Inhibition of ferroptosis via SLC25A39-NRF2 axis drives Osimertinib resistance in lung adenocarcinoma

Abstract

Objective

Overcoming resistance to Osimertinib remains a major clinical challenge in lung adenocarcinoma (LUAD). The molecular mechanisms driving this resistance are still not fully understood.

Methods

Integrated bioinformatics analysis and functional assays were performed to investigate the role of SLC25A39 in LUAD progression and drug resistance. Mechanistic studies were conducted using co-immunoprecipitation and rescue experiments. Osimertinib-resistant cell models and xenograft assays were used to evaluate therapeutic responses.

Results

SLC25A39 was significantly upregulated in LUAD and correlated with unfavorable patient outcomes. Functional studies showed that SLC25A39 depletion suppressed malignant phenotypes and enhanced ferroptosis, whereas its overexpression produced the opposite effects. Mechanistically, SLC25A39 was found to interact with NRF2 and was associated with increased NRF2 stability and transcriptional activity, leading to enhanced glutathione synthesis and attenuation of lipid peroxidation. Importantly, silencing SLC25A39 sensitized LUAD cells to Osimertinib. Consistent with this, pharmacological induction of ferroptosis using RSL3 markedly enhanced the antitumor effects of Osimertinib in both parental and resistant models, resulting in reduced tumor growth in vitro and in vivo, particularly in SLC25A39-high contexts.

Conclusion

SLC25A39 promotes LUAD progression and Osimertinib resistance by suppressing ferroptosis via NRF2. Targeting ferroptosis may represent a promising strategy to overcome Osimertinib resistance.

Graphical Abstract

SLC25A39 drives ferroptosis resistance via NRF2–GSH axis and promotes Osimertinib resistance in LUAD. SLC25A39 activates NRF2 and promotes GSH synthesis and mitochondrial import, thereby inhibiting lipid peroxidation and ferroptosis, ultimately contributing to Osimertinib resistance in LUAD

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00018-026-06313-y.

Introduction

Lung adenocarcinoma (LUAD) is the most prevalent subtype of non-small cell lung cancer (NSCLC), comprising 40%-50% of NSCLC cases [1, 2]. It primarily originates from alveolar epithelial or glandular cells, and exhibits a higher propensity for distant metastases to the brain, bones, liver, and other organs [3–5]. In recent years, the incidence of LUAD has continued to rise. Although treatment strategies have been continuously optimized, the five-year survival rate remains relatively low [6]. The advancement of molecular research has significantly contributed to the development of targeted therapies, which have dramatically improved patient survival and prognosis [7, 8].

Among the molecular targets associated with LUAD, activating mutations in the epidermal growth factor receptor (EGFR) represent the most prevalent genetic alterations in NSCLC [9]. Notably, exon 19 deletions (19del) and the exon 21 L858R point mutation, collectively termed “common mutations,” account for approximately 90% of all EGFR mutations [10]. Tyrosine kinase inhibitors (TKIs) targeting EGFR mutations have demonstrated superior efficacy over traditional chemotherapy and have become the standard first-line treatment for patients with EGFR-mutant NSCLC [11–13]. Among the various EGFR-TKIs, Osimertinib, a third-generation EGFR-TKI, has emerged as the preferred first-line therapy for patients harboring common EGFR mutations [14]. However, similar to earlier-generation EGFR-TKIs, the emergence of acquired resistance to Osimertinib remains an inevitable challenge, significantly impacting the long-term success of EGFR-mutant NSCLC treatment. Current research has identified both EGFR-dependent and EGFR-independent resistance mechanisms [15, 16]. Nevertheless, the precise resistance mechanisms remain unclear in approximately two-thirds of Osimertinib-resistant patients [17]. Therefore, further investigation into the underlying resistance mechanisms, identification of potential resistance pathways, and the development of effective therapeutic strategies are imperative to enhance long-term survival outcomes in LUAD patients.

Ferroptosis is a non-apoptotic form of cell death that is dependent on iron ions and lipid peroxidation, fundamentally distinct from conventional cell death mechanisms such as apoptosis, necrosis, and autophagy [18, 19]. First introduced by Dixon et al. in 2012 [20], ferroptosis has since emerged as a critical research focus in cancer, cardiovascular disease, and metabolic disorders [21–23]. Increasing evidence suggests that ferroptosis inducers and ferroptosis-regulating genes can effectively suppress tumor cell proliferation and, to some extent, overcome targeted therapy resistance resulting from apoptosis evasion, including in LUAD [24]. However, the precise role of ferroptosis in Osimertinib resistance in LUAD remains incompletely understood.

SLC25A39 (Solute Carrier Family 25 Member 39) belongs to the solute carrier family 25 (SLC25) and encodes a mitochondrial membrane transporter that primarily mediates iron transport across the mitochondrial membrane [25]. This protein is closely linked to glutathione (GSH) metabolism, iron homeostasis, oxidative stress, and ferroptosis regulation [26, 27]. Recent studies have demonstrated that SLC25A39 facilitates pancreatic ductal adenocarcinoma (PDAC) progression and therapeutic resistance by modulating GSH metabolism, highlighting its pivotal role in antioxidant defense and chemoresistance [28]. Despite its known role in other cancers, the specific function of SLC25A39 in LUAD, particularly its impact on ferroptosis and chemoresistance mechanisms, remains unclear and warrants further investigation.

In this study, we identified SLC25A39 as a pro-tumorigenic factor in LUAD. We observed that high SLC25A39 expression inhibits ferroptosis in tumor cells, thereby promoting cell survival and driving Osimertinib resistance. Further in vivo and in vitro experiments confirmed that co-administration of a ferroptosis activator alongside Osimertinib significantly suppresses LUAD progression induced by SLC25A39 overexpression. This finding establishes SLC25A39 as a crucial regulatory factor in LUAD and suggests a novel therapeutic strategy to overcome Osimertinib resistance.

Methods

Bioinformatics analysis

Transcriptomic data for lung adenocarcinoma (LUAD) patients were retrieved from the TCGA-LUAD cohort (https://www.cancerimagingarchive.net/collection/tcga-luad/), while RNA-seq profiles of osimertinib-treated LUAD samples were obtained from the GEO dataset GSE285298. Differentially expressed genes (DEGs) were identified using the DESeq2 package in R, applying a threshold of |log2 fold change| > 1 and p-value < 0.05. Following DEG identification, KEGG pathway enrichment and Venn diagram analyses were conducted to screen for LUAD progression-related mitochondrial genes. The prognostic relevance of SLC25A39 expression was subsequently assessed using Kaplan–Meier survival analysis.

