The leukemia cell lines HL60, K562 and Kasumi-1 (Kas-1) were purchased from Procell Life Science & Technology Co., Ltd. HL60 and Kas-1 cells were isolated from AML, whereas K562 cells were isolated from chronic myeloid leukemia. Cells were maintained in RPMI 1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.), 1% 100 U/ml penicillin G and 100 mg/ml streptomycin. Cells were cultured at 37°C with 5% CO₂. Cytarabine was purchased from MilliporeSigma. Drug-resistant leukemia cell lines were established as described previously (19). Briefly, 10⁵ cells were inoculated into a 6-well plate, and when cells reached 80% confluence, different concentrations (0, 1, 2, 5 and 10 μmol) of cytarabine were added. Subsequently, the medium was changed every 3-4 days for 3-4 weeks. The IC₅₀ concentrations of cytarabine in HL60, K562 and Kas-1 cells were 5 μmol, 100 nmol and 50 nmol, respectively.

SIRT1, HMGB1, ACSL4 and negative control (NC) siRNA sequences were obtained from Shanghai GenePharma Co., Ltd. siRNA transfection was performed using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Briefly, 10⁵ cytarabine-resistant HL60 (HL60/C) cells or HL60 cells were inoculated into a 6-well plate and were grown to 40-50% confluence. Subsequently, 50 nmol siRNA was diluted in Opti-MEM (Thermo Fisher Scientific, Inc.), mixed with 2 μl Lipofectamine 3000, added to the cell culture plate and cultured for 48 h at 37°C with 5% CO₂. The siRNA sequences were as follows: si-SIRT1, forward 5′-GUGGCAGAUUGUUAUUAAUTT-3′, reverse 5′-AUUAAUAACAAUCUGCCACTT-3′; si-SIRT1-2, forward 5′-UUCUGAAAUAUUCAAUAUCAA-3′, reverse 5′-GAUAUUGAAUAUUUCAGAAAA-3′; si-SIRT1-3, forward 5′-UUUUCCUUCCUUAUCUGACAA-3′, reverse 5′-GUCAGAUAAGGAAGGAAAACU-3′; si-HMGB1-1, forward 5′-GCUCAAGGAGAAUUUGUAATT-3′, reverse 5′-UUACAAAUUCUCCUUGAGCTT-3′; si-HMGB1-2,forward 5′-ACAAAAAAUGCAUAUGAUGAC-3′, reverse5′-CAUCAUAUGCAUUUUUUGUGC-3′; si-HMGB1-3, forward 5′-AGUUUCUUCGCAACAUCACCA-3′, reverse 5′-GUGAUGUUGCGAAGAAACUGG-3′; si-ACSL4-1, forward 5′-GAUGGAUGCUUACAGAUUA-3′, reverse 5′-UAAUCUGUAAGCAUCCAUC-3′; si-ACSL4-2, forward 5′-AUAUUGUUAUUAACAAGUGGA-3′, reverse 5′-CACUUGUUAAUAACAAUAUAC-3′; si-ACSL4-3,forward 5′-ACUGUAUAUUGUUAUUAACAA-3′, reverse 5′-GUUAAUAACAAUAUACAGUGC-3′; si-NC, forward 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse, 5′-ACGUGACACGUUCGGAGAATT-3′.

The possible protein interaction network of SIRT1 was constructed using the STRING database (https://string-db.org/).The SIRT1-HA and HMGB1-Flag overexpression vectors were constructed using pcDNA3.1 plasmids (Beyotime Institute of Biotechnology). Briefly, 1 μg SIRT1-HA and HMGB1-Flag vectors were diluted in Opti-MEM, mixed with 2 μl Lipofectamine 3000 and transfected into 293T (Procell Life Science & Technology Co., Ltd.), HL60 and HL60/C cells. After culturing at 37°C for 24 h, the cells were collected and added to 100 μl RIPA lysis buffer (Beyotime Institute of Biotechnology) on ice for 30 min, and the supernatant was collected by centrifugation at 10,000 x g for 10 min at 4°C. Subsequently, 5 μg anti-Flag (cat. no. AE063; ABclonal Biotech Co., Ltd.) and anti-HA (cat. no. AE105; ABclonal Biotech Co., Ltd.) antibodies were added to the protein A/G agarose beads (Beyotime Institute of Biotechnology) and incubated at 4°C overnight. Next, the 10 μl antibody-protein A/G agarose bead complex was added to the cell lysate and slowly shaken at 4°C for 2-4 h to conjugate the antibody to the protein A/G agarose beads. After the immunoprecipitation reaction, the agarose beads were centrifuged at 1,500 x g at 4°C for 3 min and the precipitation was collected. The agarose beads were washed with 1 ml RIPA lysis buffer (Beyotime Institute of Biotechnology) 3-4 times. Finally, SDS loading buffer was added for western blot analysis.

The Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc.) was used to determine cell viability. AML cells (5×10⁴/well) were plated in 96-well plates and treated with cytarabine (5 μmol) at 37°C for 24 h. Subsequently, 10 μl CCK-8 solution was added to each well and the cells were incubated at 37°C for 2-3 h. The absorbance was measured at 450 nm using a microplate reader.

Cells were transfected with si-NC or si-SIRT1 and were treated with cytarabine (5 μmol) at 37°C for 24 h. Cells (5×10⁴ cells/well) were then plated in 24-well plates, washed with PBS and incubated in serum-free medium containing 10 μmol/l EdU (Guangzhou RiboBio Co., Ltd.) for 2 h at 37°C. Cells were fixed in 4% paraformaldehyde (Beyotime Institute of Biotechnology) for 15 min at 4°C, after which, they were stained with Apollo solution and DNA staining solution (Beyotime Institute of Biotechnology) at room temperature. Images of the cells were then captured using a fluorescence microscope (Nikon Corporation).

AML cells (5×10⁴/well) were plated in 24-well plates and treated with cytarabine (5 μmol) at 37°C for 24 h. Flow cytometric analysis was subsequently performed using the Annexin V-FITC kit (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. Apoptotic cells were analyzed by flow cytometry (LSRFortessa; BD Biosciences) and FlowJo-V10 software (FlowJo, LLC) was used to process the experimental results.

The intracellular ROS production in AML cells was determined using the 2′,7′-dichlorofluorescein diacetate (DCFH-DA) probe. Briefly, cells were harvested and washed with serum-free medium, after which, they were incubated with 10 μM DCFH-DA diluted in serum-free medium at 37°C for 30 min. Subsequently, the cells were washed with serum-free medium to remove the unbound DCFH-DA probe, and fluorescence-labeled cells were analyzed by flow cytometry (LSRFortessa; BD Biosciences) and FlowJo-V10 software was used to process the experimental results.

After treatment, total RNA was isolated from AML cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc) and was reverse transcribed into cDNA using PrimeScript™ RT Master Mix (Takara Biotechnology Co., Ltd.) according to the manufacturer's instructions. qPCR was performed using the Fast SYBR Green Master Mix (Thermo Fisher Scientific, Inc.) on a 7900 Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). The qPCR conditions were as follows: 95°C for 10 min, followed by 35 cycles at 94°C for 30 sec, 60°C for 15 sec and 72°C for 30 sec. Relative mRNA expression was calculated using the 2^−ΔΔCq method (20). GAPDH was used as the internal control. The primer sequences were as follows: HMGB1, forward 5′-TAACTAAACATGGGCAAAGGAG-3′, reverse 5′-TAGCAGACATGGTCTTCCAC-3′; SIRT1, forward 5′-TAGCCTTGTCAGATAAGGAAGGA-3′, reverse 5′-ACAGCTTCACAGTCAACTTTGT-3′; GAPDH, forward 5′-GCACCGTCAAGGCTGAGAAC-3′ and reverse 5′-TGGTGAAGACGCCAGTGGA-3′.

AML cells were fixed in 4% formaldehyde at 4°C for 20 min followed by permeabilization with 0.1% Triton X-100 at room temperature for 10 min. The cells were then washed three times with PBS, blocked with 5% bovine serum albumin (Beyotime Institute of Biotechnology) for 30 min at room temperature, and incubated overnight at 4°C with anti-GPX4 (1:1,000; cat. no. A11243), anti-HMGB1 (1:1,000; cat. no. A2553) and anti-SIRT1 (1:1,000; cat. no. A11267) (all from ABclonal Biotech Co., Ltd.), followed by incubation with CoraLite 488-conjugated secondary antibody (1:200; cat no. SA00013-2; Wuhan Sanying Biotechnology). The nuclei were stained with DAPI at room temperature for 5 min. Cells were visualized using a confocal microscope (Olympus Corporation).

