Reduced expression of lncRNA DLEU7-AS1 is a novel favorable prognostic factor in acute myeloid leukemia

As an important group of common hematopoietic and highly heterogeneous malignancies, acute myeloid leukemia (AML) is characterized by distinct chromosomal and genetic abnormalities, which ultimately leads to the accumulation of anomalous myeloblasts. Although the prognostic importance of its high level of molecular heterogeneity has been well established over the past two decades, translation of this novel information into therapeutic strategy improvements and optimization remains at a primitive stage. The prognosis for the majority of AML subtypes remains dismal due to a lack of effective diagnostic methods and optimal long-term therapeutic regimens [1–3]. It has been frequently reported that the typical AML phenotype results from complex epigenetic/genetic aberrations, which affects cell proliferation, differentiation and apoptosis [4,5]. In particular, it is now becoming clear that cytogenetically normal AML (CN-AML) is highly heterogeneous on molecular levels according to the 2017 World Health Organization (WHO) classification of AML [6]. Cytogenetic alterations, combined with mutations in nucleophosmin 1 (NPM1), CCAAT/enhancer-binding protein α (CEBPA), fms-like tyrosine kinase 3- internal tandem duplication (FLT3-ITD) and TP53, have been incorporated into the current prognostic grading system for patients with AML [7]. In addition, the overexpression of brain and acute leukemia cytoplasmic protein (BAALC), Echovirus 11 (EVI1) and E-26 transformation-specific-related gene (ERG) has also been reported to be beneficial for outcome prediction with high accuracy [7]. Nevertheless, it remains of key importance to identify a novel set of integrated AML-related biomarkers that can have profound impacts on diagnosis, prognosis and individual treatment.

Long non-coding RNAs (lncRNAs) are non-coding RNAs that are >200 nucleotides in length and belong to an emerging category of RNAs that can serve key roles in various types of cancer [8]. Due to their tumor-specific expression patterns, lncRNAs may either serve as tumor suppressors or oncogenes, rendering them important components in the field of cancer research [9]. Previous studies have found that aberrant expression or mutation of lncRNAs can not only increase susceptibility to cancer development but also promote tumorigenesis and metastasis by regulating the expression of related protein-coding genes on different transcriptional levels [10]. As a result of next-generation sequencing technologies, large quantities of novel, distinct lncRNAs have been discovered to be abnormally expressed or mutated in a variety of cancers [11,12]. However, the majority of these lncRNAs have yet to be annotated [11,12]. Characterizing these lncRNAs in addition to understanding the mechanism underlying their function may assist in improving diagnosis and highlight novel therapeutic targets.

The lncRNA deleted in lymphocytic leukemia 7-antisense RNA 1 (DLEU1), originating from 13q.14.3 region, has been previously identified as an up-regulated oncogenic lncRNA in colorectal cancer (CRC). Following the knockdown of DLEU7-AS1 in CRC cell lines HT-29 and HCT-116, the expression levels of key proteins in the Wnt/β-catenin pathway β-catenin, c-Myc and cyclin D1 were all found to be markedly reduced [13]. This suggests that DLEU7-AS1 may promote the occurrence and development of CRC by modulating the Wnt/β-catenin signaling pathway [13]. However, the DLEU7-AS1 expression profile in hematopoietic malignancies remains poorly understood. Therefore, in the present study, we aimed to measure DLEU7-AS1 expression in patients with AML and analyze its relevant clinical significance. Using bioinformatics analysis, we also analyzed the correlation between the expression levels of DLEU7-AS1 with immunocyte infiltration and immune checkpoint genes in AML, in addition to the potential biological functions of DLEU7-AS1.

