Prostate cancer is the most common cancer diagnosed among men worldwide and the second most common cause of cancer death [1]. However, prostate cancer can be classified to castration-resistant and hormone- dependent prostate cancer. According to previous reports related to prostate cancer, apoptosis plays an important role in the chemotherapy response, suggesting a relationship between apoptosis induced by drug and treatment efficiency [2]. The resistance of drug is a critical issue for cancer treatment but its underlying mechanisms are unclear. Docetaxel is an anti-microtubule drug with antitumor activity in different cancers, such as prostate cancer, which acts by binding to tubulin and disrupting the balance between disassembly during mitosis and microtubule assembly [3]. It was previously reported which docetaxel downregulates some genes related to proliferation of cell, formation of mitotic spindle, transcription factors and oncogenesis and upregulates some genes associated with apoptosis induction and cell cycle arrest in prostate cancer cells [4]. The mechanism of resistant prostate cancer to docetaxel is related to different factors, such as expression of androgen receptor splice variant, β-tubulin expression alternations, multidrug resistance caused by unnatural expression of the ATP binding cassette (ABC) transporter family and unnatural expression of NF-κb/interleukin (IL) 6 and PI3K/AKT/mTOR signaling pathways [5]. CDH1 gene has identified in PC-3 and DU145 cells resistant to docetaxel and confirmed in prostate tumors from patients resistant to docetaxel [6]. According to the results of Jiang et al [7], mutation of the CDH1 gene is related to invasion and metastasis in various types of cancer, because it alters the transcriptional activity of epithelial cells. Previous research has shown that enhanced expression of the ABCB1 gene causes resistance to docetaxel and that ABCB1 expression is associated with prostate tumors [8]. Calcium-binding protein S100 A4 (S100A4) plays an important role in the cancer metastasis progression and has been reported to mediate migration, invasion and prostate cancer apoptosis and also plays a role in the extracellular and intracellular regions of tumors [9]. TSPAN1 is a novel member of the tetraspanin family and its role in the progression of tumor has been reported and it can enhance the proliferation and invasion of tumor cell in the laboratory and its expression is increased in human tumor cells [10]. Results from studies that combine different results in the form of meta-analysis are more reliable and provide a better answer than single results. In cancer-related studies, results from meta-analysis provide a better understanding of molecular mechanisms. Microarrays technology has allowed us to simultaneously examine the expression of more genes and identify the molecular mechanism involved in the resistance of prostate cancer cells to docetaxel. Therefore, the results from the integration of various studies could lead to the identification of important genes that are associated with the molecular mechanisms of prostate cancer cell resistance to docetaxel. Generally, this study provides information about regulatory genes, transcription factors (TFs) and gene ontology (GO) related to docetaxel resistance in prostate cancer.

To investigate the expression of microarray genes associated with docetaxel-resistant prostate cancer cells were extracted through the Gene Expression Omnibus (GEO). The samples tested in this study after quality control are listed in the Table 1. The general methodology, from raw data extraction to identification of key genes and related work in this research, is presented in Figure 1.   Table 1: Docetaxel-Resistant Prostate Cancer Cell Samples Studied from a Meta-Analysis of Microarray Data Platform Accession number Cell lines Resistant:Sensitive Status of cells selected for testing GPL571 GSE36135 DU145 cell line 22Rv1 cell line 3:3 3:3 Cell lines such as DU-145 and 22RV1 were obtained from American Type Culture Collection (ATCC) GPL6244 GSE47040 TaxR cell line C4-2B cell line 1:1 cells lines were obtained from the ATCC. GPL26944 GSE158494 DU-145 cell line PC-3 cell line 3:3 3:3 Tumor samples from patients diagnosed with mCRPC treated with D or CZ. GPL570 GSE33455 DU-145 cell line PC-3 cell line 3:3 3:3 Patients were given docetaxel-based therapy after removal of the obstruction through the urethra in a palliative manner or after biopsy after tumor spread.       