ML385 promotes ferroptosis and radiotherapy sensitivity by inhibiting the NRF2-SLC7A11 pathway in esophageal squamous cell carcinoma

Abstract

Radiotherapy is important in treating esophageal squamous cell carcinoma (ESCC) comprehensively. Resistance to radiotherapy is a prominent factor contributing to treatment failure in patients with ESCC. The objective of this study was to investigate the impact of ML385, an inhibitor of nuclear factor erythroid 2-related factor 2 (NRF2), on the radiosensitivity of ESCC and elucidate its underlying mechanism. We treated KYSE150 and KYSE510 cells with ML385 and ionising radiation separately or simultaneously, and observed the proliferation, apoptosis, cell cycle and ferroptosis of different conditions by colony formation assay and flow cytometry. Our findings reveal that NRF2 was activated by radiation and translocated from the cytoplasm to the nucleus after radiation. However, ML385 inhibited the expression and cytoplasm-to-nucleus translocation of NRF2. Compared with radiation, ML385 combined with radiation exhibited a significant inhibition on the clone formation ability of ESCC cells, induced apoptosis and promoted G2/M phase arrest. The treatment of ML385 combined with radiation markedly increased ROS and lipid peroxidation levels and decreased glutathione levels compared with the control, thus promoting the occurrence of ferroptosis. In addition, the expression trend of NRF2 was the same as that of proteins related ferroptosis, such as SLC7A11 and GPX4. After overexpression of SLC7A11, we found that significantly restored glutathione levels and alleviated ML385 combined with radiation-induced lipid peroxidation, indicating that ML385 plays a key role in radiotherapy sensitization by inhibiting the NRF2-SLC7A11 pathway. In vivo, ML385 also promoted the killing effect of radiation on xenografted tumours in nude mice. This study identifies NRF2 inhibitor ML385 as a radiosensitizer of ESCC, which highlights the therapeutic potential of the NRF2-SLC7A11 pathway and provides a deeper understanding of the mechanism of ferroptosis in esophageal squamous cell carcinoma.

Background

Radiotherapy is important in treating esophageal squamous cell carcinoma (ESCC) comprehensively. Ionising radiation (IR) results in single- or double-strand DNA breaks damaging tumour cells through direct ionising destruction and indirect effects of reactive oxygen species (ROS) [1]. However, some damaged tumour cells can survive through DNA repair, and others make errors during repair that may alter the genome, resulting in increased tumour mutational burden and the development of strong radiation resistance [2]. Resistance is the main cause of radiotherapy-related failure in patients with ESCC [3]. Consequently, there is an urgent need to identify targets or drugs that can increase IR sensitivity.

Nuclear factor erythroid 2-related factor 2 (NRF2)—a primary regulator of antioxidant response [4]—contains six highly conserved NRF2-ECH homology (Neh) domains and is regulated by Kelch-like ESH-associated protein 1 (KEAP1). During stress-free conditions, NRF2 binds to KEAP1 and is primarily located in the cytoplasm. However, when oxidative stress occurs, KEAP1 undergoes a conformational change that leads to the release of NRF2. As a result, NRF2 translocates from the cytoplasm to the nucleus [5]. Next, the small nuclear MAF protein binds to NRF2 to form a heterodimer. The NRF2-MAF complex regulates the expression of downstream target genes and protects cells from oxidative and metabolic stress by combining antioxidant response promoter elements [6]. Abnormal signal transduction of the KEAP1-NRF2 pathway activates NRF2, and the antioxidant stress effect of NRF2 can greatly enhance the adaptability of tumour cells to oxidative stress—a mechanism of radiation resistance [7, 8]. Therefore, inhibiting the activity of NRF2 is of critical importance in suppressing the proliferation of tumour cells, and studies should focus on identifying an effective NRF2 inhibitor that can increase radiosensitivity.

Ferroptosis occurs in tumour cells and is regulated by NRF2 [9]. NRF2 exerts transcriptional regulation on numerous ferroptosis-related genes, including SLC7A11, GPX4, and phase II detoxification enzymes. Notably, the SLC7A11-glutathione (GSH) system may be a key cellular mechanism in the defence against ferroptosis [10]. During oxidative stress, intracellular iron ions react with unsaturated fatty acids to produce lipid peroxides. When the intracellular GSH antioxidant system fails to detoxify the cells, the oxidative stress introduced by radiotherapy leads to lipid peroxidation and ferroptosis [11, 12]. This suggests that NRF2-regulated ferroptosis is vital in tumour suppression.