Bone marrow (BM) specimens were collected from a total of 30 healthy donors and 110 patients with AML. BM mononuclear cells (BMMNCs) were extracted by gradient centrifugation and lymphocyte separation medium (TBD Sciences, Tianjin, People’s Republic of China). According to the 2016 WHO criteria and French-American-British (FAB) classification system [14,15], all patients with de novo AML enrolled had a detailed diagnosis. Karyotypes were analyzed at first diagnosis using the conventional R-banding method whereas karyotype risk was classified into three groups (favorable, intermediate or poor) according to the 2017 European leukemiaNET (ELN) risk stratification by genetics [3]. In addition to cytogenetics, ELN 2017 assesses the risk of patients by taking into consideration the effects of mutations in certain genes, such as NPM1, CEBPA, FLT3-ITD, runt-related transcription factor 1, additional sex-combs like 1 and TP53 [3]. The detailed treatment information and inclusion criteria were kept consistent with our previously published reports [16,17]. Briefly, all patients with AML included in this study received chemotherapy, including induction therapy and subsequent consolidation treatment. Induction therapy for non-M3 AML consists of two courses of daunorubicin combined with cytarabine. Subsequent consolidation treatment included high-dose cytarabine, mitoxantrone with cytarabine and homoharringtonine combined with cytarabine. By contrast, patients with M3 were treated with all-trans retinoic acid (ATRA) together with daunorubicin in combination with cytarabine, followed by oral mercaptopurine, oral methotrexate and oral ATRA for >2 years as maintenance therapy. Our study was authorized by the Ethical Committee and Institutional Review Board of the Affiliated People’s Hospital of Jiangsu University (Zhenjiang, China). Patients and families were required to sign the informed consent permitting that their BM samples would be utilized for scientific research on the principle of voluntary participation.

In total, three publicly available AML datasets from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) were used to validate our results in this study. Of these data sets, two consisted of expression data for large primary AML samples (GSE12417 and the TCGA dataset) and one dataset contained both AML and healthy BM samples sorted by fluorescence-activated cell sorting (GSE63270). GSE42519, which includes gene expression data of normal hematopoietic cells, was used to explore the expression pattern of DLEU7-AS1 at various stages of normal hematopoiesis. The Genotype-Tissue Expression (GTEx; https://www.GTExportal.org/home/) and the Cancer Cell Line Encyclopedia (CCLE; https://www.broadinstitute.org/ccle) databases were utilized to assess the expression of DLEU7-AS1 in different human tissues and cancer cell lines.

The total RNA was separated from pre-extracted BMMNCs using TRIzol® reagent (Invitrogen, Carlsbad, CA, U.S.A.) and then reverse transcribed (MBIFermentas, Hanover, MD, U.S.A.) into complementary DNA (cDNA) in an iCycler Thermal Cycler (Eppendorf, Hamburg, Germany) as described in our previous article [18]. Subsequently, a 7500 Thermo cycler (Applied Biosystems, CA, U.S.A.) was utilized to perform qPCR, with each reaction system containing 20 μl, consisting of 0.4 μM ROX Reference Dye II (Takara Bio, Inc., Tokyo, Japan), 0.8 μM primers, 10 μM SYBR Premix TB Green and 20 ng cDNA. The thermocycling conditions for the qPCR reaction were 95°C for 5 min, followed by 40 cycles at 95°C for 10 s, 64°C for 30 s, 72°C for 30 s and 80°C to collect fluorescence for 30 s; the final step was 95°C for 15 s, 60°C for 60 s, 95°C for 15 s and 60°C for 15 s. Serial dilution was performed and calibration curves were generated to test for PCR efficiency. Positive controls and negative controls were used in each trial to exclude false negatives and false positives. The abundance of DLEU7-AS1 was calculated by normalizing to the housekeeping gene ABL1 [19]. The following primers were used for qPCR: DLEU7-AS1 reverse, 5′-AGTTCTCCCTTGCTGCACTC-3′ and forward, 5′-TATGGCCGAGCAGACATTGG-3′; ABL1 reverse, 5′-TCCAACGAGCGGCTTCAC-3′ and forward, 5′-TCCTCCAGCTGTTATCTGGAAGA-3′. The results were analyzed using the 7500 Thermo cycler software and the relative expression levels of DLEU7-AS1 were calculated using the 2^−ΔΔCq method [19].

Isocitrate dehydrogenase 1/2, C-KIT, NPM1, DNA methyltransferase 3A, U2 small nuclear RNA auxiliary factor and N/K-RAS mutations were detected using high-resolution melting analysis [20–24], as this method was demonstrated by our group to have high sensitivity and specificity for assessing these gene mutations. FLT3-ITD and CEBPA mutations were detected by RT-qPCR [25,26]. Finally, direct DNA sequencing was used to verify all the positive samples (BGI Tech Solutions Co, Shanghai, China).

