ATR Blockade Potentiates the Effects of Genotoxic Agents In Vitro and Promotes Antitumor Immunity in a Mouse Model of Non-Small Cell Lung Cancer

TL;DR

Blocking ATR in lung cancer cells boosts the effects of DNA-damaging drugs and improves immune response in mice, suggesting a new treatment approach.

Contribution

Combining ATR inhibitors, genotoxic drugs, and immune checkpoint inhibitors enhances antitumor effects in non-small cell lung cancer.

Findings

ATR inhibition with AZD6738 increases the cytotoxicity of genotoxic agents in NSCLC cells.

Combining ATR inhibitors, cisplatin, and anti-PD1 boosts T cell infiltration and antitumor efficacy in mice.

Triple therapy with AZD6738, cisplatin, and anti-PD1 achieves the strongest antitumor effect.

Abstract

The ataxia-telangiectasia mutated and Rad3-related (ATR) kinase plays a crucial role in sensing and responding to DNA damage and replication stress by coordinating cell cycle arrest, promoting DNA repair, and stabilizing replication forks. Previous studies have shown that ATR blockade can sensitize cancer cells to genotoxic agents by disrupting the DNA damage response network and by inducing immunogenicity. In this study, we found that the combination of the ATR inhibitor AZD6738 and genotoxic agents decreased the viability of non-small cell lung cancer (NSCLC) cells in vitro and increased the antitumor efficacy of an immune checkpoint inhibitor in a mouse NSCLC model. These preclinical results suggest that combining ATR inhibitors, genotoxic drugs, and immune checkpoint inhibitors may lead to the development of a new cancer treatment strategy for NSCLC.

Background/Objectives: Non-small cell lung cancer (NSCLC) is the most frequent type of lung cancer, and its main treatments include chemotherapy with genotoxic drugs and immunotherapy. Central to the cellular response to genotoxic stress is the DNA damage response (DDR) network, regulated by key kinases such as ataxia-telangiectasia mutated and Rad3-related (ATR). Herein, we tested the hypothesis that inhibition of ATR enhances the cytotoxicity of genotoxic agents and the antitumor immune response. Methods: DDR-related parameters and redox status, expressed as GSH/GSSG ratio, and apurinic/apyrimidinic lesions, were evaluated in human (A549, H1299) and murine (LLC) NSCLC cell lines after co-exposure to ATR inhibitor (AZD6738) and ultraviolet C (UVC) irradiation or cisplatin. Using a syngeneic LLC model, treatments of AZD6738 alone or in combination with cisplatin and/or anti-programmed cell death 1 antibody (anti-PD1) were examined. Results: In all cell lines, combined treatment with AZD6738 and cisplatin or UVC irradiation markedly decreased cell viability, DNA repair efficiency, and GSH/GSSG ratios; increased drug-induced DNA damage; and augmented apurinic/apyrimidinic lesions. In vivo, following treatment with AZD6738 and cisplatin, flow cytometry analysis performed in tumor cells revealed an increased infiltration of CD3+ and CD8+ T cells, with the triple combination of AZD6738, cisplatin, and anti-PD1 achieving the strongest antitumor effect. The CD3+CD4−CD8− double-negative (DN) T cell population in tumor samples also emerged as a contributing factor in this context. Conclusions: These results demonstrate that ATR blockade concurrently enhances the efficacy of genotoxic agents and immune checkpoint inhibitors, thus paving the way for combination therapies in NSCLC.

Figures

Full text

1. Introduction

Non-small cell lung cancer (NSCLC) accounts for the majority of lung cancer cases and remains a leading cause of cancer-related mortality worldwide [1]. Despite advances in targeted therapies and immunotherapy, conventional genotoxic treatments, such as platinum-based chemotherapy and radiation, remain central to NSCLC management [2]. However, their clinical efficacy is often limited by tumor resistance mechanisms and systemic toxicity, highlighting the need for strategies that enhance tumor sensitivity while minimizing adverse effects.

The DNA damage response (DDR) plays a critical role in tumor cell survival following genotoxic stress [3]. Ataxia telangiectasia and Rad3-related kinase (ATR) is a key DDR regulator that senses replication stress and coordinates cell cycle checkpoints and DNA repair [4]. By protecting cancer cells from DNA-damaging agents, ATR contributes to therapeutic resistance [5]. Preclinical studies have demonstrated that ATR inhibition can sensitize tumor cells to chemotherapeutics and radiation, suggesting a potential strategy to improve the efficacy of conventional genotoxic treatments [6,7,8,9,10]. ATR inhibition with AZD6738 synergizes with cisplatin or radiation to enhance cytotoxicity, and co-treatment with cisplatin promotes rapid regression of ataxia–telangiectasia mutated (ATM)-deficient NSCLC xenografts [9]. Furthermore, anti–PD-1 therapy shows improved antitumor activity when combined with genotoxic agents and ATR inhibition in various solid tumor models [6,11]. Clinical development of DDR inhibitors has also emerged as a promising strategy to sensitize tumors to these agents. Indeed, early-phase clinical data with the orally bioavailable ATR inhibitor ceralasertib (AZD6738) have demonstrated tolerability and signs of durable responses in patients with solid tumors characterized by genomic instability and inflammation [12].