Immunohistochemistry (IHC)

IHC staining was performed on 4-µm TMA sections using an anti-SLC25A39 antibody (PA5-55415, Invitrogen) according to standard protocols. Sections were deparaffinized, rehydrated, and subjected to antigen retrieval in citrate buffer. After blocking, sections were incubated with the primary antibody at 4 °C overnight, followed by detection with HRP-conjugated secondary antibodies. DAB was used as the chromogen, and hematoxylin was used for counterstaining. Staining intensity and the percentage of positive tumor cells were evaluated independently by two pathologists in a blinded manner.

Cell culture and transfection

Human bronchial epithelial cell line (HBE, RRID: CVCL_0112) and LUAD cell lines (H1975 [RRID: CVCL_1511], A549 [RRID: CVCL_0023] and H1299 [RRID: CVCL_0060]) were obtained from the American Type Culture Collection (ATCC, USA). All cell lines were authenticated by short tandem repeat (STR) profiling and were routinely tested for mycoplasma contamination using PCR-based assays, with negative results prior to experimental use. Cells were maintained in complete medium (DMEM supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin) under standard culture conditions (37 °C, 5% CO₂), and passaged at approximately 80% confluence for subsequent experiments.

For gene silencing, cells were transfected with either short hairpin RNA (shRNA) or small interfering RNA (siRNA) targeting SLC25A39 (shSLC25A39 or siSLC25A39), along with corresponding non-targeting controls (shNC or siNC), using Lipofectamine 3000 (Invitrogen, USA) according to the manufacturer’s instructions. Knockdown efficiency was confirmed by Western blotting. For overexpression, cells were transfected with a pcDNA3.1-SLC25A39 plasmid (SLC25A39-OE) or an empty vector (EV) control using Lipofectamine 3000, and overexpression efficiency was validated by Western blotting.

Carboxyfluorescein diacetate, succinimidyl ester (CFDA) staining

After specific treatments, A549 and H1299 cells were collected and subjected to CFDA staining to assess cell viability. Briefly, CFDA dye (C1031, Beyotime, China) was added to the cells at a final concentration of 5 µM, followed by incubation in the dark for 15 min. Excess dye was then removed by washing thoroughly with PBS. Finally, cells were visualized and imaged using a fluorescence microscope.

Western blotting

Tissues and cells were lysed in RIPA buffer with phosphatase and protease inhibitors to extract total proteins. Protein concentrations were measured using the BCA Kit (23227, ThermoFisher). Proteins were separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skim milk. Membranes were incubated with primary antibodies overnight at 4 °C, then with HRP-conjugated secondary antibodies. Protein bands were detected using an ECL system and quantified with ImageJ. Primary antibody details are in Supplementary Table 1.

Co-immunoprecipitation (Co-IP)

Co-IP assays were performed using cell lysates prepared in IP lysis buffer containing protease inhibitors. After pre-clearing with Protein A/G agarose beads, lysates were incubated overnight at 4 °C with anti-SLC25A39, anti-NRF2, or control IgG antibodies, followed by incubation with Protein A/G beads for 2 h. Immunoprecipitated proteins were washed, eluted with SDS loading buffer, and analyzed by Western blotting. For exogenous Co-IP, 293T cells were co-transfected with Flag-SLC25A39 and Myc-NRF2 plasmids, and interactions were detected using anti-Flag or anti-Myc antibodies.

Protein stability assay

Cells were treated with cycloheximide (CHX, 100 µg/mL; Sigma-Aldrich) to inhibit de novo protein synthesis and harvested at the indicated time points (0, 5, 15, 30, and 60 min). Protein levels of NRF2 were analyzed by Western blotting, and band intensities were quantified and normalized to β-actin.

Subcellular fractionation

Cytoplasmic and nuclear fractions were isolated using a Nuclear and Cytoplasmic Extraction Kit (Beyotime, China) according to the manufacturer’s instructions. NRF2 distribution was detected by Western blotting, with β-actin and Lamin B used as cytoplasmic and nuclear markers, respectively.

Dual-luciferase reporter assay

Cells were co-transfected with an antioxidant response element (ARE)-luciferase reporter plasmid and a Renilla luciferase plasmid (internal control) using Lipofectamine 3000 (Invitrogen). After 24–48 h, luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega), and firefly luciferase activity was normalized to Renilla luciferase activity.

Lipid peroxidation assay

Lipid peroxidation levels were assessed using the BODIPY 581/591 C11 probe (Thermo Fisher, USA). Treated LUAD cells (A549 and H1299) were collected and incubated with 2 µM BODIPY 581/591 C11 in serum-free DMEM at 37 °C for 30 min in the dark. After incubation, cells were washed three times with PBS to remove excess dye. The cells were immediately imaged, with oxidized BODIPY emitting green fluorescence and non-oxidized BODIPY emitting red fluorescence. The lipid peroxidation index was calculated as the ratio of green to red fluorescence intensity using ImageJ software.

Mitochondrial membrane potential and morphology analysis

Mitochondrial membrane potential (ΔΨm) and morphology were assessed using MitoTracker Red (40740ES50, YEASEN, China) and MitoTracker Green (40742ES50, YEASEN, China). Cells were incubated with 200 nM dye in serum-free DMEM at 37 °C for 30 min in the dark, followed by washing with PBS and immediate imaging.

To validate ΔΨm-dependent staining, cells were treated with the mitochondrial uncoupler FCCP (10 µM, 20 min) prior to MitoTracker Red staining as a depolarization control. Fluorescence images were acquired using identical microscope settings (laser intensity, exposure time, and gain) across all groups. For each condition, at least five random fields from three independent experiments were analyzed. Mitochondrial morphology was quantified using ImageJ software. The aspect ratio (major axis/minor axis) was used to assess mitochondrial elongation, while mitochondrial network integrity was additionally evaluated by categorizing mitochondria as interconnected or fragmented. Quantification was performed in a blinded manner.

MDA, Fe²⁺, and GSH level measurement

Malondialdehyde (MDA), ferrous ion (Fe²⁺), and glutathione (GSH) levels were measured using commercial assay kits (Beyotime, China) following the manufacturer’s protocols.

