CMI 03 (04), 363−371 CMI JOURNAL 363 Clinical Medicine Insights DOI: https://doi.org/10.52845/CMI/2023-4-1-1 CMI 04 (01), 363-371 (2023) ISSN (O) 2694-4626 RESEARCH ARTICLE Identification The Hub Genes In HL-60 Leukemia Cells On Decitabine Through Bioinformatics Analysis Fuxue Meng*, Lonakuan Li, Qin Zheng Medical Experiment Center, the Third Affiliated Hospital of Guizhou Medical University, Duyun 558000, Guizhou Province, China. Corresponding Author: Fuxue Meng Introduction Acute myeloid leukemia (AML) is a malignant clonal proliferative disease derived from hematopoietic stem cells, and its 5-year survival rate is very low. In the past few decades, the development of sequencing technology has resulted in the accumulation of a large number of omics data for various complex diseases. The subsequent development of bioinformatics revealed to us more rapid and intuitive multiple gene expression patterns [1;2] . Decitabine, as a specific inhibitor of DNA methyltransferase (DNMT), can be incorporated into DNA during the replication process, irreversibly or covalently bound to DNMT, depleting the storage of DNMT in the cell, and making DNA progressive demethylation. Decitabine can also enhance the body's anti-tumor immunity through non-methylated anti-tumor effects, stimulate the expression of tumor-associated antigens, or change the cellular immune status by regulating regulatory T cells; it can be mediated by P53 damage repair pathway promotes apoptosis of leukemia cells [3;4] . Study [5] have shown that compared with the Abstract Objective: Aims to identification the hub genes in HL-60 acute myeloid leukemia cells treated with decitabine, provide meaningful insights into the pathogenesis of AML. Methods: The gene expression profile of GSE24224 was obtained from GEO database, GO and pathway enrichment were performed though DAVID, established a PPI network from the STRING database and was displayed through the weight network diagram of omicShare tools. MiRNA targets and RBP targets prediction were carried out to further unravel the functions of hub genes identified. Results: 1558 differentially expressed genes were identified. GO and pathway analyses mainly involved citrulline metabolic process, positive regulation of neutrophil extravasation, positive regulation of cell adhesion molecule production, phospholipase A2 inhibitor activity, hematopoietic cell lineage, TGF-β signaling pathway, p53 signaling pathway. The miRNA and RBP targets prediction hinted that the 10 hub genes identified plays an important role in prognosis of AML. Conclusion: This study intimated that hub genes C3, C3AR1, FPR2, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 CXCR4 and FPR1 identified predictively targeted hsa-miR-520d-5p, hsa-miR-362-5p, hsa-miR-224-5p, hsa-miR-1913, hsa-miR-196b-5, hsa-miR-188-5p, hsa-miR-130a-5p, hsa-miR-204-5p, hsa-miR-211-5p, hsa-miR-671-5p, hsa-miR-296-3p and hsa-miR-23b-3p and ADAR1, DGCR8, DKC1, ELAVL1, FBL, FUS, HNRNPC, IGF2BP2, NOP58, TAF15, U2AF2 and UPF1 proteins may regulate the occurrence and development of AML. This study provides meaningful insights and ideas for further understanding the pathogenesis of AML. Key words: acute myeloid leukemia; hub genes; HL-60; bioinformatics analysis; decitabine Copyright : © 2021 The Authors. Published by Medical Editor and Educational Research Publishers Ltd. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/lic enses/by- nc-nd/4.0/). https://creativecommons.org/lic%20enses/by-nc-nd/4.0/ https://creativecommons.org/lic%20enses/by-nc-nd/4.0/ CMI 03 (04), 363−371 CMI JOURNAL 364 CMI JOURNAL Fuxue Meng et al. simulated control group, there are significant differences in gene expression in the HL-60 cellS group treated with decitabine, which plays a key role in the function of HL-60 cells, but the specific mechanism is still unclear. Therefore, this study utilized bioinformatics analysis methods to analyze the HL-60 cell gene chip that was acted by decitabine, in order to discover biomarkers related to the disease, and provide a