CEBPE Expression Is an Independent Prognostic Factor for Acute Myeloid Leukemia Kening Li1,2†, Yuxin Du1,2†, Dong‑Qing Wei1* and Fang Zhang3*†

Li et al. J Transl Med (2019) 17:188 https://doi.org/10.1186/s12967-019-1944-x Journal of Translational Medicine

RESEARCH Open Access CEBPE expression is an independent prognostic factor for acute myeloid leukemia Kening Li1,2†, Yuxin Du1,2†, Dong‑Qing Wei1* and Fang Zhang3*†

Abstract Background: Identifying reliable predictive markers is important to make therapeutic decisions, and determine the prognosis for acute myeloid leukemia (AML) patients. However, approximately 50% patients could not be accurately predicted by existing risk factors. It is necessary to identify novel prognostic factors to subdivide the intermediate-risk group or patients without any cytogenetic and molecular abnormalities. Methods: Kaplan–Meier and Cox regression were used for survival analyses in three independent AML datasets. Analyses integrating both bioinformatics and ChIP-qPCR experiments were performed to explore the role of CEBPE in regulating the expression of known prognostic factors. Results: CEBPE expression was an independent predictor for both overall survival (OS) and event-free survival (EFS) of AML patients. Moreover, low-expression of CEBPE was found to be associated with high relapse rate. We also proved that diferential expression of CEBPE stratifed the wild-type patients of multiple genes into good and poor outcomes. In addition, the results showed that no obvious improvement was achieved by allogeneic transplantation in CEBPE high-expressed group, while the survival rate (both OS and EFS) was signifcantly increased in transplanted patients that with low expression of CEBPE. Finally, we found that CEBPE might regulate the expression of known prognostic factors by localizing on their promoters. Conclusion: Our fndings indicated that CEBPE expression was an independent prognostic factor for AML survival, relapse and allogeneic transplantation, which will provide useful information for outcome prediction and therapeutic decisions. Keywords: Acute myeloid leukemia, Survival, Prognostic factors, Relapse, Allogeneic transplantation

Background are the major treatments of AML [3, 4]. According to the Acute myeloid leukemia (AML) is an aggressive malig- acquired cytogenetic and molecular alterations at diag- nancy and the most typical leukemia in adults, which is nosis, we could stratify the patients into diferent prog- characterized by excessive proliferation, diferentiation nostic categories, and predict the relapse risk, survival failure and apoptosis disorder, resulted in the abnormal time, drug response and whether a potentially curative accumulation of myeloblasts in the bone marrow and allogeneic transplantation is possible [5]. Terefore, iden- peripheral blood [1]. A majority of patients with AML tifying reliable predictive markers is important in person- will relapse after achieving complete remission [2]. At alized therapy of AML. present, chemotherapy and/or allogeneic transplantation Some risk factors were identified by previous studies, and used to predict treatment outcome for AML *Correspondence: [email protected]; [email protected] patients. For example, patients with genomic transloca- †Kening Li, Yuxin Du and Fang Zhang contributed equally to this work tions such as t(15;17) (lead to PML–RARa fusion pro- 1 State Key Laboratory of Microbial Metabolism and School of Life tein), t(8;21) (lead to AML1–ETO fusion protein) and Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China inv(16) (lead to CBFb–MYH11 fusion protein) were 3 Key Laboratory of Systems Biomedicine, Shanghai Center for Systems classified in the favorable-risk group, cytogenetically Biomedicine, Shanghai Jiao Tong University, Shanghai 200240, China normal AML (CN-AML) were in the intermediate-risk Full list of author information is available at the end of the article

© The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Li et al. J Transl Med (2019) 17:188 Page 2 of 11

group, and those with a complex karyotype were clas- Materials and methods sified in the adverse-risk group [6]. Moreover, some Gene expression data of AML patients molecular abnormalities, including mutations of TP53, We used three independent AML datasets in this study, CEBPA, FLT3, DNMT3A were also found to provide including Te Cancer Genome Atlas (TCGA), GSE1159 important prognostic information, especially for CN- and GSE10358. Only samples with both gene expression AML patients [7, 8]. For example, mutations in CEBPA data and clinical annotations were kept. RNA-Seq data of are associated with a good outcome [9]; internal tan- 184 clinically annotated adult cases of AML were down- dem duplications in FLT3 (FLT3–ITD) adversely affect loaded from TCGA [5]. Microarray data of 260 AML the clinical outcome [10]. Mutations with prognostic patients were downloaded from GSE1159 [19, 20]. And implications in a number of other genes (e.g., TET2 microarray data of 91 AML patients were downloaded [11], ASXL1 [12, 13], DNMT3A [14], p53 [15] and KIT from GSE10358 [21]. Microarray data and cytogenetic [12]) have also been identified. To facilitate the pre- risk of each sample in GSE14468 [22, 23] were also used diction of treatment outcome of AML, a standardized in this study. system was proposed by an international expert panel in 2010 (working on behalf of the European Leukemi- Cell culture aNet (ELN)) [16]. Based on the published data on the Te AML cell lines NB4 and Kasumi-1 were obtained prognostic significance of cytogenetic and molecu- from the American Type Culture Collection (ATCC; lar alterations, ELN stratified the patients into four Manassas, VA, USA), and cultured in RPMI 1640 groups: favorable, intermediate-I, intermediate-II and medium (Termo Fisher Scientifc, Waltham, MA, USA), adverse. This system refined the classification of AML supplemented with 10% heat-inactivated fetal bovine prognosis [6]. serum (GIBCO-BRL), 100 U/mL penicillin and 100 mg/ However, the existing risk factors still could not mL streptomycin (GIBCO-BRL). All cells were incubated effectively predict the outcome of AML patients for in a humifed 5% CO2 at 37 °C. the following reasons. Firstly, approximately 50% patients are CN-AML which is not associated with Chromatin immunoprecipitation (ChIP) assay large chromosomal abnormalities [17]. The relapse ChIP assay of NB4 and Kasumi-1 cells was conducted rate and survival time of these CN-AML patients are by the manufacturer’s Active Motif protocol. Chroma- difficult to predict because of high heterogeneity [18]. tin extracts were immunoprecipitated with anti-CEBPE Secondly, although some gene mutations have statis- (Santa Cruz Biotechnology, sc-158) and rabbit IgG tical significance in predicting survival time of AML (Abcam, ab172730) was used as negative control anti- (especially for CN-AML), the mutation rates of these bodies. ChIP-qPCR was conducted to analyze immu- genes are relatively low. For example, AML patients noprecipitated DNA using SYBR Green PCR Master with TP53 mutation are predicted to have adverse Mix (Toyobo, Osaka, Japan) and the ABI Prism 7900HT outcome, but only approximately 5% AML patients detection system (Termo Fisher Scientifc). Fold enrich- are with TP53 mutation [5]. The majority of patients ment of ChIP DNA vs. input DNA was calculated. Te are unpredictable based on gene mutation. Moreover, primers were designed to cover regions that are shown in a significant proportion of patients are classified in Additional fle 1: Table S1. intermediate-risk group according to the ELN standardized system [6], but the prognosis of these patients Statistical analyses varies, some individuals respond well to chemother- Survival was estimated according to the Kaplan–Meier apy based consolidation regimens while others may method. Te log-rank test was used to assess statistical require allogeneic transplantation. Therefore, it is nec- signifcance. Cox regression was used to assess the asso- essary to identify novel prognostic factors to subdivide ciation of a given variable with OS or EFS. Multivariable the intermediate-risk group or patients without any testing was performed using Cox proportional hazards cytogenetic and molecular abnormalities. models. P values < 0.05 were considered statistically sig- In this study, we found that CEBPE, as a master nifcant. All of the statistical analyses were conducted transcription regulator of myeloid differentiation, was using R package “Survival”. an independent predictor for both overall survival (OS) and event-free survival (EFS) of AML patients. Results Moreover, CEBPE expression was observed to have CEBPE is actively expressed in AML patients with favorable prognostic power for AML relapse. Also, CEBPE outcome expression was a potential factor for directing alloge- We collected AML gene expression data from TCGA, neic transplantation. GSE14468 and GSE1159. Te three independent datasets Li et al. J Transl Med (2019) 17:188 Page 3 of 11