After treatment, AML cells were collected and lysed in RIPA lysis buffer (Beyotime Institute of Biotechnology). The lysates were incubated on ice for 20 min and centrifuged at 10,000 x g for 30 min at 4°C, before the protein concentration was determined using a BCA kit (Beyotime Institute of Biotechnology). Protein samples (20 μg) were separated by SDS-PAGE on 10% gels and were transferred onto polyvinylidene difluoride membranes (MilliporeSigma). The membranes were blocked with 5% non-fat dry milk diluted in Tris-buffered saline-0.5% Tween-20. Subsequently, the membranes were incubated with anti-SIRT1 (1:1,000; cat. no. A11267), anti-GPX4 (1:1,000, cat. no. A11243), anti-HMGB1 (1:1,000; cat. no. A2553),anti-ACSL4 (1:1,000; cat. no. A20414) and anti-GAPDH (1:5,000; cat. no. A19056) (all from ABclonal Biotech Co., Ltd.) overnight at 4°C. After three washes, the membranes were incubated with a HRP-conjugated secondary antibody (1:10,000; cat. no. AS014; ABclonal Biotech Co., Ltd.) for 60 min at room temperature. Protein bands were visualized using a chemiluminescence imaging system (Tanon-4600; Tanon Science and Technology Co., Ltd.) and ImageJ 1.8.0.345 software (National Institutes of Health) was used for gray value analysis.

AML cells were collected and lysed in RIPA lysis buffer (Beyotime Institute of Biotechnology). The MDA, GSH and SOD levels in cell lysates were measured using MDA (cat. no. ab287797; Abcam), GSH (cat. no. E-EL-0026; Wuhan Elabscience Biotechnology Co., Ltd.) and SOD assay kits (cat. no. ab316899;Abcam) according to the manufacturers' instructions.

A total of 15 nude mice (male; age, 7-8 weeks; weight, 20-22 g) were purchased from the Guangzhou Ruige Biological Technology Co., Ltd. The mice were raised under pathogen-free conditions (temperature: 20-26°C; humidity: 40-70%) with a 12-h light/dark cycle, and had free access to water and food.

For in vivo imaging, 9 nude mice were randomly divided into three groups: Control, si-NC and si-SIRT1 (n=3). Briefly, 2×10⁶ HL60/C cells transfected with si-NC or si-SIRT1 were resuspended in 100 μl Matrigel (cat. no. 356234; Corning, Inc.) and were injected into the dorsal flanks right of the midline of nude mice. Meanwhile, the same operation with 2×10⁶ HL60 cells was performed in the control group. The mice were observed every day. At day 7, mice were intraperitoneally injected with cytarabine (20 mg/kg, three times a week) for 2 weeks. The mice in all groups underwent a luciferase activity assay. Briefly, anesthesia was induced by 2-3% isoflurane inhalation and was maintained using 1.5-2% isoflurane. The mice were intraperitoneally injected with luciferin potassium salt (150 mg/kg; Shanghai Yeasen Biotechnology Co., Ltd.)and luciferase was detected after 30 min using a live animal imaging system (AniView100; Guangzhou Boluteng Biotechnology Co., Ltd.). Subsequently, these 9 mice were euthanized by cervical dislocation.

In addition, another xenograft assay was conducted. A total of 6 nude mice were randomly divided into two groups: si-NC and si-SIRT1 (n=3); 2×10⁶ HL60/C cells transfected with si-NC or si-SIRT1 were resuspended in 100 μl Matrigel and were injected into the dorsal flanks right of the midline of nude mice (n=3). At day 7, mice were intraperitoneally injected with cytarabine (20 mg/kg, three times a week) for 2 weeks. The tumor growth was observed continuously for 21 days. Prior to sacrifice, 0.2 ml of blood was collected from the orbit of the mice, the blood was left at room temperature for 2 h, centrifuged for 10,000 x g for 5 min at 4°C and serum was collected; subsequently, these 6 mice were euthanized by cervical dislocation to collect tumor tissue samples before reaching the humane endpoint: Tumor weight did not exceed 10% of body weight.

The animal experiments were performed by Guangzhou Seyotin Biotechnology Co., Ltd. and all experiments were approved by the Animal Ethics Committee of Guangzhou Seyotin Biotechnology Co., Ltd. (approval nos. SYT20203010 and SYT2024079).