theoretical basis for revealing the molecular mechanism of the occurrence and development of AML. Methods Microarray data collection and DEG screening The publicly accessible data GSE24224 was obtained from the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc =GSE24224) that deposited by Fabiani et al [5] . Containing six samples (three decitabine treated HL-60 cells and three mock treated HL-60 cells) were utilized in the present study. Gene expression profiling (GEP) was performed using the Affymetrix U133 Plus 2.0 expression array. All hybridization reactions were performed using GeneChip Fluidics Station 450, and GeneChips were scanned using the Affymetrix GeneChip Scanner 3000 (Affymetrix, USA). The raw data of GSE24224 was processed using the Affy package pair in R, using correction, normalization and log2 conversion [6] . The differentially expressed genes (DEGs) in decitabine treated HL-60 cells compared with mock treated HL-60 cells were determined using limma package [7;8] . DEGs were screened with a false discovery rate (FDR) corrected P<0.05 and |log fold-change (FC)|>1, then were confirmed using the GEO2R application from GEO. Functional enrichment analysis Investigation into the functions of enriched DEGs may improve understanding of their involvement in AML. In the present study, functional enrichment analysis of DEGs based on DAVID (https://david.ncifcrf.gov/tools.jsp), a widely used web-based genomic functional annotation tool. DEGs were subjected to molecular function and pathway studies by Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. P < 0.05 was set as cutoff values. Protein-protein interaction (PPI) network construction and Hub genes identification Genes are likely to function together rather than alone in complex diseases. Hub nodes in the network may represent key role. In the present study, protein-protein interaction (PPI) network analysis was performed to investigate the DEGs based on the STRING online database (https://string-db.org/cgi/input.pl). The minimum required interaction score was set as medium confidence > 0.8. In addition, the network was constructed through the weight network diagram of omicShare tools (https://www.omicshare.com/tools/index.php/) to confirm the hub genes and the genes with top-ten highest- weight nodes were defined as hub genes. miRNA target prediction In this study, the targets of the hub genes were predicted using four databases: miRWalk (http://mirwalk.umm.uni-heidelberg.de/), TargetScan (http://www. targetscan.org), miRMap (https://mirmap.ezlab.org) and starBase v2.0 (http://starbase.sysu.edu.cn/). The screening criterion was that the miRNA target exists in the four databases concurrently. The Venny 2.1 Online Tool (https://bioinfogp.cnb.csic.es/tools/venny/) was used to find overlapping genes between DEGs and predictive genes of DEMs. The miRNA-gene regulatory network was depicted and visualized using Cytoscape 3.7.1 (https://cytoscape.org/). RNA binding protein (RBP) target prediction As the core position of the post-transcriptional regulatory network, RNA binding protein (RBP) participates in multiple processes of RNA processing, including alternative splicing, RNA transport and stability maintenance, RNA localization, and mRNA translation [9] . In order to further understand the function of the hub gene, we used the RBP-mRNA module on the starBase v2.0 software to performed RBP prediction. Results Differentially expressed gene acquisition By the cutoff of a 1-fold change and P < 0.05, 1592 differentially expressed were identified filtered out. After removing the genes without gene symbol, 1558 differentially expressed genes remain, of which 1368 were up-regulated genes and 190 were down-regulated genes. The heatmap with clustering of differentially CMI 03 (04), 363−371 CMI JOURNAL 365 CMI JOURNAL Fuxue Meng et al. expressed gene were illustrated, and volcano plots were generated to demonstrate the distribution of the differentially expressed gene ( Figure 1 ). Figure 1 Heatmap and volcano plots showing the