2.1 Patient samples

Bone marrow samples from AML patients and healthy controls used in this study were obtained from the routine clinical management. Written informed consent was obtained from each patient using a consent form. All sample collection methods comply with the Helsinki Declaration principles. The procedure was approved by the Medical Ethics Committee of Shengjing Hospital of China Medical University (2024PS1360K).

2.2 Isolation of CD34 + cell

CD34 + cells were isolated from bone morrow using human bone marrow mononuclear cell isolation kit (#P8560, Solarbio, Beijing, China). In brief, samples were diluted with an equal volume of tissue diluent, and reagent A and reagent D form the kit were added. The sample were spread on top of separation solution and centrifuge at 800 g for 25 min. The monocytes were collected, added 10 mL of cell washing solution, centrifuge 250 g for 10 min, and repeat twice. The mononuclear cells were collected and resuspended cells in buffer. Then, cells were cultured with 100 µL of FcR Blocking Reagent and 100 µL of CD34 MicroBeads (#130-046-702, Miltenyi, Bergisch Gladbach, Germany) in 2–8℃ for 30 min. LS column was placed in the magnetic field, and cell suspension was applied onto the column, the unlabeled cells were discarded. The buffer was pipetted into the column, the labelled cells (CD34 + cell) were collected by firmly pushing the plunger into the column.

2.3 Cell culture

KG-1, HL-60, THP-1, Kasumi-1, and MOLM-13 were purchased from iCellBioscience (Shanghai, China). KG-1 and HL-60 were cultured in IMDM (#iCell-0008, iCellBioscience) culture medium containing 20 fetal bovine serum (FBS, #11011 − 8611, Tianhang Biotechnology, Hangzhou, China) at 37 °C in a 5% CO2 incubator. THP-1 and MOLM-13 cells were cultured at 37 °C in a 5% CO2 incubator using RPMI-1640 (#31800, Solarbio, Beijing, China) medium containing 10% FBS. Kasumi-1 cells were cultured at 37 °C in a 5% CO2 incubator in RPMI-1640 medium containing 20% FBS. According to the AGAP2 expression, we choose KG-1 and HL-60 for further study. Cells were seeded in 6-well plate and infected with indicated lentivirus (MOI = 30) for 96 h for further study. Following transfection, stable cell lines were selected and maintained using puromycin to ensure the retention of the transgene. For the AMPK pathway activation, cells were treated with 30 µM GSK621 (AMPK agonist, #G872690, Macklin, China) for 48 h. Then, the cells were harvested for the following assays.

2.4 Lentivirus production and infection

The lentiviral vectors were purchased from Hunan Fenghui Biotechnology (Changsha, China). shRNA sequences were synthesized by Generalbio (Chuzhou, China). AGAP2 and NR2F2 was cloned into a lentiviral vector in order in order to generate both pLVX-IRES-puro-AGAP2, pLVX-IRES-puro-NR2F2, pLVX-shAGAP2, and pLVX-shNR2F2. HEK293T cells were used to product lentiviral constructs by co-transfection with lentiviral vector, pSPAX2, and pMD2.G. Lentivirus were obtained after 48 h and 72 h transfection and filtered through 0.45 μm filters. The target mRNA sequence was as follow: ggtgctaatccgagaggaagc (AGAP2), gagagtgagcagtacaagaaa (AGAP2), cgttgtagctgcagtgagagc (AGAP2), gccgtatatggcaattcaata (NR2F2), cctcctcagtcatagagcaat (NR2F2), and ggagcgagctgtttgtgttga (NR2F2). The shRNA that targets non-specific sequence was ttctccgaacgtgtcacgt.

2.5 Colony formation assay

Infected cells were collected and resuspended in culture medium containing 30% FBS and 0.8% methylcellulose (#M112865, Aladdin, Shanghai, China), and seeded into 6 cm plate at a density of 300 cells. After 14 days, cells were stained with iodonitrotetrazolium (#I811737, Macklin, Shanghai, China) and take a picture.