Beyond direct cytotoxicity, accumulating evidence indicates that DNA damage can trigger antitumor immune responses through pathways such as cGAS-STING and type I interferon signaling [6,13,14] and can synergize with immune checkpoint blockade by reshaping the tumor immune microenvironment [15,16]. ATR inhibitors contribute to cytosolic DNA accumulation and micronuclei formation, triggering innate immune cGAS-STING signaling and pro-inflammatory chemokine expression, thereby recruiting immune effector cells. ATR targeting also increases major histocompatibility complex class I (MHC-I) expression, thereby enhancing antigen presentation and promoting recognition by cytotoxic T lymphocytes strengthening antitumor immunity, particularly when combined with PD-1/PD-L1 blockade [6,11,14]. ATR inhibitors can also reduce DNA damage-induced PD-L1 expression, destabilizing PD-L1 via proteasomal degradation, thus restoring the susceptibility of tumor cells to T-cell-mediated killing [17]. Indeed, further findings support this synergy demonstrating that ATR inhibitors induce CD8^+^ cytotoxic T-cell-mediated antitumor responses, while CD8^+^ depletion abolishes this effect [11,14]. These mechanistic and preclinical data underscore the immunomodulatory role of ATR inhibition and have spurred significant interest in combining ATR inhibitors with PD-1/PD-L1 blockade. Indeed, this dual blockade represents a promising therapeutic strategy, but important challenges remain—including identifying predictive biomarkers for patients most likely to benefit [18].

Several clinical trials are now evaluating the combination of ATR inhibitors and immune checkpoint blockade in NSCLC and other lung malignancies [19]. One notable study is the Phase II “HUDSON” umbrella trial (NCT03334617), which tested durvalumab (anti–PD-L1) together with ceralasertib in patients with immunotherapy-resistant lung cancer. While this early-phase study suggested enhanced responses in patients with ATM-altered tumors, these findings are preliminary and hypothesis-generating, given the small cohort size, heterogeneous patient population, and exploratory biomarker analyses. These results primarily support the biological rationale for ATR inhibition to re-sensitize immunotherapy-resistant NSCLC through immune modulation and highlight the need for ongoing and planned Phase III trials to formally evaluate clinical efficacy [20].

Oxidative stress and defective processing of DNA lesions further contribute to the vulnerability of tumor cells exposed to ATR blockade. Perturbations in the cellular redox state, commonly assessed by the glutathione ratio, that is the ratio of reduced (GSH) to oxidized glutathione (GSSG), influence nucleotide excision repair (NER) capacity: glutathione depletion has been shown to modulate expression of key NER genes such as ERCC1 and to impair functional NER in oxidative-stress conditions [21]. Moreover, ATR deficiency or pharmacological inhibition has been shown to drive reactive oxygen species (ROS) hyperproduction, particularly in mitochondria, suggesting that ATR plays a protective role against oxidative stress [22]. Oxidative conditions also increase the formation of apurinic/apyrimidinic (abasic; AP) sites—major cytotoxic intermediates generated during base excision repair (BER)—and excessive accumulation of AP sites can impede replication fork progression [23,24]. While BER is the canonical pathway for AP site removal, some AP lesions require NER for resolution [25]. Emerging evidence further highlights that persistent AP sites may lead to more complex damage such as interstrand crosslinks (ICLs), amplifying replication stress [24]. ATR normally stabilizes stressed forks and facilitates the repair of bulky lesions and ICLs by coordinating repair factors at sites of damage [26]. Therefore, ATR inhibition is expected to exacerbate oxidative DNA damage, elevate AP site burden, and impair NER and ICL repair capacity. Moreover, ATR inhibition directly contributes to the generation of AP sites, inducing origin firing that depletes replication protein A (RPA) and leads to uracil DNA glycosylase (UNG2)-mediated base excision [27]. Combined, these insights underscore how redox imbalance and AP site accumulation can compromise lesion repair, and how ATR inhibition—by augmenting these stresses—could impair both NER and ICL repair capacity potentiating the effects of genotoxic agents.

Based on recent comprehensive meta-analyses and clinical trial reviews (up to September 2025), ATR inhibitor monotherapy has shown specialized, though not universal, efficacy in NSCLC. While general pooled analyses of all tumor types showed no significant improvement in response rates compared to standard therapies, stratified analysis revealed marked benefits in specific NSCLC subgroups, notably those with high replication stress, such as KRAS mutations or STK11 (serine/threonine kinase 11)/KEAP1 (Kelch-like ECH-associated protein 1) loss [28]. Interestingly, the efficacy in NSCLC is driven by ATR dependency. Tumors with high oncogene-induced replication stress [e.g., KRAS-mutant, STK11/LKB1 (liver kinase B1) loss, or KEAP1 loss] depend on the ATR-CHK1 pathway to prevent replication fork collapse. Preclinical and early-phase clinical data suggest that in these high-stress contexts, ATR inhibitors can trigger mitotic catastrophe as single agents. While monotherapy shows potential, ATR inhibitors (e.g., ceralasertib/AZD6738, berzosertib/M6620) have shown stronger, synergistic activity in combination with platinum-based chemotherapy or immune checkpoint inhibitors in NSCLC [29,30].