contained 184, 186 and 260 samples, respectively. Te TCGA (n = 184), GSE1159 (n = 260) and GSE10358 information of prognosis classifcation based on cytoge- (n = 91). In each dataset, we ranked the samples accord- netic factors was also obtained. Te results showed that ing to CEBPE expression, and samples of the top quar- CEBPE was highly expressed in patients with good prog- tile were classifed in high-expressed group, while others nosis. And this observation was confrmed in all of the were classifed in low-expressed group. As expected, three independent datasets. Te t-test P-values of CEBPE Kaplan–Meier survival analyses demonstrated that diferential expression between good and poor patients decreased expression value of CEBPE was signifcantly were 5.021e−05, 2.813e−11, 1.217e−6, respectively (P < 0.05) associated with shorter OS and EFS (Fig. 2). In (Fig. 1a). Moreover, we also found that patients with high datasets of TCGA, GSE1159 and GSE10358, the 5-year expression of CEBPE tended to have good prognosis in overall survival rates were 38%, 47% and 59% in CEBPE TCGA datasets (Fig. 1b). high-expressed group, while 17%, 29% and 35% in CEBPE low-expressed group. Signifcant diference was also CEBPE expression is an independent predictor of AML observed in OFS analysis. We then validated the prognostic impact of CEBPE In addition, univariable Cox regression analysis dem- expression in three independent AML datasets, namely onstrated that patients with higher CEBPE expression

a TCGA GSE14468 GSE1159 14 10 • • • • p=5.021e-05 p= 2.813e-11 p=1.217e-06 9 8 • 12 • • 8 6

10 7

6 • CEBPE expression(log2)

CEBPE expression(log2) CEBPE expression(log2) 4

8 5 •

Poor Good Poor Good Poor Good b

• 14 • • •

• • • •

12 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 10 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 68 • CEBPE expression • TCGA−AB−2903 TCGA−AB−3008 TCGA−AB−2829 TCGA−AB−2922 TCGA−AB−2816 TCGA−AB−2820 TCGA−AB−2938 TCGA−AB−2822 TCGA−AB−2811 TCGA−AB−2835 TCGA−AB−2825 TCGA−AB−2968 TCGA−AB−3011 TCGA−AB−2850 TCGA−AB−2941 TCGA−AB−2824 TCGA−AB−2984 TCGA−AB−2890 TCGA−AB−2807 TCGA−AB−2977 TCGA−AB−2987 TCGA−AB−2996 TCGA−AB−2979 TCGA−AB−2885 TCGA−AB−2959 TCGA−AB−2805 TCGA−AB−2926 TCGA−AB−2863 TCGA−AB−2995 TCGA−AB−2837 TCGA−AB−2817 TCGA−AB−2854 TCGA−AB−2936 TCGA−AB−2949 TCGA−AB−2923 TCGA−AB−2901 TCGA−AB−2865 TCGA−AB−2934 TCGA−AB−2871 TCGA−AB−2983 TCGA−AB−2955 TCGA−AB−2861 TCGA−AB−2970 TCGA−AB−2919 TCGA−AB−2821 TCGA−AB−3000 TCGA−AB−2969 TCGA−AB−2884 TCGA−AB−2831 TCGA−AB−2966 TCGA−AB−2904 TCGA−AB−2937 TCGA−AB−2855 TCGA−AB−2849 TCGA−AB−2895 TCGA−AB−2842 TCGA−AB−2971 TCGA−AB−2882 TCGA−AB−2830 TCGA−AB−2981 TCGA−AB−2931 TCGA−AB−2859 TCGA−AB−2833 TCGA−AB−2896 TCGA−AB−2864 TCGA−AB−2992 TCGA−AB−2851 TCGA−AB−2832 TCGA−AB−2927 TCGA−AB−2932 TCGA−AB−3002 TCGA−AB−2853 TCGA−AB−2899 TCGA−AB−2813 TCGA−AB−2988 TCGA−AB−2974 TCGA−AB−2945 TCGA−AB−2909 TCGA−AB−2942 TCGA−AB−3006 TCGA−AB−3009 TCGA−AB−2972 TCGA−AB−2843 TCGA−AB−2868 TCGA−AB−3005 TCGA−AB−2866 TCGA−AB−2965 TCGA−AB−2847 TCGA−AB−2924 TCGA−AB−2860 TCGA−AB−2925 TCGA−AB−2947 TCGA−AB−2873 TCGA−AB−2978 TCGA−AB−2879 TCGA−AB−2913 TCGA−AB−3012 TCGA−AB−2991 TCGA−AB−2950 TCGA−AB−2806 TCGA−AB−2823 TCGA−AB−2905 TCGA−AB−2812 TCGA−AB−2834 TCGA−AB−2994 TCGA−AB−2982 TCGA−AB−2920 TCGA−AB−2826 TCGA−AB−2872 TCGA−AB−2804 TCGA−AB−2836 TCGA−AB−2819 TCGA−AB−2963 TCGA−AB−3001 TCGA−AB−2911 TCGA−AB−2827 TCGA−AB−2877 TCGA−AB−2815 TCGA−AB−2803 TCGA−AB−2802 TCGA−AB−2915 TCGA−AB−2964 TCGA−AB−2957 TCGA−AB−2973 TCGA−AB−2846 TCGA−AB−2876 TCGA−AB−2898 TCGA−AB−2939 TCGA−AB−2848 TCGA−AB−2844 TCGA−AB−2878 TCGA−AB−2990 TCGA−AB−2998 TCGA−AB−2999 TCGA−AB−3007 TCGA−AB−2976 TCGA−AB−2845 TCGA−AB−2809 TCGA−AB−2967 TCGA−AB−2828 TCGA−AB−2893 TCGA−AB−2897 TCGA−AB−2952 TCGA−AB−2888 TCGA−AB−2881 TCGA−AB−2894 TCGA−AB−2916 TCGA−AB−2933 TCGA−AB−2908 TCGA−AB−2954 TCGA−AB−2840 TCGA−AB−2985 TCGA−AB−2862 TCGA−AB−2856 TCGA−AB−2874 TCGA−AB−2892 TCGA−AB−2810 TCGA−AB−2808 TCGA−AB−2838 TCGA−AB−2841 TCGA−AB−2993 TCGA−AB−2886 TCGA−AB−2857 TCGA−AB−2883 TCGA−AB−2917 TCGA−AB−2980 TCGA−AB−2910 TCGA−AB−2912 TCGA−AB−2956 TCGA−AB−2869 TCGA−AB−2839 TCGA−AB−2948 TCGA−AB−2935 TCGA−AB−2989 TCGA−AB−2914 TCGA−AB−2986 TCGA−AB−2875 TCGA−AB−2867 TCGA−AB−2858 TCGA−AB−2814 TCGA−AB−2870 TCGA−AB−2900 TCGA−AB−2889 TCGA−AB−2818