AML cells were harvested and fixed in 2.5% glutaral-dehyde for 3 h at 4°C. After three washes with 0.1 M phosphate buffer, the cells were fixed with 1% OsO₄ at room temperature for 2 h. The cells were washed again with 0.1 M phosphate buffer, and were then dehydrated with graded acetone and embedded in Epon resin for 12 h at 37°C, for 12 h at 45°C and for 24 h at 60°C, then sectioned at 50-70 nm with an ultramicrotome. The sections were double stained with 3% uranyl acetate and lead citrate (ddH₂O, 30 ml; lead nitrate, 1.33 g; sodium citrate, 1.76 g) for 30 min at room temperature and observed with a transmission electron microscope (Hitachi, Ltd.).

After treatment, nuclear and cytoplasmic protein fractions were extracted from HL60/C cells using a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Institute of Biotechnology). The protein concentration was measured using a BCA kit (MilliporeSigma). Equal amounts of cytoplasmic proteins were loaded on gels for SDS-PAGE and were transferred to membranes. Western blot analysis was performed with anti-HMGB1,anti-GAPDH and anti-Histone (1:1,000; cat. no. AF0863; Affinity Biosciences).

The experiments were repeated three times and all data are presented as the mean ± SD. GraphPad Prism 7 (Dotmatics) was used to analyze and plot the data. Multigroup comparisons were performed using one-way ANOVA followed by the Tukey post hoc multiple comparisons test. Comparisons between two groups were performed using an unpaired Student's t-test. P<0.05 was considered to indicate a statistically significant difference.

To address the involvement of SIRT1 in cytarabine resistance, its expression levels in cytarabine-resistant leukemia cells (HL-60, K562 and Kas-1) were detected. As shown in Figs. 1A and S1, the relative mRNA expression levels of SIRT1, HMGB1and ACSL4 were significantly increased in cytarabine-resistant leukemia cells compared with those in the control group. Similarly, the protein expression levels of SIRT1 in cytarabine-resistant leukemia cells were markedly increased (Fig 1B). Taken together, these findings indicated that SIRT1 may be elevated in cytarabine-resistant leukemia cells.

To confirm the potential role of SIRT1 in cytarabine-resistant leukemia cells, HL60/C cells were transfected with si-SIRT1 sequences. The knockdown of SIRT1 was confirmed in HL60/C through RT-qPCR (Fig. 2A) and western blot analysis (Fig. 2B). Among the siRNAs, si-SIRT1-1 had the best knockdown effect and was therefore used in subsequent experiments. Furthermore, cell viability was significantly reduced in the si-SIRT1 group (Fig. 2C), and the IC₅₀ value was also significantly decreased (Fig. S2). In addition, transfection of si-SIRT1 into HL60 cells was detected by RT-qPCR (Fig. S3A), and knockdown of SIRT1 did not affect the viability of HL60 parental cells (Fig. S3B). The apoptosis rate was significantly increased in the si-SIRT1 group (Fig. 2D)l; however, the apoptosis rate of HL60 parental cells was the highest (Fig. S4). Regarding the rate of apoptosis, Q2 represents late apoptosis and Q3 represents early apoptosis.

To confirm the potential role of SIRT1 in ferroptosis, HL60/C cells were transfected with si-SIRT1 sequences. The results showed that SOD and glutathione (GSH) activities were decreased, whereas MDA levels were elevated in the si-SIRT1 group compared with those in the si-NC group (Fig. 3A-C). In addition, knockdown of SIRT1 significantly enhanced ROS levels compared with those in the si-NC group (Fig. 3D). It was also observed that knockdown of SIRT1 reduced the expression levels of GPX4 in HL60/C cells (Fig. 3E and G). Moreover, mitochondrial damage is closely associated with ferroptosis, and knockdown of SIRT1 elevated mitochondrial damage in HL60/C cells compared with that in the si-NC group (Fig. 3F). These findings suggested that SIRT1 may serve a critical role in ferroptosis in cytarabine-resistant leukemia cells.

