ANTICANCER RESEARCH 36: 1683-1692 (2016)

The Effects of Estradiol and Glucocorticoid on Human Osteosarcoma Cells: Similarities and Differences

QING ZHOU1*, LIBING SHEN2*, CHAOQUN LIU1, CHUNFENG LIU1, HAO CHEN1 and JINLIAN LIU1

1Department of Orthopaedics, Suzhou Kowloon Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Jiangsu, P.R. China; 2State Key Laboratory of Genetic Engineering and MOE Key Laboratory of Contemporary Anthropology, School of Life Sciences, Fudan University, Shanghai, P.R. China

Abstract. Background: Human U2OS osteosarcoma cells on expression levels of cellular markers for proliferation and were first derived from a tibial osteosarcoma of a teenage apoptosis support this argument. Conclusion: Our study girl. They are a good cell model for osteosarcoma research suggests that glucocorticoid is more efficient in controlling in vitro. We compared the expression profiles of osteosarcoma cell proliferation and apoptosis than estradiol. osteosarcoma cells after they were treated with estradiol and glucocorticoid, respectively. Materials and Methods: We used Osteosarcoma is the most common form of malignant bone published microarray data to compare the expression profiles tumor (1). It develops from mesenchymal cells and thus is a of human osteosarcoma cells treated with estradiol and type of bone sarcoma (2). Due to the high rate of bone glucocorticoid. We investigated their differences and growth in children and teenagers, osteosarcoma is normally similarities with various bioinformatics tools. Oncogenes and found in pediatric patients and is the eighth most common tumor-suppressor differentially expressed after the pediatric cancer in the United States and the fourth most hormone treatments were identified using online databases. common type of pediatric cancer in China between 2000 and The expression levels of cellular markers for proliferation 2010 (1, 3, 4). Surgical resection and chemotherapy are both and apoptosis were also compared before and after treatment successfully used in the treatment of osteosarcoma. However, with estradiol or glucocorticoid. Results: Genes in human the long-term survival rate for patients with osteosarcoma is osteosarcoma cells differentially expressed after treatment not very optimistic. The average 5-year survival rate of with estradiol or glucocorticoid were detected. Our analysis patients diagnosed with osteosarcoma is about 60% (2, 5). showed that their similarity is more functionally prominent Hormone therapy is an important auxiliary therapy in than their difference. Both estradiol and glucocorticoid can cancer treatment (6). Exogenously administered specific inhibit purine metabolic and biosynthetic pathway in human hormones, and hormone antagonists can influence the - osteosarcoma cells. We also identified oncogenes and tumor- expression pattern of certain cancer cells and may further suppressor genes among the differentially expressed genes. stop cancer cells from growing or even induce apoptosis of The functional enrichment analyses of the identified cancer- cancer cells. It is reported that certain female patients with associated genes suggests that estradiol has antagonistic cancer with a high level of estrogen -alpha usually effects on regulation of cell proliferation, while have a favorable prognosis (7), which implies the therapeutic glucocorticoid can both arrest the cell cycle and prompt value of estrogen for certain patients. Transdermal estradiol apoptosis. The effect of estradiol or glucocorticoid treatment therapy has been used to treat the men with advance prostate cancer (8). Another steroid hormone, glucocorticoid, has been used as a curative agent for lymphoid cancer treatment for decades (9). Glucocorticoid can induce massive apoptosis *These Authors contributed equally to this work. of certain human lymphoid cell lines (10). Glucocorticoid is also employed in treating prostate cancer and other solid Correspondence to: Jinlian Liu, 1Department of Orthopaedics, tumors, although glucocorticoid therapy is usually associated Suzhou Kowloon Hospital Affiliated to Shanghai Jiaotong University with concomitant negative effects such as reducing the School of Medicine, 118 Wansheng Street, Suzhou Industrial Park, Jiangsu 215021, P.R. China. E-mail: [email protected] immune responses of patients (11). These studies and practices suggest that besides surgical resection and Key Words: Osteosarcoma cell, estradiol, glucocorticoid, oncogene, chemotherapy, estrogen- or glucocorticoid-based therapy tumor suppressor gene, cellular marker for proliferation/apoptosis. might have some value in osteosarcoma treatment.