Mouse model

Subcutaneous tumor xenografts were established by subcutaneously injecting 5 × 10⁶ SLC25A39-overexpressing A549 cells or control cells into the flank region of 6-week-old BALB/c nude mice (n = 5). When tumor volumes reached approximately 5 mm in longitudinal diameter, the mice were randomly assigned into five groups: Control, SLC25A39, SLC25A39 + Osi, SLC25A39 + RSL3, and SLC25A39 + Osi + RSL3. Osimertinib (20 mg/kg/day) was administered by oral gavage every other day to the SLC25A39 + Osi and SLC25A39 + Osi + RSL3 groups. RSL3 (100 mg/kg) was administered by intratumoral injection twice a week for two weeks to the SLC25A39 + RSL3 and SLC25A39 + Osi + RSL3 groups. The Control group received equal volume of corn oil via oral gavage. Tumor dimensions were measured twice daily using digital calipers, and tumor volumes were calculated as 1/2×length×(Width)². Body weight and health status were monitored throughout the treatment period. At the end of the study, tumors were excised, weighed, and photographed. All animal procedures were approved by the Institutional Animal Care and Use Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (TJH-202209010), and Institutional Ethics Committee of Huazhong University of Science and Technology (TJ-IRB202502107). Committee and conducted in accordance with ethical guidelines for laboratory animals.

Protein–protein docking analysis

Protein–protein interaction between SLC25A39 and NRF2 was predicted using AlphaFold-Multimer. The full-length amino acid sequences of human SLC25A39 and NRF2 were obtained from the UniProt database and submitted as input sequences. Five independent models were generated, and the top-ranked model was selected according to the AlphaFold ranking confidence score. Model confidence was evaluated using the predicted local distance difference test (pLDDT), predicted aligned error (PAE), and interface predicted template modeling score (ipTM). The interaction interface was visualized using PyMOL, and residues located at the predicted binding interface were analyzed. The binding energy of the predicted SLC25A39–NRF2 complex was further estimated using molecular interaction analysis, yielding a binding energy of − 23.0 kcal/mol. The confidence parameters of the selected model were as follows: ipTM = 0.16, and pTM = 0.32.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 10 (GraphPad Software, USA). Data are presented as mean ± SD unless otherwise indicated. Normality was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated using Levene’s test. For comparisons between two groups, unpaired two-tailed Student’s t-test was used. For multiple group comparisons, one-way ANOVA followed by Tukey’s post hoc test was applied. When variance was unequal, Welch’s correction was used. Non-parametric tests were applied when data did not meet normality assumptions. All experiments were performed with at least three independent biological replicates unless otherwise specified. The number of replicates (n) is indicated in the figure legends. A P value < 0.05 was considered statistically significant.

Results

SLC25A39 is upregulated in LUAD and correlates with poor prognosis

In an effort to uncover critical regulators involved in LUAD progression, we conducted a comprehensive analysis integrating transcriptomic data from LUAD-related datasets and RNA-seq profiles derived from osimertinib-treated LUAD patients (Fig. 1A-B). KEGG pathway enrichment analysis of DEGs revealed a pronounced involvement of mitochondrial-related pathways, implicating mitochondrial dysfunction as a potential contributor to both LUAD development and therapeutic resistance (Fig. 1C). Through intersecting these DEGs with a curated list of mitochondria-associated genes, SLC25A39 emerged as a significantly candidate (Fig. 1D). Analysis of the TCGA-LUAD cohort further demonstrated that SLC25A39 expression was markedly elevated in tumor tissues compared with adjacent normal counterparts, and high expression levels were strongly associated with unfavorable overall survival (Fig. 1E-G).

Fig. 1.

High expression of SLC25A39 in LUAD patients and its association with poor prognosis. (A) Volcano plot illustrating the DEGs from the TCGA-LUAD dataset. (B) Volcano plot showcasing the DEGs related to Osimertinib resistance. (C) Pathway enrichment analysis of the DEGs. (D) Venn analysis revealing the overlap of LUAD-DEGs, Osimertinib resistance-related DEGs, and genes associated with mitochondrial transport. (E) Expression distribution of SLC25A39 in TCGA-LUAD cohort shown by a violin plot. (F) Boxplot comparing SLC25A39 expression between LUAD tumors and adjacent normal tissues in TCGA dataset. (G) Kaplan–Meier survival analysis of TCGA-LUAD patients stratified by SLC25A39 expression levels. (H) Kaplan–Meier survival analysis of LUAD patients in the tissue microarray cohort (n = 79) based on SLC25A39 expression. (I) Western blot analysis of SLC25A39 protein levels in 10 pairs of LUAD and adjacent normal tissues, with corresponding densitometric quantification (n = 10). Each pair of points represents an individual clinical sample (Case 1–10). (J) Representative IHC images showing SLC25A39 expression in LUAD tissues (T) versus adjacent normal tissues (N) from 10 patients. Quantification of IHC-positive cell percentages is shown on the right (n = 10). Scale bar, 100 μm. Each pair of points represents an individual clinical sample (Case 1–10). Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

To reinforce these observations, we examined an independent LUAD patient cohort (n = 79) using tissue microarrays. Kaplan–Meier survival analysis indicated that patients exhibiting high SLC25A39 expression had significantly shorter overall survival, consistent with findings from the public dataset (Fig. 1H). Clinically, elevated SLC25A39 expression correlated with more advanced pathological stage and greater lymph node involvement (Table 1). At the protein level, upregulation of SLC25A39 in tumor tissues was consistently confirmed by both Western blot and IHC analyses across ten pairs of LUAD and matched adjacent tissues (Fig. 1I-J). Collectively, these results support a tumor-promoting role for SLC25A39 and highlight its potential as a novel therapeutic target in LUAD.

Table 1.

