/). Details regarding the primers used to measure gene expression are listed in Supplementary Table 1. All primers were supplied by BGI. The 2-ΔΔ CT method was used for data analysis [18].

mRNA

Total RNA was isolated as described above. First-strand cDNA was synthesized from 1 µg of total RNA using the Transcriptor First Strand cDNA Synthesis kit (Roche, Germany) as instructed by the manufacturer. qRT-PCR reactions were performed on an ABI 7500 real-time PCR system (Applied Biosystems, USA) using the following procedure: 95°C for 10 min, followed by 40 cycles of 95°C for 30 sec and 60°C for 1 min. To create a qRT-PCR standard, glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control. The sequences of the PCR primers are listed in Supplementary Table 2. The 2-ΔΔ CT method was used for data analysis [18].

Results

Microarray datasets that met the inclusion criteria

In the present study, the microRNA expression profiling datasets were GSE54665, which was generated by the Erasmus University Medical Center using the GPL16851 platform. The samples used to generate the GSE54665 dataset were cisplatin-resistant A2780 ovarian cancer cells and parental A2780 cells (three sets each). The mRNA expression profiling datasets studied were GSE33482, GSE28646, and GSE15372. GSE33482 was generated by Queen’s University using the GPL640 microarray platform. The data were derived from cisplatin-resistant A2780 ovarian cancer cells (A2780-cis) and parental A278 cells (six repeats of each). GSE28646 was generated by the Imperial College of London using the GPL570 microarray platform. The data were derived from cisplatin-resistant A2780 ovarian cancer cells (A2780-CP70) and parental A2780 cells (three repeats of each). GSE15372 originated from the Indiana University School of Medicine using the GPL570 microarray platform. The data were derived from cisplatin-resistant A2780 ovarian cancer cells (Round5-A2780) and parental A2780 cells (five repeats of each) (Table 1).

Table 1.

microRNA and gene expression microarray datasets related to the chemotherapy response in epithelial ovarian cancer

Accession number of the dataset Organization name Platforms Design Response to chemotherapy
Resistance Sensitive
microRNA GSE54665 Erasmus University Medical Center GPL16851 A2780/DDP AND A27800 3 3
mRNA GSE33482 Queen’s University GPL6480 A2780/cis AND A2780 6 6
GSE28646 Imperial College London GPL570 A2780/CP70 AND A2780 3 3
GSE15372 Indiana University GPL570 A2780-DDP AND A2780 5 5
Open in a new tab

Differentially expressed microRNAs and pathway enrichment

The drug-resistant A2780 ovarian cancer cells and the parental A2780 cells used to generate the GSE54665 expression profiling datasets served as the experimental group and the control group, respectively, and these groups were screened for differentially expressed microRNAs using the GEO2R tool. As shown in Table 2, after rigorous screening using adj. P < 0.05 and fold > 2 as thresholds, a total of nine differentially expressed microRNAs were identified from the GSE54665 microarray dataset. One microRNA (microRNA-141) was significantly upregulated in drug-resistant cells, whereas the expression levels of the remaining eight microRNAs were decreased. We evaluated the nine differentially expressed microRNAs through qRT-PCR and found that the expression levels of microRNA199a-5p, microRNA199a-3p, microRNA199b-3p, microRNA-645, microRNA-335, microRNA-18b, and microRNA-141 in SKOV3/DDP and A2780/DDP cells were in agreement with the results from the microRNA chip (Figure 1A, 1B). Among these differentially expressed microRNAs, microRNA-335 [19] and microRNA-141 [20] have been proven to be associated with the regulation of drug resistance in ovarian cancer. Pathway enrichment analysis was performed on these nine differentially expressed microRNAs using DIANA miRPath [8]. A total of 67 signaling pathways were identified (P < 0.05), and the 10 most significantly enriched pathways were selected according to their P values (Table 3).

Table 2.

Screening for differentially expressed microRNAs and their taget gens in the GSE54665 dataset

Accession number of the dataset differentially expressed microRNAs LogFC* Target genes
GSE54665 microRNA-199a-5p 5.244 C1orf21, DCLK1, EPHA7, IL20RB, MAB21L1, PI15
microRNA-199a-3p 4.103 C1orf21, DCLK1, IL20RB
microRNA-199b-3p 4.103 C1orf21, DCLK1, IL20RB
microRNA-335 3.319 C1orf21, CCR5, TPM1
microRNA-141 -3.083 C1orf21, DACT1, DCLK1, EPHA3, IL20RB, FOXP2, PI15
microRNA-215 3.02 C1orf21, CD55, CNN1, GSG1 TSLP
microRNA-363 2.389 C1orf21, DCLK1
microRNA-18b 2.069 EPHA7, PI15
microRNA-645 1.729 C1orf21, FILIP1
Open in a new tab
*

A positive logFC value indicates that the miRNA expression level was downregulated in the drug-resistant group, whereas a negative logFC value indicates that the miRNA expression level was upregulated in the drug-resistant group, Note: Thresholds used in the GEO2R analysis: adj. P < 0.05, fold > 2.