0250-7005/2016 $2.00+.40 1683 ANTICANCER RESEARCH 36: 1683-1692 (2016)

Many studies have been dedicated to identifying various Functional enrichment analysis for differentially expressed genes. hormone target genes in osteosarcoma cells (12-17). Some The differentially expressed genes were categorized into four of them were based on microarray technology and produced groups: estradiol down-regulated genes, estradiol up-regulated genes, glucocorticoid down-regulated genes and glucocorticoid up- the information resources for expression profiles of regulated genes. We used the DAVID online resource to perform the osteosarcoma cells after hormone treatment. However, due gene– (GO) term enrichment analysis for these four to the daunting amount of information or researchers’ limited gene groups (19, 20). interest, these microarray data resources are not fully utilized. The expression profiles of osteosarcoma cells after Functional module network construction. The differentially hormone treatment are especially worthy of further expressed genes were also used for building a functional module investigation because they may contain the expression network. BiNGO was used to cluster and map these genes into functional modules and build the biological network based on these information of certain oncogenes or tumor-suppressor genes modules (21). MCODE was used to extract the highly which can serve as target genes in future osteosarcoma interconnected subnetworks (22). treatment. In this study, we used published microarray data to Searching for oncogenes and tumor-suppressor genes in the group compare the expression profiles of U2OS osteosarcoma of differentially expressed genes. Online databases were used to cells after treatment with 17-β-estradiol and glucocorticoid identify oncogenes and tumor-suppressor genes in the differentially (16, 18). First, we detected the differentially expressed expressed genes. Using the differentially expressed genes as queries, we searched the Network of Cancer Genes database (NCG4.0, genes in U2OS osteosarcoma cells after treatment with 17- http://ncg.kcl.ac.uk/) for oncogenes and Tumor Suppressor Gene β-estradiol or glucocorticoid and then compared the database (TSGene, http://bioinfo.mc.vanderbilt.edu/TSGene/) for differences and similarities between four differentially tumor-suppressor genes (23, 24). expressed gene groups through functional enrichment analysis. Secondly, we identified oncogenes and tumor- Cellular markers of proliferation and apoptosis. Cellular markers suppressor genes among the differential expressed genes of proliferation and apoptosis were identified through literature- and focused on their possible biological roles in based searches. The cellular markers used in this study were antigen Ki-67 (MKI67), MYB-related B (MYBL2), apoptosis- oncogenesis. inducing factor 1 (AIFM1), tumor necrosis factor (TNFA) and the caspase gene family. The expression values of the identified cellular Materials and Methods markers were extracted from the normalized microarray data. We used the paired t-test to compare the expression values of cellular Microarray data. The microarray data used in this study were markers before and after treatment with estradiol or glucocorticoid retrieved from GEO DataSets hosted by the http:// and a p-value smaller than 0.05 was viewed as statistically www.ncbi.nlm.nih.gov/gds, which is a well-known repository of significant in comparing the expression values of these markers. gene expression data. The microarray dataset with the reference series number GSE2292 contains the gene-expression information Results for U2OS osteosarcoma cells before and after 17-β-estradiol treatment (16). The microarray dataset with the reference series number GSE6711 contains the gene-expression information of Detection of genes differentially expressed in U2OS U2OS osteosarcoma cells before and after 24-hour glucocorticoid osteosarcoma cells after 17-β-estradiol or glucocorticoid treatment (18). The dataset GSE2292 was collected using treatment. We used the paired t-test to compare the normalized Affymetrix Human HG-Focus Target Array (GPL201) and the gene counts within the two microarray datasets. A total of dataset GSE6711 was collected using Agilent-012391 Whole 1,359 genes were found to be differentially expressed in dataset Oligo Microarray G4112A (GPL708) (16, 18). GSE2292 after 17-β-estradiol treatment (p<0.01): 697 of them Both microarray datasets are not normalized. were down-regulated and 662 of them were up-regulated by 17-β-estradiol. There were 10,215 genes found differentially Statistical methods for detecting differentially expressed genes. In expressed in dataset GSE6711 after glucocorticoid treatment order to detect genes differentially expressed after hormone treatment, we first needed to normalize gene-expression counts for (p<0.01): 5,582 of them were down-regulated and 4,633 of each microarray dataset. By doing so, we made different samples them were up-regulated by glucocorticoid. comparable within each dataset. We used the mean-value normalization method to adjust the counts in each microarray DAVID analysis and functional module network analysis of sample because we only needed to compare the expression value differentially expressed genes. In order to understand the within each dataset. The paired t-test was employed to compare the possible functions of four gene groups above, we used the mean expression for each gene between hormone untreated and DAVID online resource to perform the gene-GO term treated osteosarcoma cell samples. R Package version 3.0.2 (https://cran.r-project.org/bin/windows/base/) was used to perform enrichment analysis for differentially expressed genes. The statistical analysis and a p-value smaller than 0.01 was viewed as DAVID functional enrichment analysis results for four statistically significant in detecting differentially expressed genes. differentially expressed gene groups are shown in Figure 1.