Collectively, ATR blockade may enhance tumor cell killing and increase tumor immunogenicity, providing a rationale for combining ATR inhibitors with chemo-immunotherapy. Nevertheless, the simultaneous impact of ATR inhibition on both genotoxic sensitivity and anti-tumor immunity in NSCLC remains poorly characterized. Here, we tested the hypothesis that ATR blockade both potentiates the effects of genotoxic agents in NSCLC cell lines and enhances anti-tumor immunity in a NSCLC mouse model, supporting a dual mechanism of therapeutic benefit. Combining two of the most widely used human NSCLC cell lines in pre-clinical oncology research (A549 and H1299 cells) with a syngeneic, immunocompetent mouse model of NSCLC [LLC-Ova (Lewis lung carcinoma cells expressing ovalbumin) into C57BL/6 mice] provides a robust, two-tiered preclinical platform that allows assessment of both tumor-intrinsic effects (DDR disruption, genotoxic sensitization) and tumor–immune interactions. Finally, the integration of cell line screening data with responses observed in mouse models may facilitate the identification of biomarkers of drug sensitivity or resistance, thereby supporting patient stratification for clinical trials.

2. Materials and Methods

2.1. Cell Lines

Human NSCLC A549 cells (generously offered by Dr. Panagiotis Georgiadis, National Hellenic Research Foundation, Athens, Greece) were cultured in a 1:1 mixture of DMEM and Ham’s F12 medium, supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Human epithelial-like NSCLC H1299 cells (donated by Prof. Athanassios Kotsinas, National and Kapodistrian University of Athens, Athens, Greece) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS and 1% penicillin–streptomycin. Mouse LLC-Ova cells (Lewis lung carcinoma expressing ovalbumin—kindly provided by Prof. Athanassios Kotsinas, National and Kapodistrian University of Athens, Athens, Greece), classified as NSCLC [31], were grown in DMEM supplemented with 10% FBS, 1% non-essential amino acids, and 1% penicillin–streptomycin. The partner laboratories perform routine cell line authentication and mycoplasma testing prior to distribution.

2.2. Drugs

For all experiments, cisplatin was purchased from Hospira (Pfizer, New York, NY, USA). For in vitro studies, ceralasertib (AZD6738) was purchased from Selleckchem (Houston, TX, USA, #S7693) and diluted in DMSO, according to the company’s instructions. Final drug concentration in culture was 1µM, corresponding to a DMSO concentration of 0.01%, which is far below levels reported to affect cell viability or function. Therefore, a separate DMSO-only control was not included. For the in vivo studies, AZD6738 powder was generously donated by AstraZeneca (Cambridge, UK) and was prepared according to manufacturer’s guidelines. Specifically, the drug was first dissolved in DMSO to 10% the final volume, sonicated until complete solubilization, and then added 40% of the final volume in propylene glycol, and 50% of the final volume with sterile deionized water. Anti-mouse PD1 antibody [RMP1-14] was purchased from Assay Genie (Dublin, Ireland, #IVMB0037).

2.3. Viability Assay

Drug-induced cytotoxicity and cell proliferation were assessed using the sulforhodamine B (SRB) assay [32]. Cells were subjected to the appropriate treatment, followed by fixation with 10% trichloroacetic acid and staining with 0.4% SRB in 1% acetic acid. Absorbance was measured using a microplate reader (TECAN, Männedorf, Switzerland) to estimate cell viability.

2.4. Measurement of Nucleotide Excision Repair (NER)-Alkaline Comet Assay

To evaluate NER capacity, cells were exposed to UVC irradiation (100 J/m^2^), allowed to recover in culture medium for 0–6 h at 37 °C, then collected and preserved in freezing medium at −80 °C for up to 1 month prior to analysis. Alkaline comet assay was performed to measure DNA damage as thoroughly described in our prior study [33]. In short, 10^4^ cells were put onto glass slides after being suspended in PBS (phosphate-buffered saline, pH 7.4) and 1% low-melting-point agarose. After allowing the slides to solidify for 30 min at 4 °C, they were incubated in an alkaline lysis buffer (0.01 M Tris, pH 10, 0.1 M EDTA, 2.5 M NaCl, containing 1% Triton X-100) for two hours at 4 °C before being electrophoresed for thirty minutes at 25 V and 225 mA. SYBR™ Gold nucleic acid gel stain (Thermo Fisher Scientific, Waltham, MA, USA, #S11494) was then applied to the slides, and a 10× microscopy lens was used to image them under UV light. Comet assay parameters (Olive Tail Moment, OTM) were quantified using ImageJ (version 1.46r) with the OpenComet plugin v1.3.1 (https://cometbio.org/ (accessed on 21 January 2024)). For each sample, 2 gels were scored, and the average OTM value of 150 cells was calculated.

2.5. Measurement of Gene-Specific Repair of the Interstrand Cross-Links (ICL)

Cells were treated with cisplatin (5 μΜ) for 3 h at 37 °C, followed by incubation in drug-free medium for 0–24 h, then harvested and stored in freezing medium at −80 °C for up to 1 month prior to analysis. Gene-specific ICL repair was assessed by Southern blot, as previously described [34], and N-ras alkylation analysis was performed following the protocol outlined in our recent study [33]. Briefly, following the isolation of DNA, genomic DNA was completely digested with the restriction enzyme EcoRI, and DNA was denatured before gel electrophoresis and Southern blotting. In order to detect alkylations in the N-ras gene, hybridizations were carried out as previously mentioned [35]. A 112 bp fragment of the human N-ras gene served as the N-ras-specific probe, which was amplified using earlier instructions [36]. The primer pair utilized was as follows: forward, 5′-GTT-ATA-GATGGT-GAA-ACC-TG-3′; reverse, 5′-ATA-CAC-AGA-GGAAGC-CTT-CG-3′. Gene fragments with DNA interstrand crosslinks re-anneal in the gel and move as double-stranded DNA under the conditions used. Uncrosslinked material moves as single-stranded DNA and stays denatured. The interstrand cross-link frequency was calculated directly from the denatured samples.

