Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Overexpression of ZEB2-AS1 lncRNA is associated with poor clinical outcomes in acute myeloid leukemia

Overexpression of ZEB2-AS1 lncRNA is associated with poor clinical outcomes in acute myeloid... ONCOLOGY LETTERS 17: 4935-4947, 2019 Overexpression of ZEB2‑AS1 lncRNA is associated with poor clinical outcomes in acute myeloid leukemia 1-3* 4* 1 1 1 1 XIAOLAN SHI , JIAO LI , LIANG MA , LIJUN WEN , QINRONG WANG , HONG YAO , 1,2 1-3 5 1,2 CHANGGENG RUAN , DEPEI WU , XINYOU ZHANG and SUNING CHEN Department of Hematology, Jiangsu Institute of Hematology, Key Laboratory of Thrombosis and Hemostasis of Ministry of Health, The First Affiliated Hospital of Soochow University; 2 3 Collaborative Innovation Center of Hematology, Institute of Blood and Marrow Transplantation, Soochow University, Suzhou, Jiangsu 215006; Department of Hematology, Yixing People's Hospital of Jiangsu Province, Yixing, Jiangsu 214200; Department of Hematology, Shenzhen People's Hospital, The Second Clinical Medical College of Jinan University, Shenzhen, Guangdong 518020, P.R. China Received January 10, 2018; Accepted January 28, 2019 DOI: 10.3892/ol.2019.10149 Abstract. Acute myeloid leukemia (AML) is a fatal hemato- compared with patients with a lower expression of ZEB2-AS1 poietic malignancy with poor clinical outcomes. To determine lncRNA (P= 0.031). In cases with low levels of ZEB2-AS1 whether the expression of the long non-coding (lnc)RNA zinc lncRNA, patients treated with allogenic hematopoietic stem finger E-box binding homeobox 2 (ZEB2) antisense RNA 1 cell transplantation had signic fi antly longer OS and DFS rates (ZEB2-AS1) is associated with clinical outcomes, its expres- compared with that of chemotherapy-treated patients (P=0.037 sion was analyzed in a retrospective cohort of 62 AML and 10 and P=0.049 respectively). Furthermore, the knockdown of non-malignant cases. The results revealed that the expression ZEB2-AS1 lncRNA effectively inhibited AML cell invasion and of ZEB2-AS1 lncRNA was notably high and closely associated migration, which was closely associated with the downregulation with adverse clinical outcomes in AML cases compared with the of ZEB2 and upregulation of E-cadherin expression. Collectively, non‑malignant cases, based on either modie fi d Medical Research although its independent prognostic value for survival was not Council or European Leukemia Net risk stratic fi ation systems. rigorously determined, ZEB2-AS1 lncRNA may function as a Univariate analyses indicated that patients with a higher expres- candidate marker to improve conventional risk stratification sion of ZEB2-AS1 lncRNA had significantly shorter overall systems and the evaluation of therapeutic responses for AML. survival (OS) (P=0.036) and disease-free survival (DFS) rates (P=0.039) compared with patients with a lower expression of Introduction ZEB2-AS1 lncRNA. In addition, patients with a higher expres- sion of ZEB2‑AS1 lncRNA had a signic fi ant lower complete Acute myeloid leukemia (AML) is predominantly a fatal hema- remission rate in response to induction by chemotherapy topoietic malignancy characterized by the clonal proliferation of myeloid blasts with tissue infiltration (1). It may occur at any age, with an incidence of 2-3/100,000 per annum in children <14 years old, and ~15/100,000 per annum in adults >60 years old globally (2). Despite advances in therapeutic strategies, Correspondence to: Professor Suning Chen, Department of including intensive chemotherapy and hematopoietic stem cell Hematology, Jiangsu Institute of Hematology, Key Laboratory transplantation (HSCT), the clinical outcome of AML remains of Thrombosis and Hemostasis of Ministry of Health, The First poor, particularly in older patients (>60 years old) (3-5). Affiliated Hospital of Soochow University, 188 Shizi Road, Suzhou, Considering the clonal complexity of AML, there has been Jiangsu 215006, P.R. China increasing interest in improving the prognosis and treatment E-mail: chensuning@sina.com of AML through the more extensive biological profiling of Professor Xinyou Zhang, Department of Hematology, Shenzhen cytogenetic and molecular tumor heterogeneity (6-10). The People's Hospital, The Second Clinical Medical College of Cancer Genome Atlas Research Network has reported that Jinan University, 1017 Dongmen North Road, Shenzhen, ~70% of AML cases have mutations in genes encoding epigen- Guangdong 518020, P.R. China etic modier fi s (11,12). Notably, novel data has demonstrated E-mail: zxy0518@live.cn that DNA methylation heterogeneity (epialleles) may occur with distinct kinetics and patterns that are likely to affect Contributed equally clinical outcomes. These may be hallmarks of AML and may Key words: acute myeloid leukemia, long non-coding RNA, zinc be independent of the genetic landscape (13,14). Accordingly, finger E-box binding homeobox 2 antisense RNA 1, prognosis differences in epigenetic diversity may function as molecular biomarkers to independently evaluate AML prognosis. SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Long non-coding RNAs (lncRNAs) are a class of RNAs Medical Research Council (MRC) or European Leukemia that are >200 nucleotides in length (15-17). Gene expression Net (ELN) recommendations were applied for risk stratic fi a - regulated by lncRNAs is regarded as one of the most notable tion (37-39). The 62 bone marrow specimens were collected types of epigenetic control (18,19). Recurrent mutations and/or from the pretreated patients and were frozen and archived for epigenetic alterations in the regulatory non-coding genome the following experiments. For comparison, 10 eligible bone may broadly affect lncRNA expression in numerous malignant marrow specimens were collected from patients without hema- tumor types, serving as signals for carcinogenesis, in addition topoietic malignancies and were selected as the non-malignant to providing information for prognosis and therapeutic options hemotopathy group (Table II). Clinical outcome data were in patients with cancer (20). Notably, the expression of a small updated as of April 2016. The present research was ethi- subset of lncRNAs, including nuclear paraspeckle assembly cally approved by the Institutional Review Board of the First transcript 1, have been strongly associated with treatment Affiliated Hospital of Soochow University. response and survival in cytogenetically normal older patients with AML (21). In particular, a range of lncRNAs, including AML cell line. Using the Affymetrix Human LncRNA HOXA transcript antisense RNA, myeloid‑specic fi 1 and HOX microarray analysis, it has been demonstrated that ZEB2-AS1 transcript antisense intergenic RNA myeloid 1 (HOTAIR), lncRNA is predominantly overexpressed in patients with may exert pivotal effects not only on hematopoietic stem cells AML with a karyotype of 11q23. In addition, the expression during normal hematopoiesis, but also on the cancer pheno- of ZEB2-AS1 lncRNA in THP-1 cells, also with a karyo- type during the process of leukemogenesis (22-25). type of 11q23, is signic fi antly higher compared with that of lncRNAs are categorized into antisense, bidirectional, other AML cell lines, including AP1060, NB4 and FKH-1 intronic, intergenic and overlapping lncRNAs, based on their (P=0.0020). Thus, THP-1 cells were selected for the present chromosomal location (26,27). Antisense lncRNAs are initially study and cultured in RPMI-1640 medium (GE Healthcare transcribed from the opposite strand of a protein-coding Life Sciences, Hyclone, Logan, UT, USA) supplemented with counterpart, functioning as fast regulatory mediators in 10% heat‑inactivated fetal bovine serum (FBS; Sigma‑Aldrich; self-regulatory circuits to modulate global and/or specific Merck KGaA, Darmstadt, Germany) in humidified 37˚C transcriptional outputs (28-32). Certain antisense lncRNAs, incubator containing 5% CO . including IGF1R antisense imprinted non-protein coding RNA, are downregulated in patients with high-risk AML, resulting Patient treatment. A total of 39 patients with de novo AML, in the promotion of cell growth through long-range chromatin excluding those diagnosed as the M3 subtype, received interactions with insulin like growth factor 1 receptor (33,34). front-line induction chemotherapy, including the idarubicin Using the Affymetrix Human LncRNA microarray analysis, it and cytarabine regimen, as follows: Idarubicin 8-12 mg/m was has been demonstrated that the expression of the lncRNA (days 1-3) and cytarabine 100 mg/m (days 1‑7); or the dauno - zinc finger E‑box binding homeobox 2 (ZEB2) antisense RNA 1 rubicin and cytarabine regimen, as follows: Daunorubicin 2 2 (ZEB2-AS1) is abnormally overexpressed in patients with AML 60-90 mg/m (days 1-3) and cytarabine 100 mg/m (days 1-7). (as yet unpublished). A previous study indicated that a natural Subsequent to achieving first complete remission (CR, n=24), antisense transcript, overlapping the 5' splice site in the intron patients received post-remission therapy of either several of the ZEB2 gene, may prevent splicing of the 5'-untranslated consolidation courses (n=12) or allogeneic HSCT (allo‑HSCT; region to increase ZEB2 translation and consequently down- n=12). regulate the expression of E-cadherin, which in turn induces For the treatment of AML with allo-HSCT, patients epithelial-mesenchymal transition (EMT) in a tumor (35). received an initial conditioning regimen with lomustine 2 2 However, the prognostic value of ZEB2-AS1 in AML and its (250 mg/m /day on day -10), cytarabine (2 or 4 g/m /day; days function in leukemogenesis remains to be elucidated. ‑9 to ‑8), busulfan (3.2 mg/kg/day; days ‑7 to ‑5) and cyclo- In the present study, 62 de novo patients with AML were phosphamide (1.8 g/m /day; days ‑4 to ‑3). Due to advanced retrospectively analyzed to determine if ZEB2-AS1 lncRNA patient age or the presence of other comorbidities, patients may function as a biomarker to evaluate AML prognosis. with poor responses to myeloablative conditioning received Thus, the specic fi aim of the present study was to assess the a regimen with lomustine (250 mg/m /day; day ‑10), fluda - 2 2 association between ZEB2-AS1 lncRNA expression and the rabine (30 mg/m ; days ‑10 to ‑6), cytarabine (1.5 g/m /day; clinical features of patients with AML. Additionally, the days ‑10 and ‑6) and busulfan (3.2 mg/kg/day; days ‑5 to ‑3). potential regulation of leukemic phenotypes by ZEB2-AS1 To effectively prevent graft-versus-host disease (GVHD), lncRNA was investigated. As such, the clinical and biological cyclosporine (3 mg/kg/day) was infused to achieve a target importance of ZEB2-AS1 lncRNA were evaluated. blood concentration between 200-300 ng/ml, starting on day -9 or -1 until patients switched to oral administration. For Materials and methods unrelated or haploidentical transplantation, mycophenolate mofetil (30 mg/kg/day) and rabbit anti-thymocyte globulin Patients and tissue specimens. A total of 62 eligible patients (2.5 mg/kg/day; days ‑5 to ‑2) were additionally administered with de novo AML were enrolled retrospectively in the present to prevent GVHD. In addition, methotrexate was separately study. Patients were diagnosed and classie fi d according to the administered on days +1, +3, +6, and +11, at doses of 15, 10, 10 World Health Organization (36) criteria at the First Affiliated and 10 mg/m , respectively. Hospital of Soochow University (Jiangsu, China) between May 2007 and June 2014. The clinicopathological charac- Cytogenetic and molecular genetic analysis. In the cytogenetic teristics of this cohort are summarized in Table I. Modie fi d analyses, bone marrow specimens from patients with de novo ONCOLOGY LETTERS 17: 4935-4947, 2019 Table I. Clinical, pathological and genetic characteristics of patients with AML. ZEB2-AS1 expression --------------------------------------------------------------------------------- Characteristics Patients Low level High level P-value Age (years) 0.552 Median (range) 39 (8-80) 39 (8-80) 34 (14-67) Sex 0.537 Male 32 (51.6%) 24 8 Female 30 (48.4%) 25 5 FAB Subtypes 0.006 M0 1 (1.6%) 0 1 M1 1 (1.6%) 1 0 M2 27 (43.6%) 25 2 M3 9 (14.5%) 9 0 M4 10 (16.1%) 5 5 M5 14 (22.6%) 9 5 Karyotype <0.001 Normal karyotype 16 (25.8%) 14 2 t(15;17) 9 (14.5%) 9 0 t(8;21) 13 (21.0%) 12 1 inv (16) 7 (11.3%) 5 2 t(6;9) 5 (8.0%) 5 0 11q23 8 (12.9%) 4 4 Complex karyotype 4 (6.5%) 0 4 White blood cell (x10 /l; non‑M3) 0.046 Median (range) 24.3 (1.0-190.5) 13.5 (1.0-140.2) 52.1 (1.3-190.3) Hemoglobin (g/l; non‑M3) 0.372 Median (range) 84.0 (37.0-149.0) 84.0 (37.0-149.0) 88.0 (38.0-116.0) Platelets (x10 /l; non‑M3) 0.044 Median (range) 40 (8.0-414.0) 31 (8-414) 71.5 (20-410) Blasts in bone marrow (%; non‑M3) 0.569 Median (range) 56.5 (20.5-98.0) 57 (20.5-95.5) 56.5 (25.0-98.0) Mutated gene (non-M3) 0.474 Negative 15 12 3 CEBPA 2 2 0 NPM1 1 1 0 FLT3-ITD 1 1 0 FLT3-TKD 1 1 0 DNMT3A 1 1 0 C-kit 4 1 3 C-kit/CEBPA 2 2 0 NPM1/FLT3-TKD 1 1 0 FLT3-ITD/CEBPA 1 1 0 NPM1/DNMT3A 1 1 0 DNMT3A/NPM1/FLT3-ITD 2 1 1 Modified MRC risk stratification 0.002 Favorable 29 (46.8%) 26 3 Intermediate 28 (45.1%) 22 6 Adverse 5 (8.1%) 1 4 ELN risk stratification (non‑M3) 0.028 Favorable 27 24 3 Intermediate I and II 13 8 5 Adverse 10 5 5 SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Table I. Continued. ZEB2-AS1 expression ----------------------------------------------------------- Characteristics Patients Low level High level P-value Recovery from induction chemotherapy (non-M3) 0.031 CR 24 21 3 Non-CR 15 7 8 AML, acute myeloid leukemia; FAB, French‑American‑British; MRC, Medical Research Council; ELN, European Leukemia Net; CR, complete remission; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; CEBPA, CCAAT enhancer binding protein α; NPM1, nucleophosmin 1; FLT3, fms related tyrosine kinase 3; ITD, internal tandem duplication; TKD, tyrosine kinase domain; DNMT3A, DNA methyltransferase 3α. Table II. Characteristics of 10 non-malignant hemotopathy CCAAT enhancer binding protein α (CEBPA, forward, 5'-GGC cases. G AG C AG GG T C T C C GG G T -3 ' and reverse, 5'-TGT GCT GGA A CA GGT C GG C CA- 3') and nucleophosmin 1 (NPM1, Sex Number Age Number fo r wa r d , 5 '- T TA AC T C T C T G G T G G TAG A AT GA A-3 ' and reverse, 5'-TGT TAC AGA AAT GAA ATA AGA CGG-3'). Male 3 ≥60 2 Female 7 <60 8 RNA extraction and reverse transcription‑quantitative (RT‑q) WBC 5.585 (2.6-9.18)x10/l Diagnosed Number PCR. Total RNA was extracted from patient bone marrow mononuclear cells using TRIzol (Invitrogen; Thermo Fisher HGB 74.5 (57-156)g/l IDA 8 Scientific, Inc.). RT and first strand cDNA synthesis was PLT 281 (10-324)x10 /l ITP 2 subsequently performed