ATM inhibition restores IFN-γ sensitivity and induces ferroptosis in NSCLC via DNA damage response.

1
Introduction
Although cancer immunotherapy has demonstrated promising clinical success in many tumor types [1,2], durable tumor control is achieved in only a fraction of patients, and both primary and acquired resistance to immunotherapy remain major challenges [2,3]. Interferon-γ (IFN-γ) is a key effector cytokine in anti-tumor immunity, produced by activated CD8+ cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells. IFN-γ exerts direct anti-proliferative and pro-apoptotic effects on tumor cells while also upregulating antigen presentation machinery to enhance immune recognition [4,5]. Paradoxically, tumor cells often develop mechanisms to evade these anti-tumor effects of IFN-γ. One major mechanism of IFN-γ resistance in tumor cells is the acquisition of mutations in the IFN-γ signaling pathway, such as loss-of-function mutations in JAK1/2 or STAT1 [3]. IFN-γ also induces PD-L1 expression on tumor cells to evade T cell-mediated elimination [3,6].
In addition to signaling defects in tumor cells in response to IFN-γ, recent studies have highlighted the role of DNA damage repair pathways in IFN-γ resistance. We previously reported that IFN-γ-producing antigen-specific CTLs in vivo resulted in copy-number alterations (CNAs) associated with the DNA damage response and modulation of DNA editing/repair gene expression [7]. These results suggest that enhanced genetic instability might be one of the mechanisms by which IFN-γ induces genetic evolution in tumor cells. In line with these findings, it has been reported that tumor cells with high activity in the DNA double-strand break (DSB) repair pathway tend to be resistant to IFN-γ. Inhibition of the DSB repair pathway, however, sensitizes these tumor cells to IFN-γ both in vitro and in vivo [8]. We previously found that immune-escape melanoma variants acquired resistance to the IFN-γ-induced oxidative stress response by upregulating the redox enzyme glutathione-S-transferase-4 (GSTA4) [9], suggesting that cancer cells escape IFN-γ-dependent host immunity by acquiring resistance to oxidative stress responses. Collectively, this evidence suggests that genotoxic and oxidative stress induced by IFN-γ is important for its anti-tumor function, and that DNA repair and/or redox pathways might rescue tumor cells from the IFN-γ-dependent immune-editing process.
Regarding the direct anti-tumor mechanism, it is well established that IFN-γ exhibits cytotoxic effects on tumor cells by inducing cell cycle arrest and/or apoptotic cell death [10,11]. Furthermore, an alternative form of programmed cell death, ferroptosis, is involved in IFN-γ-dependent anti-tumor immunity mediated by CD8+ T cells [5]. Ferroptosis is a non-apoptotic, iron-dependent form of programmed cell death characterized by the accumulation of lipid peroxides and dysregulation of glutathione-dependent antioxidant mechanisms, primarily due to the loss of glutathione peroxidase 4 (GPX4) activity [12]. IFN-γ downregulates the cystine/glutamate antiporter system Xc, composed of SLC7A11 and SLC3A2 subunits, in tumor cells. This downregulation limits cystine uptake, thereby impairing glutathione synthesis and leading to ferroptosis [5]. IFN-γ also upregulates acyl-CoA synthetase long-chain family member 4 (ACSL4) via IRF1, which promotes the incorporation of polyunsaturated fatty acids into cellular phospholipids [5,13]. This evidence suggests that the vulnerability of tumor cells to IFN-γ may be linked to their ability to manage genotoxic stress induced by lipid peroxidation.
In this study, we explored the mechanisms regulating IFN-γ responsiveness in the human NSCLC cell lines PC-9 and A549. PC-9 cells were more resistant to IFN-γ treatment than A549 cells, and transcriptomic analysis revealed that IFN-γ–resistant PC-9 cells were enriched for genes associated with the homologous recombination (HR) DNA repair pathway in response to IFN-γ. Given the critical role of the serine/threonine kinase ataxia telangiectasia mutated (ATM) in sensing DNA double-strand breaks and regulating their repair through HR [14], we investigated whether ATM contributes to IFN-γ responsiveness in NSCLCs by using the ATM inhibitor KU-55933.