2.6. GSH/GSSG Ratio and Apurinic/Apyrimidinic Lesions (Abasic; AP-Sites)

The ratio of reduced to oxidized glutathione (GSH/GSSG) was measured using the luminescence-based GSH/GSSG-Glo Assay from Promega (Madison, WI, USA, V6612), following the manufacturer’s instructions, allowing quantification of total glutathione (GSH + GSSG), GSSG, and the GSH/GSSG ratio. In short, 10^4^ cells were plated in a 96-well tissue culture plate (Corning Costar) that was compatible with a luminometer. The Luciferin Generation Reagent was then added in 50 μL/well, shaken briefly, and incubated at room temperature for 30 min. Following the addition of 100 μL/well of Luciferin Detection Reagent, the luminescence signal was measured in a Spectramax M3 microplate reader after 15 min of incubation (Molecular Devices LLC, San Jose, CA, USA).

Abasic sites were quantified using the OxiSelect Oxidative DNA Damage Quantitation Kit (Cell Biolabs, San Diego, CA, USA; #STA-324) according to the manufacturer’s protocol. This assay kit employs an Aldehyde Reactive Probe to specifically react with an aldehyde group on the open ring form of the apurinic/apyrimidinic sites. By doing this, the apurinic/apyrimidinic sites can be marked with biotin, which is later detected with the streptavidin–enzyme conjugate. By comparing the absorbance of the unknown DNA sample with a standard curve created from the provided DNA standard containing predetermined apurinic/apyrimidinic sites, the number of these lesions in the sample is ascertained.

2.7. In Vivo Experiments

LLC-Ova cells (5 × 10^5^) were suspended in PBS:Matrigel (3:1, BD Biosciences, Franklin Lakes, NJ, USA, #356234) and injected subcutaneously into the right flank of 6–8-week-old C57BL/6 mice. Tumor growth was measured with calipers (Mitutoyo Corporation, Kawasaki, Japan), and volume was calculated as V = (W^2^ × L)/2, where W is width and L is length. When tumors reached 150 mm^3^, mice were randomized into groups of six and treatments commenced. Mice were assigned to treatment groups using a randomization procedure designed to balance sex distribution and baseline tumor volumes across groups. One group received no treatment and served as the control. Body weight and tumor growth were monitored, with euthanasia performed if tumors reached 2 cm^3^. To ensure blinding, all tumor measurements were performed by a colleague uninvolved in group assignment. Cisplatin was administered intraperitoneally (5 mg/kg, weekly, two doses), AZD6738 orally (50 mg/kg, daily, for 3 or 14 days), and anti–PD1 intraperitoneally (10 mg/kg, every other day, one or two cycles). For the first experiment, the groups were treated as shown in Table 1.

An a priori power analysis with GPower software (version 3.1.9.4) [37], as noted in Algorithm 1, indicated a required total sample size of 36 animals (n = 9 per group across four groups). Algorithm 1. GPower analysis output for experiment in Table 1F tests-ANOVA: Fixed effects, omnibus, one-wayAnalysis: A priori: Compute required sample size Input: Effect size f = 0.6 α err prob = 0.05Power (1β err prob) = 0.8Number of groups = 4Output: Noncentrality parameter λ = 12.9600000Critical F = 2.9011196Numerator df = 3Denominator df = 32Total sample size = 36Actual power = 0.8214243

Due to ethical and practical constraints, six animals were ultimately included per group.

For the second experiment, the groups were treated as shown in Table 2 and power analysis is shown in Algorithm 2. Algorithm 2. G*Power analysis output for experiment in Table 2F tests-ANOVA: Fixed effects, omnibus, one-wayAnalysis: A priori: Compute required sample size Input: Effect size f = 0.6α err prob = 0.05Power (1-β err prob) = 0.9Number of groups = 12Output: Noncentrality parameter λ = 25.9200000Critical F = 1.9522119Numerator df = 11Denominator df = 60Total sample size = 72Actual power = 0.91496532

2.8. Tissue Dissociation

Tumors were excised from euthanized mice and dissociated using the gentleMACS Dissociator with the Mouse Tumor Dissociation Kit from Miltenyi Biotec (Bergisch Gladbach, Rhineland, Germany, #130-096-730,) according to the manufacturer’s instructions. Spleens were harvested and mechanically homogenized on a 70 μm cell strainer.

2.9. Flow Cytometry

Single-cell suspensions from tumors and spleens were centrifuged at 1200 rpm for 8 min at 4 °C and resuspended in PBS. Cells were stained with Fixable Viability Stain 780 (BD Biosciences, #565388) for 15 min at room temperature in the dark. After a wash with FACS buffer (1% FBS, 1% EDTA 0.3 M, pH 7.4 in PBS), pellets were resuspended in 100 µL of antibody cocktail. Conjugated antibodies were from BioLegend (San Diego, CA, USA, USA-PE anti-CD4, #100408BD) and BD Biosciences (FITC anti-CD45, #553080; PE-Cy5 anti-CD3ε, #555276; APC anti-CD8a, #553035; PE-CF594 anti-CD44, #562464) and diluted 1:100 in FACS buffer. Cells were incubated for 30 min at room temperature in the dark, fixed with 2% PFA (100 µL, 15 min at 4 °C), washed twice, and stored overnight at 4 °C. Prior to analysis, suspensions were filtered through 70 µm strainers and processed on a BD FACSAria™ II Cell Sorter (BD Biosciences, Franklin Lakes, NJ, USA). Data were analyzed using Beckman Coulter Kaluza Analysis v2.3.1 (Brea, CA, USA). The gating strategy for identifying T cell populations from single cell suspensions from tumors is presented in Figure S1. Flow cytometry analyses were conducted in a blinded manner by a colleague who was not involved in group assignment.