using MMLV-RT reverse transcrip- WBC, white blood cell count; HGB, hemoglobin; PLT, platelet count. tase (Promega Corporation, Madison, WI, USA 37˚C 60 min, 95˚C 5 min). RT-qPCR analysis was employed to detect levels of ZEB2-AS1. GAPDH was used as an internal reference gene. The primer sequences used were as follows: ZEB2-AS1 AML were processed in standard un-stimulated cultures for forward, 5'-GGC TGG ATA GCA AAG GAC-3' and reverse, 24 h. With standard techniques of ISCN 2016 (40) for chro- 5'-ACA CTC TTG GCG AGG T‑3'; ZEB2 forward, 5'‑GTC CAT mosome R‑banding and fluorescence in situ hybridization, GCG AAC TGC CAT CT-3' and reverse, 5'-ATC TGT CCC TGG the different karyotypes in patients with AML were routinely CTT GTG TG‑3'; E‑cadherin forward, 5'‑TGC CCA GAA AAT determined. If available, at least 20 metaphases were analyzed GAA AAA GG-3' and reverse, 5'-GTG TAT GTG GCA ATG for every bone marrow sample. For analyzing mutations in CGT TC‑3'; GAPDH forward, 5'‑CAA GGT CAT CCA TGA patients with de novo AML, a Purelink™ Genomic DNA CAA CTT TG-3', and reverse, 5'-GTC CAC CAC CCT GTT mini kit (Invitrogen; Thermo Fisher Scientic fi , Inc., Waltham, GCT GTA G-3'. SYBR Green (Taraka, Japan) RT-qPCR was MA, USA) was used to extract genomic DNA from the performed and the relative threshold cycle value normalized to 62 patients' bone marrow mononuclear cells, according to the the reference GAPDH gene was obtained (ABI 7500; Thermo manufacturer's protocol. The coding regions of mutated genes Fisher Scientic fi , Inc.) The thermo cycling conditions were as were either partially or entirely amplie fi d using a polymerase follows 50˚C 2 min, 95˚C 10 min, 95˚C 15 sec, 60˚C 1 min, - Cq ΔΔ chain reaction (PCR) in order to identify these mutations. for a total of 40 cycles. Following this, 2 was calculated The genomic DNA was extracted from the 62 patients' bone to determine relative abundance of target gene expression marrow mononuclear cells. The thermo cycling conditions between the groups (41). were as follows 95˚C 5 min, total 35 cycles of 95˚C 30 sec and 58˚C 30 sec and 72˚C 1 min, then 72˚C 10 min. Direct bidi- RNA interference. Gene-specific small interfering RNAs rectional DNA sequencing was subsequently performed. In (siRNAs) against ZEB2‑AS1 (siZEB2‑AS1; sense, 5'‑CAC the present study, a range of acute leukemia-associated muta-CUU UGG UUA CCU GAA UTT-3' and antisense, 5'-AUU CAG tions were evaluated, including fms related tyrosine kinase 3 GUA ACC AAA GGU GTT-3') and negative control (NC) siRNA (FLT3)-internal tandem duplication (FLT3-ITD, forward, (sense, 5'-UUC UCC GAA CGU GUC ACG UTT-3' and antisense, 5 '- C A A T T T AG G TAT GA A AG C C -3 ' a n d r ever s e, 5 '- G TA 5 ' - A CG UG A C A C G U U CG G A G A A T T - 3 ') w er e c om mercia l l y C C T TT C A G C A TT TT G A C - 3'), DNA methyltransferase 3α designed (Shanghai GenePharma Co., Ltd., Shanghai, China). (DNMT3A, forward, 5'-CTG CTG TGT GGT TAG ACG-3' and On the day of transfection, THP-1 cells were plated at a low reverse, 5'-TAT TTC CGC CTC TGT GGT TT-3'), FLT3-tyrosine density of 2 x 10 on the culture vessel (Corning, Corning, kinase domain (FLT3-TKD, forward, 5'-CCA GGA ACG TGC NY, USA) and subsequently transfected with 40 nM on-target T T G T CA-3 ' a n d r eve r s e, 5 '- T CA A A A AT G CAC CAC AG T siRNA using Lipofectamine™ 2000 (Invitrogen; Thermo GAG-3'), C-kit (forward, 5'-CTC CCT GAA AGC AGA AAC-3' Fisher Scientic fi , Inc.) according to the manufacturer's protocol a n d r e v e r s e, 5 ' - C AG AAA G AT AAC AC C AAA ATA G -3 ' ), cells were incubated with siRNA for 24-48 h. NC siRNA was ONCOLOGY LETTERS 17: 4935-4947, 2019 used as a transfection control in all experiments. Each experi- was considered to indicate a statistically signic fi ant difference. ment was independently repeated at least three times. SPSS statistical software version 18 (SPSS, Inc., Chicago, IL, USA) was used to perform all statistical analyses. Analyses of biological phenotype. To analyze cell migra- tion, 1x10 THP-1 cells (siZEB2-AS1 and NC groups) were Results plated into the upper chamber of Transwell cell culture inserts (24‑well; pore size, 8 µm; Corning) in serum‑free DMEM Clinical, cytogenetic and molecular features of patients with media (GE Healthcare Life Sciences). The lower chamber AML. The clinical features of the 62 AML cases enrolled in medium was supplemented with 20% FBS. To assess cell the present study were summarized in Table I. Based on the invasion, 1x10 THP-1 cells (siZEB2-AS1 and NC groups) modie fi d MRC classic fi ation, patients were categorized into a were seeded into the upper chamber of Transwell cell culture favorable risk group (n=29), an intermediate risk group (n=28) inser ts (24‑well; pore size, 8 µm; Corning) coated with and an adverse risk group (n=5). According to ELN recom- Matrigel. The lower chamber medium contained 20% FBS. mendations (12 cases missed the required mutation data and Following incubation at 37˚C for 24 h, the upper layer of were not classie fi d), patients were categorized into a favorable THP-1 cells was removed with cotton wool, and THP-1 cells risk group (n=27), an intermediate I/II risk group (n=13) and on the lower surface were fixed with 95% ethanol for 20 min an adverse risk group (n=10; Table I). The respective values of in room temperature. Invaded or migrated cells were subse- median OS and DFS rates, regarding different risk tiers and quently stained with 0.1% crystal violet for 30 min at 37˚C and treatment approaches, were summarized in Table III. observed under an IX71 inverted microscope at x200 magnifi - cation (Olympus Corporation, Tokyo, Japan). Five microscopic Expression of ZEB2‑AS1 lncRNA in patients with AML. e fi lds were counted per insert. Triplicate inserts were used for Using the Affymetrix Human LncRNA microarray, dozens each individual experiment, and each experiment was inde- of abnormally expressed lncRNAs were identie fi d in patients pendently repeated at least three times. with AML. ZEB2‑AS1 lncRNA was identie fi d to be substan - To assess cell proliferation, THP-1 cell lines (siZEB2-AS1 tially overexpressed in patients with AML with a karyotype of and NC groups) were seeded onto 96-well cell culture cluster 11q23 when compared with karyotypes such as t(15;17), t(8;21) plates (Corning) at a concentration of 5x10 cells/well in volumes and inv (16). Therefore, the ZEB-AS1 lncRNA was selected of 100 µl. A total of 10 µl Cell Counting Kit‑8 reagent (Dojindo, for further analysis in the following experiments. To further Kumamoto, Japan) was added to each well at the indicated time confirm the microarray results, RT‑qPCR was performed; the points (24, 48, 72, 96 h, 5, 7 days) and incubated in the dark results revealed that the expression levels of ZEB-AS1 lncRNA at 37˚C for a further 4 h. The absorbency was subsequently in the AML group (n=62) were signic fi antly higher compared measured at the wavelength of 450 nm with the Varioskan with that of the patients with non‑malignant hemotopathy (n=10; Flash Multimode Reader (Thermo Fisher Scientic fi , Inc.). Each P<0.001; Fig. 1A) and healthy volunteers (n= 4; P= 0.010; Fi g 1.A). experiment was independently repeated at least three times. In addition, the expression levels of ZEB-AS1 lncRNA were To detect cell apoptosis, 10X binding buffer (eBioscience; positively associated with increasing AML risk levels, according Thermo Fisher Scientic fi , Inc.) was initially diluted to 1X using to modie fi d MRC and ELN recommendations (Fig. 1B and C). distilled water (1 ml 10X binding buffer + 9 ml distilled water). Furthermore, the expression levels of ZEB-AS1 lncRNA were Cells were washed once in phosphate buffered saline and once signic fi antly higher in patients with AML that had not achieved in 1X binding buffer, prior to cell resuspension in 1X binding CR (n=15) compared with those who had (n=24) subsequent to buffer to 1-5x10 /ml. A total of 5 µl u fl orochrome‑conjugated the first induction of chemotherapy (P= 0.042; Fig. 1D). Annexin V (eBioscience; Thermo Fisher Scientic fi , Inc.) was added to 100 µl cell suspension and incubated for 10‑15 min ZEB2A ‑ S1 lncRNA expression and AML clinical outcomes. In the at room temperature. Cells were washed in 1X binding buffer present study, ZEB2-AS1 lncRNA expression levels greater than and resuspended in 200 µl 1X binding buffer. A total of 5 µl the 75th percentile were considered to be high expression levels 7-Aminoactinomycin D viability staining solution (2‑8˚C) whereas those below were considered to be low, respectively. (eBioscience; Thermo Fisher Scientific, Inc.) was added Overall, 42 AML cases had available survival data. As presented (stained within 4 h and stored at 2‑8˚C in the dark). Cells were in Fig. 2, Kaplan-Meier survival plots indicated that patients analyzed using 5-color o fl w cytometry ( type FC500, Beckman with AML with high ZEB2-AS1 lncRNA expression (n=6) had Coulter company, Fullerton, CA, USA). Each experiment was signic fi antly shorter OS (3‑year OS, 0.0 vs. 68.2%; P=0.036) and independently repeated at least three times. lower DFS rates (3‑year DFS, 25.0v s. 69.8%; P= 0.039) compared with that of the low expression subgroup (n=36). Additionally, Statistical analysis. All continuous data were expressed as according to the modie fi d MRC risk stratic fi ation, in the favor - the mean ± standard error of the mean. One-way analysis able/intermediate risk group, patients with AML with low of variance with Bonferroni's correction post-hoc test for ZEB2‑AS1 lncRNA expression (n=36) had signic fi antly longer multiple comparisons were performed. Survival probabilities OS (3-year OS, 68.2 vs. 33.3%; P= 0.026) and higher DFS rates were estimated using the Kaplan-Meier method and differ- (3‑year DFS, 69.8 vs. 33.3%; P= 0.038) compared with that of the ences between survival distributions were evaluated using the high expression subgroup (n=3; Fig. 3A). Furthermore, according log-rank test. Cox's proportional hazards model was applied to the ELN risk stratic fi ation, in the favorable /intermediate I/II to estimate the hazard ratio for disease-free survival (DFS) risk groups, patients with AML with low ZEB2-AS1 lncRNA and overall survival (OS) rates. For all analyses, the P-values expression (n=23) had significantly longer OS (3-year OS, were two‑tailed and the cond fi ence interval was 95%. P<0.05 69.9 vs. 50.0%; P= 0.034) and higher DFS rates (3‑year DFS, SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Table III. Clinical outcomes of patients with AML. OS rate (months) DFS rate (months) ----------------------------------------------------------------------- ---------------------------------------------------------------------- Groups Median (95% CI) P-value Median (95% CI) P-value Age (years) 0.187 0.199 <60 35.0 (30.33-43.89) 29.0 (27.42-41.72) ≥60 27.0 (16.47‑42.73) 24.0 (12.98‑41.82) Sex 0.254 0.246 Male 25.0 (23.99-42.01) 24.0 (20.66-39.97) Female 40.0 (30.15-47.59) 39.0 (27.44-45.60) White blood cell (x10 /l; non‑M3) 0.443 0.403 <Median 35.0 (27.74-51.39) 28.0 (24.48-48.77) ≥Median 25.0 (21.86‑41.55) 24.0 (17.79‑39.27) Hemoglobin (g/l; non‑M3) 0.988 0.924 <Median 28.0 (25.22-39.14) 27.0 (21.53-6.35) ≥Median 39.5 (25.00‑53.12) 38.5 (21.40‑50.98) Platelets (x10 /l; non‑M3) 0.199 0.262 <Median 40.0 (30.38-47.31) 39.0 (26.65-44.62) ≥Median 23.0 (16.93‑45.07) 21.0 (13.29‑43.00) Blasts in bone marrow (%; non‑M3) 0.590 0.512 <Median 33.0 (24.59-49.41) 25.5 (20.41-46.34) ≥Median 32.0 (24.53‑43.71) 29.0 (21.31‑41.87) MRC risk stratification 0.005 0.003 Favorable 38.0 (32.36-45.91) 29.0 (30.32-44.20) Intermediate 27.0 (20.92-45.32) 24.0 (17.24-43.11) Adverse 29.0 (12.02-45.98) 23.0 (5.13-40.88) ELN risk stratification (non‑M3) 0.003 0.003 Favorable 40.0 (32.29-53.18) 39.0 (29.52-51.01) Intermediate I/II 25.0 (15.25-37.35) 22.0 (10.43-35.37) Adverse 19.5 (11.51-36.99) 15.5 (6.1732.33) Treatment approaches (non-M3) 0.113 0.166 Chemotherapy 29.5 (23.15-41.05) 26.5 (19.84-38.56) Allo-HSCT 31.5 (23.80-47.77) 25.5 (19.23-44.91) ZEB2-AS1 level 0.036 0.039 Low level 36.5 (30.98-44.40) 31.5 (28.55-42.50) High level 22.0 (10.55-44.12) 20.0 (3.98-41.74) AML, acute myeloid leukemia; OS, overall survival; DFS, disease‑free survival; MRC, Medical Research Council; ELN, European Leukemia Net; HSCT, hematopoietic stem cell transplantation; CI, confidence interval; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1. 71.1 vs. 50.0%; P= 0.034) compared with that of the high expre- s ZEB2‑AS1 lncRNA expression and AML treatment response. sion subgroup (n=2; Fig. 3B). The expression levels of ZEB2‑AS1 Patients with AML with high ZEB2-AS1 lncRNA expres- lncRNA in the adverse risk group were all comparatively high. sion, excluding those classed as the M3 subtype, had a In the multivariate analyses, subsequent to control- significantly lower CR rate compared with that of the low ling for confounding variables in modified MRC expression subgroup (P= 0.031; Table I). However, differ- (favorable/intermediate vs. adverse) and ELN risk stratic fi ation ences between the OS and DFS rates in the consolidation (favorable/intermediate I/II vs. adverse) groups, high ZEB2-AS1 chemotherapy (n=10) and allo-HSCT treatment groups (n=14) lncRNA expression was determined to not be significantly were not signic fi antly different (P>0.05; Table III). Using the associated with adverse patient outcomes, including reduced OS stratic fi ation method to control for confounding variables, as (P= 0.976) and DFS rates (P= 0.725; TableI V) compared with the presented in Fig. 4, it was demonstrated that patients with a low ZEB2-AS1 lncRNA expression group. low ZEB2-AS1 lncRNA expression within the allo-HSCT ONCOLOGY LETTERS 17: 4935-4947, 2019 Figure 1. Association of the expression levels of ZEB2-AS1 lncRNA with different clinical features in patients with AML. (A) Expression levels of ZEB2-AS1 lncRNA in patients with AML were significant higher compared with that of non‑malignant hemotopathy patients and healthy volunteers ( P<0.05). (B) According to the modie fi d Medical Research Council risk stratic fi ation systems, the expression levels of ZEB2‑AS1 lncRNA exhibited a signic fi ant differ - ence between favorable, intermediate and adverse risk groups of patients with AML ( P<0.05). (C) According to the European Leukemia Net risk stratic fi ation systems, the expression levels of ZEB2‑AS1 lncRNA demonstrated a signic fi ant difference between favorable, intermediate and adverse risk groups of patients with AML ( P<0.05). (D) Expression levels of ZEB2‑AS1 lncRNA were significantly higher in patients who had not achieved CR compared with that of patients who had