Figure 1.

Figure 1

Open in a new tab

Relative expression levels of microRNAs and genes in SKOV3/DDP vs. SKOV3 cells and A2780/DDP vs. A2780 cells. A. Relative expression levels of microRNAs in SKOV3/DDP vs. SKOV3 cells. B. Relative expression levels of microRNAs in A2780/DDP vs. A2780 cells. C. Relative expression levels of genes in A2780/DDP vs. A2780 cells.

Table 3.

microRNAs and pathways related to drug resistance in ovarian cancer

Pathway Numbers of microRNAs P value Associations with drug resistance in ovarian cancer (Ref.)
microRNA199a-5p Focal adhesion 8 2.58E-13 YES [21]
microRNA199a-3p PI3K-Akt Signaling pathway 7 2.24E-12 YES [22]
microRNA199b-3p Ubiquitin mediated proteolysis 7 4.76E-12 YES [23]
microRNA215 ErbB signaling pathway 9 1.78E-09 YES [24]
microRNA363 Regulation of actin cytoskeleton 8 2.74E-09 -
microRNA335 Long-term potentiation 6 2.74E-09 -
microRNA18b Gap junction 8 2.74E-09 YES [25]
microRNA141 Ascorbate and aldarate metabolism 3 9.60E-09 -
microRNA-645 Calcium signaling pathway 8 9.60E-09 YES [26]
MAPK signaling pathway 9 4.30E-07 YES [27]
Open in a new tab

-indicates no reports or an inability to retrieve relevant reports.

Screening for differentially expressed genes in three sets of mRNA microarrays and analysis of the correlation between differentially expressed genes and drug resistance in ovarian cancer

In the present study, drug-resistant A2780 ovarian cancer cells and parental A2780 cells were used as the experimental group and the control group, respectively. The GSE33482, GSE28646, and GSE15372 datasets were screened using the GEO2R tool to identify genes that were differentially expressed between the experimental and control groups, and 1102, 761, and 927 differentially expressed genes were identified, respectively. A total of 38 genes were found to be significantly differentially expressed in all three microarray sets (Figure 2). (The list of the 38 differentially expressed genes is shown in Supplementary Table 3) Seven of the 38 genes, NHS-like 1 (NHSL1), ephrin type-A receptor 3 (EPHA3), ephrin type-A receptor 7 (EPHA7), peptidase inhibitor 15 (PI15), synuclein alpha (SNCA), ubiquitin-specific peptidase 51 (USP51), and zinc finger and SCAN domain-containing protein 4 (ZSCAN4), exhibited exactly the same expression trends in all three microarray sets (Table 4), and qRT-PCR analysis showed that the expression of NHSL1, EPHA3, USP51, ZSCAN4, EPHA7, and PI15 in A2780/DDP cells coincided with results from mRNA chips (these genes were not expressed in SKOV3/DDP or SKOV3 cells) (Figure 1C).

Figure 2.

Figure 2

Open in a new tab

Analysis of the differentially expressed genes in the three mRNA expression profiling datasets using the GEO2R tool. Note: Thresholds used in the GEO2R analysis: adj. P < 0.05, fold > 2.

Table 4.

The seven differentially expressed genes that exhibited exactly the same expression pattern trends in all three microarray datasets

Gene name LogFC* Literature related to drug resistance in tumors
GSE15372 GSE28646 GSE33482
NHSL1 -2.01 -2.79 -2.07 -
EPHA3 3.87 3.25 2.01 Related to drug resistance in melanoma [28]
EPHA7 2.84 4.72 7.77 Relate to chemosensitivity in colon cancer [29]
PI15 2.76 3.85 4.78 -
SNCA 3.2 5.51 10.57 -
USP51 2.89 2.81 2.165 -
ZSCAN4 2.03 2.14 5.02 -
Open in a new tab
*

A positive logFC value indicates that the gene is downregulated in the drug-resistant group, and a negative logFC value indicates that the gene is upregulated in the drug-resistant group.