Figure 1: Meta-Analysis Scheme of Microarray Data   Meta-Analysis and Differentially Expressed Genes (DEGs) Microarray expression data were pre-analyzed using the R language package. Data control was done using boxplot and Heatmap (S2). The background correction and quantile normalization were performed on the unnormalized data. The Limma software package was used to perform various analyses on the selected data and finally a linear model was drawn. An empirical model of simple Bayesian was applied for correct standard errors [11]. For unadjusted pooled p-values, the log-sum Fisher's exact approach was used and the ‘fdr’ method was used to apply the p.adjust function in meta-RNASeq to adjust the combined p-values for the microarray expression data (S3). The genes with p-values ≤0.01 and |log2FoldChange|>1 were used as DEGs. The method and analyses used on each GSEs (for example GSE33455) to extract DEGs are shown in Supplementary file (S1).   GO and Biological Pathways Enrichment analysis for GO was performed for genes with increased (up genes) and decreased (down genes) expression metaDEGs using g.profiler, (https://biit.cs.ut.ee/gprofiler/ gost). Shared GO terms for up and downregulated genes (metaDEGs) were identified through the web tool Venny v2.1. Biological pathways related to metaDEGs for up and downgenes were identified through Kyoto Genes and Genomes Encyclopedia (KEGG) (https://www.kegg.jp/ kegg/pathway.html).   Investigation of Motifs and TFs Related to Promoters For the analysis TFs and motifs, the enrichment score threshold of 3.5 and a minimum similarity between orthologous genes of 0.05 were chosen. For this analysis, the obtained results related to DEGs of up and downregulated genes were analyzed using iRegulon through Cytoscape 3.6.1 software.   Identifying Hub Genes Through PPI To analyze the Protein-Protein Interaction (PPI) of up and down regulated genes, the STRING (version db 12) was used and the PPI gene network was drawn using the CytoHubba plugin in Cytoscape 3.6.1. Gene network was drawn using different methods including EPC, Bottle Neck, Closeness, Degree, DMNC, EcCentricity, MNC, Betweenness, Stress, ClusteringCoefficient, Radiality and MCC for the identified up- and down-regulated genes and high-ranking genes (high degree) were selected as central genes. The CytoHubba plugin identified more relevant genes as central genes in the biological network using various methods [12].   Validation of DEGs for Up and Downregulated Genes Through Reported Studies Based on the genes obtained from this study related to DEGs, the gene expression of eight genes (four upregulated and four downregulated genes) was compared with reported qRT-PCR studies. For final confirmation, it is suggested that the expression of these genes be experimentally confirmed and tested in the laboratory so that these genes can be introduced as key and effective genes associated with docetaxel-resistant prostate cancer cells.

Identify of DEGs The present study showed 373 upregulated and 558 downregulated genes associated with prostate cancer cells resistant to docetaxel (p-value <0.05 and |log2FoldChange|> 1). Moreover, top down and upregulated meta-DEGs (|log2FoldChange|>1.5 and meta-pvalue <0.05) related to docetaxel-resistant prostate cancer cells are represented in Table 2. Accordingly, the upregulated genes involved in docetaxel-resistant prostate cancer cell lines mainly included IGFBP3, ABCB1, GSPT2, ROBO1, FSTL1, NID2, S100A4, CDH1 and ADAMTS1 and downregulated genes mainly related to CD24, ZNF587B, PARP2, MIR4271, OCLN, TSPAN1, ADGRG2, MPZL2 and POP7.   