The role of E2A in ATPR‐induced cell differentiation and cycle arrest in acute myeloid leukaemia cells

Abstract Acute myeloid leukaemia (AML) is a biologically heterogeneous disease with an overall poor prognosis; thus, novel therapeutic approaches are needed. Our previous studies showed that 4‐amino‐2‐trifluoromethyl‐phenyl retinate (ATPR), a new derivative of all‐trans retinoic acid (ATRA), could induce AML cell differentiation and cycle arrest. The current study aimed to determine the potential pharmacological mechanisms of ATPR therapies against AML. Our findings showed that E2A was overexpressed in AML specimens and cell lines, and mediate AML development by inactivating the P53 pathway. The findings indicated that E2A expression and activity decreased with ATPR treatment. Furthermore, we determined that E2A inhibition could enhance the effect of ATPR‐induced AML cell differentiation and cycle arrest, whereas E2A overexpression could reverse this effect, suggesting that the E2A gene plays a crucial role in AML. We identified P53 and c‐Myc were downstream pathways and targets for silencing E2A cells using RNA sequencing, which are involved in the progression of AML. Taken together, these results confirmed that ATPR inhibited the expression of E2A/c‐Myc, which led to the activation of the P53 pathway, and induced cell differentiation and cycle arrest in AML.

syndrome and drug resistance during the treatment process limit its application. 11 Therefore, AML therapy warrants the development of other highly effective and safe drugs. Our group via structural modification synthesized a new derivative 4-amino-2-trifluorome thyl-phenyl retinate (ATPR) of ATRA ( Figure 1A). Previous studies have confirmed enhanced therapeutic effects on human gastric cancer, hepatocellular carcinoma, gastric carcinoma, breast cancer and leukaemia from ATPR compared with ATRA. 12 Our group studies demonstrated its anti-cancer mechanism through the effect on enzyme activity, protein interaction, non-coding RNA downstream regulation, etc. [11][12][13] However, many other mechanisms need to be further explored.
As a member of the basic helix-loop-helix transcription factors (BHLH), E2A is located in the cytoplasm and regulates genes related to cell proliferation and differentiation. 14 Its BHLH domain mediates the dimer and binds to the E-box within DNA, and the Ad1/ Ad2 domains recruit p300 and GCN5 to enhance the activation of target genes. 15 E2A has been reported to be highly expressed in a variety of tumour cells, such as cervical carcinoma, 16 nasopharyngeal carcinoma 17 and is associated with poor prognosis. The E2A gene is a common target of chromosomal translocation in leukaemia and can easily produce fusion proteins, such as E2A-PBX1 18 and E2A-HLF, 19 but the mechanism of its regulation in leukaemia has not been clarified. Taken together, these results suggest that alteration of the E2A gene may play a key role in the pathogenesis of leukaemia.
To further explore the role of E2A in AML, we performed transcriptome sequencing, and the results showed that silencing E2A could specifically downregulate c-Myc. c-Myc regulates cell growth, differentiation and apoptosis, and it is also present in the nucleus of cells and highly expressed in a variety of human tumours. 20 Therefore, we hypothesized that E2A promotes cell proliferation by upregulating the expression of the c-Myc gene in AML cells. Studies have suggested that c-Myc overexpression could reverse the activation of the p53 pathway caused by silencing the SNRPG gene in glioma cells, and promoted the development of glioma. 21 In a variety of tumours, the p53 pathway participates in the regulation of biological processes, including cell cycle arrest and apoptosis, associated with cancer such as leukaemia and liver cancer. 22,23 In conclusion, we proposed a hypothesis that ATPR inhibits E2A expression in AML cells by interacting with retinoic acid receptor alpha (RARα), thereby reducing the expression of the downstream target gene c-Myc, activating the P53 pathway and inducing cell differentiation and cycle arrest.

| Materials
The design and synthesis of ATPR were performed as shown in

| Cell culture
We purchased human leukaemia cell lines THP-1 and NB4 from the in an atmosphere of 5% CO 2 .
Twenty microliters of E2A knockdown or overexpression lentiviruses was used to transfect NB4 and THP-1 cells at a density of 1 × 10 5 cells/well in 24-well plates. Three days after lentivirus infection, stable cloned cells were obtained by screening with puromycin (final concentration: 5 μg/mL Meilunbio, Dalian, China) for 2 consecutive days. 24 The shRNA target sequence for E2A was CCGGCCCGGATC AC TCA AGCA ATA AC TCGAGT TAT TGC T TGAGTGATCCGGGT   TTTTG, and the negative control (NC) sequence was AATTGAA   A A A AT TC TCCGA ACGTGTC ACGTA ATC TC T TGA AT TACGTG ACACGTTCGGAGAACG.