2.10. Statistical Analysis

Planned pairwise comparisons were conducted using unpaired two-tailed t-tests with Welch’s correction for multiple testing, with significance defined as p < 0.05. Data are presented as mean ± standard deviation (SD), and all analyses and graphical representations were conducted using GraphPad Prism v8.0.1.

3. Results

3.1. Impact of ATR Blockade on DDR-Associated Parameters in Lung Cancer Cell Lines

To investigate the effect of the combined treatment of the ATRi and cisplatin on DDR, we used two human NSCLC cell lines (A549, H1299) and the murine cell line LLC, which is classified as NSCLC [31]. Firstly, we selected the optimal concentrations of AZD6738 and cisplatin that, when administered alone, showed minimal decrease in cell viability. Using the SRB assay, we found that, after treatment with several doses of the AZD6738 (0.1, 0.5, 1, 2, 5 μM) for 72 h, or cisplatin (1, 5, 15 μM) for 3 h, the optimal concentrations were: 1 μΜ AZD6738 and 5 μΜ cisplatin (Figure S2A,B). Moreover, in all cell lines examined, combined treatment of 1 μΜ AZD6738 and 5 μΜ cisplatin had a greater effect on cell viability than administration of either compound alone (all p < 0.05; Figure S2C).

Next, cisplatin-induced DDR signals were evaluated following ATR inhibition. To assess the effect of ATR blockade on the ICL repair capacity, cells were treated with ceralasertib and cisplatin, either alone or in concurrent combination, and the ICL adducts were quantified. In all cell lines examined, ceralasertib alone did not substantially affect the ICL adduct levels (Figure 1A). Interestingly, in A549 and H1299 cells, combined treatment induced a significant increase in the ICL burden, compared with cisplatin-alone treatment (p < 0.05 and p < 0.01 respectively). In LLC cells, combination therapy also caused an increase in adduct levels, even though the increase was not significant. Moreover, in all cell lines, combination treatment resulted in a more robust reduction of the GSH/GSSG ratio, compared with the one caused by cisplatin alone (Figure 1B). In line with the GSH/GSSG ratio results, A549 cells showed a significant increase in the abasic lesions (p < 0.01; Figure 1C). Although H1299 and LLC cells also showed elevated abasic lesions after combined treatment, the increase was not significant. Importantly, AZD6738 treatment alone did not provoke any alterations in redox status or abasic site formation at any of the cell lines examined (Figure 1B,C). Together, combined treatment of ATRi with cisplatin decreased cell viability; reduced ICL repair efficiency; and increased drug-induced DNA damage, oxidative stress, and abasic lesions.

Next, the effect of ATR blockade on the efficiency of NER was estimated. To that end, A549, H1299, and LLC cell lines were pretreated with 1 μM AZD6738 for 72 h, then exposed to 100 J/m^2^ UVC, and cell viability was measured using the SRB assay. We found that combined treatment of UVC and AZD6738 had a higher impact on cell viability than administration of either agent alone (all p < 0.05; Figure S2D). Then, DNA lesions were measured up to 6 h post-irradiation via alkaline comet assay. In H1299 cells, ATR blockade showed a pronounced effect in DNA damage levels at all time-points analyzed, resulting in significantly higher UVC-induced DNA damage burden, expressed as the area under the curve (AUC) for DNA damage (p < 0.001; Figure 2A,B). On the other hand, in A549 and LLC cells, combined treatment did not differentiate the levels of UVC-induced lesions at all time-points examined (Figure S3A,B and Figure 2B). In addition, ATR inhibition followed by UVC exposure significantly reduced GSH/GSSG ratio and increased abasic sites in H1299 cells (p < 0.01; Figure 2C–E). In A549 and LLC cell lines, combined treatment had no significant effect on the GSH/GSSG ratio or the number of abasic sites (Figure S3C–F and Figure 2E). Together, ATR inhibition followed by UVC exposure decreased the viability of all tested NSCLC cell lines and markedly affected DDR-related parameters only in the p53-deficient H1299 cells.

3.2. Therapeutic Potential of ATR Inhibition in Lung Cancer

3.2.1. Combined ATR Inhibition and Cisplatin Chemotherapy In Vivo

We next evaluated the impact of ATR inhibition in vivo using a syngeneic heterotopic lung cancer model, generated by subcutaneous injection of LLC cells into C57BL/6 mice. This model was used to assess the antitumor efficacy of combined treatment with the ATR inhibitor AZD6738 and cisplatin, administered as shown in Figure 3A. No significant differences in tumor growth were observed among the experimental groups (Figure 3B,C). AZD6738 monotherapy had no effect compared to controls, while cisplatin moderately reduced tumor volume. Importantly, the combination of AZD6738 and cisplatin did not outperform cisplatin alone, yielding only a slight additional reduction in tumor growth (Figure 3B,C).

After treatment completion, tumors and spleens were collected, and single-cell suspensions were prepared for flow cytometry to evaluate T cell infiltration across treatment groups. CD3^+^ T cells increased with combination therapy compared to cisplatin alone (Figure 3D). CD4^+^ helper T cells were significantly elevated only in cisplatin-treated mice versus controls (Figure 3E). CD8^+^ cytotoxic T cells showed a significant increase with the combination therapy relative to either monotherapy, though not compared with controls (Figure 3F).