1684 Zhou et al: Differentially Expressed Genes in U2OS Cells After Estradiol and Glycorticoid Treatment

The biological processes affected by genes down-regulated Table I. The number of oncogenes and tumor-suppressor genes (TSGs) by 17-β-estradiol mainly involve the regulation of cell in osteosarcoma cells regulated by 17-β-estradiol and glucocorticoid. proliferation and the affected cellular components are mainly 17-β-Estradiol Glucocorticoid the cytosol, while those affected by up-regulated genes mainly involve the response to organic substances and the Down-regulated oncogenes 68 360 insoluble cell fraction (Figure 1A and B). The biological Up-regulated oncogenes 88 287 processes affected by genes down-regulated by Down-regulated TSGs 47 111 glucocorticoid similarly mainly involve the cell cycle and the Up-regulated TSGs 26 102 affected cellular components are mainly non-membrane- bound organelles, while those affected by up-regulated genes mainly involve the regulation of transcription (Figure 1 C and D). Phosphoproteins were present in all four groups of differentially expressed genes. regulation of cell proliferation and the regulation of cell death, Genes down-regulated by 17-β-estradiol genes and by and the Kyoto Encyclopedia of Genes and Genomes pathway glucocorticoid share a similarity in their biological process of this subgroup shows that signaling pathway is down- (cell proliferation and cell cycle) (Figure 1A and C). We regulated by 17-β-estradiol (Table III). For glucocorticoid- extracted the genes common to these two groups. A total of related subgroups, the biological processes for down-regulated 356 genes were found to be down-regulated by both 17-β- oncogenes mainly involve the cell cycle and mitosis, and their estradiol and glucocorticoid in osteosarcoma cells. We molecular function includes binding and constructed the functional module network for these transmembrane receptor protein tyrosine kinase activity (Table overlapping genes. Due to the over-complexity of the IV); the biological processes of up-regulated tumor-suppressor network, we reduced it to several highly interconnected genes involve regulation of the cell cycle, regulation of subnetworks. The most interconnected and meaningful apoptosis and cell death, and their molecular function is subnetwork mainly involves the biological pathway of primarily transcription regulatory (Table V). nucleotide (purine) metabolic and biosynthetic processes (Figure 2). Expression of cellular markers for proliferation and We similarly also extracted the genes common to those apoptosis before and after the treatment of 17-β-estradiol or up-regulated by 17-β-estradiol and by glucocorticoid. Their glucocorticoid. Antigen KI-67 and MYBL2 are cellular functional module network analysis produced no statistical markers for proliferation (25). KI-67 protein is strictly meaningful results and thus is not shown. associated with ribosomal RNA transcription and thus its gene is a standard marker for cellular proliferation (26). Identification of oncogenes and tumor-suppressor genes in MYBL2 is a member of the MYB family of transcription the four groups of differentially expressed genes. Using the factor genes which are involved in cell-cycle progression differentially expressed genes as queries, we searched (27). Our results show that expression of both markers was NCG4.0 for oncogenes and TSGene for tumor-suppressor significantly down-regulated after the treatment with genes in the four groups of differentially expressed genes. glucocorticoid, but they showed no sign of down-regulation The search results show that some genes were identified as after the treatment with 17-β-estradiol (Figure 3A and B). both oncogenes and tumor-suppressor genes by the two AIFM1, TNFA and the caspase gene family are cellular databases. We removed these genes from our results in order markers for apoptosis. AIFM1 encodes a flavoprotein to make our work more clear and explicable. essential for nuclear disassembly in apoptotic cells (28). The number of identified oncogenes and tumor suppressor TNFA is a cell-signaling protein (cytokine) responsible for genes in the four groups is shown in Table I. These systemic inflammation and acute-phase reaction (29). AIFM1 oncogenes and tumor-suppressor genes are naturally was significantly up-regulated in glucocorticoid-treated separated into eight subgroups. osteosarcoma cells, but not in estradiol-treated osteosarcoma We performed the DAVID functional enrichment analysis cells (Figure 3C). TNFA was significantly up-regulated in for each oncogene/tumor suppressor gene subgroup. Four both glucocorticoid- and estradiol-treated osteosarcoma cells subgroups yielded statistically meaningful DAVID results (Figure 3D). The caspase gene family is a family of cysteine (Tables II-V). The other four subgroups yielded no statistically proteases which is essential for apoptosis (30). Caspase meaningful DAVID results (data not shown). For 17-β- genes can be divided into two types: initiator caspases and estradiol-related subgroups, the biological processes for down- effector caspases, the former activating the latter, while the regulated oncogenes mainly involve the positive regulation of latter cleaving other protein substrates within the cell (30). cell proliferation (Table II); on the other hand, those for down- The expression level of different members of the caspase regulated tumor-suppressor genes involve the negative family before and after hormone treatment is shown in Table