To reveal the biological significance of the commonly differentially expressed genes in the regulation of ovarian cancer drug resistance at the unitary level, biological pathway enrichment and biological process annotation were performed on the above-described 38 genes using DAVID and ToppGene, respectively. DAVID-based signaling pathway enrichment analysis identified cytokine-cytokine receptor interaction pathways, whereas the ToppGene-based analysis identified a total of 32 biological processes that were significantly related to the 38 genes (P < 0.05). Among the 32 biological processes, cell adhesion and immunomodulatory responses were found to be related to the regulation of drug resistance in ovarian cancer (Figure 3A). To further demonstrate the relationship between the 38 commonly differentially expressed genes and the regulation of drug resistance in ovarian cancer, text mining was conducted using the COREMINE tool. Co-occurrence analysis of the literature was conducted using the gene names and “Ovarian Neoplasms”, “Drug Resistance”, and “Drug Resistance, Neoplasm” as keywords. The results revealed that 22 genes were present in the text-mining networks. Among these 22 genes, 13 genes were related to ovarian cancer; 11 co-occurred with tumor drug resistance and multidrug resistance in the literature; and nine (aldehyde dehydrogenase 1 family, member A1 (ALDH1A1), JUN, C-C chemokine receptor type 5 (CCR5), SNCA, lysyl oxidase (LOX), cluster of differentiation 55 (CD55), tropomyosin 1 (TPM1), filamin A-interacting protein 1 (FILIP1), and transforming growth factor beta 2 (TGFB2)) were significantly related to ovarian cancer, multidrug resistance, and tumor drug resistance. In addition to significant correlations with “Ovarian Neoplasms”, “Drug Resistance”, and “Drug Resistance, Neoplasm”, the genes were also significantly related to each other, with the relationship between ALDH1A1 and JUN presented as an example Figure 3B.

Figure 3.

Figure 3

Open in a new tab

Bioinformatic analysis of different expressed genes. A. Biological processes of different genes related to ovarian cancer. B. Diagram of the linear relationship between the differentially expressed genes retrieved using COREMINE and drug resistance and the linear relationships between differentially expressed genes (Note: DR: drug resistance; DR-N: drug resistance, Neo~; OV: ovarian carcinoma.) C. Diagram of EPHA7 protein/gene and PI15 protein/gene interactions. note: 1. response to lipid, 2. regulation of cell adhesion, 3. neuron development, 4. neruon differentiation, 5. enzyme linked receptor protein signaling, 6. positive regulation of neuron apoptotic process, 7. response to oxygen-containing compound, 8. generation of neurons, 9. neuron death, 10. dopamine biosynthetic process, 11. neuron projection development, 12. regulation of acyl-CoA biosynthetic process, 13. positive regulation of neuron death, 14. organ morphogenesis, 15. neurogenesis, 16. regulation of cafactor metabolic process, 17. regulation of coenzyme metabolic process, 18. regulation of acute inflammatory response, 19. cell projection organization, 20. regulation of swlfur metabolic process, 21. negative regulation of immune response, 22. eye development, 23. negative regulation of immune effector process, 24. microglial cell activation, 25. SMAD protein import into nucleus, 26. regulation of neuron apoptotic process, 27. catecholamine biosynthetic process, 28. catechol-containing compound biosynthetic process, 29. negative regulation of sequestering of calcium ion, 30. release of sequestered calcium ion into cytosol, 31. regulation of sequestering of calcium ion, 32. regulation of phosphorus metabolic process.

Analysis of the correlation between differentially expressed microRNAs and the differentially expressed genes associated with drug resistance in ovarian cancer

The correlation between the 9 dysregulated microRNAs (Table 2) and the 38 differentially expressed genes listed in Supplementary Table 3 was analyzed using five microRNA-mRNA interaction tools including miRanda, miRBase, TargetScan, miRWalk, and RNA22. Genes identified by at least three prediction tools were selected as candidate targets, and the intersections of the candidate target genes with the 38 common differentially expressed genes in the three microarray datasets were determined. As shown in Table 2, C1orf21 was predicted as the target of the nine microRNAs, PI15 and interleukin 20 receptor beta (IL20RB) were predicted as the target of five microRNAs including microRNA199a-5p, microRNA199a-3p, microRNA199b-3p, microRNA-141 and microRNA-363. In particular, EPHA7 and PI15 were identified as the potential target of microRNA-141, for which their expressions were negatively correlated in drug resistant ovarian cancer cells.

Analysis of interactions between either EPHA7 or PI15 and the protein-coding genes associated with drug resistance in ovarian cancer

The relationships between EPHA7 or PI15 and drug resistance-related oncogenes or tumor-suppressor genes in ovarian cancer were analyzed using the GeneMANIA tool (http://www.genemania.org/). Among the 15 tumor suppressor genes [3] and 25 oncogenes [30] that have been reported to be related to the regulation of drug resistance in ovarian cancer, 19 were found to directly interact with PI15 and EPHA7 through co-expression, genetic interaction, physical interaction, and joint participation in signaling pathways (Figure 3C). Among the 19 directly interacting genes, 10 were found to directly interact with PI15, whereas nine directly interacted with EPHA7. The oncogene CUB and zona pellucida-like domain-containing protein 1 (CUZD1) were found to directly interact with both EPHA7 and PI15. In addition, in protein networks, tumor protein p53 (TP53) [31], early growth response 1 (EGR1) [32], FOS [33], and other genes were associated with the regulation of drug resistance in ovarian cancer.