difffferential expression genes between decitabine treated HL-60 cells and group. A: The Heatmap of difffferential expression genes; B: the volcano plots difffferential expression genes (|log 2 FoldChange| ≧ 1 and P value < 0.05). GO and pathway analysis The GO functional enrichment resulted in DEGs mapped to 198 GO terms. With P<0.05 as the significant enrichment criterion, top ten of significant enriched functional clusters were screened. Of them, the DEGs GO enrichment mainly involved citrulline metabolic process, positive regulation of neutrophil extravasation, positive regulation of cell adhesion molecule production, phospholipase A2 inhibitor activity, MRF binding, phagolysosome, synaptonemal complex. KEGG enrichment analysis with P<0.05 was used as an enrichment screening standard, involved such as staphylococcus aureus infection, hematopoietic cell lineage, TGF-beta signaling pathway, p53 signaling pathway ( Figure 2 ) . Figure 2 GO and KEGG analysis of DEGs. A: KEGG enrichment histogram analysis of DEGs ; B: KEGG enrichment bubble chart of DEGs; C: GO enrichment histogram analysis of DEGs; D: GO enrichment bubble chart of DEG (P<0.05). PPI network analysis To screen and identification the hub genes in HL- 60 cells, we established a PPI network from the STRING database with scores of > 0.8. Then, the network was constructed through the weight CMI 03 (04), 363−371 CMI JOURNAL 366 CMI JOURNAL Fuxue Meng et al. network diagram of omicShare tools to confirm the hub genes and the genes with top-ten highest- weight node of DEGs such as C3, C3AR1, FPR2, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 CXCR4 and FPR1 were defined as hub genes ( Figure 3 ). Figure 3 PPI network construction and hub genes determination. B: PPI network of DEGs; A,C,D: the weight network of DEGs, C3, C3AR1, FPR2, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 CXCR4 and FPR1 identified as the hub genes (scores > 0.8 ). MiRNA target prediction For further understand the functions of the hub genes, the miRNA target prediction were carried out through four databases: miRWalk, TargetScan, miRMap and starBase v2.0. The C3 and FRP1 genes only have overlapping miRNA targets in three of the databases, and the remaining eight hub genes existed overlapping miRNA targets in the four databases, as the figure 4 shows. CMI 03 (04), 363−371 CMI JOURNAL 367 CMI JOURNAL Fuxue Meng et al. Figure 4 miRNA target prediction of hub genes. A, B: Predict miRNA targets the wenn map of C3 and FRP1 in three databases that predict miRNA targets; C: Eight hub genes for predicted miRNA targets network; D: Network of 7 hub genes with overlapping miRNA targets. RNA binding protein (RBP) target prediction RNA in the cell interacts with RBP to form a ribonucleoprotein (RNP) complex, which plays an important role in RNA synthesis, transport, stability, translation, and cellular localization. In order to further understand the function of hub genes, we carried out the RBP prediction. The results show that the top 10 RBP targets predicted by the 8 hub genes with the strongest weight were ELAVL1, FUS, DKC1, HNRNPC, ADAR, IGF2BP2, NOP58, TAF15, UPF1, U2AF2 (Figure 5). Figure 5 RBP targets prediction of hub genes. A, B: The networks of RBP targets. The larger point, darker the color, indicating the stronger the correlation and the greater the frequency of action; C: Screening of RBP targets with strong correlation. Discussion This study obtained gene expression profiles of HL-60 cells that decitabine induced and mock control group GSE24224 from GEO database and carried out DEGs screening, to understand the biological functions biological and the enrichment pathways involved through GO and KEGG analysis. Hereafter, PPI and weighted network analysis were conducted to identify the hub genes that play a key regulatory role in HL-60 cells. Further performed miRNA targets prediction and RBP targets prediction to clarify the possible mechanism of the hub gene for AML. In this study, we screened 1558 DEGs, of which 1368 were up-regulated and 190 