Table 2: Key Genes with High log2FC Related to Prostate Cancer Cells Resistant to Docetaxel Upregulated genes Gene.ID Gene Stable ID Gene name Log2FC Meta-pvalue Description 3486 ENSG00000146674 IGFBP3 2.97 0 Insulin like growth factor binding protein 3 5243 ENSG00000085563 ABCB1 2.89 1.71E-07 ATP binding cassette subfamily B member 1 23708 ENSG00000189369 GSPT2 2.66 0 G1 to S phase transition 2 5244 ENSG00000005471 ABCB4 2.63 3.11E-06 ATP binding cassette subfamily B member 4 6091 ENSG00000169855 ROBO1 2.61 0 Roundabout guidance receptor 1 11167 ENSG00000163430 FSTL1 2.23 0 Follistatin like 1 22795 ENSG00000087303 NID2 2.22 0 Nidogen 2 6275 ENSG00000196154 S100A4 2.15 0 S100 calcium binding protein A4 1010 ENSG00000154162 CDH1 2.07 0 Cadherin 1 9510 ENSG00000154734 ADAMTS1 2.06 0 ADAM metallopeptidase with thrombospondin 8277 ENSG00000007350 TKTL1 1.96 0 Transketolase like 1 5654 ENSG00000166033 HTRA1 1.91 0 HtrA serine peptidase 1 3398 ENSG00000115738 ID2 1.88 0 Inhibitor of DNA binding 2 5270 ENSG00000135919 SERPINE2 1.88 0 Serpin family E member 2 55076 ENSG00000181458 TMEM45A 1.83 0 Transmembrane protein 45A 2669 ENSG00000164949 GEM 1.81 0 GTP binding protein overexpressed in skeletal muscle 10962 ENSG00000213190 MLLT11 1.68 0 MLLT11 transcription factor 7 cofactor 57088 ENSG00000114698 PLSCR4 1.66 0 Phospholipid scramblase 4 4162 ENSG00000076706 MCAM 1.66 0 Melanoma cell adhesion molecule 2690 ENSG00000112964 GHR 1.66 0 Growth hormone receptor 668 ENSG00000183770 FOXL2 1.58 0 Forkhead box L2 79901 ENSG00000071967 CYBRD1 1.56 0 Cytochrome b reductase 1 8349 ENSG00000184678 H2BC21 1.53 0 H2B clustered histone 21 1287 ENSG00000188153 COL4A5 1.52 0 Collagen type IV alpha 5 Downregulated genes Gene.ID Gene Stable ID Gene name Log2FC Meta-pvalue Description 100133941 ENSG00000272398 CD24 -3.32 0 CD24 molecule 100293516 ENSG00000269343 ZNF587B -2.98 0 Zinc finger protein 587B 10038 ENSG00000129484 PARP2 -2.82 0 Poly (ADP-ribose) polymerase 2 100422952 ENSG00000264633 MIR4271 -2.66 0 microRNA 4271 100506658 ENSG00000197822 OCLN -2.55 0 Occludin 10103 ENSG00000117472 TSPAN1 -2.42 0 Tetraspanin 1 10149 ENSG00000173698 ADGRG2 -2.30 0 Adhesion G protein-coupled receptor G2 10205 ENSG00000149573 MPZL2 -2.21 0 Myelin protein zero like 2 10248 ENSG00000172336 POP7 -2.15 0 POP7 homolog 10279 ENSG00000112812 PRSS16 -1.85 0 Serine protease 16 10282 ENSG00000105829 BET1 -1.74 0 Bet1 golgi vesicular membrane trafficking protein 10295 ENSG00000103507 BCKDK -1.66 0 Branched chain keto acid dehydrogenase kinase 10329 ENSG00000118600 RXYLT1 -1.62 0 Ribitol xylosyltransferase 1 1033 ENSG00000100526 CDKN3 -1.58 0 Cyclin dependent kinase inhibitor 3 10360 ENSG00000107833 NPM3 -1.56 0 Nucleophosmin/nucleoplasmin 3 10371 ENSG00000075213 SEMA3A -1.54 0 Semaphorin 3A     GO and biological pathways To identify meaningful (p<0.05 and |log2FoldChange|>1) GO terms for biological processes related to resistant prostate cancer cells to docetaxel, 145 and 187 terms were enriched in down and upregulated genes, respectively (Figure 2). Shared GO terms in down and upregulated genes are mainly related to the regulation of biological processes, cell death, stimulus response and apoptotic process (Table 3). According to the top 15 GO, GO terms in the upregulated genes mostly associated with the regulation of cell differentiation, signal transduction, cell death and apoptotic process; and in the downregulated genes mainly belonged to response to chemicals, defense and immune system process, cell death and proliferation (Table 4). Table 3: Common GO Enrichment in Down and Up DEGs Involved in the Resistant Prostate Cancer Cells to Docetaxel Common GO Description Common GO Description GO:0048731 System development GO:0010648 Cell communication regulation GO:0048856 Anatomical structure development GO:0023057 Signaling regulation GO:0007275 Multicellular organism development GO:0008219 Cell death GO:0009653 Anatomical structure morphogenesis GO:0030334 Cell migration regulation GO:0032502 Developmental process GO:0023051 Signaling regulation GO:0035239 Morphogenesis of tube GO:0010646 Cell communication regulation GO:0048513 Development of animal organ GO:0048522 Cellular process regulation GO:0035295 Development of tube GO:0030855 Differentiation of epithelial cell GO:0048523 Cellular process regulation GO:0010033 Organic substance response GO:0009888 Development of tissue GO:0048583 Stimulus response regulation GO:0051239 Multicellular process regulation GO:0051716 Stimulus cellular response GO:0008283 Cell population proliferation GO:0070887 Chemical stimulus GO:0042127 Proliferation of cell population GO:2000145 Regulation of cell motility GO:0051240 Multicellular process regulation GO:0030512 Signalling pathway GO:0048519 Biological process regulation GO:0009719 Endogenous stimulus response GO:0016477 Cell migration GO:0006915 Process of apoptotic GO:0032501 Process of multicellular GO:0007154 Communication of cell Common GO Description