Poor Expression Intermediate Good 6810 12 14 Fig. 1 CEBPE expression in AML patients with diferent prognosis. a CEBPE expression of AML patients with good and poor outcomes in three independent datasets TCGA, GSE14468 and GSE1159. b CEBPE expression and prognosis classifcation based on cytogenetic factors in TCGA database Li et al. J Transl Med (2019) 17:188 Page 4 of 11

a TCGA GSE1159 GSE10358 .0 .0 .0 low,n= 138 low,n= 195 low,n= 68 high,n= 46 high,n= 65 high,n= 23 l l l 0. 81 0. 81 0. 81 .6 .6 .6 0. 40 0. 40 0. 40 Overall Surviva Overall Surviva Overall Surviva .2 .2 .2

log−rank P = 0.0064 0. 00 log−rank P = 0.0347

log−rank P = 0.0254 0. 00 0. 00 020406080 050100 150200 01020304050 Months Months Months

b TCGA GSE1159 GSE10358 .0 .0 .0 low,n= 138 low,n= 195 low,n= 68 high,n= 23 high,n= 46 l high,n= 65 0. 81 0. 81 0. 81 .6 .6 .6 0. 40 0. 40 0. 40 .2 .2 .2 Event−free Surviva Event−free Survival Event−free Survival

log−rank P = 0.044

log−rank P = 0.0275 log−rank P = 0.0484 0. 00 0. 00 0. 00 020406080100 050100 150200 01020304050 Months Months Months Fig. 2 Survival analyses of AML patients with diferential expression of CEBPE. a Overall survival (OS) analyses of three independent datasets TCGA, GSE1159, GSE10358. b Event-free survival (EFS) analyses of three independent datasets TCGA, GSE1159, GSE10358

showed lower risk. Te following variables were evalu- Table 1 Univariable analyses of overall survival (OS) ated in univariable Cox regression models for outcome: of AML patients from TCGA database CEBPE expression, age, sex, white blood cell (WBC), Variable HR 95% CI P-value peripheral blood (PB) or bone marrow (BM) blasts, the presence or absence of various chromosomal transloca- CEBPE, high vs. low 0.5 0.3–0.7 0.00029 Age, 60 vs. < 60 3.2 2.2–4.6 9.40E 10 tions [i.e., inv(16), t(8;21), t(15;17), t(9;11), t(11q23) and ≥ − log2 (WBC), each 2-unit increase 1.2 1.0–1.4 0.040 t(9;22)] and other abnormalities [+8, −3/inv(3)/t(3;3), FLT3, FLT3-ITD vs. others 1.4 0.9–2.1 0.094 −7/del(7q), −5/del(5)], and the presence or absence of gene mutations (FLT3-ITD or FLT3-TKD, DNMT3A, NPM1, mutation vs. wild-type 1.4 1.0–2.1 0.071 IDH1, IDH2, RUNX1, TET2, TP53, NRAS, CEBPA, DNMT3A, mutation vs. wild-type 1.8 1.2–2.6 0.0049 KRAS, NPM1, KIT, PHF6 and ASXL1). Variables for RUNX1, mutation vs. wild-type 1.7 1.0–2.9 0.063 TP53, mutation vs. wild-type 3.4 1.9–5.9 1.98E 05 which P < 0.1 in univariable analysis were shown in the − Table 1 (OS) and Table 2 (EFS). Hazard ratios (HR) > 1 inv(16) vs. others 0.3 0.1–0.9 0.032 or < 1 indicate, respectively, a higher or lower risk of an t(15;17) vs. others 0.3 0.1–0.7 0.0075 event for higher values of continuous variables or for t(9;11) vs. others 4.1 1.0–16.9 0.052 the frst category listed for categorical variables in OS or t(9;22) vs. others 3.4 0.8–14.1 0.086 EFS models. Accordingly, we found that age, TP53 muta- del (3) vs. others 2.0 0.8–4.1 0.097 tion, DNMT3A mutation, WBC, t(9;11), RUNX1 muta- HR hazard ratio, 95% CI 95% confdence interval, WBC white blood cell, ITD tion were risk factors, while CEBPE expression, t(15;17) internal tandem duplication and inv(16) were protective factors for AML OS and Variables for which P < 0.1 in univariable models were shown Li et al. J Transl Med (2019) 17:188 Page 5 of 11