Data analyses were performed using the IBM SPSS Statistics 25.0 software (IBM Corporation, Armonk, NY, U.S.A.). The expression values were first log₂ transformed before the differences in expression were calculated using the Wilcoxon rank-sum test. A receiver operating characteristic curve (ROC) was utilized to evaluate if DLEU7-AS1 expression could be used to distinguish patients with AML from healthy individuals. A Fisher’s exact test or Pearson's χ² analysis was applied to evaluate the association between each of the classification variables and patients in the DLEU7-AS1^high expression group or those in the DLEU7-AS1^low group. The effect of DLEU7-AS1 expression on the survival time of patients with AML was analyzed using the Kaplan–Meier method and Cox regression analysis, with the optimal cut-off point determined using X-tile analysis. The correlation between DLEU7-AS1 and DLEU7 expression was analyzed through a Pearson’s correlation test. Correlation between the expression of DLEU7-AS1 with immunocyte infiltration and immune checkpoint genes in the TCGA cohort were assessed using the Sangerbox software 3.0 (https://sangerbox.com/). Differential gene expression analysis was performed using the limma R package. Differentially expressed genes (DEGs) between patients with high and low DLEU7-AS1 expression were illustrated using a volcano plot using the “EnhancedVolcano” R package. Genes with an adjusted P-value <0.05 and |log₂ fold change| ≥0.5 were considered to be DEGs. STRING database version 11.5 (https://www.string-db.org/) was applied to analyze the protein-protein interactions (PPI) and Gene Ontology (GO) enrichment of DLEU7-AS1 co-expressed genes. For PPI analysis, a confidence score >0.9 was considered to be the judgment criterion. Cytoscape version 3.8.2 was used to present the network. All graphs were produced using GraphPad Prism® 7.0 (GraphPad Software, San Diego, California) and R version 4.0.3 (https://www.r-project.org/). P<0.05 was applied as the threshold of statistical significance in all tests.

Analysis of transcriptional data from GTEx (http://www.GTExportal.org/home/) revealed that DLEU7-AS1 expression was the highest in the bone marrow (Figure 1A). Furthermore, CCLE database analysis identified relatively high levels of DLEU7-AS1 mRNA expression in the hematopoietic cell lines (Figure 1B). It should be noted that DLEU7-AS1 expression across all tissues was fairly even and the high expression observed in hematopoietic tissue is probably a reflection of the larger expression variation across this tissue dataset.

We then integrated normal tissue transcription data from GTEx and tumor-adjacent normal tissue data from TCGA to assess if DLEU7-AS1 expression was dysregulated in different types of cancer. Although DLEU7-AS1 expression was found in AML among all malignancies analyzed, its expression was found to be markedly decreased in AML compared with that in healthy controls (Figure 2A). In addition, decreased DLEU7-AS1 expression was also observed in AML samples in an independent GEO dataset (GSE63270; P<0.001) (Figure 2B). To verify these findings, we detected DLEU7-AS1 expression in 110 patients with newly diagnosed AML and 30 normal controls using RT-qPCR as previously described [19]. The expression level of DLEU7-AS1 in the normal control group was between 0.002 and 38.349 (median, 2.953). However, the level of DLEU7-AS1 transcripts in patients with AML was between 0.015 and 143.529 (median, 0.068), which was found to be significantly lower compared with that in the normal control group according to a non-parametric test (P<0.001) (Figure 2C). DLEU7-AS1 expression was also found to be significantly lower in both non-M3 and CN-AML subgroups as compared with normal controls (Figure 2C). To validate these findings and explore the expression pattern of DLEU7-AS1 during the different stages of normal hematopoiesis, the level of DLEU7-AS1 transcripts was analyzed using another published GEO dataset (GSE42519). We found relatively low DLEU7-AS1 expression during the early stages of hematopoietic differentiation, but the quantity of this transcript was increased as the cell lineages matured, with the highest expression level observed in metamyelocytes (Figure 3).

ROC curve analysis revealed that differential DLEU7-AS1 expression levels could be used to distinguish whole cohorts of patients with AML from the normal control group (Figure 4A). The area under the curve (AUC) was 0.820 (95% CI: 0.737–0.903; P<0.001). Using the cut-off value of <0.089, the specificity and sensitivity for discriminating AML were calculated to be 90 and 55%, respectively. In addition, significant differences were also observed in non-M3 AML subgroup of patients (AUC = 0.821; 95% CI: 0.734–0.907; P<0.001) (Figure 4B) and the CN-AML subgroup of patients (AUC = 0.839; 95% CI: 0.748–0.930; P<0.001) (Figure 4C).