Common GO Description GO:0050896 Stimulus response GO:0050794 Cellular process regulation GO:0042221 Chemical response GO:0048732 Gland development GO:0040011 Locomotion GO:0035556 Signal transduction of intracellular GO:0048870 Cell motility GO:0051128 Cellular component organization regulation GO:0010941 Cell death regulation GO:0012501 Programmed cell death GO:0009968 Signal transduction regulation GO:0008284 Cell population proliferation regulation GO:0048585 Stimulus response regulation GO:0006935 Chemotaxis GO:0009966 Signal transduction regulation GO:0042330 Taxis GO:0065009 Molecular function regulation GO:0071310 Cellular response GO:0060429 Epithelium development GO:0050789 Biological process regulation GO:0048518 Biological process regulation     Table 4: Top 15 GO Annotation’s Biological Process Involved in Docetaxel-Resistant Prostate Cancer Cells Upregulate Genes GO Gene.ID Gene Stable ID Gene.Symbol adjusted p-value Description GO:0030154 2296 ENSG00000054598 FOXC1 8.82E-10 Cell differentiation GO:0009888 3913 ENSG00000172037 LAMB2 4.13E-08 Tissue development GO:0045597 2690 ENSG00000112964 GHR 8.26E-08 Cell differentiation regulation GO:0045595 2737 ENSG00000106571 GLI3 1.17E-07 Cell differentiation regulation GO:0050896 9734 ENSG00000048052 HDAC9 1.22E-07 Response to stimulus GO:0009968 26575 ENSG00000091844 RGS17 2.52E-06 Signal transduction regulation GO:0009966 81848 ENSG00000187678 SPRY4 7.57E-06 Signal transduction regulation GO:0008219 143 ENSG00000102699 PARP4 4.55E-05 Cell death GO:0060548 10979 ENSG00000073712 FERMT2 6.81E-05 Regulation of cell death GO:0048598 2737 ENSG00000106571 GLI3 08.69E-05 Embryonic morphogenesis GO:0042981 9748 ENSG00000065613 SLK 1.05E-04 Regulation of apoptotic process GO:0048589 2737 ENSG00000106571 GLI3 1.32E-04 Developmental growth GO:0006915 9821 ENSG00000023287 RB1CC1 6.97E-04 Apoptotic process GO:0071363 90 ENSG00000115170 ACVR1 1.64 E-03 Cellular response to growth factor stimulus GO:0012501 5654 ENSG00000166033 HTRA1 1.57 E-03 Programmed cell death Downregulated genes GO Gene.ID Gene Stable ID Gene.Symbol adjusted_p_value Description GO:0050896 6742 ENSG00000012048 BRCA1 5.69E-11 Response to stimulus GO:0071345 2633 ENSG00000107485 GATA3 2.02E-10 Response to cytokine stimulus GO:0002376 317749 ENSG00000029993 HMGB3 3.63E-10 Immune system process GO:0006260 6019 ENSG00000049541 RFC2 4.01E-09 DNA replication GO:0019221 79023 ENSG00000115590 IL1R2 5.86E-09 Cytokine-mediated signalling pathway GO:0033993 6821 ENSG00000072310 SREBF1 1.47E-08 Response to lipid GO:0006974 6742 ENSG00000012048 BRCA1 3.21E-08 Cellular response to DNA damage stimulus GO:0035556 55972 ENSG00000027075 PRKCH 4.79E-07 Intracellular signal transduction GO:0006952 2633 ENSG00000107485 GATA3 2.1E-06 Defence response GO:0030334 339166 ENSG00000041982 TNC 3.26E-05 Regulation of cell migration GO:0030856 2633 ENSG00000107485 GATA3 1.05E-04 Regulation of epithelial cell differentiation GO:0006915 3576 ENSG00000026103 FAS 3.07E-04 Apoptotic process GO:0009725 2274 ENSG00000115641 FHL2 3.19E-03 Response to hormone GO:0050678 6374 ENSG00000015475 BID 3.52E-03 Cell proliferation GO:0008219 2013 ENSG00000213853 EMP2 2.83E-02 Cell death       Figure 2: Common GO Term of Up and Downregulated Genes   According to our results, the down and up meta-DEGs by KEGG pathways mainly related to cancer pathways (up genes such as HMOX1, GSTP1 and CEBPA and down gene such as CSF2RA), PI3K-Akt signaling pathway (up genes such as PTK2 and ITGAV and BRCA1as down gene), cytokine-cytokine receptor interaction (up genes such as IL13RA2 and TNFSF4 and down gene such as IL15RA, CSF2RA and LTB), MAPK signaling pathway (up genes such as FOS and JUND and down gene such as FAS), cancer transcriptional misregulation (up genes such as ZEB1 and down gene such as PLAU and IL1R2), MicroRNAs in cancer (EZH2, GLS2 and KIF23 as down genes), apoptosis (up genes such as TUBA1A and LMNB1 and PARP2 as down genes), p53 signaling pathway (up genes such as TP53I3 and IGFBP3 and FAS and BID as down genes), prostate cancer (ZEB1 as up gene and PLAU and IL1R2 as down genes) and DNA replication (PRIM1 and RFC2/5 as down genes).   