Table 2 Univariable analyses of event-free survival (EFS) Low‑expression of CEBPE predicts high relapse rate of AML patients from TCGA database We evaluated the association between CEBPE expression Variable HR 95% CI P-value and relapse rates after complete remission using datasets of TCGA and GSE1159, which contained the infor- CEBPE, high vs. low 0.5 0.4–0.8 0.00098 mation of relapse. All of the samples were classifed into Age, 60 vs. < 60 2.8 2.0–4.1 2.86E 08 ≥ − CEBPE high-expressed and low-expressed groups based log2(WBC), each 2-unit increase 1.2 1.0–1.3 0.074 on k-Nearest Neighbor (KNN) approach. Te results DNMT3A, mutation vs. wild-type 1.5 1.0–2.2 0.039 showed that CEBPE expression had signifcant predic- RUNX1, mutation vs. wild-type 1.6 1.0–2.7 0.093 tive power for AML relapse (P < 0.05). Low expression TP53, mutation vs. wild-type 3.2 1.9–5.6 2.76E 05 − of CEBPE resulted in an increased incidence of relapse inv(16) vs. others 0.3 0.1–0.9 0.039 (Fig. 3). t(15;17) vs. others 0.3 0.1–0.8 0.015 t(9;11) vs. others 3.9 0.9–16.0 0.061 CEBPE expression has prognostic signifcance for wild‑type t(11q23) vs. others 2.5 1.1–5.7 0.033 AML patients of multiple genes HR hazard ratio, 95% CI 95% confdence interval, WBC white blood cell Some gene mutations were reported to be associated Variables for which P < 0.1 in univariable models were shown with poor outcome of AML, such as mutations of TP53 [24], FLT3 [25], DNMT3A [26], RUNX1 [27]. However, Table 3 Multivariable analyses of OS of AML patients the frequency of patients with these mutations was rela- from TCGA database tively low. Novel prognostic factors were required to predict the outcome of wild-type patients. We evaluated the Variable HR 95% CI P-value prognostic power of CEBPE expression for AML wild- CEBPE, high vs. low 0.6 0.4–0.9 0.034 type patients in TCGA datasets. For each gene mutation, Age, 60 vs. < 60 2.9 1.9–4.3 3.14E 07 samples were divided into four classes, namely mutated/ ≥ − log2(WBC), each 2-unit increase 1.4 1.1–1.7 0.0025 CEBPE high, mutated/CEBPE low, wild-type/CEBPE TP53, mutation vs. wild-type 4.5 2.2–9.4 6.22E 05 high, wild-type/CEBPE low. Te results showed that − t(9;11) vs. others 8.4 1.9–36.9 0.0051 CEBPE expression diferences in wild-type patients of TP53, FLT3, DNMT3A, KRAS, RUNX1 and NRAS were HR hazard ratio, 95% CI 95% confdence interval; WBC white blood cell Variables for which P < 0.05 in multivariable models were shown strongly associated with survival time (Fig. 4). Wild-type patients with high-expression of CEBPE showed longer survival than low-expressed wild-type patients. Tus, Table 4 Multivariable analyses of EFS of AML patients CEBPE expression could provide useful prognosis infor- from TCGA database mation by subdividing the wild-type patients. Variable HR 95% CI P-value CEBPE expression was a potential prognostic factor CEBPE, high vs. low 0.6 0.4–0.9 0.042 for allogeneic transplantation Age, 60 vs. < 60 2.6 1.7–3.8 2.42E 06 ≥ − We analyzed the association between CEBPE expression log2(WBC), each 2-unit increase 1.4 1.2–1.7 0.00056 and allogeneic transplantation to explore whether CEBPE TP53, mutation vs. wild-type 3.5 1.8–6.9 0.00025 expression could provide useful information for direct- HR hazard ratio, 95% CI 95% confdence interval, WBC white blood cell ing allogeneic transplantation. All samples were classifed Variables for which P < 0.05 in multivariable models were shown into CEBPE high-expressed and low-expressed groups based on KNN approach. Ten, in each group, Kaplan– Meier survival analyses were applied to compare the EFS. Trough multivariable testing, we showed that the survival diference between individuals received and not CEBPE low-expression remained signifcantly associated received transplants. Te results showed that no obvious with worse OS and EFS in TCGA datasets, after adjusting improvement was achieved by allogeneic transplanta- for all other variables that had P < 0.1 in univariable anal- tion in CEBPE high-expressed group, while the survival yses. Variables for which P < 0.05 in multivariable models rate (both OS and EFS) was signifcantly increased in were also shown in the Table 3 (OS) and Table 4 (EFS). transplanted patients that with low expression of CEBPE It turned out that age, TP53 mutation, WBC and CEBPE (Fig. 5). Tese results suggested that CEBPE expression expression were independent predictors for AML OS and would be a potential predictor for outcome of allogeneic EFS. transplantation in AML patients. Li et al. J Transl Med (2019) 17:188 Page 6 of 11

TCGA GSE1159 0. 8 .0 low,n= 95 low,n= 105 high,n= 86 high,n= 155 .6 0. 81 e e .6 0. 40 0. 40 Relapse Rat Relapse Rat .2 .2

log−rank P = 0.0117 log−rank P = 0.0481 0. 00 0. 00

020406080 050100 150200 Months Months Fig. 3 Kaplan-Meier analyses of AML relapse rates after complete remission in TCGA and GSE1159 datasets