were down- regulated. Based on GO enrichment analyses, it was found that DEGs mainly involved molecular functions including hematopoietic cell lineage, TGF-β signaling pathway, p53 signaling pathway. The TGF-β signaling pathway plays a negative role in the regulation of cell proliferation and differentiation in the hematopoietic system. When ZFYVE16 is overexpressed, the negative regulation of the TGF-β signaling pathway is enhanced, which may inhibit the malignant proliferation activated by FLT3 [10] . In addition, TGF-β stimulates the tumor pre-osteolytic and osteolytic factor production , Thereby stimulating further bone absorption [11;12] . This classifies TGF- β as as an important factor responsible for the feedforward vicious circle that drives tumor growth in the bones. In addition, abnormal reactivation of TGF-β usually leads to carcinogenic behavior [13;14] . Their role in tumorigenesis usually reflects their role in embryonic development, but also extends to other features often observed in cancer, such as cachexia and bone loss [15] . In the stage of prostate tumorigenesis, increased TGF-β production leads to the degradation of extracellular matrix, immunosuppression and angiogenesis, all of which lead to escape cell death and increase cell survival rate, which is conducive to the reproduction of cancer cells [16] . The regulation of p53 gene activity is mainly at the post- transcriptional level. After phosphorylation of p53 protein is activated, it becomes p-p53, which then CMI 03 (04), 363−371 CMI JOURNAL 368 CMI JOURNAL Fuxue Meng et al. acts by inducing cell cycle arrest and inducing apoptosis to inhibit tumor cell proliferation [17;18] . After the above research and analysis, PPI networks were constructed by STRING, and from the intersection, 10 hub genes were obtained, which were verified and show with weighted network analysis. The analytic results of C3, C3AR1, FPR2, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 CXCR4 and FPR1 were statistically significant, which suggested that these ten genes possibly play a key regulatory role in AML. In order to further understand the functions of ten hub genes, we performed miRNA targets prediction and RBP targets prediction. miRNA is an endogenous single-stranded small RNA with a length of 21-25 bases, which can regulate the expression of target genes by specifically binding to mRNA to degrade or inhibit protein translation of mRNA. Studies have shown that miRNAs are often located in tumorigenesis-related regions or fragile sites, amplification regions, heterozygous loss regions or breakpoint regions. The specificity of miRNA expression in tumor cells is specific to tumor occurrence, development, invasion, metastasis, and metastasis. It plays an important role in the prognosis and other processes, providing potential therapeutic targets or new strategies [19;20] . In this study, we selected 7 hub genes whose miRNA prediction targets were hsa- miR-520d-5p, hsa-miR-362-5p, hsa-miR-224-5p, hsa-miR-1913, hsa-miR-196b-5，hsa-miR-188- 5p, hsa-miR-130a-5p, hsa-miR-204-5p, hsa-miR- 211-5p, hsa-miR-671-5p, hsa-miR-296-3p and hsa-miR-23b-3p. Li [21] showed that hsa-miR-362- 3p was highly expressed as a marker in AML, and suggesting that it is related to the poor prognosis of AML patients. The prediction result of hsa- miR-1913 in AML is consistent with the result of Wang [22] . RBP refers to the general term for proteins that directly bind to RNA. RNA in cells interacts with RBP to form a ribonucleoprotein (RNP) complex, which plays an important role in RNA synthesis, transport, stability, translation, and cell positioning [23] . In the current study, the hub genes C3, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 and CXCR4 were predicted to be strongly associated with proteins ADAR1, DGCR8, DKC1, ELAVL1, FBL, FUS, HNRNPC, IGF2BP2, NOP58, TAF15, U2AF2 and UPF1. Interestingly, Xiao [24] and Peng [25] reported that ADAR1 plays an important role in acute myeloid leukemia. Among them, Xiao’s study demonstrate that ADAR1 may