ML385, a novel and specific NRF2 inhibitor, binds to the Neh1 domain to inhibit the nuclear translocation of NRF2, reducing the transcriptional activity of NRF2 and inhibiting the expression of NRF2 downstream target genes. Emerging research has provided compelling evidence supporting the inhibitory effects of ML385 on NRF2 expression and tumour cell proliferation, particularly in KEAP1-mutated non-small cell lung cancer (NSCLC). ML385 has demonstrated its potential in enhancing the antitumour activity in this specific context [13, 14]. However, the potential role of ML385 in tumour radiosensitization specifically in ESCC has not been extensively investigated or reported. The primary objective of this study was to explore the radiosensitization effect of ML385 on ESCC and the underlying mechanism by investigating the apoptosis, cell cycle, and ferroptosis of tumour cells under different treatments.

Materials and methods

Cell lines and reagents

The human ESCC cell lines, namely KYSE30, KYSE150, TE-1, KYSE450, KYSE410, and KYSE510, were acquired from the Key Laboratory of Shandong Cancer Hospital and Institute. To create an optimal growth environment, the cells were cultured in DMEM with a high glucose concentration, supplemented with 10% foetal bovine serum and 1% penicillin-streptomycin (Gibco Laboratories; Biosharp). The cells were then maintained at a temperature of 37 °C in a humidified incubator with a 5% CO₂ atmosphere.

ML385 (CAS: 846557-71-9, lot # 24435, purity: 99.55%) was obtained from MedChemExpress (New Jersey, NJ, USA). It was dissolved in DMSO purchased from Beijing Solarbio Science & Technology (Beijing, China) to prepare a stock solution with a concentration of 20 mM, it was stored at −80 °C for future use. To prepare ML385 for use, it was diluted in the complete cell culture medium to the desired concentrations as indicated. Primary antibodies against NRF2 (16396-1-AP), SLC7A11 (26864-1-AP), and Lamin B1 (12987-1-AP) were purchased from Protein Tech Group Inc. (Chicago, IL, USA). Additionally, the GPX4 antibody (ab125066) was procured from Abcam (Cambridge, MA, USA). Other antibodies against β-actin (#4970), anti-rabbit IgG (#7074), and anti-rabbit IgG (H + L), F(ab’)2 Fragment (#4413) were purchased from Cell Signaling Technology, Inc. (Beverly, MA, USA).

Cell viability and colony formation assay

The impact of ML385 on the viability of the KYSE150 and KYSE510 cell lines was evaluated using the Enhanced Cell Counting Kit-8 (CCK-8; Bioss, Beijing, China). In brief, cells were seeded in triplicate at a density of 2000 cells per well in 96-well plates and then incubated at 37 °C with 5% CO₂ for a period of 24 h. Next, the cells were exposed to ML385 (1, 2, 5, 10, and 20 µM) or DMSO for 48 h. Thereafter, a volume of 10 µL of CCK-8 solution was carefully added to each well of the plate and incubated at 37 °C for 2 h. Subsequently, the optical density (OD) was assessed at a wavelength of 450 nm using a microplate reader (SpectraMax; Molecular Devices, California, USA). To eliminate the influence of the medium, a blank background group consisting solely of DMEM was employed to subtract the OD value. Next, the inhibition rate of cells was determined using the following formula:

$$\text{Cell inhibition rate}\left(\%\right)=\left(\text{OD control group}-\text{OD treatment group}\right)/\left(\text{OD control group}-\text{OD blank}\right)\times \text{1}00$$

Regarding the colony formation assay, the attached cells (600 cells/well) were incubated with a fresh medium containing an appropriate amount of ML385, and the plates were irradiated in a cabinet irradiator (Rad Source Technologies, Georgia, USA) at 0, 2, 4, 6, 8 Gy after 24 h. The six-wells plates were carefully maintained and cultured for 7–14 days, allowing the colonies to grow and reach an optimal state. After discarding the medium and fixing the cells with methanol, the colonies were stained with 0.5% crystal violet for a duration of 10 min. Afterward, they were thoroughly washed with water to remove any excess staining. Lastly, the surviving colonies (over 50 cells) were counted. A triplicate of each experiment was performed. The cell survival curves were analyzed and fitted using the linear-quadratic model in GraphPad Prism 8 software (GraphPad Software, Inc., La Jolla, CA, USA).

Apoptosis and cell cycle

The Annexin V-FITC/propidium iodide (PI) Apoptosis Detection Kit from BD Biosciences was used to detect apoptosis. In a nutshell, the cultured cells were treated with ML385 or IR (8 Gy) alone or ML385 for 24 h followed by IR (8 Gy). After 24 h of IR treatment, cells were gathered and washed twice with phosphate-buffered saline (PBS) to remove any residual media or contaminants, then meticulously suspended in Annexin V binding buffer, incubated with 50 µL/mL FITC-Annexin V and 50 µL/mL PI for 15 min in the dark environment, and assessed using flow cytometer (FACSCalibur; BD Biosciences, New Jersey, USA) and the FlowJo V10 software.