CEBPE regulates known predictors of AML regulating myeloid diferentiation [29, 30], we hypoth- According to the above results, we showed that CEBPE esized that CEBPE might regulate the expression of these expression was an independent prognostic factor for known prognostic factors by localizing on their promot- AML survival, relapse and allogeneic transplantation. ers, and verifed using ChIP-qPCR assay in NB4 and Ten, we attempted to explain the molecular mecha- Kasumi-1 cells. Te results showed that CEBPE actually nism of favorable outcome induced by increase of CEBPE occupied on the promoters of known predictors, sug- expression. An international collaborative study reported gesting the regulation role of CEBPE on genes associated by Li et al. [28] identifed a 24-gene prognostic signature with AML prognosis. based on the data analyses of 1324 AML patients, and improved the established risk classifcation of AML prog- Discussion nosis. Te identifed 24 genes were ALS2CR8, ANGEL1, In the clinical setting, it is important to identify prognos- ARL6IP5, BSPRY, BTBD3, C1RL, CPT1A, DAPK1, ETFB, tic factors to direct the appropriate treatments and predict FGFR1, HEATR6, LAPTM4B, MAP7, NDFIP1, PBX3, outcomes. Patients with a molecular profle that is associ- PLA2G4A, PLOD3, PTP4A3, SLC25A12, SLC2A5, ated with a favorable risk have relatively good outcomes TMEM159, TRIM44, TRPS1, and VAV3, the increased with chemotherapy, whereas patients with an unfavorable- expression levels of which were signifcantly associated risk profle require allogeneic transplantation during the with worse (22 genes) or favorable (two genes: FGFR1 frst remission to improve their prognosis [5, 31]. Based on and PLOD3) OS of AML. We found that CEBPE expres- the analyses of several independent datasets, we identifed sion was signifcantly correlated with these known pre- CEBPE expression as an independent prognostic factors dictors of AML. As many as 13 genes were co-expressed for AML patients. Low-expression of CEBPE was found with CEBPE in TCGA dataset (P-value < 0.05, Fig. 6a left to be associated with shorter OS, EFS and higher relapse panel), and 15 genes were co-expressed with CEBPE in rate, indicating adverse outcome of AML. Importantly, GSE1159 dataset (P-value < 0.05, Fig. 6a right panel). both RNA-Seq and microarray data supported this results, Interestingly, CEBPE expression level was positively cor- suggesting that CEBPE expression was a reliable prognos- related with FGFR1 and PLOD3 in both datasets, which tic factor. In addition, CEBPE expression was proved to were reported as favorable factors, while negatively cor- have prognostic signifcance for wild type patients of vari- related with other genes which reported as predictors ous genes, providing useful information for prognosis of for poor outcome. Tis observation was consistent with patients without molecular alterations. Moreover, CEBPE our results that high expression of CEBPE predicted expression was also a potential prognostic factor for allo- longer survival and lower relapse rate. Given the fact geneic transplantation. Tis observation could be eas- that CEBPE was an important transcription factor in ily used in routine clinical practice, as CEBPE expression Li et al. J Transl Med (2019) 17:188 Page 7 of 11

TP53 FLT3 .0 .0 TP53−/CEBPE Low,n= 83 FLT3−/CEBPE Low,n= 68 TP53+/CEBPE Low,n= 9 FLT3+/CEBPE Low,n= 24 TP53−/CEBPE High,n= 86 FLT3−/CEBPE High,n= 66 0. 81 0. 81 TP53+/CEBPE High,n= 6 FLT3+/CEBPE High,n= 26 l l .6 .6 0. 40 0. 40 Overall Surviva Overall Surviva .2 .2 0. 00 0. 00 0 20 40 60 80 0 20 40 60 80 Months Months

DNMT3A KRAS .0 .0 DNMT3A−/CEBPE Low,n= 57 KRAS−/CEBPE Low,n= 87 DNMT3A+/CEBPE Low,n= 35 KRAS+/CEBPE Low,n= 5 KRAS−/CEBPE High,n= 90 DNMT3A−/CEBPE High,n= 80 KRAS+/CEBPE High,n= 2 0. 81 0. 81 DNMT3A+/CEBPE High,n= 12 l l .6 .6 0. 40 0. 40 Overall Surviva Overall Surviva .2 .2 0. 00 0. 00 0 20 40 60 80 0 20 40 60 80 Months Months

RUNX1 NRAS .0 .0 RUNX1−/CEBPE Low,n= 77 NRAS−/CEBPE Low,n= 87 RUNX1+/CEBPE Low,n= 15 NRAS+/CEBPE Low,n= 5 RUNX1−/CEBPE High,n= 89 NRAS−/CEBPE High,n= 83 0. 81 0. 81 RUNX1+/CEBPE High,n= 3 NRAS+/CEBPE High,n= 9 l l .6 .6 0. 40 0. 40 Overall Surviva Overall Surviva .2 .2 0. 00 0. 00 020406080 0 20 40 60 80 Months Months Fig. 4 CEBPE expression has prognostic signifcance for wild-type patients of multiple genes. “ ” indicates mutation and “ ” indicates wild-type. + − Diferential expression of CEBPE stratifed the wild-type patients into good and poor outcomes Li et al. J Transl Med (2019) 17:188 Page 8 of 11

ab .0 .0 no_transplant/CEBPE low,n= 55 no_transplant/CEBPE high,n= 52 transplant/CEBPE low,n= 37 transplant/CEBPE high,n= 40 0. 81 0. 81 log−rank P = 2e−16 log−rank P = 0.2178 al iv .6 al .6 iv rv rv 0. 40 0. 40 erall Su erall Su Ov Ov .2 .2 0. 00 0. 00

010203040506070 020406080 Months Months cd .0 .0 no_transplant/CEBPE low,n= 55 no_transplant/CEBPE high,n= 52 transplant/CEBPE low,n= 37 transplant/CEBPE high,n= 40 0. 81 0. 81 log−rank P = 0.5125 log−rank P = 2e−04 al al iv iv .6 .6 0. 40 0. 40 ent−free Su rv ent−free Su rv Ev .2 Ev .2 0. 00 0. 00

020406080 100 020406080 100 Months Months Fig. 5 CEBPE expression was a potential prognostic factor for allogeneic transplantation. a Overall survival analyses for CEBPE low-expressed patients received or not received allogeneic transplantation. b Overall survival analyses for CEBPE high-expressed patients received or not received allogeneic transplantation. c Event-free survival analyses for CEBPE low-expressed patients received or not received allogeneic transplantation. d Event-free survival analyses for CEBPE high-expressed patients received or not received allogeneic transplantation

could be tested before deciding if allogeneic transplanta- of birth due to the loss of mature neutrophils or eosino- tion should be implemented, and allogeneic transplan- phils [35]. Similarly, patients with a frame-shift muta- tation surgery would be recommended only for CEBPE tion in CEBPE are sufered from specifc granulocyte low-expressed patients, which will provide accurate infor- defciency disease [36]. Tese observations imply that mation for therapeutic decisions. CEBPE may play a pivotal role in granulocytic matura- Te generation and development of AML are associ- tion and exert an important function in myeloid diferen- ated with the disregulation of various transcription fac- tiation. Our observations suggested that CEBPE localized tors (TFs) [32]. Especially, the abnormal expression of on the promoters of a series of known predictors of AML TFs which are important in hematopoietic or myeloid survival, and had positive or negative co-expression rela- diferentiations would lead to the accumulation of mye- tionship with these target genes. Tis result highlighted loblasts in the bone marrow and peripheral blood [33]. the reasons of why CEBPE expression showed signifcant Previous studies suggested that CEBPE was indispensable prognostic power. Importantly, it is much more practical for myeloid normal diferentiation progress [30, 34]. For and economical to test the expression of one driver gene example, CEBPE knockout mice die within a few months (CEBPE) than to test several passenger genes. Terefore, Li et al. J Transl Med (2019) 17:188 Page 9 of 11

Fig. 6 CEBPE regulates known predictors of AML. a Heatmaps for gene expression of CEBPE and known prognostic factors of AML in TCGA and GSE1159 datasets. The Pearson correlation coefcient between expression values of CEBPE and each known predictor was listed in the brackets. b ChIP-qPCR was performed using the anti-CEBPE in Kasumi-1 and NB4 cell lines. Data are shown as fold enrichment of ChIPed DNA vs. input DNA. Error bars represent SD of triplicate measurements. NC: negative control

CEBPE expression holds great potential for clinical appli- Additional fle cation in risk stratifcation and outcome prediction in AML. Additional fle 1: Table S1. qPCR primer sequences.