2.6 Cell viability

Cell viability was determined using CCK-8 and Trypan blue. After infection for 96 h, cells were centrifuged to remove supernatant, and 1 mL of culture medium was added to resuspend the cells. Subsequently, 10 µL of cell suspension was added into 90 µL of PBS (including 10 µL of Trypan blue). After 2 min incubation at room temperature, 10 µL of mixture were loaded into a hemocytometer for cell viability and counting under a microscope (OLYMPUS, Tokyo, Japan).

Infected cells were collected and seeded into 96-well plate at a density of 6 × 103. Then, cells were cultured at 37 °C in a 5% CO₂ incubator, and tested at 0, 24, 48, and 72 h using the CCK-8 kit (#KGA9305, Keygen, Nanjing, China).

2.7 Cell apoptosis

Cell apoptosis was detected by flow cytometry using a cell apoptosis detection kit (#E-CK-A212, Elabscience, Wuhan, China). Cells were collected by centrifugation at 150 g for 5 min, and the supernatant was removed. After washing with PBS, 500 µL of Binding Buffer were added to resuspended cells, and 5 µL of AnnexinV-FITC and 5 µL of 7-AAD were added and mixed well. After incubation for 15 min at room temperature in the dark, flow cytometric analysis was performed on a NovoCyte Flow Cytometer (Agilent, CA, USA).

2.8 Animal model

To assay the effect of AGAP2 in tumor growth, subcutaneous xenograft model was used. Female NOD/SCID mice aged 8 weeks were housed in a controlled environment (12 h light/dark cycle, temperature of 22 ± 1 °C, humidity of 50 ± 5%, and free access to food and water). KG-1 cells and HL-60 cells (5 × 106) were injected subcutaneously into the right axilla of each mouse. Tumors were measured every 3 days after the tumor was visible to the naked eye. After 21 days, mice were sacrificed and tumor tissues were collected for further analysis.

To determination the effect of AGAP2 in AML development, tail vein injection leukemia model was used. Female NSG (NOD-SCID-Il2rg-/-) mice aged 8 weeks were exposed to 2.25 Gy whole-body radiation for 24 h. Next, 5 × 106 tumor cells were injected into the tail vein. During the experiment, the physical health was observed and recorded the surviving animals every week. At the end of the experiment, mice were sacrificed and bone marrow were collected from femur and tibia. All the procedures of animal experiments and reporting follow to the ARRIVE guidelines. This study was approved by the Medical Ethics Committee of Shengjing Hospital of China Medical University (2024PS1369K).

2.9 Giemsa

Ten µL of obtained cells were casted on a clean glass slide, then the slide was fixed in methanol for 15 min and allowed to air dry. The air-dried slides were stained with Giemsa solution (#G1015, Solarbio, Beijing, China) for 10 min, and slides were washed by distilled water rinse to remove excess staining solution. Subsequently, the slides were rinsed in 80% ethanol for a few seconds until clear. After rinsed with distilled water, the staining results were observed under a microscope (OLYMPUS, Tokyo, Japan).

2.10 Oil red O staining

The infected cells were collected, and oil red o fixative solution (#O0625, Sigma, USA) was added to fix the cell at room temperature for 20 min. Then, fixative solution was discarded. Wash the cells twice with PBS, centrifuged and discard the supernatant. The oil red o staining solution was added to stain the cell for 20 min, and wash the cells with PBS, the staining results were observed under a microscope (OLYMPU).

2.11 Filipin III

Cells were collected by centrifugation and washed twice for 5 min each with PBS, the cells were pelleted by centrifugation at 150 g at 5 min. Added 250 µM Filipin III solution (#GC12048, GlpBio Technology, CA, USA) to resuspend the cells and incubation for 30 min at room temperature at dark. Then, cells were washed with PBS and photographed (OLYMPUS).

2.12 Flow cytometry analysis

Following treatment, cells from each group were collected by centrifugation, after which the supernatant was discarded. Subsequently, 100 µL of staining buffer was added and mixed thoroughly, followed by the addition of 5 µL of CD11b antibody (#E-AB-F1081D, Elabscience, China). The cells were incubated with the antibody at 4 °C in the dark for 30 min. After incubation, 1 mL of staining buffer was added, and the cells were centrifuged at 150 g for 5 min. The supernatant was then removed, and an appropriate volume of buffer was added to resuspend the cells prior to flow cytometry analysis for CD11b expression.