Identification of TFs The motifs and TFs of down and up meta-DEGs in resistant prostate cancer cells to docetaxel are shown in S Table 1. The dominant TFs associated with upregulated genes included SRF, TBPs, POLR2A, TAF1 and RREB as well as downregulated genes included IRFs, ZEB1, PRDM1, SPI1, JUNs, FOSs and NFKBs. Moreover, common TFs in up and down meta-DEGs were identified as STATs, E2F1, MYB, FOXs, ZNFs and BCL6.   PPI Gene Network and Hub Genes PPI gene network for key up and down genes (30 genes) was drawn separately using CytoHubba plugin in Cytoscape 3.6.1 based on level degrees and STRING was also used for the interaction between proteins (Figure 3). The string results for up and down genes are given in the supplementary file (S4). Based on the findings, COL4A1 (collagen type IV alpha 1 chain), LOX (lysyl oxidase), SMAD2 (SMAD family member 2), TIMP1 (TIMP metallopeptidase inhibitor 1) and FOS (Fos proto-oncogene) were identified as key upregulated hub genes. Generally, the hub genes of upregulated mostly associated with collagen types and metastasis-related factors. Next, EXO1 (exonuclease 1), RRM2 (ribonucleotide reductase regulatory subunit M2), CDC45 (cell division cycle 45), KIAA0101 (PCNA clamp associated factor), MCM10 (minichromosome maintenance 10 replication initiation factor) and HMMR (hyaluronan mediated motility receptor) were reported as key downregulated hub genes. Moreover, the hub genes of downregulated mainly related to molecular processes and cancer-related factors (Table 5).     Table 5: Key Hub Genes of Up and Downregulated Genes in The Resistant Prostate Cancer Cells to Docetaxel by MCC Method Rank Gene ID Gene Stable ID Degree Gene.Symbol Description Upregulated genes 1 1289 ENSG00000130635 18 COL5A1 Collagen type V alpha 1 chain 2 1282 ENSG00000187498 17 COL4A1 Collagen type IV alpha 1 chain 3 1284 ENSG00000134871 16 COL4A2 Collagen type IV alpha 2 chain 4 5351 ENSG00000083444 15 PLOD1 Procollagen-lysine,2-oxoglutarate 5-dioxygenase 1 5 1287 ENSG00000188153 14 COL4A5 Collagen type IV alpha 5 chain 6 19316 ENSG00000117385 14 LEPRE1 Leucine proline-enriched proteoglycan (leprecan) 1 7 4015 ENSG00000113083 13 LOX Lysyl oxidase 8 4087 ENSG00000175387 13 SMAD2 SMAD family member 2 9 5054 ENSG00000106366 12 SERPINE1 Serpin family E member 1 10 871 ENSG00000149257 11 SERPINH1 Serpin family H member 1 11 19317 ENSG00000090530 10 LEPREL1 Leprecan-like 1 12 4089 ENSG00000141646 9 SMAD4 SMAD family member 4 13 2200 ENSG00000166147 9 FBN1 Fibrillin 1 14 7076 ENSG00000102265 9 TIMP1 TIMP metallopeptidase inhibitor 1 15 2353 ENSG00000170345 9 FOS Fos proto-oncogene, AP-1 TF subunit 16 4092 ENSG00000101665 7 SMAD7 SMAD family member 7 17 7077 ENSG00000035862 7 TIMP2 TIMP metallopeptidase inhibitor 2 18 6935 ENSG00000148516 7 ZEB1 Zinc finger E-box binding homeobox 1 19 6876 ENSG00000149591 7 TAGLN Transgelin 20 25937 ENSG00000018408 7 WWTR1 WW domain containing transcription regulator 1 21 8850 ENSG00000114166 7 KAT2B Lysine acetyltransferase 2B 22 5747 ENSG00000169398 6 PTK2 Protein tyrosine kinase 2 23 3685 ENSG00000138448 6 ITGAV Integrin subunit alpha V 24 2034 ENSG00000116016 6 EPAS1 Endothelial PAS domain protein 1 25 11167 ENSG00000163430 6 FSTL1 Follistatin like 1 26 3486 ENSG00000146674 5 IGFBP3 Insulin like growth factor binding protein 3 27 467 ENSG00000162772 5 ATF3 Activating transcription factor 3 28 1050 ENSG00000245848 5 CEBPA CCAAT enhancer binding protein alpha 29 6275 ENSG00000196154 5 S100A4 S100 calcium binding protein A4 30 1649 ENSG00000175197 3 DDIT3 DNA damage inducible transcript 3 Downregulated genes Rank Gene ID Gene Stable ID Degree Gene.Symbol Description 1 9388 ENSG00000174371 29 EXO1 Exonuclease 1 2 6302 ENSG00000171848 29 RRM2 Ribonucleotide reductase regulatory subunit M2 3 837 ENSG00000093009 29 CDC45 Cell division cycle 45 4 9768 ENSG00000166803 29 KIAA0101 PCNA clamp associated factor 5 5557 ENSG00000065328 29 MCM10 Minichromosome maintenance 10 replication initiation factor 6 9601 ENSG00000137807 29 KIF23 Kinesin family member 23 7 10635 ENSG00000111247 28 RAD51AP1 RAD51 associated protein 8 3198 ENSG00000072571 28 HMMR Hyaluronan mediated motility receptor 9 9111 ENSG00000198901 28 PRC1 Protein regulator of cytokinesis 1 10 6742 ENSG00000012048 28 BRCA1 BRCA1 DNA repair associated 11 55215 ENSG00000138180 28 CEP55 Centrosomal protein 55 12 55247 ENSG00000140525 28 FANCI FA complementation group I 13 9914 ENSG00000184445 27 KNTC1 Kinetochore associated 1 14 79762 ENSG00000151725 27 CENPU Centromere protein U 15 54898 ENSG00000146918 