Conclusion Abbreviations Our fndings indicated that CEBPE expression was AML: acute myeloid leukemia; OS: overall survival; EFS: event-free survival; ChIP: chromatin immunoprecipitation; WBC: white blood cell; PB: peripheral an independent prognostic factor for AML survival, blood; BM: bone marrow; HR: hazard ratios; KNN: k-Nearest Neighbor; TFs: relapse and allogeneic transplantation, which will pro- transcription factors; ELN: European LeukemiaNet; ITD: internal tandem dupli‑ vide useful information for outcome prediction and cation; CI: confdence interval. therapeutic decisions. Li et al. J Transl Med (2019) 17:188 Page 10 of 11

Acknowledgements Marra MA, Mayo M, Moore RA, Mungall K, Parker J, Pleasance E, Plettner Not applicable. P, Schein J, Stoll D, Swanson L, Tam A, Thiessen N, Varhol R, Wye N, Zhao Y, Gabriel S, Getz G, Sougnez C, Zou L, Leiserson MD, Vandin F, Wu HT, Authors’ contributions Applebaum F, Baylin SB, Akbani R, Broom BM, Chen K, Motter TC, Nguyen KNL analyzed data and drafted the article. FZ and YXD performed experi‑ K, Weinstein JN, Zhang N, Ferguson ML, Adams C, Black A, Bowen J, ments. FZ and DQW designed the project, analyzed data, revised the drafted Gastier-Foster J, Grossman T, Lichtenberg T, Wise L, Davidsen T, Demchok article. All authors read and approved the fnal manuscript. JA, Shaw KR, Sheth M, Sofa HJ, Yang L, Downing JR, et al. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl Funding J Med. 2013;368:2059–74. This work was supported by Grants from the National Natural Science 6. Mrozek K, Marcucci G, Nicolet D, Maharry KS, Becker H, Whitman SP, Foundation Grants of China (31801147), the Shanghai Sailing Program Metzeler KH, Schwind S, Wu YZ, Kohlschmidt J, Pettenati MJ, Heerema (18YF1410800), the startup fund for young teachers of Shanghai Jiao Tong NA, Block AW, Patil SR, Baer MR, Kolitz JE, Moore JO, Carroll AJ, Stone University (17X100040083), the grants from the Key Research Area Grant RM, Larson RA, Bloomfeld CD. Prognostic signifcance of the European 2016YFA0501703 of the Ministry of Science and Technology of China, the LeukemiaNet standardized system for reporting cytogenetic and National Natural Science Foundation of China (Contract no. 61832019, molecular alterations in adults with acute myeloid leukemia. J Clin Oncol. 61503244), the State Key Lab of Microbial Metabolism and Joint Research 2012;30:4515–23. Funds for Medical and Engineering and Scientifc Research at Shanghai Jiao 7. Komanduri KV, Levine RL. Diagnosis and therapy of acute myeloid Tong University (YG2017ZD14). The computations were partially performed at leukemia in the era of molecular risk stratifcation. Annu Rev Med. the Center for High Performance Computing, Shanghai Jiao Tong University. 2016;67:59–72. 8. Metzeler KH, Herold T, Rothenberg-Thurley M, Amler S, Sauerland MC, Availability of data and materials Gorlich D, Schneider S, Konstandin NP, Dufour A, Braundl K, Ksienzyk The datasets analyzed in the current study are available in The Cancer B, Zellmeier E, Hartmann L, Greif PA, Fiegl M, Subklewe M, Bohlander Genome Atlas (TCGA) (http://cancergenome.nih.gov/) and Gene Expression SK, Krug U, Faldum A, Berdel WE, Wormann B, Buchner T, Hiddemann Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/). W, Braess J, Spiekermann K, Group AS. Spectrum and prognostic relevance of driver gene mutations in acute myeloid leukemia. Blood. Ethics approval and consent to participate 2016;128:686–98. Not applicable. 9. Theis F, Corbacioglu A, Gaidzik VI, Paschka P, Weber D, Bullinger L, Heuser M, Ganser A, Thol F, Schlegelberger B, Gohring G, Kohne CH, Germing U, Consent for publication Brossart P, Horst HA, Haase D, Gotze K, Ringhofer M, Fiedler W, Nachbaur Not applicable. D, Kindler T, Held G, Lubbert M, Wattad M, Salih HR, Krauter J, Dohner H, Schlenk RF, Dohner K. Clinical impact of GATA2 mutations in acute Competing interests myeloid leukemia patients harboring CEBPA mutations: a study of the The authors declare that they have no competing interests. AML study group. Leukemia. 2016;30:2248–50. 10. Daver N, Schlenk RF, Russell NH, Levis MJ. Targeting FLT3 muta‑ Author details tions in AML: review of current knowledge and evidence. Leukemia. 1 State Key Laboratory of Microbial Metabolism and School of Life Sciences 2019;33:299–312. and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China. 