2.13 Western blot

Total protein of cells was extracted using RIPA lysis buffer (#R0010, Solarbio) with 1% PMSF (#P0100, Solarbio). Proteins were denatured at a high temperature and separated by SDS-PAGE. The concentration of stacking gel was 5%, and the concentration of separation gel was 8% and 12%, respectively. Then, proteins were transferred to the PVDF membranes (#IPVH00010, Millipore, MA, USA), and blocked with blocking solution (#A1800, Solarbio). The membranes were incubated with primary antibodies (1:1000 dilution) against AGAP2 (#14690-1-AP, Proteintech, Wuhan, China), NR2F2 (#861976, Zen-bio, Chengdu, China), p-AMPKα (#AF3423, Affinity, Nanjing, China), AMPKα (#AF6423, Affinity), p-ACC(Affinity), ACC(#AF3421, Affinity), and β-actin (sc-47778, santa cruz) followed by incubation with HRP-labelled goat anti-rabbit IgG (#SE134, Solarbio)or HRP-labelled goat anti-mouse IgG (#SE131, Solarbio) (dilution 1:3,000) for 60 min at room temperature and visualized using an enhanced chemiluminescence kit (#PE0010, Solarbio).

2.14 Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted by TRIpure (#RP1001, Bioteke, Beijing, China). All-in-One First-Strand SuperMix (#MD80101, Medgene, Guangzhou, China) was used to reverse transcribe mRNA to cDNA. qRT-PCR was conducted on Exicycler 96 (BIONEER, Daejeon, Korea) with SYBR green (#SR4110, Solarbio) and 2×Taq PCR MasterMix (#PC1150, Solarbio). The primers were synthesized by General Biology Inc (Chuzhou, China). Primers used in this study were as followed: AGAP2 F: GCCAAGCAGGCTCTAAAC, R: GCGGCTCAAAGTCCATT. NR2F2 F: GGGTGGTCGCCTTTATG, R: TGAGGAGGAGACGGTGC.

2.15 Immunohistochemistry (IHC)

The tumor tissues were fixed and dehydrated. The fixed tissues were embedded and sectioned to 5 μm thickness. The sections were deparaffinized, rehydrated, and subjected to antigen retrieval. Subsequently, sections were incubated with 3% H2O2 for 15 min to block the endogenous peroxidase activity, and sections were then incubated with 1% BSA for 15 min. Following incubating with primary antibody against AGAP2 (#14690-1-AP, Proteintech), Ki67 (#AF01998, Affinity), and CD45 (#ab40763, Abcam, Cambridge, UK). Sections were incubated with HRP-labelled goat anti-rabbit IgG (#31460, ThermoFisher, Waltham, USA) for 1 h. The DAB kit (#DAB-1031, Maxim, Fuzhou, China) was used for color development. Sections were restrained with hematoxylin and photographed under a microscope (OLYMPUS).

2.16 Dual-luciferase assay

To determine whether NR2F2 transcriptionally regulates AGAP2, we performed a dual-luciferase reporter assay. We constructed a series of reporter plasmids by cloning sequential truncations of the human AGAP2 promoter into the pGL3-Enhancer vector: -1517/+15 (P1), -1034/+15 (P2), -620/+15 (P3), and − 451/+15 (P4). An NR2F2 overexpression plasmid (NR2F2) and its corresponding empty vector (Vector) were used as effectors. HEK293T cells were co-transfected using Lipofectamine 3000. For each promoter construct, cells were transfected with either the reporter plus Vector, or the reporter plus NR2F2. All transfections included the pGL3-Enhancer plasmid for normalization. At 48 h post-transfection, luciferase activity was measured using the Dual-Luciferase Reporter Assay System.

To define the core functional element, a site-directed mutagenesis assay was conducted. Based on truncation results indicating a critical region between − 451 and + 15, a putative NR2F2-binding site (-411 to -405) was identified. A mutant reporter (MUT) was generated by altering this site within the full-length promoter (-1517/+15), alongside the wild-type (WT) construct. HEK293T cells were transfected with the following groups: WT + Vector, WT + NR2F2, and MUT + NR2F2. Luciferase activity was measured as described above. The normalized firefly luciferase activity for each group was then calculated relative to the value of the WT + Vector control group.