2
Material and methods
2.1
Reagents
IFN-γ was purchased from Biolegend, and ATM inhibitor KU-55933 was purchased from MedChem Express. Ferrostatin-1 and Liproxstatin-1 were purchased from Cayman, and the GSH/GSSG Quantification Kit and Cell Counting WST-8 were purchased from Dojindo. γH2AX antibody and β-Actin were purchased from Cell Signaling Technology.

2.2
Cells
The human non-small cell lung cancer cell lines PC-9 (kindly gifted from Dr. Kiura, Okayama University, Japan) and A549 (CCL-185, obtained from American Type Culture Collection) were used. Cells were maintained in RPMI 1640 medium supplemented with 10% FBS (Gibco), 100 U/mL Penicillin, and other additives at 37 °C in a humidified 5% CO2 incubator as previously described [9]. All experiments were performed on mycoplasma-free cells in the linear growth phase.

2.3
Cell viability assay
Cells were seeded in a 96-well plate. After a 4-h incubation, the cells were treated with IFN-γ and the ATM inhibitor for 24 h. Cell viability was measured using the WST-8 assay to quantify metabolically active cells as previously described [15]. The absorbance was measured in a microplate reader at 450/620 nm. Data was averaged from three independent experiments.

2.4
Western blotting
Cells were treated with IFN-γ and KU-55933 for 24 h. The protein lysates were prepared in whole-cell lysis buffer. An equal amount of cell lysates was subjected to SDS-PAGE electrophoresis and then electrophoretically transferred to Immobilon-P nylon membranes, as previously described [9]. The membrane was probed with the indicated primary antibodies overnight at 4 °C, followed by incubation with horseradish peroxidase-conjugated secondary antibodies. An ECL substrate was used to visualize the bands.

2.5
Glutathione assay
Intracellular glutathione levels were measured using the commercial GSH/GSSG Quantification Kit (Dojindo) according to the manufacturer's instructions. Cells were treated with IFN-γ and KU-55933 for 24 h, and total glutathione, GSSG, and reduced GSH levels were quantified as per live cells.

2.6
Statistical analysis
Results are presented as the mean ± SD of at least two independent experiments. Statistical significance was measured by Welch's one-way ANOVA followed by Dunnett's T3 multiple comparisons test. P values < 0.05 were considered to show statistical significance.

3
Results
3.1
IFN-γ responsiveness correlates with DNA damage and repair responses in NSCLC cell lines
To explore the mechanisms underlying IFN-γ responsiveness in human NSCLC cell lines, we treated the PC-9 and A549 cell lines with recombinant human IFN-γ. Although the viability of A549 cells was inhibited by IFN-γ in a dose-dependent manner, PC-9 cells exhibited resistance to IFN-γ treatment (Fig. 1A). To further investigate the molecular mechanisms underlying IFN-γ resistance in PC-9 cells, we reanalyzed the GSE180942 dataset, which contains gene expression profiles of a variety of cancer cell lines, including PC-9 and A549, following IFN-γ treatment. Reactome analysis was performed using differential CRISPR β-scores under IFN-γ treatment (Δβ = PC-9 -A549). Bars indicate the normalized enrichment score (NES) for each pathway. Positive NES denotes pathways whose constituent genes are more essential in A549, whereas negative NES denotes pathways more essential in PC-9 upon IFN-γ treatment. The analysis reveals enrichment of DNA damage response/repair pathways in PC-9, such as double-strand break repair and homologous recombination–related pathways, whereas A549 shows enrichment of stress/innate immune-associated pathways, such as pyroptosis and antimicrobial peptides (Fig. 1B). These findings suggest the involvement of DNA repair pathways in response to IFN-γ-induced DNA damage in PC-9 cells.