Discussion

Ovarian cancer is a malignancy that poses a serious threat to women’s health. Ovarian cancer has the highest degree of malignancy of all gynecological cancers [34]. Postoperative chemotherapy is the main approach employed for the treatment of ovarian cancer, and multidrug resistance is an important reason for the failure of chemotherapy [2]. Previous studies have shown that multidrug resistance in ovarian cancer is related to various factors, including differential expression of microRNAs, differential expression of genes, and microRNA-controlled differential gene expression [35,36]. Therefore, screening for and identifying microRNAs and genes that are differentially expressed between drug-resistant and drug-sensitive cells and investigating the correlations between differentially expressed microRNAs and genes may shed light on the molecular mechanisms underlying drug resistance in ovarian cancer and may provide guidance for clinicians when choosing chemotherapy regimens and predicting prognosis.

GEO2R [7] (http://www.ncbi.nlm.nih.gov/geo/geo2r/) is an R programming language-based analytical tool for studying differentially expressed genes. In the present study, microRNA expression microarray data for platinum-resistant and platinum-sensitive ovarian cancer cells (GSE54665) were analyzed, and nine differentially expressed microRNAs were identified. Eight of these microRNAs were downregulated, whereas one microRNA (microRNA-141) was upregulated (Table 2). We identified nine differentially expressed microRNAs through qRT-PCR and found that the results of microRNA199a-5p, microRNA199a-3p, microRNA199b-3p, microRNA-645, microRNA-335, microRNA-18b, and microRNA-141 expression analysis in SKOV3/DDP and A2780/DDP cells agreed with the results obtained from the microRNA chip (Figure 1A, 1B). Among the top 10 signaling pathways regulated by the nine identified microRNAs, at least seven were related to the regulation of drug resistance in ovarian cancer, including the focal adhesion pathway [21], the phosphatidylinositol 3-kinase (PI3K-Akt) signaling pathway [22], the erythroblastosis oncogene B (ErbB) signaling pathway [24], the mitogen-activated protein kinase (MAPK) signaling pathway [27]. These findings indicate that the nine microRNAs that are differentially expressed in drug-resistant ovarian cancer cells may be associated with the regulation of drug resistance in ovarian cancer. Previous studies have shown that microRNA199a-3p promotes the sensitivity of hepatoma cells to doxorubicin by targeting mammalian target of rapamycin (mToR) and c-Met [36], microRNA199a-5p is related to multidrug resistance in colon cancer [37] and cisplatin resistance in liver cancer [38]. MicroRNA-215 is related to chemoresistance and the loss of sensitivity to radiotherapy in colon cancer [39], and microRNA-141 participates in the regulation of drug resistance in ovarian cancer by regulating Kelch-like ECH-associated protein 1 (KEAP1) [20].

Three sets of mRNA expression profiles were analyzed, and a total of 38 genes were identified as differentially expressed in drug-resistant ovarian cancer cells. Enrichment analysis of biological processes and signaling pathways showed that the above-described 38 differentially expressed genes are significantly related to a series of biological processes such as cell adhesion, immunomodulation, and cytokine-cytokine receptor interactions. These processes have been proven to be closely related to the regulation of drug resistance in ovarian cancer. A previous study has demonstrated that cell adhesion is a critical step leading to the occurrence of peritoneal metastasis [40]. Peritoneal metastasis indicates that ovarian cancer is already at an advanced stage, which is frequently associated with a lack of sensitivity to chemotherapy. Cytokine-cytokine receptor interaction pathways are involved in cell growth, differentiation, and death as well as blood vessel formation, and further study had indicated that the involvement of cytokine-cytokine receptor interactions in drug resistance would proceed via tumor-stroma communication [41].

These results reveal that the 38 genes that are significantly differentially expressed in all three drug-resistant ovarian cancer cell-derived gene expression profile datasets may be involved in the regulation of drug resistance in ovarian cancer through participating in processes such as cell adhesion, immunomodulatory responses, and cytokine-cytokine receptor interactions.

Text mining was performed on the 38 genes using the COREMINE search and analysis tool. We discovered that ALDH1A1, JUN, CCR5, SNCA, LOX, CD55, TPM1, FILIP1, and TGFB2 are significantly correlated with ovarian cancer, drug resistance, and multidrug resistance in tumors. These results indicate that the above-mentioned genes may be involved in the regulation of drug resistance in ovarian cancer, which is consistent with the findings of previous studies. Four of the nine genes described above, ALDH1A1, JUN, LOX, and CD55, have been shown to be related to drug resistance in ovarian cancer [42-45].