11. Dohner H, Dolnik A, Tang L, Seymour JF, Minden MD, Stone RM, Del 2 State Key Laboratory of Medical Genomics, Shanghai Institute of Hematol‑ Castillo TB, Al-Ali HK, Santini V, Vyas P, Beach CL, MacBeth KJ, Skikne BS, ogy, Rui‑Jin Hospital, Shanghai JiaoTong University, Shanghai 200025, China. Songer S, Tu N, Bullinger L, Dombret H. Cytogenetics and gene mutations 3 Key Laboratory of Systems Biomedicine, Shanghai Center for Systems infuence survival in older patients with acute myeloid leukemia treated Biomedicine, Shanghai Jiao Tong University, Shanghai 200240, China. with azacitidine or conventional care. Leukemia. 2018;32:2546–57. 12. Jawhar M, Dohner K, Kreil S, Schwaab J, Shoumariyeh K, Meggendorfer Received: 22 March 2019 Accepted: 30 May 2019 M, Span LLF, Fuhrmann S, Naumann N, Horny HP, Sotlar K, Kubuschok B, von Bubnof N, Spiekermann K, Heuser M, Metzgeroth G, Fabarius A, Klein S, Hofmann WK, Kluin-Nelemans HC, Haferlach T, Dohner H, Cross NCP, Sperr WR, Valent P, Reiter A. KIT D816 mutated/CBF-negative acute myeloid leukemia: a poor-risk subtype associated with systemic mastocy‑ References tosis. Leukemia. 2019;33:1124–34. 1. Bullinger L, Dohner K, Dohner H. Genomics of acute myeloid leukemia 13. Metzeler KH, Becker H, Maharry K, Radmacher MD, Kohlschmidt J, Mrozek diagnosis and pathways. J Clin Oncol. 2017;35:934–46. K, Nicolet D, Whitman SP, Wu YZ, Schwind S, Powell BL, Carter TH, Wetzler 2. Szer J. The prevalent predicament of relapsed acute myeloid leukemia. M, Moore JO, Kolitz JE, Baer MR, Carroll AJ, Larson RA, Caligiuri MA, Mar‑ Hematol Am Soc Hematol Educ Progr. 2012;2012:43–8. cucci G, Bloomfeld CD. ASXL1 mutations identify a high-risk subgroup of 3. Burnett AK, Hills RK, Milligan DW, Goldstone AH, Prentice AG, McMullin older patients with primary cytogenetically normal AML within the ELN MF, Duncombe A, Gibson B, Wheatley K. Attempts to optimize induction favorable genetic category. Blood. 2011;118:6920–9. and consolidation treatment in acute myeloid leukemia: results of the 14. Morita K, Kantarjian HM, Wang F, Yan Y, Bueso-Ramos C, Sasaki K, Issa MRC AML12 trial. J Clin Oncol. 2010;28:586–95. GC, Wang S, Jorgensen J, Song X, Zhang J, Tippen S, Thornton R, Coyle 4. O’Donnell MR, Tallman MS, Abboud CN, Altman JK, Appelbaum FR, Arber M, Little L, Gumbs C, Pemmaraju N, Daver N, DiNardo CD, Konopleva DA, Bhatt V, Bixby D, Blum W, Coutre SE, De Lima M, Fathi AT, Fiorella M, M, Andreef M, Ravandi F, Cortes JE, Kadia T, Jabbour E, Garcia-Manero Foran JM, Gore SD, Hall AC, Kropf P, Lancet J, Maness LJ, Marcucci G, Mar‑ G, Patel KP, Futreal PA, Takahashi K. Clearance of somatic mutations at tin MG, Moore JO, Olin R, Peker D, Pollyea DA, Pratz K, Ravandi F, Shami PJ, remission and the risk of relapse in acute myeloid leukemia. J Clin Oncol. Stone RM, Strickland SA, Wang ES, Wieduwilt M, Gregory K, Ogba N. Acute 2018;36:1788–97. myeloid leukemia, Version 3.2017, NCCN Clinical Practice Guidelines in 15. Prochazka KT, Pregartner G, Rucker FG, Heitzer E, Pabst G, Wolfer A, Zebi‑ Oncology. J Natl Compr Cancer Netw. 2017;15:926–57. sch A, Berghold A, Dohner K, Sill H. Clinical implications of subclonal TP53 5. Cancer Genome Atlas Research, N, Ley TJ, Miller C, Ding L, Raphael BJ, mutations in acute myeloid leukemia. Haematologica. 2019;104:516–23. Mungall AJ, Robertson A, Hoadley K, Triche TJ, Jr. Laird PW, Baty JD, Fulton 16. Dohner H, Estey EH, Amadori S, Appelbaum FR, Buchner T, Burnett AK, LL, Fulton R, Heath SE, Kalicki-Veizer J, Kandoth C, Klco JM, Koboldt DC, Dombret H, Fenaux P, Grimwade D, Larson RA, Lo-Coco F, Naoe T, Nieder‑ Kanchi KL, Kulkarni S, Lamprecht TL, Larson DE, Lin L, Lu C, McLellan MD, wieser D, Ossenkoppele GJ, Sanz MA, Sierra J, Tallman MS, Lowenberg B, McMichael JF, Payton J, Schmidt H, Spencer DH, Tomasson MH, Wallis JW, Bloomfeld CD, European L. Diagnosis and management of acute myeloid Wartman LD, Watson MA, Welch J, Wendl MC, Ally A, Balasundaram M, leukemia in adults: recommendations from an international expert panel, Birol I, Butterfeld Y, Chiu R, Chu A, Chuah E, Chun HJ, Corbett R, Dhalla on behalf of the European LeukemiaNet. Blood. 2010;115:453–74. N, Guin R, He A, Hirst C, Hirst M, Holt RA, Jones S, Karsan A, Lee D, Li HI, Li et al. J Transl Med (2019) 17:188 Page 11 of 11