We divided the entire group of patients with AML from our cohort (n=110) into the DLEU7-AS1^low and DLEU7-AS1^high groups according to the cut-off value of 0.089 based on the ROC curve. A comparison of the clinical parameters of the two groups in the respective available data are shown in Table 1. We observed no significant difference in age, white blood cell (WBC) count, sex distribution, BM blasts ratio and hemoglobin levels between the two groups of patients (P>0.05). There were also no statistically significant differences in the distribution of WHO classifications or in the FAB subtypes between these two groups (P>0.05). However, the DLEU7-AS1^low group had significantly lower baseline platelet counts compared with those in the DLEU7-AS1^high group (P=0.029). In addition, a significant difference was found between the DLEU7-AS1^low and DLEU7-AS1^high groups in the distribution of karyotype classification (P=0.008). The frequency of favorable karyotype (65%, 15/23) and intermediate karyotype (58%, 44/76) was significantly higher compared with that of the poor karyotype (10%, 1/10) in the DLEU7-AS1^low group (P=0.005). The expression levels of DLEU7-AS1 in patients with favorable (median, 0.053) and intermediate karyotypes (median, 0.053) were significantly decreased compared with those in the normal group (P<0.0001). By contrast, those with poor karyotypes (median, 0.915) had a trend toward lower DLEU7-AS1 expression compared with that in the normal control group (P=0.067) (Figure 5). However, there was no significant correlation between the frequencies of common gene mutations (such as CEBPA, NPM1, and FLT3-ITD) and DLEU7-AS1 expression (P>0.05) (Table 1).

Next, the relationship between DLEU7-AS1 expression and outcome was analyzed in a total of 99 patients (our cohort) who had available follow-up data. Patients in the DLEU7-AS1^low group had significantly higher complete response (CR) rates compared with those in the DLEU7-AS1^high group (57 vs. 26%; P=0.002) (Table 1). In addition, the DLEU7-AS1^low group also had significantly higher CR rates compared with those in the DLEU7-AS1^high group in the non-M3 subgroup [46%, (19/41) vs. 20% (8/40); P=0.012]. The difference in the CR rate between patients in the DLEU7-AS1^low and DLEU7-AS1^high groups was not significant in the CN-AML subgroup [44% (14/32) vs. 22% (5/23); P=0.09]. In addition, survival analysis revealed that patients in the DLEU7-AS1^low group exhibited superior overall survival (OS) rates (P=0.004) (Figure 6A) compared with those in the DLEU7-AS1^high group in the whole AML cohort. However, the impact of DLEU7-AS1 expression on leukemia-free survival was not statistically significant (data not shown). We also tested the prognostic value of DLEU7-AS1 expression in the non-M3 (P=0.013) (Figure 6B) and CN-AML subgroups (P=0.027) (Figure 6C), where it was found to be significantly associated with longer OS. Overall, these results suggest that DLEU7-AS1 expression can be concluded to be a viable prognostic predictor in both the whole cohort in addition to the non-M3 or CN-AML subgroups. Subsequently, the prognostic value of DLEU7-AS1 expression for OS was also validated in one CN-AML cohort from the GSE12417 dataset (n=79; P=0.004) and the dataset from two large primary AML cohorts (GSE37642, n=136, P=0.005; TCGA microarray, n=183, P=0.040) (Figure 6D), where low DLEU7-AS1 expression remained an indicator of favorable OS.

Finally, common clinical parameters, including the cytogenetic risk group (favorable vs. intermediate vs. poor), WBC counting (≥30 × 10⁹/L vs. <30 × 10⁹/L), age (≤60 vs. >60 years), DLEU7-AS1 expression status and molecular mutations in our cohort were analyzed using univariate Cox models. Variables with P<0.2 in univariate analysis were further included in a multivariate model. However, multivariate analysis failed to reveal the independent prognostic significance of DLEU7-AS1 expression in patients with CN-AML (data not shown), whole-cohort AML (data not shown) or non-M3 (Table 2).