Epigenetic regulation of microRNAs plays an important role in the regulation of drug resistance in ovarian cancer. Aberrant gene expression due to the action of microRNAs is an important mechanism underlying drug resistance in ovarian cancer. Therefore, in the present study, microRNA-mRNA interaction analysis was conducted using microRNA and mRNA expression profiles to obtain additional information related to the regulation of drug resistance in ovarian cancer. By analyzing interactions between differentially expressed microRNAs and differentially expressed genes in drug-resistant and drug-sensitive ovarian cancer cells, we discovered that nine microRNAs may have regulatory effects on nearly half of the 38 identified genes. These findings indicate that the differentially expressed microRNAs and genes described above may act in concert to regulate drug resistance in ovarian cancer. MicroRNA-target relationships existed between C1orf21 and all nine microRNAs as well as between PI15 or IL20RB and three of the nine microRNAs.

Among the 38 genes that were identified to be associated with drug resistance in ovarian cancer, EPHA7 and PI15 might significantly correlate with drug resistance in ovarian cancer. Firstly, only seven (NHSL1, EPHA3, EPHA7, PI15, SNCA, USP51, and ZSCAN4) of the 38 genes exhibited exactly the same expression trends in the three mRNA microarrays, from which the EPHA7 and PI15 were included. The expression of EPHA7 and PI15 was significantly downregulated in the three independent drug-resistant A2780 ovarian cancer cells (a fold change of at least 2.7 was noted in both genes), and our qRT-PCR data showed that the expression of six genes, including NHSL1, EPHA3, USP51, ZSCAN4, EPHA7, and PI15, in A2780/DDP cells was consistent with data from mRNA chips (these genes were not expressed in SKOV3/DDP or SKOV3 cells) (Figure 1C). Secondly, the negative regulatory relationships between microRNA-141 and the two genes were predicted. MicroRNA-141 was found to be highly expressed in drug-resistant A2780 ovarian cancer cells. MicroRNA-mRNA interaction analysis showed that EPHA7 and PI15 are potential target genes of microRNA-141. Additionally, a previous study showed that microRNA-141 is significantly related to the regulation of drug resistance in ovarian cancer [20]. Third, protein/gene interaction analysis showed that PI15 and EPHA7 directly interact with multiple genes related to drug resistance in ovarian cancer through genetic co-expression, genetic interaction, physical interaction, and joint participation in signaling pathways. In a protein/gene interaction network composed of 31 genes, PI15 directly interacts with a total of 11 genes. Among these 11 genes, five are oncogenes (inhibitor of nuclear factor kappa-B kinase subunit epsilon (IKBKE), ninein-like protein (NINL), neurogenic locus notch homolog protein 3 (NOTCH3), signal transducer and activator of transcription 3 (STAT3), and CUZD1), and six are tumor-suppressor genes (cyclin-dependent kinase inhibitor 2A (CDKN2A), F-box protein 32 (FBXO32), interleukin 24 (IL24), programmed cell death protein 4 (PDCD4), sulfatase 1 (SULF1), and WW domain-containing oxidoreductase (WWOX)). EPHA7 was found to directly interact with a total of nine genes, among which seven are oncogenes (B-cell lymphoma 2 (BCL2), CUZD1, JUN, MET, MYC, remodeling and spacing factor 1 (RSF1), and SRC), and two are tumor-suppressor genes (breast cancer 2 early onset (BRCA2), and checkpoint kinase 2 (CHEK2)). The oncogene CUZD1 was found to directly interact with both PI15 and EPHA7. The above-mentioned oncogenes and proto-oncogenes have been proven to be related to the regulation of multidrug resistance in ovarian cancer [3,30]. For example, the STAT3 gene is related to platinum and paclitaxel resistance in ovarian cancer [35].

These results indicate that PI15 and EPHA7, which are significantly correlated with the 22 drug resistance-related protein-coding genes in ovarian cancer, are also likely to be associated with the regulation of drug resistance in ovarian cancer. Fourth, through a query of signaling pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG), we found that EPHA7 is upstream of the MAPK pathway, which is a critical signaling pathway that regulates drug resistance in ovarian cancer. In summary, PI15 and EPHA7 are important candidate protein-coding genes that may be involved in the regulation of drug resistance in ovarian cancer. The exact roles of PI15 and EPHA7 will require further in-depth study.

Disclosure of conflict of interest

None.