| Transcriptome sequencing
Total RNA was extracted from NB4 cells transfected with sh-NC, and sh-E2A lentiviral transfection was performed using TRIzol reagent (Invitrogen Corp). The RNA samples were stored at −80°C until analysis. RNA sequencing was performed by the Wuhan Bioacme Biological Technology Co., Ltd.

| Western blot analysis
The cells were separated into groups following the designated treatments, washed twice with prechilled phosphate-buffered saline (PBS) and lysed in RIPA buffer containing PMSF for 30

| Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was isolated from cultured NB4 and THP-1 cells that were treated with TRIzol reagent (Invitrogen Corp), and reverse transcribed into cDNA using the First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). The relative gene mRNA expression levels in different samples were evaluated via qRT-PCR using the SYBR-Green PCR kit (Takara). β-actin mRNA was used as an internal control for normalization. The sequences of primers used were as follows: β-actin, 5-CGCCGCCAGCTCACCATG-3′ (forward) and 5-CACGATGGAGGGGAAGACGG-3′ (reverse); c-Myc,
All primers were synthesized and purchased from Sangon Biotech Co., Ltd., and the results were analysed using the 2-ΔΔCT method.

| Differentiation marker analysis
The level of maturity of the cells was evaluated via the expression of the cell surface differentiation-related antigens CD11b and CD14.
The treated cells were incubated with 2 µL (CD11b-PE/CY5; CD14-FITC) antibody for 30 min in the dark, using a CytoFLEX (Becton Dickinson, USA), and the data were analysed using CytExpert software.

| Cell cycle analysis
The treated cells were harvested, washed with prechilled PBS, fixed in 75% ethanol at −20°C overnight and the cell cycle analysis

| Wright-Giemsa staining
The designated cells were separated into several groups, treated with ATPR (1 × 10 −6 M) for 3 days and washed twice with prechilled PBS.
The cells were fixed on glass slides, dried at room temperature and stained with Wright-Giemsa, and morphological differentiation was observed by fluorescence inversion microscope system (OLYMPUS).

| Nitroblue tetrazolium (NBT) reduction assay
For the NBT reduction assay, cells were treated as designated and washed with PBS. A 10-μL aliquot of the mixed solution, containing 10 mg/mL NBT (Sigma-Aldrich) and 2 μg/mL PMA (Sigma-Aldrich), was added to each tube and incubated for 30 min at 37°C.

| Double immunofluorescent staining
The cells were seeded in a six-well plate, washed twice with PBS after harvest and fixed with a fixing solution for 1 h at 4°C. After

| Tumour xenograft
We purchased 6-to 8-week-old female/male NSG mice from Nanjing Model Animal Research Institute, and randomly divided them into two groups. A tumour xenograft model was produced, and 5 × 10 6 NB4 cells transfected with sh-NC and sh-E2A were subcutaneously injected into the right flank of NSG mice to investigate the effect of E2A on AML in vivo. 25 The tumour volumes of NSG mice were measured every day, and they were killed 2 weeks later and the tumours were removed for further evaluation.

| Histopathology
To analyse the expression of related proteins in tumour tissues from NSG mice, immunohistochemistry (IHC) staining was performed following standard procedures. Human monoclonal E2A, c-Myc, cyclin A2, CDK4 and CD11b antibodies (Bioss, China) were used for IHC, diluted 1:300 and then detected and photographed using a fluorescence inversion microscope system (OLYMPUS).

| Statistical analysis
All statistical analyses were conducted using SPSS software V. 21.0, and the data are shown as the mean ± standard deviation. Student's t-test was used to compare two groups, whereas one-way analysis of variance was used for comparing multiple groups. Statistical significance was set at p < 0.05.

| High E2A expression in AML cell lines and patient samples
To determine the expression of the E2A gene in AML, PMBCs of all participants were collected, and CD34 + cells were purified from cord blood and identified using CytoFLEX ( Figure 1B). We then monitored changes in the levels of E2A protein in AML. Western blotting results showed that the level of E2A protein was significantly higher in AML specimens than in healthy controls ( Figure 1C; S1A). Furthermore, we analysed a number of AML cell lines (NB4, MOLM-13 and THP-1), and the results showed that E2A protein levels were higher than those of normal human CD34 + cells ( Figure 1D). These results suggest that high expression of E2A may be associated with the progression of AML.

| E2A expression in AML cells was inhibited by ATPR treatment
To further analyse the effect of the E2A gene on ATPR treatment of AML, we first used ATPR at different concentrations ( Together, these results indicate that ATPR induces the degradation of E2A in AML cells.