2.17 Chromatin immunoprecipitation (ChIP) -qPCR

ChIP assay was conducted using the ChIP Assay Kit following the instructions (#P2078, Beyotime, Shanghai, China). In brief, use formaldehyde w to cross-link target protein and DNA, and fragment chromatin by sonication. Then, samples were centrifuged at 12,000 g for 5 min and the supernatants were collected. For subsequent analysis, 20 µl of supernatant was used as the input sample. Except the input sample, the other samples were subsequently incubated overnight with antibodies at 4℃. Protein A + G Agarose/Salmon Sperm DNA was added and incubated at 4 °C for 60 min. The target gene was amplified using a Exicycler 96 (BIONEER). Primers were synthesized by General Biology Inc, and sequences were as followed: CHIP AGAP2-1 F: CTTTCCCACTTCATCTTCC, R: CACCAGCACCGATGTCT. CHIP AGAP2-2 F: ACCCTGCTAAGTTCCAT, R: TCCCTTTCCCACTCACA. CHIP AGAP2-3 F: GGGGACTGCTGTGACTC, R: AATGGAAACCGTCTGCT.

2.18 DNA pull-down

DNA pull-down assay was performed according to the manufacturers’ instruction (#Bes5004, Bersinbio, Guangzhou, China). In brief, AGAP2 core promoter region was tagged with biotin, and the biotin-labeled promoter was bound with streptavidin magnetic beads. Proteins were extracted from KG-1 and HL-60 cells. The proteins were incubated with streptavidin magnetic bead mixture at 4℃ for 60 min. The protein-DNA-streptavidin-agarose complex was analyzed by western blot.

2.19 mRNA-sequencing and analysis

For the RNA-seq analysis, the experiments were performed in the AML cell line KG-1. Briefly, knockdown of AGAP2 (with a non‑targeting shRNA as negative control, NC) was carried out using lentiviral transduction. Cells were harvested 96 h post‑infection and subjected to mRNA sequencing. The sequencing service was provided by Novogene (Beijing, China). Raw FASTQ files were quality‑controlled and trimmed using Fastp (version 0.23.1). Differential expression analysis was conducted with DESeq2, and visualization was performed in R. Differentially expressed genes were defined as those with |log2FC| > 1 and p‑value < 0.05.

2.20 LC-MS analysis

100 µL of cell samples were added to a glass centrifuge tube with a Teflon-lined cap, followed by adding 0.75 mL of methanol, and vortex oscillation. Add 2.5 mL of pre-cooled methyl- tert -butyl ether and incubated for 1 h at room temperature. MS-graded water was added to stratify the organic phase (upper phase), and the organic phase was collected. The lower phase was re-extracted with 1 mL of mixture buffer (methyl tert -butyl ether/methanol/water (10:3:2.5, v/v/v), and collected the organic phase. The organic phase was concentrated using a nitrogen blowing device (Reacti-Therm, Theromo). Then, 100 uL of isopropyl alcohol was used to redissolve, and the samples were detected by LC-MS/MS system. The LC-MS system included a Vanquish UHPLC system (Thermo Fisher Scientific, USA) and an Orbitrap Q Exactive™ HF/Q Exactive™ HF-X mass spectrometer (Thermo Fisher Scientific). The Lipidsearch software was used to process original data extraction, lipid and peak identification, peak alignment, and quantitative analysis. The differential lipid was identified as VIP > 1, |Log2FC > 1|, p < 0.05.

2.21 Statistical analysis

GraphPad software was used in our statistical analysis. Data represents the mean ± SD. Differences were significant statistically at p < 0.05. qRT-PCR results were calculated using 2-ΔΔCt method. Two group comparisons using the t test, and one-way analysis of variance was used for multiple group comparisons. Survival analysis of animal was performed with Kaplan‑Meier survival estimation and compared groups using the Log‑rank (Mantel‑Cox) test. For the correlation analysis of AGAP2 expression with clinical features, Chi-square test or Fisher’s exact test, and Odds Ratios (ORs) were used.