To analyse cell cycle distribution, the treated cells were collected, rinsed twice, then fixed in 70% ethyl alcohol for a duration of 24 h. Subsequently, the cells were washed and resuspended in 0.5 mL PI/RNase staining buffer. After incubation at 23 °C for 15 min in the absence of light, the cells were analysed using a flow cytometer. Additionally, cell cycle profiles were analysed using the ModFit software.

Immunofluorescence

A total of 5.0 × 10⁴ cells were cultured in 12-well plates and subjected to ML385 treatment. After 24 h of IR, the cells were washed and fixed with 4% paraformaldehyde for a duration of 20 min. Subsequently, the washed cells were blocked with 5% bovine serum albumin for a period of 30 min. Furthermore, the cells were subjected to immunodetection, which involved incubation with polyclonal anti-Nrf2 antibody (1:300) overnight at 4 °C and FITC-conjugated secondary anti-rabbit IgG antibody (1:1000) for 1 h on the subsequent day. Lastly, the washed cells were incubated with 200 µL DAPI for 5 min to effectively stain the nuclei. Fluorescence images were acquired using a fluorescence microscope.

Quantitative polymerase chain reaction (qPCR)

The Total RNA Kit (TIANGEN, Beijing, China) was used for the isolation of total RNA. The PrimeScriptTM RT Master Mix (Takara Bio Inc., Shiga, Japan) was used to reverse transcribe 500 ng of total RNA into cDNA. Following amplification, the obtained cDNA underwent qPCR analysis using TB Green Premix Ex Taq II (Takara Bio Inc., Shiga, Japan).

The primer sequences for NRF2 and β-actin were as follows: NRF2 forward, 5′-TCCAGTCAGAAACCAGTGGAT-3′ and reverse 5′-GAATGTCTGCGCCAAAAGCTG-3′; β-actin forward, 5′-CGTGGACATCCGCAAAGAC-3′ and reverse 5′-CGTCATACTCCTGCTTGCTG-3′.

Western blotting

The washed cells were lysed in sodium dodecyl sulfate (SDS) lysis buffer supplemented with the protease inhibitor phenylmethylsulfonyl fluoride (Beyotime Biotechnology, Nanjing, China). Next, a cytoplasmic and nuclear extraction kit was employed to isolate nuclear proteins. The adjusted concentration of protein extracts were loaded onto SDS-polyacrylamide gel electrophoresis membranes and transferred to polyvinylidene fluoride (PVDF) membranes. Subsequently, the PVDF membranes were incubated with primary antibody overnight at 4 °C after blocking with 5% non-fat milk powder in TBS-Tween 20 (TBST) for 1 h, washed thrice with TBST, and reincubated with horseradish peroxidase-conjugated IgG for 1 h at ambient temperature. Lastly, the bands were visualised using an ECL detection reagent (Millipore, USA).

ROS and lipid peroxidation

The cultured cells were treated with ML385 or IR (8 Gy) alone or ML385 for 24 h followed by IR (8 Gy). Next, 10 µM CM-H2DCFDA (ThermoFisher, C6827, USA) or 5 µM BODIPY 581/591 C11 dye (Invitrogen, D3861, USA) was added to the dishes containing fresh medium to detect total ROS or lipid peroxidation levels, respectively. After 30 min incubation, the washed cells were analysed via flow cytometry.

GSH assay

The cells adhering to the 96-well plate were treated with ML385, IR, or ML385 + IR separately. The treated cells were cultured in fresh medium containing 100 µL of prepared 1× GSH-Glo Reagent (Promega) at room temperature for a period of 30 min, and then 100 µL of reconstituted Luciferin Detection Reagent (Promega) was added to each well for another 15 min. Lastly, luminescence was measured and normalised to the cell viability. The relative GSH levels in cells treated with the test compounds were normalised to those in control cells.