3.2
Inhibition of ATM restores NSCLC cells to IFN-γ by inducing DNA damage response
Given the importance of ATM in DNA damage and repair responses, we next examined the involvement of ATM in IFN-γ responsiveness in NSCLC cells using an ATM inhibitor, KU-55933. In combination with IFN-γ, KU-55933 further reduced cell viability in IFN-γ-sensitive A549 cells (Fig. 2A). Furthermore, KU-55933, combined with IFN-γ, significantly reduced the viability of PC-9 cells (Fig. 2A), suggesting that ATM inhibition sensitizes NSCLC cells to IFN-γ. To further investigate whether ATM inhibition in combination with IFN-γ induces a DNA damage response, we examined γH2AX expression in both A549 and PC-9 cells. In IFN-γ-sensitive A549 cells, either IFN-γ or KU-55933 treatment alone slightly increased γH2AX expression, and their combination further increased γH2AX expression (Fig. 2B). Consistent with their resistance to IFN-γ, PC-9 cells showed no detectable γH2AX expression with either IFN-γ or KU-55933 treatment alone; however, their combination significantly increased γH2AX expression (Fig. 2B). These results indicate that ATM inhibition sensitizes NSCLC cells to IFN-γ by inducing a DNA damage response.

3.3
Inhibition of ATM in combination with IFN-γ induces ferroptosis in NSCLCs
To understand how ATM inhibition works synergistically with IFN-γ to exert an anti-tumor effect, we explored the cell death pathway involved in this combination. Previous studies have shown that ferroptosis plays a role in IFN-γ-dependent anti-tumor immune responses [16]. Importantly, the DNA damage response also plays a role in ferroptosis induction [17]. Therefore, we investigated whether ferroptosis contributes to the anti-tumor effect of ATM inhibition combined with IFN-γ using two ferroptosis inhibitors with distinct mechanisms, ferrostatin-1 (Fer-1) and liproxstatin-1 (Lip-1). In IFN-γ-sensitive A549 cells, either Fer-1 or Lip-1 treatment significantly attenuated the anti-tumor effect of ATM inhibition combined with IFN-γ (Fig. 3A). Similarly, the anti-tumor effect of ATM inhibition combined with IFN-γ in PC-9 cells was completely abolished in the presence of either Fer-1 or Lip-1 (Fig. 3B), suggesting that ATM inhibition combined with IFN-γ induces ferroptosis to exert its anti-tumor effect.

3.4
ATM inhibition in combination with IFN-γ alters glutathione metabolism in NSCLCs
To explore the mechanism by which ATM inhibition combined with IFN-γ induces ferroptosis in NSCLCs, we first measured lipid peroxidation levels in PC-9 cells treated with IFN-γ and KU-55933. Although the combination of IFN-γ and KU-55933 induced ferroptosis in PC-9 cells, there was no significant increase in lipid peroxidation (Supplementary Fig. 1). Instead, we observed a significant change in intracellular glutathione levels in PC-9 cells following combined IFN-γ and KU-55933 treatment. As shown in Fig. 4, a significant reduction in total glutathione (Fig. 4A), reduced glutathione GSH (Fig. 4B), and an increase in oxidized glutathione GSSG (Fig. 4C) were observed in response to IFN-γ combined with KU-55933. As a consequence of these changes in glutathione levels, the GSH/GSSG ratio, which reflects vulnerability to oxidative stress responses, including ferroptosis, was significantly reduced in PC-9 cells treated with IFN-γ combined with KU-55933. These results suggest that glutathione metabolism is involved in the induction of ferroptosis in NSCLCs by IFN-γ combined with an ATM inhibitor.