Supporting Information

ijcep0008-6847-f4.pdf (173.8KB, pdf)

References

  • 1.Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11–30. doi: 10.3322/caac.21166. [DOI] [PubMed] [Google Scholar]
  • 2.Gou Q, Liu L, Wang C, Wu Q, Sun L, Yang X, Xie Y, Li P, Gong C. Polymeric nanoassemblies entrapping curcumin overcome multidrug resistance in ovarian cancer. Colloids Surf B Biointerfaces. 2015;126:26–34. doi: 10.1016/j.colsurfb.2014.12.012. [DOI] [PubMed] [Google Scholar]
  • 3.Yin F, Liu X, Li D, Wang Q, Zhang W, Li L. Tumor suppressor genes associated with drug resistance in ovarian cancer (review) Oncol Rep. 2013;30:3–10. doi: 10.3892/or.2013.2446. [DOI] [PubMed] [Google Scholar]
  • 4.Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics. 2007;23:1846–1847. doi: 10.1093/bioinformatics/btm254. [DOI] [PubMed] [Google Scholar]
  • 5.Zeller C, Dai W, Steele NL, Siddiq A, Walley AJ, Wilhelm-Benartzi CS, Rizzo S, van der Zee A, Plumb JA, Brown R. Candidate DNA methylation drivers of acquired cisplatin resistance in ovarian cancer identified by methylome and expression profiling. Oncogene. 2012;31:4567–4576. doi: 10.1038/onc.2011.611. [DOI] [PubMed] [Google Scholar]
  • 6.Li M, Balch C, Montgomery JS, Jeong M, Chung JH, Yan P, Huang TH, Kim S, Nephew KP. Integrated analysis of DNA methylation and gene expression reveals specific signaling pathways associated with platinum resistance in ovarian cancer. BMC Med Genomics. 2009;2:34. doi: 10.1186/1755-8794-2-34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM, Holko M, Yefanov A, Lee H, Zhang N, Robertson CL, Serova N, Davis S, Soboleva A. NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013;41:D991–995. doi: 10.1093/nar/gks1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Vlachos IS, Kostoulas N, Vergoulis T, Georgakilas G, Reczko M, Maragkakis M, Paraskevopoulou MD, Prionidis K, Dalamagas T, Hatzigeorgiou AG. DIANA miRPath v. 2.0: investigating the combinatorial effect of microRNAs in pathways. Nucleic Acids Res. 2012;40:W498–504. doi: 10.1093/nar/gks494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4:44–57. doi: 10.1038/nprot.2008.211. [DOI] [PubMed] [Google Scholar]
  • 10.Chen J, Bardes EE, Aronow BJ, Jegga AG. ToppGene Suite for gene list enrichment analysis and candidate gene prioritization. Nucleic Acids Res. 2009;37:W305–311. doi: 10.1093/nar/gkp427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Mostafavi S, Ray D, Warde-Farley D, Grouios C, Morris Q. GeneMANIA: a real-time multiple association network integration algorithm for predicting gene function. Genome Biol. 2008;9(Suppl 1):S4. doi: 10.1186/gb-2008-9-s1-s4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.John B, Enright AJ, Aravin A, Tuschl T, Sander C, Marks DS. Human MicroRNA targets. PLoS Biol. 2004;2:e363. doi: 10.1371/journal.pbio.0020363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ. miRBase: tools for microRNA genomics. Nucleic Acids Res. 2008;36:D154–158. doi: 10.1093/nar/gkm952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell. 2005;120:15–20. doi: 10.1016/j.cell.2004.12.035. [DOI] [PubMed] [Google Scholar]
  • 15.Dweep H, Gretz N, Sticht C. miRWalk database for miRNA-target interactions. Methods Mol Biol. 2014;1182:289–305. doi: 10.1007/978-1-4939-1062-5_25. [DOI] [PubMed] [Google Scholar]
  • 16.Loher P, Rigoutsos I. Interactive exploration of RNA22 microRNA target predictions. Bioinformatics. 2012;28:3322–3323. doi: 10.1093/bioinformatics/bts615. [DOI] [PubMed] [Google Scholar]
  • 17.Li L, Luan Y, Wang G, Tang B, Li D, Zhang W, Li X, Zhao J, Ding H, Reed E, Li QQ. Development and characterization of five cell models for chemoresistance studies of human ovarian carcinoma. Int J Mol Med. 2004;14:257–264. [PubMed] [Google Scholar]
  • 18.Rao X, Huang X, Zhou Z, Lin X. An improvement of the 2ˆ(-delta delta CT) method for quantitative real-time polymerase chain reaction data analysis. Biostat Bioinforma Biomath. 2013;3:71–85. [PMC free article] [PubMed] [Google Scholar]
  • 19.Sorrentino A, Liu CG, Addario A, Peschle C, Scambia G, Ferlini C. Role of microRNAs in drug-resistant ovarian cancer cells. Gynecol Oncol. 2008;111:478–486. doi: 10.1016/j.ygyno.2008.08.017. [DOI] [PubMed] [Google Scholar]