Animal experiments

The athymic nude mice used in this study were obtained from Charles River (Beijing, China), and 2 × 10⁶ KYSE150 cells were injected subcutaneously into 6-week-old specific pathogen-free BALB/c nude mice. The tumours reached a size of approximately 50 mm³, the mice were randomly divided into four groups and intraperitoneally treated with saline or ML385 (30 mg/kg) for two consecutive days. After 24 h, the mice were anaesthetised by intraperitoneal injection of 0.1 mL 1% sodium pentobarbital. After the mice became stationary, their heads and limbs were fixed with a special mouse position fixator, and the tumours were exposed to the centre of the irradiation field, shielding the other parts with lead blocks. Irradiation was performed using an animal cell irradiator (model: Rad Source RS2000Pro, parameter setting: 225.0 KV, 17.7 mA, dose rate: 2.4 Gy/min, and total dose: 8 Gy). The tumour volume was calculated as follows:

$$\text{Volume}=\text{length}\times {\text{width}}^{\text{2}}\times \text{1}/\text{2}$$

Statistical analysis

Statistical analysis was conducted using GraphPad Prism 8. Each experiment was repeated at least three times. The results were presented as the mean ± standard deviation (SD), and error bars indicate the standard deviation. The statistical significance of the results was determined using the analysis of variance(ANOVA). Differences between groups were considered statistically significant when the p-value was less than 0.05 (p < 0.05).

Results

ML385 inhibits NRF2 activation after radiation exposure

NRF2 expression was detected using western blotting. The NRF2 expression level in KYSE30 was lowest among the six ESCC cell lines and highest in KYSE150 and KYSE510 (Fig. 1A). Therefore, KYSE150 and KYSE510 cells were selected for subsequent experiments. We detected NRF2 expression after concentration gradient treatment with ML385.

Fig. 1.

ML385 inhibits NRF2 activation after radiation exposure. A Screening cell lines from six ESCC cell lines. B ML385 inhibition was determined under the same condition for the CCK-8 assay in KYSE150 and KYSE510 cells. C Effect of different concentrations of ML385 on NRF2 expression in KYSE150 and KYSE510 cells (both cell lines were pretreated with ML385 for 48 h prior to the experiment). D NRF2 expression at different time points in KYSE150 (pretreated with 5 µM ML385) and KYSE510 cells (pretreated with 10 µM ML385). E, F The protein and mRNA expression of NRF2 in KYSE150 and KYSE510 cells under different treatments (control group: no treatment; ML385 group: treated with 5 or 10 µM ML385 for 48 h; IR group: 8 Gy radiation treatment; ML385 + IR group: pretreated with 5 or 10 µM ML385 for 24 h, followed by radiation with 8 Gy and further cultured for an additional 24 h) was detected using western blotting or qPCR. G Immunofluorescence was employed to assess the expression and localization of NRF2 in KYSE150 cells under different treatments (the treatment is the same as E–F). ESCC: esophageal squamous cell carcinoma; KYSE30: k30, KYSE150: k150, KYSE450: k450, KYSE410: k410, KYSE510: k510; IR: ionising radiation; n-NRF2: nuclear-NRF2. The statistical analysis utilized the analysis of variance, *p < 0.05, **p < 0.01, ***p < 0.001, ^#p < 0.0001, ns: no statistical significance. Bar graphs: mean ± SD, n = 3

Moreover, a proliferation assay was performed to determine ML385’s effects on KYSE150 and KYSE510. ML385 reduced cell proliferation; this effect was dose-dependent in KYSE510 (Fig. 1B). Additionally, the optimal treatment concentration was 5 µM for KYSE150 and 10 µM for KYSE510 (Fig. 1C). Furthermore, we tested NRF2 expression levels across time after treatment with 5 or 10 µM ML385. The NRF2 expression level was the lowest 48 h after treatment in KYSE150 and KYSE510 (Fig. 1D). Therefore, the ML385 treatment time in the subsequent experiments was 48 h.

Treatment of cells with ML385, with or without IR, yielded similar results for NRF2 expression, as determined using western blotting and qPCR. IR increased NRF2 nuclear transfer, which was inhibited by ML385. Additionally, the ML385 + IR group exhibited a significant decrease in NRF2 expression compared to the IR group (Fig. 1E, F).

Lastly, using immunofluorescence, we determined the expression and localisation of NRF2 (red) in KYSE150 treated with ML385 with or without IR; nuclear localisation was detected using DAPI (blue). The NRF2 fluorescence intensity increased after IR compared with the control, and the nuclear accumulation of NRF2 was enhanced. In contrast, the total and nucleus-localised NRF2 fluorescence intensity of the ML385 treatment group was weaker than that of the control, and the ML385 + IR group showed reduced nuclear accumulation of NRF2 (Fig. 1G).

These results suggest that IR effectively increased NRF2 expression at the mRNA and protein levels, whereas ML385 had the opposite effect. Additionally, ML385 significantly inhibited the nuclear translocation and accumulation of NRF2 after IR.