achieved CR subsequent to the first induction of chemotherapy ( P<0.05). Values are the mean ± the standard error of the mean. ZEB2-AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; AML, acute myeloid leukemia; CR, complete remission. Figure 2. Comparison of OS and DFS rates for patients with AML with different expression levels of ZEB2-AS1 lncRNA. The cut-off value of ZEB2-AS1 lncRNA was the 75th percentile. Expression levels greater or less than the 75th percentile were tentatively considered as the high or low expression groups, respectively. Patients with AML with a high expression of ZEB2‑AS1 lncRNA demonstrated a significantly shorter (A) OS (3‑year OS, 0.0 vs. 68.2%; P= 0.036) and (B) DFS (3‑year DFS, 25.0 vs. 69.8%; P= 0.039) compared with that of the low expression of ZEB2‑AS1 lncRNA group. OSo , verall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA. treatment group (n=11) had significantly longer OS (3‑year ZEB2‑AS1 lncRNA expression and AML cell biological OS, 75.8 vs. 28.6%; P= 0.037) and DFS rates (3‑year DFS, phenot ype. Knockdown of ZEB2-AS1 lncR NA by siR NA in 81.8 vs. 28.6%; P= 0.049) compared with that of the chemo - THP-1 cells significantly inhibited the mRNA expression therapy group (n=7). levels of ZEB2 (n= 4; P<0.05) and stimulated the mRNA SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Figure 3. Comparison of OS and DFS rates for patients with AML in favorable/intermediate risk groups. (A) According to the modie fi d Medical Research Council risk stratification recommendation, in the favorable/intermediate risk group, patients with AML with a high expression of ZEB2‑AS1 lncRNA exhibited signic fi antly shorter OS (3‑year OS, 33.3 vs. 68.2%; P= 0.026) and DFS (3‑year DFS, 33.3v s. 69.8%; P= 0.038) rates compared with that of the low expression of ZEB2‑AS1 lncRNA group. (B) According to the European Leukemia Net risk stratic fi ation recommendation, in the favorable/intermediate risk group, patients with AML with a high expression of ZEB2‑AS1 lncRNA demonstrated signic fi antly shorter OS (3‑year OS, 50.0v s. 69.9%; P= 0.034) and DFS (3‑year DFS, 50.0 vs. 71.1%; P= 0.034) rates compared with that of the low expression of ZEB2‑AS1 lncRNA group. OS, overall survival; DFS d,i sease‑free survival; AML, acute myeloid leukemia; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA. Figure 4. Comparison of OS and DFS rates for patients with AML with a low expression of ZEB2 antisense RNA 1 long non-coding RNA treated using different treatment strategies. Patients with AML receiving allo‑HSCT exhibited signic fi ant longer (A) OS (3‑year OS, 75.8 vs. 28.6%; P= 0.037) and (B) DFS (3‑year DFS, 81.8 vs. 28.6%; P= 0.049) rates compared with that of the chemotherapy treatment group. OS, overall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; allo‑HSCT, allogenic hematopoietic stem cell transplantation. expression levels of E‑cadherin (n= 4; P<0.05) compared compared with the sham and vehicle groups (P<0.05). with the sham and vehicle groups (Fig. 5A and B). As THP-1 cell proliferation (n=4) and apoptosis (n=4) were presented in Fig. 6A and B, the migration (n=4) and not significantly different following the knockdown of invasion (n=4) of THP-1 cells, respectively, were signifi- ZEB2-AS1 lncRNA compared with the sham and vehicle cantly inhibited by the knockdown of ZEB2-AS1 lncRNA groups (Fig. 7A and B). ONCOLOGY LETTERS 17: 4935-4947, 2019 Figure 5. Effects of the downregulation of ZEB2-AS1 lncRNA on the mRNA expression levels of ZEB2 and E-cadherin in THP-1 cells. Subsequent to the knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells, the mRNA expression levels of (A) ZEB2 were signic fi antly decreased and (B) E‑cadherin were signifi - cantly increased. P<0.05 vs. Sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Figure 6. Effects of ZEB2-AS1 lncRNA on the cellular migration and invasion of THP-1 cells. The biological behavior of (A) migration and (B) inva- sion were significantly inhibited following the knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells. P<0.05 vs. Sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Figure 7. Effects of ZEB2-AS1 lncRNA on cellular proliferation and apoptosis in THP-1 cells. Following the knockdown of ZEB2-AS1 lncRNA, the (A) prolif- eration and (B) apoptosis exhibited no signic fi ant differences in THP‑1 cells compared with the sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Discussion functional roles in regulating biological behaviors. By retro- spectively analyzing 62 patients with de novo AML, the present In the present study, it was proposed that the abnormal over- study contributed novel results to the literature, demonstrating expression of ZEB2-AS1 lncRNA may be closely associated that the expression of ZEB2-AS1 lncRNA was abnormally with adverse outcomes in patients with AML, and may exert elevated and may have potential as an epigenetic biomarker SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML for evaluating the clinical outcomes of AML. Furthermore, it was revealed that ZEB2-AS1 lncRNA effectively modulated leukemic phenotypes including the invasion and migration of an AML cell line. The association between ZEB2-AS1 lncRNA expression and a series of clinical features, cytogenetic characteristics, somatic mutations and clinical outcomes in patients with AML was initially investigated. The results revealed that the expression of ZEB2‑AS1 lncRNA was signic fi antly higher in the AML group compared with that of a non-malignant group (P<0.001, Fig. 1A). This indicated that the overexpression of ZEB2‑AS1 lncRNA may function as a cancer‑specic fi molec - ular signal in leukemogenesis. With respect to the identie fi d karyotype and recurrent mutations, according to either the modified MRC or ELN risk stratification recommenda- tions (37-39), the expression levels of ZEB2-AS1 lncRNA had a signic fi ant stepwise increase from the favorable to adverse risk group (all P<0.05, Fig. 1B and C). Additionally, the high expression of ZEB2-AS1 lncRNA was associated with adverse patient outcomes compared with the low expression of ZEB2-AS1 lncRNA (Table I). This suggests that the over- expression of ZEB2-AS1 lncRNA is closely associated with a higher risk in AML. Accumulating evidence has revealed that numerous lncRNAs may independently predict the prognosis for patients with cancer (23,42-44). Notably, previous research has indi- cated that lncRNAs expression profiles have the potential to independently predict clinical outcomes in AML (21). Novel studies have further demonstrated that, due to its extensive oncogenic functions, the overexpression of HOTAIR lncRNA predicts poor clinical outcomes in AML and may be a poten- tial therapeutic target (25). In the present study, to evaluate the potential of its prognostic application, the impact of ZEB2-AS1 lncRNA overexpression on OS and DFS rates in patients with AML was measured. Univariate analyses revealed that patients with high ZEB2-AS1 lncRNA expression had signic fi antly shorter OS and DFS rates compared with the low ZEB2-AS1 lncRNA expression group (all P<0.05, Table III; Fig. 2). However, using the method of multivariate analyses to control for confounding variables, the adjusted 3‑year OS and DFS rates were not signic fi antly different between patients with AML with different expression levels (high vs. low) of ZEB2-AS1 lncRNA (Table IV). Furthermore, it was demon- strated that in the favorable/intermediate risk group, patients with a higher expression of ZEB2-AS1 lncRNA exhibited signic fi antly shorter OS and DFS rates compared with those with a lower expression (all P<0.05, Fig. 3). Therefore, it was proposed that although its independent prognostic value for survival was not rigorously ascertained, the expression levels of ZEB2-AS1 lncRNA may function as a complementary factor to further improve conventional risk stratification models in AML. Treatment responses in AML are notably heteroge- neous and affected by various demographic and biological factors (45-47). To minimize influences of different treat- ment approaches on clinical outcome interpretation, patients enrolled in the present cohort were treated with an identical induction chemotherapy regimen. It was revealed that the expression levels of ZEB2‑AS1 lncRNA were significantly higher in patients who did not achieve initial CR compared with Table IV. Analysis of 3-year OS and 3-year DFS rates in patients with AML. 3-Year OS 3-Year DFS ---------------------------------------------------------------------------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------- Univariate Multivariate Univariate Multivariate ----------------------------------------------------------------- -------------------------------------------------------------- ---------------------------------------------------------------- ------------------------------------------------------------- Factors HR (95% CI) P-value HR (95% CI) P-value HR (95% CI) P-value HR (95% CI) P-value MRC: F/I vs. adverse 4.04 (1.06-227.70) 0.047 0.51 (0.05-4.78) 0.600 4.48 (1.33-373.10) 0.031 0.68 (0.08-6.18) 0.927 ELN: F/I I/II vs. adverse 3.95 (1.78-66.90) 0.011 4.07 (1.25-13.22) 0.020 4.39 (2.17-95.66) 0.006 4.52 (1.40-14.69) 0.012 ZEB2-AS1 level: low vs. high 3.20 (1.15-32.24) 0.036 1.28 (0.24-6.89) 0.976 3.17 (1.10-30.42) 0.039 1.04 (0.19-5.75) 0.725 OS, overall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; HR, hazard ratio; CI, confidence interval; MRC, Medical Research Council; F/I, favorable/intermediate; ELN, European Leukemia Net; HSCT, hematopoietic stem cell transplantation; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1. ONCOLOGY LETTERS 17: 4935-4947, 2019 those who did (P=0.042, Fig. 1D). In addition, patients with a potential leukemogenic regulator in tumorigenesis (53). In a high expression of ZEB2‑AS1 lncRNA had a signic fi antly the present study, the results revealed that the knockdown of lower CR rate compared with those with a low expression of ZEB2-AS1 lncRNA in THP-1 cells effectively downregulated ZEB2-AS1 lncRNA (P= 0.031, Table I). This demonstrated the mRNA expression of ZEB2 (Fig. 5A), which was consis- that the overexpression of ZEB2-AS1 lncRNA may be closely tent with the results of a previous study (35). In addition, the associated with chemotherapy resistance. Following front-line knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells signic fi antly induction chemotherapy, patients with AML received either increased the mRNA expression of E-cadherin compared with consolidation chemotherapy or allo-HSCT. The results the sham/vehicle groups (a alll P l P< <0 0..0 05 5, F , F Fiiig g g. 5 . 5 . 5B B B) ) ), a , a , an n nd t d t d th h he r e r e re e ep p pr r re e es- s- s- revealed that differences between the 3-year OS and DFS sion of E-cadherin has been demonstrated to trigger the EMT rates in the consolidation chemotherapy and allo-HSCT in cancer progression (54). In addition, it was revealed that treatment groups were not signic fi antly different (Table III). cellular migration and invasion were signic fi antly inhibited in Furthermore, using stratic fi ation to control for confounding THP-1 cells following ZEB2-AS1 lncRNA downregulation by variables, patients with low ZEB2-AS1 lncRNA expression si-ZEB2-AS1 (all P<0.05, Fig. 6). However, the knockdown of in the allo‑HSCT‑treated group had a signic fi antly longer OS ZEB2-AS1 lncRNA in THP-1 cells had no apparent effects and DFS compared with that of the chemotherapy group (all on proliferation and apoptosis (Fig. 7). Accordingly, it was P<0.05, Fig. 4). This suggested that patients with AML with a proposed that ZEB2-AS1 lncRNA may have upregulated low expression of ZEB2-AS1 lncRNA may be more sensitive ZEB2 expression, which in turn enhanced the motility pheno- to allo-HSCT therapy. In comparison with the relatively static type of THP-1 cells, including their invasion and migration genetic landscape, epigenetic status, including DNA methyla- abilities in vitro. These results suggested that chemotherapy tion, is altered during different phases of AML (13). However, resistance in patients with a high expression of ZEB2-AS1 dynamic changes in ZEB2-AS1 lncRNA expression in this lncRNA may be closely associated with enhanced cellular retrospective cohort could not be evaluated, and the associa- migration and invasion during leukemic progression. This tion of ZEB2-AS1 lncRNA with treatment responses requires would be consistent with the results of a previous study that further consolidation by future prospective studies. demonstrated that ZEB family protein expression may predict The limitations of the present clinical study should be differential responses to various chemotherapy drugs in hema- considered carefully. From a clinical point of view, therapeutic topoietic malignancies, including mantle cell lymphoma (51). strategies for AML have improved in previous years (47). In conclusion, to the best of our k nowledge, the present study Additionally, cytogenetic/molecular risk stratic fi ation systems was the first to evaluate the prognostic value of ZEB2‑AS1 remain controversial and require further improvement (7,9,11). lncRNA in AML. The results demonstrated that the over- These all affect the prognostic importance of ZEB2-AS1 expression of ZEB2-AS1 lncRNA was associated with poor lncRNA in AML to varying degrees. Furthermore, from a clinical outcomes in AML. Although the independent prog- statistical point of view, the retrospective cohort size in the nostic prediction for survival was not rigorously researched present study was relatively small, which may have affected in the present study, the overexpression of ZEB2-AS1 the multivariate analysis results. Thus, a future study with lncRNA may function as a candidate gene to improve cyto- a large cohort must be retrospectively and/or prospectively genetic/somatic mutation risk stratic fi ation systems in AML. analyzed to further assess the prognostic value of ZEB2-AS1 Furthermore, it was discovered that ZEB2-AS1 lncRNA lncRNA for patient survival and treatment response in AML, effectively modulated the leukemic phenotypes of invasion independent of various approved clinical factors. and migration, which may be associated with the differential Increasing evidence has demonstrated that various responses to treatment strategies. Finally, another question lncRNAs serve essential functions in not only in intrinsic must be addressed-why is ZEB2-AS1 lncRNA overexpressed cellular regulatory networks, but also in intercellular commu- in AML? It was noted that the expression levels of ZEB2-AS1 nications during tumorigenesis (23). Multiple lncRNAs