1685 ANTICANCER RESEARCH 36: 1683-1692 (2016)

Figure 1. A: Pie charts of the seven most significant Gene Ontology (GO)-terms for genes down-regulated by 17-β-estradiol genes (A), up-regulated by 17-β-estradiol genes (B), down-regulated by glucocorticoid (C), and up-regulated by glucocorticoid (D). The numbers in the pie chart are the number of genes in that GO category.

VI. In osteosarcoma cells, one initiator caspase (CASP9) was the Agilent-012391 Whole Human Genome Oligo Microarray up-regulated by estradiol, while two initiator caspases G4112A is much larger than the dataset generated with the (CASP2 and CASP9) were down-regulated by glucocorticoid; Affymetrix Human HG-Focus Target Array (GPL201) since both caspase 3 (CASP3) and caspase 7 (CASP7) were up- the former contains 44,291 genes, while the latter contains regulated by estradiol and glucocorticoid, but caspase 6 only about 8,800 genes. This study would be more (CASP6) was down-regulated only by estradiol. informative if two datasets were generated with the Agilent- 012391 Whole Human Genome Oligo Microarray G4112A. Discussion 17-β-Estradiol and glucocorticoid are both steroid hormones which share a similar basic chemical structure. Human U2OS osteosarcoma cells were originally derived The functional-enrichment analyses of the genes from the tibial sarcoma of a 15-year-old girl and are used differentially expressed after treatment with 17-β-estradiol widely in different biomedical research areas (31). We found or glucocorticoid show that these genes share differences and 1,359 genes to be differentially expressed in U2OS similarities in their expression profiles. The similarity is that osteosarcoma cells after 17-β-estradiol treatment and 10,215 17-β-estradiol can inhibit the expression of genes involved after glucocorticoid treatment. Apparently, the number of in cell proliferation and glucocorticoid can inhibit the differentially expressed genes in the 17-β-estradiol dataset is expression of genes involved in the cell cycle. This suggests significantly less than that in glucocorticoid dataset. We that both hormones could halt osteosarcoma cell growth at conclude that the difference in the number of differentially least in vitro. Our further analyses show that 17-β-estradiol expressed genes is more likely attributed to different and glucocorticoid share many target genes in osteosarcoma microarray platforms used to generate these data than the cells, especially among the down-regulated ones. The effect of the different hormones. The dataset generated with number of genes down-regulated by both 17-β-estradiol and

1686 Zhou et al: Differentially Expressed Genes in U2OS Cells After Estradiol and Glycorticoid Treatment

Figure 2. The most interconnected subnetwork found in the functional module network analysis. Circles indicate modules and arrows indicate network flows.

Figure 3. The expression levels of marker of proliferation ki-67 (MKI67) (A), v- avian myeloblastosis viral oncogene homolog-like 2 (MYBL2) (B), apoptosis-inducing factor mitochondrion-associated 1 (AIFM1) (C) and tumor necrosis factor (TNFA) (D) in osteosarcoma cells before and after estradiol or glucocorticoid treatment. Error bar indicates the standard deviation and p-values are shown.