ML385 promotes radiosensitivity in vitro

In colony formation assays, the cells were treated with ML385, either alone or in combination with IR. Results show that the number of viable cells and proliferation rate decreased significantly with increasing IR doses. The mean lethal dose (D₀) value was significantly reduced from 3.42 Gy to 2.67 Gy in KYSE150 cells, while the radiosensitizer enhancement ratio (SER) notably climbed to 1.28. Similarly, in KYSE510 cells, the D₀ value underwent a marked decrease, shifting from 4.03 Gy to 3.26 Gy, accompanied by a SER increase of 1.24. Comparatively, the ML385 + IR group exhibited a noteworthy reduction in the surviving fraction of cells when compared to the IR group, suggesting that ML385 can promote sensitivity to IR (Fig. 2A).

Fig. 2.

ML385 promotes radiosensitivity in vitro. A Effect of ML385 on clone formation ability of KYSE150 and KYSE510 cells after radiation at different doses (0, 2, 4, 6, and 8 Gy). B, C Flow cytometry was employed to assess the impact of ML385 on apoptosis and cell cycle after radiation for KYSE150 and KYSE510. The treatment of B–C is the same as E–F in Fig. 1. KYSE150: k150; KYSE510: k510; IR: ionising radiation. The statistical analysis utilized the analysis of variance, *p < 0.05, **p < 0.01, ***p < 0.001, ^#p < 0.0001, ns: no statistical significance. Bar graphs: mean ± SD, n = 3

Annexin V-FITC/PI and PI were used to detect apoptosis and cell cycle distribution, respectively. The ML385 + IR group demonstrated a significantly higher percentage of apoptosis compared to the IR group (KYSE150, p = 0.0006; KYSE510, p = 0.0015), and late-stage apoptosis was predominant (Fig. 2B). These results suggest that ML385 promotes IR-induced apoptosis. Moreover, in comparison to the IR group, the ML385 + IR group displayed a significant enhancement in the decrease of the G1 phase and arrest of the G2/M phase (Fig. 2C). Notably, in radiobiology, tumour cells exhibit higher sensitivity to IR during the G2/M phase, while they tend to be relatively resistant to IR during the S phase. Consequently, radiosensitization induced by ML385 may be caused by the promotion of apoptosis and G2/M phase arrest after ML385 treatment.

ML385 promotes radiosensitivity in vivo

Nude mice were subcutaneously injected with tumour cells. After 12 days, the volume of the transplanted tumour reached 50 mm³, and treatment with ML385, with or without IR, was performed (Fig. 3A). Within 4 days of irradiation, we observed no significant difference in tumour volume among the different treatments; however, the differences gradually increased after 4 days of irradiation (Fig. 3B). Additionally, the weight of the nude mice in each group was the lowest on the sixth day after radiation and increased afterwards (Fig. 3C). Moreover, 14 days after irradiation, the tumour volume obvious differences between the groups. Next, the tumour volume in the control group was the largest, and no obvious difference was observed between the ML385 and IR groups. Lastly, the tumours grew slowly in the ML385 + IR group, which showed a significantly smaller tumour volume than that in the IR group (Fig. 3D). This indicates that ML385 enhanced the radiation effect on xenograft tumours in nude mice and inhibited these tumours. Moreover, weight loss indicates that radiation and drugs adversely affect nude mice. Figure 3E shows the images of nude mice after sacrifice.

Fig. 3.

ML385 promotes radiosensitivity in vivo. A Time points of subcutaneous tumour growth, treatment, monitoring, and death in nude mice. B, C The volume and weight of xenografts tumours (KYSE150) treated with ML385 and/or IR at various time points. D Representative images of KYSE150 xenograft tumours with different treatment at the end of experiment. E Images of the nude mice after sacrifice. NS: normal saline, IR: ionising radiation. The statistical analysis utilized the analysis of variance, *p < 0.05, **p < 0.01, ***p < 0.001, #p < 0.0001, ns: no statistical significance. Bar graphs: mean ± SD, n = 5

ML385 promotes ferroptosis and radiosensitivity

To assess the levels of total ROS and lipid peroxidation in cells, we employed DCFH-DA and BOD-IPY™581/591C11 staining. ROS and lipid peroxidation were induced by either ML385 or IR and were significantly induced in ESCC cells by ML385 + IR treatment (Fig. 4A, B). Lipid peroxidation is a biomarker of ferroptosis [15], suggesting that the ML385 + IR induced more ferroptosis than ML385 or IR alone.

Fig. 4.