  • 20.van Jaarsveld MT, Helleman J, Boersma AW, van Kuijk PF, van Ijcken WF, Despierre E, Vergote I, Mathijssen RH, Berns EM, Verweij J, Pothof J, Wiemer EA. miR-141 regulates KEAP1 and modulates cisplatin sensitivity in ovarian cancer cells. Oncogene. 2013;32:4284–4293. doi: 10.1038/onc.2012.433. [DOI] [PubMed] [Google Scholar]
  • 21.Li J, Zhang Y, Gao Y, Cui Y, Liu H, Li M, Tian Y. Downregulation of HNF1 homeobox B is associated with drug resistance in ovarian cancer. Oncol Rep. 2014;32:979–988. doi: 10.3892/or.2014.3297. [DOI] [PubMed] [Google Scholar]
  • 22.Cheaib B, Auguste A, Leary A. The PI3K/Akt/mTOR pathway in ovarian cancer: therapeutic opportunities and challenges. Chin J Cancer. 2015;34:4–16. doi: 10.5732/cjc.014.10289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Su JL, Cheng X, Yamaguchi H, Chang YW, Hou CF, Lee DF, Ko HW, Hua KT, Wang YN, Hsiao M, Chen PB, Hsu JM, Bast RC Jr, Hortobagyi GN, Hung MC. FOXO3a-Dependent Mechanism of E1A-Induced Chemosensitization. Cancer Res. 2011;71:6878–6887. doi: 10.1158/0008-5472.CAN-11-0295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Shen G, Huang H, Zhang A, Zhao T, Hu S, Cheng L, Liu J, Xiao W, Ling B, Wu Q, Song L, Wei W. In vivo activity of novel anti-ErbB2 antibody chA21 alone and with Paclitaxel or Trastuzumab in breast and ovarian cancer xenograft models. Cancer Immunol Immunother. 2011;60:339–348. doi: 10.1007/s00262-010-0937-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Barhoumi R, Bailey HR, Hutchinson RW, Bowen JA, Burghardt RC. Enhancement of melphalan toxicity by octanol in ovarian adenocarcinoma cell lines: effects of altered cell-cell communication, glutathione levels, and plasma membrane fluidity. Fundam Appl Toxicol. 1995;25:70–79. doi: 10.1006/faat.1995.1041. [DOI] [PubMed] [Google Scholar]
  • 26.Al-Bahlani S, Fraser M, Wong AY, Sayan BS, Bergeron R, Melino G, Tsang BK. P73 regulates cisplatin-induced apoptosis in ovarian cancer cells via a calcium/calpain-dependent mechanism. Oncogene. 2011;30:4219–4230. doi: 10.1038/onc.2011.134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Campbell RM, Anderson BD, Brooks NA, Brooks HB, Chan EM, De Dios A, Gilmour R, Graff JR, Jambrina E, Mader M, McCann D, Na S, Parsons SH, Pratt SE, Shih C, Stancato LF, Starling JJ, Tate C, Velasco JA, Wang Y, Ye XS. Characterization of LY2228820 dimesylate, a potent and selective inhibitor of p38 MAPK with antitumor activity. Mol Cancer Ther. 2014;13:364–374. doi: 10.1158/1535-7163.MCT-13-0513. [DOI] [PubMed] [Google Scholar]
  • 28.Ding L, Kim M, Kanchi KL, Dees ND, Lu C, Griffith M, Fenstermacher D, Sung H, Miller CA, Goetz B, Wendl MC, Griffith O, Cornelius LA, Linette GP, McMichael JF, Sondak VK, Fields RC, Ley TJ, Mule JJ, Wilson PK, Weber JS. Clonal architectures and driver mutations in metastatic melanomas. PLoS One. 2014;9:e111153. doi: 10.1371/journal.pone.0111153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Kim JC, Kim SY, Cho DH, Roh SA, Choi EY, Jo YK, Jung SH, Na YS, Kim TW, Kim YS. Genome-wide identification of chemosensitive single nucleotide polymorphism markers in colorectal cancers. Cancer Sci. 2010;101:1007–1013. doi: 10.1111/j.1349-7006.2009.01461.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.X Liu, Y Gao, Y Lu, J Zhang, L Li, F Yin. Oncogenes associated with drug resistance in ovarian cancer. J Cancer Res Clin Oncol. 2015;141:381–395. doi: 10.1007/s00432-014-1765-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Brachova P, Mueting SR, Carlson MJ, Goodheart MJ, Button AM, Mott SL, Dai D, Thiel KW, Devor EJ, Leslie KK. TP53 oncomorphic mutations predict resistance to platinum and taxanebased standard chemotherapy in patients diagnosed with advanced serous ovarian carcinoma. Int J Oncol. 2015;46:607–618. doi: 10.3892/ijo.2014.2747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lu M, Xiao L, Hu J, Deng S, Xu Y. Targeting of p38 mitogen-activated protein kinases to early growth response gene 1 (EGR-1) in the human paclitaxel-resistance ovarian carcinoma cells. J Huazhong Univ Sci Technolog Med Sci. 2008;28:451–455. doi: 10.1007/s11596-008-0417-x. [DOI] [PubMed] [Google Scholar]