be involved in the regulation of the proliferation of AML cells partially via regulation of the Wnt signaling pathway [24] . He et al reported that IGF2BP2 overexpression indicates poor survival in patients with AML, IGF2BP2 may serve as a biomarker to predict the prognosis of AML and as a potential target in AML [26] . Conclusion In conclusion, the present study identified a panel of 10 genes in AML HL-60 cells. Gene function and pathway investigation indicated that these genes were mainly engaged in citrulline metabolic process, positive regulation of neutrophil extravasation, positive regulation of cell adhesion molecule production, phospholipase A2 inhibitor activity, hematopoietic cell lineage, TGF-β signaling pathway, p53 signaling pathway. Moreover, the miRNA and RBP targets prediction hinted that the 10 hub genes identified plays an important role in prognosis of AML. Hub genes C3, C3AR1, FPR2, GNA11, PTAFR, ITGAM, ANXA1, LPAR1 CXCR4 and FPR1 predictively targeted hsa-miR-520d-5p, hsa-miR-362-5p, hsa- miR-224-5p, hsa-miR-1913, hsa-miR-196b-5, hsa-miR-188-5p, hsa-miR-130a-5p, hsa-miR-204- 5p, hsa-miR-211-5p, hsa-miR-671-5p, hsa-miR- 296-3p and hsa-miR-23b-3p and ADAR1, DGCR8, DKC1, ELAVL1, FBL, FUS, HNRNPC, IGF2BP2, NOP58, TAF15, U2AF2 and UPF1 proteins may regulate the occurrence and development of AML, but further verification is needed. In short, this study provides meaningful insights and ideas for further understanding the pathogenesis of AML. Abbreviation AML Acute myeloid leukemia DNMT DNA methyltransferase DEGs Differentially expressed genes RBP RNA binding protein C3AR1 Complement component 3a receptor 1 FPR2 Formyl peptide receptor 1 CMI 03 (04), 363−371 CMI JOURNAL 369 CMI JOURNAL Fuxue Meng et al. GNA11 G protein subunit alpha 11 PTAFR Platelet activating factor receptor ITGAM Integrin subunit alpha M ANXA1 Annexin A1 LPAR1 Lysophosphatidic acid receptor 1 CXCR4 C-X-C motif chemokine receptor 4 FPR1 Formyl peptide receptor 1 ELAVL1 ELAV like RNA binding protein 1 FUS FUS RNA binding protein DKC1 Dyskerin pseudouridine synthase 1 HNRNPC Heterogeneous nuclear ribonucleoprotein C ADAR Adenosine deaminase, RNA specific IGF2BP2 Insulin like growth factor 2 mRNA binding protein 2 NOP58 NOP58 ribonucleoprotein TAF15 TATA-box binding protein associated factor 15 UPF1 UPF1, RNA helicase and ATPase U2AF2 U2 small nuclear RNA auxiliary factor 2 Acknowledgements We acknowledge GEO database for providing their platforms and contributors for uploading their meaningful datasets. Simultaneously, we also appreciate to the OmicShare tool that online data analysis platform. Authors contributions The design, study conduct, and financial support for this research were provided by Meng FX, Li LK and Zheng Q participated in the interpretation of data, review, and approval of the publication. Funding This study was supported by the Guizhou Provincial Health Commission Science and Technology Project (gzwjkj2019-2-006) and the Qiannan Prefecture Science and Technology Project (Qiannankeheshezi 2018 (36)). Availability of data and materials The original data of this research come from the third platform GEO database (GSE24224), and bioinformatics analysis was carried out through relevant analysis tools. Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. References 1. Pelcovits A, Niroula R. Acute Myeloid Leukemia: A Review. R I Med J. 2020, 103(3): 38-40. 2. Dombret H, Gardin C. An update of current treatments for adult acute myeloid leukemia. Blood, 2016, 127(1) : 53-61. 3. Park JW, Han JW. Targeting epigenetics for cancer therapy. Arch Pharm Res. 2019, 42(2): 159-170. 4. Geyer KK, Munshi SE, Vickers M, Squance M, Wilkinson TJ, Berrar D, Chaparro C, Swain MT, Hoffmann KF. The anti-fecundity effect of 5-azacytidine (5-AzaC) on Schistosoma mansoni is linked to dis- regulated transcription, translation and stem cell activities. Int J Parasitol Drugs Drug Resist. 2018, 8(2): 213-222. 