Relationship between SLC25A39 expression and tumor characteristics in patients with LUAD (n = 79)

Features	No. of patients	SLC25A39 expression		p value
		low	high
All patients	79	41	38
Age (years)				0.9137
<63	39	20	19
≥ 63	40	21	19
Gender				0.2289
Male	45	26	19
Female	34	15	19
Tumor size				0.7573
<4 cm	36	18	18
≥ 4 cm	43	23	20
Lymph node positive				0.0033**
≤ 1	44	29	15
> 1	34	11	23
Grade				0.4496
I & II	55	27	28
III	24	14	10
AJCC Stage				0.0169*
1	23	16	7
2	17	11	6
3 & 4	37	13	24
T Infiltrate				0.4536
T1	17	11	6
T2	44	22	22
T3 & T4	14	8	10
lymphatic metastasis(N)				0.1834
N0	33	22	11
N1	16	8	8
N2	12	4	8
N3	4	3	1
Expression of EGFR(Fish)				0.0523
negative	63	35	28
positive	12	3	9

*P < 0.05; **p < 0.01

Downregulation of SLC25A39 suppresses the malignant progression of LUAD cells

We first profiled SLC25A39 expression across LUAD cell lines and found that it was markedly upregulated in A549 and H1299 cells compared with the normal human bronchial epithelial cell line HBE (Fig. 2A). Based on their relatively high endogenous expression, A549 and H1299 cells were selected for subsequent loss-of-function studies. SLC25A39 was then silenced using two independent short hairpin RNAs (shRNAs), both of which achieved efficient knockdown as confirmed by Western blotting (Fig. 2B).

Fig. 2.

SLC25A39 knockdown inhibits proliferation, migration, and invasion of LUAD cells and increases apoptosis. (A) Western blot analysis of SLC25A39 protein expression in LUAD cell lines (H1975, H1299, and A549) and human bronchial epithelial cell line (HBE), with statistical quantification of protein expression (n = 3). (B) Western blot analysis was performed to assess the efficiency of SLC25A39 silencing in A549 and H1299 cells (n = 3). (C) CFDA-SE staining assays were conducted to evaluate cell viability between SLC25A39-silenced LUAD cells and control cells (n = 3). Scale bar, 100 μm. (D) A colony formation assay was performed to assess the proliferative capacity of control and shSLC25A39 LUAD cells (n = 3). (E) A Transwell invasion assay was conducted to determine the invasive potential of control and shSLC25A39 LUAD cells (n = 3). Scale bar, 100 μm. (F) Western blot analysis was utilized to examine the expression levels of EMT-related proteins in control and shSLC25A39 LUAD cells, with quantification of their relative protein abundance (n = 3). (G) Flow cytometry analysis was employed to measure apoptosis levels in control and shSLC25A39 LUAD cells (n = 3). Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Functionally, depletion of SLC25A39 was accompanied by a clear reduction in cell viability, which was paralleled by a marked decrease in colony-forming capacity (Fig. 2C-D). In line with this, fewer cells traversed the Transwell membrane upon SLC25A39 silencing, indicating impaired migratory ability (Fig. 2E). Alongside these changes, the proportion of apoptotic cells was notably increased following SLC25A39 knockdown (Fig. 2G). At the molecular level, SLC25A39 depletion coincided with elevated E-cadherin expression and reduced levels of N-cadherin and vimentin (Fig. 2F).

These phenotypic alterations were further corroborated using an independent siRNA-mediated silencing approach, which yielded consistent changes across cell viability, proliferation, migration, apoptosis, and EMT-associated markers (Fig. S1).

SLC25A39 significantly influences ferroptosis sensitivity in LUAD cells

Given the close association between SLC25A39 and mitochondrial function, we aimed to investigate its specific role in mitochondrial regulation in LUAD cells. To this end, we silenced SLC25A39 in A549 and H1299 cells and subsequently reintroduced its expression to evaluate its impact on mitochondrial function in LUAD cells (Fig. 3A). Preliminary CFDA staining results demonstrated that the decrease in cell viability induced by SLC25A39 silencing was effectively rescued upon its re-expression (Fig. 3B). At the mitochondrial level, loss of SLC25A39 was accompanied by a decrease in mitochondrial membrane potential (Fig. S2), along with increased mitochondrial fragmentation and a reduction in intact mitochondrial networks (Fig. 3C-D). These mitochondrial impairments were markedly reversed following SLC25A39 re-expression, further supporting its essential role in maintaining mitochondrial homeostasis.

Fig. 3.

Upregulation of SLC25A39 promotes ferroptosis resistance in LUAD cells. (A) Western blot analysis was performed to confirm the transfection efficiency of SLC25A39 overexpression in SLC25A39-silenced LUAD cells (n = 3). (B) CFDA staining was used to assess cell viability across different experimental groups (n = 3). Scale bar, 100 μm. (C) Mitotracker-Red staining was conducted to evaluate mitochondrial membrane potential alterations in each group (n = 3). Scale bar, 100 μm. (D) Mitotracker-Green staining was used to examine mitochondrial morphology and quantity variations among the groups (n = 3). Scale bar, 10 μm. (E) Correlation analysis was performed to determine the association between SLC25A39 expression and ferroptosis (n = 3). (F) Biochemical assays were utilized to measure Fe²⁺, MDA, and GSH levels in different cell groups (n = 3). (G) BODIPY 581/591 C11 staining was performed to assess intracellular lipid peroxidation levels (n = 3). Scale bar, 100 μm. (H) Western blot analysis was used to evaluate GPX4 and NRF2 protein expression levels, with quantification of their relative abundance across different groups (n = 3). Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Mitochondrial morphological and functional alterations have been widely recognized as key hallmarks of ferroptosis. Correlation analysis revealed a significant inverse association between SLC25A39 expression and ferroptosis (Fig. 3E). In parallel, the effects of SLC25A39 silencing on cell viability, proliferation, and EMT-associated changes were attenuated by ferroptosis inhibition (Fig. S3). These findings prompted further investigation into the role of SLC25A39 in ferroptosis regulation in LUAD cells. Biochemical analyses demonstrated that SLC25A39 downregulation significantly enhanced ferroptosis, as evidenced by a marked increase in Fe²⁺ and MDA levels, accompanied by a substantial depletion of GSH (Fig. 3F). Consistently, BODIPY 581/591 C11 staining further confirmed that SLC25A39 silencing led to a significant increase in intracellular lipid peroxidation, highlighting its crucial role in ferroptosis resistance (Fig. 3G). Furthermore, NRF2 and GPX4, key regulators of ferroptosis, play essential roles in antioxidant defense, iron metabolism, and lipid homeostasis. Western blot analysis revealed that SLC25A39 expression was positively correlated with NRF2 and GPX4 levels. Specifically, SLC25A39 silencing resulted in a significant downregulation of NRF2 and GPX4 protein expression, whereas reintroducing SLC25A39 effectively restored their levels (Fig. 3H). Altogether, these results demonstrate that SLC25A39 confers a survival advantage to LUAD cells by suppressing ferroptosis through mitochondrial regulation, thereby promoting LUAD progression.