have lncRNA in the AML group exhibit high heterogeneity, with been confirmed to function as original drivers and/or down - greater variability compared with that of the non-malignant stream targets in circuits involving almost all hallmarks of group. Furthermore, in patients with AML with a karyotype cancer, including proliferation, viability, immortality, motility, of 11q23, ZEB2-AS1 lncRNA expression was notably high. angiogenesis and tumor suppression (23,48). Previously, it has AML pathogenesis with a 11q23 karyotype involves the been reported that a natural antisense transcript of ZEB2 may abnormal rearrangement of the mixed lineage leukemia regulate the EMT in different tumor types, including colon gene, which is an epigenetic modifier involved in histone adenocarcinoma (35). The highly‑conserved zinc‑finger struc - methylation (55). It was hypothesized that the overexpression ture of ZEB2 binds to E-boxes located in the promoter regions of ZEB2-AS1 lncRNA may not be a key event during leuke- of certain target genes including E-cadherin, so as to further mogenesis, but a downstream target of other key oncogenic regulate EMT in cancer progression (49,50). Novel results events. As previously mentioned, the expression profiling of have demonstrated that abnormal expression of ZEB1 may lncRNAs is able to independently evaluate survival in older promote cellular proliferation and tumor growth in mantle cell patients with cytogenetically normal AML (21). A novel study lymphoma (51). Furthermore, ZEB1 expression is controlled revealed that the combination of >1 lncRNA (i.e. a six-lncRNA by growth arrest specific 5-AS1 lncRNA to modulate cell signature) may be strongly associated with survival in diffuse migration and invasion in non-small cell lung cancer (52). large B-cell lymphoma (44). This suggests that one lncRNA Notably, a novel ZEB2-BAF chromatin remodeling complex alone may be not sufci fi ent in independently predicting the subunit BCL11B fusion gene has been identie fi d in patients survival of patients with AML. These key problems require with AML with karyotype of t(2;14)(q22;q32), which may be thorough investigation in the future. SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML 4. Almeida AM and Ramos F: Acute myeloid leukemia in the older Acknowledgements adults. Leuk Res Rep 6: 1-7, 2016. 5. Percival ME, Tao L, Medeiros BC and Clarke CA: Improvements Not applicable. in the early death rate among 9,380 patients with acute myeloid leukemia after initial therapy: A SEER database analysis. Cancer 121: 2004-2012, 2015. Funding 6. Bhatnagar B and Garzon R: The use of molecular genetics to refine prognosis in acute myeloid leukemia. Curr Hematol Malig Rep 9: 148-157, 2014. This work was supported by the Priority Academic 7. Stölzel F, Mohr B, Kramer M, Oelschlägel U, Bochtler T, Program Development of Jiangsu Higher Education Berdel WE, Kaufmann M, Baldus CD, Schäfer-Eckart K, Institutions, the National Clinical Key Subject Project, Stuhlmann R, et al: Karyotype complexity and prognosis in acute myeloid leukemia. Blood Cancer J 6: e386, 2016. the Innovation Capability Development Project of Jiangsu 8. Mrózek K and Bloome fi ld CD: Chromosome aberrations, gene Province (grant no. BM2015004), the Natural Science mutations and expression changes, and prognosis in adult acute Foundation of China (grant nos. 81570139 and 81270617), myeloid leukemia. Hematology Am Soc Hematol Educ Program 2006: 169-177, 2006. the Jiangsu Provincial Special Program of Medical Science 9. Papaemmanuil E, Gerstung M, Bullinger L, Gaidzik VI, (grant no. BL2012005), Jiangsu Province's Key Medical Center Paschka P, Roberts ND, Potter NE, Heuser M, Thol F, (grant no. ZX201102), the Jiangsu Province Natural Science Bolli N, et al: Genomic classification and prognosis in acute myeloid leukemia. N Engl J Med 374: 2209-2221, Fund (grant no. BE2015639), the Project of Natural Science Foundation of Jiangsu Province (grant no. BK20141201) and 10. Byun JM, Kim YJ, Yoon HJ, Kim SY, Kim HJ, Yoon J, Min YH, the National Natural Science Foundation of China Youth Fund Cheong JW, Park J, Lee JH, et al: Cytogenetic profiles of 2,806 patients with acute myeloid leukemia-a retrospective Project (grant no. 81300424). multicenter nationwide study. Ann Hematol 95: 1223-1232, Availability of data and materials 11. Li S, Mason CE and Melnick A: Genetic and epigenetic hetero- geneity in acute myeloid leukemia. Curr Opin Genet Dev 36: 100-106, 2016. The datasets used and analyzed during the current study 12. Cancer Genome Atlas Research Network, Ley TJ, Miller C, are available from the corresponding author on reasonable Ding L, Raphael BJ, Mungall AJ, Robertson A, Hoadley K, Triche TJ Jr, Laird PW, et al: Genomic and epigenomic request. landscapes of adult de novo acute myeloid leukemia. N Engl J Med 368: 2059-2074, 2013. Authors' contributions 13. Li S, Ga r rett-Ba kelman F E, Chung SS, Sanders MA, H r ici k T, Rapapor t F, Pat el J, Di l lon R, Vijay P, Brown A L, et al: Distinct evolution and dynamics of epigenetic and genetic XS finished all the experiments, analyzed and interpreted the heterogeneity in acute myeloid leukemia. Nat Med 22: patient data, and was a major contributor in writing the manu- 792-799, 2016. script. JL finished most of the experiments. LM, LW, QW and 14. Sasca D and Huntly BJ: Independence of epigenetic and genetic diversity in AML. Nat Med 22: 708-709, 2016. HY performed some experiments. CR and DW had substantial 15. Mattick JS and Makunin IV: Non-coding RNA. Hum Mol contributions to the conception and design of the work. XZ Genet 15 Spec No 1: R17-R29, 2006. and SC designed the study and took final responsibility. All 16. Morris KV and Mattick JS: The rise of regulatory RNA. Nat Rev Genet 15: 423-437, 2014. authors read and approved the final manuscript. 17. Djeba l i S, Davis CA, Merkel A, Dobin A, Lassmann T, Mortazavi A, Tanzer A, Lagarde J, Lin W, Schlesinger F, et al: Ethics approval and consent to participate Landscape of transcription in human cells. Nature 489: 101-108, 18. Iyer MK, Niknafs YS, Malik R, Singhal U, Sahu A, Hosono Y, The present study was approved by the Institutional Review Barrette TR, Prensner JR, Evans JR, Zhao S, et al: The land- Board of the First Affiliated Hospital of Soochow University scape of long noncoding RNAs in the human transcriptome. Nat Genet 47: 199-208, 2015. Written informed consent was obtained from all participants 19. Goff LA and Rinn JL: Linking RNA biology to lncRNAs. for the use of bone marrow specimens and clinical information. Genome Res 25: 1456-1465, 2015. 20. Ling H, Vincent K, Pichler M, Fodde R, Berindan-Neagoe I, Slack FJ and Calin GA: Junk DNA and the long non-coding Patient consent for publication RNA twist in cancer genetics. Oncogene 34: 5003-5011, 2015. 21. Garzon R, Volinia S, Papaioannou D, Nicolet D, Kohlschmidt J, Not applicable. Yan PS, Mrózek K, Bucci D, Carroll AJ, Baer MR, et al: Expression and prognostic impact of lncRNAs in acute myeloid leukemia. Proc Natl Acad Sci USA 111: 18679-18684, 2014. Competing interests 22. Morlando M, Ballarino M and Fatica A: Long non-coding RNAs: New players in hematopoiesis and leukemia. Front Med (Lausanne) 2: 23, 2015. The authors declare that they have no competing interests. 23. Schmitt AM and Chang HY: Long noncoding RNAs in cancer pathways. Cancer Cell 29: 452-463, 2016. 24. Zhang X, Lian Z, Padden C, Gerstein MB, Rozowsky J, Snyder M, References Gingeras TR, Kapranov P, Weissman SM and Newburger PE: A myelopoiesis-associated regulatory intergenic noncoding 1. Döhner H, Weisdorf DJ and Bloomfield CD: Acute myeloid RNA transcript within the human HOXA cluster. Blood 113: leukemia. N Engl J Med 373: 1136-1152, 2015. 2526-2534, 2009. 2. Polednak AP: Recent improvement in completeness of inci- 25. Xing CY, Hu XQ, Xie FY, Yu ZJ, Li HY, Bin-Zhou, Wu JB, dence data on acute myeloid leukemia in US cancer registries. Tang LY and Gao SM: Long non-coding RNA HOTAIR modu- J Registry Manag 41: 77-84, 2014. lates c-KIT expression through sponging miR-193a in acute 3. Tawfik B, Pardee TS, Isom S, Sliesoraitis S, Winter A, Lawrence J, myeloid leukemia. FEBS Lett 589: 1981-1987, 2015. Powell BL and Klepin HD: Comorbidity, age, and mortality among 26. Ma L, Bajic VB and Zhang Z: On the classification of long adults treated intensively for acute myeloid leukemia (AML). J Geriatr non-coding RNAs. RNA Biol 10: 925-933, 2013. Oncol 7: 24-31, 2016. ONCOLOGY LETTERS 17: 4935-4947, 2019 27. Mattick JS and Rinn JL: Discovery and annotation of long 41. Liva k KJ and Sch m it tgen TD: Analysis of relative gene noncoding RNAs. Nat Struct Mol Biol 22: 5-7, 2015. expression data using real-time quantitative PCR and the 28. Khorkova O, Myers AJ, Hsiao J and Wahlestedt C: Natural anti- 2(-Delta Delta C(T)) method. Methods 25: 402-408, 2001. sense transcripts. Hum Mol Genet 23: R54-R63, 2014. 42. Wang KC and Chang HY: Molecular mechanisms of long 29. Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, noncoding RNAs. Mol Cell 43: 904-914, 2011. Nakamura M, Nishida H, Yap CC, Suzuki M, Kawai J, et al: 43. Bhan A and Mandal SS: lncRNA HOTAIR: A master regulator Antisense transcription in the mammalian transcriptome. of chromatin dynamics and cancer. Biochim Biophys Acta 1856: Science 309: 1564-1566, 2005. 151-164, 2015. 30. ENCODE Project Consortium. An integrated encyclopedia of 44. Sun J, Cheng L, Shi H, Zhang Z, Zhao H, Wang Z and Zhou M: A DNA elements in the human genome. Nature 489: 57-74, 2012. potential panel of six-long non-coding RNA signature to improve 31. Su WY, Xiong H and Fang JY: Natural antisense transcripts survival prediction of diffuse large-B-cell lymphoma. Sci Rep 6: regulate gene expression in an epigenetic manner. Biochem 27842, 2016. Biophys Res Commun 396: 177-181, 2010. 45. Schlenk RF, Dohner K, Krauter J, Fröhling S, Corbacioglu A, 32. Wight M and Werner A: The functions of natural antisense tran- Bullinger L, Habdank M, Späth D, Morgan M, Benner A, et al: scripts. Essays Biochem 54: 91-101, 2013. Mutations and treatment outcome in cytogenetically normal 33. Ebralidze AK, Guibal FC, Steidl U, Zhang P, Lee S, Bartholdy B, acute myeloid leukemia. N Engl J Med 358: 1909-1918, 2008. Jorda MA, Petkova V, Rosenbauer F, Huang G, et al: PU.1 46. Xu Y, Sun Y, Shen H, Ding L, Yang Z, Qiu H, Sun A, Chen S and expression is modulated by the balance of functional sense and Wu D: Allogeneic hematopoietic stem cell transplantation could antisense RNAs regulated by a shared cis-regulatory element. improve survival of cytogenetically normal adult acute myeloid Genes Dev 22: 2085-2092, 2008. leukemia patients with DNMT3A mutations. Am J Hematol 90: 34. Sun J, Li W, Sun Y, Yu D, Wen X, Wang H, Cui J, 992-997, 2015. Wang G, Hoffman AR and Hu JF: A novel antisense long 47. Dombret H and Gardin C: An update of current treatments for noncoding RNA within the IGF1R gene locus is imprinted adult acute myeloid leukemia. Blood 127: 53-61, 2016. in hematopoietic malignancies. Nucleic Acids Res 42: 48. Hanahan D and Weinberg RA: Hallmarks of cancer: The next 9588-9601, 2014. generation. Cell 144: 646-674, 2011. 35. Beltran M, Puig I, Peña C, García JM, Alvarez AB, Peña R, 49. Vandewalle C, Van Roy F and Berx G: The role of the ZEB Bonilla F and de Herreros AG: A natural antisense transcript family of transcription factors in development and disease. Cell regulates Zeb2/Sip1 gene expression during Snail1-induced Mol Life Sci 66: 773-787, 2009. epithelial-mesenchymal transition. Genes Dev 22: 756-769, 50. Hill L, Browne G and Tulchinsky E: ZEB/miR-200 feedback 2008. loop: At the crossroads of signal transduction in cancer. Int 36. A rber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, J Cancer 132: 745-754, 2013. Le Beau MM, Bloome fi ld CD, Cazzola M and Vardiman JW: The 51. Sánchez-Tilló E, Fanlo L, Siles L, Montes-Moreno S, 2016 revision to the World Health Organization classic fi ation of Moros A, Chiva-Blanch G, Estruch R, Martinez A, Colomer D, myeloid neoplasms and acute leukemia. Blood 127: 2391-2405, Győrffy B, et al: The EMT activator ZEB1 promotes tumor 2016. growth and determines differential response to chemotherapy in 37. Metzeler KH, Herold T, Rothenberg-Thurley M, Amler S, mantle cell lymphoma. Cell Death Differ 21: 247-257, 2014. Sauerland MC, Görlich D, Schneider S, Konstandin NP, 52. Wu Y, Lyu H, Liu H, Shi X, Song Y and Liu B: Downregulation Dufour A, Bräundl K, et al: Spectrum and prognostic relevance of the long noncoding RNA GAS5-AS1 contributes to tumor of driver gene mutations in acute myeloid leukemia. Blood 128: metastasis in non-small cell lung cancer. Sci Rep 6: 31093, 2016. 686-698, 2016. 53. Torkildsen S, Gorunova L, Beiske K, Tjonnfjord GE, Heim S 38. Grimwade D, Hills RK, Moorman AV, Walker H, Chatters S, and Panagopoulos I: Novel ZEB2-BCL11B fusion gene identi- Goldstone AH, Wheatley K, Har rison CJ and Burnett AK; e fi d by RNA‑sequencing in acute myeloid leukemia with t(2;14) National Cancer Research Institute Adult Leukaemia Working (q22;q32). PLoS One 10: e0132736, 2015. Group: Refinement of cytogenetic classic fi ation in acute myeloid 54. Schmalhofer O, Brabletz S and Brabletz T: E-cadherin, leukemia: Determination of prognostic significance of rare beta-catenin, and ZEB1 in malignant progression of cancer. recurring chromosomal abnormalities among 5876 younger Cancer Metastasis Rev 28: 151-166, 2009. adult patients treated in the United Kingdom Medical Research 55. Stein EM and Tallman MS: Mixed lineage rearranged leukaemia: Council trials. Blood 116: 354-365, 2010. Pathogenesis and targeting DOT1L. Curr Opin Hematol 22: 39. Röllig C, Bornhäuser M, Thiede C, Taube F, Kramer M, Mohr B, 92-96, 2015. Aulitzky W, Bodenstein H, Tischler HJ, Stuhlmann R, et al: Long-term prognosis of acute myeloid leukemia according to the new genetic risk classic fi ation of the European LeukemiaNet This work is licensed under a Creative Commons recommendations: Evaluation of the proposed reporting system. Attribution-NonCommercial-NoDerivatives 4.0 J Clin Oncol 29: 2758-2765, 2011. 40. McGowan-Jordan J, Si mons A and Sch m id M (eds): An International (CC BY-NC-ND 4.0) License. International System for Human Cytogenomic Nomenclature (2016). Cytogenetic and Genome Research 149: No. 1-2, 2016. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Oncology Letters Pubmed Central

Overexpression of ZEB2-AS1 lncRNA is associated with poor clinical outcomes in acute myeloid leukemia

Loading next page...
 
/lp/pubmed-central/overexpression-of-zeb2-as1-lncrna-is-associated-with-poor-clinical-rMVUrCFfPj

References (61)

Publisher
Pubmed Central
Copyright
Copyright: © Shi et al.