We deemed that there may be other stronger prognostic indicators. As shown in Table 2, we observed that age was an important independent risk factor in the non-M3 subgroup. Subsequently, patients who were aged ≤60 years old were separated for further study. A In this subgroup analysis, significant difference was found in the OS of patients in the DLEU7-AS1^low and DLEU7-AS1^high groups patients according to univariate analysis (P=0.047) (Table 3). Multivariate Cox analysis indicated that DLEU7-AS1 expression was also an important prognostic indicator [HR = 2.146; 95% CI (1.023–4.501); P=0.043] (Table 3) in this subset.

Since DLEU7-AS1 is the non-coding strand of the DLEU7 gene, we hypothesized that their expression might be correlated. To explore the relationship between the expression of DLEU7-AS1 and DLEU7, an independent cohort containing the expression information of 183 patients with de novo AML derived from TCGA was included in this study. Pearson’s correlation analysis revealed that DLEU7-AS1 and DLEU7 expression were positively correlated (R = 0.93; P<0.0001) (Figure 7A). Furthermore, the quantity of DLEU7 transcripts was significantly decreased in AML samples compared with that in normal controls (GSE63270; P=0.0001) (Figure 7B).

Next, we investigated the possible correlation between DLEU7-AS1 expression and immunocyte infiltration levels using CIBERSORT in the TCGA AML cohort. We found DLEU7-AS1 expression to be negatively correlated with tumor-suppressive components, such as activated CD8 T cells and natural killer cells (Figure 8A). The most significant negative association was observed for type 17 T helper cells (P=0.0016), whilst monocytes (P=0.079) and type 2 T helper cells (P=0.063) also showed a trend toward significant association with DLEU7-AS1 expression (Figure 8A).

Subsequently, we analyzed the association between DLEU7-AS1 expression and checkpoint gene expression. We found a positive association between DLEU7-AS1 expression with V-Set domain-containing T Cell activation inhibitor 1, adenosine A2A receptor, CD160 and tumor necrosis factor (TNF)-related immune genes (TNFRSF8, TNFRSF14, TNFSF14 and TNFSF15), whilst the expression of DLEU7-AS1 was negatively correlated with transmembrane and immunoglobulin domain-containing 1 (TMIGD1), HERV-H LTR-associating 2 (HHLA2; a receptor of the B7 family member). TMIGD1 and HHLA2 were previously reported to be potential therapeutic targets in cancers [27] (Figure 8B).

To further assess the role of DLEU7-AS1 in AML, we derived DLEU7-AS1-associated gene expression profiles using the TCGA microarray data (n=183). We identified 14 up-regulated and 373 down-regulated genes that were found to be significantly associated with DLEU7-AS1 expression (adjusted P<0.05; |log₂ fold change| >0.5) (Figure 9A,B). Subsequently, we constructed a PPI network using the DEGs by STRING to explore potential interactions among them. In this network, we noticed that the expression of one potential therapeutic target, EGFR, was up-regulated in the DLEU7-AS1^high group, where it directly interacted with WASP-like actin nucleation promoting factor, formin binding protein 1, LGNAI1, leucine rich repeat kinase 1, FYN, plasminogen activator, urokinase, YES1, epidermal growth factor receptor pathway substrate 8 and disabled homolog 2 (Figure 9C). In addition, GO analysis was performed to analyze the potential role of DLEU7-AS1 and the co-expressed genes. We found that several metabolism-related biological processes, such as regulation of primary metabolic process, regulation of macromolecule metabolic process, regulation of metabolic process and regulation of cellular metabolic process were enriched (Figure 9D). Moreover, cellular components, including intracellular, intracellular organelle, organelle, membrane-bounded organelle, cytoplasm, intracellular membrane-bounded organelle, cell, nuclear lumen, cytosol and intracellular organelle lumen, were significantly associated with the DLEU7-AS1 alteration (Figure 9D). However, no significant enrichment in molecular functions or pathways were found for these DEGs.

DLEU7-AS1 is an antisense lncRNA that was previously found to be uniquely expressed in specific cancer types or differentiated tissues [28]. Antisense lncRNAs are a specific type of noncoding RNA that can regulate gene expression in the nucleus and cytoplasm through RNA–DNA interactions, transcription interference or RNA–RNA interactions [29]. A number of studies have previously indicated that antisense lncRNAs are also involved in genomic imprinting [30], X-chromosome inactivation [31,32], progression and development of certain diseases [33]. However, the DLEU7-AS1 expression pattern and its downstream role in hematological malignancies remain unknown. Since negative DLEU7 expression and hypermethylation of its promoter were previously found in patients with B-cell chronic lymphocytic leukemia (B-CLL), DLEU7 was previously proposed to be a candidate cancer suppressor gene in CLL located in 13q14 [34]. The most common aberration is 13q deletion, which results in the reduced expression of microRNAs, such as miR-15a and miR16-1 [34,35]. They were also considered to be important in the pathogenesis of CLL and have an influence on TP53 expression and activity [34,35]. These results suggested that DLEU7-AS1 and DLEU7 can serve an important role in AML.