Flow cytometry of spleens showed elevated T cell levels, with cisplatin monotherapy producing the highest CD3^+^ counts and combination therapy the lowest (Figure S4A), consistent with tumor infiltration patterns (Figure 3D) and suggesting T cell redistribution toward tumors in the combination group. No significant changes were observed in CD4^+^ helper and CD8^+^ cytotoxic T cells (Figure S4B,C). Notably, CD8^+^ cells comprised ~37% of CD3^+^ populations in all spleen samples (Figure S4C), whereas only the combination-treated tumors achieved comparable CD8^+^ levels, indicating enhanced tumor infiltration of cytotoxic T cells with this regimen.

Overall, T lymphocytes represented a small fraction of immune cells within tumors, reaching a maximum of ~6% of CD45^+^ cells in the combination therapy group, whereas spleen T cells ranged from 5 to 17%, indicating peripheral activation but limited tumor infiltration. Within tumors, CD8^+^ T cells were less abundant than CD4^+^ T helper cells, and only a subset of cytotoxic CD8^+^ cells expressed the activation marker CD44. Given that the antitumor activity of ceralasertib is CD8^+^ T-cell-dependent [8,38,39] and short-course AZD6738 schedules enhance effector CD8^+^ induction [8], we next adjusted the inhibitor administration schedule in subsequent experiments.

3.2.2. In Vivo Evaluation of ATR Inhibition Combined with Chemotherapy and Immunotherapy

Subsequently, we evaluated whether ATR inhibition could enhance the efficacy of immune checkpoint blockade (ICB) in vivo. Utilizing the same murine lung cancer model, we tested the combination of AZD6738, cisplatin, and an anti-PD1 antibody, following the administration protocol shown in Figure 4A. No significant changes in body weight were observed, and overall growth remained normal. Tumor growth was tracked for 27 days (Figure 4B). Notably, on day 11, tumor progression temporarily halted in all groups receiving anti-PD1, including cisplatin + anti-PD1, AZD6738 + anti-PD1, and the triple combination (Figure 4B, red arrow). However, neither treatment cessation nor a second cycle of anti-PD1 maintained this effect, and tumors resumed growth after a brief arrest. The lowest tumor burden, as measured by the AUC [40] was observed in the group receiving the triple combination of cisplatin, AZD6738, and anti-PD1 (p < 0.05; Figure 4C).

Following treatment completion, tumors and spleens were collected, dissociated into single-cell suspensions, and analyzed by flow cytometry. CD3^+^ T cell distribution differed markedly between tumors and spleens (Figure 5A). Unexpectedly, the triple combination (cisplatin + AZD6738 + anti-PD1) showed reduced intratumoral T cell levels (Figure 5A). However, a second anti-PD1 cycle decreased splenic T cells, suggesting enhanced trafficking to the tumor (Figure S5A). Notably, the AZD6738 + cisplatin + anti-PD1 (x2) group exhibited the highest intratumoral CD3^+^ T cell infiltration, supporting synergistic immune activation (Figure 5A). Given the increased T cell infiltration in these groups, higher levels of CD4^+^ helper and/or CD8^+^ cytotoxic T cells were anticipated. However, analysis of these major T cell subsets revealed decreased proportions in tumor samples (Figure 5Β,C). In tumors, the highest CD4^+^ T cell levels were detected in groups treated with anti-PD1, AZD6738 + anti-PD1, and cisplatin + AZD6738 + anti-PD1 (Figure 5B). A second anti-PD1 cycle significantly reduced CD4^+^ T cells compared to a single dose in both monotherapy (p < 0.01) and triple treatment (p < 0.05). Similar declines were observed between combination regimens with one or two immunotherapy cycles. In spleens, CD4^+^ T cell levels remained relatively stable (~45–60%) across groups, with a modest increase following the second anti-PD1 dose (p < 0.01), opposite to the tumor trend (Figure S5B). Combination treatments did not significantly elevate CD4^+^ T cell levels relative to either monotherapy. Cytotoxic CD8^+^ T cells were generally less abundant in tumors than in spleens, suggesting that although some treatments enhanced peripheral activation, barriers to intratumoral infiltration or persistence remained (Figure 5C). Anti-PD1 monotherapy significantly increased CD8^+^ T cell levels compared to untreated controls, with cisplatin and its combinations showing similar but weaker effects. AZD6738 alone yielded the lowest CD8^+^ levels. A second anti-PD1 cycle reduced CD8^+^ frequencies in some combination groups, while all anti-PD1 (x2)-treated spleens showed significantly decreased CD8^+^ lymphocytes (Figure S5C).

Collectively, the data suggested the emergence of an unconventional double-negative (DN) T cell subset (CD3^+^CD4^−^CD8^−^), prompting further investigation. Flow cytometry revealed that DN T cells comprised ~30–80% of CD3^+^ T cells in tumors (Figure 5D), but only ~5–12% in spleens (Figure S5D). In tumors, AZD6738 and anti-PD1 (x2) induced the highest DN T cell levels (~60–80%). Combination treatments, including AZD6738 + cisplatin or the triple therapy, showed variable yet sometimes elevated DN T cell proportions. The marked tumor–spleen discrepancy suggests that DN T cell expansion is tumor-specific, potentially driven by AZD6738 and/or anti-PD1–mediated immune modulation.