1687 ANTICANCER RESEARCH 36: 1683-1692 (2016)

Table II. DAVID functional annotation result of oncogenes down-regulated by 17-β-estradiol.

Gene Ontology category Term Benjamini-corrected FDR

Biological process Regulation of cell proliferation 5.6×10–4* Enzyme linked receptor protein signaling pathway 6.8×10–4* Transmembrane receptor protein tyrosine kinase signaling pathway 1×10–3* Positive regulation of cell proliferation 1.8×10–3* Positive regulation of DNA replication 0.04* Cellular component Plasma membrane part 0.36 Integral to plasma membrane 0.32 Molecular function Transmembrane receptor protein tyrosine kinase activity 0.046* Protein tyrosine kinase activity 0.086 KEGG pathway Pathways in cancer 1.7×10−4* Prostate cancer 0.0067* ERBB signaling pathway 0.043* MAPK signaling pathway 0.089

*Benjamini-corrected false-discovery rate (FDR) is statistically significant.

Table III. DAVID functional annotation result of tumor suppressor genes down-regulated by 17-β-estradiol.

Gene Ontology category Term Benjamini-corrected FDR

Biological process Negative regulation of cell proliferation 1.81×10–4* Regulation of cell death 5.92×10–4* Regulation of programmed cell death 7.14×10–4* Regulation of apoptosis 8.65×10–4* Regulation of cell proliferation 0.001* Negative regulation of signal transduction 0.0025* Negative regulation of cell communication 0.004* Cellular component Endoplasmic reticulum 0.22 Extracellular space 0.25 Molecular function Identical protein binding 0.66 Transcription co-repressor activity 0.79 KEGG pathway P53 signaling pathway 0.018*

*Benjamini-corrected false-discovery rate (FDR) is statistically significant.

glucocorticoid is almost twice many as the number of genes to the significant difference in the number of differentially up-regulated by both (356 vs. 182). The functional module expressed genes found in the two datasets, we identified more network built on the down-regulated genes common to both oncogenes and tumor suppressor genes regulated by shows that these genes are mostly associated with the glucocorticoid than those regulated by 17-β-estradiol. 17-β- nucleotide (purine) metabolic and biosynthetic sub-network. Estradiol up-regulates more oncogenes than it down- This further suggests that 17-β-estradiol and glucocorticoid regulates, while glucocorticoid down-regulates more could halt osteosarcoma cell growth through down-regulating oncogenes than it up-regulates. This suggests that the genes in charge of nucleotide (purine) metabolic and glucocorticoid might be more effective than 17-β-estradiol in biosynthetic processes. suppressing oncogenesis. The functional enrichment analyses In our study, we were very interested in oncogenes and of the oncogenes down-regulated by 17-β-estradiol or tumor-suppressor genes that would be influenced by 17-β- glucocorticoid show that these oncogenes are involved in the estradiol and glucocorticoid in osteosarcoma cells. They can same biological processes, i.e. regulation of cell proliferation foretell the probable performance of 17-β-estradiol or and cell cycle. Both 17-β-estradiol and glucocorticoid down- glucocorticoid in osteosarcoma treatment. We searched regulate more tumor suppressor genes than they up-regulate. NCG4.0 and TSGene for oncogenes and tumor-suppressor Only 17-β-estradiol down-regulated tumor-suppressor genes genes among the differentially expressed genes (23, 24). Due and glucocorticoid up-regulated tumor suppressor genes

1688 Zhou et al: Differentially Expressed Genes in U2OS Cells After Estradiol and Glycorticoid Treatment

Table IV. DAVID functional annotation result of oncogenes down-regulated by glucocorticoid.