17. Mrozek K, Marcucci G, Paschka P, Whitman SP, Bloomfeld CD. Clinical 26. Saygin C, Hirsch C, Przychodzen B, Sekeres MA, Hamilton BK, Kalaycio relevance of mutations and gene-expression changes in adult acute M, Carraway HE, Gerds AT, Mukherjee S, Nazha A, Sobecks R, Goebel C, myeloid leukemia with normal cytogenetics: are we ready for a prognos‑ Abounader D, Maciejewski JP, Advani AS. Mutations in DNMT3A, U2AF1, tically prioritized molecular classifcation? Blood. 2007;109:431–48. and EZH2 identify intermediate-risk acute myeloid leukemia patients 18. Kihara R, Nagata Y, Kiyoi H, Kato T, Yamamoto E, Suzuki K, Chen F, Asou N, with poor outcome after CR26. Blood Cancer J. 2018;8:4. Ohtake S, Miyawaki S, Miyazaki Y, Sakura T, Ozawa Y, Usui N, Kanamori H, 27. Gaidzik VI, Teleanu V, Papaemmanuil E, Weber D, Paschka P, Hahn J, Wallra‑ Kiguchi T, Imai K, Uike N, Kimura F, Kitamura K, Nakaseko C, Onizuka M, benstein T, Kolbinger B, Kohne CH, Horst HA, Brossart P, Held G, Kundgen Takeshita A, Ishida F, Suzushima H, Kato Y, Miwa H, Shiraishi Y, Chiba K, A, Ringhofer M, Gotze K, Rummel M, Gerstung M, Campbell P, Kraus Tanaka H, Miyano S, Ogawa S, Naoe T. Comprehensive analysis of genetic JM, Kestler HA, Thol F, Heuser M, Schlegelberger B, Ganser A, Bullinger alterations and their prognostic impacts in adult acute myeloid leukemia L, Schlenk RF, Dohner K, Dohner H. RUNX1 mutations in acute myeloid patients. Leukemia. 2014;28:1586–95. leukemia are associated with distinct clinico-pathologic and genetic 19. Valk PJ, Verhaak RG, Beijen MA, Erpelinck CA, van Waalwijk Barjesteh, van features. Leukemia. 2016;30:2160–8. Doorn-Khosrovani S, Boer JM, Beverloo HB, Moorhouse MJ, van der Spek 28. Li Z, Herold T, He C, Valk PJ, Chen P, Jurinovic V, Mansmann U, Radmacher PJ, Lowenberg B, Delwel R. Prognostically useful gene-expression profles MD, Maharry KS, Sun M, Yang X, Huang H, Jiang X, Sauerland MC, Buchner in acute myeloid leukemia. N Engl J Med. 2004;350:1617–28. T, Hiddemann W, Elkahloun A, Neilly MB, Zhang Y, Larson RA, Le Beau MM, 20. Stirewalt DL, Meshinchi S, Kopecky KJ, Fan W, Pogosova-Agadjanyan Caligiuri MA, Dohner K, Bullinger L, Liu PP, Delwel R, Marcucci G, Lowen‑ EL, Engel JH, Cronk MR, Dorcy KS, McQuary AR, Hockenbery D, Wood B, berg B, Bloomfeld CD, Rowley JD, Bohlander SK, Chen J. Identifcation of Heimfeld S, Radich JP. Identifcation of genes with abnormal expres‑ a 24-gene prognostic signature that improves the European Leuke‑ sion changes in acute myeloid leukemia. Genes Chromosomes Cancer. miaNet risk classifcation of acute myeloid leukemia: an international 2008;47:8–20. collaborative study. J Clin Oncol. 2013;31:1172–81. 21. Tomasson MH, Xiang Z, Walgren R, Zhao Y, Kasai Y, Miner T, Ries RE, 29. Negrotto S, Ng KP, Jankowska AM, Bodo J, Gopalan B, Guinta K, Mulloy Lubman O, Fremont DH, McLellan MD, Payton JE, Westervelt P, DiPersio JC, Hsi E, Maciejewski J, Saunthararajah Y. CpG methylation patterns JF, Link DC, Walter MJ, Graubert TA, Watson M, Baty J, Heath S, Shannon and decitabine treatment response in acute myeloid leukemia cells and WD, Nagarajan R, Bloomfeld CD, Mardis ER, Wilson RK, Ley TJ. Somatic normal hematopoietic precursors. Leukemia. 2012;26:244–54. mutations and germline sequence variants in the expressed tyrosine 30. Marchwicka A, Marcinkowska E. Regulation of expression of CEBP genes kinase genes of patients with de novo acute myeloid leukemia. Blood. by variably expressed vitamin D receptor and retinoic acid receptor alpha 2008;111:4797–808. in human acute myeloid leukemia cell lines. Int J Mol Sci. 2018;19:E1918. 22. Wouters BJ, Lowenberg B, Erpelinck-Verschueren CA, van Putten WL, Valk 31. Maruf M, Sposto R, Oberley MJ, Kysh L, Orgel E. Therapy for children and PJ, Delwel R. Double CEBPA mutations, but not single CEBPA mutations, adults with mixed phenotype acute leukemia: a systematic review and defne a subgroup of acute myeloid leukemia with a distinctive gene meta-analysis. Leukemia. 2018;32:1515–28. expression profle that is uniquely associated with a favorable outcome. 32. Churpek JE, Bresnick EH. Transcription factor mutations as a cause of Blood. 2009;113:3088–91. familial myeloid neoplasms. J Clin Invest. 2019;129:476–88. 23. Taskesen E, Bullinger L, Corbacioglu A, Sanders MA, Erpelinck CA, Wouters 33. Takei H, Kobayashi SS. Targeting transcription factors in acute myeloid BJ, van der Poel-van de Luytgaarde SC, Damm F, Krauter J, Ganser A, leukemia. Int J Hematol. 2019;109:28–34. Schlenk RF, Lowenberg B, Delwel R, Dohner H, Valk PJ, Dohner K. Prognos‑ 34. Wada T, Akagi T. Role of the Leucine Zipper Domain of CCAAT/enhancer tic impact, concurrent genetic mutations, and gene expression features binding protein-Epsilon (C/EBPepsilon) in neutrophil-specifc granule of AML with CEBPA mutations in a cohort of 1182 cytogenetically normal defciency. Crit Rev Immunol. 2016;36:349–58. AML patients: further evidence for CEBPA double mutant AML as a 35. Priam P, Krasteva V, Rousseau P, D’Angelo G, Gaboury L, Sauvageau G, Les‑ distinctive disease entity. Blood. 2011;117:2469–75. sard JA. SMARCD2 subunit of SWI/SNF chromatin-remodeling complexes 24. Rucker FG, Schlenk RF, Bullinger L, Kayser S, Teleanu V, Kett H, Habdank M, mediates granulopoiesis through a CEBPvarepsilon dependent mecha‑ Kugler CM, Holzmann K, Gaidzik VI, Paschka P, Held G, von Lilienfeld-Toal nism. Nat Genet. 2017;49:753–64. M, Lubbert M, Frohling S, Zenz T, Krauter J, Schlegelberger B, Ganser A, 36. Gombart AF, Shiohara M, Kwok SH, Agematsu K, Komiyama A, Koefer HP. Lichter P, Dohner K, Dohner H. TP53 alterations in acute myeloid leukemia Neutrophil-specifc granule defciency: homozygous recessive inherit‑ with complex karyotype correlate with specifc copy number alterations, ance of a frameshift mutation in the gene encoding transcription factor monosomal karyotype, and dismal outcome. Blood. 2012;119:2114–21. CCAAT/enhancer binding protein–epsilon. Blood. 2001;97:2561–7. 25. Cucchi DGJ, Denys B, Kaspers GJL, Janssen J, Ossenkoppele GJ, de Haas V, Zwaan CM, van den Heuvel-Eibrink MM, Philippe J, Csikos T, Kwidama Z, de Moerloose B, de Bont E, Lissenberg-Witte BI, Zweegman S, Verwer F, Publisher’s Note Vandepoele K, Schuurhuis GJ, Sonneveld E, Cloos J. RNA-based FLT3-ITD Springer Nature remains neutral with regard to jurisdictional claims in pub‑ allelic ratio is associated with outcome and ex vivo response to FLT3 lished maps and institutional afliations. inhibitors in pediatric AML. Blood. 2018;131:2485–9.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions