Circadian clock protein Bmal1 accelerates acute myeloid leukemia by inhibiting ferroptosis through the EBF3/ALOX15 axis

Abstract

Acute myeloid leukemia (AML) is a major leukemia with high mortality. Ferroptosis is an important regulator of cancers. However, the role of ferroptosis and its regulatory mechanisms in AML remain largely unknown. In this study, we reported elevated brain and muscle ARNT‐Like protein‐1 (Bmal1) expression in AML patients and cell lines, and its upregulation indicated the poor survival of patients. The correlation analysis showed that Bmal1 expression was closely correlated with cytogenetics and the French–American–British subtypes, but was not correlated with age, gender and white blood cells. RSL3 reduced Bmal1 expression in HL‐60 and NB4 cells. Malondialdehyde, total iron, Fe²⁺, glutathione and lipid peroxidation were examined to evaluate ferroptosis. Overexpression of Bmal1 repressed RSL3‐induced ferroptosis in AML cells. Bmal1 recruited Enhancer of zeste homolog 2 (EZH2) to the Early B cell factor 3 (EBF3) promoter and enhanced its methylation, thus suppressing EBF3 expression. Moreover, the knockdown of Bmal1 sensitized AML cells to RSL3‐induced ferroptosis, and it was counteracted by EBF3 knockdown. Furthermore, EBF3 bound to the Arachidonate 15‐pipoxygenase (ALOX15) promoter to enhance its expression, and overexpression of EBF3 enhanced RSL3‐induced ferroptosis dependent on ALOX5. We established a subcutaneous AML xenograft tumor model and reported that knockdown of Bmal1 and overexpression of EBF3 restrained AML growth by promoting ALOX15‐mediated ferroptosis in vivo. Collectively, Bmal1 inhibits RSL3‐induced ferroptosis by promoting EZH2‐mediated EBF3 methylation and suppressing the expression of EBF3 and ALOX15, thus accelerating AML.

Bmal1 inhibits RSL3‐induced ferroptosis by promoting EZH2‐mediated EBF3 methylation and suppressing the expression of EBF3 and ALOX15, thus accelerating AML.

Abbreviations

ALOX15

arachidonate 15‐pipoxygenase

AML

acute myeloid leukemia

BM

bone marrow

Bmal1

Brain and muscle ARNT‐Like protein‐1

BMMNCs

bone marrow mononuclear cells

BMSCs

bone marrow‐derived mesenchymal stromal cells

Co‐IP

co‐immunoprecipitation

EBF3

Early B cell factor 3

EZH2

Enhancer of zeste homolog 2

FAB

French–American–British

GPX4

glutathione peroxidase 4

GSH

glutathione

IHC

immunohistochemistry

LSCs

leukemia stem cells

MDA

malondialdehyde

MSP

methylation‐specific PCR

qRT‐PCR

quantitative RT‐PCR

WBC

white blood cell

1. INTRODUCTION

Acute myeloid leukemia is a fast‐growing blood malignancy that starts in the BM and was characterized by arrest of the maturation of hematopoietic progenitor cells and their accumulation in the blood, BM and other tissues, causing BM failure. ¹ AML accounts for ~80% of all leukemia cases in adults. Although it only accounts for 1% of all cancers, AML shows a low 5‐year survival rate especially for elderly patients, and the burden of the incidence and AML‐related deaths is increasing worldwide. ² , ³ Chemotherapy is the standard treatment for AML, and BM or stem cell transplants may be another option for patients with chemoresistance. Novel medications, such as targeted drugs, are being investigated to further improve the therapeutic effects. ⁴ Exploring the mechanisms underlying AML pathogenesis is crucial for developing promising drugs.

As endogenous oscillators, circadian clocks govern 24‐h rhythms for biochemical, physiological and behavioral processes maintaining homeostasis. ⁵ Dysregulated circadian rhythms serve vital roles in tumorigenesis and promote the hallmarks of various cancers. ⁶ , ⁷ Bmal1 is a key circadian clock that drives circadian rhythms and confers circadian regulation of the expression of downstream target genes. ⁸ , ⁹ A recent study has proved that Bmal1 is essential for maintaining the growth of LSCs in AML, and that depletion of Bmal1 generates anti‐leukemia effects. ¹⁰ Ferroptosis is a novel form of iron‐dependent cell death and is characterized using lipid peroxidation and iron accumulation, which is emerging as a vital regulator in cancers. ¹¹ As an important ferroptosis repressor, Bmal1 is downregulated in the ferroptosis inducer RSL3‐treated AML HL‐60 cells, and suppression of Bmal1 sensitizes cancer cells to ferroptosis. ¹² Further exploration of Bmal1‐mediated repression of ferroptosis in AML contributes to developing novel therapeutic strategies that target ferroptosis.

Typical characteristics of ferroptosis include the downregulation of GPX4 and GSH and increased lipid peroxidation. ALOX15, a lipoxygenase, contributes to ferroptosis by increasing lipid peroxidation and cellular reactive oxygen species. ¹³ Ryosuke Shintoku et al. reported that suppression of ALOX15 reduced ferroptotic death of erastin and RSL3‐treated cancer cells. ¹⁴ ALOX15 is required for the survival of LSCs, ¹⁵ and inhibitors of ALOX15 reduce ferroptosis in acute lymphoblastic leukemia cells. ¹⁶ However, the role of ALOX15 in AML remains unknown.

EBF3 is a transcription factor that regulates the expression of various downstream target genes by binding to DNA. ¹⁷ Tao et al. found that the EBF3 promoter was highly methylated, and its expression was markedly downregulated in AML. Overexpression of EBF3 repressed AML cell proliferation and enhanced apoptosis. ¹⁸ We identified a potential binding site for EBF3 in the ALOX15 promoter through bioinformatics, suggesting that EBF3 might regulate ALOX15 transcription. Bmal1 directly interacts with the Polycomb group protein EZH2, ¹⁸ and EZH2 governs promoter methylation of target genes that silence gene expression. ¹⁹ Therefore, we hypothesized that Bmal1 might interact with EZH2 to promote EBF3 promoter methylation and suppress the expression of EBF3 and downstream target ALOX15, thus reducing ferroptosis and accelerating AML.

Here, we first report that AML patients with high Bmal1 expression showed poor survival. We demonstrate that circadian clock protein Bmal1 promotes AML progression by inhibiting RSL3‐induced ferroptosis through EZH2‐mediated EBF3 promoter methylation and downregulation of ALOX15. Our study not only sheds novel light on the regulation of ferroptosis in AML, but also provides potential therapeutic targets.

2. METHODS

2.1. Patients and clinical specimens

BM specimens were harvested from healthy BM transplant donors (n = 10) and patients with AML (n = 52) diagnosed between 2020 and 2022 at the Third Xiangya Hospital of Central South University. Specimens were stored at −80°C for RNA extraction. Patients (n = 46) diagnosed with AML at the Third Xiangya Hospital of Central South University in 2015 were divided into low‐ and high‐expression groups (n = 23 per group) based on the median expression of Baml1, and patient survival was monitored. The information on patients diagnosed in 2015 is shown in Table 1. BMMNCs were isolated from BM specimens using lymphocyte separation medium. This study was approved by the Ethics Committee of the Third Xiangya Hospital of Central South University. All patients and donors provided written informed consent.

TABLE 1.

Correlation between Bmal1 expression and different clinicopathological characteristics in patients with AML.

Clinical parameters	Cases (n)	Bmal1 expression		p‐value
		High (n)	Low (n)
Age
<55 years	27	12	15	0.369
≥55 years	19	11	8
Sex
Female	14	8	6	0.522
Male	32	15	17
WBC(/μL)
<10,000	27	16	11	0.134
≥10,000	19	7	12
Cytogenetics
Favorable	17	5	12	0.038*
Intermediate	19	10	9
Unfavorable	10	8	2
FAB subtype
M1–M6	34	15	19	0.179
M7	12	8	4
Response to treatment
CR	34	16	18	0.502
NR	12	7	5

Abbreviations: AML, acute myeloid leukemia; CR, complete response; FAB, French–American–British classification; NR, no response; WBC, white blood cell.

^* p<0.05.

2.2. Cell culture

Human BMSCs and AML cell lines THP‐1, HL‐60, and KG‐1 were provided by the American Type Culture Collection. NB4 cells were obtained from iCell Bioscience. Cells were maintained in RPMI 1640 medium supplemented with 10% FBS (Gibco). For inducing ferroptosis, HL‐60 and NB4 cells were treated with RSL3 (Selleck) at 0.5 μM for 12 h and collected for analysis of ferroptosis markers.

2.3. Vector construct and cell transfection

shRNAs against Bmal1 (shBmal1#1 and #2), EBF3 (shEBF3) and ALOX15 (shALOX15) were provided by RiboBio and cloned into the pLKO.1 lentiviral vector (Addgene). Coding sequences for Bmal1, EZH2 and EBF3 were inserted into the pLenti‐puro lentiviral vector (Addgene). Lentiviral particles were packaged into HEK293T cells through transfection of pLenti‐puro‐Bmal1, pLenti‐puro‐EZH2, pLenti‐puro‐EBF3 or pLKO.1‐shBaml1, psPAX2 (Addgene) and pMD2.G (Addgene) using Lipofectamine 3000 (Thermo Fisher Scientific). After 72 h, lentiviral particles were collected and filtered. Subsequently, HL‐60 and NB4 cells were infected with lentiviral particles at a 100 multiplicity of infection (MoI). After 16 h, the medium was replaced with fresh medium, and cells with stable transfection were selected using puromycin.

2.4. Cell Counting Kit‐8 assay

Cells were placed into 96‐well plates at a density of 5 × 10³ cells/well and treated with RSL3 for 12 h. After treatment, 10 μL of CCK‐8 solution was added, and cells were incubated for 3 h at 37°C. Finally, the absorbance at 450 nm was recorded using a microplate reader. The CCK‐8 kit was obtained from Beyotime.

2.5. Analysis of ferroptosis markers

After treatment, HL‐60 and NB4 cells were lysed on ice to harvest cell lysates. The levels of MDA and GSH were analyzed using the MDA Assay Kit (PromoCell) and the Glutathione Colorimetric Detection Kit (BioVision) according to their manuals. Total iron and Fe²⁺ were examined using the Iron Assay Kit (Abcam) and following the manual.

2.6. Lipid peroxidation analysis

After treatment, HL‐60 and NB4 cells were incubated with BODIPY™ 581/591 C11 (Thermo Fisher Scientific), a widely used lipid peroxidation sensor, ²⁰ at 5 μM for 30 min at 37°C. Subsequently, cells were washed and resuspended in PBS for flow cytometry analysis of green fluorescence in the FITC channel.

2.7. Methylation‐specific PCR for assessing methylation status

Genomic DNA was isolated from BMSCs, THP‐1, HL‐60, NB4 and KG‐1 cells using the PureLink™ Genomic DNA Mini Kit (Thermo Fisher Scientific) and subjected to methylation‐specific PCR. Methylation‐specific PCR was performed using the EpiScope MSP Kit provided by TaKaRa and following the manual. PCR products (5 μL) were loaded onto 3% agarose gels and electrophoresed.

2.8. Chromatin immunoprecipitation assay

HL‐60 and NB4 cells (~5 × 10⁶) were washed and incubated in 1% formaldehyde solution on ice for crosslinking. Then, the formaldehyde solution was removed, and cells were washed, detached and suspended in cell lysis buffer for 30 min on ice. Cell lysates were collected and sonicated to obtain DNA fragments (~500 bp). Rabbit anti‐human EZH2 (5 μg, Abcam), H3K27me3 (5 μg, Abcam) and EBF3 (5 μg, Thermo Fisher Scientific) and isotype control IgG (5 μg) were added to the DNA fragments and incubated for 16 h at 4°C. Subsequently, Protein A/G Magnetic Beads (Abcam) were mixed with samples and incubated for 2 h. DNA was eluted and subjected to qPCR. Primers are shown in Table S1.

2.9. Co‐immunoprecipitation assay

HL‐60 and NB4 cells (~3 × 10⁶) were washed, detached and suspended in cell lysis buffer for 30 min on ice. Cell lysates were collected, and 10% of the cell lysates were used as the input. The remaining lysates were aliquoted into two parts for IP with rabbit anti‐human Bmal1 (3 μg, Abcam) and an isotype control IgG (3 μg, Abcam). Samples were incubated at 4°C overnight, and Protein A/G Magnetic Beads were added and incubated for 2 h. Magnetic beads were washed, and immunoprecipitated complexes were recovered and subjected to western blotting to detect Bmal1 (1:1000, Abcam) and EZH2 (1:1000, Abcam).

2.10. Dual‐luciferase reporter assay

ALOX15 promoter sequences (Pro, Pro#1, 2 and 3) containing potential EBF3 binding sites (Sites 1, 2 and 3) and EBF3 promoter were inserted into the pGL3 luciferase vector provided by Promega as luciferase reporters. HL‐60 and NB4 cells were transfected with the EBF3‐overexpressing pcDNA3.1 vector and luciferase reporters. After 48 h, luciferase activity was examined using the Dual‐Glo Luciferase System (Promega) and following the manual.

2.11. A subcutaneous AML xenograft tumor mouse model

BALB/c nude mice (male, 6‐week‐old) were purchased from SLAC Jingda Laboratory Animal Co., Ltd. and divided into four groups: shNC + Vector, shNC + Vector + RSL3, shBmal1 + Vector + RSL3, and shNC + EBF3 + RSL3. HL‐60 and NB4 cells with the indicated transfection were collected, washed and resuspended in PBS. Cells in 100 μL of PBS (5 × 10⁶ cells/mouse) were implanted into the left flank of nude mice. When tumor volume reached 50 mm³, mice received intratumor injections of RSL3 at 100 mg/kg. ²¹ Tumor volume was monitored using a Vernier caliper and calculated using the formula: length × width²/2. Ultimately, mice were sacrificed, and tumors were excised for subsequent assays. Animal experiments were approved by the Animal Care and Use Committee of Third Xiangya Hospital of Central South University.

2.12. Immunohistochemistry staining

Tumors were immersed in 4% formaldehyde solution at 4°C overnight for fixation. The following day, tumors were dehydrated and subjected to paraffin embedding, the embedded material was subsequently cut into sections (5 μm). Sections were dewaxed and rehydrated followed by antigen retrieval in Antigen Retrieval Buffer (pH 6.0, Abcam). Endogenous peroxidase was inactivated in 0.5% hydrogen peroxide solution. After blocking, sections were incubated with rabbit anti‐human Ki‐67 (1:200, Abcam) and ALOX15 (1:100, Abcam) overnight. Sections were washed and incubated with a goat anti‐rabbit HRP‐conjugated secondary antibody (1:5000, Abcam). Signals were visualized using DAB (Solarbio) followed by hematoxylin staining and mounting for imaging.

2.13. RNA extraction and qRT‐PCR

Total RNA was extracted from BM specimens using the PAXgene Bone Marrow RNA Kit (QIAGEN) following the manual. Tumors were homogenized in liquid nitrogen and resuspended in TRIzol reagent (Beyotime) for RNA extraction. BMSCs, BMMNCs and AML cells were washed and resuspended in TRIzol reagent, and total RNA was isolated and subjected to RNA quantification. Subsequently, RNA was reversely transcribed into cDNA using the BeyoRT™ II First Strand cDNA Synthesis Kit (Beyotime). Quantitative PCR was applied to detect the expression of Bmal1, EBF3, EZH2 and ALOX15. Gene expression was normalized to β‐actin, and the 2^−ΔΔCt method was used for calculation. Primers are shown in Table S2.

2.14. Western blot

Tumors were homogenized and incubated in lysis buffer for 2 h on ice. BMSCs and AML cells were washed, resuspended in RIPA Lysis and Extraction Buffer (Thermo Fisher Scientific) and incubated for 30 min on ice. The supernatants from tumor homogenates and cell lysates were collected followed by protein quantification using the BCA kit from Solarbio. Protein was electrophoresed and transferred to PVDF membranes. Membranes were blocked and incubated with rabbit anti‐human Bmal1 (1: 1000), GPX4 (1:500), ALOX15 (1:1000), EBF3 (1:1000) and β‐actin (1:5000) overnight. Membranes were washed and incubated with an HRP goat anti‐rabbit secondary antibody (1:10,000) for 1 h. ECL substrates (Beyotime) were added for visualizing bands, and bands were analyzed using ImageJ software.

2.15. Statistical analysis

Data are shown as the mean ± standard deviation (SD). Student's t‐test and one‐way ANOVA were applied to analyze the variance of two and more groups. The median expression of Bmal1 in patients was used as the threshold to divide patients into low‐ and high‐expression groups. Patient survival was analyzed using Kaplan–Meier curves. A p‐value < 0.05 was considered statistically significant.

3. RESULTS

3.1. Bmal1 was upregulated in AML cells and patients and its elevated expression indicated poor survival of patients

We collected BM specimens from 52 AML patients and 10 healthy donors and found that Bmal1 was significantly upregulated in patient specimens (Figure 1A). Importantly, patients with high expression of Bmal1 showed poor survival compared with patients with low Bmal1 expression (Figure 1B), indicating that Bmal1 might contribute to AML progression. We observed that Bmal1 expression was closely correlated with cytogenetics in patients diagnosed in 2015, but it was not significantly correlated with age, gender, WBCs, the FAB subtypes and response to treatment (Table 1). We further examined Bmal1 expression in leukemia cell lines including THP‐1, HL‐60, NB4 and KG‐1. Compared with BMSCs and BMMNCs, all these cells showed elevated levels of Bmal1 (Figure 1C,D). HL‐60 and NB4 cells were used in subsequent assays as they had the highest Bmal1 expression. As Bmal1 inhibits ferroptosis in Calu‐1 and HT1080 cells, ¹² we treated cells with RSL3, an inducer of ferroptosis. ²² We observed that RSL3‐treated cells showed decreased Bmal1 expression (Figure 1E,F), indicating that Bmal1 might be implicated in the regulation of RSL3‐induced ferroptosis in AML.

FIGURE 1.

Bmal1 is upregulated in AML cells and patients and its elevated expression indicates the poor survival of patients. (A) Bmal1 expression in bone marrow (BM) specimens from patients (n = 52) diagnosed with AML between 2020 and 2022 and donors (n = 10). (B) Patients diagnosed with AML in 2015 (n = 46) were divided into low‐ or high‐expression groups (n = 23 per group) based on the median expression of Baml1, and patient survival was monitored. (C, D) qRT‐PCR and western blot analysis of Baml1 expression in BMSCs, BMMNCs, THP‐1, HL‐60, NB4 and KG‐1 cells. (E, F) Bmal1 expression in vehicle and RSL3‐treated HL‐60 and NB4 cells. *p < 0.05, **p < 0.01 and ***p < 0.001.

3.2. Overexpression of Bmal1 repressed RSL3‐induced ferroptosis of AML cells

To further confirm Bmal1‐mediated regulation of ferroptosis, Bmal1 was overexpressed in HL‐60 and NB4 cells, and its overexpression was verified at both the mRNA and protein levels (Figure 2A,B). Cells were treated with RSL3 for ferroptosis induction. We found that RSL3‐mediated suppression of cell viability was largely abrogated by Bmal1 overexpression (Figure 2C). As expected, RSL3 induced high levels of MDA, total iron and Fe²⁺, but reduced the GSH levels in HL‐60 and NB4 cells (Figure 2D–G), showing classical ferroptosis characteristics. ²³ Importantly, all these effects induced by RSL3 were reversed by overexpression of Bmal1 (Figure 2D–G). Furthermore, flow cytometry analysis showed that RSL3 enhanced lipid peroxidation, but this was inhibited by Bmal1 overexpression (Figure 2H). Also, we found that GPX4 was downregulated and ALOX15 was upregulated in RSL3‐treated cells, and overexpression of Bmal1 reversed their expression (Figure 2I). Collectively, these observations suggested that Bmal1 negatively regulated RSL3‐induced ferroptosis in AML cells.

FIGURE 2.

Overexpression of Bmal1 represses RSL3‐induced ferroptosis of AML cells. Bmal1 was overexpressed in HL‐60 and NB4 cells. Cells were treated with RSL3 at 0.5 μM for 12 h. (A, B) Bmal1 expression in Bmal1‐overexpressing and vector‐transfected HL‐60 and NB4 cells was analyzed by qRT‐PCR. (C) Cell viability analysis using CCK‐8. The levels of MDA (D), total iron (E), Fe²⁺ (F) and GSH (G) were measured. (H) Flow cytometry analysis of lipid peroxidation. (I) Protein levels of GPX4 and ALOX15 were examined using western blotting. β‐Actin was used as a reference control. *p < 0.05, **p < 0.01 and ***p < 0.001.

3.3. Bmal1 facilitated EZH2‐mediated methylation of the EBF3 promoter to inhibit EBF3 expression in AML cells

As EBF3 is involved in AML cell proliferation and apoptosis, ¹⁸ we examined its expression in patients and AML cells. EBF3 was downregulated in patients and AML cells, and its expression was negatively correlated with Bmal1 expression in patients (Figure 3A–C). Compared with BMSCs and BMMNCs, AML cells showed high methylation levels of the EBF3 promoter (Figure 3D), indicating that Bmal1 might regulate EBF3 expression by modifying its methylation. We further confirmed that EZH2 could be co‐immunoprecipitated by a Bmal1 antibody from HL‐60 and NB4 cell lysates (Figure 3E). To investigate Bmal1 and EZH2‐mediated regulation of EBF3 expression, Bmal1 was knocked down and EZH2 was overexpressed in HL‐60 and NB4 cells, and their expression levels were confirmed by qRT‐PCR and western blotting (Figure 3F–I). Knockdown of Bmal1 promoted EBF3 expression at both the mRNA and protein levels, but it was abolished by simultaneous EZH2 overexpression (Figure 3J,K). Furthermore, ChIP assays showed that the EBF3 promoter could be enriched by EZH2 and H3K27me3 antibodies, but the knockdown of Bmal1 significantly reduced the enrichment (Figure 3L). Moreover, the knockdown of Bmal1 promoted the luciferase activity of the EBF3 promoter, but this was inhibited by simultaneous EZH2 overexpression (Figure 3M). Collectively, our results implied that Bmal1 repressed EBF3 expression by recruiting EZH2 to the promoter of EBF3 and promoting its methylation in AML cells.

FIGURE 3.

Bmal1 facilitates EZH2‐mediated methylation of EBF3 promoter to inhibit EBF3 expression in AML cells. (A) EBF3 expression in patients (n = 52) and donors (n = 10). (B) Correlation analysis of Bmal1 and EBF3 expression in patients. (C) Relative expression levels of EBF3 in BMSCs, BMMNCs, THP‐1, HL‐60, NB4 and KG‐1 cells were analyzed by qRT‐PCR. (D) Methylation‐specific PCR products were subjected to electrophoresis. (E) The interaction of EZH2 and Bmal1 was analyzed through co‐IP assays. (F, G) Bmal1 was knocked down in HL‐60 and NB4 cells. (H, I) EZH2 was overexpressed in HL‐60 and NB4 cells. (J, K) qRT‐PCR and western blot analysis of EBF3 expression. β‐Actin was used as a reference control. (L) ChIP analysis of the interaction of EZH2 or H3K27me3 and the EBF3 promoter in Bmal1‐knockdown AML cells. (M) Luciferase activity of the EBF3 promoter reporter. **p < 0.01 and ***p < 0.001.

3.4. Knockdown of EBF3 abrogated Bmal1 silencing‐mediated facilitation of RSL3‐induced ferroptosis

EBF3 was knocked down in HL‐60 and NB4 cells by shEBF3 transfection (Figure 4A,B). Knockdown of Bmal1 further inhibited cell viability, elevated the levels of MDA, total iron and Fe²⁺, but reduced GSH levels in RSL3‐treated HL‐60 and NB4 cells, whereas simultaneous EBF3 silencing reversed the Bmal1 knockdown‐mediated effects (Figure 4C–G). In addition, RSL3‐induced lipid peroxidation was further promoted by Bmal1 knockdown, which was inhibited by simultaneous EBF3 silencing (Figure 4H). Furthermore, GPX4 expression was inhibited and ALOX15 expression was promoted by Bmal1 knockdown in RSL3‐treated HL‐60 and NB4 cells, and silencing of EBF3 reversed their expression (Figure 4I). These findings suggested that Bmal1 silencing‐mediated facilitation of RSL3‐induced ferroptosis in AML cells was dependent on the upregulation of EBF3.

FIGURE 4.

Knockdown of EBF3 abrogates Bmal1 silencing‐mediated facilitation of RSL3‐induced ferroptosis. Bmal1 and EBF3 were knocked down in HL‐60 and NB4 cells. Cells were treated with RSL3 at 0.5 μM for 12 h. (A, B) EBF3 expression analysis. (C) Cell viability analysis. The levels of MDA (D), total iron (E), Fe²⁺ (F) and GSH (G) were measured. (H) Flow cytometry analysis of lipid peroxidation. (I) Protein levels of GPX4 and ALOX15 were examined using western blotting. β‐Actin was used as a reference control. *p < 0.05, **p < 0.01 and ***p < 0.001.

3.5. EBF3 binds to the ALOX15 promoter to promote its expression in AML cells

We identified three potential binding sites for EBF3 in the ALOX15 promoter through JASPAR database analysis (Figure 5A). ChIP assays showed only Site 1, but not Sites 2 and 3, was enriched by the EBF3 antibody, suggesting that EBF3 could bind to Site 1 in the ALOX15 promoter (Figure 5B). Subsequently, EBF3 was overexpressed in HL‐60 and NB4 cells (Figure 5C,D), and various promoter sequences in ALOX15 (Pro: wild‐type Sites 1, 2 and 3; Pro#1: wild‐type Site 1 and mutant Sites 2 and 3; Pro#2: wild‐type Site 2 and mutant Sites 1 and 3; Pro#3: wild‐type Site 3 and mutant Sites 1 and 2) were inserted into the pGL3 vector as luciferase reporters. We found that overexpression of EBF3 significantly enhanced the luciferase activity of Pro and Pro#1 luciferase reporters, but did not affect the activity of Pro#2 and Pro#3 reporters in HL‐60 and NB4 cells (Figure 5E). In addition, the expression of ALOX15 was markedly promoted by EBF3 overexpression, but was inhibited by EBF3 knockdown in HL‐60 and NB4 cells (Figure 5F,G). Our data demonstrated that EBF3 binds directly to the ALOX15 promoter at Site 1 to facilitate its expression in AML.

FIGURE 5.

EBF3 Bound to the ALOX15 promoter to promote its expression in AML cells. (A) Three potential binding sites (Sites 1, 2 and 3) for EBF3 in the ALOX15 promoter. (B) ChIP analysis for the binding of EBF3 to Sites 1, 2 and 3. (C, D) EBF3 expression analysis in HL‐60 and NB4 cells. (E) Luciferase activity of various ALOX15 promoter reporters. (F, G) The expression of ALOX15 in EBF3‐overexpressing or knockdown cells was analyzed by qRT‐PCR and western blotting. β‐Actin was used as the reference control. *p < 0.05, **p < 0.01 and ***p < 0.001.

3.6. Knockdown of ALOX15 suppressed EBF3 overexpression‐mediated promotion of RSL3‐induced ferroptosis of AML cells

ALOX15 was knocked down in HL‐60 and NB4 cells via shALOX15 transfection (Figure 6A,B). RSL3‐induced suppression of cell viability and accumulation of MDA, total iron and Fe²⁺ were further promoted and GSH levels were reduced in HL‐60 and NB4 cells by overexpression of EBF3, whereas simultaneous knockdown of ALOX15 suppressed these effects (Figure 6C–G). Moreover, EBF3 overexpression further promoted RSL3‐induced lipid peroxidation, which was reversed by the knockdown of ALOX15 (Figure 6H). Additionally, overexpression of EBF3 suppressed GPX4 expression and promoted ALOX15 expression in RSL3‐treated cells, but these expression patterns were reversed by ALOX15 silencing (Figure 6I). These observations showed that EBF3 enhanced RSL3‐induced ferroptosis through the upregulation of ALOX15 in AML.

FIGURE 6.

Knockdown of ALOX15 suppressed EBF3 overexpression‐mediated promotion of RSL3‐induced ferroptosis of AML cells. ALOX15 was knocked down and EBF3 was overexpressed in HL‐60 and NB4 cells. Cells were treated with RSL3 at 0.5 μM for 12 h. (A, B) ALOX15 expression in shNC or shALOX15‐transfected HL‐60 and NB4 cells was analyzed by qRT‐PCR and western blotting. (C) Cell viability analysis. The levels of MDA (D), total iron (E), Fe²⁺ (F) and GSH (G) were measured. (H) Flow cytometry analysis of lipid peroxidation. (I) Protein levels of GPX4 and ALOX15 were examined by western blotting. β‐Actin was used as a reference control. *p < 0.05, **p < 0.01 and ***p < 0.001.

3.7. Silencing of Bmal1 and overexpression of EBF3 inhibited subcutaneous AML tumor growth by enhancing ALOX15‐mediated ferroptosis in mice

To evaluate the role of Bmal1 and EBF3 in vivo, we established a mouse model of a subcutaneous AML xenograft tumor through subcutaneous injection of AML cells into the left flank of mice. RSL3 administration obviously reduced tumor size, volume and weight in mice (Figure 7A–C). Intriguingly, RSL3‐mediated suppression of tumor growth was more significant in mice injected with Bmal1‐knockdown and EBF3‐overexpressing cells (Figure 7A–C), suggesting that knockdown of Bmal1 and overexpression of EBF3 enhanced RSL3‐induced ferroptosis. Further investigation showed that RSL3‐induced ALOX15 expression was further enhanced by Bmal1 knockdown and EBF3 overexpression (Figure 7D,E). By contrast, the expression of GPX4 was inhibited by RSL3 administration, and tumors with Bmal1 silencing and EBF3 overexpression showed the lowest GPX4 expression (Figure 7E). IHC staining also confirmed the expression patterns of ALOX15 (Figure 7F). In addition, RSL3 reduced the expression of Ki‐67, a classical marker for cell proliferation, in tumors, and RSL3‐mediated downregulation of Ki‐67 was reinforced by Bmal1 knockdown and EBF3 overexpression (Figure 7F). Collectively, our findings suggested that silencing of Bmal1 and overexpression of EBF3 restrained AML growth by sensitizing AML cells to RSL3‐induced ferroptosis through ALOX15 in vivo.

FIGURE 7.

Silencing of Bmal1 and overexpression of EBF3 inhibits subcutaneous AML tumor growth by enhancing ALOX15‐mediated ferroptosis in mice. A mouse model of AML xenograft tumor was established by subcutaneous injection of AML cells stably expressing shBmal1 and EBF3 into the left flank of mice. RSL3 (100 mg/kg) was intratumorally injected. (A) Photographs of excised subcutaneous tumors (n = 6 each group). (B, C) Tumor volume and weight. (D, E) The expression of ALOX15 and GPX4 in tumors was examined by qRT‐PCR and western blotting. β‐Actin was used as the reference control. (F) IHC staining of ALOX15 and Ki‐67 in tumors. *p < 0.05, **p < 0.01 and ***p < 0.001.

4. DISCUSSION

Acute myeloid leukemia is a rare but lethal cancer of the blood and BM that carries a high recurrence risk and poor survival outcomes. ²⁴ The outcomes of AML patients depend on leukemia subtypes, cytogenetics, age and drug resistance. ²⁵ , ²⁶ Survival rate is poor in elderly AML patients, and the 5‐year survival rate may be less than 5% in patients at 65 and older. ²⁷ Although therapeutic effects have been greatly improved in recent years thanks to advances in the understanding of genetic features of AML and the development of novel therapies, it is of great importance to further explore AML pathogenesis and develop novel drugs. In this study, we demonstrated that Bmal1 was upregulated in AML cells and patients, and increased Bmal1 expression promoted the progression of AML through ferroptosis suppression by facilitating EZH2‐mediated methylation of EBF3 promoter and downregulating EBF3 and ALOX15 (Figure 8).

FIGURE 8.

Schematic diagram of our study. Bmal1 suppresses ferroptosis by enhancing EZH2‐mediated methylation of EBF3 promoter and downregulating EBF3 and its downstream target ALOX15, thus promoting AML progression.

The discovery of ferroptosis has provided novel opportunities for cancer therapy. Many agents, such as RSL3, erastin, statins and sorafenib, trigger ferroptosis and suppress the growth of various tumors. ²⁸ However, ferroptosis may exert oncogenic activities by shaping the immunosuppressive tumor microenvironment. ²⁹ In AML, erastin sensitizes AML cells to chemotherapy by inducing ferroptosis. ³⁰ Moreover, dihydroartemisinin (DHA), typhaneoside (TYP), APR‐246 and circKDM4C have been reported to induce ferroptosis in AML cells. ³¹ However, the regulation of ferroptosis in AML remains largely unknown. Here, we identified a novel ferroptosis regulator in AML. We found that Bmal1 promoted AML through suppression of RSL3‐induced ferroptosis, supporting the anti‐tumor activity of ferroptosis in AML.

Bmal1 shows opposing activities, anti‐tumor and oncogenic, in various cancers. Tang et al. found that Bmal1 suppressed the progression of tongue squamous cell carcinoma. ³² However, Bmal1 facilitates the invasion and metastasis of breast cancer cells. ³³ Intriguingly, a recent study has suggested the oncogenic role of Bmal1 in AML. ¹⁰ Nevertheless, the mechanism underlying the implication of Bmal1 in AML remains largely unknown. Here, we found that Bmal1 was significantly upregulated in AML cells and patients, which is consistent with previous observations in breast cancer cells. ³³ Bmal1 is downregulated in squamous cell carcinoma and breast cancer, which may be associated with elevated lipid peroxidation. ³⁴ Intriguingly, Baml1 represses ferroptosis by suppressing EGLN2 activation and activating HIF‐1α in cancer cells. ¹² Consistently, overexpression of Bmal1 markedly repressed RSL3‐induced ferroptosis in AML cells and enhanced AML growth by reducing ferroptosis in vivo, suggesting a novel mechanism underlying the oncogenic activity of Bmal1 in AML.