| ATPR-induced AML cell differentiation and cycle arrest was enhanced in the absence of E2A
To evaluate the role of E2A in ATPR-induced cell differentiation, we used primary AML cells were harvested directly from patients with AML, and analysed by flow cytometry after ATPR treatment, and showed that increased the expression of differentiation markers CD11b/CD14, indicating that ATPR-induced primary AML cell differentiation ( Figure S2A). Next, we transfected sh-E2A into AML cells to establish AML/sh-E2A stable clone cells and used sh-NC transfected into the cells as the control ( Figure 3A). Western blotting and qRT-PCR results showed that, compared with the control group, sh-E2A cells significantly reduced the levels of E2A protein and mRNA ( Figure 3B, C). Western blotting was used to assess the expression levels of cell surface differentiation antigens (CD11b and CD14), and the results revealed that the absence of E2A indeed enhanced the expression levels of CD11b and CD14 proteins induced by ATPR ( Figure 3D).
Flow cytometry analysis further demonstrated that ATPR-induced AML cell differentiation was enhanced by the loss of E2A ( Figure 3E).
Additionally, Wright-Giemsa staining analysis revealed that the AML cells pretreated with the lentivirus group exhibited a more significant maturity, and the appearance of kidney-shaped nuclei and decreased nuclear/cytoplasm ratio could be detected compared with ATPR treatment alone ( Figure 3F). NBT staining analysis also confirmed these results ( Figure 3G). Differentiation was usually accompanied by cycle arrest, so we performed western blotting and demonstrated  that the decrease in ATPR-induced cell cycle-related proteins (cyclin D3, cyclin A2, Pr-b and CDK4) was enhanced after pretreatment with sh-E2A lentivirus ( Figure 3H). Similarly, the results of flow cytometry analysis showed that compared with ATPR treatment alone, the combined treatment of sh-E2A and ATPR revealed a significant increase in the number of arrested cells at the G0/G1 phase and a significant decrease in the S phase ( Figure 3I). Taken together, these results indicate that ATPR-induced AML cell differentiation and cycle arrest were further enhanced in the absence of E2A.

| Overexpression of E2A reversed ATPRinduced differentiation and cycle arrest of AML cells
To  Figure 5C). Finally, the most abundant upregulated MF terms were related to DNA secondary structure binding, four-way junction DNA binding, single-stranded DNA binding, coenzyme binding and damaged DNA binding ( Figure 5D).

| E2A drives AML by upregulating c-Myc and inhibiting the P53 signalling pathway, and ATPR could suppress the E2A/c-Myc axis
The oncogenic protein c-Myc may be a candidate for mediating the effect of E2A on cell differentiation and cycle arrest. P53 is a crucial tumour suppressor protein and deficiency in the majority of AML, which plays a central role in modulating diverse cell cycle and proliferation processes. 26 Next, we used western blotting to explore the expression of c-Myc/P53 in the progression of AML compared with control group, and determined that c-Myc was expressed at higher levels in AML specimens and cell lines, whereas P53 expression was inhibited ( Figure 6A, B). c-Myc was identified as one of the downregulated genes by RNA-seq after sh-E2A ( Figure 5D), and induced the development of human AML. 27 Next, by performing Co-IP analysis, we were confident that E2A could physically bind to c-Myc in AML F I G U R E 3 ATPR-induced AML cell differentiation and cycle arrest were enhanced in the absence of E2A. NB4/THP-1 cells were transfected with lentivirus containing E2A shRNA to downregulated E2A expression. After cells were exposed to ATPR (10 −6 M) for another 72 h, a series of independent experiments were conducted as follows. (A) The stable control and sh-E2A-transfected NB4/THP-1 cells were observed by an inversed fluorescent microscope. (B) After treatment with sh-E2A for 3 days, the protein expression of E2A was assessed by western blotting. (C) After treatment with sh-E2A for 3 days, the mRNA expression of E2A was assessed by qPCR. (D) After E2A was silenced, cell surface differentiation antigens CD11b and CD14 levels were assessed by western blotting. (E) After E2A was silenced, cell differentiation status was further measured using flow cytometry analysis. (F) After E2A was silenced, cell morphological assays were assessed by Wright-Giemsa staining. Arrows indicate cells with matured morphology, which exhibit kidney-shaped nucleus and decreased nuclear/cytoplasm ratio. (G) After E2A was silenced, the inversed fluorescent microscope was used to observe the positive cell by NBT