Gene Ontology category Term Benjamini-corrected FDR

Biological process Cell-cycle process 1.23×10–5* Cell cycle phase 1.24×10–5* Mitotic cell cycle 1.30×10–5* Cell cycle 3.07×10–5* Maintenance of protein location in cell 4.32×10–4* Nuclear division 5.53×10–4* Cellular component Intracellular non-membrane-bounded organelle 0.0018* Non-membrane-bounded organelle 0.0019* Hemidesmosome 0.0096* Cytoskeleton 0.0117* Plasma membrane part 0.014* Cell junction 0.016* Molecular function Actin binding 0.01* Transcription factor binding 0.039* Transmembrane receptor protein tyrosine kinase activity 0.047* Sequence-specific DNA binding 0.039* Transcription activator activity 0.039* KEGG pathway Cell cycle 0.035*

*Benjamini-corrected false-discovery rate (FDR) is statistically significant.

Table V. DAVID functional annotation result of tumor-suppressor genes up-regulated by glucocorticoid.

Gene Ontology category Term Benjamini-corrected FDR

Biological process Regulation of cell cycle 6.88×10–7* Regulation of cell proliferation 4.70×10–5* Regulation of apoptosis 2.24×10–4* Regulation of programmed cell death 1.94×10–4* Regulation of cell death 1.63×10–5* Cellular component Cytosol 0.68 Molecular function Transcription regulator activity 0.0014* Protein complex binding 0.02* Transcription activator activity 0.059 KEGG pathway Apoptosis 0.86

*Benjamini-corrected false-discovery rate (FDR) is statistically significant.

Table VI. The expression values (normalized counts, mean±SD) of caspase genes in osteosarcoma cells before and after estradiol or glucocorticoid treatment.

Estradiol Glucocorticoid

Type Caspase Before After p-Value Before After p-Value

Initiator CASP2 0.15±0.03 0.15±0.05 0.808 7.2±0.26 6.1±1.1 0.001 CASP8 0.3±0.09 0.29±0.05 0.78 4.1±0.99 4.3±0.89 0.5 CASP9 0.4±0.11 0.54±0.15 0.01 9.2±0.36 8.9±0.4 0.019 CASP10 0.1±0.04 0.14±0.06 0.075 5.8±0.8 5.4±1 0.22 Effector CASP3 0.33±0.09 0.42±0.05 0.002 10.9±0.1 12.2±0.6 2×10–7 CASP6 0.3±0.1 0.2±0.09 0.018 9.9±0.2 9.8±1.7 0.95 CASP7 0.64±0.2 1.5±0.68 0.0005 9.7±0.35 11±0.45 2×10–9

Significant up- or down-regulation of caspase genes is shown in bold.

