A total of 30 patients with leukemia (15 men and 15 women; age range, 15–35 years) admitted to Jin'an District Hospital (Fuzhou, China) between May 1, 2020, and December 1, 2020, were enrolled in the current study. Leukemia patient blood samples (n=10; 4 men and 6 women) and DOX-resistant Leukemia patient blood samples (n=20; 11 men and 14 women) were collected according to institutional regulation of the Hospital Clinic Ethical Committee and according to the declaration of Helsinki. Informed written consent was given by all patients. The blood samples were separated to obtain peripheral blood mononuclear cells (PBMCs) using a cell isolator (LTS1077-1; Tianjin Haoyang Biological Manufacture Co., Ltd.) (30). Cells were subjected to western blot analyses or reverse transcription-quantitative (RT-q) PCR analyses.

Human leukemia chronic myelogenous leukemia K562 and acute promyelocytic leukemia HL-60 cells were cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; Hyclone; Cytiva). HEK293T (Human embryonic kidney) cells were cultured in DMEM medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Hyclone; Cytiva). All cells were purchased from The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences and were grown in a medium supplemented with 100 µg/ml streptomycin (Gibco; Thermo Fisher Scientific, Inc.) and 100 U/ml penicillin (Gibco; Thermo Fisher Scientific, Inc.). Cells were maintained in a humidified 37°C incubator under a 5% CO₂ atmosphere. Cells at 80% confluence were treated with 1 µg/ml DOX (cat. no. HY-15142A; MedChemExpress) for 48 h, 40 nM berberine (cat. no. HY-N0716; MedChemExpress) for 24 h, 40 nM metformin (cat. no. HY-15763; MedChemExpress), 40 nM artemisinin (cat. no. HY-B0094; MedChemExpress) for 24 h, 40 nM curcumin (cat. no. HY-N0005; MedChemExpress) for 24 h, 2 µM erastin (cat. no. HY-B0627; MedChemExpress) for 48 h and 40 nM triptolide (cat. no. HY-32735; MedChemExpress) for 24 or 48 h at 37°C.

Recombinant lentiviruses were amplified by transfecting HEK 293T cells at 75% confluence with 6 µg pMD2.G and 6 µg psPAX2 packaging plasmids (the 2nd generation lentiviral packaging plasmid; Addgene, Inc.) and 6 µg lentivirus-based Nrf2 expression plasmid (pLVX-puro; Addgene, Inc.) or 6 µg lentiviral-based short hairpin (sh)RNAs (pLKO.1-puro; Addgene, Inc.) specific for green fluorescent protein (GFP, CAAATCACAGAATCGTCGTAT; negative control) or Nrf2 (#1, GCTCCTACTGTGATGTGAAAT; 2#, GGAGTGTCAGTATGTTGAA) using Lipofectamine^® 2000 (cat. no. 11668; Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C. Viruses were collected at 72 h after transfection. K562 or HL-60 cells at 40% confluence were infected with a recombinant lentivirus in the presence of 10 µg/ml polybrene, followed by 12 h incubation at 37°C with 5% CO₂. After 36 h of infection, cells were treated with 2 µg/ml puromycin for 24 h. The viable cells were used to further experiments. The infection efficiency was ~85%.

Cells were collected, washed with cold PBS and resuspended in EBC250 lysis buffer (250 mM NaCl, 50 mM Tris pH 8.0, 0.5% Nonidet P-40, 50 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 2 µg/ml aprotinin and 2 µg/ml leupeptin) (31). The protein concentration was assessed using the BCA method. Equal amounts of total protein (20 µg) were loaded, separated by SDS-PAGE (4% stacking gel, 10% running gel), transferred to PVDF membranes (MilliporeSigma). The membrane was then blocked with 5% skimmed milk powder for 1 h at room temperature and hybridized to an appropriate primary antibody (4°C, overnight) and HRP-conjugated secondary antibody (anti-rabbit: 1:3,000; cat. no. AB0101; anti-mouse: 1:3,000, cat. no. AB0102; Shanghai Abways Biotechnology Co., Ltd.; room temperature, 1 h) for subsequent detection by enhanced chemiluminescence using an ECL kit (cat. no. P0018S; Beyotime Institute of Biotechnology). The images was analyzed using ImageJ software 1.8.0 (National Institutes of Health). Primary antibodies for GAPDH (cat. no. AF7021; 1:1,000), Nrf2 (cat. no. AF0639; 1:1,000), catalase (cat. no. DF7545; 1:1,000), superoxide dismutase (SOD)2 (cat. no. AF5144; 1:1,000), glutathione peroxidase (GPX)4 (cat. no. DF6701; 1:1,000) and were purchased from Affinity Biosciences. Antibody for Lamin B (cat. no. ab32535; 1:1,000) was purchased from Abcam.