In the present study, we first investigated DLEU7-AS1 expression levels in patients initially diagnosed with AML and then analyzed its clinical impact. The results revealed that DLEU7-AS1 expression was dramatically down-regulated in patients with AML compared with that in normal controls. Indeed, the BMMNCs from AML and normal samples contain hematopoietic cells of nonmyeloid origin so that the results can be taken in context. It is recommended that positive sorting or lymphocyte depletion was performed prior to sample processing in future works assessing DLEU7-AS1 expression. Since the treatment outcome of M3 and non-M3 patients are very different, patients with M3 subtype are thought as confounding factors. While the CN-AML are indeed a heterogeneous group of patients with normal karyotypes. Therefore, in most analyses, we have included the whole cohort, the non-M3 and the CN-AML subsets. ROC curve analysis also revealed that differential DLEU7-AS1 expression could be used to distinguish patients with whole AML, CN-AML and non-M3 from healthy controls.

During analysis of the expression pattern of DLEU7-AS1 in the different karyotypic classifications, we revealed that DLEU7-AS1 expression was markedly reduced in patients with AML with favorable/intermediate karyotypes, but not in patients with poor karyotypes. This seemingly contradictory result may be caused by the relatively small size of our cohort, especially patients with poor karyotypes (n=10). In addition, since all patients in this study were selected retrospectively, a potential bias relating to the unbalanced clinical features with treatment heterogeneity cannot be ruled out. Finally, since patients with low DLEU7-AS1 expression generally had favorable prognosis, this may explain the decreased DLEU7-AS1 expression in patients with favorable karyotypes. Further prospective studies with larger sample sizes are required to validate this result.

Combining our investigation using Kaplan–Meier analysis together with multivariate analysis, we concluded that the patient’s age and WBC count may have had a stronger influence on the prognosis than DLEU7-AS1 expression. DLEU7-AS1 repression during leukemogenesis may depend on the context of favorable and intermediate karyotypes, which may also act as a potential prognostic factor for longer OS in AML. This can add an additional prognostic parameter for stratifying the patients. Additionally, deletion of 13q and the expression of miR-15a and miR16-1 may also be used to explain the differences in OS.

Although increased DLEU7-AS1 expression has only been reported in CRC, which correlated positively with the Wnt/β-catenin pathway [13]. This suggests that the higher the DLEU7-AS1 expression level, the poorer the outcome [13]. Our observation in AML, which is consistent with this published data with CRC, suggests that the DLEU7-AS1 expression profile and its role in tumorigenesis and development may be tissue-specific. In fact, dysregulation of Wnt/β-catenin signaling was previously found to be necessary for leukemia stem cell self-renewal and survival [36] and associated with AML initiation and progression [37]. We speculate that reduced DLEU7-AS1 expression may inhibit Wnt/β-catenin signaling through conserved binding sites in the favorable/intermediate karyotypes of patients with AML. To support this hypothesis, we performed correlation analyses between DLEU7-AS1 expression and 117 Wnt pathway genes retrieved from GSEA (https://www.gsea-msigdb.org/gsea/msigdb/cards/KEGG_WNT_SIGNALING_PATHWAY) using the TCGA dataset. The results favorably agreed with the reported role of DLEU7-AS1 in Wnt pathway. We found that DLEU7-AS1 was significantly associated with Wnt target genes such as WNT1, SOX17 and SFRP4 (P<0.05, |r| > 0.3; Supplementary Data S1). However, the positive association between the expression of DLEU7-AS1 with DLEU7 and Wnt target gens warrants further experimental verification, in addition to the regulatory relationship between DLEU7-AS1 and DLEU7 expression in AML. Functional follow ups and prospective validations will be necessary to further interpret the mechanisms of altered DLEU7-AS1 expression in AML. It is a promising observation that conserved binding sites for a number of transcription factors can be found in the DLEU7 promoter, several of which have been reported to serve significant roles in the tissue-specific expression of genes upstream of cell differentiation and proliferation [34]. However, to the best of our knowledge this has not been investigated in other cancer types. Therefore, the underlying pathological roles of DLEU7-AS1 in different malignancies remain largely unknown and require further study.