27 NCAPG2 Non-SMC condensin II complex subunit G2 16 439 ENSG00000148773 27 MKI67 Marker of proliferation Ki-67 17 11169 ENSG00000198554 26 WDHD1 WD repeat and HMG-box DNA binding protein 1 18 1033 ENSG00000100526 26 CDKN3 Cyclin dependent kinase inhibitor 3 19 54458 ENSG00000100479 26 POLE2 DNA polymerase epsilon 2 20 23397 ENSG00000121152 26 NCAPH Non-SMC condensin I complex subunit H 21 3149 ENSG00000119969 25 HELLS Helicase, lymphoid specific 22 2305 ENSG00000111206 25 FOXM1 Forkhead box M1 23 55612 ENSG00000198056 24 PRIM1 DNA primase subunit 1 24 81892 ENSG00000167513 24 CDT1 Chromatin licensing and DNA replication factor 1 25 10721 ENSG00000051341 23 POLQ DNA polymerase theta 26 7134 ENSG00000167900 22 TK1 Thymidine kinase 1 27 2146 ENSG00000106462 21 EZH2 Enhancer of zeste 2 polycomb repressive complex 2 28 6296 ENSG00000167325 20 RRM1 Ribonucleotide reductase catalytic subunit M1 29 55312 ENSG00000109674 19 NEIL3 nei like DNA glycosylase 3 30 83451 ENSG00000159259 19 CHAF1B Chromatin assembly factor 1 subunit B       Figure 3: PPI Gene Network of (a) Key Hub Genes of Upregulated and (b) Key Hub Genes of Downregulated in Docetaxel-Resistant Prostate Cancer Cells via MCC Method   Based on the results obtained from various methods (twelve methodes) used to identify hub genes through Cytohubba plugin in Cytoscape 3.6.1, LOX (lysyl oxidase), SMAD2 (SMAD family member 2), SERPINE1 (serpin family E member 1) and FOS (Fos proto-oncogene, AP-1 transcription factor subunit) in eleven methods; PTK2 (protein tyrosine kinase 2) in eight methods and TIMP1(TIMP metallopeptidase inhibitor 1) in five methods were identified as key hub genes related to upregulated genes (Figure 4).     Figure 4: PPI Gene Network of Key Hub Genes (10 Genes) of Upregulated in Resistant Prostate Cancer Cells to Docetaxel Through Twelve Methods   Moreover, BRCA1 (BRCA1 DNA repair associated) in ten methods; EGFR (epidermal growth factor receptor) in eight methods; IL6 (interleukin 6) in seven methods; CXCL8 (C-X-C motif chemokine ligand 8) in sex methods; EPRS (glutamyl-prolyl-tRNA synthetase 1), MKI67 (marker of proliferation Ki-67) and EXO1 (exonuclease 1) in five methods were identified as key hub genes related to downregulated genes (Figure 5). The results obtained from the analysis of twelve methods to identify 10 key hub genes are shown in the Table 6.     Table 6: Key Hub Genes (10 Genes) of Up and Downregulated Genes by Twelve Topological Algorithms of PPI Gene Network Analysis Through Cytohubba Methods Upregulated genes MCC COL5A1, COL4A1, COL4A2, PLOD1, COL4A5, LEPRE1, LOX, SMAD2, SERPINE1 and SERPINH1 MNC SMAD2, SMAD4, LOX, FOS, SERPINE1, TIMP1, FBN1, COL5A1, COL4A1 and PTK2 Degree SMAD2, SMAD4, LOX, SERPINE1, FOS, TIMP1, COL5A1, PTK2, FBN1 and KAT2B Closeness SMAD2, SMAD4, FOS, SERPINE1, LOX, TIMP1, PTK2, SMAD7, ZEB1 and CEBPA BottleNeck SMAD4, PTK2, KAT2B, SMAD2, SERPINE1, LOX, COPS5, CEBPA, FOS and ZEB1 Stress SMAD4, FOS, SMAD2, SERPINE1, KAT2B, LOX, COPS5, PTK2, HIST1H4F and UBXN7 DMNC LEPREL1, LEPRE1, FSTL1, COL4A5, TGFBR3, WDR45, COL4A1, COL4A2, PLOD1 and TIMP2 EcCentricity FOS, SMAD7, SMAD4, PMP22, MUC1, DPP4, ACKR3, S100A4, ACVR1 and TGFBR3 EPC SMAD2, LOX, SMAD4, SERPINE1, TIMP1, FOS, FBN1, COL5A1, COL4A1 and PTK2 Radiality SMAD4, SMAD2, FOS, SERPINE1, LOX, TIMP1, SMAD7, PTK2, ZEB1 and CEBPA Betweenness FOS, SMAD4, SMAD2, SERPINE1, LOX, HIST1H4F, PTK2, KAT2B, COPS5 and CEBPA ClusteringCoefficient LEPREL1, WDR45, NFASC, NRCAM, SLC47A1, HIST1H1C, BNIP3L, LEPRE1, FAM127B and PACSIN2 Methods Downregulated genes MCC EXO1, RRM2, CDC45, KIAA0101, MCM10, KIF23, RAD51AP1, HMMR, PRC1 and BRCA1 MNC EGFR, IL6, BRCA1, EXO1, MKI67, CXCL8, CDC45, KIF23, POLE2 and RRM2 Degree EGFR, IL6, BRCA1, EXO1, MKI67, KIF23, CDC45, CXCL8, POLE2 and RAD51AP1 Closeness EGFR, IL6, BRCA1, CXCL8, MKI67, RRM1, EZH2, EPRS, EXO1 and ICAM1 BottleNeck IL6, EGFR, EPRS, BRCA1, MKI67, RAC2, EMG1, ENO2, SSBP1 and SREBF1 Stress EGFR, IL6, EPRS, BRCA1, RAC2, CXCL8, DCXR, ME3, RRM1 and PHGDH DMNC NCAPG2, KNTC1, CDCA3, PRC1, HMMR, CHAF1B, ZWILCH, MCM10, CEP55 and NEIL3 EcCentricity IDH1, BRCA1, SOCS2, PSMB9, DENR, RFC2, FAS, VAV1, TIMM10 and MDH2 EPC BRCA1, EXO1, EGFR, RAD51AP1, NCAPH, PRC1, MCM10, KIAA0101, KIF23 and WDHD1 Radiality EGFR, IL6, BRCA1, CXCL8, MKI67, RRM1, EPRS, EZH2, PHGDH and EPCAM Betweenness EGFR, IL6, EPRS, BRCA1, RAC2, CXCL8, PHGDH, RRM1, EPCAM