ML385 promotes ferroptosis and radiosensitivity. A Effects of ML385 on reactive oxygen species in KYSE150 and KYSE510 cells following radiation. B Effects of ML385 on lipid peroxidation in KYSE150 and KYSE510 cells after radiation. C Effects of ML385 on glutathione (GSH) in KYSE150 and KYSE510 cells following radiation. D Changes in ferroptosis-related protein levels after ML385 and/or IR treatment in KYSE150 and KYSE510 cells. The treatment in A–D is the same as that in E–F in Fig. 1. KYSE150: k150; KYSE510: k510; IR: ionising radiation. The statistical analysis utilized the analysis of variance, *p < 0.05, **p < 0.01, ***p < 0.001, #p < 0.0001, ns: no statistical significance. Bar graphs: mean ± SD, n = 3

To determine the mechanism underlying ML385’s role in ferroptosis, we detected glutathione content under ML385 treatments, with or without IR, sequentially. The GSH content decreased in ML385 and IR cells, and the content in the ML385 + IR group was the lowest (Fig. 4C). Notably, ROS levels increased with GSH consumption, and the oxygen radicals generated attack polyunsaturated fatty acids, triggering the accumulation of lipid peroxidation within cells, it has been implicated in the induction of ferroptosis in tumour cells. [16]. Ferroptosis-related pathways are tightly regulated by key players such as NRF2, SLC7A11, and GPX4. Among these regulators, SLC7A11 is vital in the metabolism of ferroptosis [17]. In this study, IR induced the expression of NRF2, while ML385 inhibited it. Similarly, the protein expression levels of SLC7A11 and GPX4 also increased after IR, and decreased after ML385. It has been reported that immunohistochemical examination of esophageal cancer tissues has uncovered a positive correlation between the nuclear expression of NRF2 and SLC7A11. Furthermore, the overexpression of NRF2 was found to significantly enhance the expression of SLC7A11 in ESCC cell lines [18]. The results of this study are similar to ours, indicating that SLC7A11 and GPX4 are indeed regulated by NRF2 (Fig. 4D).

ML385 promotes ferroptosis and radiosensitivity through the NRF2-SLC7A11 pathway

To validate IR- and ML385 + IR-induced ferroptosis, we administered ferrostatin-1, a potent inhibitor of ferroptosis, to the cells, and detected the clonality of cells using a plate colony formation experiment. Ferrostatin-1 inhibited ferroptosis in the IR (2 Gy, p = 0.0335; 4 Gy, p = 0.0812) and ML385 + IR groups (2 Gy, p = 0.0002; 4 Gy, p = 0.0005), suggesting that IR induced a small amount of ferroptosis, and ML385 + IR substantially promoted radiation-induced ferroptosis (Fig. 5A). Furthermore, to validate the ML385 + IR-induced ferroptosis, radiosensitization was induced by inhibiting the NRF2-SLC7A11 pathway. We treated KYSE150 with ML385 + IR 48 h after SLC7A11 overexpression (Fig. 5B). In contrast to the control group, ML385 + IR treatment resulted in a decrease in GSH levels and an increase in lipid peroxidation. However, the overexpression of SLC7A11 effectively restored GSH levels and significantly alleviated lipid peroxidation(Figs. 5C, D).

Fig. 5.

ML385 promotes ferroptosis and radiosensitivity through the NRF2-SLC7A11 pathway. A Representative images of clonogenic survival assays in KYSE150 cells pretreated with 5 µM ML385 and/or 5 µM ferrostatin-1 for 24 h, followed by IR (2–4 Gy), respectively. The bar graph illustrates the relative fold changes in survival fraction levels for four different groups. B SLC7A11 plasmid was detected using western blotting after transfection for KYSE150 cells. C Relative levels of GSH in EV or SLC7A11 overexpressing cells (KYSE150) after treatment with DMSO or 5 µM ML385 + 8 Gy IR, respectively. D The levels of lipid peroxidation in EV or SLC7A11 overexpressing cells (KYSE150) treated with DMSO or 5 µM ML385 + 8 Gy of IR, respectively. The bar graph depicts the relative fold changes of lipid peroxidation levels per 10,000 cells induced by ML385 + IR in the indicated cells, as measured using C11-BODIPY staining. KYSE150: k150; KYSE510: k510; IR: ionising radiation; Fer-1: ferrostatin-1; EV: empty vector; SLC7A11 Over: SLC7A11 overexpression. The statistical analysis utilized the analysis of variance, *p < 0.05, **p < 0.01, ***p < 0.001, #p < 0.0001, ns: no statistical significance. Bar graphs: mean ± SD, n = 3

In summary, ML385 treatment led to a decrease in the translocation of NRF2 into the nucleus, which subsequently resulted in a reduction in the expression of SLC7A11, a downstream target gene of NRF2, thereby decreasing GSH synthesis. Due to the decrease in GSH levels, the clearance of lipid peroxides was compromised, leading to their accumulation, and the accumulated lipid peroxides resulted in ferroptosis. Our results show that SLC7A11 is vital in ML385-induced ferroptosis and radiosensitization.