Total RNA was extracted from 1×10⁶ cells using RNA easy Plus Mini kit (Qiagen) according to the manufacturer's protocol. RNA was reverse-transcribed into cDNAs using M-MLV First Strand kit (Invitrogen) according to the manufacturer's protocol. qPCR analyses of Nrf2 (forward: GGTTTCTTCGGCTACGTTT; reverse: ACTTCTTTTTCCATTGAGGGTATA), catalase (forward: CTCCGGAACAACAGCCTTCT; reverse: ATAGAATGCCCGCACCTGAG), SOD2 (forward: TAGCTCTTCAGCCTGCACTG; reverse: GCTTCCAGCAACTCCCCTTT), GPX4 (forward: TGGACGAGGGGAGGAGC; reverse: GGGACGCGCACATGGT) and GAPDH (forward: TCAAGAAGGTGGTGAAGCAGG; reverse: TCAAAGGTGGAGGAGTGGGT) were performed in CFX96 Real-Time PCR System (Bio-Rad Laboratories, Inc.) using SoFast EvaGreen Supermix (Bio-Rad Laboratories, Inc.), according to the manufacturer's protocol. The reactions were carried out in a 96-well plate at 95°C for 5 min, followed by 40 cycles of 95°C for 15 sec and 58°C for 30 sec. GAPDH expression was used as an inner control to normalize gene expression by the 2^−ΔΔCq method (31). All experiments were performed three times in triplicate.

Cell viability assay (CCK-8) was performed using a CCK-8 assay Kit (cat. no. CK04; Dojindo Molecular Technologies, Inc.) as described in the manufacturer's instruction (32). Briefly, 10 µl CCK-8 reagent was incubated with the cells at 37°C for 4 h, and absorbance was quantified at 450 nm using an ELx800 Absorbance Microplate Reader (BioTek Instruments Inc.).

K562 or HL-60 cells (2×10⁵) were seeded in 24-well plates in the absence or presence of triptolide. For the measurement of ROS levels, cells were washed and subjected to the procedures as described in the Reactive Oxygen Species Assay kit (cat. no. D6470, Beijing Solarbio Science & Technology Co., Ltd.) (33). Measurement of lipid oxidation was performed using BODIPY 581/591 C11 assay kit (cat. no. D3861; Invitrogen; Thermo Fisher Scientific, Inc.) as described in the manufacturer's instruction (33). Then the 2×10⁴ cells were analyzed using a flow cytometer (Beckman Coulter, Inc.) and the data were analyzed using FlowJo software V10 (FlowJo, LLC).

GraphPad Prism 6.0 (GraphPad Software Inc.) was used for data recording and calculation. All experiments were performed at least three times. Data were presented as means ± standard deviation. Quantitative data were analyzed statistically using unpaired Student's t-test for two groups (Figs. 1 and 2) and one-way ANOVA followed by Tukey's post-hoc test for >2 groups (Figs. 3 and 4) to assess the significance.

Chemotherapy is the first choice for leukemia treatment. DOX is a well-established chemotherapeutic drug for leukemia treatment. However, DOX resistance is a major clinical problem in leukemia treatment. Currently, the molecular mechanism(s) by which leukemia cells resistance to DOX remains to be elucidated. To explore this issue, DOX-resistant K562 cells (K562/DR) and DOX-resistant HL-60 cells (HL-60/DR) were first established upon a low dose of DOX treatment. As shown in Fig. 1A, compared with normal K562 or HL-60 cells, K562/DR or HL-60/DR cells exhibited strong resistance to DOX. Oxidative stress is an important hallmark of cancer cells (34). It is reported that DOX can induce the production of ROS to kill cancer cells (9). Next, the present study examined whether oxidative stress also served a role in leukemia cell resistance to DOX. As shown in Fig. 1B, K562/DR or HL-60/DR cells had higher Nrf2, a master regulator of the antioxidant response, protein expression than normal K562 or HL-60 cells, as evidenced by western blot analyses. It has been documented that Nrf2 is a transcription factor that regulates a series of genes expression involved in oxidative stress response, including catalase, SOD2 and GPX4. Consistent with Nrf2, K562/DR or HL-60/DR cells also exhibited higher catalase, SOD2 and GPX4 protein expression than normal K562 or HL-60 cells (Fig. 1B). In addition, qPCR analyses also showed that compared with normal K562 or HL-60 cells, K562/DR or HL-60/DR cells had higher Nrf2, catalase SOD2 and GPX4 mRNA levels (Fig. 1C and D; P<0.001). It is reported that Nrf2 can both localize in the cytoplasm and nucleus (35). However, only nuclear localization of Nrf2 can regulate antioxidant genes transcription. Therefore, the localization of Nrf2 was also examined in our system. As shown in Fig. 1E, compared with normal leukemia cells, nuclear localization of Nrf2 protein expression significantly increased in K562/DR or HL-60/DR cells. Notably, silencing of Nrf2 significantly sensitized K562/DR or HL-60/DR cells to DOX (Fig. 1F-H; P<0.05).