and ICAM1 ClusteringCoefficient C1orf112, CDCA3, IFITM2, IFITM3, IL1R2, CCL24, E2F8, NCAPG2, ISG20 and CXCL6     Figure 5: PPI Gene Network of Key Hub Genes (10 Genes) of Downregulated in Resistant Prostate Cancer Cells to Docetaxel Through Twelve Methods   To prove the results obtained from this research in the first stage; our results were compared with the results of research reported by other researchers based on gene expression data (fold-change expression) from molecular studies, including qRT-PCR results. The upregulated genes of this study including (CLU and DDIT3 [13], GSPT2 and CYBRD1 [14]) and the downregulated genes including (MPZL2 [15], FOXM1 [16], SCEL and TXNIP [14]) match and agree with the results of reported qRT-PCR studies (Figure 6). To confirm the results of this research, scientific confirmation in the laboratory is needed experimentally.   Figure 6: Comparison of the Present Results with the qRT-PCR Results of Reported Studies, (a) Up Regulated Genes and (b) Down Regulated Genes

Microarray technologies have widely been used to simultaneously examine gene on large scale, providing an efficient method for examine the expression of thousands of genes. So, the integration and analysis of data related to microarray data supply precious information for the investigation of resistance prostate cancer cells to docetaxel [17]. Based on our results, 373 upregulated genes and 558 downregulated genes were identified. Accordingly, top-upregulated genes associated with docetaxel-resistant prostate cancer cells were mostly included IGFBP3, ABCB1, GSPT2, ROBO1, FSTL1, NID2, S100A4, CDH1 and ADAMTS1 (Table 2). Insulin-like growth factor binding protein 3 (IGFBP3) inhibits adhesion of cell, endometrial cancer invasion and prostate cancer metastasis [18]. According to previous reports, ROBO1is expressed as a member of the immunoglobulin protein ROBO and its expression has been observed in primary tumors [19]. In cancer, FSTL1 may affect cancer cells by directly interacting with cells expressing TGFβ/BMP family receptors and DIP2A protein and indirectly by affecting the immune system [20]. Next, in this study top-downregulated genes mainly related to CD24, ZNF587B, PARP2, MIR4271, OCLN, TSPAN1, ADGRG2, MPZL2 and POP7 (Table 2). CD24 encodes a glycosyl phosphatidyl inositol and is most abundantly expressed in hematopoietic cells as well as has demonstrated its’ central role in tumor formation, progression and metastasis in prostate cancer [21]. ZNF587B is known as a member of the Krüppel-type zinc-finger proteins (KRAB-ZFPs) and has been shown to be associated with the regulating the proliferation of cell, cancer and apoptosis [22]. Accordingly, NF587B is a potential novel tumor suppressor for prostate cancer and may be a treatment target for this cancer. Occludin (OCLN) is a key tight junction protein which regulates remodeling of cytoskeletal and suppresses activation of the AKT/PI3K signaling pathway and proliferation of cell, thereby enhancing apoptosis of tumor in prostate cancer [23]. In the biological processes, GO enrichment in the upregulated genes mostly associated with the regulation of cell differentiation, signal transduction, cell death and process of apoptotic; and in the downregulated genes mainly belonged to response to chemicals, defense and immune system process, cell death and proliferation (Table 4). Moreover, meta-DEGs by KEGG pathways mostly associated with pathways in cancer, PI3K-Akt signaling pathway, interaction of cytokine-cytokine receptor, MAPK signaling pathway, transcriptional misregulation in cancer, MicroRNAs in cancer, apoptosis, p53 signaling pathway, prostate cancer and DNA replication. According to the KEGG analysis in prostate cancer, the DEGs were mostly related to cancer pathways, metabolic pathways, signaling pathway of PI3K‑Akt, Jak‑STAT signaling pathway, cancer proteoglycans and signaling pathway of NF‑κb [24]. In this study, the TFs associated with upregulated genes mostly identified as SRF, TBPs, POLR2A, TAF1 and RREB and downregulated genes were detected as IRFs, ZEB1, PRDM1, SPI1, JUNs, FOSs and NFKBs as well as common TFs in down and up genes identified as STATs, E2F1, MYB, FOXs, ZNFs, BCL6 (S Table 1). Activation of STAT3 has been identified in most of prostate cancer cell lines and has been shown to increase cell