Discussion

Current studies found that radiotherapy resistance of ESCC involves the dysregulation of multiple signalling pathways, containing Wnt/β-Catenin, NF-κB, JAK2/STAT3, and VEGF pathways [19–22], prompting research focus on designing radiosensitizer for different targets. Nitroimidazoles, epigenetic modifiers, and nanomaterials are the most widely studied radiosensitizer. The mechanisms of action of radiosensitizer include inhibiting free radical scavenging, promoting apoptosis, regulating the cell cycle, inhibiting damaged DNA repair, and improving the hypoxic microenvironment [23–25]. Additionally, studies reported that immunological agents combined with radiotherapy and gold nanoparticle technology could improve the sensitivity to radiotherapy by promoting ferroptosis or strengthening the free radical effect produced by radiation [26, 27]. NRF2 was proved to be a primary regulatory factor of oxidative stress related to radiosensitization [6]. Radiation overactivates NRF2 in tumour cells and overenhances the ability to remove ROS, greatly improving the adaptability of tumour cells and resulting in radiation resistance [28]. Additionally, inhibition of NRF2-dependent antioxidant defence systems reportedly improves chemosensitivity and radiosensitivity [29]. Therefore, new drugs targeting NRF2 may be effective strategies to increase the radiosensitivity of ESCC cells, especially for cancers with elevated NRF2 levels. Presently, some inhibitors targeting NRF2, such as brusatol [30], trigonelline [31], and triptolide [32], have been shown to be effective against malignant tumours. However, more evidence is needed to support NRF2 as a radiosensitization target. With the in-depth study on ROS in the mechanism of radiotherapy resistance, an increasing number of small-molecule compounds involved in oxidative stress have been used as radiosensitizer for esophageal cancer [24]. In this study, we focused on ML385, a novel small-molecule drug targeting NRF2, to verify its radiosensitising effect on ESCC and explore its underlying mechanism.

We verified that radiation increased NRF2 expression both at the mRNA and protein levels and that radiation-induced translocation of NRF2 from the cytoplasm to the nucleus. However, ML385 inhibited NRF2 expression and prevented the cytoplasmic-nuclear translocation of NRF2. Furthermore, we investigated the potential inhibitory effects of ML385 on the clonality of ESCC cells following radiation treatment, demonstrating that ML385 + IR significantly inhibited the clonality of ESCC cells compared with IR. This confirms that ML385 can be used as a radiosensitizer. In addition, brusatol (an NRF2 inhibitor) and other small-molecule drugs, such as berberine and LGK-974, could enhance the radiosensitivity of different cancer cell lines by inhibiting the NRF2 signalling pathway, as well as promoting ROS production and DNA damage [30, 33].

To explore the radiosensitization mechanism of ML385, we detected cell apoptosis and cell cycle distribution and found that ML385 + IR treatment triggered more apoptosis and promoted G2/M phase arrest compared with ML385 or IR alone. Additionally, the ROS and lipid peroxidation levels were measured using a fluorescent probe, demonstrating that the ML385 + IR group had higher ROS and lipid peroxidation levels than the IR group. Lastly, we observed the effect of ML385 on xenograft tumours in nude mice. ML385 enhanced the efficacy of radiotherapy in killing xenograft tumours and effectively inhibited tumour growth. Furthermore, we speculated that inhibiting NRF2 would result in the accumulation of ROS and lipid peroxidation, which would ultimately lead to cell death. Three classical ferroptosis regulatory pathways exist: lipid peroxidation, amino acid metabolism, and iron metabolism [34, 35]. In this study, ROS and lipid peroxidation levels were closely related to the lipid peroxidation pathway. Compared with IR, ML385 + IR increased intracellular ROS and lipid peroxidation levels, promoting ferroptosis and achieving radiosensitization in ESCC. Moreover, we investigated the expression patterns of ferroptosis-related proteins and observed that the protein levels of SLC7A11 and GPX4 increased after IR but decreased after ML385 + IR, consistent with NRF2, indicating that SLC7A11 and GPX4 are indeed regulated by NRF2.

The cystine/glutamate antiporter, also known as System Xc-, plays a crucial role in the regulation of extracellular glutamate concentration, transferring one cystine residue to the intracellular space and one glutamate residue to the extracellular space. It mainly comprises SLC7A11 and SLC3A2; notably, SLC7A11 is vital in the metabolism pathway of ferroptosis. SLC7A11 inhibition depletes cystine levels and reduces GSH synthesis, disrupting intracellular ROS homeostasis and accumulation of lipid peroxidation, which leads to ferroptosis [36, 37]. Since the SLC7A11-GSH system plays a crucial role in cellular defense against ferroptosis and can be regulated by NRF2 [38], our study mainly focused on the relationship between ML385, a specific NRF2 inhibitor, and SLC7A11. We found that ML385 reduced GSH levels, which were much lower with ML385 + IR treatment, indicating that ML385 reduced GSH synthesis by inhibiting SLC7A11 expression.