ISSN
1792-1074
eISSN
1792-1082
DOI
10.3892/ol.2019.10149
Publisher site
See Article on Publisher Site

Abstract

ONCOLOGY LETTERS 17: 4935-4947, 2019 Overexpression of ZEB2‑AS1 lncRNA is associated with poor clinical outcomes in acute myeloid leukemia 1-3* 4* 1 1 1 1 XIAOLAN SHI , JIAO LI , LIANG MA , LIJUN WEN , QINRONG WANG , HONG YAO , 1,2 1-3 5 1,2 CHANGGENG RUAN , DEPEI WU , XINYOU ZHANG and SUNING CHEN Department of Hematology, Jiangsu Institute of Hematology, Key Laboratory of Thrombosis and Hemostasis of Ministry of Health, The First Affiliated Hospital of Soochow University; 2 3 Collaborative Innovation Center of Hematology, Institute of Blood and Marrow Transplantation, Soochow University, Suzhou, Jiangsu 215006; Department of Hematology, Yixing People's Hospital of Jiangsu Province, Yixing, Jiangsu 214200; Department of Hematology, Shenzhen People's Hospital, The Second Clinical Medical College of Jinan University, Shenzhen, Guangdong 518020, P.R. China Received January 10, 2018; Accepted January 28, 2019 DOI: 10.3892/ol.2019.10149 Abstract. Acute myeloid leukemia (AML) is a fatal hemato- compared with patients with a lower expression of ZEB2-AS1 poietic malignancy with poor clinical outcomes. To determine lncRNA (P= 0.031). In cases with low levels of ZEB2-AS1 whether the expression of the long non-coding (lnc)RNA zinc lncRNA, patients treated with allogenic hematopoietic stem finger E-box binding homeobox 2 (ZEB2) antisense RNA 1 cell transplantation had signic fi antly longer OS and DFS rates (ZEB2-AS1) is associated with clinical outcomes, its expres- compared with that of chemotherapy-treated patients (P=0.037 sion was analyzed in a retrospective cohort of 62 AML and 10 and P=0.049 respectively). Furthermore, the knockdown of non-malignant cases. The results revealed that the expression ZEB2-AS1 lncRNA effectively inhibited AML cell invasion and of ZEB2-AS1 lncRNA was notably high and closely associated migration, which was closely associated with the downregulation with adverse clinical outcomes in AML cases compared with the of ZEB2 and upregulation of E-cadherin expression. Collectively, non‑malignant cases, based on either modie fi d Medical Research although its independent prognostic value for survival was not Council or European Leukemia Net risk stratic fi ation systems. rigorously determined, ZEB2-AS1 lncRNA may function as a Univariate analyses indicated that patients with a higher expres- candidate marker to improve conventional risk stratification sion of ZEB2-AS1 lncRNA had significantly shorter overall systems and the evaluation of therapeutic responses for AML. survival (OS) (P=0.036) and disease-free survival (DFS) rates (P=0.039) compared with patients with a lower expression of Introduction ZEB2-AS1 lncRNA. In addition, patients with a higher expres- sion of ZEB2‑AS1 lncRNA had a signic fi ant lower complete Acute myeloid leukemia (AML) is predominantly a fatal hema- remission rate in response to induction by chemotherapy topoietic malignancy characterized by the clonal proliferation of myeloid blasts with tissue infiltration (1). It may occur at any age, with an incidence of 2-3/100,000 per annum in children <14 years old, and ~15/100,000 per annum in adults >60 years old globally (2). Despite advances in therapeutic strategies, Correspondence to: Professor Suning Chen, Department of including intensive chemotherapy and hematopoietic stem cell Hematology, Jiangsu Institute of Hematology, Key Laboratory transplantation (HSCT), the clinical outcome of AML remains of Thrombosis and Hemostasis of Ministry of Health, The First poor, particularly in older patients (>60 years old) (3-5). Affiliated Hospital of Soochow University, 188 Shizi Road, Suzhou, Considering the clonal complexity of AML, there has been Jiangsu 215006, P.R. China increasing interest in improving the prognosis and treatment E-mail: chensuning@sina.com of AML through the more extensive biological profiling of Professor Xinyou Zhang, Department of Hematology, Shenzhen cytogenetic and molecular tumor heterogeneity (6-10). The People's Hospital, The Second Clinical Medical College of Cancer Genome Atlas Research Network has reported that Jinan University, 1017 Dongmen North Road, Shenzhen, ~70% of AML cases have mutations in genes encoding epigen- Guangdong 518020, P.R. China etic modier fi s (11,12). Notably, novel data has demonstrated E-mail: zxy0518@live.cn that DNA methylation heterogeneity (epialleles) may occur with distinct kinetics and patterns that are likely to affect Contributed equally clinical outcomes. These may be hallmarks of AML and may Key words: acute myeloid leukemia, long non-coding RNA, zinc be independent of the genetic landscape (13,14). Accordingly, finger E-box binding homeobox 2 antisense RNA 1, prognosis differences in epigenetic diversity may function as molecular biomarkers to independently evaluate AML prognosis. SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Long non-coding RNAs (lncRNAs) are a class of RNAs Medical Research Council (MRC) or European Leukemia that are >200 nucleotides in length (15-17). Gene expression Net (ELN) recommendations were applied for risk stratic fi a - regulated by lncRNAs is regarded as one of the most notable tion (37-39). The 62 bone marrow specimens were collected types of epigenetic control (18,19). Recurrent mutations and/or from the pretreated patients and were frozen and archived for epigenetic alterations in the regulatory non-coding genome the following experiments. For comparison, 10 eligible bone may broadly affect lncRNA expression in numerous malignant marrow specimens were collected from patients without hema- tumor types, serving as signals for carcinogenesis, in addition topoietic malignancies and were selected as the non-malignant to providing information for prognosis and therapeutic options hemotopathy group (Table II). Clinical outcome data were in patients with cancer (20). Notably, the expression of a small updated as of April 2016. The present research was ethi- subset of lncRNAs, including nuclear paraspeckle assembly cally approved by the Institutional Review Board of the First transcript 1, have been strongly associated with treatment Affiliated Hospital of Soochow University. response and survival in cytogenetically normal older patients with AML (21). In particular, a range of lncRNAs, including AML cell line. Using the Affymetrix Human LncRNA HOXA transcript antisense RNA, myeloid‑specic fi 1 and HOX microarray analysis, it has been demonstrated that ZEB2-AS1 transcript antisense intergenic RNA myeloid 1 (HOTAIR), lncRNA is predominantly overexpressed in patients with may exert pivotal effects not only on hematopoietic stem cells AML with a karyotype of 11q23. In addition, the expression during normal hematopoiesis, but also on the cancer pheno- of ZEB2-AS1 lncRNA in THP-1 cells, also with a karyo- type during the process of leukemogenesis (22-25). type of 11q23, is signic fi antly higher compared with that of lncRNAs are categorized into antisense, bidirectional, other AML cell lines, including AP1060, NB4 and FKH-1 intronic, intergenic and overlapping lncRNAs, based on their (P=0.0020). Thus, THP-1 cells were selected for the present chromosomal location (26,27). Antisense lncRNAs are initially study and cultured in RPMI-1640 medium (GE Healthcare transcribed from the opposite strand of a protein-coding Life Sciences, Hyclone, Logan, UT, USA) supplemented with counterpart, functioning as fast regulatory mediators in 10% heat‑inactivated fetal bovine serum (FBS; Sigma‑Aldrich; self-regulatory circuits to modulate global and/or specific Merck KGaA, Darmstadt, Germany) in humidified 37˚C transcriptional outputs (28-32). Certain antisense lncRNAs, incubator containing 5% CO . including IGF1R antisense imprinted non-protein coding RNA, are downregulated in patients with high-risk AML, resulting Patient treatment. A total of 39 patients with de novo AML, in the promotion of cell growth through long-range chromatin excluding those diagnosed as the M3 subtype, received interactions with insulin like growth factor 1 receptor (33,34). front-line induction chemotherapy, including the idarubicin Using the Affymetrix Human LncRNA microarray analysis, it and cytarabine regimen, as follows: Idarubicin 8-12 mg/m was has been demonstrated that the expression of the lncRNA (days 1-3) and cytarabine 100 mg/m (days 1‑7); or the dauno - zinc finger E‑box binding homeobox 2 (ZEB2) antisense RNA 1 rubicin and cytarabine regimen, as follows: Daunorubicin 2 2 (ZEB2-AS1) is abnormally overexpressed in patients with AML 60-90 mg/m (days 1-3) and cytarabine 100 mg/m (days 1-7). (as yet unpublished). A previous study indicated that a natural Subsequent to achieving first complete remission (CR, n=24), antisense transcript, overlapping the 5' splice site in the intron patients received post-remission therapy of either several of the ZEB2 gene, may prevent splicing of the 5'-untranslated consolidation courses (n=12) or allogeneic HSCT (allo‑HSCT; region to increase ZEB2 translation and consequently down- n=12). regulate the expression of E-cadherin, which in turn induces For the treatment of AML with allo-HSCT, patients epithelial-mesenchymal transition (EMT) in a tumor (35). received an initial conditioning regimen with lomustine 2 2 However, the prognostic value of ZEB2-AS1 in AML and its (250 mg/m /day on day -10), cytarabine (2 or 4 g/m /day; days function in leukemogenesis remains to be elucidated. ‑9 to ‑8), busulfan (3.2 mg/kg/day; days ‑7 to ‑5) and cyclo- In the present study, 62 de novo patients with AML were phosphamide (1.8 g/m /day; days ‑4 to ‑3). Due to advanced retrospectively analyzed to determine if ZEB2-AS1 lncRNA patient age or the presence of other comorbidities, patients may function as a biomarker to evaluate AML prognosis. with poor responses to myeloablative conditioning received Thus, the specic fi aim of the present study was to assess the a regimen with lomustine (250 mg/m /day; day ‑10), fluda - 2 2 association between ZEB2-AS1 lncRNA expression and the rabine (30 mg/m ; days ‑10 to ‑6), cytarabine (1.5 g/m /day; clinical features of patients with AML. Additionally, the days ‑10 and ‑6) and busulfan (3.2 mg/kg/day; days ‑5 to ‑3). potential regulation of leukemic phenotypes by ZEB2-AS1 To effectively prevent graft-versus-host disease (GVHD), lncRNA was investigated. As such, the clinical and biological cyclosporine (3 mg/kg/day) was infused to achieve a target importance of ZEB2-AS1 lncRNA were evaluated. blood concentration between 200-300 ng/ml, starting on day -9 or -1 until patients switched to oral administration. For Materials and methods unrelated or haploidentical transplantation, mycophenolate mofetil (30 mg/kg/day) and rabbit anti-thymocyte globulin Patients and tissue specimens. A total of 62 eligible patients (2.5 mg/kg/day; days ‑5 to ‑2) were additionally administered with de novo AML were enrolled retrospectively in the present to prevent GVHD. In addition, methotrexate was separately study. Patients were diagnosed and classie fi d according to the administered on days +1, +3, +6, and +11, at doses of 15, 10, 10 World Health Organization (36) criteria at the First Affiliated and 10 mg/m , respectively. Hospital of Soochow University (Jiangsu, China) between May 2007 and June 2014. The clinicopathological charac- Cytogenetic and molecular genetic analysis. In the cytogenetic teristics of this cohort are summarized in Table I. Modie fi d analyses, bone marrow specimens from patients with de novo ONCOLOGY LETTERS 17: 4935-4947, 2019 Table I. Clinical, pathological and genetic characteristics of patients with AML. ZEB2-AS1 expression --------------------------------------------------------------------------------- Characteristics Patients Low level High level P-value Age (years) 0.552 Median (range) 39 (8-80) 39 (8-80) 34 (14-67) Sex 0.537 Male 32 (51.6%) 24 8 Female 30 (48.4%) 25 5 FAB Subtypes 0.006 M0 1 (1.6%) 0 1 M1 1 (1.6%) 1 0 M2 27 (43.6%) 25 2 M3 9 (14.5%) 9 0 M4 10 (16.1%) 5 5 M5 14 (22.6%) 9 5 Karyotype <0.001 Normal karyotype 16 (25.8%) 14 2 t(15;17) 9 (14.5%) 9 0 t(8;21) 13 (21.0%) 12 1 inv (16) 7 (11.3%) 5 2 t(6;9) 5 (8.0%) 5 0 11q23 8 (12.9%) 4 4 Complex karyotype 4 (6.5%) 0 4 White blood cell (x10 /l; non‑M3) 0.046 Median (range) 24.3 (1.0-190.5) 13.5 (1.0-140.2) 52.1 (1.3-190.3) Hemoglobin (g/l; non‑M3) 0.372 Median (range) 84.0 (37.0-149.0) 84.0 (37.0-149.0) 88.0 (38.0-116.0) Platelets (x10 /l; non‑M3) 0.044 Median (range) 40 (8.0-414.0) 31 (8-414) 71.5 (20-410) Blasts in bone marrow (%; non‑M3) 0.569 Median (range) 56.5 (20.5-98.0) 57 (20.5-95.5) 56.5 (25.0-98.0) Mutated gene (non-M3) 0.474 Negative 15 12 3 CEBPA 2 2 0 NPM1 1 1 0 FLT3-ITD 1 1 0 FLT3-TKD 1 1 0 DNMT3A 1 1 0 C-kit 4 1 3 C-kit/CEBPA 2 2 0 NPM1/FLT3-TKD 1 1 0 FLT3-ITD/CEBPA 1 1 0 NPM1/DNMT3A 1 1 0 DNMT3A/NPM1/FLT3-ITD 2 1 1 Modified MRC risk stratification 0.002 Favorable 29 (46.8%) 26 3 Intermediate 28 (45.1%) 22 6 Adverse 5 (8.1%) 1 4 ELN risk stratification (non‑M3) 0.028 Favorable 27 24 3 Intermediate I and II 13 8 5 Adverse 10 5 5 SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Table I. Continued. ZEB2-AS1 expression ----------------------------------------------------------- Characteristics Patients Low level High level P-value Recovery from induction chemotherapy (non-M3) 0.031 CR 24 21 3 Non-CR 15 7 8 AML, acute myeloid leukemia; FAB, French‑American‑British; MRC, Medical Research Council; ELN, European Leukemia Net; CR, complete remission; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; CEBPA, CCAAT enhancer binding protein α; NPM1, nucleophosmin 1; FLT3, fms related tyrosine kinase 3; ITD, internal tandem duplication; TKD, tyrosine kinase domain; DNMT3A, DNA methyltransferase 3α. Table II. Characteristics of 10 non-malignant hemotopathy CCAAT enhancer binding protein α (CEBPA, forward, 5'-GGC cases. G AG C AG GG T C T C C GG G T -3 ' and reverse, 5'-TGT GCT GGA A CA GGT C GG C CA- 3') and nucleophosmin 1 (NPM1, Sex Number Age Number fo r wa r d , 5 '- T TA AC T C T C T G G T G G TAG A AT GA A-3 ' and reverse, 5'-TGT TAC AGA AAT GAA ATA AGA CGG-3'). Male 3 ≥60 2 Female 7 <60 8 RNA extraction and reverse transcription‑quantitative (RT‑q) WBC 5.585 (2.6-9.18)x10/l Diagnosed Number PCR. Total RNA was extracted from patient bone marrow mononuclear cells using TRIzol (Invitrogen; Thermo Fisher HGB 74.5 (57-156)g/l IDA 8 Scientific, Inc.). RT and first strand cDNA synthesis was PLT 281 (10-324)x10 /l ITP 2 subsequently performed using MMLV-RT reverse transcrip- WBC, white blood cell