SLC25A39-NRF2 axis regulates ferroptosis resistance in LUAD

However, the molecular mechanism by which SLC25A39 confers resistance to ferroptosis remains unclear. Protein–protein docking analysis first predicted a direct interaction between SLC25A39 and the transcription factor NRF2, with a binding energy of − 23.0 kcal/mol (Fig. 4A). This interaction was subsequently validated by endogenous Co-IP assays in A549 cells (Fig. 4B), and further confirmed by exogenous Co-IP following co-expression of Flag-SLC25A39 and Myc-NRF2 in 293T cells (Fig. 4C).

Fig. 4.

SLC25A39 interacts with NRF2 to maintain redox homeostasis and inhibit ferroptosis in LUAD cells. (A) Molecular docking analysis was performed to predict the interaction between SLC25A39 and NRF2. (B) Co-IP was conducted in A549 cells to detect the endogenous interaction between SLC25A39 and NRF2. (C) Co-IP assay was carried out in 293T cells co-transfected with Flag-SLC25A39 and Myc-NRF2 to validate their exogenous interaction. (D) qRT-PCR analysis of NRF2 mRNA levels following SLC25A39 overexpression (n = 3). (E) Western blot analysis of NRF2 protein levels upon SLC25A39 overexpression (n = 3). (F) Cycloheximide (CHX) chase assay assessing NRF2 protein stability in the presence of SLC25A39 or control. (G) Co-IP assays examining the interaction between NRF2, KEAP1, and p62 following SLC25A39 overexpression. (H) Western blot analysis of NRF2 protein levels with or without MG-132 treatment in SLC25A39-overexpressing cells. (I) Subcellular fractionation analysis showing NRF2 distribution in cytoplasmic and nuclear fractions. (J) ARE luciferase reporter assay measuring NRF2 transcriptional activity (n = 3). (K) qRT-PCR analysis of NRF2 target genes (HMOX1, NQO1, and GCLM) (n = 3). (L) BODIPY 581/591 C11 staining detecting lipid peroxidation in indicated groups (n = 3). Scale bar, 100 μm. (M) Western blot analysis of GPX4 and NRF2 expression following SLC25A39 knockdown with or without NRF2 re-expression (n = 3). (N) Measurement of Fe²⁺, MDA, and GSH levels in indicated groups (n = 3). (O) MitoTracker Red staining assessing mitochondrial membrane potential (n = 3). Scale bar, 100 μm. (P) MitoTracker Green staining evaluating mitochondrial morphology and network integrity (n = 3). Scale bar, 10 μm. Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

We next examined the regulatory effect of SLC25A39 on NRF2 expression. qPCR analysis showed that SLC25A39 overexpression had no significant effect on NRF2 mRNA levels (Fig. 4D), whereas its protein level was markedly increased (Fig. 4E). Consistently, cycloheximide (CHX) chase assays demonstrated that SLC25A39 overexpression significantly delayed NRF2 protein degradation and prolonged its half-life (Fig. 4F). Given that KEAP1-mediated ubiquitin–proteasome degradation represents the primary regulatory pathway for NRF2 turnover, we further investigated alterations in this process. Co-IP results revealed that SLC25A39 overexpression reduced the interaction between NRF2 and KEAP1, while enhancing the association between p62 and KEAP1 (Fig. 4G). Consistently, treatment with MG-132 led to further accumulation of NRF2 protein (Fig. 4H). Subcellular fractionation assays further demonstrated that SLC25A39 overexpression promoted nuclear accumulation of NRF2 (Fig. 4I). In line with this, ARE luciferase reporter activity was significantly increased (Fig. 4J), accompanied by marked upregulation of downstream target genes, including HMOX1, NQO1, and GCLM at the transcriptional level (Fig. 4K).

Functionally, the reduction in cell viability induced by SLC25A39 knockdown was partially rescued by NRF2 re-expression (Fig. S4A). Meanwhile, SLC25A39 deficiency markedly increased lipid peroxidation, which was substantially attenuated upon restoration of NRF2 (Fig. 4L, Fig. S4B). NRF2 overexpression reversed the downregulation of GPX4 caused by SLC25A39 loss (Fig. 4M), and partially restored intracellular Fe²⁺, MDA, and GSH levels (Fig. 4N). Further analysis of mitochondrial phenotypes showed that NRF2 restoration alleviated SLC25A39 depletion–induced mitochondrial dysfunction, as evidenced by improved membrane potential and reduced structural fragmentation (Fig. 4O-P, Fig.S4C).

SLC25A39 knockdown enhances the sensitivity of LUAD cells to Osimertinib

We further investigated the impact of SLC25A39 expression on the response of LUAD cells to Osimertinib treatment. To this end, SLC25A39-silenced and control A549 and H1299 cell lines were established and subsequently treated with varying concentrations of Osimertinib. CCK-8 assay results demonstrated that SLC25A39 silencing significantly enhanced the sensitivity of lung cancer cells to Osimertinib, as evidenced by a 50% reduction in cell viability at approximately 0.8 µM Osimertinib, compared to the control group (Fig. 5A). CFDA staining further validated this observation, indicating that SLC25A39 knockdown decreases cell survival and enhances the antitumor effect of Osimertinib (Fig. 5B).

Fig. 5.