Furthermore, we overexpressed SLC7A11 in KYSE150 and treated them with ML385 + IR for 48 h to further verify that ML385 increased the radiosensitivity of ESCC cells by inhibiting the NRF2-SLC7A11 pathway. While the control group experienced a reduction in GSH levels and an increase in lipid peroxidation due to ML385 + IR treatment, SLC7A11 overexpression significantly restored these levels and alleviated lipid peroxidation induced by ML385 + IR treatment. This indicates that ML385 promotes ferroptosis by inhibiting the NRF2-SLC7A11 pathway, enhancing radiosensitivity. Wohlhieter et al. [39] found that mutated KEAP1 overactivated NRF2 and increased the activity of downstream target proteins involved in ferroptosis resistance, for instance, SLC7A11 and GPX4. Additionally, Fan et al. [40] found that NRF2 overexpression or Keap1 knockout increased the SLC7A11 expression; however, NRF2 inhibition or Keap1 overexpression reduced SLC7A11 expression. This implies that the activation of the NRF2-SLC7A11 signaling pathway could cause ferroptosis resistance, and our results proved that targeted inhibition of the NRF2-SLC7A11 pathway induces ferroptosis and increases radiosensitivity.

However, the limitation of this study is that we were unable to observe the correlation between NRF2 and SLC7S11 in human tissue sections. In the future, we will conduct more analysis through staining of human tissue sections and relevant data in the database.

Conclusion

The NRF2-specific inhibitor ML385 is a novel radiosensitising agent. In this study, radiotherapy was observed to facilitate the transfer of NRF2 from the cytoplasm to the nucleus. However, ML385 inhibited NRF2 expression and prevented the cytoplasmic-nuclear translocation of NRF2. Furthermore, ML385 promoted apoptosis and G2/M phase arrest, reduced GSH levels, and facilitated the production of ROS and lipid peroxidation of tumour cells after IR by inhibiting the NRF2-SLC7A11 pathway, thereby inducing ferroptosis and increasing radiosensitivity (Fig. 6). As expected, ML385 and IR demonstrated potent anti-proliferative effects against ESCC cells in vitro and in vivo. Despite these promising findings, caution should be exercised when considering ML385 as a radiosensitizer due to its potential toxicity, which may hinder its effective use. Therefore, to address the potential toxicity of ML385 and make it a more suitable radiosensitizer for treatment, further research on its derivatives is necessary.

Fig. 6.

Mechanism of NRF2 inhibition by ML385 in regulating radiotherapy sensitivity. ML385 can induce apoptosis and G2/M phase arrest to restrain tumour, reduce GSH levels, facilitate the accumulation of reactive oxygen species and lipid peroxidation after ionising radiation (IR) by inhibiting SLC7A11 expression, ultimately promote ferroptosis. Therefore, the synergistic effect of ML385 and IR on ferroptosis can lead to potent tumour suppression in esophageal squamous cell carcinoma.

Abbreviations

NRF2

Nuclear factor erythroid 2-related factor 2

ESCC

Esophageal squamous cell carcinoma

PI

Propidium iodide

ROS

Reactive oxygen species

IR

Ionising radiation

Neh

NRF2-ECH homology

KEAP1

Kelch-like ESH-associated protein 1

GSH

Glutathione

DMEM

Dulbecco’s modified eagle medium

CCK-8

Cell counting kit-8

OD

Optical density

PBS

Phosphate-buffered saline

qPCR

Quantitative polymerase chain reaction

Author contributions

LY: conceptualization, data curation, formal analysis, project administration, writing-original draft; DH: data collection, statistics, resources; LF, ZL, CZ, JZ, XY: conceptualization, methodology, review, and editing of the manuscript; BL: obtained funding and supervised study. Submitted and published versions of the manuscript were approved by all authors.

Funding

The study was supported by National Natural Science Foundation of China (81874224), Natural Science Foundation of Shandong Province (ZR2022ZD31), Special Fund Project of the Central Government for Guiding Local Science and Technology Development of Shandong Province (YDZX2022010).

Data availability

A reasonable request can be made to the corresponding author if you would like access to the datasets used and/or analyzed during this study.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethical approval

The Shandong Cancer Hospital and Institute Ethics Committee approved this study(Approval number: SDTHEC201803).

Consent for publication

Not applicable.