count; HGB, hemoglobin; PLT, platelet count. tase (Promega Corporation, Madison, WI, USA 37˚C 60 min, 95˚C 5 min). RT-qPCR analysis was employed to detect levels of ZEB2-AS1. GAPDH was used as an internal reference gene. The primer sequences used were as follows: ZEB2-AS1 AML were processed in standard un-stimulated cultures for forward, 5'-GGC TGG ATA GCA AAG GAC-3' and reverse, 24 h. With standard techniques of ISCN 2016 (40) for chro- 5'-ACA CTC TTG GCG AGG T‑3'; ZEB2 forward, 5'‑GTC CAT mosome R‑banding and fluorescence in situ hybridization, GCG AAC TGC CAT CT-3' and reverse, 5'-ATC TGT CCC TGG the different karyotypes in patients with AML were routinely CTT GTG TG‑3'; E‑cadherin forward, 5'‑TGC CCA GAA AAT determined. If available, at least 20 metaphases were analyzed GAA AAA GG-3' and reverse, 5'-GTG TAT GTG GCA ATG for every bone marrow sample. For analyzing mutations in CGT TC‑3'; GAPDH forward, 5'‑CAA GGT CAT CCA TGA patients with de novo AML, a Purelink™ Genomic DNA CAA CTT TG-3', and reverse, 5'-GTC CAC CAC CCT GTT mini kit (Invitrogen; Thermo Fisher Scientic fi , Inc., Waltham, GCT GTA G-3'. SYBR Green (Taraka, Japan) RT-qPCR was MA, USA) was used to extract genomic DNA from the performed and the relative threshold cycle value normalized to 62 patients' bone marrow mononuclear cells, according to the the reference GAPDH gene was obtained (ABI 7500; Thermo manufacturer's protocol. The coding regions of mutated genes Fisher Scientic fi , Inc.) The thermo cycling conditions were as were either partially or entirely amplie fi d using a polymerase follows 50˚C 2 min, 95˚C 10 min, 95˚C 15 sec, 60˚C 1 min, - Cq ΔΔ chain reaction (PCR) in order to identify these mutations. for a total of 40 cycles. Following this, 2 was calculated The genomic DNA was extracted from the 62 patients' bone to determine relative abundance of target gene expression marrow mononuclear cells. The thermo cycling conditions between the groups (41). were as follows 95˚C 5 min, total 35 cycles of 95˚C 30 sec and 58˚C 30 sec and 72˚C 1 min, then 72˚C 10 min. Direct bidi- RNA interference. Gene-specific small interfering RNAs rectional DNA sequencing was subsequently performed. In (siRNAs) against ZEB2‑AS1 (siZEB2‑AS1; sense, 5'‑CAC the present study, a range of acute leukemia-associated muta-CUU UGG UUA CCU GAA UTT-3' and antisense, 5'-AUU CAG tions were evaluated, including fms related tyrosine kinase 3 GUA ACC AAA GGU GTT-3') and negative control (NC) siRNA (FLT3)-internal tandem duplication (FLT3-ITD, forward, (sense, 5'-UUC UCC GAA CGU GUC ACG UTT-3' and antisense, 5 '- C A A T T T AG G TAT GA A AG C C -3 ' a n d r ever s e, 5 '- G TA 5 ' - A CG UG A C A C G U U CG G A G A A T T - 3 ') w er e c om mercia l l y C C T TT C A G C A TT TT G A C - 3'), DNA methyltransferase 3α designed (Shanghai GenePharma Co., Ltd., Shanghai, China). (DNMT3A, forward, 5'-CTG CTG TGT GGT TAG ACG-3' and On the day of transfection, THP-1 cells were plated at a low reverse, 5'-TAT TTC CGC CTC TGT GGT TT-3'), FLT3-tyrosine density of 2 x 10 on the culture vessel (Corning, Corning, kinase domain (FLT3-TKD, forward, 5'-CCA GGA ACG TGC NY, USA) and subsequently transfected with 40 nM on-target T T G T CA-3 ' a n d r eve r s e, 5 '- T CA A A A AT G CAC CAC AG T siRNA using Lipofectamine™ 2000 (Invitrogen; Thermo GAG-3'), C-kit (forward, 5'-CTC CCT GAA AGC AGA AAC-3' Fisher Scientic fi , Inc.) according to the manufacturer's protocol a n d r e v e r s e, 5 ' - C AG AAA G AT AAC AC C AAA ATA G -3 ' ), cells were incubated with siRNA for 24-48 h. NC siRNA was ONCOLOGY LETTERS 17: 4935-4947, 2019 used as a transfection control in all experiments. Each experi- was considered to indicate a statistically signic fi ant difference. ment was independently repeated at least three times. SPSS statistical software version 18 (SPSS, Inc., Chicago, IL, USA) was used to perform all statistical analyses. Analyses of biological phenotype. To analyze cell migra- tion, 1x10 THP-1 cells (siZEB2-AS1 and NC groups) were Results plated into the upper chamber of Transwell cell culture inserts (24‑well; pore size, 8 µm; Corning) in serum‑free DMEM Clinical, cytogenetic and molecular features of patients with media (GE Healthcare Life Sciences). The lower chamber AML. The clinical features of the 62 AML cases enrolled in medium was supplemented with 20% FBS. To assess cell the present study were summarized in Table I. Based on the invasion, 1x10 THP-1 cells (siZEB2-AS1 and NC groups) modie fi d MRC classic fi ation, patients were categorized into a were seeded into the upper chamber of Transwell cell culture favorable risk group (n=29), an intermediate risk group (n=28) inser ts (24‑well; pore size, 8 µm; Corning) coated with and an adverse risk group (n=5). According to ELN recom- Matrigel. The lower chamber medium contained 20% FBS. mendations (12 cases missed the required mutation data and Following incubation at 37˚C for 24 h, the upper layer of were not classie fi d), patients were categorized into a favorable THP-1 cells was removed with cotton wool, and THP-1 cells risk group (n=27), an intermediate I/II risk group (n=13) and on the lower surface were fixed with 95% ethanol for 20 min an adverse risk group (n=10; Table I). The respective values of in room temperature. Invaded or migrated cells were subse- median OS and DFS rates, regarding different risk tiers and quently stained with 0.1% crystal violet for 30 min at 37˚C and treatment approaches, were summarized in Table III. observed under an IX71 inverted microscope at x200 magnifi - cation (Olympus Corporation, Tokyo, Japan). Five microscopic Expression of ZEB2‑AS1 lncRNA in patients with AML. e fi lds were counted per insert. Triplicate inserts were used for Using the Affymetrix Human LncRNA microarray, dozens each individual experiment, and each experiment was inde- of abnormally expressed lncRNAs were identie fi d in patients pendently repeated at least three times. with AML. ZEB2‑AS1 lncRNA was identie fi d to be substan - To assess cell proliferation, THP-1 cell lines (siZEB2-AS1 tially overexpressed in patients with AML with a karyotype of and NC groups) were seeded onto 96-well cell culture cluster 11q23 when compared with karyotypes such as t(15;17), t(8;21) plates (Corning) at a concentration of 5x10 cells/well in volumes and inv (16). Therefore, the ZEB-AS1 lncRNA was selected of 100 µl. A total of 10 µl Cell Counting Kit‑8 reagent (Dojindo, for further analysis in the following experiments. To further Kumamoto, Japan) was added to each well at the indicated time confirm the microarray results, RT‑qPCR was performed; the points (24, 48, 72, 96 h, 5, 7 days) and incubated in the dark results revealed that the expression levels of ZEB-AS1 lncRNA at 37˚C for a further 4 h. The absorbency was subsequently in the AML group (n=62) were signic fi antly higher compared measured at the wavelength of 450 nm with the Varioskan with that of the patients with non‑malignant hemotopathy (n=10; Flash Multimode Reader (Thermo Fisher Scientic fi , Inc.). Each P<0.001; Fig. 1A) and healthy volunteers (n= 4; P= 0.010; Fi g 1.A). experiment was independently repeated at least three times. In addition, the expression levels of ZEB-AS1 lncRNA were To detect cell apoptosis, 10X binding buffer (eBioscience; positively associated with increasing AML risk levels, according Thermo Fisher Scientic fi , Inc.) was initially diluted to 1X using to modie fi d MRC and ELN recommendations (Fig. 1B and C). distilled water (1 ml 10X binding buffer + 9 ml distilled water). Furthermore, the expression levels of ZEB-AS1 lncRNA were Cells were washed once in phosphate buffered saline and once signic fi antly higher in patients with AML that had not achieved in 1X binding buffer, prior to cell resuspension in 1X binding CR (n=15) compared with those who had (n=24) subsequent to buffer to 1-5x10 /ml. A total of 5 µl u fl orochrome‑conjugated the first induction of chemotherapy (P= 0.042; Fig. 1D). Annexin V (eBioscience; Thermo Fisher Scientic fi , Inc.) was added to 100 µl cell suspension and incubated for 10‑15 min ZEB2A ‑ S1 lncRNA expression and AML clinical outcomes. In the at room temperature. Cells were washed in 1X binding buffer present study, ZEB2-AS1 lncRNA expression levels greater than and resuspended in 200 µl 1X binding buffer. A total of 5 µl the 75th percentile were considered to be high expression levels 7-Aminoactinomycin D viability staining solution (2‑8˚C) whereas those below were considered to be low, respectively. (eBioscience; Thermo Fisher Scientific, Inc.) was added Overall, 42 AML cases had available survival data. As presented (stained within 4 h and stored at 2‑8˚C in the dark). Cells were in Fig. 2, Kaplan-Meier survival plots indicated that patients analyzed using 5-color o fl w cytometry ( type FC500, Beckman with AML with high ZEB2-AS1 lncRNA expression (n=6) had Coulter company, Fullerton, CA, USA). Each experiment was signic fi antly shorter OS (3‑year OS, 0.0 vs. 68.2%; P=0.036) and independently repeated at least three times. lower DFS rates (3‑year DFS, 25.0v s. 69.8%; P= 0.039) compared with that of the low expression subgroup (n=36). Additionally, Statistical analysis. All continuous data were expressed as according to the modie fi d MRC risk stratic fi ation, in the favor - the mean ± standard error of the mean. One-way analysis able/intermediate risk group, patients with AML with low of variance with Bonferroni's correction post-hoc test for ZEB2‑AS1 lncRNA expression (n=36) had signic fi antly longer multiple comparisons were performed. Survival probabilities OS (3-year OS, 68.2 vs. 33.3%; P= 0.026) and higher DFS rates were estimated using the Kaplan-Meier method and differ- (3‑year DFS, 69.8 vs. 33.3%; P= 0.038) compared with that of the ences between survival distributions were evaluated using the high expression subgroup (n=3; Fig. 3A). Furthermore, according log-rank test. Cox's proportional hazards model was applied to the ELN risk stratic fi ation, in the favorable /intermediate I/II to estimate the hazard ratio for disease-free survival (DFS) risk groups, patients with AML with low ZEB2-AS1 lncRNA and overall survival (OS) rates. For all analyses, the P-values expression (n=23) had significantly longer OS (3-year OS, were two‑tailed and the cond fi ence interval was 95%. P<0.05 69.9 vs. 50.0%; P= 0.034) and higher DFS rates (3‑year DFS, SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Table III. Clinical outcomes of patients with AML. OS rate (months) DFS rate (months) ----------------------------------------------------------------------- ---------------------------------------------------------------------- Groups Median (95% CI) P-value Median (95% CI) P-value Age (years) 0.187 0.199 <60 35.0 (30.33-43.89) 29.0 (27.42-41.72) ≥60 27.0 (16.47‑42.73) 24.0 (12.98‑41.82) Sex 0.254 0.246 Male 25.0 (23.99-42.01) 24.0 (20.66-39.97) Female 40.0 (30.15-47.59) 39.0 (27.44-45.60) White blood cell (x10 /l; non‑M3) 0.443 0.403 <Median 35.0 (27.74-51.39) 28.0 (24.48-48.77) ≥Median 25.0 (21.86‑41.55) 24.0 (17.79‑39.27) Hemoglobin (g/l; non‑M3) 0.988 0.924 <Median 28.0 (25.22-39.14) 27.0 (21.53-6.35) ≥Median 39.5 (25.00‑53.12) 38.5 (21.40‑50.98) Platelets (x10 /l; non‑M3) 0.199 0.262 <Median 40.0 (30.38-47.31) 39.0 (26.65-44.62) ≥Median 23.0 (16.93‑45.07) 21.0 (13.29‑43.00) Blasts in bone marrow (%; non‑M3) 0.590 0.512 <Median 33.0 (24.59-49.41) 25.5 (20.41-46.34) ≥Median 32.0 (24.53‑43.71) 29.0 (21.31‑41.87) MRC risk stratification 0.005 0.003 Favorable 38.0 (32.36-45.91) 29.0 (30.32-44.20) Intermediate 27.0 (20.92-45.32) 24.0 (17.24-43.11) Adverse 29.0 (12.02-45.98) 23.0 (5.13-40.88) ELN risk stratification (non‑M3) 0.003 0.003 Favorable 40.0 (32.29-53.18) 39.0 (29.52-51.01) Intermediate I/II 25.0 (15.25-37.35) 22.0 (10.43-35.37) Adverse 19.5 (11.51-36.99) 15.5 (6.1732.33) Treatment approaches (non-M3) 0.113 0.166 Chemotherapy 29.5 (23.15-41.05) 26.5 (19.84-38.56) Allo-HSCT 31.5 (23.80-47.77) 25.5 (19.23-44.91) ZEB2-AS1 level 0.036 0.039 Low level 36.5 (30.98-44.40) 31.5 (28.55-42.50) High level 22.0 (10.55-44.12) 20.0 (3.98-41.74) AML, acute myeloid leukemia; OS, overall survival; DFS, disease‑free survival; MRC, Medical Research Council; ELN, European Leukemia Net; HSCT, hematopoietic stem cell transplantation; CI, confidence interval; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1. 71.1 vs. 50.0%; P= 0.034) compared with that of the high expre- s ZEB2‑AS1 lncRNA expression and AML treatment response. sion subgroup (n=2; Fig. 3B). The expression levels of ZEB2‑AS1 Patients with AML with high ZEB2-AS1 lncRNA expres- lncRNA in the adverse risk group were all comparatively high. sion, excluding those classed as the M3 subtype, had a In the multivariate analyses, subsequent to control- significantly lower CR rate compared with that of the low ling for confounding variables in modified MRC expression subgroup (P= 0.031; Table I). However, differ- (favorable/intermediate vs. adverse) and ELN risk stratic fi ation ences between the OS and DFS rates in the consolidation (favorable/intermediate I/II vs. adverse) groups, high ZEB2-AS1 chemotherapy (n=10) and allo-HSCT treatment groups (n=14) lncRNA expression was determined to not be significantly were not signic fi antly different (P>0.05; Table III). Using the associated with adverse patient outcomes, including reduced OS stratic fi ation method to control for confounding variables, as (P= 0.976) and DFS rates (P= 0.725; TableI V) compared with the presented in Fig. 4, it was demonstrated that patients with a low ZEB2-AS1 lncRNA expression group. low ZEB2-AS1 lncRNA expression within the allo-HSCT ONCOLOGY LETTERS 17: 4935-4947, 2019 Figure 1. Association of the expression levels of ZEB2-AS1 lncRNA with different clinical features in patients with AML. (A) Expression levels of ZEB2-AS1 lncRNA in patients with AML were significant higher compared with that of non‑malignant hemotopathy patients and healthy volunteers ( P<0.05). (B) According to the modie fi d Medical Research Council risk stratic fi ation systems, the expression levels of ZEB2‑AS1 lncRNA exhibited a signic fi ant differ - ence between favorable, intermediate and adverse risk groups of patients with AML ( P<0.05). (C) According to the European Leukemia Net risk stratic fi ation systems, the expression levels of ZEB2‑AS1 lncRNA demonstrated a signic fi ant difference between favorable, intermediate and adverse risk groups of patients with AML ( P<0.05). (D) Expression levels of ZEB2‑AS1 lncRNA were significantly higher in patients who had not achieved CR compared with that of patients who had