4. Discussion

The ATR kinase pathway is essential for responding to DNA damage and replication stress, regulating cell-cycle arrest, stabilizing stalled replication forks, and maintaining genome integrity [41]. ATR activation by single-stranded DNA (ssDNA) leads to Chk1 phosphorylation and coordination of repair and cell-cycle progression. Inhibition of ATR induces DNA damage, genomic instability, and cell death, particularly in cancer cells with existing repair defects [42], and it enhances sensitivity to chemotherapy and radiation. Herein, DDR-related parameters and oxidative stress were evaluated in human (A549, H1299) and murine (LLC) NSCLC cell lines following combined treatment with the ATR inhibitor AZD6738 and genotoxic agents (cisplatin, UVC irradiation). The combination of AZD6738, cisplatin, and anti–PD1 was also analyzed in a murine NSCLC model.

Across all cell lines, co-treatment of AZD6738 and genotoxic agents markedly impaired DNA repair efficiencies of both ICL repair and NER mechanisms, increased DNA damage, elevated oxidative stress and AP-site formation, and reduced cell viability. These findings align with previous evidence that ATR is essential for ICL repair, as ICL-induced replication fork stalling activates ATR and downstream Fanconi anemia (FA) proteins [43,44,45]. ATR inhibition disrupts checkpoint activation and replication fork stability, leading to unresolved DNA lesions, genomic instability, and cancer cell death—particularly in cells carrying defects in complementary repair pathways, such as ARID1A-mutant backgrounds [46,47]. Moreover, it is widely accepted that the ATR pathway and NER machinery cooperate to preserve genomic stability by coordinating DNA repair and cell-cycle checkpoint activation. NER sensors such as DDB2 and XPC can recruit ATR/ATM to UV-induced lesions, where ATR subsequently phosphorylates factors like XPA to stabilize and enhance NER activity, particularly in G1 phase [48]. ATR inhibition disrupts these early steps by impairing the recruitment and activation of checkpoint components, including TopBP1 and the RAD9-RAD1-HUS1 (9-1-1) complex [49,50], thereby reducing NER efficiency and compromising the overall repair response.

Notably, ATR inhibitors have been shown to promote oxidative stress, particularly in combination with genotoxic agents [49]. Elevated reactive oxygen species generate additional DNA damage that normally activates ATR signaling, and inhibiting this pathway enhances genomic instability and decreases cancer cell survival. As cancer cells depend heavily on ATR to manage replication stress, this synergy is being actively explored as a therapeutic strategy [51]. In addition, ATR inhibition can lead to abasic site accumulation, particularly in cancer cells overexpressing APOBEC3A (apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like 3A) and APOBEC3B, enzymes that convert cytosine to uracil in ssDNA [52]. Unrepaired ssDNA stalls replication forks, and uracil excision by uracil-DNA glycosylase (UNG2) generates abasic sites, whose accumulation can trigger replication catastrophe and cell death.

Interestingly, ATR inhibitors reduce cell viability in cancer cells by disrupting DDR pathways and inducing oxidative stress, effects that are enhanced when combined with genotoxic agents or in cells with specific genetic deficiencies [53,54]. Consistently, combination treatment with AZD6738 and cisplatin or UVC irradiation in NSCLC cell lines caused greater reductions in viability, increased DNA damage, decreased GSH/GSSG ratio, and elevated abasic sites compared with monotherapies. Remarkably, UVC plus AZD6738 significantly affected all DDR parameters only in p53-null H1299 cells, highlighting the influence of p53 status on AZD6738 efficacy. P53 functions as the “guardian of the genome”, halting the cell cycle to permit DNA repair or triggering apoptosis in cells with irreparable damage [55]. Loss or mutation of p53 increases reliance on ATR-mediated checkpoints, rendering cells more susceptible to ATR inhibitors and promoting cell death [56]. The differential magnitude of response observed among cell lines, particularly in p53-deficient H1299 cells, underscores the context-dependent nature of treatment efficacy. Given the well-established role of p53 in regulating DNA damage responses and redox homeostasis, these findings are biologically plausible. However, it should be noted that, given the limited number of cell lines analyzed in this study, this observation should be interpreted with caution and not extrapolated beyond the specific models examined.

Following our in vitro findings, we assessed the impact of ATR inhibition in heterotopic lung cancer mouse models. Initially, we evaluated the combination of AZD6738 with cisplatin, a widely used genotoxic chemotherapeutic agent in both clinical and preclinical settings [57,58,59]. As anticipated, cisplatin monotherapy modestly reduced tumor growth, while AZD6738 alone had no effect, consistent with previous reports [11,60]. Interestingly, although the combination of AZD6738 and cisplatin demonstrated synergy in vitro, this effect did not translate into improved outcomes in vivo in the syngeneic LLC mouse model compared with cisplatin alone. This discrepancy may be explained by several factors. Flow cytometry revealed increased total T cells (CD3^+^) and specifically CD8^+^ cytotoxic T cells in tumors following combination therapy, but T cells remained a small fraction of the immune infiltrate (~6% of CD45^+^ cells), whereas spleens showed 5–17% T cells, indicating peripheral activation. Within tumors, CD8^+^ cells were less abundant than CD4^+^ helper T cells. As previously noted, the antitumor effects of ceralasertib are largely mediated by CD8^+^ T cells [38]. Short-course AZD6738 combined with radiotherapy more effectively promotes effector CD8^+^ responses than prolonged treatment [8], and intermittent dosing can improve tolerability [39]. Collectively, these observations suggest that CD8^+^ T cell exhaustion likely limited tumor growth inhibition in our initial experiments. Additionally, complex tumor–immune interactions within the tumor microenvironment (TME) may dampen the therapeutic synergy observed in vitro. Furthermore, pharmacodynamic constraints, including inadequate drug exposure, suboptimal tissue distribution, or rapid metabolic clearance, may hinder the attainment of therapeutically effective intratumoral concentrations. The dosing schedule is also crucial, as suboptimal timing or sequencing of AZD6738 and cisplatin could impair combination efficacy. Moreover, the suppressive tumor microenvironment, characterized by regulatory T cells, myeloid-derived suppressor cells, and other inhibitory factors, may limit immune-mediated anti-tumor effects, while the overall immune exhaustion induced by the combined effects of tumor and treatment can further compromise systemic anti-tumor responses. Together, these factors highlight why in vitro DDR sensitization does not always translate into in vivo efficacy, underscoring the importance of considering pharmacology, immune context, and microenvironmental suppression in preclinical combination strategies.