survival, tumor growth and apoptosis resistant [25]. Serum response factor (SRF) is a member of the TFs MADS-box family and is one of the important-known DNA-binding proteins in the human proteome. In addition, SRF plays a role in promoting cell proliferation, resistance to cell death and induction of metastasis and invasion [26]. Based on the present study, COL4A1, LOX, SMAD2, TIMP1 and FOS were identified as key hub genes of upregulated (Table 5). COL4A1 encodes the alpha1 chain of the Col IV gene, that is a dominant structural component of the microenvironment of tumor [27]. COL4A1 enhances invasion of tumor by inducing of tumor budding in various cancer cells and also involved in the migration and proliferation cancers such as to prostate cancer [28]. LOX is involved in the cross-linking of collagen and elastin as key enzymatic step and it is a copper-dependent amine oxidase [29]. The role of LOX in prostate cancer is to either enhance or suppress tumorigenesis, cell type, location and status of transformation [30]. The Small Mothers Against Decapentaplegic (SMAD) is an intracellular signal transduction protein related to the prognosis and growth of various types of tumors. SMADs have been shown to play a role in proliferation of cell, apoptosis, migration and regulation of cancer cells immune [31]. Based on our study, the hub genes of the upregulated DEGs mostly associated with collagen types and metastasis -related factors. Additionally, EXO1, RRM2, CDC45, KIAA0101, MCM10 and HMMR were detected as key hub genes of downregulated genes (Table 5). Increased expression of EXO1 has been reported to be related to invasion of tumor and metastasis. EXO1 overexpression is associated with poor survival in prostate cancer patients and increases progression of tumor and metastasis [32]. RRM2 may lead to genome instability and enhanced mutation, thereby affecting the progression of tumor [7] and overexpression of RRM2 plays an important role in the proliferation of cell, metastasis and dependence of drug in prostate cancer [33]. CDC45 plays an important role in the replication of the CDC45‑MCM‑GINS helicase holoenzyme and activates MCM2‑7, which leading to replication of DNA [34] and also, it inhibits replication of DNA and suppresses proliferation of cell in human cells [35]. Based on our study, the hub genes of downregulated genes mostly associated with DNA molecular processes and cancer-related factors. To validate our research, this study was compared with the reported qRT-PCR studies through the log2 fold change score of some upregulated genes including CLU and DDIT3 ([13], GSPT2 and CYBRD1 [14]) and downregulated genes including MPZL2 [15], FOXM1 [16], SCEL and TXNIP [14] (Figure 6). Cancer is initially caused by oncogenic changes in oncology, as well as by genes that are effective in tumor suppression [36]. Based on the reported results, no comprehensive report has been presented so far on the integration of microarray data from docetaxel-resistant prostate cancer cells for transcriptomic analyses. Accordingly, it was considered necessary to integrate and analyze data from different microarray experiments so that comprehensive results can be reported on the genes involved in docetaxel-resistant prostate cancer cells. According to this results, key upregulated consensus hub genes including LOX, SERPINE1, FOS and PTK2 and downregulated consensus hub genes including BRCA1, EGFR, IL6 and CXCL8 were identified. Based on KEGG pathways analysis, dominant biological pathways including cancer pathways, PI3K-Akt signaling pathway, cytokine-cytokine receptor interaction, MAPK signaling pathway, apoptosis, p53 signaling pathway were identified. The current confirmation is not sufficient for this research and for strong and robust validation in the future, molecular tests, including qRT-PCR are needed experimentally in the laboratory.

In conclusion, this study showed several hub genes associated with docetaxel resistance in prostate cancer cells by advanced bioinformatics analysis in Silico to understand the molecular mechanisms and can also suggested them as effective genes for treatment of docetaxel-resistance prostate cancer cells. To finally confirm this this idea, the key genes obtained need to be experimentally validated in the laboratory. Conflicts of Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.