  • 33.Moorehead RA, Singh G. Influence of the proto-oncogene c-fos on cisplatin sensitivity. Biochem Pharmacol. 2000;59:337–345. doi: 10.1016/s0006-2952(99)00333-0. [DOI] [PubMed] [Google Scholar]
  • 34.Sharifian A, Pourhoseingholi MA, Norouzinia M, Vahedi M. Ovarian cancer in Iranian women, a trend analysis of mortality and incidence. Asian Pac J Cancer Prev. 2014;15:10787–10790. doi: 10.7314/apjcp.2014.15.24.10787. [DOI] [PubMed] [Google Scholar]
  • 35.Sheng WJ, Jiang H, Wu DL, Zheng JH. Early responses of the STAT3 pathway to platinum drugs are associated with cisplatin resistance in epithelial ovarian cancer. Braz J Med Biol Res. 2013;46:650–658. doi: 10.1590/1414-431X20133003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Fornari F, Milazzo M, Chieco P, Negrini M, Calin GA, Grazi GL, Pollutri D, Croce CM, Bolondi L, Gramantieri L. MiR-199a-3p regulates mTOR and c-Met to influence the doxorubicin sensitivity of human hepatocarcinoma cells. Cancer Res. 2010;70:5184–5193. doi: 10.1158/0008-5472.CAN-10-0145. [DOI] [PubMed] [Google Scholar]
  • 37.Kong Y, Bai PS, Sun H, Nan KJ, Chen NZ, Qi XG. The deoxycholic acid targets miRNA-dependent CAC1 gene expression in multidrug resistance of human colorectal cancer. Int J Biochem Cell Biol. 2012;44:2321–2332. doi: 10.1016/j.biocel.2012.08.006. [DOI] [PubMed] [Google Scholar]
  • 38.Xu N, Zhang J, Shen C, Luo Y, Xia L, Xue F, Xia Q. Cisplatin-induced downregulation of miR-199a-5p increases drug resistance by activating autophagy in HCC cell. Biochem Biophys Res Commun. 2012;423:826–831. doi: 10.1016/j.bbrc.2012.06.048. [DOI] [PubMed] [Google Scholar]
  • 39.Svoboda M, Sana J, Fabian P, Kocakova I, Gombosova J, Nekvindova J, Radova L, Vyzula R, Slaby O. MicroRNA expression profile associated with response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer patients. Radiat Oncol. 2012;7:195. doi: 10.1186/1748-717X-7-195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Lis R, Touboul C, Halabi NM, Madduri AS, Querleu D, Mezey J, Malek JA, Suhre K, Rafii A. Mesenchymal cell interaction with ovarian cancer cells induces a background dependent pro-metastatic transcriptomic profile. J Transl Med. 2014;12:59. doi: 10.1186/1479-5876-12-59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Morales M, Planet E, Arnal-Estape A, Pavlovic M, Tarragona M, Gomis RR. Tumor-stroma interactions a trademark for metastasis. Breast. 2011;20(Suppl 3):S50–55. doi: 10.1016/S0960-9776(11)70294-6. [DOI] [PubMed] [Google Scholar]
  • 42.Meng E, Mitra A, Tripathi K, Finan MA, Scalici J, McClellan S, Madeira da Silva L, Reed E, Shevde LA, Palle K, Rocconi RP. ALDH1A1 maintains ovarian cancer stem cell-like properties by altered regulation of cell cycle checkpoint and DNA repair network signaling. PLoS One. 2014;9:e107142. doi: 10.1371/journal.pone.0107142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Amable L, Gavin E, Kudo K, Meng E, Rocconi RP, Shevde LA, Reed E. GLI1 upregulates C-JUN through a specific 130-kDa isoform. Int J Oncol. 2014;44:655–661. doi: 10.3892/ijo.2013.2222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Januchowski R, Zawierucha P, Rucinski M, Nowicki M, Zabe Ml. Extracellular matrix proteins expression profiling in chemoresistant variants of the A2780 ovarian cancer cell line. Biomed Res Int. 2014;2014:365867. doi: 10.1155/2014/365867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Cocco E, Varughese J, Buza N, Bellone S, Lin KY, Bellone M, Todeschini P, Silasi DA, Azodi M, Schwartz PE, Rutherford TJ, Carrara L, Tassi R, Pecorelli S, Lockwood CJ, Santin AD. Tissue factor expression in ovarian cancer: implications for immunotherapy with hI-con1, a factor VII-IgGF(c) chimeric protein targeting tissue factor. Clin Exp Metastasis. 2011;28:689–700. doi: 10.1007/s10585-011-9401-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ijcep0008-6847-f4.pdf (173.8KB, pdf)

Articles from International Journal of Clinical and Experimental Pathology are provided here courtesy of e-Century Publishing Corporation

RESOURCES