In conclusion, our study found that decreased DLEU7-AS1 expression was associated with favorable/intermediate karyotypes. In addition, in patients with de novo AML, downregulation of DLEU7-AS1 expression may serve as a potentially viable prognostic indicator.

The datasets analyzed in this study are available in the following open access repositories: GTEx, www.gtexportal.org/; CCLE, https://www.broadinstitute.org/ccle; TCGA, https://portal.gdc.cancer.gov/, http://www.cbioportal.org; GEO, https://www.ncbi.nlm.nih.gov/geo/ (GEO accession numbers: GSE63270, GSE42519, GSE37642, and GSE12417); CIBERSORT, https://cibersort.stanford.edu/. Other inhouse data used and/or analyzed during the current study are available from the corresponding author on reasonable request.

The authors declare that there are no competing interests associated with the manuscript.

This study was supported by the National Natural Science Foundation of China [grant number 81270630]; 333 Project of Jiangsu Province [grant number BRA2016131]; Medical Innovation Team of Jiangsu Province [grant number CXTDB2017002]; Zhenjiang Clinical Research Center of Hematology [grant number SS2018009]; Six talent peaks project in Jiangsu Province [grant number 2015-WSN-115]; Social Development Foundation of Zhenjiang [grant numbers SH2018044 and SH2019065]; Key Medical Talent Program of Zhenjiang City, “Postgraduate Research & Practice Innovation Program of Jiangsu Province” [grant number KYCX17_1821]. None of the sponsors had any involvement in the experimental design, specimen collection or data analysis. Additionally, none of the sponsors served a role in drafting the manuscript.

Cui-zhu Wang: Conceptualization, Resources, Data curation, Writing—original draft. Bei-bei Ma: Data curation, Software, Formal analysis, Project administration. Zi-jun Xu: Software, Formal analysis, Validation, Investigation. Jing-dong Zhou: Conceptualization, Resources, Data curation, Methodology, Writing—original draft. Ting-juan Zhang: Data curation, Visualization, Methodology, Writing—original draft. Qin Chen: Data curation, Formal analysis, Funding acquisition, Validation, Investigation. Dong-ming Yao: Funding acquisition, Validation, Investigation, Visualization, Methodology. Jiang Lin: Conceptualization, Funding acquisition, Investigation. Jun Qian: Methodology, Conceptualization, Investigation. Suo Sha: Formal analysis, Visualization, Methodology.

The present study was approved by the Clinical Research Ethics Committee of the Affiliated People’s Hospital of Jiangsu University (Zhenjiang, China).

Written informed consent was obtained from all enrolled individuals voluntarily before their participation.

The authors would like to thank all the patients and their families for participating in this project.

AML

acute myeloid leukemia

ATRA

all-trans retinoic acid

BAALC

brain and acute leukemia cytoplasmic protein

BM

bone marrow

BMMNC

BM mononuclear cell

CCLE

Cancer Cell Line Encyclopedia

cDNA

complementary DNA

CEBPA

CCAAT/enhancer-binding protein α

CN-AML

cytogenetically normal AML

CR

complete remission

CRC

colorectal cancer

DEG

differentially expressed gene

DLEU7-AS1

deleted in lymphocytic leukemia 7-antisense 1

DLEU7-AS1^high

high DLEU7-AS1 expression

DLEU7-AS1^low

low DLEU7-AS1 expression

ELN

European leukemiaNET

ERG

E-26 transformation-specific-related gene

EVI1

Echovirus 11

FAB

French-American-British

FLT3-ITD

fms-like tyrosine kinase 3- internal tandem duplication

GEO

Gene Expression Omnibus

GO

Gene Ontology

GTEx

Genotype-Tissue Expression project

LncRNA

long non-coding RNA

non-M3

non-acute promyelocytic leukemia

NPM1

nucleophosmin 1

OS

overall survival

PPI

protein–protein interactions

ROC

receiver operating characteristic curve

TCGA

The Cancer Genome Atlas

WBC

white blood cell

WHO

World Health Organization