4
Discussion
In this study, we demonstrated that ATM inhibition overcomes resistance to IFN-γ by inducing ferroptosis in NSCLCs. Mechanistically, the combination of IFN-γ and ATM inhibition triggered a DNA damage response and altered glutathione metabolism, resulting in a reduced GSH/GSSG ratio and increased vulnerability to oxidative stress, thereby promoting ferroptosis. These findings highlight the critical role of the DNA damage response in the anti-tumor effect of IFN-γ.
Consistent with our findings, evidence suggests that DNA repair pathways influence the responsiveness of tumor cells to anti-tumor immunity. Han et al. recently reported that activation of DSB repair pathways confers resistance to IFN-γ, whereas inhibition of DSB repair synergizes with IFN-γ to kill tumor cells [8]. Given that ATM is a central coordinator of the DSB repair response, IFN-γ-induced DNA damage may not be repaired when ATM is inhibited. Moreover, ATM functions as a redox sensor, promoting antioxidant gene expression and NADPH production to counteract ROS [14]. Considering induced ferroptosis in PC-9 cells, there was no significant increase in lipid peroxidation in the combination of IFN-γ and KU-55933 (Supplementary Fig. 1), we speculate that ATM inhibition combined with IFN-γ reduces the GSH/GSSG ratio in NSCLCs, thereby increasing susceptibility to oxidative stress and ferroptosis.
Regarding ferroptosis in anti-tumor immunity, IFN-γ has been reported to downregulate system Xc− in tumor cells, limiting cystine uptake for glutathione synthesis, and to upregulate ACSL4, increasing polyunsaturated fatty acid loading and creating a ferroptosis-prone state [5,13]. Although we did not observe changes in SLC7A11 expression or lipid peroxidation in NSCLCs, IFN-γ may downregulate cystine uptake when combined with ATM inhibition [18]. While ATM inhibitors have been clinically tested as radio- or chemosensitizers [19], and their potential in cancer immunotherapy remains unexplored. Interestingly, recent clinical data suggest that loss-of-function ATM mutations correlate with improved responses to immune checkpoint blockade therapy. Patients with ATM truncations showed better outcomes with anti-PD-1 or anti-PD-L1 treatment, associated with cGAS/STING pathway activation and an inflamed tumor microenvironment [20].
Building on these observations, our findings suggest that ATM inhibition could enhance cancer immunotherapy by sensitizing tumor cells to IFN-γ-induced DNA damage and ferroptosis. However, limitations remain. We used exogenous IFN-γ, so it is unclear whether ATM inhibition sensitizes tumor cells to immune cell-mediated cytotoxicity. It is also critical to use other ATM inhibitors or employ genetic modification of ATM to exclude the possibility of the off-target effect of KU-55933 other than ATM inhibition. Contrary to our present findings, it was reported by Chen and colleagues that ATM inhibition protects several cancer cells from ferroptosis and such protective effect of ATM against ferroptosis can be mediated by the modification of iron transport [21]. Such discrepancy between two studies might be a result of cell types tested, pathway involved in triggering ferroptosis, or other unknown mechanisms. The relevance of our findings to other tumor types also requires investigation, as DNA repair capacity varies across cancers and may influence immunotherapy resistance. Nevertheless, our results indicate that ATM is a key molecular target for regulating tumor cell responsiveness to IFN-γ. Pharmacological ATM inhibition sensitizes NSCLCs to IFN-γ by inducing DNA damage and ferroptosis.

Funding sources
This study was supported by a Grant-in-Aid for Scientific Research (21H02783, 24KK0103 25K02528), The 10.13039/501100001700Ministry of Education, Culture, Sports, Science and Technology (10.13039/501100001700MEXT), Japan (YH), and the Director Leadership Expenses, 10.13039/100016571Institute of Natural Medicine, University of Toyama.

CRediT authorship contribution statement
Muhammad Irshad Farooq: Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Sisca Ucche: Formal analysis, Investigation, Methodology, Writing – original draft. Mariho Uozumi: Formal analysis, Investigation, Methodology, Writing – review & editing. Sana Jabbar: Formal analysis, Investigation, Methodology, Writing – review & editing. Suthasinee Seephan: Formal analysis, Investigation, Methodology, Writing – review & editing. So-Ichiro Sasaki: Formal analysis, Investigation, Methodology, Writing – review & editing. Yoshihiro Hayakawa: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Yoshihiro Hayakawa reports financial support was provided by Government of Japan Ministry of Education Culture Sports Science and Technology. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.