1689 ANTICANCER RESEARCH 36: 1683-1692 (2016) yielded statistically meaningful results in the functional 3 Osteosarcoma and malignant fibrous histiocytoma of bone enrichment analyses. The DAVID result showed tumor- treatment (PDQ(r)): Health professional version. In: PDQ cancer suppressor genes down-regulated by 17-β-estradiol are information summaries. Bethesda (MD), 2002. associated with the negative regulation of cell proliferation. 4 Zheng R, Peng X, Zhang S, Chen T, Wang H and Chen W: Incidence, mortality and survival of childhood cancer in china Because 17-β-estradiol both down-regulates and up-regulates during 2000-2010 period: A population-based study. Cancer Lett cancer-associated genes regulating cell proliferation, we infer 28(363(2)): 176-180, 2015. that 17-β-estradiol might be inefficient in controlling 5 Wilkins RM, Cullen JW, Odom L, Jamroz BA, Cullen PM, Fink osteosarcoma cell growth. The DAVID result showed that K, Peck SD, Stevens SL, Kelly CM and Camozzi AB: Superior tumor suppressor genes up-regulated by glucocorticoid are survival in treatment of primary nonmetastatic pediatric osteo- associated with the regulation of apoptosis. Glucocorticoid sarcoma of the extremity. Ann Surg Oncol 10(5): 498-507, 2003. not only down-regulates oncogenes associated with cell cycle 6 DeVita VT, Hellman S and Rosenberg SA: Cancer, Principles & Practice Of Oncology. 7th edn. Lippincott Williams & Wilkins: and mitosis but also up-regulates the tumor-suppressor genes Philadelphia, PA, 2005. in charge of apoptosis. Thus, we argue that glucocorticoid is 7 Ali SH, O'Donnell AL, Balu D, Pohl MB, Seyler MJ, Mohamed a more efficient hormone than estradiol in controlling S, Mousa S and Dandona P: -alpha in the osteosarcoma cell proliferation and apoptosis. inhibition of cancer growth and angiogenesis. Cancer Res In order to support or refute this argument, we examined 60(24): 7094-7098, 2000. the expression levels of cellular markers for proliferation 8 Ockrim JL, Lalani EN, Laniado ME, Carter SS and Abel PD: and apoptosis in osteosarcoma cells before and after Transdermal estradiol therapy for advanced prostate cancer – forward to the past? J Urol 169(5): 1735-1737, 2003. treatment with 17-β-estradiol or glucocorticoid. Two 9 Pufall MA: Glucocorticoids and cancer. Adv Exp Med Biol 872: cellular markers (MKI76 and MYBL2) of proliferation were 315-333, 2015. significantly down-regulated in glucocorticoid-treated 10 Gaynon PS and Carrel AL: Glucocorticosteroid therapy in osteosarcoma cells, but not in estradiol-treated ones. This childhood acute lymphoblastic leukemia. Adv Exp Med Biol implies that glucocorticoid is more effective in controlling 457: 593-605, 1999. cell proliferation. AIFM1 was also significantly up- 11 Herr I and Pfitzenmaier J: Glucocorticoid use in prostate cancer regulated in glucocorticoid-treated osteosarcoma cells. and other solid tumours: Implications for effectiveness of AIFM1 induces cell death which is independent from the cytotoxic treatment and metastases. Lancet Oncology 7(5): 425- 430, 2006. caspase pathway (32). Two effector caspases, CASP3 and 12 Saraiva PP, Teixeira SS, Conde SJ and Nogueira CR: The CASP7, were significantly up-regulated in both importance of analysis in osteosarcoma cells glucocorticoid- and estradiol-treated cells. The analyses growth submitted to treatment with estrogen in association with based on the cellular markers of apoptosis showed that thyroid hormone. Cell Biochem Funct 26(1): 107-110, 2008. glucocorticoid not only induces apoptosis through the 13 Alokail MS and Peddie MJ: Characterisation of ligand binding caspase pathway but also through a caspase-independent to the parathyroid hormone/parathyroid hormone-related peptide pathway, while estradiol is only engaged in the caspase receptor in MCF7 breast cancer cells and saos-2 osteosarcoma cells. Cell Biochem Funct 25(2): 139-147, 2007. pathway. The evidence based on cellular markers supports 14 Valin A, Guillen C and Esbrit P: C-Terminal parathyroid our argument instead of refuting it. hormone-related protein (pthrp) (107-139) stimulates Our study proposes that glucocorticoid is more efficient intracellular Ca(2+) through a receptor different from the type 1 in controlling osteosarcoma cell growth than 17-β-estradiol. pth/pthrp receptor in osteoblastic osteosarcoma UMR 106 cells. It seems to suggest that glucocorticoid is an effective Endocrinology 142(7): 2752-2759, 2001. hormone therapeutic agent for osteosarcoma treatment. 15 Stossi F, Barnett DH, Frasor J, Komm B, Lyttle CR and However, there is evidence that glucocorticoid drugs have a Katzenellenbogen BS: Transcriptional profiling of estrogen- wide range of side-effects such as immunodeficiency and regulated gene expression via estrogen receptor (ER) alpha or erbeta in human osteosarcoma cells: Distinct and common target osteoporosis (33, 34). Some side-effects can do more harm genes for these receptors. Endocrinology 145(7): 3473-3486, 2004. than good for patients with osteosarcoma. Therefore, we 16 Monroe DG, Secreto FJ, Subramaniam M, Getz BJ, Khosla S argue that more research and evaluations of glucocorticoid and Spelsberg TC: and beta are needed before it can be used clinically in osteosarcoma heterodimers exert unique effects on estrogen- and tamoxifen- treatment. dependent gene expression in human U2OS osteosarcoma cells. Mol Endocrinol 19(6): 1555-1568, 2005. References 17 Jewell CM, Scoltock AB, Hamel BL, Yudt MR and Cidlowski JA: Complex human dim mutations 1 Jaffe N, Bruland ØS and Bielack SS: Pediatric and adolescent define glucocorticoid induced apoptotic resistance in bone cells. osteosarcoma. Springer: New York, 2009. Mol Endocrinol 26(2): 244-256, 2012. 2 Luetke A, Meyers PA, Lewis I and Juergens H: Osteosarcoma 18 Lu NZ, Collins JB, Grissom SF and Cidlowski JA: Selective treatment – Where do we stand? A state of the art review. Cancer regulation of bone cell apoptosis by translational isoforms of the Treat Rev 40(4): 523-532, 2014. glucocorticoid receptor. Mol Cell Biol 27(20): 7143-7160, 2007.