Knockdown of SLC25A39 enhances Osimertinib-induced ferroptosis in LUAD cells. (A) CCK-8 assay was performed to evaluate the sensitivity of SLC25A39-silenced and control LUAD cells to Osimertinib (n = 3). (B) CFDA staining was used to assess cell viability in control, Osimertinib (Osi), and Osi + shSLC25A39 LUAD cells (n = 3). Scale bar, 100 μm. (C) BODIPY 581/591 C11 staining was conducted to measure intracellular lipid peroxidation levels across different treatment groups (n = 3). Scale bar, 100 μm. (D-F) Biochemical assays were utilized to quantify Fe²⁺, MDA, and GSH levels in various experimental groups (n = 3). (G) Mitotracker-Green staining was performed to analyze mitochondrial morphology and quantity changes in different cell groups (n = 3).Scale bar, 10 μm. Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Given the potential role of SLC25A39 in ferroptosis regulation in LUAD cells, we next examined the effects of SLC25A39 silencing combined with Osimertinib treatment on ferroptosis-related biochemical indicators. C11-BODIPY staining results revealed that Osimertinib treatment alone significantly elevated lipid ROS levels in A549 and H1299 cells, as indicated by a reduction in red fluorescence intensity and an increase in green fluorescence, suggesting enhanced lipid peroxidation (Fig. 5C). Furthermore, MDA and Fe²⁺ levels were markedly increased following Osimertinib treatment, accompanied by a significant depletion of GSH (Fig. 5D-F). SLC25A39 knockdown further enhanced these effects, suggesting that it may potentiate Osimertinib-induced ferroptosis. Additionally, mitochondrial damage, a hallmark of ferroptosis, was also assessed. Mitotracker-Green staining results demonstrated that both Osimertinib treatment alone and its combination with SLC25A39 silencing significantly promoted mitochondrial fragmentation, suggesting structural disruption of mitochondria (Fig. 5G). Collectively, our findings highlight SLC25A39 as a potential therapeutic target in lung cancer, playing a pivotal role in Osimertinib sensitivity and ferroptosis regulation in LUAD. Targeting SLC25A39 may serve as a promising strategy to enhance the therapeutic efficacy of Osimertinib in LUAD patients.

Loss of SLC25A39 drives ferroptosis in Osimertinib-resistant LUAD cells

To explore the role of SLC25A39 in Osimertinib resistance in LUAD cells, we chronically exposed A549 and H1299 cells to Osimertinib for 24 weeks, successfully establishing Osimertinib-resistant cell lines (A549-OR / H1299-OR) (Fig. 6A). Drug sensitivity assays revealed that A549-OR and H1299-OR cells exhibited a significantly increased IC50, confirming the establishment of resistance (Fig. 6B). At the protein level, Western blot analysis demonstrated that GPX4, NRF2, and SLC25A39 were all markedly upregulated in the resistant cell lines, suggesting that ferroptosis resistance may play a crucial role in Osimertinib resistance (Fig. 6C). To further validate the function of SLC25A39 in resistant LUAD cells, we silenced SLC25A39 in A549-OR and H1299-OR cells and assessed ferroptosis-related indicators. CFDA staining results showed that SLC25A39 depletion significantly reduced the viability of resistant cells (Fig. 6D). Moreover, we observed that inhibiting SLC25A39 strongly promoted ferroptosis in the resistant cell lines. Specifically, compared to the control group, SLC25A39-silenced A549-OR and H1299-OR cells exhibited significantly higher levels of lipid peroxidation, decreased mitochondrial membrane potential, and increased mitochondrial fragmentation, further confirming that SLC25A39 contributes to ferroptosis resistance and is involved in Osimertinib resistance mechanisms (Fig. 6E-I). These findings highlight the critical role of SLC25A39 in Osimertinib-resistant LUAD cells, where its upregulation promotes ferroptosis resistance, thereby sustaining the survival of resistant cells.

Fig. 6.

Silencing SLC25A39 sensitizes LUAD cells to Osimertinib. (A) Osimertinib-resistant LUAD cell lines (A549-OR/H1299-OR) were established by continuously exposing LUAD cells to Osimertinib for 24 weeks. (B) CCK-8 assay was performed to confirm the successful establishment of Osimertinib-resistant cell lines (n = 3). (C) Western blot analysis was used to assess the protein expression levels of ferroptosis-related markers (GPX4 and NRF2) and SLC25A39 in A549-OR/H1299-OR cells compared to control cells, with quantification of relative protein abundance. (D) CFDA staining was conducted to evaluate cell viability in A549-OR/H1299-OR cells following SLC25A39 silencing (n = 3). Scale bar, 100 μm. (E) Biochemical assays were performed to measure Fe²⁺, MDA, and GSH levels across different treatment groups (n = 3). (F) Western blot analysis was used to examine the protein expression levels of GPX4, NRF2, and SLC25A39 in various experimental groups, with quantification of protein abundance (n = 3). (G) BODIPY 581/591 C11 staining was utilized to assess intracellular lipid peroxidation levels in each group (n = 3). Scale bar, 100 μm. (H) Mitotracker-Red staining was conducted to evaluate mitochondrial membrane potential alterations in different cell groups (n = 3). Scale bar, 100 μm. (I) Mitotracker-Green staining was used to analyze mitochondrial morphology and quantity variations among different experimental conditions (n = 3). Scale bar, 10 μm. Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Osimertinib combined with RSL3 enhances ferroptosis in LUAD cells

To further investigate the role of SLC25A39 in ferroptosis resistance in LUAD cells, we treated SLC25A39-overexpressing A549 and H1299 cells with Osimertinib, the ferroptosis inducer RSL3 (a GPX4 inhibitor), or their combination. Western blot analysis revealed that GPX4 protein levels were significantly downregulated following Osimertinib treatment, to a degree comparable to that observed in the RSL3-treated group (Fig. 7A). Notably, Osimertinib and RSL3 combination therapy further suppressed GPX4 expression, indicating a synergistic effect on ferroptosis induction. CFDA staining results further confirmed that Osimertinib combined with RSL3 significantly reduced the viability of SLC25A39-overexpressing A549 and H1299 cells, with a greater inhibitory effect than either treatment alone, suggesting that the combination treatment potentiates ferroptosis induction (Fig. 7B). We then assessed biochemical indicators of ferroptosis to further validate this effect. The results demonstrated that Osimertinib combined with RSL3 markedly promoted ferroptosis in SLC25A39-overexpressing LUAD cells, as indicated by increased lipid peroxidation, decreased mitochondrial membrane potential, elevated MDA and Fe²⁺ levels, and depleted GSH levels (Fig. 7C-E).

Fig. 7.