To further investigate whether the expression of Nrf2 and its downstream target genes are also upregulated in clinical DOX-resistant leukemia samples, leukemia patient blood samples (n=10) and DOX-resistant leukemia patient blood samples (n=20) were harvested. Levels of Nrf2, catalase, SOD2 and GPX4 in leukemia PBMC (n=10) and DOX-resistant leukemia PBMC (n=20) were then examined by western blot and quantitative PCR analyses. As shown in Fig. 1I-K, DOX-resistant leukemia PBMC exhibited significantly upregulated Nrf2, catalase, SOD2 and GPX4 protein and mRNA expression than normal leukemia PBMC (P<0.05).

Together, these results suggested that Nrf2 may play a role in leukemia cell resistance to DOX.

The aforementioned data indicated that Nrf2 served a critical role in leukemia cell resistance to DOX. It was thus hypothesized that inhibition of Nrf2 may be a potential strategy for overcoming DOX resistance in leukemia cells. To search for potential small chemical Nrf2 inhibitors, the effects of several known chemical compounds on Nrf2 expression were examined. As shown in Fig. 2A, triptolide, a clinical-approved chemical drug used to treat autoimmune disorders, remarkedly inhibited Nrf2 protein expression, concomitant with reduced GPX4 expression. The effects of triptolide on cellular ROS levels were also examined. As shown in Fig. 2B and C, triptolide markedly increased ROS levels in K562 and HL-60 cells, as evidenced by DCFH-DA analyses (P<0.001). Triptolide also significantly promoted lipid oxidation in K562 and HL-60 cells, as evidenced by BODIPY 581/591 C11 analyses (Fig. 2D and E; P<0.001). Since increased ROS and lipid oxidation and decreased GPX4 expression are hallmarks of ferroptosis, the effects of triptolide on leukemia cell viability were therefore examined. As shown in Fig. 2F and G, triptolide markedly inhibited K562 or HL-60 cell viability, indicating that triptolide induced leukemia cell ferroptosis. Together, these results demonstrated that triptolide can inhibit Nrf2 expression and induce leukemia cell ferroptosis.

Next, the present study investigated the role of Nrf2 in triptolide-induced leukemia cell ferroptosis. To examine this issue, K562 or HL-60 cells which stably express Nrf2 were first established. As shown in Fig. 3A, ectopic expression Nrf2 significantly increased GPX4 protein expression, consistent with the previous report (36). Furthermore, ectopic expression of Nrf2 restored GPX4 protein expression inhibited by triptolide (Fig. 3B). In addition, ectopic expression of Nrf2 also markedly reduced endogenous ROS levels and triptolide-induced ROS levels in K562 and HL-60 cells (Fig. 3C-E; P<0.001). Next, the role of Nrf2 on triptolide-induced leukemia cell lipid oxidation was examined. As shown in Fig. 3F-H, triptolide significantly induced lipid oxidation in K562 and HL-60 cells, which can be significantly rescued by ectopic expression of Nrf2 (P<0.001), suggesting that Nrf2 is critical in triptolide-induced ferroptosis. Indeed, it was found that ectopic expression of Nrf2 can inhibit triptolide-induced downregulation of K562 and HL-60 cell viability (Fig. 3I-J; P<0.001). Together, these results indicated that triptolide promotes leukemia cell ferroptosis via downregulation of Nrf2 expression.

The aforementioned data indicated that Nrf2 served a critical role in leukemia cell resistance to DOX (Fig. 1) and triptolide inhibits Nrf2 expression to induce leukemia cell ferroptosis (Figs. 2 and 3). Therefore, it was hypothesized that treatment with triptolide may overcome leukemia cell resistance to DOX. To examine this hypothesis, triptolide was used to treat K562/DR or HL-60 cells. As shown in Fig. 4A and B, triptolide significantly suppressed Nrf2 and GPX4 expression. Furthermore, DOX significantly inhibited normal K562 or HL-60 cell viability (P<0.001), but it had little effect on K562/DR or HL-60 cell viability (Fig. 4C and D). Notably, treatment with low-dose triptolide did not affect K562/DR or HL-60 cell viability, but it could re-sensitize K562/DR or HL-60 cells to DOX (Fig. 4C and D). In addition, consistent with triptolide, erastin, a ferroptosis inducer, could also re-sensitize K562/DR or HL-60 cells to DOX (Fig. 4E and F). Together, these results suggested that triptolide promotes ferroptosis via suppressing Nrf2 to sensitize leukemia cells to DOX.