One strategy to improve therapy outcome is leveraging the crosstalk between the immune system and the DDR. Using our murine lung cancer model, we evaluated the antitumor efficacy of combined AZD6738, cisplatin, and anti–PD1 therapy. We hypothesized that the synergy of platinum-based genotoxic treatment, ATR inhibition, and immune checkpoint blockade would enhance tumor immunogenicity and cytotoxicity and elicit a robust antitumor immune response. Mice were treated with various combinations of cisplatin, a brief AZD6738 regimen, and anti–PD1, administered in either one or two dosing cycles. The triple combination produced the greatest reduction in tumor growth and burden, followed by the same regimen with repeated anti–PD1 dosing, indicating synergistic effects. Notably, temporary tumor growth arrest was observed on day 11 in all anti–PD1-containing groups, but tumors resumed growth after treatment cessation or a second anti–PD1 cycle, suggesting that the initial immunogenic stimulus from DNA damage was not sustained. This may reflect reduced neoantigen exposure, diminished cGAS–STING activation, or suboptimal T cell activation leading to exhaustion. Future studies should evaluate exhaustion markers such as TIM-3 (T cell immunoglobulin and mucin domain-3) and LAG-3 (lymphocyte activation gene-3), to assess T cell functional status [61].

A key observation was that a second cycle of anti–PD1 reduced splenic T cells while increasing their tumor infiltration, indicating immune cell redistribution from peripheral lymphoid organs into the tumor microenvironment. These findings suggest that repeated anti–PD1 dosing facilitates CD3^+^ T cell migration and infiltration, potentially enhancing the overall antitumor immune response. The triple combination of AZD6738, cisplatin, and two anti–PD1 cycles produced the highest intratumoral T cell infiltration, suggesting a synergistic effect of DNA damage, ATR inhibition, and immune checkpoint blockade in enhancing T-cell-mediated antitumor responses. The pronounced CD3^+^ T cell infiltration underscores the potential of this regimen as an effective therapeutic strategy and highlights the critical role of treatment scheduling and dosing frequency in sustaining an immunostimulatory tumor microenvironment for durable tumor control.

Although total T cell infiltration increased, CD4^+^ helper and CD8^+^ cytotoxic subsets were reduced, suggesting complex or transient immune modulation within the tumor microenvironment. Intriguingly, the data indicated the emergence of an unconventional double-negative (DN) T cell subset (CD3^+^CD4^−^CD8^−^), comprising 30–80% of CD3^+^ T cells within the tumor microenvironment but only a minor fraction in the spleen, suggesting tumor-specific expansion. DN T cell enrichment was highest in mice receiving AZD6738 with dual anti–PD1 dosing, implying local expansion driven by therapy and a potential role in the antitumor immune response under specific therapeutic conditions. DN T cells remain poorly characterized, but emerging evidence across multiple tumor types indicates that they can exert either pro- or antitumor effects, depending on the tumor context [62]. Although they represent only a small fraction (1–2%) of peripheral blood in NSCLC patients [63], they have been shown to infiltrate solid tumors [64] and, in xenograft models, their adoptive transfer suppresses tumor growth and prolongs survival [64]. Thus, harnessing DN T cells for adoptive cell therapy (ACT), either alone or in combination with immune checkpoint blockade, has emerged as a promising strategy [64,65]. Notably, in nasopharyngeal carcinoma intratumoral DN T cells interact with CD4^+^ and CD8^+^ tumor-infiltrating lymphocytes (TILs) through TGF-β, IL-10, and Fas/FasL signaling, thereby restricting their expansion [66]. Further studies will be needed to elucidate the mechanisms underlying these differential responses.

5. Conclusions

In this preclinical study, we evaluated the effectiveness of the ATR inhibitor AZD6738 in NSCLC both in vitro and in vivo. We demonstrated that ATR blockade in human NSCLC cell lines potentiates the cytotoxicity of genotoxic agents by disrupting the DDR network, with this effect being most pronounced in cells that lack the tumor suppressor protein p53. Notably, ATR inhibition enhanced the immunogenic effects of genotoxic drugs in our NSCLC mouse model, with the triple combination of AZD6738, cisplatin, and anti-PD1 achieving the strongest antitumor effect. Although these findings are preclinical and require cautious interpretation—since toxicity and optimal dosing were not assessed—they highlight a potential therapeutic opportunity. Future studies leveraging biomarker-driven patient selection will be critical to translate these results into meaningful clinical benefit.