1690 Zhou et al: Differentially Expressed Genes in U2OS Cells After Estradiol and Glycorticoid Treatment

19 Huang da W, Sherman BT and Lempicki RA: Bioinformatics 29 Locksley RM, Killeen N and Lenardo MJ: The TNF and TNF enrichment tools: Paths toward the comprehensive functional receptor superfamilies: Integrating mammalian biology. Cell analysis of large gene lists. Nucleic Acids Res 37(1): 1-13, 2009. 104(4): 487-501, 2001. 20 Huang da W, Sherman BT and Lempicki RA: Systematic and 30 Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G, integrative analysis of large gene lists using david bioinformatics Thornberry NA, Wong WW and Yuan J: Human ICE/CED-3 resources. Nat Protoc 4(1): 44-57, 2009. protease nomenclature. Cell 87(2): 171, 1996. 21 Maere S, Heymans K and Kuiper M: Bingo: A cytoscape plugin 31 Niforou KM, Anagnostopoulos AK, Vougas K, Kittas C, to assess overrepresentation of gene ontology categories in Gorgoulis VG and Tsangaris GT: The proteome profile of the biological networks. Bioinformatics 21(16): 3448-3449, 2005. human osteosarcoma U2OS cell line. Cancer Genomics 22 Rivera CG, Vakil R and Bader JS: Nemo: Network module Proteomics 5(1): 63-78, 2008. identification in cytoscape. BMC Bioinformatics 11(Suppl 1): 32 Cande C, Cohen I, Daugas E, Ravagnan L, Larochette N, S61, 2010. Zamzami N and Kroemer G: Apoptosis-inducing factor (aif): A 23 An O, Pendino V, D'Antonio M, Ratti E, Gentilini M and novel caspase-independent death effector released from Ciccarelli FD: NCG 4.0: The network of cancer genes in the era mitochondria. Biochimie 84(2-3): 215-222, 2002. of massive mutational screenings of cancer genomes. Database 33 Kolesnyk MO and Dorets'kyi VV: The immunodeficiency state (Oxford) 2014: bau015, 2014. and the efficacy of the glucocorticoid therapy of patients with 24 Zhao M, Sun J and Zhao Z: Tsgene: A web resource for tumor glomerulonephritis and the nephrotic syndrome. Lik Sprava 11- suppressor genes. Nucleic Acids Res 41(Database issue): D970- 12): 53-55, 1992 (in Ukrainian). 976, 2013. 34 Blalock SJ, Norton LL, Patel RA and Dooley MA: Patient 25 Whitfield ML, George LK, Grant GD and Perou CM: Common knowledge, beliefs, and behavior concerning the prevention and markers of proliferation. Nat Rev Cancer 6(2): 99-106, 2006. treatment of glucocorticoid-induced osteoporosis. Arthritis 26 Bullwinkel J, Baron-Luhr B, Ludemann A, Wohlenberg C, Rheum 53(5): 732-739, 2005. Gerdes J and Scholzen T: Ki-67 protein is associated with ribosomal RNA transcription in quiescent and proliferating cells. J Cell Physiol 206(3): 624-635, 2006. 27 Noben-Trauth K, Copeland NG, Gilbert DJ, Jenkins NA, Sonoda G, Testa JR and Klempnauer KH: Mybl2 (bmyb) maps to mouse 2 and human chromosome 20q 13.1. Genomics 35(3): 610-612, 1996. 28 Susin SA, Lorenzo HK, Zamzami N, Marzo I, Snow BE, Brothers GM, Mangion J, Jacotot E, Costantini P, Loeffler M, Larochette N, Goodlett DR, Aebersold R, Siderovski DP, Penninger JM and Received January 8, 2016 Kroemer G: Molecular characterization of mitochondrial Revised February 23, 2016 apoptosis-inducing factor. Nature 397(6718): 441-446, 1999. Accepted February 24, 2016

1691