5. Fabiani E, Leone G, Giachelia M, D'alo' F, Greco M, Criscuolo M, Guidi F, Rutella S, Hohaus S, Voso MT. Analysis of genome- wide methylation and gene expression induced by 5-aza-2'-deoxycytidine identifies BCL2L10 as a frequent methylation target in acute myeloid leukemia. Leuk Lymphoma. 2010, 51(12): 2275-2284. 6. Gautier L, Cope L, Bolstad BM and Irizarry RA: Affy-analysis of Affymetrix Gene Chip data at the probe level. Bioinformatics, 2004, 20: 307-315. 7. Mao Y, Xue P, Li L, Xu P, Cai Y, Chu X, Jiang P, Zhu S. Bioinformatics analysis of mRNA and miRNA microarray to identify the key miRNA-gene pairs in small-cell lung cancer. Mol Med Rep. 2019, 20(3): 2199- 2208. 8. Diboun I, Wernisch L, Orengo CA and Koltzenburg M: Microarray analysis after CMI 03 (04), 363−371 CMI JOURNAL 370 CMI JOURNAL Fuxue Meng et al. RNA amplification can detect pronounced differences in gene expression using limma. BMC Genomics, 2006, 7: 252-260. 9. Li D, Kishta MS, Wang J. Regulation of pluripotency and reprogramming by RNA binding proteins. Curr Top Dev Biol. 2020,138:113-138. 10. Zhao XM, Research on the pathological and physiological mechanisms of Zfyve16 in the hematopoietic system [D]. Shanghai Jiaotong University, 2016. 11. Zi Z. Molecular Engineering of the TGF-β Signaling Pathway. J Mol Biol. 2019, 431(15): 2644-2654. 12. Haque S, Morris JC. Transforming growth factor-β: A therapeutic target for cancer. Hum Vaccin Immunother. 2017, 13(8): 1741-1750. 13. Morén A, Bellomo C, Tsubakihara Y, Kardassis D, Mikulits W, Heldin CH, Moustakas A. LXRα limits TGFβ-dependent hepatocellular carcinoma associated fibroblast differentiation. Oncogenesis. 2019, 8(6): 36- 42. 14. Chobert MN, Couchie D, Fourcot A, Zafrani ES, Laperche Y, Mavier P, Brouillet A. Liver precursor cells increase hepatic fibrosis induced by chronic carbon tetrachloride intoxication in rats. Lab Invest. 2012, 92(1):135-150. 15. Fields SZ, Parshad S, Anne M, Raftopoulos H, Alexander MJ, Sherman ML, Laadem A, Sung V, Terpos E. Activin receptor antagonists for cancer-related anemia and bone disease. Expert Opin Investig Drugs. 2013, 22(1): 87- 101. 16. Zhao M, Mishra L, Deng CX. The role of TGF-β/SMAD4 signaling in cancer. Int J Biol Sci. 201, 14(2):111-123. 17. Kanapathipillai M. Treating p53 Mutant Aggregation-Associated Cancer. Cancers (Basel). 2018, (6):154-163. 18. Lang Y, Yu C, Tang J, Li G, Bai R. Characterization of porcine p53 and its regulation by porcine Mdm2. Gene. 2020, 748:144699-144709. 19. Zheng X, Chen L, Li X, Zhang Y, Xu S, Huang X. Prediction of miRNA targets by learning from interaction sequences. PLoS One. 2020, 15(5): e0232578. 20. Fridrich A, Hazan Y, Moran Y. Too Many False Targets for MicroRNAs: Challenges and Pitfalls in Prediction of miRNA Targets and Their Gene Ontology in Model and Non- model Organisms. Bioessays. 2019, 41(4): e1800169. 21. Li G, Gao Y, Li K, Lin A, Jiang Z. Genomic analysis of biomarkers related to the prognosis of acute myeloid leukemia. Oncol Lett. 2020, 20(2): 1824-1834. 22. Wang SM, Zeng WX, Wu WS, Sun LL, Yan D. Association between a microRNA binding site polymorphism in SLCO1A2 and the risk of delayed methotrexate elimination in Chinese children with acute lymphoblastic leukemia. Leuk Res. 2018, 65:61-66. 23. Zhang Z, Yao Z, Wang L, Ding H, Shao J, Chen A, Zhang F, Zheng S. Activation of ferritinophagy is required for the RNA- binding protein ELAVL1/HuR to regulate ferroptosis in hepatic stellate cells. Autophagy. 2018, 14(12): 2083-2103. 24. Xiao H, Cheng Q, Wu X, Tang Y, Liu J, Li X. ADAR1 may be involved in the proliferation of acute myeloid leukemia cells via regulation of the Wnt pathway. Cancer Manag Res. 2019, 11:8547-8555. 25. Peng L, Yang X, Zhang Y, Hu T, Wang W, Wang X, Xu J, Cheng T, Yuan W, Gao Y. Effect of ADAR1 on the development of MLL-AF9 induced murine AML. Zhonghua Xue Ye Xue Za Zhi. 2015, 36(5): 383-388. 26. He X, Li W, Liang X, Zhu X, Zhang L, Huang Y, Yu T, Li S, Chen Z. IGF2BP2 Overexpression Indicates Poor Survival in Patients with Acute Myelocytic Leukemia. Cell Physiol Biochem. 2018, 51(4): 1945- 1956. CMI 03 (04), 363−371 CMI JOURNAL 371 CMI JOURNAL Fuxue Meng et al.