Previous studies have identified that Bmal1 directly interacted with EZH2 in various systems. Tang et al. reported that Bmal1 interacted with EZH2 to inhibit telomerase reverse transcriptase (TERT) expression in tongue squamous cell carcinoma cells. ³² Moreover, mouse EZH2 was found to coprecipitate with Bmal1 in liver extracts. ³⁵ Consistently, we found a direct interaction between Baml1 and EZH2 in AML2 cells. EZH2, a histone methyltransferase, inhibits gene transcription via trimethylation of lysine 27 on histone 3. ³⁶ In addition, EZH2 interacts with DNA methyltransferases to negatively regulate gene expression through DNA methylation of target genes in cancers. ³⁷ , ³⁸ BRCA1 promotes FOXO3 expression by regulating the methylation of the FOXO3 promoter through EZH2 in breast cancer. ³⁹ Intriguingly, we found decreased EBF3 expression in AML cells and patients, and that EZH2 could bind directly to the promoter of EBF3 for its methylation and suppression of EBF3 expression in AML.

EBF3 is a transcription factor that controls the expression of target genes and may serve as a tumor suppressor in various cancers. ⁴⁰ Tao and colleagues found that EBF3 was downregulated in AML, and its overexpression suppressed cell proliferation and enhanced apoptosis. ¹⁸ The dysregulated expression of EBF3 in cancers may be attributed to the hypermethylation of its promoter. ⁴¹ However, the downstream signaling pathway of EBF3 in tumor suppression still remains largely unknown, and the implication of EBF3 in ferroptosis has not been reported to date. Here, we observed that overexpression of EBF3 restrained AML tumor growth by enhancing RSL3‐induced ferroptosis and by binding to the AMOX15 promoter and promoting its expression, highlighting the anti‐tumor activity of EBF3 and EBF3‐mediated regulation of ferroptosis in AML. The lipoxygenase ALOX15 is required for lipid peroxidation that contributes to ferroptosis. ¹⁴ , ⁴² PD146176, an inhibitor of ALOX15, effectively suppressed lipid peroxidation and ferroptosis. ⁴³ ALOX15 functions as a tumor suppressor in colorectal cancer, ⁴⁴ whereas ALOX15 is required for LSC survival. ¹⁵ However, the role of ALOX15 in AML is unknown. We found that EBF3 bound directly to the ALOX15 promoter to enhance ALOX15 expression, thus inducing RSL3‐induced ferroptosis in AML, and supporting the notion that ALOX15 serves as a tumor suppressor in AML.

In summary, for the first time, we demonstrated that Bmal1 accelerated AML by repressing RSL3‐induced ferroptosis through EZH2‐mediated downregulation of EBF3 and ALOX15. It will be intriguing in future studies to explore the RSL3 sensitivity changes in actual patients' xenograft cells by different BMAL1 expression levels. Our findings deepen the understanding of the role of ferroptosis in AML and suggest the potential application of targeting ferroptosis for AML treatment.

AUTHOR CONTRIBUTIONS

Dan Wang: Writing; Methodology; Validation; Formal analysis; Resources. Fenglin Wang: Writing; Haixia Zhang: Data Curation; Visualization. Pan Chen: Conceptualization; Supervision; Investigation; Review and Editing. Minghua Yang: Writing; Review and Editing; Project administration; Funding acquisition.

FUNDING INFORMATION

This work was supported by grants from the National Natural Science Foundation of China (81974000, 82270185), the Natural Science Foundation of Hunan Province (2021JJ10077) and the Science and Technology Innovation Platform and Talent Program of Hunan Province (2022RC3077).

CONFLICT OF INTEREST STATEMENT

The authors declare that there are no conflicts of interest.

ETHICS STATEMENTS

Approval of the research protocol by an Institutional Reviewer Board: This study was approved by the Ethics Committee of the Third Xiangya Hospital of Central South University. All patients and donors provided written informed consent.

Informed Consent: Informed consent was obtained from study participants.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies. Animal experiments were approved by the Animal Care and Use Committee of Third Xiangya Hospital of Central South University.

Supporting information

Appendix S1.

Wang D, Wang F, Zhang H, Chen P, Yang M. Circadian clock protein Bmal1 accelerates acute myeloid leukemia by inhibiting ferroptosis through the EBF3/ALOX15 axis. Cancer Sci. 2023;114:3446‐3460. doi: 10.1111/cas.15875

DATA AVAILABILITY STATEMENT

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.