achieved CR subsequent to the first induction of chemotherapy ( P<0.05). Values are the mean ± the standard error of the mean. ZEB2-AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; AML, acute myeloid leukemia; CR, complete remission. Figure 2. Comparison of OS and DFS rates for patients with AML with different expression levels of ZEB2-AS1 lncRNA. The cut-off value of ZEB2-AS1 lncRNA was the 75th percentile. Expression levels greater or less than the 75th percentile were tentatively considered as the high or low expression groups, respectively. Patients with AML with a high expression of ZEB2‑AS1 lncRNA demonstrated a significantly shorter (A) OS (3‑year OS, 0.0 vs. 68.2%; P= 0.036) and (B) DFS (3‑year DFS, 25.0 vs. 69.8%; P= 0.039) compared with that of the low expression of ZEB2‑AS1 lncRNA group. OSo , verall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA. treatment group (n=11) had significantly longer OS (3‑year ZEB2‑AS1 lncRNA expression and AML cell biological OS, 75.8 vs. 28.6%; P= 0.037) and DFS rates (3‑year DFS, phenot ype. Knockdown of ZEB2-AS1 lncR NA by siR NA in 81.8 vs. 28.6%; P= 0.049) compared with that of the chemo - THP-1 cells significantly inhibited the mRNA expression therapy group (n=7). levels of ZEB2 (n= 4; P<0.05) and stimulated the mRNA SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML Figure 3. Comparison of OS and DFS rates for patients with AML in favorable/intermediate risk groups. (A) According to the modie fi d Medical Research Council risk stratification recommendation, in the favorable/intermediate risk group, patients with AML with a high expression of ZEB2‑AS1 lncRNA exhibited signic fi antly shorter OS (3‑year OS, 33.3 vs. 68.2%; P= 0.026) and DFS (3‑year DFS, 33.3v s. 69.8%; P= 0.038) rates compared with that of the low expression of ZEB2‑AS1 lncRNA group. (B) According to the European Leukemia Net risk stratic fi ation recommendation, in the favorable/intermediate risk group, patients with AML with a high expression of ZEB2‑AS1 lncRNA demonstrated signic fi antly shorter OS (3‑year OS, 50.0v s. 69.9%; P= 0.034) and DFS (3‑year DFS, 50.0 vs. 71.1%; P= 0.034) rates compared with that of the low expression of ZEB2‑AS1 lncRNA group. OS, overall survival; DFS d,i sease‑free survival; AML, acute myeloid leukemia; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA. Figure 4. Comparison of OS and DFS rates for patients with AML with a low expression of ZEB2 antisense RNA 1 long non-coding RNA treated using different treatment strategies. Patients with AML receiving allo‑HSCT exhibited signic fi ant longer (A) OS (3‑year OS, 75.8 vs. 28.6%; P= 0.037) and (B) DFS (3‑year DFS, 81.8 vs. 28.6%; P= 0.049) rates compared with that of the chemotherapy treatment group. OS, overall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; allo‑HSCT, allogenic hematopoietic stem cell transplantation. expression levels of E‑cadherin (n= 4; P<0.05) compared compared with the sham and vehicle groups (P<0.05). with the sham and vehicle groups (Fig. 5A and B). As THP-1 cell proliferation (n=4) and apoptosis (n=4) were presented in Fig. 6A and B, the migration (n=4) and not significantly different following the knockdown of invasion (n=4) of THP-1 cells, respectively, were signifi- ZEB2-AS1 lncRNA compared with the sham and vehicle cantly inhibited by the knockdown of ZEB2-AS1 lncRNA groups (Fig. 7A and B). ONCOLOGY LETTERS 17: 4935-4947, 2019 Figure 5. Effects of the downregulation of ZEB2-AS1 lncRNA on the mRNA expression levels of ZEB2 and E-cadherin in THP-1 cells. Subsequent to the knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells, the mRNA expression levels of (A) ZEB2 were signic fi antly decreased and (B) E‑cadherin were signifi - cantly increased. P<0.05 vs. Sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Figure 6. Effects of ZEB2-AS1 lncRNA on the cellular migration and invasion of THP-1 cells. The biological behavior of (A) migration and (B) inva- sion were significantly inhibited following the knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells. P<0.05 vs. Sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Figure 7. Effects of ZEB2-AS1 lncRNA on cellular proliferation and apoptosis in THP-1 cells. Following the knockdown of ZEB2-AS1 lncRNA, the (A) prolif- eration and (B) apoptosis exhibited no signic fi ant differences in THP‑1 cells compared with the sham and vehicle groups. Values are the mean ± standard error of the mean. ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1; lncRNA, long non‑coding RNA; si‑, small interfering RNA. Discussion functional roles in regulating biological behaviors. By retro- spectively analyzing 62 patients with de novo AML, the present In the present study, it was proposed that the abnormal over- study contributed novel results to the literature, demonstrating expression of ZEB2-AS1 lncRNA may be closely associated that the expression of ZEB2-AS1 lncRNA was abnormally with adverse outcomes in patients with AML, and may exert elevated and may have potential as an epigenetic biomarker SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML for evaluating the clinical outcomes of AML. Furthermore, it was revealed that ZEB2-AS1 lncRNA effectively modulated leukemic phenotypes including the invasion and migration of an AML cell line. The association between ZEB2-AS1 lncRNA expression and a series of clinical features, cytogenetic characteristics, somatic mutations and clinical outcomes in patients with AML was initially investigated. The results revealed that the expression of ZEB2‑AS1 lncRNA was signic fi antly higher in the AML group compared with that of a non-malignant group (P<0.001, Fig. 1A). This indicated that the overexpression of ZEB2‑AS1 lncRNA may function as a cancer‑specic fi molec - ular signal in leukemogenesis. With respect to the identie fi d karyotype and recurrent mutations, according to either the modified MRC or ELN risk stratification recommenda- tions (37-39), the expression levels of ZEB2-AS1 lncRNA had a signic fi ant stepwise increase from the favorable to adverse risk group (all P<0.05, Fig. 1B and C). Additionally, the high expression of ZEB2-AS1 lncRNA was associated with adverse patient outcomes compared with the low expression of ZEB2-AS1 lncRNA (Table I). This suggests that the over- expression of ZEB2-AS1 lncRNA is closely associated with a higher risk in AML. Accumulating evidence has revealed that numerous lncRNAs may independently predict the prognosis for patients with cancer (23,42-44). Notably, previous research has indi- cated that lncRNAs expression profiles have the potential to independently predict clinical outcomes in AML (21). Novel studies have further demonstrated that, due to its extensive oncogenic functions, the overexpression of HOTAIR lncRNA predicts poor clinical outcomes in AML and may be a poten- tial therapeutic target (25). In the present study, to evaluate the potential of its prognostic application, the impact of ZEB2-AS1 lncRNA overexpression on OS and DFS rates in patients with AML was measured. Univariate analyses revealed that patients with high ZEB2-AS1 lncRNA expression had signic fi antly shorter OS and DFS rates compared with the low ZEB2-AS1 lncRNA expression group (all P<0.05, Table III; Fig. 2). However, using the method of multivariate analyses to control for confounding variables, the adjusted 3‑year OS and DFS rates were not signic fi antly different between patients with AML with different expression levels (high vs. low) of ZEB2-AS1 lncRNA (Table IV). Furthermore, it was demon- strated that in the favorable/intermediate risk group, patients with a higher expression of ZEB2-AS1 lncRNA exhibited signic fi antly shorter OS and DFS rates compared with those with a lower expression (all P<0.05, Fig. 3). Therefore, it was proposed that although its independent prognostic value for survival was not rigorously ascertained, the expression levels of ZEB2-AS1 lncRNA may function as a complementary factor to further improve conventional risk stratification models in AML. Treatment responses in AML are notably heteroge- neous and affected by various demographic and biological factors (45-47). To minimize influences of different treat- ment approaches on clinical outcome interpretation, patients enrolled in the present cohort were treated with an identical induction chemotherapy regimen. It was revealed that the expression levels of ZEB2‑AS1 lncRNA were significantly higher in patients who did not achieve initial CR compared with Table IV. Analysis of 3-year OS and 3-year DFS rates in patients with AML. 3-Year OS 3-Year DFS ---------------------------------------------------------------------------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------- Univariate Multivariate Univariate Multivariate ----------------------------------------------------------------- -------------------------------------------------------------- ---------------------------------------------------------------- ------------------------------------------------------------- Factors HR (95% CI) P-value HR (95% CI) P-value HR (95% CI) P-value HR (95% CI) P-value MRC: F/I vs. adverse 4.04 (1.06-227.70) 0.047 0.51 (0.05-4.78) 0.600 4.48 (1.33-373.10) 0.031 0.68 (0.08-6.18) 0.927 ELN: F/I I/II vs. adverse 3.95 (1.78-66.90) 0.011 4.07 (1.25-13.22) 0.020 4.39 (2.17-95.66) 0.006 4.52 (1.40-14.69) 0.012 ZEB2-AS1 level: low vs. high 3.20 (1.15-32.24) 0.036 1.28 (0.24-6.89) 0.976 3.17 (1.10-30.42) 0.039 1.04 (0.19-5.75) 0.725 OS, overall survival; DFS, disease‑free survival; AML, acute myeloid leukemia; HR, hazard ratio; CI, confidence interval; MRC, Medical Research Council; F/I, favorable/intermediate; ELN, European Leukemia Net; HSCT, hematopoietic stem cell transplantation; ZEB2‑AS1, zinc finger E‑box binding homeobox 2 antisense RNA 1. ONCOLOGY LETTERS 17: 4935-4947, 2019 those who did (P=0.042, Fig. 1D). In addition, patients with a potential leukemogenic regulator in tumorigenesis (53). In a high expression of ZEB2‑AS1 lncRNA had a signic fi antly the present study, the results revealed that the knockdown of lower CR rate compared with those with a low expression of ZEB2-AS1 lncRNA in THP-1 cells effectively downregulated ZEB2-AS1 lncRNA (P= 0.031, Table I). This demonstrated the mRNA expression of ZEB2 (Fig. 5A), which was consis- that the overexpression of ZEB2-AS1 lncRNA may be closely tent with the results of a previous study (35). In addition, the associated with chemotherapy resistance. Following front-line knockdown of ZEB2‑AS1 lncRNA in THP‑1 cells signic fi antly induction chemotherapy, patients with AML received either increased the mRNA expression of E-cadherin compared with consolidation chemotherapy or allo-HSCT. The results the sham/vehicle groups (a alll P l P< <0 0..0 05 5, F , F Fiiig g g. 5 . 5 . 5B B B) ) ), a , a , an n nd t d t d th h he r e r e re e ep p pr r re e es- s- s- revealed that differences between the 3-year OS and DFS sion of E-cadherin has been demonstrated to trigger the EMT rates in the consolidation chemotherapy and allo-HSCT in cancer progression (54). In addition, it was revealed that treatment groups were not signic fi antly different (Table III). cellular migration and invasion were signic fi antly inhibited in Furthermore, using stratic fi ation to control for confounding THP-1 cells following ZEB2-AS1 lncRNA downregulation by variables, patients with low ZEB2-AS1 lncRNA expression si-ZEB2-AS1 (all P<0.05, Fig. 6). However, the knockdown of in the allo‑HSCT‑treated group had a signic fi antly longer OS ZEB2-AS1 lncRNA in THP-1 cells had no apparent effects and DFS compared with that of the chemotherapy group (all on proliferation and apoptosis (Fig. 7). Accordingly, it was P<0.05, Fig. 4). This suggested that patients with AML with a proposed that ZEB2-AS1 lncRNA may have upregulated low expression of ZEB2-AS1 lncRNA may be more sensitive ZEB2 expression, which in turn enhanced the motility pheno- to allo-HSCT therapy. In comparison with the relatively static type of THP-1 cells, including their invasion and migration genetic landscape, epigenetic status, including DNA methyla- abilities in vitro. These results suggested that chemotherapy tion, is altered during different phases of AML (13). However, resistance in patients with a high expression of ZEB2-AS1 dynamic changes in ZEB2-AS1 lncRNA expression in this lncRNA may be closely associated with enhanced cellular retrospective cohort could not be evaluated, and the associa- migration and invasion during leukemic progression. This tion of ZEB2-AS1 lncRNA with treatment responses requires would be consistent with the results of a previous study that further consolidation by future prospective studies. demonstrated that ZEB family protein expression may predict The limitations of the present clinical study should be differential responses to various chemotherapy drugs in hema- considered carefully. From a clinical point of view, therapeutic topoietic malignancies, including mantle cell lymphoma (51). strategies for AML have improved in previous years (47). In conclusion, to the best of our k nowledge, the present study Additionally, cytogenetic/molecular risk stratic fi ation systems was the first to evaluate the prognostic value of ZEB2‑AS1 remain controversial and require further improvement (7,9,11). lncRNA in AML. The results demonstrated that the over- These all affect the prognostic importance of ZEB2-AS1 expression of ZEB2-AS1 lncRNA was associated with poor lncRNA in AML to varying degrees. Furthermore, from a clinical outcomes in AML. Although the independent prog- statistical point of view, the retrospective cohort size in the nostic prediction for survival was not rigorously researched present study was relatively small, which may have affected in the present study, the overexpression of ZEB2-AS1 the multivariate analysis results. Thus, a future study with lncRNA may function as a candidate gene to improve cyto- a large cohort must be retrospectively and/or prospectively genetic/somatic mutation risk stratic fi ation systems in AML. analyzed to further assess the prognostic value of ZEB2-AS1 Furthermore, it was discovered that ZEB2-AS1 lncRNA lncRNA for patient survival and treatment response in AML, effectively modulated the leukemic phenotypes of invasion independent of various approved clinical factors. and migration, which may be associated with the differential Increasing evidence has demonstrated that various responses to treatment strategies. Finally, another question lncRNAs serve essential functions in not only in intrinsic must be addressed-why