The possible protein interaction network of SIRT1 was constructed using the STRING database (https://string-db.org/) and it was indicated that SIRT1 may regulate HMGB1 expression (Fig. 4A). Notably, through double immunofluorescence labelling, SIRT1 and HMGB1 were localized. The results showed that SIRT1 and HMGB1 were expressed in the nucleus, with obvious co-localization (Fig. 4B). Subsequently, the overexpression vectors SIRT1-HA and HMGB1-Flag were transfected into 293T, HL60 and HL60/C cells, and the mRNA expression levels of SIRT1 and HMGB1 were significantly upregulated (Figs. S5 and S6). Furthermore, the results of co-immunoprecipitation confirmed the interaction between SIRT1 and HMGB1 in 293T cells (Fig. 4C). In addition, the same results were detected in HL60 and HL60/C cells (Fig. 4D). Notably, it was also observed that si-SIRT1 increased the expression levels of HMGB1 in the cytoplasm and decreased the levels in the nucleus of HL60/C cells. These findings indicated that SIRT1 may interact with HMGB1 and inhibit its cytoplasmic translocation.

To confirm the potential role of HMGB1 in ferroptosis, HL60/C cells were transfected with si-HMGB1 sequences. The knockdown of HMGB1 was confirmed through RT-qPCR and western blot analysis, and si-HMGB1-1 has the best knockdown effect (Figs. 5A and S7). Subsequently, the results of the EdU proliferation assay showed that the viability of si-HMGB1-1-transfected cells was markedly higher than that in the si-NC group (Fig. 5B). The present study also examined the effects of HMGB1 knockdown on the expression of ferroptosis-related markers. The results showed that knockdown of HMGB1 elevated the activities of SOD, and the levels of GSH and GPX4, compared with those in the si-NC group (Fig. 5C, F and G). By contrast, knockdown of HMGB1 reversed the elevation of MDA and ROS in HL60/C cells (Fig. 5D and E). Taken together, these findings suggested that HMGB1 may induce the ferroptosis of HL60/C cells.

The present results suggested that, after knockdown of HMGB1, the relative mRNA and protein expression levels of ACSL4 were significantly reduced in HL60/C cells (Fig. 6A and B). To validate the relationship between SIRT1 and ASCL4, and the effect of this pathway on ferroptosis, HL60/C cells were transfected with si-SIRT1 and/or si-ACSL4. As shown in Fig. S8, si-ACSL4 significantly reduced the mRNA expression levels of ACSL4 compared with those in the si-NC group. Notably, it was observed that ROS and MDA levels were markedly increased in HL60/C cells in response to si-SIRT1; however, this effect was counteracted by knocking down ACSL4 (Fig. 6C and D). Similarly, knockdown of SIRT1 decreased SOD activity and relative expression levels of GPX4 in HL60/R cells, but both were elevated in the si-SIRT1 + si-ACSL4 group (Fig. 6E and F). Taken together, these findings suggested that knockdown of SIRT1 may induce the cytoplasmic translocation of HMGB1 and increase ferroptosis through the HMGB1/ACSL4 pathway in HL60/C cells.

To evaluate whether knockdown of SIRT1 affects the growth of HL60/C cells in vivo, in vivo imaging was used to observe the size of the tumor. In vivo imaging results showed that tumor growth was significantly inhibited in the si-SIRT1 group compared with that in the si-NC group (Fig. 7).

In order to further explore the mechanism, a further 6 nude mice were randomly divided into two groups: si-SIRT1 or si-NC (n=3); HL60/C cells transfected with si-SIRT1 or si-NC were subcutaneously injected into nude mice. At day 7, mice were intraperitoneally injected with cytarabine (20 mg/kg, three times a week) for 2 weeks. Tumor growth was measured every other day. The results showed that knockdown of SIRT1 significantly decreased the weight of tumors derived from HL60/C cells in mice (Fig. 8A). Moreover, serum and tumor tissue samples of 6 nude mice were collected, The si-SIRT1 group exhibited decreased SOD activity and increased the MDA content in serum compared with those in the si-NC group (Fig. 8B and C). Meanwhile, SIRT1 knockdown significantly inhibited GPX4 expression in tumor tissues (Fig. 8D). Furthermore, it was observed that the protein expression levels of ACSL4 in tumor tissues were increased in the si-SIRT1 group of mice compared with those in the si-NC group (Fig. 8E). In addition, compared with those in the si-NC group, the protein expression levels of HMGB1 were increased in the cytoplasm and were decreased in the nucleus of HL60/C cells after knockdown of SIRT1 in mice tumor tissues (Fig. 8F). These findings suggested that knockdown of SIRT1 may promote the translocation of HMGB1 from the nucleus to the cytoplasm.