Ferroptosis inducer enhances the antitumor effect of Osimertinib in vitro. (A) SLC25A39-overexpressing A549 and H1299 cells were treated with Osimertinib (Osi), RSL3 (a GPX4 inhibitor), or a combination of Osi + RSL3. Western blot analysis was performed to assess GPX4 protein expression in different groups, with quantification of GPX4 protein levels (n = 3). (B) CFDA staining was conducted to evaluate cell viability across different treatment groups (n = 3). Scale bar, 100 μm. (C) BODIPY 581/591 C11 staining was used to measure intracellular lipid peroxidation levels in each experimental group (n = 3). Scale bar, 100 μm. (D) Mitotracker-Red staining was performed to assess mitochondrial membrane potential alterations in response to different treatments (n = 3). Scale bar, 100 μm. (E) Biochemical assays were utilized to determine Fe²⁺, MDA, and GSH levels across various treatment groups (n = 3). Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Ferroptosis induction sensitizes resistant LUAD cells to osimertinib

We next examined whether this ferroptosis-based strategy could be extended to osimertinib-resistant LUAD cells. In A549-OR and H1299-OR cells, osimertinib or RSL3 alone moderately reduced cell viability, whereas their combination produced a stronger decrease in CFDA-SE fluorescence intensity (Fig. 8A). At the oxidative damage level, lipid ROS signals were increased by either treatment alone and were further enhanced under combination treatment (Fig. 8B). MitoTracker-Red staining showed a greater loss of mitochondrial membrane potential in the combination group than in the single-treatment groups (Fig. 8C).

Fig. 8.

RSL3 reverses Osimertinib resistance in LUAD cells. (A) A549-OR and H1299-OR cells were treated with RSL3, Osimertinib (Osi), or a combination of RSL3 + Osi, and CFDA staining was performed to assess cell viability across different treatment groups (n = 3). Scale bar, 100 μm. (B) BODIPY 581/591 C11 staining was conducted to evaluate intracellular lipid peroxidation levels in each experimental group (n = 3). Scale bar, 100 μm. (C) Mitotracker-Red staining was used to measure mitochondrial membrane potential alterations following different treatments (n = 3). Scale bar, 100 μm. (D) Biochemical assays were performed to determine Fe²⁺, MDA, and GSH levels across various treatment groups (n = 3). (E) Mitotracker-Green staining was conducted to assess mitochondrial morphology and quantity changes in response to different treatments (n = 3). Scale bar, 10 μm. Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Consistently, combined osimertinib and RSL3 treatment further increased MDA and Fe²⁺ levels while reducing GSH levels in resistant cells (Fig. 8D). MitoTracker-Green staining also revealed a more pronounced shift from intact mitochondrial networks toward fragmented mitochondria, together with a reduced aspect ratio, particularly after combination treatment (Fig. 8E). Together, these findings indicate that ferroptosis induction enhances the response to osimertinib in resistant LUAD cells.

RSL3 enhances the therapeutic efficacy of Osimertinib in LUAD in vivo

To further validate the role of ferroptosis in LUAD progression, we established a subcutaneous tumor model using A549 cells, which exhibit sensitivity to osimertinib treatment (Fig. 9A). Consistently, in control xenografts without SLC25A39 manipulation, both osimertinib and RSL3 treatments suppressed tumor growth, with the combination showing a more pronounced inhibitory effect (Fig. S5). In this context, we further evaluated tumor responses under SLC25A39 overexpression. Treatment with osimertinib or RSL3 significantly attenuated tumor progression, as reflected by reduced tumor volume and tumor weight (Fig. 9B-D).

Fig. 9.

RSL3 enhances Osimertinib sensitivity in LUAD in vivo. (A) SLC25A39-overexpressing A549 and H1299 cells were treated with varying concentrations of Osimertinib, and CCK-8 assays were performed to assess cell viability under different treatment conditions (n = 3). (B) A subcutaneous tumor model was established using SLC25A39-overexpressing A549 cells, with mice divided into five groups: Control, SLC25A39, SLC25A39 + Osi, SLC25A39 + RSL3, and SLC25A39 + Osi + RSL3. Representative tumor images from each group are shown (n = 5). (C) Tumor volume measurements were recorded and statistically analyzed for each experimental group. (D) Tumor weight was measured and compared across groups (n = 5). (E) Representative HE staining and IHC staining for Ki-67 in tumor tissues from each group, along with quantification of Ki-67-positive cells in tumor sections (n = 5). Scale bar, 100 μm. (F) Western blot analysis of SLC25A39, NRF2, and GPX4 protein levels in xenograft tumor tissues from the indicated groups, with quantification of band intensities (n = 3). Data are presented as mean ± SD. Statistical significance was determined using Student’s t-test or one-way ANOVA as indicated. *P < 0.05, **P < 0.01, ***P < 0.001

Histological analysis revealed that tumors in the SLC25A39-overexpressing group displayed a denser cellular architecture, whereas treatment with osimertinib or RSL3 was associated with increased necrotic regions and lipid vacuolation (Fig. 9E). These changes were further enhanced under combination treatment. Consistently, Ki-67 staining showed increased proliferative activity in SLC25A39-overexpressing tumors, which was reduced following treatment with osimertinib or RSL3, and further decreased upon combined treatment (Fig. 9E). At the molecular level, SLC25A39 overexpression elevated NRF2 and GPX4 protein levels in tumor tissues, whereas co-treatment with osimertinib and RSL3 reversed these changes (Fig. 9F).

Discussion

The application of EGFR-TKIs has significantly improved the clinical prognosis of patients with EGFR-mutant NSCLC, offering prolonged survival and a lower incidence of adverse events, thereby establishing itself as a first-line treatment option [29]. However, regardless of the specific EGFR-TKI used, most patients eventually develop resistance. The mechanisms underlying this resistance are highly complex and heterogeneous, often involving the coexistence and dynamic evolution of multiple resistance pathways. This complexity not only complicates the development of effective therapeutic strategies but also results in suboptimal responses to subsequent treatments in some patients [30, 31]. Osimertinib, a third-generation EGFR-TKIs, is primarily used to treat EGFR-mutant NSCLC, particularly by overcoming resistance caused by the T790M mutation [32, 33]. However, the emergence of resistance to Osimertinib severely limits its long-term efficacy, posing a significant challenge to the treatment of EGFR-mutant NSCLC [34]. Therefore, a comprehensive investigation into the mechanisms underlying Osimertinib resistance is essential for identifying key factors driving resistance and for facilitating the development of more precise therapeutic strategies. In this study, we identified SLC25A39 as a regulator of ferroptosis-related resistance to Osimertinib in LUAD (Graphical Abstract). Based on these findings, we propose that targeting SLC25A39 or employing ferroptosis inducers to reverse Osimertinib resistance may represent a promising therapeutic approach.