| Loss of E2A inhibits tumorigenesis in vivo
To further substantiate the role of E2A in tumour growth in vivo, shNC/shE2A-transfected NB4 and THP-1 cells were subcutaneously injected into NSG mice. The results showed that compared with the control group, the tumours in the sh-E2A group developed sluggishly ( Figure 7A), and the tumour volume and average tumour weight were significantly reduced ( Figure 7B and C). Western blotting showed that sh-E2A significantly increased the levels of CD11b and CD14 proteins and inhibited the expression of cyclin D3 and p-Rb compared with the control tumours ( Figure 7D). IHC staining revealed that the E2A/c-Myc and cyclin A2/CDK4 genes in the sh-E2A tumour group were significantly downregulated, whereas the level of differentiation-related protein CD11b was upregulated ( Figure 7E). Meanwhile, western blotting analysis ( Figure 7F) also revealed that sh-E2A increased the expression of P-P53/P53 and decreased c-Myc levels, reconfirming that E2A inhibits c-Myc and regulates the P53 signalling pathway. Collectively, these results indicate that E2A silencing slows tumorigenesis in vivo.

| DISCUSS ION
Acute myeloid leukaemia is a heterogeneous disease caused by multiple genetic mutations and cytogenetic abnormalities that involve aetiology, and mechanisms underlying this pathogenesis are extremely complex. 28 It is characterized by the malignant proliferation and stagnation of differentiation of myeloid progenitor cells (blasts), which result in damage to the immune system and eventually death26. Current treatment methods, such as cytarabine + anthracycline (7 + 3) and stem cell transplantation (Allo-SCT), have severe side effects and poor tolerance, which do not satisfy the treatment needs of AML27. Therefore, differentiation therapy is a novel treatment for AML. As a classic differentiation-inducing drug, ATRA has transformed APL from being highly fatal to being highly curable, and the ATRA/arsenic combination programme has cured almost all patients with standard risk APL. 29,30 Interestingly, the effect of ATRAinduced differentiation is insufficient for APL eradication, whereas only PML/RARA loss fully extinguish leukaemia-initiating activity. 31 Therefore, ATRA/arsenic combination programme exerts an anti-tumour effect by induced PML/RARA degradation in cells, from which anti-tumour efficacy could be dissociated from the trancriptomic effect. 32 This suggests that targeting the degradation of oncoproteins may be a viable therapeutic strategy for some malignancies.
However, ATRA adverse reactions, including drug resistance and retinoic acid syndrome, have promoted the development of safer and more effective therapeutic drugs. 33 ATPR is a novel derivative of ATRA, which has been proven to have a good anti-cancer effect on a variety of malignant tumours, and it has the advantages of a longer-lasting curative effect, higher solubility and lower toxicity than ATRA. Therefore, our study will further explore the mechanism of ATPR-induced differentiation and cycle arrest in AML cells and The P53 signalling pathway plays a significant role in regulating the cell cycle, proliferation and suppressing tumour expression. 37 Previous studies have indicated that the P53 signalling pathway participates in the regulation of biological processes during normal and leukaemic haematopoiesis. 38 Transcriptome sequencing data substantiated that the P53 signalling pathway was activated after E2A knockout compared to the control group. Current studies have suggested that silencing E2A could enhance cell cycle arrest and p-P53 protein levels induced by ATPR; however, the exact mechanism by which E2A regulates the P53 signalling pathway requires further investigation.
Overall, the findings of this study highlight a novel mechanism of ATPR in the treatment of AML. The present study showed that ATPR could downregulate the expression of E2A by binding to RARα, inhibit the downstream target gene c-Myc and induce cell differentiation and cycle arrest via the P53 signalling pathway (Figure 8). These studies provide evidence that E2A plays a pivotal role in the treatment of AML and provides novel insights for further study of ATPR.

ACK N OWLED G EM ENTS
We thank the First Affiliated Hospital of Anhui Medical University (Hefei, China) for providing the human PBMCs and umbilical cord blood.

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to disclose.  F I G U R E 7 Loss of E2A inhibits tumorigenesis in vivo. 24 NSG mice were randomly divided into four groups (n = 6). NB4/THP-1 cells (5 × 10 6 ) transfected with sh-NC and sh-E2A were injected subcutaneously in the right shoulder. (A) Tumour images of the xenograft mice were taken at the end of the experiment (n = 6 mice per group). (B and C) The tumour volume and weight of xenograft mice were measured during the observation period. (D) Two representative tumour tissues from each group were fixed, and immunohistochemistry staining was performed on E2A, c-Myc, CD11b, cyclin A2 and CDK4. (E) Western blotting analysis of CD11b, CD14, Pr-b and cyclin D3 in tumour tissues of sh-NC and sh-E2A groups. (F) Western blotting analysis of P-P53/p53 and c-Myc in tumour tissues of sh-NC and sh-E2A groups. β-Actin was used as an internal control. Bar graphs (mean ± SD) and representative images are shown. *p < 0.05, **p < 0.01, compared with the NC group

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are included within the article.