is ZEB2-AS1 lncRNA overexpressed cellular regulatory networks, but also in intercellular commu- in AML? It was noted that the expression levels of ZEB2-AS1 nications during tumorigenesis (23). Multiple lncRNAs have lncRNA in the AML group exhibit high heterogeneity, with been confirmed to function as original drivers and/or down - greater variability compared with that of the non-malignant stream targets in circuits involving almost all hallmarks of group. Furthermore, in patients with AML with a karyotype cancer, including proliferation, viability, immortality, motility, of 11q23, ZEB2-AS1 lncRNA expression was notably high. angiogenesis and tumor suppression (23,48). Previously, it has AML pathogenesis with a 11q23 karyotype involves the been reported that a natural antisense transcript of ZEB2 may abnormal rearrangement of the mixed lineage leukemia regulate the EMT in different tumor types, including colon gene, which is an epigenetic modifier involved in histone adenocarcinoma (35). The highly‑conserved zinc‑finger struc - methylation (55). It was hypothesized that the overexpression ture of ZEB2 binds to E-boxes located in the promoter regions of ZEB2-AS1 lncRNA may not be a key event during leuke- of certain target genes including E-cadherin, so as to further mogenesis, but a downstream target of other key oncogenic regulate EMT in cancer progression (49,50). Novel results events. As previously mentioned, the expression profiling of have demonstrated that abnormal expression of ZEB1 may lncRNAs is able to independently evaluate survival in older promote cellular proliferation and tumor growth in mantle cell patients with cytogenetically normal AML (21). A novel study lymphoma (51). Furthermore, ZEB1 expression is controlled revealed that the combination of >1 lncRNA (i.e. a six-lncRNA by growth arrest specific 5-AS1 lncRNA to modulate cell signature) may be strongly associated with survival in diffuse migration and invasion in non-small cell lung cancer (52). large B-cell lymphoma (44). This suggests that one lncRNA Notably, a novel ZEB2-BAF chromatin remodeling complex alone may be not sufci fi ent in independently predicting the subunit BCL11B fusion gene has been identie fi d in patients survival of patients with AML. These key problems require with AML with karyotype of t(2;14)(q22;q32), which may be thorough investigation in the future. SHI et al: ZEB2-AS1 lncRNA PREDICTS POOR PROGNOSIS IN AML 4. Almeida AM and Ramos F: Acute myeloid leukemia in the older Acknowledgements adults. Leuk Res Rep 6: 1-7, 2016. 5. Percival ME, Tao L, Medeiros BC and Clarke CA: Improvements Not applicable. in the early death rate among 9,380 patients with acute myeloid leukemia after initial therapy: A SEER database analysis. Cancer 121: 2004-2012, 2015. Funding 6. Bhatnagar B and Garzon R: The use of molecular genetics to refine prognosis in acute myeloid leukemia. Curr Hematol Malig Rep 9: 148-157, 2014. This work was supported by the Priority Academic 7. Stölzel F, Mohr B, Kramer M, Oelschlägel U, Bochtler T, Program Development of Jiangsu Higher Education Berdel WE, Kaufmann M, Baldus CD, Schäfer-Eckart K, Institutions, the National Clinical Key Subject Project, Stuhlmann R, et al: Karyotype complexity and prognosis in acute myeloid leukemia. Blood Cancer J 6: e386, 2016. the Innovation Capability Development Project of Jiangsu 8. Mrózek K and Bloome fi ld CD: Chromosome aberrations, gene Province (grant no. BM2015004), the Natural Science mutations and expression changes, and prognosis in adult acute Foundation of China (grant nos. 81570139 and 81270617), myeloid leukemia. Hematology Am Soc Hematol Educ Program 2006: 169-177, 2006. the Jiangsu Provincial Special Program of Medical Science 9. Papaemmanuil E, Gerstung M, Bullinger L, Gaidzik VI, (grant no. BL2012005), Jiangsu Province's Key Medical Center Paschka P, Roberts ND, Potter NE, Heuser M, Thol F, (grant no. ZX201102), the Jiangsu Province Natural Science Bolli N, et al: Genomic classification and prognosis in acute myeloid leukemia. N Engl J Med 374: 2209-2221, Fund (grant no. BE2015639), the Project of Natural Science Foundation of Jiangsu Province (grant no. BK20141201) and 10. Byun JM, Kim YJ, Yoon HJ, Kim SY, Kim HJ, Yoon J, Min YH, the National Natural Science Foundation of China Youth Fund Cheong JW, Park J, Lee JH, et al: Cytogenetic profiles of 2,806 patients with acute myeloid leukemia-a retrospective Project (grant no. 81300424). multicenter nationwide study. Ann Hematol 95: 1223-1232, Availability of data and materials 11. Li S, Mason CE and Melnick A: Genetic and epigenetic hetero- geneity in acute myeloid leukemia. Curr Opin Genet Dev 36: 100-106, 2016. The datasets used and analyzed during the current study 12. Cancer Genome Atlas Research Network, Ley TJ, Miller C, are available from the corresponding author on reasonable Ding L, Raphael BJ, Mungall AJ, Robertson A, Hoadley K, Triche TJ Jr, Laird PW, et al: Genomic and epigenomic request. landscapes of adult de novo acute myeloid leukemia. N Engl J Med 368: 2059-2074, 2013. Authors' contributions 13. Li S, Ga r rett-Ba kelman F E, Chung SS, Sanders MA, H r ici k T, Rapapor t F, Pat el J, Di l lon R, Vijay P, Brown A L, et al: Distinct evolution and dynamics of epigenetic and genetic XS finished all the experiments, analyzed and interpreted the heterogeneity in acute myeloid leukemia. Nat Med 22: patient data, and was a major contributor in writing the manu- 792-799, 2016. script. JL finished most of the experiments. LM, LW, QW and 14. Sasca D and Huntly BJ: Independence of epigenetic and genetic diversity in AML. Nat Med 22: 708-709, 2016. HY performed some experiments. CR and DW had substantial 15. Mattick JS and Makunin IV: Non-coding RNA. Hum Mol contributions to the conception and design of the work. XZ Genet 15 Spec No 1: R17-R29, 2006. and SC designed the study and took final responsibility. All 16. Morris KV and Mattick JS: The rise of regulatory RNA. Nat Rev Genet 15: 423-437, 2014. authors read and approved the final manuscript. 17. Djeba l i S, Davis CA, Merkel A, Dobin A, Lassmann T, Mortazavi A, Tanzer A, Lagarde J, Lin W, Schlesinger F, et al: Ethics approval and consent to participate Landscape of transcription in human cells. Nature 489: 101-108, 18. Iyer MK, Niknafs YS, Malik R, Singhal U, Sahu A, Hosono Y, The present study was approved by the Institutional Review Barrette TR, Prensner JR, Evans JR, Zhao S, et al: The land- Board of the First Affiliated Hospital of Soochow University scape of long noncoding RNAs in the human transcriptome. Nat Genet 47: 199-208, 2015. Written informed consent was obtained from all participants 19. Goff LA and Rinn JL: Linking RNA biology to lncRNAs. for the use of bone marrow specimens and clinical information. Genome Res 25: 1456-1465, 2015. 20. Ling H, Vincent K, Pichler M, Fodde R, Berindan-Neagoe I, Slack FJ and Calin GA: Junk DNA and the long non-coding Patient consent for publication RNA twist in cancer genetics. Oncogene 34: 5003-5011, 2015. 21. Garzon R, Volinia S, Papaioannou D, Nicolet D, Kohlschmidt J, Not applicable. Yan PS, Mrózek K, Bucci D, Carroll AJ, Baer MR, et al: Expression and prognostic impact of lncRNAs in acute myeloid leukemia. Proc Natl Acad Sci USA 111: 18679-18684, 2014. Competing interests 22. Morlando M, Ballarino M and Fatica A: Long non-coding RNAs: New players in hematopoiesis and leukemia. Front Med (Lausanne) 2: 23, 2015. The authors declare that they have no competing interests. 23. Schmitt AM and Chang HY: Long noncoding RNAs in cancer pathways. Cancer Cell 29: 452-463, 2016. 24. Zhang X, Lian Z, Padden C, Gerstein MB, Rozowsky J, Snyder M, References Gingeras TR, Kapranov P, Weissman SM and Newburger PE: A myelopoiesis-associated regulatory intergenic noncoding 1. Döhner H, Weisdorf DJ and Bloomfield CD: Acute myeloid RNA transcript within the human HOXA cluster. Blood 113: leukemia. N Engl J Med 373: 1136-1152, 2015. 2526-2534, 2009. 2. Polednak AP: Recent improvement in completeness of inci- 25. Xing CY, Hu XQ, Xie FY, Yu ZJ, Li HY, Bin-Zhou, Wu JB, dence data on acute myeloid leukemia in US cancer registries. Tang LY and Gao SM: Long non-coding RNA HOTAIR modu- J Registry Manag 41: 77-84, 2014. lates c-KIT expression through sponging miR-193a in acute 3. Tawfik B, Pardee TS, Isom S, Sliesoraitis S, Winter A, Lawrence J, myeloid leukemia. FEBS Lett 589: 1981-1987, 2015. Powell BL and Klepin HD: Comorbidity, age, and mortality among 26. Ma L, Bajic VB and Zhang Z: On the classification of long adults treated intensively for acute myeloid leukemia (AML). J Geriatr non-coding RNAs. RNA Biol 10: 925-933, 2013. Oncol 7: 24-31, 2016. ONCOLOGY LETTERS 17: 4935-4947, 2019 27. Mattick JS and Rinn JL: Discovery and annotation of long 41. Liva k KJ and Sch m it tgen TD: Analysis of relative gene noncoding RNAs. Nat Struct Mol Biol 22: 5-7, 2015. expression data using real-time quantitative PCR and the 28. Khorkova O, Myers AJ, Hsiao J and Wahlestedt C: Natural anti- 2(-Delta Delta C(T)) method. Methods 25: 402-408, 2001. sense transcripts. Hum Mol Genet 23: R54-R63, 2014. 42. Wang KC and Chang HY: Molecular mechanisms of long 29. Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, noncoding RNAs. Mol Cell 43: 904-914, 2011. Nakamura M, Nishida H, Yap CC, Suzuki M, Kawai J, et al: 43. Bhan A and Mandal SS: lncRNA HOTAIR: A master regulator Antisense transcription in the mammalian transcriptome. of chromatin dynamics and cancer. Biochim Biophys Acta 1856: Science 309: 1564-1566, 2005. 151-164, 2015. 30. ENCODE Project Consortium. An integrated encyclopedia of 44. Sun J, Cheng L, Shi H, Zhang Z, Zhao H, Wang Z and Zhou M: A DNA elements in the human genome. Nature 489: 57-74, 2012. potential panel of six-long non-coding RNA signature to improve 31. Su WY, Xiong H and Fang JY: Natural antisense transcripts survival prediction of diffuse large-B-cell lymphoma. Sci Rep 6: regulate gene expression in an epigenetic manner. Biochem 27842, 2016. Biophys Res Commun 396: 177-181, 2010. 45. Schlenk RF, Dohner K, Krauter J, Fröhling S, Corbacioglu A, 32. Wight M and Werner A: The functions of natural antisense tran- Bullinger L, Habdank M, Späth D, Morgan M, Benner A, et al: scripts. Essays Biochem 54: 91-101, 2013. Mutations and treatment outcome in cytogenetically normal 33. Ebralidze AK, Guibal FC, Steidl U, Zhang P, Lee S, Bartholdy B, acute myeloid leukemia. N Engl J Med 358: 1909-1918, 2008. Jorda MA, Petkova V, Rosenbauer F, Huang G, et al: PU.1 46. Xu Y, Sun Y, Shen H, Ding L, Yang Z, Qiu H, Sun A, Chen S and expression is modulated by the balance of functional sense and Wu D: Allogeneic hematopoietic stem cell transplantation could antisense RNAs regulated by a shared cis-regulatory element. improve survival of cytogenetically normal adult acute myeloid Genes Dev 22: 2085-2092, 2008. leukemia patients with DNMT3A mutations. Am J Hematol 90: 34. Sun J, Li W, Sun Y, Yu D, Wen X, Wang H, Cui J, 992-997, 2015. Wang G, Hoffman AR and Hu JF: A novel antisense long 47. Dombret H and Gardin C: An update of current treatments for noncoding RNA within the IGF1R gene locus is imprinted adult acute myeloid leukemia. Blood 127: 53-61, 2016. in hematopoietic malignancies. Nucleic Acids Res 42: 48. Hanahan D and Weinberg RA: Hallmarks of cancer: The next 9588-9601, 2014. generation. Cell 144: 646-674, 2011. 35. Beltran M, Puig I, Peña C, García JM, Alvarez AB, Peña R, 49. Vandewalle C, Van Roy F and Berx G: The role of the ZEB Bonilla F and de Herreros AG: A natural antisense transcript family of transcription factors in development and disease. Cell regulates Zeb2/Sip1 gene expression during Snail1-induced Mol Life Sci 66: 773-787, 2009. epithelial-mesenchymal transition. Genes Dev 22: 756-769, 50. Hill L, Browne G and Tulchinsky E: ZEB/miR-200 feedback 2008. loop: At the crossroads of signal transduction in cancer. Int 36. A rber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, J Cancer 132: 745-754, 2013. Le Beau MM, Bloome fi ld CD, Cazzola M and Vardiman JW: The 51. Sánchez-Tilló E, Fanlo L, Siles L, Montes-Moreno S, 2016 revision to the World Health Organization classic fi ation of Moros A, Chiva-Blanch G, Estruch R, Martinez A, Colomer D, myeloid neoplasms and acute leukemia. Blood 127: 2391-2405, Győrffy B, et al: The EMT activator ZEB1 promotes tumor 2016. growth and determines differential response to chemotherapy in 37. Metzeler KH, Herold T, Rothenberg-Thurley M, Amler S, mantle cell lymphoma. Cell Death Differ 21: 247-257, 2014. Sauerland MC, Görlich D, Schneider S, Konstandin NP, 52. Wu Y, Lyu H, Liu H, Shi X, Song Y and Liu B: Downregulation Dufour A, Bräundl K, et al: Spectrum and prognostic relevance of the long noncoding RNA GAS5-AS1 contributes to tumor of driver gene mutations in acute myeloid leukemia. Blood 128: metastasis in non-small cell lung cancer. Sci Rep 6: 31093, 2016. 686-698, 2016. 53. Torkildsen S, Gorunova L, Beiske K, Tjonnfjord GE, Heim S 38. Grimwade D, Hills RK, Moorman AV, Walker H, Chatters S, and Panagopoulos I: Novel ZEB2-BCL11B fusion gene identi- Goldstone AH, Wheatley K, Har rison CJ and Burnett AK; e fi d by RNA‑sequencing in acute myeloid leukemia with t(2;14) National Cancer Research Institute Adult Leukaemia Working (q22;q32). PLoS One 10: e0132736, 2015. Group: Refinement of cytogenetic classic fi ation in acute myeloid 54. Schmalhofer O, Brabletz S and Brabletz T: E-cadherin, leukemia: Determination of prognostic significance of rare beta-catenin, and ZEB1 in malignant progression of cancer. recurring chromosomal abnormalities among 5876 younger Cancer Metastasis Rev 28: 151-166, 2009. adult patients treated in the United Kingdom Medical Research 55. Stein EM and Tallman MS: Mixed lineage rearranged leukaemia: Council trials. Blood 116: 354-365, 2010. Pathogenesis and targeting DOT1L. Curr Opin Hematol 22: 39. Röllig C, Bornhäuser M, Thiede C, Taube F, Kramer M, Mohr B, 92-96, 2015. Aulitzky W, Bodenstein H, Tischler HJ, Stuhlmann R, et al: Long-term prognosis of acute myeloid leukemia according to the new genetic risk classic fi ation of the European LeukemiaNet This work is licensed under a Creative Commons recommendations: Evaluation of the proposed reporting system. Attribution-NonCommercial-NoDerivatives 4.0 J Clin Oncol 29: 2758-2765, 2011. 40. McGowan-Jordan J, Si mons A and Sch m id M (eds): An International (CC BY-NC-ND 4.0) License. International System for Human Cytogenomic Nomenclature (2016). Cytogenetic and Genome Research 149: No. 1-2, 2016.

Journal

Oncology LettersPubmed Central

Published: Mar 15, 2019

There are no references for this article.