The role of SLC25A39 in tumorigenesis has increasingly garnered attention. For instance, in colorectal cancer, SLC25A39 is upregulated and promotes tumor cell growth and metastasis by modulating ROS, highlighting its potential involvement in tumor progression [35]. Moreover, recent studies have demonstrated that SLC25A39 mediates resistance to cuproptosis in PDAC by regulating GSH metabolism. However, the functional significance of SLC25A39 in LUAD remains largely undefined, particularly regarding its role in the progression of Osimertinib resistance [28]. In this study, we identified SLC25A39 as being highly expressed in LUAD patients and significantly associated with poor prognosis and Osimertinib resistance. Further functional experiments revealed that SLC25A39 knockdown markedly suppressed the malignant phenotype of LUAD cells and enhanced their sensitivity to Osimertinib. These findings not only elucidate the critical role of SLC25A39 in the resistance mechanisms of LUAD but also suggest its potential as a biomarker in LUAD progression.

It has been reported that SLC25A39 plays a crucial role in maintaining mitochondrial homeostasis and cellular metabolism by regulating the exchange of substances between the mitochondria and the cytoplasm [36]. Wang et al. [25] demonstrated that SLC25A39 is likely involved in the mitochondrial transport of GSH, as its deletion significantly decreased mitochondrial GSH import and levels, while not affecting the overall GSH levels in the cytoplasm. Additionally, they observed that cells deficient in both SLC25A39 and its paralog SLC25A40 exhibited notable defects in the stability and activity of iron-sulfur cluster (ISC) proteins. Further supporting these findings, Shen et al. [27] confirmed that SLC25A39 plays a pivotal role in coordinating mitochondrial GSH import and iron homeostasis, which is essential for maintaining oxidative phosphorylation (OXPHOS) functionality. Notably, iron homeostasis dysregulation is intricately linked to ferroptosis, and a substantial body of research has demonstrated that ferroptosis serves as a pivotal mechanism in tumor progression and the development of drug resistance [37, 38]. For instance, USP8 has been shown to enhance GPX4 stability by deubiquitinating it, thereby suppressing ferroptosis and promoting immune therapy resistance in colorectal cancer [39]. Moreover, exploiting the metabolic vulnerability of cancer stem cells (CSCs) to ferroptosis has been demonstrated as a potential strategy to enhance the efficacy of lung cancer immunotherapy [40].

In this study, we identified a strong association between SLC25A39 expression and ferroptosis resistance in LUAD. SLC25A39 knockdown markedly promoted ferroptosis, whereas its overexpression conferred resistance. Mechanistically, our data support a model in which SLC25A39 regulates ferroptosis through modulation of the transcription factor NRF2. SLC25A39 physically interacts with NRF2 and is associated with enhanced nuclear translocation and transcriptional activity. Activated NRF2, in turn, induces the expression of genes involved in GSH biosynthesis, leading to increased intracellular GSH levels and enhanced GPX4 activity, thereby limiting lipid peroxidation and ferroptosis. While SLC25A39 has been primarily characterized as a mitochondrial GSH transporter, these findings suggest an additional role in coordinating redox homeostasis through NRF2-dependent pathways.

Notably, the mechanism by which SLC25A39 regulates NRF2 and GPX4 expression remains to be fully defined. Several non-mutually exclusive hypotheses may account for this observation. First, given its role in mitochondrial GSH transport, SLC25A39 depletion may disrupt mitochondrial redox balance, leading to oxidative stress and perturbation of the KEAP1–NRF2 regulatory axis, thereby affecting NRF2 stability and activity. Second, our observation of altered NRF2 protein stability raises the possibility that SLC25A39 influences NRF2 turnover, potentially through modulation of KEAP1-mediated ubiquitination and proteasomal degradation. Third, mitochondrial dysfunction induced by SLC25A39 deficiency may activate retrograde signaling pathways that suppress NRF2 activation. These mechanisms underscore a complex interplay between mitochondrial metabolism and redox signaling, which warrants further investigation.

Taken together, these findings position SLC25A39 as a regulator of ferroptosis sensitivity and redox homeostasis in LUAD. Importantly, both in vitro and in vivo data indicate that combining ferroptosis induction with osimertinib treatment enhances therapeutic responses, particularly in the context of resistance. These results provide a rationale for exploring ferroptosis-targeting strategies to improve the efficacy of EGFR-targeted therapies in LUAD.

Conclusion

In summary, our research identifies SLC25A39 as a key driver of Osimertinib resistance in LUAD. We demonstrated that SLC25A39 knockdown significantly suppresses LUAD progression and effectively reverses cellular resistance to Osimertinib. Mechanistically, we revealed that SLC25A39 promotes ferroptosis resistance by activating the Nrf2 pathway, which enhances GSH biosynthesis and lipid peroxide detoxification. Furthermore, the combination of Osimertinib with ferroptosis inducers further enhances antitumor efficacy, exhibiting particularly greater therapeutic potential in LUAD patients with high SLC25A39 expression. These findings establish the SLC25A39–NRF2 axis as a novel regulatory pathway underlying ferroptosis-related drug resistance. Overall, this study not only highlights SLC25A39 as a promising therapeutic target in LUAD but also provides new mechanistic insights and strategies for the personalized treatment of Osimertinib-resistant LUAD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

Changyu Liu designed the research project. Zhilin Zeng, Wenfeng Wang, Yijie Gong, and Kai Fu performed the experimental operations. Wei Wu, Xiangning Fu, and Kai Fu conducted the data analysis. Zhilin Zeng, Changyu Liu, and Kai Fu drafted the manuscript. Yijie Gong and Yixin Cai verified the data and reviewed the manuscript. All authors have read and approved the final version of the manuscript, and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

This work was supported by the National Natural Science Foundation of China - Young Scientists Fund (82102796, 22174049), the Hubei Provincial Health and Science Technology Project - Young Talent Program (2025Q063), and the Huai’er Special Fund for Cancer Prevention and Treatment Research (CXPJJH124009-090).

Data availability

The data analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Consent for publication

Not applicable.

Consent to Participate

Not applicable.

Ethics Approval

This study was approved by the Institutional Animal Care and Use Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (TJH-202209010), and Institutional Ethics Committee of Huazhong University of Science and Technology (TJ-IRB202502107).

Competing interests

The authors report no declarations of interest.